fftw-3.3.3/0000755000175000017500000000000012054411114007460 500000000000000fftw-3.3.3/m4/0000755000175000017500000000000012054411114010000 500000000000000fftw-3.3.3/m4/acx_mpi.m40000644000175000017500000000707212054410050011606 00000000000000dnl @synopsis ACX_MPI([ACTION-IF-FOUND[, ACTION-IF-NOT-FOUND]]) dnl @summary figure out how to compile/link code with MPI dnl @category InstalledPackages dnl dnl This macro tries to find out how to compile programs that dnl use MPI (Message Passing Interface), a standard API for dnl parallel process communication (see http://www-unix.mcs.anl.gov/mpi/) dnl dnl On success, it sets the MPICC, MPICXX, or MPIF77 output variable to dnl the name of the MPI compiler, depending upon the current language. dnl (This may just be $CC/$CXX/$F77, but is more often something like dnl mpicc/mpiCC/mpif77.) It also sets MPILIBS to any libraries that are dnl needed for linking MPI (e.g. -lmpi, if a special MPICC/MPICXX/MPIF77 dnl was not found). dnl dnl If you want to compile everything with MPI, you should set: dnl dnl CC="$MPICC" #OR# CXX="$MPICXX" #OR# F77="$MPIF77" dnl LIBS="$MPILIBS $LIBS" dnl dnl NOTE: The above assumes that you will use $CC (or whatever) dnl for linking as well as for compiling. (This is the dnl default for automake and most Makefiles.) dnl dnl The user can force a particular library/compiler by setting the dnl MPICC/MPICXX/MPIF77 and/or MPILIBS environment variables. dnl dnl ACTION-IF-FOUND is a list of shell commands to run if an MPI dnl library is found, and ACTION-IF-NOT-FOUND is a list of commands dnl to run it if it is not found. If ACTION-IF-FOUND is not specified, dnl the default action will define HAVE_MPI. dnl dnl @version 2005-09-02 dnl @license GPLWithACException dnl @author Steven G. Johnson AC_DEFUN([ACX_MPI], [ AC_PREREQ(2.50) dnl for AC_LANG_CASE AC_LANG_CASE([C], [ AC_REQUIRE([AC_PROG_CC]) AC_ARG_VAR(MPICC,[MPI C compiler command]) AC_CHECK_PROGS(MPICC, mpicc hcc mpcc mpcc_r mpxlc cmpicc, $CC) acx_mpi_save_CC="$CC" CC="$MPICC" AC_SUBST(MPICC) ], [C++], [ AC_REQUIRE([AC_PROG_CXX]) AC_ARG_VAR(MPICXX,[MPI C++ compiler command]) AC_CHECK_PROGS(MPICXX, mpic++ mpiCC mpicxx mpCC hcp mpxlC mpxlC_r cmpic++, $CXX) acx_mpi_save_CXX="$CXX" CXX="$MPICXX" AC_SUBST(MPICXX) ], [Fortran 77], [ AC_REQUIRE([AC_PROG_F77]) AC_ARG_VAR(MPIF77,[MPI Fortran compiler command]) AC_CHECK_PROGS(MPIF77, mpif77 hf77 mpxlf mpf77 mpif90 mpf90 mpxlf90 mpxlf95 mpxlf_r cmpifc cmpif90c, $F77) acx_mpi_save_F77="$F77" F77="$MPIF77" AC_SUBST(MPIF77) ]) if test x = x"$MPILIBS"; then AC_LANG_CASE([C], [AC_CHECK_FUNC(MPI_Init, [MPILIBS=" "])], [C++], [AC_CHECK_FUNC(MPI_Init, [MPILIBS=" "])], [Fortran 77], [AC_MSG_CHECKING([for MPI_Init]) AC_TRY_LINK([],[ call MPI_Init], [MPILIBS=" " AC_MSG_RESULT(yes)], [AC_MSG_RESULT(no)])]) fi if test x = x"$MPILIBS"; then AC_CHECK_LIB(mpi, MPI_Init, [MPILIBS="-lmpi"]) fi if test x = x"$MPILIBS"; then AC_CHECK_LIB(mpich, MPI_Init, [MPILIBS="-lmpich"]) fi dnl We have to use AC_TRY_COMPILE and not AC_CHECK_HEADER because the dnl latter uses $CPP, not $CC (which may be mpicc). AC_LANG_CASE([C], [if test x != x"$MPILIBS"; then AC_MSG_CHECKING([for mpi.h]) AC_TRY_COMPILE([#include ],[],[AC_MSG_RESULT(yes)], [MPILIBS="" AC_MSG_RESULT(no)]) fi], [C++], [if test x != x"$MPILIBS"; then AC_MSG_CHECKING([for mpi.h]) AC_TRY_COMPILE([#include ],[],[AC_MSG_RESULT(yes)], [MPILIBS="" AC_MSG_RESULT(no)]) fi]) AC_LANG_CASE([C], [CC="$acx_mpi_save_CC"], [C++], [CXX="$acx_mpi_save_CXX"], [Fortran 77], [F77="$acx_mpi_save_F77"]) AC_SUBST(MPILIBS) # Finally, execute ACTION-IF-FOUND/ACTION-IF-NOT-FOUND: if test x = x"$MPILIBS"; then $2 : else ifelse([$1],,[AC_DEFINE(HAVE_MPI,1,[Define if you have the MPI library.])],[$1]) : fi ])dnl ACX_MPI fftw-3.3.3/m4/acx_pthread.m40000644000175000017500000002301512054410050012443 00000000000000dnl @synopsis ACX_PTHREAD([ACTION-IF-FOUND[, ACTION-IF-NOT-FOUND]]) dnl @summary figure out how to build C programs using POSIX threads dnl @category InstalledPackages dnl dnl This macro figures out how to build C programs using POSIX dnl threads. It sets the PTHREAD_LIBS output variable to the threads dnl library and linker flags, and the PTHREAD_CFLAGS output variable dnl to any special C compiler flags that are needed. (The user can also dnl force certain compiler flags/libs to be tested by setting these dnl environment variables.) dnl dnl Also sets PTHREAD_CC to any special C compiler that is needed for dnl multi-threaded programs (defaults to the value of CC otherwise). dnl (This is necessary on AIX to use the special cc_r compiler alias.) dnl dnl NOTE: You are assumed to not only compile your program with these dnl flags, but also link it with them as well. e.g. you should link dnl with $PTHREAD_CC $CFLAGS $PTHREAD_CFLAGS $LDFLAGS ... $PTHREAD_LIBS $LIBS dnl dnl If you are only building threads programs, you may wish to dnl use these variables in your default LIBS, CFLAGS, and CC: dnl dnl LIBS="$PTHREAD_LIBS $LIBS" dnl CFLAGS="$CFLAGS $PTHREAD_CFLAGS" dnl CC="$PTHREAD_CC" dnl dnl In addition, if the PTHREAD_CREATE_JOINABLE thread-attribute dnl constant has a nonstandard name, defines PTHREAD_CREATE_JOINABLE dnl to that name (e.g. PTHREAD_CREATE_UNDETACHED on AIX). dnl dnl ACTION-IF-FOUND is a list of shell commands to run if a threads dnl library is found, and ACTION-IF-NOT-FOUND is a list of commands dnl to run it if it is not found. If ACTION-IF-FOUND is not specified, dnl the default action will define HAVE_PTHREAD. dnl dnl Please let the authors know if this macro fails on any platform, dnl or if you have any other suggestions or comments. This macro was dnl based on work by SGJ on autoconf scripts for FFTW (www.fftw.org) dnl (with help from M. Frigo), as well as ac_pthread and hb_pthread dnl macros posted by Alejandro Forero Cuervo to the autoconf macro dnl repository. We are also grateful for the helpful feedback of dnl numerous users. dnl dnl @version 2006-09-15 dnl @license GPLWithACException dnl @author Steven G. Johnson AC_DEFUN([ACX_PTHREAD], [ AC_REQUIRE([AC_CANONICAL_HOST]) AC_LANG_SAVE AC_LANG_C acx_pthread_ok=no # We used to check for pthread.h first, but this fails if pthread.h # requires special compiler flags (e.g. on True64 or Sequent). # It gets checked for in the link test anyway. # First of all, check if the user has set any of the PTHREAD_LIBS, # etcetera environment variables, and if threads linking works using # them: if test x"$PTHREAD_LIBS$PTHREAD_CFLAGS" != x; then save_CFLAGS="$CFLAGS" CFLAGS="$CFLAGS $PTHREAD_CFLAGS" save_LIBS="$LIBS" LIBS="$PTHREAD_LIBS $LIBS" AC_MSG_CHECKING([for pthread_join in LIBS=$PTHREAD_LIBS with CFLAGS=$PTHREAD_CFLAGS]) AC_TRY_LINK_FUNC(pthread_join, acx_pthread_ok=yes) AC_MSG_RESULT($acx_pthread_ok) if test x"$acx_pthread_ok" = xno; then PTHREAD_LIBS="" PTHREAD_CFLAGS="" fi LIBS="$save_LIBS" CFLAGS="$save_CFLAGS" fi # We must check for the threads library under a number of different # names; the ordering is very important because some systems # (e.g. DEC) have both -lpthread and -lpthreads, where one of the # libraries is broken (non-POSIX). # Create a list of thread flags to try. Items starting with a "-" are # C compiler flags, and other items are library names, except for "none" # which indicates that we try without any flags at all, and "pthread-config" # which is a program returning the flags for the Pth emulation library. acx_pthread_flags="pthreads none -Kthread -kthread lthread -pthread -pthreads -mt -mthreads pthread --thread-safe pthread-config" # The ordering *is* (sometimes) important. Some notes on the # individual items follow: # pthreads: AIX (must check this before -lpthread) # none: in case threads are in libc; should be tried before -Kthread and # other compiler flags to prevent continual compiler warnings # -Kthread: Sequent (threads in libc, but -Kthread needed for pthread.h) # -kthread: FreeBSD kernel threads (preferred to -pthread since SMP-able) # lthread: LinuxThreads port on FreeBSD (also preferred to -pthread) # -pthread: Linux/gcc (kernel threads), BSD/gcc (userland threads) # -pthreads: Solaris/gcc # -mthreads: Mingw32/gcc, Lynx/gcc # -mt: Sun Workshop C (may only link SunOS threads [-lthread], but it # doesn't hurt to check since this sometimes defines pthreads too; # also defines -D_REENTRANT) # ... -mt is also the pthreads flag for HP/aCC # (where it should come before -mthreads to avoid spurious warnings) # pthread: Linux, etcetera # --thread-safe: KAI C++ # pthread-config: use pthread-config program (for GNU Pth library) case "${host_cpu}-${host_os}" in *solaris*) # On Solaris (at least, for some versions), libc contains stubbed # (non-functional) versions of the pthreads routines, so link-based # tests will erroneously succeed. (We need to link with -pthreads/-mt/ # -lpthread.) (The stubs are missing pthread_cleanup_push, or rather # a function called by this macro, so we could check for that, but # who knows whether they'll stub that too in a future libc.) So, # we'll just look for -pthreads and -lpthread first: acx_pthread_flags="-pthreads pthread -mt -pthread $acx_pthread_flags" ;; esac if test x"$acx_pthread_ok" = xno; then for flag in $acx_pthread_flags; do case $flag in none) AC_MSG_CHECKING([whether pthreads work without any flags]) ;; -*) AC_MSG_CHECKING([whether pthreads work with $flag]) PTHREAD_CFLAGS="$flag" ;; pthread-config) AC_CHECK_PROG(acx_pthread_config, pthread-config, yes, no) if test x"$acx_pthread_config" = xno; then continue; fi PTHREAD_CFLAGS="`pthread-config --cflags`" PTHREAD_LIBS="`pthread-config --ldflags` `pthread-config --libs`" ;; *) AC_MSG_CHECKING([for the pthreads library -l$flag]) PTHREAD_LIBS="-l$flag" ;; esac save_LIBS="$LIBS" save_CFLAGS="$CFLAGS" LIBS="$PTHREAD_LIBS $LIBS" CFLAGS="$CFLAGS $PTHREAD_CFLAGS" # Check for various functions. We must include pthread.h, # since some functions may be macros. (On the Sequent, we # need a special flag -Kthread to make this header compile.) # We check for pthread_join because it is in -lpthread on IRIX # while pthread_create is in libc. We check for pthread_attr_init # due to DEC craziness with -lpthreads. We check for # pthread_cleanup_push because it is one of the few pthread # functions on Solaris that doesn't have a non-functional libc stub. # We try pthread_create on general principles. AC_TRY_LINK([#include ], [pthread_t th; pthread_join(th, (void**) 0); pthread_attr_init((pthread_attr_t*) 0); pthread_cleanup_push((void(*)(void *)) 0, (void*) 0); pthread_create((pthread_t*) 0, (pthread_attr_t*) 0, (void*(*)(void *)) 0, (void*) 0); pthread_cleanup_pop(0); ], [acx_pthread_ok=yes]) LIBS="$save_LIBS" CFLAGS="$save_CFLAGS" AC_MSG_RESULT($acx_pthread_ok) if test "x$acx_pthread_ok" = xyes; then break; fi PTHREAD_LIBS="" PTHREAD_CFLAGS="" done fi # Various other checks: if test "x$acx_pthread_ok" = xyes; then save_LIBS="$LIBS" LIBS="$PTHREAD_LIBS $LIBS" save_CFLAGS="$CFLAGS" CFLAGS="$CFLAGS $PTHREAD_CFLAGS" # Detect AIX lossage: JOINABLE attribute is called UNDETACHED. AC_MSG_CHECKING([for joinable pthread attribute]) attr_name=unknown for attr in PTHREAD_CREATE_JOINABLE PTHREAD_CREATE_UNDETACHED; do AC_TRY_LINK([#include ], [int attr=$attr; return attr;], [attr_name=$attr; break]) done AC_MSG_RESULT($attr_name) if test "$attr_name" != PTHREAD_CREATE_JOINABLE; then AC_DEFINE_UNQUOTED(PTHREAD_CREATE_JOINABLE, $attr_name, [Define to necessary symbol if this constant uses a non-standard name on your system.]) fi AC_MSG_CHECKING([if more special flags are required for pthreads]) flag=no case "${host_cpu}-${host_os}" in *-aix* | *-freebsd* | *-darwin*) flag="-D_THREAD_SAFE";; *solaris* | *-osf* | *-hpux*) flag="-D_REENTRANT";; esac AC_MSG_RESULT(${flag}) if test "x$flag" != xno; then PTHREAD_CFLAGS="$flag $PTHREAD_CFLAGS" fi LIBS="$save_LIBS" CFLAGS="$save_CFLAGS" # More AIX lossage: must compile with xlc_r or cc_r if test x"$GCC" != xyes; then AC_CHECK_PROGS(PTHREAD_CC, xlc_r cc_r, ${CC}) else PTHREAD_CC=$CC fi else PTHREAD_CC="$CC" fi AC_SUBST(PTHREAD_LIBS) AC_SUBST(PTHREAD_CFLAGS) AC_SUBST(PTHREAD_CC) # Finally, execute ACTION-IF-FOUND/ACTION-IF-NOT-FOUND: if test x"$acx_pthread_ok" = xyes; then ifelse([$1],,AC_DEFINE(HAVE_PTHREAD,1,[Define if you have POSIX threads libraries and header files.]),[$1]) : else acx_pthread_ok=no $2 fi AC_LANG_RESTORE ])dnl ACX_PTHREAD fftw-3.3.3/m4/ax_cc_maxopt.m40000644000175000017500000001167012054410050012632 00000000000000dnl @synopsis AX_CC_MAXOPT dnl @summary turn on optimization flags for the C compiler dnl @category C dnl dnl Try to turn on "good" C optimization flags for various compilers dnl and architectures, for some definition of "good". (In our case, dnl good for FFTW and hopefully for other scientific codes. Modify dnl as needed.) dnl dnl The user can override the flags by setting the CFLAGS environment dnl variable. dnl dnl Note also that the flags assume that ANSI C aliasing rules are dnl followed by the code (e.g. for gcc's -fstrict-aliasing), and that dnl floating-point computations can be re-ordered as needed. dnl dnl Requires macros: AX_CHECK_COMPILER_FLAGS, AX_COMPILER_VENDOR, dnl dnl @version 2011-06-22 dnl @license GPLWithACException dnl @author Steven G. Johnson and Matteo Frigo. AC_DEFUN([AX_CC_MAXOPT], [ AC_REQUIRE([AC_PROG_CC]) AC_REQUIRE([AX_COMPILER_VENDOR]) AC_REQUIRE([AC_CANONICAL_HOST]) # Try to determine "good" native compiler flags if none specified via CFLAGS if test "$ac_test_CFLAGS" != "set"; then CFLAGS="" case $ax_cv_c_compiler_vendor in dec) CFLAGS="-newc -w0 -O5 -ansi_alias -ansi_args -fp_reorder -tune host" ;; sun) CFLAGS="-native -fast -xO5 -dalign" ;; hp) CFLAGS="+Oall +Optrs_ansi +DSnative" ;; ibm) xlc_opt="-qtune=auto" AX_CHECK_COMPILER_FLAGS($xlc_opt, CFLAGS="-O3 -qansialias -w $xlc_opt", [CFLAGS="-O3 -qansialias -w" echo "******************************************************" echo "* You seem to have the IBM C compiler. It is *" echo "* recommended for best performance that you use: *" echo "* *" echo "* CFLAGS=-O3 -qarch=xxx -qtune=xxx -qansialias -w *" echo "* ^^^ ^^^ *" echo "* where xxx is pwr2, pwr3, 604, or whatever kind of *" echo "* CPU you have. (Set the CFLAGS environment var. *" echo "* and re-run configure.) For more info, man cc. *" echo "******************************************************"]) ;; intel) CFLAGS="-O3" # Intel seems to have changed the spelling of this flag recently icc_ansi_alias="unknown" for flag in -ansi-alias -ansi_alias; do AX_CHECK_COMPILER_FLAGS($flag, [icc_ansi_alias=$flag; break]) done if test "x$icc_ansi_alias" != xunknown; then CFLAGS="$CFLAGS $icc_ansi_alias" fi AX_CHECK_COMPILER_FLAGS(-malign-double, CFLAGS="$CFLAGS -malign-double") # We used to check for architecture flags here, e.g. -xHost etc., # but these flags are problematic. On icc-12.0.0, "-mavx -xHost" # overrides -mavx with -xHost, generating SSE2 code instead of AVX # code. ICC does not seem to support -mtune=host or equivalent # non-ABI changing flag. ;; gnu) # Default optimization flags for gcc on all systems. # Somehow -O3 does not imply -fomit-frame-pointer on ia32 CFLAGS="-O3 -fomit-frame-pointer" # tune for the host by default AX_CHECK_COMPILER_FLAGS(-mtune=native, CFLAGS="$CFLAGS -mtune=native") # -malign-double for x86 systems AX_CHECK_COMPILER_FLAGS(-malign-double, CFLAGS="$CFLAGS -malign-double") # -fstrict-aliasing for gcc-2.95+ AX_CHECK_COMPILER_FLAGS(-fstrict-aliasing, CFLAGS="$CFLAGS -fstrict-aliasing") # -fno-schedule-insns is pretty much required on all risc # processors. # # gcc performs one pass of instruction scheduling, then a pass of # register allocation, then another pass of instruction # scheduling. The first pass reorders instructions in a way that # is pretty much the worst possible for the purposes of register # allocation. We disable the first pass. AX_CHECK_COMPILER_FLAGS(-fno-schedule-insns, CFLAGS="$CFLAGS -fno-schedule-insns") # note that we enable "unsafe" fp optimization with other compilers, too AX_CHECK_COMPILER_FLAGS(-ffast-math, CFLAGS="$CFLAGS -ffast-math") ;; esac if test -z "$CFLAGS"; then echo "" echo "********************************************************" echo "* WARNING: Don't know the best CFLAGS for this system *" echo "* Use ./configure CFLAGS=... to specify your own flags *" echo "* (otherwise, a default of CFLAGS=-O3 will be used) *" echo "********************************************************" echo "" CFLAGS="-O3" fi AX_CHECK_COMPILER_FLAGS($CFLAGS, [], [ echo "" echo "********************************************************" echo "* WARNING: The guessed CFLAGS don't seem to work with *" echo "* your compiler. *" echo "* Use ./configure CFLAGS=... to specify your own flags *" echo "********************************************************" echo "" CFLAGS="" ]) fi ]) fftw-3.3.3/m4/ax_check_compiler_flags.m40000644000175000017500000000316712054410050015002 00000000000000dnl @synopsis AX_CHECK_COMPILER_FLAGS(FLAGS, [ACTION-SUCCESS], [ACTION-FAILURE]) dnl @summary check whether FLAGS are accepted by the compiler dnl @category Misc dnl dnl Check whether the given compiler FLAGS work with the current language's dnl compiler, or whether they give an error. (Warnings, however, are dnl ignored.) dnl dnl ACTION-SUCCESS/ACTION-FAILURE are shell commands to execute on dnl success/failure. dnl dnl @version 2005-05-30 dnl @license GPLWithACException dnl @author Steven G. Johnson and Matteo Frigo. AC_DEFUN([AX_CHECK_COMPILER_FLAGS], [AC_PREREQ(2.59) dnl for _AC_LANG_PREFIX AC_MSG_CHECKING([whether _AC_LANG compiler accepts $1]) dnl Some hackery here since AC_CACHE_VAL can't handle a non-literal varname: AS_LITERAL_IF([$1], [AC_CACHE_VAL(AS_TR_SH(ax_cv_[]_AC_LANG_ABBREV[]_flags_$1), [ ax_save_FLAGS=$[]_AC_LANG_PREFIX[]FLAGS _AC_LANG_PREFIX[]FLAGS="$1" AC_COMPILE_IFELSE([AC_LANG_PROGRAM()], AS_TR_SH(ax_cv_[]_AC_LANG_ABBREV[]_flags_$1)=yes, AS_TR_SH(ax_cv_[]_AC_LANG_ABBREV[]_flags_$1)=no) _AC_LANG_PREFIX[]FLAGS=$ax_save_FLAGS])], [ax_save_FLAGS=$[]_AC_LANG_PREFIX[]FLAGS _AC_LANG_PREFIX[]FLAGS="$1" AC_COMPILE_IFELSE([AC_LANG_PROGRAM()], eval AS_TR_SH(ax_cv_[]_AC_LANG_ABBREV[]_flags_$1)=yes, eval AS_TR_SH(ax_cv_[]_AC_LANG_ABBREV[]_flags_$1)=no) _AC_LANG_PREFIX[]FLAGS=$ax_save_FLAGS]) eval ax_check_compiler_flags=$AS_TR_SH(ax_cv_[]_AC_LANG_ABBREV[]_flags_$1) AC_MSG_RESULT($ax_check_compiler_flags) if test "x$ax_check_compiler_flags" = xyes; then m4_default([$2], :) else m4_default([$3], :) fi ])dnl AX_CHECK_COMPILER_FLAGS fftw-3.3.3/m4/ax_compiler_vendor.m40000644000175000017500000000272212054410050014042 00000000000000dnl @synopsis AX_COMPILER_VENDOR dnl @summary find the vendor (gnu, intel, etc.) of the C/C++ compiler dnl @category C dnl @category C++ dnl dnl Determine the vendor of the C/C++ compiler, e.g., gnu, intel, ibm, dnl sun, hp, borland, comeau, dec, cray, kai, lcc, metrowerks, sgi, dnl microsoft, watcom, etc. The vendor is returned in the cache variable dnl $ax_cv_c_compiler_vendor for C and $ax_cv_cxx_compiler_vendor for C++. dnl dnl @version 2007-08-01 dnl @license GPLWithACException dnl @author Steven G. Johnson with Matteo Frigo AC_DEFUN([AX_COMPILER_VENDOR], [ AC_CACHE_CHECK([for _AC_LANG compiler vendor], ax_cv_[]_AC_LANG_ABBREV[]_compiler_vendor, [ax_cv_[]_AC_LANG_ABBREV[]_compiler_vendor=unknown # note: don't check for gcc first since some other compilers define __GNUC__ for ventest in intel:__ICC,__ECC,__INTEL_COMPILER ibm:__xlc__,__xlC__,__IBMC__,__IBMCPP__ pathscale:__PATHCC__,__PATHSCALE__ gnu:__GNUC__ sun:__SUNPRO_C,__SUNPRO_CC hp:__HP_cc,__HP_aCC dec:__DECC,__DECCXX,__DECC_VER,__DECCXX_VER borland:__BORLANDC__,__TURBOC__ comeau:__COMO__ cray:_CRAYC kai:__KCC lcc:__LCC__ metrowerks:__MWERKS__ sgi:__sgi,sgi microsoft:_MSC_VER watcom:__WATCOMC__ portland:__PGI; do vencpp="defined("`echo $ventest | cut -d: -f2 | sed 's/,/) || defined(/g'`")" AC_COMPILE_IFELSE([AC_LANG_PROGRAM(,[ #if !($vencpp) thisisanerror; #endif ])], [ax_cv_]_AC_LANG_ABBREV[_compiler_vendor=`echo $ventest | cut -d: -f1`; break]) done ]) ]) fftw-3.3.3/m4/ax_gcc_aligns_stack.m40000644000175000017500000000407212054410050014131 00000000000000dnl @synopsis AX_GCC_ALIGNS_STACK([ACTION-IF-YES], [ACTION-IF-NO]) dnl @summary check whether gcc can align stack to 8-byte boundary dnl @category Misc dnl dnl Check to see if we are using a version of gcc that aligns the stack dnl (true in gcc-2.95+, which have the -mpreferred-stack-boundary flag). dnl Also, however, checks whether main() is correctly aligned by the dnl OS/libc/..., as well as for a bug in the stack alignment of gcc-2.95.x dnl (see http://gcc.gnu.org/ml/gcc-bugs/1999-11/msg00259.html). dnl dnl ACTION-IF-YES/ACTION-IF-NO are shell commands to execute if we are dnl using gcc and the stack is/isn't aligned, respectively. dnl dnl Requires macro: AX_CHECK_COMPILER_FLAGS, AX_GCC_VERSION dnl dnl @version 2005-05-30 dnl @license GPLWithACException dnl @author Steven G. Johnson AC_DEFUN([AX_GCC_ALIGNS_STACK], [ AC_REQUIRE([AC_PROG_CC]) ax_gcc_aligns_stack=no if test "$GCC" = "yes"; then AX_CHECK_COMPILER_FLAGS(-mpreferred-stack-boundary=4, [ AC_MSG_CHECKING([whether the stack is at least 8-byte aligned by gcc]) save_CFLAGS="$CFLAGS" CFLAGS="-O" AX_CHECK_COMPILER_FLAGS(-malign-double, CFLAGS="$CFLAGS -malign-double") AC_TRY_RUN([#include # include struct yuck { int blechh; }; int one(void) { return 1; } struct yuck ick(void) { struct yuck y; y.blechh = 3; return y; } # define CHK_ALIGN(x) if ((((long) &(x)) & 0x7)) { fprintf(stderr, "bad alignment of " #x "\n"); exit(1); } void blah(int foo) { double foobar; CHK_ALIGN(foobar); } int main2(void) {double ok1; struct yuck y; double ok2; CHK_ALIGN(ok1); CHK_ALIGN(ok2); y = ick(); blah(one()); return 0;} int main(void) { if ((((long) (__builtin_alloca(0))) & 0x7)) __builtin_alloca(4); return main2(); } ], [ax_gcc_aligns_stack=yes; ax_gcc_stack_align_bug=no], ax_gcc_stack_align_bug=yes, [AX_GCC_VERSION(3,0,0, ax_gcc_stack_align_bug=no, ax_gcc_stack_align_bug=yes)]) CFLAGS="$save_CFLAGS" AC_MSG_RESULT($ax_gcc_aligns_stack) ]) fi if test "$ax_gcc_aligns_stack" = yes; then m4_default([$1], :) else m4_default([$2], :) fi ]) fftw-3.3.3/m4/ax_gcc_version.m40000644000175000017500000000215012054410050013147 00000000000000dnl @synopsis AX_GCC_VERSION(MAJOR, MINOR, PATCHLEVEL, [ACTION-SUCCESS], [ACTION-FAILURE]) dnl @summary check wither gcc is at least version MAJOR.MINOR.PATCHLEVEL dnl @category InstalledPackages dnl dnl Check whether we are using gcc and, if so, whether its version dnl is at least MAJOR.MINOR.PATCHLEVEL dnl dnl ACTION-SUCCESS/ACTION-FAILURE are shell commands to execute on dnl success/failure. dnl dnl @version 2005-05-30 dnl @license GPLWithACException dnl @author Steven G. Johnson and Matteo Frigo. AC_DEFUN([AX_GCC_VERSION], [ AC_REQUIRE([AC_PROG_CC]) AC_CACHE_CHECK(whether we are using gcc $1.$2.$3 or later, ax_cv_gcc_$1_$2_$3, [ ax_cv_gcc_$1_$2_$3=no if test "$GCC" = "yes"; then dnl The semicolon after "yes" below is to pacify NeXT's syntax-checking cpp. AC_EGREP_CPP(yes, [ #ifdef __GNUC__ # if (__GNUC__ > $1) || (__GNUC__ == $1 && __GNUC_MINOR__ > $2) \ || (__GNUC__ == $1 && __GNUC_MINOR__ == $2 && __GNUC_PATCHLEVEL__ >= $3) yes; # endif #endif ], [ax_cv_gcc_$1_$2_$3=yes]) fi ]) if test "$ax_cv_gcc_$1_$2_$3" = yes; then m4_default([$4], :) else m4_default([$5], :) fi ]) fftw-3.3.3/m4/ax_openmp.m40000644000175000017500000000530212054410050012146 00000000000000dnl @synopsis AX_OPENMP([ACTION-IF-FOUND[, ACTION-IF-NOT-FOUND]]) dnl @summary determine how to compile programs using OpenMP dnl @category InstalledPackages dnl dnl This macro tries to find out how to compile programs that dnl use OpenMP, a standard API and set of compiler directives for dnl parallel programming (see http://www.openmp.org/). dnl dnl On success, it sets the OPENMP_CFLAGS/OPENMP_CXXFLAGS/OPENMP_FFLAGS dnl output variable to the flag (e.g. -omp) used both to compile *and* link dnl OpenMP programs in the current language. dnl dnl NOTE: You are assumed to not only compile your program with these dnl flags, but also link it with them as well. dnl dnl If you want to compile everything with OpenMP, you should set: dnl dnl CFLAGS="$CFLAGS $OPENMP_CFLAGS" dnl #OR# CXXFLAGS="$CXXFLAGS $OPENMP_CXXFLAGS" dnl #OR# FFLAGS="$FFLAGS $OPENMP_FFLAGS" dnl dnl (depending on the selected language). dnl dnl The user can override the default choice by setting the corresponding dnl environment variable (e.g. OPENMP_CFLAGS). dnl dnl ACTION-IF-FOUND is a list of shell commands to run if an OpenMP dnl flag is found, and ACTION-IF-NOT-FOUND is a list of commands dnl to run it if it is not found. If ACTION-IF-FOUND is not specified, dnl the default action will define HAVE_OPENMP. dnl dnl @version 2006-11-20 dnl @license GPLWithACException dnl @author Steven G. Johnson AC_DEFUN([AX_OPENMP], [ AC_PREREQ(2.59) dnl for _AC_LANG_PREFIX AC_CACHE_CHECK([for OpenMP flag of _AC_LANG compiler], ax_cv_[]_AC_LANG_ABBREV[]_openmp, [save[]_AC_LANG_PREFIX[]FLAGS=$[]_AC_LANG_PREFIX[]FLAGS ax_cv_[]_AC_LANG_ABBREV[]_openmp=unknown # Flags to try: -fopenmp (gcc), -openmp (icc), -mp (SGI & PGI), # -xopenmp (Sun), -omp (Tru64), -qsmp=omp (AIX), none ax_openmp_flags="-fopenmp -openmp -mp -xopenmp -omp -qsmp=omp none" if test "x$OPENMP_[]_AC_LANG_PREFIX[]FLAGS" != x; then ax_openmp_flags="$OPENMP_[]_AC_LANG_PREFIX[]FLAGS $ax_openmp_flags" fi for ax_openmp_flag in $ax_openmp_flags; do case $ax_openmp_flag in none) []_AC_LANG_PREFIX[]FLAGS=$save[]_AC_LANG_PREFIX[] ;; *) []_AC_LANG_PREFIX[]FLAGS="$save[]_AC_LANG_PREFIX[]FLAGS $ax_openmp_flag" ;; esac AC_TRY_LINK_FUNC(omp_set_num_threads, [ax_cv_[]_AC_LANG_ABBREV[]_openmp=$ax_openmp_flag; break]) done []_AC_LANG_PREFIX[]FLAGS=$save[]_AC_LANG_PREFIX[]FLAGS ]) if test "x$ax_cv_[]_AC_LANG_ABBREV[]_openmp" = "xunknown"; then m4_default([$2],:) else if test "x$ax_cv_[]_AC_LANG_ABBREV[]_openmp" != "xnone"; then OPENMP_[]_AC_LANG_PREFIX[]FLAGS=$ax_cv_[]_AC_LANG_ABBREV[]_openmp fi m4_default([$1], [AC_DEFINE(HAVE_OPENMP,1,[Define if OpenMP is enabled])]) fi AC_SUBST(OPENMP_[]_AC_LANG_PREFIX[]FLAGS) ])dnl AX_OPENMP fftw-3.3.3/m4/libtool.m40000644000175000017500000105754211765467473011677 00000000000000# libtool.m4 - Configure libtool for the host system. -*-Autoconf-*- # # Copyright (C) 1996, 1997, 1998, 1999, 2000, 2001, 2003, 2004, 2005, # 2006, 2007, 2008, 2009, 2010, 2011 Free Software # Foundation, Inc. # Written by Gordon Matzigkeit, 1996 # # This file is free software; the Free Software Foundation gives # unlimited permission to copy and/or distribute it, with or without # modifications, as long as this notice is preserved. m4_define([_LT_COPYING], [dnl # Copyright (C) 1996, 1997, 1998, 1999, 2000, 2001, 2003, 2004, 2005, # 2006, 2007, 2008, 2009, 2010, 2011 Free Software # Foundation, Inc. # Written by Gordon Matzigkeit, 1996 # # This file is part of GNU Libtool. # # GNU Libtool is free software; you can redistribute it and/or # modify it under the terms of the GNU General Public License as # published by the Free Software Foundation; either version 2 of # the License, or (at your option) any later version. # # As a special exception to the GNU General Public License, # if you distribute this file as part of a program or library that # is built using GNU Libtool, you may include this file under the # same distribution terms that you use for the rest of that program. # # GNU Libtool is distributed in the hope that it will be useful, # but WITHOUT ANY WARRANTY; without even the implied warranty of # MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the # GNU General Public License for more details. # # You should have received a copy of the GNU General Public License # along with GNU Libtool; see the file COPYING. If not, a copy # can be downloaded from http://www.gnu.org/licenses/gpl.html, or # obtained by writing to the Free Software Foundation, Inc., # 51 Franklin Street, Fifth Floor, Boston, MA 02110-1301, USA. ]) # serial 57 LT_INIT # LT_PREREQ(VERSION) # ------------------ # Complain and exit if this libtool version is less that VERSION. m4_defun([LT_PREREQ], [m4_if(m4_version_compare(m4_defn([LT_PACKAGE_VERSION]), [$1]), -1, [m4_default([$3], [m4_fatal([Libtool version $1 or higher is required], 63)])], [$2])]) # _LT_CHECK_BUILDDIR # ------------------ # Complain if the absolute build directory name contains unusual characters m4_defun([_LT_CHECK_BUILDDIR], [case `pwd` in *\ * | *\ *) AC_MSG_WARN([Libtool does not cope well with whitespace in `pwd`]) ;; esac ]) # LT_INIT([OPTIONS]) # ------------------ AC_DEFUN([LT_INIT], [AC_PREREQ([2.58])dnl We use AC_INCLUDES_DEFAULT AC_REQUIRE([AC_CONFIG_AUX_DIR_DEFAULT])dnl AC_BEFORE([$0], [LT_LANG])dnl AC_BEFORE([$0], [LT_OUTPUT])dnl AC_BEFORE([$0], [LTDL_INIT])dnl m4_require([_LT_CHECK_BUILDDIR])dnl dnl Autoconf doesn't catch unexpanded LT_ macros by default: m4_pattern_forbid([^_?LT_[A-Z_]+$])dnl m4_pattern_allow([^(_LT_EOF|LT_DLGLOBAL|LT_DLLAZY_OR_NOW|LT_MULTI_MODULE)$])dnl dnl aclocal doesn't pull ltoptions.m4, ltsugar.m4, or ltversion.m4 dnl unless we require an AC_DEFUNed macro: AC_REQUIRE([LTOPTIONS_VERSION])dnl AC_REQUIRE([LTSUGAR_VERSION])dnl AC_REQUIRE([LTVERSION_VERSION])dnl AC_REQUIRE([LTOBSOLETE_VERSION])dnl m4_require([_LT_PROG_LTMAIN])dnl _LT_SHELL_INIT([SHELL=${CONFIG_SHELL-/bin/sh}]) dnl Parse OPTIONS _LT_SET_OPTIONS([$0], [$1]) # This can be used to rebuild libtool when needed LIBTOOL_DEPS="$ltmain" # Always use our own libtool. LIBTOOL='$(SHELL) $(top_builddir)/libtool' AC_SUBST(LIBTOOL)dnl _LT_SETUP # Only expand once: m4_define([LT_INIT]) ])# LT_INIT # Old names: AU_ALIAS([AC_PROG_LIBTOOL], [LT_INIT]) AU_ALIAS([AM_PROG_LIBTOOL], [LT_INIT]) dnl aclocal-1.4 backwards compatibility: dnl AC_DEFUN([AC_PROG_LIBTOOL], []) dnl AC_DEFUN([AM_PROG_LIBTOOL], []) # _LT_CC_BASENAME(CC) # ------------------- # Calculate cc_basename. Skip known compiler wrappers and cross-prefix. m4_defun([_LT_CC_BASENAME], [for cc_temp in $1""; do case $cc_temp in compile | *[[\\/]]compile | ccache | *[[\\/]]ccache ) ;; distcc | *[[\\/]]distcc | purify | *[[\\/]]purify ) ;; \-*) ;; *) break;; esac done cc_basename=`$ECHO "$cc_temp" | $SED "s%.*/%%; s%^$host_alias-%%"` ]) # _LT_FILEUTILS_DEFAULTS # ---------------------- # It is okay to use these file commands and assume they have been set # sensibly after `m4_require([_LT_FILEUTILS_DEFAULTS])'. m4_defun([_LT_FILEUTILS_DEFAULTS], [: ${CP="cp -f"} : ${MV="mv -f"} : ${RM="rm -f"} ])# _LT_FILEUTILS_DEFAULTS # _LT_SETUP # --------- m4_defun([_LT_SETUP], [AC_REQUIRE([AC_CANONICAL_HOST])dnl AC_REQUIRE([AC_CANONICAL_BUILD])dnl AC_REQUIRE([_LT_PREPARE_SED_QUOTE_VARS])dnl AC_REQUIRE([_LT_PROG_ECHO_BACKSLASH])dnl _LT_DECL([], [PATH_SEPARATOR], [1], [The PATH separator for the build system])dnl dnl _LT_DECL([], [host_alias], [0], [The host system])dnl _LT_DECL([], [host], [0])dnl _LT_DECL([], [host_os], [0])dnl dnl _LT_DECL([], [build_alias], [0], [The build system])dnl _LT_DECL([], [build], [0])dnl _LT_DECL([], [build_os], [0])dnl dnl AC_REQUIRE([AC_PROG_CC])dnl AC_REQUIRE([LT_PATH_LD])dnl AC_REQUIRE([LT_PATH_NM])dnl dnl AC_REQUIRE([AC_PROG_LN_S])dnl test -z "$LN_S" && LN_S="ln -s" _LT_DECL([], [LN_S], [1], [Whether we need soft or hard links])dnl dnl AC_REQUIRE([LT_CMD_MAX_LEN])dnl _LT_DECL([objext], [ac_objext], [0], [Object file suffix (normally "o")])dnl _LT_DECL([], [exeext], [0], [Executable file suffix (normally "")])dnl dnl m4_require([_LT_FILEUTILS_DEFAULTS])dnl m4_require([_LT_CHECK_SHELL_FEATURES])dnl m4_require([_LT_PATH_CONVERSION_FUNCTIONS])dnl m4_require([_LT_CMD_RELOAD])dnl m4_require([_LT_CHECK_MAGIC_METHOD])dnl m4_require([_LT_CHECK_SHAREDLIB_FROM_LINKLIB])dnl m4_require([_LT_CMD_OLD_ARCHIVE])dnl m4_require([_LT_CMD_GLOBAL_SYMBOLS])dnl m4_require([_LT_WITH_SYSROOT])dnl _LT_CONFIG_LIBTOOL_INIT([ # See if we are running on zsh, and set the options which allow our # commands through without removal of \ escapes INIT. if test -n "\${ZSH_VERSION+set}" ; then setopt NO_GLOB_SUBST fi ]) if test -n "${ZSH_VERSION+set}" ; then setopt NO_GLOB_SUBST fi _LT_CHECK_OBJDIR m4_require([_LT_TAG_COMPILER])dnl case $host_os in aix3*) # AIX sometimes has problems with the GCC collect2 program. For some # reason, if we set the COLLECT_NAMES environment variable, the problems # vanish in a puff of smoke. if test "X${COLLECT_NAMES+set}" != Xset; then COLLECT_NAMES= export COLLECT_NAMES fi ;; esac # Global variables: ofile=libtool can_build_shared=yes # All known linkers require a `.a' archive for static linking (except MSVC, # which needs '.lib'). libext=a with_gnu_ld="$lt_cv_prog_gnu_ld" old_CC="$CC" old_CFLAGS="$CFLAGS" # Set sane defaults for various variables test -z "$CC" && CC=cc test -z "$LTCC" && LTCC=$CC test -z "$LTCFLAGS" && LTCFLAGS=$CFLAGS test -z "$LD" && LD=ld test -z "$ac_objext" && ac_objext=o _LT_CC_BASENAME([$compiler]) # Only perform the check for file, if the check method requires it test -z "$MAGIC_CMD" && MAGIC_CMD=file case $deplibs_check_method in file_magic*) if test "$file_magic_cmd" = '$MAGIC_CMD'; then _LT_PATH_MAGIC fi ;; esac # Use C for the default configuration in the libtool script LT_SUPPORTED_TAG([CC]) _LT_LANG_C_CONFIG _LT_LANG_DEFAULT_CONFIG _LT_CONFIG_COMMANDS ])# _LT_SETUP # _LT_PREPARE_SED_QUOTE_VARS # -------------------------- # Define a few sed substitution that help us do robust quoting. m4_defun([_LT_PREPARE_SED_QUOTE_VARS], [# Backslashify metacharacters that are still active within # double-quoted strings. sed_quote_subst='s/\([["`$\\]]\)/\\\1/g' # Same as above, but do not quote variable references. double_quote_subst='s/\([["`\\]]\)/\\\1/g' # Sed substitution to delay expansion of an escaped shell variable in a # double_quote_subst'ed string. delay_variable_subst='s/\\\\\\\\\\\$/\\\\\\$/g' # Sed substitution to delay expansion of an escaped single quote. delay_single_quote_subst='s/'\''/'\'\\\\\\\'\''/g' # Sed substitution to avoid accidental globbing in evaled expressions no_glob_subst='s/\*/\\\*/g' ]) # _LT_PROG_LTMAIN # --------------- # Note that this code is called both from `configure', and `config.status' # now that we use AC_CONFIG_COMMANDS to generate libtool. Notably, # `config.status' has no value for ac_aux_dir unless we are using Automake, # so we pass a copy along to make sure it has a sensible value anyway. m4_defun([_LT_PROG_LTMAIN], [m4_ifdef([AC_REQUIRE_AUX_FILE], [AC_REQUIRE_AUX_FILE([ltmain.sh])])dnl _LT_CONFIG_LIBTOOL_INIT([ac_aux_dir='$ac_aux_dir']) ltmain="$ac_aux_dir/ltmain.sh" ])# _LT_PROG_LTMAIN ## ------------------------------------- ## ## Accumulate code for creating libtool. ## ## ------------------------------------- ## # So that we can recreate a full libtool script including additional # tags, we accumulate the chunks of code to send to AC_CONFIG_COMMANDS # in macros and then make a single call at the end using the `libtool' # label. # _LT_CONFIG_LIBTOOL_INIT([INIT-COMMANDS]) # ---------------------------------------- # Register INIT-COMMANDS to be passed to AC_CONFIG_COMMANDS later. m4_define([_LT_CONFIG_LIBTOOL_INIT], [m4_ifval([$1], [m4_append([_LT_OUTPUT_LIBTOOL_INIT], [$1 ])])]) # Initialize. m4_define([_LT_OUTPUT_LIBTOOL_INIT]) # _LT_CONFIG_LIBTOOL([COMMANDS]) # ------------------------------ # Register COMMANDS to be passed to AC_CONFIG_COMMANDS later. m4_define([_LT_CONFIG_LIBTOOL], [m4_ifval([$1], [m4_append([_LT_OUTPUT_LIBTOOL_COMMANDS], [$1 ])])]) # Initialize. m4_define([_LT_OUTPUT_LIBTOOL_COMMANDS]) # _LT_CONFIG_SAVE_COMMANDS([COMMANDS], [INIT_COMMANDS]) # ----------------------------------------------------- m4_defun([_LT_CONFIG_SAVE_COMMANDS], [_LT_CONFIG_LIBTOOL([$1]) _LT_CONFIG_LIBTOOL_INIT([$2]) ]) # _LT_FORMAT_COMMENT([COMMENT]) # ----------------------------- # Add leading comment marks to the start of each line, and a trailing # full-stop to the whole comment if one is not present already. m4_define([_LT_FORMAT_COMMENT], [m4_ifval([$1], [ m4_bpatsubst([m4_bpatsubst([$1], [^ *], [# ])], [['`$\]], [\\\&])]m4_bmatch([$1], [[!?.]$], [], [.]) )]) ## ------------------------ ## ## FIXME: Eliminate VARNAME ## ## ------------------------ ## # _LT_DECL([CONFIGNAME], VARNAME, VALUE, [DESCRIPTION], [IS-TAGGED?]) # ------------------------------------------------------------------- # CONFIGNAME is the name given to the value in the libtool script. # VARNAME is the (base) name used in the configure script. # VALUE may be 0, 1 or 2 for a computed quote escaped value based on # VARNAME. Any other value will be used directly. m4_define([_LT_DECL], [lt_if_append_uniq([lt_decl_varnames], [$2], [, ], [lt_dict_add_subkey([lt_decl_dict], [$2], [libtool_name], [m4_ifval([$1], [$1], [$2])]) lt_dict_add_subkey([lt_decl_dict], [$2], [value], [$3]) m4_ifval([$4], [lt_dict_add_subkey([lt_decl_dict], [$2], [description], [$4])]) lt_dict_add_subkey([lt_decl_dict], [$2], [tagged?], [m4_ifval([$5], [yes], [no])])]) ]) # _LT_TAGDECL([CONFIGNAME], VARNAME, VALUE, [DESCRIPTION]) # -------------------------------------------------------- m4_define([_LT_TAGDECL], [_LT_DECL([$1], [$2], [$3], [$4], [yes])]) # lt_decl_tag_varnames([SEPARATOR], [VARNAME1...]) # ------------------------------------------------ m4_define([lt_decl_tag_varnames], [_lt_decl_filter([tagged?], [yes], $@)]) # _lt_decl_filter(SUBKEY, VALUE, [SEPARATOR], [VARNAME1..]) # --------------------------------------------------------- m4_define([_lt_decl_filter], [m4_case([$#], [0], [m4_fatal([$0: too few arguments: $#])], [1], [m4_fatal([$0: too few arguments: $#: $1])], [2], [lt_dict_filter([lt_decl_dict], [$1], [$2], [], lt_decl_varnames)], [3], [lt_dict_filter([lt_decl_dict], [$1], [$2], [$3], lt_decl_varnames)], [lt_dict_filter([lt_decl_dict], $@)])[]dnl ]) # lt_decl_quote_varnames([SEPARATOR], [VARNAME1...]) # -------------------------------------------------- m4_define([lt_decl_quote_varnames], [_lt_decl_filter([value], [1], $@)]) # lt_decl_dquote_varnames([SEPARATOR], [VARNAME1...]) # --------------------------------------------------- m4_define([lt_decl_dquote_varnames], [_lt_decl_filter([value], [2], $@)]) # lt_decl_varnames_tagged([SEPARATOR], [VARNAME1...]) # --------------------------------------------------- m4_define([lt_decl_varnames_tagged], [m4_assert([$# <= 2])dnl _$0(m4_quote(m4_default([$1], [[, ]])), m4_ifval([$2], [[$2]], [m4_dquote(lt_decl_tag_varnames)]), m4_split(m4_normalize(m4_quote(_LT_TAGS)), [ ]))]) m4_define([_lt_decl_varnames_tagged], [m4_ifval([$3], [lt_combine([$1], [$2], [_], $3)])]) # lt_decl_all_varnames([SEPARATOR], [VARNAME1...]) # ------------------------------------------------ m4_define([lt_decl_all_varnames], [_$0(m4_quote(m4_default([$1], [[, ]])), m4_if([$2], [], m4_quote(lt_decl_varnames), m4_quote(m4_shift($@))))[]dnl ]) m4_define([_lt_decl_all_varnames], [lt_join($@, lt_decl_varnames_tagged([$1], lt_decl_tag_varnames([[, ]], m4_shift($@))))dnl ]) # _LT_CONFIG_STATUS_DECLARE([VARNAME]) # ------------------------------------ # Quote a variable value, and forward it to `config.status' so that its # declaration there will have the same value as in `configure'. VARNAME # must have a single quote delimited value for this to work. m4_define([_LT_CONFIG_STATUS_DECLARE], [$1='`$ECHO "$][$1" | $SED "$delay_single_quote_subst"`']) # _LT_CONFIG_STATUS_DECLARATIONS # ------------------------------ # We delimit libtool config variables with single quotes, so when # we write them to config.status, we have to be sure to quote all # embedded single quotes properly. In configure, this macro expands # each variable declared with _LT_DECL (and _LT_TAGDECL) into: # # ='`$ECHO "$" | $SED "$delay_single_quote_subst"`' m4_defun([_LT_CONFIG_STATUS_DECLARATIONS], [m4_foreach([_lt_var], m4_quote(lt_decl_all_varnames), [m4_n([_LT_CONFIG_STATUS_DECLARE(_lt_var)])])]) # _LT_LIBTOOL_TAGS # ---------------- # Output comment and list of tags supported by the script m4_defun([_LT_LIBTOOL_TAGS], [_LT_FORMAT_COMMENT([The names of the tagged configurations supported by this script])dnl available_tags="_LT_TAGS"dnl ]) # _LT_LIBTOOL_DECLARE(VARNAME, [TAG]) # ----------------------------------- # Extract the dictionary values for VARNAME (optionally with TAG) and # expand to a commented shell variable setting: # # # Some comment about what VAR is for. # visible_name=$lt_internal_name m4_define([_LT_LIBTOOL_DECLARE], [_LT_FORMAT_COMMENT(m4_quote(lt_dict_fetch([lt_decl_dict], [$1], [description])))[]dnl m4_pushdef([_libtool_name], m4_quote(lt_dict_fetch([lt_decl_dict], [$1], [libtool_name])))[]dnl m4_case(m4_quote(lt_dict_fetch([lt_decl_dict], [$1], [value])), [0], [_libtool_name=[$]$1], [1], [_libtool_name=$lt_[]$1], [2], [_libtool_name=$lt_[]$1], [_libtool_name=lt_dict_fetch([lt_decl_dict], [$1], [value])])[]dnl m4_ifval([$2], [_$2])[]m4_popdef([_libtool_name])[]dnl ]) # _LT_LIBTOOL_CONFIG_VARS # ----------------------- # Produce commented declarations of non-tagged libtool config variables # suitable for insertion in the LIBTOOL CONFIG section of the `libtool' # script. Tagged libtool config variables (even for the LIBTOOL CONFIG # section) are produced by _LT_LIBTOOL_TAG_VARS. m4_defun([_LT_LIBTOOL_CONFIG_VARS], [m4_foreach([_lt_var], m4_quote(_lt_decl_filter([tagged?], [no], [], lt_decl_varnames)), [m4_n([_LT_LIBTOOL_DECLARE(_lt_var)])])]) # _LT_LIBTOOL_TAG_VARS(TAG) # ------------------------- m4_define([_LT_LIBTOOL_TAG_VARS], [m4_foreach([_lt_var], m4_quote(lt_decl_tag_varnames), [m4_n([_LT_LIBTOOL_DECLARE(_lt_var, [$1])])])]) # _LT_TAGVAR(VARNAME, [TAGNAME]) # ------------------------------ m4_define([_LT_TAGVAR], [m4_ifval([$2], [$1_$2], [$1])]) # _LT_CONFIG_COMMANDS # ------------------- # Send accumulated output to $CONFIG_STATUS. Thanks to the lists of # variables for single and double quote escaping we saved from calls # to _LT_DECL, we can put quote escaped variables declarations # into `config.status', and then the shell code to quote escape them in # for loops in `config.status'. Finally, any additional code accumulated # from calls to _LT_CONFIG_LIBTOOL_INIT is expanded. m4_defun([_LT_CONFIG_COMMANDS], [AC_PROVIDE_IFELSE([LT_OUTPUT], dnl If the libtool generation code has been placed in $CONFIG_LT, dnl instead of duplicating it all over again into config.status, dnl then we will have config.status run $CONFIG_LT later, so it dnl needs to know what name is stored there: [AC_CONFIG_COMMANDS([libtool], [$SHELL $CONFIG_LT || AS_EXIT(1)], [CONFIG_LT='$CONFIG_LT'])], dnl If the libtool generation code is destined for config.status, dnl expand the accumulated commands and init code now: [AC_CONFIG_COMMANDS([libtool], [_LT_OUTPUT_LIBTOOL_COMMANDS], [_LT_OUTPUT_LIBTOOL_COMMANDS_INIT])]) ])#_LT_CONFIG_COMMANDS # Initialize. m4_define([_LT_OUTPUT_LIBTOOL_COMMANDS_INIT], [ # The HP-UX ksh and POSIX shell print the target directory to stdout # if CDPATH is set. (unset CDPATH) >/dev/null 2>&1 && unset CDPATH sed_quote_subst='$sed_quote_subst' double_quote_subst='$double_quote_subst' delay_variable_subst='$delay_variable_subst' _LT_CONFIG_STATUS_DECLARATIONS LTCC='$LTCC' LTCFLAGS='$LTCFLAGS' compiler='$compiler_DEFAULT' # A function that is used when there is no print builtin or printf. func_fallback_echo () { eval 'cat <<_LTECHO_EOF \$[]1 _LTECHO_EOF' } # Quote evaled strings. for var in lt_decl_all_varnames([[ \ ]], lt_decl_quote_varnames); do case \`eval \\\\\$ECHO \\\\""\\\\\$\$var"\\\\"\` in *[[\\\\\\\`\\"\\\$]]*) eval "lt_\$var=\\\\\\"\\\`\\\$ECHO \\"\\\$\$var\\" | \\\$SED \\"\\\$sed_quote_subst\\"\\\`\\\\\\"" ;; *) eval "lt_\$var=\\\\\\"\\\$\$var\\\\\\"" ;; esac done # Double-quote double-evaled strings. for var in lt_decl_all_varnames([[ \ ]], lt_decl_dquote_varnames); do case \`eval \\\\\$ECHO \\\\""\\\\\$\$var"\\\\"\` in *[[\\\\\\\`\\"\\\$]]*) eval "lt_\$var=\\\\\\"\\\`\\\$ECHO \\"\\\$\$var\\" | \\\$SED -e \\"\\\$double_quote_subst\\" -e \\"\\\$sed_quote_subst\\" -e \\"\\\$delay_variable_subst\\"\\\`\\\\\\"" ;; *) eval "lt_\$var=\\\\\\"\\\$\$var\\\\\\"" ;; esac done _LT_OUTPUT_LIBTOOL_INIT ]) # _LT_GENERATED_FILE_INIT(FILE, [COMMENT]) # ------------------------------------ # Generate a child script FILE with all initialization necessary to # reuse the environment learned by the parent script, and make the # file executable. If COMMENT is supplied, it is inserted after the # `#!' sequence but before initialization text begins. After this # macro, additional text can be appended to FILE to form the body of # the child script. The macro ends with non-zero status if the # file could not be fully written (such as if the disk is full). m4_ifdef([AS_INIT_GENERATED], [m4_defun([_LT_GENERATED_FILE_INIT],[AS_INIT_GENERATED($@)])], [m4_defun([_LT_GENERATED_FILE_INIT], [m4_require([AS_PREPARE])]dnl [m4_pushdef([AS_MESSAGE_LOG_FD])]dnl [lt_write_fail=0 cat >$1 <<_ASEOF || lt_write_fail=1 #! $SHELL # Generated by $as_me. $2 SHELL=\${CONFIG_SHELL-$SHELL} export SHELL _ASEOF cat >>$1 <<\_ASEOF || lt_write_fail=1 AS_SHELL_SANITIZE _AS_PREPARE exec AS_MESSAGE_FD>&1 _ASEOF test $lt_write_fail = 0 && chmod +x $1[]dnl m4_popdef([AS_MESSAGE_LOG_FD])])])# _LT_GENERATED_FILE_INIT # LT_OUTPUT # --------- # This macro allows early generation of the libtool script (before # AC_OUTPUT is called), incase it is used in configure for compilation # tests. AC_DEFUN([LT_OUTPUT], [: ${CONFIG_LT=./config.lt} AC_MSG_NOTICE([creating $CONFIG_LT]) _LT_GENERATED_FILE_INIT(["$CONFIG_LT"], [# Run this file to recreate a libtool stub with the current configuration.]) cat >>"$CONFIG_LT" <<\_LTEOF lt_cl_silent=false exec AS_MESSAGE_LOG_FD>>config.log { echo AS_BOX([Running $as_me.]) } >&AS_MESSAGE_LOG_FD lt_cl_help="\ \`$as_me' creates a local libtool stub from the current configuration, for use in further configure time tests before the real libtool is generated. Usage: $[0] [[OPTIONS]] -h, --help print this help, then exit -V, --version print version number, then exit -q, --quiet do not print progress messages -d, --debug don't remove temporary files Report bugs to ." lt_cl_version="\ m4_ifset([AC_PACKAGE_NAME], [AC_PACKAGE_NAME ])config.lt[]dnl m4_ifset([AC_PACKAGE_VERSION], [ AC_PACKAGE_VERSION]) configured by $[0], generated by m4_PACKAGE_STRING. Copyright (C) 2011 Free Software Foundation, Inc. This config.lt script is free software; the Free Software Foundation gives unlimited permision to copy, distribute and modify it." while test $[#] != 0 do case $[1] in --version | --v* | -V ) echo "$lt_cl_version"; exit 0 ;; --help | --h* | -h ) echo "$lt_cl_help"; exit 0 ;; --debug | --d* | -d ) debug=: ;; --quiet | --q* | --silent | --s* | -q ) lt_cl_silent=: ;; -*) AC_MSG_ERROR([unrecognized option: $[1] Try \`$[0] --help' for more information.]) ;; *) AC_MSG_ERROR([unrecognized argument: $[1] Try \`$[0] --help' for more information.]) ;; esac shift done if $lt_cl_silent; then exec AS_MESSAGE_FD>/dev/null fi _LTEOF cat >>"$CONFIG_LT" <<_LTEOF _LT_OUTPUT_LIBTOOL_COMMANDS_INIT _LTEOF cat >>"$CONFIG_LT" <<\_LTEOF AC_MSG_NOTICE([creating $ofile]) _LT_OUTPUT_LIBTOOL_COMMANDS AS_EXIT(0) _LTEOF chmod +x "$CONFIG_LT" # configure is writing to config.log, but config.lt does its own redirection, # appending to config.log, which fails on DOS, as config.log is still kept # open by configure. Here we exec the FD to /dev/null, effectively closing # config.log, so it can be properly (re)opened and appended to by config.lt. lt_cl_success=: test "$silent" = yes && lt_config_lt_args="$lt_config_lt_args --quiet" exec AS_MESSAGE_LOG_FD>/dev/null $SHELL "$CONFIG_LT" $lt_config_lt_args || lt_cl_success=false exec AS_MESSAGE_LOG_FD>>config.log $lt_cl_success || AS_EXIT(1) ])# LT_OUTPUT # _LT_CONFIG(TAG) # --------------- # If TAG is the built-in tag, create an initial libtool script with a # default configuration from the untagged config vars. Otherwise add code # to config.status for appending the configuration named by TAG from the # matching tagged config vars. m4_defun([_LT_CONFIG], [m4_require([_LT_FILEUTILS_DEFAULTS])dnl _LT_CONFIG_SAVE_COMMANDS([ m4_define([_LT_TAG], m4_if([$1], [], [C], [$1]))dnl m4_if(_LT_TAG, [C], [ # See if we are running on zsh, and set the options which allow our # commands through without removal of \ escapes. if test -n "${ZSH_VERSION+set}" ; then setopt NO_GLOB_SUBST fi cfgfile="${ofile}T" trap "$RM \"$cfgfile\"; exit 1" 1 2 15 $RM "$cfgfile" cat <<_LT_EOF >> "$cfgfile" #! $SHELL # `$ECHO "$ofile" | sed 's%^.*/%%'` - Provide generalized library-building support services. # Generated automatically by $as_me ($PACKAGE$TIMESTAMP) $VERSION # Libtool was configured on host `(hostname || uname -n) 2>/dev/null | sed 1q`: # NOTE: Changes made to this file will be lost: look at ltmain.sh. # _LT_COPYING _LT_LIBTOOL_TAGS # ### BEGIN LIBTOOL CONFIG _LT_LIBTOOL_CONFIG_VARS _LT_LIBTOOL_TAG_VARS # ### END LIBTOOL CONFIG _LT_EOF case $host_os in aix3*) cat <<\_LT_EOF >> "$cfgfile" # AIX sometimes has problems with the GCC collect2 program. For some # reason, if we set the COLLECT_NAMES environment variable, the problems # vanish in a puff of smoke. if test "X${COLLECT_NAMES+set}" != Xset; then COLLECT_NAMES= export COLLECT_NAMES fi _LT_EOF ;; esac _LT_PROG_LTMAIN # We use sed instead of cat because bash on DJGPP gets confused if # if finds mixed CR/LF and LF-only lines. Since sed operates in # text mode, it properly converts lines to CR/LF. This bash problem # is reportedly fixed, but why not run on old versions too? sed '$q' "$ltmain" >> "$cfgfile" \ || (rm -f "$cfgfile"; exit 1) _LT_PROG_REPLACE_SHELLFNS mv -f "$cfgfile" "$ofile" || (rm -f "$ofile" && cp "$cfgfile" "$ofile" && rm -f "$cfgfile") chmod +x "$ofile" ], [cat <<_LT_EOF >> "$ofile" dnl Unfortunately we have to use $1 here, since _LT_TAG is not expanded dnl in a comment (ie after a #). # ### BEGIN LIBTOOL TAG CONFIG: $1 _LT_LIBTOOL_TAG_VARS(_LT_TAG) # ### END LIBTOOL TAG CONFIG: $1 _LT_EOF ])dnl /m4_if ], [m4_if([$1], [], [ PACKAGE='$PACKAGE' VERSION='$VERSION' TIMESTAMP='$TIMESTAMP' RM='$RM' ofile='$ofile'], []) ])dnl /_LT_CONFIG_SAVE_COMMANDS ])# _LT_CONFIG # LT_SUPPORTED_TAG(TAG) # --------------------- # Trace this macro to discover what tags are supported by the libtool # --tag option, using: # autoconf --trace 'LT_SUPPORTED_TAG:$1' AC_DEFUN([LT_SUPPORTED_TAG], []) # C support is built-in for now m4_define([_LT_LANG_C_enabled], []) m4_define([_LT_TAGS], []) # LT_LANG(LANG) # ------------- # Enable libtool support for the given language if not already enabled. AC_DEFUN([LT_LANG], [AC_BEFORE([$0], [LT_OUTPUT])dnl m4_case([$1], [C], [_LT_LANG(C)], [C++], [_LT_LANG(CXX)], [Go], [_LT_LANG(GO)], [Java], [_LT_LANG(GCJ)], [Fortran 77], [_LT_LANG(F77)], [Fortran], [_LT_LANG(FC)], [Windows Resource], [_LT_LANG(RC)], [m4_ifdef([_LT_LANG_]$1[_CONFIG], [_LT_LANG($1)], [m4_fatal([$0: unsupported language: "$1"])])])dnl ])# LT_LANG # _LT_LANG(LANGNAME) # ------------------ m4_defun([_LT_LANG], [m4_ifdef([_LT_LANG_]$1[_enabled], [], [LT_SUPPORTED_TAG([$1])dnl m4_append([_LT_TAGS], [$1 ])dnl m4_define([_LT_LANG_]$1[_enabled], [])dnl _LT_LANG_$1_CONFIG($1)])dnl ])# _LT_LANG m4_ifndef([AC_PROG_GO], [ ############################################################ # NOTE: This macro has been submitted for inclusion into # # GNU Autoconf as AC_PROG_GO. When it is available in # # a released version of Autoconf we should remove this # # macro and use it instead. # ############################################################ m4_defun([AC_PROG_GO], [AC_LANG_PUSH(Go)dnl AC_ARG_VAR([GOC], [Go compiler command])dnl AC_ARG_VAR([GOFLAGS], [Go compiler flags])dnl _AC_ARG_VAR_LDFLAGS()dnl AC_CHECK_TOOL(GOC, gccgo) if test -z "$GOC"; then if test -n "$ac_tool_prefix"; then AC_CHECK_PROG(GOC, [${ac_tool_prefix}gccgo], [${ac_tool_prefix}gccgo]) fi fi if test -z "$GOC"; then AC_CHECK_PROG(GOC, gccgo, gccgo, false) fi ])#m4_defun ])#m4_ifndef # _LT_LANG_DEFAULT_CONFIG # ----------------------- m4_defun([_LT_LANG_DEFAULT_CONFIG], [AC_PROVIDE_IFELSE([AC_PROG_CXX], [LT_LANG(CXX)], [m4_define([AC_PROG_CXX], defn([AC_PROG_CXX])[LT_LANG(CXX)])]) AC_PROVIDE_IFELSE([AC_PROG_F77], [LT_LANG(F77)], [m4_define([AC_PROG_F77], defn([AC_PROG_F77])[LT_LANG(F77)])]) AC_PROVIDE_IFELSE([AC_PROG_FC], [LT_LANG(FC)], [m4_define([AC_PROG_FC], defn([AC_PROG_FC])[LT_LANG(FC)])]) dnl The call to [A][M_PROG_GCJ] is quoted like that to stop aclocal dnl pulling things in needlessly. AC_PROVIDE_IFELSE([AC_PROG_GCJ], [LT_LANG(GCJ)], [AC_PROVIDE_IFELSE([A][M_PROG_GCJ], [LT_LANG(GCJ)], [AC_PROVIDE_IFELSE([LT_PROG_GCJ], [LT_LANG(GCJ)], [m4_ifdef([AC_PROG_GCJ], [m4_define([AC_PROG_GCJ], defn([AC_PROG_GCJ])[LT_LANG(GCJ)])]) m4_ifdef([A][M_PROG_GCJ], [m4_define([A][M_PROG_GCJ], defn([A][M_PROG_GCJ])[LT_LANG(GCJ)])]) m4_ifdef([LT_PROG_GCJ], [m4_define([LT_PROG_GCJ], defn([LT_PROG_GCJ])[LT_LANG(GCJ)])])])])]) AC_PROVIDE_IFELSE([AC_PROG_GO], [LT_LANG(GO)], [m4_define([AC_PROG_GO], defn([AC_PROG_GO])[LT_LANG(GO)])]) AC_PROVIDE_IFELSE([LT_PROG_RC], [LT_LANG(RC)], [m4_define([LT_PROG_RC], defn([LT_PROG_RC])[LT_LANG(RC)])]) ])# _LT_LANG_DEFAULT_CONFIG # Obsolete macros: AU_DEFUN([AC_LIBTOOL_CXX], [LT_LANG(C++)]) AU_DEFUN([AC_LIBTOOL_F77], [LT_LANG(Fortran 77)]) AU_DEFUN([AC_LIBTOOL_FC], [LT_LANG(Fortran)]) AU_DEFUN([AC_LIBTOOL_GCJ], [LT_LANG(Java)]) AU_DEFUN([AC_LIBTOOL_RC], [LT_LANG(Windows Resource)]) dnl aclocal-1.4 backwards compatibility: dnl AC_DEFUN([AC_LIBTOOL_CXX], []) dnl AC_DEFUN([AC_LIBTOOL_F77], []) dnl AC_DEFUN([AC_LIBTOOL_FC], []) dnl AC_DEFUN([AC_LIBTOOL_GCJ], []) dnl AC_DEFUN([AC_LIBTOOL_RC], []) # _LT_TAG_COMPILER # ---------------- m4_defun([_LT_TAG_COMPILER], [AC_REQUIRE([AC_PROG_CC])dnl _LT_DECL([LTCC], [CC], [1], [A C compiler])dnl _LT_DECL([LTCFLAGS], [CFLAGS], [1], [LTCC compiler flags])dnl _LT_TAGDECL([CC], [compiler], [1], [A language specific compiler])dnl _LT_TAGDECL([with_gcc], [GCC], [0], [Is the compiler the GNU compiler?])dnl # If no C compiler was specified, use CC. LTCC=${LTCC-"$CC"} # If no C compiler flags were specified, use CFLAGS. LTCFLAGS=${LTCFLAGS-"$CFLAGS"} # Allow CC to be a program name with arguments. compiler=$CC ])# _LT_TAG_COMPILER # _LT_COMPILER_BOILERPLATE # ------------------------ # Check for compiler boilerplate output or warnings with # the simple compiler test code. m4_defun([_LT_COMPILER_BOILERPLATE], [m4_require([_LT_DECL_SED])dnl ac_outfile=conftest.$ac_objext echo "$lt_simple_compile_test_code" >conftest.$ac_ext eval "$ac_compile" 2>&1 >/dev/null | $SED '/^$/d; /^ *+/d' >conftest.err _lt_compiler_boilerplate=`cat conftest.err` $RM conftest* ])# _LT_COMPILER_BOILERPLATE # _LT_LINKER_BOILERPLATE # ---------------------- # Check for linker boilerplate output or warnings with # the simple link test code. m4_defun([_LT_LINKER_BOILERPLATE], [m4_require([_LT_DECL_SED])dnl ac_outfile=conftest.$ac_objext echo "$lt_simple_link_test_code" >conftest.$ac_ext eval "$ac_link" 2>&1 >/dev/null | $SED '/^$/d; /^ *+/d' >conftest.err _lt_linker_boilerplate=`cat conftest.err` $RM -r conftest* ])# _LT_LINKER_BOILERPLATE # _LT_REQUIRED_DARWIN_CHECKS # ------------------------- m4_defun_once([_LT_REQUIRED_DARWIN_CHECKS],[ case $host_os in rhapsody* | darwin*) AC_CHECK_TOOL([DSYMUTIL], [dsymutil], [:]) AC_CHECK_TOOL([NMEDIT], [nmedit], [:]) AC_CHECK_TOOL([LIPO], [lipo], [:]) AC_CHECK_TOOL([OTOOL], [otool], [:]) AC_CHECK_TOOL([OTOOL64], [otool64], [:]) _LT_DECL([], [DSYMUTIL], [1], [Tool to manipulate archived DWARF debug symbol files on Mac OS X]) _LT_DECL([], [NMEDIT], [1], [Tool to change global to local symbols on Mac OS X]) _LT_DECL([], [LIPO], [1], [Tool to manipulate fat objects and archives on Mac OS X]) _LT_DECL([], [OTOOL], [1], [ldd/readelf like tool for Mach-O binaries on Mac OS X]) _LT_DECL([], [OTOOL64], [1], [ldd/readelf like tool for 64 bit Mach-O binaries on Mac OS X 10.4]) AC_CACHE_CHECK([for -single_module linker flag],[lt_cv_apple_cc_single_mod], [lt_cv_apple_cc_single_mod=no if test -z "${LT_MULTI_MODULE}"; then # By default we will add the -single_module flag. You can override # by either setting the environment variable LT_MULTI_MODULE # non-empty at configure time, or by adding -multi_module to the # link flags. rm -rf libconftest.dylib* echo "int foo(void){return 1;}" > conftest.c echo "$LTCC $LTCFLAGS $LDFLAGS -o libconftest.dylib \ -dynamiclib -Wl,-single_module conftest.c" >&AS_MESSAGE_LOG_FD $LTCC $LTCFLAGS $LDFLAGS -o libconftest.dylib \ -dynamiclib -Wl,-single_module conftest.c 2>conftest.err _lt_result=$? # If there is a non-empty error log, and "single_module" # appears in it, assume the flag caused a linker warning if test -s conftest.err && $GREP single_module conftest.err; then cat conftest.err >&AS_MESSAGE_LOG_FD # Otherwise, if the output was created with a 0 exit code from # the compiler, it worked. elif test -f libconftest.dylib && test $_lt_result -eq 0; then lt_cv_apple_cc_single_mod=yes else cat conftest.err >&AS_MESSAGE_LOG_FD fi rm -rf libconftest.dylib* rm -f conftest.* fi]) AC_CACHE_CHECK([for -exported_symbols_list linker flag], [lt_cv_ld_exported_symbols_list], [lt_cv_ld_exported_symbols_list=no save_LDFLAGS=$LDFLAGS echo "_main" > conftest.sym LDFLAGS="$LDFLAGS -Wl,-exported_symbols_list,conftest.sym" AC_LINK_IFELSE([AC_LANG_PROGRAM([],[])], [lt_cv_ld_exported_symbols_list=yes], [lt_cv_ld_exported_symbols_list=no]) LDFLAGS="$save_LDFLAGS" ]) AC_CACHE_CHECK([for -force_load linker flag],[lt_cv_ld_force_load], [lt_cv_ld_force_load=no cat > conftest.c << _LT_EOF int forced_loaded() { return 2;} _LT_EOF echo "$LTCC $LTCFLAGS -c -o conftest.o conftest.c" >&AS_MESSAGE_LOG_FD $LTCC $LTCFLAGS -c -o conftest.o conftest.c 2>&AS_MESSAGE_LOG_FD echo "$AR cru libconftest.a conftest.o" >&AS_MESSAGE_LOG_FD $AR cru libconftest.a conftest.o 2>&AS_MESSAGE_LOG_FD echo "$RANLIB libconftest.a" >&AS_MESSAGE_LOG_FD $RANLIB libconftest.a 2>&AS_MESSAGE_LOG_FD cat > conftest.c << _LT_EOF int main() { return 0;} _LT_EOF echo "$LTCC $LTCFLAGS $LDFLAGS -o conftest conftest.c -Wl,-force_load,./libconftest.a" >&AS_MESSAGE_LOG_FD $LTCC $LTCFLAGS $LDFLAGS -o conftest conftest.c -Wl,-force_load,./libconftest.a 2>conftest.err _lt_result=$? if test -s conftest.err && $GREP force_load conftest.err; then cat conftest.err >&AS_MESSAGE_LOG_FD elif test -f conftest && test $_lt_result -eq 0 && $GREP forced_load conftest >/dev/null 2>&1 ; then lt_cv_ld_force_load=yes else cat conftest.err >&AS_MESSAGE_LOG_FD fi rm -f conftest.err libconftest.a conftest conftest.c rm -rf conftest.dSYM ]) case $host_os in rhapsody* | darwin1.[[012]]) _lt_dar_allow_undefined='${wl}-undefined ${wl}suppress' ;; darwin1.*) _lt_dar_allow_undefined='${wl}-flat_namespace ${wl}-undefined ${wl}suppress' ;; darwin*) # darwin 5.x on # if running on 10.5 or later, the deployment target defaults # to the OS version, if on x86, and 10.4, the deployment # target defaults to 10.4. Don't you love it? case ${MACOSX_DEPLOYMENT_TARGET-10.0},$host in 10.0,*86*-darwin8*|10.0,*-darwin[[91]]*) _lt_dar_allow_undefined='${wl}-undefined ${wl}dynamic_lookup' ;; 10.[[012]]*) _lt_dar_allow_undefined='${wl}-flat_namespace ${wl}-undefined ${wl}suppress' ;; 10.*) _lt_dar_allow_undefined='${wl}-undefined ${wl}dynamic_lookup' ;; esac ;; esac if test "$lt_cv_apple_cc_single_mod" = "yes"; then _lt_dar_single_mod='$single_module' fi if test "$lt_cv_ld_exported_symbols_list" = "yes"; then _lt_dar_export_syms=' ${wl}-exported_symbols_list,$output_objdir/${libname}-symbols.expsym' else _lt_dar_export_syms='~$NMEDIT -s $output_objdir/${libname}-symbols.expsym ${lib}' fi if test "$DSYMUTIL" != ":" && test "$lt_cv_ld_force_load" = "no"; then _lt_dsymutil='~$DSYMUTIL $lib || :' else _lt_dsymutil= fi ;; esac ]) # _LT_DARWIN_LINKER_FEATURES([TAG]) # --------------------------------- # Checks for linker and compiler features on darwin m4_defun([_LT_DARWIN_LINKER_FEATURES], [ m4_require([_LT_REQUIRED_DARWIN_CHECKS]) _LT_TAGVAR(archive_cmds_need_lc, $1)=no _LT_TAGVAR(hardcode_direct, $1)=no _LT_TAGVAR(hardcode_automatic, $1)=yes _LT_TAGVAR(hardcode_shlibpath_var, $1)=unsupported if test "$lt_cv_ld_force_load" = "yes"; then _LT_TAGVAR(whole_archive_flag_spec, $1)='`for conv in $convenience\"\"; do test -n \"$conv\" && new_convenience=\"$new_convenience ${wl}-force_load,$conv\"; done; func_echo_all \"$new_convenience\"`' m4_case([$1], [F77], [_LT_TAGVAR(compiler_needs_object, $1)=yes], [FC], [_LT_TAGVAR(compiler_needs_object, $1)=yes]) else _LT_TAGVAR(whole_archive_flag_spec, $1)='' fi _LT_TAGVAR(link_all_deplibs, $1)=yes _LT_TAGVAR(allow_undefined_flag, $1)="$_lt_dar_allow_undefined" case $cc_basename in ifort*) _lt_dar_can_shared=yes ;; *) _lt_dar_can_shared=$GCC ;; esac if test "$_lt_dar_can_shared" = "yes"; then output_verbose_link_cmd=func_echo_all _LT_TAGVAR(archive_cmds, $1)="\$CC -dynamiclib \$allow_undefined_flag -o \$lib \$libobjs \$deplibs \$compiler_flags -install_name \$rpath/\$soname \$verstring $_lt_dar_single_mod${_lt_dsymutil}" _LT_TAGVAR(module_cmds, $1)="\$CC \$allow_undefined_flag -o \$lib -bundle \$libobjs \$deplibs \$compiler_flags${_lt_dsymutil}" _LT_TAGVAR(archive_expsym_cmds, $1)="sed 's,^,_,' < \$export_symbols > \$output_objdir/\${libname}-symbols.expsym~\$CC -dynamiclib \$allow_undefined_flag -o \$lib \$libobjs \$deplibs \$compiler_flags -install_name \$rpath/\$soname \$verstring ${_lt_dar_single_mod}${_lt_dar_export_syms}${_lt_dsymutil}" _LT_TAGVAR(module_expsym_cmds, $1)="sed -e 's,^,_,' < \$export_symbols > \$output_objdir/\${libname}-symbols.expsym~\$CC \$allow_undefined_flag -o \$lib -bundle \$libobjs \$deplibs \$compiler_flags${_lt_dar_export_syms}${_lt_dsymutil}" m4_if([$1], [CXX], [ if test "$lt_cv_apple_cc_single_mod" != "yes"; then _LT_TAGVAR(archive_cmds, $1)="\$CC -r -keep_private_externs -nostdlib -o \${lib}-master.o \$libobjs~\$CC -dynamiclib \$allow_undefined_flag -o \$lib \${lib}-master.o \$deplibs \$compiler_flags -install_name \$rpath/\$soname \$verstring${_lt_dsymutil}" _LT_TAGVAR(archive_expsym_cmds, $1)="sed 's,^,_,' < \$export_symbols > \$output_objdir/\${libname}-symbols.expsym~\$CC -r -keep_private_externs -nostdlib -o \${lib}-master.o \$libobjs~\$CC -dynamiclib \$allow_undefined_flag -o \$lib \${lib}-master.o \$deplibs \$compiler_flags -install_name \$rpath/\$soname \$verstring${_lt_dar_export_syms}${_lt_dsymutil}" fi ],[]) else _LT_TAGVAR(ld_shlibs, $1)=no fi ]) # _LT_SYS_MODULE_PATH_AIX([TAGNAME]) # ---------------------------------- # Links a minimal program and checks the executable # for the system default hardcoded library path. In most cases, # this is /usr/lib:/lib, but when the MPI compilers are used # the location of the communication and MPI libs are included too. # If we don't find anything, use the default library path according # to the aix ld manual. # Store the results from the different compilers for each TAGNAME. # Allow to override them for all tags through lt_cv_aix_libpath. m4_defun([_LT_SYS_MODULE_PATH_AIX], [m4_require([_LT_DECL_SED])dnl if test "${lt_cv_aix_libpath+set}" = set; then aix_libpath=$lt_cv_aix_libpath else AC_CACHE_VAL([_LT_TAGVAR([lt_cv_aix_libpath_], [$1])], [AC_LINK_IFELSE([AC_LANG_PROGRAM],[ lt_aix_libpath_sed='[ /Import File Strings/,/^$/ { /^0/ { s/^0 *\([^ ]*\) *$/\1/ p } }]' _LT_TAGVAR([lt_cv_aix_libpath_], [$1])=`dump -H conftest$ac_exeext 2>/dev/null | $SED -n -e "$lt_aix_libpath_sed"` # Check for a 64-bit object if we didn't find anything. if test -z "$_LT_TAGVAR([lt_cv_aix_libpath_], [$1])"; then _LT_TAGVAR([lt_cv_aix_libpath_], [$1])=`dump -HX64 conftest$ac_exeext 2>/dev/null | $SED -n -e "$lt_aix_libpath_sed"` fi],[]) if test -z "$_LT_TAGVAR([lt_cv_aix_libpath_], [$1])"; then _LT_TAGVAR([lt_cv_aix_libpath_], [$1])="/usr/lib:/lib" fi ]) aix_libpath=$_LT_TAGVAR([lt_cv_aix_libpath_], [$1]) fi ])# _LT_SYS_MODULE_PATH_AIX # _LT_SHELL_INIT(ARG) # ------------------- m4_define([_LT_SHELL_INIT], [m4_divert_text([M4SH-INIT], [$1 ])])# _LT_SHELL_INIT # _LT_PROG_ECHO_BACKSLASH # ----------------------- # Find how we can fake an echo command that does not interpret backslash. # In particular, with Autoconf 2.60 or later we add some code to the start # of the generated configure script which will find a shell with a builtin # printf (which we can use as an echo command). m4_defun([_LT_PROG_ECHO_BACKSLASH], [ECHO='\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\' ECHO=$ECHO$ECHO$ECHO$ECHO$ECHO ECHO=$ECHO$ECHO$ECHO$ECHO$ECHO$ECHO AC_MSG_CHECKING([how to print strings]) # Test print first, because it will be a builtin if present. if test "X`( print -r -- -n ) 2>/dev/null`" = X-n && \ test "X`print -r -- $ECHO 2>/dev/null`" = "X$ECHO"; then ECHO='print -r --' elif test "X`printf %s $ECHO 2>/dev/null`" = "X$ECHO"; then ECHO='printf %s\n' else # Use this function as a fallback that always works. func_fallback_echo () { eval 'cat <<_LTECHO_EOF $[]1 _LTECHO_EOF' } ECHO='func_fallback_echo' fi # func_echo_all arg... # Invoke $ECHO with all args, space-separated. func_echo_all () { $ECHO "$*" } case "$ECHO" in printf*) AC_MSG_RESULT([printf]) ;; print*) AC_MSG_RESULT([print -r]) ;; *) AC_MSG_RESULT([cat]) ;; esac m4_ifdef([_AS_DETECT_SUGGESTED], [_AS_DETECT_SUGGESTED([ test -n "${ZSH_VERSION+set}${BASH_VERSION+set}" || ( ECHO='\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\' ECHO=$ECHO$ECHO$ECHO$ECHO$ECHO ECHO=$ECHO$ECHO$ECHO$ECHO$ECHO$ECHO PATH=/empty FPATH=/empty; export PATH FPATH test "X`printf %s $ECHO`" = "X$ECHO" \ || test "X`print -r -- $ECHO`" = "X$ECHO" )])]) _LT_DECL([], [SHELL], [1], [Shell to use when invoking shell scripts]) _LT_DECL([], [ECHO], [1], [An echo program that protects backslashes]) ])# _LT_PROG_ECHO_BACKSLASH # _LT_WITH_SYSROOT # ---------------- AC_DEFUN([_LT_WITH_SYSROOT], [AC_MSG_CHECKING([for sysroot]) AC_ARG_WITH([sysroot], [ --with-sysroot[=DIR] Search for dependent libraries within DIR (or the compiler's sysroot if not specified).], [], [with_sysroot=no]) dnl lt_sysroot will always be passed unquoted. We quote it here dnl in case the user passed a directory name. lt_sysroot= case ${with_sysroot} in #( yes) if test "$GCC" = yes; then lt_sysroot=`$CC --print-sysroot 2>/dev/null` fi ;; #( /*) lt_sysroot=`echo "$with_sysroot" | sed -e "$sed_quote_subst"` ;; #( no|'') ;; #( *) AC_MSG_RESULT([${with_sysroot}]) AC_MSG_ERROR([The sysroot must be an absolute path.]) ;; esac AC_MSG_RESULT([${lt_sysroot:-no}]) _LT_DECL([], [lt_sysroot], [0], [The root where to search for ]dnl [dependent libraries, and in which our libraries should be installed.])]) # _LT_ENABLE_LOCK # --------------- m4_defun([_LT_ENABLE_LOCK], [AC_ARG_ENABLE([libtool-lock], [AS_HELP_STRING([--disable-libtool-lock], [avoid locking (might break parallel builds)])]) test "x$enable_libtool_lock" != xno && enable_libtool_lock=yes # Some flags need to be propagated to the compiler or linker for good # libtool support. case $host in ia64-*-hpux*) # Find out which ABI we are using. echo 'int i;' > conftest.$ac_ext if AC_TRY_EVAL(ac_compile); then case `/usr/bin/file conftest.$ac_objext` in *ELF-32*) HPUX_IA64_MODE="32" ;; *ELF-64*) HPUX_IA64_MODE="64" ;; esac fi rm -rf conftest* ;; *-*-irix6*) # Find out which ABI we are using. echo '[#]line '$LINENO' "configure"' > conftest.$ac_ext if AC_TRY_EVAL(ac_compile); then if test "$lt_cv_prog_gnu_ld" = yes; then case `/usr/bin/file conftest.$ac_objext` in *32-bit*) LD="${LD-ld} -melf32bsmip" ;; *N32*) LD="${LD-ld} -melf32bmipn32" ;; *64-bit*) LD="${LD-ld} -melf64bmip" ;; esac else case `/usr/bin/file conftest.$ac_objext` in *32-bit*) LD="${LD-ld} -32" ;; *N32*) LD="${LD-ld} -n32" ;; *64-bit*) LD="${LD-ld} -64" ;; esac fi fi rm -rf conftest* ;; x86_64-*kfreebsd*-gnu|x86_64-*linux*|ppc*-*linux*|powerpc*-*linux*| \ s390*-*linux*|s390*-*tpf*|sparc*-*linux*) # Find out which ABI we are using. echo 'int i;' > conftest.$ac_ext if AC_TRY_EVAL(ac_compile); then case `/usr/bin/file conftest.o` in *32-bit*) case $host in x86_64-*kfreebsd*-gnu) LD="${LD-ld} -m elf_i386_fbsd" ;; x86_64-*linux*) LD="${LD-ld} -m elf_i386" ;; ppc64-*linux*|powerpc64-*linux*) LD="${LD-ld} -m elf32ppclinux" ;; s390x-*linux*) LD="${LD-ld} -m elf_s390" ;; sparc64-*linux*) LD="${LD-ld} -m elf32_sparc" ;; esac ;; *64-bit*) case $host in x86_64-*kfreebsd*-gnu) LD="${LD-ld} -m elf_x86_64_fbsd" ;; x86_64-*linux*) LD="${LD-ld} -m elf_x86_64" ;; ppc*-*linux*|powerpc*-*linux*) LD="${LD-ld} -m elf64ppc" ;; s390*-*linux*|s390*-*tpf*) LD="${LD-ld} -m elf64_s390" ;; sparc*-*linux*) LD="${LD-ld} -m elf64_sparc" ;; esac ;; esac fi rm -rf conftest* ;; *-*-sco3.2v5*) # On SCO OpenServer 5, we need -belf to get full-featured binaries. SAVE_CFLAGS="$CFLAGS" CFLAGS="$CFLAGS -belf" AC_CACHE_CHECK([whether the C compiler needs -belf], lt_cv_cc_needs_belf, [AC_LANG_PUSH(C) AC_LINK_IFELSE([AC_LANG_PROGRAM([[]],[[]])],[lt_cv_cc_needs_belf=yes],[lt_cv_cc_needs_belf=no]) AC_LANG_POP]) if test x"$lt_cv_cc_needs_belf" != x"yes"; then # this is probably gcc 2.8.0, egcs 1.0 or newer; no need for -belf CFLAGS="$SAVE_CFLAGS" fi ;; *-*solaris*) # Find out which ABI we are using. echo 'int i;' > conftest.$ac_ext if AC_TRY_EVAL(ac_compile); then case `/usr/bin/file conftest.o` in *64-bit*) case $lt_cv_prog_gnu_ld in yes*) case $host in i?86-*-solaris*) LD="${LD-ld} -m elf_x86_64" ;; sparc*-*-solaris*) LD="${LD-ld} -m elf64_sparc" ;; esac # GNU ld 2.21 introduced _sol2 emulations. Use them if available. if ${LD-ld} -V | grep _sol2 >/dev/null 2>&1; then LD="${LD-ld}_sol2" fi ;; *) if ${LD-ld} -64 -r -o conftest2.o conftest.o >/dev/null 2>&1; then LD="${LD-ld} -64" fi ;; esac ;; esac fi rm -rf conftest* ;; esac need_locks="$enable_libtool_lock" ])# _LT_ENABLE_LOCK # _LT_PROG_AR # ----------- m4_defun([_LT_PROG_AR], [AC_CHECK_TOOLS(AR, [ar], false) : ${AR=ar} : ${AR_FLAGS=cru} _LT_DECL([], [AR], [1], [The archiver]) _LT_DECL([], [AR_FLAGS], [1], [Flags to create an archive]) AC_CACHE_CHECK([for archiver @FILE support], [lt_cv_ar_at_file], [lt_cv_ar_at_file=no AC_COMPILE_IFELSE([AC_LANG_PROGRAM], [echo conftest.$ac_objext > conftest.lst lt_ar_try='$AR $AR_FLAGS libconftest.a @conftest.lst >&AS_MESSAGE_LOG_FD' AC_TRY_EVAL([lt_ar_try]) if test "$ac_status" -eq 0; then # Ensure the archiver fails upon bogus file names. rm -f conftest.$ac_objext libconftest.a AC_TRY_EVAL([lt_ar_try]) if test "$ac_status" -ne 0; then lt_cv_ar_at_file=@ fi fi rm -f conftest.* libconftest.a ]) ]) if test "x$lt_cv_ar_at_file" = xno; then archiver_list_spec= else archiver_list_spec=$lt_cv_ar_at_file fi _LT_DECL([], [archiver_list_spec], [1], [How to feed a file listing to the archiver]) ])# _LT_PROG_AR # _LT_CMD_OLD_ARCHIVE # ------------------- m4_defun([_LT_CMD_OLD_ARCHIVE], [_LT_PROG_AR AC_CHECK_TOOL(STRIP, strip, :) test -z "$STRIP" && STRIP=: _LT_DECL([], [STRIP], [1], [A symbol stripping program]) AC_CHECK_TOOL(RANLIB, ranlib, :) test -z "$RANLIB" && RANLIB=: _LT_DECL([], [RANLIB], [1], [Commands used to install an old-style archive]) # Determine commands to create old-style static archives. old_archive_cmds='$AR $AR_FLAGS $oldlib$oldobjs' old_postinstall_cmds='chmod 644 $oldlib' old_postuninstall_cmds= if test -n "$RANLIB"; then case $host_os in openbsd*) old_postinstall_cmds="$old_postinstall_cmds~\$RANLIB -t \$tool_oldlib" ;; *) old_postinstall_cmds="$old_postinstall_cmds~\$RANLIB \$tool_oldlib" ;; esac old_archive_cmds="$old_archive_cmds~\$RANLIB \$tool_oldlib" fi case $host_os in darwin*) lock_old_archive_extraction=yes ;; *) lock_old_archive_extraction=no ;; esac _LT_DECL([], [old_postinstall_cmds], [2]) _LT_DECL([], [old_postuninstall_cmds], [2]) _LT_TAGDECL([], [old_archive_cmds], [2], [Commands used to build an old-style archive]) _LT_DECL([], [lock_old_archive_extraction], [0], [Whether to use a lock for old archive extraction]) ])# _LT_CMD_OLD_ARCHIVE # _LT_COMPILER_OPTION(MESSAGE, VARIABLE-NAME, FLAGS, # [OUTPUT-FILE], [ACTION-SUCCESS], [ACTION-FAILURE]) # ---------------------------------------------------------------- # Check whether the given compiler option works AC_DEFUN([_LT_COMPILER_OPTION], [m4_require([_LT_FILEUTILS_DEFAULTS])dnl m4_require([_LT_DECL_SED])dnl AC_CACHE_CHECK([$1], [$2], [$2=no m4_if([$4], , [ac_outfile=conftest.$ac_objext], [ac_outfile=$4]) echo "$lt_simple_compile_test_code" > conftest.$ac_ext lt_compiler_flag="$3" # Insert the option either (1) after the last *FLAGS variable, or # (2) before a word containing "conftest.", or (3) at the end. # Note that $ac_compile itself does not contain backslashes and begins # with a dollar sign (not a hyphen), so the echo should work correctly. # The option is referenced via a variable to avoid confusing sed. lt_compile=`echo "$ac_compile" | $SED \ -e 's:.*FLAGS}\{0,1\} :&$lt_compiler_flag :; t' \ -e 's: [[^ ]]*conftest\.: $lt_compiler_flag&:; t' \ -e 's:$: $lt_compiler_flag:'` (eval echo "\"\$as_me:$LINENO: $lt_compile\"" >&AS_MESSAGE_LOG_FD) (eval "$lt_compile" 2>conftest.err) ac_status=$? cat conftest.err >&AS_MESSAGE_LOG_FD echo "$as_me:$LINENO: \$? = $ac_status" >&AS_MESSAGE_LOG_FD if (exit $ac_status) && test -s "$ac_outfile"; then # The compiler can only warn and ignore the option if not recognized # So say no if there are warnings other than the usual output. $ECHO "$_lt_compiler_boilerplate" | $SED '/^$/d' >conftest.exp $SED '/^$/d; /^ *+/d' conftest.err >conftest.er2 if test ! -s conftest.er2 || diff conftest.exp conftest.er2 >/dev/null; then $2=yes fi fi $RM conftest* ]) if test x"[$]$2" = xyes; then m4_if([$5], , :, [$5]) else m4_if([$6], , :, [$6]) fi ])# _LT_COMPILER_OPTION # Old name: AU_ALIAS([AC_LIBTOOL_COMPILER_OPTION], [_LT_COMPILER_OPTION]) dnl aclocal-1.4 backwards compatibility: dnl AC_DEFUN([AC_LIBTOOL_COMPILER_OPTION], []) # _LT_LINKER_OPTION(MESSAGE, VARIABLE-NAME, FLAGS, # [ACTION-SUCCESS], [ACTION-FAILURE]) # ---------------------------------------------------- # Check whether the given linker option works AC_DEFUN([_LT_LINKER_OPTION], [m4_require([_LT_FILEUTILS_DEFAULTS])dnl m4_require([_LT_DECL_SED])dnl AC_CACHE_CHECK([$1], [$2], [$2=no save_LDFLAGS="$LDFLAGS" LDFLAGS="$LDFLAGS $3" echo "$lt_simple_link_test_code" > conftest.$ac_ext if (eval $ac_link 2>conftest.err) && test -s conftest$ac_exeext; then # The linker can only warn and ignore the option if not recognized # So say no if there are warnings if test -s conftest.err; then # Append any errors to the config.log. cat conftest.err 1>&AS_MESSAGE_LOG_FD $ECHO "$_lt_linker_boilerplate" | $SED '/^$/d' > conftest.exp $SED '/^$/d; /^ *+/d' conftest.err >conftest.er2 if diff conftest.exp conftest.er2 >/dev/null; then $2=yes fi else $2=yes fi fi $RM -r conftest* LDFLAGS="$save_LDFLAGS" ]) if test x"[$]$2" = xyes; then m4_if([$4], , :, [$4]) else m4_if([$5], , :, [$5]) fi ])# _LT_LINKER_OPTION # Old name: AU_ALIAS([AC_LIBTOOL_LINKER_OPTION], [_LT_LINKER_OPTION]) dnl aclocal-1.4 backwards compatibility: dnl AC_DEFUN([AC_LIBTOOL_LINKER_OPTION], []) # LT_CMD_MAX_LEN #--------------- AC_DEFUN([LT_CMD_MAX_LEN], [AC_REQUIRE([AC_CANONICAL_HOST])dnl # find the maximum length of command line arguments AC_MSG_CHECKING([the maximum length of command line arguments]) AC_CACHE_VAL([lt_cv_sys_max_cmd_len], [dnl i=0 teststring="ABCD" case $build_os in msdosdjgpp*) # On DJGPP, this test can blow up pretty badly due to problems in libc # (any single argument exceeding 2000 bytes causes a buffer overrun # during glob expansion). Even if it were fixed, the result of this # check would be larger than it should be. lt_cv_sys_max_cmd_len=12288; # 12K is about right ;; gnu*) # Under GNU Hurd, this test is not required because there is # no limit to the length of command line arguments. # Libtool will interpret -1 as no limit whatsoever lt_cv_sys_max_cmd_len=-1; ;; cygwin* | mingw* | cegcc*) # On Win9x/ME, this test blows up -- it succeeds, but takes # about 5 minutes as the teststring grows exponentially. # Worse, since 9x/ME are not pre-emptively multitasking, # you end up with a "frozen" computer, even though with patience # the test eventually succeeds (with a max line length of 256k). # Instead, let's just punt: use the minimum linelength reported by # all of the supported platforms: 8192 (on NT/2K/XP). lt_cv_sys_max_cmd_len=8192; ;; mint*) # On MiNT this can take a long time and run out of memory. lt_cv_sys_max_cmd_len=8192; ;; amigaos*) # On AmigaOS with pdksh, this test takes hours, literally. # So we just punt and use a minimum line length of 8192. lt_cv_sys_max_cmd_len=8192; ;; netbsd* | freebsd* | openbsd* | darwin* | dragonfly*) # This has been around since 386BSD, at least. Likely further. if test -x /sbin/sysctl; then lt_cv_sys_max_cmd_len=`/sbin/sysctl -n kern.argmax` elif test -x /usr/sbin/sysctl; then lt_cv_sys_max_cmd_len=`/usr/sbin/sysctl -n kern.argmax` else lt_cv_sys_max_cmd_len=65536 # usable default for all BSDs fi # And add a safety zone lt_cv_sys_max_cmd_len=`expr $lt_cv_sys_max_cmd_len \/ 4` lt_cv_sys_max_cmd_len=`expr $lt_cv_sys_max_cmd_len \* 3` ;; interix*) # We know the value 262144 and hardcode it with a safety zone (like BSD) lt_cv_sys_max_cmd_len=196608 ;; os2*) # The test takes a long time on OS/2. lt_cv_sys_max_cmd_len=8192 ;; osf*) # Dr. Hans Ekkehard Plesser reports seeing a kernel panic running configure # due to this test when exec_disable_arg_limit is 1 on Tru64. It is not # nice to cause kernel panics so lets avoid the loop below. # First set a reasonable default. lt_cv_sys_max_cmd_len=16384 # if test -x /sbin/sysconfig; then case `/sbin/sysconfig -q proc exec_disable_arg_limit` in *1*) lt_cv_sys_max_cmd_len=-1 ;; esac fi ;; sco3.2v5*) lt_cv_sys_max_cmd_len=102400 ;; sysv5* | sco5v6* | sysv4.2uw2*) kargmax=`grep ARG_MAX /etc/conf/cf.d/stune 2>/dev/null` if test -n "$kargmax"; then lt_cv_sys_max_cmd_len=`echo $kargmax | sed 's/.*[[ ]]//'` else lt_cv_sys_max_cmd_len=32768 fi ;; *) lt_cv_sys_max_cmd_len=`(getconf ARG_MAX) 2> /dev/null` if test -n "$lt_cv_sys_max_cmd_len"; then lt_cv_sys_max_cmd_len=`expr $lt_cv_sys_max_cmd_len \/ 4` lt_cv_sys_max_cmd_len=`expr $lt_cv_sys_max_cmd_len \* 3` else # Make teststring a little bigger before we do anything with it. # a 1K string should be a reasonable start. for i in 1 2 3 4 5 6 7 8 ; do teststring=$teststring$teststring done SHELL=${SHELL-${CONFIG_SHELL-/bin/sh}} # If test is not a shell built-in, we'll probably end up computing a # maximum length that is only half of the actual maximum length, but # we can't tell. while { test "X"`env echo "$teststring$teststring" 2>/dev/null` \ = "X$teststring$teststring"; } >/dev/null 2>&1 && test $i != 17 # 1/2 MB should be enough do i=`expr $i + 1` teststring=$teststring$teststring done # Only check the string length outside the loop. lt_cv_sys_max_cmd_len=`expr "X$teststring" : ".*" 2>&1` teststring= # Add a significant safety factor because C++ compilers can tack on # massive amounts of additional arguments before passing them to the # linker. It appears as though 1/2 is a usable value. lt_cv_sys_max_cmd_len=`expr $lt_cv_sys_max_cmd_len \/ 2` fi ;; esac ]) if test -n $lt_cv_sys_max_cmd_len ; then AC_MSG_RESULT($lt_cv_sys_max_cmd_len) else AC_MSG_RESULT(none) fi max_cmd_len=$lt_cv_sys_max_cmd_len _LT_DECL([], [max_cmd_len], [0], [What is the maximum length of a command?]) ])# LT_CMD_MAX_LEN # Old name: AU_ALIAS([AC_LIBTOOL_SYS_MAX_CMD_LEN], [LT_CMD_MAX_LEN]) dnl aclocal-1.4 backwards compatibility: dnl AC_DEFUN([AC_LIBTOOL_SYS_MAX_CMD_LEN], []) # _LT_HEADER_DLFCN # ---------------- m4_defun([_LT_HEADER_DLFCN], [AC_CHECK_HEADERS([dlfcn.h], [], [], [AC_INCLUDES_DEFAULT])dnl ])# _LT_HEADER_DLFCN # _LT_TRY_DLOPEN_SELF (ACTION-IF-TRUE, ACTION-IF-TRUE-W-USCORE, # ACTION-IF-FALSE, ACTION-IF-CROSS-COMPILING) # ---------------------------------------------------------------- m4_defun([_LT_TRY_DLOPEN_SELF], [m4_require([_LT_HEADER_DLFCN])dnl if test "$cross_compiling" = yes; then : [$4] else lt_dlunknown=0; lt_dlno_uscore=1; lt_dlneed_uscore=2 lt_status=$lt_dlunknown cat > conftest.$ac_ext <<_LT_EOF [#line $LINENO "configure" #include "confdefs.h" #if HAVE_DLFCN_H #include #endif #include #ifdef RTLD_GLOBAL # define LT_DLGLOBAL RTLD_GLOBAL #else # ifdef DL_GLOBAL # define LT_DLGLOBAL DL_GLOBAL # else # define LT_DLGLOBAL 0 # endif #endif /* We may have to define LT_DLLAZY_OR_NOW in the command line if we find out it does not work in some platform. */ #ifndef LT_DLLAZY_OR_NOW # ifdef RTLD_LAZY # define LT_DLLAZY_OR_NOW RTLD_LAZY # else # ifdef DL_LAZY # define LT_DLLAZY_OR_NOW DL_LAZY # else # ifdef RTLD_NOW # define LT_DLLAZY_OR_NOW RTLD_NOW # else # ifdef DL_NOW # define LT_DLLAZY_OR_NOW DL_NOW # else # define LT_DLLAZY_OR_NOW 0 # endif # endif # endif # endif #endif /* When -fvisbility=hidden is used, assume the code has been annotated correspondingly for the symbols needed. */ #if defined(__GNUC__) && (((__GNUC__ == 3) && (__GNUC_MINOR__ >= 3)) || (__GNUC__ > 3)) int fnord () __attribute__((visibility("default"))); #endif int fnord () { return 42; } int main () { void *self = dlopen (0, LT_DLGLOBAL|LT_DLLAZY_OR_NOW); int status = $lt_dlunknown; if (self) { if (dlsym (self,"fnord")) status = $lt_dlno_uscore; else { if (dlsym( self,"_fnord")) status = $lt_dlneed_uscore; else puts (dlerror ()); } /* dlclose (self); */ } else puts (dlerror ()); return status; }] _LT_EOF if AC_TRY_EVAL(ac_link) && test -s conftest${ac_exeext} 2>/dev/null; then (./conftest; exit; ) >&AS_MESSAGE_LOG_FD 2>/dev/null lt_status=$? case x$lt_status in x$lt_dlno_uscore) $1 ;; x$lt_dlneed_uscore) $2 ;; x$lt_dlunknown|x*) $3 ;; esac else : # compilation failed $3 fi fi rm -fr conftest* ])# _LT_TRY_DLOPEN_SELF # LT_SYS_DLOPEN_SELF # ------------------ AC_DEFUN([LT_SYS_DLOPEN_SELF], [m4_require([_LT_HEADER_DLFCN])dnl if test "x$enable_dlopen" != xyes; then enable_dlopen=unknown enable_dlopen_self=unknown enable_dlopen_self_static=unknown else lt_cv_dlopen=no lt_cv_dlopen_libs= case $host_os in beos*) lt_cv_dlopen="load_add_on" lt_cv_dlopen_libs= lt_cv_dlopen_self=yes ;; mingw* | pw32* | cegcc*) lt_cv_dlopen="LoadLibrary" lt_cv_dlopen_libs= ;; cygwin*) lt_cv_dlopen="dlopen" lt_cv_dlopen_libs= ;; darwin*) # if libdl is installed we need to link against it AC_CHECK_LIB([dl], [dlopen], [lt_cv_dlopen="dlopen" lt_cv_dlopen_libs="-ldl"],[ lt_cv_dlopen="dyld" lt_cv_dlopen_libs= lt_cv_dlopen_self=yes ]) ;; *) AC_CHECK_FUNC([shl_load], [lt_cv_dlopen="shl_load"], [AC_CHECK_LIB([dld], [shl_load], [lt_cv_dlopen="shl_load" lt_cv_dlopen_libs="-ldld"], [AC_CHECK_FUNC([dlopen], [lt_cv_dlopen="dlopen"], [AC_CHECK_LIB([dl], [dlopen], [lt_cv_dlopen="dlopen" lt_cv_dlopen_libs="-ldl"], [AC_CHECK_LIB([svld], [dlopen], [lt_cv_dlopen="dlopen" lt_cv_dlopen_libs="-lsvld"], [AC_CHECK_LIB([dld], [dld_link], [lt_cv_dlopen="dld_link" lt_cv_dlopen_libs="-ldld"]) ]) ]) ]) ]) ]) ;; esac if test "x$lt_cv_dlopen" != xno; then enable_dlopen=yes else enable_dlopen=no fi case $lt_cv_dlopen in dlopen) save_CPPFLAGS="$CPPFLAGS" test "x$ac_cv_header_dlfcn_h" = xyes && CPPFLAGS="$CPPFLAGS -DHAVE_DLFCN_H" save_LDFLAGS="$LDFLAGS" wl=$lt_prog_compiler_wl eval LDFLAGS=\"\$LDFLAGS $export_dynamic_flag_spec\" save_LIBS="$LIBS" LIBS="$lt_cv_dlopen_libs $LIBS" AC_CACHE_CHECK([whether a program can dlopen itself], lt_cv_dlopen_self, [dnl _LT_TRY_DLOPEN_SELF( lt_cv_dlopen_self=yes, lt_cv_dlopen_self=yes, lt_cv_dlopen_self=no, lt_cv_dlopen_self=cross) ]) if test "x$lt_cv_dlopen_self" = xyes; then wl=$lt_prog_compiler_wl eval LDFLAGS=\"\$LDFLAGS $lt_prog_compiler_static\" AC_CACHE_CHECK([whether a statically linked program can dlopen itself], lt_cv_dlopen_self_static, [dnl _LT_TRY_DLOPEN_SELF( lt_cv_dlopen_self_static=yes, lt_cv_dlopen_self_static=yes, lt_cv_dlopen_self_static=no, lt_cv_dlopen_self_static=cross) ]) fi CPPFLAGS="$save_CPPFLAGS" LDFLAGS="$save_LDFLAGS" LIBS="$save_LIBS" ;; esac case $lt_cv_dlopen_self in yes|no) enable_dlopen_self=$lt_cv_dlopen_self ;; *) enable_dlopen_self=unknown ;; esac case $lt_cv_dlopen_self_static in yes|no) enable_dlopen_self_static=$lt_cv_dlopen_self_static ;; *) enable_dlopen_self_static=unknown ;; esac fi _LT_DECL([dlopen_support], [enable_dlopen], [0], [Whether dlopen is supported]) _LT_DECL([dlopen_self], [enable_dlopen_self], [0], [Whether dlopen of programs is supported]) _LT_DECL([dlopen_self_static], [enable_dlopen_self_static], [0], [Whether dlopen of statically linked programs is supported]) ])# LT_SYS_DLOPEN_SELF # Old name: AU_ALIAS([AC_LIBTOOL_DLOPEN_SELF], [LT_SYS_DLOPEN_SELF]) dnl aclocal-1.4 backwards compatibility: dnl AC_DEFUN([AC_LIBTOOL_DLOPEN_SELF], []) # _LT_COMPILER_C_O([TAGNAME]) # --------------------------- # Check to see if options -c and -o are simultaneously supported by compiler. # This macro does not hard code the compiler like AC_PROG_CC_C_O. m4_defun([_LT_COMPILER_C_O], [m4_require([_LT_DECL_SED])dnl m4_require([_LT_FILEUTILS_DEFAULTS])dnl m4_require([_LT_TAG_COMPILER])dnl AC_CACHE_CHECK([if $compiler supports -c -o file.$ac_objext], [_LT_TAGVAR(lt_cv_prog_compiler_c_o, $1)], [_LT_TAGVAR(lt_cv_prog_compiler_c_o, $1)=no $RM -r conftest 2>/dev/null mkdir conftest cd conftest mkdir out echo "$lt_simple_compile_test_code" > conftest.$ac_ext lt_compiler_flag="-o out/conftest2.$ac_objext" # Insert the option either (1) after the last *FLAGS variable, or # (2) before a word containing "conftest.", or (3) at the end. # Note that $ac_compile itself does not contain backslashes and begins # with a dollar sign (not a hyphen), so the echo should work correctly. lt_compile=`echo "$ac_compile" | $SED \ -e 's:.*FLAGS}\{0,1\} :&$lt_compiler_flag :; t' \ -e 's: [[^ ]]*conftest\.: $lt_compiler_flag&:; t' \ -e 's:$: $lt_compiler_flag:'` (eval echo "\"\$as_me:$LINENO: $lt_compile\"" >&AS_MESSAGE_LOG_FD) (eval "$lt_compile" 2>out/conftest.err) ac_status=$? cat out/conftest.err >&AS_MESSAGE_LOG_FD echo "$as_me:$LINENO: \$? = $ac_status" >&AS_MESSAGE_LOG_FD if (exit $ac_status) && test -s out/conftest2.$ac_objext then # The compiler can only warn and ignore the option if not recognized # So say no if there are warnings $ECHO "$_lt_compiler_boilerplate" | $SED '/^$/d' > out/conftest.exp $SED '/^$/d; /^ *+/d' out/conftest.err >out/conftest.er2 if test ! -s out/conftest.er2 || diff out/conftest.exp out/conftest.er2 >/dev/null; then _LT_TAGVAR(lt_cv_prog_compiler_c_o, $1)=yes fi fi chmod u+w . 2>&AS_MESSAGE_LOG_FD $RM conftest* # SGI C++ compiler will create directory out/ii_files/ for # template instantiation test -d out/ii_files && $RM out/ii_files/* && rmdir out/ii_files $RM out/* && rmdir out cd .. $RM -r conftest $RM conftest* ]) _LT_TAGDECL([compiler_c_o], [lt_cv_prog_compiler_c_o], [1], [Does compiler simultaneously support -c and -o options?]) ])# _LT_COMPILER_C_O # _LT_COMPILER_FILE_LOCKS([TAGNAME]) # ---------------------------------- # Check to see if we can do hard links to lock some files if needed m4_defun([_LT_COMPILER_FILE_LOCKS], [m4_require([_LT_ENABLE_LOCK])dnl m4_require([_LT_FILEUTILS_DEFAULTS])dnl _LT_COMPILER_C_O([$1]) hard_links="nottested" if test "$_LT_TAGVAR(lt_cv_prog_compiler_c_o, $1)" = no && test "$need_locks" != no; then # do not overwrite the value of need_locks provided by the user AC_MSG_CHECKING([if we can lock with hard links]) hard_links=yes $RM conftest* ln conftest.a conftest.b 2>/dev/null && hard_links=no touch conftest.a ln conftest.a conftest.b 2>&5 || hard_links=no ln conftest.a conftest.b 2>/dev/null && hard_links=no AC_MSG_RESULT([$hard_links]) if test "$hard_links" = no; then AC_MSG_WARN([`$CC' does not support `-c -o', so `make -j' may be unsafe]) need_locks=warn fi else need_locks=no fi _LT_DECL([], [need_locks], [1], [Must we lock files when doing compilation?]) ])# _LT_COMPILER_FILE_LOCKS # _LT_CHECK_OBJDIR # ---------------- m4_defun([_LT_CHECK_OBJDIR], [AC_CACHE_CHECK([for objdir], [lt_cv_objdir], [rm -f .libs 2>/dev/null mkdir .libs 2>/dev/null if test -d .libs; then lt_cv_objdir=.libs else # MS-DOS does not allow filenames that begin with a dot. lt_cv_objdir=_libs fi rmdir .libs 2>/dev/null]) objdir=$lt_cv_objdir _LT_DECL([], [objdir], [0], [The name of the directory that contains temporary libtool files])dnl m4_pattern_allow([LT_OBJDIR])dnl AC_DEFINE_UNQUOTED(LT_OBJDIR, "$lt_cv_objdir/", [Define to the sub-directory in which libtool stores uninstalled libraries.]) ])# _LT_CHECK_OBJDIR # _LT_LINKER_HARDCODE_LIBPATH([TAGNAME]) # -------------------------------------- # Check hardcoding attributes. m4_defun([_LT_LINKER_HARDCODE_LIBPATH], [AC_MSG_CHECKING([how to hardcode library paths into programs]) _LT_TAGVAR(hardcode_action, $1)= if test -n "$_LT_TAGVAR(hardcode_libdir_flag_spec, $1)" || test -n "$_LT_TAGVAR(runpath_var, $1)" || test "X$_LT_TAGVAR(hardcode_automatic, $1)" = "Xyes" ; then # We can hardcode non-existent directories. if test "$_LT_TAGVAR(hardcode_direct, $1)" != no && # If the only mechanism to avoid hardcoding is shlibpath_var, we # have to relink, otherwise we might link with an installed library # when we should be linking with a yet-to-be-installed one ## test "$_LT_TAGVAR(hardcode_shlibpath_var, $1)" != no && test "$_LT_TAGVAR(hardcode_minus_L, $1)" != no; then # Linking always hardcodes the temporary library directory. _LT_TAGVAR(hardcode_action, $1)=relink else # We can link without hardcoding, and we can hardcode nonexisting dirs. _LT_TAGVAR(hardcode_action, $1)=immediate fi else # We cannot hardcode anything, or else we can only hardcode existing # directories. _LT_TAGVAR(hardcode_action, $1)=unsupported fi AC_MSG_RESULT([$_LT_TAGVAR(hardcode_action, $1)]) if test "$_LT_TAGVAR(hardcode_action, $1)" = relink || test "$_LT_TAGVAR(inherit_rpath, $1)" = yes; then # Fast installation is not supported enable_fast_install=no elif test "$shlibpath_overrides_runpath" = yes || test "$enable_shared" = no; then # Fast installation is not necessary enable_fast_install=needless fi _LT_TAGDECL([], [hardcode_action], [0], [How to hardcode a shared library path into an executable]) ])# _LT_LINKER_HARDCODE_LIBPATH # _LT_CMD_STRIPLIB # ---------------- m4_defun([_LT_CMD_STRIPLIB], [m4_require([_LT_DECL_EGREP]) striplib= old_striplib= AC_MSG_CHECKING([whether stripping libraries is possible]) if test -n "$STRIP" && $STRIP -V 2>&1 | $GREP "GNU strip" >/dev/null; then test -z "$old_striplib" && old_striplib="$STRIP --strip-debug" test -z "$striplib" && striplib="$STRIP --strip-unneeded" AC_MSG_RESULT([yes]) else # FIXME - insert some real tests, host_os isn't really good enough case $host_os in darwin*) if test -n "$STRIP" ; then striplib="$STRIP -x" old_striplib="$STRIP -S" AC_MSG_RESULT([yes]) else AC_MSG_RESULT([no]) fi ;; *) AC_MSG_RESULT([no]) ;; esac fi _LT_DECL([], [old_striplib], [1], [Commands to strip libraries]) _LT_DECL([], [striplib], [1]) ])# _LT_CMD_STRIPLIB # _LT_SYS_DYNAMIC_LINKER([TAG]) # ----------------------------- # PORTME Fill in your ld.so characteristics m4_defun([_LT_SYS_DYNAMIC_LINKER], [AC_REQUIRE([AC_CANONICAL_HOST])dnl m4_require([_LT_DECL_EGREP])dnl m4_require([_LT_FILEUTILS_DEFAULTS])dnl m4_require([_LT_DECL_OBJDUMP])dnl m4_require([_LT_DECL_SED])dnl m4_require([_LT_CHECK_SHELL_FEATURES])dnl AC_MSG_CHECKING([dynamic linker characteristics]) m4_if([$1], [], [ if test "$GCC" = yes; then case $host_os in darwin*) lt_awk_arg="/^libraries:/,/LR/" ;; *) lt_awk_arg="/^libraries:/" ;; esac case $host_os in mingw* | cegcc*) lt_sed_strip_eq="s,=\([[A-Za-z]]:\),\1,g" ;; *) lt_sed_strip_eq="s,=/,/,g" ;; esac lt_search_path_spec=`$CC -print-search-dirs | awk $lt_awk_arg | $SED -e "s/^libraries://" -e $lt_sed_strip_eq` case $lt_search_path_spec in *\;*) # if the path contains ";" then we assume it to be the separator # otherwise default to the standard path separator (i.e. ":") - it is # assumed that no part of a normal pathname contains ";" but that should # okay in the real world where ";" in dirpaths is itself problematic. lt_search_path_spec=`$ECHO "$lt_search_path_spec" | $SED 's/;/ /g'` ;; *) lt_search_path_spec=`$ECHO "$lt_search_path_spec" | $SED "s/$PATH_SEPARATOR/ /g"` ;; esac # Ok, now we have the path, separated by spaces, we can step through it # and add multilib dir if necessary. lt_tmp_lt_search_path_spec= lt_multi_os_dir=`$CC $CPPFLAGS $CFLAGS $LDFLAGS -print-multi-os-directory 2>/dev/null` for lt_sys_path in $lt_search_path_spec; do if test -d "$lt_sys_path/$lt_multi_os_dir"; then lt_tmp_lt_search_path_spec="$lt_tmp_lt_search_path_spec $lt_sys_path/$lt_multi_os_dir" else test -d "$lt_sys_path" && \ lt_tmp_lt_search_path_spec="$lt_tmp_lt_search_path_spec $lt_sys_path" fi done lt_search_path_spec=`$ECHO "$lt_tmp_lt_search_path_spec" | awk ' BEGIN {RS=" "; FS="/|\n";} { lt_foo=""; lt_count=0; for (lt_i = NF; lt_i > 0; lt_i--) { if ($lt_i != "" && $lt_i != ".") { if ($lt_i == "..") { lt_count++; } else { if (lt_count == 0) { lt_foo="/" $lt_i lt_foo; } else { lt_count--; } } } } if (lt_foo != "") { lt_freq[[lt_foo]]++; } if (lt_freq[[lt_foo]] == 1) { print lt_foo; } }'` # AWK program above erroneously prepends '/' to C:/dos/paths # for these hosts. case $host_os in mingw* | cegcc*) lt_search_path_spec=`$ECHO "$lt_search_path_spec" |\ $SED 's,/\([[A-Za-z]]:\),\1,g'` ;; esac sys_lib_search_path_spec=`$ECHO "$lt_search_path_spec" | $lt_NL2SP` else sys_lib_search_path_spec="/lib /usr/lib /usr/local/lib" fi]) library_names_spec= libname_spec='lib$name' soname_spec= shrext_cmds=".so" postinstall_cmds= postuninstall_cmds= finish_cmds= finish_eval= shlibpath_var= shlibpath_overrides_runpath=unknown version_type=none dynamic_linker="$host_os ld.so" sys_lib_dlsearch_path_spec="/lib /usr/lib" need_lib_prefix=unknown hardcode_into_libs=no # when you set need_version to no, make sure it does not cause -set_version # flags to be left without arguments need_version=unknown case $host_os in aix3*) version_type=linux # correct to gnu/linux during the next big refactor library_names_spec='${libname}${release}${shared_ext}$versuffix $libname.a' shlibpath_var=LIBPATH # AIX 3 has no versioning support, so we append a major version to the name. soname_spec='${libname}${release}${shared_ext}$major' ;; aix[[4-9]]*) version_type=linux # correct to gnu/linux during the next big refactor need_lib_prefix=no need_version=no hardcode_into_libs=yes if test "$host_cpu" = ia64; then # AIX 5 supports IA64 library_names_spec='${libname}${release}${shared_ext}$major ${libname}${release}${shared_ext}$versuffix $libname${shared_ext}' shlibpath_var=LD_LIBRARY_PATH else # With GCC up to 2.95.x, collect2 would create an import file # for dependence libraries. The import file would start with # the line `#! .'. This would cause the generated library to # depend on `.', always an invalid library. This was fixed in # development snapshots of GCC prior to 3.0. case $host_os in aix4 | aix4.[[01]] | aix4.[[01]].*) if { echo '#if __GNUC__ > 2 || (__GNUC__ == 2 && __GNUC_MINOR__ >= 97)' echo ' yes ' echo '#endif'; } | ${CC} -E - | $GREP yes > /dev/null; then : else can_build_shared=no fi ;; esac # AIX (on Power*) has no versioning support, so currently we can not hardcode correct # soname into executable. Probably we can add versioning support to # collect2, so additional links can be useful in future. if test "$aix_use_runtimelinking" = yes; then # If using run time linking (on AIX 4.2 or later) use lib.so # instead of lib.a to let people know that these are not # typical AIX shared libraries. library_names_spec='${libname}${release}${shared_ext}$versuffix ${libname}${release}${shared_ext}$major $libname${shared_ext}' else # We preserve .a as extension for shared libraries through AIX4.2 # and later when we are not doing run time linking. library_names_spec='${libname}${release}.a $libname.a' soname_spec='${libname}${release}${shared_ext}$major' fi shlibpath_var=LIBPATH fi ;; amigaos*) case $host_cpu in powerpc) # Since July 2007 AmigaOS4 officially supports .so libraries. # When compiling the executable, add -use-dynld -Lsobjs: to the compileline. library_names_spec='${libname}${release}${shared_ext}$versuffix ${libname}${release}${shared_ext}$major $libname${shared_ext}' ;; m68k) library_names_spec='$libname.ixlibrary $libname.a' # Create ${libname}_ixlibrary.a entries in /sys/libs. finish_eval='for lib in `ls $libdir/*.ixlibrary 2>/dev/null`; do libname=`func_echo_all "$lib" | $SED '\''s%^.*/\([[^/]]*\)\.ixlibrary$%\1%'\''`; test $RM /sys/libs/${libname}_ixlibrary.a; $show "cd /sys/libs && $LN_S $lib ${libname}_ixlibrary.a"; cd /sys/libs && $LN_S $lib ${libname}_ixlibrary.a || exit 1; done' ;; esac ;; beos*) library_names_spec='${libname}${shared_ext}' dynamic_linker="$host_os ld.so" shlibpath_var=LIBRARY_PATH ;; bsdi[[45]]*) version_type=linux # correct to gnu/linux during the next big refactor need_version=no library_names_spec='${libname}${release}${shared_ext}$versuffix ${libname}${release}${shared_ext}$major $libname${shared_ext}' soname_spec='${libname}${release}${shared_ext}$major' finish_cmds='PATH="\$PATH:/sbin" ldconfig $libdir' shlibpath_var=LD_LIBRARY_PATH sys_lib_search_path_spec="/shlib /usr/lib /usr/X11/lib /usr/contrib/lib /lib /usr/local/lib" sys_lib_dlsearch_path_spec="/shlib /usr/lib /usr/local/lib" # the default ld.so.conf also contains /usr/contrib/lib and # /usr/X11R6/lib (/usr/X11 is a link to /usr/X11R6), but let us allow # libtool to hard-code these into programs ;; cygwin* | mingw* | pw32* | cegcc*) version_type=windows shrext_cmds=".dll" need_version=no need_lib_prefix=no case $GCC,$cc_basename in yes,*) # gcc library_names_spec='$libname.dll.a' # DLL is installed to $(libdir)/../bin by postinstall_cmds postinstall_cmds='base_file=`basename \${file}`~ dlpath=`$SHELL 2>&1 -c '\''. $dir/'\''\${base_file}'\''i; echo \$dlname'\''`~ dldir=$destdir/`dirname \$dlpath`~ test -d \$dldir || mkdir -p \$dldir~ $install_prog $dir/$dlname \$dldir/$dlname~ chmod a+x \$dldir/$dlname~ if test -n '\''$stripme'\'' && test -n '\''$striplib'\''; then eval '\''$striplib \$dldir/$dlname'\'' || exit \$?; fi' postuninstall_cmds='dldll=`$SHELL 2>&1 -c '\''. $file; echo \$dlname'\''`~ dlpath=$dir/\$dldll~ $RM \$dlpath' shlibpath_overrides_runpath=yes case $host_os in cygwin*) # Cygwin DLLs use 'cyg' prefix rather than 'lib' soname_spec='`echo ${libname} | sed -e 's/^lib/cyg/'``echo ${release} | $SED -e 's/[[.]]/-/g'`${versuffix}${shared_ext}' m4_if([$1], [],[ sys_lib_search_path_spec="$sys_lib_search_path_spec /usr/lib/w32api"]) ;; mingw* | cegcc*) # MinGW DLLs use traditional 'lib' prefix soname_spec='${libname}`echo ${release} | $SED -e 's/[[.]]/-/g'`${versuffix}${shared_ext}' ;; pw32*) # pw32 DLLs use 'pw' prefix rather than 'lib' library_names_spec='`echo ${libname} | sed -e 's/^lib/pw/'``echo ${release} | $SED -e 's/[[.]]/-/g'`${versuffix}${shared_ext}' ;; esac dynamic_linker='Win32 ld.exe' ;; *,cl*) # Native MSVC libname_spec='$name' soname_spec='${libname}`echo ${release} | $SED -e 's/[[.]]/-/g'`${versuffix}${shared_ext}' library_names_spec='${libname}.dll.lib' case $build_os in mingw*) sys_lib_search_path_spec= lt_save_ifs=$IFS IFS=';' for lt_path in $LIB do IFS=$lt_save_ifs # Let DOS variable expansion print the short 8.3 style file name. lt_path=`cd "$lt_path" 2>/dev/null && cmd //C "for %i in (".") do @echo %~si"` sys_lib_search_path_spec="$sys_lib_search_path_spec $lt_path" done IFS=$lt_save_ifs # Convert to MSYS style. sys_lib_search_path_spec=`$ECHO "$sys_lib_search_path_spec" | sed -e 's|\\\\|/|g' -e 's| \\([[a-zA-Z]]\\):| /\\1|g' -e 's|^ ||'` ;; cygwin*) # Convert to unix form, then to dos form, then back to unix form # but this time dos style (no spaces!) so that the unix form looks # like /cygdrive/c/PROGRA~1:/cygdr... sys_lib_search_path_spec=`cygpath --path --unix "$LIB"` sys_lib_search_path_spec=`cygpath --path --dos "$sys_lib_search_path_spec" 2>/dev/null` sys_lib_search_path_spec=`cygpath --path --unix "$sys_lib_search_path_spec" | $SED -e "s/$PATH_SEPARATOR/ /g"` ;; *) sys_lib_search_path_spec="$LIB" if $ECHO "$sys_lib_search_path_spec" | [$GREP ';[c-zC-Z]:/' >/dev/null]; then # It is most probably a Windows format PATH. sys_lib_search_path_spec=`$ECHO "$sys_lib_search_path_spec" | $SED -e 's/;/ /g'` else sys_lib_search_path_spec=`$ECHO "$sys_lib_search_path_spec" | $SED -e "s/$PATH_SEPARATOR/ /g"` fi # FIXME: find the short name or the path components, as spaces are # common. (e.g. "Program Files" -> "PROGRA~1") ;; esac # DLL is installed to $(libdir)/../bin by postinstall_cmds postinstall_cmds='base_file=`basename \${file}`~ dlpath=`$SHELL 2>&1 -c '\''. $dir/'\''\${base_file}'\''i; echo \$dlname'\''`~ dldir=$destdir/`dirname \$dlpath`~ test -d \$dldir || mkdir -p \$dldir~ $install_prog $dir/$dlname \$dldir/$dlname' postuninstall_cmds='dldll=`$SHELL 2>&1 -c '\''. $file; echo \$dlname'\''`~ dlpath=$dir/\$dldll~ $RM \$dlpath' shlibpath_overrides_runpath=yes dynamic_linker='Win32 link.exe' ;; *) # Assume MSVC wrapper library_names_spec='${libname}`echo ${release} | $SED -e 's/[[.]]/-/g'`${versuffix}${shared_ext} $libname.lib' dynamic_linker='Win32 ld.exe' ;; esac # FIXME: first we should search . and the directory the executable is in shlibpath_var=PATH ;; darwin* | rhapsody*) dynamic_linker="$host_os dyld" version_type=darwin need_lib_prefix=no need_version=no library_names_spec='${libname}${release}${major}$shared_ext ${libname}$shared_ext' soname_spec='${libname}${release}${major}$shared_ext' shlibpath_overrides_runpath=yes shlibpath_var=DYLD_LIBRARY_PATH shrext_cmds='`test .$module = .yes && echo .so || echo .dylib`' m4_if([$1], [],[ sys_lib_search_path_spec="$sys_lib_search_path_spec /usr/local/lib"]) sys_lib_dlsearch_path_spec='/usr/local/lib /lib /usr/lib' ;; dgux*) version_type=linux # correct to gnu/linux during the next big refactor need_lib_prefix=no need_version=no library_names_spec='${libname}${release}${shared_ext}$versuffix ${libname}${release}${shared_ext}$major $libname$shared_ext' soname_spec='${libname}${release}${shared_ext}$major' shlibpath_var=LD_LIBRARY_PATH ;; freebsd* | dragonfly*) # DragonFly does not have aout. When/if they implement a new # versioning mechanism, adjust this. if test -x /usr/bin/objformat; then objformat=`/usr/bin/objformat` else case $host_os in freebsd[[23]].*) objformat=aout ;; *) objformat=elf ;; esac fi version_type=freebsd-$objformat case $version_type in freebsd-elf*) library_names_spec='${libname}${release}${shared_ext}$versuffix ${libname}${release}${shared_ext} $libname${shared_ext}' need_version=no need_lib_prefix=no ;; freebsd-*) library_names_spec='${libname}${release}${shared_ext}$versuffix $libname${shared_ext}$versuffix' need_version=yes ;; esac shlibpath_var=LD_LIBRARY_PATH case $host_os in freebsd2.*) shlibpath_overrides_runpath=yes ;; freebsd3.[[01]]* | freebsdelf3.[[01]]*) shlibpath_overrides_runpath=yes hardcode_into_libs=yes ;; freebsd3.[[2-9]]* | freebsdelf3.[[2-9]]* | \ freebsd4.[[0-5]] | freebsdelf4.[[0-5]] | freebsd4.1.1 | freebsdelf4.1.1) shlibpath_overrides_runpath=no hardcode_into_libs=yes ;; *) # from 4.6 on, and DragonFly shlibpath_overrides_runpath=yes hardcode_into_libs=yes ;; esac ;; haiku*) version_type=linux # correct to gnu/linux during the next big refactor need_lib_prefix=no need_version=no dynamic_linker="$host_os runtime_loader" library_names_spec='${libname}${release}${shared_ext}$versuffix ${libname}${release}${shared_ext}${major} ${libname}${shared_ext}' soname_spec='${libname}${release}${shared_ext}$major' shlibpath_var=LIBRARY_PATH shlibpath_overrides_runpath=yes sys_lib_dlsearch_path_spec='/boot/home/config/lib /boot/common/lib /boot/system/lib' hardcode_into_libs=yes ;; hpux9* | hpux10* | hpux11*) # Give a soname corresponding to the major version so that dld.sl refuses to # link against other versions. version_type=sunos need_lib_prefix=no need_version=no case $host_cpu in ia64*) shrext_cmds='.so' hardcode_into_libs=yes dynamic_linker="$host_os dld.so" shlibpath_var=LD_LIBRARY_PATH shlibpath_overrides_runpath=yes # Unless +noenvvar is specified. library_names_spec='${libname}${release}${shared_ext}$versuffix ${libname}${release}${shared_ext}$major $libname${shared_ext}' soname_spec='${libname}${release}${shared_ext}$major' if test "X$HPUX_IA64_MODE" = X32; then sys_lib_search_path_spec="/usr/lib/hpux32 /usr/local/lib/hpux32 /usr/local/lib" else sys_lib_search_path_spec="/usr/lib/hpux64 /usr/local/lib/hpux64" fi sys_lib_dlsearch_path_spec=$sys_lib_search_path_spec ;; hppa*64*) shrext_cmds='.sl' hardcode_into_libs=yes dynamic_linker="$host_os dld.sl" shlibpath_var=LD_LIBRARY_PATH # How should we handle SHLIB_PATH shlibpath_overrides_runpath=yes # Unless +noenvvar is specified. library_names_spec='${libname}${release}${shared_ext}$versuffix ${libname}${release}${shared_ext}$major $libname${shared_ext}' soname_spec='${libname}${release}${shared_ext}$major' sys_lib_search_path_spec="/usr/lib/pa20_64 /usr/ccs/lib/pa20_64" sys_lib_dlsearch_path_spec=$sys_lib_search_path_spec ;; *) shrext_cmds='.sl' dynamic_linker="$host_os dld.sl" shlibpath_var=SHLIB_PATH shlibpath_overrides_runpath=no # +s is required to enable SHLIB_PATH library_names_spec='${libname}${release}${shared_ext}$versuffix ${libname}${release}${shared_ext}$major $libname${shared_ext}' soname_spec='${libname}${release}${shared_ext}$major' ;; esac # HP-UX runs *really* slowly unless shared libraries are mode 555, ... postinstall_cmds='chmod 555 $lib' # or fails outright, so override atomically: install_override_mode=555 ;; interix[[3-9]]*) version_type=linux # correct to gnu/linux during the next big refactor need_lib_prefix=no need_version=no library_names_spec='${libname}${release}${shared_ext}$versuffix ${libname}${release}${shared_ext}$major ${libname}${shared_ext}' soname_spec='${libname}${release}${shared_ext}$major' dynamic_linker='Interix 3.x ld.so.1 (PE, like ELF)' shlibpath_var=LD_LIBRARY_PATH shlibpath_overrides_runpath=no hardcode_into_libs=yes ;; irix5* | irix6* | nonstopux*) case $host_os in nonstopux*) version_type=nonstopux ;; *) if test "$lt_cv_prog_gnu_ld" = yes; then version_type=linux # correct to gnu/linux during the next big refactor else version_type=irix fi ;; esac need_lib_prefix=no need_version=no soname_spec='${libname}${release}${shared_ext}$major' library_names_spec='${libname}${release}${shared_ext}$versuffix ${libname}${release}${shared_ext}$major ${libname}${release}${shared_ext} $libname${shared_ext}' case $host_os in irix5* | nonstopux*) libsuff= shlibsuff= ;; *) case $LD in # libtool.m4 will add one of these switches to LD *-32|*"-32 "|*-melf32bsmip|*"-melf32bsmip ") libsuff= shlibsuff= libmagic=32-bit;; *-n32|*"-n32 "|*-melf32bmipn32|*"-melf32bmipn32 ") libsuff=32 shlibsuff=N32 libmagic=N32;; *-64|*"-64 "|*-melf64bmip|*"-melf64bmip ") libsuff=64 shlibsuff=64 libmagic=64-bit;; *) libsuff= shlibsuff= libmagic=never-match;; esac ;; esac shlibpath_var=LD_LIBRARY${shlibsuff}_PATH shlibpath_overrides_runpath=no sys_lib_search_path_spec="/usr/lib${libsuff} /lib${libsuff} /usr/local/lib${libsuff}" sys_lib_dlsearch_path_spec="/usr/lib${libsuff} /lib${libsuff}" hardcode_into_libs=yes ;; # No shared lib support for Linux oldld, aout, or coff. linux*oldld* | linux*aout* | linux*coff*) dynamic_linker=no ;; # This must be glibc/ELF. linux* | k*bsd*-gnu | kopensolaris*-gnu | gnu*) version_type=linux # correct to gnu/linux during the next big refactor need_lib_prefix=no need_version=no library_names_spec='${libname}${release}${shared_ext}$versuffix ${libname}${release}${shared_ext}$major $libname${shared_ext}' soname_spec='${libname}${release}${shared_ext}$major' finish_cmds='PATH="\$PATH:/sbin" ldconfig -n $libdir' shlibpath_var=LD_LIBRARY_PATH shlibpath_overrides_runpath=no # Some binutils ld are patched to set DT_RUNPATH AC_CACHE_VAL([lt_cv_shlibpath_overrides_runpath], [lt_cv_shlibpath_overrides_runpath=no save_LDFLAGS=$LDFLAGS save_libdir=$libdir eval "libdir=/foo; wl=\"$_LT_TAGVAR(lt_prog_compiler_wl, $1)\"; \ LDFLAGS=\"\$LDFLAGS $_LT_TAGVAR(hardcode_libdir_flag_spec, $1)\"" AC_LINK_IFELSE([AC_LANG_PROGRAM([],[])], [AS_IF([ ($OBJDUMP -p conftest$ac_exeext) 2>/dev/null | grep "RUNPATH.*$libdir" >/dev/null], [lt_cv_shlibpath_overrides_runpath=yes])]) LDFLAGS=$save_LDFLAGS libdir=$save_libdir ]) shlibpath_overrides_runpath=$lt_cv_shlibpath_overrides_runpath # This implies no fast_install, which is unacceptable. # Some rework will be needed to allow for fast_install # before this can be enabled. hardcode_into_libs=yes # Append ld.so.conf contents to the search path if test -f /etc/ld.so.conf; then lt_ld_extra=`awk '/^include / { system(sprintf("cd /etc; cat %s 2>/dev/null", \[$]2)); skip = 1; } { if (!skip) print \[$]0; skip = 0; }' < /etc/ld.so.conf | $SED -e 's/#.*//;/^[ ]*hwcap[ ]/d;s/[:, ]/ /g;s/=[^=]*$//;s/=[^= ]* / /g;s/"//g;/^$/d' | tr '\n' ' '` sys_lib_dlsearch_path_spec="/lib /usr/lib $lt_ld_extra" fi # We used to test for /lib/ld.so.1 and disable shared libraries on # powerpc, because MkLinux only supported shared libraries with the # GNU dynamic linker. Since this was broken with cross compilers, # most powerpc-linux boxes support dynamic linking these days and # people can always --disable-shared, the test was removed, and we # assume the GNU/Linux dynamic linker is in use. dynamic_linker='GNU/Linux ld.so' ;; netbsdelf*-gnu) version_type=linux need_lib_prefix=no need_version=no library_names_spec='${libname}${release}${shared_ext}$versuffix ${libname}${release}${shared_ext}$major ${libname}${shared_ext}' soname_spec='${libname}${release}${shared_ext}$major' shlibpath_var=LD_LIBRARY_PATH shlibpath_overrides_runpath=no hardcode_into_libs=yes dynamic_linker='NetBSD ld.elf_so' ;; netbsd*) version_type=sunos need_lib_prefix=no need_version=no if echo __ELF__ | $CC -E - | $GREP __ELF__ >/dev/null; then library_names_spec='${libname}${release}${shared_ext}$versuffix ${libname}${shared_ext}$versuffix' finish_cmds='PATH="\$PATH:/sbin" ldconfig -m $libdir' dynamic_linker='NetBSD (a.out) ld.so' else library_names_spec='${libname}${release}${shared_ext}$versuffix ${libname}${release}${shared_ext}$major ${libname}${shared_ext}' soname_spec='${libname}${release}${shared_ext}$major' dynamic_linker='NetBSD ld.elf_so' fi shlibpath_var=LD_LIBRARY_PATH shlibpath_overrides_runpath=yes hardcode_into_libs=yes ;; newsos6) version_type=linux # correct to gnu/linux during the next big refactor library_names_spec='${libname}${release}${shared_ext}$versuffix ${libname}${release}${shared_ext}$major $libname${shared_ext}' shlibpath_var=LD_LIBRARY_PATH shlibpath_overrides_runpath=yes ;; *nto* | *qnx*) version_type=qnx need_lib_prefix=no need_version=no library_names_spec='${libname}${release}${shared_ext}$versuffix ${libname}${release}${shared_ext}$major $libname${shared_ext}' soname_spec='${libname}${release}${shared_ext}$major' shlibpath_var=LD_LIBRARY_PATH shlibpath_overrides_runpath=no hardcode_into_libs=yes dynamic_linker='ldqnx.so' ;; openbsd*) version_type=sunos sys_lib_dlsearch_path_spec="/usr/lib" need_lib_prefix=no # Some older versions of OpenBSD (3.3 at least) *do* need versioned libs. case $host_os in openbsd3.3 | openbsd3.3.*) need_version=yes ;; *) need_version=no ;; esac library_names_spec='${libname}${release}${shared_ext}$versuffix ${libname}${shared_ext}$versuffix' finish_cmds='PATH="\$PATH:/sbin" ldconfig -m $libdir' shlibpath_var=LD_LIBRARY_PATH if test -z "`echo __ELF__ | $CC -E - | $GREP __ELF__`" || test "$host_os-$host_cpu" = "openbsd2.8-powerpc"; then case $host_os in openbsd2.[[89]] | openbsd2.[[89]].*) shlibpath_overrides_runpath=no ;; *) shlibpath_overrides_runpath=yes ;; esac else shlibpath_overrides_runpath=yes fi ;; os2*) libname_spec='$name' shrext_cmds=".dll" need_lib_prefix=no library_names_spec='$libname${shared_ext} $libname.a' dynamic_linker='OS/2 ld.exe' shlibpath_var=LIBPATH ;; osf3* | osf4* | osf5*) version_type=osf need_lib_prefix=no need_version=no soname_spec='${libname}${release}${shared_ext}$major' library_names_spec='${libname}${release}${shared_ext}$versuffix ${libname}${release}${shared_ext}$major $libname${shared_ext}' shlibpath_var=LD_LIBRARY_PATH sys_lib_search_path_spec="/usr/shlib /usr/ccs/lib /usr/lib/cmplrs/cc /usr/lib /usr/local/lib /var/shlib" sys_lib_dlsearch_path_spec="$sys_lib_search_path_spec" ;; rdos*) dynamic_linker=no ;; solaris*) version_type=linux # correct to gnu/linux during the next big refactor need_lib_prefix=no need_version=no library_names_spec='${libname}${release}${shared_ext}$versuffix ${libname}${release}${shared_ext}$major $libname${shared_ext}' soname_spec='${libname}${release}${shared_ext}$major' shlibpath_var=LD_LIBRARY_PATH shlibpath_overrides_runpath=yes hardcode_into_libs=yes # ldd complains unless libraries are executable postinstall_cmds='chmod +x $lib' ;; sunos4*) version_type=sunos library_names_spec='${libname}${release}${shared_ext}$versuffix ${libname}${shared_ext}$versuffix' finish_cmds='PATH="\$PATH:/usr/etc" ldconfig $libdir' shlibpath_var=LD_LIBRARY_PATH shlibpath_overrides_runpath=yes if test "$with_gnu_ld" = yes; then need_lib_prefix=no fi need_version=yes ;; sysv4 | sysv4.3*) version_type=linux # correct to gnu/linux during the next big refactor library_names_spec='${libname}${release}${shared_ext}$versuffix ${libname}${release}${shared_ext}$major $libname${shared_ext}' soname_spec='${libname}${release}${shared_ext}$major' shlibpath_var=LD_LIBRARY_PATH case $host_vendor in sni) shlibpath_overrides_runpath=no need_lib_prefix=no runpath_var=LD_RUN_PATH ;; siemens) need_lib_prefix=no ;; motorola) need_lib_prefix=no need_version=no shlibpath_overrides_runpath=no sys_lib_search_path_spec='/lib /usr/lib /usr/ccs/lib' ;; esac ;; sysv4*MP*) if test -d /usr/nec ;then version_type=linux # correct to gnu/linux during the next big refactor library_names_spec='$libname${shared_ext}.$versuffix $libname${shared_ext}.$major $libname${shared_ext}' soname_spec='$libname${shared_ext}.$major' shlibpath_var=LD_LIBRARY_PATH fi ;; sysv5* | sco3.2v5* | sco5v6* | unixware* | OpenUNIX* | sysv4*uw2*) version_type=freebsd-elf need_lib_prefix=no need_version=no library_names_spec='${libname}${release}${shared_ext}$versuffix ${libname}${release}${shared_ext} $libname${shared_ext}' soname_spec='${libname}${release}${shared_ext}$major' shlibpath_var=LD_LIBRARY_PATH shlibpath_overrides_runpath=yes hardcode_into_libs=yes if test "$with_gnu_ld" = yes; then sys_lib_search_path_spec='/usr/local/lib /usr/gnu/lib /usr/ccs/lib /usr/lib /lib' else sys_lib_search_path_spec='/usr/ccs/lib /usr/lib' case $host_os in sco3.2v5*) sys_lib_search_path_spec="$sys_lib_search_path_spec /lib" ;; esac fi sys_lib_dlsearch_path_spec='/usr/lib' ;; tpf*) # TPF is a cross-target only. Preferred cross-host = GNU/Linux. version_type=linux # correct to gnu/linux during the next big refactor need_lib_prefix=no need_version=no library_names_spec='${libname}${release}${shared_ext}$versuffix ${libname}${release}${shared_ext}$major $libname${shared_ext}' shlibpath_var=LD_LIBRARY_PATH shlibpath_overrides_runpath=no hardcode_into_libs=yes ;; uts4*) version_type=linux # correct to gnu/linux during the next big refactor library_names_spec='${libname}${release}${shared_ext}$versuffix ${libname}${release}${shared_ext}$major $libname${shared_ext}' soname_spec='${libname}${release}${shared_ext}$major' shlibpath_var=LD_LIBRARY_PATH ;; *) dynamic_linker=no ;; esac AC_MSG_RESULT([$dynamic_linker]) test "$dynamic_linker" = no && can_build_shared=no variables_saved_for_relink="PATH $shlibpath_var $runpath_var" if test "$GCC" = yes; then variables_saved_for_relink="$variables_saved_for_relink GCC_EXEC_PREFIX COMPILER_PATH LIBRARY_PATH" fi if test "${lt_cv_sys_lib_search_path_spec+set}" = set; then sys_lib_search_path_spec="$lt_cv_sys_lib_search_path_spec" fi if test "${lt_cv_sys_lib_dlsearch_path_spec+set}" = set; then sys_lib_dlsearch_path_spec="$lt_cv_sys_lib_dlsearch_path_spec" fi _LT_DECL([], [variables_saved_for_relink], [1], [Variables whose values should be saved in libtool wrapper scripts and restored at link time]) _LT_DECL([], [need_lib_prefix], [0], [Do we need the "lib" prefix for modules?]) _LT_DECL([], [need_version], [0], [Do we need a version for libraries?]) _LT_DECL([], [version_type], [0], [Library versioning type]) _LT_DECL([], [runpath_var], [0], [Shared library runtime path variable]) _LT_DECL([], [shlibpath_var], [0],[Shared library path variable]) _LT_DECL([], [shlibpath_overrides_runpath], [0], [Is shlibpath searched before the hard-coded library search path?]) _LT_DECL([], [libname_spec], [1], [Format of library name prefix]) _LT_DECL([], [library_names_spec], [1], [[List of archive names. First name is the real one, the rest are links. The last name is the one that the linker finds with -lNAME]]) _LT_DECL([], [soname_spec], [1], [[The coded name of the library, if different from the real name]]) _LT_DECL([], [install_override_mode], [1], [Permission mode override for installation of shared libraries]) _LT_DECL([], [postinstall_cmds], [2], [Command to use after installation of a shared archive]) _LT_DECL([], [postuninstall_cmds], [2], [Command to use after uninstallation of a shared archive]) _LT_DECL([], [finish_cmds], [2], [Commands used to finish a libtool library installation in a directory]) _LT_DECL([], [finish_eval], [1], [[As "finish_cmds", except a single script fragment to be evaled but not shown]]) _LT_DECL([], [hardcode_into_libs], [0], [Whether we should hardcode library paths into libraries]) _LT_DECL([], [sys_lib_search_path_spec], [2], [Compile-time system search path for libraries]) _LT_DECL([], [sys_lib_dlsearch_path_spec], [2], [Run-time system search path for libraries]) ])# _LT_SYS_DYNAMIC_LINKER # _LT_PATH_TOOL_PREFIX(TOOL) # -------------------------- # find a file program which can recognize shared library AC_DEFUN([_LT_PATH_TOOL_PREFIX], [m4_require([_LT_DECL_EGREP])dnl AC_MSG_CHECKING([for $1]) AC_CACHE_VAL(lt_cv_path_MAGIC_CMD, [case $MAGIC_CMD in [[\\/*] | ?:[\\/]*]) lt_cv_path_MAGIC_CMD="$MAGIC_CMD" # Let the user override the test with a path. ;; *) lt_save_MAGIC_CMD="$MAGIC_CMD" lt_save_ifs="$IFS"; IFS=$PATH_SEPARATOR dnl $ac_dummy forces splitting on constant user-supplied paths. dnl POSIX.2 word splitting is done only on the output of word expansions, dnl not every word. This closes a longstanding sh security hole. ac_dummy="m4_if([$2], , $PATH, [$2])" for ac_dir in $ac_dummy; do IFS="$lt_save_ifs" test -z "$ac_dir" && ac_dir=. if test -f $ac_dir/$1; then lt_cv_path_MAGIC_CMD="$ac_dir/$1" if test -n "$file_magic_test_file"; then case $deplibs_check_method in "file_magic "*) file_magic_regex=`expr "$deplibs_check_method" : "file_magic \(.*\)"` MAGIC_CMD="$lt_cv_path_MAGIC_CMD" if eval $file_magic_cmd \$file_magic_test_file 2> /dev/null | $EGREP "$file_magic_regex" > /dev/null; then : else cat <<_LT_EOF 1>&2 *** Warning: the command libtool uses to detect shared libraries, *** $file_magic_cmd, produces output that libtool cannot recognize. *** The result is that libtool may fail to recognize shared libraries *** as such. This will affect the creation of libtool libraries that *** depend on shared libraries, but programs linked with such libtool *** libraries will work regardless of this problem. Nevertheless, you *** may want to report the problem to your system manager and/or to *** bug-libtool@gnu.org _LT_EOF fi ;; esac fi break fi done IFS="$lt_save_ifs" MAGIC_CMD="$lt_save_MAGIC_CMD" ;; esac]) MAGIC_CMD="$lt_cv_path_MAGIC_CMD" if test -n "$MAGIC_CMD"; then AC_MSG_RESULT($MAGIC_CMD) else AC_MSG_RESULT(no) fi _LT_DECL([], [MAGIC_CMD], [0], [Used to examine libraries when file_magic_cmd begins with "file"])dnl ])# _LT_PATH_TOOL_PREFIX # Old name: AU_ALIAS([AC_PATH_TOOL_PREFIX], [_LT_PATH_TOOL_PREFIX]) dnl aclocal-1.4 backwards compatibility: dnl AC_DEFUN([AC_PATH_TOOL_PREFIX], []) # _LT_PATH_MAGIC # -------------- # find a file program which can recognize a shared library m4_defun([_LT_PATH_MAGIC], [_LT_PATH_TOOL_PREFIX(${ac_tool_prefix}file, /usr/bin$PATH_SEPARATOR$PATH) if test -z "$lt_cv_path_MAGIC_CMD"; then if test -n "$ac_tool_prefix"; then _LT_PATH_TOOL_PREFIX(file, /usr/bin$PATH_SEPARATOR$PATH) else MAGIC_CMD=: fi fi ])# _LT_PATH_MAGIC # LT_PATH_LD # ---------- # find the pathname to the GNU or non-GNU linker AC_DEFUN([LT_PATH_LD], [AC_REQUIRE([AC_PROG_CC])dnl AC_REQUIRE([AC_CANONICAL_HOST])dnl AC_REQUIRE([AC_CANONICAL_BUILD])dnl m4_require([_LT_DECL_SED])dnl m4_require([_LT_DECL_EGREP])dnl m4_require([_LT_PROG_ECHO_BACKSLASH])dnl AC_ARG_WITH([gnu-ld], [AS_HELP_STRING([--with-gnu-ld], [assume the C compiler uses GNU ld @<:@default=no@:>@])], [test "$withval" = no || with_gnu_ld=yes], [with_gnu_ld=no])dnl ac_prog=ld if test "$GCC" = yes; then # Check if gcc -print-prog-name=ld gives a path. AC_MSG_CHECKING([for ld used by $CC]) case $host in *-*-mingw*) # gcc leaves a trailing carriage return which upsets mingw ac_prog=`($CC -print-prog-name=ld) 2>&5 | tr -d '\015'` ;; *) ac_prog=`($CC -print-prog-name=ld) 2>&5` ;; esac case $ac_prog in # Accept absolute paths. [[\\/]]* | ?:[[\\/]]*) re_direlt='/[[^/]][[^/]]*/\.\./' # Canonicalize the pathname of ld ac_prog=`$ECHO "$ac_prog"| $SED 's%\\\\%/%g'` while $ECHO "$ac_prog" | $GREP "$re_direlt" > /dev/null 2>&1; do ac_prog=`$ECHO $ac_prog| $SED "s%$re_direlt%/%"` done test -z "$LD" && LD="$ac_prog" ;; "") # If it fails, then pretend we aren't using GCC. ac_prog=ld ;; *) # If it is relative, then search for the first ld in PATH. with_gnu_ld=unknown ;; esac elif test "$with_gnu_ld" = yes; then AC_MSG_CHECKING([for GNU ld]) else AC_MSG_CHECKING([for non-GNU ld]) fi AC_CACHE_VAL(lt_cv_path_LD, [if test -z "$LD"; then lt_save_ifs="$IFS"; IFS=$PATH_SEPARATOR for ac_dir in $PATH; do IFS="$lt_save_ifs" test -z "$ac_dir" && ac_dir=. if test -f "$ac_dir/$ac_prog" || test -f "$ac_dir/$ac_prog$ac_exeext"; then lt_cv_path_LD="$ac_dir/$ac_prog" # Check to see if the program is GNU ld. I'd rather use --version, # but apparently some variants of GNU ld only accept -v. # Break only if it was the GNU/non-GNU ld that we prefer. case `"$lt_cv_path_LD" -v 2>&1 &1 /dev/null 2>&1; then lt_cv_deplibs_check_method='file_magic ^x86 archive import|^x86 DLL' lt_cv_file_magic_cmd='func_win32_libid' else # Keep this pattern in sync with the one in func_win32_libid. lt_cv_deplibs_check_method='file_magic file format (pei*-i386(.*architecture: i386)?|pe-arm-wince|pe-x86-64)' lt_cv_file_magic_cmd='$OBJDUMP -f' fi ;; cegcc*) # use the weaker test based on 'objdump'. See mingw*. lt_cv_deplibs_check_method='file_magic file format pe-arm-.*little(.*architecture: arm)?' lt_cv_file_magic_cmd='$OBJDUMP -f' ;; darwin* | rhapsody*) lt_cv_deplibs_check_method=pass_all ;; freebsd* | dragonfly*) if echo __ELF__ | $CC -E - | $GREP __ELF__ > /dev/null; then case $host_cpu in i*86 ) # Not sure whether the presence of OpenBSD here was a mistake. # Let's accept both of them until this is cleared up. lt_cv_deplibs_check_method='file_magic (FreeBSD|OpenBSD|DragonFly)/i[[3-9]]86 (compact )?demand paged shared library' lt_cv_file_magic_cmd=/usr/bin/file lt_cv_file_magic_test_file=`echo /usr/lib/libc.so.*` ;; esac else lt_cv_deplibs_check_method=pass_all fi ;; haiku*) lt_cv_deplibs_check_method=pass_all ;; hpux10.20* | hpux11*) lt_cv_file_magic_cmd=/usr/bin/file case $host_cpu in ia64*) lt_cv_deplibs_check_method='file_magic (s[[0-9]][[0-9]][[0-9]]|ELF-[[0-9]][[0-9]]) shared object file - IA64' lt_cv_file_magic_test_file=/usr/lib/hpux32/libc.so ;; hppa*64*) [lt_cv_deplibs_check_method='file_magic (s[0-9][0-9][0-9]|ELF[ -][0-9][0-9])(-bit)?( [LM]SB)? shared object( file)?[, -]* PA-RISC [0-9]\.[0-9]'] lt_cv_file_magic_test_file=/usr/lib/pa20_64/libc.sl ;; *) lt_cv_deplibs_check_method='file_magic (s[[0-9]][[0-9]][[0-9]]|PA-RISC[[0-9]]\.[[0-9]]) shared library' lt_cv_file_magic_test_file=/usr/lib/libc.sl ;; esac ;; interix[[3-9]]*) # PIC code is broken on Interix 3.x, that's why |\.a not |_pic\.a here lt_cv_deplibs_check_method='match_pattern /lib[[^/]]+(\.so|\.a)$' ;; irix5* | irix6* | nonstopux*) case $LD in *-32|*"-32 ") libmagic=32-bit;; *-n32|*"-n32 ") libmagic=N32;; *-64|*"-64 ") libmagic=64-bit;; *) libmagic=never-match;; esac lt_cv_deplibs_check_method=pass_all ;; # This must be glibc/ELF. linux* | k*bsd*-gnu | kopensolaris*-gnu | gnu*) lt_cv_deplibs_check_method=pass_all ;; netbsd* | netbsdelf*-gnu) if echo __ELF__ | $CC -E - | $GREP __ELF__ > /dev/null; then lt_cv_deplibs_check_method='match_pattern /lib[[^/]]+(\.so\.[[0-9]]+\.[[0-9]]+|_pic\.a)$' else lt_cv_deplibs_check_method='match_pattern /lib[[^/]]+(\.so|_pic\.a)$' fi ;; newos6*) lt_cv_deplibs_check_method='file_magic ELF [[0-9]][[0-9]]*-bit [[ML]]SB (executable|dynamic lib)' lt_cv_file_magic_cmd=/usr/bin/file lt_cv_file_magic_test_file=/usr/lib/libnls.so ;; *nto* | *qnx*) lt_cv_deplibs_check_method=pass_all ;; openbsd*) if test -z "`echo __ELF__ | $CC -E - | $GREP __ELF__`" || test "$host_os-$host_cpu" = "openbsd2.8-powerpc"; then lt_cv_deplibs_check_method='match_pattern /lib[[^/]]+(\.so\.[[0-9]]+\.[[0-9]]+|\.so|_pic\.a)$' else lt_cv_deplibs_check_method='match_pattern /lib[[^/]]+(\.so\.[[0-9]]+\.[[0-9]]+|_pic\.a)$' fi ;; osf3* | osf4* | osf5*) lt_cv_deplibs_check_method=pass_all ;; rdos*) lt_cv_deplibs_check_method=pass_all ;; solaris*) lt_cv_deplibs_check_method=pass_all ;; sysv5* | sco3.2v5* | sco5v6* | unixware* | OpenUNIX* | sysv4*uw2*) lt_cv_deplibs_check_method=pass_all ;; sysv4 | sysv4.3*) case $host_vendor in motorola) lt_cv_deplibs_check_method='file_magic ELF [[0-9]][[0-9]]*-bit [[ML]]SB (shared object|dynamic lib) M[[0-9]][[0-9]]* Version [[0-9]]' lt_cv_file_magic_test_file=`echo /usr/lib/libc.so*` ;; ncr) lt_cv_deplibs_check_method=pass_all ;; sequent) lt_cv_file_magic_cmd='/bin/file' lt_cv_deplibs_check_method='file_magic ELF [[0-9]][[0-9]]*-bit [[LM]]SB (shared object|dynamic lib )' ;; sni) lt_cv_file_magic_cmd='/bin/file' lt_cv_deplibs_check_method="file_magic ELF [[0-9]][[0-9]]*-bit [[LM]]SB dynamic lib" lt_cv_file_magic_test_file=/lib/libc.so ;; siemens) lt_cv_deplibs_check_method=pass_all ;; pc) lt_cv_deplibs_check_method=pass_all ;; esac ;; tpf*) lt_cv_deplibs_check_method=pass_all ;; esac ]) file_magic_glob= want_nocaseglob=no if test "$build" = "$host"; then case $host_os in mingw* | pw32*) if ( shopt | grep nocaseglob ) >/dev/null 2>&1; then want_nocaseglob=yes else file_magic_glob=`echo aAbBcCdDeEfFgGhHiIjJkKlLmMnNoOpPqQrRsStTuUvVwWxXyYzZ | $SED -e "s/\(..\)/s\/[[\1]]\/[[\1]]\/g;/g"` fi ;; esac fi file_magic_cmd=$lt_cv_file_magic_cmd deplibs_check_method=$lt_cv_deplibs_check_method test -z "$deplibs_check_method" && deplibs_check_method=unknown _LT_DECL([], [deplibs_check_method], [1], [Method to check whether dependent libraries are shared objects]) _LT_DECL([], [file_magic_cmd], [1], [Command to use when deplibs_check_method = "file_magic"]) _LT_DECL([], [file_magic_glob], [1], [How to find potential files when deplibs_check_method = "file_magic"]) _LT_DECL([], [want_nocaseglob], [1], [Find potential files using nocaseglob when deplibs_check_method = "file_magic"]) ])# _LT_CHECK_MAGIC_METHOD # LT_PATH_NM # ---------- # find the pathname to a BSD- or MS-compatible name lister AC_DEFUN([LT_PATH_NM], [AC_REQUIRE([AC_PROG_CC])dnl AC_CACHE_CHECK([for BSD- or MS-compatible name lister (nm)], lt_cv_path_NM, [if test -n "$NM"; then # Let the user override the test. lt_cv_path_NM="$NM" else lt_nm_to_check="${ac_tool_prefix}nm" if test -n "$ac_tool_prefix" && test "$build" = "$host"; then lt_nm_to_check="$lt_nm_to_check nm" fi for lt_tmp_nm in $lt_nm_to_check; do lt_save_ifs="$IFS"; IFS=$PATH_SEPARATOR for ac_dir in $PATH /usr/ccs/bin/elf /usr/ccs/bin /usr/ucb /bin; do IFS="$lt_save_ifs" test -z "$ac_dir" && ac_dir=. tmp_nm="$ac_dir/$lt_tmp_nm" if test -f "$tmp_nm" || test -f "$tmp_nm$ac_exeext" ; then # Check to see if the nm accepts a BSD-compat flag. # Adding the `sed 1q' prevents false positives on HP-UX, which says: # nm: unknown option "B" ignored # Tru64's nm complains that /dev/null is an invalid object file case `"$tmp_nm" -B /dev/null 2>&1 | sed '1q'` in */dev/null* | *'Invalid file or object type'*) lt_cv_path_NM="$tmp_nm -B" break ;; *) case `"$tmp_nm" -p /dev/null 2>&1 | sed '1q'` in */dev/null*) lt_cv_path_NM="$tmp_nm -p" break ;; *) lt_cv_path_NM=${lt_cv_path_NM="$tmp_nm"} # keep the first match, but continue # so that we can try to find one that supports BSD flags ;; esac ;; esac fi done IFS="$lt_save_ifs" done : ${lt_cv_path_NM=no} fi]) if test "$lt_cv_path_NM" != "no"; then NM="$lt_cv_path_NM" else # Didn't find any BSD compatible name lister, look for dumpbin. if test -n "$DUMPBIN"; then : # Let the user override the test. else AC_CHECK_TOOLS(DUMPBIN, [dumpbin "link -dump"], :) case `$DUMPBIN -symbols /dev/null 2>&1 | sed '1q'` in *COFF*) DUMPBIN="$DUMPBIN -symbols" ;; *) DUMPBIN=: ;; esac fi AC_SUBST([DUMPBIN]) if test "$DUMPBIN" != ":"; then NM="$DUMPBIN" fi fi test -z "$NM" && NM=nm AC_SUBST([NM]) _LT_DECL([], [NM], [1], [A BSD- or MS-compatible name lister])dnl AC_CACHE_CHECK([the name lister ($NM) interface], [lt_cv_nm_interface], [lt_cv_nm_interface="BSD nm" echo "int some_variable = 0;" > conftest.$ac_ext (eval echo "\"\$as_me:$LINENO: $ac_compile\"" >&AS_MESSAGE_LOG_FD) (eval "$ac_compile" 2>conftest.err) cat conftest.err >&AS_MESSAGE_LOG_FD (eval echo "\"\$as_me:$LINENO: $NM \\\"conftest.$ac_objext\\\"\"" >&AS_MESSAGE_LOG_FD) (eval "$NM \"conftest.$ac_objext\"" 2>conftest.err > conftest.out) cat conftest.err >&AS_MESSAGE_LOG_FD (eval echo "\"\$as_me:$LINENO: output\"" >&AS_MESSAGE_LOG_FD) cat conftest.out >&AS_MESSAGE_LOG_FD if $GREP 'External.*some_variable' conftest.out > /dev/null; then lt_cv_nm_interface="MS dumpbin" fi rm -f conftest*]) ])# LT_PATH_NM # Old names: AU_ALIAS([AM_PROG_NM], [LT_PATH_NM]) AU_ALIAS([AC_PROG_NM], [LT_PATH_NM]) dnl aclocal-1.4 backwards compatibility: dnl AC_DEFUN([AM_PROG_NM], []) dnl AC_DEFUN([AC_PROG_NM], []) # _LT_CHECK_SHAREDLIB_FROM_LINKLIB # -------------------------------- # how to determine the name of the shared library # associated with a specific link library. # -- PORTME fill in with the dynamic library characteristics m4_defun([_LT_CHECK_SHAREDLIB_FROM_LINKLIB], [m4_require([_LT_DECL_EGREP]) m4_require([_LT_DECL_OBJDUMP]) m4_require([_LT_DECL_DLLTOOL]) AC_CACHE_CHECK([how to associate runtime and link libraries], lt_cv_sharedlib_from_linklib_cmd, [lt_cv_sharedlib_from_linklib_cmd='unknown' case $host_os in cygwin* | mingw* | pw32* | cegcc*) # two different shell functions defined in ltmain.sh # decide which to use based on capabilities of $DLLTOOL case `$DLLTOOL --help 2>&1` in *--identify-strict*) lt_cv_sharedlib_from_linklib_cmd=func_cygming_dll_for_implib ;; *) lt_cv_sharedlib_from_linklib_cmd=func_cygming_dll_for_implib_fallback ;; esac ;; *) # fallback: assume linklib IS sharedlib lt_cv_sharedlib_from_linklib_cmd="$ECHO" ;; esac ]) sharedlib_from_linklib_cmd=$lt_cv_sharedlib_from_linklib_cmd test -z "$sharedlib_from_linklib_cmd" && sharedlib_from_linklib_cmd=$ECHO _LT_DECL([], [sharedlib_from_linklib_cmd], [1], [Command to associate shared and link libraries]) ])# _LT_CHECK_SHAREDLIB_FROM_LINKLIB # _LT_PATH_MANIFEST_TOOL # ---------------------- # locate the manifest tool m4_defun([_LT_PATH_MANIFEST_TOOL], [AC_CHECK_TOOL(MANIFEST_TOOL, mt, :) test -z "$MANIFEST_TOOL" && MANIFEST_TOOL=mt AC_CACHE_CHECK([if $MANIFEST_TOOL is a manifest tool], [lt_cv_path_mainfest_tool], [lt_cv_path_mainfest_tool=no echo "$as_me:$LINENO: $MANIFEST_TOOL '-?'" >&AS_MESSAGE_LOG_FD $MANIFEST_TOOL '-?' 2>conftest.err > conftest.out cat conftest.err >&AS_MESSAGE_LOG_FD if $GREP 'Manifest Tool' conftest.out > /dev/null; then lt_cv_path_mainfest_tool=yes fi rm -f conftest*]) if test "x$lt_cv_path_mainfest_tool" != xyes; then MANIFEST_TOOL=: fi _LT_DECL([], [MANIFEST_TOOL], [1], [Manifest tool])dnl ])# _LT_PATH_MANIFEST_TOOL # LT_LIB_M # -------- # check for math library AC_DEFUN([LT_LIB_M], [AC_REQUIRE([AC_CANONICAL_HOST])dnl LIBM= case $host in *-*-beos* | *-*-cegcc* | *-*-cygwin* | *-*-haiku* | *-*-pw32* | *-*-darwin*) # These system don't have libm, or don't need it ;; *-ncr-sysv4.3*) AC_CHECK_LIB(mw, _mwvalidcheckl, LIBM="-lmw") AC_CHECK_LIB(m, cos, LIBM="$LIBM -lm") ;; *) AC_CHECK_LIB(m, cos, LIBM="-lm") ;; esac AC_SUBST([LIBM]) ])# LT_LIB_M # Old name: AU_ALIAS([AC_CHECK_LIBM], [LT_LIB_M]) dnl aclocal-1.4 backwards compatibility: dnl AC_DEFUN([AC_CHECK_LIBM], []) # _LT_COMPILER_NO_RTTI([TAGNAME]) # ------------------------------- m4_defun([_LT_COMPILER_NO_RTTI], [m4_require([_LT_TAG_COMPILER])dnl _LT_TAGVAR(lt_prog_compiler_no_builtin_flag, $1)= if test "$GCC" = yes; then case $cc_basename in nvcc*) _LT_TAGVAR(lt_prog_compiler_no_builtin_flag, $1)=' -Xcompiler -fno-builtin' ;; *) _LT_TAGVAR(lt_prog_compiler_no_builtin_flag, $1)=' -fno-builtin' ;; esac _LT_COMPILER_OPTION([if $compiler supports -fno-rtti -fno-exceptions], lt_cv_prog_compiler_rtti_exceptions, [-fno-rtti -fno-exceptions], [], [_LT_TAGVAR(lt_prog_compiler_no_builtin_flag, $1)="$_LT_TAGVAR(lt_prog_compiler_no_builtin_flag, $1) -fno-rtti -fno-exceptions"]) fi _LT_TAGDECL([no_builtin_flag], [lt_prog_compiler_no_builtin_flag], [1], [Compiler flag to turn off builtin functions]) ])# _LT_COMPILER_NO_RTTI # _LT_CMD_GLOBAL_SYMBOLS # ---------------------- m4_defun([_LT_CMD_GLOBAL_SYMBOLS], [AC_REQUIRE([AC_CANONICAL_HOST])dnl AC_REQUIRE([AC_PROG_CC])dnl AC_REQUIRE([AC_PROG_AWK])dnl AC_REQUIRE([LT_PATH_NM])dnl AC_REQUIRE([LT_PATH_LD])dnl m4_require([_LT_DECL_SED])dnl m4_require([_LT_DECL_EGREP])dnl m4_require([_LT_TAG_COMPILER])dnl # Check for command to grab the raw symbol name followed by C symbol from nm. AC_MSG_CHECKING([command to parse $NM output from $compiler object]) AC_CACHE_VAL([lt_cv_sys_global_symbol_pipe], [ # These are sane defaults that work on at least a few old systems. # [They come from Ultrix. What could be older than Ultrix?!! ;)] # Character class describing NM global symbol codes. symcode='[[BCDEGRST]]' # Regexp to match symbols that can be accessed directly from C. sympat='\([[_A-Za-z]][[_A-Za-z0-9]]*\)' # Define system-specific variables. case $host_os in aix*) symcode='[[BCDT]]' ;; cygwin* | mingw* | pw32* | cegcc*) symcode='[[ABCDGISTW]]' ;; hpux*) if test "$host_cpu" = ia64; then symcode='[[ABCDEGRST]]' fi ;; irix* | nonstopux*) symcode='[[BCDEGRST]]' ;; osf*) symcode='[[BCDEGQRST]]' ;; solaris*) symcode='[[BDRT]]' ;; sco3.2v5*) symcode='[[DT]]' ;; sysv4.2uw2*) symcode='[[DT]]' ;; sysv5* | sco5v6* | unixware* | OpenUNIX*) symcode='[[ABDT]]' ;; sysv4) symcode='[[DFNSTU]]' ;; esac # If we're using GNU nm, then use its standard symbol codes. case `$NM -V 2>&1` in *GNU* | *'with BFD'*) symcode='[[ABCDGIRSTW]]' ;; esac # Transform an extracted symbol line into a proper C declaration. # Some systems (esp. on ia64) link data and code symbols differently, # so use this general approach. lt_cv_sys_global_symbol_to_cdecl="sed -n -e 's/^T .* \(.*\)$/extern int \1();/p' -e 's/^$symcode* .* \(.*\)$/extern char \1;/p'" # Transform an extracted symbol line into symbol name and symbol address lt_cv_sys_global_symbol_to_c_name_address="sed -n -e 's/^: \([[^ ]]*\)[[ ]]*$/ {\\\"\1\\\", (void *) 0},/p' -e 's/^$symcode* \([[^ ]]*\) \([[^ ]]*\)$/ {\"\2\", (void *) \&\2},/p'" lt_cv_sys_global_symbol_to_c_name_address_lib_prefix="sed -n -e 's/^: \([[^ ]]*\)[[ ]]*$/ {\\\"\1\\\", (void *) 0},/p' -e 's/^$symcode* \([[^ ]]*\) \(lib[[^ ]]*\)$/ {\"\2\", (void *) \&\2},/p' -e 's/^$symcode* \([[^ ]]*\) \([[^ ]]*\)$/ {\"lib\2\", (void *) \&\2},/p'" # Handle CRLF in mingw tool chain opt_cr= case $build_os in mingw*) opt_cr=`$ECHO 'x\{0,1\}' | tr x '\015'` # option cr in regexp ;; esac # Try without a prefix underscore, then with it. for ac_symprfx in "" "_"; do # Transform symcode, sympat, and symprfx into a raw symbol and a C symbol. symxfrm="\\1 $ac_symprfx\\2 \\2" # Write the raw and C identifiers. if test "$lt_cv_nm_interface" = "MS dumpbin"; then # Fake it for dumpbin and say T for any non-static function # and D for any global variable. # Also find C++ and __fastcall symbols from MSVC++, # which start with @ or ?. lt_cv_sys_global_symbol_pipe="$AWK ['"\ " {last_section=section; section=\$ 3};"\ " /^COFF SYMBOL TABLE/{for(i in hide) delete hide[i]};"\ " /Section length .*#relocs.*(pick any)/{hide[last_section]=1};"\ " \$ 0!~/External *\|/{next};"\ " / 0+ UNDEF /{next}; / UNDEF \([^|]\)*()/{next};"\ " {if(hide[section]) next};"\ " {f=0}; \$ 0~/\(\).*\|/{f=1}; {printf f ? \"T \" : \"D \"};"\ " {split(\$ 0, a, /\||\r/); split(a[2], s)};"\ " s[1]~/^[@?]/{print s[1], s[1]; next};"\ " s[1]~prfx {split(s[1],t,\"@\"); print t[1], substr(t[1],length(prfx))}"\ " ' prfx=^$ac_symprfx]" else lt_cv_sys_global_symbol_pipe="sed -n -e 's/^.*[[ ]]\($symcode$symcode*\)[[ ]][[ ]]*$ac_symprfx$sympat$opt_cr$/$symxfrm/p'" fi lt_cv_sys_global_symbol_pipe="$lt_cv_sys_global_symbol_pipe | sed '/ __gnu_lto/d'" # Check to see that the pipe works correctly. pipe_works=no rm -f conftest* cat > conftest.$ac_ext <<_LT_EOF #ifdef __cplusplus extern "C" { #endif char nm_test_var; void nm_test_func(void); void nm_test_func(void){} #ifdef __cplusplus } #endif int main(){nm_test_var='a';nm_test_func();return(0);} _LT_EOF if AC_TRY_EVAL(ac_compile); then # Now try to grab the symbols. nlist=conftest.nm if AC_TRY_EVAL(NM conftest.$ac_objext \| "$lt_cv_sys_global_symbol_pipe" \> $nlist) && test -s "$nlist"; then # Try sorting and uniquifying the output. if sort "$nlist" | uniq > "$nlist"T; then mv -f "$nlist"T "$nlist" else rm -f "$nlist"T fi # Make sure that we snagged all the symbols we need. if $GREP ' nm_test_var$' "$nlist" >/dev/null; then if $GREP ' nm_test_func$' "$nlist" >/dev/null; then cat <<_LT_EOF > conftest.$ac_ext /* Keep this code in sync between libtool.m4, ltmain, lt_system.h, and tests. */ #if defined(_WIN32) || defined(__CYGWIN__) || defined(_WIN32_WCE) /* DATA imports from DLLs on WIN32 con't be const, because runtime relocations are performed -- see ld's documentation on pseudo-relocs. */ # define LT@&t@_DLSYM_CONST #elif defined(__osf__) /* This system does not cope well with relocations in const data. */ # define LT@&t@_DLSYM_CONST #else # define LT@&t@_DLSYM_CONST const #endif #ifdef __cplusplus extern "C" { #endif _LT_EOF # Now generate the symbol file. eval "$lt_cv_sys_global_symbol_to_cdecl"' < "$nlist" | $GREP -v main >> conftest.$ac_ext' cat <<_LT_EOF >> conftest.$ac_ext /* The mapping between symbol names and symbols. */ LT@&t@_DLSYM_CONST struct { const char *name; void *address; } lt__PROGRAM__LTX_preloaded_symbols[[]] = { { "@PROGRAM@", (void *) 0 }, _LT_EOF $SED "s/^$symcode$symcode* \(.*\) \(.*\)$/ {\"\2\", (void *) \&\2},/" < "$nlist" | $GREP -v main >> conftest.$ac_ext cat <<\_LT_EOF >> conftest.$ac_ext {0, (void *) 0} }; /* This works around a problem in FreeBSD linker */ #ifdef FREEBSD_WORKAROUND static const void *lt_preloaded_setup() { return lt__PROGRAM__LTX_preloaded_symbols; } #endif #ifdef __cplusplus } #endif _LT_EOF # Now try linking the two files. mv conftest.$ac_objext conftstm.$ac_objext lt_globsym_save_LIBS=$LIBS lt_globsym_save_CFLAGS=$CFLAGS LIBS="conftstm.$ac_objext" CFLAGS="$CFLAGS$_LT_TAGVAR(lt_prog_compiler_no_builtin_flag, $1)" if AC_TRY_EVAL(ac_link) && test -s conftest${ac_exeext}; then pipe_works=yes fi LIBS=$lt_globsym_save_LIBS CFLAGS=$lt_globsym_save_CFLAGS else echo "cannot find nm_test_func in $nlist" >&AS_MESSAGE_LOG_FD fi else echo "cannot find nm_test_var in $nlist" >&AS_MESSAGE_LOG_FD fi else echo "cannot run $lt_cv_sys_global_symbol_pipe" >&AS_MESSAGE_LOG_FD fi else echo "$progname: failed program was:" >&AS_MESSAGE_LOG_FD cat conftest.$ac_ext >&5 fi rm -rf conftest* conftst* # Do not use the global_symbol_pipe unless it works. if test "$pipe_works" = yes; then break else lt_cv_sys_global_symbol_pipe= fi done ]) if test -z "$lt_cv_sys_global_symbol_pipe"; then lt_cv_sys_global_symbol_to_cdecl= fi if test -z "$lt_cv_sys_global_symbol_pipe$lt_cv_sys_global_symbol_to_cdecl"; then AC_MSG_RESULT(failed) else AC_MSG_RESULT(ok) fi # Response file support. if test "$lt_cv_nm_interface" = "MS dumpbin"; then nm_file_list_spec='@' elif $NM --help 2>/dev/null | grep '[[@]]FILE' >/dev/null; then nm_file_list_spec='@' fi _LT_DECL([global_symbol_pipe], [lt_cv_sys_global_symbol_pipe], [1], [Take the output of nm and produce a listing of raw symbols and C names]) _LT_DECL([global_symbol_to_cdecl], [lt_cv_sys_global_symbol_to_cdecl], [1], [Transform the output of nm in a proper C declaration]) _LT_DECL([global_symbol_to_c_name_address], [lt_cv_sys_global_symbol_to_c_name_address], [1], [Transform the output of nm in a C name address pair]) _LT_DECL([global_symbol_to_c_name_address_lib_prefix], [lt_cv_sys_global_symbol_to_c_name_address_lib_prefix], [1], [Transform the output of nm in a C name address pair when lib prefix is needed]) _LT_DECL([], [nm_file_list_spec], [1], [Specify filename containing input files for $NM]) ]) # _LT_CMD_GLOBAL_SYMBOLS # _LT_COMPILER_PIC([TAGNAME]) # --------------------------- m4_defun([_LT_COMPILER_PIC], [m4_require([_LT_TAG_COMPILER])dnl _LT_TAGVAR(lt_prog_compiler_wl, $1)= _LT_TAGVAR(lt_prog_compiler_pic, $1)= _LT_TAGVAR(lt_prog_compiler_static, $1)= m4_if([$1], [CXX], [ # C++ specific cases for pic, static, wl, etc. if test "$GXX" = yes; then _LT_TAGVAR(lt_prog_compiler_wl, $1)='-Wl,' _LT_TAGVAR(lt_prog_compiler_static, $1)='-static' case $host_os in aix*) # All AIX code is PIC. if test "$host_cpu" = ia64; then # AIX 5 now supports IA64 processor _LT_TAGVAR(lt_prog_compiler_static, $1)='-Bstatic' fi ;; amigaos*) case $host_cpu in powerpc) # see comment about AmigaOS4 .so support _LT_TAGVAR(lt_prog_compiler_pic, $1)='-fPIC' ;; m68k) # FIXME: we need at least 68020 code to build shared libraries, but # adding the `-m68020' flag to GCC prevents building anything better, # like `-m68040'. _LT_TAGVAR(lt_prog_compiler_pic, $1)='-m68020 -resident32 -malways-restore-a4' ;; esac ;; beos* | irix5* | irix6* | nonstopux* | osf3* | osf4* | osf5*) # PIC is the default for these OSes. ;; mingw* | cygwin* | os2* | pw32* | cegcc*) # This hack is so that the source file can tell whether it is being # built for inclusion in a dll (and should export symbols for example). # Although the cygwin gcc ignores -fPIC, still need this for old-style # (--disable-auto-import) libraries m4_if([$1], [GCJ], [], [_LT_TAGVAR(lt_prog_compiler_pic, $1)='-DDLL_EXPORT']) ;; darwin* | rhapsody*) # PIC is the default on this platform # Common symbols not allowed in MH_DYLIB files _LT_TAGVAR(lt_prog_compiler_pic, $1)='-fno-common' ;; *djgpp*) # DJGPP does not support shared libraries at all _LT_TAGVAR(lt_prog_compiler_pic, $1)= ;; haiku*) # PIC is the default for Haiku. # The "-static" flag exists, but is broken. _LT_TAGVAR(lt_prog_compiler_static, $1)= ;; interix[[3-9]]*) # Interix 3.x gcc -fpic/-fPIC options generate broken code. # Instead, we relocate shared libraries at runtime. ;; sysv4*MP*) if test -d /usr/nec; then _LT_TAGVAR(lt_prog_compiler_pic, $1)=-Kconform_pic fi ;; hpux*) # PIC is the default for 64-bit PA HP-UX, but not for 32-bit # PA HP-UX. On IA64 HP-UX, PIC is the default but the pic flag # sets the default TLS model and affects inlining. case $host_cpu in hppa*64*) ;; *) _LT_TAGVAR(lt_prog_compiler_pic, $1)='-fPIC' ;; esac ;; *qnx* | *nto*) # QNX uses GNU C++, but need to define -shared option too, otherwise # it will coredump. _LT_TAGVAR(lt_prog_compiler_pic, $1)='-fPIC -shared' ;; *) _LT_TAGVAR(lt_prog_compiler_pic, $1)='-fPIC' ;; esac else case $host_os in aix[[4-9]]*) # All AIX code is PIC. if test "$host_cpu" = ia64; then # AIX 5 now supports IA64 processor _LT_TAGVAR(lt_prog_compiler_static, $1)='-Bstatic' else _LT_TAGVAR(lt_prog_compiler_static, $1)='-bnso -bI:/lib/syscalls.exp' fi ;; chorus*) case $cc_basename in cxch68*) # Green Hills C++ Compiler # _LT_TAGVAR(lt_prog_compiler_static, $1)="--no_auto_instantiation -u __main -u __premain -u _abort -r $COOL_DIR/lib/libOrb.a $MVME_DIR/lib/CC/libC.a $MVME_DIR/lib/classix/libcx.s.a" ;; esac ;; mingw* | cygwin* | os2* | pw32* | cegcc*) # This hack is so that the source file can tell whether it is being # built for inclusion in a dll (and should export symbols for example). m4_if([$1], [GCJ], [], [_LT_TAGVAR(lt_prog_compiler_pic, $1)='-DDLL_EXPORT']) ;; dgux*) case $cc_basename in ec++*) _LT_TAGVAR(lt_prog_compiler_pic, $1)='-KPIC' ;; ghcx*) # Green Hills C++ Compiler _LT_TAGVAR(lt_prog_compiler_pic, $1)='-pic' ;; *) ;; esac ;; freebsd* | dragonfly*) # FreeBSD uses GNU C++ ;; hpux9* | hpux10* | hpux11*) case $cc_basename in CC*) _LT_TAGVAR(lt_prog_compiler_wl, $1)='-Wl,' _LT_TAGVAR(lt_prog_compiler_static, $1)='${wl}-a ${wl}archive' if test "$host_cpu" != ia64; then _LT_TAGVAR(lt_prog_compiler_pic, $1)='+Z' fi ;; aCC*) _LT_TAGVAR(lt_prog_compiler_wl, $1)='-Wl,' _LT_TAGVAR(lt_prog_compiler_static, $1)='${wl}-a ${wl}archive' case $host_cpu in hppa*64*|ia64*) # +Z the default ;; *) _LT_TAGVAR(lt_prog_compiler_pic, $1)='+Z' ;; esac ;; *) ;; esac ;; interix*) # This is c89, which is MS Visual C++ (no shared libs) # Anyone wants to do a port? ;; irix5* | irix6* | nonstopux*) case $cc_basename in CC*) _LT_TAGVAR(lt_prog_compiler_wl, $1)='-Wl,' _LT_TAGVAR(lt_prog_compiler_static, $1)='-non_shared' # CC pic flag -KPIC is the default. ;; *) ;; esac ;; linux* | k*bsd*-gnu | kopensolaris*-gnu | gnu*) case $cc_basename in KCC*) # KAI C++ Compiler _LT_TAGVAR(lt_prog_compiler_wl, $1)='--backend -Wl,' _LT_TAGVAR(lt_prog_compiler_pic, $1)='-fPIC' ;; ecpc* ) # old Intel C++ for x86_64 which still supported -KPIC. _LT_TAGVAR(lt_prog_compiler_wl, $1)='-Wl,' _LT_TAGVAR(lt_prog_compiler_pic, $1)='-KPIC' _LT_TAGVAR(lt_prog_compiler_static, $1)='-static' ;; icpc* ) # Intel C++, used to be incompatible with GCC. # ICC 10 doesn't accept -KPIC any more. _LT_TAGVAR(lt_prog_compiler_wl, $1)='-Wl,' _LT_TAGVAR(lt_prog_compiler_pic, $1)='-fPIC' _LT_TAGVAR(lt_prog_compiler_static, $1)='-static' ;; pgCC* | pgcpp*) # Portland Group C++ compiler _LT_TAGVAR(lt_prog_compiler_wl, $1)='-Wl,' _LT_TAGVAR(lt_prog_compiler_pic, $1)='-fpic' _LT_TAGVAR(lt_prog_compiler_static, $1)='-Bstatic' ;; cxx*) # Compaq C++ # Make sure the PIC flag is empty. It appears that all Alpha # Linux and Compaq Tru64 Unix objects are PIC. _LT_TAGVAR(lt_prog_compiler_pic, $1)= _LT_TAGVAR(lt_prog_compiler_static, $1)='-non_shared' ;; xlc* | xlC* | bgxl[[cC]]* | mpixl[[cC]]*) # IBM XL 8.0, 9.0 on PPC and BlueGene _LT_TAGVAR(lt_prog_compiler_wl, $1)='-Wl,' _LT_TAGVAR(lt_prog_compiler_pic, $1)='-qpic' _LT_TAGVAR(lt_prog_compiler_static, $1)='-qstaticlink' ;; *) case `$CC -V 2>&1 | sed 5q` in *Sun\ C*) # Sun C++ 5.9 _LT_TAGVAR(lt_prog_compiler_pic, $1)='-KPIC' _LT_TAGVAR(lt_prog_compiler_static, $1)='-Bstatic' _LT_TAGVAR(lt_prog_compiler_wl, $1)='-Qoption ld ' ;; esac ;; esac ;; lynxos*) ;; m88k*) ;; mvs*) case $cc_basename in cxx*) _LT_TAGVAR(lt_prog_compiler_pic, $1)='-W c,exportall' ;; *) ;; esac ;; netbsd* | netbsdelf*-gnu) ;; *qnx* | *nto*) # QNX uses GNU C++, but need to define -shared option too, otherwise # it will coredump. _LT_TAGVAR(lt_prog_compiler_pic, $1)='-fPIC -shared' ;; osf3* | osf4* | osf5*) case $cc_basename in KCC*) _LT_TAGVAR(lt_prog_compiler_wl, $1)='--backend -Wl,' ;; RCC*) # Rational C++ 2.4.1 _LT_TAGVAR(lt_prog_compiler_pic, $1)='-pic' ;; cxx*) # Digital/Compaq C++ _LT_TAGVAR(lt_prog_compiler_wl, $1)='-Wl,' # Make sure the PIC flag is empty. It appears that all Alpha # Linux and Compaq Tru64 Unix objects are PIC. _LT_TAGVAR(lt_prog_compiler_pic, $1)= _LT_TAGVAR(lt_prog_compiler_static, $1)='-non_shared' ;; *) ;; esac ;; psos*) ;; solaris*) case $cc_basename in CC* | sunCC*) # Sun C++ 4.2, 5.x and Centerline C++ _LT_TAGVAR(lt_prog_compiler_pic, $1)='-KPIC' _LT_TAGVAR(lt_prog_compiler_static, $1)='-Bstatic' _LT_TAGVAR(lt_prog_compiler_wl, $1)='-Qoption ld ' ;; gcx*) # Green Hills C++ Compiler _LT_TAGVAR(lt_prog_compiler_pic, $1)='-PIC' ;; *) ;; esac ;; sunos4*) case $cc_basename in CC*) # Sun C++ 4.x _LT_TAGVAR(lt_prog_compiler_pic, $1)='-pic' _LT_TAGVAR(lt_prog_compiler_static, $1)='-Bstatic' ;; lcc*) # Lucid _LT_TAGVAR(lt_prog_compiler_pic, $1)='-pic' ;; *) ;; esac ;; sysv5* | unixware* | sco3.2v5* | sco5v6* | OpenUNIX*) case $cc_basename in CC*) _LT_TAGVAR(lt_prog_compiler_wl, $1)='-Wl,' _LT_TAGVAR(lt_prog_compiler_pic, $1)='-KPIC' _LT_TAGVAR(lt_prog_compiler_static, $1)='-Bstatic' ;; esac ;; tandem*) case $cc_basename in NCC*) # NonStop-UX NCC 3.20 _LT_TAGVAR(lt_prog_compiler_pic, $1)='-KPIC' ;; *) ;; esac ;; vxworks*) ;; *) _LT_TAGVAR(lt_prog_compiler_can_build_shared, $1)=no ;; esac fi ], [ if test "$GCC" = yes; then _LT_TAGVAR(lt_prog_compiler_wl, $1)='-Wl,' _LT_TAGVAR(lt_prog_compiler_static, $1)='-static' case $host_os in aix*) # All AIX code is PIC. if test "$host_cpu" = ia64; then # AIX 5 now supports IA64 processor _LT_TAGVAR(lt_prog_compiler_static, $1)='-Bstatic' fi ;; amigaos*) case $host_cpu in powerpc) # see comment about AmigaOS4 .so support _LT_TAGVAR(lt_prog_compiler_pic, $1)='-fPIC' ;; m68k) # FIXME: we need at least 68020 code to build shared libraries, but # adding the `-m68020' flag to GCC prevents building anything better, # like `-m68040'. _LT_TAGVAR(lt_prog_compiler_pic, $1)='-m68020 -resident32 -malways-restore-a4' ;; esac ;; beos* | irix5* | irix6* | nonstopux* | osf3* | osf4* | osf5*) # PIC is the default for these OSes. ;; mingw* | cygwin* | pw32* | os2* | cegcc*) # This hack is so that the source file can tell whether it is being # built for inclusion in a dll (and should export symbols for example). # Although the cygwin gcc ignores -fPIC, still need this for old-style # (--disable-auto-import) libraries m4_if([$1], [GCJ], [], [_LT_TAGVAR(lt_prog_compiler_pic, $1)='-DDLL_EXPORT']) ;; darwin* | rhapsody*) # PIC is the default on this platform # Common symbols not allowed in MH_DYLIB files _LT_TAGVAR(lt_prog_compiler_pic, $1)='-fno-common' ;; haiku*) # PIC is the default for Haiku. # The "-static" flag exists, but is broken. _LT_TAGVAR(lt_prog_compiler_static, $1)= ;; hpux*) # PIC is the default for 64-bit PA HP-UX, but not for 32-bit # PA HP-UX. On IA64 HP-UX, PIC is the default but the pic flag # sets the default TLS model and affects inlining. case $host_cpu in hppa*64*) # +Z the default ;; *) _LT_TAGVAR(lt_prog_compiler_pic, $1)='-fPIC' ;; esac ;; interix[[3-9]]*) # Interix 3.x gcc -fpic/-fPIC options generate broken code. # Instead, we relocate shared libraries at runtime. ;; msdosdjgpp*) # Just because we use GCC doesn't mean we suddenly get shared libraries # on systems that don't support them. _LT_TAGVAR(lt_prog_compiler_can_build_shared, $1)=no enable_shared=no ;; *nto* | *qnx*) # QNX uses GNU C++, but need to define -shared option too, otherwise # it will coredump. _LT_TAGVAR(lt_prog_compiler_pic, $1)='-fPIC -shared' ;; sysv4*MP*) if test -d /usr/nec; then _LT_TAGVAR(lt_prog_compiler_pic, $1)=-Kconform_pic fi ;; *) _LT_TAGVAR(lt_prog_compiler_pic, $1)='-fPIC' ;; esac case $cc_basename in nvcc*) # Cuda Compiler Driver 2.2 _LT_TAGVAR(lt_prog_compiler_wl, $1)='-Xlinker ' if test -n "$_LT_TAGVAR(lt_prog_compiler_pic, $1)"; then _LT_TAGVAR(lt_prog_compiler_pic, $1)="-Xcompiler $_LT_TAGVAR(lt_prog_compiler_pic, $1)" fi ;; esac else # PORTME Check for flag to pass linker flags through the system compiler. case $host_os in aix*) _LT_TAGVAR(lt_prog_compiler_wl, $1)='-Wl,' if test "$host_cpu" = ia64; then # AIX 5 now supports IA64 processor _LT_TAGVAR(lt_prog_compiler_static, $1)='-Bstatic' else _LT_TAGVAR(lt_prog_compiler_static, $1)='-bnso -bI:/lib/syscalls.exp' fi ;; mingw* | cygwin* | pw32* | os2* | cegcc*) # This hack is so that the source file can tell whether it is being # built for inclusion in a dll (and should export symbols for example). m4_if([$1], [GCJ], [], [_LT_TAGVAR(lt_prog_compiler_pic, $1)='-DDLL_EXPORT']) ;; hpux9* | hpux10* | hpux11*) _LT_TAGVAR(lt_prog_compiler_wl, $1)='-Wl,' # PIC is the default for IA64 HP-UX and 64-bit HP-UX, but # not for PA HP-UX. case $host_cpu in hppa*64*|ia64*) # +Z the default ;; *) _LT_TAGVAR(lt_prog_compiler_pic, $1)='+Z' ;; esac # Is there a better lt_prog_compiler_static that works with the bundled CC? _LT_TAGVAR(lt_prog_compiler_static, $1)='${wl}-a ${wl}archive' ;; irix5* | irix6* | nonstopux*) _LT_TAGVAR(lt_prog_compiler_wl, $1)='-Wl,' # PIC (with -KPIC) is the default. _LT_TAGVAR(lt_prog_compiler_static, $1)='-non_shared' ;; linux* | k*bsd*-gnu | kopensolaris*-gnu | gnu*) case $cc_basename in # old Intel for x86_64 which still supported -KPIC. ecc*) _LT_TAGVAR(lt_prog_compiler_wl, $1)='-Wl,' _LT_TAGVAR(lt_prog_compiler_pic, $1)='-KPIC' _LT_TAGVAR(lt_prog_compiler_static, $1)='-static' ;; # icc used to be incompatible with GCC. # ICC 10 doesn't accept -KPIC any more. icc* | ifort*) _LT_TAGVAR(lt_prog_compiler_wl, $1)='-Wl,' _LT_TAGVAR(lt_prog_compiler_pic, $1)='-fPIC' _LT_TAGVAR(lt_prog_compiler_static, $1)='-static' ;; # Lahey Fortran 8.1. lf95*) _LT_TAGVAR(lt_prog_compiler_wl, $1)='-Wl,' _LT_TAGVAR(lt_prog_compiler_pic, $1)='--shared' _LT_TAGVAR(lt_prog_compiler_static, $1)='--static' ;; nagfor*) # NAG Fortran compiler _LT_TAGVAR(lt_prog_compiler_wl, $1)='-Wl,-Wl,,' _LT_TAGVAR(lt_prog_compiler_pic, $1)='-PIC' _LT_TAGVAR(lt_prog_compiler_static, $1)='-Bstatic' ;; pgcc* | pgf77* | pgf90* | pgf95* | pgfortran*) # Portland Group compilers (*not* the Pentium gcc compiler, # which looks to be a dead project) _LT_TAGVAR(lt_prog_compiler_wl, $1)='-Wl,' _LT_TAGVAR(lt_prog_compiler_pic, $1)='-fpic' _LT_TAGVAR(lt_prog_compiler_static, $1)='-Bstatic' ;; ccc*) _LT_TAGVAR(lt_prog_compiler_wl, $1)='-Wl,' # All Alpha code is PIC. _LT_TAGVAR(lt_prog_compiler_static, $1)='-non_shared' ;; xl* | bgxl* | bgf* | mpixl*) # IBM XL C 8.0/Fortran 10.1, 11.1 on PPC and BlueGene _LT_TAGVAR(lt_prog_compiler_wl, $1)='-Wl,' _LT_TAGVAR(lt_prog_compiler_pic, $1)='-qpic' _LT_TAGVAR(lt_prog_compiler_static, $1)='-qstaticlink' ;; *) case `$CC -V 2>&1 | sed 5q` in *Sun\ Ceres\ Fortran* | *Sun*Fortran*\ [[1-7]].* | *Sun*Fortran*\ 8.[[0-3]]*) # Sun Fortran 8.3 passes all unrecognized flags to the linker _LT_TAGVAR(lt_prog_compiler_pic, $1)='-KPIC' _LT_TAGVAR(lt_prog_compiler_static, $1)='-Bstatic' _LT_TAGVAR(lt_prog_compiler_wl, $1)='' ;; *Sun\ F* | *Sun*Fortran*) _LT_TAGVAR(lt_prog_compiler_pic, $1)='-KPIC' _LT_TAGVAR(lt_prog_compiler_static, $1)='-Bstatic' _LT_TAGVAR(lt_prog_compiler_wl, $1)='-Qoption ld ' ;; *Sun\ C*) # Sun C 5.9 _LT_TAGVAR(lt_prog_compiler_pic, $1)='-KPIC' _LT_TAGVAR(lt_prog_compiler_static, $1)='-Bstatic' _LT_TAGVAR(lt_prog_compiler_wl, $1)='-Wl,' ;; *Intel*\ [[CF]]*Compiler*) _LT_TAGVAR(lt_prog_compiler_wl, $1)='-Wl,' _LT_TAGVAR(lt_prog_compiler_pic, $1)='-fPIC' _LT_TAGVAR(lt_prog_compiler_static, $1)='-static' ;; *Portland\ Group*) _LT_TAGVAR(lt_prog_compiler_wl, $1)='-Wl,' _LT_TAGVAR(lt_prog_compiler_pic, $1)='-fpic' _LT_TAGVAR(lt_prog_compiler_static, $1)='-Bstatic' ;; esac ;; esac ;; newsos6) _LT_TAGVAR(lt_prog_compiler_pic, $1)='-KPIC' _LT_TAGVAR(lt_prog_compiler_static, $1)='-Bstatic' ;; *nto* | *qnx*) # QNX uses GNU C++, but need to define -shared option too, otherwise # it will coredump. _LT_TAGVAR(lt_prog_compiler_pic, $1)='-fPIC -shared' ;; osf3* | osf4* | osf5*) _LT_TAGVAR(lt_prog_compiler_wl, $1)='-Wl,' # All OSF/1 code is PIC. _LT_TAGVAR(lt_prog_compiler_static, $1)='-non_shared' ;; rdos*) _LT_TAGVAR(lt_prog_compiler_static, $1)='-non_shared' ;; solaris*) _LT_TAGVAR(lt_prog_compiler_pic, $1)='-KPIC' _LT_TAGVAR(lt_prog_compiler_static, $1)='-Bstatic' case $cc_basename in f77* | f90* | f95* | sunf77* | sunf90* | sunf95*) _LT_TAGVAR(lt_prog_compiler_wl, $1)='-Qoption ld ';; *) _LT_TAGVAR(lt_prog_compiler_wl, $1)='-Wl,';; esac ;; sunos4*) _LT_TAGVAR(lt_prog_compiler_wl, $1)='-Qoption ld ' _LT_TAGVAR(lt_prog_compiler_pic, $1)='-PIC' _LT_TAGVAR(lt_prog_compiler_static, $1)='-Bstatic' ;; sysv4 | sysv4.2uw2* | sysv4.3*) _LT_TAGVAR(lt_prog_compiler_wl, $1)='-Wl,' _LT_TAGVAR(lt_prog_compiler_pic, $1)='-KPIC' _LT_TAGVAR(lt_prog_compiler_static, $1)='-Bstatic' ;; sysv4*MP*) if test -d /usr/nec ;then _LT_TAGVAR(lt_prog_compiler_pic, $1)='-Kconform_pic' _LT_TAGVAR(lt_prog_compiler_static, $1)='-Bstatic' fi ;; sysv5* | unixware* | sco3.2v5* | sco5v6* | OpenUNIX*) _LT_TAGVAR(lt_prog_compiler_wl, $1)='-Wl,' _LT_TAGVAR(lt_prog_compiler_pic, $1)='-KPIC' _LT_TAGVAR(lt_prog_compiler_static, $1)='-Bstatic' ;; unicos*) _LT_TAGVAR(lt_prog_compiler_wl, $1)='-Wl,' _LT_TAGVAR(lt_prog_compiler_can_build_shared, $1)=no ;; uts4*) _LT_TAGVAR(lt_prog_compiler_pic, $1)='-pic' _LT_TAGVAR(lt_prog_compiler_static, $1)='-Bstatic' ;; *) _LT_TAGVAR(lt_prog_compiler_can_build_shared, $1)=no ;; esac fi ]) case $host_os in # For platforms which do not support PIC, -DPIC is meaningless: *djgpp*) _LT_TAGVAR(lt_prog_compiler_pic, $1)= ;; *) _LT_TAGVAR(lt_prog_compiler_pic, $1)="$_LT_TAGVAR(lt_prog_compiler_pic, $1)@&t@m4_if([$1],[],[ -DPIC],[m4_if([$1],[CXX],[ -DPIC],[])])" ;; esac AC_CACHE_CHECK([for $compiler option to produce PIC], [_LT_TAGVAR(lt_cv_prog_compiler_pic, $1)], [_LT_TAGVAR(lt_cv_prog_compiler_pic, $1)=$_LT_TAGVAR(lt_prog_compiler_pic, $1)]) _LT_TAGVAR(lt_prog_compiler_pic, $1)=$_LT_TAGVAR(lt_cv_prog_compiler_pic, $1) # # Check to make sure the PIC flag actually works. # if test -n "$_LT_TAGVAR(lt_prog_compiler_pic, $1)"; then _LT_COMPILER_OPTION([if $compiler PIC flag $_LT_TAGVAR(lt_prog_compiler_pic, $1) works], [_LT_TAGVAR(lt_cv_prog_compiler_pic_works, $1)], [$_LT_TAGVAR(lt_prog_compiler_pic, $1)@&t@m4_if([$1],[],[ -DPIC],[m4_if([$1],[CXX],[ -DPIC],[])])], [], [case $_LT_TAGVAR(lt_prog_compiler_pic, $1) in "" | " "*) ;; *) _LT_TAGVAR(lt_prog_compiler_pic, $1)=" $_LT_TAGVAR(lt_prog_compiler_pic, $1)" ;; esac], [_LT_TAGVAR(lt_prog_compiler_pic, $1)= _LT_TAGVAR(lt_prog_compiler_can_build_shared, $1)=no]) fi _LT_TAGDECL([pic_flag], [lt_prog_compiler_pic], [1], [Additional compiler flags for building library objects]) _LT_TAGDECL([wl], [lt_prog_compiler_wl], [1], [How to pass a linker flag through the compiler]) # # Check to make sure the static flag actually works. # wl=$_LT_TAGVAR(lt_prog_compiler_wl, $1) eval lt_tmp_static_flag=\"$_LT_TAGVAR(lt_prog_compiler_static, $1)\" _LT_LINKER_OPTION([if $compiler static flag $lt_tmp_static_flag works], _LT_TAGVAR(lt_cv_prog_compiler_static_works, $1), $lt_tmp_static_flag, [], [_LT_TAGVAR(lt_prog_compiler_static, $1)=]) _LT_TAGDECL([link_static_flag], [lt_prog_compiler_static], [1], [Compiler flag to prevent dynamic linking]) ])# _LT_COMPILER_PIC # _LT_LINKER_SHLIBS([TAGNAME]) # ---------------------------- # See if the linker supports building shared libraries. m4_defun([_LT_LINKER_SHLIBS], [AC_REQUIRE([LT_PATH_LD])dnl AC_REQUIRE([LT_PATH_NM])dnl m4_require([_LT_PATH_MANIFEST_TOOL])dnl m4_require([_LT_FILEUTILS_DEFAULTS])dnl m4_require([_LT_DECL_EGREP])dnl m4_require([_LT_DECL_SED])dnl m4_require([_LT_CMD_GLOBAL_SYMBOLS])dnl m4_require([_LT_TAG_COMPILER])dnl AC_MSG_CHECKING([whether the $compiler linker ($LD) supports shared libraries]) m4_if([$1], [CXX], [ _LT_TAGVAR(export_symbols_cmds, $1)='$NM $libobjs $convenience | $global_symbol_pipe | $SED '\''s/.* //'\'' | sort | uniq > $export_symbols' _LT_TAGVAR(exclude_expsyms, $1)=['_GLOBAL_OFFSET_TABLE_|_GLOBAL__F[ID]_.*'] case $host_os in aix[[4-9]]*) # If we're using GNU nm, then we don't want the "-C" option. # -C means demangle to AIX nm, but means don't demangle with GNU nm # Also, AIX nm treats weak defined symbols like other global defined # symbols, whereas GNU nm marks them as "W". if $NM -V 2>&1 | $GREP 'GNU' > /dev/null; then _LT_TAGVAR(export_symbols_cmds, $1)='$NM -Bpg $libobjs $convenience | awk '\''{ if (((\$ 2 == "T") || (\$ 2 == "D") || (\$ 2 == "B") || (\$ 2 == "W")) && ([substr](\$ 3,1,1) != ".")) { print \$ 3 } }'\'' | sort -u > $export_symbols' else _LT_TAGVAR(export_symbols_cmds, $1)='$NM -BCpg $libobjs $convenience | awk '\''{ if (((\$ 2 == "T") || (\$ 2 == "D") || (\$ 2 == "B")) && ([substr](\$ 3,1,1) != ".")) { print \$ 3 } }'\'' | sort -u > $export_symbols' fi ;; pw32*) _LT_TAGVAR(export_symbols_cmds, $1)="$ltdll_cmds" ;; cygwin* | mingw* | cegcc*) case $cc_basename in cl*) _LT_TAGVAR(exclude_expsyms, $1)='_NULL_IMPORT_DESCRIPTOR|_IMPORT_DESCRIPTOR_.*' ;; *) _LT_TAGVAR(export_symbols_cmds, $1)='$NM $libobjs $convenience | $global_symbol_pipe | $SED -e '\''/^[[BCDGRS]][[ ]]/s/.*[[ ]]\([[^ ]]*\)/\1 DATA/;s/^.*[[ ]]__nm__\([[^ ]]*\)[[ ]][[^ ]]*/\1 DATA/;/^I[[ ]]/d;/^[[AITW]][[ ]]/s/.* //'\'' | sort | uniq > $export_symbols' _LT_TAGVAR(exclude_expsyms, $1)=['[_]+GLOBAL_OFFSET_TABLE_|[_]+GLOBAL__[FID]_.*|[_]+head_[A-Za-z0-9_]+_dll|[A-Za-z0-9_]+_dll_iname'] ;; esac ;; linux* | k*bsd*-gnu | gnu*) _LT_TAGVAR(link_all_deplibs, $1)=no ;; *) _LT_TAGVAR(export_symbols_cmds, $1)='$NM $libobjs $convenience | $global_symbol_pipe | $SED '\''s/.* //'\'' | sort | uniq > $export_symbols' ;; esac ], [ runpath_var= _LT_TAGVAR(allow_undefined_flag, $1)= _LT_TAGVAR(always_export_symbols, $1)=no _LT_TAGVAR(archive_cmds, $1)= _LT_TAGVAR(archive_expsym_cmds, $1)= _LT_TAGVAR(compiler_needs_object, $1)=no _LT_TAGVAR(enable_shared_with_static_runtimes, $1)=no _LT_TAGVAR(export_dynamic_flag_spec, $1)= _LT_TAGVAR(export_symbols_cmds, $1)='$NM $libobjs $convenience | $global_symbol_pipe | $SED '\''s/.* //'\'' | sort | uniq > $export_symbols' _LT_TAGVAR(hardcode_automatic, $1)=no _LT_TAGVAR(hardcode_direct, $1)=no _LT_TAGVAR(hardcode_direct_absolute, $1)=no _LT_TAGVAR(hardcode_libdir_flag_spec, $1)= _LT_TAGVAR(hardcode_libdir_separator, $1)= _LT_TAGVAR(hardcode_minus_L, $1)=no _LT_TAGVAR(hardcode_shlibpath_var, $1)=unsupported _LT_TAGVAR(inherit_rpath, $1)=no _LT_TAGVAR(link_all_deplibs, $1)=unknown _LT_TAGVAR(module_cmds, $1)= _LT_TAGVAR(module_expsym_cmds, $1)= _LT_TAGVAR(old_archive_from_new_cmds, $1)= _LT_TAGVAR(old_archive_from_expsyms_cmds, $1)= _LT_TAGVAR(thread_safe_flag_spec, $1)= _LT_TAGVAR(whole_archive_flag_spec, $1)= # include_expsyms should be a list of space-separated symbols to be *always* # included in the symbol list _LT_TAGVAR(include_expsyms, $1)= # exclude_expsyms can be an extended regexp of symbols to exclude # it will be wrapped by ` (' and `)$', so one must not match beginning or # end of line. Example: `a|bc|.*d.*' will exclude the symbols `a' and `bc', # as well as any symbol that contains `d'. _LT_TAGVAR(exclude_expsyms, $1)=['_GLOBAL_OFFSET_TABLE_|_GLOBAL__F[ID]_.*'] # Although _GLOBAL_OFFSET_TABLE_ is a valid symbol C name, most a.out # platforms (ab)use it in PIC code, but their linkers get confused if # the symbol is explicitly referenced. Since portable code cannot # rely on this symbol name, it's probably fine to never include it in # preloaded symbol tables. # Exclude shared library initialization/finalization symbols. dnl Note also adjust exclude_expsyms for C++ above. extract_expsyms_cmds= case $host_os in cygwin* | mingw* | pw32* | cegcc*) # FIXME: the MSVC++ port hasn't been tested in a loooong time # When not using gcc, we currently assume that we are using # Microsoft Visual C++. if test "$GCC" != yes; then with_gnu_ld=no fi ;; interix*) # we just hope/assume this is gcc and not c89 (= MSVC++) with_gnu_ld=yes ;; openbsd*) with_gnu_ld=no ;; linux* | k*bsd*-gnu | gnu*) _LT_TAGVAR(link_all_deplibs, $1)=no ;; esac _LT_TAGVAR(ld_shlibs, $1)=yes # On some targets, GNU ld is compatible enough with the native linker # that we're better off using the native interface for both. lt_use_gnu_ld_interface=no if test "$with_gnu_ld" = yes; then case $host_os in aix*) # The AIX port of GNU ld has always aspired to compatibility # with the native linker. However, as the warning in the GNU ld # block says, versions before 2.19.5* couldn't really create working # shared libraries, regardless of the interface used. case `$LD -v 2>&1` in *\ \(GNU\ Binutils\)\ 2.19.5*) ;; *\ \(GNU\ Binutils\)\ 2.[[2-9]]*) ;; *\ \(GNU\ Binutils\)\ [[3-9]]*) ;; *) lt_use_gnu_ld_interface=yes ;; esac ;; *) lt_use_gnu_ld_interface=yes ;; esac fi if test "$lt_use_gnu_ld_interface" = yes; then # If archive_cmds runs LD, not CC, wlarc should be empty wlarc='${wl}' # Set some defaults for GNU ld with shared library support. These # are reset later if shared libraries are not supported. Putting them # here allows them to be overridden if necessary. runpath_var=LD_RUN_PATH _LT_TAGVAR(hardcode_libdir_flag_spec, $1)='${wl}-rpath ${wl}$libdir' _LT_TAGVAR(export_dynamic_flag_spec, $1)='${wl}--export-dynamic' # ancient GNU ld didn't support --whole-archive et. al. if $LD --help 2>&1 | $GREP 'no-whole-archive' > /dev/null; then _LT_TAGVAR(whole_archive_flag_spec, $1)="$wlarc"'--whole-archive$convenience '"$wlarc"'--no-whole-archive' else _LT_TAGVAR(whole_archive_flag_spec, $1)= fi supports_anon_versioning=no case `$LD -v 2>&1` in *GNU\ gold*) supports_anon_versioning=yes ;; *\ [[01]].* | *\ 2.[[0-9]].* | *\ 2.10.*) ;; # catch versions < 2.11 *\ 2.11.93.0.2\ *) supports_anon_versioning=yes ;; # RH7.3 ... *\ 2.11.92.0.12\ *) supports_anon_versioning=yes ;; # Mandrake 8.2 ... *\ 2.11.*) ;; # other 2.11 versions *) supports_anon_versioning=yes ;; esac # See if GNU ld supports shared libraries. case $host_os in aix[[3-9]]*) # On AIX/PPC, the GNU linker is very broken if test "$host_cpu" != ia64; then _LT_TAGVAR(ld_shlibs, $1)=no cat <<_LT_EOF 1>&2 *** Warning: the GNU linker, at least up to release 2.19, is reported *** to be unable to reliably create shared libraries on AIX. *** Therefore, libtool is disabling shared libraries support. If you *** really care for shared libraries, you may want to install binutils *** 2.20 or above, or modify your PATH so that a non-GNU linker is found. *** You will then need to restart the configuration process. _LT_EOF fi ;; amigaos*) case $host_cpu in powerpc) # see comment about AmigaOS4 .so support _LT_TAGVAR(archive_cmds, $1)='$CC -shared $libobjs $deplibs $compiler_flags ${wl}-soname $wl$soname -o $lib' _LT_TAGVAR(archive_expsym_cmds, $1)='' ;; m68k) _LT_TAGVAR(archive_cmds, $1)='$RM $output_objdir/a2ixlibrary.data~$ECHO "#define NAME $libname" > $output_objdir/a2ixlibrary.data~$ECHO "#define LIBRARY_ID 1" >> $output_objdir/a2ixlibrary.data~$ECHO "#define VERSION $major" >> $output_objdir/a2ixlibrary.data~$ECHO "#define REVISION $revision" >> $output_objdir/a2ixlibrary.data~$AR $AR_FLAGS $lib $libobjs~$RANLIB $lib~(cd $output_objdir && a2ixlibrary -32)' _LT_TAGVAR(hardcode_libdir_flag_spec, $1)='-L$libdir' _LT_TAGVAR(hardcode_minus_L, $1)=yes ;; esac ;; beos*) if $LD --help 2>&1 | $GREP ': supported targets:.* elf' > /dev/null; then _LT_TAGVAR(allow_undefined_flag, $1)=unsupported # Joseph Beckenbach says some releases of gcc # support --undefined. This deserves some investigation. FIXME _LT_TAGVAR(archive_cmds, $1)='$CC -nostart $libobjs $deplibs $compiler_flags ${wl}-soname $wl$soname -o $lib' else _LT_TAGVAR(ld_shlibs, $1)=no fi ;; cygwin* | mingw* | pw32* | cegcc*) # _LT_TAGVAR(hardcode_libdir_flag_spec, $1) is actually meaningless, # as there is no search path for DLLs. _LT_TAGVAR(hardcode_libdir_flag_spec, $1)='-L$libdir' _LT_TAGVAR(export_dynamic_flag_spec, $1)='${wl}--export-all-symbols' _LT_TAGVAR(allow_undefined_flag, $1)=unsupported _LT_TAGVAR(always_export_symbols, $1)=no _LT_TAGVAR(enable_shared_with_static_runtimes, $1)=yes _LT_TAGVAR(export_symbols_cmds, $1)='$NM $libobjs $convenience | $global_symbol_pipe | $SED -e '\''/^[[BCDGRS]][[ ]]/s/.*[[ ]]\([[^ ]]*\)/\1 DATA/;s/^.*[[ ]]__nm__\([[^ ]]*\)[[ ]][[^ ]]*/\1 DATA/;/^I[[ ]]/d;/^[[AITW]][[ ]]/s/.* //'\'' | sort | uniq > $export_symbols' _LT_TAGVAR(exclude_expsyms, $1)=['[_]+GLOBAL_OFFSET_TABLE_|[_]+GLOBAL__[FID]_.*|[_]+head_[A-Za-z0-9_]+_dll|[A-Za-z0-9_]+_dll_iname'] if $LD --help 2>&1 | $GREP 'auto-import' > /dev/null; then _LT_TAGVAR(archive_cmds, $1)='$CC -shared $libobjs $deplibs $compiler_flags -o $output_objdir/$soname ${wl}--enable-auto-image-base -Xlinker --out-implib -Xlinker $lib' # If the export-symbols file already is a .def file (1st line # is EXPORTS), use it as is; otherwise, prepend... _LT_TAGVAR(archive_expsym_cmds, $1)='if test "x`$SED 1q $export_symbols`" = xEXPORTS; then cp $export_symbols $output_objdir/$soname.def; else echo EXPORTS > $output_objdir/$soname.def; cat $export_symbols >> $output_objdir/$soname.def; fi~ $CC -shared $output_objdir/$soname.def $libobjs $deplibs $compiler_flags -o $output_objdir/$soname ${wl}--enable-auto-image-base -Xlinker --out-implib -Xlinker $lib' else _LT_TAGVAR(ld_shlibs, $1)=no fi ;; haiku*) _LT_TAGVAR(archive_cmds, $1)='$CC -shared $libobjs $deplibs $compiler_flags ${wl}-soname $wl$soname -o $lib' _LT_TAGVAR(link_all_deplibs, $1)=yes ;; interix[[3-9]]*) _LT_TAGVAR(hardcode_direct, $1)=no _LT_TAGVAR(hardcode_shlibpath_var, $1)=no _LT_TAGVAR(hardcode_libdir_flag_spec, $1)='${wl}-rpath,$libdir' _LT_TAGVAR(export_dynamic_flag_spec, $1)='${wl}-E' # Hack: On Interix 3.x, we cannot compile PIC because of a broken gcc. # Instead, shared libraries are loaded at an image base (0x10000000 by # default) and relocated if they conflict, which is a slow very memory # consuming and fragmenting process. To avoid this, we pick a random, # 256 KiB-aligned image base between 0x50000000 and 0x6FFC0000 at link # time. Moving up from 0x10000000 also allows more sbrk(2) space. _LT_TAGVAR(archive_cmds, $1)='$CC -shared $pic_flag $libobjs $deplibs $compiler_flags ${wl}-h,$soname ${wl}--image-base,`expr ${RANDOM-$$} % 4096 / 2 \* 262144 + 1342177280` -o $lib' _LT_TAGVAR(archive_expsym_cmds, $1)='sed "s,^,_," $export_symbols >$output_objdir/$soname.expsym~$CC -shared $pic_flag $libobjs $deplibs $compiler_flags ${wl}-h,$soname ${wl}--retain-symbols-file,$output_objdir/$soname.expsym ${wl}--image-base,`expr ${RANDOM-$$} % 4096 / 2 \* 262144 + 1342177280` -o $lib' ;; gnu* | linux* | tpf* | k*bsd*-gnu | kopensolaris*-gnu) tmp_diet=no if test "$host_os" = linux-dietlibc; then case $cc_basename in diet\ *) tmp_diet=yes;; # linux-dietlibc with static linking (!diet-dyn) esac fi if $LD --help 2>&1 | $EGREP ': supported targets:.* elf' > /dev/null \ && test "$tmp_diet" = no then tmp_addflag=' $pic_flag' tmp_sharedflag='-shared' case $cc_basename,$host_cpu in pgcc*) # Portland Group C compiler _LT_TAGVAR(whole_archive_flag_spec, $1)='${wl}--whole-archive`for conv in $convenience\"\"; do test -n \"$conv\" && new_convenience=\"$new_convenience,$conv\"; done; func_echo_all \"$new_convenience\"` ${wl}--no-whole-archive' tmp_addflag=' $pic_flag' ;; pgf77* | pgf90* | pgf95* | pgfortran*) # Portland Group f77 and f90 compilers _LT_TAGVAR(whole_archive_flag_spec, $1)='${wl}--whole-archive`for conv in $convenience\"\"; do test -n \"$conv\" && new_convenience=\"$new_convenience,$conv\"; done; func_echo_all \"$new_convenience\"` ${wl}--no-whole-archive' tmp_addflag=' $pic_flag -Mnomain' ;; ecc*,ia64* | icc*,ia64*) # Intel C compiler on ia64 tmp_addflag=' -i_dynamic' ;; efc*,ia64* | ifort*,ia64*) # Intel Fortran compiler on ia64 tmp_addflag=' -i_dynamic -nofor_main' ;; ifc* | ifort*) # Intel Fortran compiler tmp_addflag=' -nofor_main' ;; lf95*) # Lahey Fortran 8.1 _LT_TAGVAR(whole_archive_flag_spec, $1)= tmp_sharedflag='--shared' ;; xl[[cC]]* | bgxl[[cC]]* | mpixl[[cC]]*) # IBM XL C 8.0 on PPC (deal with xlf below) tmp_sharedflag='-qmkshrobj' tmp_addflag= ;; nvcc*) # Cuda Compiler Driver 2.2 _LT_TAGVAR(whole_archive_flag_spec, $1)='${wl}--whole-archive`for conv in $convenience\"\"; do test -n \"$conv\" && new_convenience=\"$new_convenience,$conv\"; done; func_echo_all \"$new_convenience\"` ${wl}--no-whole-archive' _LT_TAGVAR(compiler_needs_object, $1)=yes ;; esac case `$CC -V 2>&1 | sed 5q` in *Sun\ C*) # Sun C 5.9 _LT_TAGVAR(whole_archive_flag_spec, $1)='${wl}--whole-archive`new_convenience=; for conv in $convenience\"\"; do test -z \"$conv\" || new_convenience=\"$new_convenience,$conv\"; done; func_echo_all \"$new_convenience\"` ${wl}--no-whole-archive' _LT_TAGVAR(compiler_needs_object, $1)=yes tmp_sharedflag='-G' ;; *Sun\ F*) # Sun Fortran 8.3 tmp_sharedflag='-G' ;; esac _LT_TAGVAR(archive_cmds, $1)='$CC '"$tmp_sharedflag""$tmp_addflag"' $libobjs $deplibs $compiler_flags ${wl}-soname $wl$soname -o $lib' if test "x$supports_anon_versioning" = xyes; then _LT_TAGVAR(archive_expsym_cmds, $1)='echo "{ global:" > $output_objdir/$libname.ver~ cat $export_symbols | sed -e "s/\(.*\)/\1;/" >> $output_objdir/$libname.ver~ echo "local: *; };" >> $output_objdir/$libname.ver~ $CC '"$tmp_sharedflag""$tmp_addflag"' $libobjs $deplibs $compiler_flags ${wl}-soname $wl$soname ${wl}-version-script ${wl}$output_objdir/$libname.ver -o $lib' fi case $cc_basename in xlf* | bgf* | bgxlf* | mpixlf*) # IBM XL Fortran 10.1 on PPC cannot create shared libs itself _LT_TAGVAR(whole_archive_flag_spec, $1)='--whole-archive$convenience --no-whole-archive' _LT_TAGVAR(hardcode_libdir_flag_spec, $1)='${wl}-rpath ${wl}$libdir' _LT_TAGVAR(archive_cmds, $1)='$LD -shared $libobjs $deplibs $linker_flags -soname $soname -o $lib' if test "x$supports_anon_versioning" = xyes; then _LT_TAGVAR(archive_expsym_cmds, $1)='echo "{ global:" > $output_objdir/$libname.ver~ cat $export_symbols | sed -e "s/\(.*\)/\1;/" >> $output_objdir/$libname.ver~ echo "local: *; };" >> $output_objdir/$libname.ver~ $LD -shared $libobjs $deplibs $linker_flags -soname $soname -version-script $output_objdir/$libname.ver -o $lib' fi ;; esac else _LT_TAGVAR(ld_shlibs, $1)=no fi ;; netbsd* | netbsdelf*-gnu) if echo __ELF__ | $CC -E - | $GREP __ELF__ >/dev/null; then _LT_TAGVAR(archive_cmds, $1)='$LD -Bshareable $libobjs $deplibs $linker_flags -o $lib' wlarc= else _LT_TAGVAR(archive_cmds, $1)='$CC -shared $pic_flag $libobjs $deplibs $compiler_flags ${wl}-soname $wl$soname -o $lib' _LT_TAGVAR(archive_expsym_cmds, $1)='$CC -shared $pic_flag $libobjs $deplibs $compiler_flags ${wl}-soname $wl$soname ${wl}-retain-symbols-file $wl$export_symbols -o $lib' fi ;; solaris*) if $LD -v 2>&1 | $GREP 'BFD 2\.8' > /dev/null; then _LT_TAGVAR(ld_shlibs, $1)=no cat <<_LT_EOF 1>&2 *** Warning: The releases 2.8.* of the GNU linker cannot reliably *** create shared libraries on Solaris systems. Therefore, libtool *** is disabling shared libraries support. We urge you to upgrade GNU *** binutils to release 2.9.1 or newer. Another option is to modify *** your PATH or compiler configuration so that the native linker is *** used, and then restart. _LT_EOF elif $LD --help 2>&1 | $GREP ': supported targets:.* elf' > /dev/null; then _LT_TAGVAR(archive_cmds, $1)='$CC -shared $pic_flag $libobjs $deplibs $compiler_flags ${wl}-soname $wl$soname -o $lib' _LT_TAGVAR(archive_expsym_cmds, $1)='$CC -shared $pic_flag $libobjs $deplibs $compiler_flags ${wl}-soname $wl$soname ${wl}-retain-symbols-file $wl$export_symbols -o $lib' else _LT_TAGVAR(ld_shlibs, $1)=no fi ;; sysv5* | sco3.2v5* | sco5v6* | unixware* | OpenUNIX*) case `$LD -v 2>&1` in *\ [[01]].* | *\ 2.[[0-9]].* | *\ 2.1[[0-5]].*) _LT_TAGVAR(ld_shlibs, $1)=no cat <<_LT_EOF 1>&2 *** Warning: Releases of the GNU linker prior to 2.16.91.0.3 can not *** reliably create shared libraries on SCO systems. Therefore, libtool *** is disabling shared libraries support. We urge you to upgrade GNU *** binutils to release 2.16.91.0.3 or newer. Another option is to modify *** your PATH or compiler configuration so that the native linker is *** used, and then restart. _LT_EOF ;; *) # For security reasons, it is highly recommended that you always # use absolute paths for naming shared libraries, and exclude the # DT_RUNPATH tag from executables and libraries. But doing so # requires that you compile everything twice, which is a pain. if $LD --help 2>&1 | $GREP ': supported targets:.* elf' > /dev/null; then _LT_TAGVAR(hardcode_libdir_flag_spec, $1)='${wl}-rpath ${wl}$libdir' _LT_TAGVAR(archive_cmds, $1)='$CC -shared $libobjs $deplibs $compiler_flags ${wl}-soname $wl$soname -o $lib' _LT_TAGVAR(archive_expsym_cmds, $1)='$CC -shared $libobjs $deplibs $compiler_flags ${wl}-soname $wl$soname ${wl}-retain-symbols-file $wl$export_symbols -o $lib' else _LT_TAGVAR(ld_shlibs, $1)=no fi ;; esac ;; sunos4*) _LT_TAGVAR(archive_cmds, $1)='$LD -assert pure-text -Bshareable -o $lib $libobjs $deplibs $linker_flags' wlarc= _LT_TAGVAR(hardcode_direct, $1)=yes _LT_TAGVAR(hardcode_shlibpath_var, $1)=no ;; *) if $LD --help 2>&1 | $GREP ': supported targets:.* elf' > /dev/null; then _LT_TAGVAR(archive_cmds, $1)='$CC -shared $pic_flag $libobjs $deplibs $compiler_flags ${wl}-soname $wl$soname -o $lib' _LT_TAGVAR(archive_expsym_cmds, $1)='$CC -shared $pic_flag $libobjs $deplibs $compiler_flags ${wl}-soname $wl$soname ${wl}-retain-symbols-file $wl$export_symbols -o $lib' else _LT_TAGVAR(ld_shlibs, $1)=no fi ;; esac if test "$_LT_TAGVAR(ld_shlibs, $1)" = no; then runpath_var= _LT_TAGVAR(hardcode_libdir_flag_spec, $1)= _LT_TAGVAR(export_dynamic_flag_spec, $1)= _LT_TAGVAR(whole_archive_flag_spec, $1)= fi else # PORTME fill in a description of your system's linker (not GNU ld) case $host_os in aix3*) _LT_TAGVAR(allow_undefined_flag, $1)=unsupported _LT_TAGVAR(always_export_symbols, $1)=yes _LT_TAGVAR(archive_expsym_cmds, $1)='$LD -o $output_objdir/$soname $libobjs $deplibs $linker_flags -bE:$export_symbols -T512 -H512 -bM:SRE~$AR $AR_FLAGS $lib $output_objdir/$soname' # Note: this linker hardcodes the directories in LIBPATH if there # are no directories specified by -L. _LT_TAGVAR(hardcode_minus_L, $1)=yes if test "$GCC" = yes && test -z "$lt_prog_compiler_static"; then # Neither direct hardcoding nor static linking is supported with a # broken collect2. _LT_TAGVAR(hardcode_direct, $1)=unsupported fi ;; aix[[4-9]]*) if test "$host_cpu" = ia64; then # On IA64, the linker does run time linking by default, so we don't # have to do anything special. aix_use_runtimelinking=no exp_sym_flag='-Bexport' no_entry_flag="" else # If we're using GNU nm, then we don't want the "-C" option. # -C means demangle to AIX nm, but means don't demangle with GNU nm # Also, AIX nm treats weak defined symbols like other global # defined symbols, whereas GNU nm marks them as "W". if $NM -V 2>&1 | $GREP 'GNU' > /dev/null; then _LT_TAGVAR(export_symbols_cmds, $1)='$NM -Bpg $libobjs $convenience | awk '\''{ if (((\$ 2 == "T") || (\$ 2 == "D") || (\$ 2 == "B") || (\$ 2 == "W")) && ([substr](\$ 3,1,1) != ".")) { print \$ 3 } }'\'' | sort -u > $export_symbols' else _LT_TAGVAR(export_symbols_cmds, $1)='$NM -BCpg $libobjs $convenience | awk '\''{ if (((\$ 2 == "T") || (\$ 2 == "D") || (\$ 2 == "B")) && ([substr](\$ 3,1,1) != ".")) { print \$ 3 } }'\'' | sort -u > $export_symbols' fi aix_use_runtimelinking=no # Test if we are trying to use run time linking or normal # AIX style linking. If -brtl is somewhere in LDFLAGS, we # need to do runtime linking. case $host_os in aix4.[[23]]|aix4.[[23]].*|aix[[5-9]]*) for ld_flag in $LDFLAGS; do if (test $ld_flag = "-brtl" || test $ld_flag = "-Wl,-brtl"); then aix_use_runtimelinking=yes break fi done ;; esac exp_sym_flag='-bexport' no_entry_flag='-bnoentry' fi # When large executables or shared objects are built, AIX ld can # have problems creating the table of contents. If linking a library # or program results in "error TOC overflow" add -mminimal-toc to # CXXFLAGS/CFLAGS for g++/gcc. In the cases where that is not # enough to fix the problem, add -Wl,-bbigtoc to LDFLAGS. _LT_TAGVAR(archive_cmds, $1)='' _LT_TAGVAR(hardcode_direct, $1)=yes _LT_TAGVAR(hardcode_direct_absolute, $1)=yes _LT_TAGVAR(hardcode_libdir_separator, $1)=':' _LT_TAGVAR(link_all_deplibs, $1)=yes _LT_TAGVAR(file_list_spec, $1)='${wl}-f,' if test "$GCC" = yes; then case $host_os in aix4.[[012]]|aix4.[[012]].*) # We only want to do this on AIX 4.2 and lower, the check # below for broken collect2 doesn't work under 4.3+ collect2name=`${CC} -print-prog-name=collect2` if test -f "$collect2name" && strings "$collect2name" | $GREP resolve_lib_name >/dev/null then # We have reworked collect2 : else # We have old collect2 _LT_TAGVAR(hardcode_direct, $1)=unsupported # It fails to find uninstalled libraries when the uninstalled # path is not listed in the libpath. Setting hardcode_minus_L # to unsupported forces relinking _LT_TAGVAR(hardcode_minus_L, $1)=yes _LT_TAGVAR(hardcode_libdir_flag_spec, $1)='-L$libdir' _LT_TAGVAR(hardcode_libdir_separator, $1)= fi ;; esac shared_flag='-shared' if test "$aix_use_runtimelinking" = yes; then shared_flag="$shared_flag "'${wl}-G' fi _LT_TAGVAR(link_all_deplibs, $1)=no else # not using gcc if test "$host_cpu" = ia64; then # VisualAge C++, Version 5.5 for AIX 5L for IA-64, Beta 3 Release # chokes on -Wl,-G. The following line is correct: shared_flag='-G' else if test "$aix_use_runtimelinking" = yes; then shared_flag='${wl}-G' else shared_flag='${wl}-bM:SRE' fi fi fi _LT_TAGVAR(export_dynamic_flag_spec, $1)='${wl}-bexpall' # It seems that -bexpall does not export symbols beginning with # underscore (_), so it is better to generate a list of symbols to export. _LT_TAGVAR(always_export_symbols, $1)=yes if test "$aix_use_runtimelinking" = yes; then # Warning - without using the other runtime loading flags (-brtl), # -berok will link without error, but may produce a broken library. _LT_TAGVAR(allow_undefined_flag, $1)='-berok' # Determine the default libpath from the value encoded in an # empty executable. _LT_SYS_MODULE_PATH_AIX([$1]) _LT_TAGVAR(hardcode_libdir_flag_spec, $1)='${wl}-blibpath:$libdir:'"$aix_libpath" _LT_TAGVAR(archive_expsym_cmds, $1)='$CC -o $output_objdir/$soname $libobjs $deplibs '"\${wl}$no_entry_flag"' $compiler_flags `if test "x${allow_undefined_flag}" != "x"; then func_echo_all "${wl}${allow_undefined_flag}"; else :; fi` '"\${wl}$exp_sym_flag:\$export_symbols $shared_flag" else if test "$host_cpu" = ia64; then _LT_TAGVAR(hardcode_libdir_flag_spec, $1)='${wl}-R $libdir:/usr/lib:/lib' _LT_TAGVAR(allow_undefined_flag, $1)="-z nodefs" _LT_TAGVAR(archive_expsym_cmds, $1)="\$CC $shared_flag"' -o $output_objdir/$soname $libobjs $deplibs '"\${wl}$no_entry_flag"' $compiler_flags ${wl}${allow_undefined_flag} '"\${wl}$exp_sym_flag:\$export_symbols" else # Determine the default libpath from the value encoded in an # empty executable. _LT_SYS_MODULE_PATH_AIX([$1]) _LT_TAGVAR(hardcode_libdir_flag_spec, $1)='${wl}-blibpath:$libdir:'"$aix_libpath" # Warning - without using the other run time loading flags, # -berok will link without error, but may produce a broken library. _LT_TAGVAR(no_undefined_flag, $1)=' ${wl}-bernotok' _LT_TAGVAR(allow_undefined_flag, $1)=' ${wl}-berok' if test "$with_gnu_ld" = yes; then # We only use this code for GNU lds that support --whole-archive. _LT_TAGVAR(whole_archive_flag_spec, $1)='${wl}--whole-archive$convenience ${wl}--no-whole-archive' else # Exported symbols can be pulled into shared objects from archives _LT_TAGVAR(whole_archive_flag_spec, $1)='$convenience' fi _LT_TAGVAR(archive_cmds_need_lc, $1)=yes # This is similar to how AIX traditionally builds its shared libraries. _LT_TAGVAR(archive_expsym_cmds, $1)="\$CC $shared_flag"' -o $output_objdir/$soname $libobjs $deplibs ${wl}-bnoentry $compiler_flags ${wl}-bE:$export_symbols${allow_undefined_flag}~$AR $AR_FLAGS $output_objdir/$libname$release.a $output_objdir/$soname' fi fi ;; amigaos*) case $host_cpu in powerpc) # see comment about AmigaOS4 .so support _LT_TAGVAR(archive_cmds, $1)='$CC -shared $libobjs $deplibs $compiler_flags ${wl}-soname $wl$soname -o $lib' _LT_TAGVAR(archive_expsym_cmds, $1)='' ;; m68k) _LT_TAGVAR(archive_cmds, $1)='$RM $output_objdir/a2ixlibrary.data~$ECHO "#define NAME $libname" > $output_objdir/a2ixlibrary.data~$ECHO "#define LIBRARY_ID 1" >> $output_objdir/a2ixlibrary.data~$ECHO "#define VERSION $major" >> $output_objdir/a2ixlibrary.data~$ECHO "#define REVISION $revision" >> $output_objdir/a2ixlibrary.data~$AR $AR_FLAGS $lib $libobjs~$RANLIB $lib~(cd $output_objdir && a2ixlibrary -32)' _LT_TAGVAR(hardcode_libdir_flag_spec, $1)='-L$libdir' _LT_TAGVAR(hardcode_minus_L, $1)=yes ;; esac ;; bsdi[[45]]*) _LT_TAGVAR(export_dynamic_flag_spec, $1)=-rdynamic ;; cygwin* | mingw* | pw32* | cegcc*) # When not using gcc, we currently assume that we are using # Microsoft Visual C++. # hardcode_libdir_flag_spec is actually meaningless, as there is # no search path for DLLs. case $cc_basename in cl*) # Native MSVC _LT_TAGVAR(hardcode_libdir_flag_spec, $1)=' ' _LT_TAGVAR(allow_undefined_flag, $1)=unsupported _LT_TAGVAR(always_export_symbols, $1)=yes _LT_TAGVAR(file_list_spec, $1)='@' # Tell ltmain to make .lib files, not .a files. libext=lib # Tell ltmain to make .dll files, not .so files. shrext_cmds=".dll" # FIXME: Setting linknames here is a bad hack. _LT_TAGVAR(archive_cmds, $1)='$CC -o $output_objdir/$soname $libobjs $compiler_flags $deplibs -Wl,-dll~linknames=' _LT_TAGVAR(archive_expsym_cmds, $1)='if test "x`$SED 1q $export_symbols`" = xEXPORTS; then sed -n -e 's/\\\\\\\(.*\\\\\\\)/-link\\\ -EXPORT:\\\\\\\1/' -e '1\\\!p' < $export_symbols > $output_objdir/$soname.exp; else sed -e 's/\\\\\\\(.*\\\\\\\)/-link\\\ -EXPORT:\\\\\\\1/' < $export_symbols > $output_objdir/$soname.exp; fi~ $CC -o $tool_output_objdir$soname $libobjs $compiler_flags $deplibs "@$tool_output_objdir$soname.exp" -Wl,-DLL,-IMPLIB:"$tool_output_objdir$libname.dll.lib"~ linknames=' # The linker will not automatically build a static lib if we build a DLL. # _LT_TAGVAR(old_archive_from_new_cmds, $1)='true' _LT_TAGVAR(enable_shared_with_static_runtimes, $1)=yes _LT_TAGVAR(exclude_expsyms, $1)='_NULL_IMPORT_DESCRIPTOR|_IMPORT_DESCRIPTOR_.*' _LT_TAGVAR(export_symbols_cmds, $1)='$NM $libobjs $convenience | $global_symbol_pipe | $SED -e '\''/^[[BCDGRS]][[ ]]/s/.*[[ ]]\([[^ ]]*\)/\1,DATA/'\'' | $SED -e '\''/^[[AITW]][[ ]]/s/.*[[ ]]//'\'' | sort | uniq > $export_symbols' # Don't use ranlib _LT_TAGVAR(old_postinstall_cmds, $1)='chmod 644 $oldlib' _LT_TAGVAR(postlink_cmds, $1)='lt_outputfile="@OUTPUT@"~ lt_tool_outputfile="@TOOL_OUTPUT@"~ case $lt_outputfile in *.exe|*.EXE) ;; *) lt_outputfile="$lt_outputfile.exe" lt_tool_outputfile="$lt_tool_outputfile.exe" ;; esac~ if test "$MANIFEST_TOOL" != ":" && test -f "$lt_outputfile.manifest"; then $MANIFEST_TOOL -manifest "$lt_tool_outputfile.manifest" -outputresource:"$lt_tool_outputfile" || exit 1; $RM "$lt_outputfile.manifest"; fi' ;; *) # Assume MSVC wrapper _LT_TAGVAR(hardcode_libdir_flag_spec, $1)=' ' _LT_TAGVAR(allow_undefined_flag, $1)=unsupported # Tell ltmain to make .lib files, not .a files. libext=lib # Tell ltmain to make .dll files, not .so files. shrext_cmds=".dll" # FIXME: Setting linknames here is a bad hack. _LT_TAGVAR(archive_cmds, $1)='$CC -o $lib $libobjs $compiler_flags `func_echo_all "$deplibs" | $SED '\''s/ -lc$//'\''` -link -dll~linknames=' # The linker will automatically build a .lib file if we build a DLL. _LT_TAGVAR(old_archive_from_new_cmds, $1)='true' # FIXME: Should let the user specify the lib program. _LT_TAGVAR(old_archive_cmds, $1)='lib -OUT:$oldlib$oldobjs$old_deplibs' _LT_TAGVAR(enable_shared_with_static_runtimes, $1)=yes ;; esac ;; darwin* | rhapsody*) _LT_DARWIN_LINKER_FEATURES($1) ;; dgux*) _LT_TAGVAR(archive_cmds, $1)='$LD -G -h $soname -o $lib $libobjs $deplibs $linker_flags' _LT_TAGVAR(hardcode_libdir_flag_spec, $1)='-L$libdir' _LT_TAGVAR(hardcode_shlibpath_var, $1)=no ;; # FreeBSD 2.2.[012] allows us to include c++rt0.o to get C++ constructor # support. Future versions do this automatically, but an explicit c++rt0.o # does not break anything, and helps significantly (at the cost of a little # extra space). freebsd2.2*) _LT_TAGVAR(archive_cmds, $1)='$LD -Bshareable -o $lib $libobjs $deplibs $linker_flags /usr/lib/c++rt0.o' _LT_TAGVAR(hardcode_libdir_flag_spec, $1)='-R$libdir' _LT_TAGVAR(hardcode_direct, $1)=yes _LT_TAGVAR(hardcode_shlibpath_var, $1)=no ;; # Unfortunately, older versions of FreeBSD 2 do not have this feature. freebsd2.*) _LT_TAGVAR(archive_cmds, $1)='$LD -Bshareable -o $lib $libobjs $deplibs $linker_flags' _LT_TAGVAR(hardcode_direct, $1)=yes _LT_TAGVAR(hardcode_minus_L, $1)=yes _LT_TAGVAR(hardcode_shlibpath_var, $1)=no ;; # FreeBSD 3 and greater uses gcc -shared to do shared libraries. freebsd* | dragonfly*) _LT_TAGVAR(archive_cmds, $1)='$CC -shared $pic_flag -o $lib $libobjs $deplibs $compiler_flags' _LT_TAGVAR(hardcode_libdir_flag_spec, $1)='-R$libdir' _LT_TAGVAR(hardcode_direct, $1)=yes _LT_TAGVAR(hardcode_shlibpath_var, $1)=no ;; hpux9*) if test "$GCC" = yes; then _LT_TAGVAR(archive_cmds, $1)='$RM $output_objdir/$soname~$CC -shared $pic_flag ${wl}+b ${wl}$install_libdir -o $output_objdir/$soname $libobjs $deplibs $compiler_flags~test $output_objdir/$soname = $lib || mv $output_objdir/$soname $lib' else _LT_TAGVAR(archive_cmds, $1)='$RM $output_objdir/$soname~$LD -b +b $install_libdir -o $output_objdir/$soname $libobjs $deplibs $linker_flags~test $output_objdir/$soname = $lib || mv $output_objdir/$soname $lib' fi _LT_TAGVAR(hardcode_libdir_flag_spec, $1)='${wl}+b ${wl}$libdir' _LT_TAGVAR(hardcode_libdir_separator, $1)=: _LT_TAGVAR(hardcode_direct, $1)=yes # hardcode_minus_L: Not really in the search PATH, # but as the default location of the library. _LT_TAGVAR(hardcode_minus_L, $1)=yes _LT_TAGVAR(export_dynamic_flag_spec, $1)='${wl}-E' ;; hpux10*) if test "$GCC" = yes && test "$with_gnu_ld" = no; then _LT_TAGVAR(archive_cmds, $1)='$CC -shared $pic_flag ${wl}+h ${wl}$soname ${wl}+b ${wl}$install_libdir -o $lib $libobjs $deplibs $compiler_flags' else _LT_TAGVAR(archive_cmds, $1)='$LD -b +h $soname +b $install_libdir -o $lib $libobjs $deplibs $linker_flags' fi if test "$with_gnu_ld" = no; then _LT_TAGVAR(hardcode_libdir_flag_spec, $1)='${wl}+b ${wl}$libdir' _LT_TAGVAR(hardcode_libdir_separator, $1)=: _LT_TAGVAR(hardcode_direct, $1)=yes _LT_TAGVAR(hardcode_direct_absolute, $1)=yes _LT_TAGVAR(export_dynamic_flag_spec, $1)='${wl}-E' # hardcode_minus_L: Not really in the search PATH, # but as the default location of the library. _LT_TAGVAR(hardcode_minus_L, $1)=yes fi ;; hpux11*) if test "$GCC" = yes && test "$with_gnu_ld" = no; then case $host_cpu in hppa*64*) _LT_TAGVAR(archive_cmds, $1)='$CC -shared ${wl}+h ${wl}$soname -o $lib $libobjs $deplibs $compiler_flags' ;; ia64*) _LT_TAGVAR(archive_cmds, $1)='$CC -shared $pic_flag ${wl}+h ${wl}$soname ${wl}+nodefaultrpath -o $lib $libobjs $deplibs $compiler_flags' ;; *) _LT_TAGVAR(archive_cmds, $1)='$CC -shared $pic_flag ${wl}+h ${wl}$soname ${wl}+b ${wl}$install_libdir -o $lib $libobjs $deplibs $compiler_flags' ;; esac else case $host_cpu in hppa*64*) _LT_TAGVAR(archive_cmds, $1)='$CC -b ${wl}+h ${wl}$soname -o $lib $libobjs $deplibs $compiler_flags' ;; ia64*) _LT_TAGVAR(archive_cmds, $1)='$CC -b ${wl}+h ${wl}$soname ${wl}+nodefaultrpath -o $lib $libobjs $deplibs $compiler_flags' ;; *) m4_if($1, [], [ # Older versions of the 11.00 compiler do not understand -b yet # (HP92453-01 A.11.01.20 doesn't, HP92453-01 B.11.X.35175-35176.GP does) _LT_LINKER_OPTION([if $CC understands -b], _LT_TAGVAR(lt_cv_prog_compiler__b, $1), [-b], [_LT_TAGVAR(archive_cmds, $1)='$CC -b ${wl}+h ${wl}$soname ${wl}+b ${wl}$install_libdir -o $lib $libobjs $deplibs $compiler_flags'], [_LT_TAGVAR(archive_cmds, $1)='$LD -b +h $soname +b $install_libdir -o $lib $libobjs $deplibs $linker_flags'])], [_LT_TAGVAR(archive_cmds, $1)='$CC -b ${wl}+h ${wl}$soname ${wl}+b ${wl}$install_libdir -o $lib $libobjs $deplibs $compiler_flags']) ;; esac fi if test "$with_gnu_ld" = no; then _LT_TAGVAR(hardcode_libdir_flag_spec, $1)='${wl}+b ${wl}$libdir' _LT_TAGVAR(hardcode_libdir_separator, $1)=: case $host_cpu in hppa*64*|ia64*) _LT_TAGVAR(hardcode_direct, $1)=no _LT_TAGVAR(hardcode_shlibpath_var, $1)=no ;; *) _LT_TAGVAR(hardcode_direct, $1)=yes _LT_TAGVAR(hardcode_direct_absolute, $1)=yes _LT_TAGVAR(export_dynamic_flag_spec, $1)='${wl}-E' # hardcode_minus_L: Not really in the search PATH, # but as the default location of the library. _LT_TAGVAR(hardcode_minus_L, $1)=yes ;; esac fi ;; irix5* | irix6* | nonstopux*) if test "$GCC" = yes; then _LT_TAGVAR(archive_cmds, $1)='$CC -shared $pic_flag $libobjs $deplibs $compiler_flags ${wl}-soname ${wl}$soname `test -n "$verstring" && func_echo_all "${wl}-set_version ${wl}$verstring"` ${wl}-update_registry ${wl}${output_objdir}/so_locations -o $lib' # Try to use the -exported_symbol ld option, if it does not # work, assume that -exports_file does not work either and # implicitly export all symbols. # This should be the same for all languages, so no per-tag cache variable. AC_CACHE_CHECK([whether the $host_os linker accepts -exported_symbol], [lt_cv_irix_exported_symbol], [save_LDFLAGS="$LDFLAGS" LDFLAGS="$LDFLAGS -shared ${wl}-exported_symbol ${wl}foo ${wl}-update_registry ${wl}/dev/null" AC_LINK_IFELSE( [AC_LANG_SOURCE( [AC_LANG_CASE([C], [[int foo (void) { return 0; }]], [C++], [[int foo (void) { return 0; }]], [Fortran 77], [[ subroutine foo end]], [Fortran], [[ subroutine foo end]])])], [lt_cv_irix_exported_symbol=yes], [lt_cv_irix_exported_symbol=no]) LDFLAGS="$save_LDFLAGS"]) if test "$lt_cv_irix_exported_symbol" = yes; then _LT_TAGVAR(archive_expsym_cmds, $1)='$CC -shared $pic_flag $libobjs $deplibs $compiler_flags ${wl}-soname ${wl}$soname `test -n "$verstring" && func_echo_all "${wl}-set_version ${wl}$verstring"` ${wl}-update_registry ${wl}${output_objdir}/so_locations ${wl}-exports_file ${wl}$export_symbols -o $lib' fi else _LT_TAGVAR(archive_cmds, $1)='$CC -shared $libobjs $deplibs $compiler_flags -soname $soname `test -n "$verstring" && func_echo_all "-set_version $verstring"` -update_registry ${output_objdir}/so_locations -o $lib' _LT_TAGVAR(archive_expsym_cmds, $1)='$CC -shared $libobjs $deplibs $compiler_flags -soname $soname `test -n "$verstring" && func_echo_all "-set_version $verstring"` -update_registry ${output_objdir}/so_locations -exports_file $export_symbols -o $lib' fi _LT_TAGVAR(archive_cmds_need_lc, $1)='no' _LT_TAGVAR(hardcode_libdir_flag_spec, $1)='${wl}-rpath ${wl}$libdir' _LT_TAGVAR(hardcode_libdir_separator, $1)=: _LT_TAGVAR(inherit_rpath, $1)=yes _LT_TAGVAR(link_all_deplibs, $1)=yes ;; netbsd* | netbsdelf*-gnu) if echo __ELF__ | $CC -E - | $GREP __ELF__ >/dev/null; then _LT_TAGVAR(archive_cmds, $1)='$LD -Bshareable -o $lib $libobjs $deplibs $linker_flags' # a.out else _LT_TAGVAR(archive_cmds, $1)='$LD -shared -o $lib $libobjs $deplibs $linker_flags' # ELF fi _LT_TAGVAR(hardcode_libdir_flag_spec, $1)='-R$libdir' _LT_TAGVAR(hardcode_direct, $1)=yes _LT_TAGVAR(hardcode_shlibpath_var, $1)=no ;; newsos6) _LT_TAGVAR(archive_cmds, $1)='$LD -G -h $soname -o $lib $libobjs $deplibs $linker_flags' _LT_TAGVAR(hardcode_direct, $1)=yes _LT_TAGVAR(hardcode_libdir_flag_spec, $1)='${wl}-rpath ${wl}$libdir' _LT_TAGVAR(hardcode_libdir_separator, $1)=: _LT_TAGVAR(hardcode_shlibpath_var, $1)=no ;; *nto* | *qnx*) ;; openbsd*) if test -f /usr/libexec/ld.so; then _LT_TAGVAR(hardcode_direct, $1)=yes _LT_TAGVAR(hardcode_shlibpath_var, $1)=no _LT_TAGVAR(hardcode_direct_absolute, $1)=yes if test -z "`echo __ELF__ | $CC -E - | $GREP __ELF__`" || test "$host_os-$host_cpu" = "openbsd2.8-powerpc"; then _LT_TAGVAR(archive_cmds, $1)='$CC -shared $pic_flag -o $lib $libobjs $deplibs $compiler_flags' _LT_TAGVAR(archive_expsym_cmds, $1)='$CC -shared $pic_flag -o $lib $libobjs $deplibs $compiler_flags ${wl}-retain-symbols-file,$export_symbols' _LT_TAGVAR(hardcode_libdir_flag_spec, $1)='${wl}-rpath,$libdir' _LT_TAGVAR(export_dynamic_flag_spec, $1)='${wl}-E' else case $host_os in openbsd[[01]].* | openbsd2.[[0-7]] | openbsd2.[[0-7]].*) _LT_TAGVAR(archive_cmds, $1)='$LD -Bshareable -o $lib $libobjs $deplibs $linker_flags' _LT_TAGVAR(hardcode_libdir_flag_spec, $1)='-R$libdir' ;; *) _LT_TAGVAR(archive_cmds, $1)='$CC -shared $pic_flag -o $lib $libobjs $deplibs $compiler_flags' _LT_TAGVAR(hardcode_libdir_flag_spec, $1)='${wl}-rpath,$libdir' ;; esac fi else _LT_TAGVAR(ld_shlibs, $1)=no fi ;; os2*) _LT_TAGVAR(hardcode_libdir_flag_spec, $1)='-L$libdir' _LT_TAGVAR(hardcode_minus_L, $1)=yes _LT_TAGVAR(allow_undefined_flag, $1)=unsupported _LT_TAGVAR(archive_cmds, $1)='$ECHO "LIBRARY $libname INITINSTANCE" > $output_objdir/$libname.def~$ECHO "DESCRIPTION \"$libname\"" >> $output_objdir/$libname.def~echo DATA >> $output_objdir/$libname.def~echo " SINGLE NONSHARED" >> $output_objdir/$libname.def~echo EXPORTS >> $output_objdir/$libname.def~emxexp $libobjs >> $output_objdir/$libname.def~$CC -Zdll -Zcrtdll -o $lib $libobjs $deplibs $compiler_flags $output_objdir/$libname.def' _LT_TAGVAR(old_archive_from_new_cmds, $1)='emximp -o $output_objdir/$libname.a $output_objdir/$libname.def' ;; osf3*) if test "$GCC" = yes; then _LT_TAGVAR(allow_undefined_flag, $1)=' ${wl}-expect_unresolved ${wl}\*' _LT_TAGVAR(archive_cmds, $1)='$CC -shared${allow_undefined_flag} $libobjs $deplibs $compiler_flags ${wl}-soname ${wl}$soname `test -n "$verstring" && func_echo_all "${wl}-set_version ${wl}$verstring"` ${wl}-update_registry ${wl}${output_objdir}/so_locations -o $lib' else _LT_TAGVAR(allow_undefined_flag, $1)=' -expect_unresolved \*' _LT_TAGVAR(archive_cmds, $1)='$CC -shared${allow_undefined_flag} $libobjs $deplibs $compiler_flags -soname $soname `test -n "$verstring" && func_echo_all "-set_version $verstring"` -update_registry ${output_objdir}/so_locations -o $lib' fi _LT_TAGVAR(archive_cmds_need_lc, $1)='no' _LT_TAGVAR(hardcode_libdir_flag_spec, $1)='${wl}-rpath ${wl}$libdir' _LT_TAGVAR(hardcode_libdir_separator, $1)=: ;; osf4* | osf5*) # as osf3* with the addition of -msym flag if test "$GCC" = yes; then _LT_TAGVAR(allow_undefined_flag, $1)=' ${wl}-expect_unresolved ${wl}\*' _LT_TAGVAR(archive_cmds, $1)='$CC -shared${allow_undefined_flag} $pic_flag $libobjs $deplibs $compiler_flags ${wl}-msym ${wl}-soname ${wl}$soname `test -n "$verstring" && func_echo_all "${wl}-set_version ${wl}$verstring"` ${wl}-update_registry ${wl}${output_objdir}/so_locations -o $lib' _LT_TAGVAR(hardcode_libdir_flag_spec, $1)='${wl}-rpath ${wl}$libdir' else _LT_TAGVAR(allow_undefined_flag, $1)=' -expect_unresolved \*' _LT_TAGVAR(archive_cmds, $1)='$CC -shared${allow_undefined_flag} $libobjs $deplibs $compiler_flags -msym -soname $soname `test -n "$verstring" && func_echo_all "-set_version $verstring"` -update_registry ${output_objdir}/so_locations -o $lib' _LT_TAGVAR(archive_expsym_cmds, $1)='for i in `cat $export_symbols`; do printf "%s %s\\n" -exported_symbol "\$i" >> $lib.exp; done; printf "%s\\n" "-hidden">> $lib.exp~ $CC -shared${allow_undefined_flag} ${wl}-input ${wl}$lib.exp $compiler_flags $libobjs $deplibs -soname $soname `test -n "$verstring" && $ECHO "-set_version $verstring"` -update_registry ${output_objdir}/so_locations -o $lib~$RM $lib.exp' # Both c and cxx compiler support -rpath directly _LT_TAGVAR(hardcode_libdir_flag_spec, $1)='-rpath $libdir' fi _LT_TAGVAR(archive_cmds_need_lc, $1)='no' _LT_TAGVAR(hardcode_libdir_separator, $1)=: ;; solaris*) _LT_TAGVAR(no_undefined_flag, $1)=' -z defs' if test "$GCC" = yes; then wlarc='${wl}' _LT_TAGVAR(archive_cmds, $1)='$CC -shared $pic_flag ${wl}-z ${wl}text ${wl}-h ${wl}$soname -o $lib $libobjs $deplibs $compiler_flags' _LT_TAGVAR(archive_expsym_cmds, $1)='echo "{ global:" > $lib.exp~cat $export_symbols | $SED -e "s/\(.*\)/\1;/" >> $lib.exp~echo "local: *; };" >> $lib.exp~ $CC -shared $pic_flag ${wl}-z ${wl}text ${wl}-M ${wl}$lib.exp ${wl}-h ${wl}$soname -o $lib $libobjs $deplibs $compiler_flags~$RM $lib.exp' else case `$CC -V 2>&1` in *"Compilers 5.0"*) wlarc='' _LT_TAGVAR(archive_cmds, $1)='$LD -G${allow_undefined_flag} -h $soname -o $lib $libobjs $deplibs $linker_flags' _LT_TAGVAR(archive_expsym_cmds, $1)='echo "{ global:" > $lib.exp~cat $export_symbols | $SED -e "s/\(.*\)/\1;/" >> $lib.exp~echo "local: *; };" >> $lib.exp~ $LD -G${allow_undefined_flag} -M $lib.exp -h $soname -o $lib $libobjs $deplibs $linker_flags~$RM $lib.exp' ;; *) wlarc='${wl}' _LT_TAGVAR(archive_cmds, $1)='$CC -G${allow_undefined_flag} -h $soname -o $lib $libobjs $deplibs $compiler_flags' _LT_TAGVAR(archive_expsym_cmds, $1)='echo "{ global:" > $lib.exp~cat $export_symbols | $SED -e "s/\(.*\)/\1;/" >> $lib.exp~echo "local: *; };" >> $lib.exp~ $CC -G${allow_undefined_flag} -M $lib.exp -h $soname -o $lib $libobjs $deplibs $compiler_flags~$RM $lib.exp' ;; esac fi _LT_TAGVAR(hardcode_libdir_flag_spec, $1)='-R$libdir' _LT_TAGVAR(hardcode_shlibpath_var, $1)=no case $host_os in solaris2.[[0-5]] | solaris2.[[0-5]].*) ;; *) # The compiler driver will combine and reorder linker options, # but understands `-z linker_flag'. GCC discards it without `$wl', # but is careful enough not to reorder. # Supported since Solaris 2.6 (maybe 2.5.1?) if test "$GCC" = yes; then _LT_TAGVAR(whole_archive_flag_spec, $1)='${wl}-z ${wl}allextract$convenience ${wl}-z ${wl}defaultextract' else _LT_TAGVAR(whole_archive_flag_spec, $1)='-z allextract$convenience -z defaultextract' fi ;; esac _LT_TAGVAR(link_all_deplibs, $1)=yes ;; sunos4*) if test "x$host_vendor" = xsequent; then # Use $CC to link under sequent, because it throws in some extra .o # files that make .init and .fini sections work. _LT_TAGVAR(archive_cmds, $1)='$CC -G ${wl}-h $soname -o $lib $libobjs $deplibs $compiler_flags' else _LT_TAGVAR(archive_cmds, $1)='$LD -assert pure-text -Bstatic -o $lib $libobjs $deplibs $linker_flags' fi _LT_TAGVAR(hardcode_libdir_flag_spec, $1)='-L$libdir' _LT_TAGVAR(hardcode_direct, $1)=yes _LT_TAGVAR(hardcode_minus_L, $1)=yes _LT_TAGVAR(hardcode_shlibpath_var, $1)=no ;; sysv4) case $host_vendor in sni) _LT_TAGVAR(archive_cmds, $1)='$LD -G -h $soname -o $lib $libobjs $deplibs $linker_flags' _LT_TAGVAR(hardcode_direct, $1)=yes # is this really true??? ;; siemens) ## LD is ld it makes a PLAMLIB ## CC just makes a GrossModule. _LT_TAGVAR(archive_cmds, $1)='$LD -G -o $lib $libobjs $deplibs $linker_flags' _LT_TAGVAR(reload_cmds, $1)='$CC -r -o $output$reload_objs' _LT_TAGVAR(hardcode_direct, $1)=no ;; motorola) _LT_TAGVAR(archive_cmds, $1)='$LD -G -h $soname -o $lib $libobjs $deplibs $linker_flags' _LT_TAGVAR(hardcode_direct, $1)=no #Motorola manual says yes, but my tests say they lie ;; esac runpath_var='LD_RUN_PATH' _LT_TAGVAR(hardcode_shlibpath_var, $1)=no ;; sysv4.3*) _LT_TAGVAR(archive_cmds, $1)='$LD -G -h $soname -o $lib $libobjs $deplibs $linker_flags' _LT_TAGVAR(hardcode_shlibpath_var, $1)=no _LT_TAGVAR(export_dynamic_flag_spec, $1)='-Bexport' ;; sysv4*MP*) if test -d /usr/nec; then _LT_TAGVAR(archive_cmds, $1)='$LD -G -h $soname -o $lib $libobjs $deplibs $linker_flags' _LT_TAGVAR(hardcode_shlibpath_var, $1)=no runpath_var=LD_RUN_PATH hardcode_runpath_var=yes _LT_TAGVAR(ld_shlibs, $1)=yes fi ;; sysv4*uw2* | sysv5OpenUNIX* | sysv5UnixWare7.[[01]].[[10]]* | unixware7* | sco3.2v5.0.[[024]]*) _LT_TAGVAR(no_undefined_flag, $1)='${wl}-z,text' _LT_TAGVAR(archive_cmds_need_lc, $1)=no _LT_TAGVAR(hardcode_shlibpath_var, $1)=no runpath_var='LD_RUN_PATH' if test "$GCC" = yes; then _LT_TAGVAR(archive_cmds, $1)='$CC -shared ${wl}-h,$soname -o $lib $libobjs $deplibs $compiler_flags' _LT_TAGVAR(archive_expsym_cmds, $1)='$CC -shared ${wl}-Bexport:$export_symbols ${wl}-h,$soname -o $lib $libobjs $deplibs $compiler_flags' else _LT_TAGVAR(archive_cmds, $1)='$CC -G ${wl}-h,$soname -o $lib $libobjs $deplibs $compiler_flags' _LT_TAGVAR(archive_expsym_cmds, $1)='$CC -G ${wl}-Bexport:$export_symbols ${wl}-h,$soname -o $lib $libobjs $deplibs $compiler_flags' fi ;; sysv5* | sco3.2v5* | sco5v6*) # Note: We can NOT use -z defs as we might desire, because we do not # link with -lc, and that would cause any symbols used from libc to # always be unresolved, which means just about no library would # ever link correctly. If we're not using GNU ld we use -z text # though, which does catch some bad symbols but isn't as heavy-handed # as -z defs. _LT_TAGVAR(no_undefined_flag, $1)='${wl}-z,text' _LT_TAGVAR(allow_undefined_flag, $1)='${wl}-z,nodefs' _LT_TAGVAR(archive_cmds_need_lc, $1)=no _LT_TAGVAR(hardcode_shlibpath_var, $1)=no _LT_TAGVAR(hardcode_libdir_flag_spec, $1)='${wl}-R,$libdir' _LT_TAGVAR(hardcode_libdir_separator, $1)=':' _LT_TAGVAR(link_all_deplibs, $1)=yes _LT_TAGVAR(export_dynamic_flag_spec, $1)='${wl}-Bexport' runpath_var='LD_RUN_PATH' if test "$GCC" = yes; then _LT_TAGVAR(archive_cmds, $1)='$CC -shared ${wl}-h,$soname -o $lib $libobjs $deplibs $compiler_flags' _LT_TAGVAR(archive_expsym_cmds, $1)='$CC -shared ${wl}-Bexport:$export_symbols ${wl}-h,$soname -o $lib $libobjs $deplibs $compiler_flags' else _LT_TAGVAR(archive_cmds, $1)='$CC -G ${wl}-h,$soname -o $lib $libobjs $deplibs $compiler_flags' _LT_TAGVAR(archive_expsym_cmds, $1)='$CC -G ${wl}-Bexport:$export_symbols ${wl}-h,$soname -o $lib $libobjs $deplibs $compiler_flags' fi ;; uts4*) _LT_TAGVAR(archive_cmds, $1)='$LD -G -h $soname -o $lib $libobjs $deplibs $linker_flags' _LT_TAGVAR(hardcode_libdir_flag_spec, $1)='-L$libdir' _LT_TAGVAR(hardcode_shlibpath_var, $1)=no ;; *) _LT_TAGVAR(ld_shlibs, $1)=no ;; esac if test x$host_vendor = xsni; then case $host in sysv4 | sysv4.2uw2* | sysv4.3* | sysv5*) _LT_TAGVAR(export_dynamic_flag_spec, $1)='${wl}-Blargedynsym' ;; esac fi fi ]) AC_MSG_RESULT([$_LT_TAGVAR(ld_shlibs, $1)]) test "$_LT_TAGVAR(ld_shlibs, $1)" = no && can_build_shared=no _LT_TAGVAR(with_gnu_ld, $1)=$with_gnu_ld _LT_DECL([], [libext], [0], [Old archive suffix (normally "a")])dnl _LT_DECL([], [shrext_cmds], [1], [Shared library suffix (normally ".so")])dnl _LT_DECL([], [extract_expsyms_cmds], [2], [The commands to extract the exported symbol list from a shared archive]) # # Do we need to explicitly link libc? # case "x$_LT_TAGVAR(archive_cmds_need_lc, $1)" in x|xyes) # Assume -lc should be added _LT_TAGVAR(archive_cmds_need_lc, $1)=yes if test "$enable_shared" = yes && test "$GCC" = yes; then case $_LT_TAGVAR(archive_cmds, $1) in *'~'*) # FIXME: we may have to deal with multi-command sequences. ;; '$CC '*) # Test whether the compiler implicitly links with -lc since on some # systems, -lgcc has to come before -lc. If gcc already passes -lc # to ld, don't add -lc before -lgcc. AC_CACHE_CHECK([whether -lc should be explicitly linked in], [lt_cv_]_LT_TAGVAR(archive_cmds_need_lc, $1), [$RM conftest* echo "$lt_simple_compile_test_code" > conftest.$ac_ext if AC_TRY_EVAL(ac_compile) 2>conftest.err; then soname=conftest lib=conftest libobjs=conftest.$ac_objext deplibs= wl=$_LT_TAGVAR(lt_prog_compiler_wl, $1) pic_flag=$_LT_TAGVAR(lt_prog_compiler_pic, $1) compiler_flags=-v linker_flags=-v verstring= output_objdir=. libname=conftest lt_save_allow_undefined_flag=$_LT_TAGVAR(allow_undefined_flag, $1) _LT_TAGVAR(allow_undefined_flag, $1)= if AC_TRY_EVAL(_LT_TAGVAR(archive_cmds, $1) 2\>\&1 \| $GREP \" -lc \" \>/dev/null 2\>\&1) then lt_cv_[]_LT_TAGVAR(archive_cmds_need_lc, $1)=no else lt_cv_[]_LT_TAGVAR(archive_cmds_need_lc, $1)=yes fi _LT_TAGVAR(allow_undefined_flag, $1)=$lt_save_allow_undefined_flag else cat conftest.err 1>&5 fi $RM conftest* ]) _LT_TAGVAR(archive_cmds_need_lc, $1)=$lt_cv_[]_LT_TAGVAR(archive_cmds_need_lc, $1) ;; esac fi ;; esac _LT_TAGDECL([build_libtool_need_lc], [archive_cmds_need_lc], [0], [Whether or not to add -lc for building shared libraries]) _LT_TAGDECL([allow_libtool_libs_with_static_runtimes], [enable_shared_with_static_runtimes], [0], [Whether or not to disallow shared libs when runtime libs are static]) _LT_TAGDECL([], [export_dynamic_flag_spec], [1], [Compiler flag to allow reflexive dlopens]) _LT_TAGDECL([], [whole_archive_flag_spec], [1], [Compiler flag to generate shared objects directly from archives]) _LT_TAGDECL([], [compiler_needs_object], [1], [Whether the compiler copes with passing no objects directly]) _LT_TAGDECL([], [old_archive_from_new_cmds], [2], [Create an old-style archive from a shared archive]) _LT_TAGDECL([], [old_archive_from_expsyms_cmds], [2], [Create a temporary old-style archive to link instead of a shared archive]) _LT_TAGDECL([], [archive_cmds], [2], [Commands used to build a shared archive]) _LT_TAGDECL([], [archive_expsym_cmds], [2]) _LT_TAGDECL([], [module_cmds], [2], [Commands used to build a loadable module if different from building a shared archive.]) _LT_TAGDECL([], [module_expsym_cmds], [2]) _LT_TAGDECL([], [with_gnu_ld], [1], [Whether we are building with GNU ld or not]) _LT_TAGDECL([], [allow_undefined_flag], [1], [Flag that allows shared libraries with undefined symbols to be built]) _LT_TAGDECL([], [no_undefined_flag], [1], [Flag that enforces no undefined symbols]) _LT_TAGDECL([], [hardcode_libdir_flag_spec], [1], [Flag to hardcode $libdir into a binary during linking. This must work even if $libdir does not exist]) _LT_TAGDECL([], [hardcode_libdir_separator], [1], [Whether we need a single "-rpath" flag with a separated argument]) _LT_TAGDECL([], [hardcode_direct], [0], [Set to "yes" if using DIR/libNAME${shared_ext} during linking hardcodes DIR into the resulting binary]) _LT_TAGDECL([], [hardcode_direct_absolute], [0], [Set to "yes" if using DIR/libNAME${shared_ext} during linking hardcodes DIR into the resulting binary and the resulting library dependency is "absolute", i.e impossible to change by setting ${shlibpath_var} if the library is relocated]) _LT_TAGDECL([], [hardcode_minus_L], [0], [Set to "yes" if using the -LDIR flag during linking hardcodes DIR into the resulting binary]) _LT_TAGDECL([], [hardcode_shlibpath_var], [0], [Set to "yes" if using SHLIBPATH_VAR=DIR during linking hardcodes DIR into the resulting binary]) _LT_TAGDECL([], [hardcode_automatic], [0], [Set to "yes" if building a shared library automatically hardcodes DIR into the library and all subsequent libraries and executables linked against it]) _LT_TAGDECL([], [inherit_rpath], [0], [Set to yes if linker adds runtime paths of dependent libraries to runtime path list]) _LT_TAGDECL([], [link_all_deplibs], [0], [Whether libtool must link a program against all its dependency libraries]) _LT_TAGDECL([], [always_export_symbols], [0], [Set to "yes" if exported symbols are required]) _LT_TAGDECL([], [export_symbols_cmds], [2], [The commands to list exported symbols]) _LT_TAGDECL([], [exclude_expsyms], [1], [Symbols that should not be listed in the preloaded symbols]) _LT_TAGDECL([], [include_expsyms], [1], [Symbols that must always be exported]) _LT_TAGDECL([], [prelink_cmds], [2], [Commands necessary for linking programs (against libraries) with templates]) _LT_TAGDECL([], [postlink_cmds], [2], [Commands necessary for finishing linking programs]) _LT_TAGDECL([], [file_list_spec], [1], [Specify filename containing input files]) dnl FIXME: Not yet implemented dnl _LT_TAGDECL([], [thread_safe_flag_spec], [1], dnl [Compiler flag to generate thread safe objects]) ])# _LT_LINKER_SHLIBS # _LT_LANG_C_CONFIG([TAG]) # ------------------------ # Ensure that the configuration variables for a C compiler are suitably # defined. These variables are subsequently used by _LT_CONFIG to write # the compiler configuration to `libtool'. m4_defun([_LT_LANG_C_CONFIG], [m4_require([_LT_DECL_EGREP])dnl lt_save_CC="$CC" AC_LANG_PUSH(C) # Source file extension for C test sources. ac_ext=c # Object file extension for compiled C test sources. objext=o _LT_TAGVAR(objext, $1)=$objext # Code to be used in simple compile tests lt_simple_compile_test_code="int some_variable = 0;" # Code to be used in simple link tests lt_simple_link_test_code='int main(){return(0);}' _LT_TAG_COMPILER # Save the default compiler, since it gets overwritten when the other # tags are being tested, and _LT_TAGVAR(compiler, []) is a NOP. compiler_DEFAULT=$CC # save warnings/boilerplate of simple test code _LT_COMPILER_BOILERPLATE _LT_LINKER_BOILERPLATE ## CAVEAT EMPTOR: ## There is no encapsulation within the following macros, do not change ## the running order or otherwise move them around unless you know exactly ## what you are doing... if test -n "$compiler"; then _LT_COMPILER_NO_RTTI($1) _LT_COMPILER_PIC($1) _LT_COMPILER_C_O($1) _LT_COMPILER_FILE_LOCKS($1) _LT_LINKER_SHLIBS($1) _LT_SYS_DYNAMIC_LINKER($1) _LT_LINKER_HARDCODE_LIBPATH($1) LT_SYS_DLOPEN_SELF _LT_CMD_STRIPLIB # Report which library types will actually be built AC_MSG_CHECKING([if libtool supports shared libraries]) AC_MSG_RESULT([$can_build_shared]) AC_MSG_CHECKING([whether to build shared libraries]) test "$can_build_shared" = "no" && enable_shared=no # On AIX, shared libraries and static libraries use the same namespace, and # are all built from PIC. case $host_os in aix3*) test "$enable_shared" = yes && enable_static=no if test -n "$RANLIB"; then archive_cmds="$archive_cmds~\$RANLIB \$lib" postinstall_cmds='$RANLIB $lib' fi ;; aix[[4-9]]*) if test "$host_cpu" != ia64 && test "$aix_use_runtimelinking" = no ; then test "$enable_shared" = yes && enable_static=no fi ;; esac AC_MSG_RESULT([$enable_shared]) AC_MSG_CHECKING([whether to build static libraries]) # Make sure either enable_shared or enable_static is yes. test "$enable_shared" = yes || enable_static=yes AC_MSG_RESULT([$enable_static]) _LT_CONFIG($1) fi AC_LANG_POP CC="$lt_save_CC" ])# _LT_LANG_C_CONFIG # _LT_LANG_CXX_CONFIG([TAG]) # -------------------------- # Ensure that the configuration variables for a C++ compiler are suitably # defined. These variables are subsequently used by _LT_CONFIG to write # the compiler configuration to `libtool'. m4_defun([_LT_LANG_CXX_CONFIG], [m4_require([_LT_FILEUTILS_DEFAULTS])dnl m4_require([_LT_DECL_EGREP])dnl m4_require([_LT_PATH_MANIFEST_TOOL])dnl if test -n "$CXX" && ( test "X$CXX" != "Xno" && ( (test "X$CXX" = "Xg++" && `g++ -v >/dev/null 2>&1` ) || (test "X$CXX" != "Xg++"))) ; then AC_PROG_CXXCPP else _lt_caught_CXX_error=yes fi AC_LANG_PUSH(C++) _LT_TAGVAR(archive_cmds_need_lc, $1)=no _LT_TAGVAR(allow_undefined_flag, $1)= _LT_TAGVAR(always_export_symbols, $1)=no _LT_TAGVAR(archive_expsym_cmds, $1)= _LT_TAGVAR(compiler_needs_object, $1)=no _LT_TAGVAR(export_dynamic_flag_spec, $1)= _LT_TAGVAR(hardcode_direct, $1)=no _LT_TAGVAR(hardcode_direct_absolute, $1)=no _LT_TAGVAR(hardcode_libdir_flag_spec, $1)= _LT_TAGVAR(hardcode_libdir_separator, $1)= _LT_TAGVAR(hardcode_minus_L, $1)=no _LT_TAGVAR(hardcode_shlibpath_var, $1)=unsupported _LT_TAGVAR(hardcode_automatic, $1)=no _LT_TAGVAR(inherit_rpath, $1)=no _LT_TAGVAR(module_cmds, $1)= _LT_TAGVAR(module_expsym_cmds, $1)= _LT_TAGVAR(link_all_deplibs, $1)=unknown _LT_TAGVAR(old_archive_cmds, $1)=$old_archive_cmds _LT_TAGVAR(reload_flag, $1)=$reload_flag _LT_TAGVAR(reload_cmds, $1)=$reload_cmds _LT_TAGVAR(no_undefined_flag, $1)= _LT_TAGVAR(whole_archive_flag_spec, $1)= _LT_TAGVAR(enable_shared_with_static_runtimes, $1)=no # Source file extension for C++ test sources. ac_ext=cpp # Object file extension for compiled C++ test sources. objext=o _LT_TAGVAR(objext, $1)=$objext # No sense in running all these tests if we already determined that # the CXX compiler isn't working. Some variables (like enable_shared) # are currently assumed to apply to all compilers on this platform, # and will be corrupted by setting them based on a non-working compiler. if test "$_lt_caught_CXX_error" != yes; then # Code to be used in simple compile tests lt_simple_compile_test_code="int some_variable = 0;" # Code to be used in simple link tests lt_simple_link_test_code='int main(int, char *[[]]) { return(0); }' # ltmain only uses $CC for tagged configurations so make sure $CC is set. _LT_TAG_COMPILER # save warnings/boilerplate of simple test code _LT_COMPILER_BOILERPLATE _LT_LINKER_BOILERPLATE # Allow CC to be a program name with arguments. lt_save_CC=$CC lt_save_CFLAGS=$CFLAGS lt_save_LD=$LD lt_save_GCC=$GCC GCC=$GXX lt_save_with_gnu_ld=$with_gnu_ld lt_save_path_LD=$lt_cv_path_LD if test -n "${lt_cv_prog_gnu_ldcxx+set}"; then lt_cv_prog_gnu_ld=$lt_cv_prog_gnu_ldcxx else $as_unset lt_cv_prog_gnu_ld fi if test -n "${lt_cv_path_LDCXX+set}"; then lt_cv_path_LD=$lt_cv_path_LDCXX else $as_unset lt_cv_path_LD fi test -z "${LDCXX+set}" || LD=$LDCXX CC=${CXX-"c++"} CFLAGS=$CXXFLAGS compiler=$CC _LT_TAGVAR(compiler, $1)=$CC _LT_CC_BASENAME([$compiler]) if test -n "$compiler"; then # We don't want -fno-exception when compiling C++ code, so set the # no_builtin_flag separately if test "$GXX" = yes; then _LT_TAGVAR(lt_prog_compiler_no_builtin_flag, $1)=' -fno-builtin' else _LT_TAGVAR(lt_prog_compiler_no_builtin_flag, $1)= fi if test "$GXX" = yes; then # Set up default GNU C++ configuration LT_PATH_LD # Check if GNU C++ uses GNU ld as the underlying linker, since the # archiving commands below assume that GNU ld is being used. if test "$with_gnu_ld" = yes; then _LT_TAGVAR(archive_cmds, $1)='$CC $pic_flag -shared -nostdlib $predep_objects $libobjs $deplibs $postdep_objects $compiler_flags ${wl}-soname $wl$soname -o $lib' _LT_TAGVAR(archive_expsym_cmds, $1)='$CC $pic_flag -shared -nostdlib $predep_objects $libobjs $deplibs $postdep_objects $compiler_flags ${wl}-soname $wl$soname ${wl}-retain-symbols-file $wl$export_symbols -o $lib' _LT_TAGVAR(hardcode_libdir_flag_spec, $1)='${wl}-rpath ${wl}$libdir' _LT_TAGVAR(export_dynamic_flag_spec, $1)='${wl}--export-dynamic' # If archive_cmds runs LD, not CC, wlarc should be empty # XXX I think wlarc can be eliminated in ltcf-cxx, but I need to # investigate it a little bit more. (MM) wlarc='${wl}' # ancient GNU ld didn't support --whole-archive et. al. if eval "`$CC -print-prog-name=ld` --help 2>&1" | $GREP 'no-whole-archive' > /dev/null; then _LT_TAGVAR(whole_archive_flag_spec, $1)="$wlarc"'--whole-archive$convenience '"$wlarc"'--no-whole-archive' else _LT_TAGVAR(whole_archive_flag_spec, $1)= fi else with_gnu_ld=no wlarc= # A generic and very simple default shared library creation # command for GNU C++ for the case where it uses the native # linker, instead of GNU ld. If possible, this setting should # overridden to take advantage of the native linker features on # the platform it is being used on. _LT_TAGVAR(archive_cmds, $1)='$CC -shared -nostdlib $predep_objects $libobjs $deplibs $postdep_objects $compiler_flags -o $lib' fi # Commands to make compiler produce verbose output that lists # what "hidden" libraries, object files and flags are used when # linking a shared library. output_verbose_link_cmd='$CC -shared $CFLAGS -v conftest.$objext 2>&1 | $GREP -v "^Configured with:" | $GREP "\-L"' else GXX=no with_gnu_ld=no wlarc= fi # PORTME: fill in a description of your system's C++ link characteristics AC_MSG_CHECKING([whether the $compiler linker ($LD) supports shared libraries]) _LT_TAGVAR(ld_shlibs, $1)=yes case $host_os in aix3*) # FIXME: insert proper C++ library support _LT_TAGVAR(ld_shlibs, $1)=no ;; aix[[4-9]]*) if test "$host_cpu" = ia64; then # On IA64, the linker does run time linking by default, so we don't # have to do anything special. aix_use_runtimelinking=no exp_sym_flag='-Bexport' no_entry_flag="" else aix_use_runtimelinking=no # Test if we are trying to use run time linking or normal # AIX style linking. If -brtl is somewhere in LDFLAGS, we # need to do runtime linking. case $host_os in aix4.[[23]]|aix4.[[23]].*|aix[[5-9]]*) for ld_flag in $LDFLAGS; do case $ld_flag in *-brtl*) aix_use_runtimelinking=yes break ;; esac done ;; esac exp_sym_flag='-bexport' no_entry_flag='-bnoentry' fi # When large executables or shared objects are built, AIX ld can # have problems creating the table of contents. If linking a library # or program results in "error TOC overflow" add -mminimal-toc to # CXXFLAGS/CFLAGS for g++/gcc. In the cases where that is not # enough to fix the problem, add -Wl,-bbigtoc to LDFLAGS. _LT_TAGVAR(archive_cmds, $1)='' _LT_TAGVAR(hardcode_direct, $1)=yes _LT_TAGVAR(hardcode_direct_absolute, $1)=yes _LT_TAGVAR(hardcode_libdir_separator, $1)=':' _LT_TAGVAR(link_all_deplibs, $1)=yes _LT_TAGVAR(file_list_spec, $1)='${wl}-f,' if test "$GXX" = yes; then case $host_os in aix4.[[012]]|aix4.[[012]].*) # We only want to do this on AIX 4.2 and lower, the check # below for broken collect2 doesn't work under 4.3+ collect2name=`${CC} -print-prog-name=collect2` if test -f "$collect2name" && strings "$collect2name" | $GREP resolve_lib_name >/dev/null then # We have reworked collect2 : else # We have old collect2 _LT_TAGVAR(hardcode_direct, $1)=unsupported # It fails to find uninstalled libraries when the uninstalled # path is not listed in the libpath. Setting hardcode_minus_L # to unsupported forces relinking _LT_TAGVAR(hardcode_minus_L, $1)=yes _LT_TAGVAR(hardcode_libdir_flag_spec, $1)='-L$libdir' _LT_TAGVAR(hardcode_libdir_separator, $1)= fi esac shared_flag='-shared' if test "$aix_use_runtimelinking" = yes; then shared_flag="$shared_flag "'${wl}-G' fi else # not using gcc if test "$host_cpu" = ia64; then # VisualAge C++, Version 5.5 for AIX 5L for IA-64, Beta 3 Release # chokes on -Wl,-G. The following line is correct: shared_flag='-G' else if test "$aix_use_runtimelinking" = yes; then shared_flag='${wl}-G' else shared_flag='${wl}-bM:SRE' fi fi fi _LT_TAGVAR(export_dynamic_flag_spec, $1)='${wl}-bexpall' # It seems that -bexpall does not export symbols beginning with # underscore (_), so it is better to generate a list of symbols to # export. _LT_TAGVAR(always_export_symbols, $1)=yes if test "$aix_use_runtimelinking" = yes; then # Warning - without using the other runtime loading flags (-brtl), # -berok will link without error, but may produce a broken library. _LT_TAGVAR(allow_undefined_flag, $1)='-berok' # Determine the default libpath from the value encoded in an empty # executable. _LT_SYS_MODULE_PATH_AIX([$1]) _LT_TAGVAR(hardcode_libdir_flag_spec, $1)='${wl}-blibpath:$libdir:'"$aix_libpath" _LT_TAGVAR(archive_expsym_cmds, $1)='$CC -o $output_objdir/$soname $libobjs $deplibs '"\${wl}$no_entry_flag"' $compiler_flags `if test "x${allow_undefined_flag}" != "x"; then func_echo_all "${wl}${allow_undefined_flag}"; else :; fi` '"\${wl}$exp_sym_flag:\$export_symbols $shared_flag" else if test "$host_cpu" = ia64; then _LT_TAGVAR(hardcode_libdir_flag_spec, $1)='${wl}-R $libdir:/usr/lib:/lib' _LT_TAGVAR(allow_undefined_flag, $1)="-z nodefs" _LT_TAGVAR(archive_expsym_cmds, $1)="\$CC $shared_flag"' -o $output_objdir/$soname $libobjs $deplibs '"\${wl}$no_entry_flag"' $compiler_flags ${wl}${allow_undefined_flag} '"\${wl}$exp_sym_flag:\$export_symbols" else # Determine the default libpath from the value encoded in an # empty executable. _LT_SYS_MODULE_PATH_AIX([$1]) _LT_TAGVAR(hardcode_libdir_flag_spec, $1)='${wl}-blibpath:$libdir:'"$aix_libpath" # Warning - without using the other run time loading flags, # -berok will link without error, but may produce a broken library. _LT_TAGVAR(no_undefined_flag, $1)=' ${wl}-bernotok' _LT_TAGVAR(allow_undefined_flag, $1)=' ${wl}-berok' if test "$with_gnu_ld" = yes; then # We only use this code for GNU lds that support --whole-archive. _LT_TAGVAR(whole_archive_flag_spec, $1)='${wl}--whole-archive$convenience ${wl}--no-whole-archive' else # Exported symbols can be pulled into shared objects from archives _LT_TAGVAR(whole_archive_flag_spec, $1)='$convenience' fi _LT_TAGVAR(archive_cmds_need_lc, $1)=yes # This is similar to how AIX traditionally builds its shared # libraries. _LT_TAGVAR(archive_expsym_cmds, $1)="\$CC $shared_flag"' -o $output_objdir/$soname $libobjs $deplibs ${wl}-bnoentry $compiler_flags ${wl}-bE:$export_symbols${allow_undefined_flag}~$AR $AR_FLAGS $output_objdir/$libname$release.a $output_objdir/$soname' fi fi ;; beos*) if $LD --help 2>&1 | $GREP ': supported targets:.* elf' > /dev/null; then _LT_TAGVAR(allow_undefined_flag, $1)=unsupported # Joseph Beckenbach says some releases of gcc # support --undefined. This deserves some investigation. FIXME _LT_TAGVAR(archive_cmds, $1)='$CC -nostart $libobjs $deplibs $compiler_flags ${wl}-soname $wl$soname -o $lib' else _LT_TAGVAR(ld_shlibs, $1)=no fi ;; chorus*) case $cc_basename in *) # FIXME: insert proper C++ library support _LT_TAGVAR(ld_shlibs, $1)=no ;; esac ;; cygwin* | mingw* | pw32* | cegcc*) case $GXX,$cc_basename in ,cl* | no,cl*) # Native MSVC # hardcode_libdir_flag_spec is actually meaningless, as there is # no search path for DLLs. _LT_TAGVAR(hardcode_libdir_flag_spec, $1)=' ' _LT_TAGVAR(allow_undefined_flag, $1)=unsupported _LT_TAGVAR(always_export_symbols, $1)=yes _LT_TAGVAR(file_list_spec, $1)='@' # Tell ltmain to make .lib files, not .a files. libext=lib # Tell ltmain to make .dll files, not .so files. shrext_cmds=".dll" # FIXME: Setting linknames here is a bad hack. _LT_TAGVAR(archive_cmds, $1)='$CC -o $output_objdir/$soname $libobjs $compiler_flags $deplibs -Wl,-dll~linknames=' _LT_TAGVAR(archive_expsym_cmds, $1)='if test "x`$SED 1q $export_symbols`" = xEXPORTS; then $SED -n -e 's/\\\\\\\(.*\\\\\\\)/-link\\\ -EXPORT:\\\\\\\1/' -e '1\\\!p' < $export_symbols > $output_objdir/$soname.exp; else $SED -e 's/\\\\\\\(.*\\\\\\\)/-link\\\ -EXPORT:\\\\\\\1/' < $export_symbols > $output_objdir/$soname.exp; fi~ $CC -o $tool_output_objdir$soname $libobjs $compiler_flags $deplibs "@$tool_output_objdir$soname.exp" -Wl,-DLL,-IMPLIB:"$tool_output_objdir$libname.dll.lib"~ linknames=' # The linker will not automatically build a static lib if we build a DLL. # _LT_TAGVAR(old_archive_from_new_cmds, $1)='true' _LT_TAGVAR(enable_shared_with_static_runtimes, $1)=yes # Don't use ranlib _LT_TAGVAR(old_postinstall_cmds, $1)='chmod 644 $oldlib' _LT_TAGVAR(postlink_cmds, $1)='lt_outputfile="@OUTPUT@"~ lt_tool_outputfile="@TOOL_OUTPUT@"~ case $lt_outputfile in *.exe|*.EXE) ;; *) lt_outputfile="$lt_outputfile.exe" lt_tool_outputfile="$lt_tool_outputfile.exe" ;; esac~ func_to_tool_file "$lt_outputfile"~ if test "$MANIFEST_TOOL" != ":" && test -f "$lt_outputfile.manifest"; then $MANIFEST_TOOL -manifest "$lt_tool_outputfile.manifest" -outputresource:"$lt_tool_outputfile" || exit 1; $RM "$lt_outputfile.manifest"; fi' ;; *) # g++ # _LT_TAGVAR(hardcode_libdir_flag_spec, $1) is actually meaningless, # as there is no search path for DLLs. _LT_TAGVAR(hardcode_libdir_flag_spec, $1)='-L$libdir' _LT_TAGVAR(export_dynamic_flag_spec, $1)='${wl}--export-all-symbols' _LT_TAGVAR(allow_undefined_flag, $1)=unsupported _LT_TAGVAR(always_export_symbols, $1)=no _LT_TAGVAR(enable_shared_with_static_runtimes, $1)=yes if $LD --help 2>&1 | $GREP 'auto-import' > /dev/null; then _LT_TAGVAR(archive_cmds, $1)='$CC -shared -nostdlib $predep_objects $libobjs $deplibs $postdep_objects $compiler_flags -o $output_objdir/$soname ${wl}--enable-auto-image-base -Xlinker --out-implib -Xlinker $lib' # If the export-symbols file already is a .def file (1st line # is EXPORTS), use it as is; otherwise, prepend... _LT_TAGVAR(archive_expsym_cmds, $1)='if test "x`$SED 1q $export_symbols`" = xEXPORTS; then cp $export_symbols $output_objdir/$soname.def; else echo EXPORTS > $output_objdir/$soname.def; cat $export_symbols >> $output_objdir/$soname.def; fi~ $CC -shared -nostdlib $output_objdir/$soname.def $predep_objects $libobjs $deplibs $postdep_objects $compiler_flags -o $output_objdir/$soname ${wl}--enable-auto-image-base -Xlinker --out-implib -Xlinker $lib' else _LT_TAGVAR(ld_shlibs, $1)=no fi ;; esac ;; darwin* | rhapsody*) _LT_DARWIN_LINKER_FEATURES($1) ;; dgux*) case $cc_basename in ec++*) # FIXME: insert proper C++ library support _LT_TAGVAR(ld_shlibs, $1)=no ;; ghcx*) # Green Hills C++ Compiler # FIXME: insert proper C++ library support _LT_TAGVAR(ld_shlibs, $1)=no ;; *) # FIXME: insert proper C++ library support _LT_TAGVAR(ld_shlibs, $1)=no ;; esac ;; freebsd2.*) # C++ shared libraries reported to be fairly broken before # switch to ELF _LT_TAGVAR(ld_shlibs, $1)=no ;; freebsd-elf*) _LT_TAGVAR(archive_cmds_need_lc, $1)=no ;; freebsd* | dragonfly*) # FreeBSD 3 and later use GNU C++ and GNU ld with standard ELF # conventions _LT_TAGVAR(ld_shlibs, $1)=yes ;; haiku*) _LT_TAGVAR(archive_cmds, $1)='$CC -shared $libobjs $deplibs $compiler_flags ${wl}-soname $wl$soname -o $lib' _LT_TAGVAR(link_all_deplibs, $1)=yes ;; hpux9*) _LT_TAGVAR(hardcode_libdir_flag_spec, $1)='${wl}+b ${wl}$libdir' _LT_TAGVAR(hardcode_libdir_separator, $1)=: _LT_TAGVAR(export_dynamic_flag_spec, $1)='${wl}-E' _LT_TAGVAR(hardcode_direct, $1)=yes _LT_TAGVAR(hardcode_minus_L, $1)=yes # Not in the search PATH, # but as the default # location of the library. case $cc_basename in CC*) # FIXME: insert proper C++ library support _LT_TAGVAR(ld_shlibs, $1)=no ;; aCC*) _LT_TAGVAR(archive_cmds, $1)='$RM $output_objdir/$soname~$CC -b ${wl}+b ${wl}$install_libdir -o $output_objdir/$soname $predep_objects $libobjs $deplibs $postdep_objects $compiler_flags~test $output_objdir/$soname = $lib || mv $output_objdir/$soname $lib' # Commands to make compiler produce verbose output that lists # what "hidden" libraries, object files and flags are used when # linking a shared library. # # There doesn't appear to be a way to prevent this compiler from # explicitly linking system object files so we need to strip them # from the output so that they don't get included in the library # dependencies. output_verbose_link_cmd='templist=`($CC -b $CFLAGS -v conftest.$objext 2>&1) | $EGREP "\-L"`; list=""; for z in $templist; do case $z in conftest.$objext) list="$list $z";; *.$objext);; *) list="$list $z";;esac; done; func_echo_all "$list"' ;; *) if test "$GXX" = yes; then _LT_TAGVAR(archive_cmds, $1)='$RM $output_objdir/$soname~$CC -shared -nostdlib $pic_flag ${wl}+b ${wl}$install_libdir -o $output_objdir/$soname $predep_objects $libobjs $deplibs $postdep_objects $compiler_flags~test $output_objdir/$soname = $lib || mv $output_objdir/$soname $lib' else # FIXME: insert proper C++ library support _LT_TAGVAR(ld_shlibs, $1)=no fi ;; esac ;; hpux10*|hpux11*) if test $with_gnu_ld = no; then _LT_TAGVAR(hardcode_libdir_flag_spec, $1)='${wl}+b ${wl}$libdir' _LT_TAGVAR(hardcode_libdir_separator, $1)=: case $host_cpu in hppa*64*|ia64*) ;; *) _LT_TAGVAR(export_dynamic_flag_spec, $1)='${wl}-E' ;; esac fi case $host_cpu in hppa*64*|ia64*) _LT_TAGVAR(hardcode_direct, $1)=no _LT_TAGVAR(hardcode_shlibpath_var, $1)=no ;; *) _LT_TAGVAR(hardcode_direct, $1)=yes _LT_TAGVAR(hardcode_direct_absolute, $1)=yes _LT_TAGVAR(hardcode_minus_L, $1)=yes # Not in the search PATH, # but as the default # location of the library. ;; esac case $cc_basename in CC*) # FIXME: insert proper C++ library support _LT_TAGVAR(ld_shlibs, $1)=no ;; aCC*) case $host_cpu in hppa*64*) _LT_TAGVAR(archive_cmds, $1)='$CC -b ${wl}+h ${wl}$soname -o $lib $predep_objects $libobjs $deplibs $postdep_objects $compiler_flags' ;; ia64*) _LT_TAGVAR(archive_cmds, $1)='$CC -b ${wl}+h ${wl}$soname ${wl}+nodefaultrpath -o $lib $predep_objects $libobjs $deplibs $postdep_objects $compiler_flags' ;; *) _LT_TAGVAR(archive_cmds, $1)='$CC -b ${wl}+h ${wl}$soname ${wl}+b ${wl}$install_libdir -o $lib $predep_objects $libobjs $deplibs $postdep_objects $compiler_flags' ;; esac # Commands to make compiler produce verbose output that lists # what "hidden" libraries, object files and flags are used when # linking a shared library. # # There doesn't appear to be a way to prevent this compiler from # explicitly linking system object files so we need to strip them # from the output so that they don't get included in the library # dependencies. output_verbose_link_cmd='templist=`($CC -b $CFLAGS -v conftest.$objext 2>&1) | $GREP "\-L"`; list=""; for z in $templist; do case $z in conftest.$objext) list="$list $z";; *.$objext);; *) list="$list $z";;esac; done; func_echo_all "$list"' ;; *) if test "$GXX" = yes; then if test $with_gnu_ld = no; then case $host_cpu in hppa*64*) _LT_TAGVAR(archive_cmds, $1)='$CC -shared -nostdlib -fPIC ${wl}+h ${wl}$soname -o $lib $predep_objects $libobjs $deplibs $postdep_objects $compiler_flags' ;; ia64*) _LT_TAGVAR(archive_cmds, $1)='$CC -shared -nostdlib $pic_flag ${wl}+h ${wl}$soname ${wl}+nodefaultrpath -o $lib $predep_objects $libobjs $deplibs $postdep_objects $compiler_flags' ;; *) _LT_TAGVAR(archive_cmds, $1)='$CC -shared -nostdlib $pic_flag ${wl}+h ${wl}$soname ${wl}+b ${wl}$install_libdir -o $lib $predep_objects $libobjs $deplibs $postdep_objects $compiler_flags' ;; esac fi else # FIXME: insert proper C++ library support _LT_TAGVAR(ld_shlibs, $1)=no fi ;; esac ;; interix[[3-9]]*) _LT_TAGVAR(hardcode_direct, $1)=no _LT_TAGVAR(hardcode_shlibpath_var, $1)=no _LT_TAGVAR(hardcode_libdir_flag_spec, $1)='${wl}-rpath,$libdir' _LT_TAGVAR(export_dynamic_flag_spec, $1)='${wl}-E' # Hack: On Interix 3.x, we cannot compile PIC because of a broken gcc. # Instead, shared libraries are loaded at an image base (0x10000000 by # default) and relocated if they conflict, which is a slow very memory # consuming and fragmenting process. To avoid this, we pick a random, # 256 KiB-aligned image base between 0x50000000 and 0x6FFC0000 at link # time. Moving up from 0x10000000 also allows more sbrk(2) space. _LT_TAGVAR(archive_cmds, $1)='$CC -shared $pic_flag $libobjs $deplibs $compiler_flags ${wl}-h,$soname ${wl}--image-base,`expr ${RANDOM-$$} % 4096 / 2 \* 262144 + 1342177280` -o $lib' _LT_TAGVAR(archive_expsym_cmds, $1)='sed "s,^,_," $export_symbols >$output_objdir/$soname.expsym~$CC -shared $pic_flag $libobjs $deplibs $compiler_flags ${wl}-h,$soname ${wl}--retain-symbols-file,$output_objdir/$soname.expsym ${wl}--image-base,`expr ${RANDOM-$$} % 4096 / 2 \* 262144 + 1342177280` -o $lib' ;; irix5* | irix6*) case $cc_basename in CC*) # SGI C++ _LT_TAGVAR(archive_cmds, $1)='$CC -shared -all -multigot $predep_objects $libobjs $deplibs $postdep_objects $compiler_flags -soname $soname `test -n "$verstring" && func_echo_all "-set_version $verstring"` -update_registry ${output_objdir}/so_locations -o $lib' # Archives containing C++ object files must be created using # "CC -ar", where "CC" is the IRIX C++ compiler. This is # necessary to make sure instantiated templates are included # in the archive. _LT_TAGVAR(old_archive_cmds, $1)='$CC -ar -WR,-u -o $oldlib $oldobjs' ;; *) if test "$GXX" = yes; then if test "$with_gnu_ld" = no; then _LT_TAGVAR(archive_cmds, $1)='$CC -shared $pic_flag -nostdlib $predep_objects $libobjs $deplibs $postdep_objects $compiler_flags ${wl}-soname ${wl}$soname `test -n "$verstring" && func_echo_all "${wl}-set_version ${wl}$verstring"` ${wl}-update_registry ${wl}${output_objdir}/so_locations -o $lib' else _LT_TAGVAR(archive_cmds, $1)='$CC -shared $pic_flag -nostdlib $predep_objects $libobjs $deplibs $postdep_objects $compiler_flags ${wl}-soname ${wl}$soname `test -n "$verstring" && func_echo_all "${wl}-set_version ${wl}$verstring"` -o $lib' fi fi _LT_TAGVAR(link_all_deplibs, $1)=yes ;; esac _LT_TAGVAR(hardcode_libdir_flag_spec, $1)='${wl}-rpath ${wl}$libdir' _LT_TAGVAR(hardcode_libdir_separator, $1)=: _LT_TAGVAR(inherit_rpath, $1)=yes ;; linux* | k*bsd*-gnu | kopensolaris*-gnu | gnu*) case $cc_basename in KCC*) # Kuck and Associates, Inc. (KAI) C++ Compiler # KCC will only create a shared library if the output file # ends with ".so" (or ".sl" for HP-UX), so rename the library # to its proper name (with version) after linking. _LT_TAGVAR(archive_cmds, $1)='tempext=`echo $shared_ext | $SED -e '\''s/\([[^()0-9A-Za-z{}]]\)/\\\\\1/g'\''`; templib=`echo $lib | $SED -e "s/\${tempext}\..*/.so/"`; $CC $predep_objects $libobjs $deplibs $postdep_objects $compiler_flags --soname $soname -o \$templib; mv \$templib $lib' _LT_TAGVAR(archive_expsym_cmds, $1)='tempext=`echo $shared_ext | $SED -e '\''s/\([[^()0-9A-Za-z{}]]\)/\\\\\1/g'\''`; templib=`echo $lib | $SED -e "s/\${tempext}\..*/.so/"`; $CC $predep_objects $libobjs $deplibs $postdep_objects $compiler_flags --soname $soname -o \$templib ${wl}-retain-symbols-file,$export_symbols; mv \$templib $lib' # Commands to make compiler produce verbose output that lists # what "hidden" libraries, object files and flags are used when # linking a shared library. # # There doesn't appear to be a way to prevent this compiler from # explicitly linking system object files so we need to strip them # from the output so that they don't get included in the library # dependencies. output_verbose_link_cmd='templist=`$CC $CFLAGS -v conftest.$objext -o libconftest$shared_ext 2>&1 | $GREP "ld"`; rm -f libconftest$shared_ext; list=""; for z in $templist; do case $z in conftest.$objext) list="$list $z";; *.$objext);; *) list="$list $z";;esac; done; func_echo_all "$list"' _LT_TAGVAR(hardcode_libdir_flag_spec, $1)='${wl}-rpath,$libdir' _LT_TAGVAR(export_dynamic_flag_spec, $1)='${wl}--export-dynamic' # Archives containing C++ object files must be created using # "CC -Bstatic", where "CC" is the KAI C++ compiler. _LT_TAGVAR(old_archive_cmds, $1)='$CC -Bstatic -o $oldlib $oldobjs' ;; icpc* | ecpc* ) # Intel C++ with_gnu_ld=yes # version 8.0 and above of icpc choke on multiply defined symbols # if we add $predep_objects and $postdep_objects, however 7.1 and # earlier do not add the objects themselves. case `$CC -V 2>&1` in *"Version 7."*) _LT_TAGVAR(archive_cmds, $1)='$CC -shared $predep_objects $libobjs $deplibs $postdep_objects $compiler_flags ${wl}-soname $wl$soname -o $lib' _LT_TAGVAR(archive_expsym_cmds, $1)='$CC -shared $predep_objects $libobjs $deplibs $postdep_objects $compiler_flags ${wl}-soname $wl$soname ${wl}-retain-symbols-file $wl$export_symbols -o $lib' ;; *) # Version 8.0 or newer tmp_idyn= case $host_cpu in ia64*) tmp_idyn=' -i_dynamic';; esac _LT_TAGVAR(archive_cmds, $1)='$CC -shared'"$tmp_idyn"' $libobjs $deplibs $compiler_flags ${wl}-soname $wl$soname -o $lib' _LT_TAGVAR(archive_expsym_cmds, $1)='$CC -shared'"$tmp_idyn"' $libobjs $deplibs $compiler_flags ${wl}-soname $wl$soname ${wl}-retain-symbols-file $wl$export_symbols -o $lib' ;; esac _LT_TAGVAR(archive_cmds_need_lc, $1)=no _LT_TAGVAR(hardcode_libdir_flag_spec, $1)='${wl}-rpath,$libdir' _LT_TAGVAR(export_dynamic_flag_spec, $1)='${wl}--export-dynamic' _LT_TAGVAR(whole_archive_flag_spec, $1)='${wl}--whole-archive$convenience ${wl}--no-whole-archive' ;; pgCC* | pgcpp*) # Portland Group C++ compiler case `$CC -V` in *pgCC\ [[1-5]].* | *pgcpp\ [[1-5]].*) _LT_TAGVAR(prelink_cmds, $1)='tpldir=Template.dir~ rm -rf $tpldir~ $CC --prelink_objects --instantiation_dir $tpldir $objs $libobjs $compile_deplibs~ compile_command="$compile_command `find $tpldir -name \*.o | sort | $NL2SP`"' _LT_TAGVAR(old_archive_cmds, $1)='tpldir=Template.dir~ rm -rf $tpldir~ $CC --prelink_objects --instantiation_dir $tpldir $oldobjs$old_deplibs~ $AR $AR_FLAGS $oldlib$oldobjs$old_deplibs `find $tpldir -name \*.o | sort | $NL2SP`~ $RANLIB $oldlib' _LT_TAGVAR(archive_cmds, $1)='tpldir=Template.dir~ rm -rf $tpldir~ $CC --prelink_objects --instantiation_dir $tpldir $predep_objects $libobjs $deplibs $convenience $postdep_objects~ $CC -shared $pic_flag $predep_objects $libobjs $deplibs `find $tpldir -name \*.o | sort | $NL2SP` $postdep_objects $compiler_flags ${wl}-soname ${wl}$soname -o $lib' _LT_TAGVAR(archive_expsym_cmds, $1)='tpldir=Template.dir~ rm -rf $tpldir~ $CC --prelink_objects --instantiation_dir $tpldir $predep_objects $libobjs $deplibs $convenience $postdep_objects~ $CC -shared $pic_flag $predep_objects $libobjs $deplibs `find $tpldir -name \*.o | sort | $NL2SP` $postdep_objects $compiler_flags ${wl}-soname ${wl}$soname ${wl}-retain-symbols-file ${wl}$export_symbols -o $lib' ;; *) # Version 6 and above use weak symbols _LT_TAGVAR(archive_cmds, $1)='$CC -shared $pic_flag $predep_objects $libobjs $deplibs $postdep_objects $compiler_flags ${wl}-soname ${wl}$soname -o $lib' _LT_TAGVAR(archive_expsym_cmds, $1)='$CC -shared $pic_flag $predep_objects $libobjs $deplibs $postdep_objects $compiler_flags ${wl}-soname ${wl}$soname ${wl}-retain-symbols-file ${wl}$export_symbols -o $lib' ;; esac _LT_TAGVAR(hardcode_libdir_flag_spec, $1)='${wl}--rpath ${wl}$libdir' _LT_TAGVAR(export_dynamic_flag_spec, $1)='${wl}--export-dynamic' _LT_TAGVAR(whole_archive_flag_spec, $1)='${wl}--whole-archive`for conv in $convenience\"\"; do test -n \"$conv\" && new_convenience=\"$new_convenience,$conv\"; done; func_echo_all \"$new_convenience\"` ${wl}--no-whole-archive' ;; cxx*) # Compaq C++ _LT_TAGVAR(archive_cmds, $1)='$CC -shared $predep_objects $libobjs $deplibs $postdep_objects $compiler_flags ${wl}-soname $wl$soname -o $lib' _LT_TAGVAR(archive_expsym_cmds, $1)='$CC -shared $predep_objects $libobjs $deplibs $postdep_objects $compiler_flags ${wl}-soname $wl$soname -o $lib ${wl}-retain-symbols-file $wl$export_symbols' runpath_var=LD_RUN_PATH _LT_TAGVAR(hardcode_libdir_flag_spec, $1)='-rpath $libdir' _LT_TAGVAR(hardcode_libdir_separator, $1)=: # Commands to make compiler produce verbose output that lists # what "hidden" libraries, object files and flags are used when # linking a shared library. # # There doesn't appear to be a way to prevent this compiler from # explicitly linking system object files so we need to strip them # from the output so that they don't get included in the library # dependencies. output_verbose_link_cmd='templist=`$CC -shared $CFLAGS -v conftest.$objext 2>&1 | $GREP "ld"`; templist=`func_echo_all "$templist" | $SED "s/\(^.*ld.*\)\( .*ld .*$\)/\1/"`; list=""; for z in $templist; do case $z in conftest.$objext) list="$list $z";; *.$objext);; *) list="$list $z";;esac; done; func_echo_all "X$list" | $Xsed' ;; xl* | mpixl* | bgxl*) # IBM XL 8.0 on PPC, with GNU ld _LT_TAGVAR(hardcode_libdir_flag_spec, $1)='${wl}-rpath ${wl}$libdir' _LT_TAGVAR(export_dynamic_flag_spec, $1)='${wl}--export-dynamic' _LT_TAGVAR(archive_cmds, $1)='$CC -qmkshrobj $libobjs $deplibs $compiler_flags ${wl}-soname $wl$soname -o $lib' if test "x$supports_anon_versioning" = xyes; then _LT_TAGVAR(archive_expsym_cmds, $1)='echo "{ global:" > $output_objdir/$libname.ver~ cat $export_symbols | sed -e "s/\(.*\)/\1;/" >> $output_objdir/$libname.ver~ echo "local: *; };" >> $output_objdir/$libname.ver~ $CC -qmkshrobj $libobjs $deplibs $compiler_flags ${wl}-soname $wl$soname ${wl}-version-script ${wl}$output_objdir/$libname.ver -o $lib' fi ;; *) case `$CC -V 2>&1 | sed 5q` in *Sun\ C*) # Sun C++ 5.9 _LT_TAGVAR(no_undefined_flag, $1)=' -zdefs' _LT_TAGVAR(archive_cmds, $1)='$CC -G${allow_undefined_flag} -h$soname -o $lib $predep_objects $libobjs $deplibs $postdep_objects $compiler_flags' _LT_TAGVAR(archive_expsym_cmds, $1)='$CC -G${allow_undefined_flag} -h$soname -o $lib $predep_objects $libobjs $deplibs $postdep_objects $compiler_flags ${wl}-retain-symbols-file ${wl}$export_symbols' _LT_TAGVAR(hardcode_libdir_flag_spec, $1)='-R$libdir' _LT_TAGVAR(whole_archive_flag_spec, $1)='${wl}--whole-archive`new_convenience=; for conv in $convenience\"\"; do test -z \"$conv\" || new_convenience=\"$new_convenience,$conv\"; done; func_echo_all \"$new_convenience\"` ${wl}--no-whole-archive' _LT_TAGVAR(compiler_needs_object, $1)=yes # Not sure whether something based on # $CC $CFLAGS -v conftest.$objext -o libconftest$shared_ext 2>&1 # would be better. output_verbose_link_cmd='func_echo_all' # Archives containing C++ object files must be created using # "CC -xar", where "CC" is the Sun C++ compiler. This is # necessary to make sure instantiated templates are included # in the archive. _LT_TAGVAR(old_archive_cmds, $1)='$CC -xar -o $oldlib $oldobjs' ;; esac ;; esac ;; lynxos*) # FIXME: insert proper C++ library support _LT_TAGVAR(ld_shlibs, $1)=no ;; m88k*) # FIXME: insert proper C++ library support _LT_TAGVAR(ld_shlibs, $1)=no ;; mvs*) case $cc_basename in cxx*) # FIXME: insert proper C++ library support _LT_TAGVAR(ld_shlibs, $1)=no ;; *) # FIXME: insert proper C++ library support _LT_TAGVAR(ld_shlibs, $1)=no ;; esac ;; netbsd*) if echo __ELF__ | $CC -E - | $GREP __ELF__ >/dev/null; then _LT_TAGVAR(archive_cmds, $1)='$LD -Bshareable -o $lib $predep_objects $libobjs $deplibs $postdep_objects $linker_flags' wlarc= _LT_TAGVAR(hardcode_libdir_flag_spec, $1)='-R$libdir' _LT_TAGVAR(hardcode_direct, $1)=yes _LT_TAGVAR(hardcode_shlibpath_var, $1)=no fi # Workaround some broken pre-1.5 toolchains output_verbose_link_cmd='$CC -shared $CFLAGS -v conftest.$objext 2>&1 | $GREP conftest.$objext | $SED -e "s:-lgcc -lc -lgcc::"' ;; *nto* | *qnx*) _LT_TAGVAR(ld_shlibs, $1)=yes ;; openbsd2*) # C++ shared libraries are fairly broken _LT_TAGVAR(ld_shlibs, $1)=no ;; openbsd*) if test -f /usr/libexec/ld.so; then _LT_TAGVAR(hardcode_direct, $1)=yes _LT_TAGVAR(hardcode_shlibpath_var, $1)=no _LT_TAGVAR(hardcode_direct_absolute, $1)=yes _LT_TAGVAR(archive_cmds, $1)='$CC -shared $pic_flag $predep_objects $libobjs $deplibs $postdep_objects $compiler_flags -o $lib' _LT_TAGVAR(hardcode_libdir_flag_spec, $1)='${wl}-rpath,$libdir' if test -z "`echo __ELF__ | $CC -E - | grep __ELF__`" || test "$host_os-$host_cpu" = "openbsd2.8-powerpc"; then _LT_TAGVAR(archive_expsym_cmds, $1)='$CC -shared $pic_flag $predep_objects $libobjs $deplibs $postdep_objects $compiler_flags ${wl}-retain-symbols-file,$export_symbols -o $lib' _LT_TAGVAR(export_dynamic_flag_spec, $1)='${wl}-E' _LT_TAGVAR(whole_archive_flag_spec, $1)="$wlarc"'--whole-archive$convenience '"$wlarc"'--no-whole-archive' fi output_verbose_link_cmd=func_echo_all else _LT_TAGVAR(ld_shlibs, $1)=no fi ;; osf3* | osf4* | osf5*) case $cc_basename in KCC*) # Kuck and Associates, Inc. (KAI) C++ Compiler # KCC will only create a shared library if the output file # ends with ".so" (or ".sl" for HP-UX), so rename the library # to its proper name (with version) after linking. _LT_TAGVAR(archive_cmds, $1)='tempext=`echo $shared_ext | $SED -e '\''s/\([[^()0-9A-Za-z{}]]\)/\\\\\1/g'\''`; templib=`echo "$lib" | $SED -e "s/\${tempext}\..*/.so/"`; $CC $predep_objects $libobjs $deplibs $postdep_objects $compiler_flags --soname $soname -o \$templib; mv \$templib $lib' _LT_TAGVAR(hardcode_libdir_flag_spec, $1)='${wl}-rpath,$libdir' _LT_TAGVAR(hardcode_libdir_separator, $1)=: # Archives containing C++ object files must be created using # the KAI C++ compiler. case $host in osf3*) _LT_TAGVAR(old_archive_cmds, $1)='$CC -Bstatic -o $oldlib $oldobjs' ;; *) _LT_TAGVAR(old_archive_cmds, $1)='$CC -o $oldlib $oldobjs' ;; esac ;; RCC*) # Rational C++ 2.4.1 # FIXME: insert proper C++ library support _LT_TAGVAR(ld_shlibs, $1)=no ;; cxx*) case $host in osf3*) _LT_TAGVAR(allow_undefined_flag, $1)=' ${wl}-expect_unresolved ${wl}\*' _LT_TAGVAR(archive_cmds, $1)='$CC -shared${allow_undefined_flag} $predep_objects $libobjs $deplibs $postdep_objects $compiler_flags ${wl}-soname $soname `test -n "$verstring" && func_echo_all "${wl}-set_version $verstring"` -update_registry ${output_objdir}/so_locations -o $lib' _LT_TAGVAR(hardcode_libdir_flag_spec, $1)='${wl}-rpath ${wl}$libdir' ;; *) _LT_TAGVAR(allow_undefined_flag, $1)=' -expect_unresolved \*' _LT_TAGVAR(archive_cmds, $1)='$CC -shared${allow_undefined_flag} $predep_objects $libobjs $deplibs $postdep_objects $compiler_flags -msym -soname $soname `test -n "$verstring" && func_echo_all "-set_version $verstring"` -update_registry ${output_objdir}/so_locations -o $lib' _LT_TAGVAR(archive_expsym_cmds, $1)='for i in `cat $export_symbols`; do printf "%s %s\\n" -exported_symbol "\$i" >> $lib.exp; done~ echo "-hidden">> $lib.exp~ $CC -shared$allow_undefined_flag $predep_objects $libobjs $deplibs $postdep_objects $compiler_flags -msym -soname $soname ${wl}-input ${wl}$lib.exp `test -n "$verstring" && $ECHO "-set_version $verstring"` -update_registry ${output_objdir}/so_locations -o $lib~ $RM $lib.exp' _LT_TAGVAR(hardcode_libdir_flag_spec, $1)='-rpath $libdir' ;; esac _LT_TAGVAR(hardcode_libdir_separator, $1)=: # Commands to make compiler produce verbose output that lists # what "hidden" libraries, object files and flags are used when # linking a shared library. # # There doesn't appear to be a way to prevent this compiler from # explicitly linking system object files so we need to strip them # from the output so that they don't get included in the library # dependencies. output_verbose_link_cmd='templist=`$CC -shared $CFLAGS -v conftest.$objext 2>&1 | $GREP "ld" | $GREP -v "ld:"`; templist=`func_echo_all "$templist" | $SED "s/\(^.*ld.*\)\( .*ld.*$\)/\1/"`; list=""; for z in $templist; do case $z in conftest.$objext) list="$list $z";; *.$objext);; *) list="$list $z";;esac; done; func_echo_all "$list"' ;; *) if test "$GXX" = yes && test "$with_gnu_ld" = no; then _LT_TAGVAR(allow_undefined_flag, $1)=' ${wl}-expect_unresolved ${wl}\*' case $host in osf3*) _LT_TAGVAR(archive_cmds, $1)='$CC -shared -nostdlib ${allow_undefined_flag} $predep_objects $libobjs $deplibs $postdep_objects $compiler_flags ${wl}-soname ${wl}$soname `test -n "$verstring" && func_echo_all "${wl}-set_version ${wl}$verstring"` ${wl}-update_registry ${wl}${output_objdir}/so_locations -o $lib' ;; *) _LT_TAGVAR(archive_cmds, $1)='$CC -shared $pic_flag -nostdlib ${allow_undefined_flag} $predep_objects $libobjs $deplibs $postdep_objects $compiler_flags ${wl}-msym ${wl}-soname ${wl}$soname `test -n "$verstring" && func_echo_all "${wl}-set_version ${wl}$verstring"` ${wl}-update_registry ${wl}${output_objdir}/so_locations -o $lib' ;; esac _LT_TAGVAR(hardcode_libdir_flag_spec, $1)='${wl}-rpath ${wl}$libdir' _LT_TAGVAR(hardcode_libdir_separator, $1)=: # Commands to make compiler produce verbose output that lists # what "hidden" libraries, object files and flags are used when # linking a shared library. output_verbose_link_cmd='$CC -shared $CFLAGS -v conftest.$objext 2>&1 | $GREP -v "^Configured with:" | $GREP "\-L"' else # FIXME: insert proper C++ library support _LT_TAGVAR(ld_shlibs, $1)=no fi ;; esac ;; psos*) # FIXME: insert proper C++ library support _LT_TAGVAR(ld_shlibs, $1)=no ;; sunos4*) case $cc_basename in CC*) # Sun C++ 4.x # FIXME: insert proper C++ library support _LT_TAGVAR(ld_shlibs, $1)=no ;; lcc*) # Lucid # FIXME: insert proper C++ library support _LT_TAGVAR(ld_shlibs, $1)=no ;; *) # FIXME: insert proper C++ library support _LT_TAGVAR(ld_shlibs, $1)=no ;; esac ;; solaris*) case $cc_basename in CC* | sunCC*) # Sun C++ 4.2, 5.x and Centerline C++ _LT_TAGVAR(archive_cmds_need_lc,$1)=yes _LT_TAGVAR(no_undefined_flag, $1)=' -zdefs' _LT_TAGVAR(archive_cmds, $1)='$CC -G${allow_undefined_flag} -h$soname -o $lib $predep_objects $libobjs $deplibs $postdep_objects $compiler_flags' _LT_TAGVAR(archive_expsym_cmds, $1)='echo "{ global:" > $lib.exp~cat $export_symbols | $SED -e "s/\(.*\)/\1;/" >> $lib.exp~echo "local: *; };" >> $lib.exp~ $CC -G${allow_undefined_flag} ${wl}-M ${wl}$lib.exp -h$soname -o $lib $predep_objects $libobjs $deplibs $postdep_objects $compiler_flags~$RM $lib.exp' _LT_TAGVAR(hardcode_libdir_flag_spec, $1)='-R$libdir' _LT_TAGVAR(hardcode_shlibpath_var, $1)=no case $host_os in solaris2.[[0-5]] | solaris2.[[0-5]].*) ;; *) # The compiler driver will combine and reorder linker options, # but understands `-z linker_flag'. # Supported since Solaris 2.6 (maybe 2.5.1?) _LT_TAGVAR(whole_archive_flag_spec, $1)='-z allextract$convenience -z defaultextract' ;; esac _LT_TAGVAR(link_all_deplibs, $1)=yes output_verbose_link_cmd='func_echo_all' # Archives containing C++ object files must be created using # "CC -xar", where "CC" is the Sun C++ compiler. This is # necessary to make sure instantiated templates are included # in the archive. _LT_TAGVAR(old_archive_cmds, $1)='$CC -xar -o $oldlib $oldobjs' ;; gcx*) # Green Hills C++ Compiler _LT_TAGVAR(archive_cmds, $1)='$CC -shared $predep_objects $libobjs $deplibs $postdep_objects $compiler_flags ${wl}-h $wl$soname -o $lib' # The C++ compiler must be used to create the archive. _LT_TAGVAR(old_archive_cmds, $1)='$CC $LDFLAGS -archive -o $oldlib $oldobjs' ;; *) # GNU C++ compiler with Solaris linker if test "$GXX" = yes && test "$with_gnu_ld" = no; then _LT_TAGVAR(no_undefined_flag, $1)=' ${wl}-z ${wl}defs' if $CC --version | $GREP -v '^2\.7' > /dev/null; then _LT_TAGVAR(archive_cmds, $1)='$CC -shared $pic_flag -nostdlib $LDFLAGS $predep_objects $libobjs $deplibs $postdep_objects $compiler_flags ${wl}-h $wl$soname -o $lib' _LT_TAGVAR(archive_expsym_cmds, $1)='echo "{ global:" > $lib.exp~cat $export_symbols | $SED -e "s/\(.*\)/\1;/" >> $lib.exp~echo "local: *; };" >> $lib.exp~ $CC -shared $pic_flag -nostdlib ${wl}-M $wl$lib.exp -o $lib $predep_objects $libobjs $deplibs $postdep_objects $compiler_flags~$RM $lib.exp' # Commands to make compiler produce verbose output that lists # what "hidden" libraries, object files and flags are used when # linking a shared library. output_verbose_link_cmd='$CC -shared $CFLAGS -v conftest.$objext 2>&1 | $GREP -v "^Configured with:" | $GREP "\-L"' else # g++ 2.7 appears to require `-G' NOT `-shared' on this # platform. _LT_TAGVAR(archive_cmds, $1)='$CC -G -nostdlib $LDFLAGS $predep_objects $libobjs $deplibs $postdep_objects $compiler_flags ${wl}-h $wl$soname -o $lib' _LT_TAGVAR(archive_expsym_cmds, $1)='echo "{ global:" > $lib.exp~cat $export_symbols | $SED -e "s/\(.*\)/\1;/" >> $lib.exp~echo "local: *; };" >> $lib.exp~ $CC -G -nostdlib ${wl}-M $wl$lib.exp -o $lib $predep_objects $libobjs $deplibs $postdep_objects $compiler_flags~$RM $lib.exp' # Commands to make compiler produce verbose output that lists # what "hidden" libraries, object files and flags are used when # linking a shared library. output_verbose_link_cmd='$CC -G $CFLAGS -v conftest.$objext 2>&1 | $GREP -v "^Configured with:" | $GREP "\-L"' fi _LT_TAGVAR(hardcode_libdir_flag_spec, $1)='${wl}-R $wl$libdir' case $host_os in solaris2.[[0-5]] | solaris2.[[0-5]].*) ;; *) _LT_TAGVAR(whole_archive_flag_spec, $1)='${wl}-z ${wl}allextract$convenience ${wl}-z ${wl}defaultextract' ;; esac fi ;; esac ;; sysv4*uw2* | sysv5OpenUNIX* | sysv5UnixWare7.[[01]].[[10]]* | unixware7* | sco3.2v5.0.[[024]]*) _LT_TAGVAR(no_undefined_flag, $1)='${wl}-z,text' _LT_TAGVAR(archive_cmds_need_lc, $1)=no _LT_TAGVAR(hardcode_shlibpath_var, $1)=no runpath_var='LD_RUN_PATH' case $cc_basename in CC*) _LT_TAGVAR(archive_cmds, $1)='$CC -G ${wl}-h,$soname -o $lib $libobjs $deplibs $compiler_flags' _LT_TAGVAR(archive_expsym_cmds, $1)='$CC -G ${wl}-Bexport:$export_symbols ${wl}-h,$soname -o $lib $libobjs $deplibs $compiler_flags' ;; *) _LT_TAGVAR(archive_cmds, $1)='$CC -shared ${wl}-h,$soname -o $lib $libobjs $deplibs $compiler_flags' _LT_TAGVAR(archive_expsym_cmds, $1)='$CC -shared ${wl}-Bexport:$export_symbols ${wl}-h,$soname -o $lib $libobjs $deplibs $compiler_flags' ;; esac ;; sysv5* | sco3.2v5* | sco5v6*) # Note: We can NOT use -z defs as we might desire, because we do not # link with -lc, and that would cause any symbols used from libc to # always be unresolved, which means just about no library would # ever link correctly. If we're not using GNU ld we use -z text # though, which does catch some bad symbols but isn't as heavy-handed # as -z defs. _LT_TAGVAR(no_undefined_flag, $1)='${wl}-z,text' _LT_TAGVAR(allow_undefined_flag, $1)='${wl}-z,nodefs' _LT_TAGVAR(archive_cmds_need_lc, $1)=no _LT_TAGVAR(hardcode_shlibpath_var, $1)=no _LT_TAGVAR(hardcode_libdir_flag_spec, $1)='${wl}-R,$libdir' _LT_TAGVAR(hardcode_libdir_separator, $1)=':' _LT_TAGVAR(link_all_deplibs, $1)=yes _LT_TAGVAR(export_dynamic_flag_spec, $1)='${wl}-Bexport' runpath_var='LD_RUN_PATH' case $cc_basename in CC*) _LT_TAGVAR(archive_cmds, $1)='$CC -G ${wl}-h,$soname -o $lib $libobjs $deplibs $compiler_flags' _LT_TAGVAR(archive_expsym_cmds, $1)='$CC -G ${wl}-Bexport:$export_symbols ${wl}-h,$soname -o $lib $libobjs $deplibs $compiler_flags' _LT_TAGVAR(old_archive_cmds, $1)='$CC -Tprelink_objects $oldobjs~ '"$_LT_TAGVAR(old_archive_cmds, $1)" _LT_TAGVAR(reload_cmds, $1)='$CC -Tprelink_objects $reload_objs~ '"$_LT_TAGVAR(reload_cmds, $1)" ;; *) _LT_TAGVAR(archive_cmds, $1)='$CC -shared ${wl}-h,$soname -o $lib $libobjs $deplibs $compiler_flags' _LT_TAGVAR(archive_expsym_cmds, $1)='$CC -shared ${wl}-Bexport:$export_symbols ${wl}-h,$soname -o $lib $libobjs $deplibs $compiler_flags' ;; esac ;; tandem*) case $cc_basename in NCC*) # NonStop-UX NCC 3.20 # FIXME: insert proper C++ library support _LT_TAGVAR(ld_shlibs, $1)=no ;; *) # FIXME: insert proper C++ library support _LT_TAGVAR(ld_shlibs, $1)=no ;; esac ;; vxworks*) # FIXME: insert proper C++ library support _LT_TAGVAR(ld_shlibs, $1)=no ;; *) # FIXME: insert proper C++ library support _LT_TAGVAR(ld_shlibs, $1)=no ;; esac AC_MSG_RESULT([$_LT_TAGVAR(ld_shlibs, $1)]) test "$_LT_TAGVAR(ld_shlibs, $1)" = no && can_build_shared=no _LT_TAGVAR(GCC, $1)="$GXX" _LT_TAGVAR(LD, $1)="$LD" ## CAVEAT EMPTOR: ## There is no encapsulation within the following macros, do not change ## the running order or otherwise move them around unless you know exactly ## what you are doing... _LT_SYS_HIDDEN_LIBDEPS($1) _LT_COMPILER_PIC($1) _LT_COMPILER_C_O($1) _LT_COMPILER_FILE_LOCKS($1) _LT_LINKER_SHLIBS($1) _LT_SYS_DYNAMIC_LINKER($1) _LT_LINKER_HARDCODE_LIBPATH($1) _LT_CONFIG($1) fi # test -n "$compiler" CC=$lt_save_CC CFLAGS=$lt_save_CFLAGS LDCXX=$LD LD=$lt_save_LD GCC=$lt_save_GCC with_gnu_ld=$lt_save_with_gnu_ld lt_cv_path_LDCXX=$lt_cv_path_LD lt_cv_path_LD=$lt_save_path_LD lt_cv_prog_gnu_ldcxx=$lt_cv_prog_gnu_ld lt_cv_prog_gnu_ld=$lt_save_with_gnu_ld fi # test "$_lt_caught_CXX_error" != yes AC_LANG_POP ])# _LT_LANG_CXX_CONFIG # _LT_FUNC_STRIPNAME_CNF # ---------------------- # func_stripname_cnf prefix suffix name # strip PREFIX and SUFFIX off of NAME. # PREFIX and SUFFIX must not contain globbing or regex special # characters, hashes, percent signs, but SUFFIX may contain a leading # dot (in which case that matches only a dot). # # This function is identical to the (non-XSI) version of func_stripname, # except this one can be used by m4 code that may be executed by configure, # rather than the libtool script. m4_defun([_LT_FUNC_STRIPNAME_CNF],[dnl AC_REQUIRE([_LT_DECL_SED]) AC_REQUIRE([_LT_PROG_ECHO_BACKSLASH]) func_stripname_cnf () { case ${2} in .*) func_stripname_result=`$ECHO "${3}" | $SED "s%^${1}%%; s%\\\\${2}\$%%"`;; *) func_stripname_result=`$ECHO "${3}" | $SED "s%^${1}%%; s%${2}\$%%"`;; esac } # func_stripname_cnf ])# _LT_FUNC_STRIPNAME_CNF # _LT_SYS_HIDDEN_LIBDEPS([TAGNAME]) # --------------------------------- # Figure out "hidden" library dependencies from verbose # compiler output when linking a shared library. # Parse the compiler output and extract the necessary # objects, libraries and library flags. m4_defun([_LT_SYS_HIDDEN_LIBDEPS], [m4_require([_LT_FILEUTILS_DEFAULTS])dnl AC_REQUIRE([_LT_FUNC_STRIPNAME_CNF])dnl # Dependencies to place before and after the object being linked: _LT_TAGVAR(predep_objects, $1)= _LT_TAGVAR(postdep_objects, $1)= _LT_TAGVAR(predeps, $1)= _LT_TAGVAR(postdeps, $1)= _LT_TAGVAR(compiler_lib_search_path, $1)= dnl we can't use the lt_simple_compile_test_code here, dnl because it contains code intended for an executable, dnl not a library. It's possible we should let each dnl tag define a new lt_????_link_test_code variable, dnl but it's only used here... m4_if([$1], [], [cat > conftest.$ac_ext <<_LT_EOF int a; void foo (void) { a = 0; } _LT_EOF ], [$1], [CXX], [cat > conftest.$ac_ext <<_LT_EOF class Foo { public: Foo (void) { a = 0; } private: int a; }; _LT_EOF ], [$1], [F77], [cat > conftest.$ac_ext <<_LT_EOF subroutine foo implicit none integer*4 a a=0 return end _LT_EOF ], [$1], [FC], [cat > conftest.$ac_ext <<_LT_EOF subroutine foo implicit none integer a a=0 return end _LT_EOF ], [$1], [GCJ], [cat > conftest.$ac_ext <<_LT_EOF public class foo { private int a; public void bar (void) { a = 0; } }; _LT_EOF ], [$1], [GO], [cat > conftest.$ac_ext <<_LT_EOF package foo func foo() { } _LT_EOF ]) _lt_libdeps_save_CFLAGS=$CFLAGS case "$CC $CFLAGS " in #( *\ -flto*\ *) CFLAGS="$CFLAGS -fno-lto" ;; *\ -fwhopr*\ *) CFLAGS="$CFLAGS -fno-whopr" ;; *\ -fuse-linker-plugin*\ *) CFLAGS="$CFLAGS -fno-use-linker-plugin" ;; esac dnl Parse the compiler output and extract the necessary dnl objects, libraries and library flags. if AC_TRY_EVAL(ac_compile); then # Parse the compiler output and extract the necessary # objects, libraries and library flags. # Sentinel used to keep track of whether or not we are before # the conftest object file. pre_test_object_deps_done=no for p in `eval "$output_verbose_link_cmd"`; do case ${prev}${p} in -L* | -R* | -l*) # Some compilers place space between "-{L,R}" and the path. # Remove the space. if test $p = "-L" || test $p = "-R"; then prev=$p continue fi # Expand the sysroot to ease extracting the directories later. if test -z "$prev"; then case $p in -L*) func_stripname_cnf '-L' '' "$p"; prev=-L; p=$func_stripname_result ;; -R*) func_stripname_cnf '-R' '' "$p"; prev=-R; p=$func_stripname_result ;; -l*) func_stripname_cnf '-l' '' "$p"; prev=-l; p=$func_stripname_result ;; esac fi case $p in =*) func_stripname_cnf '=' '' "$p"; p=$lt_sysroot$func_stripname_result ;; esac if test "$pre_test_object_deps_done" = no; then case ${prev} in -L | -R) # Internal compiler library paths should come after those # provided the user. The postdeps already come after the # user supplied libs so there is no need to process them. if test -z "$_LT_TAGVAR(compiler_lib_search_path, $1)"; then _LT_TAGVAR(compiler_lib_search_path, $1)="${prev}${p}" else _LT_TAGVAR(compiler_lib_search_path, $1)="${_LT_TAGVAR(compiler_lib_search_path, $1)} ${prev}${p}" fi ;; # The "-l" case would never come before the object being # linked, so don't bother handling this case. esac else if test -z "$_LT_TAGVAR(postdeps, $1)"; then _LT_TAGVAR(postdeps, $1)="${prev}${p}" else _LT_TAGVAR(postdeps, $1)="${_LT_TAGVAR(postdeps, $1)} ${prev}${p}" fi fi prev= ;; *.lto.$objext) ;; # Ignore GCC LTO objects *.$objext) # This assumes that the test object file only shows up # once in the compiler output. if test "$p" = "conftest.$objext"; then pre_test_object_deps_done=yes continue fi if test "$pre_test_object_deps_done" = no; then if test -z "$_LT_TAGVAR(predep_objects, $1)"; then _LT_TAGVAR(predep_objects, $1)="$p" else _LT_TAGVAR(predep_objects, $1)="$_LT_TAGVAR(predep_objects, $1) $p" fi else if test -z "$_LT_TAGVAR(postdep_objects, $1)"; then _LT_TAGVAR(postdep_objects, $1)="$p" else _LT_TAGVAR(postdep_objects, $1)="$_LT_TAGVAR(postdep_objects, $1) $p" fi fi ;; *) ;; # Ignore the rest. esac done # Clean up. rm -f a.out a.exe else echo "libtool.m4: error: problem compiling $1 test program" fi $RM -f confest.$objext CFLAGS=$_lt_libdeps_save_CFLAGS # PORTME: override above test on systems where it is broken m4_if([$1], [CXX], [case $host_os in interix[[3-9]]*) # Interix 3.5 installs completely hosed .la files for C++, so rather than # hack all around it, let's just trust "g++" to DTRT. _LT_TAGVAR(predep_objects,$1)= _LT_TAGVAR(postdep_objects,$1)= _LT_TAGVAR(postdeps,$1)= ;; linux*) case `$CC -V 2>&1 | sed 5q` in *Sun\ C*) # Sun C++ 5.9 # The more standards-conforming stlport4 library is # incompatible with the Cstd library. Avoid specifying # it if it's in CXXFLAGS. Ignore libCrun as # -library=stlport4 depends on it. case " $CXX $CXXFLAGS " in *" -library=stlport4 "*) solaris_use_stlport4=yes ;; esac if test "$solaris_use_stlport4" != yes; then _LT_TAGVAR(postdeps,$1)='-library=Cstd -library=Crun' fi ;; esac ;; solaris*) case $cc_basename in CC* | sunCC*) # The more standards-conforming stlport4 library is # incompatible with the Cstd library. Avoid specifying # it if it's in CXXFLAGS. Ignore libCrun as # -library=stlport4 depends on it. case " $CXX $CXXFLAGS " in *" -library=stlport4 "*) solaris_use_stlport4=yes ;; esac # Adding this requires a known-good setup of shared libraries for # Sun compiler versions before 5.6, else PIC objects from an old # archive will be linked into the output, leading to subtle bugs. if test "$solaris_use_stlport4" != yes; then _LT_TAGVAR(postdeps,$1)='-library=Cstd -library=Crun' fi ;; esac ;; esac ]) case " $_LT_TAGVAR(postdeps, $1) " in *" -lc "*) _LT_TAGVAR(archive_cmds_need_lc, $1)=no ;; esac _LT_TAGVAR(compiler_lib_search_dirs, $1)= if test -n "${_LT_TAGVAR(compiler_lib_search_path, $1)}"; then _LT_TAGVAR(compiler_lib_search_dirs, $1)=`echo " ${_LT_TAGVAR(compiler_lib_search_path, $1)}" | ${SED} -e 's! -L! !g' -e 's!^ !!'` fi _LT_TAGDECL([], [compiler_lib_search_dirs], [1], [The directories searched by this compiler when creating a shared library]) _LT_TAGDECL([], [predep_objects], [1], [Dependencies to place before and after the objects being linked to create a shared library]) _LT_TAGDECL([], [postdep_objects], [1]) _LT_TAGDECL([], [predeps], [1]) _LT_TAGDECL([], [postdeps], [1]) _LT_TAGDECL([], [compiler_lib_search_path], [1], [The library search path used internally by the compiler when linking a shared library]) ])# _LT_SYS_HIDDEN_LIBDEPS # _LT_LANG_F77_CONFIG([TAG]) # -------------------------- # Ensure that the configuration variables for a Fortran 77 compiler are # suitably defined. These variables are subsequently used by _LT_CONFIG # to write the compiler configuration to `libtool'. m4_defun([_LT_LANG_F77_CONFIG], [AC_LANG_PUSH(Fortran 77) if test -z "$F77" || test "X$F77" = "Xno"; then _lt_disable_F77=yes fi _LT_TAGVAR(archive_cmds_need_lc, $1)=no _LT_TAGVAR(allow_undefined_flag, $1)= _LT_TAGVAR(always_export_symbols, $1)=no _LT_TAGVAR(archive_expsym_cmds, $1)= _LT_TAGVAR(export_dynamic_flag_spec, $1)= _LT_TAGVAR(hardcode_direct, $1)=no _LT_TAGVAR(hardcode_direct_absolute, $1)=no _LT_TAGVAR(hardcode_libdir_flag_spec, $1)= _LT_TAGVAR(hardcode_libdir_separator, $1)= _LT_TAGVAR(hardcode_minus_L, $1)=no _LT_TAGVAR(hardcode_automatic, $1)=no _LT_TAGVAR(inherit_rpath, $1)=no _LT_TAGVAR(module_cmds, $1)= _LT_TAGVAR(module_expsym_cmds, $1)= _LT_TAGVAR(link_all_deplibs, $1)=unknown _LT_TAGVAR(old_archive_cmds, $1)=$old_archive_cmds _LT_TAGVAR(reload_flag, $1)=$reload_flag _LT_TAGVAR(reload_cmds, $1)=$reload_cmds _LT_TAGVAR(no_undefined_flag, $1)= _LT_TAGVAR(whole_archive_flag_spec, $1)= _LT_TAGVAR(enable_shared_with_static_runtimes, $1)=no # Source file extension for f77 test sources. ac_ext=f # Object file extension for compiled f77 test sources. objext=o _LT_TAGVAR(objext, $1)=$objext # No sense in running all these tests if we already determined that # the F77 compiler isn't working. Some variables (like enable_shared) # are currently assumed to apply to all compilers on this platform, # and will be corrupted by setting them based on a non-working compiler. if test "$_lt_disable_F77" != yes; then # Code to be used in simple compile tests lt_simple_compile_test_code="\ subroutine t return end " # Code to be used in simple link tests lt_simple_link_test_code="\ program t end " # ltmain only uses $CC for tagged configurations so make sure $CC is set. _LT_TAG_COMPILER # save warnings/boilerplate of simple test code _LT_COMPILER_BOILERPLATE _LT_LINKER_BOILERPLATE # Allow CC to be a program name with arguments. lt_save_CC="$CC" lt_save_GCC=$GCC lt_save_CFLAGS=$CFLAGS CC=${F77-"f77"} CFLAGS=$FFLAGS compiler=$CC _LT_TAGVAR(compiler, $1)=$CC _LT_CC_BASENAME([$compiler]) GCC=$G77 if test -n "$compiler"; then AC_MSG_CHECKING([if libtool supports shared libraries]) AC_MSG_RESULT([$can_build_shared]) AC_MSG_CHECKING([whether to build shared libraries]) test "$can_build_shared" = "no" && enable_shared=no # On AIX, shared libraries and static libraries use the same namespace, and # are all built from PIC. case $host_os in aix3*) test "$enable_shared" = yes && enable_static=no if test -n "$RANLIB"; then archive_cmds="$archive_cmds~\$RANLIB \$lib" postinstall_cmds='$RANLIB $lib' fi ;; aix[[4-9]]*) if test "$host_cpu" != ia64 && test "$aix_use_runtimelinking" = no ; then test "$enable_shared" = yes && enable_static=no fi ;; esac AC_MSG_RESULT([$enable_shared]) AC_MSG_CHECKING([whether to build static libraries]) # Make sure either enable_shared or enable_static is yes. test "$enable_shared" = yes || enable_static=yes AC_MSG_RESULT([$enable_static]) _LT_TAGVAR(GCC, $1)="$G77" _LT_TAGVAR(LD, $1)="$LD" ## CAVEAT EMPTOR: ## There is no encapsulation within the following macros, do not change ## the running order or otherwise move them around unless you know exactly ## what you are doing... _LT_COMPILER_PIC($1) _LT_COMPILER_C_O($1) _LT_COMPILER_FILE_LOCKS($1) _LT_LINKER_SHLIBS($1) _LT_SYS_DYNAMIC_LINKER($1) _LT_LINKER_HARDCODE_LIBPATH($1) _LT_CONFIG($1) fi # test -n "$compiler" GCC=$lt_save_GCC CC="$lt_save_CC" CFLAGS="$lt_save_CFLAGS" fi # test "$_lt_disable_F77" != yes AC_LANG_POP ])# _LT_LANG_F77_CONFIG # _LT_LANG_FC_CONFIG([TAG]) # ------------------------- # Ensure that the configuration variables for a Fortran compiler are # suitably defined. These variables are subsequently used by _LT_CONFIG # to write the compiler configuration to `libtool'. m4_defun([_LT_LANG_FC_CONFIG], [AC_LANG_PUSH(Fortran) if test -z "$FC" || test "X$FC" = "Xno"; then _lt_disable_FC=yes fi _LT_TAGVAR(archive_cmds_need_lc, $1)=no _LT_TAGVAR(allow_undefined_flag, $1)= _LT_TAGVAR(always_export_symbols, $1)=no _LT_TAGVAR(archive_expsym_cmds, $1)= _LT_TAGVAR(export_dynamic_flag_spec, $1)= _LT_TAGVAR(hardcode_direct, $1)=no _LT_TAGVAR(hardcode_direct_absolute, $1)=no _LT_TAGVAR(hardcode_libdir_flag_spec, $1)= _LT_TAGVAR(hardcode_libdir_separator, $1)= _LT_TAGVAR(hardcode_minus_L, $1)=no _LT_TAGVAR(hardcode_automatic, $1)=no _LT_TAGVAR(inherit_rpath, $1)=no _LT_TAGVAR(module_cmds, $1)= _LT_TAGVAR(module_expsym_cmds, $1)= _LT_TAGVAR(link_all_deplibs, $1)=unknown _LT_TAGVAR(old_archive_cmds, $1)=$old_archive_cmds _LT_TAGVAR(reload_flag, $1)=$reload_flag _LT_TAGVAR(reload_cmds, $1)=$reload_cmds _LT_TAGVAR(no_undefined_flag, $1)= _LT_TAGVAR(whole_archive_flag_spec, $1)= _LT_TAGVAR(enable_shared_with_static_runtimes, $1)=no # Source file extension for fc test sources. ac_ext=${ac_fc_srcext-f} # Object file extension for compiled fc test sources. objext=o _LT_TAGVAR(objext, $1)=$objext # No sense in running all these tests if we already determined that # the FC compiler isn't working. Some variables (like enable_shared) # are currently assumed to apply to all compilers on this platform, # and will be corrupted by setting them based on a non-working compiler. if test "$_lt_disable_FC" != yes; then # Code to be used in simple compile tests lt_simple_compile_test_code="\ subroutine t return end " # Code to be used in simple link tests lt_simple_link_test_code="\ program t end " # ltmain only uses $CC for tagged configurations so make sure $CC is set. _LT_TAG_COMPILER # save warnings/boilerplate of simple test code _LT_COMPILER_BOILERPLATE _LT_LINKER_BOILERPLATE # Allow CC to be a program name with arguments. lt_save_CC="$CC" lt_save_GCC=$GCC lt_save_CFLAGS=$CFLAGS CC=${FC-"f95"} CFLAGS=$FCFLAGS compiler=$CC GCC=$ac_cv_fc_compiler_gnu _LT_TAGVAR(compiler, $1)=$CC _LT_CC_BASENAME([$compiler]) if test -n "$compiler"; then AC_MSG_CHECKING([if libtool supports shared libraries]) AC_MSG_RESULT([$can_build_shared]) AC_MSG_CHECKING([whether to build shared libraries]) test "$can_build_shared" = "no" && enable_shared=no # On AIX, shared libraries and static libraries use the same namespace, and # are all built from PIC. case $host_os in aix3*) test "$enable_shared" = yes && enable_static=no if test -n "$RANLIB"; then archive_cmds="$archive_cmds~\$RANLIB \$lib" postinstall_cmds='$RANLIB $lib' fi ;; aix[[4-9]]*) if test "$host_cpu" != ia64 && test "$aix_use_runtimelinking" = no ; then test "$enable_shared" = yes && enable_static=no fi ;; esac AC_MSG_RESULT([$enable_shared]) AC_MSG_CHECKING([whether to build static libraries]) # Make sure either enable_shared or enable_static is yes. test "$enable_shared" = yes || enable_static=yes AC_MSG_RESULT([$enable_static]) _LT_TAGVAR(GCC, $1)="$ac_cv_fc_compiler_gnu" _LT_TAGVAR(LD, $1)="$LD" ## CAVEAT EMPTOR: ## There is no encapsulation within the following macros, do not change ## the running order or otherwise move them around unless you know exactly ## what you are doing... _LT_SYS_HIDDEN_LIBDEPS($1) _LT_COMPILER_PIC($1) _LT_COMPILER_C_O($1) _LT_COMPILER_FILE_LOCKS($1) _LT_LINKER_SHLIBS($1) _LT_SYS_DYNAMIC_LINKER($1) _LT_LINKER_HARDCODE_LIBPATH($1) _LT_CONFIG($1) fi # test -n "$compiler" GCC=$lt_save_GCC CC=$lt_save_CC CFLAGS=$lt_save_CFLAGS fi # test "$_lt_disable_FC" != yes AC_LANG_POP ])# _LT_LANG_FC_CONFIG # _LT_LANG_GCJ_CONFIG([TAG]) # -------------------------- # Ensure that the configuration variables for the GNU Java Compiler compiler # are suitably defined. These variables are subsequently used by _LT_CONFIG # to write the compiler configuration to `libtool'. m4_defun([_LT_LANG_GCJ_CONFIG], [AC_REQUIRE([LT_PROG_GCJ])dnl AC_LANG_SAVE # Source file extension for Java test sources. ac_ext=java # Object file extension for compiled Java test sources. objext=o _LT_TAGVAR(objext, $1)=$objext # Code to be used in simple compile tests lt_simple_compile_test_code="class foo {}" # Code to be used in simple link tests lt_simple_link_test_code='public class conftest { public static void main(String[[]] argv) {}; }' # ltmain only uses $CC for tagged configurations so make sure $CC is set. _LT_TAG_COMPILER # save warnings/boilerplate of simple test code _LT_COMPILER_BOILERPLATE _LT_LINKER_BOILERPLATE # Allow CC to be a program name with arguments. lt_save_CC=$CC lt_save_CFLAGS=$CFLAGS lt_save_GCC=$GCC GCC=yes CC=${GCJ-"gcj"} CFLAGS=$GCJFLAGS compiler=$CC _LT_TAGVAR(compiler, $1)=$CC _LT_TAGVAR(LD, $1)="$LD" _LT_CC_BASENAME([$compiler]) # GCJ did not exist at the time GCC didn't implicitly link libc in. _LT_TAGVAR(archive_cmds_need_lc, $1)=no _LT_TAGVAR(old_archive_cmds, $1)=$old_archive_cmds _LT_TAGVAR(reload_flag, $1)=$reload_flag _LT_TAGVAR(reload_cmds, $1)=$reload_cmds ## CAVEAT EMPTOR: ## There is no encapsulation within the following macros, do not change ## the running order or otherwise move them around unless you know exactly ## what you are doing... if test -n "$compiler"; then _LT_COMPILER_NO_RTTI($1) _LT_COMPILER_PIC($1) _LT_COMPILER_C_O($1) _LT_COMPILER_FILE_LOCKS($1) _LT_LINKER_SHLIBS($1) _LT_LINKER_HARDCODE_LIBPATH($1) _LT_CONFIG($1) fi AC_LANG_RESTORE GCC=$lt_save_GCC CC=$lt_save_CC CFLAGS=$lt_save_CFLAGS ])# _LT_LANG_GCJ_CONFIG # _LT_LANG_GO_CONFIG([TAG]) # -------------------------- # Ensure that the configuration variables for the GNU Go compiler # are suitably defined. These variables are subsequently used by _LT_CONFIG # to write the compiler configuration to `libtool'. m4_defun([_LT_LANG_GO_CONFIG], [AC_REQUIRE([LT_PROG_GO])dnl AC_LANG_SAVE # Source file extension for Go test sources. ac_ext=go # Object file extension for compiled Go test sources. objext=o _LT_TAGVAR(objext, $1)=$objext # Code to be used in simple compile tests lt_simple_compile_test_code="package main; func main() { }" # Code to be used in simple link tests lt_simple_link_test_code='package main; func main() { }' # ltmain only uses $CC for tagged configurations so make sure $CC is set. _LT_TAG_COMPILER # save warnings/boilerplate of simple test code _LT_COMPILER_BOILERPLATE _LT_LINKER_BOILERPLATE # Allow CC to be a program name with arguments. lt_save_CC=$CC lt_save_CFLAGS=$CFLAGS lt_save_GCC=$GCC GCC=yes CC=${GOC-"gccgo"} CFLAGS=$GOFLAGS compiler=$CC _LT_TAGVAR(compiler, $1)=$CC _LT_TAGVAR(LD, $1)="$LD" _LT_CC_BASENAME([$compiler]) # Go did not exist at the time GCC didn't implicitly link libc in. _LT_TAGVAR(archive_cmds_need_lc, $1)=no _LT_TAGVAR(old_archive_cmds, $1)=$old_archive_cmds _LT_TAGVAR(reload_flag, $1)=$reload_flag _LT_TAGVAR(reload_cmds, $1)=$reload_cmds ## CAVEAT EMPTOR: ## There is no encapsulation within the following macros, do not change ## the running order or otherwise move them around unless you know exactly ## what you are doing... if test -n "$compiler"; then _LT_COMPILER_NO_RTTI($1) _LT_COMPILER_PIC($1) _LT_COMPILER_C_O($1) _LT_COMPILER_FILE_LOCKS($1) _LT_LINKER_SHLIBS($1) _LT_LINKER_HARDCODE_LIBPATH($1) _LT_CONFIG($1) fi AC_LANG_RESTORE GCC=$lt_save_GCC CC=$lt_save_CC CFLAGS=$lt_save_CFLAGS ])# _LT_LANG_GO_CONFIG # _LT_LANG_RC_CONFIG([TAG]) # ------------------------- # Ensure that the configuration variables for the Windows resource compiler # are suitably defined. These variables are subsequently used by _LT_CONFIG # to write the compiler configuration to `libtool'. m4_defun([_LT_LANG_RC_CONFIG], [AC_REQUIRE([LT_PROG_RC])dnl AC_LANG_SAVE # Source file extension for RC test sources. ac_ext=rc # Object file extension for compiled RC test sources. objext=o _LT_TAGVAR(objext, $1)=$objext # Code to be used in simple compile tests lt_simple_compile_test_code='sample MENU { MENUITEM "&Soup", 100, CHECKED }' # Code to be used in simple link tests lt_simple_link_test_code="$lt_simple_compile_test_code" # ltmain only uses $CC for tagged configurations so make sure $CC is set. _LT_TAG_COMPILER # save warnings/boilerplate of simple test code _LT_COMPILER_BOILERPLATE _LT_LINKER_BOILERPLATE # Allow CC to be a program name with arguments. lt_save_CC="$CC" lt_save_CFLAGS=$CFLAGS lt_save_GCC=$GCC GCC= CC=${RC-"windres"} CFLAGS= compiler=$CC _LT_TAGVAR(compiler, $1)=$CC _LT_CC_BASENAME([$compiler]) _LT_TAGVAR(lt_cv_prog_compiler_c_o, $1)=yes if test -n "$compiler"; then : _LT_CONFIG($1) fi GCC=$lt_save_GCC AC_LANG_RESTORE CC=$lt_save_CC CFLAGS=$lt_save_CFLAGS ])# _LT_LANG_RC_CONFIG # LT_PROG_GCJ # ----------- AC_DEFUN([LT_PROG_GCJ], [m4_ifdef([AC_PROG_GCJ], [AC_PROG_GCJ], [m4_ifdef([A][M_PROG_GCJ], [A][M_PROG_GCJ], [AC_CHECK_TOOL(GCJ, gcj,) test "x${GCJFLAGS+set}" = xset || GCJFLAGS="-g -O2" AC_SUBST(GCJFLAGS)])])[]dnl ]) # Old name: AU_ALIAS([LT_AC_PROG_GCJ], [LT_PROG_GCJ]) dnl aclocal-1.4 backwards compatibility: dnl AC_DEFUN([LT_AC_PROG_GCJ], []) # LT_PROG_GO # ---------- AC_DEFUN([LT_PROG_GO], [AC_CHECK_TOOL(GOC, gccgo,) ]) # LT_PROG_RC # ---------- AC_DEFUN([LT_PROG_RC], [AC_CHECK_TOOL(RC, windres,) ]) # Old name: AU_ALIAS([LT_AC_PROG_RC], [LT_PROG_RC]) dnl aclocal-1.4 backwards compatibility: dnl AC_DEFUN([LT_AC_PROG_RC], []) # _LT_DECL_EGREP # -------------- # If we don't have a new enough Autoconf to choose the best grep # available, choose the one first in the user's PATH. m4_defun([_LT_DECL_EGREP], [AC_REQUIRE([AC_PROG_EGREP])dnl AC_REQUIRE([AC_PROG_FGREP])dnl test -z "$GREP" && GREP=grep _LT_DECL([], [GREP], [1], [A grep program that handles long lines]) _LT_DECL([], [EGREP], [1], [An ERE matcher]) _LT_DECL([], [FGREP], [1], [A literal string matcher]) dnl Non-bleeding-edge autoconf doesn't subst GREP, so do it here too AC_SUBST([GREP]) ]) # _LT_DECL_OBJDUMP # -------------- # If we don't have a new enough Autoconf to choose the best objdump # available, choose the one first in the user's PATH. m4_defun([_LT_DECL_OBJDUMP], [AC_CHECK_TOOL(OBJDUMP, objdump, false) test -z "$OBJDUMP" && OBJDUMP=objdump _LT_DECL([], [OBJDUMP], [1], [An object symbol dumper]) AC_SUBST([OBJDUMP]) ]) # _LT_DECL_DLLTOOL # ---------------- # Ensure DLLTOOL variable is set. m4_defun([_LT_DECL_DLLTOOL], [AC_CHECK_TOOL(DLLTOOL, dlltool, false) test -z "$DLLTOOL" && DLLTOOL=dlltool _LT_DECL([], [DLLTOOL], [1], [DLL creation program]) AC_SUBST([DLLTOOL]) ]) # _LT_DECL_SED # ------------ # Check for a fully-functional sed program, that truncates # as few characters as possible. Prefer GNU sed if found. m4_defun([_LT_DECL_SED], [AC_PROG_SED test -z "$SED" && SED=sed Xsed="$SED -e 1s/^X//" _LT_DECL([], [SED], [1], [A sed program that does not truncate output]) _LT_DECL([], [Xsed], ["\$SED -e 1s/^X//"], [Sed that helps us avoid accidentally triggering echo(1) options like -n]) ])# _LT_DECL_SED m4_ifndef([AC_PROG_SED], [ ############################################################ # NOTE: This macro has been submitted for inclusion into # # GNU Autoconf as AC_PROG_SED. When it is available in # # a released version of Autoconf we should remove this # # macro and use it instead. # ############################################################ m4_defun([AC_PROG_SED], [AC_MSG_CHECKING([for a sed that does not truncate output]) AC_CACHE_VAL(lt_cv_path_SED, [# Loop through the user's path and test for sed and gsed. # Then use that list of sed's as ones to test for truncation. as_save_IFS=$IFS; IFS=$PATH_SEPARATOR for as_dir in $PATH do IFS=$as_save_IFS test -z "$as_dir" && as_dir=. for lt_ac_prog in sed gsed; do for ac_exec_ext in '' $ac_executable_extensions; do if $as_executable_p "$as_dir/$lt_ac_prog$ac_exec_ext"; then lt_ac_sed_list="$lt_ac_sed_list $as_dir/$lt_ac_prog$ac_exec_ext" fi done done done IFS=$as_save_IFS lt_ac_max=0 lt_ac_count=0 # Add /usr/xpg4/bin/sed as it is typically found on Solaris # along with /bin/sed that truncates output. for lt_ac_sed in $lt_ac_sed_list /usr/xpg4/bin/sed; do test ! -f $lt_ac_sed && continue cat /dev/null > conftest.in lt_ac_count=0 echo $ECHO_N "0123456789$ECHO_C" >conftest.in # Check for GNU sed and select it if it is found. if "$lt_ac_sed" --version 2>&1 < /dev/null | grep 'GNU' > /dev/null; then lt_cv_path_SED=$lt_ac_sed break fi while true; do cat conftest.in conftest.in >conftest.tmp mv conftest.tmp conftest.in cp conftest.in conftest.nl echo >>conftest.nl $lt_ac_sed -e 's/a$//' < conftest.nl >conftest.out || break cmp -s conftest.out conftest.nl || break # 10000 chars as input seems more than enough test $lt_ac_count -gt 10 && break lt_ac_count=`expr $lt_ac_count + 1` if test $lt_ac_count -gt $lt_ac_max; then lt_ac_max=$lt_ac_count lt_cv_path_SED=$lt_ac_sed fi done done ]) SED=$lt_cv_path_SED AC_SUBST([SED]) AC_MSG_RESULT([$SED]) ])#AC_PROG_SED ])#m4_ifndef # Old name: AU_ALIAS([LT_AC_PROG_SED], [AC_PROG_SED]) dnl aclocal-1.4 backwards compatibility: dnl AC_DEFUN([LT_AC_PROG_SED], []) # _LT_CHECK_SHELL_FEATURES # ------------------------ # Find out whether the shell is Bourne or XSI compatible, # or has some other useful features. m4_defun([_LT_CHECK_SHELL_FEATURES], [AC_MSG_CHECKING([whether the shell understands some XSI constructs]) # Try some XSI features xsi_shell=no ( _lt_dummy="a/b/c" test "${_lt_dummy##*/},${_lt_dummy%/*},${_lt_dummy#??}"${_lt_dummy%"$_lt_dummy"}, \ = c,a/b,b/c, \ && eval 'test $(( 1 + 1 )) -eq 2 \ && test "${#_lt_dummy}" -eq 5' ) >/dev/null 2>&1 \ && xsi_shell=yes AC_MSG_RESULT([$xsi_shell]) _LT_CONFIG_LIBTOOL_INIT([xsi_shell='$xsi_shell']) AC_MSG_CHECKING([whether the shell understands "+="]) lt_shell_append=no ( foo=bar; set foo baz; eval "$[1]+=\$[2]" && test "$foo" = barbaz ) \ >/dev/null 2>&1 \ && lt_shell_append=yes AC_MSG_RESULT([$lt_shell_append]) _LT_CONFIG_LIBTOOL_INIT([lt_shell_append='$lt_shell_append']) if ( (MAIL=60; unset MAIL) || exit) >/dev/null 2>&1; then lt_unset=unset else lt_unset=false fi _LT_DECL([], [lt_unset], [0], [whether the shell understands "unset"])dnl # test EBCDIC or ASCII case `echo X|tr X '\101'` in A) # ASCII based system # \n is not interpreted correctly by Solaris 8 /usr/ucb/tr lt_SP2NL='tr \040 \012' lt_NL2SP='tr \015\012 \040\040' ;; *) # EBCDIC based system lt_SP2NL='tr \100 \n' lt_NL2SP='tr \r\n \100\100' ;; esac _LT_DECL([SP2NL], [lt_SP2NL], [1], [turn spaces into newlines])dnl _LT_DECL([NL2SP], [lt_NL2SP], [1], [turn newlines into spaces])dnl ])# _LT_CHECK_SHELL_FEATURES # _LT_PROG_FUNCTION_REPLACE (FUNCNAME, REPLACEMENT-BODY) # ------------------------------------------------------ # In `$cfgfile', look for function FUNCNAME delimited by `^FUNCNAME ()$' and # '^} FUNCNAME ', and replace its body with REPLACEMENT-BODY. m4_defun([_LT_PROG_FUNCTION_REPLACE], [dnl { sed -e '/^$1 ()$/,/^} # $1 /c\ $1 ()\ {\ m4_bpatsubsts([$2], [$], [\\], [^\([ ]\)], [\\\1]) } # Extended-shell $1 implementation' "$cfgfile" > $cfgfile.tmp \ && mv -f "$cfgfile.tmp" "$cfgfile" \ || (rm -f "$cfgfile" && cp "$cfgfile.tmp" "$cfgfile" && rm -f "$cfgfile.tmp") test 0 -eq $? || _lt_function_replace_fail=: ]) # _LT_PROG_REPLACE_SHELLFNS # ------------------------- # Replace existing portable implementations of several shell functions with # equivalent extended shell implementations where those features are available.. m4_defun([_LT_PROG_REPLACE_SHELLFNS], [if test x"$xsi_shell" = xyes; then _LT_PROG_FUNCTION_REPLACE([func_dirname], [dnl case ${1} in */*) func_dirname_result="${1%/*}${2}" ;; * ) func_dirname_result="${3}" ;; esac]) _LT_PROG_FUNCTION_REPLACE([func_basename], [dnl func_basename_result="${1##*/}"]) _LT_PROG_FUNCTION_REPLACE([func_dirname_and_basename], [dnl case ${1} in */*) func_dirname_result="${1%/*}${2}" ;; * ) func_dirname_result="${3}" ;; esac func_basename_result="${1##*/}"]) _LT_PROG_FUNCTION_REPLACE([func_stripname], [dnl # pdksh 5.2.14 does not do ${X%$Y} correctly if both X and Y are # positional parameters, so assign one to ordinary parameter first. func_stripname_result=${3} func_stripname_result=${func_stripname_result#"${1}"} func_stripname_result=${func_stripname_result%"${2}"}]) _LT_PROG_FUNCTION_REPLACE([func_split_long_opt], [dnl func_split_long_opt_name=${1%%=*} func_split_long_opt_arg=${1#*=}]) _LT_PROG_FUNCTION_REPLACE([func_split_short_opt], [dnl func_split_short_opt_arg=${1#??} func_split_short_opt_name=${1%"$func_split_short_opt_arg"}]) _LT_PROG_FUNCTION_REPLACE([func_lo2o], [dnl case ${1} in *.lo) func_lo2o_result=${1%.lo}.${objext} ;; *) func_lo2o_result=${1} ;; esac]) _LT_PROG_FUNCTION_REPLACE([func_xform], [ func_xform_result=${1%.*}.lo]) _LT_PROG_FUNCTION_REPLACE([func_arith], [ func_arith_result=$(( $[*] ))]) _LT_PROG_FUNCTION_REPLACE([func_len], [ func_len_result=${#1}]) fi if test x"$lt_shell_append" = xyes; then _LT_PROG_FUNCTION_REPLACE([func_append], [ eval "${1}+=\\${2}"]) _LT_PROG_FUNCTION_REPLACE([func_append_quoted], [dnl func_quote_for_eval "${2}" dnl m4 expansion turns \\\\ into \\, and then the shell eval turns that into \ eval "${1}+=\\\\ \\$func_quote_for_eval_result"]) # Save a `func_append' function call where possible by direct use of '+=' sed -e 's%func_append \([[a-zA-Z_]]\{1,\}\) "%\1+="%g' $cfgfile > $cfgfile.tmp \ && mv -f "$cfgfile.tmp" "$cfgfile" \ || (rm -f "$cfgfile" && cp "$cfgfile.tmp" "$cfgfile" && rm -f "$cfgfile.tmp") test 0 -eq $? || _lt_function_replace_fail=: else # Save a `func_append' function call even when '+=' is not available sed -e 's%func_append \([[a-zA-Z_]]\{1,\}\) "%\1="$\1%g' $cfgfile > $cfgfile.tmp \ && mv -f "$cfgfile.tmp" "$cfgfile" \ || (rm -f "$cfgfile" && cp "$cfgfile.tmp" "$cfgfile" && rm -f "$cfgfile.tmp") test 0 -eq $? || _lt_function_replace_fail=: fi if test x"$_lt_function_replace_fail" = x":"; then AC_MSG_WARN([Unable to substitute extended shell functions in $ofile]) fi ]) # _LT_PATH_CONVERSION_FUNCTIONS # ----------------------------- # Determine which file name conversion functions should be used by # func_to_host_file (and, implicitly, by func_to_host_path). These are needed # for certain cross-compile configurations and native mingw. m4_defun([_LT_PATH_CONVERSION_FUNCTIONS], [AC_REQUIRE([AC_CANONICAL_HOST])dnl AC_REQUIRE([AC_CANONICAL_BUILD])dnl AC_MSG_CHECKING([how to convert $build file names to $host format]) AC_CACHE_VAL(lt_cv_to_host_file_cmd, [case $host in *-*-mingw* ) case $build in *-*-mingw* ) # actually msys lt_cv_to_host_file_cmd=func_convert_file_msys_to_w32 ;; *-*-cygwin* ) lt_cv_to_host_file_cmd=func_convert_file_cygwin_to_w32 ;; * ) # otherwise, assume *nix lt_cv_to_host_file_cmd=func_convert_file_nix_to_w32 ;; esac ;; *-*-cygwin* ) case $build in *-*-mingw* ) # actually msys lt_cv_to_host_file_cmd=func_convert_file_msys_to_cygwin ;; *-*-cygwin* ) lt_cv_to_host_file_cmd=func_convert_file_noop ;; * ) # otherwise, assume *nix lt_cv_to_host_file_cmd=func_convert_file_nix_to_cygwin ;; esac ;; * ) # unhandled hosts (and "normal" native builds) lt_cv_to_host_file_cmd=func_convert_file_noop ;; esac ]) to_host_file_cmd=$lt_cv_to_host_file_cmd AC_MSG_RESULT([$lt_cv_to_host_file_cmd]) _LT_DECL([to_host_file_cmd], [lt_cv_to_host_file_cmd], [0], [convert $build file names to $host format])dnl AC_MSG_CHECKING([how to convert $build file names to toolchain format]) AC_CACHE_VAL(lt_cv_to_tool_file_cmd, [#assume ordinary cross tools, or native build. lt_cv_to_tool_file_cmd=func_convert_file_noop case $host in *-*-mingw* ) case $build in *-*-mingw* ) # actually msys lt_cv_to_tool_file_cmd=func_convert_file_msys_to_w32 ;; esac ;; esac ]) to_tool_file_cmd=$lt_cv_to_tool_file_cmd AC_MSG_RESULT([$lt_cv_to_tool_file_cmd]) _LT_DECL([to_tool_file_cmd], [lt_cv_to_tool_file_cmd], [0], [convert $build files to toolchain format])dnl ])# _LT_PATH_CONVERSION_FUNCTIONS fftw-3.3.3/m4/ltoptions.m40000644000175000017500000003007311765467473012253 00000000000000# Helper functions for option handling. -*- Autoconf -*- # # Copyright (C) 2004, 2005, 2007, 2008, 2009 Free Software Foundation, # Inc. # Written by Gary V. Vaughan, 2004 # # This file is free software; the Free Software Foundation gives # unlimited permission to copy and/or distribute it, with or without # modifications, as long as this notice is preserved. # serial 7 ltoptions.m4 # This is to help aclocal find these macros, as it can't see m4_define. AC_DEFUN([LTOPTIONS_VERSION], [m4_if([1])]) # _LT_MANGLE_OPTION(MACRO-NAME, OPTION-NAME) # ------------------------------------------ m4_define([_LT_MANGLE_OPTION], [[_LT_OPTION_]m4_bpatsubst($1__$2, [[^a-zA-Z0-9_]], [_])]) # _LT_SET_OPTION(MACRO-NAME, OPTION-NAME) # --------------------------------------- # Set option OPTION-NAME for macro MACRO-NAME, and if there is a # matching handler defined, dispatch to it. Other OPTION-NAMEs are # saved as a flag. m4_define([_LT_SET_OPTION], [m4_define(_LT_MANGLE_OPTION([$1], [$2]))dnl m4_ifdef(_LT_MANGLE_DEFUN([$1], [$2]), _LT_MANGLE_DEFUN([$1], [$2]), [m4_warning([Unknown $1 option `$2'])])[]dnl ]) # _LT_IF_OPTION(MACRO-NAME, OPTION-NAME, IF-SET, [IF-NOT-SET]) # ------------------------------------------------------------ # Execute IF-SET if OPTION is set, IF-NOT-SET otherwise. m4_define([_LT_IF_OPTION], [m4_ifdef(_LT_MANGLE_OPTION([$1], [$2]), [$3], [$4])]) # _LT_UNLESS_OPTIONS(MACRO-NAME, OPTION-LIST, IF-NOT-SET) # ------------------------------------------------------- # Execute IF-NOT-SET unless all options in OPTION-LIST for MACRO-NAME # are set. m4_define([_LT_UNLESS_OPTIONS], [m4_foreach([_LT_Option], m4_split(m4_normalize([$2])), [m4_ifdef(_LT_MANGLE_OPTION([$1], _LT_Option), [m4_define([$0_found])])])[]dnl m4_ifdef([$0_found], [m4_undefine([$0_found])], [$3 ])[]dnl ]) # _LT_SET_OPTIONS(MACRO-NAME, OPTION-LIST) # ---------------------------------------- # OPTION-LIST is a space-separated list of Libtool options associated # with MACRO-NAME. If any OPTION has a matching handler declared with # LT_OPTION_DEFINE, dispatch to that macro; otherwise complain about # the unknown option and exit. m4_defun([_LT_SET_OPTIONS], [# Set options m4_foreach([_LT_Option], m4_split(m4_normalize([$2])), [_LT_SET_OPTION([$1], _LT_Option)]) m4_if([$1],[LT_INIT],[ dnl dnl Simply set some default values (i.e off) if boolean options were not dnl specified: _LT_UNLESS_OPTIONS([LT_INIT], [dlopen], [enable_dlopen=no ]) _LT_UNLESS_OPTIONS([LT_INIT], [win32-dll], [enable_win32_dll=no ]) dnl dnl If no reference was made to various pairs of opposing options, then dnl we run the default mode handler for the pair. For example, if neither dnl `shared' nor `disable-shared' was passed, we enable building of shared dnl archives by default: _LT_UNLESS_OPTIONS([LT_INIT], [shared disable-shared], [_LT_ENABLE_SHARED]) _LT_UNLESS_OPTIONS([LT_INIT], [static disable-static], [_LT_ENABLE_STATIC]) _LT_UNLESS_OPTIONS([LT_INIT], [pic-only no-pic], [_LT_WITH_PIC]) _LT_UNLESS_OPTIONS([LT_INIT], [fast-install disable-fast-install], [_LT_ENABLE_FAST_INSTALL]) ]) ])# _LT_SET_OPTIONS ## --------------------------------- ## ## Macros to handle LT_INIT options. ## ## --------------------------------- ## # _LT_MANGLE_DEFUN(MACRO-NAME, OPTION-NAME) # ----------------------------------------- m4_define([_LT_MANGLE_DEFUN], [[_LT_OPTION_DEFUN_]m4_bpatsubst(m4_toupper([$1__$2]), [[^A-Z0-9_]], [_])]) # LT_OPTION_DEFINE(MACRO-NAME, OPTION-NAME, CODE) # ----------------------------------------------- m4_define([LT_OPTION_DEFINE], [m4_define(_LT_MANGLE_DEFUN([$1], [$2]), [$3])[]dnl ])# LT_OPTION_DEFINE # dlopen # ------ LT_OPTION_DEFINE([LT_INIT], [dlopen], [enable_dlopen=yes ]) AU_DEFUN([AC_LIBTOOL_DLOPEN], [_LT_SET_OPTION([LT_INIT], [dlopen]) AC_DIAGNOSE([obsolete], [$0: Remove this warning and the call to _LT_SET_OPTION when you put the `dlopen' option into LT_INIT's first parameter.]) ]) dnl aclocal-1.4 backwards compatibility: dnl AC_DEFUN([AC_LIBTOOL_DLOPEN], []) # win32-dll # --------- # Declare package support for building win32 dll's. LT_OPTION_DEFINE([LT_INIT], [win32-dll], [enable_win32_dll=yes case $host in *-*-cygwin* | *-*-mingw* | *-*-pw32* | *-*-cegcc*) AC_CHECK_TOOL(AS, as, false) AC_CHECK_TOOL(DLLTOOL, dlltool, false) AC_CHECK_TOOL(OBJDUMP, objdump, false) ;; esac test -z "$AS" && AS=as _LT_DECL([], [AS], [1], [Assembler program])dnl test -z "$DLLTOOL" && DLLTOOL=dlltool _LT_DECL([], [DLLTOOL], [1], [DLL creation program])dnl test -z "$OBJDUMP" && OBJDUMP=objdump _LT_DECL([], [OBJDUMP], [1], [Object dumper program])dnl ])# win32-dll AU_DEFUN([AC_LIBTOOL_WIN32_DLL], [AC_REQUIRE([AC_CANONICAL_HOST])dnl _LT_SET_OPTION([LT_INIT], [win32-dll]) AC_DIAGNOSE([obsolete], [$0: Remove this warning and the call to _LT_SET_OPTION when you put the `win32-dll' option into LT_INIT's first parameter.]) ]) dnl aclocal-1.4 backwards compatibility: dnl AC_DEFUN([AC_LIBTOOL_WIN32_DLL], []) # _LT_ENABLE_SHARED([DEFAULT]) # ---------------------------- # implement the --enable-shared flag, and supports the `shared' and # `disable-shared' LT_INIT options. # DEFAULT is either `yes' or `no'. If omitted, it defaults to `yes'. m4_define([_LT_ENABLE_SHARED], [m4_define([_LT_ENABLE_SHARED_DEFAULT], [m4_if($1, no, no, yes)])dnl AC_ARG_ENABLE([shared], [AS_HELP_STRING([--enable-shared@<:@=PKGS@:>@], [build shared libraries @<:@default=]_LT_ENABLE_SHARED_DEFAULT[@:>@])], [p=${PACKAGE-default} case $enableval in yes) enable_shared=yes ;; no) enable_shared=no ;; *) enable_shared=no # Look at the argument we got. We use all the common list separators. lt_save_ifs="$IFS"; IFS="${IFS}$PATH_SEPARATOR," for pkg in $enableval; do IFS="$lt_save_ifs" if test "X$pkg" = "X$p"; then enable_shared=yes fi done IFS="$lt_save_ifs" ;; esac], [enable_shared=]_LT_ENABLE_SHARED_DEFAULT) _LT_DECL([build_libtool_libs], [enable_shared], [0], [Whether or not to build shared libraries]) ])# _LT_ENABLE_SHARED LT_OPTION_DEFINE([LT_INIT], [shared], [_LT_ENABLE_SHARED([yes])]) LT_OPTION_DEFINE([LT_INIT], [disable-shared], [_LT_ENABLE_SHARED([no])]) # Old names: AC_DEFUN([AC_ENABLE_SHARED], [_LT_SET_OPTION([LT_INIT], m4_if([$1], [no], [disable-])[shared]) ]) AC_DEFUN([AC_DISABLE_SHARED], [_LT_SET_OPTION([LT_INIT], [disable-shared]) ]) AU_DEFUN([AM_ENABLE_SHARED], [AC_ENABLE_SHARED($@)]) AU_DEFUN([AM_DISABLE_SHARED], [AC_DISABLE_SHARED($@)]) dnl aclocal-1.4 backwards compatibility: dnl AC_DEFUN([AM_ENABLE_SHARED], []) dnl AC_DEFUN([AM_DISABLE_SHARED], []) # _LT_ENABLE_STATIC([DEFAULT]) # ---------------------------- # implement the --enable-static flag, and support the `static' and # `disable-static' LT_INIT options. # DEFAULT is either `yes' or `no'. If omitted, it defaults to `yes'. m4_define([_LT_ENABLE_STATIC], [m4_define([_LT_ENABLE_STATIC_DEFAULT], [m4_if($1, no, no, yes)])dnl AC_ARG_ENABLE([static], [AS_HELP_STRING([--enable-static@<:@=PKGS@:>@], [build static libraries @<:@default=]_LT_ENABLE_STATIC_DEFAULT[@:>@])], [p=${PACKAGE-default} case $enableval in yes) enable_static=yes ;; no) enable_static=no ;; *) enable_static=no # Look at the argument we got. We use all the common list separators. lt_save_ifs="$IFS"; IFS="${IFS}$PATH_SEPARATOR," for pkg in $enableval; do IFS="$lt_save_ifs" if test "X$pkg" = "X$p"; then enable_static=yes fi done IFS="$lt_save_ifs" ;; esac], [enable_static=]_LT_ENABLE_STATIC_DEFAULT) _LT_DECL([build_old_libs], [enable_static], [0], [Whether or not to build static libraries]) ])# _LT_ENABLE_STATIC LT_OPTION_DEFINE([LT_INIT], [static], [_LT_ENABLE_STATIC([yes])]) LT_OPTION_DEFINE([LT_INIT], [disable-static], [_LT_ENABLE_STATIC([no])]) # Old names: AC_DEFUN([AC_ENABLE_STATIC], [_LT_SET_OPTION([LT_INIT], m4_if([$1], [no], [disable-])[static]) ]) AC_DEFUN([AC_DISABLE_STATIC], [_LT_SET_OPTION([LT_INIT], [disable-static]) ]) AU_DEFUN([AM_ENABLE_STATIC], [AC_ENABLE_STATIC($@)]) AU_DEFUN([AM_DISABLE_STATIC], [AC_DISABLE_STATIC($@)]) dnl aclocal-1.4 backwards compatibility: dnl AC_DEFUN([AM_ENABLE_STATIC], []) dnl AC_DEFUN([AM_DISABLE_STATIC], []) # _LT_ENABLE_FAST_INSTALL([DEFAULT]) # ---------------------------------- # implement the --enable-fast-install flag, and support the `fast-install' # and `disable-fast-install' LT_INIT options. # DEFAULT is either `yes' or `no'. If omitted, it defaults to `yes'. m4_define([_LT_ENABLE_FAST_INSTALL], [m4_define([_LT_ENABLE_FAST_INSTALL_DEFAULT], [m4_if($1, no, no, yes)])dnl AC_ARG_ENABLE([fast-install], [AS_HELP_STRING([--enable-fast-install@<:@=PKGS@:>@], [optimize for fast installation @<:@default=]_LT_ENABLE_FAST_INSTALL_DEFAULT[@:>@])], [p=${PACKAGE-default} case $enableval in yes) enable_fast_install=yes ;; no) enable_fast_install=no ;; *) enable_fast_install=no # Look at the argument we got. We use all the common list separators. lt_save_ifs="$IFS"; IFS="${IFS}$PATH_SEPARATOR," for pkg in $enableval; do IFS="$lt_save_ifs" if test "X$pkg" = "X$p"; then enable_fast_install=yes fi done IFS="$lt_save_ifs" ;; esac], [enable_fast_install=]_LT_ENABLE_FAST_INSTALL_DEFAULT) _LT_DECL([fast_install], [enable_fast_install], [0], [Whether or not to optimize for fast installation])dnl ])# _LT_ENABLE_FAST_INSTALL LT_OPTION_DEFINE([LT_INIT], [fast-install], [_LT_ENABLE_FAST_INSTALL([yes])]) LT_OPTION_DEFINE([LT_INIT], [disable-fast-install], [_LT_ENABLE_FAST_INSTALL([no])]) # Old names: AU_DEFUN([AC_ENABLE_FAST_INSTALL], [_LT_SET_OPTION([LT_INIT], m4_if([$1], [no], [disable-])[fast-install]) AC_DIAGNOSE([obsolete], [$0: Remove this warning and the call to _LT_SET_OPTION when you put the `fast-install' option into LT_INIT's first parameter.]) ]) AU_DEFUN([AC_DISABLE_FAST_INSTALL], [_LT_SET_OPTION([LT_INIT], [disable-fast-install]) AC_DIAGNOSE([obsolete], [$0: Remove this warning and the call to _LT_SET_OPTION when you put the `disable-fast-install' option into LT_INIT's first parameter.]) ]) dnl aclocal-1.4 backwards compatibility: dnl AC_DEFUN([AC_ENABLE_FAST_INSTALL], []) dnl AC_DEFUN([AM_DISABLE_FAST_INSTALL], []) # _LT_WITH_PIC([MODE]) # -------------------- # implement the --with-pic flag, and support the `pic-only' and `no-pic' # LT_INIT options. # MODE is either `yes' or `no'. If omitted, it defaults to `both'. m4_define([_LT_WITH_PIC], [AC_ARG_WITH([pic], [AS_HELP_STRING([--with-pic@<:@=PKGS@:>@], [try to use only PIC/non-PIC objects @<:@default=use both@:>@])], [lt_p=${PACKAGE-default} case $withval in yes|no) pic_mode=$withval ;; *) pic_mode=default # Look at the argument we got. We use all the common list separators. lt_save_ifs="$IFS"; IFS="${IFS}$PATH_SEPARATOR," for lt_pkg in $withval; do IFS="$lt_save_ifs" if test "X$lt_pkg" = "X$lt_p"; then pic_mode=yes fi done IFS="$lt_save_ifs" ;; esac], [pic_mode=default]) test -z "$pic_mode" && pic_mode=m4_default([$1], [default]) _LT_DECL([], [pic_mode], [0], [What type of objects to build])dnl ])# _LT_WITH_PIC LT_OPTION_DEFINE([LT_INIT], [pic-only], [_LT_WITH_PIC([yes])]) LT_OPTION_DEFINE([LT_INIT], [no-pic], [_LT_WITH_PIC([no])]) # Old name: AU_DEFUN([AC_LIBTOOL_PICMODE], [_LT_SET_OPTION([LT_INIT], [pic-only]) AC_DIAGNOSE([obsolete], [$0: Remove this warning and the call to _LT_SET_OPTION when you put the `pic-only' option into LT_INIT's first parameter.]) ]) dnl aclocal-1.4 backwards compatibility: dnl AC_DEFUN([AC_LIBTOOL_PICMODE], []) ## ----------------- ## ## LTDL_INIT Options ## ## ----------------- ## m4_define([_LTDL_MODE], []) LT_OPTION_DEFINE([LTDL_INIT], [nonrecursive], [m4_define([_LTDL_MODE], [nonrecursive])]) LT_OPTION_DEFINE([LTDL_INIT], [recursive], [m4_define([_LTDL_MODE], [recursive])]) LT_OPTION_DEFINE([LTDL_INIT], [subproject], [m4_define([_LTDL_MODE], [subproject])]) m4_define([_LTDL_TYPE], []) LT_OPTION_DEFINE([LTDL_INIT], [installable], [m4_define([_LTDL_TYPE], [installable])]) LT_OPTION_DEFINE([LTDL_INIT], [convenience], [m4_define([_LTDL_TYPE], [convenience])]) fftw-3.3.3/m4/ltsugar.m40000644000175000017500000001042411765467473011677 00000000000000# ltsugar.m4 -- libtool m4 base layer. -*-Autoconf-*- # # Copyright (C) 2004, 2005, 2007, 2008 Free Software Foundation, Inc. # Written by Gary V. Vaughan, 2004 # # This file is free software; the Free Software Foundation gives # unlimited permission to copy and/or distribute it, with or without # modifications, as long as this notice is preserved. # serial 6 ltsugar.m4 # This is to help aclocal find these macros, as it can't see m4_define. AC_DEFUN([LTSUGAR_VERSION], [m4_if([0.1])]) # lt_join(SEP, ARG1, [ARG2...]) # ----------------------------- # Produce ARG1SEPARG2...SEPARGn, omitting [] arguments and their # associated separator. # Needed until we can rely on m4_join from Autoconf 2.62, since all earlier # versions in m4sugar had bugs. m4_define([lt_join], [m4_if([$#], [1], [], [$#], [2], [[$2]], [m4_if([$2], [], [], [[$2]_])$0([$1], m4_shift(m4_shift($@)))])]) m4_define([_lt_join], [m4_if([$#$2], [2], [], [m4_if([$2], [], [], [[$1$2]])$0([$1], m4_shift(m4_shift($@)))])]) # lt_car(LIST) # lt_cdr(LIST) # ------------ # Manipulate m4 lists. # These macros are necessary as long as will still need to support # Autoconf-2.59 which quotes differently. m4_define([lt_car], [[$1]]) m4_define([lt_cdr], [m4_if([$#], 0, [m4_fatal([$0: cannot be called without arguments])], [$#], 1, [], [m4_dquote(m4_shift($@))])]) m4_define([lt_unquote], $1) # lt_append(MACRO-NAME, STRING, [SEPARATOR]) # ------------------------------------------ # Redefine MACRO-NAME to hold its former content plus `SEPARATOR'`STRING'. # Note that neither SEPARATOR nor STRING are expanded; they are appended # to MACRO-NAME as is (leaving the expansion for when MACRO-NAME is invoked). # No SEPARATOR is output if MACRO-NAME was previously undefined (different # than defined and empty). # # This macro is needed until we can rely on Autoconf 2.62, since earlier # versions of m4sugar mistakenly expanded SEPARATOR but not STRING. m4_define([lt_append], [m4_define([$1], m4_ifdef([$1], [m4_defn([$1])[$3]])[$2])]) # lt_combine(SEP, PREFIX-LIST, INFIX, SUFFIX1, [SUFFIX2...]) # ---------------------------------------------------------- # Produce a SEP delimited list of all paired combinations of elements of # PREFIX-LIST with SUFFIX1 through SUFFIXn. Each element of the list # has the form PREFIXmINFIXSUFFIXn. # Needed until we can rely on m4_combine added in Autoconf 2.62. m4_define([lt_combine], [m4_if(m4_eval([$# > 3]), [1], [m4_pushdef([_Lt_sep], [m4_define([_Lt_sep], m4_defn([lt_car]))])]]dnl [[m4_foreach([_Lt_prefix], [$2], [m4_foreach([_Lt_suffix], ]m4_dquote(m4_dquote(m4_shift(m4_shift(m4_shift($@)))))[, [_Lt_sep([$1])[]m4_defn([_Lt_prefix])[$3]m4_defn([_Lt_suffix])])])])]) # lt_if_append_uniq(MACRO-NAME, VARNAME, [SEPARATOR], [UNIQ], [NOT-UNIQ]) # ----------------------------------------------------------------------- # Iff MACRO-NAME does not yet contain VARNAME, then append it (delimited # by SEPARATOR if supplied) and expand UNIQ, else NOT-UNIQ. m4_define([lt_if_append_uniq], [m4_ifdef([$1], [m4_if(m4_index([$3]m4_defn([$1])[$3], [$3$2$3]), [-1], [lt_append([$1], [$2], [$3])$4], [$5])], [lt_append([$1], [$2], [$3])$4])]) # lt_dict_add(DICT, KEY, VALUE) # ----------------------------- m4_define([lt_dict_add], [m4_define([$1($2)], [$3])]) # lt_dict_add_subkey(DICT, KEY, SUBKEY, VALUE) # -------------------------------------------- m4_define([lt_dict_add_subkey], [m4_define([$1($2:$3)], [$4])]) # lt_dict_fetch(DICT, KEY, [SUBKEY]) # ---------------------------------- m4_define([lt_dict_fetch], [m4_ifval([$3], m4_ifdef([$1($2:$3)], [m4_defn([$1($2:$3)])]), m4_ifdef([$1($2)], [m4_defn([$1($2)])]))]) # lt_if_dict_fetch(DICT, KEY, [SUBKEY], VALUE, IF-TRUE, [IF-FALSE]) # ----------------------------------------------------------------- m4_define([lt_if_dict_fetch], [m4_if(lt_dict_fetch([$1], [$2], [$3]), [$4], [$5], [$6])]) # lt_dict_filter(DICT, [SUBKEY], VALUE, [SEPARATOR], KEY, [...]) # -------------------------------------------------------------- m4_define([lt_dict_filter], [m4_if([$5], [], [], [lt_join(m4_quote(m4_default([$4], [[, ]])), lt_unquote(m4_split(m4_normalize(m4_foreach(_Lt_key, lt_car([m4_shiftn(4, $@)]), [lt_if_dict_fetch([$1], _Lt_key, [$2], [$3], [_Lt_key ])])))))])[]dnl ]) fftw-3.3.3/m4/ltversion.m40000644000175000017500000000126211765467473012243 00000000000000# ltversion.m4 -- version numbers -*- Autoconf -*- # # Copyright (C) 2004 Free Software Foundation, Inc. # Written by Scott James Remnant, 2004 # # This file is free software; the Free Software Foundation gives # unlimited permission to copy and/or distribute it, with or without # modifications, as long as this notice is preserved. # @configure_input@ # serial 3337 ltversion.m4 # This file is part of GNU Libtool m4_define([LT_PACKAGE_VERSION], [2.4.2]) m4_define([LT_PACKAGE_REVISION], [1.3337]) AC_DEFUN([LTVERSION_VERSION], [macro_version='2.4.2' macro_revision='1.3337' _LT_DECL(, macro_version, 0, [Which release of libtool.m4 was used?]) _LT_DECL(, macro_revision, 0) ]) fftw-3.3.3/m4/lt~obsolete.m40000644000175000017500000001375611765467473012603 00000000000000# lt~obsolete.m4 -- aclocal satisfying obsolete definitions. -*-Autoconf-*- # # Copyright (C) 2004, 2005, 2007, 2009 Free Software Foundation, Inc. # Written by Scott James Remnant, 2004. # # This file is free software; the Free Software Foundation gives # unlimited permission to copy and/or distribute it, with or without # modifications, as long as this notice is preserved. # serial 5 lt~obsolete.m4 # These exist entirely to fool aclocal when bootstrapping libtool. # # In the past libtool.m4 has provided macros via AC_DEFUN (or AU_DEFUN) # which have later been changed to m4_define as they aren't part of the # exported API, or moved to Autoconf or Automake where they belong. # # The trouble is, aclocal is a bit thick. It'll see the old AC_DEFUN # in /usr/share/aclocal/libtool.m4 and remember it, then when it sees us # using a macro with the same name in our local m4/libtool.m4 it'll # pull the old libtool.m4 in (it doesn't see our shiny new m4_define # and doesn't know about Autoconf macros at all.) # # So we provide this file, which has a silly filename so it's always # included after everything else. This provides aclocal with the # AC_DEFUNs it wants, but when m4 processes it, it doesn't do anything # because those macros already exist, or will be overwritten later. # We use AC_DEFUN over AU_DEFUN for compatibility with aclocal-1.6. # # Anytime we withdraw an AC_DEFUN or AU_DEFUN, remember to add it here. # Yes, that means every name once taken will need to remain here until # we give up compatibility with versions before 1.7, at which point # we need to keep only those names which we still refer to. # This is to help aclocal find these macros, as it can't see m4_define. AC_DEFUN([LTOBSOLETE_VERSION], [m4_if([1])]) m4_ifndef([AC_LIBTOOL_LINKER_OPTION], [AC_DEFUN([AC_LIBTOOL_LINKER_OPTION])]) m4_ifndef([AC_PROG_EGREP], [AC_DEFUN([AC_PROG_EGREP])]) m4_ifndef([_LT_AC_PROG_ECHO_BACKSLASH], [AC_DEFUN([_LT_AC_PROG_ECHO_BACKSLASH])]) m4_ifndef([_LT_AC_SHELL_INIT], [AC_DEFUN([_LT_AC_SHELL_INIT])]) m4_ifndef([_LT_AC_SYS_LIBPATH_AIX], [AC_DEFUN([_LT_AC_SYS_LIBPATH_AIX])]) m4_ifndef([_LT_PROG_LTMAIN], [AC_DEFUN([_LT_PROG_LTMAIN])]) m4_ifndef([_LT_AC_TAGVAR], [AC_DEFUN([_LT_AC_TAGVAR])]) m4_ifndef([AC_LTDL_ENABLE_INSTALL], [AC_DEFUN([AC_LTDL_ENABLE_INSTALL])]) m4_ifndef([AC_LTDL_PREOPEN], [AC_DEFUN([AC_LTDL_PREOPEN])]) m4_ifndef([_LT_AC_SYS_COMPILER], [AC_DEFUN([_LT_AC_SYS_COMPILER])]) m4_ifndef([_LT_AC_LOCK], [AC_DEFUN([_LT_AC_LOCK])]) m4_ifndef([AC_LIBTOOL_SYS_OLD_ARCHIVE], [AC_DEFUN([AC_LIBTOOL_SYS_OLD_ARCHIVE])]) m4_ifndef([_LT_AC_TRY_DLOPEN_SELF], [AC_DEFUN([_LT_AC_TRY_DLOPEN_SELF])]) m4_ifndef([AC_LIBTOOL_PROG_CC_C_O], [AC_DEFUN([AC_LIBTOOL_PROG_CC_C_O])]) m4_ifndef([AC_LIBTOOL_SYS_HARD_LINK_LOCKS], [AC_DEFUN([AC_LIBTOOL_SYS_HARD_LINK_LOCKS])]) m4_ifndef([AC_LIBTOOL_OBJDIR], [AC_DEFUN([AC_LIBTOOL_OBJDIR])]) m4_ifndef([AC_LTDL_OBJDIR], [AC_DEFUN([AC_LTDL_OBJDIR])]) m4_ifndef([AC_LIBTOOL_PROG_LD_HARDCODE_LIBPATH], [AC_DEFUN([AC_LIBTOOL_PROG_LD_HARDCODE_LIBPATH])]) m4_ifndef([AC_LIBTOOL_SYS_LIB_STRIP], [AC_DEFUN([AC_LIBTOOL_SYS_LIB_STRIP])]) m4_ifndef([AC_PATH_MAGIC], [AC_DEFUN([AC_PATH_MAGIC])]) m4_ifndef([AC_PROG_LD_GNU], [AC_DEFUN([AC_PROG_LD_GNU])]) m4_ifndef([AC_PROG_LD_RELOAD_FLAG], [AC_DEFUN([AC_PROG_LD_RELOAD_FLAG])]) m4_ifndef([AC_DEPLIBS_CHECK_METHOD], [AC_DEFUN([AC_DEPLIBS_CHECK_METHOD])]) m4_ifndef([AC_LIBTOOL_PROG_COMPILER_NO_RTTI], [AC_DEFUN([AC_LIBTOOL_PROG_COMPILER_NO_RTTI])]) m4_ifndef([AC_LIBTOOL_SYS_GLOBAL_SYMBOL_PIPE], [AC_DEFUN([AC_LIBTOOL_SYS_GLOBAL_SYMBOL_PIPE])]) m4_ifndef([AC_LIBTOOL_PROG_COMPILER_PIC], [AC_DEFUN([AC_LIBTOOL_PROG_COMPILER_PIC])]) m4_ifndef([AC_LIBTOOL_PROG_LD_SHLIBS], [AC_DEFUN([AC_LIBTOOL_PROG_LD_SHLIBS])]) m4_ifndef([AC_LIBTOOL_POSTDEP_PREDEP], [AC_DEFUN([AC_LIBTOOL_POSTDEP_PREDEP])]) m4_ifndef([LT_AC_PROG_EGREP], [AC_DEFUN([LT_AC_PROG_EGREP])]) m4_ifndef([LT_AC_PROG_SED], [AC_DEFUN([LT_AC_PROG_SED])]) m4_ifndef([_LT_CC_BASENAME], [AC_DEFUN([_LT_CC_BASENAME])]) m4_ifndef([_LT_COMPILER_BOILERPLATE], [AC_DEFUN([_LT_COMPILER_BOILERPLATE])]) m4_ifndef([_LT_LINKER_BOILERPLATE], [AC_DEFUN([_LT_LINKER_BOILERPLATE])]) m4_ifndef([_AC_PROG_LIBTOOL], [AC_DEFUN([_AC_PROG_LIBTOOL])]) m4_ifndef([AC_LIBTOOL_SETUP], [AC_DEFUN([AC_LIBTOOL_SETUP])]) m4_ifndef([_LT_AC_CHECK_DLFCN], [AC_DEFUN([_LT_AC_CHECK_DLFCN])]) m4_ifndef([AC_LIBTOOL_SYS_DYNAMIC_LINKER], [AC_DEFUN([AC_LIBTOOL_SYS_DYNAMIC_LINKER])]) m4_ifndef([_LT_AC_TAGCONFIG], [AC_DEFUN([_LT_AC_TAGCONFIG])]) m4_ifndef([AC_DISABLE_FAST_INSTALL], [AC_DEFUN([AC_DISABLE_FAST_INSTALL])]) m4_ifndef([_LT_AC_LANG_CXX], [AC_DEFUN([_LT_AC_LANG_CXX])]) m4_ifndef([_LT_AC_LANG_F77], [AC_DEFUN([_LT_AC_LANG_F77])]) m4_ifndef([_LT_AC_LANG_GCJ], [AC_DEFUN([_LT_AC_LANG_GCJ])]) m4_ifndef([AC_LIBTOOL_LANG_C_CONFIG], [AC_DEFUN([AC_LIBTOOL_LANG_C_CONFIG])]) m4_ifndef([_LT_AC_LANG_C_CONFIG], [AC_DEFUN([_LT_AC_LANG_C_CONFIG])]) m4_ifndef([AC_LIBTOOL_LANG_CXX_CONFIG], [AC_DEFUN([AC_LIBTOOL_LANG_CXX_CONFIG])]) m4_ifndef([_LT_AC_LANG_CXX_CONFIG], [AC_DEFUN([_LT_AC_LANG_CXX_CONFIG])]) m4_ifndef([AC_LIBTOOL_LANG_F77_CONFIG], [AC_DEFUN([AC_LIBTOOL_LANG_F77_CONFIG])]) m4_ifndef([_LT_AC_LANG_F77_CONFIG], [AC_DEFUN([_LT_AC_LANG_F77_CONFIG])]) m4_ifndef([AC_LIBTOOL_LANG_GCJ_CONFIG], [AC_DEFUN([AC_LIBTOOL_LANG_GCJ_CONFIG])]) m4_ifndef([_LT_AC_LANG_GCJ_CONFIG], [AC_DEFUN([_LT_AC_LANG_GCJ_CONFIG])]) m4_ifndef([AC_LIBTOOL_LANG_RC_CONFIG], [AC_DEFUN([AC_LIBTOOL_LANG_RC_CONFIG])]) m4_ifndef([_LT_AC_LANG_RC_CONFIG], [AC_DEFUN([_LT_AC_LANG_RC_CONFIG])]) m4_ifndef([AC_LIBTOOL_CONFIG], [AC_DEFUN([AC_LIBTOOL_CONFIG])]) m4_ifndef([_LT_AC_FILE_LTDLL_C], [AC_DEFUN([_LT_AC_FILE_LTDLL_C])]) m4_ifndef([_LT_REQUIRED_DARWIN_CHECKS], [AC_DEFUN([_LT_REQUIRED_DARWIN_CHECKS])]) m4_ifndef([_LT_AC_PROG_CXXCPP], [AC_DEFUN([_LT_AC_PROG_CXXCPP])]) m4_ifndef([_LT_PREPARE_SED_QUOTE_VARS], [AC_DEFUN([_LT_PREPARE_SED_QUOTE_VARS])]) m4_ifndef([_LT_PROG_ECHO_BACKSLASH], [AC_DEFUN([_LT_PROG_ECHO_BACKSLASH])]) m4_ifndef([_LT_PROG_F77], [AC_DEFUN([_LT_PROG_F77])]) m4_ifndef([_LT_PROG_FC], [AC_DEFUN([_LT_PROG_FC])]) m4_ifndef([_LT_PROG_CXX], [AC_DEFUN([_LT_PROG_CXX])]) fftw-3.3.3/m4/Makefile.am0000644000175000017500000000057712054410050011763 00000000000000EXTRA_DIST = acx_mpi.m4 acx_pthread.m4 ax_cc_maxopt.m4 \ ax_check_compiler_flags.m4 ax_compiler_vendor.m4 \ ax_gcc_aligns_stack.m4 ax_gcc_version.m4 ax_openmp.m4 # libtool sticks a bunch of extra .m4 files in this directory, # but they don't seem to be needed for the distributed tarball # (they aren't needed for configure && make, and boostrapping # will regenerate them anyway). fftw-3.3.3/m4/Makefile.in0000644000175000017500000002677612054410200012002 00000000000000# Makefile.in generated by automake 1.11.6 from Makefile.am. # @configure_input@ # Copyright (C) 1994, 1995, 1996, 1997, 1998, 1999, 2000, 2001, 2002, # 2003, 2004, 2005, 2006, 2007, 2008, 2009, 2010, 2011 Free Software # Foundation, Inc. # This Makefile.in is free software; the Free Software Foundation # gives unlimited permission to copy and/or distribute it, # with or without modifications, as long as this notice is preserved. # This program is distributed in the hope that it will be useful, # but WITHOUT ANY WARRANTY, to the extent permitted by law; without # even the implied warranty of MERCHANTABILITY or FITNESS FOR A # PARTICULAR PURPOSE. @SET_MAKE@ VPATH = @srcdir@ am__make_dryrun = \ { \ am__dry=no; \ case $$MAKEFLAGS in \ *\\[\ \ ]*) \ echo 'am--echo: ; @echo "AM" OK' | $(MAKE) -f - 2>/dev/null \ | grep '^AM OK$$' >/dev/null || am__dry=yes;; \ *) \ for am__flg in $$MAKEFLAGS; do \ case $$am__flg in \ *=*|--*) ;; \ *n*) am__dry=yes; break;; \ esac; \ done;; \ esac; \ test $$am__dry = yes; \ } pkgdatadir = $(datadir)/@PACKAGE@ pkgincludedir = $(includedir)/@PACKAGE@ pkglibdir = $(libdir)/@PACKAGE@ pkglibexecdir = $(libexecdir)/@PACKAGE@ am__cd = CDPATH="$${ZSH_VERSION+.}$(PATH_SEPARATOR)" && cd install_sh_DATA = $(install_sh) -c -m 644 install_sh_PROGRAM = $(install_sh) -c install_sh_SCRIPT = $(install_sh) -c INSTALL_HEADER = $(INSTALL_DATA) transform = $(program_transform_name) NORMAL_INSTALL = : PRE_INSTALL = : POST_INSTALL = : NORMAL_UNINSTALL = : PRE_UNINSTALL = : POST_UNINSTALL = : build_triplet = @build@ host_triplet = @host@ subdir = m4 DIST_COMMON = $(srcdir)/Makefile.am $(srcdir)/Makefile.in ACLOCAL_M4 = $(top_srcdir)/aclocal.m4 am__aclocal_m4_deps = $(top_srcdir)/m4/acx_mpi.m4 \ $(top_srcdir)/m4/acx_pthread.m4 \ $(top_srcdir)/m4/ax_cc_maxopt.m4 \ $(top_srcdir)/m4/ax_check_compiler_flags.m4 \ $(top_srcdir)/m4/ax_compiler_vendor.m4 \ $(top_srcdir)/m4/ax_gcc_aligns_stack.m4 \ $(top_srcdir)/m4/ax_gcc_version.m4 \ $(top_srcdir)/m4/ax_openmp.m4 $(top_srcdir)/m4/libtool.m4 \ $(top_srcdir)/m4/ltoptions.m4 $(top_srcdir)/m4/ltsugar.m4 \ $(top_srcdir)/m4/ltversion.m4 $(top_srcdir)/m4/lt~obsolete.m4 \ $(top_srcdir)/configure.ac am__configure_deps = $(am__aclocal_m4_deps) $(CONFIGURE_DEPENDENCIES) \ $(ACLOCAL_M4) mkinstalldirs = $(install_sh) -d CONFIG_HEADER = $(top_builddir)/config.h CONFIG_CLEAN_FILES = CONFIG_CLEAN_VPATH_FILES = SOURCES = DIST_SOURCES = am__can_run_installinfo = \ case $$AM_UPDATE_INFO_DIR in \ n|no|NO) false;; \ *) (install-info --version) >/dev/null 2>&1;; \ esac DISTFILES = $(DIST_COMMON) $(DIST_SOURCES) $(TEXINFOS) $(EXTRA_DIST) ACLOCAL = @ACLOCAL@ ALLOCA = @ALLOCA@ ALTIVEC_CFLAGS = @ALTIVEC_CFLAGS@ AMTAR = @AMTAR@ AR = @AR@ AS = @AS@ AUTOCONF = @AUTOCONF@ AUTOHEADER = @AUTOHEADER@ AUTOMAKE = @AUTOMAKE@ AVX_CFLAGS = @AVX_CFLAGS@ AWK = @AWK@ CC = @CC@ CCDEPMODE = @CCDEPMODE@ CFLAGS = @CFLAGS@ CHECK_PL_OPTS = @CHECK_PL_OPTS@ CPP = @CPP@ CPPFLAGS = @CPPFLAGS@ CYGPATH_W = @CYGPATH_W@ C_FFTW_R2R_KIND = @C_FFTW_R2R_KIND@ C_MPI_FINT = @C_MPI_FINT@ DEFS = @DEFS@ DEPDIR = @DEPDIR@ DLLTOOL = @DLLTOOL@ DSYMUTIL = @DSYMUTIL@ DUMPBIN = @DUMPBIN@ ECHO_C = @ECHO_C@ ECHO_N = @ECHO_N@ ECHO_T = @ECHO_T@ EGREP = @EGREP@ EXEEXT = @EXEEXT@ F77 = @F77@ FFLAGS = @FFLAGS@ FGREP = @FGREP@ FLIBS = @FLIBS@ GREP = @GREP@ INSTALL = @INSTALL@ INSTALL_DATA = @INSTALL_DATA@ INSTALL_PROGRAM = @INSTALL_PROGRAM@ INSTALL_SCRIPT = @INSTALL_SCRIPT@ INSTALL_STRIP_PROGRAM = @INSTALL_STRIP_PROGRAM@ LD = @LD@ LDFLAGS = @LDFLAGS@ LIBOBJS = @LIBOBJS@ LIBQUADMATH = @LIBQUADMATH@ LIBS = @LIBS@ LIBTOOL = @LIBTOOL@ LIPO = @LIPO@ LN_S = @LN_S@ LTLIBOBJS = @LTLIBOBJS@ MAINT = @MAINT@ MAKEINFO = @MAKEINFO@ MANIFEST_TOOL = @MANIFEST_TOOL@ MKDIR_P = @MKDIR_P@ MPICC = @MPICC@ MPILIBS = @MPILIBS@ MPIRUN = @MPIRUN@ NEON_CFLAGS = @NEON_CFLAGS@ NM = @NM@ NMEDIT = @NMEDIT@ OBJDUMP = @OBJDUMP@ OBJEXT = @OBJEXT@ OCAMLBUILD = @OCAMLBUILD@ OPENMP_CFLAGS = @OPENMP_CFLAGS@ OTOOL = @OTOOL@ OTOOL64 = @OTOOL64@ PACKAGE = @PACKAGE@ PACKAGE_BUGREPORT = @PACKAGE_BUGREPORT@ PACKAGE_NAME = @PACKAGE_NAME@ PACKAGE_STRING = @PACKAGE_STRING@ PACKAGE_TARNAME = @PACKAGE_TARNAME@ PACKAGE_URL = @PACKAGE_URL@ PACKAGE_VERSION = @PACKAGE_VERSION@ PATH_SEPARATOR = @PATH_SEPARATOR@ POW_LIB = @POW_LIB@ PRECISION = @PRECISION@ PREC_SUFFIX = @PREC_SUFFIX@ PTHREAD_CC = @PTHREAD_CC@ PTHREAD_CFLAGS = @PTHREAD_CFLAGS@ PTHREAD_LIBS = @PTHREAD_LIBS@ RANLIB = @RANLIB@ SED = @SED@ SET_MAKE = @SET_MAKE@ SHARED_VERSION_INFO = @SHARED_VERSION_INFO@ SHELL = @SHELL@ SSE2_CFLAGS = @SSE2_CFLAGS@ STACK_ALIGN_CFLAGS = @STACK_ALIGN_CFLAGS@ STRIP = @STRIP@ THREADLIBS = @THREADLIBS@ VERSION = @VERSION@ abs_builddir = @abs_builddir@ abs_srcdir = @abs_srcdir@ abs_top_builddir = @abs_top_builddir@ abs_top_srcdir = @abs_top_srcdir@ ac_ct_AR = @ac_ct_AR@ ac_ct_CC = @ac_ct_CC@ ac_ct_DUMPBIN = @ac_ct_DUMPBIN@ ac_ct_F77 = @ac_ct_F77@ acx_pthread_config = @acx_pthread_config@ am__include = @am__include@ am__leading_dot = @am__leading_dot@ am__quote = @am__quote@ am__tar = @am__tar@ am__untar = @am__untar@ bindir = @bindir@ build = @build@ build_alias = @build_alias@ build_cpu = @build_cpu@ build_os = @build_os@ build_vendor = @build_vendor@ builddir = @builddir@ datadir = @datadir@ datarootdir = @datarootdir@ docdir = @docdir@ dvidir = @dvidir@ exec_prefix = @exec_prefix@ host = @host@ host_alias = @host_alias@ host_cpu = @host_cpu@ host_os = @host_os@ host_vendor = @host_vendor@ htmldir = @htmldir@ includedir = @includedir@ infodir = @infodir@ install_sh = @install_sh@ libdir = @libdir@ libexecdir = @libexecdir@ localedir = @localedir@ localstatedir = @localstatedir@ mandir = @mandir@ mkdir_p = @mkdir_p@ oldincludedir = @oldincludedir@ pdfdir = @pdfdir@ prefix = @prefix@ program_transform_name = @program_transform_name@ psdir = @psdir@ sbindir = @sbindir@ sharedstatedir = @sharedstatedir@ srcdir = @srcdir@ sysconfdir = @sysconfdir@ target_alias = @target_alias@ top_build_prefix = @top_build_prefix@ top_builddir = @top_builddir@ top_srcdir = @top_srcdir@ EXTRA_DIST = acx_mpi.m4 acx_pthread.m4 ax_cc_maxopt.m4 \ ax_check_compiler_flags.m4 ax_compiler_vendor.m4 \ ax_gcc_aligns_stack.m4 ax_gcc_version.m4 ax_openmp.m4 all: all-am .SUFFIXES: $(srcdir)/Makefile.in: @MAINTAINER_MODE_TRUE@ $(srcdir)/Makefile.am $(am__configure_deps) @for dep in $?; do \ case '$(am__configure_deps)' in \ *$$dep*) \ ( cd $(top_builddir) && $(MAKE) $(AM_MAKEFLAGS) am--refresh ) \ && { if test -f $@; then exit 0; else break; fi; }; \ exit 1;; \ esac; \ done; \ echo ' cd $(top_srcdir) && $(AUTOMAKE) --gnu m4/Makefile'; \ $(am__cd) $(top_srcdir) && \ $(AUTOMAKE) --gnu m4/Makefile .PRECIOUS: Makefile Makefile: $(srcdir)/Makefile.in $(top_builddir)/config.status @case '$?' in \ *config.status*) \ cd $(top_builddir) && $(MAKE) $(AM_MAKEFLAGS) am--refresh;; \ *) \ echo ' cd $(top_builddir) && $(SHELL) ./config.status $(subdir)/$@ $(am__depfiles_maybe)'; \ cd $(top_builddir) && $(SHELL) ./config.status $(subdir)/$@ $(am__depfiles_maybe);; \ esac; $(top_builddir)/config.status: $(top_srcdir)/configure $(CONFIG_STATUS_DEPENDENCIES) cd $(top_builddir) && $(MAKE) $(AM_MAKEFLAGS) am--refresh $(top_srcdir)/configure: @MAINTAINER_MODE_TRUE@ $(am__configure_deps) cd $(top_builddir) && $(MAKE) $(AM_MAKEFLAGS) am--refresh $(ACLOCAL_M4): @MAINTAINER_MODE_TRUE@ $(am__aclocal_m4_deps) cd $(top_builddir) && $(MAKE) $(AM_MAKEFLAGS) am--refresh $(am__aclocal_m4_deps): mostlyclean-libtool: -rm -f *.lo clean-libtool: -rm -rf .libs _libs tags: TAGS TAGS: ctags: CTAGS CTAGS: distdir: $(DISTFILES) @srcdirstrip=`echo "$(srcdir)" | sed 's/[].[^$$\\*]/\\\\&/g'`; \ topsrcdirstrip=`echo "$(top_srcdir)" | sed 's/[].[^$$\\*]/\\\\&/g'`; \ list='$(DISTFILES)'; \ dist_files=`for file in $$list; do echo $$file; done | \ sed -e "s|^$$srcdirstrip/||;t" \ -e "s|^$$topsrcdirstrip/|$(top_builddir)/|;t"`; \ case $$dist_files in \ */*) $(MKDIR_P) `echo "$$dist_files" | \ sed '/\//!d;s|^|$(distdir)/|;s,/[^/]*$$,,' | \ sort -u` ;; \ esac; \ for file in $$dist_files; do \ if test -f $$file || test -d $$file; then d=.; else d=$(srcdir); fi; \ if test -d $$d/$$file; then \ dir=`echo "/$$file" | sed -e 's,/[^/]*$$,,'`; \ if test -d "$(distdir)/$$file"; then \ find "$(distdir)/$$file" -type d ! -perm -700 -exec chmod u+rwx {} \;; \ fi; \ if test -d $(srcdir)/$$file && test $$d != $(srcdir); then \ cp -fpR $(srcdir)/$$file "$(distdir)$$dir" || exit 1; \ find "$(distdir)/$$file" -type d ! -perm -700 -exec chmod u+rwx {} \;; \ fi; \ cp -fpR $$d/$$file "$(distdir)$$dir" || exit 1; \ else \ test -f "$(distdir)/$$file" \ || cp -p $$d/$$file "$(distdir)/$$file" \ || exit 1; \ fi; \ done check-am: all-am check: check-am all-am: Makefile installdirs: install: install-am install-exec: install-exec-am install-data: install-data-am uninstall: uninstall-am install-am: all-am @$(MAKE) $(AM_MAKEFLAGS) install-exec-am install-data-am installcheck: installcheck-am install-strip: if test -z '$(STRIP)'; then \ $(MAKE) $(AM_MAKEFLAGS) INSTALL_PROGRAM="$(INSTALL_STRIP_PROGRAM)" \ install_sh_PROGRAM="$(INSTALL_STRIP_PROGRAM)" INSTALL_STRIP_FLAG=-s \ install; \ else \ $(MAKE) $(AM_MAKEFLAGS) INSTALL_PROGRAM="$(INSTALL_STRIP_PROGRAM)" \ install_sh_PROGRAM="$(INSTALL_STRIP_PROGRAM)" INSTALL_STRIP_FLAG=-s \ "INSTALL_PROGRAM_ENV=STRIPPROG='$(STRIP)'" install; \ fi mostlyclean-generic: clean-generic: distclean-generic: -test -z "$(CONFIG_CLEAN_FILES)" || rm -f $(CONFIG_CLEAN_FILES) -test . = "$(srcdir)" || test -z "$(CONFIG_CLEAN_VPATH_FILES)" || rm -f $(CONFIG_CLEAN_VPATH_FILES) maintainer-clean-generic: @echo "This command is intended for maintainers to use" @echo "it deletes files that may require special tools to rebuild." clean: clean-am clean-am: clean-generic clean-libtool mostlyclean-am distclean: distclean-am -rm -f Makefile distclean-am: clean-am distclean-generic dvi: dvi-am dvi-am: html: html-am html-am: info: info-am info-am: install-data-am: install-dvi: install-dvi-am install-dvi-am: install-exec-am: install-html: install-html-am install-html-am: install-info: install-info-am install-info-am: install-man: install-pdf: install-pdf-am install-pdf-am: install-ps: install-ps-am install-ps-am: installcheck-am: maintainer-clean: maintainer-clean-am -rm -f Makefile maintainer-clean-am: distclean-am maintainer-clean-generic mostlyclean: mostlyclean-am mostlyclean-am: mostlyclean-generic mostlyclean-libtool pdf: pdf-am pdf-am: ps: ps-am ps-am: uninstall-am: .MAKE: install-am install-strip .PHONY: all all-am check check-am clean clean-generic clean-libtool \ distclean distclean-generic distclean-libtool distdir dvi \ dvi-am html html-am info info-am install install-am \ install-data install-data-am install-dvi install-dvi-am \ install-exec install-exec-am install-html install-html-am \ install-info install-info-am install-man install-pdf \ install-pdf-am install-ps install-ps-am install-strip \ installcheck installcheck-am installdirs maintainer-clean \ maintainer-clean-generic mostlyclean mostlyclean-generic \ mostlyclean-libtool pdf pdf-am ps ps-am uninstall uninstall-am # libtool sticks a bunch of extra .m4 files in this directory, # but they don't seem to be needed for the distributed tarball # (they aren't needed for configure && make, and boostrapping # will regenerate them anyway). # Tell versions [3.59,3.63) of GNU make to not export all variables. # Otherwise a system limit (for SysV at least) may be exceeded. .NOEXPORT: fftw-3.3.3/README0000644000175000017500000000247212054410050010263 00000000000000FFTW is a free collection of fast C routines for computing the Discrete Fourier Transform in one or more dimensions. It includes complex, real, symmetric, and parallel transforms, and can handle arbitrary array sizes efficiently. FFTW is typically faster than other publically-available FFT implementations, and is even competitive with vendor-tuned libraries. (See our web page for extensive benchmarks.) To achieve this performance, FFTW uses novel code-generation and runtime self-optimization techniques (along with many other tricks). The doc/ directory contains the manual in texinfo, PDF, info, and HTML formats. Frequently asked questions and answers can be found in the doc/FAQ/ directory in ASCII and HTML. For a quick introduction to calling FFTW, see the "Tutorial" section of the manual. Installation instructions are provided in the manual (don't worry, it is straightforward). CONTACTS -------- FFTW was written by Matteo Frigo and Steven G. Johnson. You can contact them at fftw@fftw.org. The latest version of FFTW, benchmarks, links, and other information can be found at the FFTW home page (http://www.fftw.org). You can also sign up to the fftw-announce mailing list to receive (infrequent) updates and information about new releases; to do so, go to: http://www.fftw.org/mailman/listinfo/fftw-announce fftw-3.3.3/configure.ac0000644000175000017500000005216412054410050011674 00000000000000dnl Process this file with autoconf to produce a configure script. AC_INIT(fftw, 3.3.3, fftw@fftw.org) AC_CONFIG_SRCDIR(kernel/ifftw.h) # fftw-3.1.x was 4:X:1 # fftw-3.2.x was 5:X:2 # fftw-3.3.x was 6:X:3 SHARED_VERSION_INFO="6:2:3" # CURRENT:REVISION:AGE AM_INIT_AUTOMAKE(1.7) AM_CONFIG_HEADER(config.h) AC_CONFIG_MACRO_DIR([m4]) AM_MAINTAINER_MODE AC_SUBST(SHARED_VERSION_INFO) AC_DISABLE_SHARED dnl to hell with shared libraries AC_CANONICAL_HOST dnl configure options case "${host_cpu}" in powerpc*) have_fma=yes;; ia64*) have_fma=yes;; hppa*) have_fma=yes;; mips64*) have_fma=yes;; *) have_fma=no;; esac AC_ARG_ENABLE(fma, [AC_HELP_STRING([--enable-fma],[enable optimizations for machines with fused multiply-add])], have_fma=$enableval) if test "$have_fma"x = "yes"x; then AC_DEFINE(HAVE_FMA,1,[Define if you have a machine with fused multiply-add]) fi AC_ARG_ENABLE(debug, [AC_HELP_STRING([--enable-debug],[compile fftw with extra runtime checks for debugging])], ok=$enableval, ok=no) if test "$ok" = "yes"; then AC_DEFINE(FFTW_DEBUG,1,[Define to enable extra FFTW debugging code.]) debug_malloc=yes else debug_malloc=no fi AC_ARG_ENABLE(debug-malloc, [AC_HELP_STRING([--enable-debug-malloc],[enable malloc debugging version])], ok=$enableval, ok=$debug_malloc) if test "$ok" = "yes"; then AC_DEFINE(FFTW_DEBUG_MALLOC,1,[Define to enable debugging malloc.]) fi AC_ARG_ENABLE(debug-alignment, [AC_HELP_STRING([--enable-debug-alignment],[enable alignment debugging hacks])], ok=$enableval, ok=no) if test "$ok" = "yes"; then AC_DEFINE(FFTW_DEBUG_ALIGNMENT,1,[Define to enable alignment debugging hacks.]) fi AC_ARG_ENABLE(random-estimator, [AC_HELP_STRING([--enable-random-estimator],[enable pseudorandom estimator (debugging hack)])], ok=$enableval, ok=no) if test "$ok" = "yes"; then AC_DEFINE(FFTW_RANDOM_ESTIMATOR,1,[Define to enable pseudorandom estimate planning for debugging.]) CHECK_PL_OPTS="--estimate" fi AC_ARG_ENABLE(alloca, [AC_HELP_STRING([--disable-alloca],[disable use of the alloca() function (may be broken on mingw64)])], ok=$enableval, ok=yes) if test "$ok" = "yes"; then AC_DEFINE(FFTW_ENABLE_ALLOCA,1,[Define to enable the use of alloca().]) fi AC_ARG_ENABLE(single, [AC_HELP_STRING([--enable-single],[compile fftw in single precision])], ok=$enableval, ok=no) AC_ARG_ENABLE(float, [AC_HELP_STRING([--enable-float],[synonym for --enable-single])], ok=$enableval) if test "$ok" = "yes"; then AC_DEFINE(FFTW_SINGLE,1,[Define to compile in single precision.]) AC_DEFINE(BENCHFFT_SINGLE,1,[Define to compile in single precision.]) PRECISION=s else PRECISION=d fi AM_CONDITIONAL(SINGLE, test "$ok" = "yes") AC_ARG_ENABLE(long-double, [AC_HELP_STRING([--enable-long-double],[compile fftw in long-double precision])], ok=$enableval, ok=no) if test "$ok" = "yes"; then if test "$PRECISION" = "s"; then AC_MSG_ERROR([--enable-single/--enable-long-double conflict]) fi AC_DEFINE(FFTW_LDOUBLE,1,[Define to compile in long-double precision.]) AC_DEFINE(BENCHFFT_LDOUBLE,1,[Define to compile in long-double precision.]) PRECISION=l fi AM_CONDITIONAL(LDOUBLE, test "$ok" = "yes") AC_ARG_ENABLE(quad-precision, [AC_HELP_STRING([--enable-quad-precision],[compile fftw in quadruple precision if available])], ok=$enableval, ok=no) if test "$ok" = "yes"; then if test "$PRECISION" != "d"; then AC_MSG_ERROR([conflicting precisions specified]) fi AC_DEFINE(FFTW_QUAD,1,[Define to compile in quad precision.]) AC_DEFINE(BENCHFFT_QUAD,1,[Define to compile in quad precision.]) PRECISION=q fi AM_CONDITIONAL(QUAD, test "$ok" = "yes") AC_SUBST(PRECISION) AC_SUBST(CHECK_PL_OPTS) AC_ARG_ENABLE(sse, [AC_HELP_STRING([--enable-sse],[enable SSE optimizations])], have_sse=$enableval, have_sse=no) if test "$have_sse" = "yes"; then if test "$PRECISION" != "s"; then AC_MSG_ERROR([SSE requires single precision]) fi fi AC_ARG_ENABLE(sse2, [AC_HELP_STRING([--enable-sse2],[enable SSE/SSE2 optimizations])], have_sse2=$enableval, have_sse2=no) if test "$have_sse" = "yes"; then have_sse2=yes; fi if test "$have_sse2" = "yes"; then AC_DEFINE(HAVE_SSE2,1,[Define to enable SSE/SSE2 optimizations.]) if test "$PRECISION" != "d" -a "$PRECISION" != "s"; then AC_MSG_ERROR([SSE2 requires single or double precision]) fi fi AM_CONDITIONAL(HAVE_SSE2, test "$have_sse2" = "yes") AC_ARG_ENABLE(avx, [AC_HELP_STRING([--enable-avx],[enable AVX optimizations])], have_avx=$enableval, have_avx=no) if test "$have_avx" = "yes"; then AC_DEFINE(HAVE_AVX,1,[Define to enable AVX optimizations.]) if test "$PRECISION" != "d" -a "$PRECISION" != "s"; then AC_MSG_ERROR([AVX requires single or double precision]) fi fi AM_CONDITIONAL(HAVE_AVX, test "$have_avx" = "yes") AC_ARG_ENABLE(altivec, [AC_HELP_STRING([--enable-altivec],[enable Altivec optimizations])], have_altivec=$enableval, have_altivec=no) if test "$have_altivec" = "yes"; then AC_DEFINE(HAVE_ALTIVEC,1,[Define to enable Altivec optimizations.]) if test "$PRECISION" != "s"; then AC_MSG_ERROR([Altivec requires single precision]) fi fi AM_CONDITIONAL(HAVE_ALTIVEC, test "$have_altivec" = "yes") AC_ARG_ENABLE(neon, [AC_HELP_STRING([--enable-neon],[enable ARM NEON optimizations])], have_neon=$enableval, have_neon=no) if test "$have_neon" = "yes"; then AC_DEFINE(HAVE_NEON,1,[Define to enable ARM NEON optimizations.]) if test "$PRECISION" != "s"; then AC_MSG_ERROR([NEON requires single precision]) fi fi AM_CONDITIONAL(HAVE_NEON, test "$have_neon" = "yes") dnl FIXME: dnl AC_ARG_ENABLE(mips-ps, [AC_HELP_STRING([--enable-mips-ps],[enable MIPS pair-single optimizations])], have_mips_ps=$enableval, have_mips_ps=no) dnl if test "$have_mips_ps" = "yes"; then dnl AC_DEFINE(HAVE_MIPS_PS,1,[Define to enable MIPS paired-single optimizations.]) dnl if test "$PRECISION" != "s"; then dnl AC_MSG_ERROR([MIPS paired-single requires single precision]) dnl fi dnl fi dnl AM_CONDITIONAL(HAVE_MIPS_PS, test "$have_mips_ps" = "yes") AC_ARG_WITH(slow-timer, [AC_HELP_STRING([--with-slow-timer],[use low-precision timers (SLOW)])], with_slow_timer=$withval, with_slow_timer=no) if test "$with_slow_timer" = "yes"; then AC_DEFINE(WITH_SLOW_TIMER,1,[Use low-precision timers, making planner very slow]) fi AC_ARG_ENABLE(mips_zbus_timer, [AC_HELP_STRING([--enable-mips-zbus-timer],[use MIPS ZBus cycle-counter])], have_mips_zbus_timer=$enableval, have_mips_zbus_timer=no) if test "$have_mips_zbus_timer" = "yes"; then AC_DEFINE(HAVE_MIPS_ZBUS_TIMER,1,[Define to enable use of MIPS ZBus cycle-counter.]) fi AC_ARG_WITH(our-malloc, [AC_HELP_STRING([--with-our-malloc],[use our aligned malloc (helpful for Win32)])], with_our_malloc=$withval, with_our_malloc=no) AC_ARG_WITH(our-malloc16, [AC_HELP_STRING([--with-our-malloc16],[Obsolete alias for --with-our-malloc16])], with_our_malloc=$withval) if test "$with_our_malloc" = "yes"; then AC_DEFINE(WITH_OUR_MALLOC,1,[Use our own aligned malloc routine; mainly helpful for Windows systems lacking aligned allocation system-library routines.]) fi AC_ARG_WITH(windows-f77-mangling, [AC_HELP_STRING([--with-windows-f77-mangling],[use common Win32 Fortran interface styles])], with_windows_f77_mangling=$withval, with_windows_f77_mangling=no) if test "$with_windows_f77_mangling" = "yes"; then AC_DEFINE(WINDOWS_F77_MANGLING,1,[Use common Windows Fortran mangling styles for the Fortran interfaces.]) fi AC_ARG_WITH(incoming-stack-boundary, [AC_HELP_STRING([--with-incoming-stack-boundary=X],[Assume that stack is aligned to (1<]) CC=$save_CC if test 0 = $ac_cv_sizeof_MPI_Fint; then AC_MSG_WARN([sizeof(MPI_Fint) test failed]); dnl As a backup, assume Fortran integer == C int AC_CHECK_SIZEOF(int) if test 0 = $ac_cv_sizeof_int; then AC_MSG_ERROR([sizeof(int) test failed]); fi ac_cv_sizeof_MPI_Fint=$ac_cv_sizeof_int fi C_MPI_FINT=C_INT`expr $ac_cv_sizeof_MPI_Fint \* 8`_T AC_SUBST(C_MPI_FINT) fi AM_CONDITIONAL(MPI, test "$enable_mpi" = "yes") dnl ----------------------------------------------------------------------- dnl determine CFLAGS first AX_CC_MAXOPT case "${ax_cv_c_compiler_vendor}" in intel) # Stop icc from defining __GNUC__, except on MacOS where this fails case "${host_os}" in *darwin*) ;; # icc -no-gcc fails to compile some system headers *) AX_CHECK_COMPILER_FLAGS([-no-gcc], [CC="$CC -no-gcc"]) ;; esac ;; hp) # must (sometimes) manually increase cpp limits to handle fftw3.h AX_CHECK_COMPILER_FLAGS([-Wp,-H128000], [CC="$CC -Wp,-H128000"]) ;; portland) # -Masmkeyword required for asm("") cycle counters AX_CHECK_COMPILER_FLAGS([-Masmkeyword], [CC="$CC -Masmkeyword"]) ;; esac dnl Determine SIMD CFLAGS at least for gcc and icc case "${ax_cv_c_compiler_vendor}" in gnu|intel) # SSE/SSE2 if test "$have_sse2" = "yes" -a "x$SSE2_CFLAGS" = x; then if test "$PRECISION" = d; then flag=msse2; else flag=msse; fi AX_CHECK_COMPILER_FLAGS(-$flag, [SSE2_CFLAGS="-$flag"], [AC_MSG_ERROR([Need a version of gcc with -$flag])]) fi # AVX if test "$have_avx" = "yes" -a "x$AVX_CFLAGS" = x; then AX_CHECK_COMPILER_FLAGS(-mavx, [AVX_CFLAGS="-mavx"], [AC_MSG_ERROR([Need a version of gcc with -mavx])]) fi if test "$have_altivec" = "yes" -a "x$ALTIVEC_CFLAGS" = x; then # -DFAKE__VEC__ is a workaround because gcc-3.3 does not # #define __VEC__ with -maltivec. AX_CHECK_COMPILER_FLAGS(-faltivec, [ALTIVEC_CFLAGS="-faltivec"], [AX_CHECK_COMPILER_FLAGS(-maltivec -mabi=altivec, [ALTIVEC_CFLAGS="-maltivec -mabi=altivec -DFAKE__VEC__"], [AX_CHECK_COMPILER_FLAGS(-fvec, [ALTIVEC_CFLAGS="-fvec"], [AC_MSG_ERROR([Need a version of gcc with -maltivec])])])]) fi if test "$have_neon" = "yes" -a "x$NEON_CFLAGS" = x; then AX_CHECK_COMPILER_FLAGS(-mfpu=neon, [NEON_CFLAGS="-mfpu=neon"], [AC_MSG_ERROR([Need a version of gcc with -mfpu=neon])]) fi dnl FIXME: dnl elif test "$have_mips_ps" = "yes"; then dnl # Just punt here and use only new 4.2 compiler :( dnl # Should add section for older compilers... dnl AX_CHECK_COMPILER_FLAGS(-mpaired-single, dnl [SIMD_CFLAGS="-mpaired-single"], dnl #[AC_MSG_ERROR([Need a version of gcc with -mpaired-single])]) dnl [AX_CHECK_COMPILER_FLAGS(-march=mips64, dnl [SIMD_CFLAGS="-march=mips64"], dnl [AC_MSG_ERROR( dnl [Need a version of gcc with -mpaired-single or -march=mips64]) dnl ])]) dnl fi ;; esac AC_SUBST(SSE2_CFLAGS) AC_SUBST(AVX_CFLAGS) AC_SUBST(ALTIVEC_CFLAGS) AC_SUBST(NEON_CFLAGS) dnl add stack alignment CFLAGS if so requested if test "$with_incoming_stack_boundary"x != "no"x; then case "${ax_cv_c_compiler_vendor}" in gnu) tentative_flags="-mincoming-stack-boundary=$with_incoming_stack_boundary"; AX_CHECK_COMPILER_FLAGS($tentative_flags, [STACK_ALIGN_CFLAGS=$tentative_flags]) ;; esac fi AC_SUBST(STACK_ALIGN_CFLAGS) dnl Checks for header files. AC_HEADER_STDC AC_CHECK_HEADERS([libintl.h malloc.h stddef.h stdlib.h string.h strings.h sys/time.h unistd.h limits.h c_asm.h intrinsics.h stdint.h mach/mach_time.h sys/sysctl.h]) dnl c_asm.h: Header file for enabling asm() on Digital Unix dnl intrinsics.h: cray unicos dnl sys/sysctl.h: MacOS X altivec detection dnl altivec.h requires $ALTIVEC_CFLAGS save_CFLAGS="$CFLAGS" save_CPPFLAGS="$CPPFLAGS" CFLAGS="$CFLAGS $ALTIVEC_CFLAGS" CPPFLAGS="$CPPFLAGS $ALTIVEC_CFLAGS" AC_CHECK_HEADERS([altivec.h]) CFLAGS="$save_CFLAGS" CPPFLAGS="$save_CPPFLAGS" dnl Checks for typedefs, structures, and compiler characteristics. AC_C_CONST AC_C_INLINE AC_TYPE_SIZE_T AC_HEADER_TIME AC_CHECK_TYPE([long double], [AC_DEFINE(HAVE_LONG_DOUBLE, 1, [Define to 1 if the compiler supports `long double'])], [ if test $PRECISION = l; then AC_MSG_ERROR([long double is not a supported type with your compiler.]) fi ]) AC_CHECK_TYPE([hrtime_t],[AC_DEFINE(HAVE_HRTIME_T, 1, [Define to 1 if hrtime_t is defined in ])],, [ #if HAVE_SYS_TIME_H #include #endif ]) AC_CHECK_SIZEOF(int) AC_CHECK_SIZEOF(unsigned int) AC_CHECK_SIZEOF(long) AC_CHECK_SIZEOF(unsigned long) AC_CHECK_SIZEOF(long long) AC_CHECK_SIZEOF(unsigned long long) AC_CHECK_SIZEOF(size_t) AC_CHECK_SIZEOF(ptrdiff_t) AC_CHECK_TYPES(uintptr_t, [], [AC_CHECK_SIZEOF(void *)], [$ac_includes_default #ifdef HAVE_STDINT_H # include #endif]) AC_CHECK_SIZEOF(float) AC_CHECK_SIZEOF(double) dnl Check sizeof fftw_r2r_kind for Fortran interface [it has == sizeof(int) dnl for years, but being paranoid]. Note: the definition here must match dnl the one in api/fftw3.h! AC_CHECK_SIZEOF(fftw_r2r_kind, [], [typedef enum { FFTW_R2HC=0, FFTW_HC2R=1, FFTW_DHT=2, FFTW_REDFT00=3, FFTW_REDFT01=4, FFTW_REDFT10=5, FFTW_REDFT11=6, FFTW_RODFT00=7, FFTW_RODFT01=8, FFTW_RODFT10=9, FFTW_RODFT11=10 } fftw_r2r_kind;]) if test 0 = $ac_cv_sizeof_fftw_r2r_kind; then AC_MSG_ERROR([sizeof(fftw_r2r_kind) test failed]); fi C_FFTW_R2R_KIND=C_INT`expr $ac_cv_sizeof_fftw_r2r_kind \* 8`_T AC_SUBST(C_FFTW_R2R_KIND) dnl Checks for library functions. AC_FUNC_ALLOCA AC_FUNC_STRTOD AC_FUNC_VPRINTF AC_CHECK_LIB(m, sin) if test $PRECISION = q; then AX_GCC_VERSION(4,6,0,[],[AC_MSG_ERROR([gcc 4.6 or later required for quad precision support])]) AC_CHECK_LIB(quadmath, sinq, [], [AC_MSG_ERROR([quad precision requires libquadmath for quad-precision trigonometric routines])]) LIBQUADMATH=-lquadmath fi AC_SUBST(LIBQUADMATH) AC_CHECK_FUNCS([BSDgettimeofday gettimeofday gethrtime read_real_time time_base_to_time drand48 sqrt memset posix_memalign memalign _mm_malloc _mm_free clock_gettime mach_absolute_time sysctl abort sinl cosl snprintf]) AC_CHECK_DECLS([drand48, srand48, memalign, posix_memalign, sinl, cosl, sinq, cosq]) dnl Cray UNICOS _rtc() (real-time clock) intrinsic AC_MSG_CHECKING([for _rtc intrinsic]) rtc_ok=yes AC_TRY_LINK([#ifdef HAVE_INTRINSICS_H #include #endif], [_rtc()], [AC_DEFINE(HAVE__RTC,1,[Define if you have the UNICOS _rtc() intrinsic.])], [rtc_ok=no]) AC_MSG_RESULT($rtc_ok) if test "$PRECISION" = "l"; then AC_CHECK_FUNCS([cosl sinl tanl], [], [AC_MSG_ERROR([long-double precision requires long-double trigonometric routines])]) fi AC_MSG_CHECKING([for isnan]) AC_TRY_LINK([#include ], if (!isnan(3.14159)) isnan(2.7183);, ok=yes, ok=no) if test "$ok" = "yes"; then AC_DEFINE(HAVE_ISNAN,1,[Define if the isnan() function/macro is available.]) fi AC_MSG_RESULT(${ok}) dnl TODO AX_GCC_ALIGNS_STACK() dnl override CFLAGS selection when debugging if test "${enable_debug}" = "yes"; then CFLAGS="-g" fi dnl add gcc warnings, in debug/maintainer mode only if test "$enable_debug" = yes || test "$USE_MAINTAINER_MODE" = yes; then if test "$ac_test_CFLAGS" != "set"; then if test $ac_cv_prog_gcc = yes; then CFLAGS="$CFLAGS -Wall -W -Wcast-qual -Wpointer-arith -Wcast-align -pedantic -Wno-long-long -Wshadow -Wbad-function-cast -Wwrite-strings -Wstrict-prototypes -Wredundant-decls -Wnested-externs" # -Wundef -Wconversion -Wmissing-prototypes -Wmissing-declarations fi fi fi dnl ----------------------------------------------------------------------- AC_ARG_ENABLE(fortran, [AC_HELP_STRING([--disable-fortran],[don't include Fortran-callable wrappers])], enable_fortran=$enableval, enable_fortran=yes) if test "$enable_fortran" = "yes"; then AC_PROG_F77 if test -z "$F77"; then enable_fortran=no AC_MSG_WARN([*** Couldn't find f77 compiler; using default Fortran wrappers.]) else AC_F77_DUMMY_MAIN([], [enable_fortran=no AC_MSG_WARN([*** Couldn't figure out how to link C and Fortran; using default Fortran wrappers.])]) fi else AC_DEFINE([DISABLE_FORTRAN], 1, [Define to disable Fortran wrappers.]) fi if test "x$enable_fortran" = xyes; then AC_F77_WRAPPERS AC_F77_FUNC(f77foo) AC_F77_FUNC(f77_foo) f77_foo2=`echo $f77foo | sed 's/77/77_/'` if test "$f77_foo" = "$f77_foo2"; then AC_DEFINE(F77_FUNC_EQUIV, 1, [Define if F77_FUNC and F77_FUNC_ are equivalent.]) # Include g77 wrappers by default for GNU systems or gfortran with_g77_wrappers=$ac_cv_f77_compiler_gnu case $host_os in *gnu*) with_g77_wrappers=yes ;; esac fi else with_g77_wrappers=no fi AC_ARG_WITH(g77-wrappers, [AC_HELP_STRING([--with-g77-wrappers],[force inclusion of g77-compatible wrappers in addition to any other Fortran compiler that is detected])], with_g77_wrappers=$withval) if test "x$with_g77_wrappers" = "xyes"; then AC_DEFINE(WITH_G77_WRAPPERS,1,[Include g77-compatible wrappers in addition to any other Fortran wrappers.]) fi dnl ----------------------------------------------------------------------- have_smp="no" AC_ARG_ENABLE(openmp, [AC_HELP_STRING([--enable-openmp],[use OpenMP directives for parallelism])], enable_openmp=$enableval, enable_openmp=no) if test "$enable_openmp" = "yes"; then AC_DEFINE(HAVE_OPENMP,1,[Define to enable OpenMP]) AX_OPENMP([], [AC_MSG_ERROR([don't know how to enable OpenMP])]) fi AC_ARG_ENABLE(threads, [AC_HELP_STRING([--enable-threads],[compile FFTW SMP threads library])], enable_threads=$enableval, enable_threads=no) if test "$enable_threads" = "yes"; then AC_DEFINE(HAVE_THREADS,1,[Define to enable SMP threads]) fi AC_ARG_WITH(combined-threads, [AC_HELP_STRING([--with-combined-threads],[combine threads into main libfftw3])], with_combined_threads=$withval, with_combined_threads=no) if test "$with_combined_threads" = yes; then if test "$enable_openmp" = "yes"; then AC_MSG_ERROR([--with-combined-threads incompatible with --enable-openmp]) fi if test "$enable_threads" != "yes"; then AC_MSG_ERROR([--with-combined-threads requires --enable-threads]) fi fi dnl Check for threads library... THREADLIBS="" if test "$enable_threads" = "yes"; then # POSIX threads, the default choice: if test -z "$THREADLIBS"; then ACX_PTHREAD([THREADLIBS="$PTHREAD_LIBS " CC="$PTHREAD_CC" AC_DEFINE(USING_POSIX_THREADS, 1, [Define if we have and are using POSIX threads.])]) fi if test -z "$THREADLIBS"; then AC_MSG_CHECKING([for Win32 threads]) AC_TRY_LINK([#include ], [_beginthreadex(0,0,0,0,0,0);], [THREADLIBS=" "; AC_MSG_RESULT(yes)], [AC_MSG_RESULT(no)]) fi if test -z "$THREADLIBS"; then AC_MSG_ERROR([couldn't find threads library for --enable-threads]) fi AC_DEFINE(HAVE_THREADS, 1, [Define if we have a threads library.]) fi AC_SUBST(THREADLIBS) AM_CONDITIONAL(THREADS, test "$enable_threads" = "yes") AM_CONDITIONAL(OPENMP, test "$enable_openmp" = "yes") AM_CONDITIONAL(SMP, test "$enable_threads" = "yes" -o "$enable_openmp" = "yes") AM_CONDITIONAL(COMBINED_THREADS, test x"$with_combined_threads" = xyes) dnl ----------------------------------------------------------------------- AC_MSG_CHECKING([whether a cycle counter is available]) save_CPPFLAGS=$CPPFLAGS CPPFLAGS="$CPPFLAGS -I$srcdir/kernel" AC_TRY_CPP([#include "cycle.h" #ifndef HAVE_TICK_COUNTER # error No cycle counter #endif], [ok=yes], [ok=no]) CPPFLAGS=$save_CPPFLAGS AC_MSG_RESULT($ok) if test $ok = no && test "x$with_slow_timer" = xno; then echo "***************************************************************" echo "WARNING: No cycle counter found. FFTW will use ESTIMATE mode " echo " for all plans. See the manual for more information." echo "***************************************************************" fi dnl ----------------------------------------------------------------------- AC_DEFINE_UNQUOTED(FFTW_CC, "$CC $CFLAGS", [C compiler name and flags]) AC_CONFIG_FILES([ Makefile support/Makefile genfft/Makefile kernel/Makefile simd-support/Makefile dft/Makefile dft/scalar/Makefile dft/scalar/codelets/Makefile dft/simd/Makefile dft/simd/common/Makefile dft/simd/sse2/Makefile dft/simd/avx/Makefile dft/simd/altivec/Makefile dft/simd/neon/Makefile rdft/Makefile rdft/scalar/Makefile rdft/scalar/r2cf/Makefile rdft/scalar/r2cb/Makefile rdft/scalar/r2r/Makefile rdft/simd/Makefile rdft/simd/common/Makefile rdft/simd/sse2/Makefile rdft/simd/avx/Makefile rdft/simd/altivec/Makefile rdft/simd/neon/Makefile reodft/Makefile threads/Makefile api/Makefile mpi/Makefile libbench2/Makefile tests/Makefile doc/Makefile doc/FAQ/Makefile tools/Makefile tools/fftw_wisdom.1 tools/fftw-wisdom-to-conf m4/Makefile fftw.pc ]) AC_OUTPUT fftw-3.3.3/aclocal.m40000644000175000017500000011360512054410175011254 00000000000000# generated automatically by aclocal 1.11.6 -*- Autoconf -*- # Copyright (C) 1996, 1997, 1998, 1999, 2000, 2001, 2002, 2003, 2004, # 2005, 2006, 2007, 2008, 2009, 2010, 2011 Free Software Foundation, # Inc. # This file is free software; the Free Software Foundation # gives unlimited permission to copy and/or distribute it, # with or without modifications, as long as this notice is preserved. # This program is distributed in the hope that it will be useful, # but WITHOUT ANY WARRANTY, to the extent permitted by law; without # even the implied warranty of MERCHANTABILITY or FITNESS FOR A # PARTICULAR PURPOSE. m4_ifndef([AC_AUTOCONF_VERSION], [m4_copy([m4_PACKAGE_VERSION], [AC_AUTOCONF_VERSION])])dnl m4_if(m4_defn([AC_AUTOCONF_VERSION]), [2.69],, [m4_warning([this file was generated for autoconf 2.69. You have another version of autoconf. It may work, but is not guaranteed to. If you have problems, you may need to regenerate the build system entirely. To do so, use the procedure documented by the package, typically `autoreconf'.])]) # Copyright (C) 2002, 2003, 2005, 2006, 2007, 2008, 2011 Free Software # Foundation, Inc. # # This file is free software; the Free Software Foundation # gives unlimited permission to copy and/or distribute it, # with or without modifications, as long as this notice is preserved. # serial 1 # AM_AUTOMAKE_VERSION(VERSION) # ---------------------------- # Automake X.Y traces this macro to ensure aclocal.m4 has been # generated from the m4 files accompanying Automake X.Y. # (This private macro should not be called outside this file.) AC_DEFUN([AM_AUTOMAKE_VERSION], [am__api_version='1.11' dnl Some users find AM_AUTOMAKE_VERSION and mistake it for a way to dnl require some minimum version. Point them to the right macro. m4_if([$1], [1.11.6], [], [AC_FATAL([Do not call $0, use AM_INIT_AUTOMAKE([$1]).])])dnl ]) # _AM_AUTOCONF_VERSION(VERSION) # ----------------------------- # aclocal traces this macro to find the Autoconf version. # This is a private macro too. Using m4_define simplifies # the logic in aclocal, which can simply ignore this definition. m4_define([_AM_AUTOCONF_VERSION], []) # AM_SET_CURRENT_AUTOMAKE_VERSION # ------------------------------- # Call AM_AUTOMAKE_VERSION and AM_AUTOMAKE_VERSION so they can be traced. # This function is AC_REQUIREd by AM_INIT_AUTOMAKE. AC_DEFUN([AM_SET_CURRENT_AUTOMAKE_VERSION], [AM_AUTOMAKE_VERSION([1.11.6])dnl m4_ifndef([AC_AUTOCONF_VERSION], [m4_copy([m4_PACKAGE_VERSION], [AC_AUTOCONF_VERSION])])dnl _AM_AUTOCONF_VERSION(m4_defn([AC_AUTOCONF_VERSION]))]) # AM_AUX_DIR_EXPAND -*- Autoconf -*- # Copyright (C) 2001, 2003, 2005, 2011 Free Software Foundation, Inc. # # This file is free software; the Free Software Foundation # gives unlimited permission to copy and/or distribute it, # with or without modifications, as long as this notice is preserved. # serial 1 # For projects using AC_CONFIG_AUX_DIR([foo]), Autoconf sets # $ac_aux_dir to `$srcdir/foo'. In other projects, it is set to # `$srcdir', `$srcdir/..', or `$srcdir/../..'. # # Of course, Automake must honor this variable whenever it calls a # tool from the auxiliary directory. The problem is that $srcdir (and # therefore $ac_aux_dir as well) can be either absolute or relative, # depending on how configure is run. This is pretty annoying, since # it makes $ac_aux_dir quite unusable in subdirectories: in the top # source directory, any form will work fine, but in subdirectories a # relative path needs to be adjusted first. # # $ac_aux_dir/missing # fails when called from a subdirectory if $ac_aux_dir is relative # $top_srcdir/$ac_aux_dir/missing # fails if $ac_aux_dir is absolute, # fails when called from a subdirectory in a VPATH build with # a relative $ac_aux_dir # # The reason of the latter failure is that $top_srcdir and $ac_aux_dir # are both prefixed by $srcdir. In an in-source build this is usually # harmless because $srcdir is `.', but things will broke when you # start a VPATH build or use an absolute $srcdir. # # So we could use something similar to $top_srcdir/$ac_aux_dir/missing, # iff we strip the leading $srcdir from $ac_aux_dir. That would be: # am_aux_dir='\$(top_srcdir)/'`expr "$ac_aux_dir" : "$srcdir//*\(.*\)"` # and then we would define $MISSING as # MISSING="\${SHELL} $am_aux_dir/missing" # This will work as long as MISSING is not called from configure, because # unfortunately $(top_srcdir) has no meaning in configure. # However there are other variables, like CC, which are often used in # configure, and could therefore not use this "fixed" $ac_aux_dir. # # Another solution, used here, is to always expand $ac_aux_dir to an # absolute PATH. The drawback is that using absolute paths prevent a # configured tree to be moved without reconfiguration. AC_DEFUN([AM_AUX_DIR_EXPAND], [dnl Rely on autoconf to set up CDPATH properly. AC_PREREQ([2.50])dnl # expand $ac_aux_dir to an absolute path am_aux_dir=`cd $ac_aux_dir && pwd` ]) # AM_CONDITIONAL -*- Autoconf -*- # Copyright (C) 1997, 2000, 2001, 2003, 2004, 2005, 2006, 2008 # Free Software Foundation, Inc. # # This file is free software; the Free Software Foundation # gives unlimited permission to copy and/or distribute it, # with or without modifications, as long as this notice is preserved. # serial 9 # AM_CONDITIONAL(NAME, SHELL-CONDITION) # ------------------------------------- # Define a conditional. AC_DEFUN([AM_CONDITIONAL], [AC_PREREQ(2.52)dnl ifelse([$1], [TRUE], [AC_FATAL([$0: invalid condition: $1])], [$1], [FALSE], [AC_FATAL([$0: invalid condition: $1])])dnl AC_SUBST([$1_TRUE])dnl AC_SUBST([$1_FALSE])dnl _AM_SUBST_NOTMAKE([$1_TRUE])dnl _AM_SUBST_NOTMAKE([$1_FALSE])dnl m4_define([_AM_COND_VALUE_$1], [$2])dnl if $2; then $1_TRUE= $1_FALSE='#' else $1_TRUE='#' $1_FALSE= fi AC_CONFIG_COMMANDS_PRE( [if test -z "${$1_TRUE}" && test -z "${$1_FALSE}"; then AC_MSG_ERROR([[conditional "$1" was never defined. Usually this means the macro was only invoked conditionally.]]) fi])]) # Copyright (C) 1999, 2000, 2001, 2002, 2003, 2004, 2005, 2006, 2009, # 2010, 2011 Free Software Foundation, Inc. # # This file is free software; the Free Software Foundation # gives unlimited permission to copy and/or distribute it, # with or without modifications, as long as this notice is preserved. # serial 12 # There are a few dirty hacks below to avoid letting `AC_PROG_CC' be # written in clear, in which case automake, when reading aclocal.m4, # will think it sees a *use*, and therefore will trigger all it's # C support machinery. Also note that it means that autoscan, seeing # CC etc. in the Makefile, will ask for an AC_PROG_CC use... # _AM_DEPENDENCIES(NAME) # ---------------------- # See how the compiler implements dependency checking. # NAME is "CC", "CXX", "GCJ", or "OBJC". # We try a few techniques and use that to set a single cache variable. # # We don't AC_REQUIRE the corresponding AC_PROG_CC since the latter was # modified to invoke _AM_DEPENDENCIES(CC); we would have a circular # dependency, and given that the user is not expected to run this macro, # just rely on AC_PROG_CC. AC_DEFUN([_AM_DEPENDENCIES], [AC_REQUIRE([AM_SET_DEPDIR])dnl AC_REQUIRE([AM_OUTPUT_DEPENDENCY_COMMANDS])dnl AC_REQUIRE([AM_MAKE_INCLUDE])dnl AC_REQUIRE([AM_DEP_TRACK])dnl ifelse([$1], CC, [depcc="$CC" am_compiler_list=], [$1], CXX, [depcc="$CXX" am_compiler_list=], [$1], OBJC, [depcc="$OBJC" am_compiler_list='gcc3 gcc'], [$1], UPC, [depcc="$UPC" am_compiler_list=], [$1], GCJ, [depcc="$GCJ" am_compiler_list='gcc3 gcc'], [depcc="$$1" am_compiler_list=]) AC_CACHE_CHECK([dependency style of $depcc], [am_cv_$1_dependencies_compiler_type], [if test -z "$AMDEP_TRUE" && test -f "$am_depcomp"; then # We make a subdir and do the tests there. Otherwise we can end up # making bogus files that we don't know about and never remove. For # instance it was reported that on HP-UX the gcc test will end up # making a dummy file named `D' -- because `-MD' means `put the output # in D'. rm -rf conftest.dir mkdir conftest.dir # Copy depcomp to subdir because otherwise we won't find it if we're # using a relative directory. cp "$am_depcomp" conftest.dir cd conftest.dir # We will build objects and dependencies in a subdirectory because # it helps to detect inapplicable dependency modes. For instance # both Tru64's cc and ICC support -MD to output dependencies as a # side effect of compilation, but ICC will put the dependencies in # the current directory while Tru64 will put them in the object # directory. mkdir sub am_cv_$1_dependencies_compiler_type=none if test "$am_compiler_list" = ""; then am_compiler_list=`sed -n ['s/^#*\([a-zA-Z0-9]*\))$/\1/p'] < ./depcomp` fi am__universal=false m4_case([$1], [CC], [case " $depcc " in #( *\ -arch\ *\ -arch\ *) am__universal=true ;; esac], [CXX], [case " $depcc " in #( *\ -arch\ *\ -arch\ *) am__universal=true ;; esac]) for depmode in $am_compiler_list; do # Setup a source with many dependencies, because some compilers # like to wrap large dependency lists on column 80 (with \), and # we should not choose a depcomp mode which is confused by this. # # We need to recreate these files for each test, as the compiler may # overwrite some of them when testing with obscure command lines. # This happens at least with the AIX C compiler. : > sub/conftest.c for i in 1 2 3 4 5 6; do echo '#include "conftst'$i'.h"' >> sub/conftest.c # Using `: > sub/conftst$i.h' creates only sub/conftst1.h with # Solaris 8's {/usr,}/bin/sh. touch sub/conftst$i.h done echo "${am__include} ${am__quote}sub/conftest.Po${am__quote}" > confmf # We check with `-c' and `-o' for the sake of the "dashmstdout" # mode. It turns out that the SunPro C++ compiler does not properly # handle `-M -o', and we need to detect this. Also, some Intel # versions had trouble with output in subdirs am__obj=sub/conftest.${OBJEXT-o} am__minus_obj="-o $am__obj" case $depmode in gcc) # This depmode causes a compiler race in universal mode. test "$am__universal" = false || continue ;; nosideeffect) # after this tag, mechanisms are not by side-effect, so they'll # only be used when explicitly requested if test "x$enable_dependency_tracking" = xyes; then continue else break fi ;; msvc7 | msvc7msys | msvisualcpp | msvcmsys) # This compiler won't grok `-c -o', but also, the minuso test has # not run yet. These depmodes are late enough in the game, and # so weak that their functioning should not be impacted. am__obj=conftest.${OBJEXT-o} am__minus_obj= ;; none) break ;; esac if depmode=$depmode \ source=sub/conftest.c object=$am__obj \ depfile=sub/conftest.Po tmpdepfile=sub/conftest.TPo \ $SHELL ./depcomp $depcc -c $am__minus_obj sub/conftest.c \ >/dev/null 2>conftest.err && grep sub/conftst1.h sub/conftest.Po > /dev/null 2>&1 && grep sub/conftst6.h sub/conftest.Po > /dev/null 2>&1 && grep $am__obj sub/conftest.Po > /dev/null 2>&1 && ${MAKE-make} -s -f confmf > /dev/null 2>&1; then # icc doesn't choke on unknown options, it will just issue warnings # or remarks (even with -Werror). So we grep stderr for any message # that says an option was ignored or not supported. # When given -MP, icc 7.0 and 7.1 complain thusly: # icc: Command line warning: ignoring option '-M'; no argument required # The diagnosis changed in icc 8.0: # icc: Command line remark: option '-MP' not supported if (grep 'ignoring option' conftest.err || grep 'not supported' conftest.err) >/dev/null 2>&1; then :; else am_cv_$1_dependencies_compiler_type=$depmode break fi fi done cd .. rm -rf conftest.dir else am_cv_$1_dependencies_compiler_type=none fi ]) AC_SUBST([$1DEPMODE], [depmode=$am_cv_$1_dependencies_compiler_type]) AM_CONDITIONAL([am__fastdep$1], [ test "x$enable_dependency_tracking" != xno \ && test "$am_cv_$1_dependencies_compiler_type" = gcc3]) ]) # AM_SET_DEPDIR # ------------- # Choose a directory name for dependency files. # This macro is AC_REQUIREd in _AM_DEPENDENCIES AC_DEFUN([AM_SET_DEPDIR], [AC_REQUIRE([AM_SET_LEADING_DOT])dnl AC_SUBST([DEPDIR], ["${am__leading_dot}deps"])dnl ]) # AM_DEP_TRACK # ------------ AC_DEFUN([AM_DEP_TRACK], [AC_ARG_ENABLE(dependency-tracking, [ --disable-dependency-tracking speeds up one-time build --enable-dependency-tracking do not reject slow dependency extractors]) if test "x$enable_dependency_tracking" != xno; then am_depcomp="$ac_aux_dir/depcomp" AMDEPBACKSLASH='\' am__nodep='_no' fi AM_CONDITIONAL([AMDEP], [test "x$enable_dependency_tracking" != xno]) AC_SUBST([AMDEPBACKSLASH])dnl _AM_SUBST_NOTMAKE([AMDEPBACKSLASH])dnl AC_SUBST([am__nodep])dnl _AM_SUBST_NOTMAKE([am__nodep])dnl ]) # Generate code to set up dependency tracking. -*- Autoconf -*- # Copyright (C) 1999, 2000, 2001, 2002, 2003, 2004, 2005, 2008 # Free Software Foundation, Inc. # # This file is free software; the Free Software Foundation # gives unlimited permission to copy and/or distribute it, # with or without modifications, as long as this notice is preserved. #serial 5 # _AM_OUTPUT_DEPENDENCY_COMMANDS # ------------------------------ AC_DEFUN([_AM_OUTPUT_DEPENDENCY_COMMANDS], [{ # Autoconf 2.62 quotes --file arguments for eval, but not when files # are listed without --file. Let's play safe and only enable the eval # if we detect the quoting. case $CONFIG_FILES in *\'*) eval set x "$CONFIG_FILES" ;; *) set x $CONFIG_FILES ;; esac shift for mf do # Strip MF so we end up with the name of the file. mf=`echo "$mf" | sed -e 's/:.*$//'` # Check whether this is an Automake generated Makefile or not. # We used to match only the files named `Makefile.in', but # some people rename them; so instead we look at the file content. # Grep'ing the first line is not enough: some people post-process # each Makefile.in and add a new line on top of each file to say so. # Grep'ing the whole file is not good either: AIX grep has a line # limit of 2048, but all sed's we know have understand at least 4000. if sed -n 's,^#.*generated by automake.*,X,p' "$mf" | grep X >/dev/null 2>&1; then dirpart=`AS_DIRNAME("$mf")` else continue fi # Extract the definition of DEPDIR, am__include, and am__quote # from the Makefile without running `make'. DEPDIR=`sed -n 's/^DEPDIR = //p' < "$mf"` test -z "$DEPDIR" && continue am__include=`sed -n 's/^am__include = //p' < "$mf"` test -z "am__include" && continue am__quote=`sed -n 's/^am__quote = //p' < "$mf"` # When using ansi2knr, U may be empty or an underscore; expand it U=`sed -n 's/^U = //p' < "$mf"` # Find all dependency output files, they are included files with # $(DEPDIR) in their names. We invoke sed twice because it is the # simplest approach to changing $(DEPDIR) to its actual value in the # expansion. for file in `sed -n " s/^$am__include $am__quote\(.*(DEPDIR).*\)$am__quote"'$/\1/p' <"$mf" | \ sed -e 's/\$(DEPDIR)/'"$DEPDIR"'/g' -e 's/\$U/'"$U"'/g'`; do # Make sure the directory exists. test -f "$dirpart/$file" && continue fdir=`AS_DIRNAME(["$file"])` AS_MKDIR_P([$dirpart/$fdir]) # echo "creating $dirpart/$file" echo '# dummy' > "$dirpart/$file" done done } ])# _AM_OUTPUT_DEPENDENCY_COMMANDS # AM_OUTPUT_DEPENDENCY_COMMANDS # ----------------------------- # This macro should only be invoked once -- use via AC_REQUIRE. # # This code is only required when automatic dependency tracking # is enabled. FIXME. This creates each `.P' file that we will # need in order to bootstrap the dependency handling code. AC_DEFUN([AM_OUTPUT_DEPENDENCY_COMMANDS], [AC_CONFIG_COMMANDS([depfiles], [test x"$AMDEP_TRUE" != x"" || _AM_OUTPUT_DEPENDENCY_COMMANDS], [AMDEP_TRUE="$AMDEP_TRUE" ac_aux_dir="$ac_aux_dir"]) ]) # Copyright (C) 1996, 1997, 2000, 2001, 2003, 2005 # Free Software Foundation, Inc. # # This file is free software; the Free Software Foundation # gives unlimited permission to copy and/or distribute it, # with or without modifications, as long as this notice is preserved. # serial 8 # AM_CONFIG_HEADER is obsolete. It has been replaced by AC_CONFIG_HEADERS. AU_DEFUN([AM_CONFIG_HEADER], [AC_CONFIG_HEADERS($@)]) # Do all the work for Automake. -*- Autoconf -*- # Copyright (C) 1996, 1997, 1998, 1999, 2000, 2001, 2002, 2003, 2004, # 2005, 2006, 2008, 2009 Free Software Foundation, Inc. # # This file is free software; the Free Software Foundation # gives unlimited permission to copy and/or distribute it, # with or without modifications, as long as this notice is preserved. # serial 16 # This macro actually does too much. Some checks are only needed if # your package does certain things. But this isn't really a big deal. # AM_INIT_AUTOMAKE(PACKAGE, VERSION, [NO-DEFINE]) # AM_INIT_AUTOMAKE([OPTIONS]) # ----------------------------------------------- # The call with PACKAGE and VERSION arguments is the old style # call (pre autoconf-2.50), which is being phased out. PACKAGE # and VERSION should now be passed to AC_INIT and removed from # the call to AM_INIT_AUTOMAKE. # We support both call styles for the transition. After # the next Automake release, Autoconf can make the AC_INIT # arguments mandatory, and then we can depend on a new Autoconf # release and drop the old call support. AC_DEFUN([AM_INIT_AUTOMAKE], [AC_PREREQ([2.62])dnl dnl Autoconf wants to disallow AM_ names. We explicitly allow dnl the ones we care about. m4_pattern_allow([^AM_[A-Z]+FLAGS$])dnl AC_REQUIRE([AM_SET_CURRENT_AUTOMAKE_VERSION])dnl AC_REQUIRE([AC_PROG_INSTALL])dnl if test "`cd $srcdir && pwd`" != "`pwd`"; then # Use -I$(srcdir) only when $(srcdir) != ., so that make's output # is not polluted with repeated "-I." AC_SUBST([am__isrc], [' -I$(srcdir)'])_AM_SUBST_NOTMAKE([am__isrc])dnl # test to see if srcdir already configured if test -f $srcdir/config.status; then AC_MSG_ERROR([source directory already configured; run "make distclean" there first]) fi fi # test whether we have cygpath if test -z "$CYGPATH_W"; then if (cygpath --version) >/dev/null 2>/dev/null; then CYGPATH_W='cygpath -w' else CYGPATH_W=echo fi fi AC_SUBST([CYGPATH_W]) # Define the identity of the package. dnl Distinguish between old-style and new-style calls. m4_ifval([$2], [m4_ifval([$3], [_AM_SET_OPTION([no-define])])dnl AC_SUBST([PACKAGE], [$1])dnl AC_SUBST([VERSION], [$2])], [_AM_SET_OPTIONS([$1])dnl dnl Diagnose old-style AC_INIT with new-style AM_AUTOMAKE_INIT. m4_if(m4_ifdef([AC_PACKAGE_NAME], 1)m4_ifdef([AC_PACKAGE_VERSION], 1), 11,, [m4_fatal([AC_INIT should be called with package and version arguments])])dnl AC_SUBST([PACKAGE], ['AC_PACKAGE_TARNAME'])dnl AC_SUBST([VERSION], ['AC_PACKAGE_VERSION'])])dnl _AM_IF_OPTION([no-define],, [AC_DEFINE_UNQUOTED(PACKAGE, "$PACKAGE", [Name of package]) AC_DEFINE_UNQUOTED(VERSION, "$VERSION", [Version number of package])])dnl # Some tools Automake needs. AC_REQUIRE([AM_SANITY_CHECK])dnl AC_REQUIRE([AC_ARG_PROGRAM])dnl AM_MISSING_PROG(ACLOCAL, aclocal-${am__api_version}) AM_MISSING_PROG(AUTOCONF, autoconf) AM_MISSING_PROG(AUTOMAKE, automake-${am__api_version}) AM_MISSING_PROG(AUTOHEADER, autoheader) AM_MISSING_PROG(MAKEINFO, makeinfo) AC_REQUIRE([AM_PROG_INSTALL_SH])dnl AC_REQUIRE([AM_PROG_INSTALL_STRIP])dnl AC_REQUIRE([AM_PROG_MKDIR_P])dnl # We need awk for the "check" target. The system "awk" is bad on # some platforms. AC_REQUIRE([AC_PROG_AWK])dnl AC_REQUIRE([AC_PROG_MAKE_SET])dnl AC_REQUIRE([AM_SET_LEADING_DOT])dnl _AM_IF_OPTION([tar-ustar], [_AM_PROG_TAR([ustar])], [_AM_IF_OPTION([tar-pax], [_AM_PROG_TAR([pax])], [_AM_PROG_TAR([v7])])]) _AM_IF_OPTION([no-dependencies],, [AC_PROVIDE_IFELSE([AC_PROG_CC], [_AM_DEPENDENCIES(CC)], [define([AC_PROG_CC], defn([AC_PROG_CC])[_AM_DEPENDENCIES(CC)])])dnl AC_PROVIDE_IFELSE([AC_PROG_CXX], [_AM_DEPENDENCIES(CXX)], [define([AC_PROG_CXX], defn([AC_PROG_CXX])[_AM_DEPENDENCIES(CXX)])])dnl AC_PROVIDE_IFELSE([AC_PROG_OBJC], [_AM_DEPENDENCIES(OBJC)], [define([AC_PROG_OBJC], defn([AC_PROG_OBJC])[_AM_DEPENDENCIES(OBJC)])])dnl ]) _AM_IF_OPTION([silent-rules], [AC_REQUIRE([AM_SILENT_RULES])])dnl dnl The `parallel-tests' driver may need to know about EXEEXT, so add the dnl `am__EXEEXT' conditional if _AM_COMPILER_EXEEXT was seen. This macro dnl is hooked onto _AC_COMPILER_EXEEXT early, see below. AC_CONFIG_COMMANDS_PRE(dnl [m4_provide_if([_AM_COMPILER_EXEEXT], [AM_CONDITIONAL([am__EXEEXT], [test -n "$EXEEXT"])])])dnl ]) dnl Hook into `_AC_COMPILER_EXEEXT' early to learn its expansion. Do not dnl add the conditional right here, as _AC_COMPILER_EXEEXT may be further dnl mangled by Autoconf and run in a shell conditional statement. m4_define([_AC_COMPILER_EXEEXT], m4_defn([_AC_COMPILER_EXEEXT])[m4_provide([_AM_COMPILER_EXEEXT])]) # When config.status generates a header, we must update the stamp-h file. # This file resides in the same directory as the config header # that is generated. The stamp files are numbered to have different names. # Autoconf calls _AC_AM_CONFIG_HEADER_HOOK (when defined) in the # loop where config.status creates the headers, so we can generate # our stamp files there. AC_DEFUN([_AC_AM_CONFIG_HEADER_HOOK], [# Compute $1's index in $config_headers. _am_arg=$1 _am_stamp_count=1 for _am_header in $config_headers :; do case $_am_header in $_am_arg | $_am_arg:* ) break ;; * ) _am_stamp_count=`expr $_am_stamp_count + 1` ;; esac done echo "timestamp for $_am_arg" >`AS_DIRNAME(["$_am_arg"])`/stamp-h[]$_am_stamp_count]) # Copyright (C) 2001, 2003, 2005, 2008, 2011 Free Software Foundation, # Inc. # # This file is free software; the Free Software Foundation # gives unlimited permission to copy and/or distribute it, # with or without modifications, as long as this notice is preserved. # serial 1 # AM_PROG_INSTALL_SH # ------------------ # Define $install_sh. AC_DEFUN([AM_PROG_INSTALL_SH], [AC_REQUIRE([AM_AUX_DIR_EXPAND])dnl if test x"${install_sh}" != xset; then case $am_aux_dir in *\ * | *\ *) install_sh="\${SHELL} '$am_aux_dir/install-sh'" ;; *) install_sh="\${SHELL} $am_aux_dir/install-sh" esac fi AC_SUBST(install_sh)]) # Copyright (C) 2003, 2005 Free Software Foundation, Inc. # # This file is free software; the Free Software Foundation # gives unlimited permission to copy and/or distribute it, # with or without modifications, as long as this notice is preserved. # serial 2 # Check whether the underlying file-system supports filenames # with a leading dot. For instance MS-DOS doesn't. AC_DEFUN([AM_SET_LEADING_DOT], [rm -rf .tst 2>/dev/null mkdir .tst 2>/dev/null if test -d .tst; then am__leading_dot=. else am__leading_dot=_ fi rmdir .tst 2>/dev/null AC_SUBST([am__leading_dot])]) # Add --enable-maintainer-mode option to configure. -*- Autoconf -*- # From Jim Meyering # Copyright (C) 1996, 1998, 2000, 2001, 2002, 2003, 2004, 2005, 2008, # 2011 Free Software Foundation, Inc. # # This file is free software; the Free Software Foundation # gives unlimited permission to copy and/or distribute it, # with or without modifications, as long as this notice is preserved. # serial 5 # AM_MAINTAINER_MODE([DEFAULT-MODE]) # ---------------------------------- # Control maintainer-specific portions of Makefiles. # Default is to disable them, unless `enable' is passed literally. # For symmetry, `disable' may be passed as well. Anyway, the user # can override the default with the --enable/--disable switch. AC_DEFUN([AM_MAINTAINER_MODE], [m4_case(m4_default([$1], [disable]), [enable], [m4_define([am_maintainer_other], [disable])], [disable], [m4_define([am_maintainer_other], [enable])], [m4_define([am_maintainer_other], [enable]) m4_warn([syntax], [unexpected argument to AM@&t@_MAINTAINER_MODE: $1])]) AC_MSG_CHECKING([whether to enable maintainer-specific portions of Makefiles]) dnl maintainer-mode's default is 'disable' unless 'enable' is passed AC_ARG_ENABLE([maintainer-mode], [ --][am_maintainer_other][-maintainer-mode am_maintainer_other make rules and dependencies not useful (and sometimes confusing) to the casual installer], [USE_MAINTAINER_MODE=$enableval], [USE_MAINTAINER_MODE=]m4_if(am_maintainer_other, [enable], [no], [yes])) AC_MSG_RESULT([$USE_MAINTAINER_MODE]) AM_CONDITIONAL([MAINTAINER_MODE], [test $USE_MAINTAINER_MODE = yes]) MAINT=$MAINTAINER_MODE_TRUE AC_SUBST([MAINT])dnl ] ) AU_DEFUN([jm_MAINTAINER_MODE], [AM_MAINTAINER_MODE]) # Check to see how 'make' treats includes. -*- Autoconf -*- # Copyright (C) 2001, 2002, 2003, 2005, 2009 Free Software Foundation, Inc. # # This file is free software; the Free Software Foundation # gives unlimited permission to copy and/or distribute it, # with or without modifications, as long as this notice is preserved. # serial 4 # AM_MAKE_INCLUDE() # ----------------- # Check to see how make treats includes. AC_DEFUN([AM_MAKE_INCLUDE], [am_make=${MAKE-make} cat > confinc << 'END' am__doit: @echo this is the am__doit target .PHONY: am__doit END # If we don't find an include directive, just comment out the code. AC_MSG_CHECKING([for style of include used by $am_make]) am__include="#" am__quote= _am_result=none # First try GNU make style include. echo "include confinc" > confmf # Ignore all kinds of additional output from `make'. case `$am_make -s -f confmf 2> /dev/null` in #( *the\ am__doit\ target*) am__include=include am__quote= _am_result=GNU ;; esac # Now try BSD make style include. if test "$am__include" = "#"; then echo '.include "confinc"' > confmf case `$am_make -s -f confmf 2> /dev/null` in #( *the\ am__doit\ target*) am__include=.include am__quote="\"" _am_result=BSD ;; esac fi AC_SUBST([am__include]) AC_SUBST([am__quote]) AC_MSG_RESULT([$_am_result]) rm -f confinc confmf ]) # Copyright (C) 1999, 2000, 2001, 2003, 2004, 2005, 2008 # Free Software Foundation, Inc. # # This file is free software; the Free Software Foundation # gives unlimited permission to copy and/or distribute it, # with or without modifications, as long as this notice is preserved. # serial 6 # AM_PROG_CC_C_O # -------------- # Like AC_PROG_CC_C_O, but changed for automake. AC_DEFUN([AM_PROG_CC_C_O], [AC_REQUIRE([AC_PROG_CC_C_O])dnl AC_REQUIRE([AM_AUX_DIR_EXPAND])dnl AC_REQUIRE_AUX_FILE([compile])dnl # FIXME: we rely on the cache variable name because # there is no other way. set dummy $CC am_cc=`echo $[2] | sed ['s/[^a-zA-Z0-9_]/_/g;s/^[0-9]/_/']` eval am_t=\$ac_cv_prog_cc_${am_cc}_c_o if test "$am_t" != yes; then # Losing compiler, so override with the script. # FIXME: It is wrong to rewrite CC. # But if we don't then we get into trouble of one sort or another. # A longer-term fix would be to have automake use am__CC in this case, # and then we could set am__CC="\$(top_srcdir)/compile \$(CC)" CC="$am_aux_dir/compile $CC" fi dnl Make sure AC_PROG_CC is never called again, or it will override our dnl setting of CC. m4_define([AC_PROG_CC], [m4_fatal([AC_PROG_CC cannot be called after AM_PROG_CC_C_O])]) ]) # Fake the existence of programs that GNU maintainers use. -*- Autoconf -*- # Copyright (C) 1997, 1999, 2000, 2001, 2003, 2004, 2005, 2008 # Free Software Foundation, Inc. # # This file is free software; the Free Software Foundation # gives unlimited permission to copy and/or distribute it, # with or without modifications, as long as this notice is preserved. # serial 6 # AM_MISSING_PROG(NAME, PROGRAM) # ------------------------------ AC_DEFUN([AM_MISSING_PROG], [AC_REQUIRE([AM_MISSING_HAS_RUN]) $1=${$1-"${am_missing_run}$2"} AC_SUBST($1)]) # AM_MISSING_HAS_RUN # ------------------ # Define MISSING if not defined so far and test if it supports --run. # If it does, set am_missing_run to use it, otherwise, to nothing. AC_DEFUN([AM_MISSING_HAS_RUN], [AC_REQUIRE([AM_AUX_DIR_EXPAND])dnl AC_REQUIRE_AUX_FILE([missing])dnl if test x"${MISSING+set}" != xset; then case $am_aux_dir in *\ * | *\ *) MISSING="\${SHELL} \"$am_aux_dir/missing\"" ;; *) MISSING="\${SHELL} $am_aux_dir/missing" ;; esac fi # Use eval to expand $SHELL if eval "$MISSING --run true"; then am_missing_run="$MISSING --run " else am_missing_run= AC_MSG_WARN([`missing' script is too old or missing]) fi ]) # Copyright (C) 2003, 2004, 2005, 2006, 2011 Free Software Foundation, # Inc. # # This file is free software; the Free Software Foundation # gives unlimited permission to copy and/or distribute it, # with or without modifications, as long as this notice is preserved. # serial 1 # AM_PROG_MKDIR_P # --------------- # Check for `mkdir -p'. AC_DEFUN([AM_PROG_MKDIR_P], [AC_PREREQ([2.60])dnl AC_REQUIRE([AC_PROG_MKDIR_P])dnl dnl Automake 1.8 to 1.9.6 used to define mkdir_p. We now use MKDIR_P, dnl while keeping a definition of mkdir_p for backward compatibility. dnl @MKDIR_P@ is magic: AC_OUTPUT adjusts its value for each Makefile. dnl However we cannot define mkdir_p as $(MKDIR_P) for the sake of dnl Makefile.ins that do not define MKDIR_P, so we do our own dnl adjustment using top_builddir (which is defined more often than dnl MKDIR_P). AC_SUBST([mkdir_p], ["$MKDIR_P"])dnl case $mkdir_p in [[\\/$]]* | ?:[[\\/]]*) ;; */*) mkdir_p="\$(top_builddir)/$mkdir_p" ;; esac ]) # Helper functions for option handling. -*- Autoconf -*- # Copyright (C) 2001, 2002, 2003, 2005, 2008, 2010 Free Software # Foundation, Inc. # # This file is free software; the Free Software Foundation # gives unlimited permission to copy and/or distribute it, # with or without modifications, as long as this notice is preserved. # serial 5 # _AM_MANGLE_OPTION(NAME) # ----------------------- AC_DEFUN([_AM_MANGLE_OPTION], [[_AM_OPTION_]m4_bpatsubst($1, [[^a-zA-Z0-9_]], [_])]) # _AM_SET_OPTION(NAME) # -------------------- # Set option NAME. Presently that only means defining a flag for this option. AC_DEFUN([_AM_SET_OPTION], [m4_define(_AM_MANGLE_OPTION([$1]), 1)]) # _AM_SET_OPTIONS(OPTIONS) # ------------------------ # OPTIONS is a space-separated list of Automake options. AC_DEFUN([_AM_SET_OPTIONS], [m4_foreach_w([_AM_Option], [$1], [_AM_SET_OPTION(_AM_Option)])]) # _AM_IF_OPTION(OPTION, IF-SET, [IF-NOT-SET]) # ------------------------------------------- # Execute IF-SET if OPTION is set, IF-NOT-SET otherwise. AC_DEFUN([_AM_IF_OPTION], [m4_ifset(_AM_MANGLE_OPTION([$1]), [$2], [$3])]) # Check to make sure that the build environment is sane. -*- Autoconf -*- # Copyright (C) 1996, 1997, 2000, 2001, 2003, 2005, 2008 # Free Software Foundation, Inc. # # This file is free software; the Free Software Foundation # gives unlimited permission to copy and/or distribute it, # with or without modifications, as long as this notice is preserved. # serial 5 # AM_SANITY_CHECK # --------------- AC_DEFUN([AM_SANITY_CHECK], [AC_MSG_CHECKING([whether build environment is sane]) # Just in case sleep 1 echo timestamp > conftest.file # Reject unsafe characters in $srcdir or the absolute working directory # name. Accept space and tab only in the latter. am_lf=' ' case `pwd` in *[[\\\"\#\$\&\'\`$am_lf]]*) AC_MSG_ERROR([unsafe absolute working directory name]);; esac case $srcdir in *[[\\\"\#\$\&\'\`$am_lf\ \ ]]*) AC_MSG_ERROR([unsafe srcdir value: `$srcdir']);; esac # Do `set' in a subshell so we don't clobber the current shell's # arguments. Must try -L first in case configure is actually a # symlink; some systems play weird games with the mod time of symlinks # (eg FreeBSD returns the mod time of the symlink's containing # directory). if ( set X `ls -Lt "$srcdir/configure" conftest.file 2> /dev/null` if test "$[*]" = "X"; then # -L didn't work. set X `ls -t "$srcdir/configure" conftest.file` fi rm -f conftest.file if test "$[*]" != "X $srcdir/configure conftest.file" \ && test "$[*]" != "X conftest.file $srcdir/configure"; then # If neither matched, then we have a broken ls. This can happen # if, for instance, CONFIG_SHELL is bash and it inherits a # broken ls alias from the environment. This has actually # happened. Such a system could not be considered "sane". AC_MSG_ERROR([ls -t appears to fail. Make sure there is not a broken alias in your environment]) fi test "$[2]" = conftest.file ) then # Ok. : else AC_MSG_ERROR([newly created file is older than distributed files! Check your system clock]) fi AC_MSG_RESULT(yes)]) # Copyright (C) 2001, 2003, 2005, 2011 Free Software Foundation, Inc. # # This file is free software; the Free Software Foundation # gives unlimited permission to copy and/or distribute it, # with or without modifications, as long as this notice is preserved. # serial 1 # AM_PROG_INSTALL_STRIP # --------------------- # One issue with vendor `install' (even GNU) is that you can't # specify the program used to strip binaries. This is especially # annoying in cross-compiling environments, where the build's strip # is unlikely to handle the host's binaries. # Fortunately install-sh will honor a STRIPPROG variable, so we # always use install-sh in `make install-strip', and initialize # STRIPPROG with the value of the STRIP variable (set by the user). AC_DEFUN([AM_PROG_INSTALL_STRIP], [AC_REQUIRE([AM_PROG_INSTALL_SH])dnl # Installed binaries are usually stripped using `strip' when the user # run `make install-strip'. However `strip' might not be the right # tool to use in cross-compilation environments, therefore Automake # will honor the `STRIP' environment variable to overrule this program. dnl Don't test for $cross_compiling = yes, because it might be `maybe'. if test "$cross_compiling" != no; then AC_CHECK_TOOL([STRIP], [strip], :) fi INSTALL_STRIP_PROGRAM="\$(install_sh) -c -s" AC_SUBST([INSTALL_STRIP_PROGRAM])]) # Copyright (C) 2006, 2008, 2010 Free Software Foundation, Inc. # # This file is free software; the Free Software Foundation # gives unlimited permission to copy and/or distribute it, # with or without modifications, as long as this notice is preserved. # serial 3 # _AM_SUBST_NOTMAKE(VARIABLE) # --------------------------- # Prevent Automake from outputting VARIABLE = @VARIABLE@ in Makefile.in. # This macro is traced by Automake. AC_DEFUN([_AM_SUBST_NOTMAKE]) # AM_SUBST_NOTMAKE(VARIABLE) # -------------------------- # Public sister of _AM_SUBST_NOTMAKE. AC_DEFUN([AM_SUBST_NOTMAKE], [_AM_SUBST_NOTMAKE($@)]) # Check how to create a tarball. -*- Autoconf -*- # Copyright (C) 2004, 2005, 2012 Free Software Foundation, Inc. # # This file is free software; the Free Software Foundation # gives unlimited permission to copy and/or distribute it, # with or without modifications, as long as this notice is preserved. # serial 2 # _AM_PROG_TAR(FORMAT) # -------------------- # Check how to create a tarball in format FORMAT. # FORMAT should be one of `v7', `ustar', or `pax'. # # Substitute a variable $(am__tar) that is a command # writing to stdout a FORMAT-tarball containing the directory # $tardir. # tardir=directory && $(am__tar) > result.tar # # Substitute a variable $(am__untar) that extract such # a tarball read from stdin. # $(am__untar) < result.tar AC_DEFUN([_AM_PROG_TAR], [# Always define AMTAR for backward compatibility. Yes, it's still used # in the wild :-( We should find a proper way to deprecate it ... AC_SUBST([AMTAR], ['$${TAR-tar}']) m4_if([$1], [v7], [am__tar='$${TAR-tar} chof - "$$tardir"' am__untar='$${TAR-tar} xf -'], [m4_case([$1], [ustar],, [pax],, [m4_fatal([Unknown tar format])]) AC_MSG_CHECKING([how to create a $1 tar archive]) # Loop over all known methods to create a tar archive until one works. _am_tools='gnutar m4_if([$1], [ustar], [plaintar]) pax cpio none' _am_tools=${am_cv_prog_tar_$1-$_am_tools} # Do not fold the above two line into one, because Tru64 sh and # Solaris sh will not grok spaces in the rhs of `-'. for _am_tool in $_am_tools do case $_am_tool in gnutar) for _am_tar in tar gnutar gtar; do AM_RUN_LOG([$_am_tar --version]) && break done am__tar="$_am_tar --format=m4_if([$1], [pax], [posix], [$1]) -chf - "'"$$tardir"' am__tar_="$_am_tar --format=m4_if([$1], [pax], [posix], [$1]) -chf - "'"$tardir"' am__untar="$_am_tar -xf -" ;; plaintar) # Must skip GNU tar: if it does not support --format= it doesn't create # ustar tarball either. (tar --version) >/dev/null 2>&1 && continue am__tar='tar chf - "$$tardir"' am__tar_='tar chf - "$tardir"' am__untar='tar xf -' ;; pax) am__tar='pax -L -x $1 -w "$$tardir"' am__tar_='pax -L -x $1 -w "$tardir"' am__untar='pax -r' ;; cpio) am__tar='find "$$tardir" -print | cpio -o -H $1 -L' am__tar_='find "$tardir" -print | cpio -o -H $1 -L' am__untar='cpio -i -H $1 -d' ;; none) am__tar=false am__tar_=false am__untar=false ;; esac # If the value was cached, stop now. We just wanted to have am__tar # and am__untar set. test -n "${am_cv_prog_tar_$1}" && break # tar/untar a dummy directory, and stop if the command works rm -rf conftest.dir mkdir conftest.dir echo GrepMe > conftest.dir/file AM_RUN_LOG([tardir=conftest.dir && eval $am__tar_ >conftest.tar]) rm -rf conftest.dir if test -s conftest.tar; then AM_RUN_LOG([$am__untar /dev/null 2>&1 && break fi done rm -rf conftest.dir AC_CACHE_VAL([am_cv_prog_tar_$1], [am_cv_prog_tar_$1=$_am_tool]) AC_MSG_RESULT([$am_cv_prog_tar_$1])]) AC_SUBST([am__tar]) AC_SUBST([am__untar]) ]) # _AM_PROG_TAR m4_include([m4/acx_mpi.m4]) m4_include([m4/acx_pthread.m4]) m4_include([m4/ax_cc_maxopt.m4]) m4_include([m4/ax_check_compiler_flags.m4]) m4_include([m4/ax_compiler_vendor.m4]) m4_include([m4/ax_gcc_aligns_stack.m4]) m4_include([m4/ax_gcc_version.m4]) m4_include([m4/ax_openmp.m4]) m4_include([m4/libtool.m4]) m4_include([m4/ltoptions.m4]) m4_include([m4/ltsugar.m4]) m4_include([m4/ltversion.m4]) m4_include([m4/lt~obsolete.m4]) fftw-3.3.3/Makefile.am0000644000175000017500000000545412054410050011442 00000000000000OPTIONS_AUTOMAKE=gnu lib_LTLIBRARIES = libfftw3@PREC_SUFFIX@.la # pkgincludedir = $(includedir)/fftw3@PREC_SUFFIX@ # nodist_pkginclude_HEADERS = config.h # recompile genfft if maintainer mode is true if MAINTAINER_MODE GENFFT = genfft else GENFFT = endif ACLOCAL_AMFLAGS=-I m4 # when using combined thread libraries (necessary on Windows), we want # to build threads/ first, because libfftw3_threads is added to # libfftw3. # # Otherwise, we want to build libfftw3_threads after libfftw3 # so that we can track the fact that libfftw3_threads depends upon # libfftw3. # # This is the inescapable result of combining three bad ideas # (threads, Windows, and shared libraries). # if COMBINED_THREADS CHICKEN_EGG=threads . else CHICKEN_EGG=. threads endif SUBDIRS=support $(GENFFT) kernel simd-support dft rdft reodft api \ libbench2 $(CHICKEN_EGG) tests mpi doc tools m4 EXTRA_DIST=COPYRIGHT bootstrap.sh CONVENTIONS fftw.pc.in SIMD_LIBS = \ simd-support/libsimd_support.la \ simd-support/libsimd_sse2_nonportable.la if HAVE_SSE2 SSE2_LIBS = dft/simd/sse2/libdft_sse2_codelets.la \ rdft/simd/sse2/librdft_sse2_codelets.la endif if HAVE_AVX AVX_LIBS = dft/simd/avx/libdft_avx_codelets.la \ rdft/simd/avx/librdft_avx_codelets.la endif if HAVE_ALTIVEC ALTIVEC_LIBS = dft/simd/altivec/libdft_altivec_codelets.la \ rdft/simd/altivec/librdft_altivec_codelets.la endif if HAVE_NEON NEON_LIBS = dft/simd/neon/libdft_neon_codelets.la \ rdft/simd/neon/librdft_neon_codelets.la endif if THREADS if COMBINED_THREADS COMBINED_THREADLIBS=threads/libfftw3@PREC_SUFFIX@_threads.la endif endif libfftw3@PREC_SUFFIX@_la_SOURCES = libfftw3@PREC_SUFFIX@_la_LIBADD = \ kernel/libkernel.la \ dft/libdft.la \ dft/scalar/libdft_scalar.la \ dft/scalar/codelets/libdft_scalar_codelets.la \ rdft/librdft.la \ rdft/scalar/librdft_scalar.la \ rdft/scalar/r2cf/librdft_scalar_r2cf.la \ rdft/scalar/r2cb/librdft_scalar_r2cb.la \ rdft/scalar/r2r/librdft_scalar_r2r.la \ reodft/libreodft.la \ api/libapi.la \ $(SIMD_LIBS) $(SSE2_LIBS) $(AVX_LIBS) $(ALTIVEC_LIBS) $(NEON_LIBS) \ $(COMBINED_THREADLIBS) if QUAD # cannot use -no-undefined since dependent on libquadmath libfftw3@PREC_SUFFIX@_la_LDFLAGS = -version-info @SHARED_VERSION_INFO@ else libfftw3@PREC_SUFFIX@_la_LDFLAGS = -no-undefined -version-info \ @SHARED_VERSION_INFO@ endif fftw3@PREC_SUFFIX@.pc: fftw.pc cp -f fftw.pc fftw3@PREC_SUFFIX@.pc pkgconfigdir = $(libdir)/pkgconfig pkgconfig_DATA = fftw3@PREC_SUFFIX@.pc WISDOM_DIR = /etc/fftw WISDOM = wisdom@PREC_SUFFIX@ WISDOM_TIME=12 # default to 12-hour limit, i.e. overnight WISDOM_FLAGS=--verbose --canonical --time-limit=$(WISDOM_TIME) wisdom: tools/fftw@PREC_SUFFIX@-wisdom -o $@ $(WISDOM_FLAGS) install-wisdom: wisdom $(mkinstalldirs) $(WISDOM_DIR) $(INSTALL_DATA) wisdom $(WISDOM_DIR)/$(WISDOM) fftw-3.3.3/Makefile.in0000644000175000017500000010200512054410201011437 00000000000000# Makefile.in generated by automake 1.11.6 from Makefile.am. # @configure_input@ # Copyright (C) 1994, 1995, 1996, 1997, 1998, 1999, 2000, 2001, 2002, # 2003, 2004, 2005, 2006, 2007, 2008, 2009, 2010, 2011 Free Software # Foundation, Inc. # This Makefile.in is free software; the Free Software Foundation # gives unlimited permission to copy and/or distribute it, # with or without modifications, as long as this notice is preserved. # This program is distributed in the hope that it will be useful, # but WITHOUT ANY WARRANTY, to the extent permitted by law; without # even the implied warranty of MERCHANTABILITY or FITNESS FOR A # PARTICULAR PURPOSE. @SET_MAKE@ VPATH = @srcdir@ am__make_dryrun = \ { \ am__dry=no; \ case $$MAKEFLAGS in \ *\\[\ \ ]*) \ echo 'am--echo: ; @echo "AM" OK' | $(MAKE) -f - 2>/dev/null \ | grep '^AM OK$$' >/dev/null || am__dry=yes;; \ *) \ for am__flg in $$MAKEFLAGS; do \ case $$am__flg in \ *=*|--*) ;; \ *n*) am__dry=yes; break;; \ esac; \ done;; \ esac; \ test $$am__dry = yes; \ } pkgdatadir = $(datadir)/@PACKAGE@ pkgincludedir = $(includedir)/@PACKAGE@ pkglibdir = $(libdir)/@PACKAGE@ pkglibexecdir = $(libexecdir)/@PACKAGE@ am__cd = CDPATH="$${ZSH_VERSION+.}$(PATH_SEPARATOR)" && cd install_sh_DATA = $(install_sh) -c -m 644 install_sh_PROGRAM = $(install_sh) -c install_sh_SCRIPT = $(install_sh) -c INSTALL_HEADER = $(INSTALL_DATA) transform = $(program_transform_name) NORMAL_INSTALL = : PRE_INSTALL = : POST_INSTALL = : NORMAL_UNINSTALL = : PRE_UNINSTALL = : POST_UNINSTALL = : build_triplet = @build@ host_triplet = @host@ subdir = . DIST_COMMON = README $(am__configure_deps) $(srcdir)/Makefile.am \ $(srcdir)/Makefile.in $(srcdir)/config.h.in \ $(srcdir)/fftw.pc.in $(top_srcdir)/configure AUTHORS COPYING \ ChangeLog INSTALL NEWS TODO compile config.guess config.sub \ depcomp install-sh ltmain.sh missing ACLOCAL_M4 = $(top_srcdir)/aclocal.m4 am__aclocal_m4_deps = $(top_srcdir)/m4/acx_mpi.m4 \ $(top_srcdir)/m4/acx_pthread.m4 \ $(top_srcdir)/m4/ax_cc_maxopt.m4 \ $(top_srcdir)/m4/ax_check_compiler_flags.m4 \ $(top_srcdir)/m4/ax_compiler_vendor.m4 \ $(top_srcdir)/m4/ax_gcc_aligns_stack.m4 \ $(top_srcdir)/m4/ax_gcc_version.m4 \ $(top_srcdir)/m4/ax_openmp.m4 $(top_srcdir)/m4/libtool.m4 \ $(top_srcdir)/m4/ltoptions.m4 $(top_srcdir)/m4/ltsugar.m4 \ $(top_srcdir)/m4/ltversion.m4 $(top_srcdir)/m4/lt~obsolete.m4 \ $(top_srcdir)/configure.ac am__configure_deps = $(am__aclocal_m4_deps) $(CONFIGURE_DEPENDENCIES) \ $(ACLOCAL_M4) am__CONFIG_DISTCLEAN_FILES = config.status config.cache config.log \ configure.lineno config.status.lineno mkinstalldirs = $(install_sh) -d CONFIG_HEADER = config.h CONFIG_CLEAN_FILES = fftw.pc CONFIG_CLEAN_VPATH_FILES = am__vpath_adj_setup = srcdirstrip=`echo "$(srcdir)" | sed 's|.|.|g'`; am__vpath_adj = case $$p in \ $(srcdir)/*) f=`echo "$$p" | sed "s|^$$srcdirstrip/||"`;; \ *) f=$$p;; \ esac; am__strip_dir = f=`echo $$p | sed -e 's|^.*/||'`; am__install_max = 40 am__nobase_strip_setup = \ srcdirstrip=`echo "$(srcdir)" | sed 's/[].[^$$\\*|]/\\\\&/g'` am__nobase_strip = \ for p in $$list; do echo "$$p"; done | sed -e "s|$$srcdirstrip/||" am__nobase_list = $(am__nobase_strip_setup); \ for p in $$list; do echo "$$p $$p"; done | \ sed "s| $$srcdirstrip/| |;"' / .*\//!s/ .*/ ./; s,\( .*\)/[^/]*$$,\1,' | \ $(AWK) 'BEGIN { files["."] = "" } { files[$$2] = files[$$2] " " $$1; \ if (++n[$$2] == $(am__install_max)) \ { print $$2, files[$$2]; n[$$2] = 0; files[$$2] = "" } } \ END { for (dir in files) print dir, files[dir] }' am__base_list = \ sed '$$!N;$$!N;$$!N;$$!N;$$!N;$$!N;$$!N;s/\n/ /g' | \ sed '$$!N;$$!N;$$!N;$$!N;s/\n/ /g' am__uninstall_files_from_dir = { \ test -z "$$files" \ || { test ! -d "$$dir" && test ! -f "$$dir" && test ! -r "$$dir"; } \ || { echo " ( cd '$$dir' && rm -f" $$files ")"; \ $(am__cd) "$$dir" && rm -f $$files; }; \ } am__installdirs = "$(DESTDIR)$(libdir)" "$(DESTDIR)$(pkgconfigdir)" LTLIBRARIES = $(lib_LTLIBRARIES) libfftw3@PREC_SUFFIX@_la_DEPENDENCIES = kernel/libkernel.la \ dft/libdft.la dft/scalar/libdft_scalar.la \ dft/scalar/codelets/libdft_scalar_codelets.la rdft/librdft.la \ rdft/scalar/librdft_scalar.la \ rdft/scalar/r2cf/librdft_scalar_r2cf.la \ rdft/scalar/r2cb/librdft_scalar_r2cb.la \ rdft/scalar/r2r/librdft_scalar_r2r.la reodft/libreodft.la \ api/libapi.la $(SIMD_LIBS) $(SSE2_LIBS) $(AVX_LIBS) \ $(ALTIVEC_LIBS) $(NEON_LIBS) $(COMBINED_THREADLIBS) am_libfftw3@PREC_SUFFIX@_la_OBJECTS = libfftw3@PREC_SUFFIX@_la_OBJECTS = \ $(am_libfftw3@PREC_SUFFIX@_la_OBJECTS) libfftw3@PREC_SUFFIX@_la_LINK = $(LIBTOOL) --tag=CC $(AM_LIBTOOLFLAGS) \ $(LIBTOOLFLAGS) --mode=link $(CCLD) $(AM_CFLAGS) $(CFLAGS) \ $(libfftw3@PREC_SUFFIX@_la_LDFLAGS) $(LDFLAGS) -o $@ DEFAULT_INCLUDES = -I.@am__isrc@ COMPILE = $(CC) $(DEFS) $(DEFAULT_INCLUDES) $(INCLUDES) $(AM_CPPFLAGS) \ $(CPPFLAGS) $(AM_CFLAGS) $(CFLAGS) LTCOMPILE = $(LIBTOOL) --tag=CC $(AM_LIBTOOLFLAGS) $(LIBTOOLFLAGS) \ --mode=compile $(CC) $(DEFS) $(DEFAULT_INCLUDES) $(INCLUDES) \ $(AM_CPPFLAGS) $(CPPFLAGS) $(AM_CFLAGS) $(CFLAGS) CCLD = $(CC) LINK = $(LIBTOOL) --tag=CC $(AM_LIBTOOLFLAGS) $(LIBTOOLFLAGS) \ --mode=link $(CCLD) $(AM_CFLAGS) $(CFLAGS) $(AM_LDFLAGS) \ $(LDFLAGS) -o $@ SOURCES = $(libfftw3@PREC_SUFFIX@_la_SOURCES) DIST_SOURCES = $(libfftw3@PREC_SUFFIX@_la_SOURCES) RECURSIVE_TARGETS = all-recursive check-recursive dvi-recursive \ html-recursive info-recursive install-data-recursive \ install-dvi-recursive install-exec-recursive \ install-html-recursive install-info-recursive \ install-pdf-recursive install-ps-recursive install-recursive \ installcheck-recursive installdirs-recursive pdf-recursive \ ps-recursive uninstall-recursive am__can_run_installinfo = \ case $$AM_UPDATE_INFO_DIR in \ n|no|NO) false;; \ *) (install-info --version) >/dev/null 2>&1;; \ esac DATA = $(pkgconfig_DATA) RECURSIVE_CLEAN_TARGETS = mostlyclean-recursive clean-recursive \ distclean-recursive maintainer-clean-recursive AM_RECURSIVE_TARGETS = $(RECURSIVE_TARGETS:-recursive=) \ $(RECURSIVE_CLEAN_TARGETS:-recursive=) tags TAGS ctags CTAGS \ distdir dist dist-all distcheck ETAGS = etags CTAGS = ctags DIST_SUBDIRS = support genfft kernel simd-support dft rdft reodft api \ libbench2 . threads tests mpi doc tools m4 DISTFILES = $(DIST_COMMON) $(DIST_SOURCES) $(TEXINFOS) $(EXTRA_DIST) distdir = $(PACKAGE)-$(VERSION) top_distdir = $(distdir) am__remove_distdir = \ if test -d "$(distdir)"; then \ find "$(distdir)" -type d ! -perm -200 -exec chmod u+w {} ';' \ && rm -rf "$(distdir)" \ || { sleep 5 && rm -rf "$(distdir)"; }; \ else :; fi am__relativize = \ dir0=`pwd`; \ sed_first='s,^\([^/]*\)/.*$$,\1,'; \ sed_rest='s,^[^/]*/*,,'; \ sed_last='s,^.*/\([^/]*\)$$,\1,'; \ sed_butlast='s,/*[^/]*$$,,'; \ while test -n "$$dir1"; do \ first=`echo "$$dir1" | sed -e "$$sed_first"`; \ if test "$$first" != "."; then \ if test "$$first" = ".."; then \ dir2=`echo "$$dir0" | sed -e "$$sed_last"`/"$$dir2"; \ dir0=`echo "$$dir0" | sed -e "$$sed_butlast"`; \ else \ first2=`echo "$$dir2" | sed -e "$$sed_first"`; \ if test "$$first2" = "$$first"; then \ dir2=`echo "$$dir2" | sed -e "$$sed_rest"`; \ else \ dir2="../$$dir2"; \ fi; \ dir0="$$dir0"/"$$first"; \ fi; \ fi; \ dir1=`echo "$$dir1" | sed -e "$$sed_rest"`; \ done; \ reldir="$$dir2" DIST_ARCHIVES = $(distdir).tar.gz GZIP_ENV = --best distuninstallcheck_listfiles = find . -type f -print am__distuninstallcheck_listfiles = $(distuninstallcheck_listfiles) \ | sed 's|^\./|$(prefix)/|' | grep -v '$(infodir)/dir$$' distcleancheck_listfiles = find . -type f -print ACLOCAL = @ACLOCAL@ ALLOCA = @ALLOCA@ ALTIVEC_CFLAGS = @ALTIVEC_CFLAGS@ AMTAR = @AMTAR@ AR = @AR@ AS = @AS@ AUTOCONF = @AUTOCONF@ AUTOHEADER = @AUTOHEADER@ AUTOMAKE = @AUTOMAKE@ AVX_CFLAGS = @AVX_CFLAGS@ AWK = @AWK@ CC = @CC@ CCDEPMODE = @CCDEPMODE@ CFLAGS = @CFLAGS@ CHECK_PL_OPTS = @CHECK_PL_OPTS@ CPP = @CPP@ CPPFLAGS = @CPPFLAGS@ CYGPATH_W = @CYGPATH_W@ C_FFTW_R2R_KIND = @C_FFTW_R2R_KIND@ C_MPI_FINT = @C_MPI_FINT@ DEFS = @DEFS@ DEPDIR = @DEPDIR@ DLLTOOL = @DLLTOOL@ DSYMUTIL = @DSYMUTIL@ DUMPBIN = @DUMPBIN@ ECHO_C = @ECHO_C@ ECHO_N = @ECHO_N@ ECHO_T = @ECHO_T@ EGREP = @EGREP@ EXEEXT = @EXEEXT@ F77 = @F77@ FFLAGS = @FFLAGS@ FGREP = @FGREP@ FLIBS = @FLIBS@ GREP = @GREP@ INSTALL = @INSTALL@ INSTALL_DATA = @INSTALL_DATA@ INSTALL_PROGRAM = @INSTALL_PROGRAM@ INSTALL_SCRIPT = @INSTALL_SCRIPT@ INSTALL_STRIP_PROGRAM = @INSTALL_STRIP_PROGRAM@ LD = @LD@ LDFLAGS = @LDFLAGS@ LIBOBJS = @LIBOBJS@ LIBQUADMATH = @LIBQUADMATH@ LIBS = @LIBS@ LIBTOOL = @LIBTOOL@ LIPO = @LIPO@ LN_S = @LN_S@ LTLIBOBJS = @LTLIBOBJS@ MAINT = @MAINT@ MAKEINFO = @MAKEINFO@ MANIFEST_TOOL = @MANIFEST_TOOL@ MKDIR_P = @MKDIR_P@ MPICC = @MPICC@ MPILIBS = @MPILIBS@ MPIRUN = @MPIRUN@ NEON_CFLAGS = @NEON_CFLAGS@ NM = @NM@ NMEDIT = @NMEDIT@ OBJDUMP = @OBJDUMP@ OBJEXT = @OBJEXT@ OCAMLBUILD = @OCAMLBUILD@ OPENMP_CFLAGS = @OPENMP_CFLAGS@ OTOOL = @OTOOL@ OTOOL64 = @OTOOL64@ PACKAGE = @PACKAGE@ PACKAGE_BUGREPORT = @PACKAGE_BUGREPORT@ PACKAGE_NAME = @PACKAGE_NAME@ PACKAGE_STRING = @PACKAGE_STRING@ PACKAGE_TARNAME = @PACKAGE_TARNAME@ PACKAGE_URL = @PACKAGE_URL@ PACKAGE_VERSION = @PACKAGE_VERSION@ PATH_SEPARATOR = @PATH_SEPARATOR@ POW_LIB = @POW_LIB@ PRECISION = @PRECISION@ PREC_SUFFIX = @PREC_SUFFIX@ PTHREAD_CC = @PTHREAD_CC@ PTHREAD_CFLAGS = @PTHREAD_CFLAGS@ PTHREAD_LIBS = @PTHREAD_LIBS@ RANLIB = @RANLIB@ SED = @SED@ SET_MAKE = @SET_MAKE@ SHARED_VERSION_INFO = @SHARED_VERSION_INFO@ SHELL = @SHELL@ SSE2_CFLAGS = @SSE2_CFLAGS@ STACK_ALIGN_CFLAGS = @STACK_ALIGN_CFLAGS@ STRIP = @STRIP@ THREADLIBS = @THREADLIBS@ VERSION = @VERSION@ abs_builddir = @abs_builddir@ abs_srcdir = @abs_srcdir@ abs_top_builddir = @abs_top_builddir@ abs_top_srcdir = @abs_top_srcdir@ ac_ct_AR = @ac_ct_AR@ ac_ct_CC = @ac_ct_CC@ ac_ct_DUMPBIN = @ac_ct_DUMPBIN@ ac_ct_F77 = @ac_ct_F77@ acx_pthread_config = @acx_pthread_config@ am__include = @am__include@ am__leading_dot = @am__leading_dot@ am__quote = @am__quote@ am__tar = @am__tar@ am__untar = @am__untar@ bindir = @bindir@ build = @build@ build_alias = @build_alias@ build_cpu = @build_cpu@ build_os = @build_os@ build_vendor = @build_vendor@ builddir = @builddir@ datadir = @datadir@ datarootdir = @datarootdir@ docdir = @docdir@ dvidir = @dvidir@ exec_prefix = @exec_prefix@ host = @host@ host_alias = @host_alias@ host_cpu = @host_cpu@ host_os = @host_os@ host_vendor = @host_vendor@ htmldir = @htmldir@ includedir = @includedir@ infodir = @infodir@ install_sh = @install_sh@ libdir = @libdir@ libexecdir = @libexecdir@ localedir = @localedir@ localstatedir = @localstatedir@ mandir = @mandir@ mkdir_p = @mkdir_p@ oldincludedir = @oldincludedir@ pdfdir = @pdfdir@ prefix = @prefix@ program_transform_name = @program_transform_name@ psdir = @psdir@ sbindir = @sbindir@ sharedstatedir = @sharedstatedir@ srcdir = @srcdir@ sysconfdir = @sysconfdir@ target_alias = @target_alias@ top_build_prefix = @top_build_prefix@ top_builddir = @top_builddir@ top_srcdir = @top_srcdir@ OPTIONS_AUTOMAKE = gnu lib_LTLIBRARIES = libfftw3@PREC_SUFFIX@.la @MAINTAINER_MODE_FALSE@GENFFT = # pkgincludedir = $(includedir)/fftw3@PREC_SUFFIX@ # nodist_pkginclude_HEADERS = config.h # recompile genfft if maintainer mode is true @MAINTAINER_MODE_TRUE@GENFFT = genfft ACLOCAL_AMFLAGS = -I m4 @COMBINED_THREADS_FALSE@CHICKEN_EGG = . threads # when using combined thread libraries (necessary on Windows), we want # to build threads/ first, because libfftw3_threads is added to # libfftw3. # # Otherwise, we want to build libfftw3_threads after libfftw3 # so that we can track the fact that libfftw3_threads depends upon # libfftw3. # # This is the inescapable result of combining three bad ideas # (threads, Windows, and shared libraries). # @COMBINED_THREADS_TRUE@CHICKEN_EGG = threads . SUBDIRS = support $(GENFFT) kernel simd-support dft rdft reodft api \ libbench2 $(CHICKEN_EGG) tests mpi doc tools m4 EXTRA_DIST = COPYRIGHT bootstrap.sh CONVENTIONS fftw.pc.in SIMD_LIBS = \ simd-support/libsimd_support.la \ simd-support/libsimd_sse2_nonportable.la @HAVE_SSE2_TRUE@SSE2_LIBS = dft/simd/sse2/libdft_sse2_codelets.la \ @HAVE_SSE2_TRUE@rdft/simd/sse2/librdft_sse2_codelets.la @HAVE_AVX_TRUE@AVX_LIBS = dft/simd/avx/libdft_avx_codelets.la \ @HAVE_AVX_TRUE@rdft/simd/avx/librdft_avx_codelets.la @HAVE_ALTIVEC_TRUE@ALTIVEC_LIBS = dft/simd/altivec/libdft_altivec_codelets.la \ @HAVE_ALTIVEC_TRUE@rdft/simd/altivec/librdft_altivec_codelets.la @HAVE_NEON_TRUE@NEON_LIBS = dft/simd/neon/libdft_neon_codelets.la \ @HAVE_NEON_TRUE@rdft/simd/neon/librdft_neon_codelets.la @COMBINED_THREADS_TRUE@@THREADS_TRUE@COMBINED_THREADLIBS = threads/libfftw3@PREC_SUFFIX@_threads.la libfftw3@PREC_SUFFIX@_la_SOURCES = libfftw3@PREC_SUFFIX@_la_LIBADD = \ kernel/libkernel.la \ dft/libdft.la \ dft/scalar/libdft_scalar.la \ dft/scalar/codelets/libdft_scalar_codelets.la \ rdft/librdft.la \ rdft/scalar/librdft_scalar.la \ rdft/scalar/r2cf/librdft_scalar_r2cf.la \ rdft/scalar/r2cb/librdft_scalar_r2cb.la \ rdft/scalar/r2r/librdft_scalar_r2r.la \ reodft/libreodft.la \ api/libapi.la \ $(SIMD_LIBS) $(SSE2_LIBS) $(AVX_LIBS) $(ALTIVEC_LIBS) $(NEON_LIBS) \ $(COMBINED_THREADLIBS) @QUAD_FALSE@libfftw3@PREC_SUFFIX@_la_LDFLAGS = -no-undefined -version-info \ @QUAD_FALSE@@SHARED_VERSION_INFO@ # cannot use -no-undefined since dependent on libquadmath @QUAD_TRUE@libfftw3@PREC_SUFFIX@_la_LDFLAGS = -version-info @SHARED_VERSION_INFO@ pkgconfigdir = $(libdir)/pkgconfig pkgconfig_DATA = fftw3@PREC_SUFFIX@.pc WISDOM_DIR = /etc/fftw WISDOM = wisdom@PREC_SUFFIX@ WISDOM_TIME = 12 # default to 12-hour limit, i.e. overnight WISDOM_FLAGS = --verbose --canonical --time-limit=$(WISDOM_TIME) all: config.h $(MAKE) $(AM_MAKEFLAGS) all-recursive .SUFFIXES: am--refresh: Makefile @: $(srcdir)/Makefile.in: @MAINTAINER_MODE_TRUE@ $(srcdir)/Makefile.am $(am__configure_deps) @for dep in $?; do \ case '$(am__configure_deps)' in \ *$$dep*) \ echo ' cd $(srcdir) && $(AUTOMAKE) --gnu'; \ $(am__cd) $(srcdir) && $(AUTOMAKE) --gnu \ && exit 0; \ exit 1;; \ esac; \ done; \ echo ' cd $(top_srcdir) && $(AUTOMAKE) --gnu Makefile'; \ $(am__cd) $(top_srcdir) && \ $(AUTOMAKE) --gnu Makefile .PRECIOUS: Makefile Makefile: $(srcdir)/Makefile.in $(top_builddir)/config.status @case '$?' in \ *config.status*) \ echo ' $(SHELL) ./config.status'; \ $(SHELL) ./config.status;; \ *) \ echo ' cd $(top_builddir) && $(SHELL) ./config.status $@ $(am__depfiles_maybe)'; \ cd $(top_builddir) && $(SHELL) ./config.status $@ $(am__depfiles_maybe);; \ esac; $(top_builddir)/config.status: $(top_srcdir)/configure $(CONFIG_STATUS_DEPENDENCIES) $(SHELL) ./config.status --recheck $(top_srcdir)/configure: @MAINTAINER_MODE_TRUE@ $(am__configure_deps) $(am__cd) $(srcdir) && $(AUTOCONF) $(ACLOCAL_M4): @MAINTAINER_MODE_TRUE@ $(am__aclocal_m4_deps) $(am__cd) $(srcdir) && $(ACLOCAL) $(ACLOCAL_AMFLAGS) $(am__aclocal_m4_deps): config.h: stamp-h1 @if test ! -f $@; then rm -f stamp-h1; else :; fi @if test ! -f $@; then $(MAKE) $(AM_MAKEFLAGS) stamp-h1; else :; fi stamp-h1: $(srcdir)/config.h.in $(top_builddir)/config.status @rm -f stamp-h1 cd $(top_builddir) && $(SHELL) ./config.status config.h $(srcdir)/config.h.in: @MAINTAINER_MODE_TRUE@ $(am__configure_deps) ($(am__cd) $(top_srcdir) && $(AUTOHEADER)) rm -f stamp-h1 touch $@ distclean-hdr: -rm -f config.h stamp-h1 fftw.pc: $(top_builddir)/config.status $(srcdir)/fftw.pc.in cd $(top_builddir) && $(SHELL) ./config.status $@ install-libLTLIBRARIES: $(lib_LTLIBRARIES) @$(NORMAL_INSTALL) @list='$(lib_LTLIBRARIES)'; test -n "$(libdir)" || list=; \ list2=; for p in $$list; do \ if test -f $$p; then \ list2="$$list2 $$p"; \ else :; fi; \ done; \ test -z "$$list2" || { \ echo " $(MKDIR_P) '$(DESTDIR)$(libdir)'"; \ $(MKDIR_P) "$(DESTDIR)$(libdir)" || exit 1; \ echo " $(LIBTOOL) $(AM_LIBTOOLFLAGS) $(LIBTOOLFLAGS) --mode=install $(INSTALL) $(INSTALL_STRIP_FLAG) $$list2 '$(DESTDIR)$(libdir)'"; \ $(LIBTOOL) $(AM_LIBTOOLFLAGS) $(LIBTOOLFLAGS) --mode=install $(INSTALL) $(INSTALL_STRIP_FLAG) $$list2 "$(DESTDIR)$(libdir)"; \ } uninstall-libLTLIBRARIES: @$(NORMAL_UNINSTALL) @list='$(lib_LTLIBRARIES)'; test -n "$(libdir)" || list=; \ for p in $$list; do \ $(am__strip_dir) \ echo " $(LIBTOOL) $(AM_LIBTOOLFLAGS) $(LIBTOOLFLAGS) --mode=uninstall rm -f '$(DESTDIR)$(libdir)/$$f'"; \ $(LIBTOOL) $(AM_LIBTOOLFLAGS) $(LIBTOOLFLAGS) --mode=uninstall rm -f "$(DESTDIR)$(libdir)/$$f"; \ done clean-libLTLIBRARIES: -test -z "$(lib_LTLIBRARIES)" || rm -f $(lib_LTLIBRARIES) @list='$(lib_LTLIBRARIES)'; for p in $$list; do \ dir="`echo $$p | sed -e 's|/[^/]*$$||'`"; \ test "$$dir" != "$$p" || dir=.; \ echo "rm -f \"$${dir}/so_locations\""; \ rm -f "$${dir}/so_locations"; \ done libfftw3@PREC_SUFFIX@.la: $(libfftw3@PREC_SUFFIX@_la_OBJECTS) $(libfftw3@PREC_SUFFIX@_la_DEPENDENCIES) $(EXTRA_libfftw3@PREC_SUFFIX@_la_DEPENDENCIES) $(libfftw3@PREC_SUFFIX@_la_LINK) -rpath $(libdir) $(libfftw3@PREC_SUFFIX@_la_OBJECTS) $(libfftw3@PREC_SUFFIX@_la_LIBADD) $(LIBS) mostlyclean-compile: -rm -f *.$(OBJEXT) distclean-compile: -rm -f *.tab.c mostlyclean-libtool: -rm -f *.lo clean-libtool: -rm -rf .libs _libs distclean-libtool: -rm -f libtool config.lt install-pkgconfigDATA: $(pkgconfig_DATA) @$(NORMAL_INSTALL) @list='$(pkgconfig_DATA)'; test -n "$(pkgconfigdir)" || list=; \ if test -n "$$list"; then \ echo " $(MKDIR_P) '$(DESTDIR)$(pkgconfigdir)'"; \ $(MKDIR_P) "$(DESTDIR)$(pkgconfigdir)" || exit 1; \ fi; \ for p in $$list; do \ if test -f "$$p"; then d=; else d="$(srcdir)/"; fi; \ echo "$$d$$p"; \ done | $(am__base_list) | \ while read files; do \ echo " $(INSTALL_DATA) $$files '$(DESTDIR)$(pkgconfigdir)'"; \ $(INSTALL_DATA) $$files "$(DESTDIR)$(pkgconfigdir)" || exit $$?; \ done uninstall-pkgconfigDATA: @$(NORMAL_UNINSTALL) @list='$(pkgconfig_DATA)'; test -n "$(pkgconfigdir)" || list=; \ files=`for p in $$list; do echo $$p; done | sed -e 's|^.*/||'`; \ dir='$(DESTDIR)$(pkgconfigdir)'; $(am__uninstall_files_from_dir) # This directory's subdirectories are mostly independent; you can cd # into them and run `make' without going through this Makefile. # To change the values of `make' variables: instead of editing Makefiles, # (1) if the variable is set in `config.status', edit `config.status' # (which will cause the Makefiles to be regenerated when you run `make'); # (2) otherwise, pass the desired values on the `make' command line. $(RECURSIVE_TARGETS): @fail= failcom='exit 1'; \ for f in x $$MAKEFLAGS; do \ case $$f in \ *=* | --[!k]*);; \ *k*) failcom='fail=yes';; \ esac; \ done; \ dot_seen=no; \ target=`echo $@ | sed s/-recursive//`; \ list='$(SUBDIRS)'; for subdir in $$list; do \ echo "Making $$target in $$subdir"; \ if test "$$subdir" = "."; then \ dot_seen=yes; \ local_target="$$target-am"; \ else \ local_target="$$target"; \ fi; \ ($(am__cd) $$subdir && $(MAKE) $(AM_MAKEFLAGS) $$local_target) \ || eval $$failcom; \ done; \ if test "$$dot_seen" = "no"; then \ $(MAKE) $(AM_MAKEFLAGS) "$$target-am" || exit 1; \ fi; test -z "$$fail" $(RECURSIVE_CLEAN_TARGETS): @fail= failcom='exit 1'; \ for f in x $$MAKEFLAGS; do \ case $$f in \ *=* | --[!k]*);; \ *k*) failcom='fail=yes';; \ esac; \ done; \ dot_seen=no; \ case "$@" in \ distclean-* | maintainer-clean-*) list='$(DIST_SUBDIRS)' ;; \ *) list='$(SUBDIRS)' ;; \ esac; \ rev=''; for subdir in $$list; do \ if test "$$subdir" = "."; then :; else \ rev="$$subdir $$rev"; \ fi; \ done; \ rev="$$rev ."; \ target=`echo $@ | sed s/-recursive//`; \ for subdir in $$rev; do \ echo "Making $$target in $$subdir"; \ if test "$$subdir" = "."; then \ local_target="$$target-am"; \ else \ local_target="$$target"; \ fi; \ ($(am__cd) $$subdir && $(MAKE) $(AM_MAKEFLAGS) $$local_target) \ || eval $$failcom; \ done && test -z "$$fail" tags-recursive: list='$(SUBDIRS)'; for subdir in $$list; do \ test "$$subdir" = . || ($(am__cd) $$subdir && $(MAKE) $(AM_MAKEFLAGS) tags); \ done ctags-recursive: list='$(SUBDIRS)'; for subdir in $$list; do \ test "$$subdir" = . || ($(am__cd) $$subdir && $(MAKE) $(AM_MAKEFLAGS) ctags); \ done ID: $(HEADERS) $(SOURCES) $(LISP) $(TAGS_FILES) list='$(SOURCES) $(HEADERS) $(LISP) $(TAGS_FILES)'; \ unique=`for i in $$list; do \ if test -f "$$i"; then echo $$i; else echo $(srcdir)/$$i; fi; \ done | \ $(AWK) '{ files[$$0] = 1; nonempty = 1; } \ END { if (nonempty) { for (i in files) print i; }; }'`; \ mkid -fID $$unique tags: TAGS TAGS: tags-recursive $(HEADERS) $(SOURCES) config.h.in $(TAGS_DEPENDENCIES) \ $(TAGS_FILES) $(LISP) set x; \ here=`pwd`; \ if ($(ETAGS) --etags-include --version) >/dev/null 2>&1; then \ include_option=--etags-include; \ empty_fix=.; \ else \ include_option=--include; \ empty_fix=; \ fi; \ list='$(SUBDIRS)'; for subdir in $$list; do \ if test "$$subdir" = .; then :; else \ test ! -f $$subdir/TAGS || \ set "$$@" "$$include_option=$$here/$$subdir/TAGS"; \ fi; \ done; \ list='$(SOURCES) $(HEADERS) config.h.in $(LISP) $(TAGS_FILES)'; \ unique=`for i in $$list; do \ if test -f "$$i"; then echo $$i; else echo $(srcdir)/$$i; fi; \ done | \ $(AWK) '{ files[$$0] = 1; nonempty = 1; } \ END { if (nonempty) { for (i in files) print i; }; }'`; \ shift; \ if test -z "$(ETAGS_ARGS)$$*$$unique"; then :; else \ test -n "$$unique" || unique=$$empty_fix; \ if test $$# -gt 0; then \ $(ETAGS) $(ETAGSFLAGS) $(AM_ETAGSFLAGS) $(ETAGS_ARGS) \ "$$@" $$unique; \ else \ $(ETAGS) $(ETAGSFLAGS) $(AM_ETAGSFLAGS) $(ETAGS_ARGS) \ $$unique; \ fi; \ fi ctags: CTAGS CTAGS: ctags-recursive $(HEADERS) $(SOURCES) config.h.in $(TAGS_DEPENDENCIES) \ $(TAGS_FILES) $(LISP) list='$(SOURCES) $(HEADERS) config.h.in $(LISP) $(TAGS_FILES)'; \ unique=`for i in $$list; do \ if test -f "$$i"; then echo $$i; else echo $(srcdir)/$$i; fi; \ done | \ $(AWK) '{ files[$$0] = 1; nonempty = 1; } \ END { if (nonempty) { for (i in files) print i; }; }'`; \ test -z "$(CTAGS_ARGS)$$unique" \ || $(CTAGS) $(CTAGSFLAGS) $(AM_CTAGSFLAGS) $(CTAGS_ARGS) \ $$unique GTAGS: here=`$(am__cd) $(top_builddir) && pwd` \ && $(am__cd) $(top_srcdir) \ && gtags -i $(GTAGS_ARGS) "$$here" distclean-tags: -rm -f TAGS ID GTAGS GRTAGS GSYMS GPATH tags distdir: $(DISTFILES) $(am__remove_distdir) test -d "$(distdir)" || mkdir "$(distdir)" @srcdirstrip=`echo "$(srcdir)" | sed 's/[].[^$$\\*]/\\\\&/g'`; \ topsrcdirstrip=`echo "$(top_srcdir)" | sed 's/[].[^$$\\*]/\\\\&/g'`; \ list='$(DISTFILES)'; \ dist_files=`for file in $$list; do echo $$file; done | \ sed -e "s|^$$srcdirstrip/||;t" \ -e "s|^$$topsrcdirstrip/|$(top_builddir)/|;t"`; \ case $$dist_files in \ */*) $(MKDIR_P) `echo "$$dist_files" | \ sed '/\//!d;s|^|$(distdir)/|;s,/[^/]*$$,,' | \ sort -u` ;; \ esac; \ for file in $$dist_files; do \ if test -f $$file || test -d $$file; then d=.; else d=$(srcdir); fi; \ if test -d $$d/$$file; then \ dir=`echo "/$$file" | sed -e 's,/[^/]*$$,,'`; \ if test -d "$(distdir)/$$file"; then \ find "$(distdir)/$$file" -type d ! -perm -700 -exec chmod u+rwx {} \;; \ fi; \ if test -d $(srcdir)/$$file && test $$d != $(srcdir); then \ cp -fpR $(srcdir)/$$file "$(distdir)$$dir" || exit 1; \ find "$(distdir)/$$file" -type d ! -perm -700 -exec chmod u+rwx {} \;; \ fi; \ cp -fpR $$d/$$file "$(distdir)$$dir" || exit 1; \ else \ test -f "$(distdir)/$$file" \ || cp -p $$d/$$file "$(distdir)/$$file" \ || exit 1; \ fi; \ done @list='$(DIST_SUBDIRS)'; for subdir in $$list; do \ if test "$$subdir" = .; then :; else \ $(am__make_dryrun) \ || test -d "$(distdir)/$$subdir" \ || $(MKDIR_P) "$(distdir)/$$subdir" \ || exit 1; \ dir1=$$subdir; dir2="$(distdir)/$$subdir"; \ $(am__relativize); \ new_distdir=$$reldir; \ dir1=$$subdir; dir2="$(top_distdir)"; \ $(am__relativize); \ new_top_distdir=$$reldir; \ echo " (cd $$subdir && $(MAKE) $(AM_MAKEFLAGS) top_distdir="$$new_top_distdir" distdir="$$new_distdir" \\"; \ echo " am__remove_distdir=: am__skip_length_check=: am__skip_mode_fix=: distdir)"; \ ($(am__cd) $$subdir && \ $(MAKE) $(AM_MAKEFLAGS) \ top_distdir="$$new_top_distdir" \ distdir="$$new_distdir" \ am__remove_distdir=: \ am__skip_length_check=: \ am__skip_mode_fix=: \ distdir) \ || exit 1; \ fi; \ done -test -n "$(am__skip_mode_fix)" \ || find "$(distdir)" -type d ! -perm -755 \ -exec chmod u+rwx,go+rx {} \; -o \ ! -type d ! -perm -444 -links 1 -exec chmod a+r {} \; -o \ ! -type d ! -perm -400 -exec chmod a+r {} \; -o \ ! -type d ! -perm -444 -exec $(install_sh) -c -m a+r {} {} \; \ || chmod -R a+r "$(distdir)" dist-gzip: distdir tardir=$(distdir) && $(am__tar) | GZIP=$(GZIP_ENV) gzip -c >$(distdir).tar.gz $(am__remove_distdir) dist-bzip2: distdir tardir=$(distdir) && $(am__tar) | BZIP2=$${BZIP2--9} bzip2 -c >$(distdir).tar.bz2 $(am__remove_distdir) dist-lzip: distdir tardir=$(distdir) && $(am__tar) | lzip -c $${LZIP_OPT--9} >$(distdir).tar.lz $(am__remove_distdir) dist-lzma: distdir tardir=$(distdir) && $(am__tar) | lzma -9 -c >$(distdir).tar.lzma $(am__remove_distdir) dist-xz: distdir tardir=$(distdir) && $(am__tar) | XZ_OPT=$${XZ_OPT--e} xz -c >$(distdir).tar.xz $(am__remove_distdir) dist-tarZ: distdir tardir=$(distdir) && $(am__tar) | compress -c >$(distdir).tar.Z $(am__remove_distdir) dist-shar: distdir shar $(distdir) | GZIP=$(GZIP_ENV) gzip -c >$(distdir).shar.gz $(am__remove_distdir) dist-zip: distdir -rm -f $(distdir).zip zip -rq $(distdir).zip $(distdir) $(am__remove_distdir) dist dist-all: distdir tardir=$(distdir) && $(am__tar) | GZIP=$(GZIP_ENV) gzip -c >$(distdir).tar.gz $(am__remove_distdir) # This target untars the dist file and tries a VPATH configuration. Then # it guarantees that the distribution is self-contained by making another # tarfile. distcheck: dist case '$(DIST_ARCHIVES)' in \ *.tar.gz*) \ GZIP=$(GZIP_ENV) gzip -dc $(distdir).tar.gz | $(am__untar) ;;\ *.tar.bz2*) \ bzip2 -dc $(distdir).tar.bz2 | $(am__untar) ;;\ *.tar.lzma*) \ lzma -dc $(distdir).tar.lzma | $(am__untar) ;;\ *.tar.lz*) \ lzip -dc $(distdir).tar.lz | $(am__untar) ;;\ *.tar.xz*) \ xz -dc $(distdir).tar.xz | $(am__untar) ;;\ *.tar.Z*) \ uncompress -c $(distdir).tar.Z | $(am__untar) ;;\ *.shar.gz*) \ GZIP=$(GZIP_ENV) gzip -dc $(distdir).shar.gz | unshar ;;\ *.zip*) \ unzip $(distdir).zip ;;\ esac chmod -R a-w $(distdir); chmod u+w $(distdir) mkdir $(distdir)/_build mkdir $(distdir)/_inst chmod a-w $(distdir) test -d $(distdir)/_build || exit 0; \ dc_install_base=`$(am__cd) $(distdir)/_inst && pwd | sed -e 's,^[^:\\/]:[\\/],/,'` \ && dc_destdir="$${TMPDIR-/tmp}/am-dc-$$$$/" \ && am__cwd=`pwd` \ && $(am__cd) $(distdir)/_build \ && ../configure --srcdir=.. --prefix="$$dc_install_base" \ $(AM_DISTCHECK_CONFIGURE_FLAGS) \ $(DISTCHECK_CONFIGURE_FLAGS) \ && $(MAKE) $(AM_MAKEFLAGS) \ && $(MAKE) $(AM_MAKEFLAGS) dvi \ && $(MAKE) $(AM_MAKEFLAGS) check \ && $(MAKE) $(AM_MAKEFLAGS) install \ && $(MAKE) $(AM_MAKEFLAGS) installcheck \ && $(MAKE) $(AM_MAKEFLAGS) uninstall \ && $(MAKE) $(AM_MAKEFLAGS) distuninstallcheck_dir="$$dc_install_base" \ distuninstallcheck \ && chmod -R a-w "$$dc_install_base" \ && ({ \ (cd ../.. && umask 077 && mkdir "$$dc_destdir") \ && $(MAKE) $(AM_MAKEFLAGS) DESTDIR="$$dc_destdir" install \ && $(MAKE) $(AM_MAKEFLAGS) DESTDIR="$$dc_destdir" uninstall \ && $(MAKE) $(AM_MAKEFLAGS) DESTDIR="$$dc_destdir" \ distuninstallcheck_dir="$$dc_destdir" distuninstallcheck; \ } || { rm -rf "$$dc_destdir"; exit 1; }) \ && rm -rf "$$dc_destdir" \ && $(MAKE) $(AM_MAKEFLAGS) dist \ && rm -rf $(DIST_ARCHIVES) \ && $(MAKE) $(AM_MAKEFLAGS) distcleancheck \ && cd "$$am__cwd" \ || exit 1 $(am__remove_distdir) @(echo "$(distdir) archives ready for distribution: "; \ list='$(DIST_ARCHIVES)'; for i in $$list; do echo $$i; done) | \ sed -e 1h -e 1s/./=/g -e 1p -e 1x -e '$$p' -e '$$x' distuninstallcheck: @test -n '$(distuninstallcheck_dir)' || { \ echo 'ERROR: trying to run $@ with an empty' \ '$$(distuninstallcheck_dir)' >&2; \ exit 1; \ }; \ $(am__cd) '$(distuninstallcheck_dir)' || { \ echo 'ERROR: cannot chdir into $(distuninstallcheck_dir)' >&2; \ exit 1; \ }; \ test `$(am__distuninstallcheck_listfiles) | wc -l` -eq 0 \ || { echo "ERROR: files left after uninstall:" ; \ if test -n "$(DESTDIR)"; then \ echo " (check DESTDIR support)"; \ fi ; \ $(distuninstallcheck_listfiles) ; \ exit 1; } >&2 distcleancheck: distclean @if test '$(srcdir)' = . ; then \ echo "ERROR: distcleancheck can only run from a VPATH build" ; \ exit 1 ; \ fi @test `$(distcleancheck_listfiles) | wc -l` -eq 0 \ || { echo "ERROR: files left in build directory after distclean:" ; \ $(distcleancheck_listfiles) ; \ exit 1; } >&2 check-am: all-am check: check-recursive all-am: Makefile $(LTLIBRARIES) $(DATA) config.h installdirs: installdirs-recursive installdirs-am: for dir in "$(DESTDIR)$(libdir)" "$(DESTDIR)$(pkgconfigdir)"; do \ test -z "$$dir" || $(MKDIR_P) "$$dir"; \ done install: install-recursive install-exec: install-exec-recursive install-data: install-data-recursive uninstall: uninstall-recursive install-am: all-am @$(MAKE) $(AM_MAKEFLAGS) install-exec-am install-data-am installcheck: installcheck-recursive install-strip: if test -z '$(STRIP)'; then \ $(MAKE) $(AM_MAKEFLAGS) INSTALL_PROGRAM="$(INSTALL_STRIP_PROGRAM)" \ install_sh_PROGRAM="$(INSTALL_STRIP_PROGRAM)" INSTALL_STRIP_FLAG=-s \ install; \ else \ $(MAKE) $(AM_MAKEFLAGS) INSTALL_PROGRAM="$(INSTALL_STRIP_PROGRAM)" \ install_sh_PROGRAM="$(INSTALL_STRIP_PROGRAM)" INSTALL_STRIP_FLAG=-s \ "INSTALL_PROGRAM_ENV=STRIPPROG='$(STRIP)'" install; \ fi mostlyclean-generic: clean-generic: distclean-generic: -test -z "$(CONFIG_CLEAN_FILES)" || rm -f $(CONFIG_CLEAN_FILES) -test . = "$(srcdir)" || test -z "$(CONFIG_CLEAN_VPATH_FILES)" || rm -f $(CONFIG_CLEAN_VPATH_FILES) maintainer-clean-generic: @echo "This command is intended for maintainers to use" @echo "it deletes files that may require special tools to rebuild." clean: clean-recursive clean-am: clean-generic clean-libLTLIBRARIES clean-libtool \ mostlyclean-am distclean: distclean-recursive -rm -f $(am__CONFIG_DISTCLEAN_FILES) -rm -f Makefile distclean-am: clean-am distclean-compile distclean-generic \ distclean-hdr distclean-libtool distclean-tags dvi: dvi-recursive dvi-am: html: html-recursive html-am: info: info-recursive info-am: install-data-am: install-pkgconfigDATA install-dvi: install-dvi-recursive install-dvi-am: install-exec-am: install-libLTLIBRARIES install-html: install-html-recursive install-html-am: install-info: install-info-recursive install-info-am: install-man: install-pdf: install-pdf-recursive install-pdf-am: install-ps: install-ps-recursive install-ps-am: installcheck-am: maintainer-clean: maintainer-clean-recursive -rm -f $(am__CONFIG_DISTCLEAN_FILES) -rm -rf $(top_srcdir)/autom4te.cache -rm -f Makefile maintainer-clean-am: distclean-am maintainer-clean-generic mostlyclean: mostlyclean-recursive mostlyclean-am: mostlyclean-compile mostlyclean-generic \ mostlyclean-libtool pdf: pdf-recursive pdf-am: ps: ps-recursive ps-am: uninstall-am: uninstall-libLTLIBRARIES uninstall-pkgconfigDATA .MAKE: $(RECURSIVE_CLEAN_TARGETS) $(RECURSIVE_TARGETS) all \ ctags-recursive install-am install-strip tags-recursive .PHONY: $(RECURSIVE_CLEAN_TARGETS) $(RECURSIVE_TARGETS) CTAGS GTAGS \ all all-am am--refresh check check-am clean clean-generic \ clean-libLTLIBRARIES clean-libtool ctags ctags-recursive dist \ dist-all dist-bzip2 dist-gzip dist-lzip dist-lzma dist-shar \ dist-tarZ dist-xz dist-zip distcheck distclean \ distclean-compile distclean-generic distclean-hdr \ distclean-libtool distclean-tags distcleancheck distdir \ distuninstallcheck dvi dvi-am html html-am info info-am \ install install-am install-data install-data-am install-dvi \ install-dvi-am install-exec install-exec-am install-html \ install-html-am install-info install-info-am \ install-libLTLIBRARIES install-man install-pdf install-pdf-am \ install-pkgconfigDATA install-ps install-ps-am install-strip \ installcheck installcheck-am installdirs installdirs-am \ maintainer-clean maintainer-clean-generic mostlyclean \ mostlyclean-compile mostlyclean-generic mostlyclean-libtool \ pdf pdf-am ps ps-am tags tags-recursive uninstall uninstall-am \ uninstall-libLTLIBRARIES uninstall-pkgconfigDATA fftw3@PREC_SUFFIX@.pc: fftw.pc cp -f fftw.pc fftw3@PREC_SUFFIX@.pc wisdom: tools/fftw@PREC_SUFFIX@-wisdom -o $@ $(WISDOM_FLAGS) install-wisdom: wisdom $(mkinstalldirs) $(WISDOM_DIR) $(INSTALL_DATA) wisdom $(WISDOM_DIR)/$(WISDOM) # Tell versions [3.59,3.63) of GNU make to not export all variables. # Otherwise a system limit (for SysV at least) may be exceeded. .NOEXPORT: fftw-3.3.3/config.h.in0000644000175000017500000002503712054410177011442 00000000000000/* config.h.in. Generated from configure.ac by autoheader. */ /* Define to compile in long-double precision. */ #undef BENCHFFT_LDOUBLE /* Define to compile in quad precision. */ #undef BENCHFFT_QUAD /* Define to compile in single precision. */ #undef BENCHFFT_SINGLE /* Define to one of `_getb67', `GETB67', `getb67' for Cray-2 and Cray-YMP systems. This function is required for `alloca.c' support on those systems. */ #undef CRAY_STACKSEG_END /* Define to 1 if using `alloca.c'. */ #undef C_ALLOCA /* Define to disable Fortran wrappers. */ #undef DISABLE_FORTRAN /* Define to dummy `main' function (if any) required to link to the Fortran libraries. */ #undef F77_DUMMY_MAIN /* Define to a macro mangling the given C identifier (in lower and upper case), which must not contain underscores, for linking with Fortran. */ #undef F77_FUNC /* As F77_FUNC, but for C identifiers containing underscores. */ #undef F77_FUNC_ /* Define if F77_FUNC and F77_FUNC_ are equivalent. */ #undef F77_FUNC_EQUIV /* Define if F77 and FC dummy `main' functions are identical. */ #undef FC_DUMMY_MAIN_EQ_F77 /* C compiler name and flags */ #undef FFTW_CC /* Define to enable extra FFTW debugging code. */ #undef FFTW_DEBUG /* Define to enable alignment debugging hacks. */ #undef FFTW_DEBUG_ALIGNMENT /* Define to enable debugging malloc. */ #undef FFTW_DEBUG_MALLOC /* Define to enable the use of alloca(). */ #undef FFTW_ENABLE_ALLOCA /* Define to compile in long-double precision. */ #undef FFTW_LDOUBLE /* Define to compile in quad precision. */ #undef FFTW_QUAD /* Define to enable pseudorandom estimate planning for debugging. */ #undef FFTW_RANDOM_ESTIMATOR /* Define to compile in single precision. */ #undef FFTW_SINGLE /* Define to 1 if you have the `abort' function. */ #undef HAVE_ABORT /* Define to 1 if you have `alloca', as a function or macro. */ #undef HAVE_ALLOCA /* Define to 1 if you have and it should be used (not on Ultrix). */ #undef HAVE_ALLOCA_H /* Define to enable Altivec optimizations. */ #undef HAVE_ALTIVEC /* Define to 1 if you have the header file. */ #undef HAVE_ALTIVEC_H /* Define to enable AVX optimizations. */ #undef HAVE_AVX /* Define to 1 if you have the `BSDgettimeofday' function. */ #undef HAVE_BSDGETTIMEOFDAY /* Define to 1 if you have the `clock_gettime' function. */ #undef HAVE_CLOCK_GETTIME /* Define to 1 if you have the `cosl' function. */ #undef HAVE_COSL /* Define to 1 if you have the header file. */ #undef HAVE_C_ASM_H /* Define to 1 if you have the declaration of `cosl', and to 0 if you don't. */ #undef HAVE_DECL_COSL /* Define to 1 if you have the declaration of `cosq', and to 0 if you don't. */ #undef HAVE_DECL_COSQ /* Define to 1 if you have the declaration of `drand48', and to 0 if you don't. */ #undef HAVE_DECL_DRAND48 /* Define to 1 if you have the declaration of `memalign', and to 0 if you don't. */ #undef HAVE_DECL_MEMALIGN /* Define to 1 if you have the declaration of `posix_memalign', and to 0 if you don't. */ #undef HAVE_DECL_POSIX_MEMALIGN /* Define to 1 if you have the declaration of `sinl', and to 0 if you don't. */ #undef HAVE_DECL_SINL /* Define to 1 if you have the declaration of `sinq', and to 0 if you don't. */ #undef HAVE_DECL_SINQ /* Define to 1 if you have the declaration of `srand48', and to 0 if you don't. */ #undef HAVE_DECL_SRAND48 /* Define to 1 if you have the header file. */ #undef HAVE_DLFCN_H /* Define to 1 if you don't have `vprintf' but do have `_doprnt.' */ #undef HAVE_DOPRNT /* Define to 1 if you have the `drand48' function. */ #undef HAVE_DRAND48 /* Define if you have a machine with fused multiply-add */ #undef HAVE_FMA /* Define to 1 if you have the `gethrtime' function. */ #undef HAVE_GETHRTIME /* Define to 1 if you have the `gettimeofday' function. */ #undef HAVE_GETTIMEOFDAY /* Define to 1 if hrtime_t is defined in */ #undef HAVE_HRTIME_T /* Define to 1 if you have the header file. */ #undef HAVE_INTRINSICS_H /* Define to 1 if you have the header file. */ #undef HAVE_INTTYPES_H /* Define if the isnan() function/macro is available. */ #undef HAVE_ISNAN /* Define to 1 if you have the header file. */ #undef HAVE_LIBINTL_H /* Define to 1 if you have the `m' library (-lm). */ #undef HAVE_LIBM /* Define to 1 if you have the `quadmath' library (-lquadmath). */ #undef HAVE_LIBQUADMATH /* Define to 1 if you have the header file. */ #undef HAVE_LIMITS_H /* Define to 1 if the compiler supports `long double' */ #undef HAVE_LONG_DOUBLE /* Define to 1 if you have the `mach_absolute_time' function. */ #undef HAVE_MACH_ABSOLUTE_TIME /* Define to 1 if you have the header file. */ #undef HAVE_MACH_MACH_TIME_H /* Define to 1 if you have the header file. */ #undef HAVE_MALLOC_H /* Define to 1 if you have the `memalign' function. */ #undef HAVE_MEMALIGN /* Define to 1 if you have the header file. */ #undef HAVE_MEMORY_H /* Define to 1 if you have the `memset' function. */ #undef HAVE_MEMSET /* Define to enable use of MIPS ZBus cycle-counter. */ #undef HAVE_MIPS_ZBUS_TIMER /* Define if you have the MPI library. */ #undef HAVE_MPI /* Define to enable ARM NEON optimizations. */ #undef HAVE_NEON /* Define if OpenMP is enabled */ #undef HAVE_OPENMP /* Define to 1 if you have the `posix_memalign' function. */ #undef HAVE_POSIX_MEMALIGN /* Define if you have POSIX threads libraries and header files. */ #undef HAVE_PTHREAD /* Define to 1 if you have the `read_real_time' function. */ #undef HAVE_READ_REAL_TIME /* Define to 1 if you have the `sinl' function. */ #undef HAVE_SINL /* Define to 1 if you have the `snprintf' function. */ #undef HAVE_SNPRINTF /* Define to 1 if you have the `sqrt' function. */ #undef HAVE_SQRT /* Define to enable SSE/SSE2 optimizations. */ #undef HAVE_SSE2 /* Define to 1 if you have the header file. */ #undef HAVE_STDDEF_H /* Define to 1 if you have the header file. */ #undef HAVE_STDINT_H /* Define to 1 if you have the header file. */ #undef HAVE_STDLIB_H /* Define to 1 if you have the header file. */ #undef HAVE_STRINGS_H /* Define to 1 if you have the header file. */ #undef HAVE_STRING_H /* Define to 1 if you have the `sysctl' function. */ #undef HAVE_SYSCTL /* Define to 1 if you have the header file. */ #undef HAVE_SYS_STAT_H /* Define to 1 if you have the header file. */ #undef HAVE_SYS_SYSCTL_H /* Define to 1 if you have the header file. */ #undef HAVE_SYS_TIME_H /* Define to 1 if you have the header file. */ #undef HAVE_SYS_TYPES_H /* Define to 1 if you have the `tanl' function. */ #undef HAVE_TANL /* Define if we have a threads library. */ #undef HAVE_THREADS /* Define to 1 if you have the `time_base_to_time' function. */ #undef HAVE_TIME_BASE_TO_TIME /* Define to 1 if the system has the type `uintptr_t'. */ #undef HAVE_UINTPTR_T /* Define to 1 if you have the header file. */ #undef HAVE_UNISTD_H /* Define to 1 if you have the `vprintf' function. */ #undef HAVE_VPRINTF /* Define to 1 if you have the `_mm_free' function. */ #undef HAVE__MM_FREE /* Define to 1 if you have the `_mm_malloc' function. */ #undef HAVE__MM_MALLOC /* Define if you have the UNICOS _rtc() intrinsic. */ #undef HAVE__RTC /* Define to the sub-directory in which libtool stores uninstalled libraries. */ #undef LT_OBJDIR /* Define to 1 if your C compiler doesn't accept -c and -o together. */ #undef NO_MINUS_C_MINUS_O /* Name of package */ #undef PACKAGE /* Define to the address where bug reports for this package should be sent. */ #undef PACKAGE_BUGREPORT /* Define to the full name of this package. */ #undef PACKAGE_NAME /* Define to the full name and version of this package. */ #undef PACKAGE_STRING /* Define to the one symbol short name of this package. */ #undef PACKAGE_TARNAME /* Define to the home page for this package. */ #undef PACKAGE_URL /* Define to the version of this package. */ #undef PACKAGE_VERSION /* Define to necessary symbol if this constant uses a non-standard name on your system. */ #undef PTHREAD_CREATE_JOINABLE /* The size of `double', as computed by sizeof. */ #undef SIZEOF_DOUBLE /* The size of `fftw_r2r_kind', as computed by sizeof. */ #undef SIZEOF_FFTW_R2R_KIND /* The size of `float', as computed by sizeof. */ #undef SIZEOF_FLOAT /* The size of `int', as computed by sizeof. */ #undef SIZEOF_INT /* The size of `long', as computed by sizeof. */ #undef SIZEOF_LONG /* The size of `long long', as computed by sizeof. */ #undef SIZEOF_LONG_LONG /* The size of `MPI_Fint', as computed by sizeof. */ #undef SIZEOF_MPI_FINT /* The size of `ptrdiff_t', as computed by sizeof. */ #undef SIZEOF_PTRDIFF_T /* The size of `size_t', as computed by sizeof. */ #undef SIZEOF_SIZE_T /* The size of `unsigned int', as computed by sizeof. */ #undef SIZEOF_UNSIGNED_INT /* The size of `unsigned long', as computed by sizeof. */ #undef SIZEOF_UNSIGNED_LONG /* The size of `unsigned long long', as computed by sizeof. */ #undef SIZEOF_UNSIGNED_LONG_LONG /* The size of `void *', as computed by sizeof. */ #undef SIZEOF_VOID_P /* If using the C implementation of alloca, define if you know the direction of stack growth for your system; otherwise it will be automatically deduced at runtime. STACK_DIRECTION > 0 => grows toward higher addresses STACK_DIRECTION < 0 => grows toward lower addresses STACK_DIRECTION = 0 => direction of growth unknown */ #undef STACK_DIRECTION /* Define to 1 if you have the ANSI C header files. */ #undef STDC_HEADERS /* Define to 1 if you can safely include both and . */ #undef TIME_WITH_SYS_TIME /* Define if we have and are using POSIX threads. */ #undef USING_POSIX_THREADS /* Version number of package */ #undef VERSION /* Use common Windows Fortran mangling styles for the Fortran interfaces. */ #undef WINDOWS_F77_MANGLING /* Include g77-compatible wrappers in addition to any other Fortran wrappers. */ #undef WITH_G77_WRAPPERS /* Use our own aligned malloc routine; mainly helpful for Windows systems lacking aligned allocation system-library routines. */ #undef WITH_OUR_MALLOC /* Use low-precision timers, making planner very slow */ #undef WITH_SLOW_TIMER /* Define to empty if `const' does not conform to ANSI C. */ #undef const /* Define to `__inline__' or `__inline' if that's what the C compiler calls it, or to nothing if 'inline' is not supported under any name. */ #ifndef __cplusplus #undef inline #endif /* Define to `unsigned int' if does not define. */ #undef size_t fftw-3.3.3/fftw.pc.in0000644000175000017500000000035612054410050011301 00000000000000prefix=@prefix@ exec_prefix=@exec_prefix@ libdir=@libdir@ includedir=@includedir@ Name: FFTW Description: fast Fourier transform library Version: @VERSION@ Libs: -L${libdir} -lfftw3@PREC_SUFFIX@ @LIBQUADMATH@ -lm Cflags: -I${includedir} fftw-3.3.3/configure0000755000175000017500000246255212054410176011336 00000000000000#! /bin/sh # Guess values for system-dependent variables and create Makefiles. # Generated by GNU Autoconf 2.69 for fftw 3.3.3. # # Report bugs to . # # # Copyright (C) 1992-1996, 1998-2012 Free Software Foundation, Inc. # # # This configure script is free software; the Free Software Foundation # gives unlimited permission to copy, distribute and modify it. ## -------------------- ## ## M4sh Initialization. ## ## -------------------- ## # Be more Bourne compatible DUALCASE=1; export DUALCASE # for MKS sh if test -n "${ZSH_VERSION+set}" && (emulate sh) >/dev/null 2>&1; then : emulate sh NULLCMD=: # Pre-4.2 versions of Zsh do word splitting on ${1+"$@"}, which # is contrary to our usage. Disable this feature. alias -g '${1+"$@"}'='"$@"' setopt NO_GLOB_SUBST else case `(set -o) 2>/dev/null` in #( *posix*) : set -o posix ;; #( *) : ;; esac fi as_nl=' ' export as_nl # Printing a long string crashes Solaris 7 /usr/bin/printf. as_echo='\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\' as_echo=$as_echo$as_echo$as_echo$as_echo$as_echo as_echo=$as_echo$as_echo$as_echo$as_echo$as_echo$as_echo # Prefer a ksh shell builtin over an external printf program on Solaris, # but without wasting forks for bash or zsh. if test -z "$BASH_VERSION$ZSH_VERSION" \ && (test "X`print -r -- $as_echo`" = "X$as_echo") 2>/dev/null; then as_echo='print -r --' as_echo_n='print -rn --' elif (test "X`printf %s $as_echo`" = "X$as_echo") 2>/dev/null; then as_echo='printf %s\n' as_echo_n='printf %s' else if test "X`(/usr/ucb/echo -n -n $as_echo) 2>/dev/null`" = "X-n $as_echo"; then as_echo_body='eval /usr/ucb/echo -n "$1$as_nl"' as_echo_n='/usr/ucb/echo -n' else as_echo_body='eval expr "X$1" : "X\\(.*\\)"' as_echo_n_body='eval arg=$1; case $arg in #( *"$as_nl"*) expr "X$arg" : "X\\(.*\\)$as_nl"; arg=`expr "X$arg" : ".*$as_nl\\(.*\\)"`;; esac; expr "X$arg" : "X\\(.*\\)" | tr -d "$as_nl" ' export as_echo_n_body as_echo_n='sh -c $as_echo_n_body as_echo' fi export as_echo_body as_echo='sh -c $as_echo_body as_echo' fi # The user is always right. if test "${PATH_SEPARATOR+set}" != set; then PATH_SEPARATOR=: (PATH='/bin;/bin'; FPATH=$PATH; sh -c :) >/dev/null 2>&1 && { (PATH='/bin:/bin'; FPATH=$PATH; sh -c :) >/dev/null 2>&1 || PATH_SEPARATOR=';' } fi # IFS # We need space, tab and new line, in precisely that order. Quoting is # there to prevent editors from complaining about space-tab. # (If _AS_PATH_WALK were called with IFS unset, it would disable word # splitting by setting IFS to empty value.) IFS=" "" $as_nl" # Find who we are. Look in the path if we contain no directory separator. as_myself= case $0 in #(( *[\\/]* ) as_myself=$0 ;; *) as_save_IFS=$IFS; IFS=$PATH_SEPARATOR for as_dir in $PATH do IFS=$as_save_IFS test -z "$as_dir" && as_dir=. test -r "$as_dir/$0" && as_myself=$as_dir/$0 && break done IFS=$as_save_IFS ;; esac # We did not find ourselves, most probably we were run as `sh COMMAND' # in which case we are not to be found in the path. if test "x$as_myself" = x; then as_myself=$0 fi if test ! -f "$as_myself"; then $as_echo "$as_myself: error: cannot find myself; rerun with an absolute file name" >&2 exit 1 fi # Unset variables that we do not need and which cause bugs (e.g. in # pre-3.0 UWIN ksh). But do not cause bugs in bash 2.01; the "|| exit 1" # suppresses any "Segmentation fault" message there. '((' could # trigger a bug in pdksh 5.2.14. for as_var in BASH_ENV ENV MAIL MAILPATH do eval test x\${$as_var+set} = xset \ && ( (unset $as_var) || exit 1) >/dev/null 2>&1 && unset $as_var || : done PS1='$ ' PS2='> ' PS4='+ ' # NLS nuisances. LC_ALL=C export LC_ALL LANGUAGE=C export LANGUAGE # CDPATH. (unset CDPATH) >/dev/null 2>&1 && unset CDPATH # Use a proper internal environment variable to ensure we don't fall # into an infinite loop, continuously re-executing ourselves. if test x"${_as_can_reexec}" != xno && test "x$CONFIG_SHELL" != x; then _as_can_reexec=no; export _as_can_reexec; # We cannot yet assume a decent shell, so we have to provide a # neutralization value for shells without unset; and this also # works around shells that cannot unset nonexistent variables. # Preserve -v and -x to the replacement shell. BASH_ENV=/dev/null ENV=/dev/null (unset BASH_ENV) >/dev/null 2>&1 && unset BASH_ENV ENV case $- in # (((( *v*x* | *x*v* ) as_opts=-vx ;; *v* ) as_opts=-v ;; *x* ) as_opts=-x ;; * ) as_opts= ;; esac exec $CONFIG_SHELL $as_opts "$as_myself" ${1+"$@"} # Admittedly, this is quite paranoid, since all the known shells bail # out after a failed `exec'. $as_echo "$0: could not re-execute with $CONFIG_SHELL" >&2 as_fn_exit 255 fi # We don't want this to propagate to other subprocesses. { _as_can_reexec=; unset _as_can_reexec;} if test "x$CONFIG_SHELL" = x; then as_bourne_compatible="if test -n \"\${ZSH_VERSION+set}\" && (emulate sh) >/dev/null 2>&1; then : emulate sh NULLCMD=: # Pre-4.2 versions of Zsh do word splitting on \${1+\"\$@\"}, which # is contrary to our usage. Disable this feature. alias -g '\${1+\"\$@\"}'='\"\$@\"' setopt NO_GLOB_SUBST else case \`(set -o) 2>/dev/null\` in #( *posix*) : set -o posix ;; #( *) : ;; esac fi " as_required="as_fn_return () { (exit \$1); } as_fn_success () { as_fn_return 0; } as_fn_failure () { as_fn_return 1; } as_fn_ret_success () { return 0; } as_fn_ret_failure () { return 1; } exitcode=0 as_fn_success || { exitcode=1; echo as_fn_success failed.; } as_fn_failure && { exitcode=1; echo as_fn_failure succeeded.; } as_fn_ret_success || { exitcode=1; echo as_fn_ret_success failed.; } as_fn_ret_failure && { exitcode=1; echo as_fn_ret_failure succeeded.; } if ( set x; as_fn_ret_success y && test x = \"\$1\" ); then : else exitcode=1; echo positional parameters were not saved. fi test x\$exitcode = x0 || exit 1 test -x / || exit 1" as_suggested=" as_lineno_1=";as_suggested=$as_suggested$LINENO;as_suggested=$as_suggested" as_lineno_1a=\$LINENO as_lineno_2=";as_suggested=$as_suggested$LINENO;as_suggested=$as_suggested" as_lineno_2a=\$LINENO eval 'test \"x\$as_lineno_1'\$as_run'\" != \"x\$as_lineno_2'\$as_run'\" && test \"x\`expr \$as_lineno_1'\$as_run' + 1\`\" = \"x\$as_lineno_2'\$as_run'\"' || exit 1 test -n \"\${ZSH_VERSION+set}\${BASH_VERSION+set}\" || ( ECHO='\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\' ECHO=\$ECHO\$ECHO\$ECHO\$ECHO\$ECHO ECHO=\$ECHO\$ECHO\$ECHO\$ECHO\$ECHO\$ECHO PATH=/empty FPATH=/empty; export PATH FPATH test \"X\`printf %s \$ECHO\`\" = \"X\$ECHO\" \\ || test \"X\`print -r -- \$ECHO\`\" = \"X\$ECHO\" ) || exit 1 test \$(( 1 + 1 )) = 2 || exit 1" if (eval "$as_required") 2>/dev/null; then : as_have_required=yes else as_have_required=no fi if test x$as_have_required = xyes && (eval "$as_suggested") 2>/dev/null; then : else as_save_IFS=$IFS; IFS=$PATH_SEPARATOR as_found=false for as_dir in /bin$PATH_SEPARATOR/usr/bin$PATH_SEPARATOR$PATH do IFS=$as_save_IFS test -z "$as_dir" && as_dir=. as_found=: case $as_dir in #( /*) for as_base in sh bash ksh sh5; do # Try only shells that exist, to save several forks. as_shell=$as_dir/$as_base if { test -f "$as_shell" || test -f "$as_shell.exe"; } && { $as_echo "$as_bourne_compatible""$as_required" | as_run=a "$as_shell"; } 2>/dev/null; then : CONFIG_SHELL=$as_shell as_have_required=yes if { $as_echo "$as_bourne_compatible""$as_suggested" | as_run=a "$as_shell"; } 2>/dev/null; then : break 2 fi fi done;; esac as_found=false done $as_found || { if { test -f "$SHELL" || test -f "$SHELL.exe"; } && { $as_echo "$as_bourne_compatible""$as_required" | as_run=a "$SHELL"; } 2>/dev/null; then : CONFIG_SHELL=$SHELL as_have_required=yes fi; } IFS=$as_save_IFS if test "x$CONFIG_SHELL" != x; then : export CONFIG_SHELL # We cannot yet assume a decent shell, so we have to provide a # neutralization value for shells without unset; and this also # works around shells that cannot unset nonexistent variables. # Preserve -v and -x to the replacement shell. BASH_ENV=/dev/null ENV=/dev/null (unset BASH_ENV) >/dev/null 2>&1 && unset BASH_ENV ENV case $- in # (((( *v*x* | *x*v* ) as_opts=-vx ;; *v* ) as_opts=-v ;; *x* ) as_opts=-x ;; * ) as_opts= ;; esac exec $CONFIG_SHELL $as_opts "$as_myself" ${1+"$@"} # Admittedly, this is quite paranoid, since all the known shells bail # out after a failed `exec'. $as_echo "$0: could not re-execute with $CONFIG_SHELL" >&2 exit 255 fi if test x$as_have_required = xno; then : $as_echo "$0: This script requires a shell more modern than all" $as_echo "$0: the shells that I found on your system." if test x${ZSH_VERSION+set} = xset ; then $as_echo "$0: In particular, zsh $ZSH_VERSION has bugs and should" $as_echo "$0: be upgraded to zsh 4.3.4 or later." else $as_echo "$0: Please tell bug-autoconf@gnu.org and fftw@fftw.org $0: about your system, including any error possibly output $0: before this message. Then install a modern shell, or $0: manually run the script under such a shell if you do $0: have one." fi exit 1 fi fi fi SHELL=${CONFIG_SHELL-/bin/sh} export SHELL # Unset more variables known to interfere with behavior of common tools. CLICOLOR_FORCE= GREP_OPTIONS= unset CLICOLOR_FORCE GREP_OPTIONS ## --------------------- ## ## M4sh Shell Functions. ## ## --------------------- ## # as_fn_unset VAR # --------------- # Portably unset VAR. as_fn_unset () { { eval $1=; unset $1;} } as_unset=as_fn_unset # as_fn_set_status STATUS # ----------------------- # Set $? to STATUS, without forking. as_fn_set_status () { return $1 } # as_fn_set_status # as_fn_exit STATUS # ----------------- # Exit the shell with STATUS, even in a "trap 0" or "set -e" context. as_fn_exit () { set +e as_fn_set_status $1 exit $1 } # as_fn_exit # as_fn_mkdir_p # ------------- # Create "$as_dir" as a directory, including parents if necessary. as_fn_mkdir_p () { case $as_dir in #( -*) as_dir=./$as_dir;; esac test -d "$as_dir" || eval $as_mkdir_p || { as_dirs= while :; do case $as_dir in #( *\'*) as_qdir=`$as_echo "$as_dir" | sed "s/'/'\\\\\\\\''/g"`;; #'( *) as_qdir=$as_dir;; esac as_dirs="'$as_qdir' $as_dirs" as_dir=`$as_dirname -- "$as_dir" || $as_expr X"$as_dir" : 'X\(.*[^/]\)//*[^/][^/]*/*$' \| \ X"$as_dir" : 'X\(//\)[^/]' \| \ X"$as_dir" : 'X\(//\)$' \| \ X"$as_dir" : 'X\(/\)' \| . 2>/dev/null || $as_echo X"$as_dir" | sed '/^X\(.*[^/]\)\/\/*[^/][^/]*\/*$/{ s//\1/ q } /^X\(\/\/\)[^/].*/{ s//\1/ q } /^X\(\/\/\)$/{ s//\1/ q } /^X\(\/\).*/{ s//\1/ q } s/.*/./; q'` test -d "$as_dir" && break done test -z "$as_dirs" || eval "mkdir $as_dirs" } || test -d "$as_dir" || as_fn_error $? "cannot create directory $as_dir" } # as_fn_mkdir_p # as_fn_executable_p FILE # ----------------------- # Test if FILE is an executable regular file. as_fn_executable_p () { test -f "$1" && test -x "$1" } # as_fn_executable_p # as_fn_append VAR VALUE # ---------------------- # Append the text in VALUE to the end of the definition contained in VAR. Take # advantage of any shell optimizations that allow amortized linear growth over # repeated appends, instead of the typical quadratic growth present in naive # implementations. if (eval "as_var=1; as_var+=2; test x\$as_var = x12") 2>/dev/null; then : eval 'as_fn_append () { eval $1+=\$2 }' else as_fn_append () { eval $1=\$$1\$2 } fi # as_fn_append # as_fn_arith ARG... # ------------------ # Perform arithmetic evaluation on the ARGs, and store the result in the # global $as_val. Take advantage of shells that can avoid forks. The arguments # must be portable across $(()) and expr. if (eval "test \$(( 1 + 1 )) = 2") 2>/dev/null; then : eval 'as_fn_arith () { as_val=$(( $* )) }' else as_fn_arith () { as_val=`expr "$@" || test $? -eq 1` } fi # as_fn_arith # as_fn_error STATUS ERROR [LINENO LOG_FD] # ---------------------------------------- # Output "`basename $0`: error: ERROR" to stderr. If LINENO and LOG_FD are # provided, also output the error to LOG_FD, referencing LINENO. Then exit the # script with STATUS, using 1 if that was 0. as_fn_error () { as_status=$1; test $as_status -eq 0 && as_status=1 if test "$4"; then as_lineno=${as_lineno-"$3"} as_lineno_stack=as_lineno_stack=$as_lineno_stack $as_echo "$as_me:${as_lineno-$LINENO}: error: $2" >&$4 fi $as_echo "$as_me: error: $2" >&2 as_fn_exit $as_status } # as_fn_error if expr a : '\(a\)' >/dev/null 2>&1 && test "X`expr 00001 : '.*\(...\)'`" = X001; then as_expr=expr else as_expr=false fi if (basename -- /) >/dev/null 2>&1 && test "X`basename -- / 2>&1`" = "X/"; then as_basename=basename else as_basename=false fi if (as_dir=`dirname -- /` && test "X$as_dir" = X/) >/dev/null 2>&1; then as_dirname=dirname else as_dirname=false fi as_me=`$as_basename -- "$0" || $as_expr X/"$0" : '.*/\([^/][^/]*\)/*$' \| \ X"$0" : 'X\(//\)$' \| \ X"$0" : 'X\(/\)' \| . 2>/dev/null || $as_echo X/"$0" | sed '/^.*\/\([^/][^/]*\)\/*$/{ s//\1/ q } /^X\/\(\/\/\)$/{ s//\1/ q } /^X\/\(\/\).*/{ s//\1/ q } s/.*/./; q'` # Avoid depending upon Character Ranges. as_cr_letters='abcdefghijklmnopqrstuvwxyz' as_cr_LETTERS='ABCDEFGHIJKLMNOPQRSTUVWXYZ' as_cr_Letters=$as_cr_letters$as_cr_LETTERS as_cr_digits='0123456789' as_cr_alnum=$as_cr_Letters$as_cr_digits as_lineno_1=$LINENO as_lineno_1a=$LINENO as_lineno_2=$LINENO as_lineno_2a=$LINENO eval 'test "x$as_lineno_1'$as_run'" != "x$as_lineno_2'$as_run'" && test "x`expr $as_lineno_1'$as_run' + 1`" = "x$as_lineno_2'$as_run'"' || { # Blame Lee E. McMahon (1931-1989) for sed's syntax. :-) sed -n ' p /[$]LINENO/= ' <$as_myself | sed ' s/[$]LINENO.*/&-/ t lineno b :lineno N :loop s/[$]LINENO\([^'$as_cr_alnum'_].*\n\)\(.*\)/\2\1\2/ t loop s/-\n.*// ' >$as_me.lineno && chmod +x "$as_me.lineno" || { $as_echo "$as_me: error: cannot create $as_me.lineno; rerun with a POSIX shell" >&2; as_fn_exit 1; } # If we had to re-execute with $CONFIG_SHELL, we're ensured to have # already done that, so ensure we don't try to do so again and fall # in an infinite loop. This has already happened in practice. _as_can_reexec=no; export _as_can_reexec # Don't try to exec as it changes $[0], causing all sort of problems # (the dirname of $[0] is not the place where we might find the # original and so on. Autoconf is especially sensitive to this). . "./$as_me.lineno" # Exit status is that of the last command. exit } ECHO_C= ECHO_N= ECHO_T= case `echo -n x` in #((((( -n*) case `echo 'xy\c'` in *c*) ECHO_T=' ';; # ECHO_T is single tab character. xy) ECHO_C='\c';; *) echo `echo ksh88 bug on AIX 6.1` > /dev/null ECHO_T=' ';; esac;; *) ECHO_N='-n';; esac rm -f conf$$ conf$$.exe conf$$.file if test -d conf$$.dir; then rm -f conf$$.dir/conf$$.file else rm -f conf$$.dir mkdir conf$$.dir 2>/dev/null fi if (echo >conf$$.file) 2>/dev/null; then if ln -s conf$$.file conf$$ 2>/dev/null; then as_ln_s='ln -s' # ... but there are two gotchas: # 1) On MSYS, both `ln -s file dir' and `ln file dir' fail. # 2) DJGPP < 2.04 has no symlinks; `ln -s' creates a wrapper executable. # In both cases, we have to default to `cp -pR'. ln -s conf$$.file conf$$.dir 2>/dev/null && test ! -f conf$$.exe || as_ln_s='cp -pR' elif ln conf$$.file conf$$ 2>/dev/null; then as_ln_s=ln else as_ln_s='cp -pR' fi else as_ln_s='cp -pR' fi rm -f conf$$ conf$$.exe conf$$.dir/conf$$.file conf$$.file rmdir conf$$.dir 2>/dev/null if mkdir -p . 2>/dev/null; then as_mkdir_p='mkdir -p "$as_dir"' else test -d ./-p && rmdir ./-p as_mkdir_p=false fi as_test_x='test -x' as_executable_p=as_fn_executable_p # Sed expression to map a string onto a valid CPP name. as_tr_cpp="eval sed 'y%*$as_cr_letters%P$as_cr_LETTERS%;s%[^_$as_cr_alnum]%_%g'" # Sed expression to map a string onto a valid variable name. as_tr_sh="eval sed 'y%*+%pp%;s%[^_$as_cr_alnum]%_%g'" SHELL=${CONFIG_SHELL-/bin/sh} test -n "$DJDIR" || exec 7<&0 &1 # Name of the host. # hostname on some systems (SVR3.2, old GNU/Linux) returns a bogus exit status, # so uname gets run too. ac_hostname=`(hostname || uname -n) 2>/dev/null | sed 1q` # # Initializations. # ac_default_prefix=/usr/local ac_clean_files= ac_config_libobj_dir=. LIBOBJS= cross_compiling=no subdirs= MFLAGS= MAKEFLAGS= # Identity of this package. PACKAGE_NAME='fftw' PACKAGE_TARNAME='fftw' PACKAGE_VERSION='3.3.3' PACKAGE_STRING='fftw 3.3.3' PACKAGE_BUGREPORT='fftw@fftw.org' PACKAGE_URL='' ac_unique_file="kernel/ifftw.h" # Factoring default headers for most tests. ac_includes_default="\ #include #ifdef HAVE_SYS_TYPES_H # include #endif #ifdef HAVE_SYS_STAT_H # include #endif #ifdef STDC_HEADERS # include # include #else # ifdef HAVE_STDLIB_H # include # endif #endif #ifdef HAVE_STRING_H # if !defined STDC_HEADERS && defined HAVE_MEMORY_H # include # endif # include #endif #ifdef HAVE_STRINGS_H # include #endif #ifdef HAVE_INTTYPES_H # include #endif #ifdef HAVE_STDINT_H # include #endif #ifdef HAVE_UNISTD_H # include #endif" ac_subst_vars='am__EXEEXT_FALSE am__EXEEXT_TRUE LTLIBOBJS COMBINED_THREADS_FALSE COMBINED_THREADS_TRUE SMP_FALSE SMP_TRUE OPENMP_FALSE OPENMP_TRUE THREADS_FALSE THREADS_TRUE THREADLIBS PTHREAD_CFLAGS PTHREAD_LIBS PTHREAD_CC acx_pthread_config OPENMP_CFLAGS FLIBS ac_ct_F77 FFLAGS F77 LIBQUADMATH LIBOBJS POW_LIB ALLOCA C_FFTW_R2R_KIND STACK_ALIGN_CFLAGS NEON_CFLAGS ALTIVEC_CFLAGS AVX_CFLAGS SSE2_CFLAGS MPI_FALSE MPI_TRUE C_MPI_FINT MPIRUN MPILIBS MPICC OCAMLBUILD CPP OTOOL64 OTOOL LIPO NMEDIT DSYMUTIL MANIFEST_TOOL RANLIB ac_ct_AR AR NM ac_ct_DUMPBIN DUMPBIN LD FGREP EGREP GREP SED LIBTOOL OBJDUMP DLLTOOL AS LN_S am__fastdepCC_FALSE am__fastdepCC_TRUE CCDEPMODE am__nodep AMDEPBACKSLASH AMDEP_FALSE AMDEP_TRUE am__quote am__include DEPDIR OBJEXT EXEEXT ac_ct_CC CPPFLAGS LDFLAGS CFLAGS CC PREC_SUFFIX HAVE_NEON_FALSE HAVE_NEON_TRUE HAVE_ALTIVEC_FALSE HAVE_ALTIVEC_TRUE HAVE_AVX_FALSE HAVE_AVX_TRUE HAVE_SSE2_FALSE HAVE_SSE2_TRUE CHECK_PL_OPTS PRECISION QUAD_FALSE QUAD_TRUE LDOUBLE_FALSE LDOUBLE_TRUE SINGLE_FALSE SINGLE_TRUE host_os host_vendor host_cpu host build_os build_vendor build_cpu build SHARED_VERSION_INFO MAINT MAINTAINER_MODE_FALSE MAINTAINER_MODE_TRUE am__untar am__tar AMTAR am__leading_dot SET_MAKE AWK mkdir_p MKDIR_P INSTALL_STRIP_PROGRAM STRIP install_sh MAKEINFO AUTOHEADER AUTOMAKE AUTOCONF ACLOCAL VERSION PACKAGE CYGPATH_W am__isrc INSTALL_DATA INSTALL_SCRIPT INSTALL_PROGRAM target_alias host_alias build_alias LIBS ECHO_T ECHO_N ECHO_C DEFS mandir localedir libdir psdir pdfdir dvidir htmldir infodir docdir oldincludedir includedir localstatedir sharedstatedir sysconfdir datadir datarootdir libexecdir sbindir bindir program_transform_name prefix exec_prefix PACKAGE_URL PACKAGE_BUGREPORT PACKAGE_STRING PACKAGE_VERSION PACKAGE_TARNAME PACKAGE_NAME PATH_SEPARATOR SHELL' ac_subst_files='' ac_user_opts=' enable_option_checking enable_maintainer_mode enable_shared enable_fma enable_debug enable_debug_malloc enable_debug_alignment enable_random_estimator enable_alloca enable_single enable_float enable_long_double enable_quad_precision enable_sse enable_sse2 enable_avx enable_altivec enable_neon with_slow_timer enable_mips_zbus_timer with_our_malloc with_our_malloc16 with_windows_f77_mangling with_incoming_stack_boundary enable_dependency_tracking enable_static with_pic enable_fast_install with_gnu_ld with_sysroot enable_libtool_lock enable_mpi enable_fortran with_g77_wrappers enable_openmp enable_threads with_combined_threads ' ac_precious_vars='build_alias host_alias target_alias CC CFLAGS LDFLAGS LIBS CPPFLAGS CPP MPICC F77 FFLAGS' # Initialize some variables set by options. ac_init_help= ac_init_version=false ac_unrecognized_opts= ac_unrecognized_sep= # The variables have the same names as the options, with # dashes changed to underlines. cache_file=/dev/null exec_prefix=NONE no_create= no_recursion= prefix=NONE program_prefix=NONE program_suffix=NONE program_transform_name=s,x,x, silent= site= srcdir= verbose= x_includes=NONE x_libraries=NONE # Installation directory options. # These are left unexpanded so users can "make install exec_prefix=/foo" # and all the variables that are supposed to be based on exec_prefix # by default will actually change. # Use braces instead of parens because sh, perl, etc. also accept them. # (The list follows the same order as the GNU Coding Standards.) bindir='${exec_prefix}/bin' sbindir='${exec_prefix}/sbin' libexecdir='${exec_prefix}/libexec' datarootdir='${prefix}/share' datadir='${datarootdir}' sysconfdir='${prefix}/etc' sharedstatedir='${prefix}/com' localstatedir='${prefix}/var' includedir='${prefix}/include' oldincludedir='/usr/include' docdir='${datarootdir}/doc/${PACKAGE_TARNAME}' infodir='${datarootdir}/info' htmldir='${docdir}' dvidir='${docdir}' pdfdir='${docdir}' psdir='${docdir}' libdir='${exec_prefix}/lib' localedir='${datarootdir}/locale' mandir='${datarootdir}/man' ac_prev= ac_dashdash= for ac_option do # If the previous option needs an argument, assign it. if test -n "$ac_prev"; then eval $ac_prev=\$ac_option ac_prev= continue fi case $ac_option in *=?*) ac_optarg=`expr "X$ac_option" : '[^=]*=\(.*\)'` ;; *=) ac_optarg= ;; *) ac_optarg=yes ;; esac # Accept the important Cygnus configure options, so we can diagnose typos. case $ac_dashdash$ac_option in --) ac_dashdash=yes ;; -bindir | --bindir | --bindi | --bind | --bin | --bi) ac_prev=bindir ;; -bindir=* | --bindir=* | --bindi=* | --bind=* | --bin=* | --bi=*) bindir=$ac_optarg ;; -build | --build | --buil | --bui | --bu) ac_prev=build_alias ;; -build=* | --build=* | --buil=* | --bui=* | --bu=*) build_alias=$ac_optarg ;; -cache-file | --cache-file | --cache-fil | --cache-fi \ | --cache-f | --cache- | --cache | --cach | --cac | --ca | --c) ac_prev=cache_file ;; -cache-file=* | --cache-file=* | --cache-fil=* | --cache-fi=* \ | --cache-f=* | --cache-=* | --cache=* | --cach=* | --cac=* | --ca=* | --c=*) cache_file=$ac_optarg ;; --config-cache | -C) cache_file=config.cache ;; -datadir | --datadir | --datadi | --datad) ac_prev=datadir ;; -datadir=* | --datadir=* | --datadi=* | --datad=*) datadir=$ac_optarg ;; -datarootdir | --datarootdir | --datarootdi | --datarootd | --dataroot \ | --dataroo | --dataro | --datar) ac_prev=datarootdir ;; -datarootdir=* | --datarootdir=* | --datarootdi=* | --datarootd=* \ | --dataroot=* | --dataroo=* | --dataro=* | --datar=*) datarootdir=$ac_optarg ;; -disable-* | --disable-*) ac_useropt=`expr "x$ac_option" : 'x-*disable-\(.*\)'` # Reject names that are not valid shell variable names. expr "x$ac_useropt" : ".*[^-+._$as_cr_alnum]" >/dev/null && as_fn_error $? "invalid feature name: $ac_useropt" ac_useropt_orig=$ac_useropt ac_useropt=`$as_echo "$ac_useropt" | sed 's/[-+.]/_/g'` case $ac_user_opts in *" "enable_$ac_useropt" "*) ;; *) ac_unrecognized_opts="$ac_unrecognized_opts$ac_unrecognized_sep--disable-$ac_useropt_orig" ac_unrecognized_sep=', ';; esac eval enable_$ac_useropt=no ;; -docdir | --docdir | --docdi | --doc | --do) ac_prev=docdir ;; -docdir=* | --docdir=* | --docdi=* | --doc=* | --do=*) docdir=$ac_optarg ;; -dvidir | --dvidir | --dvidi | --dvid | --dvi | --dv) ac_prev=dvidir ;; -dvidir=* | --dvidir=* | --dvidi=* | --dvid=* | --dvi=* | --dv=*) dvidir=$ac_optarg ;; -enable-* | --enable-*) ac_useropt=`expr "x$ac_option" : 'x-*enable-\([^=]*\)'` # Reject names that are not valid shell variable names. expr "x$ac_useropt" : ".*[^-+._$as_cr_alnum]" >/dev/null && as_fn_error $? "invalid feature name: $ac_useropt" ac_useropt_orig=$ac_useropt ac_useropt=`$as_echo "$ac_useropt" | sed 's/[-+.]/_/g'` case $ac_user_opts in *" "enable_$ac_useropt" "*) ;; *) ac_unrecognized_opts="$ac_unrecognized_opts$ac_unrecognized_sep--enable-$ac_useropt_orig" ac_unrecognized_sep=', ';; esac eval enable_$ac_useropt=\$ac_optarg ;; -exec-prefix | --exec_prefix | --exec-prefix | --exec-prefi \ | --exec-pref | --exec-pre | --exec-pr | --exec-p | --exec- \ | --exec | --exe | --ex) ac_prev=exec_prefix ;; -exec-prefix=* | --exec_prefix=* | --exec-prefix=* | --exec-prefi=* \ | --exec-pref=* | --exec-pre=* | --exec-pr=* | --exec-p=* | --exec-=* \ | --exec=* | --exe=* | --ex=*) exec_prefix=$ac_optarg ;; -gas | --gas | --ga | --g) # Obsolete; use --with-gas. with_gas=yes ;; -help | --help | --hel | --he | -h) ac_init_help=long ;; -help=r* | --help=r* | --hel=r* | --he=r* | -hr*) ac_init_help=recursive ;; -help=s* | --help=s* | --hel=s* | --he=s* | -hs*) ac_init_help=short ;; -host | --host | --hos | --ho) ac_prev=host_alias ;; -host=* | --host=* | --hos=* | --ho=*) host_alias=$ac_optarg ;; -htmldir | --htmldir | --htmldi | --htmld | --html | --htm | --ht) ac_prev=htmldir ;; -htmldir=* | --htmldir=* | --htmldi=* | --htmld=* | --html=* | --htm=* \ | --ht=*) htmldir=$ac_optarg ;; -includedir | --includedir | --includedi | --included | --include \ | --includ | --inclu | --incl | --inc) ac_prev=includedir ;; -includedir=* | --includedir=* | --includedi=* | --included=* | --include=* \ | --includ=* | --inclu=* | --incl=* | --inc=*) includedir=$ac_optarg ;; -infodir | --infodir | --infodi | --infod | --info | --inf) ac_prev=infodir ;; -infodir=* | --infodir=* | --infodi=* | --infod=* | --info=* | --inf=*) infodir=$ac_optarg ;; -libdir | --libdir | --libdi | --libd) ac_prev=libdir ;; -libdir=* | --libdir=* | --libdi=* | --libd=*) libdir=$ac_optarg ;; -libexecdir | --libexecdir | --libexecdi | --libexecd | --libexec \ | --libexe | --libex | --libe) ac_prev=libexecdir ;; -libexecdir=* | --libexecdir=* | --libexecdi=* | --libexecd=* | --libexec=* \ | --libexe=* | --libex=* | --libe=*) libexecdir=$ac_optarg ;; -localedir | --localedir | --localedi | --localed | --locale) ac_prev=localedir ;; -localedir=* | --localedir=* | --localedi=* | --localed=* | --locale=*) localedir=$ac_optarg ;; -localstatedir | --localstatedir | --localstatedi | --localstated \ | --localstate | --localstat | --localsta | --localst | --locals) ac_prev=localstatedir ;; -localstatedir=* | --localstatedir=* | --localstatedi=* | --localstated=* \ | --localstate=* | --localstat=* | --localsta=* | --localst=* | --locals=*) localstatedir=$ac_optarg ;; -mandir | --mandir | --mandi | --mand | --man | --ma | --m) ac_prev=mandir ;; -mandir=* | --mandir=* | --mandi=* | --mand=* | --man=* | --ma=* | --m=*) mandir=$ac_optarg ;; -nfp | --nfp | --nf) # Obsolete; use --without-fp. with_fp=no ;; -no-create | --no-create | --no-creat | --no-crea | --no-cre \ | --no-cr | --no-c | -n) no_create=yes ;; -no-recursion | --no-recursion | --no-recursio | --no-recursi \ | --no-recurs | --no-recur | --no-recu | --no-rec | --no-re | --no-r) no_recursion=yes ;; -oldincludedir | --oldincludedir | --oldincludedi | --oldincluded \ | --oldinclude | --oldinclud | --oldinclu | --oldincl | --oldinc \ | --oldin | --oldi | --old | --ol | --o) ac_prev=oldincludedir ;; -oldincludedir=* | --oldincludedir=* | --oldincludedi=* | --oldincluded=* \ | --oldinclude=* | --oldinclud=* | --oldinclu=* | --oldincl=* | --oldinc=* \ | --oldin=* | --oldi=* | --old=* | --ol=* | --o=*) oldincludedir=$ac_optarg ;; -prefix | --prefix | --prefi | --pref | --pre | --pr | --p) ac_prev=prefix ;; -prefix=* | --prefix=* | --prefi=* | --pref=* | --pre=* | --pr=* | --p=*) prefix=$ac_optarg ;; -program-prefix | --program-prefix | --program-prefi | --program-pref \ | --program-pre | --program-pr | --program-p) ac_prev=program_prefix ;; -program-prefix=* | --program-prefix=* | --program-prefi=* \ | --program-pref=* | --program-pre=* | --program-pr=* | --program-p=*) program_prefix=$ac_optarg ;; -program-suffix | --program-suffix | --program-suffi | --program-suff \ | --program-suf | --program-su | --program-s) ac_prev=program_suffix ;; -program-suffix=* | --program-suffix=* | --program-suffi=* \ | --program-suff=* | --program-suf=* | --program-su=* | --program-s=*) program_suffix=$ac_optarg ;; -program-transform-name | --program-transform-name \ | --program-transform-nam | --program-transform-na \ | --program-transform-n | --program-transform- \ | --program-transform | --program-transfor \ | --program-transfo | --program-transf \ | --program-trans | --program-tran \ | --progr-tra | --program-tr | --program-t) ac_prev=program_transform_name ;; -program-transform-name=* | --program-transform-name=* \ | --program-transform-nam=* | --program-transform-na=* \ | --program-transform-n=* | --program-transform-=* \ | --program-transform=* | --program-transfor=* \ | --program-transfo=* | --program-transf=* \ | --program-trans=* | --program-tran=* \ | --progr-tra=* | --program-tr=* | --program-t=*) program_transform_name=$ac_optarg ;; -pdfdir | --pdfdir | --pdfdi | --pdfd | --pdf | --pd) ac_prev=pdfdir ;; -pdfdir=* | --pdfdir=* | --pdfdi=* | --pdfd=* | --pdf=* | --pd=*) pdfdir=$ac_optarg ;; -psdir | --psdir | --psdi | --psd | --ps) ac_prev=psdir ;; -psdir=* | --psdir=* | --psdi=* | --psd=* | --ps=*) psdir=$ac_optarg ;; -q | -quiet | --quiet | --quie | --qui | --qu | --q \ | -silent | --silent | --silen | --sile | --sil) silent=yes ;; -sbindir | --sbindir | --sbindi | --sbind | --sbin | --sbi | --sb) ac_prev=sbindir ;; -sbindir=* | --sbindir=* | --sbindi=* | --sbind=* | --sbin=* \ | --sbi=* | --sb=*) sbindir=$ac_optarg ;; -sharedstatedir | --sharedstatedir | --sharedstatedi \ | --sharedstated | --sharedstate | --sharedstat | --sharedsta \ | --sharedst | --shareds | --shared | --share | --shar \ | --sha | --sh) ac_prev=sharedstatedir ;; -sharedstatedir=* | --sharedstatedir=* | --sharedstatedi=* \ | --sharedstated=* | --sharedstate=* | --sharedstat=* | --sharedsta=* \ | --sharedst=* | --shareds=* | --shared=* | --share=* | --shar=* \ | --sha=* | --sh=*) sharedstatedir=$ac_optarg ;; -site | --site | --sit) ac_prev=site ;; -site=* | --site=* | --sit=*) site=$ac_optarg ;; -srcdir | --srcdir | --srcdi | --srcd | --src | --sr) ac_prev=srcdir ;; -srcdir=* | --srcdir=* | --srcdi=* | --srcd=* | --src=* | --sr=*) srcdir=$ac_optarg ;; -sysconfdir | --sysconfdir | --sysconfdi | --sysconfd | --sysconf \ | --syscon | --sysco | --sysc | --sys | --sy) ac_prev=sysconfdir ;; -sysconfdir=* | --sysconfdir=* | --sysconfdi=* | --sysconfd=* | --sysconf=* \ | --syscon=* | --sysco=* | --sysc=* | --sys=* | --sy=*) sysconfdir=$ac_optarg ;; -target | --target | --targe | --targ | --tar | --ta | --t) ac_prev=target_alias ;; -target=* | --target=* | --targe=* | --targ=* | --tar=* | --ta=* | --t=*) target_alias=$ac_optarg ;; -v | -verbose | --verbose | --verbos | --verbo | --verb) verbose=yes ;; -version | --version | --versio | --versi | --vers | -V) ac_init_version=: ;; -with-* | --with-*) ac_useropt=`expr "x$ac_option" : 'x-*with-\([^=]*\)'` # Reject names that are not valid shell variable names. expr "x$ac_useropt" : ".*[^-+._$as_cr_alnum]" >/dev/null && as_fn_error $? "invalid package name: $ac_useropt" ac_useropt_orig=$ac_useropt ac_useropt=`$as_echo "$ac_useropt" | sed 's/[-+.]/_/g'` case $ac_user_opts in *" "with_$ac_useropt" "*) ;; *) ac_unrecognized_opts="$ac_unrecognized_opts$ac_unrecognized_sep--with-$ac_useropt_orig" ac_unrecognized_sep=', ';; esac eval with_$ac_useropt=\$ac_optarg ;; -without-* | --without-*) ac_useropt=`expr "x$ac_option" : 'x-*without-\(.*\)'` # Reject names that are not valid shell variable names. expr "x$ac_useropt" : ".*[^-+._$as_cr_alnum]" >/dev/null && as_fn_error $? "invalid package name: $ac_useropt" ac_useropt_orig=$ac_useropt ac_useropt=`$as_echo "$ac_useropt" | sed 's/[-+.]/_/g'` case $ac_user_opts in *" "with_$ac_useropt" "*) ;; *) ac_unrecognized_opts="$ac_unrecognized_opts$ac_unrecognized_sep--without-$ac_useropt_orig" ac_unrecognized_sep=', ';; esac eval with_$ac_useropt=no ;; --x) # Obsolete; use --with-x. with_x=yes ;; -x-includes | --x-includes | --x-include | --x-includ | --x-inclu \ | --x-incl | --x-inc | --x-in | --x-i) ac_prev=x_includes ;; -x-includes=* | --x-includes=* | --x-include=* | --x-includ=* | --x-inclu=* \ | --x-incl=* | --x-inc=* | --x-in=* | --x-i=*) x_includes=$ac_optarg ;; -x-libraries | --x-libraries | --x-librarie | --x-librari \ | --x-librar | --x-libra | --x-libr | --x-lib | --x-li | --x-l) ac_prev=x_libraries ;; -x-libraries=* | --x-libraries=* | --x-librarie=* | --x-librari=* \ | --x-librar=* | --x-libra=* | --x-libr=* | --x-lib=* | --x-li=* | --x-l=*) x_libraries=$ac_optarg ;; -*) as_fn_error $? "unrecognized option: \`$ac_option' Try \`$0 --help' for more information" ;; *=*) ac_envvar=`expr "x$ac_option" : 'x\([^=]*\)='` # Reject names that are not valid shell variable names. case $ac_envvar in #( '' | [0-9]* | *[!_$as_cr_alnum]* ) as_fn_error $? "invalid variable name: \`$ac_envvar'" ;; esac eval $ac_envvar=\$ac_optarg export $ac_envvar ;; *) # FIXME: should be removed in autoconf 3.0. $as_echo "$as_me: WARNING: you should use --build, --host, --target" >&2 expr "x$ac_option" : ".*[^-._$as_cr_alnum]" >/dev/null && $as_echo "$as_me: WARNING: invalid host type: $ac_option" >&2 : "${build_alias=$ac_option} ${host_alias=$ac_option} ${target_alias=$ac_option}" ;; esac done if test -n "$ac_prev"; then ac_option=--`echo $ac_prev | sed 's/_/-/g'` as_fn_error $? "missing argument to $ac_option" fi if test -n "$ac_unrecognized_opts"; then case $enable_option_checking in no) ;; fatal) as_fn_error $? "unrecognized options: $ac_unrecognized_opts" ;; *) $as_echo "$as_me: WARNING: unrecognized options: $ac_unrecognized_opts" >&2 ;; esac fi # Check all directory arguments for consistency. for ac_var in exec_prefix prefix bindir sbindir libexecdir datarootdir \ datadir sysconfdir sharedstatedir localstatedir includedir \ oldincludedir docdir infodir htmldir dvidir pdfdir psdir \ libdir localedir mandir do eval ac_val=\$$ac_var # Remove trailing slashes. case $ac_val in */ ) ac_val=`expr "X$ac_val" : 'X\(.*[^/]\)' \| "X$ac_val" : 'X\(.*\)'` eval $ac_var=\$ac_val;; esac # Be sure to have absolute directory names. case $ac_val in [\\/$]* | ?:[\\/]* ) continue;; NONE | '' ) case $ac_var in *prefix ) continue;; esac;; esac as_fn_error $? "expected an absolute directory name for --$ac_var: $ac_val" done # There might be people who depend on the old broken behavior: `$host' # used to hold the argument of --host etc. # FIXME: To remove some day. build=$build_alias host=$host_alias target=$target_alias # FIXME: To remove some day. if test "x$host_alias" != x; then if test "x$build_alias" = x; then cross_compiling=maybe elif test "x$build_alias" != "x$host_alias"; then cross_compiling=yes fi fi ac_tool_prefix= test -n "$host_alias" && ac_tool_prefix=$host_alias- test "$silent" = yes && exec 6>/dev/null ac_pwd=`pwd` && test -n "$ac_pwd" && ac_ls_di=`ls -di .` && ac_pwd_ls_di=`cd "$ac_pwd" && ls -di .` || as_fn_error $? "working directory cannot be determined" test "X$ac_ls_di" = "X$ac_pwd_ls_di" || as_fn_error $? "pwd does not report name of working directory" # Find the source files, if location was not specified. if test -z "$srcdir"; then ac_srcdir_defaulted=yes # Try the directory containing this script, then the parent directory. ac_confdir=`$as_dirname -- "$as_myself" || $as_expr X"$as_myself" : 'X\(.*[^/]\)//*[^/][^/]*/*$' \| \ X"$as_myself" : 'X\(//\)[^/]' \| \ X"$as_myself" : 'X\(//\)$' \| \ X"$as_myself" : 'X\(/\)' \| . 2>/dev/null || $as_echo X"$as_myself" | sed '/^X\(.*[^/]\)\/\/*[^/][^/]*\/*$/{ s//\1/ q } /^X\(\/\/\)[^/].*/{ s//\1/ q } /^X\(\/\/\)$/{ s//\1/ q } /^X\(\/\).*/{ s//\1/ q } s/.*/./; q'` srcdir=$ac_confdir if test ! -r "$srcdir/$ac_unique_file"; then srcdir=.. fi else ac_srcdir_defaulted=no fi if test ! -r "$srcdir/$ac_unique_file"; then test "$ac_srcdir_defaulted" = yes && srcdir="$ac_confdir or .." as_fn_error $? "cannot find sources ($ac_unique_file) in $srcdir" fi ac_msg="sources are in $srcdir, but \`cd $srcdir' does not work" ac_abs_confdir=`( cd "$srcdir" && test -r "./$ac_unique_file" || as_fn_error $? "$ac_msg" pwd)` # When building in place, set srcdir=. if test "$ac_abs_confdir" = "$ac_pwd"; then srcdir=. fi # Remove unnecessary trailing slashes from srcdir. # Double slashes in file names in object file debugging info # mess up M-x gdb in Emacs. case $srcdir in */) srcdir=`expr "X$srcdir" : 'X\(.*[^/]\)' \| "X$srcdir" : 'X\(.*\)'`;; esac for ac_var in $ac_precious_vars; do eval ac_env_${ac_var}_set=\${${ac_var}+set} eval ac_env_${ac_var}_value=\$${ac_var} eval ac_cv_env_${ac_var}_set=\${${ac_var}+set} eval ac_cv_env_${ac_var}_value=\$${ac_var} done # # Report the --help message. # if test "$ac_init_help" = "long"; then # Omit some internal or obsolete options to make the list less imposing. # This message is too long to be a string in the A/UX 3.1 sh. cat <<_ACEOF \`configure' configures fftw 3.3.3 to adapt to many kinds of systems. Usage: $0 [OPTION]... [VAR=VALUE]... To assign environment variables (e.g., CC, CFLAGS...), specify them as VAR=VALUE. See below for descriptions of some of the useful variables. Defaults for the options are specified in brackets. Configuration: -h, --help display this help and exit --help=short display options specific to this package --help=recursive display the short help of all the included packages -V, --version display version information and exit -q, --quiet, --silent do not print \`checking ...' messages --cache-file=FILE cache test results in FILE [disabled] -C, --config-cache alias for \`--cache-file=config.cache' -n, --no-create do not create output files --srcdir=DIR find the sources in DIR [configure dir or \`..'] Installation directories: --prefix=PREFIX install architecture-independent files in PREFIX [$ac_default_prefix] --exec-prefix=EPREFIX install architecture-dependent files in EPREFIX [PREFIX] By default, \`make install' will install all the files in \`$ac_default_prefix/bin', \`$ac_default_prefix/lib' etc. You can specify an installation prefix other than \`$ac_default_prefix' using \`--prefix', for instance \`--prefix=\$HOME'. For better control, use the options below. Fine tuning of the installation directories: --bindir=DIR user executables [EPREFIX/bin] --sbindir=DIR system admin executables [EPREFIX/sbin] --libexecdir=DIR program executables [EPREFIX/libexec] --sysconfdir=DIR read-only single-machine data [PREFIX/etc] --sharedstatedir=DIR modifiable architecture-independent data [PREFIX/com] --localstatedir=DIR modifiable single-machine data [PREFIX/var] --libdir=DIR object code libraries [EPREFIX/lib] --includedir=DIR C header files [PREFIX/include] --oldincludedir=DIR C header files for non-gcc [/usr/include] --datarootdir=DIR read-only arch.-independent data root [PREFIX/share] --datadir=DIR read-only architecture-independent data [DATAROOTDIR] --infodir=DIR info documentation [DATAROOTDIR/info] --localedir=DIR locale-dependent data [DATAROOTDIR/locale] --mandir=DIR man documentation [DATAROOTDIR/man] --docdir=DIR documentation root [DATAROOTDIR/doc/fftw] --htmldir=DIR html documentation [DOCDIR] --dvidir=DIR dvi documentation [DOCDIR] --pdfdir=DIR pdf documentation [DOCDIR] --psdir=DIR ps documentation [DOCDIR] _ACEOF cat <<\_ACEOF Program names: --program-prefix=PREFIX prepend PREFIX to installed program names --program-suffix=SUFFIX append SUFFIX to installed program names --program-transform-name=PROGRAM run sed PROGRAM on installed program names System types: --build=BUILD configure for building on BUILD [guessed] --host=HOST cross-compile to build programs to run on HOST [BUILD] _ACEOF fi if test -n "$ac_init_help"; then case $ac_init_help in short | recursive ) echo "Configuration of fftw 3.3.3:";; esac cat <<\_ACEOF Optional Features: --disable-option-checking ignore unrecognized --enable/--with options --disable-FEATURE do not include FEATURE (same as --enable-FEATURE=no) --enable-FEATURE[=ARG] include FEATURE [ARG=yes] --enable-maintainer-mode enable make rules and dependencies not useful (and sometimes confusing) to the casual installer --enable-shared[=PKGS] build shared libraries [default=no] --enable-fma enable optimizations for machines with fused multiply-add --enable-debug compile fftw with extra runtime checks for debugging --enable-debug-malloc enable malloc debugging version --enable-debug-alignment enable alignment debugging hacks --enable-random-estimator enable pseudorandom estimator (debugging hack) --disable-alloca disable use of the alloca() function (may be broken on mingw64) --enable-single compile fftw in single precision --enable-float synonym for --enable-single --enable-long-double compile fftw in long-double precision --enable-quad-precision compile fftw in quadruple precision if available --enable-sse enable SSE optimizations --enable-sse2 enable SSE/SSE2 optimizations --enable-avx enable AVX optimizations --enable-altivec enable Altivec optimizations --enable-neon enable ARM NEON optimizations --enable-mips-zbus-timer use MIPS ZBus cycle-counter --disable-dependency-tracking speeds up one-time build --enable-dependency-tracking do not reject slow dependency extractors --enable-static[=PKGS] build static libraries [default=yes] --enable-fast-install[=PKGS] optimize for fast installation [default=yes] --disable-libtool-lock avoid locking (might break parallel builds) --enable-mpi compile FFTW MPI library --disable-fortran don't include Fortran-callable wrappers --enable-openmp use OpenMP directives for parallelism --enable-threads compile FFTW SMP threads library Optional Packages: --with-PACKAGE[=ARG] use PACKAGE [ARG=yes] --without-PACKAGE do not use PACKAGE (same as --with-PACKAGE=no) --with-slow-timer use low-precision timers (SLOW) --with-our-malloc use our aligned malloc (helpful for Win32) --with-our-malloc16 Obsolete alias for --with-our-malloc16 --with-windows-f77-mangling use common Win32 Fortran interface styles --with-incoming-stack-boundary=X Assume that stack is aligned to (1< if you have libraries in a nonstandard directory LIBS libraries to pass to the linker, e.g. -l CPPFLAGS (Objective) C/C++ preprocessor flags, e.g. -I if you have headers in a nonstandard directory CPP C preprocessor MPICC MPI C compiler command F77 Fortran 77 compiler command FFLAGS Fortran 77 compiler flags Use these variables to override the choices made by `configure' or to help it to find libraries and programs with nonstandard names/locations. Report bugs to . _ACEOF ac_status=$? fi if test "$ac_init_help" = "recursive"; then # If there are subdirs, report their specific --help. for ac_dir in : $ac_subdirs_all; do test "x$ac_dir" = x: && continue test -d "$ac_dir" || { cd "$srcdir" && ac_pwd=`pwd` && srcdir=. && test -d "$ac_dir"; } || continue ac_builddir=. case "$ac_dir" in .) ac_dir_suffix= ac_top_builddir_sub=. ac_top_build_prefix= ;; *) ac_dir_suffix=/`$as_echo "$ac_dir" | sed 's|^\.[\\/]||'` # A ".." for each directory in $ac_dir_suffix. ac_top_builddir_sub=`$as_echo "$ac_dir_suffix" | sed 's|/[^\\/]*|/..|g;s|/||'` case $ac_top_builddir_sub in "") ac_top_builddir_sub=. ac_top_build_prefix= ;; *) ac_top_build_prefix=$ac_top_builddir_sub/ ;; esac ;; esac ac_abs_top_builddir=$ac_pwd ac_abs_builddir=$ac_pwd$ac_dir_suffix # for backward compatibility: ac_top_builddir=$ac_top_build_prefix case $srcdir in .) # We are building in place. ac_srcdir=. ac_top_srcdir=$ac_top_builddir_sub ac_abs_top_srcdir=$ac_pwd ;; [\\/]* | ?:[\\/]* ) # Absolute name. ac_srcdir=$srcdir$ac_dir_suffix; ac_top_srcdir=$srcdir ac_abs_top_srcdir=$srcdir ;; *) # Relative name. ac_srcdir=$ac_top_build_prefix$srcdir$ac_dir_suffix ac_top_srcdir=$ac_top_build_prefix$srcdir ac_abs_top_srcdir=$ac_pwd/$srcdir ;; esac ac_abs_srcdir=$ac_abs_top_srcdir$ac_dir_suffix cd "$ac_dir" || { ac_status=$?; continue; } # Check for guested configure. if test -f "$ac_srcdir/configure.gnu"; then echo && $SHELL "$ac_srcdir/configure.gnu" --help=recursive elif test -f "$ac_srcdir/configure"; then echo && $SHELL "$ac_srcdir/configure" --help=recursive else $as_echo "$as_me: WARNING: no configuration information is in $ac_dir" >&2 fi || ac_status=$? cd "$ac_pwd" || { ac_status=$?; break; } done fi test -n "$ac_init_help" && exit $ac_status if $ac_init_version; then cat <<\_ACEOF fftw configure 3.3.3 generated by GNU Autoconf 2.69 Copyright (C) 2012 Free Software Foundation, Inc. This configure script is free software; the Free Software Foundation gives unlimited permission to copy, distribute and modify it. _ACEOF exit fi ## ------------------------ ## ## Autoconf initialization. ## ## ------------------------ ## # ac_fn_c_try_compile LINENO # -------------------------- # Try to compile conftest.$ac_ext, and return whether this succeeded. ac_fn_c_try_compile () { as_lineno=${as_lineno-"$1"} as_lineno_stack=as_lineno_stack=$as_lineno_stack rm -f conftest.$ac_objext if { { ac_try="$ac_compile" case "(($ac_try" in *\"* | *\`* | *\\*) ac_try_echo=\$ac_try;; *) ac_try_echo=$ac_try;; esac eval ac_try_echo="\"\$as_me:${as_lineno-$LINENO}: $ac_try_echo\"" $as_echo "$ac_try_echo"; } >&5 (eval "$ac_compile") 2>conftest.err ac_status=$? if test -s conftest.err; then grep -v '^ *+' conftest.err >conftest.er1 cat conftest.er1 >&5 mv -f conftest.er1 conftest.err fi $as_echo "$as_me:${as_lineno-$LINENO}: \$? = $ac_status" >&5 test $ac_status = 0; } && { test -z "$ac_c_werror_flag" || test ! -s conftest.err } && test -s conftest.$ac_objext; then : ac_retval=0 else $as_echo "$as_me: failed program was:" >&5 sed 's/^/| /' conftest.$ac_ext >&5 ac_retval=1 fi eval $as_lineno_stack; ${as_lineno_stack:+:} unset as_lineno as_fn_set_status $ac_retval } # ac_fn_c_try_compile # ac_fn_c_try_link LINENO # ----------------------- # Try to link conftest.$ac_ext, and return whether this succeeded. ac_fn_c_try_link () { as_lineno=${as_lineno-"$1"} as_lineno_stack=as_lineno_stack=$as_lineno_stack rm -f conftest.$ac_objext conftest$ac_exeext if { { ac_try="$ac_link" case "(($ac_try" in *\"* | *\`* | *\\*) ac_try_echo=\$ac_try;; *) ac_try_echo=$ac_try;; esac eval ac_try_echo="\"\$as_me:${as_lineno-$LINENO}: $ac_try_echo\"" $as_echo "$ac_try_echo"; } >&5 (eval "$ac_link") 2>conftest.err ac_status=$? if test -s conftest.err; then grep -v '^ *+' conftest.err >conftest.er1 cat conftest.er1 >&5 mv -f conftest.er1 conftest.err fi $as_echo "$as_me:${as_lineno-$LINENO}: \$? = $ac_status" >&5 test $ac_status = 0; } && { test -z "$ac_c_werror_flag" || test ! -s conftest.err } && test -s conftest$ac_exeext && { test "$cross_compiling" = yes || test -x conftest$ac_exeext }; then : ac_retval=0 else $as_echo "$as_me: failed program was:" >&5 sed 's/^/| /' conftest.$ac_ext >&5 ac_retval=1 fi # Delete the IPA/IPO (Inter Procedural Analysis/Optimization) information # created by the PGI compiler (conftest_ipa8_conftest.oo), as it would # interfere with the next link command; also delete a directory that is # left behind by Apple's compiler. We do this before executing the actions. rm -rf conftest.dSYM conftest_ipa8_conftest.oo eval $as_lineno_stack; ${as_lineno_stack:+:} unset as_lineno as_fn_set_status $ac_retval } # ac_fn_c_try_link # ac_fn_c_check_header_compile LINENO HEADER VAR INCLUDES # ------------------------------------------------------- # Tests whether HEADER exists and can be compiled using the include files in # INCLUDES, setting the cache variable VAR accordingly. ac_fn_c_check_header_compile () { as_lineno=${as_lineno-"$1"} as_lineno_stack=as_lineno_stack=$as_lineno_stack { $as_echo "$as_me:${as_lineno-$LINENO}: checking for $2" >&5 $as_echo_n "checking for $2... " >&6; } if eval \${$3+:} false; then : $as_echo_n "(cached) " >&6 else cat confdefs.h - <<_ACEOF >conftest.$ac_ext /* end confdefs.h. */ $4 #include <$2> _ACEOF if ac_fn_c_try_compile "$LINENO"; then : eval "$3=yes" else eval "$3=no" fi rm -f core conftest.err conftest.$ac_objext conftest.$ac_ext fi eval ac_res=\$$3 { $as_echo "$as_me:${as_lineno-$LINENO}: result: $ac_res" >&5 $as_echo "$ac_res" >&6; } eval $as_lineno_stack; ${as_lineno_stack:+:} unset as_lineno } # ac_fn_c_check_header_compile # ac_fn_c_try_cpp LINENO # ---------------------- # Try to preprocess conftest.$ac_ext, and return whether this succeeded. ac_fn_c_try_cpp () { as_lineno=${as_lineno-"$1"} as_lineno_stack=as_lineno_stack=$as_lineno_stack if { { ac_try="$ac_cpp conftest.$ac_ext" case "(($ac_try" in *\"* | *\`* | *\\*) ac_try_echo=\$ac_try;; *) ac_try_echo=$ac_try;; esac eval ac_try_echo="\"\$as_me:${as_lineno-$LINENO}: $ac_try_echo\"" $as_echo "$ac_try_echo"; } >&5 (eval "$ac_cpp conftest.$ac_ext") 2>conftest.err ac_status=$? if test -s conftest.err; then grep -v '^ *+' conftest.err >conftest.er1 cat conftest.er1 >&5 mv -f conftest.er1 conftest.err fi $as_echo "$as_me:${as_lineno-$LINENO}: \$? = $ac_status" >&5 test $ac_status = 0; } > conftest.i && { test -z "$ac_c_preproc_warn_flag$ac_c_werror_flag" || test ! -s conftest.err }; then : ac_retval=0 else $as_echo "$as_me: failed program was:" >&5 sed 's/^/| /' conftest.$ac_ext >&5 ac_retval=1 fi eval $as_lineno_stack; ${as_lineno_stack:+:} unset as_lineno as_fn_set_status $ac_retval } # ac_fn_c_try_cpp # ac_fn_c_try_run LINENO # ---------------------- # Try to link conftest.$ac_ext, and return whether this succeeded. Assumes # that executables *can* be run. ac_fn_c_try_run () { as_lineno=${as_lineno-"$1"} as_lineno_stack=as_lineno_stack=$as_lineno_stack if { { ac_try="$ac_link" case "(($ac_try" in *\"* | *\`* | *\\*) ac_try_echo=\$ac_try;; *) ac_try_echo=$ac_try;; esac eval ac_try_echo="\"\$as_me:${as_lineno-$LINENO}: $ac_try_echo\"" $as_echo "$ac_try_echo"; } >&5 (eval "$ac_link") 2>&5 ac_status=$? $as_echo "$as_me:${as_lineno-$LINENO}: \$? = $ac_status" >&5 test $ac_status = 0; } && { ac_try='./conftest$ac_exeext' { { case "(($ac_try" in *\"* | *\`* | *\\*) ac_try_echo=\$ac_try;; *) ac_try_echo=$ac_try;; esac eval ac_try_echo="\"\$as_me:${as_lineno-$LINENO}: $ac_try_echo\"" $as_echo "$ac_try_echo"; } >&5 (eval "$ac_try") 2>&5 ac_status=$? $as_echo "$as_me:${as_lineno-$LINENO}: \$? = $ac_status" >&5 test $ac_status = 0; }; }; then : ac_retval=0 else $as_echo "$as_me: program exited with status $ac_status" >&5 $as_echo "$as_me: failed program was:" >&5 sed 's/^/| /' conftest.$ac_ext >&5 ac_retval=$ac_status fi rm -rf conftest.dSYM conftest_ipa8_conftest.oo eval $as_lineno_stack; ${as_lineno_stack:+:} unset as_lineno as_fn_set_status $ac_retval } # ac_fn_c_try_run # ac_fn_c_check_func LINENO FUNC VAR # ---------------------------------- # Tests whether FUNC exists, setting the cache variable VAR accordingly ac_fn_c_check_func () { as_lineno=${as_lineno-"$1"} as_lineno_stack=as_lineno_stack=$as_lineno_stack { $as_echo "$as_me:${as_lineno-$LINENO}: checking for $2" >&5 $as_echo_n "checking for $2... " >&6; } if eval \${$3+:} false; then : $as_echo_n "(cached) " >&6 else cat confdefs.h - <<_ACEOF >conftest.$ac_ext /* end confdefs.h. */ /* Define $2 to an innocuous variant, in case declares $2. For example, HP-UX 11i declares gettimeofday. */ #define $2 innocuous_$2 /* System header to define __stub macros and hopefully few prototypes, which can conflict with char $2 (); below. Prefer to if __STDC__ is defined, since exists even on freestanding compilers. */ #ifdef __STDC__ # include #else # include #endif #undef $2 /* Override any GCC internal prototype to avoid an error. Use char because int might match the return type of a GCC builtin and then its argument prototype would still apply. */ #ifdef __cplusplus extern "C" #endif char $2 (); /* The GNU C library defines this for functions which it implements to always fail with ENOSYS. Some functions are actually named something starting with __ and the normal name is an alias. */ #if defined __stub_$2 || defined __stub___$2 choke me #endif int main () { return $2 (); ; return 0; } _ACEOF if ac_fn_c_try_link "$LINENO"; then : eval "$3=yes" else eval "$3=no" fi rm -f core conftest.err conftest.$ac_objext \ conftest$ac_exeext conftest.$ac_ext fi eval ac_res=\$$3 { $as_echo "$as_me:${as_lineno-$LINENO}: result: $ac_res" >&5 $as_echo "$ac_res" >&6; } eval $as_lineno_stack; ${as_lineno_stack:+:} unset as_lineno } # ac_fn_c_check_func # ac_fn_c_compute_int LINENO EXPR VAR INCLUDES # -------------------------------------------- # Tries to find the compile-time value of EXPR in a program that includes # INCLUDES, setting VAR accordingly. Returns whether the value could be # computed ac_fn_c_compute_int () { as_lineno=${as_lineno-"$1"} as_lineno_stack=as_lineno_stack=$as_lineno_stack if test "$cross_compiling" = yes; then # Depending upon the size, compute the lo and hi bounds. cat confdefs.h - <<_ACEOF >conftest.$ac_ext /* end confdefs.h. */ $4 int main () { static int test_array [1 - 2 * !(($2) >= 0)]; test_array [0] = 0; return test_array [0]; ; return 0; } _ACEOF if ac_fn_c_try_compile "$LINENO"; then : ac_lo=0 ac_mid=0 while :; do cat confdefs.h - <<_ACEOF >conftest.$ac_ext /* end confdefs.h. */ $4 int main () { static int test_array [1 - 2 * !(($2) <= $ac_mid)]; test_array [0] = 0; return test_array [0]; ; return 0; } _ACEOF if ac_fn_c_try_compile "$LINENO"; then : ac_hi=$ac_mid; break else as_fn_arith $ac_mid + 1 && ac_lo=$as_val if test $ac_lo -le $ac_mid; then ac_lo= ac_hi= break fi as_fn_arith 2 '*' $ac_mid + 1 && ac_mid=$as_val fi rm -f core conftest.err conftest.$ac_objext conftest.$ac_ext done else cat confdefs.h - <<_ACEOF >conftest.$ac_ext /* end confdefs.h. */ $4 int main () { static int test_array [1 - 2 * !(($2) < 0)]; test_array [0] = 0; return test_array [0]; ; return 0; } _ACEOF if ac_fn_c_try_compile "$LINENO"; then : ac_hi=-1 ac_mid=-1 while :; do cat confdefs.h - <<_ACEOF >conftest.$ac_ext /* end confdefs.h. */ $4 int main () { static int test_array [1 - 2 * !(($2) >= $ac_mid)]; test_array [0] = 0; return test_array [0]; ; return 0; } _ACEOF if ac_fn_c_try_compile "$LINENO"; then : ac_lo=$ac_mid; break else as_fn_arith '(' $ac_mid ')' - 1 && ac_hi=$as_val if test $ac_mid -le $ac_hi; then ac_lo= ac_hi= break fi as_fn_arith 2 '*' $ac_mid && ac_mid=$as_val fi rm -f core conftest.err conftest.$ac_objext conftest.$ac_ext done else ac_lo= ac_hi= fi rm -f core conftest.err conftest.$ac_objext conftest.$ac_ext fi rm -f core conftest.err conftest.$ac_objext conftest.$ac_ext # Binary search between lo and hi bounds. while test "x$ac_lo" != "x$ac_hi"; do as_fn_arith '(' $ac_hi - $ac_lo ')' / 2 + $ac_lo && ac_mid=$as_val cat confdefs.h - <<_ACEOF >conftest.$ac_ext /* end confdefs.h. */ $4 int main () { static int test_array [1 - 2 * !(($2) <= $ac_mid)]; test_array [0] = 0; return test_array [0]; ; return 0; } _ACEOF if ac_fn_c_try_compile "$LINENO"; then : ac_hi=$ac_mid else as_fn_arith '(' $ac_mid ')' + 1 && ac_lo=$as_val fi rm -f core conftest.err conftest.$ac_objext conftest.$ac_ext done case $ac_lo in #(( ?*) eval "$3=\$ac_lo"; ac_retval=0 ;; '') ac_retval=1 ;; esac else cat confdefs.h - <<_ACEOF >conftest.$ac_ext /* end confdefs.h. */ $4 static long int longval () { return $2; } static unsigned long int ulongval () { return $2; } #include #include int main () { FILE *f = fopen ("conftest.val", "w"); if (! f) return 1; if (($2) < 0) { long int i = longval (); if (i != ($2)) return 1; fprintf (f, "%ld", i); } else { unsigned long int i = ulongval (); if (i != ($2)) return 1; fprintf (f, "%lu", i); } /* Do not output a trailing newline, as this causes \r\n confusion on some platforms. */ return ferror (f) || fclose (f) != 0; ; return 0; } _ACEOF if ac_fn_c_try_run "$LINENO"; then : echo >>conftest.val; read $3 &5 $as_echo_n "checking for $2... " >&6; } if eval \${$3+:} false; then : $as_echo_n "(cached) " >&6 fi eval ac_res=\$$3 { $as_echo "$as_me:${as_lineno-$LINENO}: result: $ac_res" >&5 $as_echo "$ac_res" >&6; } else # Is the header compilable? { $as_echo "$as_me:${as_lineno-$LINENO}: checking $2 usability" >&5 $as_echo_n "checking $2 usability... " >&6; } cat confdefs.h - <<_ACEOF >conftest.$ac_ext /* end confdefs.h. */ $4 #include <$2> _ACEOF if ac_fn_c_try_compile "$LINENO"; then : ac_header_compiler=yes else ac_header_compiler=no fi rm -f core conftest.err conftest.$ac_objext conftest.$ac_ext { $as_echo "$as_me:${as_lineno-$LINENO}: result: $ac_header_compiler" >&5 $as_echo "$ac_header_compiler" >&6; } # Is the header present? { $as_echo "$as_me:${as_lineno-$LINENO}: checking $2 presence" >&5 $as_echo_n "checking $2 presence... " >&6; } cat confdefs.h - <<_ACEOF >conftest.$ac_ext /* end confdefs.h. */ #include <$2> _ACEOF if ac_fn_c_try_cpp "$LINENO"; then : ac_header_preproc=yes else ac_header_preproc=no fi rm -f conftest.err conftest.i conftest.$ac_ext { $as_echo "$as_me:${as_lineno-$LINENO}: result: $ac_header_preproc" >&5 $as_echo "$ac_header_preproc" >&6; } # So? What about this header? case $ac_header_compiler:$ac_header_preproc:$ac_c_preproc_warn_flag in #(( yes:no: ) { $as_echo "$as_me:${as_lineno-$LINENO}: WARNING: $2: accepted by the compiler, rejected by the preprocessor!" >&5 $as_echo "$as_me: WARNING: $2: accepted by the compiler, rejected by the preprocessor!" >&2;} { $as_echo "$as_me:${as_lineno-$LINENO}: WARNING: $2: proceeding with the compiler's result" >&5 $as_echo "$as_me: WARNING: $2: proceeding with the compiler's result" >&2;} ;; no:yes:* ) { $as_echo "$as_me:${as_lineno-$LINENO}: WARNING: $2: present but cannot be compiled" >&5 $as_echo "$as_me: WARNING: $2: present but cannot be compiled" >&2;} { $as_echo "$as_me:${as_lineno-$LINENO}: WARNING: $2: check for missing prerequisite headers?" >&5 $as_echo "$as_me: WARNING: $2: check for missing prerequisite headers?" >&2;} { $as_echo "$as_me:${as_lineno-$LINENO}: WARNING: $2: see the Autoconf documentation" >&5 $as_echo "$as_me: WARNING: $2: see the Autoconf documentation" >&2;} { $as_echo "$as_me:${as_lineno-$LINENO}: WARNING: $2: section \"Present But Cannot Be Compiled\"" >&5 $as_echo "$as_me: WARNING: $2: section \"Present But Cannot Be Compiled\"" >&2;} { $as_echo "$as_me:${as_lineno-$LINENO}: WARNING: $2: proceeding with the compiler's result" >&5 $as_echo "$as_me: WARNING: $2: proceeding with the compiler's result" >&2;} ( $as_echo "## ---------------------------- ## ## Report this to fftw@fftw.org ## ## ---------------------------- ##" ) | sed "s/^/$as_me: WARNING: /" >&2 ;; esac { $as_echo "$as_me:${as_lineno-$LINENO}: checking for $2" >&5 $as_echo_n "checking for $2... " >&6; } if eval \${$3+:} false; then : $as_echo_n "(cached) " >&6 else eval "$3=\$ac_header_compiler" fi eval ac_res=\$$3 { $as_echo "$as_me:${as_lineno-$LINENO}: result: $ac_res" >&5 $as_echo "$ac_res" >&6; } fi eval $as_lineno_stack; ${as_lineno_stack:+:} unset as_lineno } # ac_fn_c_check_header_mongrel # ac_fn_c_check_type LINENO TYPE VAR INCLUDES # ------------------------------------------- # Tests whether TYPE exists after having included INCLUDES, setting cache # variable VAR accordingly. ac_fn_c_check_type () { as_lineno=${as_lineno-"$1"} as_lineno_stack=as_lineno_stack=$as_lineno_stack { $as_echo "$as_me:${as_lineno-$LINENO}: checking for $2" >&5 $as_echo_n "checking for $2... " >&6; } if eval \${$3+:} false; then : $as_echo_n "(cached) " >&6 else eval "$3=no" cat confdefs.h - <<_ACEOF >conftest.$ac_ext /* end confdefs.h. */ $4 int main () { if (sizeof ($2)) return 0; ; return 0; } _ACEOF if ac_fn_c_try_compile "$LINENO"; then : cat confdefs.h - <<_ACEOF >conftest.$ac_ext /* end confdefs.h. */ $4 int main () { if (sizeof (($2))) return 0; ; return 0; } _ACEOF if ac_fn_c_try_compile "$LINENO"; then : else eval "$3=yes" fi rm -f core conftest.err conftest.$ac_objext conftest.$ac_ext fi rm -f core conftest.err conftest.$ac_objext conftest.$ac_ext fi eval ac_res=\$$3 { $as_echo "$as_me:${as_lineno-$LINENO}: result: $ac_res" >&5 $as_echo "$ac_res" >&6; } eval $as_lineno_stack; ${as_lineno_stack:+:} unset as_lineno } # ac_fn_c_check_type # ac_fn_c_check_decl LINENO SYMBOL VAR INCLUDES # --------------------------------------------- # Tests whether SYMBOL is declared in INCLUDES, setting cache variable VAR # accordingly. ac_fn_c_check_decl () { as_lineno=${as_lineno-"$1"} as_lineno_stack=as_lineno_stack=$as_lineno_stack as_decl_name=`echo $2|sed 's/ *(.*//'` as_decl_use=`echo $2|sed -e 's/(/((/' -e 's/)/) 0&/' -e 's/,/) 0& (/g'` { $as_echo "$as_me:${as_lineno-$LINENO}: checking whether $as_decl_name is declared" >&5 $as_echo_n "checking whether $as_decl_name is declared... " >&6; } if eval \${$3+:} false; then : $as_echo_n "(cached) " >&6 else cat confdefs.h - <<_ACEOF >conftest.$ac_ext /* end confdefs.h. */ $4 int main () { #ifndef $as_decl_name #ifdef __cplusplus (void) $as_decl_use; #else (void) $as_decl_name; #endif #endif ; return 0; } _ACEOF if ac_fn_c_try_compile "$LINENO"; then : eval "$3=yes" else eval "$3=no" fi rm -f core conftest.err conftest.$ac_objext conftest.$ac_ext fi eval ac_res=\$$3 { $as_echo "$as_me:${as_lineno-$LINENO}: result: $ac_res" >&5 $as_echo "$ac_res" >&6; } eval $as_lineno_stack; ${as_lineno_stack:+:} unset as_lineno } # ac_fn_c_check_decl # ac_fn_f77_try_compile LINENO # ---------------------------- # Try to compile conftest.$ac_ext, and return whether this succeeded. ac_fn_f77_try_compile () { as_lineno=${as_lineno-"$1"} as_lineno_stack=as_lineno_stack=$as_lineno_stack rm -f conftest.$ac_objext if { { ac_try="$ac_compile" case "(($ac_try" in *\"* | *\`* | *\\*) ac_try_echo=\$ac_try;; *) ac_try_echo=$ac_try;; esac eval ac_try_echo="\"\$as_me:${as_lineno-$LINENO}: $ac_try_echo\"" $as_echo "$ac_try_echo"; } >&5 (eval "$ac_compile") 2>conftest.err ac_status=$? if test -s conftest.err; then grep -v '^ *+' conftest.err >conftest.er1 cat conftest.er1 >&5 mv -f conftest.er1 conftest.err fi $as_echo "$as_me:${as_lineno-$LINENO}: \$? = $ac_status" >&5 test $ac_status = 0; } && { test -z "$ac_f77_werror_flag" || test ! -s conftest.err } && test -s conftest.$ac_objext; then : ac_retval=0 else $as_echo "$as_me: failed program was:" >&5 sed 's/^/| /' conftest.$ac_ext >&5 ac_retval=1 fi eval $as_lineno_stack; ${as_lineno_stack:+:} unset as_lineno as_fn_set_status $ac_retval } # ac_fn_f77_try_compile # ac_fn_f77_try_link LINENO # ------------------------- # Try to link conftest.$ac_ext, and return whether this succeeded. ac_fn_f77_try_link () { as_lineno=${as_lineno-"$1"} as_lineno_stack=as_lineno_stack=$as_lineno_stack rm -f conftest.$ac_objext conftest$ac_exeext if { { ac_try="$ac_link" case "(($ac_try" in *\"* | *\`* | *\\*) ac_try_echo=\$ac_try;; *) ac_try_echo=$ac_try;; esac eval ac_try_echo="\"\$as_me:${as_lineno-$LINENO}: $ac_try_echo\"" $as_echo "$ac_try_echo"; } >&5 (eval "$ac_link") 2>conftest.err ac_status=$? if test -s conftest.err; then grep -v '^ *+' conftest.err >conftest.er1 cat conftest.er1 >&5 mv -f conftest.er1 conftest.err fi $as_echo "$as_me:${as_lineno-$LINENO}: \$? = $ac_status" >&5 test $ac_status = 0; } && { test -z "$ac_f77_werror_flag" || test ! -s conftest.err } && test -s conftest$ac_exeext && { test "$cross_compiling" = yes || test -x conftest$ac_exeext }; then : ac_retval=0 else $as_echo "$as_me: failed program was:" >&5 sed 's/^/| /' conftest.$ac_ext >&5 ac_retval=1 fi # Delete the IPA/IPO (Inter Procedural Analysis/Optimization) information # created by the PGI compiler (conftest_ipa8_conftest.oo), as it would # interfere with the next link command; also delete a directory that is # left behind by Apple's compiler. We do this before executing the actions. rm -rf conftest.dSYM conftest_ipa8_conftest.oo eval $as_lineno_stack; ${as_lineno_stack:+:} unset as_lineno as_fn_set_status $ac_retval } # ac_fn_f77_try_link cat >config.log <<_ACEOF This file contains any messages produced by compilers while running configure, to aid debugging if configure makes a mistake. It was created by fftw $as_me 3.3.3, which was generated by GNU Autoconf 2.69. Invocation command line was $ $0 $@ _ACEOF exec 5>>config.log { cat <<_ASUNAME ## --------- ## ## Platform. ## ## --------- ## hostname = `(hostname || uname -n) 2>/dev/null | sed 1q` uname -m = `(uname -m) 2>/dev/null || echo unknown` uname -r = `(uname -r) 2>/dev/null || echo unknown` uname -s = `(uname -s) 2>/dev/null || echo unknown` uname -v = `(uname -v) 2>/dev/null || echo unknown` /usr/bin/uname -p = `(/usr/bin/uname -p) 2>/dev/null || echo unknown` /bin/uname -X = `(/bin/uname -X) 2>/dev/null || echo unknown` /bin/arch = `(/bin/arch) 2>/dev/null || echo unknown` /usr/bin/arch -k = `(/usr/bin/arch -k) 2>/dev/null || echo unknown` /usr/convex/getsysinfo = `(/usr/convex/getsysinfo) 2>/dev/null || echo unknown` /usr/bin/hostinfo = `(/usr/bin/hostinfo) 2>/dev/null || echo unknown` /bin/machine = `(/bin/machine) 2>/dev/null || echo unknown` /usr/bin/oslevel = `(/usr/bin/oslevel) 2>/dev/null || echo unknown` /bin/universe = `(/bin/universe) 2>/dev/null || echo unknown` _ASUNAME as_save_IFS=$IFS; IFS=$PATH_SEPARATOR for as_dir in $PATH do IFS=$as_save_IFS test -z "$as_dir" && as_dir=. $as_echo "PATH: $as_dir" done IFS=$as_save_IFS } >&5 cat >&5 <<_ACEOF ## ----------- ## ## Core tests. ## ## ----------- ## _ACEOF # Keep a trace of the command line. # Strip out --no-create and --no-recursion so they do not pile up. # Strip out --silent because we don't want to record it for future runs. # Also quote any args containing shell meta-characters. # Make two passes to allow for proper duplicate-argument suppression. ac_configure_args= ac_configure_args0= ac_configure_args1= ac_must_keep_next=false for ac_pass in 1 2 do for ac_arg do case $ac_arg in -no-create | --no-c* | -n | -no-recursion | --no-r*) continue ;; -q | -quiet | --quiet | --quie | --qui | --qu | --q \ | -silent | --silent | --silen | --sile | --sil) continue ;; *\'*) ac_arg=`$as_echo "$ac_arg" | sed "s/'/'\\\\\\\\''/g"` ;; esac case $ac_pass in 1) as_fn_append ac_configure_args0 " '$ac_arg'" ;; 2) as_fn_append ac_configure_args1 " '$ac_arg'" if test $ac_must_keep_next = true; then ac_must_keep_next=false # Got value, back to normal. else case $ac_arg in *=* | --config-cache | -C | -disable-* | --disable-* \ | -enable-* | --enable-* | -gas | --g* | -nfp | --nf* \ | -q | -quiet | --q* | -silent | --sil* | -v | -verb* \ | -with-* | --with-* | -without-* | --without-* | --x) case "$ac_configure_args0 " in "$ac_configure_args1"*" '$ac_arg' "* ) continue ;; esac ;; -* ) ac_must_keep_next=true ;; esac fi as_fn_append ac_configure_args " '$ac_arg'" ;; esac done done { ac_configure_args0=; unset ac_configure_args0;} { ac_configure_args1=; unset ac_configure_args1;} # When interrupted or exit'd, cleanup temporary files, and complete # config.log. We remove comments because anyway the quotes in there # would cause problems or look ugly. # WARNING: Use '\'' to represent an apostrophe within the trap. # WARNING: Do not start the trap code with a newline, due to a FreeBSD 4.0 bug. trap 'exit_status=$? # Save into config.log some information that might help in debugging. { echo $as_echo "## ---------------- ## ## Cache variables. ## ## ---------------- ##" echo # The following way of writing the cache mishandles newlines in values, ( for ac_var in `(set) 2>&1 | sed -n '\''s/^\([a-zA-Z_][a-zA-Z0-9_]*\)=.*/\1/p'\''`; do eval ac_val=\$$ac_var case $ac_val in #( *${as_nl}*) case $ac_var in #( *_cv_*) { $as_echo "$as_me:${as_lineno-$LINENO}: WARNING: cache variable $ac_var contains a newline" >&5 $as_echo "$as_me: WARNING: cache variable $ac_var contains a newline" >&2;} ;; esac case $ac_var in #( _ | IFS | as_nl) ;; #( BASH_ARGV | BASH_SOURCE) eval $ac_var= ;; #( *) { eval $ac_var=; unset $ac_var;} ;; esac ;; esac done (set) 2>&1 | case $as_nl`(ac_space='\'' '\''; set) 2>&1` in #( *${as_nl}ac_space=\ *) sed -n \ "s/'\''/'\''\\\\'\'''\''/g; s/^\\([_$as_cr_alnum]*_cv_[_$as_cr_alnum]*\\)=\\(.*\\)/\\1='\''\\2'\''/p" ;; #( *) sed -n "/^[_$as_cr_alnum]*_cv_[_$as_cr_alnum]*=/p" ;; esac | sort ) echo $as_echo "## ----------------- ## ## Output variables. ## ## ----------------- ##" echo for ac_var in $ac_subst_vars do eval ac_val=\$$ac_var case $ac_val in *\'\''*) ac_val=`$as_echo "$ac_val" | sed "s/'\''/'\''\\\\\\\\'\'''\''/g"`;; esac $as_echo "$ac_var='\''$ac_val'\''" done | sort echo if test -n "$ac_subst_files"; then $as_echo "## ------------------- ## ## File substitutions. ## ## ------------------- ##" echo for ac_var in $ac_subst_files do eval ac_val=\$$ac_var case $ac_val in *\'\''*) ac_val=`$as_echo "$ac_val" | sed "s/'\''/'\''\\\\\\\\'\'''\''/g"`;; esac $as_echo "$ac_var='\''$ac_val'\''" done | sort echo fi if test -s confdefs.h; then $as_echo "## ----------- ## ## confdefs.h. ## ## ----------- ##" echo cat confdefs.h echo fi test "$ac_signal" != 0 && $as_echo "$as_me: caught signal $ac_signal" $as_echo "$as_me: exit $exit_status" } >&5 rm -f core *.core core.conftest.* && rm -f -r conftest* confdefs* conf$$* $ac_clean_files && exit $exit_status ' 0 for ac_signal in 1 2 13 15; do trap 'ac_signal='$ac_signal'; as_fn_exit 1' $ac_signal done ac_signal=0 # confdefs.h avoids OS command line length limits that DEFS can exceed. rm -f -r conftest* confdefs.h $as_echo "/* confdefs.h */" > confdefs.h # Predefined preprocessor variables. cat >>confdefs.h <<_ACEOF #define PACKAGE_NAME "$PACKAGE_NAME" _ACEOF cat >>confdefs.h <<_ACEOF #define PACKAGE_TARNAME "$PACKAGE_TARNAME" _ACEOF cat >>confdefs.h <<_ACEOF #define PACKAGE_VERSION "$PACKAGE_VERSION" _ACEOF cat >>confdefs.h <<_ACEOF #define PACKAGE_STRING "$PACKAGE_STRING" _ACEOF cat >>confdefs.h <<_ACEOF #define PACKAGE_BUGREPORT "$PACKAGE_BUGREPORT" _ACEOF cat >>confdefs.h <<_ACEOF #define PACKAGE_URL "$PACKAGE_URL" _ACEOF # Let the site file select an alternate cache file if it wants to. # Prefer an explicitly selected file to automatically selected ones. ac_site_file1=NONE ac_site_file2=NONE if test -n "$CONFIG_SITE"; then # We do not want a PATH search for config.site. case $CONFIG_SITE in #(( -*) ac_site_file1=./$CONFIG_SITE;; */*) ac_site_file1=$CONFIG_SITE;; *) ac_site_file1=./$CONFIG_SITE;; esac elif test "x$prefix" != xNONE; then ac_site_file1=$prefix/share/config.site ac_site_file2=$prefix/etc/config.site else ac_site_file1=$ac_default_prefix/share/config.site ac_site_file2=$ac_default_prefix/etc/config.site fi for ac_site_file in "$ac_site_file1" "$ac_site_file2" do test "x$ac_site_file" = xNONE && continue if test /dev/null != "$ac_site_file" && test -r "$ac_site_file"; then { $as_echo "$as_me:${as_lineno-$LINENO}: loading site script $ac_site_file" >&5 $as_echo "$as_me: loading site script $ac_site_file" >&6;} sed 's/^/| /' "$ac_site_file" >&5 . "$ac_site_file" \ || { { $as_echo "$as_me:${as_lineno-$LINENO}: error: in \`$ac_pwd':" >&5 $as_echo "$as_me: error: in \`$ac_pwd':" >&2;} as_fn_error $? "failed to load site script $ac_site_file See \`config.log' for more details" "$LINENO" 5; } fi done if test -r "$cache_file"; then # Some versions of bash will fail to source /dev/null (special files # actually), so we avoid doing that. DJGPP emulates it as a regular file. if test /dev/null != "$cache_file" && test -f "$cache_file"; then { $as_echo "$as_me:${as_lineno-$LINENO}: loading cache $cache_file" >&5 $as_echo "$as_me: loading cache $cache_file" >&6;} case $cache_file in [\\/]* | ?:[\\/]* ) . "$cache_file";; *) . "./$cache_file";; esac fi else { $as_echo "$as_me:${as_lineno-$LINENO}: creating cache $cache_file" >&5 $as_echo "$as_me: creating cache $cache_file" >&6;} >$cache_file fi # Check that the precious variables saved in the cache have kept the same # value. ac_cache_corrupted=false for ac_var in $ac_precious_vars; do eval ac_old_set=\$ac_cv_env_${ac_var}_set eval ac_new_set=\$ac_env_${ac_var}_set eval ac_old_val=\$ac_cv_env_${ac_var}_value eval ac_new_val=\$ac_env_${ac_var}_value case $ac_old_set,$ac_new_set in set,) { $as_echo "$as_me:${as_lineno-$LINENO}: error: \`$ac_var' was set to \`$ac_old_val' in the previous run" >&5 $as_echo "$as_me: error: \`$ac_var' was set to \`$ac_old_val' in the previous run" >&2;} ac_cache_corrupted=: ;; ,set) { $as_echo "$as_me:${as_lineno-$LINENO}: error: \`$ac_var' was not set in the previous run" >&5 $as_echo "$as_me: error: \`$ac_var' was not set in the previous run" >&2;} ac_cache_corrupted=: ;; ,);; *) if test "x$ac_old_val" != "x$ac_new_val"; then # differences in whitespace do not lead to failure. ac_old_val_w=`echo x $ac_old_val` ac_new_val_w=`echo x $ac_new_val` if test "$ac_old_val_w" != "$ac_new_val_w"; then { $as_echo "$as_me:${as_lineno-$LINENO}: error: \`$ac_var' has changed since the previous run:" >&5 $as_echo "$as_me: error: \`$ac_var' has changed since the previous run:" >&2;} ac_cache_corrupted=: else { $as_echo "$as_me:${as_lineno-$LINENO}: warning: ignoring whitespace changes in \`$ac_var' since the previous run:" >&5 $as_echo "$as_me: warning: ignoring whitespace changes in \`$ac_var' since the previous run:" >&2;} eval $ac_var=\$ac_old_val fi { $as_echo "$as_me:${as_lineno-$LINENO}: former value: \`$ac_old_val'" >&5 $as_echo "$as_me: former value: \`$ac_old_val'" >&2;} { $as_echo "$as_me:${as_lineno-$LINENO}: current value: \`$ac_new_val'" >&5 $as_echo "$as_me: current value: \`$ac_new_val'" >&2;} fi;; esac # Pass precious variables to config.status. if test "$ac_new_set" = set; then case $ac_new_val in *\'*) ac_arg=$ac_var=`$as_echo "$ac_new_val" | sed "s/'/'\\\\\\\\''/g"` ;; *) ac_arg=$ac_var=$ac_new_val ;; esac case " $ac_configure_args " in *" '$ac_arg' "*) ;; # Avoid dups. Use of quotes ensures accuracy. *) as_fn_append ac_configure_args " '$ac_arg'" ;; esac fi done if $ac_cache_corrupted; then { $as_echo "$as_me:${as_lineno-$LINENO}: error: in \`$ac_pwd':" >&5 $as_echo "$as_me: error: in \`$ac_pwd':" >&2;} { $as_echo "$as_me:${as_lineno-$LINENO}: error: changes in the environment can compromise the build" >&5 $as_echo "$as_me: error: changes in the environment can compromise the build" >&2;} as_fn_error $? "run \`make distclean' and/or \`rm $cache_file' and start over" "$LINENO" 5 fi ## -------------------- ## ## Main body of script. ## ## -------------------- ## ac_ext=c ac_cpp='$CPP $CPPFLAGS' ac_compile='$CC -c $CFLAGS $CPPFLAGS conftest.$ac_ext >&5' ac_link='$CC -o conftest$ac_exeext $CFLAGS $CPPFLAGS $LDFLAGS conftest.$ac_ext $LIBS >&5' ac_compiler_gnu=$ac_cv_c_compiler_gnu # fftw-3.1.x was 4:X:1 # fftw-3.2.x was 5:X:2 # fftw-3.3.x was 6:X:3 SHARED_VERSION_INFO="6:2:3" # CURRENT:REVISION:AGE am__api_version='1.11' ac_aux_dir= for ac_dir in "$srcdir" "$srcdir/.." "$srcdir/../.."; do if test -f "$ac_dir/install-sh"; then ac_aux_dir=$ac_dir ac_install_sh="$ac_aux_dir/install-sh -c" break elif test -f "$ac_dir/install.sh"; then ac_aux_dir=$ac_dir ac_install_sh="$ac_aux_dir/install.sh -c" break elif test -f "$ac_dir/shtool"; then ac_aux_dir=$ac_dir ac_install_sh="$ac_aux_dir/shtool install -c" break fi done if test -z "$ac_aux_dir"; then as_fn_error $? "cannot find install-sh, install.sh, or shtool in \"$srcdir\" \"$srcdir/..\" \"$srcdir/../..\"" "$LINENO" 5 fi # These three variables are undocumented and unsupported, # and are intended to be withdrawn in a future Autoconf release. # They can cause serious problems if a builder's source tree is in a directory # whose full name contains unusual characters. ac_config_guess="$SHELL $ac_aux_dir/config.guess" # Please don't use this var. ac_config_sub="$SHELL $ac_aux_dir/config.sub" # Please don't use this var. ac_configure="$SHELL $ac_aux_dir/configure" # Please don't use this var. # Find a good install program. We prefer a C program (faster), # so one script is as good as another. But avoid the broken or # incompatible versions: # SysV /etc/install, /usr/sbin/install # SunOS /usr/etc/install # IRIX /sbin/install # AIX /bin/install # AmigaOS /C/install, which installs bootblocks on floppy discs # AIX 4 /usr/bin/installbsd, which doesn't work without a -g flag # AFS /usr/afsws/bin/install, which mishandles nonexistent args # SVR4 /usr/ucb/install, which tries to use the nonexistent group "staff" # OS/2's system install, which has a completely different semantic # ./install, which can be erroneously created by make from ./install.sh. # Reject install programs that cannot install multiple files. { $as_echo "$as_me:${as_lineno-$LINENO}: checking for a BSD-compatible install" >&5 $as_echo_n "checking for a BSD-compatible install... " >&6; } if test -z "$INSTALL"; then if ${ac_cv_path_install+:} false; then : $as_echo_n "(cached) " >&6 else as_save_IFS=$IFS; IFS=$PATH_SEPARATOR for as_dir in $PATH do IFS=$as_save_IFS test -z "$as_dir" && as_dir=. # Account for people who put trailing slashes in PATH elements. case $as_dir/ in #(( ./ | .// | /[cC]/* | \ /etc/* | /usr/sbin/* | /usr/etc/* | /sbin/* | /usr/afsws/bin/* | \ ?:[\\/]os2[\\/]install[\\/]* | ?:[\\/]OS2[\\/]INSTALL[\\/]* | \ /usr/ucb/* ) ;; *) # OSF1 and SCO ODT 3.0 have their own names for install. # Don't use installbsd from OSF since it installs stuff as root # by default. for ac_prog in ginstall scoinst install; do for ac_exec_ext in '' $ac_executable_extensions; do if as_fn_executable_p "$as_dir/$ac_prog$ac_exec_ext"; then if test $ac_prog = install && grep dspmsg "$as_dir/$ac_prog$ac_exec_ext" >/dev/null 2>&1; then # AIX install. It has an incompatible calling convention. : elif test $ac_prog = install && grep pwplus "$as_dir/$ac_prog$ac_exec_ext" >/dev/null 2>&1; then # program-specific install script used by HP pwplus--don't use. : else rm -rf conftest.one conftest.two conftest.dir echo one > conftest.one echo two > conftest.two mkdir conftest.dir if "$as_dir/$ac_prog$ac_exec_ext" -c conftest.one conftest.two "`pwd`/conftest.dir" && test -s conftest.one && test -s conftest.two && test -s conftest.dir/conftest.one && test -s conftest.dir/conftest.two then ac_cv_path_install="$as_dir/$ac_prog$ac_exec_ext -c" break 3 fi fi fi done done ;; esac done IFS=$as_save_IFS rm -rf conftest.one conftest.two conftest.dir fi if test "${ac_cv_path_install+set}" = set; then INSTALL=$ac_cv_path_install else # As a last resort, use the slow shell script. Don't cache a # value for INSTALL within a source directory, because that will # break other packages using the cache if that directory is # removed, or if the value is a relative name. INSTALL=$ac_install_sh fi fi { $as_echo "$as_me:${as_lineno-$LINENO}: result: $INSTALL" >&5 $as_echo "$INSTALL" >&6; } # Use test -z because SunOS4 sh mishandles braces in ${var-val}. # It thinks the first close brace ends the variable substitution. test -z "$INSTALL_PROGRAM" && INSTALL_PROGRAM='${INSTALL}' test -z "$INSTALL_SCRIPT" && INSTALL_SCRIPT='${INSTALL}' test -z "$INSTALL_DATA" && INSTALL_DATA='${INSTALL} -m 644' { $as_echo "$as_me:${as_lineno-$LINENO}: checking whether build environment is sane" >&5 $as_echo_n "checking whether build environment is sane... " >&6; } # Just in case sleep 1 echo timestamp > conftest.file # Reject unsafe characters in $srcdir or the absolute working directory # name. Accept space and tab only in the latter. am_lf=' ' case `pwd` in *[\\\"\#\$\&\'\`$am_lf]*) as_fn_error $? "unsafe absolute working directory name" "$LINENO" 5;; esac case $srcdir in *[\\\"\#\$\&\'\`$am_lf\ \ ]*) as_fn_error $? "unsafe srcdir value: \`$srcdir'" "$LINENO" 5;; esac # Do `set' in a subshell so we don't clobber the current shell's # arguments. Must try -L first in case configure is actually a # symlink; some systems play weird games with the mod time of symlinks # (eg FreeBSD returns the mod time of the symlink's containing # directory). if ( set X `ls -Lt "$srcdir/configure" conftest.file 2> /dev/null` if test "$*" = "X"; then # -L didn't work. set X `ls -t "$srcdir/configure" conftest.file` fi rm -f conftest.file if test "$*" != "X $srcdir/configure conftest.file" \ && test "$*" != "X conftest.file $srcdir/configure"; then # If neither matched, then we have a broken ls. This can happen # if, for instance, CONFIG_SHELL is bash and it inherits a # broken ls alias from the environment. This has actually # happened. Such a system could not be considered "sane". as_fn_error $? "ls -t appears to fail. Make sure there is not a broken alias in your environment" "$LINENO" 5 fi test "$2" = conftest.file ) then # Ok. : else as_fn_error $? "newly created file is older than distributed files! Check your system clock" "$LINENO" 5 fi { $as_echo "$as_me:${as_lineno-$LINENO}: result: yes" >&5 $as_echo "yes" >&6; } test "$program_prefix" != NONE && program_transform_name="s&^&$program_prefix&;$program_transform_name" # Use a double $ so make ignores it. test "$program_suffix" != NONE && program_transform_name="s&\$&$program_suffix&;$program_transform_name" # Double any \ or $. # By default was `s,x,x', remove it if useless. ac_script='s/[\\$]/&&/g;s/;s,x,x,$//' program_transform_name=`$as_echo "$program_transform_name" | sed "$ac_script"` # expand $ac_aux_dir to an absolute path am_aux_dir=`cd $ac_aux_dir && pwd` if test x"${MISSING+set}" != xset; then case $am_aux_dir in *\ * | *\ *) MISSING="\${SHELL} \"$am_aux_dir/missing\"" ;; *) MISSING="\${SHELL} $am_aux_dir/missing" ;; esac fi # Use eval to expand $SHELL if eval "$MISSING --run true"; then am_missing_run="$MISSING --run " else am_missing_run= { $as_echo "$as_me:${as_lineno-$LINENO}: WARNING: \`missing' script is too old or missing" >&5 $as_echo "$as_me: WARNING: \`missing' script is too old or missing" >&2;} fi if test x"${install_sh}" != xset; then case $am_aux_dir in *\ * | *\ *) install_sh="\${SHELL} '$am_aux_dir/install-sh'" ;; *) install_sh="\${SHELL} $am_aux_dir/install-sh" esac fi # Installed binaries are usually stripped using `strip' when the user # run `make install-strip'. However `strip' might not be the right # tool to use in cross-compilation environments, therefore Automake # will honor the `STRIP' environment variable to overrule this program. if test "$cross_compiling" != no; then if test -n "$ac_tool_prefix"; then # Extract the first word of "${ac_tool_prefix}strip", so it can be a program name with args. set dummy ${ac_tool_prefix}strip; ac_word=$2 { $as_echo "$as_me:${as_lineno-$LINENO}: checking for $ac_word" >&5 $as_echo_n "checking for $ac_word... " >&6; } if ${ac_cv_prog_STRIP+:} false; then : $as_echo_n "(cached) " >&6 else if test -n "$STRIP"; then ac_cv_prog_STRIP="$STRIP" # Let the user override the test. else as_save_IFS=$IFS; IFS=$PATH_SEPARATOR for as_dir in $PATH do IFS=$as_save_IFS test -z "$as_dir" && as_dir=. for ac_exec_ext in '' $ac_executable_extensions; do if as_fn_executable_p "$as_dir/$ac_word$ac_exec_ext"; then ac_cv_prog_STRIP="${ac_tool_prefix}strip" $as_echo "$as_me:${as_lineno-$LINENO}: found $as_dir/$ac_word$ac_exec_ext" >&5 break 2 fi done done IFS=$as_save_IFS fi fi STRIP=$ac_cv_prog_STRIP if test -n "$STRIP"; then { $as_echo "$as_me:${as_lineno-$LINENO}: result: $STRIP" >&5 $as_echo "$STRIP" >&6; } else { $as_echo "$as_me:${as_lineno-$LINENO}: result: no" >&5 $as_echo "no" >&6; } fi fi if test -z "$ac_cv_prog_STRIP"; then ac_ct_STRIP=$STRIP # Extract the first word of "strip", so it can be a program name with args. set dummy strip; ac_word=$2 { $as_echo "$as_me:${as_lineno-$LINENO}: checking for $ac_word" >&5 $as_echo_n "checking for $ac_word... " >&6; } if ${ac_cv_prog_ac_ct_STRIP+:} false; then : $as_echo_n "(cached) " >&6 else if test -n "$ac_ct_STRIP"; then ac_cv_prog_ac_ct_STRIP="$ac_ct_STRIP" # Let the user override the test. else as_save_IFS=$IFS; IFS=$PATH_SEPARATOR for as_dir in $PATH do IFS=$as_save_IFS test -z "$as_dir" && as_dir=. for ac_exec_ext in '' $ac_executable_extensions; do if as_fn_executable_p "$as_dir/$ac_word$ac_exec_ext"; then ac_cv_prog_ac_ct_STRIP="strip" $as_echo "$as_me:${as_lineno-$LINENO}: found $as_dir/$ac_word$ac_exec_ext" >&5 break 2 fi done done IFS=$as_save_IFS fi fi ac_ct_STRIP=$ac_cv_prog_ac_ct_STRIP if test -n "$ac_ct_STRIP"; then { $as_echo "$as_me:${as_lineno-$LINENO}: result: $ac_ct_STRIP" >&5 $as_echo "$ac_ct_STRIP" >&6; } else { $as_echo "$as_me:${as_lineno-$LINENO}: result: no" >&5 $as_echo "no" >&6; } fi if test "x$ac_ct_STRIP" = x; then STRIP=":" else case $cross_compiling:$ac_tool_warned in yes:) { $as_echo "$as_me:${as_lineno-$LINENO}: WARNING: using cross tools not prefixed with host triplet" >&5 $as_echo "$as_me: WARNING: using cross tools not prefixed with host triplet" >&2;} ac_tool_warned=yes ;; esac STRIP=$ac_ct_STRIP fi else STRIP="$ac_cv_prog_STRIP" fi fi INSTALL_STRIP_PROGRAM="\$(install_sh) -c -s" { $as_echo "$as_me:${as_lineno-$LINENO}: checking for a thread-safe mkdir -p" >&5 $as_echo_n "checking for a thread-safe mkdir -p... " >&6; } if test -z "$MKDIR_P"; then if ${ac_cv_path_mkdir+:} false; then : $as_echo_n "(cached) " >&6 else as_save_IFS=$IFS; IFS=$PATH_SEPARATOR for as_dir in $PATH$PATH_SEPARATOR/opt/sfw/bin do IFS=$as_save_IFS test -z "$as_dir" && as_dir=. for ac_prog in mkdir gmkdir; do for ac_exec_ext in '' $ac_executable_extensions; do as_fn_executable_p "$as_dir/$ac_prog$ac_exec_ext" || continue case `"$as_dir/$ac_prog$ac_exec_ext" --version 2>&1` in #( 'mkdir (GNU coreutils) '* | \ 'mkdir (coreutils) '* | \ 'mkdir (fileutils) '4.1*) ac_cv_path_mkdir=$as_dir/$ac_prog$ac_exec_ext break 3;; esac done done done IFS=$as_save_IFS fi test -d ./--version && rmdir ./--version if test "${ac_cv_path_mkdir+set}" = set; then MKDIR_P="$ac_cv_path_mkdir -p" else # As a last resort, use the slow shell script. Don't cache a # value for MKDIR_P within a source directory, because that will # break other packages using the cache if that directory is # removed, or if the value is a relative name. MKDIR_P="$ac_install_sh -d" fi fi { $as_echo "$as_me:${as_lineno-$LINENO}: result: $MKDIR_P" >&5 $as_echo "$MKDIR_P" >&6; } mkdir_p="$MKDIR_P" case $mkdir_p in [\\/$]* | ?:[\\/]*) ;; */*) mkdir_p="\$(top_builddir)/$mkdir_p" ;; esac for ac_prog in gawk mawk nawk awk do # Extract the first word of "$ac_prog", so it can be a program name with args. set dummy $ac_prog; ac_word=$2 { $as_echo "$as_me:${as_lineno-$LINENO}: checking for $ac_word" >&5 $as_echo_n "checking for $ac_word... " >&6; } if ${ac_cv_prog_AWK+:} false; then : $as_echo_n "(cached) " >&6 else if test -n "$AWK"; then ac_cv_prog_AWK="$AWK" # Let the user override the test. else as_save_IFS=$IFS; IFS=$PATH_SEPARATOR for as_dir in $PATH do IFS=$as_save_IFS test -z "$as_dir" && as_dir=. for ac_exec_ext in '' $ac_executable_extensions; do if as_fn_executable_p "$as_dir/$ac_word$ac_exec_ext"; then ac_cv_prog_AWK="$ac_prog" $as_echo "$as_me:${as_lineno-$LINENO}: found $as_dir/$ac_word$ac_exec_ext" >&5 break 2 fi done done IFS=$as_save_IFS fi fi AWK=$ac_cv_prog_AWK if test -n "$AWK"; then { $as_echo "$as_me:${as_lineno-$LINENO}: result: $AWK" >&5 $as_echo "$AWK" >&6; } else { $as_echo "$as_me:${as_lineno-$LINENO}: result: no" >&5 $as_echo "no" >&6; } fi test -n "$AWK" && break done { $as_echo "$as_me:${as_lineno-$LINENO}: checking whether ${MAKE-make} sets \$(MAKE)" >&5 $as_echo_n "checking whether ${MAKE-make} sets \$(MAKE)... " >&6; } set x ${MAKE-make} ac_make=`$as_echo "$2" | sed 's/+/p/g; s/[^a-zA-Z0-9_]/_/g'` if eval \${ac_cv_prog_make_${ac_make}_set+:} false; then : $as_echo_n "(cached) " >&6 else cat >conftest.make <<\_ACEOF SHELL = /bin/sh all: @echo '@@@%%%=$(MAKE)=@@@%%%' _ACEOF # GNU make sometimes prints "make[1]: Entering ...", which would confuse us. case `${MAKE-make} -f conftest.make 2>/dev/null` in *@@@%%%=?*=@@@%%%*) eval ac_cv_prog_make_${ac_make}_set=yes;; *) eval ac_cv_prog_make_${ac_make}_set=no;; esac rm -f conftest.make fi if eval test \$ac_cv_prog_make_${ac_make}_set = yes; then { $as_echo "$as_me:${as_lineno-$LINENO}: result: yes" >&5 $as_echo "yes" >&6; } SET_MAKE= else { $as_echo "$as_me:${as_lineno-$LINENO}: result: no" >&5 $as_echo "no" >&6; } SET_MAKE="MAKE=${MAKE-make}" fi rm -rf .tst 2>/dev/null mkdir .tst 2>/dev/null if test -d .tst; then am__leading_dot=. else am__leading_dot=_ fi rmdir .tst 2>/dev/null if test "`cd $srcdir && pwd`" != "`pwd`"; then # Use -I$(srcdir) only when $(srcdir) != ., so that make's output # is not polluted with repeated "-I." am__isrc=' -I$(srcdir)' # test to see if srcdir already configured if test -f $srcdir/config.status; then as_fn_error $? "source directory already configured; run \"make distclean\" there first" "$LINENO" 5 fi fi # test whether we have cygpath if test -z "$CYGPATH_W"; then if (cygpath --version) >/dev/null 2>/dev/null; then CYGPATH_W='cygpath -w' else CYGPATH_W=echo fi fi # Define the identity of the package. PACKAGE='fftw' VERSION='3.3.3' cat >>confdefs.h <<_ACEOF #define PACKAGE "$PACKAGE" _ACEOF cat >>confdefs.h <<_ACEOF #define VERSION "$VERSION" _ACEOF # Some tools Automake needs. ACLOCAL=${ACLOCAL-"${am_missing_run}aclocal-${am__api_version}"} AUTOCONF=${AUTOCONF-"${am_missing_run}autoconf"} AUTOMAKE=${AUTOMAKE-"${am_missing_run}automake-${am__api_version}"} AUTOHEADER=${AUTOHEADER-"${am_missing_run}autoheader"} MAKEINFO=${MAKEINFO-"${am_missing_run}makeinfo"} # We need awk for the "check" target. The system "awk" is bad on # some platforms. # Always define AMTAR for backward compatibility. Yes, it's still used # in the wild :-( We should find a proper way to deprecate it ... AMTAR='$${TAR-tar}' am__tar='$${TAR-tar} chof - "$$tardir"' am__untar='$${TAR-tar} xf -' ac_config_headers="$ac_config_headers config.h" { $as_echo "$as_me:${as_lineno-$LINENO}: checking whether to enable maintainer-specific portions of Makefiles" >&5 $as_echo_n "checking whether to enable maintainer-specific portions of Makefiles... " >&6; } # Check whether --enable-maintainer-mode was given. if test "${enable_maintainer_mode+set}" = set; then : enableval=$enable_maintainer_mode; USE_MAINTAINER_MODE=$enableval else USE_MAINTAINER_MODE=no fi { $as_echo "$as_me:${as_lineno-$LINENO}: result: $USE_MAINTAINER_MODE" >&5 $as_echo "$USE_MAINTAINER_MODE" >&6; } if test $USE_MAINTAINER_MODE = yes; then MAINTAINER_MODE_TRUE= MAINTAINER_MODE_FALSE='#' else MAINTAINER_MODE_TRUE='#' MAINTAINER_MODE_FALSE= fi MAINT=$MAINTAINER_MODE_TRUE # Check whether --enable-shared was given. if test "${enable_shared+set}" = set; then : enableval=$enable_shared; p=${PACKAGE-default} case $enableval in yes) enable_shared=yes ;; no) enable_shared=no ;; *) enable_shared=no # Look at the argument we got. We use all the common list separators. lt_save_ifs="$IFS"; IFS="${IFS}$PATH_SEPARATOR," for pkg in $enableval; do IFS="$lt_save_ifs" if test "X$pkg" = "X$p"; then enable_shared=yes fi done IFS="$lt_save_ifs" ;; esac else enable_shared=no fi # Make sure we can run config.sub. $SHELL "$ac_aux_dir/config.sub" sun4 >/dev/null 2>&1 || as_fn_error $? "cannot run $SHELL $ac_aux_dir/config.sub" "$LINENO" 5 { $as_echo "$as_me:${as_lineno-$LINENO}: checking build system type" >&5 $as_echo_n "checking build system type... " >&6; } if ${ac_cv_build+:} false; then : $as_echo_n "(cached) " >&6 else ac_build_alias=$build_alias test "x$ac_build_alias" = x && ac_build_alias=`$SHELL "$ac_aux_dir/config.guess"` test "x$ac_build_alias" = x && as_fn_error $? "cannot guess build type; you must specify one" "$LINENO" 5 ac_cv_build=`$SHELL "$ac_aux_dir/config.sub" $ac_build_alias` || as_fn_error $? "$SHELL $ac_aux_dir/config.sub $ac_build_alias failed" "$LINENO" 5 fi { $as_echo "$as_me:${as_lineno-$LINENO}: result: $ac_cv_build" >&5 $as_echo "$ac_cv_build" >&6; } case $ac_cv_build in *-*-*) ;; *) as_fn_error $? "invalid value of canonical build" "$LINENO" 5;; esac build=$ac_cv_build ac_save_IFS=$IFS; IFS='-' set x $ac_cv_build shift build_cpu=$1 build_vendor=$2 shift; shift # Remember, the first character of IFS is used to create $*, # except with old shells: build_os=$* IFS=$ac_save_IFS case $build_os in *\ *) build_os=`echo "$build_os" | sed 's/ /-/g'`;; esac { $as_echo "$as_me:${as_lineno-$LINENO}: checking host system type" >&5 $as_echo_n "checking host system type... " >&6; } if ${ac_cv_host+:} false; then : $as_echo_n "(cached) " >&6 else if test "x$host_alias" = x; then ac_cv_host=$ac_cv_build else ac_cv_host=`$SHELL "$ac_aux_dir/config.sub" $host_alias` || as_fn_error $? "$SHELL $ac_aux_dir/config.sub $host_alias failed" "$LINENO" 5 fi fi { $as_echo "$as_me:${as_lineno-$LINENO}: result: $ac_cv_host" >&5 $as_echo "$ac_cv_host" >&6; } case $ac_cv_host in *-*-*) ;; *) as_fn_error $? "invalid value of canonical host" "$LINENO" 5;; esac host=$ac_cv_host ac_save_IFS=$IFS; IFS='-' set x $ac_cv_host shift host_cpu=$1 host_vendor=$2 shift; shift # Remember, the first character of IFS is used to create $*, # except with old shells: host_os=$* IFS=$ac_save_IFS case $host_os in *\ *) host_os=`echo "$host_os" | sed 's/ /-/g'`;; esac case "${host_cpu}" in powerpc*) have_fma=yes;; ia64*) have_fma=yes;; hppa*) have_fma=yes;; mips64*) have_fma=yes;; *) have_fma=no;; esac # Check whether --enable-fma was given. if test "${enable_fma+set}" = set; then : enableval=$enable_fma; have_fma=$enableval fi if test "$have_fma"x = "yes"x; then $as_echo "#define HAVE_FMA 1" >>confdefs.h fi # Check whether --enable-debug was given. if test "${enable_debug+set}" = set; then : enableval=$enable_debug; ok=$enableval else ok=no fi if test "$ok" = "yes"; then $as_echo "#define FFTW_DEBUG 1" >>confdefs.h debug_malloc=yes else debug_malloc=no fi # Check whether --enable-debug-malloc was given. if test "${enable_debug_malloc+set}" = set; then : enableval=$enable_debug_malloc; ok=$enableval else ok=$debug_malloc fi if test "$ok" = "yes"; then $as_echo "#define FFTW_DEBUG_MALLOC 1" >>confdefs.h fi # Check whether --enable-debug-alignment was given. if test "${enable_debug_alignment+set}" = set; then : enableval=$enable_debug_alignment; ok=$enableval else ok=no fi if test "$ok" = "yes"; then $as_echo "#define FFTW_DEBUG_ALIGNMENT 1" >>confdefs.h fi # Check whether --enable-random-estimator was given. if test "${enable_random_estimator+set}" = set; then : enableval=$enable_random_estimator; ok=$enableval else ok=no fi if test "$ok" = "yes"; then $as_echo "#define FFTW_RANDOM_ESTIMATOR 1" >>confdefs.h CHECK_PL_OPTS="--estimate" fi # Check whether --enable-alloca was given. if test "${enable_alloca+set}" = set; then : enableval=$enable_alloca; ok=$enableval else ok=yes fi if test "$ok" = "yes"; then $as_echo "#define FFTW_ENABLE_ALLOCA 1" >>confdefs.h fi # Check whether --enable-single was given. if test "${enable_single+set}" = set; then : enableval=$enable_single; ok=$enableval else ok=no fi # Check whether --enable-float was given. if test "${enable_float+set}" = set; then : enableval=$enable_float; ok=$enableval fi if test "$ok" = "yes"; then $as_echo "#define FFTW_SINGLE 1" >>confdefs.h $as_echo "#define BENCHFFT_SINGLE 1" >>confdefs.h PRECISION=s else PRECISION=d fi if test "$ok" = "yes"; then SINGLE_TRUE= SINGLE_FALSE='#' else SINGLE_TRUE='#' SINGLE_FALSE= fi # Check whether --enable-long-double was given. if test "${enable_long_double+set}" = set; then : enableval=$enable_long_double; ok=$enableval else ok=no fi if test "$ok" = "yes"; then if test "$PRECISION" = "s"; then as_fn_error $? "--enable-single/--enable-long-double conflict" "$LINENO" 5 fi $as_echo "#define FFTW_LDOUBLE 1" >>confdefs.h $as_echo "#define BENCHFFT_LDOUBLE 1" >>confdefs.h PRECISION=l fi if test "$ok" = "yes"; then LDOUBLE_TRUE= LDOUBLE_FALSE='#' else LDOUBLE_TRUE='#' LDOUBLE_FALSE= fi # Check whether --enable-quad-precision was given. if test "${enable_quad_precision+set}" = set; then : enableval=$enable_quad_precision; ok=$enableval else ok=no fi if test "$ok" = "yes"; then if test "$PRECISION" != "d"; then as_fn_error $? "conflicting precisions specified" "$LINENO" 5 fi $as_echo "#define FFTW_QUAD 1" >>confdefs.h $as_echo "#define BENCHFFT_QUAD 1" >>confdefs.h PRECISION=q fi if test "$ok" = "yes"; then QUAD_TRUE= QUAD_FALSE='#' else QUAD_TRUE='#' QUAD_FALSE= fi # Check whether --enable-sse was given. if test "${enable_sse+set}" = set; then : enableval=$enable_sse; have_sse=$enableval else have_sse=no fi if test "$have_sse" = "yes"; then if test "$PRECISION" != "s"; then as_fn_error $? "SSE requires single precision" "$LINENO" 5 fi fi # Check whether --enable-sse2 was given. if test "${enable_sse2+set}" = set; then : enableval=$enable_sse2; have_sse2=$enableval else have_sse2=no fi if test "$have_sse" = "yes"; then have_sse2=yes; fi if test "$have_sse2" = "yes"; then $as_echo "#define HAVE_SSE2 1" >>confdefs.h if test "$PRECISION" != "d" -a "$PRECISION" != "s"; then as_fn_error $? "SSE2 requires single or double precision" "$LINENO" 5 fi fi if test "$have_sse2" = "yes"; then HAVE_SSE2_TRUE= HAVE_SSE2_FALSE='#' else HAVE_SSE2_TRUE='#' HAVE_SSE2_FALSE= fi # Check whether --enable-avx was given. if test "${enable_avx+set}" = set; then : enableval=$enable_avx; have_avx=$enableval else have_avx=no fi if test "$have_avx" = "yes"; then $as_echo "#define HAVE_AVX 1" >>confdefs.h if test "$PRECISION" != "d" -a "$PRECISION" != "s"; then as_fn_error $? "AVX requires single or double precision" "$LINENO" 5 fi fi if test "$have_avx" = "yes"; then HAVE_AVX_TRUE= HAVE_AVX_FALSE='#' else HAVE_AVX_TRUE='#' HAVE_AVX_FALSE= fi # Check whether --enable-altivec was given. if test "${enable_altivec+set}" = set; then : enableval=$enable_altivec; have_altivec=$enableval else have_altivec=no fi if test "$have_altivec" = "yes"; then $as_echo "#define HAVE_ALTIVEC 1" >>confdefs.h if test "$PRECISION" != "s"; then as_fn_error $? "Altivec requires single precision" "$LINENO" 5 fi fi if test "$have_altivec" = "yes"; then HAVE_ALTIVEC_TRUE= HAVE_ALTIVEC_FALSE='#' else HAVE_ALTIVEC_TRUE='#' HAVE_ALTIVEC_FALSE= fi # Check whether --enable-neon was given. if test "${enable_neon+set}" = set; then : enableval=$enable_neon; have_neon=$enableval else have_neon=no fi if test "$have_neon" = "yes"; then $as_echo "#define HAVE_NEON 1" >>confdefs.h if test "$PRECISION" != "s"; then as_fn_error $? "NEON requires single precision" "$LINENO" 5 fi fi if test "$have_neon" = "yes"; then HAVE_NEON_TRUE= HAVE_NEON_FALSE='#' else HAVE_NEON_TRUE='#' HAVE_NEON_FALSE= fi # Check whether --with-slow-timer was given. if test "${with_slow_timer+set}" = set; then : withval=$with_slow_timer; with_slow_timer=$withval else with_slow_timer=no fi if test "$with_slow_timer" = "yes"; then $as_echo "#define WITH_SLOW_TIMER 1" >>confdefs.h fi # Check whether --enable-mips_zbus_timer was given. if test "${enable_mips_zbus_timer+set}" = set; then : enableval=$enable_mips_zbus_timer; have_mips_zbus_timer=$enableval else have_mips_zbus_timer=no fi if test "$have_mips_zbus_timer" = "yes"; then $as_echo "#define HAVE_MIPS_ZBUS_TIMER 1" >>confdefs.h fi # Check whether --with-our-malloc was given. if test "${with_our_malloc+set}" = set; then : withval=$with_our_malloc; with_our_malloc=$withval else with_our_malloc=no fi # Check whether --with-our-malloc16 was given. if test "${with_our_malloc16+set}" = set; then : withval=$with_our_malloc16; with_our_malloc=$withval fi if test "$with_our_malloc" = "yes"; then $as_echo "#define WITH_OUR_MALLOC 1" >>confdefs.h fi # Check whether --with-windows-f77-mangling was given. if test "${with_windows_f77_mangling+set}" = set; then : withval=$with_windows_f77_mangling; with_windows_f77_mangling=$withval else with_windows_f77_mangling=no fi if test "$with_windows_f77_mangling" = "yes"; then $as_echo "#define WINDOWS_F77_MANGLING 1" >>confdefs.h fi # Check whether --with-incoming-stack-boundary was given. if test "${with_incoming_stack_boundary+set}" = set; then : withval=$with_incoming_stack_boundary; with_incoming_stack_boundary=$withval else with_incoming_stack_boundary=no fi case "$PRECISION" in s) PREC_SUFFIX=f;; d) PREC_SUFFIX=;; l) PREC_SUFFIX=l;; q) PREC_SUFFIX=q;; esac ac_ext=c ac_cpp='$CPP $CPPFLAGS' ac_compile='$CC -c $CFLAGS $CPPFLAGS conftest.$ac_ext >&5' ac_link='$CC -o conftest$ac_exeext $CFLAGS $CPPFLAGS $LDFLAGS conftest.$ac_ext $LIBS >&5' ac_compiler_gnu=$ac_cv_c_compiler_gnu if test -n "$ac_tool_prefix"; then # Extract the first word of "${ac_tool_prefix}gcc", so it can be a program name with args. set dummy ${ac_tool_prefix}gcc; ac_word=$2 { $as_echo "$as_me:${as_lineno-$LINENO}: checking for $ac_word" >&5 $as_echo_n "checking for $ac_word... " >&6; } if ${ac_cv_prog_CC+:} false; then : $as_echo_n "(cached) " >&6 else if test -n "$CC"; then ac_cv_prog_CC="$CC" # Let the user override the test. else as_save_IFS=$IFS; IFS=$PATH_SEPARATOR for as_dir in $PATH do IFS=$as_save_IFS test -z "$as_dir" && as_dir=. for ac_exec_ext in '' $ac_executable_extensions; do if as_fn_executable_p "$as_dir/$ac_word$ac_exec_ext"; then ac_cv_prog_CC="${ac_tool_prefix}gcc" $as_echo "$as_me:${as_lineno-$LINENO}: found $as_dir/$ac_word$ac_exec_ext" >&5 break 2 fi done done IFS=$as_save_IFS fi fi CC=$ac_cv_prog_CC if test -n "$CC"; then { $as_echo "$as_me:${as_lineno-$LINENO}: result: $CC" >&5 $as_echo "$CC" >&6; } else { $as_echo "$as_me:${as_lineno-$LINENO}: result: no" >&5 $as_echo "no" >&6; } fi fi if test -z "$ac_cv_prog_CC"; then ac_ct_CC=$CC # Extract the first word of "gcc", so it can be a program name with args. set dummy gcc; ac_word=$2 { $as_echo "$as_me:${as_lineno-$LINENO}: checking for $ac_word" >&5 $as_echo_n "checking for $ac_word... " >&6; } if ${ac_cv_prog_ac_ct_CC+:} false; then : $as_echo_n "(cached) " >&6 else if test -n "$ac_ct_CC"; then ac_cv_prog_ac_ct_CC="$ac_ct_CC" # Let the user override the test. else as_save_IFS=$IFS; IFS=$PATH_SEPARATOR for as_dir in $PATH do IFS=$as_save_IFS test -z "$as_dir" && as_dir=. for ac_exec_ext in '' $ac_executable_extensions; do if as_fn_executable_p "$as_dir/$ac_word$ac_exec_ext"; then ac_cv_prog_ac_ct_CC="gcc" $as_echo "$as_me:${as_lineno-$LINENO}: found $as_dir/$ac_word$ac_exec_ext" >&5 break 2 fi done done IFS=$as_save_IFS fi fi ac_ct_CC=$ac_cv_prog_ac_ct_CC if test -n "$ac_ct_CC"; then { $as_echo "$as_me:${as_lineno-$LINENO}: result: $ac_ct_CC" >&5 $as_echo "$ac_ct_CC" >&6; } else { $as_echo "$as_me:${as_lineno-$LINENO}: result: no" >&5 $as_echo "no" >&6; } fi if test "x$ac_ct_CC" = x; then CC="" else case $cross_compiling:$ac_tool_warned in yes:) { $as_echo "$as_me:${as_lineno-$LINENO}: WARNING: using cross tools not prefixed with host triplet" >&5 $as_echo "$as_me: WARNING: using cross tools not prefixed with host triplet" >&2;} ac_tool_warned=yes ;; esac CC=$ac_ct_CC fi else CC="$ac_cv_prog_CC" fi if test -z "$CC"; then if test -n "$ac_tool_prefix"; then # Extract the first word of "${ac_tool_prefix}cc", so it can be a program name with args. set dummy ${ac_tool_prefix}cc; ac_word=$2 { $as_echo "$as_me:${as_lineno-$LINENO}: checking for $ac_word" >&5 $as_echo_n "checking for $ac_word... " >&6; } if ${ac_cv_prog_CC+:} false; then : $as_echo_n "(cached) " >&6 else if test -n "$CC"; then ac_cv_prog_CC="$CC" # Let the user override the test. else as_save_IFS=$IFS; IFS=$PATH_SEPARATOR for as_dir in $PATH do IFS=$as_save_IFS test -z "$as_dir" && as_dir=. for ac_exec_ext in '' $ac_executable_extensions; do if as_fn_executable_p "$as_dir/$ac_word$ac_exec_ext"; then ac_cv_prog_CC="${ac_tool_prefix}cc" $as_echo "$as_me:${as_lineno-$LINENO}: found $as_dir/$ac_word$ac_exec_ext" >&5 break 2 fi done done IFS=$as_save_IFS fi fi CC=$ac_cv_prog_CC if test -n "$CC"; then { $as_echo "$as_me:${as_lineno-$LINENO}: result: $CC" >&5 $as_echo "$CC" >&6; } else { $as_echo "$as_me:${as_lineno-$LINENO}: result: no" >&5 $as_echo "no" >&6; } fi fi fi if test -z "$CC"; then # Extract the first word of "cc", so it can be a program name with args. set dummy cc; ac_word=$2 { $as_echo "$as_me:${as_lineno-$LINENO}: checking for $ac_word" >&5 $as_echo_n "checking for $ac_word... " >&6; } if ${ac_cv_prog_CC+:} false; then : $as_echo_n "(cached) " >&6 else if test -n "$CC"; then ac_cv_prog_CC="$CC" # Let the user override the test. else ac_prog_rejected=no as_save_IFS=$IFS; IFS=$PATH_SEPARATOR for as_dir in $PATH do IFS=$as_save_IFS test -z "$as_dir" && as_dir=. for ac_exec_ext in '' $ac_executable_extensions; do if as_fn_executable_p "$as_dir/$ac_word$ac_exec_ext"; then if test "$as_dir/$ac_word$ac_exec_ext" = "/usr/ucb/cc"; then ac_prog_rejected=yes continue fi ac_cv_prog_CC="cc" $as_echo "$as_me:${as_lineno-$LINENO}: found $as_dir/$ac_word$ac_exec_ext" >&5 break 2 fi done done IFS=$as_save_IFS if test $ac_prog_rejected = yes; then # We found a bogon in the path, so make sure we never use it. set dummy $ac_cv_prog_CC shift if test $# != 0; then # We chose a different compiler from the bogus one. # However, it has the same basename, so the bogon will be chosen # first if we set CC to just the basename; use the full file name. shift ac_cv_prog_CC="$as_dir/$ac_word${1+' '}$@" fi fi fi fi CC=$ac_cv_prog_CC if test -n "$CC"; then { $as_echo "$as_me:${as_lineno-$LINENO}: result: $CC" >&5 $as_echo "$CC" >&6; } else { $as_echo "$as_me:${as_lineno-$LINENO}: result: no" >&5 $as_echo "no" >&6; } fi fi if test -z "$CC"; then if test -n "$ac_tool_prefix"; then for ac_prog in cl.exe do # Extract the first word of "$ac_tool_prefix$ac_prog", so it can be a program name with args. set dummy $ac_tool_prefix$ac_prog; ac_word=$2 { $as_echo "$as_me:${as_lineno-$LINENO}: checking for $ac_word" >&5 $as_echo_n "checking for $ac_word... " >&6; } if ${ac_cv_prog_CC+:} false; then : $as_echo_n "(cached) " >&6 else if test -n "$CC"; then ac_cv_prog_CC="$CC" # Let the user override the test. else as_save_IFS=$IFS; IFS=$PATH_SEPARATOR for as_dir in $PATH do IFS=$as_save_IFS test -z "$as_dir" && as_dir=. for ac_exec_ext in '' $ac_executable_extensions; do if as_fn_executable_p "$as_dir/$ac_word$ac_exec_ext"; then ac_cv_prog_CC="$ac_tool_prefix$ac_prog" $as_echo "$as_me:${as_lineno-$LINENO}: found $as_dir/$ac_word$ac_exec_ext" >&5 break 2 fi done done IFS=$as_save_IFS fi fi CC=$ac_cv_prog_CC if test -n "$CC"; then { $as_echo "$as_me:${as_lineno-$LINENO}: result: $CC" >&5 $as_echo "$CC" >&6; } else { $as_echo "$as_me:${as_lineno-$LINENO}: result: no" >&5 $as_echo "no" >&6; } fi test -n "$CC" && break done fi if test -z "$CC"; then ac_ct_CC=$CC for ac_prog in cl.exe do # Extract the first word of "$ac_prog", so it can be a program name with args. set dummy $ac_prog; ac_word=$2 { $as_echo "$as_me:${as_lineno-$LINENO}: checking for $ac_word" >&5 $as_echo_n "checking for $ac_word... " >&6; } if ${ac_cv_prog_ac_ct_CC+:} false; then : $as_echo_n "(cached) " >&6 else if test -n "$ac_ct_CC"; then ac_cv_prog_ac_ct_CC="$ac_ct_CC" # Let the user override the test. else as_save_IFS=$IFS; IFS=$PATH_SEPARATOR for as_dir in $PATH do IFS=$as_save_IFS test -z "$as_dir" && as_dir=. for ac_exec_ext in '' $ac_executable_extensions; do if as_fn_executable_p "$as_dir/$ac_word$ac_exec_ext"; then ac_cv_prog_ac_ct_CC="$ac_prog" $as_echo "$as_me:${as_lineno-$LINENO}: found $as_dir/$ac_word$ac_exec_ext" >&5 break 2 fi done done IFS=$as_save_IFS fi fi ac_ct_CC=$ac_cv_prog_ac_ct_CC if test -n "$ac_ct_CC"; then { $as_echo "$as_me:${as_lineno-$LINENO}: result: $ac_ct_CC" >&5 $as_echo "$ac_ct_CC" >&6; } else { $as_echo "$as_me:${as_lineno-$LINENO}: result: no" >&5 $as_echo "no" >&6; } fi test -n "$ac_ct_CC" && break done if test "x$ac_ct_CC" = x; then CC="" else case $cross_compiling:$ac_tool_warned in yes:) { $as_echo "$as_me:${as_lineno-$LINENO}: WARNING: using cross tools not prefixed with host triplet" >&5 $as_echo "$as_me: WARNING: using cross tools not prefixed with host triplet" >&2;} ac_tool_warned=yes ;; esac CC=$ac_ct_CC fi fi fi test -z "$CC" && { { $as_echo "$as_me:${as_lineno-$LINENO}: error: in \`$ac_pwd':" >&5 $as_echo "$as_me: error: in \`$ac_pwd':" >&2;} as_fn_error $? "no acceptable C compiler found in \$PATH See \`config.log' for more details" "$LINENO" 5; } # Provide some information about the compiler. $as_echo "$as_me:${as_lineno-$LINENO}: checking for C compiler version" >&5 set X $ac_compile ac_compiler=$2 for ac_option in --version -v -V -qversion; do { { ac_try="$ac_compiler $ac_option >&5" case "(($ac_try" in *\"* | *\`* | *\\*) ac_try_echo=\$ac_try;; *) ac_try_echo=$ac_try;; esac eval ac_try_echo="\"\$as_me:${as_lineno-$LINENO}: $ac_try_echo\"" $as_echo "$ac_try_echo"; } >&5 (eval "$ac_compiler $ac_option >&5") 2>conftest.err ac_status=$? if test -s conftest.err; then sed '10a\ ... rest of stderr output deleted ... 10q' conftest.err >conftest.er1 cat conftest.er1 >&5 fi rm -f conftest.er1 conftest.err $as_echo "$as_me:${as_lineno-$LINENO}: \$? = $ac_status" >&5 test $ac_status = 0; } done cat confdefs.h - <<_ACEOF >conftest.$ac_ext /* end confdefs.h. */ int main () { ; return 0; } _ACEOF ac_clean_files_save=$ac_clean_files ac_clean_files="$ac_clean_files a.out a.out.dSYM a.exe b.out" # Try to create an executable without -o first, disregard a.out. # It will help us diagnose broken compilers, and finding out an intuition # of exeext. { $as_echo "$as_me:${as_lineno-$LINENO}: checking whether the C compiler works" >&5 $as_echo_n "checking whether the C compiler works... " >&6; } ac_link_default=`$as_echo "$ac_link" | sed 's/ -o *conftest[^ ]*//'` # The possible output files: ac_files="a.out conftest.exe conftest a.exe a_out.exe b.out conftest.*" ac_rmfiles= for ac_file in $ac_files do case $ac_file in *.$ac_ext | *.xcoff | *.tds | *.d | *.pdb | *.xSYM | *.bb | *.bbg | *.map | *.inf | *.dSYM | *.o | *.obj ) ;; * ) ac_rmfiles="$ac_rmfiles $ac_file";; esac done rm -f $ac_rmfiles if { { ac_try="$ac_link_default" case "(($ac_try" in *\"* | *\`* | *\\*) ac_try_echo=\$ac_try;; *) ac_try_echo=$ac_try;; esac eval ac_try_echo="\"\$as_me:${as_lineno-$LINENO}: $ac_try_echo\"" $as_echo "$ac_try_echo"; } >&5 (eval "$ac_link_default") 2>&5 ac_status=$? $as_echo "$as_me:${as_lineno-$LINENO}: \$? = $ac_status" >&5 test $ac_status = 0; }; then : # Autoconf-2.13 could set the ac_cv_exeext variable to `no'. # So ignore a value of `no', otherwise this would lead to `EXEEXT = no' # in a Makefile. We should not override ac_cv_exeext if it was cached, # so that the user can short-circuit this test for compilers unknown to # Autoconf. for ac_file in $ac_files '' do test -f "$ac_file" || continue case $ac_file in *.$ac_ext | *.xcoff | *.tds | *.d | *.pdb | *.xSYM | *.bb | *.bbg | *.map | *.inf | *.dSYM | *.o | *.obj ) ;; [ab].out ) # We found the default executable, but exeext='' is most # certainly right. break;; *.* ) if test "${ac_cv_exeext+set}" = set && test "$ac_cv_exeext" != no; then :; else ac_cv_exeext=`expr "$ac_file" : '[^.]*\(\..*\)'` fi # We set ac_cv_exeext here because the later test for it is not # safe: cross compilers may not add the suffix if given an `-o' # argument, so we may need to know it at that point already. # Even if this section looks crufty: it has the advantage of # actually working. break;; * ) break;; esac done test "$ac_cv_exeext" = no && ac_cv_exeext= else ac_file='' fi if test -z "$ac_file"; then : { $as_echo "$as_me:${as_lineno-$LINENO}: result: no" >&5 $as_echo "no" >&6; } $as_echo "$as_me: failed program was:" >&5 sed 's/^/| /' conftest.$ac_ext >&5 { { $as_echo "$as_me:${as_lineno-$LINENO}: error: in \`$ac_pwd':" >&5 $as_echo "$as_me: error: in \`$ac_pwd':" >&2;} as_fn_error 77 "C compiler cannot create executables See \`config.log' for more details" "$LINENO" 5; } else { $as_echo "$as_me:${as_lineno-$LINENO}: result: yes" >&5 $as_echo "yes" >&6; } fi { $as_echo "$as_me:${as_lineno-$LINENO}: checking for C compiler default output file name" >&5 $as_echo_n "checking for C compiler default output file name... " >&6; } { $as_echo "$as_me:${as_lineno-$LINENO}: result: $ac_file" >&5 $as_echo "$ac_file" >&6; } ac_exeext=$ac_cv_exeext rm -f -r a.out a.out.dSYM a.exe conftest$ac_cv_exeext b.out ac_clean_files=$ac_clean_files_save { $as_echo "$as_me:${as_lineno-$LINENO}: checking for suffix of executables" >&5 $as_echo_n "checking for suffix of executables... " >&6; } if { { ac_try="$ac_link" case "(($ac_try" in *\"* | *\`* | *\\*) ac_try_echo=\$ac_try;; *) ac_try_echo=$ac_try;; esac eval ac_try_echo="\"\$as_me:${as_lineno-$LINENO}: $ac_try_echo\"" $as_echo "$ac_try_echo"; } >&5 (eval "$ac_link") 2>&5 ac_status=$? $as_echo "$as_me:${as_lineno-$LINENO}: \$? = $ac_status" >&5 test $ac_status = 0; }; then : # If both `conftest.exe' and `conftest' are `present' (well, observable) # catch `conftest.exe'. For instance with Cygwin, `ls conftest' will # work properly (i.e., refer to `conftest.exe'), while it won't with # `rm'. for ac_file in conftest.exe conftest conftest.*; do test -f "$ac_file" || continue case $ac_file in *.$ac_ext | *.xcoff | *.tds | *.d | *.pdb | *.xSYM | *.bb | *.bbg | *.map | *.inf | *.dSYM | *.o | *.obj ) ;; *.* ) ac_cv_exeext=`expr "$ac_file" : '[^.]*\(\..*\)'` break;; * ) break;; esac done else { { $as_echo "$as_me:${as_lineno-$LINENO}: error: in \`$ac_pwd':" >&5 $as_echo "$as_me: error: in \`$ac_pwd':" >&2;} as_fn_error $? "cannot compute suffix of executables: cannot compile and link See \`config.log' for more details" "$LINENO" 5; } fi rm -f conftest conftest$ac_cv_exeext { $as_echo "$as_me:${as_lineno-$LINENO}: result: $ac_cv_exeext" >&5 $as_echo "$ac_cv_exeext" >&6; } rm -f conftest.$ac_ext EXEEXT=$ac_cv_exeext ac_exeext=$EXEEXT cat confdefs.h - <<_ACEOF >conftest.$ac_ext /* end confdefs.h. */ #include int main () { FILE *f = fopen ("conftest.out", "w"); return ferror (f) || fclose (f) != 0; ; return 0; } _ACEOF ac_clean_files="$ac_clean_files conftest.out" # Check that the compiler produces executables we can run. If not, either # the compiler is broken, or we cross compile. { $as_echo "$as_me:${as_lineno-$LINENO}: checking whether we are cross compiling" >&5 $as_echo_n "checking whether we are cross compiling... " >&6; } if test "$cross_compiling" != yes; then { { ac_try="$ac_link" case "(($ac_try" in *\"* | *\`* | *\\*) ac_try_echo=\$ac_try;; *) ac_try_echo=$ac_try;; esac eval ac_try_echo="\"\$as_me:${as_lineno-$LINENO}: $ac_try_echo\"" $as_echo "$ac_try_echo"; } >&5 (eval "$ac_link") 2>&5 ac_status=$? $as_echo "$as_me:${as_lineno-$LINENO}: \$? = $ac_status" >&5 test $ac_status = 0; } if { ac_try='./conftest$ac_cv_exeext' { { case "(($ac_try" in *\"* | *\`* | *\\*) ac_try_echo=\$ac_try;; *) ac_try_echo=$ac_try;; esac eval ac_try_echo="\"\$as_me:${as_lineno-$LINENO}: $ac_try_echo\"" $as_echo "$ac_try_echo"; } >&5 (eval "$ac_try") 2>&5 ac_status=$? $as_echo "$as_me:${as_lineno-$LINENO}: \$? = $ac_status" >&5 test $ac_status = 0; }; }; then cross_compiling=no else if test "$cross_compiling" = maybe; then cross_compiling=yes else { { $as_echo "$as_me:${as_lineno-$LINENO}: error: in \`$ac_pwd':" >&5 $as_echo "$as_me: error: in \`$ac_pwd':" >&2;} as_fn_error $? "cannot run C compiled programs. If you meant to cross compile, use \`--host'. See \`config.log' for more details" "$LINENO" 5; } fi fi fi { $as_echo "$as_me:${as_lineno-$LINENO}: result: $cross_compiling" >&5 $as_echo "$cross_compiling" >&6; } rm -f conftest.$ac_ext conftest$ac_cv_exeext conftest.out ac_clean_files=$ac_clean_files_save { $as_echo "$as_me:${as_lineno-$LINENO}: checking for suffix of object files" >&5 $as_echo_n "checking for suffix of object files... " >&6; } if ${ac_cv_objext+:} false; then : $as_echo_n "(cached) " >&6 else cat confdefs.h - <<_ACEOF >conftest.$ac_ext /* end confdefs.h. */ int main () { ; return 0; } _ACEOF rm -f conftest.o conftest.obj if { { ac_try="$ac_compile" case "(($ac_try" in *\"* | *\`* | *\\*) ac_try_echo=\$ac_try;; *) ac_try_echo=$ac_try;; esac eval ac_try_echo="\"\$as_me:${as_lineno-$LINENO}: $ac_try_echo\"" $as_echo "$ac_try_echo"; } >&5 (eval "$ac_compile") 2>&5 ac_status=$? $as_echo "$as_me:${as_lineno-$LINENO}: \$? = $ac_status" >&5 test $ac_status = 0; }; then : for ac_file in conftest.o conftest.obj conftest.*; do test -f "$ac_file" || continue; case $ac_file in *.$ac_ext | *.xcoff | *.tds | *.d | *.pdb | *.xSYM | *.bb | *.bbg | *.map | *.inf | *.dSYM ) ;; *) ac_cv_objext=`expr "$ac_file" : '.*\.\(.*\)'` break;; esac done else $as_echo "$as_me: failed program was:" >&5 sed 's/^/| /' conftest.$ac_ext >&5 { { $as_echo "$as_me:${as_lineno-$LINENO}: error: in \`$ac_pwd':" >&5 $as_echo "$as_me: error: in \`$ac_pwd':" >&2;} as_fn_error $? "cannot compute suffix of object files: cannot compile See \`config.log' for more details" "$LINENO" 5; } fi rm -f conftest.$ac_cv_objext conftest.$ac_ext fi { $as_echo "$as_me:${as_lineno-$LINENO}: result: $ac_cv_objext" >&5 $as_echo "$ac_cv_objext" >&6; } OBJEXT=$ac_cv_objext ac_objext=$OBJEXT { $as_echo "$as_me:${as_lineno-$LINENO}: checking whether we are using the GNU C compiler" >&5 $as_echo_n "checking whether we are using the GNU C compiler... " >&6; } if ${ac_cv_c_compiler_gnu+:} false; then : $as_echo_n "(cached) " >&6 else cat confdefs.h - <<_ACEOF >conftest.$ac_ext /* end confdefs.h. */ int main () { #ifndef __GNUC__ choke me #endif ; return 0; } _ACEOF if ac_fn_c_try_compile "$LINENO"; then : ac_compiler_gnu=yes else ac_compiler_gnu=no fi rm -f core conftest.err conftest.$ac_objext conftest.$ac_ext ac_cv_c_compiler_gnu=$ac_compiler_gnu fi { $as_echo "$as_me:${as_lineno-$LINENO}: result: $ac_cv_c_compiler_gnu" >&5 $as_echo "$ac_cv_c_compiler_gnu" >&6; } if test $ac_compiler_gnu = yes; then GCC=yes else GCC= fi ac_test_CFLAGS=${CFLAGS+set} ac_save_CFLAGS=$CFLAGS { $as_echo "$as_me:${as_lineno-$LINENO}: checking whether $CC accepts -g" >&5 $as_echo_n "checking whether $CC accepts -g... " >&6; } if ${ac_cv_prog_cc_g+:} false; then : $as_echo_n "(cached) " >&6 else ac_save_c_werror_flag=$ac_c_werror_flag ac_c_werror_flag=yes ac_cv_prog_cc_g=no CFLAGS="-g" cat confdefs.h - <<_ACEOF >conftest.$ac_ext /* end confdefs.h. */ int main () { ; return 0; } _ACEOF if ac_fn_c_try_compile "$LINENO"; then : ac_cv_prog_cc_g=yes else CFLAGS="" cat confdefs.h - <<_ACEOF >conftest.$ac_ext /* end confdefs.h. */ int main () { ; return 0; } _ACEOF if ac_fn_c_try_compile "$LINENO"; then : else ac_c_werror_flag=$ac_save_c_werror_flag CFLAGS="-g" cat confdefs.h - <<_ACEOF >conftest.$ac_ext /* end confdefs.h. */ int main () { ; return 0; } _ACEOF if ac_fn_c_try_compile "$LINENO"; then : ac_cv_prog_cc_g=yes fi rm -f core conftest.err conftest.$ac_objext conftest.$ac_ext fi rm -f core conftest.err conftest.$ac_objext conftest.$ac_ext fi rm -f core conftest.err conftest.$ac_objext conftest.$ac_ext ac_c_werror_flag=$ac_save_c_werror_flag fi { $as_echo "$as_me:${as_lineno-$LINENO}: result: $ac_cv_prog_cc_g" >&5 $as_echo "$ac_cv_prog_cc_g" >&6; } if test "$ac_test_CFLAGS" = set; then CFLAGS=$ac_save_CFLAGS elif test $ac_cv_prog_cc_g = yes; then if test "$GCC" = yes; then CFLAGS="-g -O2" else CFLAGS="-g" fi else if test "$GCC" = yes; then CFLAGS="-O2" else CFLAGS= fi fi { $as_echo "$as_me:${as_lineno-$LINENO}: checking for $CC option to accept ISO C89" >&5 $as_echo_n "checking for $CC option to accept ISO C89... " >&6; } if ${ac_cv_prog_cc_c89+:} false; then : $as_echo_n "(cached) " >&6 else ac_cv_prog_cc_c89=no ac_save_CC=$CC cat confdefs.h - <<_ACEOF >conftest.$ac_ext /* end confdefs.h. */ #include #include struct stat; /* Most of the following tests are stolen from RCS 5.7's src/conf.sh. */ struct buf { int x; }; FILE * (*rcsopen) (struct buf *, struct stat *, int); static char *e (p, i) char **p; int i; { return p[i]; } static char *f (char * (*g) (char **, int), char **p, ...) { char *s; va_list v; va_start (v,p); s = g (p, va_arg (v,int)); va_end (v); return s; } /* OSF 4.0 Compaq cc is some sort of almost-ANSI by default. It has function prototypes and stuff, but not '\xHH' hex character constants. These don't provoke an error unfortunately, instead are silently treated as 'x'. The following induces an error, until -std is added to get proper ANSI mode. Curiously '\x00'!='x' always comes out true, for an array size at least. It's necessary to write '\x00'==0 to get something that's true only with -std. */ int osf4_cc_array ['\x00' == 0 ? 1 : -1]; /* IBM C 6 for AIX is almost-ANSI by default, but it replaces macro parameters inside strings and character constants. */ #define FOO(x) 'x' int xlc6_cc_array[FOO(a) == 'x' ? 1 : -1]; int test (int i, double x); struct s1 {int (*f) (int a);}; struct s2 {int (*f) (double a);}; int pairnames (int, char **, FILE *(*)(struct buf *, struct stat *, int), int, int); int argc; char **argv; int main () { return f (e, argv, 0) != argv[0] || f (e, argv, 1) != argv[1]; ; return 0; } _ACEOF for ac_arg in '' -qlanglvl=extc89 -qlanglvl=ansi -std \ -Ae "-Aa -D_HPUX_SOURCE" "-Xc -D__EXTENSIONS__" do CC="$ac_save_CC $ac_arg" if ac_fn_c_try_compile "$LINENO"; then : ac_cv_prog_cc_c89=$ac_arg fi rm -f core conftest.err conftest.$ac_objext test "x$ac_cv_prog_cc_c89" != "xno" && break done rm -f conftest.$ac_ext CC=$ac_save_CC fi # AC_CACHE_VAL case "x$ac_cv_prog_cc_c89" in x) { $as_echo "$as_me:${as_lineno-$LINENO}: result: none needed" >&5 $as_echo "none needed" >&6; } ;; xno) { $as_echo "$as_me:${as_lineno-$LINENO}: result: unsupported" >&5 $as_echo "unsupported" >&6; } ;; *) CC="$CC $ac_cv_prog_cc_c89" { $as_echo "$as_me:${as_lineno-$LINENO}: result: $ac_cv_prog_cc_c89" >&5 $as_echo "$ac_cv_prog_cc_c89" >&6; } ;; esac if test "x$ac_cv_prog_cc_c89" != xno; then : fi ac_ext=c ac_cpp='$CPP $CPPFLAGS' ac_compile='$CC -c $CFLAGS $CPPFLAGS conftest.$ac_ext >&5' ac_link='$CC -o conftest$ac_exeext $CFLAGS $CPPFLAGS $LDFLAGS conftest.$ac_ext $LIBS >&5' ac_compiler_gnu=$ac_cv_c_compiler_gnu DEPDIR="${am__leading_dot}deps" ac_config_commands="$ac_config_commands depfiles" am_make=${MAKE-make} cat > confinc << 'END' am__doit: @echo this is the am__doit target .PHONY: am__doit END # If we don't find an include directive, just comment out the code. { $as_echo "$as_me:${as_lineno-$LINENO}: checking for style of include used by $am_make" >&5 $as_echo_n "checking for style of include used by $am_make... " >&6; } am__include="#" am__quote= _am_result=none # First try GNU make style include. echo "include confinc" > confmf # Ignore all kinds of additional output from `make'. case `$am_make -s -f confmf 2> /dev/null` in #( *the\ am__doit\ target*) am__include=include am__quote= _am_result=GNU ;; esac # Now try BSD make style include. if test "$am__include" = "#"; then echo '.include "confinc"' > confmf case `$am_make -s -f confmf 2> /dev/null` in #( *the\ am__doit\ target*) am__include=.include am__quote="\"" _am_result=BSD ;; esac fi { $as_echo "$as_me:${as_lineno-$LINENO}: result: $_am_result" >&5 $as_echo "$_am_result" >&6; } rm -f confinc confmf # Check whether --enable-dependency-tracking was given. if test "${enable_dependency_tracking+set}" = set; then : enableval=$enable_dependency_tracking; fi if test "x$enable_dependency_tracking" != xno; then am_depcomp="$ac_aux_dir/depcomp" AMDEPBACKSLASH='\' am__nodep='_no' fi if test "x$enable_dependency_tracking" != xno; then AMDEP_TRUE= AMDEP_FALSE='#' else AMDEP_TRUE='#' AMDEP_FALSE= fi depcc="$CC" am_compiler_list= { $as_echo "$as_me:${as_lineno-$LINENO}: checking dependency style of $depcc" >&5 $as_echo_n "checking dependency style of $depcc... " >&6; } if ${am_cv_CC_dependencies_compiler_type+:} false; then : $as_echo_n "(cached) " >&6 else if test -z "$AMDEP_TRUE" && test -f "$am_depcomp"; then # We make a subdir and do the tests there. Otherwise we can end up # making bogus files that we don't know about and never remove. For # instance it was reported that on HP-UX the gcc test will end up # making a dummy file named `D' -- because `-MD' means `put the output # in D'. rm -rf conftest.dir mkdir conftest.dir # Copy depcomp to subdir because otherwise we won't find it if we're # using a relative directory. cp "$am_depcomp" conftest.dir cd conftest.dir # We will build objects and dependencies in a subdirectory because # it helps to detect inapplicable dependency modes. For instance # both Tru64's cc and ICC support -MD to output dependencies as a # side effect of compilation, but ICC will put the dependencies in # the current directory while Tru64 will put them in the object # directory. mkdir sub am_cv_CC_dependencies_compiler_type=none if test "$am_compiler_list" = ""; then am_compiler_list=`sed -n 's/^#*\([a-zA-Z0-9]*\))$/\1/p' < ./depcomp` fi am__universal=false case " $depcc " in #( *\ -arch\ *\ -arch\ *) am__universal=true ;; esac for depmode in $am_compiler_list; do # Setup a source with many dependencies, because some compilers # like to wrap large dependency lists on column 80 (with \), and # we should not choose a depcomp mode which is confused by this. # # We need to recreate these files for each test, as the compiler may # overwrite some of them when testing with obscure command lines. # This happens at least with the AIX C compiler. : > sub/conftest.c for i in 1 2 3 4 5 6; do echo '#include "conftst'$i'.h"' >> sub/conftest.c # Using `: > sub/conftst$i.h' creates only sub/conftst1.h with # Solaris 8's {/usr,}/bin/sh. touch sub/conftst$i.h done echo "${am__include} ${am__quote}sub/conftest.Po${am__quote}" > confmf # We check with `-c' and `-o' for the sake of the "dashmstdout" # mode. It turns out that the SunPro C++ compiler does not properly # handle `-M -o', and we need to detect this. Also, some Intel # versions had trouble with output in subdirs am__obj=sub/conftest.${OBJEXT-o} am__minus_obj="-o $am__obj" case $depmode in gcc) # This depmode causes a compiler race in universal mode. test "$am__universal" = false || continue ;; nosideeffect) # after this tag, mechanisms are not by side-effect, so they'll # only be used when explicitly requested if test "x$enable_dependency_tracking" = xyes; then continue else break fi ;; msvc7 | msvc7msys | msvisualcpp | msvcmsys) # This compiler won't grok `-c -o', but also, the minuso test has # not run yet. These depmodes are late enough in the game, and # so weak that their functioning should not be impacted. am__obj=conftest.${OBJEXT-o} am__minus_obj= ;; none) break ;; esac if depmode=$depmode \ source=sub/conftest.c object=$am__obj \ depfile=sub/conftest.Po tmpdepfile=sub/conftest.TPo \ $SHELL ./depcomp $depcc -c $am__minus_obj sub/conftest.c \ >/dev/null 2>conftest.err && grep sub/conftst1.h sub/conftest.Po > /dev/null 2>&1 && grep sub/conftst6.h sub/conftest.Po > /dev/null 2>&1 && grep $am__obj sub/conftest.Po > /dev/null 2>&1 && ${MAKE-make} -s -f confmf > /dev/null 2>&1; then # icc doesn't choke on unknown options, it will just issue warnings # or remarks (even with -Werror). So we grep stderr for any message # that says an option was ignored or not supported. # When given -MP, icc 7.0 and 7.1 complain thusly: # icc: Command line warning: ignoring option '-M'; no argument required # The diagnosis changed in icc 8.0: # icc: Command line remark: option '-MP' not supported if (grep 'ignoring option' conftest.err || grep 'not supported' conftest.err) >/dev/null 2>&1; then :; else am_cv_CC_dependencies_compiler_type=$depmode break fi fi done cd .. rm -rf conftest.dir else am_cv_CC_dependencies_compiler_type=none fi fi { $as_echo "$as_me:${as_lineno-$LINENO}: result: $am_cv_CC_dependencies_compiler_type" >&5 $as_echo "$am_cv_CC_dependencies_compiler_type" >&6; } CCDEPMODE=depmode=$am_cv_CC_dependencies_compiler_type if test "x$enable_dependency_tracking" != xno \ && test "$am_cv_CC_dependencies_compiler_type" = gcc3; then am__fastdepCC_TRUE= am__fastdepCC_FALSE='#' else am__fastdepCC_TRUE='#' am__fastdepCC_FALSE= fi if test "x$CC" != xcc; then { $as_echo "$as_me:${as_lineno-$LINENO}: checking whether $CC and cc understand -c and -o together" >&5 $as_echo_n "checking whether $CC and cc understand -c and -o together... " >&6; } else { $as_echo "$as_me:${as_lineno-$LINENO}: checking whether cc understands -c and -o together" >&5 $as_echo_n "checking whether cc understands -c and -o together... " >&6; } fi set dummy $CC; ac_cc=`$as_echo "$2" | sed 's/[^a-zA-Z0-9_]/_/g;s/^[0-9]/_/'` if eval \${ac_cv_prog_cc_${ac_cc}_c_o+:} false; then : $as_echo_n "(cached) " >&6 else cat confdefs.h - <<_ACEOF >conftest.$ac_ext /* end confdefs.h. */ int main () { ; return 0; } _ACEOF # Make sure it works both with $CC and with simple cc. # We do the test twice because some compilers refuse to overwrite an # existing .o file with -o, though they will create one. ac_try='$CC -c conftest.$ac_ext -o conftest2.$ac_objext >&5' rm -f conftest2.* if { { case "(($ac_try" in *\"* | *\`* | *\\*) ac_try_echo=\$ac_try;; *) ac_try_echo=$ac_try;; esac eval ac_try_echo="\"\$as_me:${as_lineno-$LINENO}: $ac_try_echo\"" $as_echo "$ac_try_echo"; } >&5 (eval "$ac_try") 2>&5 ac_status=$? $as_echo "$as_me:${as_lineno-$LINENO}: \$? = $ac_status" >&5 test $ac_status = 0; } && test -f conftest2.$ac_objext && { { case "(($ac_try" in *\"* | *\`* | *\\*) ac_try_echo=\$ac_try;; *) ac_try_echo=$ac_try;; esac eval ac_try_echo="\"\$as_me:${as_lineno-$LINENO}: $ac_try_echo\"" $as_echo "$ac_try_echo"; } >&5 (eval "$ac_try") 2>&5 ac_status=$? $as_echo "$as_me:${as_lineno-$LINENO}: \$? = $ac_status" >&5 test $ac_status = 0; }; then eval ac_cv_prog_cc_${ac_cc}_c_o=yes if test "x$CC" != xcc; then # Test first that cc exists at all. if { ac_try='cc -c conftest.$ac_ext >&5' { { case "(($ac_try" in *\"* | *\`* | *\\*) ac_try_echo=\$ac_try;; *) ac_try_echo=$ac_try;; esac eval ac_try_echo="\"\$as_me:${as_lineno-$LINENO}: $ac_try_echo\"" $as_echo "$ac_try_echo"; } >&5 (eval "$ac_try") 2>&5 ac_status=$? $as_echo "$as_me:${as_lineno-$LINENO}: \$? = $ac_status" >&5 test $ac_status = 0; }; }; then ac_try='cc -c conftest.$ac_ext -o conftest2.$ac_objext >&5' rm -f conftest2.* if { { case "(($ac_try" in *\"* | *\`* | *\\*) ac_try_echo=\$ac_try;; *) ac_try_echo=$ac_try;; esac eval ac_try_echo="\"\$as_me:${as_lineno-$LINENO}: $ac_try_echo\"" $as_echo "$ac_try_echo"; } >&5 (eval "$ac_try") 2>&5 ac_status=$? $as_echo "$as_me:${as_lineno-$LINENO}: \$? = $ac_status" >&5 test $ac_status = 0; } && test -f conftest2.$ac_objext && { { case "(($ac_try" in *\"* | *\`* | *\\*) ac_try_echo=\$ac_try;; *) ac_try_echo=$ac_try;; esac eval ac_try_echo="\"\$as_me:${as_lineno-$LINENO}: $ac_try_echo\"" $as_echo "$ac_try_echo"; } >&5 (eval "$ac_try") 2>&5 ac_status=$? $as_echo "$as_me:${as_lineno-$LINENO}: \$? = $ac_status" >&5 test $ac_status = 0; }; then # cc works too. : else # cc exists but doesn't like -o. eval ac_cv_prog_cc_${ac_cc}_c_o=no fi fi fi else eval ac_cv_prog_cc_${ac_cc}_c_o=no fi rm -f core conftest* fi if eval test \$ac_cv_prog_cc_${ac_cc}_c_o = yes; then { $as_echo "$as_me:${as_lineno-$LINENO}: result: yes" >&5 $as_echo "yes" >&6; } else { $as_echo "$as_me:${as_lineno-$LINENO}: result: no" >&5 $as_echo "no" >&6; } $as_echo "#define NO_MINUS_C_MINUS_O 1" >>confdefs.h fi # FIXME: we rely on the cache variable name because # there is no other way. set dummy $CC am_cc=`echo $2 | sed 's/[^a-zA-Z0-9_]/_/g;s/^[0-9]/_/'` eval am_t=\$ac_cv_prog_cc_${am_cc}_c_o if test "$am_t" != yes; then # Losing compiler, so override with the script. # FIXME: It is wrong to rewrite CC. # But if we don't then we get into trouble of one sort or another. # A longer-term fix would be to have automake use am__CC in this case, # and then we could set am__CC="\$(top_srcdir)/compile \$(CC)" CC="$am_aux_dir/compile $CC" fi { $as_echo "$as_me:${as_lineno-$LINENO}: checking for C compiler vendor" >&5 $as_echo_n "checking for C compiler vendor... " >&6; } if ${ax_cv_c_compiler_vendor+:} false; then : $as_echo_n "(cached) " >&6 else ax_cv_c_compiler_vendor=unknown # note: don't check for gcc first since some other compilers define __GNUC__ for ventest in intel:__ICC,__ECC,__INTEL_COMPILER ibm:__xlc__,__xlC__,__IBMC__,__IBMCPP__ pathscale:__PATHCC__,__PATHSCALE__ gnu:__GNUC__ sun:__SUNPRO_C,__SUNPRO_CC hp:__HP_cc,__HP_aCC dec:__DECC,__DECCXX,__DECC_VER,__DECCXX_VER borland:__BORLANDC__,__TURBOC__ comeau:__COMO__ cray:_CRAYC kai:__KCC lcc:__LCC__ metrowerks:__MWERKS__ sgi:__sgi,sgi microsoft:_MSC_VER watcom:__WATCOMC__ portland:__PGI; do vencpp="defined("`echo $ventest | cut -d: -f2 | sed 's/,/) || defined(/g'`")" cat confdefs.h - <<_ACEOF >conftest.$ac_ext /* end confdefs.h. */ int main () { #if !($vencpp) thisisanerror; #endif ; return 0; } _ACEOF if ac_fn_c_try_compile "$LINENO"; then : ax_cv_c_compiler_vendor=`echo $ventest | cut -d: -f1`; break fi rm -f core conftest.err conftest.$ac_objext conftest.$ac_ext done fi { $as_echo "$as_me:${as_lineno-$LINENO}: result: $ax_cv_c_compiler_vendor" >&5 $as_echo "$ax_cv_c_compiler_vendor" >&6; } case $ac_cv_prog_cc_stdc in #( no) : ac_cv_prog_cc_c99=no; ac_cv_prog_cc_c89=no ;; #( *) : { $as_echo "$as_me:${as_lineno-$LINENO}: checking for $CC option to accept ISO C99" >&5 $as_echo_n "checking for $CC option to accept ISO C99... " >&6; } if ${ac_cv_prog_cc_c99+:} false; then : $as_echo_n "(cached) " >&6 else ac_cv_prog_cc_c99=no ac_save_CC=$CC cat confdefs.h - <<_ACEOF >conftest.$ac_ext /* end confdefs.h. */ #include #include #include #include #include // Check varargs macros. These examples are taken from C99 6.10.3.5. #define debug(...) fprintf (stderr, __VA_ARGS__) #define showlist(...) puts (#__VA_ARGS__) #define report(test,...) ((test) ? puts (#test) : printf (__VA_ARGS__)) static void test_varargs_macros (void) { int x = 1234; int y = 5678; debug ("Flag"); debug ("X = %d\n", x); showlist (The first, second, and third items.); report (x>y, "x is %d but y is %d", x, y); } // Check long long types. #define BIG64 18446744073709551615ull #define BIG32 4294967295ul #define BIG_OK (BIG64 / BIG32 == 4294967297ull && BIG64 % BIG32 == 0) #if !BIG_OK your preprocessor is broken; #endif #if BIG_OK #else your preprocessor is broken; #endif static long long int bignum = -9223372036854775807LL; static unsigned long long int ubignum = BIG64; struct incomplete_array { int datasize; double data[]; }; struct named_init { int number; const wchar_t *name; double average; }; typedef const char *ccp; static inline int test_restrict (ccp restrict text) { // See if C++-style comments work. // Iterate through items via the restricted pointer. // Also check for declarations in for loops. for (unsigned int i = 0; *(text+i) != '\0'; ++i) continue; return 0; } // Check varargs and va_copy. static void test_varargs (const char *format, ...) { va_list args; va_start (args, format); va_list args_copy; va_copy (args_copy, args); const char *str; int number; float fnumber; while (*format) { switch (*format++) { case 's': // string str = va_arg (args_copy, const char *); break; case 'd': // int number = va_arg (args_copy, int); break; case 'f': // float fnumber = va_arg (args_copy, double); break; default: break; } } va_end (args_copy); va_end (args); } int main () { // Check bool. _Bool success = false; // Check restrict. if (test_restrict ("String literal") == 0) success = true; char *restrict newvar = "Another string"; // Check varargs. test_varargs ("s, d' f .", "string", 65, 34.234); test_varargs_macros (); // Check flexible array members. struct incomplete_array *ia = malloc (sizeof (struct incomplete_array) + (sizeof (double) * 10)); ia->datasize = 10; for (int i = 0; i < ia->datasize; ++i) ia->data[i] = i * 1.234; // Check named initializers. struct named_init ni = { .number = 34, .name = L"Test wide string", .average = 543.34343, }; ni.number = 58; int dynamic_array[ni.number]; dynamic_array[ni.number - 1] = 543; // work around unused variable warnings return (!success || bignum == 0LL || ubignum == 0uLL || newvar[0] == 'x' || dynamic_array[ni.number - 1] != 543); ; return 0; } _ACEOF for ac_arg in '' -std=gnu99 -std=c99 -c99 -AC99 -D_STDC_C99= -qlanglvl=extc99 do CC="$ac_save_CC $ac_arg" if ac_fn_c_try_compile "$LINENO"; then : ac_cv_prog_cc_c99=$ac_arg fi rm -f core conftest.err conftest.$ac_objext test "x$ac_cv_prog_cc_c99" != "xno" && break done rm -f conftest.$ac_ext CC=$ac_save_CC fi # AC_CACHE_VAL case "x$ac_cv_prog_cc_c99" in x) { $as_echo "$as_me:${as_lineno-$LINENO}: result: none needed" >&5 $as_echo "none needed" >&6; } ;; xno) { $as_echo "$as_me:${as_lineno-$LINENO}: result: unsupported" >&5 $as_echo "unsupported" >&6; } ;; *) CC="$CC $ac_cv_prog_cc_c99" { $as_echo "$as_me:${as_lineno-$LINENO}: result: $ac_cv_prog_cc_c99" >&5 $as_echo "$ac_cv_prog_cc_c99" >&6; } ;; esac if test "x$ac_cv_prog_cc_c99" != xno; then : ac_cv_prog_cc_stdc=$ac_cv_prog_cc_c99 else { $as_echo "$as_me:${as_lineno-$LINENO}: checking for $CC option to accept ISO C89" >&5 $as_echo_n "checking for $CC option to accept ISO C89... " >&6; } if ${ac_cv_prog_cc_c89+:} false; then : $as_echo_n "(cached) " >&6 else ac_cv_prog_cc_c89=no ac_save_CC=$CC cat confdefs.h - <<_ACEOF >conftest.$ac_ext /* end confdefs.h. */ #include #include struct stat; /* Most of the following tests are stolen from RCS 5.7's src/conf.sh. */ struct buf { int x; }; FILE * (*rcsopen) (struct buf *, struct stat *, int); static char *e (p, i) char **p; int i; { return p[i]; } static char *f (char * (*g) (char **, int), char **p, ...) { char *s; va_list v; va_start (v,p); s = g (p, va_arg (v,int)); va_end (v); return s; } /* OSF 4.0 Compaq cc is some sort of almost-ANSI by default. It has function prototypes and stuff, but not '\xHH' hex character constants. These don't provoke an error unfortunately, instead are silently treated as 'x'. The following induces an error, until -std is added to get proper ANSI mode. Curiously '\x00'!='x' always comes out true, for an array size at least. It's necessary to write '\x00'==0 to get something that's true only with -std. */ int osf4_cc_array ['\x00' == 0 ? 1 : -1]; /* IBM C 6 for AIX is almost-ANSI by default, but it replaces macro parameters inside strings and character constants. */ #define FOO(x) 'x' int xlc6_cc_array[FOO(a) == 'x' ? 1 : -1]; int test (int i, double x); struct s1 {int (*f) (int a);}; struct s2 {int (*f) (double a);}; int pairnames (int, char **, FILE *(*)(struct buf *, struct stat *, int), int, int); int argc; char **argv; int main () { return f (e, argv, 0) != argv[0] || f (e, argv, 1) != argv[1]; ; return 0; } _ACEOF for ac_arg in '' -qlanglvl=extc89 -qlanglvl=ansi -std \ -Ae "-Aa -D_HPUX_SOURCE" "-Xc -D__EXTENSIONS__" do CC="$ac_save_CC $ac_arg" if ac_fn_c_try_compile "$LINENO"; then : ac_cv_prog_cc_c89=$ac_arg fi rm -f core conftest.err conftest.$ac_objext test "x$ac_cv_prog_cc_c89" != "xno" && break done rm -f conftest.$ac_ext CC=$ac_save_CC fi # AC_CACHE_VAL case "x$ac_cv_prog_cc_c89" in x) { $as_echo "$as_me:${as_lineno-$LINENO}: result: none needed" >&5 $as_echo "none needed" >&6; } ;; xno) { $as_echo "$as_me:${as_lineno-$LINENO}: result: unsupported" >&5 $as_echo "unsupported" >&6; } ;; *) CC="$CC $ac_cv_prog_cc_c89" { $as_echo "$as_me:${as_lineno-$LINENO}: result: $ac_cv_prog_cc_c89" >&5 $as_echo "$ac_cv_prog_cc_c89" >&6; } ;; esac if test "x$ac_cv_prog_cc_c89" != xno; then : ac_cv_prog_cc_stdc=$ac_cv_prog_cc_c89 else ac_cv_prog_cc_stdc=no fi fi ;; esac { $as_echo "$as_me:${as_lineno-$LINENO}: checking for $CC option to accept ISO Standard C" >&5 $as_echo_n "checking for $CC option to accept ISO Standard C... " >&6; } if ${ac_cv_prog_cc_stdc+:} false; then : $as_echo_n "(cached) " >&6 fi case $ac_cv_prog_cc_stdc in #( no) : { $as_echo "$as_me:${as_lineno-$LINENO}: result: unsupported" >&5 $as_echo "unsupported" >&6; } ;; #( '') : { $as_echo "$as_me:${as_lineno-$LINENO}: result: none needed" >&5 $as_echo "none needed" >&6; } ;; #( *) : { $as_echo "$as_me:${as_lineno-$LINENO}: result: $ac_cv_prog_cc_stdc" >&5 $as_echo "$ac_cv_prog_cc_stdc" >&6; } ;; esac { $as_echo "$as_me:${as_lineno-$LINENO}: checking whether ln -s works" >&5 $as_echo_n "checking whether ln -s works... " >&6; } LN_S=$as_ln_s if test "$LN_S" = "ln -s"; then { $as_echo "$as_me:${as_lineno-$LINENO}: result: yes" >&5 $as_echo "yes" >&6; } else { $as_echo "$as_me:${as_lineno-$LINENO}: result: no, using $LN_S" >&5 $as_echo "no, using $LN_S" >&6; } fi { $as_echo "$as_me:${as_lineno-$LINENO}: checking whether ${MAKE-make} sets \$(MAKE)" >&5 $as_echo_n "checking whether ${MAKE-make} sets \$(MAKE)... " >&6; } set x ${MAKE-make} ac_make=`$as_echo "$2" | sed 's/+/p/g; s/[^a-zA-Z0-9_]/_/g'` if eval \${ac_cv_prog_make_${ac_make}_set+:} false; then : $as_echo_n "(cached) " >&6 else cat >conftest.make <<\_ACEOF SHELL = /bin/sh all: @echo '@@@%%%=$(MAKE)=@@@%%%' _ACEOF # GNU make sometimes prints "make[1]: Entering ...", which would confuse us. case `${MAKE-make} -f conftest.make 2>/dev/null` in *@@@%%%=?*=@@@%%%*) eval ac_cv_prog_make_${ac_make}_set=yes;; *) eval ac_cv_prog_make_${ac_make}_set=no;; esac rm -f conftest.make fi if eval test \$ac_cv_prog_make_${ac_make}_set = yes; then { $as_echo "$as_me:${as_lineno-$LINENO}: result: yes" >&5 $as_echo "yes" >&6; } SET_MAKE= else { $as_echo "$as_me:${as_lineno-$LINENO}: result: no" >&5 $as_echo "no" >&6; } SET_MAKE="MAKE=${MAKE-make}" fi enable_win32_dll=yes case $host in *-*-cygwin* | *-*-mingw* | *-*-pw32* | *-*-cegcc*) if test -n "$ac_tool_prefix"; then # Extract the first word of "${ac_tool_prefix}as", so it can be a program name with args. set dummy ${ac_tool_prefix}as; ac_word=$2 { $as_echo "$as_me:${as_lineno-$LINENO}: checking for $ac_word" >&5 $as_echo_n "checking for $ac_word... " >&6; } if ${ac_cv_prog_AS+:} false; then : $as_echo_n "(cached) " >&6 else if test -n "$AS"; then ac_cv_prog_AS="$AS" # Let the user override the test. else as_save_IFS=$IFS; IFS=$PATH_SEPARATOR for as_dir in $PATH do IFS=$as_save_IFS test -z "$as_dir" && as_dir=. for ac_exec_ext in '' $ac_executable_extensions; do if as_fn_executable_p "$as_dir/$ac_word$ac_exec_ext"; then ac_cv_prog_AS="${ac_tool_prefix}as" $as_echo "$as_me:${as_lineno-$LINENO}: found $as_dir/$ac_word$ac_exec_ext" >&5 break 2 fi done done IFS=$as_save_IFS fi fi AS=$ac_cv_prog_AS if test -n "$AS"; then { $as_echo "$as_me:${as_lineno-$LINENO}: result: $AS" >&5 $as_echo "$AS" >&6; } else { $as_echo "$as_me:${as_lineno-$LINENO}: result: no" >&5 $as_echo "no" >&6; } fi fi if test -z "$ac_cv_prog_AS"; then ac_ct_AS=$AS # Extract the first word of "as", so it can be a program name with args. set dummy as; ac_word=$2 { $as_echo "$as_me:${as_lineno-$LINENO}: checking for $ac_word" >&5 $as_echo_n "checking for $ac_word... " >&6; } if ${ac_cv_prog_ac_ct_AS+:} false; then : $as_echo_n "(cached) " >&6 else if test -n "$ac_ct_AS"; then ac_cv_prog_ac_ct_AS="$ac_ct_AS" # Let the user override the test. else as_save_IFS=$IFS; IFS=$PATH_SEPARATOR for as_dir in $PATH do IFS=$as_save_IFS test -z "$as_dir" && as_dir=. for ac_exec_ext in '' $ac_executable_extensions; do if as_fn_executable_p "$as_dir/$ac_word$ac_exec_ext"; then ac_cv_prog_ac_ct_AS="as" $as_echo "$as_me:${as_lineno-$LINENO}: found $as_dir/$ac_word$ac_exec_ext" >&5 break 2 fi done done IFS=$as_save_IFS fi fi ac_ct_AS=$ac_cv_prog_ac_ct_AS if test -n "$ac_ct_AS"; then { $as_echo "$as_me:${as_lineno-$LINENO}: result: $ac_ct_AS" >&5 $as_echo "$ac_ct_AS" >&6; } else { $as_echo "$as_me:${as_lineno-$LINENO}: result: no" >&5 $as_echo "no" >&6; } fi if test "x$ac_ct_AS" = x; then AS="false" else case $cross_compiling:$ac_tool_warned in yes:) { $as_echo "$as_me:${as_lineno-$LINENO}: WARNING: using cross tools not prefixed with host triplet" >&5 $as_echo "$as_me: WARNING: using cross tools not prefixed with host triplet" >&2;} ac_tool_warned=yes ;; esac AS=$ac_ct_AS fi else AS="$ac_cv_prog_AS" fi if test -n "$ac_tool_prefix"; then # Extract the first word of "${ac_tool_prefix}dlltool", so it can be a program name with args. set dummy ${ac_tool_prefix}dlltool; ac_word=$2 { $as_echo "$as_me:${as_lineno-$LINENO}: checking for $ac_word" >&5 $as_echo_n "checking for $ac_word... " >&6; } if ${ac_cv_prog_DLLTOOL+:} false; then : $as_echo_n "(cached) " >&6 else if test -n "$DLLTOOL"; then ac_cv_prog_DLLTOOL="$DLLTOOL" # Let the user override the test. else as_save_IFS=$IFS; IFS=$PATH_SEPARATOR for as_dir in $PATH do IFS=$as_save_IFS test -z "$as_dir" && as_dir=. for ac_exec_ext in '' $ac_executable_extensions; do if as_fn_executable_p "$as_dir/$ac_word$ac_exec_ext"; then ac_cv_prog_DLLTOOL="${ac_tool_prefix}dlltool" $as_echo "$as_me:${as_lineno-$LINENO}: found $as_dir/$ac_word$ac_exec_ext" >&5 break 2 fi done done IFS=$as_save_IFS fi fi DLLTOOL=$ac_cv_prog_DLLTOOL if test -n "$DLLTOOL"; then { $as_echo "$as_me:${as_lineno-$LINENO}: result: $DLLTOOL" >&5 $as_echo "$DLLTOOL" >&6; } else { $as_echo "$as_me:${as_lineno-$LINENO}: result: no" >&5 $as_echo "no" >&6; } fi fi if test -z "$ac_cv_prog_DLLTOOL"; then ac_ct_DLLTOOL=$DLLTOOL # Extract the first word of "dlltool", so it can be a program name with args. set dummy dlltool; ac_word=$2 { $as_echo "$as_me:${as_lineno-$LINENO}: checking for $ac_word" >&5 $as_echo_n "checking for $ac_word... " >&6; } if ${ac_cv_prog_ac_ct_DLLTOOL+:} false; then : $as_echo_n "(cached) " >&6 else if test -n "$ac_ct_DLLTOOL"; then ac_cv_prog_ac_ct_DLLTOOL="$ac_ct_DLLTOOL" # Let the user override the test. else as_save_IFS=$IFS; IFS=$PATH_SEPARATOR for as_dir in $PATH do IFS=$as_save_IFS test -z "$as_dir" && as_dir=. for ac_exec_ext in '' $ac_executable_extensions; do if as_fn_executable_p "$as_dir/$ac_word$ac_exec_ext"; then ac_cv_prog_ac_ct_DLLTOOL="dlltool" $as_echo "$as_me:${as_lineno-$LINENO}: found $as_dir/$ac_word$ac_exec_ext" >&5 break 2 fi done done IFS=$as_save_IFS fi fi ac_ct_DLLTOOL=$ac_cv_prog_ac_ct_DLLTOOL if test -n "$ac_ct_DLLTOOL"; then { $as_echo "$as_me:${as_lineno-$LINENO}: result: $ac_ct_DLLTOOL" >&5 $as_echo "$ac_ct_DLLTOOL" >&6; } else { $as_echo "$as_me:${as_lineno-$LINENO}: result: no" >&5 $as_echo "no" >&6; } fi if test "x$ac_ct_DLLTOOL" = x; then DLLTOOL="false" else case $cross_compiling:$ac_tool_warned in yes:) { $as_echo "$as_me:${as_lineno-$LINENO}: WARNING: using cross tools not prefixed with host triplet" >&5 $as_echo "$as_me: WARNING: using cross tools not prefixed with host triplet" >&2;} ac_tool_warned=yes ;; esac DLLTOOL=$ac_ct_DLLTOOL fi else DLLTOOL="$ac_cv_prog_DLLTOOL" fi if test -n "$ac_tool_prefix"; then # Extract the first word of "${ac_tool_prefix}objdump", so it can be a program name with args. set dummy ${ac_tool_prefix}objdump; ac_word=$2 { $as_echo "$as_me:${as_lineno-$LINENO}: checking for $ac_word" >&5 $as_echo_n "checking for $ac_word... " >&6; } if ${ac_cv_prog_OBJDUMP+:} false; then : $as_echo_n "(cached) " >&6 else if test -n "$OBJDUMP"; then ac_cv_prog_OBJDUMP="$OBJDUMP" # Let the user override the test. else as_save_IFS=$IFS; IFS=$PATH_SEPARATOR for as_dir in $PATH do IFS=$as_save_IFS test -z "$as_dir" && as_dir=. for ac_exec_ext in '' $ac_executable_extensions; do if as_fn_executable_p "$as_dir/$ac_word$ac_exec_ext"; then ac_cv_prog_OBJDUMP="${ac_tool_prefix}objdump" $as_echo "$as_me:${as_lineno-$LINENO}: found $as_dir/$ac_word$ac_exec_ext" >&5 break 2 fi done done IFS=$as_save_IFS fi fi OBJDUMP=$ac_cv_prog_OBJDUMP if test -n "$OBJDUMP"; then { $as_echo "$as_me:${as_lineno-$LINENO}: result: $OBJDUMP" >&5 $as_echo "$OBJDUMP" >&6; } else { $as_echo "$as_me:${as_lineno-$LINENO}: result: no" >&5 $as_echo "no" >&6; } fi fi if test -z "$ac_cv_prog_OBJDUMP"; then ac_ct_OBJDUMP=$OBJDUMP # Extract the first word of "objdump", so it can be a program name with args. set dummy objdump; ac_word=$2 { $as_echo "$as_me:${as_lineno-$LINENO}: checking for $ac_word" >&5 $as_echo_n "checking for $ac_word... " >&6; } if ${ac_cv_prog_ac_ct_OBJDUMP+:} false; then : $as_echo_n "(cached) " >&6 else if test -n "$ac_ct_OBJDUMP"; then ac_cv_prog_ac_ct_OBJDUMP="$ac_ct_OBJDUMP" # Let the user override the test. else as_save_IFS=$IFS; IFS=$PATH_SEPARATOR for as_dir in $PATH do IFS=$as_save_IFS test -z "$as_dir" && as_dir=. for ac_exec_ext in '' $ac_executable_extensions; do if as_fn_executable_p "$as_dir/$ac_word$ac_exec_ext"; then ac_cv_prog_ac_ct_OBJDUMP="objdump" $as_echo "$as_me:${as_lineno-$LINENO}: found $as_dir/$ac_word$ac_exec_ext" >&5 break 2 fi done done IFS=$as_save_IFS fi fi ac_ct_OBJDUMP=$ac_cv_prog_ac_ct_OBJDUMP if test -n "$ac_ct_OBJDUMP"; then { $as_echo "$as_me:${as_lineno-$LINENO}: result: $ac_ct_OBJDUMP" >&5 $as_echo "$ac_ct_OBJDUMP" >&6; } else { $as_echo "$as_me:${as_lineno-$LINENO}: result: no" >&5 $as_echo "no" >&6; } fi if test "x$ac_ct_OBJDUMP" = x; then OBJDUMP="false" else case $cross_compiling:$ac_tool_warned in yes:) { $as_echo "$as_me:${as_lineno-$LINENO}: WARNING: using cross tools not prefixed with host triplet" >&5 $as_echo "$as_me: WARNING: using cross tools not prefixed with host triplet" >&2;} ac_tool_warned=yes ;; esac OBJDUMP=$ac_ct_OBJDUMP fi else OBJDUMP="$ac_cv_prog_OBJDUMP" fi ;; esac test -z "$AS" && AS=as test -z "$DLLTOOL" && DLLTOOL=dlltool test -z "$OBJDUMP" && OBJDUMP=objdump case `pwd` in *\ * | *\ *) { $as_echo "$as_me:${as_lineno-$LINENO}: WARNING: Libtool does not cope well with whitespace in \`pwd\`" >&5 $as_echo "$as_me: WARNING: Libtool does not cope well with whitespace in \`pwd\`" >&2;} ;; esac macro_version='2.4.2' macro_revision='1.3337' ltmain="$ac_aux_dir/ltmain.sh" # Backslashify metacharacters that are still active within # double-quoted strings. sed_quote_subst='s/\(["`$\\]\)/\\\1/g' # Same as above, but do not quote variable references. double_quote_subst='s/\(["`\\]\)/\\\1/g' # Sed substitution to delay expansion of an escaped shell variable in a # double_quote_subst'ed string. delay_variable_subst='s/\\\\\\\\\\\$/\\\\\\$/g' # Sed substitution to delay expansion of an escaped single quote. delay_single_quote_subst='s/'\''/'\'\\\\\\\'\''/g' # Sed substitution to avoid accidental globbing in evaled expressions no_glob_subst='s/\*/\\\*/g' ECHO='\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\' ECHO=$ECHO$ECHO$ECHO$ECHO$ECHO ECHO=$ECHO$ECHO$ECHO$ECHO$ECHO$ECHO { $as_echo "$as_me:${as_lineno-$LINENO}: checking how to print strings" >&5 $as_echo_n "checking how to print strings... " >&6; } # Test print first, because it will be a builtin if present. if test "X`( print -r -- -n ) 2>/dev/null`" = X-n && \ test "X`print -r -- $ECHO 2>/dev/null`" = "X$ECHO"; then ECHO='print -r --' elif test "X`printf %s $ECHO 2>/dev/null`" = "X$ECHO"; then ECHO='printf %s\n' else # Use this function as a fallback that always works. func_fallback_echo () { eval 'cat <<_LTECHO_EOF $1 _LTECHO_EOF' } ECHO='func_fallback_echo' fi # func_echo_all arg... # Invoke $ECHO with all args, space-separated. func_echo_all () { $ECHO "" } case "$ECHO" in printf*) { $as_echo "$as_me:${as_lineno-$LINENO}: result: printf" >&5 $as_echo "printf" >&6; } ;; print*) { $as_echo "$as_me:${as_lineno-$LINENO}: result: print -r" >&5 $as_echo "print -r" >&6; } ;; *) { $as_echo "$as_me:${as_lineno-$LINENO}: result: cat" >&5 $as_echo "cat" >&6; } ;; esac { $as_echo "$as_me:${as_lineno-$LINENO}: checking for a sed that does not truncate output" >&5 $as_echo_n "checking for a sed that does not truncate output... " >&6; } if ${ac_cv_path_SED+:} false; then : $as_echo_n "(cached) " >&6 else ac_script=s/aaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaa/bbbbbbbbbbbbbbbbbbbbbbbbbbbbbbbbb/ for ac_i in 1 2 3 4 5 6 7; do ac_script="$ac_script$as_nl$ac_script" done echo "$ac_script" 2>/dev/null | sed 99q >conftest.sed { ac_script=; unset ac_script;} if test -z "$SED"; then ac_path_SED_found=false # Loop through the user's path and test for each of PROGNAME-LIST as_save_IFS=$IFS; IFS=$PATH_SEPARATOR for as_dir in $PATH do IFS=$as_save_IFS test -z "$as_dir" && as_dir=. for ac_prog in sed gsed; do for ac_exec_ext in '' $ac_executable_extensions; do ac_path_SED="$as_dir/$ac_prog$ac_exec_ext" as_fn_executable_p "$ac_path_SED" || continue # Check for GNU ac_path_SED and select it if it is found. # Check for GNU $ac_path_SED case `"$ac_path_SED" --version 2>&1` in *GNU*) ac_cv_path_SED="$ac_path_SED" ac_path_SED_found=:;; *) ac_count=0 $as_echo_n 0123456789 >"conftest.in" while : do cat "conftest.in" "conftest.in" >"conftest.tmp" mv "conftest.tmp" "conftest.in" cp "conftest.in" "conftest.nl" $as_echo '' >> "conftest.nl" "$ac_path_SED" -f conftest.sed < "conftest.nl" >"conftest.out" 2>/dev/null || break diff "conftest.out" "conftest.nl" >/dev/null 2>&1 || break as_fn_arith $ac_count + 1 && ac_count=$as_val if test $ac_count -gt ${ac_path_SED_max-0}; then # Best one so far, save it but keep looking for a better one ac_cv_path_SED="$ac_path_SED" ac_path_SED_max=$ac_count fi # 10*(2^10) chars as input seems more than enough test $ac_count -gt 10 && break done rm -f conftest.in conftest.tmp conftest.nl conftest.out;; esac $ac_path_SED_found && break 3 done done done IFS=$as_save_IFS if test -z "$ac_cv_path_SED"; then as_fn_error $? "no acceptable sed could be found in \$PATH" "$LINENO" 5 fi else ac_cv_path_SED=$SED fi fi { $as_echo "$as_me:${as_lineno-$LINENO}: result: $ac_cv_path_SED" >&5 $as_echo "$ac_cv_path_SED" >&6; } SED="$ac_cv_path_SED" rm -f conftest.sed test -z "$SED" && SED=sed Xsed="$SED -e 1s/^X//" { $as_echo "$as_me:${as_lineno-$LINENO}: checking for grep that handles long lines and -e" >&5 $as_echo_n "checking for grep that handles long lines and -e... " >&6; } if ${ac_cv_path_GREP+:} false; then : $as_echo_n "(cached) " >&6 else if test -z "$GREP"; then ac_path_GREP_found=false # Loop through the user's path and test for each of PROGNAME-LIST as_save_IFS=$IFS; IFS=$PATH_SEPARATOR for as_dir in $PATH$PATH_SEPARATOR/usr/xpg4/bin do IFS=$as_save_IFS test -z "$as_dir" && as_dir=. for ac_prog in grep ggrep; do for ac_exec_ext in '' $ac_executable_extensions; do ac_path_GREP="$as_dir/$ac_prog$ac_exec_ext" as_fn_executable_p "$ac_path_GREP" || continue # Check for GNU ac_path_GREP and select it if it is found. # Check for GNU $ac_path_GREP case `"$ac_path_GREP" --version 2>&1` in *GNU*) ac_cv_path_GREP="$ac_path_GREP" ac_path_GREP_found=:;; *) ac_count=0 $as_echo_n 0123456789 >"conftest.in" while : do cat "conftest.in" "conftest.in" >"conftest.tmp" mv "conftest.tmp" "conftest.in" cp "conftest.in" "conftest.nl" $as_echo 'GREP' >> "conftest.nl" "$ac_path_GREP" -e 'GREP$' -e '-(cannot match)-' < "conftest.nl" >"conftest.out" 2>/dev/null || break diff "conftest.out" "conftest.nl" >/dev/null 2>&1 || break as_fn_arith $ac_count + 1 && ac_count=$as_val if test $ac_count -gt ${ac_path_GREP_max-0}; then # Best one so far, save it but keep looking for a better one ac_cv_path_GREP="$ac_path_GREP" ac_path_GREP_max=$ac_count fi # 10*(2^10) chars as input seems more than enough test $ac_count -gt 10 && break done rm -f conftest.in conftest.tmp conftest.nl conftest.out;; esac $ac_path_GREP_found && break 3 done done done IFS=$as_save_IFS if test -z "$ac_cv_path_GREP"; then as_fn_error $? "no acceptable grep could be found in $PATH$PATH_SEPARATOR/usr/xpg4/bin" "$LINENO" 5 fi else ac_cv_path_GREP=$GREP fi fi { $as_echo "$as_me:${as_lineno-$LINENO}: result: $ac_cv_path_GREP" >&5 $as_echo "$ac_cv_path_GREP" >&6; } GREP="$ac_cv_path_GREP" { $as_echo "$as_me:${as_lineno-$LINENO}: checking for egrep" >&5 $as_echo_n "checking for egrep... " >&6; } if ${ac_cv_path_EGREP+:} false; then : $as_echo_n "(cached) " >&6 else if echo a | $GREP -E '(a|b)' >/dev/null 2>&1 then ac_cv_path_EGREP="$GREP -E" else if test -z "$EGREP"; then ac_path_EGREP_found=false # Loop through the user's path and test for each of PROGNAME-LIST as_save_IFS=$IFS; IFS=$PATH_SEPARATOR for as_dir in $PATH$PATH_SEPARATOR/usr/xpg4/bin do IFS=$as_save_IFS test -z "$as_dir" && as_dir=. for ac_prog in egrep; do for ac_exec_ext in '' $ac_executable_extensions; do ac_path_EGREP="$as_dir/$ac_prog$ac_exec_ext" as_fn_executable_p "$ac_path_EGREP" || continue # Check for GNU ac_path_EGREP and select it if it is found. # Check for GNU $ac_path_EGREP case `"$ac_path_EGREP" --version 2>&1` in *GNU*) ac_cv_path_EGREP="$ac_path_EGREP" ac_path_EGREP_found=:;; *) ac_count=0 $as_echo_n 0123456789 >"conftest.in" while : do cat "conftest.in" "conftest.in" >"conftest.tmp" mv "conftest.tmp" "conftest.in" cp "conftest.in" "conftest.nl" $as_echo 'EGREP' >> "conftest.nl" "$ac_path_EGREP" 'EGREP$' < "conftest.nl" >"conftest.out" 2>/dev/null || break diff "conftest.out" "conftest.nl" >/dev/null 2>&1 || break as_fn_arith $ac_count + 1 && ac_count=$as_val if test $ac_count -gt ${ac_path_EGREP_max-0}; then # Best one so far, save it but keep looking for a better one ac_cv_path_EGREP="$ac_path_EGREP" ac_path_EGREP_max=$ac_count fi # 10*(2^10) chars as input seems more than enough test $ac_count -gt 10 && break done rm -f conftest.in conftest.tmp conftest.nl conftest.out;; esac $ac_path_EGREP_found && break 3 done done done IFS=$as_save_IFS if test -z "$ac_cv_path_EGREP"; then as_fn_error $? "no acceptable egrep could be found in $PATH$PATH_SEPARATOR/usr/xpg4/bin" "$LINENO" 5 fi else ac_cv_path_EGREP=$EGREP fi fi fi { $as_echo "$as_me:${as_lineno-$LINENO}: result: $ac_cv_path_EGREP" >&5 $as_echo "$ac_cv_path_EGREP" >&6; } EGREP="$ac_cv_path_EGREP" { $as_echo "$as_me:${as_lineno-$LINENO}: checking for fgrep" >&5 $as_echo_n "checking for fgrep... " >&6; } if ${ac_cv_path_FGREP+:} false; then : $as_echo_n "(cached) " >&6 else if echo 'ab*c' | $GREP -F 'ab*c' >/dev/null 2>&1 then ac_cv_path_FGREP="$GREP -F" else if test -z "$FGREP"; then ac_path_FGREP_found=false # Loop through the user's path and test for each of PROGNAME-LIST as_save_IFS=$IFS; IFS=$PATH_SEPARATOR for as_dir in $PATH$PATH_SEPARATOR/usr/xpg4/bin do IFS=$as_save_IFS test -z "$as_dir" && as_dir=. for ac_prog in fgrep; do for ac_exec_ext in '' $ac_executable_extensions; do ac_path_FGREP="$as_dir/$ac_prog$ac_exec_ext" as_fn_executable_p "$ac_path_FGREP" || continue # Check for GNU ac_path_FGREP and select it if it is found. # Check for GNU $ac_path_FGREP case `"$ac_path_FGREP" --version 2>&1` in *GNU*) ac_cv_path_FGREP="$ac_path_FGREP" ac_path_FGREP_found=:;; *) ac_count=0 $as_echo_n 0123456789 >"conftest.in" while : do cat "conftest.in" "conftest.in" >"conftest.tmp" mv "conftest.tmp" "conftest.in" cp "conftest.in" "conftest.nl" $as_echo 'FGREP' >> "conftest.nl" "$ac_path_FGREP" FGREP < "conftest.nl" >"conftest.out" 2>/dev/null || break diff "conftest.out" "conftest.nl" >/dev/null 2>&1 || break as_fn_arith $ac_count + 1 && ac_count=$as_val if test $ac_count -gt ${ac_path_FGREP_max-0}; then # Best one so far, save it but keep looking for a better one ac_cv_path_FGREP="$ac_path_FGREP" ac_path_FGREP_max=$ac_count fi # 10*(2^10) chars as input seems more than enough test $ac_count -gt 10 && break done rm -f conftest.in conftest.tmp conftest.nl conftest.out;; esac $ac_path_FGREP_found && break 3 done done done IFS=$as_save_IFS if test -z "$ac_cv_path_FGREP"; then as_fn_error $? "no acceptable fgrep could be found in $PATH$PATH_SEPARATOR/usr/xpg4/bin" "$LINENO" 5 fi else ac_cv_path_FGREP=$FGREP fi fi fi { $as_echo "$as_me:${as_lineno-$LINENO}: result: $ac_cv_path_FGREP" >&5 $as_echo "$ac_cv_path_FGREP" >&6; } FGREP="$ac_cv_path_FGREP" test -z "$GREP" && GREP=grep # Check whether --with-gnu-ld was given. if test "${with_gnu_ld+set}" = set; then : withval=$with_gnu_ld; test "$withval" = no || with_gnu_ld=yes else with_gnu_ld=no fi ac_prog=ld if test "$GCC" = yes; then # Check if gcc -print-prog-name=ld gives a path. { $as_echo "$as_me:${as_lineno-$LINENO}: checking for ld used by $CC" >&5 $as_echo_n "checking for ld used by $CC... " >&6; } case $host in *-*-mingw*) # gcc leaves a trailing carriage return which upsets mingw ac_prog=`($CC -print-prog-name=ld) 2>&5 | tr -d '\015'` ;; *) ac_prog=`($CC -print-prog-name=ld) 2>&5` ;; esac case $ac_prog in # Accept absolute paths. [\\/]* | ?:[\\/]*) re_direlt='/[^/][^/]*/\.\./' # Canonicalize the pathname of ld ac_prog=`$ECHO "$ac_prog"| $SED 's%\\\\%/%g'` while $ECHO "$ac_prog" | $GREP "$re_direlt" > /dev/null 2>&1; do ac_prog=`$ECHO $ac_prog| $SED "s%$re_direlt%/%"` done test -z "$LD" && LD="$ac_prog" ;; "") # If it fails, then pretend we aren't using GCC. ac_prog=ld ;; *) # If it is relative, then search for the first ld in PATH. with_gnu_ld=unknown ;; esac elif test "$with_gnu_ld" = yes; then { $as_echo "$as_me:${as_lineno-$LINENO}: checking for GNU ld" >&5 $as_echo_n "checking for GNU ld... " >&6; } else { $as_echo "$as_me:${as_lineno-$LINENO}: checking for non-GNU ld" >&5 $as_echo_n "checking for non-GNU ld... " >&6; } fi if ${lt_cv_path_LD+:} false; then : $as_echo_n "(cached) " >&6 else if test -z "$LD"; then lt_save_ifs="$IFS"; IFS=$PATH_SEPARATOR for ac_dir in $PATH; do IFS="$lt_save_ifs" test -z "$ac_dir" && ac_dir=. if test -f "$ac_dir/$ac_prog" || test -f "$ac_dir/$ac_prog$ac_exeext"; then lt_cv_path_LD="$ac_dir/$ac_prog" # Check to see if the program is GNU ld. I'd rather use --version, # but apparently some variants of GNU ld only accept -v. # Break only if it was the GNU/non-GNU ld that we prefer. case `"$lt_cv_path_LD" -v 2>&1 &5 $as_echo "$LD" >&6; } else { $as_echo "$as_me:${as_lineno-$LINENO}: result: no" >&5 $as_echo "no" >&6; } fi test -z "$LD" && as_fn_error $? "no acceptable ld found in \$PATH" "$LINENO" 5 { $as_echo "$as_me:${as_lineno-$LINENO}: checking if the linker ($LD) is GNU ld" >&5 $as_echo_n "checking if the linker ($LD) is GNU ld... " >&6; } if ${lt_cv_prog_gnu_ld+:} false; then : $as_echo_n "(cached) " >&6 else # I'd rather use --version here, but apparently some GNU lds only accept -v. case `$LD -v 2>&1 &5 $as_echo "$lt_cv_prog_gnu_ld" >&6; } with_gnu_ld=$lt_cv_prog_gnu_ld { $as_echo "$as_me:${as_lineno-$LINENO}: checking for BSD- or MS-compatible name lister (nm)" >&5 $as_echo_n "checking for BSD- or MS-compatible name lister (nm)... " >&6; } if ${lt_cv_path_NM+:} false; then : $as_echo_n "(cached) " >&6 else if test -n "$NM"; then # Let the user override the test. lt_cv_path_NM="$NM" else lt_nm_to_check="${ac_tool_prefix}nm" if test -n "$ac_tool_prefix" && test "$build" = "$host"; then lt_nm_to_check="$lt_nm_to_check nm" fi for lt_tmp_nm in $lt_nm_to_check; do lt_save_ifs="$IFS"; IFS=$PATH_SEPARATOR for ac_dir in $PATH /usr/ccs/bin/elf /usr/ccs/bin /usr/ucb /bin; do IFS="$lt_save_ifs" test -z "$ac_dir" && ac_dir=. tmp_nm="$ac_dir/$lt_tmp_nm" if test -f "$tmp_nm" || test -f "$tmp_nm$ac_exeext" ; then # Check to see if the nm accepts a BSD-compat flag. # Adding the `sed 1q' prevents false positives on HP-UX, which says: # nm: unknown option "B" ignored # Tru64's nm complains that /dev/null is an invalid object file case `"$tmp_nm" -B /dev/null 2>&1 | sed '1q'` in */dev/null* | *'Invalid file or object type'*) lt_cv_path_NM="$tmp_nm -B" break ;; *) case `"$tmp_nm" -p /dev/null 2>&1 | sed '1q'` in */dev/null*) lt_cv_path_NM="$tmp_nm -p" break ;; *) lt_cv_path_NM=${lt_cv_path_NM="$tmp_nm"} # keep the first match, but continue # so that we can try to find one that supports BSD flags ;; esac ;; esac fi done IFS="$lt_save_ifs" done : ${lt_cv_path_NM=no} fi fi { $as_echo "$as_me:${as_lineno-$LINENO}: result: $lt_cv_path_NM" >&5 $as_echo "$lt_cv_path_NM" >&6; } if test "$lt_cv_path_NM" != "no"; then NM="$lt_cv_path_NM" else # Didn't find any BSD compatible name lister, look for dumpbin. if test -n "$DUMPBIN"; then : # Let the user override the test. else if test -n "$ac_tool_prefix"; then for ac_prog in dumpbin "link -dump" do # Extract the first word of "$ac_tool_prefix$ac_prog", so it can be a program name with args. set dummy $ac_tool_prefix$ac_prog; ac_word=$2 { $as_echo "$as_me:${as_lineno-$LINENO}: checking for $ac_word" >&5 $as_echo_n "checking for $ac_word... " >&6; } if ${ac_cv_prog_DUMPBIN+:} false; then : $as_echo_n "(cached) " >&6 else if test -n "$DUMPBIN"; then ac_cv_prog_DUMPBIN="$DUMPBIN" # Let the user override the test. else as_save_IFS=$IFS; IFS=$PATH_SEPARATOR for as_dir in $PATH do IFS=$as_save_IFS test -z "$as_dir" && as_dir=. for ac_exec_ext in '' $ac_executable_extensions; do if as_fn_executable_p "$as_dir/$ac_word$ac_exec_ext"; then ac_cv_prog_DUMPBIN="$ac_tool_prefix$ac_prog" $as_echo "$as_me:${as_lineno-$LINENO}: found $as_dir/$ac_word$ac_exec_ext" >&5 break 2 fi done done IFS=$as_save_IFS fi fi DUMPBIN=$ac_cv_prog_DUMPBIN if test -n "$DUMPBIN"; then { $as_echo "$as_me:${as_lineno-$LINENO}: result: $DUMPBIN" >&5 $as_echo "$DUMPBIN" >&6; } else { $as_echo "$as_me:${as_lineno-$LINENO}: result: no" >&5 $as_echo "no" >&6; } fi test -n "$DUMPBIN" && break done fi if test -z "$DUMPBIN"; then ac_ct_DUMPBIN=$DUMPBIN for ac_prog in dumpbin "link -dump" do # Extract the first word of "$ac_prog", so it can be a program name with args. set dummy $ac_prog; ac_word=$2 { $as_echo "$as_me:${as_lineno-$LINENO}: checking for $ac_word" >&5 $as_echo_n "checking for $ac_word... " >&6; } if ${ac_cv_prog_ac_ct_DUMPBIN+:} false; then : $as_echo_n "(cached) " >&6 else if test -n "$ac_ct_DUMPBIN"; then ac_cv_prog_ac_ct_DUMPBIN="$ac_ct_DUMPBIN" # Let the user override the test. else as_save_IFS=$IFS; IFS=$PATH_SEPARATOR for as_dir in $PATH do IFS=$as_save_IFS test -z "$as_dir" && as_dir=. for ac_exec_ext in '' $ac_executable_extensions; do if as_fn_executable_p "$as_dir/$ac_word$ac_exec_ext"; then ac_cv_prog_ac_ct_DUMPBIN="$ac_prog" $as_echo "$as_me:${as_lineno-$LINENO}: found $as_dir/$ac_word$ac_exec_ext" >&5 break 2 fi done done IFS=$as_save_IFS fi fi ac_ct_DUMPBIN=$ac_cv_prog_ac_ct_DUMPBIN if test -n "$ac_ct_DUMPBIN"; then { $as_echo "$as_me:${as_lineno-$LINENO}: result: $ac_ct_DUMPBIN" >&5 $as_echo "$ac_ct_DUMPBIN" >&6; } else { $as_echo "$as_me:${as_lineno-$LINENO}: result: no" >&5 $as_echo "no" >&6; } fi test -n "$ac_ct_DUMPBIN" && break done if test "x$ac_ct_DUMPBIN" = x; then DUMPBIN=":" else case $cross_compiling:$ac_tool_warned in yes:) { $as_echo "$as_me:${as_lineno-$LINENO}: WARNING: using cross tools not prefixed with host triplet" >&5 $as_echo "$as_me: WARNING: using cross tools not prefixed with host triplet" >&2;} ac_tool_warned=yes ;; esac DUMPBIN=$ac_ct_DUMPBIN fi fi case `$DUMPBIN -symbols /dev/null 2>&1 | sed '1q'` in *COFF*) DUMPBIN="$DUMPBIN -symbols" ;; *) DUMPBIN=: ;; esac fi if test "$DUMPBIN" != ":"; then NM="$DUMPBIN" fi fi test -z "$NM" && NM=nm { $as_echo "$as_me:${as_lineno-$LINENO}: checking the name lister ($NM) interface" >&5 $as_echo_n "checking the name lister ($NM) interface... " >&6; } if ${lt_cv_nm_interface+:} false; then : $as_echo_n "(cached) " >&6 else lt_cv_nm_interface="BSD nm" echo "int some_variable = 0;" > conftest.$ac_ext (eval echo "\"\$as_me:$LINENO: $ac_compile\"" >&5) (eval "$ac_compile" 2>conftest.err) cat conftest.err >&5 (eval echo "\"\$as_me:$LINENO: $NM \\\"conftest.$ac_objext\\\"\"" >&5) (eval "$NM \"conftest.$ac_objext\"" 2>conftest.err > conftest.out) cat conftest.err >&5 (eval echo "\"\$as_me:$LINENO: output\"" >&5) cat conftest.out >&5 if $GREP 'External.*some_variable' conftest.out > /dev/null; then lt_cv_nm_interface="MS dumpbin" fi rm -f conftest* fi { $as_echo "$as_me:${as_lineno-$LINENO}: result: $lt_cv_nm_interface" >&5 $as_echo "$lt_cv_nm_interface" >&6; } # find the maximum length of command line arguments { $as_echo "$as_me:${as_lineno-$LINENO}: checking the maximum length of command line arguments" >&5 $as_echo_n "checking the maximum length of command line arguments... " >&6; } if ${lt_cv_sys_max_cmd_len+:} false; then : $as_echo_n "(cached) " >&6 else i=0 teststring="ABCD" case $build_os in msdosdjgpp*) # On DJGPP, this test can blow up pretty badly due to problems in libc # (any single argument exceeding 2000 bytes causes a buffer overrun # during glob expansion). Even if it were fixed, the result of this # check would be larger than it should be. lt_cv_sys_max_cmd_len=12288; # 12K is about right ;; gnu*) # Under GNU Hurd, this test is not required because there is # no limit to the length of command line arguments. # Libtool will interpret -1 as no limit whatsoever lt_cv_sys_max_cmd_len=-1; ;; cygwin* | mingw* | cegcc*) # On Win9x/ME, this test blows up -- it succeeds, but takes # about 5 minutes as the teststring grows exponentially. # Worse, since 9x/ME are not pre-emptively multitasking, # you end up with a "frozen" computer, even though with patience # the test eventually succeeds (with a max line length of 256k). # Instead, let's just punt: use the minimum linelength reported by # all of the supported platforms: 8192 (on NT/2K/XP). lt_cv_sys_max_cmd_len=8192; ;; mint*) # On MiNT this can take a long time and run out of memory. lt_cv_sys_max_cmd_len=8192; ;; amigaos*) # On AmigaOS with pdksh, this test takes hours, literally. # So we just punt and use a minimum line length of 8192. lt_cv_sys_max_cmd_len=8192; ;; netbsd* | freebsd* | openbsd* | darwin* | dragonfly*) # This has been around since 386BSD, at least. Likely further. if test -x /sbin/sysctl; then lt_cv_sys_max_cmd_len=`/sbin/sysctl -n kern.argmax` elif test -x /usr/sbin/sysctl; then lt_cv_sys_max_cmd_len=`/usr/sbin/sysctl -n kern.argmax` else lt_cv_sys_max_cmd_len=65536 # usable default for all BSDs fi # And add a safety zone lt_cv_sys_max_cmd_len=`expr $lt_cv_sys_max_cmd_len \/ 4` lt_cv_sys_max_cmd_len=`expr $lt_cv_sys_max_cmd_len \* 3` ;; interix*) # We know the value 262144 and hardcode it with a safety zone (like BSD) lt_cv_sys_max_cmd_len=196608 ;; os2*) # The test takes a long time on OS/2. lt_cv_sys_max_cmd_len=8192 ;; osf*) # Dr. Hans Ekkehard Plesser reports seeing a kernel panic running configure # due to this test when exec_disable_arg_limit is 1 on Tru64. It is not # nice to cause kernel panics so lets avoid the loop below. # First set a reasonable default. lt_cv_sys_max_cmd_len=16384 # if test -x /sbin/sysconfig; then case `/sbin/sysconfig -q proc exec_disable_arg_limit` in *1*) lt_cv_sys_max_cmd_len=-1 ;; esac fi ;; sco3.2v5*) lt_cv_sys_max_cmd_len=102400 ;; sysv5* | sco5v6* | sysv4.2uw2*) kargmax=`grep ARG_MAX /etc/conf/cf.d/stune 2>/dev/null` if test -n "$kargmax"; then lt_cv_sys_max_cmd_len=`echo $kargmax | sed 's/.*[ ]//'` else lt_cv_sys_max_cmd_len=32768 fi ;; *) lt_cv_sys_max_cmd_len=`(getconf ARG_MAX) 2> /dev/null` if test -n "$lt_cv_sys_max_cmd_len"; then lt_cv_sys_max_cmd_len=`expr $lt_cv_sys_max_cmd_len \/ 4` lt_cv_sys_max_cmd_len=`expr $lt_cv_sys_max_cmd_len \* 3` else # Make teststring a little bigger before we do anything with it. # a 1K string should be a reasonable start. for i in 1 2 3 4 5 6 7 8 ; do teststring=$teststring$teststring done SHELL=${SHELL-${CONFIG_SHELL-/bin/sh}} # If test is not a shell built-in, we'll probably end up computing a # maximum length that is only half of the actual maximum length, but # we can't tell. while { test "X"`env echo "$teststring$teststring" 2>/dev/null` \ = "X$teststring$teststring"; } >/dev/null 2>&1 && test $i != 17 # 1/2 MB should be enough do i=`expr $i + 1` teststring=$teststring$teststring done # Only check the string length outside the loop. lt_cv_sys_max_cmd_len=`expr "X$teststring" : ".*" 2>&1` teststring= # Add a significant safety factor because C++ compilers can tack on # massive amounts of additional arguments before passing them to the # linker. It appears as though 1/2 is a usable value. lt_cv_sys_max_cmd_len=`expr $lt_cv_sys_max_cmd_len \/ 2` fi ;; esac fi if test -n $lt_cv_sys_max_cmd_len ; then { $as_echo "$as_me:${as_lineno-$LINENO}: result: $lt_cv_sys_max_cmd_len" >&5 $as_echo "$lt_cv_sys_max_cmd_len" >&6; } else { $as_echo "$as_me:${as_lineno-$LINENO}: result: none" >&5 $as_echo "none" >&6; } fi max_cmd_len=$lt_cv_sys_max_cmd_len : ${CP="cp -f"} : ${MV="mv -f"} : ${RM="rm -f"} { $as_echo "$as_me:${as_lineno-$LINENO}: checking whether the shell understands some XSI constructs" >&5 $as_echo_n "checking whether the shell understands some XSI constructs... " >&6; } # Try some XSI features xsi_shell=no ( _lt_dummy="a/b/c" test "${_lt_dummy##*/},${_lt_dummy%/*},${_lt_dummy#??}"${_lt_dummy%"$_lt_dummy"}, \ = c,a/b,b/c, \ && eval 'test $(( 1 + 1 )) -eq 2 \ && test "${#_lt_dummy}" -eq 5' ) >/dev/null 2>&1 \ && xsi_shell=yes { $as_echo "$as_me:${as_lineno-$LINENO}: result: $xsi_shell" >&5 $as_echo "$xsi_shell" >&6; } { $as_echo "$as_me:${as_lineno-$LINENO}: checking whether the shell understands \"+=\"" >&5 $as_echo_n "checking whether the shell understands \"+=\"... " >&6; } lt_shell_append=no ( foo=bar; set foo baz; eval "$1+=\$2" && test "$foo" = barbaz ) \ >/dev/null 2>&1 \ && lt_shell_append=yes { $as_echo "$as_me:${as_lineno-$LINENO}: result: $lt_shell_append" >&5 $as_echo "$lt_shell_append" >&6; } if ( (MAIL=60; unset MAIL) || exit) >/dev/null 2>&1; then lt_unset=unset else lt_unset=false fi # test EBCDIC or ASCII case `echo X|tr X '\101'` in A) # ASCII based system # \n is not interpreted correctly by Solaris 8 /usr/ucb/tr lt_SP2NL='tr \040 \012' lt_NL2SP='tr \015\012 \040\040' ;; *) # EBCDIC based system lt_SP2NL='tr \100 \n' lt_NL2SP='tr \r\n \100\100' ;; esac { $as_echo "$as_me:${as_lineno-$LINENO}: checking how to convert $build file names to $host format" >&5 $as_echo_n "checking how to convert $build file names to $host format... " >&6; } if ${lt_cv_to_host_file_cmd+:} false; then : $as_echo_n "(cached) " >&6 else case $host in *-*-mingw* ) case $build in *-*-mingw* ) # actually msys lt_cv_to_host_file_cmd=func_convert_file_msys_to_w32 ;; *-*-cygwin* ) lt_cv_to_host_file_cmd=func_convert_file_cygwin_to_w32 ;; * ) # otherwise, assume *nix lt_cv_to_host_file_cmd=func_convert_file_nix_to_w32 ;; esac ;; *-*-cygwin* ) case $build in *-*-mingw* ) # actually msys lt_cv_to_host_file_cmd=func_convert_file_msys_to_cygwin ;; *-*-cygwin* ) lt_cv_to_host_file_cmd=func_convert_file_noop ;; * ) # otherwise, assume *nix lt_cv_to_host_file_cmd=func_convert_file_nix_to_cygwin ;; esac ;; * ) # unhandled hosts (and "normal" native builds) lt_cv_to_host_file_cmd=func_convert_file_noop ;; esac fi to_host_file_cmd=$lt_cv_to_host_file_cmd { $as_echo "$as_me:${as_lineno-$LINENO}: result: $lt_cv_to_host_file_cmd" >&5 $as_echo "$lt_cv_to_host_file_cmd" >&6; } { $as_echo "$as_me:${as_lineno-$LINENO}: checking how to convert $build file names to toolchain format" >&5 $as_echo_n "checking how to convert $build file names to toolchain format... " >&6; } if ${lt_cv_to_tool_file_cmd+:} false; then : $as_echo_n "(cached) " >&6 else #assume ordinary cross tools, or native build. lt_cv_to_tool_file_cmd=func_convert_file_noop case $host in *-*-mingw* ) case $build in *-*-mingw* ) # actually msys lt_cv_to_tool_file_cmd=func_convert_file_msys_to_w32 ;; esac ;; esac fi to_tool_file_cmd=$lt_cv_to_tool_file_cmd { $as_echo "$as_me:${as_lineno-$LINENO}: result: $lt_cv_to_tool_file_cmd" >&5 $as_echo "$lt_cv_to_tool_file_cmd" >&6; } { $as_echo "$as_me:${as_lineno-$LINENO}: checking for $LD option to reload object files" >&5 $as_echo_n "checking for $LD option to reload object files... " >&6; } if ${lt_cv_ld_reload_flag+:} false; then : $as_echo_n "(cached) " >&6 else lt_cv_ld_reload_flag='-r' fi { $as_echo "$as_me:${as_lineno-$LINENO}: result: $lt_cv_ld_reload_flag" >&5 $as_echo "$lt_cv_ld_reload_flag" >&6; } reload_flag=$lt_cv_ld_reload_flag case $reload_flag in "" | " "*) ;; *) reload_flag=" $reload_flag" ;; esac reload_cmds='$LD$reload_flag -o $output$reload_objs' case $host_os in cygwin* | mingw* | pw32* | cegcc*) if test "$GCC" != yes; then reload_cmds=false fi ;; darwin*) if test "$GCC" = yes; then reload_cmds='$LTCC $LTCFLAGS -nostdlib ${wl}-r -o $output$reload_objs' else reload_cmds='$LD$reload_flag -o $output$reload_objs' fi ;; esac if test -n "$ac_tool_prefix"; then # Extract the first word of "${ac_tool_prefix}objdump", so it can be a program name with args. set dummy ${ac_tool_prefix}objdump; ac_word=$2 { $as_echo "$as_me:${as_lineno-$LINENO}: checking for $ac_word" >&5 $as_echo_n "checking for $ac_word... " >&6; } if ${ac_cv_prog_OBJDUMP+:} false; then : $as_echo_n "(cached) " >&6 else if test -n "$OBJDUMP"; then ac_cv_prog_OBJDUMP="$OBJDUMP" # Let the user override the test. else as_save_IFS=$IFS; IFS=$PATH_SEPARATOR for as_dir in $PATH do IFS=$as_save_IFS test -z "$as_dir" && as_dir=. for ac_exec_ext in '' $ac_executable_extensions; do if as_fn_executable_p "$as_dir/$ac_word$ac_exec_ext"; then ac_cv_prog_OBJDUMP="${ac_tool_prefix}objdump" $as_echo "$as_me:${as_lineno-$LINENO}: found $as_dir/$ac_word$ac_exec_ext" >&5 break 2 fi done done IFS=$as_save_IFS fi fi OBJDUMP=$ac_cv_prog_OBJDUMP if test -n "$OBJDUMP"; then { $as_echo "$as_me:${as_lineno-$LINENO}: result: $OBJDUMP" >&5 $as_echo "$OBJDUMP" >&6; } else { $as_echo "$as_me:${as_lineno-$LINENO}: result: no" >&5 $as_echo "no" >&6; } fi fi if test -z "$ac_cv_prog_OBJDUMP"; then ac_ct_OBJDUMP=$OBJDUMP # Extract the first word of "objdump", so it can be a program name with args. set dummy objdump; ac_word=$2 { $as_echo "$as_me:${as_lineno-$LINENO}: checking for $ac_word" >&5 $as_echo_n "checking for $ac_word... " >&6; } if ${ac_cv_prog_ac_ct_OBJDUMP+:} false; then : $as_echo_n "(cached) " >&6 else if test -n "$ac_ct_OBJDUMP"; then ac_cv_prog_ac_ct_OBJDUMP="$ac_ct_OBJDUMP" # Let the user override the test. else as_save_IFS=$IFS; IFS=$PATH_SEPARATOR for as_dir in $PATH do IFS=$as_save_IFS test -z "$as_dir" && as_dir=. for ac_exec_ext in '' $ac_executable_extensions; do if as_fn_executable_p "$as_dir/$ac_word$ac_exec_ext"; then ac_cv_prog_ac_ct_OBJDUMP="objdump" $as_echo "$as_me:${as_lineno-$LINENO}: found $as_dir/$ac_word$ac_exec_ext" >&5 break 2 fi done done IFS=$as_save_IFS fi fi ac_ct_OBJDUMP=$ac_cv_prog_ac_ct_OBJDUMP if test -n "$ac_ct_OBJDUMP"; then { $as_echo "$as_me:${as_lineno-$LINENO}: result: $ac_ct_OBJDUMP" >&5 $as_echo "$ac_ct_OBJDUMP" >&6; } else { $as_echo "$as_me:${as_lineno-$LINENO}: result: no" >&5 $as_echo "no" >&6; } fi if test "x$ac_ct_OBJDUMP" = x; then OBJDUMP="false" else case $cross_compiling:$ac_tool_warned in yes:) { $as_echo "$as_me:${as_lineno-$LINENO}: WARNING: using cross tools not prefixed with host triplet" >&5 $as_echo "$as_me: WARNING: using cross tools not prefixed with host triplet" >&2;} ac_tool_warned=yes ;; esac OBJDUMP=$ac_ct_OBJDUMP fi else OBJDUMP="$ac_cv_prog_OBJDUMP" fi test -z "$OBJDUMP" && OBJDUMP=objdump { $as_echo "$as_me:${as_lineno-$LINENO}: checking how to recognize dependent libraries" >&5 $as_echo_n "checking how to recognize dependent libraries... " >&6; } if ${lt_cv_deplibs_check_method+:} false; then : $as_echo_n "(cached) " >&6 else lt_cv_file_magic_cmd='$MAGIC_CMD' lt_cv_file_magic_test_file= lt_cv_deplibs_check_method='unknown' # Need to set the preceding variable on all platforms that support # interlibrary dependencies. # 'none' -- dependencies not supported. # `unknown' -- same as none, but documents that we really don't know. # 'pass_all' -- all dependencies passed with no checks. # 'test_compile' -- check by making test program. # 'file_magic [[regex]]' -- check by looking for files in library path # which responds to the $file_magic_cmd with a given extended regex. # If you have `file' or equivalent on your system and you're not sure # whether `pass_all' will *always* work, you probably want this one. case $host_os in aix[4-9]*) lt_cv_deplibs_check_method=pass_all ;; beos*) lt_cv_deplibs_check_method=pass_all ;; bsdi[45]*) lt_cv_deplibs_check_method='file_magic ELF [0-9][0-9]*-bit [ML]SB (shared object|dynamic lib)' lt_cv_file_magic_cmd='/usr/bin/file -L' lt_cv_file_magic_test_file=/shlib/libc.so ;; cygwin*) # func_win32_libid is a shell function defined in ltmain.sh lt_cv_deplibs_check_method='file_magic ^x86 archive import|^x86 DLL' lt_cv_file_magic_cmd='func_win32_libid' ;; mingw* | pw32*) # Base MSYS/MinGW do not provide the 'file' command needed by # func_win32_libid shell function, so use a weaker test based on 'objdump', # unless we find 'file', for example because we are cross-compiling. # func_win32_libid assumes BSD nm, so disallow it if using MS dumpbin. if ( test "$lt_cv_nm_interface" = "BSD nm" && file / ) >/dev/null 2>&1; then lt_cv_deplibs_check_method='file_magic ^x86 archive import|^x86 DLL' lt_cv_file_magic_cmd='func_win32_libid' else # Keep this pattern in sync with the one in func_win32_libid. lt_cv_deplibs_check_method='file_magic file format (pei*-i386(.*architecture: i386)?|pe-arm-wince|pe-x86-64)' lt_cv_file_magic_cmd='$OBJDUMP -f' fi ;; cegcc*) # use the weaker test based on 'objdump'. See mingw*. lt_cv_deplibs_check_method='file_magic file format pe-arm-.*little(.*architecture: arm)?' lt_cv_file_magic_cmd='$OBJDUMP -f' ;; darwin* | rhapsody*) lt_cv_deplibs_check_method=pass_all ;; freebsd* | dragonfly*) if echo __ELF__ | $CC -E - | $GREP __ELF__ > /dev/null; then case $host_cpu in i*86 ) # Not sure whether the presence of OpenBSD here was a mistake. # Let's accept both of them until this is cleared up. lt_cv_deplibs_check_method='file_magic (FreeBSD|OpenBSD|DragonFly)/i[3-9]86 (compact )?demand paged shared library' lt_cv_file_magic_cmd=/usr/bin/file lt_cv_file_magic_test_file=`echo /usr/lib/libc.so.*` ;; esac else lt_cv_deplibs_check_method=pass_all fi ;; haiku*) lt_cv_deplibs_check_method=pass_all ;; hpux10.20* | hpux11*) lt_cv_file_magic_cmd=/usr/bin/file case $host_cpu in ia64*) lt_cv_deplibs_check_method='file_magic (s[0-9][0-9][0-9]|ELF-[0-9][0-9]) shared object file - IA64' lt_cv_file_magic_test_file=/usr/lib/hpux32/libc.so ;; hppa*64*) lt_cv_deplibs_check_method='file_magic (s[0-9][0-9][0-9]|ELF[ -][0-9][0-9])(-bit)?( [LM]SB)? shared object( file)?[, -]* PA-RISC [0-9]\.[0-9]' lt_cv_file_magic_test_file=/usr/lib/pa20_64/libc.sl ;; *) lt_cv_deplibs_check_method='file_magic (s[0-9][0-9][0-9]|PA-RISC[0-9]\.[0-9]) shared library' lt_cv_file_magic_test_file=/usr/lib/libc.sl ;; esac ;; interix[3-9]*) # PIC code is broken on Interix 3.x, that's why |\.a not |_pic\.a here lt_cv_deplibs_check_method='match_pattern /lib[^/]+(\.so|\.a)$' ;; irix5* | irix6* | nonstopux*) case $LD in *-32|*"-32 ") libmagic=32-bit;; *-n32|*"-n32 ") libmagic=N32;; *-64|*"-64 ") libmagic=64-bit;; *) libmagic=never-match;; esac lt_cv_deplibs_check_method=pass_all ;; # This must be glibc/ELF. linux* | k*bsd*-gnu | kopensolaris*-gnu | gnu*) lt_cv_deplibs_check_method=pass_all ;; netbsd* | netbsdelf*-gnu) if echo __ELF__ | $CC -E - | $GREP __ELF__ > /dev/null; then lt_cv_deplibs_check_method='match_pattern /lib[^/]+(\.so\.[0-9]+\.[0-9]+|_pic\.a)$' else lt_cv_deplibs_check_method='match_pattern /lib[^/]+(\.so|_pic\.a)$' fi ;; newos6*) lt_cv_deplibs_check_method='file_magic ELF [0-9][0-9]*-bit [ML]SB (executable|dynamic lib)' lt_cv_file_magic_cmd=/usr/bin/file lt_cv_file_magic_test_file=/usr/lib/libnls.so ;; *nto* | *qnx*) lt_cv_deplibs_check_method=pass_all ;; openbsd*) if test -z "`echo __ELF__ | $CC -E - | $GREP __ELF__`" || test "$host_os-$host_cpu" = "openbsd2.8-powerpc"; then lt_cv_deplibs_check_method='match_pattern /lib[^/]+(\.so\.[0-9]+\.[0-9]+|\.so|_pic\.a)$' else lt_cv_deplibs_check_method='match_pattern /lib[^/]+(\.so\.[0-9]+\.[0-9]+|_pic\.a)$' fi ;; osf3* | osf4* | osf5*) lt_cv_deplibs_check_method=pass_all ;; rdos*) lt_cv_deplibs_check_method=pass_all ;; solaris*) lt_cv_deplibs_check_method=pass_all ;; sysv5* | sco3.2v5* | sco5v6* | unixware* | OpenUNIX* | sysv4*uw2*) lt_cv_deplibs_check_method=pass_all ;; sysv4 | sysv4.3*) case $host_vendor in motorola) lt_cv_deplibs_check_method='file_magic ELF [0-9][0-9]*-bit [ML]SB (shared object|dynamic lib) M[0-9][0-9]* Version [0-9]' lt_cv_file_magic_test_file=`echo /usr/lib/libc.so*` ;; ncr) lt_cv_deplibs_check_method=pass_all ;; sequent) lt_cv_file_magic_cmd='/bin/file' lt_cv_deplibs_check_method='file_magic ELF [0-9][0-9]*-bit [LM]SB (shared object|dynamic lib )' ;; sni) lt_cv_file_magic_cmd='/bin/file' lt_cv_deplibs_check_method="file_magic ELF [0-9][0-9]*-bit [LM]SB dynamic lib" lt_cv_file_magic_test_file=/lib/libc.so ;; siemens) lt_cv_deplibs_check_method=pass_all ;; pc) lt_cv_deplibs_check_method=pass_all ;; esac ;; tpf*) lt_cv_deplibs_check_method=pass_all ;; esac fi { $as_echo "$as_me:${as_lineno-$LINENO}: result: $lt_cv_deplibs_check_method" >&5 $as_echo "$lt_cv_deplibs_check_method" >&6; } file_magic_glob= want_nocaseglob=no if test "$build" = "$host"; then case $host_os in mingw* | pw32*) if ( shopt | grep nocaseglob ) >/dev/null 2>&1; then want_nocaseglob=yes else file_magic_glob=`echo aAbBcCdDeEfFgGhHiIjJkKlLmMnNoOpPqQrRsStTuUvVwWxXyYzZ | $SED -e "s/\(..\)/s\/[\1]\/[\1]\/g;/g"` fi ;; esac fi file_magic_cmd=$lt_cv_file_magic_cmd deplibs_check_method=$lt_cv_deplibs_check_method test -z "$deplibs_check_method" && deplibs_check_method=unknown if test -n "$ac_tool_prefix"; then # Extract the first word of "${ac_tool_prefix}dlltool", so it can be a program name with args. set dummy ${ac_tool_prefix}dlltool; ac_word=$2 { $as_echo "$as_me:${as_lineno-$LINENO}: checking for $ac_word" >&5 $as_echo_n "checking for $ac_word... " >&6; } if ${ac_cv_prog_DLLTOOL+:} false; then : $as_echo_n "(cached) " >&6 else if test -n "$DLLTOOL"; then ac_cv_prog_DLLTOOL="$DLLTOOL" # Let the user override the test. else as_save_IFS=$IFS; IFS=$PATH_SEPARATOR for as_dir in $PATH do IFS=$as_save_IFS test -z "$as_dir" && as_dir=. for ac_exec_ext in '' $ac_executable_extensions; do if as_fn_executable_p "$as_dir/$ac_word$ac_exec_ext"; then ac_cv_prog_DLLTOOL="${ac_tool_prefix}dlltool" $as_echo "$as_me:${as_lineno-$LINENO}: found $as_dir/$ac_word$ac_exec_ext" >&5 break 2 fi done done IFS=$as_save_IFS fi fi DLLTOOL=$ac_cv_prog_DLLTOOL if test -n "$DLLTOOL"; then { $as_echo "$as_me:${as_lineno-$LINENO}: result: $DLLTOOL" >&5 $as_echo "$DLLTOOL" >&6; } else { $as_echo "$as_me:${as_lineno-$LINENO}: result: no" >&5 $as_echo "no" >&6; } fi fi if test -z "$ac_cv_prog_DLLTOOL"; then ac_ct_DLLTOOL=$DLLTOOL # Extract the first word of "dlltool", so it can be a program name with args. set dummy dlltool; ac_word=$2 { $as_echo "$as_me:${as_lineno-$LINENO}: checking for $ac_word" >&5 $as_echo_n "checking for $ac_word... " >&6; } if ${ac_cv_prog_ac_ct_DLLTOOL+:} false; then : $as_echo_n "(cached) " >&6 else if test -n "$ac_ct_DLLTOOL"; then ac_cv_prog_ac_ct_DLLTOOL="$ac_ct_DLLTOOL" # Let the user override the test. else as_save_IFS=$IFS; IFS=$PATH_SEPARATOR for as_dir in $PATH do IFS=$as_save_IFS test -z "$as_dir" && as_dir=. for ac_exec_ext in '' $ac_executable_extensions; do if as_fn_executable_p "$as_dir/$ac_word$ac_exec_ext"; then ac_cv_prog_ac_ct_DLLTOOL="dlltool" $as_echo "$as_me:${as_lineno-$LINENO}: found $as_dir/$ac_word$ac_exec_ext" >&5 break 2 fi done done IFS=$as_save_IFS fi fi ac_ct_DLLTOOL=$ac_cv_prog_ac_ct_DLLTOOL if test -n "$ac_ct_DLLTOOL"; then { $as_echo "$as_me:${as_lineno-$LINENO}: result: $ac_ct_DLLTOOL" >&5 $as_echo "$ac_ct_DLLTOOL" >&6; } else { $as_echo "$as_me:${as_lineno-$LINENO}: result: no" >&5 $as_echo "no" >&6; } fi if test "x$ac_ct_DLLTOOL" = x; then DLLTOOL="false" else case $cross_compiling:$ac_tool_warned in yes:) { $as_echo "$as_me:${as_lineno-$LINENO}: WARNING: using cross tools not prefixed with host triplet" >&5 $as_echo "$as_me: WARNING: using cross tools not prefixed with host triplet" >&2;} ac_tool_warned=yes ;; esac DLLTOOL=$ac_ct_DLLTOOL fi else DLLTOOL="$ac_cv_prog_DLLTOOL" fi test -z "$DLLTOOL" && DLLTOOL=dlltool { $as_echo "$as_me:${as_lineno-$LINENO}: checking how to associate runtime and link libraries" >&5 $as_echo_n "checking how to associate runtime and link libraries... " >&6; } if ${lt_cv_sharedlib_from_linklib_cmd+:} false; then : $as_echo_n "(cached) " >&6 else lt_cv_sharedlib_from_linklib_cmd='unknown' case $host_os in cygwin* | mingw* | pw32* | cegcc*) # two different shell functions defined in ltmain.sh # decide which to use based on capabilities of $DLLTOOL case `$DLLTOOL --help 2>&1` in *--identify-strict*) lt_cv_sharedlib_from_linklib_cmd=func_cygming_dll_for_implib ;; *) lt_cv_sharedlib_from_linklib_cmd=func_cygming_dll_for_implib_fallback ;; esac ;; *) # fallback: assume linklib IS sharedlib lt_cv_sharedlib_from_linklib_cmd="$ECHO" ;; esac fi { $as_echo "$as_me:${as_lineno-$LINENO}: result: $lt_cv_sharedlib_from_linklib_cmd" >&5 $as_echo "$lt_cv_sharedlib_from_linklib_cmd" >&6; } sharedlib_from_linklib_cmd=$lt_cv_sharedlib_from_linklib_cmd test -z "$sharedlib_from_linklib_cmd" && sharedlib_from_linklib_cmd=$ECHO if test -n "$ac_tool_prefix"; then for ac_prog in ar do # Extract the first word of "$ac_tool_prefix$ac_prog", so it can be a program name with args. set dummy $ac_tool_prefix$ac_prog; ac_word=$2 { $as_echo "$as_me:${as_lineno-$LINENO}: checking for $ac_word" >&5 $as_echo_n "checking for $ac_word... " >&6; } if ${ac_cv_prog_AR+:} false; then : $as_echo_n "(cached) " >&6 else if test -n "$AR"; then ac_cv_prog_AR="$AR" # Let the user override the test. else as_save_IFS=$IFS; IFS=$PATH_SEPARATOR for as_dir in $PATH do IFS=$as_save_IFS test -z "$as_dir" && as_dir=. for ac_exec_ext in '' $ac_executable_extensions; do if as_fn_executable_p "$as_dir/$ac_word$ac_exec_ext"; then ac_cv_prog_AR="$ac_tool_prefix$ac_prog" $as_echo "$as_me:${as_lineno-$LINENO}: found $as_dir/$ac_word$ac_exec_ext" >&5 break 2 fi done done IFS=$as_save_IFS fi fi AR=$ac_cv_prog_AR if test -n "$AR"; then { $as_echo "$as_me:${as_lineno-$LINENO}: result: $AR" >&5 $as_echo "$AR" >&6; } else { $as_echo "$as_me:${as_lineno-$LINENO}: result: no" >&5 $as_echo "no" >&6; } fi test -n "$AR" && break done fi if test -z "$AR"; then ac_ct_AR=$AR for ac_prog in ar do # Extract the first word of "$ac_prog", so it can be a program name with args. set dummy $ac_prog; ac_word=$2 { $as_echo "$as_me:${as_lineno-$LINENO}: checking for $ac_word" >&5 $as_echo_n "checking for $ac_word... " >&6; } if ${ac_cv_prog_ac_ct_AR+:} false; then : $as_echo_n "(cached) " >&6 else if test -n "$ac_ct_AR"; then ac_cv_prog_ac_ct_AR="$ac_ct_AR" # Let the user override the test. else as_save_IFS=$IFS; IFS=$PATH_SEPARATOR for as_dir in $PATH do IFS=$as_save_IFS test -z "$as_dir" && as_dir=. for ac_exec_ext in '' $ac_executable_extensions; do if as_fn_executable_p "$as_dir/$ac_word$ac_exec_ext"; then ac_cv_prog_ac_ct_AR="$ac_prog" $as_echo "$as_me:${as_lineno-$LINENO}: found $as_dir/$ac_word$ac_exec_ext" >&5 break 2 fi done done IFS=$as_save_IFS fi fi ac_ct_AR=$ac_cv_prog_ac_ct_AR if test -n "$ac_ct_AR"; then { $as_echo "$as_me:${as_lineno-$LINENO}: result: $ac_ct_AR" >&5 $as_echo "$ac_ct_AR" >&6; } else { $as_echo "$as_me:${as_lineno-$LINENO}: result: no" >&5 $as_echo "no" >&6; } fi test -n "$ac_ct_AR" && break done if test "x$ac_ct_AR" = x; then AR="false" else case $cross_compiling:$ac_tool_warned in yes:) { $as_echo "$as_me:${as_lineno-$LINENO}: WARNING: using cross tools not prefixed with host triplet" >&5 $as_echo "$as_me: WARNING: using cross tools not prefixed with host triplet" >&2;} ac_tool_warned=yes ;; esac AR=$ac_ct_AR fi fi : ${AR=ar} : ${AR_FLAGS=cru} { $as_echo "$as_me:${as_lineno-$LINENO}: checking for archiver @FILE support" >&5 $as_echo_n "checking for archiver @FILE support... " >&6; } if ${lt_cv_ar_at_file+:} false; then : $as_echo_n "(cached) " >&6 else lt_cv_ar_at_file=no cat confdefs.h - <<_ACEOF >conftest.$ac_ext /* end confdefs.h. */ int main () { ; return 0; } _ACEOF if ac_fn_c_try_compile "$LINENO"; then : echo conftest.$ac_objext > conftest.lst lt_ar_try='$AR $AR_FLAGS libconftest.a @conftest.lst >&5' { { eval echo "\"\$as_me\":${as_lineno-$LINENO}: \"$lt_ar_try\""; } >&5 (eval $lt_ar_try) 2>&5 ac_status=$? $as_echo "$as_me:${as_lineno-$LINENO}: \$? = $ac_status" >&5 test $ac_status = 0; } if test "$ac_status" -eq 0; then # Ensure the archiver fails upon bogus file names. rm -f conftest.$ac_objext libconftest.a { { eval echo "\"\$as_me\":${as_lineno-$LINENO}: \"$lt_ar_try\""; } >&5 (eval $lt_ar_try) 2>&5 ac_status=$? $as_echo "$as_me:${as_lineno-$LINENO}: \$? = $ac_status" >&5 test $ac_status = 0; } if test "$ac_status" -ne 0; then lt_cv_ar_at_file=@ fi fi rm -f conftest.* libconftest.a fi rm -f core conftest.err conftest.$ac_objext conftest.$ac_ext fi { $as_echo "$as_me:${as_lineno-$LINENO}: result: $lt_cv_ar_at_file" >&5 $as_echo "$lt_cv_ar_at_file" >&6; } if test "x$lt_cv_ar_at_file" = xno; then archiver_list_spec= else archiver_list_spec=$lt_cv_ar_at_file fi if test -n "$ac_tool_prefix"; then # Extract the first word of "${ac_tool_prefix}strip", so it can be a program name with args. set dummy ${ac_tool_prefix}strip; ac_word=$2 { $as_echo "$as_me:${as_lineno-$LINENO}: checking for $ac_word" >&5 $as_echo_n "checking for $ac_word... " >&6; } if ${ac_cv_prog_STRIP+:} false; then : $as_echo_n "(cached) " >&6 else if test -n "$STRIP"; then ac_cv_prog_STRIP="$STRIP" # Let the user override the test. else as_save_IFS=$IFS; IFS=$PATH_SEPARATOR for as_dir in $PATH do IFS=$as_save_IFS test -z "$as_dir" && as_dir=. for ac_exec_ext in '' $ac_executable_extensions; do if as_fn_executable_p "$as_dir/$ac_word$ac_exec_ext"; then ac_cv_prog_STRIP="${ac_tool_prefix}strip" $as_echo "$as_me:${as_lineno-$LINENO}: found $as_dir/$ac_word$ac_exec_ext" >&5 break 2 fi done done IFS=$as_save_IFS fi fi STRIP=$ac_cv_prog_STRIP if test -n "$STRIP"; then { $as_echo "$as_me:${as_lineno-$LINENO}: result: $STRIP" >&5 $as_echo "$STRIP" >&6; } else { $as_echo "$as_me:${as_lineno-$LINENO}: result: no" >&5 $as_echo "no" >&6; } fi fi if test -z "$ac_cv_prog_STRIP"; then ac_ct_STRIP=$STRIP # Extract the first word of "strip", so it can be a program name with args. set dummy strip; ac_word=$2 { $as_echo "$as_me:${as_lineno-$LINENO}: checking for $ac_word" >&5 $as_echo_n "checking for $ac_word... " >&6; } if ${ac_cv_prog_ac_ct_STRIP+:} false; then : $as_echo_n "(cached) " >&6 else if test -n "$ac_ct_STRIP"; then ac_cv_prog_ac_ct_STRIP="$ac_ct_STRIP" # Let the user override the test. else as_save_IFS=$IFS; IFS=$PATH_SEPARATOR for as_dir in $PATH do IFS=$as_save_IFS test -z "$as_dir" && as_dir=. for ac_exec_ext in '' $ac_executable_extensions; do if as_fn_executable_p "$as_dir/$ac_word$ac_exec_ext"; then ac_cv_prog_ac_ct_STRIP="strip" $as_echo "$as_me:${as_lineno-$LINENO}: found $as_dir/$ac_word$ac_exec_ext" >&5 break 2 fi done done IFS=$as_save_IFS fi fi ac_ct_STRIP=$ac_cv_prog_ac_ct_STRIP if test -n "$ac_ct_STRIP"; then { $as_echo "$as_me:${as_lineno-$LINENO}: result: $ac_ct_STRIP" >&5 $as_echo "$ac_ct_STRIP" >&6; } else { $as_echo "$as_me:${as_lineno-$LINENO}: result: no" >&5 $as_echo "no" >&6; } fi if test "x$ac_ct_STRIP" = x; then STRIP=":" else case $cross_compiling:$ac_tool_warned in yes:) { $as_echo "$as_me:${as_lineno-$LINENO}: WARNING: using cross tools not prefixed with host triplet" >&5 $as_echo "$as_me: WARNING: using cross tools not prefixed with host triplet" >&2;} ac_tool_warned=yes ;; esac STRIP=$ac_ct_STRIP fi else STRIP="$ac_cv_prog_STRIP" fi test -z "$STRIP" && STRIP=: if test -n "$ac_tool_prefix"; then # Extract the first word of "${ac_tool_prefix}ranlib", so it can be a program name with args. set dummy ${ac_tool_prefix}ranlib; ac_word=$2 { $as_echo "$as_me:${as_lineno-$LINENO}: checking for $ac_word" >&5 $as_echo_n "checking for $ac_word... " >&6; } if ${ac_cv_prog_RANLIB+:} false; then : $as_echo_n "(cached) " >&6 else if test -n "$RANLIB"; then ac_cv_prog_RANLIB="$RANLIB" # Let the user override the test. else as_save_IFS=$IFS; IFS=$PATH_SEPARATOR for as_dir in $PATH do IFS=$as_save_IFS test -z "$as_dir" && as_dir=. for ac_exec_ext in '' $ac_executable_extensions; do if as_fn_executable_p "$as_dir/$ac_word$ac_exec_ext"; then ac_cv_prog_RANLIB="${ac_tool_prefix}ranlib" $as_echo "$as_me:${as_lineno-$LINENO}: found $as_dir/$ac_word$ac_exec_ext" >&5 break 2 fi done done IFS=$as_save_IFS fi fi RANLIB=$ac_cv_prog_RANLIB if test -n "$RANLIB"; then { $as_echo "$as_me:${as_lineno-$LINENO}: result: $RANLIB" >&5 $as_echo "$RANLIB" >&6; } else { $as_echo "$as_me:${as_lineno-$LINENO}: result: no" >&5 $as_echo "no" >&6; } fi fi if test -z "$ac_cv_prog_RANLIB"; then ac_ct_RANLIB=$RANLIB # Extract the first word of "ranlib", so it can be a program name with args. set dummy ranlib; ac_word=$2 { $as_echo "$as_me:${as_lineno-$LINENO}: checking for $ac_word" >&5 $as_echo_n "checking for $ac_word... " >&6; } if ${ac_cv_prog_ac_ct_RANLIB+:} false; then : $as_echo_n "(cached) " >&6 else if test -n "$ac_ct_RANLIB"; then ac_cv_prog_ac_ct_RANLIB="$ac_ct_RANLIB" # Let the user override the test. else as_save_IFS=$IFS; IFS=$PATH_SEPARATOR for as_dir in $PATH do IFS=$as_save_IFS test -z "$as_dir" && as_dir=. for ac_exec_ext in '' $ac_executable_extensions; do if as_fn_executable_p "$as_dir/$ac_word$ac_exec_ext"; then ac_cv_prog_ac_ct_RANLIB="ranlib" $as_echo "$as_me:${as_lineno-$LINENO}: found $as_dir/$ac_word$ac_exec_ext" >&5 break 2 fi done done IFS=$as_save_IFS fi fi ac_ct_RANLIB=$ac_cv_prog_ac_ct_RANLIB if test -n "$ac_ct_RANLIB"; then { $as_echo "$as_me:${as_lineno-$LINENO}: result: $ac_ct_RANLIB" >&5 $as_echo "$ac_ct_RANLIB" >&6; } else { $as_echo "$as_me:${as_lineno-$LINENO}: result: no" >&5 $as_echo "no" >&6; } fi if test "x$ac_ct_RANLIB" = x; then RANLIB=":" else case $cross_compiling:$ac_tool_warned in yes:) { $as_echo "$as_me:${as_lineno-$LINENO}: WARNING: using cross tools not prefixed with host triplet" >&5 $as_echo "$as_me: WARNING: using cross tools not prefixed with host triplet" >&2;} ac_tool_warned=yes ;; esac RANLIB=$ac_ct_RANLIB fi else RANLIB="$ac_cv_prog_RANLIB" fi test -z "$RANLIB" && RANLIB=: # Determine commands to create old-style static archives. old_archive_cmds='$AR $AR_FLAGS $oldlib$oldobjs' old_postinstall_cmds='chmod 644 $oldlib' old_postuninstall_cmds= if test -n "$RANLIB"; then case $host_os in openbsd*) old_postinstall_cmds="$old_postinstall_cmds~\$RANLIB -t \$tool_oldlib" ;; *) old_postinstall_cmds="$old_postinstall_cmds~\$RANLIB \$tool_oldlib" ;; esac old_archive_cmds="$old_archive_cmds~\$RANLIB \$tool_oldlib" fi case $host_os in darwin*) lock_old_archive_extraction=yes ;; *) lock_old_archive_extraction=no ;; esac # If no C compiler was specified, use CC. LTCC=${LTCC-"$CC"} # If no C compiler flags were specified, use CFLAGS. LTCFLAGS=${LTCFLAGS-"$CFLAGS"} # Allow CC to be a program name with arguments. compiler=$CC # Check for command to grab the raw symbol name followed by C symbol from nm. { $as_echo "$as_me:${as_lineno-$LINENO}: checking command to parse $NM output from $compiler object" >&5 $as_echo_n "checking command to parse $NM output from $compiler object... " >&6; } if ${lt_cv_sys_global_symbol_pipe+:} false; then : $as_echo_n "(cached) " >&6 else # These are sane defaults that work on at least a few old systems. # [They come from Ultrix. What could be older than Ultrix?!! ;)] # Character class describing NM global symbol codes. symcode='[BCDEGRST]' # Regexp to match symbols that can be accessed directly from C. sympat='\([_A-Za-z][_A-Za-z0-9]*\)' # Define system-specific variables. case $host_os in aix*) symcode='[BCDT]' ;; cygwin* | mingw* | pw32* | cegcc*) symcode='[ABCDGISTW]' ;; hpux*) if test "$host_cpu" = ia64; then symcode='[ABCDEGRST]' fi ;; irix* | nonstopux*) symcode='[BCDEGRST]' ;; osf*) symcode='[BCDEGQRST]' ;; solaris*) symcode='[BDRT]' ;; sco3.2v5*) symcode='[DT]' ;; sysv4.2uw2*) symcode='[DT]' ;; sysv5* | sco5v6* | unixware* | OpenUNIX*) symcode='[ABDT]' ;; sysv4) symcode='[DFNSTU]' ;; esac # If we're using GNU nm, then use its standard symbol codes. case `$NM -V 2>&1` in *GNU* | *'with BFD'*) symcode='[ABCDGIRSTW]' ;; esac # Transform an extracted symbol line into a proper C declaration. # Some systems (esp. on ia64) link data and code symbols differently, # so use this general approach. lt_cv_sys_global_symbol_to_cdecl="sed -n -e 's/^T .* \(.*\)$/extern int \1();/p' -e 's/^$symcode* .* \(.*\)$/extern char \1;/p'" # Transform an extracted symbol line into symbol name and symbol address lt_cv_sys_global_symbol_to_c_name_address="sed -n -e 's/^: \([^ ]*\)[ ]*$/ {\\\"\1\\\", (void *) 0},/p' -e 's/^$symcode* \([^ ]*\) \([^ ]*\)$/ {\"\2\", (void *) \&\2},/p'" lt_cv_sys_global_symbol_to_c_name_address_lib_prefix="sed -n -e 's/^: \([^ ]*\)[ ]*$/ {\\\"\1\\\", (void *) 0},/p' -e 's/^$symcode* \([^ ]*\) \(lib[^ ]*\)$/ {\"\2\", (void *) \&\2},/p' -e 's/^$symcode* \([^ ]*\) \([^ ]*\)$/ {\"lib\2\", (void *) \&\2},/p'" # Handle CRLF in mingw tool chain opt_cr= case $build_os in mingw*) opt_cr=`$ECHO 'x\{0,1\}' | tr x '\015'` # option cr in regexp ;; esac # Try without a prefix underscore, then with it. for ac_symprfx in "" "_"; do # Transform symcode, sympat, and symprfx into a raw symbol and a C symbol. symxfrm="\\1 $ac_symprfx\\2 \\2" # Write the raw and C identifiers. if test "$lt_cv_nm_interface" = "MS dumpbin"; then # Fake it for dumpbin and say T for any non-static function # and D for any global variable. # Also find C++ and __fastcall symbols from MSVC++, # which start with @ or ?. lt_cv_sys_global_symbol_pipe="$AWK '"\ " {last_section=section; section=\$ 3};"\ " /^COFF SYMBOL TABLE/{for(i in hide) delete hide[i]};"\ " /Section length .*#relocs.*(pick any)/{hide[last_section]=1};"\ " \$ 0!~/External *\|/{next};"\ " / 0+ UNDEF /{next}; / UNDEF \([^|]\)*()/{next};"\ " {if(hide[section]) next};"\ " {f=0}; \$ 0~/\(\).*\|/{f=1}; {printf f ? \"T \" : \"D \"};"\ " {split(\$ 0, a, /\||\r/); split(a[2], s)};"\ " s[1]~/^[@?]/{print s[1], s[1]; next};"\ " s[1]~prfx {split(s[1],t,\"@\"); print t[1], substr(t[1],length(prfx))}"\ " ' prfx=^$ac_symprfx" else lt_cv_sys_global_symbol_pipe="sed -n -e 's/^.*[ ]\($symcode$symcode*\)[ ][ ]*$ac_symprfx$sympat$opt_cr$/$symxfrm/p'" fi lt_cv_sys_global_symbol_pipe="$lt_cv_sys_global_symbol_pipe | sed '/ __gnu_lto/d'" # Check to see that the pipe works correctly. pipe_works=no rm -f conftest* cat > conftest.$ac_ext <<_LT_EOF #ifdef __cplusplus extern "C" { #endif char nm_test_var; void nm_test_func(void); void nm_test_func(void){} #ifdef __cplusplus } #endif int main(){nm_test_var='a';nm_test_func();return(0);} _LT_EOF if { { eval echo "\"\$as_me\":${as_lineno-$LINENO}: \"$ac_compile\""; } >&5 (eval $ac_compile) 2>&5 ac_status=$? $as_echo "$as_me:${as_lineno-$LINENO}: \$? = $ac_status" >&5 test $ac_status = 0; }; then # Now try to grab the symbols. nlist=conftest.nm if { { eval echo "\"\$as_me\":${as_lineno-$LINENO}: \"$NM conftest.$ac_objext \| "$lt_cv_sys_global_symbol_pipe" \> $nlist\""; } >&5 (eval $NM conftest.$ac_objext \| "$lt_cv_sys_global_symbol_pipe" \> $nlist) 2>&5 ac_status=$? $as_echo "$as_me:${as_lineno-$LINENO}: \$? = $ac_status" >&5 test $ac_status = 0; } && test -s "$nlist"; then # Try sorting and uniquifying the output. if sort "$nlist" | uniq > "$nlist"T; then mv -f "$nlist"T "$nlist" else rm -f "$nlist"T fi # Make sure that we snagged all the symbols we need. if $GREP ' nm_test_var$' "$nlist" >/dev/null; then if $GREP ' nm_test_func$' "$nlist" >/dev/null; then cat <<_LT_EOF > conftest.$ac_ext /* Keep this code in sync between libtool.m4, ltmain, lt_system.h, and tests. */ #if defined(_WIN32) || defined(__CYGWIN__) || defined(_WIN32_WCE) /* DATA imports from DLLs on WIN32 con't be const, because runtime relocations are performed -- see ld's documentation on pseudo-relocs. */ # define LT_DLSYM_CONST #elif defined(__osf__) /* This system does not cope well with relocations in const data. */ # define LT_DLSYM_CONST #else # define LT_DLSYM_CONST const #endif #ifdef __cplusplus extern "C" { #endif _LT_EOF # Now generate the symbol file. eval "$lt_cv_sys_global_symbol_to_cdecl"' < "$nlist" | $GREP -v main >> conftest.$ac_ext' cat <<_LT_EOF >> conftest.$ac_ext /* The mapping between symbol names and symbols. */ LT_DLSYM_CONST struct { const char *name; void *address; } lt__PROGRAM__LTX_preloaded_symbols[] = { { "@PROGRAM@", (void *) 0 }, _LT_EOF $SED "s/^$symcode$symcode* \(.*\) \(.*\)$/ {\"\2\", (void *) \&\2},/" < "$nlist" | $GREP -v main >> conftest.$ac_ext cat <<\_LT_EOF >> conftest.$ac_ext {0, (void *) 0} }; /* This works around a problem in FreeBSD linker */ #ifdef FREEBSD_WORKAROUND static const void *lt_preloaded_setup() { return lt__PROGRAM__LTX_preloaded_symbols; } #endif #ifdef __cplusplus } #endif _LT_EOF # Now try linking the two files. mv conftest.$ac_objext conftstm.$ac_objext lt_globsym_save_LIBS=$LIBS lt_globsym_save_CFLAGS=$CFLAGS LIBS="conftstm.$ac_objext" CFLAGS="$CFLAGS$lt_prog_compiler_no_builtin_flag" if { { eval echo "\"\$as_me\":${as_lineno-$LINENO}: \"$ac_link\""; } >&5 (eval $ac_link) 2>&5 ac_status=$? $as_echo "$as_me:${as_lineno-$LINENO}: \$? = $ac_status" >&5 test $ac_status = 0; } && test -s conftest${ac_exeext}; then pipe_works=yes fi LIBS=$lt_globsym_save_LIBS CFLAGS=$lt_globsym_save_CFLAGS else echo "cannot find nm_test_func in $nlist" >&5 fi else echo "cannot find nm_test_var in $nlist" >&5 fi else echo "cannot run $lt_cv_sys_global_symbol_pipe" >&5 fi else echo "$progname: failed program was:" >&5 cat conftest.$ac_ext >&5 fi rm -rf conftest* conftst* # Do not use the global_symbol_pipe unless it works. if test "$pipe_works" = yes; then break else lt_cv_sys_global_symbol_pipe= fi done fi if test -z "$lt_cv_sys_global_symbol_pipe"; then lt_cv_sys_global_symbol_to_cdecl= fi if test -z "$lt_cv_sys_global_symbol_pipe$lt_cv_sys_global_symbol_to_cdecl"; then { $as_echo "$as_me:${as_lineno-$LINENO}: result: failed" >&5 $as_echo "failed" >&6; } else { $as_echo "$as_me:${as_lineno-$LINENO}: result: ok" >&5 $as_echo "ok" >&6; } fi # Response file support. if test "$lt_cv_nm_interface" = "MS dumpbin"; then nm_file_list_spec='@' elif $NM --help 2>/dev/null | grep '[@]FILE' >/dev/null; then nm_file_list_spec='@' fi { $as_echo "$as_me:${as_lineno-$LINENO}: checking for sysroot" >&5 $as_echo_n "checking for sysroot... " >&6; } # Check whether --with-sysroot was given. if test "${with_sysroot+set}" = set; then : withval=$with_sysroot; else with_sysroot=no fi lt_sysroot= case ${with_sysroot} in #( yes) if test "$GCC" = yes; then lt_sysroot=`$CC --print-sysroot 2>/dev/null` fi ;; #( /*) lt_sysroot=`echo "$with_sysroot" | sed -e "$sed_quote_subst"` ;; #( no|'') ;; #( *) { $as_echo "$as_me:${as_lineno-$LINENO}: result: ${with_sysroot}" >&5 $as_echo "${with_sysroot}" >&6; } as_fn_error $? "The sysroot must be an absolute path." "$LINENO" 5 ;; esac { $as_echo "$as_me:${as_lineno-$LINENO}: result: ${lt_sysroot:-no}" >&5 $as_echo "${lt_sysroot:-no}" >&6; } # Check whether --enable-libtool-lock was given. if test "${enable_libtool_lock+set}" = set; then : enableval=$enable_libtool_lock; fi test "x$enable_libtool_lock" != xno && enable_libtool_lock=yes # Some flags need to be propagated to the compiler or linker for good # libtool support. case $host in ia64-*-hpux*) # Find out which ABI we are using. echo 'int i;' > conftest.$ac_ext if { { eval echo "\"\$as_me\":${as_lineno-$LINENO}: \"$ac_compile\""; } >&5 (eval $ac_compile) 2>&5 ac_status=$? $as_echo "$as_me:${as_lineno-$LINENO}: \$? = $ac_status" >&5 test $ac_status = 0; }; then case `/usr/bin/file conftest.$ac_objext` in *ELF-32*) HPUX_IA64_MODE="32" ;; *ELF-64*) HPUX_IA64_MODE="64" ;; esac fi rm -rf conftest* ;; *-*-irix6*) # Find out which ABI we are using. echo '#line '$LINENO' "configure"' > conftest.$ac_ext if { { eval echo "\"\$as_me\":${as_lineno-$LINENO}: \"$ac_compile\""; } >&5 (eval $ac_compile) 2>&5 ac_status=$? $as_echo "$as_me:${as_lineno-$LINENO}: \$? = $ac_status" >&5 test $ac_status = 0; }; then if test "$lt_cv_prog_gnu_ld" = yes; then case `/usr/bin/file conftest.$ac_objext` in *32-bit*) LD="${LD-ld} -melf32bsmip" ;; *N32*) LD="${LD-ld} -melf32bmipn32" ;; *64-bit*) LD="${LD-ld} -melf64bmip" ;; esac else case `/usr/bin/file conftest.$ac_objext` in *32-bit*) LD="${LD-ld} -32" ;; *N32*) LD="${LD-ld} -n32" ;; *64-bit*) LD="${LD-ld} -64" ;; esac fi fi rm -rf conftest* ;; x86_64-*kfreebsd*-gnu|x86_64-*linux*|ppc*-*linux*|powerpc*-*linux*| \ s390*-*linux*|s390*-*tpf*|sparc*-*linux*) # Find out which ABI we are using. echo 'int i;' > conftest.$ac_ext if { { eval echo "\"\$as_me\":${as_lineno-$LINENO}: \"$ac_compile\""; } >&5 (eval $ac_compile) 2>&5 ac_status=$? $as_echo "$as_me:${as_lineno-$LINENO}: \$? = $ac_status" >&5 test $ac_status = 0; }; then case `/usr/bin/file conftest.o` in *32-bit*) case $host in x86_64-*kfreebsd*-gnu) LD="${LD-ld} -m elf_i386_fbsd" ;; x86_64-*linux*) LD="${LD-ld} -m elf_i386" ;; ppc64-*linux*|powerpc64-*linux*) LD="${LD-ld} -m elf32ppclinux" ;; s390x-*linux*) LD="${LD-ld} -m elf_s390" ;; sparc64-*linux*) LD="${LD-ld} -m elf32_sparc" ;; esac ;; *64-bit*) case $host in x86_64-*kfreebsd*-gnu) LD="${LD-ld} -m elf_x86_64_fbsd" ;; x86_64-*linux*) LD="${LD-ld} -m elf_x86_64" ;; ppc*-*linux*|powerpc*-*linux*) LD="${LD-ld} -m elf64ppc" ;; s390*-*linux*|s390*-*tpf*) LD="${LD-ld} -m elf64_s390" ;; sparc*-*linux*) LD="${LD-ld} -m elf64_sparc" ;; esac ;; esac fi rm -rf conftest* ;; *-*-sco3.2v5*) # On SCO OpenServer 5, we need -belf to get full-featured binaries. SAVE_CFLAGS="$CFLAGS" CFLAGS="$CFLAGS -belf" { $as_echo "$as_me:${as_lineno-$LINENO}: checking whether the C compiler needs -belf" >&5 $as_echo_n "checking whether the C compiler needs -belf... " >&6; } if ${lt_cv_cc_needs_belf+:} false; then : $as_echo_n "(cached) " >&6 else ac_ext=c ac_cpp='$CPP $CPPFLAGS' ac_compile='$CC -c $CFLAGS $CPPFLAGS conftest.$ac_ext >&5' ac_link='$CC -o conftest$ac_exeext $CFLAGS $CPPFLAGS $LDFLAGS conftest.$ac_ext $LIBS >&5' ac_compiler_gnu=$ac_cv_c_compiler_gnu cat confdefs.h - <<_ACEOF >conftest.$ac_ext /* end confdefs.h. */ int main () { ; return 0; } _ACEOF if ac_fn_c_try_link "$LINENO"; then : lt_cv_cc_needs_belf=yes else lt_cv_cc_needs_belf=no fi rm -f core conftest.err conftest.$ac_objext \ conftest$ac_exeext conftest.$ac_ext ac_ext=c ac_cpp='$CPP $CPPFLAGS' ac_compile='$CC -c $CFLAGS $CPPFLAGS conftest.$ac_ext >&5' ac_link='$CC -o conftest$ac_exeext $CFLAGS $CPPFLAGS $LDFLAGS conftest.$ac_ext $LIBS >&5' ac_compiler_gnu=$ac_cv_c_compiler_gnu fi { $as_echo "$as_me:${as_lineno-$LINENO}: result: $lt_cv_cc_needs_belf" >&5 $as_echo "$lt_cv_cc_needs_belf" >&6; } if test x"$lt_cv_cc_needs_belf" != x"yes"; then # this is probably gcc 2.8.0, egcs 1.0 or newer; no need for -belf CFLAGS="$SAVE_CFLAGS" fi ;; *-*solaris*) # Find out which ABI we are using. echo 'int i;' > conftest.$ac_ext if { { eval echo "\"\$as_me\":${as_lineno-$LINENO}: \"$ac_compile\""; } >&5 (eval $ac_compile) 2>&5 ac_status=$? $as_echo "$as_me:${as_lineno-$LINENO}: \$? = $ac_status" >&5 test $ac_status = 0; }; then case `/usr/bin/file conftest.o` in *64-bit*) case $lt_cv_prog_gnu_ld in yes*) case $host in i?86-*-solaris*) LD="${LD-ld} -m elf_x86_64" ;; sparc*-*-solaris*) LD="${LD-ld} -m elf64_sparc" ;; esac # GNU ld 2.21 introduced _sol2 emulations. Use them if available. if ${LD-ld} -V | grep _sol2 >/dev/null 2>&1; then LD="${LD-ld}_sol2" fi ;; *) if ${LD-ld} -64 -r -o conftest2.o conftest.o >/dev/null 2>&1; then LD="${LD-ld} -64" fi ;; esac ;; esac fi rm -rf conftest* ;; esac need_locks="$enable_libtool_lock" if test -n "$ac_tool_prefix"; then # Extract the first word of "${ac_tool_prefix}mt", so it can be a program name with args. set dummy ${ac_tool_prefix}mt; ac_word=$2 { $as_echo "$as_me:${as_lineno-$LINENO}: checking for $ac_word" >&5 $as_echo_n "checking for $ac_word... " >&6; } if ${ac_cv_prog_MANIFEST_TOOL+:} false; then : $as_echo_n "(cached) " >&6 else if test -n "$MANIFEST_TOOL"; then ac_cv_prog_MANIFEST_TOOL="$MANIFEST_TOOL" # Let the user override the test. else as_save_IFS=$IFS; IFS=$PATH_SEPARATOR for as_dir in $PATH do IFS=$as_save_IFS test -z "$as_dir" && as_dir=. for ac_exec_ext in '' $ac_executable_extensions; do if as_fn_executable_p "$as_dir/$ac_word$ac_exec_ext"; then ac_cv_prog_MANIFEST_TOOL="${ac_tool_prefix}mt" $as_echo "$as_me:${as_lineno-$LINENO}: found $as_dir/$ac_word$ac_exec_ext" >&5 break 2 fi done done IFS=$as_save_IFS fi fi MANIFEST_TOOL=$ac_cv_prog_MANIFEST_TOOL if test -n "$MANIFEST_TOOL"; then { $as_echo "$as_me:${as_lineno-$LINENO}: result: $MANIFEST_TOOL" >&5 $as_echo "$MANIFEST_TOOL" >&6; } else { $as_echo "$as_me:${as_lineno-$LINENO}: result: no" >&5 $as_echo "no" >&6; } fi fi if test -z "$ac_cv_prog_MANIFEST_TOOL"; then ac_ct_MANIFEST_TOOL=$MANIFEST_TOOL # Extract the first word of "mt", so it can be a program name with args. set dummy mt; ac_word=$2 { $as_echo "$as_me:${as_lineno-$LINENO}: checking for $ac_word" >&5 $as_echo_n "checking for $ac_word... " >&6; } if ${ac_cv_prog_ac_ct_MANIFEST_TOOL+:} false; then : $as_echo_n "(cached) " >&6 else if test -n "$ac_ct_MANIFEST_TOOL"; then ac_cv_prog_ac_ct_MANIFEST_TOOL="$ac_ct_MANIFEST_TOOL" # Let the user override the test. else as_save_IFS=$IFS; IFS=$PATH_SEPARATOR for as_dir in $PATH do IFS=$as_save_IFS test -z "$as_dir" && as_dir=. for ac_exec_ext in '' $ac_executable_extensions; do if as_fn_executable_p "$as_dir/$ac_word$ac_exec_ext"; then ac_cv_prog_ac_ct_MANIFEST_TOOL="mt" $as_echo "$as_me:${as_lineno-$LINENO}: found $as_dir/$ac_word$ac_exec_ext" >&5 break 2 fi done done IFS=$as_save_IFS fi fi ac_ct_MANIFEST_TOOL=$ac_cv_prog_ac_ct_MANIFEST_TOOL if test -n "$ac_ct_MANIFEST_TOOL"; then { $as_echo "$as_me:${as_lineno-$LINENO}: result: $ac_ct_MANIFEST_TOOL" >&5 $as_echo "$ac_ct_MANIFEST_TOOL" >&6; } else { $as_echo "$as_me:${as_lineno-$LINENO}: result: no" >&5 $as_echo "no" >&6; } fi if test "x$ac_ct_MANIFEST_TOOL" = x; then MANIFEST_TOOL=":" else case $cross_compiling:$ac_tool_warned in yes:) { $as_echo "$as_me:${as_lineno-$LINENO}: WARNING: using cross tools not prefixed with host triplet" >&5 $as_echo "$as_me: WARNING: using cross tools not prefixed with host triplet" >&2;} ac_tool_warned=yes ;; esac MANIFEST_TOOL=$ac_ct_MANIFEST_TOOL fi else MANIFEST_TOOL="$ac_cv_prog_MANIFEST_TOOL" fi test -z "$MANIFEST_TOOL" && MANIFEST_TOOL=mt { $as_echo "$as_me:${as_lineno-$LINENO}: checking if $MANIFEST_TOOL is a manifest tool" >&5 $as_echo_n "checking if $MANIFEST_TOOL is a manifest tool... " >&6; } if ${lt_cv_path_mainfest_tool+:} false; then : $as_echo_n "(cached) " >&6 else lt_cv_path_mainfest_tool=no echo "$as_me:$LINENO: $MANIFEST_TOOL '-?'" >&5 $MANIFEST_TOOL '-?' 2>conftest.err > conftest.out cat conftest.err >&5 if $GREP 'Manifest Tool' conftest.out > /dev/null; then lt_cv_path_mainfest_tool=yes fi rm -f conftest* fi { $as_echo "$as_me:${as_lineno-$LINENO}: result: $lt_cv_path_mainfest_tool" >&5 $as_echo "$lt_cv_path_mainfest_tool" >&6; } if test "x$lt_cv_path_mainfest_tool" != xyes; then MANIFEST_TOOL=: fi case $host_os in rhapsody* | darwin*) if test -n "$ac_tool_prefix"; then # Extract the first word of "${ac_tool_prefix}dsymutil", so it can be a program name with args. set dummy ${ac_tool_prefix}dsymutil; ac_word=$2 { $as_echo "$as_me:${as_lineno-$LINENO}: checking for $ac_word" >&5 $as_echo_n "checking for $ac_word... " >&6; } if ${ac_cv_prog_DSYMUTIL+:} false; then : $as_echo_n "(cached) " >&6 else if test -n "$DSYMUTIL"; then ac_cv_prog_DSYMUTIL="$DSYMUTIL" # Let the user override the test. else as_save_IFS=$IFS; IFS=$PATH_SEPARATOR for as_dir in $PATH do IFS=$as_save_IFS test -z "$as_dir" && as_dir=. for ac_exec_ext in '' $ac_executable_extensions; do if as_fn_executable_p "$as_dir/$ac_word$ac_exec_ext"; then ac_cv_prog_DSYMUTIL="${ac_tool_prefix}dsymutil" $as_echo "$as_me:${as_lineno-$LINENO}: found $as_dir/$ac_word$ac_exec_ext" >&5 break 2 fi done done IFS=$as_save_IFS fi fi DSYMUTIL=$ac_cv_prog_DSYMUTIL if test -n "$DSYMUTIL"; then { $as_echo "$as_me:${as_lineno-$LINENO}: result: $DSYMUTIL" >&5 $as_echo "$DSYMUTIL" >&6; } else { $as_echo "$as_me:${as_lineno-$LINENO}: result: no" >&5 $as_echo "no" >&6; } fi fi if test -z "$ac_cv_prog_DSYMUTIL"; then ac_ct_DSYMUTIL=$DSYMUTIL # Extract the first word of "dsymutil", so it can be a program name with args. set dummy dsymutil; ac_word=$2 { $as_echo "$as_me:${as_lineno-$LINENO}: checking for $ac_word" >&5 $as_echo_n "checking for $ac_word... " >&6; } if ${ac_cv_prog_ac_ct_DSYMUTIL+:} false; then : $as_echo_n "(cached) " >&6 else if test -n "$ac_ct_DSYMUTIL"; then ac_cv_prog_ac_ct_DSYMUTIL="$ac_ct_DSYMUTIL" # Let the user override the test. else as_save_IFS=$IFS; IFS=$PATH_SEPARATOR for as_dir in $PATH do IFS=$as_save_IFS test -z "$as_dir" && as_dir=. for ac_exec_ext in '' $ac_executable_extensions; do if as_fn_executable_p "$as_dir/$ac_word$ac_exec_ext"; then ac_cv_prog_ac_ct_DSYMUTIL="dsymutil" $as_echo "$as_me:${as_lineno-$LINENO}: found $as_dir/$ac_word$ac_exec_ext" >&5 break 2 fi done done IFS=$as_save_IFS fi fi ac_ct_DSYMUTIL=$ac_cv_prog_ac_ct_DSYMUTIL if test -n "$ac_ct_DSYMUTIL"; then { $as_echo "$as_me:${as_lineno-$LINENO}: result: $ac_ct_DSYMUTIL" >&5 $as_echo "$ac_ct_DSYMUTIL" >&6; } else { $as_echo "$as_me:${as_lineno-$LINENO}: result: no" >&5 $as_echo "no" >&6; } fi if test "x$ac_ct_DSYMUTIL" = x; then DSYMUTIL=":" else case $cross_compiling:$ac_tool_warned in yes:) { $as_echo "$as_me:${as_lineno-$LINENO}: WARNING: using cross tools not prefixed with host triplet" >&5 $as_echo "$as_me: WARNING: using cross tools not prefixed with host triplet" >&2;} ac_tool_warned=yes ;; esac DSYMUTIL=$ac_ct_DSYMUTIL fi else DSYMUTIL="$ac_cv_prog_DSYMUTIL" fi if test -n "$ac_tool_prefix"; then # Extract the first word of "${ac_tool_prefix}nmedit", so it can be a program name with args. set dummy ${ac_tool_prefix}nmedit; ac_word=$2 { $as_echo "$as_me:${as_lineno-$LINENO}: checking for $ac_word" >&5 $as_echo_n "checking for $ac_word... " >&6; } if ${ac_cv_prog_NMEDIT+:} false; then : $as_echo_n "(cached) " >&6 else if test -n "$NMEDIT"; then ac_cv_prog_NMEDIT="$NMEDIT" # Let the user override the test. else as_save_IFS=$IFS; IFS=$PATH_SEPARATOR for as_dir in $PATH do IFS=$as_save_IFS test -z "$as_dir" && as_dir=. for ac_exec_ext in '' $ac_executable_extensions; do if as_fn_executable_p "$as_dir/$ac_word$ac_exec_ext"; then ac_cv_prog_NMEDIT="${ac_tool_prefix}nmedit" $as_echo "$as_me:${as_lineno-$LINENO}: found $as_dir/$ac_word$ac_exec_ext" >&5 break 2 fi done done IFS=$as_save_IFS fi fi NMEDIT=$ac_cv_prog_NMEDIT if test -n "$NMEDIT"; then { $as_echo "$as_me:${as_lineno-$LINENO}: result: $NMEDIT" >&5 $as_echo "$NMEDIT" >&6; } else { $as_echo "$as_me:${as_lineno-$LINENO}: result: no" >&5 $as_echo "no" >&6; } fi fi if test -z "$ac_cv_prog_NMEDIT"; then ac_ct_NMEDIT=$NMEDIT # Extract the first word of "nmedit", so it can be a program name with args. set dummy nmedit; ac_word=$2 { $as_echo "$as_me:${as_lineno-$LINENO}: checking for $ac_word" >&5 $as_echo_n "checking for $ac_word... " >&6; } if ${ac_cv_prog_ac_ct_NMEDIT+:} false; then : $as_echo_n "(cached) " >&6 else if test -n "$ac_ct_NMEDIT"; then ac_cv_prog_ac_ct_NMEDIT="$ac_ct_NMEDIT" # Let the user override the test. else as_save_IFS=$IFS; IFS=$PATH_SEPARATOR for as_dir in $PATH do IFS=$as_save_IFS test -z "$as_dir" && as_dir=. for ac_exec_ext in '' $ac_executable_extensions; do if as_fn_executable_p "$as_dir/$ac_word$ac_exec_ext"; then ac_cv_prog_ac_ct_NMEDIT="nmedit" $as_echo "$as_me:${as_lineno-$LINENO}: found $as_dir/$ac_word$ac_exec_ext" >&5 break 2 fi done done IFS=$as_save_IFS fi fi ac_ct_NMEDIT=$ac_cv_prog_ac_ct_NMEDIT if test -n "$ac_ct_NMEDIT"; then { $as_echo "$as_me:${as_lineno-$LINENO}: result: $ac_ct_NMEDIT" >&5 $as_echo "$ac_ct_NMEDIT" >&6; } else { $as_echo "$as_me:${as_lineno-$LINENO}: result: no" >&5 $as_echo "no" >&6; } fi if test "x$ac_ct_NMEDIT" = x; then NMEDIT=":" else case $cross_compiling:$ac_tool_warned in yes:) { $as_echo "$as_me:${as_lineno-$LINENO}: WARNING: using cross tools not prefixed with host triplet" >&5 $as_echo "$as_me: WARNING: using cross tools not prefixed with host triplet" >&2;} ac_tool_warned=yes ;; esac NMEDIT=$ac_ct_NMEDIT fi else NMEDIT="$ac_cv_prog_NMEDIT" fi if test -n "$ac_tool_prefix"; then # Extract the first word of "${ac_tool_prefix}lipo", so it can be a program name with args. set dummy ${ac_tool_prefix}lipo; ac_word=$2 { $as_echo "$as_me:${as_lineno-$LINENO}: checking for $ac_word" >&5 $as_echo_n "checking for $ac_word... " >&6; } if ${ac_cv_prog_LIPO+:} false; then : $as_echo_n "(cached) " >&6 else if test -n "$LIPO"; then ac_cv_prog_LIPO="$LIPO" # Let the user override the test. else as_save_IFS=$IFS; IFS=$PATH_SEPARATOR for as_dir in $PATH do IFS=$as_save_IFS test -z "$as_dir" && as_dir=. for ac_exec_ext in '' $ac_executable_extensions; do if as_fn_executable_p "$as_dir/$ac_word$ac_exec_ext"; then ac_cv_prog_LIPO="${ac_tool_prefix}lipo" $as_echo "$as_me:${as_lineno-$LINENO}: found $as_dir/$ac_word$ac_exec_ext" >&5 break 2 fi done done IFS=$as_save_IFS fi fi LIPO=$ac_cv_prog_LIPO if test -n "$LIPO"; then { $as_echo "$as_me:${as_lineno-$LINENO}: result: $LIPO" >&5 $as_echo "$LIPO" >&6; } else { $as_echo "$as_me:${as_lineno-$LINENO}: result: no" >&5 $as_echo "no" >&6; } fi fi if test -z "$ac_cv_prog_LIPO"; then ac_ct_LIPO=$LIPO # Extract the first word of "lipo", so it can be a program name with args. set dummy lipo; ac_word=$2 { $as_echo "$as_me:${as_lineno-$LINENO}: checking for $ac_word" >&5 $as_echo_n "checking for $ac_word... " >&6; } if ${ac_cv_prog_ac_ct_LIPO+:} false; then : $as_echo_n "(cached) " >&6 else if test -n "$ac_ct_LIPO"; then ac_cv_prog_ac_ct_LIPO="$ac_ct_LIPO" # Let the user override the test. else as_save_IFS=$IFS; IFS=$PATH_SEPARATOR for as_dir in $PATH do IFS=$as_save_IFS test -z "$as_dir" && as_dir=. for ac_exec_ext in '' $ac_executable_extensions; do if as_fn_executable_p "$as_dir/$ac_word$ac_exec_ext"; then ac_cv_prog_ac_ct_LIPO="lipo" $as_echo "$as_me:${as_lineno-$LINENO}: found $as_dir/$ac_word$ac_exec_ext" >&5 break 2 fi done done IFS=$as_save_IFS fi fi ac_ct_LIPO=$ac_cv_prog_ac_ct_LIPO if test -n "$ac_ct_LIPO"; then { $as_echo "$as_me:${as_lineno-$LINENO}: result: $ac_ct_LIPO" >&5 $as_echo "$ac_ct_LIPO" >&6; } else { $as_echo "$as_me:${as_lineno-$LINENO}: result: no" >&5 $as_echo "no" >&6; } fi if test "x$ac_ct_LIPO" = x; then LIPO=":" else case $cross_compiling:$ac_tool_warned in yes:) { $as_echo "$as_me:${as_lineno-$LINENO}: WARNING: using cross tools not prefixed with host triplet" >&5 $as_echo "$as_me: WARNING: using cross tools not prefixed with host triplet" >&2;} ac_tool_warned=yes ;; esac LIPO=$ac_ct_LIPO fi else LIPO="$ac_cv_prog_LIPO" fi if test -n "$ac_tool_prefix"; then # Extract the first word of "${ac_tool_prefix}otool", so it can be a program name with args. set dummy ${ac_tool_prefix}otool; ac_word=$2 { $as_echo "$as_me:${as_lineno-$LINENO}: checking for $ac_word" >&5 $as_echo_n "checking for $ac_word... " >&6; } if ${ac_cv_prog_OTOOL+:} false; then : $as_echo_n "(cached) " >&6 else if test -n "$OTOOL"; then ac_cv_prog_OTOOL="$OTOOL" # Let the user override the test. else as_save_IFS=$IFS; IFS=$PATH_SEPARATOR for as_dir in $PATH do IFS=$as_save_IFS test -z "$as_dir" && as_dir=. for ac_exec_ext in '' $ac_executable_extensions; do if as_fn_executable_p "$as_dir/$ac_word$ac_exec_ext"; then ac_cv_prog_OTOOL="${ac_tool_prefix}otool" $as_echo "$as_me:${as_lineno-$LINENO}: found $as_dir/$ac_word$ac_exec_ext" >&5 break 2 fi done done IFS=$as_save_IFS fi fi OTOOL=$ac_cv_prog_OTOOL if test -n "$OTOOL"; then { $as_echo "$as_me:${as_lineno-$LINENO}: result: $OTOOL" >&5 $as_echo "$OTOOL" >&6; } else { $as_echo "$as_me:${as_lineno-$LINENO}: result: no" >&5 $as_echo "no" >&6; } fi fi if test -z "$ac_cv_prog_OTOOL"; then ac_ct_OTOOL=$OTOOL # Extract the first word of "otool", so it can be a program name with args. set dummy otool; ac_word=$2 { $as_echo "$as_me:${as_lineno-$LINENO}: checking for $ac_word" >&5 $as_echo_n "checking for $ac_word... " >&6; } if ${ac_cv_prog_ac_ct_OTOOL+:} false; then : $as_echo_n "(cached) " >&6 else if test -n "$ac_ct_OTOOL"; then ac_cv_prog_ac_ct_OTOOL="$ac_ct_OTOOL" # Let the user override the test. else as_save_IFS=$IFS; IFS=$PATH_SEPARATOR for as_dir in $PATH do IFS=$as_save_IFS test -z "$as_dir" && as_dir=. for ac_exec_ext in '' $ac_executable_extensions; do if as_fn_executable_p "$as_dir/$ac_word$ac_exec_ext"; then ac_cv_prog_ac_ct_OTOOL="otool" $as_echo "$as_me:${as_lineno-$LINENO}: found $as_dir/$ac_word$ac_exec_ext" >&5 break 2 fi done done IFS=$as_save_IFS fi fi ac_ct_OTOOL=$ac_cv_prog_ac_ct_OTOOL if test -n "$ac_ct_OTOOL"; then { $as_echo "$as_me:${as_lineno-$LINENO}: result: $ac_ct_OTOOL" >&5 $as_echo "$ac_ct_OTOOL" >&6; } else { $as_echo "$as_me:${as_lineno-$LINENO}: result: no" >&5 $as_echo "no" >&6; } fi if test "x$ac_ct_OTOOL" = x; then OTOOL=":" else case $cross_compiling:$ac_tool_warned in yes:) { $as_echo "$as_me:${as_lineno-$LINENO}: WARNING: using cross tools not prefixed with host triplet" >&5 $as_echo "$as_me: WARNING: using cross tools not prefixed with host triplet" >&2;} ac_tool_warned=yes ;; esac OTOOL=$ac_ct_OTOOL fi else OTOOL="$ac_cv_prog_OTOOL" fi if test -n "$ac_tool_prefix"; then # Extract the first word of "${ac_tool_prefix}otool64", so it can be a program name with args. set dummy ${ac_tool_prefix}otool64; ac_word=$2 { $as_echo "$as_me:${as_lineno-$LINENO}: checking for $ac_word" >&5 $as_echo_n "checking for $ac_word... " >&6; } if ${ac_cv_prog_OTOOL64+:} false; then : $as_echo_n "(cached) " >&6 else if test -n "$OTOOL64"; then ac_cv_prog_OTOOL64="$OTOOL64" # Let the user override the test. else as_save_IFS=$IFS; IFS=$PATH_SEPARATOR for as_dir in $PATH do IFS=$as_save_IFS test -z "$as_dir" && as_dir=. for ac_exec_ext in '' $ac_executable_extensions; do if as_fn_executable_p "$as_dir/$ac_word$ac_exec_ext"; then ac_cv_prog_OTOOL64="${ac_tool_prefix}otool64" $as_echo "$as_me:${as_lineno-$LINENO}: found $as_dir/$ac_word$ac_exec_ext" >&5 break 2 fi done done IFS=$as_save_IFS fi fi OTOOL64=$ac_cv_prog_OTOOL64 if test -n "$OTOOL64"; then { $as_echo "$as_me:${as_lineno-$LINENO}: result: $OTOOL64" >&5 $as_echo "$OTOOL64" >&6; } else { $as_echo "$as_me:${as_lineno-$LINENO}: result: no" >&5 $as_echo "no" >&6; } fi fi if test -z "$ac_cv_prog_OTOOL64"; then ac_ct_OTOOL64=$OTOOL64 # Extract the first word of "otool64", so it can be a program name with args. set dummy otool64; ac_word=$2 { $as_echo "$as_me:${as_lineno-$LINENO}: checking for $ac_word" >&5 $as_echo_n "checking for $ac_word... " >&6; } if ${ac_cv_prog_ac_ct_OTOOL64+:} false; then : $as_echo_n "(cached) " >&6 else if test -n "$ac_ct_OTOOL64"; then ac_cv_prog_ac_ct_OTOOL64="$ac_ct_OTOOL64" # Let the user override the test. else as_save_IFS=$IFS; IFS=$PATH_SEPARATOR for as_dir in $PATH do IFS=$as_save_IFS test -z "$as_dir" && as_dir=. for ac_exec_ext in '' $ac_executable_extensions; do if as_fn_executable_p "$as_dir/$ac_word$ac_exec_ext"; then ac_cv_prog_ac_ct_OTOOL64="otool64" $as_echo "$as_me:${as_lineno-$LINENO}: found $as_dir/$ac_word$ac_exec_ext" >&5 break 2 fi done done IFS=$as_save_IFS fi fi ac_ct_OTOOL64=$ac_cv_prog_ac_ct_OTOOL64 if test -n "$ac_ct_OTOOL64"; then { $as_echo "$as_me:${as_lineno-$LINENO}: result: $ac_ct_OTOOL64" >&5 $as_echo "$ac_ct_OTOOL64" >&6; } else { $as_echo "$as_me:${as_lineno-$LINENO}: result: no" >&5 $as_echo "no" >&6; } fi if test "x$ac_ct_OTOOL64" = x; then OTOOL64=":" else case $cross_compiling:$ac_tool_warned in yes:) { $as_echo "$as_me:${as_lineno-$LINENO}: WARNING: using cross tools not prefixed with host triplet" >&5 $as_echo "$as_me: WARNING: using cross tools not prefixed with host triplet" >&2;} ac_tool_warned=yes ;; esac OTOOL64=$ac_ct_OTOOL64 fi else OTOOL64="$ac_cv_prog_OTOOL64" fi { $as_echo "$as_me:${as_lineno-$LINENO}: checking for -single_module linker flag" >&5 $as_echo_n "checking for -single_module linker flag... " >&6; } if ${lt_cv_apple_cc_single_mod+:} false; then : $as_echo_n "(cached) " >&6 else lt_cv_apple_cc_single_mod=no if test -z "${LT_MULTI_MODULE}"; then # By default we will add the -single_module flag. You can override # by either setting the environment variable LT_MULTI_MODULE # non-empty at configure time, or by adding -multi_module to the # link flags. rm -rf libconftest.dylib* echo "int foo(void){return 1;}" > conftest.c echo "$LTCC $LTCFLAGS $LDFLAGS -o libconftest.dylib \ -dynamiclib -Wl,-single_module conftest.c" >&5 $LTCC $LTCFLAGS $LDFLAGS -o libconftest.dylib \ -dynamiclib -Wl,-single_module conftest.c 2>conftest.err _lt_result=$? # If there is a non-empty error log, and "single_module" # appears in it, assume the flag caused a linker warning if test -s conftest.err && $GREP single_module conftest.err; then cat conftest.err >&5 # Otherwise, if the output was created with a 0 exit code from # the compiler, it worked. elif test -f libconftest.dylib && test $_lt_result -eq 0; then lt_cv_apple_cc_single_mod=yes else cat conftest.err >&5 fi rm -rf libconftest.dylib* rm -f conftest.* fi fi { $as_echo "$as_me:${as_lineno-$LINENO}: result: $lt_cv_apple_cc_single_mod" >&5 $as_echo "$lt_cv_apple_cc_single_mod" >&6; } { $as_echo "$as_me:${as_lineno-$LINENO}: checking for -exported_symbols_list linker flag" >&5 $as_echo_n "checking for -exported_symbols_list linker flag... " >&6; } if ${lt_cv_ld_exported_symbols_list+:} false; then : $as_echo_n "(cached) " >&6 else lt_cv_ld_exported_symbols_list=no save_LDFLAGS=$LDFLAGS echo "_main" > conftest.sym LDFLAGS="$LDFLAGS -Wl,-exported_symbols_list,conftest.sym" cat confdefs.h - <<_ACEOF >conftest.$ac_ext /* end confdefs.h. */ int main () { ; return 0; } _ACEOF if ac_fn_c_try_link "$LINENO"; then : lt_cv_ld_exported_symbols_list=yes else lt_cv_ld_exported_symbols_list=no fi rm -f core conftest.err conftest.$ac_objext \ conftest$ac_exeext conftest.$ac_ext LDFLAGS="$save_LDFLAGS" fi { $as_echo "$as_me:${as_lineno-$LINENO}: result: $lt_cv_ld_exported_symbols_list" >&5 $as_echo "$lt_cv_ld_exported_symbols_list" >&6; } { $as_echo "$as_me:${as_lineno-$LINENO}: checking for -force_load linker flag" >&5 $as_echo_n "checking for -force_load linker flag... " >&6; } if ${lt_cv_ld_force_load+:} false; then : $as_echo_n "(cached) " >&6 else lt_cv_ld_force_load=no cat > conftest.c << _LT_EOF int forced_loaded() { return 2;} _LT_EOF echo "$LTCC $LTCFLAGS -c -o conftest.o conftest.c" >&5 $LTCC $LTCFLAGS -c -o conftest.o conftest.c 2>&5 echo "$AR cru libconftest.a conftest.o" >&5 $AR cru libconftest.a conftest.o 2>&5 echo "$RANLIB libconftest.a" >&5 $RANLIB libconftest.a 2>&5 cat > conftest.c << _LT_EOF int main() { return 0;} _LT_EOF echo "$LTCC $LTCFLAGS $LDFLAGS -o conftest conftest.c -Wl,-force_load,./libconftest.a" >&5 $LTCC $LTCFLAGS $LDFLAGS -o conftest conftest.c -Wl,-force_load,./libconftest.a 2>conftest.err _lt_result=$? if test -s conftest.err && $GREP force_load conftest.err; then cat conftest.err >&5 elif test -f conftest && test $_lt_result -eq 0 && $GREP forced_load conftest >/dev/null 2>&1 ; then lt_cv_ld_force_load=yes else cat conftest.err >&5 fi rm -f conftest.err libconftest.a conftest conftest.c rm -rf conftest.dSYM fi { $as_echo "$as_me:${as_lineno-$LINENO}: result: $lt_cv_ld_force_load" >&5 $as_echo "$lt_cv_ld_force_load" >&6; } case $host_os in rhapsody* | darwin1.[012]) _lt_dar_allow_undefined='${wl}-undefined ${wl}suppress' ;; darwin1.*) _lt_dar_allow_undefined='${wl}-flat_namespace ${wl}-undefined ${wl}suppress' ;; darwin*) # darwin 5.x on # if running on 10.5 or later, the deployment target defaults # to the OS version, if on x86, and 10.4, the deployment # target defaults to 10.4. Don't you love it? case ${MACOSX_DEPLOYMENT_TARGET-10.0},$host in 10.0,*86*-darwin8*|10.0,*-darwin[91]*) _lt_dar_allow_undefined='${wl}-undefined ${wl}dynamic_lookup' ;; 10.[012]*) _lt_dar_allow_undefined='${wl}-flat_namespace ${wl}-undefined ${wl}suppress' ;; 10.*) _lt_dar_allow_undefined='${wl}-undefined ${wl}dynamic_lookup' ;; esac ;; esac if test "$lt_cv_apple_cc_single_mod" = "yes"; then _lt_dar_single_mod='$single_module' fi if test "$lt_cv_ld_exported_symbols_list" = "yes"; then _lt_dar_export_syms=' ${wl}-exported_symbols_list,$output_objdir/${libname}-symbols.expsym' else _lt_dar_export_syms='~$NMEDIT -s $output_objdir/${libname}-symbols.expsym ${lib}' fi if test "$DSYMUTIL" != ":" && test "$lt_cv_ld_force_load" = "no"; then _lt_dsymutil='~$DSYMUTIL $lib || :' else _lt_dsymutil= fi ;; esac ac_ext=c ac_cpp='$CPP $CPPFLAGS' ac_compile='$CC -c $CFLAGS $CPPFLAGS conftest.$ac_ext >&5' ac_link='$CC -o conftest$ac_exeext $CFLAGS $CPPFLAGS $LDFLAGS conftest.$ac_ext $LIBS >&5' ac_compiler_gnu=$ac_cv_c_compiler_gnu { $as_echo "$as_me:${as_lineno-$LINENO}: checking how to run the C preprocessor" >&5 $as_echo_n "checking how to run the C preprocessor... " >&6; } # On Suns, sometimes $CPP names a directory. if test -n "$CPP" && test -d "$CPP"; then CPP= fi if test -z "$CPP"; then if ${ac_cv_prog_CPP+:} false; then : $as_echo_n "(cached) " >&6 else # Double quotes because CPP needs to be expanded for CPP in "$CC -E" "$CC -E -traditional-cpp" "/lib/cpp" do ac_preproc_ok=false for ac_c_preproc_warn_flag in '' yes do # Use a header file that comes with gcc, so configuring glibc # with a fresh cross-compiler works. # Prefer to if __STDC__ is defined, since # exists even on freestanding compilers. # On the NeXT, cc -E runs the code through the compiler's parser, # not just through cpp. "Syntax error" is here to catch this case. cat confdefs.h - <<_ACEOF >conftest.$ac_ext /* end confdefs.h. */ #ifdef __STDC__ # include #else # include #endif Syntax error _ACEOF if ac_fn_c_try_cpp "$LINENO"; then : else # Broken: fails on valid input. continue fi rm -f conftest.err conftest.i conftest.$ac_ext # OK, works on sane cases. Now check whether nonexistent headers # can be detected and how. cat confdefs.h - <<_ACEOF >conftest.$ac_ext /* end confdefs.h. */ #include _ACEOF if ac_fn_c_try_cpp "$LINENO"; then : # Broken: success on invalid input. continue else # Passes both tests. ac_preproc_ok=: break fi rm -f conftest.err conftest.i conftest.$ac_ext done # Because of `break', _AC_PREPROC_IFELSE's cleaning code was skipped. rm -f conftest.i conftest.err conftest.$ac_ext if $ac_preproc_ok; then : break fi done ac_cv_prog_CPP=$CPP fi CPP=$ac_cv_prog_CPP else ac_cv_prog_CPP=$CPP fi { $as_echo "$as_me:${as_lineno-$LINENO}: result: $CPP" >&5 $as_echo "$CPP" >&6; } ac_preproc_ok=false for ac_c_preproc_warn_flag in '' yes do # Use a header file that comes with gcc, so configuring glibc # with a fresh cross-compiler works. # Prefer to if __STDC__ is defined, since # exists even on freestanding compilers. # On the NeXT, cc -E runs the code through the compiler's parser, # not just through cpp. "Syntax error" is here to catch this case. cat confdefs.h - <<_ACEOF >conftest.$ac_ext /* end confdefs.h. */ #ifdef __STDC__ # include #else # include #endif Syntax error _ACEOF if ac_fn_c_try_cpp "$LINENO"; then : else # Broken: fails on valid input. continue fi rm -f conftest.err conftest.i conftest.$ac_ext # OK, works on sane cases. Now check whether nonexistent headers # can be detected and how. cat confdefs.h - <<_ACEOF >conftest.$ac_ext /* end confdefs.h. */ #include _ACEOF if ac_fn_c_try_cpp "$LINENO"; then : # Broken: success on invalid input. continue else # Passes both tests. ac_preproc_ok=: break fi rm -f conftest.err conftest.i conftest.$ac_ext done # Because of `break', _AC_PREPROC_IFELSE's cleaning code was skipped. rm -f conftest.i conftest.err conftest.$ac_ext if $ac_preproc_ok; then : else { { $as_echo "$as_me:${as_lineno-$LINENO}: error: in \`$ac_pwd':" >&5 $as_echo "$as_me: error: in \`$ac_pwd':" >&2;} as_fn_error $? "C preprocessor \"$CPP\" fails sanity check See \`config.log' for more details" "$LINENO" 5; } fi ac_ext=c ac_cpp='$CPP $CPPFLAGS' ac_compile='$CC -c $CFLAGS $CPPFLAGS conftest.$ac_ext >&5' ac_link='$CC -o conftest$ac_exeext $CFLAGS $CPPFLAGS $LDFLAGS conftest.$ac_ext $LIBS >&5' ac_compiler_gnu=$ac_cv_c_compiler_gnu { $as_echo "$as_me:${as_lineno-$LINENO}: checking for ANSI C header files" >&5 $as_echo_n "checking for ANSI C header files... " >&6; } if ${ac_cv_header_stdc+:} false; then : $as_echo_n "(cached) " >&6 else cat confdefs.h - <<_ACEOF >conftest.$ac_ext /* end confdefs.h. */ #include #include #include #include int main () { ; return 0; } _ACEOF if ac_fn_c_try_compile "$LINENO"; then : ac_cv_header_stdc=yes else ac_cv_header_stdc=no fi rm -f core conftest.err conftest.$ac_objext conftest.$ac_ext if test $ac_cv_header_stdc = yes; then # SunOS 4.x string.h does not declare mem*, contrary to ANSI. cat confdefs.h - <<_ACEOF >conftest.$ac_ext /* end confdefs.h. */ #include _ACEOF if (eval "$ac_cpp conftest.$ac_ext") 2>&5 | $EGREP "memchr" >/dev/null 2>&1; then : else ac_cv_header_stdc=no fi rm -f conftest* fi if test $ac_cv_header_stdc = yes; then # ISC 2.0.2 stdlib.h does not declare free, contrary to ANSI. cat confdefs.h - <<_ACEOF >conftest.$ac_ext /* end confdefs.h. */ #include _ACEOF if (eval "$ac_cpp conftest.$ac_ext") 2>&5 | $EGREP "free" >/dev/null 2>&1; then : else ac_cv_header_stdc=no fi rm -f conftest* fi if test $ac_cv_header_stdc = yes; then # /bin/cc in Irix-4.0.5 gets non-ANSI ctype macros unless using -ansi. if test "$cross_compiling" = yes; then : : else cat confdefs.h - <<_ACEOF >conftest.$ac_ext /* end confdefs.h. */ #include #include #if ((' ' & 0x0FF) == 0x020) # define ISLOWER(c) ('a' <= (c) && (c) <= 'z') # define TOUPPER(c) (ISLOWER(c) ? 'A' + ((c) - 'a') : (c)) #else # define ISLOWER(c) \ (('a' <= (c) && (c) <= 'i') \ || ('j' <= (c) && (c) <= 'r') \ || ('s' <= (c) && (c) <= 'z')) # define TOUPPER(c) (ISLOWER(c) ? ((c) | 0x40) : (c)) #endif #define XOR(e, f) (((e) && !(f)) || (!(e) && (f))) int main () { int i; for (i = 0; i < 256; i++) if (XOR (islower (i), ISLOWER (i)) || toupper (i) != TOUPPER (i)) return 2; return 0; } _ACEOF if ac_fn_c_try_run "$LINENO"; then : else ac_cv_header_stdc=no fi rm -f core *.core core.conftest.* gmon.out bb.out conftest$ac_exeext \ conftest.$ac_objext conftest.beam conftest.$ac_ext fi fi fi { $as_echo "$as_me:${as_lineno-$LINENO}: result: $ac_cv_header_stdc" >&5 $as_echo "$ac_cv_header_stdc" >&6; } if test $ac_cv_header_stdc = yes; then $as_echo "#define STDC_HEADERS 1" >>confdefs.h fi # On IRIX 5.3, sys/types and inttypes.h are conflicting. for ac_header in sys/types.h sys/stat.h stdlib.h string.h memory.h strings.h \ inttypes.h stdint.h unistd.h do : as_ac_Header=`$as_echo "ac_cv_header_$ac_header" | $as_tr_sh` ac_fn_c_check_header_compile "$LINENO" "$ac_header" "$as_ac_Header" "$ac_includes_default " if eval test \"x\$"$as_ac_Header"\" = x"yes"; then : cat >>confdefs.h <<_ACEOF #define `$as_echo "HAVE_$ac_header" | $as_tr_cpp` 1 _ACEOF fi done for ac_header in dlfcn.h do : ac_fn_c_check_header_compile "$LINENO" "dlfcn.h" "ac_cv_header_dlfcn_h" "$ac_includes_default " if test "x$ac_cv_header_dlfcn_h" = xyes; then : cat >>confdefs.h <<_ACEOF #define HAVE_DLFCN_H 1 _ACEOF fi done # Set options enable_dlopen=no # Check whether --enable-static was given. if test "${enable_static+set}" = set; then : enableval=$enable_static; p=${PACKAGE-default} case $enableval in yes) enable_static=yes ;; no) enable_static=no ;; *) enable_static=no # Look at the argument we got. We use all the common list separators. lt_save_ifs="$IFS"; IFS="${IFS}$PATH_SEPARATOR," for pkg in $enableval; do IFS="$lt_save_ifs" if test "X$pkg" = "X$p"; then enable_static=yes fi done IFS="$lt_save_ifs" ;; esac else enable_static=yes fi # Check whether --with-pic was given. if test "${with_pic+set}" = set; then : withval=$with_pic; lt_p=${PACKAGE-default} case $withval in yes|no) pic_mode=$withval ;; *) pic_mode=default # Look at the argument we got. We use all the common list separators. lt_save_ifs="$IFS"; IFS="${IFS}$PATH_SEPARATOR," for lt_pkg in $withval; do IFS="$lt_save_ifs" if test "X$lt_pkg" = "X$lt_p"; then pic_mode=yes fi done IFS="$lt_save_ifs" ;; esac else pic_mode=default fi test -z "$pic_mode" && pic_mode=default # Check whether --enable-fast-install was given. if test "${enable_fast_install+set}" = set; then : enableval=$enable_fast_install; p=${PACKAGE-default} case $enableval in yes) enable_fast_install=yes ;; no) enable_fast_install=no ;; *) enable_fast_install=no # Look at the argument we got. We use all the common list separators. lt_save_ifs="$IFS"; IFS="${IFS}$PATH_SEPARATOR," for pkg in $enableval; do IFS="$lt_save_ifs" if test "X$pkg" = "X$p"; then enable_fast_install=yes fi done IFS="$lt_save_ifs" ;; esac else enable_fast_install=yes fi # This can be used to rebuild libtool when needed LIBTOOL_DEPS="$ltmain" # Always use our own libtool. LIBTOOL='$(SHELL) $(top_builddir)/libtool' test -z "$LN_S" && LN_S="ln -s" if test -n "${ZSH_VERSION+set}" ; then setopt NO_GLOB_SUBST fi { $as_echo "$as_me:${as_lineno-$LINENO}: checking for objdir" >&5 $as_echo_n "checking for objdir... " >&6; } if ${lt_cv_objdir+:} false; then : $as_echo_n "(cached) " >&6 else rm -f .libs 2>/dev/null mkdir .libs 2>/dev/null if test -d .libs; then lt_cv_objdir=.libs else # MS-DOS does not allow filenames that begin with a dot. lt_cv_objdir=_libs fi rmdir .libs 2>/dev/null fi { $as_echo "$as_me:${as_lineno-$LINENO}: result: $lt_cv_objdir" >&5 $as_echo "$lt_cv_objdir" >&6; } objdir=$lt_cv_objdir cat >>confdefs.h <<_ACEOF #define LT_OBJDIR "$lt_cv_objdir/" _ACEOF case $host_os in aix3*) # AIX sometimes has problems with the GCC collect2 program. For some # reason, if we set the COLLECT_NAMES environment variable, the problems # vanish in a puff of smoke. if test "X${COLLECT_NAMES+set}" != Xset; then COLLECT_NAMES= export COLLECT_NAMES fi ;; esac # Global variables: ofile=libtool can_build_shared=yes # All known linkers require a `.a' archive for static linking (except MSVC, # which needs '.lib'). libext=a with_gnu_ld="$lt_cv_prog_gnu_ld" old_CC="$CC" old_CFLAGS="$CFLAGS" # Set sane defaults for various variables test -z "$CC" && CC=cc test -z "$LTCC" && LTCC=$CC test -z "$LTCFLAGS" && LTCFLAGS=$CFLAGS test -z "$LD" && LD=ld test -z "$ac_objext" && ac_objext=o for cc_temp in $compiler""; do case $cc_temp in compile | *[\\/]compile | ccache | *[\\/]ccache ) ;; distcc | *[\\/]distcc | purify | *[\\/]purify ) ;; \-*) ;; *) break;; esac done cc_basename=`$ECHO "$cc_temp" | $SED "s%.*/%%; s%^$host_alias-%%"` # Only perform the check for file, if the check method requires it test -z "$MAGIC_CMD" && MAGIC_CMD=file case $deplibs_check_method in file_magic*) if test "$file_magic_cmd" = '$MAGIC_CMD'; then { $as_echo "$as_me:${as_lineno-$LINENO}: checking for ${ac_tool_prefix}file" >&5 $as_echo_n "checking for ${ac_tool_prefix}file... " >&6; } if ${lt_cv_path_MAGIC_CMD+:} false; then : $as_echo_n "(cached) " >&6 else case $MAGIC_CMD in [\\/*] | ?:[\\/]*) lt_cv_path_MAGIC_CMD="$MAGIC_CMD" # Let the user override the test with a path. ;; *) lt_save_MAGIC_CMD="$MAGIC_CMD" lt_save_ifs="$IFS"; IFS=$PATH_SEPARATOR ac_dummy="/usr/bin$PATH_SEPARATOR$PATH" for ac_dir in $ac_dummy; do IFS="$lt_save_ifs" test -z "$ac_dir" && ac_dir=. if test -f $ac_dir/${ac_tool_prefix}file; then lt_cv_path_MAGIC_CMD="$ac_dir/${ac_tool_prefix}file" if test -n "$file_magic_test_file"; then case $deplibs_check_method in "file_magic "*) file_magic_regex=`expr "$deplibs_check_method" : "file_magic \(.*\)"` MAGIC_CMD="$lt_cv_path_MAGIC_CMD" if eval $file_magic_cmd \$file_magic_test_file 2> /dev/null | $EGREP "$file_magic_regex" > /dev/null; then : else cat <<_LT_EOF 1>&2 *** Warning: the command libtool uses to detect shared libraries, *** $file_magic_cmd, produces output that libtool cannot recognize. *** The result is that libtool may fail to recognize shared libraries *** as such. This will affect the creation of libtool libraries that *** depend on shared libraries, but programs linked with such libtool *** libraries will work regardless of this problem. Nevertheless, you *** may want to report the problem to your system manager and/or to *** bug-libtool@gnu.org _LT_EOF fi ;; esac fi break fi done IFS="$lt_save_ifs" MAGIC_CMD="$lt_save_MAGIC_CMD" ;; esac fi MAGIC_CMD="$lt_cv_path_MAGIC_CMD" if test -n "$MAGIC_CMD"; then { $as_echo "$as_me:${as_lineno-$LINENO}: result: $MAGIC_CMD" >&5 $as_echo "$MAGIC_CMD" >&6; } else { $as_echo "$as_me:${as_lineno-$LINENO}: result: no" >&5 $as_echo "no" >&6; } fi if test -z "$lt_cv_path_MAGIC_CMD"; then if test -n "$ac_tool_prefix"; then { $as_echo "$as_me:${as_lineno-$LINENO}: checking for file" >&5 $as_echo_n "checking for file... " >&6; } if ${lt_cv_path_MAGIC_CMD+:} false; then : $as_echo_n "(cached) " >&6 else case $MAGIC_CMD in [\\/*] | ?:[\\/]*) lt_cv_path_MAGIC_CMD="$MAGIC_CMD" # Let the user override the test with a path. ;; *) lt_save_MAGIC_CMD="$MAGIC_CMD" lt_save_ifs="$IFS"; IFS=$PATH_SEPARATOR ac_dummy="/usr/bin$PATH_SEPARATOR$PATH" for ac_dir in $ac_dummy; do IFS="$lt_save_ifs" test -z "$ac_dir" && ac_dir=. if test -f $ac_dir/file; then lt_cv_path_MAGIC_CMD="$ac_dir/file" if test -n "$file_magic_test_file"; then case $deplibs_check_method in "file_magic "*) file_magic_regex=`expr "$deplibs_check_method" : "file_magic \(.*\)"` MAGIC_CMD="$lt_cv_path_MAGIC_CMD" if eval $file_magic_cmd \$file_magic_test_file 2> /dev/null | $EGREP "$file_magic_regex" > /dev/null; then : else cat <<_LT_EOF 1>&2 *** Warning: the command libtool uses to detect shared libraries, *** $file_magic_cmd, produces output that libtool cannot recognize. *** The result is that libtool may fail to recognize shared libraries *** as such. This will affect the creation of libtool libraries that *** depend on shared libraries, but programs linked with such libtool *** libraries will work regardless of this problem. Nevertheless, you *** may want to report the problem to your system manager and/or to *** bug-libtool@gnu.org _LT_EOF fi ;; esac fi break fi done IFS="$lt_save_ifs" MAGIC_CMD="$lt_save_MAGIC_CMD" ;; esac fi MAGIC_CMD="$lt_cv_path_MAGIC_CMD" if test -n "$MAGIC_CMD"; then { $as_echo "$as_me:${as_lineno-$LINENO}: result: $MAGIC_CMD" >&5 $as_echo "$MAGIC_CMD" >&6; } else { $as_echo "$as_me:${as_lineno-$LINENO}: result: no" >&5 $as_echo "no" >&6; } fi else MAGIC_CMD=: fi fi fi ;; esac # Use C for the default configuration in the libtool script lt_save_CC="$CC" ac_ext=c ac_cpp='$CPP $CPPFLAGS' ac_compile='$CC -c $CFLAGS $CPPFLAGS conftest.$ac_ext >&5' ac_link='$CC -o conftest$ac_exeext $CFLAGS $CPPFLAGS $LDFLAGS conftest.$ac_ext $LIBS >&5' ac_compiler_gnu=$ac_cv_c_compiler_gnu # Source file extension for C test sources. ac_ext=c # Object file extension for compiled C test sources. objext=o objext=$objext # Code to be used in simple compile tests lt_simple_compile_test_code="int some_variable = 0;" # Code to be used in simple link tests lt_simple_link_test_code='int main(){return(0);}' # If no C compiler was specified, use CC. LTCC=${LTCC-"$CC"} # If no C compiler flags were specified, use CFLAGS. LTCFLAGS=${LTCFLAGS-"$CFLAGS"} # Allow CC to be a program name with arguments. compiler=$CC # Save the default compiler, since it gets overwritten when the other # tags are being tested, and _LT_TAGVAR(compiler, []) is a NOP. compiler_DEFAULT=$CC # save warnings/boilerplate of simple test code ac_outfile=conftest.$ac_objext echo "$lt_simple_compile_test_code" >conftest.$ac_ext eval "$ac_compile" 2>&1 >/dev/null | $SED '/^$/d; /^ *+/d' >conftest.err _lt_compiler_boilerplate=`cat conftest.err` $RM conftest* ac_outfile=conftest.$ac_objext echo "$lt_simple_link_test_code" >conftest.$ac_ext eval "$ac_link" 2>&1 >/dev/null | $SED '/^$/d; /^ *+/d' >conftest.err _lt_linker_boilerplate=`cat conftest.err` $RM -r conftest* ## CAVEAT EMPTOR: ## There is no encapsulation within the following macros, do not change ## the running order or otherwise move them around unless you know exactly ## what you are doing... if test -n "$compiler"; then lt_prog_compiler_no_builtin_flag= if test "$GCC" = yes; then case $cc_basename in nvcc*) lt_prog_compiler_no_builtin_flag=' -Xcompiler -fno-builtin' ;; *) lt_prog_compiler_no_builtin_flag=' -fno-builtin' ;; esac { $as_echo "$as_me:${as_lineno-$LINENO}: checking if $compiler supports -fno-rtti -fno-exceptions" >&5 $as_echo_n "checking if $compiler supports -fno-rtti -fno-exceptions... " >&6; } if ${lt_cv_prog_compiler_rtti_exceptions+:} false; then : $as_echo_n "(cached) " >&6 else lt_cv_prog_compiler_rtti_exceptions=no ac_outfile=conftest.$ac_objext echo "$lt_simple_compile_test_code" > conftest.$ac_ext lt_compiler_flag="-fno-rtti -fno-exceptions" # Insert the option either (1) after the last *FLAGS variable, or # (2) before a word containing "conftest.", or (3) at the end. # Note that $ac_compile itself does not contain backslashes and begins # with a dollar sign (not a hyphen), so the echo should work correctly. # The option is referenced via a variable to avoid confusing sed. lt_compile=`echo "$ac_compile" | $SED \ -e 's:.*FLAGS}\{0,1\} :&$lt_compiler_flag :; t' \ -e 's: [^ ]*conftest\.: $lt_compiler_flag&:; t' \ -e 's:$: $lt_compiler_flag:'` (eval echo "\"\$as_me:$LINENO: $lt_compile\"" >&5) (eval "$lt_compile" 2>conftest.err) ac_status=$? cat conftest.err >&5 echo "$as_me:$LINENO: \$? = $ac_status" >&5 if (exit $ac_status) && test -s "$ac_outfile"; then # The compiler can only warn and ignore the option if not recognized # So say no if there are warnings other than the usual output. $ECHO "$_lt_compiler_boilerplate" | $SED '/^$/d' >conftest.exp $SED '/^$/d; /^ *+/d' conftest.err >conftest.er2 if test ! -s conftest.er2 || diff conftest.exp conftest.er2 >/dev/null; then lt_cv_prog_compiler_rtti_exceptions=yes fi fi $RM conftest* fi { $as_echo "$as_me:${as_lineno-$LINENO}: result: $lt_cv_prog_compiler_rtti_exceptions" >&5 $as_echo "$lt_cv_prog_compiler_rtti_exceptions" >&6; } if test x"$lt_cv_prog_compiler_rtti_exceptions" = xyes; then lt_prog_compiler_no_builtin_flag="$lt_prog_compiler_no_builtin_flag -fno-rtti -fno-exceptions" else : fi fi lt_prog_compiler_wl= lt_prog_compiler_pic= lt_prog_compiler_static= if test "$GCC" = yes; then lt_prog_compiler_wl='-Wl,' lt_prog_compiler_static='-static' case $host_os in aix*) # All AIX code is PIC. if test "$host_cpu" = ia64; then # AIX 5 now supports IA64 processor lt_prog_compiler_static='-Bstatic' fi ;; amigaos*) case $host_cpu in powerpc) # see comment about AmigaOS4 .so support lt_prog_compiler_pic='-fPIC' ;; m68k) # FIXME: we need at least 68020 code to build shared libraries, but # adding the `-m68020' flag to GCC prevents building anything better, # like `-m68040'. lt_prog_compiler_pic='-m68020 -resident32 -malways-restore-a4' ;; esac ;; beos* | irix5* | irix6* | nonstopux* | osf3* | osf4* | osf5*) # PIC is the default for these OSes. ;; mingw* | cygwin* | pw32* | os2* | cegcc*) # This hack is so that the source file can tell whether it is being # built for inclusion in a dll (and should export symbols for example). # Although the cygwin gcc ignores -fPIC, still need this for old-style # (--disable-auto-import) libraries lt_prog_compiler_pic='-DDLL_EXPORT' ;; darwin* | rhapsody*) # PIC is the default on this platform # Common symbols not allowed in MH_DYLIB files lt_prog_compiler_pic='-fno-common' ;; haiku*) # PIC is the default for Haiku. # The "-static" flag exists, but is broken. lt_prog_compiler_static= ;; hpux*) # PIC is the default for 64-bit PA HP-UX, but not for 32-bit # PA HP-UX. On IA64 HP-UX, PIC is the default but the pic flag # sets the default TLS model and affects inlining. case $host_cpu in hppa*64*) # +Z the default ;; *) lt_prog_compiler_pic='-fPIC' ;; esac ;; interix[3-9]*) # Interix 3.x gcc -fpic/-fPIC options generate broken code. # Instead, we relocate shared libraries at runtime. ;; msdosdjgpp*) # Just because we use GCC doesn't mean we suddenly get shared libraries # on systems that don't support them. lt_prog_compiler_can_build_shared=no enable_shared=no ;; *nto* | *qnx*) # QNX uses GNU C++, but need to define -shared option too, otherwise # it will coredump. lt_prog_compiler_pic='-fPIC -shared' ;; sysv4*MP*) if test -d /usr/nec; then lt_prog_compiler_pic=-Kconform_pic fi ;; *) lt_prog_compiler_pic='-fPIC' ;; esac case $cc_basename in nvcc*) # Cuda Compiler Driver 2.2 lt_prog_compiler_wl='-Xlinker ' if test -n "$lt_prog_compiler_pic"; then lt_prog_compiler_pic="-Xcompiler $lt_prog_compiler_pic" fi ;; esac else # PORTME Check for flag to pass linker flags through the system compiler. case $host_os in aix*) lt_prog_compiler_wl='-Wl,' if test "$host_cpu" = ia64; then # AIX 5 now supports IA64 processor lt_prog_compiler_static='-Bstatic' else lt_prog_compiler_static='-bnso -bI:/lib/syscalls.exp' fi ;; mingw* | cygwin* | pw32* | os2* | cegcc*) # This hack is so that the source file can tell whether it is being # built for inclusion in a dll (and should export symbols for example). lt_prog_compiler_pic='-DDLL_EXPORT' ;; hpux9* | hpux10* | hpux11*) lt_prog_compiler_wl='-Wl,' # PIC is the default for IA64 HP-UX and 64-bit HP-UX, but # not for PA HP-UX. case $host_cpu in hppa*64*|ia64*) # +Z the default ;; *) lt_prog_compiler_pic='+Z' ;; esac # Is there a better lt_prog_compiler_static that works with the bundled CC? lt_prog_compiler_static='${wl}-a ${wl}archive' ;; irix5* | irix6* | nonstopux*) lt_prog_compiler_wl='-Wl,' # PIC (with -KPIC) is the default. lt_prog_compiler_static='-non_shared' ;; linux* | k*bsd*-gnu | kopensolaris*-gnu | gnu*) case $cc_basename in # old Intel for x86_64 which still supported -KPIC. ecc*) lt_prog_compiler_wl='-Wl,' lt_prog_compiler_pic='-KPIC' lt_prog_compiler_static='-static' ;; # icc used to be incompatible with GCC. # ICC 10 doesn't accept -KPIC any more. icc* | ifort*) lt_prog_compiler_wl='-Wl,' lt_prog_compiler_pic='-fPIC' lt_prog_compiler_static='-static' ;; # Lahey Fortran 8.1. lf95*) lt_prog_compiler_wl='-Wl,' lt_prog_compiler_pic='--shared' lt_prog_compiler_static='--static' ;; nagfor*) # NAG Fortran compiler lt_prog_compiler_wl='-Wl,-Wl,,' lt_prog_compiler_pic='-PIC' lt_prog_compiler_static='-Bstatic' ;; pgcc* | pgf77* | pgf90* | pgf95* | pgfortran*) # Portland Group compilers (*not* the Pentium gcc compiler, # which looks to be a dead project) lt_prog_compiler_wl='-Wl,' lt_prog_compiler_pic='-fpic' lt_prog_compiler_static='-Bstatic' ;; ccc*) lt_prog_compiler_wl='-Wl,' # All Alpha code is PIC. lt_prog_compiler_static='-non_shared' ;; xl* | bgxl* | bgf* | mpixl*) # IBM XL C 8.0/Fortran 10.1, 11.1 on PPC and BlueGene lt_prog_compiler_wl='-Wl,' lt_prog_compiler_pic='-qpic' lt_prog_compiler_static='-qstaticlink' ;; *) case `$CC -V 2>&1 | sed 5q` in *Sun\ Ceres\ Fortran* | *Sun*Fortran*\ [1-7].* | *Sun*Fortran*\ 8.[0-3]*) # Sun Fortran 8.3 passes all unrecognized flags to the linker lt_prog_compiler_pic='-KPIC' lt_prog_compiler_static='-Bstatic' lt_prog_compiler_wl='' ;; *Sun\ F* | *Sun*Fortran*) lt_prog_compiler_pic='-KPIC' lt_prog_compiler_static='-Bstatic' lt_prog_compiler_wl='-Qoption ld ' ;; *Sun\ C*) # Sun C 5.9 lt_prog_compiler_pic='-KPIC' lt_prog_compiler_static='-Bstatic' lt_prog_compiler_wl='-Wl,' ;; *Intel*\ [CF]*Compiler*) lt_prog_compiler_wl='-Wl,' lt_prog_compiler_pic='-fPIC' lt_prog_compiler_static='-static' ;; *Portland\ Group*) lt_prog_compiler_wl='-Wl,' lt_prog_compiler_pic='-fpic' lt_prog_compiler_static='-Bstatic' ;; esac ;; esac ;; newsos6) lt_prog_compiler_pic='-KPIC' lt_prog_compiler_static='-Bstatic' ;; *nto* | *qnx*) # QNX uses GNU C++, but need to define -shared option too, otherwise # it will coredump. lt_prog_compiler_pic='-fPIC -shared' ;; osf3* | osf4* | osf5*) lt_prog_compiler_wl='-Wl,' # All OSF/1 code is PIC. lt_prog_compiler_static='-non_shared' ;; rdos*) lt_prog_compiler_static='-non_shared' ;; solaris*) lt_prog_compiler_pic='-KPIC' lt_prog_compiler_static='-Bstatic' case $cc_basename in f77* | f90* | f95* | sunf77* | sunf90* | sunf95*) lt_prog_compiler_wl='-Qoption ld ';; *) lt_prog_compiler_wl='-Wl,';; esac ;; sunos4*) lt_prog_compiler_wl='-Qoption ld ' lt_prog_compiler_pic='-PIC' lt_prog_compiler_static='-Bstatic' ;; sysv4 | sysv4.2uw2* | sysv4.3*) lt_prog_compiler_wl='-Wl,' lt_prog_compiler_pic='-KPIC' lt_prog_compiler_static='-Bstatic' ;; sysv4*MP*) if test -d /usr/nec ;then lt_prog_compiler_pic='-Kconform_pic' lt_prog_compiler_static='-Bstatic' fi ;; sysv5* | unixware* | sco3.2v5* | sco5v6* | OpenUNIX*) lt_prog_compiler_wl='-Wl,' lt_prog_compiler_pic='-KPIC' lt_prog_compiler_static='-Bstatic' ;; unicos*) lt_prog_compiler_wl='-Wl,' lt_prog_compiler_can_build_shared=no ;; uts4*) lt_prog_compiler_pic='-pic' lt_prog_compiler_static='-Bstatic' ;; *) lt_prog_compiler_can_build_shared=no ;; esac fi case $host_os in # For platforms which do not support PIC, -DPIC is meaningless: *djgpp*) lt_prog_compiler_pic= ;; *) lt_prog_compiler_pic="$lt_prog_compiler_pic -DPIC" ;; esac { $as_echo "$as_me:${as_lineno-$LINENO}: checking for $compiler option to produce PIC" >&5 $as_echo_n "checking for $compiler option to produce PIC... " >&6; } if ${lt_cv_prog_compiler_pic+:} false; then : $as_echo_n "(cached) " >&6 else lt_cv_prog_compiler_pic=$lt_prog_compiler_pic fi { $as_echo "$as_me:${as_lineno-$LINENO}: result: $lt_cv_prog_compiler_pic" >&5 $as_echo "$lt_cv_prog_compiler_pic" >&6; } lt_prog_compiler_pic=$lt_cv_prog_compiler_pic # # Check to make sure the PIC flag actually works. # if test -n "$lt_prog_compiler_pic"; then { $as_echo "$as_me:${as_lineno-$LINENO}: checking if $compiler PIC flag $lt_prog_compiler_pic works" >&5 $as_echo_n "checking if $compiler PIC flag $lt_prog_compiler_pic works... " >&6; } if ${lt_cv_prog_compiler_pic_works+:} false; then : $as_echo_n "(cached) " >&6 else lt_cv_prog_compiler_pic_works=no ac_outfile=conftest.$ac_objext echo "$lt_simple_compile_test_code" > conftest.$ac_ext lt_compiler_flag="$lt_prog_compiler_pic -DPIC" # Insert the option either (1) after the last *FLAGS variable, or # (2) before a word containing "conftest.", or (3) at the end. # Note that $ac_compile itself does not contain backslashes and begins # with a dollar sign (not a hyphen), so the echo should work correctly. # The option is referenced via a variable to avoid confusing sed. lt_compile=`echo "$ac_compile" | $SED \ -e 's:.*FLAGS}\{0,1\} :&$lt_compiler_flag :; t' \ -e 's: [^ ]*conftest\.: $lt_compiler_flag&:; t' \ -e 's:$: $lt_compiler_flag:'` (eval echo "\"\$as_me:$LINENO: $lt_compile\"" >&5) (eval "$lt_compile" 2>conftest.err) ac_status=$? cat conftest.err >&5 echo "$as_me:$LINENO: \$? = $ac_status" >&5 if (exit $ac_status) && test -s "$ac_outfile"; then # The compiler can only warn and ignore the option if not recognized # So say no if there are warnings other than the usual output. $ECHO "$_lt_compiler_boilerplate" | $SED '/^$/d' >conftest.exp $SED '/^$/d; /^ *+/d' conftest.err >conftest.er2 if test ! -s conftest.er2 || diff conftest.exp conftest.er2 >/dev/null; then lt_cv_prog_compiler_pic_works=yes fi fi $RM conftest* fi { $as_echo "$as_me:${as_lineno-$LINENO}: result: $lt_cv_prog_compiler_pic_works" >&5 $as_echo "$lt_cv_prog_compiler_pic_works" >&6; } if test x"$lt_cv_prog_compiler_pic_works" = xyes; then case $lt_prog_compiler_pic in "" | " "*) ;; *) lt_prog_compiler_pic=" $lt_prog_compiler_pic" ;; esac else lt_prog_compiler_pic= lt_prog_compiler_can_build_shared=no fi fi # # Check to make sure the static flag actually works. # wl=$lt_prog_compiler_wl eval lt_tmp_static_flag=\"$lt_prog_compiler_static\" { $as_echo "$as_me:${as_lineno-$LINENO}: checking if $compiler static flag $lt_tmp_static_flag works" >&5 $as_echo_n "checking if $compiler static flag $lt_tmp_static_flag works... " >&6; } if ${lt_cv_prog_compiler_static_works+:} false; then : $as_echo_n "(cached) " >&6 else lt_cv_prog_compiler_static_works=no save_LDFLAGS="$LDFLAGS" LDFLAGS="$LDFLAGS $lt_tmp_static_flag" echo "$lt_simple_link_test_code" > conftest.$ac_ext if (eval $ac_link 2>conftest.err) && test -s conftest$ac_exeext; then # The linker can only warn and ignore the option if not recognized # So say no if there are warnings if test -s conftest.err; then # Append any errors to the config.log. cat conftest.err 1>&5 $ECHO "$_lt_linker_boilerplate" | $SED '/^$/d' > conftest.exp $SED '/^$/d; /^ *+/d' conftest.err >conftest.er2 if diff conftest.exp conftest.er2 >/dev/null; then lt_cv_prog_compiler_static_works=yes fi else lt_cv_prog_compiler_static_works=yes fi fi $RM -r conftest* LDFLAGS="$save_LDFLAGS" fi { $as_echo "$as_me:${as_lineno-$LINENO}: result: $lt_cv_prog_compiler_static_works" >&5 $as_echo "$lt_cv_prog_compiler_static_works" >&6; } if test x"$lt_cv_prog_compiler_static_works" = xyes; then : else lt_prog_compiler_static= fi { $as_echo "$as_me:${as_lineno-$LINENO}: checking if $compiler supports -c -o file.$ac_objext" >&5 $as_echo_n "checking if $compiler supports -c -o file.$ac_objext... " >&6; } if ${lt_cv_prog_compiler_c_o+:} false; then : $as_echo_n "(cached) " >&6 else lt_cv_prog_compiler_c_o=no $RM -r conftest 2>/dev/null mkdir conftest cd conftest mkdir out echo "$lt_simple_compile_test_code" > conftest.$ac_ext lt_compiler_flag="-o out/conftest2.$ac_objext" # Insert the option either (1) after the last *FLAGS variable, or # (2) before a word containing "conftest.", or (3) at the end. # Note that $ac_compile itself does not contain backslashes and begins # with a dollar sign (not a hyphen), so the echo should work correctly. lt_compile=`echo "$ac_compile" | $SED \ -e 's:.*FLAGS}\{0,1\} :&$lt_compiler_flag :; t' \ -e 's: [^ ]*conftest\.: $lt_compiler_flag&:; t' \ -e 's:$: $lt_compiler_flag:'` (eval echo "\"\$as_me:$LINENO: $lt_compile\"" >&5) (eval "$lt_compile" 2>out/conftest.err) ac_status=$? cat out/conftest.err >&5 echo "$as_me:$LINENO: \$? = $ac_status" >&5 if (exit $ac_status) && test -s out/conftest2.$ac_objext then # The compiler can only warn and ignore the option if not recognized # So say no if there are warnings $ECHO "$_lt_compiler_boilerplate" | $SED '/^$/d' > out/conftest.exp $SED '/^$/d; /^ *+/d' out/conftest.err >out/conftest.er2 if test ! -s out/conftest.er2 || diff out/conftest.exp out/conftest.er2 >/dev/null; then lt_cv_prog_compiler_c_o=yes fi fi chmod u+w . 2>&5 $RM conftest* # SGI C++ compiler will create directory out/ii_files/ for # template instantiation test -d out/ii_files && $RM out/ii_files/* && rmdir out/ii_files $RM out/* && rmdir out cd .. $RM -r conftest $RM conftest* fi { $as_echo "$as_me:${as_lineno-$LINENO}: result: $lt_cv_prog_compiler_c_o" >&5 $as_echo "$lt_cv_prog_compiler_c_o" >&6; } { $as_echo "$as_me:${as_lineno-$LINENO}: checking if $compiler supports -c -o file.$ac_objext" >&5 $as_echo_n "checking if $compiler supports -c -o file.$ac_objext... " >&6; } if ${lt_cv_prog_compiler_c_o+:} false; then : $as_echo_n "(cached) " >&6 else lt_cv_prog_compiler_c_o=no $RM -r conftest 2>/dev/null mkdir conftest cd conftest mkdir out echo "$lt_simple_compile_test_code" > conftest.$ac_ext lt_compiler_flag="-o out/conftest2.$ac_objext" # Insert the option either (1) after the last *FLAGS variable, or # (2) before a word containing "conftest.", or (3) at the end. # Note that $ac_compile itself does not contain backslashes and begins # with a dollar sign (not a hyphen), so the echo should work correctly. lt_compile=`echo "$ac_compile" | $SED \ -e 's:.*FLAGS}\{0,1\} :&$lt_compiler_flag :; t' \ -e 's: [^ ]*conftest\.: $lt_compiler_flag&:; t' \ -e 's:$: $lt_compiler_flag:'` (eval echo "\"\$as_me:$LINENO: $lt_compile\"" >&5) (eval "$lt_compile" 2>out/conftest.err) ac_status=$? cat out/conftest.err >&5 echo "$as_me:$LINENO: \$? = $ac_status" >&5 if (exit $ac_status) && test -s out/conftest2.$ac_objext then # The compiler can only warn and ignore the option if not recognized # So say no if there are warnings $ECHO "$_lt_compiler_boilerplate" | $SED '/^$/d' > out/conftest.exp $SED '/^$/d; /^ *+/d' out/conftest.err >out/conftest.er2 if test ! -s out/conftest.er2 || diff out/conftest.exp out/conftest.er2 >/dev/null; then lt_cv_prog_compiler_c_o=yes fi fi chmod u+w . 2>&5 $RM conftest* # SGI C++ compiler will create directory out/ii_files/ for # template instantiation test -d out/ii_files && $RM out/ii_files/* && rmdir out/ii_files $RM out/* && rmdir out cd .. $RM -r conftest $RM conftest* fi { $as_echo "$as_me:${as_lineno-$LINENO}: result: $lt_cv_prog_compiler_c_o" >&5 $as_echo "$lt_cv_prog_compiler_c_o" >&6; } hard_links="nottested" if test "$lt_cv_prog_compiler_c_o" = no && test "$need_locks" != no; then # do not overwrite the value of need_locks provided by the user { $as_echo "$as_me:${as_lineno-$LINENO}: checking if we can lock with hard links" >&5 $as_echo_n "checking if we can lock with hard links... " >&6; } hard_links=yes $RM conftest* ln conftest.a conftest.b 2>/dev/null && hard_links=no touch conftest.a ln conftest.a conftest.b 2>&5 || hard_links=no ln conftest.a conftest.b 2>/dev/null && hard_links=no { $as_echo "$as_me:${as_lineno-$LINENO}: result: $hard_links" >&5 $as_echo "$hard_links" >&6; } if test "$hard_links" = no; then { $as_echo "$as_me:${as_lineno-$LINENO}: WARNING: \`$CC' does not support \`-c -o', so \`make -j' may be unsafe" >&5 $as_echo "$as_me: WARNING: \`$CC' does not support \`-c -o', so \`make -j' may be unsafe" >&2;} need_locks=warn fi else need_locks=no fi { $as_echo "$as_me:${as_lineno-$LINENO}: checking whether the $compiler linker ($LD) supports shared libraries" >&5 $as_echo_n "checking whether the $compiler linker ($LD) supports shared libraries... " >&6; } runpath_var= allow_undefined_flag= always_export_symbols=no archive_cmds= archive_expsym_cmds= compiler_needs_object=no enable_shared_with_static_runtimes=no export_dynamic_flag_spec= export_symbols_cmds='$NM $libobjs $convenience | $global_symbol_pipe | $SED '\''s/.* //'\'' | sort | uniq > $export_symbols' hardcode_automatic=no hardcode_direct=no hardcode_direct_absolute=no hardcode_libdir_flag_spec= hardcode_libdir_separator= hardcode_minus_L=no hardcode_shlibpath_var=unsupported inherit_rpath=no link_all_deplibs=unknown module_cmds= module_expsym_cmds= old_archive_from_new_cmds= old_archive_from_expsyms_cmds= thread_safe_flag_spec= whole_archive_flag_spec= # include_expsyms should be a list of space-separated symbols to be *always* # included in the symbol list include_expsyms= # exclude_expsyms can be an extended regexp of symbols to exclude # it will be wrapped by ` (' and `)$', so one must not match beginning or # end of line. Example: `a|bc|.*d.*' will exclude the symbols `a' and `bc', # as well as any symbol that contains `d'. exclude_expsyms='_GLOBAL_OFFSET_TABLE_|_GLOBAL__F[ID]_.*' # Although _GLOBAL_OFFSET_TABLE_ is a valid symbol C name, most a.out # platforms (ab)use it in PIC code, but their linkers get confused if # the symbol is explicitly referenced. Since portable code cannot # rely on this symbol name, it's probably fine to never include it in # preloaded symbol tables. # Exclude shared library initialization/finalization symbols. extract_expsyms_cmds= case $host_os in cygwin* | mingw* | pw32* | cegcc*) # FIXME: the MSVC++ port hasn't been tested in a loooong time # When not using gcc, we currently assume that we are using # Microsoft Visual C++. if test "$GCC" != yes; then with_gnu_ld=no fi ;; interix*) # we just hope/assume this is gcc and not c89 (= MSVC++) with_gnu_ld=yes ;; openbsd*) with_gnu_ld=no ;; linux* | k*bsd*-gnu | gnu*) link_all_deplibs=no ;; esac ld_shlibs=yes # On some targets, GNU ld is compatible enough with the native linker # that we're better off using the native interface for both. lt_use_gnu_ld_interface=no if test "$with_gnu_ld" = yes; then case $host_os in aix*) # The AIX port of GNU ld has always aspired to compatibility # with the native linker. However, as the warning in the GNU ld # block says, versions before 2.19.5* couldn't really create working # shared libraries, regardless of the interface used. case `$LD -v 2>&1` in *\ \(GNU\ Binutils\)\ 2.19.5*) ;; *\ \(GNU\ Binutils\)\ 2.[2-9]*) ;; *\ \(GNU\ Binutils\)\ [3-9]*) ;; *) lt_use_gnu_ld_interface=yes ;; esac ;; *) lt_use_gnu_ld_interface=yes ;; esac fi if test "$lt_use_gnu_ld_interface" = yes; then # If archive_cmds runs LD, not CC, wlarc should be empty wlarc='${wl}' # Set some defaults for GNU ld with shared library support. These # are reset later if shared libraries are not supported. Putting them # here allows them to be overridden if necessary. runpath_var=LD_RUN_PATH hardcode_libdir_flag_spec='${wl}-rpath ${wl}$libdir' export_dynamic_flag_spec='${wl}--export-dynamic' # ancient GNU ld didn't support --whole-archive et. al. if $LD --help 2>&1 | $GREP 'no-whole-archive' > /dev/null; then whole_archive_flag_spec="$wlarc"'--whole-archive$convenience '"$wlarc"'--no-whole-archive' else whole_archive_flag_spec= fi supports_anon_versioning=no case `$LD -v 2>&1` in *GNU\ gold*) supports_anon_versioning=yes ;; *\ [01].* | *\ 2.[0-9].* | *\ 2.10.*) ;; # catch versions < 2.11 *\ 2.11.93.0.2\ *) supports_anon_versioning=yes ;; # RH7.3 ... *\ 2.11.92.0.12\ *) supports_anon_versioning=yes ;; # Mandrake 8.2 ... *\ 2.11.*) ;; # other 2.11 versions *) supports_anon_versioning=yes ;; esac # See if GNU ld supports shared libraries. case $host_os in aix[3-9]*) # On AIX/PPC, the GNU linker is very broken if test "$host_cpu" != ia64; then ld_shlibs=no cat <<_LT_EOF 1>&2 *** Warning: the GNU linker, at least up to release 2.19, is reported *** to be unable to reliably create shared libraries on AIX. *** Therefore, libtool is disabling shared libraries support. If you *** really care for shared libraries, you may want to install binutils *** 2.20 or above, or modify your PATH so that a non-GNU linker is found. *** You will then need to restart the configuration process. _LT_EOF fi ;; amigaos*) case $host_cpu in powerpc) # see comment about AmigaOS4 .so support archive_cmds='$CC -shared $libobjs $deplibs $compiler_flags ${wl}-soname $wl$soname -o $lib' archive_expsym_cmds='' ;; m68k) archive_cmds='$RM $output_objdir/a2ixlibrary.data~$ECHO "#define NAME $libname" > $output_objdir/a2ixlibrary.data~$ECHO "#define LIBRARY_ID 1" >> $output_objdir/a2ixlibrary.data~$ECHO "#define VERSION $major" >> $output_objdir/a2ixlibrary.data~$ECHO "#define REVISION $revision" >> $output_objdir/a2ixlibrary.data~$AR $AR_FLAGS $lib $libobjs~$RANLIB $lib~(cd $output_objdir && a2ixlibrary -32)' hardcode_libdir_flag_spec='-L$libdir' hardcode_minus_L=yes ;; esac ;; beos*) if $LD --help 2>&1 | $GREP ': supported targets:.* elf' > /dev/null; then allow_undefined_flag=unsupported # Joseph Beckenbach says some releases of gcc # support --undefined. This deserves some investigation. FIXME archive_cmds='$CC -nostart $libobjs $deplibs $compiler_flags ${wl}-soname $wl$soname -o $lib' else ld_shlibs=no fi ;; cygwin* | mingw* | pw32* | cegcc*) # _LT_TAGVAR(hardcode_libdir_flag_spec, ) is actually meaningless, # as there is no search path for DLLs. hardcode_libdir_flag_spec='-L$libdir' export_dynamic_flag_spec='${wl}--export-all-symbols' allow_undefined_flag=unsupported always_export_symbols=no enable_shared_with_static_runtimes=yes export_symbols_cmds='$NM $libobjs $convenience | $global_symbol_pipe | $SED -e '\''/^[BCDGRS][ ]/s/.*[ ]\([^ ]*\)/\1 DATA/;s/^.*[ ]__nm__\([^ ]*\)[ ][^ ]*/\1 DATA/;/^I[ ]/d;/^[AITW][ ]/s/.* //'\'' | sort | uniq > $export_symbols' exclude_expsyms='[_]+GLOBAL_OFFSET_TABLE_|[_]+GLOBAL__[FID]_.*|[_]+head_[A-Za-z0-9_]+_dll|[A-Za-z0-9_]+_dll_iname' if $LD --help 2>&1 | $GREP 'auto-import' > /dev/null; then archive_cmds='$CC -shared $libobjs $deplibs $compiler_flags -o $output_objdir/$soname ${wl}--enable-auto-image-base -Xlinker --out-implib -Xlinker $lib' # If the export-symbols file already is a .def file (1st line # is EXPORTS), use it as is; otherwise, prepend... archive_expsym_cmds='if test "x`$SED 1q $export_symbols`" = xEXPORTS; then cp $export_symbols $output_objdir/$soname.def; else echo EXPORTS > $output_objdir/$soname.def; cat $export_symbols >> $output_objdir/$soname.def; fi~ $CC -shared $output_objdir/$soname.def $libobjs $deplibs $compiler_flags -o $output_objdir/$soname ${wl}--enable-auto-image-base -Xlinker --out-implib -Xlinker $lib' else ld_shlibs=no fi ;; haiku*) archive_cmds='$CC -shared $libobjs $deplibs $compiler_flags ${wl}-soname $wl$soname -o $lib' link_all_deplibs=yes ;; interix[3-9]*) hardcode_direct=no hardcode_shlibpath_var=no hardcode_libdir_flag_spec='${wl}-rpath,$libdir' export_dynamic_flag_spec='${wl}-E' # Hack: On Interix 3.x, we cannot compile PIC because of a broken gcc. # Instead, shared libraries are loaded at an image base (0x10000000 by # default) and relocated if they conflict, which is a slow very memory # consuming and fragmenting process. To avoid this, we pick a random, # 256 KiB-aligned image base between 0x50000000 and 0x6FFC0000 at link # time. Moving up from 0x10000000 also allows more sbrk(2) space. archive_cmds='$CC -shared $pic_flag $libobjs $deplibs $compiler_flags ${wl}-h,$soname ${wl}--image-base,`expr ${RANDOM-$$} % 4096 / 2 \* 262144 + 1342177280` -o $lib' archive_expsym_cmds='sed "s,^,_," $export_symbols >$output_objdir/$soname.expsym~$CC -shared $pic_flag $libobjs $deplibs $compiler_flags ${wl}-h,$soname ${wl}--retain-symbols-file,$output_objdir/$soname.expsym ${wl}--image-base,`expr ${RANDOM-$$} % 4096 / 2 \* 262144 + 1342177280` -o $lib' ;; gnu* | linux* | tpf* | k*bsd*-gnu | kopensolaris*-gnu) tmp_diet=no if test "$host_os" = linux-dietlibc; then case $cc_basename in diet\ *) tmp_diet=yes;; # linux-dietlibc with static linking (!diet-dyn) esac fi if $LD --help 2>&1 | $EGREP ': supported targets:.* elf' > /dev/null \ && test "$tmp_diet" = no then tmp_addflag=' $pic_flag' tmp_sharedflag='-shared' case $cc_basename,$host_cpu in pgcc*) # Portland Group C compiler whole_archive_flag_spec='${wl}--whole-archive`for conv in $convenience\"\"; do test -n \"$conv\" && new_convenience=\"$new_convenience,$conv\"; done; func_echo_all \"$new_convenience\"` ${wl}--no-whole-archive' tmp_addflag=' $pic_flag' ;; pgf77* | pgf90* | pgf95* | pgfortran*) # Portland Group f77 and f90 compilers whole_archive_flag_spec='${wl}--whole-archive`for conv in $convenience\"\"; do test -n \"$conv\" && new_convenience=\"$new_convenience,$conv\"; done; func_echo_all \"$new_convenience\"` ${wl}--no-whole-archive' tmp_addflag=' $pic_flag -Mnomain' ;; ecc*,ia64* | icc*,ia64*) # Intel C compiler on ia64 tmp_addflag=' -i_dynamic' ;; efc*,ia64* | ifort*,ia64*) # Intel Fortran compiler on ia64 tmp_addflag=' -i_dynamic -nofor_main' ;; ifc* | ifort*) # Intel Fortran compiler tmp_addflag=' -nofor_main' ;; lf95*) # Lahey Fortran 8.1 whole_archive_flag_spec= tmp_sharedflag='--shared' ;; xl[cC]* | bgxl[cC]* | mpixl[cC]*) # IBM XL C 8.0 on PPC (deal with xlf below) tmp_sharedflag='-qmkshrobj' tmp_addflag= ;; nvcc*) # Cuda Compiler Driver 2.2 whole_archive_flag_spec='${wl}--whole-archive`for conv in $convenience\"\"; do test -n \"$conv\" && new_convenience=\"$new_convenience,$conv\"; done; func_echo_all \"$new_convenience\"` ${wl}--no-whole-archive' compiler_needs_object=yes ;; esac case `$CC -V 2>&1 | sed 5q` in *Sun\ C*) # Sun C 5.9 whole_archive_flag_spec='${wl}--whole-archive`new_convenience=; for conv in $convenience\"\"; do test -z \"$conv\" || new_convenience=\"$new_convenience,$conv\"; done; func_echo_all \"$new_convenience\"` ${wl}--no-whole-archive' compiler_needs_object=yes tmp_sharedflag='-G' ;; *Sun\ F*) # Sun Fortran 8.3 tmp_sharedflag='-G' ;; esac archive_cmds='$CC '"$tmp_sharedflag""$tmp_addflag"' $libobjs $deplibs $compiler_flags ${wl}-soname $wl$soname -o $lib' if test "x$supports_anon_versioning" = xyes; then archive_expsym_cmds='echo "{ global:" > $output_objdir/$libname.ver~ cat $export_symbols | sed -e "s/\(.*\)/\1;/" >> $output_objdir/$libname.ver~ echo "local: *; };" >> $output_objdir/$libname.ver~ $CC '"$tmp_sharedflag""$tmp_addflag"' $libobjs $deplibs $compiler_flags ${wl}-soname $wl$soname ${wl}-version-script ${wl}$output_objdir/$libname.ver -o $lib' fi case $cc_basename in xlf* | bgf* | bgxlf* | mpixlf*) # IBM XL Fortran 10.1 on PPC cannot create shared libs itself whole_archive_flag_spec='--whole-archive$convenience --no-whole-archive' hardcode_libdir_flag_spec='${wl}-rpath ${wl}$libdir' archive_cmds='$LD -shared $libobjs $deplibs $linker_flags -soname $soname -o $lib' if test "x$supports_anon_versioning" = xyes; then archive_expsym_cmds='echo "{ global:" > $output_objdir/$libname.ver~ cat $export_symbols | sed -e "s/\(.*\)/\1;/" >> $output_objdir/$libname.ver~ echo "local: *; };" >> $output_objdir/$libname.ver~ $LD -shared $libobjs $deplibs $linker_flags -soname $soname -version-script $output_objdir/$libname.ver -o $lib' fi ;; esac else ld_shlibs=no fi ;; netbsd* | netbsdelf*-gnu) if echo __ELF__ | $CC -E - | $GREP __ELF__ >/dev/null; then archive_cmds='$LD -Bshareable $libobjs $deplibs $linker_flags -o $lib' wlarc= else archive_cmds='$CC -shared $pic_flag $libobjs $deplibs $compiler_flags ${wl}-soname $wl$soname -o $lib' archive_expsym_cmds='$CC -shared $pic_flag $libobjs $deplibs $compiler_flags ${wl}-soname $wl$soname ${wl}-retain-symbols-file $wl$export_symbols -o $lib' fi ;; solaris*) if $LD -v 2>&1 | $GREP 'BFD 2\.8' > /dev/null; then ld_shlibs=no cat <<_LT_EOF 1>&2 *** Warning: The releases 2.8.* of the GNU linker cannot reliably *** create shared libraries on Solaris systems. Therefore, libtool *** is disabling shared libraries support. We urge you to upgrade GNU *** binutils to release 2.9.1 or newer. Another option is to modify *** your PATH or compiler configuration so that the native linker is *** used, and then restart. _LT_EOF elif $LD --help 2>&1 | $GREP ': supported targets:.* elf' > /dev/null; then archive_cmds='$CC -shared $pic_flag $libobjs $deplibs $compiler_flags ${wl}-soname $wl$soname -o $lib' archive_expsym_cmds='$CC -shared $pic_flag $libobjs $deplibs $compiler_flags ${wl}-soname $wl$soname ${wl}-retain-symbols-file $wl$export_symbols -o $lib' else ld_shlibs=no fi ;; sysv5* | sco3.2v5* | sco5v6* | unixware* | OpenUNIX*) case `$LD -v 2>&1` in *\ [01].* | *\ 2.[0-9].* | *\ 2.1[0-5].*) ld_shlibs=no cat <<_LT_EOF 1>&2 *** Warning: Releases of the GNU linker prior to 2.16.91.0.3 can not *** reliably create shared libraries on SCO systems. Therefore, libtool *** is disabling shared libraries support. We urge you to upgrade GNU *** binutils to release 2.16.91.0.3 or newer. Another option is to modify *** your PATH or compiler configuration so that the native linker is *** used, and then restart. _LT_EOF ;; *) # For security reasons, it is highly recommended that you always # use absolute paths for naming shared libraries, and exclude the # DT_RUNPATH tag from executables and libraries. But doing so # requires that you compile everything twice, which is a pain. if $LD --help 2>&1 | $GREP ': supported targets:.* elf' > /dev/null; then hardcode_libdir_flag_spec='${wl}-rpath ${wl}$libdir' archive_cmds='$CC -shared $libobjs $deplibs $compiler_flags ${wl}-soname $wl$soname -o $lib' archive_expsym_cmds='$CC -shared $libobjs $deplibs $compiler_flags ${wl}-soname $wl$soname ${wl}-retain-symbols-file $wl$export_symbols -o $lib' else ld_shlibs=no fi ;; esac ;; sunos4*) archive_cmds='$LD -assert pure-text -Bshareable -o $lib $libobjs $deplibs $linker_flags' wlarc= hardcode_direct=yes hardcode_shlibpath_var=no ;; *) if $LD --help 2>&1 | $GREP ': supported targets:.* elf' > /dev/null; then archive_cmds='$CC -shared $pic_flag $libobjs $deplibs $compiler_flags ${wl}-soname $wl$soname -o $lib' archive_expsym_cmds='$CC -shared $pic_flag $libobjs $deplibs $compiler_flags ${wl}-soname $wl$soname ${wl}-retain-symbols-file $wl$export_symbols -o $lib' else ld_shlibs=no fi ;; esac if test "$ld_shlibs" = no; then runpath_var= hardcode_libdir_flag_spec= export_dynamic_flag_spec= whole_archive_flag_spec= fi else # PORTME fill in a description of your system's linker (not GNU ld) case $host_os in aix3*) allow_undefined_flag=unsupported always_export_symbols=yes archive_expsym_cmds='$LD -o $output_objdir/$soname $libobjs $deplibs $linker_flags -bE:$export_symbols -T512 -H512 -bM:SRE~$AR $AR_FLAGS $lib $output_objdir/$soname' # Note: this linker hardcodes the directories in LIBPATH if there # are no directories specified by -L. hardcode_minus_L=yes if test "$GCC" = yes && test -z "$lt_prog_compiler_static"; then # Neither direct hardcoding nor static linking is supported with a # broken collect2. hardcode_direct=unsupported fi ;; aix[4-9]*) if test "$host_cpu" = ia64; then # On IA64, the linker does run time linking by default, so we don't # have to do anything special. aix_use_runtimelinking=no exp_sym_flag='-Bexport' no_entry_flag="" else # If we're using GNU nm, then we don't want the "-C" option. # -C means demangle to AIX nm, but means don't demangle with GNU nm # Also, AIX nm treats weak defined symbols like other global # defined symbols, whereas GNU nm marks them as "W". if $NM -V 2>&1 | $GREP 'GNU' > /dev/null; then export_symbols_cmds='$NM -Bpg $libobjs $convenience | awk '\''{ if (((\$ 2 == "T") || (\$ 2 == "D") || (\$ 2 == "B") || (\$ 2 == "W")) && (substr(\$ 3,1,1) != ".")) { print \$ 3 } }'\'' | sort -u > $export_symbols' else export_symbols_cmds='$NM -BCpg $libobjs $convenience | awk '\''{ if (((\$ 2 == "T") || (\$ 2 == "D") || (\$ 2 == "B")) && (substr(\$ 3,1,1) != ".")) { print \$ 3 } }'\'' | sort -u > $export_symbols' fi aix_use_runtimelinking=no # Test if we are trying to use run time linking or normal # AIX style linking. If -brtl is somewhere in LDFLAGS, we # need to do runtime linking. case $host_os in aix4.[23]|aix4.[23].*|aix[5-9]*) for ld_flag in $LDFLAGS; do if (test $ld_flag = "-brtl" || test $ld_flag = "-Wl,-brtl"); then aix_use_runtimelinking=yes break fi done ;; esac exp_sym_flag='-bexport' no_entry_flag='-bnoentry' fi # When large executables or shared objects are built, AIX ld can # have problems creating the table of contents. If linking a library # or program results in "error TOC overflow" add -mminimal-toc to # CXXFLAGS/CFLAGS for g++/gcc. In the cases where that is not # enough to fix the problem, add -Wl,-bbigtoc to LDFLAGS. archive_cmds='' hardcode_direct=yes hardcode_direct_absolute=yes hardcode_libdir_separator=':' link_all_deplibs=yes file_list_spec='${wl}-f,' if test "$GCC" = yes; then case $host_os in aix4.[012]|aix4.[012].*) # We only want to do this on AIX 4.2 and lower, the check # below for broken collect2 doesn't work under 4.3+ collect2name=`${CC} -print-prog-name=collect2` if test -f "$collect2name" && strings "$collect2name" | $GREP resolve_lib_name >/dev/null then # We have reworked collect2 : else # We have old collect2 hardcode_direct=unsupported # It fails to find uninstalled libraries when the uninstalled # path is not listed in the libpath. Setting hardcode_minus_L # to unsupported forces relinking hardcode_minus_L=yes hardcode_libdir_flag_spec='-L$libdir' hardcode_libdir_separator= fi ;; esac shared_flag='-shared' if test "$aix_use_runtimelinking" = yes; then shared_flag="$shared_flag "'${wl}-G' fi link_all_deplibs=no else # not using gcc if test "$host_cpu" = ia64; then # VisualAge C++, Version 5.5 for AIX 5L for IA-64, Beta 3 Release # chokes on -Wl,-G. The following line is correct: shared_flag='-G' else if test "$aix_use_runtimelinking" = yes; then shared_flag='${wl}-G' else shared_flag='${wl}-bM:SRE' fi fi fi export_dynamic_flag_spec='${wl}-bexpall' # It seems that -bexpall does not export symbols beginning with # underscore (_), so it is better to generate a list of symbols to export. always_export_symbols=yes if test "$aix_use_runtimelinking" = yes; then # Warning - without using the other runtime loading flags (-brtl), # -berok will link without error, but may produce a broken library. allow_undefined_flag='-berok' # Determine the default libpath from the value encoded in an # empty executable. if test "${lt_cv_aix_libpath+set}" = set; then aix_libpath=$lt_cv_aix_libpath else if ${lt_cv_aix_libpath_+:} false; then : $as_echo_n "(cached) " >&6 else cat confdefs.h - <<_ACEOF >conftest.$ac_ext /* end confdefs.h. */ int main () { ; return 0; } _ACEOF if ac_fn_c_try_link "$LINENO"; then : lt_aix_libpath_sed=' /Import File Strings/,/^$/ { /^0/ { s/^0 *\([^ ]*\) *$/\1/ p } }' lt_cv_aix_libpath_=`dump -H conftest$ac_exeext 2>/dev/null | $SED -n -e "$lt_aix_libpath_sed"` # Check for a 64-bit object if we didn't find anything. if test -z "$lt_cv_aix_libpath_"; then lt_cv_aix_libpath_=`dump -HX64 conftest$ac_exeext 2>/dev/null | $SED -n -e "$lt_aix_libpath_sed"` fi fi rm -f core conftest.err conftest.$ac_objext \ conftest$ac_exeext conftest.$ac_ext if test -z "$lt_cv_aix_libpath_"; then lt_cv_aix_libpath_="/usr/lib:/lib" fi fi aix_libpath=$lt_cv_aix_libpath_ fi hardcode_libdir_flag_spec='${wl}-blibpath:$libdir:'"$aix_libpath" archive_expsym_cmds='$CC -o $output_objdir/$soname $libobjs $deplibs '"\${wl}$no_entry_flag"' $compiler_flags `if test "x${allow_undefined_flag}" != "x"; then func_echo_all "${wl}${allow_undefined_flag}"; else :; fi` '"\${wl}$exp_sym_flag:\$export_symbols $shared_flag" else if test "$host_cpu" = ia64; then hardcode_libdir_flag_spec='${wl}-R $libdir:/usr/lib:/lib' allow_undefined_flag="-z nodefs" archive_expsym_cmds="\$CC $shared_flag"' -o $output_objdir/$soname $libobjs $deplibs '"\${wl}$no_entry_flag"' $compiler_flags ${wl}${allow_undefined_flag} '"\${wl}$exp_sym_flag:\$export_symbols" else # Determine the default libpath from the value encoded in an # empty executable. if test "${lt_cv_aix_libpath+set}" = set; then aix_libpath=$lt_cv_aix_libpath else if ${lt_cv_aix_libpath_+:} false; then : $as_echo_n "(cached) " >&6 else cat confdefs.h - <<_ACEOF >conftest.$ac_ext /* end confdefs.h. */ int main () { ; return 0; } _ACEOF if ac_fn_c_try_link "$LINENO"; then : lt_aix_libpath_sed=' /Import File Strings/,/^$/ { /^0/ { s/^0 *\([^ ]*\) *$/\1/ p } }' lt_cv_aix_libpath_=`dump -H conftest$ac_exeext 2>/dev/null | $SED -n -e "$lt_aix_libpath_sed"` # Check for a 64-bit object if we didn't find anything. if test -z "$lt_cv_aix_libpath_"; then lt_cv_aix_libpath_=`dump -HX64 conftest$ac_exeext 2>/dev/null | $SED -n -e "$lt_aix_libpath_sed"` fi fi rm -f core conftest.err conftest.$ac_objext \ conftest$ac_exeext conftest.$ac_ext if test -z "$lt_cv_aix_libpath_"; then lt_cv_aix_libpath_="/usr/lib:/lib" fi fi aix_libpath=$lt_cv_aix_libpath_ fi hardcode_libdir_flag_spec='${wl}-blibpath:$libdir:'"$aix_libpath" # Warning - without using the other run time loading flags, # -berok will link without error, but may produce a broken library. no_undefined_flag=' ${wl}-bernotok' allow_undefined_flag=' ${wl}-berok' if test "$with_gnu_ld" = yes; then # We only use this code for GNU lds that support --whole-archive. whole_archive_flag_spec='${wl}--whole-archive$convenience ${wl}--no-whole-archive' else # Exported symbols can be pulled into shared objects from archives whole_archive_flag_spec='$convenience' fi archive_cmds_need_lc=yes # This is similar to how AIX traditionally builds its shared libraries. archive_expsym_cmds="\$CC $shared_flag"' -o $output_objdir/$soname $libobjs $deplibs ${wl}-bnoentry $compiler_flags ${wl}-bE:$export_symbols${allow_undefined_flag}~$AR $AR_FLAGS $output_objdir/$libname$release.a $output_objdir/$soname' fi fi ;; amigaos*) case $host_cpu in powerpc) # see comment about AmigaOS4 .so support archive_cmds='$CC -shared $libobjs $deplibs $compiler_flags ${wl}-soname $wl$soname -o $lib' archive_expsym_cmds='' ;; m68k) archive_cmds='$RM $output_objdir/a2ixlibrary.data~$ECHO "#define NAME $libname" > $output_objdir/a2ixlibrary.data~$ECHO "#define LIBRARY_ID 1" >> $output_objdir/a2ixlibrary.data~$ECHO "#define VERSION $major" >> $output_objdir/a2ixlibrary.data~$ECHO "#define REVISION $revision" >> $output_objdir/a2ixlibrary.data~$AR $AR_FLAGS $lib $libobjs~$RANLIB $lib~(cd $output_objdir && a2ixlibrary -32)' hardcode_libdir_flag_spec='-L$libdir' hardcode_minus_L=yes ;; esac ;; bsdi[45]*) export_dynamic_flag_spec=-rdynamic ;; cygwin* | mingw* | pw32* | cegcc*) # When not using gcc, we currently assume that we are using # Microsoft Visual C++. # hardcode_libdir_flag_spec is actually meaningless, as there is # no search path for DLLs. case $cc_basename in cl*) # Native MSVC hardcode_libdir_flag_spec=' ' allow_undefined_flag=unsupported always_export_symbols=yes file_list_spec='@' # Tell ltmain to make .lib files, not .a files. libext=lib # Tell ltmain to make .dll files, not .so files. shrext_cmds=".dll" # FIXME: Setting linknames here is a bad hack. archive_cmds='$CC -o $output_objdir/$soname $libobjs $compiler_flags $deplibs -Wl,-dll~linknames=' archive_expsym_cmds='if test "x`$SED 1q $export_symbols`" = xEXPORTS; then sed -n -e 's/\\\\\\\(.*\\\\\\\)/-link\\\ -EXPORT:\\\\\\\1/' -e '1\\\!p' < $export_symbols > $output_objdir/$soname.exp; else sed -e 's/\\\\\\\(.*\\\\\\\)/-link\\\ -EXPORT:\\\\\\\1/' < $export_symbols > $output_objdir/$soname.exp; fi~ $CC -o $tool_output_objdir$soname $libobjs $compiler_flags $deplibs "@$tool_output_objdir$soname.exp" -Wl,-DLL,-IMPLIB:"$tool_output_objdir$libname.dll.lib"~ linknames=' # The linker will not automatically build a static lib if we build a DLL. # _LT_TAGVAR(old_archive_from_new_cmds, )='true' enable_shared_with_static_runtimes=yes exclude_expsyms='_NULL_IMPORT_DESCRIPTOR|_IMPORT_DESCRIPTOR_.*' export_symbols_cmds='$NM $libobjs $convenience | $global_symbol_pipe | $SED -e '\''/^[BCDGRS][ ]/s/.*[ ]\([^ ]*\)/\1,DATA/'\'' | $SED -e '\''/^[AITW][ ]/s/.*[ ]//'\'' | sort | uniq > $export_symbols' # Don't use ranlib old_postinstall_cmds='chmod 644 $oldlib' postlink_cmds='lt_outputfile="@OUTPUT@"~ lt_tool_outputfile="@TOOL_OUTPUT@"~ case $lt_outputfile in *.exe|*.EXE) ;; *) lt_outputfile="$lt_outputfile.exe" lt_tool_outputfile="$lt_tool_outputfile.exe" ;; esac~ if test "$MANIFEST_TOOL" != ":" && test -f "$lt_outputfile.manifest"; then $MANIFEST_TOOL -manifest "$lt_tool_outputfile.manifest" -outputresource:"$lt_tool_outputfile" || exit 1; $RM "$lt_outputfile.manifest"; fi' ;; *) # Assume MSVC wrapper hardcode_libdir_flag_spec=' ' allow_undefined_flag=unsupported # Tell ltmain to make .lib files, not .a files. libext=lib # Tell ltmain to make .dll files, not .so files. shrext_cmds=".dll" # FIXME: Setting linknames here is a bad hack. archive_cmds='$CC -o $lib $libobjs $compiler_flags `func_echo_all "$deplibs" | $SED '\''s/ -lc$//'\''` -link -dll~linknames=' # The linker will automatically build a .lib file if we build a DLL. old_archive_from_new_cmds='true' # FIXME: Should let the user specify the lib program. old_archive_cmds='lib -OUT:$oldlib$oldobjs$old_deplibs' enable_shared_with_static_runtimes=yes ;; esac ;; darwin* | rhapsody*) archive_cmds_need_lc=no hardcode_direct=no hardcode_automatic=yes hardcode_shlibpath_var=unsupported if test "$lt_cv_ld_force_load" = "yes"; then whole_archive_flag_spec='`for conv in $convenience\"\"; do test -n \"$conv\" && new_convenience=\"$new_convenience ${wl}-force_load,$conv\"; done; func_echo_all \"$new_convenience\"`' else whole_archive_flag_spec='' fi link_all_deplibs=yes allow_undefined_flag="$_lt_dar_allow_undefined" case $cc_basename in ifort*) _lt_dar_can_shared=yes ;; *) _lt_dar_can_shared=$GCC ;; esac if test "$_lt_dar_can_shared" = "yes"; then output_verbose_link_cmd=func_echo_all archive_cmds="\$CC -dynamiclib \$allow_undefined_flag -o \$lib \$libobjs \$deplibs \$compiler_flags -install_name \$rpath/\$soname \$verstring $_lt_dar_single_mod${_lt_dsymutil}" module_cmds="\$CC \$allow_undefined_flag -o \$lib -bundle \$libobjs \$deplibs \$compiler_flags${_lt_dsymutil}" archive_expsym_cmds="sed 's,^,_,' < \$export_symbols > \$output_objdir/\${libname}-symbols.expsym~\$CC -dynamiclib \$allow_undefined_flag -o \$lib \$libobjs \$deplibs \$compiler_flags -install_name \$rpath/\$soname \$verstring ${_lt_dar_single_mod}${_lt_dar_export_syms}${_lt_dsymutil}" module_expsym_cmds="sed -e 's,^,_,' < \$export_symbols > \$output_objdir/\${libname}-symbols.expsym~\$CC \$allow_undefined_flag -o \$lib -bundle \$libobjs \$deplibs \$compiler_flags${_lt_dar_export_syms}${_lt_dsymutil}" else ld_shlibs=no fi ;; dgux*) archive_cmds='$LD -G -h $soname -o $lib $libobjs $deplibs $linker_flags' hardcode_libdir_flag_spec='-L$libdir' hardcode_shlibpath_var=no ;; # FreeBSD 2.2.[012] allows us to include c++rt0.o to get C++ constructor # support. Future versions do this automatically, but an explicit c++rt0.o # does not break anything, and helps significantly (at the cost of a little # extra space). freebsd2.2*) archive_cmds='$LD -Bshareable -o $lib $libobjs $deplibs $linker_flags /usr/lib/c++rt0.o' hardcode_libdir_flag_spec='-R$libdir' hardcode_direct=yes hardcode_shlibpath_var=no ;; # Unfortunately, older versions of FreeBSD 2 do not have this feature. freebsd2.*) archive_cmds='$LD -Bshareable -o $lib $libobjs $deplibs $linker_flags' hardcode_direct=yes hardcode_minus_L=yes hardcode_shlibpath_var=no ;; # FreeBSD 3 and greater uses gcc -shared to do shared libraries. freebsd* | dragonfly*) archive_cmds='$CC -shared $pic_flag -o $lib $libobjs $deplibs $compiler_flags' hardcode_libdir_flag_spec='-R$libdir' hardcode_direct=yes hardcode_shlibpath_var=no ;; hpux9*) if test "$GCC" = yes; then archive_cmds='$RM $output_objdir/$soname~$CC -shared $pic_flag ${wl}+b ${wl}$install_libdir -o $output_objdir/$soname $libobjs $deplibs $compiler_flags~test $output_objdir/$soname = $lib || mv $output_objdir/$soname $lib' else archive_cmds='$RM $output_objdir/$soname~$LD -b +b $install_libdir -o $output_objdir/$soname $libobjs $deplibs $linker_flags~test $output_objdir/$soname = $lib || mv $output_objdir/$soname $lib' fi hardcode_libdir_flag_spec='${wl}+b ${wl}$libdir' hardcode_libdir_separator=: hardcode_direct=yes # hardcode_minus_L: Not really in the search PATH, # but as the default location of the library. hardcode_minus_L=yes export_dynamic_flag_spec='${wl}-E' ;; hpux10*) if test "$GCC" = yes && test "$with_gnu_ld" = no; then archive_cmds='$CC -shared $pic_flag ${wl}+h ${wl}$soname ${wl}+b ${wl}$install_libdir -o $lib $libobjs $deplibs $compiler_flags' else archive_cmds='$LD -b +h $soname +b $install_libdir -o $lib $libobjs $deplibs $linker_flags' fi if test "$with_gnu_ld" = no; then hardcode_libdir_flag_spec='${wl}+b ${wl}$libdir' hardcode_libdir_separator=: hardcode_direct=yes hardcode_direct_absolute=yes export_dynamic_flag_spec='${wl}-E' # hardcode_minus_L: Not really in the search PATH, # but as the default location of the library. hardcode_minus_L=yes fi ;; hpux11*) if test "$GCC" = yes && test "$with_gnu_ld" = no; then case $host_cpu in hppa*64*) archive_cmds='$CC -shared ${wl}+h ${wl}$soname -o $lib $libobjs $deplibs $compiler_flags' ;; ia64*) archive_cmds='$CC -shared $pic_flag ${wl}+h ${wl}$soname ${wl}+nodefaultrpath -o $lib $libobjs $deplibs $compiler_flags' ;; *) archive_cmds='$CC -shared $pic_flag ${wl}+h ${wl}$soname ${wl}+b ${wl}$install_libdir -o $lib $libobjs $deplibs $compiler_flags' ;; esac else case $host_cpu in hppa*64*) archive_cmds='$CC -b ${wl}+h ${wl}$soname -o $lib $libobjs $deplibs $compiler_flags' ;; ia64*) archive_cmds='$CC -b ${wl}+h ${wl}$soname ${wl}+nodefaultrpath -o $lib $libobjs $deplibs $compiler_flags' ;; *) # Older versions of the 11.00 compiler do not understand -b yet # (HP92453-01 A.11.01.20 doesn't, HP92453-01 B.11.X.35175-35176.GP does) { $as_echo "$as_me:${as_lineno-$LINENO}: checking if $CC understands -b" >&5 $as_echo_n "checking if $CC understands -b... " >&6; } if ${lt_cv_prog_compiler__b+:} false; then : $as_echo_n "(cached) " >&6 else lt_cv_prog_compiler__b=no save_LDFLAGS="$LDFLAGS" LDFLAGS="$LDFLAGS -b" echo "$lt_simple_link_test_code" > conftest.$ac_ext if (eval $ac_link 2>conftest.err) && test -s conftest$ac_exeext; then # The linker can only warn and ignore the option if not recognized # So say no if there are warnings if test -s conftest.err; then # Append any errors to the config.log. cat conftest.err 1>&5 $ECHO "$_lt_linker_boilerplate" | $SED '/^$/d' > conftest.exp $SED '/^$/d; /^ *+/d' conftest.err >conftest.er2 if diff conftest.exp conftest.er2 >/dev/null; then lt_cv_prog_compiler__b=yes fi else lt_cv_prog_compiler__b=yes fi fi $RM -r conftest* LDFLAGS="$save_LDFLAGS" fi { $as_echo "$as_me:${as_lineno-$LINENO}: result: $lt_cv_prog_compiler__b" >&5 $as_echo "$lt_cv_prog_compiler__b" >&6; } if test x"$lt_cv_prog_compiler__b" = xyes; then archive_cmds='$CC -b ${wl}+h ${wl}$soname ${wl}+b ${wl}$install_libdir -o $lib $libobjs $deplibs $compiler_flags' else archive_cmds='$LD -b +h $soname +b $install_libdir -o $lib $libobjs $deplibs $linker_flags' fi ;; esac fi if test "$with_gnu_ld" = no; then hardcode_libdir_flag_spec='${wl}+b ${wl}$libdir' hardcode_libdir_separator=: case $host_cpu in hppa*64*|ia64*) hardcode_direct=no hardcode_shlibpath_var=no ;; *) hardcode_direct=yes hardcode_direct_absolute=yes export_dynamic_flag_spec='${wl}-E' # hardcode_minus_L: Not really in the search PATH, # but as the default location of the library. hardcode_minus_L=yes ;; esac fi ;; irix5* | irix6* | nonstopux*) if test "$GCC" = yes; then archive_cmds='$CC -shared $pic_flag $libobjs $deplibs $compiler_flags ${wl}-soname ${wl}$soname `test -n "$verstring" && func_echo_all "${wl}-set_version ${wl}$verstring"` ${wl}-update_registry ${wl}${output_objdir}/so_locations -o $lib' # Try to use the -exported_symbol ld option, if it does not # work, assume that -exports_file does not work either and # implicitly export all symbols. # This should be the same for all languages, so no per-tag cache variable. { $as_echo "$as_me:${as_lineno-$LINENO}: checking whether the $host_os linker accepts -exported_symbol" >&5 $as_echo_n "checking whether the $host_os linker accepts -exported_symbol... " >&6; } if ${lt_cv_irix_exported_symbol+:} false; then : $as_echo_n "(cached) " >&6 else save_LDFLAGS="$LDFLAGS" LDFLAGS="$LDFLAGS -shared ${wl}-exported_symbol ${wl}foo ${wl}-update_registry ${wl}/dev/null" cat confdefs.h - <<_ACEOF >conftest.$ac_ext /* end confdefs.h. */ int foo (void) { return 0; } _ACEOF if ac_fn_c_try_link "$LINENO"; then : lt_cv_irix_exported_symbol=yes else lt_cv_irix_exported_symbol=no fi rm -f core conftest.err conftest.$ac_objext \ conftest$ac_exeext conftest.$ac_ext LDFLAGS="$save_LDFLAGS" fi { $as_echo "$as_me:${as_lineno-$LINENO}: result: $lt_cv_irix_exported_symbol" >&5 $as_echo "$lt_cv_irix_exported_symbol" >&6; } if test "$lt_cv_irix_exported_symbol" = yes; then archive_expsym_cmds='$CC -shared $pic_flag $libobjs $deplibs $compiler_flags ${wl}-soname ${wl}$soname `test -n "$verstring" && func_echo_all "${wl}-set_version ${wl}$verstring"` ${wl}-update_registry ${wl}${output_objdir}/so_locations ${wl}-exports_file ${wl}$export_symbols -o $lib' fi else archive_cmds='$CC -shared $libobjs $deplibs $compiler_flags -soname $soname `test -n "$verstring" && func_echo_all "-set_version $verstring"` -update_registry ${output_objdir}/so_locations -o $lib' archive_expsym_cmds='$CC -shared $libobjs $deplibs $compiler_flags -soname $soname `test -n "$verstring" && func_echo_all "-set_version $verstring"` -update_registry ${output_objdir}/so_locations -exports_file $export_symbols -o $lib' fi archive_cmds_need_lc='no' hardcode_libdir_flag_spec='${wl}-rpath ${wl}$libdir' hardcode_libdir_separator=: inherit_rpath=yes link_all_deplibs=yes ;; netbsd* | netbsdelf*-gnu) if echo __ELF__ | $CC -E - | $GREP __ELF__ >/dev/null; then archive_cmds='$LD -Bshareable -o $lib $libobjs $deplibs $linker_flags' # a.out else archive_cmds='$LD -shared -o $lib $libobjs $deplibs $linker_flags' # ELF fi hardcode_libdir_flag_spec='-R$libdir' hardcode_direct=yes hardcode_shlibpath_var=no ;; newsos6) archive_cmds='$LD -G -h $soname -o $lib $libobjs $deplibs $linker_flags' hardcode_direct=yes hardcode_libdir_flag_spec='${wl}-rpath ${wl}$libdir' hardcode_libdir_separator=: hardcode_shlibpath_var=no ;; *nto* | *qnx*) ;; openbsd*) if test -f /usr/libexec/ld.so; then hardcode_direct=yes hardcode_shlibpath_var=no hardcode_direct_absolute=yes if test -z "`echo __ELF__ | $CC -E - | $GREP __ELF__`" || test "$host_os-$host_cpu" = "openbsd2.8-powerpc"; then archive_cmds='$CC -shared $pic_flag -o $lib $libobjs $deplibs $compiler_flags' archive_expsym_cmds='$CC -shared $pic_flag -o $lib $libobjs $deplibs $compiler_flags ${wl}-retain-symbols-file,$export_symbols' hardcode_libdir_flag_spec='${wl}-rpath,$libdir' export_dynamic_flag_spec='${wl}-E' else case $host_os in openbsd[01].* | openbsd2.[0-7] | openbsd2.[0-7].*) archive_cmds='$LD -Bshareable -o $lib $libobjs $deplibs $linker_flags' hardcode_libdir_flag_spec='-R$libdir' ;; *) archive_cmds='$CC -shared $pic_flag -o $lib $libobjs $deplibs $compiler_flags' hardcode_libdir_flag_spec='${wl}-rpath,$libdir' ;; esac fi else ld_shlibs=no fi ;; os2*) hardcode_libdir_flag_spec='-L$libdir' hardcode_minus_L=yes allow_undefined_flag=unsupported archive_cmds='$ECHO "LIBRARY $libname INITINSTANCE" > $output_objdir/$libname.def~$ECHO "DESCRIPTION \"$libname\"" >> $output_objdir/$libname.def~echo DATA >> $output_objdir/$libname.def~echo " SINGLE NONSHARED" >> $output_objdir/$libname.def~echo EXPORTS >> $output_objdir/$libname.def~emxexp $libobjs >> $output_objdir/$libname.def~$CC -Zdll -Zcrtdll -o $lib $libobjs $deplibs $compiler_flags $output_objdir/$libname.def' old_archive_from_new_cmds='emximp -o $output_objdir/$libname.a $output_objdir/$libname.def' ;; osf3*) if test "$GCC" = yes; then allow_undefined_flag=' ${wl}-expect_unresolved ${wl}\*' archive_cmds='$CC -shared${allow_undefined_flag} $libobjs $deplibs $compiler_flags ${wl}-soname ${wl}$soname `test -n "$verstring" && func_echo_all "${wl}-set_version ${wl}$verstring"` ${wl}-update_registry ${wl}${output_objdir}/so_locations -o $lib' else allow_undefined_flag=' -expect_unresolved \*' archive_cmds='$CC -shared${allow_undefined_flag} $libobjs $deplibs $compiler_flags -soname $soname `test -n "$verstring" && func_echo_all "-set_version $verstring"` -update_registry ${output_objdir}/so_locations -o $lib' fi archive_cmds_need_lc='no' hardcode_libdir_flag_spec='${wl}-rpath ${wl}$libdir' hardcode_libdir_separator=: ;; osf4* | osf5*) # as osf3* with the addition of -msym flag if test "$GCC" = yes; then allow_undefined_flag=' ${wl}-expect_unresolved ${wl}\*' archive_cmds='$CC -shared${allow_undefined_flag} $pic_flag $libobjs $deplibs $compiler_flags ${wl}-msym ${wl}-soname ${wl}$soname `test -n "$verstring" && func_echo_all "${wl}-set_version ${wl}$verstring"` ${wl}-update_registry ${wl}${output_objdir}/so_locations -o $lib' hardcode_libdir_flag_spec='${wl}-rpath ${wl}$libdir' else allow_undefined_flag=' -expect_unresolved \*' archive_cmds='$CC -shared${allow_undefined_flag} $libobjs $deplibs $compiler_flags -msym -soname $soname `test -n "$verstring" && func_echo_all "-set_version $verstring"` -update_registry ${output_objdir}/so_locations -o $lib' archive_expsym_cmds='for i in `cat $export_symbols`; do printf "%s %s\\n" -exported_symbol "\$i" >> $lib.exp; done; printf "%s\\n" "-hidden">> $lib.exp~ $CC -shared${allow_undefined_flag} ${wl}-input ${wl}$lib.exp $compiler_flags $libobjs $deplibs -soname $soname `test -n "$verstring" && $ECHO "-set_version $verstring"` -update_registry ${output_objdir}/so_locations -o $lib~$RM $lib.exp' # Both c and cxx compiler support -rpath directly hardcode_libdir_flag_spec='-rpath $libdir' fi archive_cmds_need_lc='no' hardcode_libdir_separator=: ;; solaris*) no_undefined_flag=' -z defs' if test "$GCC" = yes; then wlarc='${wl}' archive_cmds='$CC -shared $pic_flag ${wl}-z ${wl}text ${wl}-h ${wl}$soname -o $lib $libobjs $deplibs $compiler_flags' archive_expsym_cmds='echo "{ global:" > $lib.exp~cat $export_symbols | $SED -e "s/\(.*\)/\1;/" >> $lib.exp~echo "local: *; };" >> $lib.exp~ $CC -shared $pic_flag ${wl}-z ${wl}text ${wl}-M ${wl}$lib.exp ${wl}-h ${wl}$soname -o $lib $libobjs $deplibs $compiler_flags~$RM $lib.exp' else case `$CC -V 2>&1` in *"Compilers 5.0"*) wlarc='' archive_cmds='$LD -G${allow_undefined_flag} -h $soname -o $lib $libobjs $deplibs $linker_flags' archive_expsym_cmds='echo "{ global:" > $lib.exp~cat $export_symbols | $SED -e "s/\(.*\)/\1;/" >> $lib.exp~echo "local: *; };" >> $lib.exp~ $LD -G${allow_undefined_flag} -M $lib.exp -h $soname -o $lib $libobjs $deplibs $linker_flags~$RM $lib.exp' ;; *) wlarc='${wl}' archive_cmds='$CC -G${allow_undefined_flag} -h $soname -o $lib $libobjs $deplibs $compiler_flags' archive_expsym_cmds='echo "{ global:" > $lib.exp~cat $export_symbols | $SED -e "s/\(.*\)/\1;/" >> $lib.exp~echo "local: *; };" >> $lib.exp~ $CC -G${allow_undefined_flag} -M $lib.exp -h $soname -o $lib $libobjs $deplibs $compiler_flags~$RM $lib.exp' ;; esac fi hardcode_libdir_flag_spec='-R$libdir' hardcode_shlibpath_var=no case $host_os in solaris2.[0-5] | solaris2.[0-5].*) ;; *) # The compiler driver will combine and reorder linker options, # but understands `-z linker_flag'. GCC discards it without `$wl', # but is careful enough not to reorder. # Supported since Solaris 2.6 (maybe 2.5.1?) if test "$GCC" = yes; then whole_archive_flag_spec='${wl}-z ${wl}allextract$convenience ${wl}-z ${wl}defaultextract' else whole_archive_flag_spec='-z allextract$convenience -z defaultextract' fi ;; esac link_all_deplibs=yes ;; sunos4*) if test "x$host_vendor" = xsequent; then # Use $CC to link under sequent, because it throws in some extra .o # files that make .init and .fini sections work. archive_cmds='$CC -G ${wl}-h $soname -o $lib $libobjs $deplibs $compiler_flags' else archive_cmds='$LD -assert pure-text -Bstatic -o $lib $libobjs $deplibs $linker_flags' fi hardcode_libdir_flag_spec='-L$libdir' hardcode_direct=yes hardcode_minus_L=yes hardcode_shlibpath_var=no ;; sysv4) case $host_vendor in sni) archive_cmds='$LD -G -h $soname -o $lib $libobjs $deplibs $linker_flags' hardcode_direct=yes # is this really true??? ;; siemens) ## LD is ld it makes a PLAMLIB ## CC just makes a GrossModule. archive_cmds='$LD -G -o $lib $libobjs $deplibs $linker_flags' reload_cmds='$CC -r -o $output$reload_objs' hardcode_direct=no ;; motorola) archive_cmds='$LD -G -h $soname -o $lib $libobjs $deplibs $linker_flags' hardcode_direct=no #Motorola manual says yes, but my tests say they lie ;; esac runpath_var='LD_RUN_PATH' hardcode_shlibpath_var=no ;; sysv4.3*) archive_cmds='$LD -G -h $soname -o $lib $libobjs $deplibs $linker_flags' hardcode_shlibpath_var=no export_dynamic_flag_spec='-Bexport' ;; sysv4*MP*) if test -d /usr/nec; then archive_cmds='$LD -G -h $soname -o $lib $libobjs $deplibs $linker_flags' hardcode_shlibpath_var=no runpath_var=LD_RUN_PATH hardcode_runpath_var=yes ld_shlibs=yes fi ;; sysv4*uw2* | sysv5OpenUNIX* | sysv5UnixWare7.[01].[10]* | unixware7* | sco3.2v5.0.[024]*) no_undefined_flag='${wl}-z,text' archive_cmds_need_lc=no hardcode_shlibpath_var=no runpath_var='LD_RUN_PATH' if test "$GCC" = yes; then archive_cmds='$CC -shared ${wl}-h,$soname -o $lib $libobjs $deplibs $compiler_flags' archive_expsym_cmds='$CC -shared ${wl}-Bexport:$export_symbols ${wl}-h,$soname -o $lib $libobjs $deplibs $compiler_flags' else archive_cmds='$CC -G ${wl}-h,$soname -o $lib $libobjs $deplibs $compiler_flags' archive_expsym_cmds='$CC -G ${wl}-Bexport:$export_symbols ${wl}-h,$soname -o $lib $libobjs $deplibs $compiler_flags' fi ;; sysv5* | sco3.2v5* | sco5v6*) # Note: We can NOT use -z defs as we might desire, because we do not # link with -lc, and that would cause any symbols used from libc to # always be unresolved, which means just about no library would # ever link correctly. If we're not using GNU ld we use -z text # though, which does catch some bad symbols but isn't as heavy-handed # as -z defs. no_undefined_flag='${wl}-z,text' allow_undefined_flag='${wl}-z,nodefs' archive_cmds_need_lc=no hardcode_shlibpath_var=no hardcode_libdir_flag_spec='${wl}-R,$libdir' hardcode_libdir_separator=':' link_all_deplibs=yes export_dynamic_flag_spec='${wl}-Bexport' runpath_var='LD_RUN_PATH' if test "$GCC" = yes; then archive_cmds='$CC -shared ${wl}-h,$soname -o $lib $libobjs $deplibs $compiler_flags' archive_expsym_cmds='$CC -shared ${wl}-Bexport:$export_symbols ${wl}-h,$soname -o $lib $libobjs $deplibs $compiler_flags' else archive_cmds='$CC -G ${wl}-h,$soname -o $lib $libobjs $deplibs $compiler_flags' archive_expsym_cmds='$CC -G ${wl}-Bexport:$export_symbols ${wl}-h,$soname -o $lib $libobjs $deplibs $compiler_flags' fi ;; uts4*) archive_cmds='$LD -G -h $soname -o $lib $libobjs $deplibs $linker_flags' hardcode_libdir_flag_spec='-L$libdir' hardcode_shlibpath_var=no ;; *) ld_shlibs=no ;; esac if test x$host_vendor = xsni; then case $host in sysv4 | sysv4.2uw2* | sysv4.3* | sysv5*) export_dynamic_flag_spec='${wl}-Blargedynsym' ;; esac fi fi { $as_echo "$as_me:${as_lineno-$LINENO}: result: $ld_shlibs" >&5 $as_echo "$ld_shlibs" >&6; } test "$ld_shlibs" = no && can_build_shared=no with_gnu_ld=$with_gnu_ld # # Do we need to explicitly link libc? # case "x$archive_cmds_need_lc" in x|xyes) # Assume -lc should be added archive_cmds_need_lc=yes if test "$enable_shared" = yes && test "$GCC" = yes; then case $archive_cmds in *'~'*) # FIXME: we may have to deal with multi-command sequences. ;; '$CC '*) # Test whether the compiler implicitly links with -lc since on some # systems, -lgcc has to come before -lc. If gcc already passes -lc # to ld, don't add -lc before -lgcc. { $as_echo "$as_me:${as_lineno-$LINENO}: checking whether -lc should be explicitly linked in" >&5 $as_echo_n "checking whether -lc should be explicitly linked in... " >&6; } if ${lt_cv_archive_cmds_need_lc+:} false; then : $as_echo_n "(cached) " >&6 else $RM conftest* echo "$lt_simple_compile_test_code" > conftest.$ac_ext if { { eval echo "\"\$as_me\":${as_lineno-$LINENO}: \"$ac_compile\""; } >&5 (eval $ac_compile) 2>&5 ac_status=$? $as_echo "$as_me:${as_lineno-$LINENO}: \$? = $ac_status" >&5 test $ac_status = 0; } 2>conftest.err; then soname=conftest lib=conftest libobjs=conftest.$ac_objext deplibs= wl=$lt_prog_compiler_wl pic_flag=$lt_prog_compiler_pic compiler_flags=-v linker_flags=-v verstring= output_objdir=. libname=conftest lt_save_allow_undefined_flag=$allow_undefined_flag allow_undefined_flag= if { { eval echo "\"\$as_me\":${as_lineno-$LINENO}: \"$archive_cmds 2\>\&1 \| $GREP \" -lc \" \>/dev/null 2\>\&1\""; } >&5 (eval $archive_cmds 2\>\&1 \| $GREP \" -lc \" \>/dev/null 2\>\&1) 2>&5 ac_status=$? $as_echo "$as_me:${as_lineno-$LINENO}: \$? = $ac_status" >&5 test $ac_status = 0; } then lt_cv_archive_cmds_need_lc=no else lt_cv_archive_cmds_need_lc=yes fi allow_undefined_flag=$lt_save_allow_undefined_flag else cat conftest.err 1>&5 fi $RM conftest* fi { $as_echo "$as_me:${as_lineno-$LINENO}: result: $lt_cv_archive_cmds_need_lc" >&5 $as_echo "$lt_cv_archive_cmds_need_lc" >&6; } archive_cmds_need_lc=$lt_cv_archive_cmds_need_lc ;; esac fi ;; esac { $as_echo "$as_me:${as_lineno-$LINENO}: checking dynamic linker characteristics" >&5 $as_echo_n "checking dynamic linker characteristics... " >&6; } if test "$GCC" = yes; then case $host_os in darwin*) lt_awk_arg="/^libraries:/,/LR/" ;; *) lt_awk_arg="/^libraries:/" ;; esac case $host_os in mingw* | cegcc*) lt_sed_strip_eq="s,=\([A-Za-z]:\),\1,g" ;; *) lt_sed_strip_eq="s,=/,/,g" ;; esac lt_search_path_spec=`$CC -print-search-dirs | awk $lt_awk_arg | $SED -e "s/^libraries://" -e $lt_sed_strip_eq` case $lt_search_path_spec in *\;*) # if the path contains ";" then we assume it to be the separator # otherwise default to the standard path separator (i.e. ":") - it is # assumed that no part of a normal pathname contains ";" but that should # okay in the real world where ";" in dirpaths is itself problematic. lt_search_path_spec=`$ECHO "$lt_search_path_spec" | $SED 's/;/ /g'` ;; *) lt_search_path_spec=`$ECHO "$lt_search_path_spec" | $SED "s/$PATH_SEPARATOR/ /g"` ;; esac # Ok, now we have the path, separated by spaces, we can step through it # and add multilib dir if necessary. lt_tmp_lt_search_path_spec= lt_multi_os_dir=`$CC $CPPFLAGS $CFLAGS $LDFLAGS -print-multi-os-directory 2>/dev/null` for lt_sys_path in $lt_search_path_spec; do if test -d "$lt_sys_path/$lt_multi_os_dir"; then lt_tmp_lt_search_path_spec="$lt_tmp_lt_search_path_spec $lt_sys_path/$lt_multi_os_dir" else test -d "$lt_sys_path" && \ lt_tmp_lt_search_path_spec="$lt_tmp_lt_search_path_spec $lt_sys_path" fi done lt_search_path_spec=`$ECHO "$lt_tmp_lt_search_path_spec" | awk ' BEGIN {RS=" "; FS="/|\n";} { lt_foo=""; lt_count=0; for (lt_i = NF; lt_i > 0; lt_i--) { if ($lt_i != "" && $lt_i != ".") { if ($lt_i == "..") { lt_count++; } else { if (lt_count == 0) { lt_foo="/" $lt_i lt_foo; } else { lt_count--; } } } } if (lt_foo != "") { lt_freq[lt_foo]++; } if (lt_freq[lt_foo] == 1) { print lt_foo; } }'` # AWK program above erroneously prepends '/' to C:/dos/paths # for these hosts. case $host_os in mingw* | cegcc*) lt_search_path_spec=`$ECHO "$lt_search_path_spec" |\ $SED 's,/\([A-Za-z]:\),\1,g'` ;; esac sys_lib_search_path_spec=`$ECHO "$lt_search_path_spec" | $lt_NL2SP` else sys_lib_search_path_spec="/lib /usr/lib /usr/local/lib" fi library_names_spec= libname_spec='lib$name' soname_spec= shrext_cmds=".so" postinstall_cmds= postuninstall_cmds= finish_cmds= finish_eval= shlibpath_var= shlibpath_overrides_runpath=unknown version_type=none dynamic_linker="$host_os ld.so" sys_lib_dlsearch_path_spec="/lib /usr/lib" need_lib_prefix=unknown hardcode_into_libs=no # when you set need_version to no, make sure it does not cause -set_version # flags to be left without arguments need_version=unknown case $host_os in aix3*) version_type=linux # correct to gnu/linux during the next big refactor library_names_spec='${libname}${release}${shared_ext}$versuffix $libname.a' shlibpath_var=LIBPATH # AIX 3 has no versioning support, so we append a major version to the name. soname_spec='${libname}${release}${shared_ext}$major' ;; aix[4-9]*) version_type=linux # correct to gnu/linux during the next big refactor need_lib_prefix=no need_version=no hardcode_into_libs=yes if test "$host_cpu" = ia64; then # AIX 5 supports IA64 library_names_spec='${libname}${release}${shared_ext}$major ${libname}${release}${shared_ext}$versuffix $libname${shared_ext}' shlibpath_var=LD_LIBRARY_PATH else # With GCC up to 2.95.x, collect2 would create an import file # for dependence libraries. The import file would start with # the line `#! .'. This would cause the generated library to # depend on `.', always an invalid library. This was fixed in # development snapshots of GCC prior to 3.0. case $host_os in aix4 | aix4.[01] | aix4.[01].*) if { echo '#if __GNUC__ > 2 || (__GNUC__ == 2 && __GNUC_MINOR__ >= 97)' echo ' yes ' echo '#endif'; } | ${CC} -E - | $GREP yes > /dev/null; then : else can_build_shared=no fi ;; esac # AIX (on Power*) has no versioning support, so currently we can not hardcode correct # soname into executable. Probably we can add versioning support to # collect2, so additional links can be useful in future. if test "$aix_use_runtimelinking" = yes; then # If using run time linking (on AIX 4.2 or later) use lib.so # instead of lib.a to let people know that these are not # typical AIX shared libraries. library_names_spec='${libname}${release}${shared_ext}$versuffix ${libname}${release}${shared_ext}$major $libname${shared_ext}' else # We preserve .a as extension for shared libraries through AIX4.2 # and later when we are not doing run time linking. library_names_spec='${libname}${release}.a $libname.a' soname_spec='${libname}${release}${shared_ext}$major' fi shlibpath_var=LIBPATH fi ;; amigaos*) case $host_cpu in powerpc) # Since July 2007 AmigaOS4 officially supports .so libraries. # When compiling the executable, add -use-dynld -Lsobjs: to the compileline. library_names_spec='${libname}${release}${shared_ext}$versuffix ${libname}${release}${shared_ext}$major $libname${shared_ext}' ;; m68k) library_names_spec='$libname.ixlibrary $libname.a' # Create ${libname}_ixlibrary.a entries in /sys/libs. finish_eval='for lib in `ls $libdir/*.ixlibrary 2>/dev/null`; do libname=`func_echo_all "$lib" | $SED '\''s%^.*/\([^/]*\)\.ixlibrary$%\1%'\''`; test $RM /sys/libs/${libname}_ixlibrary.a; $show "cd /sys/libs && $LN_S $lib ${libname}_ixlibrary.a"; cd /sys/libs && $LN_S $lib ${libname}_ixlibrary.a || exit 1; done' ;; esac ;; beos*) library_names_spec='${libname}${shared_ext}' dynamic_linker="$host_os ld.so" shlibpath_var=LIBRARY_PATH ;; bsdi[45]*) version_type=linux # correct to gnu/linux during the next big refactor need_version=no library_names_spec='${libname}${release}${shared_ext}$versuffix ${libname}${release}${shared_ext}$major $libname${shared_ext}' soname_spec='${libname}${release}${shared_ext}$major' finish_cmds='PATH="\$PATH:/sbin" ldconfig $libdir' shlibpath_var=LD_LIBRARY_PATH sys_lib_search_path_spec="/shlib /usr/lib /usr/X11/lib /usr/contrib/lib /lib /usr/local/lib" sys_lib_dlsearch_path_spec="/shlib /usr/lib /usr/local/lib" # the default ld.so.conf also contains /usr/contrib/lib and # /usr/X11R6/lib (/usr/X11 is a link to /usr/X11R6), but let us allow # libtool to hard-code these into programs ;; cygwin* | mingw* | pw32* | cegcc*) version_type=windows shrext_cmds=".dll" need_version=no need_lib_prefix=no case $GCC,$cc_basename in yes,*) # gcc library_names_spec='$libname.dll.a' # DLL is installed to $(libdir)/../bin by postinstall_cmds postinstall_cmds='base_file=`basename \${file}`~ dlpath=`$SHELL 2>&1 -c '\''. $dir/'\''\${base_file}'\''i; echo \$dlname'\''`~ dldir=$destdir/`dirname \$dlpath`~ test -d \$dldir || mkdir -p \$dldir~ $install_prog $dir/$dlname \$dldir/$dlname~ chmod a+x \$dldir/$dlname~ if test -n '\''$stripme'\'' && test -n '\''$striplib'\''; then eval '\''$striplib \$dldir/$dlname'\'' || exit \$?; fi' postuninstall_cmds='dldll=`$SHELL 2>&1 -c '\''. $file; echo \$dlname'\''`~ dlpath=$dir/\$dldll~ $RM \$dlpath' shlibpath_overrides_runpath=yes case $host_os in cygwin*) # Cygwin DLLs use 'cyg' prefix rather than 'lib' soname_spec='`echo ${libname} | sed -e 's/^lib/cyg/'``echo ${release} | $SED -e 's/[.]/-/g'`${versuffix}${shared_ext}' sys_lib_search_path_spec="$sys_lib_search_path_spec /usr/lib/w32api" ;; mingw* | cegcc*) # MinGW DLLs use traditional 'lib' prefix soname_spec='${libname}`echo ${release} | $SED -e 's/[.]/-/g'`${versuffix}${shared_ext}' ;; pw32*) # pw32 DLLs use 'pw' prefix rather than 'lib' library_names_spec='`echo ${libname} | sed -e 's/^lib/pw/'``echo ${release} | $SED -e 's/[.]/-/g'`${versuffix}${shared_ext}' ;; esac dynamic_linker='Win32 ld.exe' ;; *,cl*) # Native MSVC libname_spec='$name' soname_spec='${libname}`echo ${release} | $SED -e 's/[.]/-/g'`${versuffix}${shared_ext}' library_names_spec='${libname}.dll.lib' case $build_os in mingw*) sys_lib_search_path_spec= lt_save_ifs=$IFS IFS=';' for lt_path in $LIB do IFS=$lt_save_ifs # Let DOS variable expansion print the short 8.3 style file name. lt_path=`cd "$lt_path" 2>/dev/null && cmd //C "for %i in (".") do @echo %~si"` sys_lib_search_path_spec="$sys_lib_search_path_spec $lt_path" done IFS=$lt_save_ifs # Convert to MSYS style. sys_lib_search_path_spec=`$ECHO "$sys_lib_search_path_spec" | sed -e 's|\\\\|/|g' -e 's| \\([a-zA-Z]\\):| /\\1|g' -e 's|^ ||'` ;; cygwin*) # Convert to unix form, then to dos form, then back to unix form # but this time dos style (no spaces!) so that the unix form looks # like /cygdrive/c/PROGRA~1:/cygdr... sys_lib_search_path_spec=`cygpath --path --unix "$LIB"` sys_lib_search_path_spec=`cygpath --path --dos "$sys_lib_search_path_spec" 2>/dev/null` sys_lib_search_path_spec=`cygpath --path --unix "$sys_lib_search_path_spec" | $SED -e "s/$PATH_SEPARATOR/ /g"` ;; *) sys_lib_search_path_spec="$LIB" if $ECHO "$sys_lib_search_path_spec" | $GREP ';[c-zC-Z]:/' >/dev/null; then # It is most probably a Windows format PATH. sys_lib_search_path_spec=`$ECHO "$sys_lib_search_path_spec" | $SED -e 's/;/ /g'` else sys_lib_search_path_spec=`$ECHO "$sys_lib_search_path_spec" | $SED -e "s/$PATH_SEPARATOR/ /g"` fi # FIXME: find the short name or the path components, as spaces are # common. (e.g. "Program Files" -> "PROGRA~1") ;; esac # DLL is installed to $(libdir)/../bin by postinstall_cmds postinstall_cmds='base_file=`basename \${file}`~ dlpath=`$SHELL 2>&1 -c '\''. $dir/'\''\${base_file}'\''i; echo \$dlname'\''`~ dldir=$destdir/`dirname \$dlpath`~ test -d \$dldir || mkdir -p \$dldir~ $install_prog $dir/$dlname \$dldir/$dlname' postuninstall_cmds='dldll=`$SHELL 2>&1 -c '\''. $file; echo \$dlname'\''`~ dlpath=$dir/\$dldll~ $RM \$dlpath' shlibpath_overrides_runpath=yes dynamic_linker='Win32 link.exe' ;; *) # Assume MSVC wrapper library_names_spec='${libname}`echo ${release} | $SED -e 's/[.]/-/g'`${versuffix}${shared_ext} $libname.lib' dynamic_linker='Win32 ld.exe' ;; esac # FIXME: first we should search . and the directory the executable is in shlibpath_var=PATH ;; darwin* | rhapsody*) dynamic_linker="$host_os dyld" version_type=darwin need_lib_prefix=no need_version=no library_names_spec='${libname}${release}${major}$shared_ext ${libname}$shared_ext' soname_spec='${libname}${release}${major}$shared_ext' shlibpath_overrides_runpath=yes shlibpath_var=DYLD_LIBRARY_PATH shrext_cmds='`test .$module = .yes && echo .so || echo .dylib`' sys_lib_search_path_spec="$sys_lib_search_path_spec /usr/local/lib" sys_lib_dlsearch_path_spec='/usr/local/lib /lib /usr/lib' ;; dgux*) version_type=linux # correct to gnu/linux during the next big refactor need_lib_prefix=no need_version=no library_names_spec='${libname}${release}${shared_ext}$versuffix ${libname}${release}${shared_ext}$major $libname$shared_ext' soname_spec='${libname}${release}${shared_ext}$major' shlibpath_var=LD_LIBRARY_PATH ;; freebsd* | dragonfly*) # DragonFly does not have aout. When/if they implement a new # versioning mechanism, adjust this. if test -x /usr/bin/objformat; then objformat=`/usr/bin/objformat` else case $host_os in freebsd[23].*) objformat=aout ;; *) objformat=elf ;; esac fi version_type=freebsd-$objformat case $version_type in freebsd-elf*) library_names_spec='${libname}${release}${shared_ext}$versuffix ${libname}${release}${shared_ext} $libname${shared_ext}' need_version=no need_lib_prefix=no ;; freebsd-*) library_names_spec='${libname}${release}${shared_ext}$versuffix $libname${shared_ext}$versuffix' need_version=yes ;; esac shlibpath_var=LD_LIBRARY_PATH case $host_os in freebsd2.*) shlibpath_overrides_runpath=yes ;; freebsd3.[01]* | freebsdelf3.[01]*) shlibpath_overrides_runpath=yes hardcode_into_libs=yes ;; freebsd3.[2-9]* | freebsdelf3.[2-9]* | \ freebsd4.[0-5] | freebsdelf4.[0-5] | freebsd4.1.1 | freebsdelf4.1.1) shlibpath_overrides_runpath=no hardcode_into_libs=yes ;; *) # from 4.6 on, and DragonFly shlibpath_overrides_runpath=yes hardcode_into_libs=yes ;; esac ;; haiku*) version_type=linux # correct to gnu/linux during the next big refactor need_lib_prefix=no need_version=no dynamic_linker="$host_os runtime_loader" library_names_spec='${libname}${release}${shared_ext}$versuffix ${libname}${release}${shared_ext}${major} ${libname}${shared_ext}' soname_spec='${libname}${release}${shared_ext}$major' shlibpath_var=LIBRARY_PATH shlibpath_overrides_runpath=yes sys_lib_dlsearch_path_spec='/boot/home/config/lib /boot/common/lib /boot/system/lib' hardcode_into_libs=yes ;; hpux9* | hpux10* | hpux11*) # Give a soname corresponding to the major version so that dld.sl refuses to # link against other versions. version_type=sunos need_lib_prefix=no need_version=no case $host_cpu in ia64*) shrext_cmds='.so' hardcode_into_libs=yes dynamic_linker="$host_os dld.so" shlibpath_var=LD_LIBRARY_PATH shlibpath_overrides_runpath=yes # Unless +noenvvar is specified. library_names_spec='${libname}${release}${shared_ext}$versuffix ${libname}${release}${shared_ext}$major $libname${shared_ext}' soname_spec='${libname}${release}${shared_ext}$major' if test "X$HPUX_IA64_MODE" = X32; then sys_lib_search_path_spec="/usr/lib/hpux32 /usr/local/lib/hpux32 /usr/local/lib" else sys_lib_search_path_spec="/usr/lib/hpux64 /usr/local/lib/hpux64" fi sys_lib_dlsearch_path_spec=$sys_lib_search_path_spec ;; hppa*64*) shrext_cmds='.sl' hardcode_into_libs=yes dynamic_linker="$host_os dld.sl" shlibpath_var=LD_LIBRARY_PATH # How should we handle SHLIB_PATH shlibpath_overrides_runpath=yes # Unless +noenvvar is specified. library_names_spec='${libname}${release}${shared_ext}$versuffix ${libname}${release}${shared_ext}$major $libname${shared_ext}' soname_spec='${libname}${release}${shared_ext}$major' sys_lib_search_path_spec="/usr/lib/pa20_64 /usr/ccs/lib/pa20_64" sys_lib_dlsearch_path_spec=$sys_lib_search_path_spec ;; *) shrext_cmds='.sl' dynamic_linker="$host_os dld.sl" shlibpath_var=SHLIB_PATH shlibpath_overrides_runpath=no # +s is required to enable SHLIB_PATH library_names_spec='${libname}${release}${shared_ext}$versuffix ${libname}${release}${shared_ext}$major $libname${shared_ext}' soname_spec='${libname}${release}${shared_ext}$major' ;; esac # HP-UX runs *really* slowly unless shared libraries are mode 555, ... postinstall_cmds='chmod 555 $lib' # or fails outright, so override atomically: install_override_mode=555 ;; interix[3-9]*) version_type=linux # correct to gnu/linux during the next big refactor need_lib_prefix=no need_version=no library_names_spec='${libname}${release}${shared_ext}$versuffix ${libname}${release}${shared_ext}$major ${libname}${shared_ext}' soname_spec='${libname}${release}${shared_ext}$major' dynamic_linker='Interix 3.x ld.so.1 (PE, like ELF)' shlibpath_var=LD_LIBRARY_PATH shlibpath_overrides_runpath=no hardcode_into_libs=yes ;; irix5* | irix6* | nonstopux*) case $host_os in nonstopux*) version_type=nonstopux ;; *) if test "$lt_cv_prog_gnu_ld" = yes; then version_type=linux # correct to gnu/linux during the next big refactor else version_type=irix fi ;; esac need_lib_prefix=no need_version=no soname_spec='${libname}${release}${shared_ext}$major' library_names_spec='${libname}${release}${shared_ext}$versuffix ${libname}${release}${shared_ext}$major ${libname}${release}${shared_ext} $libname${shared_ext}' case $host_os in irix5* | nonstopux*) libsuff= shlibsuff= ;; *) case $LD in # libtool.m4 will add one of these switches to LD *-32|*"-32 "|*-melf32bsmip|*"-melf32bsmip ") libsuff= shlibsuff= libmagic=32-bit;; *-n32|*"-n32 "|*-melf32bmipn32|*"-melf32bmipn32 ") libsuff=32 shlibsuff=N32 libmagic=N32;; *-64|*"-64 "|*-melf64bmip|*"-melf64bmip ") libsuff=64 shlibsuff=64 libmagic=64-bit;; *) libsuff= shlibsuff= libmagic=never-match;; esac ;; esac shlibpath_var=LD_LIBRARY${shlibsuff}_PATH shlibpath_overrides_runpath=no sys_lib_search_path_spec="/usr/lib${libsuff} /lib${libsuff} /usr/local/lib${libsuff}" sys_lib_dlsearch_path_spec="/usr/lib${libsuff} /lib${libsuff}" hardcode_into_libs=yes ;; # No shared lib support for Linux oldld, aout, or coff. linux*oldld* | linux*aout* | linux*coff*) dynamic_linker=no ;; # This must be glibc/ELF. linux* | k*bsd*-gnu | kopensolaris*-gnu | gnu*) version_type=linux # correct to gnu/linux during the next big refactor need_lib_prefix=no need_version=no library_names_spec='${libname}${release}${shared_ext}$versuffix ${libname}${release}${shared_ext}$major $libname${shared_ext}' soname_spec='${libname}${release}${shared_ext}$major' finish_cmds='PATH="\$PATH:/sbin" ldconfig -n $libdir' shlibpath_var=LD_LIBRARY_PATH shlibpath_overrides_runpath=no # Some binutils ld are patched to set DT_RUNPATH if ${lt_cv_shlibpath_overrides_runpath+:} false; then : $as_echo_n "(cached) " >&6 else lt_cv_shlibpath_overrides_runpath=no save_LDFLAGS=$LDFLAGS save_libdir=$libdir eval "libdir=/foo; wl=\"$lt_prog_compiler_wl\"; \ LDFLAGS=\"\$LDFLAGS $hardcode_libdir_flag_spec\"" cat confdefs.h - <<_ACEOF >conftest.$ac_ext /* end confdefs.h. */ int main () { ; return 0; } _ACEOF if ac_fn_c_try_link "$LINENO"; then : if ($OBJDUMP -p conftest$ac_exeext) 2>/dev/null | grep "RUNPATH.*$libdir" >/dev/null; then : lt_cv_shlibpath_overrides_runpath=yes fi fi rm -f core conftest.err conftest.$ac_objext \ conftest$ac_exeext conftest.$ac_ext LDFLAGS=$save_LDFLAGS libdir=$save_libdir fi shlibpath_overrides_runpath=$lt_cv_shlibpath_overrides_runpath # This implies no fast_install, which is unacceptable. # Some rework will be needed to allow for fast_install # before this can be enabled. hardcode_into_libs=yes # Append ld.so.conf contents to the search path if test -f /etc/ld.so.conf; then lt_ld_extra=`awk '/^include / { system(sprintf("cd /etc; cat %s 2>/dev/null", \$2)); skip = 1; } { if (!skip) print \$0; skip = 0; }' < /etc/ld.so.conf | $SED -e 's/#.*//;/^[ ]*hwcap[ ]/d;s/[:, ]/ /g;s/=[^=]*$//;s/=[^= ]* / /g;s/"//g;/^$/d' | tr '\n' ' '` sys_lib_dlsearch_path_spec="/lib /usr/lib $lt_ld_extra" fi # We used to test for /lib/ld.so.1 and disable shared libraries on # powerpc, because MkLinux only supported shared libraries with the # GNU dynamic linker. Since this was broken with cross compilers, # most powerpc-linux boxes support dynamic linking these days and # people can always --disable-shared, the test was removed, and we # assume the GNU/Linux dynamic linker is in use. dynamic_linker='GNU/Linux ld.so' ;; netbsdelf*-gnu) version_type=linux need_lib_prefix=no need_version=no library_names_spec='${libname}${release}${shared_ext}$versuffix ${libname}${release}${shared_ext}$major ${libname}${shared_ext}' soname_spec='${libname}${release}${shared_ext}$major' shlibpath_var=LD_LIBRARY_PATH shlibpath_overrides_runpath=no hardcode_into_libs=yes dynamic_linker='NetBSD ld.elf_so' ;; netbsd*) version_type=sunos need_lib_prefix=no need_version=no if echo __ELF__ | $CC -E - | $GREP __ELF__ >/dev/null; then library_names_spec='${libname}${release}${shared_ext}$versuffix ${libname}${shared_ext}$versuffix' finish_cmds='PATH="\$PATH:/sbin" ldconfig -m $libdir' dynamic_linker='NetBSD (a.out) ld.so' else library_names_spec='${libname}${release}${shared_ext}$versuffix ${libname}${release}${shared_ext}$major ${libname}${shared_ext}' soname_spec='${libname}${release}${shared_ext}$major' dynamic_linker='NetBSD ld.elf_so' fi shlibpath_var=LD_LIBRARY_PATH shlibpath_overrides_runpath=yes hardcode_into_libs=yes ;; newsos6) version_type=linux # correct to gnu/linux during the next big refactor library_names_spec='${libname}${release}${shared_ext}$versuffix ${libname}${release}${shared_ext}$major $libname${shared_ext}' shlibpath_var=LD_LIBRARY_PATH shlibpath_overrides_runpath=yes ;; *nto* | *qnx*) version_type=qnx need_lib_prefix=no need_version=no library_names_spec='${libname}${release}${shared_ext}$versuffix ${libname}${release}${shared_ext}$major $libname${shared_ext}' soname_spec='${libname}${release}${shared_ext}$major' shlibpath_var=LD_LIBRARY_PATH shlibpath_overrides_runpath=no hardcode_into_libs=yes dynamic_linker='ldqnx.so' ;; openbsd*) version_type=sunos sys_lib_dlsearch_path_spec="/usr/lib" need_lib_prefix=no # Some older versions of OpenBSD (3.3 at least) *do* need versioned libs. case $host_os in openbsd3.3 | openbsd3.3.*) need_version=yes ;; *) need_version=no ;; esac library_names_spec='${libname}${release}${shared_ext}$versuffix ${libname}${shared_ext}$versuffix' finish_cmds='PATH="\$PATH:/sbin" ldconfig -m $libdir' shlibpath_var=LD_LIBRARY_PATH if test -z "`echo __ELF__ | $CC -E - | $GREP __ELF__`" || test "$host_os-$host_cpu" = "openbsd2.8-powerpc"; then case $host_os in openbsd2.[89] | openbsd2.[89].*) shlibpath_overrides_runpath=no ;; *) shlibpath_overrides_runpath=yes ;; esac else shlibpath_overrides_runpath=yes fi ;; os2*) libname_spec='$name' shrext_cmds=".dll" need_lib_prefix=no library_names_spec='$libname${shared_ext} $libname.a' dynamic_linker='OS/2 ld.exe' shlibpath_var=LIBPATH ;; osf3* | osf4* | osf5*) version_type=osf need_lib_prefix=no need_version=no soname_spec='${libname}${release}${shared_ext}$major' library_names_spec='${libname}${release}${shared_ext}$versuffix ${libname}${release}${shared_ext}$major $libname${shared_ext}' shlibpath_var=LD_LIBRARY_PATH sys_lib_search_path_spec="/usr/shlib /usr/ccs/lib /usr/lib/cmplrs/cc /usr/lib /usr/local/lib /var/shlib" sys_lib_dlsearch_path_spec="$sys_lib_search_path_spec" ;; rdos*) dynamic_linker=no ;; solaris*) version_type=linux # correct to gnu/linux during the next big refactor need_lib_prefix=no need_version=no library_names_spec='${libname}${release}${shared_ext}$versuffix ${libname}${release}${shared_ext}$major $libname${shared_ext}' soname_spec='${libname}${release}${shared_ext}$major' shlibpath_var=LD_LIBRARY_PATH shlibpath_overrides_runpath=yes hardcode_into_libs=yes # ldd complains unless libraries are executable postinstall_cmds='chmod +x $lib' ;; sunos4*) version_type=sunos library_names_spec='${libname}${release}${shared_ext}$versuffix ${libname}${shared_ext}$versuffix' finish_cmds='PATH="\$PATH:/usr/etc" ldconfig $libdir' shlibpath_var=LD_LIBRARY_PATH shlibpath_overrides_runpath=yes if test "$with_gnu_ld" = yes; then need_lib_prefix=no fi need_version=yes ;; sysv4 | sysv4.3*) version_type=linux # correct to gnu/linux during the next big refactor library_names_spec='${libname}${release}${shared_ext}$versuffix ${libname}${release}${shared_ext}$major $libname${shared_ext}' soname_spec='${libname}${release}${shared_ext}$major' shlibpath_var=LD_LIBRARY_PATH case $host_vendor in sni) shlibpath_overrides_runpath=no need_lib_prefix=no runpath_var=LD_RUN_PATH ;; siemens) need_lib_prefix=no ;; motorola) need_lib_prefix=no need_version=no shlibpath_overrides_runpath=no sys_lib_search_path_spec='/lib /usr/lib /usr/ccs/lib' ;; esac ;; sysv4*MP*) if test -d /usr/nec ;then version_type=linux # correct to gnu/linux during the next big refactor library_names_spec='$libname${shared_ext}.$versuffix $libname${shared_ext}.$major $libname${shared_ext}' soname_spec='$libname${shared_ext}.$major' shlibpath_var=LD_LIBRARY_PATH fi ;; sysv5* | sco3.2v5* | sco5v6* | unixware* | OpenUNIX* | sysv4*uw2*) version_type=freebsd-elf need_lib_prefix=no need_version=no library_names_spec='${libname}${release}${shared_ext}$versuffix ${libname}${release}${shared_ext} $libname${shared_ext}' soname_spec='${libname}${release}${shared_ext}$major' shlibpath_var=LD_LIBRARY_PATH shlibpath_overrides_runpath=yes hardcode_into_libs=yes if test "$with_gnu_ld" = yes; then sys_lib_search_path_spec='/usr/local/lib /usr/gnu/lib /usr/ccs/lib /usr/lib /lib' else sys_lib_search_path_spec='/usr/ccs/lib /usr/lib' case $host_os in sco3.2v5*) sys_lib_search_path_spec="$sys_lib_search_path_spec /lib" ;; esac fi sys_lib_dlsearch_path_spec='/usr/lib' ;; tpf*) # TPF is a cross-target only. Preferred cross-host = GNU/Linux. version_type=linux # correct to gnu/linux during the next big refactor need_lib_prefix=no need_version=no library_names_spec='${libname}${release}${shared_ext}$versuffix ${libname}${release}${shared_ext}$major $libname${shared_ext}' shlibpath_var=LD_LIBRARY_PATH shlibpath_overrides_runpath=no hardcode_into_libs=yes ;; uts4*) version_type=linux # correct to gnu/linux during the next big refactor library_names_spec='${libname}${release}${shared_ext}$versuffix ${libname}${release}${shared_ext}$major $libname${shared_ext}' soname_spec='${libname}${release}${shared_ext}$major' shlibpath_var=LD_LIBRARY_PATH ;; *) dynamic_linker=no ;; esac { $as_echo "$as_me:${as_lineno-$LINENO}: result: $dynamic_linker" >&5 $as_echo "$dynamic_linker" >&6; } test "$dynamic_linker" = no && can_build_shared=no variables_saved_for_relink="PATH $shlibpath_var $runpath_var" if test "$GCC" = yes; then variables_saved_for_relink="$variables_saved_for_relink GCC_EXEC_PREFIX COMPILER_PATH LIBRARY_PATH" fi if test "${lt_cv_sys_lib_search_path_spec+set}" = set; then sys_lib_search_path_spec="$lt_cv_sys_lib_search_path_spec" fi if test "${lt_cv_sys_lib_dlsearch_path_spec+set}" = set; then sys_lib_dlsearch_path_spec="$lt_cv_sys_lib_dlsearch_path_spec" fi { $as_echo "$as_me:${as_lineno-$LINENO}: checking how to hardcode library paths into programs" >&5 $as_echo_n "checking how to hardcode library paths into programs... " >&6; } hardcode_action= if test -n "$hardcode_libdir_flag_spec" || test -n "$runpath_var" || test "X$hardcode_automatic" = "Xyes" ; then # We can hardcode non-existent directories. if test "$hardcode_direct" != no && # If the only mechanism to avoid hardcoding is shlibpath_var, we # have to relink, otherwise we might link with an installed library # when we should be linking with a yet-to-be-installed one ## test "$_LT_TAGVAR(hardcode_shlibpath_var, )" != no && test "$hardcode_minus_L" != no; then # Linking always hardcodes the temporary library directory. hardcode_action=relink else # We can link without hardcoding, and we can hardcode nonexisting dirs. hardcode_action=immediate fi else # We cannot hardcode anything, or else we can only hardcode existing # directories. hardcode_action=unsupported fi { $as_echo "$as_me:${as_lineno-$LINENO}: result: $hardcode_action" >&5 $as_echo "$hardcode_action" >&6; } if test "$hardcode_action" = relink || test "$inherit_rpath" = yes; then # Fast installation is not supported enable_fast_install=no elif test "$shlibpath_overrides_runpath" = yes || test "$enable_shared" = no; then # Fast installation is not necessary enable_fast_install=needless fi if test "x$enable_dlopen" != xyes; then enable_dlopen=unknown enable_dlopen_self=unknown enable_dlopen_self_static=unknown else lt_cv_dlopen=no lt_cv_dlopen_libs= case $host_os in beos*) lt_cv_dlopen="load_add_on" lt_cv_dlopen_libs= lt_cv_dlopen_self=yes ;; mingw* | pw32* | cegcc*) lt_cv_dlopen="LoadLibrary" lt_cv_dlopen_libs= ;; cygwin*) lt_cv_dlopen="dlopen" lt_cv_dlopen_libs= ;; darwin*) # if libdl is installed we need to link against it { $as_echo "$as_me:${as_lineno-$LINENO}: checking for dlopen in -ldl" >&5 $as_echo_n "checking for dlopen in -ldl... " >&6; } if ${ac_cv_lib_dl_dlopen+:} false; then : $as_echo_n "(cached) " >&6 else ac_check_lib_save_LIBS=$LIBS LIBS="-ldl $LIBS" cat confdefs.h - <<_ACEOF >conftest.$ac_ext /* end confdefs.h. */ /* Override any GCC internal prototype to avoid an error. Use char because int might match the return type of a GCC builtin and then its argument prototype would still apply. */ #ifdef __cplusplus extern "C" #endif char dlopen (); int main () { return dlopen (); ; return 0; } _ACEOF if ac_fn_c_try_link "$LINENO"; then : ac_cv_lib_dl_dlopen=yes else ac_cv_lib_dl_dlopen=no fi rm -f core conftest.err conftest.$ac_objext \ conftest$ac_exeext conftest.$ac_ext LIBS=$ac_check_lib_save_LIBS fi { $as_echo "$as_me:${as_lineno-$LINENO}: result: $ac_cv_lib_dl_dlopen" >&5 $as_echo "$ac_cv_lib_dl_dlopen" >&6; } if test "x$ac_cv_lib_dl_dlopen" = xyes; then : lt_cv_dlopen="dlopen" lt_cv_dlopen_libs="-ldl" else lt_cv_dlopen="dyld" lt_cv_dlopen_libs= lt_cv_dlopen_self=yes fi ;; *) ac_fn_c_check_func "$LINENO" "shl_load" "ac_cv_func_shl_load" if test "x$ac_cv_func_shl_load" = xyes; then : lt_cv_dlopen="shl_load" else { $as_echo "$as_me:${as_lineno-$LINENO}: checking for shl_load in -ldld" >&5 $as_echo_n "checking for shl_load in -ldld... " >&6; } if ${ac_cv_lib_dld_shl_load+:} false; then : $as_echo_n "(cached) " >&6 else ac_check_lib_save_LIBS=$LIBS LIBS="-ldld $LIBS" cat confdefs.h - <<_ACEOF >conftest.$ac_ext /* end confdefs.h. */ /* Override any GCC internal prototype to avoid an error. Use char because int might match the return type of a GCC builtin and then its argument prototype would still apply. */ #ifdef __cplusplus extern "C" #endif char shl_load (); int main () { return shl_load (); ; return 0; } _ACEOF if ac_fn_c_try_link "$LINENO"; then : ac_cv_lib_dld_shl_load=yes else ac_cv_lib_dld_shl_load=no fi rm -f core conftest.err conftest.$ac_objext \ conftest$ac_exeext conftest.$ac_ext LIBS=$ac_check_lib_save_LIBS fi { $as_echo "$as_me:${as_lineno-$LINENO}: result: $ac_cv_lib_dld_shl_load" >&5 $as_echo "$ac_cv_lib_dld_shl_load" >&6; } if test "x$ac_cv_lib_dld_shl_load" = xyes; then : lt_cv_dlopen="shl_load" lt_cv_dlopen_libs="-ldld" else ac_fn_c_check_func "$LINENO" "dlopen" "ac_cv_func_dlopen" if test "x$ac_cv_func_dlopen" = xyes; then : lt_cv_dlopen="dlopen" else { $as_echo "$as_me:${as_lineno-$LINENO}: checking for dlopen in -ldl" >&5 $as_echo_n "checking for dlopen in -ldl... " >&6; } if ${ac_cv_lib_dl_dlopen+:} false; then : $as_echo_n "(cached) " >&6 else ac_check_lib_save_LIBS=$LIBS LIBS="-ldl $LIBS" cat confdefs.h - <<_ACEOF >conftest.$ac_ext /* end confdefs.h. */ /* Override any GCC internal prototype to avoid an error. Use char because int might match the return type of a GCC builtin and then its argument prototype would still apply. */ #ifdef __cplusplus extern "C" #endif char dlopen (); int main () { return dlopen (); ; return 0; } _ACEOF if ac_fn_c_try_link "$LINENO"; then : ac_cv_lib_dl_dlopen=yes else ac_cv_lib_dl_dlopen=no fi rm -f core conftest.err conftest.$ac_objext \ conftest$ac_exeext conftest.$ac_ext LIBS=$ac_check_lib_save_LIBS fi { $as_echo "$as_me:${as_lineno-$LINENO}: result: $ac_cv_lib_dl_dlopen" >&5 $as_echo "$ac_cv_lib_dl_dlopen" >&6; } if test "x$ac_cv_lib_dl_dlopen" = xyes; then : lt_cv_dlopen="dlopen" lt_cv_dlopen_libs="-ldl" else { $as_echo "$as_me:${as_lineno-$LINENO}: checking for dlopen in -lsvld" >&5 $as_echo_n "checking for dlopen in -lsvld... " >&6; } if ${ac_cv_lib_svld_dlopen+:} false; then : $as_echo_n "(cached) " >&6 else ac_check_lib_save_LIBS=$LIBS LIBS="-lsvld $LIBS" cat confdefs.h - <<_ACEOF >conftest.$ac_ext /* end confdefs.h. */ /* Override any GCC internal prototype to avoid an error. Use char because int might match the return type of a GCC builtin and then its argument prototype would still apply. */ #ifdef __cplusplus extern "C" #endif char dlopen (); int main () { return dlopen (); ; return 0; } _ACEOF if ac_fn_c_try_link "$LINENO"; then : ac_cv_lib_svld_dlopen=yes else ac_cv_lib_svld_dlopen=no fi rm -f core conftest.err conftest.$ac_objext \ conftest$ac_exeext conftest.$ac_ext LIBS=$ac_check_lib_save_LIBS fi { $as_echo "$as_me:${as_lineno-$LINENO}: result: $ac_cv_lib_svld_dlopen" >&5 $as_echo "$ac_cv_lib_svld_dlopen" >&6; } if test "x$ac_cv_lib_svld_dlopen" = xyes; then : lt_cv_dlopen="dlopen" lt_cv_dlopen_libs="-lsvld" else { $as_echo "$as_me:${as_lineno-$LINENO}: checking for dld_link in -ldld" >&5 $as_echo_n "checking for dld_link in -ldld... " >&6; } if ${ac_cv_lib_dld_dld_link+:} false; then : $as_echo_n "(cached) " >&6 else ac_check_lib_save_LIBS=$LIBS LIBS="-ldld $LIBS" cat confdefs.h - <<_ACEOF >conftest.$ac_ext /* end confdefs.h. */ /* Override any GCC internal prototype to avoid an error. Use char because int might match the return type of a GCC builtin and then its argument prototype would still apply. */ #ifdef __cplusplus extern "C" #endif char dld_link (); int main () { return dld_link (); ; return 0; } _ACEOF if ac_fn_c_try_link "$LINENO"; then : ac_cv_lib_dld_dld_link=yes else ac_cv_lib_dld_dld_link=no fi rm -f core conftest.err conftest.$ac_objext \ conftest$ac_exeext conftest.$ac_ext LIBS=$ac_check_lib_save_LIBS fi { $as_echo "$as_me:${as_lineno-$LINENO}: result: $ac_cv_lib_dld_dld_link" >&5 $as_echo "$ac_cv_lib_dld_dld_link" >&6; } if test "x$ac_cv_lib_dld_dld_link" = xyes; then : lt_cv_dlopen="dld_link" lt_cv_dlopen_libs="-ldld" fi fi fi fi fi fi ;; esac if test "x$lt_cv_dlopen" != xno; then enable_dlopen=yes else enable_dlopen=no fi case $lt_cv_dlopen in dlopen) save_CPPFLAGS="$CPPFLAGS" test "x$ac_cv_header_dlfcn_h" = xyes && CPPFLAGS="$CPPFLAGS -DHAVE_DLFCN_H" save_LDFLAGS="$LDFLAGS" wl=$lt_prog_compiler_wl eval LDFLAGS=\"\$LDFLAGS $export_dynamic_flag_spec\" save_LIBS="$LIBS" LIBS="$lt_cv_dlopen_libs $LIBS" { $as_echo "$as_me:${as_lineno-$LINENO}: checking whether a program can dlopen itself" >&5 $as_echo_n "checking whether a program can dlopen itself... " >&6; } if ${lt_cv_dlopen_self+:} false; then : $as_echo_n "(cached) " >&6 else if test "$cross_compiling" = yes; then : lt_cv_dlopen_self=cross else lt_dlunknown=0; lt_dlno_uscore=1; lt_dlneed_uscore=2 lt_status=$lt_dlunknown cat > conftest.$ac_ext <<_LT_EOF #line $LINENO "configure" #include "confdefs.h" #if HAVE_DLFCN_H #include #endif #include #ifdef RTLD_GLOBAL # define LT_DLGLOBAL RTLD_GLOBAL #else # ifdef DL_GLOBAL # define LT_DLGLOBAL DL_GLOBAL # else # define LT_DLGLOBAL 0 # endif #endif /* We may have to define LT_DLLAZY_OR_NOW in the command line if we find out it does not work in some platform. */ #ifndef LT_DLLAZY_OR_NOW # ifdef RTLD_LAZY # define LT_DLLAZY_OR_NOW RTLD_LAZY # else # ifdef DL_LAZY # define LT_DLLAZY_OR_NOW DL_LAZY # else # ifdef RTLD_NOW # define LT_DLLAZY_OR_NOW RTLD_NOW # else # ifdef DL_NOW # define LT_DLLAZY_OR_NOW DL_NOW # else # define LT_DLLAZY_OR_NOW 0 # endif # endif # endif # endif #endif /* When -fvisbility=hidden is used, assume the code has been annotated correspondingly for the symbols needed. */ #if defined(__GNUC__) && (((__GNUC__ == 3) && (__GNUC_MINOR__ >= 3)) || (__GNUC__ > 3)) int fnord () __attribute__((visibility("default"))); #endif int fnord () { return 42; } int main () { void *self = dlopen (0, LT_DLGLOBAL|LT_DLLAZY_OR_NOW); int status = $lt_dlunknown; if (self) { if (dlsym (self,"fnord")) status = $lt_dlno_uscore; else { if (dlsym( self,"_fnord")) status = $lt_dlneed_uscore; else puts (dlerror ()); } /* dlclose (self); */ } else puts (dlerror ()); return status; } _LT_EOF if { { eval echo "\"\$as_me\":${as_lineno-$LINENO}: \"$ac_link\""; } >&5 (eval $ac_link) 2>&5 ac_status=$? $as_echo "$as_me:${as_lineno-$LINENO}: \$? = $ac_status" >&5 test $ac_status = 0; } && test -s conftest${ac_exeext} 2>/dev/null; then (./conftest; exit; ) >&5 2>/dev/null lt_status=$? case x$lt_status in x$lt_dlno_uscore) lt_cv_dlopen_self=yes ;; x$lt_dlneed_uscore) lt_cv_dlopen_self=yes ;; x$lt_dlunknown|x*) lt_cv_dlopen_self=no ;; esac else : # compilation failed lt_cv_dlopen_self=no fi fi rm -fr conftest* fi { $as_echo "$as_me:${as_lineno-$LINENO}: result: $lt_cv_dlopen_self" >&5 $as_echo "$lt_cv_dlopen_self" >&6; } if test "x$lt_cv_dlopen_self" = xyes; then wl=$lt_prog_compiler_wl eval LDFLAGS=\"\$LDFLAGS $lt_prog_compiler_static\" { $as_echo "$as_me:${as_lineno-$LINENO}: checking whether a statically linked program can dlopen itself" >&5 $as_echo_n "checking whether a statically linked program can dlopen itself... " >&6; } if ${lt_cv_dlopen_self_static+:} false; then : $as_echo_n "(cached) " >&6 else if test "$cross_compiling" = yes; then : lt_cv_dlopen_self_static=cross else lt_dlunknown=0; lt_dlno_uscore=1; lt_dlneed_uscore=2 lt_status=$lt_dlunknown cat > conftest.$ac_ext <<_LT_EOF #line $LINENO "configure" #include "confdefs.h" #if HAVE_DLFCN_H #include #endif #include #ifdef RTLD_GLOBAL # define LT_DLGLOBAL RTLD_GLOBAL #else # ifdef DL_GLOBAL # define LT_DLGLOBAL DL_GLOBAL # else # define LT_DLGLOBAL 0 # endif #endif /* We may have to define LT_DLLAZY_OR_NOW in the command line if we find out it does not work in some platform. */ #ifndef LT_DLLAZY_OR_NOW # ifdef RTLD_LAZY # define LT_DLLAZY_OR_NOW RTLD_LAZY # else # ifdef DL_LAZY # define LT_DLLAZY_OR_NOW DL_LAZY # else # ifdef RTLD_NOW # define LT_DLLAZY_OR_NOW RTLD_NOW # else # ifdef DL_NOW # define LT_DLLAZY_OR_NOW DL_NOW # else # define LT_DLLAZY_OR_NOW 0 # endif # endif # endif # endif #endif /* When -fvisbility=hidden is used, assume the code has been annotated correspondingly for the symbols needed. */ #if defined(__GNUC__) && (((__GNUC__ == 3) && (__GNUC_MINOR__ >= 3)) || (__GNUC__ > 3)) int fnord () __attribute__((visibility("default"))); #endif int fnord () { return 42; } int main () { void *self = dlopen (0, LT_DLGLOBAL|LT_DLLAZY_OR_NOW); int status = $lt_dlunknown; if (self) { if (dlsym (self,"fnord")) status = $lt_dlno_uscore; else { if (dlsym( self,"_fnord")) status = $lt_dlneed_uscore; else puts (dlerror ()); } /* dlclose (self); */ } else puts (dlerror ()); return status; } _LT_EOF if { { eval echo "\"\$as_me\":${as_lineno-$LINENO}: \"$ac_link\""; } >&5 (eval $ac_link) 2>&5 ac_status=$? $as_echo "$as_me:${as_lineno-$LINENO}: \$? = $ac_status" >&5 test $ac_status = 0; } && test -s conftest${ac_exeext} 2>/dev/null; then (./conftest; exit; ) >&5 2>/dev/null lt_status=$? case x$lt_status in x$lt_dlno_uscore) lt_cv_dlopen_self_static=yes ;; x$lt_dlneed_uscore) lt_cv_dlopen_self_static=yes ;; x$lt_dlunknown|x*) lt_cv_dlopen_self_static=no ;; esac else : # compilation failed lt_cv_dlopen_self_static=no fi fi rm -fr conftest* fi { $as_echo "$as_me:${as_lineno-$LINENO}: result: $lt_cv_dlopen_self_static" >&5 $as_echo "$lt_cv_dlopen_self_static" >&6; } fi CPPFLAGS="$save_CPPFLAGS" LDFLAGS="$save_LDFLAGS" LIBS="$save_LIBS" ;; esac case $lt_cv_dlopen_self in yes|no) enable_dlopen_self=$lt_cv_dlopen_self ;; *) enable_dlopen_self=unknown ;; esac case $lt_cv_dlopen_self_static in yes|no) enable_dlopen_self_static=$lt_cv_dlopen_self_static ;; *) enable_dlopen_self_static=unknown ;; esac fi striplib= old_striplib= { $as_echo "$as_me:${as_lineno-$LINENO}: checking whether stripping libraries is possible" >&5 $as_echo_n "checking whether stripping libraries is possible... " >&6; } if test -n "$STRIP" && $STRIP -V 2>&1 | $GREP "GNU strip" >/dev/null; then test -z "$old_striplib" && old_striplib="$STRIP --strip-debug" test -z "$striplib" && striplib="$STRIP --strip-unneeded" { $as_echo "$as_me:${as_lineno-$LINENO}: result: yes" >&5 $as_echo "yes" >&6; } else # FIXME - insert some real tests, host_os isn't really good enough case $host_os in darwin*) if test -n "$STRIP" ; then striplib="$STRIP -x" old_striplib="$STRIP -S" { $as_echo "$as_me:${as_lineno-$LINENO}: result: yes" >&5 $as_echo "yes" >&6; } else { $as_echo "$as_me:${as_lineno-$LINENO}: result: no" >&5 $as_echo "no" >&6; } fi ;; *) { $as_echo "$as_me:${as_lineno-$LINENO}: result: no" >&5 $as_echo "no" >&6; } ;; esac fi # Report which library types will actually be built { $as_echo "$as_me:${as_lineno-$LINENO}: checking if libtool supports shared libraries" >&5 $as_echo_n "checking if libtool supports shared libraries... " >&6; } { $as_echo "$as_me:${as_lineno-$LINENO}: result: $can_build_shared" >&5 $as_echo "$can_build_shared" >&6; } { $as_echo "$as_me:${as_lineno-$LINENO}: checking whether to build shared libraries" >&5 $as_echo_n "checking whether to build shared libraries... " >&6; } test "$can_build_shared" = "no" && enable_shared=no # On AIX, shared libraries and static libraries use the same namespace, and # are all built from PIC. case $host_os in aix3*) test "$enable_shared" = yes && enable_static=no if test -n "$RANLIB"; then archive_cmds="$archive_cmds~\$RANLIB \$lib" postinstall_cmds='$RANLIB $lib' fi ;; aix[4-9]*) if test "$host_cpu" != ia64 && test "$aix_use_runtimelinking" = no ; then test "$enable_shared" = yes && enable_static=no fi ;; esac { $as_echo "$as_me:${as_lineno-$LINENO}: result: $enable_shared" >&5 $as_echo "$enable_shared" >&6; } { $as_echo "$as_me:${as_lineno-$LINENO}: checking whether to build static libraries" >&5 $as_echo_n "checking whether to build static libraries... " >&6; } # Make sure either enable_shared or enable_static is yes. test "$enable_shared" = yes || enable_static=yes { $as_echo "$as_me:${as_lineno-$LINENO}: result: $enable_static" >&5 $as_echo "$enable_static" >&6; } fi ac_ext=c ac_cpp='$CPP $CPPFLAGS' ac_compile='$CC -c $CFLAGS $CPPFLAGS conftest.$ac_ext >&5' ac_link='$CC -o conftest$ac_exeext $CFLAGS $CPPFLAGS $LDFLAGS conftest.$ac_ext $LIBS >&5' ac_compiler_gnu=$ac_cv_c_compiler_gnu CC="$lt_save_CC" ac_config_commands="$ac_config_commands libtool" # Only expand once: # Extract the first word of "ocamlbuild", so it can be a program name with args. set dummy ocamlbuild; ac_word=$2 { $as_echo "$as_me:${as_lineno-$LINENO}: checking for $ac_word" >&5 $as_echo_n "checking for $ac_word... " >&6; } if ${ac_cv_prog_OCAMLBUILD+:} false; then : $as_echo_n "(cached) " >&6 else if test -n "$OCAMLBUILD"; then ac_cv_prog_OCAMLBUILD="$OCAMLBUILD" # Let the user override the test. else as_save_IFS=$IFS; IFS=$PATH_SEPARATOR for as_dir in $PATH do IFS=$as_save_IFS test -z "$as_dir" && as_dir=. for ac_exec_ext in '' $ac_executable_extensions; do if as_fn_executable_p "$as_dir/$ac_word$ac_exec_ext"; then ac_cv_prog_OCAMLBUILD="ocamlbuild" $as_echo "$as_me:${as_lineno-$LINENO}: found $as_dir/$ac_word$ac_exec_ext" >&5 break 2 fi done done IFS=$as_save_IFS fi fi OCAMLBUILD=$ac_cv_prog_OCAMLBUILD if test -n "$OCAMLBUILD"; then { $as_echo "$as_me:${as_lineno-$LINENO}: result: $OCAMLBUILD" >&5 $as_echo "$OCAMLBUILD" >&6; } else { $as_echo "$as_me:${as_lineno-$LINENO}: result: no" >&5 $as_echo "no" >&6; } fi # Check whether --enable-mpi was given. if test "${enable_mpi+set}" = set; then : enableval=$enable_mpi; enable_mpi=$enableval else enable_mpi=no fi if test "$enable_mpi" = "yes"; then if test $PRECISION = q; then as_fn_error $? "quad precision is not supported in MPI" "$LINENO" 5 fi for ac_prog in mpicc hcc mpcc mpcc_r mpxlc cmpicc do # Extract the first word of "$ac_prog", so it can be a program name with args. set dummy $ac_prog; ac_word=$2 { $as_echo "$as_me:${as_lineno-$LINENO}: checking for $ac_word" >&5 $as_echo_n "checking for $ac_word... " >&6; } if ${ac_cv_prog_MPICC+:} false; then : $as_echo_n "(cached) " >&6 else if test -n "$MPICC"; then ac_cv_prog_MPICC="$MPICC" # Let the user override the test. else as_save_IFS=$IFS; IFS=$PATH_SEPARATOR for as_dir in $PATH do IFS=$as_save_IFS test -z "$as_dir" && as_dir=. for ac_exec_ext in '' $ac_executable_extensions; do if as_fn_executable_p "$as_dir/$ac_word$ac_exec_ext"; then ac_cv_prog_MPICC="$ac_prog" $as_echo "$as_me:${as_lineno-$LINENO}: found $as_dir/$ac_word$ac_exec_ext" >&5 break 2 fi done done IFS=$as_save_IFS fi fi MPICC=$ac_cv_prog_MPICC if test -n "$MPICC"; then { $as_echo "$as_me:${as_lineno-$LINENO}: result: $MPICC" >&5 $as_echo "$MPICC" >&6; } else { $as_echo "$as_me:${as_lineno-$LINENO}: result: no" >&5 $as_echo "no" >&6; } fi test -n "$MPICC" && break done test -n "$MPICC" || MPICC="$CC" acx_mpi_save_CC="$CC" CC="$MPICC" if test x = x"$MPILIBS"; then ac_fn_c_check_func "$LINENO" "MPI_Init" "ac_cv_func_MPI_Init" if test "x$ac_cv_func_MPI_Init" = xyes; then : MPILIBS=" " fi fi if test x = x"$MPILIBS"; then { $as_echo "$as_me:${as_lineno-$LINENO}: checking for MPI_Init in -lmpi" >&5 $as_echo_n "checking for MPI_Init in -lmpi... " >&6; } if ${ac_cv_lib_mpi_MPI_Init+:} false; then : $as_echo_n "(cached) " >&6 else ac_check_lib_save_LIBS=$LIBS LIBS="-lmpi $LIBS" cat confdefs.h - <<_ACEOF >conftest.$ac_ext /* end confdefs.h. */ /* Override any GCC internal prototype to avoid an error. Use char because int might match the return type of a GCC builtin and then its argument prototype would still apply. */ #ifdef __cplusplus extern "C" #endif char MPI_Init (); int main () { return MPI_Init (); ; return 0; } _ACEOF if ac_fn_c_try_link "$LINENO"; then : ac_cv_lib_mpi_MPI_Init=yes else ac_cv_lib_mpi_MPI_Init=no fi rm -f core conftest.err conftest.$ac_objext \ conftest$ac_exeext conftest.$ac_ext LIBS=$ac_check_lib_save_LIBS fi { $as_echo "$as_me:${as_lineno-$LINENO}: result: $ac_cv_lib_mpi_MPI_Init" >&5 $as_echo "$ac_cv_lib_mpi_MPI_Init" >&6; } if test "x$ac_cv_lib_mpi_MPI_Init" = xyes; then : MPILIBS="-lmpi" fi fi if test x = x"$MPILIBS"; then { $as_echo "$as_me:${as_lineno-$LINENO}: checking for MPI_Init in -lmpich" >&5 $as_echo_n "checking for MPI_Init in -lmpich... " >&6; } if ${ac_cv_lib_mpich_MPI_Init+:} false; then : $as_echo_n "(cached) " >&6 else ac_check_lib_save_LIBS=$LIBS LIBS="-lmpich $LIBS" cat confdefs.h - <<_ACEOF >conftest.$ac_ext /* end confdefs.h. */ /* Override any GCC internal prototype to avoid an error. Use char because int might match the return type of a GCC builtin and then its argument prototype would still apply. */ #ifdef __cplusplus extern "C" #endif char MPI_Init (); int main () { return MPI_Init (); ; return 0; } _ACEOF if ac_fn_c_try_link "$LINENO"; then : ac_cv_lib_mpich_MPI_Init=yes else ac_cv_lib_mpich_MPI_Init=no fi rm -f core conftest.err conftest.$ac_objext \ conftest$ac_exeext conftest.$ac_ext LIBS=$ac_check_lib_save_LIBS fi { $as_echo "$as_me:${as_lineno-$LINENO}: result: $ac_cv_lib_mpich_MPI_Init" >&5 $as_echo "$ac_cv_lib_mpich_MPI_Init" >&6; } if test "x$ac_cv_lib_mpich_MPI_Init" = xyes; then : MPILIBS="-lmpich" fi fi if test x != x"$MPILIBS"; then { $as_echo "$as_me:${as_lineno-$LINENO}: checking for mpi.h" >&5 $as_echo_n "checking for mpi.h... " >&6; } cat confdefs.h - <<_ACEOF >conftest.$ac_ext /* end confdefs.h. */ #include int main () { ; return 0; } _ACEOF if ac_fn_c_try_compile "$LINENO"; then : { $as_echo "$as_me:${as_lineno-$LINENO}: result: yes" >&5 $as_echo "yes" >&6; } else MPILIBS="" { $as_echo "$as_me:${as_lineno-$LINENO}: result: no" >&5 $as_echo "no" >&6; } fi rm -f core conftest.err conftest.$ac_objext conftest.$ac_ext fi CC="$acx_mpi_save_CC" # Finally, execute ACTION-IF-FOUND/ACTION-IF-NOT-FOUND: if test x = x"$MPILIBS"; then as_fn_error $? "could not find mpi library for --enable-mpi" "$LINENO" 5 : else $as_echo "#define HAVE_MPI 1" >>confdefs.h : fi # Extract the first word of "mpirun", so it can be a program name with args. set dummy mpirun; ac_word=$2 { $as_echo "$as_me:${as_lineno-$LINENO}: checking for $ac_word" >&5 $as_echo_n "checking for $ac_word... " >&6; } if ${ac_cv_prog_MPIRUN+:} false; then : $as_echo_n "(cached) " >&6 else if test -n "$MPIRUN"; then ac_cv_prog_MPIRUN="$MPIRUN" # Let the user override the test. else as_save_IFS=$IFS; IFS=$PATH_SEPARATOR for as_dir in $PATH do IFS=$as_save_IFS test -z "$as_dir" && as_dir=. for ac_exec_ext in '' $ac_executable_extensions; do if as_fn_executable_p "$as_dir/$ac_word$ac_exec_ext"; then ac_cv_prog_MPIRUN="mpirun" $as_echo "$as_me:${as_lineno-$LINENO}: found $as_dir/$ac_word$ac_exec_ext" >&5 break 2 fi done done IFS=$as_save_IFS fi fi MPIRUN=$ac_cv_prog_MPIRUN if test -n "$MPIRUN"; then { $as_echo "$as_me:${as_lineno-$LINENO}: result: $MPIRUN" >&5 $as_echo "$MPIRUN" >&6; } else { $as_echo "$as_me:${as_lineno-$LINENO}: result: no" >&5 $as_echo "no" >&6; } fi save_CC=$CC CC=$MPICC # The cast to long int works around a bug in the HP C Compiler # version HP92453-01 B.11.11.23709.GP, which incorrectly rejects # declarations like `int a3[[(sizeof (unsigned char)) >= 0]];'. # This bug is HP SR number 8606223364. { $as_echo "$as_me:${as_lineno-$LINENO}: checking size of MPI_Fint" >&5 $as_echo_n "checking size of MPI_Fint... " >&6; } if ${ac_cv_sizeof_MPI_Fint+:} false; then : $as_echo_n "(cached) " >&6 else if ac_fn_c_compute_int "$LINENO" "(long int) (sizeof (MPI_Fint))" "ac_cv_sizeof_MPI_Fint" "#include "; then : else if test "$ac_cv_type_MPI_Fint" = yes; then { { $as_echo "$as_me:${as_lineno-$LINENO}: error: in \`$ac_pwd':" >&5 $as_echo "$as_me: error: in \`$ac_pwd':" >&2;} as_fn_error 77 "cannot compute sizeof (MPI_Fint) See \`config.log' for more details" "$LINENO" 5; } else ac_cv_sizeof_MPI_Fint=0 fi fi fi { $as_echo "$as_me:${as_lineno-$LINENO}: result: $ac_cv_sizeof_MPI_Fint" >&5 $as_echo "$ac_cv_sizeof_MPI_Fint" >&6; } cat >>confdefs.h <<_ACEOF #define SIZEOF_MPI_FINT $ac_cv_sizeof_MPI_Fint _ACEOF CC=$save_CC if test 0 = $ac_cv_sizeof_MPI_Fint; then { $as_echo "$as_me:${as_lineno-$LINENO}: WARNING: sizeof(MPI_Fint) test failed" >&5 $as_echo "$as_me: WARNING: sizeof(MPI_Fint) test failed" >&2;}; # The cast to long int works around a bug in the HP C Compiler # version HP92453-01 B.11.11.23709.GP, which incorrectly rejects # declarations like `int a3[[(sizeof (unsigned char)) >= 0]];'. # This bug is HP SR number 8606223364. { $as_echo "$as_me:${as_lineno-$LINENO}: checking size of int" >&5 $as_echo_n "checking size of int... " >&6; } if ${ac_cv_sizeof_int+:} false; then : $as_echo_n "(cached) " >&6 else if ac_fn_c_compute_int "$LINENO" "(long int) (sizeof (int))" "ac_cv_sizeof_int" "$ac_includes_default"; then : else if test "$ac_cv_type_int" = yes; then { { $as_echo "$as_me:${as_lineno-$LINENO}: error: in \`$ac_pwd':" >&5 $as_echo "$as_me: error: in \`$ac_pwd':" >&2;} as_fn_error 77 "cannot compute sizeof (int) See \`config.log' for more details" "$LINENO" 5; } else ac_cv_sizeof_int=0 fi fi fi { $as_echo "$as_me:${as_lineno-$LINENO}: result: $ac_cv_sizeof_int" >&5 $as_echo "$ac_cv_sizeof_int" >&6; } cat >>confdefs.h <<_ACEOF #define SIZEOF_INT $ac_cv_sizeof_int _ACEOF if test 0 = $ac_cv_sizeof_int; then as_fn_error $? "sizeof(int) test failed" "$LINENO" 5; fi ac_cv_sizeof_MPI_Fint=$ac_cv_sizeof_int fi C_MPI_FINT=C_INT`expr $ac_cv_sizeof_MPI_Fint \* 8`_T fi if test "$enable_mpi" = "yes"; then MPI_TRUE= MPI_FALSE='#' else MPI_TRUE='#' MPI_FALSE= fi # Try to determine "good" native compiler flags if none specified via CFLAGS if test "$ac_test_CFLAGS" != "set"; then CFLAGS="" case $ax_cv_c_compiler_vendor in dec) CFLAGS="-newc -w0 -O5 -ansi_alias -ansi_args -fp_reorder -tune host" ;; sun) CFLAGS="-native -fast -xO5 -dalign" ;; hp) CFLAGS="+Oall +Optrs_ansi +DSnative" ;; ibm) xlc_opt="-qtune=auto" { $as_echo "$as_me:${as_lineno-$LINENO}: checking whether C compiler accepts $xlc_opt" >&5 $as_echo_n "checking whether C compiler accepts $xlc_opt... " >&6; } ax_save_FLAGS=$CFLAGS CFLAGS="$xlc_opt" cat confdefs.h - <<_ACEOF >conftest.$ac_ext /* end confdefs.h. */ int main () { ; return 0; } _ACEOF if ac_fn_c_try_compile "$LINENO"; then : eval `$as_echo "ax_cv_c_flags_$xlc_opt" | $as_tr_sh`=yes else eval `$as_echo "ax_cv_c_flags_$xlc_opt" | $as_tr_sh`=no fi rm -f core conftest.err conftest.$ac_objext conftest.$ac_ext CFLAGS=$ax_save_FLAGS eval ax_check_compiler_flags=$`$as_echo "ax_cv_c_flags_$xlc_opt" | $as_tr_sh` { $as_echo "$as_me:${as_lineno-$LINENO}: result: $ax_check_compiler_flags" >&5 $as_echo "$ax_check_compiler_flags" >&6; } if test "x$ax_check_compiler_flags" = xyes; then CFLAGS="-O3 -qansialias -w $xlc_opt" else CFLAGS="-O3 -qansialias -w" echo "******************************************************" echo "* You seem to have the IBM C compiler. It is *" echo "* recommended for best performance that you use: *" echo "* *" echo "* CFLAGS=-O3 -qarch=xxx -qtune=xxx -qansialias -w *" echo "* ^^^ ^^^ *" echo "* where xxx is pwr2, pwr3, 604, or whatever kind of *" echo "* CPU you have. (Set the CFLAGS environment var. *" echo "* and re-run configure.) For more info, man cc. *" echo "******************************************************" fi ;; intel) CFLAGS="-O3" # Intel seems to have changed the spelling of this flag recently icc_ansi_alias="unknown" for flag in -ansi-alias -ansi_alias; do { $as_echo "$as_me:${as_lineno-$LINENO}: checking whether C compiler accepts $flag" >&5 $as_echo_n "checking whether C compiler accepts $flag... " >&6; } ax_save_FLAGS=$CFLAGS CFLAGS="$flag" cat confdefs.h - <<_ACEOF >conftest.$ac_ext /* end confdefs.h. */ int main () { ; return 0; } _ACEOF if ac_fn_c_try_compile "$LINENO"; then : eval `$as_echo "ax_cv_c_flags_$flag" | $as_tr_sh`=yes else eval `$as_echo "ax_cv_c_flags_$flag" | $as_tr_sh`=no fi rm -f core conftest.err conftest.$ac_objext conftest.$ac_ext CFLAGS=$ax_save_FLAGS eval ax_check_compiler_flags=$`$as_echo "ax_cv_c_flags_$flag" | $as_tr_sh` { $as_echo "$as_me:${as_lineno-$LINENO}: result: $ax_check_compiler_flags" >&5 $as_echo "$ax_check_compiler_flags" >&6; } if test "x$ax_check_compiler_flags" = xyes; then icc_ansi_alias=$flag; break else : fi done if test "x$icc_ansi_alias" != xunknown; then CFLAGS="$CFLAGS $icc_ansi_alias" fi { $as_echo "$as_me:${as_lineno-$LINENO}: checking whether C compiler accepts -malign-double" >&5 $as_echo_n "checking whether C compiler accepts -malign-double... " >&6; } if ${ax_cv_c_flags__malign_double+:} false; then : $as_echo_n "(cached) " >&6 else ax_save_FLAGS=$CFLAGS CFLAGS="-malign-double" cat confdefs.h - <<_ACEOF >conftest.$ac_ext /* end confdefs.h. */ int main () { ; return 0; } _ACEOF if ac_fn_c_try_compile "$LINENO"; then : ax_cv_c_flags__malign_double=yes else ax_cv_c_flags__malign_double=no fi rm -f core conftest.err conftest.$ac_objext conftest.$ac_ext CFLAGS=$ax_save_FLAGS fi eval ax_check_compiler_flags=$ax_cv_c_flags__malign_double { $as_echo "$as_me:${as_lineno-$LINENO}: result: $ax_check_compiler_flags" >&5 $as_echo "$ax_check_compiler_flags" >&6; } if test "x$ax_check_compiler_flags" = xyes; then CFLAGS="$CFLAGS -malign-double" else : fi # We used to check for architecture flags here, e.g. -xHost etc., # but these flags are problematic. On icc-12.0.0, "-mavx -xHost" # overrides -mavx with -xHost, generating SSE2 code instead of AVX # code. ICC does not seem to support -mtune=host or equivalent # non-ABI changing flag. ;; gnu) # Default optimization flags for gcc on all systems. # Somehow -O3 does not imply -fomit-frame-pointer on ia32 CFLAGS="-O3 -fomit-frame-pointer" # tune for the host by default { $as_echo "$as_me:${as_lineno-$LINENO}: checking whether C compiler accepts -mtune=native" >&5 $as_echo_n "checking whether C compiler accepts -mtune=native... " >&6; } if ${ax_cv_c_flags__mtune_native+:} false; then : $as_echo_n "(cached) " >&6 else ax_save_FLAGS=$CFLAGS CFLAGS="-mtune=native" cat confdefs.h - <<_ACEOF >conftest.$ac_ext /* end confdefs.h. */ int main () { ; return 0; } _ACEOF if ac_fn_c_try_compile "$LINENO"; then : ax_cv_c_flags__mtune_native=yes else ax_cv_c_flags__mtune_native=no fi rm -f core conftest.err conftest.$ac_objext conftest.$ac_ext CFLAGS=$ax_save_FLAGS fi eval ax_check_compiler_flags=$ax_cv_c_flags__mtune_native { $as_echo "$as_me:${as_lineno-$LINENO}: result: $ax_check_compiler_flags" >&5 $as_echo "$ax_check_compiler_flags" >&6; } if test "x$ax_check_compiler_flags" = xyes; then CFLAGS="$CFLAGS -mtune=native" else : fi # -malign-double for x86 systems { $as_echo "$as_me:${as_lineno-$LINENO}: checking whether C compiler accepts -malign-double" >&5 $as_echo_n "checking whether C compiler accepts -malign-double... " >&6; } if ${ax_cv_c_flags__malign_double+:} false; then : $as_echo_n "(cached) " >&6 else ax_save_FLAGS=$CFLAGS CFLAGS="-malign-double" cat confdefs.h - <<_ACEOF >conftest.$ac_ext /* end confdefs.h. */ int main () { ; return 0; } _ACEOF if ac_fn_c_try_compile "$LINENO"; then : ax_cv_c_flags__malign_double=yes else ax_cv_c_flags__malign_double=no fi rm -f core conftest.err conftest.$ac_objext conftest.$ac_ext CFLAGS=$ax_save_FLAGS fi eval ax_check_compiler_flags=$ax_cv_c_flags__malign_double { $as_echo "$as_me:${as_lineno-$LINENO}: result: $ax_check_compiler_flags" >&5 $as_echo "$ax_check_compiler_flags" >&6; } if test "x$ax_check_compiler_flags" = xyes; then CFLAGS="$CFLAGS -malign-double" else : fi # -fstrict-aliasing for gcc-2.95+ { $as_echo "$as_me:${as_lineno-$LINENO}: checking whether C compiler accepts -fstrict-aliasing" >&5 $as_echo_n "checking whether C compiler accepts -fstrict-aliasing... " >&6; } if ${ax_cv_c_flags__fstrict_aliasing+:} false; then : $as_echo_n "(cached) " >&6 else ax_save_FLAGS=$CFLAGS CFLAGS="-fstrict-aliasing" cat confdefs.h - <<_ACEOF >conftest.$ac_ext /* end confdefs.h. */ int main () { ; return 0; } _ACEOF if ac_fn_c_try_compile "$LINENO"; then : ax_cv_c_flags__fstrict_aliasing=yes else ax_cv_c_flags__fstrict_aliasing=no fi rm -f core conftest.err conftest.$ac_objext conftest.$ac_ext CFLAGS=$ax_save_FLAGS fi eval ax_check_compiler_flags=$ax_cv_c_flags__fstrict_aliasing { $as_echo "$as_me:${as_lineno-$LINENO}: result: $ax_check_compiler_flags" >&5 $as_echo "$ax_check_compiler_flags" >&6; } if test "x$ax_check_compiler_flags" = xyes; then CFLAGS="$CFLAGS -fstrict-aliasing" else : fi # -fno-schedule-insns is pretty much required on all risc # processors. # # gcc performs one pass of instruction scheduling, then a pass of # register allocation, then another pass of instruction # scheduling. The first pass reorders instructions in a way that # is pretty much the worst possible for the purposes of register # allocation. We disable the first pass. { $as_echo "$as_me:${as_lineno-$LINENO}: checking whether C compiler accepts -fno-schedule-insns" >&5 $as_echo_n "checking whether C compiler accepts -fno-schedule-insns... " >&6; } if ${ax_cv_c_flags__fno_schedule_insns+:} false; then : $as_echo_n "(cached) " >&6 else ax_save_FLAGS=$CFLAGS CFLAGS="-fno-schedule-insns" cat confdefs.h - <<_ACEOF >conftest.$ac_ext /* end confdefs.h. */ int main () { ; return 0; } _ACEOF if ac_fn_c_try_compile "$LINENO"; then : ax_cv_c_flags__fno_schedule_insns=yes else ax_cv_c_flags__fno_schedule_insns=no fi rm -f core conftest.err conftest.$ac_objext conftest.$ac_ext CFLAGS=$ax_save_FLAGS fi eval ax_check_compiler_flags=$ax_cv_c_flags__fno_schedule_insns { $as_echo "$as_me:${as_lineno-$LINENO}: result: $ax_check_compiler_flags" >&5 $as_echo "$ax_check_compiler_flags" >&6; } if test "x$ax_check_compiler_flags" = xyes; then CFLAGS="$CFLAGS -fno-schedule-insns" else : fi # note that we enable "unsafe" fp optimization with other compilers, too { $as_echo "$as_me:${as_lineno-$LINENO}: checking whether C compiler accepts -ffast-math" >&5 $as_echo_n "checking whether C compiler accepts -ffast-math... " >&6; } if ${ax_cv_c_flags__ffast_math+:} false; then : $as_echo_n "(cached) " >&6 else ax_save_FLAGS=$CFLAGS CFLAGS="-ffast-math" cat confdefs.h - <<_ACEOF >conftest.$ac_ext /* end confdefs.h. */ int main () { ; return 0; } _ACEOF if ac_fn_c_try_compile "$LINENO"; then : ax_cv_c_flags__ffast_math=yes else ax_cv_c_flags__ffast_math=no fi rm -f core conftest.err conftest.$ac_objext conftest.$ac_ext CFLAGS=$ax_save_FLAGS fi eval ax_check_compiler_flags=$ax_cv_c_flags__ffast_math { $as_echo "$as_me:${as_lineno-$LINENO}: result: $ax_check_compiler_flags" >&5 $as_echo "$ax_check_compiler_flags" >&6; } if test "x$ax_check_compiler_flags" = xyes; then CFLAGS="$CFLAGS -ffast-math" else : fi ;; esac if test -z "$CFLAGS"; then echo "" echo "********************************************************" echo "* WARNING: Don't know the best CFLAGS for this system *" echo "* Use ./configure CFLAGS=... to specify your own flags *" echo "* (otherwise, a default of CFLAGS=-O3 will be used) *" echo "********************************************************" echo "" CFLAGS="-O3" fi { $as_echo "$as_me:${as_lineno-$LINENO}: checking whether C compiler accepts $CFLAGS" >&5 $as_echo_n "checking whether C compiler accepts $CFLAGS... " >&6; } ax_save_FLAGS=$CFLAGS CFLAGS="$CFLAGS" cat confdefs.h - <<_ACEOF >conftest.$ac_ext /* end confdefs.h. */ int main () { ; return 0; } _ACEOF if ac_fn_c_try_compile "$LINENO"; then : eval `$as_echo "ax_cv_c_flags_$CFLAGS" | $as_tr_sh`=yes else eval `$as_echo "ax_cv_c_flags_$CFLAGS" | $as_tr_sh`=no fi rm -f core conftest.err conftest.$ac_objext conftest.$ac_ext CFLAGS=$ax_save_FLAGS eval ax_check_compiler_flags=$`$as_echo "ax_cv_c_flags_$CFLAGS" | $as_tr_sh` { $as_echo "$as_me:${as_lineno-$LINENO}: result: $ax_check_compiler_flags" >&5 $as_echo "$ax_check_compiler_flags" >&6; } if test "x$ax_check_compiler_flags" = xyes; then : else echo "" echo "********************************************************" echo "* WARNING: The guessed CFLAGS don't seem to work with *" echo "* your compiler. *" echo "* Use ./configure CFLAGS=... to specify your own flags *" echo "********************************************************" echo "" CFLAGS="" fi fi case "${ax_cv_c_compiler_vendor}" in intel) # Stop icc from defining __GNUC__, except on MacOS where this fails case "${host_os}" in *darwin*) ;; # icc -no-gcc fails to compile some system headers *) { $as_echo "$as_me:${as_lineno-$LINENO}: checking whether C compiler accepts -no-gcc" >&5 $as_echo_n "checking whether C compiler accepts -no-gcc... " >&6; } if ${ax_cv_c_flags__no_gcc+:} false; then : $as_echo_n "(cached) " >&6 else ax_save_FLAGS=$CFLAGS CFLAGS="-no-gcc" cat confdefs.h - <<_ACEOF >conftest.$ac_ext /* end confdefs.h. */ int main () { ; return 0; } _ACEOF if ac_fn_c_try_compile "$LINENO"; then : ax_cv_c_flags__no_gcc=yes else ax_cv_c_flags__no_gcc=no fi rm -f core conftest.err conftest.$ac_objext conftest.$ac_ext CFLAGS=$ax_save_FLAGS fi eval ax_check_compiler_flags=$ax_cv_c_flags__no_gcc { $as_echo "$as_me:${as_lineno-$LINENO}: result: $ax_check_compiler_flags" >&5 $as_echo "$ax_check_compiler_flags" >&6; } if test "x$ax_check_compiler_flags" = xyes; then CC="$CC -no-gcc" else : fi ;; esac ;; hp) # must (sometimes) manually increase cpp limits to handle fftw3.h { $as_echo "$as_me:${as_lineno-$LINENO}: checking whether C compiler accepts -Wp,-H128000" >&5 $as_echo_n "checking whether C compiler accepts -Wp,-H128000... " >&6; } if ${ax_cv_c_flags__Wp+:} false; then : $as_echo_n "(cached) " >&6 else ax_save_FLAGS=$CFLAGS CFLAGS="-Wp,-H128000" cat confdefs.h - <<_ACEOF >conftest.$ac_ext /* end confdefs.h. */ int main () { ; return 0; } _ACEOF if ac_fn_c_try_compile "$LINENO"; then : ax_cv_c_flags__Wp=yes else ax_cv_c_flags__Wp=no fi rm -f core conftest.err conftest.$ac_objext conftest.$ac_ext CFLAGS=$ax_save_FLAGS fi eval ax_check_compiler_flags=$ax_cv_c_flags__Wp { $as_echo "$as_me:${as_lineno-$LINENO}: result: $ax_check_compiler_flags" >&5 $as_echo "$ax_check_compiler_flags" >&6; } if test "x$ax_check_compiler_flags" = xyes; then CC="$CC -Wp,-H128000" else : fi ;; portland) # -Masmkeyword required for asm("") cycle counters { $as_echo "$as_me:${as_lineno-$LINENO}: checking whether C compiler accepts -Masmkeyword" >&5 $as_echo_n "checking whether C compiler accepts -Masmkeyword... " >&6; } if ${ax_cv_c_flags__Masmkeyword+:} false; then : $as_echo_n "(cached) " >&6 else ax_save_FLAGS=$CFLAGS CFLAGS="-Masmkeyword" cat confdefs.h - <<_ACEOF >conftest.$ac_ext /* end confdefs.h. */ int main () { ; return 0; } _ACEOF if ac_fn_c_try_compile "$LINENO"; then : ax_cv_c_flags__Masmkeyword=yes else ax_cv_c_flags__Masmkeyword=no fi rm -f core conftest.err conftest.$ac_objext conftest.$ac_ext CFLAGS=$ax_save_FLAGS fi eval ax_check_compiler_flags=$ax_cv_c_flags__Masmkeyword { $as_echo "$as_me:${as_lineno-$LINENO}: result: $ax_check_compiler_flags" >&5 $as_echo "$ax_check_compiler_flags" >&6; } if test "x$ax_check_compiler_flags" = xyes; then CC="$CC -Masmkeyword" else : fi ;; esac case "${ax_cv_c_compiler_vendor}" in gnu|intel) # SSE/SSE2 if test "$have_sse2" = "yes" -a "x$SSE2_CFLAGS" = x; then if test "$PRECISION" = d; then flag=msse2; else flag=msse; fi { $as_echo "$as_me:${as_lineno-$LINENO}: checking whether C compiler accepts -$flag" >&5 $as_echo_n "checking whether C compiler accepts -$flag... " >&6; } ax_save_FLAGS=$CFLAGS CFLAGS="-$flag" cat confdefs.h - <<_ACEOF >conftest.$ac_ext /* end confdefs.h. */ int main () { ; return 0; } _ACEOF if ac_fn_c_try_compile "$LINENO"; then : eval `$as_echo "ax_cv_c_flags_-$flag" | $as_tr_sh`=yes else eval `$as_echo "ax_cv_c_flags_-$flag" | $as_tr_sh`=no fi rm -f core conftest.err conftest.$ac_objext conftest.$ac_ext CFLAGS=$ax_save_FLAGS eval ax_check_compiler_flags=$`$as_echo "ax_cv_c_flags_-$flag" | $as_tr_sh` { $as_echo "$as_me:${as_lineno-$LINENO}: result: $ax_check_compiler_flags" >&5 $as_echo "$ax_check_compiler_flags" >&6; } if test "x$ax_check_compiler_flags" = xyes; then SSE2_CFLAGS="-$flag" else as_fn_error $? "Need a version of gcc with -$flag" "$LINENO" 5 fi fi # AVX if test "$have_avx" = "yes" -a "x$AVX_CFLAGS" = x; then { $as_echo "$as_me:${as_lineno-$LINENO}: checking whether C compiler accepts -mavx" >&5 $as_echo_n "checking whether C compiler accepts -mavx... " >&6; } if ${ax_cv_c_flags__mavx+:} false; then : $as_echo_n "(cached) " >&6 else ax_save_FLAGS=$CFLAGS CFLAGS="-mavx" cat confdefs.h - <<_ACEOF >conftest.$ac_ext /* end confdefs.h. */ int main () { ; return 0; } _ACEOF if ac_fn_c_try_compile "$LINENO"; then : ax_cv_c_flags__mavx=yes else ax_cv_c_flags__mavx=no fi rm -f core conftest.err conftest.$ac_objext conftest.$ac_ext CFLAGS=$ax_save_FLAGS fi eval ax_check_compiler_flags=$ax_cv_c_flags__mavx { $as_echo "$as_me:${as_lineno-$LINENO}: result: $ax_check_compiler_flags" >&5 $as_echo "$ax_check_compiler_flags" >&6; } if test "x$ax_check_compiler_flags" = xyes; then AVX_CFLAGS="-mavx" else as_fn_error $? "Need a version of gcc with -mavx" "$LINENO" 5 fi fi if test "$have_altivec" = "yes" -a "x$ALTIVEC_CFLAGS" = x; then # -DFAKE__VEC__ is a workaround because gcc-3.3 does not # #define __VEC__ with -maltivec. { $as_echo "$as_me:${as_lineno-$LINENO}: checking whether C compiler accepts -faltivec" >&5 $as_echo_n "checking whether C compiler accepts -faltivec... " >&6; } if ${ax_cv_c_flags__faltivec+:} false; then : $as_echo_n "(cached) " >&6 else ax_save_FLAGS=$CFLAGS CFLAGS="-faltivec" cat confdefs.h - <<_ACEOF >conftest.$ac_ext /* end confdefs.h. */ int main () { ; return 0; } _ACEOF if ac_fn_c_try_compile "$LINENO"; then : ax_cv_c_flags__faltivec=yes else ax_cv_c_flags__faltivec=no fi rm -f core conftest.err conftest.$ac_objext conftest.$ac_ext CFLAGS=$ax_save_FLAGS fi eval ax_check_compiler_flags=$ax_cv_c_flags__faltivec { $as_echo "$as_me:${as_lineno-$LINENO}: result: $ax_check_compiler_flags" >&5 $as_echo "$ax_check_compiler_flags" >&6; } if test "x$ax_check_compiler_flags" = xyes; then ALTIVEC_CFLAGS="-faltivec" else { $as_echo "$as_me:${as_lineno-$LINENO}: checking whether C compiler accepts -maltivec -mabi=altivec" >&5 $as_echo_n "checking whether C compiler accepts -maltivec -mabi=altivec... " >&6; } if ${ax_cv_c_flags__maltivec__mabi_altivec+:} false; then : $as_echo_n "(cached) " >&6 else ax_save_FLAGS=$CFLAGS CFLAGS="-maltivec -mabi=altivec" cat confdefs.h - <<_ACEOF >conftest.$ac_ext /* end confdefs.h. */ int main () { ; return 0; } _ACEOF if ac_fn_c_try_compile "$LINENO"; then : ax_cv_c_flags__maltivec__mabi_altivec=yes else ax_cv_c_flags__maltivec__mabi_altivec=no fi rm -f core conftest.err conftest.$ac_objext conftest.$ac_ext CFLAGS=$ax_save_FLAGS fi eval ax_check_compiler_flags=$ax_cv_c_flags__maltivec__mabi_altivec { $as_echo "$as_me:${as_lineno-$LINENO}: result: $ax_check_compiler_flags" >&5 $as_echo "$ax_check_compiler_flags" >&6; } if test "x$ax_check_compiler_flags" = xyes; then ALTIVEC_CFLAGS="-maltivec -mabi=altivec -DFAKE__VEC__" else { $as_echo "$as_me:${as_lineno-$LINENO}: checking whether C compiler accepts -fvec" >&5 $as_echo_n "checking whether C compiler accepts -fvec... " >&6; } if ${ax_cv_c_flags__fvec+:} false; then : $as_echo_n "(cached) " >&6 else ax_save_FLAGS=$CFLAGS CFLAGS="-fvec" cat confdefs.h - <<_ACEOF >conftest.$ac_ext /* end confdefs.h. */ int main () { ; return 0; } _ACEOF if ac_fn_c_try_compile "$LINENO"; then : ax_cv_c_flags__fvec=yes else ax_cv_c_flags__fvec=no fi rm -f core conftest.err conftest.$ac_objext conftest.$ac_ext CFLAGS=$ax_save_FLAGS fi eval ax_check_compiler_flags=$ax_cv_c_flags__fvec { $as_echo "$as_me:${as_lineno-$LINENO}: result: $ax_check_compiler_flags" >&5 $as_echo "$ax_check_compiler_flags" >&6; } if test "x$ax_check_compiler_flags" = xyes; then ALTIVEC_CFLAGS="-fvec" else as_fn_error $? "Need a version of gcc with -maltivec" "$LINENO" 5 fi fi fi fi if test "$have_neon" = "yes" -a "x$NEON_CFLAGS" = x; then { $as_echo "$as_me:${as_lineno-$LINENO}: checking whether C compiler accepts -mfpu=neon" >&5 $as_echo_n "checking whether C compiler accepts -mfpu=neon... " >&6; } if ${ax_cv_c_flags__mfpu_neon+:} false; then : $as_echo_n "(cached) " >&6 else ax_save_FLAGS=$CFLAGS CFLAGS="-mfpu=neon" cat confdefs.h - <<_ACEOF >conftest.$ac_ext /* end confdefs.h. */ int main () { ; return 0; } _ACEOF if ac_fn_c_try_compile "$LINENO"; then : ax_cv_c_flags__mfpu_neon=yes else ax_cv_c_flags__mfpu_neon=no fi rm -f core conftest.err conftest.$ac_objext conftest.$ac_ext CFLAGS=$ax_save_FLAGS fi eval ax_check_compiler_flags=$ax_cv_c_flags__mfpu_neon { $as_echo "$as_me:${as_lineno-$LINENO}: result: $ax_check_compiler_flags" >&5 $as_echo "$ax_check_compiler_flags" >&6; } if test "x$ax_check_compiler_flags" = xyes; then NEON_CFLAGS="-mfpu=neon" else as_fn_error $? "Need a version of gcc with -mfpu=neon" "$LINENO" 5 fi fi ;; esac if test "$with_incoming_stack_boundary"x != "no"x; then case "${ax_cv_c_compiler_vendor}" in gnu) tentative_flags="-mincoming-stack-boundary=$with_incoming_stack_boundary"; { $as_echo "$as_me:${as_lineno-$LINENO}: checking whether C compiler accepts $tentative_flags" >&5 $as_echo_n "checking whether C compiler accepts $tentative_flags... " >&6; } ax_save_FLAGS=$CFLAGS CFLAGS="$tentative_flags" cat confdefs.h - <<_ACEOF >conftest.$ac_ext /* end confdefs.h. */ int main () { ; return 0; } _ACEOF if ac_fn_c_try_compile "$LINENO"; then : eval `$as_echo "ax_cv_c_flags_$tentative_flags" | $as_tr_sh`=yes else eval `$as_echo "ax_cv_c_flags_$tentative_flags" | $as_tr_sh`=no fi rm -f core conftest.err conftest.$ac_objext conftest.$ac_ext CFLAGS=$ax_save_FLAGS eval ax_check_compiler_flags=$`$as_echo "ax_cv_c_flags_$tentative_flags" | $as_tr_sh` { $as_echo "$as_me:${as_lineno-$LINENO}: result: $ax_check_compiler_flags" >&5 $as_echo "$ax_check_compiler_flags" >&6; } if test "x$ax_check_compiler_flags" = xyes; then STACK_ALIGN_CFLAGS=$tentative_flags else : fi ;; esac fi { $as_echo "$as_me:${as_lineno-$LINENO}: checking for ANSI C header files" >&5 $as_echo_n "checking for ANSI C header files... " >&6; } if ${ac_cv_header_stdc+:} false; then : $as_echo_n "(cached) " >&6 else cat confdefs.h - <<_ACEOF >conftest.$ac_ext /* end confdefs.h. */ #include #include #include #include int main () { ; return 0; } _ACEOF if ac_fn_c_try_compile "$LINENO"; then : ac_cv_header_stdc=yes else ac_cv_header_stdc=no fi rm -f core conftest.err conftest.$ac_objext conftest.$ac_ext if test $ac_cv_header_stdc = yes; then # SunOS 4.x string.h does not declare mem*, contrary to ANSI. cat confdefs.h - <<_ACEOF >conftest.$ac_ext /* end confdefs.h. */ #include _ACEOF if (eval "$ac_cpp conftest.$ac_ext") 2>&5 | $EGREP "memchr" >/dev/null 2>&1; then : else ac_cv_header_stdc=no fi rm -f conftest* fi if test $ac_cv_header_stdc = yes; then # ISC 2.0.2 stdlib.h does not declare free, contrary to ANSI. cat confdefs.h - <<_ACEOF >conftest.$ac_ext /* end confdefs.h. */ #include _ACEOF if (eval "$ac_cpp conftest.$ac_ext") 2>&5 | $EGREP "free" >/dev/null 2>&1; then : else ac_cv_header_stdc=no fi rm -f conftest* fi if test $ac_cv_header_stdc = yes; then # /bin/cc in Irix-4.0.5 gets non-ANSI ctype macros unless using -ansi. if test "$cross_compiling" = yes; then : : else cat confdefs.h - <<_ACEOF >conftest.$ac_ext /* end confdefs.h. */ #include #include #if ((' ' & 0x0FF) == 0x020) # define ISLOWER(c) ('a' <= (c) && (c) <= 'z') # define TOUPPER(c) (ISLOWER(c) ? 'A' + ((c) - 'a') : (c)) #else # define ISLOWER(c) \ (('a' <= (c) && (c) <= 'i') \ || ('j' <= (c) && (c) <= 'r') \ || ('s' <= (c) && (c) <= 'z')) # define TOUPPER(c) (ISLOWER(c) ? ((c) | 0x40) : (c)) #endif #define XOR(e, f) (((e) && !(f)) || (!(e) && (f))) int main () { int i; for (i = 0; i < 256; i++) if (XOR (islower (i), ISLOWER (i)) || toupper (i) != TOUPPER (i)) return 2; return 0; } _ACEOF if ac_fn_c_try_run "$LINENO"; then : else ac_cv_header_stdc=no fi rm -f core *.core core.conftest.* gmon.out bb.out conftest$ac_exeext \ conftest.$ac_objext conftest.beam conftest.$ac_ext fi fi fi { $as_echo "$as_me:${as_lineno-$LINENO}: result: $ac_cv_header_stdc" >&5 $as_echo "$ac_cv_header_stdc" >&6; } if test $ac_cv_header_stdc = yes; then $as_echo "#define STDC_HEADERS 1" >>confdefs.h fi for ac_header in libintl.h malloc.h stddef.h stdlib.h string.h strings.h sys/time.h unistd.h limits.h c_asm.h intrinsics.h stdint.h mach/mach_time.h sys/sysctl.h do : as_ac_Header=`$as_echo "ac_cv_header_$ac_header" | $as_tr_sh` ac_fn_c_check_header_mongrel "$LINENO" "$ac_header" "$as_ac_Header" "$ac_includes_default" if eval test \"x\$"$as_ac_Header"\" = x"yes"; then : cat >>confdefs.h <<_ACEOF #define `$as_echo "HAVE_$ac_header" | $as_tr_cpp` 1 _ACEOF fi done save_CFLAGS="$CFLAGS" save_CPPFLAGS="$CPPFLAGS" CFLAGS="$CFLAGS $ALTIVEC_CFLAGS" CPPFLAGS="$CPPFLAGS $ALTIVEC_CFLAGS" for ac_header in altivec.h do : ac_fn_c_check_header_mongrel "$LINENO" "altivec.h" "ac_cv_header_altivec_h" "$ac_includes_default" if test "x$ac_cv_header_altivec_h" = xyes; then : cat >>confdefs.h <<_ACEOF #define HAVE_ALTIVEC_H 1 _ACEOF fi done CFLAGS="$save_CFLAGS" CPPFLAGS="$save_CPPFLAGS" { $as_echo "$as_me:${as_lineno-$LINENO}: checking for an ANSI C-conforming const" >&5 $as_echo_n "checking for an ANSI C-conforming const... " >&6; } if ${ac_cv_c_const+:} false; then : $as_echo_n "(cached) " >&6 else cat confdefs.h - <<_ACEOF >conftest.$ac_ext /* end confdefs.h. */ int main () { #ifndef __cplusplus /* Ultrix mips cc rejects this sort of thing. */ typedef int charset[2]; const charset cs = { 0, 0 }; /* SunOS 4.1.1 cc rejects this. */ char const *const *pcpcc; char **ppc; /* NEC SVR4.0.2 mips cc rejects this. */ struct point {int x, y;}; static struct point const zero = {0,0}; /* AIX XL C 1.02.0.0 rejects this. It does not let you subtract one const X* pointer from another in an arm of an if-expression whose if-part is not a constant expression */ const char *g = "string"; pcpcc = &g + (g ? g-g : 0); /* HPUX 7.0 cc rejects these. */ ++pcpcc; ppc = (char**) pcpcc; pcpcc = (char const *const *) ppc; { /* SCO 3.2v4 cc rejects this sort of thing. */ char tx; char *t = &tx; char const *s = 0 ? (char *) 0 : (char const *) 0; *t++ = 0; if (s) return 0; } { /* Someone thinks the Sun supposedly-ANSI compiler will reject this. */ int x[] = {25, 17}; const int *foo = &x[0]; ++foo; } { /* Sun SC1.0 ANSI compiler rejects this -- but not the above. */ typedef const int *iptr; iptr p = 0; ++p; } { /* AIX XL C 1.02.0.0 rejects this sort of thing, saying "k.c", line 2.27: 1506-025 (S) Operand must be a modifiable lvalue. */ struct s { int j; const int *ap[3]; } bx; struct s *b = &bx; b->j = 5; } { /* ULTRIX-32 V3.1 (Rev 9) vcc rejects this */ const int foo = 10; if (!foo) return 0; } return !cs[0] && !zero.x; #endif ; return 0; } _ACEOF if ac_fn_c_try_compile "$LINENO"; then : ac_cv_c_const=yes else ac_cv_c_const=no fi rm -f core conftest.err conftest.$ac_objext conftest.$ac_ext fi { $as_echo "$as_me:${as_lineno-$LINENO}: result: $ac_cv_c_const" >&5 $as_echo "$ac_cv_c_const" >&6; } if test $ac_cv_c_const = no; then $as_echo "#define const /**/" >>confdefs.h fi { $as_echo "$as_me:${as_lineno-$LINENO}: checking for inline" >&5 $as_echo_n "checking for inline... " >&6; } if ${ac_cv_c_inline+:} false; then : $as_echo_n "(cached) " >&6 else ac_cv_c_inline=no for ac_kw in inline __inline__ __inline; do cat confdefs.h - <<_ACEOF >conftest.$ac_ext /* end confdefs.h. */ #ifndef __cplusplus typedef int foo_t; static $ac_kw foo_t static_foo () {return 0; } $ac_kw foo_t foo () {return 0; } #endif _ACEOF if ac_fn_c_try_compile "$LINENO"; then : ac_cv_c_inline=$ac_kw fi rm -f core conftest.err conftest.$ac_objext conftest.$ac_ext test "$ac_cv_c_inline" != no && break done fi { $as_echo "$as_me:${as_lineno-$LINENO}: result: $ac_cv_c_inline" >&5 $as_echo "$ac_cv_c_inline" >&6; } case $ac_cv_c_inline in inline | yes) ;; *) case $ac_cv_c_inline in no) ac_val=;; *) ac_val=$ac_cv_c_inline;; esac cat >>confdefs.h <<_ACEOF #ifndef __cplusplus #define inline $ac_val #endif _ACEOF ;; esac ac_fn_c_check_type "$LINENO" "size_t" "ac_cv_type_size_t" "$ac_includes_default" if test "x$ac_cv_type_size_t" = xyes; then : else cat >>confdefs.h <<_ACEOF #define size_t unsigned int _ACEOF fi { $as_echo "$as_me:${as_lineno-$LINENO}: checking whether time.h and sys/time.h may both be included" >&5 $as_echo_n "checking whether time.h and sys/time.h may both be included... " >&6; } if ${ac_cv_header_time+:} false; then : $as_echo_n "(cached) " >&6 else cat confdefs.h - <<_ACEOF >conftest.$ac_ext /* end confdefs.h. */ #include #include #include int main () { if ((struct tm *) 0) return 0; ; return 0; } _ACEOF if ac_fn_c_try_compile "$LINENO"; then : ac_cv_header_time=yes else ac_cv_header_time=no fi rm -f core conftest.err conftest.$ac_objext conftest.$ac_ext fi { $as_echo "$as_me:${as_lineno-$LINENO}: result: $ac_cv_header_time" >&5 $as_echo "$ac_cv_header_time" >&6; } if test $ac_cv_header_time = yes; then $as_echo "#define TIME_WITH_SYS_TIME 1" >>confdefs.h fi ac_fn_c_check_type "$LINENO" "long double" "ac_cv_type_long_double" "$ac_includes_default" if test "x$ac_cv_type_long_double" = xyes; then : $as_echo "#define HAVE_LONG_DOUBLE 1" >>confdefs.h else if test $PRECISION = l; then as_fn_error $? "long double is not a supported type with your compiler." "$LINENO" 5 fi fi ac_fn_c_check_type "$LINENO" "hrtime_t" "ac_cv_type_hrtime_t" " #if HAVE_SYS_TIME_H #include #endif " if test "x$ac_cv_type_hrtime_t" = xyes; then : $as_echo "#define HAVE_HRTIME_T 1" >>confdefs.h fi # The cast to long int works around a bug in the HP C Compiler # version HP92453-01 B.11.11.23709.GP, which incorrectly rejects # declarations like `int a3[[(sizeof (unsigned char)) >= 0]];'. # This bug is HP SR number 8606223364. { $as_echo "$as_me:${as_lineno-$LINENO}: checking size of int" >&5 $as_echo_n "checking size of int... " >&6; } if ${ac_cv_sizeof_int+:} false; then : $as_echo_n "(cached) " >&6 else if ac_fn_c_compute_int "$LINENO" "(long int) (sizeof (int))" "ac_cv_sizeof_int" "$ac_includes_default"; then : else if test "$ac_cv_type_int" = yes; then { { $as_echo "$as_me:${as_lineno-$LINENO}: error: in \`$ac_pwd':" >&5 $as_echo "$as_me: error: in \`$ac_pwd':" >&2;} as_fn_error 77 "cannot compute sizeof (int) See \`config.log' for more details" "$LINENO" 5; } else ac_cv_sizeof_int=0 fi fi fi { $as_echo "$as_me:${as_lineno-$LINENO}: result: $ac_cv_sizeof_int" >&5 $as_echo "$ac_cv_sizeof_int" >&6; } cat >>confdefs.h <<_ACEOF #define SIZEOF_INT $ac_cv_sizeof_int _ACEOF # The cast to long int works around a bug in the HP C Compiler # version HP92453-01 B.11.11.23709.GP, which incorrectly rejects # declarations like `int a3[[(sizeof (unsigned char)) >= 0]];'. # This bug is HP SR number 8606223364. { $as_echo "$as_me:${as_lineno-$LINENO}: checking size of unsigned int" >&5 $as_echo_n "checking size of unsigned int... " >&6; } if ${ac_cv_sizeof_unsigned_int+:} false; then : $as_echo_n "(cached) " >&6 else if ac_fn_c_compute_int "$LINENO" "(long int) (sizeof (unsigned int))" "ac_cv_sizeof_unsigned_int" "$ac_includes_default"; then : else if test "$ac_cv_type_unsigned_int" = yes; then { { $as_echo "$as_me:${as_lineno-$LINENO}: error: in \`$ac_pwd':" >&5 $as_echo "$as_me: error: in \`$ac_pwd':" >&2;} as_fn_error 77 "cannot compute sizeof (unsigned int) See \`config.log' for more details" "$LINENO" 5; } else ac_cv_sizeof_unsigned_int=0 fi fi fi { $as_echo "$as_me:${as_lineno-$LINENO}: result: $ac_cv_sizeof_unsigned_int" >&5 $as_echo "$ac_cv_sizeof_unsigned_int" >&6; } cat >>confdefs.h <<_ACEOF #define SIZEOF_UNSIGNED_INT $ac_cv_sizeof_unsigned_int _ACEOF # The cast to long int works around a bug in the HP C Compiler # version HP92453-01 B.11.11.23709.GP, which incorrectly rejects # declarations like `int a3[[(sizeof (unsigned char)) >= 0]];'. # This bug is HP SR number 8606223364. { $as_echo "$as_me:${as_lineno-$LINENO}: checking size of long" >&5 $as_echo_n "checking size of long... " >&6; } if ${ac_cv_sizeof_long+:} false; then : $as_echo_n "(cached) " >&6 else if ac_fn_c_compute_int "$LINENO" "(long int) (sizeof (long))" "ac_cv_sizeof_long" "$ac_includes_default"; then : else if test "$ac_cv_type_long" = yes; then { { $as_echo "$as_me:${as_lineno-$LINENO}: error: in \`$ac_pwd':" >&5 $as_echo "$as_me: error: in \`$ac_pwd':" >&2;} as_fn_error 77 "cannot compute sizeof (long) See \`config.log' for more details" "$LINENO" 5; } else ac_cv_sizeof_long=0 fi fi fi { $as_echo "$as_me:${as_lineno-$LINENO}: result: $ac_cv_sizeof_long" >&5 $as_echo "$ac_cv_sizeof_long" >&6; } cat >>confdefs.h <<_ACEOF #define SIZEOF_LONG $ac_cv_sizeof_long _ACEOF # The cast to long int works around a bug in the HP C Compiler # version HP92453-01 B.11.11.23709.GP, which incorrectly rejects # declarations like `int a3[[(sizeof (unsigned char)) >= 0]];'. # This bug is HP SR number 8606223364. { $as_echo "$as_me:${as_lineno-$LINENO}: checking size of unsigned long" >&5 $as_echo_n "checking size of unsigned long... " >&6; } if ${ac_cv_sizeof_unsigned_long+:} false; then : $as_echo_n "(cached) " >&6 else if ac_fn_c_compute_int "$LINENO" "(long int) (sizeof (unsigned long))" "ac_cv_sizeof_unsigned_long" "$ac_includes_default"; then : else if test "$ac_cv_type_unsigned_long" = yes; then { { $as_echo "$as_me:${as_lineno-$LINENO}: error: in \`$ac_pwd':" >&5 $as_echo "$as_me: error: in \`$ac_pwd':" >&2;} as_fn_error 77 "cannot compute sizeof (unsigned long) See \`config.log' for more details" "$LINENO" 5; } else ac_cv_sizeof_unsigned_long=0 fi fi fi { $as_echo "$as_me:${as_lineno-$LINENO}: result: $ac_cv_sizeof_unsigned_long" >&5 $as_echo "$ac_cv_sizeof_unsigned_long" >&6; } cat >>confdefs.h <<_ACEOF #define SIZEOF_UNSIGNED_LONG $ac_cv_sizeof_unsigned_long _ACEOF # The cast to long int works around a bug in the HP C Compiler # version HP92453-01 B.11.11.23709.GP, which incorrectly rejects # declarations like `int a3[[(sizeof (unsigned char)) >= 0]];'. # This bug is HP SR number 8606223364. { $as_echo "$as_me:${as_lineno-$LINENO}: checking size of long long" >&5 $as_echo_n "checking size of long long... " >&6; } if ${ac_cv_sizeof_long_long+:} false; then : $as_echo_n "(cached) " >&6 else if ac_fn_c_compute_int "$LINENO" "(long int) (sizeof (long long))" "ac_cv_sizeof_long_long" "$ac_includes_default"; then : else if test "$ac_cv_type_long_long" = yes; then { { $as_echo "$as_me:${as_lineno-$LINENO}: error: in \`$ac_pwd':" >&5 $as_echo "$as_me: error: in \`$ac_pwd':" >&2;} as_fn_error 77 "cannot compute sizeof (long long) See \`config.log' for more details" "$LINENO" 5; } else ac_cv_sizeof_long_long=0 fi fi fi { $as_echo "$as_me:${as_lineno-$LINENO}: result: $ac_cv_sizeof_long_long" >&5 $as_echo "$ac_cv_sizeof_long_long" >&6; } cat >>confdefs.h <<_ACEOF #define SIZEOF_LONG_LONG $ac_cv_sizeof_long_long _ACEOF # The cast to long int works around a bug in the HP C Compiler # version HP92453-01 B.11.11.23709.GP, which incorrectly rejects # declarations like `int a3[[(sizeof (unsigned char)) >= 0]];'. # This bug is HP SR number 8606223364. { $as_echo "$as_me:${as_lineno-$LINENO}: checking size of unsigned long long" >&5 $as_echo_n "checking size of unsigned long long... " >&6; } if ${ac_cv_sizeof_unsigned_long_long+:} false; then : $as_echo_n "(cached) " >&6 else if ac_fn_c_compute_int "$LINENO" "(long int) (sizeof (unsigned long long))" "ac_cv_sizeof_unsigned_long_long" "$ac_includes_default"; then : else if test "$ac_cv_type_unsigned_long_long" = yes; then { { $as_echo "$as_me:${as_lineno-$LINENO}: error: in \`$ac_pwd':" >&5 $as_echo "$as_me: error: in \`$ac_pwd':" >&2;} as_fn_error 77 "cannot compute sizeof (unsigned long long) See \`config.log' for more details" "$LINENO" 5; } else ac_cv_sizeof_unsigned_long_long=0 fi fi fi { $as_echo "$as_me:${as_lineno-$LINENO}: result: $ac_cv_sizeof_unsigned_long_long" >&5 $as_echo "$ac_cv_sizeof_unsigned_long_long" >&6; } cat >>confdefs.h <<_ACEOF #define SIZEOF_UNSIGNED_LONG_LONG $ac_cv_sizeof_unsigned_long_long _ACEOF # The cast to long int works around a bug in the HP C Compiler # version HP92453-01 B.11.11.23709.GP, which incorrectly rejects # declarations like `int a3[[(sizeof (unsigned char)) >= 0]];'. # This bug is HP SR number 8606223364. { $as_echo "$as_me:${as_lineno-$LINENO}: checking size of size_t" >&5 $as_echo_n "checking size of size_t... " >&6; } if ${ac_cv_sizeof_size_t+:} false; then : $as_echo_n "(cached) " >&6 else if ac_fn_c_compute_int "$LINENO" "(long int) (sizeof (size_t))" "ac_cv_sizeof_size_t" "$ac_includes_default"; then : else if test "$ac_cv_type_size_t" = yes; then { { $as_echo "$as_me:${as_lineno-$LINENO}: error: in \`$ac_pwd':" >&5 $as_echo "$as_me: error: in \`$ac_pwd':" >&2;} as_fn_error 77 "cannot compute sizeof (size_t) See \`config.log' for more details" "$LINENO" 5; } else ac_cv_sizeof_size_t=0 fi fi fi { $as_echo "$as_me:${as_lineno-$LINENO}: result: $ac_cv_sizeof_size_t" >&5 $as_echo "$ac_cv_sizeof_size_t" >&6; } cat >>confdefs.h <<_ACEOF #define SIZEOF_SIZE_T $ac_cv_sizeof_size_t _ACEOF # The cast to long int works around a bug in the HP C Compiler # version HP92453-01 B.11.11.23709.GP, which incorrectly rejects # declarations like `int a3[[(sizeof (unsigned char)) >= 0]];'. # This bug is HP SR number 8606223364. { $as_echo "$as_me:${as_lineno-$LINENO}: checking size of ptrdiff_t" >&5 $as_echo_n "checking size of ptrdiff_t... " >&6; } if ${ac_cv_sizeof_ptrdiff_t+:} false; then : $as_echo_n "(cached) " >&6 else if ac_fn_c_compute_int "$LINENO" "(long int) (sizeof (ptrdiff_t))" "ac_cv_sizeof_ptrdiff_t" "$ac_includes_default"; then : else if test "$ac_cv_type_ptrdiff_t" = yes; then { { $as_echo "$as_me:${as_lineno-$LINENO}: error: in \`$ac_pwd':" >&5 $as_echo "$as_me: error: in \`$ac_pwd':" >&2;} as_fn_error 77 "cannot compute sizeof (ptrdiff_t) See \`config.log' for more details" "$LINENO" 5; } else ac_cv_sizeof_ptrdiff_t=0 fi fi fi { $as_echo "$as_me:${as_lineno-$LINENO}: result: $ac_cv_sizeof_ptrdiff_t" >&5 $as_echo "$ac_cv_sizeof_ptrdiff_t" >&6; } cat >>confdefs.h <<_ACEOF #define SIZEOF_PTRDIFF_T $ac_cv_sizeof_ptrdiff_t _ACEOF ac_fn_c_check_type "$LINENO" "uintptr_t" "ac_cv_type_uintptr_t" "$ac_includes_default #ifdef HAVE_STDINT_H # include #endif " if test "x$ac_cv_type_uintptr_t" = xyes; then : cat >>confdefs.h <<_ACEOF #define HAVE_UINTPTR_T 1 _ACEOF else # The cast to long int works around a bug in the HP C Compiler # version HP92453-01 B.11.11.23709.GP, which incorrectly rejects # declarations like `int a3[[(sizeof (unsigned char)) >= 0]];'. # This bug is HP SR number 8606223364. { $as_echo "$as_me:${as_lineno-$LINENO}: checking size of void *" >&5 $as_echo_n "checking size of void *... " >&6; } if ${ac_cv_sizeof_void_p+:} false; then : $as_echo_n "(cached) " >&6 else if ac_fn_c_compute_int "$LINENO" "(long int) (sizeof (void *))" "ac_cv_sizeof_void_p" "$ac_includes_default"; then : else if test "$ac_cv_type_void_p" = yes; then { { $as_echo "$as_me:${as_lineno-$LINENO}: error: in \`$ac_pwd':" >&5 $as_echo "$as_me: error: in \`$ac_pwd':" >&2;} as_fn_error 77 "cannot compute sizeof (void *) See \`config.log' for more details" "$LINENO" 5; } else ac_cv_sizeof_void_p=0 fi fi fi { $as_echo "$as_me:${as_lineno-$LINENO}: result: $ac_cv_sizeof_void_p" >&5 $as_echo "$ac_cv_sizeof_void_p" >&6; } cat >>confdefs.h <<_ACEOF #define SIZEOF_VOID_P $ac_cv_sizeof_void_p _ACEOF fi # The cast to long int works around a bug in the HP C Compiler # version HP92453-01 B.11.11.23709.GP, which incorrectly rejects # declarations like `int a3[[(sizeof (unsigned char)) >= 0]];'. # This bug is HP SR number 8606223364. { $as_echo "$as_me:${as_lineno-$LINENO}: checking size of float" >&5 $as_echo_n "checking size of float... " >&6; } if ${ac_cv_sizeof_float+:} false; then : $as_echo_n "(cached) " >&6 else if ac_fn_c_compute_int "$LINENO" "(long int) (sizeof (float))" "ac_cv_sizeof_float" "$ac_includes_default"; then : else if test "$ac_cv_type_float" = yes; then { { $as_echo "$as_me:${as_lineno-$LINENO}: error: in \`$ac_pwd':" >&5 $as_echo "$as_me: error: in \`$ac_pwd':" >&2;} as_fn_error 77 "cannot compute sizeof (float) See \`config.log' for more details" "$LINENO" 5; } else ac_cv_sizeof_float=0 fi fi fi { $as_echo "$as_me:${as_lineno-$LINENO}: result: $ac_cv_sizeof_float" >&5 $as_echo "$ac_cv_sizeof_float" >&6; } cat >>confdefs.h <<_ACEOF #define SIZEOF_FLOAT $ac_cv_sizeof_float _ACEOF # The cast to long int works around a bug in the HP C Compiler # version HP92453-01 B.11.11.23709.GP, which incorrectly rejects # declarations like `int a3[[(sizeof (unsigned char)) >= 0]];'. # This bug is HP SR number 8606223364. { $as_echo "$as_me:${as_lineno-$LINENO}: checking size of double" >&5 $as_echo_n "checking size of double... " >&6; } if ${ac_cv_sizeof_double+:} false; then : $as_echo_n "(cached) " >&6 else if ac_fn_c_compute_int "$LINENO" "(long int) (sizeof (double))" "ac_cv_sizeof_double" "$ac_includes_default"; then : else if test "$ac_cv_type_double" = yes; then { { $as_echo "$as_me:${as_lineno-$LINENO}: error: in \`$ac_pwd':" >&5 $as_echo "$as_me: error: in \`$ac_pwd':" >&2;} as_fn_error 77 "cannot compute sizeof (double) See \`config.log' for more details" "$LINENO" 5; } else ac_cv_sizeof_double=0 fi fi fi { $as_echo "$as_me:${as_lineno-$LINENO}: result: $ac_cv_sizeof_double" >&5 $as_echo "$ac_cv_sizeof_double" >&6; } cat >>confdefs.h <<_ACEOF #define SIZEOF_DOUBLE $ac_cv_sizeof_double _ACEOF # The cast to long int works around a bug in the HP C Compiler # version HP92453-01 B.11.11.23709.GP, which incorrectly rejects # declarations like `int a3[[(sizeof (unsigned char)) >= 0]];'. # This bug is HP SR number 8606223364. { $as_echo "$as_me:${as_lineno-$LINENO}: checking size of fftw_r2r_kind" >&5 $as_echo_n "checking size of fftw_r2r_kind... " >&6; } if ${ac_cv_sizeof_fftw_r2r_kind+:} false; then : $as_echo_n "(cached) " >&6 else if ac_fn_c_compute_int "$LINENO" "(long int) (sizeof (fftw_r2r_kind))" "ac_cv_sizeof_fftw_r2r_kind" "typedef enum { FFTW_R2HC=0, FFTW_HC2R=1, FFTW_DHT=2, FFTW_REDFT00=3, FFTW_REDFT01=4, FFTW_REDFT10=5, FFTW_REDFT11=6, FFTW_RODFT00=7, FFTW_RODFT01=8, FFTW_RODFT10=9, FFTW_RODFT11=10 } fftw_r2r_kind; "; then : else if test "$ac_cv_type_fftw_r2r_kind" = yes; then { { $as_echo "$as_me:${as_lineno-$LINENO}: error: in \`$ac_pwd':" >&5 $as_echo "$as_me: error: in \`$ac_pwd':" >&2;} as_fn_error 77 "cannot compute sizeof (fftw_r2r_kind) See \`config.log' for more details" "$LINENO" 5; } else ac_cv_sizeof_fftw_r2r_kind=0 fi fi fi { $as_echo "$as_me:${as_lineno-$LINENO}: result: $ac_cv_sizeof_fftw_r2r_kind" >&5 $as_echo "$ac_cv_sizeof_fftw_r2r_kind" >&6; } cat >>confdefs.h <<_ACEOF #define SIZEOF_FFTW_R2R_KIND $ac_cv_sizeof_fftw_r2r_kind _ACEOF if test 0 = $ac_cv_sizeof_fftw_r2r_kind; then as_fn_error $? "sizeof(fftw_r2r_kind) test failed" "$LINENO" 5; fi C_FFTW_R2R_KIND=C_INT`expr $ac_cv_sizeof_fftw_r2r_kind \* 8`_T # The Ultrix 4.2 mips builtin alloca declared by alloca.h only works # for constant arguments. Useless! { $as_echo "$as_me:${as_lineno-$LINENO}: checking for working alloca.h" >&5 $as_echo_n "checking for working alloca.h... " >&6; } if ${ac_cv_working_alloca_h+:} false; then : $as_echo_n "(cached) " >&6 else cat confdefs.h - <<_ACEOF >conftest.$ac_ext /* end confdefs.h. */ #include int main () { char *p = (char *) alloca (2 * sizeof (int)); if (p) return 0; ; return 0; } _ACEOF if ac_fn_c_try_link "$LINENO"; then : ac_cv_working_alloca_h=yes else ac_cv_working_alloca_h=no fi rm -f core conftest.err conftest.$ac_objext \ conftest$ac_exeext conftest.$ac_ext fi { $as_echo "$as_me:${as_lineno-$LINENO}: result: $ac_cv_working_alloca_h" >&5 $as_echo "$ac_cv_working_alloca_h" >&6; } if test $ac_cv_working_alloca_h = yes; then $as_echo "#define HAVE_ALLOCA_H 1" >>confdefs.h fi { $as_echo "$as_me:${as_lineno-$LINENO}: checking for alloca" >&5 $as_echo_n "checking for alloca... " >&6; } if ${ac_cv_func_alloca_works+:} false; then : $as_echo_n "(cached) " >&6 else cat confdefs.h - <<_ACEOF >conftest.$ac_ext /* end confdefs.h. */ #ifdef __GNUC__ # define alloca __builtin_alloca #else # ifdef _MSC_VER # include # define alloca _alloca # else # ifdef HAVE_ALLOCA_H # include # else # ifdef _AIX #pragma alloca # else # ifndef alloca /* predefined by HP cc +Olibcalls */ void *alloca (size_t); # endif # endif # endif # endif #endif int main () { char *p = (char *) alloca (1); if (p) return 0; ; return 0; } _ACEOF if ac_fn_c_try_link "$LINENO"; then : ac_cv_func_alloca_works=yes else ac_cv_func_alloca_works=no fi rm -f core conftest.err conftest.$ac_objext \ conftest$ac_exeext conftest.$ac_ext fi { $as_echo "$as_me:${as_lineno-$LINENO}: result: $ac_cv_func_alloca_works" >&5 $as_echo "$ac_cv_func_alloca_works" >&6; } if test $ac_cv_func_alloca_works = yes; then $as_echo "#define HAVE_ALLOCA 1" >>confdefs.h else # The SVR3 libPW and SVR4 libucb both contain incompatible functions # that cause trouble. Some versions do not even contain alloca or # contain a buggy version. If you still want to use their alloca, # use ar to extract alloca.o from them instead of compiling alloca.c. ALLOCA=\${LIBOBJDIR}alloca.$ac_objext $as_echo "#define C_ALLOCA 1" >>confdefs.h { $as_echo "$as_me:${as_lineno-$LINENO}: checking whether \`alloca.c' needs Cray hooks" >&5 $as_echo_n "checking whether \`alloca.c' needs Cray hooks... " >&6; } if ${ac_cv_os_cray+:} false; then : $as_echo_n "(cached) " >&6 else cat confdefs.h - <<_ACEOF >conftest.$ac_ext /* end confdefs.h. */ #if defined CRAY && ! defined CRAY2 webecray #else wenotbecray #endif _ACEOF if (eval "$ac_cpp conftest.$ac_ext") 2>&5 | $EGREP "webecray" >/dev/null 2>&1; then : ac_cv_os_cray=yes else ac_cv_os_cray=no fi rm -f conftest* fi { $as_echo "$as_me:${as_lineno-$LINENO}: result: $ac_cv_os_cray" >&5 $as_echo "$ac_cv_os_cray" >&6; } if test $ac_cv_os_cray = yes; then for ac_func in _getb67 GETB67 getb67; do as_ac_var=`$as_echo "ac_cv_func_$ac_func" | $as_tr_sh` ac_fn_c_check_func "$LINENO" "$ac_func" "$as_ac_var" if eval test \"x\$"$as_ac_var"\" = x"yes"; then : cat >>confdefs.h <<_ACEOF #define CRAY_STACKSEG_END $ac_func _ACEOF break fi done fi { $as_echo "$as_me:${as_lineno-$LINENO}: checking stack direction for C alloca" >&5 $as_echo_n "checking stack direction for C alloca... " >&6; } if ${ac_cv_c_stack_direction+:} false; then : $as_echo_n "(cached) " >&6 else if test "$cross_compiling" = yes; then : ac_cv_c_stack_direction=0 else cat confdefs.h - <<_ACEOF >conftest.$ac_ext /* end confdefs.h. */ $ac_includes_default int find_stack_direction (int *addr, int depth) { int dir, dummy = 0; if (! addr) addr = &dummy; *addr = addr < &dummy ? 1 : addr == &dummy ? 0 : -1; dir = depth ? find_stack_direction (addr, depth - 1) : 0; return dir + dummy; } int main (int argc, char **argv) { return find_stack_direction (0, argc + !argv + 20) < 0; } _ACEOF if ac_fn_c_try_run "$LINENO"; then : ac_cv_c_stack_direction=1 else ac_cv_c_stack_direction=-1 fi rm -f core *.core core.conftest.* gmon.out bb.out conftest$ac_exeext \ conftest.$ac_objext conftest.beam conftest.$ac_ext fi fi { $as_echo "$as_me:${as_lineno-$LINENO}: result: $ac_cv_c_stack_direction" >&5 $as_echo "$ac_cv_c_stack_direction" >&6; } cat >>confdefs.h <<_ACEOF #define STACK_DIRECTION $ac_cv_c_stack_direction _ACEOF fi { $as_echo "$as_me:${as_lineno-$LINENO}: checking for working strtod" >&5 $as_echo_n "checking for working strtod... " >&6; } if ${ac_cv_func_strtod+:} false; then : $as_echo_n "(cached) " >&6 else if test "$cross_compiling" = yes; then : ac_cv_func_strtod=no else cat confdefs.h - <<_ACEOF >conftest.$ac_ext /* end confdefs.h. */ $ac_includes_default #ifndef strtod double strtod (); #endif int main() { { /* Some versions of Linux strtod mis-parse strings with leading '+'. */ char *string = " +69"; char *term; double value; value = strtod (string, &term); if (value != 69 || term != (string + 4)) return 1; } { /* Under Solaris 2.4, strtod returns the wrong value for the terminating character under some conditions. */ char *string = "NaN"; char *term; strtod (string, &term); if (term != string && *(term - 1) == 0) return 1; } return 0; } _ACEOF if ac_fn_c_try_run "$LINENO"; then : ac_cv_func_strtod=yes else ac_cv_func_strtod=no fi rm -f core *.core core.conftest.* gmon.out bb.out conftest$ac_exeext \ conftest.$ac_objext conftest.beam conftest.$ac_ext fi fi { $as_echo "$as_me:${as_lineno-$LINENO}: result: $ac_cv_func_strtod" >&5 $as_echo "$ac_cv_func_strtod" >&6; } if test $ac_cv_func_strtod = no; then case " $LIBOBJS " in *" strtod.$ac_objext "* ) ;; *) LIBOBJS="$LIBOBJS strtod.$ac_objext" ;; esac ac_fn_c_check_func "$LINENO" "pow" "ac_cv_func_pow" if test "x$ac_cv_func_pow" = xyes; then : fi if test $ac_cv_func_pow = no; then { $as_echo "$as_me:${as_lineno-$LINENO}: checking for pow in -lm" >&5 $as_echo_n "checking for pow in -lm... " >&6; } if ${ac_cv_lib_m_pow+:} false; then : $as_echo_n "(cached) " >&6 else ac_check_lib_save_LIBS=$LIBS LIBS="-lm $LIBS" cat confdefs.h - <<_ACEOF >conftest.$ac_ext /* end confdefs.h. */ /* Override any GCC internal prototype to avoid an error. Use char because int might match the return type of a GCC builtin and then its argument prototype would still apply. */ #ifdef __cplusplus extern "C" #endif char pow (); int main () { return pow (); ; return 0; } _ACEOF if ac_fn_c_try_link "$LINENO"; then : ac_cv_lib_m_pow=yes else ac_cv_lib_m_pow=no fi rm -f core conftest.err conftest.$ac_objext \ conftest$ac_exeext conftest.$ac_ext LIBS=$ac_check_lib_save_LIBS fi { $as_echo "$as_me:${as_lineno-$LINENO}: result: $ac_cv_lib_m_pow" >&5 $as_echo "$ac_cv_lib_m_pow" >&6; } if test "x$ac_cv_lib_m_pow" = xyes; then : POW_LIB=-lm else { $as_echo "$as_me:${as_lineno-$LINENO}: WARNING: cannot find library containing definition of pow" >&5 $as_echo "$as_me: WARNING: cannot find library containing definition of pow" >&2;} fi fi fi for ac_func in vprintf do : ac_fn_c_check_func "$LINENO" "vprintf" "ac_cv_func_vprintf" if test "x$ac_cv_func_vprintf" = xyes; then : cat >>confdefs.h <<_ACEOF #define HAVE_VPRINTF 1 _ACEOF ac_fn_c_check_func "$LINENO" "_doprnt" "ac_cv_func__doprnt" if test "x$ac_cv_func__doprnt" = xyes; then : $as_echo "#define HAVE_DOPRNT 1" >>confdefs.h fi fi done { $as_echo "$as_me:${as_lineno-$LINENO}: checking for sin in -lm" >&5 $as_echo_n "checking for sin in -lm... " >&6; } if ${ac_cv_lib_m_sin+:} false; then : $as_echo_n "(cached) " >&6 else ac_check_lib_save_LIBS=$LIBS LIBS="-lm $LIBS" cat confdefs.h - <<_ACEOF >conftest.$ac_ext /* end confdefs.h. */ /* Override any GCC internal prototype to avoid an error. Use char because int might match the return type of a GCC builtin and then its argument prototype would still apply. */ #ifdef __cplusplus extern "C" #endif char sin (); int main () { return sin (); ; return 0; } _ACEOF if ac_fn_c_try_link "$LINENO"; then : ac_cv_lib_m_sin=yes else ac_cv_lib_m_sin=no fi rm -f core conftest.err conftest.$ac_objext \ conftest$ac_exeext conftest.$ac_ext LIBS=$ac_check_lib_save_LIBS fi { $as_echo "$as_me:${as_lineno-$LINENO}: result: $ac_cv_lib_m_sin" >&5 $as_echo "$ac_cv_lib_m_sin" >&6; } if test "x$ac_cv_lib_m_sin" = xyes; then : cat >>confdefs.h <<_ACEOF #define HAVE_LIBM 1 _ACEOF LIBS="-lm $LIBS" fi if test $PRECISION = q; then { $as_echo "$as_me:${as_lineno-$LINENO}: checking whether we are using gcc 4.6.0 or later" >&5 $as_echo_n "checking whether we are using gcc 4.6.0 or later... " >&6; } if ${ax_cv_gcc_4_6_0+:} false; then : $as_echo_n "(cached) " >&6 else ax_cv_gcc_4_6_0=no if test "$GCC" = "yes"; then cat confdefs.h - <<_ACEOF >conftest.$ac_ext /* end confdefs.h. */ #ifdef __GNUC__ # if (__GNUC__ > 4) || (__GNUC__ == 4 && __GNUC_MINOR__ > 6) \ || (__GNUC__ == 4 && __GNUC_MINOR__ == 6 && __GNUC_PATCHLEVEL__ >= 0) yes; # endif #endif _ACEOF if (eval "$ac_cpp conftest.$ac_ext") 2>&5 | $EGREP "yes" >/dev/null 2>&1; then : ax_cv_gcc_4_6_0=yes fi rm -f conftest* fi fi { $as_echo "$as_me:${as_lineno-$LINENO}: result: $ax_cv_gcc_4_6_0" >&5 $as_echo "$ax_cv_gcc_4_6_0" >&6; } if test "$ax_cv_gcc_4_6_0" = yes; then : else as_fn_error $? "gcc 4.6 or later required for quad precision support" "$LINENO" 5 fi { $as_echo "$as_me:${as_lineno-$LINENO}: checking for sinq in -lquadmath" >&5 $as_echo_n "checking for sinq in -lquadmath... " >&6; } if ${ac_cv_lib_quadmath_sinq+:} false; then : $as_echo_n "(cached) " >&6 else ac_check_lib_save_LIBS=$LIBS LIBS="-lquadmath $LIBS" cat confdefs.h - <<_ACEOF >conftest.$ac_ext /* end confdefs.h. */ /* Override any GCC internal prototype to avoid an error. Use char because int might match the return type of a GCC builtin and then its argument prototype would still apply. */ #ifdef __cplusplus extern "C" #endif char sinq (); int main () { return sinq (); ; return 0; } _ACEOF if ac_fn_c_try_link "$LINENO"; then : ac_cv_lib_quadmath_sinq=yes else ac_cv_lib_quadmath_sinq=no fi rm -f core conftest.err conftest.$ac_objext \ conftest$ac_exeext conftest.$ac_ext LIBS=$ac_check_lib_save_LIBS fi { $as_echo "$as_me:${as_lineno-$LINENO}: result: $ac_cv_lib_quadmath_sinq" >&5 $as_echo "$ac_cv_lib_quadmath_sinq" >&6; } if test "x$ac_cv_lib_quadmath_sinq" = xyes; then : cat >>confdefs.h <<_ACEOF #define HAVE_LIBQUADMATH 1 _ACEOF LIBS="-lquadmath $LIBS" else as_fn_error $? "quad precision requires libquadmath for quad-precision trigonometric routines" "$LINENO" 5 fi LIBQUADMATH=-lquadmath fi for ac_func in BSDgettimeofday gettimeofday gethrtime read_real_time time_base_to_time drand48 sqrt memset posix_memalign memalign _mm_malloc _mm_free clock_gettime mach_absolute_time sysctl abort sinl cosl snprintf do : as_ac_var=`$as_echo "ac_cv_func_$ac_func" | $as_tr_sh` ac_fn_c_check_func "$LINENO" "$ac_func" "$as_ac_var" if eval test \"x\$"$as_ac_var"\" = x"yes"; then : cat >>confdefs.h <<_ACEOF #define `$as_echo "HAVE_$ac_func" | $as_tr_cpp` 1 _ACEOF fi done ac_fn_c_check_decl "$LINENO" "drand48" "ac_cv_have_decl_drand48" "$ac_includes_default" if test "x$ac_cv_have_decl_drand48" = xyes; then : ac_have_decl=1 else ac_have_decl=0 fi cat >>confdefs.h <<_ACEOF #define HAVE_DECL_DRAND48 $ac_have_decl _ACEOF ac_fn_c_check_decl "$LINENO" "srand48" "ac_cv_have_decl_srand48" "$ac_includes_default" if test "x$ac_cv_have_decl_srand48" = xyes; then : ac_have_decl=1 else ac_have_decl=0 fi cat >>confdefs.h <<_ACEOF #define HAVE_DECL_SRAND48 $ac_have_decl _ACEOF ac_fn_c_check_decl "$LINENO" "memalign" "ac_cv_have_decl_memalign" "$ac_includes_default" if test "x$ac_cv_have_decl_memalign" = xyes; then : ac_have_decl=1 else ac_have_decl=0 fi cat >>confdefs.h <<_ACEOF #define HAVE_DECL_MEMALIGN $ac_have_decl _ACEOF ac_fn_c_check_decl "$LINENO" "posix_memalign" "ac_cv_have_decl_posix_memalign" "$ac_includes_default" if test "x$ac_cv_have_decl_posix_memalign" = xyes; then : ac_have_decl=1 else ac_have_decl=0 fi cat >>confdefs.h <<_ACEOF #define HAVE_DECL_POSIX_MEMALIGN $ac_have_decl _ACEOF ac_fn_c_check_decl "$LINENO" "sinl" "ac_cv_have_decl_sinl" "$ac_includes_default" if test "x$ac_cv_have_decl_sinl" = xyes; then : ac_have_decl=1 else ac_have_decl=0 fi cat >>confdefs.h <<_ACEOF #define HAVE_DECL_SINL $ac_have_decl _ACEOF ac_fn_c_check_decl "$LINENO" "cosl" "ac_cv_have_decl_cosl" "$ac_includes_default" if test "x$ac_cv_have_decl_cosl" = xyes; then : ac_have_decl=1 else ac_have_decl=0 fi cat >>confdefs.h <<_ACEOF #define HAVE_DECL_COSL $ac_have_decl _ACEOF ac_fn_c_check_decl "$LINENO" "sinq" "ac_cv_have_decl_sinq" "$ac_includes_default" if test "x$ac_cv_have_decl_sinq" = xyes; then : ac_have_decl=1 else ac_have_decl=0 fi cat >>confdefs.h <<_ACEOF #define HAVE_DECL_SINQ $ac_have_decl _ACEOF ac_fn_c_check_decl "$LINENO" "cosq" "ac_cv_have_decl_cosq" "$ac_includes_default" if test "x$ac_cv_have_decl_cosq" = xyes; then : ac_have_decl=1 else ac_have_decl=0 fi cat >>confdefs.h <<_ACEOF #define HAVE_DECL_COSQ $ac_have_decl _ACEOF { $as_echo "$as_me:${as_lineno-$LINENO}: checking for _rtc intrinsic" >&5 $as_echo_n "checking for _rtc intrinsic... " >&6; } rtc_ok=yes cat confdefs.h - <<_ACEOF >conftest.$ac_ext /* end confdefs.h. */ #ifdef HAVE_INTRINSICS_H #include #endif int main () { _rtc() ; return 0; } _ACEOF if ac_fn_c_try_link "$LINENO"; then : $as_echo "#define HAVE__RTC 1" >>confdefs.h else rtc_ok=no fi rm -f core conftest.err conftest.$ac_objext \ conftest$ac_exeext conftest.$ac_ext { $as_echo "$as_me:${as_lineno-$LINENO}: result: $rtc_ok" >&5 $as_echo "$rtc_ok" >&6; } if test "$PRECISION" = "l"; then for ac_func in cosl sinl tanl do : as_ac_var=`$as_echo "ac_cv_func_$ac_func" | $as_tr_sh` ac_fn_c_check_func "$LINENO" "$ac_func" "$as_ac_var" if eval test \"x\$"$as_ac_var"\" = x"yes"; then : cat >>confdefs.h <<_ACEOF #define `$as_echo "HAVE_$ac_func" | $as_tr_cpp` 1 _ACEOF else as_fn_error $? "long-double precision requires long-double trigonometric routines" "$LINENO" 5 fi done fi { $as_echo "$as_me:${as_lineno-$LINENO}: checking for isnan" >&5 $as_echo_n "checking for isnan... " >&6; } cat confdefs.h - <<_ACEOF >conftest.$ac_ext /* end confdefs.h. */ #include int main () { if (!isnan(3.14159)) isnan(2.7183); ; return 0; } _ACEOF if ac_fn_c_try_link "$LINENO"; then : ok=yes else ok=no fi rm -f core conftest.err conftest.$ac_objext \ conftest$ac_exeext conftest.$ac_ext if test "$ok" = "yes"; then $as_echo "#define HAVE_ISNAN 1" >>confdefs.h fi { $as_echo "$as_me:${as_lineno-$LINENO}: result: ${ok}" >&5 $as_echo "${ok}" >&6; } ax_gcc_aligns_stack=no if test "$GCC" = "yes"; then { $as_echo "$as_me:${as_lineno-$LINENO}: checking whether C compiler accepts -mpreferred-stack-boundary=4" >&5 $as_echo_n "checking whether C compiler accepts -mpreferred-stack-boundary=4... " >&6; } if ${ax_cv_c_flags__mpreferred_stack_boundary_4+:} false; then : $as_echo_n "(cached) " >&6 else ax_save_FLAGS=$CFLAGS CFLAGS="-mpreferred-stack-boundary=4" cat confdefs.h - <<_ACEOF >conftest.$ac_ext /* end confdefs.h. */ int main () { ; return 0; } _ACEOF if ac_fn_c_try_compile "$LINENO"; then : ax_cv_c_flags__mpreferred_stack_boundary_4=yes else ax_cv_c_flags__mpreferred_stack_boundary_4=no fi rm -f core conftest.err conftest.$ac_objext conftest.$ac_ext CFLAGS=$ax_save_FLAGS fi eval ax_check_compiler_flags=$ax_cv_c_flags__mpreferred_stack_boundary_4 { $as_echo "$as_me:${as_lineno-$LINENO}: result: $ax_check_compiler_flags" >&5 $as_echo "$ax_check_compiler_flags" >&6; } if test "x$ax_check_compiler_flags" = xyes; then { $as_echo "$as_me:${as_lineno-$LINENO}: checking whether the stack is at least 8-byte aligned by gcc" >&5 $as_echo_n "checking whether the stack is at least 8-byte aligned by gcc... " >&6; } save_CFLAGS="$CFLAGS" CFLAGS="-O" { $as_echo "$as_me:${as_lineno-$LINENO}: checking whether C compiler accepts -malign-double" >&5 $as_echo_n "checking whether C compiler accepts -malign-double... " >&6; } if ${ax_cv_c_flags__malign_double+:} false; then : $as_echo_n "(cached) " >&6 else ax_save_FLAGS=$CFLAGS CFLAGS="-malign-double" cat confdefs.h - <<_ACEOF >conftest.$ac_ext /* end confdefs.h. */ int main () { ; return 0; } _ACEOF if ac_fn_c_try_compile "$LINENO"; then : ax_cv_c_flags__malign_double=yes else ax_cv_c_flags__malign_double=no fi rm -f core conftest.err conftest.$ac_objext conftest.$ac_ext CFLAGS=$ax_save_FLAGS fi eval ax_check_compiler_flags=$ax_cv_c_flags__malign_double { $as_echo "$as_me:${as_lineno-$LINENO}: result: $ax_check_compiler_flags" >&5 $as_echo "$ax_check_compiler_flags" >&6; } if test "x$ax_check_compiler_flags" = xyes; then CFLAGS="$CFLAGS -malign-double" else : fi if test "$cross_compiling" = yes; then : { $as_echo "$as_me:${as_lineno-$LINENO}: checking whether we are using gcc 3.0.0 or later" >&5 $as_echo_n "checking whether we are using gcc 3.0.0 or later... " >&6; } if ${ax_cv_gcc_3_0_0+:} false; then : $as_echo_n "(cached) " >&6 else ax_cv_gcc_3_0_0=no if test "$GCC" = "yes"; then cat confdefs.h - <<_ACEOF >conftest.$ac_ext /* end confdefs.h. */ #ifdef __GNUC__ # if (__GNUC__ > 3) || (__GNUC__ == 3 && __GNUC_MINOR__ > 0) \ || (__GNUC__ == 3 && __GNUC_MINOR__ == 0 && __GNUC_PATCHLEVEL__ >= 0) yes; # endif #endif _ACEOF if (eval "$ac_cpp conftest.$ac_ext") 2>&5 | $EGREP "yes" >/dev/null 2>&1; then : ax_cv_gcc_3_0_0=yes fi rm -f conftest* fi fi { $as_echo "$as_me:${as_lineno-$LINENO}: result: $ax_cv_gcc_3_0_0" >&5 $as_echo "$ax_cv_gcc_3_0_0" >&6; } if test "$ax_cv_gcc_3_0_0" = yes; then ax_gcc_stack_align_bug=no else ax_gcc_stack_align_bug=yes fi else cat confdefs.h - <<_ACEOF >conftest.$ac_ext /* end confdefs.h. */ #include # include struct yuck { int blechh; }; int one(void) { return 1; } struct yuck ick(void) { struct yuck y; y.blechh = 3; return y; } # define CHK_ALIGN(x) if ((((long) &(x)) & 0x7)) { fprintf(stderr, "bad alignment of " #x "\n"); exit(1); } void blah(int foo) { double foobar; CHK_ALIGN(foobar); } int main2(void) {double ok1; struct yuck y; double ok2; CHK_ALIGN(ok1); CHK_ALIGN(ok2); y = ick(); blah(one()); return 0;} int main(void) { if ((((long) (__builtin_alloca(0))) & 0x7)) __builtin_alloca(4); return main2(); } _ACEOF if ac_fn_c_try_run "$LINENO"; then : ax_gcc_aligns_stack=yes; ax_gcc_stack_align_bug=no else ax_gcc_stack_align_bug=yes fi rm -f core *.core core.conftest.* gmon.out bb.out conftest$ac_exeext \ conftest.$ac_objext conftest.beam conftest.$ac_ext fi CFLAGS="$save_CFLAGS" { $as_echo "$as_me:${as_lineno-$LINENO}: result: $ax_gcc_aligns_stack" >&5 $as_echo "$ax_gcc_aligns_stack" >&6; } else : fi fi if test "$ax_gcc_aligns_stack" = yes; then : else : fi if test "${enable_debug}" = "yes"; then CFLAGS="-g" fi if test "$enable_debug" = yes || test "$USE_MAINTAINER_MODE" = yes; then if test "$ac_test_CFLAGS" != "set"; then if test $ac_cv_c_compiler_gnu = yes; then CFLAGS="$CFLAGS -Wall -W -Wcast-qual -Wpointer-arith -Wcast-align -pedantic -Wno-long-long -Wshadow -Wbad-function-cast -Wwrite-strings -Wstrict-prototypes -Wredundant-decls -Wnested-externs" # -Wundef -Wconversion -Wmissing-prototypes -Wmissing-declarations fi fi fi # Check whether --enable-fortran was given. if test "${enable_fortran+set}" = set; then : enableval=$enable_fortran; enable_fortran=$enableval else enable_fortran=yes fi if test "$enable_fortran" = "yes"; then ac_ext=f ac_compile='$F77 -c $FFLAGS conftest.$ac_ext >&5' ac_link='$F77 -o conftest$ac_exeext $FFLAGS $LDFLAGS conftest.$ac_ext $LIBS >&5' ac_compiler_gnu=$ac_cv_f77_compiler_gnu if test -n "$ac_tool_prefix"; then for ac_prog in g77 xlf f77 frt pgf77 cf77 fort77 fl32 af77 xlf90 f90 pgf90 pghpf epcf90 gfortran g95 xlf95 f95 fort ifort ifc efc pgfortran pgf95 lf95 ftn nagfor do # Extract the first word of "$ac_tool_prefix$ac_prog", so it can be a program name with args. set dummy $ac_tool_prefix$ac_prog; ac_word=$2 { $as_echo "$as_me:${as_lineno-$LINENO}: checking for $ac_word" >&5 $as_echo_n "checking for $ac_word... " >&6; } if ${ac_cv_prog_F77+:} false; then : $as_echo_n "(cached) " >&6 else if test -n "$F77"; then ac_cv_prog_F77="$F77" # Let the user override the test. else as_save_IFS=$IFS; IFS=$PATH_SEPARATOR for as_dir in $PATH do IFS=$as_save_IFS test -z "$as_dir" && as_dir=. for ac_exec_ext in '' $ac_executable_extensions; do if as_fn_executable_p "$as_dir/$ac_word$ac_exec_ext"; then ac_cv_prog_F77="$ac_tool_prefix$ac_prog" $as_echo "$as_me:${as_lineno-$LINENO}: found $as_dir/$ac_word$ac_exec_ext" >&5 break 2 fi done done IFS=$as_save_IFS fi fi F77=$ac_cv_prog_F77 if test -n "$F77"; then { $as_echo "$as_me:${as_lineno-$LINENO}: result: $F77" >&5 $as_echo "$F77" >&6; } else { $as_echo "$as_me:${as_lineno-$LINENO}: result: no" >&5 $as_echo "no" >&6; } fi test -n "$F77" && break done fi if test -z "$F77"; then ac_ct_F77=$F77 for ac_prog in g77 xlf f77 frt pgf77 cf77 fort77 fl32 af77 xlf90 f90 pgf90 pghpf epcf90 gfortran g95 xlf95 f95 fort ifort ifc efc pgfortran pgf95 lf95 ftn nagfor do # Extract the first word of "$ac_prog", so it can be a program name with args. set dummy $ac_prog; ac_word=$2 { $as_echo "$as_me:${as_lineno-$LINENO}: checking for $ac_word" >&5 $as_echo_n "checking for $ac_word... " >&6; } if ${ac_cv_prog_ac_ct_F77+:} false; then : $as_echo_n "(cached) " >&6 else if test -n "$ac_ct_F77"; then ac_cv_prog_ac_ct_F77="$ac_ct_F77" # Let the user override the test. else as_save_IFS=$IFS; IFS=$PATH_SEPARATOR for as_dir in $PATH do IFS=$as_save_IFS test -z "$as_dir" && as_dir=. for ac_exec_ext in '' $ac_executable_extensions; do if as_fn_executable_p "$as_dir/$ac_word$ac_exec_ext"; then ac_cv_prog_ac_ct_F77="$ac_prog" $as_echo "$as_me:${as_lineno-$LINENO}: found $as_dir/$ac_word$ac_exec_ext" >&5 break 2 fi done done IFS=$as_save_IFS fi fi ac_ct_F77=$ac_cv_prog_ac_ct_F77 if test -n "$ac_ct_F77"; then { $as_echo "$as_me:${as_lineno-$LINENO}: result: $ac_ct_F77" >&5 $as_echo "$ac_ct_F77" >&6; } else { $as_echo "$as_me:${as_lineno-$LINENO}: result: no" >&5 $as_echo "no" >&6; } fi test -n "$ac_ct_F77" && break done if test "x$ac_ct_F77" = x; then F77="" else case $cross_compiling:$ac_tool_warned in yes:) { $as_echo "$as_me:${as_lineno-$LINENO}: WARNING: using cross tools not prefixed with host triplet" >&5 $as_echo "$as_me: WARNING: using cross tools not prefixed with host triplet" >&2;} ac_tool_warned=yes ;; esac F77=$ac_ct_F77 fi fi # Provide some information about the compiler. $as_echo "$as_me:${as_lineno-$LINENO}: checking for Fortran 77 compiler version" >&5 set X $ac_compile ac_compiler=$2 for ac_option in --version -v -V -qversion; do { { ac_try="$ac_compiler $ac_option >&5" case "(($ac_try" in *\"* | *\`* | *\\*) ac_try_echo=\$ac_try;; *) ac_try_echo=$ac_try;; esac eval ac_try_echo="\"\$as_me:${as_lineno-$LINENO}: $ac_try_echo\"" $as_echo "$ac_try_echo"; } >&5 (eval "$ac_compiler $ac_option >&5") 2>conftest.err ac_status=$? if test -s conftest.err; then sed '10a\ ... rest of stderr output deleted ... 10q' conftest.err >conftest.er1 cat conftest.er1 >&5 fi rm -f conftest.er1 conftest.err $as_echo "$as_me:${as_lineno-$LINENO}: \$? = $ac_status" >&5 test $ac_status = 0; } done rm -f a.out # If we don't use `.F' as extension, the preprocessor is not run on the # input file. (Note that this only needs to work for GNU compilers.) ac_save_ext=$ac_ext ac_ext=F { $as_echo "$as_me:${as_lineno-$LINENO}: checking whether we are using the GNU Fortran 77 compiler" >&5 $as_echo_n "checking whether we are using the GNU Fortran 77 compiler... " >&6; } if ${ac_cv_f77_compiler_gnu+:} false; then : $as_echo_n "(cached) " >&6 else cat > conftest.$ac_ext <<_ACEOF program main #ifndef __GNUC__ choke me #endif end _ACEOF if ac_fn_f77_try_compile "$LINENO"; then : ac_compiler_gnu=yes else ac_compiler_gnu=no fi rm -f core conftest.err conftest.$ac_objext conftest.$ac_ext ac_cv_f77_compiler_gnu=$ac_compiler_gnu fi { $as_echo "$as_me:${as_lineno-$LINENO}: result: $ac_cv_f77_compiler_gnu" >&5 $as_echo "$ac_cv_f77_compiler_gnu" >&6; } ac_ext=$ac_save_ext ac_test_FFLAGS=${FFLAGS+set} ac_save_FFLAGS=$FFLAGS FFLAGS= { $as_echo "$as_me:${as_lineno-$LINENO}: checking whether $F77 accepts -g" >&5 $as_echo_n "checking whether $F77 accepts -g... " >&6; } if ${ac_cv_prog_f77_g+:} false; then : $as_echo_n "(cached) " >&6 else FFLAGS=-g cat > conftest.$ac_ext <<_ACEOF program main end _ACEOF if ac_fn_f77_try_compile "$LINENO"; then : ac_cv_prog_f77_g=yes else ac_cv_prog_f77_g=no fi rm -f core conftest.err conftest.$ac_objext conftest.$ac_ext fi { $as_echo "$as_me:${as_lineno-$LINENO}: result: $ac_cv_prog_f77_g" >&5 $as_echo "$ac_cv_prog_f77_g" >&6; } if test "$ac_test_FFLAGS" = set; then FFLAGS=$ac_save_FFLAGS elif test $ac_cv_prog_f77_g = yes; then if test "x$ac_cv_f77_compiler_gnu" = xyes; then FFLAGS="-g -O2" else FFLAGS="-g" fi else if test "x$ac_cv_f77_compiler_gnu" = xyes; then FFLAGS="-O2" else FFLAGS= fi fi if test $ac_compiler_gnu = yes; then G77=yes else G77= fi ac_ext=c ac_cpp='$CPP $CPPFLAGS' ac_compile='$CC -c $CFLAGS $CPPFLAGS conftest.$ac_ext >&5' ac_link='$CC -o conftest$ac_exeext $CFLAGS $CPPFLAGS $LDFLAGS conftest.$ac_ext $LIBS >&5' ac_compiler_gnu=$ac_cv_c_compiler_gnu ac_ext=f ac_compile='$F77 -c $FFLAGS conftest.$ac_ext >&5' ac_link='$F77 -o conftest$ac_exeext $FFLAGS $LDFLAGS conftest.$ac_ext $LIBS >&5' ac_compiler_gnu=$ac_cv_f77_compiler_gnu if test -z "$F77" || test "X$F77" = "Xno"; then _lt_disable_F77=yes fi archive_cmds_need_lc_F77=no allow_undefined_flag_F77= always_export_symbols_F77=no archive_expsym_cmds_F77= export_dynamic_flag_spec_F77= hardcode_direct_F77=no hardcode_direct_absolute_F77=no hardcode_libdir_flag_spec_F77= hardcode_libdir_separator_F77= hardcode_minus_L_F77=no hardcode_automatic_F77=no inherit_rpath_F77=no module_cmds_F77= module_expsym_cmds_F77= link_all_deplibs_F77=unknown old_archive_cmds_F77=$old_archive_cmds reload_flag_F77=$reload_flag reload_cmds_F77=$reload_cmds no_undefined_flag_F77= whole_archive_flag_spec_F77= enable_shared_with_static_runtimes_F77=no # Source file extension for f77 test sources. ac_ext=f # Object file extension for compiled f77 test sources. objext=o objext_F77=$objext # No sense in running all these tests if we already determined that # the F77 compiler isn't working. Some variables (like enable_shared) # are currently assumed to apply to all compilers on this platform, # and will be corrupted by setting them based on a non-working compiler. if test "$_lt_disable_F77" != yes; then # Code to be used in simple compile tests lt_simple_compile_test_code="\ subroutine t return end " # Code to be used in simple link tests lt_simple_link_test_code="\ program t end " # ltmain only uses $CC for tagged configurations so make sure $CC is set. # If no C compiler was specified, use CC. LTCC=${LTCC-"$CC"} # If no C compiler flags were specified, use CFLAGS. LTCFLAGS=${LTCFLAGS-"$CFLAGS"} # Allow CC to be a program name with arguments. compiler=$CC # save warnings/boilerplate of simple test code ac_outfile=conftest.$ac_objext echo "$lt_simple_compile_test_code" >conftest.$ac_ext eval "$ac_compile" 2>&1 >/dev/null | $SED '/^$/d; /^ *+/d' >conftest.err _lt_compiler_boilerplate=`cat conftest.err` $RM conftest* ac_outfile=conftest.$ac_objext echo "$lt_simple_link_test_code" >conftest.$ac_ext eval "$ac_link" 2>&1 >/dev/null | $SED '/^$/d; /^ *+/d' >conftest.err _lt_linker_boilerplate=`cat conftest.err` $RM -r conftest* # Allow CC to be a program name with arguments. lt_save_CC="$CC" lt_save_GCC=$GCC lt_save_CFLAGS=$CFLAGS CC=${F77-"f77"} CFLAGS=$FFLAGS compiler=$CC compiler_F77=$CC for cc_temp in $compiler""; do case $cc_temp in compile | *[\\/]compile | ccache | *[\\/]ccache ) ;; distcc | *[\\/]distcc | purify | *[\\/]purify ) ;; \-*) ;; *) break;; esac done cc_basename=`$ECHO "$cc_temp" | $SED "s%.*/%%; s%^$host_alias-%%"` GCC=$G77 if test -n "$compiler"; then { $as_echo "$as_me:${as_lineno-$LINENO}: checking if libtool supports shared libraries" >&5 $as_echo_n "checking if libtool supports shared libraries... " >&6; } { $as_echo "$as_me:${as_lineno-$LINENO}: result: $can_build_shared" >&5 $as_echo "$can_build_shared" >&6; } { $as_echo "$as_me:${as_lineno-$LINENO}: checking whether to build shared libraries" >&5 $as_echo_n "checking whether to build shared libraries... " >&6; } test "$can_build_shared" = "no" && enable_shared=no # On AIX, shared libraries and static libraries use the same namespace, and # are all built from PIC. case $host_os in aix3*) test "$enable_shared" = yes && enable_static=no if test -n "$RANLIB"; then archive_cmds="$archive_cmds~\$RANLIB \$lib" postinstall_cmds='$RANLIB $lib' fi ;; aix[4-9]*) if test "$host_cpu" != ia64 && test "$aix_use_runtimelinking" = no ; then test "$enable_shared" = yes && enable_static=no fi ;; esac { $as_echo "$as_me:${as_lineno-$LINENO}: result: $enable_shared" >&5 $as_echo "$enable_shared" >&6; } { $as_echo "$as_me:${as_lineno-$LINENO}: checking whether to build static libraries" >&5 $as_echo_n "checking whether to build static libraries... " >&6; } # Make sure either enable_shared or enable_static is yes. test "$enable_shared" = yes || enable_static=yes { $as_echo "$as_me:${as_lineno-$LINENO}: result: $enable_static" >&5 $as_echo "$enable_static" >&6; } GCC_F77="$G77" LD_F77="$LD" ## CAVEAT EMPTOR: ## There is no encapsulation within the following macros, do not change ## the running order or otherwise move them around unless you know exactly ## what you are doing... lt_prog_compiler_wl_F77= lt_prog_compiler_pic_F77= lt_prog_compiler_static_F77= if test "$GCC" = yes; then lt_prog_compiler_wl_F77='-Wl,' lt_prog_compiler_static_F77='-static' case $host_os in aix*) # All AIX code is PIC. if test "$host_cpu" = ia64; then # AIX 5 now supports IA64 processor lt_prog_compiler_static_F77='-Bstatic' fi ;; amigaos*) case $host_cpu in powerpc) # see comment about AmigaOS4 .so support lt_prog_compiler_pic_F77='-fPIC' ;; m68k) # FIXME: we need at least 68020 code to build shared libraries, but # adding the `-m68020' flag to GCC prevents building anything better, # like `-m68040'. lt_prog_compiler_pic_F77='-m68020 -resident32 -malways-restore-a4' ;; esac ;; beos* | irix5* | irix6* | nonstopux* | osf3* | osf4* | osf5*) # PIC is the default for these OSes. ;; mingw* | cygwin* | pw32* | os2* | cegcc*) # This hack is so that the source file can tell whether it is being # built for inclusion in a dll (and should export symbols for example). # Although the cygwin gcc ignores -fPIC, still need this for old-style # (--disable-auto-import) libraries lt_prog_compiler_pic_F77='-DDLL_EXPORT' ;; darwin* | rhapsody*) # PIC is the default on this platform # Common symbols not allowed in MH_DYLIB files lt_prog_compiler_pic_F77='-fno-common' ;; haiku*) # PIC is the default for Haiku. # The "-static" flag exists, but is broken. lt_prog_compiler_static_F77= ;; hpux*) # PIC is the default for 64-bit PA HP-UX, but not for 32-bit # PA HP-UX. On IA64 HP-UX, PIC is the default but the pic flag # sets the default TLS model and affects inlining. case $host_cpu in hppa*64*) # +Z the default ;; *) lt_prog_compiler_pic_F77='-fPIC' ;; esac ;; interix[3-9]*) # Interix 3.x gcc -fpic/-fPIC options generate broken code. # Instead, we relocate shared libraries at runtime. ;; msdosdjgpp*) # Just because we use GCC doesn't mean we suddenly get shared libraries # on systems that don't support them. lt_prog_compiler_can_build_shared_F77=no enable_shared=no ;; *nto* | *qnx*) # QNX uses GNU C++, but need to define -shared option too, otherwise # it will coredump. lt_prog_compiler_pic_F77='-fPIC -shared' ;; sysv4*MP*) if test -d /usr/nec; then lt_prog_compiler_pic_F77=-Kconform_pic fi ;; *) lt_prog_compiler_pic_F77='-fPIC' ;; esac case $cc_basename in nvcc*) # Cuda Compiler Driver 2.2 lt_prog_compiler_wl_F77='-Xlinker ' if test -n "$lt_prog_compiler_pic_F77"; then lt_prog_compiler_pic_F77="-Xcompiler $lt_prog_compiler_pic_F77" fi ;; esac else # PORTME Check for flag to pass linker flags through the system compiler. case $host_os in aix*) lt_prog_compiler_wl_F77='-Wl,' if test "$host_cpu" = ia64; then # AIX 5 now supports IA64 processor lt_prog_compiler_static_F77='-Bstatic' else lt_prog_compiler_static_F77='-bnso -bI:/lib/syscalls.exp' fi ;; mingw* | cygwin* | pw32* | os2* | cegcc*) # This hack is so that the source file can tell whether it is being # built for inclusion in a dll (and should export symbols for example). lt_prog_compiler_pic_F77='-DDLL_EXPORT' ;; hpux9* | hpux10* | hpux11*) lt_prog_compiler_wl_F77='-Wl,' # PIC is the default for IA64 HP-UX and 64-bit HP-UX, but # not for PA HP-UX. case $host_cpu in hppa*64*|ia64*) # +Z the default ;; *) lt_prog_compiler_pic_F77='+Z' ;; esac # Is there a better lt_prog_compiler_static that works with the bundled CC? lt_prog_compiler_static_F77='${wl}-a ${wl}archive' ;; irix5* | irix6* | nonstopux*) lt_prog_compiler_wl_F77='-Wl,' # PIC (with -KPIC) is the default. lt_prog_compiler_static_F77='-non_shared' ;; linux* | k*bsd*-gnu | kopensolaris*-gnu | gnu*) case $cc_basename in # old Intel for x86_64 which still supported -KPIC. ecc*) lt_prog_compiler_wl_F77='-Wl,' lt_prog_compiler_pic_F77='-KPIC' lt_prog_compiler_static_F77='-static' ;; # icc used to be incompatible with GCC. # ICC 10 doesn't accept -KPIC any more. icc* | ifort*) lt_prog_compiler_wl_F77='-Wl,' lt_prog_compiler_pic_F77='-fPIC' lt_prog_compiler_static_F77='-static' ;; # Lahey Fortran 8.1. lf95*) lt_prog_compiler_wl_F77='-Wl,' lt_prog_compiler_pic_F77='--shared' lt_prog_compiler_static_F77='--static' ;; nagfor*) # NAG Fortran compiler lt_prog_compiler_wl_F77='-Wl,-Wl,,' lt_prog_compiler_pic_F77='-PIC' lt_prog_compiler_static_F77='-Bstatic' ;; pgcc* | pgf77* | pgf90* | pgf95* | pgfortran*) # Portland Group compilers (*not* the Pentium gcc compiler, # which looks to be a dead project) lt_prog_compiler_wl_F77='-Wl,' lt_prog_compiler_pic_F77='-fpic' lt_prog_compiler_static_F77='-Bstatic' ;; ccc*) lt_prog_compiler_wl_F77='-Wl,' # All Alpha code is PIC. lt_prog_compiler_static_F77='-non_shared' ;; xl* | bgxl* | bgf* | mpixl*) # IBM XL C 8.0/Fortran 10.1, 11.1 on PPC and BlueGene lt_prog_compiler_wl_F77='-Wl,' lt_prog_compiler_pic_F77='-qpic' lt_prog_compiler_static_F77='-qstaticlink' ;; *) case `$CC -V 2>&1 | sed 5q` in *Sun\ Ceres\ Fortran* | *Sun*Fortran*\ [1-7].* | *Sun*Fortran*\ 8.[0-3]*) # Sun Fortran 8.3 passes all unrecognized flags to the linker lt_prog_compiler_pic_F77='-KPIC' lt_prog_compiler_static_F77='-Bstatic' lt_prog_compiler_wl_F77='' ;; *Sun\ F* | *Sun*Fortran*) lt_prog_compiler_pic_F77='-KPIC' lt_prog_compiler_static_F77='-Bstatic' lt_prog_compiler_wl_F77='-Qoption ld ' ;; *Sun\ C*) # Sun C 5.9 lt_prog_compiler_pic_F77='-KPIC' lt_prog_compiler_static_F77='-Bstatic' lt_prog_compiler_wl_F77='-Wl,' ;; *Intel*\ [CF]*Compiler*) lt_prog_compiler_wl_F77='-Wl,' lt_prog_compiler_pic_F77='-fPIC' lt_prog_compiler_static_F77='-static' ;; *Portland\ Group*) lt_prog_compiler_wl_F77='-Wl,' lt_prog_compiler_pic_F77='-fpic' lt_prog_compiler_static_F77='-Bstatic' ;; esac ;; esac ;; newsos6) lt_prog_compiler_pic_F77='-KPIC' lt_prog_compiler_static_F77='-Bstatic' ;; *nto* | *qnx*) # QNX uses GNU C++, but need to define -shared option too, otherwise # it will coredump. lt_prog_compiler_pic_F77='-fPIC -shared' ;; osf3* | osf4* | osf5*) lt_prog_compiler_wl_F77='-Wl,' # All OSF/1 code is PIC. lt_prog_compiler_static_F77='-non_shared' ;; rdos*) lt_prog_compiler_static_F77='-non_shared' ;; solaris*) lt_prog_compiler_pic_F77='-KPIC' lt_prog_compiler_static_F77='-Bstatic' case $cc_basename in f77* | f90* | f95* | sunf77* | sunf90* | sunf95*) lt_prog_compiler_wl_F77='-Qoption ld ';; *) lt_prog_compiler_wl_F77='-Wl,';; esac ;; sunos4*) lt_prog_compiler_wl_F77='-Qoption ld ' lt_prog_compiler_pic_F77='-PIC' lt_prog_compiler_static_F77='-Bstatic' ;; sysv4 | sysv4.2uw2* | sysv4.3*) lt_prog_compiler_wl_F77='-Wl,' lt_prog_compiler_pic_F77='-KPIC' lt_prog_compiler_static_F77='-Bstatic' ;; sysv4*MP*) if test -d /usr/nec ;then lt_prog_compiler_pic_F77='-Kconform_pic' lt_prog_compiler_static_F77='-Bstatic' fi ;; sysv5* | unixware* | sco3.2v5* | sco5v6* | OpenUNIX*) lt_prog_compiler_wl_F77='-Wl,' lt_prog_compiler_pic_F77='-KPIC' lt_prog_compiler_static_F77='-Bstatic' ;; unicos*) lt_prog_compiler_wl_F77='-Wl,' lt_prog_compiler_can_build_shared_F77=no ;; uts4*) lt_prog_compiler_pic_F77='-pic' lt_prog_compiler_static_F77='-Bstatic' ;; *) lt_prog_compiler_can_build_shared_F77=no ;; esac fi case $host_os in # For platforms which do not support PIC, -DPIC is meaningless: *djgpp*) lt_prog_compiler_pic_F77= ;; *) lt_prog_compiler_pic_F77="$lt_prog_compiler_pic_F77" ;; esac { $as_echo "$as_me:${as_lineno-$LINENO}: checking for $compiler option to produce PIC" >&5 $as_echo_n "checking for $compiler option to produce PIC... " >&6; } if ${lt_cv_prog_compiler_pic_F77+:} false; then : $as_echo_n "(cached) " >&6 else lt_cv_prog_compiler_pic_F77=$lt_prog_compiler_pic_F77 fi { $as_echo "$as_me:${as_lineno-$LINENO}: result: $lt_cv_prog_compiler_pic_F77" >&5 $as_echo "$lt_cv_prog_compiler_pic_F77" >&6; } lt_prog_compiler_pic_F77=$lt_cv_prog_compiler_pic_F77 # # Check to make sure the PIC flag actually works. # if test -n "$lt_prog_compiler_pic_F77"; then { $as_echo "$as_me:${as_lineno-$LINENO}: checking if $compiler PIC flag $lt_prog_compiler_pic_F77 works" >&5 $as_echo_n "checking if $compiler PIC flag $lt_prog_compiler_pic_F77 works... " >&6; } if ${lt_cv_prog_compiler_pic_works_F77+:} false; then : $as_echo_n "(cached) " >&6 else lt_cv_prog_compiler_pic_works_F77=no ac_outfile=conftest.$ac_objext echo "$lt_simple_compile_test_code" > conftest.$ac_ext lt_compiler_flag="$lt_prog_compiler_pic_F77" # Insert the option either (1) after the last *FLAGS variable, or # (2) before a word containing "conftest.", or (3) at the end. # Note that $ac_compile itself does not contain backslashes and begins # with a dollar sign (not a hyphen), so the echo should work correctly. # The option is referenced via a variable to avoid confusing sed. lt_compile=`echo "$ac_compile" | $SED \ -e 's:.*FLAGS}\{0,1\} :&$lt_compiler_flag :; t' \ -e 's: [^ ]*conftest\.: $lt_compiler_flag&:; t' \ -e 's:$: $lt_compiler_flag:'` (eval echo "\"\$as_me:$LINENO: $lt_compile\"" >&5) (eval "$lt_compile" 2>conftest.err) ac_status=$? cat conftest.err >&5 echo "$as_me:$LINENO: \$? = $ac_status" >&5 if (exit $ac_status) && test -s "$ac_outfile"; then # The compiler can only warn and ignore the option if not recognized # So say no if there are warnings other than the usual output. $ECHO "$_lt_compiler_boilerplate" | $SED '/^$/d' >conftest.exp $SED '/^$/d; /^ *+/d' conftest.err >conftest.er2 if test ! -s conftest.er2 || diff conftest.exp conftest.er2 >/dev/null; then lt_cv_prog_compiler_pic_works_F77=yes fi fi $RM conftest* fi { $as_echo "$as_me:${as_lineno-$LINENO}: result: $lt_cv_prog_compiler_pic_works_F77" >&5 $as_echo "$lt_cv_prog_compiler_pic_works_F77" >&6; } if test x"$lt_cv_prog_compiler_pic_works_F77" = xyes; then case $lt_prog_compiler_pic_F77 in "" | " "*) ;; *) lt_prog_compiler_pic_F77=" $lt_prog_compiler_pic_F77" ;; esac else lt_prog_compiler_pic_F77= lt_prog_compiler_can_build_shared_F77=no fi fi # # Check to make sure the static flag actually works. # wl=$lt_prog_compiler_wl_F77 eval lt_tmp_static_flag=\"$lt_prog_compiler_static_F77\" { $as_echo "$as_me:${as_lineno-$LINENO}: checking if $compiler static flag $lt_tmp_static_flag works" >&5 $as_echo_n "checking if $compiler static flag $lt_tmp_static_flag works... " >&6; } if ${lt_cv_prog_compiler_static_works_F77+:} false; then : $as_echo_n "(cached) " >&6 else lt_cv_prog_compiler_static_works_F77=no save_LDFLAGS="$LDFLAGS" LDFLAGS="$LDFLAGS $lt_tmp_static_flag" echo "$lt_simple_link_test_code" > conftest.$ac_ext if (eval $ac_link 2>conftest.err) && test -s conftest$ac_exeext; then # The linker can only warn and ignore the option if not recognized # So say no if there are warnings if test -s conftest.err; then # Append any errors to the config.log. cat conftest.err 1>&5 $ECHO "$_lt_linker_boilerplate" | $SED '/^$/d' > conftest.exp $SED '/^$/d; /^ *+/d' conftest.err >conftest.er2 if diff conftest.exp conftest.er2 >/dev/null; then lt_cv_prog_compiler_static_works_F77=yes fi else lt_cv_prog_compiler_static_works_F77=yes fi fi $RM -r conftest* LDFLAGS="$save_LDFLAGS" fi { $as_echo "$as_me:${as_lineno-$LINENO}: result: $lt_cv_prog_compiler_static_works_F77" >&5 $as_echo "$lt_cv_prog_compiler_static_works_F77" >&6; } if test x"$lt_cv_prog_compiler_static_works_F77" = xyes; then : else lt_prog_compiler_static_F77= fi { $as_echo "$as_me:${as_lineno-$LINENO}: checking if $compiler supports -c -o file.$ac_objext" >&5 $as_echo_n "checking if $compiler supports -c -o file.$ac_objext... " >&6; } if ${lt_cv_prog_compiler_c_o_F77+:} false; then : $as_echo_n "(cached) " >&6 else lt_cv_prog_compiler_c_o_F77=no $RM -r conftest 2>/dev/null mkdir conftest cd conftest mkdir out echo "$lt_simple_compile_test_code" > conftest.$ac_ext lt_compiler_flag="-o out/conftest2.$ac_objext" # Insert the option either (1) after the last *FLAGS variable, or # (2) before a word containing "conftest.", or (3) at the end. # Note that $ac_compile itself does not contain backslashes and begins # with a dollar sign (not a hyphen), so the echo should work correctly. lt_compile=`echo "$ac_compile" | $SED \ -e 's:.*FLAGS}\{0,1\} :&$lt_compiler_flag :; t' \ -e 's: [^ ]*conftest\.: $lt_compiler_flag&:; t' \ -e 's:$: $lt_compiler_flag:'` (eval echo "\"\$as_me:$LINENO: $lt_compile\"" >&5) (eval "$lt_compile" 2>out/conftest.err) ac_status=$? cat out/conftest.err >&5 echo "$as_me:$LINENO: \$? = $ac_status" >&5 if (exit $ac_status) && test -s out/conftest2.$ac_objext then # The compiler can only warn and ignore the option if not recognized # So say no if there are warnings $ECHO "$_lt_compiler_boilerplate" | $SED '/^$/d' > out/conftest.exp $SED '/^$/d; /^ *+/d' out/conftest.err >out/conftest.er2 if test ! -s out/conftest.er2 || diff out/conftest.exp out/conftest.er2 >/dev/null; then lt_cv_prog_compiler_c_o_F77=yes fi fi chmod u+w . 2>&5 $RM conftest* # SGI C++ compiler will create directory out/ii_files/ for # template instantiation test -d out/ii_files && $RM out/ii_files/* && rmdir out/ii_files $RM out/* && rmdir out cd .. $RM -r conftest $RM conftest* fi { $as_echo "$as_me:${as_lineno-$LINENO}: result: $lt_cv_prog_compiler_c_o_F77" >&5 $as_echo "$lt_cv_prog_compiler_c_o_F77" >&6; } { $as_echo "$as_me:${as_lineno-$LINENO}: checking if $compiler supports -c -o file.$ac_objext" >&5 $as_echo_n "checking if $compiler supports -c -o file.$ac_objext... " >&6; } if ${lt_cv_prog_compiler_c_o_F77+:} false; then : $as_echo_n "(cached) " >&6 else lt_cv_prog_compiler_c_o_F77=no $RM -r conftest 2>/dev/null mkdir conftest cd conftest mkdir out echo "$lt_simple_compile_test_code" > conftest.$ac_ext lt_compiler_flag="-o out/conftest2.$ac_objext" # Insert the option either (1) after the last *FLAGS variable, or # (2) before a word containing "conftest.", or (3) at the end. # Note that $ac_compile itself does not contain backslashes and begins # with a dollar sign (not a hyphen), so the echo should work correctly. lt_compile=`echo "$ac_compile" | $SED \ -e 's:.*FLAGS}\{0,1\} :&$lt_compiler_flag :; t' \ -e 's: [^ ]*conftest\.: $lt_compiler_flag&:; t' \ -e 's:$: $lt_compiler_flag:'` (eval echo "\"\$as_me:$LINENO: $lt_compile\"" >&5) (eval "$lt_compile" 2>out/conftest.err) ac_status=$? cat out/conftest.err >&5 echo "$as_me:$LINENO: \$? = $ac_status" >&5 if (exit $ac_status) && test -s out/conftest2.$ac_objext then # The compiler can only warn and ignore the option if not recognized # So say no if there are warnings $ECHO "$_lt_compiler_boilerplate" | $SED '/^$/d' > out/conftest.exp $SED '/^$/d; /^ *+/d' out/conftest.err >out/conftest.er2 if test ! -s out/conftest.er2 || diff out/conftest.exp out/conftest.er2 >/dev/null; then lt_cv_prog_compiler_c_o_F77=yes fi fi chmod u+w . 2>&5 $RM conftest* # SGI C++ compiler will create directory out/ii_files/ for # template instantiation test -d out/ii_files && $RM out/ii_files/* && rmdir out/ii_files $RM out/* && rmdir out cd .. $RM -r conftest $RM conftest* fi { $as_echo "$as_me:${as_lineno-$LINENO}: result: $lt_cv_prog_compiler_c_o_F77" >&5 $as_echo "$lt_cv_prog_compiler_c_o_F77" >&6; } hard_links="nottested" if test "$lt_cv_prog_compiler_c_o_F77" = no && test "$need_locks" != no; then # do not overwrite the value of need_locks provided by the user { $as_echo "$as_me:${as_lineno-$LINENO}: checking if we can lock with hard links" >&5 $as_echo_n "checking if we can lock with hard links... " >&6; } hard_links=yes $RM conftest* ln conftest.a conftest.b 2>/dev/null && hard_links=no touch conftest.a ln conftest.a conftest.b 2>&5 || hard_links=no ln conftest.a conftest.b 2>/dev/null && hard_links=no { $as_echo "$as_me:${as_lineno-$LINENO}: result: $hard_links" >&5 $as_echo "$hard_links" >&6; } if test "$hard_links" = no; then { $as_echo "$as_me:${as_lineno-$LINENO}: WARNING: \`$CC' does not support \`-c -o', so \`make -j' may be unsafe" >&5 $as_echo "$as_me: WARNING: \`$CC' does not support \`-c -o', so \`make -j' may be unsafe" >&2;} need_locks=warn fi else need_locks=no fi { $as_echo "$as_me:${as_lineno-$LINENO}: checking whether the $compiler linker ($LD) supports shared libraries" >&5 $as_echo_n "checking whether the $compiler linker ($LD) supports shared libraries... " >&6; } runpath_var= allow_undefined_flag_F77= always_export_symbols_F77=no archive_cmds_F77= archive_expsym_cmds_F77= compiler_needs_object_F77=no enable_shared_with_static_runtimes_F77=no export_dynamic_flag_spec_F77= export_symbols_cmds_F77='$NM $libobjs $convenience | $global_symbol_pipe | $SED '\''s/.* //'\'' | sort | uniq > $export_symbols' hardcode_automatic_F77=no hardcode_direct_F77=no hardcode_direct_absolute_F77=no hardcode_libdir_flag_spec_F77= hardcode_libdir_separator_F77= hardcode_minus_L_F77=no hardcode_shlibpath_var_F77=unsupported inherit_rpath_F77=no link_all_deplibs_F77=unknown module_cmds_F77= module_expsym_cmds_F77= old_archive_from_new_cmds_F77= old_archive_from_expsyms_cmds_F77= thread_safe_flag_spec_F77= whole_archive_flag_spec_F77= # include_expsyms should be a list of space-separated symbols to be *always* # included in the symbol list include_expsyms_F77= # exclude_expsyms can be an extended regexp of symbols to exclude # it will be wrapped by ` (' and `)$', so one must not match beginning or # end of line. Example: `a|bc|.*d.*' will exclude the symbols `a' and `bc', # as well as any symbol that contains `d'. exclude_expsyms_F77='_GLOBAL_OFFSET_TABLE_|_GLOBAL__F[ID]_.*' # Although _GLOBAL_OFFSET_TABLE_ is a valid symbol C name, most a.out # platforms (ab)use it in PIC code, but their linkers get confused if # the symbol is explicitly referenced. Since portable code cannot # rely on this symbol name, it's probably fine to never include it in # preloaded symbol tables. # Exclude shared library initialization/finalization symbols. extract_expsyms_cmds= case $host_os in cygwin* | mingw* | pw32* | cegcc*) # FIXME: the MSVC++ port hasn't been tested in a loooong time # When not using gcc, we currently assume that we are using # Microsoft Visual C++. if test "$GCC" != yes; then with_gnu_ld=no fi ;; interix*) # we just hope/assume this is gcc and not c89 (= MSVC++) with_gnu_ld=yes ;; openbsd*) with_gnu_ld=no ;; linux* | k*bsd*-gnu | gnu*) link_all_deplibs_F77=no ;; esac ld_shlibs_F77=yes # On some targets, GNU ld is compatible enough with the native linker # that we're better off using the native interface for both. lt_use_gnu_ld_interface=no if test "$with_gnu_ld" = yes; then case $host_os in aix*) # The AIX port of GNU ld has always aspired to compatibility # with the native linker. However, as the warning in the GNU ld # block says, versions before 2.19.5* couldn't really create working # shared libraries, regardless of the interface used. case `$LD -v 2>&1` in *\ \(GNU\ Binutils\)\ 2.19.5*) ;; *\ \(GNU\ Binutils\)\ 2.[2-9]*) ;; *\ \(GNU\ Binutils\)\ [3-9]*) ;; *) lt_use_gnu_ld_interface=yes ;; esac ;; *) lt_use_gnu_ld_interface=yes ;; esac fi if test "$lt_use_gnu_ld_interface" = yes; then # If archive_cmds runs LD, not CC, wlarc should be empty wlarc='${wl}' # Set some defaults for GNU ld with shared library support. These # are reset later if shared libraries are not supported. Putting them # here allows them to be overridden if necessary. runpath_var=LD_RUN_PATH hardcode_libdir_flag_spec_F77='${wl}-rpath ${wl}$libdir' export_dynamic_flag_spec_F77='${wl}--export-dynamic' # ancient GNU ld didn't support --whole-archive et. al. if $LD --help 2>&1 | $GREP 'no-whole-archive' > /dev/null; then whole_archive_flag_spec_F77="$wlarc"'--whole-archive$convenience '"$wlarc"'--no-whole-archive' else whole_archive_flag_spec_F77= fi supports_anon_versioning=no case `$LD -v 2>&1` in *GNU\ gold*) supports_anon_versioning=yes ;; *\ [01].* | *\ 2.[0-9].* | *\ 2.10.*) ;; # catch versions < 2.11 *\ 2.11.93.0.2\ *) supports_anon_versioning=yes ;; # RH7.3 ... *\ 2.11.92.0.12\ *) supports_anon_versioning=yes ;; # Mandrake 8.2 ... *\ 2.11.*) ;; # other 2.11 versions *) supports_anon_versioning=yes ;; esac # See if GNU ld supports shared libraries. case $host_os in aix[3-9]*) # On AIX/PPC, the GNU linker is very broken if test "$host_cpu" != ia64; then ld_shlibs_F77=no cat <<_LT_EOF 1>&2 *** Warning: the GNU linker, at least up to release 2.19, is reported *** to be unable to reliably create shared libraries on AIX. *** Therefore, libtool is disabling shared libraries support. If you *** really care for shared libraries, you may want to install binutils *** 2.20 or above, or modify your PATH so that a non-GNU linker is found. *** You will then need to restart the configuration process. _LT_EOF fi ;; amigaos*) case $host_cpu in powerpc) # see comment about AmigaOS4 .so support archive_cmds_F77='$CC -shared $libobjs $deplibs $compiler_flags ${wl}-soname $wl$soname -o $lib' archive_expsym_cmds_F77='' ;; m68k) archive_cmds_F77='$RM $output_objdir/a2ixlibrary.data~$ECHO "#define NAME $libname" > $output_objdir/a2ixlibrary.data~$ECHO "#define LIBRARY_ID 1" >> $output_objdir/a2ixlibrary.data~$ECHO "#define VERSION $major" >> $output_objdir/a2ixlibrary.data~$ECHO "#define REVISION $revision" >> $output_objdir/a2ixlibrary.data~$AR $AR_FLAGS $lib $libobjs~$RANLIB $lib~(cd $output_objdir && a2ixlibrary -32)' hardcode_libdir_flag_spec_F77='-L$libdir' hardcode_minus_L_F77=yes ;; esac ;; beos*) if $LD --help 2>&1 | $GREP ': supported targets:.* elf' > /dev/null; then allow_undefined_flag_F77=unsupported # Joseph Beckenbach says some releases of gcc # support --undefined. This deserves some investigation. FIXME archive_cmds_F77='$CC -nostart $libobjs $deplibs $compiler_flags ${wl}-soname $wl$soname -o $lib' else ld_shlibs_F77=no fi ;; cygwin* | mingw* | pw32* | cegcc*) # _LT_TAGVAR(hardcode_libdir_flag_spec, F77) is actually meaningless, # as there is no search path for DLLs. hardcode_libdir_flag_spec_F77='-L$libdir' export_dynamic_flag_spec_F77='${wl}--export-all-symbols' allow_undefined_flag_F77=unsupported always_export_symbols_F77=no enable_shared_with_static_runtimes_F77=yes export_symbols_cmds_F77='$NM $libobjs $convenience | $global_symbol_pipe | $SED -e '\''/^[BCDGRS][ ]/s/.*[ ]\([^ ]*\)/\1 DATA/;s/^.*[ ]__nm__\([^ ]*\)[ ][^ ]*/\1 DATA/;/^I[ ]/d;/^[AITW][ ]/s/.* //'\'' | sort | uniq > $export_symbols' exclude_expsyms_F77='[_]+GLOBAL_OFFSET_TABLE_|[_]+GLOBAL__[FID]_.*|[_]+head_[A-Za-z0-9_]+_dll|[A-Za-z0-9_]+_dll_iname' if $LD --help 2>&1 | $GREP 'auto-import' > /dev/null; then archive_cmds_F77='$CC -shared $libobjs $deplibs $compiler_flags -o $output_objdir/$soname ${wl}--enable-auto-image-base -Xlinker --out-implib -Xlinker $lib' # If the export-symbols file already is a .def file (1st line # is EXPORTS), use it as is; otherwise, prepend... archive_expsym_cmds_F77='if test "x`$SED 1q $export_symbols`" = xEXPORTS; then cp $export_symbols $output_objdir/$soname.def; else echo EXPORTS > $output_objdir/$soname.def; cat $export_symbols >> $output_objdir/$soname.def; fi~ $CC -shared $output_objdir/$soname.def $libobjs $deplibs $compiler_flags -o $output_objdir/$soname ${wl}--enable-auto-image-base -Xlinker --out-implib -Xlinker $lib' else ld_shlibs_F77=no fi ;; haiku*) archive_cmds_F77='$CC -shared $libobjs $deplibs $compiler_flags ${wl}-soname $wl$soname -o $lib' link_all_deplibs_F77=yes ;; interix[3-9]*) hardcode_direct_F77=no hardcode_shlibpath_var_F77=no hardcode_libdir_flag_spec_F77='${wl}-rpath,$libdir' export_dynamic_flag_spec_F77='${wl}-E' # Hack: On Interix 3.x, we cannot compile PIC because of a broken gcc. # Instead, shared libraries are loaded at an image base (0x10000000 by # default) and relocated if they conflict, which is a slow very memory # consuming and fragmenting process. To avoid this, we pick a random, # 256 KiB-aligned image base between 0x50000000 and 0x6FFC0000 at link # time. Moving up from 0x10000000 also allows more sbrk(2) space. archive_cmds_F77='$CC -shared $pic_flag $libobjs $deplibs $compiler_flags ${wl}-h,$soname ${wl}--image-base,`expr ${RANDOM-$$} % 4096 / 2 \* 262144 + 1342177280` -o $lib' archive_expsym_cmds_F77='sed "s,^,_," $export_symbols >$output_objdir/$soname.expsym~$CC -shared $pic_flag $libobjs $deplibs $compiler_flags ${wl}-h,$soname ${wl}--retain-symbols-file,$output_objdir/$soname.expsym ${wl}--image-base,`expr ${RANDOM-$$} % 4096 / 2 \* 262144 + 1342177280` -o $lib' ;; gnu* | linux* | tpf* | k*bsd*-gnu | kopensolaris*-gnu) tmp_diet=no if test "$host_os" = linux-dietlibc; then case $cc_basename in diet\ *) tmp_diet=yes;; # linux-dietlibc with static linking (!diet-dyn) esac fi if $LD --help 2>&1 | $EGREP ': supported targets:.* elf' > /dev/null \ && test "$tmp_diet" = no then tmp_addflag=' $pic_flag' tmp_sharedflag='-shared' case $cc_basename,$host_cpu in pgcc*) # Portland Group C compiler whole_archive_flag_spec_F77='${wl}--whole-archive`for conv in $convenience\"\"; do test -n \"$conv\" && new_convenience=\"$new_convenience,$conv\"; done; func_echo_all \"$new_convenience\"` ${wl}--no-whole-archive' tmp_addflag=' $pic_flag' ;; pgf77* | pgf90* | pgf95* | pgfortran*) # Portland Group f77 and f90 compilers whole_archive_flag_spec_F77='${wl}--whole-archive`for conv in $convenience\"\"; do test -n \"$conv\" && new_convenience=\"$new_convenience,$conv\"; done; func_echo_all \"$new_convenience\"` ${wl}--no-whole-archive' tmp_addflag=' $pic_flag -Mnomain' ;; ecc*,ia64* | icc*,ia64*) # Intel C compiler on ia64 tmp_addflag=' -i_dynamic' ;; efc*,ia64* | ifort*,ia64*) # Intel Fortran compiler on ia64 tmp_addflag=' -i_dynamic -nofor_main' ;; ifc* | ifort*) # Intel Fortran compiler tmp_addflag=' -nofor_main' ;; lf95*) # Lahey Fortran 8.1 whole_archive_flag_spec_F77= tmp_sharedflag='--shared' ;; xl[cC]* | bgxl[cC]* | mpixl[cC]*) # IBM XL C 8.0 on PPC (deal with xlf below) tmp_sharedflag='-qmkshrobj' tmp_addflag= ;; nvcc*) # Cuda Compiler Driver 2.2 whole_archive_flag_spec_F77='${wl}--whole-archive`for conv in $convenience\"\"; do test -n \"$conv\" && new_convenience=\"$new_convenience,$conv\"; done; func_echo_all \"$new_convenience\"` ${wl}--no-whole-archive' compiler_needs_object_F77=yes ;; esac case `$CC -V 2>&1 | sed 5q` in *Sun\ C*) # Sun C 5.9 whole_archive_flag_spec_F77='${wl}--whole-archive`new_convenience=; for conv in $convenience\"\"; do test -z \"$conv\" || new_convenience=\"$new_convenience,$conv\"; done; func_echo_all \"$new_convenience\"` ${wl}--no-whole-archive' compiler_needs_object_F77=yes tmp_sharedflag='-G' ;; *Sun\ F*) # Sun Fortran 8.3 tmp_sharedflag='-G' ;; esac archive_cmds_F77='$CC '"$tmp_sharedflag""$tmp_addflag"' $libobjs $deplibs $compiler_flags ${wl}-soname $wl$soname -o $lib' if test "x$supports_anon_versioning" = xyes; then archive_expsym_cmds_F77='echo "{ global:" > $output_objdir/$libname.ver~ cat $export_symbols | sed -e "s/\(.*\)/\1;/" >> $output_objdir/$libname.ver~ echo "local: *; };" >> $output_objdir/$libname.ver~ $CC '"$tmp_sharedflag""$tmp_addflag"' $libobjs $deplibs $compiler_flags ${wl}-soname $wl$soname ${wl}-version-script ${wl}$output_objdir/$libname.ver -o $lib' fi case $cc_basename in xlf* | bgf* | bgxlf* | mpixlf*) # IBM XL Fortran 10.1 on PPC cannot create shared libs itself whole_archive_flag_spec_F77='--whole-archive$convenience --no-whole-archive' hardcode_libdir_flag_spec_F77='${wl}-rpath ${wl}$libdir' archive_cmds_F77='$LD -shared $libobjs $deplibs $linker_flags -soname $soname -o $lib' if test "x$supports_anon_versioning" = xyes; then archive_expsym_cmds_F77='echo "{ global:" > $output_objdir/$libname.ver~ cat $export_symbols | sed -e "s/\(.*\)/\1;/" >> $output_objdir/$libname.ver~ echo "local: *; };" >> $output_objdir/$libname.ver~ $LD -shared $libobjs $deplibs $linker_flags -soname $soname -version-script $output_objdir/$libname.ver -o $lib' fi ;; esac else ld_shlibs_F77=no fi ;; netbsd* | netbsdelf*-gnu) if echo __ELF__ | $CC -E - | $GREP __ELF__ >/dev/null; then archive_cmds_F77='$LD -Bshareable $libobjs $deplibs $linker_flags -o $lib' wlarc= else archive_cmds_F77='$CC -shared $pic_flag $libobjs $deplibs $compiler_flags ${wl}-soname $wl$soname -o $lib' archive_expsym_cmds_F77='$CC -shared $pic_flag $libobjs $deplibs $compiler_flags ${wl}-soname $wl$soname ${wl}-retain-symbols-file $wl$export_symbols -o $lib' fi ;; solaris*) if $LD -v 2>&1 | $GREP 'BFD 2\.8' > /dev/null; then ld_shlibs_F77=no cat <<_LT_EOF 1>&2 *** Warning: The releases 2.8.* of the GNU linker cannot reliably *** create shared libraries on Solaris systems. Therefore, libtool *** is disabling shared libraries support. We urge you to upgrade GNU *** binutils to release 2.9.1 or newer. Another option is to modify *** your PATH or compiler configuration so that the native linker is *** used, and then restart. _LT_EOF elif $LD --help 2>&1 | $GREP ': supported targets:.* elf' > /dev/null; then archive_cmds_F77='$CC -shared $pic_flag $libobjs $deplibs $compiler_flags ${wl}-soname $wl$soname -o $lib' archive_expsym_cmds_F77='$CC -shared $pic_flag $libobjs $deplibs $compiler_flags ${wl}-soname $wl$soname ${wl}-retain-symbols-file $wl$export_symbols -o $lib' else ld_shlibs_F77=no fi ;; sysv5* | sco3.2v5* | sco5v6* | unixware* | OpenUNIX*) case `$LD -v 2>&1` in *\ [01].* | *\ 2.[0-9].* | *\ 2.1[0-5].*) ld_shlibs_F77=no cat <<_LT_EOF 1>&2 *** Warning: Releases of the GNU linker prior to 2.16.91.0.3 can not *** reliably create shared libraries on SCO systems. Therefore, libtool *** is disabling shared libraries support. We urge you to upgrade GNU *** binutils to release 2.16.91.0.3 or newer. Another option is to modify *** your PATH or compiler configuration so that the native linker is *** used, and then restart. _LT_EOF ;; *) # For security reasons, it is highly recommended that you always # use absolute paths for naming shared libraries, and exclude the # DT_RUNPATH tag from executables and libraries. But doing so # requires that you compile everything twice, which is a pain. if $LD --help 2>&1 | $GREP ': supported targets:.* elf' > /dev/null; then hardcode_libdir_flag_spec_F77='${wl}-rpath ${wl}$libdir' archive_cmds_F77='$CC -shared $libobjs $deplibs $compiler_flags ${wl}-soname $wl$soname -o $lib' archive_expsym_cmds_F77='$CC -shared $libobjs $deplibs $compiler_flags ${wl}-soname $wl$soname ${wl}-retain-symbols-file $wl$export_symbols -o $lib' else ld_shlibs_F77=no fi ;; esac ;; sunos4*) archive_cmds_F77='$LD -assert pure-text -Bshareable -o $lib $libobjs $deplibs $linker_flags' wlarc= hardcode_direct_F77=yes hardcode_shlibpath_var_F77=no ;; *) if $LD --help 2>&1 | $GREP ': supported targets:.* elf' > /dev/null; then archive_cmds_F77='$CC -shared $pic_flag $libobjs $deplibs $compiler_flags ${wl}-soname $wl$soname -o $lib' archive_expsym_cmds_F77='$CC -shared $pic_flag $libobjs $deplibs $compiler_flags ${wl}-soname $wl$soname ${wl}-retain-symbols-file $wl$export_symbols -o $lib' else ld_shlibs_F77=no fi ;; esac if test "$ld_shlibs_F77" = no; then runpath_var= hardcode_libdir_flag_spec_F77= export_dynamic_flag_spec_F77= whole_archive_flag_spec_F77= fi else # PORTME fill in a description of your system's linker (not GNU ld) case $host_os in aix3*) allow_undefined_flag_F77=unsupported always_export_symbols_F77=yes archive_expsym_cmds_F77='$LD -o $output_objdir/$soname $libobjs $deplibs $linker_flags -bE:$export_symbols -T512 -H512 -bM:SRE~$AR $AR_FLAGS $lib $output_objdir/$soname' # Note: this linker hardcodes the directories in LIBPATH if there # are no directories specified by -L. hardcode_minus_L_F77=yes if test "$GCC" = yes && test -z "$lt_prog_compiler_static"; then # Neither direct hardcoding nor static linking is supported with a # broken collect2. hardcode_direct_F77=unsupported fi ;; aix[4-9]*) if test "$host_cpu" = ia64; then # On IA64, the linker does run time linking by default, so we don't # have to do anything special. aix_use_runtimelinking=no exp_sym_flag='-Bexport' no_entry_flag="" else # If we're using GNU nm, then we don't want the "-C" option. # -C means demangle to AIX nm, but means don't demangle with GNU nm # Also, AIX nm treats weak defined symbols like other global # defined symbols, whereas GNU nm marks them as "W". if $NM -V 2>&1 | $GREP 'GNU' > /dev/null; then export_symbols_cmds_F77='$NM -Bpg $libobjs $convenience | awk '\''{ if (((\$ 2 == "T") || (\$ 2 == "D") || (\$ 2 == "B") || (\$ 2 == "W")) && (substr(\$ 3,1,1) != ".")) { print \$ 3 } }'\'' | sort -u > $export_symbols' else export_symbols_cmds_F77='$NM -BCpg $libobjs $convenience | awk '\''{ if (((\$ 2 == "T") || (\$ 2 == "D") || (\$ 2 == "B")) && (substr(\$ 3,1,1) != ".")) { print \$ 3 } }'\'' | sort -u > $export_symbols' fi aix_use_runtimelinking=no # Test if we are trying to use run time linking or normal # AIX style linking. If -brtl is somewhere in LDFLAGS, we # need to do runtime linking. case $host_os in aix4.[23]|aix4.[23].*|aix[5-9]*) for ld_flag in $LDFLAGS; do if (test $ld_flag = "-brtl" || test $ld_flag = "-Wl,-brtl"); then aix_use_runtimelinking=yes break fi done ;; esac exp_sym_flag='-bexport' no_entry_flag='-bnoentry' fi # When large executables or shared objects are built, AIX ld can # have problems creating the table of contents. If linking a library # or program results in "error TOC overflow" add -mminimal-toc to # CXXFLAGS/CFLAGS for g++/gcc. In the cases where that is not # enough to fix the problem, add -Wl,-bbigtoc to LDFLAGS. archive_cmds_F77='' hardcode_direct_F77=yes hardcode_direct_absolute_F77=yes hardcode_libdir_separator_F77=':' link_all_deplibs_F77=yes file_list_spec_F77='${wl}-f,' if test "$GCC" = yes; then case $host_os in aix4.[012]|aix4.[012].*) # We only want to do this on AIX 4.2 and lower, the check # below for broken collect2 doesn't work under 4.3+ collect2name=`${CC} -print-prog-name=collect2` if test -f "$collect2name" && strings "$collect2name" | $GREP resolve_lib_name >/dev/null then # We have reworked collect2 : else # We have old collect2 hardcode_direct_F77=unsupported # It fails to find uninstalled libraries when the uninstalled # path is not listed in the libpath. Setting hardcode_minus_L # to unsupported forces relinking hardcode_minus_L_F77=yes hardcode_libdir_flag_spec_F77='-L$libdir' hardcode_libdir_separator_F77= fi ;; esac shared_flag='-shared' if test "$aix_use_runtimelinking" = yes; then shared_flag="$shared_flag "'${wl}-G' fi link_all_deplibs_F77=no else # not using gcc if test "$host_cpu" = ia64; then # VisualAge C++, Version 5.5 for AIX 5L for IA-64, Beta 3 Release # chokes on -Wl,-G. The following line is correct: shared_flag='-G' else if test "$aix_use_runtimelinking" = yes; then shared_flag='${wl}-G' else shared_flag='${wl}-bM:SRE' fi fi fi export_dynamic_flag_spec_F77='${wl}-bexpall' # It seems that -bexpall does not export symbols beginning with # underscore (_), so it is better to generate a list of symbols to export. always_export_symbols_F77=yes if test "$aix_use_runtimelinking" = yes; then # Warning - without using the other runtime loading flags (-brtl), # -berok will link without error, but may produce a broken library. allow_undefined_flag_F77='-berok' # Determine the default libpath from the value encoded in an # empty executable. if test "${lt_cv_aix_libpath+set}" = set; then aix_libpath=$lt_cv_aix_libpath else if ${lt_cv_aix_libpath__F77+:} false; then : $as_echo_n "(cached) " >&6 else cat > conftest.$ac_ext <<_ACEOF program main end _ACEOF if ac_fn_f77_try_link "$LINENO"; then : lt_aix_libpath_sed=' /Import File Strings/,/^$/ { /^0/ { s/^0 *\([^ ]*\) *$/\1/ p } }' lt_cv_aix_libpath__F77=`dump -H conftest$ac_exeext 2>/dev/null | $SED -n -e "$lt_aix_libpath_sed"` # Check for a 64-bit object if we didn't find anything. if test -z "$lt_cv_aix_libpath__F77"; then lt_cv_aix_libpath__F77=`dump -HX64 conftest$ac_exeext 2>/dev/null | $SED -n -e "$lt_aix_libpath_sed"` fi fi rm -f core conftest.err conftest.$ac_objext \ conftest$ac_exeext conftest.$ac_ext if test -z "$lt_cv_aix_libpath__F77"; then lt_cv_aix_libpath__F77="/usr/lib:/lib" fi fi aix_libpath=$lt_cv_aix_libpath__F77 fi hardcode_libdir_flag_spec_F77='${wl}-blibpath:$libdir:'"$aix_libpath" archive_expsym_cmds_F77='$CC -o $output_objdir/$soname $libobjs $deplibs '"\${wl}$no_entry_flag"' $compiler_flags `if test "x${allow_undefined_flag}" != "x"; then func_echo_all "${wl}${allow_undefined_flag}"; else :; fi` '"\${wl}$exp_sym_flag:\$export_symbols $shared_flag" else if test "$host_cpu" = ia64; then hardcode_libdir_flag_spec_F77='${wl}-R $libdir:/usr/lib:/lib' allow_undefined_flag_F77="-z nodefs" archive_expsym_cmds_F77="\$CC $shared_flag"' -o $output_objdir/$soname $libobjs $deplibs '"\${wl}$no_entry_flag"' $compiler_flags ${wl}${allow_undefined_flag} '"\${wl}$exp_sym_flag:\$export_symbols" else # Determine the default libpath from the value encoded in an # empty executable. if test "${lt_cv_aix_libpath+set}" = set; then aix_libpath=$lt_cv_aix_libpath else if ${lt_cv_aix_libpath__F77+:} false; then : $as_echo_n "(cached) " >&6 else cat > conftest.$ac_ext <<_ACEOF program main end _ACEOF if ac_fn_f77_try_link "$LINENO"; then : lt_aix_libpath_sed=' /Import File Strings/,/^$/ { /^0/ { s/^0 *\([^ ]*\) *$/\1/ p } }' lt_cv_aix_libpath__F77=`dump -H conftest$ac_exeext 2>/dev/null | $SED -n -e "$lt_aix_libpath_sed"` # Check for a 64-bit object if we didn't find anything. if test -z "$lt_cv_aix_libpath__F77"; then lt_cv_aix_libpath__F77=`dump -HX64 conftest$ac_exeext 2>/dev/null | $SED -n -e "$lt_aix_libpath_sed"` fi fi rm -f core conftest.err conftest.$ac_objext \ conftest$ac_exeext conftest.$ac_ext if test -z "$lt_cv_aix_libpath__F77"; then lt_cv_aix_libpath__F77="/usr/lib:/lib" fi fi aix_libpath=$lt_cv_aix_libpath__F77 fi hardcode_libdir_flag_spec_F77='${wl}-blibpath:$libdir:'"$aix_libpath" # Warning - without using the other run time loading flags, # -berok will link without error, but may produce a broken library. no_undefined_flag_F77=' ${wl}-bernotok' allow_undefined_flag_F77=' ${wl}-berok' if test "$with_gnu_ld" = yes; then # We only use this code for GNU lds that support --whole-archive. whole_archive_flag_spec_F77='${wl}--whole-archive$convenience ${wl}--no-whole-archive' else # Exported symbols can be pulled into shared objects from archives whole_archive_flag_spec_F77='$convenience' fi archive_cmds_need_lc_F77=yes # This is similar to how AIX traditionally builds its shared libraries. archive_expsym_cmds_F77="\$CC $shared_flag"' -o $output_objdir/$soname $libobjs $deplibs ${wl}-bnoentry $compiler_flags ${wl}-bE:$export_symbols${allow_undefined_flag}~$AR $AR_FLAGS $output_objdir/$libname$release.a $output_objdir/$soname' fi fi ;; amigaos*) case $host_cpu in powerpc) # see comment about AmigaOS4 .so support archive_cmds_F77='$CC -shared $libobjs $deplibs $compiler_flags ${wl}-soname $wl$soname -o $lib' archive_expsym_cmds_F77='' ;; m68k) archive_cmds_F77='$RM $output_objdir/a2ixlibrary.data~$ECHO "#define NAME $libname" > $output_objdir/a2ixlibrary.data~$ECHO "#define LIBRARY_ID 1" >> $output_objdir/a2ixlibrary.data~$ECHO "#define VERSION $major" >> $output_objdir/a2ixlibrary.data~$ECHO "#define REVISION $revision" >> $output_objdir/a2ixlibrary.data~$AR $AR_FLAGS $lib $libobjs~$RANLIB $lib~(cd $output_objdir && a2ixlibrary -32)' hardcode_libdir_flag_spec_F77='-L$libdir' hardcode_minus_L_F77=yes ;; esac ;; bsdi[45]*) export_dynamic_flag_spec_F77=-rdynamic ;; cygwin* | mingw* | pw32* | cegcc*) # When not using gcc, we currently assume that we are using # Microsoft Visual C++. # hardcode_libdir_flag_spec is actually meaningless, as there is # no search path for DLLs. case $cc_basename in cl*) # Native MSVC hardcode_libdir_flag_spec_F77=' ' allow_undefined_flag_F77=unsupported always_export_symbols_F77=yes file_list_spec_F77='@' # Tell ltmain to make .lib files, not .a files. libext=lib # Tell ltmain to make .dll files, not .so files. shrext_cmds=".dll" # FIXME: Setting linknames here is a bad hack. archive_cmds_F77='$CC -o $output_objdir/$soname $libobjs $compiler_flags $deplibs -Wl,-dll~linknames=' archive_expsym_cmds_F77='if test "x`$SED 1q $export_symbols`" = xEXPORTS; then sed -n -e 's/\\\\\\\(.*\\\\\\\)/-link\\\ -EXPORT:\\\\\\\1/' -e '1\\\!p' < $export_symbols > $output_objdir/$soname.exp; else sed -e 's/\\\\\\\(.*\\\\\\\)/-link\\\ -EXPORT:\\\\\\\1/' < $export_symbols > $output_objdir/$soname.exp; fi~ $CC -o $tool_output_objdir$soname $libobjs $compiler_flags $deplibs "@$tool_output_objdir$soname.exp" -Wl,-DLL,-IMPLIB:"$tool_output_objdir$libname.dll.lib"~ linknames=' # The linker will not automatically build a static lib if we build a DLL. # _LT_TAGVAR(old_archive_from_new_cmds, F77)='true' enable_shared_with_static_runtimes_F77=yes exclude_expsyms_F77='_NULL_IMPORT_DESCRIPTOR|_IMPORT_DESCRIPTOR_.*' export_symbols_cmds_F77='$NM $libobjs $convenience | $global_symbol_pipe | $SED -e '\''/^[BCDGRS][ ]/s/.*[ ]\([^ ]*\)/\1,DATA/'\'' | $SED -e '\''/^[AITW][ ]/s/.*[ ]//'\'' | sort | uniq > $export_symbols' # Don't use ranlib old_postinstall_cmds_F77='chmod 644 $oldlib' postlink_cmds_F77='lt_outputfile="@OUTPUT@"~ lt_tool_outputfile="@TOOL_OUTPUT@"~ case $lt_outputfile in *.exe|*.EXE) ;; *) lt_outputfile="$lt_outputfile.exe" lt_tool_outputfile="$lt_tool_outputfile.exe" ;; esac~ if test "$MANIFEST_TOOL" != ":" && test -f "$lt_outputfile.manifest"; then $MANIFEST_TOOL -manifest "$lt_tool_outputfile.manifest" -outputresource:"$lt_tool_outputfile" || exit 1; $RM "$lt_outputfile.manifest"; fi' ;; *) # Assume MSVC wrapper hardcode_libdir_flag_spec_F77=' ' allow_undefined_flag_F77=unsupported # Tell ltmain to make .lib files, not .a files. libext=lib # Tell ltmain to make .dll files, not .so files. shrext_cmds=".dll" # FIXME: Setting linknames here is a bad hack. archive_cmds_F77='$CC -o $lib $libobjs $compiler_flags `func_echo_all "$deplibs" | $SED '\''s/ -lc$//'\''` -link -dll~linknames=' # The linker will automatically build a .lib file if we build a DLL. old_archive_from_new_cmds_F77='true' # FIXME: Should let the user specify the lib program. old_archive_cmds_F77='lib -OUT:$oldlib$oldobjs$old_deplibs' enable_shared_with_static_runtimes_F77=yes ;; esac ;; darwin* | rhapsody*) archive_cmds_need_lc_F77=no hardcode_direct_F77=no hardcode_automatic_F77=yes hardcode_shlibpath_var_F77=unsupported if test "$lt_cv_ld_force_load" = "yes"; then whole_archive_flag_spec_F77='`for conv in $convenience\"\"; do test -n \"$conv\" && new_convenience=\"$new_convenience ${wl}-force_load,$conv\"; done; func_echo_all \"$new_convenience\"`' compiler_needs_object_F77=yes else whole_archive_flag_spec_F77='' fi link_all_deplibs_F77=yes allow_undefined_flag_F77="$_lt_dar_allow_undefined" case $cc_basename in ifort*) _lt_dar_can_shared=yes ;; *) _lt_dar_can_shared=$GCC ;; esac if test "$_lt_dar_can_shared" = "yes"; then output_verbose_link_cmd=func_echo_all archive_cmds_F77="\$CC -dynamiclib \$allow_undefined_flag -o \$lib \$libobjs \$deplibs \$compiler_flags -install_name \$rpath/\$soname \$verstring $_lt_dar_single_mod${_lt_dsymutil}" module_cmds_F77="\$CC \$allow_undefined_flag -o \$lib -bundle \$libobjs \$deplibs \$compiler_flags${_lt_dsymutil}" archive_expsym_cmds_F77="sed 's,^,_,' < \$export_symbols > \$output_objdir/\${libname}-symbols.expsym~\$CC -dynamiclib \$allow_undefined_flag -o \$lib \$libobjs \$deplibs \$compiler_flags -install_name \$rpath/\$soname \$verstring ${_lt_dar_single_mod}${_lt_dar_export_syms}${_lt_dsymutil}" module_expsym_cmds_F77="sed -e 's,^,_,' < \$export_symbols > \$output_objdir/\${libname}-symbols.expsym~\$CC \$allow_undefined_flag -o \$lib -bundle \$libobjs \$deplibs \$compiler_flags${_lt_dar_export_syms}${_lt_dsymutil}" else ld_shlibs_F77=no fi ;; dgux*) archive_cmds_F77='$LD -G -h $soname -o $lib $libobjs $deplibs $linker_flags' hardcode_libdir_flag_spec_F77='-L$libdir' hardcode_shlibpath_var_F77=no ;; # FreeBSD 2.2.[012] allows us to include c++rt0.o to get C++ constructor # support. Future versions do this automatically, but an explicit c++rt0.o # does not break anything, and helps significantly (at the cost of a little # extra space). freebsd2.2*) archive_cmds_F77='$LD -Bshareable -o $lib $libobjs $deplibs $linker_flags /usr/lib/c++rt0.o' hardcode_libdir_flag_spec_F77='-R$libdir' hardcode_direct_F77=yes hardcode_shlibpath_var_F77=no ;; # Unfortunately, older versions of FreeBSD 2 do not have this feature. freebsd2.*) archive_cmds_F77='$LD -Bshareable -o $lib $libobjs $deplibs $linker_flags' hardcode_direct_F77=yes hardcode_minus_L_F77=yes hardcode_shlibpath_var_F77=no ;; # FreeBSD 3 and greater uses gcc -shared to do shared libraries. freebsd* | dragonfly*) archive_cmds_F77='$CC -shared $pic_flag -o $lib $libobjs $deplibs $compiler_flags' hardcode_libdir_flag_spec_F77='-R$libdir' hardcode_direct_F77=yes hardcode_shlibpath_var_F77=no ;; hpux9*) if test "$GCC" = yes; then archive_cmds_F77='$RM $output_objdir/$soname~$CC -shared $pic_flag ${wl}+b ${wl}$install_libdir -o $output_objdir/$soname $libobjs $deplibs $compiler_flags~test $output_objdir/$soname = $lib || mv $output_objdir/$soname $lib' else archive_cmds_F77='$RM $output_objdir/$soname~$LD -b +b $install_libdir -o $output_objdir/$soname $libobjs $deplibs $linker_flags~test $output_objdir/$soname = $lib || mv $output_objdir/$soname $lib' fi hardcode_libdir_flag_spec_F77='${wl}+b ${wl}$libdir' hardcode_libdir_separator_F77=: hardcode_direct_F77=yes # hardcode_minus_L: Not really in the search PATH, # but as the default location of the library. hardcode_minus_L_F77=yes export_dynamic_flag_spec_F77='${wl}-E' ;; hpux10*) if test "$GCC" = yes && test "$with_gnu_ld" = no; then archive_cmds_F77='$CC -shared $pic_flag ${wl}+h ${wl}$soname ${wl}+b ${wl}$install_libdir -o $lib $libobjs $deplibs $compiler_flags' else archive_cmds_F77='$LD -b +h $soname +b $install_libdir -o $lib $libobjs $deplibs $linker_flags' fi if test "$with_gnu_ld" = no; then hardcode_libdir_flag_spec_F77='${wl}+b ${wl}$libdir' hardcode_libdir_separator_F77=: hardcode_direct_F77=yes hardcode_direct_absolute_F77=yes export_dynamic_flag_spec_F77='${wl}-E' # hardcode_minus_L: Not really in the search PATH, # but as the default location of the library. hardcode_minus_L_F77=yes fi ;; hpux11*) if test "$GCC" = yes && test "$with_gnu_ld" = no; then case $host_cpu in hppa*64*) archive_cmds_F77='$CC -shared ${wl}+h ${wl}$soname -o $lib $libobjs $deplibs $compiler_flags' ;; ia64*) archive_cmds_F77='$CC -shared $pic_flag ${wl}+h ${wl}$soname ${wl}+nodefaultrpath -o $lib $libobjs $deplibs $compiler_flags' ;; *) archive_cmds_F77='$CC -shared $pic_flag ${wl}+h ${wl}$soname ${wl}+b ${wl}$install_libdir -o $lib $libobjs $deplibs $compiler_flags' ;; esac else case $host_cpu in hppa*64*) archive_cmds_F77='$CC -b ${wl}+h ${wl}$soname -o $lib $libobjs $deplibs $compiler_flags' ;; ia64*) archive_cmds_F77='$CC -b ${wl}+h ${wl}$soname ${wl}+nodefaultrpath -o $lib $libobjs $deplibs $compiler_flags' ;; *) archive_cmds_F77='$CC -b ${wl}+h ${wl}$soname ${wl}+b ${wl}$install_libdir -o $lib $libobjs $deplibs $compiler_flags' ;; esac fi if test "$with_gnu_ld" = no; then hardcode_libdir_flag_spec_F77='${wl}+b ${wl}$libdir' hardcode_libdir_separator_F77=: case $host_cpu in hppa*64*|ia64*) hardcode_direct_F77=no hardcode_shlibpath_var_F77=no ;; *) hardcode_direct_F77=yes hardcode_direct_absolute_F77=yes export_dynamic_flag_spec_F77='${wl}-E' # hardcode_minus_L: Not really in the search PATH, # but as the default location of the library. hardcode_minus_L_F77=yes ;; esac fi ;; irix5* | irix6* | nonstopux*) if test "$GCC" = yes; then archive_cmds_F77='$CC -shared $pic_flag $libobjs $deplibs $compiler_flags ${wl}-soname ${wl}$soname `test -n "$verstring" && func_echo_all "${wl}-set_version ${wl}$verstring"` ${wl}-update_registry ${wl}${output_objdir}/so_locations -o $lib' # Try to use the -exported_symbol ld option, if it does not # work, assume that -exports_file does not work either and # implicitly export all symbols. # This should be the same for all languages, so no per-tag cache variable. { $as_echo "$as_me:${as_lineno-$LINENO}: checking whether the $host_os linker accepts -exported_symbol" >&5 $as_echo_n "checking whether the $host_os linker accepts -exported_symbol... " >&6; } if ${lt_cv_irix_exported_symbol+:} false; then : $as_echo_n "(cached) " >&6 else save_LDFLAGS="$LDFLAGS" LDFLAGS="$LDFLAGS -shared ${wl}-exported_symbol ${wl}foo ${wl}-update_registry ${wl}/dev/null" cat > conftest.$ac_ext <<_ACEOF subroutine foo end _ACEOF if ac_fn_f77_try_link "$LINENO"; then : lt_cv_irix_exported_symbol=yes else lt_cv_irix_exported_symbol=no fi rm -f core conftest.err conftest.$ac_objext \ conftest$ac_exeext conftest.$ac_ext LDFLAGS="$save_LDFLAGS" fi { $as_echo "$as_me:${as_lineno-$LINENO}: result: $lt_cv_irix_exported_symbol" >&5 $as_echo "$lt_cv_irix_exported_symbol" >&6; } if test "$lt_cv_irix_exported_symbol" = yes; then archive_expsym_cmds_F77='$CC -shared $pic_flag $libobjs $deplibs $compiler_flags ${wl}-soname ${wl}$soname `test -n "$verstring" && func_echo_all "${wl}-set_version ${wl}$verstring"` ${wl}-update_registry ${wl}${output_objdir}/so_locations ${wl}-exports_file ${wl}$export_symbols -o $lib' fi else archive_cmds_F77='$CC -shared $libobjs $deplibs $compiler_flags -soname $soname `test -n "$verstring" && func_echo_all "-set_version $verstring"` -update_registry ${output_objdir}/so_locations -o $lib' archive_expsym_cmds_F77='$CC -shared $libobjs $deplibs $compiler_flags -soname $soname `test -n "$verstring" && func_echo_all "-set_version $verstring"` -update_registry ${output_objdir}/so_locations -exports_file $export_symbols -o $lib' fi archive_cmds_need_lc_F77='no' hardcode_libdir_flag_spec_F77='${wl}-rpath ${wl}$libdir' hardcode_libdir_separator_F77=: inherit_rpath_F77=yes link_all_deplibs_F77=yes ;; netbsd* | netbsdelf*-gnu) if echo __ELF__ | $CC -E - | $GREP __ELF__ >/dev/null; then archive_cmds_F77='$LD -Bshareable -o $lib $libobjs $deplibs $linker_flags' # a.out else archive_cmds_F77='$LD -shared -o $lib $libobjs $deplibs $linker_flags' # ELF fi hardcode_libdir_flag_spec_F77='-R$libdir' hardcode_direct_F77=yes hardcode_shlibpath_var_F77=no ;; newsos6) archive_cmds_F77='$LD -G -h $soname -o $lib $libobjs $deplibs $linker_flags' hardcode_direct_F77=yes hardcode_libdir_flag_spec_F77='${wl}-rpath ${wl}$libdir' hardcode_libdir_separator_F77=: hardcode_shlibpath_var_F77=no ;; *nto* | *qnx*) ;; openbsd*) if test -f /usr/libexec/ld.so; then hardcode_direct_F77=yes hardcode_shlibpath_var_F77=no hardcode_direct_absolute_F77=yes if test -z "`echo __ELF__ | $CC -E - | $GREP __ELF__`" || test "$host_os-$host_cpu" = "openbsd2.8-powerpc"; then archive_cmds_F77='$CC -shared $pic_flag -o $lib $libobjs $deplibs $compiler_flags' archive_expsym_cmds_F77='$CC -shared $pic_flag -o $lib $libobjs $deplibs $compiler_flags ${wl}-retain-symbols-file,$export_symbols' hardcode_libdir_flag_spec_F77='${wl}-rpath,$libdir' export_dynamic_flag_spec_F77='${wl}-E' else case $host_os in openbsd[01].* | openbsd2.[0-7] | openbsd2.[0-7].*) archive_cmds_F77='$LD -Bshareable -o $lib $libobjs $deplibs $linker_flags' hardcode_libdir_flag_spec_F77='-R$libdir' ;; *) archive_cmds_F77='$CC -shared $pic_flag -o $lib $libobjs $deplibs $compiler_flags' hardcode_libdir_flag_spec_F77='${wl}-rpath,$libdir' ;; esac fi else ld_shlibs_F77=no fi ;; os2*) hardcode_libdir_flag_spec_F77='-L$libdir' hardcode_minus_L_F77=yes allow_undefined_flag_F77=unsupported archive_cmds_F77='$ECHO "LIBRARY $libname INITINSTANCE" > $output_objdir/$libname.def~$ECHO "DESCRIPTION \"$libname\"" >> $output_objdir/$libname.def~echo DATA >> $output_objdir/$libname.def~echo " SINGLE NONSHARED" >> $output_objdir/$libname.def~echo EXPORTS >> $output_objdir/$libname.def~emxexp $libobjs >> $output_objdir/$libname.def~$CC -Zdll -Zcrtdll -o $lib $libobjs $deplibs $compiler_flags $output_objdir/$libname.def' old_archive_from_new_cmds_F77='emximp -o $output_objdir/$libname.a $output_objdir/$libname.def' ;; osf3*) if test "$GCC" = yes; then allow_undefined_flag_F77=' ${wl}-expect_unresolved ${wl}\*' archive_cmds_F77='$CC -shared${allow_undefined_flag} $libobjs $deplibs $compiler_flags ${wl}-soname ${wl}$soname `test -n "$verstring" && func_echo_all "${wl}-set_version ${wl}$verstring"` ${wl}-update_registry ${wl}${output_objdir}/so_locations -o $lib' else allow_undefined_flag_F77=' -expect_unresolved \*' archive_cmds_F77='$CC -shared${allow_undefined_flag} $libobjs $deplibs $compiler_flags -soname $soname `test -n "$verstring" && func_echo_all "-set_version $verstring"` -update_registry ${output_objdir}/so_locations -o $lib' fi archive_cmds_need_lc_F77='no' hardcode_libdir_flag_spec_F77='${wl}-rpath ${wl}$libdir' hardcode_libdir_separator_F77=: ;; osf4* | osf5*) # as osf3* with the addition of -msym flag if test "$GCC" = yes; then allow_undefined_flag_F77=' ${wl}-expect_unresolved ${wl}\*' archive_cmds_F77='$CC -shared${allow_undefined_flag} $pic_flag $libobjs $deplibs $compiler_flags ${wl}-msym ${wl}-soname ${wl}$soname `test -n "$verstring" && func_echo_all "${wl}-set_version ${wl}$verstring"` ${wl}-update_registry ${wl}${output_objdir}/so_locations -o $lib' hardcode_libdir_flag_spec_F77='${wl}-rpath ${wl}$libdir' else allow_undefined_flag_F77=' -expect_unresolved \*' archive_cmds_F77='$CC -shared${allow_undefined_flag} $libobjs $deplibs $compiler_flags -msym -soname $soname `test -n "$verstring" && func_echo_all "-set_version $verstring"` -update_registry ${output_objdir}/so_locations -o $lib' archive_expsym_cmds_F77='for i in `cat $export_symbols`; do printf "%s %s\\n" -exported_symbol "\$i" >> $lib.exp; done; printf "%s\\n" "-hidden">> $lib.exp~ $CC -shared${allow_undefined_flag} ${wl}-input ${wl}$lib.exp $compiler_flags $libobjs $deplibs -soname $soname `test -n "$verstring" && $ECHO "-set_version $verstring"` -update_registry ${output_objdir}/so_locations -o $lib~$RM $lib.exp' # Both c and cxx compiler support -rpath directly hardcode_libdir_flag_spec_F77='-rpath $libdir' fi archive_cmds_need_lc_F77='no' hardcode_libdir_separator_F77=: ;; solaris*) no_undefined_flag_F77=' -z defs' if test "$GCC" = yes; then wlarc='${wl}' archive_cmds_F77='$CC -shared $pic_flag ${wl}-z ${wl}text ${wl}-h ${wl}$soname -o $lib $libobjs $deplibs $compiler_flags' archive_expsym_cmds_F77='echo "{ global:" > $lib.exp~cat $export_symbols | $SED -e "s/\(.*\)/\1;/" >> $lib.exp~echo "local: *; };" >> $lib.exp~ $CC -shared $pic_flag ${wl}-z ${wl}text ${wl}-M ${wl}$lib.exp ${wl}-h ${wl}$soname -o $lib $libobjs $deplibs $compiler_flags~$RM $lib.exp' else case `$CC -V 2>&1` in *"Compilers 5.0"*) wlarc='' archive_cmds_F77='$LD -G${allow_undefined_flag} -h $soname -o $lib $libobjs $deplibs $linker_flags' archive_expsym_cmds_F77='echo "{ global:" > $lib.exp~cat $export_symbols | $SED -e "s/\(.*\)/\1;/" >> $lib.exp~echo "local: *; };" >> $lib.exp~ $LD -G${allow_undefined_flag} -M $lib.exp -h $soname -o $lib $libobjs $deplibs $linker_flags~$RM $lib.exp' ;; *) wlarc='${wl}' archive_cmds_F77='$CC -G${allow_undefined_flag} -h $soname -o $lib $libobjs $deplibs $compiler_flags' archive_expsym_cmds_F77='echo "{ global:" > $lib.exp~cat $export_symbols | $SED -e "s/\(.*\)/\1;/" >> $lib.exp~echo "local: *; };" >> $lib.exp~ $CC -G${allow_undefined_flag} -M $lib.exp -h $soname -o $lib $libobjs $deplibs $compiler_flags~$RM $lib.exp' ;; esac fi hardcode_libdir_flag_spec_F77='-R$libdir' hardcode_shlibpath_var_F77=no case $host_os in solaris2.[0-5] | solaris2.[0-5].*) ;; *) # The compiler driver will combine and reorder linker options, # but understands `-z linker_flag'. GCC discards it without `$wl', # but is careful enough not to reorder. # Supported since Solaris 2.6 (maybe 2.5.1?) if test "$GCC" = yes; then whole_archive_flag_spec_F77='${wl}-z ${wl}allextract$convenience ${wl}-z ${wl}defaultextract' else whole_archive_flag_spec_F77='-z allextract$convenience -z defaultextract' fi ;; esac link_all_deplibs_F77=yes ;; sunos4*) if test "x$host_vendor" = xsequent; then # Use $CC to link under sequent, because it throws in some extra .o # files that make .init and .fini sections work. archive_cmds_F77='$CC -G ${wl}-h $soname -o $lib $libobjs $deplibs $compiler_flags' else archive_cmds_F77='$LD -assert pure-text -Bstatic -o $lib $libobjs $deplibs $linker_flags' fi hardcode_libdir_flag_spec_F77='-L$libdir' hardcode_direct_F77=yes hardcode_minus_L_F77=yes hardcode_shlibpath_var_F77=no ;; sysv4) case $host_vendor in sni) archive_cmds_F77='$LD -G -h $soname -o $lib $libobjs $deplibs $linker_flags' hardcode_direct_F77=yes # is this really true??? ;; siemens) ## LD is ld it makes a PLAMLIB ## CC just makes a GrossModule. archive_cmds_F77='$LD -G -o $lib $libobjs $deplibs $linker_flags' reload_cmds_F77='$CC -r -o $output$reload_objs' hardcode_direct_F77=no ;; motorola) archive_cmds_F77='$LD -G -h $soname -o $lib $libobjs $deplibs $linker_flags' hardcode_direct_F77=no #Motorola manual says yes, but my tests say they lie ;; esac runpath_var='LD_RUN_PATH' hardcode_shlibpath_var_F77=no ;; sysv4.3*) archive_cmds_F77='$LD -G -h $soname -o $lib $libobjs $deplibs $linker_flags' hardcode_shlibpath_var_F77=no export_dynamic_flag_spec_F77='-Bexport' ;; sysv4*MP*) if test -d /usr/nec; then archive_cmds_F77='$LD -G -h $soname -o $lib $libobjs $deplibs $linker_flags' hardcode_shlibpath_var_F77=no runpath_var=LD_RUN_PATH hardcode_runpath_var=yes ld_shlibs_F77=yes fi ;; sysv4*uw2* | sysv5OpenUNIX* | sysv5UnixWare7.[01].[10]* | unixware7* | sco3.2v5.0.[024]*) no_undefined_flag_F77='${wl}-z,text' archive_cmds_need_lc_F77=no hardcode_shlibpath_var_F77=no runpath_var='LD_RUN_PATH' if test "$GCC" = yes; then archive_cmds_F77='$CC -shared ${wl}-h,$soname -o $lib $libobjs $deplibs $compiler_flags' archive_expsym_cmds_F77='$CC -shared ${wl}-Bexport:$export_symbols ${wl}-h,$soname -o $lib $libobjs $deplibs $compiler_flags' else archive_cmds_F77='$CC -G ${wl}-h,$soname -o $lib $libobjs $deplibs $compiler_flags' archive_expsym_cmds_F77='$CC -G ${wl}-Bexport:$export_symbols ${wl}-h,$soname -o $lib $libobjs $deplibs $compiler_flags' fi ;; sysv5* | sco3.2v5* | sco5v6*) # Note: We can NOT use -z defs as we might desire, because we do not # link with -lc, and that would cause any symbols used from libc to # always be unresolved, which means just about no library would # ever link correctly. If we're not using GNU ld we use -z text # though, which does catch some bad symbols but isn't as heavy-handed # as -z defs. no_undefined_flag_F77='${wl}-z,text' allow_undefined_flag_F77='${wl}-z,nodefs' archive_cmds_need_lc_F77=no hardcode_shlibpath_var_F77=no hardcode_libdir_flag_spec_F77='${wl}-R,$libdir' hardcode_libdir_separator_F77=':' link_all_deplibs_F77=yes export_dynamic_flag_spec_F77='${wl}-Bexport' runpath_var='LD_RUN_PATH' if test "$GCC" = yes; then archive_cmds_F77='$CC -shared ${wl}-h,$soname -o $lib $libobjs $deplibs $compiler_flags' archive_expsym_cmds_F77='$CC -shared ${wl}-Bexport:$export_symbols ${wl}-h,$soname -o $lib $libobjs $deplibs $compiler_flags' else archive_cmds_F77='$CC -G ${wl}-h,$soname -o $lib $libobjs $deplibs $compiler_flags' archive_expsym_cmds_F77='$CC -G ${wl}-Bexport:$export_symbols ${wl}-h,$soname -o $lib $libobjs $deplibs $compiler_flags' fi ;; uts4*) archive_cmds_F77='$LD -G -h $soname -o $lib $libobjs $deplibs $linker_flags' hardcode_libdir_flag_spec_F77='-L$libdir' hardcode_shlibpath_var_F77=no ;; *) ld_shlibs_F77=no ;; esac if test x$host_vendor = xsni; then case $host in sysv4 | sysv4.2uw2* | sysv4.3* | sysv5*) export_dynamic_flag_spec_F77='${wl}-Blargedynsym' ;; esac fi fi { $as_echo "$as_me:${as_lineno-$LINENO}: result: $ld_shlibs_F77" >&5 $as_echo "$ld_shlibs_F77" >&6; } test "$ld_shlibs_F77" = no && can_build_shared=no with_gnu_ld_F77=$with_gnu_ld # # Do we need to explicitly link libc? # case "x$archive_cmds_need_lc_F77" in x|xyes) # Assume -lc should be added archive_cmds_need_lc_F77=yes if test "$enable_shared" = yes && test "$GCC" = yes; then case $archive_cmds_F77 in *'~'*) # FIXME: we may have to deal with multi-command sequences. ;; '$CC '*) # Test whether the compiler implicitly links with -lc since on some # systems, -lgcc has to come before -lc. If gcc already passes -lc # to ld, don't add -lc before -lgcc. { $as_echo "$as_me:${as_lineno-$LINENO}: checking whether -lc should be explicitly linked in" >&5 $as_echo_n "checking whether -lc should be explicitly linked in... " >&6; } if ${lt_cv_archive_cmds_need_lc_F77+:} false; then : $as_echo_n "(cached) " >&6 else $RM conftest* echo "$lt_simple_compile_test_code" > conftest.$ac_ext if { { eval echo "\"\$as_me\":${as_lineno-$LINENO}: \"$ac_compile\""; } >&5 (eval $ac_compile) 2>&5 ac_status=$? $as_echo "$as_me:${as_lineno-$LINENO}: \$? = $ac_status" >&5 test $ac_status = 0; } 2>conftest.err; then soname=conftest lib=conftest libobjs=conftest.$ac_objext deplibs= wl=$lt_prog_compiler_wl_F77 pic_flag=$lt_prog_compiler_pic_F77 compiler_flags=-v linker_flags=-v verstring= output_objdir=. libname=conftest lt_save_allow_undefined_flag=$allow_undefined_flag_F77 allow_undefined_flag_F77= if { { eval echo "\"\$as_me\":${as_lineno-$LINENO}: \"$archive_cmds_F77 2\>\&1 \| $GREP \" -lc \" \>/dev/null 2\>\&1\""; } >&5 (eval $archive_cmds_F77 2\>\&1 \| $GREP \" -lc \" \>/dev/null 2\>\&1) 2>&5 ac_status=$? $as_echo "$as_me:${as_lineno-$LINENO}: \$? = $ac_status" >&5 test $ac_status = 0; } then lt_cv_archive_cmds_need_lc_F77=no else lt_cv_archive_cmds_need_lc_F77=yes fi allow_undefined_flag_F77=$lt_save_allow_undefined_flag else cat conftest.err 1>&5 fi $RM conftest* fi { $as_echo "$as_me:${as_lineno-$LINENO}: result: $lt_cv_archive_cmds_need_lc_F77" >&5 $as_echo "$lt_cv_archive_cmds_need_lc_F77" >&6; } archive_cmds_need_lc_F77=$lt_cv_archive_cmds_need_lc_F77 ;; esac fi ;; esac { $as_echo "$as_me:${as_lineno-$LINENO}: checking dynamic linker characteristics" >&5 $as_echo_n "checking dynamic linker characteristics... " >&6; } library_names_spec= libname_spec='lib$name' soname_spec= shrext_cmds=".so" postinstall_cmds= postuninstall_cmds= finish_cmds= finish_eval= shlibpath_var= shlibpath_overrides_runpath=unknown version_type=none dynamic_linker="$host_os ld.so" sys_lib_dlsearch_path_spec="/lib /usr/lib" need_lib_prefix=unknown hardcode_into_libs=no # when you set need_version to no, make sure it does not cause -set_version # flags to be left without arguments need_version=unknown case $host_os in aix3*) version_type=linux # correct to gnu/linux during the next big refactor library_names_spec='${libname}${release}${shared_ext}$versuffix $libname.a' shlibpath_var=LIBPATH # AIX 3 has no versioning support, so we append a major version to the name. soname_spec='${libname}${release}${shared_ext}$major' ;; aix[4-9]*) version_type=linux # correct to gnu/linux during the next big refactor need_lib_prefix=no need_version=no hardcode_into_libs=yes if test "$host_cpu" = ia64; then # AIX 5 supports IA64 library_names_spec='${libname}${release}${shared_ext}$major ${libname}${release}${shared_ext}$versuffix $libname${shared_ext}' shlibpath_var=LD_LIBRARY_PATH else # With GCC up to 2.95.x, collect2 would create an import file # for dependence libraries. The import file would start with # the line `#! .'. This would cause the generated library to # depend on `.', always an invalid library. This was fixed in # development snapshots of GCC prior to 3.0. case $host_os in aix4 | aix4.[01] | aix4.[01].*) if { echo '#if __GNUC__ > 2 || (__GNUC__ == 2 && __GNUC_MINOR__ >= 97)' echo ' yes ' echo '#endif'; } | ${CC} -E - | $GREP yes > /dev/null; then : else can_build_shared=no fi ;; esac # AIX (on Power*) has no versioning support, so currently we can not hardcode correct # soname into executable. Probably we can add versioning support to # collect2, so additional links can be useful in future. if test "$aix_use_runtimelinking" = yes; then # If using run time linking (on AIX 4.2 or later) use lib.so # instead of lib.a to let people know that these are not # typical AIX shared libraries. library_names_spec='${libname}${release}${shared_ext}$versuffix ${libname}${release}${shared_ext}$major $libname${shared_ext}' else # We preserve .a as extension for shared libraries through AIX4.2 # and later when we are not doing run time linking. library_names_spec='${libname}${release}.a $libname.a' soname_spec='${libname}${release}${shared_ext}$major' fi shlibpath_var=LIBPATH fi ;; amigaos*) case $host_cpu in powerpc) # Since July 2007 AmigaOS4 officially supports .so libraries. # When compiling the executable, add -use-dynld -Lsobjs: to the compileline. library_names_spec='${libname}${release}${shared_ext}$versuffix ${libname}${release}${shared_ext}$major $libname${shared_ext}' ;; m68k) library_names_spec='$libname.ixlibrary $libname.a' # Create ${libname}_ixlibrary.a entries in /sys/libs. finish_eval='for lib in `ls $libdir/*.ixlibrary 2>/dev/null`; do libname=`func_echo_all "$lib" | $SED '\''s%^.*/\([^/]*\)\.ixlibrary$%\1%'\''`; test $RM /sys/libs/${libname}_ixlibrary.a; $show "cd /sys/libs && $LN_S $lib ${libname}_ixlibrary.a"; cd /sys/libs && $LN_S $lib ${libname}_ixlibrary.a || exit 1; done' ;; esac ;; beos*) library_names_spec='${libname}${shared_ext}' dynamic_linker="$host_os ld.so" shlibpath_var=LIBRARY_PATH ;; bsdi[45]*) version_type=linux # correct to gnu/linux during the next big refactor need_version=no library_names_spec='${libname}${release}${shared_ext}$versuffix ${libname}${release}${shared_ext}$major $libname${shared_ext}' soname_spec='${libname}${release}${shared_ext}$major' finish_cmds='PATH="\$PATH:/sbin" ldconfig $libdir' shlibpath_var=LD_LIBRARY_PATH sys_lib_search_path_spec="/shlib /usr/lib /usr/X11/lib /usr/contrib/lib /lib /usr/local/lib" sys_lib_dlsearch_path_spec="/shlib /usr/lib /usr/local/lib" # the default ld.so.conf also contains /usr/contrib/lib and # /usr/X11R6/lib (/usr/X11 is a link to /usr/X11R6), but let us allow # libtool to hard-code these into programs ;; cygwin* | mingw* | pw32* | cegcc*) version_type=windows shrext_cmds=".dll" need_version=no need_lib_prefix=no case $GCC,$cc_basename in yes,*) # gcc library_names_spec='$libname.dll.a' # DLL is installed to $(libdir)/../bin by postinstall_cmds postinstall_cmds='base_file=`basename \${file}`~ dlpath=`$SHELL 2>&1 -c '\''. $dir/'\''\${base_file}'\''i; echo \$dlname'\''`~ dldir=$destdir/`dirname \$dlpath`~ test -d \$dldir || mkdir -p \$dldir~ $install_prog $dir/$dlname \$dldir/$dlname~ chmod a+x \$dldir/$dlname~ if test -n '\''$stripme'\'' && test -n '\''$striplib'\''; then eval '\''$striplib \$dldir/$dlname'\'' || exit \$?; fi' postuninstall_cmds='dldll=`$SHELL 2>&1 -c '\''. $file; echo \$dlname'\''`~ dlpath=$dir/\$dldll~ $RM \$dlpath' shlibpath_overrides_runpath=yes case $host_os in cygwin*) # Cygwin DLLs use 'cyg' prefix rather than 'lib' soname_spec='`echo ${libname} | sed -e 's/^lib/cyg/'``echo ${release} | $SED -e 's/[.]/-/g'`${versuffix}${shared_ext}' ;; mingw* | cegcc*) # MinGW DLLs use traditional 'lib' prefix soname_spec='${libname}`echo ${release} | $SED -e 's/[.]/-/g'`${versuffix}${shared_ext}' ;; pw32*) # pw32 DLLs use 'pw' prefix rather than 'lib' library_names_spec='`echo ${libname} | sed -e 's/^lib/pw/'``echo ${release} | $SED -e 's/[.]/-/g'`${versuffix}${shared_ext}' ;; esac dynamic_linker='Win32 ld.exe' ;; *,cl*) # Native MSVC libname_spec='$name' soname_spec='${libname}`echo ${release} | $SED -e 's/[.]/-/g'`${versuffix}${shared_ext}' library_names_spec='${libname}.dll.lib' case $build_os in mingw*) sys_lib_search_path_spec= lt_save_ifs=$IFS IFS=';' for lt_path in $LIB do IFS=$lt_save_ifs # Let DOS variable expansion print the short 8.3 style file name. lt_path=`cd "$lt_path" 2>/dev/null && cmd //C "for %i in (".") do @echo %~si"` sys_lib_search_path_spec="$sys_lib_search_path_spec $lt_path" done IFS=$lt_save_ifs # Convert to MSYS style. sys_lib_search_path_spec=`$ECHO "$sys_lib_search_path_spec" | sed -e 's|\\\\|/|g' -e 's| \\([a-zA-Z]\\):| /\\1|g' -e 's|^ ||'` ;; cygwin*) # Convert to unix form, then to dos form, then back to unix form # but this time dos style (no spaces!) so that the unix form looks # like /cygdrive/c/PROGRA~1:/cygdr... sys_lib_search_path_spec=`cygpath --path --unix "$LIB"` sys_lib_search_path_spec=`cygpath --path --dos "$sys_lib_search_path_spec" 2>/dev/null` sys_lib_search_path_spec=`cygpath --path --unix "$sys_lib_search_path_spec" | $SED -e "s/$PATH_SEPARATOR/ /g"` ;; *) sys_lib_search_path_spec="$LIB" if $ECHO "$sys_lib_search_path_spec" | $GREP ';[c-zC-Z]:/' >/dev/null; then # It is most probably a Windows format PATH. sys_lib_search_path_spec=`$ECHO "$sys_lib_search_path_spec" | $SED -e 's/;/ /g'` else sys_lib_search_path_spec=`$ECHO "$sys_lib_search_path_spec" | $SED -e "s/$PATH_SEPARATOR/ /g"` fi # FIXME: find the short name or the path components, as spaces are # common. (e.g. "Program Files" -> "PROGRA~1") ;; esac # DLL is installed to $(libdir)/../bin by postinstall_cmds postinstall_cmds='base_file=`basename \${file}`~ dlpath=`$SHELL 2>&1 -c '\''. $dir/'\''\${base_file}'\''i; echo \$dlname'\''`~ dldir=$destdir/`dirname \$dlpath`~ test -d \$dldir || mkdir -p \$dldir~ $install_prog $dir/$dlname \$dldir/$dlname' postuninstall_cmds='dldll=`$SHELL 2>&1 -c '\''. $file; echo \$dlname'\''`~ dlpath=$dir/\$dldll~ $RM \$dlpath' shlibpath_overrides_runpath=yes dynamic_linker='Win32 link.exe' ;; *) # Assume MSVC wrapper library_names_spec='${libname}`echo ${release} | $SED -e 's/[.]/-/g'`${versuffix}${shared_ext} $libname.lib' dynamic_linker='Win32 ld.exe' ;; esac # FIXME: first we should search . and the directory the executable is in shlibpath_var=PATH ;; darwin* | rhapsody*) dynamic_linker="$host_os dyld" version_type=darwin need_lib_prefix=no need_version=no library_names_spec='${libname}${release}${major}$shared_ext ${libname}$shared_ext' soname_spec='${libname}${release}${major}$shared_ext' shlibpath_overrides_runpath=yes shlibpath_var=DYLD_LIBRARY_PATH shrext_cmds='`test .$module = .yes && echo .so || echo .dylib`' sys_lib_dlsearch_path_spec='/usr/local/lib /lib /usr/lib' ;; dgux*) version_type=linux # correct to gnu/linux during the next big refactor need_lib_prefix=no need_version=no library_names_spec='${libname}${release}${shared_ext}$versuffix ${libname}${release}${shared_ext}$major $libname$shared_ext' soname_spec='${libname}${release}${shared_ext}$major' shlibpath_var=LD_LIBRARY_PATH ;; freebsd* | dragonfly*) # DragonFly does not have aout. When/if they implement a new # versioning mechanism, adjust this. if test -x /usr/bin/objformat; then objformat=`/usr/bin/objformat` else case $host_os in freebsd[23].*) objformat=aout ;; *) objformat=elf ;; esac fi version_type=freebsd-$objformat case $version_type in freebsd-elf*) library_names_spec='${libname}${release}${shared_ext}$versuffix ${libname}${release}${shared_ext} $libname${shared_ext}' need_version=no need_lib_prefix=no ;; freebsd-*) library_names_spec='${libname}${release}${shared_ext}$versuffix $libname${shared_ext}$versuffix' need_version=yes ;; esac shlibpath_var=LD_LIBRARY_PATH case $host_os in freebsd2.*) shlibpath_overrides_runpath=yes ;; freebsd3.[01]* | freebsdelf3.[01]*) shlibpath_overrides_runpath=yes hardcode_into_libs=yes ;; freebsd3.[2-9]* | freebsdelf3.[2-9]* | \ freebsd4.[0-5] | freebsdelf4.[0-5] | freebsd4.1.1 | freebsdelf4.1.1) shlibpath_overrides_runpath=no hardcode_into_libs=yes ;; *) # from 4.6 on, and DragonFly shlibpath_overrides_runpath=yes hardcode_into_libs=yes ;; esac ;; haiku*) version_type=linux # correct to gnu/linux during the next big refactor need_lib_prefix=no need_version=no dynamic_linker="$host_os runtime_loader" library_names_spec='${libname}${release}${shared_ext}$versuffix ${libname}${release}${shared_ext}${major} ${libname}${shared_ext}' soname_spec='${libname}${release}${shared_ext}$major' shlibpath_var=LIBRARY_PATH shlibpath_overrides_runpath=yes sys_lib_dlsearch_path_spec='/boot/home/config/lib /boot/common/lib /boot/system/lib' hardcode_into_libs=yes ;; hpux9* | hpux10* | hpux11*) # Give a soname corresponding to the major version so that dld.sl refuses to # link against other versions. version_type=sunos need_lib_prefix=no need_version=no case $host_cpu in ia64*) shrext_cmds='.so' hardcode_into_libs=yes dynamic_linker="$host_os dld.so" shlibpath_var=LD_LIBRARY_PATH shlibpath_overrides_runpath=yes # Unless +noenvvar is specified. library_names_spec='${libname}${release}${shared_ext}$versuffix ${libname}${release}${shared_ext}$major $libname${shared_ext}' soname_spec='${libname}${release}${shared_ext}$major' if test "X$HPUX_IA64_MODE" = X32; then sys_lib_search_path_spec="/usr/lib/hpux32 /usr/local/lib/hpux32 /usr/local/lib" else sys_lib_search_path_spec="/usr/lib/hpux64 /usr/local/lib/hpux64" fi sys_lib_dlsearch_path_spec=$sys_lib_search_path_spec ;; hppa*64*) shrext_cmds='.sl' hardcode_into_libs=yes dynamic_linker="$host_os dld.sl" shlibpath_var=LD_LIBRARY_PATH # How should we handle SHLIB_PATH shlibpath_overrides_runpath=yes # Unless +noenvvar is specified. library_names_spec='${libname}${release}${shared_ext}$versuffix ${libname}${release}${shared_ext}$major $libname${shared_ext}' soname_spec='${libname}${release}${shared_ext}$major' sys_lib_search_path_spec="/usr/lib/pa20_64 /usr/ccs/lib/pa20_64" sys_lib_dlsearch_path_spec=$sys_lib_search_path_spec ;; *) shrext_cmds='.sl' dynamic_linker="$host_os dld.sl" shlibpath_var=SHLIB_PATH shlibpath_overrides_runpath=no # +s is required to enable SHLIB_PATH library_names_spec='${libname}${release}${shared_ext}$versuffix ${libname}${release}${shared_ext}$major $libname${shared_ext}' soname_spec='${libname}${release}${shared_ext}$major' ;; esac # HP-UX runs *really* slowly unless shared libraries are mode 555, ... postinstall_cmds='chmod 555 $lib' # or fails outright, so override atomically: install_override_mode=555 ;; interix[3-9]*) version_type=linux # correct to gnu/linux during the next big refactor need_lib_prefix=no need_version=no library_names_spec='${libname}${release}${shared_ext}$versuffix ${libname}${release}${shared_ext}$major ${libname}${shared_ext}' soname_spec='${libname}${release}${shared_ext}$major' dynamic_linker='Interix 3.x ld.so.1 (PE, like ELF)' shlibpath_var=LD_LIBRARY_PATH shlibpath_overrides_runpath=no hardcode_into_libs=yes ;; irix5* | irix6* | nonstopux*) case $host_os in nonstopux*) version_type=nonstopux ;; *) if test "$lt_cv_prog_gnu_ld" = yes; then version_type=linux # correct to gnu/linux during the next big refactor else version_type=irix fi ;; esac need_lib_prefix=no need_version=no soname_spec='${libname}${release}${shared_ext}$major' library_names_spec='${libname}${release}${shared_ext}$versuffix ${libname}${release}${shared_ext}$major ${libname}${release}${shared_ext} $libname${shared_ext}' case $host_os in irix5* | nonstopux*) libsuff= shlibsuff= ;; *) case $LD in # libtool.m4 will add one of these switches to LD *-32|*"-32 "|*-melf32bsmip|*"-melf32bsmip ") libsuff= shlibsuff= libmagic=32-bit;; *-n32|*"-n32 "|*-melf32bmipn32|*"-melf32bmipn32 ") libsuff=32 shlibsuff=N32 libmagic=N32;; *-64|*"-64 "|*-melf64bmip|*"-melf64bmip ") libsuff=64 shlibsuff=64 libmagic=64-bit;; *) libsuff= shlibsuff= libmagic=never-match;; esac ;; esac shlibpath_var=LD_LIBRARY${shlibsuff}_PATH shlibpath_overrides_runpath=no sys_lib_search_path_spec="/usr/lib${libsuff} /lib${libsuff} /usr/local/lib${libsuff}" sys_lib_dlsearch_path_spec="/usr/lib${libsuff} /lib${libsuff}" hardcode_into_libs=yes ;; # No shared lib support for Linux oldld, aout, or coff. linux*oldld* | linux*aout* | linux*coff*) dynamic_linker=no ;; # This must be glibc/ELF. linux* | k*bsd*-gnu | kopensolaris*-gnu | gnu*) version_type=linux # correct to gnu/linux during the next big refactor need_lib_prefix=no need_version=no library_names_spec='${libname}${release}${shared_ext}$versuffix ${libname}${release}${shared_ext}$major $libname${shared_ext}' soname_spec='${libname}${release}${shared_ext}$major' finish_cmds='PATH="\$PATH:/sbin" ldconfig -n $libdir' shlibpath_var=LD_LIBRARY_PATH shlibpath_overrides_runpath=no # Some binutils ld are patched to set DT_RUNPATH if ${lt_cv_shlibpath_overrides_runpath+:} false; then : $as_echo_n "(cached) " >&6 else lt_cv_shlibpath_overrides_runpath=no save_LDFLAGS=$LDFLAGS save_libdir=$libdir eval "libdir=/foo; wl=\"$lt_prog_compiler_wl_F77\"; \ LDFLAGS=\"\$LDFLAGS $hardcode_libdir_flag_spec_F77\"" cat > conftest.$ac_ext <<_ACEOF program main end _ACEOF if ac_fn_f77_try_link "$LINENO"; then : if ($OBJDUMP -p conftest$ac_exeext) 2>/dev/null | grep "RUNPATH.*$libdir" >/dev/null; then : lt_cv_shlibpath_overrides_runpath=yes fi fi rm -f core conftest.err conftest.$ac_objext \ conftest$ac_exeext conftest.$ac_ext LDFLAGS=$save_LDFLAGS libdir=$save_libdir fi shlibpath_overrides_runpath=$lt_cv_shlibpath_overrides_runpath # This implies no fast_install, which is unacceptable. # Some rework will be needed to allow for fast_install # before this can be enabled. hardcode_into_libs=yes # Append ld.so.conf contents to the search path if test -f /etc/ld.so.conf; then lt_ld_extra=`awk '/^include / { system(sprintf("cd /etc; cat %s 2>/dev/null", \$2)); skip = 1; } { if (!skip) print \$0; skip = 0; }' < /etc/ld.so.conf | $SED -e 's/#.*//;/^[ ]*hwcap[ ]/d;s/[:, ]/ /g;s/=[^=]*$//;s/=[^= ]* / /g;s/"//g;/^$/d' | tr '\n' ' '` sys_lib_dlsearch_path_spec="/lib /usr/lib $lt_ld_extra" fi # We used to test for /lib/ld.so.1 and disable shared libraries on # powerpc, because MkLinux only supported shared libraries with the # GNU dynamic linker. Since this was broken with cross compilers, # most powerpc-linux boxes support dynamic linking these days and # people can always --disable-shared, the test was removed, and we # assume the GNU/Linux dynamic linker is in use. dynamic_linker='GNU/Linux ld.so' ;; netbsdelf*-gnu) version_type=linux need_lib_prefix=no need_version=no library_names_spec='${libname}${release}${shared_ext}$versuffix ${libname}${release}${shared_ext}$major ${libname}${shared_ext}' soname_spec='${libname}${release}${shared_ext}$major' shlibpath_var=LD_LIBRARY_PATH shlibpath_overrides_runpath=no hardcode_into_libs=yes dynamic_linker='NetBSD ld.elf_so' ;; netbsd*) version_type=sunos need_lib_prefix=no need_version=no if echo __ELF__ | $CC -E - | $GREP __ELF__ >/dev/null; then library_names_spec='${libname}${release}${shared_ext}$versuffix ${libname}${shared_ext}$versuffix' finish_cmds='PATH="\$PATH:/sbin" ldconfig -m $libdir' dynamic_linker='NetBSD (a.out) ld.so' else library_names_spec='${libname}${release}${shared_ext}$versuffix ${libname}${release}${shared_ext}$major ${libname}${shared_ext}' soname_spec='${libname}${release}${shared_ext}$major' dynamic_linker='NetBSD ld.elf_so' fi shlibpath_var=LD_LIBRARY_PATH shlibpath_overrides_runpath=yes hardcode_into_libs=yes ;; newsos6) version_type=linux # correct to gnu/linux during the next big refactor library_names_spec='${libname}${release}${shared_ext}$versuffix ${libname}${release}${shared_ext}$major $libname${shared_ext}' shlibpath_var=LD_LIBRARY_PATH shlibpath_overrides_runpath=yes ;; *nto* | *qnx*) version_type=qnx need_lib_prefix=no need_version=no library_names_spec='${libname}${release}${shared_ext}$versuffix ${libname}${release}${shared_ext}$major $libname${shared_ext}' soname_spec='${libname}${release}${shared_ext}$major' shlibpath_var=LD_LIBRARY_PATH shlibpath_overrides_runpath=no hardcode_into_libs=yes dynamic_linker='ldqnx.so' ;; openbsd*) version_type=sunos sys_lib_dlsearch_path_spec="/usr/lib" need_lib_prefix=no # Some older versions of OpenBSD (3.3 at least) *do* need versioned libs. case $host_os in openbsd3.3 | openbsd3.3.*) need_version=yes ;; *) need_version=no ;; esac library_names_spec='${libname}${release}${shared_ext}$versuffix ${libname}${shared_ext}$versuffix' finish_cmds='PATH="\$PATH:/sbin" ldconfig -m $libdir' shlibpath_var=LD_LIBRARY_PATH if test -z "`echo __ELF__ | $CC -E - | $GREP __ELF__`" || test "$host_os-$host_cpu" = "openbsd2.8-powerpc"; then case $host_os in openbsd2.[89] | openbsd2.[89].*) shlibpath_overrides_runpath=no ;; *) shlibpath_overrides_runpath=yes ;; esac else shlibpath_overrides_runpath=yes fi ;; os2*) libname_spec='$name' shrext_cmds=".dll" need_lib_prefix=no library_names_spec='$libname${shared_ext} $libname.a' dynamic_linker='OS/2 ld.exe' shlibpath_var=LIBPATH ;; osf3* | osf4* | osf5*) version_type=osf need_lib_prefix=no need_version=no soname_spec='${libname}${release}${shared_ext}$major' library_names_spec='${libname}${release}${shared_ext}$versuffix ${libname}${release}${shared_ext}$major $libname${shared_ext}' shlibpath_var=LD_LIBRARY_PATH sys_lib_search_path_spec="/usr/shlib /usr/ccs/lib /usr/lib/cmplrs/cc /usr/lib /usr/local/lib /var/shlib" sys_lib_dlsearch_path_spec="$sys_lib_search_path_spec" ;; rdos*) dynamic_linker=no ;; solaris*) version_type=linux # correct to gnu/linux during the next big refactor need_lib_prefix=no need_version=no library_names_spec='${libname}${release}${shared_ext}$versuffix ${libname}${release}${shared_ext}$major $libname${shared_ext}' soname_spec='${libname}${release}${shared_ext}$major' shlibpath_var=LD_LIBRARY_PATH shlibpath_overrides_runpath=yes hardcode_into_libs=yes # ldd complains unless libraries are executable postinstall_cmds='chmod +x $lib' ;; sunos4*) version_type=sunos library_names_spec='${libname}${release}${shared_ext}$versuffix ${libname}${shared_ext}$versuffix' finish_cmds='PATH="\$PATH:/usr/etc" ldconfig $libdir' shlibpath_var=LD_LIBRARY_PATH shlibpath_overrides_runpath=yes if test "$with_gnu_ld" = yes; then need_lib_prefix=no fi need_version=yes ;; sysv4 | sysv4.3*) version_type=linux # correct to gnu/linux during the next big refactor library_names_spec='${libname}${release}${shared_ext}$versuffix ${libname}${release}${shared_ext}$major $libname${shared_ext}' soname_spec='${libname}${release}${shared_ext}$major' shlibpath_var=LD_LIBRARY_PATH case $host_vendor in sni) shlibpath_overrides_runpath=no need_lib_prefix=no runpath_var=LD_RUN_PATH ;; siemens) need_lib_prefix=no ;; motorola) need_lib_prefix=no need_version=no shlibpath_overrides_runpath=no sys_lib_search_path_spec='/lib /usr/lib /usr/ccs/lib' ;; esac ;; sysv4*MP*) if test -d /usr/nec ;then version_type=linux # correct to gnu/linux during the next big refactor library_names_spec='$libname${shared_ext}.$versuffix $libname${shared_ext}.$major $libname${shared_ext}' soname_spec='$libname${shared_ext}.$major' shlibpath_var=LD_LIBRARY_PATH fi ;; sysv5* | sco3.2v5* | sco5v6* | unixware* | OpenUNIX* | sysv4*uw2*) version_type=freebsd-elf need_lib_prefix=no need_version=no library_names_spec='${libname}${release}${shared_ext}$versuffix ${libname}${release}${shared_ext} $libname${shared_ext}' soname_spec='${libname}${release}${shared_ext}$major' shlibpath_var=LD_LIBRARY_PATH shlibpath_overrides_runpath=yes hardcode_into_libs=yes if test "$with_gnu_ld" = yes; then sys_lib_search_path_spec='/usr/local/lib /usr/gnu/lib /usr/ccs/lib /usr/lib /lib' else sys_lib_search_path_spec='/usr/ccs/lib /usr/lib' case $host_os in sco3.2v5*) sys_lib_search_path_spec="$sys_lib_search_path_spec /lib" ;; esac fi sys_lib_dlsearch_path_spec='/usr/lib' ;; tpf*) # TPF is a cross-target only. Preferred cross-host = GNU/Linux. version_type=linux # correct to gnu/linux during the next big refactor need_lib_prefix=no need_version=no library_names_spec='${libname}${release}${shared_ext}$versuffix ${libname}${release}${shared_ext}$major $libname${shared_ext}' shlibpath_var=LD_LIBRARY_PATH shlibpath_overrides_runpath=no hardcode_into_libs=yes ;; uts4*) version_type=linux # correct to gnu/linux during the next big refactor library_names_spec='${libname}${release}${shared_ext}$versuffix ${libname}${release}${shared_ext}$major $libname${shared_ext}' soname_spec='${libname}${release}${shared_ext}$major' shlibpath_var=LD_LIBRARY_PATH ;; *) dynamic_linker=no ;; esac { $as_echo "$as_me:${as_lineno-$LINENO}: result: $dynamic_linker" >&5 $as_echo "$dynamic_linker" >&6; } test "$dynamic_linker" = no && can_build_shared=no variables_saved_for_relink="PATH $shlibpath_var $runpath_var" if test "$GCC" = yes; then variables_saved_for_relink="$variables_saved_for_relink GCC_EXEC_PREFIX COMPILER_PATH LIBRARY_PATH" fi if test "${lt_cv_sys_lib_search_path_spec+set}" = set; then sys_lib_search_path_spec="$lt_cv_sys_lib_search_path_spec" fi if test "${lt_cv_sys_lib_dlsearch_path_spec+set}" = set; then sys_lib_dlsearch_path_spec="$lt_cv_sys_lib_dlsearch_path_spec" fi { $as_echo "$as_me:${as_lineno-$LINENO}: checking how to hardcode library paths into programs" >&5 $as_echo_n "checking how to hardcode library paths into programs... " >&6; } hardcode_action_F77= if test -n "$hardcode_libdir_flag_spec_F77" || test -n "$runpath_var_F77" || test "X$hardcode_automatic_F77" = "Xyes" ; then # We can hardcode non-existent directories. if test "$hardcode_direct_F77" != no && # If the only mechanism to avoid hardcoding is shlibpath_var, we # have to relink, otherwise we might link with an installed library # when we should be linking with a yet-to-be-installed one ## test "$_LT_TAGVAR(hardcode_shlibpath_var, F77)" != no && test "$hardcode_minus_L_F77" != no; then # Linking always hardcodes the temporary library directory. hardcode_action_F77=relink else # We can link without hardcoding, and we can hardcode nonexisting dirs. hardcode_action_F77=immediate fi else # We cannot hardcode anything, or else we can only hardcode existing # directories. hardcode_action_F77=unsupported fi { $as_echo "$as_me:${as_lineno-$LINENO}: result: $hardcode_action_F77" >&5 $as_echo "$hardcode_action_F77" >&6; } if test "$hardcode_action_F77" = relink || test "$inherit_rpath_F77" = yes; then # Fast installation is not supported enable_fast_install=no elif test "$shlibpath_overrides_runpath" = yes || test "$enable_shared" = no; then # Fast installation is not necessary enable_fast_install=needless fi fi # test -n "$compiler" GCC=$lt_save_GCC CC="$lt_save_CC" CFLAGS="$lt_save_CFLAGS" fi # test "$_lt_disable_F77" != yes ac_ext=c ac_cpp='$CPP $CPPFLAGS' ac_compile='$CC -c $CFLAGS $CPPFLAGS conftest.$ac_ext >&5' ac_link='$CC -o conftest$ac_exeext $CFLAGS $CPPFLAGS $LDFLAGS conftest.$ac_ext $LIBS >&5' ac_compiler_gnu=$ac_cv_c_compiler_gnu if test -z "$F77"; then enable_fortran=no { $as_echo "$as_me:${as_lineno-$LINENO}: WARNING: *** Couldn't find f77 compiler; using default Fortran wrappers." >&5 $as_echo "$as_me: WARNING: *** Couldn't find f77 compiler; using default Fortran wrappers." >&2;} else ac_ext=f ac_compile='$F77 -c $FFLAGS conftest.$ac_ext >&5' ac_link='$F77 -o conftest$ac_exeext $FFLAGS $LDFLAGS conftest.$ac_ext $LIBS >&5' ac_compiler_gnu=$ac_cv_f77_compiler_gnu { $as_echo "$as_me:${as_lineno-$LINENO}: checking how to get verbose linking output from $F77" >&5 $as_echo_n "checking how to get verbose linking output from $F77... " >&6; } if ${ac_cv_prog_f77_v+:} false; then : $as_echo_n "(cached) " >&6 else cat > conftest.$ac_ext <<_ACEOF program main end _ACEOF if ac_fn_f77_try_compile "$LINENO"; then : ac_cv_prog_f77_v= # Try some options frequently used verbose output for ac_verb in -v -verbose --verbose -V -\#\#\#; do cat > conftest.$ac_ext <<_ACEOF program main end _ACEOF # Compile and link our simple test program by passing a flag (argument # 1 to this macro) to the Fortran compiler in order to get # "verbose" output that we can then parse for the Fortran linker # flags. ac_save_FFLAGS=$FFLAGS FFLAGS="$FFLAGS $ac_verb" eval "set x $ac_link" shift $as_echo "$as_me:${as_lineno-$LINENO}: $*" >&5 # gfortran 4.3 outputs lines setting COLLECT_GCC_OPTIONS, COMPILER_PATH, # LIBRARY_PATH; skip all such settings. ac_f77_v_output=`eval $ac_link 5>&1 2>&1 | sed '/^Driving:/d; /^Configured with:/d; '"/^[_$as_cr_Letters][_$as_cr_alnum]*=/d"` $as_echo "$ac_f77_v_output" >&5 FFLAGS=$ac_save_FFLAGS rm -rf conftest* # On HP/UX there is a line like: "LPATH is: /foo:/bar:/baz" where # /foo, /bar, and /baz are search directories for the Fortran linker. # Here, we change these into -L/foo -L/bar -L/baz (and put it first): ac_f77_v_output="`echo $ac_f77_v_output | grep 'LPATH is:' | sed 's|.*LPATH is\(: *[^ ]*\).*|\1|;s|: */| -L/|g'` $ac_f77_v_output" # FIXME: we keep getting bitten by quoted arguments; a more general fix # that detects unbalanced quotes in FLIBS should be implemented # and (ugh) tested at some point. case $ac_f77_v_output in # With xlf replace commas with spaces, # and remove "-link" and closing parenthesis. *xlfentry*) ac_f77_v_output=`echo $ac_f77_v_output | sed ' s/,/ /g s/ -link / /g s/) *$// ' ` ;; # With Intel ifc, ignore the quoted -mGLOB_options_string stuff (quoted # $LIBS confuse us, and the libraries appear later in the output anyway). *mGLOB_options_string*) ac_f77_v_output=`echo $ac_f77_v_output | sed 's/"-mGLOB[^"]*"/ /g'` ;; # Portland Group compiler has singly- or doubly-quoted -cmdline argument # Singly-quoted arguments were reported for versions 5.2-4 and 6.0-4. # Doubly-quoted arguments were reported for "PGF90/x86 Linux/x86 5.0-2". *-cmdline\ * | *-ignore\ * | *-def\ *) ac_f77_v_output=`echo $ac_f77_v_output | sed "\ s/-cmdline *'[^']*'/ /g; s/-cmdline *\"[^\"]*\"/ /g s/-ignore *'[^']*'/ /g; s/-ignore *\"[^\"]*\"/ /g s/-def *'[^']*'/ /g; s/-def *\"[^\"]*\"/ /g"` ;; # If we are using fort77 (the f2c wrapper) then filter output and delete quotes. *fort77*f2c*gcc*) ac_f77_v_output=`echo "$ac_f77_v_output" | sed -n ' /:[ ]\+Running[ ]\{1,\}"gcc"/{ /"-c"/d /[.]c"*/d s/^.*"gcc"/"gcc"/ s/"//gp }'` ;; # If we are using Cray Fortran then delete quotes. *cft90*) ac_f77_v_output=`echo $ac_f77_v_output | sed 's/"//g'` ;; esac # look for -l* and *.a constructs in the output for ac_arg in $ac_f77_v_output; do case $ac_arg in [\\/]*.a | ?:[\\/]*.a | -[lLRu]*) ac_cv_prog_f77_v=$ac_verb break 2 ;; esac done done if test -z "$ac_cv_prog_f77_v"; then { $as_echo "$as_me:${as_lineno-$LINENO}: WARNING: cannot determine how to obtain linking information from $F77" >&5 $as_echo "$as_me: WARNING: cannot determine how to obtain linking information from $F77" >&2;} fi else { $as_echo "$as_me:${as_lineno-$LINENO}: WARNING: compilation failed" >&5 $as_echo "$as_me: WARNING: compilation failed" >&2;} fi rm -f core conftest.err conftest.$ac_objext conftest.$ac_ext fi { $as_echo "$as_me:${as_lineno-$LINENO}: result: $ac_cv_prog_f77_v" >&5 $as_echo "$ac_cv_prog_f77_v" >&6; } { $as_echo "$as_me:${as_lineno-$LINENO}: checking for Fortran 77 libraries of $F77" >&5 $as_echo_n "checking for Fortran 77 libraries of $F77... " >&6; } if ${ac_cv_f77_libs+:} false; then : $as_echo_n "(cached) " >&6 else if test "x$FLIBS" != "x"; then ac_cv_f77_libs="$FLIBS" # Let the user override the test. else cat > conftest.$ac_ext <<_ACEOF program main end _ACEOF # Compile and link our simple test program by passing a flag (argument # 1 to this macro) to the Fortran compiler in order to get # "verbose" output that we can then parse for the Fortran linker # flags. ac_save_FFLAGS=$FFLAGS FFLAGS="$FFLAGS $ac_cv_prog_f77_v" eval "set x $ac_link" shift $as_echo "$as_me:${as_lineno-$LINENO}: $*" >&5 # gfortran 4.3 outputs lines setting COLLECT_GCC_OPTIONS, COMPILER_PATH, # LIBRARY_PATH; skip all such settings. ac_f77_v_output=`eval $ac_link 5>&1 2>&1 | sed '/^Driving:/d; /^Configured with:/d; '"/^[_$as_cr_Letters][_$as_cr_alnum]*=/d"` $as_echo "$ac_f77_v_output" >&5 FFLAGS=$ac_save_FFLAGS rm -rf conftest* # On HP/UX there is a line like: "LPATH is: /foo:/bar:/baz" where # /foo, /bar, and /baz are search directories for the Fortran linker. # Here, we change these into -L/foo -L/bar -L/baz (and put it first): ac_f77_v_output="`echo $ac_f77_v_output | grep 'LPATH is:' | sed 's|.*LPATH is\(: *[^ ]*\).*|\1|;s|: */| -L/|g'` $ac_f77_v_output" # FIXME: we keep getting bitten by quoted arguments; a more general fix # that detects unbalanced quotes in FLIBS should be implemented # and (ugh) tested at some point. case $ac_f77_v_output in # With xlf replace commas with spaces, # and remove "-link" and closing parenthesis. *xlfentry*) ac_f77_v_output=`echo $ac_f77_v_output | sed ' s/,/ /g s/ -link / /g s/) *$// ' ` ;; # With Intel ifc, ignore the quoted -mGLOB_options_string stuff (quoted # $LIBS confuse us, and the libraries appear later in the output anyway). *mGLOB_options_string*) ac_f77_v_output=`echo $ac_f77_v_output | sed 's/"-mGLOB[^"]*"/ /g'` ;; # Portland Group compiler has singly- or doubly-quoted -cmdline argument # Singly-quoted arguments were reported for versions 5.2-4 and 6.0-4. # Doubly-quoted arguments were reported for "PGF90/x86 Linux/x86 5.0-2". *-cmdline\ * | *-ignore\ * | *-def\ *) ac_f77_v_output=`echo $ac_f77_v_output | sed "\ s/-cmdline *'[^']*'/ /g; s/-cmdline *\"[^\"]*\"/ /g s/-ignore *'[^']*'/ /g; s/-ignore *\"[^\"]*\"/ /g s/-def *'[^']*'/ /g; s/-def *\"[^\"]*\"/ /g"` ;; # If we are using fort77 (the f2c wrapper) then filter output and delete quotes. *fort77*f2c*gcc*) ac_f77_v_output=`echo "$ac_f77_v_output" | sed -n ' /:[ ]\+Running[ ]\{1,\}"gcc"/{ /"-c"/d /[.]c"*/d s/^.*"gcc"/"gcc"/ s/"//gp }'` ;; # If we are using Cray Fortran then delete quotes. *cft90*) ac_f77_v_output=`echo $ac_f77_v_output | sed 's/"//g'` ;; esac ac_cv_f77_libs= # Save positional arguments (if any) ac_save_positional="$@" set X $ac_f77_v_output while test $# != 1; do shift ac_arg=$1 case $ac_arg in [\\/]*.a | ?:[\\/]*.a) ac_exists=false for ac_i in $ac_cv_f77_libs; do if test x"$ac_arg" = x"$ac_i"; then ac_exists=true break fi done if test x"$ac_exists" = xtrue; then : else ac_cv_f77_libs="$ac_cv_f77_libs $ac_arg" fi ;; -bI:*) ac_exists=false for ac_i in $ac_cv_f77_libs; do if test x"$ac_arg" = x"$ac_i"; then ac_exists=true break fi done if test x"$ac_exists" = xtrue; then : else if test "$ac_compiler_gnu" = yes; then for ac_link_opt in $ac_arg; do ac_cv_f77_libs="$ac_cv_f77_libs -Xlinker $ac_link_opt" done else ac_cv_f77_libs="$ac_cv_f77_libs $ac_arg" fi fi ;; # Ignore these flags. -lang* | -lcrt*.o | -lc | -lgcc* | -lSystem | -libmil | -little \ |-LANG:=* | -LIST:* | -LNO:* | -link) ;; -lkernel32) case $host_os in *cygwin*) ;; *) ac_cv_f77_libs="$ac_cv_f77_libs $ac_arg" ;; esac ;; -[LRuYz]) # These flags, when seen by themselves, take an argument. # We remove the space between option and argument and re-iterate # unless we find an empty arg or a new option (starting with -) case $2 in "" | -*);; *) ac_arg="$ac_arg$2" shift; shift set X $ac_arg "$@" ;; esac ;; -YP,*) for ac_j in `$as_echo "$ac_arg" | sed -e 's/-YP,/-L/;s/:/ -L/g'`; do ac_exists=false for ac_i in $ac_cv_f77_libs; do if test x"$ac_j" = x"$ac_i"; then ac_exists=true break fi done if test x"$ac_exists" = xtrue; then : else ac_arg="$ac_arg $ac_j" ac_cv_f77_libs="$ac_cv_f77_libs $ac_j" fi done ;; -[lLR]*) ac_exists=false for ac_i in $ac_cv_f77_libs; do if test x"$ac_arg" = x"$ac_i"; then ac_exists=true break fi done if test x"$ac_exists" = xtrue; then : else ac_cv_f77_libs="$ac_cv_f77_libs $ac_arg" fi ;; -zallextract*| -zdefaultextract) ac_cv_f77_libs="$ac_cv_f77_libs $ac_arg" ;; # Ignore everything else. esac done # restore positional arguments set X $ac_save_positional; shift # We only consider "LD_RUN_PATH" on Solaris systems. If this is seen, # then we insist that the "run path" must be an absolute path (i.e. it # must begin with a "/"). case `(uname -sr) 2>/dev/null` in "SunOS 5"*) ac_ld_run_path=`$as_echo "$ac_f77_v_output" | sed -n 's,^.*LD_RUN_PATH *= *\(/[^ ]*\).*$,-R\1,p'` test "x$ac_ld_run_path" != x && if test "$ac_compiler_gnu" = yes; then for ac_link_opt in $ac_ld_run_path; do ac_cv_f77_libs="$ac_cv_f77_libs -Xlinker $ac_link_opt" done else ac_cv_f77_libs="$ac_cv_f77_libs $ac_ld_run_path" fi ;; esac fi # test "x$[]_AC_LANG_PREFIX[]LIBS" = "x" fi { $as_echo "$as_me:${as_lineno-$LINENO}: result: $ac_cv_f77_libs" >&5 $as_echo "$ac_cv_f77_libs" >&6; } FLIBS="$ac_cv_f77_libs" ac_ext=c ac_cpp='$CPP $CPPFLAGS' ac_compile='$CC -c $CFLAGS $CPPFLAGS conftest.$ac_ext >&5' ac_link='$CC -o conftest$ac_exeext $CFLAGS $CPPFLAGS $LDFLAGS conftest.$ac_ext $LIBS >&5' ac_compiler_gnu=$ac_cv_c_compiler_gnu ac_ext=f ac_compile='$F77 -c $FFLAGS conftest.$ac_ext >&5' ac_link='$F77 -o conftest$ac_exeext $FFLAGS $LDFLAGS conftest.$ac_ext $LIBS >&5' ac_compiler_gnu=$ac_cv_f77_compiler_gnu { $as_echo "$as_me:${as_lineno-$LINENO}: checking for dummy main to link with Fortran 77 libraries" >&5 $as_echo_n "checking for dummy main to link with Fortran 77 libraries... " >&6; } if ${ac_cv_f77_dummy_main+:} false; then : $as_echo_n "(cached) " >&6 else ac_f77_dm_save_LIBS=$LIBS LIBS="$LIBS $FLIBS" ac_fortran_dm_var=F77_DUMMY_MAIN ac_ext=c ac_cpp='$CPP $CPPFLAGS' ac_compile='$CC -c $CFLAGS $CPPFLAGS conftest.$ac_ext >&5' ac_link='$CC -o conftest$ac_exeext $CFLAGS $CPPFLAGS $LDFLAGS conftest.$ac_ext $LIBS >&5' ac_compiler_gnu=$ac_cv_c_compiler_gnu # First, try linking without a dummy main: cat confdefs.h - <<_ACEOF >conftest.$ac_ext /* end confdefs.h. */ #ifdef F77_DUMMY_MAIN # ifdef __cplusplus extern "C" # endif int F77_DUMMY_MAIN() { return 1; } #endif int main () { ; return 0; } _ACEOF if ac_fn_c_try_link "$LINENO"; then : ac_cv_fortran_dummy_main=none else ac_cv_fortran_dummy_main=unknown fi rm -f core conftest.err conftest.$ac_objext \ conftest$ac_exeext conftest.$ac_ext if test $ac_cv_fortran_dummy_main = unknown; then for ac_func in MAIN__ MAIN_ __main MAIN _MAIN __MAIN main_ main__ _main; do cat confdefs.h - <<_ACEOF >conftest.$ac_ext /* end confdefs.h. */ #define $ac_fortran_dm_var $ac_func #ifdef F77_DUMMY_MAIN # ifdef __cplusplus extern "C" # endif int F77_DUMMY_MAIN() { return 1; } #endif int main () { ; return 0; } _ACEOF if ac_fn_c_try_link "$LINENO"; then : ac_cv_fortran_dummy_main=$ac_func; break fi rm -f core conftest.err conftest.$ac_objext \ conftest$ac_exeext conftest.$ac_ext done fi ac_ext=f ac_compile='$F77 -c $FFLAGS conftest.$ac_ext >&5' ac_link='$F77 -o conftest$ac_exeext $FFLAGS $LDFLAGS conftest.$ac_ext $LIBS >&5' ac_compiler_gnu=$ac_cv_f77_compiler_gnu ac_cv_f77_dummy_main=$ac_cv_fortran_dummy_main rm -rf conftest* LIBS=$ac_f77_dm_save_LIBS fi { $as_echo "$as_me:${as_lineno-$LINENO}: result: $ac_cv_f77_dummy_main" >&5 $as_echo "$ac_cv_f77_dummy_main" >&6; } F77_DUMMY_MAIN=$ac_cv_f77_dummy_main if test "$F77_DUMMY_MAIN" != unknown; then : if test $F77_DUMMY_MAIN != none; then cat >>confdefs.h <<_ACEOF #define F77_DUMMY_MAIN $F77_DUMMY_MAIN _ACEOF if test "x$ac_cv_fc_dummy_main" = "x$ac_cv_f77_dummy_main"; then $as_echo "#define FC_DUMMY_MAIN_EQ_F77 1" >>confdefs.h fi fi else enable_fortran=no { $as_echo "$as_me:${as_lineno-$LINENO}: WARNING: *** Couldn't figure out how to link C and Fortran; using default Fortran wrappers." >&5 $as_echo "$as_me: WARNING: *** Couldn't figure out how to link C and Fortran; using default Fortran wrappers." >&2;} fi ac_ext=c ac_cpp='$CPP $CPPFLAGS' ac_compile='$CC -c $CFLAGS $CPPFLAGS conftest.$ac_ext >&5' ac_link='$CC -o conftest$ac_exeext $CFLAGS $CPPFLAGS $LDFLAGS conftest.$ac_ext $LIBS >&5' ac_compiler_gnu=$ac_cv_c_compiler_gnu fi else $as_echo "#define DISABLE_FORTRAN 1" >>confdefs.h fi if test "x$enable_fortran" = xyes; then ac_ext=f ac_compile='$F77 -c $FFLAGS conftest.$ac_ext >&5' ac_link='$F77 -o conftest$ac_exeext $FFLAGS $LDFLAGS conftest.$ac_ext $LIBS >&5' ac_compiler_gnu=$ac_cv_f77_compiler_gnu { $as_echo "$as_me:${as_lineno-$LINENO}: checking for Fortran 77 name-mangling scheme" >&5 $as_echo_n "checking for Fortran 77 name-mangling scheme... " >&6; } if ${ac_cv_f77_mangling+:} false; then : $as_echo_n "(cached) " >&6 else cat > conftest.$ac_ext <<_ACEOF subroutine foobar() return end subroutine foo_bar() return end _ACEOF if ac_fn_f77_try_compile "$LINENO"; then : mv conftest.$ac_objext cfortran_test.$ac_objext ac_save_LIBS=$LIBS LIBS="cfortran_test.$ac_objext $LIBS $FLIBS" ac_ext=c ac_cpp='$CPP $CPPFLAGS' ac_compile='$CC -c $CFLAGS $CPPFLAGS conftest.$ac_ext >&5' ac_link='$CC -o conftest$ac_exeext $CFLAGS $CPPFLAGS $LDFLAGS conftest.$ac_ext $LIBS >&5' ac_compiler_gnu=$ac_cv_c_compiler_gnu ac_success=no for ac_foobar in foobar FOOBAR; do for ac_underscore in "" "_"; do ac_func="$ac_foobar$ac_underscore" cat confdefs.h - <<_ACEOF >conftest.$ac_ext /* end confdefs.h. */ /* Override any GCC internal prototype to avoid an error. Use char because int might match the return type of a GCC builtin and then its argument prototype would still apply. */ #ifdef __cplusplus extern "C" #endif char $ac_func (); #ifdef F77_DUMMY_MAIN # ifdef __cplusplus extern "C" # endif int F77_DUMMY_MAIN() { return 1; } #endif int main () { return $ac_func (); ; return 0; } _ACEOF if ac_fn_c_try_link "$LINENO"; then : ac_success=yes; break 2 fi rm -f core conftest.err conftest.$ac_objext \ conftest$ac_exeext conftest.$ac_ext done done ac_ext=f ac_compile='$F77 -c $FFLAGS conftest.$ac_ext >&5' ac_link='$F77 -o conftest$ac_exeext $FFLAGS $LDFLAGS conftest.$ac_ext $LIBS >&5' ac_compiler_gnu=$ac_cv_f77_compiler_gnu if test "$ac_success" = "yes"; then case $ac_foobar in foobar) ac_case=lower ac_foo_bar=foo_bar ;; FOOBAR) ac_case=upper ac_foo_bar=FOO_BAR ;; esac ac_ext=c ac_cpp='$CPP $CPPFLAGS' ac_compile='$CC -c $CFLAGS $CPPFLAGS conftest.$ac_ext >&5' ac_link='$CC -o conftest$ac_exeext $CFLAGS $CPPFLAGS $LDFLAGS conftest.$ac_ext $LIBS >&5' ac_compiler_gnu=$ac_cv_c_compiler_gnu ac_success_extra=no for ac_extra in "" "_"; do ac_func="$ac_foo_bar$ac_underscore$ac_extra" cat confdefs.h - <<_ACEOF >conftest.$ac_ext /* end confdefs.h. */ /* Override any GCC internal prototype to avoid an error. Use char because int might match the return type of a GCC builtin and then its argument prototype would still apply. */ #ifdef __cplusplus extern "C" #endif char $ac_func (); #ifdef F77_DUMMY_MAIN # ifdef __cplusplus extern "C" # endif int F77_DUMMY_MAIN() { return 1; } #endif int main () { return $ac_func (); ; return 0; } _ACEOF if ac_fn_c_try_link "$LINENO"; then : ac_success_extra=yes; break fi rm -f core conftest.err conftest.$ac_objext \ conftest$ac_exeext conftest.$ac_ext done ac_ext=f ac_compile='$F77 -c $FFLAGS conftest.$ac_ext >&5' ac_link='$F77 -o conftest$ac_exeext $FFLAGS $LDFLAGS conftest.$ac_ext $LIBS >&5' ac_compiler_gnu=$ac_cv_f77_compiler_gnu if test "$ac_success_extra" = "yes"; then ac_cv_f77_mangling="$ac_case case" if test -z "$ac_underscore"; then ac_cv_f77_mangling="$ac_cv_f77_mangling, no underscore" else ac_cv_f77_mangling="$ac_cv_f77_mangling, underscore" fi if test -z "$ac_extra"; then ac_cv_f77_mangling="$ac_cv_f77_mangling, no extra underscore" else ac_cv_f77_mangling="$ac_cv_f77_mangling, extra underscore" fi else ac_cv_f77_mangling="unknown" fi else ac_cv_f77_mangling="unknown" fi LIBS=$ac_save_LIBS rm -rf conftest* rm -f cfortran_test* else { { $as_echo "$as_me:${as_lineno-$LINENO}: error: in \`$ac_pwd':" >&5 $as_echo "$as_me: error: in \`$ac_pwd':" >&2;} as_fn_error $? "cannot compile a simple Fortran program See \`config.log' for more details" "$LINENO" 5; } fi rm -f core conftest.err conftest.$ac_objext conftest.$ac_ext fi { $as_echo "$as_me:${as_lineno-$LINENO}: result: $ac_cv_f77_mangling" >&5 $as_echo "$ac_cv_f77_mangling" >&6; } ac_ext=c ac_cpp='$CPP $CPPFLAGS' ac_compile='$CC -c $CFLAGS $CPPFLAGS conftest.$ac_ext >&5' ac_link='$CC -o conftest$ac_exeext $CFLAGS $CPPFLAGS $LDFLAGS conftest.$ac_ext $LIBS >&5' ac_compiler_gnu=$ac_cv_c_compiler_gnu ac_ext=f ac_compile='$F77 -c $FFLAGS conftest.$ac_ext >&5' ac_link='$F77 -o conftest$ac_exeext $FFLAGS $LDFLAGS conftest.$ac_ext $LIBS >&5' ac_compiler_gnu=$ac_cv_f77_compiler_gnu case $ac_cv_f77_mangling in "lower case, no underscore, no extra underscore") $as_echo "#define F77_FUNC(name,NAME) name" >>confdefs.h $as_echo "#define F77_FUNC_(name,NAME) name" >>confdefs.h ;; "lower case, no underscore, extra underscore") $as_echo "#define F77_FUNC(name,NAME) name" >>confdefs.h $as_echo "#define F77_FUNC_(name,NAME) name ## _" >>confdefs.h ;; "lower case, underscore, no extra underscore") $as_echo "#define F77_FUNC(name,NAME) name ## _" >>confdefs.h $as_echo "#define F77_FUNC_(name,NAME) name ## _" >>confdefs.h ;; "lower case, underscore, extra underscore") $as_echo "#define F77_FUNC(name,NAME) name ## _" >>confdefs.h $as_echo "#define F77_FUNC_(name,NAME) name ## __" >>confdefs.h ;; "upper case, no underscore, no extra underscore") $as_echo "#define F77_FUNC(name,NAME) NAME" >>confdefs.h $as_echo "#define F77_FUNC_(name,NAME) NAME" >>confdefs.h ;; "upper case, no underscore, extra underscore") $as_echo "#define F77_FUNC(name,NAME) NAME" >>confdefs.h $as_echo "#define F77_FUNC_(name,NAME) NAME ## _" >>confdefs.h ;; "upper case, underscore, no extra underscore") $as_echo "#define F77_FUNC(name,NAME) NAME ## _" >>confdefs.h $as_echo "#define F77_FUNC_(name,NAME) NAME ## _" >>confdefs.h ;; "upper case, underscore, extra underscore") $as_echo "#define F77_FUNC(name,NAME) NAME ## _" >>confdefs.h $as_echo "#define F77_FUNC_(name,NAME) NAME ## __" >>confdefs.h ;; *) { $as_echo "$as_me:${as_lineno-$LINENO}: WARNING: unknown Fortran name-mangling scheme" >&5 $as_echo "$as_me: WARNING: unknown Fortran name-mangling scheme" >&2;} ;; esac ac_ext=c ac_cpp='$CPP $CPPFLAGS' ac_compile='$CC -c $CFLAGS $CPPFLAGS conftest.$ac_ext >&5' ac_link='$CC -o conftest$ac_exeext $CFLAGS $CPPFLAGS $LDFLAGS conftest.$ac_ext $LIBS >&5' ac_compiler_gnu=$ac_cv_c_compiler_gnu ac_ext=f ac_compile='$F77 -c $FFLAGS conftest.$ac_ext >&5' ac_link='$F77 -o conftest$ac_exeext $FFLAGS $LDFLAGS conftest.$ac_ext $LIBS >&5' ac_compiler_gnu=$ac_cv_f77_compiler_gnu case $ac_cv_f77_mangling in upper*) ac_val="F77FOO" ;; lower*) ac_val="f77foo" ;; *) ac_val="unknown" ;; esac case $ac_cv_f77_mangling in *," underscore"*) ac_val="$ac_val"_ ;; esac f77foo="$ac_val" ac_ext=c ac_cpp='$CPP $CPPFLAGS' ac_compile='$CC -c $CFLAGS $CPPFLAGS conftest.$ac_ext >&5' ac_link='$CC -o conftest$ac_exeext $CFLAGS $CPPFLAGS $LDFLAGS conftest.$ac_ext $LIBS >&5' ac_compiler_gnu=$ac_cv_c_compiler_gnu ac_ext=f ac_compile='$F77 -c $FFLAGS conftest.$ac_ext >&5' ac_link='$F77 -o conftest$ac_exeext $FFLAGS $LDFLAGS conftest.$ac_ext $LIBS >&5' ac_compiler_gnu=$ac_cv_f77_compiler_gnu case $ac_cv_f77_mangling in upper*) ac_val="F77_FOO" ;; lower*) ac_val="f77_foo" ;; *) ac_val="unknown" ;; esac case $ac_cv_f77_mangling in *," underscore"*) ac_val="$ac_val"_ ;; esac case $ac_cv_f77_mangling in *," extra underscore"*) ac_val="$ac_val"_ ;; esac f77_foo="$ac_val" ac_ext=c ac_cpp='$CPP $CPPFLAGS' ac_compile='$CC -c $CFLAGS $CPPFLAGS conftest.$ac_ext >&5' ac_link='$CC -o conftest$ac_exeext $CFLAGS $CPPFLAGS $LDFLAGS conftest.$ac_ext $LIBS >&5' ac_compiler_gnu=$ac_cv_c_compiler_gnu f77_foo2=`echo $f77foo | sed 's/77/77_/'` if test "$f77_foo" = "$f77_foo2"; then $as_echo "#define F77_FUNC_EQUIV 1" >>confdefs.h # Include g77 wrappers by default for GNU systems or gfortran with_g77_wrappers=$ac_cv_f77_compiler_gnu case $host_os in *gnu*) with_g77_wrappers=yes ;; esac fi else with_g77_wrappers=no fi # Check whether --with-g77-wrappers was given. if test "${with_g77_wrappers+set}" = set; then : withval=$with_g77_wrappers; with_g77_wrappers=$withval fi if test "x$with_g77_wrappers" = "xyes"; then $as_echo "#define WITH_G77_WRAPPERS 1" >>confdefs.h fi have_smp="no" # Check whether --enable-openmp was given. if test "${enable_openmp+set}" = set; then : enableval=$enable_openmp; enable_openmp=$enableval else enable_openmp=no fi if test "$enable_openmp" = "yes"; then $as_echo "#define HAVE_OPENMP 1" >>confdefs.h { $as_echo "$as_me:${as_lineno-$LINENO}: checking for OpenMP flag of C compiler" >&5 $as_echo_n "checking for OpenMP flag of C compiler... " >&6; } if ${ax_cv_c_openmp+:} false; then : $as_echo_n "(cached) " >&6 else saveCFLAGS=$CFLAGS ax_cv_c_openmp=unknown # Flags to try: -fopenmp (gcc), -openmp (icc), -mp (SGI & PGI), # -xopenmp (Sun), -omp (Tru64), -qsmp=omp (AIX), none ax_openmp_flags="-fopenmp -openmp -mp -xopenmp -omp -qsmp=omp none" if test "x$OPENMP_CFLAGS" != x; then ax_openmp_flags="$OPENMP_CFLAGS $ax_openmp_flags" fi for ax_openmp_flag in $ax_openmp_flags; do case $ax_openmp_flag in none) CFLAGS=$saveC ;; *) CFLAGS="$saveCFLAGS $ax_openmp_flag" ;; esac cat confdefs.h - <<_ACEOF >conftest.$ac_ext /* end confdefs.h. */ /* Override any GCC internal prototype to avoid an error. Use char because int might match the return type of a GCC builtin and then its argument prototype would still apply. */ #ifdef __cplusplus extern "C" #endif char omp_set_num_threads (); #ifdef F77_DUMMY_MAIN # ifdef __cplusplus extern "C" # endif int F77_DUMMY_MAIN() { return 1; } #endif int main () { return omp_set_num_threads (); ; return 0; } _ACEOF if ac_fn_c_try_link "$LINENO"; then : ax_cv_c_openmp=$ax_openmp_flag; break fi rm -f core conftest.err conftest.$ac_objext \ conftest$ac_exeext conftest.$ac_ext done CFLAGS=$saveCFLAGS fi { $as_echo "$as_me:${as_lineno-$LINENO}: result: $ax_cv_c_openmp" >&5 $as_echo "$ax_cv_c_openmp" >&6; } if test "x$ax_cv_c_openmp" = "xunknown"; then as_fn_error $? "don't know how to enable OpenMP" "$LINENO" 5 else if test "x$ax_cv_c_openmp" != "xnone"; then OPENMP_CFLAGS=$ax_cv_c_openmp fi $as_echo "#define HAVE_OPENMP 1" >>confdefs.h fi fi # Check whether --enable-threads was given. if test "${enable_threads+set}" = set; then : enableval=$enable_threads; enable_threads=$enableval else enable_threads=no fi if test "$enable_threads" = "yes"; then $as_echo "#define HAVE_THREADS 1" >>confdefs.h fi # Check whether --with-combined-threads was given. if test "${with_combined_threads+set}" = set; then : withval=$with_combined_threads; with_combined_threads=$withval else with_combined_threads=no fi if test "$with_combined_threads" = yes; then if test "$enable_openmp" = "yes"; then as_fn_error $? "--with-combined-threads incompatible with --enable-openmp" "$LINENO" 5 fi if test "$enable_threads" != "yes"; then as_fn_error $? "--with-combined-threads requires --enable-threads" "$LINENO" 5 fi fi THREADLIBS="" if test "$enable_threads" = "yes"; then # POSIX threads, the default choice: if test -z "$THREADLIBS"; then ac_ext=c ac_cpp='$CPP $CPPFLAGS' ac_compile='$CC -c $CFLAGS $CPPFLAGS conftest.$ac_ext >&5' ac_link='$CC -o conftest$ac_exeext $CFLAGS $CPPFLAGS $LDFLAGS conftest.$ac_ext $LIBS >&5' ac_compiler_gnu=$ac_cv_c_compiler_gnu acx_pthread_ok=no # We used to check for pthread.h first, but this fails if pthread.h # requires special compiler flags (e.g. on True64 or Sequent). # It gets checked for in the link test anyway. # First of all, check if the user has set any of the PTHREAD_LIBS, # etcetera environment variables, and if threads linking works using # them: if test x"$PTHREAD_LIBS$PTHREAD_CFLAGS" != x; then save_CFLAGS="$CFLAGS" CFLAGS="$CFLAGS $PTHREAD_CFLAGS" save_LIBS="$LIBS" LIBS="$PTHREAD_LIBS $LIBS" { $as_echo "$as_me:${as_lineno-$LINENO}: checking for pthread_join in LIBS=$PTHREAD_LIBS with CFLAGS=$PTHREAD_CFLAGS" >&5 $as_echo_n "checking for pthread_join in LIBS=$PTHREAD_LIBS with CFLAGS=$PTHREAD_CFLAGS... " >&6; } cat confdefs.h - <<_ACEOF >conftest.$ac_ext /* end confdefs.h. */ /* Override any GCC internal prototype to avoid an error. Use char because int might match the return type of a GCC builtin and then its argument prototype would still apply. */ #ifdef __cplusplus extern "C" #endif char pthread_join (); #ifdef F77_DUMMY_MAIN # ifdef __cplusplus extern "C" # endif int F77_DUMMY_MAIN() { return 1; } #endif int main () { return pthread_join (); ; return 0; } _ACEOF if ac_fn_c_try_link "$LINENO"; then : acx_pthread_ok=yes fi rm -f core conftest.err conftest.$ac_objext \ conftest$ac_exeext conftest.$ac_ext { $as_echo "$as_me:${as_lineno-$LINENO}: result: $acx_pthread_ok" >&5 $as_echo "$acx_pthread_ok" >&6; } if test x"$acx_pthread_ok" = xno; then PTHREAD_LIBS="" PTHREAD_CFLAGS="" fi LIBS="$save_LIBS" CFLAGS="$save_CFLAGS" fi # We must check for the threads library under a number of different # names; the ordering is very important because some systems # (e.g. DEC) have both -lpthread and -lpthreads, where one of the # libraries is broken (non-POSIX). # Create a list of thread flags to try. Items starting with a "-" are # C compiler flags, and other items are library names, except for "none" # which indicates that we try without any flags at all, and "pthread-config" # which is a program returning the flags for the Pth emulation library. acx_pthread_flags="pthreads none -Kthread -kthread lthread -pthread -pthreads -mt -mthreads pthread --thread-safe pthread-config" # The ordering *is* (sometimes) important. Some notes on the # individual items follow: # pthreads: AIX (must check this before -lpthread) # none: in case threads are in libc; should be tried before -Kthread and # other compiler flags to prevent continual compiler warnings # -Kthread: Sequent (threads in libc, but -Kthread needed for pthread.h) # -kthread: FreeBSD kernel threads (preferred to -pthread since SMP-able) # lthread: LinuxThreads port on FreeBSD (also preferred to -pthread) # -pthread: Linux/gcc (kernel threads), BSD/gcc (userland threads) # -pthreads: Solaris/gcc # -mthreads: Mingw32/gcc, Lynx/gcc # -mt: Sun Workshop C (may only link SunOS threads [-lthread], but it # doesn't hurt to check since this sometimes defines pthreads too; # also defines -D_REENTRANT) # ... -mt is also the pthreads flag for HP/aCC # (where it should come before -mthreads to avoid spurious warnings) # pthread: Linux, etcetera # --thread-safe: KAI C++ # pthread-config: use pthread-config program (for GNU Pth library) case "${host_cpu}-${host_os}" in *solaris*) # On Solaris (at least, for some versions), libc contains stubbed # (non-functional) versions of the pthreads routines, so link-based # tests will erroneously succeed. (We need to link with -pthreads/-mt/ # -lpthread.) (The stubs are missing pthread_cleanup_push, or rather # a function called by this macro, so we could check for that, but # who knows whether they'll stub that too in a future libc.) So, # we'll just look for -pthreads and -lpthread first: acx_pthread_flags="-pthreads pthread -mt -pthread $acx_pthread_flags" ;; esac if test x"$acx_pthread_ok" = xno; then for flag in $acx_pthread_flags; do case $flag in none) { $as_echo "$as_me:${as_lineno-$LINENO}: checking whether pthreads work without any flags" >&5 $as_echo_n "checking whether pthreads work without any flags... " >&6; } ;; -*) { $as_echo "$as_me:${as_lineno-$LINENO}: checking whether pthreads work with $flag" >&5 $as_echo_n "checking whether pthreads work with $flag... " >&6; } PTHREAD_CFLAGS="$flag" ;; pthread-config) # Extract the first word of "pthread-config", so it can be a program name with args. set dummy pthread-config; ac_word=$2 { $as_echo "$as_me:${as_lineno-$LINENO}: checking for $ac_word" >&5 $as_echo_n "checking for $ac_word... " >&6; } if ${ac_cv_prog_acx_pthread_config+:} false; then : $as_echo_n "(cached) " >&6 else if test -n "$acx_pthread_config"; then ac_cv_prog_acx_pthread_config="$acx_pthread_config" # Let the user override the test. else as_save_IFS=$IFS; IFS=$PATH_SEPARATOR for as_dir in $PATH do IFS=$as_save_IFS test -z "$as_dir" && as_dir=. for ac_exec_ext in '' $ac_executable_extensions; do if as_fn_executable_p "$as_dir/$ac_word$ac_exec_ext"; then ac_cv_prog_acx_pthread_config="yes" $as_echo "$as_me:${as_lineno-$LINENO}: found $as_dir/$ac_word$ac_exec_ext" >&5 break 2 fi done done IFS=$as_save_IFS test -z "$ac_cv_prog_acx_pthread_config" && ac_cv_prog_acx_pthread_config="no" fi fi acx_pthread_config=$ac_cv_prog_acx_pthread_config if test -n "$acx_pthread_config"; then { $as_echo "$as_me:${as_lineno-$LINENO}: result: $acx_pthread_config" >&5 $as_echo "$acx_pthread_config" >&6; } else { $as_echo "$as_me:${as_lineno-$LINENO}: result: no" >&5 $as_echo "no" >&6; } fi if test x"$acx_pthread_config" = xno; then continue; fi PTHREAD_CFLAGS="`pthread-config --cflags`" PTHREAD_LIBS="`pthread-config --ldflags` `pthread-config --libs`" ;; *) { $as_echo "$as_me:${as_lineno-$LINENO}: checking for the pthreads library -l$flag" >&5 $as_echo_n "checking for the pthreads library -l$flag... " >&6; } PTHREAD_LIBS="-l$flag" ;; esac save_LIBS="$LIBS" save_CFLAGS="$CFLAGS" LIBS="$PTHREAD_LIBS $LIBS" CFLAGS="$CFLAGS $PTHREAD_CFLAGS" # Check for various functions. We must include pthread.h, # since some functions may be macros. (On the Sequent, we # need a special flag -Kthread to make this header compile.) # We check for pthread_join because it is in -lpthread on IRIX # while pthread_create is in libc. We check for pthread_attr_init # due to DEC craziness with -lpthreads. We check for # pthread_cleanup_push because it is one of the few pthread # functions on Solaris that doesn't have a non-functional libc stub. # We try pthread_create on general principles. cat confdefs.h - <<_ACEOF >conftest.$ac_ext /* end confdefs.h. */ #include #ifdef F77_DUMMY_MAIN # ifdef __cplusplus extern "C" # endif int F77_DUMMY_MAIN() { return 1; } #endif int main () { pthread_t th; pthread_join(th, (void**) 0); pthread_attr_init((pthread_attr_t*) 0); pthread_cleanup_push((void(*)(void *)) 0, (void*) 0); pthread_create((pthread_t*) 0, (pthread_attr_t*) 0, (void*(*)(void *)) 0, (void*) 0); pthread_cleanup_pop(0); ; return 0; } _ACEOF if ac_fn_c_try_link "$LINENO"; then : acx_pthread_ok=yes fi rm -f core conftest.err conftest.$ac_objext \ conftest$ac_exeext conftest.$ac_ext LIBS="$save_LIBS" CFLAGS="$save_CFLAGS" { $as_echo "$as_me:${as_lineno-$LINENO}: result: $acx_pthread_ok" >&5 $as_echo "$acx_pthread_ok" >&6; } if test "x$acx_pthread_ok" = xyes; then break; fi PTHREAD_LIBS="" PTHREAD_CFLAGS="" done fi # Various other checks: if test "x$acx_pthread_ok" = xyes; then save_LIBS="$LIBS" LIBS="$PTHREAD_LIBS $LIBS" save_CFLAGS="$CFLAGS" CFLAGS="$CFLAGS $PTHREAD_CFLAGS" # Detect AIX lossage: JOINABLE attribute is called UNDETACHED. { $as_echo "$as_me:${as_lineno-$LINENO}: checking for joinable pthread attribute" >&5 $as_echo_n "checking for joinable pthread attribute... " >&6; } attr_name=unknown for attr in PTHREAD_CREATE_JOINABLE PTHREAD_CREATE_UNDETACHED; do cat confdefs.h - <<_ACEOF >conftest.$ac_ext /* end confdefs.h. */ #include #ifdef F77_DUMMY_MAIN # ifdef __cplusplus extern "C" # endif int F77_DUMMY_MAIN() { return 1; } #endif int main () { int attr=$attr; return attr; ; return 0; } _ACEOF if ac_fn_c_try_link "$LINENO"; then : attr_name=$attr; break fi rm -f core conftest.err conftest.$ac_objext \ conftest$ac_exeext conftest.$ac_ext done { $as_echo "$as_me:${as_lineno-$LINENO}: result: $attr_name" >&5 $as_echo "$attr_name" >&6; } if test "$attr_name" != PTHREAD_CREATE_JOINABLE; then cat >>confdefs.h <<_ACEOF #define PTHREAD_CREATE_JOINABLE $attr_name _ACEOF fi { $as_echo "$as_me:${as_lineno-$LINENO}: checking if more special flags are required for pthreads" >&5 $as_echo_n "checking if more special flags are required for pthreads... " >&6; } flag=no case "${host_cpu}-${host_os}" in *-aix* | *-freebsd* | *-darwin*) flag="-D_THREAD_SAFE";; *solaris* | *-osf* | *-hpux*) flag="-D_REENTRANT";; esac { $as_echo "$as_me:${as_lineno-$LINENO}: result: ${flag}" >&5 $as_echo "${flag}" >&6; } if test "x$flag" != xno; then PTHREAD_CFLAGS="$flag $PTHREAD_CFLAGS" fi LIBS="$save_LIBS" CFLAGS="$save_CFLAGS" # More AIX lossage: must compile with xlc_r or cc_r if test x"$GCC" != xyes; then for ac_prog in xlc_r cc_r do # Extract the first word of "$ac_prog", so it can be a program name with args. set dummy $ac_prog; ac_word=$2 { $as_echo "$as_me:${as_lineno-$LINENO}: checking for $ac_word" >&5 $as_echo_n "checking for $ac_word... " >&6; } if ${ac_cv_prog_PTHREAD_CC+:} false; then : $as_echo_n "(cached) " >&6 else if test -n "$PTHREAD_CC"; then ac_cv_prog_PTHREAD_CC="$PTHREAD_CC" # Let the user override the test. else as_save_IFS=$IFS; IFS=$PATH_SEPARATOR for as_dir in $PATH do IFS=$as_save_IFS test -z "$as_dir" && as_dir=. for ac_exec_ext in '' $ac_executable_extensions; do if as_fn_executable_p "$as_dir/$ac_word$ac_exec_ext"; then ac_cv_prog_PTHREAD_CC="$ac_prog" $as_echo "$as_me:${as_lineno-$LINENO}: found $as_dir/$ac_word$ac_exec_ext" >&5 break 2 fi done done IFS=$as_save_IFS fi fi PTHREAD_CC=$ac_cv_prog_PTHREAD_CC if test -n "$PTHREAD_CC"; then { $as_echo "$as_me:${as_lineno-$LINENO}: result: $PTHREAD_CC" >&5 $as_echo "$PTHREAD_CC" >&6; } else { $as_echo "$as_me:${as_lineno-$LINENO}: result: no" >&5 $as_echo "no" >&6; } fi test -n "$PTHREAD_CC" && break done test -n "$PTHREAD_CC" || PTHREAD_CC="${CC}" else PTHREAD_CC=$CC fi else PTHREAD_CC="$CC" fi # Finally, execute ACTION-IF-FOUND/ACTION-IF-NOT-FOUND: if test x"$acx_pthread_ok" = xyes; then THREADLIBS="$PTHREAD_LIBS " CC="$PTHREAD_CC" $as_echo "#define USING_POSIX_THREADS 1" >>confdefs.h : else acx_pthread_ok=no fi ac_ext=c ac_cpp='$CPP $CPPFLAGS' ac_compile='$CC -c $CFLAGS $CPPFLAGS conftest.$ac_ext >&5' ac_link='$CC -o conftest$ac_exeext $CFLAGS $CPPFLAGS $LDFLAGS conftest.$ac_ext $LIBS >&5' ac_compiler_gnu=$ac_cv_c_compiler_gnu fi if test -z "$THREADLIBS"; then { $as_echo "$as_me:${as_lineno-$LINENO}: checking for Win32 threads" >&5 $as_echo_n "checking for Win32 threads... " >&6; } cat confdefs.h - <<_ACEOF >conftest.$ac_ext /* end confdefs.h. */ #include #ifdef F77_DUMMY_MAIN # ifdef __cplusplus extern "C" # endif int F77_DUMMY_MAIN() { return 1; } #endif int main () { _beginthreadex(0,0,0,0,0,0); ; return 0; } _ACEOF if ac_fn_c_try_link "$LINENO"; then : THREADLIBS=" "; { $as_echo "$as_me:${as_lineno-$LINENO}: result: yes" >&5 $as_echo "yes" >&6; } else { $as_echo "$as_me:${as_lineno-$LINENO}: result: no" >&5 $as_echo "no" >&6; } fi rm -f core conftest.err conftest.$ac_objext \ conftest$ac_exeext conftest.$ac_ext fi if test -z "$THREADLIBS"; then as_fn_error $? "couldn't find threads library for --enable-threads" "$LINENO" 5 fi $as_echo "#define HAVE_THREADS 1" >>confdefs.h fi if test "$enable_threads" = "yes"; then THREADS_TRUE= THREADS_FALSE='#' else THREADS_TRUE='#' THREADS_FALSE= fi if test "$enable_openmp" = "yes"; then OPENMP_TRUE= OPENMP_FALSE='#' else OPENMP_TRUE='#' OPENMP_FALSE= fi if test "$enable_threads" = "yes" -o "$enable_openmp" = "yes"; then SMP_TRUE= SMP_FALSE='#' else SMP_TRUE='#' SMP_FALSE= fi if test x"$with_combined_threads" = xyes; then COMBINED_THREADS_TRUE= COMBINED_THREADS_FALSE='#' else COMBINED_THREADS_TRUE='#' COMBINED_THREADS_FALSE= fi { $as_echo "$as_me:${as_lineno-$LINENO}: checking whether a cycle counter is available" >&5 $as_echo_n "checking whether a cycle counter is available... " >&6; } save_CPPFLAGS=$CPPFLAGS CPPFLAGS="$CPPFLAGS -I$srcdir/kernel" cat confdefs.h - <<_ACEOF >conftest.$ac_ext /* end confdefs.h. */ #include "cycle.h" #ifndef HAVE_TICK_COUNTER # error No cycle counter #endif _ACEOF if ac_fn_c_try_cpp "$LINENO"; then : ok=yes else ok=no fi rm -f conftest.err conftest.i conftest.$ac_ext CPPFLAGS=$save_CPPFLAGS { $as_echo "$as_me:${as_lineno-$LINENO}: result: $ok" >&5 $as_echo "$ok" >&6; } if test $ok = no && test "x$with_slow_timer" = xno; then echo "***************************************************************" echo "WARNING: No cycle counter found. FFTW will use ESTIMATE mode " echo " for all plans. See the manual for more information." echo "***************************************************************" fi cat >>confdefs.h <<_ACEOF #define FFTW_CC "$CC $CFLAGS" _ACEOF ac_config_files="$ac_config_files Makefile support/Makefile genfft/Makefile kernel/Makefile simd-support/Makefile dft/Makefile dft/scalar/Makefile dft/scalar/codelets/Makefile dft/simd/Makefile dft/simd/common/Makefile dft/simd/sse2/Makefile dft/simd/avx/Makefile dft/simd/altivec/Makefile dft/simd/neon/Makefile rdft/Makefile rdft/scalar/Makefile rdft/scalar/r2cf/Makefile rdft/scalar/r2cb/Makefile rdft/scalar/r2r/Makefile rdft/simd/Makefile rdft/simd/common/Makefile rdft/simd/sse2/Makefile rdft/simd/avx/Makefile rdft/simd/altivec/Makefile rdft/simd/neon/Makefile reodft/Makefile threads/Makefile api/Makefile mpi/Makefile libbench2/Makefile tests/Makefile doc/Makefile doc/FAQ/Makefile tools/Makefile tools/fftw_wisdom.1 tools/fftw-wisdom-to-conf m4/Makefile fftw.pc" cat >confcache <<\_ACEOF # This file is a shell script that caches the results of configure # tests run on this system so they can be shared between configure # scripts and configure runs, see configure's option --config-cache. # It is not useful on other systems. If it contains results you don't # want to keep, you may remove or edit it. # # config.status only pays attention to the cache file if you give it # the --recheck option to rerun configure. # # `ac_cv_env_foo' variables (set or unset) will be overridden when # loading this file, other *unset* `ac_cv_foo' will be assigned the # following values. _ACEOF # The following way of writing the cache mishandles newlines in values, # but we know of no workaround that is simple, portable, and efficient. # So, we kill variables containing newlines. # Ultrix sh set writes to stderr and can't be redirected directly, # and sets the high bit in the cache file unless we assign to the vars. ( for ac_var in `(set) 2>&1 | sed -n 's/^\([a-zA-Z_][a-zA-Z0-9_]*\)=.*/\1/p'`; do eval ac_val=\$$ac_var case $ac_val in #( *${as_nl}*) case $ac_var in #( *_cv_*) { $as_echo "$as_me:${as_lineno-$LINENO}: WARNING: cache variable $ac_var contains a newline" >&5 $as_echo "$as_me: WARNING: cache variable $ac_var contains a newline" >&2;} ;; esac case $ac_var in #( _ | IFS | as_nl) ;; #( BASH_ARGV | BASH_SOURCE) eval $ac_var= ;; #( *) { eval $ac_var=; unset $ac_var;} ;; esac ;; esac done (set) 2>&1 | case $as_nl`(ac_space=' '; set) 2>&1` in #( *${as_nl}ac_space=\ *) # `set' does not quote correctly, so add quotes: double-quote # substitution turns \\\\ into \\, and sed turns \\ into \. sed -n \ "s/'/'\\\\''/g; s/^\\([_$as_cr_alnum]*_cv_[_$as_cr_alnum]*\\)=\\(.*\\)/\\1='\\2'/p" ;; #( *) # `set' quotes correctly as required by POSIX, so do not add quotes. sed -n "/^[_$as_cr_alnum]*_cv_[_$as_cr_alnum]*=/p" ;; esac | sort ) | sed ' /^ac_cv_env_/b end t clear :clear s/^\([^=]*\)=\(.*[{}].*\)$/test "${\1+set}" = set || &/ t end s/^\([^=]*\)=\(.*\)$/\1=${\1=\2}/ :end' >>confcache if diff "$cache_file" confcache >/dev/null 2>&1; then :; else if test -w "$cache_file"; then if test "x$cache_file" != "x/dev/null"; then { $as_echo "$as_me:${as_lineno-$LINENO}: updating cache $cache_file" >&5 $as_echo "$as_me: updating cache $cache_file" >&6;} if test ! -f "$cache_file" || test -h "$cache_file"; then cat confcache >"$cache_file" else case $cache_file in #( */* | ?:*) mv -f confcache "$cache_file"$$ && mv -f "$cache_file"$$ "$cache_file" ;; #( *) mv -f confcache "$cache_file" ;; esac fi fi else { $as_echo "$as_me:${as_lineno-$LINENO}: not updating unwritable cache $cache_file" >&5 $as_echo "$as_me: not updating unwritable cache $cache_file" >&6;} fi fi rm -f confcache test "x$prefix" = xNONE && prefix=$ac_default_prefix # Let make expand exec_prefix. test "x$exec_prefix" = xNONE && exec_prefix='${prefix}' DEFS=-DHAVE_CONFIG_H ac_libobjs= ac_ltlibobjs= U= for ac_i in : $LIBOBJS; do test "x$ac_i" = x: && continue # 1. Remove the extension, and $U if already installed. ac_script='s/\$U\././;s/\.o$//;s/\.obj$//' ac_i=`$as_echo "$ac_i" | sed "$ac_script"` # 2. Prepend LIBOBJDIR. When used with automake>=1.10 LIBOBJDIR # will be set to the directory where LIBOBJS objects are built. as_fn_append ac_libobjs " \${LIBOBJDIR}$ac_i\$U.$ac_objext" as_fn_append ac_ltlibobjs " \${LIBOBJDIR}$ac_i"'$U.lo' done LIBOBJS=$ac_libobjs LTLIBOBJS=$ac_ltlibobjs if test -n "$EXEEXT"; then am__EXEEXT_TRUE= am__EXEEXT_FALSE='#' else am__EXEEXT_TRUE='#' am__EXEEXT_FALSE= fi if test -z "${MAINTAINER_MODE_TRUE}" && test -z "${MAINTAINER_MODE_FALSE}"; then as_fn_error $? "conditional \"MAINTAINER_MODE\" was never defined. Usually this means the macro was only invoked conditionally." "$LINENO" 5 fi if test -z "${SINGLE_TRUE}" && test -z "${SINGLE_FALSE}"; then as_fn_error $? "conditional \"SINGLE\" was never defined. Usually this means the macro was only invoked conditionally." "$LINENO" 5 fi if test -z "${LDOUBLE_TRUE}" && test -z "${LDOUBLE_FALSE}"; then as_fn_error $? "conditional \"LDOUBLE\" was never defined. Usually this means the macro was only invoked conditionally." "$LINENO" 5 fi if test -z "${QUAD_TRUE}" && test -z "${QUAD_FALSE}"; then as_fn_error $? "conditional \"QUAD\" was never defined. Usually this means the macro was only invoked conditionally." "$LINENO" 5 fi if test -z "${HAVE_SSE2_TRUE}" && test -z "${HAVE_SSE2_FALSE}"; then as_fn_error $? "conditional \"HAVE_SSE2\" was never defined. Usually this means the macro was only invoked conditionally." "$LINENO" 5 fi if test -z "${HAVE_AVX_TRUE}" && test -z "${HAVE_AVX_FALSE}"; then as_fn_error $? "conditional \"HAVE_AVX\" was never defined. Usually this means the macro was only invoked conditionally." "$LINENO" 5 fi if test -z "${HAVE_ALTIVEC_TRUE}" && test -z "${HAVE_ALTIVEC_FALSE}"; then as_fn_error $? "conditional \"HAVE_ALTIVEC\" was never defined. Usually this means the macro was only invoked conditionally." "$LINENO" 5 fi if test -z "${HAVE_NEON_TRUE}" && test -z "${HAVE_NEON_FALSE}"; then as_fn_error $? "conditional \"HAVE_NEON\" was never defined. Usually this means the macro was only invoked conditionally." "$LINENO" 5 fi if test -z "${AMDEP_TRUE}" && test -z "${AMDEP_FALSE}"; then as_fn_error $? "conditional \"AMDEP\" was never defined. Usually this means the macro was only invoked conditionally." "$LINENO" 5 fi if test -z "${am__fastdepCC_TRUE}" && test -z "${am__fastdepCC_FALSE}"; then as_fn_error $? "conditional \"am__fastdepCC\" was never defined. Usually this means the macro was only invoked conditionally." "$LINENO" 5 fi if test -z "${MPI_TRUE}" && test -z "${MPI_FALSE}"; then as_fn_error $? "conditional \"MPI\" was never defined. Usually this means the macro was only invoked conditionally." "$LINENO" 5 fi if test -z "${THREADS_TRUE}" && test -z "${THREADS_FALSE}"; then as_fn_error $? "conditional \"THREADS\" was never defined. Usually this means the macro was only invoked conditionally." "$LINENO" 5 fi if test -z "${OPENMP_TRUE}" && test -z "${OPENMP_FALSE}"; then as_fn_error $? "conditional \"OPENMP\" was never defined. Usually this means the macro was only invoked conditionally." "$LINENO" 5 fi if test -z "${SMP_TRUE}" && test -z "${SMP_FALSE}"; then as_fn_error $? "conditional \"SMP\" was never defined. Usually this means the macro was only invoked conditionally." "$LINENO" 5 fi if test -z "${COMBINED_THREADS_TRUE}" && test -z "${COMBINED_THREADS_FALSE}"; then as_fn_error $? "conditional \"COMBINED_THREADS\" was never defined. Usually this means the macro was only invoked conditionally." "$LINENO" 5 fi : "${CONFIG_STATUS=./config.status}" ac_write_fail=0 ac_clean_files_save=$ac_clean_files ac_clean_files="$ac_clean_files $CONFIG_STATUS" { $as_echo "$as_me:${as_lineno-$LINENO}: creating $CONFIG_STATUS" >&5 $as_echo "$as_me: creating $CONFIG_STATUS" >&6;} as_write_fail=0 cat >$CONFIG_STATUS <<_ASEOF || as_write_fail=1 #! $SHELL # Generated by $as_me. # Run this file to recreate the current configuration. # Compiler output produced by configure, useful for debugging # configure, is in config.log if it exists. debug=false ac_cs_recheck=false ac_cs_silent=false SHELL=\${CONFIG_SHELL-$SHELL} export SHELL _ASEOF cat >>$CONFIG_STATUS <<\_ASEOF || as_write_fail=1 ## -------------------- ## ## M4sh Initialization. ## ## -------------------- ## # Be more Bourne compatible DUALCASE=1; export DUALCASE # for MKS sh if test -n "${ZSH_VERSION+set}" && (emulate sh) >/dev/null 2>&1; then : emulate sh NULLCMD=: # Pre-4.2 versions of Zsh do word splitting on ${1+"$@"}, which # is contrary to our usage. Disable this feature. alias -g '${1+"$@"}'='"$@"' setopt NO_GLOB_SUBST else case `(set -o) 2>/dev/null` in #( *posix*) : set -o posix ;; #( *) : ;; esac fi as_nl=' ' export as_nl # Printing a long string crashes Solaris 7 /usr/bin/printf. as_echo='\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\' as_echo=$as_echo$as_echo$as_echo$as_echo$as_echo as_echo=$as_echo$as_echo$as_echo$as_echo$as_echo$as_echo # Prefer a ksh shell builtin over an external printf program on Solaris, # but without wasting forks for bash or zsh. if test -z "$BASH_VERSION$ZSH_VERSION" \ && (test "X`print -r -- $as_echo`" = "X$as_echo") 2>/dev/null; then as_echo='print -r --' as_echo_n='print -rn --' elif (test "X`printf %s $as_echo`" = "X$as_echo") 2>/dev/null; then as_echo='printf %s\n' as_echo_n='printf %s' else if test "X`(/usr/ucb/echo -n -n $as_echo) 2>/dev/null`" = "X-n $as_echo"; then as_echo_body='eval /usr/ucb/echo -n "$1$as_nl"' as_echo_n='/usr/ucb/echo -n' else as_echo_body='eval expr "X$1" : "X\\(.*\\)"' as_echo_n_body='eval arg=$1; case $arg in #( *"$as_nl"*) expr "X$arg" : "X\\(.*\\)$as_nl"; arg=`expr "X$arg" : ".*$as_nl\\(.*\\)"`;; esac; expr "X$arg" : "X\\(.*\\)" | tr -d "$as_nl" ' export as_echo_n_body as_echo_n='sh -c $as_echo_n_body as_echo' fi export as_echo_body as_echo='sh -c $as_echo_body as_echo' fi # The user is always right. if test "${PATH_SEPARATOR+set}" != set; then PATH_SEPARATOR=: (PATH='/bin;/bin'; FPATH=$PATH; sh -c :) >/dev/null 2>&1 && { (PATH='/bin:/bin'; FPATH=$PATH; sh -c :) >/dev/null 2>&1 || PATH_SEPARATOR=';' } fi # IFS # We need space, tab and new line, in precisely that order. Quoting is # there to prevent editors from complaining about space-tab. # (If _AS_PATH_WALK were called with IFS unset, it would disable word # splitting by setting IFS to empty value.) IFS=" "" $as_nl" # Find who we are. Look in the path if we contain no directory separator. as_myself= case $0 in #(( *[\\/]* ) as_myself=$0 ;; *) as_save_IFS=$IFS; IFS=$PATH_SEPARATOR for as_dir in $PATH do IFS=$as_save_IFS test -z "$as_dir" && as_dir=. test -r "$as_dir/$0" && as_myself=$as_dir/$0 && break done IFS=$as_save_IFS ;; esac # We did not find ourselves, most probably we were run as `sh COMMAND' # in which case we are not to be found in the path. if test "x$as_myself" = x; then as_myself=$0 fi if test ! -f "$as_myself"; then $as_echo "$as_myself: error: cannot find myself; rerun with an absolute file name" >&2 exit 1 fi # Unset variables that we do not need and which cause bugs (e.g. in # pre-3.0 UWIN ksh). But do not cause bugs in bash 2.01; the "|| exit 1" # suppresses any "Segmentation fault" message there. '((' could # trigger a bug in pdksh 5.2.14. for as_var in BASH_ENV ENV MAIL MAILPATH do eval test x\${$as_var+set} = xset \ && ( (unset $as_var) || exit 1) >/dev/null 2>&1 && unset $as_var || : done PS1='$ ' PS2='> ' PS4='+ ' # NLS nuisances. LC_ALL=C export LC_ALL LANGUAGE=C export LANGUAGE # CDPATH. (unset CDPATH) >/dev/null 2>&1 && unset CDPATH # as_fn_error STATUS ERROR [LINENO LOG_FD] # ---------------------------------------- # Output "`basename $0`: error: ERROR" to stderr. If LINENO and LOG_FD are # provided, also output the error to LOG_FD, referencing LINENO. Then exit the # script with STATUS, using 1 if that was 0. as_fn_error () { as_status=$1; test $as_status -eq 0 && as_status=1 if test "$4"; then as_lineno=${as_lineno-"$3"} as_lineno_stack=as_lineno_stack=$as_lineno_stack $as_echo "$as_me:${as_lineno-$LINENO}: error: $2" >&$4 fi $as_echo "$as_me: error: $2" >&2 as_fn_exit $as_status } # as_fn_error # as_fn_set_status STATUS # ----------------------- # Set $? to STATUS, without forking. as_fn_set_status () { return $1 } # as_fn_set_status # as_fn_exit STATUS # ----------------- # Exit the shell with STATUS, even in a "trap 0" or "set -e" context. as_fn_exit () { set +e as_fn_set_status $1 exit $1 } # as_fn_exit # as_fn_unset VAR # --------------- # Portably unset VAR. as_fn_unset () { { eval $1=; unset $1;} } as_unset=as_fn_unset # as_fn_append VAR VALUE # ---------------------- # Append the text in VALUE to the end of the definition contained in VAR. Take # advantage of any shell optimizations that allow amortized linear growth over # repeated appends, instead of the typical quadratic growth present in naive # implementations. if (eval "as_var=1; as_var+=2; test x\$as_var = x12") 2>/dev/null; then : eval 'as_fn_append () { eval $1+=\$2 }' else as_fn_append () { eval $1=\$$1\$2 } fi # as_fn_append # as_fn_arith ARG... # ------------------ # Perform arithmetic evaluation on the ARGs, and store the result in the # global $as_val. Take advantage of shells that can avoid forks. The arguments # must be portable across $(()) and expr. if (eval "test \$(( 1 + 1 )) = 2") 2>/dev/null; then : eval 'as_fn_arith () { as_val=$(( $* )) }' else as_fn_arith () { as_val=`expr "$@" || test $? -eq 1` } fi # as_fn_arith if expr a : '\(a\)' >/dev/null 2>&1 && test "X`expr 00001 : '.*\(...\)'`" = X001; then as_expr=expr else as_expr=false fi if (basename -- /) >/dev/null 2>&1 && test "X`basename -- / 2>&1`" = "X/"; then as_basename=basename else as_basename=false fi if (as_dir=`dirname -- /` && test "X$as_dir" = X/) >/dev/null 2>&1; then as_dirname=dirname else as_dirname=false fi as_me=`$as_basename -- "$0" || $as_expr X/"$0" : '.*/\([^/][^/]*\)/*$' \| \ X"$0" : 'X\(//\)$' \| \ X"$0" : 'X\(/\)' \| . 2>/dev/null || $as_echo X/"$0" | sed '/^.*\/\([^/][^/]*\)\/*$/{ s//\1/ q } /^X\/\(\/\/\)$/{ s//\1/ q } /^X\/\(\/\).*/{ s//\1/ q } s/.*/./; q'` # Avoid depending upon Character Ranges. as_cr_letters='abcdefghijklmnopqrstuvwxyz' as_cr_LETTERS='ABCDEFGHIJKLMNOPQRSTUVWXYZ' as_cr_Letters=$as_cr_letters$as_cr_LETTERS as_cr_digits='0123456789' as_cr_alnum=$as_cr_Letters$as_cr_digits ECHO_C= ECHO_N= ECHO_T= case `echo -n x` in #((((( -n*) case `echo 'xy\c'` in *c*) ECHO_T=' ';; # ECHO_T is single tab character. xy) ECHO_C='\c';; *) echo `echo ksh88 bug on AIX 6.1` > /dev/null ECHO_T=' ';; esac;; *) ECHO_N='-n';; esac rm -f conf$$ conf$$.exe conf$$.file if test -d conf$$.dir; then rm -f conf$$.dir/conf$$.file else rm -f conf$$.dir mkdir conf$$.dir 2>/dev/null fi if (echo >conf$$.file) 2>/dev/null; then if ln -s conf$$.file conf$$ 2>/dev/null; then as_ln_s='ln -s' # ... but there are two gotchas: # 1) On MSYS, both `ln -s file dir' and `ln file dir' fail. # 2) DJGPP < 2.04 has no symlinks; `ln -s' creates a wrapper executable. # In both cases, we have to default to `cp -pR'. ln -s conf$$.file conf$$.dir 2>/dev/null && test ! -f conf$$.exe || as_ln_s='cp -pR' elif ln conf$$.file conf$$ 2>/dev/null; then as_ln_s=ln else as_ln_s='cp -pR' fi else as_ln_s='cp -pR' fi rm -f conf$$ conf$$.exe conf$$.dir/conf$$.file conf$$.file rmdir conf$$.dir 2>/dev/null # as_fn_mkdir_p # ------------- # Create "$as_dir" as a directory, including parents if necessary. as_fn_mkdir_p () { case $as_dir in #( -*) as_dir=./$as_dir;; esac test -d "$as_dir" || eval $as_mkdir_p || { as_dirs= while :; do case $as_dir in #( *\'*) as_qdir=`$as_echo "$as_dir" | sed "s/'/'\\\\\\\\''/g"`;; #'( *) as_qdir=$as_dir;; esac as_dirs="'$as_qdir' $as_dirs" as_dir=`$as_dirname -- "$as_dir" || $as_expr X"$as_dir" : 'X\(.*[^/]\)//*[^/][^/]*/*$' \| \ X"$as_dir" : 'X\(//\)[^/]' \| \ X"$as_dir" : 'X\(//\)$' \| \ X"$as_dir" : 'X\(/\)' \| . 2>/dev/null || $as_echo X"$as_dir" | sed '/^X\(.*[^/]\)\/\/*[^/][^/]*\/*$/{ s//\1/ q } /^X\(\/\/\)[^/].*/{ s//\1/ q } /^X\(\/\/\)$/{ s//\1/ q } /^X\(\/\).*/{ s//\1/ q } s/.*/./; q'` test -d "$as_dir" && break done test -z "$as_dirs" || eval "mkdir $as_dirs" } || test -d "$as_dir" || as_fn_error $? "cannot create directory $as_dir" } # as_fn_mkdir_p if mkdir -p . 2>/dev/null; then as_mkdir_p='mkdir -p "$as_dir"' else test -d ./-p && rmdir ./-p as_mkdir_p=false fi # as_fn_executable_p FILE # ----------------------- # Test if FILE is an executable regular file. as_fn_executable_p () { test -f "$1" && test -x "$1" } # as_fn_executable_p as_test_x='test -x' as_executable_p=as_fn_executable_p # Sed expression to map a string onto a valid CPP name. as_tr_cpp="eval sed 'y%*$as_cr_letters%P$as_cr_LETTERS%;s%[^_$as_cr_alnum]%_%g'" # Sed expression to map a string onto a valid variable name. as_tr_sh="eval sed 'y%*+%pp%;s%[^_$as_cr_alnum]%_%g'" exec 6>&1 ## ----------------------------------- ## ## Main body of $CONFIG_STATUS script. ## ## ----------------------------------- ## _ASEOF test $as_write_fail = 0 && chmod +x $CONFIG_STATUS || ac_write_fail=1 cat >>$CONFIG_STATUS <<\_ACEOF || ac_write_fail=1 # Save the log message, to keep $0 and so on meaningful, and to # report actual input values of CONFIG_FILES etc. instead of their # values after options handling. ac_log=" This file was extended by fftw $as_me 3.3.3, which was generated by GNU Autoconf 2.69. Invocation command line was CONFIG_FILES = $CONFIG_FILES CONFIG_HEADERS = $CONFIG_HEADERS CONFIG_LINKS = $CONFIG_LINKS CONFIG_COMMANDS = $CONFIG_COMMANDS $ $0 $@ on `(hostname || uname -n) 2>/dev/null | sed 1q` " _ACEOF case $ac_config_files in *" "*) set x $ac_config_files; shift; ac_config_files=$*;; esac case $ac_config_headers in *" "*) set x $ac_config_headers; shift; ac_config_headers=$*;; esac cat >>$CONFIG_STATUS <<_ACEOF || ac_write_fail=1 # Files that config.status was made for. config_files="$ac_config_files" config_headers="$ac_config_headers" config_commands="$ac_config_commands" _ACEOF cat >>$CONFIG_STATUS <<\_ACEOF || ac_write_fail=1 ac_cs_usage="\ \`$as_me' instantiates files and other configuration actions from templates according to the current configuration. Unless the files and actions are specified as TAGs, all are instantiated by default. Usage: $0 [OPTION]... [TAG]... -h, --help print this help, then exit -V, --version print version number and configuration settings, then exit --config print configuration, then exit -q, --quiet, --silent do not print progress messages -d, --debug don't remove temporary files --recheck update $as_me by reconfiguring in the same conditions --file=FILE[:TEMPLATE] instantiate the configuration file FILE --header=FILE[:TEMPLATE] instantiate the configuration header FILE Configuration files: $config_files Configuration headers: $config_headers Configuration commands: $config_commands Report bugs to ." _ACEOF cat >>$CONFIG_STATUS <<_ACEOF || ac_write_fail=1 ac_cs_config="`$as_echo "$ac_configure_args" | sed 's/^ //; s/[\\""\`\$]/\\\\&/g'`" ac_cs_version="\\ fftw config.status 3.3.3 configured by $0, generated by GNU Autoconf 2.69, with options \\"\$ac_cs_config\\" Copyright (C) 2012 Free Software Foundation, Inc. This config.status script is free software; the Free Software Foundation gives unlimited permission to copy, distribute and modify it." ac_pwd='$ac_pwd' srcdir='$srcdir' INSTALL='$INSTALL' MKDIR_P='$MKDIR_P' AWK='$AWK' test -n "\$AWK" || AWK=awk _ACEOF cat >>$CONFIG_STATUS <<\_ACEOF || ac_write_fail=1 # The default lists apply if the user does not specify any file. ac_need_defaults=: while test $# != 0 do case $1 in --*=?*) ac_option=`expr "X$1" : 'X\([^=]*\)='` ac_optarg=`expr "X$1" : 'X[^=]*=\(.*\)'` ac_shift=: ;; --*=) ac_option=`expr "X$1" : 'X\([^=]*\)='` ac_optarg= ac_shift=: ;; *) ac_option=$1 ac_optarg=$2 ac_shift=shift ;; esac case $ac_option in # Handling of the options. -recheck | --recheck | --rechec | --reche | --rech | --rec | --re | --r) ac_cs_recheck=: ;; --version | --versio | --versi | --vers | --ver | --ve | --v | -V ) $as_echo "$ac_cs_version"; exit ;; --config | --confi | --conf | --con | --co | --c ) $as_echo "$ac_cs_config"; exit ;; --debug | --debu | --deb | --de | --d | -d ) debug=: ;; --file | --fil | --fi | --f ) $ac_shift case $ac_optarg in *\'*) ac_optarg=`$as_echo "$ac_optarg" | sed "s/'/'\\\\\\\\''/g"` ;; '') as_fn_error $? "missing file argument" ;; esac as_fn_append CONFIG_FILES " '$ac_optarg'" ac_need_defaults=false;; --header | --heade | --head | --hea ) $ac_shift case $ac_optarg in *\'*) ac_optarg=`$as_echo "$ac_optarg" | sed "s/'/'\\\\\\\\''/g"` ;; esac as_fn_append CONFIG_HEADERS " '$ac_optarg'" ac_need_defaults=false;; --he | --h) # Conflict between --help and --header as_fn_error $? "ambiguous option: \`$1' Try \`$0 --help' for more information.";; --help | --hel | -h ) $as_echo "$ac_cs_usage"; exit ;; -q | -quiet | --quiet | --quie | --qui | --qu | --q \ | -silent | --silent | --silen | --sile | --sil | --si | --s) ac_cs_silent=: ;; # This is an error. -*) as_fn_error $? "unrecognized option: \`$1' Try \`$0 --help' for more information." ;; *) as_fn_append ac_config_targets " $1" ac_need_defaults=false ;; esac shift done ac_configure_extra_args= if $ac_cs_silent; then exec 6>/dev/null ac_configure_extra_args="$ac_configure_extra_args --silent" fi _ACEOF cat >>$CONFIG_STATUS <<_ACEOF || ac_write_fail=1 if \$ac_cs_recheck; then set X $SHELL '$0' $ac_configure_args \$ac_configure_extra_args --no-create --no-recursion shift \$as_echo "running CONFIG_SHELL=$SHELL \$*" >&6 CONFIG_SHELL='$SHELL' export CONFIG_SHELL exec "\$@" fi _ACEOF cat >>$CONFIG_STATUS <<\_ACEOF || ac_write_fail=1 exec 5>>config.log { echo sed 'h;s/./-/g;s/^.../## /;s/...$/ ##/;p;x;p;x' <<_ASBOX ## Running $as_me. ## _ASBOX $as_echo "$ac_log" } >&5 _ACEOF cat >>$CONFIG_STATUS <<_ACEOF || ac_write_fail=1 # # INIT-COMMANDS # AMDEP_TRUE="$AMDEP_TRUE" ac_aux_dir="$ac_aux_dir" # The HP-UX ksh and POSIX shell print the target directory to stdout # if CDPATH is set. (unset CDPATH) >/dev/null 2>&1 && unset CDPATH sed_quote_subst='$sed_quote_subst' double_quote_subst='$double_quote_subst' delay_variable_subst='$delay_variable_subst' enable_shared='`$ECHO "$enable_shared" | $SED "$delay_single_quote_subst"`' AS='`$ECHO "$AS" | $SED "$delay_single_quote_subst"`' DLLTOOL='`$ECHO "$DLLTOOL" | $SED "$delay_single_quote_subst"`' OBJDUMP='`$ECHO "$OBJDUMP" | $SED "$delay_single_quote_subst"`' macro_version='`$ECHO "$macro_version" | $SED "$delay_single_quote_subst"`' macro_revision='`$ECHO "$macro_revision" | $SED "$delay_single_quote_subst"`' enable_static='`$ECHO "$enable_static" | $SED "$delay_single_quote_subst"`' pic_mode='`$ECHO "$pic_mode" | $SED "$delay_single_quote_subst"`' enable_fast_install='`$ECHO "$enable_fast_install" | $SED "$delay_single_quote_subst"`' SHELL='`$ECHO "$SHELL" | $SED "$delay_single_quote_subst"`' ECHO='`$ECHO "$ECHO" | $SED "$delay_single_quote_subst"`' PATH_SEPARATOR='`$ECHO "$PATH_SEPARATOR" | $SED "$delay_single_quote_subst"`' host_alias='`$ECHO "$host_alias" | $SED "$delay_single_quote_subst"`' host='`$ECHO "$host" | $SED "$delay_single_quote_subst"`' host_os='`$ECHO "$host_os" | $SED "$delay_single_quote_subst"`' build_alias='`$ECHO "$build_alias" | $SED "$delay_single_quote_subst"`' build='`$ECHO "$build" | $SED "$delay_single_quote_subst"`' build_os='`$ECHO "$build_os" | $SED "$delay_single_quote_subst"`' SED='`$ECHO "$SED" | $SED "$delay_single_quote_subst"`' Xsed='`$ECHO "$Xsed" | $SED "$delay_single_quote_subst"`' GREP='`$ECHO "$GREP" | $SED "$delay_single_quote_subst"`' EGREP='`$ECHO "$EGREP" | $SED "$delay_single_quote_subst"`' FGREP='`$ECHO "$FGREP" | $SED "$delay_single_quote_subst"`' LD='`$ECHO "$LD" | $SED "$delay_single_quote_subst"`' NM='`$ECHO "$NM" | $SED "$delay_single_quote_subst"`' LN_S='`$ECHO "$LN_S" | $SED "$delay_single_quote_subst"`' max_cmd_len='`$ECHO "$max_cmd_len" | $SED "$delay_single_quote_subst"`' ac_objext='`$ECHO "$ac_objext" | $SED "$delay_single_quote_subst"`' exeext='`$ECHO "$exeext" | $SED "$delay_single_quote_subst"`' lt_unset='`$ECHO "$lt_unset" | $SED "$delay_single_quote_subst"`' lt_SP2NL='`$ECHO "$lt_SP2NL" | $SED "$delay_single_quote_subst"`' lt_NL2SP='`$ECHO "$lt_NL2SP" | $SED "$delay_single_quote_subst"`' lt_cv_to_host_file_cmd='`$ECHO "$lt_cv_to_host_file_cmd" | $SED "$delay_single_quote_subst"`' lt_cv_to_tool_file_cmd='`$ECHO "$lt_cv_to_tool_file_cmd" | $SED "$delay_single_quote_subst"`' reload_flag='`$ECHO "$reload_flag" | $SED "$delay_single_quote_subst"`' reload_cmds='`$ECHO "$reload_cmds" | $SED "$delay_single_quote_subst"`' deplibs_check_method='`$ECHO "$deplibs_check_method" | $SED "$delay_single_quote_subst"`' file_magic_cmd='`$ECHO "$file_magic_cmd" | $SED "$delay_single_quote_subst"`' file_magic_glob='`$ECHO "$file_magic_glob" | $SED "$delay_single_quote_subst"`' want_nocaseglob='`$ECHO "$want_nocaseglob" | $SED "$delay_single_quote_subst"`' sharedlib_from_linklib_cmd='`$ECHO "$sharedlib_from_linklib_cmd" | $SED "$delay_single_quote_subst"`' AR='`$ECHO "$AR" | $SED "$delay_single_quote_subst"`' AR_FLAGS='`$ECHO "$AR_FLAGS" | $SED "$delay_single_quote_subst"`' archiver_list_spec='`$ECHO "$archiver_list_spec" | $SED "$delay_single_quote_subst"`' STRIP='`$ECHO "$STRIP" | $SED "$delay_single_quote_subst"`' RANLIB='`$ECHO "$RANLIB" | $SED "$delay_single_quote_subst"`' old_postinstall_cmds='`$ECHO "$old_postinstall_cmds" | $SED "$delay_single_quote_subst"`' old_postuninstall_cmds='`$ECHO "$old_postuninstall_cmds" | $SED "$delay_single_quote_subst"`' old_archive_cmds='`$ECHO "$old_archive_cmds" | $SED "$delay_single_quote_subst"`' lock_old_archive_extraction='`$ECHO "$lock_old_archive_extraction" | $SED "$delay_single_quote_subst"`' CC='`$ECHO "$CC" | $SED "$delay_single_quote_subst"`' CFLAGS='`$ECHO "$CFLAGS" | $SED "$delay_single_quote_subst"`' compiler='`$ECHO "$compiler" | $SED "$delay_single_quote_subst"`' GCC='`$ECHO "$GCC" | $SED "$delay_single_quote_subst"`' lt_cv_sys_global_symbol_pipe='`$ECHO "$lt_cv_sys_global_symbol_pipe" | $SED "$delay_single_quote_subst"`' lt_cv_sys_global_symbol_to_cdecl='`$ECHO "$lt_cv_sys_global_symbol_to_cdecl" | $SED "$delay_single_quote_subst"`' lt_cv_sys_global_symbol_to_c_name_address='`$ECHO "$lt_cv_sys_global_symbol_to_c_name_address" | $SED "$delay_single_quote_subst"`' lt_cv_sys_global_symbol_to_c_name_address_lib_prefix='`$ECHO "$lt_cv_sys_global_symbol_to_c_name_address_lib_prefix" | $SED "$delay_single_quote_subst"`' nm_file_list_spec='`$ECHO "$nm_file_list_spec" | $SED "$delay_single_quote_subst"`' lt_sysroot='`$ECHO "$lt_sysroot" | $SED "$delay_single_quote_subst"`' objdir='`$ECHO "$objdir" | $SED "$delay_single_quote_subst"`' MAGIC_CMD='`$ECHO "$MAGIC_CMD" | $SED "$delay_single_quote_subst"`' lt_prog_compiler_no_builtin_flag='`$ECHO "$lt_prog_compiler_no_builtin_flag" | $SED "$delay_single_quote_subst"`' lt_prog_compiler_pic='`$ECHO "$lt_prog_compiler_pic" | $SED "$delay_single_quote_subst"`' lt_prog_compiler_wl='`$ECHO "$lt_prog_compiler_wl" | $SED "$delay_single_quote_subst"`' lt_prog_compiler_static='`$ECHO "$lt_prog_compiler_static" | $SED "$delay_single_quote_subst"`' lt_cv_prog_compiler_c_o='`$ECHO "$lt_cv_prog_compiler_c_o" | $SED "$delay_single_quote_subst"`' need_locks='`$ECHO "$need_locks" | $SED "$delay_single_quote_subst"`' MANIFEST_TOOL='`$ECHO "$MANIFEST_TOOL" | $SED "$delay_single_quote_subst"`' DSYMUTIL='`$ECHO "$DSYMUTIL" | $SED "$delay_single_quote_subst"`' NMEDIT='`$ECHO "$NMEDIT" | $SED "$delay_single_quote_subst"`' LIPO='`$ECHO "$LIPO" | $SED "$delay_single_quote_subst"`' OTOOL='`$ECHO "$OTOOL" | $SED "$delay_single_quote_subst"`' OTOOL64='`$ECHO "$OTOOL64" | $SED "$delay_single_quote_subst"`' libext='`$ECHO "$libext" | $SED "$delay_single_quote_subst"`' shrext_cmds='`$ECHO "$shrext_cmds" | $SED "$delay_single_quote_subst"`' extract_expsyms_cmds='`$ECHO "$extract_expsyms_cmds" | $SED "$delay_single_quote_subst"`' archive_cmds_need_lc='`$ECHO "$archive_cmds_need_lc" | $SED "$delay_single_quote_subst"`' enable_shared_with_static_runtimes='`$ECHO "$enable_shared_with_static_runtimes" | $SED "$delay_single_quote_subst"`' export_dynamic_flag_spec='`$ECHO "$export_dynamic_flag_spec" | $SED "$delay_single_quote_subst"`' whole_archive_flag_spec='`$ECHO "$whole_archive_flag_spec" | $SED "$delay_single_quote_subst"`' compiler_needs_object='`$ECHO "$compiler_needs_object" | $SED "$delay_single_quote_subst"`' old_archive_from_new_cmds='`$ECHO "$old_archive_from_new_cmds" | $SED "$delay_single_quote_subst"`' old_archive_from_expsyms_cmds='`$ECHO "$old_archive_from_expsyms_cmds" | $SED "$delay_single_quote_subst"`' archive_cmds='`$ECHO "$archive_cmds" | $SED "$delay_single_quote_subst"`' archive_expsym_cmds='`$ECHO "$archive_expsym_cmds" | $SED "$delay_single_quote_subst"`' module_cmds='`$ECHO "$module_cmds" | $SED "$delay_single_quote_subst"`' module_expsym_cmds='`$ECHO "$module_expsym_cmds" | $SED "$delay_single_quote_subst"`' with_gnu_ld='`$ECHO "$with_gnu_ld" | $SED "$delay_single_quote_subst"`' allow_undefined_flag='`$ECHO "$allow_undefined_flag" | $SED "$delay_single_quote_subst"`' no_undefined_flag='`$ECHO "$no_undefined_flag" | $SED "$delay_single_quote_subst"`' hardcode_libdir_flag_spec='`$ECHO "$hardcode_libdir_flag_spec" | $SED "$delay_single_quote_subst"`' hardcode_libdir_separator='`$ECHO "$hardcode_libdir_separator" | $SED "$delay_single_quote_subst"`' hardcode_direct='`$ECHO "$hardcode_direct" | $SED "$delay_single_quote_subst"`' hardcode_direct_absolute='`$ECHO "$hardcode_direct_absolute" | $SED "$delay_single_quote_subst"`' hardcode_minus_L='`$ECHO "$hardcode_minus_L" | $SED "$delay_single_quote_subst"`' hardcode_shlibpath_var='`$ECHO "$hardcode_shlibpath_var" | $SED "$delay_single_quote_subst"`' hardcode_automatic='`$ECHO "$hardcode_automatic" | $SED "$delay_single_quote_subst"`' inherit_rpath='`$ECHO "$inherit_rpath" | $SED "$delay_single_quote_subst"`' link_all_deplibs='`$ECHO "$link_all_deplibs" | $SED "$delay_single_quote_subst"`' always_export_symbols='`$ECHO "$always_export_symbols" | $SED "$delay_single_quote_subst"`' export_symbols_cmds='`$ECHO "$export_symbols_cmds" | $SED "$delay_single_quote_subst"`' exclude_expsyms='`$ECHO "$exclude_expsyms" | $SED "$delay_single_quote_subst"`' include_expsyms='`$ECHO "$include_expsyms" | $SED "$delay_single_quote_subst"`' prelink_cmds='`$ECHO "$prelink_cmds" | $SED "$delay_single_quote_subst"`' postlink_cmds='`$ECHO "$postlink_cmds" | $SED "$delay_single_quote_subst"`' file_list_spec='`$ECHO "$file_list_spec" | $SED "$delay_single_quote_subst"`' variables_saved_for_relink='`$ECHO "$variables_saved_for_relink" | $SED "$delay_single_quote_subst"`' need_lib_prefix='`$ECHO "$need_lib_prefix" | $SED "$delay_single_quote_subst"`' need_version='`$ECHO "$need_version" | $SED "$delay_single_quote_subst"`' version_type='`$ECHO "$version_type" | $SED "$delay_single_quote_subst"`' runpath_var='`$ECHO "$runpath_var" | $SED "$delay_single_quote_subst"`' shlibpath_var='`$ECHO "$shlibpath_var" | $SED "$delay_single_quote_subst"`' shlibpath_overrides_runpath='`$ECHO "$shlibpath_overrides_runpath" | $SED "$delay_single_quote_subst"`' libname_spec='`$ECHO "$libname_spec" | $SED "$delay_single_quote_subst"`' library_names_spec='`$ECHO "$library_names_spec" | $SED "$delay_single_quote_subst"`' soname_spec='`$ECHO "$soname_spec" | $SED "$delay_single_quote_subst"`' install_override_mode='`$ECHO "$install_override_mode" | $SED "$delay_single_quote_subst"`' postinstall_cmds='`$ECHO "$postinstall_cmds" | $SED "$delay_single_quote_subst"`' postuninstall_cmds='`$ECHO "$postuninstall_cmds" | $SED "$delay_single_quote_subst"`' finish_cmds='`$ECHO "$finish_cmds" | $SED "$delay_single_quote_subst"`' finish_eval='`$ECHO "$finish_eval" | $SED "$delay_single_quote_subst"`' hardcode_into_libs='`$ECHO "$hardcode_into_libs" | $SED "$delay_single_quote_subst"`' sys_lib_search_path_spec='`$ECHO "$sys_lib_search_path_spec" | $SED "$delay_single_quote_subst"`' sys_lib_dlsearch_path_spec='`$ECHO "$sys_lib_dlsearch_path_spec" | $SED "$delay_single_quote_subst"`' hardcode_action='`$ECHO "$hardcode_action" | $SED "$delay_single_quote_subst"`' enable_dlopen='`$ECHO "$enable_dlopen" | $SED "$delay_single_quote_subst"`' enable_dlopen_self='`$ECHO "$enable_dlopen_self" | $SED "$delay_single_quote_subst"`' enable_dlopen_self_static='`$ECHO "$enable_dlopen_self_static" | $SED "$delay_single_quote_subst"`' old_striplib='`$ECHO "$old_striplib" | $SED "$delay_single_quote_subst"`' striplib='`$ECHO "$striplib" | $SED "$delay_single_quote_subst"`' LD_F77='`$ECHO "$LD_F77" | $SED "$delay_single_quote_subst"`' reload_flag_F77='`$ECHO "$reload_flag_F77" | $SED "$delay_single_quote_subst"`' reload_cmds_F77='`$ECHO "$reload_cmds_F77" | $SED "$delay_single_quote_subst"`' old_archive_cmds_F77='`$ECHO "$old_archive_cmds_F77" | $SED "$delay_single_quote_subst"`' compiler_F77='`$ECHO "$compiler_F77" | $SED "$delay_single_quote_subst"`' GCC_F77='`$ECHO "$GCC_F77" | $SED "$delay_single_quote_subst"`' lt_prog_compiler_no_builtin_flag_F77='`$ECHO "$lt_prog_compiler_no_builtin_flag_F77" | $SED "$delay_single_quote_subst"`' lt_prog_compiler_pic_F77='`$ECHO "$lt_prog_compiler_pic_F77" | $SED "$delay_single_quote_subst"`' lt_prog_compiler_wl_F77='`$ECHO "$lt_prog_compiler_wl_F77" | $SED "$delay_single_quote_subst"`' lt_prog_compiler_static_F77='`$ECHO "$lt_prog_compiler_static_F77" | $SED "$delay_single_quote_subst"`' lt_cv_prog_compiler_c_o_F77='`$ECHO "$lt_cv_prog_compiler_c_o_F77" | $SED "$delay_single_quote_subst"`' archive_cmds_need_lc_F77='`$ECHO "$archive_cmds_need_lc_F77" | $SED "$delay_single_quote_subst"`' enable_shared_with_static_runtimes_F77='`$ECHO "$enable_shared_with_static_runtimes_F77" | $SED "$delay_single_quote_subst"`' export_dynamic_flag_spec_F77='`$ECHO "$export_dynamic_flag_spec_F77" | $SED "$delay_single_quote_subst"`' whole_archive_flag_spec_F77='`$ECHO "$whole_archive_flag_spec_F77" | $SED "$delay_single_quote_subst"`' compiler_needs_object_F77='`$ECHO "$compiler_needs_object_F77" | $SED "$delay_single_quote_subst"`' old_archive_from_new_cmds_F77='`$ECHO "$old_archive_from_new_cmds_F77" | $SED "$delay_single_quote_subst"`' old_archive_from_expsyms_cmds_F77='`$ECHO "$old_archive_from_expsyms_cmds_F77" | $SED "$delay_single_quote_subst"`' archive_cmds_F77='`$ECHO "$archive_cmds_F77" | $SED "$delay_single_quote_subst"`' archive_expsym_cmds_F77='`$ECHO "$archive_expsym_cmds_F77" | $SED "$delay_single_quote_subst"`' module_cmds_F77='`$ECHO "$module_cmds_F77" | $SED "$delay_single_quote_subst"`' module_expsym_cmds_F77='`$ECHO "$module_expsym_cmds_F77" | $SED "$delay_single_quote_subst"`' with_gnu_ld_F77='`$ECHO "$with_gnu_ld_F77" | $SED "$delay_single_quote_subst"`' allow_undefined_flag_F77='`$ECHO "$allow_undefined_flag_F77" | $SED "$delay_single_quote_subst"`' no_undefined_flag_F77='`$ECHO "$no_undefined_flag_F77" | $SED "$delay_single_quote_subst"`' hardcode_libdir_flag_spec_F77='`$ECHO "$hardcode_libdir_flag_spec_F77" | $SED "$delay_single_quote_subst"`' hardcode_libdir_separator_F77='`$ECHO "$hardcode_libdir_separator_F77" | $SED "$delay_single_quote_subst"`' hardcode_direct_F77='`$ECHO "$hardcode_direct_F77" | $SED "$delay_single_quote_subst"`' hardcode_direct_absolute_F77='`$ECHO "$hardcode_direct_absolute_F77" | $SED "$delay_single_quote_subst"`' hardcode_minus_L_F77='`$ECHO "$hardcode_minus_L_F77" | $SED "$delay_single_quote_subst"`' hardcode_shlibpath_var_F77='`$ECHO "$hardcode_shlibpath_var_F77" | $SED "$delay_single_quote_subst"`' hardcode_automatic_F77='`$ECHO "$hardcode_automatic_F77" | $SED "$delay_single_quote_subst"`' inherit_rpath_F77='`$ECHO "$inherit_rpath_F77" | $SED "$delay_single_quote_subst"`' link_all_deplibs_F77='`$ECHO "$link_all_deplibs_F77" | $SED "$delay_single_quote_subst"`' always_export_symbols_F77='`$ECHO "$always_export_symbols_F77" | $SED "$delay_single_quote_subst"`' export_symbols_cmds_F77='`$ECHO "$export_symbols_cmds_F77" | $SED "$delay_single_quote_subst"`' exclude_expsyms_F77='`$ECHO "$exclude_expsyms_F77" | $SED "$delay_single_quote_subst"`' include_expsyms_F77='`$ECHO "$include_expsyms_F77" | $SED "$delay_single_quote_subst"`' prelink_cmds_F77='`$ECHO "$prelink_cmds_F77" | $SED "$delay_single_quote_subst"`' postlink_cmds_F77='`$ECHO "$postlink_cmds_F77" | $SED "$delay_single_quote_subst"`' file_list_spec_F77='`$ECHO "$file_list_spec_F77" | $SED "$delay_single_quote_subst"`' hardcode_action_F77='`$ECHO "$hardcode_action_F77" | $SED "$delay_single_quote_subst"`' LTCC='$LTCC' LTCFLAGS='$LTCFLAGS' compiler='$compiler_DEFAULT' # A function that is used when there is no print builtin or printf. func_fallback_echo () { eval 'cat <<_LTECHO_EOF \$1 _LTECHO_EOF' } # Quote evaled strings. for var in AS \ DLLTOOL \ OBJDUMP \ SHELL \ ECHO \ PATH_SEPARATOR \ SED \ GREP \ EGREP \ FGREP \ LD \ NM \ LN_S \ lt_SP2NL \ lt_NL2SP \ reload_flag \ deplibs_check_method \ file_magic_cmd \ file_magic_glob \ want_nocaseglob \ sharedlib_from_linklib_cmd \ AR \ AR_FLAGS \ archiver_list_spec \ STRIP \ RANLIB \ CC \ CFLAGS \ compiler \ lt_cv_sys_global_symbol_pipe \ lt_cv_sys_global_symbol_to_cdecl \ lt_cv_sys_global_symbol_to_c_name_address \ lt_cv_sys_global_symbol_to_c_name_address_lib_prefix \ nm_file_list_spec \ lt_prog_compiler_no_builtin_flag \ lt_prog_compiler_pic \ lt_prog_compiler_wl \ lt_prog_compiler_static \ lt_cv_prog_compiler_c_o \ need_locks \ MANIFEST_TOOL \ DSYMUTIL \ NMEDIT \ LIPO \ OTOOL \ OTOOL64 \ shrext_cmds \ export_dynamic_flag_spec \ whole_archive_flag_spec \ compiler_needs_object \ with_gnu_ld \ allow_undefined_flag \ no_undefined_flag \ hardcode_libdir_flag_spec \ hardcode_libdir_separator \ exclude_expsyms \ include_expsyms \ file_list_spec \ variables_saved_for_relink \ libname_spec \ library_names_spec \ soname_spec \ install_override_mode \ finish_eval \ old_striplib \ striplib \ LD_F77 \ reload_flag_F77 \ compiler_F77 \ lt_prog_compiler_no_builtin_flag_F77 \ lt_prog_compiler_pic_F77 \ lt_prog_compiler_wl_F77 \ lt_prog_compiler_static_F77 \ lt_cv_prog_compiler_c_o_F77 \ export_dynamic_flag_spec_F77 \ whole_archive_flag_spec_F77 \ compiler_needs_object_F77 \ with_gnu_ld_F77 \ allow_undefined_flag_F77 \ no_undefined_flag_F77 \ hardcode_libdir_flag_spec_F77 \ hardcode_libdir_separator_F77 \ exclude_expsyms_F77 \ include_expsyms_F77 \ file_list_spec_F77; do case \`eval \\\\\$ECHO \\\\""\\\\\$\$var"\\\\"\` in *[\\\\\\\`\\"\\\$]*) eval "lt_\$var=\\\\\\"\\\`\\\$ECHO \\"\\\$\$var\\" | \\\$SED \\"\\\$sed_quote_subst\\"\\\`\\\\\\"" ;; *) eval "lt_\$var=\\\\\\"\\\$\$var\\\\\\"" ;; esac done # Double-quote double-evaled strings. for var in reload_cmds \ old_postinstall_cmds \ old_postuninstall_cmds \ old_archive_cmds \ extract_expsyms_cmds \ old_archive_from_new_cmds \ old_archive_from_expsyms_cmds \ archive_cmds \ archive_expsym_cmds \ module_cmds \ module_expsym_cmds \ export_symbols_cmds \ prelink_cmds \ postlink_cmds \ postinstall_cmds \ postuninstall_cmds \ finish_cmds \ sys_lib_search_path_spec \ sys_lib_dlsearch_path_spec \ reload_cmds_F77 \ old_archive_cmds_F77 \ old_archive_from_new_cmds_F77 \ old_archive_from_expsyms_cmds_F77 \ archive_cmds_F77 \ archive_expsym_cmds_F77 \ module_cmds_F77 \ module_expsym_cmds_F77 \ export_symbols_cmds_F77 \ prelink_cmds_F77 \ postlink_cmds_F77; do case \`eval \\\\\$ECHO \\\\""\\\\\$\$var"\\\\"\` in *[\\\\\\\`\\"\\\$]*) eval "lt_\$var=\\\\\\"\\\`\\\$ECHO \\"\\\$\$var\\" | \\\$SED -e \\"\\\$double_quote_subst\\" -e \\"\\\$sed_quote_subst\\" -e \\"\\\$delay_variable_subst\\"\\\`\\\\\\"" ;; *) eval "lt_\$var=\\\\\\"\\\$\$var\\\\\\"" ;; esac done ac_aux_dir='$ac_aux_dir' xsi_shell='$xsi_shell' lt_shell_append='$lt_shell_append' # See if we are running on zsh, and set the options which allow our # commands through without removal of \ escapes INIT. if test -n "\${ZSH_VERSION+set}" ; then setopt NO_GLOB_SUBST fi PACKAGE='$PACKAGE' VERSION='$VERSION' TIMESTAMP='$TIMESTAMP' RM='$RM' ofile='$ofile' _ACEOF cat >>$CONFIG_STATUS <<\_ACEOF || ac_write_fail=1 # Handling of arguments. for ac_config_target in $ac_config_targets do case $ac_config_target in "config.h") CONFIG_HEADERS="$CONFIG_HEADERS config.h" ;; "depfiles") CONFIG_COMMANDS="$CONFIG_COMMANDS depfiles" ;; "libtool") CONFIG_COMMANDS="$CONFIG_COMMANDS libtool" ;; "Makefile") CONFIG_FILES="$CONFIG_FILES Makefile" ;; "support/Makefile") CONFIG_FILES="$CONFIG_FILES support/Makefile" ;; "genfft/Makefile") CONFIG_FILES="$CONFIG_FILES genfft/Makefile" ;; "kernel/Makefile") CONFIG_FILES="$CONFIG_FILES kernel/Makefile" ;; "simd-support/Makefile") CONFIG_FILES="$CONFIG_FILES simd-support/Makefile" ;; "dft/Makefile") CONFIG_FILES="$CONFIG_FILES dft/Makefile" ;; "dft/scalar/Makefile") CONFIG_FILES="$CONFIG_FILES dft/scalar/Makefile" ;; "dft/scalar/codelets/Makefile") CONFIG_FILES="$CONFIG_FILES dft/scalar/codelets/Makefile" ;; "dft/simd/Makefile") CONFIG_FILES="$CONFIG_FILES dft/simd/Makefile" ;; "dft/simd/common/Makefile") CONFIG_FILES="$CONFIG_FILES dft/simd/common/Makefile" ;; "dft/simd/sse2/Makefile") CONFIG_FILES="$CONFIG_FILES dft/simd/sse2/Makefile" ;; "dft/simd/avx/Makefile") CONFIG_FILES="$CONFIG_FILES dft/simd/avx/Makefile" ;; "dft/simd/altivec/Makefile") CONFIG_FILES="$CONFIG_FILES dft/simd/altivec/Makefile" ;; "dft/simd/neon/Makefile") CONFIG_FILES="$CONFIG_FILES dft/simd/neon/Makefile" ;; "rdft/Makefile") CONFIG_FILES="$CONFIG_FILES rdft/Makefile" ;; "rdft/scalar/Makefile") CONFIG_FILES="$CONFIG_FILES rdft/scalar/Makefile" ;; "rdft/scalar/r2cf/Makefile") CONFIG_FILES="$CONFIG_FILES rdft/scalar/r2cf/Makefile" ;; "rdft/scalar/r2cb/Makefile") CONFIG_FILES="$CONFIG_FILES rdft/scalar/r2cb/Makefile" ;; "rdft/scalar/r2r/Makefile") CONFIG_FILES="$CONFIG_FILES rdft/scalar/r2r/Makefile" ;; "rdft/simd/Makefile") CONFIG_FILES="$CONFIG_FILES rdft/simd/Makefile" ;; "rdft/simd/common/Makefile") CONFIG_FILES="$CONFIG_FILES rdft/simd/common/Makefile" ;; "rdft/simd/sse2/Makefile") CONFIG_FILES="$CONFIG_FILES rdft/simd/sse2/Makefile" ;; "rdft/simd/avx/Makefile") CONFIG_FILES="$CONFIG_FILES rdft/simd/avx/Makefile" ;; "rdft/simd/altivec/Makefile") CONFIG_FILES="$CONFIG_FILES rdft/simd/altivec/Makefile" ;; "rdft/simd/neon/Makefile") CONFIG_FILES="$CONFIG_FILES rdft/simd/neon/Makefile" ;; "reodft/Makefile") CONFIG_FILES="$CONFIG_FILES reodft/Makefile" ;; "threads/Makefile") CONFIG_FILES="$CONFIG_FILES threads/Makefile" ;; "api/Makefile") CONFIG_FILES="$CONFIG_FILES api/Makefile" ;; "mpi/Makefile") CONFIG_FILES="$CONFIG_FILES mpi/Makefile" ;; "libbench2/Makefile") CONFIG_FILES="$CONFIG_FILES libbench2/Makefile" ;; "tests/Makefile") CONFIG_FILES="$CONFIG_FILES tests/Makefile" ;; "doc/Makefile") CONFIG_FILES="$CONFIG_FILES doc/Makefile" ;; "doc/FAQ/Makefile") CONFIG_FILES="$CONFIG_FILES doc/FAQ/Makefile" ;; "tools/Makefile") CONFIG_FILES="$CONFIG_FILES tools/Makefile" ;; "tools/fftw_wisdom.1") CONFIG_FILES="$CONFIG_FILES tools/fftw_wisdom.1" ;; "tools/fftw-wisdom-to-conf") CONFIG_FILES="$CONFIG_FILES tools/fftw-wisdom-to-conf" ;; "m4/Makefile") CONFIG_FILES="$CONFIG_FILES m4/Makefile" ;; "fftw.pc") CONFIG_FILES="$CONFIG_FILES fftw.pc" ;; *) as_fn_error $? "invalid argument: \`$ac_config_target'" "$LINENO" 5;; esac done # If the user did not use the arguments to specify the items to instantiate, # then the envvar interface is used. Set only those that are not. # We use the long form for the default assignment because of an extremely # bizarre bug on SunOS 4.1.3. if $ac_need_defaults; then test "${CONFIG_FILES+set}" = set || CONFIG_FILES=$config_files test "${CONFIG_HEADERS+set}" = set || CONFIG_HEADERS=$config_headers test "${CONFIG_COMMANDS+set}" = set || CONFIG_COMMANDS=$config_commands fi # Have a temporary directory for convenience. Make it in the build tree # simply because there is no reason against having it here, and in addition, # creating and moving files from /tmp can sometimes cause problems. # Hook for its removal unless debugging. # Note that there is a small window in which the directory will not be cleaned: # after its creation but before its name has been assigned to `$tmp'. $debug || { tmp= ac_tmp= trap 'exit_status=$? : "${ac_tmp:=$tmp}" { test ! -d "$ac_tmp" || rm -fr "$ac_tmp"; } && exit $exit_status ' 0 trap 'as_fn_exit 1' 1 2 13 15 } # Create a (secure) tmp directory for tmp files. { tmp=`(umask 077 && mktemp -d "./confXXXXXX") 2>/dev/null` && test -d "$tmp" } || { tmp=./conf$$-$RANDOM (umask 077 && mkdir "$tmp") } || as_fn_error $? "cannot create a temporary directory in ." "$LINENO" 5 ac_tmp=$tmp # Set up the scripts for CONFIG_FILES section. # No need to generate them if there are no CONFIG_FILES. # This happens for instance with `./config.status config.h'. if test -n "$CONFIG_FILES"; then ac_cr=`echo X | tr X '\015'` # On cygwin, bash can eat \r inside `` if the user requested igncr. # But we know of no other shell where ac_cr would be empty at this # point, so we can use a bashism as a fallback. if test "x$ac_cr" = x; then eval ac_cr=\$\'\\r\' fi ac_cs_awk_cr=`$AWK 'BEGIN { print "a\rb" }' /dev/null` if test "$ac_cs_awk_cr" = "a${ac_cr}b"; then ac_cs_awk_cr='\\r' else ac_cs_awk_cr=$ac_cr fi echo 'BEGIN {' >"$ac_tmp/subs1.awk" && _ACEOF { echo "cat >conf$$subs.awk <<_ACEOF" && echo "$ac_subst_vars" | sed 's/.*/&!$&$ac_delim/' && echo "_ACEOF" } >conf$$subs.sh || as_fn_error $? "could not make $CONFIG_STATUS" "$LINENO" 5 ac_delim_num=`echo "$ac_subst_vars" | grep -c '^'` ac_delim='%!_!# ' for ac_last_try in false false false false false :; do . ./conf$$subs.sh || as_fn_error $? "could not make $CONFIG_STATUS" "$LINENO" 5 ac_delim_n=`sed -n "s/.*$ac_delim\$/X/p" conf$$subs.awk | grep -c X` if test $ac_delim_n = $ac_delim_num; then break elif $ac_last_try; then as_fn_error $? "could not make $CONFIG_STATUS" "$LINENO" 5 else ac_delim="$ac_delim!$ac_delim _$ac_delim!! " fi done rm -f conf$$subs.sh cat >>$CONFIG_STATUS <<_ACEOF || ac_write_fail=1 cat >>"\$ac_tmp/subs1.awk" <<\\_ACAWK && _ACEOF sed -n ' h s/^/S["/; s/!.*/"]=/ p g s/^[^!]*!// :repl t repl s/'"$ac_delim"'$// t delim :nl h s/\(.\{148\}\)..*/\1/ t more1 s/["\\]/\\&/g; s/^/"/; s/$/\\n"\\/ p n b repl :more1 s/["\\]/\\&/g; s/^/"/; s/$/"\\/ p g s/.\{148\}// t nl :delim h s/\(.\{148\}\)..*/\1/ t more2 s/["\\]/\\&/g; s/^/"/; s/$/"/ p b :more2 s/["\\]/\\&/g; s/^/"/; s/$/"\\/ p g s/.\{148\}// t delim ' >$CONFIG_STATUS || ac_write_fail=1 rm -f conf$$subs.awk cat >>$CONFIG_STATUS <<_ACEOF || ac_write_fail=1 _ACAWK cat >>"\$ac_tmp/subs1.awk" <<_ACAWK && for (key in S) S_is_set[key] = 1 FS = "" } { line = $ 0 nfields = split(line, field, "@") substed = 0 len = length(field[1]) for (i = 2; i < nfields; i++) { key = field[i] keylen = length(key) if (S_is_set[key]) { value = S[key] line = substr(line, 1, len) "" value "" substr(line, len + keylen + 3) len += length(value) + length(field[++i]) substed = 1 } else len += 1 + keylen } print line } _ACAWK _ACEOF cat >>$CONFIG_STATUS <<\_ACEOF || ac_write_fail=1 if sed "s/$ac_cr//" < /dev/null > /dev/null 2>&1; then sed "s/$ac_cr\$//; s/$ac_cr/$ac_cs_awk_cr/g" else cat fi < "$ac_tmp/subs1.awk" > "$ac_tmp/subs.awk" \ || as_fn_error $? "could not setup config files machinery" "$LINENO" 5 _ACEOF # VPATH may cause trouble with some makes, so we remove sole $(srcdir), # ${srcdir} and @srcdir@ entries from VPATH if srcdir is ".", strip leading and # trailing colons and then remove the whole line if VPATH becomes empty # (actually we leave an empty line to preserve line numbers). if test "x$srcdir" = x.; then ac_vpsub='/^[ ]*VPATH[ ]*=[ ]*/{ h s/// s/^/:/ s/[ ]*$/:/ s/:\$(srcdir):/:/g s/:\${srcdir}:/:/g s/:@srcdir@:/:/g s/^:*// s/:*$// x s/\(=[ ]*\).*/\1/ G s/\n// s/^[^=]*=[ ]*$// }' fi cat >>$CONFIG_STATUS <<\_ACEOF || ac_write_fail=1 fi # test -n "$CONFIG_FILES" # Set up the scripts for CONFIG_HEADERS section. # No need to generate them if there are no CONFIG_HEADERS. # This happens for instance with `./config.status Makefile'. if test -n "$CONFIG_HEADERS"; then cat >"$ac_tmp/defines.awk" <<\_ACAWK || BEGIN { _ACEOF # Transform confdefs.h into an awk script `defines.awk', embedded as # here-document in config.status, that substitutes the proper values into # config.h.in to produce config.h. # Create a delimiter string that does not exist in confdefs.h, to ease # handling of long lines. ac_delim='%!_!# ' for ac_last_try in false false :; do ac_tt=`sed -n "/$ac_delim/p" confdefs.h` if test -z "$ac_tt"; then break elif $ac_last_try; then as_fn_error $? "could not make $CONFIG_HEADERS" "$LINENO" 5 else ac_delim="$ac_delim!$ac_delim _$ac_delim!! " fi done # For the awk script, D is an array of macro values keyed by name, # likewise P contains macro parameters if any. Preserve backslash # newline sequences. ac_word_re=[_$as_cr_Letters][_$as_cr_alnum]* sed -n ' s/.\{148\}/&'"$ac_delim"'/g t rset :rset s/^[ ]*#[ ]*define[ ][ ]*/ / t def d :def s/\\$// t bsnl s/["\\]/\\&/g s/^ \('"$ac_word_re"'\)\(([^()]*)\)[ ]*\(.*\)/P["\1"]="\2"\ D["\1"]=" \3"/p s/^ \('"$ac_word_re"'\)[ ]*\(.*\)/D["\1"]=" \2"/p d :bsnl s/["\\]/\\&/g s/^ \('"$ac_word_re"'\)\(([^()]*)\)[ ]*\(.*\)/P["\1"]="\2"\ D["\1"]=" \3\\\\\\n"\\/p t cont s/^ \('"$ac_word_re"'\)[ ]*\(.*\)/D["\1"]=" \2\\\\\\n"\\/p t cont d :cont n s/.\{148\}/&'"$ac_delim"'/g t clear :clear s/\\$// t bsnlc s/["\\]/\\&/g; s/^/"/; s/$/"/p d :bsnlc s/["\\]/\\&/g; s/^/"/; s/$/\\\\\\n"\\/p b cont ' >$CONFIG_STATUS || ac_write_fail=1 cat >>$CONFIG_STATUS <<_ACEOF || ac_write_fail=1 for (key in D) D_is_set[key] = 1 FS = "" } /^[\t ]*#[\t ]*(define|undef)[\t ]+$ac_word_re([\t (]|\$)/ { line = \$ 0 split(line, arg, " ") if (arg[1] == "#") { defundef = arg[2] mac1 = arg[3] } else { defundef = substr(arg[1], 2) mac1 = arg[2] } split(mac1, mac2, "(") #) macro = mac2[1] prefix = substr(line, 1, index(line, defundef) - 1) if (D_is_set[macro]) { # Preserve the white space surrounding the "#". print prefix "define", macro P[macro] D[macro] next } else { # Replace #undef with comments. This is necessary, for example, # in the case of _POSIX_SOURCE, which is predefined and required # on some systems where configure will not decide to define it. if (defundef == "undef") { print "/*", prefix defundef, macro, "*/" next } } } { print } _ACAWK _ACEOF cat >>$CONFIG_STATUS <<\_ACEOF || ac_write_fail=1 as_fn_error $? "could not setup config headers machinery" "$LINENO" 5 fi # test -n "$CONFIG_HEADERS" eval set X " :F $CONFIG_FILES :H $CONFIG_HEADERS :C $CONFIG_COMMANDS" shift for ac_tag do case $ac_tag in :[FHLC]) ac_mode=$ac_tag; continue;; esac case $ac_mode$ac_tag in :[FHL]*:*);; :L* | :C*:*) as_fn_error $? "invalid tag \`$ac_tag'" "$LINENO" 5;; :[FH]-) ac_tag=-:-;; :[FH]*) ac_tag=$ac_tag:$ac_tag.in;; esac ac_save_IFS=$IFS IFS=: set x $ac_tag IFS=$ac_save_IFS shift ac_file=$1 shift case $ac_mode in :L) ac_source=$1;; :[FH]) ac_file_inputs= for ac_f do case $ac_f in -) ac_f="$ac_tmp/stdin";; *) # Look for the file first in the build tree, then in the source tree # (if the path is not absolute). The absolute path cannot be DOS-style, # because $ac_f cannot contain `:'. test -f "$ac_f" || case $ac_f in [\\/$]*) false;; *) test -f "$srcdir/$ac_f" && ac_f="$srcdir/$ac_f";; esac || as_fn_error 1 "cannot find input file: \`$ac_f'" "$LINENO" 5;; esac case $ac_f in *\'*) ac_f=`$as_echo "$ac_f" | sed "s/'/'\\\\\\\\''/g"`;; esac as_fn_append ac_file_inputs " '$ac_f'" done # Let's still pretend it is `configure' which instantiates (i.e., don't # use $as_me), people would be surprised to read: # /* config.h. Generated by config.status. */ configure_input='Generated from '` $as_echo "$*" | sed 's|^[^:]*/||;s|:[^:]*/|, |g' `' by configure.' if test x"$ac_file" != x-; then configure_input="$ac_file. $configure_input" { $as_echo "$as_me:${as_lineno-$LINENO}: creating $ac_file" >&5 $as_echo "$as_me: creating $ac_file" >&6;} fi # Neutralize special characters interpreted by sed in replacement strings. case $configure_input in #( *\&* | *\|* | *\\* ) ac_sed_conf_input=`$as_echo "$configure_input" | sed 's/[\\\\&|]/\\\\&/g'`;; #( *) ac_sed_conf_input=$configure_input;; esac case $ac_tag in *:-:* | *:-) cat >"$ac_tmp/stdin" \ || as_fn_error $? "could not create $ac_file" "$LINENO" 5 ;; esac ;; esac ac_dir=`$as_dirname -- "$ac_file" || $as_expr X"$ac_file" : 'X\(.*[^/]\)//*[^/][^/]*/*$' \| \ X"$ac_file" : 'X\(//\)[^/]' \| \ X"$ac_file" : 'X\(//\)$' \| \ X"$ac_file" : 'X\(/\)' \| . 2>/dev/null || $as_echo X"$ac_file" | sed '/^X\(.*[^/]\)\/\/*[^/][^/]*\/*$/{ s//\1/ q } /^X\(\/\/\)[^/].*/{ s//\1/ q } /^X\(\/\/\)$/{ s//\1/ q } /^X\(\/\).*/{ s//\1/ q } s/.*/./; q'` as_dir="$ac_dir"; as_fn_mkdir_p ac_builddir=. case "$ac_dir" in .) ac_dir_suffix= ac_top_builddir_sub=. ac_top_build_prefix= ;; *) ac_dir_suffix=/`$as_echo "$ac_dir" | sed 's|^\.[\\/]||'` # A ".." for each directory in $ac_dir_suffix. ac_top_builddir_sub=`$as_echo "$ac_dir_suffix" | sed 's|/[^\\/]*|/..|g;s|/||'` case $ac_top_builddir_sub in "") ac_top_builddir_sub=. ac_top_build_prefix= ;; *) ac_top_build_prefix=$ac_top_builddir_sub/ ;; esac ;; esac ac_abs_top_builddir=$ac_pwd ac_abs_builddir=$ac_pwd$ac_dir_suffix # for backward compatibility: ac_top_builddir=$ac_top_build_prefix case $srcdir in .) # We are building in place. ac_srcdir=. ac_top_srcdir=$ac_top_builddir_sub ac_abs_top_srcdir=$ac_pwd ;; [\\/]* | ?:[\\/]* ) # Absolute name. ac_srcdir=$srcdir$ac_dir_suffix; ac_top_srcdir=$srcdir ac_abs_top_srcdir=$srcdir ;; *) # Relative name. ac_srcdir=$ac_top_build_prefix$srcdir$ac_dir_suffix ac_top_srcdir=$ac_top_build_prefix$srcdir ac_abs_top_srcdir=$ac_pwd/$srcdir ;; esac ac_abs_srcdir=$ac_abs_top_srcdir$ac_dir_suffix case $ac_mode in :F) # # CONFIG_FILE # case $INSTALL in [\\/$]* | ?:[\\/]* ) ac_INSTALL=$INSTALL ;; *) ac_INSTALL=$ac_top_build_prefix$INSTALL ;; esac ac_MKDIR_P=$MKDIR_P case $MKDIR_P in [\\/$]* | ?:[\\/]* ) ;; */*) ac_MKDIR_P=$ac_top_build_prefix$MKDIR_P ;; esac _ACEOF cat >>$CONFIG_STATUS <<\_ACEOF || ac_write_fail=1 # If the template does not know about datarootdir, expand it. # FIXME: This hack should be removed a few years after 2.60. ac_datarootdir_hack=; ac_datarootdir_seen= ac_sed_dataroot=' /datarootdir/ { p q } /@datadir@/p /@docdir@/p /@infodir@/p /@localedir@/p /@mandir@/p' case `eval "sed -n \"\$ac_sed_dataroot\" $ac_file_inputs"` in *datarootdir*) ac_datarootdir_seen=yes;; *@datadir@*|*@docdir@*|*@infodir@*|*@localedir@*|*@mandir@*) { $as_echo "$as_me:${as_lineno-$LINENO}: WARNING: $ac_file_inputs seems to ignore the --datarootdir setting" >&5 $as_echo "$as_me: WARNING: $ac_file_inputs seems to ignore the --datarootdir setting" >&2;} _ACEOF cat >>$CONFIG_STATUS <<_ACEOF || ac_write_fail=1 ac_datarootdir_hack=' s&@datadir@&$datadir&g s&@docdir@&$docdir&g s&@infodir@&$infodir&g s&@localedir@&$localedir&g s&@mandir@&$mandir&g s&\\\${datarootdir}&$datarootdir&g' ;; esac _ACEOF # Neutralize VPATH when `$srcdir' = `.'. # Shell code in configure.ac might set extrasub. # FIXME: do we really want to maintain this feature? cat >>$CONFIG_STATUS <<_ACEOF || ac_write_fail=1 ac_sed_extra="$ac_vpsub $extrasub _ACEOF cat >>$CONFIG_STATUS <<\_ACEOF || ac_write_fail=1 :t /@[a-zA-Z_][a-zA-Z_0-9]*@/!b s|@configure_input@|$ac_sed_conf_input|;t t s&@top_builddir@&$ac_top_builddir_sub&;t t s&@top_build_prefix@&$ac_top_build_prefix&;t t s&@srcdir@&$ac_srcdir&;t t s&@abs_srcdir@&$ac_abs_srcdir&;t t s&@top_srcdir@&$ac_top_srcdir&;t t s&@abs_top_srcdir@&$ac_abs_top_srcdir&;t t s&@builddir@&$ac_builddir&;t t s&@abs_builddir@&$ac_abs_builddir&;t t s&@abs_top_builddir@&$ac_abs_top_builddir&;t t s&@INSTALL@&$ac_INSTALL&;t t s&@MKDIR_P@&$ac_MKDIR_P&;t t $ac_datarootdir_hack " eval sed \"\$ac_sed_extra\" "$ac_file_inputs" | $AWK -f "$ac_tmp/subs.awk" \ >$ac_tmp/out || as_fn_error $? "could not create $ac_file" "$LINENO" 5 test -z "$ac_datarootdir_hack$ac_datarootdir_seen" && { ac_out=`sed -n '/\${datarootdir}/p' "$ac_tmp/out"`; test -n "$ac_out"; } && { ac_out=`sed -n '/^[ ]*datarootdir[ ]*:*=/p' \ "$ac_tmp/out"`; test -z "$ac_out"; } && { $as_echo "$as_me:${as_lineno-$LINENO}: WARNING: $ac_file contains a reference to the variable \`datarootdir' which seems to be undefined. Please make sure it is defined" >&5 $as_echo "$as_me: WARNING: $ac_file contains a reference to the variable \`datarootdir' which seems to be undefined. Please make sure it is defined" >&2;} rm -f "$ac_tmp/stdin" case $ac_file in -) cat "$ac_tmp/out" && rm -f "$ac_tmp/out";; *) rm -f "$ac_file" && mv "$ac_tmp/out" "$ac_file";; esac \ || as_fn_error $? "could not create $ac_file" "$LINENO" 5 ;; :H) # # CONFIG_HEADER # if test x"$ac_file" != x-; then { $as_echo "/* $configure_input */" \ && eval '$AWK -f "$ac_tmp/defines.awk"' "$ac_file_inputs" } >"$ac_tmp/config.h" \ || as_fn_error $? "could not create $ac_file" "$LINENO" 5 if diff "$ac_file" "$ac_tmp/config.h" >/dev/null 2>&1; then { $as_echo "$as_me:${as_lineno-$LINENO}: $ac_file is unchanged" >&5 $as_echo "$as_me: $ac_file is unchanged" >&6;} else rm -f "$ac_file" mv "$ac_tmp/config.h" "$ac_file" \ || as_fn_error $? "could not create $ac_file" "$LINENO" 5 fi else $as_echo "/* $configure_input */" \ && eval '$AWK -f "$ac_tmp/defines.awk"' "$ac_file_inputs" \ || as_fn_error $? "could not create -" "$LINENO" 5 fi # Compute "$ac_file"'s index in $config_headers. _am_arg="$ac_file" _am_stamp_count=1 for _am_header in $config_headers :; do case $_am_header in $_am_arg | $_am_arg:* ) break ;; * ) _am_stamp_count=`expr $_am_stamp_count + 1` ;; esac done echo "timestamp for $_am_arg" >`$as_dirname -- "$_am_arg" || $as_expr X"$_am_arg" : 'X\(.*[^/]\)//*[^/][^/]*/*$' \| \ X"$_am_arg" : 'X\(//\)[^/]' \| \ X"$_am_arg" : 'X\(//\)$' \| \ X"$_am_arg" : 'X\(/\)' \| . 2>/dev/null || $as_echo X"$_am_arg" | sed '/^X\(.*[^/]\)\/\/*[^/][^/]*\/*$/{ s//\1/ q } /^X\(\/\/\)[^/].*/{ s//\1/ q } /^X\(\/\/\)$/{ s//\1/ q } /^X\(\/\).*/{ s//\1/ q } s/.*/./; q'`/stamp-h$_am_stamp_count ;; :C) { $as_echo "$as_me:${as_lineno-$LINENO}: executing $ac_file commands" >&5 $as_echo "$as_me: executing $ac_file commands" >&6;} ;; esac case $ac_file$ac_mode in "depfiles":C) test x"$AMDEP_TRUE" != x"" || { # Autoconf 2.62 quotes --file arguments for eval, but not when files # are listed without --file. Let's play safe and only enable the eval # if we detect the quoting. case $CONFIG_FILES in *\'*) eval set x "$CONFIG_FILES" ;; *) set x $CONFIG_FILES ;; esac shift for mf do # Strip MF so we end up with the name of the file. mf=`echo "$mf" | sed -e 's/:.*$//'` # Check whether this is an Automake generated Makefile or not. # We used to match only the files named `Makefile.in', but # some people rename them; so instead we look at the file content. # Grep'ing the first line is not enough: some people post-process # each Makefile.in and add a new line on top of each file to say so. # Grep'ing the whole file is not good either: AIX grep has a line # limit of 2048, but all sed's we know have understand at least 4000. if sed -n 's,^#.*generated by automake.*,X,p' "$mf" | grep X >/dev/null 2>&1; then dirpart=`$as_dirname -- "$mf" || $as_expr X"$mf" : 'X\(.*[^/]\)//*[^/][^/]*/*$' \| \ X"$mf" : 'X\(//\)[^/]' \| \ X"$mf" : 'X\(//\)$' \| \ X"$mf" : 'X\(/\)' \| . 2>/dev/null || $as_echo X"$mf" | sed '/^X\(.*[^/]\)\/\/*[^/][^/]*\/*$/{ s//\1/ q } /^X\(\/\/\)[^/].*/{ s//\1/ q } /^X\(\/\/\)$/{ s//\1/ q } /^X\(\/\).*/{ s//\1/ q } s/.*/./; q'` else continue fi # Extract the definition of DEPDIR, am__include, and am__quote # from the Makefile without running `make'. DEPDIR=`sed -n 's/^DEPDIR = //p' < "$mf"` test -z "$DEPDIR" && continue am__include=`sed -n 's/^am__include = //p' < "$mf"` test -z "am__include" && continue am__quote=`sed -n 's/^am__quote = //p' < "$mf"` # When using ansi2knr, U may be empty or an underscore; expand it U=`sed -n 's/^U = //p' < "$mf"` # Find all dependency output files, they are included files with # $(DEPDIR) in their names. We invoke sed twice because it is the # simplest approach to changing $(DEPDIR) to its actual value in the # expansion. for file in `sed -n " s/^$am__include $am__quote\(.*(DEPDIR).*\)$am__quote"'$/\1/p' <"$mf" | \ sed -e 's/\$(DEPDIR)/'"$DEPDIR"'/g' -e 's/\$U/'"$U"'/g'`; do # Make sure the directory exists. test -f "$dirpart/$file" && continue fdir=`$as_dirname -- "$file" || $as_expr X"$file" : 'X\(.*[^/]\)//*[^/][^/]*/*$' \| \ X"$file" : 'X\(//\)[^/]' \| \ X"$file" : 'X\(//\)$' \| \ X"$file" : 'X\(/\)' \| . 2>/dev/null || $as_echo X"$file" | sed '/^X\(.*[^/]\)\/\/*[^/][^/]*\/*$/{ s//\1/ q } /^X\(\/\/\)[^/].*/{ s//\1/ q } /^X\(\/\/\)$/{ s//\1/ q } /^X\(\/\).*/{ s//\1/ q } s/.*/./; q'` as_dir=$dirpart/$fdir; as_fn_mkdir_p # echo "creating $dirpart/$file" echo '# dummy' > "$dirpart/$file" done done } ;; "libtool":C) # See if we are running on zsh, and set the options which allow our # commands through without removal of \ escapes. if test -n "${ZSH_VERSION+set}" ; then setopt NO_GLOB_SUBST fi cfgfile="${ofile}T" trap "$RM \"$cfgfile\"; exit 1" 1 2 15 $RM "$cfgfile" cat <<_LT_EOF >> "$cfgfile" #! $SHELL # `$ECHO "$ofile" | sed 's%^.*/%%'` - Provide generalized library-building support services. # Generated automatically by $as_me ($PACKAGE$TIMESTAMP) $VERSION # Libtool was configured on host `(hostname || uname -n) 2>/dev/null | sed 1q`: # NOTE: Changes made to this file will be lost: look at ltmain.sh. # # Copyright (C) 1996, 1997, 1998, 1999, 2000, 2001, 2003, 2004, 2005, # 2006, 2007, 2008, 2009, 2010, 2011 Free Software # Foundation, Inc. # Written by Gordon Matzigkeit, 1996 # # This file is part of GNU Libtool. # # GNU Libtool is free software; you can redistribute it and/or # modify it under the terms of the GNU General Public License as # published by the Free Software Foundation; either version 2 of # the License, or (at your option) any later version. # # As a special exception to the GNU General Public License, # if you distribute this file as part of a program or library that # is built using GNU Libtool, you may include this file under the # same distribution terms that you use for the rest of that program. # # GNU Libtool is distributed in the hope that it will be useful, # but WITHOUT ANY WARRANTY; without even the implied warranty of # MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the # GNU General Public License for more details. # # You should have received a copy of the GNU General Public License # along with GNU Libtool; see the file COPYING. If not, a copy # can be downloaded from http://www.gnu.org/licenses/gpl.html, or # obtained by writing to the Free Software Foundation, Inc., # 51 Franklin Street, Fifth Floor, Boston, MA 02110-1301, USA. # The names of the tagged configurations supported by this script. available_tags="F77 " # ### BEGIN LIBTOOL CONFIG # Whether or not to build shared libraries. build_libtool_libs=$enable_shared # Assembler program. AS=$lt_AS # DLL creation program. DLLTOOL=$lt_DLLTOOL # Object dumper program. OBJDUMP=$lt_OBJDUMP # Which release of libtool.m4 was used? macro_version=$macro_version macro_revision=$macro_revision # Whether or not to build static libraries. build_old_libs=$enable_static # What type of objects to build. pic_mode=$pic_mode # Whether or not to optimize for fast installation. fast_install=$enable_fast_install # Shell to use when invoking shell scripts. SHELL=$lt_SHELL # An echo program that protects backslashes. ECHO=$lt_ECHO # The PATH separator for the build system. PATH_SEPARATOR=$lt_PATH_SEPARATOR # The host system. host_alias=$host_alias host=$host host_os=$host_os # The build system. build_alias=$build_alias build=$build build_os=$build_os # A sed program that does not truncate output. SED=$lt_SED # Sed that helps us avoid accidentally triggering echo(1) options like -n. Xsed="\$SED -e 1s/^X//" # A grep program that handles long lines. GREP=$lt_GREP # An ERE matcher. EGREP=$lt_EGREP # A literal string matcher. FGREP=$lt_FGREP # A BSD- or MS-compatible name lister. NM=$lt_NM # Whether we need soft or hard links. LN_S=$lt_LN_S # What is the maximum length of a command? max_cmd_len=$max_cmd_len # Object file suffix (normally "o"). objext=$ac_objext # Executable file suffix (normally ""). exeext=$exeext # whether the shell understands "unset". lt_unset=$lt_unset # turn spaces into newlines. SP2NL=$lt_lt_SP2NL # turn newlines into spaces. NL2SP=$lt_lt_NL2SP # convert \$build file names to \$host format. to_host_file_cmd=$lt_cv_to_host_file_cmd # convert \$build files to toolchain format. to_tool_file_cmd=$lt_cv_to_tool_file_cmd # Method to check whether dependent libraries are shared objects. deplibs_check_method=$lt_deplibs_check_method # Command to use when deplibs_check_method = "file_magic". file_magic_cmd=$lt_file_magic_cmd # How to find potential files when deplibs_check_method = "file_magic". file_magic_glob=$lt_file_magic_glob # Find potential files using nocaseglob when deplibs_check_method = "file_magic". want_nocaseglob=$lt_want_nocaseglob # Command to associate shared and link libraries. sharedlib_from_linklib_cmd=$lt_sharedlib_from_linklib_cmd # The archiver. AR=$lt_AR # Flags to create an archive. AR_FLAGS=$lt_AR_FLAGS # How to feed a file listing to the archiver. archiver_list_spec=$lt_archiver_list_spec # A symbol stripping program. STRIP=$lt_STRIP # Commands used to install an old-style archive. RANLIB=$lt_RANLIB old_postinstall_cmds=$lt_old_postinstall_cmds old_postuninstall_cmds=$lt_old_postuninstall_cmds # Whether to use a lock for old archive extraction. lock_old_archive_extraction=$lock_old_archive_extraction # A C compiler. LTCC=$lt_CC # LTCC compiler flags. LTCFLAGS=$lt_CFLAGS # Take the output of nm and produce a listing of raw symbols and C names. global_symbol_pipe=$lt_lt_cv_sys_global_symbol_pipe # Transform the output of nm in a proper C declaration. global_symbol_to_cdecl=$lt_lt_cv_sys_global_symbol_to_cdecl # Transform the output of nm in a C name address pair. global_symbol_to_c_name_address=$lt_lt_cv_sys_global_symbol_to_c_name_address # Transform the output of nm in a C name address pair when lib prefix is needed. global_symbol_to_c_name_address_lib_prefix=$lt_lt_cv_sys_global_symbol_to_c_name_address_lib_prefix # Specify filename containing input files for \$NM. nm_file_list_spec=$lt_nm_file_list_spec # The root where to search for dependent libraries,and in which our libraries should be installed. lt_sysroot=$lt_sysroot # The name of the directory that contains temporary libtool files. objdir=$objdir # Used to examine libraries when file_magic_cmd begins with "file". MAGIC_CMD=$MAGIC_CMD # Must we lock files when doing compilation? need_locks=$lt_need_locks # Manifest tool. MANIFEST_TOOL=$lt_MANIFEST_TOOL # Tool to manipulate archived DWARF debug symbol files on Mac OS X. DSYMUTIL=$lt_DSYMUTIL # Tool to change global to local symbols on Mac OS X. NMEDIT=$lt_NMEDIT # Tool to manipulate fat objects and archives on Mac OS X. LIPO=$lt_LIPO # ldd/readelf like tool for Mach-O binaries on Mac OS X. OTOOL=$lt_OTOOL # ldd/readelf like tool for 64 bit Mach-O binaries on Mac OS X 10.4. OTOOL64=$lt_OTOOL64 # Old archive suffix (normally "a"). libext=$libext # Shared library suffix (normally ".so"). shrext_cmds=$lt_shrext_cmds # The commands to extract the exported symbol list from a shared archive. extract_expsyms_cmds=$lt_extract_expsyms_cmds # Variables whose values should be saved in libtool wrapper scripts and # restored at link time. variables_saved_for_relink=$lt_variables_saved_for_relink # Do we need the "lib" prefix for modules? need_lib_prefix=$need_lib_prefix # Do we need a version for libraries? need_version=$need_version # Library versioning type. version_type=$version_type # Shared library runtime path variable. runpath_var=$runpath_var # Shared library path variable. shlibpath_var=$shlibpath_var # Is shlibpath searched before the hard-coded library search path? shlibpath_overrides_runpath=$shlibpath_overrides_runpath # Format of library name prefix. libname_spec=$lt_libname_spec # List of archive names. First name is the real one, the rest are links. # The last name is the one that the linker finds with -lNAME library_names_spec=$lt_library_names_spec # The coded name of the library, if different from the real name. soname_spec=$lt_soname_spec # Permission mode override for installation of shared libraries. install_override_mode=$lt_install_override_mode # Command to use after installation of a shared archive. postinstall_cmds=$lt_postinstall_cmds # Command to use after uninstallation of a shared archive. postuninstall_cmds=$lt_postuninstall_cmds # Commands used to finish a libtool library installation in a directory. finish_cmds=$lt_finish_cmds # As "finish_cmds", except a single script fragment to be evaled but # not shown. finish_eval=$lt_finish_eval # Whether we should hardcode library paths into libraries. hardcode_into_libs=$hardcode_into_libs # Compile-time system search path for libraries. sys_lib_search_path_spec=$lt_sys_lib_search_path_spec # Run-time system search path for libraries. sys_lib_dlsearch_path_spec=$lt_sys_lib_dlsearch_path_spec # Whether dlopen is supported. dlopen_support=$enable_dlopen # Whether dlopen of programs is supported. dlopen_self=$enable_dlopen_self # Whether dlopen of statically linked programs is supported. dlopen_self_static=$enable_dlopen_self_static # Commands to strip libraries. old_striplib=$lt_old_striplib striplib=$lt_striplib # The linker used to build libraries. LD=$lt_LD # How to create reloadable object files. reload_flag=$lt_reload_flag reload_cmds=$lt_reload_cmds # Commands used to build an old-style archive. old_archive_cmds=$lt_old_archive_cmds # A language specific compiler. CC=$lt_compiler # Is the compiler the GNU compiler? with_gcc=$GCC # Compiler flag to turn off builtin functions. no_builtin_flag=$lt_lt_prog_compiler_no_builtin_flag # Additional compiler flags for building library objects. pic_flag=$lt_lt_prog_compiler_pic # How to pass a linker flag through the compiler. wl=$lt_lt_prog_compiler_wl # Compiler flag to prevent dynamic linking. link_static_flag=$lt_lt_prog_compiler_static # Does compiler simultaneously support -c and -o options? compiler_c_o=$lt_lt_cv_prog_compiler_c_o # Whether or not to add -lc for building shared libraries. build_libtool_need_lc=$archive_cmds_need_lc # Whether or not to disallow shared libs when runtime libs are static. allow_libtool_libs_with_static_runtimes=$enable_shared_with_static_runtimes # Compiler flag to allow reflexive dlopens. export_dynamic_flag_spec=$lt_export_dynamic_flag_spec # Compiler flag to generate shared objects directly from archives. whole_archive_flag_spec=$lt_whole_archive_flag_spec # Whether the compiler copes with passing no objects directly. compiler_needs_object=$lt_compiler_needs_object # Create an old-style archive from a shared archive. old_archive_from_new_cmds=$lt_old_archive_from_new_cmds # Create a temporary old-style archive to link instead of a shared archive. old_archive_from_expsyms_cmds=$lt_old_archive_from_expsyms_cmds # Commands used to build a shared archive. archive_cmds=$lt_archive_cmds archive_expsym_cmds=$lt_archive_expsym_cmds # Commands used to build a loadable module if different from building # a shared archive. module_cmds=$lt_module_cmds module_expsym_cmds=$lt_module_expsym_cmds # Whether we are building with GNU ld or not. with_gnu_ld=$lt_with_gnu_ld # Flag that allows shared libraries with undefined symbols to be built. allow_undefined_flag=$lt_allow_undefined_flag # Flag that enforces no undefined symbols. no_undefined_flag=$lt_no_undefined_flag # Flag to hardcode \$libdir into a binary during linking. # This must work even if \$libdir does not exist hardcode_libdir_flag_spec=$lt_hardcode_libdir_flag_spec # Whether we need a single "-rpath" flag with a separated argument. hardcode_libdir_separator=$lt_hardcode_libdir_separator # Set to "yes" if using DIR/libNAME\${shared_ext} during linking hardcodes # DIR into the resulting binary. hardcode_direct=$hardcode_direct # Set to "yes" if using DIR/libNAME\${shared_ext} during linking hardcodes # DIR into the resulting binary and the resulting library dependency is # "absolute",i.e impossible to change by setting \${shlibpath_var} if the # library is relocated. hardcode_direct_absolute=$hardcode_direct_absolute # Set to "yes" if using the -LDIR flag during linking hardcodes DIR # into the resulting binary. hardcode_minus_L=$hardcode_minus_L # Set to "yes" if using SHLIBPATH_VAR=DIR during linking hardcodes DIR # into the resulting binary. hardcode_shlibpath_var=$hardcode_shlibpath_var # Set to "yes" if building a shared library automatically hardcodes DIR # into the library and all subsequent libraries and executables linked # against it. hardcode_automatic=$hardcode_automatic # Set to yes if linker adds runtime paths of dependent libraries # to runtime path list. inherit_rpath=$inherit_rpath # Whether libtool must link a program against all its dependency libraries. link_all_deplibs=$link_all_deplibs # Set to "yes" if exported symbols are required. always_export_symbols=$always_export_symbols # The commands to list exported symbols. export_symbols_cmds=$lt_export_symbols_cmds # Symbols that should not be listed in the preloaded symbols. exclude_expsyms=$lt_exclude_expsyms # Symbols that must always be exported. include_expsyms=$lt_include_expsyms # Commands necessary for linking programs (against libraries) with templates. prelink_cmds=$lt_prelink_cmds # Commands necessary for finishing linking programs. postlink_cmds=$lt_postlink_cmds # Specify filename containing input files. file_list_spec=$lt_file_list_spec # How to hardcode a shared library path into an executable. hardcode_action=$hardcode_action # ### END LIBTOOL CONFIG _LT_EOF case $host_os in aix3*) cat <<\_LT_EOF >> "$cfgfile" # AIX sometimes has problems with the GCC collect2 program. For some # reason, if we set the COLLECT_NAMES environment variable, the problems # vanish in a puff of smoke. if test "X${COLLECT_NAMES+set}" != Xset; then COLLECT_NAMES= export COLLECT_NAMES fi _LT_EOF ;; esac ltmain="$ac_aux_dir/ltmain.sh" # We use sed instead of cat because bash on DJGPP gets confused if # if finds mixed CR/LF and LF-only lines. Since sed operates in # text mode, it properly converts lines to CR/LF. This bash problem # is reportedly fixed, but why not run on old versions too? sed '$q' "$ltmain" >> "$cfgfile" \ || (rm -f "$cfgfile"; exit 1) if test x"$xsi_shell" = xyes; then sed -e '/^func_dirname ()$/,/^} # func_dirname /c\ func_dirname ()\ {\ \ case ${1} in\ \ */*) func_dirname_result="${1%/*}${2}" ;;\ \ * ) func_dirname_result="${3}" ;;\ \ esac\ } # Extended-shell func_dirname implementation' "$cfgfile" > $cfgfile.tmp \ && mv -f "$cfgfile.tmp" "$cfgfile" \ || (rm -f "$cfgfile" && cp "$cfgfile.tmp" "$cfgfile" && rm -f "$cfgfile.tmp") test 0 -eq $? || _lt_function_replace_fail=: sed -e '/^func_basename ()$/,/^} # func_basename /c\ func_basename ()\ {\ \ func_basename_result="${1##*/}"\ } # Extended-shell func_basename implementation' "$cfgfile" > $cfgfile.tmp \ && mv -f "$cfgfile.tmp" "$cfgfile" \ || (rm -f "$cfgfile" && cp "$cfgfile.tmp" "$cfgfile" && rm -f "$cfgfile.tmp") test 0 -eq $? || _lt_function_replace_fail=: sed -e '/^func_dirname_and_basename ()$/,/^} # func_dirname_and_basename /c\ func_dirname_and_basename ()\ {\ \ case ${1} in\ \ */*) func_dirname_result="${1%/*}${2}" ;;\ \ * ) func_dirname_result="${3}" ;;\ \ esac\ \ func_basename_result="${1##*/}"\ } # Extended-shell func_dirname_and_basename implementation' "$cfgfile" > $cfgfile.tmp \ && mv -f "$cfgfile.tmp" "$cfgfile" \ || (rm -f "$cfgfile" && cp "$cfgfile.tmp" "$cfgfile" && rm -f "$cfgfile.tmp") test 0 -eq $? || _lt_function_replace_fail=: sed -e '/^func_stripname ()$/,/^} # func_stripname /c\ func_stripname ()\ {\ \ # pdksh 5.2.14 does not do ${X%$Y} correctly if both X and Y are\ \ # positional parameters, so assign one to ordinary parameter first.\ \ func_stripname_result=${3}\ \ func_stripname_result=${func_stripname_result#"${1}"}\ \ func_stripname_result=${func_stripname_result%"${2}"}\ } # Extended-shell func_stripname implementation' "$cfgfile" > $cfgfile.tmp \ && mv -f "$cfgfile.tmp" "$cfgfile" \ || (rm -f "$cfgfile" && cp "$cfgfile.tmp" "$cfgfile" && rm -f "$cfgfile.tmp") test 0 -eq $? || _lt_function_replace_fail=: sed -e '/^func_split_long_opt ()$/,/^} # func_split_long_opt /c\ func_split_long_opt ()\ {\ \ func_split_long_opt_name=${1%%=*}\ \ func_split_long_opt_arg=${1#*=}\ } # Extended-shell func_split_long_opt implementation' "$cfgfile" > $cfgfile.tmp \ && mv -f "$cfgfile.tmp" "$cfgfile" \ || (rm -f "$cfgfile" && cp "$cfgfile.tmp" "$cfgfile" && rm -f "$cfgfile.tmp") test 0 -eq $? || _lt_function_replace_fail=: sed -e '/^func_split_short_opt ()$/,/^} # func_split_short_opt /c\ func_split_short_opt ()\ {\ \ func_split_short_opt_arg=${1#??}\ \ func_split_short_opt_name=${1%"$func_split_short_opt_arg"}\ } # Extended-shell func_split_short_opt implementation' "$cfgfile" > $cfgfile.tmp \ && mv -f "$cfgfile.tmp" "$cfgfile" \ || (rm -f "$cfgfile" && cp "$cfgfile.tmp" "$cfgfile" && rm -f "$cfgfile.tmp") test 0 -eq $? || _lt_function_replace_fail=: sed -e '/^func_lo2o ()$/,/^} # func_lo2o /c\ func_lo2o ()\ {\ \ case ${1} in\ \ *.lo) func_lo2o_result=${1%.lo}.${objext} ;;\ \ *) func_lo2o_result=${1} ;;\ \ esac\ } # Extended-shell func_lo2o implementation' "$cfgfile" > $cfgfile.tmp \ && mv -f "$cfgfile.tmp" "$cfgfile" \ || (rm -f "$cfgfile" && cp "$cfgfile.tmp" "$cfgfile" && rm -f "$cfgfile.tmp") test 0 -eq $? || _lt_function_replace_fail=: sed -e '/^func_xform ()$/,/^} # func_xform /c\ func_xform ()\ {\ func_xform_result=${1%.*}.lo\ } # Extended-shell func_xform implementation' "$cfgfile" > $cfgfile.tmp \ && mv -f "$cfgfile.tmp" "$cfgfile" \ || (rm -f "$cfgfile" && cp "$cfgfile.tmp" "$cfgfile" && rm -f "$cfgfile.tmp") test 0 -eq $? || _lt_function_replace_fail=: sed -e '/^func_arith ()$/,/^} # func_arith /c\ func_arith ()\ {\ func_arith_result=$(( $* ))\ } # Extended-shell func_arith implementation' "$cfgfile" > $cfgfile.tmp \ && mv -f "$cfgfile.tmp" "$cfgfile" \ || (rm -f "$cfgfile" && cp "$cfgfile.tmp" "$cfgfile" && rm -f "$cfgfile.tmp") test 0 -eq $? || _lt_function_replace_fail=: sed -e '/^func_len ()$/,/^} # func_len /c\ func_len ()\ {\ func_len_result=${#1}\ } # Extended-shell func_len implementation' "$cfgfile" > $cfgfile.tmp \ && mv -f "$cfgfile.tmp" "$cfgfile" \ || (rm -f "$cfgfile" && cp "$cfgfile.tmp" "$cfgfile" && rm -f "$cfgfile.tmp") test 0 -eq $? || _lt_function_replace_fail=: fi if test x"$lt_shell_append" = xyes; then sed -e '/^func_append ()$/,/^} # func_append /c\ func_append ()\ {\ eval "${1}+=\\${2}"\ } # Extended-shell func_append implementation' "$cfgfile" > $cfgfile.tmp \ && mv -f "$cfgfile.tmp" "$cfgfile" \ || (rm -f "$cfgfile" && cp "$cfgfile.tmp" "$cfgfile" && rm -f "$cfgfile.tmp") test 0 -eq $? || _lt_function_replace_fail=: sed -e '/^func_append_quoted ()$/,/^} # func_append_quoted /c\ func_append_quoted ()\ {\ \ func_quote_for_eval "${2}"\ \ eval "${1}+=\\\\ \\$func_quote_for_eval_result"\ } # Extended-shell func_append_quoted implementation' "$cfgfile" > $cfgfile.tmp \ && mv -f "$cfgfile.tmp" "$cfgfile" \ || (rm -f "$cfgfile" && cp "$cfgfile.tmp" "$cfgfile" && rm -f "$cfgfile.tmp") test 0 -eq $? || _lt_function_replace_fail=: # Save a `func_append' function call where possible by direct use of '+=' sed -e 's%func_append \([a-zA-Z_]\{1,\}\) "%\1+="%g' $cfgfile > $cfgfile.tmp \ && mv -f "$cfgfile.tmp" "$cfgfile" \ || (rm -f "$cfgfile" && cp "$cfgfile.tmp" "$cfgfile" && rm -f "$cfgfile.tmp") test 0 -eq $? || _lt_function_replace_fail=: else # Save a `func_append' function call even when '+=' is not available sed -e 's%func_append \([a-zA-Z_]\{1,\}\) "%\1="$\1%g' $cfgfile > $cfgfile.tmp \ && mv -f "$cfgfile.tmp" "$cfgfile" \ || (rm -f "$cfgfile" && cp "$cfgfile.tmp" "$cfgfile" && rm -f "$cfgfile.tmp") test 0 -eq $? || _lt_function_replace_fail=: fi if test x"$_lt_function_replace_fail" = x":"; then { $as_echo "$as_me:${as_lineno-$LINENO}: WARNING: Unable to substitute extended shell functions in $ofile" >&5 $as_echo "$as_me: WARNING: Unable to substitute extended shell functions in $ofile" >&2;} fi mv -f "$cfgfile" "$ofile" || (rm -f "$ofile" && cp "$cfgfile" "$ofile" && rm -f "$cfgfile") chmod +x "$ofile" cat <<_LT_EOF >> "$ofile" # ### BEGIN LIBTOOL TAG CONFIG: F77 # The linker used to build libraries. LD=$lt_LD_F77 # How to create reloadable object files. reload_flag=$lt_reload_flag_F77 reload_cmds=$lt_reload_cmds_F77 # Commands used to build an old-style archive. old_archive_cmds=$lt_old_archive_cmds_F77 # A language specific compiler. CC=$lt_compiler_F77 # Is the compiler the GNU compiler? with_gcc=$GCC_F77 # Compiler flag to turn off builtin functions. no_builtin_flag=$lt_lt_prog_compiler_no_builtin_flag_F77 # Additional compiler flags for building library objects. pic_flag=$lt_lt_prog_compiler_pic_F77 # How to pass a linker flag through the compiler. wl=$lt_lt_prog_compiler_wl_F77 # Compiler flag to prevent dynamic linking. link_static_flag=$lt_lt_prog_compiler_static_F77 # Does compiler simultaneously support -c and -o options? compiler_c_o=$lt_lt_cv_prog_compiler_c_o_F77 # Whether or not to add -lc for building shared libraries. build_libtool_need_lc=$archive_cmds_need_lc_F77 # Whether or not to disallow shared libs when runtime libs are static. allow_libtool_libs_with_static_runtimes=$enable_shared_with_static_runtimes_F77 # Compiler flag to allow reflexive dlopens. export_dynamic_flag_spec=$lt_export_dynamic_flag_spec_F77 # Compiler flag to generate shared objects directly from archives. whole_archive_flag_spec=$lt_whole_archive_flag_spec_F77 # Whether the compiler copes with passing no objects directly. compiler_needs_object=$lt_compiler_needs_object_F77 # Create an old-style archive from a shared archive. old_archive_from_new_cmds=$lt_old_archive_from_new_cmds_F77 # Create a temporary old-style archive to link instead of a shared archive. old_archive_from_expsyms_cmds=$lt_old_archive_from_expsyms_cmds_F77 # Commands used to build a shared archive. archive_cmds=$lt_archive_cmds_F77 archive_expsym_cmds=$lt_archive_expsym_cmds_F77 # Commands used to build a loadable module if different from building # a shared archive. module_cmds=$lt_module_cmds_F77 module_expsym_cmds=$lt_module_expsym_cmds_F77 # Whether we are building with GNU ld or not. with_gnu_ld=$lt_with_gnu_ld_F77 # Flag that allows shared libraries with undefined symbols to be built. allow_undefined_flag=$lt_allow_undefined_flag_F77 # Flag that enforces no undefined symbols. no_undefined_flag=$lt_no_undefined_flag_F77 # Flag to hardcode \$libdir into a binary during linking. # This must work even if \$libdir does not exist hardcode_libdir_flag_spec=$lt_hardcode_libdir_flag_spec_F77 # Whether we need a single "-rpath" flag with a separated argument. hardcode_libdir_separator=$lt_hardcode_libdir_separator_F77 # Set to "yes" if using DIR/libNAME\${shared_ext} during linking hardcodes # DIR into the resulting binary. hardcode_direct=$hardcode_direct_F77 # Set to "yes" if using DIR/libNAME\${shared_ext} during linking hardcodes # DIR into the resulting binary and the resulting library dependency is # "absolute",i.e impossible to change by setting \${shlibpath_var} if the # library is relocated. hardcode_direct_absolute=$hardcode_direct_absolute_F77 # Set to "yes" if using the -LDIR flag during linking hardcodes DIR # into the resulting binary. hardcode_minus_L=$hardcode_minus_L_F77 # Set to "yes" if using SHLIBPATH_VAR=DIR during linking hardcodes DIR # into the resulting binary. hardcode_shlibpath_var=$hardcode_shlibpath_var_F77 # Set to "yes" if building a shared library automatically hardcodes DIR # into the library and all subsequent libraries and executables linked # against it. hardcode_automatic=$hardcode_automatic_F77 # Set to yes if linker adds runtime paths of dependent libraries # to runtime path list. inherit_rpath=$inherit_rpath_F77 # Whether libtool must link a program against all its dependency libraries. link_all_deplibs=$link_all_deplibs_F77 # Set to "yes" if exported symbols are required. always_export_symbols=$always_export_symbols_F77 # The commands to list exported symbols. export_symbols_cmds=$lt_export_symbols_cmds_F77 # Symbols that should not be listed in the preloaded symbols. exclude_expsyms=$lt_exclude_expsyms_F77 # Symbols that must always be exported. include_expsyms=$lt_include_expsyms_F77 # Commands necessary for linking programs (against libraries) with templates. prelink_cmds=$lt_prelink_cmds_F77 # Commands necessary for finishing linking programs. postlink_cmds=$lt_postlink_cmds_F77 # Specify filename containing input files. file_list_spec=$lt_file_list_spec_F77 # How to hardcode a shared library path into an executable. hardcode_action=$hardcode_action_F77 # ### END LIBTOOL TAG CONFIG: F77 _LT_EOF ;; esac done # for ac_tag as_fn_exit 0 _ACEOF ac_clean_files=$ac_clean_files_save test $ac_write_fail = 0 || as_fn_error $? "write failure creating $CONFIG_STATUS" "$LINENO" 5 # configure is writing to config.log, and then calls config.status. # config.status does its own redirection, appending to config.log. # Unfortunately, on DOS this fails, as config.log is still kept open # by configure, so config.status won't be able to write to it; its # output is simply discarded. So we exec the FD to /dev/null, # effectively closing config.log, so it can be properly (re)opened and # appended to by config.status. When coming back to configure, we # need to make the FD available again. if test "$no_create" != yes; then ac_cs_success=: ac_config_status_args= test "$silent" = yes && ac_config_status_args="$ac_config_status_args --quiet" exec 5>/dev/null $SHELL $CONFIG_STATUS $ac_config_status_args || ac_cs_success=false exec 5>>config.log # Use ||, not &&, to avoid exiting from the if with $? = 1, which # would make configure fail if this is the last instruction. $ac_cs_success || as_fn_exit 1 fi if test -n "$ac_unrecognized_opts" && test "$enable_option_checking" != no; then { $as_echo "$as_me:${as_lineno-$LINENO}: WARNING: unrecognized options: $ac_unrecognized_opts" >&5 $as_echo "$as_me: WARNING: unrecognized options: $ac_unrecognized_opts" >&2;} fi fftw-3.3.3/AUTHORS0000644000175000017500000000072612054410050010453 00000000000000Authors of FFTW (reachable at fftw@fftw.org): Matteo Frigo Stevenj G. Johnson Stefan Kral wrote genfft-k7/*.ml*, which was added in fftw-3.0 and removed in fftw-3.2. Support for the Cell Broadband Engine was graciously donated by the IBM Austin Research Lab, which was added in fftw-3.2 and removed in fftw-3.3. Support for MIPS64 paired-single SIMD instructions was graciously donated by CodeSourcery, Inc. fftw-3.3.3/COPYING0000644000175000017500000004312212054410050010433 00000000000000 GNU GENERAL PUBLIC LICENSE Version 2, June 1991 Copyright (C) 1989, 1991 Free Software Foundation, Inc. 51 Franklin Street, Fifth Floor, Boston, MA 02110-1301 USA Everyone is permitted to copy and distribute verbatim copies of this license document, but changing it is not allowed. Preamble The licenses for most software are designed to take away your freedom to share and change it. By contrast, the GNU General Public License is intended to guarantee your freedom to share and change free software--to make sure the software is free for all its users. This General Public License applies to most of the Free Software Foundation's software and to any other program whose authors commit to using it. (Some other Free Software Foundation software is covered by the GNU Library General Public License instead.) You can apply it to your programs, too. When we speak of free software, we are referring to freedom, not price. Our General Public Licenses are designed to make sure that you have the freedom to distribute copies of free software (and charge for this service if you wish), that you receive source code or can get it if you want it, that you can change the software or use pieces of it in new free programs; and that you know you can do these things. To protect your rights, we need to make restrictions that forbid anyone to deny you these rights or to ask you to surrender the rights. These restrictions translate to certain responsibilities for you if you distribute copies of the software, or if you modify it. For example, if you distribute copies of such a program, whether gratis or for a fee, you must give the recipients all the rights that you have. You must make sure that they, too, receive or can get the source code. And you must show them these terms so they know their rights. We protect your rights with two steps: (1) copyright the software, and (2) offer you this license which gives you legal permission to copy, distribute and/or modify the software. Also, for each author's protection and ours, we want to make certain that everyone understands that there is no warranty for this free software. If the software is modified by someone else and passed on, we want its recipients to know that what they have is not the original, so that any problems introduced by others will not reflect on the original authors' reputations. Finally, any free program is threatened constantly by software patents. We wish to avoid the danger that redistributors of a free program will individually obtain patent licenses, in effect making the program proprietary. To prevent this, we have made it clear that any patent must be licensed for everyone's free use or not licensed at all. The precise terms and conditions for copying, distribution and modification follow. GNU GENERAL PUBLIC LICENSE TERMS AND CONDITIONS FOR COPYING, DISTRIBUTION AND MODIFICATION 0. This License applies to any program or other work which contains a notice placed by the copyright holder saying it may be distributed under the terms of this General Public License. The "Program", below, refers to any such program or work, and a "work based on the Program" means either the Program or any derivative work under copyright law: that is to say, a work containing the Program or a portion of it, either verbatim or with modifications and/or translated into another language. (Hereinafter, translation is included without limitation in the term "modification".) Each licensee is addressed as "you". Activities other than copying, distribution and modification are not covered by this License; they are outside its scope. The act of running the Program is not restricted, and the output from the Program is covered only if its contents constitute a work based on the Program (independent of having been made by running the Program). Whether that is true depends on what the Program does. 1. You may copy and distribute verbatim copies of the Program's source code as you receive it, in any medium, provided that you conspicuously and appropriately publish on each copy an appropriate copyright notice and disclaimer of warranty; keep intact all the notices that refer to this License and to the absence of any warranty; and give any other recipients of the Program a copy of this License along with the Program. You may charge a fee for the physical act of transferring a copy, and you may at your option offer warranty protection in exchange for a fee. 2. You may modify your copy or copies of the Program or any portion of it, thus forming a work based on the Program, and copy and distribute such modifications or work under the terms of Section 1 above, provided that you also meet all of these conditions: a) You must cause the modified files to carry prominent notices stating that you changed the files and the date of any change. b) You must cause any work that you distribute or publish, that in whole or in part contains or is derived from the Program or any part thereof, to be licensed as a whole at no charge to all third parties under the terms of this License. c) If the modified program normally reads commands interactively when run, you must cause it, when started running for such interactive use in the most ordinary way, to print or display an announcement including an appropriate copyright notice and a notice that there is no warranty (or else, saying that you provide a warranty) and that users may redistribute the program under these conditions, and telling the user how to view a copy of this License. (Exception: if the Program itself is interactive but does not normally print such an announcement, your work based on the Program is not required to print an announcement.) These requirements apply to the modified work as a whole. If identifiable sections of that work are not derived from the Program, and can be reasonably considered independent and separate works in themselves, then this License, and its terms, do not apply to those sections when you distribute them as separate works. But when you distribute the same sections as part of a whole which is a work based on the Program, the distribution of the whole must be on the terms of this License, whose permissions for other licensees extend to the entire whole, and thus to each and every part regardless of who wrote it. Thus, it is not the intent of this section to claim rights or contest your rights to work written entirely by you; rather, the intent is to exercise the right to control the distribution of derivative or collective works based on the Program. In addition, mere aggregation of another work not based on the Program with the Program (or with a work based on the Program) on a volume of a storage or distribution medium does not bring the other work under the scope of this License. 3. You may copy and distribute the Program (or a work based on it, under Section 2) in object code or executable form under the terms of Sections 1 and 2 above provided that you also do one of the following: a) Accompany it with the complete corresponding machine-readable source code, which must be distributed under the terms of Sections 1 and 2 above on a medium customarily used for software interchange; or, b) Accompany it with a written offer, valid for at least three years, to give any third party, for a charge no more than your cost of physically performing source distribution, a complete machine-readable copy of the corresponding source code, to be distributed under the terms of Sections 1 and 2 above on a medium customarily used for software interchange; or, c) Accompany it with the information you received as to the offer to distribute corresponding source code. (This alternative is allowed only for noncommercial distribution and only if you received the program in object code or executable form with such an offer, in accord with Subsection b above.) The source code for a work means the preferred form of the work for making modifications to it. For an executable work, complete source code means all the source code for all modules it contains, plus any associated interface definition files, plus the scripts used to control compilation and installation of the executable. However, as a special exception, the source code distributed need not include anything that is normally distributed (in either source or binary form) with the major components (compiler, kernel, and so on) of the operating system on which the executable runs, unless that component itself accompanies the executable. If distribution of executable or object code is made by offering access to copy from a designated place, then offering equivalent access to copy the source code from the same place counts as distribution of the source code, even though third parties are not compelled to copy the source along with the object code. 4. You may not copy, modify, sublicense, or distribute the Program except as expressly provided under this License. Any attempt otherwise to copy, modify, sublicense or distribute the Program is void, and will automatically terminate your rights under this License. However, parties who have received copies, or rights, from you under this License will not have their licenses terminated so long as such parties remain in full compliance. 5. You are not required to accept this License, since you have not signed it. However, nothing else grants you permission to modify or distribute the Program or its derivative works. These actions are prohibited by law if you do not accept this License. Therefore, by modifying or distributing the Program (or any work based on the Program), you indicate your acceptance of this License to do so, and all its terms and conditions for copying, distributing or modifying the Program or works based on it. 6. Each time you redistribute the Program (or any work based on the Program), the recipient automatically receives a license from the original licensor to copy, distribute or modify the Program subject to these terms and conditions. You may not impose any further restrictions on the recipients' exercise of the rights granted herein. You are not responsible for enforcing compliance by third parties to this License. 7. If, as a consequence of a court judgment or allegation of patent infringement or for any other reason (not limited to patent issues), conditions are imposed on you (whether by court order, agreement or otherwise) that contradict the conditions of this License, they do not excuse you from the conditions of this License. If you cannot distribute so as to satisfy simultaneously your obligations under this License and any other pertinent obligations, then as a consequence you may not distribute the Program at all. For example, if a patent license would not permit royalty-free redistribution of the Program by all those who receive copies directly or indirectly through you, then the only way you could satisfy both it and this License would be to refrain entirely from distribution of the Program. If any portion of this section is held invalid or unenforceable under any particular circumstance, the balance of the section is intended to apply and the section as a whole is intended to apply in other circumstances. It is not the purpose of this section to induce you to infringe any patents or other property right claims or to contest validity of any such claims; this section has the sole purpose of protecting the integrity of the free software distribution system, which is implemented by public license practices. Many people have made generous contributions to the wide range of software distributed through that system in reliance on consistent application of that system; it is up to the author/donor to decide if he or she is willing to distribute software through any other system and a licensee cannot impose that choice. This section is intended to make thoroughly clear what is believed to be a consequence of the rest of this License. 8. If the distribution and/or use of the Program is restricted in certain countries either by patents or by copyrighted interfaces, the original copyright holder who places the Program under this License may add an explicit geographical distribution limitation excluding those countries, so that distribution is permitted only in or among countries not thus excluded. In such case, this License incorporates the limitation as if written in the body of this License. 9. The Free Software Foundation may publish revised and/or new versions of the General Public License from time to time. Such new versions will be similar in spirit to the present version, but may differ in detail to address new problems or concerns. Each version is given a distinguishing version number. If the Program specifies a version number of this License which applies to it and "any later version", you have the option of following the terms and conditions either of that version or of any later version published by the Free Software Foundation. If the Program does not specify a version number of this License, you may choose any version ever published by the Free Software Foundation. 10. If you wish to incorporate parts of the Program into other free programs whose distribution conditions are different, write to the author to ask for permission. For software which is copyrighted by the Free Software Foundation, write to the Free Software Foundation; we sometimes make exceptions for this. Our decision will be guided by the two goals of preserving the free status of all derivatives of our free software and of promoting the sharing and reuse of software generally. NO WARRANTY 11. BECAUSE THE PROGRAM IS LICENSED FREE OF CHARGE, THERE IS NO WARRANTY FOR THE PROGRAM, TO THE EXTENT PERMITTED BY APPLICABLE LAW. EXCEPT WHEN OTHERWISE STATED IN WRITING THE COPYRIGHT HOLDERS AND/OR OTHER PARTIES PROVIDE THE PROGRAM "AS IS" WITHOUT WARRANTY OF ANY KIND, EITHER EXPRESSED OR IMPLIED, INCLUDING, BUT NOT LIMITED TO, THE IMPLIED WARRANTIES OF MERCHANTABILITY AND FITNESS FOR A PARTICULAR PURPOSE. THE ENTIRE RISK AS TO THE QUALITY AND PERFORMANCE OF THE PROGRAM IS WITH YOU. SHOULD THE PROGRAM PROVE DEFECTIVE, YOU ASSUME THE COST OF ALL NECESSARY SERVICING, REPAIR OR CORRECTION. 12. IN NO EVENT UNLESS REQUIRED BY APPLICABLE LAW OR AGREED TO IN WRITING WILL ANY COPYRIGHT HOLDER, OR ANY OTHER PARTY WHO MAY MODIFY AND/OR REDISTRIBUTE THE PROGRAM AS PERMITTED ABOVE, BE LIABLE TO YOU FOR DAMAGES, INCLUDING ANY GENERAL, SPECIAL, INCIDENTAL OR CONSEQUENTIAL DAMAGES ARISING OUT OF THE USE OR INABILITY TO USE THE PROGRAM (INCLUDING BUT NOT LIMITED TO LOSS OF DATA OR DATA BEING RENDERED INACCURATE OR LOSSES SUSTAINED BY YOU OR THIRD PARTIES OR A FAILURE OF THE PROGRAM TO OPERATE WITH ANY OTHER PROGRAMS), EVEN IF SUCH HOLDER OR OTHER PARTY HAS BEEN ADVISED OF THE POSSIBILITY OF SUCH DAMAGES. END OF TERMS AND CONDITIONS How to Apply These Terms to Your New Programs If you develop a new program, and you want it to be of the greatest possible use to the public, the best way to achieve this is to make it free software which everyone can redistribute and change under these terms. To do so, attach the following notices to the program. It is safest to attach them to the start of each source file to most effectively convey the exclusion of warranty; and each file should have at least the "copyright" line and a pointer to where the full notice is found. Copyright (C) This program is free software; you can redistribute it and/or modify it under the terms of the GNU General Public License as published by the Free Software Foundation; either version 2 of the License, or (at your option) any later version. This program is distributed in the hope that it will be useful, but WITHOUT ANY WARRANTY; without even the implied warranty of MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the GNU General Public License for more details. You should have received a copy of the GNU General Public License along with this program; if not, write to the Free Software Foundation, Inc., 51 Franklin Street, Fifth Floor, Boston, MA 02110-1301 USA Also add information on how to contact you by electronic and paper mail. If the program is interactive, make it output a short notice like this when it starts in an interactive mode: Gnomovision version 69, Copyright (C) year name of author Gnomovision comes with ABSOLUTELY NO WARRANTY; for details type `show w'. This is free software, and you are welcome to redistribute it under certain conditions; type `show c' for details. The hypothetical commands `show w' and `show c' should show the appropriate parts of the General Public License. Of course, the commands you use may be called something other than `show w' and `show c'; they could even be mouse-clicks or menu items--whatever suits your program. You should also get your employer (if you work as a programmer) or your school, if any, to sign a "copyright disclaimer" for the program, if necessary. Here is a sample; alter the names: Yoyodyne, Inc., hereby disclaims all copyright interest in the program `Gnomovision' (which makes passes at compilers) written by James Hacker. , 1 April 1989 Ty Coon, President of Vice This General Public License does not permit incorporating your program into proprietary programs. If your program is a subroutine library, you may consider it more useful to permit linking proprietary applications with the library. If this is what you want to do, use the GNU Library General Public License instead of this License. fftw-3.3.3/ChangeLog0000644000175000017500000300224012054410135011155 00000000000000Sat Nov 24 22:37:54 EST 2012 stevenj@fftw.org * fixed deadlock bug caused by bogosity flag getting out of synch between processes; thanks to Michael Pippig for the bug report M ./kernel/ifftw.h +1 M ./kernel/planner.c -3 +6 M ./mpi/api.c +12 Wed Nov 21 18:34:29 EST 2012 athena@fftw.org * Updated NEWS M ./NEWS -2 +7 Wed Nov 21 18:33:15 EST 2012 athena@fftw.org * use 2x2 AVX transposition instead of individual stores. This seems to improve single-precision AVX on Sandy Bridge machines. M ./simd-support/simd-avx.h -2 +14 Tue Nov 20 12:18:00 EST 2012 stevenj@fftw.org * revert part of Taylor patch to acx_mpi.m4: do not link -lmpi if mpicc works without libraries, as -lmpi may be some completely different MPI implementation M ./m4/acx_mpi.m4 -3 +3 Tue Nov 20 11:44:57 EST 2012 stevenj@fftw.org * fix deadlock bug (thanks to Michael Pippig for the bug report and patch, and to Graham Dennis for the bug report) in which some processes called MPI_Alltoall and some called MPI_Alltoallv M ./mpi/transpose-alltoall.c -3 +2 Mon Oct 29 15:20:01 EDT 2012 athena@fftw.org * fix texinfo quirk M ./doc/tutorial.texi -2 +2 Mon Oct 29 09:16:43 EDT 2012 athena@fftw.org * clarify that padding only applies to in-place transforms M ./doc/tutorial.texi -5 +10 Sun Oct 28 18:42:48 EDT 2012 athena@fftw.org * make the index-computation logic less paranoid The problem is that for each K and for each expression of the form P[I + STRIDE * K] in a loop, most compilers will try to lift an induction variable PK := &P[I + STRIDE * K]. In large codelets we have many such values of K. For example, a codelet of size 32 with 4 input pointers will generate O(128) induction variables, which will likely overflow the register set, which is likely worse than doing the index computation in the first place. In the past we (wisely and correctly) assumed that compilers will do the wrong thing, and consequently we disabled the induction-variable "optimization" altogether by setting STRIDE ^= ZERO, where ZERO is a value guaranteed to be 0. Since the compiler does not know that ZERO=0, it cannot perform its "optimization" and it is forced to behave sensibly. With this patch, FFTW is a little bit less paranoid. FFTW now disables the induction-variable optimization" only when we estimate that the codelet uses more than ESTIMATED_AVAILABLE_INDEX_REGISTERS induction variables. Currently we set ESTIMATED_AVAILABLE_INDEX_REGISTERS=16. 16 registers ought to be enough for anybody (or so the amd64 and ARM ISA's seem to imply). M ./genfft/gen_hc2c.ml -1 +1 M ./genfft/gen_hc2cdft.ml -1 +1 M ./genfft/gen_hc2cdft_c.ml -1 +1 M ./genfft/gen_hc2hc.ml -1 +1 M ./genfft/gen_notw.ml -2 +2 M ./genfft/gen_notw_c.ml -2 +2 M ./genfft/gen_r2cb.ml -3 +3 M ./genfft/gen_r2cf.ml -3 +3 M ./genfft/gen_r2r.ml -2 +2 M ./genfft/gen_twiddle.ml -1 +1 M ./genfft/gen_twiddle_c.ml -1 +1 M ./genfft/gen_twidsq.ml -2 +2 M ./genfft/gen_twidsq_c.ml -2 +2 M ./genfft/genutil.ml -1 +2 M ./kernel/ifftw.h -3 +20 Sun Oct 28 18:33:24 EDT 2012 athena@fftw.org * silence warnings M ./kernel/buffered.c +1 M ./rdft/rank0.c +1 Sat Oct 27 09:58:49 EDT 2012 athena@fftw.org * bump version to 3.3.3 M ./NEWS +7 M ./configure.ac -1 +1 Sat Oct 27 09:55:15 EDT 2012 athena@fftw.org * evaluate plans for >1ms when using gettimeofday() The previous limit 10ms was too paranoid, and it made life difficult on machines without an "official" cycle counter, such as ARM. M ./kernel/timer.c -1 +1 Sat Oct 27 09:46:04 EDT 2012 athena@fftw.org * use 4-way NEON SIMD instead of 2-way Kai-Uwe Bloem tried to warn me a year ago that 128-bit NEON was better than 64-bit NEON even on machines with a 64-bit pipe, but I foolishly did not listen. Now that 128-bit NEON pipes are starting to appear on the market it is definitely time to switch. M ./simd-support/simd-neon.h -55 +100 Wed Sep 26 14:21:12 EDT 2012 athena@fftw.org * Note that fftw-3.3 includes MPI support M ./doc/intro.texi -5 +4 Wed Jul 18 11:25:40 EDT 2012 athena@fftw.org * remove obsolete unused function M ./dft/bluestein.c -14 Fri Jun 29 15:57:14 EDT 2012 stevenj@fftw.org * whoops, call omp_get_max_threads; thanks to Hanno Rein for the bug report M ./doc/threads.texi -1 +1 Sat Apr 28 10:55:09 EDT 2012 athena@fftw.org * Fix libfftw3/libfftw3_threads chicken-egg problem On most systems we want to build libfftw3 first, so that libfftw3_threads can depend upon libfftw3. When producing a single combined-thread library (e.g. on Windows) we want the opposite, so that libfftw3 can include libfftw3_threads. M ./Makefile.am -1 +18 M ./threads/Makefile.am +5 Sat Apr 28 10:11:28 EDT 2012 athena@fftw.org * updated NEWS for 3.3.2 M ./NEWS +15 Thu Apr 26 19:36:11 EDT 2012 athena@fftw.org * change revision to 3.3.2 M ./configure.ac -2 +2 Thu Apr 26 19:31:02 EDT 2012 athena@fftw.org * Remove old aligned_main() hack. On i386, in our benchmark program we used to manually aligned the stack to 16-byte boundary via asm trickery. This was a good idea in 1999 (and it was actually necessary to make things work) but the hack is now obsolete and it seems to break gcc-4.7. So the hack is now gone. M ./libbench2/Makefile.am -6 +6 R ./libbench2/aligned-main.c M ./libbench2/main.c -1 +1 Thu Mar 29 16:26:16 EDT 2012 athena@fftw.org * Bugfix: a couple of uninitialized values in the benchmark program M ./libbench2/problem.c -1 +1 Tue Mar 20 19:03:47 EDT 2012 athena@fftw.org * make libfftw{threads,mpi} depend upon libfftw for libtool purposes Thanks Julian Taylor for the patch M ./Makefile.am -1 +1 M ./m4/acx_mpi.m4 -3 +3 M ./mpi/Makefile.am +1 M ./threads/Makefile.am +1 Tue Mar 6 04:44:00 EST 2012 stevenj@fftw.org * formatting tweak M ./doc/reference.texi -1 +1 Mon Mar 5 21:05:27 EST 2012 athena@fftw.org * destroying => overwriting M ./doc/tutorial.texi -1 +1 Fri Mar 2 10:31:20 EST 2012 stevenj@fftw.org * note that WISDOM_ONLY is a documented flag M ./api/fftw3.h -1 +1 Fri Mar 2 10:27:08 EST 2012 stevenj@fftw.org * check for icc pretending to be gcc before including quad-precision decls; thanks to Michael Anselmi for the bug report M ./api/fftw3.h +1 Fri Mar 2 10:23:19 EST 2012 stevenj@fftw.org * foo_CFLAGS needs to manually include AM_CFLAGS; thanks to Henry Gomersall for the Windows bug report M ./threads/Makefile.am -2 +2 Sat Feb 25 15:21:39 EST 2012 athena@fftw.org * update for latest mingw M ./support/BUILD-MINGW32.sh -3 +3 M ./support/BUILD-MINGW64.sh -2 +7 Sat Feb 25 15:21:17 EST 2012 athena@fftw.org tagged fftw-3.3.1 Mon Feb 20 23:06:13 EST 2012 stevenj@fftw.org * added Fortran NEWS M ./NEWS +6 Mon Feb 20 23:00:13 EST 2012 stevenj@fftw.org * move non-portable extended/quad precision F03 interfaces into separate .f03 files (while keeping double/single in fftw3.f03 for minimal ABI breakage) M ./api/Makefile.am -3 +10 M ./api/f03api.sh -2 +4 M ./api/genf03.pl -5 +5 M ./doc/modern-fortran.texi -2 +39 M ./doc/mpi.texi -2 +2 M ./mpi/Makefile.am -5 +11 M ./mpi/f03api.sh -2 +4 Mon Feb 20 11:21:57 EST 2012 athena@fftw.org * rm mpi/fftw3-mpi.f03 at make clean time; thanks Tyler Luchko for the bug report. M ./mpi/Makefile.am +1 Mon Feb 20 11:18:24 EST 2012 athena@fftw.org * Disable a Visual Studion warning that was obnoxious enough for Sebastian Schuberth to send us a patch. M ./simd-support/simd-sse2.h +14 Mon Feb 20 11:18:06 EST 2012 athena@fftw.org * Change version to 3.3.1 M ./configure.ac -1 +1 Mon Feb 20 11:03:15 EST 2012 athena@fftw.org * Integrated Visual Studio AVX patches by Carsten Steger M ./NEWS +3 M ./simd-support/amd64-cpuid.h -10 +40 M ./simd-support/simd-avx.h +6 M ./simd-support/simd-sse2.h +8 M ./simd-support/x86-cpuid.h +7 Wed Nov 9 10:13:32 EST 2011 stevenj@fftw.org * typo M ./doc/modern-fortran.texi -2 +2 Tue Nov 8 22:45:09 EST 2011 stevenj@fftw.org * add missing F77 set_timelimit function; thanks to Martin Diehl for the bug repory M ./api/f77funcs.h +5 Sun Sep 25 10:54:56 EDT 2011 athena@fftw.org * note requirement of /machine:x64 in windows x64 README M ./support/BUILD-MINGW64.sh +7 Sun Sep 18 09:28:20 EDT 2011 athena@fftw.org * AVX detection for MSVC M ./simd-support/amd64-cpuid.h -2 +10 Tue Sep 13 14:58:29 EDT 2011 athena@fftw.org * compile with C89 M ./dft/bluestein.c -1 +1 M ./dft/generic.c -1 +1 M ./rdft/dht-rader.c -1 +1 Sat Sep 3 16:25:50 EDT 2011 athena@fftw.org * use the same search pruning heuristics for threaded plans as for nonthreaded plans M ./threads/ct.c -1 +3 M ./threads/hc2hc.c -1 +3 Sat Sep 3 16:12:11 EDT 2011 athena@fftw.org * shorten ESTIMATE planning time for certain weird sizes FFTW includes a collection of "solvers" that apply to a subset of "problems". Assume for simplicity that a "problem" is a single 1D complex transform of size N, even though real "problems" are much more general than that. FFTW includes three "prime" solvers called "generic", "bluestein", and "rader", which implement different algorithms for prime sizes. Now, for a "problem" of size 13 (say) FFTW also includes special code that handles that size at high speed. It would be a waste of time to measure the execution time of the prime solvers, since we know that the special code is way faster. However, FFTW is modular and one may or may not include the special code for size 13, in which case we must resort to one of the "prime" solvers. To address this issue, the "prime" solvers (and others) are proclaimed to be SLOW". When planning, FFTW first tries to produce a plan ignoring all the SLOW solvers, and if this fails FFTW tries again allowing SLOW solvers. This heuristic works ok unless the sizes are too large. For example for 1044000=2*2*2*2*2*3*3*5*5*5*29 FFTW explores a huge search tree of all zillion factorizations of 1044000/29, failing every time because 29 is SLOW; then it finally allows SLOW solvers and finds a solution immediately. This patch proclaims solvers to be SLOW only for small values of N. For example, the "generic" solver implements an O(n^2) DFT algorithm; we say that it is SLOW only for N<=16. The side effects of this choice are as follows. If one modifies FFTW to include a fast solver of size 17, then planning for N=17*K will be slower than today, because FFTW till try both the fast solver and the generic solver (which is SLOW today and therefore not tried, but is no longer SLOW after the patch). If one removes a fast solver, of size say 13, then he may still fall into the current exponential-search behavior for "problems" of size 13*HIGHLY_FACTORIZABLE_N. If somebody had compleined about transforms of size 1044000 ten years ago, "don't do that" would have been an acceptable answer. I guess the bar is higher today, so I am going to include this patch in our 3.3.1 release despite their side-effects for people who want to modify FFTW. M ./dft/bluestein.c -5 +14 M ./dft/generic.c -15 +6 M ./dft/rader.c -7 +7 M ./kernel/ifftw.h +8 M ./kernel/primes.c +6 M ./rdft/dht-rader.c -7 +8 M ./rdft/generic.c -14 +4 Sat Aug 27 13:55:24 EDT 2011 athena@fftw.org * Fix typo fftw_execute_dft_r2r => fftw_execute_r2r Thanks KIU Shueng Chuan for the bug report. M ./doc/reference.texi -1 +1 Fri Aug 26 06:13:55 EDT 2011 athena@fftw.org * In Rader's algorithm, compute the generator lazily. The planner was spending a lot of time computing generators for plans that were immediately discarded. Now we compute generators only when absolutely needed. M ./NEWS +5 M ./dft/rader.c -3 +4 M ./rdft/dht-rader.c -3 +4 Sun Aug 21 16:27:31 EDT 2011 athena@fftw.org tagged fftw-3.3.1-beta1 Sun Aug 21 16:16:38 EDT 2011 athena@fftw.org * Release notes for 3.3.1-beta1 M ./NEWS +4 M ./configure.ac -1 +1 M ./doc/install.texi -8 +18 M ./doc/other.texi -2 +3 Fri Aug 19 19:59:17 EDT 2011 stevenj@alum.mit.edu * make fftw_mpi_block routine 10x faster, since it is being called zillions of times (thanks to Tom Vacek for the profiling) M ./mpi/block.c -6 +2 Thu Aug 18 14:19:36 EDT 2011 athena@fftw.org * Implement autodetection of NEON extensions M ./simd-support/neon.c -4 +51 Sun Aug 14 14:12:29 EDT 2011 athena@fftw.org * Update the FSF address. The FSF moved downtown. M ./COPYRIGHT -1 +1 M ./api/api.h -1 +1 M ./api/apiplan.c -1 +1 M ./api/configure.c -1 +1 M ./api/execute-dft-c2r.c -1 +1 M ./api/execute-dft-r2c.c -1 +1 M ./api/execute-dft.c -1 +1 M ./api/execute-r2r.c -1 +1 M ./api/execute-split-dft-c2r.c -1 +1 M ./api/execute-split-dft-r2c.c -1 +1 M ./api/execute-split-dft.c -1 +1 M ./api/execute.c -1 +1 M ./api/export-wisdom-to-file.c -1 +1 M ./api/export-wisdom-to-string.c -1 +1 M ./api/export-wisdom.c -1 +1 M ./api/f77api.c -1 +1 M ./api/f77funcs.h -1 +1 M ./api/flops.c -1 +1 M ./api/forget-wisdom.c -1 +1 M ./api/import-system-wisdom.c -1 +1 M ./api/import-wisdom-from-file.c -1 +1 M ./api/import-wisdom-from-string.c -1 +1 M ./api/import-wisdom.c -1 +1 M ./api/malloc.c -1 +1 M ./api/map-r2r-kind.c -1 +1 M ./api/mapflags.c -1 +1 M ./api/mkprinter-file.c -1 +1 M ./api/mktensor-iodims.h -1 +1 M ./api/mktensor-rowmajor.c -1 +1 M ./api/plan-dft-1d.c -1 +1 M ./api/plan-dft-2d.c -1 +1 M ./api/plan-dft-3d.c -1 +1 M ./api/plan-dft-c2r-1d.c -1 +1 M ./api/plan-dft-c2r-2d.c -1 +1 M ./api/plan-dft-c2r-3d.c -1 +1 M ./api/plan-dft-c2r.c -1 +1 M ./api/plan-dft-r2c-1d.c -1 +1 M ./api/plan-dft-r2c-2d.c -1 +1 M ./api/plan-dft-r2c-3d.c -1 +1 M ./api/plan-dft-r2c.c -1 +1 M ./api/plan-dft.c -1 +1 M ./api/plan-guru-dft-c2r.h -1 +1 M ./api/plan-guru-dft-r2c.h -1 +1 M ./api/plan-guru-dft.h -1 +1 M ./api/plan-guru-r2r.h -1 +1 M ./api/plan-guru-split-dft-c2r.h -1 +1 M ./api/plan-guru-split-dft-r2c.h -1 +1 M ./api/plan-guru-split-dft.h -1 +1 M ./api/plan-many-dft-c2r.c -1 +1 M ./api/plan-many-dft-r2c.c -1 +1 M ./api/plan-many-dft.c -1 +1 M ./api/plan-many-r2r.c -1 +1 M ./api/plan-r2r-1d.c -1 +1 M ./api/plan-r2r-2d.c -1 +1 M ./api/plan-r2r-3d.c -1 +1 M ./api/plan-r2r.c -1 +1 M ./api/print-plan.c -1 +1 M ./api/rdft2-pad.c -1 +1 M ./api/the-planner.c -1 +1 M ./api/version.c -1 +1 M ./api/x77.h -1 +1 M ./commercialize.sh -2 +2 M ./dft/bluestein.c -1 +1 M ./dft/buffered.c -1 +1 M ./dft/codelet-dft.h -1 +1 M ./dft/conf.c -1 +1 M ./dft/ct.c -1 +1 M ./dft/ct.h -1 +1 M ./dft/dft.h -1 +1 M ./dft/dftw-direct.c -1 +1 M ./dft/dftw-directsq.c -1 +1 M ./dft/dftw-generic.c -1 +1 M ./dft/dftw-genericbuf.c -1 +1 M ./dft/direct.c -1 +1 M ./dft/generic.c -1 +1 M ./dft/indirect-transpose.c -1 +1 M ./dft/indirect.c -1 +1 M ./dft/kdft-dif.c -1 +1 M ./dft/kdft-difsq.c -1 +1 M ./dft/kdft-dit.c -1 +1 M ./dft/kdft.c -1 +1 M ./dft/nop.c -1 +1 M ./dft/plan.c -1 +1 M ./dft/problem.c -1 +1 M ./dft/rader.c -1 +1 M ./dft/rank-geq2.c -1 +1 M ./dft/scalar/n.c -1 +1 M ./dft/scalar/n.h -1 +1 M ./dft/scalar/t.c -1 +1 M ./dft/scalar/t.h -1 +1 M ./dft/simd/common/genus.c -1 +1 M ./dft/simd/n1b.h -1 +1 M ./dft/simd/n1f.h -1 +1 M ./dft/simd/n2b.h -1 +1 M ./dft/simd/n2f.h -1 +1 M ./dft/simd/n2s.h -1 +1 M ./dft/simd/q1b.h -1 +1 M ./dft/simd/q1f.h -1 +1 M ./dft/simd/t1b.h -1 +1 M ./dft/simd/t1bu.h -1 +1 M ./dft/simd/t1f.h -1 +1 M ./dft/simd/t1fu.h -1 +1 M ./dft/simd/t2b.h -1 +1 M ./dft/simd/t2f.h -1 +1 M ./dft/simd/t3b.h -1 +1 M ./dft/simd/t3f.h -1 +1 M ./dft/simd/ts.h -1 +1 M ./dft/solve.c -1 +1 M ./dft/vrank-geq1.c -1 +1 M ./dft/zero.c -1 +1 M ./doc/f77_wisdom.f -1 +1 M ./doc/license.texi -3 +3 M ./genfft/algsimp.ml -1 +1 M ./genfft/algsimp.mli -1 +1 M ./genfft/annotate.ml -1 +1 M ./genfft/annotate.mli -1 +1 M ./genfft/assoctable.ml -1 +1 M ./genfft/assoctable.mli -1 +1 M ./genfft/c.ml -1 +1 M ./genfft/c.mli -1 +1 M ./genfft/complex.ml -1 +1 M ./genfft/complex.mli -1 +1 M ./genfft/conv.ml -1 +1 M ./genfft/conv.mli -1 +1 M ./genfft/dag.ml -1 +1 M ./genfft/dag.mli -1 +1 M ./genfft/expr.ml -1 +1 M ./genfft/expr.mli -1 +1 M ./genfft/fft.ml -1 +1 M ./genfft/fft.mli -1 +1 M ./genfft/gen_hc2c.ml -1 +1 M ./genfft/gen_hc2cdft.ml -1 +1 M ./genfft/gen_hc2cdft_c.ml -1 +1 M ./genfft/gen_hc2hc.ml -1 +1 M ./genfft/gen_mdct.ml -1 +1 M ./genfft/gen_notw.ml -1 +1 M ./genfft/gen_notw_c.ml -1 +1 M ./genfft/gen_r2cb.ml -1 +1 M ./genfft/gen_r2cf.ml -1 +1 M ./genfft/gen_r2r.ml -1 +1 M ./genfft/gen_twiddle.ml -1 +1 M ./genfft/gen_twiddle_c.ml -1 +1 M ./genfft/gen_twidsq.ml -1 +1 M ./genfft/gen_twidsq_c.ml -1 +1 M ./genfft/genutil.ml -1 +1 M ./genfft/littlesimp.ml -1 +1 M ./genfft/littlesimp.mli -1 +1 M ./genfft/magic.ml -1 +1 M ./genfft/monads.ml -1 +1 M ./genfft/number.ml -1 +1 M ./genfft/number.mli -1 +1 M ./genfft/oracle.ml -1 +1 M ./genfft/oracle.mli -1 +1 M ./genfft/schedule.ml -1 +1 M ./genfft/schedule.mli -1 +1 M ./genfft/simd.ml -1 +1 M ./genfft/simd.mli -1 +1 M ./genfft/simdmagic.ml -1 +1 M ./genfft/to_alist.ml -1 +1 M ./genfft/to_alist.mli -1 +1 M ./genfft/trig.ml -1 +1 M ./genfft/trig.mli -1 +1 M ./genfft/twiddle.ml -1 +1 M ./genfft/twiddle.mli -1 +1 M ./genfft/unique.ml -1 +1 M ./genfft/unique.mli -1 +1 M ./genfft/util.ml -1 +1 M ./genfft/util.mli -1 +1 M ./genfft/variable.ml -1 +1 M ./genfft/variable.mli -1 +1 M ./kernel/align.c -1 +1 M ./kernel/alloc.c -1 +1 M ./kernel/assert.c -1 +1 M ./kernel/awake.c -1 +1 M ./kernel/buffered.c -1 +1 M ./kernel/cpy1d.c -1 +1 M ./kernel/cpy2d-pair.c -1 +1 M ./kernel/cpy2d.c -1 +1 M ./kernel/ct.c -1 +1 M ./kernel/debug.c -1 +1 M ./kernel/extract-reim.c -1 +1 M ./kernel/hash.c -1 +1 M ./kernel/iabs.c -1 +1 M ./kernel/ifftw.h -1 +1 M ./kernel/kalloc.c -1 +1 M ./kernel/md5-1.c -1 +1 M ./kernel/md5.c -1 +1 M ./kernel/minmax.c -1 +1 M ./kernel/ops.c -1 +1 M ./kernel/pickdim.c -1 +1 M ./kernel/plan.c -1 +1 M ./kernel/planner.c -1 +1 M ./kernel/primes.c -1 +1 M ./kernel/print.c -1 +1 M ./kernel/problem.c -1 +1 M ./kernel/rader.c -1 +1 M ./kernel/scan.c -1 +1 M ./kernel/solver.c -1 +1 M ./kernel/solvtab.c -1 +1 M ./kernel/stride.c -1 +1 M ./kernel/tensor.c -1 +1 M ./kernel/tensor1.c -1 +1 M ./kernel/tensor2.c -1 +1 M ./kernel/tensor3.c -1 +1 M ./kernel/tensor4.c -1 +1 M ./kernel/tensor5.c -1 +1 M ./kernel/tensor7.c -1 +1 M ./kernel/tensor8.c -1 +1 M ./kernel/tensor9.c -1 +1 M ./kernel/tile2d.c -1 +1 M ./kernel/timer.c -1 +1 M ./kernel/transpose.c -1 +1 M ./kernel/trig.c -1 +1 M ./kernel/twiddle.c -1 +1 M ./libbench2/aligned-main.c -1 +1 M ./libbench2/bench-main.c -1 +1 M ./libbench2/bench-user.h -1 +1 M ./libbench2/bench.h -1 +1 M ./libbench2/can-do.c -1 +1 M ./libbench2/dotens2.c -1 +1 M ./libbench2/info.c -1 +1 M ./libbench2/main.c -1 +1 M ./libbench2/my-getopt.c -1 +1 M ./libbench2/my-getopt.h -1 +1 M ./libbench2/problem.c -1 +1 M ./libbench2/report.c -1 +1 M ./libbench2/speed.c -1 +1 M ./libbench2/tensor.c -1 +1 M ./libbench2/timer.c -1 +1 M ./libbench2/useropt.c -1 +1 M ./libbench2/util.c -1 +1 M ./libbench2/verify-dft.c -1 +1 M ./libbench2/verify-lib.c -1 +1 M ./libbench2/verify-r2r.c -1 +1 M ./libbench2/verify-rdft2.c -1 +1 M ./libbench2/verify.c -1 +1 M ./libbench2/verify.h -1 +1 M ./libbench2/zero.c -1 +1 M ./mpi/any-true.c -1 +1 M ./mpi/api.c -1 +1 M ./mpi/block.c -1 +1 M ./mpi/choose-radix.c -1 +1 M ./mpi/conf.c -1 +1 M ./mpi/dft-problem.c -1 +1 M ./mpi/dft-rank-geq2-transposed.c -1 +1 M ./mpi/dft-rank-geq2.c -1 +1 M ./mpi/dft-rank1-bigvec.c -1 +1 M ./mpi/dft-rank1.c -1 +1 M ./mpi/dft-serial.c -1 +1 M ./mpi/dft-solve.c -1 +1 M ./mpi/dtensor.c -1 +1 M ./mpi/ifftw-mpi.h -1 +1 M ./mpi/mpi-dft.h -1 +1 M ./mpi/mpi-rdft.h -1 +1 M ./mpi/mpi-rdft2.h -1 +1 M ./mpi/mpi-transpose.h -1 +1 M ./mpi/rdft-problem.c -1 +1 M ./mpi/rdft-rank-geq2-transposed.c -1 +1 M ./mpi/rdft-rank-geq2.c -1 +1 M ./mpi/rdft-rank1-bigvec.c -1 +1 M ./mpi/rdft-serial.c -1 +1 M ./mpi/rdft-solve.c -1 +1 M ./mpi/rdft2-problem.c -1 +1 M ./mpi/rdft2-rank-geq2-transposed.c -1 +1 M ./mpi/rdft2-rank-geq2.c -1 +1 M ./mpi/rdft2-serial.c -1 +1 M ./mpi/rdft2-solve.c -1 +1 M ./mpi/rearrange.c -1 +1 M ./mpi/testsched.c -1 +1 M ./mpi/transpose-alltoall.c -1 +1 M ./mpi/transpose-pairwise.c -1 +1 M ./mpi/transpose-problem.c -1 +1 M ./mpi/transpose-recurse.c -1 +1 M ./mpi/transpose-solve.c -1 +1 M ./mpi/wisdom-api.c -1 +1 M ./rdft/buffered.c -1 +1 M ./rdft/buffered2.c -1 +1 M ./rdft/codelet-rdft.h -1 +1 M ./rdft/conf.c -1 +1 M ./rdft/ct-hc2c-direct.c -1 +1 M ./rdft/ct-hc2c.c -1 +1 M ./rdft/ct-hc2c.h -1 +1 M ./rdft/dft-r2hc.c -1 +1 M ./rdft/dht-r2hc.c -1 +1 M ./rdft/dht-rader.c -1 +1 M ./rdft/direct-r2c.c -1 +1 M ./rdft/direct-r2r.c -1 +1 M ./rdft/direct2.c -1 +1 M ./rdft/generic.c -1 +1 M ./rdft/hc2hc-direct.c -1 +1 M ./rdft/hc2hc-generic.c -1 +1 M ./rdft/hc2hc.c -1 +1 M ./rdft/hc2hc.h -1 +1 M ./rdft/indirect.c -1 +1 M ./rdft/khc2c.c -1 +1 M ./rdft/khc2hc.c -1 +1 M ./rdft/kr2c.c -1 +1 M ./rdft/kr2r.c -1 +1 M ./rdft/nop.c -1 +1 M ./rdft/nop2.c -1 +1 M ./rdft/plan.c -1 +1 M ./rdft/plan2.c -1 +1 M ./rdft/problem.c -1 +1 M ./rdft/problem2.c -1 +1 M ./rdft/rank-geq2-rdft2.c -1 +1 M ./rdft/rank-geq2.c -1 +1 M ./rdft/rank0-rdft2.c -1 +1 M ./rdft/rank0.c -1 +1 M ./rdft/rdft-dht.c -1 +1 M ./rdft/rdft.h -1 +1 M ./rdft/rdft2-inplace-strides.c -1 +1 M ./rdft/rdft2-rdft.c -1 +1 M ./rdft/rdft2-strides.c -1 +1 M ./rdft/rdft2-tensor-max-index.c -1 +1 M ./rdft/scalar/hb.h -1 +1 M ./rdft/scalar/hc2c.c -1 +1 M ./rdft/scalar/hc2cb.h -1 +1 M ./rdft/scalar/hc2cf.h -1 +1 M ./rdft/scalar/hf.h -1 +1 M ./rdft/scalar/hfb.c -1 +1 M ./rdft/scalar/r2c.c -1 +1 M ./rdft/scalar/r2cb.h -1 +1 M ./rdft/scalar/r2cbIII.h -1 +1 M ./rdft/scalar/r2cf.h -1 +1 M ./rdft/scalar/r2cfII.h -1 +1 M ./rdft/scalar/r2r.c -1 +1 M ./rdft/scalar/r2r.h -1 +1 M ./rdft/simd/common/genus.c -1 +1 M ./rdft/simd/hc2cbv.h -1 +1 M ./rdft/simd/hc2cfv.h -1 +1 M ./rdft/solve.c -1 +1 M ./rdft/solve2.c -1 +1 M ./rdft/vrank-geq1-rdft2.c -1 +1 M ./rdft/vrank-geq1.c -1 +1 M ./rdft/vrank3-transpose.c -1 +1 M ./reodft/conf.c -1 +1 M ./reodft/redft00e-r2hc-pad.c -1 +1 M ./reodft/redft00e-r2hc.c -1 +1 M ./reodft/reodft.h -1 +1 M ./reodft/reodft00e-splitradix.c -1 +1 M ./reodft/reodft010e-r2hc.c -1 +1 M ./reodft/reodft11e-r2hc-odd.c -1 +1 M ./reodft/reodft11e-r2hc.c -1 +1 M ./reodft/reodft11e-radix2.c -1 +1 M ./reodft/rodft00e-r2hc-pad.c -1 +1 M ./reodft/rodft00e-r2hc.c -1 +1 M ./simd-support/altivec.c -1 +1 M ./simd-support/amd64-cpuid.h -1 +1 M ./simd-support/avx.c -1 +1 M ./simd-support/neon.c -1 +1 M ./simd-support/simd-altivec.h -1 +1 M ./simd-support/simd-avx.h -1 +1 M ./simd-support/simd-common.h -1 +1 M ./simd-support/simd-neon.h -1 +1 M ./simd-support/simd-sse2.h -1 +1 M ./simd-support/sse2-nonportable.c -1 +1 M ./simd-support/sse2.c -1 +1 M ./simd-support/taint.c -1 +1 M ./simd-support/x86-cpuid.h -1 +1 M ./simd/altivec.c -1 +1 M ./simd/mips_ps.c -1 +1 M ./simd/mips_ps.h -1 +1 M ./simd/nonportable/sse.c -1 +1 M ./simd/nonportable/sse2.c -1 +1 M ./simd/simd-altivec.h -1 +1 M ./simd/simd-mips_ps.h -1 +1 M ./simd/simd-sse.h -1 +1 M ./simd/simd-sse2.h -1 +1 M ./simd/simd.h -1 +1 M ./simd/sse.c -1 +1 M ./simd/sse2.c -1 +1 M ./simd/taint.c -1 +1 M ./simd/x86-cpuid.h -1 +1 M ./threads/api.c -1 +1 M ./threads/conf.c -1 +1 M ./threads/ct.c -1 +1 M ./threads/dft-vrank-geq1.c -1 +1 M ./threads/f77api.c -1 +1 M ./threads/f77funcs.h -1 +1 M ./threads/hc2hc.c -1 +1 M ./threads/openmp.c -1 +1 M ./threads/rdft-vrank-geq1.c -1 +1 M ./threads/threads.c -1 +1 M ./threads/threads.h -1 +1 M ./threads/vrank-geq1-rdft2.c -1 +1 M ./tools/fftw-wisdom-to-conf.1 -1 +1 M ./tools/fftw-wisdom-to-conf.in -1 +1 M ./tools/fftw-wisdom.c -1 +1 M ./tools/fftw_wisdom.1.in -1 +1 Thu Aug 11 14:54:38 EDT 2011 stevenj@fftw.org * allow specifying TRANSPOSE_{IN/OUT} transpose plans, since libbench does not canonicalize rnk=1 n=1 plans as rnk=0 M ./mpi/mpi-bench.c -2 +2 Thu Aug 11 14:17:24 EDT 2011 stevenj@fftw.org * check.pl should occasionally check DESTROY_INPUT problems too (especially since those enable slightly different algorithms in MPI) M ./tests/check.pl +1 Thu Aug 11 12:37:51 EDT 2011 stevenj@fftw.org * unify post-MPI transpose handling in pairwise and alltoall solvers; should make the former faster in the destroy-input out-of-place case, and the latter more widely applicable M ./mpi/mpi-transpose.h +5 M ./mpi/transpose-alltoall.c -69 +24 M ./mpi/transpose-pairwise.c -66 +97 Mon Aug 8 10:06:14 EDT 2011 athena@fftw.org * Add support for ARM NEON M ./Makefile.am -1 +6 M ./api/version.c +4 M ./configure.ac -5 +23 M ./dft/codelet-dft.h +1 M ./dft/conf.c +4 M ./dft/simd/Makefile.am -1 +1 A ./dft/simd/neon/ A ./dft/simd/neon/Makefile.am M ./kernel/ifftw.h +2 M ./m4/ax_cc_maxopt.m4 +10 M ./rdft/codelet-rdft.h +1 M ./rdft/conf.c +4 M ./rdft/simd/Makefile.am -1 +1 A ./rdft/simd/neon/ A ./rdft/simd/neon/Makefile.am M ./simd-support/Makefile.am -1 +2 A ./simd-support/neon.c M ./simd-support/simd-common.h +3 A ./simd-support/simd-neon.h Fri Aug 5 17:25:32 EDT 2011 stevenj@fftw.org * more C++ paranoia M ./tests/fftw-bench.h +10 Fri Aug 5 17:02:00 EDT 2011 stevenj@fftw.org * tentative version bump for 3.3.1 M ./NEWS +5 M ./configure.ac -2 +2 Fri Aug 5 16:52:28 EDT 2011 stevenj@fftw.org * fixes so that MPI code compiles when MPICC is a C++ compiler, even if the serial code is compiled with a C compiler; thanks to Kyle Spyksma for the bug report M ./api/api.h +12 M ./api/plan-guru-r2r.h -2 M ./api/plan-many-r2r.c -2 M ./dft/dft.h +9 M ./kernel/ifftw.h +9 M ./mpi/api.c -3 +1 M ./mpi/dtensor.c -1 +1 M ./mpi/ifftw-mpi.h -1 +4 M ./mpi/mpi-bench.c -1 +1 M ./rdft/rdft.h +9 Fri Aug 5 16:04:06 EDT 2011 stevenj@fftw.org * use correct precision in f03-wrap.c, avoiding a (harmless) implicit pointer cast that prevented compilation under C++; thanks to Kyle Spyksma for the bug report M ./mpi/genf03-wrap.pl +2 Fri Aug 5 14:04:32 EDT 2011 stevenj@fftw.org * manual typo M ./doc/modern-fortran.texi -1 +1 Wed Jul 27 12:50:25 EDT 2011 athena@fftw.org tagged fftw-3.3 Tue Jul 26 20:55:45 EDT 2011 athena@fftw.org * Honor WITH_OUR_MALLOC in libbench2 M ./libbench2/util.c -6 +5 Tue Jul 26 20:27:28 EDT 2011 athena@fftw.org * fixed typo: incorrect name of combined threads library on Windows M ./Makefile.am -1 +1 Mon Jul 25 14:38:20 EDT 2011 stevenj@fftw.org * 3.3 version bump & NEWS M ./NEWS -5 +16 M ./configure.ac -1 +1 Mon Jul 25 14:37:48 EDT 2011 stevenj@fftw.org * use int(..., C_SIZE_T) rather than declaring another variable in the Fortran examples M ./doc/modern-fortran.texi -12 +9 Wed Jul 13 05:02:32 EDT 2011 stevenj@fftw.org * typo, thanks to Rhys Ulerich for the comment M ./doc/mpi.texi -1 +1 Mon Jul 11 14:39:52 EDT 2011 athena@fftw.org * Fix bug in bubblesort Bubblesort was not sorting. This was a bug in the benchmark library (not in FFTW per se), and it impacted the benchmark program with --report-time and --report-mflops causing it to output an incorrect value for the median. (The minimum, maximum, and average value were correct.) Thanks Dima Baksheev of Intel for reporting this bug. M ./libbench2/report.c -2 +2 Fri Jul 8 13:35:59 EDT 2011 stevenj@fftw.org * small manual typos M ./doc/modern-fortran.texi -2 +2 Wed Jul 6 10:49:40 EDT 2011 athena@fftw.org * Detection of altivec.h requires $ALTIVEC_CFLAGS M ./configure.ac -5 +4 Tue Jul 5 19:58:47 EDT 2011 athena@fftw.org * Introduce fake dependency so that my-getopt.c is recompiled my-getopt.c does not depend on anything, and so it is not rebuilt when reconfiguring for a different ISA (e.g., CC="gcc -m32" vs CC="gcc -m64"). Add a fake dependency on so that the file is recompiled. M ./libbench2/my-getopt.c +1 Tue Jul 5 18:53:36 EDT 2011 stevenj@fftw.org * support compiling/installing --enable-threads --enable-openmp at the same time, although in this case the test program only uses the threads variety. Update documentation accordingly, and in general expand the documentation of the OpenMP support M ./Makefile.am -1 +1 M ./configure.ac -21 +13 M ./doc/install.texi -14 +13 M ./doc/threads.texi -19 +43 M ./mpi/Makefile.am -2 +6 M ./tests/Makefile.am -2 +6 M ./tests/fftw-bench.c -1 +3 M ./threads/Makefile.am -6 +20 M ./threads/conf.c -4 M ./threads/openmp.c -4 M ./threads/threads.c -3 M ./tools/Makefile.am -4 +8 Tue Jul 5 16:04:03 EDT 2011 stevenj@fftw.org * call omp_set_num_threads in fftw-bench so that the number of OpenMP threads corresponds with the number of FFTW threads M ./tests/fftw-bench.c +5 Tue Jul 5 16:03:06 EDT 2011 stevenj@fftw.org * when --enable-openmp, install as fftw3_omp rather than fftw3_threads, so that both the POSIX threads and OpenMP variants of FFTW can be installed at once M ./Makefile.am -1 +1 M ./configure.ac +4 M ./threads/Makefile.am -4 +4 Sat Jul 2 02:21:22 EDT 2011 stevenj@fftw.org * don't even declare an fftw_execute interface in Fortran, since it is unsafe and we recommend against it anyway; thanks to Arjen Markus for the suggestion M ./api/f03api.sh -1 +1 M ./doc/modern-fortran.texi -2 +4 Fri Jul 1 14:35:44 EDT 2011 athena@fftw.org * consistently use the order single, double, long double M ./support/BUILD-MINGW32.sh -1 +1 M ./support/BUILD-MINGW64.sh -1 +1 Wed Jun 29 17:27:06 EDT 2011 athena@fftw.org * MSVC AVX 64-bit detection does not work, punt for now. M ./simd-support/amd64-cpuid.h -16 +2 Wed Jun 29 15:52:27 EDT 2011 stevenj@fftw.org * fixed typo, added note on transposed flags for r2c/c2r; thanks to Rhys Ulerich for the suggestions M ./doc/mpi.texi -1 +14 Wed Jun 29 09:41:39 EDT 2011 athena@fftw.org * fixes for compiling with MSVC (untested) M ./simd-support/amd64-cpuid.h -2 +16 M ./simd-support/simd-avx.h -11 +16 M ./simd-support/x86-cpuid.h -1 +7 Tue Jun 28 16:48:36 EDT 2011 athena@fftw.org * comment M ./simd-support/sse2-nonportable.c +3 Mon Jun 27 21:01:56 EDT 2011 stevenj@fftw.org * rm extraneous line break in HTML output ... I hate texinfo M ./doc/reference.texi -6 +3 Mon Jun 27 06:43:57 EDT 2011 athena@fftw.org tagged fftw-3.3-beta1 Mon Jun 27 00:47:33 EDT 2011 stevenj@fftw.org * maintainer-clean should delete html directory (otherwwise we keep obsolete HTML files in the dist tarball, sigh) M ./doc/Makefile.am -2 +2 Mon Jun 27 00:45:28 EDT 2011 stevenj@fftw.org tagged fftw-3.3-beta1 Sun Jun 26 23:36:32 EDT 2011 stevenj@fftw.org * update copyright year in manual M ./doc/license.texi -2 +2 Sun Jun 26 22:52:54 EDT 2011 stevenj@fftw.org * whoops, don't dist .f03 headers, since those are built by the user's Makefile M ./api/Makefile.am -1 +2 M ./mpi/Makefile.am -1 +2 Sun Jun 26 22:43:49 EDT 2011 stevenj@fftw.org * fix embarrassing deadlock/crashing bug in my previous nowisdom_hook fix -- I forgot to handle the case where one process has wisdom and another one doesn't, requiring a nowisdom_hook in the latter case; this should only affect MPI transforms since otherwise these hook functions are NULL M ./kernel/ifftw.h +1 M ./kernel/planner.c -44 +48 M ./mpi/api.c +14 Sun Jun 26 21:02:15 EDT 2011 stevenj@fftw.org * subsubheadings, MPI transpose reference M ./doc/mpi.texi -7 +53 Sun Jun 26 20:48:53 EDT 2011 stevenj@fftw.org * add MPI plan reference M ./doc/mpi.texi -5 +197 Sun Jun 26 17:07:21 EDT 2011 stevenj@fftw.org * portions of MPI reference docs; tweaks to NEWS M ./NEWS -2 +3 M ./doc/mpi.texi -12 +264 Sun Jun 26 12:40:43 EDT 2011 stevenj@fftw.org * use $(CHECK_PL_OPTS) more consistently M ./tests/Makefile.am -1 +1 Sun Jun 26 10:04:54 EDT 2011 athena@fftw.org * accept \r\n as well as \n. Grrr... M ./tests/check.pl -1 +2 Sun Jun 26 09:52:11 EDT 2011 athena@fftw.org * new configure option --with-incoming-stack-boundary=N This option selects CFLAGS to align the stack at all externally-callable functions. This currently comprises api/* and threads/* M ./api/Makefile.am +1 M ./configure.ac +14 M ./support/BUILD-MINGW32.sh -1 +1 M ./threads/Makefile.am +1 Sun Jun 26 09:51:37 EDT 2011 athena@fftw.org * add -fomit-frame-pointer back Somehow -O3 does not imply -fomit-frame-pointer on ia32 M ./m4/ax_cc_maxopt.m4 -2 +3 Sun Jun 26 07:20:27 EDT 2011 athena@fftw.org * Note that removal of mips-ps is temporary. M ./NEWS -3 +5 Sat Jun 25 23:15:03 EDT 2011 stevenj@fftw.org * update copyright year M ./COPYRIGHT -2 +2 M ./api/api.h -2 +2 M ./api/apiplan.c -2 +2 M ./api/configure.c -2 +2 M ./api/execute-dft-c2r.c -2 +2 M ./api/execute-dft-r2c.c -2 +2 M ./api/execute-dft.c -2 +2 M ./api/execute-r2r.c -2 +2 M ./api/execute-split-dft-c2r.c -2 +2 M ./api/execute-split-dft-r2c.c -2 +2 M ./api/execute-split-dft.c -2 +2 M ./api/execute.c -2 +2 M ./api/export-wisdom-to-file.c -2 +2 M ./api/export-wisdom-to-string.c -2 +2 M ./api/export-wisdom.c -2 +2 M ./api/f77api.c -2 +2 M ./api/f77funcs.h -2 +2 M ./api/fftw3.h -2 +2 M ./api/flops.c -2 +2 M ./api/forget-wisdom.c -2 +2 M ./api/import-system-wisdom.c -2 +2 M ./api/import-wisdom-from-file.c -2 +2 M ./api/import-wisdom-from-string.c -2 +2 M ./api/import-wisdom.c -2 +2 M ./api/malloc.c -2 +2 M ./api/map-r2r-kind.c -2 +2 M ./api/mapflags.c -2 +2 M ./api/mkprinter-file.c -2 +2 M ./api/mktensor-iodims.h -2 +2 M ./api/mktensor-rowmajor.c -2 +2 M ./api/plan-dft-1d.c -2 +2 M ./api/plan-dft-2d.c -2 +2 M ./api/plan-dft-3d.c -2 +2 M ./api/plan-dft-c2r-1d.c -2 +2 M ./api/plan-dft-c2r-2d.c -2 +2 M ./api/plan-dft-c2r-3d.c -2 +2 M ./api/plan-dft-c2r.c -2 +2 M ./api/plan-dft-r2c-1d.c -2 +2 M ./api/plan-dft-r2c-2d.c -2 +2 M ./api/plan-dft-r2c-3d.c -2 +2 M ./api/plan-dft-r2c.c -2 +2 M ./api/plan-dft.c -2 +2 M ./api/plan-guru-dft-c2r.h -2 +2 M ./api/plan-guru-dft-r2c.h -2 +2 M ./api/plan-guru-dft.h -2 +2 M ./api/plan-guru-r2r.h -2 +2 M ./api/plan-guru-split-dft-c2r.h -2 +2 M ./api/plan-guru-split-dft-r2c.h -2 +2 M ./api/plan-guru-split-dft.h -2 +2 M ./api/plan-many-dft-c2r.c -2 +2 M ./api/plan-many-dft-r2c.c -2 +2 M ./api/plan-many-dft.c -2 +2 M ./api/plan-many-r2r.c -2 +2 M ./api/plan-r2r-1d.c -2 +2 M ./api/plan-r2r-2d.c -2 +2 M ./api/plan-r2r-3d.c -2 +2 M ./api/plan-r2r.c -2 +2 M ./api/print-plan.c -2 +2 M ./api/rdft2-pad.c -2 +2 M ./api/the-planner.c -2 +2 M ./api/version.c -2 +2 M ./api/x77.h -2 +2 M ./dft/bluestein.c -2 +2 M ./dft/buffered.c -2 +2 M ./dft/codelet-dft.h -2 +2 M ./dft/conf.c -2 +2 M ./dft/ct.c -2 +2 M ./dft/ct.h -2 +2 M ./dft/dft.h -2 +2 M ./dft/dftw-direct.c -2 +2 M ./dft/dftw-directsq.c -2 +2 M ./dft/dftw-generic.c -2 +2 M ./dft/dftw-genericbuf.c -2 +2 M ./dft/direct.c -2 +2 M ./dft/generic.c -2 +2 M ./dft/indirect-transpose.c -2 +2 M ./dft/indirect.c -2 +2 M ./dft/kdft-dif.c -2 +2 M ./dft/kdft-difsq.c -2 +2 M ./dft/kdft-dit.c -2 +2 M ./dft/kdft.c -2 +2 M ./dft/nop.c -2 +2 M ./dft/plan.c -2 +2 M ./dft/problem.c -2 +2 M ./dft/rader.c -2 +2 M ./dft/rank-geq2.c -2 +2 M ./dft/scalar/n.c -2 +2 M ./dft/scalar/n.h -2 +2 M ./dft/scalar/t.c -2 +2 M ./dft/scalar/t.h -2 +2 M ./dft/simd/common/genus.c -2 +2 M ./dft/simd/n1b.h -2 +2 M ./dft/simd/n1f.h -2 +2 M ./dft/simd/n2b.h -2 +2 M ./dft/simd/n2f.h -2 +2 M ./dft/simd/n2s.h -2 +2 M ./dft/simd/q1b.h -2 +2 M ./dft/simd/q1f.h -2 +2 M ./dft/simd/t1b.h -2 +2 M ./dft/simd/t1bu.h -2 +2 M ./dft/simd/t1f.h -2 +2 M ./dft/simd/t1fu.h -2 +2 M ./dft/simd/t2b.h -2 +2 M ./dft/simd/t2f.h -2 +2 M ./dft/simd/t3b.h -2 +2 M ./dft/simd/t3f.h -2 +2 M ./dft/simd/ts.h -2 +2 M ./dft/solve.c -2 +2 M ./dft/vrank-geq1.c -2 +2 M ./dft/zero.c -2 +2 M ./doc/f77_wisdom.f -2 +2 M ./genfft/algsimp.ml -2 +2 M ./genfft/algsimp.mli -2 +2 M ./genfft/annotate.ml -2 +2 M ./genfft/annotate.mli -2 +2 M ./genfft/assoctable.ml -2 +2 M ./genfft/assoctable.mli -2 +2 M ./genfft/c.ml -2 +2 M ./genfft/c.mli -2 +2 M ./genfft/complex.ml -2 +2 M ./genfft/complex.mli -2 +2 M ./genfft/conv.ml -2 +2 M ./genfft/conv.mli -2 +2 M ./genfft/dag.ml -2 +2 M ./genfft/dag.mli -2 +2 M ./genfft/expr.ml -2 +2 M ./genfft/expr.mli -2 +2 M ./genfft/fft.ml -2 +2 M ./genfft/fft.mli -2 +2 M ./genfft/gen_hc2c.ml -2 +2 M ./genfft/gen_hc2cdft.ml -2 +2 M ./genfft/gen_hc2cdft_c.ml -2 +2 M ./genfft/gen_hc2hc.ml -2 +2 M ./genfft/gen_mdct.ml -2 +2 M ./genfft/gen_notw.ml -2 +2 M ./genfft/gen_notw_c.ml -2 +2 M ./genfft/gen_r2cb.ml -2 +2 M ./genfft/gen_r2cf.ml -2 +2 M ./genfft/gen_r2r.ml -2 +2 M ./genfft/gen_twiddle.ml -2 +2 M ./genfft/gen_twiddle_c.ml -2 +2 M ./genfft/gen_twidsq.ml -2 +2 M ./genfft/gen_twidsq_c.ml -2 +2 M ./genfft/genutil.ml -2 +2 M ./genfft/littlesimp.ml -2 +2 M ./genfft/littlesimp.mli -2 +2 M ./genfft/magic.ml -2 +2 M ./genfft/monads.ml -2 +2 M ./genfft/number.ml -2 +2 M ./genfft/number.mli -2 +2 M ./genfft/oracle.ml -2 +2 M ./genfft/oracle.mli -2 +2 M ./genfft/schedule.ml -2 +2 M ./genfft/schedule.mli -2 +2 M ./genfft/simd.ml -2 +2 M ./genfft/simd.mli -2 +2 M ./genfft/simdmagic.ml -2 +2 M ./genfft/to_alist.ml -2 +2 M ./genfft/to_alist.mli -2 +2 M ./genfft/trig.ml -2 +2 M ./genfft/trig.mli -2 +2 M ./genfft/twiddle.ml -2 +2 M ./genfft/twiddle.mli -2 +2 M ./genfft/unique.ml -2 +2 M ./genfft/unique.mli -2 +2 M ./genfft/util.ml -2 +2 M ./genfft/util.mli -2 +2 M ./genfft/variable.ml -2 +2 M ./genfft/variable.mli -2 +2 M ./kernel/align.c -2 +2 M ./kernel/alloc.c -2 +2 M ./kernel/assert.c -2 +2 M ./kernel/awake.c -2 +2 M ./kernel/buffered.c -2 +2 M ./kernel/cpy1d.c -2 +2 M ./kernel/cpy2d-pair.c -2 +2 M ./kernel/cpy2d.c -2 +2 M ./kernel/ct.c -2 +2 M ./kernel/cycle.h -2 +2 M ./kernel/debug.c -2 +2 M ./kernel/extract-reim.c -2 +2 M ./kernel/hash.c -2 +2 M ./kernel/iabs.c -2 +2 M ./kernel/ifftw.h -2 +2 M ./kernel/kalloc.c -2 +2 M ./kernel/md5-1.c -2 +2 M ./kernel/md5.c -2 +2 M ./kernel/minmax.c -2 +2 M ./kernel/ops.c -2 +2 M ./kernel/pickdim.c -2 +2 M ./kernel/plan.c -2 +2 M ./kernel/primes.c -2 +2 M ./kernel/print.c -2 +2 M ./kernel/problem.c -2 +2 M ./kernel/rader.c -2 +2 M ./kernel/scan.c -2 +2 M ./kernel/solver.c -2 +2 M ./kernel/solvtab.c -2 +2 M ./kernel/stride.c -2 +2 M ./kernel/tensor.c -2 +2 M ./kernel/tensor1.c -2 +2 M ./kernel/tensor2.c -2 +2 M ./kernel/tensor3.c -2 +2 M ./kernel/tensor4.c -2 +2 M ./kernel/tensor5.c -2 +2 M ./kernel/tensor7.c -2 +2 M ./kernel/tensor8.c -2 +2 M ./kernel/tensor9.c -2 +2 M ./kernel/tile2d.c -2 +2 M ./kernel/timer.c -2 +2 M ./kernel/transpose.c -2 +2 M ./kernel/trig.c -2 +2 M ./kernel/twiddle.c -2 +2 M ./libbench2/dotens2.c -2 +2 M ./libbench2/my-getopt.c -2 +2 M ./libbench2/my-getopt.h -2 +2 M ./libbench2/verify-dft.c -2 +2 M ./libbench2/verify-lib.c -2 +2 M ./libbench2/verify-r2r.c -2 +2 M ./libbench2/verify-rdft2.c -2 +2 M ./libbench2/verify.h -2 +2 M ./mpi/any-true.c -2 +2 M ./mpi/api.c -2 +2 M ./mpi/block.c -2 +2 M ./mpi/choose-radix.c -2 +2 M ./mpi/conf.c -2 +2 M ./mpi/dft-problem.c -2 +2 M ./mpi/dft-rank-geq2-transposed.c -2 +2 M ./mpi/dft-rank-geq2.c -2 +2 M ./mpi/dft-rank1-bigvec.c -2 +2 M ./mpi/dft-rank1.c -2 +2 M ./mpi/dft-serial.c -2 +2 M ./mpi/dft-solve.c -2 +2 M ./mpi/dtensor.c -2 +2 M ./mpi/fftw3-mpi.h -2 +2 M ./mpi/ifftw-mpi.h -2 +2 M ./mpi/mpi-dft.h -2 +2 M ./mpi/mpi-rdft.h -2 +2 M ./mpi/mpi-rdft2.h -2 +2 M ./mpi/mpi-transpose.h -2 +2 M ./mpi/rdft-problem.c -2 +2 M ./mpi/rdft-rank-geq2-transposed.c -2 +2 M ./mpi/rdft-rank-geq2.c -2 +2 M ./mpi/rdft-rank1-bigvec.c -2 +2 M ./mpi/rdft-serial.c -2 +2 M ./mpi/rdft-solve.c -2 +2 M ./mpi/rdft2-problem.c -2 +2 M ./mpi/rdft2-rank-geq2-transposed.c -2 +2 M ./mpi/rdft2-rank-geq2.c -2 +2 M ./mpi/rdft2-serial.c -2 +2 M ./mpi/rdft2-solve.c -2 +2 M ./mpi/rearrange.c -2 +2 M ./mpi/testsched.c -1 +1 M ./mpi/transpose-alltoall.c -2 +2 M ./mpi/transpose-pairwise.c -2 +2 M ./mpi/transpose-problem.c -2 +2 M ./mpi/transpose-recurse.c -2 +2 M ./mpi/transpose-solve.c -2 +2 M ./mpi/wisdom-api.c -2 +2 M ./rdft/buffered.c -2 +2 M ./rdft/buffered2.c -2 +2 M ./rdft/codelet-rdft.h -2 +2 M ./rdft/conf.c -2 +2 M ./rdft/ct-hc2c-direct.c -2 +2 M ./rdft/ct-hc2c.c -2 +2 M ./rdft/ct-hc2c.h -2 +2 M ./rdft/dft-r2hc.c -2 +2 M ./rdft/dht-r2hc.c -2 +2 M ./rdft/dht-rader.c -2 +2 M ./rdft/direct-r2c.c -2 +2 M ./rdft/direct-r2r.c -2 +2 M ./rdft/direct2.c -2 +2 M ./rdft/generic.c -2 +2 M ./rdft/hc2hc-direct.c -2 +2 M ./rdft/hc2hc-generic.c -2 +2 M ./rdft/hc2hc.c -2 +2 M ./rdft/hc2hc.h -2 +2 M ./rdft/indirect.c -2 +2 M ./rdft/khc2c.c -2 +2 M ./rdft/khc2hc.c -2 +2 M ./rdft/kr2c.c -2 +2 M ./rdft/kr2r.c -2 +2 M ./rdft/nop.c -2 +2 M ./rdft/nop2.c -2 +2 M ./rdft/plan.c -2 +2 M ./rdft/plan2.c -2 +2 M ./rdft/problem.c -2 +2 M ./rdft/problem2.c -2 +2 M ./rdft/rank-geq2-rdft2.c -2 +2 M ./rdft/rank-geq2.c -2 +2 M ./rdft/rank0-rdft2.c -2 +2 M ./rdft/rank0.c -2 +2 M ./rdft/rdft-dht.c -2 +2 M ./rdft/rdft.h -2 +2 M ./rdft/rdft2-inplace-strides.c -2 +2 M ./rdft/rdft2-rdft.c -2 +2 M ./rdft/rdft2-strides.c -2 +2 M ./rdft/rdft2-tensor-max-index.c -2 +2 M ./rdft/scalar/hb.h -2 +2 M ./rdft/scalar/hc2c.c -2 +2 M ./rdft/scalar/hc2cb.h -2 +2 M ./rdft/scalar/hc2cf.h -2 +2 M ./rdft/scalar/hf.h -2 +2 M ./rdft/scalar/hfb.c -2 +2 M ./rdft/scalar/r2c.c -2 +2 M ./rdft/scalar/r2cb.h -2 +2 M ./rdft/scalar/r2cbIII.h -2 +2 M ./rdft/scalar/r2cf.h -2 +2 M ./rdft/scalar/r2cfII.h -2 +2 M ./rdft/scalar/r2r.c -2 +2 M ./rdft/scalar/r2r.h -2 +2 M ./rdft/simd/common/genus.c -2 +2 M ./rdft/simd/hc2cbv.h -2 +2 M ./rdft/simd/hc2cfv.h -2 +2 M ./rdft/solve.c -2 +2 M ./rdft/solve2.c -2 +2 M ./rdft/vrank-geq1-rdft2.c -2 +2 M ./rdft/vrank-geq1.c -2 +2 M ./rdft/vrank3-transpose.c -2 +2 M ./reodft/conf.c -2 +2 M ./reodft/redft00e-r2hc-pad.c -2 +2 M ./reodft/redft00e-r2hc.c -2 +2 M ./reodft/reodft.h -2 +2 M ./reodft/reodft010e-r2hc.c -2 +2 M ./reodft/reodft11e-r2hc-odd.c -2 +2 M ./reodft/reodft11e-r2hc.c -2 +2 M ./reodft/reodft11e-radix2.c -2 +2 M ./reodft/rodft00e-r2hc-pad.c -2 +2 M ./reodft/rodft00e-r2hc.c -2 +2 M ./simd/altivec.c -2 +2 M ./simd/nonportable/sse.c -2 +2 M ./simd/nonportable/sse2.c -2 +2 M ./simd/simd-altivec.h -2 +2 M ./simd/simd-sse.h -2 +2 M ./simd/simd-sse2.h -2 +2 M ./simd/simd.h -2 +2 M ./simd/sse.c -2 +2 M ./simd/sse2.c -2 +2 M ./simd/taint.c -2 +2 M ./simd/x86-cpuid.h -2 +2 M ./simd-support/altivec.c -2 +2 M ./simd-support/amd64-cpuid.h -2 +2 M ./simd-support/avx.c -2 +2 M ./simd-support/simd-altivec.h -2 +2 M ./simd-support/simd-avx.h -2 +2 M ./simd-support/simd-common.h -2 +2 M ./simd-support/simd-sse2.h -2 +2 M ./simd-support/sse2-nonportable.c -2 +2 M ./simd-support/sse2.c -2 +2 M ./simd-support/taint.c -2 +2 M ./simd-support/x86-cpuid.h -2 +2 M ./threads/api.c -2 +2 M ./threads/conf.c -2 +2 M ./threads/ct.c -2 +2 M ./threads/dft-vrank-geq1.c -2 +2 M ./threads/f77api.c -2 +2 M ./threads/f77funcs.h -2 +2 M ./threads/hc2hc.c -2 +2 M ./threads/openmp.c -2 +2 M ./threads/rdft-vrank-geq1.c -2 +2 M ./threads/threads.c -2 +2 M ./threads/threads.h -2 +2 M ./threads/vrank-geq1-rdft2.c -2 +2 M ./tools/fftw-wisdom-to-conf.1 -4 +4 M ./tools/fftw-wisdom-to-conf.in -2 +2 M ./tools/fftw-wisdom.c -2 +2 M ./tools/fftw_wisdom.1.in -4 +4 Sat Jun 25 21:33:13 EDT 2011 stevenj@fftw.org * updated NEWS M ./NEWS -4 +53 Sat Jun 25 20:29:55 EDT 2011 stevenj@fftw.org * fixes to Fortran interface and docs M ./api/genf03.pl -12 +22 M ./doc/modern-fortran.texi -5 +14 M ./doc/mpi.texi -13 +16 Sat Jun 25 17:43:31 EDT 2011 stevenj@fftw.org * initial stab at MPI Fortran docs M ./doc/mpi.texi +138 Sat Jun 25 16:43:31 EDT 2011 stevenj@fftw.org * correct description of what MPI standard says about I/O (I can't believe this crap) M ./doc/mpi.texi -11 +21 Sat Jun 25 15:14:07 EDT 2011 stevenj@fftw.org * more MPI documentation; mention `fftw_alloc' functions earlier in the manual M ./doc/fftw3.texi +12 M ./doc/mpi.texi -30 +158 M ./doc/other.texi +9 M ./doc/tutorial.texi -11 +14 Sat Jun 25 13:40:19 EDT 2011 stevenj@fftw.org * clarification about --enable-sse2 M ./doc/install.texi -6 +8 Sat Jun 25 13:31:25 EDT 2011 athena@fftw.org * Update mingw build scripts for fftw-3.3 M ./kernel/kalloc.c -2 +2 M ./support/BUILD-MINGW32.sh -3 +3 M ./support/BUILD-MINGW64.sh -4 +4 Sat Jun 25 08:52:13 EDT 2011 athena@fftw.org * Fix typo: EXTRADIST => EXTRA_DIST M ./mpi/Makefile.am -1 +1 Fri Jun 24 23:52:19 EDT 2011 stevenj@fftw.org * finished draft "modern fortran" chapter M ./doc/modern-fortran.texi -9 +91 Fri Jun 24 20:47:49 EDT 2011 stevenj@fftw.org * include FFTW_EXTERN prototypes for wrappers, so that they are properly exported to DLLs on Windows (sigh) M ./mpi/f03-wrap.sh +4 Fri Jun 24 16:52:30 EDT 2011 athena@fftw.org * use malloc() instead of alloca() for large buffers The proximate cause for this patch is that OpenBSD/i386 reserves 256KB stack size per thread. We were allocating a buffer of size 128*130*sizeof(fftw_complex) that exceeds the stack. While 128*130*sizeof(fftw_complex) = 260KiB is the worst case for normal configurations, it is a good idea to limit stack allocation just in case. Also, the generic solver might in principle generate unbounded buffers, even though it is normally disabled for n > 137. So, as an added precaution, we now never stack-allocate buffers larger than 64KiB, which ought to be enough for anybody. M ./dft/dftw-direct.c -2 +3 M ./dft/direct.c -2 +3 M ./dft/generic.c -2 +3 M ./kernel/ifftw.h -7 +31 M ./rdft/ct-hc2c-direct.c -2 +3 M ./rdft/direct-r2c.c -2 +3 M ./rdft/generic.c -4 +6 M ./rdft/hc2hc-direct.c -2 +3 Fri Jun 24 16:32:30 EDT 2011 stevenj@fftw.org * don't imply that AVX is available on Pentium III; note that MIPS Paired Single is currently only in FFTW 3.2.x M ./doc/install.texi -3 +2 M ./doc/other.texi -3 +4 Fri Jun 24 16:05:27 EDT 2011 stevenj@fftw.org * silence annoying gfortran warnings M ./api/genf03.pl -1 +6 M ./configure.ac +14 M ./mpi/Makefile.am -3 +6 Fri Jun 24 14:59:30 EDT 2011 stevenj@fftw.org * a couple MPI Fortran 2003 fixes; changed MPI flags to not use 1<<31 since Fortran (not having unsigned integers) does not allow us to declare that constant in a portable way M ./mpi/api.c -1 +1 M ./mpi/f03api.sh -2 +2 M ./mpi/fftw3-mpi.h -4 +4 Fri Jun 24 15:05:05 EDT 2011 athena@fftw.org * Fix libtool shared version info. FFTW-3.3.x should be a direct drop-in replacement for all FFTW-3.x.y versions. M ./configure.ac -1 +4 Fri Jun 24 14:38:47 EDT 2011 stevenj@fftw.org * add MPI Fortran API and wrappers M ./api/Makefile.am -2 +3 M ./api/f03api.sh -3 +5 M ./api/genf03.pl -2 +14 M ./mpi/Makefile.am -3 +14 R ./mpi/README A ./mpi/f03-wrap.sh A ./mpi/f03api.sh M ./mpi/fftw3-mpi.h -4 +4 A ./mpi/genf03-wrap.pl Fri Jun 24 14:51:12 EDT 2011 athena@fftw.org * Do not require fig2dev on the user's machine Distribute the manual's figures in PDF/PS/PNG form instead. M ./doc/Makefile.am -10 +16 Fri Jun 24 11:52:44 EDT 2011 athena@fftw.org * Remove --enable-portable-binary, --with-gcc-arch from documentation. M ./NEWS +3 M ./doc/install.texi -23 +2 Fri Jun 24 11:48:48 EDT 2011 athena@fftw.org * Forget about specifying nonportable CFLAGS. Let the user do it if he wants. M ./m4/Makefile.am -4 +3 M ./m4/ax_cc_maxopt.m4 -24 +10 R ./m4/ax_gcc_archflag.m4 R ./m4/ax_gcc_x86_cpuid.m4 Fri Jun 24 11:48:25 EDT 2011 athena@fftw.org * Add "-avx" to version string when appropriate. M ./api/version.c +4 Fri Jun 24 10:26:38 EDT 2011 athena@fftw.org * change 3.3-alpha => 3.3-beta1 M ./configure.ac -1 +1 Fri Jun 24 09:25:49 EDT 2011 athena@fftw.org * Extend OUR_MALLOC16 to larger alignments Make it work for 32-byte alignment and beyond, as needed by AVX. Rename --with-our-malloc16 to --with-our-malloc. Keep old --with-our-malloc16 flag for compatibility. M ./configure.ac -3 +4 M ./kernel/kalloc.c -9 +10 Fri Jun 24 09:19:38 EDT 2011 athena@fftw.org * Fix typo M ./doc/install.texi -1 +1 Fri Jun 24 09:10:26 EDT 2011 athena@fftw.org * One pass over the manual. M ./NEWS -1 +6 M ./doc/Makefile.am -5 +2 M ./doc/install.texi -16 +10 M ./doc/intro.texi -1 +1 M ./doc/other.texi -2 +2 M ./doc/reference.texi -5 +5 M ./doc/tutorial.texi -32 +32 Fri Jun 24 08:19:03 EDT 2011 athena@fftw.org * eliminate the WITH_ALIGNED_STACK hack This is 2011 and I have no system with incorrect stack alignment. M ./TODO -2 M ./api/apiplan.c -9 +4 M ./api/execute-dft-c2r.c -2 +2 M ./api/execute-dft-r2c.c -2 +2 M ./api/execute-dft.c -2 +2 M ./api/execute-r2r.c -2 +2 M ./api/execute-split-dft-c2r.c -2 +2 M ./api/execute-split-dft-r2c.c -2 +2 M ./api/execute-split-dft.c -2 +2 M ./api/execute.c -2 +2 M ./api/f77funcs.h -16 +16 M ./doc/Makefile.am -2 +2 M ./doc/fftw3.texi -1 +1 M ./doc/intro.texi -3 +4 M ./doc/legacy-fortran.texi -1 +1 M ./doc/other.texi -51 +4 M ./doc/reference.texi -1 +1 M ./doc/tutorial.texi -9 +5 M ./kernel/ifftw.h -51 M ./threads/ct.c -2 +2 M ./threads/dft-vrank-geq1.c -2 +2 M ./threads/hc2hc.c -2 +2 M ./threads/rdft-vrank-geq1.c -2 +2 M ./threads/vrank-geq1-rdft2.c -2 +2 Fri Jun 24 07:49:47 EDT 2011 athena@fftw.org * enable both threaded and unthreaded wisdom in tools/fftw-wisdom M ./TODO -7 +2 M ./tools/fftw-wisdom.c -18 +14 Fri Jun 24 02:40:04 EDT 2011 stevenj@fftw.org * clarification M ./doc/modern-fortran.texi -2 +2 Fri Jun 24 02:24:01 EDT 2011 stevenj@fftw.org * check for error code in example M ./doc/modern-fortran.texi +2 Fri Jun 24 02:22:18 EDT 2011 stevenj@fftw.org * cleanup - since NATIVE_MALLOC is always malloc, delete this #define M ./api/export-wisdom-to-string.c -1 +1 M ./kernel/ifftw.h -2 Fri Jun 24 02:19:44 EDT 2011 stevenj@fftw.org * document wisdom string import/export in Fortran M ./doc/modern-fortran.texi +49 Fri Jun 24 02:11:40 EDT 2011 stevenj@fftw.org * bug fix - NATIVE_MALLOC should always be plain malloc, even in debug_malloc mode, because it is used in the API to return things that should be deallocated with free(); correspondingly, be sure to use free() ansd not X(free) with this M ./kernel/ifftw.h -1 +1 M ./mpi/wisdom-api.c -1 +1 Fri Jun 24 01:35:27 EDT 2011 stevenj@fftw.org * declaration style M ./doc/modern-fortran.texi -2 +4 Fri Jun 24 01:25:36 EDT 2011 stevenj@fftw.org * document wisdom file export/import from Fortran; add export/import_to/from_filename functions for convenience M ./api/export-wisdom-to-file.c +11 M ./api/fftw3.h +2 M ./api/import-wisdom-from-file.c +10 M ./doc/modern-fortran.texi -9 +79 M ./doc/other.texi -5 +6 M ./doc/reference.texi -11 +22 Thu Jun 23 19:19:43 EDT 2011 stevenj@fftw.org * more fortran docs M ./doc/legacy-fortran.texi -5 +6 M ./doc/modern-fortran.texi -10 +167 M ./doc/reference.texi +2 Thu Jun 23 17:50:30 EDT 2011 stevenj@fftw.org * enforce 132-character line-length limit that is the default in Fortran M ./api/genf03.pl -3 +26 Wed Jun 22 23:27:31 EDT 2011 stevenj@fftw.org * the F03 standard is ambiguous about whether types can be assigned to wider types as formal parameters with VALUE attributes, and e.g. gfortran interprets it to disallow this code M ./doc/modern-fortran.texi -2 +4 Thu Jun 23 18:12:10 EDT 2011 athena@fftw.org * Add md5 hash of fftw's configuration to wisdom file People were already confused by threaded vs unthreaded wisdom, and now things will be even worse because we enable/disable AVX codelets at runtime. Accept incoming wisdom only if it was produced by the same configuration (modulo MD5). M ./kernel/planner.c -2 +33 Thu Jun 23 09:01:27 EDT 2011 athena@fftw.org * distribute fftw3.f03.in M ./api/Makefile.am -1 +1 Wed Jun 22 22:02:18 EDT 2011 stevenj@fftw.org * more Fortran documentation M ./doc/modern-fortran.texi -1 +95 Wed Jun 22 20:10:39 EDT 2011 stevenj@fftw.org * correct comment M ./api/genf03.pl -1 +2 Wed Jun 22 20:26:18 EDT 2011 athena@fftw.org * Use "sh FOO.sh" instead of "./FOO.sh" to avoid chmod +x. M ./api/Makefile.am -1 +1 Wed Jun 22 19:19:05 EDT 2011 stevenj@fftw.org * document fftw_alloc_real/complex ... should we switch to using these in the tutorial examples? M ./doc/reference.texi +19 Wed Jun 22 19:07:49 EDT 2011 stevenj@fftw.org * whoops, added missing file A ./api/genf03.pl Wed Jun 22 18:46:01 EDT 2011 athena@fftw.org * Note addition of AVX. M ./NEWS +2 Wed Jun 22 18:43:43 EDT 2011 athena@fftw.org * In SSE2, AVX: use FMA macros when applicable. Makes it easier to play with fma4 and fma3 when it comes out. M ./simd-support/simd-avx.h -21 +21 M ./simd-support/simd-sse2.h -19 +19 Wed Jun 22 18:16:45 EDT 2011 stevenj@fftw.org * all modern Fortran compilers can call FFTW's C interfface directly -- support this, and in particular generate a Fortran 2003 interface file from fftw3.h so that Fortran code calling FFTW can be typechecked ((addressing the source of a lot of Fortran-user problems) ./doc/fortran.texi -> ./doc/legacy-fortran.texi M ./api/Makefile.am -3 +12 A ./api/f03api.sh M ./api/fftw3.h -3 +11 M ./api/malloc.c +18 M ./configure.ac +12 M ./doc/Makefile.am -1 +1 M ./doc/fftw3.texi -2 +4 M ./doc/install.texi -2 +2 M ./doc/intro.texi -5 +6 M ./doc/legacy-fortran.texi -23 +38 A ./doc/modern-fortran.texi M ./doc/mpi.texi -1 +1 M ./doc/other.texi -2 +4 M ./doc/upgrading.texi -1 +1 Wed Jun 22 13:10:02 EDT 2011 athena@fftw.org * some cleanup of SSE2 macros M ./simd-support/simd-sse2.h -27 +16 Wed Jun 22 07:38:18 EDT 2011 athena@fftw.org * don't use -xHost on ICC -xHost with ICC is problematic. On icc-12.0.0, "-mavx -xHost" overrides -mavx with -xHost, generating SSE2 code instead of AVX code. ICC does not seem to support -mtune=host or equivalent non-ABI changing flag. M ./m4/ax_cc_maxopt.m4 -31 +5 Tue Jun 21 20:35:36 EDT 2011 athena@fftw.org * Complete AVX implementation for split codelets M ./configure.ac -36 +38 M ./simd-support/simd-avx.h -41 +26 Tue Jun 21 19:37:14 EDT 2011 stevenj@fftw.org * whoops, missing altivec conf patches M ./dft/codelet-dft.h +1 M ./dft/conf.c +4 M ./rdft/codelet-rdft.h +1 M ./rdft/conf.c +4 Tue Jun 21 19:12:45 EDT 2011 stevenj@fftw.org * some BSD ar versions (e.g. on MacOS X) give an error if there are no object files, so we cannot build empty libraries M ./dft/simd/altivec/Makefile.am -4 +3 M ./dft/simd/avx/Makefile.am -3 +4 M ./dft/simd/sse2/Makefile.am -4 +3 M ./rdft/simd/altivec/Makefile.am -4 +3 M ./rdft/simd/avx/Makefile.am -3 +4 M ./rdft/simd/sse2/Makefile.am -3 +4 Tue Jun 21 19:12:12 EDT 2011 stevenj@fftw.org * re-insertion of Altivec code M ./Makefile.am -1 +6 M ./configure.ac -8 +13 M ./dft/simd/Makefile.am -1 +1 A ./dft/simd/altivec/ A ./dft/simd/altivec/Makefile.am M ./kernel/ifftw.h -2 +9 M ./rdft/simd/Makefile.am -1 +1 A ./rdft/simd/altivec/ A ./rdft/simd/altivec/Makefile.am M ./simd-support/Makefile.am -1 +1 A ./simd-support/altivec.c A ./simd-support/simd-altivec.h M ./simd-support/simd-common.h +3 Tue Jun 21 16:26:09 EDT 2011 athena@fftw.org * Implement faster AVX loads/stores. M ./simd-support/simd-avx.h -21 +28 Tue Jun 21 16:03:24 EDT 2011 athena@fftw.org * Initial AVX256/single implementation This should be correct but slow. I need to figure out how to implement noncontiguous loads/stores efficiently. M ./simd-support/simd-avx.h -53 +133 Tue Jun 21 14:13:57 EDT 2011 athena@fftw.org * fix AVX alignment M ./simd-support/simd-common.h -6 +8 Tue Jun 21 14:07:28 EDT 2011 athena@fftw.org * rename avx256d -> avx AVX will work in both double and single precision, like SSE2. ./dft/simd/avx256d -> ./dft/simd/avx ./rdft/simd/avx256d -> ./rdft/simd/avx ./simd-support/avx256d.c -> ./simd-support/avx.c ./simd-support/simd-avx256d.h -> ./simd-support/simd-avx.h M ./Makefile.am -4 +4 M ./configure.ac -11 +6 M ./dft/codelet-dft.h -1 +1 M ./dft/conf.c -3 +3 M ./dft/simd/Makefile.am -1 +1 M ./dft/simd/avx/Makefile.am -4 +4 M ./kernel/ifftw.h -2 +2 M ./rdft/codelet-rdft.h -1 +1 M ./rdft/conf.c -3 +3 M ./rdft/simd/Makefile.am -1 +1 M ./rdft/simd/avx/Makefile.am -4 +4 M ./simd-support/Makefile.am -1 +1 M ./simd-support/avx.c -3 +3 M ./simd-support/simd-avx.h -2 +2 Tue Jun 21 13:52:20 EDT 2011 athena@fftw.org * remove CODELET_OPTIM In the old 32-bit gcc-3.x days we used to play games with gcc to force it to produce decent code. Now gcc has gotten smarter and it produces indecent code no matter what we do, so it is safe to remove these hacks. M ./api/version.c -5 +5 M ./configure.ac -40 M ./support/Makefile.codelets -6 Tue Jun 21 09:57:31 EDT 2011 athena@fftw.org * work around gcc/icc quirks M ./simd-support/simd-avx256d.h -14 +45 Tue Jun 21 09:56:07 EDT 2011 athena@fftw.org * Add remarks in places where we work around gcc quirks M ./simd-support/simd-sse2.h -7 +16 Mon Jun 20 21:17:59 EDT 2011 stevenj@fftw.org * remove the libbench directory (which we have kept lingering in the repository for years due to CVS's inability to remove directories) R ./libbench/Makefile.am R ./libbench/accopy-from.c R ./libbench/accopy-to.c R ./libbench/acopy.c R ./libbench/allocate.c R ./libbench/ascale.c R ./libbench/aset.c R ./libbench/bench-main.c R ./libbench/bench-user.h R ./libbench/bench.h R ./libbench/caadd.c R ./libbench/cacopy.c R ./libbench/can-do.c R ./libbench/cascale.c R ./libbench/caset.c R ./libbench/casub.c R ./libbench/ccopy-from.c R ./libbench/ccopy-to.c R ./libbench/copy-c2c-from.c R ./libbench/copy-c2c-to.c R ./libbench/copy-c2h-1d-fftpack.c R ./libbench/copy-c2h-1d-halfcomplex.c R ./libbench/copy-c2h-1d-packed.c R ./libbench/copy-c2h-1d-unpacked-ri.c R ./libbench/copy-c2h-unpacked.c R ./libbench/copy-c2h.c R ./libbench/copy-c2r-packed.c R ./libbench/copy-c2r-unpacked.c R ./libbench/copy-c2r.c R ./libbench/copy-c2ri.c R ./libbench/copy-h2c-1d-fftpack.c R ./libbench/copy-h2c-1d-halfcomplex.c R ./libbench/copy-h2c-1d-packed.c R ./libbench/copy-h2c-1d-unpacked-ri.c R ./libbench/copy-h2c-unpacked.c R ./libbench/copy-h2c.c R ./libbench/copy-r2c-packed.c R ./libbench/copy-r2c-unpacked.c R ./libbench/copy-r2c.c R ./libbench/copy-ri2c.c R ./libbench/deallocate.c R ./libbench/getopt-utils.c R ./libbench/getopt.c R ./libbench/getopt.h R ./libbench/getopt1.c R ./libbench/info.c R ./libbench/log2.c R ./libbench/main.c R ./libbench/mflops.c R ./libbench/mp.c R ./libbench/ovtpvt.c R ./libbench/pow2.c R ./libbench/prime.c R ./libbench/problem.c R ./libbench/report.c R ./libbench/speed.c R ./libbench/timer.c R ./libbench/unnormalize.c R ./libbench/util.c R ./libbench/verify.c R ./libbench/zero.c R ./libbench/ Mon Jun 20 21:17:14 EDT 2011 stevenj@fftw.org * update URLs M ./doc/install.texi -2 +2 M ./doc/intro.texi -1 +1 M ./doc/license.texi -1 +1 M ./doc/mpi.texi -1 +1 M ./doc/other.texi -1 +1 Mon Jun 20 20:53:31 EDT 2011 stevenj@fftw.org * whoops, forgot to check in alignment change M ./simd-support/simd-common.h -3 +6 Mon Jun 20 20:22:23 EDT 2011 athena@fftw.org * "test X = Y" requires spaces around "=" M ./configure.ac -1 +1 Mon Jun 20 19:18:52 EDT 2011 stevenj@fftw.org * indenting M ./simd-support/simd-sse2.h -3 +3 Mon Jun 20 18:57:10 EDT 2011 stevenj@fftw.org * merge back in SSE support, now combined with SSE2; --enable-sse2 now works in both single and double precision, and simd-sse2.h contains both the double- and single-precision code (which overlap a lot); in single precision it is still compiled for SSE-only (SSE2 is only required for double) M ./configure.ac -16 +9 M ./simd-support/simd-sse2.h -54 +181 M ./simd-support/sse2-nonportable.c -2 +6 M ./simd-support/sse2.c -4 +11 Mon Jun 20 16:02:07 EDT 2011 athena@fftw.org * Implement AVX autodetection (gcc-only so far) M ./simd-support/Makefile.am -1 +1 A ./simd-support/amd64-cpuid.h M ./simd-support/avx256d.c -2 +32 M ./simd-support/x86-cpuid.h +33 Mon Jun 20 14:25:54 EDT 2011 athena@fftw.org * Add VZEROUPPER at the end of AVX codelets If the Intel Optimization Manual is to be believed, we need to wave a dead chicken before transitioning from AVX code to SSE code. I am supposed to believe that there is a transition penalty for doing so, unless one uses a magic VZEROUPPER instruction that apparently has zero cost. Whatever. M ./genfft/c.ml +2 M ./genfft/c.mli +1 M ./genfft/gen_hc2c.ml -1 +1 M ./genfft/gen_hc2cdft.ml -1 +1 M ./genfft/gen_hc2cdft_c.ml -1 +1 M ./genfft/gen_hc2hc.ml -1 +1 M ./genfft/gen_mdct.ml -1 +1 M ./genfft/gen_notw.ml -1 +1 M ./genfft/gen_notw_c.ml -1 +1 M ./genfft/gen_r2cb.ml -1 +1 M ./genfft/gen_r2cf.ml -1 +1 M ./genfft/gen_r2r.ml -1 +1 M ./genfft/gen_twiddle.ml -1 +1 M ./genfft/gen_twiddle_c.ml -1 +1 M ./genfft/gen_twidsq.ml -1 +1 M ./genfft/gen_twidsq_c.ml -1 +1 M ./genfft/genutil.ml -2 +2 M ./genfft/simd.ml +1 M ./simd-support/simd-avx256d.h +5 M ./simd-support/simd-sse2.h +2 Mon Jun 20 10:21:25 EDT 2011 athena@fftw.org * Move RDFT to new simd scheme A ./rdft/simd/avx256d/ R ./rdft/simd/codelets/Makefile.am R ./rdft/simd/codelets/ A ./rdft/simd/common/ A ./rdft/simd/sse2/ M ./Makefile.am -2 +4 M ./configure.ac +4 M ./dft/simd/Makefile.am -1 +1 M ./genfft/gen_hc2cdft_c.ml -2 +2 M ./kernel/ifftw.h +1 M ./rdft/Makefile.am -1 +1 M ./rdft/codelet-rdft.h -1 +2 M ./rdft/conf.c -2 +7 M ./rdft/simd/Makefile.am -6 +2 A ./rdft/simd/avx256d/Makefile.am A ./rdft/simd/codlist.mk A ./rdft/simd/common/Makefile.am A ./rdft/simd/common/genus.c R ./rdft/simd/hc2cbv.c M ./rdft/simd/hc2cbv.h -2 +2 R ./rdft/simd/hc2cfv.c M ./rdft/simd/hc2cfv.h -2 +2 A ./rdft/simd/simd.mk A ./rdft/simd/sse2/Makefile.am Mon Jun 20 09:23:38 EDT 2011 athena@fftw.org * New SIMD build system We now support multiple SIMD extensions in the same binary, e.g. --enable-sse2 --enable-avx. This patch adds the necessary infrastructure for SSE2/AVX and complex DFT. Later patches will add RDFT and SSE/ALTIVEC/etc. A ./dft/simd/avx256d/ A ./simd-support/ R ./dft/simd/codelets/Makefile.am R ./dft/simd/codelets/ A ./dft/simd/common/ A ./dft/simd/sse2/ M ./Makefile.am -11 +13 M ./configure.ac -47 +74 M ./dft/codelet-dft.h -4 +2 M ./dft/conf.c -2 +7 M ./dft/scalar/codelets/Makefile.am -1 +1 M ./dft/simd/Makefile.am -6 +3 A ./dft/simd/avx256d/Makefile.am A ./dft/simd/codlist.mk A ./dft/simd/common/Makefile.am A ./dft/simd/common/genus.c R ./dft/simd/n1b.c M ./dft/simd/n1b.h -2 +2 R ./dft/simd/n1f.c M ./dft/simd/n1f.h -2 +2 R ./dft/simd/n2b.c M ./dft/simd/n2b.h -2 +2 R ./dft/simd/n2f.c M ./dft/simd/n2f.h -2 +2 R ./dft/simd/n2s.c M ./dft/simd/n2s.h -2 +2 R ./dft/simd/q1b.c M ./dft/simd/q1b.h -2 +2 R ./dft/simd/q1f.c M ./dft/simd/q1f.h -2 +2 A ./dft/simd/simd.mk A ./dft/simd/sse2/Makefile.am R ./dft/simd/t.c M ./dft/simd/t1b.h -2 +2 M ./dft/simd/t1bu.h -2 +2 M ./dft/simd/t1f.h -2 +2 M ./dft/simd/t1fu.h -2 +2 M ./dft/simd/t2b.h -2 +2 M ./dft/simd/t2f.h -2 +2 M ./dft/simd/t3b.h -2 +2 M ./dft/simd/t3f.h -2 +2 R ./dft/simd/ts.c M ./dft/simd/ts.h -2 +2 M ./genfft/gen_twiddle.ml -2 +2 M ./genfft/genutil.ml -2 +5 M ./kernel/ifftw.h -2 +6 M ./rdft/Makefile.am -1 +1 M ./rdft/conf.c -1 +1 M ./rdft/scalar/r2cb/Makefile.am +1 M ./rdft/scalar/r2cf/Makefile.am +1 M ./rdft/scalar/r2r/Makefile.am +1 A ./simd-support/Makefile.am A ./simd-support/avx256d.c A ./simd-support/simd-avx256d.h A ./simd-support/simd-common.h A ./simd-support/simd-sse2.h A ./simd-support/sse2-nonportable.c A ./simd-support/sse2.c A ./simd-support/taint.c A ./simd-support/x86-cpuid.h M ./support/Makefile.codelets -21 +22 Sun Jun 19 12:29:27 EDT 2011 stevenj@fftw.org * italicize Latin quote M ./doc/reference.texi -2 +2 Sun Jun 19 12:26:34 EDT 2011 stevenj@fftw.org * work around incredibly annoying makeinfo bug -- for HTML output, in any paragraph ending with an @index command, two blank lines are needed to create a paragraph break ... our HTML output has apparently been screwed up for years M ./doc/acknowledgements.texi +2 M ./doc/fortran.texi +1 M ./doc/install.texi +2 M ./doc/intro.texi +1 M ./doc/mpi.texi +20 M ./doc/other.texi +2 M ./doc/reference.texi +21 M ./doc/threads.texi +2 M ./doc/tutorial.texi +13 Sun Jun 19 12:01:39 EDT 2011 stevenj@fftw.org * conjugate-pair algorithm turns out not to be due to djb, but it was pointed out to us by djb M ./doc/intro.texi -3 +4 Sun Jun 19 11:59:21 EDT 2011 stevenj@fftw.org * new-array execute functions are *not* the same for MPI, since the problems are different M ./doc/mpi.texi -2 +3 Sun Jun 19 11:57:52 EDT 2011 stevenj@fftw.org * tweaks to MPI manual M ./doc/mpi.texi -19 +44 M ./doc/reference.texi -1 +1 Sun Jun 19 00:47:23 EDT 2011 stevenj@fftw.org * only call MPI_Init_thread for MPI version >= 2 M ./mpi/mpi-bench.c +5 Sun Jun 19 00:35:44 EDT 2011 stevenj@fftw.org * document quad precision in Fortran M ./doc/fortran.texi -9 +12 Sun Jun 19 00:16:10 EDT 2011 stevenj@fftw.org * use -lquadmath for quad-precision library in pkgconfig file M ./configure.ac +2 M ./fftw.pc.in -1 +1 Sun Jun 19 00:10:33 EDT 2011 stevenj@fftw.org * document quad precision M ./doc/install.texi -1 +11 M ./doc/reference.texi +8 Sat Jun 18 21:26:24 EDT 2011 stevenj@fftw.org * there is currently no standard quad-precision type in MPI, so don't allow MPI support to be compiled with --enable-quad-precision M ./configure.ac +3 M ./mpi/ifftw-mpi.h +2 M ./mpi/mpi-bench.c +2 Sat Jun 18 21:25:51 EDT 2011 stevenj@fftw.org * make quad-precision library libfftwq M ./configure.ac +1 Sat Jun 18 21:19:50 EDT 2011 stevenj@fftw.org * use --estimate in check script with --enable-random-estimator M ./configure.ac +2 M ./mpi/Makefile.am -1 +1 M ./tests/Makefile.am -5 +5 Sat Jun 18 20:36:51 EDT 2011 stevenj@fftw.org * bug fix - correct crashing interaction between threads and debug-malloc M ./threads/threads.c +4 Sat Jun 18 18:41:48 EDT 2011 stevenj@fftw.org * whoops, typo M ./doc/reference.texi -1 +1 Sat Jun 18 18:35:14 EDT 2011 stevenj@fftw.org * clarification of when fftw_cost may return 0 M ./doc/reference.texi -2 +5 Sat Jun 18 18:31:26 EDT 2011 stevenj@fftw.org * corrected manual and test program for proper interaction of MPI and threads M ./doc/mpi.texi -20 +59 M ./mpi/mpi-bench.c +13 M ./tests/fftw-bench.c -4 +10 M ./tests/fftw-bench.h +5 Sat Jun 18 17:13:52 EDT 2011 stevenj@fftw.org * need --mpi restriction in mpi+threads check M ./mpi/Makefile.am -1 +1 Sat Jun 18 12:09:04 EDT 2011 stevenj@fftw.org * split fftw3.texi into multiple files for ease of editing M ./doc/Makefile.am -2 +1 A ./doc/acknowledgements.texi A ./doc/cindex.texi M ./doc/fftw3.texi -6158 +14 A ./doc/findex.texi A ./doc/fortran.texi A ./doc/install.texi A ./doc/intro.texi A ./doc/license.texi A ./doc/mpi.texi A ./doc/other.texi A ./doc/reference.texi A ./doc/threads.texi A ./doc/tutorial.texi A ./doc/upgrading.texi Fri Jun 17 17:26:50 EDT 2011 stevenj@fftw.org * merge recent Cell deletion with MPI branch M ./Makefile.am -3 +2 M ./NEWS +40 M ./configure.ac -1 +1 M ./doc/fftw3.texi -95 +909 Fri Jun 17 01:52:59 EDT 2011 stevenj@fftw.org * whoops, incorrect assertion M ./mpi/transpose-problem.c -2 +2 Fri Jun 17 01:52:51 EDT 2011 stevenj@fftw.org * comment fix M ./mpi/block.c -2 +2 Thu Jun 16 23:30:27 EDT 2011 stevenj@fftw.org * check if pln creation failed (e.g. for split input) bbefore calling setup_gather_scatter, to prevent crashes M ./mpi/mpi-bench.c -1 +1 Thu Jun 16 23:26:48 EDT 2011 stevenj@fftw.org * bug fix -- transpose-recurse is only applicable if subtransposes fit in the same space (unless I change the allocation routine, but this would seem to require looking at all possible recursive invocations of transpose-recurse) M ./mpi/transpose-recurse.c +28 Mon Apr 11 17:58:03 EDT 2011 stevenj@fftw.org * yikes, any_true check on subplan creation should be in comm, not comm2, so that all processes know if failure occurred M ./mpi/transpose-recurse.c -6 +4 Mon Apr 11 17:00:46 EDT 2011 stevenj@fftw.org * add wisdom_ok_hook to enforce wisdom synchronization on MPI problems, apparently fixing a longstanding deadlock/crash bug M ./kernel/ifftw.h +1 M ./kernel/planner.c +7 M ./mpi/api.c -13 +61 Mon Apr 11 15:10:22 EDT 2011 stevenj@fftw.org * add a check (in DEBUG mode only) that all processes produce the same hash of MPI problems; don't include alignment in MPI problem hash because it may differ between processes for unaligned malloc M ./mpi/any-true.c +21 M ./mpi/dft-problem.c -2 +6 M ./mpi/ifftw-mpi.h +1 M ./mpi/rdft-problem.c -2 +6 M ./mpi/rdft2-problem.c -2 +6 M ./mpi/transpose-problem.c -2 +6 Fri Apr 8 18:46:54 EDT 2011 stevenj@fftw.org * use cost_hook in random_estimate M ./kernel/planner.c -1 +1 Sun Mar 6 23:33:53 EST 2011 stevenj@fftw.org * added mpi new-array execute functions; thanks to Guo Luo for the bug report M ./mpi/api.c +23 M ./mpi/fftw3-mpi.h -1 +6 Wed Feb 9 21:29:17 EST 2011 stevenj@fftw.org * MPI may not support tags > 2^15-1 (e.g. Cray MPI requires tags < 2^24); thanks to Jonathan Bentz for the bug report. M ./mpi/transpose-pairwise.c -4 +4 Wed Feb 2 12:21:30 EST 2011 stevenj@fftw.org * fix merge conflicts M! ./configure.ac -1 +1 Sat Nov 15 22:33:20 EST 2008 stevenj@fftw.org tagged fftw-3.3alpha1 Sat Nov 15 20:33:33 EST 2008 stevenj@fftw.org * version bump for 3.3alpha1 M! ./NEWS -40 M! ./configure.ac -2 +2 Sun Oct 26 22:47:07 EDT 2008 stevenj@fftw.org * re-added mpi/Makefile M ./configure.ac +2 Sat Oct 25 17:14:42 EDT 2008 stevenj@fftw.org * re-add MPI to dist M! ./Makefile.am -2 +3 M! ./doc/fftw3.texi -9 +134 Sat Jun 18 08:50:13 EDT 2011 athena@fftw.org * remove obsolete Cell code M ./kernel/align.c -1 +1 M ./kernel/ifftw.h -9 +2 M ./rdft/rank0.c -83 M ./simd/simd.h -6 M ./simd/taint.c -1 +1 M ./tests/fftw-bench.c -7 Fri Jun 17 23:31:33 EDT 2011 stevenj@fftw.org * bug fix in accuracy test, which prevented us from consiistently determining accuracy in > double precision M ./libbench2/mp.c -2 +2 Fri Jun 17 20:05:13 EDT 2011 athena@fftw.org * do not check for gcc version before checking for gcc M ./configure.ac -1 +1 Fri Jun 17 18:56:37 EDT 2011 stevenj@fftw.org * require gcc 4.6.0 or later for --enable-quad-precision, to match fftw3.h header file; no need to mark this as EXPERIMENTAL (make check passes, and support in gcc 4.6 seems reasonably complete) M ./configure.ac -1 +2 Fri Jun 17 18:51:41 EDT 2011 stevenj@fftw.org * need ugly __attribute__ to use __float128 with _Complex, ugh M ./api/fftw3.h +9 Fri Jun 17 18:23:05 EDT 2011 stevenj@fftw.org * --verify tolerance in quad precision changed to 1e-29 M ./libbench2/bench-main.c -1 +1 Fri Jun 17 18:22:38 EDT 2011 stevenj@fftw.org * quad-precision F77 api should use "qfftw" prefix M ./api/x77.h +3 Fri Jun 17 18:22:27 EDT 2011 stevenj@fftw.org * rm extraneous space from fftw3.h M ./api/fftw3.h -1 +1 Fri Jun 17 18:05:10 EDT 2011 stevenj@fftw.org * use cosq etcetera with libquadmath in libbench2, so that --verify correctly gives ~33 decimal places in shift test M ./libbench2/verify.h -1 +4 Fri Jun 17 17:52:51 EDT 2011 stevenj@fftw.org * libquadmath ships with gcc 4.6.0, so we should require this library for sinq/cosq with --enable-quad-precision; also, include the __float128 FFTW functions in the header file for gcc >= 4.6 on i86/x86??_64/ia64 M ./Makefile.am +5 M ./api/fftw3.h -1 +6 M ./configure.ac -1 +5 M ./kernel/trig.c -13 +4 Fri Jun 17 16:54:01 EDT 2011 stevenj@fftw.org * typo in manual for fftw_cost M ./doc/fftw3.texi -1 +1 Fri Jun 17 16:48:24 EDT 2011 stevenj@fftw.org * fix fftw_cost function: pcost needs to be saved in mkapiplan, since the plan is re-created from wisdom M ./api/apiplan.c +5 M ./tests/fftw-bench.c -2 +3 Fri Jun 17 16:42:25 EDT 2011 athena@fftw.org * removed support for the Cell Broadband Engine R ./cell/spu/Makefile.am R ./cell/spu/alloc.spuc R ./cell/spu/copy.spuc R ./cell/spu/dft.spuc R ./cell/spu/dma.spuc R ./cell/spu/execute.spuc R ./cell/spu/fftw-spu.h R ./cell/spu/main.spuc R ./cell/spu/planner.spuc R ./cell/spu/spu-double.h R ./cell/spu/spu-single.h R ./cell/spu/transpose.spuc R ./cell/spu/ R ./cell/Makefile.am R ./cell/cell.c R ./cell/conf.c R ./cell/copy.c R ./cell/dft-direct-cell.c R ./cell/fftw-cell.h R ./cell/plans-double.c R ./cell/plans-single.c R ./cell/spufftw-embed.S R ./cell/transpose.c R ./cell/ M ./AUTHORS -2 +3 M ./Makefile.am -11 +5 M ./NEWS +2 R ./README.Cell M ./api/configure.c -3 M ./configure.ac -18 +1 M ./dft/dft.h -4 M ./dft/rank-geq2.c -6 M ./doc/fftw3.texi -128 +4 Tue May 24 06:51:07 EDT 2011 athena@fftw.org * Undo previous change; the typo was not a typo after all. M ./doc/fftw3.texi -1 +1 Mon May 23 05:08:05 EDT 2011 athena@fftw.org * Fix typo in manual M ./doc/fftw3.texi -1 +1 Sat May 21 17:37:50 EDT 2011 athena@fftw.org * clarify intent about canonicalization of tensor in tensor_compress_contiguous() M ./kernel/tensor7.c -1 +9 Sat May 21 17:30:31 EDT 2011 athena@fftw.org * avoid useless canonicalization in tensor_compress_contiguous() M ./kernel/tensor7.c -1 +1 Sat May 21 17:24:57 EDT 2011 athena@fftw.org * Fix tensor_compress_contiguous tensor_compress_contiguous() was supposed to sort dimensions by descending istride, and then compress adjacent dimensions. This property was lost once we changed the canonical order of strides to be sorted by descending min{istride,ostride}. Change tensor_compress_contiguous() to sort by descending istride again, which is necessary for its correctness, and then canonicalize at the end. M ./kernel/tensor7.c -20 +41 Sun May 8 18:47:26 EDT 2011 athena@fftw.org * Don't distribute obsolete .depend M ./genfft/Makefile.am -2 +2 Sun May 8 18:05:36 EDT 2011 athena@fftw.org * Use ocamlbuild for building genfft Remove the old Makefile cruft to support ocaml, and use ocamlbuild instead. M ./bootstrap.sh -2 M ./configure.ac -10 +1 M ./genfft/Makefile.am -151 +8 M ./m4/Makefile.am -1 +1 R ./m4/ocaml.m4 M ./support/Makefile.codelets -13 +13 Sun May 8 18:03:07 EDT 2011 athena@fftw.org * Do not use __float128 unless BENCHFFT_QUAD is defined Otherwise, compilation fails on compilers that do not support __float128. M ./libbench2/bench-user.h +5 Fri Apr 8 13:15:54 EDT 2011 stevenj@fftw.org * fix configure --help string for --disable-alloca (since default is enabled) M ./configure.ac -1 +1 Fri Apr 8 13:09:56 EDT 2011 stevenj@fftw.org * add "random estimator" for debugging purposes; note that this is best used with ESTIMATE_PATIENT mode M ./api/apiplan.c +4 M ./configure.ac +5 M ./kernel/ifftw.h +4 M ./kernel/planner.c +27 M ./tests/fftw-bench.c +7 Tue Apr 5 14:47:56 EDT 2011 stevenj@fftw.org * add AC_CHECK_DECLS for srand48; thanks to Ralf Wildenhues for the bug report M ./configure.ac -1 +1 M ./libbench2/util.c +3 Sat Feb 5 17:00:40 EST 2011 stevenj@fftw.org * experimental support for gcc's __float128 quad-precision type M ./api/fftw3.h +3 M ./configure.ac -2 +14 M ./kernel/ifftw.h -2 +10 M ./kernel/trig.c -1 +16 M ./libbench2/bench-main.c +2 M ./libbench2/bench-user.h +4 M ./libbench2/info.c -1 +2 M ./libbench2/verify.h +6 M ./simd/simd-sse2.h -1 +1 M ./tests/fftw-bench.h +2 M ./tools/fftw-wisdom.c +2 Sun Oct 24 14:33:59 EDT 2010 athena@fftw.org * guarantee that "timelimit < 0" means "no timeout" "timelimit < 0" was always meant to be equivalent to "timelimit = HUGENUM", but this was not true in all cases, causing some obscure wisdom behavior. Thanks William Andrew Burnson for the bug report. M ./api/mapflags.c -1 +1 Sun Oct 24 14:32:20 EDT 2010 athena@fftw.org * compile with --enable-fma and SSE, SSE2 Allow compilation with --enable-fma and --enable-sse, --enable-sse2. This is a bad idea performance-wise, but people will try anyway. M ./simd/simd-sse.h +5 M ./simd/simd-sse2.h +5 Sun Jul 11 13:34:06 EDT 2010 athena@fftw.org * Make threads.c compiler with c++ M ./threads/threads.c -1 +1 Sun Jul 11 10:05:05 EDT 2010 athena@fftw.org * Attempt at clarifying the advanced interface doc. M ./doc/fftw3.texi -72 +151 Sun Jul 11 07:37:27 EDT 2010 athena@fftw.org * rename rfftwnd html picture It turns out that texinfo with pdf output reads .png files in preference to .pdf files (when did this change?). I renamed the .png figure to avoid producing an ugly pdf file. M ./doc/Makefile.am -4 +4 M ./doc/fftw3.texi +5 Tue Mar 30 19:43:22 EDT 2010 stevenj@fftw.org * added fftw_cost function; this is the second time people have asked for this, and there is a reasonable use for it in comparing e.g. oout-of-place vs. in-place plans M ./NEWS +4 M ./api/f77funcs.h +10 M ./api/fftw3.h +1 M ./api/flops.c +5 M ./doc/fftw3.texi +17 Tue Mar 2 18:55:49 EST 2010 stevenj@fftw.org * documented that --enable-debug-malloc causes fftw_execute to be thread-unsafe (thanks to Alexis Rohou for the problem report) M ./doc/fftw3.texi +8 Fri Jan 22 19:42:08 EST 2010 athena@fftw.org * Added FAQ about how to transpose matrices using FFTW. M ./doc/FAQ/fftw-faq.bfnn +33 Thu Jan 7 20:16:57 EST 2010 stevenj@fftw.org * catch FMS (instead of generating FMA(_,_,NEG(_)) with h -generic-arith option M ./genfft/c.ml +2 Fri Dec 11 07:01:26 EST 2009 athena@fftw.org * note future wisdom enhancements. M ./TODO +5 Mon Oct 19 20:21:05 EDT 2009 athena@fftw.org * Use SIMD flags when checking for xmmintrin.h This prevents an obnoxious warning from configure. M ./configure.ac -3 +3 Sat Aug 29 20:47:56 EDT 2009 athena@fftw.org * new bug M ./TODO +4 Sun Jul 26 00:40:11 EDT 2009 stevenj@fftw.org * typo (s/man1/many) M ./doc/fftw3.texi -1 +1 Tue Jul 14 14:19:08 EDT 2009 stevenj@fftw.org * BUILD-MINGW32 script, updated Windows README A ./support/BUILD-MINGW32.sh M ./support/BUILD-MINGW64.sh -4 +4 Mon Jul 13 09:40:38 EDT 2009 fftw@fftw.org * cleanup BUILD-MINGW64.sh M ./support/BUILD-MINGW64.sh -19 +27 Sun Jul 12 06:34:46 EDT 2009 athena@fftw.org * Update NEWS, version number for 3.2.2 release. M ./NEWS +3 M ./configure.ac -2 +2 Sat Jul 11 22:28:38 EDT 2009 athena@fftw.org * Reintroduce the pruning heuristic in ESTIMATE mode for r2r problems. Somehow, we lost this feature between fftw-3.1.3 and fftw-3.2. M ./rdft/hc2hc-direct.c +3 Thu Jun 25 07:39:04 EDT 2009 athena@fftw.org * don't use pshared=1 in sem_init pshared is really not necessary, and it is not supported on GNU/kFreeBSD. Thanks Petr Salinger for the bug report. M ./threads/threads.c -2 +2 Thu Jun 11 19:35:40 EDT 2009 fftw@fftw.org * Add mingw64 build file so that we can track it. A ./support/BUILD-MINGW64.sh Wed Jun 10 12:10:58 EDT 2009 fftw@fftw.org * note 3.2.2 NEWS M ./NEWS +7 Wed Jun 10 12:04:54 EDT 2009 fftw@fftw.org * add --disable-alloca to configure It looks like alloca() is broken on mingw64, and thus we need to disable it explicitly. M ./configure.ac -1 +6 M ./kernel/ifftw.h -1 +1 Sun Apr 26 16:33:10 EDT 2009 athena@fftw.org * Note in FAQ that --enable-k7 has been discontinued. M ./doc/FAQ/fftw-faq.bfnn -4 +7 Tue Mar 24 09:16:18 EDT 2009 athena@fftw.org * clarified small confusion in fftw_cleanup documentation M ./doc/fftw3.texi -2 +3 Thu Mar 19 13:18:06 EDT 2009 stevenj@fftw.org * fix documentation of dfftw_init_threads to indicate thaat it takes an argument (since the C version returns a value); thanks t Hans Johnnston for the bug report M ./doc/fftw3.texi -1 +5 Thu Mar 12 13:12:13 EDT 2009 fftw@fftw.org * if possible, use a 128-bit type for copy M ./configure.ac -1 +1 M ./kernel/cpy2d.c -9 +38 Tue Mar 10 12:49:51 EDT 2009 fftw@fftw.org * add size-128 simd codelets It's about time M ./dft/simd/codelets/Makefile.am -2 +2 Mon Mar 9 20:29:16 EDT 2009 athena@fftw.org * copy two floats as a double when possible Resurrect the old hack of copying two floats as a double, which makes some difference in these days of 64 bit boxes. M ./kernel/cpy2d.c -16 +34 Sun Mar 8 18:08:04 EDT 2009 athena@fftw.org * fixed (harmless) confusion of strides RS and VS were swapped in dftw-direct.c. This is a bug, but it is harmless unless one uses fixed-stride codelets, which we do not. M ./dft/dftw-direct.c -7 +7 Sun Mar 8 10:29:49 EDT 2009 athena@fftw.org * oops, I checked in debug code accidentally. M ./kernel/ct.c -1 Sat Feb 14 19:01:00 EST 2009 athena@fftw.org * Change TLO email address since Magdalen is no longer there. M ./commercialize.sh -1 +1 Sat Feb 14 18:18:45 EST 2009 stevenj@fftw.org * quote arguments to bench in test script on the off-chance that '*' would be expanded by the shell into a valid filename, and also to avoid shell confusion on Cygwin that "//" begins the name of a Windows network mountpoint M ./tests/check.pl -1 +1 Sat Feb 14 18:17:23 EST 2009 stevenj@fftw.org tagged fftw-3.2.1 Sun Mar 8 10:02:59 EDT 2009 athena@fftw.org * stricter conditions for Cooley-Tukey being ugly It turns out that m=2 in the leaf of Cooley-Tukey may be advantageous in certain cases, eg. i512v512 on AMD Shanghai: (dft-buffered-512-x128/512-6 (dft-ct-dit/4 (dftw-direct-4/24-x128 "t2fv_4") (dft-vrank>=1-x4/1 (dft-ct-dit/64 (dftw-direct-64/504-x128 "t2fv_64") (dft-vrank>=1-x64/1 (dft-direct-2-x128 "n2fv_2"))))) (dft-r2hc-1 (rdft-rank0-tiled/2-x128-x512)) (dft-nop)) Presumably this works around the 2 way associativity of the L1 cache. M ./dft/dftw-direct.c -1 +1 M ./kernel/ct.c -2 +3 M ./kernel/ifftw.h -1 +1 M ./rdft/ct-hc2c-direct.c -1 +1 M ./rdft/hc2hc-direct.c -3 +3 Mon Feb 9 19:46:00 EST 2009 stevenj@fftw.org * disable Windows QueryPerformanceCounter code, since it requires us to pull in windows.h in ifftw.h and causes namespace conflicts; gettimeofday seems to work well enough and has had few complaints M ./kernel/ifftw.h -4 +1 M ./kernel/timer.c -1 +8 Wed Feb 4 22:55:54 EST 2009 stevenj@fftw.org * version bump for 3.2.1, updated NEWS M ./NEWS +10 M ./configure.ac -1 +1 Wed Feb 4 22:27:28 EST 2009 stevenj@fftw.org * recommend that users avoid fftw_execute in Fortran, instead using dfftw_execute_dft and friends so that the compiler knows that the input/output arrays are used M ./doc/fftw3.texi -10 +80 Wed Jan 21 16:02:08 EST 2009 stevenj@fftw.org * prefer windows queryperformancecounter to gettimeofday on Windows, thanks to David Price for the suggestion M ./kernel/ifftw.h -1 +4 M ./kernel/timer.c -1 +20 Sun Feb 1 14:34:49 EST 2009 athena@fftw.org * compilation fixes in case snprintf() is defined as a macro. M ./libbench2/report.c -8 +7 Wed Jan 28 20:19:04 EST 2009 athena@fftw.org * Automake does not like continuation lines beginning with a comment. M ./rdft/scalar/r2cb/Makefile.am -1 +1 Wed Jan 28 18:24:39 EST 2009 athena@fftw.org * Add r2cb_2.c r2cb_2.c is needed for problem rb2, which is not equivalent to rf2 (unlike kb2, which is equivalent to kf2). This change would not matter much except that rb2 is generated when reducing backward rdft2 to dft, and the absence of the codelet was preventing radix 2 from being employed at all in this case. M ./rdft/scalar/r2cb/Makefile.am -6 +4 Sat Jan 10 06:47:22 EST 2009 athena@fftw.org * handle the case vecsz->rnk == 0 correctly. M ./dft/buffered.c -1 +3 M ./rdft/buffered.c -1 +3 M ./rdft/buffered2.c -1 +3 Fri Dec 19 15:20:36 EST 2008 stevenj@fftw.org * Macs are no longer ppc-based; thanks to Charles Collicutt for the FAQ update M ./doc/FAQ/fftw-faq.bfnn -3 +3 Mon Dec 8 18:08:33 EST 2008 stevenj@fftw.org * use new multiple-nbuf code in rdft/buffered, like for dft/buffered M ./rdft/buffered.c -11 +25 M ./rdft/buffered2.c -12 +26 Sat Dec 6 16:34:36 EST 2008 stevenj@fftw.org * make x86_cpuid macro work on x86_64 M ./m4/ax_gcc_x86_cpuid.m4 -1 +15 Sat Dec 6 09:20:37 EST 2008 athena@fftw.org * Allow automatic choice of buffer size in dft/buffered.c Try a couple of different buffer sizes in buffered transforms, since this seems to make a difference on some Core2 models. M ./api/mapflags.c -2 M ./dft/buffered.c -10 +25 M ./kernel/buffered.c -3 +19 M ./kernel/ifftw.h -1 +5 M ./rdft/buffered.c -2 +3 M ./rdft/buffered2.c -2 +3 M ./rdft/rdft2-rdft.c -1 +1 Tue Dec 2 19:18:30 EST 2008 athena@fftw.org * libbench2: do not assume that split-complex arrays are stride-1 M ./libbench2/verify-dft.c -4 +3 M ./libbench2/verify.c -1 +3 M ./tests/bench.c -3 +2 Tue Dec 2 18:39:43 EST 2008 stevenj@fftw.org * updated NEWS M ./NEWS +14 Tue Dec 2 18:30:00 EST 2008 stevenj@fftw.org * date fix M ./m4/ax_gcc_archflag.m4 -1 +1 Tue Dec 2 18:29:06 EST 2008 stevenj@fftw.org * updated icc flags -- now prefer -xHost (-xN etc. seem t be obsolete), check for new spelling -ansi-alias, and use -malign-double like we do for gcc M ./m4/ax_cc_maxopt.m4 -2 +14 Tue Dec 2 18:28:03 EST 2008 stevenj@fftw.org * use $ax_cv_c_compiler_vendor rather than $GCC, as the former is more reliable (icc incorrectly self-identifies as gcc on MacOS where we don't use -no-gcc) M ./configure.ac -1 +1 Tue Dec 2 17:55:36 EST 2008 stevenj@fftw.org * don't use -no-gcc for icc on MacOS M ./configure.ac -2 +7 Tue Dec 2 17:34:04 EST 2008 stevenj@fftw.org * document some more bench options M ./tests/README -1 +17 Wed Nov 19 16:55:13 EST 2008 stevenj@fftw.org * make it clearer that --enable-openmp and --enable-threads are mutually exclusive; thanks to Long To for his comments M ./doc/fftw3.texi -6 +10 Mon Nov 17 20:16:28 EST 2008 stevenj@fftw.org * version bump to 3.2.1, use explicit Makefile.am for m4 subdirectory so that tarball does not include random files in there when you do 'make dist' M ./Makefile.am -3 +2 M ./configure.ac -2 +4 A ./m4/Makefile.am Sat Nov 15 21:12:58 EST 2008 stevenj@fftw.org * document behavior of FFTW guru arrays, and in particular the odd behavior of the plan_guru_r2r routine in Fortran (thanks to Alexander Pozdneev for the bug report) M ./doc/fftw3.texi +10 Sat Nov 15 21:03:14 EST 2008 stevenj@fftw.org tagged fftw-3.2 Mon Nov 10 20:21:32 EST 2008 stevenj@fftw.org * version bump to 3.2, updated copyright year M ./COPYRIGHT -2 +2 M ./NEWS -1 +6 M ./api/api.h -2 +2 M ./api/apiplan.c -2 +2 M ./api/configure.c -2 +2 M ./api/execute-dft-c2r.c -2 +2 M ./api/execute-dft-r2c.c -2 +2 M ./api/execute-dft.c -2 +2 M ./api/execute-r2r.c -2 +2 M ./api/execute-split-dft-c2r.c -2 +2 M ./api/execute-split-dft-r2c.c -2 +2 M ./api/execute-split-dft.c -2 +2 M ./api/execute.c -2 +2 M ./api/export-wisdom-to-file.c -2 +2 M ./api/export-wisdom-to-string.c -2 +2 M ./api/export-wisdom.c -2 +2 M ./api/f77api.c -2 +2 M ./api/f77funcs.h -2 +2 M ./api/fftw3.h -2 +2 M ./api/flops.c -2 +2 M ./api/forget-wisdom.c -2 +2 M ./api/import-system-wisdom.c -2 +2 M ./api/import-wisdom-from-file.c -2 +2 M ./api/import-wisdom-from-string.c -2 +2 M ./api/import-wisdom.c -2 +2 M ./api/malloc.c -2 +2 M ./api/map-r2r-kind.c -2 +2 M ./api/mapflags.c -2 +2 M ./api/mkprinter-file.c -2 +2 M ./api/mktensor-iodims.h -2 +2 M ./api/mktensor-rowmajor.c -2 +2 M ./api/plan-dft-1d.c -2 +2 M ./api/plan-dft-2d.c -2 +2 M ./api/plan-dft-3d.c -2 +2 M ./api/plan-dft-c2r-1d.c -2 +2 M ./api/plan-dft-c2r-2d.c -2 +2 M ./api/plan-dft-c2r-3d.c -2 +2 M ./api/plan-dft-c2r.c -2 +2 M ./api/plan-dft-r2c-1d.c -2 +2 M ./api/plan-dft-r2c-2d.c -2 +2 M ./api/plan-dft-r2c-3d.c -2 +2 M ./api/plan-dft-r2c.c -2 +2 M ./api/plan-dft.c -2 +2 M ./api/plan-guru-dft-c2r.h -2 +2 M ./api/plan-guru-dft-r2c.h -2 +2 M ./api/plan-guru-dft.h -2 +2 M ./api/plan-guru-r2r.h -2 +2 M ./api/plan-guru-split-dft-c2r.h -2 +2 M ./api/plan-guru-split-dft-r2c.h -2 +2 M ./api/plan-guru-split-dft.h -2 +2 M ./api/plan-many-dft-c2r.c -2 +2 M ./api/plan-many-dft-r2c.c -2 +2 M ./api/plan-many-dft.c -2 +2 M ./api/plan-many-r2r.c -2 +2 M ./api/plan-r2r-1d.c -2 +2 M ./api/plan-r2r-2d.c -2 +2 M ./api/plan-r2r-3d.c -2 +2 M ./api/plan-r2r.c -2 +2 M ./api/print-plan.c -2 +2 M ./api/rdft2-pad.c -2 +2 M ./api/the-planner.c -2 +2 M ./api/version.c -2 +2 M ./api/x77.h -2 +2 M ./configure.ac -2 +2 M ./dft/bluestein.c -2 +2 M ./dft/buffered.c -2 +2 M ./dft/codelet-dft.h -2 +2 M ./dft/conf.c -2 +2 M ./dft/ct.c -2 +2 M ./dft/ct.h -2 +2 M ./dft/dft.h -2 +2 M ./dft/dftw-direct.c -2 +2 M ./dft/dftw-directsq.c -2 +2 M ./dft/dftw-generic.c -2 +2 M ./dft/dftw-genericbuf.c -2 +2 M ./dft/direct.c -2 +2 M ./dft/generic.c -2 +2 M ./dft/indirect-transpose.c -2 +2 M ./dft/indirect.c -2 +2 M ./dft/kdft-dif.c -2 +2 M ./dft/kdft-difsq.c -2 +2 M ./dft/kdft-dit.c -2 +2 M ./dft/kdft.c -2 +2 M ./dft/nop.c -2 +2 M ./dft/plan.c -2 +2 M ./dft/problem.c -2 +2 M ./dft/rader.c -2 +2 M ./dft/rank-geq2.c -2 +2 M ./dft/scalar/n.c -2 +2 M ./dft/scalar/n.h -2 +2 M ./dft/scalar/t.c -2 +2 M ./dft/scalar/t.h -2 +2 M ./dft/simd/n1b.c -2 +2 M ./dft/simd/n1b.h -2 +2 M ./dft/simd/n1f.c -2 +2 M ./dft/simd/n1f.h -2 +2 M ./dft/simd/n2b.c -2 +2 M ./dft/simd/n2b.h -2 +2 M ./dft/simd/n2f.c -2 +2 M ./dft/simd/n2f.h -2 +2 M ./dft/simd/n2s.c -2 +2 M ./dft/simd/n2s.h -2 +2 M ./dft/simd/q1b.c -2 +2 M ./dft/simd/q1b.h -2 +2 M ./dft/simd/q1f.c -2 +2 M ./dft/simd/q1f.h -2 +2 M ./dft/simd/t.c -2 +2 M ./dft/simd/t1b.h -2 +2 M ./dft/simd/t1bu.h -2 +2 M ./dft/simd/t1f.h -2 +2 M ./dft/simd/t1fu.h -2 +2 M ./dft/simd/t2b.h -2 +2 M ./dft/simd/t2f.h -2 +2 M ./dft/simd/t3b.h -2 +2 M ./dft/simd/t3f.h -2 +2 M ./dft/simd/ts.c -2 +2 M ./dft/simd/ts.h -2 +2 M ./dft/solve.c -2 +2 M ./dft/vrank-geq1.c -2 +2 M ./dft/zero.c -2 +2 M ./doc/f77_wisdom.f -2 +2 M ./genfft/algsimp.ml -2 +2 M ./genfft/algsimp.mli -2 +2 M ./genfft/annotate.ml -2 +2 M ./genfft/annotate.mli -2 +2 M ./genfft/assoctable.ml -2 +2 M ./genfft/assoctable.mli -2 +2 M ./genfft/c.ml -2 +2 M ./genfft/c.mli -2 +2 M ./genfft/complex.ml -2 +2 M ./genfft/complex.mli -2 +2 M ./genfft/conv.ml -2 +2 M ./genfft/conv.mli -2 +2 M ./genfft/dag.ml -2 +2 M ./genfft/dag.mli -2 +2 M ./genfft/expr.ml -2 +2 M ./genfft/expr.mli -2 +2 M ./genfft/fft.ml -2 +2 M ./genfft/fft.mli -2 +2 M ./genfft/gen_hc2c.ml -2 +2 M ./genfft/gen_hc2cdft.ml -2 +2 M ./genfft/gen_hc2cdft_c.ml -2 +2 M ./genfft/gen_hc2hc.ml -2 +2 M ./genfft/gen_mdct.ml -2 +2 M ./genfft/gen_notw.ml -2 +2 M ./genfft/gen_notw_c.ml -2 +2 M ./genfft/gen_r2cb.ml -2 +2 M ./genfft/gen_r2cf.ml -2 +2 M ./genfft/gen_r2r.ml -2 +2 M ./genfft/gen_twiddle.ml -2 +2 M ./genfft/gen_twiddle_c.ml -2 +2 M ./genfft/gen_twidsq.ml -2 +2 M ./genfft/gen_twidsq_c.ml -2 +2 M ./genfft/genutil.ml -2 +2 M ./genfft/littlesimp.ml -2 +2 M ./genfft/littlesimp.mli -2 +2 M ./genfft/magic.ml -2 +2 M ./genfft/monads.ml -2 +2 M ./genfft/number.ml -2 +2 M ./genfft/number.mli -2 +2 M ./genfft/oracle.ml -2 +2 M ./genfft/oracle.mli -2 +2 M ./genfft/schedule.ml -2 +2 M ./genfft/schedule.mli -2 +2 M ./genfft/simd.ml -2 +2 M ./genfft/simd.mli -2 +2 M ./genfft/simdmagic.ml -2 +2 M ./genfft/to_alist.ml -2 +2 M ./genfft/to_alist.mli -2 +2 M ./genfft/trig.ml -2 +2 M ./genfft/trig.mli -2 +2 M ./genfft/twiddle.ml -2 +2 M ./genfft/twiddle.mli -2 +2 M ./genfft/unique.ml -2 +2 M ./genfft/unique.mli -2 +2 M ./genfft/util.ml -2 +2 M ./genfft/util.mli -2 +2 M ./genfft/variable.ml -2 +2 M ./genfft/variable.mli -2 +2 M ./kernel/align.c -2 +2 M ./kernel/alloc.c -2 +2 M ./kernel/assert.c -2 +2 M ./kernel/awake.c -2 +2 M ./kernel/buffered.c -2 +2 M ./kernel/cpy1d.c -2 +2 M ./kernel/cpy2d-pair.c -2 +2 M ./kernel/cpy2d.c -2 +2 M ./kernel/ct.c -2 +2 M ./kernel/cycle.h -2 +2 M ./kernel/debug.c -2 +2 M ./kernel/extract-reim.c -2 +2 M ./kernel/hash.c -2 +2 M ./kernel/iabs.c -2 +2 M ./kernel/ifftw.h -2 +2 M ./kernel/kalloc.c -2 +2 M ./kernel/md5-1.c -2 +2 M ./kernel/md5.c -2 +2 M ./kernel/minmax.c -2 +2 M ./kernel/ops.c -2 +2 M ./kernel/pickdim.c -2 +2 M ./kernel/plan.c -2 +2 M ./kernel/primes.c -2 +2 M ./kernel/print.c -2 +2 M ./kernel/problem.c -2 +2 M ./kernel/rader.c -2 +2 M ./kernel/scan.c -2 +2 M ./kernel/solver.c -2 +2 M ./kernel/solvtab.c -2 +2 M ./kernel/stride.c -2 +2 M ./kernel/tensor.c -2 +2 M ./kernel/tensor1.c -2 +2 M ./kernel/tensor2.c -2 +2 M ./kernel/tensor3.c -2 +2 M ./kernel/tensor4.c -2 +2 M ./kernel/tensor5.c -2 +2 M ./kernel/tensor7.c -2 +2 M ./kernel/tensor8.c -2 +2 M ./kernel/tensor9.c -2 +2 M ./kernel/tile2d.c -2 +2 M ./kernel/timer.c -2 +2 M ./kernel/transpose.c -2 +2 M ./kernel/trig.c -2 +2 M ./kernel/twiddle.c -2 +2 M ./libbench2/dotens2.c -2 +2 M ./libbench2/my-getopt.c -2 +2 M ./libbench2/my-getopt.h -2 +2 M ./libbench2/verify-dft.c -2 +2 M ./libbench2/verify-lib.c -2 +2 M ./libbench2/verify-r2r.c -2 +2 M ./libbench2/verify-rdft2.c -2 +2 M ./libbench2/verify.h -2 +2 M ./mpi/any-true.c -2 +2 M ./mpi/api.c -2 +2 M ./mpi/block.c -2 +2 M ./mpi/choose-radix.c -2 +2 M ./mpi/conf.c -2 +2 M ./mpi/dft-problem.c -2 +2 M ./mpi/dft-rank-geq2-transposed.c -2 +2 M ./mpi/dft-rank-geq2.c -2 +2 M ./mpi/dft-rank1-bigvec.c -2 +2 M ./mpi/dft-rank1.c -2 +2 M ./mpi/dft-serial.c -2 +2 M ./mpi/dft-solve.c -2 +2 M ./mpi/dtensor.c -2 +2 M ./mpi/fftw3-mpi.h -2 +2 M ./mpi/ifftw-mpi.h -2 +2 M ./mpi/mpi-dft.h -2 +2 M ./mpi/mpi-rdft.h -2 +2 M ./mpi/mpi-rdft2.h -2 +2 M ./mpi/mpi-transpose.h -2 +2 M ./mpi/rdft-problem.c -2 +2 M ./mpi/rdft-rank-geq2-transposed.c -2 +2 M ./mpi/rdft-rank-geq2.c -2 +2 M ./mpi/rdft-rank1-bigvec.c -2 +2 M ./mpi/rdft-serial.c -2 +2 M ./mpi/rdft-solve.c -2 +2 M ./mpi/rdft2-problem.c -2 +2 M ./mpi/rdft2-rank-geq2-transposed.c -2 +2 M ./mpi/rdft2-rank-geq2.c -2 +2 M ./mpi/rdft2-serial.c -2 +2 M ./mpi/rdft2-solve.c -2 +2 M ./mpi/rearrange.c -2 +2 M ./mpi/testsched.c -2 +2 M ./mpi/transpose-alltoall.c -2 +2 M ./mpi/transpose-pairwise.c -2 +2 M ./mpi/transpose-problem.c -2 +2 M ./mpi/transpose-recurse.c -2 +2 M ./mpi/transpose-solve.c -2 +2 M ./mpi/wisdom-api.c -2 +2 M ./rdft/buffered.c -2 +2 M ./rdft/buffered2.c -2 +2 M ./rdft/codelet-rdft.h -2 +2 M ./rdft/conf.c -2 +2 M ./rdft/ct-hc2c-direct.c -2 +2 M ./rdft/ct-hc2c.c -2 +2 M ./rdft/ct-hc2c.h -2 +2 M ./rdft/dft-r2hc.c -2 +2 M ./rdft/dht-r2hc.c -2 +2 M ./rdft/dht-rader.c -2 +2 M ./rdft/direct-r2c.c -2 +2 M ./rdft/direct-r2r.c -2 +2 M ./rdft/direct2.c -2 +2 M ./rdft/generic.c -2 +2 M ./rdft/hc2hc-direct.c -2 +2 M ./rdft/hc2hc-generic.c -2 +2 M ./rdft/hc2hc.c -2 +2 M ./rdft/hc2hc.h -2 +2 M ./rdft/indirect.c -2 +2 M ./rdft/khc2c.c -2 +2 M ./rdft/khc2hc.c -2 +2 M ./rdft/kr2c.c -2 +2 M ./rdft/kr2r.c -2 +2 M ./rdft/nop.c -2 +2 M ./rdft/nop2.c -2 +2 M ./rdft/plan.c -2 +2 M ./rdft/plan2.c -2 +2 M ./rdft/problem.c -2 +2 M ./rdft/problem2.c -2 +2 M ./rdft/rank-geq2-rdft2.c -2 +2 M ./rdft/rank-geq2.c -2 +2 M ./rdft/rank0-rdft2.c -2 +2 M ./rdft/rank0.c -2 +2 M ./rdft/rdft-dht.c -2 +2 M ./rdft/rdft.h -2 +2 M ./rdft/rdft2-inplace-strides.c -2 +2 M ./rdft/rdft2-rdft.c -2 +2 M ./rdft/rdft2-strides.c -2 +2 M ./rdft/rdft2-tensor-max-index.c -2 +2 M ./rdft/scalar/hb.h -2 +2 M ./rdft/scalar/hc2c.c -2 +2 M ./rdft/scalar/hc2cb.h -2 +2 M ./rdft/scalar/hc2cf.h -2 +2 M ./rdft/scalar/hf.h -2 +2 M ./rdft/scalar/hfb.c -2 +2 M ./rdft/scalar/r2c.c -2 +2 M ./rdft/scalar/r2cb.h -2 +2 M ./rdft/scalar/r2cbIII.h -2 +2 M ./rdft/scalar/r2cf.h -2 +2 M ./rdft/scalar/r2cfII.h -2 +2 M ./rdft/scalar/r2r.c -2 +2 M ./rdft/scalar/r2r.h -2 +2 M ./rdft/simd/hc2cbv.c -2 +2 M ./rdft/simd/hc2cbv.h -2 +2 M ./rdft/simd/hc2cfv.c -2 +2 M ./rdft/simd/hc2cfv.h -2 +2 M ./rdft/solve.c -2 +2 M ./rdft/solve2.c -2 +2 M ./rdft/vrank-geq1-rdft2.c -2 +2 M ./rdft/vrank-geq1.c -2 +2 M ./rdft/vrank3-transpose.c -2 +2 M ./reodft/conf.c -2 +2 M ./reodft/redft00e-r2hc-pad.c -2 +2 M ./reodft/redft00e-r2hc.c -2 +2 M ./reodft/reodft.h -2 +2 M ./reodft/reodft010e-r2hc.c -2 +2 M ./reodft/reodft11e-r2hc-odd.c -2 +2 M ./reodft/reodft11e-r2hc.c -2 +2 M ./reodft/reodft11e-radix2.c -2 +2 M ./reodft/rodft00e-r2hc-pad.c -2 +2 M ./reodft/rodft00e-r2hc.c -2 +2 M ./simd/altivec.c -2 +2 M ./simd/nonportable/sse.c -2 +2 M ./simd/nonportable/sse2.c -2 +2 M ./simd/simd-altivec.h -2 +2 M ./simd/simd-sse.h -2 +2 M ./simd/simd-sse2.h -2 +2 M ./simd/simd.h -2 +2 M ./simd/sse.c -2 +2 M ./simd/sse2.c -2 +2 M ./simd/taint.c -2 +2 M ./simd/x86-cpuid.h -2 +2 M ./threads/api.c -2 +2 M ./threads/conf.c -2 +2 M ./threads/ct.c -2 +2 M ./threads/dft-vrank-geq1.c -2 +2 M ./threads/f77api.c -2 +2 M ./threads/f77funcs.h -2 +2 M ./threads/hc2hc.c -2 +2 M ./threads/openmp.c -2 +2 M ./threads/rdft-vrank-geq1.c -2 +2 M ./threads/threads.c -2 +2 M ./threads/threads.h -2 +2 M ./threads/vrank-geq1-rdft2.c -2 +2 M ./tools/fftw-wisdom-to-conf.1 -4 +4 M ./tools/fftw-wisdom-to-conf.in -2 +2 M ./tools/fftw-wisdom.c -2 +2 M ./tools/fftw_wisdom.1.in -4 +4 Wed Nov 5 16:40:31 EST 2008 athena@fftw.org * Store GPLv2 in darcs because automake installs GPLv3 these days. A ./COPYING Thu Oct 30 15:03:41 EDT 2008 athena@fftw.org * stylistic changes, comments M ./threads/threads.c -13 +17 Thu Oct 30 14:40:14 EDT 2008 athena@fftw.org * oops M ./threads/threads.c -27 +28 Thu Oct 30 14:30:08 EDT 2008 athena@fftw.org * simplification of the threading machinery M ./threads/threads.c -12 +7 Thu Oct 30 14:22:40 EDT 2008 athena@fftw.org * typo M ./threads/threads.c -1 +1 Thu Oct 30 13:42:07 EDT 2008 athena@fftw.org * [SECOND ATTEMPT] do not assume that a semaphore can be freed just because nobody is using it Let S be a semaphore, initially 0. Let thread A execute UP(S); let thread B execute DOWN(S); free(&S); It is unclear whether this code is correct with posix semaphores. The problem is whether UP() uses S after allowing DOWN() to continue; this seems to be the case in the glibc-2.7 implementation, and thus the pattern above seems to be incorrect. Avoid using such a pattern, and introduce a global semaphore for the unavoidable case when nothing else can be depended upon. M ./threads/threads.c -61 +80 Wed Oct 29 20:09:39 EDT 2008 stevenj@fftw.org * updated cpu codes from x86-1.21 M ./m4/ax_gcc_archflag.m4 -5 +7 Wed Oct 29 16:24:16 EDT 2008 athena@fftw.org * Previous change was bogus, need to find another way. M ./threads/threads.c -56 +42 Wed Oct 29 12:22:20 EDT 2008 athena@fftw.org * do not assume that a semaphore can be freed just because nobody is using it Let S be a semaphore, initially 0. Let thread A execute UP(S); let thread B execute DOWN(S); free(&S); It is unclear whether this code is correct with posix semaphores. The problem is whether UP() uses S after allowing DOWN() to continue; this seems to be the case in the glibc-2.7 implementation, and thus the pattern above seems to be incorrect. Avoid using such a pattern, and introduce a global semaphore for the unavoidable case when nothing else can be depended upon. M ./threads/threads.c -42 +56 Mon Oct 27 23:38:02 EDT 2008 stevenj@fftw.org * don't need PROG_AS any more M ./configure.ac -1 R ./m4/amx_prog_as.m4 Sun Oct 26 23:41:11 EDT 2008 stevenj@fftw.org * use AC_CONFIG_MACRO_DIR macro M ./Makefile.am -1 +1 M ./configure.ac +1 Sun Oct 26 10:08:44 EDT 2008 athena@fftw.org * Remove mpi/Makefile from configure.ac Otherwise, the tarball breaks because mpi/ is not in the distribution. M ./configure.ac -2 Sat Oct 25 17:13:50 EDT 2008 stevenj@fftw.org * remove MPI from dist until FFTW 3.3 M ./Makefile.am -1 +1 M ./NEWS +3 M ./doc/fftw3.texi -948 +9 Sat Oct 25 17:12:35 EDT 2008 stevenj@fftw.org * use MPIRUN even for -np 1 M ./mpi/Makefile.am -1 +1 Fri Jul 18 17:17:08 EDT 2008 stevenj@fftw.org * use new gcc arch=native flag as fallback M ./m4/ax_gcc_archflag.m4 -5 +5 Sat Oct 25 13:36:40 EDT 2008 athena@fftw.org * Use sem_t to implement mutexes Use sem_t instead of pthread_mutex_t to implement mutexes. It seems like pthread mutexes hang on linux-2.6.22 after several days of tests; the hang does not occur on linux >= 2.6.24 or when we use sem_t instead of pthread_mutex_t. The situation is still quite mysterious but this code seems to work. M ./threads/threads.c -22 +34 Thu Oct 23 13:32:45 EDT 2008 athena@fftw.org * print informative message when pstring is NULL. M ./libbench2/verify.c -1 +2 Sun Oct 19 16:00:07 EDT 2008 athena@fftw.org * Fix incorrect alignment in dftw-generic. Multithreaded dftw-generic is supposed to process only a slice of the array, but we were planning with the alignment of the original array rather than the slice. This led to unaligned accesses in certain obscure situations. M ./NEWS +3 M ./configure.ac -1 +1 M ./dft/dftw-generic.c -1 +2 Mon Aug 18 17:27:26 EDT 2008 Matteo Frigo * Paranoia: do not create OS threads while holding locks. Glibc at least plays silly games such as keeping a global variable that records whether there is more than one thread in the process, and it does not perform atomic operations if the variable says that there is only one thread. Who knows how this interacts with creating threads while holding a lock. Some day some genius will come up with some ``optimization'' that breaks everything. M ./threads/threads.c -7 +13 Wed Aug 6 07:41:46 EDT 2008 athena@fftw.org * Welcome to the quadcore era M ./mkdist.sh -1 +1 Mon Jun 16 16:46:39 EDT 2008 stevenj@fftw.org * backslash is technically not allowed in "echo" arguments; thanks to Debian Bug#486046 for pointing out problem and solution (and Raphael Geissert and Vincent Zweije, in particular) M ./tools/fftw-wisdom-to-conf.in -1 +1 Sun May 4 12:15:24 EDT 2008 stevenj@fftw.org * note problem with test program in gcc 4.1.2-4.2; thanks to Raymond Rogers for reporting it M ./doc/FAQ/fftw-faq.bfnn +6 Fri May 2 19:21:30 EDT 2008 stevenj@fftw.org * output count of constants along with other statistics M ./genfft/c.ml -3 +8 Sat Apr 19 14:15:03 EDT 2008 athena@fftw.org * Lower priority of unaligned SIMD codelets. List t1[fb]uv_* codelets before the corresponding aligned codelets, since the estimator picks the latter ones in case of a tie and aligned codelets are preferable. In other words, this is a hack. M ./dft/simd/codelets/Makefile.am -3 +2 Sat Apr 19 08:55:46 EDT 2008 Matteo Frigo * There is no point in using higher radices for unaligned codelets. M ./dft/simd/codelets/Makefile.am -4 +2 Fri Apr 18 19:01:27 EDT 2008 stevenj@fftw.org * support generating loopless, strideless r2r codelets M ./genfft/gen_r2r.ml -7 +17 Fri Apr 18 19:00:25 EDT 2008 stevenj@fftw.org * added Magic.threemult to use 3+3 complex-multiply variant when possible M ./genfft/complex.ml -2 +13 M ./genfft/magic.ml +4 Thu Apr 10 19:53:31 EDT 2008 stevenj@fftw.org * fix documentation bug - export_wisdom_to_string returns a string that should be deallocated with free, not fftw_free (thanks to Stein Vidar Hagfors Haugan for the bug report) M ./doc/fftw3.texi -1 +1 Mon Jan 21 01:11:44 EST 2008 stevenj@fftw.org * bsd calls x86_64 "amd64"; thanks to Fernando Herrero Carron for the bug report M ./m4/ax_gcc_archflag.m4 -3 +3 Tue Jan 1 12:29:56 EST 2008 stevenj@fftw.org * fix typo in manual, thanks to Yinon Ehrlich M ./doc/fftw3.texi -1 +1 Mon Dec 3 13:57:13 EST 2007 stevenj@fftw.org * note problem with gcc 3.4.4 on x86_64, thanks to Uwe Hollerbach for the report M ./doc/FAQ/fftw-faq.bfnn -2 +4 Tue Nov 13 16:19:22 EST 2007 stevenj@fftw.org * bump shared-lib revision M ./configure.ac -1 +1 Tue Nov 13 16:16:49 EST 2007 stevenj@fftw.org * update NEWS for alpha3 M ./NEWS +23 Tue Oct 2 13:53:04 EDT 2007 stevenj@fftw.org * fixed URL M ./m4/ax_openmp.m4 -1 +1 Mon Sep 17 19:38:29 EDT 2007 stevenj@fftw.org * added missing prototype M ./tests/fftw-bench.c +4 Tue Aug 14 22:35:06 EDT 2007 stevenj@fftw.org * terminology tweak M ./doc/fftw3.texi -1 +1 Wed Aug 1 18:44:21 EDT 2007 stevenj@fftw.org * check for pathscale compilers (thanks to Julian Cummings) M ./m4/ax_compiler_vendor.m4 -2 +2 Sat Sep 15 18:02:32 EDT 2007 athena@fftw.org * Avoid possible conflict with Windows include files. M ./threads/threads.c +5 Tue Aug 7 21:26:05 EDT 2007 athena@fftw.org * Distribute codlist.c for SIMD codelets in the commercial tarball. M ./commercialize.sh -1 +1 Wed Aug 1 10:33:41 EDT 2007 stevenj@fftw.org * some documentation clarifications, and documented FFTW_WISDOM_ONLY, at the suggestion of Mario Emmenlauer and Phil Dumont M ./api/apiplan.c -3 +3 M ./doc/fftw3.texi -3 +24 Tue Jul 31 16:52:56 EDT 2007 stevenj@fftw.org * bug fix in test program for vrank-3 transpose plans with vl=1 M ./mpi/mpi-bench.c -1 +1 Sun Jul 29 17:02:46 EDT 2007 stevenj@fftw.org * only run mpi checks for --enable-mpi M ./mpi/Makefile.am +4 Sun Jul 29 16:45:30 EDT 2007 stevenj@fftw.org * check for NULL return from spe_context_create in case SPE_MAP_PS not supported M ./cell/cell.c -4 +9 Sun Jul 29 15:56:57 EDT 2007 stevenj@fftw.org * use problem-state pointer to write SPE mailbox with lower latency (makes a significant performance difference for N < 32k), thanks to Jan Wagner for suggestion M ./cell/cell.c +16 Sun Jul 29 14:22:08 EDT 2007 stevenj@fftw.org * port cell code to SDK2.1 (libspe2), since libspe1 API is deprecated and can't be used in code that also uses libspe2 API M ./cell/cell.c -3 +49 M ./cell/copy.c -1 +1 M ./cell/dft-direct-cell.c -1 +1 M ./cell/fftw-cell.h -4 +5 M ./cell/spu/main.spuc -4 +4 M ./cell/transpose.c -1 +1 M ./configure.ac -2 +2 Sun Jul 29 11:46:24 EDT 2007 stevenj@fftw.org * bug fix: ego->W allocated with cell_aligned_malloc, so deallocate with free, not X(ifree0) M ./cell/dft-direct-cell.c -1 +1 Mon Jul 2 15:57:12 EDT 2007 stevenj@fftw.org * removed obsolete reference to CVS id M ./api/fftw3.h -4 Mon May 21 14:25:39 EDT 2007 athena@fftw.org * cycle counter for sun compiler M ./kernel/cycle.h -1 +1 Wed May 9 19:49:11 EDT 2007 stevenj@fftw.org * use __inline instead of inline for AIX routines (__inline is supported by gcc and xlc, whereas apparently "inline" is only supported by xlc if you specify -qlanglvl=stdc99 or similar); thanks to Jeff Haferman for the bug report M ./kernel/cycle.h -2 +2 Mon Apr 30 15:37:56 EDT 2007 stevenj@fftw.org * fixed incorrect type prefix (fftw_ vs. X(...)) in mpi/wisdom-api.c; thanks to Eric A. Borisch for the bug report M ./mpi/wisdom-api.c -6 +6 Wed Apr 25 21:21:39 EDT 2007 stevenj@fftw.org * some cleanups in MPI make check M ./mpi/Makefile.am -7 +11 Wed Apr 25 21:19:27 EDT 2007 stevenj@fftw.org * re-enable heuristic in the common case where we are not compiling for Cell M ./dft/rank-geq2.c -1 +3 Tue Apr 24 17:42:43 EDT 2007 athena@fftw.org * Removed duplicate codelet names, was breaking linker. M ./dft/simd/codelets/Makefile.am -4 +4 Tue Apr 24 11:38:16 EDT 2007 stevenj@fftw.org * added more codelets of sizes 5/10/20/25 to improve speed for round decimal sizes (speed improvements of 10-20%, at cost of 10-30% in library size) M ./dft/scalar/codelets/Makefile.am -3 +6 M ./dft/simd/codelets/Makefile.am -14 +26 M ./rdft/scalar/r2cb/Makefile.am -8 +16 M ./rdft/scalar/r2cf/Makefile.am -8 +16 M ./rdft/simd/codelets/Makefile.am -2 +4 Sat Mar 24 18:40:47 EDT 2007 stevenj@fftw.org * for 1d prime sizes, punt and return serial plan M ./mpi/api.c +4 Sat Mar 24 18:24:55 EDT 2007 stevenj@fftw.org * output reminders of the problem during bench --verify M ./libbench2/verify.c -2 +5 Sat Mar 24 18:10:24 EDT 2007 stevenj@fftw.org * bug fix - missing solver->destroy initializer in rdft2-rdft M ./rdft/rdft2-rdft.c -1 +1 Fri Mar 23 11:12:19 EDT 2007 stevenj@fftw.org * -static, in --enable-debug, doesn't work on MacOS X (according to Daniel Oberhoff) M ./configure.ac -1 Wed Mar 21 22:23:06 EDT 2007 stevenj@fftw.org * fix MPI r2c/c2r to work with howmany > 1 M ./mpi/rdft2-rank-geq2-transposed.c -1 +2 M ./mpi/rdft2-rank-geq2.c -1 +2 M ./mpi/rdft2-serial.c -2 +3 M ./tests/check.pl -1 Wed Mar 21 18:44:41 EDT 2007 stevenj@fftw.org * rm MPI version from TODO M ./TODO -2 Wed Mar 21 18:34:40 EDT 2007 stevenj@fftw.org * added 'make bigcheck' for MPI (no paranoid-check, unfortunately), and properly get MPIRUN from configure M ./configure.ac +2 M ./mpi/Makefile.am -7 +10 Wed Mar 21 18:23:18 EDT 2007 stevenj@fftw.org * bug fix - incorrect local_size returned for 1d bigvec case M ./mpi/api.c -1 +1 Wed Mar 21 03:13:54 EDT 2007 stevenj@fftw.org * hack to specify MPI_TRANSPOSED_IN/OUT via "[" and "]" in libbench2 problem M ./libbench2/bench-user.h +1 M ./libbench2/problem.c +6 M ./mpi/mpi-bench.c +12 M ./tests/check.pl -2 +7 Wed Mar 21 02:58:11 EDT 2007 stevenj@fftw.org * added MPI 'make check', still needs a bit of work M ./mpi/Makefile.am +28 M ./tests/check.pl +31 Wed Mar 21 02:47:10 EDT 2007 stevenj@fftw.org * bug fix in r2r transposed-input case M ./mpi/rdft-problem.c +3 Wed Mar 21 02:46:25 EDT 2007 stevenj@fftw.org * don't output more than 300 erroneous outputs (unless verbose > 2) M ./libbench2/verify-lib.c -2 +3 M ./libbench2/verify-r2r.c -2 +3 Wed Mar 21 01:48:54 EDT 2007 stevenj@fftw.org * fixed bug in transposed-in c2r MPI transforms ... seems to be working, finally M ./mpi/rdft2-problem.c -5 +6 Wed Mar 21 00:41:32 EDT 2007 stevenj@fftw.org * some fixes to MPI r2c/c2r transforms with transposed output/input M ./mpi/rdft2-rank-geq2-transposed.c -11 +17 Wed Mar 21 00:40:25 EDT 2007 stevenj@fftw.org * typos M ./NEWS -4 +4 Tue Mar 20 19:53:02 EDT 2007 stevenj@fftw.org * bug fix for mpi-bench with r2c/c2r: allocate a little bit extra to make sure that padding is allocated M ./libbench2/allocate.c -2 +4 Tue Mar 20 19:19:13 EDT 2007 stevenj@fftw.org * fix typo, thanks to Ernest Turro for the bug report M ./m4/ax_gcc_archflag.m4 -2 +2 Tue Mar 20 01:39:06 EDT 2007 stevenj@fftw.org * spacing tweaks M ./doc/fftw3.texi -2 +2 Tue Mar 20 00:53:11 EDT 2007 stevenj@fftw.org * Ralf Wildenhues is the one who pointed out that the self-communication could fill in the stalls in the pairwise schedule M ./mpi/testsched.c -3 +2 Tue Mar 20 00:22:25 EDT 2007 stevenj@fftw.org * add TRANSPOSED_OUT/IN support for r2c/c2r, respectively M ./doc/fftw3.texi -4 +1 M ./mpi/Makefile.am -1 +1 M ./mpi/README -3 M ./mpi/api.c -6 +15 M ./mpi/conf.c -1 +1 M ./mpi/rdft2-problem.c -18 +9 A ./mpi/rdft2-rank-geq2-transposed.c Mon Mar 19 21:45:34 EDT 2007 stevenj@fftw.org * yikes! fixed likely deadlock bug in MPI M ./mpi/transpose-recurse.c -2 +2 Mon Mar 19 21:38:52 EDT 2007 stevenj@fftw.org * comment M ./mpi/transpose-pairwise.c -2 +8 Mon Mar 19 21:30:44 EDT 2007 stevenj@fftw.org * s/alpha1/alpha2/ M ./NEWS -1 +1 Mon Mar 19 00:39:47 EDT 2007 stevenj@fftw.org * include README in dist tarball M ./mpi/Makefile.am -1 +1 Mon Mar 19 00:35:43 EDT 2007 stevenj@fftw.org * added MPI r2c/c2r transforms, some more documentation M ./NEWS -4 +9 M ./doc/fftw3.texi -6 +73 M ./libbench2/bench-user.h +2 M ./libbench2/problem.c -2 +4 M ./mpi/Makefile.am -1 +2 A ./mpi/README M ./mpi/api.c +140 M ./mpi/conf.c +4 M ./mpi/fftw3-mpi.h -88 +116 M ./mpi/mpi-bench.c -2 +102 A ./mpi/mpi-rdft2.h A ./mpi/rdft2-problem.c A ./mpi/rdft2-rank-geq2.c A ./mpi/rdft2-serial.c A ./mpi/rdft2-solve.c Sun Mar 18 23:14:29 EDT 2007 stevenj@fftw.org * set version to 3.2alpha2 M ./configure.ac -1 +1 Sun Mar 18 19:12:18 EDT 2007 stevenj@fftw.org * changed --enable-mips_ps to --enable-mips-ps; added Cell section to manual (from README.Cell); many minor updates to manual M ./configure.ac -1 +1 M ./doc/fftw3.texi -15 +243 Sun Mar 18 15:27:06 EDT 2007 stevenj@fftw.org * whoops, need to sync costs in problem_mpi_rdft M ./mpi/api.c +3 Sun Mar 18 12:44:49 EDT 2007 stevenj@fftw.org * documented guru64 interface M ./NEWS -1 +1 M ./doc/fftw3.texi -11 +81 Sun Mar 18 02:57:46 EDT 2007 stevenj@fftw.org * typo M ./doc/fftw3.texi -1 +1 Sun Mar 18 02:45:09 EDT 2007 stevenj@fftw.org * bumped copyright year to 2007 M ./COPYRIGHT -2 +2 M ./api/api.h -2 +2 M ./api/apiplan.c -2 +2 M ./api/configure.c -2 +2 M ./api/execute-dft-c2r.c -2 +2 M ./api/execute-dft-r2c.c -2 +2 M ./api/execute-dft.c -2 +2 M ./api/execute-r2r.c -2 +2 M ./api/execute-split-dft-c2r.c -2 +2 M ./api/execute-split-dft-r2c.c -2 +2 M ./api/execute-split-dft.c -2 +2 M ./api/execute.c -2 +2 M ./api/export-wisdom-to-file.c -2 +2 M ./api/export-wisdom-to-string.c -2 +2 M ./api/export-wisdom.c -2 +2 M ./api/f77api.c -2 +2 M ./api/f77funcs.h -2 +2 M ./api/fftw3.h -2 +2 M ./api/flops.c -2 +2 M ./api/forget-wisdom.c -2 +2 M ./api/import-system-wisdom.c -2 +2 M ./api/import-wisdom-from-file.c -2 +2 M ./api/import-wisdom-from-string.c -2 +2 M ./api/import-wisdom.c -2 +2 M ./api/malloc.c -2 +2 M ./api/map-r2r-kind.c -2 +2 M ./api/mapflags.c -2 +2 M ./api/mkprinter-file.c -2 +2 M ./api/mktensor-iodims.h -2 +2 M ./api/mktensor-rowmajor.c -2 +2 M ./api/plan-dft-1d.c -2 +2 M ./api/plan-dft-2d.c -2 +2 M ./api/plan-dft-3d.c -2 +2 M ./api/plan-dft-c2r-1d.c -2 +2 M ./api/plan-dft-c2r-2d.c -2 +2 M ./api/plan-dft-c2r-3d.c -2 +2 M ./api/plan-dft-c2r.c -2 +2 M ./api/plan-dft-r2c-1d.c -2 +2 M ./api/plan-dft-r2c-2d.c -2 +2 M ./api/plan-dft-r2c-3d.c -2 +2 M ./api/plan-dft-r2c.c -2 +2 M ./api/plan-dft.c -2 +2 M ./api/plan-guru-dft-c2r.h -2 +2 M ./api/plan-guru-dft-r2c.h -2 +2 M ./api/plan-guru-dft.h -2 +2 M ./api/plan-guru-r2r.h -2 +2 M ./api/plan-guru-split-dft-c2r.h -2 +2 M ./api/plan-guru-split-dft-r2c.h -2 +2 M ./api/plan-guru-split-dft.h -2 +2 M ./api/plan-many-dft-c2r.c -2 +2 M ./api/plan-many-dft-r2c.c -2 +2 M ./api/plan-many-dft.c -2 +2 M ./api/plan-many-r2r.c -2 +2 M ./api/plan-r2r-1d.c -2 +2 M ./api/plan-r2r-2d.c -2 +2 M ./api/plan-r2r-3d.c -2 +2 M ./api/plan-r2r.c -2 +2 M ./api/print-plan.c -2 +2 M ./api/rdft2-pad.c -2 +2 M ./api/the-planner.c -2 +2 M ./api/version.c -2 +2 M ./api/x77.h -2 +2 M ./dft/bluestein.c -2 +2 M ./dft/buffered.c -2 +2 M ./dft/codelet-dft.h -2 +2 M ./dft/conf.c -2 +2 M ./dft/ct.c -2 +2 M ./dft/ct.h -2 +2 M ./dft/dft.h -2 +2 M ./dft/dftw-direct.c -2 +2 M ./dft/dftw-directsq.c -2 +2 M ./dft/dftw-generic.c -2 +2 M ./dft/dftw-genericbuf.c -2 +2 M ./dft/direct.c -2 +2 M ./dft/generic.c -2 +2 M ./dft/indirect-transpose.c -2 +2 M ./dft/indirect.c -2 +2 M ./dft/kdft-dif.c -2 +2 M ./dft/kdft-difsq.c -2 +2 M ./dft/kdft-dit.c -2 +2 M ./dft/kdft.c -2 +2 M ./dft/nop.c -2 +2 M ./dft/plan.c -2 +2 M ./dft/problem.c -2 +2 M ./dft/rader.c -2 +2 M ./dft/rank-geq2.c -2 +2 M ./dft/scalar/n.c -2 +2 M ./dft/scalar/n.h -2 +2 M ./dft/scalar/t.c -2 +2 M ./dft/scalar/t.h -2 +2 M ./dft/simd/n1b.c -2 +2 M ./dft/simd/n1b.h -2 +2 M ./dft/simd/n1f.c -2 +2 M ./dft/simd/n1f.h -2 +2 M ./dft/simd/n2b.c -2 +2 M ./dft/simd/n2b.h -2 +2 M ./dft/simd/n2f.c -2 +2 M ./dft/simd/n2f.h -2 +2 M ./dft/simd/n2s.c -2 +2 M ./dft/simd/n2s.h -2 +2 M ./dft/simd/q1b.c -2 +2 M ./dft/simd/q1b.h -2 +2 M ./dft/simd/q1f.c -2 +2 M ./dft/simd/q1f.h -2 +2 M ./dft/simd/t.c -2 +2 M ./dft/simd/t1b.h -2 +2 M ./dft/simd/t1bu.h -2 +2 M ./dft/simd/t1f.h -2 +2 M ./dft/simd/t1fu.h -2 +2 M ./dft/simd/t2b.h -2 +2 M ./dft/simd/t2f.h -2 +2 M ./dft/simd/t3b.h -2 +2 M ./dft/simd/t3f.h -2 +2 M ./dft/simd/ts.c -2 +2 M ./dft/simd/ts.h -2 +2 M ./dft/solve.c -2 +2 M ./dft/vrank-geq1.c -2 +2 M ./dft/zero.c -2 +2 M ./doc/f77_wisdom.f -2 +2 M ./genfft/algsimp.ml -2 +2 M ./genfft/algsimp.mli -2 +2 M ./genfft/annotate.ml -2 +2 M ./genfft/annotate.mli -2 +2 M ./genfft/assoctable.ml -2 +2 M ./genfft/assoctable.mli -2 +2 M ./genfft/c.ml -2 +2 M ./genfft/c.mli -2 +2 M ./genfft/complex.ml -2 +2 M ./genfft/complex.mli -2 +2 M ./genfft/conv.ml -2 +2 M ./genfft/conv.mli -2 +2 M ./genfft/dag.ml -2 +2 M ./genfft/dag.mli -2 +2 M ./genfft/expr.ml -2 +2 M ./genfft/expr.mli -2 +2 M ./genfft/fft.ml -2 +2 M ./genfft/fft.mli -2 +2 M ./genfft/gen_hc2c.ml -2 +2 M ./genfft/gen_hc2cdft.ml -2 +2 M ./genfft/gen_hc2cdft_c.ml -2 +2 M ./genfft/gen_hc2hc.ml -2 +2 M ./genfft/gen_mdct.ml -2 +2 M ./genfft/gen_notw.ml -2 +2 M ./genfft/gen_notw_c.ml -2 +2 M ./genfft/gen_r2cb.ml -2 +2 M ./genfft/gen_r2cf.ml -2 +2 M ./genfft/gen_r2r.ml -2 +2 M ./genfft/gen_twiddle.ml -2 +2 M ./genfft/gen_twiddle_c.ml -2 +2 M ./genfft/gen_twidsq.ml -2 +2 M ./genfft/gen_twidsq_c.ml -2 +2 M ./genfft/genutil.ml -2 +2 M ./genfft/littlesimp.ml -2 +2 M ./genfft/littlesimp.mli -2 +2 M ./genfft/magic.ml -2 +2 M ./genfft/monads.ml -2 +2 M ./genfft/number.ml -2 +2 M ./genfft/number.mli -2 +2 M ./genfft/oracle.ml -2 +2 M ./genfft/oracle.mli -2 +2 M ./genfft/schedule.ml -2 +2 M ./genfft/schedule.mli -2 +2 M ./genfft/simd.ml -2 +2 M ./genfft/simd.mli -2 +2 M ./genfft/simdmagic.ml -2 +2 M ./genfft/to_alist.ml -2 +2 M ./genfft/to_alist.mli -2 +2 M ./genfft/trig.ml -2 +2 M ./genfft/trig.mli -2 +2 M ./genfft/twiddle.ml -2 +2 M ./genfft/twiddle.mli -2 +2 M ./genfft/unique.ml -2 +2 M ./genfft/unique.mli -2 +2 M ./genfft/util.ml -2 +2 M ./genfft/util.mli -2 +2 M ./genfft/variable.ml -2 +2 M ./genfft/variable.mli -2 +2 M ./kernel/align.c -2 +2 M ./kernel/alloc.c -2 +2 M ./kernel/assert.c -2 +2 M ./kernel/awake.c -2 +2 M ./kernel/buffered.c -2 +2 M ./kernel/cpy1d.c -2 +2 M ./kernel/cpy2d-pair.c -2 +2 M ./kernel/cpy2d.c -2 +2 M ./kernel/ct.c -2 +2 M ./kernel/cycle.h -2 +2 M ./kernel/debug.c -2 +2 M ./kernel/extract-reim.c -2 +2 M ./kernel/hash.c -2 +2 M ./kernel/iabs.c -2 +2 M ./kernel/ifftw.h -2 +2 M ./kernel/kalloc.c -2 +2 M ./kernel/md5-1.c -2 +2 M ./kernel/md5.c -2 +2 M ./kernel/minmax.c -2 +2 M ./kernel/ops.c -2 +2 M ./kernel/pickdim.c -2 +2 M ./kernel/plan.c -2 +2 M ./kernel/primes.c -2 +2 M ./kernel/print.c -2 +2 M ./kernel/problem.c -2 +2 M ./kernel/rader.c -2 +2 M ./kernel/scan.c -2 +2 M ./kernel/solver.c -2 +2 M ./kernel/solvtab.c -2 +2 M ./kernel/stride.c -2 +2 M ./kernel/tensor.c -2 +2 M ./kernel/tensor1.c -2 +2 M ./kernel/tensor2.c -2 +2 M ./kernel/tensor3.c -2 +2 M ./kernel/tensor4.c -2 +2 M ./kernel/tensor5.c -2 +2 M ./kernel/tensor7.c -2 +2 M ./kernel/tensor8.c -2 +2 M ./kernel/tensor9.c -2 +2 M ./kernel/tile2d.c -2 +2 M ./kernel/timer.c -2 +2 M ./kernel/transpose.c -2 +2 M ./kernel/trig.c -2 +2 M ./kernel/twiddle.c -2 +2 M ./libbench2/dotens2.c -2 +2 M ./libbench2/my-getopt.c -2 +2 M ./libbench2/my-getopt.h -2 +2 M ./libbench2/verify-dft.c -2 +2 M ./libbench2/verify-lib.c -2 +2 M ./libbench2/verify-r2r.c -2 +2 M ./libbench2/verify-rdft2.c -2 +2 M ./libbench2/verify.h -2 +2 M ./mpi/any-true.c -2 +2 M ./mpi/api.c -2 +2 M ./mpi/block.c -2 +2 M ./mpi/choose-radix.c -2 +2 M ./mpi/conf.c -2 +2 M ./mpi/dft-problem.c -2 +2 M ./mpi/dft-rank-geq2-transposed.c -2 +2 M ./mpi/dft-rank-geq2.c -2 +2 M ./mpi/dft-rank1-bigvec.c -2 +2 M ./mpi/dft-rank1.c -2 +2 M ./mpi/dft-serial.c -2 +2 M ./mpi/dft-solve.c -2 +2 M ./mpi/dtensor.c -2 +2 M ./mpi/fftw3-mpi.h -2 +2 M ./mpi/ifftw-mpi.h -2 +2 M ./mpi/mpi-dft.h -2 +2 M ./mpi/mpi-rdft.h -2 +2 M ./mpi/mpi-transpose.h -2 +2 M ./mpi/rdft-problem.c -2 +2 M ./mpi/rdft-rank-geq2-transposed.c -2 +2 M ./mpi/rdft-rank-geq2.c -2 +2 M ./mpi/rdft-rank1-bigvec.c -2 +2 M ./mpi/rdft-serial.c -2 +2 M ./mpi/rdft-solve.c -2 +2 M ./mpi/rearrange.c -2 +2 M ./mpi/testsched.c -2 +2 M ./mpi/transpose-alltoall.c -2 +2 M ./mpi/transpose-pairwise.c -2 +2 M ./mpi/transpose-problem.c -2 +2 M ./mpi/transpose-recurse.c -2 +2 M ./mpi/transpose-solve.c -2 +2 M ./mpi/wisdom-api.c -2 +2 M ./rdft/buffered.c -2 +2 M ./rdft/buffered2.c -2 +2 M ./rdft/codelet-rdft.h -2 +2 M ./rdft/conf.c -2 +2 M ./rdft/ct-hc2c-direct.c -2 +2 M ./rdft/ct-hc2c.c -2 +2 M ./rdft/ct-hc2c.h -2 +2 M ./rdft/dft-r2hc.c -2 +2 M ./rdft/dht-r2hc.c -2 +2 M ./rdft/dht-rader.c -2 +2 M ./rdft/direct-r2c.c -2 +2 M ./rdft/direct-r2r.c -2 +2 M ./rdft/direct2.c -2 +2 M ./rdft/generic.c -2 +2 M ./rdft/hc2hc-direct.c -2 +2 M ./rdft/hc2hc-generic.c -2 +2 M ./rdft/hc2hc.c -2 +2 M ./rdft/hc2hc.h -2 +2 M ./rdft/indirect.c -2 +2 M ./rdft/khc2c.c -2 +2 M ./rdft/khc2hc.c -2 +2 M ./rdft/kr2c.c -2 +2 M ./rdft/kr2r.c -2 +2 M ./rdft/nop.c -2 +2 M ./rdft/nop2.c -2 +2 M ./rdft/plan.c -2 +2 M ./rdft/plan2.c -2 +2 M ./rdft/problem.c -2 +2 M ./rdft/problem2.c -2 +2 M ./rdft/rank-geq2-rdft2.c -2 +2 M ./rdft/rank-geq2.c -2 +2 M ./rdft/rank0-rdft2.c -2 +2 M ./rdft/rank0.c -2 +2 M ./rdft/rdft-dht.c -2 +2 M ./rdft/rdft.h -2 +2 M ./rdft/rdft2-inplace-strides.c -2 +2 M ./rdft/rdft2-rdft.c -2 +2 M ./rdft/rdft2-strides.c -2 +2 M ./rdft/rdft2-tensor-max-index.c -2 +2 M ./rdft/scalar/hb.h -2 +2 M ./rdft/scalar/hc2c.c -2 +2 M ./rdft/scalar/hc2cb.h -2 +2 M ./rdft/scalar/hc2cf.h -2 +2 M ./rdft/scalar/hf.h -2 +2 M ./rdft/scalar/hfb.c -2 +2 M ./rdft/scalar/r2c.c -2 +2 M ./rdft/scalar/r2cb.h -2 +2 M ./rdft/scalar/r2cbIII.h -2 +2 M ./rdft/scalar/r2cf.h -2 +2 M ./rdft/scalar/r2cfII.h -2 +2 M ./rdft/scalar/r2r.c -2 +2 M ./rdft/scalar/r2r.h -2 +2 M ./rdft/simd/hc2cbv.c -2 +2 M ./rdft/simd/hc2cbv.h -2 +2 M ./rdft/simd/hc2cfv.c -2 +2 M ./rdft/simd/hc2cfv.h -2 +2 M ./rdft/solve.c -2 +2 M ./rdft/solve2.c -2 +2 M ./rdft/vrank-geq1-rdft2.c -2 +2 M ./rdft/vrank-geq1.c -2 +2 M ./rdft/vrank3-transpose.c -2 +2 M ./reodft/conf.c -2 +2 M ./reodft/redft00e-r2hc-pad.c -2 +2 M ./reodft/redft00e-r2hc.c -2 +2 M ./reodft/reodft.h -2 +2 M ./reodft/reodft010e-r2hc.c -2 +2 M ./reodft/reodft11e-r2hc-odd.c -2 +2 M ./reodft/reodft11e-r2hc.c -2 +2 M ./reodft/reodft11e-radix2.c -2 +2 M ./reodft/rodft00e-r2hc-pad.c -2 +2 M ./reodft/rodft00e-r2hc.c -2 +2 M ./simd/altivec.c -2 +2 M ./simd/nonportable/sse.c -2 +2 M ./simd/nonportable/sse2.c -2 +2 M ./simd/simd-altivec.h -2 +2 M ./simd/simd-sse.h -2 +2 M ./simd/simd-sse2.h -2 +2 M ./simd/simd.h -2 +2 M ./simd/sse.c -2 +2 M ./simd/sse2.c -2 +2 M ./simd/taint.c -2 +2 M ./simd/x86-cpuid.h -2 +2 M ./threads/api.c -2 +2 M ./threads/conf.c -2 +2 M ./threads/ct.c -2 +2 M ./threads/dft-vrank-geq1.c -2 +2 M ./threads/f77api.c -2 +2 M ./threads/f77funcs.h -2 +2 M ./threads/hc2hc.c -2 +2 M ./threads/openmp.c -2 +2 M ./threads/rdft-vrank-geq1.c -2 +2 M ./threads/threads.c -2 +2 M ./threads/threads.h -2 +2 M ./threads/vrank-geq1-rdft2.c -2 +2 M ./tools/fftw-wisdom-to-conf.1 -4 +4 M ./tools/fftw-wisdom-to-conf.in -2 +2 M ./tools/fftw-wisdom.c -2 +2 M ./tools/fftw_wisdom.1.in -4 +4 Sun Mar 18 01:41:40 EDT 2007 stevenj@fftw.org * noted CodeSourcery in AUTHORS M ./AUTHORS -1 +2 Sun Mar 18 01:25:00 EDT 2007 stevenj@fftw.org * more MPI documentation M ./doc/fftw3.texi -2 +61 Sat Mar 17 23:15:04 EDT 2007 stevenj@fftw.org * added MPI multi-dimensional rdft solvers & tests M ./mpi/Makefile.am -2 +4 M ./mpi/api.c +104 M ./mpi/conf.c +5 M ./mpi/fftw3-mpi.h -73 +90 M ./mpi/mpi-bench.c -1 +94 A ./mpi/mpi-rdft.h A ./mpi/rdft-problem.c A ./mpi/rdft-rank-geq2-transposed.c A ./mpi/rdft-rank-geq2.c A ./mpi/rdft-rank1-bigvec.c A ./mpi/rdft-serial.c A ./mpi/rdft-solve.c Sat Mar 17 22:52:00 EDT 2007 stevenj@fftw.org * whoops M ./mpi/dft-rank1-bigvec.c -1 +1 Sat Mar 17 22:43:54 EDT 2007 stevenj@fftw.org * clarification - fftw_mpi_init should be called before importing wisdom M ./doc/fftw3.texi -2 +4 Sat Mar 17 19:49:37 EDT 2007 stevenj@fftw.org * kindx/y/z -> kind0/1/2 for consistency M ./api/fftw3.h -3 +3 M ./doc/fftw3.texi -9 +9 Sat Mar 17 19:34:02 EDT 2007 stevenj@fftw.org * typo M ./mpi/dft-rank-geq2-transposed.c -1 +1 Sat Mar 17 19:14:16 EDT 2007 stevenj@fftw.org * some refactoring in preparation for mpi-rdft M ./mpi/Makefile.am -1 +1 M ./mpi/api.c -1 +1 M ./mpi/dft-problem.c -2 +2 M ./mpi/dft-rank1-bigvec.c -53 +5 M ./mpi/dtensor.c -3 +3 M ./mpi/ifftw-mpi.h -1 +29 A ./mpi/rearrange.c Sat Mar 17 18:12:45 EDT 2007 stevenj@fftw.org * documented more stuff for MPI M ./doc/fftw3.texi -2 +112 Sat Mar 17 15:41:23 EDT 2007 stevenj@fftw.org * added NEWS for 3.2alpha M ./NEWS -1 +37 Sat Mar 17 14:50:22 EDT 2007 stevenj@fftw.org * documented MPI transpose routines M ./doc/fftw3.texi -1 +130 Sat Mar 17 08:57:30 EDT 2007 athena@fftw.org * Removed unused variables M ./rdft/direct-r2c.c -3 Fri Mar 16 14:47:10 EDT 2007 athena@fftw.org * Preparing for interim release of Cell code. M ./AUTHORS +2 M ./configure.ac -1 +1 M ./doc/fftw3.texi +3 Thu Feb 8 12:23:43 EST 2007 athena@fftw.org * Added README.Cell A ./README.Cell M ./Makefile.am -1 +2 M ./README.Cell +103 Sat Mar 10 19:17:40 EST 2007 athena@fftw.org * Synchronized with main branch M ./simd/simd.h +10 Mon Jan 22 17:43:56 EST 2007 athena@fftw.org * Adapted vrecur heuristic to Cell. M ./cell/dft-direct-cell.c -1 +28 M ./dft/dft.h +4 Sun Jan 21 19:09:33 EST 2007 athena@fftw.org * synchronize with main branch. Sun Jan 21 14:42:00 EST 2007 athena@fftw.org * synchronized with main branch Thu Jan 18 20:29:22 EST 2007 athena@fftw.org * Increased MAX_N to 32K/sizeof(R). M ./cell/dft-direct-cell.c -23 +35 M ./cell/fftw-cell.h -3 +3 M ./cell/plans-double.c -9 +1033 M ./cell/plans-single.c -13 +525 M ./cell/spu/dft.spuc -2 +8 Thu Jan 18 13:43:51 EST 2007 Matteo Frigo * Added pointer to solver->destroy which is used in the Cell branch. M ./mpi/dft-rank-geq2-transposed.c -1 +1 M ./mpi/dft-rank1-bigvec.c -1 +1 M ./mpi/dft-rank1.c -1 +1 M ./mpi/transpose-recurse.c -1 +1 Thu Jan 18 12:09:26 EST 2007 athena@fftw.org * Updated copyright notices R ./mpi/transpose-radix2.c M ./cell/cell.c -2 +1 M ./cell/conf.c -2 +1 M ./cell/copy.c -2 +1 M ./cell/dft-direct-cell.c -2 +1 M ./cell/fftw-cell.h +19 M ./cell/spu/alloc.spuc +19 M ./cell/spu/copy.spuc +19 M ./cell/spu/dft.spuc +18 M ./cell/spu/dma.spuc +19 M ./cell/spu/execute.spuc +19 M ./cell/spu/fftw-spu.h +19 M ./cell/spu/main.spuc +19 M ./cell/spu/planner.spuc +19 M ./cell/spu/spu-double.h +19 M ./cell/spu/spu-single.h +19 M ./cell/spu/transpose.spuc +19 M ./cell/spufftw-embed.S +19 M ./cell/transpose.c -2 +1 Fri Jan 12 12:54:43 EST 2007 athena@fftw.org * Use mfc_read_tag_status_all() instead of spu_mfcstat(2), since the former seems to be standardized. M ./cell/spu/dma.spuc -1 +1 Thu Jan 11 14:55:08 EST 2007 athena@fftw.org * Silence some int/INT warnings. M ./cell/dft-direct-cell.c -21 +24 Wed Jan 10 18:19:53 EST 2007 athena@fftw.org * Note incompatibility of --enable-cell with --enable-threads M ./configure.ac +3 Wed Jan 10 17:57:10 EST 2007 athena@fftw.org * forgot to add file A ./cell/spufftw-embed.S Wed Jan 10 17:45:16 EST 2007 athena@fftw.org * 64-bit cleanup M ./cell/Makefile.am -4 +3 M ./cell/cell.c -7 +21 M ./cell/copy.c -2 +2 M ./cell/dft-direct-cell.c -11 +28 M ./cell/fftw-cell.h -8 +10 M ./cell/spu/dft.spuc -3 +3 M ./cell/spu/dma.spuc -7 +9 M ./cell/spu/fftw-spu.h -2 +2 M ./cell/transpose.c -1 +1 M ./configure.ac +3 M ./m4/ax_gcc_archflag.m4 -1 +1 Wed Jan 10 13:47:20 EST 2007 athena@fftw.org * Use -mcpu=cell where appropriate. M ./m4/ax_gcc_archflag.m4 +1 Tue Dec 26 21:35:59 EST 2006 athena@fftw.org * synchronized with main M ./dft/buffered.c -14 +1 Sun Dec 24 20:58:25 EST 2006 athena@fftw.org * synchronized with main branch M ./Makefile.am -12 +14 Sun Dec 24 13:47:37 EST 2006 athena@fftw.org * synchronized with main branch, updated to new sdk. M! ./cell/spu/Makefile.am -4 +3 M! ./cell/spu/dma.spuc -1 M! ./cell/spu/execute.spuc -2 +2 M! ./cell/spu/fftw-spu.h -29 +29 Thu Dec 21 17:17:41 EST 2006 athena@fftw.org * removed obsolete file R ./rdft/rdft2-radix2.c Tue Dec 19 15:17:20 EST 2006 athena@fftw.org * synchronized with main branch R ./rdft/direct.c M ./rdft/direct-r2c.c -1 +1 M ./rdft/direct-r2r.c -1 +1 Tue Dec 19 11:27:38 EST 2006 athena@fftw.org * Synchronized with main branch M! ./rdft/ct-hc2c.c -1 +1 M! ./rdft/direct2.c -13 +1 Fri Dec 15 16:04:31 EST 2006 athena@fftw.org * resolved conflict with main branch M ./tests/fftw-bench.c -3 +2 Fri Dec 8 14:43:50 EST 2006 athena@fftw.org * Fixes for compilation in subdirectories M ./cell/spu/Makefile.am -5 +3 Fri Dec 8 12:46:00 EST 2006 athena@fftw.org * Silence warning M ./rdft/rank0.c -2 Fri Dec 8 12:24:19 EST 2006 athena@fftw.org * silence warning M ./rdft/rank0.c -1 +1 Thu Dec 7 15:18:17 EST 2006 athena@fftw.org * Commented a particularly obscure piece of code. M ./cell/spu/dma.spuc -1 +14 Thu Dec 7 11:53:29 EST 2006 athena@fftw.org * Reorganized, clarified conditions for applicability of the DFT solver. M ./cell/dft-direct-cell.c -193 +178 Mon Dec 4 21:33:49 EST 2006 athena@fftw.org * Minor changes M ./cell/dft-direct-cell.c -1 +6 M ./cell/spu/dft.spuc -1 +7 Mon Dec 4 17:43:28 EST 2006 athena@fftw.org * Clarified comment M ./cell/dft-direct-cell.c -2 +2 Mon Dec 4 16:49:06 EST 2006 athena@fftw.org * Less incorrect conditions for fitting into local store. M ./cell/dft-direct-cell.c -5 +12 Mon Dec 4 16:08:24 EST 2006 athena@fftw.org * Implemented DECDIF+TRANSPOSE on Cell R ./dft/ctsq.c M ./cell/dft-direct-cell.c -93 +142 M ./cell/fftw-cell.h -8 +3 M ./cell/spu/dft.spuc -12 +21 M ./dft/indirect-transpose.c -1 +2 Fri Dec 1 17:42:55 EST 2006 athena@fftw.org * relaxed conditions of applicability of SPE M ./cell/dft-direct-cell.c -31 +42 Fri Dec 1 16:28:10 EST 2006 athena@fftw.org * tweaks M ./cell/dft-direct-cell.c -3 +2 Fri Dec 1 14:35:17 EST 2006 athena@fftw.org * Implemented Cell opcounts M ./cell/dft-direct-cell.c -4 +55 M ./cell/spu/execute.spuc +1 Fri Dec 1 13:38:44 EST 2006 athena@fftw.org * minor cleanup M ./cell/dft-direct-cell.c -9 +4 Fri Dec 1 11:16:52 EST 2006 athena@fftw.org * use [c0 s0 c1 s1] format for Cell twiddle factors, rather than [c0 c1 s0 s1]. This makes life easier and there is no speed penalty on Cell (unlike Altivec). M ./cell/copy.c -6 +6 M ./cell/dft-direct-cell.c -46 +15 M ./cell/spu/planner.spuc -2 +3 M ./cell/spu/spu-single.h -4 +4 Wed Nov 29 18:02:54 EST 2006 athena@fftw.org * Implemented SPE-accelerated copies A ./cell/copy.c M ./cell/Makefile.am -2 +2 M ./cell/copy.c +81 M ./cell/fftw-cell.h +14 M ./cell/spu/copy.spuc -69 +31 M ./cell/spu/dma.spuc -2 +78 M ./cell/spu/fftw-spu.h -4 +6 M ./cell/spu/main.spuc +4 M ./cell/spu/transpose.spuc -14 +6 M ./cell/transpose.c -4 +4 M ./kernel/ifftw.h -1 +3 M ./rdft/rank0.c +52 Wed Nov 29 12:11:08 EST 2006 athena@fftw.org * allow SPEs to compute vrank-0 problems. M ./cell/dft-direct-cell.c -10 +8 M ./cell/spu/dma.spuc -4 +4 Tue Nov 28 18:03:07 EST 2006 athena@fftw.org * eliminated DMA lists M ./cell/spu/dft.spuc -4 +4 M ./cell/spu/dma.spuc -119 +62 M ./cell/spu/fftw-spu.h -3 M ./cell/spu/main.spuc -2 +2 M ./cell/spu/transpose.spuc -6 +6 Tue Nov 28 14:22:05 EST 2006 athena@fftw.org * Conservatively force all dimensions to be 0 (mod VL) in cell, since otherwise it is too hard to get all cases right. M ./cell/dft-direct-cell.c -11 +10 M ./cell/spu/dft.spuc -1 +2 Tue Nov 28 12:39:01 EST 2006 athena@fftw.org * Check alignment of strides when transposing on Cell. M ./rdft/rank0.c -1 +10 Tue Nov 28 12:19:09 EST 2006 athena@fftw.org * consistent usage of FFT_SIGN M ./cell/dft-direct-cell.c -3 +3 M ./cell/spu/dft.spuc -2 +2 Tue Nov 28 11:35:38 EST 2006 athena@fftw.org * clever transposition algorithm without buffering M ./cell/spu/copy.spuc -178 +59 M ./cell/spu/dft.spuc -52 +8 M ./cell/spu/dma.spuc -54 +97 M ./cell/spu/fftw-spu.h -13 +3 M ./cell/spu/transpose.spuc -13 +11 Mon Nov 27 14:08:28 EST 2006 athena@fftw.org * Fixed tracking of dependencies M ./cell/spu/Makefile.am -1 +3 Mon Nov 27 14:03:53 EST 2006 athena@fftw.org * implemented 1D transforms, various tweaks M ./cell/cell.c -4 +14 M ./cell/conf.c +1 M ./cell/dft-direct-cell.c -15 +196 M ./cell/fftw-cell.h -2 +13 M ./cell/spu/Makefile.am +1 M ./cell/spu/alloc.spuc +1 M ./cell/spu/copy.spuc -34 +151 M ./cell/spu/dft.spuc -2 +37 M ./cell/spu/dma.spuc -13 +9 M ./cell/spu/fftw-spu.h -5 +9 M ./cell/spu/main.spuc -6 +3 Wed Nov 22 15:43:36 EST 2006 athena@fftw.org * no need to poll mailbox on spu side M ./cell/spu/main.spuc -1 Wed Nov 22 14:08:24 EST 2006 athena@fftw.org * increased maximum size handled by spe M ./cell/fftw-cell.h -2 +2 M ./cell/plans-double.c -2 +1027 M ./cell/plans-single.c -2 +514 M ./cell/spu/alloc.spuc -1 +1 M ./cell/spu/planner.spuc -1 +1 Tue Nov 21 16:23:17 EST 2006 athena@fftw.org * allow vrank<=2 problems in SPEs to avoid the vecloop overhead (grrr...) M ./cell/dft-direct-cell.c -40 +114 M ./cell/fftw-cell.h -1 +15 M ./cell/spu/dft.spuc -27 +37 M ./dft/rank-geq2.c +4 Mon Nov 20 14:41:45 EST 2006 athena@fftw.org * added emacs mode M ./cell/spu/copy.spuc +1 Mon Nov 20 09:34:12 EST 2006 athena@fftw.org * revised transpose, cleanup A ./cell/spu/transpose.spuc A ./cell/spu/dft.spuc M ./cell/spu/Makefile.am -1 +1 M ./cell/spu/dft.spuc +165 M ./cell/spu/fftw-spu.h +5 M ./cell/spu/main.spuc -224 +2 M ./cell/spu/transpose.spuc +70 Sun Nov 19 20:20:23 EST 2006 athena@fftw.org * added file A ./cell/spu/Makefile.am Sun Nov 19 20:18:35 EST 2006 athena@fftw.org * removed file R ./cell/spu/generate.sh Sun Nov 19 20:15:38 EST 2006 athena@fftw.org * better automake integration M! ./cell/Makefile.am -9 +4 R! ./cell/spu/alloc.c A! ./cell/spu/alloc.spuc R! ./cell/spu/copy.c A! ./cell/spu/copy.spuc R! ./cell/spu/dma.c A! ./cell/spu/dma.spuc R! ./cell/spu/execute.c A! ./cell/spu/execute.spuc R! ./cell/spu/main.c A! ./cell/spu/main.spuc R! ./cell/spu/planner.c A! ./cell/spu/planner.spuc R! ./cell/spu/spu_n2fv_10.c R! ./cell/spu/spu_n2fv_11.c R! ./cell/spu/spu_n2fv_12.c R! ./cell/spu/spu_n2fv_13.c R! ./cell/spu/spu_n2fv_14.c R! ./cell/spu/spu_n2fv_15.c R! ./cell/spu/spu_n2fv_16.c R! ./cell/spu/spu_n2fv_2.c R! ./cell/spu/spu_n2fv_3.c R! ./cell/spu/spu_n2fv_32.c R! ./cell/spu/spu_n2fv_4.c R! ./cell/spu/spu_n2fv_5.c R! ./cell/spu/spu_n2fv_6.c R! ./cell/spu/spu_n2fv_7.c R! ./cell/spu/spu_n2fv_8.c R! ./cell/spu/spu_n2fv_9.c R! ./cell/spu/spu_t1fv_10.c R! ./cell/spu/spu_t1fv_12.c R! ./cell/spu/spu_t1fv_15.c R! ./cell/spu/spu_t1fv_16.c R! ./cell/spu/spu_t1fv_2.c R! ./cell/spu/spu_t1fv_3.c R! ./cell/spu/spu_t1fv_32.c R! ./cell/spu/spu_t1fv_4.c R! ./cell/spu/spu_t1fv_5.c R! ./cell/spu/spu_t1fv_6.c R! ./cell/spu/spu_t1fv_7.c R! ./cell/spu/spu_t1fv_8.c R! ./cell/spu/spu_t1fv_9.c M! ./dft/indirect-transpose.c -1 +1 M! ./rdft/direct2.c -1 +13 Sat Nov 18 20:14:29 EST 2006 athena@fftw.org * changed algorithm for computing chunk size M ./cell/spu/main.c -32 +36 Sat Nov 18 19:18:11 EST 2006 athena@fftw.org * implemented transpose, various fixes. M! ./cell/Makefile.am -2 +2 M! ./cell/cell.c -2 +10 M! ./cell/conf.c -1 +1 M! ./cell/dft-direct-cell.c -13 +29 M! ./cell/fftw-cell.h -3 +13 A! ./cell/spu/copy.c A! ./cell/spu/dma.c M! ./cell/spu/fftw-spu.h +27 M! ./cell/spu/main.c -169 +83 A! ./cell/transpose.c M! ./configure.ac -2 +5 M! ./kernel/ifftw.h +5 M! ./rdft/rank0.c +24 M! ./tests/fftw-bench.c +4 Thu Nov 16 16:33:50 EST 2006 athena@fftw.org * Added explicit destructor to all solvers to help with the cell port. M! ./Makefile.am -2 +6 M! ./cell/cell.c -4 +22 M! ./cell/dft-direct-cell.c -1 +8 M! ./cell/fftw-cell.h -1 +3 M! ./configure.ac +1 M! ./dft/bluestein.c -1 +1 M! ./dft/buffered.c -1 +1 M! ./dft/ct.c -1 +1 A! ./dft/ctsq.c M! ./dft/direct.c -1 +1 M! ./dft/generic.c -1 +1 M! ./dft/indirect-transpose.c -1 +1 M! ./dft/indirect.c -1 +1 M! ./dft/nop.c -1 +1 M! ./dft/rader.c -1 +1 M! ./dft/rank-geq2.c -1 +1 M! ./dft/vrank-geq1.c -1 +1 M! ./kernel/ifftw.h +1 M! ./kernel/solver.c -1 +4 M! ./mpi/dft-rank-geq2.c -1 +1 M! ./mpi/dft-serial.c -1 +1 M! ./mpi/transpose-alltoall.c -1 +1 M! ./mpi/transpose-pairwise.c -1 +1 A! ./mpi/transpose-radix2.c M! ./rdft/buffered.c -1 +1 M! ./rdft/buffered2.c -1 +1 M! ./rdft/dft-r2hc.c -1 +1 M! ./rdft/dht-r2hc.c -1 +1 M! ./rdft/dht-rader.c -1 +1 A! ./rdft/direct.c M! ./rdft/direct2.c -1 +1 M! ./rdft/generic.c -1 +1 M! ./rdft/hc2hc.c -1 +1 M! ./rdft/indirect.c -1 +1 M! ./rdft/nop.c -1 +1 M! ./rdft/nop2.c -1 +1 M! ./rdft/rank-geq2-rdft2.c -1 +1 M! ./rdft/rank-geq2.c -1 +1 M! ./rdft/rank0-rdft2.c -1 +1 M! ./rdft/rank0.c -1 +1 M! ./rdft/rdft-dht.c -1 +1 A! ./rdft/rdft2-radix2.c M! ./rdft/vrank-geq1-rdft2.c -1 +1 M! ./rdft/vrank-geq1.c -1 +1 M! ./rdft/vrank3-transpose.c -1 +1 M! ./reodft/redft00e-r2hc-pad.c -1 +1 M! ./reodft/redft00e-r2hc.c -1 +1 M! ./reodft/reodft00e-splitradix.c -1 +1 M! ./reodft/reodft010e-r2hc.c -1 +1 M! ./reodft/reodft11e-r2hc-odd.c -1 +1 M! ./reodft/reodft11e-r2hc.c -1 +1 M! ./reodft/reodft11e-radix2.c -1 +1 M! ./reodft/rodft00e-r2hc-pad.c -1 +1 M! ./reodft/rodft00e-r2hc.c -1 +1 M! ./tests/fftw-bench.c -1 M! ./threads/ct.c -1 +1 M! ./threads/dft-vrank-geq1.c -1 +1 M! ./threads/hc2hc.c -1 +1 M! ./threads/rdft-vrank-geq1.c -1 +1 M! ./threads/vrank-geq1-rdft2.c -1 +1 Thu Nov 16 15:22:15 EST 2006 athena@fftw.org * consistent use of #if vs. #ifdef M ./api/configure.c +2 M ./cell/cell.c -1 +1 M ./cell/conf.c -1 +1 M ./cell/dft-direct-cell.c -1 +1 M ./cell/plans-double.c -1 +1 M ./cell/plans-single.c -1 +1 Thu Nov 16 15:15:34 EST 2006 athena@fftw.org * Additional Cell double codelets, better automake integration A! ./cell/spu/spu_n2fv_11.c A! ./cell/spu/spu_n2fv_13.c A! ./cell/spu/spu_n2fv_15.c A! ./cell/spu/spu_n2fv_3.c A! ./cell/spu/spu_n2fv_5.c A! ./cell/spu/spu_n2fv_7.c A! ./cell/spu/spu_n2fv_9.c M! ./Makefile.am -7 +5 M! ./cell/Makefile.am -2 +7 M! ./cell/cell.c -11 +7 M! ./cell/conf.c +26 M! ./cell/dft-direct-cell.c -9 +14 M! ./cell/fftw-cell.h -2 M! ./cell/plans-double.c -196 +201 M! ./cell/plans-single.c -3 +9 M! ./cell/spu/execute.c +10 M! ./cell/spu/fftw-spu.h -8 +22 M! ./cell/spu/generate.sh -1 +6 M! ./cell/spu/main.c -2 +1 M! ./cell/spu/planner.c +3 M! ./cell/spu/spu_n2fv_11.c +118 M! ./cell/spu/spu_n2fv_13.c +155 M! ./cell/spu/spu_n2fv_15.c +132 M! ./cell/spu/spu_n2fv_3.c +37 M! ./cell/spu/spu_n2fv_5.c +52 M! ./cell/spu/spu_n2fv_7.c +68 M! ./cell/spu/spu_n2fv_9.c +103 M! ./configure.ac +6 M! ./dft/buffered.c -1 +14 M! ./kernel/align.c -1 +1 M! ./kernel/ifftw.h -2 +2 M! ./simd/simd.h -4 M! ./simd/taint.c -1 +1 Thu Nov 16 12:43:34 EST 2006 athena@fftw.org * Use dma lists. M ./cell/spu/main.c -21 +86 Thu Nov 16 11:03:46 EST 2006 athena@fftw.org * converted to automake R ./cell/spu/Makefile.in M ./cell/Makefile.am +1 Wed Nov 15 18:00:12 EST 2006 athena@fftw.org * Initial port to Cell Broadband Engine. A! ./cell/ A! ./cell/spu/ A! ./cell/Makefile.am A! ./cell/fftw-cell.h A! ./cell/plans-single.c A! ./cell/spu/Makefile.in A! ./cell/spu/execute.c A! ./cell/spu/fftw-spu.h A! ./cell/spu/generate.sh A! ./cell/spu/main.c A! ./cell/spu/planner.c A! ./cell/spu/spu-single.h A! ./cell/spu/spu_n2fv_10.c A! ./cell/spu/spu_n2fv_12.c A! ./cell/spu/spu_n2fv_14.c A! ./cell/spu/spu_n2fv_16.c A! ./cell/spu/spu_n2fv_2.c A! ./cell/spu/spu_n2fv_32.c A! ./cell/spu/spu_n2fv_4.c A! ./cell/spu/spu_n2fv_6.c A! ./cell/spu/spu_n2fv_8.c A! ./cell/spu/spu_t1fv_10.c A! ./cell/spu/spu_t1fv_12.c A! ./cell/spu/spu_t1fv_15.c A! ./cell/spu/spu_t1fv_16.c A! ./cell/spu/spu_t1fv_2.c A! ./cell/spu/spu_t1fv_3.c A! ./cell/spu/spu_t1fv_32.c A! ./cell/spu/spu_t1fv_4.c A! ./cell/spu/spu_t1fv_5.c A! ./cell/spu/spu_t1fv_6.c A! ./cell/spu/spu_t1fv_7.c A! ./cell/spu/spu_t1fv_8.c A! ./cell/spu/spu_t1fv_9.c A! ./cell/cell.c A! ./cell/conf.c A! ./cell/dft-direct-cell.c A! ./cell/spu/spu-double.h A! ./cell/plans-double.c A! ./cell/spu/alloc.c M! ./Makefile.am -14 +14 M! ./api/configure.c +1 M! ./cell/Makefile.am +22 M! ./cell/cell.c +90 M! ./cell/conf.c +12 M! ./cell/dft-direct-cell.c +302 M! ./cell/fftw-cell.h +64 M! ./cell/plans-double.c +2056 M! ./cell/plans-single.c +1032 M! ./cell/spu/Makefile.in +35 M! ./cell/spu/alloc.c +24 M! ./cell/spu/execute.c +68 M! ./cell/spu/fftw-spu.h +86 M! ./cell/spu/generate.sh +16 M! ./cell/spu/main.c +281 M! ./cell/spu/planner.c +187 M! ./cell/spu/spu-double.h +105 M! ./cell/spu/spu-single.h +120 M! ./cell/spu/spu_n2fv_10.c +100 M! ./cell/spu/spu_n2fv_12.c +109 M! ./cell/spu/spu_n2fv_14.c +142 M! ./cell/spu/spu_n2fv_16.c +146 M! ./cell/spu/spu_n2fv_2.c +33 M! ./cell/spu/spu_n2fv_32.c +313 M! ./cell/spu/spu_n2fv_4.c +44 M! ./cell/spu/spu_n2fv_6.c +62 M! ./cell/spu/spu_n2fv_8.c +75 M! ./cell/spu/spu_t1fv_10.c +104 M! ./cell/spu/spu_t1fv_12.c +114 M! ./cell/spu/spu_t1fv_15.c +162 M! ./cell/spu/spu_t1fv_16.c +153 M! ./cell/spu/spu_t1fv_2.c +32 M! ./cell/spu/spu_t1fv_3.c +41 M! ./cell/spu/spu_t1fv_32.c +329 M! ./cell/spu/spu_t1fv_4.c +44 M! ./cell/spu/spu_t1fv_5.c +60 M! ./cell/spu/spu_t1fv_6.c +63 M! ./cell/spu/spu_t1fv_7.c +81 M! ./cell/spu/spu_t1fv_8.c +77 M! ./cell/spu/spu_t1fv_9.c +120 M! ./configure.ac +2 M! ./kernel/ifftw.h -17 +21 M! ./libbench2/util.c -1 +1 M! ./tests/fftw-bench.c +1 Wed Mar 14 10:19:53 EDT 2007 athena@fftw.org * Remove Codesourcery contributions from commercial tarball. M ./commercialize.sh +7 Wed Mar 14 08:59:18 EDT 2007 athena@fftw.org * Added FFTW_WISDOM_ONLY, at the request of Phil Dumont. M ./api/apiplan.c -19 +29 M ./api/fftw3.h +1 M ./tests/fftw-bench.c +1 Tue Mar 13 00:32:05 EDT 2007 stevenj@fftw.org * fixed potential MPI deadlock if timer misbehaves M ./kernel/timer.c -3 +3 Mon Mar 12 23:31:52 EDT 2007 stevenj@fftw.org * more work on MPI documentation M ./doc/fftw3.texi -31 +309 M ./mpi/api.c -3 +4 M ./mpi/fftw3-mpi.h -19 +19 Tue Feb 27 13:48:43 EST 2007 stevenj@fftw.org * index M ./doc/fftw3.texi +1 Tue Feb 27 13:46:45 EST 2007 stevenj@fftw.org * rename "new-data execute" to "new-array execute", since of course you do not need a new array to have new data M ./doc/FAQ/fftw-faq.bfnn -1 +2 M ./doc/fftw3.texi -12 +12 Tue Feb 27 13:43:55 EST 2007 stevenj@fftw.org * consistency with manual (guru execute -> new-data execute) M ./doc/FAQ/fftw-faq.bfnn -1 +1 Tue Feb 27 13:42:24 EST 2007 stevenj@fftw.org * texinfo fixes; renamed "guru execute" section to "new-data execute", since previously it seemed to lead to endless confusion with the guru planner API M ./doc/fftw3.texi -60 +83 Mon Feb 26 18:57:11 EST 2007 stevenj@fftw.org * consistently use n0/n1/.. everywhere instead of nx/ny/... (for consistency with d-dimensional case n[0], n[1], ...) ... first start at MPI documentation M ./api/fftw3.h -10 +10 M ./doc/fftw3.texi -138 +374 M ./mpi/fftw3-mpi.h -27 +27 Sat Mar 10 18:48:05 EST 2007 athena@fftw.org * Changed C++-style comment into K&R M ./kernel/cycle.h -1 +1 Sat Mar 10 18:47:12 EST 2007 athena@fftw.org * Forgot to add file A ./simd/simd-mips_ps.h Sat Mar 10 18:44:39 EST 2007 athena@fftw.org * Note removal of K7 support. M ./AUTHORS -1 +4 Sat Mar 10 18:41:52 EST 2007 athena@fftw.org * Updated manual for MIPS PS M ./doc/fftw3.texi -9 +12 Sat Mar 10 18:37:07 EST 2007 athena@fftw.org * Adopted MIPS_PS patches from Codesourcery. A ./simd/mips_ps.c A ./simd/mips_ps.h M ./configure.ac +23 M ./kernel/cycle.h +43 M ./kernel/ifftw.h -1 +2 M ./kernel/timer.c -1 +3 M ./simd/Makefile.am -1 +1 M ./simd/mips_ps.c +45 M ./simd/mips_ps.h +273 M ./simd/simd.h +4 Sun Feb 25 11:34:51 EST 2007 athena@fftw.org * Incorrect initialization of win32 semaphores M ./threads/threads.c -1 +1 Tue Jan 30 11:43:09 EST 2007 stevenj@fftw.org * win32 fixes (I think, still untested) M ./threads/threads.c -9 +10 Fri Jan 19 17:31:47 EST 2007 stevenj@fftw.org * message-size heuristic in tranpose-recurse M ./mpi/transpose-recurse.c -3 +15 Tue Jan 30 08:53:55 EST 2007 athena@fftw.org * Threading layer for Win32, completely untested. M ./threads/threads.c +69 Mon Jan 29 14:26:30 EST 2007 athena@fftw.org * Check for EINTR after sem_wait(), as suggested by Chip Salzenberg. M ./threads/threads.c -1 +11 Mon Jan 22 13:58:23 EST 2007 athena@fftw.org * Force vector recursion by means of a separate function pointer. I need this for Cell. M ./dft/ct.c -3 +7 M ./dft/ct.h -2 +8 M ./dft/dft.h -15 M ./dft/dftw-direct.c -2 +2 M ./dft/dftw-directsq.c -2 +3 M ./dft/dftw-generic.c -2 +2 M ./dft/dftw-genericbuf.c -2 +2 M ./dft/vrank-geq1.c -4 M ./threads/ct.c -1 +3 M ./threads/threads.h -1 +3 Mon Jan 22 09:28:35 EST 2007 athena@fftw.org * Merge multiplications by twiddle with multiplications by i for faster r2c transforms. M ./genfft/algsimp.ml +4 M ./genfft/simd.ml +4 M ./simd/simd-altivec.h +24 M ./simd/simd-sse.h +18 M ./simd/simd-sse2.h +18 Sun Jan 21 19:02:44 EST 2007 athena@fftw.org * Disabled vector recursion, too messy. M ./dft/dft.h -2 +13 Sun Jan 21 14:23:35 EST 2007 athena@fftw.org * Changed heuristics for vector recursion. As in fftw-3.1, NO_VRECURSE disables vector recursion. As an exception, however, vector recursion is allowed when the predicate VRECURSE_ANYWAYP is true. We need some form of vector recursion to obtain decent plans on Cell, and this solution captures the common cases without increasing planning time too much. M ./api/fftw3.h -1 +1 M ./api/mapflags.c -2 +2 M ./dft/ct.c -10 +6 M ./dft/dft.h +4 M ./dft/vrank-geq1.c -6 +3 M ./kernel/ifftw.h -2 +2 M ./kernel/planner.c -1 +1 M ./rdft/ct-hc2c.c -9 +4 M ./rdft/hc2hc.c -9 +4 M ./rdft/vrank-geq1.c -7 Sun Dec 17 22:31:17 EST 2006 athena@fftw.org * fixed hc2c for vector-recursion branch M ./rdft/ct-hc2c.c -2 +7 Tue Dec 5 12:52:36 EST 2006 athena@fftw.org * switch to default vector recursion M ./api/fftw3.h -1 +1 M ./api/mapflags.c -2 +2 M ./dft/ct.c -4 +6 M ./dft/vrank-geq1.c +7 M ./kernel/ifftw.h -2 +2 M ./kernel/planner.c -1 +1 M ./rdft/hc2hc.c -2 +7 M ./rdft/vrank-geq1.c -2 +9 Sat Jan 20 23:37:33 EST 2007 athena@fftw.org * Smarter algorithm for selection of nbuf. M ./kernel/buffered.c -8 +6 Sat Jan 20 22:15:33 EST 2007 athena@fftw.org * Increased buffer sizes according to Moore's law. M ./kernel/buffered.c -3 +5 Fri Jan 19 16:02:00 EST 2007 stevenj@fftw.org * fix another MPI synchronization bug -- several more places where cost_hook must be called to synchronized process timings (sigh) M ./api/flops.c -3 +3 M ./kernel/ifftw.h -4 +5 M ./kernel/planner.c -10 +6 M ./kernel/timer.c -4 +15 M ./mpi/api.c -3 +1 Fri Jan 19 12:08:07 EST 2007 athena@fftw.org * Set havewisdom=0 when calling forget_wisdom() in the test program. M ./tests/fftw-bench.c -1 +3 Fri Jan 19 10:29:56 EST 2007 stevenj@fftw.org * remove redundant check M ./kernel/planner.c -4 +1 Thu Jan 18 22:37:59 EST 2007 stevenj@fftw.org * fixed potential (unlikely) bug in wisdom import (triggered when importing impatient wisdom after creating more patient plans, but apparently only for nonstandard configure.c configurations) M ./kernel/planner.c -1 +5 Thu Jan 18 21:50:14 EST 2007 stevenj@fftw.org * added functions to gather/broadcast wisdom for MPI M ./configure.ac -3 +8 M ./mpi/Makefile.am -1 +1 M ./mpi/api.c -3 +3 M ./mpi/fftw3-mpi.h -1 +5 M ./mpi/mpi-bench.c -19 +40 A ./mpi/wisdom-api.c M ./tests/bench.c +10 M ./tests/fftw-bench.c -3 +3 M ./tests/fftw-bench.h +2 Thu Jan 11 18:33:17 EST 2007 stevenj@fftw.org * whoops, another int/INT bug M ./mpi/transpose-pairwise.c -1 +1 Thu Jan 11 17:42:24 EST 2007 stevenj@fftw.org * whoops, fixed bug in transpose-recurse for r != m M ./mpi/transpose-recurse.c -10 +10 Thu Jan 11 17:25:36 EST 2007 stevenj@fftw.org * canonicalize mpi-transposed flags by setting TRANSPOSED_IN/OUT where possible M ./mpi/transpose-problem.c -1 +8 Thu Jan 11 17:16:24 EST 2007 stevenj@fftw.org * replace transpose-radix2 with much more general transpose-recurse solver M ./kernel/Makefile.am -3 +3 M ./kernel/ifftw.h +5 M ./kernel/tensor2.c -21 A ./kernel/tensor3.c M ./mpi/Makefile.am -1 +1 M ./mpi/conf.c -1 +1 M ./mpi/dft-rank-geq2-transposed.c -1 M ./mpi/mpi-bench.c +1 M ./mpi/mpi-transpose.h -1 +1 R ./mpi/transpose-radix2.c A ./mpi/transpose-recurse.c Wed Jan 10 20:23:48 EST 2007 stevenj@fftw.org * rename transpose-inplace to transpose-pairwise, as the algorithm is not restricted to inplace operation ./mpi/transpose-inplace.c -> ./mpi/transpose-pairwise.c M ./mpi/Makefile.am -1 +1 M ./mpi/conf.c -1 +1 M ./mpi/mpi-transpose.h -1 +1 M ./mpi/testsched.c -1 +1 M ./mpi/transpose-pairwise.c -9 +9 Wed Jan 10 14:39:08 EST 2007 stevenj@fftw.org * whoops, some int/INT bugs M ./mpi/api.c -2 +2 M ./mpi/block.c -3 +3 M ./mpi/ifftw-mpi.h -5 +5 Tue Jan 9 18:50:07 EST 2007 stevenj@fftw.org * fix FAQ Makefile for vpath builds M ./doc/FAQ/Makefile.am -4 +7 M ./doc/FAQ/fftw-faq.bfnn -1 +1 Tue Jan 9 20:22:11 EST 2007 athena@fftw.org * Missing ``static'' keyword. M ./rdft/direct-r2c.c -2 +2 Tue Jan 9 20:13:18 EST 2007 athena@fftw.org * Minor cleanup. M ./dft/direct.c -27 +16 Tue Jan 9 00:04:03 EST 2007 stevenj@fftw.org * interleave twiddle mults with DFTs (should we use dftw?) M ./mpi/dft-rank1.c -41 +44 Mon Jan 8 18:35:41 EST 2007 stevenj@fftw.org * simplified (and somewhat sped up) dft-rank1 by exploiting dft-rank1-bigvec M ./mpi/dft-rank-geq2-transposed.c -1 +1 M ./mpi/dft-rank-geq2.c -1 +1 M ./mpi/dft-rank1-bigvec.c -2 +4 M ./mpi/dft-rank1.c -175 +181 M ./mpi/transpose-alltoall.c -6 +3 M ./mpi/transpose-inplace.c -6 +3 M ./mpi/transpose-radix2.c -7 +2 Sun Jan 7 00:31:31 EST 2007 stevenj@fftw.org * rearranged TRANSPOSED format, numerous speedups Split the TRANSPOSED and non-TRANSPOSED rank-geq2 solvers, and changed the DFT TRANSPOSED format to be more like fftw2 (both globally and locally transposed). In general, more emphasis on arranging the data contiguously for the DFTs, and more flexibility in intermediate transposed formats. Also disable NO_SLOW when planning transposes, since otherwise non-square in-place transposes gratuitously put the planner in SLOW mode. Currently, dft-rank1-bigvec has 5 variants (or 10, if DESTROY_INPUT). It looks like only 2 of these are commonly used, so I should probably add some UGLY tags once I do more benchmarking. M ./mpi/Makefile.am -1 +1 M ./mpi/api.c -14 +34 M ./mpi/conf.c +1 M ./mpi/dft-problem.c -2 +12 A ./mpi/dft-rank-geq2-transposed.c M ./mpi/dft-rank-geq2.c -131 +60 M ./mpi/dft-rank1-bigvec.c -77 +123 M ./mpi/dft-rank1.c -2 +2 M ./mpi/ifftw-mpi.h -2 +3 M ./mpi/mpi-bench.c +13 M ./mpi/mpi-dft.h -1 +2 M ./mpi/transpose-alltoall.c -32 +39 M ./mpi/transpose-inplace.c -53 +69 M ./mpi/transpose-radix2.c -37 +43 Thu Jan 4 19:13:17 EST 2007 stevenj@fftw.org * add bench_cost_postprocess to prevent deadlocks in mpi-bench M ./libbench2/Makefile.am -5 +6 A ./libbench2/bench-cost-postprocess.c M ./libbench2/bench-user.h +1 M ./libbench2/speed.c -2 +2 M ./mpi/mpi-bench.c +7 Thu Jan 4 16:46:29 EST 2007 stevenj@fftw.org * whoops M ./m4/acx_pthread.m4 -1 +1 Wed Jan 3 14:23:42 EST 2007 stevenj@fftw.org * pass proper pointer types as arguments, so that ACX_PTHREAD still works with C++ and -Werror (thanks to Ewald Arnold for the suggestion) M ./m4/acx_pthread.m4 -3 +6 Mon Jan 1 19:30:43 EST 2007 athena@fftw.org * Renamed [io]vs => [io]vs_by_nbuf, which is more appropriate and would have saved me 30mins debugging. M ./dft/buffered.c -6 +6 Mon Jan 1 18:52:38 EST 2007 stevenj@fftw.org * add --with-g77-wrappers option & always include g77 wrappers on GNU systems and/or with gfortran Upcoming GNU/Linux distros will most likely switch to configuring FFTW with gfortran by default, since g77 isn't even included with recent gcc versions. However, we still want to include g77-compatible wrappers in this case (two underscores) in addition to gfortran wrappers (one underscore) lest we silently break binary compatibility and provoke lots of annoying emails. M ./api/x77.h +6 M ./configure.ac +11 Mon Jan 1 16:48:36 EST 2007 stevenj@fftw.org * use AC_HELP_STRING for --disable-fortran M ./configure.ac -1 +1 Mon Jan 1 15:56:12 EST 2007 stevenj@fftw.org * terminology M ./doc/FAQ/fftw-faq.bfnn -1 +1 Sat Dec 30 16:18:35 EST 2006 athena@fftw.org * Free buffers before calling cldrest. M ./dft/buffered.c -2 +2 M ./rdft/buffered.c -4 +4 Fri Dec 29 10:52:15 EST 2006 athena@fftw.org * Removed obsolete code. M ./simd/simd-sse.h -22 +9 Thu Dec 28 21:37:48 EST 2006 athena@fftw.org * Attempt to work-around old gcc bugs in a more efficient fashion that does not lose performance on newer gcc's. M ./simd/simd-sse.h -13 +24 Thu Dec 28 16:10:33 EST 2006 athena@fftw.org * Make sure that the speed() input is zero even in paranoid mode. M ./libbench2/speed.c +6 Thu Dec 28 11:41:46 EST 2006 athena@fftw.org * cld0 and cldm problems must be tainted because they are used in a v-loop. M ./rdft/ct-hc2c-direct.c -2 +5 M ./rdft/hc2hc-direct.c -2 +3 Wed Dec 27 17:17:45 EST 2006 athena@fftw.org * Run paranoid-check in patient mode. M ./tests/Makefile.am -5 +5 Wed Dec 27 10:51:42 EST 2006 athena@fftw.org * Fixed incorrect initialization to zero. M ./rdft/problem2.c -1 +1 Wed Dec 27 09:33:02 EST 2006 athena@fftw.org * Fixed wrong TAINT() M ./rdft/buffered2.c -1 +1 Tue Dec 26 22:50:28 EST 2006 athena@fftw.org * Grrrrr... M ./libbench2/verify-rdft2.c -1 +1 Tue Dec 26 22:48:44 EST 2006 athena@fftw.org * Give up trying to verify rdft2 when vrank=-infinity. M ./libbench2/verify-rdft2.c +3 Tue Dec 26 22:31:38 EST 2006 athena@fftw.org * typo M ./dft/direct.c -1 +2 Tue Dec 26 21:54:53 EST 2006 athena@fftw.org * Correctly verify rdft2 when vrank = -infinity. M ./libbench2/verify-rdft2.c -1 +1 Tue Dec 26 21:25:02 EST 2006 athena@fftw.org * rdft/buffered2.c now generates rdft2 subproblems, not rdft. The old rdft2->rdft reduction is now in rdft/rdft2-rdft.c and still does way too much. M ./TODO -3 M ./rdft/Makefile.am -1 +1 M ./rdft/buffered.c -1 +1 M ./rdft/buffered2.c -190 +200 M ./rdft/conf.c +1 M ./rdft/problem2.c -5 +6 M ./rdft/rdft.h -1 +2 M ./rdft/rdft2-inplace-strides.c -5 +5 A ./rdft/rdft2-rdft.c M ./rdft/rdft2-strides.c -5 +5 M ./rdft/vrank-geq1-rdft2.c -11 +10 Tue Dec 26 14:03:27 EST 2006 athena@fftw.org * Buffer the input in hc2r problems, as opposed to the output. M ./rdft/buffered.c -34 +104 M ./rdft/buffered2.c -1 Tue Dec 26 10:02:59 EST 2006 athena@fftw.org * streamlined buffered solvers M ./dft/buffered.c -66 +14 M ./kernel/buffered.c -4 +25 M ./kernel/ifftw.h -1 +4 M ./kernel/primes.c +11 M ./rdft/buffered.c -63 +13 M ./rdft/buffered2.c -54 +7 Mon Dec 25 16:08:22 EST 2006 athena@fftw.org * c++ compatibility M ./libbench2/verify-r2r.c -2 +2 Sun Dec 24 20:27:23 EST 2006 athena@fftw.org * Gratuitous renaming of directories and files since the old naming was becoming too inconsistent for my taste. ./dft/codelets -> ./dft/scalar ./dft/scalar/standard -> ./dft/scalar/codelets ./rdft/codelets -> ./rdft/scalar M ./Makefile.am -14 +15 M ./configure.ac -6 +6 M ./dft/Makefile.am -1 +1 M ./dft/scalar/Makefile.am -3 +3 M ./dft/scalar/codelets/Makefile.am -3 +3 M ./dft/simd/codelets/Makefile.am -2 +2 M ./doc/fftw3.texi -4 +4 M ./rdft/Makefile.am -1 +1 M ./rdft/scalar/Makefile.am -2 +2 M ./rdft/scalar/r2cb/Makefile.am -3 +3 M ./rdft/scalar/r2cf/Makefile.am -3 +3 M ./rdft/scalar/r2r/Makefile.am -3 +3 Sun Dec 24 20:11:50 EST 2006 athena@fftw.org * Fixed another dftw bug (sigh) M ./threads/ct.c -7 +3 Sun Dec 24 11:48:01 EST 2006 athena@fftw.org * Removed debugging leftovers. M ./dft/dftw-genericbuf.c -2 +2 Sun Dec 24 11:34:17 EST 2006 athena@fftw.org * Moved dftw-generic* to new dftw protocol. M ./dft/dftw-generic.c -30 +24 M ./dft/dftw-genericbuf.c -13 +13 Sun Dec 24 09:37:19 EST 2006 athena@fftw.org * Oops M ./tests/hook.c -1 +1 Sun Dec 24 09:31:46 EST 2006 athena@fftw.org * Fixed wrong verification of rank-1 rdft2 M ./tests/hook.c -4 +6 Sat Dec 23 20:11:29 EST 2006 athena@fftw.org * minor tweaks M ./dft/dftw-direct.c -2 +4 Sat Dec 23 19:50:03 EST 2006 athena@fftw.org * Removed obsolete items. M ./TODO -13 Sat Dec 23 17:56:37 EST 2006 athena@fftw.org * Modified the problem_dftw invocation protocol. apply() now requires pointers to the beginning of the full array. Each thread processes a slice mb <= m < me. This protocol is consistent with the one used in hc2hc, where there is no other choice. M ./dft/dftw-direct.c -27 +31 M ./dft/dftw-directsq.c -1 +3 M ./threads/ct.c -7 +2 Sat Dec 23 16:18:25 EST 2006 athena@fftw.org * typo M ./simd/simd-altivec.h -1 +1 Sat Dec 23 16:06:56 EST 2006 athena@fftw.org * changed hc2hc twiddle storage to be the same as hc2c M ./genfft/gen_hc2hc.ml -2 +3 M ./rdft/hc2hc-direct.c -14 +11 Sat Dec 23 15:16:36 EST 2006 athena@fftw.org * Allowed extra_iter in dftw-direct. Rationalized twiddle factors in hc2c. M ./dft/dftw-direct.c -10 +38 M ./dft/simd/Makefile.am -1 +1 M ./dft/simd/codelets/Makefile.am -1 +12 M ./dft/simd/t.c -2 +45 A ./dft/simd/t1bu.h A ./dft/simd/t1fu.h M ./genfft/gen_hc2c.ml -2 +3 M ./genfft/gen_hc2cdft.ml -2 +3 M ./genfft/gen_hc2cdft_c.ml -1 +1 M ./genfft/gen_hc2hc.ml -1 +1 M ./genfft/gen_twiddle.ml -1 +1 M ./genfft/gen_twiddle_c.ml -1 +1 M ./genfft/gen_twidsq.ml -1 +1 M ./genfft/gen_twidsq_c.ml -1 +1 M ./genfft/twiddle.ml -12 +10 M ./genfft/twiddle.mli -1 +1 M ./kernel/twiddle.c -4 +8 M ./rdft/ct-hc2c-direct.c -4 +7 M ./rdft/simd/hc2cbv.h -2 +2 M ./rdft/simd/hc2cfv.h -2 +2 M ./simd/simd-altivec.h -12 +9 M ./simd/simd-sse.h -12 +9 M ./simd/simd-sse2.h -11 +6 Sat Dec 23 10:37:11 EST 2006 athena@fftw.org * Implemented unmentionable hack to use 4-way SIMD with an odd number of iterations. M ./dft/direct.c -9 +34 M ./rdft/ct-hc2c-direct.c -43 +88 M ./simd/simd-altivec.h -1 +3 M ./simd/simd-sse.h -1 +3 Fri Dec 22 22:13:30 EST 2006 athena@fftw.org * altivec support for new codelets M ./simd/simd-altivec.h -7 +28 Fri Dec 22 19:09:15 EST 2006 athena@fftw.org * fixed incorrect computation of W M ./genfft/gen_hc2cdft_c.ml -1 +1 Fri Dec 22 18:51:22 EST 2006 athena@fftw.org * Implemented 4-way simd hc2cdftv Also eliminated the twiddle_shift hack. A zillion changes dictated by this choice, which was in turn necessary for the hc2cdftv thing to work. M ./Makefile.am -13 +8 M ./configure.ac -3 M ./dft/codelet-dft.h -9 +9 M ./dft/codelets/t.c -4 +4 M ./dft/dftw-direct.c -35 +30 M ./dft/dftw-directsq.c -7 +4 M ./dft/simd/q1b.c -6 +6 M ./dft/simd/q1f.c -6 +6 M ./dft/simd/t.c -14 +14 M ./dft/simd/t3b.h +1 M ./dft/simd/t3f.h +1 M ./dft/simd/ts.c -6 +6 M ./dft/simd/ts.h +1 M ./genfft/Makefile.am -17 +10 R ./genfft/gen_conv.ml M ./genfft/gen_hc2c.ml -25 +19 M ./genfft/gen_hc2cdft.ml -19 +19 M ./genfft/gen_hc2cdft_c.ml -20 +24 M ./genfft/gen_hc2hc.ml -23 +17 M ./genfft/gen_mdct.ml -9 +9 M ./genfft/gen_notw.ml -8 +8 M ./genfft/gen_notw_c.ml -6 +6 M ./genfft/gen_r2cb.ml -9 +9 M ./genfft/gen_r2cf.ml -9 +9 M ./genfft/gen_r2r.ml -6 +6 M ./genfft/gen_twiddle.ml -37 +39 M ./genfft/gen_twiddle_c.ml -38 +42 M ./genfft/gen_twidsq.ml -42 +44 M ./genfft/gen_twidsq_c.ml -44 +47 M ./genfft/genutil.ml -6 +6 M ./genfft/simd.ml -6 +7 M ./genfft/simd.mli -2 +1 M ./genfft/variable.ml -7 +12 M ./genfft/variable.mli -1 +3 M ./kernel/ifftw.h -7 M ./kernel/twiddle.c -14 M ./rdft/codelet-rdft.h -5 +5 M ./rdft/codelets/hc2c.c -2 +2 M ./rdft/conf.c -1 +1 M ./rdft/ct-hc2c-direct.c -47 +36 M ./rdft/ct-hc2c.c -1 +3 M ./rdft/hc2hc-direct.c -39 +31 M ./rdft/simd/codelets/Makefile.am -2 +2 M ./rdft/simd/hc2cbv.c -7 +8 M ./rdft/simd/hc2cbv.h -4 +3 M ./rdft/simd/hc2cfv.c -7 +8 M ./rdft/simd/hc2cfv.h -4 +3 M ./simd/simd-sse.h +4 M ./simd/simd-sse2.h -3 +5 M ./simd/simd.h -2 M ./tests/Makefile.am +9 Fri Dec 22 08:45:46 EST 2006 athena@fftw.org * Fixed verification of rdft2 problems with new format. M ./tests/hook.c -1 +5 Fri Dec 22 00:05:59 EST 2006 athena@fftw.org * Added file A ./rdft/simd/Makefile.am Fri Dec 22 00:02:50 EST 2006 athena@fftw.org * Hmm, previous commit did not work A ./rdft/simd/ A ./rdft/simd/codelets/ A ./rdft/simd/codelets/Makefile.am A ./rdft/simd/hc2hcv.h ./rdft/simd/hc2hcv.h -> ./rdft/simd/hc2cfv.h A ./rdft/simd/hc2cbv.h A ./rdft/simd/hc2cv.c ./rdft/simd/hc2cv.c -> ./rdft/simd/hc2cfv.c M ./rdft/simd/codelets/Makefile.am +51 A ./rdft/simd/hc2cbv.c M ./rdft/simd/hc2cbv.h +29 M ./rdft/simd/hc2cfv.c +41 M ./rdft/simd/hc2cfv.h +29 Thu Dec 21 23:58:33 EST 2006 athena@fftw.org * Added SIMD r2cdft codelets. A ./genfft/gen_hc2cdft_c.ml M ./Makefile.am -14 +22 M ./configure.ac +5 M ./dft/simd/n1b.h -1 M ./dft/simd/n1f.h -1 M ./dft/simd/n2b.h -1 M ./dft/simd/n2f.h -1 M ./dft/simd/n2s.h -1 M ./genfft/Makefile.am -8 +14 M ./genfft/algsimp.ml -2 +12 M ./genfft/c.ml +1 M ./genfft/expr.ml -1 +3 M ./genfft/expr.mli -1 +1 M ./genfft/gen_hc2cdft_c.ml +217 M ./genfft/simd.ml -2 +9 M ./genfft/to_alist.ml +1 M ./kernel/ifftw.h +6 M ./rdft/Makefile.am -1 +1 M ./rdft/codelet-rdft.h +5 M ./rdft/codelets/hc2c.c -2 +12 M ./rdft/conf.c +4 M ./rdft/ct-hc2c-direct.c -5 +60 M ./rdft/hc2hc-direct.c -5 M ./rdft/hc2hc.h -2 M ./rdft/khc2hc.c -1 M ./simd/nonportable/sse.c -9 +4 M ./simd/nonportable/sse2.c -9 +4 M ./simd/simd-sse.h -4 +4 M ./simd/simd-sse2.h -4 +7 M ./simd/sse.c -2 +2 M ./simd/sse2.c -2 +2 M ./support/Makefile.codelets +1 Thu Dec 21 21:19:21 EST 2006 athena@fftw.org * Bug in buffering, grrr... M ./rdft/ct-hc2c-direct.c -1 +1 Thu Dec 21 20:58:14 EST 2006 athena@fftw.org * Oops, memory leak. M ./rdft/hc2hc-direct.c +1 Thu Dec 21 17:12:31 EST 2006 athena@fftw.org * minor changes, cleanup. M ./rdft/ct-hc2c.c -2 M ./rdft/hc2hc.c -16 +10 M ./threads/hc2hc.c -17 +12 Wed Dec 20 22:09:28 EST 2006 athena@fftw.org * Unified hc2hc-direct, hc2hc-directbuf. Cleanup. M ./rdft/Makefile.am -4 +3 M ./rdft/ct-hc2c-direct.c -8 +8 R ./rdft/hc2hc-common.c M ./rdft/hc2hc-direct.c -44 +153 R ./rdft/hc2hc-directbuf.c M ./rdft/hc2hc.h -4 Wed Dec 20 17:55:56 EST 2006 athena@fftw.org * removed obsolete rdft2-radix2 M ./rdft/Makefile.am -1 +1 M ./rdft/conf.c -1 M ./rdft/rdft.h -1 R ./rdft/rdft2-radix2.c Wed Dec 20 17:51:20 EST 2006 athena@fftw.org * implemented reduction rdft2->dft A ./genfft/gen_hc2cdft.ml M ./genfft/Makefile.am -9 +16 M ./genfft/gen_hc2c.ml -1 +1 M ./genfft/gen_hc2cdft.ml +207 M ./rdft/codelet-rdft.h -2 +7 M ./rdft/codelets/r2cb/Makefile.am -1 +12 M ./rdft/codelets/r2cf/Makefile.am -1 +12 M ./rdft/ct-hc2c-direct.c -5 +8 M ./rdft/ct-hc2c.c -24 +95 M ./rdft/ct-hc2c.h -2 +6 M ./rdft/khc2c.c -2 +3 M ./support/Makefile.codelets +1 Wed Dec 20 09:29:39 EST 2006 athena@fftw.org * Implemented buffered direct-r2c, direct-hc2c. Also, removed some old cruft: * okp() functions were never used and a pain to maintain---now they are gone. * ``m'' in hc2hc and hc2c codelets is now the number of iterations, not the ``logical'' m. M ./genfft/gen_hc2c.ml -5 +3 M ./genfft/gen_hc2hc.ml -9 +9 M ./genfft/gen_r2cb.ml -4 +1 M ./genfft/gen_r2cf.ml -4 +1 M ./genfft/gen_r2r.ml -5 +1 M ./rdft/codelet-rdft.h -30 +4 M ./rdft/codelets/hc2c.c -14 +2 M ./rdft/codelets/hfb.c -14 +2 M ./rdft/codelets/r2c.c -18 +4 M ./rdft/codelets/r2r.c -15 +1 M ./rdft/ct-hc2c-direct.c -37 +127 M ./rdft/direct-r2c.c -41 +189 M ./rdft/direct-r2r.c -15 +6 M ./rdft/direct2.c -20 +11 M ./rdft/hc2hc-direct.c -23 +9 M ./rdft/hc2hc-directbuf.c -79 +42 M ./rdft/kr2c.c +1 M ./rdft/rdft.h +1 Tue Dec 19 17:07:04 EST 2006 stevenj@fftw.org * added memcpy-loop rank0 solver (it makes a 5-20% difference for transposes of large tuples) M ./rdft/rank0.c +33 Tue Dec 19 16:15:54 EST 2006 stevenj@fftw.org * new variable to disable libbench2's problem allocation during speed benchmarking (to benchmark MPI transforms where the array does not fit into the memory of a single process) M ./libbench2/bench-user.h +2 M ./libbench2/speed.c -3 +8 M ./mpi/mpi-bench.c -5 +15 Tue Dec 19 14:55:08 EST 2006 stevenj@fftw.org * allow transpose-inplace to use input as scratch for DESTROY_INPUT plans (to avoid non-square in-place transpositions) ... on supersgj, the planner often prefers transpose-inplace to transpose-alltoall in this case (apparently MPI_Alltoall in LAM MPI isn't that great) M ./mpi/transpose-inplace.c -52 +99 Tue Dec 19 17:07:14 EST 2006 athena@fftw.org * For some reason HB2 codelets were not generated. M ./rdft/codelets/r2cb/Makefile.am -2 +2 M ./rdft/codelets/r2cf/Makefile.am -1 +1 Tue Dec 19 15:12:39 EST 2006 athena@fftw.org * split rdft/direct.c into direct-r2r and direct-r2c, since the file was getting out of control. A ./rdft/direct-r2c.c A ./rdft/direct-r2r.c R ./rdft/direct.c M ./rdft/Makefile.am -4 +5 M ./rdft/direct-r2c.c +195 M ./rdft/direct-r2r.c +154 Tue Dec 19 02:59:35 EST 2006 stevenj@fftw.org * added dft-rank1 solver - MPI now supports 1d complex DFTs! M ./mpi/Makefile.am -2 +2 M ./mpi/api.c -9 +54 A ./mpi/choose-radix.c M ./mpi/conf.c +1 M ./mpi/dft-problem.c -2 +2 A ./mpi/dft-rank1.c M ./mpi/fftw3-mpi.h -2 +8 M ./mpi/ifftw-mpi.h +4 M ./mpi/mpi-bench.c -25 +24 M ./mpi/mpi-dft.h -4 +5 M ./mpi/mpi-transpose.h -3 +3 M ./mpi/transpose-problem.c -1 +1 M ./tests/bench.c -8 +8 M ./tests/fftw-bench.h -1 +1 Tue Dec 19 01:27:20 EST 2006 stevenj@fftw.org * fftw_flops must call cost_hook directly; iestimate_cost always uses COST_MAX M ./api/flops.c -1 +7 M ./kernel/ifftw.h -2 +2 M ./kernel/planner.c -4 +3 Tue Dec 19 01:16:54 EST 2006 stevenj@fftw.org * fftw_flops and fftw_estimate_cost must now be called from every process, to prevent deadlocks in the MPI code (since they sum/max the cost over all processes) M ./tests/fftw-bench.c -7 +11 Tue Dec 19 00:55:34 EST 2006 stevenj@fftw.org * whoops, typo in assert M ./mpi/block.c -1 +1 Tue Dec 19 00:51:07 EST 2006 stevenj@fftw.org * remove multiplication by FFT_SIGN M ./kernel/trig.c -2 +7 Mon Dec 18 22:51:45 EST 2006 stevenj@fftw.org * need to synchronize ESTIMATE costs in MPI planner, and sum ESTIMATE costs for flop reporting: generalize measure_hook to cost_hook(..., {COST_SUM, COST_MAX}) M ./api/flops.c -1 +1 M ./kernel/ifftw.h -3 +6 M ./kernel/planner.c -8 +12 M ./mpi/api.c -6 +7 Mon Dec 18 15:36:15 EST 2006 stevenj@fftw.org * previous patch slowed down transpose-alltoall when TRANSPOSED_IN and DESTROY_INPUT; now allow planner to choose old behavior in this case M ./mpi/transpose-alltoall.c -11 +31 Mon Dec 18 15:15:39 EST 2006 stevenj@fftw.org * transpose-alltoall doesn't require input to be destroyed if TRANSPOSED_IN is set M ./mpi/transpose-alltoall.c -26 +41 Mon Dec 18 17:41:25 EST 2006 athena@fftw.org * Added t2-style hc2c codelets, fixed typos. M ./api/plan-guru-dft-c2r.h -1 +1 M ./api/plan-guru-dft-r2c.h -1 +1 M ./api/plan-guru-dft.h -2 +2 M ./rdft/codelets/r2cb/Makefile.am -1 +9 M ./rdft/codelets/r2cf/Makefile.am -1 +9 Mon Dec 18 16:02:42 EST 2006 athena@fftw.org * Renamed certain variables to avoid calling an output stride `is'. M ./rdft/codelet-rdft.h -1 +1 M ./rdft/direct.c -17 +17 M ./rdft/direct2.c -7 +7 Mon Dec 18 15:54:43 EST 2006 athena@fftw.org * Oops---wrong stride M ./rdft/ct-hc2c.c -1 +1 Mon Dec 18 14:59:16 EST 2006 athena@fftw.org * Oops, forgot to add file A ./rdft/kr2c.c Mon Dec 18 14:17:02 EST 2006 athena@fftw.org * Renamed r2hc/hc2r codelets to r2c After the recent changes, r2hc/hc2r codelets became rdft2 problems, so I renamed them accordingly to r2cf/r2cb. Codelet parameters are now a real array and a complex array, instead of an input array and an output array, and forward and backward codelets have the same type, which removes some clutter from the rdft code. ./genfft/gen_hc2r.ml -> ./genfft/gen_r2cb.ml ./genfft/gen_r2hc.ml -> ./genfft/gen_r2cf.ml ./rdft/codelets/hc2r -> ./rdft/codelets/r2cb ./rdft/codelets/r2hc -> ./rdft/codelets/r2cf A ./rdft/codelets/r2c.c A ./rdft/codelets/r2cb.h A ./rdft/codelets/r2cbIII.h A ./rdft/codelets/r2cf.h A ./rdft/codelets/r2cfII.h M ./Makefile.am -2 +2 M ./configure.ac -2 +2 M ./doc/fftw3.texi -1 +1 M ./genfft/Makefile.am -9 +9 M ./genfft/gen_r2cb.ml -58 +51 M ./genfft/gen_r2cf.ml -58 +51 M ./rdft/Makefile.am -3 +3 M ./rdft/codelet-rdft.h -59 +16 M ./rdft/codelets/Makefile.am -3 +3 R ./rdft/codelets/hc2r.c R ./rdft/codelets/hc2r.h R ./rdft/codelets/hc2rIII.h M ./rdft/codelets/r2c.c +51 M ./rdft/codelets/r2cb/Makefile.am -21 +21 M ./rdft/codelets/r2cb.h +23 M ./rdft/codelets/r2cbIII.h +23 M ./rdft/codelets/r2cf/Makefile.am -20 +20 M ./rdft/codelets/r2cf.h +23 M ./rdft/codelets/r2cfII.h +23 R ./rdft/codelets/r2hc.c R ./rdft/codelets/r2hc.h R ./rdft/codelets/r2hcII.h M ./rdft/conf.c -2 +2 M ./rdft/ct-hc2c.c -2 +2 M ./rdft/direct.c -46 +28 M ./rdft/direct2.c -72 +23 R ./rdft/khc2r.c R ./rdft/kr2hc.c M ./rdft/rdft.h -4 +2 M ./support/Makefile.codelets -2 +2 Mon Dec 18 10:48:07 EST 2006 athena@fftw.org * Implemented backward radix-2k rdft2. A ./rdft/codelets/hc2cf.h A ./rdft/codelets/hc2cb.h M ./genfft/gen_hc2c.ml -26 +30 M ./rdft/codelets/Makefile.am -1 +1 M ./rdft/codelets/hc2c.c -1 +5 R ./rdft/codelets/hc2c.h M ./rdft/codelets/hc2cb.h +23 M ./rdft/codelets/hc2cf.h +23 M ./rdft/codelets/hc2r/Makefile.am -1 +11 M ./rdft/codelets/r2hc/Makefile.am -7 +7 M ./rdft/ct-hc2c.c -3 +25 M ./rdft/rdft2-radix2.c -72 +1 Mon Dec 18 13:56:09 EST 2006 stevenj@fftw.org * move extract_reim into kernel, since it is used by internal MPI stuff and not just in the API code any more ./api/extract-reim.c -> ./kernel/extract-reim.c M ./api/Makefile.am -1 +1 M ./api/api.h -1 +1 M ./api/plan-many-dft-c2r.c -1 +1 M ./api/plan-many-dft-r2c.c -1 +1 M ./api/plan-many-dft.c -2 +2 M ./kernel/Makefile.am -5 +6 M ./kernel/extract-reim.c -6 +6 M ./kernel/ifftw.h +1 M ./mpi/dft-rank-geq2.c -3 M ./mpi/dft-rank1-bigvec.c -3 M ./mpi/dft-serial.c -3 Mon Dec 18 08:40:14 EST 2006 athena@fftw.org * Do not check r1==cr unless rnk>0 M ./rdft/problem2.c -1 +5 Sun Dec 17 21:03:50 EST 2006 athena@fftw.org * Implemented radix-2k RDFT2, forward only for now M ./api/execute-dft-c2r.c -1 +2 M ./api/execute-dft-r2c.c -1 +2 M ./api/execute-split-dft-c2r.c -1 +2 M ./api/execute-split-dft-r2c.c -1 +2 M ./api/f77funcs.h -4 +8 M ./api/plan-guru-dft-c2r.h -8 +8 M ./api/plan-guru-dft-r2c.h -9 +9 M ./api/plan-guru-split-dft-c2r.h -6 +6 M ./api/plan-guru-split-dft-r2c.h -9 +9 M ./api/plan-many-dft-c2r.c -3 +3 M ./api/plan-many-dft-r2c.c -1 +1 M ./dft/problem.c -3 +5 M ./genfft/Makefile.am -5 +11 A ./genfft/gen_hc2c.ml M ./genfft/gen_hc2hc.ml -30 +30 M ./genfft/gen_hc2r.ml -7 +16 M ./genfft/gen_r2hc.ml -8 +18 M ./rdft/Makefile.am -1 +2 M ./rdft/buffered2.c -31 +45 M ./rdft/codelet-rdft.h -11 +41 M ./rdft/codelets/Makefile.am -1 +1 A ./rdft/codelets/hc2c.c A ./rdft/codelets/hc2c.h M ./rdft/codelets/hc2r.c -4 +5 M ./rdft/codelets/hfb.c -4 +4 M ./rdft/codelets/r2hc.c -4 +5 M ./rdft/codelets/r2hc/Makefile.am -1 +11 A ./rdft/ct-hc2c-direct.c A ./rdft/ct-hc2c.c A ./rdft/ct-hc2c.h M ./rdft/direct.c -16 +21 M ./rdft/direct2.c -15 +32 M ./rdft/hc2hc-common.c -2 +2 M ./rdft/hc2hc-direct.c -4 +4 M ./rdft/hc2hc-directbuf.c -8 +8 M ./rdft/hc2hc-generic.c -4 +3 M ./rdft/hc2hc.c -4 +4 A ./rdft/khc2c.c M ./rdft/nop2.c -7 +10 M ./rdft/problem.c -2 +1 M ./rdft/problem2.c -38 +117 M ./rdft/rank-geq2-rdft2.c -13 +17 M ./rdft/rank0-rdft2.c -35 +44 M ./rdft/rdft.h -8 +27 M ./rdft/rdft2-inplace-strides.c -2 +6 M ./rdft/rdft2-radix2.c -131 +37 M ./rdft/solve2.c -1 +3 M ./rdft/vrank-geq1-rdft2.c -7 +7 M ./support/Makefile.codelets +1 M ./tests/hook.c -8 +23 M ./threads/vrank-geq1-rdft2.c -10 +10 Mon Dec 18 01:23:45 EST 2006 stevenj@fftw.org * separate TRANSPOSED/SCRAMBLED flags internally (this is required so that dft-rank1-bigvec and the future dft-rank1 won't have incompatible SCRAMBLED formats) M ./mpi/api.c -16 +2 M ./mpi/dft-rank-geq2.c -7 +7 M ./mpi/dft-rank1-bigvec.c -8 +9 M ./mpi/dft-serial.c -1 +1 M ./mpi/ifftw-mpi.h -1 +8 M ./mpi/mpi-dft.h -1 +2 M ./mpi/mpi-transpose.h -2 +2 M ./mpi/transpose-alltoall.c -3 +4 M ./mpi/transpose-inplace.c -6 +7 M ./mpi/transpose-radix2.c -5 +6 Mon Dec 18 01:02:27 EST 2006 stevenj@fftw.org * ops_add -> ops_add2 where possible, to shrink code M ./mpi/dft-rank-geq2.c -8 +4 M ./mpi/dft-rank1-bigvec.c -4 +2 M ./mpi/transpose-alltoall.c -6 +3 M ./mpi/transpose-inplace.c -8 +4 M ./mpi/transpose-radix2.c -10 +5 M ./rdft/vrank3-transpose.c -9 +5 Mon Dec 18 00:43:02 EST 2006 stevenj@fftw.org * added dft-rank1-bigvec solver (easy case for 1d parallel transforms) M ./mpi/Makefile.am -1 +1 M ./mpi/api.c -2 +14 M ./mpi/conf.c +1 A ./mpi/dft-rank1-bigvec.c M ./mpi/mpi-dft.h +1 Sun Dec 17 20:42:21 EST 2006 stevenj@fftw.org * rewrote MPI stuff to use dtensor data structure A dtensor is an ordered tuple of triplets (n, ib, ob) giving the size of a dimension (n) and its input and output block sizes of a distributed row-major multi-dimensional array. An MPI DFT (etc.) is now specified in terms of dtensors, which provide a much more flexible data layout. For example, we can now describe multidimensional block distributions, which are important if the number of processors is greater than the size of any given dimension. Currently, we only have solvers for 1d slab distributions, and this is all that is supported in the basic and advanced APIs. The guru API allows one to specify more general distributions, however, which will be useful when/if we have solvers for this case. We now also don't need a TRANSPOSED flag, at least internally, since TRANSPOSED multi-dimensional DFT plans just correspond to dtensors where the input and output block distributions are different. Other changes include the use of the XM(foo) macro for X(mpi_foo). M ./mpi/Makefile.am -1 +1 M ./mpi/any-true.c -1 +1 M ./mpi/api.c -129 +335 M ./mpi/block.c -32 +95 M ./mpi/conf.c -6 +6 M ./mpi/dft-problem.c -46 +33 M ./mpi/dft-rank-geq2.c -60 +72 M ./mpi/dft-serial.c -26 +20 M ./mpi/dft-solve.c -1 +1 A ./mpi/dtensor.c M ./mpi/fftw3-mpi.h -24 +38 M ./mpi/ifftw-mpi.h -22 +60 M ./mpi/mpi-bench.c -115 +289 M ./mpi/mpi-dft.h -12 +13 M ./mpi/mpi-transpose.h -10 +11 M ./mpi/transpose-alltoall.c -17 +16 M ./mpi/transpose-inplace.c -24 +21 M ./mpi/transpose-problem.c -19 +18 M ./mpi/transpose-radix2.c -13 +11 M ./mpi/transpose-solve.c -1 +1 M ./tests/bench.c +4 M ./tests/fftw-bench.c +2 M ./tests/fftw-bench.h +1 Fri Dec 15 16:01:23 EST 2006 athena@fftw.org * Distinguished mutexes from semaphores. The distinction is useful because the linux implementation of sem_post() in unnecessarily slow when semaphores are used for mutual exclusion. This change made spinlocks messier to implement, so I excised them. M ./tests/fftw-bench.c -8 M ./threads/threads.c -105 +40 Fri Dec 15 12:46:11 EST 2006 athena@fftw.org * Use posix semaphores where available. Paranoid declaration of all shared variables as ``volatile''. Paranoid initialization of all shared variables within locks. M ./threads/threads.c -36 +68 Wed Dec 13 20:12:13 EST 2006 Matteo Frigo * paranoia M ./threads/threads.c -2 +2 Tue Dec 12 17:28:13 EST 2006 stevenj@fftw.org * punt on detecting unsolvable rdft2 problems; make r==iio rdft2 problems unsolvable, since it doesn't look like we've consistently checked for this case and it's not clear why we would want to support it (it was also not documented in the manual) M ./doc/fftw3.texi -1 +2 M ./rdft/problem2.c -30 +2 Sat Dec 9 12:14:13 EST 2006 athena@fftw.org * Obey stupid const rules M ./api/api.h -2 +2 M ./api/apiplan.c -1 +1 M ./dft/dft.h -3 +3 M ./dft/indirect.c -4 +4 M ./dft/problem.c -9 +8 M ./kernel/ifftw.h -5 +5 M ./kernel/planner.c -2 +2 M ./kernel/problem.c -4 +4 M ./rdft/buffered2.c -1 +1 M ./rdft/dht-rader.c -1 +1 M ./rdft/indirect.c -4 +4 M ./rdft/problem.c -14 +13 M ./rdft/problem2.c -9 +8 M ./rdft/rank-geq2-rdft2.c -1 +1 M ./rdft/rdft-dht.c -1 +1 M ./rdft/rdft.h -13 +13 M ./rdft/rdft2-radix2.c -5 +5 Fri Dec 8 18:21:50 EST 2006 stevenj@fftw.org * added unsolvable check for rdft2 problem An in-place rdft2 problem is ill-formed if the real data, including the extra "padding" elements, do not coincide with the complex data. CHANGE: the new code considers all in-place split r2c and c2r problems to be ill-formed. Previously, these could be done, but only if the entire multi-dimensional array fit into the buffer, which is kind of stupid. I'm not sure it's worth it to even try to support the split in-place r2c case. M ./rdft/problem2.c -1 +33 Fri Dec 8 13:47:53 EST 2006 stevenj@fftw.org * check in-placeness after joining taints M ./dft/problem.c -9 +9 Fri Dec 8 13:43:44 EST 2006 athena@fftw.org * Grrr... paranoid-check was not testing in exhaustive mode M ./tests/Makefile.am -5 +5 Fri Dec 8 10:00:30 EST 2006 Matteo Frigo * Implemented PROBLEM_UNSOLVABLE. In-place DFT and RDFT problems with inconsistent I/O strides are now unsolvable, and we don't check for them any longer in solvers. While I was at it, declared all problem pointers to be ``const'' for extra safety. M ./api/api.h -2 +2 M ./api/apiplan.c -3 +5 M ./dft/buffered.c -4 +1 M ./dft/dft.h -4 +4 M ./dft/direct.c -8 +6 M ./dft/indirect.c -4 +4 M ./dft/problem.c -11 +15 M ./dft/vrank-geq1.c -3 M ./kernel/ifftw.h -6 +10 M ./kernel/planner.c -6 +6 M ./kernel/problem.c -1 +40 M ./rdft/buffered.c -4 +1 M ./rdft/buffered2.c -1 +1 M ./rdft/dht-rader.c -1 +1 M ./rdft/direct.c -4 +3 M ./rdft/indirect.c -4 +4 M ./rdft/problem.c -14 +17 M ./rdft/problem2.c -10 +10 M ./rdft/rank-geq2-rdft2.c -1 +1 M ./rdft/rdft-dht.c -1 +1 M ./rdft/rdft.h -13 +13 M ./rdft/rdft2-radix2.c -5 +5 M ./rdft/vrank-geq1.c -2 Thu Dec 7 20:13:46 EST 2006 Matteo Frigo * Avoid qsort'ing one element. M ./kernel/tensor7.c -6 +2 Thu Dec 7 18:25:47 EST 2006 Matteo Frigo * In-place vrank>=1 is now applicable only if the problem is really in-place. M ./dft/buffered.c -7 +12 M ./dft/direct.c -11 +11 M ./dft/vrank-geq1.c +3 M ./kernel/ifftw.h +2 M ./kernel/tensor7.c +41 M ./rdft/buffered.c -7 +12 M ./rdft/direct.c -7 +5 M ./rdft/vrank-geq1.c +2 Tue Dec 5 12:21:38 EST 2006 athena@fftw.org * unused variable M ./kernel/align.c +1 Sun Dec 3 19:16:33 EST 2006 Matteo Frigo * Removed CVS $Id$ everywhere, since darcs does not update them. M ./api/fftw3.h -1 M ./api/version.c -1 M ./commercialize.sh +1 M ./dft/buffered.c -1 M ./dft/codelet-dft.h -1 M ./dft/codelets/standard/Makefile.am +1 M ./dft/conf.c -1 M ./dft/ct.c -1 M ./dft/dft.h -1 M ./dft/dftw-direct.c -1 M ./dft/dftw-directsq.c -1 M ./dft/direct.c -1 M ./dft/indirect.c -1 M ./dft/kdft-dif.c -1 M ./dft/kdft-difsq.c -1 M ./dft/kdft-dit.c -1 M ./dft/kdft.c -1 M ./dft/nop.c -1 M ./dft/plan.c -1 M ./dft/problem.c -1 M ./dft/rank-geq2.c -1 M ./dft/solve.c -1 M ./dft/vrank-geq1.c -1 M ./dft/zero.c -1 M ./doc/fftw3.texi -1 M ./genfft/algsimp.ml -2 M ./genfft/algsimp.mli -2 M ./genfft/annotate.ml -2 M ./genfft/annotate.mli -1 M ./genfft/assoctable.ml -1 M ./genfft/assoctable.mli -1 M ./genfft/c.ml -1 M ./genfft/c.mli -1 M ./genfft/complex.ml -1 M ./genfft/complex.mli -1 M ./genfft/conv.mli -1 M ./genfft/dag.ml -1 M ./genfft/dag.mli -1 M ./genfft/expr.ml -1 M ./genfft/expr.mli -1 M ./genfft/fft.ml -2 M ./genfft/fft.mli -2 M ./genfft/gen_conv.ml -3 +1 M ./genfft/gen_hc2hc.ml -3 +1 M ./genfft/gen_hc2r.ml -3 +1 M ./genfft/gen_mdct.ml -3 +1 M ./genfft/gen_notw.ml -3 +1 M ./genfft/gen_notw_c.ml -3 +1 M ./genfft/gen_r2hc.ml -3 +1 M ./genfft/gen_r2r.ml -3 +1 M ./genfft/gen_twiddle.ml -3 +1 M ./genfft/gen_twiddle_c.ml -3 +1 M ./genfft/gen_twidsq.ml -3 +1 M ./genfft/gen_twidsq_c.ml -3 +1 M ./genfft/genutil.ml -8 +1 M ./genfft/littlesimp.ml -1 M ./genfft/littlesimp.mli -1 M ./genfft/magic.ml -1 M ./genfft/monads.ml -1 M ./genfft/number.ml -1 M ./genfft/number.mli -1 M ./genfft/oracle.ml -1 M ./genfft/oracle.mli -1 M ./genfft/schedule.ml -1 M ./genfft/schedule.mli -1 M ./genfft/simd.ml -1 M ./genfft/simd.mli -1 M ./genfft/simdmagic.ml -1 M ./genfft/to_alist.ml -1 M ./genfft/to_alist.mli -1 M ./genfft/trig.ml -1 M ./genfft/trig.mli -1 M ./genfft/twiddle.ml -1 M ./genfft/twiddle.mli -1 M ./genfft/unique.ml -1 M ./genfft/unique.mli -1 M ./genfft/util.ml -1 M ./genfft/util.mli -1 M ./genfft/variable.ml -1 M ./genfft/variable.mli -1 M ./kernel/align.c -1 M ./kernel/alloc.c -1 M ./kernel/assert.c -1 M ./kernel/awake.c -1 M ./kernel/cycle.h -1 M ./kernel/debug.c -1 M ./kernel/iabs.c -1 M ./kernel/ifftw.h -1 M ./kernel/kalloc.c -1 M ./kernel/minmax.c -1 M ./kernel/ops.c -1 M ./kernel/pickdim.c -1 M ./kernel/plan.c -1 M ./kernel/planner.c -1 M ./kernel/primes.c -1 M ./kernel/print.c -1 M ./kernel/problem.c -1 M ./kernel/scan.c -1 M ./kernel/solver.c -1 M ./kernel/solvtab.c -1 M ./kernel/stride.c -1 M ./kernel/tensor.c -1 M ./kernel/tensor1.c -1 M ./kernel/tensor2.c -1 M ./kernel/tensor4.c -1 M ./kernel/tensor5.c -1 M ./kernel/tensor7.c -1 M ./kernel/tensor8.c -1 M ./kernel/tensor9.c -1 M ./kernel/timer.c -1 M ./kernel/trig.c -1 M ./kernel/twiddle.c -1 M ./libbench/accopy-from.c -1 M ./libbench/accopy-to.c -1 M ./libbench/allocate.c -1 M ./libbench/bench-main.c -1 M ./libbench/bench-user.h -1 M ./libbench/bench.h -1 M ./libbench/can-do.c -1 M ./libbench/ccopy-from.c -1 M ./libbench/ccopy-to.c -1 M ./libbench/deallocate.c -1 M ./libbench/getopt-utils.c -1 M ./libbench/info.c -1 M ./libbench/main.c -1 M ./libbench/prime.c -1 M ./libbench/problem.c -1 M ./libbench/report.c -1 M ./libbench/speed.c -1 M ./libbench/timer.c -1 M ./libbench/verify.c -1 M ./libbench/zero.c -1 M ./libbench2/aligned-main.c -1 M ./libbench2/allocate.c -1 M ./libbench2/bench-main.c -1 M ./libbench2/bench-user.h -1 M ./libbench2/bench.h -1 M ./libbench2/can-do.c -1 M ./libbench2/dotens2.c -1 M ./libbench2/info.c -1 M ./libbench2/main.c -1 M ./libbench2/problem.c -1 M ./libbench2/report.c -1 M ./libbench2/speed.c -1 M ./libbench2/tensor.c -1 M ./libbench2/timer.c -1 M ./libbench2/useropt.c -1 M ./libbench2/verify-dft.c -1 M ./libbench2/verify-lib.c -1 M ./libbench2/verify-rdft2.c -1 M ./libbench2/verify.c -1 M ./libbench2/zero.c -1 M ./mpi/conf.c -1 M ./rdft/buffered.c -1 M ./rdft/buffered2.c -1 M ./rdft/codelets/hc2r/Makefile.am +1 M ./rdft/codelets/r2hc/Makefile.am +1 M ./rdft/conf.c -1 M ./rdft/dft-r2hc.c -1 M ./rdft/dht-r2hc.c -1 M ./rdft/direct.c -1 M ./rdft/direct2.c -1 M ./rdft/hc2hc-direct.c -1 M ./rdft/hc2hc-directbuf.c -1 M ./rdft/indirect.c -1 M ./rdft/khc2hc.c -1 M ./rdft/khc2r.c -1 M ./rdft/kr2hc.c -1 M ./rdft/kr2r.c -1 M ./rdft/nop.c -1 M ./rdft/nop2.c -1 M ./rdft/plan.c -1 M ./rdft/plan2.c -1 M ./rdft/problem.c -1 M ./rdft/problem2.c -1 M ./rdft/rank-geq2-rdft2.c -1 M ./rdft/rank-geq2.c -1 M ./rdft/rank0-rdft2.c -1 M ./rdft/rank0.c -1 M ./rdft/rdft-dht.c -1 M ./rdft/rdft2-inplace-strides.c -1 M ./rdft/rdft2-radix2.c -1 M ./rdft/rdft2-tensor-max-index.c -1 M ./rdft/solve.c -1 M ./rdft/solve2.c -1 M ./rdft/vrank-geq1-rdft2.c -1 M ./rdft/vrank-geq1.c -1 M ./rdft/vrank3-transpose.c -1 M ./reodft/conf.c -1 M ./reodft/redft00e-r2hc-pad.c -1 M ./reodft/redft00e-r2hc.c -1 M ./reodft/reodft00e-splitradix.c -1 M ./reodft/reodft010e-r2hc.c -1 M ./reodft/reodft11e-r2hc-odd.c -1 M ./reodft/reodft11e-r2hc.c -1 M ./reodft/reodft11e-radix2.c -1 M ./reodft/rodft00e-r2hc-pad.c -1 M ./reodft/rodft00e-r2hc.c -1 M ./simd/altivec.c -1 M ./simd/nonportable/sse.c -1 M ./simd/nonportable/sse2.c -1 M ./simd/sse.c -1 M ./simd/sse2.c -1 M ./simd/taint.c -1 M ./simd/x86-cpuid.h -1 M ./threads/conf.c -1 M ./threads/ct.c -1 M ./threads/dft-vrank-geq1.c -1 M ./threads/rdft-vrank-geq1.c -1 M ./threads/vrank-geq1-rdft2.c -1 Sun Dec 3 16:11:17 EST 2006 Matteo Frigo * generalized dftw to encompass q codelets. As a side effect, q codelets are now threaded. R ./dft/ctsq.c A ./dft/dftw-directsq.c M ./dft/Makefile.am -2 +2 M ./dft/ct.c -25 +55 M ./dft/ct.h -6 +9 M ./dft/dftw-direct.c -59 +72 M ./dft/dftw-directsq.c +163 M ./dft/dftw-generic.c -24 +32 M ./dft/dftw-genericbuf.c -30 +37 M ./dft/indirect-transpose.c -1 +1 M ./dft/kdft-difsq.c -1 +1 M ./libbench2/my-getopt.c -3 +3 M ./threads/ct.c -28 +53 Sat Nov 25 16:34:38 EST 2006 stevenj@fftw.org * add missing __declspec attribute to threads API functions when compiling for Windows (thanks to Robert O. Morris for the bug report) M ./api/f77api.c -29 +1 M ./api/x77.h +28 M ./threads/f77api.c +13 M ./threads/f77funcs.h -3 +3 Mon Nov 20 17:39:20 EST 2006 stevenj@fftw.org * add AC_SUBST to AX_OPENMP, thanks to Sebastien Maret for the suggestion M ./m4/ax_openmp.m4 -1 +2 Mon Oct 16 23:02:29 EDT 2006 stevenj@fftw.org * not gcc bug for MIPS (thanks to Jonathan Day) M ./doc/FAQ/fftw-faq.bfnn +4 Sat Sep 23 17:52:36 EDT 2006 stevenj@fftw.org * in maintainer/debug mode, don't modify CFLAGS if they were explicitly set (-pedantic seems to cause problems with LAM's mpicc, so I need a way to override) M ./configure.ac +2 Wed Nov 22 18:14:47 EST 2006 athena@fftw.org * Removed obsolete comment. M ./dft/dftw-generic.c -1 Sun Nov 19 11:21:44 EST 2006 athena@fftw.org * Use p->v when comparing TW_FULL fields. M ./kernel/twiddle.c -2 +1 Thu Nov 16 14:49:05 EST 2006 athena@fftw.org * removed useless definition M ./simd/simd.h -1 Mon Nov 13 09:18:32 EST 2006 athena@fftw.org * paranoid avoidance of integer overflows M ./libbench2/timer.c -3 +3 Mon Nov 13 09:00:11 EST 2006 athena@fftw.org * avoid potential overflows in cycle counters At the suggestion of Alex Cichowski, convert all ticks to double before operating on them, to avoid potential signed/unsigned confusion and integer overflow. M ./kernel/cycle.h -5 +8 Sun Nov 5 09:00:52 EST 2006 Matteo Frigo * Removed unused struct field M ./threads/threads.c -1 Sat Nov 4 09:43:13 EST 2006 Matteo Frigo * use pthread condition variables instead of semaphores Condition variables are more likely to be portable everywhere, and somehow they appear to introduce less overhead at least on my linux box. M ./threads/threads.c -6 +19 Tue Oct 31 20:45:24 EST 2006 Matteo Frigo * Imprecise help message. M ./libbench2/my-getopt.c -1 +1 Mon Oct 30 20:13:35 EST 2006 athena@fftw.org * Experimental implementation of spinlocks. This patch implements spinlocks via a semi-portable hack, and adds the -ospinlocks option to the bench program so that we can play with them. M ./tests/fftw-bench.c +8 M ./threads/threads.c -28 +104 Thu Oct 26 22:29:18 EDT 2006 athena@fftw.org * Updated manual for new openmp configure options. M ./doc/fftw3.texi -20 +17 Thu Oct 26 21:52:39 EDT 2006 athena@fftw.org * Added back openmp. Rationalized threads naming conventions: * threads explicitly managed by us are enabled by --enable-threads, predicated on HAVE_THREADS, etc. * openmp is enabled by --enable-openmp, predicated on HAVE_OPENMP, etc. * SMP denotes either THREADS or OPENMP. M ./Makefile.am -1 +1 M ./configure.ac -10 +27 M ./kernel/alloc.c -1 +1 M ./kernel/ifftw.h -1 +1 M ./mpi/Makefile.am -1 +1 M ./tests/Makefile.am -5 +5 M ./tests/fftw-bench.c -3 +7 M ./threads/Makefile.am -2 +2 M ./threads/api.c -1 +16 M ./threads/conf.c -2 +2 A ./threads/openmp.c M ./threads/threads.c -13 +19 M ./threads/threads.h -2 +2 M ./tools/Makefile.am -1 +1 M ./tools/fftw-wisdom.c -5 +9 Mon Oct 23 20:14:31 EDT 2006 athena@fftw.org * different thread protocols M ./threads/threads.c -80 +71 Sun Oct 22 14:49:32 EDT 2006 athena@fftw.org * fix memory leak Added pthread_attr_destroy to avoid memory leak. M ./threads/threads.c +1 Sun Oct 22 14:23:30 EDT 2006 athena@fftw.org * Experimental new pthread implementation that recycles threads. M ./threads/threads.c -528 +155 Tue Oct 24 23:28:10 EDT 2006 athena@fftw.org * switched buddies Switched order of buddies in rdft2 rank-geq2 for consistency with analogous dft and rdft solvers. Furthermore, this change reduces the MEASURE planning time for rank == 3. M ./rdft/rank-geq2-rdft2.c -4 +1 Sat Oct 21 10:10:00 EDT 2006 Matteo Frigo * Typo. M ./tools/fftw_wisdom.1.in -1 +1 Sat Oct 21 09:56:50 EDT 2006 Matteo Frigo * Out of place is the default. Thanks to Kirk Kern for pointing this out. M ./tests/README -2 +2 Fri Sep 29 01:36:11 EDT 2006 stevenj@fftw.org * rename "test" to "tst", since a user (Igor Levicki) reports that "test" is a reserved words in some x86 assemblers M ./simd/x86-cpuid.h -10 +10 Tue Sep 26 09:01:08 EDT 2006 Matteo Frigo * Stylistic change. M ./kernel/timer.c -2 +1 Tue Sep 26 08:45:37 EDT 2006 athena@fftw.org * Do not set tmin=1e10, since a large FFT may take longer than that. M ./kernel/timer.c -5 +4 Sat Sep 23 22:07:10 EDT 2006 athena@fftw.org * Disable certain gcc optimizations. When PRECOMPUTE_ARRAY_INDICES is #define'd, array indices have the form array[stride[k]] for compile-time constant k. Apparently new gcc's copy stride[k] onto the stack before the codelet loop, which is an idiotic optimization if ever there was one. This patch confuses gcc enough to prevent this optimization. M ./kernel/ifftw.h -3 +6 M ./kernel/stride.c -1 +1 Sat Sep 23 13:02:58 EDT 2006 stevenj@fftw.org * re-enable TOMS algorithm - it is the best for large vector lengths, since for such sizes the cache line is not an issue and the bookkeeping overhead is negligible M ./rdft/vrank3-transpose.c -18 +15 Thu Sep 21 15:40:15 EDT 2006 stevenj@fftw.org * add measure_hook so that MPI can synchronize timing measurements (otherwise different processors might end up with different MPI plans, yikes!) M ./kernel/ifftw.h +2 M ./kernel/planner.c +5 M ./mpi/api.c -2 +26 Tue Sep 19 21:26:19 EDT 2006 stevenj@fftw.org * added O(p log p) transpose algorithm (radix 2) M ./mpi/Makefile.am -1 +1 M ./mpi/conf.c +1 M ./mpi/mpi-transpose.h +1 A ./mpi/transpose-radix2.c Tue Sep 19 21:05:09 EDT 2006 stevenj@fftw.org * comments M ./mpi/testsched.c -8 +10 Tue Sep 19 19:54:58 EDT 2006 stevenj@fftw.org * whoops ./mpi/any_true.c -> ./mpi/any-true.c M ./mpi/Makefile.am -1 +1 Tue Sep 19 19:49:01 EDT 2006 stevenj@fftw.org * synchronize planning so that if one process fails to create a plan then all of them do. M ./mpi/Makefile.am -1 +1 A ./mpi/any_true.c M ./mpi/dft-rank-geq2.c -4 +4 M ./mpi/dft-serial.c -1 +1 M ./mpi/ifftw-mpi.h +3 M ./mpi/transpose-alltoall.c -5 +5 M ./mpi/transpose-inplace.c -5 +5 Tue Sep 19 18:17:38 EDT 2006 stevenj@fftw.org * call MPI_Alltoall instead of MPI_Alltoallv for equal-blocks case, in case MPI implementation has special optimizations for the common case of equal sizes M ./mpi/transpose-alltoall.c -6 +19 Tue Sep 19 12:07:35 EDT 2006 stevenj@fftw.org * whoops M ./tools/Makefile.am -1 +1 Tue Sep 19 02:20:06 EDT 2006 stevenj@fftw.org * typo in comment M ./mpi/testsched.c -3 +3 Tue Sep 19 01:58:55 EDT 2006 stevenj@fftw.org * more filename simplifications ./mpi/mpi-dft-rank-geq2.c -> ./mpi/dft-rank-geq2.c ./mpi/mpi-dft-serial.c -> ./mpi/dft-serial.c ./mpi/mpi-transpose-alltoall.c -> ./mpi/transpose-alltoall.c ./mpi/mpi-transpose-inplace.c -> ./mpi/transpose-inplace.c ./mpi/problem.c -> ./mpi/dft-problem.c ./mpi/solve.c -> ./mpi/dft-solve.c ./mpi/tproblem.c -> ./mpi/transpose-problem.c ./mpi/tsolve.c -> ./mpi/transpose-solve.c M ./mpi/Makefile.am -3 +3 Tue Sep 19 01:49:52 EDT 2006 stevenj@fftw.org * canonicalize file names (hyphens, not underscores) ./mpi/fftw3_mpi.h -> ./mpi/fftw3-mpi.h ./mpi/ifftw_mpi.h -> ./mpi/ifftw-mpi.h ./mpi/mpi_bench.c -> ./mpi/mpi-bench.c ./mpi/mpi_dft.h -> ./mpi/mpi-dft.h ./mpi/mpi_transpose.h -> ./mpi/mpi-transpose.h ./mpi/mpi_transpose_alltoall.c -> ./mpi/mpi-transpose-alltoall.c ./mpi/mpi_transpose_inplace.c -> ./mpi/mpi-transpose-inplace.c ./tests/fftw_bench.h -> ./tests/fftw-bench.h ./tests/fftw_bench_common.c -> ./tests/fftw-bench.c M ./mpi/Makefile.am -5 +5 M ./mpi/api.c -4 +4 M ./mpi/block.c -1 +1 M ./mpi/conf.c -2 +2 M ./mpi/mpi-bench.c -2 +2 M ./mpi/mpi-dft-rank-geq2.c -2 +2 M ./mpi/mpi-dft-serial.c -1 +1 M ./mpi/mpi-dft.h -1 +1 M ./mpi/mpi-transpose-alltoall.c -1 +1 M ./mpi/mpi-transpose-inplace.c -1 +1 M ./mpi/mpi-transpose.h -1 +1 M ./mpi/problem.c -1 +1 M ./mpi/solve.c -1 +1 M ./mpi/tproblem.c -1 +1 M ./mpi/tsolve.c -1 +1 M ./tests/Makefile.am -1 +1 M ./tests/bench.c -2 +2 M ./tests/fftw-bench.c -1 +1 Tue Sep 19 01:34:07 EDT 2006 stevenj@fftw.org * add mpi-dft-serial M ./mpi/Makefile.am -1 +1 M ./mpi/conf.c +1 M ./mpi/mpi-dft-rank-geq2.c -3 +9 A ./mpi/mpi-dft-serial.c M ./mpi/mpi_dft.h +1 M ./mpi/mpi_transpose_alltoall.c -1 +1 M ./mpi/mpi_transpose_inplace.c -1 +1 Tue Sep 19 01:21:47 EDT 2006 stevenj@fftw.org * silence warnings M ./dft/indirect-transpose.c -1 M ./libbench2/problem.c -2 Tue Sep 19 00:31:59 EDT 2006 stevenj@fftw.org * make "t" problem semantics match FFTW_MPI_TRANSPOSED M ./libbench2/problem.c -4 +3 M ./mpi/mpi_bench.c -8 +18 Mon Sep 18 23:50:43 EDT 2006 stevenj@fftw.org * whoops, fixed backwards mpi_dft M ./mpi/mpi-dft-rank-geq2.c -7 +7 Mon Sep 18 22:26:31 EDT 2006 stevenj@fftw.org * initial stab at rank-geq2 mpi-dft; seems to be mostly working M ./libbench2/bench-user.h -1 +1 M ./mpi/Makefile.am -1 +4 M ./mpi/api.c -26 +180 M ./mpi/conf.c +2 M ./mpi/fftw3_mpi.h -11 +51 M ./mpi/ifftw_mpi.h -4 +1 A ./mpi/mpi-dft-rank-geq2.c M ./mpi/mpi_bench.c -55 +166 M ./mpi/mpi_dft.h -2 +2 M ./mpi/mpi_transpose.h -6 +1 M ./mpi/mpi_transpose_alltoall.c -7 +10 M ./mpi/mpi_transpose_inplace.c -7 +8 M ./mpi/problem.c -2 +7 A ./mpi/solve.c M ./mpi/tproblem.c -2 +2 Sun Sep 17 13:41:32 EDT 2006 stevenj@fftw.org * support SCRAMBLED_OUT in alltoall transpose M ./kernel/ifftw.h +4 M ./kernel/tensor2.c +20 M ./mpi/ifftw_mpi.h -1 +16 M ./mpi/mpi_dft.h -1 +1 M ./mpi/mpi_transpose_alltoall.c -23 +48 M ./mpi/mpi_transpose_inplace.c -23 +2 M ./mpi/problem.c -2 +3 Sun Sep 17 12:34:30 EDT 2006 stevenj@fftw.org * skeleton of future support for block-cyclic M ./mpi/block.c -7 +27 M ./mpi/ifftw_mpi.h +1 M ./mpi/mpi_transpose_alltoall.c +2 M ./mpi/mpi_transpose_inplace.c -2 +5 Sun Sep 17 12:11:19 EDT 2006 stevenj@fftw.org * test program now checks scrambled in/out via -obflag=28/29 M ./mpi/api.c -1 +1 M ./mpi/fftw3_mpi.h -1 +1 M ./mpi/mpi_bench.c -13 +59 M ./mpi/problem.c -1 Sun Sep 17 11:58:36 EDT 2006 stevenj@fftw.org * added -obflag to make it easier to set high-order bits M ./tests/fftw_bench_common.c +1 Sun Sep 17 01:30:51 EDT 2006 stevenj@fftw.org * use proper child plans for 2nd transpose in transpose_alltoall; implement opcount in transpose_inplace M ./mpi/mpi_transpose_alltoall.c -47 +47 M ./mpi/mpi_transpose_inplace.c -1 +10 Sun Sep 17 01:08:01 EDT 2006 stevenj@fftw.org * fix in test program for transposes of vectors -- transpose routines seem to completely work now (except for scrambled in/out, which is untested) M ./mpi/mpi_bench.c -1 +1 Sun Sep 17 01:01:16 EDT 2006 stevenj@fftw.org * fixed bug in transpose_alltoall for unequal blocks M ./mpi/mpi_transpose_alltoall.c -2 +2 M ./mpi/mpi_transpose_inplace.c -2 +2 Sat Sep 16 15:29:46 EDT 2006 stevenj@fftw.org * correctly handle cld2rest M ./mpi/mpi_transpose_inplace.c -24 +28 Sat Sep 16 15:29:31 EDT 2006 stevenj@fftw.org * some debugging code and other fixes M ./mpi/mpi_bench.c -13 +34 Sat Sep 16 14:54:30 EDT 2006 stevenj@fftw.org * whoops, forgot to check in mpi_bench.c file A ./mpi/mpi_bench.c Sat Sep 16 14:54:02 EDT 2006 stevenj@fftw.org * added bench_exit routine so that it can be overridden (by MPI_Abort) if needed M ./libbench2/Makefile.am -5 +5 A ./libbench2/bench-exit.c M ./libbench2/bench-user.h +1 M ./libbench2/util.c -1 +1 M ./libbench2/verify-lib.c -1 +1 M ./libbench2/verify-r2r.c -1 +1 Sat Sep 16 14:52:56 EDT 2006 stevenj@fftw.org * bug fix in mpi_transpose_inplace for case where some processors are idle M ./mpi/mpi_transpose_inplace.c -3 +4 Fri Sep 15 18:47:13 EDT 2006 stevenj@fftw.org * allow vecloop for sz->rnk==0 in exceptional (SLOW) cases, e.g. it is necessary for loops of non-square transposes (otherwise e.g. ik1v5:200:200x10:20:1x20:1:10 planning fails) M ./rdft/vrank-geq1.c -2 +6 Thu Sep 14 23:36:48 EDT 2006 stevenj@fftw.org * first pass at working mpi_bench test program; transpose seems to work iff dimensions are divisible by #processors M ./libbench2/Makefile.am -1 +3 A ./libbench2/after-ccopy-from.c A ./libbench2/after-ccopy-to.c A ./libbench2/after-hccopy-from.c A ./libbench2/after-hccopy-to.c A ./libbench2/after-rcopy-from.c A ./libbench2/after-rcopy-to.c M ./libbench2/bench-main.c -4 +6 M ./libbench2/bench-user.h +10 M ./libbench2/verify-dft.c +2 M ./libbench2/verify-r2r.c +4 M ./libbench2/verify-rdft2.c +4 M ./mpi/api.c -2 +6 Wed Sep 13 17:28:07 EDT 2006 stevenj@fftw.org * whoops M ./mpi/mpi_transpose_inplace.c -2 +2 Tue Sep 12 22:27:03 EDT 2006 stevenj@fftw.org * initial stub for mpi_bench M ./Makefile.am -2 +2 M ./libbench2/bench-main.c +3 M ./libbench2/bench-user.h +1 M ./libbench2/verify-lib.c -1 +1 M ./mpi/Makefile.am -1 +14 M ./tests/bench.c +9 M ./tests/fftw_bench.h +1 M ./tests/fftw_bench_common.c +2 Tue Sep 12 21:54:31 EDT 2006 stevenj@fftw.org * do no output at all if verbose < 0 (for use with MPI, where we only want output from process 0) M ./libbench2/bench-main.c -1 +1 M ./libbench2/bench.h +1 M ./libbench2/ovtpvt.c -1 +15 M ./libbench2/useropt.c -1 +1 M ./libbench2/util.c -2 +1 M ./libbench2/verify-lib.c -5 +4 M ./libbench2/verify-r2r.c -4 +4 Tue Sep 12 21:39:15 EDT 2006 stevenj@fftw.org * whoops M ./tests/fftw_bench_common.c +3 Tue Sep 12 21:31:40 EDT 2006 stevenj@fftw.org * split bench.c into bench.c and fftw_bench_common.c so that we can re-use some of the code in the MPI test program M ./tests/Makefile.am -1 +1 M ./tests/bench.c -210 +6 A ./tests/fftw_bench.h A ./tests/fftw_bench_common.c M ./tools/Makefile.am -1 +2 Tue Sep 12 21:00:36 EDT 2006 stevenj@fftw.org * MPI stuff at least compiles now M ./configure.ac -9 +11 M ./kernel/ifftw.h +7 M ./mpi/api.c -3 +4 M ./mpi/fftw3_mpi.h +1 M ./mpi/mpi_transpose_alltoall.c -10 +12 M ./mpi/mpi_transpose_inplace.c -11 +14 M ./mpi/problem.c -5 +4 M ./mpi/tproblem.c -4 +3 Mon Sep 11 22:26:36 EDT 2006 stevenj@fftw.org * initial (nonfunctional) start at MPI support (similar to FFTW 2.x in spirit, but mostly rewritten) A ./mpi/ M ./Makefile.am -1 +1 M ./configure.ac +11 A ./m4/acx_mpi.m4 A ./mpi/Makefile.am A ./mpi/api.c A ./mpi/block.c A ./mpi/conf.c A ./mpi/fftw3_mpi.h A ./mpi/ifftw_mpi.h A ./mpi/mpi_dft.h A ./mpi/mpi_transpose.h A ./mpi/mpi_transpose_alltoall.c A ./mpi/mpi_transpose_inplace.c A ./mpi/problem.c A ./mpi/testsched.c A ./mpi/tproblem.c A ./mpi/tsolve.c Mon Sep 11 22:25:38 EDT 2006 stevenj@fftw.org * make X(plan_awake) work for NULL argument to reduce code size M ./kernel/plan.c -3 +5 M ./rdft/vrank3-transpose.c -3 +3 Fri Sep 15 23:47:08 EDT 2006 stevenj@fftw.org * -mt should go before -mthreads to avoid spurious warnings on HPUX (thanks to Peter O'Gorman for the bug report) M ./m4/acx_pthread.m4 -2 +3 Mon Sep 11 13:53:44 EDT 2006 stevenj@fftw.org * Fortran init_threads wrapper didn't return result; thanks to Markus Wetzstein for the bug report M ./threads/f77funcs.h -2 +2 Thu Sep 7 18:43:55 EDT 2006 stevenj@fftw.org * make sure wrappers are included even if Fortran compiler was not detected (unless --disable-fortran was specified explicitly) ... this was supposed to be done before, but the definition was in the wrong place, grr M ./api/f77api.c -21 +21 Thu Aug 31 19:33:29 EDT 2006 stevenj@fftw.org * I'm sick of answering this question about non-deterministic results M ./doc/FAQ/fftw-faq.bfnn +13 Tue Aug 22 21:27:29 EDT 2006 Matteo Frigo * Add --tag=CC flag to libtool. This change is consistent with the libtool invocation in the latest automake, and is required to compile with (some version of) xlc. M ./support/Makefile.codelets -2 +3 Mon Aug 21 21:40:36 EDT 2006 athena@fftw.org * avoid ``fma'' because it is defined in c99. M ./tests/bench.c -3 +3 Sun Aug 20 11:40:53 EDT 2006 Matteo Frigo * Obey -standalone flag. M ./genfft/gen_twiddle.ml -2 +4 Sat Aug 19 13:34:27 EDT 2006 Matteo Frigo * obey -standalone when generating simd codelets M ./genfft/gen_notw_c.ml -2 +1 M ./genfft/gen_twiddle_c.ml -2 +3 Sat Aug 19 13:33:43 EDT 2006 Matteo Frigo * removed obsolete athfft M ./genfft/Makefile.am -10 +9 R ./genfft/gen_athnotw.ml R ./genfft/gen_athtw.ml Thu Aug 17 21:50:50 EDT 2006 stevenj@fftw.org * updated citation to Proc. IEEE paper M ./doc/fftw3.texi -8 +5 Thu Aug 17 21:47:05 EDT 2006 stevenj@fftw.org * use darcs changes --summary to make nice changelog; emacs fill-region hack is obsolete M ./mkdist.sh -3 +1 Mon Aug 14 17:53:19 EDT 2006 athena@fftw.org * removed timer calibration Timer calibration seems not to work any longer on recent processors--- too much noise. I have remove it completely. M ./libbench2/Makefile.am -1 +1 M ./libbench2/bench.h -2 M ./libbench2/timer.c -70 +3 R ./libbench2/timer2.c Mon Aug 14 10:47:15 EDT 2006 Matteo Frigo * removed k7 Removed obsolete k7 support. M ./Makefile.am -10 +2 M ./NEWS +3 M ./TODO -5 M ./api/version.c -3 M ./bootstrap.sh -2 M ./commercialize.sh -8 +1 M ./configure.ac -13 M ./dft/Makefile.am -1 +1 M ./dft/codelet-dft.h -4 M ./dft/conf.c -3 R ./dft/k7/Makefile.am R ./dft/k7/codelets/Makefile.am R ./dft/k7/codelets/ R ./dft/k7/k7.c R ./dft/k7/ M ./doc/fftw3.texi -31 +22 R ./genfft-k7/Makefile.am R ./genfft-k7/algsimp.ml R ./genfft-k7/algsimp.mli R ./genfft-k7/assignmentsToVfpinstrs.ml R ./genfft-k7/assignmentsToVfpinstrs.mli R ./genfft-k7/assoctable.ml R ./genfft-k7/assoctable.mli R ./genfft-k7/balanceVfpinstrs.ml R ./genfft-k7/balanceVfpinstrs.mli R ./genfft-k7/complex.ml R ./genfft-k7/complex.mli R ./genfft-k7/expr.ml R ./genfft-k7/expr.mli R ./genfft-k7/fft.ml R ./genfft-k7/fft.mli R ./genfft-k7/genUtil.ml R ./genfft-k7/gen_notw.ml R ./genfft-k7/gen_twiddle.ml R ./genfft-k7/id.ml R ./genfft-k7/id.mli R ./genfft-k7/k7Basics.ml R ./genfft-k7/k7Basics.mli R ./genfft-k7/k7ExecutionModel.ml R ./genfft-k7/k7ExecutionModel.mli R ./genfft-k7/k7FlatInstructionScheduling.ml R ./genfft-k7/k7FlatInstructionScheduling.mli R ./genfft-k7/k7InstructionSchedulingBasics.ml R ./genfft-k7/k7InstructionSchedulingBasics.mli R ./genfft-k7/k7RegisterAllocationBasics.ml R ./genfft-k7/k7RegisterAllocationBasics.mli R ./genfft-k7/k7RegisterAllocator.ml R ./genfft-k7/k7RegisterAllocator.mli R ./genfft-k7/k7RegisterAllocatorEATranslation.ml R ./genfft-k7/k7RegisterAllocatorEATranslation.mli R ./genfft-k7/k7RegisterAllocatorInit.ml R ./genfft-k7/k7RegisterAllocatorInit.mli R ./genfft-k7/k7RegisterReallocation.ml R ./genfft-k7/k7RegisterReallocation.mli R ./genfft-k7/k7Translate.ml R ./genfft-k7/k7Translate.mli R ./genfft-k7/k7Unparsing.ml R ./genfft-k7/k7Unparsing.mli R ./genfft-k7/k7Vectorization.ml R ./genfft-k7/k7Vectorization.mli R ./genfft-k7/littlesimp.ml R ./genfft-k7/littlesimp.mli R ./genfft-k7/magic.ml R ./genfft-k7/memoMonad.ml R ./genfft-k7/memoMonad.mli R ./genfft-k7/monads.ml R ./genfft-k7/nonDetMonad.ml R ./genfft-k7/nonDetMonad.mli R ./genfft-k7/nullVectorization.ml R ./genfft-k7/nullVectorization.mli R ./genfft-k7/number.ml R ./genfft-k7/number.mli R ./genfft-k7/oracle.ml R ./genfft-k7/oracle.mli R ./genfft-k7/stateMonad.ml R ./genfft-k7/stateMonad.mli R ./genfft-k7/to_alist.ml R ./genfft-k7/to_alist.mli R ./genfft-k7/twiddle.ml R ./genfft-k7/twiddle.mli R ./genfft-k7/util.ml R ./genfft-k7/util.mli R ./genfft-k7/vAnnotatedScheduler.ml R ./genfft-k7/vAnnotatedScheduler.mli R ./genfft-k7/vDag.ml R ./genfft-k7/vDag.mli R ./genfft-k7/vFpBasics.ml R ./genfft-k7/vFpBasics.mli R ./genfft-k7/vFpUnparsing.ml R ./genfft-k7/vFpUnparsing.mli R ./genfft-k7/vImproveSchedule.ml R ./genfft-k7/vImproveSchedule.mli R ./genfft-k7/vK7Optimization.ml R ./genfft-k7/vK7Optimization.mli R ./genfft-k7/vScheduler.ml R ./genfft-k7/vScheduler.mli R ./genfft-k7/vSimdBasics.ml R ./genfft-k7/vSimdBasics.mli R ./genfft-k7/vSimdIndexing.ml R ./genfft-k7/vSimdIndexing.mli R ./genfft-k7/vSimdUnparsing.ml R ./genfft-k7/vSimdUnparsing.mli R ./genfft-k7/variable.ml R ./genfft-k7/variable.mli R ./genfft-k7/ M ./kernel/align.c -2 M ./kernel/ifftw.h -5 +1 M ./mkdist.sh -1 +1 M ./support/Makefile.am -1 +1 M ./support/Makefile.codelets -6 +1 R ./support/codelet_asmprelude Sun Aug 13 11:02:11 EDT 2006 athena@fftw.org * Use darcs instead of cvs. M ./mkdist.sh -5 +2 Wed Jul 19 08:52:15 EDT 2006 athena * [project @ 2006-07-19 12:52:15 by athena] Treat a the string "-" as a nonoption. M ./libbench2/my-getopt.c -1 +4 Tue Jul 4 17:10:47 EDT 2006 stevenj * [project @ 2006-07-04 21:10:47 by stevenj] comment out pkginclude dir for now M ./Makefile.am -2 +2 M ./api/Makefile.am -2 +2 M ./dft/Makefile.am -2 +2 M ./kernel/Makefile.am -2 +2 M ./rdft/Makefile.am -2 +2 M ./reodft/Makefile.am -2 +2 M ./threads/Makefile.am -2 +2 Mon Jul 3 20:51:08 EDT 2006 stevenj * [project @ 2006-07-04 00:51:08 by stevenj] make sure CCAS = CC to avoid libtool confusion M ./configure.ac -1 +1 Fri Jun 23 04:07:31 EDT 2006 stevenj * [project @ 2006-06-23 08:07:31 by stevenj] install x77.h guru.h guru64.h in pkgincludedir M ./api/Makefile.am -1 +1 Fri Jun 23 04:03:42 EDT 2006 stevenj * [project @ 2006-06-23 08:03:42 by stevenj] whitespace M ./configure.ac -3 Fri Jun 23 02:33:45 EDT 2006 stevenj * [project @ 2006-06-23 06:33:45 by stevenj] support cycle counter with xlc on Linux/ppc M ./kernel/cycle.h -2 +2 Tue Jun 20 08:16:08 EDT 2006 athena * [project @ 2006-06-20 12:16:08 by athena] Stylistic change. M ./tools/fftw-wisdom.c -1 +1 Tue Jun 20 02:20:34 EDT 2006 stevenj * [project @ 2006-06-20 06:20:34 by stevenj] bump date M ./m4/ax_cc_maxopt.m4 -1 +1 Tue Jun 20 02:20:06 EDT 2006 stevenj * [project @ 2006-06-20 06:20:06 by stevenj] correct bug reported by Andrew Salamon ... --enable-portable-binary was ignored (or rather, treated unpredictably) due to typo, grrr M ./m4/ax_cc_maxopt.m4 -1 +1 Thu Jun 1 20:30:06 EDT 2006 stevenj * [project @ 2006-06-02 00:30:06 by stevenj] install 'internal' header files into includedir/fftw3/, includedir/fftw3f/, etcetera....this will make it easier to write external libraries that plug into FFTW internals, e.g. to add new solvers M ./Makefile.am +3 M ./api/Makefile.am +3 M ./dft/Makefile.am +3 M ./kernel/Makefile.am +3 M ./rdft/Makefile.am +3 M ./reodft/Makefile.am +3 M ./threads/Makefile.am +3 Mon May 29 23:59:19 EDT 2006 stevenj * [project @ 2006-05-30 03:59:19 by stevenj] bug fix, thanks to James Donald for the bug report (only affects experimental semaphore stuff) M ./threads/threads.c +1 Mon May 29 23:58:16 EDT 2006 stevenj * [project @ 2006-05-30 03:58:16 by stevenj] comment M ./NEWS -3 +4 Mon May 29 21:02:50 EDT 2006 stevenj * [project @ 2006-05-30 01:02:50 by stevenj] whoops M ./m4/acx_pthread.m4 -1 +3 Sat May 27 19:36:15 EDT 2006 stevenj * [project @ 2006-05-27 23:36:15 by stevenj] version bump M ./m4/acx_pthread.m4 -1 +1 Sat May 27 14:54:47 EDT 2006 stevenj * [project @ 2006-05-27 18:54:47 by stevenj] only check for xlc_r/cc_r if we are not using gcc M ./m4/acx_pthread.m4 -2 +3 Fri May 26 15:00:38 EDT 2006 stevenj * [project @ 2006-05-26 19:00:38 by stevenj] use ptrdiff_t (it's C89 and standard C++, hooray) M ./api/fftw3.h -10 +5 Fri May 26 12:59:33 EDT 2006 stevenj * [project @ 2006-05-26 16:59:33 by stevenj] version bump M ./configure.ac -2 +2 Fri May 26 12:57:32 EDT 2006 stevenj * [project @ 2006-05-26 16:57:32 by stevenj] noted 64-bit guru API M ./NEWS +7 Fri May 26 12:53:09 EDT 2006 stevenj * [project @ 2006-05-26 16:53:09 by stevenj] note that newer versions of VC++ support long long M ./api/fftw3.h -2 +3 Fri May 26 12:46:09 EDT 2006 stevenj * [project @ 2006-05-26 16:46:09 by stevenj] try harder to get a portable 64-bit type M ./api/fftw3.h -2 +6 Thu May 25 22:04:18 EDT 2006 stevenj * [project @ 2006-05-26 02:04:18 by stevenj] added draft guru64 API A ./api/guru.h A ./api/guru64.h A ./api/mktensor-iodims.h A ./api/mktensor-iodims64.c A ./api/plan-guru-dft-c2r.h A ./api/plan-guru-dft-r2c.h A ./api/plan-guru-dft.h A ./api/plan-guru-r2r.h A ./api/plan-guru-split-dft-c2r.h A ./api/plan-guru-split-dft-r2c.h A ./api/plan-guru-split-dft.h A ./api/plan-guru64-dft-c2r.c A ./api/plan-guru64-dft-r2c.c A ./api/plan-guru64-dft.c A ./api/plan-guru64-r2r.c A ./api/plan-guru64-split-dft-c2r.c A ./api/plan-guru64-split-dft-r2c.c A ./api/plan-guru64-split-dft.c M ./api/Makefile.am -1 +7 M ./api/api.h +3 M ./api/fftw3.h -2 +56 M ./api/guru.h +4 M ./api/guru64.h +4 M ./api/mktensor-iodims.c -62 +2 M ./api/mktensor-iodims.h +62 M ./api/mktensor-iodims64.c +2 M ./api/plan-guru-dft-c2r.c -44 +2 M ./api/plan-guru-dft-c2r.h +44 M ./api/plan-guru-dft-r2c.c -43 +2 M ./api/plan-guru-dft-r2c.h +43 M ./api/plan-guru-dft.c -44 +2 M ./api/plan-guru-dft.h +44 M ./api/plan-guru-r2r.c -47 +2 M ./api/plan-guru-r2r.h +47 M ./api/plan-guru-split-dft-c2r.c -40 +2 M ./api/plan-guru-split-dft-c2r.h +40 M ./api/plan-guru-split-dft-r2c.c -39 +2 M ./api/plan-guru-split-dft-r2c.h +39 M ./api/plan-guru-split-dft.c -39 +2 M ./api/plan-guru-split-dft.h +39 M ./api/plan-guru64-dft-c2r.c +2 M ./api/plan-guru64-dft-r2c.c +2 M ./api/plan-guru64-dft.c +2 M ./api/plan-guru64-r2r.c +2 M ./api/plan-guru64-split-dft-c2r.c +2 M ./api/plan-guru64-split-dft-r2c.c +2 M ./api/plan-guru64-split-dft.c +2 Mon May 22 16:41:44 EDT 2006 stevenj * [project @ 2006-05-22 20:41:44 by stevenj] added FIXME note M ./m4/acx_pthread.m4 +1 Mon May 22 16:40:30 EDT 2006 stevenj * [project @ 2006-05-22 20:40:30 by stevenj] check for xlc_r in addition to cc_r; thanks to Guy Moebs for the bug report M ./m4/acx_pthread.m4 -3 +3 Fri Apr 21 12:35:25 EDT 2006 stevenj * [project @ 2006-04-21 16:35:25 by stevenj] added note about gcc 4.0.1 on MacOS/Intel M ./doc/FAQ/fftw-faq.bfnn -1 +4 Thu Apr 20 23:08:42 EDT 2006 stevenj * [project @ 2006-04-21 03:08:42 by stevenj] added code for Core Duo; thanks to Eric Branlund M ./m4/ax_gcc_archflag.m4 -2 +3 Thu Apr 20 20:21:03 EDT 2006 stevenj * [project @ 2006-04-21 00:21:03 by stevenj] fixed failure for -fPIC or for gcc-4 on Apple Intel machines; thanks to Eric Branlund for the bug report M ./m4/ax_gcc_x86_cpuid.m4 -4 +12 Tue Apr 11 20:00:31 EDT 2006 athena * [project @ 2006-04-12 00:00:31 by athena] Use -maltivec when checking for altivec.h. M ./configure.ac -1 +15 Mon Apr 3 15:52:44 EDT 2006 stevenj * [project @ 2006-04-03 19:52:44 by stevenj] note planner overwriting input in planner-flags reference M ./doc/fftw3.texi -1 +6 Tue Mar 28 09:05:26 EST 2006 athena * [project @ 2006-03-28 14:05:26 by athena] FAQ entry about --enable-k7 in 64-bit mode. M ./doc/FAQ/fftw-faq.bfnn +11 Mon Mar 27 23:41:05 EST 2006 stevenj * [project @ 2006-03-28 04:41:05 by stevenj] sprintf -> snprintf, to avoid (harmless) complaints by users/compilers M ./configure.ac -1 +1 M ./libbench2/report.c -12 +19 M ./tools/fftw-wisdom.c +4 Mon Mar 27 23:30:22 EST 2006 stevenj * [project @ 2006-03-28 04:30:22 by stevenj] silence compiler warning M ./kernel/align.c -5 +6 Fri Mar 17 09:20:10 EST 2006 athena * [project @ 2006-03-17 14:20:10 by athena] Remove dft/codelets/inplace, add simd/nonportable to list of directories to be compiled on non-unix systems. M ./doc/fftw3.texi -13 +14 Sat Aug 12 23:52:49 EDT 2006 Unknown tagger tagged fftw-3-1-1 Sat Mar 4 16:17:56 EST 2006 stevenj * [project @ 2006-03-04 21:17:56 by stevenj] whoops M ./doc/fftw3.texi -2 +2 Sat Mar 4 16:13:08 EST 2006 stevenj * [project @ 2006-03-04 21:13:08 by stevenj] note that we align the stack ourselves if necessary, with gcc and icc M ./doc/fftw3.texi -8 +10 Sat Mar 4 16:08:16 EST 2006 stevenj * [project @ 2006-03-04 21:08:16 by stevenj] clearer distinction between static and automatic storage in C M ./doc/fftw3.texi -29 +30 Sat Feb 25 20:27:01 EST 2006 stevenj * [project @ 2006-02-26 01:27:01 by stevenj] rm unused var M ./libbench2/verify-lib.c -2 +1 Sat Feb 25 17:30:28 EST 2006 athena * [project @ 2006-02-25 22:30:28 by athena] Improved usage of goto (Dijkstra miserere nostri) M ./libbench2/my-getopt.c -18 +17 Sat Feb 25 14:19:15 EST 2006 stevenj * [project @ 2006-02-25 19:19:15 by stevenj] boilerplate M ./libbench2/my-getopt.h +13 Sat Feb 25 14:14:40 EST 2006 stevenj * [project @ 2006-02-25 19:14:40 by stevenj] update for upcoming 3.1.1 M ./NEWS +17 Sat Feb 25 13:57:34 EST 2006 stevenj * [project @ 2006-02-25 18:57:34 by stevenj] replace obsolete IMPATIENT with MEASURE M ./tools/fftw-wisdom.c -1 +3 M ./tools/fftw_wisdom.1.in -2 +2 Sat Feb 25 13:52:25 EST 2006 stevenj * [project @ 2006-02-25 18:52:25 by stevenj] corrected comment M ./tools/fftw-wisdom.c -1 +1 Sat Feb 25 10:19:26 EST 2006 athena * [project @ 2006-02-25 15:19:26 by athena] -v does not take an argument. M ./tools/fftw-wisdom.c -6 +3 Sat Feb 25 10:17:18 EST 2006 athena * [project @ 2006-02-25 15:17:18 by athena] Obey the unix convention that -ab = -a -b M ./libbench2/my-getopt.c -15 +24 Fri Feb 24 23:13:49 EST 2006 stevenj * [project @ 2006-02-25 04:13:49 by stevenj] minor fixes (return error on unrecognized option) M ./libbench2/bench-main.c -2 +3 M ./libbench2/my-getopt.c +5 M ./tools/fftw-wisdom.c -3 +4 Fri Feb 24 22:46:12 EST 2006 stevenj * [project @ 2006-02-25 03:46:12 by stevenj] ugh M ./tools/fftw-wisdom.c -34 +30 Fri Feb 24 21:42:56 EST 2006 athena * [project @ 2006-02-25 02:42:56 by athena] require exact match for long options. M ./libbench2/my-getopt.c -1 +2 Fri Feb 24 21:38:02 EST 2006 athena * [project @ 2006-02-25 02:38:02 by athena] better fix M ./libbench2/my-getopt.c -1 +2 Fri Feb 24 21:37:06 EST 2006 athena * [project @ 2006-02-25 02:37:06 by athena] Fix M ./libbench2/my-getopt.c -1 +1 Fri Feb 24 21:25:48 EST 2006 athena * [project @ 2006-02-25 02:25:48 by athena] nothing A ./libbench2/my-getopt.c A ./libbench2/my-getopt.h M ./libbench2/Makefile.am -5 +4 M ./libbench2/bench-main.c -53 +48 M ./libbench2/bench.h -5 +1 R ./libbench2/getopt-utils.c R ./libbench2/getopt.c R ./libbench2/getopt.h R ./libbench2/getopt1.c M ./libbench2/my-getopt.c +153 M ./libbench2/my-getopt.h +33 Mon Feb 20 17:37:21 EST 2006 stevenj * [project @ 2006-02-20 22:37:21 by stevenj] rm transpose-indirect-inplace solver, which was buggy M ./dft/indirect-transpose.c -54 +11 Wed Feb 15 08:43:05 EST 2006 athena * [project @ 2006-02-15 13:43:05 by athena] Comment fix. M ./kernel/cycle.h -2 +2 Wed Feb 15 08:18:41 EST 2006 athena * [project @ 2006-02-15 13:18:41 by athena] Cycle counter for Visual C++ x86-64, courtesy of Dirk Michaelis M ./kernel/cycle.h -4 +5 Tue Feb 14 19:17:30 EST 2006 stevenj * [project @ 2006-02-15 00:17:30 by stevenj] rfftwnd.png is in builddir M ./doc/Makefile.am -1 +1 Tue Feb 14 19:03:27 EST 2006 stevenj * [project @ 2006-02-15 00:03:27 by stevenj] fixed typo: --enable-portable-binary, not --with M ./doc/fftw3.texi -3 +3 Mon Feb 13 07:59:06 EST 2006 athena * [project @ 2006-02-13 12:59:06 by athena] estimator tweaks. M ./dft/dftw-direct.c -2 +3 M ./rdft/hc2hc-direct.c -2 +2 Sun Feb 12 20:43:39 EST 2006 athena * [project @ 2006-02-13 01:43:39 by athena] sse/sse2 support for t3?v codelets M ./simd/simd-sse.h +22 M ./simd/simd-sse2.h -8 +20 Sun Feb 12 20:39:22 EST 2006 athena * [project @ 2006-02-13 01:39:22 by athena] Use CEXP instead of SIN/COS. M ./simd/simd-altivec.h -1 +1 Sun Feb 12 20:12:10 EST 2006 athena * [project @ 2006-02-13 01:12:10 by athena] bug in randomized cse eliminator. M ./genfft/oracle.ml -3 +5 Sun Feb 12 18:34:12 EST 2006 athena * [project @ 2006-02-12 23:34:12 by athena] Added support for t2-style simd codelets. This is altivec only for now; sse/sse2 don't even compile yet. A ./dft/simd/t3b.h A ./dft/simd/t3f.h M ./dft/simd/Makefile.am -2 +2 M ./dft/simd/codelets/Makefile.am -2 +11 M ./dft/simd/t3b.h +34 M ./dft/simd/t3f.h +34 M ./genfft/algsimp.ml -4 +26 M ./genfft/annotate.ml -5 +3 M ./genfft/c.ml -10 +9 M ./genfft/c.mli -2 +2 M ./genfft/complex.ml -93 +6 M ./genfft/complex.mli -5 +3 M ./genfft/expr.ml -6 +15 M ./genfft/expr.mli -2 +4 M ./genfft/gen_athtw.ml -3 +3 M ./genfft/gen_conv.ml -5 +3 M ./genfft/gen_hc2hc.ml -4 +3 M ./genfft/gen_hc2r.ml -5 +3 M ./genfft/gen_mdct.ml -7 +3 M ./genfft/gen_notw.ml -5 +2 M ./genfft/gen_notw_c.ml -5 +2 M ./genfft/gen_r2hc.ml -4 +3 M ./genfft/gen_r2r.ml -4 +3 M ./genfft/gen_twiddle.ml -3 +3 M ./genfft/gen_twiddle_c.ml -10 +4 M ./genfft/gen_twidsq.ml -7 +3 M ./genfft/gen_twidsq_c.ml -15 +5 M ./genfft/magic.ml -4 +4 M ./genfft/oracle.ml -1 +5 M ./genfft/schedule.ml -3 +3 M ./genfft/simd.ml -5 +11 M ./genfft/to_alist.ml -4 +16 M ./genfft/trig.ml -3 +2 M ./genfft/twiddle.ml -265 +35 M ./genfft/twiddle.mli -4 +4 M ./simd/simd-altivec.h +28 Sun Feb 12 15:30:27 EST 2006 athena * [project @ 2006-02-12 20:30:27 by athena] Added support for t2-style simd split-complex codelets. A ./dft/simd/ts.c A ./dft/simd/ts.h M ./dft/simd/Makefile.am -2 +2 M ./dft/simd/codelets/Makefile.am -2 +7 R ./dft/simd/t1s.c R ./dft/simd/t1s.h M ./dft/simd/ts.c +50 M ./dft/simd/ts.h +33 M ./genfft/twiddle.ml -5 +3 Fri Feb 10 18:21:28 EST 2006 stevenj * [project @ 2006-02-10 23:21:28 by stevenj] M ./m4/ax_openmp.m4 -1 +1 Fri Feb 10 18:19:46 EST 2006 stevenj * [project @ 2006-02-10 23:19:46 by stevenj] punctuation M ./m4/ax_openmp.m4 -2 +2 Fri Feb 10 18:00:35 EST 2006 stevenj * [project @ 2006-02-10 23:00:35 by stevenj] windows DLL stuff for Fortran interface M ./api/f77api.c +7 M ./api/f77funcs.h -43 +43 Fri Feb 10 09:48:52 EST 2006 athena * [project @ 2006-02-10 14:48:52 by athena] Bumped version to 3.1.1 M ./configure.ac -2 +2 Fri Feb 10 09:18:39 EST 2006 athena * [project @ 2006-02-10 14:18:39 by athena] Precompute array indices on x86-64. Speeds up Pentium IV and makes no appreciable difference on AMD. M ./kernel/ifftw.h -2 +2 Tue Feb 7 22:01:36 EST 2006 athena * [project @ 2006-02-08 03:01:36 by athena] Check whether the processor supports CPUID before issuing the instruction. (Grrr...) Code contributed by Eric J. Korpela. A ./simd/x86-cpuid.h M ./simd/Makefile.am -1 +1 M ./simd/sse.c -24 +6 M ./simd/sse2.c -26 +8 M ./simd/x86-cpuid.h +127 Tue Feb 7 21:36:47 EST 2006 athena * [project @ 2006-02-08 02:36:47 by athena] icc supports x86_64 these days. M ./kernel/cycle.h -2 +2 Sun Feb 5 18:19:55 EST 2006 athena * [project @ 2006-02-05 23:19:55 by athena] Paranoia. M ./kernel/primes.c -2 +2 Mon Jan 30 15:27:53 EST 2006 stevenj * [project @ 2006-01-30 20:27:53 by stevenj] whoops, fixed assert (y <= x) M ./kernel/primes.c -2 +2 Mon Jan 30 15:26:22 EST 2006 stevenj * [project @ 2006-01-30 20:26:22 by stevenj] note that safe_mulmod requires {x,y} < p (or at least < 2p), and added assert M ./kernel/primes.c -2 +4 Mon Jan 30 11:09:32 EST 2006 athena * [project @ 2006-01-30 16:09:32 by athena] fixed aix/xlc lossage M ./libbench2/bench-user.h -2 +2 M ./libbench2/timer.c -4 +4 Sun Jan 29 20:42:51 EST 2006 athena * [project @ 2006-01-30 01:42:51 by athena] In the impuse test, normalize the impulse so that the impulse and the random vectors have roughly the same L2 norm. This change reduces the number of bits that we lose because of floating-point cancellation, so that we can focus on the bits that we lose because of bugs. M ./libbench2/verify-lib.c -3 +3 Sun Jan 29 20:37:47 EST 2006 athena * [project @ 2006-01-30 01:37:47 by athena] Compute omega in trigreal precision, as opposed to R. M ./rdft/dht-rader.c -2 +2 Sat Aug 12 23:51:14 EDT 2006 Unknown tagger tagged fftw-3-1 Fri Jan 27 19:16:22 EST 2006 stevenj * [project @ 2006-01-28 00:16:22 by stevenj] add --with-combined-threads option as workaround to Windows inability to build shared libs with dependencies M ./Makefile.am -1 +7 M ./configure.ac -1 +3 M ./tests/Makefile.am +2 M ./threads/Makefile.am +4 M ./tools/Makefile.am +2 Fri Jan 27 17:20:45 EST 2006 stevenj * [project @ 2006-01-27 22:20:45 by stevenj] libfftw3_threads should *not* used -no-undefined because, in fact, it is not true -- this library depends on -lfftw3, and is not self-contained M ./threads/Makefile.am -1 +1 Thu Jan 26 22:04:34 EST 2006 stevenj * [project @ 2006-01-27 03:04:34 by stevenj] updated M ./NEWS -3 +10 Thu Jan 26 21:10:50 EST 2006 athena * [project @ 2006-01-27 02:10:50 by athena] Added paranoid stack alignment when awaking plans. While I was at it, removed obsolete, redundant AWAKE macro. M ./api/apiplan.c -3 +8 M ./dft/bluestein.c -1 +1 M ./dft/buffered.c -4 +4 M ./dft/ct.c -3 +3 M ./dft/ctsq.c -2 +2 M ./dft/dftw-generic.c -1 +1 M ./dft/dftw-genericbuf.c -1 +1 M ./dft/indirect-transpose.c -3 +3 M ./dft/indirect.c -3 +3 M ./dft/rader.c -3 +3 M ./dft/rank-geq2.c -3 +3 M ./dft/vrank-geq1.c -2 +2 M ./kernel/ifftw.h -2 +1 M ./kernel/timer.c -3 +3 M ./rdft/buffered.c -4 +4 M ./rdft/buffered2.c -3 +3 M ./rdft/dft-r2hc.c -2 +2 M ./rdft/dht-r2hc.c -2 +2 M ./rdft/dht-rader.c -3 +3 M ./rdft/hc2hc-direct.c -3 +3 M ./rdft/hc2hc-directbuf.c -3 +3 M ./rdft/hc2hc-generic.c -2 +2 M ./rdft/hc2hc.c -2 +2 M ./rdft/indirect.c -3 +3 M ./rdft/rank-geq2-rdft2.c -3 +3 M ./rdft/rank-geq2.c -3 +3 M ./rdft/rank0-rdft2.c -2 +2 M ./rdft/rdft-dht.c -2 +2 M ./rdft/rdft2-radix2.c -2 +2 M ./rdft/vrank-geq1-rdft2.c -2 +2 M ./rdft/vrank-geq1.c -2 +2 M ./rdft/vrank3-transpose.c -4 +4 M ./reodft/redft00e-r2hc-pad.c -3 +3 M ./reodft/redft00e-r2hc.c -2 +2 M ./reodft/reodft00e-splitradix.c -3 +3 M ./reodft/reodft010e-r2hc.c -2 +2 M ./reodft/reodft11e-r2hc-odd.c -2 +2 M ./reodft/reodft11e-r2hc.c -2 +2 M ./reodft/reodft11e-radix2.c -2 +2 M ./reodft/rodft00e-r2hc-pad.c -3 +3 M ./reodft/rodft00e-r2hc.c -2 +2 M ./tests/hook.c -2 +2 M ./threads/ct.c -3 +3 M ./threads/dft-vrank-geq1.c -2 +2 M ./threads/hc2hc.c -2 +2 M ./threads/rdft-vrank-geq1.c -2 +2 M ./threads/vrank-geq1-rdft2.c -2 +2 Thu Jan 26 20:54:39 EST 2006 athena * [project @ 2006-01-27 01:54:39 by athena] Updated for 3.1. M ./NEWS -1 +1 Thu Jan 26 19:15:12 EST 2006 athena * [project @ 2006-01-27 00:15:12 by athena] ditched one alignment check and noted that we should eliminate the rest as well M ./TODO +3 M ./libbench2/bench-main.c -13 +1 Wed Jan 25 23:02:19 EST 2006 athena * [project @ 2006-01-26 04:02:19 by athena] alignment hack M ./libbench2/bench-main.c -2 +4 Wed Jan 25 22:05:11 EST 2006 athena * [project @ 2006-01-26 03:05:11 by athena] detect pentium M M ./m4/ax_gcc_archflag.m4 -1 +2 Wed Jan 25 18:42:58 EST 2006 stevenj * [project @ 2006-01-25 23:42:58 by stevenj] don't trust host_cpu if it claims we are on i386/i486, and call cpuid anyway (if it fails we use no arch flag). This is needed on FreeBSD M ./m4/ax_gcc_archflag.m4 -3 +1 Wed Jan 25 18:00:04 EST 2006 stevenj * [project @ 2006-01-25 23:00:04 by stevenj] suggest --with-our-malloc16 in error message M ./kernel/kalloc.c -2 +2 Tue Jan 24 19:53:34 EST 2006 stevenj * [project @ 2006-01-25 00:53:34 by stevenj] ditto for -no-gcc M ./configure.ac -2 +4 Tue Jan 24 19:51:08 EST 2006 stevenj * [project @ 2006-01-25 00:51:08 by stevenj] flags required for successfull compilation should be added even if the user overrides CFLAGS M ./configure.ac -9 +12 Tue Jan 24 18:43:59 EST 2006 stevenj * [project @ 2006-01-24 23:43:59 by stevenj] upcoming gcc OpenMP support uses -fopenmp M ./m4/ax_openmp.m4 -4 +4 Tue Jan 24 18:26:59 EST 2006 stevenj * [project @ 2006-01-24 23:26:59 by stevenj] note that PGI uses -mp as well M ./m4/ax_openmp.m4 -1 +1 Mon Jan 23 15:31:24 EST 2006 athena * [project @ 2006-01-23 20:31:24 by athena] my best guess at how to fix the microsoft crap du jour M ./kernel/cycle.h -2 +2 M ./simd/sse.c -65 +70 M ./simd/sse2.c -62 +67 Mon Jan 23 14:05:14 EST 2006 stevenj * [project @ 2006-01-23 19:05:14 by stevenj] use -Masmkeyword for PGI cycle counter, grr M ./configure.ac +5 M ./kernel/cycle.h -2 +4 Sun Jan 22 18:09:06 EST 2006 athena * [project @ 2006-01-22 23:09:06 by athena] Bumped version number to 3.1. M ./configure.ac -1 +1 Sat Jan 21 10:03:59 EST 2006 athena * [project @ 2006-01-21 15:03:59 by athena] Report that --enable-k7 is incompatible with --enable-shared. M ./configure.ac +3 Sat Jan 21 09:17:54 EST 2006 athena * [project @ 2006-01-21 14:17:54 by athena] Do not use empty libraries in LIBADD, since otherwise the linker fails on Solaris. M ./Makefile.am -7 +20 Wed Jan 18 10:47:59 EST 2006 stevenj * [project @ 2006-01-18 15:47:59 by stevenj] warn end-users away from this file M ./bootstrap.sh -1 +10 Tue Jan 17 16:16:42 EST 2006 athena * [project @ 2006-01-17 21:16:42 by athena] Gcc sucks. M ./simd/simd-sse.h -1 +2 Tue Jan 17 11:48:55 EST 2006 athena * [project @ 2006-01-17 16:48:55 by athena] Disabled checks that may turn out to be too paranoid. M ./tests/hook.c -1 +2 Tue Jan 17 10:35:03 EST 2006 athena * [project @ 2006-01-17 15:35:03 by athena] Some paranoid checks. M ./tests/hook.c +16 Tue Jan 17 09:31:08 EST 2006 athena * [project @ 2006-01-17 14:31:08 by athena] Flush stdout after printing. M ./libbench2/ovtpvt.c +1 Tue Jan 17 08:28:18 EST 2006 athena * [project @ 2006-01-17 13:28:18 by athena] Run the leak detector in all cases, not just when verbose > 2. M ./kernel/alloc.c -2 +7 M ./tests/bench.c -2 +1 Tue Jan 17 08:11:41 EST 2006 athena * [project @ 2006-01-17 13:11:41 by athena] Eliminate calls to pow(), rint(). M ./api/mapflags.c -8 +3 Tue Jan 17 00:45:06 EST 2006 stevenj * [project @ 2006-01-17 05:45:06 by stevenj] put # in first column, for stylistic consistency M ./kernel/ifftw.h -3 +3 Tue Jan 17 00:17:27 EST 2006 athena * [project @ 2006-01-17 05:17:27 by athena] Made timeout part of impatience flags, in order to improve the usability of wisdom. Also, fixed bogus error recovery logic in planner.c:imprt(). M ./api/mapflags.c -1 +36 M ./kernel/ifftw.h -4 +7 M ./kernel/planner.c -23 +70 Mon Jan 16 23:03:34 EST 2006 stevenj * [project @ 2006-01-17 04:03:33 by stevenj] make timelimit < 0 .eq. FFTW_NO_TIMELIMIT M ./api/apiplan.c -1 +1 M ./api/fftw3.h -1 +3 M ./doc/fftw3.texi -5 +9 M ./kernel/planner.c -3 +3 Mon Jan 16 21:52:01 EST 2006 athena * [project @ 2006-01-17 02:52:01 by athena] Eliminated the FFTW_TIMELIMIT flag in favor of this simpler logic: fftw_set_timelimit(0) disables time limit. fftw_set_timelimit(X), X>0 sets the time limit to X. M ./api/apiplan.c -6 +1 M ./api/fftw3.h -2 +1 M ./api/the-planner.c +7 M ./doc/fftw3.texi -30 +27 M ./kernel/planner.c -2 +4 M ./tests/bench.c -1 Mon Jan 16 08:38:04 EST 2006 athena * [project @ 2006-01-16 13:38:04 by athena] Force the use of the estimator when wisdom fails because of md5 collisions, otherwise the planner takes forever. M ./api/apiplan.c -2 +10 Sun Jan 15 21:30:31 EST 2006 athena * [project @ 2006-01-16 02:30:31 by athena] Ranted about how broken gcc-4 is. M ./kernel/ifftw.h -5 +40 Sun Jan 15 19:59:38 EST 2006 stevenj * [project @ 2006-01-16 00:59:38 by stevenj] change fftw_timelimit global var to fftw_set_timelimit(double) function, for simpler usage with shared libraries and for consistency with e.g. set_numthreads M ./api/apiplan.c -2 +4 M ./api/fftw3.h -2 +3 M ./doc/fftw3.texi -11 +11 M ./tests/bench.c -1 +1 Sun Jan 15 19:32:27 EST 2006 athena * [project @ 2006-01-16 00:32:27 by athena] Minor tweaks. M ./doc/fftw3.texi -8 +11 Sun Jan 15 16:32:54 EST 2006 athena * [project @ 2006-01-15 21:32:54 by athena] tweaks to make sure that time_n() is always called from the same stack position. M ./libbench2/timer.c -12 +12 Sun Jan 15 16:09:53 EST 2006 athena * [project @ 2006-01-15 21:09:53 by athena] Major simplification of the timer calibration logic. Also, use an FFT as a unit of work instead of the old pointer chasing, because God knows how pointer chasing interacts with the idiotic cache-hit speculation on the Pentium IV. M ./libbench2/bench.h -2 +3 M ./libbench2/speed.c -5 +8 M ./libbench2/timer.c -96 +35 M ./libbench2/timer2.c -7 +99 Sun Jan 15 15:12:08 EST 2006 athena * [project @ 2006-01-15 20:12:08 by athena] Fixed broken aligment checks when sizeof(R)==12. M ./kernel/align.c -3 +8 Sun Jan 15 10:36:40 EST 2006 athena * [project @ 2006-01-15 15:36:40 by athena] Manual unrolling of loop. M ./libbench2/timer2.c -5 +2 Sun Jan 15 10:12:55 EST 2006 athena * [project @ 2006-01-15 15:12:55 by athena] Various improvements to timer calibration routines. A ./libbench2/timer2.c M ./libbench2/Makefile.am -1 +1 M ./libbench2/bench.h -1 +2 M ./libbench2/timer.c -18 +32 M ./libbench2/timer2.c +36 Sat Jan 14 22:16:09 EST 2006 athena * [project @ 2006-01-15 03:16:09 by athena] cygwin defines __CYGWIN__, not __WIN32__ etc. M ./libbench2/timer.c -16 +17 Sat Jan 14 20:40:12 EST 2006 athena * [project @ 2006-01-15 01:40:12 by athena] fixed confusion between libbench and user timers M ./libbench2/bench-user.h -3 +7 M ./libbench2/speed.c -5 +5 M ./libbench2/timer.c -6 +8 M ./tests/bench.c -8 +8 Sat Jan 14 12:32:44 EST 2006 stevenj * [project @ 2006-01-14 17:32:44 by stevenj] update M ./NEWS -2 +4 Sat Jan 14 10:24:11 EST 2006 athena * [project @ 2006-01-14 15:24:11 by athena] Comment. M ./simd/simd-sse.h +5 Sat Jan 14 10:19:28 EST 2006 athena * [project @ 2006-01-14 15:19:28 by athena] Workaround gcc bug. M ./simd/simd-sse.h -2 +3 Fri Jan 13 19:13:18 EST 2006 athena * [project @ 2006-01-14 00:13:18 by athena] Switched to -beta2. M ./configure.ac -1 +1 Thu Jan 12 22:21:57 EST 2006 athena * [project @ 2006-01-13 03:21:57 by athena] Fixed technically correct but highly obfuscated use of the enum tag R2HC as a null pointer. M ./rdft/buffered.c -5 +3 M ./rdft/indirect.c -4 +3 M ./rdft/problem.c -1 +8 M ./rdft/rank0-rdft2.c -4 +3 M ./rdft/rdft.h +1 M ./rdft/vrank3-transpose.c -25 +13 Thu Jan 12 19:25:20 EST 2006 stevenj * [project @ 2006-01-13 00:25:20 by stevenj] --enable-unsafe-mulmod is obsolete M ./configure.ac -4 Thu Jan 12 19:23:18 EST 2006 athena * [project @ 2006-01-13 00:23:18 by athena] More thoughts. M ./TODO +2 Thu Jan 12 19:17:57 EST 2006 athena * [project @ 2006-01-13 00:17:57 by athena] Removed loop unrolling because it slows things down on at least one powerpc and it generates clumsy x86 code. M ./rdft/buffered2.c -50 +12 Thu Jan 12 19:17:35 EST 2006 stevenj * [project @ 2006-01-13 00:17:35 by stevenj] tweaks M ./kernel/kalloc.c -6 +6 Thu Jan 12 15:55:52 EST 2006 stevenj * [project @ 2006-01-12 20:55:52 by stevenj] MacOSX x86 ABI specifies that the stack is kept 16-byte aligned M ./kernel/ifftw.h -2 +3 Thu Jan 12 12:46:49 EST 2006 athena * [project @ 2006-01-12 17:46:49 by athena] ``ret'' is a reserved word in the evil empire. M ./kernel/cycle.h -5 +5 Thu Jan 12 08:31:43 EST 2006 athena * [project @ 2006-01-12 13:31:09 by athena] Changed ret => result because ret ``is a reserved word'' in the evil empire. M ./simd/sse.c -4 +4 M ./simd/sse2.c -4 +4 Wed Jan 11 19:30:42 EST 2006 athena * [project @ 2006-01-12 00:30:42 by athena] Workaround Visual c++ lossage. M ./simd/simd-sse2.h -5 +4 Wed Jan 11 19:26:16 EST 2006 athena * [project @ 2006-01-12 00:26:16 by athena] Workaround visual c++ lossage. M ./simd/simd-sse.h +38 Wed Jan 11 19:10:52 EST 2006 athena * [project @ 2006-01-12 00:10:52 by athena] isprint() is guaranteed to work for unsigned char + EOF only. M ./libbench2/getopt-utils.c -2 +3 Wed Jan 11 13:47:49 EST 2006 stevenj * [project @ 2006-01-11 18:47:49 by stevenj] rm obsolete fixme M ./rdft/vrank3-transpose.c -2 +2 Wed Jan 11 13:38:46 EST 2006 stevenj * [project @ 2006-01-11 18:38:46 by stevenj] M ./rdft/vrank3-transpose.c -2 +2 Wed Jan 11 13:32:26 EST 2006 stevenj * [project @ 2006-01-11 18:32:26 by stevenj] fix comment M ./rdft/vrank3-transpose.c -2 +3 Wed Jan 11 12:27:05 EST 2006 athena * [project @ 2006-01-11 17:27:05 by athena] Paranoid use of K(x) for all constants x, to avoid runtime double->float conversions on sufficiently stupid compilers. M ./dft/bluestein.c -2 +2 M ./rdft/buffered2.c -2 +2 M ./rdft/dht-rader.c -1 +1 M ./rdft/rank0-rdft2.c -11 +11 M ./reodft/rodft00e-r2hc-pad.c -3 +3 Tue Jan 10 20:10:38 EST 2006 athena * [project @ 2006-01-11 01:10:38 by athena] Workaround to gcc nonsense. M ./simd/simd-sse.h -1 +15 Tue Jan 10 18:44:28 EST 2006 stevenj * [project @ 2006-01-10 23:44:28 by stevenj] bug fix: infinite loop in transpose-cut planning M ./rdft/vrank3-transpose.c -6 +14 Tue Jan 10 18:12:14 EST 2006 stevenj * [project @ 2006-01-10 23:12:14 by stevenj] clarified comment M ./api/fftw3.h -6 +6 Tue Jan 10 18:10:32 EST 2006 stevenj * [project @ 2006-01-10 23:10:32 by stevenj] more Windows decorations M ./tests/bench.c -1 +1 Tue Jan 10 17:57:45 EST 2006 stevenj * [project @ 2006-01-10 22:57:45 by stevenj] added FIXME comment M ./support/Makefile.codelets +1 Tue Jan 10 17:52:07 EST 2006 stevenj * [project @ 2006-01-10 22:52:07 by stevenj] 'make clean' should not delete codlist.c since it is included in the dist tarball M ./support/Makefile.codelets -2 +3 Tue Jan 10 17:50:12 EST 2006 athena * [project @ 2006-01-10 22:50:12 by athena] Change threshold for ``large'' Cooley-Tukey to 256K from 64K, since it seems to benefit the Pentium IV with sse and the planning cost is not too horrible. M ./dft/dftw-direct.c -2 +2 Tue Jan 10 17:45:11 EST 2006 stevenj * [project @ 2006-01-10 22:45:11 by stevenj] more missing Windows DLL decorations M ./kernel/ifftw.h -3 +3 Tue Jan 10 17:41:28 EST 2006 stevenj * [project @ 2006-01-10 22:41:28 by stevenj] remove unused var M ./rdft/dht-rader.c -2 Tue Jan 10 14:00:50 EST 2006 stevenj * [project @ 2006-01-10 19:00:50 by stevenj] allow compiler threads, if enabled, to take precedence over explicit threads M ./threads/threads.c -17 +17 Tue Jan 10 12:30:09 EST 2006 stevenj * [project @ 2006-01-10 17:30:09 by stevenj] M ./api/api.h -2 +2 Tue Jan 10 12:21:56 EST 2006 stevenj * [project @ 2006-01-10 17:21:56 by stevenj] M ./kernel/planner.c -2 +2 Tue Jan 10 09:13:20 EST 2006 athena * [project @ 2006-01-10 14:13:20 by athena] Fixed comment typo. M ./kernel/planner.c -2 +2 Tue Jan 10 08:59:22 EST 2006 athena * [project @ 2006-01-10 13:59:22 by athena] Rearranged timeout checks so as to eliminate one of them. M ./kernel/planner.c -9 +4 Tue Jan 10 08:56:55 EST 2006 athena * [project @ 2006-01-10 13:56:55 by athena] Converted residual CK() -> A(). M ./kernel/plan.c -2 +2 Tue Jan 10 08:36:13 EST 2006 athena * [project @ 2006-01-10 13:36:13 by athena] Maintain the invariant TIMED_OUT ==> NEED_TIMEOUT_CHECK. M ./kernel/planner.c -5 +10 Tue Jan 10 08:24:41 EST 2006 athena * [project @ 2006-01-10 13:24:41 by athena] silence some 64-bit warnings M ./api/mapflags.c -1 +1 M ./dft/rank-geq2.c -2 +2 M ./dft/vrank-geq1.c -2 +2 M ./kernel/buffered.c -1 +1 M ./kernel/md5.c -1 +1 M ./kernel/scan.c -2 +3 M ./rdft/rank-geq2-rdft2.c -2 +2 M ./rdft/rank-geq2.c -2 +2 M ./rdft/vrank-geq1-rdft2.c -2 +2 M ./rdft/vrank-geq1.c -2 +2 M ./threads/dft-vrank-geq1.c -2 +2 M ./threads/rdft-vrank-geq1.c -2 +2 M ./threads/vrank-geq1-rdft2.c -2 +2 Tue Jan 10 07:58:48 EST 2006 athena * [project @ 2006-01-10 12:58:48 by athena] Assertions. M ./tests/hook.c +3 Tue Jan 10 00:14:00 EST 2006 stevenj * [project @ 2006-01-10 05:14:00 by stevenj] some condensing M ./kernel/timer.c -10 +5 Tue Jan 10 00:03:32 EST 2006 stevenj * [project @ 2006-01-10 05:03:32 by stevenj] eliminate X(seconds) in favor of X(elapsed_since), in paranoia of clock wrap M ./api/apiplan.c -2 +2 M ./kernel/ifftw.h -3 +33 M ./kernel/planner.c -4 +4 M ./kernel/timer.c -58 +25 Mon Jan 9 23:21:21 EST 2006 stevenj * [project @ 2006-01-10 04:21:21 by stevenj] M ./kernel/timer.c -2 +2 Mon Jan 9 23:21:06 EST 2006 stevenj * [project @ 2006-01-10 04:21:06 by stevenj] hmm, a bit more pessimistic about clock wrapping M ./kernel/timer.c -4 +4 Mon Jan 9 23:20:26 EST 2006 athena * [project @ 2006-01-10 04:20:26 by athena] Revert to md5uint = unsigned int whenever possible, so as to avoid wasting space for unsigned long on 64-bit machines. M ./configure.ac +2 M ./kernel/ifftw.h -1 +5 Mon Jan 9 23:12:27 EST 2006 stevenj * [project @ 2006-01-10 04:12:27 by stevenj] note why clock() wrap should not be a concern M ./kernel/timer.c -1 +6 Mon Jan 9 22:57:16 EST 2006 stevenj * [project @ 2006-01-10 03:57:16 by stevenj] bugfix in recent timeout changes - check for case where last solver times out M ./kernel/planner.c -2 +8 Mon Jan 9 22:40:26 EST 2006 stevenj * [project @ 2006-01-10 03:40:26 by stevenj] started changes list from beta M ./NEWS -1 +11 Mon Jan 9 22:34:13 EST 2006 athena * [project @ 2006-01-10 03:34:13 by athena] Paranoia. M ./api/mapflags.c +4 Mon Jan 9 22:27:37 EST 2006 athena * [project @ 2006-01-10 03:27:37 by athena] Paranoid assertions. M ./kernel/planner.c -3 +7 Mon Jan 9 22:13:32 EST 2006 athena * [project @ 2006-01-10 03:13:32 by athena] Added FIXME comment stating the 64-bit uncleaniness of fftw_tensor_to_bench_tensor(). M ./tests/hook.c +1 Mon Jan 9 22:06:05 EST 2006 athena * [project @ 2006-01-10 03:06:05 by athena] Another 64-bit bug. M ./dft/simd/t.c -1 +1 Mon Jan 9 21:54:07 EST 2006 stevenj * [project @ 2006-01-10 02:54:07 by stevenj] more Windows DLL nonsense M ./api/api.h -5 +15 M ./kernel/ifftw.h -3 +5 M ./tests/hook.c +2 Mon Jan 9 21:18:25 EST 2006 stevenj * [project @ 2006-01-10 02:18:25 by stevenj] some additional dllexport tags required to build the test program, due to internal stuff called by hook.c M ./api/api.h -2 +3 M ./kernel/ifftw.h -6 +17 Mon Jan 9 20:31:15 EST 2006 stevenj * [project @ 2006-01-10 01:31:15 by stevenj] M ./api/fftw3.h -2 +2 Mon Jan 9 20:30:19 EST 2006 stevenj * [project @ 2006-01-10 01:30:19 by stevenj] comment M ./api/fftw3.h -4 +6 Mon Jan 9 20:20:28 EST 2006 stevenj * [project @ 2006-01-10 01:17:11 by stevenj] M ./api/api.h -7 +7 M ./api/fftw3.h -4 +3 Mon Jan 9 20:16:50 EST 2006 stevenj * [project @ 2006-01-10 01:16:50 by stevenj] clarification M ./api/fftw3.h -3 +3 Mon Jan 9 20:12:23 EST 2006 stevenj * [project @ 2006-01-10 01:12:23 by stevenj] define FFTW_DLL if DLL_EXPORT (defined by libtool) is supplied M ./api/api.h +10 Mon Jan 9 20:05:11 EST 2006 stevenj * [project @ 2006-01-10 01:05:11 by stevenj] whoops M ./api/fftw3.h -2 +2 Mon Jan 9 20:00:47 EST 2006 stevenj * [project @ 2006-01-10 01:00:47 by stevenj] another stab at Windows DLL mess M ./api/fftw3.h -172 +185 Mon Jan 9 19:23:42 EST 2006 athena * [project @ 2006-01-10 00:23:42 by athena] 64-bit clean SIMD header file. I missed those because sparse does not know vector types. Grrr... M ./simd/simd-altivec.h -11 +11 M ./simd/simd-sse.h -6 +6 M ./simd/simd-sse2.h -4 +4 Mon Jan 9 19:08:36 EST 2006 stevenj * [project @ 2006-01-10 00:08:36 by stevenj] this option is called AC_DISABLE_SHARED in the documentation M ./configure.ac -1 +1 Mon Jan 9 17:34:13 EST 2006 stevenj * [project @ 2006-01-09 22:34:13 by stevenj] fixed --with-gcc-arch to work when cross-compiling M ./m4/ax_gcc_archflag.m4 -3 +4 Mon Jan 9 12:04:04 EST 2006 athena * [project @ 2006-01-09 17:04:04 by athena] Moved the timeout check back into the search loop, sicut erat in principio. This gives us a precise control over the timeout. To avoid the overhead of X(seconds)(), only call X(seconds)() if some time measurement was taken since the last call to X(seconds)(). M ./api/apiplan.c -11 +1 M ./kernel/ifftw.h -1 +2 M ./kernel/planner.c -5 +36 Mon Jan 9 00:07:40 EST 2006 stevenj * [project @ 2006-01-09 05:07:40 by stevenj] comments M ./rdft/vrank3-transpose.c -5 +5 Sun Jan 8 23:58:23 EST 2006 stevenj * [project @ 2006-01-09 04:58:23 by stevenj] generalized transpose-cut routine to be able to call transpose-gcd recursivly; TOMS follow-the-cycles algorithm now seems to be completely superseded M ./rdft/vrank3-transpose.c -111 +142 Sun Jan 8 20:53:18 EST 2006 stevenj * [project @ 2006-01-09 01:53:18 by stevenj] M ./threads/threads.c -3 +2 Sun Jan 8 20:52:16 EST 2006 stevenj * [project @ 2006-01-09 01:52:16 by stevenj] ignore errors from setscope -- POSIX standard does not require PTHREAD_SCOPE_SYSTEM to be supported, and PTHREAD_SCOPE_PROCESS is usually okay in that case M ./threads/threads.c -7 +9 Sun Jan 8 15:58:40 EST 2006 stevenj * [project @ 2006-01-08 20:58:40 by stevenj] added TODO comment M ./rdft/vrank3-transpose.c -1 +7 Sun Jan 8 15:39:28 EST 2006 stevenj * [project @ 2006-01-08 20:39:28 by stevenj] whoops M ./rdft/vrank3-transpose.c -8 +18 Sun Jan 8 14:44:23 EST 2006 athena * [project @ 2006-01-08 19:44:23 by athena] Boasted ``much faster altivec performance''. M ./NEWS -1 +1 Sun Jan 8 11:44:52 EST 2006 athena * [project @ 2006-01-08 16:44:52 by athena] Added a new pass to the generator to schedule for the pipeline latency. (This schedule modifies the ``optimal'' cache-oblivious schedule and hence it uses more registers.) This pass is currently: * disabled for non-fma code, under the assumption that this will run on a register-starved fma. * enabled for non-simd fma code, under the assumption that this will run on a processor with 32 or more FP registers. The latency of 4 is conservative and does not introduce too much register pressure. * enabled for simd fma code, under the assumption that this will run on altivec. The latency of 8 seems to produce the best results. M ./configure.ac -1 +1 M ./dft/simd/codelets/Makefile.am -1 +1 M ./genfft/annotate.ml -3 +58 M ./genfft/magic.ml -3 +5 M ./genfft/schedule.ml -33 +5 M ./support/Makefile.codelets -2 +2 M ./support/twovers.sh -1 +1 Sun Jan 8 03:13:53 EST 2006 stevenj * [project @ 2006-01-08 08:13:53 by stevenj] fixed estimator for vrank3-transpose M ./rdft/vrank3-transpose.c -6 +22 Sun Jan 8 02:02:11 EST 2006 stevenj * [project @ 2006-01-08 07:02:11 by stevenj] more detail on VC++ workaround M ./NEWS -1 +2 Sun Jan 8 00:19:19 EST 2006 stevenj * [project @ 2006-01-08 05:19:19 by stevenj] typo M ./rdft/vrank3-transpose.c -2 +2 Sun Jan 8 00:16:20 EST 2006 stevenj * [project @ 2006-01-08 05:16:20 by stevenj] screw it, just use planner for all sub-transposes in vrank3-transpose (still just use memcpy for contiguous copies, though) M ./rdft/vrank3-transpose.c -95 +185 Sat Jan 7 23:13:45 EST 2006 stevenj * [project @ 2006-01-08 04:13:45 by stevenj] add an assert M ./kernel/tile2d.c +2 Sat Jan 7 21:57:34 EST 2006 stevenj * [project @ 2006-01-08 02:57:34 by stevenj] vrank3-transpose now uses planner to decide whether to use cpy2d, cpy2d_tiled, etc. M ./kernel/ifftw.h -1 +7 M ./rdft/rank0.c -7 +5 M ./rdft/vrank3-transpose.c -217 +231 Sat Jan 7 20:57:16 EST 2006 stevenj * [project @ 2006-01-08 01:57:16 by stevenj] too annoying to have isqrt unexpectedly fail for n==0 M ./kernel/primes.c -2 +4 Sat Jan 7 17:49:37 EST 2006 stevenj * [project @ 2006-01-07 22:49:37 by stevenj] clarifications M ./NEWS -1 +1 M ./doc/fftw3.texi -2 +5 Sat Jan 7 16:39:20 EST 2006 stevenj * [project @ 2006-01-07 21:39:20 by stevenj] comment fix M ./rdft/vrank3-transpose.c -4 +4 Sat Jan 7 15:16:22 EST 2006 stevenj * [project @ 2006-01-07 20:16:22 by stevenj] more faq updates M ./doc/FAQ/fftw-faq.bfnn -8 +9 Sat Jan 7 15:12:16 EST 2006 stevenj * [project @ 2006-01-07 20:12:16 by stevenj] enable fma on hppa, update FAQ entry M ./configure.ac +1 M ./doc/FAQ/fftw-faq.bfnn -3 +11 Sat Aug 12 23:47:06 EDT 2006 Unknown tagger tagged fftw-3-1-beta1 Sat Jan 7 14:06:31 EST 2006 athena * [project @ 2006-01-07 19:06:31 by athena] Accomodate different semantics of 'const' in C and C++ M ./dft/simd/t.c +4 Fri Jan 6 23:40:53 EST 2006 athena * [project @ 2006-01-07 04:40:53 by athena] Altivec is called VMX in IBM land. M ./NEWS -1 +1 Fri Jan 6 23:40:16 EST 2006 athena * [project @ 2006-01-07 04:40:16 by athena] Noted faster altivec support. M ./NEWS +2 Fri Jan 6 21:49:10 EST 2006 stevenj * [project @ 2006-01-07 02:49:10 by stevenj] updated icc flag detection M ./m4/ax_cc_maxopt.m4 -1 +1 Fri Jan 6 10:01:50 EST 2006 athena * [project @ 2006-01-06 15:01:50 by athena] Note ``memoize triggen''. M ./TODO -2 +1 Fri Jan 6 09:36:51 EST 2006 athena * [project @ 2006-01-06 14:36:51 by athena] Use --enable-threads to generate dependencies in the threads/ directory. M ./mkdist.sh -1 +1 Fri Jan 6 09:26:29 EST 2006 athena * [project @ 2006-01-06 14:26:29 by athena] Workaround to icc #defining __GNUC__. M ./kernel/ifftw.h -14 +14 Fri Jan 6 09:21:19 EST 2006 athena * [project @ 2006-01-06 14:21:19 by athena] Switched name to 3.1-beta1. M ./configure.ac -1 +1 Thu Jan 5 23:08:44 EST 2006 athena * [project @ 2006-01-06 04:08:44 by athena] More thoughts. M ./TODO +12 Thu Jan 5 22:30:51 EST 2006 athena * [project @ 2006-01-06 03:30:51 by athena] Note wish that (block_size % 4) == 0. M ./TODO +3 Thu Jan 5 22:19:09 EST 2006 athena * [project @ 2006-01-06 03:19:09 by athena] Check alignment of mstart, mcount in SIMD codelets. M ./dft/codelet-dft.h -2 +3 M ./dft/codelets/t.c -2 +3 M ./dft/ctsq.c -2 +3 M ./dft/dftw-direct.c -11 +17 M ./dft/k7/k7.c -3 +3 M ./dft/simd/q1b.c -1 +3 M ./dft/simd/q1f.c -1 +3 M ./dft/simd/t.c -9 +11 M ./dft/simd/t1s.c -3 +5 M ./threads/ct.c -3 +1 M ./threads/hc2hc.c -2 Thu Jan 5 21:56:19 EST 2006 athena * [project @ 2006-01-06 02:56:19 by athena] Enable threads at bootstrap time, so I get the compiler warnings that I would otherwise ignore. M ./bootstrap.sh -1 +1 Thu Jan 5 18:23:15 EST 2006 athena * [project @ 2006-01-05 23:23:15 by athena] made compilable by c++ M ./threads/dft-vrank-geq1.c -2 +2 M ./threads/rdft-vrank-geq1.c -2 +2 M ./threads/vrank-geq1-rdft2.c -2 +2 Thu Jan 5 17:39:02 EST 2006 athena * [project @ 2006-01-05 22:39:02 by athena] FIXED: incorrect twiddle_shift() M ./kernel/twiddle.c -13 +19 Thu Jan 5 16:01:51 EST 2006 athena * [project @ 2006-01-05 21:01:51 by athena] Replaced remnants of awake flag with the new enum wakefulness type. M ./reodft/redft00e-r2hc.c -4 +5 M ./reodft/reodft11e-r2hc.c -5 +7 M ./reodft/rodft00e-r2hc.c -4 +5 M ./threads/ct.c -4 +4 M ./threads/dft-vrank-geq1.c -3 +3 M ./threads/hc2hc.c -3 +3 M ./threads/rdft-vrank-geq1.c -3 +3 M ./threads/vrank-geq1-rdft2.c -3 +3 Thu Jan 5 11:20:59 EST 2006 athena * [project @ 2006-01-05 16:20:59 by athena] Oops---there is no need to find a free slot. M ./kernel/planner.c -10 +1 Thu Jan 5 09:41:58 EST 2006 athena * [project @ 2006-01-05 14:41:58 by athena] Assertions. M ./kernel/planner.c -1 +3 Thu Jan 5 09:29:55 EST 2006 athena * [project @ 2006-01-05 14:29:55 by athena] Commented the hash table lookup algorithm. M ./kernel/planner.c -10 +14 Thu Jan 5 09:12:00 EST 2006 athena * [project @ 2006-01-05 14:12:00 by athena] Fixed infinite loop in hashtable lookup/insert. Grrr... M ./kernel/planner.c -12 +31 Wed Jan 4 22:04:28 EST 2006 stevenj * [project @ 2006-01-05 03:04:26 by stevenj] updated copyright years to 2006 M ./COPYRIGHT -2 +2 M ./api/api.h -2 +2 M ./api/apiplan.c -2 +2 M ./api/configure.c -2 +2 M ./api/execute-dft-c2r.c -2 +2 M ./api/execute-dft-r2c.c -2 +2 M ./api/execute-dft.c -2 +2 M ./api/execute-r2r.c -2 +2 M ./api/execute-split-dft-c2r.c -2 +2 M ./api/execute-split-dft-r2c.c -2 +2 M ./api/execute-split-dft.c -2 +2 M ./api/execute.c -2 +2 M ./api/export-wisdom-to-file.c -2 +2 M ./api/export-wisdom-to-string.c -2 +2 M ./api/export-wisdom.c -2 +2 M ./api/extract-reim.c -2 +2 M ./api/f77api.c -2 +2 M ./api/f77funcs.h -2 +2 M ./api/fftw3.h -3 +3 M ./api/flops.c -2 +2 M ./api/forget-wisdom.c -2 +2 M ./api/import-system-wisdom.c -2 +2 M ./api/import-wisdom-from-file.c -2 +2 M ./api/import-wisdom-from-string.c -2 +2 M ./api/import-wisdom.c -2 +2 M ./api/malloc.c -2 +2 M ./api/map-r2r-kind.c -2 +2 M ./api/mapflags.c -2 +2 M ./api/mkprinter-file.c -2 +2 M ./api/mktensor-iodims.c -2 +2 M ./api/mktensor-rowmajor.c -2 +2 M ./api/plan-dft-1d.c -2 +2 M ./api/plan-dft-2d.c -2 +2 M ./api/plan-dft-3d.c -2 +2 M ./api/plan-dft-c2r-1d.c -2 +2 M ./api/plan-dft-c2r-2d.c -2 +2 M ./api/plan-dft-c2r-3d.c -2 +2 M ./api/plan-dft-c2r.c -2 +2 M ./api/plan-dft-r2c-1d.c -2 +2 M ./api/plan-dft-r2c-2d.c -2 +2 M ./api/plan-dft-r2c-3d.c -2 +2 M ./api/plan-dft-r2c.c -2 +2 M ./api/plan-dft.c -2 +2 M ./api/plan-guru-dft-c2r.c -2 +2 M ./api/plan-guru-dft-r2c.c -2 +2 M ./api/plan-guru-dft.c -2 +2 M ./api/plan-guru-r2r.c -2 +2 M ./api/plan-guru-split-dft-c2r.c -2 +2 M ./api/plan-guru-split-dft-r2c.c -2 +2 M ./api/plan-guru-split-dft.c -2 +2 M ./api/plan-many-dft-c2r.c -2 +2 M ./api/plan-many-dft-r2c.c -2 +2 M ./api/plan-many-dft.c -2 +2 M ./api/plan-many-r2r.c -2 +2 M ./api/plan-r2r-1d.c -2 +2 M ./api/plan-r2r-2d.c -2 +2 M ./api/plan-r2r-3d.c -2 +2 M ./api/plan-r2r.c -2 +2 M ./api/print-plan.c -2 +2 M ./api/rdft2-pad.c -2 +2 M ./api/the-planner.c -2 +2 M ./api/version.c -3 +3 M ./api/x77.h -2 +2 M ./dft/bluestein.c -2 +2 M ./dft/buffered.c -3 +3 M ./dft/codelet-dft.h -3 +3 M ./dft/codelets/n.c -2 +2 M ./dft/codelets/n.h -2 +2 M ./dft/codelets/t.c -2 +2 M ./dft/codelets/t.h -2 +2 M ./dft/conf.c -3 +3 M ./dft/ct.c -3 +3 M ./dft/ct.h -2 +2 M ./dft/ctsq.c -3 +3 M ./dft/dft.h -3 +3 M ./dft/dftw-direct.c -3 +3 M ./dft/dftw-generic.c -2 +2 M ./dft/dftw-genericbuf.c -2 +2 M ./dft/direct.c -3 +3 M ./dft/generic.c -2 +2 M ./dft/indirect-transpose.c -2 +2 M ./dft/indirect.c -3 +3 M ./dft/k7/k7.c -3 +3 M ./dft/kdft-dif.c -3 +3 M ./dft/kdft-difsq.c -3 +3 M ./dft/kdft-dit.c -3 +3 M ./dft/kdft.c -3 +3 M ./dft/nop.c -3 +3 M ./dft/plan.c -3 +3 M ./dft/problem.c -3 +3 M ./dft/rader.c -2 +2 M ./dft/rank-geq2.c -3 +3 M ./dft/simd/n1b.c -2 +2 M ./dft/simd/n1b.h -2 +2 M ./dft/simd/n1f.c -2 +2 M ./dft/simd/n1f.h -2 +2 M ./dft/simd/n2b.c -2 +2 M ./dft/simd/n2b.h -2 +2 M ./dft/simd/n2f.c -2 +2 M ./dft/simd/n2f.h -2 +2 M ./dft/simd/n2s.c -2 +2 M ./dft/simd/n2s.h -2 +2 M ./dft/simd/q1b.c -2 +2 M ./dft/simd/q1b.h -2 +2 M ./dft/simd/q1f.c -2 +2 M ./dft/simd/q1f.h -2 +2 M ./dft/simd/t.c -2 +2 M ./dft/simd/t1b.h -2 +2 M ./dft/simd/t1f.h -2 +2 M ./dft/simd/t1s.c -2 +2 M ./dft/simd/t1s.h -2 +2 M ./dft/simd/t2b.h -2 +2 M ./dft/simd/t2f.h -2 +2 M ./dft/solve.c -3 +3 M ./dft/vrank-geq1.c -3 +3 M ./dft/zero.c -3 +3 M ./doc/f77_wisdom.f -2 +2 M ./doc/fftw3.texi -1 +1 M ./genfft/algsimp.ml -4 +4 M ./genfft/algsimp.mli -3 +3 M ./genfft/annotate.ml -4 +4 M ./genfft/annotate.mli -3 +3 M ./genfft/assoctable.ml -3 +3 M ./genfft/assoctable.mli -3 +3 M ./genfft/c.ml -3 +3 M ./genfft/c.mli -3 +3 M ./genfft/complex.ml -3 +3 M ./genfft/complex.mli -3 +3 M ./genfft/conv.ml -2 +2 M ./genfft/conv.mli -3 +3 M ./genfft/dag.ml -3 +3 M ./genfft/dag.mli -3 +3 M ./genfft/expr.ml -3 +3 M ./genfft/expr.mli -3 +3 M ./genfft/fft.ml -4 +4 M ./genfft/fft.mli -3 +3 M ./genfft/gen_athnotw.ml -4 +4 M ./genfft/gen_athtw.ml -4 +4 M ./genfft/gen_conv.ml -4 +4 M ./genfft/gen_hc2hc.ml -4 +4 M ./genfft/gen_hc2r.ml -4 +4 M ./genfft/gen_mdct.ml -4 +4 M ./genfft/gen_notw.ml -4 +4 M ./genfft/gen_notw_c.ml -4 +4 M ./genfft/gen_r2hc.ml -4 +4 M ./genfft/gen_r2r.ml -4 +4 M ./genfft/gen_twiddle.ml -4 +4 M ./genfft/gen_twiddle_c.ml -4 +4 M ./genfft/gen_twidsq.ml -4 +4 M ./genfft/gen_twidsq_c.ml -4 +4 M ./genfft/genutil.ml -3 +3 M ./genfft/littlesimp.ml -3 +3 M ./genfft/littlesimp.mli -3 +3 M ./genfft/magic.ml -3 +3 M ./genfft/monads.ml -3 +3 M ./genfft/number.ml -3 +3 M ./genfft/number.mli -3 +3 M ./genfft/oracle.ml -3 +3 M ./genfft/oracle.mli -3 +3 M ./genfft/schedule.ml -3 +3 M ./genfft/schedule.mli -3 +3 M ./genfft/simd.ml -3 +3 M ./genfft/simd.mli -3 +3 M ./genfft/simdmagic.ml -3 +3 M ./genfft/to_alist.ml -3 +3 M ./genfft/to_alist.mli -3 +3 M ./genfft/trig.ml -3 +3 M ./genfft/trig.mli -3 +3 M ./genfft/twiddle.ml -3 +3 M ./genfft/twiddle.mli -3 +3 M ./genfft/unique.ml -3 +3 M ./genfft/unique.mli -3 +3 M ./genfft/util.ml -3 +3 M ./genfft/util.mli -3 +3 M ./genfft/variable.ml -3 +3 M ./genfft/variable.mli -3 +3 M ./genfft-k7/algsimp.ml -4 +4 M ./genfft-k7/algsimp.mli -3 +3 M ./genfft-k7/assoctable.ml -3 +3 M ./genfft-k7/assoctable.mli -3 +3 M ./genfft-k7/complex.ml -1 +1 M ./genfft-k7/complex.mli -1 +1 M ./genfft-k7/expr.ml -3 +3 M ./genfft-k7/expr.mli -3 +3 M ./genfft-k7/fft.ml -4 +4 M ./genfft-k7/gen_notw.ml -1 +1 M ./genfft-k7/littlesimp.ml -3 +3 M ./genfft-k7/littlesimp.mli -3 +3 M ./genfft-k7/monads.ml -3 +3 M ./genfft-k7/number.ml -3 +3 M ./genfft-k7/number.mli -3 +3 M ./genfft-k7/oracle.ml -3 +3 M ./genfft-k7/oracle.mli -3 +3 M ./genfft-k7/to_alist.ml -3 +3 M ./genfft-k7/to_alist.mli -3 +3 M ./genfft-k7/twiddle.ml -3 +3 M ./genfft-k7/twiddle.mli -3 +3 M ./genfft-k7/vScheduler.mli -1 +1 M ./kernel/align.c -3 +3 M ./kernel/alloc.c -3 +3 M ./kernel/assert.c -3 +3 M ./kernel/awake.c -3 +3 M ./kernel/buffered.c -2 +2 M ./kernel/cpy1d.c -2 +2 M ./kernel/cpy2d-pair.c -2 +2 M ./kernel/cpy2d.c -2 +2 M ./kernel/ct.c -2 +2 M ./kernel/cycle.h -3 +3 M ./kernel/debug.c -3 +3 M ./kernel/hash.c -2 +2 M ./kernel/iabs.c -3 +3 M ./kernel/ifftw.h -3 +3 M ./kernel/kalloc.c -3 +3 M ./kernel/md5-1.c -2 +2 M ./kernel/md5.c -2 +2 M ./kernel/minmax.c -3 +3 M ./kernel/ops.c -3 +3 M ./kernel/pickdim.c -3 +3 M ./kernel/plan.c -3 +3 M ./kernel/primes.c -3 +3 M ./kernel/print.c -3 +3 M ./kernel/problem.c -3 +3 M ./kernel/rader.c -2 +2 M ./kernel/scan.c -3 +3 M ./kernel/solver.c -3 +3 M ./kernel/solvtab.c -3 +3 M ./kernel/stride.c -3 +3 M ./kernel/tensor.c -3 +3 M ./kernel/tensor1.c -3 +3 M ./kernel/tensor2.c -3 +3 M ./kernel/tensor4.c -3 +3 M ./kernel/tensor5.c -3 +3 M ./kernel/tensor7.c -3 +3 M ./kernel/tensor8.c -3 +3 M ./kernel/tensor9.c -3 +3 M ./kernel/tile2d.c -2 +2 M ./kernel/timer.c -3 +3 M ./kernel/transpose.c -2 +2 M ./kernel/trig.c -3 +3 M ./kernel/twiddle.c -3 +3 M ./libbench/accopy-from.c -1 +1 M ./libbench/accopy-to.c -1 +1 M ./libbench/allocate.c -1 +1 M ./libbench/bench-main.c -1 +1 M ./libbench/bench-user.h -1 +1 M ./libbench/bench.h -1 +1 M ./libbench/can-do.c -1 +1 M ./libbench/ccopy-from.c -1 +1 M ./libbench/ccopy-to.c -1 +1 M ./libbench/deallocate.c -1 +1 M ./libbench/getopt-utils.c -1 +1 M ./libbench/info.c -1 +1 M ./libbench/main.c -1 +1 M ./libbench/prime.c -1 +1 M ./libbench/problem.c -1 +1 M ./libbench/report.c -1 +1 M ./libbench/speed.c -1 +1 M ./libbench/timer.c -1 +1 M ./libbench/verify.c -1 +1 M ./libbench/zero.c -1 +1 M ./libbench2/aligned-main.c -1 +1 M ./libbench2/allocate.c -1 +1 M ./libbench2/can-do.c -1 +1 M ./libbench2/dotens2.c -3 +3 M ./libbench2/getopt-utils.c -1 +1 M ./libbench2/info.c -1 +1 M ./libbench2/main.c -1 +1 M ./libbench2/report.c -1 +1 M ./libbench2/tensor.c -1 +1 M ./libbench2/useropt.c -1 +1 M ./libbench2/verify-dft.c -3 +3 M ./libbench2/verify-lib.c -3 +3 M ./libbench2/verify-r2r.c -2 +2 M ./libbench2/verify-rdft2.c -3 +3 M ./libbench2/verify.c -1 +1 M ./libbench2/verify.h -2 +2 M ./libbench2/zero.c -1 +1 M ./m4/ax_gcc_archflag.m4 -1 +1 M ./rdft/buffered.c -3 +3 M ./rdft/buffered2.c -3 +3 M ./rdft/codelet-rdft.h -2 +2 M ./rdft/codelets/hb.h -2 +2 M ./rdft/codelets/hc2r.c -2 +2 M ./rdft/codelets/hc2r.h -2 +2 M ./rdft/codelets/hc2rIII.h -2 +2 M ./rdft/codelets/hf.h -2 +2 M ./rdft/codelets/hfb.c -2 +2 M ./rdft/codelets/r2hc.c -2 +2 M ./rdft/codelets/r2hc.h -2 +2 M ./rdft/codelets/r2hcII.h -2 +2 M ./rdft/codelets/r2r.c -2 +2 M ./rdft/codelets/r2r.h -2 +2 M ./rdft/conf.c -3 +3 M ./rdft/dft-r2hc.c -3 +3 M ./rdft/dht-r2hc.c -3 +3 M ./rdft/dht-rader.c -2 +2 M ./rdft/direct.c -3 +3 M ./rdft/direct2.c -3 +3 M ./rdft/generic.c -2 +2 M ./rdft/hc2hc-common.c -2 +2 M ./rdft/hc2hc-direct.c -3 +3 M ./rdft/hc2hc-directbuf.c -3 +3 M ./rdft/hc2hc-generic.c -2 +2 M ./rdft/hc2hc.c -2 +2 M ./rdft/hc2hc.h -2 +2 M ./rdft/indirect.c -3 +3 M ./rdft/khc2hc.c -3 +3 M ./rdft/khc2r.c -3 +3 M ./rdft/kr2hc.c -3 +3 M ./rdft/kr2r.c -3 +3 M ./rdft/nop.c -3 +3 M ./rdft/nop2.c -3 +3 M ./rdft/plan.c -3 +3 M ./rdft/plan2.c -3 +3 M ./rdft/problem.c -3 +3 M ./rdft/problem2.c -3 +3 M ./rdft/rank-geq2-rdft2.c -3 +3 M ./rdft/rank-geq2.c -3 +3 M ./rdft/rank0-rdft2.c -3 +3 M ./rdft/rank0.c -3 +3 M ./rdft/rdft-dht.c -3 +3 M ./rdft/rdft.h -2 +2 M ./rdft/rdft2-inplace-strides.c -3 +3 M ./rdft/rdft2-radix2.c -3 +3 M ./rdft/rdft2-strides.c -2 +2 M ./rdft/rdft2-tensor-max-index.c -3 +3 M ./rdft/solve.c -3 +3 M ./rdft/solve2.c -3 +3 M ./rdft/vrank-geq1-rdft2.c -3 +3 M ./rdft/vrank-geq1.c -3 +3 M ./rdft/vrank3-transpose.c -3 +3 M ./reodft/conf.c -3 +3 M ./reodft/redft00e-r2hc-pad.c -3 +3 M ./reodft/redft00e-r2hc.c -3 +3 M ./reodft/reodft.h -2 +2 M ./reodft/reodft010e-r2hc.c -3 +3 M ./reodft/reodft11e-r2hc-odd.c -3 +3 M ./reodft/reodft11e-r2hc.c -3 +3 M ./reodft/reodft11e-radix2.c -3 +3 M ./reodft/rodft00e-r2hc-pad.c -3 +3 M ./reodft/rodft00e-r2hc.c -3 +3 M ./simd/altivec.c -3 +3 M ./simd/nonportable/sse.c -3 +3 M ./simd/nonportable/sse2.c -3 +3 M ./simd/simd-altivec.h -2 +2 M ./simd/simd-sse.h -2 +2 M ./simd/simd-sse2.h -2 +2 M ./simd/simd.h -2 +2 M ./simd/sse.c -3 +3 M ./simd/sse2.c -3 +3 M ./simd/taint.c -3 +3 M ./threads/api.c -2 +2 M ./threads/conf.c -3 +3 M ./threads/ct.c -3 +3 M ./threads/dft-vrank-geq1.c -3 +3 M ./threads/f77api.c -2 +2 M ./threads/f77funcs.h -2 +2 M ./threads/hc2hc.c -2 +2 M ./threads/rdft-vrank-geq1.c -3 +3 M ./threads/threads.c -2 +2 M ./threads/threads.h -2 +2 M ./threads/vrank-geq1-rdft2.c -3 +3 M ./tools/fftw-wisdom-to-conf.1 -4 +4 M ./tools/fftw-wisdom-to-conf.in -2 +2 M ./tools/fftw-wisdom.c -2 +2 M ./tools/fftw_wisdom.1.in -4 +4 Wed Jan 4 21:57:23 EST 2006 stevenj * [project @ 2006-01-05 02:57:23 by stevenj] whoops M ./m4/ax_gcc_archflag.m4 -7 +4 Wed Jan 4 21:52:18 EST 2006 stevenj * [project @ 2006-01-05 02:52:18 by stevenj] whoops M ./m4/ax_gcc_archflag.m4 -1 +1 Wed Jan 4 21:51:40 EST 2006 stevenj * [project @ 2006-01-05 02:51:40 by stevenj] more updates for recent pentia/amd M ./m4/ax_gcc_archflag.m4 -5 +12 Wed Jan 4 20:57:47 EST 2006 athena * [project @ 2006-01-05 01:57:47 by athena] Pruned TODO. M ./TODO -6 +2 Wed Jan 4 20:43:41 EST 2006 athena * [project @ 2006-01-05 01:43:41 by athena] Prototype of problem_destroy() M ./libbench2/bench-user.h -1 +2 M ./libbench2/bench.h -2 +1 Wed Jan 4 20:43:13 EST 2006 stevenj * [project @ 2006-01-05 01:43:13 by stevenj] rm obsoleted TODOs M ./TODO -5 Wed Jan 4 20:37:24 EST 2006 athena * [project @ 2006-01-05 01:37:24 by athena] Fallback to 970 if neither -mcpu=power5 nor -mcpu=power4 are supported. M ./m4/ax_gcc_archflag.m4 -3 +3 Wed Jan 4 20:29:07 EST 2006 stevenj * [project @ 2006-01-05 01:29:07 by stevenj] NEWS updates, clarifications, and reorganization M ./NEWS -6 +10 Wed Jan 4 19:54:41 EST 2006 stevenj * [project @ 2006-01-05 00:54:40 by stevenj] remove some compiler warnings, add an assert check, make estimator work properly for nop plans M ./dft/dftw-genericbuf.c +3 M ./kernel/planner.c -1 +2 M ./kernel/trig.c -1 +3 M ./m4/ax_gcc_x86_cpuid.m4 +2 M ./rdft/dft-r2hc.c -1 +2 Tue Jan 3 19:34:04 EST 2006 athena * [project @ 2006-01-04 00:34:03 by athena] Two big changes: 1) revised the twiddle generation machinery, to avoid generating twiddles when measuring, and to use a faster O(sqrt(N)) table when this entails no loss of precision. 2) implemented new ALLOW_PRUNING estimator hack. M ./api/apiplan.c -2 +10 M ./api/fftw3.h -1 +2 M ./api/mapflags.c -1 +4 M ./configure.ac -8 +9 M ./dft/bluestein.c -12 +18 M ./dft/buffered.c -5 +5 M ./dft/ct.c -4 +7 M ./dft/ctsq.c -4 +4 M ./dft/dftw-direct.c -3 +4 M ./dft/dftw-generic.c -131 +19 M ./dft/dftw-genericbuf.c -76 +36 M ./dft/direct.c -1 +2 M ./dft/generic.c -2 +2 M ./dft/indirect-transpose.c -4 +4 M ./dft/indirect.c -4 +4 M ./dft/rader.c -17 +22 M ./dft/rank-geq2.c -4 +4 M ./dft/vrank-geq1.c -3 +3 M ./genfft/twiddle.ml -22 +12 M ./kernel/awake.c -3 +3 M ./kernel/ifftw.h -23 +48 M ./kernel/plan.c -14 +9 M ./kernel/planner.c -1 +8 M ./kernel/timer.c -3 +3 M ./kernel/trig.c -41 +166 M ./kernel/twiddle.c -64 +103 M ./libbench2/bench-main.c -3 +8 M ./libbench2/bench.h -2 +2 M ./libbench2/problem.c -4 +3 M ./libbench2/speed.c -2 +8 M ./rdft/buffered.c -5 +5 M ./rdft/buffered2.c -4 +4 M ./rdft/dft-r2hc.c -3 +3 M ./rdft/dht-r2hc.c -3 +3 M ./rdft/dht-rader.c -16 +23 M ./rdft/direct.c -1 +3 M ./rdft/direct2.c -1 +2 M ./rdft/generic.c -2 +2 M ./rdft/hc2hc-direct.c -5 +7 M ./rdft/hc2hc-directbuf.c -5 +5 M ./rdft/hc2hc-generic.c -6 +7 M ./rdft/hc2hc.c -3 +7 M ./rdft/indirect.c -4 +4 M ./rdft/rank-geq2-rdft2.c -4 +4 M ./rdft/rank-geq2.c -4 +4 M ./rdft/rank0-rdft2.c -3 +3 M ./rdft/rdft-dht.c -3 +3 M ./rdft/rdft2-radix2.c -4 +5 M ./rdft/vrank-geq1-rdft2.c -3 +3 M ./rdft/vrank-geq1.c -3 +3 M ./reodft/redft00e-r2hc-pad.c -4 +4 M ./reodft/reodft00e-splitradix.c -5 +6 M ./reodft/reodft010e-r2hc.c -4 +5 M ./reodft/reodft11e-r2hc-odd.c -3 +3 M ./reodft/reodft11e-radix2.c -5 +7 M ./reodft/rodft00e-r2hc-pad.c -4 +4 M ./tests/hook.c -3 +6 Sat Dec 24 22:08:29 EST 2005 athena * [project @ 2005-12-25 03:08:29 by athena] Estimator tweaks, mostly to favor generic over rader for small n. M ./dft/generic.c +2 M ./rdft/generic.c +2 Sat Dec 24 17:55:47 EST 2005 athena * [project @ 2005-12-24 22:55:47 by athena] Grrr... missing break statement in switch. M ./tests/hook.c +1 Sat Dec 24 16:08:50 EST 2005 athena * [project @ 2005-12-24 21:08:49 by athena] Swapped fields TW and OPS in struct ct_desc_s, to make k7 asm code insensitive to -malign-double. For consistency, changed struct hc2hc_desc_s in the same way. M ./dft/codelet-dft.h -2 +2 M ./genfft/gen_hc2hc.ml -3 +3 M ./genfft/gen_twiddle.ml -3 +3 M ./genfft/gen_twiddle_c.ml -3 +3 M ./genfft/gen_twidsq.ml -3 +3 M ./genfft/gen_twidsq_c.ml -3 +3 M ./genfft-k7/gen_twiddle.ml -3 +2 M ./rdft/codelet-rdft.h -1 +1 Sat Dec 24 16:00:42 EST 2005 athena * [project @ 2005-12-24 21:00:42 by athena] Wrong check for infeasible slvndx in imprt(). M ./kernel/planner.c -4 +4 Sat Dec 24 15:56:59 EST 2005 athena * [project @ 2005-12-24 20:56:59 by athena] Removed obsolete function invoke_solver_if_correct_kind(). M ./kernel/planner.c -10 +1 Sat Dec 24 14:22:12 EST 2005 athena * [project @ 2005-12-24 19:22:12 by athena] Faster implementation of safe_mulmod(), avoiding divisions altogether. Works for 0 <= p <= INT_MAX. M ./kernel/primes.c -13 +22 Sat Dec 24 12:05:54 EST 2005 athena * [project @ 2005-12-24 17:05:54 by athena] FFTW_ALLOW_LARGE_GENERIC must belong to flags->l, it cannot be overridden by fftw. M ./api/mapflags.c -2 +2 Fri Dec 23 20:46:24 EST 2005 stevenj * [project @ 2005-12-24 01:46:24 by stevenj] no more need for limits.h, add some explanatory comments M ./kernel/primes.c -4 +9 Fri Dec 23 17:50:25 EST 2005 athena * [project @ 2005-12-23 22:50:25 by athena] Paranoia. M ./dft/k7/k7.c -1 +8 Fri Dec 23 17:40:41 EST 2005 athena * [project @ 2005-12-23 22:40:41 by athena] Fixed subtle bug involving overflow of the slvndx field in flags_t. M ./kernel/ifftw.h -5 +6 M ./kernel/planner.c -18 +26 Fri Dec 23 16:33:56 EST 2005 athena * [project @ 2005-12-23 21:33:56 by athena] Note 64-bit clean. M ./NEWS +2 Fri Dec 23 15:34:32 EST 2005 athena * [project @ 2005-12-23 20:34:32 by athena] Threads are now 64-bit clean M ./threads/ct.c -12 +13 M ./threads/dft-vrank-geq1.c -20 +17 M ./threads/hc2hc.c -7 +8 M ./threads/rdft-vrank-geq1.c -18 +15 M ./threads/threads.h -2 +2 M ./threads/vrank-geq1-rdft2.c -20 +20 Fri Dec 23 13:00:31 EST 2005 athena * [project @ 2005-12-23 18:00:31 by athena] Restored the old numbering TW_NEXT=3 etc, because the k7 code depends on it. M ./kernel/ifftw.h -3 +3 Fri Dec 23 11:58:00 EST 2005 athena * [project @ 2005-12-23 16:58:00 by athena] Portable implementation of MULMOD() and safe_mulmod(). Removed all unnecessary AC_CHECK_SIZEOF() from configure.ac. M ./configure.ac -5 M ./kernel/ifftw.h -27 +3 M ./kernel/primes.c -8 +3 Thu Dec 22 11:12:29 EST 2005 athena * [project @ 2005-12-22 16:12:29 by athena] Inline the loop body in r2r codelets like we do everywhere else. M ./genfft/gen_r2r.ml -44 +34 Thu Dec 22 10:48:53 EST 2005 athena * [project @ 2005-12-22 15:48:53 by athena] Oops. M ./dft/conf.c -3 +3 Thu Dec 22 10:25:15 EST 2005 athena * [project @ 2005-12-22 15:25:15 by athena] Renamed X(sin_and_cos)() to X(cexp)(). M ./dft/bluestein.c -1 +1 M ./dft/dftw-generic.c -2 +2 M ./dft/dftw-genericbuf.c -2 +2 M ./dft/rader.c -1 +1 M ./kernel/ifftw.h -2 +2 M ./kernel/trig.c -2 +2 M ./kernel/twiddle.c -4 +3 M ./rdft/dht-rader.c -1 +1 Wed Dec 21 22:49:58 EST 2005 athena * [project @ 2005-12-22 03:49:58 by athena] Somewhat faster generation of twiddle factors. M ./dft/bluestein.c -2 +1 M ./dft/conf.c -3 +3 M ./dft/dftw-generic.c -8 +5 M ./dft/dftw-genericbuf.c -8 +6 M ./dft/rader.c -3 +4 M ./kernel/Makefile.am -1 +1 M ./kernel/ifftw.h -4 +3 M ./kernel/trig.c -5 +67 R ./kernel/trig1.c M ./kernel/twiddle.c -11 +15 M ./rdft/dht-rader.c -2 +4 Tue Dec 20 23:50:01 EST 2005 athena * [project @ 2005-12-21 04:50:01 by athena] tweaks M ./kernel/md5.c -2 +1 Tue Dec 20 22:29:19 EST 2005 athena * [project @ 2005-12-21 03:29:19 by athena] Sped up planner, esp. in estimate mode. The planner now classifies all solvers into DFT, RDFT, and RDFT2, and it only invokes solvers appropriate for the problem being planned. Because we have several hundred solvers, the overhead of calling irrelevant solvers is significant, and this modification mitigates the issue somewhat. M ./dft/bluestein.c -14 +10 M ./dft/buffered.c -31 +30 M ./dft/ct.c -16 +13 M ./dft/ctsq.c -23 +20 M ./dft/dft.h -4 +1 M ./dft/direct.c -69 +62 M ./dft/generic.c -12 +8 M ./dft/indirect-transpose.c -23 +19 M ./dft/indirect.c -37 +33 M ./dft/nop.c -16 +15 M ./dft/problem.c -6 +2 M ./dft/rader.c -11 +7 M ./dft/rank-geq2.c -13 +9 M ./dft/vrank-geq1.c -16 +12 M ./kernel/ifftw.h -1 +25 M ./kernel/planner.c -5 +29 M ./rdft/buffered.c -31 +30 M ./rdft/buffered2.c -8 +5 M ./rdft/dft-r2hc.c -10 +6 M ./rdft/dht-r2hc.c -12 +9 M ./rdft/dht-rader.c -13 +9 M ./rdft/direct.c -49 +45 M ./rdft/direct2.c -45 +41 M ./rdft/generic.c -13 +9 M ./rdft/hc2hc.c -19 +15 M ./rdft/indirect.c -30 +26 M ./rdft/nop.c -16 +13 M ./rdft/nop2.c -17 +15 M ./rdft/problem.c -6 +2 M ./rdft/problem2.c -6 +2 M ./rdft/rank-geq2-rdft2.c -20 +16 M ./rdft/rank-geq2.c -13 +9 M ./rdft/rank0-rdft2.c -13 +10 M ./rdft/rank0.c -14 +10 M ./rdft/rdft-dht.c -15 +13 M ./rdft/rdft.h -6 M ./rdft/rdft2-radix2.c -25 +17 M ./rdft/vrank-geq1-rdft2.c -13 +12 M ./rdft/vrank-geq1.c -15 +11 M ./rdft/vrank3-transpose.c -27 +24 M ./reodft/redft00e-r2hc-pad.c -12 +9 M ./reodft/redft00e-r2hc.c -12 +9 M ./reodft/reodft00e-splitradix.c -17 +14 M ./reodft/reodft010e-r2hc.c -12 +9 M ./reodft/reodft11e-r2hc-odd.c -12 +9 M ./reodft/reodft11e-r2hc.c -11 +9 M ./reodft/reodft11e-radix2.c -12 +9 M ./reodft/rodft00e-r2hc-pad.c -12 +8 M ./reodft/rodft00e-r2hc.c -11 +8 M ./tests/hook.c -92 +98 Mon Dec 19 22:04:00 EST 2005 athena * [project @ 2005-12-20 03:04:00 by athena] Eliminated all calls to sprintf() in favor of own routines, so as not to force users to link stdio and the associated locale/pthreads crap. M ./kernel/print.c -35 +50 Mon Dec 19 21:27:25 EST 2005 athena * [project @ 2005-12-20 02:27:25 by athena] Implemented routine to print INT, removing the need for c99's %td format. M ./kernel/ifftw.h -2 +1 M ./kernel/print.c -7 +27 Mon Dec 19 12:06:33 EST 2005 athena * [project @ 2005-12-19 17:06:33 by athena] info->n is size_t M ./kernel/alloc.c -2 +2 Sun Dec 18 18:15:04 EST 2005 athena * [project @ 2005-12-18 23:15:04 by athena] Explicit casts in front of pointer difference in printf() context, just in case INT != ptrdiff_t. M ./configure.ac -1 +1 M ./dft/problem.c -3 +3 M ./rdft/problem.c -2 +2 M ./rdft/problem2.c -3 +3 Sun Dec 18 16:52:38 EST 2005 athena * [project @ 2005-12-18 21:52:38 by athena] Forgot to add %D to print.c M ./kernel/print.c -1 +6 Sun Dec 18 16:43:26 EST 2005 athena * [project @ 2005-12-18 21:43:26 by athena] Use %D as format character for type INT. M ./dft/bluestein.c -1 +1 M ./dft/buffered.c -2 +2 M ./dft/ct.c -2 +2 M ./dft/ctsq.c -2 +2 M ./dft/dftw-direct.c -3 +3 M ./dft/dftw-generic.c -1 +1 M ./dft/dftw-genericbuf.c -1 +1 M ./dft/direct.c -3 +3 M ./dft/generic.c -1 +1 M ./dft/problem.c -2 +2 M ./dft/rader.c -1 +1 M ./dft/vrank-geq1.c -2 +2 M ./kernel/print.c -9 +1 M ./kernel/tensor.c -2 +2 M ./rdft/buffered.c -2 +2 M ./rdft/buffered2.c -2 +2 M ./rdft/dft-r2hc.c -2 +2 M ./rdft/dht-r2hc.c -2 +2 M ./rdft/dht-rader.c -1 +1 M ./rdft/direct.c -2 +2 M ./rdft/direct2.c -2 +2 M ./rdft/generic.c -1 +1 M ./rdft/hc2hc-direct.c -2 +2 M ./rdft/hc2hc-directbuf.c -2 +2 M ./rdft/hc2hc-generic.c -1 +1 M ./rdft/hc2hc.c -1 +1 M ./rdft/problem.c -2 +2 M ./rdft/problem2.c -2 +2 M ./rdft/rank0.c -2 +2 M ./rdft/rdft-dht.c -2 +2 M ./rdft/rdft2-radix2.c -2 +2 M ./rdft/vrank-geq1-rdft2.c -2 +2 M ./rdft/vrank-geq1.c -2 +2 M ./rdft/vrank3-transpose.c -2 +2 M ./reodft/redft00e-r2hc-pad.c -2 +2 M ./reodft/redft00e-r2hc.c -2 +2 M ./reodft/reodft00e-splitradix.c -3 +3 M ./reodft/reodft010e-r2hc.c -2 +2 M ./reodft/reodft11e-r2hc-odd.c -2 +2 M ./reodft/reodft11e-r2hc.c -2 +2 M ./reodft/reodft11e-radix2.c -2 +2 M ./reodft/rodft00e-r2hc-pad.c -2 +2 M ./reodft/rodft00e-r2hc.c -2 +2 Sun Dec 18 15:14:03 EST 2005 athena * [project @ 2005-12-18 20:14:03 by athena] Changed type of an_int_guaranteed_to_be_zero. Changed name as well. M ./kernel/ifftw.h -3 +3 M ./kernel/stride.c -2 +2 Sun Dec 18 14:41:31 EST 2005 athena * [project @ 2005-12-18 19:41:31 by athena] converted %o -> INT M ./kernel/ifftw.h -1 +2 M ./kernel/planner.c -2 +2 M ./kernel/print.c -8 +6 Sat Dec 17 20:28:50 EST 2005 athena * [project @ 2005-12-18 01:28:50 by athena] Major 64-bit cleanup. M ./dft/bluestein.c -14 +15 M ./dft/buffered.c -20 +20 M ./dft/codelet-dft.h -17 +17 M ./dft/codelets/n.c -1 +1 M ./dft/codelets/t.c -1 +1 M ./dft/ct.c -6 +6 M ./dft/ct.h -5 +5 M ./dft/ctsq.c -6 +6 M ./dft/dftw-direct.c -17 +17 M ./dft/dftw-generic.c -28 +28 M ./dft/dftw-genericbuf.c -25 +25 M ./dft/direct.c -16 +14 M ./dft/generic.c -8 +8 M ./dft/indirect-transpose.c -5 +5 M ./dft/problem.c -3 +3 M ./dft/rader.c -10 +10 M ./dft/simd/n1b.c -1 +1 M ./dft/simd/n1f.c -1 +1 M ./dft/simd/n2b.c -1 +1 M ./dft/simd/n2f.c -1 +1 M ./dft/simd/n2s.c -1 +1 M ./dft/simd/q1b.c -1 +1 M ./dft/simd/q1f.c -1 +1 M ./dft/simd/t.c -5 +5 M ./dft/simd/t1s.c -1 +1 M ./dft/vrank-geq1.c -5 +5 M ./dft/zero.c -3 +3 M ./genfft/gen_hc2hc.ml -5 +5 M ./genfft/gen_hc2r.ml -6 +6 M ./genfft/gen_notw.ml -6 +6 M ./genfft/gen_notw_c.ml -6 +6 M ./genfft/gen_r2hc.ml -6 +6 M ./genfft/gen_r2r.ml -6 +6 M ./genfft/gen_twiddle.ml -5 +5 M ./genfft/gen_twiddle_c.ml -5 +5 M ./genfft/gen_twidsq.ml -5 +5 M ./genfft/gen_twidsq_c.ml -5 +5 M ./kernel/buffered.c -2 +2 M ./kernel/cpy1d.c -2 +2 M ./kernel/cpy2d-pair.c -9 +7 M ./kernel/cpy2d.c -20 +19 M ./kernel/ct.c -1 +1 M ./kernel/iabs.c -3 +3 M ./kernel/ifftw.h -81 +99 M ./kernel/md5-1.c -5 +5 M ./kernel/minmax.c -3 +3 M ./kernel/ops.c -4 +4 M ./kernel/planner.c -2 +2 M ./kernel/primes.c -27 +29 M ./kernel/rader.c -3 +3 M ./kernel/solvtab.c -1 +2 M ./kernel/stride.c -3 +3 M ./kernel/tensor.c -7 +8 M ./kernel/tensor1.c -2 +2 M ./kernel/tensor2.c -3 +21 M ./kernel/tensor4.c -8 +8 M ./kernel/tensor7.c -10 +16 M ./kernel/tile2d.c -31 +8 M ./kernel/transpose.c -12 +12 M ./kernel/trig.c -13 +3 M ./kernel/twiddle.c -23 +23 M ./rdft/buffered.c -15 +15 M ./rdft/buffered2.c -30 +30 M ./rdft/codelet-rdft.h -29 +29 M ./rdft/codelets/hc2r.c -1 +1 M ./rdft/codelets/hfb.c -1 +1 M ./rdft/codelets/r2hc.c -1 +1 M ./rdft/codelets/r2r.c -1 +1 M ./rdft/dft-r2hc.c -12 +12 M ./rdft/dht-r2hc.c -5 +5 M ./rdft/dht-rader.c -12 +12 M ./rdft/direct.c -11 +11 M ./rdft/direct2.c -8 +8 M ./rdft/generic.c -14 +14 M ./rdft/hc2hc-common.c -3 +3 M ./rdft/hc2hc-direct.c -13 +13 M ./rdft/hc2hc-directbuf.c -24 +24 M ./rdft/hc2hc-generic.c -48 +31 M ./rdft/hc2hc.c -5 +5 M ./rdft/hc2hc.h -7 +7 M ./rdft/problem.c -3 +3 M ./rdft/problem2.c -2 +2 M ./rdft/rank0-rdft2.c -7 +7 M ./rdft/rank0.c -11 +11 M ./rdft/rdft-dht.c -9 +9 M ./rdft/rdft.h -2 +2 M ./rdft/rdft2-inplace-strides.c -3 +3 M ./rdft/rdft2-radix2.c -23 +23 M ./rdft/rdft2-strides.c -1 +1 M ./rdft/rdft2-tensor-max-index.c -4 +4 M ./rdft/vrank-geq1-rdft2.c -6 +6 M ./rdft/vrank-geq1.c -5 +5 M ./rdft/vrank3-transpose.c -53 +54 M ./reodft/redft00e-r2hc-pad.c -11 +11 M ./reodft/redft00e-r2hc.c -11 +11 M ./reodft/reodft00e-splitradix.c -16 +17 M ./reodft/reodft010e-r2hc.c -26 +26 M ./reodft/reodft11e-r2hc-odd.c -19 +19 M ./reodft/reodft11e-r2hc.c -16 +16 M ./reodft/reodft11e-radix2.c -25 +25 M ./reodft/rodft00e-r2hc-pad.c -11 +11 M ./reodft/rodft00e-r2hc.c -11 +11 M ./simd/sse2.c -2 +3 M ./simd/taint.c -2 +2 Sat Aug 12 23:34:43 EDT 2006 Unknown tagger tagged before-64bit-rewrite Wed Dec 7 22:39:01 EST 2005 stevenj * [project @ 2005-12-08 03:39:01 by stevenj] PGI x86-64 cycle counter, courtesy Cristiano Calonaci M ./kernel/cycle.h -1 +12 Mon Dec 5 21:25:57 EST 2005 athena * [project @ 2005-12-06 02:25:57 by athena] Must insert into hash table when wisdom_state == WISDOM_ONLY, otherwise wisdom does not work. M ./kernel/planner.c -2 +3 Sat Oct 8 18:08:44 EDT 2005 stevenj * [project @ 2005-10-08 22:07:37 by stevenj] comment M ./m4/acx_pthread.m4 -1 +2 Sun Oct 2 11:49:13 EDT 2005 athena * [project @ 2005-10-02 15:49:13 by athena] Paranoia: made planner robust against MD5 collisions. M ./api/apiplan.c -6 +37 M ./kernel/ifftw.h -1 +19 M ./kernel/planner.c -25 +55 Tue Sep 27 22:33:18 EDT 2005 athena * [project @ 2005-09-28 02:33:18 by athena] Note that --enable-3dnow is unsupported. M ./doc/FAQ/fftw-faq.bfnn -2 +2 Tue Sep 27 22:31:04 EDT 2005 athena * [project @ 2005-09-28 02:31:04 by athena] * Removed --enable-3dnow support. * SIMD support for split complex arrays. M ./NEWS +4 Tue Sep 27 22:28:41 EDT 2005 athena * [project @ 2005-09-28 02:28:40 by athena] Removed --enabled-3dnow, since it is becoming useless as the world moves to x86-64, and it is a pain to maintain. (We should probably remove the k7 stuff as well.) M ./api/version.c -5 +1 M ./configure.ac -9 M ./genfft/gen_notw.ml -6 +4 M ./genfft/gen_notw_c.ml -4 +2 M ./genfft/gen_twiddle.ml -5 +3 M ./genfft/gen_twiddle_c.ml -4 +2 M ./genfft/gen_twidsq_c.ml -4 +2 M ./kernel/align.c -4 +2 M ./kernel/ifftw.h -2 +2 R ./simd/3dnow.c M ./simd/Makefile.am -2 +2 R ./simd/simd-3dnow.h M ./simd/simd-altivec.h -3 M ./simd/simd-sse.h -3 M ./simd/simd-sse2.h -3 M ./simd/simd.h -4 Tue Sep 27 21:59:16 EDT 2005 athena * [project @ 2005-09-28 01:59:16 by athena] Missing BEGIN_SIMD(), END_SIMD() statements. M ./genfft/gen_notw.ml -4 +6 M ./genfft/gen_twiddle.ml -3 +5 Tue Sep 27 12:16:08 EDT 2005 athena * [project @ 2005-09-27 16:16:08 by athena] Tweaks M ./simd/simd-sse.h -12 +9 Tue Sep 27 10:04:32 EDT 2005 athena * [project @ 2005-09-27 14:04:32 by athena] Fixed wrong opcount for simd codelets. M ./genfft/to_alist.ml -9 +9 Tue Sep 27 09:25:50 EDT 2005 athena * [project @ 2005-09-27 13:25:50 by athena] Fixed wrong opcount for simd codelets. M ./dft/dftw-direct.c -2 +2 Mon Sep 26 22:58:19 EDT 2005 athena * [project @ 2005-09-27 02:58:19 by athena] fixed flop counts M ./genfft/c.ml -23 +17 M ./simd/simd-altivec.h -2 +2 M ./simd/simd-sse2.h -9 +9 Mon Sep 26 22:34:40 EDT 2005 athena * [project @ 2005-09-27 02:34:40 by athena] Silence warnings M ./simd/simd-sse2.h +6 Mon Sep 26 20:52:36 EDT 2005 athena * [project @ 2005-09-27 00:52:36 by athena] Implemented split-complex SIMD codelets A ./dft/simd/n2s.c A ./dft/simd/n2s.h A ./dft/simd/t1s.c A ./dft/simd/t1s.h M ./dft/simd/Makefile.am -2 +3 M ./dft/simd/codelets/Makefile.am -6 +18 M ./dft/simd/n2s.c +50 M ./dft/simd/n2s.h +28 M ./dft/simd/t1s.c +48 M ./dft/simd/t1s.h +33 M ./genfft/annotate.ml -3 +3 M ./genfft/c.ml -1 +3 M ./genfft/gen_hc2hc.ml -3 +3 M ./genfft/gen_notw.ml -5 +26 M ./genfft/gen_twiddle.ml -4 +4 M ./genfft/gen_twiddle_c.ml -3 +3 M ./genfft/gen_twidsq_c.ml -3 +3 M ./genfft/genutil.ml -3 +3 M ./genfft/simd.ml -14 +8 M ./genfft/twiddle.ml -3 +3 M ./genfft/twiddle.mli -2 +2 M ./simd/simd-altivec.h -7 +33 M ./simd/simd-sse.h +33 M ./simd/simd-sse2.h -2 +38 M ./simd/simd.h +2 Sun Sep 25 22:25:35 EDT 2005 athena * [project @ 2005-09-26 02:25:35 by athena] Generalized the ``store pairs'' trick (now called ``store multiple''). M ./dft/simd/codelets/Makefile.am -2 +2 M ./genfft/annotate.ml -37 +25 M ./genfft/annotate.mli -4 +2 M ./genfft/expr.ml -6 +6 M ./genfft/expr.mli -2 +2 M ./genfft/gen_notw_c.ml -11 +13 M ./genfft/simd.ml -6 +10 M ./genfft/simdmagic.ml -3 +3 M ./simd/simd-3dnow.h -2 +2 M ./simd/simd-altivec.h -2 +4 M ./simd/simd-sse.h -3 +3 M ./simd/simd-sse2.h -2 +2 Sun Sep 25 18:58:20 EDT 2005 athena * [project @ 2005-09-25 22:58:20 by athena] Silence some warnings. M ./simd/simd-altivec.h -4 +4 Sat Sep 24 12:37:16 EDT 2005 athena * [project @ 2005-09-24 16:37:16 by athena] Removed obsolete cruft M ./simd/simd-altivec.h -6 +6 Mon Sep 19 22:55:19 EDT 2005 athena * [project @ 2005-09-20 02:55:19 by athena] Re-enabled check for because OSX requires it. M ./configure.ac -1 +1 M ./simd/simd-altivec.h +2 Sun Sep 11 11:03:03 EDT 2005 athena * [project @ 2005-09-11 15:03:03 by athena] Check for sizeof(unsigned int) unconditionally, because the result is used by ifftw.h. M ./configure.ac -2 +2 Sun Sep 11 10:59:40 EDT 2005 athena * [project @ 2005-09-11 14:59:40 by athena] Higher size limit for t2 codelets. M ./dft/simd/t.c -1 +1 Sun Sep 11 10:50:37 EDT 2005 athena * [project @ 2005-09-11 14:50:37 by athena] Heuristic: do not use t2 simd codelets for N>1024. A ./dft/simd/t.c M ./dft/simd/Makefile.am -2 +2 M ./dft/simd/t.c +98 R ./dft/simd/t1b.c R ./dft/simd/t1f.c M ./dft/simd/t2b.h -2 +1 M ./dft/simd/t2f.h -2 +1 Mon Sep 5 22:22:50 EDT 2005 athena * [project @ 2005-09-06 02:22:50 by athena] Larger tolerance in timer calibration routine. M ./libbench2/timer.c -2 +2 Mon Sep 5 16:03:33 EDT 2005 athena * [project @ 2005-09-05 20:03:33 by athena] #include unconditionally. (There is no point in checking.) M ./configure.ac -2 +1 M ./simd/simd-altivec.h -2 Mon Sep 5 15:23:27 EDT 2005 athena * [project @ 2005-09-05 19:23:27 by athena] Removed SSE and SSE2 asm because it was bitrotting. Use the Intel API instead, which seems to be supported by gcc >= 3.3. Moved files that require -msse, -msse2 to new directory. A ./simd/nonportable/ A ./simd/nonportable/Makefile.am A ./simd/nonportable/sse.c A ./simd/nonportable/sse2.c M ./Makefile.am +1 M ./configure.ac +1 M ./simd/Makefile.am -4 +3 M ./simd/nonportable/Makefile.am +8 M ./simd/nonportable/sse.c +43 M ./simd/nonportable/sse2.c +43 M ./simd/simd-sse.h -101 +19 M ./simd/simd-sse2.h -75 +7 R ./simd/sse-aux.c M ./simd/sse.c -5 +7 R ./simd/sse2-aux.c M ./simd/sse2.c -5 +7 Mon Sep 5 12:56:28 EDT 2005 athena * [project @ 2005-09-05 16:56:28 by athena] Parse cputypes of the form 7447A,altivecsupported M ./m4/ax_gcc_archflag.m4 -1 +1 Mon Sep 5 12:52:30 EDT 2005 athena * [project @ 2005-09-05 16:52:30 by athena] Distinguish powerpc 7400 from the 7450, which has a different pipeline. M ./m4/ax_gcc_archflag.m4 +2 Mon Sep 5 12:46:00 EDT 2005 athena * [project @ 2005-09-05 16:46:00 by athena] Paranoia: define RIGHT_CPU unconditionally. M ./simd/simd-altivec.h -3 +3 Thu Aug 11 20:56:41 EDT 2005 athena * [project @ 2005-08-12 00:56:41 by athena] Removed obsolete name fftw-wisdom2c. M ./tools/fftw-wisdom-to-conf.in -1 +1 Thu Aug 11 20:55:59 EDT 2005 athena * [project @ 2005-08-12 00:55:59 by athena] Avoid creation of temporary files---use cpp magic instead. This fix solves a security bug and avoids nonportable tempfile creation hacks. M ./tools/fftw-wisdom-to-conf.in -13 +20 Fri Aug 5 10:03:02 EDT 2005 athena * [project @ 2005-08-05 14:03:02 by athena] Workaround for with gcc-3.3 altivec bug. M ./configure.ac -1 +3 M ./simd/altivec.c -3 +2 M ./simd/simd-altivec.h -2 +2 Wed Jun 15 21:36:46 EDT 2005 stevenj * [project @ 2005-06-16 01:36:46 by stevenj] solaris fix: check -pthreads first since gcc does not like -pthread but chokes due to stubbed libc (grr) M ./m4/acx_pthread.m4 -2 +2 Fri Jun 3 17:19:56 EDT 2005 stevenj * [project @ 2005-06-03 21:19:56 by stevenj] note that VC++ bug was fixed in 2005 M ./doc/FAQ/fftw-faq.bfnn -4 +4 Mon May 30 16:30:45 EDT 2005 stevenj * [project @ 2005-05-30 20:30:32 by stevenj] generalized ax_cc_vendor to ax_compiler_vendor A ./m4/ax_compiler_vendor.m4 M ./configure.ac -2 +2 M ./m4/ax_cc_maxopt.m4 -5 +5 R ./m4/ax_cc_vendor.m4 M ./m4/ax_compiler_vendor.m4 +30 Mon May 30 15:55:07 EDT 2005 stevenj * [project @ 2005-05-30 19:55:07 by stevenj] updated message M ./m4/ax_cc_maxopt.m4 -2 +2 Mon May 30 15:45:14 EDT 2005 stevenj * [project @ 2005-05-30 19:45:14 by stevenj] update for new AC archive format M ./m4/acx_pthread.m4 -1 +4 M ./m4/ax_cc_maxopt.m4 -4 +9 M ./m4/ax_cc_vendor.m4 +6 M ./m4/ax_check_compiler_flags.m4 -1 +4 M ./m4/ax_gcc_aligns_stack.m4 -1 +4 M ./m4/ax_gcc_archflag.m4 -2 +5 M ./m4/ax_gcc_version.m4 -1 +4 M ./m4/ax_gcc_x86_cpuid.m4 -1 +4 M ./m4/ax_openmp.m4 -1 +4 Mon May 23 23:12:22 EDT 2005 stevenj * [project @ 2005-05-24 03:12:22 by stevenj] M ./api/fftw3.h -2 +2 Mon May 23 18:17:38 EDT 2005 stevenj * [project @ 2005-05-23 22:17:38 by stevenj] M ./NEWS -1 +1 Mon May 23 18:13:08 EDT 2005 stevenj * [project @ 2005-05-23 22:13:08 by stevenj] more notes M ./NEWS -1 +9 Sun May 22 23:37:08 EDT 2005 stevenj * [project @ 2005-05-23 03:37:08 by stevenj] whoops M ./m4/ax_cc_maxopt.m4 -8 +8 Sun May 22 22:37:50 EDT 2005 stevenj * [project @ 2005-05-23 02:37:50 by stevenj] note icc 8.x annoyance M ./doc/FAQ/fftw-faq.bfnn -1 +3 Sun May 22 22:36:04 EDT 2005 stevenj * [project @ 2005-05-23 02:36:04 by stevenj] M ./doc/FAQ/fftw-faq.bfnn -1 +1 Sun May 22 22:35:34 EDT 2005 stevenj * [project @ 2005-05-23 02:35:34 by stevenj] note gcc 3.4.[0123] bug, which is fixed in gcc 3.4.4 M ./doc/FAQ/fftw-faq.bfnn +5 Sun May 22 22:21:26 EDT 2005 stevenj * [project @ 2005-05-23 02:21:26 by stevenj] added automatic detection of icc architecture flag M ./m4/ax_cc_maxopt.m4 -1 +34 Sun May 22 21:47:19 EDT 2005 stevenj * [project @ 2005-05-23 01:47:19 by stevenj] add -no-gcc to icc flags...even if it is Intel's fault, I'm sick of dealing with bug reports about this M ./configure.ac +2 Sun May 22 21:40:59 EDT 2005 stevenj * [project @ 2005-05-23 01:40:59 by stevenj] added @cindex portability M ./doc/fftw3.texi -2 +11 Sun May 22 21:34:10 EDT 2005 stevenj * [project @ 2005-05-23 01:34:10 by stevenj] note --without-gcc-arch M ./doc/fftw3.texi -2 +4 Sun May 22 20:54:54 EDT 2005 stevenj * [project @ 2005-05-23 00:54:54 by stevenj] bsd ppc detection; some odd 603 types M ./m4/ax_gcc_archflag.m4 -2 +3 Sun May 22 11:53:20 EDT 2005 stevenj * [project @ 2005-05-22 15:53:20 by stevenj] M ./m4/ax_gcc_archflag.m4 -2 +2 Sat May 21 20:34:52 EDT 2005 stevenj * [project @ 2005-05-22 00:34:52 by stevenj] ensure no spaces in cputype M ./m4/ax_gcc_archflag.m4 -2 +2 Sat May 21 20:31:41 EDT 2005 stevenj * [project @ 2005-05-22 00:31:41 by stevenj] nevermind M ./m4/ax_gcc_archflag.m4 -2 +2 Sat May 21 20:30:08 EDT 2005 stevenj * [project @ 2005-05-22 00:30:08 by stevenj] more bsd stuff M ./m4/ax_gcc_archflag.m4 -2 +2 Sat May 21 20:28:40 EDT 2005 stevenj * [project @ 2005-05-22 00:28:40 by stevenj] added BSD cpu detection for SPARC and better super/hypersparc detection M ./m4/ax_gcc_archflag.m4 -5 +5 Sat May 21 20:22:11 EDT 2005 stevenj * [project @ 2005-05-22 00:22:11 by stevenj] comment M ./m4/ax_gcc_archflag.m4 -1 +4 Fri May 20 19:40:09 EDT 2005 stevenj * [project @ 2005-05-20 23:40:09 by stevenj] "alternate" == "alternative" is US-centric M ./doc/fftw3.texi -3 +3 Fri May 20 19:36:26 EDT 2005 stevenj * [project @ 2005-05-20 23:36:26 by stevenj] typo M ./doc/fftw3.texi -2 +2 Fri May 20 01:28:34 EDT 2005 stevenj * [project @ 2005-05-20 05:28:34 by stevenj] clarification M ./doc/FAQ/fftw-faq.bfnn -2 +2 Tue May 17 18:56:46 EDT 2005 stevenj * [project @ 2005-05-17 22:56:46 by stevenj] print out estimate-planner time from can_do in verbose>2 mode M ./tests/bench.c -1 +7 Mon May 9 00:47:19 EDT 2005 stevenj * [project @ 2005-05-09 04:47:19 by stevenj] comment M ./m4/ax_cc_vendor.m4 +1 Thu May 5 23:47:55 EDT 2005 stevenj * [project @ 2005-05-06 03:47:55 by stevenj] fixes for building Windows DLLs with Cygwin; thanks in part to Stephane Fillod M ./Makefile.am -1 +1 M ./api/api.h +2 M ./api/fftw3.h -2 +12 M ./configure.ac -2 +4 M ./threads/Makefile.am -1 +1 Fri Apr 22 19:47:43 EDT 2005 stevenj * [project @ 2005-04-22 23:47:43 by stevenj] -ffast-math seems to produce code that is either about the same speed or slightly faster (gcc 3.3 and 4.0, x86) M ./m4/ax_cc_maxopt.m4 -1 +4 Fri Apr 22 19:18:23 EDT 2005 stevenj * [project @ 2005-04-22 23:18:23 by stevenj] power5 fallback to power4 sched for older gcc's M ./m4/ax_gcc_archflag.m4 -2 +2 Fri Apr 22 19:14:53 EDT 2005 stevenj * [project @ 2005-04-22 23:14:53 by stevenj] check for power5 M ./m4/ax_gcc_archflag.m4 -1 +2 Tue Apr 19 21:55:13 EDT 2005 athena * [project @ 2005-04-20 01:55:13 by athena] Removed clause #3 M ./api/fftw3.h -5 +1 Tue Apr 19 21:44:57 EDT 2005 stevenj * [project @ 2005-04-20 01:44:57 by stevenj] license clarification M ./api/fftw3.h -1 +4 Tue Apr 19 21:42:51 EDT 2005 athena * [project @ 2005-04-20 01:42:51 by athena] Changed license of fftw3.h to X11. M ./api/fftw3.h -12 +24 Mon Apr 11 13:15:12 EDT 2005 stevenj * [project @ 2005-04-11 17:15:12 by stevenj] delete fixed-input code M ./genfft/gen_conv.ml -3 +2 Sun Apr 10 16:33:24 EDT 2005 athena * [project @ 2005-04-10 20:33:24 by athena] joned L-U-planner branch M ./api/apiplan.c -2 +2 M ./api/fftw3.h -4 +5 M ./api/mapflags.c -19 +35 M ./dft/bluestein.c -6 +7 M ./dft/buffered.c -18 +11 M ./dft/ct.c -7 +5 M ./dft/dftw-direct.c -1 +4 M ./dft/dftw-generic.c -2 +4 M ./dft/generic.c -1 +1 M ./dft/indirect-transpose.c -1 +1 M ./dft/indirect.c -8 +8 M ./dft/rader.c -14 +16 M ./dft/vrank-geq1.c -2 +2 M ./kernel/ifftw.h -49 +71 M ./kernel/planner.c -97 +146 M ./rdft/buffered.c -15 +9 M ./rdft/buffered2.c -9 +6 M ./rdft/dft-r2hc.c -7 +4 M ./rdft/dht-r2hc.c -7 +7 M ./rdft/dht-rader.c -13 +15 M ./rdft/generic.c -1 +1 M ./rdft/hc2hc-generic.c -1 +1 M ./rdft/hc2hc.c -3 +2 M ./rdft/indirect.c -6 +4 M ./rdft/rank-geq2-rdft2.c -2 +2 M ./rdft/rdft-dht.c -5 +5 M ./rdft/rdft2-radix2.c -2 +2 M ./rdft/vrank-geq1-rdft2.c -2 +2 M ./rdft/vrank-geq1.c -2 +2 M ./rdft/vrank3-transpose.c -6 +6 M ./reodft/redft00e-r2hc-pad.c -2 +2 M ./reodft/redft00e-r2hc.c -2 +2 M ./reodft/reodft00e-splitradix.c -34 +17 M ./reodft/reodft010e-r2hc.c -2 +2 M ./reodft/reodft11e-r2hc-odd.c -2 +2 M ./reodft/reodft11e-r2hc.c -2 +2 M ./reodft/reodft11e-radix2.c -2 +2 M ./reodft/rodft00e-r2hc-pad.c -2 +2 M ./reodft/rodft00e-r2hc.c -2 +2 M ./tests/bench.c -1 +1 M ./tests/hook.c -1 +1 Thu Apr 7 23:15:02 EDT 2005 stevenj * [project @ 2005-04-08 03:15:02 by stevenj] ref M ./reodft/reodft00e-splitradix.c -16 +33 Thu Apr 7 00:11:13 EDT 2005 stevenj * [project @ 2005-04-07 04:10:26 by stevenj] whoops M ./genfft/gen_r2r.ml -3 +3 Wed Apr 6 22:06:21 EDT 2005 stevenj * [project @ 2005-04-07 02:06:21 by stevenj] added (optional) new split-radix algorithm, enabled with -newsplit; also new -standalone option to omit desc; also -unitary, -normalization, and -normsqr options to generate r2r codelets with various normalization (to match lit. in DCT-II, use: -unitary -normsqr 2) M ./genfft/complex.ml -1 +22 M ./genfft/complex.mli -1 +7 M ./genfft/fft.ml -4 +86 M ./genfft/gen_athtw.ml -3 +3 M ./genfft/gen_hc2r.ml -4 +4 M ./genfft/gen_notw.ml -5 +4 M ./genfft/gen_notw_c.ml -3 +3 M ./genfft/gen_r2hc.ml -4 +4 M ./genfft/gen_r2r.ml -19 +43 M ./genfft/magic.ml -1 +6 M ./genfft/number.ml -4 +15 M ./genfft/number.mli -1 +2 Fri Mar 25 08:59:43 EST 2005 athena * [project @ 2005-03-25 13:59:43 by athena] Moved timeout check outside the search loop, because X(seconds) is expensive. M ./kernel/planner.c -9 +7 Sun Mar 20 18:35:53 EST 2005 athena * [project @ 2005-03-20 23:35:53 by athena] Enable vector recursion for in-place problems, otherwise dftw-genericbuf works only in PATIENT mode. M ./dft/ct.c -2 +4 Sun Mar 20 17:53:58 EST 2005 athena * [project @ 2005-03-20 22:53:58 by athena] oops M ./dft/dftw-genericbuf.c -1 +1 Sun Mar 20 17:49:13 EST 2005 athena * [project @ 2005-03-20 22:49:13 by athena] make solver UGLY for small N M ./dft/dftw-genericbuf.c -3 +13 Sun Mar 20 17:16:37 EST 2005 athena * [project @ 2005-03-20 22:16:37 by athena] new dftw-genericbuf solver A ./dft/dftw-genericbuf.c Sun Mar 20 16:12:44 EST 2005 athena * [project @ 2005-03-20 21:12:44 by athena] new dftw-genericbuf solver M ./dft/Makefile.am -4 +4 M ./dft/conf.c -1 +2 M ./dft/dft.h -1 +2 Thu Mar 17 21:48:19 EST 2005 athena * [project @ 2005-03-18 02:48:19 by athena] Hmm... what was I thinking? M ./simd/sse2-aux.c -2 +2 Thu Mar 17 19:20:54 EST 2005 athena * [project @ 2005-03-18 00:20:54 by athena] Workaround for a MSVC bug. M ./simd/simd-sse2.h -1 +6 M ./simd/sse2-aux.c -1 +8 Thu Mar 17 08:18:39 EST 2005 athena * [project @ 2005-03-17 13:18:39 by athena] Workaround for a MSVC bug that was reported by Eddie Yee. M ./simd/simd-sse.h -1 +6 M ./simd/sse-aux.c -1 +8 Tue Mar 15 13:25:53 EST 2005 athena * [project @ 2005-03-15 18:25:53 by athena] try both contiguous input and contiguous output when in doubt M ./rdft/rank0.c -3 +26 Tue Mar 15 08:44:41 EST 2005 athena * [project @ 2005-03-15 13:44:41 by athena] Added genfft flag -precompute-twiddles which moves the computation of the twiddle factors before the main schedule. This flag produces smaller code everywhere, and slightly faster code on powerpc. I observe no speed difference on x86. M ./dft/codelets/standard/Makefile.am -3 +3 M ./genfft/genutil.ml -3 +5 M ./genfft/magic.ml -1 +5 M ./genfft/schedule.ml -1 +53 M ./genfft/schedule.mli -1 +2 M ./rdft/codelets/hc2r/Makefile.am -1 +1 M ./rdft/codelets/r2hc/Makefile.am -1 +1 Mon Mar 14 21:43:53 EST 2005 stevenj * [project @ 2005-03-15 02:43:53 by stevenj] sp M ./kernel/kalloc.c -2 +2 Mon Mar 14 21:43:05 EST 2005 stevenj * [project @ 2005-03-15 02:43:05 by stevenj] whoops, spelling error (thanks to Steve Eddins for bug report) M ./kernel/alloc.c -2 +2 Sat Mar 12 15:03:47 EST 2005 athena * [project @ 2005-03-12 20:03:45 by athena] Do not approximate pcost = vl * child->pcost unless child is guaranteed not to be a simple codelet. M ./dft/vrank-geq1.c -2 +4 M ./rdft/vrank-geq1-rdft2.c -2 +4 M ./rdft/vrank-geq1.c -2 +4 Wed Mar 9 20:00:02 EST 2005 athena * [project @ 2005-03-10 01:00:02 by athena] Relaxed applicability conditions. M ./dft/direct.c -6 +3 Wed Mar 9 00:05:47 EST 2005 athena * [project @ 2005-03-09 05:05:47 by athena] Minor optimization M ./dft/dftw-generic.c -8 +8 Tue Mar 8 22:14:02 EST 2005 athena * [project @ 2005-03-09 03:14:02 by athena] Interpret K to mean *1024. Similarly for M. M ./libbench2/problem.c -1 +12 Tue Mar 8 20:44:25 EST 2005 athena * [project @ 2005-03-09 01:44:25 by athena] Hmm... somehow some previous commit got lost. M ./kernel/primes.c -2 +8 Tue Mar 8 20:30:42 EST 2005 athena * [project @ 2005-03-09 01:30:42 by athena] Paranoia M ./dft/ct.c -2 +2 Mon Mar 7 14:30:01 EST 2005 stevenj * [project @ 2005-03-07 19:30:01 by stevenj] whoops M ./configure.ac +1 Mon Mar 7 14:29:43 EST 2005 stevenj * [project @ 2005-03-07 19:29:43 by stevenj] move fftw-specific HP/UX tweak into configure.ac M ./configure.ac +3 M ./m4/ax_cc_maxopt.m4 -2 +2 Mon Mar 7 14:19:24 EST 2005 stevenj * [project @ 2005-03-07 19:19:24 by stevenj] ax_cc_family -> ax_cc_vendor (vendor names are easier to remember), add checks for many new compilers, use in ax_cc_maxopt A ./m4/ax_cc_vendor.m4 M ./configure.ac -3 +3 R ./m4/ax_cc_family.m4 M ./m4/ax_cc_maxopt.m4 -65 +42 M ./m4/ax_cc_vendor.m4 +23 Sun Mar 6 21:36:05 EST 2005 athena * [project @ 2005-03-07 02:36:05 by athena] Count FMA as one flop in estimator when HAVE_FMA M ./kernel/planner.c -1 +7 Sun Mar 6 19:16:06 EST 2005 athena * [project @ 2005-03-07 00:16:06 by athena] Do not try radix-2 generic. M ./dft/dftw-generic.c -1 +1 Sun Mar 6 13:04:23 EST 2005 athena * [project @ 2005-03-06 18:04:23 by athena] Use -O3 for xlc now that we use -O for CODELET_OPTIM M ./m4/ax_cc_maxopt.m4 -4 +4 Sun Mar 6 13:02:41 EST 2005 athena * [project @ 2005-03-06 18:02:41 by athena] New AX_CC_FAMILY macro, that detects the compiler based on symbols that it defines (as opposed to the name of the compiler). We need to start use this strategy everywhere else. A ./m4/ax_cc_family.m4 M ./configure.ac -25 +29 M ./m4/ax_cc_family.m4 +18 Sun Mar 6 11:33:15 EST 2005 athena * [project @ 2005-03-06 16:33:15 by athena] Runtime checks to guarantee small strides. M ./dft/direct.c -2 +18 Sat Mar 5 20:09:25 EST 2005 athena * [project @ 2005-03-06 01:09:25 by athena] Reduced the search space for rank-0 transforms M ./dft/vrank-geq1.c -6 +6 M ./kernel/tensor7.c -8 +20 M ./rdft/rank0.c -50 +48 M ./rdft/vrank-geq1.c -6 +5 Fri Mar 4 17:50:29 EST 2005 stevenj * [project @ 2005-03-04 22:50:29 by stevenj] little assert M ./kernel/primes.c -8 +3 Tue Mar 1 09:19:16 EST 2005 athena * [project @ 2005-03-01 14:19:16 by athena] Implemented directbuf, enabled for now. M ./dft/dft.h -1 +2 M ./dft/dftw-direct.c -10 +7 M ./dft/direct.c -11 +135 M ./dft/kdft.c -3 +3 Mon Feb 28 22:21:14 EST 2005 athena * [project @ 2005-03-01 03:21:14 by athena] Unified dftw-direct, dftw-directbuf in an attempt to tame code growth M ./dft/Makefile.am -4 +4 M ./dft/dftw-direct.c -15 +130 R ./dft/dftw-directbuf.c M ./dft/kdft-dif.c -2 +1 M ./dft/kdft-dit.c -2 +1 Sun Feb 27 13:51:24 EST 2005 stevenj * [project @ 2005-02-27 18:51:24 by stevenj] fixed copyright M ./doc/fftw3.texi -2 +2 Sat Feb 26 22:21:03 EST 2005 athena * [project @ 2005-02-27 03:21:03 by athena] silence warnings M ./rdft/rank0.c -4 +5 Sat Feb 26 22:19:16 EST 2005 athena * [project @ 2005-02-27 03:19:16 by athena] oops M ./rdft/rank0.c -2 +2 Sat Feb 26 21:28:39 EST 2005 athena * [project @ 2005-02-27 02:28:39 by athena] Tweaking while thinking about a higher-rank transposer (bitreverser) M ./rdft/rank0.c -44 +63 Sat Feb 26 20:06:49 EST 2005 athena * [project @ 2005-02-27 01:06:49 by athena] Transposed the buffer, and skewed it. This allows for contiguous copy operations, and the codelet should not incur associativity conflicts if the buffer is large. M ./dft/dftw-directbuf.c -20 +26 Sat Feb 26 18:14:11 EST 2005 stevenj * [project @ 2005-02-26 23:14:11 by stevenj] make tensor_max_index more reasonable (take maximum of input and output max indices, computed separately) M ./kernel/tensor4.c -4 +5 Sat Feb 26 10:04:30 EST 2005 athena * [project @ 2005-02-26 15:04:30 by athena] Use cpy2d instead of cpy2d_tiled, because vl may be too large. M ./rdft/vrank3-transpose.c -5 +5 Sat Feb 26 00:31:52 EST 2005 athena * [project @ 2005-02-26 05:31:52 by athena] Fixed old bug that was introduced with yesterday's changes. M ./genfft/annotate.ml -4 +6 Fri Feb 25 21:54:23 EST 2005 athena * [project @ 2005-02-26 02:54:23 by athena] ``Interesting'' switch statement. M ./kernel/cpy1d.c -3 +23 Fri Feb 25 12:29:54 EST 2005 athena * [project @ 2005-02-25 17:29:54 by athena] Disabled -reorder-loads -reorder-stores, since they seem to do nothing. M ./support/Makefile.codelets -1 +1 Fri Feb 25 12:19:10 EST 2005 stevenj * [project @ 2005-02-25 17:19:03 by stevenj] Because of the recent changes to kernel/pickdim.c, splitrnk=0 is no longer equivalent to splitrnk=1 for rnk < 4, where the latter is the FFTW2 behavior. For small rnk, however, I observe the planner to pretty consistently choose the FFTW2 behavior (splitrnk=1), despite its not being asymptotically optimal in the cache oblivious sense. So, make splitrnk=1 instead of splitrnk=0 the default in FFTW_MEASURE and FFTW_ESTIMATE modes (rnk > 3 is pretty rare in practice anyway). M ./dft/rank-geq2.c -5 +2 M ./rdft/rank-geq2.c -5 +2 Fri Feb 25 00:33:27 EST 2005 stevenj * [project @ 2005-02-25 05:33:27 by stevenj] tweak M ./dft/indirect-transpose.c -3 +4 Fri Feb 25 00:29:09 EST 2005 stevenj * [project @ 2005-02-25 05:29:09 by stevenj] slight relaxation M ./dft/indirect-transpose.c -2 +4 Fri Feb 25 00:21:00 EST 2005 stevenj * [project @ 2005-02-25 05:21:00 by stevenj] cruft M ./dft/indirect-transpose.c -2 Fri Feb 25 00:03:14 EST 2005 stevenj * [project @ 2005-02-25 05:03:13 by stevenj] added experimental indirect-transpose solver: when transforming the columns of the matrix, allow us to do a transpose to make the DFTs contiguous A ./dft/indirect-transpose.c M ./dft/Makefile.am -3 +3 M ./dft/conf.c -1 +2 M ./dft/dft.h -1 +2 M ./dft/indirect-transpose.c +280 M ./dft/indirect.c -3 +10 M ./kernel/ifftw.h -1 +3 M ./kernel/tensor4.c -1 +32 Thu Feb 24 23:04:58 EST 2005 stevenj * [project @ 2005-02-25 04:04:58 by stevenj] check for abort() M ./configure.ac -1 +1 Thu Feb 24 23:04:43 EST 2005 stevenj * [project @ 2005-02-25 04:04:43 by stevenj] call abort() on failed assertion M ./kernel/assert.c -1 +5 Thu Feb 24 21:17:23 EST 2005 athena * [project @ 2005-02-25 02:17:23 by athena] Forgot to change X(isqrt) -> isqrt_maybe M ./kernel/primes.c -2 +2 Thu Feb 24 20:18:59 EST 2005 stevenj * [project @ 2005-02-25 01:17:59 by stevenj] require finite_rnk M ./dft/rank-geq2.c -1 +2 M ./rdft/rank-geq2-rdft2.c -1 +2 M ./rdft/rank-geq2.c -1 +2 Thu Feb 24 20:07:38 EST 2005 stevenj * [project @ 2005-02-25 01:07:38 by stevenj] #ifdef HAVE_STRING_H must come after rdft.h so that we get config.h M ./rdft/vrank3-transpose.c -3 +3 Thu Feb 24 18:59:40 EST 2005 athena * [project @ 2005-02-24 23:59:38 by athena] Implemented reordering of loads and stores so that the real and imaginary part are loaded/stored together. This should improve out-of-cache performance in the presence of associativity conflicts, and maybe worsen in-cache performance because of worse scheduling. Enabled for now, for experimental purposes. M ./genfft/annotate.ml -30 +76 M ./genfft/magic.ml -4 +10 M ./support/Makefile.codelets -1 +1 M ./support/twovers.sh -1 +1 Thu Feb 24 18:10:49 EST 2005 stevenj * [project @ 2005-02-24 23:10:49 by stevenj] fix comment M ./m4/ax_gcc_aligns_stack.m4 -2 +2 Thu Feb 24 18:10:23 EST 2005 stevenj * [project @ 2005-02-24 23:10:23 by stevenj] better message M ./m4/ax_gcc_aligns_stack.m4 -2 +2 Thu Feb 24 18:08:36 EST 2005 stevenj * [project @ 2005-02-24 23:08:36 by stevenj] use gcc version > 3.0 as fallback in check for alignment bug M ./m4/ax_gcc_aligns_stack.m4 -2 +2 Thu Feb 24 18:02:31 EST 2005 stevenj * [project @ 2005-02-24 23:02:31 by stevenj] don't use -malign-double unconditionally (it is only available on x86) M ./m4/ax_gcc_aligns_stack.m4 -2 +3 Thu Feb 24 12:03:30 EST 2005 athena * [project @ 2005-02-24 17:03:30 by athena] Subtler selection of tilesz. M ./kernel/transpose.c -3 +8 Thu Feb 24 11:52:25 EST 2005 athena * [project @ 2005-02-24 16:52:25 by athena] Call cpy2d_tiledbuf, not cpy2d_tiled. M ./rdft/rank0.c -5 +5 Thu Feb 24 11:29:28 EST 2005 athena * [project @ 2005-02-24 16:29:28 by athena] buffer sizes were wrong :-( M ./kernel/cpy2d.c -1 +1 M ./kernel/transpose.c -2 +2 Thu Feb 24 11:19:01 EST 2005 athena * [project @ 2005-02-24 16:19:01 by athena] Single function for computing tile size. Eliminate spurious assertions. M ./kernel/cpy2d.c -3 +8 M ./kernel/ifftw.h -1 +3 M ./kernel/tile2d.c +5 M ./kernel/transpose.c -4 +6 M ./rdft/rank0.c -3 +1 Thu Feb 24 10:00:02 EST 2005 athena * [project @ 2005-02-24 15:00:02 by athena] Do tiling recursively. M ./kernel/tile2d.c +28 Thu Feb 24 09:40:30 EST 2005 athena * [project @ 2005-02-24 14:40:30 by athena] Reworked tiled transposes; provide tiling with and without buffering. I can't believe that one has to waste his life with this @#$%. A ./kernel/tile2d.c M ./kernel/Makefile.am -1 +1 M ./kernel/cpy2d.c -36 +60 M ./kernel/ifftw.h -2 +10 M ./kernel/tile2d.c +41 M ./kernel/transpose.c -64 +122 M ./rdft/rank0.c -6 +41 M ./rdft/vrank3-transpose.c -2 +2 Wed Feb 23 22:21:19 EST 2005 athena * [project @ 2005-02-24 03:21:19 by athena] Clarified logic. I am not sure why the code was so confusing to begin with. The computation of *dp in the which_dim == 0 case was also wrong, returning e.g. *dp == -1 if sz->rnk == 1. M ./kernel/pickdim.c -6 +6 Wed Feb 23 22:00:15 EST 2005 athena * [project @ 2005-02-24 03:00:15 by athena] Enable aggressive inlining in codelets only, to avoid code bloat. M ./configure.ac -9 +10 Wed Feb 23 21:51:50 EST 2005 athena * [project @ 2005-02-24 02:51:50 by athena] Removed cache-oblivious copy/transpose algorithms in favor of explicitly blocked algorithms. The cache-oblivious algorithms fail if there are associativity conflicts, in which case buffering is necessary, as per Carter and Gatlin. Once you set the buffer size, there is no point whatsoever to do the algorithm recursively, and you may as well use blocking. M ./kernel/Makefile.am -2 +1 M ./kernel/cpy2d.c -62 +40 M ./kernel/ifftw.h -11 +11 M ./kernel/primes.c -4 +10 R ./kernel/transpose-rec.c M ./kernel/transpose.c +54 M ./rdft/rank0.c -49 +28 M ./rdft/vrank3-transpose.c -6 +6 Wed Feb 23 18:46:12 EST 2005 stevenj * [project @ 2005-02-23 23:46:12 by stevenj] --disable-fortran now differs from --enable-fortran that fails M ./configure.ac -2 +4 Wed Feb 23 18:42:21 EST 2005 stevenj * [project @ 2005-02-23 23:42:21 by stevenj] comment tweak M ./api/f77api.c -1 +1 Wed Feb 23 18:41:14 EST 2005 stevenj * [project @ 2005-02-23 23:41:14 by stevenj] If a Fortran compiler was not detected, just make our best guess at what wrappers to use...I'm sick of dealing with user complaints from cases where wrapper detection fails for whatever reason. M ./api/f77api.c +21 Wed Feb 23 18:10:40 EST 2005 stevenj * [project @ 2005-02-23 23:10:40 by stevenj] fflush(stdout) after print_plan, in case F77 doesn't M ./api/f77funcs.h +1 Tue Feb 22 22:54:42 EST 2005 athena * [project @ 2005-02-23 03:54:42 by athena] --enable-sse is necessary after all, to generate all dependencies correctly. M ./mkdist.sh -1 +1 Tue Feb 22 22:32:06 EST 2005 athena * [project @ 2005-02-23 03:32:06 by athena] Put cpy2d_pair into its own file, so that I can experiment with buffering of nontwiddle codelets. A ./kernel/cpy2d-pair.c M ./dft/dftw-directbuf.c -28 +7 M ./kernel/Makefile.am -5 +6 M ./kernel/cpy2d-pair.c +61 M ./kernel/ifftw.h -1 +11 Tue Feb 22 20:07:11 EST 2005 athena * [project @ 2005-02-23 01:07:11 by athena] Copy rfftwnd.png from ${srcdir}, not $PWD M ./doc/Makefile.am -1 +1 Tue Feb 22 17:08:48 EST 2005 athena * [project @ 2005-02-22 22:08:48 by athena] Do not bother memcpy-ing complex numbers. M ./rdft/rank0.c -2 +6 Tue Feb 22 16:20:46 EST 2005 athena * [project @ 2005-02-22 21:20:46 by athena] Tighther layout of buffers. I am not sure it matters, but just in case... M ./kernel/cpy2d.c -2 +2 M ./kernel/transpose-rec.c -4 +4 Tue Feb 22 10:13:02 EST 2005 athena * [project @ 2005-02-22 15:13:02 by athena] Usec cpy1d for rank-0 copies M ./rdft/rank0.c -3 +2 Tue Feb 22 10:06:13 EST 2005 athena * [project @ 2005-02-22 15:06:13 by athena] Implemented in-place transposes with buffering. Moved copy/transposition routines into own files, so that we can reuse them from multiple places. TODO: merge vrank3-transpose.c with rank0.c, or rename vrank3-transpose.c to rank0-fancy.c or something like that; decide whether square in-place transposes should be in rank0.c or vrank3-transpose.c; apply FIXME's in vrank3-transpose.c. A ./kernel/cpy1d.c A ./kernel/cpy2d.c A ./kernel/transpose-rec.c A ./kernel/transpose.c M ./kernel/Makefile.am -5 +5 M ./kernel/cpy1d.c +50 M ./kernel/cpy2d.c +154 M ./kernel/ifftw.h -1 +28 M ./kernel/transpose-rec.c +143 M ./kernel/transpose.c +72 M ./rdft/rank0.c -160 +85 M ./rdft/vrank3-transpose.c -231 +24 Mon Feb 21 23:29:52 EST 2005 athena * [project @ 2005-02-22 04:29:52 by athena] Indentation should be printed after newline, not at the beginning of print() M ./kernel/print.c -6 +11 Mon Feb 21 10:07:24 EST 2005 athena * [project @ 2005-02-21 15:07:24 by athena] generalized in anticipation of more complicated solvers. M ./rdft/rank0.c -24 +45 Sun Feb 20 22:18:59 EST 2005 athena * [project @ 2005-02-21 03:18:59 by athena] Implemented buffered recursive transpose M ./rdft/rank0.c -7 +85 Sun Feb 20 18:27:29 EST 2005 athena * [project @ 2005-02-20 23:27:29 by athena] Fixed comment M ./rdft/rank0.c -2 +2 Sun Feb 20 18:22:15 EST 2005 athena * [project @ 2005-02-20 23:22:15 by athena] grand unification of rank0 solvers M ./rdft/Makefile.am -3 +2 M ./rdft/conf.c -2 +1 R ./rdft/rank0-vrank2.c M ./rdft/rank0.c -93 +173 M ./rdft/rdft.h -1 Sun Feb 20 15:35:24 EST 2005 athena * [project @ 2005-02-20 20:35:24 by athena] manual tail-recursion optimization M ./rdft/vrank3-transpose.c -7 +11 Sat Feb 19 17:57:44 EST 2005 athena * [project @ 2005-02-19 22:57:44 by athena] implemented check for transpositions M ./libbench2/verify-lib.c -11 +11 M ./libbench2/verify-r2r.c -1 +1 M ./tests/check.pl -1 +19 Sat Feb 19 17:28:43 EST 2005 athena * [project @ 2005-02-19 22:28:43 by athena] Previous fix was wrong for rdft2 problems. M ./libbench2/verify-lib.c -3 +3 Sat Feb 19 17:23:36 EST 2005 athena * [project @ 2005-02-19 22:23:36 by athena] vecsz->rnk must be finite for this solver to apply. M ./rdft/dft-r2hc.c -2 +2 Sat Feb 19 17:15:19 EST 2005 athena * [project @ 2005-02-19 22:15:19 by athena] unified the various simple'' transposers M ./rdft/vrank3-transpose.c -101 +51 Sat Feb 19 16:55:29 EST 2005 athena * [project @ 2005-02-19 21:55:29 by athena] Fixed stupid bug in rec_transpose_swap. Fixed stupid verifier that did not catch the bug. M ./libbench2/verify-lib.c -17 +26 M ./libbench2/verify-r2r.c -2 +3 M ./rdft/vrank3-transpose.c -5 +5 Sat Feb 19 15:24:03 EST 2005 athena * [project @ 2005-02-19 20:24:03 by athena] Minor cleanup of transposition routines. M ./rdft/vrank3-transpose.c -188 +83 Sat Feb 19 09:31:14 EST 2005 athena * [project @ 2005-02-19 14:31:14 by athena] Make the batch size B=Theta(r) instead of B=Theta(1) in buffered twiddle solvers. Theory: for cache line size L, we want B = Omega(L) to utilize the cache line fully. We also want B*r =O(Z), where Z is the size of the cache. It is safe to assume that Z = Theta(L^2): cache designers will tend to make L as large as they can get away with, because they don't have to program the machines that they build, and Z < Theta(L^2) will screw up the little matrix transposition benchmarks that they use to design the cache. Hence, B=Theta(r) is the right number. M ./dft/dftw-directbuf.c -12 +17 M ./rdft/hc2hc-directbuf.c -16 +20 Fri Feb 18 23:47:22 EST 2005 stevenj * [project @ 2005-02-19 04:47:22 by stevenj] for --enable-portable-binary, only try -mcpu=$arch and -m$arch on x86, since these generate non-portable code on every other target (and some other targets, like Alpha, don't support -mtune=$arch). M ./m4/ax_gcc_archflag.m4 -2 +5 Thu Feb 17 21:15:42 EST 2005 athena * [project @ 2005-02-18 02:15:42 by athena] gcc/aix defines _POWER, not __powerpc__ like the rest of the world does. M ./kernel/ifftw.h -2 +2 Wed Feb 16 22:30:27 EST 2005 athena * [project @ 2005-02-17 03:30:27 by athena] enable fma for ia64, since it seems to help with the hpux compiler. M ./configure.ac +1 Wed Feb 16 21:47:48 EST 2005 athena * [project @ 2005-02-17 02:47:48 by athena] M ./TODO -1 +2 Wed Feb 16 15:27:18 EST 2005 athena * [project @ 2005-02-16 20:27:18 by athena] Fixes for darwin M ./simd/simd-altivec.h -2 +2 Wed Feb 16 14:27:42 EST 2005 athena * [project @ 2005-02-16 19:27:42 by athena] Made the correctness of the code more obvious. M ./api/apiplan.c -14 +16 Wed Feb 16 12:30:29 EST 2005 stevenj * [project @ 2005-02-16 17:30:29 by stevenj] s/with-portable-binary/enable-portable-binary/ to be GNUlly correct; I'm sticking with --with-gcc-arch=arch, however, as --enable-gcc-arch=arch has the wrong connotations for me M ./NEWS -2 +2 M ./m4/ax_cc_maxopt.m4 -3 +3 Wed Feb 16 11:44:48 EST 2005 stevenj * [project @ 2005-02-16 16:44:48 by stevenj] whoops M ./api/apiplan.c -2 +2 Wed Feb 16 11:23:38 EST 2005 stevenj * [project @ 2005-02-16 16:23:38 by stevenj] bless wisdom with patience used to create it M ./api/apiplan.c -1 +1 Wed Feb 16 11:18:56 EST 2005 stevenj * [project @ 2005-02-16 16:18:56 by stevenj] whoops M ./api/apiplan.c -7 +10 Wed Feb 16 10:50:28 EST 2005 stevenj * [project @ 2005-02-16 15:50:28 by stevenj] whoops M ./api/apiplan.c -1 +2 Tue Feb 15 23:53:53 EST 2005 stevenj * [project @ 2005-02-16 04:53:53 by stevenj] added 'timed' planner option M ./NEWS +3 M ./TODO -6 M ./api/apiplan.c -16 +56 M ./api/fftw3.h -1 +3 M ./doc/fftw3.texi -1 +27 M ./kernel/ifftw.h -1 +6 M ./kernel/planner.c -2 +14 M ./kernel/timer.c -1 +13 M ./tests/bench.c +5 Tue Feb 15 23:08:29 EST 2005 athena * [project @ 2005-02-16 04:08:27 by athena] Do not use SIMD_CFLAGS. The theory is that if taint.c is unsafe with SIMD_CFLAGS, then all files in this directory are as well. Conversely, if these files require SIMD_CFLAGS because they include "simd.h", then taint.c requires SIMD_CFLAGS as well, and thus we need some other hack. M ./dft/simd/Makefile.am -1 M ./simd/Makefile.am -8 +3 Tue Feb 15 22:49:05 EST 2005 athena * [project @ 2005-02-16 03:49:05 by athena] Do not override CFLAGS in Makefile.am. M ./dft/codelets/standard/Makefile.am -1 M ./dft/simd/Makefile.am -1 +1 M ./dft/simd/codelets/Makefile.am -1 +1 M ./rdft/codelets/hc2r/Makefile.am -1 M ./rdft/codelets/r2hc/Makefile.am -1 M ./rdft/codelets/r2r/Makefile.am -1 M ./support/Makefile.codelets -1 +6 Tue Feb 15 10:30:12 EST 2005 athena * [project @ 2005-02-15 15:30:12 by athena] Allow users to build long double version even if sizeof(long double) == sizeof(double) M ./configure.ac -3 Mon Feb 14 19:55:38 EST 2005 athena * [project @ 2005-02-15 00:55:38 by athena] Updated for 3.1 M ./commercialize.sh -12 +20 Mon Feb 14 19:07:14 EST 2005 athena * [project @ 2005-02-15 00:07:14 by athena] Oops, version.h is no longer used M ./api/version.c -2 +1 Mon Feb 14 18:51:05 EST 2005 athena * [project @ 2005-02-14 23:51:05 by athena] unified fma and non-fma versions A ./support/twovers.sh M ./api/Makefile.am -5 +2 M ./api/version.c -2 +6 M ./configure.ac -5 +10 M ./dft/codelets/standard/Makefile.am -7 +7 M ./dft/simd/codelets/Makefile.am -10 +10 M ./m4/ocaml.m4 -9 +3 M ./mkdist.sh -5 M ./rdft/codelets/hc2r/Makefile.am -4 +4 M ./rdft/codelets/r2hc/Makefile.am -4 +4 M ./rdft/codelets/r2r/Makefile.am -9 +9 M ./support/Makefile.am -2 +2 M ./support/Makefile.codelets -7 +2 M ./support/twovers.sh +17 Mon Feb 14 14:12:09 EST 2005 athena * [project @ 2005-02-14 19:12:09 by athena] forgot to remove inplace/Makefile from configure.ac M ./configure.ac -1 Mon Feb 14 12:08:52 EST 2005 athena * [project @ 2005-02-14 17:08:49 by athena] Merged dft/codelets/inplace with the main dft/codelets/standard directory. This step makes dft codelets consistent with the rest of the naming conventions, and will simplify the eventual merge of fma and non-fma codelets. M ./Makefile.am -1 M ./dft/codelet-dft.h -2 +1 M ./dft/codelets/Makefile.am -1 +1 R ./dft/codelets/inplace/Makefile.am R ./dft/codelets/inplace/ M ./dft/codelets/standard/Makefile.am -9 +31 M ./dft/conf.c -2 +1 Mon Feb 14 11:16:15 EST 2005 athena * [project @ 2005-02-14 16:16:15 by athena] inline altivec constants, since gcc seems to generate better code this way. M ./simd/altivec.c -21 +1 M ./simd/simd-altivec.h -16 +18 Sun Feb 13 18:17:32 EST 2005 athena * [project @ 2005-02-13 23:17:32 by athena] group altivec constants into a single array, for faster access M ./simd/altivec.c -11 +17 M ./simd/simd-altivec.h -20 +11 Sun Feb 13 18:15:37 EST 2005 athena * [project @ 2005-02-13 23:15:37 by athena] code cleanup M ./genfft/c.ml -23 +9 M ./genfft/c.mli -2 +1 M ./genfft/simd.ml -6 +6 Sun Feb 13 10:29:32 EST 2005 athena * [project @ 2005-02-13 15:29:32 by athena] removed some unused stuff M ./genfft/c.ml -3 +1 M ./genfft/c.mli -2 +1 Sat Feb 12 22:04:40 EST 2005 athena * [project @ 2005-02-13 03:04:40 by athena] New twiddle scheme for altivec, 3dnow M ./simd/simd-3dnow.h -4 +34 M ./simd/simd-altivec.h -4 +27 Sat Feb 12 20:17:35 EST 2005 athena * [project @ 2005-02-13 01:17:35 by athena] Implemented new twiddle scheme for sse2 M ./simd/simd-sse2.h -4 +26 Sat Feb 12 19:57:46 EST 2005 athena * [project @ 2005-02-13 00:57:40 by athena] Implemented experimental t2* codelets, which store twiddle factors in a more convenient format, at the expense of twice the storage. Currently only SSE works; I have to port SSE2, altivec, etc. to the new scheme. After this, we will decide whether these codelets are worth the price. A ./dft/simd/t2b.h A ./dft/simd/t2f.h M ./dft/simd/Makefile.am -1 +1 M ./dft/simd/codelets/Makefile.am -2 +14 M ./dft/simd/q1b.h +5 M ./dft/simd/q1f.h +5 M ./dft/simd/t1b.h +5 M ./dft/simd/t1f.h +5 M ./dft/simd/t2b.h +36 M ./dft/simd/t2f.h +36 M ./simd/simd-sse.h -36 +26 Fri Feb 11 08:07:12 EST 2005 athena * [project @ 2005-02-11 13:07:12 by athena] Forgot to define SIMD_STRIDE_OKPAIR M ./simd/simd-altivec.h +1 Thu Feb 10 22:20:00 EST 2005 athena * [project @ 2005-02-11 03:20:00 by athena] fixed sse2, 3dnow, and altivec, as promised M ./simd/simd-3dnow.h -5 +3 M ./simd/simd-altivec.h -1 +4 M ./simd/simd-sse.h +1 M ./simd/simd-sse2.h -5 +3 Thu Feb 10 21:47:40 EST 2005 athena * [project @ 2005-02-11 02:47:33 by athena] Generate n2?v_* codelets in such a way that we may or may not pair stores, depending on which mode happens to work best on a particular SIMD implementation. sse2, 3dnow, and altivec are currently broken---will fix soon. M ./dft/simd/n2b.c -1 +1 M ./dft/simd/n2f.c -1 +1 M ./genfft/annotate.ml -7 +9 M ./genfft/expr.ml -4 +6 M ./genfft/expr.mli -2 +2 M ./genfft/simd.ml -3 +7 M ./simd/simd-sse.h -1 +6 Thu Feb 10 08:53:22 EST 2005 athena * [project @ 2005-02-10 13:53:22 by athena] instantiate altivec constants only once M ./simd/altivec.c -1 +6 M ./simd/simd-altivec.h -4 +10 Thu Feb 10 06:37:56 EST 2005 athena * [project @ 2005-02-10 11:37:56 by athena] Fixed alignment checks for new SIMD scheme M ./dft/simd/n2b.c -3 +3 M ./dft/simd/n2f.c -3 +3 Wed Feb 9 21:35:01 EST 2005 athena * [project @ 2005-02-10 02:35:01 by athena] Change n2?v_* codelets to store pairs of vectors, with implicit 2x2 transposition. Works for 2-way SIMD as well. Tested with sse and sse2. I haven't tried altivec yet, but I observed a huge speedup when I transformed one codelet by hand. M ./dft/simd/codelets/Makefile.am -8 +6 M ./genfft/annotate.ml -3 +48 M ./genfft/annotate.mli -2 +5 M ./genfft/expr.ml -2 +5 M ./genfft/expr.mli -2 +2 M ./genfft/gen_notw_c.ml -3 +17 M ./genfft/genutil.ml -3 +2 M ./genfft/simd.ml -5 +9 M ./genfft/simdmagic.ml -1 +3 M ./simd/simd-3dnow.h +6 M ./simd/simd-altivec.h +10 M ./simd/simd-sse.h +6 M ./simd/simd-sse2.h +6 Tue Feb 8 21:28:38 EST 2005 athena * [project @ 2005-02-09 02:28:38 by athena] Resurrected old DIF codelets for experimental purposes. They are disabled for now, but I am keeping the setup around for future reference. M ./dft/codelets/standard/Makefile.am -1 +14 Tue Feb 8 20:10:19 EST 2005 stevenj * [project @ 2005-02-09 01:09:12 by stevenj] M ./doc/fftw3.texi -7 +7 Tue Feb 8 19:37:09 EST 2005 stevenj * [project @ 2005-02-09 00:37:09 by stevenj] clarifications, document --with-portable-binary and --with-gcc-arch M ./doc/fftw3.texi -17 +30 Tue Feb 8 19:23:41 EST 2005 stevenj * [project @ 2005-02-09 00:20:56 by stevenj] M ./NEWS -7 +7 Tue Feb 8 01:36:22 EST 2005 stevenj * [project @ 2005-02-08 06:36:22 by stevenj] more change comments M ./NEWS +61 Tue Feb 8 00:41:38 EST 2005 stevenj * [project @ 2005-02-08 05:41:38 by stevenj] fma is definitely beneficial on Itanium with the HP/UX compiler M ./doc/FAQ/fftw-faq.bfnn -1 +1 Mon Feb 7 22:58:47 EST 2005 athena * [project @ 2005-02-08 03:58:47 by athena] Silence warnings. M ./libbench2/bench-main.c -3 +4 Mon Feb 7 22:55:49 EST 2005 stevenj * [project @ 2005-02-08 03:55:49 by stevenj] when we compile our own getopt, change symbol names to avoid conflicts (e.g. avoid build failure on MacOS X with --enable-shared) M ./libbench2/getopt.h +26 Mon Feb 7 22:36:42 EST 2005 stevenj * [project @ 2005-02-08 03:36:42 by stevenj] grr, more bugfixes for in-place case M ./reodft/reodft00e-splitradix.c -9 +13 Mon Feb 7 22:29:35 EST 2005 athena * [project @ 2005-02-08 03:29:35 by athena] removed relics of FRANZ mode M ./dft/codelets/standard/Makefile.am -19 Mon Feb 7 18:48:36 EST 2005 athena * [project @ 2005-02-07 23:48:36 by athena] Somehow xlc does not like ``vector int dummy;'' M ./simd/altivec.c -2 +2 Mon Feb 7 13:59:47 EST 2005 athena * [project @ 2005-02-07 18:59:47 by athena] There is no need to enable sse to make the distribution. This might have been true in the past but not anymore. M ./mkdist.sh -6 +8 Mon Feb 7 13:55:17 EST 2005 athena * [project @ 2005-02-07 18:55:17 by athena] Oops---included fortran file in C sources M ./api/Makefile.am -1 +1 Mon Feb 7 13:42:45 EST 2005 athena * [project @ 2005-02-07 18:42:45 by athena] Set version string at ``make dist'' time, not at ``configure'' time, so we know whether a user is using the fma version or not. M ./api/Makefile.am -2 +9 M ./api/version.c -2 +3 Sun Feb 6 17:00:33 EST 2005 athena * [project @ 2005-02-06 22:00:33 by athena] Removed useless files R ./genfft/gen_hc2r_noinline.ml R ./genfft/gen_notw_noinline.ml R ./genfft/gen_notw_noinline_c.ml R ./genfft/gen_r2hc_noinline.ml Sun Feb 6 16:59:39 EST 2005 athena * [project @ 2005-02-06 21:59:39 by athena] Different (simpler?) way to prevent the compiler from optimizing loop inductive variables. We now explicitly corrupt stride variables by xor-ing them with another variable that happens to be zero (but the compiler does not know it). In this way, the compiler does not attempt to extract a zillion loop indices from codelets, which would overflow the register set. Set the -fno-loop-optimize flag to further help the process. Consequences: removed m* codelets. Smaller library size. Slightly faster code with gcc/powerpc (including altivec). Much faster code with xlc/powerpc. No changes for gcc/pentium. Maybe slightly faster with icc/pentium. M ./configure.ac -11 +9 M ./dft/codelets/standard/Makefile.am -52 +3 M ./dft/simd/codelets/Makefile.am -30 +6 M ./genfft/Makefile.am -36 +10 M ./genfft/c.ml -1 +4 M ./genfft/c.mli -1 +2 M ./genfft/gen_hc2hc.ml -3 +5 M ./genfft/gen_hc2r.ml -3 +7 M ./genfft/gen_notw.ml -3 +6 M ./genfft/gen_notw_c.ml -3 +6 M ./genfft/gen_r2hc.ml -3 +7 M ./genfft/gen_twiddle.ml -3 +5 M ./genfft/gen_twiddle_c.ml -3 +5 M ./genfft/gen_twidsq.ml -3 +6 M ./genfft/gen_twidsq_c.ml -3 +6 M ./genfft/genutil.ml -1 +2 M ./genfft/simd.ml -1 +2 M ./kernel/ifftw.h -3 +7 M ./kernel/stride.c -1 +3 M ./rdft/codelets/hc2r/Makefile.am -13 +3 M ./rdft/codelets/r2hc/Makefile.am -13 +3 M ./support/Makefile.codelets -4 Sat Feb 5 18:51:08 EST 2005 stevenj * [project @ 2005-02-05 23:51:08 by stevenj] paranoia about in-place rodft00 plans M ./reodft/reodft00e-splitradix.c -3 +17 Sat Feb 5 18:39:55 EST 2005 stevenj * [project @ 2005-02-05 23:39:55 by stevenj] don't believe pcost when using the estimator...there is no point, and it screws up estimator hacks to prefer in-codelet loops to vecloops M ./kernel/planner.c -2 +2 Sat Feb 5 18:34:25 EST 2005 athena * [project @ 2005-02-05 23:34:25 by athena] Reduced optimization level from -O3 to -O for xlc, since -O generates faster code. M ./m4/ax_cc_maxopt.m4 -4 +4 Sat Feb 5 16:26:58 EST 2005 stevenj * [project @ 2005-02-05 21:26:58 by stevenj] whoops, only applicable to redft00/rodft00 plans M ./reodft/reodft00e-splitradix.c -1 +2 Sat Feb 5 16:22:39 EST 2005 stevenj * [project @ 2005-02-05 21:22:39 by stevenj] fixed in-place operation, and don't create size-0 sub-plans M ./reodft/reodft00e-splitradix.c -17 +16 Fri Feb 4 11:30:30 EST 2005 athena * [project @ 2005-02-04 16:30:30 by athena] Autodetect altivec on linux. This code works with gcc-3.4 and -maltivec, with or without -mabi=altivec. The code *should* work with gcc-3.3 without -mabi=altivec. However, disabling -mabi=altivec on gcc-3.4 produces much worse code (I don't know why). M ./simd/altivec.c -5 +31 Fri Jan 28 00:04:58 EST 2005 stevenj * [project @ 2005-01-28 05:04:58 by stevenj] update reference M ./doc/fftw3.texi -3 +3 Thu Jan 27 15:48:28 EST 2005 stevenj * [project @ 2005-01-27 20:48:28 by stevenj] note that DCT-II/III are often called the'' DCT/DCT M ./doc/fftw3.texi -11 +17 Fri Jan 21 14:42:04 EST 2005 stevenj * [project @ 2005-01-21 19:42:04 by stevenj] added MSVC++ for ia64 (based on information at http://www.intel.com/cd/ids/developer/asmo-na/eng/19949.htm?prn=Y) M ./kernel/cycle.h -1 +21 Fri Jan 21 14:22:50 EST 2005 stevenj * [project @ 2005-01-21 19:22:50 by stevenj] vc++ defines _M_AMD64 on x86-64, apparently M ./kernel/cycle.h -3 +3 Tue Jan 18 22:30:27 EST 2005 stevenj * [project @ 2005-01-19 03:30:27 by stevenj] avoid gratuitous breakage with -Werror, requested by Simon Perreault M ./m4/acx_pthread.m4 -2 +2 Mon Jan 17 18:54:55 EST 2005 stevenj * [project @ 2005-01-17 23:54:55 by stevenj] comment typo M ./m4/ax_gcc_aligns_stack.m4 -2 +2 Sat Jan 15 16:56:23 EST 2005 stevenj * [project @ 2005-01-15 21:56:23 by stevenj] bumped shared-lib revision# M ./configure.ac -1 +1 Sat Jan 15 16:35:42 EST 2005 stevenj * [project @ 2005-01-15 21:35:42 by stevenj] add X(estimate_cost) to get estimator cost, and print from bench, to aid in tweaking estimator M ./api/fftw3.h -1 +2 M ./api/flops.c +5 M ./kernel/ifftw.h -1 +2 M ./kernel/planner.c -6 +11 M ./tests/bench.c -1 +1 Sat Jan 15 14:57:56 EST 2005 stevenj * [project @ 2005-01-15 19:57:56 by stevenj] M ./doc/fftw3.texi -3 +3 Sat Jan 15 12:57:07 EST 2005 stevenj * [project @ 2005-01-15 17:57:07 by stevenj] formatting fix M ./doc/fftw3.texi -3 +9 Sat Jan 15 12:31:28 EST 2005 stevenj * [project @ 2005-01-15 17:31:28 by stevenj] tweaks M ./doc/fftw3.texi -17 +17 M ./reodft/Makefile.am +1 M ./reodft/conf.c -4 +7 Sat Jan 15 12:03:24 EST 2005 stevenj * [project @ 2005-01-15 17:03:24 by stevenj] use less buffer space M ./reodft/reodft00e-splitradix.c -22 +23 Sat Jan 15 01:41:58 EST 2005 stevenj * [project @ 2005-01-15 06:41:58 by stevenj] added split-radix-based dct/dst I for odd n A ./reodft/reodft00e-splitradix.c M ./doc/fftw3.texi -16 +21 M ./reodft/Makefile.am -1 +1 M ./reodft/conf.c -1 +2 M ./reodft/redft00e-r2hc.c -2 +4 M ./reodft/reodft.h +1 M ./reodft/reodft00e-splitradix.c +337 M ./reodft/rodft00e-r2hc.c -2 +4 Fri Jan 14 21:50:08 EST 2005 stevenj * [project @ 2005-01-15 02:50:08 by stevenj] M ./api/fftw3.h -3 +3 Fri Jan 14 21:49:55 EST 2005 stevenj * [project @ 2005-01-15 02:49:55 by stevenj] warn silly users who confuse CVS id with FFTW version M ./api/fftw3.h -1 +3 Fri Jan 14 16:57:36 EST 2005 stevenj * [project @ 2005-01-14 21:57:36 by stevenj] get sparc cpu type on solaris as well as with linux M ./m4/ax_gcc_archflag.m4 -13 +11 Thu Jan 13 19:21:58 EST 2005 stevenj * [project @ 2005-01-14 00:21:58 by stevenj] detect prescott mobile (f37) M ./m4/ax_gcc_archflag.m4 -2 +2 Thu Jan 13 18:09:52 EST 2005 stevenj * [project @ 2005-01-13 23:09:52 by stevenj] use cpuid for x86_64 as well as i[56]86 M ./bootstrap.sh +1 M ./m4/ax_gcc_archflag.m4 -2 +2 Thu Jan 13 17:59:55 EST 2005 stevenj * [project @ 2005-01-13 22:59:55 by stevenj] update with x86info 1.7 and other sources (identify k8, nocona, etc), handle nonzero leading bytes in eax M ./m4/ax_gcc_archflag.m4 -23 +27 Thu Jan 13 16:30:33 EST 2005 stevenj * [project @ 2005-01-13 21:30:33 by stevenj] compactified check for JOINABLE; use AC_DEFINE_UNQUOTED instead of AC_DEFINE for PTHREAD_CREATE_JOINABLE (thanks to Oliver Niekrenz for the bug report) M ./m4/acx_pthread.m4 -23 +16 Wed Jan 12 12:22:13 EST 2005 athena * [project @ 2005-01-12 17:22:13 by athena] The scheduler hack was incorrect because it swapped instructions of the form A = *B and *B = C. Fixed. M ./genfft/annotate.ml -7 +13 Tue Jan 11 22:13:24 EST 2005 athena * [project @ 2005-01-12 03:13:24 by athena] Quote expressions such as ``if test $FOO = yes'' when $FOO may be empty. Also, $GCC is set to either ``yes'' or empty, never to ``no''. M ./m4/ax_cc_maxopt.m4 -2 +2 M ./m4/ax_gcc_archflag.m4 -2 +2 Tue Jan 11 19:30:47 EST 2005 athena * [project @ 2005-01-12 00:30:47 by athena] Hmm---somehow the previous commit did not work. M ./TODO -6 M ./configure.ac +9 M ./simd/altivec.c -3 +3 Tue Jan 11 16:54:45 EST 2005 athena * [project @ 2005-01-11 21:54:45 by athena] Fixed various gcc-related problems on powerpc: - gcc-3.4 becomes totally confused by expressions like vec_add(a, vec_add(b, vec_add(c, ...))) The compiler uses gigabytes of memory and then crashes, presumably because of the exponential-time search problem involved in typing the above expression (since vec_add can take either ints or floats). I changed VADD and similar macros to be inline functions, thus constraining the type system. - New flags --param inline-unit-growth=1000 --param large-function-growth=1000 to work around limitations of the gcc-3.4 inliner. M ./simd/simd-altivec.h -14 +12 Mon Jan 10 21:27:24 EST 2005 athena * [project @ 2005-01-11 02:27:24 by athena] Check for HAVE_ALTIVEC_H M ./simd/simd-altivec.h -1 +2 Mon Jan 10 21:09:30 EST 2005 athena * [project @ 2005-01-11 02:09:30 by athena] Remove support for altivec using gcc builtins, since these keep changing across gcc versions. These changes work on gcc-3.4/linux; I haven't tried MacOS X yet. (The altivec ``spec'' differs between Motorola/Apple and gcc, grrr...) M ./configure.ac -1 +2 M ./simd/altivec.c -8 +5 M ./simd/simd-altivec.h -69 +5 Mon Jan 10 18:57:30 EST 2005 athena * [project @ 2005-01-10 23:57:30 by athena] Stylistic changes M ./rdft/rank0-vrank2.c -21 +21 Mon Jan 10 17:34:41 EST 2005 athena * [project @ 2005-01-10 22:34:41 by athena] Changed incorrect ugliness condition. M ./rdft/dft-r2hc.c -4 +2 Mon Jan 10 16:09:43 EST 2005 stevenj * [project @ 2005-01-10 21:09:43 by stevenj] note x86info version number that was used, to make it easier to update the cpuid for changes in later versions M ./m4/ax_gcc_archflag.m4 -2 +2 Mon Jan 10 15:00:51 EST 2005 athena * [project @ 2005-01-10 20:00:51 by athena] Make dft-r2hc non-UGLY for rank-0 problems M ./rdft/dft-r2hc.c -1 +6 Mon Jan 10 14:50:23 EST 2005 athena * [project @ 2005-01-10 19:50:23 by athena] Do not use -mcpu=970 on power4 processors, because power4 does not have altivec. M ./m4/ax_gcc_archflag.m4 -2 +2 Mon Jan 10 14:48:47 EST 2005 athena * [project @ 2005-01-10 19:48:47 by athena] Note gcc-3.4 problem with inlining. M ./TODO +5 Mon Jan 10 13:51:08 EST 2005 athena * [project @ 2005-01-10 18:51:08 by athena] Oops, forgot to remove ``static'' from the declaration of noninlinable functions. M ./genfft/gen_hc2r_noinline.ml -3 +3 M ./genfft/gen_notw_noinline_c.ml -3 +3 Mon Jan 10 12:31:26 EST 2005 athena * [project @ 2005-01-10 17:31:26 by athena] Recognize power4. Use ``head -n COUNT'' instead of obsolete ``head -COUNT'' (which fails on gentoo). M ./m4/ax_gcc_archflag.m4 -3 +4 Sun Jan 9 22:12:16 EST 2005 athena * [project @ 2005-01-10 03:12:16 by athena] Remind to add FAQ entry concerning gcc-3.4.[1-3] crashes. M ./TODO +4 Sun Jan 9 21:53:08 EST 2005 stevenj * [project @ 2005-01-10 02:53:08 by stevenj] whoops M ./m4/ax_gcc_version.m4 -3 +3 Sun Jan 9 21:48:02 EST 2005 stevenj * [project @ 2005-01-10 02:48:02 by stevenj] support checking for major.minor.patchlevel M ./m4/ax_gcc_version.m4 -10 +11 Sun Jan 9 21:40:18 EST 2005 athena * [project @ 2005-01-10 02:40:18 by athena] Revert CODELET_OPTIM to -O on IA32, which is faster than -O2. M ./configure.ac -3 +4 Sun Jan 9 20:30:12 EST 2005 athena * [project @ 2005-01-10 01:30:12 by athena] /bin/sh allows no spaces in assignments. M ./configure.ac -1 +1 Sun Jan 9 20:05:55 EST 2005 athena * [project @ 2005-01-10 01:05:55 by athena] Make non-inlinable functions external, so that gcc becomes confused and does not try to inline them. M ./genfft/gen_hc2r_noinline.ml -5 +5 M ./genfft/gen_notw_noinline.ml -5 +5 M ./genfft/gen_notw_noinline_c.ml -5 +5 Sun Jan 9 13:44:25 EST 2005 athena * [project @ 2005-01-09 18:44:25 by athena] Add -fno-web to CFLAGS, because -fweb destroys FMAs. M ./configure.ac -13 +18 Sun Jan 9 10:31:47 EST 2005 athena * [project @ 2005-01-09 15:31:47 by athena] Allow -mcpu=970 besides -mcpu=G5 M ./m4/ax_gcc_archflag.m4 -2 +2 Sun Jan 9 10:26:20 EST 2005 athena * [project @ 2005-01-09 15:26:20 by athena] configure was not using -fno-schedule-insns :-( M ./configure.ac -2 +1 Sun Jan 9 08:52:40 EST 2005 athena * [project @ 2005-01-09 13:52:40 by athena] In mkplan() and elsewhere, use solver index instead of solver *pointer*, which looks marginally clearer. M ./kernel/planner.c -18 +22 Sun Jan 9 08:15:36 EST 2005 athena * [project @ 2005-01-09 13:15:36 by athena] Split planner hash table into two tables, for blessed and unblessed solutions respectively. Now an unblessed solution never overwrites a blessed solution, thus avoiding wisdom leakage by construction. Further, forget() is now a O(1) operation, which speeds up the estimator when the wisdom table is large. M ./TODO -6 M ./kernel/ifftw.h -14 +17 M ./kernel/planner.c -157 +126 Sat Jan 8 21:19:45 EST 2005 athena * [project @ 2005-01-09 02:19:45 by athena] New TODO idea. M ./TODO +7 Thu Jan 6 11:02:29 EST 2005 athena * [project @ 2005-01-06 16:02:29 by athena] Split search() into two routines to make the UGLY/NO_UGLY logic obvious. M ./kernel/planner.c -41 +48 Fri Dec 17 16:08:54 EST 2004 stevenj * [project @ 2004-12-17 21:08:54 by stevenj] push/pop 64-bit registers on ia64; thanks to Orion Poplawski for the fix M ./simd/3dnow.c -1 +13 M ./simd/sse.c -1 +7 M ./simd/sse2.c -1 +7 Thu Dec 9 21:41:09 EST 2004 stevenj * [project @ 2004-12-10 02:41:09 by stevenj] patch from FreeBSD ports - FreeBSD does not have memalign, but its malloc is 16-byte aligned M ./kernel/kalloc.c -1 +5 Tue Nov 23 17:06:47 EST 2004 stevenj * [project @ 2004-11-23 22:06:47 by stevenj] don't compile taint.c with SIMD_CFLAGS (fixed Debian bug #259612) M ./simd/Makefile.am -3 +10 Thu Nov 18 11:37:32 EST 2004 stevenj * [project @ 2004-11-18 16:37:32 by stevenj] revert incorrect change -- codlist.c should be rebuilt, but it is built in the build directory and not in the source directory M ./support/Makefile.codelets -1 +1 Wed Nov 17 22:53:53 EST 2004 stevenj * [project @ 2004-11-18 03:53:53 by stevenj] $(CODLIST) should be rebuilt only if Makefile.am changes, or alternatively only in maintainer mode, to prevent stomping in the source directory during user builds. (Thanks to Grant Cook for the bug report.) M ./support/Makefile.codelets -1 +1 Sat Nov 13 13:43:01 EST 2004 stevenj * [project @ 2004-11-13 18:43:01 by stevenj] corrected #ifdef for icc/ia64, thanks to Matt Boman M ./kernel/cycle.h -16 +18 Sat Nov 13 13:34:55 EST 2004 stevenj * [project @ 2004-11-13 18:34:55 by stevenj] spelling correction (Larsen, not Larson) M ./NEWS -1 +1 Mon Nov 8 22:12:39 EST 2004 stevenj * [project @ 2004-11-09 03:12:39 by stevenj] use standard withval M ./m4/ax_gcc_archflag.m4 -3 +3 Mon Nov 8 22:09:16 EST 2004 stevenj * [project @ 2004-11-09 03:09:16 by stevenj] match doc M ./m4/ax_gcc_x86_cpuid.m4 -2 +2 Mon Nov 8 22:00:34 EST 2004 stevenj * [project @ 2004-11-09 03:00:34 by stevenj] formatting M ./m4/ax_openmp.m4 -3 +3 Mon Nov 8 21:59:33 EST 2004 stevenj * [project @ 2004-11-09 02:59:33 by stevenj] make sure OPENMP_CFLAGS environment variable is used correctly M ./m4/ax_openmp.m4 -2 +6 Mon Nov 8 21:46:50 EST 2004 stevenj * [project @ 2004-11-09 02:46:50 by stevenj] replace ax_check_cc_flags with more generic ax_check_compiler_flags A ./m4/ax_check_compiler_flags.m4 M ./configure.ac -13 +8 M ./m4/ax_cc_maxopt.m4 -7 +6 R ./m4/ax_check_cc_flags.m4 M ./m4/ax_check_compiler_flags.m4 +37 M ./m4/ax_gcc_aligns_stack.m4 -5 +4 M ./m4/ax_gcc_archflag.m4 -3 +3 Mon Nov 8 17:49:42 EST 2004 stevenj * [project @ 2004-11-08 22:49:42 by stevenj] separate macro for OpenMP test A ./m4/ax_openmp.m4 M ./configure.ac -40 +6 M ./m4/ax_cc_maxopt.m4 -3 +5 M ./m4/ax_openmp.m4 +58 Fri Nov 5 16:24:22 EST 2004 stevenj * [project @ 2004-11-05 21:24:22 by stevenj] typo M ./doc/fftw3.texi -2 +2 Fri Oct 29 00:48:13 EDT 2004 stevenj * [project @ 2004-10-29 04:48:13 by stevenj] M ./configure.ac -12 +19 Thu Oct 28 00:09:38 EDT 2004 stevenj * [project @ 2004-10-28 04:09:38 by stevenj] better guessing of sparc type on Linux M ./m4/ax_gcc_archflag.m4 -2 +15 Wed Oct 27 13:44:08 EDT 2004 stevenj * [project @ 2004-10-27 17:44:08 by stevenj] note default M ./m4/ax_gcc_archflag.m4 -2 +2 Wed Oct 27 13:41:57 EDT 2004 stevenj * [project @ 2004-10-27 17:41:57 by stevenj] tweak M ./m4/ax_gcc_archflag.m4 -2 +2 Wed Oct 27 13:34:25 EDT 2004 stevenj * [project @ 2004-10-27 17:34:25 by stevenj] comment M ./m4/ax_gcc_x86_cpuid.m4 -2 +3 Wed Oct 27 13:31:10 EDT 2004 stevenj * [project @ 2004-10-27 17:31:10 by stevenj] whoops, m4 is EXTRA_DIST, not SUBDIR, since it doesn't have a Makefile M ./Makefile.am -2 +2 Wed Oct 27 13:16:57 EDT 2004 stevenj * [project @ 2004-10-27 17:16:57 by stevenj] silence warnings M ./m4/ocaml.m4 -2 +2 Wed Oct 27 13:14:22 EDT 2004 stevenj * [project @ 2004-10-27 17:14:22 by stevenj] clean up m4 macros; try to detect correct gcc -march flag on x86; new --with-portable-binary, --with-gcc-arch= flags; use -O2 for codelets with gcc 3.4 to work around bug A ./m4/ A ./m4/acx_pthread.m4 A ./m4/amx_prog_as.m4 A ./m4/ax_cc_maxopt.m4 A ./m4/ax_check_cc_flags.m4 A ./m4/ax_gcc_aligns_stack.m4 A ./m4/ax_gcc_archflag.m4 A ./m4/ax_gcc_version.m4 A ./m4/ax_gcc_x86_cpuid.m4 A ./m4/ocaml.m4 M ./Makefile.am -2 +4 R ./acinclude.m4 R ./acx_pthread.m4 M ./bootstrap.sh -12 +10 M ./configure.ac -11 +26 M ./m4/acx_pthread.m4 +240 M ./m4/amx_prog_as.m4 +11 M ./m4/ax_cc_maxopt.m4 +126 M ./m4/ax_check_cc_flags.m4 +37 M ./m4/ax_gcc_aligns_stack.m4 +47 M ./m4/ax_gcc_archflag.m4 +147 M ./m4/ax_gcc_version.m4 +34 M ./m4/ax_gcc_x86_cpuid.m4 +37 M ./m4/ocaml.m4 +84 Tue Oct 26 16:46:14 EDT 2004 stevenj * [project @ 2004-10-26 20:46:14 by stevenj] rename cexp -> mcexp to avoid conflict with C99 builtin M ./libbench2/mp.c -3 +3 Mon Oct 25 16:58:23 EDT 2004 stevenj * [project @ 2004-10-25 20:58:23 by stevenj] use basename , w/o args, for compiler-name comparisons; also detect Compaq ccc on alpha-linus M ./acinclude.m4 -4 +10 Sun Oct 24 22:05:10 EDT 2004 stevenj * [project @ 2004-10-25 02:05:10 by stevenj] note recent icc problems M ./doc/FAQ/fftw-faq.bfnn -6 +9 Sun Oct 24 02:10:12 EDT 2004 stevenj * [project @ 2004-10-24 06:10:12 by stevenj] whoops, disable semaphores again (for now) M ./threads/threads.c -1 +1 Sun Oct 24 02:04:58 EDT 2004 stevenj * [project @ 2004-10-24 06:04:58 by stevenj] POSIX semaphores are *not* the same as SYSV semaphores M ./threads/threads.c -2 +2 Sun Oct 24 01:18:14 EDT 2004 stevenj * [project @ 2004-10-24 05:18:14 by stevenj] re-implement threaded stuff; dftw now takes parameters to indicate a portion of m loop A ./rdft/hc2hc.c A ./rdft/hc2hc.h A ./threads/ct.c A ./threads/hc2hc.c M ./dft/conf.c -2 +2 M ./dft/ct.c -8 +11 M ./dft/ct.h -8 +11 M ./dft/ctsq.c -2 +2 M ./dft/dft.h -5 +2 M ./dft/dftw-direct.c -15 +29 M ./dft/dftw-directbuf.c -15 +30 M ./dft/dftw-generic.c -14 +31 M ./dft/kdft-dif.c -7 +3 M ./dft/kdft-difsq.c -2 +2 M ./dft/kdft-dit.c -7 +3 M ./kernel/ifftw.h -1 +2 M ./kernel/twiddle.c -1 +10 M ./rdft/Makefile.am -1 +1 M ./rdft/conf.c -2 +2 R ./rdft/ct.c R ./rdft/ct.h M ./rdft/hc2hc-common.c -4 +33 M ./rdft/hc2hc-direct.c -26 +42 M ./rdft/hc2hc-directbuf.c -25 +44 M ./rdft/hc2hc-generic.c -37 +104 M ./rdft/hc2hc.c +221 M ./rdft/hc2hc.h +60 M ./rdft/khc2hc.c -8 +4 M ./rdft/rdft.h -3 +1 M ./threads/Makefile.am -2 +2 R ./threads/ct-dit.c M ./threads/ct.c +253 M ./threads/dft-vrank-geq1.c -2 +2 R ./threads/hc2hc-dif.c R ./threads/hc2hc-dit.c M ./threads/hc2hc.c +238 M ./threads/threads.c -27 +8 M ./threads/threads.h -5 +4 Thu Oct 21 20:44:51 EDT 2004 stevenj * [project @ 2004-10-22 00:44:51 by stevenj] more C++ notes M ./doc/fftw3.texi -4 +11 Thu Oct 14 09:50:38 EDT 2004 stevenj * [project @ 2004-10-14 13:50:38 by stevenj] note bug report for VC++ 6.0 from Dale Dickerhoof M ./doc/FAQ/fftw-faq.bfnn -1 +4 Fri Oct 1 16:06:59 EDT 2004 stevenj * [project @ 2004-10-01 20:06:59 by stevenj] fmt M ./api/fftw3.h -3 +3 Fri Oct 1 15:59:17 EDT 2004 stevenj * [project @ 2004-10-01 19:58:50 by stevenj] comment typo M ./rdft/vrank3-transpose.c -3 +3 Fri Oct 1 15:48:09 EDT 2004 stevenj * [project @ 2004-10-01 19:48:09 by stevenj] bug fix -- ishift/oshift only apply to execution of child plan M ./rdft/dft-r2hc.c -9 +4 Thu Sep 30 21:12:47 EDT 2004 athena * [project @ 2004-10-01 01:12:47 by athena] New planner that tries never to lose wisdom. M ./api/fftw3.h -2 +1 M ./api/mapflags.c -1 M ./kernel/ifftw.h -6 +4 M ./kernel/planner.c -137 +183 Thu Sep 30 13:36:43 EDT 2004 athena * [project @ 2004-09-30 17:36:43 by athena] Nested comment was triggering a warning. M ./api/fftw3.h -2 +2 Fri Sep 10 15:20:07 EDT 2004 stevenj * [project @ 2004-09-10 19:20:07 by stevenj] system "root" under dgjpp is /dev/env/DJDIR, not /dev/env/DJGPP, according to djgpp's libc.info; patch confirmed with J. M. Guerrero M ./api/import-system-wisdom.c -1 +1 Wed Sep 8 18:50:03 EDT 2004 stevenj * [project @ 2004-09-08 22:50:03 by stevenj] some minor portability fixes for djgpp; thanks to Juan Manuel Guerrero for the patch M ./api/import-system-wisdom.c -1 +8 M ./tests/Makefile.am -4 +4 M ./tools/fftw-wisdom-to-conf.in -1 +1 Thu Aug 19 12:41:23 EDT 2004 stevenj * [project @ 2004-08-19 16:41:23 by stevenj] pointer to tutorial for quick start M ./README +3 Thu Aug 19 12:39:50 EDT 2004 stevenj * [project @ 2004-08-19 16:39:50 by stevenj] point users to manual M ./api/fftw3.h -1 +16 Sat Aug 7 13:42:22 EDT 2004 stevenj * [project @ 2004-08-07 17:42:22 by stevenj] minor typo M ./doc/fftw3.texi -2 +2 Sun Jul 18 18:54:18 EDT 2004 stevenj * [project @ 2004-07-18 22:54:18 by stevenj] use __DECCXX for Compaq cxx, not Linux-specific symbol M ./kernel/cycle.h -2 +2 Fri Jul 16 13:55:25 EDT 2004 stevenj * [project @ 2004-07-16 17:55:25 by stevenj] patch by John Bowman to make cycle counter work with DEC cxx under Linux M ./kernel/cycle.h -2 +2 Wed Jun 30 00:45:10 EDT 2004 stevenj * [project @ 2004-06-30 04:45:10 by stevenj] updated pruned FFT discussion, with link to further details on www.fftw.org/pruned.html M ./doc/FAQ/fftw-faq.bfnn -8 +16 M ./doc/FAQ/html.refs +1 Mon Jun 14 20:08:27 EDT 2004 stevenj * [project @ 2004-06-15 00:08:27 by stevenj] darwin is based on freebsd M ./acx_pthread.m4 -2 +2 Thu Jun 3 14:23:41 EDT 2004 stevenj * [project @ 2004-06-03 18:23:41 by stevenj] in --with-windows-f77-mangling, add lowercase + single underscore for Intel compilers, etc. (thanks to David Gomez for the bug report) M ./api/f77api.c +5 Wed Apr 7 00:46:07 EDT 2004 stevenj * [project @ 2004-04-07 04:46:07 by stevenj] whoops, extra alignment check M ./rdft/rank0-vrank2.c -1 +5 Wed Apr 7 00:16:49 EDT 2004 stevenj * [project @ 2004-04-07 04:16:49 by stevenj] disable most 2-float-as-double copying, add alignment check in one remaining place M ./kernel/ifftw.h -1 +11 M ./rdft/rank0-vrank2.c -22 +25 M ./rdft/vrank3-transpose.c -6 +6 Tue Apr 6 13:49:13 EDT 2004 stevenj * [project @ 2004-04-06 17:49:13 by stevenj] make sure it is clear that real-even/odd refers to symmetry, not size M ./doc/fftw3.texi -7 +7 Mon Apr 5 20:18:29 EDT 2004 stevenj * [project @ 2004-04-06 00:18:29 by stevenj] optimization M ./rdft/vrank3-transpose.c -11 +48 Fri Apr 2 21:31:00 EST 2004 stevenj * [project @ 2004-04-03 02:31:00 by stevenj] separate cutoff for ugliness...these cutoffs are still not ideal M ./rdft/vrank3-transpose.c -3 +4 Fri Apr 2 21:30:17 EST 2004 stevenj * [project @ 2004-04-03 02:30:17 by stevenj] transpose.c is gone M ./kernel/ifftw.h -15 +1 Fri Apr 2 21:18:27 EST 2004 stevenj * [project @ 2004-04-03 02:18:27 by stevenj] move all rank0 transforms to rdft A ./rdft/rank0-vrank2.c A ./rdft/vrank3-transpose.c M ./configure.ac +1 M ./dft/Makefile.am -2 +2 M ./dft/conf.c -4 +1 R ./dft/rank0.c R ./dft/vrank2-transpose.c R ./dft/vrank3-transpose.c M ./kernel/Makefile.am -2 +2 R ./kernel/transpose.c M ./rdft/Makefile.am -2 +3 M ./rdft/conf.c -1 +3 M ./rdft/dft-r2hc.c -20 +36 M ./rdft/rank0-vrank2.c +289 M ./rdft/rdft.h -2 +2 M ./rdft/vrank3-transpose.c +943 Fri Apr 2 20:35:35 EST 2004 stevenj * [project @ 2004-04-03 01:35:35 by stevenj] enable fp-moves/us comparison of rank-0 transforms M ./libbench2/mflops.c -2 +8 M ./libbench2/report.c -3 +7 Thu Apr 1 16:13:22 EST 2004 stevenj * [project @ 2004-04-01 21:13:22 by stevenj] whoops M ./kernel/transpose.c -1 +1 Thu Apr 1 15:25:30 EST 2004 stevenj * [project @ 2004-04-01 20:25:30 by stevenj] whoops M ./kernel/tensor7.c -3 +3 Wed Mar 31 18:11:02 EST 2004 stevenj * [project @ 2004-03-31 23:11:02 by stevenj] sort tensor dims by stride absolute values, not strides M ./kernel/tensor7.c -5 +7 Tue Mar 30 20:22:50 EST 2004 stevenj * [project @ 2004-03-31 01:22:50 by stevenj] M ./kernel/transpose.c -2 +1 Tue Mar 30 19:44:54 EST 2004 stevenj * [project @ 2004-03-31 00:44:54 by stevenj] added improved transpose algorithm for N x M where |N-M| is small M ./dft/dftw-generic.c +1 M ./dft/vrank2-transpose.c -7 +10 M ./dft/vrank3-transpose.c -6 +7 M ./kernel/ifftw.h -2 +2 M ./kernel/transpose.c -15 +73 M ./libbench2/problem.c -2 +4 Tue Mar 30 19:41:14 EST 2004 stevenj * [project @ 2004-03-31 00:41:14 by stevenj] check to make sure SIMD matches precision, and make sure user doesn't select both SSE and SSE2 M ./configure.ac -2 +7 Sun Mar 28 09:26:38 EST 2004 athena * [project @ 2004-03-28 14:26:38 by athena] Implemented hc2hc-generic hc2r. M ./rdft/hc2hc-generic.c -29 +79 Thu Mar 25 11:19:25 EST 2004 athena * [project @ 2004-03-25 16:19:25 by athena] Inverted loop for stride-1 access. M ./rdft/hc2hc-generic.c -12 +18 Thu Mar 25 11:18:49 EST 2004 athena * [project @ 2004-03-25 16:18:49 by athena] Swapped j <-> k for consistency M ./dft/dftw-generic.c -3 +3 Tue Mar 23 12:08:07 EST 2004 athena * [project @ 2004-03-23 17:08:07 by athena] Require that R be odd M ./rdft/hc2hc-generic.c -1 +1 Tue Mar 23 11:49:01 EST 2004 athena * [project @ 2004-03-23 16:49:01 by athena] Implemented hc2hc-generic (DIT only for now). A ./rdft/hc2hc-generic.c M ./rdft/Makefile.am -3 +3 M ./rdft/conf.c -5 +4 M ./rdft/dft-r2hc.c -2 +2 M ./rdft/hc2hc-generic.c +216 M ./rdft/rdft.h +1 Mon Mar 22 14:43:16 EST 2004 athena * [project @ 2004-03-22 19:43:16 by athena] Relax equality of twiddle description, since the `i' field is not used by TW_FULL or TW_HALF. M ./kernel/twiddle.c -5 +17 Mon Mar 22 13:22:44 EST 2004 athena * [project @ 2004-03-22 18:22:41 by athena] Do not allocate tw_instr's on the stack. Thus, the ``consistency check'' in twiddle.c becomes wrong. M ./dft/dftw-generic.c -8 +1 M ./kernel/twiddle.c -2 +1 Mon Mar 22 13:21:28 EST 2004 athena * [project @ 2004-03-22 18:21:28 by athena] Fixed incorrect malloc()/free() logic. M ./libbench2/mp.c -3 +6 Mon Mar 22 09:04:37 EST 2004 athena * [project @ 2004-03-22 14:04:37 by athena] Silence warnings M ./rdft/hc2hc-directbuf.c -2 +2 Mon Mar 22 09:02:55 EST 2004 athena * [project @ 2004-03-22 14:02:55 by athena] Separate file for hc2hc common routines A ./rdft/hc2hc-common.c M ./rdft/Makefile.am -4 +4 M ./rdft/ct.c -35 M ./rdft/hc2hc-common.c +58 Mon Mar 22 08:23:56 EST 2004 athena * [project @ 2004-03-22 13:23:56 by athena] (re)Implemented buffered hc2hc. Slight simplification of twiddle-factors management. A ./rdft/hc2hc-directbuf.c M ./dft/dftw-directbuf.c -2 +2 M ./rdft/Makefile.am -2 +3 M ./rdft/ct.h +2 M ./rdft/hc2hc-direct.c -16 +8 M ./rdft/hc2hc-directbuf.c +238 M ./rdft/khc2hc.c -1 +2 Sun Mar 21 19:53:05 EST 2004 athena * [project @ 2004-03-22 00:53:05 by athena] Incremented libtool revision number before we forget. M ./configure.ac -1 +1 Sun Mar 21 19:25:56 EST 2004 athena * [project @ 2004-03-22 00:25:56 by athena] Fixed opcnt M ./rdft/hc2hc-direct.c -2 +5 Sun Mar 21 17:56:15 EST 2004 athena * [project @ 2004-03-21 22:56:15 by athena] Renamed files. These solvers are not really cooley-tukey. A ./dft/dftw-direct.c A ./dft/dftw-directbuf.c A ./dft/dftw-generic.c M ./dft/Makefile.am -5 +5 R ./dft/ct-directw.c R ./dft/ct-directwbuf.c R ./dft/ct-generic.c M ./dft/dftw-direct.c +153 M ./dft/dftw-directbuf.c +213 M ./dft/dftw-generic.c +304 Sun Mar 21 12:38:45 EST 2004 athena * [project @ 2004-03-21 17:38:45 by athena] Started moving rdft/ to the new cooley-tukey ontology A ./rdft/ct.c A ./rdft/ct.h A ./rdft/hc2hc-direct.c A ./rdft/khc2hc.c M ./dft/ct.h -1 +1 M ./genfft/gen_hc2hc.ml -6 +3 M ./rdft/Makefile.am -5 +4 M ./rdft/codelet-rdft.h -3 +1 M ./rdft/ct.c +253 M ./rdft/ct.h +54 R ./rdft/hc2hc-buf.c R ./rdft/hc2hc-dif.c M ./rdft/hc2hc-direct.c +182 R ./rdft/hc2hc-dit.c R ./rdft/hc2hc.c R ./rdft/hc2hc.h R ./rdft/khc2hc-dif.c R ./rdft/khc2hc-dit.c M ./rdft/khc2hc.c +32 M ./rdft/rdft.h -6 +1 Sun Mar 21 10:59:42 EST 2004 athena * [project @ 2004-03-21 15:59:42 by athena] Plans in ct-*.c are subtypes of plan_dftw, not plan_dft M ./dft/ct-directw.c -2 +2 M ./dft/ct-directwbuf.c -2 +2 M ./dft/ct-generic.c -1 +1 Sun Mar 21 10:38:18 EST 2004 athena * [project @ 2004-03-21 15:38:18 by athena] Slight simplification M ./dft/ct-directw.c -2 +2 Sun Mar 21 10:20:06 EST 2004 athena * [project @ 2004-03-21 15:20:06 by athena] Minor simplification M ./dft/ct.c -4 +2 Sat Mar 20 08:43:57 EST 2004 athena * [project @ 2004-03-20 13:43:57 by athena] Workarounds for icc-8.0 nonsense. M ./simd/simd-sse.h +6 M ./simd/simd-sse2.h +6 Sun Mar 7 07:56:08 EST 2004 athena * [project @ 2004-03-07 12:56:08 by athena] FFTW_FORWARD is not technically an ``option''. M ./doc/fftw3.texi -2 +2 Tue Feb 24 12:17:06 EST 2004 stevenj * [project @ 2004-02-24 17:17:06 by stevenj] Alejandro requested that his name be removed from @author M ./acx_pthread.m4 -4 +5 Mon Feb 23 17:42:56 EST 2004 stevenj * [project @ 2004-02-23 22:42:56 by stevenj] GNU Pth emulation library check M ./acx_pthread.m4 -3 +12 Sat Feb 21 17:51:13 EST 2004 stevenj * [project @ 2004-02-21 22:51:13 by stevenj] calling can-do calls the estimating-planner, which creates wisdom that we don't want ...we should be able to do all of the documented problems, anyway M ./tools/fftw-wisdom.c -1 +1 Sat Feb 21 17:46:06 EST 2004 stevenj * [project @ 2004-02-21 22:46:06 by stevenj] don't forget_wisdom because of side effects M ./tests/bench.c -1 Sat Feb 21 17:42:47 EST 2004 stevenj * [project @ 2004-02-21 22:42:47 by stevenj] forget wisdom from can_do M ./tests/bench.c +1 Thu Feb 19 14:11:14 EST 2004 stevenj * [project @ 2004-02-19 19:11:14 by stevenj] parenthesization M ./api/malloc.c -1 +1 Fri Feb 13 07:20:31 EST 2004 athena * [project @ 2004-02-13 12:20:31 by athena] Split malloc into kernel_malloc and API malloc A ./api/malloc.c A ./kernel/kalloc.c M ./api/Makefile.am -14 +14 M ./api/malloc.c +32 M ./kernel/Makefile.am -5 +5 M ./kernel/alloc.c -127 +5 M ./kernel/ifftw.h -1 +6 M ./kernel/kalloc.c +140 M ./tests/bench.c +4 Thu Feb 12 15:42:20 EST 2004 stevenj * [project @ 2004-02-12 20:42:20 by stevenj] X(malloc) must be extern "C" M ./kernel/alloc.c -1 +4 Thu Feb 12 15:41:44 EST 2004 stevenj * [project @ 2004-02-12 20:41:44 by stevenj] satsify C++ compiler M ./dft/bluestein.c -2 +2 Thu Feb 5 20:39:14 EST 2004 stevenj * [project @ 2004-02-06 01:39:14 by stevenj] with the new flags, fma is definitely beneficial on PA-RISC with HP/UX cc M ./doc/FAQ/fftw-faq.bfnn -6 +7 Thu Feb 5 19:52:17 EST 2004 stevenj * [project @ 2004-02-06 00:52:17 by stevenj] grr, Ofaster etcetera are not supported under older versions of the compiler. Note that +Ofltacc *disables* fp-reordering optimizations (which are enabled by +Oall). +Optrs_ansi is the older version of the aliasing stuff M ./acinclude.m4 -1 +1 Thu Feb 5 19:26:01 EST 2004 stevenj * [project @ 2004-02-06 00:26:01 by stevenj] +Otype_safety=ansi on hpux M ./acinclude.m4 -1 +1 Thu Feb 5 19:22:34 EST 2004 stevenj * [project @ 2004-02-06 00:22:34 by stevenj] just use +Ofaster on hpux (+O3 +Onolimit +Olibcalls +Ofltacc=relaxed -Wl,+mergeseg) M ./acinclude.m4 -1 +1 Fri Jan 30 14:17:15 EST 2004 stevenj * [project @ 2004-01-30 19:17:15 by stevenj] check for win32 threads for mingw32; thanks to Alessio Massaro M ./configure.ac +8 Thu Jan 29 15:23:33 EST 2004 stevenj * [project @ 2004-01-29 20:23:33 by stevenj] added missing 'static', thanks to Alessio Massaro M ./threads/threads.c -3 +3 Fri Jan 9 16:36:48 EST 2004 stevenj * [project @ 2004-01-09 21:36:48 by stevenj] print more like bluestein M ./rdft/dht-rader.c -1 +1 Fri Jan 9 15:45:22 EST 2004 stevenj * [project @ 2004-01-09 20:45:22 by stevenj] fixed op count for R2HC_ONLY_CONV M ./rdft/dht-rader.c -3 +2 Fri Jan 9 15:41:50 EST 2004 stevenj * [project @ 2004-01-09 20:41:50 by stevenj] include DESTROY_INPUT in buffered flags for in-place...otherwise in-place hc2r uses rdft-dhtcvs diff M ./dft/buffered.c -1 +10 M ./rdft/buffered.c -1 +10 M ./rdft/buffered2.c -1 +9 Fri Jan 9 15:41:09 EST 2004 stevenj * [project @ 2004-01-09 20:41:09 by stevenj] resurrected R2HC_ONLY_CONV option to share plans and save on planning time M ./rdft/dht-rader.c -3 +46 Fri Jan 9 14:47:00 EST 2004 stevenj * [project @ 2004-01-09 19:47:00 by stevenj] precompute folding for cyclic convolution M ./rdft/dht-rader.c -13 +9 Wed Jan 7 16:48:39 EST 2004 stevenj * [project @ 2004-01-07 21:48:39 by stevenj] minor M ./doc/FAQ/fftw-faq.bfnn -1 +1 Wed Jan 7 16:48:25 EST 2004 stevenj * [project @ 2004-01-07 21:48:25 by stevenj] note reports of successful compilation on Windows M ./doc/FAQ/fftw-faq.bfnn -5 +5 Wed Jan 7 14:16:16 EST 2004 stevenj * [project @ 2004-01-07 19:16:16 by stevenj] citation year M ./reodft/reodft010e-r2hc.c -2 +2 Tue Jan 6 01:07:36 EST 2004 stevenj * [project @ 2004-01-06 06:07:36 by stevenj] comment M ./rdft/dht-rader.c -1 +1 Tue Jan 6 01:07:08 EST 2004 stevenj * [project @ 2004-01-06 06:07:08 by stevenj] comment fix M ./rdft/dht-rader.c -1 +1 Tue Jan 6 01:06:57 EST 2004 stevenj * [project @ 2004-01-06 06:06:57 by stevenj] fixed naming cruft M ./rdft/dht-rader.c -18 +18 Tue Jan 6 00:56:16 EST 2004 stevenj * [project @ 2004-01-06 05:56:16 by stevenj] space M ./rdft/dht-rader.c -1 Tue Jan 6 00:55:53 EST 2004 stevenj * [project @ 2004-01-06 05:55:45 by stevenj] comment M ./rdft/dht-rader.c -1 +1 Tue Jan 6 00:54:07 EST 2004 stevenj * [project @ 2004-01-06 05:54:07 by stevenj] moved assert M ./rdft/dht-rader.c -1 +1 Tue Jan 6 00:49:16 EST 2004 stevenj * [project @ 2004-01-06 05:49:16 by stevenj] comment M ./rdft/dht-rader.c +5 Tue Jan 6 00:41:06 EST 2004 stevenj * [project @ 2004-01-06 05:41:06 by stevenj] delete old R2HC_ONLY_CONV hack, now defunct M ./rdft/dht-rader.c -46 Tue Jan 6 00:32:58 EST 2004 stevenj * [project @ 2004-01-06 05:32:58 by stevenj] added padded real rader M ./rdft/dht-rader.c -56 +95 Mon Jan 5 22:56:58 EST 2004 stevenj * [project @ 2004-01-06 03:56:58 by stevenj] removed unused var M ./rdft/generic.c -1 +1 Mon Jan 5 21:20:29 EST 2004 stevenj * [project @ 2004-01-06 02:20:29 by stevenj] handle both FFT_SIGN values M ./rdft/generic.c -2 +14 Fri Jan 2 06:07:51 EST 2004 athena * [project @ 2004-01-02 11:07:51 by athena] Oops: d->ros ==> d->ios M ./rdft/codelets/r2hc.c -1 +1 Fri Jan 2 06:05:10 EST 2004 athena * [project @ 2004-01-02 11:05:10 by athena] Oops: d->ris should have been d->iis M ./rdft/codelets/hc2r.c -1 +1 Thu Jan 1 16:00:07 EST 2004 athena * [project @ 2004-01-01 21:00:07 by athena] Removed rdft rader cooley-tukey, to be superseded by a generic reduction of rdft twiddle problems to dft + pre/post processing M ./dft/Makefile.am -2 +2 M ./dft/dft.h -5 +1 R ./dft/rader-omega.c M ./dft/rader.c -2 +38 M ./rdft/Makefile.am -2 +2 M ./rdft/conf.c -2 +1 R ./rdft/rader-hc2hc.c Thu Jan 1 15:44:09 EST 2004 athena * [project @ 2004-01-01 20:44:09 by athena] In anticipation of the upcoming revision of rdft, removed rdft generic dit/dif cooley-tukey, in favor of generic rh2c and hc2r solvers. Cleaned up stuff that became unused after this change, such as TW_GENERIC. M ./dft/ct.c -33 +3 M ./dft/generic.c -1 +2 M ./kernel/ifftw.h -2 +3 M ./kernel/primes.c -1 +32 M ./kernel/twiddle.c -18 +1 M ./rdft/generic.c -258 +103 Thu Jan 1 12:59:30 EST 2004 athena * [project @ 2004-01-01 17:59:30 by athena] Removed useless file M ./kernel/Makefile.am -3 +3 M ./kernel/ifftw.h -2 +1 R ./kernel/square.c Fri Dec 26 13:54:00 EST 2003 stevenj * [project @ 2003-12-26 18:54:00 by stevenj] whoops, don't call AC_F77_DUMMY_MAIN if no Fortran compiler is found; thanks to Charles Radley for the bug report. M ./configure.ac -2 +3 Fri Dec 19 13:58:05 EST 2003 stevenj * [project @ 2003-12-19 18:58:05 by stevenj] guess good flags for Solaris/intel, suggested by J. Gregory Wright M ./acinclude.m4 -1 +1 Fri Dec 5 19:55:13 EST 2003 stevenj * [project @ 2003-12-06 00:55:13 by stevenj] blah M ./doc/FAQ/fftw-faq.bfnn -1 +1 M ./doc/FAQ/html.refs -1 +1 Sun Nov 30 06:59:41 EST 2003 athena * [project @ 2003-11-30 11:59:41 by athena] DIF generic solver was destroying the input. M ./rdft/generic.c -2 +16 Sat Nov 29 19:28:39 EST 2003 athena * [project @ 2003-11-30 00:28:39 by athena] Fixed bug that caused HC2R transforms to destroy the input in certain cases, even if the user specified FFTW_PRESERVE_INPUT. M ./NEWS +5 M ./rdft/rader-hc2hc.c -2 +14 Sat Nov 29 16:49:01 EST 2003 athena * [project @ 2003-11-29 21:49:01 by athena] Implemented swap_io hack for r2r verifier. M ./libbench2/verify-r2r.c -1 +8 Thu Nov 20 22:00:53 EST 2003 stevenj * [project @ 2003-11-21 03:00:53 by stevenj] citation M ./reodft/reodft010e-r2hc.c -2 +6 Fri Nov 14 20:57:55 EST 2003 athena * [project @ 2003-11-15 01:57:55 by athena] Trying to get ``make paranoid-check'' to work. (Still broken.) M ./kernel/ifftw.h -2 +3 M ./kernel/planner.c -2 +2 M ./tests/hook.c -9 +8 Fri Nov 14 20:05:54 EST 2003 stevenj * [project @ 2003-11-15 01:05:54 by stevenj] fixes for input-preservation tests M ./libbench2/bench-user.h -1 +2 M ./libbench2/tensor.c -1 +15 M ./libbench2/verify-dft.c -3 +12 M ./libbench2/verify-lib.c -1 +4 M ./libbench2/verify-r2r.c -13 +33 M ./libbench2/verify-rdft2.c -6 +22 M ./libbench2/verify.h -1 +1 Fri Nov 14 19:19:31 EST 2003 athena * [project @ 2003-11-15 00:19:31 by athena] Assume FFTW_PRESERVE_INPUT unless either the `d' flag is given in the problem, or the problem is multidimensional c2r (which fftw3 cannot without destroying the input). With this change, we can at least test that FFTW_PRESERVE_INPUT works in the c2r 1d case. M ./tests/bench.c -12 +21 Fri Nov 14 19:14:40 EST 2003 stevenj * [project @ 2003-11-15 00:14:40 by stevenj] apply should copy back input for input-preservation check M ./libbench2/verify-dft.c -1 +3 M ./libbench2/verify-r2r.c +13 M ./libbench2/verify-rdft2.c -1 +4 Fri Nov 14 19:01:36 EST 2003 athena * [project @ 2003-11-15 00:01:14 by athena] Undone previous bogus changes M ./rdft/rank-geq2-rdft2.c -1 +10 M ./tests/bench.c -14 +10 M ./tests/check.pl -1 Fri Nov 14 18:27:12 EST 2003 athena * [project @ 2003-11-14 23:27:12 by athena] Check dr[fb] in addition to r[fb] M ./tests/check.pl +1 Fri Nov 14 17:33:44 EST 2003 athena * [project @ 2003-11-14 22:33:42 by athena] Fixed conditions under which the rank-geq2-rdft2 solver is applicable. The old solver was not applicable for out-of-place problems unless DESTROY_INPUT. This is bogus. As long as the subsolvers honor !DESTROY_INPUT, the solver is always applicable. Changed semantics of test program, so that PRESERVE_INPUT is always true unless the problem specifies destroy_input explicitly. Without this change, there is no way to test the new solver. M ./rdft/rank-geq2-rdft2.c -10 +1 M ./tests/bench.c -10 +14 Thu Oct 30 15:10:42 EST 2003 stevenj * [project @ 2003-10-30 20:10:42 by stevenj] added AIX OpenMP (-qsmp=omp) support; thanks to Greg Bauer M ./configure.ac +5 Thu Oct 30 10:11:39 EST 2003 athena * [project @ 2003-10-30 15:11:39 by athena] G5 CFLAGS M ./acinclude.m4 +6 Fri Oct 24 04:17:39 EDT 2003 stevenj * [project @ 2003-10-24 08:17:39 by stevenj] western FAQ M ./doc/FAQ/fftw-faq.bfnn +5 Thu Oct 23 11:34:11 EDT 2003 athena * [project @ 2003-10-23 15:34:11 by athena] Oops. M ./simd/altivec.c -1 +2 Thu Oct 23 11:28:28 EDT 2003 athena * [project @ 2003-10-23 15:28:27 by athena] Autodetect altivec M ./configure.ac -2 +3 M ./simd/altivec.c -1 +28 M ./simd/simd-altivec.h -1 +2 Wed Oct 22 01:14:10 EDT 2003 stevenj * [project @ 2003-10-22 05:14:10 by stevenj] MinGW gets confused by a single / M ./tests/check.pl -4 +4 Fri Oct 17 10:46:41 EDT 2003 athena * [project @ 2003-10-17 14:46:41 by athena] Paranoid portability fix M ./libbench2/mp.c -2 +2 Thu Oct 16 11:07:46 EDT 2003 athena * [project @ 2003-10-16 15:07:46 by athena] size -> length, which should make clear that we are not talking about arbitrary precision. M ./doc/fftw3.texi -4 +4 Wed Oct 15 15:01:40 EDT 2003 stevenj * [project @ 2003-10-15 19:01:40 by stevenj] pruned transforms are a FAQ M ./doc/FAQ/fftw-faq.bfnn +11 Wed Oct 8 23:54:17 EDT 2003 stevenj * [project @ 2003-10-09 03:54:17 by stevenj] NO_SEARCH has already been mapped to FFTW_WISDOM_ONLY M ./TODO -1 +1 Wed Oct 8 23:53:19 EDT 2003 stevenj * [project @ 2003-10-09 03:53:19 by stevenj] newline M ./TODO +1 Sat Sep 27 20:27:32 EDT 2003 stevenj * [project @ 2003-09-28 00:27:32 by stevenj] fix M ./doc/fftw3.texi -2 +2 Sat Sep 27 20:24:39 EDT 2003 stevenj * [project @ 2003-09-28 00:24:39 by stevenj] clarification M ./doc/fftw3.texi -2 +4 Sat Sep 27 17:43:57 EDT 2003 stevenj * [project @ 2003-09-27 21:43:57 by stevenj] minor fix M ./doc/fftw3.texi -2 +2 Sat Sep 27 17:42:30 EDT 2003 stevenj * [project @ 2003-09-27 21:42:30 by stevenj] grammar M ./doc/fftw3.texi -2 +2 Sat Sep 27 17:29:04 EDT 2003 stevenj * [project @ 2003-09-27 21:29:04 by stevenj] html output fix M ./doc/fftw3.texi -5 +5 Sat Sep 27 17:22:48 EDT 2003 stevenj * [project @ 2003-09-27 21:22:48 by stevenj] mentioned sqrt(2) factors for DCT/DST M ./doc/fftw3.texi -5 +48 Sat Sep 27 17:07:18 EDT 2003 stevenj * [project @ 2003-09-27 21:07:18 by stevenj] FFTW_WISDOM_ONLY flag (undocumented for now), suggested by Phil Dumont M ./api/fftw3.h -1 +2 M ./api/mapflags.c +1 Tue Sep 23 23:36:19 EDT 2003 stevenj * [project @ 2003-09-24 03:36:19 by stevenj] removed UpTime code M ./kernel/cycle.h -20 +3 Tue Sep 23 23:27:29 EDT 2003 stevenj * [project @ 2003-09-24 03:27:29 by stevenj] updated documentation for mach_absolute_time M ./kernel/cycle.h -3 +3 Tue Sep 23 23:25:52 EDT 2003 stevenj * [project @ 2003-09-24 03:25:52 by stevenj] use mach_absolute_time on MacOS/Darwin, as a fallback; don't bother checking for UpTime since it requires extra libs M ./configure.ac -3 +2 M ./kernel/cycle.h -1 +11 Tue Sep 23 22:59:29 EDT 2003 stevenj * [project @ 2003-09-24 02:59:29 by stevenj] support Apple UpTime function for asm-less xlc, grrr... M ./configure.ac -2 +3 M ./kernel/cycle.h -1 +19 Tue Sep 23 15:42:29 EDT 2003 stevenj * [project @ 2003-09-23 19:42:29 by stevenj] additional paranoia for xlc etc. M ./api/api.h -1 +1 M ./api/fftw3.h -2 +2 Mon Sep 22 15:28:56 EDT 2003 stevenj * [project @ 2003-09-22 19:28:56 by stevenj] work around _Complex_I weirdness in xlc, reported by Greg Allen M ./api/api.h -2 +4 M ./api/fftw3.h -2 +2 Fri Sep 5 18:03:11 EDT 2003 stevenj * [project @ 2003-09-05 22:03:11 by stevenj] typo M ./doc/FAQ/fftw-faq.bfnn -2 +2 Fri Sep 5 13:11:40 EDT 2003 athena * [project @ 2003-09-05 17:11:40 by athena] New script that produces commercial version. A ./commercialize.sh Fri Sep 5 07:27:06 EDT 2003 athena * [project @ 2003-09-05 11:27:06 by athena] Noted that VC++ is buggy. Noted that we know nothing about Windows. Noted that the sky is blue as well. M ./doc/FAQ/fftw-faq.bfnn -10 +22 Tue Sep 2 09:04:19 EDT 2003 athena * [project @ 2003-09-02 13:04:19 by athena] Noted that certain arrays are no longer used after the planner has completed. M ./doc/fftw3.texi -2 +12 Tue Aug 26 08:22:38 EDT 2003 athena * [project @ 2003-08-26 12:22:38 by athena] Typo M ./doc/fftw3.texi -2 +2 Mon Aug 25 21:27:43 EDT 2003 athena * [project @ 2003-08-26 01:27:43 by athena] New item M ./TODO -2 +1 Thu Aug 21 17:36:08 EDT 2003 stevenj * [project @ 2003-08-21 21:36:08 by stevenj] try creating output file before planning (thanks to Phil Dumont for the suggestion) M ./tools/fftw-wisdom.c -10 +10 Tue Aug 19 10:08:07 EDT 2003 athena * [project @ 2003-08-19 14:08:07 by athena] Clarified fftw_cleanup() M ./doc/fftw3.texi -4 +7 Sat Aug 16 03:13:41 EDT 2003 stevenj * [project @ 2003-08-16 07:13:41 by stevenj] typo M ./doc/fftw3.texi -2 +2 Mon Jul 28 18:01:13 EDT 2003 stevenj * [project @ 2003-07-28 22:01:13 by stevenj] use time() instead of clock() (FIXME: what to do for non-POSIX systems?) ...thanks to JP Sugarbroad and James A. Treacy for the bug report M ./tools/fftw-wisdom.c -4 +4 Thu Jul 24 18:58:10 EDT 2003 athena * [project @ 2003-07-24 22:58:10 by athena] Need __volatile__ in sparc cycle counter. This is why the debian port hangs. M ./kernel/cycle.h -2 +2 Sun Jul 20 16:02:43 EDT 2003 stevenj * [project @ 2003-07-20 20:02:43 by stevenj] merged 3.0.1 notes M ./NEWS +30 Sun Jul 13 20:57:34 EDT 2003 stevenj * [project @ 2003-07-14 00:57:34 by stevenj] whoops M ./libbench2/bench-main.c -2 +2 Thu Jul 10 11:48:50 EDT 2003 athena * [project @ 2003-07-10 15:48:50 by athena] Dealing with constants in a way that seems to confuse gcc less. M ./simd/simd-sse.h -5 +2 M ./simd/simd-sse2.h -6 +3 Wed Jul 9 17:39:23 EDT 2003 athena * [project @ 2003-07-09 21:39:16 by athena] Enabled scheduler hack for FMA, where it seems to help. M ./genfft/annotate.ml -7 +3 M ./genfft/magic.ml -1 +5 M ./support/Makefile.codelets -1 +1 Wed Jul 9 12:57:38 EDT 2003 athena * [project @ 2003-07-09 16:57:38 by athena] Hmm---the new scheduler seems make things worse for gcc/x86, better for gcc/ppc, and about the same for icc/x86. Disabled for now. M ./genfft/annotate.ml -3 +6 Wed Jul 9 08:09:53 EDT 2003 athena * [project @ 2003-07-09 12:09:53 by athena] New scheduling pass that keeps ``x = a + b'' and ``y = a - b'' close together. This property was no longer automatic for the dags generated in SIMD mode. I cannot measure any speed difference due to this change. However, the change is justified by a minimal-screwup argument. Moreover, the sse2 fftw library is now 1% smaller than it was before. M ./genfft/annotate.ml -32 +70 Tue Jul 8 20:42:22 EDT 2003 athena * [project @ 2003-07-09 00:42:22 by athena] -(FNMS()) => FMS() M ./genfft/c.ml -1 +2 Sun Jul 6 13:53:23 EDT 2003 stevenj * [project @ 2003-07-06 17:53:23 by stevenj] added more convenient target name M ./doc/FAQ/Makefile.am +2 Sat Jul 5 13:30:10 EDT 2003 stevenj * [project @ 2003-07-05 17:30:10 by stevenj] typo M ./kernel/ifftw.h -2 +2 Sat Jul 5 13:19:36 EDT 2003 athena * [project @ 2003-07-05 17:19:36 by athena] Consistent naming M ./dft/ct-generic.c -2 +2 Sat Jul 5 13:05:51 EDT 2003 athena * [project @ 2003-07-05 17:05:51 by athena] Got rid of problemw. A ./dft/ct-directw.c A ./dft/ct-directwbuf.c A ./dft/ct-generic.c A ./dft/ct.h M ./dft/Makefile.am -5 +4 M ./dft/conf.c -3 +2 M ./dft/ct-directw.c +153 M ./dft/ct-directwbuf.c +213 M ./dft/ct-generic.c +304 M ./dft/ct.c -62 +24 M ./dft/ct.h +56 M ./dft/ctsq.c -3 +3 M ./dft/dft.h -39 +2 R ./dft/dftw-dft.c M ./dft/direct.c -2 +1 R ./dft/directw.c R ./dft/directwbuf.c M ./dft/generic.c -2 M ./dft/kdft-dif.c -4 +4 M ./dft/kdft-difsq.c -3 +3 M ./dft/kdft-dit.c -4 +4 M ./dft/plan.c -11 +1 R ./dft/problemw.c M ./dft/rader.c -2 M ./dft/solve.c -8 +1 Fri Jul 4 06:56:26 EDT 2003 athena * [project @ 2003-07-04 10:56:26 by athena] Increase TIME_MIN on intel only M ./kernel/cycle.h -1 +3 M ./kernel/timer.c -2 +2 Fri Jul 4 06:36:02 EDT 2003 athena * [project @ 2003-07-04 10:36:02 by athena] A little hack to get more consistent scheduling. M ./genfft/schedule.ml -7 +8 Thu Jul 3 16:47:42 EDT 2003 athena * [project @ 2003-07-03 20:47:42 by athena] New experimental scheduler (currently disabled). The old scheduler is ``optimal'' in the sense that it minimizes register pressure. The only way to reduce register pressure is to schedule dependent instructions as closely as possible, so as to minimize the life time of registers. This strategy maximizes the number of pipeline stalls, however. With enough registers and short enough pipelines, this tradeoff is fine. This is no longer the case for the devilish pipeline of the Pentium IV or (probably) the PowerPC 970. The new scheduler switches to a ``list scheduler'' for dags smaller than a specified size. The list scheduler executes a butterfly left to right one column at the time. This amounts to the best possible pipeline utilization, and the worst possible register pressure. The ``specified size'' defaults to 0, i.e., no change from fftw2 and fftw-3.0. It seems like a value of 7--10 produces the best results for Pentium IV (probably screwing the G3/G4 powerpcs and sparc, but I haven't tried.) As time goes by, we may want to increase this number to favor newer processors over older processors. M ./genfft/magic.ml -1 +3 M ./genfft/schedule.ml -9 +35 Wed Jun 25 17:43:59 EDT 2003 stevenj * [project @ 2003-06-25 21:43:59 by stevenj] remove non-portable use of tempfile; thanks to Nicolas Decoster for the patch M ./tools/fftw-wisdom-to-conf.in -1 +1 Wed Jun 25 17:14:03 EDT 2003 stevenj * [project @ 2003-06-25 21:14:03 by stevenj] increase stupid HP preprocessor limits M ./acinclude.m4 -1 +1 Thu Jun 19 15:21:52 EDT 2003 athena * [project @ 2003-06-19 19:21:52 by athena] Distribute gen_mdct.ml M ./genfft/Makefile.am -8 +8 Wed Jun 11 06:55:21 EDT 2003 athena * [project @ 2003-06-11 10:55:21 by athena] Cleared int/ptrdiff_t confusions M ./rdft/buffered2.c -2 +3 M ./rdft/rdft2-radix2.c -5 +5 Tue Jun 10 22:15:42 EDT 2003 athena * [project @ 2003-06-11 02:15:41 by athena] Cleared int/ptrdiff_t confusion M ./dft/dftw-dft.c -1 +1 M ./dft/directwbuf.c -2 +2 M ./dft/rank0.c -2 +2 M ./dft/vrank2-transpose.c -2 +2 M ./dft/vrank3-transpose.c -2 +2 M ./kernel/planner.c -2 +2 Sun Jun 8 09:52:57 EDT 2003 athena * [project @ 2003-06-08 13:52:57 by athena] Increased TIME_MIN. This seems to produce more reliable plans on Pentium IV. M ./kernel/timer.c -2 +2 Sat Jun 7 21:43:00 EDT 2003 athena * [project @ 2003-06-08 01:43:00 by athena] Removed relic -trivial-stores, which dates back to Franz's early experiments. Speed improved on SSE2, both with gcc and icc. M ./dft/simd/codelets/Makefile.am -1 +1 Thu Jun 5 22:29:52 EDT 2003 stevenj * [project @ 2003-06-06 02:29:52 by stevenj] fix direntry M ./doc/fftw3.texi -2 +2 Thu Jun 5 13:41:34 EDT 2003 stevenj * [project @ 2003-06-05 17:41:34 by stevenj] added imdct M ./genfft/gen_mdct.ml -13 +66 Wed Jun 4 19:54:38 EDT 2003 athena * [project @ 2003-06-04 23:54:38 by athena] Collect pattern (a * b) +- (c * d) in generic-arith, because this operation can usually be computed with one rounding in fixed-point (and it possibly exposes a FMA instruction) M ./genfft/c.ml -5 +15 Wed Jun 4 15:11:29 EDT 2003 athena * [project @ 2003-06-04 19:11:29 by athena] Generic-arithmetic unparser M ./genfft/c.ml -8 +31 M ./genfft/magic.ml -1 +5 Sun Jun 1 09:05:30 EDT 2003 athena * [project @ 2003-06-01 13:05:30 by athena] Oops---randomized CSE was using the same random numbers over and over M ./genfft/oracle.ml -8 +6 M ./genfft-k7/oracle.ml -11 +11 Sun Jun 1 07:01:17 EDT 2003 athena * [project @ 2003-06-01 11:01:17 by athena] Paranoia. M ./genfft/c.ml -3 +3 Sun Jun 1 07:00:54 EDT 2003 athena * [project @ 2003-06-01 11:00:54 by athena] Use relative error instead of absolute error, to avoid problems when normalization factors are used. M ./genfft/oracle.ml -3 +5 Sat May 31 22:11:28 EDT 2003 stevenj * [project @ 2003-06-01 02:11:28 by stevenj] slight opt M ./reodft/reodft11e-radix2.c -3 +3 Sat May 31 22:10:45 EDT 2003 stevenj * [project @ 2003-06-01 02:10:45 by stevenj] slight optimization M ./reodft/reodft11e-radix2.c -3 +3 Sat May 31 20:43:31 EDT 2003 stevenj * [project @ 2003-06-01 00:43:31 by stevenj] *W is const M ./genfft/gen_mdct.ml -3 +3 Sat May 31 20:41:15 EDT 2003 stevenj * [project @ 2003-06-01 00:41:15 by stevenj] comment M ./genfft/gen_mdct.ml -2 +4 Thu May 29 21:31:31 EDT 2003 stevenj * [project @ 2003-05-30 01:31:31 by stevenj] added experimental MDCT A ./genfft/gen_mdct.ml M ./genfft/Makefile.am -1 +7 M ./genfft/gen_mdct.ml +208 Wed May 28 22:01:37 EDT 2003 stevenj * [project @ 2003-05-29 02:01:37 by stevenj] altivec (fma) needs simd codlist.c too M ./mkdist.sh -1 +1 Wed May 28 22:00:49 EDT 2003 stevenj * [project @ 2003-05-29 02:00:49 by stevenj] make sure we include SIMD codlist.c for non-Unix folks M ./mkdist.sh -1 +1 Tue May 27 20:31:25 EDT 2003 stevenj * [project @ 2003-05-28 00:31:25 by stevenj] noted howmany_rank == 0 is a single transform M ./doc/fftw3.texi -1 +3 Tue May 27 20:02:31 EDT 2003 stevenj * [project @ 2003-05-28 00:02:31 by stevenj] further stride clarification M ./doc/fftw3.texi -4 +4 Mon May 26 10:21:22 EDT 2003 athena * [project @ 2003-05-26 14:21:22 by athena] Removed transposed dftw problems. I now consider transposed dftw a Bad Idea, since it does not apply to the case that it was originally meant for (speed up four-step) and it complicates the implementation of the other thing I want to try (dftw m-slices). M ./dft/Makefile.am -3 +3 M ./dft/conf.c -2 +1 M ./dft/ct.c -5 +3 M ./dft/ctsq.c -74 +57 M ./dft/dft.h -9 +5 M ./dft/dftw-dft.c -83 M ./dft/directw.c -5 +1 M ./dft/directwbuf.c -5 +1 R ./dft/directwsq.c M ./dft/kdft-difsq.c -2 +2 M ./dft/problemw.c -13 +9 Mon May 26 07:22:59 EDT 2003 athena * [project @ 2003-05-26 11:22:59 by athena] Obsolete comment M ./dft/buffered.c -6 +1 Sat May 24 15:00:53 EDT 2003 athena * [project @ 2003-05-24 19:00:53 by athena] comment M ./dft/ct.c -2 +2 Sat May 24 07:20:35 EDT 2003 athena * [project @ 2003-05-24 11:20:35 by athena] Oops---wrong test NO_UGLYP instead of !NO_UGLYP M ./dft/dftw-dft.c -3 +3 Sat May 24 07:05:34 EDT 2003 athena * [project @ 2003-05-24 11:05:34 by athena] Implemented radix r, where n=r^2 * p M ./dft/ct.c -19 +53 Wed May 21 01:54:32 EDT 2003 stevenj * [project @ 2003-05-21 05:54:32 by stevenj] xlc seems to properly use fma as well M ./doc/FAQ/fftw-faq.bfnn -2 +2 Tue May 20 23:07:43 EDT 2003 stevenj * [project @ 2003-05-21 03:07:43 by stevenj] print warning if there is no cycle counter M ./configure.ac +18 M ./doc/fftw3.texi -2 +3 Tue May 20 17:32:04 EDT 2003 stevenj * [project @ 2003-05-20 21:32:04 by stevenj] updated Funda reference M ./libbench2/verify-lib.c -1 +4 M ./libbench2/verify-r2r.c +3 Mon May 19 20:12:36 EDT 2003 athena * [project @ 2003-05-20 00:12:36 by athena] const M ./dft/ct.c -2 +2 Mon May 19 15:41:09 EDT 2003 athena * [project @ 2003-05-19 19:41:09 by athena] Implemented generic dif square transposed (q-style) solver. M ./dft/dftw-dft.c -8 +88 M ./dft/directwsq.c -2 +2 Mon May 19 07:00:36 EDT 2003 athena * [project @ 2003-05-19 11:00:36 by athena] applicable() is now a property of the solver (in anticipation of transposed solvers) M ./dft/dftw-dft.c -35 +35 Mon May 19 06:33:40 EDT 2003 athena * [project @ 2003-05-19 10:33:40 by athena] Slight cleanup M ./dft/dftw-dft.c -7 +7 Sun May 18 13:05:51 EDT 2003 athena * [project @ 2003-05-18 17:05:51 by athena] Nothing, really M ./dft/bluestein.c -9 +3 M ./kernel/ifftw.h -1 +2 M ./kernel/primes.c -1 +9 Sun May 18 09:05:20 EDT 2003 athena * [project @ 2003-05-18 13:05:20 by athena] Moved vector loop inside bytwiddle(), in anticipation of a q-style dftw-dit transposed solver. M ./dft/dftw-dft.c -38 +41 Sun May 18 08:52:02 EDT 2003 athena * [project @ 2003-05-18 12:52:02 by athena] Fixed flops count M ./dft/dftw-dft.c -1 +1 Sun May 18 08:47:20 EDT 2003 athena * [project @ 2003-05-18 12:47:20 by athena] style M ./dft/dftw-dft.c -5 +3 Sun May 18 07:16:34 EDT 2003 athena * [project @ 2003-05-18 11:16:34 by athena] Faster inner loop. M ./dft/dftw-dft.c -13 +25 Sat May 17 08:02:38 EDT 2003 athena * [project @ 2003-05-17 12:02:38 by athena] Print vector length M ./dft/dftw-dft.c -2 +2 Sat May 17 07:55:33 EDT 2003 athena * [project @ 2003-05-17 11:55:33 by athena] Oops M ./dft/dftw-dft.c -7 +7 Sat May 17 07:50:35 EDT 2003 athena * [project @ 2003-05-17 11:50:35 by athena] Allow vl > 1 M ./dft/dftw-dft.c -7 +16 Sat May 17 07:01:42 EDT 2003 athena * [project @ 2003-05-17 11:01:42 by athena] Radix can be derived from problem---no need to pre-specify it. M ./dft/ctsq.c -21 +12 Fri May 16 22:50:50 EDT 2003 stevenj * [project @ 2003-05-17 02:50:50 by stevenj] fixed comment M ./kernel/transpose.c -1 +1 Fri May 16 22:48:20 EDT 2003 stevenj * [project @ 2003-05-17 02:48:20 by stevenj] whoops, gcd should be static M ./kernel/transpose.c -1 +1 Fri May 16 22:40:32 EDT 2003 stevenj * [project @ 2003-05-17 02:40:32 by stevenj] more unrolling M ./kernel/transpose.c -7 +31 Fri May 16 20:03:48 EDT 2003 athena * [project @ 2003-05-17 00:03:48 by athena] Hack to avoid infinite recursion. M ./dft/bluestein.c +3 Fri May 16 19:52:43 EDT 2003 stevenj * [project @ 2003-05-16 23:52:43 by stevenj] consistency M ./dft/codelet-dft.h -4 +4 Fri May 16 19:45:15 EDT 2003 athena * [project @ 2003-05-16 23:45:15 by athena] Wrong comment. M ./dft/bluestein.c -1 +1 Fri May 16 19:45:03 EDT 2003 athena * [project @ 2003-05-16 23:45:03 by athena] Style. M ./dft/bluestein.c -3 +3 Fri May 16 18:35:27 EDT 2003 stevenj * [project @ 2003-05-16 22:35:27 by stevenj] punctuation M ./doc/FAQ/fftw-faq.bfnn -2 +2 Fri May 16 18:33:45 EDT 2003 stevenj * [project @ 2003-05-16 22:33:45 by stevenj] added allzero FAQ M ./doc/FAQ/fftw-faq.bfnn +7 Fri May 16 18:22:45 EDT 2003 stevenj * [project @ 2003-05-16 22:22:45 by stevenj] simplification: instead of cldb, just use cldf with inputs/output values swapped M ./dft/bluestein.c -22 +11 Fri May 16 15:47:17 EDT 2003 athena * [project @ 2003-05-16 19:47:17 by athena] Allow more general transform sizes. M ./dft/bluestein.c -15 +21 Fri May 16 14:22:37 EDT 2003 stevenj * [project @ 2003-05-16 18:22:37 by stevenj] slight change M ./kernel/ifftw.h -4 +4 Fri May 16 14:22:05 EDT 2003 stevenj * [project @ 2003-05-16 18:22:05 by stevenj] MS has __int64 type, not long long (grr) M ./kernel/ifftw.h -1 +4 Fri May 16 13:34:16 EDT 2003 athena * [project @ 2003-05-16 17:34:16 by athena] Fixed printout M ./dft/ct.c -2 +4 Fri May 16 13:23:00 EDT 2003 athena * [project @ 2003-05-16 17:23:00 by athena] Fixed flop count M ./dft/bluestein.c +5 Fri May 16 13:02:06 EDT 2003 athena * [project @ 2003-05-16 17:02:06 by athena] New bluestein solver A ./dft/bluestein.c M ./dft/Makefile.am -5 +6 M ./dft/bluestein.c +255 M ./dft/conf.c -1 +2 M ./dft/dft.h -1 +2 Fri May 16 09:51:05 EDT 2003 athena * [project @ 2003-05-16 13:51:05 by athena] Implemented generic radix. M ./dft/ct.c -14 +35 Fri May 16 08:19:38 EDT 2003 athena * [project @ 2003-05-16 12:19:38 by athena] Removed conditional branch from inner loop in generic.c M ./dft/generic.c -18 +20 M ./kernel/ifftw.h -2 +2 M ./kernel/twiddle.c -1 +18 Fri May 16 07:48:28 EDT 2003 athena * [project @ 2003-05-16 11:48:28 by athena] Simplified indexing M ./dft/generic.c -29 +20 Fri May 16 06:53:56 EDT 2003 athena * [project @ 2003-05-16 10:53:56 by athena] Better still. M ./dft/generic.c -26 +37 Fri May 16 06:24:31 EDT 2003 athena * [project @ 2003-05-16 10:24:31 by athena] Further improvement of generic solver M ./dft/generic.c -8 +8 Fri May 16 05:57:07 EDT 2003 athena * [project @ 2003-05-16 09:57:07 by athena] Cleanup M ./dft/rader.c -54 +51 Fri May 16 05:42:57 EDT 2003 athena * [project @ 2003-05-16 09:42:57 by athena] Cleanup M ./dft/dftw-dft.c -30 +25 Fri May 16 05:31:40 EDT 2003 athena * [project @ 2003-05-16 09:31:40 by athena] Generic now only works for odd sized. Added check. M ./dft/generic.c -1 +2 Thu May 15 21:53:25 EDT 2003 athena * [project @ 2003-05-16 01:53:25 by athena] Increased GENERIC_MIN_BAD because of new algorithm. M ./kernel/ifftw.h -2 +2 Thu May 15 21:40:27 EDT 2003 athena * [project @ 2003-05-16 01:40:27 by athena] Much, much better. M ./dft/generic.c -9 +30 Thu May 15 21:25:00 EDT 2003 athena * [project @ 2003-05-16 01:25:00 by athena] Still trying to understand why rdft-generic-dit is faster then dft-generic... M ./dft/generic.c -9 +8 Thu May 15 21:04:33 EDT 2003 athena * [project @ 2003-05-16 01:04:33 by athena] Nothing, really M ./dft/generic.c -2 +2 Thu May 15 20:59:45 EDT 2003 athena * [project @ 2003-05-16 00:59:45 by athena] Never be clever for the sake of being clever. M ./dft/generic.c -1 +1 Thu May 15 20:58:06 EDT 2003 athena * [project @ 2003-05-16 00:58:06 by athena] Simplified. generic-dit is gone. The solver is now out-of-place only---buffering is done by the buffered solver. M ./dft/generic.c -94 +45 Thu May 15 19:18:18 EDT 2003 athena * [project @ 2003-05-15 23:18:18 by athena] rader-dit is gone. M ./dft/rader.c -200 Thu May 15 19:13:03 EDT 2003 athena * [project @ 2003-05-15 23:13:03 by athena] Cast M ./dft/plan.c -2 +2 Thu May 15 19:09:07 EDT 2003 athena * [project @ 2003-05-15 23:09:07 by athena] Introduced twiddle problem ``dftw''. Changed most other things to deal with this change. A ./dft/ctsq.c A ./dft/dftw-dft.c A ./dft/directw.c A ./dft/directwbuf.c A ./dft/directwsq.c A ./dft/problemw.c M ./configure.ac -1 +1 M ./dft/Makefile.am -5 +5 M ./dft/buffered.c -2 +1 M ./dft/codelet-dft.h -13 +9 M ./dft/conf.c -1 +4 R ./dft/ct-dif.c R ./dft/ct-dit.c R ./dft/ct-ditbuf.c R ./dft/ct-ditf.c M ./dft/ct.c -91 +157 R ./dft/ct.h M ./dft/ctsq.c +213 M ./dft/dft.h -7 +45 M ./dft/dftw-dft.c +297 M ./dft/directw.c +172 M ./dft/directwbuf.c +230 M ./dft/directwsq.c +148 M ./dft/kdft-dif.c -4 +5 M ./dft/kdft-difsq.c -3 +3 M ./dft/kdft-dit.c -5 +5 M ./dft/plan.c -1 +11 M ./dft/problemw.c +103 M ./dft/rader.c -1 +1 M ./dft/solve.c -1 +8 Sat Aug 12 23:17:14 EDT 2006 Unknown tagger tagged fftw-3-0-1-branch Thu May 15 18:47:18 EDT 2003 stevenj * [project @ 2003-05-15 22:47:18 by stevenj] whoops, X(safe_mulmod) not fftw_safe_mulmod M ./kernel/primes.c -3 +3 Thu May 15 16:53:16 EDT 2003 stevenj * [project @ 2003-05-15 20:53:16 by stevenj] add VC++ versions of asm M ./simd/sse.c -1 +17 M ./simd/sse2.c -1 +17 Thu May 15 15:03:06 EDT 2003 stevenj * [project @ 2003-05-15 19:03:05 by stevenj] VC++ reportedly supports the intel intrinsics, but requires __inline instead of __inline__ M ./simd/simd-sse.h -1 +4 M ./simd/simd-sse2.h -1 +5 Thu May 15 14:32:06 EDT 2003 stevenj * [project @ 2003-05-15 18:32:06 by stevenj] precompute array indices with VC++ M ./kernel/ifftw.h -2 +2 Wed May 14 21:57:39 EDT 2003 stevenj * [project @ 2003-05-15 01:57:39 by stevenj] added doc note M ./acx_pthread.m4 -1 +5 Wed May 14 19:45:54 EDT 2003 stevenj * [project @ 2003-05-14 23:45:54 by stevenj] autodetect windows M ./threads/threads.c -3 +4 Wed May 14 15:08:49 EDT 2003 stevenj * [project @ 2003-05-14 19:08:49 by stevenj] don't bother with #ifdef HAVE_CONFIG_H, since non-Unix users always forget to define it M ./libbench2/getopt.c -2 Tue May 13 16:58:07 EDT 2003 stevenj * [project @ 2003-05-13 20:58:07 by stevenj] VC++ uses __inline M ./kernel/cycle.h -2 +2 Tue May 13 14:51:26 EDT 2003 stevenj * [project @ 2003-05-13 18:51:26 by stevenj] added leak question M ./doc/FAQ/fftw-faq.bfnn +8 Mon May 12 18:26:51 EDT 2003 stevenj * [project @ 2003-05-12 22:26:51 by stevenj] LARGE_INTEGER needs windows.h (supposedly, there is some problem converting _itnt64 to double...damn MS and their nonstandard types) M ./kernel/cycle.h -1 +2 Mon May 12 18:22:16 EDT 2003 stevenj * [project @ 2003-05-12 22:22:16 by stevenj] whoops M ./libbench2/timer.c -2 +2 Mon May 12 17:16:19 EDT 2003 stevenj * [project @ 2003-05-12 21:16:19 by stevenj] added 256x256 to canonical list M ./tools/fftw-wisdom.c -1 +1 Mon May 12 07:02:06 EDT 2003 athena * [project @ 2003-05-12 11:02:06 by athena] Oops... M ./kernel/transpose.c +1 Sun May 11 11:04:46 EDT 2003 athena * [project @ 2003-05-11 15:04:46 by athena] Unrolled loops, changed cutoff M ./kernel/transpose.c -23 +80 Sun May 11 10:20:04 EDT 2003 athena * [project @ 2003-05-11 14:20:04 by athena] Do not multiply strides by 2 twice. M ./tests/bench.c -16 +15 Wed May 7 21:09:43 EDT 2003 stevenj * [project @ 2003-05-08 01:09:43 by stevenj] added 'make smallcheck' M ./tests/Makefile.am +13 Wed May 7 20:46:10 EDT 2003 stevenj * [project @ 2003-05-08 00:46:10 by stevenj] --without-cycle-counter becomes --with-slow-timer, updated docs M ./configure.ac -3 +3 M ./doc/fftw3.texi -15 +15 M ./kernel/timer.c -3 +3 Wed May 7 18:05:29 EDT 2003 stevenj * [project @ 2003-05-07 22:05:29 by stevenj] remove duplicate -openmp check; Sun requires -xopenmp M ./configure.ac -5 +5 Wed May 7 17:59:23 EDT 2003 stevenj * [project @ 2003-05-07 21:59:23 by stevenj] fixed compilation under Sun C++ M ./dft/ct-ditbuf.c -3 +3 M ./rdft/hc2hc-buf.c -3 +3 Wed May 7 14:24:46 EDT 2003 athena * [project @ 2003-05-07 18:24:46 by athena] Use estimator if cycle counter is unavailable, regardless of the FFTW_MEASURE/ESTIMATE setting. M ./kernel/planner.c -3 +13 M ./kernel/timer.c -57 +67 Tue May 6 23:15:34 EDT 2003 stevenj * [project @ 2003-05-07 03:15:34 by stevenj] _WIN32 (not __WIN32__) is always defined M ./kernel/cycle.h -3 +2 Tue May 6 23:11:52 EDT 2003 stevenj * [project @ 2003-05-07 03:11:52 by stevenj] minor cleanup M ./kernel/cycle.h -2 +2 Tue May 6 22:50:07 EDT 2003 stevenj * [project @ 2003-05-07 02:50:07 by stevenj] tentative VC++ stuff, some consolidation M ./kernel/cycle.h -51 +56 Tue May 6 12:17:56 EDT 2003 stevenj * [project @ 2003-05-06 16:17:56 by stevenj] made cycle.h more self-contained M ./kernel/cycle.h -3 +13 M ./kernel/timer.c -15 +18 Tue May 6 08:30:39 EDT 2003 athena * [project @ 2003-05-06 12:30:39 by athena] Use ``%'' flag to denote commutative operations. M ./simd/simd-3dnow.h -3 +3 M ./simd/simd-sse.h -3 +3 M ./simd/simd-sse2.h -3 +3 Mon May 5 20:42:30 EDT 2003 stevenj * [project @ 2003-05-06 00:42:30 by stevenj] MIT license, brief documentation M ./kernel/cycle.h -12 +41 Mon May 5 20:31:16 EDT 2003 stevenj * [project @ 2003-05-06 00:31:16 by stevenj] whoops, forgot f77_wisdom.f M ./doc/Makefile.am -1 +1 Sun May 4 19:37:09 EDT 2003 athena * [project @ 2003-05-04 23:37:09 by athena] Improved speed of accuracy test. M ./dft/problem.c -6 +3 M ./libbench/mp.c -16 +26 M ./libbench2/bench.h -1 +2 M ./libbench2/mp.c -46 +92 M ./libbench2/verify-lib.c -1 +3 M ./rdft/problem2.c -3 +2 Tue Apr 29 11:45:34 EDT 2003 athena * [project @ 2003-04-29 15:45:34 by athena] s390 cycle counter M ./kernel/cycle.h -1 +21 Sat Apr 26 12:26:15 EDT 2003 stevenj * [project @ 2003-04-26 16:26:15 by stevenj] forgot r2r directory M ./doc/fftw3.texi -10 +10 Fri Apr 25 20:52:23 EDT 2003 stevenj * [project @ 2003-04-26 00:52:23 by stevenj] delete unused files, since they don't compile any more M ./rdft/Makefile.am -2 R ./rdft/vrank2-transpose.c R ./rdft/vrank3-transpose.c Thu Apr 24 06:37:41 EDT 2003 athena * [project @ 2003-04-24 10:37:41 by athena] Better gcc code generation M ./simd/simd-sse2.h -3 +5 Wed Apr 23 15:30:50 EDT 2003 stevenj * [project @ 2003-04-23 19:30:50 by stevenj] ccc is the Compaq C compiler on Linux/alpha M ./acinclude.m4 -1 +1 Wed Apr 23 00:06:03 EDT 2003 stevenj * [project @ 2003-04-23 04:06:03 by stevenj] whoops M ./doc/fftw3.texi -3 +3 Sat Aug 12 23:16:25 EDT 2006 Unknown tagger tagged fftw-3-0 Sat Apr 19 09:18:25 EDT 2003 athena * [project @ 2003-04-19 13:18:25 by athena] ia64 cycle counter with intel compiler. M ./kernel/cycle.h -1 +19 Fri Apr 18 18:27:30 EDT 2003 athena * [project @ 2003-04-18 22:27:30 by athena] More gcc bugs. Sigh. M ./doc/FAQ/fftw-faq.bfnn -4 +12 Fri Apr 18 18:01:49 EDT 2003 athena * [project @ 2003-04-18 22:01:49 by athena] touch ChangeLog to observe GNU standards M ./bootstrap.sh +1 Fri Apr 18 18:01:12 EDT 2003 athena * [project @ 2003-04-18 22:01:12 by athena] We now build ChangeLog automatically at distribution time R ./ChangeLog Fri Apr 18 18:00:17 EDT 2003 athena * [project @ 2003-04-18 22:00:17 by athena] Automatic ChangeLog hackery M ./mkdist.sh +7 Fri Apr 18 13:25:26 EDT 2003 stevenj * [project @ 2003-04-18 17:25:26 by stevenj] plural M ./doc/FAQ/fftw-faq.bfnn -1 +1 Fri Apr 18 13:25:01 EDT 2003 stevenj * [project @ 2003-04-18 17:25:01 by stevenj] updated M ./NEWS +12 Fri Apr 18 12:59:41 EDT 2003 athena * [project @ 2003-04-18 16:59:41 by athena] Updated M ./ChangeLog -6789 +3415 Fri Apr 18 11:48:39 EDT 2003 stevenj * [project @ 2003-04-18 15:48:39 by stevenj] a -> an M ./doc/FAQ/fftw-faq.bfnn -1 +1 Fri Apr 18 11:47:56 EDT 2003 stevenj * [project @ 2003-04-18 15:47:56 by stevenj] hyphen M ./doc/FAQ/fftw-faq.bfnn -1 +1 Fri Apr 18 11:47:28 EDT 2003 stevenj * [project @ 2003-04-18 15:47:28 by stevenj] comma M ./doc/FAQ/fftw-faq.bfnn -1 +1 Fri Apr 18 11:46:59 EDT 2003 stevenj * [project @ 2003-04-18 15:46:59 by stevenj] minor M ./doc/FAQ/fftw-faq.bfnn -4 +4 Fri Apr 18 10:37:31 EDT 2003 athena * [project @ 2003-04-18 14:37:31 by athena] Updated M ./doc/FAQ/fftw-faq.bfnn -9 +36 Fri Apr 18 10:14:59 EDT 2003 athena * [project @ 2003-04-18 14:14:59 by athena] New script that builds the distributions A ./mkdist.sh Fri Apr 18 08:51:07 EDT 2003 athena * [project @ 2003-04-18 12:51:07 by athena] Oops again M ./dft/simd/codelets/Makefile.am -4 +4 Fri Apr 18 08:39:05 EDT 2003 athena * [project @ 2003-04-18 12:39:05 by athena] Oops, forgot -sign 1 M ./dft/simd/codelets/Makefile.am -4 +4 Fri Apr 18 08:28:25 EDT 2003 athena * [project @ 2003-04-18 12:28:25 by athena] Reorganization of simd codelets M ./configure.ac -4 M ./dft/simd/codelets/Makefile.am -136 +22 M ./dft/simd/n1b.c -3 +3 M ./dft/simd/n1b.h -3 M ./dft/simd/n1f.c -3 +3 M ./dft/simd/n1f.h -3 M ./dft/simd/n2b.c -3 +3 M ./dft/simd/n2b.h +3 M ./dft/simd/n2f.c -3 +3 M ./dft/simd/n2f.h +3 Thu Apr 17 21:21:45 EDT 2003 athena * [project @ 2003-04-18 01:21:45 by athena] k7 assembly was not updated after conversion of opcnt from int to double M ./genfft-k7/gen_notw.ml -5 +5 M ./genfft-k7/gen_twiddle.ml -5 +6 Thu Apr 17 19:15:53 EDT 2003 athena * [project @ 2003-04-17 23:15:53 by athena] Capital `X' looks bad in all-lowercase plans M ./dft/vrank2-transpose.c -2 +2 M ./dft/vrank3-transpose.c -2 +2 Thu Apr 17 18:53:29 EDT 2003 athena * [project @ 2003-04-17 22:53:29 by athena] Removed redundant inline/noinline codelets M ./dft/codelets/standard/Makefile.am -3 +2 M ./dft/simd/codelets/Makefile.am -27 +12 M ./rdft/codelets/hc2r/Makefile.am -8 +4 M ./rdft/codelets/r2hc/Makefile.am -8 +4 Thu Apr 17 15:25:50 EDT 2003 athena * [project @ 2003-04-17 19:25:50 by athena] New noinline Noinline real codelets A ./genfft/gen_hc2r_noinline.ml A ./genfft/gen_r2hc_noinline.ml M ./genfft/Makefile.am -10 +24 M ./genfft/gen_hc2hc.ml -5 +5 M ./genfft/gen_hc2r.ml -48 +31 M ./genfft/gen_hc2r_noinline.ml +185 M ./genfft/gen_r2hc.ml -49 +31 M ./genfft/gen_r2hc_noinline.ml +181 M ./genfft/gen_r2r.ml -3 +3 M ./kernel/ifftw.h -4 +1 M ./rdft/codelets/hc2r/Makefile.am -1 +15 M ./rdft/codelets/r2hc/Makefile.am -1 +15 M ./support/Makefile.codelets -1 +3 Thu Apr 17 15:23:03 EDT 2003 stevenj * [project @ 2003-04-17 19:23:03 by stevenj] more ideas M ./TODO +6 Thu Apr 17 13:18:45 EDT 2003 athena * [project @ 2003-04-17 17:18:45 by athena] Removed duplicate rules. M ./dft/simd/codelets/Makefile.am -6 Thu Apr 17 10:51:09 EDT 2003 athena * [project @ 2003-04-17 14:51:09 by athena] acx_pthread.m4 was not distributed M ./Makefile.am -1 +1 Thu Apr 17 07:21:17 EDT 2003 athena * [project @ 2003-04-17 11:21:17 by athena] Oops M ./support/Makefile.codelets -1 +1 Thu Apr 17 07:07:19 EDT 2003 athena * [project @ 2003-04-17 11:07:19 by athena] Both inlined and non-inlined notw codelets. A ./genfft/gen_notw_noinline.ml M ./dft/codelets/standard/Makefile.am -1 +10 M ./genfft/Makefile.am -11 +17 M ./genfft/gen_notw.ml -51 +37 M ./genfft/gen_notw_noinline.ml +164 M ./support/Makefile.codelets +1 Thu Apr 17 06:44:21 EDT 2003 athena * [project @ 2003-04-17 10:44:21 by athena] Initial experiment with both inlined and non-inlined simd codelets. Both are included for now. A ./genfft/gen_notw_noinline_c.ml M ./dft/simd/codelets/Makefile.am -3 +68 M ./genfft/Makefile.am -9 +17 M ./genfft/gen_notw_noinline_c.ml +160 M ./support/Makefile.codelets -1 +2 Thu Apr 17 05:57:36 EDT 2003 athena * [project @ 2003-04-17 09:57:36 by athena] --enable-fma to build FMA distribution M ./configure.ac +7 M ./support/Makefile.codelets -1 +7 Wed Apr 16 17:21:53 EDT 2003 athena * [project @ 2003-04-16 21:21:53 by athena] Inline SIMD nontwiddle codelets M ./genfft/gen_notw_c.ml -47 +41 Wed Apr 16 16:18:29 EDT 2003 athena * [project @ 2003-04-16 20:18:29 by athena] Pathetic attempt at saving a couple of registers... M ./simd/simd-sse.h -6 +9 M ./simd/simd-sse2.h -6 +9 Wed Apr 16 15:51:27 EDT 2003 athena * [project @ 2003-04-16 19:51:27 by athena] for (i = 0; i < m; ++i) ==> for (i = m; i > 0; --i) No proof of evidence that this is any faster, but just in case... M ./genfft/gen_hc2r.ml -3 +3 M ./genfft/gen_notw.ml -3 +3 M ./genfft/gen_notw_c.ml -3 +3 M ./genfft/gen_r2hc.ml -3 +3 M ./genfft/gen_r2r.ml -3 +3 M ./genfft/gen_twiddle.ml -5 +5 M ./genfft/gen_twiddle_c.ml -5 +5 Tue Apr 15 15:03:20 EDT 2003 stevenj * [project @ 2003-04-15 19:03:20 by stevenj] added hack to make sure that codelet loops are preferred to vecloop solvers in the estimator M ./dft/vrank-geq1.c -1 +2 M ./rdft/vrank-geq1-rdft2.c -1 +2 M ./rdft/vrank-geq1.c -1 +2 Tue Apr 15 14:53:44 EDT 2003 stevenj * [project @ 2003-04-15 18:53:44 by stevenj] use double for flops M ./api/f77funcs.h -1 +1 M ./api/fftw3.h -2 +2 M ./api/flops.c -1 +1 M ./doc/fftw3.texi -3 +6 M ./kernel/ifftw.h -5 +5 M ./tests/bench.c -2 +2 Tue Apr 15 14:51:50 EDT 2003 stevenj * [project @ 2003-04-15 18:51:50 by stevenj] metrowerks reportedly supports gcc assembly extensions on ppc M ./kernel/cycle.h -2 +2 Mon Apr 14 15:00:50 EDT 2003 athena * [project @ 2003-04-14 19:00:48 by athena] foo_CFLAGS generates some automake junk that breaks the build on Redhat 7.3. Screw it. M ./dft/simd/Makefile.am -1 +1 M ./simd/Makefile.am -1 +1 Mon Apr 14 12:22:59 EDT 2003 athena * [project @ 2003-04-14 16:22:59 by athena] Carefully check return status M ./tests/check.pl -1 +21 Sun Apr 13 16:46:12 EDT 2003 athena * [project @ 2003-04-13 20:46:12 by athena] Removed annoying -FMA() expressions. M ./genfft/c.ml -2 +6 M ./genfft/simd.ml -2 +2 M ./kernel/ifftw.h -1 +9 M ./support/Makefile.codelets -1 +1 Sat Apr 12 14:32:22 EDT 2003 athena * [project @ 2003-04-12 18:32:22 by athena] Major fma hackery M ./kernel/ifftw.h -5 +17 Sat Apr 12 14:25:43 EDT 2003 athena * [project @ 2003-04-12 18:25:43 by athena] Slight cleanup M ./api/apiplan.c -8 +5 Sat Apr 12 10:04:51 EDT 2003 athena * [project @ 2003-04-12 14:04:51 by athena] Updated version number M ./configure.ac -1 +1 Sat Apr 12 08:03:07 EDT 2003 athena * [project @ 2003-04-12 12:03:07 by athena] Damn autoconf M ./acinclude.m4 -2 +2 Sat Apr 12 07:54:20 EDT 2003 athena * [project @ 2003-04-12 11:54:20 by athena] Recognize all 74xx processors M ./acinclude.m4 -4 +5 Sat Apr 12 07:35:17 EDT 2003 athena * [project @ 2003-04-12 11:35:17 by athena] Detect 7400 processor. M ./acinclude.m4 -3 +7 Fri Apr 11 20:42:11 EDT 2003 athena * [project @ 2003-04-12 00:42:11 by athena] No need to check for gcc-2.95 M ./acinclude.m4 -4 +2 Fri Apr 11 16:14:39 EDT 2003 stevenj * [project @ 2003-04-11 20:14:39 by stevenj] removed duplicate M ./NEWS -3 Fri Apr 11 08:45:37 EDT 2003 athena * [project @ 2003-04-11 12:45:37 by athena] mflops ==> ``mflops'' M ./libbench2/report.c -2 +2 Fri Apr 11 07:00:53 EDT 2003 athena * [project @ 2003-04-11 11:00:53 by athena] Print setup time as well M ./libbench2/report.c -3 +5 Sat Aug 12 23:14:42 EDT 2006 Unknown tagger tagged fftw-3-0-beta3 Thu Apr 10 15:36:18 EDT 2003 athena * [project @ 2003-04-10 19:36:15 by athena] Enforce pointer equality for in-place problems. M ./dft/problem.c -4 +8 M ./kernel/ifftw.h -1 +4 M ./rdft/problem.c -4 +3 M ./rdft/problem2.c -4 +7 M ./simd/taint.c -1 +8 Wed Apr 9 17:47:54 EDT 2003 stevenj * [project @ 2003-04-09 21:47:26 by stevenj] updated M ./ChangeLog +406 M ./NEWS +28 Wed Apr 9 14:53:38 EDT 2003 stevenj * [project @ 2003-04-09 18:53:38 by stevenj] cross-ref fftw-wisdom man page M ./tests/README -1 +1 Wed Apr 9 10:13:00 EDT 2003 athena * [project @ 2003-04-09 14:13:00 by athena] Undone previous change, committed by mistake. M ./kernel/planner.c -2 +1 Wed Apr 9 10:12:24 EDT 2003 athena * [project @ 2003-04-09 14:12:18 by athena] Quick and dirty README for bench A ./tests/README M ./kernel/planner.c -1 +2 M ./tests/Makefile.am -1 +1 M ./tests/README +57 Wed Apr 9 08:50:25 EDT 2003 athena * [project @ 2003-04-09 12:50:25 by athena] Consider additional command-line arguments as problems to be benchmarked. M ./libbench2/bench-main.c -7 +6 M ./libbench2/timer.c -1 +7 Wed Apr 9 08:44:13 EDT 2003 athena * [project @ 2003-04-09 12:44:13 by athena] Default report format is now human-readable. Removed unnecessary complexity in benchmark reporting. M ./libbench2/bench-main.c -18 +7 M ./libbench2/bench.h -5 +2 M ./libbench2/report.c -71 +39 Wed Apr 9 06:10:40 EDT 2003 athena * [project @ 2003-04-09 10:10:40 by athena] Updated for new interleaved/split api. M ./doc/fftw3.texi -100 +180 Wed Apr 9 03:01:03 EDT 2003 stevenj * [project @ 2003-04-09 07:01:03 by stevenj] updated citation M ./doc/fftw3.texi -3 +3 Tue Apr 8 19:35:59 EDT 2003 athena * [project @ 2003-04-08 23:35:59 by athena] Time for beta3 M ./configure.ac -1 +1 Tue Apr 8 17:40:59 EDT 2003 stevenj * [project @ 2003-04-08 21:40:59 by stevenj] whoops, added A ./reodft/redft00e-r2hc-pad.c Tue Apr 8 17:33:47 EDT 2003 stevenj * [project @ 2003-04-08 21:33:47 by stevenj] more comparison of different R*DFT types M ./doc/fftw3.texi -1 +32 Tue Apr 8 16:48:08 EDT 2003 stevenj * [project @ 2003-04-08 20:48:08 by stevenj] comments M ./reodft/redft00e-r2hc.c -5 +18 M ./reodft/rodft00e-r2hc.c -5 +18 Tue Apr 8 16:19:39 EDT 2003 stevenj * [project @ 2003-04-08 20:19:39 by stevenj] more accurate DCT-I and DST-I, at the expense of up to a factor of 2 in speed and memory A ./reodft/rodft00e-r2hc-pad.c M ./reodft/Makefile.am -4 +5 M ./reodft/conf.c -3 +5 M ./reodft/reodft.h +2 M ./reodft/rodft00e-r2hc-pad.c +200 Tue Apr 8 05:38:09 EDT 2003 athena * [project @ 2003-04-08 09:38:09 by athena] Workaround gcc/sparc bug M ./kernel/planner.c -2 +3 Tue Apr 8 01:34:12 EDT 2003 stevenj * [project @ 2003-04-08 05:34:12 by stevenj] rumors M ./doc/fftw3.texi -2 +2 Mon Apr 7 18:54:11 EDT 2003 stevenj * [project @ 2003-04-07 22:54:11 by stevenj] added rdft2 paranoid mode M ./tests/hook.c +33 Mon Apr 7 18:47:37 EDT 2003 stevenj * [project @ 2003-04-07 22:47:37 by stevenj] added paranoid mode for r2r M ./tests/hook.c +51 Mon Apr 7 15:10:08 EDT 2003 stevenj * [project @ 2003-04-07 19:10:08 by stevenj] whoops, sincos is predefined on some systems M ./libbench2/verify-r2r.c -3 +3 Sat Apr 5 16:50:57 EST 2003 athena * [project @ 2003-04-05 21:50:57 by athena] bp->destroy_input was not initialized M ./tests/hook.c +1 Sat Apr 5 09:29:11 EST 2003 athena * [project @ 2003-04-05 14:29:11 by athena] Asserted correctness conditions for tainted pointers. (For now, use CK() while we test. They should be changed into A() at some point.) M ./dft/problem.c -7 +16 M ./kernel/ifftw.h -1 +3 M ./rdft/problem.c -3 +7 M ./rdft/problem2.c -6 +12 Sat Apr 5 08:18:23 EST 2003 athena * [project @ 2003-04-05 13:18:23 by athena] Untaint pointers before zero'ing arrays and before hashing M ./dft/problem.c -8 +10 M ./rdft/problem.c -4 +4 M ./rdft/problem2.c -9 +9 Sat Apr 5 07:11:56 EST 2003 athena * [project @ 2003-04-05 12:11:56 by athena] Alignment check did not work with icc, which seems to be confused by the fact that the variable is not used. M ./libbench2/bench-main.c -5 +4 Sat Apr 5 06:41:20 EST 2003 athena * [project @ 2003-04-05 11:41:20 by athena] More paranoid paranoid-check M ./tests/Makefile.am -1 +7 Sat Apr 5 06:19:25 EST 2003 athena * [project @ 2003-04-05 11:19:25 by athena] 0 == x & 7 parses as (0 == x) & 7, which is wrong M ./kernel/ifftw.h -2 +2 Fri Apr 4 21:35:49 EST 2003 stevenj * [project @ 2003-04-05 02:35:49 by stevenj] alignment checks M ./dft/direct.c -1 +2 M ./kernel/ifftw.h -1 +12 M ./kernel/planner.c -1 +2 M ./libbench2/bench-main.c -1 +3 M ./rdft/direct.c -1 +4 M ./rdft/direct2.c -1 +3 Fri Apr 4 21:04:14 EST 2003 stevenj * [project @ 2003-04-05 02:04:14 by stevenj] prevent infinite loops in exhaustive planning M ./rdft/rdft-dht.c -1 +6 Fri Apr 4 20:58:20 EST 2003 stevenj * [project @ 2003-04-05 01:58:20 by stevenj] split/unsplit guru interface A ./api/execute-split-dft-c2r.c A ./api/execute-split-dft-r2c.c A ./api/execute-split-dft.c A ./api/plan-guru-split-dft-c2r.c A ./api/plan-guru-split-dft-r2c.c A ./api/plan-guru-split-dft.c M ./api/Makefile.am -1 +3 M ./api/api.h -2 +3 M ./api/apiplan.c -3 +4 M ./api/execute-dft-c2r.c -2 +2 M ./api/execute-dft-r2c.c -2 +2 M ./api/execute-dft.c -3 +6 M ./api/execute-split-dft-c2r.c +29 M ./api/execute-split-dft-r2c.c +29 M ./api/execute-split-dft.c +29 M ./api/f77funcs.h -8 +73 M ./api/fftw3.h -5 +25 M ./api/mktensor-iodims.c -3 +3 M ./api/plan-guru-dft-c2r.c -4 +9 M ./api/plan-guru-dft-r2c.c -4 +9 M ./api/plan-guru-dft.c -4 +10 M ./api/plan-guru-r2r.c -3 +4 M ./api/plan-guru-split-dft-c2r.c +40 M ./api/plan-guru-split-dft-r2c.c +39 M ./api/plan-guru-split-dft.c +39 M ./api/plan-many-dft-c2r.c -1 +1 M ./api/plan-many-dft-r2c.c -1 +1 M ./api/plan-many-dft.c -1 +1 M ./api/plan-many-r2r.c -1 +1 M ./tests/bench.c -29 +15 Fri Apr 4 20:39:55 EST 2003 athena * [project @ 2003-04-05 01:39:55 by athena] Need UNTAINT in verifier too. M ./tests/hook.c -4 +4 Fri Apr 4 19:36:46 EST 2003 athena * [project @ 2003-04-05 00:36:46 by athena] Forgot #if HAVE_SIMD M ./simd/taint.c -1 +5 Fri Apr 4 19:30:37 EST 2003 athena * [project @ 2003-04-05 00:30:37 by athena] Keep track of two separate taint bits A ./simd/taint.c M ./api/fftw3.h -2 +2 M ./kernel/align.c -15 +1 M ./kernel/ifftw.h -2 +2 M ./simd/Makefile.am -1 +1 M ./simd/simd.h -2 +16 M ./simd/taint.c +33 Fri Apr 4 19:16:32 EST 2003 stevenj * [project @ 2003-04-05 00:16:32 by stevenj] added NO_SIMD problem flag, made UNALIGNED an API issue (taints input pointers) M ./api/api.h +2 M ./api/fftw3.h -1 +3 M ./api/mapflags.c -1 +1 M ./api/plan-guru-dft-c2r.c -1 +3 M ./api/plan-guru-dft-r2c.c -1 +3 M ./api/plan-guru-dft.c -1 +4 M ./api/plan-guru-r2r.c -1 +2 M ./api/plan-many-dft-c2r.c -1 +3 M ./api/plan-many-dft-r2c.c -1 +3 M ./api/plan-many-dft.c -1 +4 M ./api/plan-many-r2r.c -1 +2 M ./dft/k7/k7.c -5 +13 M ./dft/simd/n1b.c -1 +1 M ./dft/simd/n1f.c -1 +1 M ./dft/simd/n2b.c +1 M ./dft/simd/n2f.c +1 M ./dft/simd/q1b.c -1 +1 M ./dft/simd/q1f.c -1 +1 M ./dft/simd/t1b.c -1 +1 M ./dft/simd/t1f.c -1 +1 M ./kernel/ifftw.h -3 +3 M ./tests/bench.c +1 Fri Apr 4 18:14:14 EST 2003 stevenj * [project @ 2003-04-04 23:14:14 by stevenj] bugfix in buffered: wrong pointers passed for cldrest; also use TAINT instead of UNALIGNED in buffered2 M ./dft/buffered.c -6 +10 M ./rdft/buffered.c -6 +10 M ./rdft/buffered2.c -7 +5 Fri Apr 4 17:19:51 EST 2003 athena * [project @ 2003-04-04 22:19:51 by athena] Reverted previous change, committed accidentally M ./dft/vrank-geq1.c -3 +3 Fri Apr 4 17:18:39 EST 2003 athena * [project @ 2003-04-04 22:18:39 by athena] What was I thinking? M ./kernel/align.c -1 +2 Fri Apr 4 17:18:21 EST 2003 athena * [project @ 2003-04-04 22:18:21 by athena] M ./dft/vrank-geq1.c -3 +3 Fri Apr 4 16:48:32 EST 2003 stevenj * [project @ 2003-04-04 21:48:32 by stevenj] added --enable-debug-alignment M ./configure.ac +5 M ./libbench2/aligned-main.c -1 +7 Fri Apr 4 16:29:43 EST 2003 stevenj * [project @ 2003-04-04 21:29:43 by stevenj] X(taint) prototype, define corresponding function only if HAVE_SIMD M ./kernel/align.c -1 +3 M ./kernel/ifftw.h -1 +2 Fri Apr 4 16:15:53 EST 2003 athena * [project @ 2003-04-04 21:15:53 by athena] Initial checkin of tained pointers M ./dft/buffered.c -4 +9 M ./dft/solve.c -2 +4 M ./dft/vrank-geq1.c -3 +3 M ./kernel/align.c -6 +4 M ./kernel/ifftw.h -2 +9 M ./rdft/buffered.c -4 +4 M ./rdft/buffered2.c -3 +2 M ./rdft/solve.c -2 +2 M ./rdft/solve2.c -2 +2 M ./rdft/vrank-geq1-rdft2.c -3 +4 M ./rdft/vrank-geq1.c -3 +3 Fri Apr 4 13:12:58 EST 2003 athena * [project @ 2003-04-04 18:12:52 by athena] More conservative preservation of alignment M ./dft/buffered.c -1 +2 M ./dft/rader.c -4 +3 M ./dft/simd/n2b.c -1 M ./dft/simd/n2f.c -1 M ./dft/vrank-geq1.c -6 +4 M ./kernel/align.c -35 +8 M ./kernel/ifftw.h -3 +2 M ./rdft/buffered.c -1 +2 M ./rdft/buffered2.c -1 +4 M ./rdft/dht-rader.c -2 +2 M ./rdft/vrank-geq1-rdft2.c -5 +3 M ./rdft/vrank-geq1.c -5 +5 Thu Apr 3 23:16:27 EST 2003 stevenj * [project @ 2003-04-04 04:16:27 by stevenj] plan/execute with aligned stack M ./api/apiplan.c -19 +24 M ./api/execute-dft-c2r.c -2 +2 M ./api/execute-dft-r2c.c -2 +2 M ./api/execute-dft.c -2 +2 M ./api/execute-r2r.c -2 +2 M ./api/execute.c -2 +2 M ./api/f77funcs.h -10 +10 Thu Apr 3 15:40:01 EST 2003 stevenj * [project @ 2003-04-03 20:40:01 by stevenj] whoops, missed FFTW_MEASURE in fftw3.f M ./api/Makefile.am -1 +1 Thu Apr 3 13:44:46 EST 2003 stevenj * [project @ 2003-04-03 18:44:46 by stevenj] use WITH_ALIGNED_STACK for experimental semaphore stuff, too M ./threads/threads.c -8 +3 Thu Apr 3 09:04:23 EST 2003 athena * [project @ 2003-04-03 14:04:23 by athena] Removed old file R ./kernel/stack.c Thu Apr 3 07:50:43 EST 2003 athena * [project @ 2003-04-03 12:50:43 by athena] Improved stack-alignment hack M ./kernel/Makefile.am -3 +3 M ./kernel/ifftw.h -4 +52 M ./threads/ct-dit.c -7 +6 M ./threads/dft-vrank-geq1.c -7 +6 M ./threads/hc2hc-dif.c -7 +6 M ./threads/hc2hc-dit.c -7 +6 M ./threads/rdft-vrank-geq1.c -6 +4 M ./threads/threads.h -5 M ./threads/vrank-geq1-rdft2.c -7 +6 Thu Apr 3 02:37:57 EST 2003 stevenj * [project @ 2003-04-03 07:37:57 by stevenj] use aligned stack for experimental semaphores, too M ./threads/threads.c -2 +8 Thu Apr 3 02:17:58 EST 2003 stevenj * [project @ 2003-04-03 07:17:58 by stevenj] whoops M ./kernel/ifftw.h -2 +4 M ./kernel/stack.c -2 +2 M ./threads/ct-dit.c -2 +4 M ./threads/dft-vrank-geq1.c -2 +4 M ./threads/hc2hc-dif.c -2 +4 M ./threads/hc2hc-dit.c -2 +4 M ./threads/rdft-vrank-geq1.c -2 +4 M ./threads/threads.c -24 +3 M ./threads/threads.h +5 M ./threads/vrank-geq1-rdft2.c -2 +4 Thu Apr 3 01:58:32 EST 2003 stevenj * [project @ 2003-04-03 06:58:32 by stevenj] fix(?) for SIMD thread problems M ./kernel/ifftw.h -2 +2 M ./kernel/stack.c -3 +3 M ./threads/ct-dit.c -7 +10 M ./threads/hc2hc-dif.c -2 +2 M ./threads/hc2hc-dit.c -2 +2 M ./threads/threads.c -6 +27 Wed Apr 2 20:33:12 EST 2003 stevenj * [project @ 2003-04-03 01:33:12 by stevenj] noted n=1 REDFT01 case M ./doc/fftw3.texi -1 +11 Wed Apr 2 20:32:07 EST 2003 stevenj * [project @ 2003-04-03 01:32:07 by stevenj] note about n=2 REDFT00 formula M ./doc/fftw3.texi -2 +3 Wed Apr 2 20:30:10 EST 2003 stevenj * [project @ 2003-04-03 01:30:10 by stevenj] note about undefined REDFT00 M ./doc/fftw3.texi -1 +2 Wed Apr 2 20:18:03 EST 2003 stevenj * [project @ 2003-04-03 01:18:03 by stevenj] noted n=1 RODFT01 case M ./doc/fftw3.texi -2 +12 Wed Apr 2 20:14:07 EST 2003 stevenj * [project @ 2003-04-03 01:14:07 by stevenj] corrected definitions M ./doc/equation-redft11.png M ./doc/equation-rodft01.png M ./doc/equation-rodft11.png M ./doc/fftw3.texi -6 +6 Wed Apr 2 19:43:59 EST 2003 stevenj * [project @ 2003-04-03 00:43:59 by stevenj] added REODFT_KINDP, fixed nontrivial test for R2HC11 and HC2R11 (not that we support these yet anyway) M ./rdft/codelet-rdft.h +1 M ./rdft/problem.c -3 +3 M ./rdft/vrank-geq1.c -3 +3 Wed Apr 2 19:16:54 EST 2003 stevenj * [project @ 2003-04-03 00:16:53 by stevenj] size 2 hc2r and dht are equivalent to r2hc M ./rdft/codelets/hc2r/Makefile.am -3 +5 M ./rdft/problem.c -3 +5 Wed Apr 2 15:09:08 EST 2003 stevenj * [project @ 2003-04-02 20:09:08 by stevenj] noted overwriting in upgrading section M ./doc/fftw3.texi -1 +4 Wed Apr 2 05:25:56 EST 2003 athena * [project @ 2003-04-02 10:25:56 by athena] Moved with_aligned_stack to its own file A ./kernel/stack.c M ./kernel/Makefile.am -3 +3 M ./kernel/align.c -46 +1 M ./kernel/stack.c +67 Tue Apr 1 21:11:31 EST 2003 athena * [project @ 2003-04-02 02:11:31 by athena] Fixed comments M ./kernel/align.c -6 +2 M ./libbench2/aligned-main.c -6 +2 Tue Apr 1 20:57:39 EST 2003 athena * [project @ 2003-04-02 01:57:39 by athena] Alignment hacks M ./kernel/align.c -1 +50 M ./kernel/ifftw.h -2 +3 M ./libbench2/aligned-main.c -26 +14 M ./libbench2/bench-main.c -1 +9 Tue Apr 1 14:26:48 EST 2003 stevenj * [project @ 2003-04-01 19:26:48 by stevenj] phew, no, previous version was okay M ./threads/threads.c -1 +1 Tue Apr 1 14:26:15 EST 2003 stevenj * [project @ 2003-04-01 19:26:15 by stevenj] whoops, crap M ./threads/threads.c -2 +2 Tue Apr 1 08:01:06 EST 2003 athena * [project @ 2003-04-01 13:01:06 by athena] support sse2 in forthcoming gcc-3.3 M ./simd/simd-sse2.h -3 +11 Tue Apr 1 01:17:15 EST 2003 stevenj * [project @ 2003-04-01 06:17:15 by stevenj] comment M ./kernel/cycle.h -2 +2 Tue Apr 1 01:16:46 EST 2003 stevenj * [project @ 2003-04-01 06:16:46 by stevenj] noted ac_check_headers M ./kernel/cycle.h -1 +4 Tue Apr 1 01:11:31 EST 2003 stevenj * [project @ 2003-04-01 06:11:31 by stevenj] comment M ./kernel/cycle.h -1 +2 Tue Apr 1 01:06:53 EST 2003 stevenj * [project @ 2003-04-01 06:06:53 by stevenj] documented autoconf tests, so that cycle.h can be distributed separately M ./kernel/cycle.h -1 +27 Sat Aug 12 23:11:17 EDT 2006 Unknown tagger tagged fftw-3-0-beta2 Mon Mar 31 22:12:02 EST 2003 stevenj * [project @ 2003-04-01 03:12:02 by stevenj] IRIX is all-caps M ./NEWS -1 +1 Mon Mar 31 22:11:42 EST 2003 stevenj * [project @ 2003-04-01 03:11:42 by stevenj] noted Irix fix M ./NEWS +3 Mon Mar 31 22:10:33 EST 2003 stevenj * [project @ 2003-04-01 03:10:33 by stevenj] whoops M ./threads/api.c -1 +1 M ./threads/threads.h -1 +1 Mon Mar 31 22:04:35 EST 2003 stevenj * [project @ 2003-04-01 03:04:35 by stevenj] use ithreads_init so as not to confuse fftw 2 users M ./threads/threads.c -6 +8 Mon Mar 31 22:00:42 EST 2003 stevenj * [project @ 2003-04-01 03:00:42 by stevenj] IRIX lossage M ./threads/threads.c -1 +5 Mon Mar 31 21:19:20 EST 2003 stevenj * [project @ 2003-04-01 02:19:20 by stevenj] check for -openmp (icc) among the OpenMP flags (TODO: make this a separate macro, with a loop instead of repeated checks) M ./configure.ac +5 Mon Mar 31 17:12:19 EST 2003 stevenj * [project @ 2003-03-31 22:12:19 by stevenj] clarification M ./doc/fftw3.texi -1 +4 Mon Mar 31 17:01:16 EST 2003 athena * [project @ 2003-03-31 22:01:16 by athena] More liberal test for solaris CC M ./acinclude.m4 -1 +1 Mon Mar 31 15:13:33 EST 2003 athena * [project @ 2003-03-31 20:13:33 by athena] Allow x86-64 simd M ./simd/simd-sse.h -1 +1 M ./simd/simd-sse2.h -1 +1 Mon Mar 31 15:13:21 EST 2003 athena * [project @ 2003-03-31 20:13:21 by athena] Added x86-64 timer code M ./kernel/cycle.h -1 +22 Mon Mar 31 13:10:54 EST 2003 stevenj * [project @ 2003-03-31 18:10:54 by stevenj] updated M ./NEWS +4 Mon Mar 31 13:07:19 EST 2003 stevenj * [project @ 2003-03-31 18:07:19 by stevenj] updated M ./ChangeLog -1378 +3497 Mon Mar 31 13:05:27 EST 2003 stevenj * [project @ 2003-03-31 18:05:27 by stevenj] colon M ./doc/FAQ/fftw-faq.bfnn -1 +1 Mon Mar 31 07:20:20 EST 2003 athena * [project @ 2003-03-31 12:20:20 by athena] Reorganized compiler bugs section (which is growing out of control) M ./doc/FAQ/fftw-faq.bfnn -11 +18 Mon Mar 31 07:15:20 EST 2003 athena * [project @ 2003-03-31 12:15:20 by athena] solaris gcc bug appears to be also in 2.95.2 M ./doc/FAQ/fftw-faq.bfnn -1 +1 Mon Mar 31 07:13:45 EST 2003 athena * [project @ 2003-03-31 12:13:45 by athena] Workaround works---there is another gcc/sparc bug elsehwere M ./kernel/planner.c -3 +3 Mon Mar 31 07:08:56 EST 2003 athena * [project @ 2003-03-31 12:08:56 by athena] Grrr, workaround does not work. M ./kernel/planner.c -3 +3 Mon Mar 31 07:02:23 EST 2003 athena * [project @ 2003-03-31 12:02:23 by athena] ADDMOD is now function, which seems to avoid gcc bugs. M ./kernel/planner.c -10 +16 Sun Mar 30 16:40:26 EST 2003 athena * [project @ 2003-03-30 21:40:26 by athena] Workaround sparc gcc bug M ./kernel/planner.c -3 +5 Sun Mar 30 15:51:59 EST 2003 stevenj * [project @ 2003-03-30 20:50:59 by stevenj] note M ./doc/fftw3.texi -4 +5 Sun Mar 30 15:34:57 EST 2003 stevenj * [project @ 2003-03-30 20:34:57 by stevenj] make non-square UGLY, for now M ./dft/vrank2-transpose.c -4 +4 M ./dft/vrank3-transpose.c -1 +4 Sun Mar 30 15:33:57 EST 2003 stevenj * [project @ 2003-03-30 20:33:57 by stevenj] added -o amnesia to forget_wisdom before each plan M ./tests/bench.c +5 Sun Mar 30 09:41:27 EST 2003 athena * [project @ 2003-03-30 14:41:27 by athena] Report setup time in benchmark M ./libbench2/bench-user.h -1 +4 M ./libbench2/report.c -2 +2 M ./libbench2/speed.c -1 +4 Sat Mar 29 20:21:15 EST 2003 stevenj * [project @ 2003-03-30 01:20:52 by stevenj] comment M ./kernel/transpose.c -1 +2 Sat Mar 29 19:11:10 EST 2003 stevenj * [project @ 2003-03-30 00:11:10 by stevenj] slight change M ./doc/fftw3.texi -4 +4 Sat Mar 29 18:46:16 EST 2003 athena * [project @ 2003-03-29 23:46:16 by athena] More relaxed definition of UGLYness M ./kernel/ct.c -1 +1 Sat Mar 29 15:28:01 EST 2003 stevenj * [project @ 2003-03-29 20:28:01 by stevenj] no more cvs id strings in header files...I'm tired of having to rebuild everything after a commit M ./rdft/codelet-rdft.h -2 M ./rdft/hc2hc.h -2 M ./rdft/rdft.h -2 M ./reodft/reodft.h -2 M ./threads/threads.h -2 Sat Mar 29 15:22:28 EST 2003 stevenj * [project @ 2003-03-29 20:22:28 by stevenj] rdft2 stride unification A ./rdft/rdft2-strides.c M ./rdft/Makefile.am -1 +1 M ./rdft/buffered2.c -20 +11 M ./rdft/direct2.c -5 +5 M ./rdft/rdft.h -1 +2 M ./rdft/rdft2-inplace-strides.c -9 +2 M ./rdft/rdft2-strides.c +38 M ./rdft/rdft2-tensor-max-index.c -7 +4 M ./rdft/vrank-geq1-rdft2.c -26 +9 M ./threads/vrank-geq1-rdft2.c -9 +3 Sat Mar 29 14:38:23 EST 2003 stevenj * [project @ 2003-03-29 19:38:23 by stevenj] preserve in-place-ness M ./rdft/vrank-geq1-rdft2.c -3 +3 Sat Mar 29 14:23:31 EST 2003 stevenj * [project @ 2003-03-29 19:23:31 by stevenj] make nowisdom the default M ./tests/Makefile.am -7 +7 M ./tests/bench.c -2 +2 M ./tests/check.pl -3 +3 Sat Mar 29 14:13:18 EST 2003 athena * [project @ 2003-03-29 19:13:18 by athena] --verbose in paranoid-check produces too much output. Make it quiet. M ./tests/Makefile.am -1 +1 Sat Mar 29 13:45:13 EST 2003 stevenj * [project @ 2003-03-29 18:45:13 by stevenj] fixed transpose bugs...need to check ri-ii before deciding whether Ntuple fits M ./dft/vrank2-transpose.c -5 +9 M ./dft/vrank3-transpose.c -7 +13 M ./kernel/ifftw.h -3 +5 M ./kernel/transpose.c -11 +21 Sat Mar 29 08:10:40 EST 2003 athena * [project @ 2003-03-29 13:10:40 by athena] try more 2^k M ./tests/check.pl +2 Sat Mar 29 08:05:41 EST 2003 athena * [project @ 2003-03-29 13:05:41 by athena] MIN_ALIGNMENT was defined after being used, causing crash in sse2. M ./kernel/ifftw.h -4 +4 Sat Mar 29 03:07:34 EST 2003 stevenj * [project @ 2003-03-29 08:07:34 by stevenj] real transposes are currently unused, and are not needed for MPI code either M ./kernel/Makefile.am -2 +2 M ./kernel/ifftw.h -2 +1 R ./kernel/tensor10.c M ./kernel/transpose.c -180 +18 M ./rdft/Makefile.am -2 +3 M ./rdft/conf.c -3 +5 Sat Mar 29 02:58:39 EST 2003 stevenj * [project @ 2003-03-29 07:58:39 by stevenj] added general transpose A ./kernel/transpose.c M ./dft/vrank2-transpose.c -9 +43 M ./dft/vrank3-transpose.c -25 +57 M ./kernel/Makefile.am -2 +2 M ./kernel/ifftw.h -1 +13 M ./kernel/transpose.c +523 Fri Mar 28 22:49:04 EST 2003 stevenj * [project @ 2003-03-29 03:49:04 by stevenj] added transposition option M ./libbench2/problem.c -1 +18 Fri Mar 28 22:09:22 EST 2003 stevenj * [project @ 2003-03-29 03:09:22 by stevenj] yikes, fixed incorrect applicability of transpose plans A ./kernel/tensor10.c M ./dft/vrank2-transpose.c -3 +3 M ./dft/vrank3-transpose.c -2 +2 M ./kernel/Makefile.am -2 +2 M ./kernel/ifftw.h -1 +2 M ./kernel/tensor10.c +31 M ./rdft/vrank2-transpose.c -2 +2 M ./rdft/vrank3-transpose.c -2 +2 Fri Mar 28 22:06:14 EST 2003 stevenj * [project @ 2003-03-29 03:06:14 by stevenj] in the future, we might want to allow sz->rnk == 0, vecsz->rnk arbitrary to be converted to r2hc (the apply function already should work for this case)...disabled for now, though M ./rdft/dft-r2hc.c -7 +21 Fri Mar 28 19:12:08 EST 2003 stevenj * [project @ 2003-03-29 00:12:08 by stevenj] use most_unaligned in rdft2 M ./kernel/align.c -6 +1 M ./kernel/ifftw.h -2 +1 M ./rdft/vrank-geq1-rdft2.c -6 +11 Fri Mar 28 19:11:47 EST 2003 stevenj * [project @ 2003-03-29 00:11:47 by stevenj] slight change M ./tests/Makefile.am -2 +2 Fri Mar 28 19:00:21 EST 2003 stevenj * [project @ 2003-03-29 00:00:21 by stevenj] output message when checks pass M ./tests/Makefile.am +12 Fri Mar 28 17:21:47 EST 2003 stevenj * [project @ 2003-03-28 22:21:47 by stevenj] added ifndef alloca around alloca stuff M ./kernel/ifftw.h -1 +3 Fri Mar 28 13:45:50 EST 2003 athena * [project @ 2003-03-28 18:45:43 by athena] Proper alignment in rader M ./dft/rader.c -2 +4 M ./dft/vrank-geq1.c -4 +3 M ./kernel/align.c -3 +2 M ./kernel/ifftw.h -2 +2 M ./rdft/dht-rader.c -2 +5 Fri Mar 28 12:43:23 EST 2003 stevenj * [project @ 2003-03-28 17:43:23 by stevenj] whitespace M ./kernel/ifftw.h -1 +2 Fri Mar 28 12:41:39 EST 2003 stevenj * [project @ 2003-03-28 17:41:39 by stevenj] whoops, alloca stuff inside HAVE_ALLOCA M ./kernel/ifftw.h -3 +4 Fri Mar 28 12:35:21 EST 2003 stevenj * [project @ 2003-03-28 17:35:21 by stevenj] make check can afford to be a little bigger M ./tests/Makefile.am -2 +2 Fri Mar 28 12:31:32 EST 2003 stevenj * [project @ 2003-03-28 17:31:32 by stevenj] use same alloca macrology as configure script M ./kernel/ifftw.h -28 +45 Fri Mar 28 03:05:15 EST 2003 stevenj * [project @ 2003-03-28 08:05:15 by stevenj] fallback is no longer needed for mingw M ./kernel/ifftw.h -6 +1 Fri Mar 28 02:58:45 EST 2003 stevenj * [project @ 2003-03-28 07:58:45 by stevenj] alloca fallback for gcc M ./kernel/ifftw.h -3 +4 Fri Mar 28 02:49:59 EST 2003 stevenj * [project @ 2003-03-28 07:49:59 by stevenj] _alloca was added for MinGW, but it causes problems there M ./configure.ac -1 +1 Thu Mar 27 22:06:07 EST 2003 stevenj * [project @ 2003-03-28 03:06:07 by stevenj] fixed most_unaligned for split format M ./kernel/align.c -5 +8 Thu Mar 27 19:01:58 EST 2003 stevenj * [project @ 2003-03-28 00:01:58 by stevenj] whoops M ./Makefile.am -1 +1 Thu Mar 27 19:00:20 EST 2003 stevenj * [project @ 2003-03-28 00:00:20 by stevenj] added pkg-config A ./fftw.pc.in M ./Makefile.am -1 +6 M ./configure.ac +1 M ./fftw.pc.in +10 Thu Mar 27 15:59:01 EST 2003 stevenj * [project @ 2003-03-27 20:59:01 by stevenj] fixed asserts M ./dft/vrank-geq1.c -2 +2 M ./rdft/vrank-geq1-rdft2.c -2 +3 M ./rdft/vrank-geq1.c -2 +2 Thu Mar 27 15:49:53 EST 2003 athena * [project @ 2003-03-27 20:49:53 by athena] Do not adjust r/i pointers separately. M ./kernel/align.c -4 +5 Thu Mar 27 15:17:40 EST 2003 athena * [project @ 2003-03-27 20:17:40 by athena] iForgot to add files A ./dft/simd/n2b.h A ./dft/simd/n2f.h M ./dft/simd/n2b.h +25 M ./dft/simd/n2f.h +25 Thu Mar 27 15:10:41 EST 2003 athena * [project @ 2003-03-27 20:10:36 by athena] Specialized n simd codelets for unit vector stride. A ./dft/simd/n2b.c A ./dft/simd/n2f.c M ./configure.ac +4 M ./dft/simd/Makefile.am -2 +2 M ./dft/simd/codelets/Makefile.am -8 +54 M ./dft/simd/n1b.c -3 +3 M ./dft/simd/n1b.h +4 M ./dft/simd/n1f.c -3 +3 M ./dft/simd/n1f.h +4 M ./dft/simd/n2b.c +49 M ./dft/simd/n2f.c +49 Thu Mar 27 08:22:03 EST 2003 athena * [project @ 2003-03-27 13:22:03 by athena] Changed version number to beta2 M ./configure.ac -1 +1 Thu Mar 27 06:37:07 EST 2003 athena * [project @ 2003-03-27 11:37:07 by athena] Changed alignment requirements for n1 simd codelets. Changed mechanism for detecting lack of alignment. M ./api/mapflags.c -1 +1 M ./dft/simd/n1b.c -4 +4 M ./dft/simd/n1b.h -3 M ./dft/simd/n1f.c -4 +4 M ./dft/simd/n1f.h -3 M ./dft/simd/q1b.c -1 +1 M ./dft/simd/q1f.c -1 +1 M ./dft/simd/t1b.c -1 +1 M ./dft/simd/t1f.c -1 +1 M ./dft/vrank-geq1.c -7 +7 M ./kernel/align.c -5 +42 M ./kernel/ifftw.h -3 +5 M ./rdft/vrank-geq1-rdft2.c -4 +7 M ./rdft/vrank-geq1.c -5 +5 Thu Mar 27 04:25:06 EST 2003 athena * [project @ 2003-03-27 09:25:06 by athena] Oops, wrong place for hook M ./tests/bench.c -2 +1 Thu Mar 27 02:37:52 EST 2003 stevenj * [project @ 2003-03-27 07:37:52 by stevenj] added comments to codelet makefiles, to aid people wanting to generate their own code M ./dft/codelets/inplace/Makefile.am -1 +23 M ./dft/codelets/standard/Makefile.am +24 M ./dft/k7/codelets/Makefile.am +25 M ./dft/simd/codelets/Makefile.am -1 +32 M ./rdft/codelets/hc2r/Makefile.am +25 M ./rdft/codelets/r2hc/Makefile.am +25 M ./rdft/codelets/r2r/Makefile.am +18 Thu Mar 27 01:42:27 EST 2003 stevenj * [project @ 2003-03-27 06:42:27 by stevenj] Matteo is also a copyright holder M ./doc/FAQ/fftw-faq.bfnn -1 +1 Thu Mar 27 01:41:08 EST 2003 stevenj * [project @ 2003-03-27 06:41:08 by stevenj] FORTRAN is officially Fortran, these days M ./doc/FAQ/fftw-faq.bfnn -2 +2 Thu Mar 27 01:40:32 EST 2003 stevenj * [project @ 2003-03-27 06:40:32 by stevenj] punctuation M ./doc/FAQ/fftw-faq.bfnn -1 +1 Thu Mar 27 01:40:14 EST 2003 stevenj * [project @ 2003-03-27 06:40:14 by stevenj] don't use "wrapper" M ./doc/FAQ/fftw-faq.bfnn -1 +1 Thu Mar 27 01:37:53 EST 2003 stevenj * [project @ 2003-03-27 06:37:53 by stevenj] plural M ./doc/FAQ/fftw-faq.bfnn -1 +1 Thu Mar 27 01:35:32 EST 2003 stevenj * [project @ 2003-03-27 06:35:32 by stevenj] grammar M ./doc/FAQ/fftw-faq.bfnn -1 +1 Thu Mar 27 01:33:35 EST 2003 stevenj * [project @ 2003-03-27 06:33:35 by stevenj] better phrasing M ./doc/FAQ/fftw-faq.bfnn -1 +1 Wed Mar 26 22:47:58 EST 2003 stevenj * [project @ 2003-03-27 03:47:58 by stevenj] stddef.h should not be needed anymore for this file M ./kernel/align.c -2 +1 Wed Mar 26 22:13:48 EST 2003 stevenj * [project @ 2003-03-27 03:13:48 by stevenj] added comments for Franz mode M ./dft/codelets/standard/Makefile.am -5 +5 Wed Mar 26 22:11:58 EST 2003 stevenj * [project @ 2003-03-27 03:10:11 by stevenj] clarification M ./dft/simd/codelets/Makefile.am -3 +4 Wed Mar 26 22:08:51 EST 2003 stevenj * [project @ 2003-03-27 03:08:51 by stevenj] commented on FRANZ codelets M ./dft/simd/codelets/Makefile.am -5 +5 Wed Mar 26 22:06:45 EST 2003 stevenj * [project @ 2003-03-27 03:06:45 by stevenj] updated M ./NEWS +23 Wed Mar 26 21:51:15 EST 2003 stevenj * [project @ 2003-03-27 02:51:15 by stevenj] disable DIF codelets, since they are never used (apparently) except for some non-power-of-two sizes...improve support for the latter by adding size 3, 5, and 6 q^2 codelets. M ./dft/codelets/inplace/Makefile.am -4 +8 Wed Mar 26 20:07:11 EST 2003 stevenj * [project @ 2003-03-27 01:07:11 by stevenj] DHT has no forward/backward M ./doc/fftw3.texi -3 +3 Wed Mar 26 19:46:12 EST 2003 fftw * [project @ 2003-03-27 00:46:12 by fftw] added hacky way to use an arbitrary flag M ./tests/bench.c +1 Wed Mar 26 19:44:31 EST 2003 athena * [project @ 2003-03-27 00:44:31 by athena] Better place to install hook M ./tests/bench.c -2 +2 Wed Mar 26 19:40:28 EST 2003 stevenj * [project @ 2003-03-27 00:40:28 by stevenj] noted that the user should run make check if they think FFTW has a bug M ./doc/FAQ/fftw-faq.bfnn -6 +8 Wed Mar 26 17:31:16 EST 2003 athena * [project @ 2003-03-26 22:31:16 by athena] Oops, what am I thinking M ./kernel/planner.c -5 +8 Wed Mar 26 17:23:56 EST 2003 athena * [project @ 2003-03-26 22:23:56 by athena] Grrr.... fixed bug in estimator M ./kernel/planner.c -8 +5 Wed Mar 26 17:16:19 EST 2003 athena * [project @ 2003-03-26 22:16:19 by athena] Oops---the flop count was right. The estimator is broken elsewhere. M ./genfft/c.ml -5 +2 Wed Mar 26 14:28:41 EST 2003 athena * [project @ 2003-03-26 19:28:41 by athena] Fixed SIMD estimator M ./genfft/c.ml -2 +5 Wed Mar 26 07:45:03 EST 2003 athena * [project @ 2003-03-26 12:45:03 by athena] Added twidsq simd codelets A ./dft/simd/q1b.c A ./dft/simd/q1b.h A ./dft/simd/q1f.c A ./dft/simd/q1f.h A ./genfft/gen_twidsq_c.ml M ./dft/simd/Makefile.am -1 +2 M ./dft/simd/codelets/Makefile.am -1 +12 M ./dft/simd/q1b.c +44 M ./dft/simd/q1b.h +25 M ./dft/simd/q1f.c +44 M ./dft/simd/q1f.h +25 M ./dft/simd/t1b.c +1 M ./dft/simd/t1f.c +1 M ./genfft/Makefile.am -8 +14 M ./genfft/gen_twiddle_c.ml -3 +3 M ./genfft/gen_twidsq_c.ml +195 M ./support/Makefile.codelets -1 +2 Tue Mar 25 23:33:03 EST 2003 stevenj * [project @ 2003-03-26 04:33:03 by stevenj] gensrc -> genfft M ./doc/fftw3.texi -2 +2 Tue Mar 25 23:32:16 EST 2003 stevenj * [project @ 2003-03-26 04:32:16 by stevenj] newline M ./TODO +1 Tue Mar 25 19:17:08 EST 2003 athena * [project @ 2003-03-26 00:17:08 by athena] Noted need to add dif simd codelets M ./TODO +1 Tue Mar 25 13:03:47 EST 2003 stevenj * [project @ 2003-03-25 18:03:47 by stevenj] noted shift M ./doc/fftw3.texi -3 +2 Tue Mar 25 13:02:47 EST 2003 stevenj * [project @ 2003-03-25 18:02:47 by stevenj] clarification M ./doc/fftw3.texi -4 +14 Tue Mar 25 12:46:44 EST 2003 stevenj * [project @ 2003-03-25 17:46:44 by stevenj] need make after bootstrap M ./doc/fftw3.texi -2 +3 Tue Mar 25 12:31:49 EST 2003 stevenj * [project @ 2003-03-25 17:31:49 by stevenj] slight change M ./doc/fftw3.texi -3 +3 Tue Mar 25 12:30:56 EST 2003 stevenj * [project @ 2003-03-25 17:30:56 by stevenj] libtool is also needed M ./doc/fftw3.texi -4 +4 Tue Mar 25 12:29:52 EST 2003 stevenj * [project @ 2003-03-25 17:29:52 by stevenj] added code generator introduction M ./doc/fftw3.texi -13 +63 Tue Mar 25 11:51:49 EST 2003 stevenj * [project @ 2003-03-25 16:51:49 by stevenj] added support for REDFT/RODFT/DHT direct codelets A ./genfft/gen_r2r.ml A ./rdft/kr2r.c A ./rdft/codelets/r2r/ A ./rdft/codelets/r2r.c A ./rdft/codelets/r2r.h A ./rdft/codelets/r2r/Makefile.am M ./Makefile.am +1 M ./configure.ac +1 M ./genfft/Makefile.am -6 +6 M ./genfft/complex.ml -1 +8 M ./genfft/complex.mli -1 +2 M ./genfft/gen_r2r.ml +240 R ./genfft/gen_trig.ml M ./genfft/trig.ml -60 +47 M ./rdft/Makefile.am -3 +3 M ./rdft/codelet-rdft.h -2 +2 M ./rdft/codelets/Makefile.am -2 +2 M ./rdft/codelets/r2r/Makefile.am +72 M ./rdft/codelets/r2r.c +38 M ./rdft/codelets/r2r.h +23 M ./rdft/conf.c -1 +2 M ./rdft/direct.c -16 +58 M ./rdft/kr2r.c +28 M ./rdft/rdft.h -1 +2 M ./support/Makefile.codelets -1 +2 Tue Mar 25 11:29:29 EST 2003 stevenj * [project @ 2003-03-25 16:29:29 by stevenj] noted ARM bug; thanks to Jay Treacy M ./doc/FAQ/fftw-faq.bfnn -1 +2 Tue Mar 25 07:55:54 EST 2003 athena * [project @ 2003-03-25 12:55:54 by athena] bugfix from Stefan M ./genfft-k7/vK7Optimization.ml -1 +2 Mon Mar 24 15:59:08 EST 2003 stevenj * [project @ 2003-03-24 20:59:08 by stevenj] slight change M ./doc/fftw3.texi -2 +2 Mon Mar 24 15:58:44 EST 2003 stevenj * [project @ 2003-03-24 20:58:44 by stevenj] caveat M ./doc/fftw3.texi -2 +2 Mon Mar 24 15:58:04 EST 2003 stevenj * [project @ 2003-03-24 20:58:04 by stevenj] warning about DHT M ./doc/fftw3.texi -1 +8 Mon Mar 24 08:34:14 EST 2003 athena * [project @ 2003-03-24 13:34:14 by athena] Oops M ./dft/k7/codelets/Makefile.am -2 +2 Mon Mar 24 08:13:15 EST 2003 athena * [project @ 2003-03-24 13:13:15 by athena] Regression test for p4fftwgel M ./dft/k7/codelets/Makefile.am -2 +2 M ./tests/Makefile.am +3 M ./tests/check.pl +4 Mon Mar 24 03:09:06 EST 2003 stevenj * [project @ 2003-03-24 08:09:06 by stevenj] make check is faster, old tests are in make bigcheck M ./tests/Makefile.am -1 +6 Sat Mar 22 00:41:21 EST 2003 stevenj * [project @ 2003-03-22 05:41:21 by stevenj] note M ./doc/FAQ/fftw-faq.bfnn -1 +1 Sat Mar 22 00:40:05 EST 2003 stevenj * [project @ 2003-03-22 05:40:05 by stevenj] whoops, line wrapping M ./doc/FAQ/fftw-faq.bfnn -2 +1 Fri Mar 21 15:10:00 EST 2003 athena * [project @ 2003-03-21 20:10:00 by athena] Franz-mode codelets even without SIMD. (disabled) M ./dft/codelets/standard/Makefile.am -2 +59 M ./genfft/gen_notw.ml -5 +6 M ./genfft/gen_twiddle.ml -4 +5 Fri Mar 21 09:09:30 EST 2003 athena * [project @ 2003-03-21 14:09:30 by athena] Bug is in netbsd-1.6, not 1.5 M ./doc/FAQ/fftw-faq.bfnn -1 +1 Fri Mar 21 07:45:48 EST 2003 athena * [project @ 2003-03-21 12:45:48 by athena] const cast, should placate c++ compilers. M ./simd/simd-altivec.h -1 +1 Thu Mar 20 18:49:49 EST 2003 stevenj * [project @ 2003-03-20 23:49:49 by stevenj] added FAQ on why plans are array-specific M ./doc/FAQ/fftw-faq.bfnn +26 Thu Mar 20 16:12:56 EST 2003 stevenj * [project @ 2003-03-20 21:12:56 by stevenj] comment fix M ./reodft/reodft010e-r2hc.c -3 +3 Thu Mar 20 16:12:15 EST 2003 stevenj * [project @ 2003-03-20 21:12:15 by stevenj] noted comparison to NR M ./reodft/reodft010e-r2hc.c -1 +5 Wed Mar 19 20:13:16 EST 2003 stevenj * [project @ 2003-03-20 01:13:16 by stevenj] whoops, C99 complex didn't work if complex is a macro (as it is with glibc); thanks to Keh-Cheng Chu for the bug report M ./api/fftw3.h -5 +4 Wed Mar 19 16:52:54 EST 2003 stevenj * [project @ 2003-03-19 21:52:54 by stevenj] noted in help that --enable-k7 enables 3dnow, and that --enable-3dnow is only a fallback M ./configure.ac -2 +2 Wed Mar 19 15:09:52 EST 2003 athena * [project @ 2003-03-19 20:09:52 by athena] New gcc bug. html.refs was not in repository/distribution. A ./doc/FAQ/html.refs M ./doc/FAQ/Makefile.am -1 +1 M ./doc/FAQ/fftw-faq.bfnn -1 +13 M ./doc/FAQ/html.refs +6 Wed Mar 19 10:09:16 EST 2003 athena * [project @ 2003-03-19 15:09:16 by athena] Don't write wisdom if you don't have it. M ./tests/bench.c -1 +1 Tue Mar 18 15:44:41 EST 2003 athena * [project @ 2003-03-18 20:44:41 by athena] Added index entries for DHT. Similarly for DCT, DST M ./doc/fftw3.texi -1 +6 Tue Mar 18 14:50:04 EST 2003 stevenj * [project @ 2003-03-18 19:50:04 by stevenj] execute should not go through C api, for efficiency M ./api/f77api.c +2 M ./api/f77funcs.h -5 +10 Tue Mar 18 06:14:51 EST 2003 athena * [project @ 2003-03-18 11:14:51 by athena] Renamed FFTW_IODIM, FFTW_R2R_KIND M ./api/fftw3.h -6 +6 Tue Mar 18 00:30:17 EST 2003 stevenj * [project @ 2003-03-18 05:30:17 by stevenj] added rfftwnd.eps to dist, so that transfig is not required for people trying to build other formats (e.g. ps); thanks to Brian Gough for the bug report M ./doc/Makefile.am -1 +1 Mon Mar 17 15:17:59 EST 2003 stevenj * [project @ 2003-03-17 20:17:59 by stevenj] pointer to upgrading section from tutorial M ./doc/fftw3.texi -1 +4 Mon Mar 17 14:44:40 EST 2003 stevenj * [project @ 2003-03-17 19:44:40 by stevenj] make print_plan and fprint_plan, so that the former can be more easily called from other languages M ./api/f77funcs.h -2 +2 M ./api/fftw3.h -2 +3 M ./api/print-plan.c -1 +6 M ./doc/fftw3.texi -3 +5 M ./tests/bench.c -1 +1 Mon Mar 17 14:19:10 EST 2003 stevenj * [project @ 2003-03-17 19:19:10 by stevenj] whoops, forgot to change equation image links to .png M ./doc/fftw3.texi -14 +14 Mon Mar 17 04:15:50 EST 2003 athena * [project @ 2003-03-17 09:15:50 by athena] fixed c++ linkage problems M ./api/fftw3.h -4 +4 M ./api/version.c -8 +5 M ./support/Makefile.codelets -1 +2 Mon Mar 17 03:25:17 EST 2003 athena * [project @ 2003-03-17 08:25:17 by athena] Removed ``const'', otherwise c++ link fails M ./api/fftw3.h -4 +4 M ./api/version.c -4 +4 Sun Mar 16 20:24:31 EST 2003 stevenj * [project @ 2003-03-17 01:24:31 by stevenj] fixed C++ annoyances: void* casts, and global variables are static by default(?!?) M ./api/f77api.c -4 +5 M ./api/f77funcs.h -3 +5 M ./api/version.c -4 +4 M ./libbench2/allocate.c -9 +9 M ./libbench2/getopt-utils.c -2 +2 M ./libbench2/problem.c -3 +3 M ./libbench2/speed.c -2 +2 M ./libbench2/timer.c -2 +2 M ./libbench2/verify-r2r.c -3 +3 M ./libbench2/zero.c -9 +9 M ./support/Makefile.codelets -1 +1 M ./tests/bench.c -37 +66 M ./tests/hook.c -2 +2 M ./tools/fftw-wisdom.c -1 +1 Sat Aug 12 23:06:56 EDT 2006 Unknown tagger tagged fftw-3-0-beta1 Sun Mar 16 15:29:11 EST 2003 stevenj * [project @ 2003-03-16 20:29:11 by stevenj] ranlib bug is in binutils M ./doc/FAQ/fftw-faq.bfnn -4 +5 Sun Mar 16 15:26:42 EST 2003 stevenj * [project @ 2003-03-16 20:26:42 by stevenj] ranlib Irix bug M ./doc/FAQ/fftw-faq.bfnn +5 Sun Mar 16 15:13:35 EST 2003 stevenj * [project @ 2003-03-16 20:13:35 by stevenj] start with random tests M ./tests/check.pl -1 +1 Sun Mar 16 15:00:04 EST 2003 stevenj * [project @ 2003-03-16 20:00:03 by stevenj] silenced some compiler warnings, eliminated unused variables, and fixed Makefile.am for f77funcs.h M ./api/Makefile.am -5 +1 M ./dft/direct.c -3 +2 M ./kernel/ifftw.h -4 +4 M ./libbench2/verify-r2r.c -3 M ./rdft/direct.c -3 +2 M ./rdft/direct2.c -3 +2 M ./threads/Makefile.am -3 +1 Sun Mar 16 14:55:13 EST 2003 stevenj * [project @ 2003-03-16 19:55:13 by stevenj] whoops M ./doc/FAQ/fftw-faq.bfnn -1 +1 Sun Mar 16 14:28:22 EST 2003 stevenj * [project @ 2003-03-16 19:28:22 by stevenj] 3dnow is float M ./doc/fftw3.texi -2 +2 Sun Mar 16 14:27:45 EST 2003 stevenj * [project @ 2003-03-16 19:27:45 by stevenj] fixed k7 docs M ./doc/fftw3.texi -17 +10 Sun Mar 16 14:19:10 EST 2003 stevenj * [project @ 2003-03-16 19:19:10 by stevenj] SGI compilers now support inline M ./kernel/cycle.h -3 +3 Sun Mar 16 14:18:32 EST 2003 stevenj * [project @ 2003-03-16 19:18:32 by stevenj] cruft M ./kernel/cycle.h -2 +1 Sun Mar 16 14:15:47 EST 2003 stevenj * [project @ 2003-03-16 19:15:47 by stevenj] texinfo doesn't like commas in nodes M ./doc/fftw3.texi -16 +15 Sun Mar 16 13:52:04 EST 2003 stevenj * [project @ 2003-03-16 18:48:30 by stevenj] updated M ./ChangeLog -47 +1389 M ./README +28 Sun Mar 16 13:47:44 EST 2003 stevenj * [project @ 2003-03-16 18:47:44 by stevenj] f77funcs.c -> f77funcs.h so that people don't try to compile it A ./api/f77funcs.h A ./threads/f77funcs.h M ./api/f77api.c -4 +4 R ./api/f77funcs.c M ./api/f77funcs.h +366 M ./threads/f77api.c -4 +4 R ./threads/f77funcs.c M ./threads/f77funcs.h +39 Sun Mar 16 13:46:11 EST 2003 stevenj * [project @ 2003-03-16 18:46:11 by stevenj] minor changes M ./doc/FAQ/fftw-faq.bfnn -7 +4 Sun Mar 16 13:39:24 EST 2003 stevenj * [project @ 2003-03-16 18:39:24 by stevenj] updated compiler bug list M ./doc/FAQ/fftw-faq.bfnn -40 +20 Sun Mar 16 13:39:07 EST 2003 stevenj * [project @ 2003-03-16 18:39:07 by stevenj] noted how to set CC M ./doc/fftw3.texi -4 +10 Sun Mar 16 13:01:01 EST 2003 stevenj * [project @ 2003-03-16 18:01:01 by stevenj] TODONE M ./TODO -10 Sun Mar 16 13:00:42 EST 2003 stevenj * [project @ 2003-03-16 18:00:42 by stevenj] yikes, bugfix M ./threads/vrank-geq1-rdft2.c -19 +11 Sun Mar 16 10:26:28 EST 2003 stevenj * [project @ 2003-03-16 15:26:28 by stevenj] whoops M ./kernel/ifftw.h -2 +2 Sun Mar 16 09:24:19 EST 2003 athena * [project @ 2003-03-16 14:24:19 by athena] Report SIMD extensions in version string M ./api/version.c -2 +26 Sat Mar 15 18:56:11 EST 2003 stevenj * [project @ 2003-03-15 23:56:11 by stevenj] more verbose output M ./tests/bench.c +2 Sat Mar 15 17:41:25 EST 2003 stevenj * [project @ 2003-03-15 22:41:25 by stevenj] a couple of additional non-Unix instructions M ./doc/fftw3.texi -1 +9 Sat Mar 15 17:15:26 EST 2003 stevenj * [project @ 2003-03-15 22:15:26 by stevenj] hyphen M ./doc/FAQ/fftw-faq.bfnn -1 +1 Sat Mar 15 17:12:29 EST 2003 stevenj * [project @ 2003-03-15 22:12:29 by stevenj] softened M ./doc/FAQ/fftw-faq.bfnn -3 +3 Sat Mar 15 17:09:44 EST 2003 stevenj * [project @ 2003-03-15 22:09:44 by stevenj] added FAQ, used PNGs A ./doc/equation-dft.png A ./doc/equation-dht.png A ./doc/equation-idft.png A ./doc/equation-redft00.png A ./doc/equation-redft01.png A ./doc/equation-redft10.png A ./doc/equation-redft11.png A ./doc/equation-rodft00.png A ./doc/equation-rodft01.png A ./doc/equation-rodft10.png A ./doc/FAQ/ A ./doc/FAQ/Makefile.am A ./doc/FAQ/bfnnconv.pl A ./doc/FAQ/fftw-faq.bfnn A ./doc/FAQ/m-ascii.pl A ./doc/FAQ/m-html.pl A ./doc/FAQ/m-info.pl A ./doc/FAQ/m-lout.pl A ./doc/FAQ/m-post.pl A ./doc/equation-rodft11.png M ./configure.ac +1 M ./doc/FAQ/Makefile.am +14 M ./doc/FAQ/bfnnconv.pl +298 M ./doc/FAQ/fftw-faq.bfnn +492 M ./doc/FAQ/m-ascii.pl +189 M ./doc/FAQ/m-html.pl +337 M ./doc/FAQ/m-info.pl +226 M ./doc/FAQ/m-lout.pl +242 M ./doc/FAQ/m-post.pl +189 M ./doc/Makefile.am -4 +6 R ./doc/equation-dft.gif M ./doc/equation-dft.png R ./doc/equation-dht.gif M ./doc/equation-dht.png R ./doc/equation-idft.gif M ./doc/equation-idft.png R ./doc/equation-redft00.gif M ./doc/equation-redft00.png R ./doc/equation-redft01.gif M ./doc/equation-redft01.png R ./doc/equation-redft10.gif M ./doc/equation-redft10.png R ./doc/equation-redft11.gif M ./doc/equation-redft11.png R ./doc/equation-rodft00.gif M ./doc/equation-rodft00.png R ./doc/equation-rodft01.gif M ./doc/equation-rodft01.png R ./doc/equation-rodft10.gif M ./doc/equation-rodft10.png R ./doc/equation-rodft11.gif M ./doc/equation-rodft11.png Sat Mar 15 15:29:43 EST 2003 stevenj * [project @ 2003-03-15 20:29:42 by stevenj] great copyright update M ./COPYRIGHT -2 +2 M ./TODO -6 M ./api/api.h -2 +2 M ./api/apiplan.c -2 +2 M ./api/configure.c -2 +2 M ./api/execute-dft-c2r.c -2 +2 M ./api/execute-dft-r2c.c -2 +2 M ./api/execute-dft.c -2 +2 M ./api/execute-r2r.c -2 +2 M ./api/execute.c -2 +2 M ./api/export-wisdom-to-file.c -2 +2 M ./api/export-wisdom-to-string.c -2 +2 M ./api/export-wisdom.c -2 +2 M ./api/extract-reim.c -2 +2 M ./api/f77api.c -2 +2 M ./api/f77funcs.c -2 +2 M ./api/fftw3.h -3 +3 M ./api/flops.c -2 +2 M ./api/forget-wisdom.c -2 +2 M ./api/import-system-wisdom.c -2 +2 M ./api/import-wisdom-from-file.c -2 +2 M ./api/import-wisdom-from-string.c -2 +2 M ./api/import-wisdom.c -2 +2 M ./api/map-r2r-kind.c -2 +2 M ./api/mapflags.c -2 +2 M ./api/mkprinter-file.c -2 +2 M ./api/mktensor-iodims.c -2 +2 M ./api/mktensor-rowmajor.c -2 +2 M ./api/plan-dft-1d.c -2 +2 M ./api/plan-dft-2d.c -2 +2 M ./api/plan-dft-3d.c -2 +2 M ./api/plan-dft-c2r-1d.c -2 +2 M ./api/plan-dft-c2r-2d.c -2 +2 M ./api/plan-dft-c2r-3d.c -2 +2 M ./api/plan-dft-c2r.c -2 +2 M ./api/plan-dft-r2c-1d.c -2 +2 M ./api/plan-dft-r2c-2d.c -2 +2 M ./api/plan-dft-r2c-3d.c -2 +2 M ./api/plan-dft-r2c.c -2 +2 M ./api/plan-dft.c -2 +2 M ./api/plan-guru-dft-c2r.c -2 +2 M ./api/plan-guru-dft-r2c.c -2 +2 M ./api/plan-guru-dft.c -2 +2 M ./api/plan-guru-r2r.c -2 +2 M ./api/plan-many-dft-c2r.c -2 +2 M ./api/plan-many-dft-r2c.c -2 +2 M ./api/plan-many-dft.c -2 +2 M ./api/plan-many-r2r.c -2 +2 M ./api/plan-r2r-1d.c -2 +2 M ./api/plan-r2r-2d.c -2 +2 M ./api/plan-r2r-3d.c -2 +2 M ./api/plan-r2r.c -2 +2 M ./api/print-plan.c -2 +2 M ./api/rdft2-pad.c -2 +2 M ./api/the-planner.c -2 +2 M ./api/version.c -3 +3 M ./api/x77.h -2 +2 M ./dft/buffered.c -3 +3 M ./dft/codelet-dft.h -3 +3 M ./dft/codelets/n.c -2 +2 M ./dft/codelets/n.h -2 +2 M ./dft/codelets/t.c -2 +2 M ./dft/codelets/t.h -2 +2 M ./dft/conf.c -3 +3 M ./dft/ct-dif.c -3 +3 M ./dft/ct-dit.c -3 +3 M ./dft/ct-ditbuf.c -3 +3 M ./dft/ct-ditf.c -3 +3 M ./dft/ct.c -3 +3 M ./dft/ct.h -3 +3 M ./dft/dft.h -3 +3 M ./dft/direct.c -3 +3 M ./dft/generic.c -2 +2 M ./dft/indirect.c -3 +3 M ./dft/k7/k7.c -3 +3 M ./dft/kdft-dif.c -3 +3 M ./dft/kdft-difsq.c -3 +3 M ./dft/kdft-dit.c -3 +3 M ./dft/kdft.c -3 +3 M ./dft/nop.c -3 +3 M ./dft/plan.c -3 +3 M ./dft/problem.c -3 +3 M ./dft/rader-omega.c -2 +2 M ./dft/rader.c -2 +2 M ./dft/rank-geq2.c -3 +3 M ./dft/rank0.c -3 +3 M ./dft/simd/n1b.c -2 +2 M ./dft/simd/n1b.h -2 +2 M ./dft/simd/n1f.c -2 +2 M ./dft/simd/n1f.h -2 +2 M ./dft/simd/t1b.c -2 +2 M ./dft/simd/t1b.h -2 +2 M ./dft/simd/t1f.c -2 +2 M ./dft/simd/t1f.h -2 +2 M ./dft/solve.c -3 +3 M ./dft/vrank-geq1.c -3 +3 M ./dft/vrank2-transpose.c -3 +3 M ./dft/vrank3-transpose.c -3 +3 M ./dft/zero.c -3 +3 M ./doc/f77_wisdom.f -2 +2 M ./doc/fftw3.texi -4 +4 M ./genfft/algsimp.ml -4 +4 M ./genfft/algsimp.mli -3 +3 M ./genfft/annotate.ml -4 +4 M ./genfft/annotate.mli -3 +3 M ./genfft/assoctable.ml -3 +3 M ./genfft/assoctable.mli -3 +3 M ./genfft/c.ml -3 +3 M ./genfft/c.mli -3 +3 M ./genfft/complex.ml -3 +3 M ./genfft/complex.mli -3 +3 M ./genfft/conv.ml -2 +2 M ./genfft/conv.mli -3 +3 M ./genfft/dag.ml -3 +3 M ./genfft/dag.mli -3 +3 M ./genfft/expr.ml -3 +3 M ./genfft/expr.mli -3 +3 M ./genfft/fft.ml -4 +4 M ./genfft/fft.mli -3 +3 M ./genfft/gen_athnotw.ml -4 +4 M ./genfft/gen_athtw.ml -4 +4 M ./genfft/gen_conv.ml -4 +4 M ./genfft/gen_hc2hc.ml -4 +4 M ./genfft/gen_hc2r.ml -4 +4 M ./genfft/gen_notw.ml -4 +4 M ./genfft/gen_notw_c.ml -4 +4 M ./genfft/gen_r2hc.ml -4 +4 M ./genfft/gen_trig.ml -4 +4 M ./genfft/gen_twiddle.ml -4 +4 M ./genfft/gen_twiddle_c.ml -4 +4 M ./genfft/gen_twidsq.ml -4 +4 M ./genfft/genutil.ml -3 +3 M ./genfft/littlesimp.ml -3 +3 M ./genfft/littlesimp.mli -3 +3 M ./genfft/magic.ml -3 +3 M ./genfft/monads.ml -3 +3 M ./genfft/number.ml -3 +3 M ./genfft/number.mli -3 +3 M ./genfft/oracle.ml -3 +3 M ./genfft/oracle.mli -3 +3 M ./genfft/schedule.ml -3 +3 M ./genfft/schedule.mli -3 +3 M ./genfft/simd.ml -3 +3 M ./genfft/simd.mli -3 +3 M ./genfft/simdmagic.ml -3 +3 M ./genfft/to_alist.ml -3 +3 M ./genfft/to_alist.mli -3 +3 M ./genfft/trig.ml -3 +3 M ./genfft/trig.mli -3 +3 M ./genfft/twiddle.ml -3 +3 M ./genfft/twiddle.mli -3 +3 M ./genfft/unique.ml -3 +3 M ./genfft/unique.mli -3 +3 M ./genfft/util.ml -3 +3 M ./genfft/util.mli -3 +3 M ./genfft/variable.ml -3 +3 M ./genfft/variable.mli -3 +3 M ./genfft-k7/algsimp.ml -4 +4 M ./genfft-k7/algsimp.mli -3 +3 M ./genfft-k7/assoctable.ml -3 +3 M ./genfft-k7/assoctable.mli -3 +3 M ./genfft-k7/expr.ml -3 +3 M ./genfft-k7/expr.mli -3 +3 M ./genfft-k7/fft.ml -4 +4 M ./genfft-k7/littlesimp.ml -3 +3 M ./genfft-k7/littlesimp.mli -3 +3 M ./genfft-k7/monads.ml -3 +3 M ./genfft-k7/number.ml -3 +3 M ./genfft-k7/number.mli -3 +3 M ./genfft-k7/oracle.ml -3 +3 M ./genfft-k7/oracle.mli -3 +3 M ./genfft-k7/to_alist.ml -3 +3 M ./genfft-k7/to_alist.mli -3 +3 M ./genfft-k7/twiddle.ml -3 +3 M ./genfft-k7/twiddle.mli -3 +3 M ./kernel/align.c -3 +3 M ./kernel/alloc.c -3 +3 M ./kernel/assert.c -3 +3 M ./kernel/awake.c -3 +3 M ./kernel/buffered.c -2 +2 M ./kernel/ct.c -2 +2 M ./kernel/cycle.h -3 +3 M ./kernel/debug.c -3 +3 M ./kernel/hash.c -2 +2 M ./kernel/iabs.c -3 +3 M ./kernel/ifftw.h -3 +3 M ./kernel/md5-1.c -2 +2 M ./kernel/md5.c -2 +2 M ./kernel/minmax.c -3 +3 M ./kernel/ops.c -3 +3 M ./kernel/pickdim.c -3 +3 M ./kernel/plan.c -3 +3 M ./kernel/planner.c -2 +2 M ./kernel/primes.c -3 +3 M ./kernel/print.c -3 +3 M ./kernel/problem.c -3 +3 M ./kernel/rader.c -2 +2 M ./kernel/scan.c -3 +3 M ./kernel/solver.c -3 +3 M ./kernel/solvtab.c -3 +3 M ./kernel/square.c -3 +3 M ./kernel/stride.c -3 +3 M ./kernel/tensor.c -3 +3 M ./kernel/tensor1.c -3 +3 M ./kernel/tensor2.c -3 +3 M ./kernel/tensor4.c -3 +3 M ./kernel/tensor5.c -3 +3 M ./kernel/tensor7.c -3 +3 M ./kernel/tensor8.c -3 +3 M ./kernel/tensor9.c -3 +3 M ./kernel/timer.c -3 +3 M ./kernel/trig.c -3 +3 M ./kernel/trig1.c -3 +3 M ./kernel/twiddle.c -3 +3 M ./libbench/bench-main.c -2 +2 M ./libbench/bench-user.h -2 +2 M ./libbench/bench.h -2 +2 M ./libbench/can-do.c -2 +2 M ./libbench/getopt-utils.c -2 +2 M ./libbench/info.c -2 +2 M ./libbench/main.c -2 +2 M ./libbench/prime.c -2 +2 M ./libbench/problem.c -2 +2 M ./libbench/report.c -2 +2 M ./libbench/speed.c -2 +2 M ./libbench/timer.c -2 +2 M ./libbench/util.c -1 +1 M ./libbench/verify.c -2 +2 M ./libbench/zero.c -2 +2 M ./libbench2/aligned-main.c -2 +2 M ./libbench2/bench-main.c -2 +2 M ./libbench2/bench-user.h -2 +2 M ./libbench2/bench.h -2 +2 M ./libbench2/can-do.c -2 +2 M ./libbench2/dotens2.c -3 +3 M ./libbench2/getopt-utils.c -2 +2 M ./libbench2/info.c -2 +2 M ./libbench2/main.c -2 +2 M ./libbench2/problem.c -2 +2 M ./libbench2/report.c -2 +2 M ./libbench2/speed.c -2 +2 M ./libbench2/tensor.c -2 +2 M ./libbench2/timer.c -2 +2 M ./libbench2/useropt.c -2 +2 M ./libbench2/util.c -1 +1 M ./libbench2/verify-dft.c -3 +3 M ./libbench2/verify-lib.c -3 +3 M ./libbench2/verify-r2r.c -2 +2 M ./libbench2/verify-rdft2.c -3 +3 M ./libbench2/verify.c -2 +2 M ./libbench2/verify.h -2 +2 M ./libbench2/zero.c -2 +2 M ./rdft/buffered.c -3 +3 M ./rdft/buffered2.c -3 +3 M ./rdft/codelet-rdft.h -3 +3 M ./rdft/codelets/hb.h -2 +2 M ./rdft/codelets/hc2r.c -2 +2 M ./rdft/codelets/hc2r.h -2 +2 M ./rdft/codelets/hc2rIII.h -2 +2 M ./rdft/codelets/hf.h -2 +2 M ./rdft/codelets/hfb.c -2 +2 M ./rdft/codelets/r2hc.c -2 +2 M ./rdft/codelets/r2hc.h -2 +2 M ./rdft/codelets/r2hcII.h -2 +2 M ./rdft/conf.c -3 +3 M ./rdft/dft-r2hc.c -3 +3 M ./rdft/dht-r2hc.c -3 +3 M ./rdft/dht-rader.c -2 +2 M ./rdft/direct.c -3 +3 M ./rdft/direct2.c -3 +3 M ./rdft/generic.c -2 +2 M ./rdft/hc2hc-buf.c -3 +3 M ./rdft/hc2hc-dif.c -3 +3 M ./rdft/hc2hc-dit.c -3 +3 M ./rdft/hc2hc.c -3 +3 M ./rdft/hc2hc.h -3 +3 M ./rdft/indirect.c -3 +3 M ./rdft/khc2hc-dif.c -3 +3 M ./rdft/khc2hc-dit.c -3 +3 M ./rdft/khc2r.c -3 +3 M ./rdft/kr2hc.c -3 +3 M ./rdft/nop.c -3 +3 M ./rdft/nop2.c -3 +3 M ./rdft/plan.c -3 +3 M ./rdft/plan2.c -3 +3 M ./rdft/problem.c -3 +3 M ./rdft/problem2.c -3 +3 M ./rdft/rader-hc2hc.c -2 +2 M ./rdft/rank-geq2-rdft2.c -3 +3 M ./rdft/rank-geq2.c -3 +3 M ./rdft/rank0-rdft2.c -3 +3 M ./rdft/rank0.c -3 +3 M ./rdft/rdft-dht.c -3 +3 M ./rdft/rdft.h -3 +3 M ./rdft/rdft2-inplace-strides.c -3 +3 M ./rdft/rdft2-radix2.c -3 +3 M ./rdft/rdft2-tensor-max-index.c -3 +3 M ./rdft/solve.c -3 +3 M ./rdft/solve2.c -3 +3 M ./rdft/vrank-geq1-rdft2.c -3 +3 M ./rdft/vrank-geq1.c -3 +3 M ./rdft/vrank2-transpose.c -3 +3 M ./rdft/vrank3-transpose.c -3 +3 M ./reodft/conf.c -3 +3 M ./reodft/redft00e-r2hc.c -3 +3 M ./reodft/reodft.h -3 +3 M ./reodft/reodft010e-r2hc.c -3 +3 M ./reodft/reodft11e-r2hc-odd.c -3 +3 M ./reodft/reodft11e-r2hc.c -3 +3 M ./reodft/reodft11e-radix2.c -3 +3 M ./reodft/rodft00e-r2hc.c -3 +3 M ./simd/3dnow.c -3 +3 M ./simd/altivec.c -3 +3 M ./simd/simd-3dnow.h -2 +2 M ./simd/simd-altivec.h -2 +2 M ./simd/simd-sse.h -2 +2 M ./simd/simd-sse2.h -2 +2 M ./simd/simd.h -2 +2 M ./simd/sse-aux.c -3 +3 M ./simd/sse.c -3 +3 M ./simd/sse2-aux.c -3 +3 M ./simd/sse2.c -3 +3 M ./threads/api.c -2 +2 M ./threads/conf.c -3 +3 M ./threads/ct-dit.c -3 +3 M ./threads/dft-vrank-geq1.c -3 +3 M ./threads/f77api.c -2 +2 M ./threads/f77funcs.c -2 +2 M ./threads/hc2hc-dif.c -3 +3 M ./threads/hc2hc-dit.c -3 +3 M ./threads/rdft-vrank-geq1.c -3 +3 M ./threads/threads.c -2 +2 M ./threads/threads.h -3 +3 M ./threads/vrank-geq1-rdft2.c -3 +3 M ./tools/fftw-wisdom-to-conf.1 -4 +4 M ./tools/fftw-wisdom-to-conf.in -2 +2 M ./tools/fftw-wisdom.c -2 +2 M ./tools/fftw_wisdom.1.in -4 +4 Sat Mar 15 15:14:02 EST 2003 stevenj * [project @ 2003-03-15 20:14:02 by stevenj] threads in make check M ./TODO -4 M ./tests/Makefile.am -1 +5 M ./tests/check.pl +3 Sat Mar 15 15:11:24 EST 2003 stevenj * [project @ 2003-03-15 20:11:24 by stevenj] fixed const warnings M ./threads/ct-dit.c -2 +2 M ./threads/hc2hc-dif.c -2 +2 M ./threads/hc2hc-dit.c -2 +2 Sat Mar 15 15:08:25 EST 2003 stevenj * [project @ 2003-03-15 20:08:25 by stevenj] make sure spawn_loop size > 1 (it has to be at least > 0 lest we crash, but > 1 is an optimization) M ./threads/ct-dit.c -1 +2 M ./threads/hc2hc-dif.c -1 +2 M ./threads/hc2hc-dit.c -1 +2 Sat Mar 15 14:00:17 EST 2003 athena * [project @ 2003-03-15 19:00:17 by athena] hpux seems to want machine/sys/inline.h as opposed to machine/inline.h. M ./kernel/cycle.h -2 +2 Sat Mar 15 13:36:56 EST 2003 stevenj * [project @ 2003-03-15 18:36:56 by stevenj] Sourceforge is really SourceForge.net, and is run by VA M ./doc/fftw3.texi -3 +4 Sat Mar 15 13:34:05 EST 2003 stevenj * [project @ 2003-03-15 18:33:07 by stevenj] comma M ./doc/fftw3.texi -2 +2 Sat Mar 15 13:31:42 EST 2003 stevenj * [project @ 2003-03-15 18:31:42 by stevenj] fixed AMD company name M ./doc/fftw3.texi -2 +2 Sat Mar 15 13:29:41 EST 2003 stevenj * [project @ 2003-03-15 18:29:41 by stevenj] minor changes M ./doc/fftw3.texi -19 +19 Sat Mar 15 13:13:55 EST 2003 stevenj * [project @ 2003-03-15 18:13:55 by stevenj] more emitter->read_char renaming M ./api/f77api.c -5 +5 M ./api/f77funcs.c -4 +4 Sat Mar 15 13:08:45 EST 2003 stevenj * [project @ 2003-03-15 18:08:45 by stevenj] more wisdom docs, noted wisdom utilities M ./doc/fftw3.texi -35 +95 Sat Mar 15 11:41:32 EST 2003 stevenj * [project @ 2003-03-15 16:41:32 by stevenj] compound adjectives are hyphenated M ./doc/fftw3.texi -2 +2 Sat Mar 15 11:40:30 EST 2003 stevenj * [project @ 2003-03-15 16:40:30 by stevenj] fftw does support another type of packed array via r2r M ./doc/fftw3.texi -2 +2 Sat Mar 15 11:29:12 EST 2003 stevenj * [project @ 2003-03-15 16:29:12 by stevenj] write_char/read_char for export/import functions M ./api/export-wisdom.c -4 +4 M ./api/f77api.c -5 +5 M ./api/f77funcs.c -4 +4 M ./api/fftw3.h -3 +3 M ./api/import-wisdom.c -4 +4 M ./doc/f77_wisdom.f -6 +6 M ./doc/fftw3.texi -5 +5 Sat Mar 15 11:19:19 EST 2003 stevenj * [project @ 2003-03-15 16:19:19 by stevenj] comments M ./threads/threads.c -4 +8 Sat Mar 15 10:08:26 EST 2003 athena * [project @ 2003-03-15 15:08:26 by athena] Enabled randomized-cse M ./support/Makefile.codelets -1 +1 Sat Mar 15 09:47:49 EST 2003 athena * [project @ 2003-03-15 14:47:49 by athena] Changed to 3.0-beta1 M ./configure.ac -1 +1 Sat Mar 15 09:07:31 EST 2003 athena * [project @ 2003-03-15 14:07:31 by athena] First complete draft M ./doc/fftw3.texi -18 +131 Sat Mar 15 08:37:52 EST 2003 athena * [project @ 2003-03-15 13:37:52 by athena] EMITTER is a misnomer M ./api/fftw3.h -2 +2 M ./api/import-wisdom.c -8 +6 Sat Mar 15 05:50:50 EST 2003 athena * [project @ 2003-03-15 10:50:50 by athena] Revision, wisdom tutorial, acks. M ./doc/fftw3.texi -44 +219 Fri Mar 14 22:59:04 EST 2003 stevenj * [project @ 2003-03-15 03:59:04 by stevenj] noted OpenMP M ./NEWS +2 Fri Mar 14 22:38:49 EST 2003 stevenj * [project @ 2003-03-15 03:38:49 by stevenj] comment M ./threads/threads.c -1 +1 Fri Mar 14 22:38:30 EST 2003 stevenj * [project @ 2003-03-15 03:38:30 by stevenj] comments M ./threads/threads.c -2 +2 Fri Mar 14 22:38:05 EST 2003 stevenj * [project @ 2003-03-15 03:38:05 by stevenj] reformatting M ./threads/threads.c -3 +1 Fri Mar 14 22:26:28 EST 2003 stevenj * [project @ 2003-03-15 03:26:28 by stevenj] whoops M ./threads/threads.c -2 +2 Fri Mar 14 22:11:23 EST 2003 stevenj * [project @ 2003-03-15 03:11:23 by stevenj] some threads fixes, and added experimental semaphore (pre-thread-spawning) and Linux spinlock support M ./tests/bench.c -1 +10 M ./threads/api.c +1 M ./threads/threads.c -8 +159 M ./threads/threads.h -1 +2 Fri Mar 14 20:50:46 EST 2003 stevenj * [project @ 2003-03-15 01:50:46 by stevenj] whoops M ./threads/f77funcs.c -2 +2 Fri Mar 14 18:23:03 EST 2003 stevenj * [project @ 2003-03-14 23:23:03 by stevenj] added note that FFTW_PATIENT will disable threads if they are not beneficial M ./doc/fftw3.texi -2 +5 Fri Mar 14 18:20:44 EST 2003 stevenj * [project @ 2003-03-14 23:20:44 by stevenj] made fftw_cleanup* more restrictive, in that we don't want to guarantee that previously created plans will still work (they won't, in the case of threaded plans and fftw_cleanup_threads), and there is no reason to provide such a guarantee anyway. M ./doc/fftw3.texi -6 +14 Fri Mar 14 17:23:13 EST 2003 athena * [project @ 2003-03-14 22:23:13 by athena] Moved version.c from kernel/ into api/ A ./api/version.c M ./api/Makefile.am -1 +2 M ./api/version.c +28 M ./kernel/Makefile.am -1 +1 M ./kernel/ifftw.h -4 +1 R ./kernel/version.c Fri Mar 14 17:19:50 EST 2003 athena * [project @ 2003-03-14 22:19:50 by athena] icc-7.0 requires -openmp M ./configure.ac +5 Fri Mar 14 14:47:52 EST 2003 athena * [project @ 2003-03-14 19:47:52 by athena] Ensure that one can do make dist given the distribution M ./doc/Makefile.am -5 +5 Fri Mar 14 14:38:11 EST 2003 athena * [project @ 2003-03-14 19:38:11 by athena] Dist fftw3.pdf, not fftw.pdf M ./doc/Makefile.am -1 +2 Fri Mar 14 14:36:25 EST 2003 athena * [project @ 2003-03-14 19:36:25 by athena] Support -onthreads=%d M ./tests/bench.c -4 +8 Fri Mar 14 14:34:21 EST 2003 stevenj * [project @ 2003-03-14 19:34:21 by stevenj] comment M ./kernel/alloc.c -2 +2 Fri Mar 14 14:33:27 EST 2003 stevenj * [project @ 2003-03-14 19:33:27 by stevenj] whoops M ./threads/Makefile.am -1 +3 Fri Mar 14 12:32:18 EST 2003 stevenj * [project @ 2003-03-14 17:32:18 by stevenj] fftw_real is gone M ./doc/rfftwnd.fig -1 +1 Fri Mar 14 12:26:04 EST 2003 stevenj * [project @ 2003-03-14 17:26:04 by stevenj] typos M ./doc/fftw3.texi -3 +3 Fri Mar 14 06:21:43 EST 2003 athena * [project @ 2003-03-14 11:21:43 by athena] More BENCH_DOC strings M ./api/fftw3.h -3 +8 M ./tests/bench.c -4 +11 Fri Mar 14 05:58:53 EST 2003 athena * [project @ 2003-03-14 10:58:53 by athena] Fixed xref's M ./doc/fftw3.texi -15 +15 Fri Mar 14 05:38:26 EST 2003 athena * [project @ 2003-03-14 10:38:26 by athena] Revised manual (esp. intro and tutorial), fixed texinfo hackery for figures. M ./doc/Makefile.am -8 +13 M ./doc/fftw3.texi -165 +280 R ./doc/rfftwnd.gif Wed Mar 12 02:42:33 EST 2003 stevenj * [project @ 2003-03-12 07:42:33 by stevenj] redirect users from guru execute to advanced interface, if possible M ./doc/fftw3.texi -1 +5 Wed Mar 12 02:35:22 EST 2003 stevenj * [project @ 2003-03-12 07:35:22 by stevenj] punctuation M ./doc/fftw3.texi -2 +2 Wed Mar 12 02:28:51 EST 2003 stevenj * [project @ 2003-03-12 07:28:51 by stevenj] use correct heading level M ./doc/fftw3.texi -7 +7 Wed Mar 12 02:24:37 EST 2003 stevenj * [project @ 2003-03-12 07:24:37 by stevenj] html generation M ./doc/Makefile.am +15 M ./doc/fftw3.texi -91 +90 Wed Mar 12 01:44:00 EST 2003 stevenj * [project @ 2003-03-12 06:44:00 by stevenj] added equation GIFs A ./doc/equation-dft.gif A ./doc/equation-dht.gif A ./doc/equation-idft.gif A ./doc/equation-redft00.gif A ./doc/equation-redft01.gif A ./doc/equation-redft10.gif A ./doc/equation-redft11.gif A ./doc/equation-rodft00.gif A ./doc/equation-rodft01.gif A ./doc/equation-rodft10.gif A ./doc/equation-rodft11.gif M ./doc/equation-dft.gif M ./doc/equation-dht.gif M ./doc/equation-idft.gif M ./doc/equation-redft00.gif M ./doc/equation-redft01.gif M ./doc/equation-redft10.gif M ./doc/equation-redft11.gif M ./doc/equation-rodft00.gif M ./doc/equation-rodft01.gif M ./doc/equation-rodft10.gif M ./doc/equation-rodft11.gif Wed Mar 12 01:43:27 EST 2003 stevenj * [project @ 2003-03-12 06:43:27 by stevenj] punctuation M ./doc/fftw3.texi -18 +18 Wed Mar 12 01:26:46 EST 2003 stevenj * [project @ 2003-03-12 06:26:46 by stevenj] punctuation M ./doc/fftw3.texi -14 +14 Wed Mar 12 01:25:12 EST 2003 stevenj * [project @ 2003-03-12 06:25:12 by stevenj] added multi-dimensional transform definitions M ./doc/fftw3.texi -3 +124 Wed Mar 12 00:14:03 EST 2003 stevenj * [project @ 2003-03-12 05:14:03 by stevenj] slight changes M ./doc/fftw3.texi -3 +5 Wed Mar 12 00:06:34 EST 2003 stevenj * [project @ 2003-03-12 05:06:34 by stevenj] typo M ./doc/fftw3.texi -2 +2 Tue Mar 11 23:50:43 EST 2003 stevenj * [project @ 2003-03-12 04:50:43 by stevenj] added 1d version of What FFTW Really Computes M ./doc/fftw3.texi -1 +432 Tue Mar 11 21:17:54 EST 2003 stevenj * [project @ 2003-03-12 02:17:54 by stevenj] note in upgrading section about FFTW_PATIENT M ./doc/fftw3.texi -1 +7 Tue Mar 11 15:18:39 EST 2003 stevenj * [project @ 2003-03-11 20:18:39 by stevenj] added cycle-counter section M ./doc/fftw3.texi -1 +32 Tue Mar 11 14:53:44 EST 2003 stevenj * [project @ 2003-03-11 19:53:44 by stevenj] more ideas M ./TODO +7 Mon Mar 10 17:41:35 EST 2003 stevenj * [project @ 2003-03-10 22:41:35 by stevenj] noted that indirect should probably be merged with rank-geq2, to make a rank-split solver M ./dft/indirect.c -1 +4 M ./rdft/indirect.c -1 +4 Fri Mar 7 03:01:52 EST 2003 stevenj * [project @ 2003-03-07 08:01:52 by stevenj] added non-Unix installation instructions M ./doc/fftw3.texi -1 +38 Fri Mar 7 02:30:59 EST 2003 stevenj * [project @ 2003-03-07 07:30:59 by stevenj] also talk about stack alignment with SSE/SSE2 M ./doc/fftw3.texi -1 +7 Fri Mar 7 02:24:07 EST 2003 stevenj * [project @ 2003-03-07 07:24:07 by stevenj] made warning more dire M ./doc/fftw3.texi -3 +4 Fri Mar 7 02:13:25 EST 2003 stevenj * [project @ 2003-03-07 07:13:25 by stevenj] fix M ./doc/fftw3.texi -2 +2 Fri Mar 7 02:09:55 EST 2003 stevenj * [project @ 2003-03-07 07:09:55 by stevenj] number M ./doc/fftw3.texi -2 +2 Fri Mar 7 02:09:08 EST 2003 stevenj * [project @ 2003-03-07 07:09:08 by stevenj] fix M ./doc/fftw3.texi -3 +3 Fri Mar 7 02:08:01 EST 2003 stevenj * [project @ 2003-03-07 07:08:01 by stevenj] minor M ./doc/fftw3.texi -2 +2 Fri Mar 7 02:04:45 EST 2003 stevenj * [project @ 2003-03-07 07:04:45 by stevenj] minor fix M ./doc/fftw3.texi -3 +3 Fri Mar 7 01:58:15 EST 2003 stevenj * [project @ 2003-03-07 06:58:15 by stevenj] cross-ref M ./doc/fftw3.texi -2 +3 Fri Mar 7 01:57:31 EST 2003 stevenj * [project @ 2003-03-07 06:57:31 by stevenj] minor M ./doc/fftw3.texi -7 +8 Fri Mar 7 01:53:28 EST 2003 stevenj * [project @ 2003-03-07 06:53:28 by stevenj] more installation manual M ./doc/fftw3.texi -7 +151 Fri Mar 7 00:43:40 EST 2003 stevenj * [project @ 2003-03-07 05:43:40 by stevenj] GNU-lly correct M ./doc/fftw3.texi -2 +2 Fri Mar 7 00:38:48 EST 2003 stevenj * [project @ 2003-03-07 05:38:48 by stevenj] started installation section M ./doc/fftw3.texi -3 +33 Fri Mar 7 00:25:02 EST 2003 stevenj * [project @ 2003-03-07 05:25:02 by stevenj] added --without-cycle-counter option as a last resort M ./configure.ac +5 M ./kernel/timer.c -11 +24 Fri Mar 7 00:07:12 EST 2003 stevenj * [project @ 2003-03-07 05:07:12 by stevenj] macros with () arguments were only standardized in C99, and we don't need them anyway M ./kernel/cycle.h -3 +3 Thu Mar 6 23:10:41 EST 2003 stevenj * [project @ 2003-03-07 04:10:41 by stevenj] wording M ./doc/fftw3.texi -2 +2 Thu Mar 6 23:03:03 EST 2003 stevenj * [project @ 2003-03-07 04:03:03 by stevenj] parallelism M ./doc/fftw3.texi -2 +2 Thu Mar 6 23:01:47 EST 2003 stevenj * [project @ 2003-03-07 04:01:47 by stevenj] additions to upgrading chapter M ./doc/fftw3.texi -1 +31 Thu Mar 6 22:39:36 EST 2003 stevenj * [project @ 2003-03-07 03:39:36 by stevenj] noted additional humility of FFTW 3 wisdom M ./doc/fftw3.texi -1 +9 Thu Mar 6 22:32:44 EST 2003 stevenj * [project @ 2003-03-07 03:32:44 by stevenj] renaming M ./doc/fftw3.texi -6 +6 Thu Mar 6 22:31:00 EST 2003 stevenj * [project @ 2003-03-07 03:31:00 by stevenj] added placeholder for wisdom reference M ./doc/fftw3.texi -3 +10 Thu Mar 6 22:29:38 EST 2003 stevenj * [project @ 2003-03-07 03:29:38 by stevenj] wrote upgrading chapter M ./doc/fftw3.texi -1 +139 Thu Mar 6 18:01:10 EST 2003 stevenj * [project @ 2003-03-06 23:01:10 by stevenj] slight change M ./doc/fftw3.texi -6 +6 Thu Mar 6 18:00:43 EST 2003 stevenj * [project @ 2003-03-06 23:00:43 by stevenj] placeholder for upgrade chapter M ./doc/fftw3.texi -3 +18 Thu Mar 6 13:47:49 EST 2003 stevenj * [project @ 2003-03-06 18:47:49 by stevenj] whoops M ./tools/fftw-wisdom.c -2 +2 Thu Mar 6 13:36:38 EST 2003 stevenj * [project @ 2003-03-06 18:36:38 by stevenj] strengthed warning about time M ./tools/fftw_wisdom.1.in -1 +1 Thu Mar 6 13:35:42 EST 2003 stevenj * [project @ 2003-03-06 18:35:42 by stevenj] noted -t in example M ./tools/fftw_wisdom.1.in -1 +3 Thu Mar 6 13:21:03 EST 2003 stevenj * [project @ 2003-03-06 18:21:03 by stevenj] pay attention to WINDOWS_F77_MANGLING M ./threads/f77api.c -3 +26 Thu Mar 6 02:52:30 EST 2003 stevenj * [project @ 2003-03-06 07:52:30 by stevenj] punctuation M ./doc/fftw3.texi -2 +2 Thu Mar 6 02:51:02 EST 2003 stevenj * [project @ 2003-03-06 07:51:02 by stevenj] index M ./doc/fftw3.texi -1 +3 Thu Mar 6 02:50:38 EST 2003 stevenj * [project @ 2003-03-06 07:50:38 by stevenj] documented C++ usage M ./doc/fftw3.texi -3 +18 Thu Mar 6 02:25:32 EST 2003 stevenj * [project @ 2003-03-06 07:25:32 by stevenj] got rid of overfull hbox TeX warnings M ./doc/fftw3.texi -6 +6 Thu Mar 6 02:20:38 EST 2003 stevenj * [project @ 2003-03-06 07:20:38 by stevenj] whoops M ./doc/fftw3.texi -2 +2 Thu Mar 6 02:20:13 EST 2003 stevenj * [project @ 2003-03-06 07:20:13 by stevenj] noted fftw_iodim split for Fortran guru interface M ./doc/fftw3.texi -5 +16 Thu Mar 6 02:14:21 EST 2003 stevenj * [project @ 2003-03-06 07:14:21 by stevenj] added guru reference M ./doc/fftw3.texi -3 +187 Wed Mar 5 22:56:05 EST 2003 stevenj * [project @ 2003-03-06 03:56:05 by stevenj] minor M ./doc/fftw3.texi -3 +3 Wed Mar 5 22:45:31 EST 2003 stevenj * [project @ 2003-03-06 03:45:31 by stevenj] use @r{...} for comment text in code examples M ./doc/fftw3.texi -3 +3 Wed Mar 5 13:14:04 EST 2003 stevenj * [project @ 2003-03-05 18:14:04 by stevenj] eliminate warning M ./simd/sse.c -1 +2 Wed Mar 5 13:12:56 EST 2003 stevenj * [project @ 2003-03-05 18:12:56 by stevenj] SIMD_CFLAGS only for simd code M ./configure.ac -5 +8 M ./dft/simd/Makefile.am -1 +1 M ./dft/simd/codelets/Makefile.am -1 +1 M ./kernel/align.c -12 +9 M ./simd/Makefile.am +1 Wed Mar 5 11:06:41 EST 2003 athena * [project @ 2003-03-05 16:06:41 by athena] Minor changes. M ./doc/fftw3.texi -89 +96 Wed Mar 5 02:13:34 EST 2003 stevenj * [project @ 2003-03-05 07:13:34 by stevenj] cross-compiling with MinGW can't detect f77 mangling, so add an option to use what seems to be the most common styles M ./api/f77api.c -3 +34 M ./configure.ac +5 Tue Mar 4 20:00:31 EST 2003 stevenj * [project @ 2003-03-05 01:00:31 by stevenj] comment M ./libbench2/util.c -2 +3 Tue Mar 4 20:00:13 EST 2003 stevenj * [project @ 2003-03-05 01:00:13 by stevenj] we only use our-malloc-16 on machines where size_t == uintptr_t, so don't bother doing the right thing with the benchmark M ./libbench2/util.c -1 +1 Tue Mar 4 19:46:09 EST 2003 stevenj * [project @ 2003-03-05 00:46:09 by stevenj] support WITH_OUR_MALLOC16 M ./libbench2/util.c -2 +19 Tue Mar 4 18:50:53 EST 2003 fftw * [project @ 2003-03-04 23:50:53 by fftw] automatically add -msse etcetera for --enable-sse etcetera M ./configure.ac +22 Tue Mar 4 18:24:26 EST 2003 fftw * [project @ 2003-03-04 23:24:26 by fftw] got rid of const warning M ./tools/fftw-wisdom.c -2 +2 Tue Mar 4 18:22:48 EST 2003 fftw * [project @ 2003-03-04 23:22:48 by fftw] missing header M ./libbench2/problem.c -1 +2 Tue Mar 4 15:55:47 EST 2003 stevenj * [project @ 2003-03-04 20:55:47 by stevenj] fixes M ./doc/fftw3.texi -15 +49 Tue Mar 4 15:53:26 EST 2003 stevenj * [project @ 2003-03-04 20:53:26 by stevenj] whoops M ./api/import-system-wisdom.c -1 +1 Tue Mar 4 02:22:14 EST 2003 stevenj * [project @ 2003-03-04 07:22:14 by stevenj] started guru reference M ./doc/fftw3.texi -1 +110 Tue Mar 4 01:44:09 EST 2003 stevenj * [project @ 2003-03-04 06:44:09 by stevenj] use same FFTW_IODIM between precisions M ./api/fftw3.h -6 +8 Tue Mar 4 00:25:57 EST 2003 stevenj * [project @ 2003-03-04 05:25:57 by stevenj] renamed section M ./doc/fftw3.texi -5 +5 Tue Mar 4 00:21:49 EST 2003 stevenj * [project @ 2003-03-04 05:21:49 by stevenj] no need for "advanced" in subheadings M ./doc/fftw3.texi -4 +4 Tue Mar 4 00:20:05 EST 2003 stevenj * [project @ 2003-03-04 05:20:05 by stevenj] typo M ./doc/fftw3.texi -2 +2 Tue Mar 4 00:17:23 EST 2003 stevenj * [project @ 2003-03-04 05:17:23 by stevenj] finished advanced interface M ./doc/fftw3.texi -1 +34 Mon Mar 3 23:26:12 EST 2003 stevenj * [project @ 2003-03-04 04:26:12 by stevenj] more advance interface docs M ./doc/fftw3.texi -1 +46 Mon Mar 3 23:12:09 EST 2003 stevenj * [project @ 2003-03-04 04:12:09 by stevenj] fail for win32 M ./api/import-system-wisdom.c -2 +2 Mon Mar 3 17:18:48 EST 2003 fftw * [project @ 2003-03-03 22:18:48 by fftw] shortened help string M ./configure.ac -1 +1 Mon Mar 3 17:16:17 EST 2003 fftw * [project @ 2003-03-03 22:16:17 by fftw] fixed cross-refs M ./doc/fftw3.texi -3 +3 Mon Mar 3 17:07:27 EST 2003 fftw * [project @ 2003-03-03 22:07:27 by fftw] FFTW_POSSIBLY_UNALIGNED -> simpler FFTW_UNALIGNED in API, added bench option M ./api/fftw3.h -2 +2 M ./api/mapflags.c -1 +1 M ./doc/fftw3.texi -3 +3 M ./tests/bench.c +1 Mon Mar 3 16:58:07 EST 2003 fftw * [project @ 2003-03-03 21:58:07 by fftw] whoops M ./kernel/alloc.c -3 +3 Mon Mar 3 16:52:58 EST 2003 fftw * [project @ 2003-03-03 21:52:58 by fftw] noted assumption M ./kernel/alloc.c -2 +2 Mon Mar 3 16:50:33 EST 2003 fftw * [project @ 2003-03-03 21:50:33 by fftw] provide our own malloc16 routine because of Windows lossage M ./configure.ac +5 M ./kernel/alloc.c -3 +35 Mon Mar 3 13:28:12 EST 2003 stevenj * [project @ 2003-03-03 18:28:12 by stevenj] capitalization M ./doc/fftw3.texi -6 +6 Mon Mar 3 13:26:32 EST 2003 stevenj * [project @ 2003-03-03 18:26:32 by stevenj] whoops M ./doc/fftw3.texi -2 +2 Mon Mar 3 12:55:57 EST 2003 stevenj * [project @ 2003-03-03 17:55:57 by stevenj] vertical skip looks better than indenting for setting off short paragraphs M ./doc/fftw3.texi -1 +6 Mon Mar 3 06:34:09 EST 2003 athena * [project @ 2003-03-03 11:34:09 by athena] Removed franz-mode. Automake was distributing franz files whether franz mode was enabled or not. M ./configure.ac -3 M ./dft/simd/codelets/Makefile.am -7 +2 Mon Mar 3 01:44:00 EST 2003 stevenj * [project @ 2003-03-03 06:44:00 by stevenj] made output boundary conditions more prominent; they are important, because they make the different transform types inequivalent in parity M ./doc/fftw3.texi -4 +8 Mon Mar 3 01:17:28 EST 2003 stevenj * [project @ 2003-03-03 06:17:28 by stevenj] clarification M ./doc/fftw3.texi -2 +2 Mon Mar 3 01:17:07 EST 2003 stevenj * [project @ 2003-03-03 06:17:07 by stevenj] typo M ./doc/fftw3.texi -2 +2 Mon Mar 3 01:10:28 EST 2003 stevenj * [project @ 2003-03-03 06:10:28 by stevenj] started advanced reference M ./doc/fftw3.texi -6 +51 Mon Mar 3 00:52:02 EST 2003 stevenj * [project @ 2003-03-03 05:52:02 by stevenj] r2r reference M ./doc/fftw3.texi -2 +209 Sun Mar 2 23:51:21 EST 2003 stevenj * [project @ 2003-03-03 04:51:21 by stevenj] workaround for info formatting bug M ./doc/fftw3.texi -5 +4 Sun Mar 2 23:47:19 EST 2003 stevenj * [project @ 2003-03-03 04:47:19 by stevenj] noted lack of fftw_malloc in Fortran M ./doc/fftw3.texi -3 +9 Sun Mar 2 23:42:52 EST 2003 stevenj * [project @ 2003-03-03 04:42:32 by stevenj] parallelism M ./doc/fftw3.texi -4 +4 Sun Mar 2 23:39:54 EST 2003 stevenj * [project @ 2003-03-03 04:39:05 by stevenj] whoops M ./doc/fftw3.texi -7 +7 Sun Mar 2 23:33:02 EST 2003 stevenj * [project @ 2003-03-03 04:33:02 by stevenj] r2c/c2r reference M ./doc/fftw3.texi -40 +248 Sun Mar 2 22:44:10 EST 2003 stevenj * [project @ 2003-03-03 03:44:10 by stevenj] table of contents was being included twice M ./doc/fftw3.texi -2 +1 Sun Mar 2 22:42:29 EST 2003 stevenj * [project @ 2003-03-03 03:42:29 by stevenj] minor changes M ./doc/fftw3.texi -15 +16 Sun Mar 2 21:54:13 EST 2003 stevenj * [project @ 2003-03-03 02:54:13 by stevenj] started reference section M ./doc/fftw3.texi -2 +238 Sun Mar 2 19:10:02 EST 2003 stevenj * [project @ 2003-03-03 00:10:02 by stevenj] whoops M ./doc/Makefile.am -2 +2 Sun Mar 2 19:03:23 EST 2003 stevenj * [project @ 2003-03-03 00:03:23 by stevenj] started ref. section M ./doc/fftw3.texi -2 +146 Sun Mar 2 18:50:58 EST 2003 stevenj * [project @ 2003-03-02 23:50:58 by stevenj] fftw_flops takes const plan M ./api/fftw3.h -2 +2 M ./api/flops.c -1 +1 Sun Mar 2 15:54:14 EST 2003 stevenj * [project @ 2003-03-02 20:54:14 by stevenj] typo M ./doc/fftw3.texi -2 +2 Sun Mar 2 15:52:41 EST 2003 stevenj * [project @ 2003-03-02 20:52:41 by stevenj] added "Wisdom of Fortran?" section M ./doc/fftw3.texi -1 +39 Sun Mar 2 15:50:37 EST 2003 stevenj * [project @ 2003-03-02 20:50:37 by stevenj] typo M ./doc/f77_wisdom.f -2 +2 Sun Mar 2 15:49:57 EST 2003 stevenj * [project @ 2003-03-02 20:49:57 by stevenj] wording M ./doc/f77_wisdom.f -1 +1 Sun Mar 2 15:46:13 EST 2003 stevenj * [project @ 2003-03-02 20:46:13 by stevenj] added comments M ./doc/f77_wisdom.f +25 Sun Mar 2 15:44:01 EST 2003 stevenj * [project @ 2003-03-02 20:44:01 by stevenj] added example file A ./doc/f77_wisdom.f Sun Mar 2 15:37:32 EST 2003 stevenj * [project @ 2003-03-02 20:37:32 by stevenj] don't print out READ WISDOM unless we have M ./tests/bench.c -6 +11 Sun Mar 2 15:36:28 EST 2003 stevenj * [project @ 2003-03-02 20:36:28 by stevenj] EOF is not a space M ./kernel/scan.c -2 +2 Sun Mar 2 09:14:37 EST 2003 athena * [project @ 2003-03-02 14:14:37 by athena] Turn on inline by default M ./kernel/ifftw.h -2 +3 Sun Mar 2 07:11:56 EST 2003 athena * [project @ 2003-03-02 12:11:56 by athena] Optionally inline loop in notw codelets M ./genfft/gen_hc2r.ml -3 +3 M ./genfft/gen_notw.ml -3 +3 M ./genfft/gen_notw_c.ml -3 +3 M ./genfft/gen_r2hc.ml -3 +3 M ./kernel/ifftw.h -1 +3 Sun Mar 2 01:37:41 EST 2003 stevenj * [project @ 2003-03-02 06:37:41 by stevenj] updated nodes M ./doc/fftw3.texi -5 +9 Sun Mar 2 01:37:19 EST 2003 stevenj * [project @ 2003-03-02 06:37:19 by stevenj] wrote most of Fortran chapter M ./doc/fftw3.texi -1 +208 Sun Mar 2 00:58:37 EST 2003 stevenj * [project @ 2003-03-02 05:58:37 by stevenj] citation M ./doc/fftw3.texi -2 +2 Sun Mar 2 00:57:22 EST 2003 stevenj * [project @ 2003-03-02 05:57:22 by stevenj] added parallel FFTW chapter M ./doc/fftw3.texi -3 +209 Sat Mar 1 20:42:23 EST 2003 stevenj * [project @ 2003-03-02 01:42:23 by stevenj] typo M ./doc/fftw3.texi -2 +2 Sat Mar 1 20:34:38 EST 2003 stevenj * [project @ 2003-03-02 01:34:38 by stevenj] added inlining to TODO M ./TODO +2 Sat Mar 1 19:36:26 EST 2003 stevenj * [project @ 2003-03-02 00:36:26 by stevenj] added K M ./CONVENTIONS -1 +2 Sat Mar 1 19:15:18 EST 2003 stevenj * [project @ 2003-03-02 00:15:18 by stevenj] use K for constants M ./dft/zero.c -3 +3 M ./kernel/trig1.c -5 +5 M ./rdft/generic.c -8 +8 M ./rdft/problem.c -3 +3 M ./rdft/rdft-dht.c -3 +3 M ./rdft/rdft2-radix2.c -11 +11 M ./reodft/redft00e-r2hc.c -3 +3 M ./reodft/reodft010e-r2hc.c -11 +11 M ./reodft/reodft11e-r2hc.c -7 +7 M ./reodft/reodft11e-radix2.c -11 +11 M ./reodft/rodft00e-r2hc.c -3 +3 Sat Mar 1 19:14:54 EST 2003 stevenj * [project @ 2003-03-02 00:14:54 by stevenj] fixed cross-ref M ./doc/fftw3.texi -2 +2 Sat Mar 1 19:14:16 EST 2003 stevenj * [project @ 2003-03-02 00:14:16 by stevenj] whoops M ./doc/fftw3.texi -6 +6 Sat Mar 1 18:50:43 EST 2003 stevenj * [project @ 2003-03-01 23:50:43 by stevenj] cleanup M ./doc/fftw3.texi -8 +8 Sat Mar 1 18:46:38 EST 2003 stevenj * [project @ 2003-03-01 23:46:38 by stevenj] "words of wisdom" by itself is a little too obscure M ./doc/fftw3.texi -6 +6 Sat Mar 1 18:43:21 EST 2003 stevenj * [project @ 2003-03-01 23:43:21 by stevenj] re-added multi-dimensional array stuff M ./doc/fftw3.texi -1 +203 Sat Mar 1 18:15:22 EST 2003 stevenj * [project @ 2003-03-01 23:15:22 by stevenj] added alignment section M ./doc/fftw3.texi -4 +93 Sat Mar 1 16:34:21 EST 2003 stevenj * [project @ 2003-03-01 21:34:21 by stevenj] shrunk code M ./reodft/reodft11e-r2hc-odd.c -147 +57 Fri Feb 28 20:22:00 EST 2003 stevenj * [project @ 2003-03-01 01:22:00 by stevenj] slight compression M ./reodft/reodft11e-r2hc-odd.c -146 +122 Fri Feb 28 19:01:20 EST 2003 stevenj * [project @ 2003-03-01 00:01:20 by stevenj] style M ./doc/fftw3.texi -7 +7 M ./reodft/reodft11e-radix2.c -2 +2 Fri Feb 28 18:46:53 EST 2003 stevenj * [project @ 2003-02-28 23:46:53 by stevenj] noted not in API M ./CONVENTIONS -1 +1 Fri Feb 28 18:43:14 EST 2003 stevenj * [project @ 2003-02-28 23:43:14 by stevenj] more updates M ./CONVENTIONS -3 +4 Fri Feb 28 18:38:42 EST 2003 stevenj * [project @ 2003-02-28 23:38:42 by stevenj] slight updates M ./CONVENTIONS -3 +4 Fri Feb 28 18:28:58 EST 2003 stevenj * [project @ 2003-02-28 23:28:58 by stevenj] great const-ification of apply/solve and print M ./api/f77funcs.c -1 +1 M ./api/fftw3.h -2 +2 M ./api/print-plan.c -1 +1 M ./dft/buffered.c -5 +5 M ./dft/ct-dif.c -3 +3 M ./dft/ct-dit.c -3 +3 M ./dft/ct-ditbuf.c -3 +3 M ./dft/ct-ditf.c -3 +3 M ./dft/ct.c -3 +3 M ./dft/dft.h -3 +3 M ./dft/direct.c -5 +5 M ./dft/generic.c -4 +4 M ./dft/indirect.c -7 +7 M ./dft/nop.c -3 +3 M ./dft/rader.c -12 +12 M ./dft/rank-geq2.c -5 +5 M ./dft/rank0.c -12 +12 M ./dft/solve.c -3 +3 M ./dft/vrank-geq1.c -5 +5 M ./dft/vrank2-transpose.c -5 +5 M ./dft/vrank3-transpose.c -5 +5 M ./kernel/ifftw.h -3 +3 M ./rdft/buffered.c -5 +5 M ./rdft/buffered2.c -7 +7 M ./rdft/dft-r2hc.c -5 +5 M ./rdft/dht-r2hc.c -5 +5 M ./rdft/dht-rader.c -4 +4 M ./rdft/direct.c -7 +7 M ./rdft/direct2.c -7 +7 M ./rdft/generic.c -6 +6 M ./rdft/hc2hc-buf.c -5 +5 M ./rdft/hc2hc-dif.c -3 +3 M ./rdft/hc2hc-dit.c -3 +3 M ./rdft/hc2hc.c -3 +3 M ./rdft/indirect.c -7 +7 M ./rdft/nop.c -3 +3 M ./rdft/nop2.c -3 +3 M ./rdft/rader-hc2hc.c -6 +6 M ./rdft/rank-geq2-rdft2.c -7 +7 M ./rdft/rank-geq2.c -5 +5 M ./rdft/rank0-rdft2.c -9 +9 M ./rdft/rank0.c -8 +8 M ./rdft/rdft-dht.c -9 +9 M ./rdft/rdft.h -5 +5 M ./rdft/rdft2-radix2.c -12 +12 M ./rdft/solve.c -3 +3 M ./rdft/solve2.c -3 +3 M ./rdft/vrank-geq1-rdft2.c -7 +7 M ./rdft/vrank-geq1.c -5 +5 M ./rdft/vrank2-transpose.c -5 +5 M ./rdft/vrank3-transpose.c -5 +5 M ./reodft/redft00e-r2hc.c -5 +5 M ./reodft/reodft010e-r2hc.c -11 +11 M ./reodft/reodft11e-r2hc-odd.c -7 +7 M ./reodft/reodft11e-r2hc.c -7 +7 M ./reodft/reodft11e-radix2.c -9 +9 M ./reodft/rodft00e-r2hc.c -5 +5 M ./threads/ct-dit.c -3 +3 M ./threads/dft-vrank-geq1.c -5 +5 M ./threads/hc2hc-dif.c -3 +3 M ./threads/hc2hc-dit.c -3 +3 M ./threads/rdft-vrank-geq1.c -5 +5 M ./threads/vrank-geq1-rdft2.c -7 +7 Fri Feb 28 17:51:15 EST 2003 stevenj * [project @ 2003-02-28 22:51:15 by stevenj] make fftw_execute take a const plan, to remind the user that it is re-entrant (or should be)... M ./api/execute-dft-c2r.c -1 +1 M ./api/execute-dft-r2c.c -1 +1 M ./api/execute-dft.c -1 +1 M ./api/execute-r2r.c -1 +1 M ./api/execute.c -1 +1 M ./api/f77funcs.c -5 +6 M ./api/fftw3.h -6 +6 M ./doc/fftw3.texi -2 +2 Fri Feb 28 17:29:40 EST 2003 stevenj * [project @ 2003-02-28 22:29:40 by stevenj] weakening M ./doc/fftw3.texi -2 +2 Fri Feb 28 17:28:48 EST 2003 stevenj * [project @ 2003-02-28 22:28:48 by stevenj] note M ./doc/fftw3.texi -4 +4 Fri Feb 28 17:27:10 EST 2003 stevenj * [project @ 2003-02-28 22:27:10 by stevenj] footnote about why DHT is provided M ./doc/fftw3.texi -7 +12 Fri Feb 28 15:07:03 EST 2003 stevenj * [project @ 2003-02-28 20:07:03 by stevenj] index M ./doc/fftw3.texi -1 +2 Fri Feb 28 15:05:48 EST 2003 stevenj * [project @ 2003-02-28 20:05:48 by stevenj] added DHT tutorial M ./doc/fftw3.texi -2 +32 Fri Feb 28 14:36:45 EST 2003 stevenj * [project @ 2003-02-28 19:36:45 by stevenj] fixed O(n log n) M ./doc/fftw3.texi -4 +16 Fri Feb 28 14:12:15 EST 2003 stevenj * [project @ 2003-02-28 19:08:03 by stevenj] whoops M ./doc/fftw3.texi -4 +10 Fri Feb 28 14:06:22 EST 2003 stevenj * [project @ 2003-02-28 19:06:22 by stevenj] slight improvements M ./doc/fftw3.texi -2 +12 Fri Feb 28 00:55:50 EST 2003 stevenj * [project @ 2003-02-28 05:55:50 by stevenj] addition M ./doc/fftw3.texi -3 +5 Fri Feb 28 00:54:09 EST 2003 stevenj * [project @ 2003-02-28 05:54:09 by stevenj] clarification M ./doc/fftw3.texi -7 +8 Thu Feb 27 23:49:37 EST 2003 stevenj * [project @ 2003-02-28 04:49:37 by stevenj] fix M ./doc/fftw3.texi -2 +2 Thu Feb 27 23:43:56 EST 2003 stevenj * [project @ 2003-02-28 04:43:56 by stevenj] slight changes M ./doc/fftw3.texi -15 +21 Thu Feb 27 23:27:48 EST 2003 stevenj * [project @ 2003-02-28 04:27:48 by stevenj] added R{E,O}DFTab tutorial M ./doc/fftw3.texi -9 +105 Thu Feb 27 17:24:20 EST 2003 stevenj * [project @ 2003-02-27 22:24:20 by stevenj] fixes M ./doc/fftw3.texi -12 +11 Thu Feb 27 17:20:42 EST 2003 stevenj * [project @ 2003-02-27 22:20:42 by stevenj] fixes M ./doc/fftw3.texi -24 +26 Thu Feb 27 17:11:54 EST 2003 stevenj * [project @ 2003-02-27 22:11:22 by stevenj] slight change M ./doc/fftw3.texi -4 +4 Thu Feb 27 17:07:45 EST 2003 stevenj * [project @ 2003-02-27 22:07:45 by stevenj] documented r2hc/hc2r M ./doc/fftw3.texi -6 +86 Thu Feb 27 16:19:16 EST 2003 stevenj * [project @ 2003-02-27 21:19:16 by stevenj] minor changes M ./doc/fftw3.texi -24 +26 Thu Feb 27 13:54:06 EST 2003 stevenj * [project @ 2003-02-27 18:54:06 by stevenj] timed planner and unifying radix-2 butterfly loops are not critical for release M ./TODO -8 +8 Thu Feb 27 13:51:20 EST 2003 stevenj * [project @ 2003-02-27 18:51:20 by stevenj] reodft/verify.c no longer exists M ./TODO -2 Thu Feb 27 13:44:19 EST 2003 stevenj * [project @ 2003-02-27 18:44:19 by stevenj] optimization: REDFT00 of size 2 is same as R2HC M ./rdft/problem.c -1 +5 Thu Feb 27 12:35:33 EST 2003 stevenj * [project @ 2003-02-27 17:35:33 by stevenj] R{E,O}DFT01 of size-1 is identity M ./rdft/problem.c -3 +4 Thu Feb 27 12:15:10 EST 2003 stevenj * [project @ 2003-02-27 17:15:10 by stevenj] minor simplification M ./reodft/reodft11e-r2hc-odd.c -31 +31 Thu Feb 27 02:46:31 EST 2003 stevenj * [project @ 2003-02-27 07:46:31 by stevenj] fixed add count M ./reodft/reodft11e-r2hc-odd.c -2 +2 Thu Feb 27 02:25:04 EST 2003 stevenj * [project @ 2003-02-27 07:25:04 by stevenj] whoops M ./reodft/reodft11e-r2hc-odd.c -5 +5 Thu Feb 27 02:22:03 EST 2003 stevenj * [project @ 2003-02-27 07:22:03 by stevenj] another optimization M ./reodft/reodft11e-r2hc-odd.c -53 +53 Thu Feb 27 01:43:00 EST 2003 stevenj * [project @ 2003-02-27 06:43:00 by stevenj] added op counts M ./reodft/reodft11e-r2hc-odd.c -2 +4 M ./reodft/reodft11e-radix2.c -2 +9 Thu Feb 27 01:29:32 EST 2003 stevenj * [project @ 2003-02-27 06:29:32 by stevenj] cleanup M ./reodft/reodft11e-r2hc-odd.c -53 +55 Thu Feb 27 01:17:23 EST 2003 stevenj * [project @ 2003-02-27 06:17:23 by stevenj] typo in comment M ./reodft/reodft11e-r2hc-odd.c -2 +2 Thu Feb 27 01:13:49 EST 2003 stevenj * [project @ 2003-02-27 06:13:49 by stevenj] fixed comment M ./reodft/reodft11e-r2hc-odd.c -2 +2 Thu Feb 27 01:12:05 EST 2003 stevenj * [project @ 2003-02-27 06:12:05 by stevenj] use E instead of R M ./reodft/reodft11e-r2hc-odd.c -9 +9 Thu Feb 27 01:05:39 EST 2003 stevenj * [project @ 2003-02-27 06:05:39 by stevenj] more unrolling to eliminate if statements in loops, for speedups of 25-40% M ./reodft/reodft11e-r2hc-odd.c -25 +28 Thu Feb 27 00:27:00 EST 2003 stevenj * [project @ 2003-02-27 05:27:00 by stevenj] some loop splitting to touch each element of output buf only once and eliminate some conditionals...speeds up by 30-40% M ./reodft/reodft11e-r2hc-odd.c -31 +167 Wed Feb 26 17:48:26 EST 2003 stevenj * [project @ 2003-02-26 22:48:26 by stevenj] comma M ./reodft/reodft11e-r2hc-odd.c -2 +2 Wed Feb 26 17:46:17 EST 2003 stevenj * [project @ 2003-02-26 22:46:17 by stevenj] pointer to odd case M ./reodft/reodft11e-radix2.c -1 +3 Wed Feb 26 17:40:54 EST 2003 stevenj * [project @ 2003-02-26 22:40:54 by stevenj] precision -> accuracy (c.f. Kahan) M ./reodft/reodft11e-r2hc.c -2 +2 Wed Feb 26 17:36:13 EST 2003 stevenj * [project @ 2003-02-26 22:36:13 by stevenj] added time limit for wisdom generation M ./Makefile.am -1 +4 M ./libbench2/bench-user.h -1 +2 M ./libbench2/problem.c -1 +4 M ./tools/fftw-wisdom.c -16 +59 M ./tools/fftw_wisdom.1.in +6 Wed Feb 26 13:24:36 EST 2003 stevenj * [project @ 2003-02-26 18:24:36 by stevenj] caps M ./reodft/reodft11e-r2hc-odd.c -2 +2 Tue Feb 25 20:56:01 EST 2003 stevenj * [project @ 2003-02-26 01:56:01 by stevenj] another note M ./reodft/reodft11e-r2hc-odd.c -2 +3 Tue Feb 25 20:54:57 EST 2003 stevenj * [project @ 2003-02-26 01:54:57 by stevenj] note M ./reodft/reodft11e-r2hc-odd.c -4 +5 Tue Feb 25 20:42:08 EST 2003 stevenj * [project @ 2003-02-26 01:42:08 by stevenj] added new, more accurate (hopefully) reodft11 algorithms; added --disable-debug-malloc; added --impulse-accuracy-rounds=rounds flags to libbench2 for impulse-response accuracy tests A ./reodft/reodft11e-r2hc-odd.c A ./reodft/reodft11e-radix2.c M ./configure.ac +8 M ./kernel/alloc.c -2 +2 M ./kernel/ifftw.h -3 +3 M ./libbench2/bench-main.c -2 +8 M ./libbench2/bench.h -2 +2 M ./libbench2/verify-dft.c -3 +4 M ./libbench2/verify-lib.c -18 +76 M ./libbench2/verify-r2r.c -7 +39 M ./libbench2/verify-rdft2.c -3 +4 M ./libbench2/verify.c -7 +10 M ./libbench2/verify.h -4 +7 M ./reodft/Makefile.am -1 +4 M ./reodft/conf.c -2 +4 M ./reodft/reodft.h -1 +3 M ./reodft/reodft11e-r2hc-odd.c +273 M ./reodft/reodft11e-r2hc.c -2 +8 M ./reodft/reodft11e-radix2.c +506 M ./tests/bench.c -1 +1 Sun Feb 23 14:07:48 EST 2003 athena * [project @ 2003-02-23 19:07:48 by athena] fftw_wisdom.1 is in $builddir, not $srcdir M ./tools/Makefile.am -1 +1 Mon Feb 17 03:42:19 EST 2003 stevenj * [project @ 2003-02-17 08:42:19 by stevenj] pde M ./doc/fftw3.texi -2 +2 Mon Feb 17 03:40:19 EST 2003 stevenj * [project @ 2003-02-17 08:40:19 by stevenj] consistent number M ./doc/fftw3.texi -4 +4 Mon Feb 17 03:39:02 EST 2003 stevenj * [project @ 2003-02-17 08:39:02 by stevenj] started r2r doc M ./doc/fftw3.texi -1 +65 Mon Feb 17 02:31:51 EST 2003 stevenj * [project @ 2003-02-17 07:31:50 by stevenj] rfftwnd A ./doc/rfftwnd.fig A ./doc/rfftwnd.gif M ./doc/Makefile.am +11 M ./doc/fftw3.texi -26 +147 M ./doc/rfftwnd.fig +1148 M ./doc/rfftwnd.gif Sat Feb 15 17:02:07 EST 2003 stevenj * [project @ 2003-02-15 22:02:07 by stevenj] continued M ./doc/fftw3.texi -4 +32 Sat Feb 15 15:16:26 EST 2003 stevenj * [project @ 2003-02-15 20:16:26 by stevenj] started r2c/c2r docs M ./doc/fftw3.texi -8 +87 Sat Feb 15 01:12:52 EST 2003 stevenj * [project @ 2003-02-15 06:12:52 by stevenj] added r{e,o}dft11 accuracy test M ./libbench2/verify-r2r.c -23 +45 Sat Feb 15 00:42:48 EST 2003 stevenj * [project @ 2003-02-15 05:42:48 by stevenj] added more r2r accuracy checks M ./libbench2/verify-dft.c -2 +2 M ./libbench2/verify-lib.c -10 +4 M ./libbench2/verify-r2r.c -1 +127 M ./libbench2/verify-rdft2.c -2 +2 M ./libbench2/verify.h -2 +3 Fri Feb 14 19:19:54 EST 2003 athena * [project @ 2003-02-15 00:19:54 by athena] $< is a GNUism M ./tools/Makefile.am -1 +1 Wed Feb 12 21:02:16 EST 2003 stevenj * [project @ 2003-02-13 02:02:16 by stevenj] r2r test cases are in M ./TODO -1 +1 Wed Feb 12 21:01:28 EST 2003 stevenj * [project @ 2003-02-13 02:01:28 by stevenj] added vector radix to TODO M ./TODO +2 Wed Feb 12 17:21:33 EST 2003 stevenj * [project @ 2003-02-12 22:21:33 by stevenj] fixed cross-ref M ./tools/fftw_wisdom.1.in -1 +1 Wed Feb 12 17:19:56 EST 2003 stevenj * [project @ 2003-02-12 22:19:56 by stevenj] shorter synopsis M ./tools/fftw_wisdom.1.in -1 +1 Wed Feb 12 12:53:19 EST 2003 stevenj * [project @ 2003-02-12 17:53:19 by stevenj] obsolete R ./tests/debug.h Wed Feb 12 12:52:53 EST 2003 stevenj * [project @ 2003-02-12 17:52:53 by stevenj] removed old dotens R ./tests/dotens.c R ./tests/dotens2.c Wed Feb 12 12:52:16 EST 2003 stevenj * [project @ 2003-02-12 17:52:16 by stevenj] removed old verify files R ./tests/verify-dft.c R ./tests/verify-lib.c R ./tests/verify-rdft.c R ./tests/verify-reodft.c R ./tests/verify.h Wed Feb 12 12:37:17 EST 2003 stevenj * [project @ 2003-02-12 17:37:17 by stevenj] disable threads support by default M ./tools/fftw-wisdom.c -4 +11 M ./tools/fftw_wisdom.1.in -7 Wed Feb 12 11:03:28 EST 2003 athena * [project @ 2003-02-12 16:03:28 by athena] Removed old test program M ./tests/bench.c -390 Tue Feb 11 22:30:55 EST 2003 stevenj * [project @ 2003-02-12 03:30:55 by stevenj] joke M ./tools/fftw-wisdom-to-conf.in -1 +1 Tue Feb 11 22:27:44 EST 2003 stevenj * [project @ 2003-02-12 03:27:44 by stevenj] add --help and --version, to be GNU-lly correct M ./tools/fftw-wisdom-to-conf.1 +7 M ./tools/fftw-wisdom-to-conf.in +36 Tue Feb 11 22:27:18 EST 2003 stevenj * [project @ 2003-02-12 03:27:18 by stevenj] whoops M ./tools/fftw_wisdom.1.in -1 Tue Feb 11 22:17:35 EST 2003 stevenj * [project @ 2003-02-12 03:15:03 by stevenj] better help M ./tools/fftw-wisdom.c -10 +43 Tue Feb 11 21:47:35 EST 2003 stevenj * [project @ 2003-02-12 02:47:35 by stevenj] comma M ./tools/fftw-wisdom-to-conf.1 -2 +1 Tue Feb 11 21:46:12 EST 2003 stevenj * [project @ 2003-02-12 02:46:12 by stevenj] formatting M ./tools/fftw-wisdom-to-conf.1 -1 +1 Tue Feb 11 21:45:23 EST 2003 stevenj * [project @ 2003-02-12 02:45:23 by stevenj] man pages for tools A ./tools/fftw-wisdom-to-conf.1 A ./tools/fftw_wisdom.1.in M ./configure.ac +1 M ./tools/Makefile.am -1 +7 M ./tools/fftw-wisdom-to-conf.1 +85 M ./tools/fftw_wisdom.1.in +190 Tue Feb 11 19:07:12 EST 2003 stevenj * [project @ 2003-02-12 00:07:12 by stevenj] added -V M ./tools/fftw-wisdom.c +25 Tue Feb 11 18:42:17 EST 2003 stevenj * [project @ 2003-02-11 23:42:17 by stevenj] added install-wisdom target M ./Makefile.am +10 Tue Feb 11 18:23:02 EST 2003 stevenj * [project @ 2003-02-11 23:23:02 by stevenj] another note M ./NEWS -1 +2 Tue Feb 11 17:32:56 EST 2003 stevenj * [project @ 2003-02-11 22:32:56 by stevenj] started r2r accuracy tests (only three kinds covered so far) M ./libbench2/verify-dft.c -2 +2 M ./libbench2/verify-lib.c -10 +13 M ./libbench2/verify-r2r.c +103 M ./libbench2/verify-rdft2.c -2 +2 M ./libbench2/verify.c -2 +2 M ./libbench2/verify.h -1 +4 Mon Feb 10 22:04:18 EST 2003 stevenj * [project @ 2003-02-11 03:04:18 by stevenj] silence warning M ./kernel/ifftw.h -3 +3 Mon Feb 10 20:55:20 EST 2003 athena * [project @ 2003-02-11 01:55:20 by athena] gcc bug is now avoided. M ./TODO -5 Mon Feb 10 20:37:54 EST 2003 athena * [project @ 2003-02-11 01:37:54 by athena] Accuracy test A ./libbench2/mp.c M ./libbench2/Makefile.am -5 +6 M ./libbench2/bench-user.h -7 +1 M ./libbench2/mp.c +592 M ./libbench2/verify-dft.c -1 +22 M ./libbench2/verify-lib.c -16 +39 M ./libbench2/verify-r2r.c -17 +2 M ./libbench2/verify-rdft2.c -1 +23 M ./libbench2/verify.c -2 +21 M ./libbench2/verify.h +27 Mon Feb 10 07:59:57 EST 2003 athena * [project @ 2003-02-10 12:59:57 by athena] There is no point in precomputing strides for the long-double code, as multiplication by sizeof(long double) cannot be folded into the addressing mode. This change also fixes the gcc-2.95 bug that causes miscompilation of certain codelets. M ./kernel/ifftw.h -2 +2 Mon Feb 10 02:54:35 EST 2003 stevenj * [project @ 2003-02-10 07:54:35 by stevenj] added random r2r tests M ./tests/check.pl -3 +19 Mon Feb 10 02:44:58 EST 2003 stevenj * [project @ 2003-02-10 07:44:58 by stevenj] whoops, bugfix: missing stride for ro10 M ./reodft/reodft010e-r2hc.c -2 +2 Mon Feb 10 02:21:50 EST 2003 stevenj * [project @ 2003-02-10 07:21:50 by stevenj] formatting M ./api/mapflags.c -1 +2 Sun Feb 9 23:24:52 EST 2003 stevenj * [project @ 2003-02-10 04:24:52 by stevenj] flop counts for reodft M ./reodft/redft00e-r2hc.c -5 +12 M ./reodft/reodft010e-r2hc.c -3 +16 M ./reodft/reodft11e-r2hc.c -3 +10 M ./reodft/rodft00e-r2hc.c -3 +12 Sun Feb 9 23:22:15 EST 2003 stevenj * [project @ 2003-02-10 04:22:15 by stevenj] declare aligned_main M ./libbench2/bench.h -1 +2 Sun Feb 9 20:56:06 EST 2003 stevenj * [project @ 2003-02-10 01:56:06 by stevenj] corrected rader op counts M ./rdft/dht-rader.c -6 +5 M ./rdft/rader-hc2hc.c -2 +2 Sun Feb 9 20:25:32 EST 2003 stevenj * [project @ 2003-02-10 01:25:32 by stevenj] punctuation M ./TODO -1 +1 Sun Feb 9 20:25:17 EST 2003 stevenj * [project @ 2003-02-10 01:25:17 by stevenj] noted need for better estimator M ./TODO +4 Sun Feb 9 19:58:59 EST 2003 stevenj * [project @ 2003-02-10 00:58:59 by stevenj] noted F77 api fix for g77 mangling incompatibility M ./NEWS +3 Sun Feb 9 19:30:55 EST 2003 stevenj * [project @ 2003-02-10 00:30:55 by stevenj] build f77 header file of constants from fftw3.h M ./api/Makefile.am -1 +10 Sun Feb 9 19:04:53 EST 2003 stevenj * [project @ 2003-02-10 00:04:53 by stevenj] updates M ./TODO -7 +7 Sun Feb 9 19:03:34 EST 2003 stevenj * [project @ 2003-02-10 00:03:34 by stevenj] threads f77 api A ./api/x77.h A ./threads/f77api.c A ./threads/f77funcs.c M ./api/Makefile.am -1 +1 M ./api/f77api.c -12 +1 M ./api/x77.h +32 M ./threads/Makefile.am -1 +1 M ./threads/f77api.c +39 M ./threads/f77funcs.c +39 Sun Feb 9 18:54:00 EST 2003 stevenj * [project @ 2003-02-09 23:54:00 by stevenj] finished f77 serial api M ./api/f77api.c +15 M ./api/f77funcs.c +220 Sun Feb 9 18:32:26 EST 2003 stevenj * [project @ 2003-02-09 23:32:26 by stevenj] added flops, slight cleanups M ./api/f77api.c -8 +5 M ./api/f77funcs.c -21 +27 Sun Feb 9 18:11:48 EST 2003 athena * [project @ 2003-02-09 23:11:48 by athena] Oops, forgot #include M ./libbench2/aligned-main.c -1 +3 Sun Feb 9 18:08:26 EST 2003 athena * [project @ 2003-02-09 23:08:26 by athena] Removed duplication of stack-alignment code A ./libbench2/aligned-main.c M ./libbench2/Makefile.am -5 +5 M ./libbench2/aligned-main.c +81 M ./libbench2/bench-main.c -62 +2 M ./libbench2/main.c -2 +2 M ./tools/fftw-wisdom.c -61 +1 Sun Feb 9 15:48:15 EST 2003 stevenj * [project @ 2003-02-09 20:48:15 by stevenj] allow - to read problems from stdin M ./tools/fftw-wisdom.c -2 +14 Sun Feb 9 15:22:23 EST 2003 stevenj * [project @ 2003-02-09 20:22:23 by stevenj] added fftw-wisdom tool A ./tools/fftw-wisdom.c M ./tools/Makefile.am +15 M ./tools/fftw-wisdom.c +298 Sun Feb 9 15:06:38 EST 2003 stevenj * [project @ 2003-02-09 20:06:38 by stevenj] elim. warning M ./tests/bench.c -1 +1 Sun Feb 9 14:24:19 EST 2003 stevenj * [project @ 2003-02-09 19:24:19 by stevenj] destroy_input should not contaminate flags of other problems M ./tests/bench.c -1 +5 Sun Feb 9 13:06:11 EST 2003 stevenj * [project @ 2003-02-09 18:06:11 by stevenj] updated M ./ChangeLog -53 +1740 Sun Feb 9 13:01:45 EST 2003 stevenj * [project @ 2003-02-09 18:01:44 by stevenj] removed overzealous inplace check, which caused problems for rdft2 M ./dft/rank-geq2.c -12 +1 M ./rdft/rank-geq2-rdft2.c -3 +2 M ./rdft/rank-geq2.c -12 +1 Sun Feb 9 08:14:03 EST 2003 athena * [project @ 2003-02-09 13:14:03 by athena] Consistent syntax for RNK_MINFTY tensors M ./kernel/tensor.c -4 +4 Sun Feb 9 07:31:13 EST 2003 athena * [project @ 2003-02-09 12:31:13 by athena] lisply-correct tensor print. We no longer need to parse tensors. M ./kernel/tensor.c -3 +9 Sun Feb 9 03:35:56 EST 2003 stevenj * [project @ 2003-02-09 08:35:56 by stevenj] removed completed items M ./TODO -4 Sun Feb 9 03:27:56 EST 2003 stevenj * [project @ 2003-02-09 08:27:56 by stevenj] slight renaming M ./libbench2/verify-r2r.c -4 +4 Sun Feb 9 03:15:28 EST 2003 stevenj * [project @ 2003-02-09 08:15:28 by stevenj] multi-dimensional r2r verifier M ./libbench2/problem.c -2 +2 M ./libbench2/verify-r2r.c -81 +118 Sun Feb 9 02:40:22 EST 2003 stevenj * [project @ 2003-02-09 07:40:22 by stevenj] comments M ./libbench2/verify-r2r.c -1 +3 Sun Feb 9 02:38:26 EST 2003 stevenj * [project @ 2003-02-09 07:38:26 by stevenj] slight simplification M ./libbench2/verify-r2r.c -7 +2 Sun Feb 9 02:36:25 EST 2003 stevenj * [project @ 2003-02-09 07:36:25 by stevenj] added 1d r2r verifier (triple ugh) A ./libbench2/verify-r2r.c M ./libbench2/Makefile.am -2 +2 M ./libbench2/allocate.c -1 +17 M ./libbench2/bench-user.h -2 +10 M ./libbench2/mflops.c +1 M ./libbench2/problem.c -7 +68 M ./libbench2/verify-r2r.c +616 M ./libbench2/verify.c -2 +2 M ./libbench2/zero.c -1 +4 M ./tests/bench.c +101 Sat Feb 8 22:23:00 EST 2003 stevenj * [project @ 2003-02-09 03:23:00 by stevenj] added vector transforms to random tests M ./tests/check.pl -3 +11 Sat Feb 8 20:59:07 EST 2003 stevenj * [project @ 2003-02-09 01:59:07 by stevenj] whoops M ./rdft/direct2.c -2 +1 Sat Feb 8 19:52:58 EST 2003 stevenj * [project @ 2003-02-09 00:52:58 by stevenj] fixed interaction between dwims for sz/vecsz with rdft2 transforms M ./libbench2/problem.c -15 +17 Sat Feb 8 19:35:56 EST 2003 stevenj * [project @ 2003-02-09 00:35:56 by stevenj] added destroy_input flag/check M ./libbench2/bench-user.h -1 +2 M ./libbench2/problem.c -1 +3 M ./libbench2/verify-dft.c -1 +4 M ./libbench2/verify-lib.c -1 +35 M ./libbench2/verify-rdft2.c -1 +5 M ./libbench2/verify.h +2 M ./tests/bench.c +7 Sat Feb 8 19:11:58 EST 2003 stevenj * [project @ 2003-02-09 00:11:57 by stevenj] added rdft2 verifier A ./api/extract-reim.c A ./libbench2/aset.c A ./libbench2/verify-rdft2.c M ./api/Makefile.am -3 +3 R ./api/dfthelp.c M ./api/extract-reim.c +36 M ./api/plan-guru-dft-c2r.c -1 +2 M ./api/plan-many-dft-c2r.c +2 M ./libbench2/Makefile.am -5 +5 M ./libbench2/allocate.c -7 +44 M ./libbench2/aset.c +10 M ./libbench2/bench-user.h -1 +4 M ./libbench2/bench.h -1 +2 M ./libbench2/problem.c -10 +42 M ./libbench2/tensor.c -1 +47 M ./libbench2/verify-dft.c -1 +3 M ./libbench2/verify-lib.c -16 +13 M ./libbench2/verify-rdft2.c +255 M ./libbench2/verify.c -2 +2 M ./libbench2/verify.h -1 +1 M ./libbench2/zero.c -2 +8 M ./tests/bench.c -8 +189 M ./tests/check.pl -1 +7 Sat Feb 8 13:31:14 EST 2003 stevenj * [project @ 2003-02-08 18:31:14 by stevenj] an additional check for in-place case M ./rdft/rdft2-radix2.c -3 +7 Fri Feb 7 17:36:56 EST 2003 stevenj * [project @ 2003-02-07 22:36:56 by stevenj] slight fix: hc2r constraints are mostly determined by sub-plan M ./rdft/rank0-rdft2.c -6 +7 Fri Feb 7 16:28:55 EST 2003 stevenj * [project @ 2003-02-07 21:28:55 by stevenj] make radix2-dft inapplicable to in-place/split case (r == rio, iio >= rio + n/2+1 != r + 1) M ./rdft/rdft2-radix2.c -3 +22 Tue Feb 4 06:36:29 EST 2003 athena * [project @ 2003-02-04 11:36:29 by athena] Allow plnr->hook to be 0 M ./kernel/planner.c -12 +12 M ./tests/hook.c -7 +2 Tue Feb 4 03:25:36 EST 2003 stevenj * [project @ 2003-02-04 08:25:36 by stevenj] moved dft stuff into verify-dft M ./libbench2/bench-user.h -4 +2 M ./libbench2/verify-dft.c -9 +84 M ./libbench2/verify.c -84 +2 Tue Feb 4 03:25:00 EST 2003 stevenj * [project @ 2003-02-04 08:25:00 by stevenj] cruft M ./tests/hook.c -1 Tue Feb 4 03:18:28 EST 2003 stevenj * [project @ 2003-02-04 08:18:28 by stevenj] further unify libbench2 and paranoid verifiers M ./libbench2/bench-user.h -1 +6 M ./libbench2/problem.c -1 +2 M ./libbench2/verify.c -16 +21 M ./tests/bench.c -3 +5 M ./tests/hook.c -125 +58 Sun Feb 2 01:45:37 EST 2003 stevenj * [project @ 2003-02-02 06:45:37 by stevenj] typo in comment M ./api/import-wisdom-from-file.c -1 +1 Sat Feb 1 09:30:03 EST 2003 athena * [project @ 2003-02-01 14:30:03 by athena] Fixed p==2 case M ./kernel/primes.c -1 +4 Sat Feb 1 09:23:43 EST 2003 athena * [project @ 2003-02-01 14:23:43 by athena] Incorporated new find_generator by Greg Dionne. M ./kernel/primes.c -21 +35 Fri Jan 31 20:46:24 EST 2003 athena * [project @ 2003-02-01 01:46:24 by athena] Removed nonportable call to gettext() M ./libbench2/getopt.c +7 Wed Jan 29 19:03:43 EST 2003 athena * [project @ 2003-01-30 00:03:43 by athena] uintptr_t is in in openbsd M ./kernel/ifftw.h -1 +5 Wed Jan 29 15:41:56 EST 2003 athena * [project @ 2003-01-29 20:41:56 by athena] Huge speedups in wisdom I/O. M ./api/export-wisdom-to-string.c -2 +2 M ./api/export-wisdom.c -1 +1 M ./api/import-wisdom-from-file.c -4 +27 M ./api/mkprinter-file.c -5 +25 M ./kernel/debug.c -2 +2 M ./kernel/ifftw.h -2 +5 M ./kernel/planner.c -5 +9 M ./kernel/print.c -2 +7 M ./kernel/scan.c -4 +25 M ./tests/bench.c -1 +5 Tue Jan 28 19:36:51 EST 2003 athena * [project @ 2003-01-29 00:36:51 by athena] Added appropriate warning against likely future bug. M ./kernel/planner.c -1 +4 Tue Jan 28 19:00:24 EST 2003 athena * [project @ 2003-01-29 00:00:24 by athena] Don't attempt to remove bogus wisdom entries. M ./kernel/planner.c -10 +3 Tue Jan 28 18:16:24 EST 2003 athena * [project @ 2003-01-28 23:16:24 by athena] Fixed a couple of very very very nasty bugs---pointers became invalid after the hash table was relocated. M ./kernel/planner.c -19 +29 Tue Jan 28 07:34:10 EST 2003 athena * [project @ 2003-01-28 12:34:10 by athena] Read wisdom at can_do() time, otherwise wisdom is destroyed. M ./tests/bench.c -2 +5 Tue Jan 28 06:54:38 EST 2003 athena * [project @ 2003-01-28 11:54:38 by athena] More conservative inheritance of blessings M ./kernel/planner.c -19 +24 Tue Jan 28 06:50:20 EST 2003 athena * [project @ 2003-01-28 11:50:20 by athena] Print the same info as it is hashed M ./dft/problem.c -3 +4 Tue Jan 28 06:49:48 EST 2003 athena * [project @ 2003-01-28 11:49:48 by athena] Print name of executable when FAILURE M ./tests/check.pl -2 +2 Mon Jan 27 06:59:40 EST 2003 athena * [project @ 2003-01-27 11:59:40 by athena] New NO_SEARCH planner flag, which avoids searching altogether. A wisdom entry must lead to a NO_SEARCH-grade plan, or else the wisdom entry is bogus. M ./kernel/ifftw.h -5 +10 M ./kernel/planner.c -13 +30 Sun Jan 26 20:45:21 EST 2003 athena * [project @ 2003-01-27 01:45:21 by athena] Use cosl()/sinl() when appropriate M ./libbench2/verify-lib.c -3 +20 Sun Jan 26 16:29:18 EST 2003 athena * [project @ 2003-01-26 21:29:18 by athena] Use null pointers when estimating. The estimator should never time anything. M ./kernel/planner.c -1 +6 M ./libbench2/problem.c -2 +1 M ./libbench2/speed.c -1 +2 M ./libbench2/verify.c -1 +3 Sun Jan 26 15:19:01 EST 2003 stevenj * [project @ 2003-01-26 20:19:01 by stevenj] note M ./api/f77api.c -1 +1 Sun Jan 26 15:16:22 EST 2003 stevenj * [project @ 2003-01-26 20:16:22 by stevenj] support multiple mangling schemes with g77 A ./api/f77funcs.c M ./api/Makefile.am +4 M ./api/f77api.c -130 +41 M ./api/f77funcs.c +139 M ./configure.ac +6 Sun Jan 26 12:58:57 EST 2003 stevenj * [project @ 2003-01-26 17:58:57 by stevenj] fixed verbose, made random tests only use selected rank, use rank <= 4, fixed final flush_problems call M ./tests/check.pl -4 +9 Sun Jan 26 12:42:49 EST 2003 stevenj * [project @ 2003-01-26 17:42:49 by stevenj] fixed typo (count instead of maxcount) M ./tests/check.pl -2 +2 Sun Jan 26 12:12:07 EST 2003 stevenj * [project @ 2003-01-26 17:12:07 by stevenj] hypot is no longer used M ./configure.ac -1 +1 Sun Jan 26 12:07:43 EST 2003 stevenj * [project @ 2003-01-26 17:07:43 by stevenj] check for _alloca (MSVC) M ./configure.ac -1 +1 M ./kernel/ifftw.h -1 +5 Sun Jan 26 11:56:53 EST 2003 stevenj * [project @ 2003-01-26 16:56:53 by stevenj] slight fix in assert M ./kernel/alloc.c -2 +2 Sun Jan 26 11:55:39 EST 2003 athena * [project @ 2003-01-26 16:55:39 by athena] Allocate problem in all cases--- can_do may need correct pointers. M ./libbench2/problem.c -1 +2 M ./libbench2/speed.c -2 +1 M ./libbench2/verify.c -3 +1 M ./tests/bench.c -6 Sun Jan 26 11:51:27 EST 2003 athena * [project @ 2003-01-26 16:51:27 by athena] Nastier checks M ./tests/bench.c -1 +8 M ./tests/check.pl -9 +23 Sun Jan 26 11:51:16 EST 2003 athena * [project @ 2003-01-26 16:51:16 by athena] X(use_plan) is a relic. M ./kernel/ifftw.h -3 +1 M ./kernel/plan.c -11 +3 M ./kernel/planner.c -5 +2 Sun Jan 26 09:23:16 EST 2003 athena * [project @ 2003-01-26 14:23:16 by athena] Print full pathname of the bench executable, so that I don't get confused when running multiple tests for different configurations. M ./tests/Makefile.am -1 +1 Sun Jan 26 07:35:46 EST 2003 athena * [project @ 2003-01-26 12:35:46 by athena] Split done() into done() and cleanup(), in order to test multiple problems with the same planner from the command line. M ./libbench2/bench-main.c -1 +2 M ./libbench2/bench-user.h -1 +2 M ./tests/bench.c -1 +8 Sat Jan 25 20:44:49 EST 2003 athena * [project @ 2003-01-26 01:44:49 by athena] Improved readability M ./kernel/alloc.c -6 +16 Sat Jan 25 19:17:26 EST 2003 stevenj * [project @ 2003-01-26 00:17:26 by stevenj] comment M ./kernel/alloc.c -3 +3 Sat Jan 25 19:16:53 EST 2003 stevenj * [project @ 2003-01-26 00:16:53 by stevenj] added macos9 mpallocatealigned function M ./kernel/alloc.c -1 +19 Sat Jan 25 18:59:55 EST 2003 stevenj * [project @ 2003-01-25 23:59:55 by stevenj] sometimes __APPLE__ is defined instead of __MACOSX__ M ./kernel/alloc.c -2 +3 Sat Jan 25 18:54:39 EST 2003 stevenj * [project @ 2003-01-25 23:54:39 by stevenj] macos x malloc is already 16-byte aligned M ./kernel/alloc.c -2 +3 Sat Jan 25 13:38:32 EST 2003 athena * [project @ 2003-01-25 18:38:32 by athena] Include because uintptr_t is defined there on solaris. M ./kernel/ifftw.h -2 +6 Sat Jan 25 13:22:59 EST 2003 athena * [project @ 2003-01-25 18:22:59 by athena] Oops---forgot getopt_long A ./libbench2/getopt1.c M ./libbench2/Makefile.am -2 +2 M ./libbench2/getopt1.c +188 Sat Jan 25 13:17:29 EST 2003 athena * [project @ 2003-01-25 18:17:29 by athena] Include default includes when checking for uintptr_t. (Otherwise solaris breaks.) M ./configure.ac -1 +1 Sat Jan 25 12:39:52 EST 2003 athena * [project @ 2003-01-25 17:39:52 by athena] distribute check.pl M ./tests/Makefile.am +1 Sat Jan 25 12:38:34 EST 2003 athena * [project @ 2003-01-25 17:38:34 by athena] Check split format, too. M ./tests/check.pl +4 Sat Jan 25 11:48:19 EST 2003 athena * [project @ 2003-01-25 16:48:19 by athena] New tests, added make check M ./tests/Makefile.am -1 +2 M ./tests/check.pl -3 +66 Thu Jan 23 08:34:24 EST 2003 athena * [project @ 2003-01-23 13:34:24 by athena] More tests M ./tests/check.pl -17 +83 Tue Jan 21 20:32:12 EST 2003 athena * [project @ 2003-01-22 01:32:09 by athena] Deal with rnk(sz)=-infinity M ./api/mktensor-iodims.c -4 +9 M ./api/mktensor-rowmajor.c -4 +6 M ./libbench2/problem.c -5 +2 Tue Jan 21 10:07:16 EST 2003 athena * [project @ 2003-01-21 15:07:16 by athena] Crazy idea M ./TODO +3 Tue Jan 21 07:14:22 EST 2003 athena * [project @ 2003-01-21 12:14:22 by athena] Test program, still barely worthy of the name. A ./tests/check.pl Mon Jan 20 08:29:21 EST 2003 athena * [project @ 2003-01-20 13:29:21 by athena] Stylistic changes M ./libbench2/problem.c -15 +10 Mon Jan 20 07:03:38 EST 2003 athena * [project @ 2003-01-20 12:03:38 by athena] Implemented flops api A ./api/flops.c M ./api/Makefile.am -10 +10 M ./api/fftw3.h -2 +4 M ./api/flops.c +27 M ./tests/bench.c +3 Sun Jan 19 14:27:21 EST 2003 stevenj * [project @ 2003-01-19 19:27:21 by stevenj] cleanup M ./libbench2/problem.c -31 +24 Sun Jan 19 14:14:49 EST 2003 stevenj * [project @ 2003-01-19 19:14:49 by stevenj] 'v' syntax now defaults to an 'internal' (stride 1) vector, which is a more interesting case and corresponds more closely to the intuitive notion of a 'vector' transform, while '*' does the old 'external' (stride n) vector M ./libbench2/problem.c -4 +24 Sun Jan 19 13:55:35 EST 2003 stevenj * [project @ 2003-01-19 18:55:35 by stevenj] removed '/' overloading M ./libbench2/problem.c -2 +2 Sun Jan 19 13:52:09 EST 2003 stevenj * [project @ 2003-01-19 18:52:09 by stevenj] get rid of '*' and ',' synonyms for 'x' in problem parser; there's no need to clutter the namespace with syntax we never use M ./libbench2/problem.c -2 +2 Sun Jan 19 07:28:27 EST 2003 athena * [project @ 2003-01-19 12:28:27 by athena] Signed/unsigned fixes. M ./kernel/planner.c -4 +4 Sun Jan 19 07:09:54 EST 2003 athena * [project @ 2003-01-19 12:09:54 by athena] Test split arrays. M ./libbench2/bench-user.h -3 +9 M ./libbench2/verify-dft.c -14 +11 M ./libbench2/verify.c -13 +43 M ./libbench2/verify.h +1 M ./tests/bench.c -13 +50 M ./tests/hook.c -1 +4 Sat Jan 18 23:46:57 EST 2003 stevenj * [project @ 2003-01-19 04:46:57 by stevenj] clarification M ./doc/fftw3.texi -17 +16 Sat Jan 18 21:53:18 EST 2003 stevenj * [project @ 2003-01-19 02:53:18 by stevenj] caps M ./doc/fftw3.texi -2 +2 Sat Jan 18 21:52:51 EST 2003 stevenj * [project @ 2003-01-19 02:52:51 by stevenj] brackets M ./doc/fftw3.texi -2 +2 Sat Jan 18 21:52:32 EST 2003 stevenj * [project @ 2003-01-19 02:52:32 by stevenj] quote M ./doc/fftw3.texi -1 +7 Sat Jan 18 20:53:11 EST 2003 stevenj * [project @ 2003-01-19 01:53:11 by stevenj] referencing M ./doc/fftw3.texi -4 +7 Sat Jan 18 20:33:28 EST 2003 stevenj * [project @ 2003-01-19 01:33:28 by stevenj] fix M ./doc/fftw3.texi -2 +2 Sat Jan 18 20:31:41 EST 2003 stevenj * [project @ 2003-01-19 01:31:41 by stevenj] slight change M ./doc/fftw3.texi -8 +8 Sat Jan 18 20:31:22 EST 2003 athena * [project @ 2003-01-19 01:31:22 by athena] Print errors when --verify. M ./libbench2/verify-dft.c -7 +12 M ./libbench2/verify-lib.c -30 +38 M ./libbench2/verify.h -12 +14 M ./tests/bench.c -5 +5 Sat Jan 18 20:30:27 EST 2003 stevenj * [project @ 2003-01-19 01:30:27 by stevenj] improved description, noted that FFTW_ESTIMATE does not destroy arrays M ./doc/fftw3.texi -7 +8 Sat Jan 18 20:23:12 EST 2003 stevenj * [project @ 2003-01-19 01:23:12 by stevenj] FFTW_DEFAULTS isn't really needed M ./api/fftw3.h -2 +1 Sat Jan 18 20:21:09 EST 2003 stevenj * [project @ 2003-01-19 01:21:09 by stevenj] added FFTW_MEASURE synonym for FFTW_DEFAULTS M ./api/fftw3.h -1 +2 M ./doc/fftw3.texi -6 +6 Sat Jan 18 20:18:29 EST 2003 stevenj * [project @ 2003-01-19 01:18:29 by stevenj] slight change M ./kernel/alloc.c -2 +2 Sat Jan 18 20:16:08 EST 2003 athena * [project @ 2003-01-19 01:16:08 by athena] Clearer name M ./tests/bench.c -2 +2 Sat Jan 18 20:13:14 EST 2003 athena * [project @ 2003-01-19 01:13:14 by athena] Completed dft api test M ./api/fftw3.h -12 +12 M ./libbench2/tensor.c -3 +3 M ./tests/bench.c -8 +65 Sat Jan 18 20:07:33 EST 2003 stevenj * [project @ 2003-01-19 01:07:33 by stevenj] index M ./doc/fftw3.texi -1 +2 Sat Jan 18 20:05:50 EST 2003 stevenj * [project @ 2003-01-19 01:05:50 by stevenj] fix M ./doc/fftw3.texi -2 +2 Sat Jan 18 20:04:11 EST 2003 stevenj * [project @ 2003-01-19 01:04:11 by stevenj] parallel structure M ./doc/fftw3.texi -2 +2 Sat Jan 18 20:03:18 EST 2003 stevenj * [project @ 2003-01-19 01:03:18 by stevenj] fix M ./doc/fftw3.texi -4 +4 Sat Jan 18 20:00:24 EST 2003 stevenj * [project @ 2003-01-19 01:00:24 by stevenj] joke M ./doc/fftw3.texi -5 +6 Sat Jan 18 19:59:28 EST 2003 stevenj * [project @ 2003-01-19 00:59:28 by stevenj] recommendation to read tutorial in-order M ./doc/fftw3.texi -1 +6 Sat Jan 18 19:54:55 EST 2003 stevenj * [project @ 2003-01-19 00:54:55 by stevenj] expanded outline M ./doc/fftw3.texi -7 +42 Sat Jan 18 19:35:52 EST 2003 stevenj * [project @ 2003-01-19 00:35:52 by stevenj] clarification M ./doc/fftw3.texi -3 +5 Sat Jan 18 19:17:27 EST 2003 stevenj * [project @ 2003-01-19 00:17:27 by stevenj] draft complex-dft tutorial M ./doc/fftw3.texi -25 +183 Sat Jan 18 17:27:15 EST 2003 athena * [project @ 2003-01-18 22:27:15 by athena] Paranoid mode is back. Fixed dwim to do what I mean. A ./tests/hook.c M ./libbench2/allocate.c -2 +1 M ./libbench2/bench-main.c -2 +1 M ./libbench2/bench-user.h -8 +1 M ./libbench2/bench.h -2 +1 M ./libbench2/can-do.c -2 +1 M ./libbench2/dotens2.c -2 +2 M ./libbench2/info.c -2 +1 M ./libbench2/problem.c -18 +21 M ./libbench2/report.c -2 +1 M ./libbench2/speed.c -2 +1 M ./libbench2/timer.c -2 +1 M ./libbench2/util.c -1 M ./libbench2/verify.c -4 +3 M ./libbench2/verify.h +7 M ./libbench2/zero.c -2 +1 M ./tests/Makefile.am -1 +1 M ./tests/bench.c +4 M ./tests/hook.c +191 Sat Jan 18 17:13:51 EST 2003 stevenj * [project @ 2003-01-18 22:13:51 by stevenj] started tut. M ./doc/fftw3.texi -14 +46 Sat Jan 18 16:13:15 EST 2003 athena * [project @ 2003-01-18 21:13:15 by athena] Great renaming, so that we can include both bench-user.h and ifftw.h to implement the paranoid-mode hook. M ./libbench2/allocate.c -5 +5 M ./libbench2/bench-user.h -30 +28 M ./libbench2/bench.h -12 +12 M ./libbench2/can-do.c -2 +2 M ./libbench2/dotens2.c -3 +3 M ./libbench2/mflops.c -1 +1 M ./libbench2/problem.c -12 +12 M ./libbench2/report.c -10 +10 M ./libbench2/speed.c -2 +2 M ./libbench2/tensor.c -22 +22 M ./libbench2/verify-dft.c -2 +2 M ./libbench2/verify-lib.c -6 +6 M ./libbench2/verify.c -9 +9 M ./libbench2/verify.h -2 +2 M ./libbench2/zero.c -2 +2 M ./tests/bench.c -29 +29 Sat Jan 18 15:41:18 EST 2003 athena * [project @ 2003-01-18 20:41:18 by athena] Trying to tweak the verifier so that I can use it in bench.c for paranoid mode M ./libbench2/bench-user.h -1 +23 M ./libbench2/problem.c -3 +3 M ./libbench2/verify-dft.c -66 +9 M ./libbench2/verify-lib.c -21 +18 M ./libbench2/verify.c -5 +62 M ./libbench2/verify.h -18 +3 Sat Jan 18 10:24:05 EST 2003 athena * [project @ 2003-01-18 15:24:05 by athena] Added stride_factor for complex arrays. M ./tests/bench.c -5 +5 Sat Jan 18 10:02:11 EST 2003 athena * [project @ 2003-01-18 15:02:11 by athena] can_do now calls the planner. M ./tests/bench.c -1 +6 Sat Jan 18 09:59:24 EST 2003 athena * [project @ 2003-01-18 14:59:24 by athena] Call guru api in bench.c M ./api/plan-guru-dft.c -2 +1 M ./tests/bench.c -3 +47 Sat Jan 18 08:17:23 EST 2003 athena * [project @ 2003-01-18 13:17:23 by athena] Fixed prototype. M ./libbench2/bench.h -1 +3 M ./libbench2/zero.c -2 +2 Sat Jan 18 08:14:48 EST 2003 athena * [project @ 2003-01-18 13:14:48 by athena] Attempt to make the signed/unsigned use of flags consistent. M ./api/api.h -2 +2 M ./api/apiplan.c -1 +1 M ./api/fftw3.h -25 +25 M ./api/mapflags.c -8 +10 M ./api/plan-dft-1d.c -1 +1 M ./api/plan-dft-2d.c -1 +1 M ./api/plan-dft-3d.c -1 +1 M ./api/plan-dft-c2r-1d.c -1 +1 M ./api/plan-dft-c2r-2d.c -1 +1 M ./api/plan-dft-c2r-3d.c -1 +1 M ./api/plan-dft-c2r.c -1 +1 M ./api/plan-dft-r2c-1d.c -1 +1 M ./api/plan-dft-r2c-2d.c -1 +1 M ./api/plan-dft-r2c-3d.c -1 +1 M ./api/plan-dft-r2c.c -1 +1 M ./api/plan-dft.c -1 +1 M ./api/plan-guru-dft-c2r.c -1 +1 M ./api/plan-guru-dft-r2c.c -1 +1 M ./api/plan-guru-dft.c -1 +1 M ./api/plan-guru-r2r.c -1 +1 M ./api/plan-many-dft-c2r.c -1 +1 M ./api/plan-many-dft-r2c.c -1 +1 M ./api/plan-many-dft.c -1 +1 M ./api/plan-many-r2r.c -1 +1 M ./api/plan-r2r-1d.c -1 +1 M ./api/plan-r2r-2d.c -1 +1 M ./api/plan-r2r-3d.c -1 +1 M ./api/plan-r2r.c -1 +1 M ./kernel/ifftw.h -2 +2 Sat Jan 18 08:03:07 EST 2003 athena * [project @ 2003-01-18 13:03:07 by athena] Implemented useropt. A ./libbench2/useropt.c M ./libbench2/Makefile.am -2 +3 M ./libbench2/bench-main.c -1 +5 M ./libbench2/bench-user.h -1 +2 M ./libbench2/useropt.c +30 M ./tests/bench.c -7 +24 Sat Jan 18 08:02:05 EST 2003 athena * [project @ 2003-01-18 13:02:05 by athena] The first map_flags pass must be transitive, i.e., always use the latest flags value as opposed to the original value. (I think.) M ./api/mapflags.c -11 +12 Sat Jan 18 07:20:19 EST 2003 athena * [project @ 2003-01-18 12:20:18 by athena] Started working on verifier A ./libbench2/dotens2.c A ./libbench2/verify-dft.c A ./libbench2/verify-lib.c A ./libbench2/verify.h M ./libbench2/Makefile.am -3 +3 M ./libbench2/bench-user.h -1 +4 M ./libbench2/dotens2.c +55 M ./libbench2/tensor.c -1 +8 M ./libbench2/verify-dft.c +119 M ./libbench2/verify-lib.c +399 M ./libbench2/verify.c -2 +9 M ./libbench2/verify.h +65 M ./tests/Makefile.am -4 +1 Fri Jan 17 14:53:28 EST 2003 stevenj * [project @ 2003-01-17 19:53:28 by stevenj] added X(threads_cleanup) M ./api/fftw3.h -1 +2 M ./threads/api.c +9 M ./threads/threads.c -6 +19 M ./threads/threads.h -1 +2 Fri Jan 17 10:35:56 EST 2003 athena * [project @ 2003-01-17 15:35:56 by athena] Use C style for upper and lower array bounds. Free tensors properly. M ./libbench2/allocate.c -3 +5 M ./libbench2/tensor.c -2 +2 Fri Jan 17 08:50:42 EST 2003 athena * [project @ 2003-01-17 13:50:42 by athena] Fixed ambiguous syntax M ./libbench2/problem.c -2 +2 Fri Jan 17 08:20:57 EST 2003 athena * [project @ 2003-01-17 13:20:57 by athena] Parse minus sign, bugfixes M ./libbench2/problem.c -4 +17 Fri Jan 17 08:11:56 EST 2003 athena * [project @ 2003-01-17 13:11:56 by athena] Skeleton libbench2 implemented (probably still buggy) A ./libbench2/ A ./libbench2/Makefile.am A ./libbench2/allocate.c A ./libbench2/bench-main.c A ./libbench2/bench-user.h A ./libbench2/bench.h A ./libbench2/can-do.c A ./libbench2/caset.c A ./libbench2/getopt-utils.c A ./libbench2/getopt.c A ./libbench2/getopt.h A ./libbench2/info.c A ./libbench2/main.c A ./libbench2/mflops.c A ./libbench2/ovtpvt.c A ./libbench2/pow2.c A ./libbench2/problem.c A ./libbench2/report.c A ./libbench2/speed.c A ./libbench2/tensor.c A ./libbench2/timer.c A ./libbench2/util.c A ./libbench2/verify.c A ./libbench2/zero.c M ./Makefile.am -1 +1 M ./configure.ac -1 +1 M ./libbench2/Makefile.am +13 M ./libbench2/allocate.c +55 M ./libbench2/bench-main.c +250 M ./libbench2/bench-user.h +206 M ./libbench2/bench.h +67 M ./libbench2/can-do.c +33 M ./libbench2/caset.c +12 M ./libbench2/getopt-utils.c +104 M ./libbench2/getopt.c +1062 M ./libbench2/getopt.h +180 M ./libbench2/info.c +59 M ./libbench2/main.c +40 M ./libbench2/mflops.c +25 M ./libbench2/ovtpvt.c +13 M ./libbench2/pow2.c +6 M ./libbench2/problem.c +172 M ./libbench2/report.c +159 M ./libbench2/speed.c +73 M ./libbench2/tensor.c +173 M ./libbench2/timer.c +241 M ./libbench2/util.c +216 M ./libbench2/verify.c +53 M ./libbench2/zero.c +36 M ./tests/Makefile.am -2 +2 M ./tests/bench.c -56 +93 Fri Jan 17 04:23:37 EST 2003 athena * [project @ 2003-01-17 09:23:37 by athena] Formatting M ./kernel/tensor4.c -10 +10 Fri Jan 17 03:15:24 EST 2003 fftw * [project @ 2003-01-17 08:15:24 by fftw] slight updates M ./doc/fftw3.texi -27 +34 Fri Jan 17 01:44:44 EST 2003 stevenj * [project @ 2003-01-17 06:44:44 by stevenj] eliminated obsolete uimin/uimax M ./dft/vrank-geq1.c -2 +2 M ./kernel/buffered.c -2 +2 M ./kernel/ifftw.h -3 +1 M ./kernel/minmax.c -11 +1 M ./kernel/tensor4.c -4 +4 M ./rdft/buffered2.c -4 +4 M ./rdft/rdft2-inplace-strides.c -2 +2 M ./rdft/rdft2-tensor-max-index.c -4 +4 M ./rdft/vrank-geq1-rdft2.c -2 +2 M ./rdft/vrank-geq1.c -2 +2 Fri Jan 17 01:40:10 EST 2003 stevenj * [project @ 2003-01-17 06:40:10 by stevenj] threads needs to have its own library, lest all programs linking to libfftw3.so need -lpthread A ./threads/api.c M ./Makefile.am -1 M ./api/Makefile.am -4 +3 M ./api/configure.c -2 M ./api/fftw3.h -2 +3 R ./api/plan-with-nthreads.c M ./tests/Makefile.am -1 +8 M ./tests/bench.c +5 M ./threads/Makefile.am -9 +6 M ./threads/api.c +56 M ./threads/threads.c -1 +1 Thu Jan 16 19:53:46 EST 2003 stevenj * [project @ 2003-01-17 00:53:46 by stevenj] whoops M ./api/f77api.c -2 +2 Thu Jan 16 19:53:30 EST 2003 stevenj * [project @ 2003-01-17 00:53:30 by stevenj] better name M ./api/f77api.c -2 +2 Thu Jan 16 19:52:36 EST 2003 stevenj * [project @ 2003-01-17 00:52:36 by stevenj] added more functions M ./api/f77api.c +68 Thu Jan 16 16:57:06 EST 2003 stevenj * [project @ 2003-01-16 21:57:06 by stevenj] if 'long' is big enough, use it for mulmod in preference to 'long long' M ./kernel/ifftw.h -1 +3 Thu Jan 16 14:53:41 EST 2003 stevenj * [project @ 2003-01-16 19:53:41 by stevenj] use uintptr_t for pointer alignment arithmetic M ./configure.ac -2 +9 M ./kernel/align.c -2 +2 M ./kernel/ifftw.h -5 +22 Thu Jan 16 07:58:28 EST 2003 athena * [project @ 2003-01-16 12:58:28 by athena] More signed/unsigned cleanup M ./kernel/planner.c -10 +10 M ./kernel/print.c -3 +3 M ./kernel/tensor.c -2 +2 M ./kernel/twiddle.c -2 +2 M ./rdft/problem.c -2 +2 Thu Jan 16 07:57:40 EST 2003 athena * [project @ 2003-01-16 12:57:40 by athena] null function pointers are technically nonportable M ./kernel/solvtab.c -2 +2 Thu Jan 16 07:17:45 EST 2003 athena * [project @ 2003-01-16 12:17:45 by athena] Free short_options M ./libbench/bench-main.c -2 +3 Thu Jan 16 05:48:30 EST 2003 athena * [project @ 2003-01-16 10:48:30 by athena] Oops, forgot STACK_FREE M ./kernel/alloc.c -15 +17 M ./kernel/ifftw.h -2 +3 M ./tests/bench.c -1 +1 Thu Jan 16 05:40:39 EST 2003 athena * [project @ 2003-01-16 10:40:39 by athena] Do not require memalign() unless HAVE_SIMD M ./kernel/alloc.c -6 +15 M ./kernel/ifftw.h -13 +18 Thu Jan 16 01:03:31 EST 2003 stevenj * [project @ 2003-01-16 06:03:31 by stevenj] MS VC++ _aligned_malloc M ./kernel/alloc.c -1 +7 Thu Jan 16 00:44:45 EST 2003 stevenj * [project @ 2003-01-16 05:44:45 by stevenj] added api fftw_malloc/free M ./api/fftw3.h -2 +5 M ./kernel/alloc.c -2 +10 Thu Jan 16 00:43:48 EST 2003 stevenj * [project @ 2003-01-16 05:43:48 by stevenj] silence warning M ./api/map-r2r-kind.c -1 +1 Wed Jan 15 22:39:04 EST 2003 stevenj * [project @ 2003-01-16 03:39:04 by stevenj] send error output to stderr M ./tools/fftw-wisdom-to-conf.in -1 +1 Wed Jan 15 13:20:35 EST 2003 athena * [project @ 2003-01-15 18:20:35 by athena] Pure paranoia. M ./kernel/tensor7.c -3 +6 Wed Jan 15 06:51:34 EST 2003 athena * [project @ 2003-01-15 11:51:34 by athena] Fixed formatting that was messed up by the conversion uint->int. Ensure that iodims etc are kosher. A ./kernel/tensor9.c M ./api/api.h -13 +20 M ./api/apiplan.c -9 +9 M ./api/configure.c -5 +5 M ./api/dfthelp.c -1 +1 M ./api/execute-dft-c2r.c -1 +1 M ./api/execute-dft-r2c.c -1 +1 M ./api/execute-dft.c -1 +1 M ./api/execute-r2r.c -1 +1 M ./api/execute.c -1 +1 M ./api/export-wisdom-to-file.c -4 +4 M ./api/export-wisdom-to-string.c -6 +6 M ./api/export-wisdom.c -6 +6 M ./api/f77api.c -58 +46 M ./api/fftw3.h -14 +11 M ./api/forget-wisdom.c -2 +2 M ./api/import-system-wisdom.c -2 +2 M ./api/import-wisdom-from-file.c -5 +5 M ./api/import-wisdom-from-string.c -4 +4 M ./api/import-wisdom.c -6 +6 M ./api/map-r2r-kind.c -37 +17 M ./api/mapflags.c -4 +4 M ./api/mkprinter-file.c -2 +2 M ./api/mktensor-iodims.c -2 +23 M ./api/mktensor-rowmajor.c -4 +20 M ./api/plan-dft-1d.c -2 +2 M ./api/plan-dft-2d.c -2 +2 M ./api/plan-dft-3d.c -3 +3 M ./api/plan-dft-c2r-1d.c -2 +2 M ./api/plan-dft-c2r-2d.c -2 +2 M ./api/plan-dft-c2r-3d.c -3 +3 M ./api/plan-dft-c2r.c -3 +3 M ./api/plan-dft-r2c-1d.c -2 +2 M ./api/plan-dft-r2c-2d.c -2 +2 M ./api/plan-dft-r2c-3d.c -3 +3 M ./api/plan-dft-r2c.c -3 +5 M ./api/plan-dft.c -4 +6 M ./api/plan-guru-dft-c2r.c -10 +10 M ./api/plan-guru-dft-r2c.c -10 +11 M ./api/plan-guru-dft.c -10 +11 M ./api/plan-guru-r2r.c -14 +16 M ./api/plan-many-dft-c2r.c -34 +21 M ./api/plan-many-dft-r2c.c -34 +21 M ./api/plan-many-dft.c -18 +19 M ./api/plan-many-r2r.c -22 +21 M ./api/plan-r2r-1d.c -2 +2 M ./api/plan-r2r-2d.c -3 +3 M ./api/plan-r2r-3d.c -4 +4 M ./api/plan-r2r.c -4 +4 M ./api/plan-with-nthreads.c -3 +3 M ./api/print-plan.c -3 +3 M ./api/rdft2-pad.c -2 +2 M ./api/the-planner.c -5 +5 M ./dft/buffered.c -2 +2 M ./dft/ct.c -2 +2 M ./dft/direct.c -2 +2 M ./dft/generic.c -1 +1 M ./dft/problem.c -4 +5 M ./dft/rader.c -1 +1 M ./dft/vrank-geq1.c -2 +2 M ./dft/vrank2-transpose.c -2 +2 M ./dft/vrank3-transpose.c -2 +2 M ./kernel/Makefile.am -2 +2 M ./kernel/alloc.c -2 +2 M ./kernel/ifftw.h -1 +2 M ./kernel/planner.c -10 +10 M ./kernel/print.c -3 +3 M ./kernel/tensor.c -2 +4 M ./kernel/tensor9.c +37 M ./rdft/buffered.c -2 +2 M ./rdft/buffered2.c -2 +2 M ./rdft/dft-r2hc.c -2 +2 M ./rdft/dht-r2hc.c -2 +2 M ./rdft/dht-rader.c -1 +1 M ./rdft/direct.c -2 +2 M ./rdft/direct2.c -2 +2 M ./rdft/generic.c -1 +1 M ./rdft/hc2hc.c -2 +2 M ./rdft/problem.c -2 +6 M ./rdft/problem2.c -2 +5 M ./rdft/rader-hc2hc.c -1 +1 M ./rdft/rdft-dht.c -2 +2 M ./rdft/rdft2-radix2.c -2 +2 M ./rdft/vrank-geq1-rdft2.c -2 +2 M ./rdft/vrank-geq1.c -2 +2 M ./rdft/vrank2-transpose.c -2 +2 M ./rdft/vrank3-transpose.c -2 +2 M ./reodft/redft00e-r2hc.c -2 +2 M ./reodft/reodft010e-r2hc.c -2 +2 M ./reodft/reodft11e-r2hc.c -2 +2 M ./reodft/rodft00e-r2hc.c -2 +2 M ./threads/dft-vrank-geq1.c -2 +2 M ./threads/rdft-vrank-geq1.c -2 +2 M ./threads/vrank-geq1-rdft2.c -2 +2 Wed Jan 15 01:32:18 EST 2003 stevenj * [project @ 2003-01-15 06:32:18 by stevenj] added version stamp M ./tools/fftw-wisdom-to-conf.in -1 +2 Wed Jan 15 01:28:20 EST 2003 stevenj * [project @ 2003-01-15 06:28:20 by stevenj] added warning M ./tools/fftw-wisdom-to-conf.in +1 Wed Jan 15 01:23:25 EST 2003 stevenj * [project @ 2003-01-15 06:23:25 by stevenj] add fftw-wisdom-to-conf to BUILT_SOURCES M ./tools/Makefile.am -1 +1 Wed Jan 15 01:09:29 EST 2003 stevenj * [project @ 2003-01-15 06:09:29 by stevenj] added const M ./tools/fftw-wisdom-to-conf.in -2 +2 Wed Jan 15 01:04:10 EST 2003 stevenj * [project @ 2003-01-15 06:04:10 by stevenj] added wisdom-to-conf A ./tools/ A ./tools/Makefile.am A ./tools/fftw-wisdom-to-conf.in M ./Makefile.am -1 +1 M ./configure.ac +3 M ./tools/Makefile.am +3 M ./tools/fftw-wisdom-to-conf.in +37 Wed Jan 15 00:23:36 EST 2003 stevenj * [project @ 2003-01-15 05:23:36 by stevenj] include type prefix in wisdom preamble M ./kernel/planner.c -2 +2 Wed Jan 15 00:02:31 EST 2003 stevenj * [project @ 2003-01-15 05:02:31 by stevenj] updates M ./TODO -4 +2 Tue Jan 14 23:59:26 EST 2003 stevenj * [project @ 2003-01-15 04:59:26 by stevenj] check the_plan before printing M ./tests/bench.c -2 +3 Tue Jan 14 21:10:25 EST 2003 athena * [project @ 2003-01-15 02:10:25 by athena] Eliminated those unsigned values that would break LP64 machines. M ./ChangeLog -1 +1 M ./api/api.h -16 +16 M ./api/apiplan.c -19 +19 M ./api/configure.c -5 +5 M ./api/dfthelp.c -5 +5 M ./api/execute-dft-c2r.c -1 +1 M ./api/execute-dft-r2c.c -1 +1 M ./api/execute-dft.c -1 +1 M ./api/execute-r2r.c -1 +1 M ./api/execute.c -1 +1 M ./api/export-wisdom-to-file.c -4 +4 M ./api/export-wisdom-to-string.c -13 +13 M ./api/export-wisdom.c -7 +7 M ./api/f77api.c -71 +69 M ./api/fftw3.h -134 +136 M ./api/forget-wisdom.c -2 +2 M ./api/import-system-wisdom.c -8 +8 M ./api/import-wisdom-from-file.c -6 +6 M ./api/import-wisdom-from-string.c -6 +6 M ./api/import-wisdom.c -9 +11 M ./api/map-r2r-kind.c -17 +40 M ./api/mapflags.c -18 +18 M ./api/mkprinter-file.c -3 +3 M ./api/mktensor-iodims.c -9 +9 M ./api/mktensor-rowmajor.c -6 +5 M ./api/plan-dft-1d.c -3 +2 M ./api/plan-dft-2d.c -5 +5 M ./api/plan-dft-3d.c -5 +7 M ./api/plan-dft-c2r-1d.c -2 +2 M ./api/plan-dft-c2r-2d.c -5 +5 M ./api/plan-dft-c2r-3d.c -5 +7 M ./api/plan-dft-c2r.c -6 +3 M ./api/plan-dft-r2c-1d.c -2 +2 M ./api/plan-dft-r2c-2d.c -5 +5 M ./api/plan-dft-r2c-3d.c -5 +7 M ./api/plan-dft-r2c.c -6 +3 M ./api/plan-dft.c -6 +4 M ./api/plan-guru-dft-c2r.c -10 +10 M ./api/plan-guru-dft-r2c.c -10 +10 M ./api/plan-guru-dft.c -10 +10 M ./api/plan-guru-r2r.c -13 +14 M ./api/plan-many-dft-c2r.c -23 +36 M ./api/plan-many-dft-r2c.c -23 +36 M ./api/plan-many-dft.c -17 +18 M ./api/plan-many-r2r.c -18 +22 M ./api/plan-r2r-1d.c -3 +2 M ./api/plan-r2r-2d.c -7 +8 M ./api/plan-r2r-3d.c -8 +11 M ./api/plan-r2r.c -3 +4 M ./api/plan-with-nthreads.c -7 +7 M ./api/print-plan.c -3 +3 M ./api/rdft2-pad.c -10 +9 M ./api/the-planner.c -7 +7 M ./configure.ac -9 +3 M ./dft/buffered.c -11 +11 M ./dft/codelet-dft.h -11 +11 M ./dft/codelets/n.c -1 +1 M ./dft/codelets/t.c -1 +1 M ./dft/ct-dif.c -4 +4 M ./dft/ct-dit.c -4 +4 M ./dft/ct-ditbuf.c -6 +6 M ./dft/ct-ditf.c -2 +2 M ./dft/ct.c -5 +5 M ./dft/ct.h -2 +2 M ./dft/dft.h -2 +2 M ./dft/direct.c -3 +3 M ./dft/generic.c -5 +5 M ./dft/indirect.c -2 +2 M ./dft/k7/k7.c -9 +9 M ./dft/problem.c -3 +3 M ./dft/rader-omega.c -2 +2 M ./dft/rader.c -13 +13 M ./dft/rank-geq2.c -9 +9 M ./dft/rank0.c -7 +7 M ./dft/simd/n1b.c -1 +1 M ./dft/simd/n1f.c -1 +1 M ./dft/simd/t1b.c -1 +1 M ./dft/simd/t1f.c -1 +1 M ./dft/vrank-geq1.c -11 +11 M ./dft/vrank2-transpose.c -4 +4 M ./dft/vrank3-transpose.c -10 +10 M ./dft/zero.c -3 +3 M ./genfft/gen_hc2hc.ml -4 +4 M ./genfft/gen_hc2r.ml -4 +4 M ./genfft/gen_notw.ml -4 +4 M ./genfft/gen_notw_c.ml -4 +4 M ./genfft/gen_r2hc.ml -4 +4 M ./genfft/gen_twiddle.ml -4 +4 M ./genfft/gen_twiddle_c.ml -4 +4 M ./genfft/gen_twidsq.ml -4 +4 M ./kernel/align.c -5 +5 M ./kernel/buffered.c -3 +3 M ./kernel/ct.c -1 +1 M ./kernel/hash.c -3 +3 M ./kernel/iabs.c -3 +3 M ./kernel/ifftw.h -71 +70 M ./kernel/md5-1.c -3 +3 M ./kernel/md5.c -4 +4 M ./kernel/minmax.c -3 +3 M ./kernel/ops.c -4 +4 M ./kernel/pickdim.c -6 +6 M ./kernel/planner.c -32 +32 M ./kernel/primes.c -14 +14 M ./kernel/print.c -5 +5 M ./kernel/rader.c -3 +3 M ./kernel/scan.c -3 +3 M ./kernel/tensor.c -9 +9 M ./kernel/tensor1.c -2 +2 M ./kernel/tensor2.c -3 +3 M ./kernel/tensor4.c -10 +10 M ./kernel/tensor5.c -6 +6 M ./kernel/tensor7.c -4 +4 M ./kernel/trig.c -4 +4 M ./kernel/twiddle.c -14 +14 M ./libbench/acopy.c -2 +2 M ./libbench/allocate.c -2 +2 M ./libbench/ascale.c -2 +2 M ./libbench/aset.c -2 +2 M ./libbench/bench-user.h -23 +23 M ./libbench/bench.h -2 +2 M ./libbench/caadd.c -2 +2 M ./libbench/cacopy.c -2 +2 M ./libbench/cascale.c -2 +2 M ./libbench/caset.c -2 +2 M ./libbench/casub.c -2 +2 M ./libbench/copy-c2h-1d-fftpack.c -1 +1 M ./libbench/copy-c2h-1d-halfcomplex.c -1 +1 M ./libbench/copy-c2h-1d-packed.c -2 +2 M ./libbench/copy-c2h-1d-unpacked-ri.c -1 +1 M ./libbench/copy-c2h-unpacked.c -1 +1 M ./libbench/copy-c2r-packed.c -2 +2 M ./libbench/copy-c2r-unpacked.c -1 +1 M ./libbench/copy-c2ri.c -2 +2 M ./libbench/copy-h2c-1d-fftpack.c -1 +1 M ./libbench/copy-h2c-1d-halfcomplex.c -1 +1 M ./libbench/copy-h2c-1d-packed.c -2 +2 M ./libbench/copy-h2c-1d-unpacked-ri.c -1 +1 M ./libbench/copy-h2c-unpacked.c -9 +9 M ./libbench/copy-r2c-packed.c -2 +2 M ./libbench/copy-r2c-unpacked.c -1 +1 M ./libbench/copy-ri2c.c -2 +2 M ./libbench/getopt-utils.c -2 +2 M ./libbench/getopt.c -3 +3 M ./libbench/log2.c -2 +2 M ./libbench/mp.c -20 +20 M ./libbench/pow2.c -1 +1 M ./libbench/prime.c -4 +4 M ./libbench/problem.c -2 +2 M ./libbench/timer.c -2 +2 M ./libbench/verify.c -36 +36 M ./rdft/buffered.c -10 +10 M ./rdft/buffered2.c -22 +22 M ./rdft/codelet-rdft.h -17 +17 M ./rdft/codelets/hc2r.c -1 +1 M ./rdft/codelets/hfb.c -1 +1 M ./rdft/codelets/r2hc.c -1 +1 M ./rdft/dft-r2hc.c -4 +4 M ./rdft/dht-r2hc.c -3 +3 M ./rdft/dht-rader.c -6 +6 M ./rdft/direct.c -6 +6 M ./rdft/direct2.c -5 +5 M ./rdft/generic.c -8 +8 M ./rdft/hc2hc-buf.c -7 +7 M ./rdft/hc2hc-dif.c -4 +4 M ./rdft/hc2hc-dit.c -4 +4 M ./rdft/hc2hc.c -5 +5 M ./rdft/hc2hc.h -2 +2 M ./rdft/indirect.c -2 +2 M ./rdft/problem.c -11 +11 M ./rdft/problem2.c -3 +3 M ./rdft/rader-hc2hc.c -12 +12 M ./rdft/rank-geq2-rdft2.c -9 +9 M ./rdft/rank-geq2.c -9 +9 M ./rdft/rank0-rdft2.c -4 +4 M ./rdft/rank0.c -5 +5 M ./rdft/rdft-dht.c -5 +5 M ./rdft/rdft.h -4 +4 M ./rdft/rdft2-inplace-strides.c -4 +4 M ./rdft/rdft2-radix2.c -15 +15 M ./rdft/rdft2-tensor-max-index.c -4 +4 M ./rdft/vrank-geq1-rdft2.c -12 +12 M ./rdft/vrank-geq1.c -11 +11 M ./rdft/vrank2-transpose.c -4 +4 M ./rdft/vrank3-transpose.c -10 +10 M ./reodft/redft00e-r2hc.c -7 +7 M ./reodft/reodft010e-r2hc.c -16 +16 M ./reodft/reodft11e-r2hc.c -10 +10 M ./reodft/rodft00e-r2hc.c -7 +7 M ./simd/3dnow.c -5 +5 M ./simd/sse.c -3 +3 M ./simd/sse2.c -3 +3 M ./tests/bench.c -13 +7 M ./tests/dotens.c -3 +3 M ./tests/dotens2.c -3 +3 M ./tests/trigtest.c -7 +7 M ./tests/verify-dft.c -4 +4 M ./tests/verify-lib.c -41 +41 M ./tests/verify-rdft.c -13 +13 M ./tests/verify-reodft.c -48 +48 M ./tests/verify.h -15 +15 M ./threads/ct-dit.c -5 +5 M ./threads/dft-vrank-geq1.c -15 +15 M ./threads/hc2hc-dif.c -6 +6 M ./threads/hc2hc-dit.c -6 +6 M ./threads/rdft-vrank-geq1.c -15 +15 M ./threads/threads.c -4 +4 M ./threads/threads.h -3 +3 M ./threads/vrank-geq1-rdft2.c -15 +15 Tue Jan 14 15:14:29 EST 2003 stevenj * [project @ 2003-01-14 20:14:29 by stevenj] comments M ./kernel/primes.c -4 +4 Tue Jan 14 08:00:08 EST 2003 athena * [project @ 2003-01-14 13:00:08 by athena] Oops M ./dft/generic.c -1 +1 M ./rdft/generic.c -1 +1 Tue Jan 14 07:59:14 EST 2003 athena * [project @ 2003-01-14 12:59:14 by athena] int/uint confusion M ./dft/generic.c -1 +2 M ./rdft/generic.c -1 +2 Tue Jan 14 02:25:33 EST 2003 stevenj * [project @ 2003-01-14 07:25:33 by stevenj] updated introduction and some organization M ./doc/fftw3.texi -79 +117 Tue Jan 14 01:34:46 EST 2003 stevenj * [project @ 2003-01-14 06:34:46 by stevenj] whoops M ./api/f77api.c -18 +18 Tue Jan 14 01:33:04 EST 2003 stevenj * [project @ 2003-01-14 06:33:04 by stevenj] newline M ./Makefile.am +1 Tue Jan 14 00:23:04 EST 2003 stevenj * [project @ 2003-01-14 05:23:04 by stevenj] added win32 timer M ./libbench/timer.c -1 +23 Tue Jan 14 00:12:21 EST 2003 stevenj * [project @ 2003-01-14 05:12:21 by stevenj] sync with kernel/alloc.c M ./libbench/util.c -2 +5 Tue Jan 14 00:03:20 EST 2003 stevenj * [project @ 2003-01-14 05:03:20 by stevenj] handle missing F77_FUNC_ M ./api/f77api.c +4 Mon Jan 13 17:42:50 EST 2003 stevenj * [project @ 2003-01-13 22:42:50 by stevenj] used fint instead of int to make Fortran integer type easier to change M ./api/f77api.c -24 +31 Mon Jan 13 17:38:56 EST 2003 stevenj * [project @ 2003-01-13 22:38:56 by stevenj] slight abbreviation M ./api/f77api.c -4 +4 Mon Jan 13 17:35:20 EST 2003 stevenj * [project @ 2003-01-13 22:35:20 by stevenj] the great lengthening, part I: int -> long in api; mv mktensor-rowmajor to api A ./api/mktensor-rowmajor.c M ./api/Makefile.am -2 +2 M ./api/api.h -2 +5 M ./api/f77api.c -7 +7 M ./api/fftw3.h -44 +46 M ./api/mktensor-rowmajor.c +44 M ./api/plan-dft-1d.c -1 +1 M ./api/plan-dft-2d.c -2 +2 M ./api/plan-dft-3d.c -2 +2 M ./api/plan-dft-c2r-1d.c -1 +1 M ./api/plan-dft-c2r-2d.c -2 +2 M ./api/plan-dft-c2r-3d.c -2 +2 M ./api/plan-dft-c2r.c -1 +1 M ./api/plan-dft-r2c-1d.c -1 +1 M ./api/plan-dft-r2c-2d.c -2 +2 M ./api/plan-dft-r2c-3d.c -2 +2 M ./api/plan-dft-r2c.c -1 +1 M ./api/plan-dft.c -1 +1 M ./api/plan-many-dft-c2r.c -7 +7 M ./api/plan-many-dft-r2c.c -7 +7 M ./api/plan-many-dft.c -6 +6 M ./api/plan-many-r2r.c -6 +6 M ./api/plan-r2r-1d.c -1 +1 M ./api/plan-r2r-2d.c -2 +2 M ./api/plan-r2r-3d.c -2 +2 M ./api/plan-r2r.c -1 +1 M ./api/rdft2-pad.c -4 +4 M ./kernel/Makefile.am -3 +3 M ./kernel/ifftw.h -4 +1 R ./kernel/tensor3.c M ./tests/bench.c -1 +9 Mon Jan 13 15:23:22 EST 2003 stevenj * [project @ 2003-01-13 20:23:22 by stevenj] long types M ./configure.ac +4 Mon Jan 13 04:20:37 EST 2003 athena * [project @ 2003-01-13 09:20:36 by athena] Renamed fftw_malloc -> MALLOC, X(free) -> X(ifree), X(free0) -> X(ifree0), non_fftw_malloc -> NATIVE_MALLOC M ./api/apiplan.c -2 +2 M ./api/export-wisdom-to-string.c -1 +1 M ./api/f77api.c -8 +8 M ./api/map-r2r-kind.c -1 +1 M ./api/plan-guru-r2r.c -1 +1 M ./api/plan-many-dft-c2r.c -2 +2 M ./api/plan-many-dft-r2c.c -2 +2 M ./api/plan-many-r2r.c -1 +1 M ./api/rdft2-pad.c -1 +1 M ./dft/buffered.c -6 +6 M ./dft/generic.c -1 +1 M ./dft/problem.c -2 +2 M ./dft/rader-omega.c -1 +1 M ./dft/rader.c -10 +10 M ./kernel/alloc.c -6 +6 M ./kernel/ifftw.h -11 +11 M ./kernel/plan.c -3 +3 M ./kernel/planner.c -12 +12 M ./kernel/print.c -3 +3 M ./kernel/problem.c -2 +2 M ./kernel/rader.c -3 +3 M ./kernel/scan.c -3 +3 M ./kernel/solver.c -3 +3 M ./kernel/stride.c -3 +3 M ./kernel/tensor.c -9 +9 M ./kernel/twiddle.c -5 +5 M ./rdft/buffered.c -6 +6 M ./rdft/buffered2.c -8 +8 M ./rdft/dht-rader.c -6 +6 M ./rdft/generic.c -1 +1 M ./rdft/problem.c -4 +4 M ./rdft/problem2.c -2 +2 M ./rdft/rader-hc2hc.c -10 +10 M ./reodft/redft00e-r2hc.c -5 +5 M ./reodft/reodft010e-r2hc.c -11 +11 M ./reodft/reodft11e-r2hc.c -7 +7 M ./reodft/rodft00e-r2hc.c -5 +5 M ./tests/verify-dft.c -15 +15 M ./tests/verify-rdft.c -29 +29 M ./tests/verify-reodft.c -15 +15 M ./threads/dft-vrank-geq1.c -4 +4 M ./threads/rdft-vrank-geq1.c -4 +4 M ./threads/threads.c -1 +1 M ./threads/vrank-geq1-rdft2.c -4 +4 Mon Jan 13 02:37:22 EST 2003 stevenj * [project @ 2003-01-13 07:37:22 by stevenj] added beginning of Fortran interface A ./api/f77api.c M ./api/Makefile.am -1 +1 M ./api/f77api.c +145 Mon Jan 13 01:05:29 EST 2003 stevenj * [project @ 2003-01-13 06:05:29 by stevenj] add fortran mangling check M ./configure.ac +18 Mon Jan 13 00:33:28 EST 2003 stevenj * [project @ 2003-01-13 05:33:28 by stevenj] added guru r2r interface A ./api/execute-r2r.c A ./api/plan-guru-r2r.c M ./api/Makefile.am -3 +4 M ./api/execute-r2r.c +29 M ./api/fftw3.h -1 +8 M ./api/plan-guru-r2r.c +42 Mon Jan 13 00:23:26 EST 2003 stevenj * [project @ 2003-01-13 05:23:26 by stevenj] whoops M ./api/fftw3.h -5 +7 M ./api/plan-r2r-1d.c -2 +2 M ./api/plan-r2r-2d.c -1 +4 M ./api/plan-r2r-3d.c -2 +5 Mon Jan 13 00:16:20 EST 2003 stevenj * [project @ 2003-01-13 05:16:20 by stevenj] added r2r planner A ./api/map-r2r-kind.c A ./api/plan-many-r2r.c A ./api/plan-r2r-1d.c A ./api/plan-r2r-2d.c A ./api/plan-r2r-3d.c A ./api/plan-r2r.c M ./api/Makefile.am -2 +3 M ./api/fftw3.h -1 +29 M ./api/map-r2r-kind.c +47 M ./api/plan-many-r2r.c +48 M ./api/plan-r2r-1d.c +27 M ./api/plan-r2r-2d.c +29 M ./api/plan-r2r-3d.c +30 M ./api/plan-r2r.c +27 Sun Jan 12 22:58:18 EST 2003 stevenj * [project @ 2003-01-13 03:58:18 by stevenj] more long-double checks M ./configure.ac -2 +13 Sun Jan 12 20:01:51 EST 2003 stevenj * [project @ 2003-01-13 01:01:51 by stevenj] slight regrouping M ./kernel/planner.c -5 +7 Sun Jan 12 19:58:46 EST 2003 stevenj * [project @ 2003-01-13 00:58:46 by stevenj] added joke M ./kernel/planner.c -2 +2 Sun Jan 12 19:53:58 EST 2003 stevenj * [project @ 2003-01-13 00:53:58 by stevenj] simplified rdft2 padding A ./api/rdft2-pad.c M ./api/Makefile.am -7 +7 M ./api/api.h -4 +2 R ./api/mktensor-rowmajor-pad.c M ./api/plan-many-dft-c2r.c -19 +16 M ./api/plan-many-dft-r2c.c -6 +18 M ./api/rdft2-pad.c +40 Sun Jan 12 19:02:09 EST 2003 stevenj * [project @ 2003-01-13 00:02:09 by stevenj] added comment M ./api/fftw3.h -2 +2 Sun Jan 12 18:54:49 EST 2003 stevenj * [project @ 2003-01-12 23:54:49 by stevenj] use latest api M ./tests/bench.c -2 +1 Sun Jan 12 18:49:58 EST 2003 stevenj * [project @ 2003-01-12 23:49:58 by stevenj] nembed should only be in advanced (many) interface, not basic interface...only a handful of people over the years have ever requested that functionality. M ./api/fftw3.h -10 +4 M ./api/plan-dft-1d.c -1 +1 M ./api/plan-dft-2d.c -1 +1 M ./api/plan-dft-3d.c -1 +1 M ./api/plan-dft-c2r-1d.c -1 +1 M ./api/plan-dft-c2r-2d.c -1 +1 M ./api/plan-dft-c2r-3d.c -1 +1 M ./api/plan-dft-c2r.c -5 +3 M ./api/plan-dft-r2c-1d.c -1 +1 M ./api/plan-dft-r2c-2d.c -1 +1 M ./api/plan-dft-r2c-3d.c -1 +1 M ./api/plan-dft-r2c.c -5 +3 M ./api/plan-dft.c -5 +3 Sun Jan 12 18:41:57 EST 2003 stevenj * [project @ 2003-01-12 23:41:57 by stevenj] impatient is default; generalize mapping functions using xor trick M ./api/fftw3.h -2 +2 M ./api/mapflags.c -55 +61 Sun Jan 12 14:39:42 EST 2003 stevenj * [project @ 2003-01-12 19:39:41 by stevenj] use NULL nembed to signal padding M ./api/mktensor-rowmajor-pad.c -5 +9 M ./api/plan-dft-c2r-1d.c -1 +1 M ./api/plan-dft-c2r-2d.c -1 +1 M ./api/plan-dft-c2r-3d.c -1 +1 M ./api/plan-dft-r2c-1d.c -1 +1 M ./api/plan-dft-r2c-2d.c -1 +1 M ./api/plan-dft-r2c-3d.c -1 +1 Sun Jan 12 14:23:00 EST 2003 stevenj * [project @ 2003-01-12 19:23:00 by stevenj] accept NULL nembed M ./api/plan-many-dft.c -1 +4 Sun Jan 12 13:57:13 EST 2003 stevenj * [project @ 2003-01-12 18:57:13 by stevenj] added execute-dft-r2c/c2r A ./api/execute-dft-c2r.c A ./api/execute-dft-r2c.c M ./api/Makefile.am -5 +6 M ./api/execute-dft-c2r.c +29 M ./api/execute-dft-r2c.c +29 M ./api/fftw3.h -1 +4 Sun Jan 12 13:43:20 EST 2003 stevenj * [project @ 2003-01-12 18:43:20 by stevenj] don't need dft.h M ./api/plan-dft.c -1 Sun Jan 12 13:22:14 EST 2003 stevenj * [project @ 2003-01-12 18:22:14 by stevenj] tensors are compressed in the problem, duh M ./api/plan-many-dft-c2r.c -2 +1 M ./api/plan-many-dft-r2c.c -2 +1 Sun Jan 12 12:45:26 EST 2003 stevenj * [project @ 2003-01-12 17:45:26 by stevenj] noted that posix_memalign bug is now fixed, thanks to bug report by yours truly M ./kernel/alloc.c -2 +3 Sun Jan 12 12:44:43 EST 2003 athena * [project @ 2003-01-12 17:44:43 by athena] Bug: n[3] instead of n[2]. Bug was propagated by copy-and-paste. Grrr... M ./api/plan-dft-3d.c -1 +1 M ./api/plan-dft-c2r-3d.c -1 +1 M ./api/plan-dft-r2c-3d.c -1 +1 Sun Jan 12 12:41:43 EST 2003 athena * [project @ 2003-01-12 17:41:43 by athena] Express plan_dft() in terms of plan_many_dft() M ./api/plan-dft.c -11 +4 Sun Jan 12 12:19:53 EST 2003 stevenj * [project @ 2003-01-12 17:19:38 by stevenj] whoops A ./api/plan-guru-dft-c2r.c A ./api/plan-guru-dft-r2c.c A ./api/plan-guru-dft.c M ./api/plan-guru-dft-c2r.c +36 M ./api/plan-guru-dft-r2c.c +35 M ./api/plan-guru-dft.c +35 Sun Jan 12 06:00:46 EST 2003 athena * [project @ 2003-01-12 11:00:46 by athena] Manual skeleton. A ./doc/ A ./doc/Makefile.am A ./doc/fftw3.texi M ./Makefile.am -1 +1 M ./configure.ac +1 M ./doc/Makefile.am -1 +2 M ./doc/fftw3.texi +318 M ./genfft-k7/vK7Optimization.ml -1 +1 Sat Jan 11 23:46:34 EST 2003 stevenj * [project @ 2003-01-12 04:46:34 by stevenj] added r2c/c2r guru api M ./api/Makefile.am -1 +2 M ./api/fftw3.h -1 +12 Sat Jan 11 23:42:10 EST 2003 stevenj * [project @ 2003-01-12 04:42:10 by stevenj] FFTW_DESTROY_INPUT is default for c2r transforms M ./api/plan-many-dft-c2r.c -1 +1 Sat Jan 11 23:36:26 EST 2003 stevenj * [project @ 2003-01-12 04:36:26 by stevenj] added more of r2c/c2r api A ./api/plan-dft-c2r-1d.c A ./api/plan-dft-c2r-2d.c A ./api/plan-dft-c2r-3d.c A ./api/plan-dft-c2r.c A ./api/plan-dft-r2c-1d.c A ./api/plan-dft-r2c-2d.c A ./api/plan-dft-r2c-3d.c A ./api/plan-dft-r2c.c M ./api/Makefile.am -1 +3 M ./api/fftw3.h -69 +90 M ./api/plan-dft-c2r-1d.c +26 M ./api/plan-dft-c2r-2d.c +29 M ./api/plan-dft-c2r-3d.c +29 M ./api/plan-dft-c2r.c +32 M ./api/plan-dft-r2c-1d.c +26 M ./api/plan-dft-r2c-2d.c +29 M ./api/plan-dft-r2c-3d.c +29 M ./api/plan-dft-r2c.c +32 Sat Jan 11 21:09:41 EST 2003 stevenj * [project @ 2003-01-12 02:09:41 by stevenj] r2c doesn't have adjustible sign M ./api/fftw3.h -3 +3 M ./api/plan-many-dft-c2r.c -2 +2 M ./api/plan-many-dft-r2c.c -2 +2 Sat Jan 11 21:07:55 EST 2003 stevenj * [project @ 2003-01-12 02:07:55 by stevenj] note that copyright year is out of date M ./TODO -1 +1 Sat Jan 11 21:04:23 EST 2003 stevenj * [project @ 2003-01-12 02:04:23 by stevenj] updated api for r2c M ./api/fftw3.h -1 +17 Sat Jan 11 21:00:07 EST 2003 stevenj * [project @ 2003-01-12 02:00:07 by stevenj] removed annoying nophys == niphys case M ./api/mktensor-rowmajor-pad.c -11 Sat Jan 11 20:58:13 EST 2003 stevenj * [project @ 2003-01-12 01:58:13 by stevenj] added basic r2c/c2r planner A ./api/mktensor-rowmajor-pad.c A ./api/plan-many-dft-c2r.c A ./api/plan-many-dft-r2c.c M ./api/Makefile.am -2 +3 M ./api/api.h +5 M ./api/mktensor-rowmajor-pad.c +88 M ./api/plan-many-dft-c2r.c +59 M ./api/plan-many-dft-r2c.c +44 Sat Jan 11 19:34:14 EST 2003 stevenj * [project @ 2003-01-12 00:34:14 by stevenj] dist should be in terms of complex values M ./api/plan-many-dft.c -1 +1 Sat Jan 11 19:14:24 EST 2003 stevenj * [project @ 2003-01-12 00:14:24 by stevenj] added plan-with-nthreads M ./api/fftw3.h -2 +5 Sat Jan 11 19:12:51 EST 2003 stevenj * [project @ 2003-01-12 00:12:51 by stevenj] added function to set nthr A ./api/plan-with-nthreads.c M ./api/Makefile.am -1 +1 M ./api/plan-with-nthreads.c +39 Sat Jan 11 18:04:57 EST 2003 stevenj * [project @ 2003-01-11 23:04:57 by stevenj] slight cleanup M ./api/fftw3.h -4 +5 Sat Jan 11 17:57:29 EST 2003 stevenj * [project @ 2003-01-11 22:57:29 by stevenj] whoops A ./api/mktensor-iodims.c Sat Jan 11 17:55:39 EST 2003 stevenj * [project @ 2003-01-11 22:55:39 by stevenj] maxlen is maximum string length, not including null termination M ./kernel/scan.c -2 +2 Sat Jan 11 17:50:49 EST 2003 stevenj * [project @ 2003-01-11 22:50:49 by stevenj] imprt reverts hashtable on failure M ./kernel/planner.c -3 +14 Sat Jan 11 16:43:54 EST 2003 stevenj * [project @ 2003-01-11 21:43:54 by stevenj] slight move M ./api/fftw3.h -14 +14 Sat Jan 11 16:34:56 EST 2003 stevenj * [project @ 2003-01-11 21:34:56 by stevenj] stdio.h should be inlcuded outside of extern "C" M ./api/fftw3.h -3 +3 Sat Jan 11 16:26:35 EST 2003 stevenj * [project @ 2003-01-11 21:26:35 by stevenj] added guru planner API M ./api/Makefile.am -2 +3 M ./api/api.h +1 M ./api/fftw3.h -1 +15 Sat Jan 11 15:54:57 EST 2003 stevenj * [project @ 2003-01-11 20:54:57 by stevenj] added FFTW_FORWARD/BACKWARD M ./api/fftw3.h -1 +4 Sat Jan 11 15:52:17 EST 2003 stevenj * [project @ 2003-01-11 20:52:17 by stevenj] added plan_many_dft A ./api/plan-many-dft.c M ./api/Makefile.am -1 +1 M ./api/fftw3.h -1 +9 M ./api/plan-many-dft.c +43 Sat Jan 11 15:44:37 EST 2003 stevenj * [project @ 2003-01-11 20:44:37 by stevenj] indenting M ./kernel/tensor3.c -3 +3 Sat Jan 11 14:49:08 EST 2003 athena * [project @ 2003-01-11 19:49:08 by athena] Final \n M ./tests/bench.c -2 +4 Sat Jan 11 14:47:31 EST 2003 athena * [project @ 2003-01-11 19:47:31 by athena] Do not compile if not defined(FFTW_DEBUG), in order to avoid unused code in the shared library. M ./kernel/debug.c -1 +3 Sat Jan 11 14:45:56 EST 2003 athena * [project @ 2003-01-11 19:45:56 by athena] Implemented print_plan() A ./api/mkprinter-file.c A ./api/print-plan.c M ./api/Makefile.am -4 +4 M ./api/api.h +1 M ./api/export-wisdom-to-file.c -19 +1 M ./api/fftw3.h -2 +3 M ./api/mkprinter-file.c +39 M ./api/print-plan.c +29 M ./tests/bench.c -2 +3 Sat Jan 11 13:12:01 EST 2003 stevenj * [project @ 2003-01-11 18:12:01 by stevenj] changed the OOP-like plan_destroy to the more-grammatical destroy_plan M ./api/apiplan.c -1 +1 M ./api/fftw3.h -2 +2 M ./tests/bench.c -1 +1 Sat Jan 11 12:58:04 EST 2003 stevenj * [project @ 2003-01-11 17:55:49 by stevenj] added guru execute_dft A ./api/execute-dft.c M ./api/Makefile.am -4 +5 M ./api/execute-dft.c +29 M ./api/fftw3.h -2 +3 Sat Jan 11 12:38:40 EST 2003 stevenj * [project @ 2003-01-11 17:38:40 by stevenj] allow for malloc errors in wisdom string, since non-fftw-malloc M ./api/export-wisdom-to-string.c -4 +5 Sat Jan 11 12:16:05 EST 2003 stevenj * [project @ 2003-01-11 17:16:05 by stevenj] cleanup should reset plnr to zero so that fftw can be restarted M ./api/the-planner.c -1 +3 Sat Jan 11 12:13:18 EST 2003 stevenj * [project @ 2003-01-11 17:13:18 by stevenj] NO_UGLY is an internal planner flag M ./api/fftw3.h -4 +3 M ./api/mapflags.c -3 Sat Jan 11 11:23:13 EST 2003 athena * [project @ 2003-01-11 16:23:13 by athena] Written 1d api in terms of generic n-d api. The code is less compact but easier to test M ./api/plan-dft-1d.c -9 +1 M ./tests/bench.c -27 +36 Sat Jan 11 11:07:25 EST 2003 athena * [project @ 2003-01-11 16:07:24 by athena] Added wisdom to header file, made scanners/printer static. stdio.h no longer needed in fftw.h, removed. Probably the printer_file should be reintroduced in a separate file if we ever want to print plans... M ./api/export-wisdom-to-file.c -2 +2 M ./api/export-wisdom-to-string.c -4 +4 M ./api/fftw3.h -2 +13 M ./api/import-wisdom-from-file.c -2 +2 M ./api/import-wisdom-from-string.c -2 +2 M ./kernel/alloc.c -1 +3 M ./kernel/assert.c -2 +2 M ./kernel/debug.c -2 +22 M ./kernel/ifftw.h -15 +1 M ./kernel/print.c -1 +2 M ./kernel/scan.c -1 +2 M ./tests/bench.c -13 +49 M ./tests/verify-lib.c -1 +2 M ./tests/verify-reodft.c -1 +2 Sat Jan 11 09:49:30 EST 2003 athena * [project @ 2003-01-11 14:49:30 by athena] Implemented more APIs A ./api/plan-dft-2d.c A ./api/plan-dft-3d.c A ./api/plan-dft.c M ./api/Makefile.am -6 +5 M ./api/apiplan.c -1 +1 M ./api/fftw3.h -13 +25 M ./api/plan-dft-2d.c +30 M ./api/plan-dft-3d.c +30 M ./api/plan-dft.c +40 M ./tests/bench.c +18 Sat Jan 11 09:21:53 EST 2003 athena * [project @ 2003-01-11 14:21:53 by athena] Added cleanup() to API M ./api/fftw3.h -3 +3 M ./api/the-planner.c +6 M ./tests/bench.c +1 Sat Jan 11 09:17:34 EST 2003 athena * [project @ 2003-01-11 14:17:34 by athena] Started new bench.c. I had to rename plan_destroy -> plan_destroy_internal to avoid conflicts with API M ./api/api.h -1 M ./api/apiplan.c -3 +3 M ./api/fftw3.h -1 +3 M ./dft/buffered.c -7 +7 M ./dft/ct.c -2 +2 M ./dft/generic.c -2 +2 M ./dft/indirect.c -5 +5 M ./dft/rader.c -8 +8 M ./dft/rank-geq2.c -5 +5 M ./dft/vrank-geq1.c -2 +2 M ./kernel/ifftw.h -2 +2 M ./kernel/plan.c -2 +2 M ./kernel/planner.c -3 +3 M ./libbench/bench-user.h -1 +4 M ./libbench/bench.h -3 +1 M ./rdft/buffered.c -7 +7 M ./rdft/buffered2.c -5 +5 M ./rdft/dft-r2hc.c -2 +2 M ./rdft/dht-r2hc.c -2 +2 M ./rdft/dht-rader.c -6 +6 M ./rdft/generic.c -2 +2 M ./rdft/hc2hc.c -7 +7 M ./rdft/indirect.c -5 +5 M ./rdft/rader-hc2hc.c -7 +7 M ./rdft/rank-geq2-rdft2.c -5 +5 M ./rdft/rank-geq2.c -5 +5 M ./rdft/rank0-rdft2.c -2 +2 M ./rdft/rdft-dht.c -2 +2 M ./rdft/rdft2-radix2.c -2 +2 M ./rdft/vrank-geq1-rdft2.c -2 +2 M ./rdft/vrank-geq1.c -2 +2 M ./reodft/redft00e-r2hc.c -2 +2 M ./reodft/reodft010e-r2hc.c -2 +2 M ./reodft/reodft11e-r2hc.c -2 +2 M ./reodft/rodft00e-r2hc.c -2 +2 M ./tests/Makefile.am -1 +1 M ./tests/bench.c -2 +103 M ./threads/dft-vrank-geq1.c -3 +3 M ./threads/rdft-vrank-geq1.c -3 +3 M ./threads/vrank-geq1-rdft2.c -3 +3 Sat Jan 11 02:45:39 EST 2003 stevenj * [project @ 2003-01-11 07:45:39 by stevenj] fix types A ./api/import-system-wisdom.c M ./api/Makefile.am -3 +3 M ./api/export-wisdom.c -2 +2 M ./api/import-system-wisdom.c +46 M ./api/import-wisdom-from-file.c -2 +3 M ./api/import-wisdom-from-string.c -3 +4 M ./api/import-wisdom.c -3 +5 Sat Jan 11 02:13:25 EST 2003 stevenj * [project @ 2003-01-11 07:13:25 by stevenj] whoops M ./api/export-wisdom-to-string.c -1 +2 Sat Jan 11 02:10:50 EST 2003 stevenj * [project @ 2003-01-11 07:10:50 by stevenj] added wisdom api A ./api/export-wisdom-to-file.c A ./api/export-wisdom-to-string.c A ./api/export-wisdom.c A ./api/forget-wisdom.c A ./api/import-wisdom-from-file.c A ./api/import-wisdom-from-string.c A ./api/import-wisdom.c M ./api/Makefile.am -2 +6 M ./api/export-wisdom-to-file.c +47 M ./api/export-wisdom-to-string.c +80 M ./api/export-wisdom.c +44 M ./api/forget-wisdom.c +27 M ./api/import-wisdom-from-file.c +47 M ./api/import-wisdom-from-string.c +49 M ./api/import-wisdom.c +44 M ./kernel/Makefile.am -4 +4 M ./kernel/ifftw.h -1 +3 R ./kernel/printers.c R ./kernel/scanners.c Sat Jan 11 01:01:17 EST 2003 stevenj * [project @ 2003-01-11 06:01:17 by stevenj] grammar M ./api/mapflags.c -1 +1 Sat Jan 11 00:54:54 EST 2003 stevenj * [project @ 2003-01-11 05:54:54 by stevenj] slight change M ./api/mapflags.c -1 +1 Sat Jan 11 00:52:04 EST 2003 stevenj * [project @ 2003-01-11 05:52:04 by stevenj] implemented api/mapflags M ./api/fftw3.h -1 +23 M ./api/mapflags.c -1 +83 Sat Jan 11 00:48:27 EST 2003 stevenj * [project @ 2003-01-11 05:48:27 by stevenj] IMPATIENT is an api issue M ./kernel/ifftw.h -11 +1 Fri Jan 10 01:57:41 EST 2003 stevenj * [project @ 2003-01-10 06:57:41 by stevenj] removed un-needed headers M ./api/the-planner.c -4 Fri Jan 10 01:56:59 EST 2003 stevenj * [project @ 2003-01-10 06:56:59 by stevenj] mkplanner initializes nthr to 1 already M ./api/the-planner.c -1 Thu Jan 9 18:53:09 EST 2003 stevenj * [project @ 2003-01-09 23:53:09 by stevenj] boilerplate M ./api/fftw3.h -1 +14 Thu Jan 9 18:16:39 EST 2003 stevenj * [project @ 2003-01-09 23:16:39 by stevenj] fold vecloop into r{e,o}dft apply function to share buffer, etcetera M ./rdft/vrank-geq1.c -1 +6 M ./reodft/redft00e-r2hc.c -42 +45 M ./reodft/reodft010e-r2hc.c -138 +153 M ./reodft/reodft11e-r2hc.c -98 +107 M ./reodft/rodft00e-r2hc.c -39 +42 Thu Jan 9 18:10:19 EST 2003 stevenj * [project @ 2003-01-09 23:10:19 by stevenj] whoops, bugfix in impulse test for vecn > 1 M ./tests/verify-reodft.c -2 +2 Thu Jan 9 14:23:51 EST 2003 stevenj * [project @ 2003-01-09 19:23:51 by stevenj] bugfix, grr M ./rdft/hc2hc-buf.c -2 +2 Thu Jan 9 14:21:16 EST 2003 stevenj * [project @ 2003-01-09 19:21:16 by stevenj] fixed signed-ness enum problem M ./rdft/codelet-rdft.h -2 +5 Thu Jan 9 14:12:42 EST 2003 athena * [project @ 2003-01-09 19:12:42 by athena] Explicit cast M ./kernel/md5-1.c -1 +1 Thu Jan 9 13:41:51 EST 2003 athena * [project @ 2003-01-09 18:41:51 by athena] Added configure_planner(). mkplan() behaves properly when plan is null. A ./api/configure.c M ./api/Makefile.am -2 +2 M ./api/api.h +1 M ./api/apiplan.c -14 +21 M ./api/configure.c +33 M ./api/fftw3.h -3 +3 M ./api/the-planner.c -4 +1 Thu Jan 9 06:48:53 EST 2003 athena * [project @ 2003-01-09 11:48:53 by athena] More API work A ./api/apiplan.c A ./api/execute.c A ./api/mapflags.c M ./api/Makefile.am -4 +2 M ./api/api.h -2 +3 M ./api/apiplan.c +60 M ./api/execute.c +27 M ./api/fftw3.h -7 +3 M ./api/mapflags.c +26 M ./api/plan-dft-1d.c -11 +4 M ./tests/bench.c -1 +1 Thu Jan 9 05:40:34 EST 2003 athena * [project @ 2003-01-09 10:40:34 by athena] First skeleton of API infrastructure A ./api/ A ./api/Makefile.am A ./api/api.h A ./api/dfthelp.c A ./api/fftw3.h A ./api/plan-dft-1d.c A ./api/the-planner.c M ./Makefile.am -5 +7 M ./api/Makefile.am +10 M ./api/api.h +51 M ./api/dfthelp.c +37 M ./api/fftw3.h +67 M ./api/plan-dft-1d.c +43 M ./api/the-planner.c +42 M ./configure.ac +2 M ./kernel/ifftw.h -1 +2 M ./kernel/trig.c -3 +1 M ./tests/Makefile.am -1 +1 Thu Jan 9 03:19:35 EST 2003 stevenj * [project @ 2003-01-09 08:19:35 by stevenj] unsigned strikes again M ./rdft/rdft2-tensor-max-index.c -3 +3 Thu Jan 9 01:51:45 EST 2003 stevenj * [project @ 2003-01-09 06:51:45 by stevenj] put rdft2_inplace_strides and rdft2_tensor_max_index in their own files for tighter linking A ./rdft/rdft2-inplace-strides.c A ./rdft/rdft2-tensor-max-index.c M ./rdft/Makefile.am -1 +2 M ./rdft/problem2.c -46 +1 M ./rdft/rdft2-inplace-strides.c +68 M ./rdft/rdft2-tensor-max-index.c +47 M ./rdft/vrank-geq1-rdft2.c -25 +1 Thu Jan 9 01:43:13 EST 2003 stevenj * [project @ 2003-01-09 06:43:13 by stevenj] added rdft2_tensor_max_index...incorrect use of tensor_max_index was preventing proper loop ordering for rnk > 2 rdft2 M ./rdft/rank-geq2-rdft2.c -2 +3 M ./rdft/rdft.h -1 +2 M ./rdft/vrank-geq1-rdft2.c -4 +27 Thu Jan 9 00:44:45 EST 2003 stevenj * [project @ 2003-01-09 05:44:45 by stevenj] arbitrary spltrnk in rdft2 rank-geq2 M ./rdft/rank-geq2-rdft2.c -11 +53 Thu Jan 9 00:40:17 EST 2003 stevenj * [project @ 2003-01-09 05:40:17 by stevenj] don't mention wisdom when non-verbose M ./tests/bench.c -2 +4 Thu Jan 9 00:02:35 EST 2003 stevenj * [project @ 2003-01-09 05:02:35 by stevenj] bug fix: printing %T should pass tensor *, not tensor ** M ./dft/problem.c -3 +3 M ./rdft/problem.c -3 +3 M ./rdft/problem2.c -3 +3 Wed Jan 8 23:40:48 EST 2003 stevenj * [project @ 2003-01-09 04:40:48 by stevenj] correct(?) normalization for rodft00 ... all of the even/odd transforms should be normalized according to the expanded'' DFT of ~twice the length M ./reodft/rodft00e-r2hc.c -4 +4 M ./tests/verify-reodft.c -2 +2 Wed Jan 8 23:18:23 EST 2003 stevenj * [project @ 2003-01-09 04:18:23 by stevenj] fixed tests for n=1 M ./tests/verify-reodft.c -3 +6 Wed Jan 8 22:10:08 EST 2003 stevenj * [project @ 2003-01-09 03:10:08 by stevenj] fixed bug in vector tests for rdft(2) M ./tests/bench.c -3 +5 Wed Jan 8 20:12:00 EST 2003 stevenj * [project @ 2003-01-09 01:12:00 by stevenj] fixed handling when first rnk-1 dimensions compress to nothing (ugh) M ./rdft/problem2.c -3 +6 Wed Jan 8 20:02:35 EST 2003 stevenj * [project @ 2003-01-09 01:02:35 by stevenj] fixed incorrect/missing rdft2 rank-0 handling A ./rdft/rank0-rdft2.c M ./rdft/Makefile.am -2 +2 M ./rdft/conf.c -1 +2 M ./rdft/nop2.c -3 +6 M ./rdft/rank0-rdft2.c +194 M ./rdft/rdft.h -1 +2 Wed Jan 8 19:49:05 EST 2003 stevenj * [project @ 2003-01-09 00:49:05 by stevenj] bug fix: for rnk > 1, must compress rnk-1 dims separately (ugh) M ./rdft/problem2.c -2 +11 Wed Jan 8 17:39:14 EST 2003 stevenj * [project @ 2003-01-08 22:39:14 by stevenj] added trailing newline M ./configure.ac +1 Wed Jan 8 17:38:02 EST 2003 stevenj * [project @ 2003-01-08 22:38:02 by stevenj] updated M ./ChangeLog +66 Wed Jan 8 16:53:16 EST 2003 stevenj * [project @ 2003-01-08 21:53:16 by stevenj] got rid of compiler warning M ./rdft/problem.c -2 +2 Wed Jan 8 16:49:48 EST 2003 stevenj * [project @ 2003-01-08 21:49:48 by stevenj] whoops, test r2hc and not rodft00 by default M ./tests/bench.c -1 +1 Wed Jan 8 16:46:24 EST 2003 stevenj * [project @ 2003-01-08 21:46:24 by stevenj] got rid of real_n...use physical n everywhere in rdft; fixed rdft sz compression; fixed rodft00 verify bug M ./rdft/buffered.c -3 +2 M ./rdft/indirect.c -10 +6 M ./rdft/problem.c -44 +43 M ./rdft/rank-geq2.c -10 +6 M ./rdft/rdft.h -3 +1 M ./reodft/redft00e-r2hc.c -4 +5 M ./reodft/reodft010e-r2hc.c -4 +3 M ./reodft/reodft11e-r2hc.c -4 +3 M ./reodft/rodft00e-r2hc.c -4 +3 M ./tests/bench.c -2 +2 M ./tests/verify-reodft.c -33 +35 Wed Jan 8 07:20:47 EST 2003 athena * [project @ 2003-01-08 12:20:47 by athena] icc-6.0 bug workaround A ./simd/sse-aux.c A ./simd/sse2-aux.c M ./simd/Makefile.am -2 +2 M ./simd/sse-aux.c +33 M ./simd/sse.c -3 +1 M ./simd/sse2-aux.c +34 M ./simd/sse2.c -3 +1 Wed Jan 8 04:21:40 EST 2003 athena * [project @ 2003-01-08 09:21:40 by athena] Reclaimed the fftw_real identifier, because I need it for the API M ./kernel/ifftw.h -12 +9 M ./rdft/buffered2.c -5 +5 M ./rdft/rader-hc2hc.c -1 +1 M ./tests/bench.c +1 Wed Jan 8 04:14:55 EST 2003 athena * [project @ 2003-01-08 09:14:55 by athena] Use recommended AC_OUTPUT syntax M ./configure.ac -1 +2 Wed Jan 8 04:00:22 EST 2003 athena * [project @ 2003-01-08 09:00:22 by athena] Removed FFTW(foo) as a synonym for X(foo). This is an API issue. M ./kernel/ifftw.h -8 +7 M ./tests/bench.c +1 Tue Jan 7 17:45:52 EST 2003 stevenj * [project @ 2003-01-07 22:45:52 by stevenj] get rid of warning M ./simd/sse2.c -1 +2 Tue Jan 7 16:22:39 EST 2003 athena * [project @ 2003-01-07 21:22:39 by athena] Renamed conflicting files */codelet.h into dft/codelet-dft.h and rdft/codelet-rdft.h A ./dft/codelet-dft.h A ./rdft/codelet-rdft.h A ./support/codelet_prelude.dft A ./support/codelet_prelude.rdft M ./dft/Makefile.am -1 +1 M ./dft/codelet-dft.h +113 R ./dft/codelet.h M ./dft/codelets/inplace/Makefile.am -4 +4 M ./dft/codelets/n.c -1 +1 M ./dft/codelets/standard/Makefile.am -3 +3 M ./dft/codelets/t.c -1 +1 M ./dft/dft.h -2 +2 M ./dft/simd/codelets/Makefile.am -6 +6 M ./dft/simd/n1b.c -1 +1 M ./dft/simd/n1f.c -1 +1 M ./dft/simd/t1b.c -1 +1 M ./dft/simd/t1f.c -1 +1 M ./rdft/Makefile.am -5 +5 M ./rdft/codelet-rdft.h +190 R ./rdft/codelet.h M ./rdft/codelets/hc2r/Makefile.am -4 +4 M ./rdft/codelets/hc2r.c -1 +1 M ./rdft/codelets/hfb.c -1 +1 M ./rdft/codelets/r2hc/Makefile.am -4 +4 M ./rdft/codelets/r2hc.c -1 +1 M ./rdft/rdft.h -2 +2 M ./support/Makefile.am -1 +2 M ./support/Makefile.codelets -2 +4 R ./support/codelet_prelude M ./support/codelet_prelude.dft +8 M ./support/codelet_prelude.rdft +8 Tue Jan 7 16:21:16 EST 2003 stevenj * [project @ 2003-01-07 21:21:16 by stevenj] updated M ./ChangeLog +737 Tue Jan 7 15:47:24 EST 2003 athena * [project @ 2003-01-07 20:47:24 by athena] Silence warnings M ./simd/simd-3dnow.h -1 +1 M ./simd/simd-sse.h -1 M ./simd/simd-sse2.h -1 +1 M ./simd/sse2.c -2 +1 Tue Jan 7 15:00:14 EST 2003 stevenj * [project @ 2003-01-07 20:00:14 by stevenj] fftw2 used spltrnk=1 M ./dft/rank-geq2.c -2 +4 M ./rdft/rank-geq2.c -2 +4 Tue Jan 7 14:32:06 EST 2003 athena * [project @ 2003-01-07 19:32:06 by athena] Silence warning M ./dft/codelet.h -2 +1 M ./rdft/codelet.h -2 +1 M ./simd/simd-sse.h -1 +2 M ./simd/sse.c -2 +1 Tue Jan 7 12:13:50 EST 2003 stevenj * [project @ 2003-01-07 17:13:50 by stevenj] noted deficiency M ./TODO +2 Tue Jan 7 07:18:51 EST 2003 athena * [project @ 2003-01-07 12:18:51 by athena] Strengthened conditions for a problem to be POSSIBLY_UNALIGNED M ./rdft/vrank-geq1-rdft2.c -4 +2 M ./rdft/vrank-geq1.c -2 +2 Tue Jan 7 05:09:42 EST 2003 athena * [project @ 2003-01-07 10:09:42 by athena] Strengthened conditions for a plan to be POSSIBLY_UNALIGNED M ./dft/vrank-geq1.c -5 +2 M ./kernel/align.c -1 +6 M ./kernel/ifftw.h -1 +2 Sun Jan 5 02:43:45 EST 2003 stevenj * [project @ 2003-01-05 07:43:45 by stevenj] added copyright todo M ./TODO +2 Sun Jan 5 02:37:31 EST 2003 stevenj * [project @ 2003-01-05 07:37:31 by stevenj] modified comment M ./kernel/planner.c -2 +3 Sun Jan 5 02:34:36 EST 2003 stevenj * [project @ 2003-01-05 07:33:41 by stevenj] fixed comment M ./tests/verify-rdft.c -3 +3 Sun Jan 5 02:31:56 EST 2003 stevenj * [project @ 2003-01-05 07:31:56 by stevenj] implemented rdft2 verify M ./TODO -4 +2 M ./tests/verify-rdft.c -5 +15 Sat Jan 4 16:20:42 EST 2003 stevenj * [project @ 2003-01-04 21:20:42 by stevenj] fix --enable-single M ./configure.ac -1 +1 Wed Oct 23 12:59:12 EDT 2002 stevenj * [project @ 2002-10-23 16:59:12 by stevenj] slight fixes M ./threads/threads.c -20 +14 Wed Oct 23 12:42:39 EDT 2002 stevenj * [project @ 2002-10-23 16:42:39 by stevenj] typo M ./threads/threads.c -1 +1 Tue Oct 1 09:32:56 EDT 2002 athena * [project @ 2002-10-01 13:32:56 by athena] Experimental stuff M ./genfft/annotate.ml -2 +9 M ./genfft/annotate.mli -3 +2 M ./genfft/c.ml -4 +16 M ./genfft/genutil.ml -1 +11 M ./genfft/magic.ml -1 +5 Sat Sep 28 13:03:53 EDT 2002 athena * [project @ 2002-09-28 17:03:53 by athena] Experimental Franz mode M ./configure.ac +3 M ./dft/simd/codelets/Makefile.am -3 +68 M ./genfft/gen_notw_c.ml -7 +8 M ./genfft/gen_twiddle_c.ml -6 +7 M ./genfft/genutil.ml -2 +10 Thu Sep 26 15:14:38 EDT 2002 athena * [project @ 2002-09-26 19:14:38 by athena] const-correct M ./kernel/tensor.c -2 +2 Thu Sep 26 15:06:38 EDT 2002 athena * [project @ 2002-09-26 19:06:38 by athena] Reuse dimcmp routine for other purposes M ./dft/vrank2-transpose.c -4 +3 M ./dft/vrank3-transpose.c -4 +2 M ./kernel/ifftw.h -1 +2 M ./kernel/tensor7.c -5 +5 M ./rdft/vrank2-transpose.c -4 +2 M ./rdft/vrank3-transpose.c -4 +2 Wed Sep 25 07:37:38 EDT 2002 athena * [project @ 2002-09-25 11:37:38 by athena] Use tornk1 correctly. M ./dft/direct.c -3 +3 M ./kernel/ifftw.h -2 +2 M ./kernel/tensor.c -2 +3 M ./rdft/direct.c -3 +3 M ./rdft/direct2.c -3 +3 Wed Sep 25 07:36:38 EDT 2002 athena * [project @ 2002-09-25 11:36:38 by athena] Hmm... I thought I had fixed this before... M ./rdft/rdft2-radix2.c -2 +2 Tue Sep 24 21:27:49 EDT 2002 athena * [project @ 2002-09-25 01:27:49 by athena] Collect more common idioms M ./dft/buffered.c -5 +3 M ./dft/rank0.c -16 +3 M ./kernel/tensor.c -2 +2 M ./rdft/buffered.c -5 +2 M ./rdft/buffered2.c -5 +2 M ./rdft/rank0.c -16 +3 Tue Sep 24 21:15:57 EDT 2002 athena * [project @ 2002-09-25 01:15:57 by athena] Still collecting common idioms... M ./dft/direct.c -10 +3 M ./rdft/direct.c -10 +2 M ./rdft/direct2.c -9 +2 Tue Sep 24 21:13:00 EDT 2002 athena * [project @ 2002-09-25 01:13:00 by athena] More garbage collection. M ./dft/direct.c -6 +5 M ./rdft/direct.c -6 +5 M ./rdft/direct2.c -6 +5 Tue Sep 24 21:08:19 EDT 2002 athena * [project @ 2002-09-25 01:08:19 by athena] More compact code M ./dft/buffered.c -6 +3 Tue Sep 24 20:54:43 EDT 2002 athena * [project @ 2002-09-25 00:54:43 by athena] Collect common pattern if (foo) free(foo) ==> free0(foo) M ./dft/buffered.c -7 +6 M ./dft/generic.c -2 +1 M ./dft/rader.c -6 +3 M ./kernel/alloc.c -1 +7 M ./kernel/ifftw.h -1 +2 M ./kernel/planner.c -8 +4 M ./kernel/stride.c -3 +2 M ./kernel/tensor.c -5 +2 M ./rdft/buffered.c -3 +2 M ./rdft/buffered2.c -3 +2 M ./rdft/dht-rader.c -2 +1 M ./rdft/generic.c -2 +1 M ./rdft/problem.c -3 +2 M ./rdft/rader-hc2hc.c -6 +3 Tue Sep 24 20:08:44 EDT 2002 athena * [project @ 2002-09-25 00:08:44 by athena] Collect some common code in */buffered*.c A ./kernel/buffered.c M ./dft/buffered.c -18 +2 M ./kernel/Makefile.am -7 +6 M ./kernel/buffered.c +44 M ./kernel/ifftw.h -1 +2 M ./rdft/buffered.c -18 +3 M ./rdft/buffered2.c -18 +2 Tue Sep 24 19:39:22 EDT 2002 stevenj * [project @ 2002-09-24 23:39:22 by stevenj] use STRUCT_HACK #define to determing rdft kind[] allocation M ./rdft/problem.c -8 +23 M ./rdft/rdft.h -2 +8 Tue Sep 24 17:21:09 EDT 2002 stevenj * [project @ 2002-09-24 21:21:09 by stevenj] report total pcost of measured/estimated plans...epcost is especially useful to estimate the effects of various impatience flags on planning time for large transforms M ./kernel/ifftw.h -1 +2 M ./kernel/planner.c -1 +6 Mon Sep 23 18:49:10 EDT 2002 athena * [project @ 2002-09-23 22:49:10 by athena] Prevent unwanted inlining A ./kernel/trig1.c M ./kernel/Makefile.am -1 +1 M ./kernel/trig.c -54 +7 M ./kernel/trig1.c +70 Mon Sep 23 18:37:59 EDT 2002 athena * [project @ 2002-09-23 22:37:59 by athena] Space compaction M ./kernel/ifftw.h -2 +1 M ./kernel/trig.c -21 +27 Mon Sep 23 11:49:32 EDT 2002 athena * [project @ 2002-09-23 15:49:32 by athena] Still reducing size A ./kernel/hash.c M ./kernel/Makefile.am -1 +1 M ./kernel/hash.c +31 M ./kernel/ifftw.h -1 +2 M ./kernel/md5-1.c +1 M ./kernel/planner.c -12 +3 M ./kernel/scan.c -22 +1 Sun Sep 22 16:03:30 EDT 2002 athena * [project @ 2002-09-22 20:03:30 by athena] Saved another 5KB by redesigning opcnt protocol. (gasp!) M ./dft/buffered.c -5 +6 M ./dft/ct-dif.c -4 +3 M ./dft/ct-dit.c -4 +3 M ./dft/ct-ditbuf.c -6 +6 M ./dft/ct-ditf.c -4 +3 M ./dft/direct.c -2 +3 M ./dft/generic.c -4 +2 M ./dft/indirect.c -2 +2 M ./dft/nop.c -2 +2 M ./dft/rader.c -4 +3 M ./dft/rank-geq2.c -2 +2 M ./dft/rank0.c -2 +2 M ./dft/vrank-geq1.c -2 +3 M ./dft/vrank2-transpose.c -2 +2 M ./dft/vrank3-transpose.c -2 +2 M ./kernel/ifftw.h -6 +14 M ./kernel/ops.c -23 +28 M ./kernel/plan.c -2 +2 M ./rdft/buffered.c -5 +6 M ./rdft/buffered2.c -3 +3 M ./rdft/dht-rader.c -1 +1 M ./rdft/direct.c -3 +8 M ./rdft/direct2.c -3 +8 M ./rdft/generic.c -4 +3 M ./rdft/hc2hc-buf.c -10 +9 M ./rdft/hc2hc-dif.c -8 +8 M ./rdft/hc2hc-dit.c -8 +8 M ./rdft/indirect.c -2 +2 M ./rdft/nop.c -2 +2 M ./rdft/nop2.c -2 +2 M ./rdft/rader-hc2hc.c -8 +5 M ./rdft/rank-geq2-rdft2.c -2 +2 M ./rdft/rank-geq2.c -2 +2 M ./rdft/rank0.c -2 +2 M ./rdft/rdft2-radix2.c -5 +4 M ./rdft/vrank-geq1-rdft2.c -2 +3 M ./rdft/vrank-geq1.c -2 +3 M ./rdft/vrank2-transpose.c -2 +2 M ./rdft/vrank3-transpose.c -2 +2 M ./threads/ct-dit.c -4 +3 M ./threads/dft-vrank-geq1.c -4 +3 M ./threads/hc2hc-dif.c -8 +8 M ./threads/hc2hc-dit.c -8 +8 M ./threads/rdft-vrank-geq1.c -4 +3 M ./threads/vrank-geq1-rdft2.c -4 +3 Sun Sep 22 15:00:59 EDT 2002 athena * [project @ 2002-09-22 19:00:59 by athena] More code compression A ./kernel/tensor8.c M ./dft/buffered.c -4 +3 M ./dft/direct.c -3 +2 M ./dft/indirect.c -4 +3 M ./dft/problem.c -5 +3 M ./dft/rank-geq2.c -9 +3 M ./kernel/Makefile.am -2 +2 M ./kernel/ifftw.h -1 +5 M ./kernel/tensor1.c -1 +6 M ./kernel/tensor4.c -1 +6 M ./kernel/tensor8.c +35 M ./rdft/buffered.c -4 +3 M ./rdft/dft-r2hc.c -2 +2 M ./rdft/direct.c -3 +2 M ./rdft/hc2hc.c -7 +6 M ./rdft/indirect.c -4 +3 M ./rdft/problem.c -7 +4 M ./rdft/problem2.c -5 +3 M ./rdft/rank-geq2-rdft2.c -9 +3 M ./rdft/rank-geq2.c -13 +5 Sun Sep 22 13:27:46 EDT 2002 athena * [project @ 2002-09-22 17:27:46 by athena] Smaller code size. M ./kernel/ifftw.h -2 +4 M ./kernel/solver.c -1 +6 Sun Sep 22 12:50:36 EDT 2002 athena * [project @ 2002-09-22 16:50:36 by athena] Started unification of rader A ./dft/rader-omega.c M ./dft/Makefile.am -3 +3 M ./dft/dft.h -1 +5 M ./dft/rader-omega.c +57 M ./dft/rader.c -38 +3 M ./rdft/rader-hc2hc.c -37 +3 Sun Sep 22 12:35:30 EDT 2002 athena * [project @ 2002-09-22 16:35:30 by athena] Typo M ./rdft/rdft2-radix2.c -2 +2 Sun Sep 22 12:25:20 EDT 2002 athena * [project @ 2002-09-22 16:25:20 by athena] Changed protocol for destroy_plan so as to save space. M ./dft/buffered.c -8 +4 M ./dft/ct.c -2 +1 M ./dft/direct.c -2 +1 M ./dft/generic.c -3 +1 M ./dft/indirect.c -6 +3 M ./dft/nop.c -7 +2 M ./dft/rader.c -7 +3 M ./dft/rank-geq2.c -6 +3 M ./dft/rank0.c -7 +2 M ./dft/vrank-geq1.c -2 +1 M ./dft/vrank2-transpose.c -7 +2 M ./dft/vrank3-transpose.c -7 +2 M ./kernel/ifftw.h -2 +2 M ./kernel/plan.c -3 +11 M ./kernel/problem.c -2 +3 M ./rdft/buffered.c -8 +4 M ./rdft/buffered2.c -6 +3 M ./rdft/dft-r2hc.c -2 +1 M ./rdft/dht-r2hc.c -2 +1 M ./rdft/dht-rader.c -7 +3 M ./rdft/direct.c -2 +1 M ./rdft/direct2.c -2 +1 M ./rdft/generic.c -3 +1 M ./rdft/hc2hc.c -14 +7 M ./rdft/indirect.c -6 +3 M ./rdft/nop.c -7 +2 M ./rdft/nop2.c -7 +2 M ./rdft/rader-hc2hc.c -9 +4 M ./rdft/rank-geq2-rdft2.c -6 +3 M ./rdft/rank-geq2.c -6 +3 M ./rdft/rank0.c -7 +2 M ./rdft/rdft-dht.c -2 +1 M ./rdft/rdft2-radix2.c -2 +1 M ./rdft/vrank-geq1-rdft2.c -2 +1 M ./rdft/vrank-geq1.c -2 +1 M ./rdft/vrank2-transpose.c -7 +2 M ./rdft/vrank3-transpose.c -7 +2 M ./reodft/redft00e-r2hc.c -2 +1 M ./reodft/reodft010e-r2hc.c -2 +1 M ./reodft/reodft11e-r2hc.c -2 +1 M ./reodft/rodft00e-r2hc.c -2 +1 M ./threads/dft-vrank-geq1.c -4 +2 M ./threads/rdft-vrank-geq1.c -4 +2 M ./threads/vrank-geq1-rdft2.c -4 +2 Sun Sep 22 11:08:57 EDT 2002 athena * [project @ 2002-09-22 15:08:57 by athena] Introduced convenient function X(mkplan_d) M ./dft/buffered.c -31 +18 M ./dft/ct.c -5 +2 M ./dft/generic.c -10 +5 M ./dft/indirect.c -14 +8 M ./dft/rader.c -38 +20 M ./dft/rank-geq2.c -16 +12 M ./dft/vrank-geq1.c -9 +7 M ./kernel/ifftw.h -2 +4 M ./kernel/planner.c -1 +8 M ./rdft/buffered.c -30 +20 M ./rdft/buffered2.c -9 +3 M ./rdft/dft-r2hc.c -7 +5 M ./rdft/dht-r2hc.c -7 +4 M ./rdft/dht-rader.c -23 +12 M ./rdft/generic.c -4 +1 M ./rdft/hc2hc.c -16 +10 M ./rdft/indirect.c -14 +9 M ./rdft/rader-hc2hc.c -41 +21 M ./rdft/rank-geq2-rdft2.c -12 +8 M ./rdft/rank-geq2.c -16 +12 M ./rdft/rdft-dht.c -5 +3 M ./rdft/rdft2-radix2.c -7 +3 M ./rdft/vrank-geq1-rdft2.c -9 +7 M ./rdft/vrank-geq1.c -10 +7 M ./reodft/redft00e-r2hc.c -3 +2 M ./reodft/reodft010e-r2hc.c -3 +2 M ./reodft/reodft11e-r2hc.c -3 +2 M ./reodft/rodft00e-r2hc.c -3 +2 M ./threads/ct-dit.c -4 +4 M ./threads/dft-vrank-geq1.c -12 +10 M ./threads/hc2hc-dif.c -4 +4 M ./threads/hc2hc-dit.c -4 +4 M ./threads/rdft-vrank-geq1.c -11 +9 M ./threads/vrank-geq1-rdft2.c -11 +9 Sun Sep 22 10:21:36 EDT 2002 athena * [project @ 2002-09-22 14:21:36 by athena] Split tensor/md5 into separate files to allow independent linking and/or prevent undesidred inlining A ./kernel/md5-1.c A ./kernel/tensor1.c A ./kernel/tensor2.c A ./kernel/tensor3.c A ./kernel/tensor4.c A ./kernel/tensor5.c A ./kernel/tensor7.c M ./kernel/Makefile.am -3 +4 M ./kernel/md5-1.c +53 M ./kernel/md5.c -31 M ./kernel/tensor.c -270 +2 M ./kernel/tensor1.c +32 M ./kernel/tensor2.c +37 M ./kernel/tensor3.c +46 M ./kernel/tensor4.c +68 M ./kernel/tensor5.c +93 M ./kernel/tensor7.c +127 Sun Sep 22 09:49:09 EDT 2002 athena * [project @ 2002-09-22 13:49:08 by athena] Treat all tensors as dynamically allocated objects. They were dynamically allocated in part anyway, so there is no point in complicating the object code with the clumsy calling conventions for by-value structs. M ./dft/buffered.c -19 +19 M ./dft/ct-dif.c -5 +5 M ./dft/ct-dit.c -5 +5 M ./dft/ct-ditbuf.c -4 +5 M ./dft/ct-ditf.c -5 +5 M ./dft/ct.c -14 +14 M ./dft/dft.h -4 +4 M ./dft/direct.c -13 +13 M ./dft/generic.c -7 +7 M ./dft/indirect.c -18 +16 M ./dft/nop.c -5 +5 M ./dft/problem.c -12 +12 M ./dft/rader.c -12 +12 M ./dft/rank-geq2.c -22 +22 M ./dft/rank0.c -14 +14 M ./dft/vrank-geq1.c -11 +11 M ./dft/vrank2-transpose.c -9 +9 M ./dft/vrank3-transpose.c -12 +12 M ./dft/zero.c -18 +20 M ./kernel/ifftw.h -13 +22 M ./kernel/tensor.c -92 +102 M ./rdft/buffered.c -23 +21 M ./rdft/buffered2.c -24 +24 M ./rdft/dft-r2hc.c -11 +11 M ./rdft/dht-r2hc.c -6 +6 M ./rdft/dht-rader.c -7 +7 M ./rdft/direct.c -18 +18 M ./rdft/direct2.c -16 +16 M ./rdft/generic.c -8 +8 M ./rdft/hc2hc-buf.c -4 +5 M ./rdft/hc2hc-dif.c -5 +5 M ./rdft/hc2hc-dit.c -5 +5 M ./rdft/hc2hc.c -23 +23 M ./rdft/indirect.c -20 +18 M ./rdft/nop.c -5 +5 M ./rdft/nop2.c -4 +4 M ./rdft/problem.c -39 +40 M ./rdft/problem2.c -34 +34 M ./rdft/rader-hc2hc.c -11 +11 M ./rdft/rank-geq2-rdft2.c -23 +23 M ./rdft/rank-geq2.c -28 +28 M ./rdft/rank0.c -10 +10 M ./rdft/rdft-dht.c -10 +10 M ./rdft/rdft.h -9 +9 M ./rdft/rdft2-radix2.c -21 +21 M ./rdft/vrank-geq1-rdft2.c -11 +11 M ./rdft/vrank-geq1.c -11 +11 M ./rdft/vrank2-transpose.c -9 +9 M ./rdft/vrank3-transpose.c -12 +12 M ./reodft/redft00e-r2hc.c -10 +10 M ./reodft/reodft010e-r2hc.c -10 +10 M ./reodft/reodft11e-r2hc.c -10 +10 M ./reodft/rodft00e-r2hc.c -10 +10 M ./tests/debug.h -2 +2 M ./tests/dotens.c -4 +4 M ./tests/dotens2.c -5 +5 M ./tests/verify-dft.c -10 +11 M ./tests/verify-lib.c -15 +16 M ./tests/verify-rdft.c -43 +47 M ./tests/verify-reodft.c -28 +28 M ./tests/verify.h -3 +3 Sat Sep 21 18:24:55 EDT 2002 stevenj * [project @ 2002-09-21 22:24:55 by stevenj] typo M ./kernel/ifftw.h -2 +2 Sat Sep 21 18:10:07 EDT 2002 athena * [project @ 2002-09-21 22:10:07 by athena] Avoid generating NaN when n = 0. M ./tests/verify-lib.c -13 +16 Sat Sep 21 18:04:05 EDT 2002 athena * [project @ 2002-09-21 22:04:05 by athena] Saved more. M ./dft/dft.h -2 +2 M ./dft/problem.c -5 +5 M ./dft/rank-geq2.c -2 +2 M ./rdft/dft-r2hc.c -2 +2 M ./rdft/dht-r2hc.c -2 +2 M ./rdft/hc2hc.c -5 +7 M ./rdft/problem.c -11 +11 M ./rdft/problem2.c -5 +5 M ./rdft/rank-geq2.c -2 +2 M ./rdft/rdft-dht.c -3 +3 M ./rdft/rdft.h -4 +4 M ./reodft/redft00e-r2hc.c -2 +2 M ./reodft/reodft010e-r2hc.c -2 +2 M ./reodft/reodft11e-r2hc.c -2 +2 M ./reodft/rodft00e-r2hc.c -2 +2 M ./threads/dft-vrank-geq1.c -2 +2 Sat Sep 21 17:47:36 EDT 2002 athena * [project @ 2002-09-21 21:47:35 by athena] Save 1200 bytes of object code. Do not pass structs by value whenever practical, because the calling protocol generates clumsy code. M ./dft/buffered.c -6 +6 M ./dft/ct.c -5 +5 M ./dft/direct.c -3 +3 M ./dft/indirect.c -12 +12 M ./dft/nop.c -2 +2 M ./dft/problem.c -11 +11 M ./dft/rank-geq2.c -21 +21 M ./dft/vrank-geq1.c -6 +6 M ./dft/vrank3-transpose.c -8 +8 M ./kernel/ifftw.h -20 +22 M ./kernel/pickdim.c -9 +9 M ./kernel/print.c -2 +2 M ./kernel/tensor.c -69 +71 M ./rdft/buffered.c -6 +6 M ./rdft/buffered2.c -2 +2 M ./rdft/dft-r2hc.c -3 +3 M ./rdft/direct.c -3 +3 M ./rdft/hc2hc.c -9 +9 M ./rdft/indirect.c -14 +14 M ./rdft/nop.c -2 +2 M ./rdft/problem.c -15 +15 M ./rdft/problem2.c -15 +15 M ./rdft/rank-geq2-rdft2.c -19 +19 M ./rdft/rank-geq2.c -27 +27 M ./rdft/rdft-dht.c -3 +3 M ./rdft/rdft.h -2 +2 M ./rdft/rdft2-radix2.c -7 +7 M ./rdft/vrank-geq1-rdft2.c -6 +7 M ./rdft/vrank-geq1.c -6 +6 M ./rdft/vrank3-transpose.c -8 +8 M ./reodft/redft00e-r2hc.c -2 +2 M ./reodft/reodft010e-r2hc.c -2 +2 M ./reodft/reodft11e-r2hc.c -2 +2 M ./reodft/rodft00e-r2hc.c -2 +2 M ./tests/verify-dft.c -6 +6 M ./tests/verify-lib.c -2 +2 M ./tests/verify-rdft.c -21 +21 M ./tests/verify-reodft.c -10 +10 M ./threads/dft-vrank-geq1.c -3 +3 M ./threads/rdft-vrank-geq1.c -3 +3 M ./threads/vrank-geq1-rdft2.c -3 +4 Sat Sep 21 12:10:21 EDT 2002 athena * [project @ 2002-09-21 16:10:21 by athena] Do not allocate buffers for rader omegas. Let the planner do it if necessary. M ./rdft/dht-rader.c -15 +6 Sat Sep 21 12:03:46 EDT 2002 athena * [project @ 2002-09-21 16:03:46 by athena] Check rank *before* reading kind[0], which may be undefined if rnk < 1 M ./tests/verify-rdft.c -2 +2 M ./tests/verify-reodft.c -2 +2 Sat Sep 21 11:48:50 EDT 2002 athena * [project @ 2002-09-21 15:48:50 by athena] Second step towards rader unification. M ./dft/rader.c -17 +7 M ./rdft/rader-hc2hc.c -1 +1 Sat Sep 21 11:37:06 EDT 2002 athena * [project @ 2002-09-21 15:37:06 by athena] First step towards unification of Rader code A ./kernel/rader.c M ./dft/rader.c -57 +12 M ./kernel/Makefile.am -3 +3 M ./kernel/ifftw.h -1 +9 M ./kernel/rader.c +68 M ./rdft/dht-rader.c -50 +6 M ./rdft/rader-hc2hc.c -57 +11 Sat Sep 21 07:58:11 EDT 2002 athena * [project @ 2002-09-21 11:58:11 by athena] Fix ugliness condition for cooley-tukey. A ./kernel/ct.c M ./dft/ct-dif.c -5 +3 M ./dft/ct-dit.c -4 +2 M ./dft/ct-ditbuf.c -5 +3 M ./kernel/Makefile.am -4 +5 M ./kernel/ct.c +31 M ./kernel/ifftw.h -1 +2 M ./kernel/planner.c -3 +2 M ./rdft/dht-r2hc.c -3 +3 M ./rdft/dht-rader.c -3 +3 M ./rdft/hc2hc-buf.c -9 +3 M ./rdft/hc2hc-dif.c -4 +2 M ./rdft/hc2hc-dit.c -4 +2 M ./rdft/rdft-dht.c -4 +2 M ./reodft/redft00e-r2hc.c -4 +2 M ./reodft/reodft010e-r2hc.c -4 +2 M ./reodft/reodft11e-r2hc.c -4 +2 M ./reodft/rodft00e-r2hc.c -4 +2 M ./threads/ct-dit.c -5 +3 M ./threads/hc2hc-dif.c -5 +3 M ./threads/hc2hc-dit.c -5 +3 Fri Sep 20 16:53:45 EDT 2002 athena * [project @ 2002-09-20 20:53:45 by athena] Removed RADER_MIN_GOOD and associated machinery M ./dft/rader.c -29 +6 M ./kernel/ifftw.h -2 +1 M ./rdft/dht-rader.c -14 +3 M ./rdft/rader-hc2hc.c -17 +5 Fri Sep 20 14:49:12 EDT 2002 athena * [project @ 2002-09-20 18:49:12 by athena] Proper cast M ./rdft/dht-r2hc.c -4 +3 Fri Sep 20 14:45:54 EDT 2002 athena * [project @ 2002-09-20 18:45:54 by athena] Typo M ./kernel/planner.c -2 +2 Fri Sep 20 14:38:13 EDT 2002 athena * [project @ 2002-09-20 18:38:13 by athena] Implemented NO_LARGE_GENERIC M ./dft/generic.c -1 +8 M ./kernel/ifftw.h -20 +23 M ./rdft/dht-rader.c -1 +1 M ./rdft/generic.c -2 +8 M ./rdft/rdft-dht.c -10 +2 M ./tests/bench.c +1 Thu Sep 19 07:48:25 EDT 2002 athena * [project @ 2002-09-19 11:48:24 by athena] Consistent macroization of NO_DHT_R2HC M ./kernel/ifftw.h -1 +2 M ./rdft/dht-r2hc.c -4 +4 Wed Sep 18 21:47:17 EDT 2002 athena * [project @ 2002-09-19 01:47:17 by athena] NO_DHT_R2HC is a planner flag, otherwise the EXHAUSTIVE planner loops. M ./kernel/ifftw.h -7 +6 M ./kernel/planner.c -10 +6 M ./rdft/dht-r2hc.c -3 +3 M ./tests/bench.c +1 Wed Sep 18 20:47:31 EDT 2002 athena * [project @ 2002-09-19 00:47:31 by athena] Resurrected NO_EXHAUSTIVE M ./kernel/ifftw.h -1 +3 M ./kernel/planner.c -8 +24 Wed Sep 18 19:31:57 EDT 2002 stevenj * [project @ 2002-09-18 23:31:57 by stevenj] au revoir, score() M ./threads/ct-dit.c -25 +22 M ./threads/dft-vrank-geq1.c -14 +10 M ./threads/hc2hc-dif.c -25 +21 M ./threads/hc2hc-dit.c -25 +21 M ./threads/rdft-vrank-geq1.c -14 +10 M ./threads/vrank-geq1-rdft2.c -14 +10 Wed Sep 18 19:31:05 EDT 2002 stevenj * [project @ 2002-09-18 23:31:05 by stevenj] eliminated unused M ./tests/bench.c +2 M ./tests/verify-reodft.c -7 +7 Wed Sep 18 18:28:44 EDT 2002 stevenj * [project @ 2002-09-18 22:28:44 by stevenj] capitalize and parenthesize SUBSUMES M ./kernel/planner.c -9 +8 Wed Sep 18 18:26:58 EDT 2002 stevenj * [project @ 2002-09-18 22:26:58 by stevenj] comment M ./kernel/ifftw.h -2 +2 Wed Sep 18 18:03:18 EDT 2002 athena * [project @ 2002-09-18 22:03:18 by athena] Use flags from wisdom if wisdom is applicable. M ./kernel/ifftw.h -2 +3 M ./kernel/planner.c -70 +32 Wed Sep 18 17:16:17 EDT 2002 athena * [project @ 2002-09-18 21:16:16 by athena] Removed score() machinery M ./dft/buffered.c -16 +11 M ./dft/ct-dif.c -23 +21 M ./dft/ct-dit.c -29 +25 M ./dft/ct-ditbuf.c -30 +24 M ./dft/ct-ditf.c -10 +2 M ./dft/direct.c -8 +2 M ./dft/generic.c -9 +7 M ./dft/indirect.c -9 +12 M ./dft/nop.c -8 +2 M ./dft/rader.c -23 +25 M ./dft/rank-geq2.c -14 +13 M ./dft/rank0.c -8 +2 M ./dft/vrank-geq1.c -30 +29 M ./dft/vrank2-transpose.c -8 +2 M ./dft/vrank3-transpose.c -13 +15 M ./kernel/ifftw.h -12 +2 M ./kernel/planner.c -31 +19 M ./rdft/buffered.c -18 +11 M ./rdft/buffered2.c -18 +11 M ./rdft/dft-r2hc.c -12 +14 M ./rdft/dht-r2hc.c -8 +8 M ./rdft/dht-rader.c -12 +12 M ./rdft/direct.c -9 +3 M ./rdft/direct2.c -9 +3 M ./rdft/generic.c -8 +7 M ./rdft/hc2hc-buf.c -30 +29 M ./rdft/hc2hc-dif.c -33 +25 M ./rdft/hc2hc-dit.c -32 +27 M ./rdft/indirect.c -9 +14 M ./rdft/nop.c -8 +2 M ./rdft/nop2.c -8 +2 M ./rdft/rader-hc2hc.c -13 +12 M ./rdft/rank-geq2-rdft2.c -15 +17 M ./rdft/rank-geq2.c -17 +19 M ./rdft/rank0.c -8 +2 M ./rdft/rdft-dht.c -11 +13 M ./rdft/rdft2-radix2.c -12 +2 M ./rdft/vrank-geq1-rdft2.c -32 +29 M ./rdft/vrank-geq1.c -31 +31 M ./rdft/vrank2-transpose.c -8 +2 M ./rdft/vrank3-transpose.c -18 +14 M ./reodft/redft00e-r2hc.c -6 +7 M ./reodft/reodft010e-r2hc.c -6 +7 M ./reodft/reodft11e-r2hc.c -6 +7 M ./reodft/rodft00e-r2hc.c -6 +7 M ./tests/bench.c -2 Wed Sep 18 14:12:21 EDT 2002 athena * [project @ 2002-09-18 18:12:21 by athena] Revised planner hack M ./kernel/planner.c -4 +7 Wed Sep 18 10:14:41 EDT 2002 athena * [project @ 2002-09-18 14:14:41 by athena] Fix warning M ./simd/simd-altivec.h +2 Tue Sep 17 17:54:07 EDT 2002 athena * [project @ 2002-09-17 21:54:07 by athena] Type qualifiers. M ./dft/indirect.c -2 +2 M ./rdft/indirect.c -2 +2 Tue Sep 17 16:17:55 EDT 2002 athena * [project @ 2002-09-17 20:17:55 by athena] ESTIMATE is no longer subsumed by everything else. M ./kernel/planner.c -2 +1 Tue Sep 17 10:55:15 EDT 2002 athena * [project @ 2002-09-17 14:55:15 by athena] NO_BUFFERING is a planner flag, not a problem flag M ./dft/indirect.c -2 +2 M ./rdft/indirect.c -2 +2 Tue Sep 17 09:36:16 EDT 2002 athena * [project @ 2002-09-17 13:36:16 by athena] Maintain flags in canonical form. M ./kernel/ifftw.h -4 +6 M ./kernel/planner.c -13 +15 Tue Sep 17 09:09:57 EDT 2002 athena * [project @ 2002-09-17 13:09:56 by athena] In dramatic break with tradition, SUBSUME is now a partial order. I swear. M ./kernel/ifftw.h -4 +1 M ./kernel/planner.c -15 +54 Tue Sep 17 07:29:00 EDT 2002 athena * [project @ 2002-09-17 11:29:00 by athena] Added comment M ./kernel/planner.c -1 +3 Tue Sep 17 07:27:17 EDT 2002 athena * [project @ 2002-09-17 11:27:17 by athena] Inverted ESTIMATE flag, renamed USE_SCORE for consistency with the convention that 0 subsumes 1. M ./kernel/ifftw.h -9 +6 M ./kernel/planner.c -10 +7 M ./tests/bench.c +2 Tue Sep 17 02:50:15 EDT 2002 stevenj * [project @ 2002-09-17 06:50:15 by stevenj] NO_INDIRECT -> NO_INDIRECT_OP (out-of-place only) M ./dft/indirect.c -4 +5 M ./kernel/ifftw.h -3 +3 M ./rdft/indirect.c -4 +5 M ./tests/bench.c -1 +1 Tue Sep 17 00:40:04 EDT 2002 stevenj * [project @ 2002-09-17 04:40:04 by stevenj] hpux needs -D_REENTRANT (thanks to Clinton Roy for the bug report) M ./acx_pthread.m4 -2 +2 Mon Sep 16 23:54:34 EDT 2002 athena * [project @ 2002-09-17 03:54:34 by athena] Oops. M ./kernel/planner.c -2 +2 Mon Sep 16 23:44:47 EDT 2002 athena * [project @ 2002-09-17 03:44:47 by athena] Yet another attempt at getting the planner right. M ./kernel/ifftw.h -3 +3 M ./kernel/planner.c -28 +27 Mon Sep 16 21:56:14 EDT 2002 athena * [project @ 2002-09-17 01:56:14 by athena] Better coding. M ./kernel/planner.c -21 +11 Mon Sep 16 21:51:06 EDT 2002 athena * [project @ 2002-09-17 01:51:06 by athena] NO_UGLY is no longer a flag, but a separate planner field that does not interfere with wisdom. M ./kernel/ifftw.h -3 +2 M ./kernel/planner.c -14 +15 Mon Sep 16 19:04:41 EDT 2002 athena * [project @ 2002-09-16 23:04:41 by athena] Did not compile without FFTW_DEBUG M ./tests/verify-reodft.c -3 +1 Mon Sep 16 18:37:06 EDT 2002 athena * [project @ 2002-09-16 22:37:06 by athena] Changed scoring mechanism. M ./kernel/ifftw.h -5 +4 M ./kernel/plan.c -5 +1 M ./kernel/planner.c -51 +28 M ./tests/bench.c -5 +2 Mon Sep 16 17:13:45 EDT 2002 athena * [project @ 2002-09-16 21:13:45 by athena] Count infeasible plans M ./kernel/planner.c -4 +12 Mon Sep 16 16:36:12 EDT 2002 athena * [project @ 2002-09-16 20:36:12 by athena] curse subsumed plans before export M ./kernel/planner.c -27 +35 Mon Sep 16 15:40:46 EDT 2002 stevenj * [project @ 2002-09-16 19:40:46 by stevenj] removed ESTIMATE_BIT vs. ESTIMATE... ESTIMATE | IMPATIENT is a UI issue M ./kernel/ifftw.h -6 +4 M ./kernel/planner.c -2 +2 Mon Sep 16 15:31:39 EDT 2002 stevenj * [project @ 2002-09-16 19:31:39 by stevenj] cleanup M ./rdft/buffered2.c -6 +3 Mon Sep 16 15:28:47 EDT 2002 stevenj * [project @ 2002-09-16 19:28:47 by stevenj] use CONSERVE_MEMORY flag to prevent buffered for large sizes M ./dft/buffered.c -4 +7 M ./rdft/buffered.c -4 +7 M ./rdft/buffered2.c -5 +9 Mon Sep 16 15:16:16 EDT 2002 stevenj * [project @ 2002-09-16 19:16:16 by stevenj] moved NO_DHT_R2HC back into planner flags: there's no reason we would want this flag to block plan reuse M ./kernel/ifftw.h -4 +4 Mon Sep 16 14:59:14 EDT 2002 stevenj * [project @ 2002-09-16 18:59:14 by stevenj] whoops, commas M ./kernel/ifftw.h -3 +3 Mon Sep 16 14:58:26 EDT 2002 stevenj * [project @ 2002-09-16 18:58:26 by stevenj] problem_flags == checked in applicable, planner_flags == checked in score M ./kernel/ifftw.h -5 +7 Mon Sep 16 14:53:16 EDT 2002 stevenj * [project @ 2002-09-16 18:53:16 by stevenj] ESTIMATE should not *include* all impatience flags, even if it subsumes them; some impatience flags, like NO_INDIRECT, might make a problem unsolvable M ./kernel/ifftw.h -6 +6 M ./kernel/planner.c -3 +4 Mon Sep 16 00:56:29 EDT 2002 stevenj * [project @ 2002-09-16 04:56:29 by stevenj] quotatio marks M ./kernel/planner.c -3 +3 Sun Sep 15 23:55:44 EDT 2002 stevenj * [project @ 2002-09-16 03:55:44 by stevenj] delete blank line M ./kernel/planner.c -2 +1 Sun Sep 15 23:51:14 EDT 2002 stevenj * [project @ 2002-09-16 03:51:14 by stevenj] substitution M ./kernel/planner.c -2 +2 Sun Sep 15 23:49:50 EDT 2002 stevenj * [project @ 2002-09-16 03:49:50 by stevenj] note that we are not GNUlly correct M ./kernel/planner.c -1 +10 Sun Sep 15 23:41:01 EDT 2002 stevenj * [project @ 2002-09-16 03:41:01 by stevenj] indenting M ./kernel/planner.c -3 +3 Sun Sep 15 23:37:46 EDT 2002 stevenj * [project @ 2002-09-16 03:37:46 by stevenj] more jokes M ./kernel/planner.c -2 +5 Sun Sep 15 23:20:14 EDT 2002 stevenj * [project @ 2002-09-16 03:20:14 by stevenj] NONTHREADED_ICKYP includes nthr > 1 check M ./dft/ct-dit.c -2 +2 M ./dft/vrank-geq1.c -2 +2 M ./kernel/ifftw.h -2 +3 M ./rdft/hc2hc-dif.c -2 +2 M ./rdft/hc2hc-dit.c -2 +2 M ./rdft/vrank-geq1-rdft2.c -2 +2 M ./rdft/vrank-geq1.c -2 +2 Sun Sep 15 22:56:44 EDT 2002 stevenj * [project @ 2002-09-16 02:56:44 by stevenj] use md5sig M ./kernel/md5.c -1 +1 Sun Sep 15 22:55:41 EDT 2002 stevenj * [project @ 2002-09-16 02:55:41 by stevenj] md5sig typedef M ./kernel/ifftw.h -2 +4 M ./kernel/planner.c -9 +9 Sun Sep 15 22:35:13 EDT 2002 stevenj * [project @ 2002-09-16 02:35:13 by stevenj] updated M ./ChangeLog +324 Sun Sep 15 22:30:26 EDT 2002 stevenj * [project @ 2002-09-16 02:30:26 by stevenj] partially-ordered impatience M ./dft/buffered.c -2 +2 M ./dft/ct-dif.c -3 +3 M ./dft/ct-dit.c -3 +3 M ./dft/ct-ditbuf.c -2 +2 M ./dft/ct.c -5 +1 M ./dft/indirect.c -4 +4 M ./dft/rank-geq2.c -3 +2 M ./dft/vrank-geq1.c -13 +3 M ./kernel/ifftw.h -18 +51 M ./kernel/planner.c -11 +49 M ./rdft/buffered.c -2 +2 M ./rdft/buffered2.c -2 +2 M ./rdft/dft-r2hc.c -2 +2 M ./rdft/dht-r2hc.c -3 +3 M ./rdft/hc2hc-buf.c -4 +3 M ./rdft/hc2hc-dif.c -4 +4 M ./rdft/hc2hc-dit.c -3 +3 M ./rdft/hc2hc.c -5 +1 M ./rdft/indirect.c -4 +4 M ./rdft/rank-geq2-rdft2.c -3 +2 M ./rdft/rank-geq2.c -3 +2 M ./rdft/rdft-dht.c -3 +3 M ./rdft/rdft2-radix2.c -2 +2 M ./rdft/vrank-geq1-rdft2.c -13 +3 M ./rdft/vrank-geq1.c -13 +3 M ./tests/bench.c -3 M ./threads/dft-vrank-geq1.c -13 +2 M ./threads/hc2hc-dif.c -2 +2 M ./threads/rdft-vrank-geq1.c -12 +2 M ./threads/vrank-geq1-rdft2.c -12 +2 Sat Sep 14 19:47:56 EDT 2002 athena * [project @ 2002-09-14 23:47:56 by athena] Removed all that planner inheritance crap. M ./kernel/Makefile.am -4 +3 M ./kernel/ifftw.h -15 +5 R ./kernel/planner-naive.c R ./kernel/planner-score.c M ./kernel/planner.c -35 +128 M ./tests/bench.c -2 +2 Sat Sep 14 16:35:28 EDT 2002 stevenj * [project @ 2002-09-14 20:35:28 by stevenj] string.h is used for more than strlen M ./kernel/planner.c -2 +2 Sat Sep 14 12:19:13 EDT 2002 athena * [project @ 2002-09-14 16:19:13 by athena] Reduced hashtable size by 1/6 (on 32-bit machines) at the expense of messier planner. M ./kernel/ifftw.h -13 +17 M ./kernel/planner.c -80 +87 Sat Sep 14 08:31:29 EDT 2002 athena * [project @ 2002-09-14 12:31:29 by athena] Only print wisdom if verbose > 3 M ./tests/bench.c -2 +4 Sat Sep 14 07:56:56 EDT 2002 athena * [project @ 2002-09-14 11:56:56 by athena] Changed syntax of temporaries to avoid shadowing library functions (which is harmless but I hate the warning) M ./genfft/variable.ml -2 +2 M ./genfft-k7/variable.ml -1 +1 Fri Sep 13 23:07:39 EDT 2002 stevenj * [project @ 2002-09-14 03:07:39 by stevenj] only add warnings in debug/maintainer mode, and add a few more warning flags; eliminate more warnings; add support for posix_memalign (broken in glibc, grrr) M ./acinclude.m4 -1 +1 M ./configure.ac -1 +9 M ./dft/rader.c -1 +1 M ./kernel/alloc.c -3 +16 M ./kernel/assert.c -2 +2 M ./kernel/ifftw.h -5 +5 M ./kernel/md5.c -1 +1 M ./kernel/planner-score.c -11 +13 M ./kernel/primes.c -4 +4 M ./kernel/scan.c -2 +2 M ./libbench/bench-user.h -2 +2 M ./libbench/bench.h -3 +4 M ./libbench/report.c -5 +5 M ./libbench/timer.c -1 +2 M ./libbench/util.c -14 +26 M ./libbench/verify.c -4 +1 M ./rdft/rader-hc2hc.c -6 +6 M ./tests/bench.c -3 M ./tests/verify-lib.c -1 +4 Fri Sep 13 21:57:50 EDT 2002 athena * [project @ 2002-09-14 01:57:50 by athena] Explicit cast M ./kernel/twiddle.c -2 +2 Fri Sep 13 21:54:50 EDT 2002 athena * [project @ 2002-09-14 01:54:50 by athena] Use double-hashing. This allows a slightly higher load factor at the expense of a messier computation of the hashtable size. M ./kernel/ifftw.h -1 +2 M ./kernel/planner.c -57 +63 M ./kernel/primes.c -1 +7 Fri Sep 13 17:53:13 EDT 2002 stevenj * [project @ 2002-09-13 21:53:13 by stevenj] typo M ./genfft/magic.ml -2 +2 Fri Sep 13 15:36:07 EDT 2002 athena * [project @ 2002-09-13 19:36:07 by athena] Slight change in hash table growth functions. M ./kernel/planner.c -3 +12 Fri Sep 13 14:58:22 EDT 2002 athena * [project @ 2002-09-13 18:58:22 by athena] More statistics. M ./kernel/ifftw.h -1 +2 M ./kernel/planner.c -1 +8 Fri Sep 13 10:13:02 EDT 2002 athena * [project @ 2002-09-13 14:13:02 by athena] Clearer logic. M ./kernel/planner.c -2 +3 Fri Sep 13 10:11:10 EDT 2002 athena * [project @ 2002-09-13 14:11:10 by athena] Oops. M ./kernel/planner.c -2 +1 Fri Sep 13 09:31:46 EDT 2002 athena * [project @ 2002-09-13 13:31:46 by athena] Cleaned up M ./kernel/planner.c -5 +6 Fri Sep 13 09:16:07 EDT 2002 athena * [project @ 2002-09-13 13:16:07 by athena] Deal properly with infeasible problems. M ./kernel/planner.c -4 +7 Fri Sep 13 07:15:06 EDT 2002 athena * [project @ 2002-09-13 11:15:06 by athena] Redundantly initialize hash table to prevent valgrind warnings. M ./kernel/planner.c -3 +7 Thu Sep 12 19:00:22 EDT 2002 athena * [project @ 2002-09-12 23:00:22 by athena] Removed relics from past. M ./kernel/md5.c -5 +3 Thu Sep 12 18:53:44 EDT 2002 athena * [project @ 2002-09-12 22:53:44 by athena] md5hash a problem only once. M ./kernel/ifftw.h -6 +11 M ./kernel/planner.c -39 +32 Thu Sep 12 16:33:49 EDT 2002 athena * [project @ 2002-09-12 20:33:49 by athena] Renamed k7 codelets M ./genfft-k7/genUtil.ml -1 +1 M ./genfft-k7/gen_notw.ml -3 +3 M ./genfft-k7/gen_twiddle.ml -4 +4 Thu Sep 12 16:32:03 EDT 2002 stevenj * [project @ 2002-09-12 20:32:03 by stevenj] FORBID_DHT_R2HC -> DHT_R2HC_VERBOTEN for consistency M ./kernel/ifftw.h -2 +2 M ./rdft/dht-r2hc.c -3 +3 Thu Sep 12 16:28:43 EDT 2002 stevenj * [project @ 2002-09-12 20:28:43 by stevenj] removed obsolete macro M ./kernel/ifftw.h -3 +1 Thu Sep 12 16:20:39 EDT 2002 athena * [project @ 2002-09-12 20:20:39 by athena] Split flags in SIMD code. M ./dft/simd/n1b.c -1 +1 M ./dft/simd/n1f.c -1 +1 M ./dft/simd/t1b.c -1 +1 M ./dft/simd/t1f.c -1 +1 Thu Sep 12 16:18:51 EDT 2002 athena * [project @ 2002-09-12 20:18:51 by athena] Forgot to fix threads M ./threads/dft-vrank-geq1.c -5 +6 M ./threads/hc2hc-dif.c -2 +2 M ./threads/rdft-vrank-geq1.c -5 +6 M ./threads/vrank-geq1-rdft2.c -5 +6 Thu Sep 12 16:10:05 EDT 2002 athena * [project @ 2002-09-12 20:10:05 by athena] Split flags into planner_flags and problem_flags M ./dft/buffered.c -2 +2 M ./dft/ct-dif.c -3 +3 M ./dft/ct-dit.c -3 +3 M ./dft/ct-ditbuf.c -2 +2 M ./dft/ct.c -3 +3 M ./dft/indirect.c -4 +4 M ./dft/rader.c -1 +1 M ./dft/rank-geq2.c -2 +3 M ./dft/rank0.c -2 +2 M ./dft/vrank-geq1.c -7 +8 M ./kernel/ifftw.h -22 +21 M ./kernel/planner-naive.c -3 +3 M ./kernel/planner-score.c -3 +3 M ./kernel/planner.c -33 +28 M ./rdft/buffered.c -2 +2 M ./rdft/buffered2.c -3 +3 M ./rdft/dft-r2hc.c -2 +2 M ./rdft/dht-r2hc.c -3 +5 M ./rdft/dht-rader.c -1 +1 M ./rdft/hc2hc-buf.c -3 +3 M ./rdft/hc2hc-dif.c -4 +4 M ./rdft/hc2hc-dit.c -3 +3 M ./rdft/hc2hc.c -3 +3 M ./rdft/indirect.c -4 +4 M ./rdft/rank-geq2-rdft2.c -2 +3 M ./rdft/rank-geq2.c -2 +3 M ./rdft/rdft-dht.c -3 +3 M ./rdft/rdft2-radix2.c -2 +2 M ./rdft/vrank-geq1-rdft2.c -7 +8 M ./rdft/vrank-geq1.c -7 +8 M ./tests/bench.c -8 +8 Thu Sep 12 15:46:56 EDT 2002 stevenj * [project @ 2002-09-12 19:46:56 by stevenj] tetrameter M ./kernel/planner.c -2 +3 Thu Sep 12 15:11:21 EDT 2002 athena * [project @ 2002-09-12 19:11:21 by athena] Overwrite less impatient solutions properly. M ./kernel/planner.c -23 +15 Thu Sep 12 11:29:16 EDT 2002 athena * [project @ 2002-09-12 15:29:16 by athena] Oops. M ./kernel/planner.c -3 +3 Thu Sep 12 10:58:56 EDT 2002 athena * [project @ 2002-09-12 14:58:56 by athena] Keep less impatient solution in case of conflict. Paranoid cast to uint in certain places. M ./kernel/planner.c -4 +5 Thu Sep 12 10:02:51 EDT 2002 athena * [project @ 2002-09-12 14:02:51 by athena] Complete reimplementation of planner hash table. M ./kernel/ifftw.h -6 +7 M ./kernel/planner.c -143 +146 M ./tests/bench.c -2 +4 Thu Sep 12 07:58:45 EDT 2002 athena * [project @ 2002-09-12 11:58:45 by athena] planner->cnt was not properly decremented. M ./kernel/planner.c -13 +14 Wed Sep 11 17:52:39 EDT 2002 stevenj * [project @ 2002-09-11 21:52:39 by stevenj] typo M ./NEWS -1 +1 Mon Sep 9 17:10:45 EDT 2002 athena * [project @ 2002-09-09 21:10:45 by athena] Simplified M ./kernel/planner.c -15 +5 Mon Sep 9 17:03:32 EDT 2002 athena * [project @ 2002-09-09 21:03:32 by athena] Always overwrite old wisdom with new, in case the old is corrupt/conclicting. M ./kernel/planner.c -16 +8 Mon Sep 9 16:56:03 EDT 2002 stevenj * [project @ 2002-09-09 20:56:03 by stevenj] added quote/joke M ./kernel/plan.c -1 +5 Mon Sep 9 15:04:47 EDT 2002 athena * [project @ 2002-09-09 19:04:47 by athena] Completed wisdom import M ./kernel/ifftw.h -10 +11 M ./kernel/md5.c -1 +1 M ./kernel/planner.c -37 +54 M ./kernel/print.c -2 +3 M ./kernel/scan.c -73 +25 M ./tests/bench.c -3 +3 Mon Sep 9 10:14:22 EDT 2002 athena * [project @ 2002-09-09 14:14:22 by athena] Slight cleanup of md5 interface. M ./dft/problem.c -2 +2 M ./kernel/ifftw.h -2 +4 M ./kernel/md5.c -12 +24 M ./rdft/problem.c -2 +2 M ./rdft/problem2.c -2 +2 Tue Sep 3 22:32:43 EDT 2002 athena * [project @ 2002-09-04 02:32:43 by athena] More consistent protocol between planner and inferior. M ./kernel/planner-naive.c -2 +5 M ./kernel/planner-score.c -1 +2 M ./kernel/planner.c -3 +1 Tue Sep 3 21:08:30 EDT 2002 athena * [project @ 2002-09-04 01:08:30 by athena] I can't think of any situation where saving infeasible problems would be desirable. Removed relevant code. M ./kernel/planner.c -9 +3 Tue Sep 3 20:57:03 EDT 2002 athena * [project @ 2002-09-04 00:57:03 by athena] Encoder registrar's names in wisdom. Remove export_conf, since a separate program can now generate it. M ./kernel/ifftw.h -5 +5 M ./kernel/planner.c -108 +32 M ./kernel/solvtab.c -1 +2 M ./tests/bench.c -3 +1 Tue Sep 3 15:11:06 EDT 2002 athena * [project @ 2002-09-03 19:11:06 by athena] Fixed typo M ./kernel/planner.c -2 +2 Tue Sep 3 14:52:45 EDT 2002 athena * [project @ 2002-09-03 18:52:45 by athena] Fixed broken trochaic meter. M ./kernel/planner.c -2 +2 Tue Sep 3 09:49:50 EDT 2002 athena * [project @ 2002-09-03 13:49:50 by athena] Initialize planner->score. It is correct to leave it uninitialized, but I don't want people to send reports about purify complaining. M ./kernel/planner.c -1 +2 Tue Sep 3 09:03:46 EDT 2002 athena * [project @ 2002-09-03 13:03:46 by athena] More latin silliness M ./kernel/planner.c -3 +7 Mon Sep 2 17:57:32 EDT 2002 stevenj * [project @ 2002-09-02 21:57:32 by stevenj] updated M ./ChangeLog +193 Mon Sep 2 17:33:49 EDT 2002 stevenj * [project @ 2002-09-02 21:33:49 by stevenj] added clock() getseconds timer M ./kernel/timer.c -1 +16 Mon Sep 2 16:16:58 EDT 2002 athena * [project @ 2002-09-02 20:16:58 by athena] Oops M ./rdft/indirect.c -2 +1 Mon Sep 2 15:58:19 EDT 2002 athena * [project @ 2002-09-02 19:58:19 by athena] Experimental INDIRECT_VERBOTEN flag (not used) M ./dft/indirect.c -1 +3 M ./kernel/ifftw.h -5 +6 M ./rdft/indirect.c -3 +4 Mon Sep 2 15:36:21 EDT 2002 athena * [project @ 2002-09-02 19:36:21 by athena] Do not allow buffering in children of indirect solvers. M ./dft/buffered.c -1 +4 M ./dft/indirect.c -1 +3 M ./kernel/ifftw.h -1 +2 M ./rdft/buffered.c -1 +4 M ./rdft/buffered2.c -1 +4 M ./rdft/indirect.c -1 +3 Mon Sep 2 15:02:11 EDT 2002 athena * [project @ 2002-09-02 19:02:11 by athena] Oops M ./kernel/planner.c -2 +2 Mon Sep 2 14:32:28 EDT 2002 athena * [project @ 2002-09-02 18:32:28 by athena] Hash sizeof(R) as part of wisdom. M ./kernel/planner.c -1 +2 Mon Sep 2 13:47:57 EDT 2002 stevenj * [project @ 2002-09-02 17:47:57 by stevenj] added --enable-float synonym for --enable-single (since with have --enable-long-double) M ./configure.ac +1 Mon Sep 2 13:46:08 EDT 2002 athena * [project @ 2002-09-02 17:46:08 by athena] zerotens is now in its own file, so it does not cause dft to be linked in if only rdft is used. A ./dft/zero.c M ./dft/Makefile.am -1 +1 M ./dft/problem.c -25 +1 M ./dft/zero.c +49 Mon Sep 2 11:56:37 EDT 2002 athena * [project @ 2002-09-02 15:56:37 by athena] Removed unused var. M ./kernel/planner.c -2 +1 Mon Sep 2 11:55:33 EDT 2002 athena * [project @ 2002-09-02 15:55:33 by athena] Split insert() in preparation for wisdom import M ./kernel/planner.c -5 +11 Mon Sep 2 11:46:57 EDT 2002 athena * [project @ 2002-09-02 15:46:57 by athena] Moved debugging infrastructure to test directory so that it is not linked into the shared library. A ./tests/debug.h A ./tests/dotens.c A ./tests/dotens2.c A ./tests/verify-dft.c A ./tests/verify-lib.c A ./tests/verify-rdft.c A ./tests/verify-reodft.c A ./tests/verify.h M ./dft/Makefile.am -1 +1 M ./dft/dft.h -4 +1 R ./dft/verify.c M ./kernel/Makefile.am -6 +5 R ./kernel/dotens.c R ./kernel/dotens2.c M ./kernel/ifftw.h -17 +1 R ./kernel/verify-lib.c R ./kernel/verify.h M ./rdft/Makefile.am -2 +2 M ./rdft/rdft.h -4 +1 R ./rdft/verify.c M ./reodft/Makefile.am -1 +1 M ./reodft/reodft.h -4 +1 R ./reodft/verify.c M ./tests/Makefile.am -2 +7 M ./tests/bench.c +3 M ./tests/debug.h +18 M ./tests/dotens.c +48 M ./tests/dotens2.c +56 M ./tests/verify-dft.c +131 M ./tests/verify-lib.c +386 M ./tests/verify-rdft.c +420 M ./tests/verify-reodft.c +536 M ./tests/verify.h +83 Mon Sep 2 11:04:54 EDT 2002 athena * [project @ 2002-09-02 15:04:53 by athena] Reactivated wisdom export M ./kernel/planner.c -42 +9 M ./kernel/print.c -1 +12 Sun Sep 1 21:30:58 EDT 2002 athena * [project @ 2002-09-02 01:30:58 by athena] Dump errors to stderr, not stdout. M ./kernel/verify-lib.c -2 +3 Sun Sep 1 21:26:38 EDT 2002 athena * [project @ 2002-09-02 01:26:38 by athena] Removed traverse.c. traverse.c is no longer need for plan blessing. I figured out a way to avoid using it in planner-score.c, so the file is now redundant. M ./kernel/Makefile.am -2 +2 M ./kernel/ifftw.h -9 +2 M ./kernel/planner-score.c -30 +37 R ./kernel/traverse.c M ./tests/bench.c -20 Sun Sep 1 19:51:50 EDT 2002 athena * [project @ 2002-09-01 23:51:50 by athena] Removed code made obsolete by new MD5 scheme: problem equality tests, scanners, and associated list of problem kinds. M ./dft/conf.c -3 +1 M ./dft/dft.h -3 +1 M ./dft/problem.c -58 +3 M ./kernel/align.c -7 +1 M ./kernel/ifftw.h -24 +4 M ./kernel/planner.c -25 +10 M ./kernel/problem.c -11 +2 M ./kernel/scan.c -44 +2 M ./kernel/scanners.c -5 +5 M ./kernel/tensor.c -46 +1 M ./rdft/conf.c -4 +1 M ./rdft/problem.c -78 +3 M ./rdft/problem2.c -58 +3 M ./rdft/rdft.h -4 +1 Sun Sep 1 19:22:54 EDT 2002 athena * [project @ 2002-09-01 23:22:53 by athena] Started md5 implementation A ./kernel/md5.c M ./dft/problem.c -9 +9 M ./kernel/Makefile.am -1 +1 M ./kernel/ifftw.h -3 +29 M ./kernel/md5.c +164 M ./kernel/planner-naive.c -3 +3 M ./kernel/planner-score.c -7 +11 M ./kernel/planner.c -27 +47 M ./kernel/tensor.c -9 +7 M ./rdft/problem.c -12 +11 M ./rdft/problem2.c -10 +11 Sat Aug 31 14:00:04 EDT 2002 athena * [project @ 2002-08-31 18:00:04 by athena] Keep track of hit rate M ./kernel/ifftw.h -1 +2 M ./kernel/planner.c -3 +9 Sat Aug 31 12:44:04 EDT 2002 athena * [project @ 2002-08-31 16:44:04 by athena] Only dump when verbose > 4 M ./kernel/planner.c -2 +2 Sat Aug 31 09:55:57 EDT 2002 athena * [project @ 2002-08-31 13:55:57 by athena] Debugging infrastructure M ./dft/indirect.c -2 +1 M ./kernel/ifftw.h -5 +6 M ./kernel/plan.c -1 +6 M ./kernel/planner.c -7 +12 M ./tests/bench.c -3 +7 Sat Aug 31 09:21:48 EDT 2002 athena * [project @ 2002-08-31 13:21:48 by athena] Use debug infrastructure to dump planner. M ./kernel/planner.c -19 +10 M ./kernel/print.c -2 +5 Fri Aug 30 21:29:10 EDT 2002 athena * [project @ 2002-08-31 01:29:10 by athena] Do not store plans in planner, plus general planner cleanup. M ./kernel/alloc.c -2 +2 M ./kernel/ifftw.h -34 +37 M ./kernel/plan.c -15 +1 M ./kernel/planner-naive.c -7 +7 M ./kernel/planner-score.c -8 +7 M ./kernel/planner.c -103 +80 M ./kernel/scan.c -3 +3 M ./kernel/scanners.c -3 +3 M ./tests/bench.c -3 +16 Fri Aug 30 18:07:52 EDT 2002 stevenj * [project @ 2002-08-30 22:07:52 by stevenj] renamed IN_DHT_R2HC to the more general FORBID_DHT_R2HC M ./kernel/ifftw.h -2 +2 M ./rdft/dht-r2hc.c -3 +3 Fri Aug 30 18:07:21 EDT 2002 stevenj * [project @ 2002-08-30 22:07:21 by stevenj] eliminated unused var M ./kernel/planner.c -2 +1 Fri Aug 30 12:09:48 EDT 2002 athena * [project @ 2002-08-30 16:09:48 by athena] Score planner was not working correctly when using wisdom. Fixed. M ./kernel/planner-naive.c -2 +8 M ./kernel/planner-score.c -2 +13 M ./kernel/planner.c -13 +18 Fri Aug 30 08:20:48 EDT 2002 athena * [project @ 2002-08-30 12:20:48 by athena] Use hash table in debug malloc M ./kernel/alloc.c -12 +25 M ./kernel/ifftw.h -1 +2 M ./kernel/stride.c -2 +2 Fri Aug 30 02:45:15 EDT 2002 stevenj * [project @ 2002-08-30 06:45:15 by stevenj] listed some good stuff M ./NEWS +47 Fri Aug 30 02:17:30 EDT 2002 stevenj * [project @ 2002-08-30 06:17:30 by stevenj] timed planner M ./TODO +4 Fri Aug 30 02:07:00 EDT 2002 stevenj * [project @ 2002-08-30 06:07:00 by stevenj] fma? M ./TODO +2 Fri Aug 30 02:05:55 EDT 2002 stevenj * [project @ 2002-08-30 06:05:55 by stevenj] update M ./TODO -3 +3 Fri Aug 30 01:31:47 EDT 2002 stevenj * [project @ 2002-08-30 05:31:47 by stevenj] rader-dht -> dht-rader A ./rdft/dht-rader.c M ./rdft/Makefile.am -4 +4 M ./rdft/conf.c -2 +2 M ./rdft/dht-rader.c +422 R ./rdft/rader-dht.c M ./rdft/rdft.h -2 +2 Fri Aug 30 01:21:37 EDT 2002 stevenj * [project @ 2002-08-30 05:21:37 by stevenj] add DHT solver, and break up rader-dht and r2hc-hc2r A ./rdft/dht-r2hc.c A ./rdft/rdft-dht.c M ./kernel/ifftw.h -2 +3 M ./rdft/Makefile.am -6 +6 M ./rdft/buffered2.c -2 +4 M ./rdft/conf.c -4 +5 M ./rdft/dht-r2hc.c +151 R ./rdft/r2hc-hc2r.c M ./rdft/rader-dht.c -73 +21 M ./rdft/rank-geq2.c -1 +12 M ./rdft/rdft-dht.c +229 M ./rdft/rdft.h -3 +4 Thu Aug 29 23:20:35 EDT 2002 stevenj * [project @ 2002-08-30 03:20:35 by stevenj] another option M ./tests/bench.c +1 Thu Aug 29 22:55:29 EDT 2002 stevenj * [project @ 2002-08-30 02:55:29 by stevenj] generalized indirect solvers for fftw2-like buffering and more M ./dft/indirect.c -22 +35 M ./kernel/ifftw.h -1 +3 M ./kernel/tensor.c -13 +17 M ./rdft/indirect.c -18 +35 Thu Aug 29 18:08:16 EDT 2002 stevenj * [project @ 2002-08-29 22:08:16 by stevenj] tensor_max_index and tensor_min_stride are now both unsigned M ./dft/vrank-geq1.c -2 +3 M ./kernel/ifftw.h -3 +3 M ./kernel/tensor.c -11 +7 M ./rdft/vrank-geq1-rdft2.c -2 +3 M ./rdft/vrank-geq1.c -2 +3 Thu Aug 29 17:58:35 EDT 2002 stevenj * [project @ 2002-08-29 21:58:35 by stevenj] added iabs.c, and tensor_min_stride returns min absolute value A ./kernel/iabs.c M ./kernel/Makefile.am -5 +5 M ./kernel/iabs.c +28 M ./kernel/ifftw.h -1 +5 M ./kernel/tensor.c -12 +6 M ./rdft/buffered2.c -9 +4 M ./rdft/problem2.c -8 +3 Thu Aug 29 17:31:39 EDT 2002 stevenj * [project @ 2002-08-29 21:31:39 by stevenj] bug fix in cldrest hc2c/c2hc copy loops M ./rdft/buffered2.c -9 +14 Thu Aug 29 13:45:08 EDT 2002 athena * [project @ 2002-08-29 17:45:08 by athena] Added things to do. M ./TODO -1 +1 Thu Aug 29 13:10:04 EDT 2002 stevenj * [project @ 2002-08-29 17:10:04 by stevenj] added automake prereq M ./configure.ac -1 +1 Thu Aug 29 08:36:36 EDT 2002 athena * [project @ 2002-08-29 12:36:36 by athena] Use indexed addressing M ./rdft/rdft2-radix2.c -61 +61 Thu Aug 29 08:20:55 EDT 2002 athena * [project @ 2002-08-29 12:20:55 by athena] Ooops M ./libbench/verify.c -3 +6 M ./rdft/rdft2-radix2.c -4 +7 Thu Aug 29 07:45:37 EDT 2002 athena * [project @ 2002-08-29 11:45:37 by athena] Oops M ./kernel/ifftw.h -10 +10 Thu Aug 29 02:32:13 EDT 2002 stevenj * [project @ 2002-08-29 06:32:13 by stevenj] updates to win32 threads code (ick) M ./threads/threads.c -4 +22 Thu Aug 29 01:44:33 EDT 2002 stevenj * [project @ 2002-08-29 05:44:33 by stevenj] added threaded version A ./acx_pthread.m4 A ./threads/ A ./threads/Makefile.am A ./threads/conf.c A ./threads/ct-dit.c A ./threads/dft-vrank-geq1.c A ./threads/hc2hc-dif.c A ./threads/hc2hc-dit.c A ./threads/rdft-vrank-geq1.c A ./threads/threads.c A ./threads/threads.h A ./threads/vrank-geq1-rdft2.c M ./Makefile.am -2 +3 M ./acx_pthread.m4 +226 M ./configure.ac +56 M ./dft/ct-dif.c -2 +2 M ./dft/ct-dit.c -2 +5 M ./dft/ct-ditbuf.c -2 +2 M ./dft/ct-ditf.c -2 +2 M ./dft/ct.c -2 +3 M ./dft/ct.h -1 +2 M ./dft/dft.h -1 +4 M ./dft/kdft-dif.c -1 +5 M ./dft/kdft-dit.c -1 +5 M ./dft/vrank-geq1.c -1 +4 M ./kernel/alloc.c -42 +56 M ./kernel/ifftw.h -7 +18 M ./kernel/planner.c -15 +24 M ./rdft/hc2hc-buf.c -1 +3 M ./rdft/hc2hc-dif.c -1 +5 M ./rdft/hc2hc-dit.c -1 +5 M ./rdft/hc2hc.c -2 +3 M ./rdft/hc2hc.h -1 +2 M ./rdft/khc2hc-dif.c -1 +5 M ./rdft/khc2hc-dit.c -1 +5 M ./rdft/rdft.h -1 +4 M ./rdft/vrank-geq1-rdft2.c -1 +4 M ./rdft/vrank-geq1.c -1 +4 M ./tests/Makefile.am -2 +4 M ./tests/bench.c -1 +5 M ./threads/Makefile.am +15 M ./threads/conf.c +41 M ./threads/ct-dit.c +151 M ./threads/dft-vrank-geq1.c +249 M ./threads/hc2hc-dif.c +168 M ./threads/hc2hc-dit.c +166 M ./threads/rdft-vrank-geq1.c +246 M ./threads/threads.c +467 M ./threads/threads.h +53 M ./threads/vrank-geq1-rdft2.c +267 Wed Aug 28 19:47:21 EDT 2002 stevenj * [project @ 2002-08-28 23:47:21 by stevenj] fix make dist M ./kernel/Makefile.am -2 +3 Wed Aug 28 15:09:03 EDT 2002 stevenj * [project @ 2002-08-28 19:09:03 by stevenj] whoops, bugfix for inverse M ./rdft/rank-geq2-rdft2.c -4 +9 Wed Aug 28 14:50:34 EDT 2002 athena * [project @ 2002-08-28 18:50:34 by athena] Use C9x convention for naming (fftwf etc.). Removed installable header files since they will be part of the API. M ./Makefile.am -5 +5 M ./configure.ac -2 +9 M ./kernel/Makefile.am -1 R ./kernel/dfftw3.h R ./kernel/fftw3.h M ./kernel/ifftw.h -2 +17 R ./kernel/lfftw3.h R ./kernel/sfftw3.h M ./tests/Makefile.am -1 +1 Tue Aug 27 23:34:00 EDT 2002 stevenj * [project @ 2002-08-28 03:34:00 by stevenj] allow _1 variants to accept rnk 0 (sz 1) problems M ./rdft/problem.c -3 +3 Tue Aug 27 15:56:09 EDT 2002 stevenj * [project @ 2002-08-27 19:56:09 by stevenj] updated M ./ChangeLog +619 Mon Aug 26 20:14:56 EDT 2002 athena * [project @ 2002-08-27 00:14:56 by athena] Loop unroll is useless M ./dft/rank0.c -28 +6 Mon Aug 26 20:00:41 EDT 2002 athena * [project @ 2002-08-27 00:00:41 by athena] Use indexed addressing M ./dft/ct-ditbuf.c -12 +9 Mon Aug 26 19:46:46 EDT 2002 athena * [project @ 2002-08-26 23:46:46 by athena] Use indexed addressing in transpose routines. (Seems to be slightly better on athlon.) M ./dft/vrank2-transpose.c -16 +8 M ./dft/vrank3-transpose.c -15 +8 Mon Aug 26 12:59:44 EDT 2002 stevenj * [project @ 2002-08-26 16:59:44 by stevenj] added comment about stability M ./reodft/redft00e-r2hc.c -1 +2 M ./reodft/reodft11e-r2hc.c -1 +3 M ./reodft/rodft00e-r2hc.c -1 +2 Mon Aug 26 07:43:53 EDT 2002 athena * [project @ 2002-08-26 11:43:53 by athena] Approximate opcount M ./rdft/rdft2-radix2.c -6 +10 Mon Aug 26 06:38:49 EDT 2002 athena * [project @ 2002-08-26 10:38:49 by athena] Finished rdft2 via dft/rdft M ./dft/rank-geq2.c -2 +2 M ./rdft/rank-geq2.c -2 +2 M ./rdft/rdft2-radix2.c -9 +149 Mon Aug 26 00:15:59 EDT 2002 stevenj * [project @ 2002-08-26 04:15:59 by stevenj] some updates M ./TODO -9 +9 Mon Aug 26 00:05:53 EDT 2002 stevenj * [project @ 2002-08-26 04:05:52 by stevenj] rdft kind is now per-dimension, added rdft/rank-geq2 A ./rdft/rank-geq2.c M ./rdft/Makefile.am -2 +3 M ./rdft/buffered.c -3 +4 M ./rdft/buffered2.c -5 +5 M ./rdft/conf.c -3 +2 M ./rdft/dft-r2hc.c -2 +2 M ./rdft/direct.c -6 +6 M ./rdft/generic.c -5 +5 M ./rdft/hc2hc.c -10 +10 M ./rdft/indirect.c -2 +2 M ./rdft/problem.c -21 +75 M ./rdft/r2hc-hc2r.c -4 +4 M ./rdft/rader-dht.c -4 +4 M ./rdft/rader-hc2hc.c -4 +4 M ./rdft/rank-geq2.c +231 M ./rdft/rdft.h -4 +9 M ./rdft/rdft2-radix2.c -2 +2 M ./rdft/verify.c -5 +5 M ./reodft/redft00e-r2hc.c -3 +3 M ./reodft/reodft010e-r2hc.c -6 +6 M ./reodft/reodft11e-r2hc.c -5 +5 M ./reodft/rodft00e-r2hc.c -3 +3 M ./reodft/verify.c -4 +4 M ./tests/bench.c -2 +4 Sun Aug 25 22:45:38 EDT 2002 stevenj * [project @ 2002-08-26 02:45:38 by stevenj] added note M ./rdft/problem.c -2 +3 Sun Aug 25 22:28:12 EDT 2002 stevenj * [project @ 2002-08-26 02:28:12 by stevenj] must zero real sz M ./rdft/problem.c -2 +4 Sun Aug 25 22:06:52 EDT 2002 stevenj * [project @ 2002-08-26 02:06:52 by stevenj] unified pickdim funcs A ./kernel/pickdim.c M ./dft/rank-geq2.c -32 +10 M ./dft/vrank-geq1.c -45 +3 M ./kernel/Makefile.am -1 +1 M ./kernel/ifftw.h -1 +6 M ./kernel/pickdim.c +82 M ./rdft/vrank-geq1-rdft2.c -45 +3 M ./rdft/vrank-geq1.c -45 +3 M ./tests/bench.c -1 +1 Sun Aug 25 14:10:55 EDT 2002 fftw * [project @ 2002-08-25 18:10:55 by fftw] silence warnings M ./libbench/mp.c -14 M ./rdft/codelet.h -2 +2 M ./rdft/indirect.c -3 +1 M ./rdft/rank-geq2-rdft2.c -5 +2 M ./rdft/verify.c -10 +7 M ./reodft/reodft010e-r2hc.c -2 +2 M ./reodft/reodft11e-r2hc.c -7 +2 M ./reodft/verify.c -2 +2 Sun Aug 25 13:16:49 EDT 2002 athena * [project @ 2002-08-25 17:16:49 by athena] I had to add another planner flag to record whether pointers could become unaligned because of vrank-geq1 solvers (these solvers only plan the first element of a vector problem, but the second element may have a different alignment). This addition is ugly, but I don't see any way around it. M ./dft/codelet.h -3 +5 M ./dft/codelets/n.c -2 +3 M ./dft/codelets/t.c -2 +2 M ./dft/ct-dif.c -3 +3 M ./dft/ct-dit.c -3 +3 M ./dft/ct-ditbuf.c -3 +3 M ./dft/ct-ditf.c -2 +2 M ./dft/direct.c -5 +6 M ./dft/simd/n1b.c -1 +3 M ./dft/simd/n1f.c -1 +3 M ./dft/simd/t1b.c -1 +3 M ./dft/simd/t1f.c -1 +3 M ./dft/vrank-geq1.c -2 +11 M ./kernel/ifftw.h -4 +8 M ./rdft/vrank-geq1-rdft2.c -2 +8 M ./rdft/vrank-geq1.c -2 +7 Sun Aug 25 10:18:25 EDT 2002 athena * [project @ 2002-08-25 14:18:25 by athena] Added thoughts M ./TODO -1 +2 Sun Aug 25 10:08:59 EDT 2002 athena * [project @ 2002-08-25 14:08:59 by athena] Implemented rdft2 via vector rdft + radix2 step A ./rdft/rdft2-radix2.c M ./rdft/Makefile.am -1 +1 M ./rdft/conf.c -2 +2 M ./rdft/rdft.h -2 +2 R ./rdft/rdft2-dft.c M ./rdft/rdft2-radix2.c +325 Sat Aug 24 17:43:54 EDT 2002 athena * [project @ 2002-08-24 21:43:54 by athena] Stylistic changes M ./rdft/rdft2-dft.c -5 +5 Sat Aug 24 11:19:30 EDT 2002 athena * [project @ 2002-08-24 15:19:30 by athena] Simplified mktwiddle interface M ./dft/ct.c -7 +3 M ./dft/generic.c -5 +2 M ./kernel/ifftw.h -1 +3 M ./kernel/twiddle.c -1 +11 M ./rdft/generic.c -6 +3 M ./rdft/rdft2-dft.c -11 +5 M ./reodft/redft00e-r2hc.c -6 +2 M ./reodft/reodft010e-r2hc.c -5 +2 M ./reodft/reodft11e-r2hc.c -8 +3 M ./reodft/rodft00e-r2hc.c -5 +2 Sat Aug 24 11:05:08 EDT 2002 athena * [project @ 2002-08-24 15:05:08 by athena] Unification of certain vector computations. rdft2-dft is now a vector transform. M ./dft/ct-dif.c -2 +2 M ./dft/ct-dit.c -2 +2 M ./dft/ct.c -16 +2 M ./dft/ct.h -4 +1 M ./kernel/ifftw.h -1 +2 M ./kernel/tensor.c -1 +17 M ./rdft/hc2hc-dif.c -2 +2 M ./rdft/hc2hc-dit.c -2 +2 M ./rdft/hc2hc.c -18 +2 M ./rdft/hc2hc.h -4 +1 M ./rdft/rdft2-dft.c -54 +66 Fri Aug 23 20:21:25 EDT 2002 athena * [project @ 2002-08-24 00:21:25 by athena] Intel compiler seems to be still buggy M ./configure.ac +1 M ./simd/sse.c -1 +3 M ./simd/sse2.c -1 +3 Fri Aug 23 16:07:12 EDT 2002 athena * [project @ 2002-08-23 20:07:12 by athena] Streamlined twiddle protocol M ./dft/ct-dif.c -2 +3 M ./dft/ct-dit.c -2 +3 M ./dft/ct-ditbuf.c -2 +2 M ./dft/ct-ditf.c -2 +2 M ./dft/ct.c -13 +6 M ./dft/ct.h -3 +2 M ./dft/generic.c -16 +6 M ./dft/indirect.c -3 +1 M ./kernel/ifftw.h -3 +3 M ./kernel/twiddle.c -7 +14 M ./rdft/generic.c -18 +8 M ./rdft/hc2hc.c -10 +6 M ./rdft/rdft2-dft.c -19 +10 M ./reodft/redft00e-r2hc.c -15 +6 M ./reodft/reodft010e-r2hc.c -19 +10 M ./reodft/reodft11e-r2hc.c -21 +9 M ./reodft/rodft00e-r2hc.c -15 +6 Fri Aug 23 13:22:17 EDT 2002 athena * [project @ 2002-08-23 17:22:17 by athena] Implemented rdft2 via dft (forward only for now) A ./rdft/rdft2-dft.c M ./libbench/verify.c -1 +10 M ./rdft/Makefile.am -1 +1 M ./rdft/conf.c -1 +2 M ./rdft/rdft.h -1 +2 M ./rdft/rdft2-dft.c +213 Thu Aug 22 11:29:29 EDT 2002 athena * [project @ 2002-08-22 15:29:29 by athena] More cleanup of verify M ./kernel/verify-lib.c -21 +13 M ./libbench/verify.c -45 +12 Thu Aug 22 11:16:03 EDT 2002 athena * [project @ 2002-08-22 15:16:03 by athena] Changed error criterion because old one was too strict M ./kernel/verify-lib.c -40 +21 Thu Aug 22 11:15:17 EDT 2002 athena * [project @ 2002-08-22 15:15:17 by athena] Disable shared M ./bootstrap.sh -1 +1 Thu Aug 22 09:19:12 EDT 2002 athena * [project @ 2002-08-22 13:19:12 by athena] Added thoughts M ./TODO +4 Thu Aug 22 09:17:28 EDT 2002 athena * [project @ 2002-08-22 13:17:28 by athena] Oops M ./dft/generic.c -1 +2 Thu Aug 22 09:11:34 EDT 2002 athena * [project @ 2002-08-22 13:11:34 by athena] Do not use inline. Minor changes. M ./dft/generic.c -1 +1 M ./kernel/alloc.c -2 +2 M ./kernel/planner-score.c -2 +2 M ./kernel/tensor.c -3 +3 Wed Aug 21 16:23:26 EDT 2002 stevenj * [project @ 2002-08-21 20:23:26 by stevenj] more commented flags M ./tests/bench.c +2 Tue Aug 20 19:44:43 EDT 2002 stevenj * [project @ 2002-08-20 23:44:43 by stevenj] added DCT-IV and DST-IV A ./reodft/reodft11e-r2hc.c M ./reodft/Makefile.am -1 +2 M ./reodft/conf.c -1 +2 M ./reodft/reodft11e-r2hc.c +297 M ./reodft/verify.c -1 +15 M ./tests/bench.c -1 +5 Tue Aug 20 16:01:36 EDT 2002 athena * [project @ 2002-08-20 20:01:36 by athena] Slight improvement in twiddle scheme M ./genfft/twiddle.ml -2 +9 Tue Aug 20 15:31:54 EDT 2002 stevenj * [project @ 2002-08-20 19:31:54 by stevenj] name fix M ./reodft/conf.c -2 +2 M ./reodft/reodft.h -6 +4 M ./reodft/reodft010e-r2hc.c -2 +2 Tue Aug 20 15:16:48 EDT 2002 stevenj * [project @ 2002-08-20 19:16:48 by stevenj] removed extraneous variable M ./reodft/reodft010e-r2hc.c -5 +1 Tue Aug 20 11:46:29 EDT 2002 athena * [project @ 2002-08-20 15:46:29 by athena] Oops M ./libbench/mp.c -33 +79 M ./libbench/verify.c -33 +25 Tue Aug 20 08:37:45 EDT 2002 athena * [project @ 2002-08-20 12:37:45 by athena] Still playing around M ./genfft/twiddle.ml -35 +62 M ./kernel/trig.c -1 +3 Mon Aug 19 19:56:29 EDT 2002 athena * [project @ 2002-08-19 23:56:29 by athena] Playing around with addition chain M ./TODO -6 +11 M ./genfft/algsimp.ml -3 +4 M ./genfft/expr.ml -1 +6 M ./genfft/expr.mli -1 +2 M ./genfft/twiddle.ml -46 +64 M ./support/addchain.c -8 +25 Mon Aug 19 19:48:56 EDT 2002 stevenj * [project @ 2002-08-19 23:48:56 by stevenj] comments M ./reodft/redft00e-r2hc.c -1 +4 M ./reodft/rodft00e-r2hc.c -1 +4 Mon Aug 19 19:45:35 EDT 2002 stevenj * [project @ 2002-08-19 23:45:35 by stevenj] comment fixes M ./reodft/reodft010e-r2hc.c -5 +2 Mon Aug 19 19:40:18 EDT 2002 stevenj * [project @ 2002-08-19 23:40:18 by stevenj] added reodft stuff A ./reodft/ A ./reodft/Makefile.am A ./reodft/conf.c A ./reodft/redft00e-r2hc.c A ./reodft/reodft.h A ./reodft/reodft010e-r2hc.c A ./reodft/rodft00e-r2hc.c A ./reodft/verify.c M ./Makefile.am -2 +3 M ./configure.ac +2 M ./dft/dft.h -1 +5 M ./rdft/rdft.h -1 +5 M ./reodft/Makefile.am +6 M ./reodft/conf.c +37 M ./reodft/redft00e-r2hc.c +204 M ./reodft/reodft.h +44 M ./reodft/reodft010e-r2hc.c +400 M ./reodft/rodft00e-r2hc.c +199 M ./reodft/verify.c +521 M ./tests/Makefile.am -1 +1 M ./tests/bench.c -2 +11 Sun Aug 18 19:44:14 EDT 2002 athena * [project @ 2002-08-18 23:44:14 by athena] Sync with nbenchfft M ./libbench/Makefile.am +4 M ./libbench/verify.c -4 +3 Sun Aug 18 16:02:37 EDT 2002 athena * [project @ 2002-08-18 20:02:37 by athena] Economy of thought M ./genfft/complex.ml -23 +12 M ./genfft/complex.mli -3 +3 M ./genfft/twiddle.ml -5 +5 Sat Aug 17 15:52:05 EDT 2002 stevenj * [project @ 2002-08-17 19:52:05 by stevenj] distribute addchain.c M ./support/Makefile.am -1 +1 Sat Aug 17 14:09:11 EDT 2002 athena * [project @ 2002-08-17 18:09:11 by athena] Nothing serious M ./support/addchain.c -4 +7 Sat Aug 17 10:47:59 EDT 2002 athena * [project @ 2002-08-17 14:47:59 by athena] New twiddle policy (disabled for now) A ./support/addchain.c M ./genfft/twiddle.ml -1 +98 M ./support/addchain.c +151 Fri Aug 16 23:44:28 EDT 2002 stevenj * [project @ 2002-08-17 03:44:28 by stevenj] bug fix for hc2r (must use inverse dft) M ./rdft/rank-geq2-rdft2.c -2 +2 Fri Aug 16 20:27:10 EDT 2002 athena * [project @ 2002-08-17 00:27:10 by athena] New log3 twiddle policy M ./dft/codelets/inplace/Makefile.am -2 +2 M ./dft/codelets/standard/Makefile.am -1 +1 M ./genfft/twiddle.ml -9 +88 M ./rdft/codelets/hc2r/Makefile.am +6 M ./rdft/codelets/r2hc/Makefile.am -1 +1 Fri Aug 16 18:10:33 EDT 2002 athena * [project @ 2002-08-16 22:10:33 by athena] More verify cleanup M ./dft/verify.c -20 +2 M ./kernel/verify-lib.c -1 +19 M ./kernel/verify.h +3 M ./rdft/verify.c -23 +5 Fri Aug 16 16:31:19 EDT 2002 athena * [project @ 2002-08-16 20:31:19 by athena] Oops M ./rdft/verify.c -3 +3 Fri Aug 16 15:22:36 EDT 2002 athena * [project @ 2002-08-16 19:22:36 by athena] Economy of thought (and code) A ./kernel/verify-lib.c A ./kernel/verify.h M ./dft/verify.c -305 +12 M ./kernel/Makefile.am -2 +2 M ./kernel/verify-lib.c +394 M ./kernel/verify.h +80 M ./rdft/verify.c -364 +23 Fri Aug 16 14:05:45 EDT 2002 athena * [project @ 2002-08-16 18:05:45 by athena] Added comment M ./TODO +2 Fri Aug 16 12:57:43 EDT 2002 athena * [project @ 2002-08-16 16:57:43 by athena] Cleaner rounding algorithm M ./libbench/mp.c -16 +27 Fri Aug 16 11:27:43 EDT 2002 athena * [project @ 2002-08-16 15:27:43 by athena] Can get away with shorter length in bluestein (I think). M ./libbench/mp.c -1 +1 Fri Aug 16 11:08:09 EDT 2002 athena * [project @ 2002-08-16 15:08:09 by athena] Portability improvements M ./libbench/mp.c -11 +4 Fri Aug 16 08:06:31 EDT 2002 athena * [project @ 2002-08-16 12:06:31 by athena] Optionally average accuracy test over many rounds M ./libbench/bench-main.c -2 +9 M ./libbench/bench.h -2 +2 M ./libbench/verify.c -29 +43 Fri Aug 16 07:50:24 EDT 2002 athena * [project @ 2002-08-16 11:50:24 by athena] More accurate formula for trig tables M ./dft/rader.c -3 +3 M ./rdft/rader-dht.c -3 +3 M ./rdft/rader-hc2hc.c -3 +3 Fri Aug 16 06:42:02 EDT 2002 athena * [project @ 2002-08-16 10:42:02 by athena] Implemented accuracy test for all integers M ./libbench/mp.c -6 +123 M ./libbench/verify.c -2 +1 Thu Aug 15 18:54:44 EDT 2002 athena * [project @ 2002-08-15 22:54:44 by athena] inv, neg: make static M ./libbench/mp.c -2 +2 Thu Aug 15 17:25:37 EDT 2002 athena * [project @ 2002-08-15 21:25:37 by athena] Verify was not complete for real transforms M ./libbench/verify.c -9 +34 Thu Aug 15 16:30:03 EDT 2002 athena * [project @ 2002-08-15 20:30:03 by athena] Oops M ./libbench/verify.c -3 +1 Thu Aug 15 16:29:16 EDT 2002 athena * [project @ 2002-08-15 20:29:16 by athena] Fixed hb codelets M ./genfft/gen_hc2hc.ml -3 +5 M ./libbench/verify.c -1 +4 Thu Aug 15 14:10:45 EDT 2002 athena * [project @ 2002-08-15 18:10:45 by athena] Changed twiddle policy M ./dft/codelets/inplace/Makefile.am -2 +2 M ./dft/codelets/standard/Makefile.am -1 +1 M ./rdft/codelets/r2hc/Makefile.am -1 +1 Thu Aug 15 13:32:24 EDT 2002 stevenj * [project @ 2002-08-15 17:32:24 by stevenj] whoops M ./rdft/direct2.c -3 +3 Thu Aug 15 11:01:04 EDT 2002 athena * [project @ 2002-08-15 15:01:04 by athena] No point in libbench being a shared library M ./libbench/Makefile.am -2 +2 M ./tests/Makefile.am -1 +1 Thu Aug 15 09:48:37 EDT 2002 athena * [project @ 2002-08-15 13:48:37 by athena] Moved accuracy test to libbench A ./libbench/mp.c M ./libbench/Makefile.am -1 +1 M ./libbench/bench-main.c -4 +13 M ./libbench/bench.h -1 +3 M ./libbench/mp.c +439 M ./libbench/util.c +8 M ./libbench/verify.c -3 +66 M ./tests/Makefile.am -5 +1 R ./tests/accuracy.c R ./tests/mp.c Wed Aug 14 19:48:23 EDT 2002 athena * [project @ 2002-08-14 23:48:23 by athena] Modified accuracy test M ./tests/accuracy.c -33 +20 Wed Aug 14 08:34:26 EDT 2002 athena * [project @ 2002-08-14 12:34:26 by athena] Fixes for long double M ./tests/accuracy.c -2 +3 M ./tests/mp.c -1 +1 Wed Aug 14 08:17:57 EDT 2002 athena * [project @ 2002-08-14 12:17:57 by athena] Normalize input M ./tests/accuracy.c -3 +17 Wed Aug 14 07:26:41 EDT 2002 athena * [project @ 2002-08-14 11:26:41 by athena] Oops M ./tests/accuracy.c +1 Wed Aug 14 07:25:34 EDT 2002 athena * [project @ 2002-08-14 11:25:34 by athena] Also compute relative error M ./tests/accuracy.c -2 +13 Wed Aug 14 07:08:20 EDT 2002 athena * [project @ 2002-08-14 11:08:20 by athena] Loop over N M ./tests/accuracy.c -24 +32 Wed Aug 14 06:54:50 EDT 2002 athena * [project @ 2002-08-14 10:54:50 by athena] simple-minded accuracy test A ./tests/accuracy.c A ./tests/mp.c M ./tests/Makefile.am -1 +6 M ./tests/accuracy.c +48 M ./tests/mp.c +434 Wed Aug 14 03:26:06 EDT 2002 stevenj * [project @ 2002-08-14 07:26:06 by stevenj] whoops A ./rdft/rank-geq2-rdft2.c Tue Aug 13 11:42:41 EDT 2002 athena * [project @ 2002-08-13 15:42:41 by athena] fma() stuff is too nonportable, removed M ./kernel/trig.c -45 +1 Mon Aug 12 14:07:44 EDT 2002 stevenj * [project @ 2002-08-12 18:07:44 by stevenj] slight fix M ./rdft/problem.c -2 +2 Mon Aug 12 14:07:18 EDT 2002 stevenj * [project @ 2002-08-12 18:07:18 by stevenj] use table for rdft_kind_str M ./rdft/problem.c -26 +12 Mon Aug 12 13:43:08 EDT 2002 stevenj * [project @ 2002-08-12 17:43:08 by stevenj] slight fixes M ./rdft/problem2.c -6 +11 Mon Aug 12 13:31:37 EDT 2002 stevenj * [project @ 2002-08-12 17:31:37 by stevenj] multidimensional rdft2 M ./kernel/ifftw.h -2 +3 M ./kernel/planner.c -1 +2 M ./kernel/tensor.c -5 +6 M ./rdft/Makefile.am -2 +2 M ./rdft/buffered2.c -13 +18 M ./rdft/conf.c -1 +2 M ./rdft/direct2.c -9 +14 M ./rdft/nop2.c -2 +12 M ./rdft/problem2.c -36 +41 M ./rdft/rdft.h -7 +9 M ./rdft/vrank-geq1-rdft2.c -6 +23 M ./tests/bench.c -11 +36 Sat Aug 10 19:33:23 EDT 2002 stevenj * [project @ 2002-08-10 23:33:23 by stevenj] use tensor_copy_inplace M ./rdft/indirect.c -13 +5 Sat Aug 10 19:32:03 EDT 2002 stevenj * [project @ 2002-08-10 23:32:03 by stevenj] bugfix, use tensor_copy_inplace M ./dft/rank-geq2.c -4 +10 Sat Aug 10 19:30:39 EDT 2002 stevenj * [project @ 2002-08-10 23:30:39 by stevenj] use tensor_copy_inplace M ./dft/indirect.c -13 +5 Sat Aug 10 19:28:07 EDT 2002 stevenj * [project @ 2002-08-10 23:28:07 by stevenj] added tensor_copy_inplace M ./kernel/ifftw.h -1 +3 M ./kernel/tensor.c -1 +19 Sat Aug 10 19:25:50 EDT 2002 stevenj * [project @ 2002-08-10 23:25:50 by stevenj] fixed trig-function table type M ./kernel/twiddle.c -2 +2 Sat Aug 10 14:41:04 EDT 2002 athena * [project @ 2002-08-10 18:41:04 by athena] Improved trig scheme M ./kernel/trig.c -8 +53 M ./tests/trigtest.c -8 +74 Fri Aug 9 21:05:01 EDT 2002 athena * [project @ 2002-08-10 01:05:01 by athena] Allow for testing using long double instead of pari M ./tests/trigtest.c -15 +27 Fri Aug 9 20:49:32 EDT 2002 athena * [project @ 2002-08-10 00:49:32 by athena] Yet another trig scheme. M ./kernel/trig.c -25 +24 M ./tests/trigtest.c -25 +22 Fri Aug 9 20:38:07 EDT 2002 athena * [project @ 2002-08-10 00:38:07 by athena] Yet another scheme M ./kernel/trig.c -4 +12 M ./tests/trigtest.c +7 Fri Aug 9 20:31:16 EDT 2002 athena * [project @ 2002-08-10 00:31:16 by athena] Careful with overflow M ./kernel/ifftw.h -4 +4 M ./kernel/trig.c -23 +23 M ./tests/trigtest.c -24 +17 Fri Aug 9 20:16:23 EDT 2002 athena * [project @ 2002-08-10 00:16:23 by athena] Avoid overflow M ./kernel/ifftw.h -4 +4 M ./kernel/trig.c -20 +24 M ./tests/trigtest.c -21 +43 Fri Aug 9 19:26:57 EDT 2002 athena * [project @ 2002-08-09 23:26:57 by athena] New(er) trig routines M ./dft/rader.c -8 +6 M ./dft/verify.c -4 +3 M ./kernel/ifftw.h -4 +4 M ./kernel/trig.c -22 +17 M ./kernel/twiddle.c -9 +7 M ./rdft/rader-dht.c -4 +3 M ./rdft/rader-hc2hc.c -8 +6 M ./rdft/verify.c -4 +3 M ./tests/trigtest.c -20 +19 Fri Aug 9 19:25:44 EDT 2002 athena * [project @ 2002-08-09 23:25:44 by athena] Oops M ./tests/bench.c -1 +1 Fri Aug 9 18:49:04 EDT 2002 athena * [project @ 2002-08-09 22:49:04 by athena] New file A ./tests/trigtest.c Fri Aug 9 13:04:00 EDT 2002 athena * [project @ 2002-08-09 17:04:00 by athena] Commented about likely gcc bug M ./TODO +4 Fri Aug 9 13:01:49 EDT 2002 athena * [project @ 2002-08-09 17:01:49 by athena] Improved accuracy of twiddle factors A ./kernel/trig.c M ./dft/rader.c -8 +8 M ./dft/verify.c -5 +4 M ./kernel/Makefile.am -1 +1 M ./kernel/ifftw.h -10 +7 M ./kernel/trig.c +79 M ./kernel/twiddle.c -8 +10 M ./rdft/rader-dht.c -4 +4 M ./rdft/rader-hc2hc.c -8 +8 M ./rdft/verify.c -5 +4 M ./tests/bench.c -1 +1 Thu Aug 8 06:36:23 EDT 2002 athena * [project @ 2002-08-08 10:36:23 by athena] Wrong comment M ./simd/simd-3dnow.h -1 Wed Aug 7 17:14:09 EDT 2002 athena * [project @ 2002-08-07 21:14:09 by athena] Experimental 3dnow port using gcc, to compare it with Stefan's stuff. A ./simd/3dnow.c A ./simd/simd-3dnow.h M ./configure.ac +6 M ./genfft/gen_notw_c.ml -3 +6 M ./genfft/gen_twiddle_c.ml -2 +4 M ./kernel/ifftw.h -2 +2 M ./simd/3dnow.c +66 M ./simd/Makefile.am -2 +2 M ./simd/simd-3dnow.h +164 M ./simd/simd-altivec.h +2 M ./simd/simd-sse.h +3 M ./simd/simd-sse2.h +3 M ./simd/simd.h +4 Wed Aug 7 12:58:10 EDT 2002 athena * [project @ 2002-08-07 16:58:10 by athena] End of AREF experiment M ./genfft/c.ml -2 +2 M ./kernel/ifftw.h -3 +1 Wed Aug 7 07:47:19 EDT 2002 athena * [project @ 2002-08-07 11:47:19 by athena] Oops M ./configure.ac -4 +1 Wed Aug 7 07:46:38 EDT 2002 athena * [project @ 2002-08-07 11:46:38 by athena] Pathetic attempt to reduce size of configure script M ./configure.ac -16 +8 Tue Aug 6 20:38:11 EDT 2002 athena * [project @ 2002-08-07 00:38:11 by athena] Changed array syntax for experiments. M ./genfft/c.ml -8 +10 M ./kernel/ifftw.h -1 +3 Tue Aug 6 19:58:20 EDT 2002 athena * [project @ 2002-08-06 23:58:20 by athena] Fix warning M ./simd/simd-sse2.h +2 Tue Aug 6 13:35:28 EDT 2002 athena * [project @ 2002-08-06 17:35:28 by athena] Move nonportable stuff in one place. M ./dft/problem.c -2 +2 M ./kernel/align.c -1 +8 M ./kernel/ifftw.h -1 +2 M ./rdft/problem.c -2 +2 M ./rdft/problem2.c -2 +2 Tue Aug 6 10:32:53 EDT 2002 athena * [project @ 2002-08-06 14:32:53 by athena] Economy of thought: I didn't like having two algorithms for removing solutions, both correct. At least now we have the same algorithm copied twice. M ./kernel/planner.c -12 +22 Tue Aug 6 09:12:21 EDT 2002 athena * [project @ 2002-08-06 13:12:21 by athena] Added things to do M ./TODO -1 +1 Mon Aug 5 19:54:31 EDT 2002 stevenj * [project @ 2002-08-05 23:54:31 by stevenj] improved interaction of planner with patience flags M ./kernel/ifftw.h -1 +2 M ./kernel/planner.c -12 +34 Mon Aug 5 14:17:58 EDT 2002 stevenj * [project @ 2002-08-05 18:17:58 by stevenj] set up for real-even/odd DFTs, where n is not the size of the data M ./rdft/buffered.c -3 +3 M ./rdft/codelet.h -7 +25 M ./rdft/indirect.c -4 +8 M ./rdft/problem.c -13 +58 M ./rdft/rader-hc2hc.c -1 +1 M ./rdft/rdft.h -1 +3 M ./tests/bench.c -2 +3 Sun Aug 4 23:57:51 EDT 2002 stevenj * [project @ 2002-08-05 03:57:51 by stevenj] DESTROY_INPUT flag M ./dft/ct-dif.c -4 +5 M ./dft/ct-dit.c -3 +5 M ./dft/ct-ditbuf.c -3 +5 M ./dft/ct-ditf.c -3 +5 M ./dft/ct.c -2 +2 M ./dft/ct.h -2 +3 M ./kernel/ifftw.h -2 +3 M ./rdft/hc2hc-buf.c -3 +6 M ./rdft/hc2hc-dif.c -3 +5 M ./rdft/hc2hc-dit.c -3 +5 M ./rdft/hc2hc.c -2 +2 M ./rdft/hc2hc.h -2 +3 M ./rdft/r2hc-hc2r.c -13 +20 M ./tests/bench.c +2 Sun Aug 4 22:50:19 EDT 2002 stevenj * [project @ 2002-08-05 02:50:19 by stevenj] CLASSIC -> IMPATIENT M ./dft/rank-geq2.c -2 +2 M ./dft/vrank-geq1.c -2 +2 M ./kernel/ifftw.h -4 +4 M ./kernel/planner.c -3 +3 M ./rdft/dft-r2hc.c -2 +2 M ./rdft/vrank-geq1-rdft2.c -2 +2 M ./rdft/vrank-geq1.c -2 +2 M ./tests/bench.c -1 +1 Sun Aug 4 19:05:43 EDT 2002 athena * [project @ 2002-08-04 23:05:43 by athena] Require make maintainer-clean to remove the generator, as opposed to make clean. In this way we can type make clean without regenerating all codelets. M ./genfft/Makefile.am -11 +11 M ./genfft-k7/Makefile.am -3 +3 Sun Aug 4 17:34:04 EDT 2002 stevenj * [project @ 2002-08-04 21:34:04 by stevenj] ESTIMATE plans are not blessed M ./kernel/planner.c -10 +13 Sun Aug 4 17:24:37 EDT 2002 stevenj * [project @ 2002-08-04 21:24:37 by stevenj] use flags in wisdom M ./kernel/ifftw.h -2 +4 M ./kernel/planner.c -9 +14 Sun Aug 4 17:03:45 EDT 2002 stevenj * [project @ 2002-08-04 21:03:45 by stevenj] score now takes plnr, not flags, as arg M ./dft/buffered.c -3 +3 M ./dft/ct-dif.c -4 +3 M ./dft/ct-dit.c -3 +3 M ./dft/ct-ditbuf.c -3 +3 M ./dft/ct-ditf.c -3 +3 M ./dft/direct.c -3 +3 M ./dft/generic.c -2 +2 M ./dft/indirect.c -3 +3 M ./dft/nop.c -3 +3 M ./dft/rader.c -4 +4 M ./dft/rank-geq2.c -3 +3 M ./dft/rank0.c -3 +3 M ./dft/vrank-geq1.c -4 +4 M ./dft/vrank2-transpose.c -3 +3 M ./dft/vrank3-transpose.c -3 +3 M ./kernel/ifftw.h -2 +2 M ./kernel/planner-score.c -4 +3 M ./kernel/tensor.c -3 +3 M ./rdft/buffered.c -3 +3 M ./rdft/buffered2.c -3 +3 M ./rdft/dft-r2hc.c -3 +3 M ./rdft/direct.c -3 +3 M ./rdft/direct2.c -3 +3 M ./rdft/generic.c -2 +2 M ./rdft/hc2hc-buf.c -3 +3 M ./rdft/hc2hc-dif.c -3 +3 M ./rdft/hc2hc-dit.c -3 +3 M ./rdft/indirect.c -3 +3 M ./rdft/nop.c -3 +3 M ./rdft/nop2.c -3 +3 M ./rdft/r2hc-hc2r.c -3 +3 M ./rdft/rader-dht.c -2 +2 M ./rdft/rader-hc2hc.c -2 +2 M ./rdft/rank0.c -3 +3 M ./rdft/vrank-geq1-rdft2.c -4 +4 M ./rdft/vrank-geq1.c -4 +4 M ./rdft/vrank2-transpose.c -3 +3 M ./rdft/vrank3-transpose.c -3 +3 M ./tests/bench.c -1 +1 Sun Aug 4 16:37:46 EDT 2002 stevenj * [project @ 2002-08-04 20:37:46 by stevenj] align initial stack in alignment check, which should now pass for gcc 3.1.1 M ./acinclude.m4 -2 +3 Sat Aug 3 20:04:57 EDT 2002 athena * [project @ 2002-08-04 00:04:57 by athena] Detect ultrasparc (sort of) M ./acinclude.m4 +3 Sat Aug 3 19:38:17 EDT 2002 stevenj * [project @ 2002-08-03 23:38:17 by stevenj] added solvtab_rdft_r2r placeholder M ./rdft/codelet.h -1 +3 Sat Aug 3 19:34:49 EDT 2002 athena * [project @ 2002-08-03 23:34:49 by athena] Damn solaris M ./support/Makefile.codelets -2 +2 Sat Aug 3 17:55:44 EDT 2002 stevenj * [project @ 2002-08-03 21:55:44 by stevenj] use E extended precision in solvers M ./rdft/problem.c -7 +9 Sat Aug 3 17:53:29 EDT 2002 stevenj * [project @ 2002-08-03 21:53:29 by stevenj] an alternative notation for D{C,S}T: DXTio, where i/o are {0,1} according to whether the input/output are shifted, respectively. Alternatively, io is the binary representation of the usual DXT-{I,II,III,IV} nomenclature, minus 1. M ./rdft/codelet.h -2 +2 Sat Aug 3 17:49:11 EDT 2002 stevenj * [project @ 2002-08-03 21:49:11 by stevenj] use E extended precision in solvers M ./dft/generic.c -5 +5 M ./dft/rader.c -2 +2 M ./rdft/generic.c -21 +21 M ./rdft/r2hc-hc2r.c -3 +3 M ./rdft/rader-dht.c -4 +4 Sat Aug 3 15:39:49 EDT 2002 athena * [project @ 2002-08-03 19:39:49 by athena] More portability fixes, compiler bugs workarounds, etc. M ./configure.ac -2 +4 M ./kernel/cycle.h -2 +2 M ./kernel/planner.c -2 +2 M ./rdft/problem2.c -2 +4 Sat Aug 3 15:09:56 EDT 2002 athena * [project @ 2002-08-03 19:09:56 by athena] More portability work M ./configure.ac -2 +7 M ./kernel/cycle.h -2 +2 M ./kernel/ifftw.h -9 +2 Sat Aug 3 14:33:40 EDT 2002 athena * [project @ 2002-08-03 18:33:40 by athena] Improved portability, removed gnu make dependencies M ./acinclude.m4 -1 +1 M ./configure.ac -4 +3 M ./kernel/cycle.h -2 +2 M ./kernel/ifftw.h -2 +2 M ./support/Makefile.codelets -14 +6 Sat Aug 3 13:48:53 EDT 2002 athena * [project @ 2002-08-03 17:48:53 by athena] Remember to thank XXX M ./TODO +8 Fri Aug 2 17:38:18 EDT 2002 athena * [project @ 2002-08-02 21:38:18 by athena] Multiplication on altivec requires FMA with -0.0 to be IEEE754 compliant. M ./simd/simd-altivec.h -1 +1 Fri Aug 2 15:26:37 EDT 2002 athena * [project @ 2002-08-02 19:26:37 by athena] Allow for extended precision in codelets M ./genfft/c.ml -3 +4 M ./kernel/ifftw.h -7 +9 Fri Aug 2 08:52:04 EDT 2002 athena * [project @ 2002-08-02 12:52:04 by athena] Shortened names M ./dft/codelets/inplace/Makefile.am -5 +5 Fri Aug 2 03:49:09 EDT 2002 stevenj * [project @ 2002-08-02 07:49:09 by stevenj] added infrastructure for future r2r transforms M ./TODO -3 +1 M ./rdft/codelet.h -3 +33 M ./rdft/problem.c -1 +8 Thu Aug 1 21:29:14 EDT 2002 athena * [project @ 2002-08-02 01:29:05 by athena] Version info M ./Makefile.am +1 M ./configure.ac +2 Thu Aug 1 21:06:22 EDT 2002 athena * [project @ 2002-08-02 01:06:22 by athena] Listened to one customer and added radix-12. Added radix-15 for consistency (whatever that is) M ./dft/codelets/inplace/Makefile.am -1 +1 M ./dft/codelets/standard/Makefile.am -1 +1 M ./dft/k7/codelets/Makefile.am -3 +4 M ./dft/simd/codelets/Makefile.am -2 +2 M ./kernel/align.c -2 +8 M ./rdft/codelets/hc2r/Makefile.am -3 +3 M ./rdft/codelets/r2hc/Makefile.am -3 +3 Thu Aug 1 19:50:53 EDT 2002 stevenj * [project @ 2002-08-01 23:50:53 by stevenj] whoops again, fixed the wrong line M ./kernel/cycle.h -3 +3 Thu Aug 1 19:50:16 EDT 2002 stevenj * [project @ 2002-08-01 23:50:16 by stevenj] whoops M ./kernel/cycle.h -2 +2 Thu Aug 1 16:01:15 EDT 2002 stevenj * [project @ 2002-08-01 20:01:15 by stevenj] use new AC_INIT and add VERSION to wisdom M ./configure.ac -3 +4 M ./kernel/planner.c -4 +6 Thu Aug 1 14:56:45 EDT 2002 stevenj * [project @ 2002-08-01 18:56:45 by stevenj] mygetR -> getR M ./kernel/scan.c -3 +3 Thu Aug 1 14:56:02 EDT 2002 stevenj * [project @ 2002-08-01 18:56:02 by stevenj] scanner cleanups: just return 0/1, simplify integer reads M ./dft/problem.c -6 +4 M ./kernel/planner.c -14 +12 M ./kernel/scan.c -78 +42 M ./kernel/tensor.c -9 +8 M ./rdft/problem.c -7 +5 M ./rdft/problem2.c -7 +5 M ./tests/bench.c +7 Thu Aug 1 08:04:01 EDT 2002 athena * [project @ 2002-08-01 12:04:01 by athena] Reverted back to casting pointer to ulong M ./kernel/align.c -2 +2 Thu Aug 1 08:03:46 EDT 2002 athena * [project @ 2002-08-01 12:03:46 by athena] Cast to unsigned long, not long M ./kernel/ifftw.h -6 +6 Thu Aug 1 03:14:50 EDT 2002 stevenj * [project @ 2002-08-01 07:14:50 by stevenj] additional comment M ./kernel/scan.c -2 +4 Thu Aug 1 03:12:37 EDT 2002 stevenj * [project @ 2002-08-01 07:12:37 by stevenj] added comment M ./kernel/scan.c -1 +6 Thu Aug 1 03:03:18 EDT 2002 stevenj * [project @ 2002-08-01 07:03:18 by stevenj] added wisdom import A ./kernel/printers.c A ./kernel/scan.c A ./kernel/scanners.c M ./dft/conf.c -1 +3 M ./dft/dft.h -1 +3 M ./dft/problem.c -2 +31 M ./dft/verify.c -2 +1 M ./kernel/Makefile.am -3 +4 M ./kernel/alloc.c -2 +1 M ./kernel/assert.c -2 +1 M ./kernel/debug.c -9 +2 M ./kernel/ifftw.h -6 +53 M ./kernel/planner.c -22 +106 M ./kernel/print.c -2 +1 M ./kernel/printers.c +80 M ./kernel/problem.c -1 +2 M ./kernel/scan.c +322 M ./kernel/scanners.c +61 M ./kernel/tensor.c -4 +30 M ./kernel/timer.c -2 +1 M ./rdft/conf.c -1 +4 M ./rdft/problem.c -2 +32 M ./rdft/problem2.c -2 +32 M ./rdft/rdft.h -1 +4 M ./rdft/verify.c -2 +1 M ./tests/bench.c -9 +31 Wed Jul 31 23:12:05 EDT 2002 stevenj * [project @ 2002-08-01 03:12:05 by stevenj] whoops M ./kernel/align.c -1 +2 Wed Jul 31 22:06:46 EDT 2002 stevenj * [project @ 2002-08-01 02:06:46 by stevenj] use %u for alignment_of M ./dft/problem.c -3 +3 M ./rdft/problem.c -2 +2 M ./rdft/problem2.c -2 +2 Wed Jul 31 21:47:15 EDT 2002 stevenj * [project @ 2002-08-01 01:47:15 by stevenj] ptrdiff_t form M ./kernel/align.c -2 +2 Wed Jul 31 21:33:35 EDT 2002 athena * [project @ 2002-08-01 01:33:35 by athena] Cast to avoid warning from C++ compiler M ./kernel/ifftw.h -2 +2 Wed Jul 31 18:57:04 EDT 2002 athena * [project @ 2002-07-31 22:57:04 by athena] Make problem equality depend on alignments. A ./kernel/align.c M ./dft/problem.c -4 +10 M ./kernel/Makefile.am -5 +6 M ./kernel/align.c +35 M ./kernel/ifftw.h -1 +2 M ./rdft/problem.c -4 +9 M ./rdft/problem2.c -8 +13 M ./simd/simd.h +2 Wed Jul 31 15:45:31 EDT 2002 athena * [project @ 2002-07-31 19:45:31 by athena] Shorter names M ./dft/simd/codelets/Makefile.am -22 +18 Wed Jul 31 14:38:00 EDT 2002 athena * [project @ 2002-07-31 18:38:00 by athena] Oops M ./simd/simd-sse.h -1 +1 Wed Jul 31 14:37:19 EDT 2002 athena * [project @ 2002-07-31 18:37:19 by athena] Fix warning M ./simd/simd-sse.h +2 Wed Jul 31 07:52:53 EDT 2002 athena * [project @ 2002-07-31 11:52:53 by athena] Removed silly abstraction barrier. Also, cons() terminology was no longer appropriate. M ./kernel/alloc.c -2 +2 M ./kernel/ifftw.h -16 +19 M ./kernel/planner-naive.c -2 +2 M ./kernel/planner-score.c -2 +2 M ./kernel/planner.c -26 +13 Tue Jul 30 22:35:24 EDT 2002 stevenj * [project @ 2002-07-31 02:35:24 by stevenj] removed register_registrar and solvtab_exec_reverse hacks M ./kernel/ifftw.h -4 +3 M ./kernel/planner.c -24 +15 M ./kernel/solvtab.c -19 +6 Tue Jul 30 19:54:41 EDT 2002 stevenj * [project @ 2002-07-30 23:54:41 by stevenj] register_registrar doesn't search whole solver list (maybe we should change register_solver instead) M ./kernel/planner.c -1 +3 Tue Jul 30 19:36:37 EDT 2002 stevenj * [project @ 2002-07-30 23:36:37 by stevenj] credit M ./kernel/cycle.h -2 +2 Tue Jul 30 19:34:16 EDT 2002 stevenj * [project @ 2002-07-30 23:34:16 by stevenj] added HP/UX ia64 support, courtesy of Teresa L. Johnson M ./kernel/cycle.h -1 +23 Tue Jul 30 13:28:33 EDT 2002 athena * [project @ 2002-07-30 17:28:33 by athena] Fixed alignment checks M ./dft/simd/n1b.c -1 +1 M ./dft/simd/n1f.c -1 +1 M ./dft/simd/t1b.c -1 +1 M ./dft/simd/t1f.c -1 +1 M ./kernel/alloc.c -1 +2 Tue Jul 30 01:20:11 EDT 2002 stevenj * [project @ 2002-07-30 05:20:11 by stevenj] ugh, wisdom id fixes in exprt_conf M ./kernel/ifftw.h -2 +4 M ./kernel/planner.c -2 +4 M ./kernel/solvtab.c -5 +18 Tue Jul 30 00:41:15 EDT 2002 stevenj * [project @ 2002-07-30 04:41:15 by stevenj] exprt_registrars -> exprt_conf, added missing SOLVTAB_END M ./kernel/ifftw.h -2 +2 M ./kernel/planner.c -5 +6 M ./tests/bench.c -1 +1 Tue Jul 30 00:36:26 EDT 2002 stevenj * [project @ 2002-07-30 04:36:26 by stevenj] exprt_registrars should output self-contained configuration M ./kernel/planner.c -9 +20 Mon Jul 29 23:52:07 EDT 2002 stevenj * [project @ 2002-07-30 03:52:07 by stevenj] added exprt_registrars M ./dft/conf.c -12 +12 M ./kernel/ifftw.h -2 +7 M ./kernel/planner.c -4 +91 M ./kernel/solvtab.c -3 +5 M ./rdft/conf.c -18 +18 M ./support/Makefile.codelets -3 +3 M ./tests/bench.c +2 Mon Jul 29 23:42:27 EDT 2002 stevenj * [project @ 2002-07-30 03:42:27 by stevenj] whoops M ./kernel/print.c -2 +3 Mon Jul 29 21:24:51 EDT 2002 athena * [project @ 2002-07-30 01:24:51 by athena] More stringent requirements on strides for SIMD codelets M ./dft/simd/n1b.c -4 +4 M ./dft/simd/n1b.h +3 M ./dft/simd/n1f.c -4 +4 M ./dft/simd/n1f.h +3 M ./dft/simd/t1b.c -2 +2 M ./dft/simd/t1b.h -1 +6 M ./dft/simd/t1f.c -2 +2 M ./dft/simd/t1f.h +5 M ./simd/simd-altivec.h +2 M ./simd/simd-sse.h +2 M ./simd/simd-sse2.h +2 M ./simd/simd.h -2 +3 Mon Jul 29 21:05:49 EDT 2002 stevenj * [project @ 2002-07-30 01:05:49 by stevenj] remove warning M ./rdft/buffered2.c -2 +3 Mon Jul 29 20:51:19 EDT 2002 stevenj * [project @ 2002-07-30 00:51:01 by stevenj] use %td for ptrdiff_t and %T for tensors M ./dft/problem.c -2 +2 M ./kernel/print.c -3 +13 M ./kernel/traverse.c -2 +7 M ./rdft/problem.c -2 +2 M ./rdft/problem2.c -2 +2 Mon Jul 29 16:17:11 EDT 2002 athena * [project @ 2002-07-29 20:17:11 by athena] Fix for SIMD M ./dft/buffered.c -1 +5 Mon Jul 29 15:40:53 EDT 2002 athena * [project @ 2002-07-29 19:40:53 by athena] Missing lfftw_mkstride and lfftw_stride_destroy M ./kernel/ifftw.h -1 +3 Mon Jul 29 14:34:46 EDT 2002 athena * [project @ 2002-07-29 18:34:46 by athena] Implement LDA/STA M ./simd/simd-altivec.h +10 Mon Jul 29 14:19:21 EDT 2002 athena * [project @ 2002-07-29 18:19:21 by athena] More SIMD work M ./dft/simd/n1b.c -4 +4 M ./dft/simd/n1f.c -4 +4 M ./dft/simd/t1b.c -2 +2 M ./dft/simd/t1f.c -2 +2 M ./simd/simd-altivec.h -3 +2 M ./simd/simd-sse.h -3 +2 M ./simd/simd-sse2.h -2 +1 M ./simd/simd.h +3 Mon Jul 29 13:16:12 EDT 2002 athena * [project @ 2002-07-29 17:16:12 by athena] Cleanup M ./simd/simd-altivec.h -9 +9 M ./simd/simd-sse.h -16 +28 M ./simd/simd-sse2.h -13 +25 Mon Jul 29 13:02:38 EDT 2002 stevenj * [project @ 2002-07-29 17:02:38 by stevenj] update M ./ChangeLog +461 Mon Jul 29 12:45:33 EDT 2002 athena * [project @ 2002-07-29 16:45:33 by athena] Also check strides in SIMD codelets M ./dft/simd/n1b.c +4 M ./dft/simd/n1f.c +4 M ./dft/simd/t1b.c +2 M ./dft/simd/t1f.c +2 M ./simd/simd-altivec.h -1 +2 M ./simd/simd-sse.h +2 M ./simd/simd-sse2.h +2 Mon Jul 29 11:26:08 EDT 2002 athena * [project @ 2002-07-29 15:26:08 by athena] Minor changes, mostly for consistency with the big-endian processor M ./simd/simd-altivec.h -18 +18 Mon Jul 29 00:50:06 EDT 2002 stevenj * [project @ 2002-07-29 04:50:06 by stevenj] added comment M ./rdft/rader-dht.c +1 Sun Jul 28 21:19:35 EDT 2002 stevenj * [project @ 2002-07-29 01:19:35 by stevenj] added code for icc's _mm_malloc (memalign replacement) M ./configure.ac -1 +1 M ./kernel/alloc.c -4 +11 Sun Jul 28 17:33:07 EDT 2002 stevenj * [project @ 2002-07-28 21:33:07 by stevenj] slight fixes M ./rdft/problem2.c -6 +11 M ./rdft/verify.c -4 +7 M ./tests/bench.c -1 +2 Sun Jul 28 16:28:43 EDT 2002 stevenj * [project @ 2002-07-28 20:28:43 by stevenj] whoops M ./rdft/problem2.c -3 +3 Sun Jul 28 16:13:19 EDT 2002 athena * [project @ 2002-07-28 20:13:19 by athena] Use vec_xor to change sign M ./simd/altivec.c -2 +4 M ./simd/simd-altivec.h -3 +10 Sun Jul 28 16:10:59 EDT 2002 stevenj * [project @ 2002-07-28 20:10:59 by stevenj] added rdft2 A ./rdft/buffered2.c A ./rdft/direct2.c A ./rdft/nop2.c A ./rdft/plan2.c A ./rdft/problem2.c A ./rdft/solve2.c A ./rdft/vrank-geq1-rdft2.c M ./rdft/Makefile.am -2 +4 M ./rdft/buffered2.c +451 M ./rdft/conf.c -1 +7 M ./rdft/direct2.c +221 M ./rdft/khc2r.c -3 +3 M ./rdft/kr2hc.c -3 +3 M ./rdft/nop2.c +88 M ./rdft/plan2.c +33 M ./rdft/problem2.c +175 M ./rdft/rdft.h -1 +53 M ./rdft/solve2.c +31 M ./rdft/verify.c -17 +195 M ./rdft/vrank-geq1-rdft2.c +259 M ./tests/bench.c -6 +35 Sun Jul 28 15:45:54 EDT 2002 athena * [project @ 2002-07-28 19:45:54 by athena] Optimized M ./simd/simd-altivec.h -17 +10 Sun Jul 28 15:11:14 EDT 2002 athena * [project @ 2002-07-28 19:11:14 by athena] Changed ALIGNMENT M ./simd/simd-altivec.h -1 +1 Sun Jul 28 15:09:40 EDT 2002 athena * [project @ 2002-07-28 19:09:40 by athena] alignment := 8 M ./simd/simd-sse.h -1 +1 Sun Jul 28 14:57:22 EDT 2002 athena * [project @ 2002-07-28 18:57:22 by athena] Avoid warning M ./simd/simd-altivec.h -1 +1 Sun Jul 28 14:53:03 EDT 2002 athena * [project @ 2002-07-28 18:53:03 by athena] Oops M ./simd/simd-sse2.h -2 +2 Sun Jul 28 14:50:09 EDT 2002 athena * [project @ 2002-07-28 18:50:09 by athena] New altivec experiment M ./genfft/annotate.ml -5 +5 M ./genfft/genutil.ml -7 +7 M ./genfft/simd.ml -3 +6 M ./genfft/variable.ml -38 +16 M ./genfft/variable.mli -8 +4 M ./simd/simd-altivec.h -15 +15 M ./simd/simd-sse.h -3 +5 M ./simd/simd-sse2.h -2 +2 Sun Jul 28 13:48:20 EDT 2002 athena * [project @ 2002-07-28 17:48:20 by athena] Nothing M ./simd/simd-altivec.h -6 +6 Sun Jul 28 13:47:50 EDT 2002 athena * [project @ 2002-07-28 17:47:50 by athena] Oops M ./simd/simd-altivec.h -26 +19 Sun Jul 28 13:44:28 EDT 2002 athena * [project @ 2002-07-28 17:44:28 by athena] Nothing M ./simd/simd-altivec.h -19 +26 Sun Jul 28 10:38:10 EDT 2002 athena * [project @ 2002-07-28 14:38:10 by athena] Constants are now in separate file. A ./simd/altivec.c M ./simd/Makefile.am -1 +1 M ./simd/altivec.c +38 M ./simd/simd-altivec.h -20 +13 Sun Jul 28 07:58:37 EDT 2002 athena * [project @ 2002-07-28 11:58:37 by athena] More precise comment M ./simd/simd-altivec.h -1 +1 Sun Jul 28 07:56:40 EDT 2002 athena * [project @ 2002-07-28 11:56:40 by athena] gcc-3.1 bug workaround M ./simd/simd-altivec.h -42 +24 Sun Jul 28 01:39:54 EDT 2002 stevenj * [project @ 2002-07-28 05:39:54 by stevenj] slight optimization, and exported zerotens functions M ./dft/buffered.c -4 +4 M ./dft/dft.h -1 +2 M ./dft/problem.c -4 +4 M ./rdft/buffered.c -4 +4 M ./rdft/problem.c -4 +4 M ./rdft/rdft.h -1 +2 Sun Jul 28 00:54:59 EDT 2002 stevenj * [project @ 2002-07-28 04:54:59 by stevenj] should be a plan_dft, not a plan_rdft M ./rdft/dft-r2hc.c -2 +2 Sat Jul 27 21:36:46 EDT 2002 athena * [project @ 2002-07-28 01:36:46 by athena] Optimizations. Make it work with vanilla non-Apple gcc. M ./simd/simd-altivec.h -21 +82 Sat Jul 27 19:20:09 EDT 2002 stevenj * [project @ 2002-07-27 23:20:09 by stevenj] whoops M ./rdft/generic.c -5 Sat Jul 27 18:54:01 EDT 2002 stevenj * [project @ 2002-07-27 22:54:01 by stevenj] added hc2r (dif) M ./rdft/generic.c -13 +122 Sat Jul 27 18:31:43 EDT 2002 stevenj * [project @ 2002-07-27 22:31:43 by stevenj] add hc2r (dif) case M ./rdft/rader-hc2hc.c -2 +158 Sat Jul 27 15:09:40 EDT 2002 athena * [project @ 2002-07-27 19:09:40 by athena] Altivec port M ./simd/simd-altivec.h -38 +131 M ./support/Makefile.codelets -3 +3 Sat Jul 27 15:06:21 EDT 2002 athena * [project @ 2002-07-27 19:06:21 by athena] Fixed signed/unsigned bug. M ./kernel/twiddle.c -2 +3 Thu Jul 25 20:11:26 EDT 2002 athena * [project @ 2002-07-26 00:11:26 by athena] Make rank0 unapplicable to in-place problems. M ./dft/rank0.c -1 +2 M ./rdft/rank0.c -1 +2 Thu Jul 25 17:10:52 EDT 2002 stevenj * [project @ 2002-07-25 21:10:52 by stevenj] only works for r odd M ./rdft/generic.c -1 +1 Thu Jul 25 15:30:06 EDT 2002 athena * [project @ 2002-07-25 19:30:06 by athena] Reinserted much better timing-avoidance heuristic M ./kernel/planner-score.c -2 +7 Thu Jul 25 15:21:13 EDT 2002 athena * [project @ 2002-07-25 19:21:13 by athena] Score is now a property of the plan, not of the solver. Revised representation of closures. M ./dft/buffered.c -5 +1 M ./kernel/ifftw.h -2 +7 M ./kernel/plan.c -4 +6 M ./kernel/planner-score.c -17 +30 M ./kernel/traverse.c -17 +16 M ./rdft/buffered.c -5 +1 M ./tests/bench.c -7 +14 Thu Jul 25 06:36:51 EDT 2002 athena * [project @ 2002-07-25 10:36:51 by athena] Cosmetic changes. Added hc2r_128.c M ./genfft/gen_hc2r.ml -3 +3 M ./genfft/gen_r2hc.ml -3 +3 M ./rdft/codelets/hc2r/Makefile.am -1 +1 Thu Jul 25 01:37:53 EDT 2002 stevenj * [project @ 2002-07-25 05:37:53 by stevenj] added hc2r M ./rdft/rader-dht.c -16 +55 Thu Jul 25 00:51:45 EDT 2002 stevenj * [project @ 2002-07-25 04:51:45 by stevenj] added hc2hc-difbuf A ./rdft/hc2hc-buf.c M ./rdft/Makefile.am -3 +3 M ./rdft/hc2hc-buf.c +248 R ./rdft/hc2hc-ditbuf.c M ./rdft/khc2hc-dif.c -3 +1 Thu Jul 25 00:25:06 EDT 2002 stevenj * [project @ 2002-07-25 04:25:06 by stevenj] added rdft-dif A ./rdft/hc2hc-dif.c M ./rdft/Makefile.am -3 +3 M ./rdft/hc2hc-dif.c +128 M ./rdft/hc2hc.c -3 +5 M ./rdft/khc2hc-dif.c -4 +3 M ./rdft/rdft.h -1 +3 Thu Jul 25 00:22:36 EDT 2002 stevenj * [project @ 2002-07-25 04:22:36 by stevenj] whoops, hc2r must be conjugated to have right sign M ./rdft/verify.c -3 +3 Wed Jul 24 23:27:45 EDT 2002 stevenj * [project @ 2002-07-25 03:27:45 by stevenj] slight change M ./dft/ct-dif.c -4 +6 Wed Jul 24 23:24:24 EDT 2002 stevenj * [project @ 2002-07-25 03:24:24 by stevenj] whoops M ./rdft/verify.c -3 +2 Wed Jul 24 22:46:39 EDT 2002 stevenj * [project @ 2002-07-25 02:46:39 by stevenj] support hc2r codelets A ./rdft/direct.c M ./rdft/Makefile.am -5 +5 M ./rdft/codelet.h -1 +3 R ./rdft/direct-r2hc.c M ./rdft/direct.c +227 M ./rdft/khc2r.c -5 +2 M ./rdft/rdft.h -2 +2 Wed Jul 24 22:01:53 EDT 2002 stevenj * [project @ 2002-07-25 02:01:53 by stevenj] use vector plan for r/i instead of two separate plans M ./rdft/dft-r2hc.c -36 +20 Wed Jul 24 20:36:34 EDT 2002 stevenj * [project @ 2002-07-25 00:36:34 by stevenj] hack to allow rader/generic to work in-place for small prime sizes, instead of always using buffered M ./dft/buffered.c -1 +5 M ./dft/rader.c -2 +2 M ./kernel/ifftw.h -1 +3 M ./rdft/buffered.c -1 +5 M ./rdft/rader-dht.c -1 +1 M ./rdft/rader-hc2hc.c -1 +1 Wed Jul 24 18:04:41 EDT 2002 stevenj * [project @ 2002-07-24 22:04:41 by stevenj] added rdft-generic A ./rdft/generic.c M ./rdft/Makefile.am -4 +4 M ./rdft/conf.c -2 +2 M ./rdft/generic.c +281 Wed Jul 24 17:27:34 EDT 2002 stevenj * [project @ 2002-07-24 21:27:34 by stevenj] fixed add count M ./dft/generic.c -1 +1 Wed Jul 24 14:52:26 EDT 2002 stevenj * [project @ 2002-07-24 18:52:26 by stevenj] again M ./rdft/rader-hc2hc.c -2 +2 Wed Jul 24 14:51:58 EDT 2002 stevenj * [project @ 2002-07-24 18:51:58 by stevenj] slight fix M ./rdft/rader-hc2hc.c -4 +5 Wed Jul 24 14:51:07 EDT 2002 stevenj * [project @ 2002-07-24 18:51:07 by stevenj] fixed comment M ./rdft/rader-hc2hc.c -3 +3 Wed Jul 24 14:41:24 EDT 2002 stevenj * [project @ 2002-07-24 18:41:24 by stevenj] whoops M ./tests/bench.c -2 Wed Jul 24 14:38:15 EDT 2002 stevenj * [project @ 2002-07-24 18:38:15 by stevenj] added rader-hc2hc A ./rdft/rader-hc2hc.c M ./rdft/Makefile.am -2 +2 M ./rdft/conf.c -2 +2 M ./rdft/rader-hc2hc.c +482 M ./rdft/rdft.h -2 +2 M ./tests/bench.c +2 Wed Jul 24 00:07:59 EDT 2002 stevenj * [project @ 2002-07-24 04:07:59 by stevenj] whoops, initialize W M ./dft/rader.c +1 Tue Jul 23 23:03:09 EDT 2002 stevenj * [project @ 2002-07-24 03:03:09 by stevenj] strides should not be unsigned M ./rdft/rader-dht.c -2 +4 Tue Jul 23 23:02:08 EDT 2002 stevenj * [project @ 2002-07-24 03:02:08 by stevenj] more stride sign fixes M ./dft/rader.c -3 +5 Tue Jul 23 23:01:04 EDT 2002 stevenj * [project @ 2002-07-24 03:01:04 by stevenj] strides should not be unsigned! M ./dft/rader.c -1 +2 Tue Jul 23 14:55:25 EDT 2002 stevenj * [project @ 2002-07-23 18:55:25 by stevenj] added comment M ./rdft/dft-r2hc.c -1 +4 Tue Jul 23 14:52:04 EDT 2002 stevenj * [project @ 2002-07-23 18:52:04 by stevenj] another fix to op count M ./rdft/r2hc-hc2r.c -1 +4 Tue Jul 23 14:51:01 EDT 2002 stevenj * [project @ 2002-07-23 18:51:01 by stevenj] whoops M ./rdft/r2hc-hc2r.c -2 +2 Tue Jul 23 14:49:43 EDT 2002 stevenj * [project @ 2002-07-23 18:49:43 by stevenj] slight fix to op counts M ./rdft/dft-r2hc.c -3 +3 M ./rdft/r2hc-hc2r.c -3 +3 Tue Jul 23 14:09:19 EDT 2002 stevenj * [project @ 2002-07-23 18:09:18 by stevenj] added dft-r2hc A ./rdft/dft-r2hc.c M ./rdft/Makefile.am -5 +6 M ./rdft/conf.c -2 +3 M ./rdft/dft-r2hc.c +187 M ./rdft/rdft.h -1 +2 Tue Jul 23 02:50:12 EDT 2002 stevenj * [project @ 2002-07-23 06:50:12 by stevenj] better comment and var. name M ./rdft/rader-dht.c -7 +7 Tue Jul 23 02:39:11 EDT 2002 stevenj * [project @ 2002-07-23 06:39:11 by stevenj] fixed tests for hc2r, and added r2hc-hc2r A ./rdft/r2hc-hc2r.c M ./rdft/Makefile.am -3 +3 M ./rdft/conf.c -1 +2 M ./rdft/r2hc-hc2r.c +182 M ./rdft/rdft.h -1 +2 M ./rdft/verify.c -1 +46 M ./tests/bench.c -1 +6 Tue Jul 23 00:45:23 EDT 2002 stevenj * [project @ 2002-07-23 04:45:23 by stevenj] added rader-dht A ./rdft/rader-dht.c M ./rdft/Makefile.am -2 +3 M ./rdft/conf.c -1 +2 M ./rdft/rader-dht.c +433 M ./rdft/rdft.h -1 +2 Mon Jul 22 21:05:12 EDT 2002 athena * [project @ 2002-07-23 01:05:12 by athena] Added r2hc_128, what the hell. M ./rdft/codelets/r2hc/Makefile.am -1 +1 Mon Jul 22 20:48:59 EDT 2002 athena * [project @ 2002-07-23 00:48:59 by athena] Added codelets that compute twiddle factors M ./rdft/codelets/r2hc/Makefile.am -1 +7 Mon Jul 22 19:57:16 EDT 2002 stevenj * [project @ 2002-07-22 23:57:16 by stevenj] added rdft-buffered A ./rdft/buffered.c M ./rdft/Makefile.am -2 +2 M ./rdft/buffered.c +330 M ./rdft/conf.c -2 +2 Mon Jul 22 19:43:39 EDT 2002 stevenj * [project @ 2002-07-22 23:43:39 by stevenj] added hc2hc-ditbuf A ./rdft/hc2hc-ditbuf.c M ./rdft/Makefile.am -3 +3 M ./rdft/hc2hc-ditbuf.c +189 M ./rdft/khc2hc-dit.c -3 +1 Mon Jul 22 14:29:04 EDT 2002 stevenj * [project @ 2002-07-22 18:29:04 by stevenj] use STACK_MALLOC (alloca), since generic radix is always small M ./dft/generic.c -3 +3 Mon Jul 22 14:22:43 EDT 2002 stevenj * [project @ 2002-07-22 18:22:43 by stevenj] small cleanup M ./rdft/hc2hc-dit.c -2 +2 Mon Jul 22 07:42:13 EDT 2002 athena * [project @ 2002-07-22 11:42:13 by athena] What the hell was I thinking? M ./rdft/problem.c -18 +8 Mon Jul 22 07:37:12 EDT 2002 athena * [project @ 2002-07-22 11:37:12 by athena] Reduced code size by using table instead of switch statement. M ./rdft/problem.c -12 +16 Mon Jul 22 07:27:06 EDT 2002 athena * [project @ 2002-07-22 11:27:06 by athena] Changed hash function to avoid collisions with DFT. M ./rdft/problem.c -4 +4 Mon Jul 22 01:37:06 EDT 2002 stevenj * [project @ 2002-07-22 05:37:06 by stevenj] added missing file, whoops A ./rdft/hc2hc-dit.c Mon Jul 22 01:24:17 EDT 2002 stevenj * [project @ 2002-07-22 05:24:17 by stevenj] whoops, generate enough twiddles for odd m M ./rdft/hc2hc.c -2 +2 Mon Jul 22 01:10:21 EDT 2002 stevenj * [project @ 2002-07-22 05:10:21 by stevenj] don't try to verify R2HCII or HC2RIII plans M ./rdft/verify.c -5 +7 Mon Jul 22 01:05:00 EDT 2002 stevenj * [project @ 2002-07-22 05:05:00 by stevenj] recursive case now works, I think M ./rdft/hc2hc.c -9 +4 Mon Jul 22 01:04:40 EDT 2002 stevenj * [project @ 2002-07-22 05:04:40 by stevenj] add extra impulse test for debugging M ./rdft/verify.c -1 +4 Mon Jul 22 01:02:38 EDT 2002 stevenj * [project @ 2002-07-22 05:02:38 by stevenj] whoops, multiply ios offset by stride (and rename to ioffset) M ./rdft/direct-r2hc.c -4 +5 Mon Jul 22 00:22:02 EDT 2002 stevenj * [project @ 2002-07-22 04:22:02 by stevenj] whoops M ./rdft/verify.c -2 +2 Sun Jul 21 23:58:14 EDT 2002 stevenj * [project @ 2002-07-22 03:58:14 by stevenj] whoops M ./rdft/verify.c -2 +2 Sun Jul 21 23:43:03 EDT 2002 stevenj * [project @ 2002-07-22 03:43:03 by stevenj] added hc2hc-dit M ./genfft/gen_hc2hc.ml -3 +3 M ./rdft/Makefile.am -2 +2 M ./rdft/hc2hc.c -10 +21 M ./rdft/khc2hc-dit.c -4 +3 Sun Jul 21 23:15:12 EDT 2002 stevenj * [project @ 2002-07-22 03:15:12 by stevenj] twiddles can be shared with smaller m's M ./kernel/twiddle.c -4 +4 Sun Jul 21 22:34:28 EDT 2002 stevenj * [project @ 2002-07-22 02:34:28 by stevenj] preparing for recursive rdft... A ./rdft/hc2hc.c A ./rdft/hc2hc.h M ./rdft/Makefile.am -3 +3 M ./rdft/codelet.h -1 +2 M ./rdft/codelets/hfb.c -2 +2 M ./rdft/hc2hc.c +249 M ./rdft/hc2hc.h +69 Sun Jul 21 19:31:22 EDT 2002 stevenj * [project @ 2002-07-21 23:31:22 by stevenj] slight fix, to match libbench/verify.c M ./rdft/verify.c -3 +3 Sun Jul 21 18:43:12 EDT 2002 stevenj * [project @ 2002-07-21 22:43:12 by stevenj] r2hcII has imag parts offset by n-1, not n. We can also allocate fewer strides. M ./rdft/direct-r2hc.c -4 +7 Sun Jul 21 18:27:09 EDT 2002 stevenj * [project @ 2002-07-21 22:27:09 by stevenj] delete unused var M ./rdft/rank0.c -2 +2 Sun Jul 21 02:06:53 EDT 2002 stevenj * [project @ 2002-07-21 06:06:53 by stevenj] added some rdft solvers A ./rdft/direct-r2hc.c A ./rdft/indirect.c A ./rdft/nop.c A ./rdft/rank0.c A ./rdft/vrank-geq1.c A ./rdft/vrank2-transpose.c A ./rdft/vrank3-transpose.c M ./rdft/Makefile.am -2 +4 M ./rdft/codelet.h -3 +8 M ./rdft/codelets/hc2r.c -2 +2 M ./rdft/codelets/r2hc.c -2 +2 M ./rdft/conf.c -4 +4 M ./rdft/direct-r2hc.c +174 M ./rdft/indirect.c +235 R ./rdft/khc2rIII.c M ./rdft/kr2hc.c -5 +2 R ./rdft/kr2hcII.c M ./rdft/nop.c +97 M ./rdft/problem.c -3 +20 M ./rdft/rank0.c +217 M ./rdft/rdft.h -5 +4 M ./rdft/vrank-geq1.c +260 M ./rdft/vrank2-transpose.c +130 M ./rdft/vrank3-transpose.c +171 Sun Jul 21 01:52:54 EDT 2002 stevenj * [project @ 2002-07-21 05:52:54 by stevenj] pass identifier in FFTW() through another macro so that the mangled name can itself be a preprocessor symbol M ./kernel/fftw3.h -4 +5 Sun Jul 21 01:05:21 EDT 2002 stevenj * [project @ 2002-07-21 05:05:21 by stevenj] fix in comment M ./dft/vrank-geq1.c -2 +2 Sun Jul 21 00:47:03 EDT 2002 stevenj * [project @ 2002-07-21 04:47:03 by stevenj] bench tests rdft plans M ./Makefile.am -2 +5 M ./rdft/rdft.h -2 +3 M ./tests/bench.c -16 +36 Sun Jul 21 00:22:14 EDT 2002 stevenj * [project @ 2002-07-21 04:22:14 by stevenj] make rdft.h and dft.h compatible M ./rdft/codelet.h -4 +4 M ./tests/Makefile.am -1 +1 M ./tests/bench.c -1 +1 Sun Jul 21 00:12:19 EDT 2002 stevenj * [project @ 2002-07-21 04:12:19 by stevenj] first-draft rdft verify A ./rdft/verify.c M ./rdft/Makefile.am -1 +1 M ./rdft/problem.c -8 +8 M ./rdft/rdft.h -5 +5 M ./rdft/verify.c +554 Sat Jul 20 22:09:15 EDT 2002 stevenj * [project @ 2002-07-21 02:09:15 by stevenj] got rid of annoying warnings M ./rdft/khc2hc-dif.c -1 +2 M ./rdft/khc2hc-dit.c -1 +2 M ./rdft/khc2r.c -1 +2 M ./rdft/khc2rIII.c -1 +2 M ./rdft/kr2hc.c -1 +2 M ./rdft/kr2hcII.c -1 +2 Sat Jul 20 22:07:37 EDT 2002 stevenj * [project @ 2002-07-21 02:07:37 by stevenj] added stub codelet registration for linking purposes A ./rdft/khc2hc-dif.c A ./rdft/khc2hc-dit.c A ./rdft/khc2r.c A ./rdft/khc2rIII.c A ./rdft/kr2hc.c A ./rdft/kr2hcII.c M ./rdft/Makefile.am -1 +2 M ./rdft/khc2hc-dif.c +31 M ./rdft/khc2hc-dit.c +31 M ./rdft/khc2r.c +31 M ./rdft/khc2rIII.c +31 M ./rdft/kr2hc.c +31 M ./rdft/kr2hcII.c +31 M ./rdft/rdft.h -2 +2 Sat Jul 20 21:46:03 EDT 2002 stevenj * [project @ 2002-07-21 01:46:03 by stevenj] basic rdft stuff A ./rdft/conf.c A ./rdft/plan.c A ./rdft/problem.c A ./rdft/rdft.h A ./rdft/solve.c M ./rdft/Makefile.am -1 +1 M ./rdft/conf.c +47 M ./rdft/plan.c +33 M ./rdft/problem.c +145 M ./rdft/rdft.h +82 M ./rdft/solve.c +31 Sat Jul 20 21:06:50 EDT 2002 stevenj * [project @ 2002-07-21 01:06:50 by stevenj] rdft codelets now compile A ./rdft/Makefile.am A ./rdft/codelet.h A ./rdft/codelets/Makefile.am A ./rdft/codelets/hb.h A ./rdft/codelets/hc2r.c A ./rdft/codelets/hc2r.h A ./rdft/codelets/hc2rIII.h A ./rdft/codelets/hf.h A ./rdft/codelets/hfb.c A ./rdft/codelets/r2hc.c A ./rdft/codelets/r2hc.h A ./rdft/codelets/r2hcII.h M ./Makefile.am -1 +1 M ./configure.ac +2 M ./dft/codelet.h -33 +1 M ./genfft/gen_hc2hc.ml -3 +3 M ./kernel/ifftw.h -1 +35 M ./rdft/Makefile.am +6 M ./rdft/codelet.h +133 M ./rdft/codelets/Makefile.am +7 M ./rdft/codelets/hb.h +23 M ./rdft/codelets/hc2r/Makefile.am -3 +3 M ./rdft/codelets/hc2r.c +44 M ./rdft/codelets/hc2r.h +23 M ./rdft/codelets/hc2rIII.h +23 M ./rdft/codelets/hf.h +23 M ./rdft/codelets/hfb.c +41 M ./rdft/codelets/r2hc/Makefile.am -2 +2 M ./rdft/codelets/r2hc.c +44 M ./rdft/codelets/r2hc.h +23 M ./rdft/codelets/r2hcII.h +23 Sat Jul 20 18:40:31 EDT 2002 athena * [project @ 2002-07-20 22:40:31 by athena] Oops, was generating rdfts instead of hdfts M ./genfft/gen_hc2r.ml -3 +3 Sat Jul 20 18:25:47 EDT 2002 athena * [project @ 2002-07-20 22:25:47 by athena] Added hc2r codelets A ./rdft/codelets/hc2r/ A ./rdft/codelets/hc2r/Makefile.am M ./TODO +4 M ./configure.ac +1 M ./genfft/twiddle.ml -7 +14 M ./genfft-k7/twiddle.ml -12 +7 M ./kernel/twiddle.c -26 +39 M ./rdft/codelets/hc2r/Makefile.am -1 +42 Sat Jul 20 17:54:39 EDT 2002 athena * [project @ 2002-07-20 21:54:39 by athena] return W in hc2hc codelets M ./genfft/gen_hc2hc.ml -3 +4 Sat Jul 20 17:51:06 EDT 2002 athena * [project @ 2002-07-20 21:51:06 by athena] Some work on rdft codelets A ./rdft/ A ./rdft/codelets/ A ./rdft/codelets/r2hc/ A ./rdft/codelets/r2hc/Makefile.am M ./configure.ac +2 M ./dft/codelets/inplace/Makefile.am +4 M ./dft/codelets/standard/Makefile.am +3 M ./dft/simd/codelets/Makefile.am -1 +1 M ./genfft/gen_hc2hc.ml -62 +18 M ./genfft/gen_hc2r.ml -5 +24 M ./genfft/gen_r2hc.ml -5 +21 M ./genfft/trig.ml -22 +22 M ./rdft/codelets/r2hc/Makefile.am -1 +42 M ./support/Makefile.codelets -9 +6 Tue Jul 16 13:55:50 EDT 2002 athena * [project @ 2002-07-16 17:55:50 by athena] fix const M ./kernel/fftw3.h -3 +4 Tue Jul 16 07:00:10 EDT 2002 athena * [project @ 2002-07-16 11:00:10 by athena] Separate CFLAGS in codelets. Fix const in certain places. M ./acinclude.m4 -2 +6 M ./configure.ac +1 M ./dft/codelets/inplace/Makefile.am +1 M ./dft/codelets/standard/Makefile.am +1 M ./dft/indirect.c -2 +2 M ./dft/rank0.c -2 +2 M ./dft/simd/codelets/Makefile.am +1 M ./kernel/version.c -3 +4 M ./tests/bench.c -2 +9 Mon Jul 15 21:10:42 EDT 2002 stevenj * [project @ 2002-07-16 01:10:42 by stevenj] note buffering problem M ./TODO +2 Mon Jul 15 20:27:51 EDT 2002 athena * [project @ 2002-07-16 00:27:51 by athena] Removed unpredictable branch from inner loop M ./dft/generic.c -2 +2 Mon Jul 15 19:35:04 EDT 2002 stevenj * [project @ 2002-07-15 23:35:04 by stevenj] update M ./TODO -2 Mon Jul 15 19:31:39 EDT 2002 stevenj * [project @ 2002-07-15 23:31:39 by stevenj] optimization M ./dft/generic.c -1 +2 Mon Jul 15 19:28:30 EDT 2002 stevenj * [project @ 2002-07-15 23:28:30 by stevenj] added generic dit A ./dft/generic.c M ./dft/Makefile.am -4 +4 M ./dft/conf.c -1 +2 M ./dft/dft.h -1 +2 M ./dft/generic.c +219 M ./kernel/ifftw.h -2 +3 M ./kernel/twiddle.c -1 +13 Mon Jul 15 17:03:53 EDT 2002 stevenj * [project @ 2002-07-15 21:03:53 by stevenj] whoops, mksolver should be static M ./dft/rader.c -2 +2 Mon Jul 15 16:46:36 EDT 2002 athena * [project @ 2002-07-15 20:46:35 by athena] First implementation of gen_hc2hc, probably still buggy. A ./genfft/gen_hc2hc.ml M ./genfft/Makefile.am -9 +15 M ./genfft/algsimp.ml -4 +5 M ./genfft/c.ml -3 +10 M ./genfft/c.mli -1 +2 M ./genfft/gen_hc2hc.ml +220 M ./genfft/gen_hc2r.ml -24 +69 M ./genfft/gen_r2hc.ml -22 +66 M ./genfft/genutil.ml -13 +17 Mon Jul 15 16:40:23 EDT 2002 stevenj * [project @ 2002-07-15 20:40:23 by stevenj] don't count loading of twiddle factors in ops.other, since it isn't counted for the codelets M ./dft/rader.c -1 Mon Jul 15 15:13:19 EDT 2002 stevenj * [project @ 2002-07-15 19:13:19 by stevenj] plan_destroy puts plan to sleep before deallocating it, to eliminate duplicate free calls in solvers M ./dft/ct.c -2 +1 M ./dft/rader.c -2 M ./kernel/plan.c -2 +5 Mon Jul 15 15:07:41 EDT 2002 stevenj * [project @ 2002-07-15 19:07:41 by stevenj] fftw2-like vector recursion flag M ./dft/ct-dif.c -1 +6 M ./dft/ct-dit.c -2 +2 M ./dft/ct-ditbuf.c -2 +2 M ./dft/ct.c -1 +5 M ./dft/vrank-geq1.c -2 +10 M ./kernel/ifftw.h -2 +6 M ./tests/bench.c -2 +1 Sun Jul 14 21:01:44 EDT 2002 athena * [project @ 2002-07-15 01:01:44 by athena] More jokes M ./kernel/planner.c -1 +2 Sun Jul 14 20:36:01 EDT 2002 athena * [project @ 2002-07-15 00:36:01 by athena] Bless plan for testing purposes M ./tests/bench.c +1 Sun Jul 14 20:35:49 EDT 2002 athena * [project @ 2002-07-15 00:35:49 by athena] Canonical linked-list deletion (hope it is right) M ./kernel/planner.c -12 +13 Sun Jul 14 18:26:19 EDT 2002 stevenj * [project @ 2002-07-14 22:26:19 by stevenj] use estimating planner for cld_omega M ./dft/rader.c -1 +1 Sun Jul 14 18:10:56 EDT 2002 stevenj * [project @ 2002-07-14 22:10:56 by stevenj] better internal naming M ./dft/rader.c -3 +3 Sun Jul 14 18:10:01 EDT 2002 stevenj * [project @ 2002-07-14 22:10:01 by stevenj] printing should really be fixed now, grrr M ./dft/rader.c -10 +12 Sun Jul 14 17:57:12 EDT 2002 stevenj * [project @ 2002-07-14 21:57:12 by stevenj] print all distinct child plans M ./dft/rader.c -2 +7 Sun Jul 14 17:49:21 EDT 2002 stevenj * [project @ 2002-07-14 21:49:21 by stevenj] whoops M ./tests/bench.c -1 +1 Sun Jul 14 17:45:54 EDT 2002 stevenj * [project @ 2002-07-14 21:45:54 by stevenj] whoops, destroy should delete twiddle/omega from list M ./dft/rader.c -4 +2 Sun Jul 14 17:33:02 EDT 2002 stevenj * [project @ 2002-07-14 21:33:02 by stevenj] whoops M ./kernel/planner.c -3 +6 Sun Jul 14 17:12:14 EDT 2002 stevenj * [project @ 2002-07-14 21:12:14 by stevenj] added plan_bless and FORGET_ACCURSED M ./kernel/ifftw.h -2 +6 M ./kernel/plan.c -1 +14 M ./kernel/planner.c -13 +24 Sun Jul 14 16:15:43 EDT 2002 stevenj * [project @ 2002-07-14 20:15:43 by stevenj] further cleanup M ./kernel/traverse.c -6 +4 Sun Jul 14 16:14:15 EDT 2002 stevenj * [project @ 2002-07-14 20:14:15 by stevenj] slight cleanup M ./kernel/traverse.c -8 +7 Sun Jul 14 16:09:17 EDT 2002 stevenj * [project @ 2002-07-14 20:09:17 by stevenj] added traverse_plan via print (ugh) A ./kernel/traverse.c M ./kernel/Makefile.am -3 +3 M ./kernel/ifftw.h -1 +5 M ./kernel/traverse.c +115 M ./tests/bench.c +12 Sun Jul 14 15:08:29 EDT 2002 stevenj * [project @ 2002-07-14 19:08:29 by stevenj] added TW_FULL, and additional n parameter for twiddles M ./dft/ct.c -3 +4 M ./kernel/ifftw.h -5 +5 M ./kernel/twiddle.c -18 +34 Sun Jul 14 15:03:51 EDT 2002 stevenj * [project @ 2002-07-14 19:03:51 by stevenj] whoops M ./kernel/planner.c -2 +4 Sun Jul 14 13:49:20 EDT 2002 stevenj * [project @ 2002-07-14 17:49:20 by stevenj] save flags before invoking solver mkplan M ./kernel/ifftw.h -1 +2 M ./kernel/planner-naive.c -2 +2 M ./kernel/planner-score.c -2 +2 M ./kernel/planner.c -3 +10 Sun Jul 14 09:28:37 EDT 2002 athena * [project @ 2002-07-14 13:28:37 by athena] M ./TODO -6 +1 Sat Jul 13 22:17:29 EDT 2002 stevenj * [project @ 2002-07-14 02:17:29 by stevenj] added support for UNICOS _rtc() real-time-clock intrinsic function M ./configure.ac -1 +10 M ./kernel/cycle.h -2 +20 Sat Jul 13 22:06:35 EDT 2002 stevenj * [project @ 2002-07-14 02:06:35 by stevenj] fixed typo: HAVE_TIME_H should include time.h, not sys/time.h M ./kernel/timer.c -2 +2 Sat Jul 13 21:46:02 EDT 2002 stevenj * [project @ 2002-07-14 01:46:02 by stevenj] support AIX read_real_time timer M ./configure.ac +1 M ./kernel/cycle.h -1 +24 Sat Jul 13 17:02:51 EDT 2002 stevenj * [project @ 2002-07-13 21:02:51 by stevenj] use && instead of the (sigh) unportable -a M ./configure.ac -1 +1 Sat Jul 13 16:38:18 EDT 2002 stevenj * [project @ 2002-07-13 20:38:18 by stevenj] use AC_HELP_STRING M ./configure.ac -11 +11 Sat Jul 13 16:05:43 EDT 2002 stevenj * [project @ 2002-07-13 20:05:43 by stevenj] support long-double precision A ./kernel/lfftw3.h M ./configure.ac -2 +22 M ./dft/codelet.h -2 +6 M ./dft/verify.c -10 +29 M ./kernel/Makefile.am -3 +3 M ./kernel/dfftw3.h -1 +2 M ./kernel/fftw3.h -2 +5 M ./kernel/ifftw.h -9 +14 M ./kernel/lfftw3.h +25 M ./kernel/sfftw3.h -1 +2 M ./libbench/bench-main.c -3 +5 M ./libbench/bench-user.h -4 +8 M ./libbench/info.c -2 +4 M ./libbench/verify.c -10 +38 M ./simd/simd-sse2.h -1 +1 Sat Jul 13 15:48:10 EDT 2002 stevenj * [project @ 2002-07-13 19:48:10 by stevenj] whoops whoops M ./dft/rader.c -2 +2 Sat Jul 13 15:47:39 EDT 2002 stevenj * [project @ 2002-07-13 19:47:39 by stevenj] whoops M ./dft/rader.c -1 +1 Sat Jul 13 14:13:42 EDT 2002 stevenj * [project @ 2002-07-13 18:13:42 by stevenj] buffered solver strides have been fixed M ./TODO -7 Sat Jul 13 13:48:13 EDT 2002 stevenj * [project @ 2002-07-13 17:48:13 by stevenj] convention M ./dft/rader.c -1 +1 Sat Jul 13 12:50:06 EDT 2002 stevenj * [project @ 2002-07-13 16:50:06 by stevenj] share twiddle arrays in Rader M ./TODO -2 M ./dft/rader.c -5 +68 Sat Jul 13 12:48:10 EDT 2002 stevenj * [project @ 2002-07-13 16:48:10 by stevenj] call done() after verify M ./libbench/verify.c -1 +2 Fri Jul 12 15:42:04 EDT 2002 stevenj * [project @ 2002-07-12 19:42:04 by stevenj] output planner time with -v M ./tests/bench.c +7 Fri Jul 12 15:40:14 EDT 2002 stevenj * [project @ 2002-07-12 19:40:14 by stevenj] support double outputs M ./kernel/print.c -2 +9 Fri Jul 12 15:09:19 EDT 2002 stevenj * [project @ 2002-07-12 19:09:19 by stevenj] removed extraneous parens M ./dft/vrank-geq1.c -2 +2 Fri Jul 12 15:08:13 EDT 2002 stevenj * [project @ 2002-07-12 19:08:13 by stevenj] increase maxbufsz to 64k; makes a big difference for large 2d transforms M ./dft/buffered.c -2 +2 Fri Jul 12 05:59:26 EDT 2002 athena * [project @ 2002-07-12 09:59:26 by athena] Fix M ./dft/vrank-geq1.c -2 +2 Fri Jul 12 01:22:38 EDT 2002 stevenj * [project @ 2002-07-12 05:22:38 by stevenj] fix comment M ./dft/rank-geq2.c -2 +2 Fri Jul 12 00:59:29 EDT 2002 stevenj * [project @ 2002-07-12 04:59:29 by stevenj] fix in comment M ./kernel/tensor.c -2 +2 Fri Jul 12 00:13:13 EDT 2002 stevenj * [project @ 2002-07-12 04:13:13 by stevenj] updated M ./ChangeLog +873 Thu Jul 11 23:39:27 EDT 2002 stevenj * [project @ 2002-07-12 03:39:27 by stevenj] buffered malloc's buffers M ./TODO -3 Thu Jul 11 23:30:26 EDT 2002 stevenj * [project @ 2002-07-12 03:30:26 by stevenj] share more code between apply and apply_dit in Rader M ./TODO -2 M ./dft/rader.c -69 +30 Mon Jul 8 12:30:34 EDT 2002 athena * [project @ 2002-07-08 16:30:34 by athena] Polished M ./simd/simd-sse.h -18 +28 M ./simd/simd-sse2.h -25 +31 M ./simd/sse.c -4 +2 M ./simd/sse2.c -3 +2 Mon Jul 8 09:47:11 EDT 2002 athena * [project @ 2002-07-08 13:47:11 by athena] M ./support/Makefile.codelets -1 +1 Mon Jul 8 09:42:08 EDT 2002 athena * [project @ 2002-07-08 13:42:08 by athena] SIMD/FMA stuff M ./dft/simd/codelets/Makefile.am -1 +1 M ./genfft/c.ml -7 +2 M ./genfft/gen_notw_c.ml -2 +3 M ./genfft/simd.ml -2 +13 M ./genfft/to_alist.ml -1 +3 M ./genfft/to_alist.mli -1 +2 M ./simd/simd-sse.h +3 M ./simd/simd-sse2.h +3 M ./support/Makefile.codelets -1 +2 Mon Jul 8 07:43:51 EDT 2002 athena * [project @ 2002-07-08 11:43:51 by athena] Avoid code duplication M ./simd/simd-sse.h -12 +6 Sun Jul 7 20:56:15 EDT 2002 athena * [project @ 2002-07-08 00:56:15 by athena] Fixes for FMA+SIMD M ./genfft/Makefile.am -3 +3 M ./genfft/to_alist.ml -7 +14 Sun Jul 7 20:32:01 EDT 2002 athena * [project @ 2002-07-08 00:32:01 by athena] Major changes in SIMD fftw A ./dft/simd/n1b.c A ./dft/simd/n1b.h A ./dft/simd/n1f.c A ./dft/simd/n1f.h A ./dft/simd/t1b.c A ./dft/simd/t1b.h A ./dft/simd/t1f.c A ./dft/simd/t1f.h A ./genfft/gen_notw_c.ml A ./genfft/gen_twiddle_c.ml M ./dft/buffered.c -2 +2 M ./dft/codelets/standard/Makefile.am -6 +5 M ./dft/simd/Makefile.am -3 +1 R ./dft/simd/NAMING M ./dft/simd/codelets/Makefile.am -26 +23 M ./dft/simd/n1b.c +43 M ./dft/simd/n1b.h +24 M ./dft/simd/n1f.c +43 M ./dft/simd/n1f.h +24 R ./dft/simd/n2f.c R ./dft/simd/n2f.h R ./dft/simd/n3f.h R ./dft/simd/n4.c R ./dft/simd/n4.h M ./dft/simd/t1b.c +38 M ./dft/simd/t1b.h +25 M ./dft/simd/t1f.c +38 M ./dft/simd/t1f.h +25 R ./dft/simd/t2f.c R ./dft/simd/t2f.h R ./dft/simd/t3f.h R ./dft/simd/t4.c R ./dft/simd/t4.h M ./genfft/Makefile.am -29 +27 M ./genfft/algsimp.ml -16 +27 M ./genfft/annotate.ml -131 +11 M ./genfft/annotate.mli -17 +2 M ./genfft/c.ml -7 +23 M ./genfft/complex.ml -1 +3 M ./genfft/complex.mli -1 +2 M ./genfft/expr.ml -3 +21 M ./genfft/expr.mli -1 +6 M ./genfft/gen_athnotw.ml -3 +3 M ./genfft/gen_athtw.ml -3 +3 M ./genfft/gen_conv.ml -4 +4 M ./genfft/gen_hc2r.ml -4 +4 M ./genfft/gen_notw.ml -4 +4 M ./genfft/gen_notw_c.ml +155 M ./genfft/gen_r2hc.ml -4 +4 M ./genfft/gen_trig.ml -4 +4 M ./genfft/gen_twiddle.ml -4 +4 M ./genfft/gen_twiddle_c.ml +165 M ./genfft/gen_twidsq.ml -4 +4 M ./genfft/genutil.ml -4 +5 M ./genfft/oracle.ml -1 +2 M ./genfft/simd.ml -177 +95 M ./genfft/simd.mli -2 +2 M ./genfft/simdmagic.ml -47 +1 M ./genfft/to_alist.ml -3 +9 M ./genfft/trig.ml -1 +10 M ./genfft/trig.mli -1 +2 M ./genfft/twiddle.ml -5 +6 M ./kernel/ifftw.h -2 +2 M ./simd/simd-sse.h -61 +88 M ./simd/simd-sse2.h -67 +57 M ./simd/sse.c -1 +4 M ./simd/sse2.c -1 +4 M ./support/Makefile.codelets -67 +2 Fri Jul 5 17:32:09 EDT 2002 athena * [project @ 2002-07-05 21:32:09 by athena] Use unpck instructions instead of shuffles M ./dft/buffered.c -2 +2 M ./simd/simd-altivec.h -13 +13 M ./simd/simd-sse.h -23 +10 Fri Jul 5 15:49:14 EDT 2002 athena * [project @ 2002-07-05 19:49:14 by athena] Minor tweaks M ./dft/codelets/n.c +1 M ./dft/codelets/t.c +1 M ./dft/ct-ditbuf.c -2 +2 M ./dft/verify.c -4 +1 M ./kernel/ifftw.h -10 +9 M ./kernel/planner.c -5 +4 M ./tests/bench.c -2 +2 Fri Jul 5 15:02:54 EDT 2002 athena * [project @ 2002-07-05 19:02:54 by athena] Use score planner M ./tests/bench.c -1 +1 Fri Jul 5 14:49:59 EDT 2002 athena * [project @ 2002-07-05 18:49:59 by athena] Added verifier A ./dft/verify.c A ./kernel/dotens.c A ./kernel/dotens2.c M ./CONVENTIONS +1 M ./dft/Makefile.am -1 +1 M ./dft/dft.h -1 +4 M ./dft/verify.c +428 M ./kernel/Makefile.am -1 +2 M ./kernel/debug.c -2 +2 M ./kernel/dotens.c +47 M ./kernel/dotens2.c +55 M ./kernel/ifftw.h -8 +24 M ./tests/bench.c -8 +14 Wed Jul 3 20:32:28 EDT 2002 athena * [project @ 2002-07-04 00:32:28 by athena] More simd codelets A ./dft/simd/n2f.c A ./dft/simd/n2f.h A ./dft/simd/n3f.h A ./dft/simd/t2f.c A ./dft/simd/t2f.h A ./dft/simd/t3f.h M ./dft/buffered.c -3 +3 M ./dft/codelet.h -1 +3 M ./dft/codelets/n.c +1 M ./dft/codelets/t.c +1 M ./dft/ct-dif.c -1 +6 M ./dft/ct-dit.c -1 +6 M ./dft/ct-ditbuf.c -9 +19 M ./dft/ct.c -11 +17 M ./dft/ct.h -3 +5 M ./dft/simd/Makefile.am -1 +2 M ./dft/simd/NAMING -7 +20 M ./dft/simd/codelets/Makefile.am -4 +27 M ./dft/simd/n2f.c +43 M ./dft/simd/n2f.h +25 M ./dft/simd/n3f.h +1 M ./dft/simd/n4.c -2 +1 M ./dft/simd/t2f.c +40 M ./dft/simd/t2f.h +25 M ./dft/simd/t3f.h +1 M ./dft/simd/t4.c -1 +1 M ./genfft/annotate.ml -3 +3 M ./genfft/gen_notw.ml -9 +23 M ./genfft/gen_twiddle.ml -6 +14 M ./genfft/gen_twidsq.ml -3 +9 M ./genfft/simd.ml -43 +69 M ./genfft/simdmagic.ml -11 +36 M ./genfft-k7/gen_notw.ml -1 +2 M ./genfft-k7/gen_twiddle.ml -1 +2 M ./kernel/alloc.c -5 +2 M ./kernel/ifftw.h -3 +9 M ./simd/simd-sse.h -28 +106 M ./simd/simd-sse2.h -24 +47 Tue Jul 2 16:18:09 EDT 2002 athena * [project @ 2002-07-02 20:18:09 by athena] Oops M ./dft/rank-geq2.c -3 +4 Tue Jul 2 16:13:24 EDT 2002 athena * [project @ 2002-07-02 20:13:24 by athena] Fixed classic mode M ./dft/rank-geq2.c -6 +5 M ./dft/vrank-geq1.c -6 +6 Tue Jul 2 15:38:36 EDT 2002 athena * [project @ 2002-07-02 19:38:36 by athena] Use LDK for constants so that we can play games. M ./genfft/simd.ml -2 +2 M ./simd/simd-altivec.h +1 M ./simd/simd-sse.h -5 +5 M ./simd/simd-sse2.h -2 +5 Tue Jul 2 13:15:58 EDT 2002 athena * [project @ 2002-07-02 17:15:58 by athena] Improved support for fixed strides M ./dft/codelet.h -1 +2 M ./dft/codelets/n.c -1 +2 M ./dft/simd/n4.c +3 M ./dft/simd/t4.c +2 M ./genfft/gen_notw.ml -8 +21 M ./genfft/genutil.ml -1 +6 M ./genfft/simd.ml -7 +7 M ./genfft/simd.mli -3 +3 M ./genfft-k7/gen_notw.ml -1 +2 M ./simd/simd-sse.h -17 +54 Tue Jul 2 10:30:58 EDT 2002 athena * [project @ 2002-07-02 14:30:58 by athena] Changed accounting of flops M ./dft/codelet.h -12 +25 M ./dft/codelets/n.c -4 +5 M ./dft/codelets/n.h -6 +2 M ./dft/codelets/t.c -3 +5 M ./dft/codelets/t.h -3 +2 M ./dft/ct-dif.c -3 +5 M ./dft/ct-dit.c -3 +5 M ./dft/ct-ditbuf.c -3 +4 M ./dft/ct-ditf.c -3 +6 M ./dft/direct.c -5 +5 M ./dft/k7/k7.c -14 +23 M ./dft/simd/n4.c -3 +5 M ./dft/simd/n4.h -5 +2 M ./dft/simd/t4.c -3 +5 M ./dft/simd/t4.h -3 +2 M ./genfft/gen_notw.ml -3 +3 M ./genfft/gen_twiddle.ml -3 +3 M ./genfft/gen_twidsq.ml -3 +3 M ./genfft-k7/gen_notw.ml -2 +2 M ./genfft-k7/gen_twiddle.ml -2 +2 Tue Jul 2 08:51:38 EDT 2002 athena * [project @ 2002-07-02 12:51:38 by athena] Wrong code in non-fma mode M ./genfft/algsimp.ml -11 +15 M ./genfft-k7/algsimp.ml -8 +24 M ./genfft-k7/to_alist.ml -3 +3 M ./simd/simd-sse2.h -4 +9 Mon Jul 1 23:17:06 EDT 2002 athena * [project @ 2002-07-02 03:17:06 by athena] sse2 stuff A ./simd/simd-sse2.h A ./simd/sse2.c M ./genfft/simdmagic.ml -2 +2 M ./kernel/alloc.c -2 +2 M ./simd/Makefile.am -1 +2 M ./simd/simd-sse2.h +125 M ./simd/sse2.c +73 Mon Jul 1 14:05:56 EDT 2002 athena * [project @ 2002-07-01 18:05:56 by athena] Identify CPUs for special codelets A ./simd/sse.c M ./Makefile.am +1 M ./dft/ct.c -4 +4 M ./dft/direct.c -2 +2 M ./dft/k7/k7.c -5 +43 M ./dft/simd/n4.c -1 +1 M ./dft/simd/t4.c -1 +1 M ./kernel/alloc.c -2 +2 M ./simd/Makefile.am -1 +3 M ./simd/simd-altivec.h +2 M ./simd/simd-sse.h +3 M ./simd/sse.c +73 Mon Jul 1 09:26:42 EDT 2002 athena * [project @ 2002-07-01 13:26:42 by athena] Change split problem syntax M ./libbench/problem.c -2 +2 Mon Jul 1 09:11:39 EDT 2002 athena * [project @ 2002-07-01 13:11:39 by athena] Removed -fma flag M ./dft/simd/codelets/Makefile.am -1 +1 Sun Jun 30 20:08:26 EDT 2002 athena * [project @ 2002-07-01 00:08:26 by athena] Work around gcc bug M ./simd/simd-altivec.h -3 +17 Sun Jun 30 18:34:06 EDT 2002 athena * [project @ 2002-06-30 22:34:06 by athena] New simd stuff M ./genfft/algsimp.ml -8 +20 M ./genfft/magic.ml -1 +3 M ./genfft/oracle.ml -2 +2 M ./genfft/simd.ml -2 +4 M ./genfft/to_alist.ml -3 +3 Sun Jun 30 17:00:09 EDT 2002 athena * [project @ 2002-06-30 21:00:09 by athena] Added altivec support A ./simd/simd-altivec.h M ./dft/simd/codelets/Makefile.am -1 +1 M ./simd/Makefile.am -1 +1 M ./simd/simd-altivec.h +66 M ./simd/simd-sse.h -1 Sun Jun 30 14:47:47 EDT 2002 athena * [project @ 2002-06-30 18:47:47 by athena] Forgot file A ./dft/simd/t4.c Sun Jun 30 14:37:55 EDT 2002 athena * [project @ 2002-06-30 18:37:55 by athena] Progress towards simd implementation A ./dft/codelets/f.h A ./dft/codelets/n.c A ./dft/codelets/n.h A ./dft/codelets/q.h A ./dft/codelets/t.c A ./dft/codelets/t.h A ./dft/k7/k7.c A ./dft/simd/ A ./dft/simd/Makefile.am A ./dft/simd/NAMING A ./dft/simd/codelets/ A ./dft/simd/codelets/Makefile.am A ./dft/simd/n4.c A ./dft/simd/n4.h A ./dft/simd/t4.h A ./simd/ A ./simd/Makefile.am A ./simd/README A ./simd/simd-sse.h A ./simd/simd.h M ./Makefile.am -2 +5 M ./configure.ac -6 +32 M ./dft/Makefile.am -3 +3 R ./dft/codelet-k7.h M ./dft/codelet.h -11 +23 M ./dft/codelets/Makefile.am +4 M ./dft/codelets/f.h +1 M ./dft/codelets/inplace/Makefile.am -13 +14 M ./dft/codelets/n.c +34 M ./dft/codelets/n.h +27 M ./dft/codelets/q.h +1 M ./dft/codelets/standard/Makefile.am -13 +14 M ./dft/codelets/t.c +32 M ./dft/codelets/t.h +24 M ./dft/conf.c -2 +5 M ./dft/ct-dif.c -3 +3 M ./dft/ct-dit.c -5 +3 M ./dft/ct-ditbuf.c -2 +2 M ./dft/ct-ditf.c -4 +3 M ./dft/ct.c -6 +6 M ./dft/ct.h -3 +1 M ./dft/direct.c -6 +11 M ./dft/k7/Makefile.am -2 +1 M ./dft/k7/codelets/Makefile.am -33 +29 R ./dft/k7/ct-dif.c R ./dft/k7/ct-dit.c R ./dft/k7/ct-ditbuf.c R ./dft/k7/direct.c M ./dft/k7/k7.c +69 R ./dft/k7/kdft-dif.c R ./dft/k7/kdft-dit.c R ./dft/k7/kdft.c M ./dft/simd/Makefile.am +6 M ./dft/simd/NAMING +17 M ./dft/simd/codelets/Makefile.am +36 M ./dft/simd/n4.c +39 M ./dft/simd/n4.h +28 M ./dft/simd/t4.h +26 M ./genfft/annotate.ml -7 +7 M ./genfft/c.ml -1 +6 M ./genfft/c.mli -1 +2 M ./genfft/gen_notw.ml -12 +20 M ./genfft/gen_twiddle.ml -14 +18 M ./genfft/gen_twidsq.ml -7 +7 M ./genfft/genutil.ml -4 +17 M ./genfft/magic.ml -1 +3 M ./genfft/simd.ml -66 +34 M ./genfft/simd.mli -1 +7 M ./genfft/simdmagic.ml -2 +12 M ./genfft/twiddle.ml -1 +12 M ./genfft/twiddle.mli -1 +2 M ./genfft-k7/genUtil.ml -1 +1 M ./genfft-k7/gen_notw.ml -11 +18 M ./genfft-k7/gen_twiddle.ml -12 +17 M ./kernel/alloc.c -9 +28 M ./kernel/ifftw.h -10 +10 M ./libbench/bench-user.h -1 +2 M ./libbench/problem.c -1 +3 M ./libbench/util.c -1 M ./simd/Makefile.am +1 M ./simd/README +2 M ./simd/simd-sse.h +94 M ./simd/simd.h +33 M ./support/Makefile.codelets -33 +36 M ./tests/Makefile.am +1 M ./tests/bench.c -14 +63 Tue Jun 25 20:23:29 EDT 2002 athena * [project @ 2002-06-26 00:23:29 by athena] Add 128- codelet M ./dft/k7/codelets/Makefile.am -2 +3 Sat Jun 22 20:47:28 EDT 2002 athena * [project @ 2002-06-23 00:47:28 by athena] More simd changes. Ensure proper stack alignment in k7 codelets. M ./configure.ac +1 M ./genfft/c.ml -25 +4 M ./genfft/c.mli -2 +4 M ./genfft/expr.ml -1 +18 M ./genfft/expr.mli -1 +4 M ./genfft/gen_hc2r.ml -3 +3 M ./genfft/gen_notw.ml -3 +3 M ./genfft/gen_r2hc.ml -3 +3 M ./genfft/gen_trig.ml -3 +3 M ./genfft/gen_twiddle.ml -3 +3 M ./genfft/gen_twidsq.ml -3 +3 M ./genfft/genutil.ml -1 +13 M ./genfft/simd.ml -1 +9 M ./genfft/simd.mli -1 +2 M ./genfft-k7/genUtil.ml -34 +6 M ./genfft-k7/gen_notw.ml -2 +2 M ./genfft-k7/gen_twiddle.ml -2 +2 M ./libbench/bench-main.c -1 +11 Sat Jun 22 13:01:33 EDT 2002 athena * [project @ 2002-06-22 17:01:33 by athena] Fixed prototypes M ./kernel/ifftw.h -2 +2 M ./kernel/solvtab.c -2 +2 Sat Jun 22 12:53:26 EDT 2002 athena * [project @ 2002-06-22 16:53:26 by athena] Sparc cycle counter requires v9 M ./kernel/cycle.h -2 +2 Sat Jun 22 11:45:48 EDT 2002 athena * [project @ 2002-06-22 15:45:48 by athena] Minor fixes M ./configure.ac -1 +1 M ./kernel/cycle.h -15 +36 M ./kernel/ifftw.h -1 +5 Sat Jun 22 11:11:46 EDT 2002 athena * [project @ 2002-06-22 15:11:46 by athena] Fixed ev67 detection M ./acinclude.m4 -1 +1 Sat Jun 22 10:19:11 EDT 2002 athena * [project @ 2002-06-22 14:19:11 by athena] Print flops M ./tests/bench.c +2 Sat Jun 22 09:36:47 EDT 2002 athena * [project @ 2002-06-22 13:36:47 by athena] Nothing really M ./genfft/simd.ml -7 +3 Fri Jun 21 22:19:20 EDT 2002 athena * [project @ 2002-06-22 02:19:20 by athena] More simd work M ./dft/codelet-k7.h -2 +2 M ./dft/codelet.h -3 +3 M ./genfft/Makefile.am -2 +2 M ./genfft/c.ml -19 +1 M ./genfft/c.mli -2 +2 M ./genfft/gen_conv.ml -3 +3 M ./genfft/gen_hc2r.ml -3 +3 M ./genfft/gen_notw.ml -3 +3 M ./genfft/gen_r2hc.ml -3 +3 M ./genfft/gen_trig.ml -3 +3 M ./genfft/gen_twiddle.ml -3 +3 M ./genfft/gen_twidsq.ml -3 +3 M ./genfft/genutil.ml -2 +20 M ./genfft/magic.ml -1 +5 M ./genfft/simd.ml -47 +39 M ./genfft/simd.mli -1 +3 M ./genfft/simdmagic.ml -4 +2 M ./genfft/to_alist.ml -2 +5 M ./genfft-k7/magic.ml +4 M ./genfft-k7/to_alist.ml -2 +5 M ./genfft-k7/to_alist.mli -1 +1 M ./kernel/ifftw.h -2 +2 Thu Jun 20 21:22:41 EDT 2002 athena * [project @ 2002-06-21 01:22:41 by athena] More simd work A ./genfft/simd.mli A ./genfft/simdmagic.ml M ./genfft/Makefile.am -7 +8 M ./genfft/annotate.ml -13 +13 M ./genfft/annotate.mli -1 +2 M ./genfft/magic.ml -21 +1 M ./genfft/simd.ml -21 +10 M ./genfft/simd.mli +21 M ./genfft/simdmagic.ml +43 Thu Jun 20 18:51:33 EDT 2002 athena * [project @ 2002-06-20 22:51:33 by athena] More simd work A ./genfft/simd.ml M ./genfft/Makefile.am -23 +7 M ./genfft/annotate.ml -18 +2 M ./genfft/c.ml -6 +1 M ./genfft/c.mli -5 +8 M ./genfft/magic.ml -1 +6 M ./genfft/simd.ml +288 M ./genfft/variable.ml -1 +18 M ./genfft/variable.mli -1 +4 Thu Jun 20 15:04:37 EDT 2002 athena * [project @ 2002-06-20 19:04:37 by athena] Moving towards incorporation of simd stuff M ./genfft/annotate.ml -11 +144 M ./genfft/annotate.mli -2 +16 M ./genfft/gen_athnotw.ml -5 +5 M ./genfft/gen_athtw.ml -6 +6 M ./genfft/gen_conv.ml -3 +3 M ./genfft/gen_hc2r.ml -3 +3 M ./genfft/gen_notw.ml -3 +3 M ./genfft/gen_r2hc.ml -3 +3 M ./genfft/gen_trig.ml -3 +3 M ./genfft/gen_twiddle.ml -4 +4 M ./genfft/gen_twidsq.ml -6 +5 M ./genfft/genutil.ml -26 +17 M ./genfft/magic.ml -2 +17 M ./genfft/twiddle.ml -4 +4 M ./genfft/twiddle.mli -2 +2 M ./genfft/variable.ml -12 +22 M ./genfft/variable.mli -3 +6 Wed Jun 19 18:47:55 EDT 2002 athena * [project @ 2002-06-19 22:47:55 by athena] Reorganized k7 stuff into own directory A ./dft/k7/ A ./dft/k7/Makefile.am A ./dft/k7/codelets/ A ./dft/k7/codelets/Makefile.am A ./dft/k7/ct-dif.c A ./dft/k7/ct-dit.c A ./dft/k7/ct-ditbuf.c A ./dft/k7/direct.c A ./dft/k7/kdft-dif.c A ./dft/k7/kdft-dit.c A ./dft/k7/kdft.c M ./Makefile.am -1 +2 M ./configure.ac -2 +3 M ./dft/Makefile.am -6 +5 M ./dft/codelets/Makefile.am -1 +1 R ./dft/ct-dif-k7.c R ./dft/ct-dit-k7.c R ./dft/ct-ditbuf-k7.c R ./dft/direct-k7.c M ./dft/k7/Makefile.am +7 M ./dft/k7/codelets/Makefile.am +74 M ./dft/k7/ct-dif.c +129 M ./dft/k7/ct-dit.c +137 M ./dft/k7/ct-ditbuf.c +176 M ./dft/k7/direct.c +172 M ./dft/k7/kdft-dif.c +30 M ./dft/k7/kdft-dit.c +31 M ./dft/k7/kdft.c +31 R ./dft/kdft-dif-k7.c R ./dft/kdft-dit-k7.c R ./dft/kdft-k7.c Wed Jun 19 13:21:13 EDT 2002 athena * [project @ 2002-06-19 17:20:37 by athena] Minor experimental stuff M ./genfft/expr.ml -3 +2 M ./genfft/expr.mli -2 +2 M ./genfft/genutil.ml -2 +11 M ./genfft/magic.ml -1 +3 M ./genfft-k7/expr.ml -1 +3 M ./genfft-k7/expr.mli -1 +3 Wed Jun 19 11:20:29 EDT 2002 athena * [project @ 2002-06-19 15:20:29 by athena] Cosmetic changes M ./genfft/expr.ml -1 +4 M ./genfft/expr.mli -1 +3 M ./genfft/genutil.ml -9 +2 Wed Jun 19 01:43:31 EDT 2002 fftw * [project @ 2002-06-19 05:43:31 by fftw] allocate buffers on the fly M ./dft/buffered.c -18 +7 M ./dft/rader.c -17 +20 Tue Jun 18 17:48:41 EDT 2002 athena * [project @ 2002-06-18 21:48:41 by athena] Added ct-ditbuf-k7.c . Major changes required in generator. A ./dft/ct-ditbuf-k7.c M ./dft/Makefile.am -1 +1 M ./dft/codelet-k7.h -1 +2 M ./dft/ct-dif-k7.c -17 +3 M ./dft/ct-dif.c -16 +2 M ./dft/ct-dit-k7.c -18 +10 M ./dft/ct-dit.c -16 +2 M ./dft/ct-ditbuf-k7.c +176 M ./dft/ct-ditbuf.c -16 +2 M ./dft/ct.c -1 +30 M ./dft/ct.h -1 +3 M ./dft/kdft-dit-k7.c -1 +2 M ./dft/rader.c -1 +1 M ./genfft/number.ml -2 +2 M ./genfft/to_alist.ml -2 +2 M ./genfft-k7/Makefile.am -4 +4 M ./genfft-k7/assignmentsToVfpinstrs.ml -1 +4 M ./genfft-k7/gen_twiddle.ml -2 +5 M ./genfft-k7/k7Basics.ml -1 +1 M ./genfft-k7/k7Basics.mli +1 M ./genfft-k7/k7RegisterAllocationBasics.ml +3 M ./genfft-k7/k7RegisterAllocationBasics.mli -2 +4 M ./genfft-k7/k7RegisterAllocator.ml -15 +28 M ./genfft-k7/k7RegisterAllocatorInit.ml -6 +12 M ./genfft-k7/number.ml -2 +2 M ./genfft-k7/to_alist.ml -2 +2 M ./kernel/ifftw.h -2 +2 M ./kernel/planner.c -3 +5 M ./kernel/primes.c -2 +2 Tue Jun 18 11:55:57 EDT 2002 athena * [project @ 2002-06-18 15:55:57 by athena] Nothing, really M ./genfft-k7/gen_twiddle.ml -2 +2 M ./kernel/ifftw.h -18 +19 Tue Jun 18 11:19:59 EDT 2002 athena * [project @ 2002-06-18 15:19:59 by athena] !SINGLE ==> !K7_MODE (for some reason the contrapositive sounds wrong) M ./configure.ac +1 Tue Jun 18 11:07:13 EDT 2002 athena * [project @ 2002-06-18 15:07:13 by athena] Buffer is now symmetric wrt forward/backward transform M ./dft/buffered.c -12 +22 Tue Jun 18 10:33:58 EDT 2002 athena * [project @ 2002-06-18 14:33:58 by athena] Fixed applicable() in indirect.c A ./kernel/debug.c M ./dft/ct-dif.c -2 +1 M ./dft/indirect.c -3 +3 M ./dft/vrank2-transpose.c -10 +14 M ./dft/vrank3-transpose.c -10 +13 M ./kernel/Makefile.am -1 +1 M ./kernel/debug.c +39 M ./kernel/ifftw.h -1 +5 M ./kernel/print.c -4 +10 Tue Jun 18 08:41:18 EDT 2002 athena * [project @ 2002-06-18 12:41:18 by athena] Fixed attempt to free() uninitialized pointer. M ./dft/rader.c -2 +3 Tue Jun 18 08:13:55 EDT 2002 athena * [project @ 2002-06-18 12:13:55 by athena] Added reference counts for awake() M ./CONVENTIONS -1 +1 M ./TODO +5 M ./dft/rader.c -11 +3 M ./kernel/ifftw.h -4 +7 M ./kernel/plan.c -2 +15 M ./kernel/planner.c -3 +3 M ./tests/bench.c -2 +3 Mon Jun 17 20:49:05 EDT 2002 stevenj * [project @ 2002-06-18 00:49:05 by stevenj] updated comment M ./dft/rader.c -2 +2 Mon Jun 17 20:46:45 EDT 2002 stevenj * [project @ 2002-06-18 00:46:45 by stevenj] slight update M ./TODO -1 +3 Mon Jun 17 02:30:16 EDT 2002 fftw * [project @ 2002-06-17 06:30:16 by fftw] moved prime-number stuff into primes.c, so it can be shared with generic codelet and with rfftw rader A ./kernel/primes.c M ./dft/rader.c -104 +5 M ./kernel/Makefile.am -3 +3 M ./kernel/ifftw.h -1 +24 M ./kernel/primes.c +112 Mon Jun 17 02:01:58 EDT 2002 fftw * [project @ 2002-06-17 06:01:58 by fftw] added comment M ./dft/rader.c -6 +9 Mon Jun 17 01:39:55 EDT 2002 fftw * [project @ 2002-06-17 05:39:55 by fftw] added rader-dit M ./dft/rader.c -66 +308 M ./kernel/ifftw.h -1 +10 M ./kernel/twiddle.c -8 +1 Sun Jun 16 23:50:16 EDT 2002 fftw * [project @ 2002-06-17 03:50:16 by fftw] added initial Rader (no DIT yet) A ./dft/rader.c M ./configure.ac +8 M ./dft/Makefile.am -1 +1 M ./dft/conf.c -1 +2 M ./dft/dft.h -1 +2 M ./dft/rader.c +457 Sun Jun 16 22:29:51 EDT 2002 fftw * [project @ 2002-06-17 02:29:51 by fftw] don't warn about long long M ./acinclude.m4 -1 +1 Sun Jun 16 21:30:42 EDT 2002 athena * [project @ 2002-06-17 01:30:42 by athena] Added k7 DIF codelets A ./dft/ct-dif-k7.c A ./dft/kdft-dif-k7.c M ./dft/Makefile.am -5 +6 M ./dft/codelet-k7.h -5 +10 M ./dft/ct-dif-k7.c +143 M ./dft/ct.h -1 +2 M ./dft/kdft-dif-k7.c +30 M ./genfft-k7/gen_twiddle.ml -4 +25 M ./kernel/planner.c -3 +2 M ./tests/bench.c +1 Sun Jun 16 19:13:31 EDT 2002 athena * [project @ 2002-06-16 23:13:31 by athena] Added stuff to do M ./TODO +6 Sun Jun 16 19:05:58 EDT 2002 athena * [project @ 2002-06-16 23:05:58 by athena] Handle dual case R = I + 1 M ./dft/rank0.c -6 +30 Sun Jun 16 18:54:31 EDT 2002 athena * [project @ 2002-06-16 22:54:31 by athena] Removed useless flag M ./bootstrap.sh -1 +1 Sun Jun 16 18:30:32 EDT 2002 athena * [project @ 2002-06-16 22:30:32 by athena] Removed useless file R ./mkdist.sh Sun Jun 16 18:30:18 EDT 2002 athena * [project @ 2002-06-16 22:30:18 by athena] More k7 work. Switched to runtime CLASSIC mode. A ./dft/ct-dit-k7.c A ./dft/kdft-dit-k7.c R ./CLASSIC-MODE M ./Makefile.am -10 +1 M ./configure.ac -21 M ./dft/Makefile.am -5 +5 M ./dft/buffered.c -2 +4 M ./dft/codelet-k7.h -5 +7 M ./dft/codelet.h -1 +2 M ./dft/codelets/inplace/Makefile.am -2 M ./dft/ct-dif.c -3 +4 M ./dft/ct-dit-k7.c +145 M ./dft/ct-dit.c -6 +9 M ./dft/ct-ditbuf.c -9 +11 M ./dft/ct-ditf.c -2 +3 M ./dft/ct.h -5 +3 M ./dft/direct-k7.c -2 +3 M ./dft/direct.c -2 +3 M ./dft/indirect.c -2 +3 M ./dft/kdft-dit-k7.c +30 M ./dft/kdft-dit.c -4 +2 M ./dft/nop.c -2 +3 M ./dft/rank-geq2.c -8 +8 M ./dft/rank0.c -2 +3 M ./dft/vrank-geq1.c -6 +7 M ./dft/vrank2-transpose.c -3 +3 M ./dft/vrank3-transpose.c -3 +3 M ./genfft/gen_twiddle.ml -4 +4 M ./genfft/gen_twidsq.ml -4 +4 M ./genfft/twiddle.ml -7 +28 M ./genfft/twiddle.mli -2 +5 M ./genfft-k7/Makefile.am -25 +24 M ./genfft-k7/genUtil.ml -2 +3 M ./genfft-k7/gen_notw.ml -1 +2 M ./genfft-k7/gen_twiddle.ml -23 +29 M ./genfft-k7/twiddle.ml -7 +36 M ./genfft-k7/twiddle.mli -3 +7 M ./kernel/ifftw.h -15 +13 M ./kernel/planner-naive.c -3 +3 M ./kernel/planner-score.c -5 +6 M ./kernel/planner.c -6 +5 M ./kernel/version.c -7 +2 M ./support/Makefile.codelets +1 M ./tests/bench.c -12 +2 Sun Jun 16 17:15:18 EDT 2002 stevenj * [project @ 2002-06-16 21:15:18 by stevenj] spelling M ./kernel/tensor.c -2 +2 Sun Jun 16 15:51:44 EDT 2002 athena * [project @ 2002-06-16 19:51:44 by athena] Do not compile if not K7_MODE M ./dft/kdft-k7.c -1 +3 Sun Jun 16 15:35:02 EDT 2002 athena * [project @ 2002-06-16 19:35:02 by athena] Do not require K7 definitions to compile M ./dft/codelet-k7.h -1 +2 M ./dft/dft.h -2 +1 Sun Jun 16 08:05:17 EDT 2002 athena * [project @ 2002-06-16 12:05:17 by athena] More k7 stuff A ./dft/codelet-k7.h A ./genfft-k7/gen_twiddle.ml M ./dft/Makefile.am -4 +4 M ./dft/codelet-k7.h +40 M ./dft/codelet.h -20 +5 M ./dft/direct-k7.c -1 +5 M ./genfft-k7/Makefile.am -1 +9 M ./genfft-k7/complex.ml -1 +2 M ./genfft-k7/complex.mli -1 +2 M ./genfft-k7/genUtil.ml -1 +12 M ./genfft-k7/gen_notw.ml -5 +1 M ./genfft-k7/gen_twiddle.ml +149 Sat Jun 15 18:30:43 EDT 2002 athena * [project @ 2002-06-15 22:30:43 by athena] Try to be compatible with automake-1.6 M ./acinclude.m4 -1 +5 Sat Jun 15 18:23:40 EDT 2002 athena * [project @ 2002-06-15 22:23:40 by athena] More merging of Stefan's generator with main genfft branch A ./genfft-k7/algsimp.ml A ./genfft-k7/algsimp.mli A ./genfft-k7/assoctable.ml A ./genfft-k7/assoctable.mli A ./genfft-k7/littlesimp.ml A ./genfft-k7/littlesimp.mli A ./genfft-k7/monads.ml A ./genfft-k7/oracle.ml A ./genfft-k7/oracle.mli A ./genfft-k7/to_alist.ml A ./genfft-k7/to_alist.mli A ./genfft-k7/twiddle.mli A ./support/codelet_asmprelude M ./acinclude.m4 +8 M ./configure.ac +1 M ./genfft/number.ml -7 +8 M ./genfft-k7/Makefile.am -23 +27 M ./genfft-k7/algsimp.ml +517 M ./genfft-k7/algsimp.mli +24 M ./genfft-k7/assignmentsToVfpinstrs.ml -11 +10 M ./genfft-k7/assoctable.ml +66 M ./genfft-k7/assoctable.mli +30 M ./genfft-k7/complex.ml -4 +19 M ./genfft-k7/complex.mli -5 +6 M ./genfft-k7/expr.ml -27 +58 M ./genfft-k7/expr.mli -4 +11 R ./genfft-k7/exprdag.ml R ./genfft-k7/exprdag.mli M ./genfft-k7/genUtil.ml -13 +18 M ./genfft-k7/gen_notw.ml -16 +31 M ./genfft-k7/k7Unparsing.ml -2 +2 M ./genfft-k7/littlesimp.ml +72 M ./genfft-k7/littlesimp.mli +26 M ./genfft-k7/magic.ml -16 +41 M ./genfft-k7/monads.ml +76 M ./genfft-k7/number.ml -9 +5 M ./genfft-k7/number.mli -8 +4 M ./genfft-k7/oracle.ml +138 M ./genfft-k7/oracle.mli +25 M ./genfft-k7/to_alist.ml +258 M ./genfft-k7/to_alist.mli +24 M ./genfft-k7/twiddle.ml -5 +22 M ./genfft-k7/twiddle.mli +29 M ./genfft-k7/util.ml -4 +4 M ./genfft-k7/util.mli -2 +1 M ./genfft-k7/vFpUnparsing.ml -1 +1 M ./genfft-k7/vSimdBasics.ml -2 +3 M ./genfft-k7/vSimdUnparsing.ml -1 +1 M ./genfft-k7/variable.ml +1 M ./genfft-k7/variable.mli +1 M ./support/Makefile.am -1 +1 M ./support/Makefile.codelets -1 +13 M ./support/codelet_asmprelude +8 Sat Jun 15 13:51:39 EDT 2002 athena * [project @ 2002-06-15 17:51:39 by athena] Slowly merging genfft-k7 with main genfft branch A ./genfft-k7/gen_notw.ml M ./genfft/expr.ml -1 +3 M ./genfft/expr.mli -1 +2 M ./genfft/genutil.ml -1 +9 M ./genfft-k7/Makefile.am -26 +23 M ./genfft-k7/complex.ml -34 +92 M ./genfft-k7/complex.mli -1 +11 M ./genfft-k7/expr.ml -1 +19 M ./genfft-k7/expr.mli -4 +3 M ./genfft-k7/exprdag.ml -73 +103 M ./genfft-k7/exprdag.mli -1 M ./genfft-k7/fft.ml -167 +83 M ./genfft-k7/fft.mli -29 +1 M ./genfft-k7/genUtil.ml -56 +33 R ./genfft-k7/gen_hc2hc.ml R ./genfft-k7/gen_hc2real.ml M ./genfft-k7/gen_notw.ml +143 R ./genfft-k7/gen_notwiddle.ml R ./genfft-k7/gen_notwiddle_fixedstride.ml R ./genfft-k7/gen_real2hc.ml R ./genfft-k7/gen_realeven.ml R ./genfft-k7/gen_realeven2.ml R ./genfft-k7/gen_realodd.ml R ./genfft-k7/gen_realodd2.ml R ./genfft-k7/gen_twiddle.ml M ./genfft-k7/magic.ml -11 +35 R ./genfft-k7/symmetry.ml M ./genfft-k7/twiddle.ml -63 +95 M ./genfft-k7/util.ml +15 M ./genfft-k7/util.mli +3 M ./genfft-k7/variable.ml +12 M ./genfft-k7/variable.mli +3 M ./support/Makefile.codelets -1 +1 Fri Jun 14 21:33:02 EDT 2002 athena * [project @ 2002-06-15 01:33:02 by athena] Fixed, really M ./genfft-k7/Makefile.am -1 +1 M ./genfft-k7/genUtil.ml -25 M ./genfft-k7/magic.ml -1 +1 R ./genfft-k7/magic.mli M ./genfft-k7/twiddle.ml -2 +2 M ./support/Makefile.codelets -4 +8 Fri Jun 14 21:27:12 EDT 2002 athena * [project @ 2002-06-15 01:27:12 by athena] Oops... M ./support/Makefile.codelets -1 +1 Fri Jun 14 21:25:34 EDT 2002 athena * [project @ 2002-06-15 01:25:34 by athena] Work properly when $(ALL_CODELETS) = "" M ./support/Makefile.codelets -4 +4 Fri Jun 14 21:11:16 EDT 2002 athena * [project @ 2002-06-15 01:11:16 by athena] Fixed k7 build machinery M ./Makefile.am -1 +2 M ./configure.ac +8 M ./dft/codelet.h -1 +4 M ./dft/codelets/Makefile.am -1 +1 M ./dft/conf.c -1 +4 M ./genfft-k7/gen_notwiddle.ml -1 +3 M ./kernel/ifftw.h -1 +10 M ./support/Makefile.codelets +2 Fri Jun 14 17:42:35 EDT 2002 athena * [project @ 2002-06-14 21:42:35 by athena] More work on k7 stuff M ./Makefile.am -1 +1 M ./configure.ac +1 M ./dft/codelet.h -1 +4 M ./dft/direct-k7.c -3 +3 M ./genfft-k7/Makefile.am -24 +24 R ./genfft-k7/codeletMisc.ml R ./genfft-k7/codeletMisc.mli M ./genfft-k7/genUtil.ml -10 +128 R ./genfft-k7/genUtil.mli R ./genfft-k7/gen_hc2hc.mli R ./genfft-k7/gen_hc2real.mli M ./genfft-k7/gen_notwiddle.ml -3 +33 R ./genfft-k7/gen_notwiddle.mli R ./genfft-k7/gen_real2hc.mli R ./genfft-k7/gen_realeven.mli R ./genfft-k7/gen_realeven2.mli R ./genfft-k7/gen_realodd.mli R ./genfft-k7/gen_realodd2.mli R ./genfft-k7/gen_twiddle.mli R ./genfft-k7/genfft.ml M ./genfft-k7/k7Basics.ml +19 M ./genfft-k7/k7Basics.mli +2 M ./genfft-k7/k7Unparsing.ml -1 +1 M ./genfft-k7/magic.ml -1 M ./genfft-k7/magic.mli -1 M ./kernel/ifftw.h -1 +4 M ./libbench/bench-user.h -1 +3 M ./support/Makefile.am -1 +2 Fri Jun 14 15:54:29 EDT 2002 athena * [project @ 2002-06-14 19:54:29 by athena] More work on k7 stuff A ./dft/direct-k7.c A ./dft/kdft-k7.c M ./dft/Makefile.am -1 +2 M ./dft/codelet.h -1 +13 M ./dft/conf.c -9 +15 M ./dft/dft.h -1 +2 M ./dft/direct-k7.c +167 M ./dft/kdft-k7.c +29 M ./genfft-k7/codeletMisc.ml -91 M ./genfft-k7/codeletMisc.mli -13 +1 M ./genfft-k7/genUtil.ml -24 +19 M ./genfft-k7/gen_notwiddle.ml -2 +44 M ./genfft-k7/genfft.ml +4 M ./genfft-k7/magic.ml +1 M ./genfft-k7/magic.mli +1 M ./support/Makefile.codelets -1 +1 Fri Jun 14 14:18:15 EDT 2002 athena * [project @ 2002-06-14 18:18:15 by athena] Changed my mind again M ./dft/codelet.h -4 +1 M ./dft/direct.c -7 +4 M ./genfft/gen_notw.ml -5 +4 Fri Jun 14 11:53:09 EDT 2002 athena * [project @ 2002-06-14 15:53:09 by athena] Removed some useless stuff. M ./genfft-k7/gen_notwiddle.ml -8 +3 Fri Jun 14 11:01:39 EDT 2002 athena * [project @ 2002-06-14 15:01:39 by athena] Hmm... M ./genfft-k7/gen_notwiddle.ml -3 +8 Fri Jun 14 10:28:12 EDT 2002 athena * [project @ 2002-06-14 14:28:12 by athena] More work in preparation for k7 stuff M ./dft/codelet.h -1 +2 M ./dft/direct.c -3 +4 M ./genfft/gen_notw.ml -4 +5 M ./genfft-k7/gen_notwiddle.ml -18 +13 Fri Jun 14 07:25:28 EDT 2002 athena * [project @ 2002-06-14 11:25:28 by athena] Still preparing to include k7 stuff M ./TODO +4 M ./dft/codelet.h -4 +6 M ./dft/direct.c -1 +5 M ./genfft/gen_notw.ml -7 +9 Fri Jun 14 07:06:02 EDT 2002 athena * [project @ 2002-06-14 11:06:02 by athena] Create .depend M ./bootstrap.sh -2 +2 Fri Jun 14 06:56:15 EDT 2002 athena * [project @ 2002-06-14 10:56:14 by athena] Imported Stefan's K7 generator A ./genfft-k7/ A ./genfft-k7/Makefile.am A ./genfft-k7/assignmentsToVfpinstrs.ml A ./genfft-k7/assignmentsToVfpinstrs.mli A ./genfft-k7/balanceVfpinstrs.ml A ./genfft-k7/balanceVfpinstrs.mli A ./genfft-k7/codeletMisc.ml A ./genfft-k7/codeletMisc.mli A ./genfft-k7/complex.ml A ./genfft-k7/complex.mli A ./genfft-k7/expr.ml A ./genfft-k7/expr.mli A ./genfft-k7/exprdag.ml A ./genfft-k7/exprdag.mli A ./genfft-k7/fft.ml A ./genfft-k7/fft.mli A ./genfft-k7/genUtil.ml A ./genfft-k7/genUtil.mli A ./genfft-k7/gen_hc2hc.ml A ./genfft-k7/gen_hc2hc.mli A ./genfft-k7/gen_hc2real.ml A ./genfft-k7/gen_hc2real.mli A ./genfft-k7/gen_notwiddle.ml A ./genfft-k7/gen_notwiddle.mli A ./genfft-k7/gen_notwiddle_fixedstride.ml A ./genfft-k7/gen_real2hc.ml A ./genfft-k7/gen_real2hc.mli A ./genfft-k7/gen_realeven.ml A ./genfft-k7/gen_realeven.mli A ./genfft-k7/gen_realeven2.ml A ./genfft-k7/gen_realeven2.mli A ./genfft-k7/gen_realodd.ml A ./genfft-k7/gen_realodd.mli A ./genfft-k7/gen_realodd2.ml A ./genfft-k7/gen_realodd2.mli A ./genfft-k7/gen_twiddle.ml A ./genfft-k7/gen_twiddle.mli A ./genfft-k7/genfft.ml A ./genfft-k7/id.ml A ./genfft-k7/id.mli A ./genfft-k7/k7Basics.ml A ./genfft-k7/k7Basics.mli A ./genfft-k7/k7ExecutionModel.ml A ./genfft-k7/k7ExecutionModel.mli A ./genfft-k7/k7FlatInstructionScheduling.ml A ./genfft-k7/k7FlatInstructionScheduling.mli A ./genfft-k7/k7InstructionSchedulingBasics.ml A ./genfft-k7/k7InstructionSchedulingBasics.mli A ./genfft-k7/k7RegisterAllocationBasics.ml A ./genfft-k7/k7RegisterAllocationBasics.mli A ./genfft-k7/k7RegisterAllocator.ml A ./genfft-k7/k7RegisterAllocator.mli A ./genfft-k7/k7RegisterAllocatorEATranslation.ml A ./genfft-k7/k7RegisterAllocatorEATranslation.mli A ./genfft-k7/k7RegisterAllocatorInit.ml A ./genfft-k7/k7RegisterAllocatorInit.mli A ./genfft-k7/k7RegisterReallocation.ml A ./genfft-k7/k7RegisterReallocation.mli A ./genfft-k7/k7Translate.ml A ./genfft-k7/k7Translate.mli A ./genfft-k7/k7Unparsing.ml A ./genfft-k7/k7Unparsing.mli A ./genfft-k7/k7Vectorization.ml A ./genfft-k7/k7Vectorization.mli A ./genfft-k7/magic.ml A ./genfft-k7/magic.mli A ./genfft-k7/memoMonad.ml A ./genfft-k7/memoMonad.mli A ./genfft-k7/nonDetMonad.ml A ./genfft-k7/nonDetMonad.mli A ./genfft-k7/nullVectorization.ml A ./genfft-k7/nullVectorization.mli A ./genfft-k7/number.ml A ./genfft-k7/number.mli A ./genfft-k7/stateMonad.ml A ./genfft-k7/stateMonad.mli A ./genfft-k7/symmetry.ml A ./genfft-k7/twiddle.ml A ./genfft-k7/util.ml A ./genfft-k7/util.mli A ./genfft-k7/vAnnotatedScheduler.ml A ./genfft-k7/vAnnotatedScheduler.mli A ./genfft-k7/vDag.ml A ./genfft-k7/vDag.mli A ./genfft-k7/vFpBasics.ml A ./genfft-k7/vFpBasics.mli A ./genfft-k7/vFpUnparsing.ml A ./genfft-k7/vFpUnparsing.mli A ./genfft-k7/vImproveSchedule.ml A ./genfft-k7/vImproveSchedule.mli A ./genfft-k7/vK7Optimization.ml A ./genfft-k7/vK7Optimization.mli A ./genfft-k7/vScheduler.ml A ./genfft-k7/vScheduler.mli A ./genfft-k7/vSimdBasics.ml A ./genfft-k7/vSimdBasics.mli A ./genfft-k7/vSimdIndexing.ml A ./genfft-k7/vSimdIndexing.mli A ./genfft-k7/vSimdUnparsing.ml A ./genfft-k7/vSimdUnparsing.mli A ./genfft-k7/variable.ml A ./genfft-k7/variable.mli M ./AUTHORS -2 +5 M ./Makefile.am -1 +1 M ./bootstrap.sh +2 M ./configure.ac +1 M ./genfft-k7/Makefile.am +108 M ./genfft-k7/assignmentsToVfpinstrs.ml +225 M ./genfft-k7/assignmentsToVfpinstrs.mli +38 M ./genfft-k7/balanceVfpinstrs.ml +195 M ./genfft-k7/balanceVfpinstrs.mli +29 M ./genfft-k7/codeletMisc.ml +127 M ./genfft-k7/codeletMisc.mli +45 M ./genfft-k7/complex.ml +142 M ./genfft-k7/complex.mli +44 M ./genfft-k7/expr.ml +46 M ./genfft-k7/expr.mli -1 +34 M ./genfft-k7/exprdag.ml +879 M ./genfft-k7/exprdag.mli +42 M ./genfft-k7/fft.ml +311 M ./genfft-k7/fft.mli +50 M ./genfft-k7/genUtil.ml +377 M ./genfft-k7/genUtil.mli +41 M ./genfft-k7/gen_hc2hc.ml +133 M ./genfft-k7/gen_hc2hc.mli +25 M ./genfft-k7/gen_hc2real.ml +77 M ./genfft-k7/gen_hc2real.mli +24 M ./genfft-k7/gen_notwiddle.ml +69 M ./genfft-k7/gen_notwiddle.mli +25 M ./genfft-k7/gen_notwiddle_fixedstride.ml +45 M ./genfft-k7/gen_real2hc.ml +72 M ./genfft-k7/gen_real2hc.mli +24 M ./genfft-k7/gen_realeven.ml +55 M ./genfft-k7/gen_realeven.mli +24 M ./genfft-k7/gen_realeven2.ml +57 M ./genfft-k7/gen_realeven2.mli +24 M ./genfft-k7/gen_realodd.ml +60 M ./genfft-k7/gen_realodd.mli +24 M ./genfft-k7/gen_realodd2.ml +60 M ./genfft-k7/gen_realodd2.mli +24 M ./genfft-k7/gen_twiddle.ml +98 M ./genfft-k7/gen_twiddle.mli +25 M ./genfft-k7/genfft.ml +282 M ./genfft-k7/id.ml +61 M ./genfft-k7/id.mli +79 M ./genfft-k7/k7Basics.ml +524 M ./genfft-k7/k7Basics.mli +219 M ./genfft-k7/k7ExecutionModel.ml +111 M ./genfft-k7/k7ExecutionModel.mli +24 M ./genfft-k7/k7FlatInstructionScheduling.ml +177 M ./genfft-k7/k7FlatInstructionScheduling.mli +22 M ./genfft-k7/k7InstructionSchedulingBasics.ml +205 M ./genfft-k7/k7InstructionSchedulingBasics.mli +74 M ./genfft-k7/k7RegisterAllocationBasics.ml +162 M ./genfft-k7/k7RegisterAllocationBasics.mli +88 M ./genfft-k7/k7RegisterAllocator.ml +516 M ./genfft-k7/k7RegisterAllocator.mli +23 M ./genfft-k7/k7RegisterAllocatorEATranslation.ml +134 M ./genfft-k7/k7RegisterAllocatorEATranslation.mli +27 M ./genfft-k7/k7RegisterAllocatorInit.ml +97 M ./genfft-k7/k7RegisterAllocatorInit.mli +30 M ./genfft-k7/k7RegisterReallocation.ml +259 M ./genfft-k7/k7RegisterReallocation.mli +22 M ./genfft-k7/k7Translate.ml +142 M ./genfft-k7/k7Translate.mli +87 M ./genfft-k7/k7Unparsing.ml +326 M ./genfft-k7/k7Unparsing.mli +53 M ./genfft-k7/k7Vectorization.ml +739 M ./genfft-k7/k7Vectorization.mli +25 M ./genfft-k7/magic.ml +80 M ./genfft-k7/magic.mli +55 M ./genfft-k7/memoMonad.ml +32 M ./genfft-k7/memoMonad.mli +24 M ./genfft-k7/nonDetMonad.ml +136 M ./genfft-k7/nonDetMonad.mli +94 M ./genfft-k7/nullVectorization.ml +78 M ./genfft-k7/nullVectorization.mli +24 M ./genfft-k7/number.ml +158 M ./genfft-k7/number.mli +53 M ./genfft-k7/stateMonad.ml +71 M ./genfft-k7/stateMonad.mli +41 M ./genfft-k7/symmetry.ml +314 M ./genfft-k7/twiddle.ml +116 M ./genfft-k7/util.ml -1 +322 M ./genfft-k7/util.mli +124 M ./genfft-k7/vAnnotatedScheduler.ml +182 M ./genfft-k7/vAnnotatedScheduler.mli +39 M ./genfft-k7/vDag.ml +123 M ./genfft-k7/vDag.mli +52 M ./genfft-k7/vFpBasics.ml +203 M ./genfft-k7/vFpBasics.mli +106 M ./genfft-k7/vFpUnparsing.ml +79 M ./genfft-k7/vFpUnparsing.mli +28 M ./genfft-k7/vImproveSchedule.ml +155 M ./genfft-k7/vImproveSchedule.mli +22 M ./genfft-k7/vK7Optimization.ml +626 M ./genfft-k7/vK7Optimization.mli +22 M ./genfft-k7/vScheduler.ml +166 M ./genfft-k7/vScheduler.mli +31 M ./genfft-k7/vSimdBasics.ml +296 M ./genfft-k7/vSimdBasics.mli +247 M ./genfft-k7/vSimdIndexing.ml +119 M ./genfft-k7/vSimdIndexing.mli +68 M ./genfft-k7/vSimdUnparsing.ml +104 M ./genfft-k7/vSimdUnparsing.mli +33 M ./genfft-k7/variable.ml -1 +119 M ./genfft-k7/variable.mli +53 Thu Jun 13 15:30:41 EDT 2002 athena * [project @ 2002-06-13 19:30:41 by athena] Generator for real->halfcomplex and halfcomplex->real codelets A ./genfft/gen_hc2r.ml A ./genfft/gen_r2hc.ml M ./genfft/Makefile.am -9 +27 M ./genfft/c.ml -2 +2 M ./genfft/complex.ml -1 +8 M ./genfft/complex.mli -1 +3 M ./genfft/gen_hc2r.ml +121 M ./genfft/gen_r2hc.ml +121 M ./genfft/genutil.ml -8 +9 M ./genfft/trig.ml -2 +2 Thu Jun 13 11:54:02 EDT 2002 athena * [project @ 2002-06-13 15:54:02 by athena] Improved hash functions, printers M ./dft/problem.c -5 +10 M ./kernel/planner.c -11 +5 M ./kernel/tensor.c -7 +6 M ./tests/bench.c -1 +1 Thu Jun 13 11:17:31 EDT 2002 athena * [project @ 2002-06-13 15:17:31 by athena] Only regenerate codlist.c in maintainer mode M ./support/Makefile.codelets -18 +18 Thu Jun 13 11:04:24 EDT 2002 athena * [project @ 2002-06-13 15:04:24 by athena] Planner can export solution list M ./dft/problem.c -10 +21 M ./dft/rank-geq2.c -2 +2 M ./kernel/ifftw.h -11 +13 M ./kernel/planner-naive.c -7 +7 M ./kernel/planner-score.c -8 +7 M ./kernel/planner.c -42 +47 M ./kernel/print.c -9 +27 M ./kernel/tensor.c -4 +4 M ./tests/bench.c -5 +6 Thu Jun 13 08:59:53 EDT 2002 athena * [project @ 2002-06-13 12:59:53 by athena] Fixed for intel compiler M ./dft/ct-ditbuf.c -2 +3 M ./dft/dft.h -2 +1 M ./dft/direct.c -3 +1 M ./kernel/cycle.h -2 +2 M ./libbench/bench-user.h -12 +1 Thu Jun 13 08:48:51 EDT 2002 athena * [project @ 2002-06-13 12:48:51 by athena] Revised strategy for constants in codelets M ./dft/codelet.h -13 +2 M ./genfft/c.ml -50 +43 M ./genfft/c.mli -1 +4 M ./genfft/gen_notw.ml -5 +5 M ./genfft/gen_trig.ml -3 +3 M ./genfft/gen_twiddle.ml -3 +3 M ./genfft/gen_twidsq.ml -3 +3 M ./genfft/magic.ml -5 +1 Thu Jun 13 06:21:31 EDT 2002 athena * [project @ 2002-06-13 10:21:31 by athena] Enable score planner in classic mode, naive planner in pro mode. M ./tests/bench.c -3 +3 Wed Jun 12 19:18:18 EDT 2002 athena * [project @ 2002-06-12 23:18:18 by athena] Report classic/pro M ./tests/bench.c -2 +7 Wed Jun 12 19:07:48 EDT 2002 athena * [project @ 2002-06-12 23:07:48 by athena] Fixed behavior of buffered solver for large buffers. M ./dft/buffered.c -5 +5 M ./tests/bench.c -1 +1 Wed Jun 12 18:57:19 EDT 2002 athena * [project @ 2002-06-12 22:57:19 by athena] Make assumption COST(vector) = length * COST(scalar) in classic mode. M ./dft/rank-geq2.c -2 +2 M ./dft/vrank-geq1.c -1 +2 M ./kernel/ifftw.h -4 +7 M ./kernel/planner-naive.c -4 +2 M ./kernel/planner-score.c -5 +3 M ./kernel/planner.c -13 +20 M ./kernel/timer.c -8 +2 M ./libbench/timer.c -2 +2 M ./tests/bench.c -1 +1 Wed Jun 12 18:19:48 EDT 2002 athena * [project @ 2002-06-12 22:19:48 by athena] Revised planner implementation in preparation for wisdom. M ./kernel/ifftw.h -5 +6 M ./kernel/plan.c -1 +2 M ./kernel/planner-naive.c -3 +7 M ./kernel/planner-score.c -3 +7 M ./kernel/planner.c -46 +81 M ./support/Makefile.codelets -19 +19 Wed Jun 12 08:27:36 EDT 2002 athena * [project @ 2002-06-12 12:27:36 by athena] Manually hoist loop invariants. M ./dft/ct-ditbuf.c -5 +11 Wed Jun 12 07:47:41 EDT 2002 athena * [project @ 2002-06-12 11:47:41 by athena] Revised loop to compile better with gcc -O M ./dft/rank-geq2.c -16 +8 M ./dft/rank0.c -23 +22 M ./dft/vrank-geq1.c -2 +2 Tue Jun 11 17:24:09 EDT 2002 athena * [project @ 2002-06-11 21:24:09 by athena] Changed tensor syntax M ./kernel/tensor.c -4 +5 Tue Jun 11 16:39:45 EDT 2002 athena * [project @ 2002-06-11 20:39:45 by athena] Added stuff to do. M ./TODO +15 Tue Jun 11 16:28:14 EDT 2002 athena * [project @ 2002-06-11 20:28:14 by athena] Report classic/pro in version number M ./kernel/version.c -2 +7 Tue Jun 11 14:22:49 EDT 2002 athena * [project @ 2002-06-11 18:22:49 by athena] Renamed versions into classic/pro A ./CLASSIC-MODE M ./Makefile.am -7 +7 R ./RESEARCH-MODE M ./bootstrap.sh -1 +1 M ./configure.ac -13 +13 M ./dft/codelets/inplace/Makefile.am -1 +1 M ./dft/ct-dit.c -2 +2 M ./dft/ct-ditbuf.c -2 +2 M ./dft/kdft-dit.c -2 +2 M ./dft/rank-geq2.c -4 +4 M ./dft/vrank-geq1.c -4 +4 M ./kernel/ifftw.h -3 +3 M ./mkdist.sh -2 +2 M ./tests/bench.c -1 +1 Tue Jun 11 14:06:06 EDT 2002 athena * [project @ 2002-06-11 18:06:06 by athena] Revised planners, estimator M ./kernel/Makefile.am -4 +4 M ./kernel/ifftw.h -6 +7 R ./kernel/planner-estimate.c M ./kernel/planner-naive.c -4 +4 M ./kernel/planner-score.c -4 +4 M ./kernel/planner.c -2 +17 M ./tests/bench.c -5 +5 Tue Jun 11 11:45:41 EDT 2002 athena * [project @ 2002-06-11 15:45:41 by athena] I don't know what I am doing. M ./Makefile.am -2 +2 M ./dft/buffered.c -3 +2 M ./dft/ct-dit.c -4 +5 M ./dft/ct-ditbuf.c -4 +6 M ./dft/kdft-dif.c -3 +2 M ./dft/kdft-difsq.c -3 +2 M ./dft/kdft-dit.c -7 +4 M ./kernel/ifftw.h -3 +2 Tue Jun 11 10:35:52 EDT 2002 athena * [project @ 2002-06-11 14:35:52 by athena] Massive revision of estimator A ./kernel/ops.c M ./Makefile.am -1 +1 M ./dft/buffered.c -8 +7 M ./dft/codelet.h -3 +3 M ./dft/ct-dif.c -5 +4 M ./dft/ct-dit.c -5 +7 M ./dft/ct-ditbuf.c -4 +9 M ./dft/ct-ditf.c -4 +3 M ./dft/ct.c -6 +1 M ./dft/dft.h -1 +3 M ./dft/direct.c -3 +2 M ./dft/indirect.c -3 +2 M ./dft/nop.c -1 +3 M ./dft/rank-geq2.c -3 +7 M ./dft/rank0.c -3 +4 M ./dft/vrank-geq1.c -10 +18 M ./dft/vrank2-transpose.c -3 +3 M ./dft/vrank3-transpose.c -3 +4 M ./genfft/c.ml -2 +2 M ./kernel/Makefile.am -1 +1 M ./kernel/cycle.h -3 +3 R ./kernel/flops.c M ./kernel/ifftw.h -13 +21 M ./kernel/ops.c +58 M ./kernel/plan.c -2 +2 M ./kernel/planner-estimate.c -2 +9 M ./kernel/planner-naive.c -3 +3 M ./kernel/planner-score.c -4 +4 M ./tests/bench.c -1 +3 Tue Jun 11 07:32:20 EDT 2002 athena * [project @ 2002-06-11 11:32:20 by athena] Many changes A ./dft/vrank-geq1.c M ./dft/Makefile.am -1 +1 M ./dft/buffered.c -9 +8 M ./dft/codelets/inplace/Makefile.am -2 +3 M ./dft/codelets/standard/Makefile.am -2 M ./dft/conf.c -2 +2 M ./dft/ct-dit.c -2 +5 M ./dft/ct.c -2 +2 M ./dft/dft.h -6 +4 M ./dft/indirect.c -2 +2 M ./dft/problem.c -1 +13 M ./dft/rank-geq2.c -5 +4 R ./dft/vecloop.c M ./dft/vrank-geq1.c +243 M ./kernel/ifftw.h -2 +6 M ./kernel/planner-estimate.c -2 +2 M ./kernel/planner-naive.c -2 +2 M ./kernel/planner-score.c -4 +17 M ./kernel/planner.c -4 +6 M ./kernel/print.c -6 +27 M ./kernel/tensor.c -1 +15 M ./kernel/timer.c -2 +2 M ./libbench/bench-main.c -3 +6 M ./tests/bench.c -15 +15 Mon Jun 10 21:35:29 EDT 2002 athena * [project @ 2002-06-11 01:35:29 by athena] Keep it simple, stupid. M ./dft/ct-ditbuf.c -21 +4 Mon Jun 10 19:24:28 EDT 2002 athena * [project @ 2002-06-10 23:24:28 by athena] Fixed when #undef PRECOMPUTE_ARRAY_INDICES M ./kernel/ifftw.h -3 +5 Mon Jun 10 17:58:13 EDT 2002 athena * [project @ 2002-06-10 21:58:13 by athena] Minor changes M ./dft/vrank3-transpose.c -5 +3 M ./kernel/print.c -2 +2 Mon Jun 10 16:30:37 EDT 2002 athena * [project @ 2002-06-10 20:30:37 by athena] Added ct-ditbuf.c, many changes everywhere A ./dft/ct-ditbuf.c M ./CONVENTIONS -2 +3 M ./configure.ac +2 M ./dft/Makefile.am -3 +3 M ./dft/buffered.c -11 +11 M ./dft/ct-dif.c -5 +3 M ./dft/ct-dit.c -6 +4 M ./dft/ct-ditbuf.c +187 M ./dft/ct.c -3 +3 M ./dft/ct.h -1 +3 M ./dft/dft.h -1 +2 M ./dft/direct.c -6 +6 M ./dft/indirect.c -6 +6 M ./dft/kdft-dif.c -2 +2 M ./dft/kdft-difsq.c -2 +2 M ./dft/kdft-dit.c -3 +8 M ./dft/kdft.c -2 +2 M ./dft/nop.c -2 +2 M ./dft/rank-geq2.c -6 +6 M ./dft/rank0.c -2 +2 M ./dft/vecloop.c -10 +8 M ./dft/vrank2-transpose.c -2 +2 M ./dft/vrank3-transpose.c -2 +2 M ./kernel/ifftw.h -3 +25 M ./kernel/plan.c -5 +4 M ./kernel/problem.c -2 +2 M ./kernel/timer.c -3 +3 M ./tests/bench.c -7 +8 Mon Jun 10 10:55:40 EDT 2002 athena * [project @ 2002-06-10 14:55:40 by athena] More name mangling M ./kernel/ifftw.h -2 +2 M ./kernel/planner.c -2 +2 M ./tests/bench.c -58 +47 Mon Jun 10 10:08:27 EDT 2002 athena * [project @ 2002-06-10 14:08:27 by athena] Fixed build system for single/double precision M ./Makefile.am -4 +3 M ./acinclude.m4 -13 M ./configure.ac +4 M ./tests/Makefile.am -1 +1 Mon Jun 10 09:04:21 EDT 2002 athena * [project @ 2002-06-10 13:04:21 by athena] Massive renaming to support both single and double precision. (Must recompile everything twice). A ./kernel/dfftw3.h A ./kernel/fftw3.h A ./kernel/sfftw3.h M ./CONVENTIONS -3 +9 M ./configure.ac -1 +1 M ./dft/buffered.c -96 +102 M ./dft/codelet.h -16 +16 M ./dft/codelets/inplace/Makefile.am -1 +1 M ./dft/codelets/standard/Makefile.am -1 +1 M ./dft/conf.c -12 +12 M ./dft/ct-dif.c -31 +33 M ./dft/ct-dit.c -29 +31 M ./dft/ct-ditf.c -27 +29 M ./dft/ct.c -40 +41 M ./dft/ct.h -10 +12 M ./dft/dft.h -25 +27 M ./dft/direct.c -26 +26 M ./dft/indirect.c -50 +63 M ./dft/kdft-dif.c -3 +3 M ./dft/kdft-difsq.c -3 +3 M ./dft/kdft-dit.c -3 +3 M ./dft/kdft.c -3 +3 M ./dft/nop.c -11 +11 M ./dft/plan.c -3 +3 M ./dft/problem.c -39 +40 M ./dft/rank-geq2.c -57 +59 M ./dft/rank0.c -54 +62 M ./dft/solve.c -2 +2 M ./dft/vecloop.c -49 +49 M ./dft/vrank2-transpose.c -27 +28 M ./dft/vrank3-transpose.c -44 +47 M ./genfft/gen_notw.ml -3 +3 M ./genfft/gen_twiddle.ml -4 +4 M ./genfft/gen_twidsq.ml -4 +4 M ./genfft/genutil.ml -2 +2 M ./kernel/Makefile.am -1 +2 M ./kernel/alloc.c -76 +76 M ./kernel/assert.c -2 +2 M ./kernel/awake.c -2 +2 M ./kernel/cycle.h -11 +11 M ./kernel/dfftw3.h +24 R ./kernel/fftw.h M ./kernel/fftw3.h +39 M ./kernel/flops.c -4 +6 M ./kernel/ifftw.h -90 +89 M ./kernel/minmax.c -5 +5 M ./kernel/plan.c -6 +6 M ./kernel/planner-estimate.c -6 +6 M ./kernel/planner-naive.c -7 +7 M ./kernel/planner-score.c -11 +11 M ./kernel/planner.c -73 +71 M ./kernel/print.c -12 +14 M ./kernel/problem.c -5 +5 M ./kernel/sfftw3.h +24 M ./kernel/solver.c -5 +5 M ./kernel/solvtab.c -4 +4 M ./kernel/square.c -2 +2 M ./kernel/stride.c -6 +6 M ./kernel/tensor.c -96 +96 M ./kernel/timer.c -39 +40 M ./kernel/twiddle.c -33 +38 M ./kernel/version.c -3 +4 M ./support/Makefile.codelets -3 +3 M ./tests/bench.c -39 +53 Mon Jun 10 06:49:55 EDT 2002 athena * [project @ 2002-06-10 10:49:55 by athena] Preliminary crude support for vector transforms in benchmark library. M ./libbench/allocate.c -7 +8 M ./libbench/bench-user.h -2 +5 M ./libbench/mflops.c -2 +2 M ./libbench/problem.c -2 +31 M ./tests/bench.c -1 +2 Sun Jun 9 21:11:51 EDT 2002 athena * [project @ 2002-06-10 01:11:51 by athena] Wrong cast M ./kernel/tensor.c -3 +3 Sun Jun 9 16:48:54 EDT 2002 athena * [project @ 2002-06-09 20:48:54 by athena] Added things to do. M ./TODO +1 Sun Jun 9 16:07:12 EDT 2002 athena * [project @ 2002-06-09 20:07:12 by athena] twlen0: make static M ./kernel/twiddle.c -2 +2 Sun Jun 9 15:34:54 EDT 2002 athena * [project @ 2002-06-09 19:34:54 by athena] Nothing M ./dft/buffered.c -3 +3 Sun Jun 9 15:30:13 EDT 2002 athena * [project @ 2002-06-09 19:30:13 by athena] Forgot break in switch statement. M ./kernel/print.c -1 +2 Sun Jun 9 15:27:24 EDT 2002 athena * [project @ 2002-06-09 19:27:24 by athena] Fix for c++ compatibility M ./kernel/print.c -4 +6 Sun Jun 9 15:16:43 EDT 2002 athena * [project @ 2002-06-09 19:16:43 by athena] Added printer, changed everything A ./kernel/print.c M ./TODO -4 M ./dft/buffered.c -26 +22 M ./dft/ct.c -10 +5 M ./dft/direct.c -8 +3 M ./dft/indirect.c -20 +16 M ./dft/nop.c -4 +6 M ./dft/rank-geq2.c -7 +3 M ./dft/rank0.c -6 +3 M ./dft/vecloop.c -5 +3 M ./dft/vrank2-transpose.c -3 +3 M ./dft/vrank3-transpose.c -3 +3 M ./kernel/Makefile.am -2 +2 M ./kernel/ifftw.h -4 +14 M ./kernel/planner.c -3 +10 M ./kernel/print.c +122 M ./tests/bench.c -2 +22 Sun Jun 9 11:37:07 EDT 2002 athena * [project @ 2002-06-09 15:37:07 by athena] Removed redundant nop solver M ./dft/buffered.c -8 +3 M ./dft/nop.c -2 +12 M ./dft/rank0.c -25 +3 M ./tests/bench.c -1 +1 Sun Jun 9 11:06:31 EDT 2002 athena * [project @ 2002-06-09 15:06:31 by athena] More things to do M ./TODO +10 Sun Jun 9 11:01:41 EDT 2002 athena * [project @ 2002-06-09 15:01:41 by athena] Introduced idea of rank -infinity and associated NOP plans A ./TODO A ./dft/nop.c M ./TODO +4 M ./dft/Makefile.am -4 +4 M ./dft/buffered.c -88 +66 M ./dft/conf.c -1 +2 M ./dft/dft.h -1 +2 M ./dft/direct.c -2 +3 M ./dft/indirect.c -1 +2 M ./dft/nop.c +84 M ./dft/problem.c -2 +5 M ./dft/rank0.c -3 +12 M ./dft/vecloop.c -1 +2 M ./kernel/ifftw.h -1 +11 M ./kernel/tensor.c -11 +39 Sun Jun 9 08:36:27 EDT 2002 athena * [project @ 2002-06-09 12:36:27 by athena] Fixed comment M ./dft/buffered.c -2 +2 Sun Jun 9 08:20:13 EDT 2002 athena * [project @ 2002-06-09 12:20:13 by athena] Removed useless assertions. M ./kernel/tensor.c -4 +3 Sun Jun 9 08:19:26 EDT 2002 athena * [project @ 2002-06-09 12:19:26 by athena] Don't malloc(0). M ./kernel/tensor.c -3 +7 Sun Jun 9 08:08:13 EDT 2002 athena * [project @ 2002-06-09 12:08:13 by athena] Fixed signed/unsigned puns M ./dft/buffered.c -11 +11 Sun Jun 9 07:52:22 EDT 2002 athena * [project @ 2002-06-09 11:52:22 by athena] Added buffered.c A ./dft/buffered.c M ./dft/Makefile.am -1 +1 M ./dft/buffered.c +374 M ./dft/conf.c -1 +2 M ./dft/ct-dif.c -2 +2 M ./dft/ct-dit.c -2 +2 M ./dft/ct-ditf.c -2 +2 M ./dft/dft.h -1 +2 M ./dft/direct.c -2 +2 M ./dft/indirect.c -3 +3 M ./dft/rank-geq2.c -2 +2 M ./dft/rank0.c -6 +6 M ./dft/vecloop.c -2 +2 M ./dft/vrank2-transpose.c -2 +2 M ./dft/vrank3-transpose.c -2 +2 M ./libbench/bench-main.c -2 +7 M ./libbench/bench-user.h -1 +3 M ./tests/bench.c -1 +1 Sat Jun 8 16:57:54 EDT 2002 athena * [project @ 2002-06-08 20:57:54 by athena] Fixed printout M ./dft/ct.c -6 +5 Sat Jun 8 16:42:52 EDT 2002 athena * [project @ 2002-06-08 20:42:52 by athena] Fixed comment M ./dft/vrank3-transpose.c -2 +2 Sat Jun 8 16:40:58 EDT 2002 athena * [project @ 2002-06-08 20:40:58 by athena] Added vrank3-transpose, renamed vrank0-transpose -> vrank2-transpose A ./dft/vrank2-transpose.c A ./dft/vrank3-transpose.c M ./dft/Makefile.am -2 +2 M ./dft/conf.c -2 +3 M ./dft/dft.h -1 +3 R ./dft/vrank0-transpose.c M ./dft/vrank2-transpose.c +132 M ./dft/vrank3-transpose.c +173 M ./tests/bench.c -1 +1 Sat Jun 8 15:51:46 EDT 2002 athena * [project @ 2002-06-08 19:51:46 by athena] Added vrank0-transpose A ./dft/rank-geq2.c A ./dft/vrank0-transpose.c M ./bootstrap.sh -1 +1 M ./dft/Makefile.am -1 +2 M ./dft/conf.c -1 +2 M ./dft/direct.c -4 +4 M ./dft/rank-geq2.c +257 M ./dft/rank0.c -2 +2 R ./dft/rank_geq2.c M ./dft/vrank0-transpose.c +132 M ./tests/bench.c -1 +2 Sat Jun 8 15:11:09 EDT 2002 athena * [project @ 2002-06-08 19:11:09 by athena] Added planner-score.c A ./kernel/planner-score.c M ./dft/Makefile.am -2 +2 M ./dft/ct-dif.c -2 +2 M ./dft/ct-dit.c -2 +2 M ./dft/ct-ditf.c -2 +2 M ./dft/direct.c -2 +2 M ./dft/indirect.c -2 +2 M ./dft/rank0.c -2 +2 M ./dft/rank_geq2.c -2 +2 M ./dft/vecloop.c -2 +2 M ./kernel/Makefile.am -3 +3 M ./kernel/ifftw.h -4 +5 M ./kernel/planner-score.c +70 M ./kernel/planner.c -12 +13 M ./tests/bench.c -1 +4 Sat Jun 8 11:10:44 EDT 2002 athena * [project @ 2002-06-08 15:10:44 by athena] Added indirect.c A ./dft/indirect.c M ./dft/Makefile.am -1 +1 M ./dft/conf.c -1 +2 M ./dft/ct-dif.c -2 +2 M ./dft/ct-dit.c -2 +2 M ./dft/ct-ditf.c -4 +6 M ./dft/dft.h -1 +3 M ./dft/indirect.c +227 M ./dft/rank_geq2.c -5 +5 M ./dft/vecloop.c -2 +1 Sat Jun 8 09:34:58 EDT 2002 athena * [project @ 2002-06-08 13:34:58 by athena] dif, ditf solvers A ./mkdist.sh A ./dft/ct-dif.c A ./dft/ct-ditf.c A ./dft/kdft-dif.c A ./dft/kdft-difsq.c M ./Makefile.am -4 +6 M ./dft/Makefile.am -5 +3 M ./dft/codelet.h -2 +3 M ./dft/codelets/Makefile.am -5 M ./dft/codelets/inplace/Makefile.am -2 +4 M ./dft/conf.c -2 +3 M ./dft/ct-dif.c +120 M ./dft/ct-dit.c -2 +1 M ./dft/ct-ditf.c +106 M ./dft/dft.h -1 +2 M ./dft/direct.c -2 +2 M ./dft/kdft-dif.c +29 M ./dft/kdft-difsq.c +29 M ./mkdist.sh +6 M ./tests/Makefile.am -2 +2 Fri Jun 7 18:07:53 EDT 2002 athena * [project @ 2002-06-07 22:07:53 by athena] Implemented rank_geq2. Revised build system A ./RESEARCH-MODE A ./dft/rank_geq2.c M ./Makefile.am -3 +12 M ./RESEARCH-MODE +1 M ./bootstrap.sh -1 +1 M ./configure.ac -1 +10 M ./dft/Makefile.am -1 +3 M ./dft/conf.c -1 +2 M ./dft/dft.h -1 +2 M ./dft/rank_geq2.c +257 M ./dft/vecloop.c -42 +48 M ./kernel/ifftw.h -1 +3 M ./kernel/minmax.c -1 +11 M ./kernel/planner.c -1 +2 M ./support/Makefile.codelets -1 +1 Fri Jun 7 07:12:25 EDT 2002 athena * [project @ 2002-06-07 11:12:25 by athena] Fixed printout M ./kernel/alloc.c -2 +2 Fri Jun 7 07:07:46 EDT 2002 athena * [project @ 2002-06-07 11:07:46 by athena] Added rank0. Revised codelet organization. A ./dft/codelet.h A ./dft/conf.c A ./dft/codelets/ A ./dft/codelets/Makefile.am A ./dft/codelets/inplace/ A ./dft/codelets/inplace/Makefile.am A ./dft/codelets/standard/ A ./dft/codelets/standard/Makefile.am A ./dft/rank0.c M ./Makefile.am -3 +2 M ./bootstrap.sh -5 +1 M ./configure.ac -3 +18 M ./dft/Makefile.am -3 +4 M ./dft/codelet.h +118 M ./dft/codelets/Makefile.am +7 M ./dft/codelets/inplace/Makefile.am -1 +35 M ./dft/codelets/standard/Makefile.am -1 +37 M ./dft/conf.c +30 M ./dft/dft.h -1 +5 M ./dft/rank0.c +263 M ./genfft/gen_notw.ml -3 +3 M ./tests/Makefile.am -4 +5 M ./tests/bench.c -6 +1 Thu Jun 6 18:03:17 EDT 2002 athena * [project @ 2002-06-06 22:03:17 by athena] Added memoization M ./dft/ct.c -2 +2 M ./dft/vecloop.c -2 +2 M ./genfft/trig.ml -5 +5 M ./kernel/ifftw.h -2 +13 M ./kernel/planner-estimate.c -2 +2 M ./kernel/planner-naive.c -2 +2 M ./kernel/planner.c -12 +191 M ./libbench/bench-user.h -1 +2 M ./tests/bench.c -2 +16 Thu Jun 6 08:07:33 EDT 2002 athena * [project @ 2002-06-06 12:07:33 by athena] Added vecloop A ./dft/vecloop.c M ./dft/Makefile.am -1 +1 M ./dft/dft.h -1 +3 M ./dft/direct.c -9 +7 M ./dft/vecloop.c +242 M ./kernel/alloc.c -11 +38 M ./kernel/ifftw.h -2 +4 M ./kernel/planner.c -2 +2 M ./tests/bench.c +5 Wed Jun 5 19:02:56 EDT 2002 athena * [project @ 2002-06-05 23:02:56 by athena] First DIT solver/plan A ./dft/ct-dit.c M ./dft/Makefile.am -2 +2 M ./dft/ct-dit.c +118 M ./dft/ct.c -9 +16 M ./dft/ct.h -3 +14 M ./dft/dft.h -1 +2 M ./dft/direct.c -10 +6 M ./dft/kdft-dit.c -2 +3 M ./kernel/alloc.c -1 +2 M ./kernel/twiddle.c -3 +11 Wed Jun 5 16:03:44 EDT 2002 athena * [project @ 2002-06-05 20:03:44 by athena] More work on ct A ./dft/ct.c A ./dft/ct.h M ./dft/Makefile.am -2 +2 M ./dft/ct.c +166 M ./dft/ct.h +56 M ./kernel/ifftw.h -1 +2 M ./kernel/stride.c -2 +3 M ./kernel/twiddle.c -16 +26 Wed Jun 5 11:28:09 EDT 2002 athena * [project @ 2002-06-05 15:28:09 by athena] Only use cycle counters M ./kernel/ifftw.h -2 +2 M ./kernel/planner-naive.c -2 +2 M ./kernel/timer.c -166 +18 Tue Jun 4 20:22:23 EDT 2002 athena * [project @ 2002-06-05 00:22:23 by athena] Signed/unsigned fixup M ./CONVENTIONS +2 M ./bootstrap.sh -1 +1 M ./kernel/ifftw.h -7 +6 M ./kernel/twiddle.c -28 +16 Tue Jun 4 20:03:56 EDT 2002 athena * [project @ 2002-06-05 00:03:56 by athena] New file twiddle.c A ./kernel/twiddle.c M ./kernel/Makefile.am -1 +1 M ./kernel/ifftw.h -1 +24 M ./kernel/twiddle.c +137 Tue Jun 4 17:49:39 EDT 2002 athena * [project @ 2002-06-04 21:49:39 by athena] Made tensor ranks and vector lengths unsigned. Hopefully fixed all places where it matters. A ./dft/kdft-dit.c M ./configure.ac +1 M ./dft/Makefile.am -1 +2 M ./dft/direct.c -3 +4 M ./dft/kdft-dit.c +28 M ./dft/problem.c -5 +7 M ./genfft/gen_notw.ml -4 +4 M ./genfft/gen_twiddle.ml -4 +4 M ./genfft/gen_twidsq.ml -4 +4 M ./kernel/Makefile.am -2 +2 R ./kernel/codelet.h M ./kernel/flops.c -2 +2 M ./kernel/ifftw.h -17 +20 M ./kernel/tensor.c -35 +38 M ./kernel/timer.c -3 +2 M ./tests/bench.c +1 Tue Jun 4 16:28:58 EDT 2002 athena * [project @ 2002-06-04 20:28:58 by athena] System is in working state now (but very incomplete) A ./dft/kdft.c A ./kernel/planner-estimate.c A ./kernel/solvtab.c A ./tests/ A ./tests/Makefile.am A ./tests/bench.c M ./Makefile.am -1 +1 M ./configure.ac +2 M ./dft/Makefile.am -2 +2 M ./dft/dft.h -1 +10 M ./dft/direct.c -3 +2 M ./dft/kdft.c +29 M ./kernel/Makefile.am -3 +4 M ./kernel/codelet.h -1 +3 M ./kernel/fftw.h -1 +5 M ./kernel/ifftw.h -2 +21 M ./kernel/planner-estimate.c +55 M ./kernel/planner-naive.c -9 +9 M ./kernel/solvtab.c +30 M ./libbench/Makefile.am -2 +2 M ./support/Makefile.codelets +23 M ./tests/Makefile.am -1 +7 M ./tests/bench.c +92 Mon Jun 3 18:10:12 EDT 2002 athena * [project @ 2002-06-03 22:10:12 by athena] Started implementing planners A ./kernel/planner-naive.c A ./kernel/planner.c M ./CONVENTIONS +2 M ./kernel/Makefile.am -1 +1 M ./kernel/ifftw.h -6 +55 M ./kernel/planner-naive.c +58 M ./kernel/planner.c +113 Mon Jun 3 11:44:18 EDT 2002 athena * [project @ 2002-06-03 15:44:18 by athena] Imported libbench from the new benchfft. We will use libbench for benchmarking and testing. A ./libbench/ A ./libbench/Makefile.am A ./libbench/accopy-from.c A ./libbench/accopy-to.c A ./libbench/acopy.c A ./libbench/allocate.c A ./libbench/ascale.c A ./libbench/aset.c A ./libbench/bench-main.c A ./libbench/bench-user.h A ./libbench/bench.h A ./libbench/caadd.c A ./libbench/cacopy.c A ./libbench/can-do.c A ./libbench/cascale.c A ./libbench/caset.c A ./libbench/casub.c A ./libbench/ccopy-from.c A ./libbench/ccopy-to.c A ./libbench/copy-c2c-from.c A ./libbench/copy-c2c-to.c A ./libbench/copy-c2h-1d-fftpack.c A ./libbench/copy-c2h-1d-halfcomplex.c A ./libbench/copy-c2h-1d-packed.c A ./libbench/copy-c2h-1d-unpacked-ri.c A ./libbench/copy-c2h-unpacked.c A ./libbench/copy-c2h.c A ./libbench/copy-c2r-packed.c A ./libbench/copy-c2r-unpacked.c A ./libbench/copy-c2r.c A ./libbench/copy-c2ri.c A ./libbench/copy-h2c-1d-fftpack.c A ./libbench/copy-h2c-1d-halfcomplex.c A ./libbench/copy-h2c-1d-packed.c A ./libbench/copy-h2c-1d-unpacked-ri.c A ./libbench/copy-h2c-unpacked.c A ./libbench/copy-h2c.c A ./libbench/copy-r2c-packed.c A ./libbench/copy-r2c-unpacked.c A ./libbench/copy-r2c.c A ./libbench/copy-ri2c.c A ./libbench/deallocate.c A ./libbench/getopt-utils.c A ./libbench/getopt.c A ./libbench/getopt.h A ./libbench/getopt1.c A ./libbench/info.c A ./libbench/log2.c A ./libbench/main.c A ./libbench/mflops.c A ./libbench/ovtpvt.c A ./libbench/pow2.c A ./libbench/prime.c A ./libbench/problem.c A ./libbench/report.c A ./libbench/speed.c A ./libbench/timer.c A ./libbench/unnormalize.c A ./libbench/util.c A ./libbench/verify.c A ./libbench/zero.c M ./Makefile.am -1 +1 M ./configure.ac -1 +4 M ./libbench/Makefile.am +18 M ./libbench/accopy-from.c +10 M ./libbench/accopy-to.c +10 M ./libbench/acopy.c +11 M ./libbench/allocate.c +40 M ./libbench/ascale.c +13 M ./libbench/aset.c +10 M ./libbench/bench-main.c +218 M ./libbench/bench-user.h +247 M ./libbench/bench.h +66 M ./libbench/caadd.c +15 M ./libbench/cacopy.c +11 M ./libbench/can-do.c +33 M ./libbench/cascale.c +15 M ./libbench/caset.c +10 M ./libbench/casub.c +15 M ./libbench/ccopy-from.c +21 M ./libbench/ccopy-to.c +20 M ./libbench/copy-c2c-from.c +7 M ./libbench/copy-c2c-to.c +7 M ./libbench/copy-c2h-1d-fftpack.c +29 M ./libbench/copy-c2h-1d-halfcomplex.c +29 M ./libbench/copy-c2h-1d-packed.c +40 M ./libbench/copy-c2h-1d-unpacked-ri.c +25 M ./libbench/copy-c2h-unpacked.c +32 M ./libbench/copy-c2h.c +10 M ./libbench/copy-c2r-packed.c +11 M ./libbench/copy-c2r-unpacked.c +24 M ./libbench/copy-c2r.c +6 M ./libbench/copy-c2ri.c +15 M ./libbench/copy-h2c-1d-fftpack.c +31 M ./libbench/copy-h2c-1d-halfcomplex.c +34 M ./libbench/copy-h2c-1d-packed.c +44 M ./libbench/copy-h2c-1d-unpacked-ri.c +35 M ./libbench/copy-h2c-unpacked.c +51 M ./libbench/copy-h2c.c +10 M ./libbench/copy-r2c-packed.c +13 M ./libbench/copy-r2c-unpacked.c +26 M ./libbench/copy-r2c.c +6 M ./libbench/copy-ri2c.c +15 M ./libbench/deallocate.c +14 M ./libbench/getopt-utils.c +104 M ./libbench/getopt.c +1062 M ./libbench/getopt.h +180 M ./libbench/getopt1.c +188 M ./libbench/info.c +57 M ./libbench/log2.c +13 M ./libbench/main.c +40 M ./libbench/mflops.c +22 M ./libbench/ovtpvt.c +13 M ./libbench/pow2.c +6 M ./libbench/prime.c +49 M ./libbench/problem.c +126 M ./libbench/report.c +159 M ./libbench/speed.c +73 M ./libbench/timer.c +218 M ./libbench/unnormalize.c +13 M ./libbench/util.c +189 M ./libbench/verify.c +408 M ./libbench/zero.c +37 Mon Jun 3 09:18:46 EDT 2002 athena * [project @ 2002-06-03 13:18:46 by athena] Removed useless rand.c M ./kernel/Makefile.am -2 +2 M ./kernel/ifftw.h -4 +1 R ./kernel/rand.c M ./kernel/timer.c -2 +2 Mon Jun 3 08:09:05 EDT 2002 athena * [project @ 2002-06-03 12:09:05 by athena] Added timer A ./kernel/cycle.h A ./kernel/timer.c M ./CONVENTIONS +1 M ./dft/problem.c -2 +2 M ./kernel/Makefile.am -1 +1 M ./kernel/alloc.c -2 +2 M ./kernel/cycle.h +204 M ./kernel/ifftw.h -2 +10 M ./kernel/plan.c -2 +4 M ./kernel/timer.c +293 Sun Jun 2 21:03:51 EDT 2002 athena * [project @ 2002-06-03 01:03:51 by athena] Split codelets into standard and inplace M ./configure.ac -2 +4 Sun Jun 2 19:49:03 EDT 2002 athena * [project @ 2002-06-02 23:49:03 by athena] Many many changes A ./CONVENTIONS A ./dft/direct.c A ./dft/plan.c A ./dft/solve.c A ./kernel/awake.c A ./kernel/square.c M ./CONVENTIONS +49 M ./Makefile.am -4 +4 M ./dft/Makefile.am -1 +1 M ./dft/dft.h -3 +22 M ./dft/direct.c +171 M ./dft/plan.c +33 M ./dft/problem.c -10 +10 M ./dft/solve.c +31 M ./kernel/Makefile.am -1 +1 M ./kernel/awake.c +30 M ./kernel/ifftw.h -13 +33 M ./kernel/square.c +28 Sun Jun 2 15:00:11 EDT 2002 athena * [project @ 2002-06-02 19:00:11 by athena] Fixed anachronism M ./kernel/codelet.h -4 +4 Sat Aug 12 21:43:16 EDT 2006 Unknown tagger tagged bar Sun Jun 2 14:42:32 EDT 2002 athena * [project @ 2002-06-02 18:42:32 by athena] Initial revision A ./AUTHORS A ./COPYRIGHT A ./ChangeLog A ./Makefile.am A ./NEWS A ./README A ./acinclude.m4 A ./bootstrap.sh A ./configure.ac A ./dft/ A ./dft/Makefile.am A ./dft/dft.h A ./dft/problem.c A ./genfft/ A ./genfft/Makefile.am A ./genfft/algsimp.ml A ./genfft/algsimp.mli A ./genfft/annotate.ml A ./genfft/annotate.mli A ./genfft/assoctable.ml A ./genfft/assoctable.mli A ./genfft/c.ml A ./genfft/c.mli A ./genfft/complex.ml A ./genfft/complex.mli A ./genfft/conv.ml A ./genfft/conv.mli A ./genfft/dag.ml A ./genfft/dag.mli A ./genfft/expr.ml A ./genfft/expr.mli A ./genfft/fft.ml A ./genfft/fft.mli A ./genfft/gen_athnotw.ml A ./genfft/gen_athtw.ml A ./genfft/gen_conv.ml A ./genfft/gen_notw.ml A ./genfft/gen_trig.ml A ./genfft/gen_twiddle.ml A ./genfft/gen_twidsq.ml A ./genfft/genutil.ml A ./genfft/littlesimp.ml A ./genfft/littlesimp.mli A ./genfft/magic.ml A ./genfft/monads.ml A ./genfft/number.ml A ./genfft/number.mli A ./genfft/oracle.ml A ./genfft/oracle.mli A ./genfft/schedule.ml A ./genfft/schedule.mli A ./genfft/to_alist.ml A ./genfft/to_alist.mli A ./genfft/trig.ml A ./genfft/trig.mli A ./genfft/twiddle.ml A ./genfft/twiddle.mli A ./genfft/unique.ml A ./genfft/unique.mli A ./genfft/util.ml A ./genfft/util.mli A ./genfft/variable.ml A ./genfft/variable.mli A ./kernel/ A ./kernel/Makefile.am A ./kernel/alloc.c A ./kernel/assert.c A ./kernel/codelet.h A ./kernel/fftw.h A ./kernel/flops.c A ./kernel/ifftw.h A ./kernel/minmax.c A ./kernel/plan.c A ./kernel/problem.c A ./kernel/rand.c A ./kernel/solver.c A ./kernel/stride.c A ./kernel/tensor.c A ./kernel/version.c A ./support/ A ./support/Makefile.am A ./support/Makefile.codelets A ./support/codelet_prelude M ./AUTHORS +4 M ./COPYRIGHT +19 M ./Makefile.am +21 M ./acinclude.m4 +356 M ./bootstrap.sh +26 M ./configure.ac +92 M ./dft/Makefile.am +6 M ./dft/dft.h +30 M ./dft/problem.c +121 M ./genfft/Makefile.am +136 M ./genfft/algsimp.ml +517 M ./genfft/algsimp.mli +24 M ./genfft/annotate.ml +180 M ./genfft/annotate.mli +37 M ./genfft/assoctable.ml +66 M ./genfft/assoctable.mli +30 M ./genfft/c.ml +445 M ./genfft/c.mli +65 M ./genfft/complex.ml +198 M ./genfft/complex.mli +61 M ./genfft/conv.ml +130 M ./genfft/conv.mli +23 M ./genfft/dag.ml +110 M ./genfft/dag.mli +44 M ./genfft/expr.ml +93 M ./genfft/expr.mli +38 M ./genfft/fft.ml +227 M ./genfft/fft.mli +24 M ./genfft/gen_athnotw.ml +80 M ./genfft/gen_athtw.ml +108 M ./genfft/gen_conv.ml +89 M ./genfft/gen_notw.ml +126 M ./genfft/gen_trig.ml +159 M ./genfft/gen_twiddle.ml +144 M ./genfft/gen_twidsq.ml +172 M ./genfft/genutil.ml +256 M ./genfft/littlesimp.ml +72 M ./genfft/littlesimp.mli +26 M ./genfft/magic.ml +121 M ./genfft/monads.ml +76 M ./genfft/number.ml +153 M ./genfft/number.mli +49 M ./genfft/oracle.ml +138 M ./genfft/oracle.mli +25 M ./genfft/schedule.ml +186 M ./genfft/schedule.mli +30 M ./genfft/to_alist.ml +258 M ./genfft/to_alist.mli +24 M ./genfft/trig.ml +158 M ./genfft/trig.mli +35 M ./genfft/twiddle.ml +165 M ./genfft/twiddle.mli +29 M ./genfft/unique.ml +39 M ./genfft/unique.mli +25 M ./genfft/util.ml +177 M ./genfft/util.mli +50 M ./genfft/variable.ml +99 M ./genfft/variable.mli +35 M ./kernel/Makefile.am +5 M ./kernel/alloc.c +217 M ./kernel/assert.c +32 M ./kernel/codelet.h +126 M ./kernel/fftw.h +29 M ./kernel/flops.c +41 M ./kernel/ifftw.h +215 M ./kernel/minmax.c +33 M ./kernel/plan.c +50 M ./kernel/problem.c +47 M ./kernel/rand.c +36 M ./kernel/solver.c +43 M ./kernel/stride.c +41 M ./kernel/tensor.c +318 M ./kernel/version.c +26 M ./support/Makefile.am -1 +1 M ./support/Makefile.codelets +99 M ./support/codelet_prelude +8 fftw-3.3.3/INSTALL0000644000175000017500000003660011777117217010460 00000000000000Installation Instructions ************************* Copyright (C) 1994-1996, 1999-2002, 2004-2011 Free Software Foundation, Inc. Copying and distribution of this file, with or without modification, are permitted in any medium without royalty provided the copyright notice and this notice are preserved. This file is offered as-is, without warranty of any kind. Basic Installation ================== Briefly, the shell commands `./configure; make; make install' should configure, build, and install this package. The following more-detailed instructions are generic; see the `README' file for instructions specific to this package. Some packages provide this `INSTALL' file but do not implement all of the features documented below. The lack of an optional feature in a given package is not necessarily a bug. More recommendations for GNU packages can be found in *note Makefile Conventions: (standards)Makefile Conventions. The `configure' shell script attempts to guess correct values for various system-dependent variables used during compilation. It uses those values to create a `Makefile' in each directory of the package. It may also create one or more `.h' files containing system-dependent definitions. Finally, it creates a shell script `config.status' that you can run in the future to recreate the current configuration, and a file `config.log' containing compiler output (useful mainly for debugging `configure'). It can also use an optional file (typically called `config.cache' and enabled with `--cache-file=config.cache' or simply `-C') that saves the results of its tests to speed up reconfiguring. Caching is disabled by default to prevent problems with accidental use of stale cache files. If you need to do unusual things to compile the package, please try to figure out how `configure' could check whether to do them, and mail diffs or instructions to the address given in the `README' so they can be considered for the next release. If you are using the cache, and at some point `config.cache' contains results you don't want to keep, you may remove or edit it. The file `configure.ac' (or `configure.in') is used to create `configure' by a program called `autoconf'. You need `configure.ac' if you want to change it or regenerate `configure' using a newer version of `autoconf'. The simplest way to compile this package is: 1. `cd' to the directory containing the package's source code and type `./configure' to configure the package for your system. Running `configure' might take a while. While running, it prints some messages telling which features it is checking for. 2. Type `make' to compile the package. 3. Optionally, type `make check' to run any self-tests that come with the package, generally using the just-built uninstalled binaries. 4. Type `make install' to install the programs and any data files and documentation. When installing into a prefix owned by root, it is recommended that the package be configured and built as a regular user, and only the `make install' phase executed with root privileges. 5. Optionally, type `make installcheck' to repeat any self-tests, but this time using the binaries in their final installed location. This target does not install anything. Running this target as a regular user, particularly if the prior `make install' required root privileges, verifies that the installation completed correctly. 6. You can remove the program binaries and object files from the source code directory by typing `make clean'. To also remove the files that `configure' created (so you can compile the package for a different kind of computer), type `make distclean'. There is also a `make maintainer-clean' target, but that is intended mainly for the package's developers. If you use it, you may have to get all sorts of other programs in order to regenerate files that came with the distribution. 7. Often, you can also type `make uninstall' to remove the installed files again. In practice, not all packages have tested that uninstallation works correctly, even though it is required by the GNU Coding Standards. 8. Some packages, particularly those that use Automake, provide `make distcheck', which can by used by developers to test that all other targets like `make install' and `make uninstall' work correctly. This target is generally not run by end users. Compilers and Options ===================== Some systems require unusual options for compilation or linking that the `configure' script does not know about. Run `./configure --help' for details on some of the pertinent environment variables. You can give `configure' initial values for configuration parameters by setting variables in the command line or in the environment. Here is an example: ./configure CC=c99 CFLAGS=-g LIBS=-lposix *Note Defining Variables::, for more details. Compiling For Multiple Architectures ==================================== You can compile the package for more than one kind of computer at the same time, by placing the object files for each architecture in their own directory. To do this, you can use GNU `make'. `cd' to the directory where you want the object files and executables to go and run the `configure' script. `configure' automatically checks for the source code in the directory that `configure' is in and in `..'. This is known as a "VPATH" build. With a non-GNU `make', it is safer to compile the package for one architecture at a time in the source code directory. After you have installed the package for one architecture, use `make distclean' before reconfiguring for another architecture. On MacOS X 10.5 and later systems, you can create libraries and executables that work on multiple system types--known as "fat" or "universal" binaries--by specifying multiple `-arch' options to the compiler but only a single `-arch' option to the preprocessor. Like this: ./configure CC="gcc -arch i386 -arch x86_64 -arch ppc -arch ppc64" \ CXX="g++ -arch i386 -arch x86_64 -arch ppc -arch ppc64" \ CPP="gcc -E" CXXCPP="g++ -E" This is not guaranteed to produce working output in all cases, you may have to build one architecture at a time and combine the results using the `lipo' tool if you have problems. Installation Names ================== By default, `make install' installs the package's commands under `/usr/local/bin', include files under `/usr/local/include', etc. You can specify an installation prefix other than `/usr/local' by giving `configure' the option `--prefix=PREFIX', where PREFIX must be an absolute file name. You can specify separate installation prefixes for architecture-specific files and architecture-independent files. If you pass the option `--exec-prefix=PREFIX' to `configure', the package uses PREFIX as the prefix for installing programs and libraries. Documentation and other data files still use the regular prefix. In addition, if you use an unusual directory layout you can give options like `--bindir=DIR' to specify different values for particular kinds of files. Run `configure --help' for a list of the directories you can set and what kinds of files go in them. In general, the default for these options is expressed in terms of `${prefix}', so that specifying just `--prefix' will affect all of the other directory specifications that were not explicitly provided. The most portable way to affect installation locations is to pass the correct locations to `configure'; however, many packages provide one or both of the following shortcuts of passing variable assignments to the `make install' command line to change installation locations without having to reconfigure or recompile. The first method involves providing an override variable for each affected directory. For example, `make install prefix=/alternate/directory' will choose an alternate location for all directory configuration variables that were expressed in terms of `${prefix}'. Any directories that were specified during `configure', but not in terms of `${prefix}', must each be overridden at install time for the entire installation to be relocated. The approach of makefile variable overrides for each directory variable is required by the GNU Coding Standards, and ideally causes no recompilation. However, some platforms have known limitations with the semantics of shared libraries that end up requiring recompilation when using this method, particularly noticeable in packages that use GNU Libtool. The second method involves providing the `DESTDIR' variable. For example, `make install DESTDIR=/alternate/directory' will prepend `/alternate/directory' before all installation names. The approach of `DESTDIR' overrides is not required by the GNU Coding Standards, and does not work on platforms that have drive letters. On the other hand, it does better at avoiding recompilation issues, and works well even when some directory options were not specified in terms of `${prefix}' at `configure' time. Optional Features ================= If the package supports it, you can cause programs to be installed with an extra prefix or suffix on their names by giving `configure' the option `--program-prefix=PREFIX' or `--program-suffix=SUFFIX'. Some packages pay attention to `--enable-FEATURE' options to `configure', where FEATURE indicates an optional part of the package. They may also pay attention to `--with-PACKAGE' options, where PACKAGE is something like `gnu-as' or `x' (for the X Window System). The `README' should mention any `--enable-' and `--with-' options that the package recognizes. For packages that use the X Window System, `configure' can usually find the X include and library files automatically, but if it doesn't, you can use the `configure' options `--x-includes=DIR' and `--x-libraries=DIR' to specify their locations. Some packages offer the ability to configure how verbose the execution of `make' will be. For these packages, running `./configure --enable-silent-rules' sets the default to minimal output, which can be overridden with `make V=1'; while running `./configure --disable-silent-rules' sets the default to verbose, which can be overridden with `make V=0'. Particular systems ================== On HP-UX, the default C compiler is not ANSI C compatible. If GNU CC is not installed, it is recommended to use the following options in order to use an ANSI C compiler: ./configure CC="cc -Ae -D_XOPEN_SOURCE=500" and if that doesn't work, install pre-built binaries of GCC for HP-UX. HP-UX `make' updates targets which have the same time stamps as their prerequisites, which makes it generally unusable when shipped generated files such as `configure' are involved. Use GNU `make' instead. On OSF/1 a.k.a. Tru64, some versions of the default C compiler cannot parse its `' header file. The option `-nodtk' can be used as a workaround. If GNU CC is not installed, it is therefore recommended to try ./configure CC="cc" and if that doesn't work, try ./configure CC="cc -nodtk" On Solaris, don't put `/usr/ucb' early in your `PATH'. This directory contains several dysfunctional programs; working variants of these programs are available in `/usr/bin'. So, if you need `/usr/ucb' in your `PATH', put it _after_ `/usr/bin'. On Haiku, software installed for all users goes in `/boot/common', not `/usr/local'. It is recommended to use the following options: ./configure --prefix=/boot/common Specifying the System Type ========================== There may be some features `configure' cannot figure out automatically, but needs to determine by the type of machine the package will run on. Usually, assuming the package is built to be run on the _same_ architectures, `configure' can figure that out, but if it prints a message saying it cannot guess the machine type, give it the `--build=TYPE' option. TYPE can either be a short name for the system type, such as `sun4', or a canonical name which has the form: CPU-COMPANY-SYSTEM where SYSTEM can have one of these forms: OS KERNEL-OS See the file `config.sub' for the possible values of each field. If `config.sub' isn't included in this package, then this package doesn't need to know the machine type. If you are _building_ compiler tools for cross-compiling, you should use the option `--target=TYPE' to select the type of system they will produce code for. If you want to _use_ a cross compiler, that generates code for a platform different from the build platform, you should specify the "host" platform (i.e., that on which the generated programs will eventually be run) with `--host=TYPE'. Sharing Defaults ================ If you want to set default values for `configure' scripts to share, you can create a site shell script called `config.site' that gives default values for variables like `CC', `cache_file', and `prefix'. `configure' looks for `PREFIX/share/config.site' if it exists, then `PREFIX/etc/config.site' if it exists. Or, you can set the `CONFIG_SITE' environment variable to the location of the site script. A warning: not all `configure' scripts look for a site script. Defining Variables ================== Variables not defined in a site shell script can be set in the environment passed to `configure'. However, some packages may run configure again during the build, and the customized values of these variables may be lost. In order to avoid this problem, you should set them in the `configure' command line, using `VAR=value'. For example: ./configure CC=/usr/local2/bin/gcc causes the specified `gcc' to be used as the C compiler (unless it is overridden in the site shell script). Unfortunately, this technique does not work for `CONFIG_SHELL' due to an Autoconf bug. Until the bug is fixed you can use this workaround: CONFIG_SHELL=/bin/bash /bin/bash ./configure CONFIG_SHELL=/bin/bash `configure' Invocation ====================== `configure' recognizes the following options to control how it operates. `--help' `-h' Print a summary of all of the options to `configure', and exit. `--help=short' `--help=recursive' Print a summary of the options unique to this package's `configure', and exit. The `short' variant lists options used only in the top level, while the `recursive' variant lists options also present in any nested packages. `--version' `-V' Print the version of Autoconf used to generate the `configure' script, and exit. `--cache-file=FILE' Enable the cache: use and save the results of the tests in FILE, traditionally `config.cache'. FILE defaults to `/dev/null' to disable caching. `--config-cache' `-C' Alias for `--cache-file=config.cache'. `--quiet' `--silent' `-q' Do not print messages saying which checks are being made. To suppress all normal output, redirect it to `/dev/null' (any error messages will still be shown). `--srcdir=DIR' Look for the package's source code in directory DIR. Usually `configure' can determine that directory automatically. `--prefix=DIR' Use DIR as the installation prefix. *note Installation Names:: for more details, including other options available for fine-tuning the installation locations. `--no-create' `-n' Run the configure checks, but stop before creating any output files. `configure' also accepts some other, not widely useful, options. Run `configure --help' for more details. fftw-3.3.3/NEWS0000644000175000017500000004571612054410050010112 00000000000000FFTW 3.3.3 * Fix deadlock bug in MPI transforms (thanks to Michael Pippig for the bug report and patch, and to Graham Dennis for the bug report). * Use 128-bit ARM NEON instructions instead of 64-bits. This change appears to speed up even ARM processors with a 64-bit NEON pipe. * Speed improvements for single-precision AVX. * Speed up planner on machines without "official" cycle counters, such as ARM. FFTW 3.3.2 * Removed an archaic stack-alignment hack that was failing with gcc-4.7/i386. * Added stack-alignment hack necessary for gcc on Windows/i386. We will regret this in ten years (see previous change). * Fix incompatibility with Intel icc which pretends to be gcc but does not support quad precision. * make libfftw{threads,mpi} depend upon libfftw when using libtool; this is consistent with most other libraries and simplifies the life of various distributors of GNU/Linux. FFTW 3.3.1 * Changes since 3.3.1-beta1: - Reduced planning time in estimate mode for sizes with large prime factors. - Added AVX autodetection under Visual Studio. Thanks Carsten Steger for submitting the necessary code. - Modern Fortran interface now uses a separate fftw3l.f03 interface file for the long double interface, which is not supported by some Fortran compilers. Provided new fftw3q.f03 interface file to access the quadruple-precision FFTW routines with recent versions of gcc/gfortran. * Added support for the NEON extensions to the ARM ISA. (Note to beta users: an ARM cycle counter is not yet implemented; please contact fftw@fftw.org if you know how to do it right.) * MPI code now compiles even if mpicc is a C++ compiler; thanks to Kyle Spyksma for the bug report. FFTW 3.3 * Changes since 3.3-beta1: - Compiling OpenMP support (--enable-openmp) now installs a fftw3_omp library, instead of fftw3_threads, so that OpenMP and POSIX threads (--enable-threads) libraries can be built and installed at the same time. - Various minor compilation fixes, corrections of manual typos, and improvements to the benchmark test program. * Add support for the AVX extensions to x86 and x86-64. The AVX code works with 16-byte alignment (as opposed to 32-byte alignment), so there is no ABI change compared to FFTW 3.2.2. * Added Fortran 2003 interface, which should be usable on most modern Fortran compilers (e.g. gfortran) and provides type-checked access to the the C FFTW interface. (The legacy Fortran-77 interface is still included also.) * Added MPI distributed-memory transforms. Compared to 3.3alpha, the major changes in the MPI transforms are: - Fixed some deadlock and crashing bugs. - Added Fortran 2003 interface. - Added new-array execute functions for MPI plans. - Eliminated use of large MPI tags, since Cray MPI requires tags < 2^24; thanks to Jonathan Bentz for the bug report. - Expanded documentation. - 'make check' now runs MPI tests - Some ABI changes - not binary-compatible with 3.3alpha MPI. * Add support for quad-precision __float128 in gcc 4.6 or later (on x86. x86-64, and Itanium). The new routines use the fftwq_ prefix. * Removed support for MIPS paired-single instructions due to lack of available hardware for testing. Users who want this functionality should continue using FFTW 3.2.x. (Note that FFTW 3.3 still works on MIPS; this only concerns special instructions available on some MIPS chips.) * Removed support for the Cell Broadband Engine. Cell users should use FFTW 3.2.x. * New convenience functions fftw_alloc_real and fftw_alloc_complex to use fftw_malloc for real and complex arrays without typecasts or sizeof. * New convenience functions fftw_export_wisdom_to_filename and fftw_import_wisdom_from_filename that export/import wisdom to a file, which don't require you to open/close the file yourself. * New function fftw_cost to return FFTW's internal cost metric for a given plan; thanks to Rhys Ulerich and Nathanael Schaeffer for the suggestion. * The --enable-sse2 configure flag now works in both double and single precision (and is equivalent to --enable-sse in the latter case). * Remove --enable-portable-binary flag: we new produce portable binaries by default. * Remove the automatic detection of native architecture flag for gcc which was introduced in fftw-3.1, since new gcc supports -mtune=native. Remove the --with-gcc-arch flag; if you want to specify a particlar arch to configure, use ./configure CC="gcc -mtune=...". * --with-our-malloc16 configure flag is now renamed --with-our-malloc. * Fixed build problem failure when srand48 declaration is missing; thanks to Ralf Wildenhues for the bug report. * Fixed bug in fftw_set_timelimit: ensure that a negative timelimit is equivalent to no timelimit in all cases. Thanks to William Andrew Burnson for the bug report. * Fixed stack-overflow problem on OpenBSD caused by using alloca with too large a buffer. FFTW 3.2.2 * Improve performance of some copy operations of complex arrays on x86 machines. * Add configure flag to disable alloca(), which is broken in mingw64. * Planning in FFTW_ESTIMATE mode for r2r transforms became slower between fftw-3.1.3 and 3.2. This regression has now been fixed. FFTW 3.2.1 * Performance improvements for some multidimensional r2c/c2r transforms; thanks to Eugene Miloslavsky for his benchmark reports. * Compile with icc on MacOS X, use better icc compiler flags. * Compilation fixes for systems where snprintf is defined as a macro; thanks to Marcus Mae for the bug report. * Fortran documentation now recommends not using dfftw_execute, because of reports of problems with various Fortran compilers; it is better to use dfftw_execute_dft etcetera. * Some documentation clarifications, e.g. of fact that --enable-openmp and --enable-threads are mutually exclusive (thanks to Long To), and document slightly odd behavior of plan_guru_r2r in Fortran (thanks to Alexander Pozdneev). * FAQ was accidentally omitted from 3.2 tarball. * Remove some extraneous (harmless) files accidentally included in a subdirectory of the 3.2 tarball. FFTW 3.2 * Worked around apparent glibc bug that leads to rare hangs when freeing semaphores. * Fixed segfault due to unaligned access in certain obscure problems that use SSE and multiple threads. * MPI transforms not included, as they are still in alpha; the alpha versions of the MPI transforms have been moved to FFTW 3.3alpha1. FFTW 3.2alpha3 * Performance improvements for sizes with factors of 5 and 10. * Documented FFTW_WISDOM_ONLY flag, at the suggestion of Mario Emmenlauer and Phil Dumont. * Port Cell code to SDK2.1 (libspe2), as opposed to the old libspe1 code. * Performance improvements in Cell code for N < 32k, thanks to Jan Wagner for the suggestions. * Cycle counter for Sun x86_64 compiler, and compilation fix in cycle counter for AIX/xlc (thanks to Jeff Haferman for the bug report). * Fixed incorrect type prefix in MPI code that prevented wisdom routines from working in single precision (thanks to Eric A. Borisch for the report). * Added 'make check' for MPI code (which still fails in a couple corner cases, but should be much better than in alpha2). * Many other small fixes. FFTW 3.2alpha2 * Support for the Cell processor, donated by IBM Research; see README.Cell and the Cell section of the manual. * New 64-bit API: for every "plan_guru" function there is a new "plan_guru64" function with the same semantics, but which takes fftw_iodim64 instead of fftw_iodim. fftw_iodim64 is the same as fftw_iodim, except that it takes ptrdiff_t integer types as parameters, which is a 64-bit type on 64-bit machines. This is only useful for specifying very large transforms on 64-bit machines. (Internally, FFTW uses ptrdiff_t everywhere regardless of what API you choose.) * Experimental MPI support. Complex one- and multi-dimensional FFTs, multi-dimensional r2r, multi-dimensional r2c/c2r transforms, and distributed transpose operations, with 1d block distributions. (This is an alpha preview: routines have not been exhaustively tested, documentation is incomplete, and some functionality is missing, e.g. Fortran support.) See mpi/README and also the MPI section of the manual. * Significantly faster r2c/c2r transforms, especially on machines with SIMD. * Rewritten multi-threaded support for better performance by re-using a fixed pool of threads rather than continually respawning and joining (which nowadays is much slower). * Support for MIPS paired-single SIMD instructions, donated by Codesourcery. * FFTW_WISDOM_ONLY planner flag, to create plan only if wisdom is available and return NULL otherwise. * Removed k7 support, which only worked in 32-bit mode and is becoming obsolete. Use --enable-sse instead. * Added --with-g77-wrappers configure option to force inclusion of g77 wrappers, in addition to whatever is needed for the detected Fortran compilers. This is mainly intended for GNU/Linux distros switching to gfortran that wish to include both gfortran and g77 support in FFTW. * In manual, renamed "guru execute" functions to "new-array execute" functions, to reduce confusion with the guru planner interface. (The programming interface is unchanged.) * Add missing __declspec attribute to threads API functions when compiling for Windows; thanks to Robert O. Morris for the bug report. * Fixed missing return value from dfftw_init_threads in Fortran; thanks to Markus Wetzstein for the bug report. FFTW 3.1.1 * Performance improvements for Intel EMT64. * Performance improvements for large-size transforms with SIMD. * Cycle counter support for Intel icc and Visual C++ on x86-64. * In fftw-wisdom tool, replaced obsolete --impatient with --measure. * Fixed compilation failure with AIX/xlc; thanks to Joseph Thomas. * Windows DLL support for Fortran API (added missing __declspec(dllexport)). * SSE/SSE2 code works properly (i.e. disables itself) on older 386 and 486 CPUs lacking a CPUID instruction; thanks to Eric Korpela. FFTW 3.1 * Faster FFTW_ESTIMATE planner. * New (faster) algorithm for REDFT00/RODFT00 (type-I DCT/DST) of odd size. * "4-step" algorithm for faster FFTs of very large sizes (> 2^18). * Faster in-place real-data DFTs (for R2HC and HC2R r2r formats). * Faster in-place non-square transpositions (FFTW uses these internally for in-place FFTs, and you can also perform them explicitly using the guru interface). * Faster prime-size DFTs: implemented Bluestein's algorithm, as well as a zero-padded Rader variant to limit recursive use of Rader's algorithm. * SIMD support for split complex arrays. * Much faster Altivec/VMX performance. * New fftw_set_timelimit function to specify a (rough) upper bound to the planning time (does not affect ESTIMATE mode). * Removed --enable-3dnow support; use --enable-k7 instead. * FMA (fused multiply-add) version is now included in "standard" FFTW, and is enabled with --enable-fma (the default on PowerPC and Itanium). * Automatic detection of native architecture flag for gcc. New configure options: --enable-portable-binary and --with-gcc-arch=, for people distributing compiled binaries of FFTW (see manual). * Automatic detection of Altivec under Linux with gcc 3.4 (so that same binary should work on both Altivec and non-Altivec PowerPCs). * Compiler-specific tweaks/flags/workarounds for gcc 3.4, xlc, HP/UX, Solaris/Intel. * Various documentation clarifications. * 64-bit clean. (Fixes a bug affecting the split guru planner on 64-bit machines, reported by David Necas.) * Fixed Debian bug #259612: inadvertent use of SSE instructions on non-SSE machines (causing a crash) for --enable-sse binaries. * Fixed bug that caused HC2R transforms to destroy the input in certain cases, even if the user specified FFTW_PRESERVE_INPUT. * Fixed bug where wisdom would be lost under rare circumstances, causing excessive planning time. * FAQ notes bug in gcc-3.4.[1-3] that causes FFTW to crash with SSE/SSE2. * Fixed accidentally exported symbol that prohibited simultaneous linking to double/single multithreaded FFTW (thanks to Alessio Massaro). * Support Win32 threads under MinGW (thanks to Alessio Massaro). * Fixed problem with building DLL under Cygwin; thanks to Stephane Fillod. * Fix build failure if no Fortran compiler is found (thanks to Charles Radley for the bug report). * Fixed compilation failure with icc 8.0 and SSE/SSE2. Automatic detection of icc architecture flag (e.g. -xW). * Fixed compilation with OpenMP on AIX (thanks to Greg Bauer). * Fixed compilation failure on x86-64 with gcc (thanks to Orion Poplawski). * Incorporated patch from FreeBSD ports (FreeBSD does not have memalign, but its malloc is 16-byte aligned). * Cycle-counter compilation fixes for Itanium, Alpha, x86-64, Sparc, MacOS (thanks to Matt Boman, John Bowman, and James A. Treacy for reports/fixes). Added x86-64 cycle counter for PGI compilers, courtesy Cristiano Calonaci. * Fix compilation problem in test program due to C99 conflict. * Portability fix for import_system_wisdom with djgpp (thanks to Juan Manuel Guerrero). * Fixed compilation failure on MacOS 10.3 due to getopt conflict. * Work around Visual C++ (version 6/7) bug in SSE compilation; thanks to Eddie Yee for his detailed report. Changes from FFTW 3.1 beta 2: * Several minor compilation fixes. * Eliminate FFTW_TIMELIMIT flag and replace fftw_timelimit global with fftw_set_timelimit function. Make wisdom work with time-limited plans. Changes from FFTW 3.1 beta 1: * Fixes for creating DLLs under Windows; thanks to John Pavel for his feedback. * Fixed more 64-bit problems, thanks to John Pavel for the bug report. * Further speed improvements for Altivec/VMX. * Further speed improvements for non-square transpositions. * Many minor tweaks. FFTW 3.0.1 * Some speed improvements in SIMD code. * --without-cycle-counter option is removed. If no cycle counter is found, then the estimator is always used. A --with-slow-timer option is provided to force the use of lower-resolution timers. * Several fixes for compilation under Visual C++, with help from Stefane Ruel. * Added x86 cycle counter for Visual C++, with help from Morten Nissov. * Added S390 cycle counter, courtesy of James Treacy. * Added missing static keyword that prevented simultaneous linkage of different-precision versions; thanks to Rasmus Larsen for the bug report. * Corrected accidental omission of f77_wisdom.f file; thanks to Alan Watson. * Support -xopenmp flag for SunOS; thanks to John Lou for the bug report. * Compilation with HP/UX cc requires -Wp,-H128000 flag to increase preprocessor limits; thanks to Peter Vouras for the bug report. * Removed non-portable use of 'tempfile' in fftw-wisdom-to-conf script; thanks to Nicolas Decoster for the patch. * Added 'make smallcheck' target in tests/ directory, at the request of James Treacy. FFTW 3.0 Major goals of this release: * Speed: often 20% or more faster than FFTW 2.x, even without SIMD (see below). * Complete rewrite, to make it easier to add new algorithms and transforms. * New API, to support more general semantics. Other enhancements: * SIMD acceleration on supporting CPUs (SSE, SSE2, 3DNow!, and AltiVec). (With special thanks to Franz Franchetti for many experimental prototypes and to Stefan Kral for the vectorizing generator from fftwgel.) * True in-place 1d transforms of large sizes (as well as compressed twiddle tables for additional memory/cache savings). * More arbitrary placement of real & imaginary data, e.g. including interleaved (as in FFTW 2.x) as well as separate real/imag arrays. * Efficient prime-size transforms of real data. * Multidimensional transforms can operate on a subset of a larger matrix, and/or transform selected dimensions of a multidimensional array. * By popular demand, simultaneous linking to double precision (fftw), single precision (fftwf), and long-double precision (fftwl) versions of FFTW is now supported. * Cycle counters (on all modern CPUs) are exploited to speed planning. * Efficient transforms of real even/odd arrays, a.k.a. discrete cosine/sine transforms (types I-IV). (Currently work via pre/post processing of real transforms, ala FFTPACK, so are not optimal.) * DHTs (Discrete Hartley Transforms), again via post-processing of real transforms (and thus suboptimal, for now). * Support for linking to just those parts of FFTW that you need, greatly reducing the size of statically linked programs when only a limited set of transform sizes/types are required. * Canonical global wisdom file (/etc/fftw/wisdom) on Unix, along with a command-line tool (fftw-wisdom) to generate/update it. * Fortran API can be used with both g77 and non-g77 compilers simultaneously. * Multi-threaded version has optional OpenMP support. * Authors' good looks have greatly improved with age. Changes from 3.0beta3: * Separate FMA distribution to better exploit fused multiply-add instructions on PowerPC (and possibly other) architectures. * Performance improvements via some inlining tweaks. * fftw_flops now returns double arguments, not int, to avoid overflows for large sizes. * Workarounds for automake bugs. Changes from 3.0beta2: * The standard REDFT00/RODFT00 (DCT-I/DST-I) algorithm (used in FFTPACK, NR, etcetera) turns out to have poor numerical accuracy, so we replaced it with a slower routine that is more accurate. * The guru planner and execute functions now have two variants, one that takes complex arguments and one that takes separate real/imag pointers. * Execute and planner routines now automatically align the stack on x86, in case the calling program is misaligned. * README file for test program. * Fixed bugs in the combination of SIMD with multi-threaded transforms. * Eliminated internal fftw_threads_init function, which some people were calling accidentally instead of the fftw_init_threads API function. * Check for -openmp flag (Intel C compiler) when --enable-openmp is used. * Support AMD x86-64 SIMD and cycle counter. * Support SSE2 intrinsics in forthcoming gcc 3.3. Changes from 3.0beta1: * Faster in-place 1d transforms of non-power-of-two sizes. * SIMD improvements for in-place, multi-dimensional, and/or non-FFTW_PATIENT transforms. * Added support for hard-coded DCT/DST/DHT codelets of small sizes; the default distribution only includes hard-coded size-8 DCT-II/III, however. * Many minor improvements to the manual. Added section on using the codelet generator to customize and enhance FFTW. * The default 'make check' should now only take a few minutes; for more strenuous tests (which may take a day or so), do 'cd tests; make bigcheck'. * fftw_print_plan is split into fftw_fprint_plan and fftw_print_plan, where the latter uses stdout. * Fixed ability to compile with a C++ compiler. * Fixed support for C99 complex type under glibc. * Fixed problems with alloca under MinGW, AIX. * Workaround for gcc/SPARC bug. * Fixed multi-threaded initialization failure on IRIX due to lack of user-accessible PTHREAD_SCOPE_SYSTEM there. fftw-3.3.3/TODO0000644000175000017500000000234512054410050010072 00000000000000TODO before FFTW-$2\pi$: * Wisdom: make it clear that it is specific to the exact fftw version and configuration. Report error codes when reading wisdom. Maybe have multiple system wisdom files, one per version? * DCT/DST codelets? which kinds? * investigate the addition-chain trig computation * I can't believe that there isn't a closed form for the omega array in Rader. * convolution problem type(s) * Explore the idea of having n < 0 in tensors, possibly to mean inverse DFT. * better estimator: possibly, let "other" cost be coef * n, where coef is a per-solver constant determined via some big numerical optimization/fit. * vector radix, multidimensional codelets * it may be a good idea to unify all those little loops that do copying, (X[i], X[n-i]) <- (X[i] + X[n-i], X[i] - X[n-i]), and multiplication of vectors by twiddle factors. * Pruned FFTs (basically, a vecloop that skips zeros). * Try FFTPACK-style back-and-forth (Stockham) FFT. (We tried this a few years ago and it was slower, but perhaps matters have changed.) * Generate assembly directly for more processors, or maybe fork gcc. =) * ensure that threaded solvers generate (block_size % 4 == 0) to allow SIMD to be used. * memoize triggen. fftw-3.3.3/compile0000755000175000017500000001615211777117217011005 00000000000000#! /bin/sh # Wrapper for compilers which do not understand '-c -o'. scriptversion=2012-03-05.13; # UTC # Copyright (C) 1999, 2000, 2003, 2004, 2005, 2009, 2010, 2012 Free # Software Foundation, Inc. # Written by Tom Tromey . # # This program is free software; you can redistribute it and/or modify # it under the terms of the GNU General Public License as published by # the Free Software Foundation; either version 2, or (at your option) # any later version. # # This program is distributed in the hope that it will be useful, # but WITHOUT ANY WARRANTY; without even the implied warranty of # MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the # GNU General Public License for more details. # # You should have received a copy of the GNU General Public License # along with this program. If not, see . # As a special exception to the GNU General Public License, if you # distribute this file as part of a program that contains a # configuration script generated by Autoconf, you may include it under # the same distribution terms that you use for the rest of that program. # This file is maintained in Automake, please report # bugs to or send patches to # . nl=' ' # We need space, tab and new line, in precisely that order. Quoting is # there to prevent tools from complaining about whitespace usage. IFS=" "" $nl" file_conv= # func_file_conv build_file lazy # Convert a $build file to $host form and store it in $file # Currently only supports Windows hosts. If the determined conversion # type is listed in (the comma separated) LAZY, no conversion will # take place. func_file_conv () { file=$1 case $file in / | /[!/]*) # absolute file, and not a UNC file if test -z "$file_conv"; then # lazily determine how to convert abs files case `uname -s` in MINGW*) file_conv=mingw ;; CYGWIN*) file_conv=cygwin ;; *) file_conv=wine ;; esac fi case $file_conv/,$2, in *,$file_conv,*) ;; mingw/*) file=`cmd //C echo "$file " | sed -e 's/"\(.*\) " *$/\1/'` ;; cygwin/*) file=`cygpath -m "$file" || echo "$file"` ;; wine/*) file=`winepath -w "$file" || echo "$file"` ;; esac ;; esac } # func_cl_dashL linkdir # Make cl look for libraries in LINKDIR func_cl_dashL () { func_file_conv "$1" if test -z "$lib_path"; then lib_path=$file else lib_path="$lib_path;$file" fi linker_opts="$linker_opts -LIBPATH:$file" } # func_cl_dashl library # Do a library search-path lookup for cl func_cl_dashl () { lib=$1 found=no save_IFS=$IFS IFS=';' for dir in $lib_path $LIB do IFS=$save_IFS if $shared && test -f "$dir/$lib.dll.lib"; then found=yes lib=$dir/$lib.dll.lib break fi if test -f "$dir/$lib.lib"; then found=yes lib=$dir/$lib.lib break fi done IFS=$save_IFS if test "$found" != yes; then lib=$lib.lib fi } # func_cl_wrapper cl arg... # Adjust compile command to suit cl func_cl_wrapper () { # Assume a capable shell lib_path= shared=: linker_opts= for arg do if test -n "$eat"; then eat= else case $1 in -o) # configure might choose to run compile as 'compile cc -o foo foo.c'. eat=1 case $2 in *.o | *.[oO][bB][jJ]) func_file_conv "$2" set x "$@" -Fo"$file" shift ;; *) func_file_conv "$2" set x "$@" -Fe"$file" shift ;; esac ;; -I) eat=1 func_file_conv "$2" mingw set x "$@" -I"$file" shift ;; -I*) func_file_conv "${1#-I}" mingw set x "$@" -I"$file" shift ;; -l) eat=1 func_cl_dashl "$2" set x "$@" "$lib" shift ;; -l*) func_cl_dashl "${1#-l}" set x "$@" "$lib" shift ;; -L) eat=1 func_cl_dashL "$2" ;; -L*) func_cl_dashL "${1#-L}" ;; -static) shared=false ;; -Wl,*) arg=${1#-Wl,} save_ifs="$IFS"; IFS=',' for flag in $arg; do IFS="$save_ifs" linker_opts="$linker_opts $flag" done IFS="$save_ifs" ;; -Xlinker) eat=1 linker_opts="$linker_opts $2" ;; -*) set x "$@" "$1" shift ;; *.cc | *.CC | *.cxx | *.CXX | *.[cC]++) func_file_conv "$1" set x "$@" -Tp"$file" shift ;; *.c | *.cpp | *.CPP | *.lib | *.LIB | *.Lib | *.OBJ | *.obj | *.[oO]) func_file_conv "$1" mingw set x "$@" "$file" shift ;; *) set x "$@" "$1" shift ;; esac fi shift done if test -n "$linker_opts"; then linker_opts="-link$linker_opts" fi exec "$@" $linker_opts exit 1 } eat= case $1 in '') echo "$0: No command. Try '$0 --help' for more information." 1>&2 exit 1; ;; -h | --h*) cat <<\EOF Usage: compile [--help] [--version] PROGRAM [ARGS] Wrapper for compilers which do not understand '-c -o'. Remove '-o dest.o' from ARGS, run PROGRAM with the remaining arguments, and rename the output as expected. If you are trying to build a whole package this is not the right script to run: please start by reading the file 'INSTALL'. Report bugs to . EOF exit $? ;; -v | --v*) echo "compile $scriptversion" exit $? ;; cl | *[/\\]cl | cl.exe | *[/\\]cl.exe ) func_cl_wrapper "$@" # Doesn't return... ;; esac ofile= cfile= for arg do if test -n "$eat"; then eat= else case $1 in -o) # configure might choose to run compile as 'compile cc -o foo foo.c'. # So we strip '-o arg' only if arg is an object. eat=1 case $2 in *.o | *.obj) ofile=$2 ;; *) set x "$@" -o "$2" shift ;; esac ;; *.c) cfile=$1 set x "$@" "$1" shift ;; *) set x "$@" "$1" shift ;; esac fi shift done if test -z "$ofile" || test -z "$cfile"; then # If no '-o' option was seen then we might have been invoked from a # pattern rule where we don't need one. That is ok -- this is a # normal compilation that the losing compiler can handle. If no # '.c' file was seen then we are probably linking. That is also # ok. exec "$@" fi # Name of file we expect compiler to create. cofile=`echo "$cfile" | sed 's|^.*[\\/]||; s|^[a-zA-Z]:||; s/\.c$/.o/'` # Create the lock directory. # Note: use '[/\\:.-]' here to ensure that we don't use the same name # that we are using for the .o file. Also, base the name on the expected # object file name, since that is what matters with a parallel build. lockdir=`echo "$cofile" | sed -e 's|[/\\:.-]|_|g'`.d while true; do if mkdir "$lockdir" >/dev/null 2>&1; then break fi sleep 1 done # FIXME: race condition here if user kills between mkdir and trap. trap "rmdir '$lockdir'; exit 1" 1 2 15 # Run the compile. "$@" ret=$? if test -f "$cofile"; then test "$cofile" = "$ofile" || mv "$cofile" "$ofile" elif test -f "${cofile}bj"; then test "${cofile}bj" = "$ofile" || mv "${cofile}bj" "$ofile" fi rmdir "$lockdir" exit $ret # Local Variables: # mode: shell-script # sh-indentation: 2 # eval: (add-hook 'write-file-hooks 'time-stamp) # time-stamp-start: "scriptversion=" # time-stamp-format: "%:y-%02m-%02d.%02H" # time-stamp-time-zone: "UTC" # time-stamp-end: "; # UTC" # End: fftw-3.3.3/config.guess0000755000175000017500000012743211764422452011747 00000000000000#! /bin/sh # Attempt to guess a canonical system name. # Copyright (C) 1992, 1993, 1994, 1995, 1996, 1997, 1998, 1999, # 2000, 2001, 2002, 2003, 2004, 2005, 2006, 2007, 2008, 2009, 2010, # 2011, 2012 Free Software Foundation, Inc. timestamp='2012-02-10' # This file is free software; you can redistribute it and/or modify it # under the terms of the GNU General Public License as published by # the Free Software Foundation; either version 2 of the License, or # (at your option) any later version. # # This program is distributed in the hope that it will be useful, but # WITHOUT ANY WARRANTY; without even the implied warranty of # MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the GNU # General Public License for more details. # # You should have received a copy of the GNU General Public License # along with this program; if not, see . # # As a special exception to the GNU General Public License, if you # distribute this file as part of a program that contains a # configuration script generated by Autoconf, you may include it under # the same distribution terms that you use for the rest of that program. # Originally written by Per Bothner. Please send patches (context # diff format) to and include a ChangeLog # entry. # # This script attempts to guess a canonical system name similar to # config.sub. If it succeeds, it prints the system name on stdout, and # exits with 0. Otherwise, it exits with 1. # # You can get the latest version of this script from: # http://git.savannah.gnu.org/gitweb/?p=config.git;a=blob_plain;f=config.guess;hb=HEAD me=`echo "$0" | sed -e 's,.*/,,'` usage="\ Usage: $0 [OPTION] Output the configuration name of the system \`$me' is run on. Operation modes: -h, --help print this help, then exit -t, --time-stamp print date of last modification, then exit -v, --version print version number, then exit Report bugs and patches to ." version="\ GNU config.guess ($timestamp) Originally written by Per Bothner. Copyright (C) 1992, 1993, 1994, 1995, 1996, 1997, 1998, 1999, 2000, 2001, 2002, 2003, 2004, 2005, 2006, 2007, 2008, 2009, 2010, 2011, 2012 Free Software Foundation, Inc. This is free software; see the source for copying conditions. There is NO warranty; not even for MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE." help=" Try \`$me --help' for more information." # Parse command line while test $# -gt 0 ; do case $1 in --time-stamp | --time* | -t ) echo "$timestamp" ; exit ;; --version | -v ) echo "$version" ; exit ;; --help | --h* | -h ) echo "$usage"; exit ;; -- ) # Stop option processing shift; break ;; - ) # Use stdin as input. break ;; -* ) echo "$me: invalid option $1$help" >&2 exit 1 ;; * ) break ;; esac done if test $# != 0; then echo "$me: too many arguments$help" >&2 exit 1 fi trap 'exit 1' 1 2 15 # CC_FOR_BUILD -- compiler used by this script. Note that the use of a # compiler to aid in system detection is discouraged as it requires # temporary files to be created and, as you can see below, it is a # headache to deal with in a portable fashion. # Historically, `CC_FOR_BUILD' used to be named `HOST_CC'. We still # use `HOST_CC' if defined, but it is deprecated. # Portable tmp directory creation inspired by the Autoconf team. set_cc_for_build=' trap "exitcode=\$?; (rm -f \$tmpfiles 2>/dev/null; rmdir \$tmp 2>/dev/null) && exit \$exitcode" 0 ; trap "rm -f \$tmpfiles 2>/dev/null; rmdir \$tmp 2>/dev/null; exit 1" 1 2 13 15 ; : ${TMPDIR=/tmp} ; { tmp=`(umask 077 && mktemp -d "$TMPDIR/cgXXXXXX") 2>/dev/null` && test -n "$tmp" && test -d "$tmp" ; } || { test -n "$RANDOM" && tmp=$TMPDIR/cg$$-$RANDOM && (umask 077 && mkdir $tmp) ; } || { tmp=$TMPDIR/cg-$$ && (umask 077 && mkdir $tmp) && echo "Warning: creating insecure temp directory" >&2 ; } || { echo "$me: cannot create a temporary directory in $TMPDIR" >&2 ; exit 1 ; } ; dummy=$tmp/dummy ; tmpfiles="$dummy.c $dummy.o $dummy.rel $dummy" ; case $CC_FOR_BUILD,$HOST_CC,$CC in ,,) echo "int x;" > $dummy.c ; for c in cc gcc c89 c99 ; do if ($c -c -o $dummy.o $dummy.c) >/dev/null 2>&1 ; then CC_FOR_BUILD="$c"; break ; fi ; done ; if test x"$CC_FOR_BUILD" = x ; then CC_FOR_BUILD=no_compiler_found ; fi ;; ,,*) CC_FOR_BUILD=$CC ;; ,*,*) CC_FOR_BUILD=$HOST_CC ;; esac ; set_cc_for_build= ;' # This is needed to find uname on a Pyramid OSx when run in the BSD universe. # (ghazi@noc.rutgers.edu 1994-08-24) if (test -f /.attbin/uname) >/dev/null 2>&1 ; then PATH=$PATH:/.attbin ; export PATH fi UNAME_MACHINE=`(uname -m) 2>/dev/null` || UNAME_MACHINE=unknown UNAME_RELEASE=`(uname -r) 2>/dev/null` || UNAME_RELEASE=unknown UNAME_SYSTEM=`(uname -s) 2>/dev/null` || UNAME_SYSTEM=unknown UNAME_VERSION=`(uname -v) 2>/dev/null` || UNAME_VERSION=unknown # Note: order is significant - the case branches are not exclusive. case "${UNAME_MACHINE}:${UNAME_SYSTEM}:${UNAME_RELEASE}:${UNAME_VERSION}" in *:NetBSD:*:*) # NetBSD (nbsd) targets should (where applicable) match one or # more of the tuples: *-*-netbsdelf*, *-*-netbsdaout*, # *-*-netbsdecoff* and *-*-netbsd*. For targets that recently # switched to ELF, *-*-netbsd* would select the old # object file format. This provides both forward # compatibility and a consistent mechanism for selecting the # object file format. # # Note: NetBSD doesn't particularly care about the vendor # portion of the name. We always set it to "unknown". sysctl="sysctl -n hw.machine_arch" UNAME_MACHINE_ARCH=`(/sbin/$sysctl 2>/dev/null || \ /usr/sbin/$sysctl 2>/dev/null || echo unknown)` case "${UNAME_MACHINE_ARCH}" in armeb) machine=armeb-unknown ;; arm*) machine=arm-unknown ;; sh3el) machine=shl-unknown ;; sh3eb) machine=sh-unknown ;; sh5el) machine=sh5le-unknown ;; *) machine=${UNAME_MACHINE_ARCH}-unknown ;; esac # The Operating System including object format, if it has switched # to ELF recently, or will in the future. case "${UNAME_MACHINE_ARCH}" in arm*|i386|m68k|ns32k|sh3*|sparc|vax) eval $set_cc_for_build if echo __ELF__ | $CC_FOR_BUILD -E - 2>/dev/null \ | grep -q __ELF__ then # Once all utilities can be ECOFF (netbsdecoff) or a.out (netbsdaout). # Return netbsd for either. FIX? os=netbsd else os=netbsdelf fi ;; *) os=netbsd ;; esac # The OS release # Debian GNU/NetBSD machines have a different userland, and # thus, need a distinct triplet. However, they do not need # kernel version information, so it can be replaced with a # suitable tag, in the style of linux-gnu. case "${UNAME_VERSION}" in Debian*) release='-gnu' ;; *) release=`echo ${UNAME_RELEASE}|sed -e 's/[-_].*/\./'` ;; esac # Since CPU_TYPE-MANUFACTURER-KERNEL-OPERATING_SYSTEM: # contains redundant information, the shorter form: # CPU_TYPE-MANUFACTURER-OPERATING_SYSTEM is used. echo "${machine}-${os}${release}" exit ;; *:OpenBSD:*:*) UNAME_MACHINE_ARCH=`arch | sed 's/OpenBSD.//'` echo ${UNAME_MACHINE_ARCH}-unknown-openbsd${UNAME_RELEASE} exit ;; *:ekkoBSD:*:*) echo ${UNAME_MACHINE}-unknown-ekkobsd${UNAME_RELEASE} exit ;; *:SolidBSD:*:*) echo ${UNAME_MACHINE}-unknown-solidbsd${UNAME_RELEASE} exit ;; macppc:MirBSD:*:*) echo powerpc-unknown-mirbsd${UNAME_RELEASE} exit ;; *:MirBSD:*:*) echo ${UNAME_MACHINE}-unknown-mirbsd${UNAME_RELEASE} exit ;; alpha:OSF1:*:*) case $UNAME_RELEASE in *4.0) UNAME_RELEASE=`/usr/sbin/sizer -v | awk '{print $3}'` ;; *5.*) UNAME_RELEASE=`/usr/sbin/sizer -v | awk '{print $4}'` ;; esac # According to Compaq, /usr/sbin/psrinfo has been available on # OSF/1 and Tru64 systems produced since 1995. I hope that # covers most systems running today. This code pipes the CPU # types through head -n 1, so we only detect the type of CPU 0. ALPHA_CPU_TYPE=`/usr/sbin/psrinfo -v | sed -n -e 's/^ The alpha \(.*\) processor.*$/\1/p' | head -n 1` case "$ALPHA_CPU_TYPE" in "EV4 (21064)") UNAME_MACHINE="alpha" ;; "EV4.5 (21064)") UNAME_MACHINE="alpha" ;; "LCA4 (21066/21068)") UNAME_MACHINE="alpha" ;; "EV5 (21164)") UNAME_MACHINE="alphaev5" ;; "EV5.6 (21164A)") UNAME_MACHINE="alphaev56" ;; "EV5.6 (21164PC)") UNAME_MACHINE="alphapca56" ;; "EV5.7 (21164PC)") UNAME_MACHINE="alphapca57" ;; "EV6 (21264)") UNAME_MACHINE="alphaev6" ;; "EV6.7 (21264A)") UNAME_MACHINE="alphaev67" ;; "EV6.8CB (21264C)") UNAME_MACHINE="alphaev68" ;; "EV6.8AL (21264B)") UNAME_MACHINE="alphaev68" ;; "EV6.8CX (21264D)") UNAME_MACHINE="alphaev68" ;; "EV6.9A (21264/EV69A)") UNAME_MACHINE="alphaev69" ;; "EV7 (21364)") UNAME_MACHINE="alphaev7" ;; "EV7.9 (21364A)") UNAME_MACHINE="alphaev79" ;; esac # A Pn.n version is a patched version. # A Vn.n version is a released version. # A Tn.n version is a released field test version. # A Xn.n version is an unreleased experimental baselevel. # 1.2 uses "1.2" for uname -r. echo ${UNAME_MACHINE}-dec-osf`echo ${UNAME_RELEASE} | sed -e 's/^[PVTX]//' | tr 'ABCDEFGHIJKLMNOPQRSTUVWXYZ' 'abcdefghijklmnopqrstuvwxyz'` # Reset EXIT trap before exiting to avoid spurious non-zero exit code. exitcode=$? trap '' 0 exit $exitcode ;; Alpha\ *:Windows_NT*:*) # How do we know it's Interix rather than the generic POSIX subsystem? # Should we change UNAME_MACHINE based on the output of uname instead # of the specific Alpha model? echo alpha-pc-interix exit ;; 21064:Windows_NT:50:3) echo alpha-dec-winnt3.5 exit ;; Amiga*:UNIX_System_V:4.0:*) echo m68k-unknown-sysv4 exit ;; *:[Aa]miga[Oo][Ss]:*:*) echo ${UNAME_MACHINE}-unknown-amigaos exit ;; *:[Mm]orph[Oo][Ss]:*:*) echo ${UNAME_MACHINE}-unknown-morphos exit ;; *:OS/390:*:*) echo i370-ibm-openedition exit ;; *:z/VM:*:*) echo s390-ibm-zvmoe exit ;; *:OS400:*:*) echo powerpc-ibm-os400 exit ;; arm:RISC*:1.[012]*:*|arm:riscix:1.[012]*:*) echo arm-acorn-riscix${UNAME_RELEASE} exit ;; arm:riscos:*:*|arm:RISCOS:*:*) echo arm-unknown-riscos exit ;; SR2?01:HI-UX/MPP:*:* | SR8000:HI-UX/MPP:*:*) echo hppa1.1-hitachi-hiuxmpp exit ;; Pyramid*:OSx*:*:* | MIS*:OSx*:*:* | MIS*:SMP_DC-OSx*:*:*) # akee@wpdis03.wpafb.af.mil (Earle F. Ake) contributed MIS and NILE. if test "`(/bin/universe) 2>/dev/null`" = att ; then echo pyramid-pyramid-sysv3 else echo pyramid-pyramid-bsd fi exit ;; NILE*:*:*:dcosx) echo pyramid-pyramid-svr4 exit ;; DRS?6000:unix:4.0:6*) echo sparc-icl-nx6 exit ;; DRS?6000:UNIX_SV:4.2*:7* | DRS?6000:isis:4.2*:7*) case `/usr/bin/uname -p` in sparc) echo sparc-icl-nx7; exit ;; esac ;; s390x:SunOS:*:*) echo ${UNAME_MACHINE}-ibm-solaris2`echo ${UNAME_RELEASE}|sed -e 's/[^.]*//'` exit ;; sun4H:SunOS:5.*:*) echo sparc-hal-solaris2`echo ${UNAME_RELEASE}|sed -e 's/[^.]*//'` exit ;; sun4*:SunOS:5.*:* | tadpole*:SunOS:5.*:*) echo sparc-sun-solaris2`echo ${UNAME_RELEASE}|sed -e 's/[^.]*//'` exit ;; i86pc:AuroraUX:5.*:* | i86xen:AuroraUX:5.*:*) echo i386-pc-auroraux${UNAME_RELEASE} exit ;; i86pc:SunOS:5.*:* | i86xen:SunOS:5.*:*) eval $set_cc_for_build SUN_ARCH="i386" # If there is a compiler, see if it is configured for 64-bit objects. # Note that the Sun cc does not turn __LP64__ into 1 like gcc does. # This test works for both compilers. if [ "$CC_FOR_BUILD" != 'no_compiler_found' ]; then if (echo '#ifdef __amd64'; echo IS_64BIT_ARCH; echo '#endif') | \ (CCOPTS= $CC_FOR_BUILD -E - 2>/dev/null) | \ grep IS_64BIT_ARCH >/dev/null then SUN_ARCH="x86_64" fi fi echo ${SUN_ARCH}-pc-solaris2`echo ${UNAME_RELEASE}|sed -e 's/[^.]*//'` exit ;; sun4*:SunOS:6*:*) # According to config.sub, this is the proper way to canonicalize # SunOS6. Hard to guess exactly what SunOS6 will be like, but # it's likely to be more like Solaris than SunOS4. echo sparc-sun-solaris3`echo ${UNAME_RELEASE}|sed -e 's/[^.]*//'` exit ;; sun4*:SunOS:*:*) case "`/usr/bin/arch -k`" in Series*|S4*) UNAME_RELEASE=`uname -v` ;; esac # Japanese Language versions have a version number like `4.1.3-JL'. echo sparc-sun-sunos`echo ${UNAME_RELEASE}|sed -e 's/-/_/'` exit ;; sun3*:SunOS:*:*) echo m68k-sun-sunos${UNAME_RELEASE} exit ;; sun*:*:4.2BSD:*) UNAME_RELEASE=`(sed 1q /etc/motd | awk '{print substr($5,1,3)}') 2>/dev/null` test "x${UNAME_RELEASE}" = "x" && UNAME_RELEASE=3 case "`/bin/arch`" in sun3) echo m68k-sun-sunos${UNAME_RELEASE} ;; sun4) echo sparc-sun-sunos${UNAME_RELEASE} ;; esac exit ;; aushp:SunOS:*:*) echo sparc-auspex-sunos${UNAME_RELEASE} exit ;; # The situation for MiNT is a little confusing. The machine name # can be virtually everything (everything which is not # "atarist" or "atariste" at least should have a processor # > m68000). The system name ranges from "MiNT" over "FreeMiNT" # to the lowercase version "mint" (or "freemint"). Finally # the system name "TOS" denotes a system which is actually not # MiNT. But MiNT is downward compatible to TOS, so this should # be no problem. atarist[e]:*MiNT:*:* | atarist[e]:*mint:*:* | atarist[e]:*TOS:*:*) echo m68k-atari-mint${UNAME_RELEASE} exit ;; atari*:*MiNT:*:* | atari*:*mint:*:* | atarist[e]:*TOS:*:*) echo m68k-atari-mint${UNAME_RELEASE} exit ;; *falcon*:*MiNT:*:* | *falcon*:*mint:*:* | *falcon*:*TOS:*:*) echo m68k-atari-mint${UNAME_RELEASE} exit ;; milan*:*MiNT:*:* | milan*:*mint:*:* | *milan*:*TOS:*:*) echo m68k-milan-mint${UNAME_RELEASE} exit ;; hades*:*MiNT:*:* | hades*:*mint:*:* | *hades*:*TOS:*:*) echo m68k-hades-mint${UNAME_RELEASE} exit ;; *:*MiNT:*:* | *:*mint:*:* | *:*TOS:*:*) echo m68k-unknown-mint${UNAME_RELEASE} exit ;; m68k:machten:*:*) echo m68k-apple-machten${UNAME_RELEASE} exit ;; powerpc:machten:*:*) echo powerpc-apple-machten${UNAME_RELEASE} exit ;; RISC*:Mach:*:*) echo mips-dec-mach_bsd4.3 exit ;; RISC*:ULTRIX:*:*) echo mips-dec-ultrix${UNAME_RELEASE} exit ;; VAX*:ULTRIX*:*:*) echo vax-dec-ultrix${UNAME_RELEASE} exit ;; 2020:CLIX:*:* | 2430:CLIX:*:*) echo clipper-intergraph-clix${UNAME_RELEASE} exit ;; mips:*:*:UMIPS | mips:*:*:RISCos) eval $set_cc_for_build sed 's/^ //' << EOF >$dummy.c #ifdef __cplusplus #include /* for printf() prototype */ int main (int argc, char *argv[]) { #else int main (argc, argv) int argc; char *argv[]; { #endif #if defined (host_mips) && defined (MIPSEB) #if defined (SYSTYPE_SYSV) printf ("mips-mips-riscos%ssysv\n", argv[1]); exit (0); #endif #if defined (SYSTYPE_SVR4) printf ("mips-mips-riscos%ssvr4\n", argv[1]); exit (0); #endif #if defined (SYSTYPE_BSD43) || defined(SYSTYPE_BSD) printf ("mips-mips-riscos%sbsd\n", argv[1]); exit (0); #endif #endif exit (-1); } EOF $CC_FOR_BUILD -o $dummy $dummy.c && dummyarg=`echo "${UNAME_RELEASE}" | sed -n 's/\([0-9]*\).*/\1/p'` && SYSTEM_NAME=`$dummy $dummyarg` && { echo "$SYSTEM_NAME"; exit; } echo mips-mips-riscos${UNAME_RELEASE} exit ;; Motorola:PowerMAX_OS:*:*) echo powerpc-motorola-powermax exit ;; Motorola:*:4.3:PL8-*) echo powerpc-harris-powermax exit ;; Night_Hawk:*:*:PowerMAX_OS | Synergy:PowerMAX_OS:*:*) echo powerpc-harris-powermax exit ;; Night_Hawk:Power_UNIX:*:*) echo powerpc-harris-powerunix exit ;; m88k:CX/UX:7*:*) echo m88k-harris-cxux7 exit ;; m88k:*:4*:R4*) echo m88k-motorola-sysv4 exit ;; m88k:*:3*:R3*) echo m88k-motorola-sysv3 exit ;; AViiON:dgux:*:*) # DG/UX returns AViiON for all architectures UNAME_PROCESSOR=`/usr/bin/uname -p` if [ $UNAME_PROCESSOR = mc88100 ] || [ $UNAME_PROCESSOR = mc88110 ] then if [ ${TARGET_BINARY_INTERFACE}x = m88kdguxelfx ] || \ [ ${TARGET_BINARY_INTERFACE}x = x ] then echo m88k-dg-dgux${UNAME_RELEASE} else echo m88k-dg-dguxbcs${UNAME_RELEASE} fi else echo i586-dg-dgux${UNAME_RELEASE} fi exit ;; M88*:DolphinOS:*:*) # DolphinOS (SVR3) echo m88k-dolphin-sysv3 exit ;; M88*:*:R3*:*) # Delta 88k system running SVR3 echo m88k-motorola-sysv3 exit ;; XD88*:*:*:*) # Tektronix XD88 system running UTekV (SVR3) echo m88k-tektronix-sysv3 exit ;; Tek43[0-9][0-9]:UTek:*:*) # Tektronix 4300 system running UTek (BSD) echo m68k-tektronix-bsd exit ;; *:IRIX*:*:*) echo mips-sgi-irix`echo ${UNAME_RELEASE}|sed -e 's/-/_/g'` exit ;; ????????:AIX?:[12].1:2) # AIX 2.2.1 or AIX 2.1.1 is RT/PC AIX. echo romp-ibm-aix # uname -m gives an 8 hex-code CPU id exit ;; # Note that: echo "'`uname -s`'" gives 'AIX ' i*86:AIX:*:*) echo i386-ibm-aix exit ;; ia64:AIX:*:*) if [ -x /usr/bin/oslevel ] ; then IBM_REV=`/usr/bin/oslevel` else IBM_REV=${UNAME_VERSION}.${UNAME_RELEASE} fi echo ${UNAME_MACHINE}-ibm-aix${IBM_REV} exit ;; *:AIX:2:3) if grep bos325 /usr/include/stdio.h >/dev/null 2>&1; then eval $set_cc_for_build sed 's/^ //' << EOF >$dummy.c #include main() { if (!__power_pc()) exit(1); puts("powerpc-ibm-aix3.2.5"); exit(0); } EOF if $CC_FOR_BUILD -o $dummy $dummy.c && SYSTEM_NAME=`$dummy` then echo "$SYSTEM_NAME" else echo rs6000-ibm-aix3.2.5 fi elif grep bos324 /usr/include/stdio.h >/dev/null 2>&1; then echo rs6000-ibm-aix3.2.4 else echo rs6000-ibm-aix3.2 fi exit ;; *:AIX:*:[4567]) IBM_CPU_ID=`/usr/sbin/lsdev -C -c processor -S available | sed 1q | awk '{ print $1 }'` if /usr/sbin/lsattr -El ${IBM_CPU_ID} | grep ' POWER' >/dev/null 2>&1; then IBM_ARCH=rs6000 else IBM_ARCH=powerpc fi if [ -x /usr/bin/oslevel ] ; then IBM_REV=`/usr/bin/oslevel` else IBM_REV=${UNAME_VERSION}.${UNAME_RELEASE} fi echo ${IBM_ARCH}-ibm-aix${IBM_REV} exit ;; *:AIX:*:*) echo rs6000-ibm-aix exit ;; ibmrt:4.4BSD:*|romp-ibm:BSD:*) echo romp-ibm-bsd4.4 exit ;; ibmrt:*BSD:*|romp-ibm:BSD:*) # covers RT/PC BSD and echo romp-ibm-bsd${UNAME_RELEASE} # 4.3 with uname added to exit ;; # report: romp-ibm BSD 4.3 *:BOSX:*:*) echo rs6000-bull-bosx exit ;; DPX/2?00:B.O.S.:*:*) echo m68k-bull-sysv3 exit ;; 9000/[34]??:4.3bsd:1.*:*) echo m68k-hp-bsd exit ;; hp300:4.4BSD:*:* | 9000/[34]??:4.3bsd:2.*:*) echo m68k-hp-bsd4.4 exit ;; 9000/[34678]??:HP-UX:*:*) HPUX_REV=`echo ${UNAME_RELEASE}|sed -e 's/[^.]*.[0B]*//'` case "${UNAME_MACHINE}" in 9000/31? ) HP_ARCH=m68000 ;; 9000/[34]?? ) HP_ARCH=m68k ;; 9000/[678][0-9][0-9]) if [ -x /usr/bin/getconf ]; then sc_cpu_version=`/usr/bin/getconf SC_CPU_VERSION 2>/dev/null` sc_kernel_bits=`/usr/bin/getconf SC_KERNEL_BITS 2>/dev/null` case "${sc_cpu_version}" in 523) HP_ARCH="hppa1.0" ;; # CPU_PA_RISC1_0 528) HP_ARCH="hppa1.1" ;; # CPU_PA_RISC1_1 532) # CPU_PA_RISC2_0 case "${sc_kernel_bits}" in 32) HP_ARCH="hppa2.0n" ;; 64) HP_ARCH="hppa2.0w" ;; '') HP_ARCH="hppa2.0" ;; # HP-UX 10.20 esac ;; esac fi if [ "${HP_ARCH}" = "" ]; then eval $set_cc_for_build sed 's/^ //' << EOF >$dummy.c #define _HPUX_SOURCE #include #include int main () { #if defined(_SC_KERNEL_BITS) long bits = sysconf(_SC_KERNEL_BITS); #endif long cpu = sysconf (_SC_CPU_VERSION); switch (cpu) { case CPU_PA_RISC1_0: puts ("hppa1.0"); break; case CPU_PA_RISC1_1: puts ("hppa1.1"); break; case CPU_PA_RISC2_0: #if defined(_SC_KERNEL_BITS) switch (bits) { case 64: puts ("hppa2.0w"); break; case 32: puts ("hppa2.0n"); break; default: puts ("hppa2.0"); break; } break; #else /* !defined(_SC_KERNEL_BITS) */ puts ("hppa2.0"); break; #endif default: puts ("hppa1.0"); break; } exit (0); } EOF (CCOPTS= $CC_FOR_BUILD -o $dummy $dummy.c 2>/dev/null) && HP_ARCH=`$dummy` test -z "$HP_ARCH" && HP_ARCH=hppa fi ;; esac if [ ${HP_ARCH} = "hppa2.0w" ] then eval $set_cc_for_build # hppa2.0w-hp-hpux* has a 64-bit kernel and a compiler generating # 32-bit code. hppa64-hp-hpux* has the same kernel and a compiler # generating 64-bit code. GNU and HP use different nomenclature: # # $ CC_FOR_BUILD=cc ./config.guess # => hppa2.0w-hp-hpux11.23 # $ CC_FOR_BUILD="cc +DA2.0w" ./config.guess # => hppa64-hp-hpux11.23 if echo __LP64__ | (CCOPTS= $CC_FOR_BUILD -E - 2>/dev/null) | grep -q __LP64__ then HP_ARCH="hppa2.0w" else HP_ARCH="hppa64" fi fi echo ${HP_ARCH}-hp-hpux${HPUX_REV} exit ;; ia64:HP-UX:*:*) HPUX_REV=`echo ${UNAME_RELEASE}|sed -e 's/[^.]*.[0B]*//'` echo ia64-hp-hpux${HPUX_REV} exit ;; 3050*:HI-UX:*:*) eval $set_cc_for_build sed 's/^ //' << EOF >$dummy.c #include int main () { long cpu = sysconf (_SC_CPU_VERSION); /* The order matters, because CPU_IS_HP_MC68K erroneously returns true for CPU_PA_RISC1_0. CPU_IS_PA_RISC returns correct results, however. */ if (CPU_IS_PA_RISC (cpu)) { switch (cpu) { case CPU_PA_RISC1_0: puts ("hppa1.0-hitachi-hiuxwe2"); break; case CPU_PA_RISC1_1: puts ("hppa1.1-hitachi-hiuxwe2"); break; case CPU_PA_RISC2_0: puts ("hppa2.0-hitachi-hiuxwe2"); break; default: puts ("hppa-hitachi-hiuxwe2"); break; } } else if (CPU_IS_HP_MC68K (cpu)) puts ("m68k-hitachi-hiuxwe2"); else puts ("unknown-hitachi-hiuxwe2"); exit (0); } EOF $CC_FOR_BUILD -o $dummy $dummy.c && SYSTEM_NAME=`$dummy` && { echo "$SYSTEM_NAME"; exit; } echo unknown-hitachi-hiuxwe2 exit ;; 9000/7??:4.3bsd:*:* | 9000/8?[79]:4.3bsd:*:* ) echo hppa1.1-hp-bsd exit ;; 9000/8??:4.3bsd:*:*) echo hppa1.0-hp-bsd exit ;; *9??*:MPE/iX:*:* | *3000*:MPE/iX:*:*) echo hppa1.0-hp-mpeix exit ;; hp7??:OSF1:*:* | hp8?[79]:OSF1:*:* ) echo hppa1.1-hp-osf exit ;; hp8??:OSF1:*:*) echo hppa1.0-hp-osf exit ;; i*86:OSF1:*:*) if [ -x /usr/sbin/sysversion ] ; then echo ${UNAME_MACHINE}-unknown-osf1mk else echo ${UNAME_MACHINE}-unknown-osf1 fi exit ;; parisc*:Lites*:*:*) echo hppa1.1-hp-lites exit ;; C1*:ConvexOS:*:* | convex:ConvexOS:C1*:*) echo c1-convex-bsd exit ;; C2*:ConvexOS:*:* | convex:ConvexOS:C2*:*) if getsysinfo -f scalar_acc then echo c32-convex-bsd else echo c2-convex-bsd fi exit ;; C34*:ConvexOS:*:* | convex:ConvexOS:C34*:*) echo c34-convex-bsd exit ;; C38*:ConvexOS:*:* | convex:ConvexOS:C38*:*) echo c38-convex-bsd exit ;; C4*:ConvexOS:*:* | convex:ConvexOS:C4*:*) echo c4-convex-bsd exit ;; CRAY*Y-MP:*:*:*) echo ymp-cray-unicos${UNAME_RELEASE} | sed -e 's/\.[^.]*$/.X/' exit ;; CRAY*[A-Z]90:*:*:*) echo ${UNAME_MACHINE}-cray-unicos${UNAME_RELEASE} \ | sed -e 's/CRAY.*\([A-Z]90\)/\1/' \ -e y/ABCDEFGHIJKLMNOPQRSTUVWXYZ/abcdefghijklmnopqrstuvwxyz/ \ -e 's/\.[^.]*$/.X/' exit ;; CRAY*TS:*:*:*) echo t90-cray-unicos${UNAME_RELEASE} | sed -e 's/\.[^.]*$/.X/' exit ;; CRAY*T3E:*:*:*) echo alphaev5-cray-unicosmk${UNAME_RELEASE} | sed -e 's/\.[^.]*$/.X/' exit ;; CRAY*SV1:*:*:*) echo sv1-cray-unicos${UNAME_RELEASE} | sed -e 's/\.[^.]*$/.X/' exit ;; *:UNICOS/mp:*:*) echo craynv-cray-unicosmp${UNAME_RELEASE} | sed -e 's/\.[^.]*$/.X/' exit ;; F30[01]:UNIX_System_V:*:* | F700:UNIX_System_V:*:*) FUJITSU_PROC=`uname -m | tr 'ABCDEFGHIJKLMNOPQRSTUVWXYZ' 'abcdefghijklmnopqrstuvwxyz'` FUJITSU_SYS=`uname -p | tr 'ABCDEFGHIJKLMNOPQRSTUVWXYZ' 'abcdefghijklmnopqrstuvwxyz' | sed -e 's/\///'` FUJITSU_REL=`echo ${UNAME_RELEASE} | sed -e 's/ /_/'` echo "${FUJITSU_PROC}-fujitsu-${FUJITSU_SYS}${FUJITSU_REL}" exit ;; 5000:UNIX_System_V:4.*:*) FUJITSU_SYS=`uname -p | tr 'ABCDEFGHIJKLMNOPQRSTUVWXYZ' 'abcdefghijklmnopqrstuvwxyz' | sed -e 's/\///'` FUJITSU_REL=`echo ${UNAME_RELEASE} | tr 'ABCDEFGHIJKLMNOPQRSTUVWXYZ' 'abcdefghijklmnopqrstuvwxyz' | sed -e 's/ /_/'` echo "sparc-fujitsu-${FUJITSU_SYS}${FUJITSU_REL}" exit ;; i*86:BSD/386:*:* | i*86:BSD/OS:*:* | *:Ascend\ Embedded/OS:*:*) echo ${UNAME_MACHINE}-pc-bsdi${UNAME_RELEASE} exit ;; sparc*:BSD/OS:*:*) echo sparc-unknown-bsdi${UNAME_RELEASE} exit ;; *:BSD/OS:*:*) echo ${UNAME_MACHINE}-unknown-bsdi${UNAME_RELEASE} exit ;; *:FreeBSD:*:*) UNAME_PROCESSOR=`/usr/bin/uname -p` case ${UNAME_PROCESSOR} in amd64) echo x86_64-unknown-freebsd`echo ${UNAME_RELEASE}|sed -e 's/[-(].*//'` ;; *) echo ${UNAME_PROCESSOR}-unknown-freebsd`echo ${UNAME_RELEASE}|sed -e 's/[-(].*//'` ;; esac exit ;; i*:CYGWIN*:*) echo ${UNAME_MACHINE}-pc-cygwin exit ;; *:MINGW*:*) echo ${UNAME_MACHINE}-pc-mingw32 exit ;; i*:MSYS*:*) echo ${UNAME_MACHINE}-pc-msys exit ;; i*:windows32*:*) # uname -m includes "-pc" on this system. echo ${UNAME_MACHINE}-mingw32 exit ;; i*:PW*:*) echo ${UNAME_MACHINE}-pc-pw32 exit ;; *:Interix*:*) case ${UNAME_MACHINE} in x86) echo i586-pc-interix${UNAME_RELEASE} exit ;; authenticamd | genuineintel | EM64T) echo x86_64-unknown-interix${UNAME_RELEASE} exit ;; IA64) echo ia64-unknown-interix${UNAME_RELEASE} exit ;; esac ;; [345]86:Windows_95:* | [345]86:Windows_98:* | [345]86:Windows_NT:*) echo i${UNAME_MACHINE}-pc-mks exit ;; 8664:Windows_NT:*) echo x86_64-pc-mks exit ;; i*:Windows_NT*:* | Pentium*:Windows_NT*:*) # How do we know it's Interix rather than the generic POSIX subsystem? # It also conflicts with pre-2.0 versions of AT&T UWIN. Should we # UNAME_MACHINE based on the output of uname instead of i386? echo i586-pc-interix exit ;; i*:UWIN*:*) echo ${UNAME_MACHINE}-pc-uwin exit ;; amd64:CYGWIN*:*:* | x86_64:CYGWIN*:*:*) echo x86_64-unknown-cygwin exit ;; p*:CYGWIN*:*) echo powerpcle-unknown-cygwin exit ;; prep*:SunOS:5.*:*) echo powerpcle-unknown-solaris2`echo ${UNAME_RELEASE}|sed -e 's/[^.]*//'` exit ;; *:GNU:*:*) # the GNU system echo `echo ${UNAME_MACHINE}|sed -e 's,[-/].*$,,'`-unknown-gnu`echo ${UNAME_RELEASE}|sed -e 's,/.*$,,'` exit ;; *:GNU/*:*:*) # other systems with GNU libc and userland echo ${UNAME_MACHINE}-unknown-`echo ${UNAME_SYSTEM} | sed 's,^[^/]*/,,' | tr '[A-Z]' '[a-z]'``echo ${UNAME_RELEASE}|sed -e 's/[-(].*//'`-gnu exit ;; i*86:Minix:*:*) echo ${UNAME_MACHINE}-pc-minix exit ;; aarch64:Linux:*:*) echo ${UNAME_MACHINE}-unknown-linux-gnu exit ;; aarch64_be:Linux:*:*) UNAME_MACHINE=aarch64_be echo ${UNAME_MACHINE}-unknown-linux-gnu exit ;; alpha:Linux:*:*) case `sed -n '/^cpu model/s/^.*: \(.*\)/\1/p' < /proc/cpuinfo` in EV5) UNAME_MACHINE=alphaev5 ;; EV56) UNAME_MACHINE=alphaev56 ;; PCA56) UNAME_MACHINE=alphapca56 ;; PCA57) UNAME_MACHINE=alphapca56 ;; EV6) UNAME_MACHINE=alphaev6 ;; EV67) UNAME_MACHINE=alphaev67 ;; EV68*) UNAME_MACHINE=alphaev68 ;; esac objdump --private-headers /bin/sh | grep -q ld.so.1 if test "$?" = 0 ; then LIBC="libc1" ; else LIBC="" ; fi echo ${UNAME_MACHINE}-unknown-linux-gnu${LIBC} exit ;; arm*:Linux:*:*) eval $set_cc_for_build if echo __ARM_EABI__ | $CC_FOR_BUILD -E - 2>/dev/null \ | grep -q __ARM_EABI__ then echo ${UNAME_MACHINE}-unknown-linux-gnu else if echo __ARM_PCS_VFP | $CC_FOR_BUILD -E - 2>/dev/null \ | grep -q __ARM_PCS_VFP then echo ${UNAME_MACHINE}-unknown-linux-gnueabi else echo ${UNAME_MACHINE}-unknown-linux-gnueabihf fi fi exit ;; avr32*:Linux:*:*) echo ${UNAME_MACHINE}-unknown-linux-gnu exit ;; cris:Linux:*:*) echo ${UNAME_MACHINE}-axis-linux-gnu exit ;; crisv32:Linux:*:*) echo ${UNAME_MACHINE}-axis-linux-gnu exit ;; frv:Linux:*:*) echo ${UNAME_MACHINE}-unknown-linux-gnu exit ;; hexagon:Linux:*:*) echo ${UNAME_MACHINE}-unknown-linux-gnu exit ;; i*86:Linux:*:*) LIBC=gnu eval $set_cc_for_build sed 's/^ //' << EOF >$dummy.c #ifdef __dietlibc__ LIBC=dietlibc #endif EOF eval `$CC_FOR_BUILD -E $dummy.c 2>/dev/null | grep '^LIBC'` echo "${UNAME_MACHINE}-pc-linux-${LIBC}" exit ;; ia64:Linux:*:*) echo ${UNAME_MACHINE}-unknown-linux-gnu exit ;; m32r*:Linux:*:*) echo ${UNAME_MACHINE}-unknown-linux-gnu exit ;; m68*:Linux:*:*) echo ${UNAME_MACHINE}-unknown-linux-gnu exit ;; mips:Linux:*:* | mips64:Linux:*:*) eval $set_cc_for_build sed 's/^ //' << EOF >$dummy.c #undef CPU #undef ${UNAME_MACHINE} #undef ${UNAME_MACHINE}el #if defined(__MIPSEL__) || defined(__MIPSEL) || defined(_MIPSEL) || defined(MIPSEL) CPU=${UNAME_MACHINE}el #else #if defined(__MIPSEB__) || defined(__MIPSEB) || defined(_MIPSEB) || defined(MIPSEB) CPU=${UNAME_MACHINE} #else CPU= #endif #endif EOF eval `$CC_FOR_BUILD -E $dummy.c 2>/dev/null | grep '^CPU'` test x"${CPU}" != x && { echo "${CPU}-unknown-linux-gnu"; exit; } ;; or32:Linux:*:*) echo ${UNAME_MACHINE}-unknown-linux-gnu exit ;; padre:Linux:*:*) echo sparc-unknown-linux-gnu exit ;; parisc64:Linux:*:* | hppa64:Linux:*:*) echo hppa64-unknown-linux-gnu exit ;; parisc:Linux:*:* | hppa:Linux:*:*) # Look for CPU level case `grep '^cpu[^a-z]*:' /proc/cpuinfo 2>/dev/null | cut -d' ' -f2` in PA7*) echo hppa1.1-unknown-linux-gnu ;; PA8*) echo hppa2.0-unknown-linux-gnu ;; *) echo hppa-unknown-linux-gnu ;; esac exit ;; ppc64:Linux:*:*) echo powerpc64-unknown-linux-gnu exit ;; ppc:Linux:*:*) echo powerpc-unknown-linux-gnu exit ;; s390:Linux:*:* | s390x:Linux:*:*) echo ${UNAME_MACHINE}-ibm-linux exit ;; sh64*:Linux:*:*) echo ${UNAME_MACHINE}-unknown-linux-gnu exit ;; sh*:Linux:*:*) echo ${UNAME_MACHINE}-unknown-linux-gnu exit ;; sparc:Linux:*:* | sparc64:Linux:*:*) echo ${UNAME_MACHINE}-unknown-linux-gnu exit ;; tile*:Linux:*:*) echo ${UNAME_MACHINE}-unknown-linux-gnu exit ;; vax:Linux:*:*) echo ${UNAME_MACHINE}-dec-linux-gnu exit ;; x86_64:Linux:*:*) echo ${UNAME_MACHINE}-unknown-linux-gnu exit ;; xtensa*:Linux:*:*) echo ${UNAME_MACHINE}-unknown-linux-gnu exit ;; i*86:DYNIX/ptx:4*:*) # ptx 4.0 does uname -s correctly, with DYNIX/ptx in there. # earlier versions are messed up and put the nodename in both # sysname and nodename. echo i386-sequent-sysv4 exit ;; i*86:UNIX_SV:4.2MP:2.*) # Unixware is an offshoot of SVR4, but it has its own version # number series starting with 2... # I am not positive that other SVR4 systems won't match this, # I just have to hope. -- rms. # Use sysv4.2uw... so that sysv4* matches it. echo ${UNAME_MACHINE}-pc-sysv4.2uw${UNAME_VERSION} exit ;; i*86:OS/2:*:*) # If we were able to find `uname', then EMX Unix compatibility # is probably installed. echo ${UNAME_MACHINE}-pc-os2-emx exit ;; i*86:XTS-300:*:STOP) echo ${UNAME_MACHINE}-unknown-stop exit ;; i*86:atheos:*:*) echo ${UNAME_MACHINE}-unknown-atheos exit ;; i*86:syllable:*:*) echo ${UNAME_MACHINE}-pc-syllable exit ;; i*86:LynxOS:2.*:* | i*86:LynxOS:3.[01]*:* | i*86:LynxOS:4.[02]*:*) echo i386-unknown-lynxos${UNAME_RELEASE} exit ;; i*86:*DOS:*:*) echo ${UNAME_MACHINE}-pc-msdosdjgpp exit ;; i*86:*:4.*:* | i*86:SYSTEM_V:4.*:*) UNAME_REL=`echo ${UNAME_RELEASE} | sed 's/\/MP$//'` if grep Novell /usr/include/link.h >/dev/null 2>/dev/null; then echo ${UNAME_MACHINE}-univel-sysv${UNAME_REL} else echo ${UNAME_MACHINE}-pc-sysv${UNAME_REL} fi exit ;; i*86:*:5:[678]*) # UnixWare 7.x, OpenUNIX and OpenServer 6. case `/bin/uname -X | grep "^Machine"` in *486*) UNAME_MACHINE=i486 ;; *Pentium) UNAME_MACHINE=i586 ;; *Pent*|*Celeron) UNAME_MACHINE=i686 ;; esac echo ${UNAME_MACHINE}-unknown-sysv${UNAME_RELEASE}${UNAME_SYSTEM}${UNAME_VERSION} exit ;; i*86:*:3.2:*) if test -f /usr/options/cb.name; then UNAME_REL=`sed -n 's/.*Version //p' /dev/null >/dev/null ; then UNAME_REL=`(/bin/uname -X|grep Release|sed -e 's/.*= //')` (/bin/uname -X|grep i80486 >/dev/null) && UNAME_MACHINE=i486 (/bin/uname -X|grep '^Machine.*Pentium' >/dev/null) \ && UNAME_MACHINE=i586 (/bin/uname -X|grep '^Machine.*Pent *II' >/dev/null) \ && UNAME_MACHINE=i686 (/bin/uname -X|grep '^Machine.*Pentium Pro' >/dev/null) \ && UNAME_MACHINE=i686 echo ${UNAME_MACHINE}-pc-sco$UNAME_REL else echo ${UNAME_MACHINE}-pc-sysv32 fi exit ;; pc:*:*:*) # Left here for compatibility: # uname -m prints for DJGPP always 'pc', but it prints nothing about # the processor, so we play safe by assuming i586. # Note: whatever this is, it MUST be the same as what config.sub # prints for the "djgpp" host, or else GDB configury will decide that # this is a cross-build. echo i586-pc-msdosdjgpp exit ;; Intel:Mach:3*:*) echo i386-pc-mach3 exit ;; paragon:*:*:*) echo i860-intel-osf1 exit ;; i860:*:4.*:*) # i860-SVR4 if grep Stardent /usr/include/sys/uadmin.h >/dev/null 2>&1 ; then echo i860-stardent-sysv${UNAME_RELEASE} # Stardent Vistra i860-SVR4 else # Add other i860-SVR4 vendors below as they are discovered. echo i860-unknown-sysv${UNAME_RELEASE} # Unknown i860-SVR4 fi exit ;; mini*:CTIX:SYS*5:*) # "miniframe" echo m68010-convergent-sysv exit ;; mc68k:UNIX:SYSTEM5:3.51m) echo m68k-convergent-sysv exit ;; M680?0:D-NIX:5.3:*) echo m68k-diab-dnix exit ;; M68*:*:R3V[5678]*:*) test -r /sysV68 && { echo 'm68k-motorola-sysv'; exit; } ;; 3[345]??:*:4.0:3.0 | 3[34]??A:*:4.0:3.0 | 3[34]??,*:*:4.0:3.0 | 3[34]??/*:*:4.0:3.0 | 4400:*:4.0:3.0 | 4850:*:4.0:3.0 | SKA40:*:4.0:3.0 | SDS2:*:4.0:3.0 | SHG2:*:4.0:3.0 | S7501*:*:4.0:3.0) OS_REL='' test -r /etc/.relid \ && OS_REL=.`sed -n 's/[^ ]* [^ ]* \([0-9][0-9]\).*/\1/p' < /etc/.relid` /bin/uname -p 2>/dev/null | grep 86 >/dev/null \ && { echo i486-ncr-sysv4.3${OS_REL}; exit; } /bin/uname -p 2>/dev/null | /bin/grep entium >/dev/null \ && { echo i586-ncr-sysv4.3${OS_REL}; exit; } ;; 3[34]??:*:4.0:* | 3[34]??,*:*:4.0:*) /bin/uname -p 2>/dev/null | grep 86 >/dev/null \ && { echo i486-ncr-sysv4; exit; } ;; NCR*:*:4.2:* | MPRAS*:*:4.2:*) OS_REL='.3' test -r /etc/.relid \ && OS_REL=.`sed -n 's/[^ ]* [^ ]* \([0-9][0-9]\).*/\1/p' < /etc/.relid` /bin/uname -p 2>/dev/null | grep 86 >/dev/null \ && { echo i486-ncr-sysv4.3${OS_REL}; exit; } /bin/uname -p 2>/dev/null | /bin/grep entium >/dev/null \ && { echo i586-ncr-sysv4.3${OS_REL}; exit; } /bin/uname -p 2>/dev/null | /bin/grep pteron >/dev/null \ && { echo i586-ncr-sysv4.3${OS_REL}; exit; } ;; m68*:LynxOS:2.*:* | m68*:LynxOS:3.0*:*) echo m68k-unknown-lynxos${UNAME_RELEASE} exit ;; mc68030:UNIX_System_V:4.*:*) echo m68k-atari-sysv4 exit ;; TSUNAMI:LynxOS:2.*:*) echo sparc-unknown-lynxos${UNAME_RELEASE} exit ;; rs6000:LynxOS:2.*:*) echo rs6000-unknown-lynxos${UNAME_RELEASE} exit ;; PowerPC:LynxOS:2.*:* | PowerPC:LynxOS:3.[01]*:* | PowerPC:LynxOS:4.[02]*:*) echo powerpc-unknown-lynxos${UNAME_RELEASE} exit ;; SM[BE]S:UNIX_SV:*:*) echo mips-dde-sysv${UNAME_RELEASE} exit ;; RM*:ReliantUNIX-*:*:*) echo mips-sni-sysv4 exit ;; RM*:SINIX-*:*:*) echo mips-sni-sysv4 exit ;; *:SINIX-*:*:*) if uname -p 2>/dev/null >/dev/null ; then UNAME_MACHINE=`(uname -p) 2>/dev/null` echo ${UNAME_MACHINE}-sni-sysv4 else echo ns32k-sni-sysv fi exit ;; PENTIUM:*:4.0*:*) # Unisys `ClearPath HMP IX 4000' SVR4/MP effort # says echo i586-unisys-sysv4 exit ;; *:UNIX_System_V:4*:FTX*) # From Gerald Hewes . # How about differentiating between stratus architectures? -djm echo hppa1.1-stratus-sysv4 exit ;; *:*:*:FTX*) # From seanf@swdc.stratus.com. echo i860-stratus-sysv4 exit ;; i*86:VOS:*:*) # From Paul.Green@stratus.com. echo ${UNAME_MACHINE}-stratus-vos exit ;; *:VOS:*:*) # From Paul.Green@stratus.com. echo hppa1.1-stratus-vos exit ;; mc68*:A/UX:*:*) echo m68k-apple-aux${UNAME_RELEASE} exit ;; news*:NEWS-OS:6*:*) echo mips-sony-newsos6 exit ;; R[34]000:*System_V*:*:* | R4000:UNIX_SYSV:*:* | R*000:UNIX_SV:*:*) if [ -d /usr/nec ]; then echo mips-nec-sysv${UNAME_RELEASE} else echo mips-unknown-sysv${UNAME_RELEASE} fi exit ;; BeBox:BeOS:*:*) # BeOS running on hardware made by Be, PPC only. echo powerpc-be-beos exit ;; BeMac:BeOS:*:*) # BeOS running on Mac or Mac clone, PPC only. echo powerpc-apple-beos exit ;; BePC:BeOS:*:*) # BeOS running on Intel PC compatible. echo i586-pc-beos exit ;; BePC:Haiku:*:*) # Haiku running on Intel PC compatible. echo i586-pc-haiku exit ;; SX-4:SUPER-UX:*:*) echo sx4-nec-superux${UNAME_RELEASE} exit ;; SX-5:SUPER-UX:*:*) echo sx5-nec-superux${UNAME_RELEASE} exit ;; SX-6:SUPER-UX:*:*) echo sx6-nec-superux${UNAME_RELEASE} exit ;; SX-7:SUPER-UX:*:*) echo sx7-nec-superux${UNAME_RELEASE} exit ;; SX-8:SUPER-UX:*:*) echo sx8-nec-superux${UNAME_RELEASE} exit ;; SX-8R:SUPER-UX:*:*) echo sx8r-nec-superux${UNAME_RELEASE} exit ;; Power*:Rhapsody:*:*) echo powerpc-apple-rhapsody${UNAME_RELEASE} exit ;; *:Rhapsody:*:*) echo ${UNAME_MACHINE}-apple-rhapsody${UNAME_RELEASE} exit ;; *:Darwin:*:*) UNAME_PROCESSOR=`uname -p` || UNAME_PROCESSOR=unknown case $UNAME_PROCESSOR in i386) eval $set_cc_for_build if [ "$CC_FOR_BUILD" != 'no_compiler_found' ]; then if (echo '#ifdef __LP64__'; echo IS_64BIT_ARCH; echo '#endif') | \ (CCOPTS= $CC_FOR_BUILD -E - 2>/dev/null) | \ grep IS_64BIT_ARCH >/dev/null then UNAME_PROCESSOR="x86_64" fi fi ;; unknown) UNAME_PROCESSOR=powerpc ;; esac echo ${UNAME_PROCESSOR}-apple-darwin${UNAME_RELEASE} exit ;; *:procnto*:*:* | *:QNX:[0123456789]*:*) UNAME_PROCESSOR=`uname -p` if test "$UNAME_PROCESSOR" = "x86"; then UNAME_PROCESSOR=i386 UNAME_MACHINE=pc fi echo ${UNAME_PROCESSOR}-${UNAME_MACHINE}-nto-qnx${UNAME_RELEASE} exit ;; *:QNX:*:4*) echo i386-pc-qnx exit ;; NEO-?:NONSTOP_KERNEL:*:*) echo neo-tandem-nsk${UNAME_RELEASE} exit ;; NSE-?:NONSTOP_KERNEL:*:*) echo nse-tandem-nsk${UNAME_RELEASE} exit ;; NSR-?:NONSTOP_KERNEL:*:*) echo nsr-tandem-nsk${UNAME_RELEASE} exit ;; *:NonStop-UX:*:*) echo mips-compaq-nonstopux exit ;; BS2000:POSIX*:*:*) echo bs2000-siemens-sysv exit ;; DS/*:UNIX_System_V:*:*) echo ${UNAME_MACHINE}-${UNAME_SYSTEM}-${UNAME_RELEASE} exit ;; *:Plan9:*:*) # "uname -m" is not consistent, so use $cputype instead. 386 # is converted to i386 for consistency with other x86 # operating systems. if test "$cputype" = "386"; then UNAME_MACHINE=i386 else UNAME_MACHINE="$cputype" fi echo ${UNAME_MACHINE}-unknown-plan9 exit ;; *:TOPS-10:*:*) echo pdp10-unknown-tops10 exit ;; *:TENEX:*:*) echo pdp10-unknown-tenex exit ;; KS10:TOPS-20:*:* | KL10:TOPS-20:*:* | TYPE4:TOPS-20:*:*) echo pdp10-dec-tops20 exit ;; XKL-1:TOPS-20:*:* | TYPE5:TOPS-20:*:*) echo pdp10-xkl-tops20 exit ;; *:TOPS-20:*:*) echo pdp10-unknown-tops20 exit ;; *:ITS:*:*) echo pdp10-unknown-its exit ;; SEI:*:*:SEIUX) echo mips-sei-seiux${UNAME_RELEASE} exit ;; *:DragonFly:*:*) echo ${UNAME_MACHINE}-unknown-dragonfly`echo ${UNAME_RELEASE}|sed -e 's/[-(].*//'` exit ;; *:*VMS:*:*) UNAME_MACHINE=`(uname -p) 2>/dev/null` case "${UNAME_MACHINE}" in A*) echo alpha-dec-vms ; exit ;; I*) echo ia64-dec-vms ; exit ;; V*) echo vax-dec-vms ; exit ;; esac ;; *:XENIX:*:SysV) echo i386-pc-xenix exit ;; i*86:skyos:*:*) echo ${UNAME_MACHINE}-pc-skyos`echo ${UNAME_RELEASE}` | sed -e 's/ .*$//' exit ;; i*86:rdos:*:*) echo ${UNAME_MACHINE}-pc-rdos exit ;; i*86:AROS:*:*) echo ${UNAME_MACHINE}-pc-aros exit ;; x86_64:VMkernel:*:*) echo ${UNAME_MACHINE}-unknown-esx exit ;; esac #echo '(No uname command or uname output not recognized.)' 1>&2 #echo "${UNAME_MACHINE}:${UNAME_SYSTEM}:${UNAME_RELEASE}:${UNAME_VERSION}" 1>&2 eval $set_cc_for_build cat >$dummy.c < # include #endif main () { #if defined (sony) #if defined (MIPSEB) /* BFD wants "bsd" instead of "newsos". Perhaps BFD should be changed, I don't know.... */ printf ("mips-sony-bsd\n"); exit (0); #else #include printf ("m68k-sony-newsos%s\n", #ifdef NEWSOS4 "4" #else "" #endif ); exit (0); #endif #endif #if defined (__arm) && defined (__acorn) && defined (__unix) printf ("arm-acorn-riscix\n"); exit (0); #endif #if defined (hp300) && !defined (hpux) printf ("m68k-hp-bsd\n"); exit (0); #endif #if defined (NeXT) #if !defined (__ARCHITECTURE__) #define __ARCHITECTURE__ "m68k" #endif int version; version=`(hostinfo | sed -n 's/.*NeXT Mach \([0-9]*\).*/\1/p') 2>/dev/null`; if (version < 4) printf ("%s-next-nextstep%d\n", __ARCHITECTURE__, version); else printf ("%s-next-openstep%d\n", __ARCHITECTURE__, version); exit (0); #endif #if defined (MULTIMAX) || defined (n16) #if defined (UMAXV) printf ("ns32k-encore-sysv\n"); exit (0); #else #if defined (CMU) printf ("ns32k-encore-mach\n"); exit (0); #else printf ("ns32k-encore-bsd\n"); exit (0); #endif #endif #endif #if defined (__386BSD__) printf ("i386-pc-bsd\n"); exit (0); #endif #if defined (sequent) #if defined (i386) printf ("i386-sequent-dynix\n"); exit (0); #endif #if defined (ns32000) printf ("ns32k-sequent-dynix\n"); exit (0); #endif #endif #if defined (_SEQUENT_) struct utsname un; uname(&un); if (strncmp(un.version, "V2", 2) == 0) { printf ("i386-sequent-ptx2\n"); exit (0); } if (strncmp(un.version, "V1", 2) == 0) { /* XXX is V1 correct? */ printf ("i386-sequent-ptx1\n"); exit (0); } printf ("i386-sequent-ptx\n"); exit (0); #endif #if defined (vax) # if !defined (ultrix) # include # if defined (BSD) # if BSD == 43 printf ("vax-dec-bsd4.3\n"); exit (0); # else # if BSD == 199006 printf ("vax-dec-bsd4.3reno\n"); exit (0); # else printf ("vax-dec-bsd\n"); exit (0); # endif # endif # else printf ("vax-dec-bsd\n"); exit (0); # endif # else printf ("vax-dec-ultrix\n"); exit (0); # endif #endif #if defined (alliant) && defined (i860) printf ("i860-alliant-bsd\n"); exit (0); #endif exit (1); } EOF $CC_FOR_BUILD -o $dummy $dummy.c 2>/dev/null && SYSTEM_NAME=`$dummy` && { echo "$SYSTEM_NAME"; exit; } # Apollos put the system type in the environment. test -d /usr/apollo && { echo ${ISP}-apollo-${SYSTYPE}; exit; } # Convex versions that predate uname can use getsysinfo(1) if [ -x /usr/convex/getsysinfo ] then case `getsysinfo -f cpu_type` in c1*) echo c1-convex-bsd exit ;; c2*) if getsysinfo -f scalar_acc then echo c32-convex-bsd else echo c2-convex-bsd fi exit ;; c34*) echo c34-convex-bsd exit ;; c38*) echo c38-convex-bsd exit ;; c4*) echo c4-convex-bsd exit ;; esac fi cat >&2 < in order to provide the needed information to handle your system. config.guess timestamp = $timestamp uname -m = `(uname -m) 2>/dev/null || echo unknown` uname -r = `(uname -r) 2>/dev/null || echo unknown` uname -s = `(uname -s) 2>/dev/null || echo unknown` uname -v = `(uname -v) 2>/dev/null || echo unknown` /usr/bin/uname -p = `(/usr/bin/uname -p) 2>/dev/null` /bin/uname -X = `(/bin/uname -X) 2>/dev/null` hostinfo = `(hostinfo) 2>/dev/null` /bin/universe = `(/bin/universe) 2>/dev/null` /usr/bin/arch -k = `(/usr/bin/arch -k) 2>/dev/null` /bin/arch = `(/bin/arch) 2>/dev/null` /usr/bin/oslevel = `(/usr/bin/oslevel) 2>/dev/null` /usr/convex/getsysinfo = `(/usr/convex/getsysinfo) 2>/dev/null` UNAME_MACHINE = ${UNAME_MACHINE} UNAME_RELEASE = ${UNAME_RELEASE} UNAME_SYSTEM = ${UNAME_SYSTEM} UNAME_VERSION = ${UNAME_VERSION} EOF exit 1 # Local variables: # eval: (add-hook 'write-file-hooks 'time-stamp) # time-stamp-start: "timestamp='" # time-stamp-format: "%:y-%02m-%02d" # time-stamp-end: "'" # End: fftw-3.3.3/config.sub0000755000175000017500000010532711764422452011411 00000000000000#! /bin/sh # Configuration validation subroutine script. # Copyright (C) 1992, 1993, 1994, 1995, 1996, 1997, 1998, 1999, # 2000, 2001, 2002, 2003, 2004, 2005, 2006, 2007, 2008, 2009, 2010, # 2011, 2012 Free Software Foundation, Inc. timestamp='2012-04-18' # This file is (in principle) common to ALL GNU software. # The presence of a machine in this file suggests that SOME GNU software # can handle that machine. It does not imply ALL GNU software can. # # This file is free software; you can redistribute it and/or modify # it under the terms of the GNU General Public License as published by # the Free Software Foundation; either version 2 of the License, or # (at your option) any later version. # # This program is distributed in the hope that it will be useful, # but WITHOUT ANY WARRANTY; without even the implied warranty of # MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the # GNU General Public License for more details. # # You should have received a copy of the GNU General Public License # along with this program; if not, see . # # As a special exception to the GNU General Public License, if you # distribute this file as part of a program that contains a # configuration script generated by Autoconf, you may include it under # the same distribution terms that you use for the rest of that program. # Please send patches to . Submit a context # diff and a properly formatted GNU ChangeLog entry. # # Configuration subroutine to validate and canonicalize a configuration type. # Supply the specified configuration type as an argument. # If it is invalid, we print an error message on stderr and exit with code 1. # Otherwise, we print the canonical config type on stdout and succeed. # You can get the latest version of this script from: # http://git.savannah.gnu.org/gitweb/?p=config.git;a=blob_plain;f=config.sub;hb=HEAD # This file is supposed to be the same for all GNU packages # and recognize all the CPU types, system types and aliases # that are meaningful with *any* GNU software. # Each package is responsible for reporting which valid configurations # it does not support. The user should be able to distinguish # a failure to support a valid configuration from a meaningless # configuration. # The goal of this file is to map all the various variations of a given # machine specification into a single specification in the form: # CPU_TYPE-MANUFACTURER-OPERATING_SYSTEM # or in some cases, the newer four-part form: # CPU_TYPE-MANUFACTURER-KERNEL-OPERATING_SYSTEM # It is wrong to echo any other type of specification. me=`echo "$0" | sed -e 's,.*/,,'` usage="\ Usage: $0 [OPTION] CPU-MFR-OPSYS $0 [OPTION] ALIAS Canonicalize a configuration name. Operation modes: -h, --help print this help, then exit -t, --time-stamp print date of last modification, then exit -v, --version print version number, then exit Report bugs and patches to ." version="\ GNU config.sub ($timestamp) Copyright (C) 1992, 1993, 1994, 1995, 1996, 1997, 1998, 1999, 2000, 2001, 2002, 2003, 2004, 2005, 2006, 2007, 2008, 2009, 2010, 2011, 2012 Free Software Foundation, Inc. This is free software; see the source for copying conditions. There is NO warranty; not even for MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE." help=" Try \`$me --help' for more information." # Parse command line while test $# -gt 0 ; do case $1 in --time-stamp | --time* | -t ) echo "$timestamp" ; exit ;; --version | -v ) echo "$version" ; exit ;; --help | --h* | -h ) echo "$usage"; exit ;; -- ) # Stop option processing shift; break ;; - ) # Use stdin as input. break ;; -* ) echo "$me: invalid option $1$help" exit 1 ;; *local*) # First pass through any local machine types. echo $1 exit ;; * ) break ;; esac done case $# in 0) echo "$me: missing argument$help" >&2 exit 1;; 1) ;; *) echo "$me: too many arguments$help" >&2 exit 1;; esac # Separate what the user gave into CPU-COMPANY and OS or KERNEL-OS (if any). # Here we must recognize all the valid KERNEL-OS combinations. maybe_os=`echo $1 | sed 's/^\(.*\)-\([^-]*-[^-]*\)$/\2/'` case $maybe_os in nto-qnx* | linux-gnu* | linux-android* | linux-dietlibc | linux-newlib* | \ linux-uclibc* | uclinux-uclibc* | uclinux-gnu* | kfreebsd*-gnu* | \ knetbsd*-gnu* | netbsd*-gnu* | \ kopensolaris*-gnu* | \ storm-chaos* | os2-emx* | rtmk-nova*) os=-$maybe_os basic_machine=`echo $1 | sed 's/^\(.*\)-\([^-]*-[^-]*\)$/\1/'` ;; android-linux) os=-linux-android basic_machine=`echo $1 | sed 's/^\(.*\)-\([^-]*-[^-]*\)$/\1/'`-unknown ;; *) basic_machine=`echo $1 | sed 's/-[^-]*$//'` if [ $basic_machine != $1 ] then os=`echo $1 | sed 's/.*-/-/'` else os=; fi ;; esac ### Let's recognize common machines as not being operating systems so ### that things like config.sub decstation-3100 work. We also ### recognize some manufacturers as not being operating systems, so we ### can provide default operating systems below. case $os in -sun*os*) # Prevent following clause from handling this invalid input. ;; -dec* | -mips* | -sequent* | -encore* | -pc532* | -sgi* | -sony* | \ -att* | -7300* | -3300* | -delta* | -motorola* | -sun[234]* | \ -unicom* | -ibm* | -next | -hp | -isi* | -apollo | -altos* | \ -convergent* | -ncr* | -news | -32* | -3600* | -3100* | -hitachi* |\ -c[123]* | -convex* | -sun | -crds | -omron* | -dg | -ultra | -tti* | \ -harris | -dolphin | -highlevel | -gould | -cbm | -ns | -masscomp | \ -apple | -axis | -knuth | -cray | -microblaze) os= basic_machine=$1 ;; -bluegene*) os=-cnk ;; -sim | -cisco | -oki | -wec | -winbond) os= basic_machine=$1 ;; -scout) ;; -wrs) os=-vxworks basic_machine=$1 ;; -chorusos*) os=-chorusos basic_machine=$1 ;; -chorusrdb) os=-chorusrdb basic_machine=$1 ;; -hiux*) os=-hiuxwe2 ;; -sco6) os=-sco5v6 basic_machine=`echo $1 | sed -e 's/86-.*/86-pc/'` ;; -sco5) os=-sco3.2v5 basic_machine=`echo $1 | sed -e 's/86-.*/86-pc/'` ;; -sco4) os=-sco3.2v4 basic_machine=`echo $1 | sed -e 's/86-.*/86-pc/'` ;; -sco3.2.[4-9]*) os=`echo $os | sed -e 's/sco3.2./sco3.2v/'` basic_machine=`echo $1 | sed -e 's/86-.*/86-pc/'` ;; -sco3.2v[4-9]*) # Don't forget version if it is 3.2v4 or newer. basic_machine=`echo $1 | sed -e 's/86-.*/86-pc/'` ;; -sco5v6*) # Don't forget version if it is 3.2v4 or newer. basic_machine=`echo $1 | sed -e 's/86-.*/86-pc/'` ;; -sco*) os=-sco3.2v2 basic_machine=`echo $1 | sed -e 's/86-.*/86-pc/'` ;; -udk*) basic_machine=`echo $1 | sed -e 's/86-.*/86-pc/'` ;; -isc) os=-isc2.2 basic_machine=`echo $1 | sed -e 's/86-.*/86-pc/'` ;; -clix*) basic_machine=clipper-intergraph ;; -isc*) basic_machine=`echo $1 | sed -e 's/86-.*/86-pc/'` ;; -lynx*178) os=-lynxos178 ;; -lynx*5) os=-lynxos5 ;; -lynx*) os=-lynxos ;; -ptx*) basic_machine=`echo $1 | sed -e 's/86-.*/86-sequent/'` ;; -windowsnt*) os=`echo $os | sed -e 's/windowsnt/winnt/'` ;; -psos*) os=-psos ;; -mint | -mint[0-9]*) basic_machine=m68k-atari os=-mint ;; esac # Decode aliases for certain CPU-COMPANY combinations. case $basic_machine in # Recognize the basic CPU types without company name. # Some are omitted here because they have special meanings below. 1750a | 580 \ | a29k \ | aarch64 | aarch64_be \ | alpha | alphaev[4-8] | alphaev56 | alphaev6[78] | alphapca5[67] \ | alpha64 | alpha64ev[4-8] | alpha64ev56 | alpha64ev6[78] | alpha64pca5[67] \ | am33_2.0 \ | arc | arm | arm[bl]e | arme[lb] | armv[2345] | armv[345][lb] | avr | avr32 \ | be32 | be64 \ | bfin \ | c4x | clipper \ | d10v | d30v | dlx | dsp16xx \ | epiphany \ | fido | fr30 | frv \ | h8300 | h8500 | hppa | hppa1.[01] | hppa2.0 | hppa2.0[nw] | hppa64 \ | hexagon \ | i370 | i860 | i960 | ia64 \ | ip2k | iq2000 \ | le32 | le64 \ | lm32 \ | m32c | m32r | m32rle | m68000 | m68k | m88k \ | maxq | mb | microblaze | mcore | mep | metag \ | mips | mipsbe | mipseb | mipsel | mipsle \ | mips16 \ | mips64 | mips64el \ | mips64octeon | mips64octeonel \ | mips64orion | mips64orionel \ | mips64r5900 | mips64r5900el \ | mips64vr | mips64vrel \ | mips64vr4100 | mips64vr4100el \ | mips64vr4300 | mips64vr4300el \ | mips64vr5000 | mips64vr5000el \ | mips64vr5900 | mips64vr5900el \ | mipsisa32 | mipsisa32el \ | mipsisa32r2 | mipsisa32r2el \ | mipsisa64 | mipsisa64el \ | mipsisa64r2 | mipsisa64r2el \ | mipsisa64sb1 | mipsisa64sb1el \ | mipsisa64sr71k | mipsisa64sr71kel \ | mipstx39 | mipstx39el \ | mn10200 | mn10300 \ | moxie \ | mt \ | msp430 \ | nds32 | nds32le | nds32be \ | nios | nios2 \ | ns16k | ns32k \ | open8 \ | or32 \ | pdp10 | pdp11 | pj | pjl \ | powerpc | powerpc64 | powerpc64le | powerpcle \ | pyramid \ | rl78 | rx \ | score \ | sh | sh[1234] | sh[24]a | sh[24]aeb | sh[23]e | sh[34]eb | sheb | shbe | shle | sh[1234]le | sh3ele \ | sh64 | sh64le \ | sparc | sparc64 | sparc64b | sparc64v | sparc86x | sparclet | sparclite \ | sparcv8 | sparcv9 | sparcv9b | sparcv9v \ | spu \ | tahoe | tic4x | tic54x | tic55x | tic6x | tic80 | tron \ | ubicom32 \ | v850 | v850e | v850e1 | v850e2 | v850es | v850e2v3 \ | we32k \ | x86 | xc16x | xstormy16 | xtensa \ | z8k | z80) basic_machine=$basic_machine-unknown ;; c54x) basic_machine=tic54x-unknown ;; c55x) basic_machine=tic55x-unknown ;; c6x) basic_machine=tic6x-unknown ;; m6811 | m68hc11 | m6812 | m68hc12 | m68hcs12x | picochip) basic_machine=$basic_machine-unknown os=-none ;; m88110 | m680[12346]0 | m683?2 | m68360 | m5200 | v70 | w65 | z8k) ;; ms1) basic_machine=mt-unknown ;; strongarm | thumb | xscale) basic_machine=arm-unknown ;; xgate) basic_machine=$basic_machine-unknown os=-none ;; xscaleeb) basic_machine=armeb-unknown ;; xscaleel) basic_machine=armel-unknown ;; # We use `pc' rather than `unknown' # because (1) that's what they normally are, and # (2) the word "unknown" tends to confuse beginning users. i*86 | x86_64) basic_machine=$basic_machine-pc ;; # Object if more than one company name word. *-*-*) echo Invalid configuration \`$1\': machine \`$basic_machine\' not recognized 1>&2 exit 1 ;; # Recognize the basic CPU types with company name. 580-* \ | a29k-* \ | aarch64-* | aarch64_be-* \ | alpha-* | alphaev[4-8]-* | alphaev56-* | alphaev6[78]-* \ | alpha64-* | alpha64ev[4-8]-* | alpha64ev56-* | alpha64ev6[78]-* \ | alphapca5[67]-* | alpha64pca5[67]-* | arc-* \ | arm-* | armbe-* | armle-* | armeb-* | armv*-* \ | avr-* | avr32-* \ | be32-* | be64-* \ | bfin-* | bs2000-* \ | c[123]* | c30-* | [cjt]90-* | c4x-* \ | clipper-* | craynv-* | cydra-* \ | d10v-* | d30v-* | dlx-* \ | elxsi-* \ | f30[01]-* | f700-* | fido-* | fr30-* | frv-* | fx80-* \ | h8300-* | h8500-* \ | hppa-* | hppa1.[01]-* | hppa2.0-* | hppa2.0[nw]-* | hppa64-* \ | hexagon-* \ | i*86-* | i860-* | i960-* | ia64-* \ | ip2k-* | iq2000-* \ | le32-* | le64-* \ | lm32-* \ | m32c-* | m32r-* | m32rle-* \ | m68000-* | m680[012346]0-* | m68360-* | m683?2-* | m68k-* \ | m88110-* | m88k-* | maxq-* | mcore-* | metag-* | microblaze-* \ | mips-* | mipsbe-* | mipseb-* | mipsel-* | mipsle-* \ | mips16-* \ | mips64-* | mips64el-* \ | mips64octeon-* | mips64octeonel-* \ | mips64orion-* | mips64orionel-* \ | mips64r5900-* | mips64r5900el-* \ | mips64vr-* | mips64vrel-* \ | mips64vr4100-* | mips64vr4100el-* \ | mips64vr4300-* | mips64vr4300el-* \ | mips64vr5000-* | mips64vr5000el-* \ | mips64vr5900-* | mips64vr5900el-* \ | mipsisa32-* | mipsisa32el-* \ | mipsisa32r2-* | mipsisa32r2el-* \ | mipsisa64-* | mipsisa64el-* \ | mipsisa64r2-* | mipsisa64r2el-* \ | mipsisa64sb1-* | mipsisa64sb1el-* \ | mipsisa64sr71k-* | mipsisa64sr71kel-* \ | mipstx39-* | mipstx39el-* \ | mmix-* \ | mt-* \ | msp430-* \ | nds32-* | nds32le-* | nds32be-* \ | nios-* | nios2-* \ | none-* | np1-* | ns16k-* | ns32k-* \ | open8-* \ | orion-* \ | pdp10-* | pdp11-* | pj-* | pjl-* | pn-* | power-* \ | powerpc-* | powerpc64-* | powerpc64le-* | powerpcle-* \ | pyramid-* \ | rl78-* | romp-* | rs6000-* | rx-* \ | sh-* | sh[1234]-* | sh[24]a-* | sh[24]aeb-* | sh[23]e-* | sh[34]eb-* | sheb-* | shbe-* \ | shle-* | sh[1234]le-* | sh3ele-* | sh64-* | sh64le-* \ | sparc-* | sparc64-* | sparc64b-* | sparc64v-* | sparc86x-* | sparclet-* \ | sparclite-* \ | sparcv8-* | sparcv9-* | sparcv9b-* | sparcv9v-* | sv1-* | sx?-* \ | tahoe-* \ | tic30-* | tic4x-* | tic54x-* | tic55x-* | tic6x-* | tic80-* \ | tile*-* \ | tron-* \ | ubicom32-* \ | v850-* | v850e-* | v850e1-* | v850es-* | v850e2-* | v850e2v3-* \ | vax-* \ | we32k-* \ | x86-* | x86_64-* | xc16x-* | xps100-* \ | xstormy16-* | xtensa*-* \ | ymp-* \ | z8k-* | z80-*) ;; # Recognize the basic CPU types without company name, with glob match. xtensa*) basic_machine=$basic_machine-unknown ;; # Recognize the various machine names and aliases which stand # for a CPU type and a company and sometimes even an OS. 386bsd) basic_machine=i386-unknown os=-bsd ;; 3b1 | 7300 | 7300-att | att-7300 | pc7300 | safari | unixpc) basic_machine=m68000-att ;; 3b*) basic_machine=we32k-att ;; a29khif) basic_machine=a29k-amd os=-udi ;; abacus) basic_machine=abacus-unknown ;; adobe68k) basic_machine=m68010-adobe os=-scout ;; alliant | fx80) basic_machine=fx80-alliant ;; altos | altos3068) basic_machine=m68k-altos ;; am29k) basic_machine=a29k-none os=-bsd ;; amd64) basic_machine=x86_64-pc ;; amd64-*) basic_machine=x86_64-`echo $basic_machine | sed 's/^[^-]*-//'` ;; amdahl) basic_machine=580-amdahl os=-sysv ;; amiga | amiga-*) basic_machine=m68k-unknown ;; amigaos | amigados) basic_machine=m68k-unknown os=-amigaos ;; amigaunix | amix) basic_machine=m68k-unknown os=-sysv4 ;; apollo68) basic_machine=m68k-apollo os=-sysv ;; apollo68bsd) basic_machine=m68k-apollo os=-bsd ;; aros) basic_machine=i386-pc os=-aros ;; aux) basic_machine=m68k-apple os=-aux ;; balance) basic_machine=ns32k-sequent os=-dynix ;; blackfin) basic_machine=bfin-unknown os=-linux ;; blackfin-*) basic_machine=bfin-`echo $basic_machine | sed 's/^[^-]*-//'` os=-linux ;; bluegene*) basic_machine=powerpc-ibm os=-cnk ;; c54x-*) basic_machine=tic54x-`echo $basic_machine | sed 's/^[^-]*-//'` ;; c55x-*) basic_machine=tic55x-`echo $basic_machine | sed 's/^[^-]*-//'` ;; c6x-*) basic_machine=tic6x-`echo $basic_machine | sed 's/^[^-]*-//'` ;; c90) basic_machine=c90-cray os=-unicos ;; cegcc) basic_machine=arm-unknown os=-cegcc ;; convex-c1) basic_machine=c1-convex os=-bsd ;; convex-c2) basic_machine=c2-convex os=-bsd ;; convex-c32) basic_machine=c32-convex os=-bsd ;; convex-c34) basic_machine=c34-convex os=-bsd ;; convex-c38) basic_machine=c38-convex os=-bsd ;; cray | j90) basic_machine=j90-cray os=-unicos ;; craynv) basic_machine=craynv-cray os=-unicosmp ;; cr16 | cr16-*) basic_machine=cr16-unknown os=-elf ;; crds | unos) basic_machine=m68k-crds ;; crisv32 | crisv32-* | etraxfs*) basic_machine=crisv32-axis ;; cris | cris-* | etrax*) basic_machine=cris-axis ;; crx) basic_machine=crx-unknown os=-elf ;; da30 | da30-*) basic_machine=m68k-da30 ;; decstation | decstation-3100 | pmax | pmax-* | pmin | dec3100 | decstatn) basic_machine=mips-dec ;; decsystem10* | dec10*) basic_machine=pdp10-dec os=-tops10 ;; decsystem20* | dec20*) basic_machine=pdp10-dec os=-tops20 ;; delta | 3300 | motorola-3300 | motorola-delta \ | 3300-motorola | delta-motorola) basic_machine=m68k-motorola ;; delta88) basic_machine=m88k-motorola os=-sysv3 ;; dicos) basic_machine=i686-pc os=-dicos ;; djgpp) basic_machine=i586-pc os=-msdosdjgpp ;; dpx20 | dpx20-*) basic_machine=rs6000-bull os=-bosx ;; dpx2* | dpx2*-bull) basic_machine=m68k-bull os=-sysv3 ;; ebmon29k) basic_machine=a29k-amd os=-ebmon ;; elxsi) basic_machine=elxsi-elxsi os=-bsd ;; encore | umax | mmax) basic_machine=ns32k-encore ;; es1800 | OSE68k | ose68k | ose | OSE) basic_machine=m68k-ericsson os=-ose ;; fx2800) basic_machine=i860-alliant ;; genix) basic_machine=ns32k-ns ;; gmicro) basic_machine=tron-gmicro os=-sysv ;; go32) basic_machine=i386-pc os=-go32 ;; h3050r* | hiux*) basic_machine=hppa1.1-hitachi os=-hiuxwe2 ;; h8300hms) basic_machine=h8300-hitachi os=-hms ;; h8300xray) basic_machine=h8300-hitachi os=-xray ;; h8500hms) basic_machine=h8500-hitachi os=-hms ;; harris) basic_machine=m88k-harris os=-sysv3 ;; hp300-*) basic_machine=m68k-hp ;; hp300bsd) basic_machine=m68k-hp os=-bsd ;; hp300hpux) basic_machine=m68k-hp os=-hpux ;; hp3k9[0-9][0-9] | hp9[0-9][0-9]) basic_machine=hppa1.0-hp ;; hp9k2[0-9][0-9] | hp9k31[0-9]) basic_machine=m68000-hp ;; hp9k3[2-9][0-9]) basic_machine=m68k-hp ;; hp9k6[0-9][0-9] | hp6[0-9][0-9]) basic_machine=hppa1.0-hp ;; hp9k7[0-79][0-9] | hp7[0-79][0-9]) basic_machine=hppa1.1-hp ;; hp9k78[0-9] | hp78[0-9]) # FIXME: really hppa2.0-hp basic_machine=hppa1.1-hp ;; hp9k8[67]1 | hp8[67]1 | hp9k80[24] | hp80[24] | hp9k8[78]9 | hp8[78]9 | hp9k893 | hp893) # FIXME: really hppa2.0-hp basic_machine=hppa1.1-hp ;; hp9k8[0-9][13679] | hp8[0-9][13679]) basic_machine=hppa1.1-hp ;; hp9k8[0-9][0-9] | hp8[0-9][0-9]) basic_machine=hppa1.0-hp ;; hppa-next) os=-nextstep3 ;; hppaosf) basic_machine=hppa1.1-hp os=-osf ;; hppro) basic_machine=hppa1.1-hp os=-proelf ;; i370-ibm* | ibm*) basic_machine=i370-ibm ;; i*86v32) basic_machine=`echo $1 | sed -e 's/86.*/86-pc/'` os=-sysv32 ;; i*86v4*) basic_machine=`echo $1 | sed -e 's/86.*/86-pc/'` os=-sysv4 ;; i*86v) basic_machine=`echo $1 | sed -e 's/86.*/86-pc/'` os=-sysv ;; i*86sol2) basic_machine=`echo $1 | sed -e 's/86.*/86-pc/'` os=-solaris2 ;; i386mach) basic_machine=i386-mach os=-mach ;; i386-vsta | vsta) basic_machine=i386-unknown os=-vsta ;; iris | iris4d) basic_machine=mips-sgi case $os in -irix*) ;; *) os=-irix4 ;; esac ;; isi68 | isi) basic_machine=m68k-isi os=-sysv ;; m68knommu) basic_machine=m68k-unknown os=-linux ;; m68knommu-*) basic_machine=m68k-`echo $basic_machine | sed 's/^[^-]*-//'` os=-linux ;; m88k-omron*) basic_machine=m88k-omron ;; magnum | m3230) basic_machine=mips-mips os=-sysv ;; merlin) basic_machine=ns32k-utek os=-sysv ;; microblaze) basic_machine=microblaze-xilinx ;; mingw32) basic_machine=i386-pc os=-mingw32 ;; mingw32ce) basic_machine=arm-unknown os=-mingw32ce ;; miniframe) basic_machine=m68000-convergent ;; *mint | -mint[0-9]* | *MiNT | *MiNT[0-9]*) basic_machine=m68k-atari os=-mint ;; mips3*-*) basic_machine=`echo $basic_machine | sed -e 's/mips3/mips64/'` ;; mips3*) basic_machine=`echo $basic_machine | sed -e 's/mips3/mips64/'`-unknown ;; monitor) basic_machine=m68k-rom68k os=-coff ;; morphos) basic_machine=powerpc-unknown os=-morphos ;; msdos) basic_machine=i386-pc os=-msdos ;; ms1-*) basic_machine=`echo $basic_machine | sed -e 's/ms1-/mt-/'` ;; msys) basic_machine=i386-pc os=-msys ;; mvs) basic_machine=i370-ibm os=-mvs ;; nacl) basic_machine=le32-unknown os=-nacl ;; ncr3000) basic_machine=i486-ncr os=-sysv4 ;; netbsd386) basic_machine=i386-unknown os=-netbsd ;; netwinder) basic_machine=armv4l-rebel os=-linux ;; news | news700 | news800 | news900) basic_machine=m68k-sony os=-newsos ;; news1000) basic_machine=m68030-sony os=-newsos ;; news-3600 | risc-news) basic_machine=mips-sony os=-newsos ;; necv70) basic_machine=v70-nec os=-sysv ;; next | m*-next ) basic_machine=m68k-next case $os in -nextstep* ) ;; -ns2*) os=-nextstep2 ;; *) os=-nextstep3 ;; esac ;; nh3000) basic_machine=m68k-harris os=-cxux ;; nh[45]000) basic_machine=m88k-harris os=-cxux ;; nindy960) basic_machine=i960-intel os=-nindy ;; mon960) basic_machine=i960-intel os=-mon960 ;; nonstopux) basic_machine=mips-compaq os=-nonstopux ;; np1) basic_machine=np1-gould ;; neo-tandem) basic_machine=neo-tandem ;; nse-tandem) basic_machine=nse-tandem ;; nsr-tandem) basic_machine=nsr-tandem ;; op50n-* | op60c-*) basic_machine=hppa1.1-oki os=-proelf ;; openrisc | openrisc-*) basic_machine=or32-unknown ;; os400) basic_machine=powerpc-ibm os=-os400 ;; OSE68000 | ose68000) basic_machine=m68000-ericsson os=-ose ;; os68k) basic_machine=m68k-none os=-os68k ;; pa-hitachi) basic_machine=hppa1.1-hitachi os=-hiuxwe2 ;; paragon) basic_machine=i860-intel os=-osf ;; parisc) basic_machine=hppa-unknown os=-linux ;; parisc-*) basic_machine=hppa-`echo $basic_machine | sed 's/^[^-]*-//'` os=-linux ;; pbd) basic_machine=sparc-tti ;; pbb) basic_machine=m68k-tti ;; pc532 | pc532-*) basic_machine=ns32k-pc532 ;; pc98) basic_machine=i386-pc ;; pc98-*) basic_machine=i386-`echo $basic_machine | sed 's/^[^-]*-//'` ;; pentium | p5 | k5 | k6 | nexgen | viac3) basic_machine=i586-pc ;; pentiumpro | p6 | 6x86 | athlon | athlon_*) basic_machine=i686-pc ;; pentiumii | pentium2 | pentiumiii | pentium3) basic_machine=i686-pc ;; pentium4) basic_machine=i786-pc ;; pentium-* | p5-* | k5-* | k6-* | nexgen-* | viac3-*) basic_machine=i586-`echo $basic_machine | sed 's/^[^-]*-//'` ;; pentiumpro-* | p6-* | 6x86-* | athlon-*) basic_machine=i686-`echo $basic_machine | sed 's/^[^-]*-//'` ;; pentiumii-* | pentium2-* | pentiumiii-* | pentium3-*) basic_machine=i686-`echo $basic_machine | sed 's/^[^-]*-//'` ;; pentium4-*) basic_machine=i786-`echo $basic_machine | sed 's/^[^-]*-//'` ;; pn) basic_machine=pn-gould ;; power) basic_machine=power-ibm ;; ppc | ppcbe) basic_machine=powerpc-unknown ;; ppc-* | ppcbe-*) basic_machine=powerpc-`echo $basic_machine | sed 's/^[^-]*-//'` ;; ppcle | powerpclittle | ppc-le | powerpc-little) basic_machine=powerpcle-unknown ;; ppcle-* | powerpclittle-*) basic_machine=powerpcle-`echo $basic_machine | sed 's/^[^-]*-//'` ;; ppc64) basic_machine=powerpc64-unknown ;; ppc64-*) basic_machine=powerpc64-`echo $basic_machine | sed 's/^[^-]*-//'` ;; ppc64le | powerpc64little | ppc64-le | powerpc64-little) basic_machine=powerpc64le-unknown ;; ppc64le-* | powerpc64little-*) basic_machine=powerpc64le-`echo $basic_machine | sed 's/^[^-]*-//'` ;; ps2) basic_machine=i386-ibm ;; pw32) basic_machine=i586-unknown os=-pw32 ;; rdos) basic_machine=i386-pc os=-rdos ;; rom68k) basic_machine=m68k-rom68k os=-coff ;; rm[46]00) basic_machine=mips-siemens ;; rtpc | rtpc-*) basic_machine=romp-ibm ;; s390 | s390-*) basic_machine=s390-ibm ;; s390x | s390x-*) basic_machine=s390x-ibm ;; sa29200) basic_machine=a29k-amd os=-udi ;; sb1) basic_machine=mipsisa64sb1-unknown ;; sb1el) basic_machine=mipsisa64sb1el-unknown ;; sde) basic_machine=mipsisa32-sde os=-elf ;; sei) basic_machine=mips-sei os=-seiux ;; sequent) basic_machine=i386-sequent ;; sh) basic_machine=sh-hitachi os=-hms ;; sh5el) basic_machine=sh5le-unknown ;; sh64) basic_machine=sh64-unknown ;; sparclite-wrs | simso-wrs) basic_machine=sparclite-wrs os=-vxworks ;; sps7) basic_machine=m68k-bull os=-sysv2 ;; spur) basic_machine=spur-unknown ;; st2000) basic_machine=m68k-tandem ;; stratus) basic_machine=i860-stratus os=-sysv4 ;; strongarm-* | thumb-*) basic_machine=arm-`echo $basic_machine | sed 's/^[^-]*-//'` ;; sun2) basic_machine=m68000-sun ;; sun2os3) basic_machine=m68000-sun os=-sunos3 ;; sun2os4) basic_machine=m68000-sun os=-sunos4 ;; sun3os3) basic_machine=m68k-sun os=-sunos3 ;; sun3os4) basic_machine=m68k-sun os=-sunos4 ;; sun4os3) basic_machine=sparc-sun os=-sunos3 ;; sun4os4) basic_machine=sparc-sun os=-sunos4 ;; sun4sol2) basic_machine=sparc-sun os=-solaris2 ;; sun3 | sun3-*) basic_machine=m68k-sun ;; sun4) basic_machine=sparc-sun ;; sun386 | sun386i | roadrunner) basic_machine=i386-sun ;; sv1) basic_machine=sv1-cray os=-unicos ;; symmetry) basic_machine=i386-sequent os=-dynix ;; t3e) basic_machine=alphaev5-cray os=-unicos ;; t90) basic_machine=t90-cray os=-unicos ;; tile*) basic_machine=$basic_machine-unknown os=-linux-gnu ;; tx39) basic_machine=mipstx39-unknown ;; tx39el) basic_machine=mipstx39el-unknown ;; toad1) basic_machine=pdp10-xkl os=-tops20 ;; tower | tower-32) basic_machine=m68k-ncr ;; tpf) basic_machine=s390x-ibm os=-tpf ;; udi29k) basic_machine=a29k-amd os=-udi ;; ultra3) basic_machine=a29k-nyu os=-sym1 ;; v810 | necv810) basic_machine=v810-nec os=-none ;; vaxv) basic_machine=vax-dec os=-sysv ;; vms) basic_machine=vax-dec os=-vms ;; vpp*|vx|vx-*) basic_machine=f301-fujitsu ;; vxworks960) basic_machine=i960-wrs os=-vxworks ;; vxworks68) basic_machine=m68k-wrs os=-vxworks ;; vxworks29k) basic_machine=a29k-wrs os=-vxworks ;; w65*) basic_machine=w65-wdc os=-none ;; w89k-*) basic_machine=hppa1.1-winbond os=-proelf ;; xbox) basic_machine=i686-pc os=-mingw32 ;; xps | xps100) basic_machine=xps100-honeywell ;; xscale-* | xscalee[bl]-*) basic_machine=`echo $basic_machine | sed 's/^xscale/arm/'` ;; ymp) basic_machine=ymp-cray os=-unicos ;; z8k-*-coff) basic_machine=z8k-unknown os=-sim ;; z80-*-coff) basic_machine=z80-unknown os=-sim ;; none) basic_machine=none-none os=-none ;; # Here we handle the default manufacturer of certain CPU types. It is in # some cases the only manufacturer, in others, it is the most popular. w89k) basic_machine=hppa1.1-winbond ;; op50n) basic_machine=hppa1.1-oki ;; op60c) basic_machine=hppa1.1-oki ;; romp) basic_machine=romp-ibm ;; mmix) basic_machine=mmix-knuth ;; rs6000) basic_machine=rs6000-ibm ;; vax) basic_machine=vax-dec ;; pdp10) # there are many clones, so DEC is not a safe bet basic_machine=pdp10-unknown ;; pdp11) basic_machine=pdp11-dec ;; we32k) basic_machine=we32k-att ;; sh[1234] | sh[24]a | sh[24]aeb | sh[34]eb | sh[1234]le | sh[23]ele) basic_machine=sh-unknown ;; sparc | sparcv8 | sparcv9 | sparcv9b | sparcv9v) basic_machine=sparc-sun ;; cydra) basic_machine=cydra-cydrome ;; orion) basic_machine=orion-highlevel ;; orion105) basic_machine=clipper-highlevel ;; mac | mpw | mac-mpw) basic_machine=m68k-apple ;; pmac | pmac-mpw) basic_machine=powerpc-apple ;; *-unknown) # Make sure to match an already-canonicalized machine name. ;; *) echo Invalid configuration \`$1\': machine \`$basic_machine\' not recognized 1>&2 exit 1 ;; esac # Here we canonicalize certain aliases for manufacturers. case $basic_machine in *-digital*) basic_machine=`echo $basic_machine | sed 's/digital.*/dec/'` ;; *-commodore*) basic_machine=`echo $basic_machine | sed 's/commodore.*/cbm/'` ;; *) ;; esac # Decode manufacturer-specific aliases for certain operating systems. if [ x"$os" != x"" ] then case $os in # First match some system type aliases # that might get confused with valid system types. # -solaris* is a basic system type, with this one exception. -auroraux) os=-auroraux ;; -solaris1 | -solaris1.*) os=`echo $os | sed -e 's|solaris1|sunos4|'` ;; -solaris) os=-solaris2 ;; -svr4*) os=-sysv4 ;; -unixware*) os=-sysv4.2uw ;; -gnu/linux*) os=`echo $os | sed -e 's|gnu/linux|linux-gnu|'` ;; # First accept the basic system types. # The portable systems comes first. # Each alternative MUST END IN A *, to match a version number. # -sysv* is not here because it comes later, after sysvr4. -gnu* | -bsd* | -mach* | -minix* | -genix* | -ultrix* | -irix* \ | -*vms* | -sco* | -esix* | -isc* | -aix* | -cnk* | -sunos | -sunos[34]*\ | -hpux* | -unos* | -osf* | -luna* | -dgux* | -auroraux* | -solaris* \ | -sym* | -kopensolaris* \ | -amigaos* | -amigados* | -msdos* | -newsos* | -unicos* | -aof* \ | -aos* | -aros* \ | -nindy* | -vxsim* | -vxworks* | -ebmon* | -hms* | -mvs* \ | -clix* | -riscos* | -uniplus* | -iris* | -rtu* | -xenix* \ | -hiux* | -386bsd* | -knetbsd* | -mirbsd* | -netbsd* \ | -openbsd* | -solidbsd* \ | -ekkobsd* | -kfreebsd* | -freebsd* | -riscix* | -lynxos* \ | -bosx* | -nextstep* | -cxux* | -aout* | -elf* | -oabi* \ | -ptx* | -coff* | -ecoff* | -winnt* | -domain* | -vsta* \ | -udi* | -eabi* | -lites* | -ieee* | -go32* | -aux* \ | -chorusos* | -chorusrdb* | -cegcc* \ | -cygwin* | -msys* | -pe* | -psos* | -moss* | -proelf* | -rtems* \ | -mingw32* | -linux-gnu* | -linux-android* \ | -linux-newlib* | -linux-uclibc* \ | -uxpv* | -beos* | -mpeix* | -udk* \ | -interix* | -uwin* | -mks* | -rhapsody* | -darwin* | -opened* \ | -openstep* | -oskit* | -conix* | -pw32* | -nonstopux* \ | -storm-chaos* | -tops10* | -tenex* | -tops20* | -its* \ | -os2* | -vos* | -palmos* | -uclinux* | -nucleus* \ | -morphos* | -superux* | -rtmk* | -rtmk-nova* | -windiss* \ | -powermax* | -dnix* | -nx6 | -nx7 | -sei* | -dragonfly* \ | -skyos* | -haiku* | -rdos* | -toppers* | -drops* | -es*) # Remember, each alternative MUST END IN *, to match a version number. ;; -qnx*) case $basic_machine in x86-* | i*86-*) ;; *) os=-nto$os ;; esac ;; -nto-qnx*) ;; -nto*) os=`echo $os | sed -e 's|nto|nto-qnx|'` ;; -sim | -es1800* | -hms* | -xray | -os68k* | -none* | -v88r* \ | -windows* | -osx | -abug | -netware* | -os9* | -beos* | -haiku* \ | -macos* | -mpw* | -magic* | -mmixware* | -mon960* | -lnews*) ;; -mac*) os=`echo $os | sed -e 's|mac|macos|'` ;; -linux-dietlibc) os=-linux-dietlibc ;; -linux*) os=`echo $os | sed -e 's|linux|linux-gnu|'` ;; -sunos5*) os=`echo $os | sed -e 's|sunos5|solaris2|'` ;; -sunos6*) os=`echo $os | sed -e 's|sunos6|solaris3|'` ;; -opened*) os=-openedition ;; -os400*) os=-os400 ;; -wince*) os=-wince ;; -osfrose*) os=-osfrose ;; -osf*) os=-osf ;; -utek*) os=-bsd ;; -dynix*) os=-bsd ;; -acis*) os=-aos ;; -atheos*) os=-atheos ;; -syllable*) os=-syllable ;; -386bsd) os=-bsd ;; -ctix* | -uts*) os=-sysv ;; -nova*) os=-rtmk-nova ;; -ns2 ) os=-nextstep2 ;; -nsk*) os=-nsk ;; # Preserve the version number of sinix5. -sinix5.*) os=`echo $os | sed -e 's|sinix|sysv|'` ;; -sinix*) os=-sysv4 ;; -tpf*) os=-tpf ;; -triton*) os=-sysv3 ;; -oss*) os=-sysv3 ;; -svr4) os=-sysv4 ;; -svr3) os=-sysv3 ;; -sysvr4) os=-sysv4 ;; # This must come after -sysvr4. -sysv*) ;; -ose*) os=-ose ;; -es1800*) os=-ose ;; -xenix) os=-xenix ;; -*mint | -mint[0-9]* | -*MiNT | -MiNT[0-9]*) os=-mint ;; -aros*) os=-aros ;; -kaos*) os=-kaos ;; -zvmoe) os=-zvmoe ;; -dicos*) os=-dicos ;; -nacl*) ;; -none) ;; *) # Get rid of the `-' at the beginning of $os. os=`echo $os | sed 's/[^-]*-//'` echo Invalid configuration \`$1\': system \`$os\' not recognized 1>&2 exit 1 ;; esac else # Here we handle the default operating systems that come with various machines. # The value should be what the vendor currently ships out the door with their # machine or put another way, the most popular os provided with the machine. # Note that if you're going to try to match "-MANUFACTURER" here (say, # "-sun"), then you have to tell the case statement up towards the top # that MANUFACTURER isn't an operating system. Otherwise, code above # will signal an error saying that MANUFACTURER isn't an operating # system, and we'll never get to this point. case $basic_machine in score-*) os=-elf ;; spu-*) os=-elf ;; *-acorn) os=-riscix1.2 ;; arm*-rebel) os=-linux ;; arm*-semi) os=-aout ;; c4x-* | tic4x-*) os=-coff ;; hexagon-*) os=-elf ;; tic54x-*) os=-coff ;; tic55x-*) os=-coff ;; tic6x-*) os=-coff ;; # This must come before the *-dec entry. pdp10-*) os=-tops20 ;; pdp11-*) os=-none ;; *-dec | vax-*) os=-ultrix4.2 ;; m68*-apollo) os=-domain ;; i386-sun) os=-sunos4.0.2 ;; m68000-sun) os=-sunos3 ;; m68*-cisco) os=-aout ;; mep-*) os=-elf ;; mips*-cisco) os=-elf ;; mips*-*) os=-elf ;; or32-*) os=-coff ;; *-tti) # must be before sparc entry or we get the wrong os. os=-sysv3 ;; sparc-* | *-sun) os=-sunos4.1.1 ;; *-be) os=-beos ;; *-haiku) os=-haiku ;; *-ibm) os=-aix ;; *-knuth) os=-mmixware ;; *-wec) os=-proelf ;; *-winbond) os=-proelf ;; *-oki) os=-proelf ;; *-hp) os=-hpux ;; *-hitachi) os=-hiux ;; i860-* | *-att | *-ncr | *-altos | *-motorola | *-convergent) os=-sysv ;; *-cbm) os=-amigaos ;; *-dg) os=-dgux ;; *-dolphin) os=-sysv3 ;; m68k-ccur) os=-rtu ;; m88k-omron*) os=-luna ;; *-next ) os=-nextstep ;; *-sequent) os=-ptx ;; *-crds) os=-unos ;; *-ns) os=-genix ;; i370-*) os=-mvs ;; *-next) os=-nextstep3 ;; *-gould) os=-sysv ;; *-highlevel) os=-bsd ;; *-encore) os=-bsd ;; *-sgi) os=-irix ;; *-siemens) os=-sysv4 ;; *-masscomp) os=-rtu ;; f30[01]-fujitsu | f700-fujitsu) os=-uxpv ;; *-rom68k) os=-coff ;; *-*bug) os=-coff ;; *-apple) os=-macos ;; *-atari*) os=-mint ;; *) os=-none ;; esac fi # Here we handle the case where we know the os, and the CPU type, but not the # manufacturer. We pick the logical manufacturer. vendor=unknown case $basic_machine in *-unknown) case $os in -riscix*) vendor=acorn ;; -sunos*) vendor=sun ;; -cnk*|-aix*) vendor=ibm ;; -beos*) vendor=be ;; -hpux*) vendor=hp ;; -mpeix*) vendor=hp ;; -hiux*) vendor=hitachi ;; -unos*) vendor=crds ;; -dgux*) vendor=dg ;; -luna*) vendor=omron ;; -genix*) vendor=ns ;; -mvs* | -opened*) vendor=ibm ;; -os400*) vendor=ibm ;; -ptx*) vendor=sequent ;; -tpf*) vendor=ibm ;; -vxsim* | -vxworks* | -windiss*) vendor=wrs ;; -aux*) vendor=apple ;; -hms*) vendor=hitachi ;; -mpw* | -macos*) vendor=apple ;; -*mint | -mint[0-9]* | -*MiNT | -MiNT[0-9]*) vendor=atari ;; -vos*) vendor=stratus ;; esac basic_machine=`echo $basic_machine | sed "s/unknown/$vendor/"` ;; esac echo $basic_machine$os exit # Local variables: # eval: (add-hook 'write-file-hooks 'time-stamp) # time-stamp-start: "timestamp='" # time-stamp-format: "%:y-%02m-%02d" # time-stamp-end: "'" # End: fftw-3.3.3/depcomp0000755000175000017500000005064311777117217011007 00000000000000#! /bin/sh # depcomp - compile a program generating dependencies as side-effects scriptversion=2012-03-27.16; # UTC # Copyright (C) 1999, 2000, 2003, 2004, 2005, 2006, 2007, 2009, 2010, # 2011, 2012 Free Software Foundation, Inc. # This program is free software; you can redistribute it and/or modify # it under the terms of the GNU General Public License as published by # the Free Software Foundation; either version 2, or (at your option) # any later version. # This program is distributed in the hope that it will be useful, # but WITHOUT ANY WARRANTY; without even the implied warranty of # MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the # GNU General Public License for more details. # You should have received a copy of the GNU General Public License # along with this program. If not, see . # As a special exception to the GNU General Public License, if you # distribute this file as part of a program that contains a # configuration script generated by Autoconf, you may include it under # the same distribution terms that you use for the rest of that program. # Originally written by Alexandre Oliva . case $1 in '') echo "$0: No command. Try '$0 --help' for more information." 1>&2 exit 1; ;; -h | --h*) cat <<\EOF Usage: depcomp [--help] [--version] PROGRAM [ARGS] Run PROGRAMS ARGS to compile a file, generating dependencies as side-effects. Environment variables: depmode Dependency tracking mode. source Source file read by 'PROGRAMS ARGS'. object Object file output by 'PROGRAMS ARGS'. DEPDIR directory where to store dependencies. depfile Dependency file to output. tmpdepfile Temporary file to use when outputting dependencies. libtool Whether libtool is used (yes/no). Report bugs to . EOF exit $? ;; -v | --v*) echo "depcomp $scriptversion" exit $? ;; esac # A tabulation character. tab=' ' # A newline character. nl=' ' if test -z "$depmode" || test -z "$source" || test -z "$object"; then echo "depcomp: Variables source, object and depmode must be set" 1>&2 exit 1 fi # Dependencies for sub/bar.o or sub/bar.obj go into sub/.deps/bar.Po. depfile=${depfile-`echo "$object" | sed 's|[^\\/]*$|'${DEPDIR-.deps}'/&|;s|\.\([^.]*\)$|.P\1|;s|Pobj$|Po|'`} tmpdepfile=${tmpdepfile-`echo "$depfile" | sed 's/\.\([^.]*\)$/.T\1/'`} rm -f "$tmpdepfile" # Some modes work just like other modes, but use different flags. We # parameterize here, but still list the modes in the big case below, # to make depend.m4 easier to write. Note that we *cannot* use a case # here, because this file can only contain one case statement. if test "$depmode" = hp; then # HP compiler uses -M and no extra arg. gccflag=-M depmode=gcc fi if test "$depmode" = dashXmstdout; then # This is just like dashmstdout with a different argument. dashmflag=-xM depmode=dashmstdout fi cygpath_u="cygpath -u -f -" if test "$depmode" = msvcmsys; then # This is just like msvisualcpp but w/o cygpath translation. # Just convert the backslash-escaped backslashes to single forward # slashes to satisfy depend.m4 cygpath_u='sed s,\\\\,/,g' depmode=msvisualcpp fi if test "$depmode" = msvc7msys; then # This is just like msvc7 but w/o cygpath translation. # Just convert the backslash-escaped backslashes to single forward # slashes to satisfy depend.m4 cygpath_u='sed s,\\\\,/,g' depmode=msvc7 fi if test "$depmode" = xlc; then # IBM C/C++ Compilers xlc/xlC can output gcc-like dependency informations. gccflag=-qmakedep=gcc,-MF depmode=gcc fi case "$depmode" in gcc3) ## gcc 3 implements dependency tracking that does exactly what ## we want. Yay! Note: for some reason libtool 1.4 doesn't like ## it if -MD -MP comes after the -MF stuff. Hmm. ## Unfortunately, FreeBSD c89 acceptance of flags depends upon ## the command line argument order; so add the flags where they ## appear in depend2.am. Note that the slowdown incurred here ## affects only configure: in makefiles, %FASTDEP% shortcuts this. for arg do case $arg in -c) set fnord "$@" -MT "$object" -MD -MP -MF "$tmpdepfile" "$arg" ;; *) set fnord "$@" "$arg" ;; esac shift # fnord shift # $arg done "$@" stat=$? if test $stat -eq 0; then : else rm -f "$tmpdepfile" exit $stat fi mv "$tmpdepfile" "$depfile" ;; gcc) ## There are various ways to get dependency output from gcc. Here's ## why we pick this rather obscure method: ## - Don't want to use -MD because we'd like the dependencies to end ## up in a subdir. Having to rename by hand is ugly. ## (We might end up doing this anyway to support other compilers.) ## - The DEPENDENCIES_OUTPUT environment variable makes gcc act like ## -MM, not -M (despite what the docs say). ## - Using -M directly means running the compiler twice (even worse ## than renaming). if test -z "$gccflag"; then gccflag=-MD, fi "$@" -Wp,"$gccflag$tmpdepfile" stat=$? if test $stat -eq 0; then : else rm -f "$tmpdepfile" exit $stat fi rm -f "$depfile" echo "$object : \\" > "$depfile" alpha=ABCDEFGHIJKLMNOPQRSTUVWXYZabcdefghijklmnopqrstuvwxyz ## The second -e expression handles DOS-style file names with drive letters. sed -e 's/^[^:]*: / /' \ -e 's/^['$alpha']:\/[^:]*: / /' < "$tmpdepfile" >> "$depfile" ## This next piece of magic avoids the "deleted header file" problem. ## The problem is that when a header file which appears in a .P file ## is deleted, the dependency causes make to die (because there is ## typically no way to rebuild the header). We avoid this by adding ## dummy dependencies for each header file. Too bad gcc doesn't do ## this for us directly. tr ' ' "$nl" < "$tmpdepfile" | ## Some versions of gcc put a space before the ':'. On the theory ## that the space means something, we add a space to the output as ## well. hp depmode also adds that space, but also prefixes the VPATH ## to the object. Take care to not repeat it in the output. ## Some versions of the HPUX 10.20 sed can't process this invocation ## correctly. Breaking it into two sed invocations is a workaround. sed -e 's/^\\$//' -e '/^$/d' -e "s|.*$object$||" -e '/:$/d' \ | sed -e 's/$/ :/' >> "$depfile" rm -f "$tmpdepfile" ;; hp) # This case exists only to let depend.m4 do its work. It works by # looking at the text of this script. This case will never be run, # since it is checked for above. exit 1 ;; sgi) if test "$libtool" = yes; then "$@" "-Wp,-MDupdate,$tmpdepfile" else "$@" -MDupdate "$tmpdepfile" fi stat=$? if test $stat -eq 0; then : else rm -f "$tmpdepfile" exit $stat fi rm -f "$depfile" if test -f "$tmpdepfile"; then # yes, the sourcefile depend on other files echo "$object : \\" > "$depfile" # Clip off the initial element (the dependent). Don't try to be # clever and replace this with sed code, as IRIX sed won't handle # lines with more than a fixed number of characters (4096 in # IRIX 6.2 sed, 8192 in IRIX 6.5). We also remove comment lines; # the IRIX cc adds comments like '#:fec' to the end of the # dependency line. tr ' ' "$nl" < "$tmpdepfile" \ | sed -e 's/^.*\.o://' -e 's/#.*$//' -e '/^$/ d' | \ tr "$nl" ' ' >> "$depfile" echo >> "$depfile" # The second pass generates a dummy entry for each header file. tr ' ' "$nl" < "$tmpdepfile" \ | sed -e 's/^.*\.o://' -e 's/#.*$//' -e '/^$/ d' -e 's/$/:/' \ >> "$depfile" else # The sourcefile does not contain any dependencies, so just # store a dummy comment line, to avoid errors with the Makefile # "include basename.Plo" scheme. echo "#dummy" > "$depfile" fi rm -f "$tmpdepfile" ;; xlc) # This case exists only to let depend.m4 do its work. It works by # looking at the text of this script. This case will never be run, # since it is checked for above. exit 1 ;; aix) # The C for AIX Compiler uses -M and outputs the dependencies # in a .u file. In older versions, this file always lives in the # current directory. Also, the AIX compiler puts '$object:' at the # start of each line; $object doesn't have directory information. # Version 6 uses the directory in both cases. dir=`echo "$object" | sed -e 's|/[^/]*$|/|'` test "x$dir" = "x$object" && dir= base=`echo "$object" | sed -e 's|^.*/||' -e 's/\.o$//' -e 's/\.lo$//'` if test "$libtool" = yes; then tmpdepfile1=$dir$base.u tmpdepfile2=$base.u tmpdepfile3=$dir.libs/$base.u "$@" -Wc,-M else tmpdepfile1=$dir$base.u tmpdepfile2=$dir$base.u tmpdepfile3=$dir$base.u "$@" -M fi stat=$? if test $stat -eq 0; then : else rm -f "$tmpdepfile1" "$tmpdepfile2" "$tmpdepfile3" exit $stat fi for tmpdepfile in "$tmpdepfile1" "$tmpdepfile2" "$tmpdepfile3" do test -f "$tmpdepfile" && break done if test -f "$tmpdepfile"; then # Each line is of the form 'foo.o: dependent.h'. # Do two passes, one to just change these to # '$object: dependent.h' and one to simply 'dependent.h:'. sed -e "s,^.*\.[a-z]*:,$object:," < "$tmpdepfile" > "$depfile" sed -e 's,^.*\.[a-z]*:['"$tab"' ]*,,' -e 's,$,:,' < "$tmpdepfile" >> "$depfile" else # The sourcefile does not contain any dependencies, so just # store a dummy comment line, to avoid errors with the Makefile # "include basename.Plo" scheme. echo "#dummy" > "$depfile" fi rm -f "$tmpdepfile" ;; icc) # Intel's C compiler anf tcc (Tiny C Compiler) understand '-MD -MF file'. # However on # $CC -MD -MF foo.d -c -o sub/foo.o sub/foo.c # ICC 7.0 will fill foo.d with something like # foo.o: sub/foo.c # foo.o: sub/foo.h # which is wrong. We want # sub/foo.o: sub/foo.c # sub/foo.o: sub/foo.h # sub/foo.c: # sub/foo.h: # ICC 7.1 will output # foo.o: sub/foo.c sub/foo.h # and will wrap long lines using '\': # foo.o: sub/foo.c ... \ # sub/foo.h ... \ # ... # tcc 0.9.26 (FIXME still under development at the moment of writing) # will emit a similar output, but also prepend the continuation lines # with horizontal tabulation characters. "$@" -MD -MF "$tmpdepfile" stat=$? if test $stat -eq 0; then : else rm -f "$tmpdepfile" exit $stat fi rm -f "$depfile" # Each line is of the form 'foo.o: dependent.h', # or 'foo.o: dep1.h dep2.h \', or ' dep3.h dep4.h \'. # Do two passes, one to just change these to # '$object: dependent.h' and one to simply 'dependent.h:'. sed -e "s/^[ $tab][ $tab]*/ /" -e "s,^[^:]*:,$object :," \ < "$tmpdepfile" > "$depfile" sed ' s/[ '"$tab"'][ '"$tab"']*/ /g s/^ *// s/ *\\*$// s/^[^:]*: *// /^$/d /:$/d s/$/ :/ ' < "$tmpdepfile" >> "$depfile" rm -f "$tmpdepfile" ;; hp2) # The "hp" stanza above does not work with aCC (C++) and HP's ia64 # compilers, which have integrated preprocessors. The correct option # to use with these is +Maked; it writes dependencies to a file named # 'foo.d', which lands next to the object file, wherever that # happens to be. # Much of this is similar to the tru64 case; see comments there. dir=`echo "$object" | sed -e 's|/[^/]*$|/|'` test "x$dir" = "x$object" && dir= base=`echo "$object" | sed -e 's|^.*/||' -e 's/\.o$//' -e 's/\.lo$//'` if test "$libtool" = yes; then tmpdepfile1=$dir$base.d tmpdepfile2=$dir.libs/$base.d "$@" -Wc,+Maked else tmpdepfile1=$dir$base.d tmpdepfile2=$dir$base.d "$@" +Maked fi stat=$? if test $stat -eq 0; then : else rm -f "$tmpdepfile1" "$tmpdepfile2" exit $stat fi for tmpdepfile in "$tmpdepfile1" "$tmpdepfile2" do test -f "$tmpdepfile" && break done if test -f "$tmpdepfile"; then sed -e "s,^.*\.[a-z]*:,$object:," "$tmpdepfile" > "$depfile" # Add 'dependent.h:' lines. sed -ne '2,${ s/^ *// s/ \\*$// s/$/:/ p }' "$tmpdepfile" >> "$depfile" else echo "#dummy" > "$depfile" fi rm -f "$tmpdepfile" "$tmpdepfile2" ;; tru64) # The Tru64 compiler uses -MD to generate dependencies as a side # effect. 'cc -MD -o foo.o ...' puts the dependencies into 'foo.o.d'. # At least on Alpha/Redhat 6.1, Compaq CCC V6.2-504 seems to put # dependencies in 'foo.d' instead, so we check for that too. # Subdirectories are respected. dir=`echo "$object" | sed -e 's|/[^/]*$|/|'` test "x$dir" = "x$object" && dir= base=`echo "$object" | sed -e 's|^.*/||' -e 's/\.o$//' -e 's/\.lo$//'` if test "$libtool" = yes; then # With Tru64 cc, shared objects can also be used to make a # static library. This mechanism is used in libtool 1.4 series to # handle both shared and static libraries in a single compilation. # With libtool 1.4, dependencies were output in $dir.libs/$base.lo.d. # # With libtool 1.5 this exception was removed, and libtool now # generates 2 separate objects for the 2 libraries. These two # compilations output dependencies in $dir.libs/$base.o.d and # in $dir$base.o.d. We have to check for both files, because # one of the two compilations can be disabled. We should prefer # $dir$base.o.d over $dir.libs/$base.o.d because the latter is # automatically cleaned when .libs/ is deleted, while ignoring # the former would cause a distcleancheck panic. tmpdepfile1=$dir.libs/$base.lo.d # libtool 1.4 tmpdepfile2=$dir$base.o.d # libtool 1.5 tmpdepfile3=$dir.libs/$base.o.d # libtool 1.5 tmpdepfile4=$dir.libs/$base.d # Compaq CCC V6.2-504 "$@" -Wc,-MD else tmpdepfile1=$dir$base.o.d tmpdepfile2=$dir$base.d tmpdepfile3=$dir$base.d tmpdepfile4=$dir$base.d "$@" -MD fi stat=$? if test $stat -eq 0; then : else rm -f "$tmpdepfile1" "$tmpdepfile2" "$tmpdepfile3" "$tmpdepfile4" exit $stat fi for tmpdepfile in "$tmpdepfile1" "$tmpdepfile2" "$tmpdepfile3" "$tmpdepfile4" do test -f "$tmpdepfile" && break done if test -f "$tmpdepfile"; then sed -e "s,^.*\.[a-z]*:,$object:," < "$tmpdepfile" > "$depfile" sed -e 's,^.*\.[a-z]*:['"$tab"' ]*,,' -e 's,$,:,' < "$tmpdepfile" >> "$depfile" else echo "#dummy" > "$depfile" fi rm -f "$tmpdepfile" ;; msvc7) if test "$libtool" = yes; then showIncludes=-Wc,-showIncludes else showIncludes=-showIncludes fi "$@" $showIncludes > "$tmpdepfile" stat=$? grep -v '^Note: including file: ' "$tmpdepfile" if test "$stat" = 0; then : else rm -f "$tmpdepfile" exit $stat fi rm -f "$depfile" echo "$object : \\" > "$depfile" # The first sed program below extracts the file names and escapes # backslashes for cygpath. The second sed program outputs the file # name when reading, but also accumulates all include files in the # hold buffer in order to output them again at the end. This only # works with sed implementations that can handle large buffers. sed < "$tmpdepfile" -n ' /^Note: including file: *\(.*\)/ { s//\1/ s/\\/\\\\/g p }' | $cygpath_u | sort -u | sed -n ' s/ /\\ /g s/\(.*\)/'"$tab"'\1 \\/p s/.\(.*\) \\/\1:/ H $ { s/.*/'"$tab"'/ G p }' >> "$depfile" rm -f "$tmpdepfile" ;; msvc7msys) # This case exists only to let depend.m4 do its work. It works by # looking at the text of this script. This case will never be run, # since it is checked for above. exit 1 ;; #nosideeffect) # This comment above is used by automake to tell side-effect # dependency tracking mechanisms from slower ones. dashmstdout) # Important note: in order to support this mode, a compiler *must* # always write the preprocessed file to stdout, regardless of -o. "$@" || exit $? # Remove the call to Libtool. if test "$libtool" = yes; then while test "X$1" != 'X--mode=compile'; do shift done shift fi # Remove '-o $object'. IFS=" " for arg do case $arg in -o) shift ;; $object) shift ;; *) set fnord "$@" "$arg" shift # fnord shift # $arg ;; esac done test -z "$dashmflag" && dashmflag=-M # Require at least two characters before searching for ':' # in the target name. This is to cope with DOS-style filenames: # a dependency such as 'c:/foo/bar' could be seen as target 'c' otherwise. "$@" $dashmflag | sed 's:^['"$tab"' ]*[^:'"$tab"' ][^:][^:]*\:['"$tab"' ]*:'"$object"'\: :' > "$tmpdepfile" rm -f "$depfile" cat < "$tmpdepfile" > "$depfile" tr ' ' "$nl" < "$tmpdepfile" | \ ## Some versions of the HPUX 10.20 sed can't process this invocation ## correctly. Breaking it into two sed invocations is a workaround. sed -e 's/^\\$//' -e '/^$/d' -e '/:$/d' | sed -e 's/$/ :/' >> "$depfile" rm -f "$tmpdepfile" ;; dashXmstdout) # This case only exists to satisfy depend.m4. It is never actually # run, as this mode is specially recognized in the preamble. exit 1 ;; makedepend) "$@" || exit $? # Remove any Libtool call if test "$libtool" = yes; then while test "X$1" != 'X--mode=compile'; do shift done shift fi # X makedepend shift cleared=no eat=no for arg do case $cleared in no) set ""; shift cleared=yes ;; esac if test $eat = yes; then eat=no continue fi case "$arg" in -D*|-I*) set fnord "$@" "$arg"; shift ;; # Strip any option that makedepend may not understand. Remove # the object too, otherwise makedepend will parse it as a source file. -arch) eat=yes ;; -*|$object) ;; *) set fnord "$@" "$arg"; shift ;; esac done obj_suffix=`echo "$object" | sed 's/^.*\././'` touch "$tmpdepfile" ${MAKEDEPEND-makedepend} -o"$obj_suffix" -f"$tmpdepfile" "$@" rm -f "$depfile" # makedepend may prepend the VPATH from the source file name to the object. # No need to regex-escape $object, excess matching of '.' is harmless. sed "s|^.*\($object *:\)|\1|" "$tmpdepfile" > "$depfile" sed '1,2d' "$tmpdepfile" | tr ' ' "$nl" | \ ## Some versions of the HPUX 10.20 sed can't process this invocation ## correctly. Breaking it into two sed invocations is a workaround. sed -e 's/^\\$//' -e '/^$/d' -e '/:$/d' | sed -e 's/$/ :/' >> "$depfile" rm -f "$tmpdepfile" "$tmpdepfile".bak ;; cpp) # Important note: in order to support this mode, a compiler *must* # always write the preprocessed file to stdout. "$@" || exit $? # Remove the call to Libtool. if test "$libtool" = yes; then while test "X$1" != 'X--mode=compile'; do shift done shift fi # Remove '-o $object'. IFS=" " for arg do case $arg in -o) shift ;; $object) shift ;; *) set fnord "$@" "$arg" shift # fnord shift # $arg ;; esac done "$@" -E | sed -n -e '/^# [0-9][0-9]* "\([^"]*\)".*/ s:: \1 \\:p' \ -e '/^#line [0-9][0-9]* "\([^"]*\)".*/ s:: \1 \\:p' | sed '$ s: \\$::' > "$tmpdepfile" rm -f "$depfile" echo "$object : \\" > "$depfile" cat < "$tmpdepfile" >> "$depfile" sed < "$tmpdepfile" '/^$/d;s/^ //;s/ \\$//;s/$/ :/' >> "$depfile" rm -f "$tmpdepfile" ;; msvisualcpp) # Important note: in order to support this mode, a compiler *must* # always write the preprocessed file to stdout. "$@" || exit $? # Remove the call to Libtool. if test "$libtool" = yes; then while test "X$1" != 'X--mode=compile'; do shift done shift fi IFS=" " for arg do case "$arg" in -o) shift ;; $object) shift ;; "-Gm"|"/Gm"|"-Gi"|"/Gi"|"-ZI"|"/ZI") set fnord "$@" shift shift ;; *) set fnord "$@" "$arg" shift shift ;; esac done "$@" -E 2>/dev/null | sed -n '/^#line [0-9][0-9]* "\([^"]*\)"/ s::\1:p' | $cygpath_u | sort -u > "$tmpdepfile" rm -f "$depfile" echo "$object : \\" > "$depfile" sed < "$tmpdepfile" -n -e 's% %\\ %g' -e '/^\(.*\)$/ s::'"$tab"'\1 \\:p' >> "$depfile" echo "$tab" >> "$depfile" sed < "$tmpdepfile" -n -e 's% %\\ %g' -e '/^\(.*\)$/ s::\1\::p' >> "$depfile" rm -f "$tmpdepfile" ;; msvcmsys) # This case exists only to let depend.m4 do its work. It works by # looking at the text of this script. This case will never be run, # since it is checked for above. exit 1 ;; none) exec "$@" ;; *) echo "Unknown depmode $depmode" 1>&2 exit 1 ;; esac exit 0 # Local Variables: # mode: shell-script # sh-indentation: 2 # eval: (add-hook 'write-file-hooks 'time-stamp) # time-stamp-start: "scriptversion=" # time-stamp-format: "%:y-%02m-%02d.%02H" # time-stamp-time-zone: "UTC" # time-stamp-end: "; # UTC" # End: fftw-3.3.3/install-sh0000755000175000017500000003325611777117217011437 00000000000000#!/bin/sh # install - install a program, script, or datafile scriptversion=2011-01-19.21; # UTC # This originates from X11R5 (mit/util/scripts/install.sh), which was # later released in X11R6 (xc/config/util/install.sh) with the # following copyright and license. # # Copyright (C) 1994 X Consortium # # Permission is hereby granted, free of charge, to any person obtaining a copy # of this software and associated documentation files (the "Software"), to # deal in the Software without restriction, including without limitation the # rights to use, copy, modify, merge, publish, distribute, sublicense, and/or # sell copies of the Software, and to permit persons to whom the Software is # furnished to do so, subject to the following conditions: # # The above copyright notice and this permission notice shall be included in # all copies or substantial portions of the Software. # # THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS OR # IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY, # FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT. IN NO EVENT SHALL THE # X CONSORTIUM BE LIABLE FOR ANY CLAIM, DAMAGES OR OTHER LIABILITY, WHETHER IN # AN ACTION OF CONTRACT, TORT OR OTHERWISE, ARISING FROM, OUT OF OR IN CONNEC- # TION WITH THE SOFTWARE OR THE USE OR OTHER DEALINGS IN THE SOFTWARE. # # Except as contained in this notice, the name of the X Consortium shall not # be used in advertising or otherwise to promote the sale, use or other deal- # ings in this Software without prior written authorization from the X Consor- # tium. # # # FSF changes to this file are in the public domain. # # Calling this script install-sh is preferred over install.sh, to prevent # `make' implicit rules from creating a file called install from it # when there is no Makefile. # # This script is compatible with the BSD install script, but was written # from scratch. nl=' ' IFS=" "" $nl" # set DOITPROG to echo to test this script # Don't use :- since 4.3BSD and earlier shells don't like it. doit=${DOITPROG-} if test -z "$doit"; then doit_exec=exec else doit_exec=$doit fi # Put in absolute file names if you don't have them in your path; # or use environment vars. chgrpprog=${CHGRPPROG-chgrp} chmodprog=${CHMODPROG-chmod} chownprog=${CHOWNPROG-chown} cmpprog=${CMPPROG-cmp} cpprog=${CPPROG-cp} mkdirprog=${MKDIRPROG-mkdir} mvprog=${MVPROG-mv} rmprog=${RMPROG-rm} stripprog=${STRIPPROG-strip} posix_glob='?' initialize_posix_glob=' test "$posix_glob" != "?" || { if (set -f) 2>/dev/null; then posix_glob= else posix_glob=: fi } ' posix_mkdir= # Desired mode of installed file. mode=0755 chgrpcmd= chmodcmd=$chmodprog chowncmd= mvcmd=$mvprog rmcmd="$rmprog -f" stripcmd= src= dst= dir_arg= dst_arg= copy_on_change=false no_target_directory= usage="\ Usage: $0 [OPTION]... [-T] SRCFILE DSTFILE or: $0 [OPTION]... SRCFILES... DIRECTORY or: $0 [OPTION]... -t DIRECTORY SRCFILES... or: $0 [OPTION]... -d DIRECTORIES... In the 1st form, copy SRCFILE to DSTFILE. In the 2nd and 3rd, copy all SRCFILES to DIRECTORY. In the 4th, create DIRECTORIES. Options: --help display this help and exit. --version display version info and exit. -c (ignored) -C install only if different (preserve the last data modification time) -d create directories instead of installing files. -g GROUP $chgrpprog installed files to GROUP. -m MODE $chmodprog installed files to MODE. -o USER $chownprog installed files to USER. -s $stripprog installed files. -t DIRECTORY install into DIRECTORY. -T report an error if DSTFILE is a directory. Environment variables override the default commands: CHGRPPROG CHMODPROG CHOWNPROG CMPPROG CPPROG MKDIRPROG MVPROG RMPROG STRIPPROG " while test $# -ne 0; do case $1 in -c) ;; -C) copy_on_change=true;; -d) dir_arg=true;; -g) chgrpcmd="$chgrpprog $2" shift;; --help) echo "$usage"; exit $?;; -m) mode=$2 case $mode in *' '* | *' '* | *' '* | *'*'* | *'?'* | *'['*) echo "$0: invalid mode: $mode" >&2 exit 1;; esac shift;; -o) chowncmd="$chownprog $2" shift;; -s) stripcmd=$stripprog;; -t) dst_arg=$2 # Protect names problematic for `test' and other utilities. case $dst_arg in -* | [=\(\)!]) dst_arg=./$dst_arg;; esac shift;; -T) no_target_directory=true;; --version) echo "$0 $scriptversion"; exit $?;; --) shift break;; -*) echo "$0: invalid option: $1" >&2 exit 1;; *) break;; esac shift done if test $# -ne 0 && test -z "$dir_arg$dst_arg"; then # When -d is used, all remaining arguments are directories to create. # When -t is used, the destination is already specified. # Otherwise, the last argument is the destination. Remove it from $@. for arg do if test -n "$dst_arg"; then # $@ is not empty: it contains at least $arg. set fnord "$@" "$dst_arg" shift # fnord fi shift # arg dst_arg=$arg # Protect names problematic for `test' and other utilities. case $dst_arg in -* | [=\(\)!]) dst_arg=./$dst_arg;; esac done fi if test $# -eq 0; then if test -z "$dir_arg"; then echo "$0: no input file specified." >&2 exit 1 fi # It's OK to call `install-sh -d' without argument. # This can happen when creating conditional directories. exit 0 fi if test -z "$dir_arg"; then do_exit='(exit $ret); exit $ret' trap "ret=129; $do_exit" 1 trap "ret=130; $do_exit" 2 trap "ret=141; $do_exit" 13 trap "ret=143; $do_exit" 15 # Set umask so as not to create temps with too-generous modes. # However, 'strip' requires both read and write access to temps. case $mode in # Optimize common cases. *644) cp_umask=133;; *755) cp_umask=22;; *[0-7]) if test -z "$stripcmd"; then u_plus_rw= else u_plus_rw='% 200' fi cp_umask=`expr '(' 777 - $mode % 1000 ')' $u_plus_rw`;; *) if test -z "$stripcmd"; then u_plus_rw= else u_plus_rw=,u+rw fi cp_umask=$mode$u_plus_rw;; esac fi for src do # Protect names problematic for `test' and other utilities. case $src in -* | [=\(\)!]) src=./$src;; esac if test -n "$dir_arg"; then dst=$src dstdir=$dst test -d "$dstdir" dstdir_status=$? else # Waiting for this to be detected by the "$cpprog $src $dsttmp" command # might cause directories to be created, which would be especially bad # if $src (and thus $dsttmp) contains '*'. if test ! -f "$src" && test ! -d "$src"; then echo "$0: $src does not exist." >&2 exit 1 fi if test -z "$dst_arg"; then echo "$0: no destination specified." >&2 exit 1 fi dst=$dst_arg # If destination is a directory, append the input filename; won't work # if double slashes aren't ignored. if test -d "$dst"; then if test -n "$no_target_directory"; then echo "$0: $dst_arg: Is a directory" >&2 exit 1 fi dstdir=$dst dst=$dstdir/`basename "$src"` dstdir_status=0 else # Prefer dirname, but fall back on a substitute if dirname fails. dstdir=` (dirname "$dst") 2>/dev/null || expr X"$dst" : 'X\(.*[^/]\)//*[^/][^/]*/*$' \| \ X"$dst" : 'X\(//\)[^/]' \| \ X"$dst" : 'X\(//\)$' \| \ X"$dst" : 'X\(/\)' \| . 2>/dev/null || echo X"$dst" | sed '/^X\(.*[^/]\)\/\/*[^/][^/]*\/*$/{ s//\1/ q } /^X\(\/\/\)[^/].*/{ s//\1/ q } /^X\(\/\/\)$/{ s//\1/ q } /^X\(\/\).*/{ s//\1/ q } s/.*/./; q' ` test -d "$dstdir" dstdir_status=$? fi fi obsolete_mkdir_used=false if test $dstdir_status != 0; then case $posix_mkdir in '') # Create intermediate dirs using mode 755 as modified by the umask. # This is like FreeBSD 'install' as of 1997-10-28. umask=`umask` case $stripcmd.$umask in # Optimize common cases. *[2367][2367]) mkdir_umask=$umask;; .*0[02][02] | .[02][02] | .[02]) mkdir_umask=22;; *[0-7]) mkdir_umask=`expr $umask + 22 \ - $umask % 100 % 40 + $umask % 20 \ - $umask % 10 % 4 + $umask % 2 `;; *) mkdir_umask=$umask,go-w;; esac # With -d, create the new directory with the user-specified mode. # Otherwise, rely on $mkdir_umask. if test -n "$dir_arg"; then mkdir_mode=-m$mode else mkdir_mode= fi posix_mkdir=false case $umask in *[123567][0-7][0-7]) # POSIX mkdir -p sets u+wx bits regardless of umask, which # is incompatible with FreeBSD 'install' when (umask & 300) != 0. ;; *) tmpdir=${TMPDIR-/tmp}/ins$RANDOM-$$ trap 'ret=$?; rmdir "$tmpdir/d" "$tmpdir" 2>/dev/null; exit $ret' 0 if (umask $mkdir_umask && exec $mkdirprog $mkdir_mode -p -- "$tmpdir/d") >/dev/null 2>&1 then if test -z "$dir_arg" || { # Check for POSIX incompatibilities with -m. # HP-UX 11.23 and IRIX 6.5 mkdir -m -p sets group- or # other-writeable bit of parent directory when it shouldn't. # FreeBSD 6.1 mkdir -m -p sets mode of existing directory. ls_ld_tmpdir=`ls -ld "$tmpdir"` case $ls_ld_tmpdir in d????-?r-*) different_mode=700;; d????-?--*) different_mode=755;; *) false;; esac && $mkdirprog -m$different_mode -p -- "$tmpdir" && { ls_ld_tmpdir_1=`ls -ld "$tmpdir"` test "$ls_ld_tmpdir" = "$ls_ld_tmpdir_1" } } then posix_mkdir=: fi rmdir "$tmpdir/d" "$tmpdir" else # Remove any dirs left behind by ancient mkdir implementations. rmdir ./$mkdir_mode ./-p ./-- 2>/dev/null fi trap '' 0;; esac;; esac if $posix_mkdir && ( umask $mkdir_umask && $doit_exec $mkdirprog $mkdir_mode -p -- "$dstdir" ) then : else # The umask is ridiculous, or mkdir does not conform to POSIX, # or it failed possibly due to a race condition. Create the # directory the slow way, step by step, checking for races as we go. case $dstdir in /*) prefix='/';; [-=\(\)!]*) prefix='./';; *) prefix='';; esac eval "$initialize_posix_glob" oIFS=$IFS IFS=/ $posix_glob set -f set fnord $dstdir shift $posix_glob set +f IFS=$oIFS prefixes= for d do test X"$d" = X && continue prefix=$prefix$d if test -d "$prefix"; then prefixes= else if $posix_mkdir; then (umask=$mkdir_umask && $doit_exec $mkdirprog $mkdir_mode -p -- "$dstdir") && break # Don't fail if two instances are running concurrently. test -d "$prefix" || exit 1 else case $prefix in *\'*) qprefix=`echo "$prefix" | sed "s/'/'\\\\\\\\''/g"`;; *) qprefix=$prefix;; esac prefixes="$prefixes '$qprefix'" fi fi prefix=$prefix/ done if test -n "$prefixes"; then # Don't fail if two instances are running concurrently. (umask $mkdir_umask && eval "\$doit_exec \$mkdirprog $prefixes") || test -d "$dstdir" || exit 1 obsolete_mkdir_used=true fi fi fi if test -n "$dir_arg"; then { test -z "$chowncmd" || $doit $chowncmd "$dst"; } && { test -z "$chgrpcmd" || $doit $chgrpcmd "$dst"; } && { test "$obsolete_mkdir_used$chowncmd$chgrpcmd" = false || test -z "$chmodcmd" || $doit $chmodcmd $mode "$dst"; } || exit 1 else # Make a couple of temp file names in the proper directory. dsttmp=$dstdir/_inst.$$_ rmtmp=$dstdir/_rm.$$_ # Trap to clean up those temp files at exit. trap 'ret=$?; rm -f "$dsttmp" "$rmtmp" && exit $ret' 0 # Copy the file name to the temp name. (umask $cp_umask && $doit_exec $cpprog "$src" "$dsttmp") && # and set any options; do chmod last to preserve setuid bits. # # If any of these fail, we abort the whole thing. If we want to # ignore errors from any of these, just make sure not to ignore # errors from the above "$doit $cpprog $src $dsttmp" command. # { test -z "$chowncmd" || $doit $chowncmd "$dsttmp"; } && { test -z "$chgrpcmd" || $doit $chgrpcmd "$dsttmp"; } && { test -z "$stripcmd" || $doit $stripcmd "$dsttmp"; } && { test -z "$chmodcmd" || $doit $chmodcmd $mode "$dsttmp"; } && # If -C, don't bother to copy if it wouldn't change the file. if $copy_on_change && old=`LC_ALL=C ls -dlL "$dst" 2>/dev/null` && new=`LC_ALL=C ls -dlL "$dsttmp" 2>/dev/null` && eval "$initialize_posix_glob" && $posix_glob set -f && set X $old && old=:$2:$4:$5:$6 && set X $new && new=:$2:$4:$5:$6 && $posix_glob set +f && test "$old" = "$new" && $cmpprog "$dst" "$dsttmp" >/dev/null 2>&1 then rm -f "$dsttmp" else # Rename the file to the real destination. $doit $mvcmd -f "$dsttmp" "$dst" 2>/dev/null || # The rename failed, perhaps because mv can't rename something else # to itself, or perhaps because mv is so ancient that it does not # support -f. { # Now remove or move aside any old file at destination location. # We try this two ways since rm can't unlink itself on some # systems and the destination file might be busy for other # reasons. In this case, the final cleanup might fail but the new # file should still install successfully. { test ! -f "$dst" || $doit $rmcmd -f "$dst" 2>/dev/null || { $doit $mvcmd -f "$dst" "$rmtmp" 2>/dev/null && { $doit $rmcmd -f "$rmtmp" 2>/dev/null; :; } } || { echo "$0: cannot unlink or rename $dst" >&2 (exit 1); exit 1 } } && # Now rename the file to the real destination. $doit $mvcmd "$dsttmp" "$dst" } fi || exit 1 trap '' 0 fi done # Local variables: # eval: (add-hook 'write-file-hooks 'time-stamp) # time-stamp-start: "scriptversion=" # time-stamp-format: "%:y-%02m-%02d.%02H" # time-stamp-time-zone: "UTC" # time-stamp-end: "; # UTC" # End: fftw-3.3.3/ltmain.sh0000644000175000017500000105202611765467473011261 00000000000000 # libtool (GNU libtool) 2.4.2 # Written by Gordon Matzigkeit , 1996 # Copyright (C) 1996, 1997, 1998, 1999, 2000, 2001, 2003, 2004, 2005, 2006, # 2007, 2008, 2009, 2010, 2011 Free Software Foundation, Inc. # This is free software; see the source for copying conditions. There is NO # warranty; not even for MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. # GNU Libtool is free software; you can redistribute it and/or modify # it under the terms of the GNU General Public License as published by # the Free Software Foundation; either version 2 of the License, or # (at your option) any later version. # # As a special exception to the GNU General Public License, # if you distribute this file as part of a program or library that # is built using GNU Libtool, you may include this file under the # same distribution terms that you use for the rest of that program. # # GNU Libtool is distributed in the hope that it will be useful, but # WITHOUT ANY WARRANTY; without even the implied warranty of # MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the GNU # General Public License for more details. # # You should have received a copy of the GNU General Public License # along with GNU Libtool; see the file COPYING. If not, a copy # can be downloaded from http://www.gnu.org/licenses/gpl.html, # or obtained by writing to the Free Software Foundation, Inc., # 51 Franklin Street, Fifth Floor, Boston, MA 02110-1301, USA. # Usage: $progname [OPTION]... [MODE-ARG]... # # Provide generalized library-building support services. # # --config show all configuration variables # --debug enable verbose shell tracing # -n, --dry-run display commands without modifying any files # --features display basic configuration information and exit # --mode=MODE use operation mode MODE # --preserve-dup-deps don't remove duplicate dependency libraries # --quiet, --silent don't print informational messages # --no-quiet, --no-silent # print informational messages (default) # --no-warn don't display warning messages # --tag=TAG use configuration variables from tag TAG # -v, --verbose print more informational messages than default # --no-verbose don't print the extra informational messages # --version print version information # -h, --help, --help-all print short, long, or detailed help message # # MODE must be one of the following: # # clean remove files from the build directory # compile compile a source file into a libtool object # execute automatically set library path, then run a program # finish complete the installation of libtool libraries # install install libraries or executables # link create a library or an executable # uninstall remove libraries from an installed directory # # MODE-ARGS vary depending on the MODE. When passed as first option, # `--mode=MODE' may be abbreviated as `MODE' or a unique abbreviation of that. # Try `$progname --help --mode=MODE' for a more detailed description of MODE. # # When reporting a bug, please describe a test case to reproduce it and # include the following information: # # host-triplet: $host # shell: $SHELL # compiler: $LTCC # compiler flags: $LTCFLAGS # linker: $LD (gnu? $with_gnu_ld) # $progname: (GNU libtool) 2.4.2 Debian-2.4.2-1.1 # automake: $automake_version # autoconf: $autoconf_version # # Report bugs to . # GNU libtool home page: . # General help using GNU software: . PROGRAM=libtool PACKAGE=libtool VERSION="2.4.2 Debian-2.4.2-1.1" TIMESTAMP="" package_revision=1.3337 # Be Bourne compatible if test -n "${ZSH_VERSION+set}" && (emulate sh) >/dev/null 2>&1; then emulate sh NULLCMD=: # Zsh 3.x and 4.x performs word splitting on ${1+"$@"}, which # is contrary to our usage. Disable this feature. alias -g '${1+"$@"}'='"$@"' setopt NO_GLOB_SUBST else case `(set -o) 2>/dev/null` in *posix*) set -o posix;; esac fi BIN_SH=xpg4; export BIN_SH # for Tru64 DUALCASE=1; export DUALCASE # for MKS sh # A function that is used when there is no print builtin or printf. func_fallback_echo () { eval 'cat <<_LTECHO_EOF $1 _LTECHO_EOF' } # NLS nuisances: We save the old values to restore during execute mode. lt_user_locale= lt_safe_locale= for lt_var in LANG LANGUAGE LC_ALL LC_CTYPE LC_COLLATE LC_MESSAGES do eval "if test \"\${$lt_var+set}\" = set; then save_$lt_var=\$$lt_var $lt_var=C export $lt_var lt_user_locale=\"$lt_var=\\\$save_\$lt_var; \$lt_user_locale\" lt_safe_locale=\"$lt_var=C; \$lt_safe_locale\" fi" done LC_ALL=C LANGUAGE=C export LANGUAGE LC_ALL $lt_unset CDPATH # Work around backward compatibility issue on IRIX 6.5. On IRIX 6.4+, sh # is ksh but when the shell is invoked as "sh" and the current value of # the _XPG environment variable is not equal to 1 (one), the special # positional parameter $0, within a function call, is the name of the # function. progpath="$0" : ${CP="cp -f"} test "${ECHO+set}" = set || ECHO=${as_echo-'printf %s\n'} : ${MAKE="make"} : ${MKDIR="mkdir"} : ${MV="mv -f"} : ${RM="rm -f"} : ${SHELL="${CONFIG_SHELL-/bin/sh}"} : ${Xsed="$SED -e 1s/^X//"} # Global variables: EXIT_SUCCESS=0 EXIT_FAILURE=1 EXIT_MISMATCH=63 # $? = 63 is used to indicate version mismatch to missing. EXIT_SKIP=77 # $? = 77 is used to indicate a skipped test to automake. exit_status=$EXIT_SUCCESS # Make sure IFS has a sensible default lt_nl=' ' IFS=" $lt_nl" dirname="s,/[^/]*$,," basename="s,^.*/,," # func_dirname file append nondir_replacement # Compute the dirname of FILE. If nonempty, add APPEND to the result, # otherwise set result to NONDIR_REPLACEMENT. func_dirname () { func_dirname_result=`$ECHO "${1}" | $SED "$dirname"` if test "X$func_dirname_result" = "X${1}"; then func_dirname_result="${3}" else func_dirname_result="$func_dirname_result${2}" fi } # func_dirname may be replaced by extended shell implementation # func_basename file func_basename () { func_basename_result=`$ECHO "${1}" | $SED "$basename"` } # func_basename may be replaced by extended shell implementation # func_dirname_and_basename file append nondir_replacement # perform func_basename and func_dirname in a single function # call: # dirname: Compute the dirname of FILE. If nonempty, # add APPEND to the result, otherwise set result # to NONDIR_REPLACEMENT. # value returned in "$func_dirname_result" # basename: Compute filename of FILE. # value retuned in "$func_basename_result" # Implementation must be kept synchronized with func_dirname # and func_basename. For efficiency, we do not delegate to # those functions but instead duplicate the functionality here. func_dirname_and_basename () { # Extract subdirectory from the argument. func_dirname_result=`$ECHO "${1}" | $SED -e "$dirname"` if test "X$func_dirname_result" = "X${1}"; then func_dirname_result="${3}" else func_dirname_result="$func_dirname_result${2}" fi func_basename_result=`$ECHO "${1}" | $SED -e "$basename"` } # func_dirname_and_basename may be replaced by extended shell implementation # func_stripname prefix suffix name # strip PREFIX and SUFFIX off of NAME. # PREFIX and SUFFIX must not contain globbing or regex special # characters, hashes, percent signs, but SUFFIX may contain a leading # dot (in which case that matches only a dot). # func_strip_suffix prefix name func_stripname () { case ${2} in .*) func_stripname_result=`$ECHO "${3}" | $SED "s%^${1}%%; s%\\\\${2}\$%%"`;; *) func_stripname_result=`$ECHO "${3}" | $SED "s%^${1}%%; s%${2}\$%%"`;; esac } # func_stripname may be replaced by extended shell implementation # These SED scripts presuppose an absolute path with a trailing slash. pathcar='s,^/\([^/]*\).*$,\1,' pathcdr='s,^/[^/]*,,' removedotparts=':dotsl s@/\./@/@g t dotsl s,/\.$,/,' collapseslashes='s@/\{1,\}@/@g' finalslash='s,/*$,/,' # func_normal_abspath PATH # Remove doubled-up and trailing slashes, "." path components, # and cancel out any ".." path components in PATH after making # it an absolute path. # value returned in "$func_normal_abspath_result" func_normal_abspath () { # Start from root dir and reassemble the path. func_normal_abspath_result= func_normal_abspath_tpath=$1 func_normal_abspath_altnamespace= case $func_normal_abspath_tpath in "") # Empty path, that just means $cwd. func_stripname '' '/' "`pwd`" func_normal_abspath_result=$func_stripname_result return ;; # The next three entries are used to spot a run of precisely # two leading slashes without using negated character classes; # we take advantage of case's first-match behaviour. ///*) # Unusual form of absolute path, do nothing. ;; //*) # Not necessarily an ordinary path; POSIX reserves leading '//' # and for example Cygwin uses it to access remote file shares # over CIFS/SMB, so we conserve a leading double slash if found. func_normal_abspath_altnamespace=/ ;; /*) # Absolute path, do nothing. ;; *) # Relative path, prepend $cwd. func_normal_abspath_tpath=`pwd`/$func_normal_abspath_tpath ;; esac # Cancel out all the simple stuff to save iterations. We also want # the path to end with a slash for ease of parsing, so make sure # there is one (and only one) here. func_normal_abspath_tpath=`$ECHO "$func_normal_abspath_tpath" | $SED \ -e "$removedotparts" -e "$collapseslashes" -e "$finalslash"` while :; do # Processed it all yet? if test "$func_normal_abspath_tpath" = / ; then # If we ascended to the root using ".." the result may be empty now. if test -z "$func_normal_abspath_result" ; then func_normal_abspath_result=/ fi break fi func_normal_abspath_tcomponent=`$ECHO "$func_normal_abspath_tpath" | $SED \ -e "$pathcar"` func_normal_abspath_tpath=`$ECHO "$func_normal_abspath_tpath" | $SED \ -e "$pathcdr"` # Figure out what to do with it case $func_normal_abspath_tcomponent in "") # Trailing empty path component, ignore it. ;; ..) # Parent dir; strip last assembled component from result. func_dirname "$func_normal_abspath_result" func_normal_abspath_result=$func_dirname_result ;; *) # Actual path component, append it. func_normal_abspath_result=$func_normal_abspath_result/$func_normal_abspath_tcomponent ;; esac done # Restore leading double-slash if one was found on entry. func_normal_abspath_result=$func_normal_abspath_altnamespace$func_normal_abspath_result } # func_relative_path SRCDIR DSTDIR # generates a relative path from SRCDIR to DSTDIR, with a trailing # slash if non-empty, suitable for immediately appending a filename # without needing to append a separator. # value returned in "$func_relative_path_result" func_relative_path () { func_relative_path_result= func_normal_abspath "$1" func_relative_path_tlibdir=$func_normal_abspath_result func_normal_abspath "$2" func_relative_path_tbindir=$func_normal_abspath_result # Ascend the tree starting from libdir while :; do # check if we have found a prefix of bindir case $func_relative_path_tbindir in $func_relative_path_tlibdir) # found an exact match func_relative_path_tcancelled= break ;; $func_relative_path_tlibdir*) # found a matching prefix func_stripname "$func_relative_path_tlibdir" '' "$func_relative_path_tbindir" func_relative_path_tcancelled=$func_stripname_result if test -z "$func_relative_path_result"; then func_relative_path_result=. fi break ;; *) func_dirname $func_relative_path_tlibdir func_relative_path_tlibdir=${func_dirname_result} if test "x$func_relative_path_tlibdir" = x ; then # Have to descend all the way to the root! func_relative_path_result=../$func_relative_path_result func_relative_path_tcancelled=$func_relative_path_tbindir break fi func_relative_path_result=../$func_relative_path_result ;; esac done # Now calculate path; take care to avoid doubling-up slashes. func_stripname '' '/' "$func_relative_path_result" func_relative_path_result=$func_stripname_result func_stripname '/' '/' "$func_relative_path_tcancelled" if test "x$func_stripname_result" != x ; then func_relative_path_result=${func_relative_path_result}/${func_stripname_result} fi # Normalisation. If bindir is libdir, return empty string, # else relative path ending with a slash; either way, target # file name can be directly appended. if test ! -z "$func_relative_path_result"; then func_stripname './' '' "$func_relative_path_result/" func_relative_path_result=$func_stripname_result fi } # The name of this program: func_dirname_and_basename "$progpath" progname=$func_basename_result # Make sure we have an absolute path for reexecution: case $progpath in [\\/]*|[A-Za-z]:\\*) ;; *[\\/]*) progdir=$func_dirname_result progdir=`cd "$progdir" && pwd` progpath="$progdir/$progname" ;; *) save_IFS="$IFS" IFS=${PATH_SEPARATOR-:} for progdir in $PATH; do IFS="$save_IFS" test -x "$progdir/$progname" && break done IFS="$save_IFS" test -n "$progdir" || progdir=`pwd` progpath="$progdir/$progname" ;; esac # Sed substitution that helps us do robust quoting. It backslashifies # metacharacters that are still active within double-quoted strings. Xsed="${SED}"' -e 1s/^X//' sed_quote_subst='s/\([`"$\\]\)/\\\1/g' # Same as above, but do not quote variable references. double_quote_subst='s/\(["`\\]\)/\\\1/g' # Sed substitution that turns a string into a regex matching for the # string literally. sed_make_literal_regex='s,[].[^$\\*\/],\\&,g' # Sed substitution that converts a w32 file name or path # which contains forward slashes, into one that contains # (escaped) backslashes. A very naive implementation. lt_sed_naive_backslashify='s|\\\\*|\\|g;s|/|\\|g;s|\\|\\\\|g' # Re-`\' parameter expansions in output of double_quote_subst that were # `\'-ed in input to the same. If an odd number of `\' preceded a '$' # in input to double_quote_subst, that '$' was protected from expansion. # Since each input `\' is now two `\'s, look for any number of runs of # four `\'s followed by two `\'s and then a '$'. `\' that '$'. bs='\\' bs2='\\\\' bs4='\\\\\\\\' dollar='\$' sed_double_backslash="\ s/$bs4/&\\ /g s/^$bs2$dollar/$bs&/ s/\\([^$bs]\\)$bs2$dollar/\\1$bs2$bs$dollar/g s/\n//g" # Standard options: opt_dry_run=false opt_help=false opt_quiet=false opt_verbose=false opt_warning=: # func_echo arg... # Echo program name prefixed message, along with the current mode # name if it has been set yet. func_echo () { $ECHO "$progname: ${opt_mode+$opt_mode: }$*" } # func_verbose arg... # Echo program name prefixed message in verbose mode only. func_verbose () { $opt_verbose && func_echo ${1+"$@"} # A bug in bash halts the script if the last line of a function # fails when set -e is in force, so we need another command to # work around that: : } # func_echo_all arg... # Invoke $ECHO with all args, space-separated. func_echo_all () { $ECHO "$*" } # func_error arg... # Echo program name prefixed message to standard error. func_error () { $ECHO "$progname: ${opt_mode+$opt_mode: }"${1+"$@"} 1>&2 } # func_warning arg... # Echo program name prefixed warning message to standard error. func_warning () { $opt_warning && $ECHO "$progname: ${opt_mode+$opt_mode: }warning: "${1+"$@"} 1>&2 # bash bug again: : } # func_fatal_error arg... # Echo program name prefixed message to standard error, and exit. func_fatal_error () { func_error ${1+"$@"} exit $EXIT_FAILURE } # func_fatal_help arg... # Echo program name prefixed message to standard error, followed by # a help hint, and exit. func_fatal_help () { func_error ${1+"$@"} func_fatal_error "$help" } help="Try \`$progname --help' for more information." ## default # func_grep expression filename # Check whether EXPRESSION matches any line of FILENAME, without output. func_grep () { $GREP "$1" "$2" >/dev/null 2>&1 } # func_mkdir_p directory-path # Make sure the entire path to DIRECTORY-PATH is available. func_mkdir_p () { my_directory_path="$1" my_dir_list= if test -n "$my_directory_path" && test "$opt_dry_run" != ":"; then # Protect directory names starting with `-' case $my_directory_path in -*) my_directory_path="./$my_directory_path" ;; esac # While some portion of DIR does not yet exist... while test ! -d "$my_directory_path"; do # ...make a list in topmost first order. Use a colon delimited # list incase some portion of path contains whitespace. my_dir_list="$my_directory_path:$my_dir_list" # If the last portion added has no slash in it, the list is done case $my_directory_path in */*) ;; *) break ;; esac # ...otherwise throw away the child directory and loop my_directory_path=`$ECHO "$my_directory_path" | $SED -e "$dirname"` done my_dir_list=`$ECHO "$my_dir_list" | $SED 's,:*$,,'` save_mkdir_p_IFS="$IFS"; IFS=':' for my_dir in $my_dir_list; do IFS="$save_mkdir_p_IFS" # mkdir can fail with a `File exist' error if two processes # try to create one of the directories concurrently. Don't # stop in that case! $MKDIR "$my_dir" 2>/dev/null || : done IFS="$save_mkdir_p_IFS" # Bail out if we (or some other process) failed to create a directory. test -d "$my_directory_path" || \ func_fatal_error "Failed to create \`$1'" fi } # func_mktempdir [string] # Make a temporary directory that won't clash with other running # libtool processes, and avoids race conditions if possible. If # given, STRING is the basename for that directory. func_mktempdir () { my_template="${TMPDIR-/tmp}/${1-$progname}" if test "$opt_dry_run" = ":"; then # Return a directory name, but don't create it in dry-run mode my_tmpdir="${my_template}-$$" else # If mktemp works, use that first and foremost my_tmpdir=`mktemp -d "${my_template}-XXXXXXXX" 2>/dev/null` if test ! -d "$my_tmpdir"; then # Failing that, at least try and use $RANDOM to avoid a race my_tmpdir="${my_template}-${RANDOM-0}$$" save_mktempdir_umask=`umask` umask 0077 $MKDIR "$my_tmpdir" umask $save_mktempdir_umask fi # If we're not in dry-run mode, bomb out on failure test -d "$my_tmpdir" || \ func_fatal_error "cannot create temporary directory \`$my_tmpdir'" fi $ECHO "$my_tmpdir" } # func_quote_for_eval arg # Aesthetically quote ARG to be evaled later. # This function returns two values: FUNC_QUOTE_FOR_EVAL_RESULT # is double-quoted, suitable for a subsequent eval, whereas # FUNC_QUOTE_FOR_EVAL_UNQUOTED_RESULT has merely all characters # which are still active within double quotes backslashified. func_quote_for_eval () { case $1 in *[\\\`\"\$]*) func_quote_for_eval_unquoted_result=`$ECHO "$1" | $SED "$sed_quote_subst"` ;; *) func_quote_for_eval_unquoted_result="$1" ;; esac case $func_quote_for_eval_unquoted_result in # Double-quote args containing shell metacharacters to delay # word splitting, command substitution and and variable # expansion for a subsequent eval. # Many Bourne shells cannot handle close brackets correctly # in scan sets, so we specify it separately. *[\[\~\#\^\&\*\(\)\{\}\|\;\<\>\?\'\ \ ]*|*]*|"") func_quote_for_eval_result="\"$func_quote_for_eval_unquoted_result\"" ;; *) func_quote_for_eval_result="$func_quote_for_eval_unquoted_result" esac } # func_quote_for_expand arg # Aesthetically quote ARG to be evaled later; same as above, # but do not quote variable references. func_quote_for_expand () { case $1 in *[\\\`\"]*) my_arg=`$ECHO "$1" | $SED \ -e "$double_quote_subst" -e "$sed_double_backslash"` ;; *) my_arg="$1" ;; esac case $my_arg in # Double-quote args containing shell metacharacters to delay # word splitting and command substitution for a subsequent eval. # Many Bourne shells cannot handle close brackets correctly # in scan sets, so we specify it separately. *[\[\~\#\^\&\*\(\)\{\}\|\;\<\>\?\'\ \ ]*|*]*|"") my_arg="\"$my_arg\"" ;; esac func_quote_for_expand_result="$my_arg" } # func_show_eval cmd [fail_exp] # Unless opt_silent is true, then output CMD. Then, if opt_dryrun is # not true, evaluate CMD. If the evaluation of CMD fails, and FAIL_EXP # is given, then evaluate it. func_show_eval () { my_cmd="$1" my_fail_exp="${2-:}" ${opt_silent-false} || { func_quote_for_expand "$my_cmd" eval "func_echo $func_quote_for_expand_result" } if ${opt_dry_run-false}; then :; else eval "$my_cmd" my_status=$? if test "$my_status" -eq 0; then :; else eval "(exit $my_status); $my_fail_exp" fi fi } # func_show_eval_locale cmd [fail_exp] # Unless opt_silent is true, then output CMD. Then, if opt_dryrun is # not true, evaluate CMD. If the evaluation of CMD fails, and FAIL_EXP # is given, then evaluate it. Use the saved locale for evaluation. func_show_eval_locale () { my_cmd="$1" my_fail_exp="${2-:}" ${opt_silent-false} || { func_quote_for_expand "$my_cmd" eval "func_echo $func_quote_for_expand_result" } if ${opt_dry_run-false}; then :; else eval "$lt_user_locale $my_cmd" my_status=$? eval "$lt_safe_locale" if test "$my_status" -eq 0; then :; else eval "(exit $my_status); $my_fail_exp" fi fi } # func_tr_sh # Turn $1 into a string suitable for a shell variable name. # Result is stored in $func_tr_sh_result. All characters # not in the set a-zA-Z0-9_ are replaced with '_'. Further, # if $1 begins with a digit, a '_' is prepended as well. func_tr_sh () { case $1 in [0-9]* | *[!a-zA-Z0-9_]*) func_tr_sh_result=`$ECHO "$1" | $SED 's/^\([0-9]\)/_\1/; s/[^a-zA-Z0-9_]/_/g'` ;; * ) func_tr_sh_result=$1 ;; esac } # func_version # Echo version message to standard output and exit. func_version () { $opt_debug $SED -n '/(C)/!b go :more /\./!{ N s/\n# / / b more } :go /^# '$PROGRAM' (GNU /,/# warranty; / { s/^# // s/^# *$// s/\((C)\)[ 0-9,-]*\( [1-9][0-9]*\)/\1\2/ p }' < "$progpath" exit $? } # func_usage # Echo short help message to standard output and exit. func_usage () { $opt_debug $SED -n '/^# Usage:/,/^# *.*--help/ { s/^# // s/^# *$// s/\$progname/'$progname'/ p }' < "$progpath" echo $ECHO "run \`$progname --help | more' for full usage" exit $? } # func_help [NOEXIT] # Echo long help message to standard output and exit, # unless 'noexit' is passed as argument. func_help () { $opt_debug $SED -n '/^# Usage:/,/# Report bugs to/ { :print s/^# // s/^# *$// s*\$progname*'$progname'* s*\$host*'"$host"'* s*\$SHELL*'"$SHELL"'* s*\$LTCC*'"$LTCC"'* s*\$LTCFLAGS*'"$LTCFLAGS"'* s*\$LD*'"$LD"'* s/\$with_gnu_ld/'"$with_gnu_ld"'/ s/\$automake_version/'"`(${AUTOMAKE-automake} --version) 2>/dev/null |$SED 1q`"'/ s/\$autoconf_version/'"`(${AUTOCONF-autoconf} --version) 2>/dev/null |$SED 1q`"'/ p d } /^# .* home page:/b print /^# General help using/b print ' < "$progpath" ret=$? if test -z "$1"; then exit $ret fi } # func_missing_arg argname # Echo program name prefixed message to standard error and set global # exit_cmd. func_missing_arg () { $opt_debug func_error "missing argument for $1." exit_cmd=exit } # func_split_short_opt shortopt # Set func_split_short_opt_name and func_split_short_opt_arg shell # variables after splitting SHORTOPT after the 2nd character. func_split_short_opt () { my_sed_short_opt='1s/^\(..\).*$/\1/;q' my_sed_short_rest='1s/^..\(.*\)$/\1/;q' func_split_short_opt_name=`$ECHO "$1" | $SED "$my_sed_short_opt"` func_split_short_opt_arg=`$ECHO "$1" | $SED "$my_sed_short_rest"` } # func_split_short_opt may be replaced by extended shell implementation # func_split_long_opt longopt # Set func_split_long_opt_name and func_split_long_opt_arg shell # variables after splitting LONGOPT at the `=' sign. func_split_long_opt () { my_sed_long_opt='1s/^\(--[^=]*\)=.*/\1/;q' my_sed_long_arg='1s/^--[^=]*=//' func_split_long_opt_name=`$ECHO "$1" | $SED "$my_sed_long_opt"` func_split_long_opt_arg=`$ECHO "$1" | $SED "$my_sed_long_arg"` } # func_split_long_opt may be replaced by extended shell implementation exit_cmd=: magic="%%%MAGIC variable%%%" magic_exe="%%%MAGIC EXE variable%%%" # Global variables. nonopt= preserve_args= lo2o="s/\\.lo\$/.${objext}/" o2lo="s/\\.${objext}\$/.lo/" extracted_archives= extracted_serial=0 # If this variable is set in any of the actions, the command in it # will be execed at the end. This prevents here-documents from being # left over by shells. exec_cmd= # func_append var value # Append VALUE to the end of shell variable VAR. func_append () { eval "${1}=\$${1}\${2}" } # func_append may be replaced by extended shell implementation # func_append_quoted var value # Quote VALUE and append to the end of shell variable VAR, separated # by a space. func_append_quoted () { func_quote_for_eval "${2}" eval "${1}=\$${1}\\ \$func_quote_for_eval_result" } # func_append_quoted may be replaced by extended shell implementation # func_arith arithmetic-term... func_arith () { func_arith_result=`expr "${@}"` } # func_arith may be replaced by extended shell implementation # func_len string # STRING may not start with a hyphen. func_len () { func_len_result=`expr "${1}" : ".*" 2>/dev/null || echo $max_cmd_len` } # func_len may be replaced by extended shell implementation # func_lo2o object func_lo2o () { func_lo2o_result=`$ECHO "${1}" | $SED "$lo2o"` } # func_lo2o may be replaced by extended shell implementation # func_xform libobj-or-source func_xform () { func_xform_result=`$ECHO "${1}" | $SED 's/\.[^.]*$/.lo/'` } # func_xform may be replaced by extended shell implementation # func_fatal_configuration arg... # Echo program name prefixed message to standard error, followed by # a configuration failure hint, and exit. func_fatal_configuration () { func_error ${1+"$@"} func_error "See the $PACKAGE documentation for more information." func_fatal_error "Fatal configuration error." } # func_config # Display the configuration for all the tags in this script. func_config () { re_begincf='^# ### BEGIN LIBTOOL' re_endcf='^# ### END LIBTOOL' # Default configuration. $SED "1,/$re_begincf CONFIG/d;/$re_endcf CONFIG/,\$d" < "$progpath" # Now print the configurations for the tags. for tagname in $taglist; do $SED -n "/$re_begincf TAG CONFIG: $tagname\$/,/$re_endcf TAG CONFIG: $tagname\$/p" < "$progpath" done exit $? } # func_features # Display the features supported by this script. func_features () { echo "host: $host" if test "$build_libtool_libs" = yes; then echo "enable shared libraries" else echo "disable shared libraries" fi if test "$build_old_libs" = yes; then echo "enable static libraries" else echo "disable static libraries" fi exit $? } # func_enable_tag tagname # Verify that TAGNAME is valid, and either flag an error and exit, or # enable the TAGNAME tag. We also add TAGNAME to the global $taglist # variable here. func_enable_tag () { # Global variable: tagname="$1" re_begincf="^# ### BEGIN LIBTOOL TAG CONFIG: $tagname\$" re_endcf="^# ### END LIBTOOL TAG CONFIG: $tagname\$" sed_extractcf="/$re_begincf/,/$re_endcf/p" # Validate tagname. case $tagname in *[!-_A-Za-z0-9,/]*) func_fatal_error "invalid tag name: $tagname" ;; esac # Don't test for the "default" C tag, as we know it's # there but not specially marked. case $tagname in CC) ;; *) if $GREP "$re_begincf" "$progpath" >/dev/null 2>&1; then taglist="$taglist $tagname" # Evaluate the configuration. Be careful to quote the path # and the sed script, to avoid splitting on whitespace, but # also don't use non-portable quotes within backquotes within # quotes we have to do it in 2 steps: extractedcf=`$SED -n -e "$sed_extractcf" < "$progpath"` eval "$extractedcf" else func_error "ignoring unknown tag $tagname" fi ;; esac } # func_check_version_match # Ensure that we are using m4 macros, and libtool script from the same # release of libtool. func_check_version_match () { if test "$package_revision" != "$macro_revision"; then if test "$VERSION" != "$macro_version"; then if test -z "$macro_version"; then cat >&2 <<_LT_EOF $progname: Version mismatch error. This is $PACKAGE $VERSION, but the $progname: definition of this LT_INIT comes from an older release. $progname: You should recreate aclocal.m4 with macros from $PACKAGE $VERSION $progname: and run autoconf again. _LT_EOF else cat >&2 <<_LT_EOF $progname: Version mismatch error. This is $PACKAGE $VERSION, but the $progname: definition of this LT_INIT comes from $PACKAGE $macro_version. $progname: You should recreate aclocal.m4 with macros from $PACKAGE $VERSION $progname: and run autoconf again. _LT_EOF fi else cat >&2 <<_LT_EOF $progname: Version mismatch error. This is $PACKAGE $VERSION, revision $package_revision, $progname: but the definition of this LT_INIT comes from revision $macro_revision. $progname: You should recreate aclocal.m4 with macros from revision $package_revision $progname: of $PACKAGE $VERSION and run autoconf again. _LT_EOF fi exit $EXIT_MISMATCH fi } # Shorthand for --mode=foo, only valid as the first argument case $1 in clean|clea|cle|cl) shift; set dummy --mode clean ${1+"$@"}; shift ;; compile|compil|compi|comp|com|co|c) shift; set dummy --mode compile ${1+"$@"}; shift ;; execute|execut|execu|exec|exe|ex|e) shift; set dummy --mode execute ${1+"$@"}; shift ;; finish|finis|fini|fin|fi|f) shift; set dummy --mode finish ${1+"$@"}; shift ;; install|instal|insta|inst|ins|in|i) shift; set dummy --mode install ${1+"$@"}; shift ;; link|lin|li|l) shift; set dummy --mode link ${1+"$@"}; shift ;; uninstall|uninstal|uninsta|uninst|unins|unin|uni|un|u) shift; set dummy --mode uninstall ${1+"$@"}; shift ;; esac # Option defaults: opt_debug=: opt_dry_run=false opt_config=false opt_preserve_dup_deps=false opt_features=false opt_finish=false opt_help=false opt_help_all=false opt_silent=: opt_warning=: opt_verbose=: opt_silent=false opt_verbose=false # Parse options once, thoroughly. This comes as soon as possible in the # script to make things like `--version' happen as quickly as we can. { # this just eases exit handling while test $# -gt 0; do opt="$1" shift case $opt in --debug|-x) opt_debug='set -x' func_echo "enabling shell trace mode" $opt_debug ;; --dry-run|--dryrun|-n) opt_dry_run=: ;; --config) opt_config=: func_config ;; --dlopen|-dlopen) optarg="$1" opt_dlopen="${opt_dlopen+$opt_dlopen }$optarg" shift ;; --preserve-dup-deps) opt_preserve_dup_deps=: ;; --features) opt_features=: func_features ;; --finish) opt_finish=: set dummy --mode finish ${1+"$@"}; shift ;; --help) opt_help=: ;; --help-all) opt_help_all=: opt_help=': help-all' ;; --mode) test $# = 0 && func_missing_arg $opt && break optarg="$1" opt_mode="$optarg" case $optarg in # Valid mode arguments: clean|compile|execute|finish|install|link|relink|uninstall) ;; # Catch anything else as an error *) func_error "invalid argument for $opt" exit_cmd=exit break ;; esac shift ;; --no-silent|--no-quiet) opt_silent=false func_append preserve_args " $opt" ;; --no-warning|--no-warn) opt_warning=false func_append preserve_args " $opt" ;; --no-verbose) opt_verbose=false func_append preserve_args " $opt" ;; --silent|--quiet) opt_silent=: func_append preserve_args " $opt" opt_verbose=false ;; --verbose|-v) opt_verbose=: func_append preserve_args " $opt" opt_silent=false ;; --tag) test $# = 0 && func_missing_arg $opt && break optarg="$1" opt_tag="$optarg" func_append preserve_args " $opt $optarg" func_enable_tag "$optarg" shift ;; -\?|-h) func_usage ;; --help) func_help ;; --version) func_version ;; # Separate optargs to long options: --*=*) func_split_long_opt "$opt" set dummy "$func_split_long_opt_name" "$func_split_long_opt_arg" ${1+"$@"} shift ;; # Separate non-argument short options: -\?*|-h*|-n*|-v*) func_split_short_opt "$opt" set dummy "$func_split_short_opt_name" "-$func_split_short_opt_arg" ${1+"$@"} shift ;; --) break ;; -*) func_fatal_help "unrecognized option \`$opt'" ;; *) set dummy "$opt" ${1+"$@"}; shift; break ;; esac done # Validate options: # save first non-option argument if test "$#" -gt 0; then nonopt="$opt" shift fi # preserve --debug test "$opt_debug" = : || func_append preserve_args " --debug" case $host in *cygwin* | *mingw* | *pw32* | *cegcc*) # don't eliminate duplications in $postdeps and $predeps opt_duplicate_compiler_generated_deps=: ;; *) opt_duplicate_compiler_generated_deps=$opt_preserve_dup_deps ;; esac $opt_help || { # Sanity checks first: func_check_version_match if test "$build_libtool_libs" != yes && test "$build_old_libs" != yes; then func_fatal_configuration "not configured to build any kind of library" fi # Darwin sucks eval std_shrext=\"$shrext_cmds\" # Only execute mode is allowed to have -dlopen flags. if test -n "$opt_dlopen" && test "$opt_mode" != execute; then func_error "unrecognized option \`-dlopen'" $ECHO "$help" 1>&2 exit $EXIT_FAILURE fi # Change the help message to a mode-specific one. generic_help="$help" help="Try \`$progname --help --mode=$opt_mode' for more information." } # Bail if the options were screwed $exit_cmd $EXIT_FAILURE } ## ----------- ## ## Main. ## ## ----------- ## # func_lalib_p file # True iff FILE is a libtool `.la' library or `.lo' object file. # This function is only a basic sanity check; it will hardly flush out # determined imposters. func_lalib_p () { test -f "$1" && $SED -e 4q "$1" 2>/dev/null \ | $GREP "^# Generated by .*$PACKAGE" > /dev/null 2>&1 } # func_lalib_unsafe_p file # True iff FILE is a libtool `.la' library or `.lo' object file. # This function implements the same check as func_lalib_p without # resorting to external programs. To this end, it redirects stdin and # closes it afterwards, without saving the original file descriptor. # As a safety measure, use it only where a negative result would be # fatal anyway. Works if `file' does not exist. func_lalib_unsafe_p () { lalib_p=no if test -f "$1" && test -r "$1" && exec 5<&0 <"$1"; then for lalib_p_l in 1 2 3 4 do read lalib_p_line case "$lalib_p_line" in \#\ Generated\ by\ *$PACKAGE* ) lalib_p=yes; break;; esac done exec 0<&5 5<&- fi test "$lalib_p" = yes } # func_ltwrapper_script_p file # True iff FILE is a libtool wrapper script # This function is only a basic sanity check; it will hardly flush out # determined imposters. func_ltwrapper_script_p () { func_lalib_p "$1" } # func_ltwrapper_executable_p file # True iff FILE is a libtool wrapper executable # This function is only a basic sanity check; it will hardly flush out # determined imposters. func_ltwrapper_executable_p () { func_ltwrapper_exec_suffix= case $1 in *.exe) ;; *) func_ltwrapper_exec_suffix=.exe ;; esac $GREP "$magic_exe" "$1$func_ltwrapper_exec_suffix" >/dev/null 2>&1 } # func_ltwrapper_scriptname file # Assumes file is an ltwrapper_executable # uses $file to determine the appropriate filename for a # temporary ltwrapper_script. func_ltwrapper_scriptname () { func_dirname_and_basename "$1" "" "." func_stripname '' '.exe' "$func_basename_result" func_ltwrapper_scriptname_result="$func_dirname_result/$objdir/${func_stripname_result}_ltshwrapper" } # func_ltwrapper_p file # True iff FILE is a libtool wrapper script or wrapper executable # This function is only a basic sanity check; it will hardly flush out # determined imposters. func_ltwrapper_p () { func_ltwrapper_script_p "$1" || func_ltwrapper_executable_p "$1" } # func_execute_cmds commands fail_cmd # Execute tilde-delimited COMMANDS. # If FAIL_CMD is given, eval that upon failure. # FAIL_CMD may read-access the current command in variable CMD! func_execute_cmds () { $opt_debug save_ifs=$IFS; IFS='~' for cmd in $1; do IFS=$save_ifs eval cmd=\"$cmd\" func_show_eval "$cmd" "${2-:}" done IFS=$save_ifs } # func_source file # Source FILE, adding directory component if necessary. # Note that it is not necessary on cygwin/mingw to append a dot to # FILE even if both FILE and FILE.exe exist: automatic-append-.exe # behavior happens only for exec(3), not for open(2)! Also, sourcing # `FILE.' does not work on cygwin managed mounts. func_source () { $opt_debug case $1 in */* | *\\*) . "$1" ;; *) . "./$1" ;; esac } # func_resolve_sysroot PATH # Replace a leading = in PATH with a sysroot. Store the result into # func_resolve_sysroot_result func_resolve_sysroot () { func_resolve_sysroot_result=$1 case $func_resolve_sysroot_result in =*) func_stripname '=' '' "$func_resolve_sysroot_result" func_resolve_sysroot_result=$lt_sysroot$func_stripname_result ;; esac } # func_replace_sysroot PATH # If PATH begins with the sysroot, replace it with = and # store the result into func_replace_sysroot_result. func_replace_sysroot () { case "$lt_sysroot:$1" in ?*:"$lt_sysroot"*) func_stripname "$lt_sysroot" '' "$1" func_replace_sysroot_result="=$func_stripname_result" ;; *) # Including no sysroot. func_replace_sysroot_result=$1 ;; esac } # func_infer_tag arg # Infer tagged configuration to use if any are available and # if one wasn't chosen via the "--tag" command line option. # Only attempt this if the compiler in the base compile # command doesn't match the default compiler. # arg is usually of the form 'gcc ...' func_infer_tag () { $opt_debug if test -n "$available_tags" && test -z "$tagname"; then CC_quoted= for arg in $CC; do func_append_quoted CC_quoted "$arg" done CC_expanded=`func_echo_all $CC` CC_quoted_expanded=`func_echo_all $CC_quoted` case $@ in # Blanks in the command may have been stripped by the calling shell, # but not from the CC environment variable when configure was run. " $CC "* | "$CC "* | " $CC_expanded "* | "$CC_expanded "* | \ " $CC_quoted"* | "$CC_quoted "* | " $CC_quoted_expanded "* | "$CC_quoted_expanded "*) ;; # Blanks at the start of $base_compile will cause this to fail # if we don't check for them as well. *) for z in $available_tags; do if $GREP "^# ### BEGIN LIBTOOL TAG CONFIG: $z$" < "$progpath" > /dev/null; then # Evaluate the configuration. eval "`${SED} -n -e '/^# ### BEGIN LIBTOOL TAG CONFIG: '$z'$/,/^# ### END LIBTOOL TAG CONFIG: '$z'$/p' < $progpath`" CC_quoted= for arg in $CC; do # Double-quote args containing other shell metacharacters. func_append_quoted CC_quoted "$arg" done CC_expanded=`func_echo_all $CC` CC_quoted_expanded=`func_echo_all $CC_quoted` case "$@ " in " $CC "* | "$CC "* | " $CC_expanded "* | "$CC_expanded "* | \ " $CC_quoted"* | "$CC_quoted "* | " $CC_quoted_expanded "* | "$CC_quoted_expanded "*) # The compiler in the base compile command matches # the one in the tagged configuration. # Assume this is the tagged configuration we want. tagname=$z break ;; esac fi done # If $tagname still isn't set, then no tagged configuration # was found and let the user know that the "--tag" command # line option must be used. if test -z "$tagname"; then func_echo "unable to infer tagged configuration" func_fatal_error "specify a tag with \`--tag'" # else # func_verbose "using $tagname tagged configuration" fi ;; esac fi } # func_write_libtool_object output_name pic_name nonpic_name # Create a libtool object file (analogous to a ".la" file), # but don't create it if we're doing a dry run. func_write_libtool_object () { write_libobj=${1} if test "$build_libtool_libs" = yes; then write_lobj=\'${2}\' else write_lobj=none fi if test "$build_old_libs" = yes; then write_oldobj=\'${3}\' else write_oldobj=none fi $opt_dry_run || { cat >${write_libobj}T </dev/null` if test "$?" -eq 0 && test -n "${func_convert_core_file_wine_to_w32_tmp}"; then func_convert_core_file_wine_to_w32_result=`$ECHO "$func_convert_core_file_wine_to_w32_tmp" | $SED -e "$lt_sed_naive_backslashify"` else func_convert_core_file_wine_to_w32_result= fi fi } # end: func_convert_core_file_wine_to_w32 # func_convert_core_path_wine_to_w32 ARG # Helper function used by path conversion functions when $build is *nix, and # $host is mingw, cygwin, or some other w32 environment. Relies on a correctly # configured wine environment available, with the winepath program in $build's # $PATH. Assumes ARG has no leading or trailing path separator characters. # # ARG is path to be converted from $build format to win32. # Result is available in $func_convert_core_path_wine_to_w32_result. # Unconvertible file (directory) names in ARG are skipped; if no directory names # are convertible, then the result may be empty. func_convert_core_path_wine_to_w32 () { $opt_debug # unfortunately, winepath doesn't convert paths, only file names func_convert_core_path_wine_to_w32_result="" if test -n "$1"; then oldIFS=$IFS IFS=: for func_convert_core_path_wine_to_w32_f in $1; do IFS=$oldIFS func_convert_core_file_wine_to_w32 "$func_convert_core_path_wine_to_w32_f" if test -n "$func_convert_core_file_wine_to_w32_result" ; then if test -z "$func_convert_core_path_wine_to_w32_result"; then func_convert_core_path_wine_to_w32_result="$func_convert_core_file_wine_to_w32_result" else func_append func_convert_core_path_wine_to_w32_result ";$func_convert_core_file_wine_to_w32_result" fi fi done IFS=$oldIFS fi } # end: func_convert_core_path_wine_to_w32 # func_cygpath ARGS... # Wrapper around calling the cygpath program via LT_CYGPATH. This is used when # when (1) $build is *nix and Cygwin is hosted via a wine environment; or (2) # $build is MSYS and $host is Cygwin, or (3) $build is Cygwin. In case (1) or # (2), returns the Cygwin file name or path in func_cygpath_result (input # file name or path is assumed to be in w32 format, as previously converted # from $build's *nix or MSYS format). In case (3), returns the w32 file name # or path in func_cygpath_result (input file name or path is assumed to be in # Cygwin format). Returns an empty string on error. # # ARGS are passed to cygpath, with the last one being the file name or path to # be converted. # # Specify the absolute *nix (or w32) name to cygpath in the LT_CYGPATH # environment variable; do not put it in $PATH. func_cygpath () { $opt_debug if test -n "$LT_CYGPATH" && test -f "$LT_CYGPATH"; then func_cygpath_result=`$LT_CYGPATH "$@" 2>/dev/null` if test "$?" -ne 0; then # on failure, ensure result is empty func_cygpath_result= fi else func_cygpath_result= func_error "LT_CYGPATH is empty or specifies non-existent file: \`$LT_CYGPATH'" fi } #end: func_cygpath # func_convert_core_msys_to_w32 ARG # Convert file name or path ARG from MSYS format to w32 format. Return # result in func_convert_core_msys_to_w32_result. func_convert_core_msys_to_w32 () { $opt_debug # awkward: cmd appends spaces to result func_convert_core_msys_to_w32_result=`( cmd //c echo "$1" ) 2>/dev/null | $SED -e 's/[ ]*$//' -e "$lt_sed_naive_backslashify"` } #end: func_convert_core_msys_to_w32 # func_convert_file_check ARG1 ARG2 # Verify that ARG1 (a file name in $build format) was converted to $host # format in ARG2. Otherwise, emit an error message, but continue (resetting # func_to_host_file_result to ARG1). func_convert_file_check () { $opt_debug if test -z "$2" && test -n "$1" ; then func_error "Could not determine host file name corresponding to" func_error " \`$1'" func_error "Continuing, but uninstalled executables may not work." # Fallback: func_to_host_file_result="$1" fi } # end func_convert_file_check # func_convert_path_check FROM_PATHSEP TO_PATHSEP FROM_PATH TO_PATH # Verify that FROM_PATH (a path in $build format) was converted to $host # format in TO_PATH. Otherwise, emit an error message, but continue, resetting # func_to_host_file_result to a simplistic fallback value (see below). func_convert_path_check () { $opt_debug if test -z "$4" && test -n "$3"; then func_error "Could not determine the host path corresponding to" func_error " \`$3'" func_error "Continuing, but uninstalled executables may not work." # Fallback. This is a deliberately simplistic "conversion" and # should not be "improved". See libtool.info. if test "x$1" != "x$2"; then lt_replace_pathsep_chars="s|$1|$2|g" func_to_host_path_result=`echo "$3" | $SED -e "$lt_replace_pathsep_chars"` else func_to_host_path_result="$3" fi fi } # end func_convert_path_check # func_convert_path_front_back_pathsep FRONTPAT BACKPAT REPL ORIG # Modifies func_to_host_path_result by prepending REPL if ORIG matches FRONTPAT # and appending REPL if ORIG matches BACKPAT. func_convert_path_front_back_pathsep () { $opt_debug case $4 in $1 ) func_to_host_path_result="$3$func_to_host_path_result" ;; esac case $4 in $2 ) func_append func_to_host_path_result "$3" ;; esac } # end func_convert_path_front_back_pathsep ################################################## # $build to $host FILE NAME CONVERSION FUNCTIONS # ################################################## # invoked via `$to_host_file_cmd ARG' # # In each case, ARG is the path to be converted from $build to $host format. # Result will be available in $func_to_host_file_result. # func_to_host_file ARG # Converts the file name ARG from $build format to $host format. Return result # in func_to_host_file_result. func_to_host_file () { $opt_debug $to_host_file_cmd "$1" } # end func_to_host_file # func_to_tool_file ARG LAZY # converts the file name ARG from $build format to toolchain format. Return # result in func_to_tool_file_result. If the conversion in use is listed # in (the comma separated) LAZY, no conversion takes place. func_to_tool_file () { $opt_debug case ,$2, in *,"$to_tool_file_cmd",*) func_to_tool_file_result=$1 ;; *) $to_tool_file_cmd "$1" func_to_tool_file_result=$func_to_host_file_result ;; esac } # end func_to_tool_file # func_convert_file_noop ARG # Copy ARG to func_to_host_file_result. func_convert_file_noop () { func_to_host_file_result="$1" } # end func_convert_file_noop # func_convert_file_msys_to_w32 ARG # Convert file name ARG from (mingw) MSYS to (mingw) w32 format; automatic # conversion to w32 is not available inside the cwrapper. Returns result in # func_to_host_file_result. func_convert_file_msys_to_w32 () { $opt_debug func_to_host_file_result="$1" if test -n "$1"; then func_convert_core_msys_to_w32 "$1" func_to_host_file_result="$func_convert_core_msys_to_w32_result" fi func_convert_file_check "$1" "$func_to_host_file_result" } # end func_convert_file_msys_to_w32 # func_convert_file_cygwin_to_w32 ARG # Convert file name ARG from Cygwin to w32 format. Returns result in # func_to_host_file_result. func_convert_file_cygwin_to_w32 () { $opt_debug func_to_host_file_result="$1" if test -n "$1"; then # because $build is cygwin, we call "the" cygpath in $PATH; no need to use # LT_CYGPATH in this case. func_to_host_file_result=`cygpath -m "$1"` fi func_convert_file_check "$1" "$func_to_host_file_result" } # end func_convert_file_cygwin_to_w32 # func_convert_file_nix_to_w32 ARG # Convert file name ARG from *nix to w32 format. Requires a wine environment # and a working winepath. Returns result in func_to_host_file_result. func_convert_file_nix_to_w32 () { $opt_debug func_to_host_file_result="$1" if test -n "$1"; then func_convert_core_file_wine_to_w32 "$1" func_to_host_file_result="$func_convert_core_file_wine_to_w32_result" fi func_convert_file_check "$1" "$func_to_host_file_result" } # end func_convert_file_nix_to_w32 # func_convert_file_msys_to_cygwin ARG # Convert file name ARG from MSYS to Cygwin format. Requires LT_CYGPATH set. # Returns result in func_to_host_file_result. func_convert_file_msys_to_cygwin () { $opt_debug func_to_host_file_result="$1" if test -n "$1"; then func_convert_core_msys_to_w32 "$1" func_cygpath -u "$func_convert_core_msys_to_w32_result" func_to_host_file_result="$func_cygpath_result" fi func_convert_file_check "$1" "$func_to_host_file_result" } # end func_convert_file_msys_to_cygwin # func_convert_file_nix_to_cygwin ARG # Convert file name ARG from *nix to Cygwin format. Requires Cygwin installed # in a wine environment, working winepath, and LT_CYGPATH set. Returns result # in func_to_host_file_result. func_convert_file_nix_to_cygwin () { $opt_debug func_to_host_file_result="$1" if test -n "$1"; then # convert from *nix to w32, then use cygpath to convert from w32 to cygwin. func_convert_core_file_wine_to_w32 "$1" func_cygpath -u "$func_convert_core_file_wine_to_w32_result" func_to_host_file_result="$func_cygpath_result" fi func_convert_file_check "$1" "$func_to_host_file_result" } # end func_convert_file_nix_to_cygwin ############################################# # $build to $host PATH CONVERSION FUNCTIONS # ############################################# # invoked via `$to_host_path_cmd ARG' # # In each case, ARG is the path to be converted from $build to $host format. # The result will be available in $func_to_host_path_result. # # Path separators are also converted from $build format to $host format. If # ARG begins or ends with a path separator character, it is preserved (but # converted to $host format) on output. # # All path conversion functions are named using the following convention: # file name conversion function : func_convert_file_X_to_Y () # path conversion function : func_convert_path_X_to_Y () # where, for any given $build/$host combination the 'X_to_Y' value is the # same. If conversion functions are added for new $build/$host combinations, # the two new functions must follow this pattern, or func_init_to_host_path_cmd # will break. # func_init_to_host_path_cmd # Ensures that function "pointer" variable $to_host_path_cmd is set to the # appropriate value, based on the value of $to_host_file_cmd. to_host_path_cmd= func_init_to_host_path_cmd () { $opt_debug if test -z "$to_host_path_cmd"; then func_stripname 'func_convert_file_' '' "$to_host_file_cmd" to_host_path_cmd="func_convert_path_${func_stripname_result}" fi } # func_to_host_path ARG # Converts the path ARG from $build format to $host format. Return result # in func_to_host_path_result. func_to_host_path () { $opt_debug func_init_to_host_path_cmd $to_host_path_cmd "$1" } # end func_to_host_path # func_convert_path_noop ARG # Copy ARG to func_to_host_path_result. func_convert_path_noop () { func_to_host_path_result="$1" } # end func_convert_path_noop # func_convert_path_msys_to_w32 ARG # Convert path ARG from (mingw) MSYS to (mingw) w32 format; automatic # conversion to w32 is not available inside the cwrapper. Returns result in # func_to_host_path_result. func_convert_path_msys_to_w32 () { $opt_debug func_to_host_path_result="$1" if test -n "$1"; then # Remove leading and trailing path separator characters from ARG. MSYS # behavior is inconsistent here; cygpath turns them into '.;' and ';.'; # and winepath ignores them completely. func_stripname : : "$1" func_to_host_path_tmp1=$func_stripname_result func_convert_core_msys_to_w32 "$func_to_host_path_tmp1" func_to_host_path_result="$func_convert_core_msys_to_w32_result" func_convert_path_check : ";" \ "$func_to_host_path_tmp1" "$func_to_host_path_result" func_convert_path_front_back_pathsep ":*" "*:" ";" "$1" fi } # end func_convert_path_msys_to_w32 # func_convert_path_cygwin_to_w32 ARG # Convert path ARG from Cygwin to w32 format. Returns result in # func_to_host_file_result. func_convert_path_cygwin_to_w32 () { $opt_debug func_to_host_path_result="$1" if test -n "$1"; then # See func_convert_path_msys_to_w32: func_stripname : : "$1" func_to_host_path_tmp1=$func_stripname_result func_to_host_path_result=`cygpath -m -p "$func_to_host_path_tmp1"` func_convert_path_check : ";" \ "$func_to_host_path_tmp1" "$func_to_host_path_result" func_convert_path_front_back_pathsep ":*" "*:" ";" "$1" fi } # end func_convert_path_cygwin_to_w32 # func_convert_path_nix_to_w32 ARG # Convert path ARG from *nix to w32 format. Requires a wine environment and # a working winepath. Returns result in func_to_host_file_result. func_convert_path_nix_to_w32 () { $opt_debug func_to_host_path_result="$1" if test -n "$1"; then # See func_convert_path_msys_to_w32: func_stripname : : "$1" func_to_host_path_tmp1=$func_stripname_result func_convert_core_path_wine_to_w32 "$func_to_host_path_tmp1" func_to_host_path_result="$func_convert_core_path_wine_to_w32_result" func_convert_path_check : ";" \ "$func_to_host_path_tmp1" "$func_to_host_path_result" func_convert_path_front_back_pathsep ":*" "*:" ";" "$1" fi } # end func_convert_path_nix_to_w32 # func_convert_path_msys_to_cygwin ARG # Convert path ARG from MSYS to Cygwin format. Requires LT_CYGPATH set. # Returns result in func_to_host_file_result. func_convert_path_msys_to_cygwin () { $opt_debug func_to_host_path_result="$1" if test -n "$1"; then # See func_convert_path_msys_to_w32: func_stripname : : "$1" func_to_host_path_tmp1=$func_stripname_result func_convert_core_msys_to_w32 "$func_to_host_path_tmp1" func_cygpath -u -p "$func_convert_core_msys_to_w32_result" func_to_host_path_result="$func_cygpath_result" func_convert_path_check : : \ "$func_to_host_path_tmp1" "$func_to_host_path_result" func_convert_path_front_back_pathsep ":*" "*:" : "$1" fi } # end func_convert_path_msys_to_cygwin # func_convert_path_nix_to_cygwin ARG # Convert path ARG from *nix to Cygwin format. Requires Cygwin installed in a # a wine environment, working winepath, and LT_CYGPATH set. Returns result in # func_to_host_file_result. func_convert_path_nix_to_cygwin () { $opt_debug func_to_host_path_result="$1" if test -n "$1"; then # Remove leading and trailing path separator characters from # ARG. msys behavior is inconsistent here, cygpath turns them # into '.;' and ';.', and winepath ignores them completely. func_stripname : : "$1" func_to_host_path_tmp1=$func_stripname_result func_convert_core_path_wine_to_w32 "$func_to_host_path_tmp1" func_cygpath -u -p "$func_convert_core_path_wine_to_w32_result" func_to_host_path_result="$func_cygpath_result" func_convert_path_check : : \ "$func_to_host_path_tmp1" "$func_to_host_path_result" func_convert_path_front_back_pathsep ":*" "*:" : "$1" fi } # end func_convert_path_nix_to_cygwin # func_mode_compile arg... func_mode_compile () { $opt_debug # Get the compilation command and the source file. base_compile= srcfile="$nonopt" # always keep a non-empty value in "srcfile" suppress_opt=yes suppress_output= arg_mode=normal libobj= later= pie_flag= for arg do case $arg_mode in arg ) # do not "continue". Instead, add this to base_compile lastarg="$arg" arg_mode=normal ;; target ) libobj="$arg" arg_mode=normal continue ;; normal ) # Accept any command-line options. case $arg in -o) test -n "$libobj" && \ func_fatal_error "you cannot specify \`-o' more than once" arg_mode=target continue ;; -pie | -fpie | -fPIE) func_append pie_flag " $arg" continue ;; -shared | -static | -prefer-pic | -prefer-non-pic) func_append later " $arg" continue ;; -no-suppress) suppress_opt=no continue ;; -Xcompiler) arg_mode=arg # the next one goes into the "base_compile" arg list continue # The current "srcfile" will either be retained or ;; # replaced later. I would guess that would be a bug. -Wc,*) func_stripname '-Wc,' '' "$arg" args=$func_stripname_result lastarg= save_ifs="$IFS"; IFS=',' for arg in $args; do IFS="$save_ifs" func_append_quoted lastarg "$arg" done IFS="$save_ifs" func_stripname ' ' '' "$lastarg" lastarg=$func_stripname_result # Add the arguments to base_compile. func_append base_compile " $lastarg" continue ;; *) # Accept the current argument as the source file. # The previous "srcfile" becomes the current argument. # lastarg="$srcfile" srcfile="$arg" ;; esac # case $arg ;; esac # case $arg_mode # Aesthetically quote the previous argument. func_append_quoted base_compile "$lastarg" done # for arg case $arg_mode in arg) func_fatal_error "you must specify an argument for -Xcompile" ;; target) func_fatal_error "you must specify a target with \`-o'" ;; *) # Get the name of the library object. test -z "$libobj" && { func_basename "$srcfile" libobj="$func_basename_result" } ;; esac # Recognize several different file suffixes. # If the user specifies -o file.o, it is replaced with file.lo case $libobj in *.[cCFSifmso] | \ *.ada | *.adb | *.ads | *.asm | \ *.c++ | *.cc | *.ii | *.class | *.cpp | *.cxx | \ *.[fF][09]? | *.for | *.java | *.go | *.obj | *.sx | *.cu | *.cup) func_xform "$libobj" libobj=$func_xform_result ;; esac case $libobj in *.lo) func_lo2o "$libobj"; obj=$func_lo2o_result ;; *) func_fatal_error "cannot determine name of library object from \`$libobj'" ;; esac func_infer_tag $base_compile for arg in $later; do case $arg in -shared) test "$build_libtool_libs" != yes && \ func_fatal_configuration "can not build a shared library" build_old_libs=no continue ;; -static) build_libtool_libs=no build_old_libs=yes continue ;; -prefer-pic) pic_mode=yes continue ;; -prefer-non-pic) pic_mode=no continue ;; esac done func_quote_for_eval "$libobj" test "X$libobj" != "X$func_quote_for_eval_result" \ && $ECHO "X$libobj" | $GREP '[]~#^*{};<>?"'"'"' &()|`$[]' \ && func_warning "libobj name \`$libobj' may not contain shell special characters." func_dirname_and_basename "$obj" "/" "" objname="$func_basename_result" xdir="$func_dirname_result" lobj=${xdir}$objdir/$objname test -z "$base_compile" && \ func_fatal_help "you must specify a compilation command" # Delete any leftover library objects. if test "$build_old_libs" = yes; then removelist="$obj $lobj $libobj ${libobj}T" else removelist="$lobj $libobj ${libobj}T" fi # On Cygwin there's no "real" PIC flag so we must build both object types case $host_os in cygwin* | mingw* | pw32* | os2* | cegcc*) pic_mode=default ;; esac if test "$pic_mode" = no && test "$deplibs_check_method" != pass_all; then # non-PIC code in shared libraries is not supported pic_mode=default fi # Calculate the filename of the output object if compiler does # not support -o with -c if test "$compiler_c_o" = no; then output_obj=`$ECHO "$srcfile" | $SED 's%^.*/%%; s%\.[^.]*$%%'`.${objext} lockfile="$output_obj.lock" else output_obj= need_locks=no lockfile= fi # Lock this critical section if it is needed # We use this script file to make the link, it avoids creating a new file if test "$need_locks" = yes; then until $opt_dry_run || ln "$progpath" "$lockfile" 2>/dev/null; do func_echo "Waiting for $lockfile to be removed" sleep 2 done elif test "$need_locks" = warn; then if test -f "$lockfile"; then $ECHO "\ *** ERROR, $lockfile exists and contains: `cat $lockfile 2>/dev/null` This indicates that another process is trying to use the same temporary object file, and libtool could not work around it because your compiler does not support \`-c' and \`-o' together. If you repeat this compilation, it may succeed, by chance, but you had better avoid parallel builds (make -j) in this platform, or get a better compiler." $opt_dry_run || $RM $removelist exit $EXIT_FAILURE fi func_append removelist " $output_obj" $ECHO "$srcfile" > "$lockfile" fi $opt_dry_run || $RM $removelist func_append removelist " $lockfile" trap '$opt_dry_run || $RM $removelist; exit $EXIT_FAILURE' 1 2 15 func_to_tool_file "$srcfile" func_convert_file_msys_to_w32 srcfile=$func_to_tool_file_result func_quote_for_eval "$srcfile" qsrcfile=$func_quote_for_eval_result # Only build a PIC object if we are building libtool libraries. if test "$build_libtool_libs" = yes; then # Without this assignment, base_compile gets emptied. fbsd_hideous_sh_bug=$base_compile if test "$pic_mode" != no; then command="$base_compile $qsrcfile $pic_flag" else # Don't build PIC code command="$base_compile $qsrcfile" fi func_mkdir_p "$xdir$objdir" if test -z "$output_obj"; then # Place PIC objects in $objdir func_append command " -o $lobj" fi func_show_eval_locale "$command" \ 'test -n "$output_obj" && $RM $removelist; exit $EXIT_FAILURE' if test "$need_locks" = warn && test "X`cat $lockfile 2>/dev/null`" != "X$srcfile"; then $ECHO "\ *** ERROR, $lockfile contains: `cat $lockfile 2>/dev/null` but it should contain: $srcfile This indicates that another process is trying to use the same temporary object file, and libtool could not work around it because your compiler does not support \`-c' and \`-o' together. If you repeat this compilation, it may succeed, by chance, but you had better avoid parallel builds (make -j) in this platform, or get a better compiler." $opt_dry_run || $RM $removelist exit $EXIT_FAILURE fi # Just move the object if needed, then go on to compile the next one if test -n "$output_obj" && test "X$output_obj" != "X$lobj"; then func_show_eval '$MV "$output_obj" "$lobj"' \ 'error=$?; $opt_dry_run || $RM $removelist; exit $error' fi # Allow error messages only from the first compilation. if test "$suppress_opt" = yes; then suppress_output=' >/dev/null 2>&1' fi fi # Only build a position-dependent object if we build old libraries. if test "$build_old_libs" = yes; then if test "$pic_mode" != yes; then # Don't build PIC code command="$base_compile $qsrcfile$pie_flag" else command="$base_compile $qsrcfile $pic_flag" fi if test "$compiler_c_o" = yes; then func_append command " -o $obj" fi # Suppress compiler output if we already did a PIC compilation. func_append command "$suppress_output" func_show_eval_locale "$command" \ '$opt_dry_run || $RM $removelist; exit $EXIT_FAILURE' if test "$need_locks" = warn && test "X`cat $lockfile 2>/dev/null`" != "X$srcfile"; then $ECHO "\ *** ERROR, $lockfile contains: `cat $lockfile 2>/dev/null` but it should contain: $srcfile This indicates that another process is trying to use the same temporary object file, and libtool could not work around it because your compiler does not support \`-c' and \`-o' together. If you repeat this compilation, it may succeed, by chance, but you had better avoid parallel builds (make -j) in this platform, or get a better compiler." $opt_dry_run || $RM $removelist exit $EXIT_FAILURE fi # Just move the object if needed if test -n "$output_obj" && test "X$output_obj" != "X$obj"; then func_show_eval '$MV "$output_obj" "$obj"' \ 'error=$?; $opt_dry_run || $RM $removelist; exit $error' fi fi $opt_dry_run || { func_write_libtool_object "$libobj" "$objdir/$objname" "$objname" # Unlock the critical section if it was locked if test "$need_locks" != no; then removelist=$lockfile $RM "$lockfile" fi } exit $EXIT_SUCCESS } $opt_help || { test "$opt_mode" = compile && func_mode_compile ${1+"$@"} } func_mode_help () { # We need to display help for each of the modes. case $opt_mode in "") # Generic help is extracted from the usage comments # at the start of this file. func_help ;; clean) $ECHO \ "Usage: $progname [OPTION]... --mode=clean RM [RM-OPTION]... FILE... Remove files from the build directory. RM is the name of the program to use to delete files associated with each FILE (typically \`/bin/rm'). RM-OPTIONS are options (such as \`-f') to be passed to RM. If FILE is a libtool library, object or program, all the files associated with it are deleted. Otherwise, only FILE itself is deleted using RM." ;; compile) $ECHO \ "Usage: $progname [OPTION]... --mode=compile COMPILE-COMMAND... SOURCEFILE Compile a source file into a libtool library object. This mode accepts the following additional options: -o OUTPUT-FILE set the output file name to OUTPUT-FILE -no-suppress do not suppress compiler output for multiple passes -prefer-pic try to build PIC objects only -prefer-non-pic try to build non-PIC objects only -shared do not build a \`.o' file suitable for static linking -static only build a \`.o' file suitable for static linking -Wc,FLAG pass FLAG directly to the compiler COMPILE-COMMAND is a command to be used in creating a \`standard' object file from the given SOURCEFILE. The output file name is determined by removing the directory component from SOURCEFILE, then substituting the C source code suffix \`.c' with the library object suffix, \`.lo'." ;; execute) $ECHO \ "Usage: $progname [OPTION]... --mode=execute COMMAND [ARGS]... Automatically set library path, then run a program. This mode accepts the following additional options: -dlopen FILE add the directory containing FILE to the library path This mode sets the library path environment variable according to \`-dlopen' flags. If any of the ARGS are libtool executable wrappers, then they are translated into their corresponding uninstalled binary, and any of their required library directories are added to the library path. Then, COMMAND is executed, with ARGS as arguments." ;; finish) $ECHO \ "Usage: $progname [OPTION]... --mode=finish [LIBDIR]... Complete the installation of libtool libraries. Each LIBDIR is a directory that contains libtool libraries. The commands that this mode executes may require superuser privileges. Use the \`--dry-run' option if you just want to see what would be executed." ;; install) $ECHO \ "Usage: $progname [OPTION]... --mode=install INSTALL-COMMAND... Install executables or libraries. INSTALL-COMMAND is the installation command. The first component should be either the \`install' or \`cp' program. The following components of INSTALL-COMMAND are treated specially: -inst-prefix-dir PREFIX-DIR Use PREFIX-DIR as a staging area for installation The rest of the components are interpreted as arguments to that command (only BSD-compatible install options are recognized)." ;; link) $ECHO \ "Usage: $progname [OPTION]... --mode=link LINK-COMMAND... Link object files or libraries together to form another library, or to create an executable program. LINK-COMMAND is a command using the C compiler that you would use to create a program from several object files. The following components of LINK-COMMAND are treated specially: -all-static do not do any dynamic linking at all -avoid-version do not add a version suffix if possible -bindir BINDIR specify path to binaries directory (for systems where libraries must be found in the PATH setting at runtime) -dlopen FILE \`-dlpreopen' FILE if it cannot be dlopened at runtime -dlpreopen FILE link in FILE and add its symbols to lt_preloaded_symbols -export-dynamic allow symbols from OUTPUT-FILE to be resolved with dlsym(3) -export-symbols SYMFILE try to export only the symbols listed in SYMFILE -export-symbols-regex REGEX try to export only the symbols matching REGEX -LLIBDIR search LIBDIR for required installed libraries -lNAME OUTPUT-FILE requires the installed library libNAME -module build a library that can dlopened -no-fast-install disable the fast-install mode -no-install link a not-installable executable -no-undefined declare that a library does not refer to external symbols -o OUTPUT-FILE create OUTPUT-FILE from the specified objects -objectlist FILE Use a list of object files found in FILE to specify objects -precious-files-regex REGEX don't remove output files matching REGEX -release RELEASE specify package release information -rpath LIBDIR the created library will eventually be installed in LIBDIR -R[ ]LIBDIR add LIBDIR to the runtime path of programs and libraries -shared only do dynamic linking of libtool libraries -shrext SUFFIX override the standard shared library file extension -static do not do any dynamic linking of uninstalled libtool libraries -static-libtool-libs do not do any dynamic linking of libtool libraries -version-info CURRENT[:REVISION[:AGE]] specify library version info [each variable defaults to 0] -weak LIBNAME declare that the target provides the LIBNAME interface -Wc,FLAG -Xcompiler FLAG pass linker-specific FLAG directly to the compiler -Wl,FLAG -Xlinker FLAG pass linker-specific FLAG directly to the linker -XCClinker FLAG pass link-specific FLAG to the compiler driver (CC) All other options (arguments beginning with \`-') are ignored. Every other argument is treated as a filename. Files ending in \`.la' are treated as uninstalled libtool libraries, other files are standard or library object files. If the OUTPUT-FILE ends in \`.la', then a libtool library is created, only library objects (\`.lo' files) may be specified, and \`-rpath' is required, except when creating a convenience library. If OUTPUT-FILE ends in \`.a' or \`.lib', then a standard library is created using \`ar' and \`ranlib', or on Windows using \`lib'. If OUTPUT-FILE ends in \`.lo' or \`.${objext}', then a reloadable object file is created, otherwise an executable program is created." ;; uninstall) $ECHO \ "Usage: $progname [OPTION]... --mode=uninstall RM [RM-OPTION]... FILE... Remove libraries from an installation directory. RM is the name of the program to use to delete files associated with each FILE (typically \`/bin/rm'). RM-OPTIONS are options (such as \`-f') to be passed to RM. If FILE is a libtool library, all the files associated with it are deleted. Otherwise, only FILE itself is deleted using RM." ;; *) func_fatal_help "invalid operation mode \`$opt_mode'" ;; esac echo $ECHO "Try \`$progname --help' for more information about other modes." } # Now that we've collected a possible --mode arg, show help if necessary if $opt_help; then if test "$opt_help" = :; then func_mode_help else { func_help noexit for opt_mode in compile link execute install finish uninstall clean; do func_mode_help done } | sed -n '1p; 2,$s/^Usage:/ or: /p' { func_help noexit for opt_mode in compile link execute install finish uninstall clean; do echo func_mode_help done } | sed '1d /^When reporting/,/^Report/{ H d } $x /information about other modes/d /more detailed .*MODE/d s/^Usage:.*--mode=\([^ ]*\) .*/Description of \1 mode:/' fi exit $? fi # func_mode_execute arg... func_mode_execute () { $opt_debug # The first argument is the command name. cmd="$nonopt" test -z "$cmd" && \ func_fatal_help "you must specify a COMMAND" # Handle -dlopen flags immediately. for file in $opt_dlopen; do test -f "$file" \ || func_fatal_help "\`$file' is not a file" dir= case $file in *.la) func_resolve_sysroot "$file" file=$func_resolve_sysroot_result # Check to see that this really is a libtool archive. func_lalib_unsafe_p "$file" \ || func_fatal_help "\`$lib' is not a valid libtool archive" # Read the libtool library. dlname= library_names= func_source "$file" # Skip this library if it cannot be dlopened. if test -z "$dlname"; then # Warn if it was a shared library. test -n "$library_names" && \ func_warning "\`$file' was not linked with \`-export-dynamic'" continue fi func_dirname "$file" "" "." dir="$func_dirname_result" if test -f "$dir/$objdir/$dlname"; then func_append dir "/$objdir" else if test ! -f "$dir/$dlname"; then func_fatal_error "cannot find \`$dlname' in \`$dir' or \`$dir/$objdir'" fi fi ;; *.lo) # Just add the directory containing the .lo file. func_dirname "$file" "" "." dir="$func_dirname_result" ;; *) func_warning "\`-dlopen' is ignored for non-libtool libraries and objects" continue ;; esac # Get the absolute pathname. absdir=`cd "$dir" && pwd` test -n "$absdir" && dir="$absdir" # Now add the directory to shlibpath_var. if eval "test -z \"\$$shlibpath_var\""; then eval "$shlibpath_var=\"\$dir\"" else eval "$shlibpath_var=\"\$dir:\$$shlibpath_var\"" fi done # This variable tells wrapper scripts just to set shlibpath_var # rather than running their programs. libtool_execute_magic="$magic" # Check if any of the arguments is a wrapper script. args= for file do case $file in -* | *.la | *.lo ) ;; *) # Do a test to see if this is really a libtool program. if func_ltwrapper_script_p "$file"; then func_source "$file" # Transform arg to wrapped name. file="$progdir/$program" elif func_ltwrapper_executable_p "$file"; then func_ltwrapper_scriptname "$file" func_source "$func_ltwrapper_scriptname_result" # Transform arg to wrapped name. file="$progdir/$program" fi ;; esac # Quote arguments (to preserve shell metacharacters). func_append_quoted args "$file" done if test "X$opt_dry_run" = Xfalse; then if test -n "$shlibpath_var"; then # Export the shlibpath_var. eval "export $shlibpath_var" fi # Restore saved environment variables for lt_var in LANG LANGUAGE LC_ALL LC_CTYPE LC_COLLATE LC_MESSAGES do eval "if test \"\${save_$lt_var+set}\" = set; then $lt_var=\$save_$lt_var; export $lt_var else $lt_unset $lt_var fi" done # Now prepare to actually exec the command. exec_cmd="\$cmd$args" else # Display what would be done. if test -n "$shlibpath_var"; then eval "\$ECHO \"\$shlibpath_var=\$$shlibpath_var\"" echo "export $shlibpath_var" fi $ECHO "$cmd$args" exit $EXIT_SUCCESS fi } test "$opt_mode" = execute && func_mode_execute ${1+"$@"} # func_mode_finish arg... func_mode_finish () { $opt_debug libs= libdirs= admincmds= for opt in "$nonopt" ${1+"$@"} do if test -d "$opt"; then func_append libdirs " $opt" elif test -f "$opt"; then if func_lalib_unsafe_p "$opt"; then func_append libs " $opt" else func_warning "\`$opt' is not a valid libtool archive" fi else func_fatal_error "invalid argument \`$opt'" fi done if test -n "$libs"; then if test -n "$lt_sysroot"; then sysroot_regex=`$ECHO "$lt_sysroot" | $SED "$sed_make_literal_regex"` sysroot_cmd="s/\([ ']\)$sysroot_regex/\1/g;" else sysroot_cmd= fi # Remove sysroot references if $opt_dry_run; then for lib in $libs; do echo "removing references to $lt_sysroot and \`=' prefixes from $lib" done else tmpdir=`func_mktempdir` for lib in $libs; do sed -e "${sysroot_cmd} s/\([ ']-[LR]\)=/\1/g; s/\([ ']\)=/\1/g" $lib \ > $tmpdir/tmp-la mv -f $tmpdir/tmp-la $lib done ${RM}r "$tmpdir" fi fi if test -n "$finish_cmds$finish_eval" && test -n "$libdirs"; then for libdir in $libdirs; do if test -n "$finish_cmds"; then # Do each command in the finish commands. func_execute_cmds "$finish_cmds" 'admincmds="$admincmds '"$cmd"'"' fi if test -n "$finish_eval"; then # Do the single finish_eval. eval cmds=\"$finish_eval\" $opt_dry_run || eval "$cmds" || func_append admincmds " $cmds" fi done fi # Exit here if they wanted silent mode. $opt_silent && exit $EXIT_SUCCESS if test -n "$finish_cmds$finish_eval" && test -n "$libdirs"; then echo "----------------------------------------------------------------------" echo "Libraries have been installed in:" for libdir in $libdirs; do $ECHO " $libdir" done echo echo "If you ever happen to want to link against installed libraries" echo "in a given directory, LIBDIR, you must either use libtool, and" echo "specify the full pathname of the library, or use the \`-LLIBDIR'" echo "flag during linking and do at least one of the following:" if test -n "$shlibpath_var"; then echo " - add LIBDIR to the \`$shlibpath_var' environment variable" echo " during execution" fi if test -n "$runpath_var"; then echo " - add LIBDIR to the \`$runpath_var' environment variable" echo " during linking" fi if test -n "$hardcode_libdir_flag_spec"; then libdir=LIBDIR eval flag=\"$hardcode_libdir_flag_spec\" $ECHO " - use the \`$flag' linker flag" fi if test -n "$admincmds"; then $ECHO " - have your system administrator run these commands:$admincmds" fi if test -f /etc/ld.so.conf; then echo " - have your system administrator add LIBDIR to \`/etc/ld.so.conf'" fi echo echo "See any operating system documentation about shared libraries for" case $host in solaris2.[6789]|solaris2.1[0-9]) echo "more information, such as the ld(1), crle(1) and ld.so(8) manual" echo "pages." ;; *) echo "more information, such as the ld(1) and ld.so(8) manual pages." ;; esac echo "----------------------------------------------------------------------" fi exit $EXIT_SUCCESS } test "$opt_mode" = finish && func_mode_finish ${1+"$@"} # func_mode_install arg... func_mode_install () { $opt_debug # There may be an optional sh(1) argument at the beginning of # install_prog (especially on Windows NT). if test "$nonopt" = "$SHELL" || test "$nonopt" = /bin/sh || # Allow the use of GNU shtool's install command. case $nonopt in *shtool*) :;; *) false;; esac; then # Aesthetically quote it. func_quote_for_eval "$nonopt" install_prog="$func_quote_for_eval_result " arg=$1 shift else install_prog= arg=$nonopt fi # The real first argument should be the name of the installation program. # Aesthetically quote it. func_quote_for_eval "$arg" func_append install_prog "$func_quote_for_eval_result" install_shared_prog=$install_prog case " $install_prog " in *[\\\ /]cp\ *) install_cp=: ;; *) install_cp=false ;; esac # We need to accept at least all the BSD install flags. dest= files= opts= prev= install_type= isdir=no stripme= no_mode=: for arg do arg2= if test -n "$dest"; then func_append files " $dest" dest=$arg continue fi case $arg in -d) isdir=yes ;; -f) if $install_cp; then :; else prev=$arg fi ;; -g | -m | -o) prev=$arg ;; -s) stripme=" -s" continue ;; -*) ;; *) # If the previous option needed an argument, then skip it. if test -n "$prev"; then if test "x$prev" = x-m && test -n "$install_override_mode"; then arg2=$install_override_mode no_mode=false fi prev= else dest=$arg continue fi ;; esac # Aesthetically quote the argument. func_quote_for_eval "$arg" func_append install_prog " $func_quote_for_eval_result" if test -n "$arg2"; then func_quote_for_eval "$arg2" fi func_append install_shared_prog " $func_quote_for_eval_result" done test -z "$install_prog" && \ func_fatal_help "you must specify an install program" test -n "$prev" && \ func_fatal_help "the \`$prev' option requires an argument" if test -n "$install_override_mode" && $no_mode; then if $install_cp; then :; else func_quote_for_eval "$install_override_mode" func_append install_shared_prog " -m $func_quote_for_eval_result" fi fi if test -z "$files"; then if test -z "$dest"; then func_fatal_help "no file or destination specified" else func_fatal_help "you must specify a destination" fi fi # Strip any trailing slash from the destination. func_stripname '' '/' "$dest" dest=$func_stripname_result # Check to see that the destination is a directory. test -d "$dest" && isdir=yes if test "$isdir" = yes; then destdir="$dest" destname= else func_dirname_and_basename "$dest" "" "." destdir="$func_dirname_result" destname="$func_basename_result" # Not a directory, so check to see that there is only one file specified. set dummy $files; shift test "$#" -gt 1 && \ func_fatal_help "\`$dest' is not a directory" fi case $destdir in [\\/]* | [A-Za-z]:[\\/]*) ;; *) for file in $files; do case $file in *.lo) ;; *) func_fatal_help "\`$destdir' must be an absolute directory name" ;; esac done ;; esac # This variable tells wrapper scripts just to set variables rather # than running their programs. libtool_install_magic="$magic" staticlibs= future_libdirs= current_libdirs= for file in $files; do # Do each installation. case $file in *.$libext) # Do the static libraries later. func_append staticlibs " $file" ;; *.la) func_resolve_sysroot "$file" file=$func_resolve_sysroot_result # Check to see that this really is a libtool archive. func_lalib_unsafe_p "$file" \ || func_fatal_help "\`$file' is not a valid libtool archive" library_names= old_library= relink_command= func_source "$file" # Add the libdir to current_libdirs if it is the destination. if test "X$destdir" = "X$libdir"; then case "$current_libdirs " in *" $libdir "*) ;; *) func_append current_libdirs " $libdir" ;; esac else # Note the libdir as a future libdir. case "$future_libdirs " in *" $libdir "*) ;; *) func_append future_libdirs " $libdir" ;; esac fi func_dirname "$file" "/" "" dir="$func_dirname_result" func_append dir "$objdir" if test -n "$relink_command"; then # Determine the prefix the user has applied to our future dir. inst_prefix_dir=`$ECHO "$destdir" | $SED -e "s%$libdir\$%%"` # Don't allow the user to place us outside of our expected # location b/c this prevents finding dependent libraries that # are installed to the same prefix. # At present, this check doesn't affect windows .dll's that # are installed into $libdir/../bin (currently, that works fine) # but it's something to keep an eye on. test "$inst_prefix_dir" = "$destdir" && \ func_fatal_error "error: cannot install \`$file' to a directory not ending in $libdir" if test -n "$inst_prefix_dir"; then # Stick the inst_prefix_dir data into the link command. relink_command=`$ECHO "$relink_command" | $SED "s%@inst_prefix_dir@%-inst-prefix-dir $inst_prefix_dir%"` else relink_command=`$ECHO "$relink_command" | $SED "s%@inst_prefix_dir@%%"` fi func_warning "relinking \`$file'" func_show_eval "$relink_command" \ 'func_fatal_error "error: relink \`$file'\'' with the above command before installing it"' fi # See the names of the shared library. set dummy $library_names; shift if test -n "$1"; then realname="$1" shift srcname="$realname" test -n "$relink_command" && srcname="$realname"T # Install the shared library and build the symlinks. func_show_eval "$install_shared_prog $dir/$srcname $destdir/$realname" \ 'exit $?' tstripme="$stripme" case $host_os in cygwin* | mingw* | pw32* | cegcc*) case $realname in *.dll.a) tstripme="" ;; esac ;; esac if test -n "$tstripme" && test -n "$striplib"; then func_show_eval "$striplib $destdir/$realname" 'exit $?' fi if test "$#" -gt 0; then # Delete the old symlinks, and create new ones. # Try `ln -sf' first, because the `ln' binary might depend on # the symlink we replace! Solaris /bin/ln does not understand -f, # so we also need to try rm && ln -s. for linkname do test "$linkname" != "$realname" \ && func_show_eval "(cd $destdir && { $LN_S -f $realname $linkname || { $RM $linkname && $LN_S $realname $linkname; }; })" done fi # Do each command in the postinstall commands. lib="$destdir/$realname" func_execute_cmds "$postinstall_cmds" 'exit $?' fi # Install the pseudo-library for information purposes. func_basename "$file" name="$func_basename_result" instname="$dir/$name"i func_show_eval "$install_prog $instname $destdir/$name" 'exit $?' # Maybe install the static library, too. test -n "$old_library" && func_append staticlibs " $dir/$old_library" ;; *.lo) # Install (i.e. copy) a libtool object. # Figure out destination file name, if it wasn't already specified. if test -n "$destname"; then destfile="$destdir/$destname" else func_basename "$file" destfile="$func_basename_result" destfile="$destdir/$destfile" fi # Deduce the name of the destination old-style object file. case $destfile in *.lo) func_lo2o "$destfile" staticdest=$func_lo2o_result ;; *.$objext) staticdest="$destfile" destfile= ;; *) func_fatal_help "cannot copy a libtool object to \`$destfile'" ;; esac # Install the libtool object if requested. test -n "$destfile" && \ func_show_eval "$install_prog $file $destfile" 'exit $?' # Install the old object if enabled. if test "$build_old_libs" = yes; then # Deduce the name of the old-style object file. func_lo2o "$file" staticobj=$func_lo2o_result func_show_eval "$install_prog \$staticobj \$staticdest" 'exit $?' fi exit $EXIT_SUCCESS ;; *) # Figure out destination file name, if it wasn't already specified. if test -n "$destname"; then destfile="$destdir/$destname" else func_basename "$file" destfile="$func_basename_result" destfile="$destdir/$destfile" fi # If the file is missing, and there is a .exe on the end, strip it # because it is most likely a libtool script we actually want to # install stripped_ext="" case $file in *.exe) if test ! -f "$file"; then func_stripname '' '.exe' "$file" file=$func_stripname_result stripped_ext=".exe" fi ;; esac # Do a test to see if this is really a libtool program. case $host in *cygwin* | *mingw*) if func_ltwrapper_executable_p "$file"; then func_ltwrapper_scriptname "$file" wrapper=$func_ltwrapper_scriptname_result else func_stripname '' '.exe' "$file" wrapper=$func_stripname_result fi ;; *) wrapper=$file ;; esac if func_ltwrapper_script_p "$wrapper"; then notinst_deplibs= relink_command= func_source "$wrapper" # Check the variables that should have been set. test -z "$generated_by_libtool_version" && \ func_fatal_error "invalid libtool wrapper script \`$wrapper'" finalize=yes for lib in $notinst_deplibs; do # Check to see that each library is installed. libdir= if test -f "$lib"; then func_source "$lib" fi libfile="$libdir/"`$ECHO "$lib" | $SED 's%^.*/%%g'` ### testsuite: skip nested quoting test if test -n "$libdir" && test ! -f "$libfile"; then func_warning "\`$lib' has not been installed in \`$libdir'" finalize=no fi done relink_command= func_source "$wrapper" outputname= if test "$fast_install" = no && test -n "$relink_command"; then $opt_dry_run || { if test "$finalize" = yes; then tmpdir=`func_mktempdir` func_basename "$file$stripped_ext" file="$func_basename_result" outputname="$tmpdir/$file" # Replace the output file specification. relink_command=`$ECHO "$relink_command" | $SED 's%@OUTPUT@%'"$outputname"'%g'` $opt_silent || { func_quote_for_expand "$relink_command" eval "func_echo $func_quote_for_expand_result" } if eval "$relink_command"; then : else func_error "error: relink \`$file' with the above command before installing it" $opt_dry_run || ${RM}r "$tmpdir" continue fi file="$outputname" else func_warning "cannot relink \`$file'" fi } else # Install the binary that we compiled earlier. file=`$ECHO "$file$stripped_ext" | $SED "s%\([^/]*\)$%$objdir/\1%"` fi fi # remove .exe since cygwin /usr/bin/install will append another # one anyway case $install_prog,$host in */usr/bin/install*,*cygwin*) case $file:$destfile in *.exe:*.exe) # this is ok ;; *.exe:*) destfile=$destfile.exe ;; *:*.exe) func_stripname '' '.exe' "$destfile" destfile=$func_stripname_result ;; esac ;; esac func_show_eval "$install_prog\$stripme \$file \$destfile" 'exit $?' $opt_dry_run || if test -n "$outputname"; then ${RM}r "$tmpdir" fi ;; esac done for file in $staticlibs; do func_basename "$file" name="$func_basename_result" # Set up the ranlib parameters. oldlib="$destdir/$name" func_to_tool_file "$oldlib" func_convert_file_msys_to_w32 tool_oldlib=$func_to_tool_file_result func_show_eval "$install_prog \$file \$oldlib" 'exit $?' if test -n "$stripme" && test -n "$old_striplib"; then func_show_eval "$old_striplib $tool_oldlib" 'exit $?' fi # Do each command in the postinstall commands. func_execute_cmds "$old_postinstall_cmds" 'exit $?' done test -n "$future_libdirs" && \ func_warning "remember to run \`$progname --finish$future_libdirs'" if test -n "$current_libdirs"; then # Maybe just do a dry run. $opt_dry_run && current_libdirs=" -n$current_libdirs" exec_cmd='$SHELL $progpath $preserve_args --finish$current_libdirs' else exit $EXIT_SUCCESS fi } test "$opt_mode" = install && func_mode_install ${1+"$@"} # func_generate_dlsyms outputname originator pic_p # Extract symbols from dlprefiles and create ${outputname}S.o with # a dlpreopen symbol table. func_generate_dlsyms () { $opt_debug my_outputname="$1" my_originator="$2" my_pic_p="${3-no}" my_prefix=`$ECHO "$my_originator" | sed 's%[^a-zA-Z0-9]%_%g'` my_dlsyms= if test -n "$dlfiles$dlprefiles" || test "$dlself" != no; then if test -n "$NM" && test -n "$global_symbol_pipe"; then my_dlsyms="${my_outputname}S.c" else func_error "not configured to extract global symbols from dlpreopened files" fi fi if test -n "$my_dlsyms"; then case $my_dlsyms in "") ;; *.c) # Discover the nlist of each of the dlfiles. nlist="$output_objdir/${my_outputname}.nm" func_show_eval "$RM $nlist ${nlist}S ${nlist}T" # Parse the name list into a source file. func_verbose "creating $output_objdir/$my_dlsyms" $opt_dry_run || $ECHO > "$output_objdir/$my_dlsyms" "\ /* $my_dlsyms - symbol resolution table for \`$my_outputname' dlsym emulation. */ /* Generated by $PROGRAM (GNU $PACKAGE$TIMESTAMP) $VERSION */ #ifdef __cplusplus extern \"C\" { #endif #if defined(__GNUC__) && (((__GNUC__ == 4) && (__GNUC_MINOR__ >= 4)) || (__GNUC__ > 4)) #pragma GCC diagnostic ignored \"-Wstrict-prototypes\" #endif /* Keep this code in sync between libtool.m4, ltmain, lt_system.h, and tests. */ #if defined(_WIN32) || defined(__CYGWIN__) || defined(_WIN32_WCE) /* DATA imports from DLLs on WIN32 con't be const, because runtime relocations are performed -- see ld's documentation on pseudo-relocs. */ # define LT_DLSYM_CONST #elif defined(__osf__) /* This system does not cope well with relocations in const data. */ # define LT_DLSYM_CONST #else # define LT_DLSYM_CONST const #endif /* External symbol declarations for the compiler. */\ " if test "$dlself" = yes; then func_verbose "generating symbol list for \`$output'" $opt_dry_run || echo ': @PROGRAM@ ' > "$nlist" # Add our own program objects to the symbol list. progfiles=`$ECHO "$objs$old_deplibs" | $SP2NL | $SED "$lo2o" | $NL2SP` for progfile in $progfiles; do func_to_tool_file "$progfile" func_convert_file_msys_to_w32 func_verbose "extracting global C symbols from \`$func_to_tool_file_result'" $opt_dry_run || eval "$NM $func_to_tool_file_result | $global_symbol_pipe >> '$nlist'" done if test -n "$exclude_expsyms"; then $opt_dry_run || { eval '$EGREP -v " ($exclude_expsyms)$" "$nlist" > "$nlist"T' eval '$MV "$nlist"T "$nlist"' } fi if test -n "$export_symbols_regex"; then $opt_dry_run || { eval '$EGREP -e "$export_symbols_regex" "$nlist" > "$nlist"T' eval '$MV "$nlist"T "$nlist"' } fi # Prepare the list of exported symbols if test -z "$export_symbols"; then export_symbols="$output_objdir/$outputname.exp" $opt_dry_run || { $RM $export_symbols eval "${SED} -n -e '/^: @PROGRAM@ $/d' -e 's/^.* \(.*\)$/\1/p' "'< "$nlist" > "$export_symbols"' case $host in *cygwin* | *mingw* | *cegcc* ) eval "echo EXPORTS "'> "$output_objdir/$outputname.def"' eval 'cat "$export_symbols" >> "$output_objdir/$outputname.def"' ;; esac } else $opt_dry_run || { eval "${SED} -e 's/\([].[*^$]\)/\\\\\1/g' -e 's/^/ /' -e 's/$/$/'"' < "$export_symbols" > "$output_objdir/$outputname.exp"' eval '$GREP -f "$output_objdir/$outputname.exp" < "$nlist" > "$nlist"T' eval '$MV "$nlist"T "$nlist"' case $host in *cygwin* | *mingw* | *cegcc* ) eval "echo EXPORTS "'> "$output_objdir/$outputname.def"' eval 'cat "$nlist" >> "$output_objdir/$outputname.def"' ;; esac } fi fi for dlprefile in $dlprefiles; do func_verbose "extracting global C symbols from \`$dlprefile'" func_basename "$dlprefile" name="$func_basename_result" case $host in *cygwin* | *mingw* | *cegcc* ) # if an import library, we need to obtain dlname if func_win32_import_lib_p "$dlprefile"; then func_tr_sh "$dlprefile" eval "curr_lafile=\$libfile_$func_tr_sh_result" dlprefile_dlbasename="" if test -n "$curr_lafile" && func_lalib_p "$curr_lafile"; then # Use subshell, to avoid clobbering current variable values dlprefile_dlname=`source "$curr_lafile" && echo "$dlname"` if test -n "$dlprefile_dlname" ; then func_basename "$dlprefile_dlname" dlprefile_dlbasename="$func_basename_result" else # no lafile. user explicitly requested -dlpreopen . $sharedlib_from_linklib_cmd "$dlprefile" dlprefile_dlbasename=$sharedlib_from_linklib_result fi fi $opt_dry_run || { if test -n "$dlprefile_dlbasename" ; then eval '$ECHO ": $dlprefile_dlbasename" >> "$nlist"' else func_warning "Could not compute DLL name from $name" eval '$ECHO ": $name " >> "$nlist"' fi func_to_tool_file "$dlprefile" func_convert_file_msys_to_w32 eval "$NM \"$func_to_tool_file_result\" 2>/dev/null | $global_symbol_pipe | $SED -e '/I __imp/d' -e 's/I __nm_/D /;s/_nm__//' >> '$nlist'" } else # not an import lib $opt_dry_run || { eval '$ECHO ": $name " >> "$nlist"' func_to_tool_file "$dlprefile" func_convert_file_msys_to_w32 eval "$NM \"$func_to_tool_file_result\" 2>/dev/null | $global_symbol_pipe >> '$nlist'" } fi ;; *) $opt_dry_run || { eval '$ECHO ": $name " >> "$nlist"' func_to_tool_file "$dlprefile" func_convert_file_msys_to_w32 eval "$NM \"$func_to_tool_file_result\" 2>/dev/null | $global_symbol_pipe >> '$nlist'" } ;; esac done $opt_dry_run || { # Make sure we have at least an empty file. test -f "$nlist" || : > "$nlist" if test -n "$exclude_expsyms"; then $EGREP -v " ($exclude_expsyms)$" "$nlist" > "$nlist"T $MV "$nlist"T "$nlist" fi # Try sorting and uniquifying the output. if $GREP -v "^: " < "$nlist" | if sort -k 3 /dev/null 2>&1; then sort -k 3 else sort +2 fi | uniq > "$nlist"S; then : else $GREP -v "^: " < "$nlist" > "$nlist"S fi if test -f "$nlist"S; then eval "$global_symbol_to_cdecl"' < "$nlist"S >> "$output_objdir/$my_dlsyms"' else echo '/* NONE */' >> "$output_objdir/$my_dlsyms" fi echo >> "$output_objdir/$my_dlsyms" "\ /* The mapping between symbol names and symbols. */ typedef struct { const char *name; void *address; } lt_dlsymlist; extern LT_DLSYM_CONST lt_dlsymlist lt_${my_prefix}_LTX_preloaded_symbols[]; LT_DLSYM_CONST lt_dlsymlist lt_${my_prefix}_LTX_preloaded_symbols[] = {\ { \"$my_originator\", (void *) 0 }," case $need_lib_prefix in no) eval "$global_symbol_to_c_name_address" < "$nlist" >> "$output_objdir/$my_dlsyms" ;; *) eval "$global_symbol_to_c_name_address_lib_prefix" < "$nlist" >> "$output_objdir/$my_dlsyms" ;; esac echo >> "$output_objdir/$my_dlsyms" "\ {0, (void *) 0} }; /* This works around a problem in FreeBSD linker */ #ifdef FREEBSD_WORKAROUND static const void *lt_preloaded_setup() { return lt_${my_prefix}_LTX_preloaded_symbols; } #endif #ifdef __cplusplus } #endif\ " } # !$opt_dry_run pic_flag_for_symtable= case "$compile_command " in *" -static "*) ;; *) case $host in # compiling the symbol table file with pic_flag works around # a FreeBSD bug that causes programs to crash when -lm is # linked before any other PIC object. But we must not use # pic_flag when linking with -static. The problem exists in # FreeBSD 2.2.6 and is fixed in FreeBSD 3.1. *-*-freebsd2.*|*-*-freebsd3.0*|*-*-freebsdelf3.0*) pic_flag_for_symtable=" $pic_flag -DFREEBSD_WORKAROUND" ;; *-*-hpux*) pic_flag_for_symtable=" $pic_flag" ;; *) if test "X$my_pic_p" != Xno; then pic_flag_for_symtable=" $pic_flag" fi ;; esac ;; esac symtab_cflags= for arg in $LTCFLAGS; do case $arg in -pie | -fpie | -fPIE) ;; *) func_append symtab_cflags " $arg" ;; esac done # Now compile the dynamic symbol file. func_show_eval '(cd $output_objdir && $LTCC$symtab_cflags -c$no_builtin_flag$pic_flag_for_symtable "$my_dlsyms")' 'exit $?' # Clean up the generated files. func_show_eval '$RM "$output_objdir/$my_dlsyms" "$nlist" "${nlist}S" "${nlist}T"' # Transform the symbol file into the correct name. symfileobj="$output_objdir/${my_outputname}S.$objext" case $host in *cygwin* | *mingw* | *cegcc* ) if test -f "$output_objdir/$my_outputname.def"; then compile_command=`$ECHO "$compile_command" | $SED "s%@SYMFILE@%$output_objdir/$my_outputname.def $symfileobj%"` finalize_command=`$ECHO "$finalize_command" | $SED "s%@SYMFILE@%$output_objdir/$my_outputname.def $symfileobj%"` else compile_command=`$ECHO "$compile_command" | $SED "s%@SYMFILE@%$symfileobj%"` finalize_command=`$ECHO "$finalize_command" | $SED "s%@SYMFILE@%$symfileobj%"` fi ;; *) compile_command=`$ECHO "$compile_command" | $SED "s%@SYMFILE@%$symfileobj%"` finalize_command=`$ECHO "$finalize_command" | $SED "s%@SYMFILE@%$symfileobj%"` ;; esac ;; *) func_fatal_error "unknown suffix for \`$my_dlsyms'" ;; esac else # We keep going just in case the user didn't refer to # lt_preloaded_symbols. The linker will fail if global_symbol_pipe # really was required. # Nullify the symbol file. compile_command=`$ECHO "$compile_command" | $SED "s% @SYMFILE@%%"` finalize_command=`$ECHO "$finalize_command" | $SED "s% @SYMFILE@%%"` fi } # func_win32_libid arg # return the library type of file 'arg' # # Need a lot of goo to handle *both* DLLs and import libs # Has to be a shell function in order to 'eat' the argument # that is supplied when $file_magic_command is called. # Despite the name, also deal with 64 bit binaries. func_win32_libid () { $opt_debug win32_libid_type="unknown" win32_fileres=`file -L $1 2>/dev/null` case $win32_fileres in *ar\ archive\ import\ library*) # definitely import win32_libid_type="x86 archive import" ;; *ar\ archive*) # could be an import, or static # Keep the egrep pattern in sync with the one in _LT_CHECK_MAGIC_METHOD. if eval $OBJDUMP -f $1 | $SED -e '10q' 2>/dev/null | $EGREP 'file format (pei*-i386(.*architecture: i386)?|pe-arm-wince|pe-x86-64)' >/dev/null; then func_to_tool_file "$1" func_convert_file_msys_to_w32 win32_nmres=`eval $NM -f posix -A \"$func_to_tool_file_result\" | $SED -n -e ' 1,100{ / I /{ s,.*,import, p q } }'` case $win32_nmres in import*) win32_libid_type="x86 archive import";; *) win32_libid_type="x86 archive static";; esac fi ;; *DLL*) win32_libid_type="x86 DLL" ;; *executable*) # but shell scripts are "executable" too... case $win32_fileres in *MS\ Windows\ PE\ Intel*) win32_libid_type="x86 DLL" ;; esac ;; esac $ECHO "$win32_libid_type" } # func_cygming_dll_for_implib ARG # # Platform-specific function to extract the # name of the DLL associated with the specified # import library ARG. # Invoked by eval'ing the libtool variable # $sharedlib_from_linklib_cmd # Result is available in the variable # $sharedlib_from_linklib_result func_cygming_dll_for_implib () { $opt_debug sharedlib_from_linklib_result=`$DLLTOOL --identify-strict --identify "$1"` } # func_cygming_dll_for_implib_fallback_core SECTION_NAME LIBNAMEs # # The is the core of a fallback implementation of a # platform-specific function to extract the name of the # DLL associated with the specified import library LIBNAME. # # SECTION_NAME is either .idata$6 or .idata$7, depending # on the platform and compiler that created the implib. # # Echos the name of the DLL associated with the # specified import library. func_cygming_dll_for_implib_fallback_core () { $opt_debug match_literal=`$ECHO "$1" | $SED "$sed_make_literal_regex"` $OBJDUMP -s --section "$1" "$2" 2>/dev/null | $SED '/^Contents of section '"$match_literal"':/{ # Place marker at beginning of archive member dllname section s/.*/====MARK====/ p d } # These lines can sometimes be longer than 43 characters, but # are always uninteresting /:[ ]*file format pe[i]\{,1\}-/d /^In archive [^:]*:/d # Ensure marker is printed /^====MARK====/p # Remove all lines with less than 43 characters /^.\{43\}/!d # From remaining lines, remove first 43 characters s/^.\{43\}//' | $SED -n ' # Join marker and all lines until next marker into a single line /^====MARK====/ b para H $ b para b :para x s/\n//g # Remove the marker s/^====MARK====// # Remove trailing dots and whitespace s/[\. \t]*$// # Print /./p' | # we now have a list, one entry per line, of the stringified # contents of the appropriate section of all members of the # archive which possess that section. Heuristic: eliminate # all those which have a first or second character that is # a '.' (that is, objdump's representation of an unprintable # character.) This should work for all archives with less than # 0x302f exports -- but will fail for DLLs whose name actually # begins with a literal '.' or a single character followed by # a '.'. # # Of those that remain, print the first one. $SED -e '/^\./d;/^.\./d;q' } # func_cygming_gnu_implib_p ARG # This predicate returns with zero status (TRUE) if # ARG is a GNU/binutils-style import library. Returns # with nonzero status (FALSE) otherwise. func_cygming_gnu_implib_p () { $opt_debug func_to_tool_file "$1" func_convert_file_msys_to_w32 func_cygming_gnu_implib_tmp=`$NM "$func_to_tool_file_result" | eval "$global_symbol_pipe" | $EGREP ' (_head_[A-Za-z0-9_]+_[ad]l*|[A-Za-z0-9_]+_[ad]l*_iname)$'` test -n "$func_cygming_gnu_implib_tmp" } # func_cygming_ms_implib_p ARG # This predicate returns with zero status (TRUE) if # ARG is an MS-style import library. Returns # with nonzero status (FALSE) otherwise. func_cygming_ms_implib_p () { $opt_debug func_to_tool_file "$1" func_convert_file_msys_to_w32 func_cygming_ms_implib_tmp=`$NM "$func_to_tool_file_result" | eval "$global_symbol_pipe" | $GREP '_NULL_IMPORT_DESCRIPTOR'` test -n "$func_cygming_ms_implib_tmp" } # func_cygming_dll_for_implib_fallback ARG # Platform-specific function to extract the # name of the DLL associated with the specified # import library ARG. # # This fallback implementation is for use when $DLLTOOL # does not support the --identify-strict option. # Invoked by eval'ing the libtool variable # $sharedlib_from_linklib_cmd # Result is available in the variable # $sharedlib_from_linklib_result func_cygming_dll_for_implib_fallback () { $opt_debug if func_cygming_gnu_implib_p "$1" ; then # binutils import library sharedlib_from_linklib_result=`func_cygming_dll_for_implib_fallback_core '.idata$7' "$1"` elif func_cygming_ms_implib_p "$1" ; then # ms-generated import library sharedlib_from_linklib_result=`func_cygming_dll_for_implib_fallback_core '.idata$6' "$1"` else # unknown sharedlib_from_linklib_result="" fi } # func_extract_an_archive dir oldlib func_extract_an_archive () { $opt_debug f_ex_an_ar_dir="$1"; shift f_ex_an_ar_oldlib="$1" if test "$lock_old_archive_extraction" = yes; then lockfile=$f_ex_an_ar_oldlib.lock until $opt_dry_run || ln "$progpath" "$lockfile" 2>/dev/null; do func_echo "Waiting for $lockfile to be removed" sleep 2 done fi func_show_eval "(cd \$f_ex_an_ar_dir && $AR x \"\$f_ex_an_ar_oldlib\")" \ 'stat=$?; rm -f "$lockfile"; exit $stat' if test "$lock_old_archive_extraction" = yes; then $opt_dry_run || rm -f "$lockfile" fi if ($AR t "$f_ex_an_ar_oldlib" | sort | sort -uc >/dev/null 2>&1); then : else func_fatal_error "object name conflicts in archive: $f_ex_an_ar_dir/$f_ex_an_ar_oldlib" fi } # func_extract_archives gentop oldlib ... func_extract_archives () { $opt_debug my_gentop="$1"; shift my_oldlibs=${1+"$@"} my_oldobjs="" my_xlib="" my_xabs="" my_xdir="" for my_xlib in $my_oldlibs; do # Extract the objects. case $my_xlib in [\\/]* | [A-Za-z]:[\\/]*) my_xabs="$my_xlib" ;; *) my_xabs=`pwd`"/$my_xlib" ;; esac func_basename "$my_xlib" my_xlib="$func_basename_result" my_xlib_u=$my_xlib while :; do case " $extracted_archives " in *" $my_xlib_u "*) func_arith $extracted_serial + 1 extracted_serial=$func_arith_result my_xlib_u=lt$extracted_serial-$my_xlib ;; *) break ;; esac done extracted_archives="$extracted_archives $my_xlib_u" my_xdir="$my_gentop/$my_xlib_u" func_mkdir_p "$my_xdir" case $host in *-darwin*) func_verbose "Extracting $my_xabs" # Do not bother doing anything if just a dry run $opt_dry_run || { darwin_orig_dir=`pwd` cd $my_xdir || exit $? darwin_archive=$my_xabs darwin_curdir=`pwd` darwin_base_archive=`basename "$darwin_archive"` darwin_arches=`$LIPO -info "$darwin_archive" 2>/dev/null | $GREP Architectures 2>/dev/null || true` if test -n "$darwin_arches"; then darwin_arches=`$ECHO "$darwin_arches" | $SED -e 's/.*are://'` darwin_arch= func_verbose "$darwin_base_archive has multiple architectures $darwin_arches" for darwin_arch in $darwin_arches ; do func_mkdir_p "unfat-$$/${darwin_base_archive}-${darwin_arch}" $LIPO -thin $darwin_arch -output "unfat-$$/${darwin_base_archive}-${darwin_arch}/${darwin_base_archive}" "${darwin_archive}" cd "unfat-$$/${darwin_base_archive}-${darwin_arch}" func_extract_an_archive "`pwd`" "${darwin_base_archive}" cd "$darwin_curdir" $RM "unfat-$$/${darwin_base_archive}-${darwin_arch}/${darwin_base_archive}" done # $darwin_arches ## Okay now we've a bunch of thin objects, gotta fatten them up :) darwin_filelist=`find unfat-$$ -type f -name \*.o -print -o -name \*.lo -print | $SED -e "$basename" | sort -u` darwin_file= darwin_files= for darwin_file in $darwin_filelist; do darwin_files=`find unfat-$$ -name $darwin_file -print | sort | $NL2SP` $LIPO -create -output "$darwin_file" $darwin_files done # $darwin_filelist $RM -rf unfat-$$ cd "$darwin_orig_dir" else cd $darwin_orig_dir func_extract_an_archive "$my_xdir" "$my_xabs" fi # $darwin_arches } # !$opt_dry_run ;; *) func_extract_an_archive "$my_xdir" "$my_xabs" ;; esac my_oldobjs="$my_oldobjs "`find $my_xdir -name \*.$objext -print -o -name \*.lo -print | sort | $NL2SP` done func_extract_archives_result="$my_oldobjs" } # func_emit_wrapper [arg=no] # # Emit a libtool wrapper script on stdout. # Don't directly open a file because we may want to # incorporate the script contents within a cygwin/mingw # wrapper executable. Must ONLY be called from within # func_mode_link because it depends on a number of variables # set therein. # # ARG is the value that the WRAPPER_SCRIPT_BELONGS_IN_OBJDIR # variable will take. If 'yes', then the emitted script # will assume that the directory in which it is stored is # the $objdir directory. This is a cygwin/mingw-specific # behavior. func_emit_wrapper () { func_emit_wrapper_arg1=${1-no} $ECHO "\ #! $SHELL # $output - temporary wrapper script for $objdir/$outputname # Generated by $PROGRAM (GNU $PACKAGE$TIMESTAMP) $VERSION # # The $output program cannot be directly executed until all the libtool # libraries that it depends on are installed. # # This wrapper script should never be moved out of the build directory. # If it is, it will not operate correctly. # Sed substitution that helps us do robust quoting. It backslashifies # metacharacters that are still active within double-quoted strings. sed_quote_subst='$sed_quote_subst' # Be Bourne compatible if test -n \"\${ZSH_VERSION+set}\" && (emulate sh) >/dev/null 2>&1; then emulate sh NULLCMD=: # Zsh 3.x and 4.x performs word splitting on \${1+\"\$@\"}, which # is contrary to our usage. Disable this feature. alias -g '\${1+\"\$@\"}'='\"\$@\"' setopt NO_GLOB_SUBST else case \`(set -o) 2>/dev/null\` in *posix*) set -o posix;; esac fi BIN_SH=xpg4; export BIN_SH # for Tru64 DUALCASE=1; export DUALCASE # for MKS sh # The HP-UX ksh and POSIX shell print the target directory to stdout # if CDPATH is set. (unset CDPATH) >/dev/null 2>&1 && unset CDPATH relink_command=\"$relink_command\" # This environment variable determines our operation mode. if test \"\$libtool_install_magic\" = \"$magic\"; then # install mode needs the following variables: generated_by_libtool_version='$macro_version' notinst_deplibs='$notinst_deplibs' else # When we are sourced in execute mode, \$file and \$ECHO are already set. if test \"\$libtool_execute_magic\" != \"$magic\"; then file=\"\$0\"" qECHO=`$ECHO "$ECHO" | $SED "$sed_quote_subst"` $ECHO "\ # A function that is used when there is no print builtin or printf. func_fallback_echo () { eval 'cat <<_LTECHO_EOF \$1 _LTECHO_EOF' } ECHO=\"$qECHO\" fi # Very basic option parsing. These options are (a) specific to # the libtool wrapper, (b) are identical between the wrapper # /script/ and the wrapper /executable/ which is used only on # windows platforms, and (c) all begin with the string "--lt-" # (application programs are unlikely to have options which match # this pattern). # # There are only two supported options: --lt-debug and # --lt-dump-script. There is, deliberately, no --lt-help. # # The first argument to this parsing function should be the # script's $0 value, followed by "$@". lt_option_debug= func_parse_lt_options () { lt_script_arg0=\$0 shift for lt_opt do case \"\$lt_opt\" in --lt-debug) lt_option_debug=1 ;; --lt-dump-script) lt_dump_D=\`\$ECHO \"X\$lt_script_arg0\" | $SED -e 's/^X//' -e 's%/[^/]*$%%'\` test \"X\$lt_dump_D\" = \"X\$lt_script_arg0\" && lt_dump_D=. lt_dump_F=\`\$ECHO \"X\$lt_script_arg0\" | $SED -e 's/^X//' -e 's%^.*/%%'\` cat \"\$lt_dump_D/\$lt_dump_F\" exit 0 ;; --lt-*) \$ECHO \"Unrecognized --lt- option: '\$lt_opt'\" 1>&2 exit 1 ;; esac done # Print the debug banner immediately: if test -n \"\$lt_option_debug\"; then echo \"${outputname}:${output}:\${LINENO}: libtool wrapper (GNU $PACKAGE$TIMESTAMP) $VERSION\" 1>&2 fi } # Used when --lt-debug. Prints its arguments to stdout # (redirection is the responsibility of the caller) func_lt_dump_args () { lt_dump_args_N=1; for lt_arg do \$ECHO \"${outputname}:${output}:\${LINENO}: newargv[\$lt_dump_args_N]: \$lt_arg\" lt_dump_args_N=\`expr \$lt_dump_args_N + 1\` done } # Core function for launching the target application func_exec_program_core () { " case $host in # Backslashes separate directories on plain windows *-*-mingw | *-*-os2* | *-cegcc*) $ECHO "\ if test -n \"\$lt_option_debug\"; then \$ECHO \"${outputname}:${output}:\${LINENO}: newargv[0]: \$progdir\\\\\$program\" 1>&2 func_lt_dump_args \${1+\"\$@\"} 1>&2 fi exec \"\$progdir\\\\\$program\" \${1+\"\$@\"} " ;; *) $ECHO "\ if test -n \"\$lt_option_debug\"; then \$ECHO \"${outputname}:${output}:\${LINENO}: newargv[0]: \$progdir/\$program\" 1>&2 func_lt_dump_args \${1+\"\$@\"} 1>&2 fi exec \"\$progdir/\$program\" \${1+\"\$@\"} " ;; esac $ECHO "\ \$ECHO \"\$0: cannot exec \$program \$*\" 1>&2 exit 1 } # A function to encapsulate launching the target application # Strips options in the --lt-* namespace from \$@ and # launches target application with the remaining arguments. func_exec_program () { case \" \$* \" in *\\ --lt-*) for lt_wr_arg do case \$lt_wr_arg in --lt-*) ;; *) set x \"\$@\" \"\$lt_wr_arg\"; shift;; esac shift done ;; esac func_exec_program_core \${1+\"\$@\"} } # Parse options func_parse_lt_options \"\$0\" \${1+\"\$@\"} # Find the directory that this script lives in. thisdir=\`\$ECHO \"\$file\" | $SED 's%/[^/]*$%%'\` test \"x\$thisdir\" = \"x\$file\" && thisdir=. # Follow symbolic links until we get to the real thisdir. file=\`ls -ld \"\$file\" | $SED -n 's/.*-> //p'\` while test -n \"\$file\"; do destdir=\`\$ECHO \"\$file\" | $SED 's%/[^/]*\$%%'\` # If there was a directory component, then change thisdir. if test \"x\$destdir\" != \"x\$file\"; then case \"\$destdir\" in [\\\\/]* | [A-Za-z]:[\\\\/]*) thisdir=\"\$destdir\" ;; *) thisdir=\"\$thisdir/\$destdir\" ;; esac fi file=\`\$ECHO \"\$file\" | $SED 's%^.*/%%'\` file=\`ls -ld \"\$thisdir/\$file\" | $SED -n 's/.*-> //p'\` done # Usually 'no', except on cygwin/mingw when embedded into # the cwrapper. WRAPPER_SCRIPT_BELONGS_IN_OBJDIR=$func_emit_wrapper_arg1 if test \"\$WRAPPER_SCRIPT_BELONGS_IN_OBJDIR\" = \"yes\"; then # special case for '.' if test \"\$thisdir\" = \".\"; then thisdir=\`pwd\` fi # remove .libs from thisdir case \"\$thisdir\" in *[\\\\/]$objdir ) thisdir=\`\$ECHO \"\$thisdir\" | $SED 's%[\\\\/][^\\\\/]*$%%'\` ;; $objdir ) thisdir=. ;; esac fi # Try to get the absolute directory name. absdir=\`cd \"\$thisdir\" && pwd\` test -n \"\$absdir\" && thisdir=\"\$absdir\" " if test "$fast_install" = yes; then $ECHO "\ program=lt-'$outputname'$exeext progdir=\"\$thisdir/$objdir\" if test ! -f \"\$progdir/\$program\" || { file=\`ls -1dt \"\$progdir/\$program\" \"\$progdir/../\$program\" 2>/dev/null | ${SED} 1q\`; \\ test \"X\$file\" != \"X\$progdir/\$program\"; }; then file=\"\$\$-\$program\" if test ! -d \"\$progdir\"; then $MKDIR \"\$progdir\" else $RM \"\$progdir/\$file\" fi" $ECHO "\ # relink executable if necessary if test -n \"\$relink_command\"; then if relink_command_output=\`eval \$relink_command 2>&1\`; then : else $ECHO \"\$relink_command_output\" >&2 $RM \"\$progdir/\$file\" exit 1 fi fi $MV \"\$progdir/\$file\" \"\$progdir/\$program\" 2>/dev/null || { $RM \"\$progdir/\$program\"; $MV \"\$progdir/\$file\" \"\$progdir/\$program\"; } $RM \"\$progdir/\$file\" fi" else $ECHO "\ program='$outputname' progdir=\"\$thisdir/$objdir\" " fi $ECHO "\ if test -f \"\$progdir/\$program\"; then" # fixup the dll searchpath if we need to. # # Fix the DLL searchpath if we need to. Do this before prepending # to shlibpath, because on Windows, both are PATH and uninstalled # libraries must come first. if test -n "$dllsearchpath"; then $ECHO "\ # Add the dll search path components to the executable PATH PATH=$dllsearchpath:\$PATH " fi # Export our shlibpath_var if we have one. if test "$shlibpath_overrides_runpath" = yes && test -n "$shlibpath_var" && test -n "$temp_rpath"; then $ECHO "\ # Add our own library path to $shlibpath_var $shlibpath_var=\"$temp_rpath\$$shlibpath_var\" # Some systems cannot cope with colon-terminated $shlibpath_var # The second colon is a workaround for a bug in BeOS R4 sed $shlibpath_var=\`\$ECHO \"\$$shlibpath_var\" | $SED 's/::*\$//'\` export $shlibpath_var " fi $ECHO "\ if test \"\$libtool_execute_magic\" != \"$magic\"; then # Run the actual program with our arguments. func_exec_program \${1+\"\$@\"} fi else # The program doesn't exist. \$ECHO \"\$0: error: \\\`\$progdir/\$program' does not exist\" 1>&2 \$ECHO \"This script is just a wrapper for \$program.\" 1>&2 \$ECHO \"See the $PACKAGE documentation for more information.\" 1>&2 exit 1 fi fi\ " } # func_emit_cwrapperexe_src # emit the source code for a wrapper executable on stdout # Must ONLY be called from within func_mode_link because # it depends on a number of variable set therein. func_emit_cwrapperexe_src () { cat < #include #ifdef _MSC_VER # include # include # include #else # include # include # ifdef __CYGWIN__ # include # endif #endif #include #include #include #include #include #include #include #include /* declarations of non-ANSI functions */ #if defined(__MINGW32__) # ifdef __STRICT_ANSI__ int _putenv (const char *); # endif #elif defined(__CYGWIN__) # ifdef __STRICT_ANSI__ char *realpath (const char *, char *); int putenv (char *); int setenv (const char *, const char *, int); # endif /* #elif defined (other platforms) ... */ #endif /* portability defines, excluding path handling macros */ #if defined(_MSC_VER) # define setmode _setmode # define stat _stat # define chmod _chmod # define getcwd _getcwd # define putenv _putenv # define S_IXUSR _S_IEXEC # ifndef _INTPTR_T_DEFINED # define _INTPTR_T_DEFINED # define intptr_t int # endif #elif defined(__MINGW32__) # define setmode _setmode # define stat _stat # define chmod _chmod # define getcwd _getcwd # define putenv _putenv #elif defined(__CYGWIN__) # define HAVE_SETENV # define FOPEN_WB "wb" /* #elif defined (other platforms) ... */ #endif #if defined(PATH_MAX) # define LT_PATHMAX PATH_MAX #elif defined(MAXPATHLEN) # define LT_PATHMAX MAXPATHLEN #else # define LT_PATHMAX 1024 #endif #ifndef S_IXOTH # define S_IXOTH 0 #endif #ifndef S_IXGRP # define S_IXGRP 0 #endif /* path handling portability macros */ #ifndef DIR_SEPARATOR # define DIR_SEPARATOR '/' # define PATH_SEPARATOR ':' #endif #if defined (_WIN32) || defined (__MSDOS__) || defined (__DJGPP__) || \ defined (__OS2__) # define HAVE_DOS_BASED_FILE_SYSTEM # define FOPEN_WB "wb" # ifndef DIR_SEPARATOR_2 # define DIR_SEPARATOR_2 '\\' # endif # ifndef PATH_SEPARATOR_2 # define PATH_SEPARATOR_2 ';' # endif #endif #ifndef DIR_SEPARATOR_2 # define IS_DIR_SEPARATOR(ch) ((ch) == DIR_SEPARATOR) #else /* DIR_SEPARATOR_2 */ # define IS_DIR_SEPARATOR(ch) \ (((ch) == DIR_SEPARATOR) || ((ch) == DIR_SEPARATOR_2)) #endif /* DIR_SEPARATOR_2 */ #ifndef PATH_SEPARATOR_2 # define IS_PATH_SEPARATOR(ch) ((ch) == PATH_SEPARATOR) #else /* PATH_SEPARATOR_2 */ # define IS_PATH_SEPARATOR(ch) ((ch) == PATH_SEPARATOR_2) #endif /* PATH_SEPARATOR_2 */ #ifndef FOPEN_WB # define FOPEN_WB "w" #endif #ifndef _O_BINARY # define _O_BINARY 0 #endif #define XMALLOC(type, num) ((type *) xmalloc ((num) * sizeof(type))) #define XFREE(stale) do { \ if (stale) { free ((void *) stale); stale = 0; } \ } while (0) #if defined(LT_DEBUGWRAPPER) static int lt_debug = 1; #else static int lt_debug = 0; #endif const char *program_name = "libtool-wrapper"; /* in case xstrdup fails */ void *xmalloc (size_t num); char *xstrdup (const char *string); const char *base_name (const char *name); char *find_executable (const char *wrapper); char *chase_symlinks (const char *pathspec); int make_executable (const char *path); int check_executable (const char *path); char *strendzap (char *str, const char *pat); void lt_debugprintf (const char *file, int line, const char *fmt, ...); void lt_fatal (const char *file, int line, const char *message, ...); static const char *nonnull (const char *s); static const char *nonempty (const char *s); void lt_setenv (const char *name, const char *value); char *lt_extend_str (const char *orig_value, const char *add, int to_end); void lt_update_exe_path (const char *name, const char *value); void lt_update_lib_path (const char *name, const char *value); char **prepare_spawn (char **argv); void lt_dump_script (FILE *f); EOF cat <= 0) && (st.st_mode & (S_IXUSR | S_IXGRP | S_IXOTH))) return 1; else return 0; } int make_executable (const char *path) { int rval = 0; struct stat st; lt_debugprintf (__FILE__, __LINE__, "(make_executable): %s\n", nonempty (path)); if ((!path) || (!*path)) return 0; if (stat (path, &st) >= 0) { rval = chmod (path, st.st_mode | S_IXOTH | S_IXGRP | S_IXUSR); } return rval; } /* Searches for the full path of the wrapper. Returns newly allocated full path name if found, NULL otherwise Does not chase symlinks, even on platforms that support them. */ char * find_executable (const char *wrapper) { int has_slash = 0; const char *p; const char *p_next; /* static buffer for getcwd */ char tmp[LT_PATHMAX + 1]; int tmp_len; char *concat_name; lt_debugprintf (__FILE__, __LINE__, "(find_executable): %s\n", nonempty (wrapper)); if ((wrapper == NULL) || (*wrapper == '\0')) return NULL; /* Absolute path? */ #if defined (HAVE_DOS_BASED_FILE_SYSTEM) if (isalpha ((unsigned char) wrapper[0]) && wrapper[1] == ':') { concat_name = xstrdup (wrapper); if (check_executable (concat_name)) return concat_name; XFREE (concat_name); } else { #endif if (IS_DIR_SEPARATOR (wrapper[0])) { concat_name = xstrdup (wrapper); if (check_executable (concat_name)) return concat_name; XFREE (concat_name); } #if defined (HAVE_DOS_BASED_FILE_SYSTEM) } #endif for (p = wrapper; *p; p++) if (*p == '/') { has_slash = 1; break; } if (!has_slash) { /* no slashes; search PATH */ const char *path = getenv ("PATH"); if (path != NULL) { for (p = path; *p; p = p_next) { const char *q; size_t p_len; for (q = p; *q; q++) if (IS_PATH_SEPARATOR (*q)) break; p_len = q - p; p_next = (*q == '\0' ? q : q + 1); if (p_len == 0) { /* empty path: current directory */ if (getcwd (tmp, LT_PATHMAX) == NULL) lt_fatal (__FILE__, __LINE__, "getcwd failed: %s", nonnull (strerror (errno))); tmp_len = strlen (tmp); concat_name = XMALLOC (char, tmp_len + 1 + strlen (wrapper) + 1); memcpy (concat_name, tmp, tmp_len); concat_name[tmp_len] = '/'; strcpy (concat_name + tmp_len + 1, wrapper); } else { concat_name = XMALLOC (char, p_len + 1 + strlen (wrapper) + 1); memcpy (concat_name, p, p_len); concat_name[p_len] = '/'; strcpy (concat_name + p_len + 1, wrapper); } if (check_executable (concat_name)) return concat_name; XFREE (concat_name); } } /* not found in PATH; assume curdir */ } /* Relative path | not found in path: prepend cwd */ if (getcwd (tmp, LT_PATHMAX) == NULL) lt_fatal (__FILE__, __LINE__, "getcwd failed: %s", nonnull (strerror (errno))); tmp_len = strlen (tmp); concat_name = XMALLOC (char, tmp_len + 1 + strlen (wrapper) + 1); memcpy (concat_name, tmp, tmp_len); concat_name[tmp_len] = '/'; strcpy (concat_name + tmp_len + 1, wrapper); if (check_executable (concat_name)) return concat_name; XFREE (concat_name); return NULL; } char * chase_symlinks (const char *pathspec) { #ifndef S_ISLNK return xstrdup (pathspec); #else char buf[LT_PATHMAX]; struct stat s; char *tmp_pathspec = xstrdup (pathspec); char *p; int has_symlinks = 0; while (strlen (tmp_pathspec) && !has_symlinks) { lt_debugprintf (__FILE__, __LINE__, "checking path component for symlinks: %s\n", tmp_pathspec); if (lstat (tmp_pathspec, &s) == 0) { if (S_ISLNK (s.st_mode) != 0) { has_symlinks = 1; break; } /* search backwards for last DIR_SEPARATOR */ p = tmp_pathspec + strlen (tmp_pathspec) - 1; while ((p > tmp_pathspec) && (!IS_DIR_SEPARATOR (*p))) p--; if ((p == tmp_pathspec) && (!IS_DIR_SEPARATOR (*p))) { /* no more DIR_SEPARATORS left */ break; } *p = '\0'; } else { lt_fatal (__FILE__, __LINE__, "error accessing file \"%s\": %s", tmp_pathspec, nonnull (strerror (errno))); } } XFREE (tmp_pathspec); if (!has_symlinks) { return xstrdup (pathspec); } tmp_pathspec = realpath (pathspec, buf); if (tmp_pathspec == 0) { lt_fatal (__FILE__, __LINE__, "could not follow symlinks for %s", pathspec); } return xstrdup (tmp_pathspec); #endif } char * strendzap (char *str, const char *pat) { size_t len, patlen; assert (str != NULL); assert (pat != NULL); len = strlen (str); patlen = strlen (pat); if (patlen <= len) { str += len - patlen; if (strcmp (str, pat) == 0) *str = '\0'; } return str; } void lt_debugprintf (const char *file, int line, const char *fmt, ...) { va_list args; if (lt_debug) { (void) fprintf (stderr, "%s:%s:%d: ", program_name, file, line); va_start (args, fmt); (void) vfprintf (stderr, fmt, args); va_end (args); } } static void lt_error_core (int exit_status, const char *file, int line, const char *mode, const char *message, va_list ap) { fprintf (stderr, "%s:%s:%d: %s: ", program_name, file, line, mode); vfprintf (stderr, message, ap); fprintf (stderr, ".\n"); if (exit_status >= 0) exit (exit_status); } void lt_fatal (const char *file, int line, const char *message, ...) { va_list ap; va_start (ap, message); lt_error_core (EXIT_FAILURE, file, line, "FATAL", message, ap); va_end (ap); } static const char * nonnull (const char *s) { return s ? s : "(null)"; } static const char * nonempty (const char *s) { return (s && !*s) ? "(empty)" : nonnull (s); } void lt_setenv (const char *name, const char *value) { lt_debugprintf (__FILE__, __LINE__, "(lt_setenv) setting '%s' to '%s'\n", nonnull (name), nonnull (value)); { #ifdef HAVE_SETENV /* always make a copy, for consistency with !HAVE_SETENV */ char *str = xstrdup (value); setenv (name, str, 1); #else int len = strlen (name) + 1 + strlen (value) + 1; char *str = XMALLOC (char, len); sprintf (str, "%s=%s", name, value); if (putenv (str) != EXIT_SUCCESS) { XFREE (str); } #endif } } char * lt_extend_str (const char *orig_value, const char *add, int to_end) { char *new_value; if (orig_value && *orig_value) { int orig_value_len = strlen (orig_value); int add_len = strlen (add); new_value = XMALLOC (char, add_len + orig_value_len + 1); if (to_end) { strcpy (new_value, orig_value); strcpy (new_value + orig_value_len, add); } else { strcpy (new_value, add); strcpy (new_value + add_len, orig_value); } } else { new_value = xstrdup (add); } return new_value; } void lt_update_exe_path (const char *name, const char *value) { lt_debugprintf (__FILE__, __LINE__, "(lt_update_exe_path) modifying '%s' by prepending '%s'\n", nonnull (name), nonnull (value)); if (name && *name && value && *value) { char *new_value = lt_extend_str (getenv (name), value, 0); /* some systems can't cope with a ':'-terminated path #' */ int len = strlen (new_value); while (((len = strlen (new_value)) > 0) && IS_PATH_SEPARATOR (new_value[len-1])) { new_value[len-1] = '\0'; } lt_setenv (name, new_value); XFREE (new_value); } } void lt_update_lib_path (const char *name, const char *value) { lt_debugprintf (__FILE__, __LINE__, "(lt_update_lib_path) modifying '%s' by prepending '%s'\n", nonnull (name), nonnull (value)); if (name && *name && value && *value) { char *new_value = lt_extend_str (getenv (name), value, 0); lt_setenv (name, new_value); XFREE (new_value); } } EOF case $host_os in mingw*) cat <<"EOF" /* Prepares an argument vector before calling spawn(). Note that spawn() does not by itself call the command interpreter (getenv ("COMSPEC") != NULL ? getenv ("COMSPEC") : ({ OSVERSIONINFO v; v.dwOSVersionInfoSize = sizeof(OSVERSIONINFO); GetVersionEx(&v); v.dwPlatformId == VER_PLATFORM_WIN32_NT; }) ? "cmd.exe" : "command.com"). Instead it simply concatenates the arguments, separated by ' ', and calls CreateProcess(). We must quote the arguments since Win32 CreateProcess() interprets characters like ' ', '\t', '\\', '"' (but not '<' and '>') in a special way: - Space and tab are interpreted as delimiters. They are not treated as delimiters if they are surrounded by double quotes: "...". - Unescaped double quotes are removed from the input. Their only effect is that within double quotes, space and tab are treated like normal characters. - Backslashes not followed by double quotes are not special. - But 2*n+1 backslashes followed by a double quote become n backslashes followed by a double quote (n >= 0): \" -> " \\\" -> \" \\\\\" -> \\" */ #define SHELL_SPECIAL_CHARS "\"\\ \001\002\003\004\005\006\007\010\011\012\013\014\015\016\017\020\021\022\023\024\025\026\027\030\031\032\033\034\035\036\037" #define SHELL_SPACE_CHARS " \001\002\003\004\005\006\007\010\011\012\013\014\015\016\017\020\021\022\023\024\025\026\027\030\031\032\033\034\035\036\037" char ** prepare_spawn (char **argv) { size_t argc; char **new_argv; size_t i; /* Count number of arguments. */ for (argc = 0; argv[argc] != NULL; argc++) ; /* Allocate new argument vector. */ new_argv = XMALLOC (char *, argc + 1); /* Put quoted arguments into the new argument vector. */ for (i = 0; i < argc; i++) { const char *string = argv[i]; if (string[0] == '\0') new_argv[i] = xstrdup ("\"\""); else if (strpbrk (string, SHELL_SPECIAL_CHARS) != NULL) { int quote_around = (strpbrk (string, SHELL_SPACE_CHARS) != NULL); size_t length; unsigned int backslashes; const char *s; char *quoted_string; char *p; length = 0; backslashes = 0; if (quote_around) length++; for (s = string; *s != '\0'; s++) { char c = *s; if (c == '"') length += backslashes + 1; length++; if (c == '\\') backslashes++; else backslashes = 0; } if (quote_around) length += backslashes + 1; quoted_string = XMALLOC (char, length + 1); p = quoted_string; backslashes = 0; if (quote_around) *p++ = '"'; for (s = string; *s != '\0'; s++) { char c = *s; if (c == '"') { unsigned int j; for (j = backslashes + 1; j > 0; j--) *p++ = '\\'; } *p++ = c; if (c == '\\') backslashes++; else backslashes = 0; } if (quote_around) { unsigned int j; for (j = backslashes; j > 0; j--) *p++ = '\\'; *p++ = '"'; } *p = '\0'; new_argv[i] = quoted_string; } else new_argv[i] = (char *) string; } new_argv[argc] = NULL; return new_argv; } EOF ;; esac cat <<"EOF" void lt_dump_script (FILE* f) { EOF func_emit_wrapper yes | $SED -n -e ' s/^\(.\{79\}\)\(..*\)/\1\ \2/ h s/\([\\"]\)/\\\1/g s/$/\\n/ s/\([^\n]*\).*/ fputs ("\1", f);/p g D' cat <<"EOF" } EOF } # end: func_emit_cwrapperexe_src # func_win32_import_lib_p ARG # True if ARG is an import lib, as indicated by $file_magic_cmd func_win32_import_lib_p () { $opt_debug case `eval $file_magic_cmd \"\$1\" 2>/dev/null | $SED -e 10q` in *import*) : ;; *) false ;; esac } # func_mode_link arg... func_mode_link () { $opt_debug case $host in *-*-cygwin* | *-*-mingw* | *-*-pw32* | *-*-os2* | *-cegcc*) # It is impossible to link a dll without this setting, and # we shouldn't force the makefile maintainer to figure out # which system we are compiling for in order to pass an extra # flag for every libtool invocation. # allow_undefined=no # FIXME: Unfortunately, there are problems with the above when trying # to make a dll which has undefined symbols, in which case not # even a static library is built. For now, we need to specify # -no-undefined on the libtool link line when we can be certain # that all symbols are satisfied, otherwise we get a static library. allow_undefined=yes ;; *) allow_undefined=yes ;; esac libtool_args=$nonopt base_compile="$nonopt $@" compile_command=$nonopt finalize_command=$nonopt compile_rpath= finalize_rpath= compile_shlibpath= finalize_shlibpath= convenience= old_convenience= deplibs= old_deplibs= compiler_flags= linker_flags= dllsearchpath= lib_search_path=`pwd` inst_prefix_dir= new_inherited_linker_flags= avoid_version=no bindir= dlfiles= dlprefiles= dlself=no export_dynamic=no export_symbols= export_symbols_regex= generated= libobjs= ltlibs= module=no no_install=no objs= non_pic_objects= precious_files_regex= prefer_static_libs=no preload=no prev= prevarg= release= rpath= xrpath= perm_rpath= temp_rpath= thread_safe=no vinfo= vinfo_number=no weak_libs= single_module="${wl}-single_module" func_infer_tag $base_compile # We need to know -static, to get the right output filenames. for arg do case $arg in -shared) test "$build_libtool_libs" != yes && \ func_fatal_configuration "can not build a shared library" build_old_libs=no break ;; -all-static | -static | -static-libtool-libs) case $arg in -all-static) if test "$build_libtool_libs" = yes && test -z "$link_static_flag"; then func_warning "complete static linking is impossible in this configuration" fi if test -n "$link_static_flag"; then dlopen_self=$dlopen_self_static fi prefer_static_libs=yes ;; -static) if test -z "$pic_flag" && test -n "$link_static_flag"; then dlopen_self=$dlopen_self_static fi prefer_static_libs=built ;; -static-libtool-libs) if test -z "$pic_flag" && test -n "$link_static_flag"; then dlopen_self=$dlopen_self_static fi prefer_static_libs=yes ;; esac build_libtool_libs=no build_old_libs=yes break ;; esac done # See if our shared archives depend on static archives. test -n "$old_archive_from_new_cmds" && build_old_libs=yes # Go through the arguments, transforming them on the way. while test "$#" -gt 0; do arg="$1" shift func_quote_for_eval "$arg" qarg=$func_quote_for_eval_unquoted_result func_append libtool_args " $func_quote_for_eval_result" # If the previous option needs an argument, assign it. if test -n "$prev"; then case $prev in output) func_append compile_command " @OUTPUT@" func_append finalize_command " @OUTPUT@" ;; esac case $prev in bindir) bindir="$arg" prev= continue ;; dlfiles|dlprefiles) if test "$preload" = no; then # Add the symbol object into the linking commands. func_append compile_command " @SYMFILE@" func_append finalize_command " @SYMFILE@" preload=yes fi case $arg in *.la | *.lo) ;; # We handle these cases below. force) if test "$dlself" = no; then dlself=needless export_dynamic=yes fi prev= continue ;; self) if test "$prev" = dlprefiles; then dlself=yes elif test "$prev" = dlfiles && test "$dlopen_self" != yes; then dlself=yes else dlself=needless export_dynamic=yes fi prev= continue ;; *) if test "$prev" = dlfiles; then func_append dlfiles " $arg" else func_append dlprefiles " $arg" fi prev= continue ;; esac ;; expsyms) export_symbols="$arg" test -f "$arg" \ || func_fatal_error "symbol file \`$arg' does not exist" prev= continue ;; expsyms_regex) export_symbols_regex="$arg" prev= continue ;; framework) case $host in *-*-darwin*) case "$deplibs " in *" $qarg.ltframework "*) ;; *) func_append deplibs " $qarg.ltframework" # this is fixed later ;; esac ;; esac prev= continue ;; inst_prefix) inst_prefix_dir="$arg" prev= continue ;; objectlist) if test -f "$arg"; then save_arg=$arg moreargs= for fil in `cat "$save_arg"` do # func_append moreargs " $fil" arg=$fil # A libtool-controlled object. # Check to see that this really is a libtool object. if func_lalib_unsafe_p "$arg"; then pic_object= non_pic_object= # Read the .lo file func_source "$arg" if test -z "$pic_object" || test -z "$non_pic_object" || test "$pic_object" = none && test "$non_pic_object" = none; then func_fatal_error "cannot find name of object for \`$arg'" fi # Extract subdirectory from the argument. func_dirname "$arg" "/" "" xdir="$func_dirname_result" if test "$pic_object" != none; then # Prepend the subdirectory the object is found in. pic_object="$xdir$pic_object" if test "$prev" = dlfiles; then if test "$build_libtool_libs" = yes && test "$dlopen_support" = yes; then func_append dlfiles " $pic_object" prev= continue else # If libtool objects are unsupported, then we need to preload. prev=dlprefiles fi fi # CHECK ME: I think I busted this. -Ossama if test "$prev" = dlprefiles; then # Preload the old-style object. func_append dlprefiles " $pic_object" prev= fi # A PIC object. func_append libobjs " $pic_object" arg="$pic_object" fi # Non-PIC object. if test "$non_pic_object" != none; then # Prepend the subdirectory the object is found in. non_pic_object="$xdir$non_pic_object" # A standard non-PIC object func_append non_pic_objects " $non_pic_object" if test -z "$pic_object" || test "$pic_object" = none ; then arg="$non_pic_object" fi else # If the PIC object exists, use it instead. # $xdir was prepended to $pic_object above. non_pic_object="$pic_object" func_append non_pic_objects " $non_pic_object" fi else # Only an error if not doing a dry-run. if $opt_dry_run; then # Extract subdirectory from the argument. func_dirname "$arg" "/" "" xdir="$func_dirname_result" func_lo2o "$arg" pic_object=$xdir$objdir/$func_lo2o_result non_pic_object=$xdir$func_lo2o_result func_append libobjs " $pic_object" func_append non_pic_objects " $non_pic_object" else func_fatal_error "\`$arg' is not a valid libtool object" fi fi done else func_fatal_error "link input file \`$arg' does not exist" fi arg=$save_arg prev= continue ;; precious_regex) precious_files_regex="$arg" prev= continue ;; release) release="-$arg" prev= continue ;; rpath | xrpath) # We need an absolute path. case $arg in [\\/]* | [A-Za-z]:[\\/]*) ;; *) func_fatal_error "only absolute run-paths are allowed" ;; esac if test "$prev" = rpath; then case "$rpath " in *" $arg "*) ;; *) func_append rpath " $arg" ;; esac else case "$xrpath " in *" $arg "*) ;; *) func_append xrpath " $arg" ;; esac fi prev= continue ;; shrext) shrext_cmds="$arg" prev= continue ;; weak) func_append weak_libs " $arg" prev= continue ;; xcclinker) func_append linker_flags " $qarg" func_append compiler_flags " $qarg" prev= func_append compile_command " $qarg" func_append finalize_command " $qarg" continue ;; xcompiler) func_append compiler_flags " $qarg" prev= func_append compile_command " $qarg" func_append finalize_command " $qarg" continue ;; xlinker) func_append linker_flags " $qarg" func_append compiler_flags " $wl$qarg" prev= func_append compile_command " $wl$qarg" func_append finalize_command " $wl$qarg" continue ;; *) eval "$prev=\"\$arg\"" prev= continue ;; esac fi # test -n "$prev" prevarg="$arg" case $arg in -all-static) if test -n "$link_static_flag"; then # See comment for -static flag below, for more details. func_append compile_command " $link_static_flag" func_append finalize_command " $link_static_flag" fi continue ;; -allow-undefined) # FIXME: remove this flag sometime in the future. func_fatal_error "\`-allow-undefined' must not be used because it is the default" ;; -avoid-version) avoid_version=yes continue ;; -bindir) prev=bindir continue ;; -dlopen) prev=dlfiles continue ;; -dlpreopen) prev=dlprefiles continue ;; -export-dynamic) export_dynamic=yes continue ;; -export-symbols | -export-symbols-regex) if test -n "$export_symbols" || test -n "$export_symbols_regex"; then func_fatal_error "more than one -exported-symbols argument is not allowed" fi if test "X$arg" = "X-export-symbols"; then prev=expsyms else prev=expsyms_regex fi continue ;; -framework) prev=framework continue ;; -inst-prefix-dir) prev=inst_prefix continue ;; # The native IRIX linker understands -LANG:*, -LIST:* and -LNO:* # so, if we see these flags be careful not to treat them like -L -L[A-Z][A-Z]*:*) case $with_gcc/$host in no/*-*-irix* | /*-*-irix*) func_append compile_command " $arg" func_append finalize_command " $arg" ;; esac continue ;; -L*) func_stripname "-L" '' "$arg" if test -z "$func_stripname_result"; then if test "$#" -gt 0; then func_fatal_error "require no space between \`-L' and \`$1'" else func_fatal_error "need path for \`-L' option" fi fi func_resolve_sysroot "$func_stripname_result" dir=$func_resolve_sysroot_result # We need an absolute path. case $dir in [\\/]* | [A-Za-z]:[\\/]*) ;; *) absdir=`cd "$dir" && pwd` test -z "$absdir" && \ func_fatal_error "cannot determine absolute directory name of \`$dir'" dir="$absdir" ;; esac case "$deplibs " in *" -L$dir "* | *" $arg "*) # Will only happen for absolute or sysroot arguments ;; *) # Preserve sysroot, but never include relative directories case $dir in [\\/]* | [A-Za-z]:[\\/]* | =*) func_append deplibs " $arg" ;; *) func_append deplibs " -L$dir" ;; esac func_append lib_search_path " $dir" ;; esac case $host in *-*-cygwin* | *-*-mingw* | *-*-pw32* | *-*-os2* | *-cegcc*) testbindir=`$ECHO "$dir" | $SED 's*/lib$*/bin*'` case :$dllsearchpath: in *":$dir:"*) ;; ::) dllsearchpath=$dir;; *) func_append dllsearchpath ":$dir";; esac case :$dllsearchpath: in *":$testbindir:"*) ;; ::) dllsearchpath=$testbindir;; *) func_append dllsearchpath ":$testbindir";; esac ;; esac continue ;; -l*) if test "X$arg" = "X-lc" || test "X$arg" = "X-lm"; then case $host in *-*-cygwin* | *-*-mingw* | *-*-pw32* | *-*-beos* | *-cegcc* | *-*-haiku*) # These systems don't actually have a C or math library (as such) continue ;; *-*-os2*) # These systems don't actually have a C library (as such) test "X$arg" = "X-lc" && continue ;; *-*-openbsd* | *-*-freebsd* | *-*-dragonfly*) # Do not include libc due to us having libc/libc_r. test "X$arg" = "X-lc" && continue ;; *-*-rhapsody* | *-*-darwin1.[012]) # Rhapsody C and math libraries are in the System framework func_append deplibs " System.ltframework" continue ;; *-*-sco3.2v5* | *-*-sco5v6*) # Causes problems with __ctype test "X$arg" = "X-lc" && continue ;; *-*-sysv4.2uw2* | *-*-sysv5* | *-*-unixware* | *-*-OpenUNIX*) # Compiler inserts libc in the correct place for threads to work test "X$arg" = "X-lc" && continue ;; esac elif test "X$arg" = "X-lc_r"; then case $host in *-*-openbsd* | *-*-freebsd* | *-*-dragonfly*) # Do not include libc_r directly, use -pthread flag. continue ;; esac fi func_append deplibs " $arg" continue ;; -module) module=yes continue ;; # Tru64 UNIX uses -model [arg] to determine the layout of C++ # classes, name mangling, and exception handling. # Darwin uses the -arch flag to determine output architecture. -model|-arch|-isysroot|--sysroot) func_append compiler_flags " $arg" func_append compile_command " $arg" func_append finalize_command " $arg" prev=xcompiler continue ;; -mt|-mthreads|-kthread|-Kthread|-pthread|-pthreads|--thread-safe \ |-threads|-fopenmp|-openmp|-mp|-xopenmp|-omp|-qsmp=*) func_append compiler_flags " $arg" func_append compile_command " $arg" func_append finalize_command " $arg" case "$new_inherited_linker_flags " in *" $arg "*) ;; * ) func_append new_inherited_linker_flags " $arg" ;; esac continue ;; -multi_module) single_module="${wl}-multi_module" continue ;; -no-fast-install) fast_install=no continue ;; -no-install) case $host in *-*-cygwin* | *-*-mingw* | *-*-pw32* | *-*-os2* | *-*-darwin* | *-cegcc*) # The PATH hackery in wrapper scripts is required on Windows # and Darwin in order for the loader to find any dlls it needs. func_warning "\`-no-install' is ignored for $host" func_warning "assuming \`-no-fast-install' instead" fast_install=no ;; *) no_install=yes ;; esac continue ;; -no-undefined) allow_undefined=no continue ;; -objectlist) prev=objectlist continue ;; -o) prev=output ;; -precious-files-regex) prev=precious_regex continue ;; -release) prev=release continue ;; -rpath) prev=rpath continue ;; -R) prev=xrpath continue ;; -R*) func_stripname '-R' '' "$arg" dir=$func_stripname_result # We need an absolute path. case $dir in [\\/]* | [A-Za-z]:[\\/]*) ;; =*) func_stripname '=' '' "$dir" dir=$lt_sysroot$func_stripname_result ;; *) func_fatal_error "only absolute run-paths are allowed" ;; esac case "$xrpath " in *" $dir "*) ;; *) func_append xrpath " $dir" ;; esac continue ;; -shared) # The effects of -shared are defined in a previous loop. continue ;; -shrext) prev=shrext continue ;; -static | -static-libtool-libs) # The effects of -static are defined in a previous loop. # We used to do the same as -all-static on platforms that # didn't have a PIC flag, but the assumption that the effects # would be equivalent was wrong. It would break on at least # Digital Unix and AIX. continue ;; -thread-safe) thread_safe=yes continue ;; -version-info) prev=vinfo continue ;; -version-number) prev=vinfo vinfo_number=yes continue ;; -weak) prev=weak continue ;; -Wc,*) func_stripname '-Wc,' '' "$arg" args=$func_stripname_result arg= save_ifs="$IFS"; IFS=',' for flag in $args; do IFS="$save_ifs" func_quote_for_eval "$flag" func_append arg " $func_quote_for_eval_result" func_append compiler_flags " $func_quote_for_eval_result" done IFS="$save_ifs" func_stripname ' ' '' "$arg" arg=$func_stripname_result ;; -Wl,*) func_stripname '-Wl,' '' "$arg" args=$func_stripname_result arg= save_ifs="$IFS"; IFS=',' for flag in $args; do IFS="$save_ifs" func_quote_for_eval "$flag" func_append arg " $wl$func_quote_for_eval_result" func_append compiler_flags " $wl$func_quote_for_eval_result" func_append linker_flags " $func_quote_for_eval_result" done IFS="$save_ifs" func_stripname ' ' '' "$arg" arg=$func_stripname_result ;; -Xcompiler) prev=xcompiler continue ;; -Xlinker) prev=xlinker continue ;; -XCClinker) prev=xcclinker continue ;; # -msg_* for osf cc -msg_*) func_quote_for_eval "$arg" arg="$func_quote_for_eval_result" ;; # Flags to be passed through unchanged, with rationale: # -64, -mips[0-9] enable 64-bit mode for the SGI compiler # -r[0-9][0-9]* specify processor for the SGI compiler # -xarch=*, -xtarget=* enable 64-bit mode for the Sun compiler # +DA*, +DD* enable 64-bit mode for the HP compiler # -q* compiler args for the IBM compiler # -m*, -t[45]*, -txscale* architecture-specific flags for GCC # -F/path path to uninstalled frameworks, gcc on darwin # -p, -pg, --coverage, -fprofile-* profiling flags for GCC # @file GCC response files # -tp=* Portland pgcc target processor selection # --sysroot=* for sysroot support # -O*, -flto*, -fwhopr*, -fuse-linker-plugin GCC link-time optimization -64|-mips[0-9]|-r[0-9][0-9]*|-xarch=*|-xtarget=*|+DA*|+DD*|-q*|-m*| \ -t[45]*|-txscale*|-p|-pg|--coverage|-fprofile-*|-F*|@*|-tp=*|--sysroot=*| \ -O*|-flto*|-fwhopr*|-fuse-linker-plugin) func_quote_for_eval "$arg" arg="$func_quote_for_eval_result" func_append compile_command " $arg" func_append finalize_command " $arg" func_append compiler_flags " $arg" continue ;; # Some other compiler flag. -* | +*) func_quote_for_eval "$arg" arg="$func_quote_for_eval_result" ;; *.$objext) # A standard object. func_append objs " $arg" ;; *.lo) # A libtool-controlled object. # Check to see that this really is a libtool object. if func_lalib_unsafe_p "$arg"; then pic_object= non_pic_object= # Read the .lo file func_source "$arg" if test -z "$pic_object" || test -z "$non_pic_object" || test "$pic_object" = none && test "$non_pic_object" = none; then func_fatal_error "cannot find name of object for \`$arg'" fi # Extract subdirectory from the argument. func_dirname "$arg" "/" "" xdir="$func_dirname_result" if test "$pic_object" != none; then # Prepend the subdirectory the object is found in. pic_object="$xdir$pic_object" if test "$prev" = dlfiles; then if test "$build_libtool_libs" = yes && test "$dlopen_support" = yes; then func_append dlfiles " $pic_object" prev= continue else # If libtool objects are unsupported, then we need to preload. prev=dlprefiles fi fi # CHECK ME: I think I busted this. -Ossama if test "$prev" = dlprefiles; then # Preload the old-style object. func_append dlprefiles " $pic_object" prev= fi # A PIC object. func_append libobjs " $pic_object" arg="$pic_object" fi # Non-PIC object. if test "$non_pic_object" != none; then # Prepend the subdirectory the object is found in. non_pic_object="$xdir$non_pic_object" # A standard non-PIC object func_append non_pic_objects " $non_pic_object" if test -z "$pic_object" || test "$pic_object" = none ; then arg="$non_pic_object" fi else # If the PIC object exists, use it instead. # $xdir was prepended to $pic_object above. non_pic_object="$pic_object" func_append non_pic_objects " $non_pic_object" fi else # Only an error if not doing a dry-run. if $opt_dry_run; then # Extract subdirectory from the argument. func_dirname "$arg" "/" "" xdir="$func_dirname_result" func_lo2o "$arg" pic_object=$xdir$objdir/$func_lo2o_result non_pic_object=$xdir$func_lo2o_result func_append libobjs " $pic_object" func_append non_pic_objects " $non_pic_object" else func_fatal_error "\`$arg' is not a valid libtool object" fi fi ;; *.$libext) # An archive. func_append deplibs " $arg" func_append old_deplibs " $arg" continue ;; *.la) # A libtool-controlled library. func_resolve_sysroot "$arg" if test "$prev" = dlfiles; then # This library was specified with -dlopen. func_append dlfiles " $func_resolve_sysroot_result" prev= elif test "$prev" = dlprefiles; then # The library was specified with -dlpreopen. func_append dlprefiles " $func_resolve_sysroot_result" prev= else func_append deplibs " $func_resolve_sysroot_result" fi continue ;; # Some other compiler argument. *) # Unknown arguments in both finalize_command and compile_command need # to be aesthetically quoted because they are evaled later. func_quote_for_eval "$arg" arg="$func_quote_for_eval_result" ;; esac # arg # Now actually substitute the argument into the commands. if test -n "$arg"; then func_append compile_command " $arg" func_append finalize_command " $arg" fi done # argument parsing loop test -n "$prev" && \ func_fatal_help "the \`$prevarg' option requires an argument" if test "$export_dynamic" = yes && test -n "$export_dynamic_flag_spec"; then eval arg=\"$export_dynamic_flag_spec\" func_append compile_command " $arg" func_append finalize_command " $arg" fi oldlibs= # calculate the name of the file, without its directory func_basename "$output" outputname="$func_basename_result" libobjs_save="$libobjs" if test -n "$shlibpath_var"; then # get the directories listed in $shlibpath_var eval shlib_search_path=\`\$ECHO \"\${$shlibpath_var}\" \| \$SED \'s/:/ /g\'\` else shlib_search_path= fi eval sys_lib_search_path=\"$sys_lib_search_path_spec\" eval sys_lib_dlsearch_path=\"$sys_lib_dlsearch_path_spec\" func_dirname "$output" "/" "" output_objdir="$func_dirname_result$objdir" func_to_tool_file "$output_objdir/" tool_output_objdir=$func_to_tool_file_result # Create the object directory. func_mkdir_p "$output_objdir" # Determine the type of output case $output in "") func_fatal_help "you must specify an output file" ;; *.$libext) linkmode=oldlib ;; *.lo | *.$objext) linkmode=obj ;; *.la) linkmode=lib ;; *) linkmode=prog ;; # Anything else should be a program. esac specialdeplibs= libs= # Find all interdependent deplibs by searching for libraries # that are linked more than once (e.g. -la -lb -la) for deplib in $deplibs; do if $opt_preserve_dup_deps ; then case "$libs " in *" $deplib "*) func_append specialdeplibs " $deplib" ;; esac fi func_append libs " $deplib" done if test "$linkmode" = lib; then libs="$predeps $libs $compiler_lib_search_path $postdeps" # Compute libraries that are listed more than once in $predeps # $postdeps and mark them as special (i.e., whose duplicates are # not to be eliminated). pre_post_deps= if $opt_duplicate_compiler_generated_deps; then for pre_post_dep in $predeps $postdeps; do case "$pre_post_deps " in *" $pre_post_dep "*) func_append specialdeplibs " $pre_post_deps" ;; esac func_append pre_post_deps " $pre_post_dep" done fi pre_post_deps= fi deplibs= newdependency_libs= newlib_search_path= need_relink=no # whether we're linking any uninstalled libtool libraries notinst_deplibs= # not-installed libtool libraries notinst_path= # paths that contain not-installed libtool libraries case $linkmode in lib) passes="conv dlpreopen link" for file in $dlfiles $dlprefiles; do case $file in *.la) ;; *) func_fatal_help "libraries can \`-dlopen' only libtool libraries: $file" ;; esac done ;; prog) compile_deplibs= finalize_deplibs= alldeplibs=no newdlfiles= newdlprefiles= passes="conv scan dlopen dlpreopen link" ;; *) passes="conv" ;; esac for pass in $passes; do # The preopen pass in lib mode reverses $deplibs; put it back here # so that -L comes before libs that need it for instance... if test "$linkmode,$pass" = "lib,link"; then ## FIXME: Find the place where the list is rebuilt in the wrong ## order, and fix it there properly tmp_deplibs= for deplib in $deplibs; do tmp_deplibs="$deplib $tmp_deplibs" done deplibs="$tmp_deplibs" fi if test "$linkmode,$pass" = "lib,link" || test "$linkmode,$pass" = "prog,scan"; then libs="$deplibs" deplibs= fi if test "$linkmode" = prog; then case $pass in dlopen) libs="$dlfiles" ;; dlpreopen) libs="$dlprefiles" ;; link) libs="$deplibs %DEPLIBS%" test "X$link_all_deplibs" != Xno && libs="$libs $dependency_libs" ;; esac fi if test "$linkmode,$pass" = "lib,dlpreopen"; then # Collect and forward deplibs of preopened libtool libs for lib in $dlprefiles; do # Ignore non-libtool-libs dependency_libs= func_resolve_sysroot "$lib" case $lib in *.la) func_source "$func_resolve_sysroot_result" ;; esac # Collect preopened libtool deplibs, except any this library # has declared as weak libs for deplib in $dependency_libs; do func_basename "$deplib" deplib_base=$func_basename_result case " $weak_libs " in *" $deplib_base "*) ;; *) func_append deplibs " $deplib" ;; esac done done libs="$dlprefiles" fi if test "$pass" = dlopen; then # Collect dlpreopened libraries save_deplibs="$deplibs" deplibs= fi for deplib in $libs; do lib= found=no case $deplib in -mt|-mthreads|-kthread|-Kthread|-pthread|-pthreads|--thread-safe \ |-threads|-fopenmp|-openmp|-mp|-xopenmp|-omp|-qsmp=*) if test "$linkmode,$pass" = "prog,link"; then compile_deplibs="$deplib $compile_deplibs" finalize_deplibs="$deplib $finalize_deplibs" else func_append compiler_flags " $deplib" if test "$linkmode" = lib ; then case "$new_inherited_linker_flags " in *" $deplib "*) ;; * ) func_append new_inherited_linker_flags " $deplib" ;; esac fi fi continue ;; -l*) if test "$linkmode" != lib && test "$linkmode" != prog; then func_warning "\`-l' is ignored for archives/objects" continue fi func_stripname '-l' '' "$deplib" name=$func_stripname_result if test "$linkmode" = lib; then searchdirs="$newlib_search_path $lib_search_path $compiler_lib_search_dirs $sys_lib_search_path $shlib_search_path" else searchdirs="$newlib_search_path $lib_search_path $sys_lib_search_path $shlib_search_path" fi for searchdir in $searchdirs; do for search_ext in .la $std_shrext .so .a; do # Search the libtool library lib="$searchdir/lib${name}${search_ext}" if test -f "$lib"; then if test "$search_ext" = ".la"; then found=yes else found=no fi break 2 fi done done if test "$found" != yes; then # deplib doesn't seem to be a libtool library if test "$linkmode,$pass" = "prog,link"; then compile_deplibs="$deplib $compile_deplibs" finalize_deplibs="$deplib $finalize_deplibs" else deplibs="$deplib $deplibs" test "$linkmode" = lib && newdependency_libs="$deplib $newdependency_libs" fi continue else # deplib is a libtool library # If $allow_libtool_libs_with_static_runtimes && $deplib is a stdlib, # We need to do some special things here, and not later. if test "X$allow_libtool_libs_with_static_runtimes" = "Xyes" ; then case " $predeps $postdeps " in *" $deplib "*) if func_lalib_p "$lib"; then library_names= old_library= func_source "$lib" for l in $old_library $library_names; do ll="$l" done if test "X$ll" = "X$old_library" ; then # only static version available found=no func_dirname "$lib" "" "." ladir="$func_dirname_result" lib=$ladir/$old_library if test "$linkmode,$pass" = "prog,link"; then compile_deplibs="$deplib $compile_deplibs" finalize_deplibs="$deplib $finalize_deplibs" else deplibs="$deplib $deplibs" test "$linkmode" = lib && newdependency_libs="$deplib $newdependency_libs" fi continue fi fi ;; *) ;; esac fi fi ;; # -l *.ltframework) if test "$linkmode,$pass" = "prog,link"; then compile_deplibs="$deplib $compile_deplibs" finalize_deplibs="$deplib $finalize_deplibs" else deplibs="$deplib $deplibs" if test "$linkmode" = lib ; then case "$new_inherited_linker_flags " in *" $deplib "*) ;; * ) func_append new_inherited_linker_flags " $deplib" ;; esac fi fi continue ;; -L*) case $linkmode in lib) deplibs="$deplib $deplibs" test "$pass" = conv && continue newdependency_libs="$deplib $newdependency_libs" func_stripname '-L' '' "$deplib" func_resolve_sysroot "$func_stripname_result" func_append newlib_search_path " $func_resolve_sysroot_result" ;; prog) if test "$pass" = conv; then deplibs="$deplib $deplibs" continue fi if test "$pass" = scan; then deplibs="$deplib $deplibs" else compile_deplibs="$deplib $compile_deplibs" finalize_deplibs="$deplib $finalize_deplibs" fi func_stripname '-L' '' "$deplib" func_resolve_sysroot "$func_stripname_result" func_append newlib_search_path " $func_resolve_sysroot_result" ;; *) func_warning "\`-L' is ignored for archives/objects" ;; esac # linkmode continue ;; # -L -R*) if test "$pass" = link; then func_stripname '-R' '' "$deplib" func_resolve_sysroot "$func_stripname_result" dir=$func_resolve_sysroot_result # Make sure the xrpath contains only unique directories. case "$xrpath " in *" $dir "*) ;; *) func_append xrpath " $dir" ;; esac fi deplibs="$deplib $deplibs" continue ;; *.la) func_resolve_sysroot "$deplib" lib=$func_resolve_sysroot_result ;; *.$libext) if test "$pass" = conv; then deplibs="$deplib $deplibs" continue fi case $linkmode in lib) # Linking convenience modules into shared libraries is allowed, # but linking other static libraries is non-portable. case " $dlpreconveniencelibs " in *" $deplib "*) ;; *) valid_a_lib=no case $deplibs_check_method in match_pattern*) set dummy $deplibs_check_method; shift match_pattern_regex=`expr "$deplibs_check_method" : "$1 \(.*\)"` if eval "\$ECHO \"$deplib\"" 2>/dev/null | $SED 10q \ | $EGREP "$match_pattern_regex" > /dev/null; then valid_a_lib=yes fi ;; pass_all) valid_a_lib=yes ;; esac if test "$valid_a_lib" != yes; then echo $ECHO "*** Warning: Trying to link with static lib archive $deplib." echo "*** I have the capability to make that library automatically link in when" echo "*** you link to this library. But I can only do this if you have a" echo "*** shared version of the library, which you do not appear to have" echo "*** because the file extensions .$libext of this argument makes me believe" echo "*** that it is just a static archive that I should not use here." else echo $ECHO "*** Warning: Linking the shared library $output against the" $ECHO "*** static library $deplib is not portable!" deplibs="$deplib $deplibs" fi ;; esac continue ;; prog) if test "$pass" != link; then deplibs="$deplib $deplibs" else compile_deplibs="$deplib $compile_deplibs" finalize_deplibs="$deplib $finalize_deplibs" fi continue ;; esac # linkmode ;; # *.$libext *.lo | *.$objext) if test "$pass" = conv; then deplibs="$deplib $deplibs" elif test "$linkmode" = prog; then if test "$pass" = dlpreopen || test "$dlopen_support" != yes || test "$build_libtool_libs" = no; then # If there is no dlopen support or we're linking statically, # we need to preload. func_append newdlprefiles " $deplib" compile_deplibs="$deplib $compile_deplibs" finalize_deplibs="$deplib $finalize_deplibs" else func_append newdlfiles " $deplib" fi fi continue ;; %DEPLIBS%) alldeplibs=yes continue ;; esac # case $deplib if test "$found" = yes || test -f "$lib"; then : else func_fatal_error "cannot find the library \`$lib' or unhandled argument \`$deplib'" fi # Check to see that this really is a libtool archive. func_lalib_unsafe_p "$lib" \ || func_fatal_error "\`$lib' is not a valid libtool archive" func_dirname "$lib" "" "." ladir="$func_dirname_result" dlname= dlopen= dlpreopen= libdir= library_names= old_library= inherited_linker_flags= # If the library was installed with an old release of libtool, # it will not redefine variables installed, or shouldnotlink installed=yes shouldnotlink=no avoidtemprpath= # Read the .la file func_source "$lib" # Convert "-framework foo" to "foo.ltframework" if test -n "$inherited_linker_flags"; then tmp_inherited_linker_flags=`$ECHO "$inherited_linker_flags" | $SED 's/-framework \([^ $]*\)/\1.ltframework/g'` for tmp_inherited_linker_flag in $tmp_inherited_linker_flags; do case " $new_inherited_linker_flags " in *" $tmp_inherited_linker_flag "*) ;; *) func_append new_inherited_linker_flags " $tmp_inherited_linker_flag";; esac done fi dependency_libs=`$ECHO " $dependency_libs" | $SED 's% \([^ $]*\).ltframework% -framework \1%g'` if test "$linkmode,$pass" = "lib,link" || test "$linkmode,$pass" = "prog,scan" || { test "$linkmode" != prog && test "$linkmode" != lib; }; then test -n "$dlopen" && func_append dlfiles " $dlopen" test -n "$dlpreopen" && func_append dlprefiles " $dlpreopen" fi if test "$pass" = conv; then # Only check for convenience libraries deplibs="$lib $deplibs" if test -z "$libdir"; then if test -z "$old_library"; then func_fatal_error "cannot find name of link library for \`$lib'" fi # It is a libtool convenience library, so add in its objects. func_append convenience " $ladir/$objdir/$old_library" func_append old_convenience " $ladir/$objdir/$old_library" tmp_libs= for deplib in $dependency_libs; do deplibs="$deplib $deplibs" if $opt_preserve_dup_deps ; then case "$tmp_libs " in *" $deplib "*) func_append specialdeplibs " $deplib" ;; esac fi func_append tmp_libs " $deplib" done elif test "$linkmode" != prog && test "$linkmode" != lib; then func_fatal_error "\`$lib' is not a convenience library" fi continue fi # $pass = conv # Get the name of the library we link against. linklib= if test -n "$old_library" && { test "$prefer_static_libs" = yes || test "$prefer_static_libs,$installed" = "built,no"; }; then linklib=$old_library else for l in $old_library $library_names; do linklib="$l" done fi if test -z "$linklib"; then func_fatal_error "cannot find name of link library for \`$lib'" fi # This library was specified with -dlopen. if test "$pass" = dlopen; then if test -z "$libdir"; then func_fatal_error "cannot -dlopen a convenience library: \`$lib'" fi if test -z "$dlname" || test "$dlopen_support" != yes || test "$build_libtool_libs" = no; then # If there is no dlname, no dlopen support or we're linking # statically, we need to preload. We also need to preload any # dependent libraries so libltdl's deplib preloader doesn't # bomb out in the load deplibs phase. func_append dlprefiles " $lib $dependency_libs" else func_append newdlfiles " $lib" fi continue fi # $pass = dlopen # We need an absolute path. case $ladir in [\\/]* | [A-Za-z]:[\\/]*) abs_ladir="$ladir" ;; *) abs_ladir=`cd "$ladir" && pwd` if test -z "$abs_ladir"; then func_warning "cannot determine absolute directory name of \`$ladir'" func_warning "passing it literally to the linker, although it might fail" abs_ladir="$ladir" fi ;; esac func_basename "$lib" laname="$func_basename_result" # Find the relevant object directory and library name. if test "X$installed" = Xyes; then if test ! -f "$lt_sysroot$libdir/$linklib" && test -f "$abs_ladir/$linklib"; then func_warning "library \`$lib' was moved." dir="$ladir" absdir="$abs_ladir" libdir="$abs_ladir" else dir="$lt_sysroot$libdir" absdir="$lt_sysroot$libdir" fi test "X$hardcode_automatic" = Xyes && avoidtemprpath=yes else if test ! -f "$ladir/$objdir/$linklib" && test -f "$abs_ladir/$linklib"; then dir="$ladir" absdir="$abs_ladir" # Remove this search path later func_append notinst_path " $abs_ladir" else dir="$ladir/$objdir" absdir="$abs_ladir/$objdir" # Remove this search path later func_append notinst_path " $abs_ladir" fi fi # $installed = yes func_stripname 'lib' '.la' "$laname" name=$func_stripname_result # This library was specified with -dlpreopen. if test "$pass" = dlpreopen; then if test -z "$libdir" && test "$linkmode" = prog; then func_fatal_error "only libraries may -dlpreopen a convenience library: \`$lib'" fi case "$host" in # special handling for platforms with PE-DLLs. *cygwin* | *mingw* | *cegcc* ) # Linker will automatically link against shared library if both # static and shared are present. Therefore, ensure we extract # symbols from the import library if a shared library is present # (otherwise, the dlopen module name will be incorrect). We do # this by putting the import library name into $newdlprefiles. # We recover the dlopen module name by 'saving' the la file # name in a special purpose variable, and (later) extracting the # dlname from the la file. if test -n "$dlname"; then func_tr_sh "$dir/$linklib" eval "libfile_$func_tr_sh_result=\$abs_ladir/\$laname" func_append newdlprefiles " $dir/$linklib" else func_append newdlprefiles " $dir/$old_library" # Keep a list of preopened convenience libraries to check # that they are being used correctly in the link pass. test -z "$libdir" && \ func_append dlpreconveniencelibs " $dir/$old_library" fi ;; * ) # Prefer using a static library (so that no silly _DYNAMIC symbols # are required to link). if test -n "$old_library"; then func_append newdlprefiles " $dir/$old_library" # Keep a list of preopened convenience libraries to check # that they are being used correctly in the link pass. test -z "$libdir" && \ func_append dlpreconveniencelibs " $dir/$old_library" # Otherwise, use the dlname, so that lt_dlopen finds it. elif test -n "$dlname"; then func_append newdlprefiles " $dir/$dlname" else func_append newdlprefiles " $dir/$linklib" fi ;; esac fi # $pass = dlpreopen if test -z "$libdir"; then # Link the convenience library if test "$linkmode" = lib; then deplibs="$dir/$old_library $deplibs" elif test "$linkmode,$pass" = "prog,link"; then compile_deplibs="$dir/$old_library $compile_deplibs" finalize_deplibs="$dir/$old_library $finalize_deplibs" else deplibs="$lib $deplibs" # used for prog,scan pass fi continue fi if test "$linkmode" = prog && test "$pass" != link; then func_append newlib_search_path " $ladir" deplibs="$lib $deplibs" linkalldeplibs=no if test "$link_all_deplibs" != no || test -z "$library_names" || test "$build_libtool_libs" = no; then linkalldeplibs=yes fi tmp_libs= for deplib in $dependency_libs; do case $deplib in -L*) func_stripname '-L' '' "$deplib" func_resolve_sysroot "$func_stripname_result" func_append newlib_search_path " $func_resolve_sysroot_result" ;; esac # Need to link against all dependency_libs? if test "$linkalldeplibs" = yes; then deplibs="$deplib $deplibs" else # Need to hardcode shared library paths # or/and link against static libraries newdependency_libs="$deplib $newdependency_libs" fi if $opt_preserve_dup_deps ; then case "$tmp_libs " in *" $deplib "*) func_append specialdeplibs " $deplib" ;; esac fi func_append tmp_libs " $deplib" done # for deplib continue fi # $linkmode = prog... if test "$linkmode,$pass" = "prog,link"; then if test -n "$library_names" && { { test "$prefer_static_libs" = no || test "$prefer_static_libs,$installed" = "built,yes"; } || test -z "$old_library"; }; then # We need to hardcode the library path if test -n "$shlibpath_var" && test -z "$avoidtemprpath" ; then # Make sure the rpath contains only unique directories. case "$temp_rpath:" in *"$absdir:"*) ;; *) func_append temp_rpath "$absdir:" ;; esac fi # Hardcode the library path. # Skip directories that are in the system default run-time # search path. case " $sys_lib_dlsearch_path " in *" $absdir "*) ;; *) case "$compile_rpath " in *" $absdir "*) ;; *) func_append compile_rpath " $absdir" ;; esac ;; esac case " $sys_lib_dlsearch_path " in *" $libdir "*) ;; *) case "$finalize_rpath " in *" $libdir "*) ;; *) func_append finalize_rpath " $libdir" ;; esac ;; esac fi # $linkmode,$pass = prog,link... if test "$alldeplibs" = yes && { test "$deplibs_check_method" = pass_all || { test "$build_libtool_libs" = yes && test -n "$library_names"; }; }; then # We only need to search for static libraries continue fi fi link_static=no # Whether the deplib will be linked statically use_static_libs=$prefer_static_libs if test "$use_static_libs" = built && test "$installed" = yes; then use_static_libs=no fi if test -n "$library_names" && { test "$use_static_libs" = no || test -z "$old_library"; }; then case $host in *cygwin* | *mingw* | *cegcc*) # No point in relinking DLLs because paths are not encoded func_append notinst_deplibs " $lib" need_relink=no ;; *) if test "$installed" = no; then func_append notinst_deplibs " $lib" need_relink=yes fi ;; esac # This is a shared library # Warn about portability, can't link against -module's on some # systems (darwin). Don't bleat about dlopened modules though! dlopenmodule="" for dlpremoduletest in $dlprefiles; do if test "X$dlpremoduletest" = "X$lib"; then dlopenmodule="$dlpremoduletest" break fi done if test -z "$dlopenmodule" && test "$shouldnotlink" = yes && test "$pass" = link; then echo if test "$linkmode" = prog; then $ECHO "*** Warning: Linking the executable $output against the loadable module" else $ECHO "*** Warning: Linking the shared library $output against the loadable module" fi $ECHO "*** $linklib is not portable!" fi if test "$linkmode" = lib && test "$hardcode_into_libs" = yes; then # Hardcode the library path. # Skip directories that are in the system default run-time # search path. case " $sys_lib_dlsearch_path " in *" $absdir "*) ;; *) case "$compile_rpath " in *" $absdir "*) ;; *) func_append compile_rpath " $absdir" ;; esac ;; esac case " $sys_lib_dlsearch_path " in *" $libdir "*) ;; *) case "$finalize_rpath " in *" $libdir "*) ;; *) func_append finalize_rpath " $libdir" ;; esac ;; esac fi if test -n "$old_archive_from_expsyms_cmds"; then # figure out the soname set dummy $library_names shift realname="$1" shift libname=`eval "\\$ECHO \"$libname_spec\""` # use dlname if we got it. it's perfectly good, no? if test -n "$dlname"; then soname="$dlname" elif test -n "$soname_spec"; then # bleh windows case $host in *cygwin* | mingw* | *cegcc*) func_arith $current - $age major=$func_arith_result versuffix="-$major" ;; esac eval soname=\"$soname_spec\" else soname="$realname" fi # Make a new name for the extract_expsyms_cmds to use soroot="$soname" func_basename "$soroot" soname="$func_basename_result" func_stripname 'lib' '.dll' "$soname" newlib=libimp-$func_stripname_result.a # If the library has no export list, then create one now if test -f "$output_objdir/$soname-def"; then : else func_verbose "extracting exported symbol list from \`$soname'" func_execute_cmds "$extract_expsyms_cmds" 'exit $?' fi # Create $newlib if test -f "$output_objdir/$newlib"; then :; else func_verbose "generating import library for \`$soname'" func_execute_cmds "$old_archive_from_expsyms_cmds" 'exit $?' fi # make sure the library variables are pointing to the new library dir=$output_objdir linklib=$newlib fi # test -n "$old_archive_from_expsyms_cmds" if test "$linkmode" = prog || test "$opt_mode" != relink; then add_shlibpath= add_dir= add= lib_linked=yes case $hardcode_action in immediate | unsupported) if test "$hardcode_direct" = no; then add="$dir/$linklib" case $host in *-*-sco3.2v5.0.[024]*) add_dir="-L$dir" ;; *-*-sysv4*uw2*) add_dir="-L$dir" ;; *-*-sysv5OpenUNIX* | *-*-sysv5UnixWare7.[01].[10]* | \ *-*-unixware7*) add_dir="-L$dir" ;; *-*-darwin* ) # if the lib is a (non-dlopened) module then we can not # link against it, someone is ignoring the earlier warnings if /usr/bin/file -L $add 2> /dev/null | $GREP ": [^:]* bundle" >/dev/null ; then if test "X$dlopenmodule" != "X$lib"; then $ECHO "*** Warning: lib $linklib is a module, not a shared library" if test -z "$old_library" ; then echo echo "*** And there doesn't seem to be a static archive available" echo "*** The link will probably fail, sorry" else add="$dir/$old_library" fi elif test -n "$old_library"; then add="$dir/$old_library" fi fi esac elif test "$hardcode_minus_L" = no; then case $host in *-*-sunos*) add_shlibpath="$dir" ;; esac add_dir="-L$dir" add="-l$name" elif test "$hardcode_shlibpath_var" = no; then add_shlibpath="$dir" add="-l$name" else lib_linked=no fi ;; relink) if test "$hardcode_direct" = yes && test "$hardcode_direct_absolute" = no; then add="$dir/$linklib" elif test "$hardcode_minus_L" = yes; then add_dir="-L$absdir" # Try looking first in the location we're being installed to. if test -n "$inst_prefix_dir"; then case $libdir in [\\/]*) func_append add_dir " -L$inst_prefix_dir$libdir" ;; esac fi add="-l$name" elif test "$hardcode_shlibpath_var" = yes; then add_shlibpath="$dir" add="-l$name" else lib_linked=no fi ;; *) lib_linked=no ;; esac if test "$lib_linked" != yes; then func_fatal_configuration "unsupported hardcode properties" fi if test -n "$add_shlibpath"; then case :$compile_shlibpath: in *":$add_shlibpath:"*) ;; *) func_append compile_shlibpath "$add_shlibpath:" ;; esac fi if test "$linkmode" = prog; then test -n "$add_dir" && compile_deplibs="$add_dir $compile_deplibs" test -n "$add" && compile_deplibs="$add $compile_deplibs" else test -n "$add_dir" && deplibs="$add_dir $deplibs" test -n "$add" && deplibs="$add $deplibs" if test "$hardcode_direct" != yes && test "$hardcode_minus_L" != yes && test "$hardcode_shlibpath_var" = yes; then case :$finalize_shlibpath: in *":$libdir:"*) ;; *) func_append finalize_shlibpath "$libdir:" ;; esac fi fi fi if test "$linkmode" = prog || test "$opt_mode" = relink; then add_shlibpath= add_dir= add= # Finalize command for both is simple: just hardcode it. if test "$hardcode_direct" = yes && test "$hardcode_direct_absolute" = no; then add="$libdir/$linklib" elif test "$hardcode_minus_L" = yes; then add_dir="-L$libdir" add="-l$name" elif test "$hardcode_shlibpath_var" = yes; then case :$finalize_shlibpath: in *":$libdir:"*) ;; *) func_append finalize_shlibpath "$libdir:" ;; esac add="-l$name" elif test "$hardcode_automatic" = yes; then if test -n "$inst_prefix_dir" && test -f "$inst_prefix_dir$libdir/$linklib" ; then add="$inst_prefix_dir$libdir/$linklib" else add="$libdir/$linklib" fi else # We cannot seem to hardcode it, guess we'll fake it. add_dir="-L$libdir" # Try looking first in the location we're being installed to. if test -n "$inst_prefix_dir"; then case $libdir in [\\/]*) func_append add_dir " -L$inst_prefix_dir$libdir" ;; esac fi add="-l$name" fi if test "$linkmode" = prog; then test -n "$add_dir" && finalize_deplibs="$add_dir $finalize_deplibs" test -n "$add" && finalize_deplibs="$add $finalize_deplibs" else test -n "$add_dir" && deplibs="$add_dir $deplibs" test -n "$add" && deplibs="$add $deplibs" fi fi elif test "$linkmode" = prog; then # Here we assume that one of hardcode_direct or hardcode_minus_L # is not unsupported. This is valid on all known static and # shared platforms. if test "$hardcode_direct" != unsupported; then test -n "$old_library" && linklib="$old_library" compile_deplibs="$dir/$linklib $compile_deplibs" finalize_deplibs="$dir/$linklib $finalize_deplibs" else compile_deplibs="-l$name -L$dir $compile_deplibs" finalize_deplibs="-l$name -L$dir $finalize_deplibs" fi elif test "$build_libtool_libs" = yes; then # Not a shared library if test "$deplibs_check_method" != pass_all; then # We're trying link a shared library against a static one # but the system doesn't support it. # Just print a warning and add the library to dependency_libs so # that the program can be linked against the static library. echo $ECHO "*** Warning: This system can not link to static lib archive $lib." echo "*** I have the capability to make that library automatically link in when" echo "*** you link to this library. But I can only do this if you have a" echo "*** shared version of the library, which you do not appear to have." if test "$module" = yes; then echo "*** But as you try to build a module library, libtool will still create " echo "*** a static module, that should work as long as the dlopening application" echo "*** is linked with the -dlopen flag to resolve symbols at runtime." if test -z "$global_symbol_pipe"; then echo echo "*** However, this would only work if libtool was able to extract symbol" echo "*** lists from a program, using \`nm' or equivalent, but libtool could" echo "*** not find such a program. So, this module is probably useless." echo "*** \`nm' from GNU binutils and a full rebuild may help." fi if test "$build_old_libs" = no; then build_libtool_libs=module build_old_libs=yes else build_libtool_libs=no fi fi else deplibs="$dir/$old_library $deplibs" link_static=yes fi fi # link shared/static library? if test "$linkmode" = lib; then if test -n "$dependency_libs" && { test "$hardcode_into_libs" != yes || test "$build_old_libs" = yes || test "$link_static" = yes; }; then # Extract -R from dependency_libs temp_deplibs= for libdir in $dependency_libs; do case $libdir in -R*) func_stripname '-R' '' "$libdir" temp_xrpath=$func_stripname_result case " $xrpath " in *" $temp_xrpath "*) ;; *) func_append xrpath " $temp_xrpath";; esac;; *) func_append temp_deplibs " $libdir";; esac done dependency_libs="$temp_deplibs" fi func_append newlib_search_path " $absdir" # Link against this library test "$link_static" = no && newdependency_libs="$abs_ladir/$laname $newdependency_libs" # ... and its dependency_libs tmp_libs= for deplib in $dependency_libs; do newdependency_libs="$deplib $newdependency_libs" case $deplib in -L*) func_stripname '-L' '' "$deplib" func_resolve_sysroot "$func_stripname_result";; *) func_resolve_sysroot "$deplib" ;; esac if $opt_preserve_dup_deps ; then case "$tmp_libs " in *" $func_resolve_sysroot_result "*) func_append specialdeplibs " $func_resolve_sysroot_result" ;; esac fi func_append tmp_libs " $func_resolve_sysroot_result" done if test "$link_all_deplibs" != no; then # Add the search paths of all dependency libraries for deplib in $dependency_libs; do path= case $deplib in -L*) path="$deplib" ;; *.la) func_resolve_sysroot "$deplib" deplib=$func_resolve_sysroot_result func_dirname "$deplib" "" "." dir=$func_dirname_result # We need an absolute path. case $dir in [\\/]* | [A-Za-z]:[\\/]*) absdir="$dir" ;; *) absdir=`cd "$dir" && pwd` if test -z "$absdir"; then func_warning "cannot determine absolute directory name of \`$dir'" absdir="$dir" fi ;; esac if $GREP "^installed=no" $deplib > /dev/null; then case $host in *-*-darwin*) depdepl= eval deplibrary_names=`${SED} -n -e 's/^library_names=\(.*\)$/\1/p' $deplib` if test -n "$deplibrary_names" ; then for tmp in $deplibrary_names ; do depdepl=$tmp done if test -f "$absdir/$objdir/$depdepl" ; then depdepl="$absdir/$objdir/$depdepl" darwin_install_name=`${OTOOL} -L $depdepl | awk '{if (NR == 2) {print $1;exit}}'` if test -z "$darwin_install_name"; then darwin_install_name=`${OTOOL64} -L $depdepl | awk '{if (NR == 2) {print $1;exit}}'` fi func_append compiler_flags " ${wl}-dylib_file ${wl}${darwin_install_name}:${depdepl}" func_append linker_flags " -dylib_file ${darwin_install_name}:${depdepl}" path= fi fi ;; *) path="-L$absdir/$objdir" ;; esac else eval libdir=`${SED} -n -e 's/^libdir=\(.*\)$/\1/p' $deplib` test -z "$libdir" && \ func_fatal_error "\`$deplib' is not a valid libtool archive" test "$absdir" != "$libdir" && \ func_warning "\`$deplib' seems to be moved" path="-L$absdir" fi ;; esac case " $deplibs " in *" $path "*) ;; *) deplibs="$path $deplibs" ;; esac done fi # link_all_deplibs != no fi # linkmode = lib done # for deplib in $libs if test "$pass" = link; then if test "$linkmode" = "prog"; then compile_deplibs="$new_inherited_linker_flags $compile_deplibs" finalize_deplibs="$new_inherited_linker_flags $finalize_deplibs" else compiler_flags="$compiler_flags "`$ECHO " $new_inherited_linker_flags" | $SED 's% \([^ $]*\).ltframework% -framework \1%g'` fi fi dependency_libs="$newdependency_libs" if test "$pass" = dlpreopen; then # Link the dlpreopened libraries before other libraries for deplib in $save_deplibs; do deplibs="$deplib $deplibs" done fi if test "$pass" != dlopen; then if test "$pass" != conv; then # Make sure lib_search_path contains only unique directories. lib_search_path= for dir in $newlib_search_path; do case "$lib_search_path " in *" $dir "*) ;; *) func_append lib_search_path " $dir" ;; esac done newlib_search_path= fi if test "$linkmode,$pass" != "prog,link"; then vars="deplibs" else vars="compile_deplibs finalize_deplibs" fi for var in $vars dependency_libs; do # Add libraries to $var in reverse order eval tmp_libs=\"\$$var\" new_libs= for deplib in $tmp_libs; do # FIXME: Pedantically, this is the right thing to do, so # that some nasty dependency loop isn't accidentally # broken: #new_libs="$deplib $new_libs" # Pragmatically, this seems to cause very few problems in # practice: case $deplib in -L*) new_libs="$deplib $new_libs" ;; -R*) ;; *) # And here is the reason: when a library appears more # than once as an explicit dependence of a library, or # is implicitly linked in more than once by the # compiler, it is considered special, and multiple # occurrences thereof are not removed. Compare this # with having the same library being listed as a # dependency of multiple other libraries: in this case, # we know (pedantically, we assume) the library does not # need to be listed more than once, so we keep only the # last copy. This is not always right, but it is rare # enough that we require users that really mean to play # such unportable linking tricks to link the library # using -Wl,-lname, so that libtool does not consider it # for duplicate removal. case " $specialdeplibs " in *" $deplib "*) new_libs="$deplib $new_libs" ;; *) case " $new_libs " in *" $deplib "*) ;; *) new_libs="$deplib $new_libs" ;; esac ;; esac ;; esac done tmp_libs= for deplib in $new_libs; do case $deplib in -L*) case " $tmp_libs " in *" $deplib "*) ;; *) func_append tmp_libs " $deplib" ;; esac ;; *) func_append tmp_libs " $deplib" ;; esac done eval $var=\"$tmp_libs\" done # for var fi # Last step: remove runtime libs from dependency_libs # (they stay in deplibs) tmp_libs= for i in $dependency_libs ; do case " $predeps $postdeps $compiler_lib_search_path " in *" $i "*) i="" ;; esac if test -n "$i" ; then func_append tmp_libs " $i" fi done dependency_libs=$tmp_libs done # for pass if test "$linkmode" = prog; then dlfiles="$newdlfiles" fi if test "$linkmode" = prog || test "$linkmode" = lib; then dlprefiles="$newdlprefiles" fi case $linkmode in oldlib) if test -n "$dlfiles$dlprefiles" || test "$dlself" != no; then func_warning "\`-dlopen' is ignored for archives" fi case " $deplibs" in *\ -l* | *\ -L*) func_warning "\`-l' and \`-L' are ignored for archives" ;; esac test -n "$rpath" && \ func_warning "\`-rpath' is ignored for archives" test -n "$xrpath" && \ func_warning "\`-R' is ignored for archives" test -n "$vinfo" && \ func_warning "\`-version-info/-version-number' is ignored for archives" test -n "$release" && \ func_warning "\`-release' is ignored for archives" test -n "$export_symbols$export_symbols_regex" && \ func_warning "\`-export-symbols' is ignored for archives" # Now set the variables for building old libraries. build_libtool_libs=no oldlibs="$output" func_append objs "$old_deplibs" ;; lib) # Make sure we only generate libraries of the form `libNAME.la'. case $outputname in lib*) func_stripname 'lib' '.la' "$outputname" name=$func_stripname_result eval shared_ext=\"$shrext_cmds\" eval libname=\"$libname_spec\" ;; *) test "$module" = no && \ func_fatal_help "libtool library \`$output' must begin with \`lib'" if test "$need_lib_prefix" != no; then # Add the "lib" prefix for modules if required func_stripname '' '.la' "$outputname" name=$func_stripname_result eval shared_ext=\"$shrext_cmds\" eval libname=\"$libname_spec\" else func_stripname '' '.la' "$outputname" libname=$func_stripname_result fi ;; esac if test -n "$objs"; then if test "$deplibs_check_method" != pass_all; then func_fatal_error "cannot build libtool library \`$output' from non-libtool objects on this host:$objs" else echo $ECHO "*** Warning: Linking the shared library $output against the non-libtool" $ECHO "*** objects $objs is not portable!" func_append libobjs " $objs" fi fi test "$dlself" != no && \ func_warning "\`-dlopen self' is ignored for libtool libraries" set dummy $rpath shift test "$#" -gt 1 && \ func_warning "ignoring multiple \`-rpath's for a libtool library" install_libdir="$1" oldlibs= if test -z "$rpath"; then if test "$build_libtool_libs" = yes; then # Building a libtool convenience library. # Some compilers have problems with a `.al' extension so # convenience libraries should have the same extension an # archive normally would. oldlibs="$output_objdir/$libname.$libext $oldlibs" build_libtool_libs=convenience build_old_libs=yes fi test -n "$vinfo" && \ func_warning "\`-version-info/-version-number' is ignored for convenience libraries" test -n "$release" && \ func_warning "\`-release' is ignored for convenience libraries" else # Parse the version information argument. save_ifs="$IFS"; IFS=':' set dummy $vinfo 0 0 0 shift IFS="$save_ifs" test -n "$7" && \ func_fatal_help "too many parameters to \`-version-info'" # convert absolute version numbers to libtool ages # this retains compatibility with .la files and attempts # to make the code below a bit more comprehensible case $vinfo_number in yes) number_major="$1" number_minor="$2" number_revision="$3" # # There are really only two kinds -- those that # use the current revision as the major version # and those that subtract age and use age as # a minor version. But, then there is irix # which has an extra 1 added just for fun # case $version_type in # correct linux to gnu/linux during the next big refactor darwin|linux|osf|windows|none) func_arith $number_major + $number_minor current=$func_arith_result age="$number_minor" revision="$number_revision" ;; freebsd-aout|freebsd-elf|qnx|sunos) current="$number_major" revision="$number_minor" age="0" ;; irix|nonstopux) func_arith $number_major + $number_minor current=$func_arith_result age="$number_minor" revision="$number_minor" lt_irix_increment=no ;; *) func_fatal_configuration "$modename: unknown library version type \`$version_type'" ;; esac ;; no) current="$1" revision="$2" age="$3" ;; esac # Check that each of the things are valid numbers. case $current in 0|[1-9]|[1-9][0-9]|[1-9][0-9][0-9]|[1-9][0-9][0-9][0-9]|[1-9][0-9][0-9][0-9][0-9]) ;; *) func_error "CURRENT \`$current' must be a nonnegative integer" func_fatal_error "\`$vinfo' is not valid version information" ;; esac case $revision in 0|[1-9]|[1-9][0-9]|[1-9][0-9][0-9]|[1-9][0-9][0-9][0-9]|[1-9][0-9][0-9][0-9][0-9]) ;; *) func_error "REVISION \`$revision' must be a nonnegative integer" func_fatal_error "\`$vinfo' is not valid version information" ;; esac case $age in 0|[1-9]|[1-9][0-9]|[1-9][0-9][0-9]|[1-9][0-9][0-9][0-9]|[1-9][0-9][0-9][0-9][0-9]) ;; *) func_error "AGE \`$age' must be a nonnegative integer" func_fatal_error "\`$vinfo' is not valid version information" ;; esac if test "$age" -gt "$current"; then func_error "AGE \`$age' is greater than the current interface number \`$current'" func_fatal_error "\`$vinfo' is not valid version information" fi # Calculate the version variables. major= versuffix= verstring= case $version_type in none) ;; darwin) # Like Linux, but with the current version available in # verstring for coding it into the library header func_arith $current - $age major=.$func_arith_result versuffix="$major.$age.$revision" # Darwin ld doesn't like 0 for these options... func_arith $current + 1 minor_current=$func_arith_result xlcverstring="${wl}-compatibility_version ${wl}$minor_current ${wl}-current_version ${wl}$minor_current.$revision" verstring="-compatibility_version $minor_current -current_version $minor_current.$revision" ;; freebsd-aout) major=".$current" versuffix=".$current.$revision"; ;; freebsd-elf) major=".$current" versuffix=".$current" ;; irix | nonstopux) if test "X$lt_irix_increment" = "Xno"; then func_arith $current - $age else func_arith $current - $age + 1 fi major=$func_arith_result case $version_type in nonstopux) verstring_prefix=nonstopux ;; *) verstring_prefix=sgi ;; esac verstring="$verstring_prefix$major.$revision" # Add in all the interfaces that we are compatible with. loop=$revision while test "$loop" -ne 0; do func_arith $revision - $loop iface=$func_arith_result func_arith $loop - 1 loop=$func_arith_result verstring="$verstring_prefix$major.$iface:$verstring" done # Before this point, $major must not contain `.'. major=.$major versuffix="$major.$revision" ;; linux) # correct to gnu/linux during the next big refactor func_arith $current - $age major=.$func_arith_result versuffix="$major.$age.$revision" ;; osf) func_arith $current - $age major=.$func_arith_result versuffix=".$current.$age.$revision" verstring="$current.$age.$revision" # Add in all the interfaces that we are compatible with. loop=$age while test "$loop" -ne 0; do func_arith $current - $loop iface=$func_arith_result func_arith $loop - 1 loop=$func_arith_result verstring="$verstring:${iface}.0" done # Make executables depend on our current version. func_append verstring ":${current}.0" ;; qnx) major=".$current" versuffix=".$current" ;; sunos) major=".$current" versuffix=".$current.$revision" ;; windows) # Use '-' rather than '.', since we only want one # extension on DOS 8.3 filesystems. func_arith $current - $age major=$func_arith_result versuffix="-$major" ;; *) func_fatal_configuration "unknown library version type \`$version_type'" ;; esac # Clear the version info if we defaulted, and they specified a release. if test -z "$vinfo" && test -n "$release"; then major= case $version_type in darwin) # we can't check for "0.0" in archive_cmds due to quoting # problems, so we reset it completely verstring= ;; *) verstring="0.0" ;; esac if test "$need_version" = no; then versuffix= else versuffix=".0.0" fi fi # Remove version info from name if versioning should be avoided if test "$avoid_version" = yes && test "$need_version" = no; then major= versuffix= verstring="" fi # Check to see if the archive will have undefined symbols. if test "$allow_undefined" = yes; then if test "$allow_undefined_flag" = unsupported; then func_warning "undefined symbols not allowed in $host shared libraries" build_libtool_libs=no build_old_libs=yes fi else # Don't allow undefined symbols. allow_undefined_flag="$no_undefined_flag" fi fi func_generate_dlsyms "$libname" "$libname" "yes" func_append libobjs " $symfileobj" test "X$libobjs" = "X " && libobjs= if test "$opt_mode" != relink; then # Remove our outputs, but don't remove object files since they # may have been created when compiling PIC objects. removelist= tempremovelist=`$ECHO "$output_objdir/*"` for p in $tempremovelist; do case $p in *.$objext | *.gcno) ;; $output_objdir/$outputname | $output_objdir/$libname.* | $output_objdir/${libname}${release}.*) if test "X$precious_files_regex" != "X"; then if $ECHO "$p" | $EGREP -e "$precious_files_regex" >/dev/null 2>&1 then continue fi fi func_append removelist " $p" ;; *) ;; esac done test -n "$removelist" && \ func_show_eval "${RM}r \$removelist" fi # Now set the variables for building old libraries. if test "$build_old_libs" = yes && test "$build_libtool_libs" != convenience ; then func_append oldlibs " $output_objdir/$libname.$libext" # Transform .lo files to .o files. oldobjs="$objs "`$ECHO "$libobjs" | $SP2NL | $SED "/\.${libext}$/d; $lo2o" | $NL2SP` fi # Eliminate all temporary directories. #for path in $notinst_path; do # lib_search_path=`$ECHO "$lib_search_path " | $SED "s% $path % %g"` # deplibs=`$ECHO "$deplibs " | $SED "s% -L$path % %g"` # dependency_libs=`$ECHO "$dependency_libs " | $SED "s% -L$path % %g"` #done if test -n "$xrpath"; then # If the user specified any rpath flags, then add them. temp_xrpath= for libdir in $xrpath; do func_replace_sysroot "$libdir" func_append temp_xrpath " -R$func_replace_sysroot_result" case "$finalize_rpath " in *" $libdir "*) ;; *) func_append finalize_rpath " $libdir" ;; esac done if test "$hardcode_into_libs" != yes || test "$build_old_libs" = yes; then dependency_libs="$temp_xrpath $dependency_libs" fi fi # Make sure dlfiles contains only unique files that won't be dlpreopened old_dlfiles="$dlfiles" dlfiles= for lib in $old_dlfiles; do case " $dlprefiles $dlfiles " in *" $lib "*) ;; *) func_append dlfiles " $lib" ;; esac done # Make sure dlprefiles contains only unique files old_dlprefiles="$dlprefiles" dlprefiles= for lib in $old_dlprefiles; do case "$dlprefiles " in *" $lib "*) ;; *) func_append dlprefiles " $lib" ;; esac done if test "$build_libtool_libs" = yes; then if test -n "$rpath"; then case $host in *-*-cygwin* | *-*-mingw* | *-*-pw32* | *-*-os2* | *-*-beos* | *-cegcc* | *-*-haiku*) # these systems don't actually have a c library (as such)! ;; *-*-rhapsody* | *-*-darwin1.[012]) # Rhapsody C library is in the System framework func_append deplibs " System.ltframework" ;; *-*-netbsd*) # Don't link with libc until the a.out ld.so is fixed. ;; *-*-openbsd* | *-*-freebsd* | *-*-dragonfly*) # Do not include libc due to us having libc/libc_r. ;; *-*-sco3.2v5* | *-*-sco5v6*) # Causes problems with __ctype ;; *-*-sysv4.2uw2* | *-*-sysv5* | *-*-unixware* | *-*-OpenUNIX*) # Compiler inserts libc in the correct place for threads to work ;; *) # Add libc to deplibs on all other systems if necessary. if test "$build_libtool_need_lc" = "yes"; then func_append deplibs " -lc" fi ;; esac fi # Transform deplibs into only deplibs that can be linked in shared. name_save=$name libname_save=$libname release_save=$release versuffix_save=$versuffix major_save=$major # I'm not sure if I'm treating the release correctly. I think # release should show up in the -l (ie -lgmp5) so we don't want to # add it in twice. Is that correct? release="" versuffix="" major="" newdeplibs= droppeddeps=no case $deplibs_check_method in pass_all) # Don't check for shared/static. Everything works. # This might be a little naive. We might want to check # whether the library exists or not. But this is on # osf3 & osf4 and I'm not really sure... Just # implementing what was already the behavior. newdeplibs=$deplibs ;; test_compile) # This code stresses the "libraries are programs" paradigm to its # limits. Maybe even breaks it. We compile a program, linking it # against the deplibs as a proxy for the library. Then we can check # whether they linked in statically or dynamically with ldd. $opt_dry_run || $RM conftest.c cat > conftest.c </dev/null` $nocaseglob else potential_libs=`ls $i/$libnameglob[.-]* 2>/dev/null` fi for potent_lib in $potential_libs; do # Follow soft links. if ls -lLd "$potent_lib" 2>/dev/null | $GREP " -> " >/dev/null; then continue fi # The statement above tries to avoid entering an # endless loop below, in case of cyclic links. # We might still enter an endless loop, since a link # loop can be closed while we follow links, # but so what? potlib="$potent_lib" while test -h "$potlib" 2>/dev/null; do potliblink=`ls -ld $potlib | ${SED} 's/.* -> //'` case $potliblink in [\\/]* | [A-Za-z]:[\\/]*) potlib="$potliblink";; *) potlib=`$ECHO "$potlib" | $SED 's,[^/]*$,,'`"$potliblink";; esac done if eval $file_magic_cmd \"\$potlib\" 2>/dev/null | $SED -e 10q | $EGREP "$file_magic_regex" > /dev/null; then func_append newdeplibs " $a_deplib" a_deplib="" break 2 fi done done fi if test -n "$a_deplib" ; then droppeddeps=yes echo $ECHO "*** Warning: linker path does not have real file for library $a_deplib." echo "*** I have the capability to make that library automatically link in when" echo "*** you link to this library. But I can only do this if you have a" echo "*** shared version of the library, which you do not appear to have" echo "*** because I did check the linker path looking for a file starting" if test -z "$potlib" ; then $ECHO "*** with $libname but no candidates were found. (...for file magic test)" else $ECHO "*** with $libname and none of the candidates passed a file format test" $ECHO "*** using a file magic. Last file checked: $potlib" fi fi ;; *) # Add a -L argument. func_append newdeplibs " $a_deplib" ;; esac done # Gone through all deplibs. ;; match_pattern*) set dummy $deplibs_check_method; shift match_pattern_regex=`expr "$deplibs_check_method" : "$1 \(.*\)"` for a_deplib in $deplibs; do case $a_deplib in -l*) func_stripname -l '' "$a_deplib" name=$func_stripname_result if test "X$allow_libtool_libs_with_static_runtimes" = "Xyes" ; then case " $predeps $postdeps " in *" $a_deplib "*) func_append newdeplibs " $a_deplib" a_deplib="" ;; esac fi if test -n "$a_deplib" ; then libname=`eval "\\$ECHO \"$libname_spec\""` for i in $lib_search_path $sys_lib_search_path $shlib_search_path; do potential_libs=`ls $i/$libname[.-]* 2>/dev/null` for potent_lib in $potential_libs; do potlib="$potent_lib" # see symlink-check above in file_magic test if eval "\$ECHO \"$potent_lib\"" 2>/dev/null | $SED 10q | \ $EGREP "$match_pattern_regex" > /dev/null; then func_append newdeplibs " $a_deplib" a_deplib="" break 2 fi done done fi if test -n "$a_deplib" ; then droppeddeps=yes echo $ECHO "*** Warning: linker path does not have real file for library $a_deplib." echo "*** I have the capability to make that library automatically link in when" echo "*** you link to this library. But I can only do this if you have a" echo "*** shared version of the library, which you do not appear to have" echo "*** because I did check the linker path looking for a file starting" if test -z "$potlib" ; then $ECHO "*** with $libname but no candidates were found. (...for regex pattern test)" else $ECHO "*** with $libname and none of the candidates passed a file format test" $ECHO "*** using a regex pattern. Last file checked: $potlib" fi fi ;; *) # Add a -L argument. func_append newdeplibs " $a_deplib" ;; esac done # Gone through all deplibs. ;; none | unknown | *) newdeplibs="" tmp_deplibs=`$ECHO " $deplibs" | $SED 's/ -lc$//; s/ -[LR][^ ]*//g'` if test "X$allow_libtool_libs_with_static_runtimes" = "Xyes" ; then for i in $predeps $postdeps ; do # can't use Xsed below, because $i might contain '/' tmp_deplibs=`$ECHO " $tmp_deplibs" | $SED "s,$i,,"` done fi case $tmp_deplibs in *[!\ \ ]*) echo if test "X$deplibs_check_method" = "Xnone"; then echo "*** Warning: inter-library dependencies are not supported in this platform." else echo "*** Warning: inter-library dependencies are not known to be supported." fi echo "*** All declared inter-library dependencies are being dropped." droppeddeps=yes ;; esac ;; esac versuffix=$versuffix_save major=$major_save release=$release_save libname=$libname_save name=$name_save case $host in *-*-rhapsody* | *-*-darwin1.[012]) # On Rhapsody replace the C library with the System framework newdeplibs=`$ECHO " $newdeplibs" | $SED 's/ -lc / System.ltframework /'` ;; esac if test "$droppeddeps" = yes; then if test "$module" = yes; then echo echo "*** Warning: libtool could not satisfy all declared inter-library" $ECHO "*** dependencies of module $libname. Therefore, libtool will create" echo "*** a static module, that should work as long as the dlopening" echo "*** application is linked with the -dlopen flag." if test -z "$global_symbol_pipe"; then echo echo "*** However, this would only work if libtool was able to extract symbol" echo "*** lists from a program, using \`nm' or equivalent, but libtool could" echo "*** not find such a program. So, this module is probably useless." echo "*** \`nm' from GNU binutils and a full rebuild may help." fi if test "$build_old_libs" = no; then oldlibs="$output_objdir/$libname.$libext" build_libtool_libs=module build_old_libs=yes else build_libtool_libs=no fi else echo "*** The inter-library dependencies that have been dropped here will be" echo "*** automatically added whenever a program is linked with this library" echo "*** or is declared to -dlopen it." if test "$allow_undefined" = no; then echo echo "*** Since this library must not contain undefined symbols," echo "*** because either the platform does not support them or" echo "*** it was explicitly requested with -no-undefined," echo "*** libtool will only create a static version of it." if test "$build_old_libs" = no; then oldlibs="$output_objdir/$libname.$libext" build_libtool_libs=module build_old_libs=yes else build_libtool_libs=no fi fi fi fi # Done checking deplibs! deplibs=$newdeplibs fi # Time to change all our "foo.ltframework" stuff back to "-framework foo" case $host in *-*-darwin*) newdeplibs=`$ECHO " $newdeplibs" | $SED 's% \([^ $]*\).ltframework% -framework \1%g'` new_inherited_linker_flags=`$ECHO " $new_inherited_linker_flags" | $SED 's% \([^ $]*\).ltframework% -framework \1%g'` deplibs=`$ECHO " $deplibs" | $SED 's% \([^ $]*\).ltframework% -framework \1%g'` ;; esac # move library search paths that coincide with paths to not yet # installed libraries to the beginning of the library search list new_libs= for path in $notinst_path; do case " $new_libs " in *" -L$path/$objdir "*) ;; *) case " $deplibs " in *" -L$path/$objdir "*) func_append new_libs " -L$path/$objdir" ;; esac ;; esac done for deplib in $deplibs; do case $deplib in -L*) case " $new_libs " in *" $deplib "*) ;; *) func_append new_libs " $deplib" ;; esac ;; *) func_append new_libs " $deplib" ;; esac done deplibs="$new_libs" # All the library-specific variables (install_libdir is set above). library_names= old_library= dlname= # Test again, we may have decided not to build it any more if test "$build_libtool_libs" = yes; then # Remove ${wl} instances when linking with ld. # FIXME: should test the right _cmds variable. case $archive_cmds in *\$LD\ *) wl= ;; esac if test "$hardcode_into_libs" = yes; then # Hardcode the library paths hardcode_libdirs= dep_rpath= rpath="$finalize_rpath" test "$opt_mode" != relink && rpath="$compile_rpath$rpath" for libdir in $rpath; do if test -n "$hardcode_libdir_flag_spec"; then if test -n "$hardcode_libdir_separator"; then func_replace_sysroot "$libdir" libdir=$func_replace_sysroot_result if test -z "$hardcode_libdirs"; then hardcode_libdirs="$libdir" else # Just accumulate the unique libdirs. case $hardcode_libdir_separator$hardcode_libdirs$hardcode_libdir_separator in *"$hardcode_libdir_separator$libdir$hardcode_libdir_separator"*) ;; *) func_append hardcode_libdirs "$hardcode_libdir_separator$libdir" ;; esac fi else eval flag=\"$hardcode_libdir_flag_spec\" func_append dep_rpath " $flag" fi elif test -n "$runpath_var"; then case "$perm_rpath " in *" $libdir "*) ;; *) func_append perm_rpath " $libdir" ;; esac fi done # Substitute the hardcoded libdirs into the rpath. if test -n "$hardcode_libdir_separator" && test -n "$hardcode_libdirs"; then libdir="$hardcode_libdirs" eval "dep_rpath=\"$hardcode_libdir_flag_spec\"" fi if test -n "$runpath_var" && test -n "$perm_rpath"; then # We should set the runpath_var. rpath= for dir in $perm_rpath; do func_append rpath "$dir:" done eval "$runpath_var='$rpath\$$runpath_var'; export $runpath_var" fi test -n "$dep_rpath" && deplibs="$dep_rpath $deplibs" fi shlibpath="$finalize_shlibpath" test "$opt_mode" != relink && shlibpath="$compile_shlibpath$shlibpath" if test -n "$shlibpath"; then eval "$shlibpath_var='$shlibpath\$$shlibpath_var'; export $shlibpath_var" fi # Get the real and link names of the library. eval shared_ext=\"$shrext_cmds\" eval library_names=\"$library_names_spec\" set dummy $library_names shift realname="$1" shift if test -n "$soname_spec"; then eval soname=\"$soname_spec\" else soname="$realname" fi if test -z "$dlname"; then dlname=$soname fi lib="$output_objdir/$realname" linknames= for link do func_append linknames " $link" done # Use standard objects if they are pic test -z "$pic_flag" && libobjs=`$ECHO "$libobjs" | $SP2NL | $SED "$lo2o" | $NL2SP` test "X$libobjs" = "X " && libobjs= delfiles= if test -n "$export_symbols" && test -n "$include_expsyms"; then $opt_dry_run || cp "$export_symbols" "$output_objdir/$libname.uexp" export_symbols="$output_objdir/$libname.uexp" func_append delfiles " $export_symbols" fi orig_export_symbols= case $host_os in cygwin* | mingw* | cegcc*) if test -n "$export_symbols" && test -z "$export_symbols_regex"; then # exporting using user supplied symfile if test "x`$SED 1q $export_symbols`" != xEXPORTS; then # and it's NOT already a .def file. Must figure out # which of the given symbols are data symbols and tag # them as such. So, trigger use of export_symbols_cmds. # export_symbols gets reassigned inside the "prepare # the list of exported symbols" if statement, so the # include_expsyms logic still works. orig_export_symbols="$export_symbols" export_symbols= always_export_symbols=yes fi fi ;; esac # Prepare the list of exported symbols if test -z "$export_symbols"; then if test "$always_export_symbols" = yes || test -n "$export_symbols_regex"; then func_verbose "generating symbol list for \`$libname.la'" export_symbols="$output_objdir/$libname.exp" $opt_dry_run || $RM $export_symbols cmds=$export_symbols_cmds save_ifs="$IFS"; IFS='~' for cmd1 in $cmds; do IFS="$save_ifs" # Take the normal branch if the nm_file_list_spec branch # doesn't work or if tool conversion is not needed. case $nm_file_list_spec~$to_tool_file_cmd in *~func_convert_file_noop | *~func_convert_file_msys_to_w32 | ~*) try_normal_branch=yes eval cmd=\"$cmd1\" func_len " $cmd" len=$func_len_result ;; *) try_normal_branch=no ;; esac if test "$try_normal_branch" = yes \ && { test "$len" -lt "$max_cmd_len" \ || test "$max_cmd_len" -le -1; } then func_show_eval "$cmd" 'exit $?' skipped_export=false elif test -n "$nm_file_list_spec"; then func_basename "$output" output_la=$func_basename_result save_libobjs=$libobjs save_output=$output output=${output_objdir}/${output_la}.nm func_to_tool_file "$output" libobjs=$nm_file_list_spec$func_to_tool_file_result func_append delfiles " $output" func_verbose "creating $NM input file list: $output" for obj in $save_libobjs; do func_to_tool_file "$obj" $ECHO "$func_to_tool_file_result" done > "$output" eval cmd=\"$cmd1\" func_show_eval "$cmd" 'exit $?' output=$save_output libobjs=$save_libobjs skipped_export=false else # The command line is too long to execute in one step. func_verbose "using reloadable object file for export list..." skipped_export=: # Break out early, otherwise skipped_export may be # set to false by a later but shorter cmd. break fi done IFS="$save_ifs" if test -n "$export_symbols_regex" && test "X$skipped_export" != "X:"; then func_show_eval '$EGREP -e "$export_symbols_regex" "$export_symbols" > "${export_symbols}T"' func_show_eval '$MV "${export_symbols}T" "$export_symbols"' fi fi fi if test -n "$export_symbols" && test -n "$include_expsyms"; then tmp_export_symbols="$export_symbols" test -n "$orig_export_symbols" && tmp_export_symbols="$orig_export_symbols" $opt_dry_run || eval '$ECHO "$include_expsyms" | $SP2NL >> "$tmp_export_symbols"' fi if test "X$skipped_export" != "X:" && test -n "$orig_export_symbols"; then # The given exports_symbols file has to be filtered, so filter it. func_verbose "filter symbol list for \`$libname.la' to tag DATA exports" # FIXME: $output_objdir/$libname.filter potentially contains lots of # 's' commands which not all seds can handle. GNU sed should be fine # though. Also, the filter scales superlinearly with the number of # global variables. join(1) would be nice here, but unfortunately # isn't a blessed tool. $opt_dry_run || $SED -e '/[ ,]DATA/!d;s,\(.*\)\([ \,].*\),s|^\1$|\1\2|,' < $export_symbols > $output_objdir/$libname.filter func_append delfiles " $export_symbols $output_objdir/$libname.filter" export_symbols=$output_objdir/$libname.def $opt_dry_run || $SED -f $output_objdir/$libname.filter < $orig_export_symbols > $export_symbols fi tmp_deplibs= for test_deplib in $deplibs; do case " $convenience " in *" $test_deplib "*) ;; *) func_append tmp_deplibs " $test_deplib" ;; esac done deplibs="$tmp_deplibs" if test -n "$convenience"; then if test -n "$whole_archive_flag_spec" && test "$compiler_needs_object" = yes && test -z "$libobjs"; then # extract the archives, so we have objects to list. # TODO: could optimize this to just extract one archive. whole_archive_flag_spec= fi if test -n "$whole_archive_flag_spec"; then save_libobjs=$libobjs eval libobjs=\"\$libobjs $whole_archive_flag_spec\" test "X$libobjs" = "X " && libobjs= else gentop="$output_objdir/${outputname}x" func_append generated " $gentop" func_extract_archives $gentop $convenience func_append libobjs " $func_extract_archives_result" test "X$libobjs" = "X " && libobjs= fi fi if test "$thread_safe" = yes && test -n "$thread_safe_flag_spec"; then eval flag=\"$thread_safe_flag_spec\" func_append linker_flags " $flag" fi # Make a backup of the uninstalled library when relinking if test "$opt_mode" = relink; then $opt_dry_run || eval '(cd $output_objdir && $RM ${realname}U && $MV $realname ${realname}U)' || exit $? fi # Do each of the archive commands. if test "$module" = yes && test -n "$module_cmds" ; then if test -n "$export_symbols" && test -n "$module_expsym_cmds"; then eval test_cmds=\"$module_expsym_cmds\" cmds=$module_expsym_cmds else eval test_cmds=\"$module_cmds\" cmds=$module_cmds fi else if test -n "$export_symbols" && test -n "$archive_expsym_cmds"; then eval test_cmds=\"$archive_expsym_cmds\" cmds=$archive_expsym_cmds else eval test_cmds=\"$archive_cmds\" cmds=$archive_cmds fi fi if test "X$skipped_export" != "X:" && func_len " $test_cmds" && len=$func_len_result && test "$len" -lt "$max_cmd_len" || test "$max_cmd_len" -le -1; then : else # The command line is too long to link in one step, link piecewise # or, if using GNU ld and skipped_export is not :, use a linker # script. # Save the value of $output and $libobjs because we want to # use them later. If we have whole_archive_flag_spec, we # want to use save_libobjs as it was before # whole_archive_flag_spec was expanded, because we can't # assume the linker understands whole_archive_flag_spec. # This may have to be revisited, in case too many # convenience libraries get linked in and end up exceeding # the spec. if test -z "$convenience" || test -z "$whole_archive_flag_spec"; then save_libobjs=$libobjs fi save_output=$output func_basename "$output" output_la=$func_basename_result # Clear the reloadable object creation command queue and # initialize k to one. test_cmds= concat_cmds= objlist= last_robj= k=1 if test -n "$save_libobjs" && test "X$skipped_export" != "X:" && test "$with_gnu_ld" = yes; then output=${output_objdir}/${output_la}.lnkscript func_verbose "creating GNU ld script: $output" echo 'INPUT (' > $output for obj in $save_libobjs do func_to_tool_file "$obj" $ECHO "$func_to_tool_file_result" >> $output done echo ')' >> $output func_append delfiles " $output" func_to_tool_file "$output" output=$func_to_tool_file_result elif test -n "$save_libobjs" && test "X$skipped_export" != "X:" && test "X$file_list_spec" != X; then output=${output_objdir}/${output_la}.lnk func_verbose "creating linker input file list: $output" : > $output set x $save_libobjs shift firstobj= if test "$compiler_needs_object" = yes; then firstobj="$1 " shift fi for obj do func_to_tool_file "$obj" $ECHO "$func_to_tool_file_result" >> $output done func_append delfiles " $output" func_to_tool_file "$output" output=$firstobj\"$file_list_spec$func_to_tool_file_result\" else if test -n "$save_libobjs"; then func_verbose "creating reloadable object files..." output=$output_objdir/$output_la-${k}.$objext eval test_cmds=\"$reload_cmds\" func_len " $test_cmds" len0=$func_len_result len=$len0 # Loop over the list of objects to be linked. for obj in $save_libobjs do func_len " $obj" func_arith $len + $func_len_result len=$func_arith_result if test "X$objlist" = X || test "$len" -lt "$max_cmd_len"; then func_append objlist " $obj" else # The command $test_cmds is almost too long, add a # command to the queue. if test "$k" -eq 1 ; then # The first file doesn't have a previous command to add. reload_objs=$objlist eval concat_cmds=\"$reload_cmds\" else # All subsequent reloadable object files will link in # the last one created. reload_objs="$objlist $last_robj" eval concat_cmds=\"\$concat_cmds~$reload_cmds~\$RM $last_robj\" fi last_robj=$output_objdir/$output_la-${k}.$objext func_arith $k + 1 k=$func_arith_result output=$output_objdir/$output_la-${k}.$objext objlist=" $obj" func_len " $last_robj" func_arith $len0 + $func_len_result len=$func_arith_result fi done # Handle the remaining objects by creating one last # reloadable object file. All subsequent reloadable object # files will link in the last one created. test -z "$concat_cmds" || concat_cmds=$concat_cmds~ reload_objs="$objlist $last_robj" eval concat_cmds=\"\${concat_cmds}$reload_cmds\" if test -n "$last_robj"; then eval concat_cmds=\"\${concat_cmds}~\$RM $last_robj\" fi func_append delfiles " $output" else output= fi if ${skipped_export-false}; then func_verbose "generating symbol list for \`$libname.la'" export_symbols="$output_objdir/$libname.exp" $opt_dry_run || $RM $export_symbols libobjs=$output # Append the command to create the export file. test -z "$concat_cmds" || concat_cmds=$concat_cmds~ eval concat_cmds=\"\$concat_cmds$export_symbols_cmds\" if test -n "$last_robj"; then eval concat_cmds=\"\$concat_cmds~\$RM $last_robj\" fi fi test -n "$save_libobjs" && func_verbose "creating a temporary reloadable object file: $output" # Loop through the commands generated above and execute them. save_ifs="$IFS"; IFS='~' for cmd in $concat_cmds; do IFS="$save_ifs" $opt_silent || { func_quote_for_expand "$cmd" eval "func_echo $func_quote_for_expand_result" } $opt_dry_run || eval "$cmd" || { lt_exit=$? # Restore the uninstalled library and exit if test "$opt_mode" = relink; then ( cd "$output_objdir" && \ $RM "${realname}T" && \ $MV "${realname}U" "$realname" ) fi exit $lt_exit } done IFS="$save_ifs" if test -n "$export_symbols_regex" && ${skipped_export-false}; then func_show_eval '$EGREP -e "$export_symbols_regex" "$export_symbols" > "${export_symbols}T"' func_show_eval '$MV "${export_symbols}T" "$export_symbols"' fi fi if ${skipped_export-false}; then if test -n "$export_symbols" && test -n "$include_expsyms"; then tmp_export_symbols="$export_symbols" test -n "$orig_export_symbols" && tmp_export_symbols="$orig_export_symbols" $opt_dry_run || eval '$ECHO "$include_expsyms" | $SP2NL >> "$tmp_export_symbols"' fi if test -n "$orig_export_symbols"; then # The given exports_symbols file has to be filtered, so filter it. func_verbose "filter symbol list for \`$libname.la' to tag DATA exports" # FIXME: $output_objdir/$libname.filter potentially contains lots of # 's' commands which not all seds can handle. GNU sed should be fine # though. Also, the filter scales superlinearly with the number of # global variables. join(1) would be nice here, but unfortunately # isn't a blessed tool. $opt_dry_run || $SED -e '/[ ,]DATA/!d;s,\(.*\)\([ \,].*\),s|^\1$|\1\2|,' < $export_symbols > $output_objdir/$libname.filter func_append delfiles " $export_symbols $output_objdir/$libname.filter" export_symbols=$output_objdir/$libname.def $opt_dry_run || $SED -f $output_objdir/$libname.filter < $orig_export_symbols > $export_symbols fi fi libobjs=$output # Restore the value of output. output=$save_output if test -n "$convenience" && test -n "$whole_archive_flag_spec"; then eval libobjs=\"\$libobjs $whole_archive_flag_spec\" test "X$libobjs" = "X " && libobjs= fi # Expand the library linking commands again to reset the # value of $libobjs for piecewise linking. # Do each of the archive commands. if test "$module" = yes && test -n "$module_cmds" ; then if test -n "$export_symbols" && test -n "$module_expsym_cmds"; then cmds=$module_expsym_cmds else cmds=$module_cmds fi else if test -n "$export_symbols" && test -n "$archive_expsym_cmds"; then cmds=$archive_expsym_cmds else cmds=$archive_cmds fi fi fi if test -n "$delfiles"; then # Append the command to remove temporary files to $cmds. eval cmds=\"\$cmds~\$RM $delfiles\" fi # Add any objects from preloaded convenience libraries if test -n "$dlprefiles"; then gentop="$output_objdir/${outputname}x" func_append generated " $gentop" func_extract_archives $gentop $dlprefiles func_append libobjs " $func_extract_archives_result" test "X$libobjs" = "X " && libobjs= fi save_ifs="$IFS"; IFS='~' for cmd in $cmds; do IFS="$save_ifs" eval cmd=\"$cmd\" $opt_silent || { func_quote_for_expand "$cmd" eval "func_echo $func_quote_for_expand_result" } $opt_dry_run || eval "$cmd" || { lt_exit=$? # Restore the uninstalled library and exit if test "$opt_mode" = relink; then ( cd "$output_objdir" && \ $RM "${realname}T" && \ $MV "${realname}U" "$realname" ) fi exit $lt_exit } done IFS="$save_ifs" # Restore the uninstalled library and exit if test "$opt_mode" = relink; then $opt_dry_run || eval '(cd $output_objdir && $RM ${realname}T && $MV $realname ${realname}T && $MV ${realname}U $realname)' || exit $? if test -n "$convenience"; then if test -z "$whole_archive_flag_spec"; then func_show_eval '${RM}r "$gentop"' fi fi exit $EXIT_SUCCESS fi # Create links to the real library. for linkname in $linknames; do if test "$realname" != "$linkname"; then func_show_eval '(cd "$output_objdir" && $RM "$linkname" && $LN_S "$realname" "$linkname")' 'exit $?' fi done # If -module or -export-dynamic was specified, set the dlname. if test "$module" = yes || test "$export_dynamic" = yes; then # On all known operating systems, these are identical. dlname="$soname" fi fi ;; obj) if test -n "$dlfiles$dlprefiles" || test "$dlself" != no; then func_warning "\`-dlopen' is ignored for objects" fi case " $deplibs" in *\ -l* | *\ -L*) func_warning "\`-l' and \`-L' are ignored for objects" ;; esac test -n "$rpath" && \ func_warning "\`-rpath' is ignored for objects" test -n "$xrpath" && \ func_warning "\`-R' is ignored for objects" test -n "$vinfo" && \ func_warning "\`-version-info' is ignored for objects" test -n "$release" && \ func_warning "\`-release' is ignored for objects" case $output in *.lo) test -n "$objs$old_deplibs" && \ func_fatal_error "cannot build library object \`$output' from non-libtool objects" libobj=$output func_lo2o "$libobj" obj=$func_lo2o_result ;; *) libobj= obj="$output" ;; esac # Delete the old objects. $opt_dry_run || $RM $obj $libobj # Objects from convenience libraries. This assumes # single-version convenience libraries. Whenever we create # different ones for PIC/non-PIC, this we'll have to duplicate # the extraction. reload_conv_objs= gentop= # reload_cmds runs $LD directly, so let us get rid of # -Wl from whole_archive_flag_spec and hope we can get by with # turning comma into space.. wl= if test -n "$convenience"; then if test -n "$whole_archive_flag_spec"; then eval tmp_whole_archive_flags=\"$whole_archive_flag_spec\" reload_conv_objs=$reload_objs\ `$ECHO "$tmp_whole_archive_flags" | $SED 's|,| |g'` else gentop="$output_objdir/${obj}x" func_append generated " $gentop" func_extract_archives $gentop $convenience reload_conv_objs="$reload_objs $func_extract_archives_result" fi fi # If we're not building shared, we need to use non_pic_objs test "$build_libtool_libs" != yes && libobjs="$non_pic_objects" # Create the old-style object. reload_objs="$objs$old_deplibs "`$ECHO "$libobjs" | $SP2NL | $SED "/\.${libext}$/d; /\.lib$/d; $lo2o" | $NL2SP`" $reload_conv_objs" ### testsuite: skip nested quoting test output="$obj" func_execute_cmds "$reload_cmds" 'exit $?' # Exit if we aren't doing a library object file. if test -z "$libobj"; then if test -n "$gentop"; then func_show_eval '${RM}r "$gentop"' fi exit $EXIT_SUCCESS fi if test "$build_libtool_libs" != yes; then if test -n "$gentop"; then func_show_eval '${RM}r "$gentop"' fi # Create an invalid libtool object if no PIC, so that we don't # accidentally link it into a program. # $show "echo timestamp > $libobj" # $opt_dry_run || eval "echo timestamp > $libobj" || exit $? exit $EXIT_SUCCESS fi if test -n "$pic_flag" || test "$pic_mode" != default; then # Only do commands if we really have different PIC objects. reload_objs="$libobjs $reload_conv_objs" output="$libobj" func_execute_cmds "$reload_cmds" 'exit $?' fi if test -n "$gentop"; then func_show_eval '${RM}r "$gentop"' fi exit $EXIT_SUCCESS ;; prog) case $host in *cygwin*) func_stripname '' '.exe' "$output" output=$func_stripname_result.exe;; esac test -n "$vinfo" && \ func_warning "\`-version-info' is ignored for programs" test -n "$release" && \ func_warning "\`-release' is ignored for programs" test "$preload" = yes \ && test "$dlopen_support" = unknown \ && test "$dlopen_self" = unknown \ && test "$dlopen_self_static" = unknown && \ func_warning "\`LT_INIT([dlopen])' not used. Assuming no dlopen support." case $host in *-*-rhapsody* | *-*-darwin1.[012]) # On Rhapsody replace the C library is the System framework compile_deplibs=`$ECHO " $compile_deplibs" | $SED 's/ -lc / System.ltframework /'` finalize_deplibs=`$ECHO " $finalize_deplibs" | $SED 's/ -lc / System.ltframework /'` ;; esac case $host in *-*-darwin*) # Don't allow lazy linking, it breaks C++ global constructors # But is supposedly fixed on 10.4 or later (yay!). if test "$tagname" = CXX ; then case ${MACOSX_DEPLOYMENT_TARGET-10.0} in 10.[0123]) func_append compile_command " ${wl}-bind_at_load" func_append finalize_command " ${wl}-bind_at_load" ;; esac fi # Time to change all our "foo.ltframework" stuff back to "-framework foo" compile_deplibs=`$ECHO " $compile_deplibs" | $SED 's% \([^ $]*\).ltframework% -framework \1%g'` finalize_deplibs=`$ECHO " $finalize_deplibs" | $SED 's% \([^ $]*\).ltframework% -framework \1%g'` ;; esac # move library search paths that coincide with paths to not yet # installed libraries to the beginning of the library search list new_libs= for path in $notinst_path; do case " $new_libs " in *" -L$path/$objdir "*) ;; *) case " $compile_deplibs " in *" -L$path/$objdir "*) func_append new_libs " -L$path/$objdir" ;; esac ;; esac done for deplib in $compile_deplibs; do case $deplib in -L*) case " $new_libs " in *" $deplib "*) ;; *) func_append new_libs " $deplib" ;; esac ;; *) func_append new_libs " $deplib" ;; esac done compile_deplibs="$new_libs" func_append compile_command " $compile_deplibs" func_append finalize_command " $finalize_deplibs" if test -n "$rpath$xrpath"; then # If the user specified any rpath flags, then add them. for libdir in $rpath $xrpath; do # This is the magic to use -rpath. case "$finalize_rpath " in *" $libdir "*) ;; *) func_append finalize_rpath " $libdir" ;; esac done fi # Now hardcode the library paths rpath= hardcode_libdirs= for libdir in $compile_rpath $finalize_rpath; do if test -n "$hardcode_libdir_flag_spec"; then if test -n "$hardcode_libdir_separator"; then if test -z "$hardcode_libdirs"; then hardcode_libdirs="$libdir" else # Just accumulate the unique libdirs. case $hardcode_libdir_separator$hardcode_libdirs$hardcode_libdir_separator in *"$hardcode_libdir_separator$libdir$hardcode_libdir_separator"*) ;; *) func_append hardcode_libdirs "$hardcode_libdir_separator$libdir" ;; esac fi else eval flag=\"$hardcode_libdir_flag_spec\" func_append rpath " $flag" fi elif test -n "$runpath_var"; then case "$perm_rpath " in *" $libdir "*) ;; *) func_append perm_rpath " $libdir" ;; esac fi case $host in *-*-cygwin* | *-*-mingw* | *-*-pw32* | *-*-os2* | *-cegcc*) testbindir=`${ECHO} "$libdir" | ${SED} -e 's*/lib$*/bin*'` case :$dllsearchpath: in *":$libdir:"*) ;; ::) dllsearchpath=$libdir;; *) func_append dllsearchpath ":$libdir";; esac case :$dllsearchpath: in *":$testbindir:"*) ;; ::) dllsearchpath=$testbindir;; *) func_append dllsearchpath ":$testbindir";; esac ;; esac done # Substitute the hardcoded libdirs into the rpath. if test -n "$hardcode_libdir_separator" && test -n "$hardcode_libdirs"; then libdir="$hardcode_libdirs" eval rpath=\" $hardcode_libdir_flag_spec\" fi compile_rpath="$rpath" rpath= hardcode_libdirs= for libdir in $finalize_rpath; do if test -n "$hardcode_libdir_flag_spec"; then if test -n "$hardcode_libdir_separator"; then if test -z "$hardcode_libdirs"; then hardcode_libdirs="$libdir" else # Just accumulate the unique libdirs. case $hardcode_libdir_separator$hardcode_libdirs$hardcode_libdir_separator in *"$hardcode_libdir_separator$libdir$hardcode_libdir_separator"*) ;; *) func_append hardcode_libdirs "$hardcode_libdir_separator$libdir" ;; esac fi else eval flag=\"$hardcode_libdir_flag_spec\" func_append rpath " $flag" fi elif test -n "$runpath_var"; then case "$finalize_perm_rpath " in *" $libdir "*) ;; *) func_append finalize_perm_rpath " $libdir" ;; esac fi done # Substitute the hardcoded libdirs into the rpath. if test -n "$hardcode_libdir_separator" && test -n "$hardcode_libdirs"; then libdir="$hardcode_libdirs" eval rpath=\" $hardcode_libdir_flag_spec\" fi finalize_rpath="$rpath" if test -n "$libobjs" && test "$build_old_libs" = yes; then # Transform all the library objects into standard objects. compile_command=`$ECHO "$compile_command" | $SP2NL | $SED "$lo2o" | $NL2SP` finalize_command=`$ECHO "$finalize_command" | $SP2NL | $SED "$lo2o" | $NL2SP` fi func_generate_dlsyms "$outputname" "@PROGRAM@" "no" # template prelinking step if test -n "$prelink_cmds"; then func_execute_cmds "$prelink_cmds" 'exit $?' fi wrappers_required=yes case $host in *cegcc* | *mingw32ce*) # Disable wrappers for cegcc and mingw32ce hosts, we are cross compiling anyway. wrappers_required=no ;; *cygwin* | *mingw* ) if test "$build_libtool_libs" != yes; then wrappers_required=no fi ;; *) if test "$need_relink" = no || test "$build_libtool_libs" != yes; then wrappers_required=no fi ;; esac if test "$wrappers_required" = no; then # Replace the output file specification. compile_command=`$ECHO "$compile_command" | $SED 's%@OUTPUT@%'"$output"'%g'` link_command="$compile_command$compile_rpath" # We have no uninstalled library dependencies, so finalize right now. exit_status=0 func_show_eval "$link_command" 'exit_status=$?' if test -n "$postlink_cmds"; then func_to_tool_file "$output" postlink_cmds=`func_echo_all "$postlink_cmds" | $SED -e 's%@OUTPUT@%'"$output"'%g' -e 's%@TOOL_OUTPUT@%'"$func_to_tool_file_result"'%g'` func_execute_cmds "$postlink_cmds" 'exit $?' fi # Delete the generated files. if test -f "$output_objdir/${outputname}S.${objext}"; then func_show_eval '$RM "$output_objdir/${outputname}S.${objext}"' fi exit $exit_status fi if test -n "$compile_shlibpath$finalize_shlibpath"; then compile_command="$shlibpath_var=\"$compile_shlibpath$finalize_shlibpath\$$shlibpath_var\" $compile_command" fi if test -n "$finalize_shlibpath"; then finalize_command="$shlibpath_var=\"$finalize_shlibpath\$$shlibpath_var\" $finalize_command" fi compile_var= finalize_var= if test -n "$runpath_var"; then if test -n "$perm_rpath"; then # We should set the runpath_var. rpath= for dir in $perm_rpath; do func_append rpath "$dir:" done compile_var="$runpath_var=\"$rpath\$$runpath_var\" " fi if test -n "$finalize_perm_rpath"; then # We should set the runpath_var. rpath= for dir in $finalize_perm_rpath; do func_append rpath "$dir:" done finalize_var="$runpath_var=\"$rpath\$$runpath_var\" " fi fi if test "$no_install" = yes; then # We don't need to create a wrapper script. link_command="$compile_var$compile_command$compile_rpath" # Replace the output file specification. link_command=`$ECHO "$link_command" | $SED 's%@OUTPUT@%'"$output"'%g'` # Delete the old output file. $opt_dry_run || $RM $output # Link the executable and exit func_show_eval "$link_command" 'exit $?' if test -n "$postlink_cmds"; then func_to_tool_file "$output" postlink_cmds=`func_echo_all "$postlink_cmds" | $SED -e 's%@OUTPUT@%'"$output"'%g' -e 's%@TOOL_OUTPUT@%'"$func_to_tool_file_result"'%g'` func_execute_cmds "$postlink_cmds" 'exit $?' fi exit $EXIT_SUCCESS fi if test "$hardcode_action" = relink; then # Fast installation is not supported link_command="$compile_var$compile_command$compile_rpath" relink_command="$finalize_var$finalize_command$finalize_rpath" func_warning "this platform does not like uninstalled shared libraries" func_warning "\`$output' will be relinked during installation" else if test "$fast_install" != no; then link_command="$finalize_var$compile_command$finalize_rpath" if test "$fast_install" = yes; then relink_command=`$ECHO "$compile_var$compile_command$compile_rpath" | $SED 's%@OUTPUT@%\$progdir/\$file%g'` else # fast_install is set to needless relink_command= fi else link_command="$compile_var$compile_command$compile_rpath" relink_command="$finalize_var$finalize_command$finalize_rpath" fi fi # Replace the output file specification. link_command=`$ECHO "$link_command" | $SED 's%@OUTPUT@%'"$output_objdir/$outputname"'%g'` # Delete the old output files. $opt_dry_run || $RM $output $output_objdir/$outputname $output_objdir/lt-$outputname func_show_eval "$link_command" 'exit $?' if test -n "$postlink_cmds"; then func_to_tool_file "$output_objdir/$outputname" postlink_cmds=`func_echo_all "$postlink_cmds" | $SED -e 's%@OUTPUT@%'"$output_objdir/$outputname"'%g' -e 's%@TOOL_OUTPUT@%'"$func_to_tool_file_result"'%g'` func_execute_cmds "$postlink_cmds" 'exit $?' fi # Now create the wrapper script. func_verbose "creating $output" # Quote the relink command for shipping. if test -n "$relink_command"; then # Preserve any variables that may affect compiler behavior for var in $variables_saved_for_relink; do if eval test -z \"\${$var+set}\"; then relink_command="{ test -z \"\${$var+set}\" || $lt_unset $var || { $var=; export $var; }; }; $relink_command" elif eval var_value=\$$var; test -z "$var_value"; then relink_command="$var=; export $var; $relink_command" else func_quote_for_eval "$var_value" relink_command="$var=$func_quote_for_eval_result; export $var; $relink_command" fi done relink_command="(cd `pwd`; $relink_command)" relink_command=`$ECHO "$relink_command" | $SED "$sed_quote_subst"` fi # Only actually do things if not in dry run mode. $opt_dry_run || { # win32 will think the script is a binary if it has # a .exe suffix, so we strip it off here. case $output in *.exe) func_stripname '' '.exe' "$output" output=$func_stripname_result ;; esac # test for cygwin because mv fails w/o .exe extensions case $host in *cygwin*) exeext=.exe func_stripname '' '.exe' "$outputname" outputname=$func_stripname_result ;; *) exeext= ;; esac case $host in *cygwin* | *mingw* ) func_dirname_and_basename "$output" "" "." output_name=$func_basename_result output_path=$func_dirname_result cwrappersource="$output_path/$objdir/lt-$output_name.c" cwrapper="$output_path/$output_name.exe" $RM $cwrappersource $cwrapper trap "$RM $cwrappersource $cwrapper; exit $EXIT_FAILURE" 1 2 15 func_emit_cwrapperexe_src > $cwrappersource # The wrapper executable is built using the $host compiler, # because it contains $host paths and files. If cross- # compiling, it, like the target executable, must be # executed on the $host or under an emulation environment. $opt_dry_run || { $LTCC $LTCFLAGS -o $cwrapper $cwrappersource $STRIP $cwrapper } # Now, create the wrapper script for func_source use: func_ltwrapper_scriptname $cwrapper $RM $func_ltwrapper_scriptname_result trap "$RM $func_ltwrapper_scriptname_result; exit $EXIT_FAILURE" 1 2 15 $opt_dry_run || { # note: this script will not be executed, so do not chmod. if test "x$build" = "x$host" ; then $cwrapper --lt-dump-script > $func_ltwrapper_scriptname_result else func_emit_wrapper no > $func_ltwrapper_scriptname_result fi } ;; * ) $RM $output trap "$RM $output; exit $EXIT_FAILURE" 1 2 15 func_emit_wrapper no > $output chmod +x $output ;; esac } exit $EXIT_SUCCESS ;; esac # See if we need to build an old-fashioned archive. for oldlib in $oldlibs; do if test "$build_libtool_libs" = convenience; then oldobjs="$libobjs_save $symfileobj" addlibs="$convenience" build_libtool_libs=no else if test "$build_libtool_libs" = module; then oldobjs="$libobjs_save" build_libtool_libs=no else oldobjs="$old_deplibs $non_pic_objects" if test "$preload" = yes && test -f "$symfileobj"; then func_append oldobjs " $symfileobj" fi fi addlibs="$old_convenience" fi if test -n "$addlibs"; then gentop="$output_objdir/${outputname}x" func_append generated " $gentop" func_extract_archives $gentop $addlibs func_append oldobjs " $func_extract_archives_result" fi # Do each command in the archive commands. if test -n "$old_archive_from_new_cmds" && test "$build_libtool_libs" = yes; then cmds=$old_archive_from_new_cmds else # Add any objects from preloaded convenience libraries if test -n "$dlprefiles"; then gentop="$output_objdir/${outputname}x" func_append generated " $gentop" func_extract_archives $gentop $dlprefiles func_append oldobjs " $func_extract_archives_result" fi # POSIX demands no paths to be encoded in archives. We have # to avoid creating archives with duplicate basenames if we # might have to extract them afterwards, e.g., when creating a # static archive out of a convenience library, or when linking # the entirety of a libtool archive into another (currently # not supported by libtool). if (for obj in $oldobjs do func_basename "$obj" $ECHO "$func_basename_result" done | sort | sort -uc >/dev/null 2>&1); then : else echo "copying selected object files to avoid basename conflicts..." gentop="$output_objdir/${outputname}x" func_append generated " $gentop" func_mkdir_p "$gentop" save_oldobjs=$oldobjs oldobjs= counter=1 for obj in $save_oldobjs do func_basename "$obj" objbase="$func_basename_result" case " $oldobjs " in " ") oldobjs=$obj ;; *[\ /]"$objbase "*) while :; do # Make sure we don't pick an alternate name that also # overlaps. newobj=lt$counter-$objbase func_arith $counter + 1 counter=$func_arith_result case " $oldobjs " in *[\ /]"$newobj "*) ;; *) if test ! -f "$gentop/$newobj"; then break; fi ;; esac done func_show_eval "ln $obj $gentop/$newobj || cp $obj $gentop/$newobj" func_append oldobjs " $gentop/$newobj" ;; *) func_append oldobjs " $obj" ;; esac done fi func_to_tool_file "$oldlib" func_convert_file_msys_to_w32 tool_oldlib=$func_to_tool_file_result eval cmds=\"$old_archive_cmds\" func_len " $cmds" len=$func_len_result if test "$len" -lt "$max_cmd_len" || test "$max_cmd_len" -le -1; then cmds=$old_archive_cmds elif test -n "$archiver_list_spec"; then func_verbose "using command file archive linking..." for obj in $oldobjs do func_to_tool_file "$obj" $ECHO "$func_to_tool_file_result" done > $output_objdir/$libname.libcmd func_to_tool_file "$output_objdir/$libname.libcmd" oldobjs=" $archiver_list_spec$func_to_tool_file_result" cmds=$old_archive_cmds else # the command line is too long to link in one step, link in parts func_verbose "using piecewise archive linking..." save_RANLIB=$RANLIB RANLIB=: objlist= concat_cmds= save_oldobjs=$oldobjs oldobjs= # Is there a better way of finding the last object in the list? for obj in $save_oldobjs do last_oldobj=$obj done eval test_cmds=\"$old_archive_cmds\" func_len " $test_cmds" len0=$func_len_result len=$len0 for obj in $save_oldobjs do func_len " $obj" func_arith $len + $func_len_result len=$func_arith_result func_append objlist " $obj" if test "$len" -lt "$max_cmd_len"; then : else # the above command should be used before it gets too long oldobjs=$objlist if test "$obj" = "$last_oldobj" ; then RANLIB=$save_RANLIB fi test -z "$concat_cmds" || concat_cmds=$concat_cmds~ eval concat_cmds=\"\${concat_cmds}$old_archive_cmds\" objlist= len=$len0 fi done RANLIB=$save_RANLIB oldobjs=$objlist if test "X$oldobjs" = "X" ; then eval cmds=\"\$concat_cmds\" else eval cmds=\"\$concat_cmds~\$old_archive_cmds\" fi fi fi func_execute_cmds "$cmds" 'exit $?' done test -n "$generated" && \ func_show_eval "${RM}r$generated" # Now create the libtool archive. case $output in *.la) old_library= test "$build_old_libs" = yes && old_library="$libname.$libext" func_verbose "creating $output" # Preserve any variables that may affect compiler behavior for var in $variables_saved_for_relink; do if eval test -z \"\${$var+set}\"; then relink_command="{ test -z \"\${$var+set}\" || $lt_unset $var || { $var=; export $var; }; }; $relink_command" elif eval var_value=\$$var; test -z "$var_value"; then relink_command="$var=; export $var; $relink_command" else func_quote_for_eval "$var_value" relink_command="$var=$func_quote_for_eval_result; export $var; $relink_command" fi done # Quote the link command for shipping. relink_command="(cd `pwd`; $SHELL $progpath $preserve_args --mode=relink $libtool_args @inst_prefix_dir@)" relink_command=`$ECHO "$relink_command" | $SED "$sed_quote_subst"` if test "$hardcode_automatic" = yes ; then relink_command= fi # Only create the output if not a dry run. $opt_dry_run || { for installed in no yes; do if test "$installed" = yes; then if test -z "$install_libdir"; then break fi output="$output_objdir/$outputname"i # Replace all uninstalled libtool libraries with the installed ones newdependency_libs= for deplib in $dependency_libs; do case $deplib in *.la) func_basename "$deplib" name="$func_basename_result" func_resolve_sysroot "$deplib" eval libdir=`${SED} -n -e 's/^libdir=\(.*\)$/\1/p' $func_resolve_sysroot_result` test -z "$libdir" && \ func_fatal_error "\`$deplib' is not a valid libtool archive" func_append newdependency_libs " ${lt_sysroot:+=}$libdir/$name" ;; -L*) func_stripname -L '' "$deplib" func_replace_sysroot "$func_stripname_result" func_append newdependency_libs " -L$func_replace_sysroot_result" ;; -R*) func_stripname -R '' "$deplib" func_replace_sysroot "$func_stripname_result" func_append newdependency_libs " -R$func_replace_sysroot_result" ;; *) func_append newdependency_libs " $deplib" ;; esac done dependency_libs="$newdependency_libs" newdlfiles= for lib in $dlfiles; do case $lib in *.la) func_basename "$lib" name="$func_basename_result" eval libdir=`${SED} -n -e 's/^libdir=\(.*\)$/\1/p' $lib` test -z "$libdir" && \ func_fatal_error "\`$lib' is not a valid libtool archive" func_append newdlfiles " ${lt_sysroot:+=}$libdir/$name" ;; *) func_append newdlfiles " $lib" ;; esac done dlfiles="$newdlfiles" newdlprefiles= for lib in $dlprefiles; do case $lib in *.la) # Only pass preopened files to the pseudo-archive (for # eventual linking with the app. that links it) if we # didn't already link the preopened objects directly into # the library: func_basename "$lib" name="$func_basename_result" eval libdir=`${SED} -n -e 's/^libdir=\(.*\)$/\1/p' $lib` test -z "$libdir" && \ func_fatal_error "\`$lib' is not a valid libtool archive" func_append newdlprefiles " ${lt_sysroot:+=}$libdir/$name" ;; esac done dlprefiles="$newdlprefiles" else newdlfiles= for lib in $dlfiles; do case $lib in [\\/]* | [A-Za-z]:[\\/]*) abs="$lib" ;; *) abs=`pwd`"/$lib" ;; esac func_append newdlfiles " $abs" done dlfiles="$newdlfiles" newdlprefiles= for lib in $dlprefiles; do case $lib in [\\/]* | [A-Za-z]:[\\/]*) abs="$lib" ;; *) abs=`pwd`"/$lib" ;; esac func_append newdlprefiles " $abs" done dlprefiles="$newdlprefiles" fi $RM $output # place dlname in correct position for cygwin # In fact, it would be nice if we could use this code for all target # systems that can't hard-code library paths into their executables # and that have no shared library path variable independent of PATH, # but it turns out we can't easily determine that from inspecting # libtool variables, so we have to hard-code the OSs to which it # applies here; at the moment, that means platforms that use the PE # object format with DLL files. See the long comment at the top of # tests/bindir.at for full details. tdlname=$dlname case $host,$output,$installed,$module,$dlname in *cygwin*,*lai,yes,no,*.dll | *mingw*,*lai,yes,no,*.dll | *cegcc*,*lai,yes,no,*.dll) # If a -bindir argument was supplied, place the dll there. if test "x$bindir" != x ; then func_relative_path "$install_libdir" "$bindir" tdlname=$func_relative_path_result$dlname else # Otherwise fall back on heuristic. tdlname=../bin/$dlname fi ;; esac $ECHO > $output "\ # $outputname - a libtool library file # Generated by $PROGRAM (GNU $PACKAGE$TIMESTAMP) $VERSION # # Please DO NOT delete this file! # It is necessary for linking the library. # The name that we can dlopen(3). dlname='$tdlname' # Names of this library. library_names='$library_names' # The name of the static archive. old_library='$old_library' # Linker flags that can not go in dependency_libs. inherited_linker_flags='$new_inherited_linker_flags' # Libraries that this one depends upon. dependency_libs='$dependency_libs' # Names of additional weak libraries provided by this library weak_library_names='$weak_libs' # Version information for $libname. current=$current age=$age revision=$revision # Is this an already installed library? installed=$installed # Should we warn about portability when linking against -modules? shouldnotlink=$module # Files to dlopen/dlpreopen dlopen='$dlfiles' dlpreopen='$dlprefiles' # Directory that this library needs to be installed in: libdir='$install_libdir'" if test "$installed" = no && test "$need_relink" = yes; then $ECHO >> $output "\ relink_command=\"$relink_command\"" fi done } # Do a symbolic link so that the libtool archive can be found in # LD_LIBRARY_PATH before the program is installed. func_show_eval '( cd "$output_objdir" && $RM "$outputname" && $LN_S "../$outputname" "$outputname" )' 'exit $?' ;; esac exit $EXIT_SUCCESS } { test "$opt_mode" = link || test "$opt_mode" = relink; } && func_mode_link ${1+"$@"} # func_mode_uninstall arg... func_mode_uninstall () { $opt_debug RM="$nonopt" files= rmforce= exit_status=0 # This variable tells wrapper scripts just to set variables rather # than running their programs. libtool_install_magic="$magic" for arg do case $arg in -f) func_append RM " $arg"; rmforce=yes ;; -*) func_append RM " $arg" ;; *) func_append files " $arg" ;; esac done test -z "$RM" && \ func_fatal_help "you must specify an RM program" rmdirs= for file in $files; do func_dirname "$file" "" "." dir="$func_dirname_result" if test "X$dir" = X.; then odir="$objdir" else odir="$dir/$objdir" fi func_basename "$file" name="$func_basename_result" test "$opt_mode" = uninstall && odir="$dir" # Remember odir for removal later, being careful to avoid duplicates if test "$opt_mode" = clean; then case " $rmdirs " in *" $odir "*) ;; *) func_append rmdirs " $odir" ;; esac fi # Don't error if the file doesn't exist and rm -f was used. if { test -L "$file"; } >/dev/null 2>&1 || { test -h "$file"; } >/dev/null 2>&1 || test -f "$file"; then : elif test -d "$file"; then exit_status=1 continue elif test "$rmforce" = yes; then continue fi rmfiles="$file" case $name in *.la) # Possibly a libtool archive, so verify it. if func_lalib_p "$file"; then func_source $dir/$name # Delete the libtool libraries and symlinks. for n in $library_names; do func_append rmfiles " $odir/$n" done test -n "$old_library" && func_append rmfiles " $odir/$old_library" case "$opt_mode" in clean) case " $library_names " in *" $dlname "*) ;; *) test -n "$dlname" && func_append rmfiles " $odir/$dlname" ;; esac test -n "$libdir" && func_append rmfiles " $odir/$name $odir/${name}i" ;; uninstall) if test -n "$library_names"; then # Do each command in the postuninstall commands. func_execute_cmds "$postuninstall_cmds" 'test "$rmforce" = yes || exit_status=1' fi if test -n "$old_library"; then # Do each command in the old_postuninstall commands. func_execute_cmds "$old_postuninstall_cmds" 'test "$rmforce" = yes || exit_status=1' fi # FIXME: should reinstall the best remaining shared library. ;; esac fi ;; *.lo) # Possibly a libtool object, so verify it. if func_lalib_p "$file"; then # Read the .lo file func_source $dir/$name # Add PIC object to the list of files to remove. if test -n "$pic_object" && test "$pic_object" != none; then func_append rmfiles " $dir/$pic_object" fi # Add non-PIC object to the list of files to remove. if test -n "$non_pic_object" && test "$non_pic_object" != none; then func_append rmfiles " $dir/$non_pic_object" fi fi ;; *) if test "$opt_mode" = clean ; then noexename=$name case $file in *.exe) func_stripname '' '.exe' "$file" file=$func_stripname_result func_stripname '' '.exe' "$name" noexename=$func_stripname_result # $file with .exe has already been added to rmfiles, # add $file without .exe func_append rmfiles " $file" ;; esac # Do a test to see if this is a libtool program. if func_ltwrapper_p "$file"; then if func_ltwrapper_executable_p "$file"; then func_ltwrapper_scriptname "$file" relink_command= func_source $func_ltwrapper_scriptname_result func_append rmfiles " $func_ltwrapper_scriptname_result" else relink_command= func_source $dir/$noexename fi # note $name still contains .exe if it was in $file originally # as does the version of $file that was added into $rmfiles func_append rmfiles " $odir/$name $odir/${name}S.${objext}" if test "$fast_install" = yes && test -n "$relink_command"; then func_append rmfiles " $odir/lt-$name" fi if test "X$noexename" != "X$name" ; then func_append rmfiles " $odir/lt-${noexename}.c" fi fi fi ;; esac func_show_eval "$RM $rmfiles" 'exit_status=1' done # Try to remove the ${objdir}s in the directories where we deleted files for dir in $rmdirs; do if test -d "$dir"; then func_show_eval "rmdir $dir >/dev/null 2>&1" fi done exit $exit_status } { test "$opt_mode" = uninstall || test "$opt_mode" = clean; } && func_mode_uninstall ${1+"$@"} test -z "$opt_mode" && { help="$generic_help" func_fatal_help "you must specify a MODE" } test -z "$exec_cmd" && \ func_fatal_help "invalid operation mode \`$opt_mode'" if test -n "$exec_cmd"; then eval exec "$exec_cmd" exit $EXIT_FAILURE fi exit $exit_status # The TAGs below are defined such that we never get into a situation # in which we disable both kinds of libraries. Given conflicting # choices, we go for a static library, that is the most portable, # since we can't tell whether shared libraries were disabled because # the user asked for that or because the platform doesn't support # them. This is particularly important on AIX, because we don't # support having both static and shared libraries enabled at the same # time on that platform, so we default to a shared-only configuration. # If a disable-shared tag is given, we'll fallback to a static-only # configuration. But we'll never go from static-only to shared-only. # ### BEGIN LIBTOOL TAG CONFIG: disable-shared build_libtool_libs=no build_old_libs=yes # ### END LIBTOOL TAG CONFIG: disable-shared # ### BEGIN LIBTOOL TAG CONFIG: disable-static build_old_libs=`case $build_libtool_libs in yes) echo no;; *) echo yes;; esac` # ### END LIBTOOL TAG CONFIG: disable-static # Local Variables: # mode:shell-script # sh-indentation:2 # End: # vi:sw=2 fftw-3.3.3/missing0000755000175000017500000002415211777117217011025 00000000000000#! /bin/sh # Common stub for a few missing GNU programs while installing. scriptversion=2012-01-06.13; # UTC # Copyright (C) 1996, 1997, 1999, 2000, 2002, 2003, 2004, 2005, 2006, # 2008, 2009, 2010, 2011, 2012 Free Software Foundation, Inc. # Originally by Fran,cois Pinard , 1996. # This program is free software; you can redistribute it and/or modify # it under the terms of the GNU General Public License as published by # the Free Software Foundation; either version 2, or (at your option) # any later version. # This program is distributed in the hope that it will be useful, # but WITHOUT ANY WARRANTY; without even the implied warranty of # MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the # GNU General Public License for more details. # You should have received a copy of the GNU General Public License # along with this program. If not, see . # As a special exception to the GNU General Public License, if you # distribute this file as part of a program that contains a # configuration script generated by Autoconf, you may include it under # the same distribution terms that you use for the rest of that program. if test $# -eq 0; then echo 1>&2 "Try \`$0 --help' for more information" exit 1 fi run=: sed_output='s/.* --output[ =]\([^ ]*\).*/\1/p' sed_minuso='s/.* -o \([^ ]*\).*/\1/p' # In the cases where this matters, `missing' is being run in the # srcdir already. if test -f configure.ac; then configure_ac=configure.ac else configure_ac=configure.in fi msg="missing on your system" case $1 in --run) # Try to run requested program, and just exit if it succeeds. run= shift "$@" && exit 0 # Exit code 63 means version mismatch. This often happens # when the user try to use an ancient version of a tool on # a file that requires a minimum version. In this case we # we should proceed has if the program had been absent, or # if --run hadn't been passed. if test $? = 63; then run=: msg="probably too old" fi ;; -h|--h|--he|--hel|--help) echo "\ $0 [OPTION]... PROGRAM [ARGUMENT]... Handle \`PROGRAM [ARGUMENT]...' for when PROGRAM is missing, or return an error status if there is no known handling for PROGRAM. Options: -h, --help display this help and exit -v, --version output version information and exit --run try to run the given command, and emulate it if it fails Supported PROGRAM values: aclocal touch file \`aclocal.m4' autoconf touch file \`configure' autoheader touch file \`config.h.in' autom4te touch the output file, or create a stub one automake touch all \`Makefile.in' files bison create \`y.tab.[ch]', if possible, from existing .[ch] flex create \`lex.yy.c', if possible, from existing .c help2man touch the output file lex create \`lex.yy.c', if possible, from existing .c makeinfo touch the output file yacc create \`y.tab.[ch]', if possible, from existing .[ch] Version suffixes to PROGRAM as well as the prefixes \`gnu-', \`gnu', and \`g' are ignored when checking the name. Send bug reports to ." exit $? ;; -v|--v|--ve|--ver|--vers|--versi|--versio|--version) echo "missing $scriptversion (GNU Automake)" exit $? ;; -*) echo 1>&2 "$0: Unknown \`$1' option" echo 1>&2 "Try \`$0 --help' for more information" exit 1 ;; esac # normalize program name to check for. program=`echo "$1" | sed ' s/^gnu-//; t s/^gnu//; t s/^g//; t'` # Now exit if we have it, but it failed. Also exit now if we # don't have it and --version was passed (most likely to detect # the program). This is about non-GNU programs, so use $1 not # $program. case $1 in lex*|yacc*) # Not GNU programs, they don't have --version. ;; *) if test -z "$run" && ($1 --version) > /dev/null 2>&1; then # We have it, but it failed. exit 1 elif test "x$2" = "x--version" || test "x$2" = "x--help"; then # Could not run --version or --help. This is probably someone # running `$TOOL --version' or `$TOOL --help' to check whether # $TOOL exists and not knowing $TOOL uses missing. exit 1 fi ;; esac # If it does not exist, or fails to run (possibly an outdated version), # try to emulate it. case $program in aclocal*) echo 1>&2 "\ WARNING: \`$1' is $msg. You should only need it if you modified \`acinclude.m4' or \`${configure_ac}'. You might want to install the \`Automake' and \`Perl' packages. Grab them from any GNU archive site." touch aclocal.m4 ;; autoconf*) echo 1>&2 "\ WARNING: \`$1' is $msg. You should only need it if you modified \`${configure_ac}'. You might want to install the \`Autoconf' and \`GNU m4' packages. Grab them from any GNU archive site." touch configure ;; autoheader*) echo 1>&2 "\ WARNING: \`$1' is $msg. You should only need it if you modified \`acconfig.h' or \`${configure_ac}'. You might want to install the \`Autoconf' and \`GNU m4' packages. Grab them from any GNU archive site." files=`sed -n 's/^[ ]*A[CM]_CONFIG_HEADER(\([^)]*\)).*/\1/p' ${configure_ac}` test -z "$files" && files="config.h" touch_files= for f in $files; do case $f in *:*) touch_files="$touch_files "`echo "$f" | sed -e 's/^[^:]*://' -e 's/:.*//'`;; *) touch_files="$touch_files $f.in";; esac done touch $touch_files ;; automake*) echo 1>&2 "\ WARNING: \`$1' is $msg. You should only need it if you modified \`Makefile.am', \`acinclude.m4' or \`${configure_ac}'. You might want to install the \`Automake' and \`Perl' packages. Grab them from any GNU archive site." find . -type f -name Makefile.am -print | sed 's/\.am$/.in/' | while read f; do touch "$f"; done ;; autom4te*) echo 1>&2 "\ WARNING: \`$1' is needed, but is $msg. You might have modified some files without having the proper tools for further handling them. You can get \`$1' as part of \`Autoconf' from any GNU archive site." file=`echo "$*" | sed -n "$sed_output"` test -z "$file" && file=`echo "$*" | sed -n "$sed_minuso"` if test -f "$file"; then touch $file else test -z "$file" || exec >$file echo "#! /bin/sh" echo "# Created by GNU Automake missing as a replacement of" echo "# $ $@" echo "exit 0" chmod +x $file exit 1 fi ;; bison*|yacc*) echo 1>&2 "\ WARNING: \`$1' $msg. You should only need it if you modified a \`.y' file. You may need the \`Bison' package in order for those modifications to take effect. You can get \`Bison' from any GNU archive site." rm -f y.tab.c y.tab.h if test $# -ne 1; then eval LASTARG=\${$#} case $LASTARG in *.y) SRCFILE=`echo "$LASTARG" | sed 's/y$/c/'` if test -f "$SRCFILE"; then cp "$SRCFILE" y.tab.c fi SRCFILE=`echo "$LASTARG" | sed 's/y$/h/'` if test -f "$SRCFILE"; then cp "$SRCFILE" y.tab.h fi ;; esac fi if test ! -f y.tab.h; then echo >y.tab.h fi if test ! -f y.tab.c; then echo 'main() { return 0; }' >y.tab.c fi ;; lex*|flex*) echo 1>&2 "\ WARNING: \`$1' is $msg. You should only need it if you modified a \`.l' file. You may need the \`Flex' package in order for those modifications to take effect. You can get \`Flex' from any GNU archive site." rm -f lex.yy.c if test $# -ne 1; then eval LASTARG=\${$#} case $LASTARG in *.l) SRCFILE=`echo "$LASTARG" | sed 's/l$/c/'` if test -f "$SRCFILE"; then cp "$SRCFILE" lex.yy.c fi ;; esac fi if test ! -f lex.yy.c; then echo 'main() { return 0; }' >lex.yy.c fi ;; help2man*) echo 1>&2 "\ WARNING: \`$1' is $msg. You should only need it if you modified a dependency of a manual page. You may need the \`Help2man' package in order for those modifications to take effect. You can get \`Help2man' from any GNU archive site." file=`echo "$*" | sed -n "$sed_output"` test -z "$file" && file=`echo "$*" | sed -n "$sed_minuso"` if test -f "$file"; then touch $file else test -z "$file" || exec >$file echo ".ab help2man is required to generate this page" exit $? fi ;; makeinfo*) echo 1>&2 "\ WARNING: \`$1' is $msg. You should only need it if you modified a \`.texi' or \`.texinfo' file, or any other file indirectly affecting the aspect of the manual. The spurious call might also be the consequence of using a buggy \`make' (AIX, DU, IRIX). You might want to install the \`Texinfo' package or the \`GNU make' package. Grab either from any GNU archive site." # The file to touch is that specified with -o ... file=`echo "$*" | sed -n "$sed_output"` test -z "$file" && file=`echo "$*" | sed -n "$sed_minuso"` if test -z "$file"; then # ... or it is the one specified with @setfilename ... infile=`echo "$*" | sed 's/.* \([^ ]*\) *$/\1/'` file=`sed -n ' /^@setfilename/{ s/.* \([^ ]*\) *$/\1/ p q }' $infile` # ... or it is derived from the source name (dir/f.texi becomes f.info) test -z "$file" && file=`echo "$infile" | sed 's,.*/,,;s,.[^.]*$,,'`.info fi # If the file does not exist, the user really needs makeinfo; # let's fail without touching anything. test -f $file || exit 1 touch $file ;; *) echo 1>&2 "\ WARNING: \`$1' is needed, and is $msg. You might have modified some files without having the proper tools for further handling them. Check the \`README' file, it often tells you about the needed prerequisites for installing this package. You may also peek at any GNU archive site, in case some other package would contain this missing \`$1' program." exit 1 ;; esac exit 0 # Local variables: # eval: (add-hook 'write-file-hooks 'time-stamp) # time-stamp-start: "scriptversion=" # time-stamp-format: "%:y-%02m-%02d.%02H" # time-stamp-time-zone: "UTC" # time-stamp-end: "; # UTC" # End: fftw-3.3.3/COPYRIGHT0000644000175000017500000000151412054410050010672 00000000000000/* * Copyright (c) 2003, 2007-11 Matteo Frigo * Copyright (c) 2003, 2007-11 Massachusetts Institute of Technology * * This program is free software; you can redistribute it and/or modify * it under the terms of the GNU General Public License as published by * the Free Software Foundation; either version 2 of the License, or * (at your option) any later version. * * This program is distributed in the hope that it will be useful, * but WITHOUT ANY WARRANTY; without even the implied warranty of * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the * GNU General Public License for more details. * * You should have received a copy of the GNU General Public License * along with this program; if not, write to the Free Software * Foundation, Inc., 51 Franklin Street, Fifth Floor, Boston, MA 02110-1301 USA * */ fftw-3.3.3/bootstrap.sh0000644000175000017500000000175412054410050011756 00000000000000#! /bin/sh ############################################################################ # # NOTE: If you just want to build FFTW, do not use this file. Just use # the ordinary ./configure && make commmands as described in the installation # section of the manual. # # This file is only for users that want to generate their own codelets, # as described in the "generating your own code" section of the manual. # ############################################################################ touch ChangeLog echo "PLEASE IGNORE WARNINGS AND ERRORS" # paranoia: sometimes autoconf doesn't get things right the first time rm -rf autom4te.cache autoreconf --verbose --install --symlink --force autoreconf --verbose --install --symlink --force autoreconf --verbose --install --symlink --force rm -f config.cache # --enable-maintainer-mode enables build of genfft and automatic # rebuild of codelets whenever genfft changes ( ./configure --disable-shared --enable-maintainer-mode --enable-threads $* ) fftw-3.3.3/CONVENTIONS0000644000175000017500000000435512054410050011135 00000000000000Code conventions used internally by fftw3 (not in API): LEARN FROM THE MASTERS: read Ken Thompson's C compiler in Plan 9. Avoid learning from C++/Java programs. INDENTATION: K&R, 5 spaces/tab. In case of doubt, indent -kr -i5. NAMES: keep them short. Shorter than you think. The Bible was written without vowels. Don't outsmart the Bible. Common names: R : real type, aka fftw_real E : real type for local variables (possibly extra precision) C : complex type sz : size vecsz : vector size is, os : input/output stride ri, ii : real/imag input (complex data) ro, io : real/imag output (complex data) I, O : real input/output (real data) A : assert CK : check S : solver, defined internally to each solver file P : plan, defined internally to each solver file k : codelet X(...) : used for mangling of external names (see below) K(...) : floating-point constant, in E precision If a name is used often and must have the form fftw_foo to avoid namespace pollution, #define FOO fftw_foo and use the short name. Leave that hungarian crap to MS. foo_t counts as hungarian: use foo instead. foo is lowercase so that it does not look like a DOS program. Exception: typedef struct foo_s {...} foo; instead of typedef struct foo {...} foo; for C++ compatibility. NAME MANGLING: use X(foo) for external names instead of fftw_foo. X(foo) expands to fftwf_foo or fftw_foo, depending on the precision. (Unfortunately, this is a ugly form of hungarian notation. Grrr...) Names that are not exported do not need to be mangled. REPEATED CODE: favor a table. E.g., do not write foo("xxx", 1); foo("yyy", 2); foo("zzz", -1); Instead write struct { const char *nam, int arg } footab[] = { { "xxx", 1 }, { "yyy", 2 }, { "zzz", -1 } }; and loop over footab. Rationale: it saves code space. Similarly, replace a switch statement with a table whenever possible. C++: The code should compile as a C++ program. Run the code through gcc -xc++ . The extra C++ restrictions are unnecessary, of course, but this will save us from a flood of complaints when we release the code. fftw-3.3.3/support/0000755000175000017500000000000012054411105011174 500000000000000fftw-3.3.3/support/Makefile.am0000644000175000017500000000014012054410050013141 00000000000000EXTRA_DIST = Makefile.codelets codelet_prelude.dft codelet_prelude.rdft \ addchain.c twovers.sh fftw-3.3.3/support/Makefile.in0000644000175000017500000002635612054410200013170 00000000000000# Makefile.in generated by automake 1.11.6 from Makefile.am. # @configure_input@ # Copyright (C) 1994, 1995, 1996, 1997, 1998, 1999, 2000, 2001, 2002, # 2003, 2004, 2005, 2006, 2007, 2008, 2009, 2010, 2011 Free Software # Foundation, Inc. # This Makefile.in is free software; the Free Software Foundation # gives unlimited permission to copy and/or distribute it, # with or without modifications, as long as this notice is preserved. # This program is distributed in the hope that it will be useful, # but WITHOUT ANY WARRANTY, to the extent permitted by law; without # even the implied warranty of MERCHANTABILITY or FITNESS FOR A # PARTICULAR PURPOSE. @SET_MAKE@ VPATH = @srcdir@ am__make_dryrun = \ { \ am__dry=no; \ case $$MAKEFLAGS in \ *\\[\ \ ]*) \ echo 'am--echo: ; @echo "AM" OK' | $(MAKE) -f - 2>/dev/null \ | grep '^AM OK$$' >/dev/null || am__dry=yes;; \ *) \ for am__flg in $$MAKEFLAGS; do \ case $$am__flg in \ *=*|--*) ;; \ *n*) am__dry=yes; break;; \ esac; \ done;; \ esac; \ test $$am__dry = yes; \ } pkgdatadir = $(datadir)/@PACKAGE@ pkgincludedir = $(includedir)/@PACKAGE@ pkglibdir = $(libdir)/@PACKAGE@ pkglibexecdir = $(libexecdir)/@PACKAGE@ am__cd = CDPATH="$${ZSH_VERSION+.}$(PATH_SEPARATOR)" && cd install_sh_DATA = $(install_sh) -c -m 644 install_sh_PROGRAM = $(install_sh) -c install_sh_SCRIPT = $(install_sh) -c INSTALL_HEADER = $(INSTALL_DATA) transform = $(program_transform_name) NORMAL_INSTALL = : PRE_INSTALL = : POST_INSTALL = : NORMAL_UNINSTALL = : PRE_UNINSTALL = : POST_UNINSTALL = : build_triplet = @build@ host_triplet = @host@ subdir = support DIST_COMMON = $(srcdir)/Makefile.am $(srcdir)/Makefile.in ACLOCAL_M4 = $(top_srcdir)/aclocal.m4 am__aclocal_m4_deps = $(top_srcdir)/m4/acx_mpi.m4 \ $(top_srcdir)/m4/acx_pthread.m4 \ $(top_srcdir)/m4/ax_cc_maxopt.m4 \ $(top_srcdir)/m4/ax_check_compiler_flags.m4 \ $(top_srcdir)/m4/ax_compiler_vendor.m4 \ $(top_srcdir)/m4/ax_gcc_aligns_stack.m4 \ $(top_srcdir)/m4/ax_gcc_version.m4 \ $(top_srcdir)/m4/ax_openmp.m4 $(top_srcdir)/m4/libtool.m4 \ $(top_srcdir)/m4/ltoptions.m4 $(top_srcdir)/m4/ltsugar.m4 \ $(top_srcdir)/m4/ltversion.m4 $(top_srcdir)/m4/lt~obsolete.m4 \ $(top_srcdir)/configure.ac am__configure_deps = $(am__aclocal_m4_deps) $(CONFIGURE_DEPENDENCIES) \ $(ACLOCAL_M4) mkinstalldirs = $(install_sh) -d CONFIG_HEADER = $(top_builddir)/config.h CONFIG_CLEAN_FILES = CONFIG_CLEAN_VPATH_FILES = SOURCES = DIST_SOURCES = am__can_run_installinfo = \ case $$AM_UPDATE_INFO_DIR in \ n|no|NO) false;; \ *) (install-info --version) >/dev/null 2>&1;; \ esac DISTFILES = $(DIST_COMMON) $(DIST_SOURCES) $(TEXINFOS) $(EXTRA_DIST) ACLOCAL = @ACLOCAL@ ALLOCA = @ALLOCA@ ALTIVEC_CFLAGS = @ALTIVEC_CFLAGS@ AMTAR = @AMTAR@ AR = @AR@ AS = @AS@ AUTOCONF = @AUTOCONF@ AUTOHEADER = @AUTOHEADER@ AUTOMAKE = @AUTOMAKE@ AVX_CFLAGS = @AVX_CFLAGS@ AWK = @AWK@ CC = @CC@ CCDEPMODE = @CCDEPMODE@ CFLAGS = @CFLAGS@ CHECK_PL_OPTS = @CHECK_PL_OPTS@ CPP = @CPP@ CPPFLAGS = @CPPFLAGS@ CYGPATH_W = @CYGPATH_W@ C_FFTW_R2R_KIND = @C_FFTW_R2R_KIND@ C_MPI_FINT = @C_MPI_FINT@ DEFS = @DEFS@ DEPDIR = @DEPDIR@ DLLTOOL = @DLLTOOL@ DSYMUTIL = @DSYMUTIL@ DUMPBIN = @DUMPBIN@ ECHO_C = @ECHO_C@ ECHO_N = @ECHO_N@ ECHO_T = @ECHO_T@ EGREP = @EGREP@ EXEEXT = @EXEEXT@ F77 = @F77@ FFLAGS = @FFLAGS@ FGREP = @FGREP@ FLIBS = @FLIBS@ GREP = @GREP@ INSTALL = @INSTALL@ INSTALL_DATA = @INSTALL_DATA@ INSTALL_PROGRAM = @INSTALL_PROGRAM@ INSTALL_SCRIPT = @INSTALL_SCRIPT@ INSTALL_STRIP_PROGRAM = @INSTALL_STRIP_PROGRAM@ LD = @LD@ LDFLAGS = @LDFLAGS@ LIBOBJS = @LIBOBJS@ LIBQUADMATH = @LIBQUADMATH@ LIBS = @LIBS@ LIBTOOL = @LIBTOOL@ LIPO = @LIPO@ LN_S = @LN_S@ LTLIBOBJS = @LTLIBOBJS@ MAINT = @MAINT@ MAKEINFO = @MAKEINFO@ MANIFEST_TOOL = @MANIFEST_TOOL@ MKDIR_P = @MKDIR_P@ MPICC = @MPICC@ MPILIBS = @MPILIBS@ MPIRUN = @MPIRUN@ NEON_CFLAGS = @NEON_CFLAGS@ NM = @NM@ NMEDIT = @NMEDIT@ OBJDUMP = @OBJDUMP@ OBJEXT = @OBJEXT@ OCAMLBUILD = @OCAMLBUILD@ OPENMP_CFLAGS = @OPENMP_CFLAGS@ OTOOL = @OTOOL@ OTOOL64 = @OTOOL64@ PACKAGE = @PACKAGE@ PACKAGE_BUGREPORT = @PACKAGE_BUGREPORT@ PACKAGE_NAME = @PACKAGE_NAME@ PACKAGE_STRING = @PACKAGE_STRING@ PACKAGE_TARNAME = @PACKAGE_TARNAME@ PACKAGE_URL = @PACKAGE_URL@ PACKAGE_VERSION = @PACKAGE_VERSION@ PATH_SEPARATOR = @PATH_SEPARATOR@ POW_LIB = @POW_LIB@ PRECISION = @PRECISION@ PREC_SUFFIX = @PREC_SUFFIX@ PTHREAD_CC = @PTHREAD_CC@ PTHREAD_CFLAGS = @PTHREAD_CFLAGS@ PTHREAD_LIBS = @PTHREAD_LIBS@ RANLIB = @RANLIB@ SED = @SED@ SET_MAKE = @SET_MAKE@ SHARED_VERSION_INFO = @SHARED_VERSION_INFO@ SHELL = @SHELL@ SSE2_CFLAGS = @SSE2_CFLAGS@ STACK_ALIGN_CFLAGS = @STACK_ALIGN_CFLAGS@ STRIP = @STRIP@ THREADLIBS = @THREADLIBS@ VERSION = @VERSION@ abs_builddir = @abs_builddir@ abs_srcdir = @abs_srcdir@ abs_top_builddir = @abs_top_builddir@ abs_top_srcdir = @abs_top_srcdir@ ac_ct_AR = @ac_ct_AR@ ac_ct_CC = @ac_ct_CC@ ac_ct_DUMPBIN = @ac_ct_DUMPBIN@ ac_ct_F77 = @ac_ct_F77@ acx_pthread_config = @acx_pthread_config@ am__include = @am__include@ am__leading_dot = @am__leading_dot@ am__quote = @am__quote@ am__tar = @am__tar@ am__untar = @am__untar@ bindir = @bindir@ build = @build@ build_alias = @build_alias@ build_cpu = @build_cpu@ build_os = @build_os@ build_vendor = @build_vendor@ builddir = @builddir@ datadir = @datadir@ datarootdir = @datarootdir@ docdir = @docdir@ dvidir = @dvidir@ exec_prefix = @exec_prefix@ host = @host@ host_alias = @host_alias@ host_cpu = @host_cpu@ host_os = @host_os@ host_vendor = @host_vendor@ htmldir = @htmldir@ includedir = @includedir@ infodir = @infodir@ install_sh = @install_sh@ libdir = @libdir@ libexecdir = @libexecdir@ localedir = @localedir@ localstatedir = @localstatedir@ mandir = @mandir@ mkdir_p = @mkdir_p@ oldincludedir = @oldincludedir@ pdfdir = @pdfdir@ prefix = @prefix@ program_transform_name = @program_transform_name@ psdir = @psdir@ sbindir = @sbindir@ sharedstatedir = @sharedstatedir@ srcdir = @srcdir@ sysconfdir = @sysconfdir@ target_alias = @target_alias@ top_build_prefix = @top_build_prefix@ top_builddir = @top_builddir@ top_srcdir = @top_srcdir@ EXTRA_DIST = Makefile.codelets codelet_prelude.dft codelet_prelude.rdft \ addchain.c twovers.sh all: all-am .SUFFIXES: $(srcdir)/Makefile.in: @MAINTAINER_MODE_TRUE@ $(srcdir)/Makefile.am $(am__configure_deps) @for dep in $?; do \ case '$(am__configure_deps)' in \ *$$dep*) \ ( cd $(top_builddir) && $(MAKE) $(AM_MAKEFLAGS) am--refresh ) \ && { if test -f $@; then exit 0; else break; fi; }; \ exit 1;; \ esac; \ done; \ echo ' cd $(top_srcdir) && $(AUTOMAKE) --gnu support/Makefile'; \ $(am__cd) $(top_srcdir) && \ $(AUTOMAKE) --gnu support/Makefile .PRECIOUS: Makefile Makefile: $(srcdir)/Makefile.in $(top_builddir)/config.status @case '$?' in \ *config.status*) \ cd $(top_builddir) && $(MAKE) $(AM_MAKEFLAGS) am--refresh;; \ *) \ echo ' cd $(top_builddir) && $(SHELL) ./config.status $(subdir)/$@ $(am__depfiles_maybe)'; \ cd $(top_builddir) && $(SHELL) ./config.status $(subdir)/$@ $(am__depfiles_maybe);; \ esac; $(top_builddir)/config.status: $(top_srcdir)/configure $(CONFIG_STATUS_DEPENDENCIES) cd $(top_builddir) && $(MAKE) $(AM_MAKEFLAGS) am--refresh $(top_srcdir)/configure: @MAINTAINER_MODE_TRUE@ $(am__configure_deps) cd $(top_builddir) && $(MAKE) $(AM_MAKEFLAGS) am--refresh $(ACLOCAL_M4): @MAINTAINER_MODE_TRUE@ $(am__aclocal_m4_deps) cd $(top_builddir) && $(MAKE) $(AM_MAKEFLAGS) am--refresh $(am__aclocal_m4_deps): mostlyclean-libtool: -rm -f *.lo clean-libtool: -rm -rf .libs _libs tags: TAGS TAGS: ctags: CTAGS CTAGS: distdir: $(DISTFILES) @srcdirstrip=`echo "$(srcdir)" | sed 's/[].[^$$\\*]/\\\\&/g'`; \ topsrcdirstrip=`echo "$(top_srcdir)" | sed 's/[].[^$$\\*]/\\\\&/g'`; \ list='$(DISTFILES)'; \ dist_files=`for file in $$list; do echo $$file; done | \ sed -e "s|^$$srcdirstrip/||;t" \ -e "s|^$$topsrcdirstrip/|$(top_builddir)/|;t"`; \ case $$dist_files in \ */*) $(MKDIR_P) `echo "$$dist_files" | \ sed '/\//!d;s|^|$(distdir)/|;s,/[^/]*$$,,' | \ sort -u` ;; \ esac; \ for file in $$dist_files; do \ if test -f $$file || test -d $$file; then d=.; else d=$(srcdir); fi; \ if test -d $$d/$$file; then \ dir=`echo "/$$file" | sed -e 's,/[^/]*$$,,'`; \ if test -d "$(distdir)/$$file"; then \ find "$(distdir)/$$file" -type d ! -perm -700 -exec chmod u+rwx {} \;; \ fi; \ if test -d $(srcdir)/$$file && test $$d != $(srcdir); then \ cp -fpR $(srcdir)/$$file "$(distdir)$$dir" || exit 1; \ find "$(distdir)/$$file" -type d ! -perm -700 -exec chmod u+rwx {} \;; \ fi; \ cp -fpR $$d/$$file "$(distdir)$$dir" || exit 1; \ else \ test -f "$(distdir)/$$file" \ || cp -p $$d/$$file "$(distdir)/$$file" \ || exit 1; \ fi; \ done check-am: all-am check: check-am all-am: Makefile installdirs: install: install-am install-exec: install-exec-am install-data: install-data-am uninstall: uninstall-am install-am: all-am @$(MAKE) $(AM_MAKEFLAGS) install-exec-am install-data-am installcheck: installcheck-am install-strip: if test -z '$(STRIP)'; then \ $(MAKE) $(AM_MAKEFLAGS) INSTALL_PROGRAM="$(INSTALL_STRIP_PROGRAM)" \ install_sh_PROGRAM="$(INSTALL_STRIP_PROGRAM)" INSTALL_STRIP_FLAG=-s \ install; \ else \ $(MAKE) $(AM_MAKEFLAGS) INSTALL_PROGRAM="$(INSTALL_STRIP_PROGRAM)" \ install_sh_PROGRAM="$(INSTALL_STRIP_PROGRAM)" INSTALL_STRIP_FLAG=-s \ "INSTALL_PROGRAM_ENV=STRIPPROG='$(STRIP)'" install; \ fi mostlyclean-generic: clean-generic: distclean-generic: -test -z "$(CONFIG_CLEAN_FILES)" || rm -f $(CONFIG_CLEAN_FILES) -test . = "$(srcdir)" || test -z "$(CONFIG_CLEAN_VPATH_FILES)" || rm -f $(CONFIG_CLEAN_VPATH_FILES) maintainer-clean-generic: @echo "This command is intended for maintainers to use" @echo "it deletes files that may require special tools to rebuild." clean: clean-am clean-am: clean-generic clean-libtool mostlyclean-am distclean: distclean-am -rm -f Makefile distclean-am: clean-am distclean-generic dvi: dvi-am dvi-am: html: html-am html-am: info: info-am info-am: install-data-am: install-dvi: install-dvi-am install-dvi-am: install-exec-am: install-html: install-html-am install-html-am: install-info: install-info-am install-info-am: install-man: install-pdf: install-pdf-am install-pdf-am: install-ps: install-ps-am install-ps-am: installcheck-am: maintainer-clean: maintainer-clean-am -rm -f Makefile maintainer-clean-am: distclean-am maintainer-clean-generic mostlyclean: mostlyclean-am mostlyclean-am: mostlyclean-generic mostlyclean-libtool pdf: pdf-am pdf-am: ps: ps-am ps-am: uninstall-am: .MAKE: install-am install-strip .PHONY: all all-am check check-am clean clean-generic clean-libtool \ distclean distclean-generic distclean-libtool distdir dvi \ dvi-am html html-am info info-am install install-am \ install-data install-data-am install-dvi install-dvi-am \ install-exec install-exec-am install-html install-html-am \ install-info install-info-am install-man install-pdf \ install-pdf-am install-ps install-ps-am install-strip \ installcheck installcheck-am installdirs maintainer-clean \ maintainer-clean-generic mostlyclean mostlyclean-generic \ mostlyclean-libtool pdf pdf-am ps ps-am uninstall uninstall-am # Tell versions [3.59,3.63) of GNU make to not export all variables. # Otherwise a system limit (for SysV at least) may be exceeded. .NOEXPORT: fftw-3.3.3/support/Makefile.codelets0000644000175000017500000000474412054410050014364 00000000000000# -*- makefile -*- # This file contains special make rules to generate codelets. # Most of this file requires GNU make . CODLIST = codlist.c CODELET_NAME=codelet_ # rule to build codlist $(CODLIST): Makefile ( \ echo "#include \"ifftw.h\""; \ echo $(INCLUDE_SIMD_HEADER); \ echo; \ for i in $(ALL_CODELETS) NIL; do \ if test "$$i" != NIL; then \ j=`basename $$i | sed -e 's/[.][cS]$$//g'`; \ echo "extern void $(XRENAME)($(CODELET_NAME)$$j)(planner *);"; \ fi \ done; \ echo; \ echo; \ echo "extern const solvtab $(SOLVTAB_NAME);"; \ echo "const solvtab $(SOLVTAB_NAME) = {"; \ for i in $(ALL_CODELETS) NIL; do \ if test "$$i" != NIL; then \ j=`basename $$i | sed -e 's/[.][cS]$$//g'`; \ echo " SOLVTAB($(XRENAME)($(CODELET_NAME)$$j)),"; \ fi \ done; \ echo " SOLVTAB_END"; \ echo "};"; \ ) >$@ # only delete codlist.c in maintainer-mode, since it is included in the dist # FIXME: is there a way to delete in 'make clean' only when builddir != srcdir? maintainer-clean-local: rm -f $(CODLIST) if MAINTAINER_MODE INDENT = indent -kr -cs -i5 -l800 -fca -nfc1 -sc -sob -cli4 -TR -Tplanner -TV TWOVERS = sh ${top_srcdir}/support/twovers.sh GENFFTDIR = ${top_builddir}/genfft GEN_NOTW = ${GENFFTDIR}/gen_notw.native GEN_NOTW_C = ${GENFFTDIR}/gen_notw_c.native GEN_TWIDDLE = ${GENFFTDIR}/gen_twiddle.native GEN_TWIDDLE_C = ${GENFFTDIR}/gen_twiddle_c.native GEN_TWIDSQ = ${GENFFTDIR}/gen_twidsq.native GEN_TWIDSQ_C = ${GENFFTDIR}/gen_twidsq_c.native GEN_R2CF = ${GENFFTDIR}/gen_r2cf.native GEN_R2CB = ${GENFFTDIR}/gen_r2cb.native GEN_HC2HC = ${GENFFTDIR}/gen_hc2hc.native GEN_HC2C = ${GENFFTDIR}/gen_hc2c.native GEN_HC2CDFT = ${GENFFTDIR}/gen_hc2cdft.native GEN_HC2CDFT_C = ${GENFFTDIR}/gen_hc2cdft_c.native GEN_R2R = ${GENFFTDIR}/gen_r2r.native PRELUDE_DFT = ${top_srcdir}/support/codelet_prelude.dft PRELUDE_RDFT = ${top_srcdir}/support/codelet_prelude.rdft ADD_DATE = sed -e s/@DATE@/"`date`"/ COPYRIGHT=${top_srcdir}/COPYRIGHT CODELET_DEPS=$(COPYRIGHT) $(PRELUDE) PRELUDE_COMMANDS_DFT=cat $(COPYRIGHT) $(PRELUDE_DFT) PRELUDE_COMMANDS_RDFT=cat $(COPYRIGHT) $(PRELUDE_RDFT) FLAGS_COMMON = -compact -variables 4 DFT_FLAGS_COMMON = $(FLAGS_COMMON) -pipeline-latency 4 RDFT_FLAGS_COMMON = $(FLAGS_COMMON) -pipeline-latency 4 # cancel the hideous builtin rules that cause an infinite loop %: %.o %: %.s %: %.c %: %.S endif # MAINTAINER_MODE fftw-3.3.3/support/codelet_prelude.dft0000644000175000017500000000016412054410050014751 00000000000000 /* This file was automatically generated --- DO NOT EDIT */ /* Generated on @DATE@ */ #include "codelet-dft.h" fftw-3.3.3/support/codelet_prelude.rdft0000644000175000017500000000016512054410050015134 00000000000000 /* This file was automatically generated --- DO NOT EDIT */ /* Generated on @DATE@ */ #include "codelet-rdft.h" fftw-3.3.3/support/addchain.c0000644000175000017500000000617712054410050013024 00000000000000/* addition-chain optimizer */ #include #include #include static int verbose; static int mulcost = 18; static int ldcost = 2; static int sqcost = 10; static int reflcost = 8; #define INFTY 100000 static int *answer; static int best_so_far; static void print_answer(int n, int t) { int i; printf("| (%d, %d) -> [", n, t); for (i = 0; i < t; ++i) printf("%d;", answer[i]); printf("] (* %d *)\n", best_so_far); } #define DO(i, j, k, cst) \ if (k < n) { \ int c = A[i] + A[j] + cst; \ if (c < A[k]) { \ A[k] = c; \ changed = 1; \ } \ } #define DO3(i, j, l, k, cst) \ if (k < n) { \ int c = A[i] + A[j] + A[l] + cst; \ if (c < A[k]) { \ A[k] = c; \ changed = 1; \ } \ } static int optimize(int n, int *A) { int i, j, k, changed, cst, cstmax; do { changed = 0; for (i = 0; i < n; ++i) { k = i + i; DO(i, i, k, sqcost); } for (i = 0; i < n; ++i) { for (j = 0; j <= i; ++j) { k = i + j; DO(i, j, k, mulcost); k = i - j; DO(i, j, k, mulcost); k = i + j; DO3(i, j, i - j, k, reflcost); } } } while (changed); cst = cstmax = 0; for (i = 0; i < n; ++i) { cst += A[i]; if (A[i] > cstmax) cstmax = A[i]; } /* return cstmax; */ return cst; } static void search(int n, int t, int *A, int *B, int depth) { if (depth == 0) { int i, tc; for (i = 0; i < n; ++i) A[i] = INFTY; A[0] = 0; /* always free */ for (i = 1; i <= t; ++i) A[B[-i]] = ldcost; tc = optimize(n, A); if (tc < best_so_far) { best_so_far = tc; for (i = 1; i <= t; ++i) answer[t - i] = B[-i]; if (verbose) print_answer(n, t); } } else { for (B[0] = B[-1] + 1; B[0] < n; ++B[0]) search(n, t, A, B + 1, depth - 1); } } static void doit(int n, int t) { int *A; int *B; A = malloc(n * sizeof(int)); B = malloc((t + 1) * sizeof(int)); answer = malloc(t * sizeof(int)); B[0] = 0; best_so_far = INFTY; search(n, t, A, B + 1, t); print_answer(n, t); free(A); free(B); free(answer); } int main(int argc, char *argv[]) { int n = 32; int t = 3; int all; int ch; verbose = 0; all = 0; while ((ch = getopt(argc, argv, "n:t:m:l:r:s:va")) != -1) { switch (ch) { case 'n': n = atoi(optarg); break; case 't': t = atoi(optarg); break; case 'm': mulcost = atoi(optarg); break; case 'l': ldcost = atoi(optarg); break; case 's': sqcost = atoi(optarg); break; case 'r': reflcost = atoi(optarg); break; case 'v': ++verbose; break; case 'a': ++all; break; case '?': fprintf(stderr, "use the source\n"); exit(1); } } if (all) { for (n = 4; n <= 64; n *= 2) { int n1 = n - 1; if (n1 > 7) n1 = 7; for (t = 1; t <= n1; ++t) doit(n, t); } } else { doit(n, t); } return 0; } fftw-3.3.3/support/twovers.sh0000644000175000017500000000041212054410050013154 00000000000000#! /bin/sh # wrapper to generate two codelet versions, with and without # fma genfft=$1 shift echo "#ifdef HAVE_FMA" echo $genfft -fma -reorder-insns -schedule-for-pipeline $* echo echo "#else /* HAVE_FMA */" echo $genfft $* echo echo "#endif /* HAVE_FMA */" fftw-3.3.3/genfft/0000755000175000017500000000000012054411106010732 500000000000000fftw-3.3.3/genfft/Makefile.am0000644000175000017500000000216612054410050012710 00000000000000# this makefile requires GNU make. EXTRA_DIST = algsimp.ml annotate.ml assoctable.ml c.ml complex.ml \ conv.ml dag.ml expr.ml fft.ml gen_hc2c.ml gen_hc2cdft.ml \ gen_hc2cdft_c.ml gen_hc2hc.ml gen_r2cb.ml gen_mdct.ml gen_notw.ml \ gen_notw_c.ml gen_r2cf.ml gen_r2r.ml gen_twiddle.ml gen_twiddle_c.ml \ gen_twidsq.ml gen_twidsq_c.ml genutil.ml littlesimp.ml magic.ml \ monads.ml number.ml oracle.ml schedule.ml simd.ml simdmagic.ml \ to_alist.ml trig.ml twiddle.ml unique.ml util.ml variable.ml \ algsimp.mli annotate.mli assoctable.mli c.mli complex.mli conv.mli \ dag.mli expr.mli fft.mli littlesimp.mli number.mli oracle.mli \ schedule.mli simd.mli to_alist.mli trig.mli twiddle.mli unique.mli \ util.mli variable.mli GENFFT_NATIVE=gen_notw.native gen_notw_c.native gen_twiddle.native \ gen_twiddle_c.native gen_twidsq.native gen_twidsq_c.native \ gen_r2r.native gen_r2cf.native gen_r2cb.native gen_hc2c.native \ gen_hc2cdft.native gen_hc2cdft_c.native gen_hc2hc.native \ gen_mdct.native all-local:: $(OCAMLBUILD) -classic-display -libs unix,nums $(GENFFT_NATIVE) maintainer-clean-local:: $(OCAMLBUILD) -classic-display -clean fftw-3.3.3/genfft/Makefile.in0000644000175000017500000003051312054410200012713 00000000000000# Makefile.in generated by automake 1.11.6 from Makefile.am. # @configure_input@ # Copyright (C) 1994, 1995, 1996, 1997, 1998, 1999, 2000, 2001, 2002, # 2003, 2004, 2005, 2006, 2007, 2008, 2009, 2010, 2011 Free Software # Foundation, Inc. # This Makefile.in is free software; the Free Software Foundation # gives unlimited permission to copy and/or distribute it, # with or without modifications, as long as this notice is preserved. # This program is distributed in the hope that it will be useful, # but WITHOUT ANY WARRANTY, to the extent permitted by law; without # even the implied warranty of MERCHANTABILITY or FITNESS FOR A # PARTICULAR PURPOSE. @SET_MAKE@ # this makefile requires GNU make. VPATH = @srcdir@ am__make_dryrun = \ { \ am__dry=no; \ case $$MAKEFLAGS in \ *\\[\ \ ]*) \ echo 'am--echo: ; @echo "AM" OK' | $(MAKE) -f - 2>/dev/null \ | grep '^AM OK$$' >/dev/null || am__dry=yes;; \ *) \ for am__flg in $$MAKEFLAGS; do \ case $$am__flg in \ *=*|--*) ;; \ *n*) am__dry=yes; break;; \ esac; \ done;; \ esac; \ test $$am__dry = yes; \ } pkgdatadir = $(datadir)/@PACKAGE@ pkgincludedir = $(includedir)/@PACKAGE@ pkglibdir = $(libdir)/@PACKAGE@ pkglibexecdir = $(libexecdir)/@PACKAGE@ am__cd = CDPATH="$${ZSH_VERSION+.}$(PATH_SEPARATOR)" && cd install_sh_DATA = $(install_sh) -c -m 644 install_sh_PROGRAM = $(install_sh) -c install_sh_SCRIPT = $(install_sh) -c INSTALL_HEADER = $(INSTALL_DATA) transform = $(program_transform_name) NORMAL_INSTALL = : PRE_INSTALL = : POST_INSTALL = : NORMAL_UNINSTALL = : PRE_UNINSTALL = : POST_UNINSTALL = : build_triplet = @build@ host_triplet = @host@ subdir = genfft DIST_COMMON = $(srcdir)/Makefile.am $(srcdir)/Makefile.in ACLOCAL_M4 = $(top_srcdir)/aclocal.m4 am__aclocal_m4_deps = $(top_srcdir)/m4/acx_mpi.m4 \ $(top_srcdir)/m4/acx_pthread.m4 \ $(top_srcdir)/m4/ax_cc_maxopt.m4 \ $(top_srcdir)/m4/ax_check_compiler_flags.m4 \ $(top_srcdir)/m4/ax_compiler_vendor.m4 \ $(top_srcdir)/m4/ax_gcc_aligns_stack.m4 \ $(top_srcdir)/m4/ax_gcc_version.m4 \ $(top_srcdir)/m4/ax_openmp.m4 $(top_srcdir)/m4/libtool.m4 \ $(top_srcdir)/m4/ltoptions.m4 $(top_srcdir)/m4/ltsugar.m4 \ $(top_srcdir)/m4/ltversion.m4 $(top_srcdir)/m4/lt~obsolete.m4 \ $(top_srcdir)/configure.ac am__configure_deps = $(am__aclocal_m4_deps) $(CONFIGURE_DEPENDENCIES) \ $(ACLOCAL_M4) mkinstalldirs = $(install_sh) -d CONFIG_HEADER = $(top_builddir)/config.h CONFIG_CLEAN_FILES = CONFIG_CLEAN_VPATH_FILES = SOURCES = DIST_SOURCES = am__can_run_installinfo = \ case $$AM_UPDATE_INFO_DIR in \ n|no|NO) false;; \ *) (install-info --version) >/dev/null 2>&1;; \ esac DISTFILES = $(DIST_COMMON) $(DIST_SOURCES) $(TEXINFOS) $(EXTRA_DIST) ACLOCAL = @ACLOCAL@ ALLOCA = @ALLOCA@ ALTIVEC_CFLAGS = @ALTIVEC_CFLAGS@ AMTAR = @AMTAR@ AR = @AR@ AS = @AS@ AUTOCONF = @AUTOCONF@ AUTOHEADER = @AUTOHEADER@ AUTOMAKE = @AUTOMAKE@ AVX_CFLAGS = @AVX_CFLAGS@ AWK = @AWK@ CC = @CC@ CCDEPMODE = @CCDEPMODE@ CFLAGS = @CFLAGS@ CHECK_PL_OPTS = @CHECK_PL_OPTS@ CPP = @CPP@ CPPFLAGS = @CPPFLAGS@ CYGPATH_W = @CYGPATH_W@ C_FFTW_R2R_KIND = @C_FFTW_R2R_KIND@ C_MPI_FINT = @C_MPI_FINT@ DEFS = @DEFS@ DEPDIR = @DEPDIR@ DLLTOOL = @DLLTOOL@ DSYMUTIL = @DSYMUTIL@ DUMPBIN = @DUMPBIN@ ECHO_C = @ECHO_C@ ECHO_N = @ECHO_N@ ECHO_T = @ECHO_T@ EGREP = @EGREP@ EXEEXT = @EXEEXT@ F77 = @F77@ FFLAGS = @FFLAGS@ FGREP = @FGREP@ FLIBS = @FLIBS@ GREP = @GREP@ INSTALL = @INSTALL@ INSTALL_DATA = @INSTALL_DATA@ INSTALL_PROGRAM = @INSTALL_PROGRAM@ INSTALL_SCRIPT = @INSTALL_SCRIPT@ INSTALL_STRIP_PROGRAM = @INSTALL_STRIP_PROGRAM@ LD = @LD@ LDFLAGS = @LDFLAGS@ LIBOBJS = @LIBOBJS@ LIBQUADMATH = @LIBQUADMATH@ LIBS = @LIBS@ LIBTOOL = @LIBTOOL@ LIPO = @LIPO@ LN_S = @LN_S@ LTLIBOBJS = @LTLIBOBJS@ MAINT = @MAINT@ MAKEINFO = @MAKEINFO@ MANIFEST_TOOL = @MANIFEST_TOOL@ MKDIR_P = @MKDIR_P@ MPICC = @MPICC@ MPILIBS = @MPILIBS@ MPIRUN = @MPIRUN@ NEON_CFLAGS = @NEON_CFLAGS@ NM = @NM@ NMEDIT = @NMEDIT@ OBJDUMP = @OBJDUMP@ OBJEXT = @OBJEXT@ OCAMLBUILD = @OCAMLBUILD@ OPENMP_CFLAGS = @OPENMP_CFLAGS@ OTOOL = @OTOOL@ OTOOL64 = @OTOOL64@ PACKAGE = @PACKAGE@ PACKAGE_BUGREPORT = @PACKAGE_BUGREPORT@ PACKAGE_NAME = @PACKAGE_NAME@ PACKAGE_STRING = @PACKAGE_STRING@ PACKAGE_TARNAME = @PACKAGE_TARNAME@ PACKAGE_URL = @PACKAGE_URL@ PACKAGE_VERSION = @PACKAGE_VERSION@ PATH_SEPARATOR = @PATH_SEPARATOR@ POW_LIB = @POW_LIB@ PRECISION = @PRECISION@ PREC_SUFFIX = @PREC_SUFFIX@ PTHREAD_CC = @PTHREAD_CC@ PTHREAD_CFLAGS = @PTHREAD_CFLAGS@ PTHREAD_LIBS = @PTHREAD_LIBS@ RANLIB = @RANLIB@ SED = @SED@ SET_MAKE = @SET_MAKE@ SHARED_VERSION_INFO = @SHARED_VERSION_INFO@ SHELL = @SHELL@ SSE2_CFLAGS = @SSE2_CFLAGS@ STACK_ALIGN_CFLAGS = @STACK_ALIGN_CFLAGS@ STRIP = @STRIP@ THREADLIBS = @THREADLIBS@ VERSION = @VERSION@ abs_builddir = @abs_builddir@ abs_srcdir = @abs_srcdir@ abs_top_builddir = @abs_top_builddir@ abs_top_srcdir = @abs_top_srcdir@ ac_ct_AR = @ac_ct_AR@ ac_ct_CC = @ac_ct_CC@ ac_ct_DUMPBIN = @ac_ct_DUMPBIN@ ac_ct_F77 = @ac_ct_F77@ acx_pthread_config = @acx_pthread_config@ am__include = @am__include@ am__leading_dot = @am__leading_dot@ am__quote = @am__quote@ am__tar = @am__tar@ am__untar = @am__untar@ bindir = @bindir@ build = @build@ build_alias = @build_alias@ build_cpu = @build_cpu@ build_os = @build_os@ build_vendor = @build_vendor@ builddir = @builddir@ datadir = @datadir@ datarootdir = @datarootdir@ docdir = @docdir@ dvidir = @dvidir@ exec_prefix = @exec_prefix@ host = @host@ host_alias = @host_alias@ host_cpu = @host_cpu@ host_os = @host_os@ host_vendor = @host_vendor@ htmldir = @htmldir@ includedir = @includedir@ infodir = @infodir@ install_sh = @install_sh@ libdir = @libdir@ libexecdir = @libexecdir@ localedir = @localedir@ localstatedir = @localstatedir@ mandir = @mandir@ mkdir_p = @mkdir_p@ oldincludedir = @oldincludedir@ pdfdir = @pdfdir@ prefix = @prefix@ program_transform_name = @program_transform_name@ psdir = @psdir@ sbindir = @sbindir@ sharedstatedir = @sharedstatedir@ srcdir = @srcdir@ sysconfdir = @sysconfdir@ target_alias = @target_alias@ top_build_prefix = @top_build_prefix@ top_builddir = @top_builddir@ top_srcdir = @top_srcdir@ EXTRA_DIST = algsimp.ml annotate.ml assoctable.ml c.ml complex.ml \ conv.ml dag.ml expr.ml fft.ml gen_hc2c.ml gen_hc2cdft.ml \ gen_hc2cdft_c.ml gen_hc2hc.ml gen_r2cb.ml gen_mdct.ml gen_notw.ml \ gen_notw_c.ml gen_r2cf.ml gen_r2r.ml gen_twiddle.ml gen_twiddle_c.ml \ gen_twidsq.ml gen_twidsq_c.ml genutil.ml littlesimp.ml magic.ml \ monads.ml number.ml oracle.ml schedule.ml simd.ml simdmagic.ml \ to_alist.ml trig.ml twiddle.ml unique.ml util.ml variable.ml \ algsimp.mli annotate.mli assoctable.mli c.mli complex.mli conv.mli \ dag.mli expr.mli fft.mli littlesimp.mli number.mli oracle.mli \ schedule.mli simd.mli to_alist.mli trig.mli twiddle.mli unique.mli \ util.mli variable.mli GENFFT_NATIVE = gen_notw.native gen_notw_c.native gen_twiddle.native \ gen_twiddle_c.native gen_twidsq.native gen_twidsq_c.native \ gen_r2r.native gen_r2cf.native gen_r2cb.native gen_hc2c.native \ gen_hc2cdft.native gen_hc2cdft_c.native gen_hc2hc.native \ gen_mdct.native all: all-am .SUFFIXES: $(srcdir)/Makefile.in: @MAINTAINER_MODE_TRUE@ $(srcdir)/Makefile.am $(am__configure_deps) @for dep in $?; do \ case '$(am__configure_deps)' in \ *$$dep*) \ ( cd $(top_builddir) && $(MAKE) $(AM_MAKEFLAGS) am--refresh ) \ && { if test -f $@; then exit 0; else break; fi; }; \ exit 1;; \ esac; \ done; \ echo ' cd $(top_srcdir) && $(AUTOMAKE) --gnu genfft/Makefile'; \ $(am__cd) $(top_srcdir) && \ $(AUTOMAKE) --gnu genfft/Makefile .PRECIOUS: Makefile Makefile: $(srcdir)/Makefile.in $(top_builddir)/config.status @case '$?' in \ *config.status*) \ cd $(top_builddir) && $(MAKE) $(AM_MAKEFLAGS) am--refresh;; \ *) \ echo ' cd $(top_builddir) && $(SHELL) ./config.status $(subdir)/$@ $(am__depfiles_maybe)'; \ cd $(top_builddir) && $(SHELL) ./config.status $(subdir)/$@ $(am__depfiles_maybe);; \ esac; $(top_builddir)/config.status: $(top_srcdir)/configure $(CONFIG_STATUS_DEPENDENCIES) cd $(top_builddir) && $(MAKE) $(AM_MAKEFLAGS) am--refresh $(top_srcdir)/configure: @MAINTAINER_MODE_TRUE@ $(am__configure_deps) cd $(top_builddir) && $(MAKE) $(AM_MAKEFLAGS) am--refresh $(ACLOCAL_M4): @MAINTAINER_MODE_TRUE@ $(am__aclocal_m4_deps) cd $(top_builddir) && $(MAKE) $(AM_MAKEFLAGS) am--refresh $(am__aclocal_m4_deps): mostlyclean-libtool: -rm -f *.lo clean-libtool: -rm -rf .libs _libs tags: TAGS TAGS: ctags: CTAGS CTAGS: distdir: $(DISTFILES) @srcdirstrip=`echo "$(srcdir)" | sed 's/[].[^$$\\*]/\\\\&/g'`; \ topsrcdirstrip=`echo "$(top_srcdir)" | sed 's/[].[^$$\\*]/\\\\&/g'`; \ list='$(DISTFILES)'; \ dist_files=`for file in $$list; do echo $$file; done | \ sed -e "s|^$$srcdirstrip/||;t" \ -e "s|^$$topsrcdirstrip/|$(top_builddir)/|;t"`; \ case $$dist_files in \ */*) $(MKDIR_P) `echo "$$dist_files" | \ sed '/\//!d;s|^|$(distdir)/|;s,/[^/]*$$,,' | \ sort -u` ;; \ esac; \ for file in $$dist_files; do \ if test -f $$file || test -d $$file; then d=.; else d=$(srcdir); fi; \ if test -d $$d/$$file; then \ dir=`echo "/$$file" | sed -e 's,/[^/]*$$,,'`; \ if test -d "$(distdir)/$$file"; then \ find "$(distdir)/$$file" -type d ! -perm -700 -exec chmod u+rwx {} \;; \ fi; \ if test -d $(srcdir)/$$file && test $$d != $(srcdir); then \ cp -fpR $(srcdir)/$$file "$(distdir)$$dir" || exit 1; \ find "$(distdir)/$$file" -type d ! -perm -700 -exec chmod u+rwx {} \;; \ fi; \ cp -fpR $$d/$$file "$(distdir)$$dir" || exit 1; \ else \ test -f "$(distdir)/$$file" \ || cp -p $$d/$$file "$(distdir)/$$file" \ || exit 1; \ fi; \ done check-am: all-am check: check-am all-am: Makefile all-local installdirs: install: install-am install-exec: install-exec-am install-data: install-data-am uninstall: uninstall-am install-am: all-am @$(MAKE) $(AM_MAKEFLAGS) install-exec-am install-data-am installcheck: installcheck-am install-strip: if test -z '$(STRIP)'; then \ $(MAKE) $(AM_MAKEFLAGS) INSTALL_PROGRAM="$(INSTALL_STRIP_PROGRAM)" \ install_sh_PROGRAM="$(INSTALL_STRIP_PROGRAM)" INSTALL_STRIP_FLAG=-s \ install; \ else \ $(MAKE) $(AM_MAKEFLAGS) INSTALL_PROGRAM="$(INSTALL_STRIP_PROGRAM)" \ install_sh_PROGRAM="$(INSTALL_STRIP_PROGRAM)" INSTALL_STRIP_FLAG=-s \ "INSTALL_PROGRAM_ENV=STRIPPROG='$(STRIP)'" install; \ fi mostlyclean-generic: clean-generic: distclean-generic: -test -z "$(CONFIG_CLEAN_FILES)" || rm -f $(CONFIG_CLEAN_FILES) -test . = "$(srcdir)" || test -z "$(CONFIG_CLEAN_VPATH_FILES)" || rm -f $(CONFIG_CLEAN_VPATH_FILES) maintainer-clean-generic: @echo "This command is intended for maintainers to use" @echo "it deletes files that may require special tools to rebuild." clean: clean-am clean-am: clean-generic clean-libtool mostlyclean-am distclean: distclean-am -rm -f Makefile distclean-am: clean-am distclean-generic dvi: dvi-am dvi-am: html: html-am html-am: info: info-am info-am: install-data-am: install-dvi: install-dvi-am install-dvi-am: install-exec-am: install-html: install-html-am install-html-am: install-info: install-info-am install-info-am: install-man: install-pdf: install-pdf-am install-pdf-am: install-ps: install-ps-am install-ps-am: installcheck-am: maintainer-clean: maintainer-clean-am -rm -f Makefile maintainer-clean-am: distclean-am maintainer-clean-generic \ maintainer-clean-local mostlyclean: mostlyclean-am mostlyclean-am: mostlyclean-generic mostlyclean-libtool pdf: pdf-am pdf-am: ps: ps-am ps-am: uninstall-am: .MAKE: install-am install-strip .PHONY: all all-am all-local check check-am clean clean-generic \ clean-libtool distclean distclean-generic distclean-libtool \ distdir dvi dvi-am html html-am info info-am install \ install-am install-data install-data-am install-dvi \ install-dvi-am install-exec install-exec-am install-html \ install-html-am install-info install-info-am install-man \ install-pdf install-pdf-am install-ps install-ps-am \ install-strip installcheck installcheck-am installdirs \ maintainer-clean maintainer-clean-generic \ maintainer-clean-local mostlyclean mostlyclean-generic \ mostlyclean-libtool pdf pdf-am ps ps-am uninstall uninstall-am all-local:: $(OCAMLBUILD) -classic-display -libs unix,nums $(GENFFT_NATIVE) maintainer-clean-local:: $(OCAMLBUILD) -classic-display -clean # Tell versions [3.59,3.63) of GNU make to not export all variables. # Otherwise a system limit (for SysV at least) may be exceeded. .NOEXPORT: fftw-3.3.3/genfft/algsimp.ml0000644000175000017500000004426312054410050012646 00000000000000(* * Copyright (c) 1997-1999 Massachusetts Institute of Technology * Copyright (c) 2003, 2007-11 Matteo Frigo * Copyright (c) 2003, 2007-11 Massachusetts Institute of Technology * * This program is free software; you can redistribute it and/or modify * it under the terms of the GNU General Public License as published by * the Free Software Foundation; either version 2 of the License, or * (at your option) any later version. * * This program is distributed in the hope that it will be useful, * but WITHOUT ANY WARRANTY; without even the implied warranty of * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the * GNU General Public License for more details. * * You should have received a copy of the GNU General Public License * along with this program; if not, write to the Free Software * Foundation, Inc., 51 Franklin Street, Fifth Floor, Boston, MA 02110-1301 USA * *) open Util open Expr let node_insert x = Assoctable.insert Expr.hash x let node_lookup x = Assoctable.lookup Expr.hash (==) x (************************************************************* * Algebraic simplifier/elimination of common subexpressions *************************************************************) module AlgSimp : sig val algsimp : expr list -> expr list end = struct open Monads.StateMonad open Monads.MemoMonad open Assoctable let fetchSimp = fetchState >>= fun (s, _) -> returnM s let storeSimp s = fetchState >>= (fun (_, c) -> storeState (s, c)) let lookupSimpM key = fetchSimp >>= fun table -> returnM (node_lookup key table) let insertSimpM key value = fetchSimp >>= fun table -> storeSimp (node_insert key value table) let subset a b = List.for_all (fun x -> List.exists (fun y -> x == y) b) a let structurallyEqualCSE a b = match (a, b) with | (Num a, Num b) -> Number.equal a b | (NaN a, NaN b) -> a == b | (Load a, Load b) -> Variable.same a b | (Times (a, a'), Times (b, b')) -> ((a == b) && (a' == b')) or ((a == b') && (a' == b)) | (CTimes (a, a'), CTimes (b, b')) -> ((a == b) && (a' == b')) or ((a == b') && (a' == b)) | (CTimesJ (a, a'), CTimesJ (b, b')) -> ((a == b) && (a' == b')) | (Plus a, Plus b) -> subset a b && subset b a | (Uminus a, Uminus b) -> (a == b) | _ -> false let hashCSE x = if (!Magic.randomized_cse) then Oracle.hash x else Expr.hash x let equalCSE a b = if (!Magic.randomized_cse) then (structurallyEqualCSE a b || Oracle.likely_equal a b) else structurallyEqualCSE a b let fetchCSE = fetchState >>= fun (_, c) -> returnM c let storeCSE c = fetchState >>= (fun (s, _) -> storeState (s, c)) let lookupCSEM key = fetchCSE >>= fun table -> returnM (Assoctable.lookup hashCSE equalCSE key table) let insertCSEM key value = fetchCSE >>= fun table -> storeCSE (Assoctable.insert hashCSE key value table) (* memoize both x and Uminus x (unless x is already negated) *) let identityM x = let memo x = memoizing lookupCSEM insertCSEM returnM x in match x with Uminus _ -> memo x | _ -> memo x >>= fun x' -> memo (Uminus x') >> returnM x' let makeNode = identityM (* simplifiers for various kinds of nodes *) let rec snumM = function n when Number.is_zero n -> makeNode (Num (Number.zero)) | n when Number.negative n -> makeNode (Num (Number.negate n)) >>= suminusM | n -> makeNode (Num n) and suminusM = function Uminus x -> makeNode x | Num a when (Number.is_zero a) -> snumM Number.zero | a -> makeNode (Uminus a) and stimesM = function | (Uminus a, b) -> stimesM (a, b) >>= suminusM | (a, Uminus b) -> stimesM (a, b) >>= suminusM | (NaN I, CTimes (a, b)) -> stimesM (NaN I, b) >>= fun ib -> sctimesM (a, ib) | (NaN I, CTimesJ (a, b)) -> stimesM (NaN I, b) >>= fun ib -> sctimesjM (a, ib) | (Num a, Num b) -> snumM (Number.mul a b) | (Num a, Times (Num b, c)) -> snumM (Number.mul a b) >>= fun x -> stimesM (x, c) | (Num a, b) when Number.is_zero a -> snumM Number.zero | (Num a, b) when Number.is_one a -> makeNode b | (Num a, b) when Number.is_mone a -> suminusM b | (a, b) when is_known_constant b && not (is_known_constant a) -> stimesM (b, a) | (a, b) -> makeNode (Times (a, b)) and sctimesM = function | (Uminus a, b) -> sctimesM (a, b) >>= suminusM | (a, Uminus b) -> sctimesM (a, b) >>= suminusM | (a, b) -> makeNode (CTimes (a, b)) and sctimesjM = function | (Uminus a, b) -> sctimesjM (a, b) >>= suminusM | (a, Uminus b) -> sctimesjM (a, b) >>= suminusM | (a, b) -> makeNode (CTimesJ (a, b)) and reduce_sumM x = match x with [] -> returnM [] | [Num a] -> if (Number.is_zero a) then returnM [] else returnM x | [Uminus (Num a)] -> if (Number.is_zero a) then returnM [] else returnM x | (Num a) :: (Num b) :: s -> snumM (Number.add a b) >>= fun x -> reduce_sumM (x :: s) | (Num a) :: (Uminus (Num b)) :: s -> snumM (Number.sub a b) >>= fun x -> reduce_sumM (x :: s) | (Uminus (Num a)) :: (Num b) :: s -> snumM (Number.sub b a) >>= fun x -> reduce_sumM (x :: s) | (Uminus (Num a)) :: (Uminus (Num b)) :: s -> snumM (Number.add a b) >>= suminusM >>= fun x -> reduce_sumM (x :: s) | ((Num _) as a) :: b :: s -> reduce_sumM (b :: a :: s) | ((Uminus (Num _)) as a) :: b :: s -> reduce_sumM (b :: a :: s) | a :: s -> reduce_sumM s >>= fun s' -> returnM (a :: s') and collectible1 = function | NaN _ -> false | Uminus x -> collectible1 x | _ -> true and collectible (a, b) = collectible1 a (* collect common factors: ax + bx -> (a+b)x *) and collectM which x = let rec findCoeffM which = function | Times (a, b) when collectible (which (a, b)) -> returnM (which (a, b)) | Uminus x -> findCoeffM which x >>= fun (coeff, b) -> suminusM coeff >>= fun mcoeff -> returnM (mcoeff, b) | x -> snumM Number.one >>= fun one -> returnM (one, x) and separateM xpr = function [] -> returnM ([], []) | a :: b -> separateM xpr b >>= fun (w, wo) -> (* try first factor *) findCoeffM (fun (a, b) -> (a, b)) a >>= fun (c, x) -> if (xpr == x) && collectible (c, x) then returnM (c :: w, wo) else (* try second factor *) findCoeffM (fun (a, b) -> (b, a)) a >>= fun (c, x) -> if (xpr == x) && collectible (c, x) then returnM (c :: w, wo) else returnM (w, a :: wo) in match x with [] -> returnM x | [a] -> returnM x | a :: b -> findCoeffM which a >>= fun (_, xpr) -> separateM xpr x >>= fun (w, wo) -> collectM which wo >>= fun wo' -> splusM w >>= fun w' -> stimesM (w', xpr) >>= fun t' -> returnM (t':: wo') and mangleSumM x = returnM x >>= reduce_sumM >>= collectM (fun (a, b) -> (a, b)) >>= collectM (fun (a, b) -> (b, a)) >>= reduce_sumM >>= deepCollectM !Magic.deep_collect_depth >>= reduce_sumM and reorder_uminus = function (* push all Uminuses to the end *) [] -> [] | ((Uminus _) as a' :: b) -> (reorder_uminus b) @ [a'] | (a :: b) -> a :: (reorder_uminus b) and canonicalizeM = function [] -> snumM Number.zero | [a] -> makeNode a (* one term *) | a -> generateFusedMultAddM (reorder_uminus a) and generateFusedMultAddM = let rec is_multiplication = function | Times (Num a, b) -> true | Uminus (Times (Num a, b)) -> true | _ -> false and separate = function [] -> ([], [], Number.zero) | (Times (Num a, b)) as this :: c -> let (x, y, max) = separate c in let newmax = if (Number.greater a max) then a else max in (this :: x, y, newmax) | (Uminus (Times (Num a, b))) as this :: c -> let (x, y, max) = separate c in let newmax = if (Number.greater a max) then a else max in (this :: x, y, newmax) | this :: c -> let (x, y, max) = separate c in (x, this :: y, max) in fun l -> if !Magic.enable_fma && count is_multiplication l >= 2 then let (w, wo, max) = separate l in snumM (Number.div Number.one max) >>= fun invmax' -> snumM max >>= fun max' -> mapM (fun x -> stimesM (invmax', x)) w >>= splusM >>= fun pw' -> stimesM (max', pw') >>= fun mw' -> splusM (wo @ [mw']) else makeNode (Plus l) and negative = function Uminus _ -> true | _ -> false (* * simplify patterns of the form * * ((c_1 * a + ...) + ...) + (c_2 * a + ...) * * The pattern includes arbitrary coefficients and minus signs. * A common case of this pattern is the butterfly * (a + b) + (a - b) * (a + b) - (a - b) *) (* this whole procedure needs much more thought *) and deepCollectM maxdepth l = let rec findTerms depth x = match x with | Uminus x -> findTerms depth x | Times (Num _, b) -> (findTerms (depth - 1) b) | Plus l when depth > 0 -> x :: List.flatten (List.map (findTerms (depth - 1)) l) | x -> [x] and duplicates = function [] -> [] | a :: b -> if List.memq a b then a :: duplicates b else duplicates b in let rec splitDuplicates depth d x = if (List.memq x d) then snumM (Number.zero) >>= fun zero -> returnM (zero, x) else match x with | Times (a, b) -> splitDuplicates (depth - 1) d a >>= fun (a', xa) -> splitDuplicates (depth - 1) d b >>= fun (b', xb) -> stimesM (a', b') >>= fun ab -> stimesM (a, xb) >>= fun xb' -> stimesM (xa, b) >>= fun xa' -> stimesM (xa, xb) >>= fun xab -> splusM [xa'; xb'; xab] >>= fun x -> returnM (ab, x) | Uminus a -> splitDuplicates depth d a >>= fun (x, y) -> suminusM x >>= fun ux -> suminusM y >>= fun uy -> returnM (ux, uy) | Plus l when depth > 0 -> mapM (splitDuplicates (depth - 1) d) l >>= fun ld -> let (l', d') = List.split ld in splusM l' >>= fun p -> splusM d' >>= fun d'' -> returnM (p, d'') | x -> snumM (Number.zero) >>= fun zero' -> returnM (x, zero') in let l' = List.flatten (List.map (findTerms maxdepth) l) in match duplicates l' with | [] -> returnM l | d -> mapM (splitDuplicates maxdepth d) l >>= fun ld -> let (l', d') = List.split ld in splusM l' >>= fun l'' -> let rec flattenPlusM = function | Plus l -> returnM l | Uminus x -> flattenPlusM x >>= mapM suminusM | x -> returnM [x] in mapM flattenPlusM d' >>= fun d'' -> splusM (List.flatten d'') >>= fun d''' -> mangleSumM [l''; d'''] and splusM l = let fma_heuristics x = if !Magic.enable_fma then match x with | [Uminus (Times _); Times _] -> Some false | [Times _; Uminus (Times _)] -> Some false | [Uminus (_); Times _] -> Some true | [Times _; Uminus (Plus _)] -> Some true | [_; Uminus (Times _)] -> Some false | [Uminus (Times _); _] -> Some false | _ -> None else None in mangleSumM l >>= fun l' -> (* no terms are negative. Don't do anything *) if not (List.exists negative l') then canonicalizeM l' (* all terms are negative. Negate them all and collect the minus sign *) else if List.for_all negative l' then mapM suminusM l' >>= splusM >>= suminusM else match fma_heuristics l' with | Some true -> mapM suminusM l' >>= splusM >>= suminusM | Some false -> canonicalizeM l' | None -> (* Ask the Oracle for the canonical form *) if (not !Magic.randomized_cse) && Oracle.should_flip_sign (Plus l') then mapM suminusM l' >>= splusM >>= suminusM else canonicalizeM l' (* monadic style algebraic simplifier for the dag *) let rec algsimpM x = memoizing lookupSimpM insertSimpM (function | Num a -> snumM a | NaN _ as x -> makeNode x | Plus a -> mapM algsimpM a >>= splusM | Times (a, b) -> (algsimpM a >>= fun a' -> algsimpM b >>= fun b' -> stimesM (a', b')) | CTimes (a, b) -> (algsimpM a >>= fun a' -> algsimpM b >>= fun b' -> sctimesM (a', b')) | CTimesJ (a, b) -> (algsimpM a >>= fun a' -> algsimpM b >>= fun b' -> sctimesjM (a', b')) | Uminus a -> algsimpM a >>= suminusM | Store (v, a) -> algsimpM a >>= fun a' -> makeNode (Store (v, a')) | Load _ as x -> makeNode x) x let initialTable = (empty, empty) let simp_roots = mapM algsimpM let algsimp = runM initialTable simp_roots end (************************************************************* * Network transposition algorithm *************************************************************) module Transpose = struct open Monads.StateMonad open Monads.MemoMonad open Littlesimp let fetchDuals = fetchState let storeDuals = storeState let lookupDualsM key = fetchDuals >>= fun table -> returnM (node_lookup key table) let insertDualsM key value = fetchDuals >>= fun table -> storeDuals (node_insert key value table) let rec visit visited vtable parent_table = function [] -> (visited, parent_table) | node :: rest -> match node_lookup node vtable with | Some _ -> visit visited vtable parent_table rest | None -> let children = match node with | Store (v, n) -> [n] | Plus l -> l | Times (a, b) -> [a; b] | CTimes (a, b) -> [a; b] | CTimesJ (a, b) -> [a; b] | Uminus x -> [x] | _ -> [] in let rec loop t = function [] -> t | a :: rest -> (match node_lookup a t with None -> loop (node_insert a [node] t) rest | Some c -> loop (node_insert a (node :: c) t) rest) in (visit (node :: visited) (node_insert node () vtable) (loop parent_table children) (children @ rest)) let make_transposer parent_table = let rec termM node candidate_parent = match candidate_parent with | Store (_, n) when n == node -> dualM candidate_parent >>= fun x' -> returnM [x'] | Plus (l) when List.memq node l -> dualM candidate_parent >>= fun x' -> returnM [x'] | Times (a, b) when b == node -> dualM candidate_parent >>= fun x' -> returnM [makeTimes (a, x')] | CTimes (a, b) when b == node -> dualM candidate_parent >>= fun x' -> returnM [CTimes (a, x')] | CTimesJ (a, b) when b == node -> dualM candidate_parent >>= fun x' -> returnM [CTimesJ (a, x')] | Uminus n when n == node -> dualM candidate_parent >>= fun x' -> returnM [makeUminus x'] | _ -> returnM [] and dualExpressionM this_node = mapM (termM this_node) (match node_lookup this_node parent_table with | Some a -> a | None -> failwith "bug in dualExpressionM" ) >>= fun l -> returnM (makePlus (List.flatten l)) and dualM this_node = memoizing lookupDualsM insertDualsM (function | Load v as x -> if (Variable.is_constant v) then returnM (Load v) else (dualExpressionM x >>= fun d -> returnM (Store (v, d))) | Store (v, x) -> returnM (Load v) | x -> dualExpressionM x) this_node in dualM let is_store = function | Store _ -> true | _ -> false let transpose dag = let _ = Util.info "begin transpose" in let (all_nodes, parent_table) = visit [] Assoctable.empty Assoctable.empty dag in let transposerM = make_transposer parent_table in let mapTransposerM = mapM transposerM in let duals = runM Assoctable.empty mapTransposerM all_nodes in let roots = List.filter is_store duals in let _ = Util.info "end transpose" in roots end (************************************************************* * Various dag statistics *************************************************************) module Stats : sig type complexity val complexity : Expr.expr list -> complexity val same_complexity : complexity -> complexity -> bool val leq_complexity : complexity -> complexity -> bool val to_string : complexity -> string end = struct type complexity = int * int * int * int * int * int let rec visit visited vtable = function [] -> visited | node :: rest -> match node_lookup node vtable with Some _ -> visit visited vtable rest | None -> let children = match node with Store (v, n) -> [n] | Plus l -> l | Times (a, b) -> [a; b] | Uminus x -> [x] | _ -> [] in visit (node :: visited) (node_insert node () vtable) (children @ rest) let complexity dag = let rec loop (load, store, plus, times, uminus, num) = function [] -> (load, store, plus, times, uminus, num) | node :: rest -> loop (match node with | Load _ -> (load + 1, store, plus, times, uminus, num) | Store _ -> (load, store + 1, plus, times, uminus, num) | Plus x -> (load, store, plus + (List.length x - 1), times, uminus, num) | Times _ -> (load, store, plus, times + 1, uminus, num) | Uminus _ -> (load, store, plus, times, uminus + 1, num) | Num _ -> (load, store, plus, times, uminus, num + 1) | CTimes _ -> (load, store, plus, times, uminus, num) | CTimesJ _ -> (load, store, plus, times, uminus, num) | NaN _ -> (load, store, plus, times, uminus, num)) rest in let (l, s, p, t, u, n) = loop (0, 0, 0, 0, 0, 0) (visit [] Assoctable.empty dag) in (l, s, p, t, u, n) let weight (l, s, p, t, u, n) = l + s + 10 * p + 20 * t + u + n let same_complexity a b = weight a = weight b let leq_complexity a b = weight a <= weight b let to_string (l, s, p, t, u, n) = Printf.sprintf "ld=%d st=%d add=%d mul=%d uminus=%d num=%d\n" l s p t u n end (* simplify the dag *) let algsimp v = let rec simplification_loop v = let () = Util.info "simplification step" in let complexity = Stats.complexity v in let () = Util.info ("complexity = " ^ (Stats.to_string complexity)) in let v = (AlgSimp.algsimp @@ Transpose.transpose @@ AlgSimp.algsimp @@ Transpose.transpose) v in let complexity' = Stats.complexity v in let () = Util.info ("complexity = " ^ (Stats.to_string complexity')) in if (Stats.leq_complexity complexity' complexity) then let () = Util.info "end algsimp" in v else simplification_loop v in let () = Util.info "begin algsimp" in let v = AlgSimp.algsimp v in if !Magic.network_transposition then simplification_loop v else v fftw-3.3.3/genfft/annotate.ml0000644000175000017500000002700412054410050013015 00000000000000(* * Copyright (c) 1997-1999 Massachusetts Institute of Technology * Copyright (c) 2003, 2007-11 Matteo Frigo * Copyright (c) 2003, 2007-11 Massachusetts Institute of Technology * * This program is free software; you can redistribute it and/or modify * it under the terms of the GNU General Public License as published by * the Free Software Foundation; either version 2 of the License, or * (at your option) any later version. * * This program is distributed in the hope that it will be useful, * but WITHOUT ANY WARRANTY; without even the implied warranty of * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the * GNU General Public License for more details. * * You should have received a copy of the GNU General Public License * along with this program; if not, write to the Free Software * Foundation, Inc., 51 Franklin Street, Fifth Floor, Boston, MA 02110-1301 USA * *) (* Here, we take a schedule (produced by schedule.ml) ordering a sequence of instructions, and produce an annotated schedule. The annotated schedule has the same ordering as the original schedule, but is additionally partitioned into nested blocks of temporary variables. The partitioning is computed via a heuristic algorithm. The blocking allows the C code that we generate to consist of nested blocks that help communicate variable lifetimes to the compiler. *) open Schedule open Expr open Variable type annotated_schedule = Annotate of variable list * variable list * variable list * int * aschedule and aschedule = ADone | AInstr of assignment | ASeq of (annotated_schedule * annotated_schedule) let addelem a set = if not (List.memq a set) then a :: set else set let union l = let f x = addelem x (* let is source of polymorphism *) in List.fold_right f l (* set difference a - b *) let diff a b = List.filter (fun x -> not (List.memq x b)) a let rec minimize f = function [] -> failwith "minimize" | [n] -> n | n :: rest -> let x = minimize f rest in if (f x) >= (f n) then n else x (* find all variables used inside a scheduling unit *) let rec find_block_vars = function Done -> [] | (Instr (Assign (v, x))) -> v :: (find_vars x) | Par a -> List.flatten (List.map find_block_vars a) | Seq (a, b) -> (find_block_vars a) @ (find_block_vars b) let uniq l = List.fold_right (fun a b -> if List.memq a b then b else a :: b) l [] let has_related x = List.exists (Variable.same_class x) let rec overlap a b = Util.count (fun y -> has_related y b) a (* reorder a list of schedules so as to maximize overlap of variables *) let reorder l = let rec loop = function [] -> [] | (a, va) :: b -> let c = List.map (fun (a, x) -> ((a, x), (overlap va x, List.length x))) b in let c' = Sort.list (fun (_, (a, la)) (_, (b, lb)) -> la < lb or a > b) c in let b' = List.map (fun (a, _) -> a) c' in a :: (loop b') in let l' = List.map (fun x -> x, uniq (find_block_vars x)) l in (* start with smallest block --- does this matter ? *) match l' with [] -> [] | _ -> let m = minimize (fun (_, x) -> (List.length x)) l' in let l'' = Util.remove m l' in loop (m :: l'') (* remove Par blocks *) let rec linearize = function | Seq (a, Done) -> linearize a | Seq (Done, a) -> linearize a | Seq (a, b) -> Seq (linearize a, linearize b) (* try to balance nested Par blocks *) | Par [a] -> linearize a | Par l -> let n2 = (List.length l) / 2 in let rec loop n a b = if n = 0 then (List.rev b, a) else match a with [] -> failwith "loop" | x :: y -> loop (n - 1) y (x :: b) in let (a, b) = loop n2 (reorder l) [] in linearize (Seq (Par a, Par b)) | x -> x let subset a b = List.for_all (fun x -> List.exists (fun y -> x == y) b) a let use_same_vars (Assign (av, ax)) (Assign (bv, bx)) = is_temporary av && is_temporary bv && (let va = Expr.find_vars ax and vb = Expr.find_vars bx in subset va vb && subset vb va) let store_to_same_class (Assign (av, ax)) (Assign (bv, bx)) = is_locative av && is_locative bv && Variable.same_class av bv let loads_from_same_class (Assign (av, ax)) (Assign (bv, bx)) = match (ax, bx) with | (Load a), (Load b) when Variable.is_locative a && Variable.is_locative b -> Variable.same_class a b | _ -> false (* extract instructions from schedule *) let rec sched_to_ilist = function | Done -> [] | Instr a -> [a] | Seq (a, b) -> (sched_to_ilist a) @ (sched_to_ilist b) | _ -> failwith "sched_to_ilist" (* Par blocks removed by linearize *) let rec find_friends friendp insn friends foes = function | [] -> (friends, foes) | a :: b -> if (a == insn) || (friendp a insn) then find_friends friendp insn (a :: friends) foes b else find_friends friendp insn friends (a :: foes) b (* schedule all instructions in the equivalence class determined by friendp at the point where the last one is executed *) let rec delay_friends friendp sched = let rec recur insns = function | Done -> (Done, insns) | Instr a -> let (friends, foes) = find_friends friendp a [] [] insns in (Schedule.sequentially friends), foes | Seq (a, b) -> let (b', insnsb) = recur insns b in let (a', insnsa) = recur insnsb a in (Seq (a', b')), insnsa | _ -> failwith "delay_friends" in match recur (sched_to_ilist sched) sched with | (s, []) -> s (* assert that all insns have been used *) | _ -> failwith "delay_friends" (* schedule all instructions in the equivalence class determined by friendp at the point where the first one is executed *) let rec anticipate_friends friendp sched = let rec recur insns = function | Done -> (Done, insns) | Instr a -> let (friends, foes) = find_friends friendp a [] [] insns in (Schedule.sequentially friends), foes | Seq (a, b) -> let (a', insnsa) = recur insns a in let (b', insnsb) = recur insnsa b in (Seq (a', b')), insnsb | _ -> failwith "anticipate_friends" in match recur (sched_to_ilist sched) sched with | (s, []) -> s (* assert that all insns have been used *) | _ -> failwith "anticipate_friends" let collect_buddy_stores buddy_list sched = let rec recur sched delayed_stores = match sched with | Done -> (sched, delayed_stores) | Instr (Assign (v, x)) -> begin try let buddies = List.find (List.memq v) buddy_list in let tmp = Variable.make_temporary () in let i = Seq(Instr (Assign (tmp, x)), Instr (Assign (v, Times (NaN MULTI_A, Load tmp)))) and delayed_stores = (v, Load tmp) :: delayed_stores in try (Seq (i, Instr (Assign (List.hd buddies, Times (NaN MULTI_B, Plus (List.map (fun buddy -> List.assq buddy delayed_stores) buddies))) ))) , delayed_stores with Not_found -> (i, delayed_stores) with Not_found -> (sched, delayed_stores) end | Seq (a, b) -> let (newa, delayed_stores) = recur a delayed_stores in let (newb, delayed_stores) = recur b delayed_stores in (Seq (newa, newb), delayed_stores) | _ -> failwith "collect_buddy_stores" in let (sched, _) = recur sched [] in sched let schedule_for_pipeline sched = let update_readytimes t (Assign (v, _)) ready_times = (v, (t + !Magic.pipeline_latency)) :: ready_times and readyp t ready_times (Assign (_, x)) = List.for_all (fun var -> try (List.assq var ready_times) <= t with Not_found -> false) (List.filter Variable.is_temporary (Expr.find_vars x)) in let rec recur sched t ready_times delayed_instructions = let (ready, not_ready) = List.partition (readyp t ready_times) delayed_instructions in match ready with | a :: b -> let (sched, t, ready_times, delayed_instructions) = recur sched (t+1) (update_readytimes t a ready_times) (b @ not_ready) in (Seq (Instr a, sched)), t, ready_times, delayed_instructions | _ -> (match sched with | Done -> (sched, t, ready_times, delayed_instructions) | Instr a -> if (readyp t ready_times a) then (sched, (t+1), (update_readytimes t a ready_times), delayed_instructions) else (Done, t, ready_times, (a :: delayed_instructions)) | Seq (a, b) -> let (a, t, ready_times, delayed_instructions) = recur a t ready_times delayed_instructions in let (b, t, ready_times, delayed_instructions) = recur b t ready_times delayed_instructions in (Seq (a, b)), t, ready_times, delayed_instructions | _ -> failwith "schedule_for_pipeline") in let rec recur_until_done sched t ready_times delayed_instructions = let (sched, t, ready_times, delayed_instructions) = recur sched t ready_times delayed_instructions in match delayed_instructions with | [] -> sched | _ -> (Seq (sched, (recur_until_done Done (t+1) ready_times delayed_instructions))) in recur_until_done sched 0 [] [] let rec rewrite_declarations force_declarations (Annotate (_, _, declared, _, what)) = let m = !Magic.number_of_variables in let declare_it declared = if (force_declarations or List.length declared >= m) then ([], declared) else (declared, []) in match what with ADone -> Annotate ([], [], [], 0, what) | AInstr i -> let (u, d) = declare_it declared in Annotate ([], u, d, 0, what) | ASeq (a, b) -> let ma = rewrite_declarations false a and mb = rewrite_declarations false b in let Annotate (_, ua, _, _, _) = ma and Annotate (_, ub, _, _, _) = mb in let (u, d) = declare_it (declared @ ua @ ub) in Annotate ([], u, d, 0, ASeq (ma, mb)) let annotate list_of_buddy_stores schedule = let rec analyze live_at_end = function Done -> Annotate (live_at_end, [], [], 0, ADone) | Instr i -> (match i with Assign (v, x) -> let vars = (find_vars x) in Annotate (Util.remove v (union live_at_end vars), [v], [], 0, AInstr i)) | Seq (a, b) -> let ab = analyze live_at_end b in let Annotate (live_at_begin_b, defined_b, _, depth_a, _) = ab in let aa = analyze live_at_begin_b a in let Annotate (live_at_begin_a, defined_a, _, depth_b, _) = aa in let defined = List.filter is_temporary (defined_a @ defined_b) in let declarable = diff defined live_at_end in let undeclarable = diff defined declarable and maxdepth = max depth_a depth_b in Annotate (live_at_begin_a, undeclarable, declarable, List.length declarable + maxdepth, ASeq (aa, ab)) | _ -> failwith "really_analyze" in let () = Util.info "begin annotate" in let x = linearize schedule in let x = if (!Magic.schedule_for_pipeline && !Magic.pipeline_latency > 0) then schedule_for_pipeline x else x in let x = if !Magic.reorder_insns then linearize(anticipate_friends use_same_vars x) else x in (* delay stores to the real and imaginary parts of the same number *) let x = if !Magic.reorder_stores then linearize(delay_friends store_to_same_class x) else x in (* move loads of the real and imaginary parts of the same number *) let x = if !Magic.reorder_loads then linearize(anticipate_friends loads_from_same_class x) else x in let x = collect_buddy_stores list_of_buddy_stores x in let x = analyze [] x in let res = rewrite_declarations true x in let () = Util.info "end annotate" in res let rec dump print (Annotate (_, _, _, _, code)) = dump_code print code and dump_code print = function | ADone -> () | AInstr x -> print ((assignment_to_string x) ^ "\n") | ASeq (a, b) -> dump print a; dump print b fftw-3.3.3/genfft/assoctable.ml0000644000175000017500000000450212054410050013322 00000000000000(* * Copyright (c) 1997-1999 Massachusetts Institute of Technology * Copyright (c) 2003, 2007-11 Matteo Frigo * Copyright (c) 2003, 2007-11 Massachusetts Institute of Technology * * This program is free software; you can redistribute it and/or modify * it under the terms of the GNU General Public License as published by * the Free Software Foundation; either version 2 of the License, or * (at your option) any later version. * * This program is distributed in the hope that it will be useful, * but WITHOUT ANY WARRANTY; without even the implied warranty of * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the * GNU General Public License for more details. * * You should have received a copy of the GNU General Public License * along with this program; if not, write to the Free Software * Foundation, Inc., 51 Franklin Street, Fifth Floor, Boston, MA 02110-1301 USA * *) (************************************************************* * Functional associative table *************************************************************) (* * this module implements a functional associative table. * The table is parametrized by an equality predicate and * a hash function, with the restriction that (equal a b) ==> * hash a == hash b. * The table is purely functional and implemented using a binary * search tree (not balanced for now) *) type ('a, 'b) elem = Leaf | Node of int * ('a, 'b) elem * ('a, 'b) elem * ('a * 'b) list let empty = Leaf let lookup hash equal key table = let h = hash key in let rec look = function Leaf -> None | Node (hash_key, left, right, this_list) -> if (hash_key < h) then look left else if (hash_key > h) then look right else let rec loop = function [] -> None | (a, b) :: rest -> if (equal key a) then Some b else loop rest in loop this_list in look table let insert hash key value table = let h = hash key in let rec ins = function Leaf -> Node (h, Leaf, Leaf, [(key, value)]) | Node (hash_key, left, right, this_list) -> if (hash_key < h) then Node (hash_key, ins left, right, this_list) else if (hash_key > h) then Node (hash_key, left, ins right, this_list) else Node (hash_key, left, right, (key, value) :: this_list) in ins table fftw-3.3.3/genfft/c.ml0000644000175000017500000003615212054410050011432 00000000000000(* * Copyright (c) 1997-1999 Massachusetts Institute of Technology * Copyright (c) 2003, 2007-11 Matteo Frigo * Copyright (c) 2003, 2007-11 Massachusetts Institute of Technology * * This program is free software; you can redistribute it and/or modify * it under the terms of the GNU General Public License as published by * the Free Software Foundation; either version 2 of the License, or * (at your option) any later version. * * This program is distributed in the hope that it will be useful, * but WITHOUT ANY WARRANTY; without even the implied warranty of * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the * GNU General Public License for more details. * * You should have received a copy of the GNU General Public License * along with this program; if not, write to the Free Software * Foundation, Inc., 51 Franklin Street, Fifth Floor, Boston, MA 02110-1301 USA * *) (* * This module contains the definition of a C-like abstract * syntax tree, and functions to convert ML values into C * programs *) open Expr open Annotate open List let realtype = "R" let realtypep = realtype ^ " *" let extended_realtype = "E" let constrealtype = "const " ^ realtype let constrealtypep = constrealtype ^ " *" let stridetype = "stride" (*********************************** * C program structure ***********************************) type c_decl = | Decl of string * string | Tdecl of string (* arbitrary text declaration *) and c_ast = | Asch of annotated_schedule | Simd_leavefun | Return of c_ast | For of c_ast * c_ast * c_ast * c_ast | If of c_ast * c_ast | Block of (c_decl list) * (c_ast list) | Binop of string * c_ast * c_ast | Expr_assign of c_ast * c_ast | Stmt_assign of c_ast * c_ast | Comma of c_ast * c_ast | Integer of int | CVar of string | CCall of string * c_ast | CPlus of c_ast list | ITimes of c_ast * c_ast | CUminus of c_ast and c_fcn = Fcn of string * string * (c_decl list) * c_ast let ctimes = function | (Integer 1), a -> a | a, (Integer 1) -> a | a, b -> ITimes (a, b) (* * C AST unparser *) let foldr_string_concat l = fold_right (^) l "" let rec unparse_expr_c = let yes x = x and no x = "" in let rec unparse_plus maybe = let maybep = maybe " + " in function | [] -> "" | (Uminus (Times (a, b))) :: (Uminus c) :: d -> maybep ^ (op "FNMA" a b c) ^ (unparse_plus yes d) | (Uminus c) :: (Uminus (Times (a, b))) :: d -> maybep ^ (op "FNMA" a b c) ^ (unparse_plus yes d) | (Uminus (Times (a, b))) :: c :: d -> maybep ^ (op "FNMS" a b c) ^ (unparse_plus yes d) | c :: (Uminus (Times (a, b))) :: d -> maybep ^ (op "FNMS" a b c) ^ (unparse_plus yes d) | (Times (a, b)) :: (Uminus c) :: d -> maybep ^ (op "FMS" a b c) ^ (unparse_plus yes d) | (Uminus c) :: (Times (a, b)) :: d -> maybep ^ (op "FMS" a b c) ^ (unparse_plus yes d) | (Times (a, b)) :: c :: d -> maybep ^ (op "FMA" a b c) ^ (unparse_plus yes d) | c :: (Times (a, b)) :: d -> maybep ^ (op "FMA" a b c) ^ (unparse_plus yes d) | (Uminus a :: b) -> " - " ^ (parenthesize a) ^ (unparse_plus yes b) | (a :: b) -> maybep ^ (parenthesize a) ^ (unparse_plus yes b) and parenthesize x = match x with | (Load _) -> unparse_expr_c x | (Num _) -> unparse_expr_c x | _ -> "(" ^ (unparse_expr_c x) ^ ")" and op nam a b c = nam ^ "(" ^ (unparse_expr_c a) ^ ", " ^ (unparse_expr_c b) ^ ", " ^ (unparse_expr_c c) ^ ")" in function | Load v -> Variable.unparse v | Num n -> Number.to_konst n | Plus [] -> "0.0 /* bug */" | Plus [a] -> " /* bug */ " ^ (unparse_expr_c a) | Plus a -> (unparse_plus no a) | Times (a, b) -> (parenthesize a) ^ " * " ^ (parenthesize b) | Uminus (Plus [a; Uminus b]) -> unparse_plus no [b; Uminus a] | Uminus a -> "- " ^ (parenthesize a) | _ -> failwith "unparse_expr_c" and unparse_expr_generic = let rec u x = unparse_expr_generic x and unary op a = Printf.sprintf "%s(%s)" op (u a) and binary op a b = Printf.sprintf "%s(%s, %s)" op (u a) (u b) and ternary op a b c = Printf.sprintf "%s(%s, %s, %s)" op (u a) (u b) (u c) and quaternary op a b c d = Printf.sprintf "%s(%s, %s, %s, %s)" op (u a) (u b) (u c) (u d) and unparse_plus = function | [(Uminus (Times (a, b))); Times (c, d)] -> quaternary "FNMMS" a b c d | [Times (c, d); (Uminus (Times (a, b)))] -> quaternary "FNMMS" a b c d | [Times (c, d); (Times (a, b))] -> quaternary "FMMA" a b c d | [(Uminus (Times (a, b))); c] -> ternary "FNMS" a b c | [c; (Uminus (Times (a, b)))] -> ternary "FNMS" a b c | [(Uminus c); (Times (a, b))] -> ternary "FMS" a b c | [(Times (a, b)); (Uminus c)] -> ternary "FMS" a b c | [c; (Times (a, b))] -> ternary "FMA" a b c | [(Times (a, b)); c] -> ternary "FMA" a b c | [a; Uminus b] -> binary "SUB" a b | [a; b] -> binary "ADD" a b | a :: b :: c -> binary "ADD" a (Plus (b :: c)) | _ -> failwith "unparse_plus" in function | Load v -> Variable.unparse v | Num n -> Number.to_konst n | Plus a -> unparse_plus a | Times (a, b) -> binary "MUL" a b | Uminus a -> unary "NEG" a | _ -> failwith "unparse_expr" and unparse_expr x = if !Magic.generic_arith then unparse_expr_generic x else unparse_expr_c x and unparse_assignment (Assign (v, x)) = (Variable.unparse v) ^ " = " ^ (unparse_expr x) ^ ";\n" and unparse_annotated force_bracket = let rec unparse_code = function ADone -> "" | AInstr i -> unparse_assignment i | ASeq (a, b) -> (unparse_annotated false a) ^ (unparse_annotated false b) and declare_variables l = let rec uvar = function [] -> failwith "uvar" | [v] -> (Variable.unparse v) ^ ";\n" | a :: b -> (Variable.unparse a) ^ ", " ^ (uvar b) in let rec vvar l = let s = if !Magic.compact then 15 else 1 in if (List.length l <= s) then match l with [] -> "" | _ -> extended_realtype ^ " " ^ (uvar l) else (vvar (Util.take s l)) ^ (vvar (Util.drop s l)) in vvar (List.filter Variable.is_temporary l) in function Annotate (_, _, decl, _, code) -> if (not force_bracket) && (Util.null decl) then unparse_code code else "{\n" ^ (declare_variables decl) ^ (unparse_code code) ^ "}\n" and unparse_decl = function | Decl (a, b) -> a ^ " " ^ b ^ ";\n" | Tdecl x -> x and unparse_ast = let rec unparse_plus = function | [] -> "" | (CUminus a :: b) -> " - " ^ (parenthesize a) ^ (unparse_plus b) | (a :: b) -> " + " ^ (parenthesize a) ^ (unparse_plus b) and parenthesize x = match x with | (CVar _) -> unparse_ast x | (CCall _) -> unparse_ast x | (Integer _) -> unparse_ast x | _ -> "(" ^ (unparse_ast x) ^ ")" in function | Asch a -> (unparse_annotated true a) | Simd_leavefun -> "" (* used only in SIMD code *) | Return x -> "return " ^ unparse_ast x ^ ";" | For (a, b, c, d) -> "for (" ^ unparse_ast a ^ "; " ^ unparse_ast b ^ "; " ^ unparse_ast c ^ ")" ^ unparse_ast d | If (a, d) -> "if (" ^ unparse_ast a ^ ")" ^ unparse_ast d | Block (d, s) -> if (s == []) then "" else "{\n" ^ foldr_string_concat (map unparse_decl d) ^ foldr_string_concat (map unparse_ast s) ^ "}\n" | Binop (op, a, b) -> (unparse_ast a) ^ op ^ (unparse_ast b) | Expr_assign (a, b) -> (unparse_ast a) ^ " = " ^ (unparse_ast b) | Stmt_assign (a, b) -> (unparse_ast a) ^ " = " ^ (unparse_ast b) ^ ";\n" | Comma (a, b) -> (unparse_ast a) ^ ", " ^ (unparse_ast b) | Integer i -> string_of_int i | CVar s -> s | CCall (s, x) -> s ^ "(" ^ (unparse_ast x) ^ ")" | CPlus [] -> "0 /* bug */" | CPlus [a] -> " /* bug */ " ^ (unparse_ast a) | CPlus (a::b) -> (parenthesize a) ^ (unparse_plus b) | ITimes (a, b) -> (parenthesize a) ^ " * " ^ (parenthesize b) | CUminus a -> "- " ^ (parenthesize a) and unparse_function = function Fcn (typ, name, args, body) -> let rec unparse_args = function [Decl (a, b)] -> a ^ " " ^ b | (Decl (a, b)) :: s -> a ^ " " ^ b ^ ", " ^ unparse_args s | [] -> "" | _ -> failwith "unparse_function" in (typ ^ " " ^ name ^ "(" ^ unparse_args args ^ ")\n" ^ unparse_ast body) (************************************************************* * traverse a a function and return a list of all expressions, * in the execution order **************************************************************) let rec fcn_to_expr_list = fun (Fcn (_, _, _, body)) -> ast_to_expr_list body and acode_to_expr_list = function AInstr (Assign (_, x)) -> [x] | ASeq (a, b) -> (asched_to_expr_list a) @ (asched_to_expr_list b) | _ -> [] and asched_to_expr_list (Annotate (_, _, _, _, code)) = acode_to_expr_list code and ast_to_expr_list = function Asch a -> asched_to_expr_list a | Block (_, a) -> flatten (map ast_to_expr_list a) | For (_, _, _, body) -> ast_to_expr_list body | If (_, body) -> ast_to_expr_list body | _ -> [] (*********************** * Extracting Constants ***********************) (* add a new key & value to a list of (key,value) pairs, where the keys are floats and each key is unique up to almost_equal *) let extract_constants f = let constlist = flatten (map expr_to_constants (ast_to_expr_list f)) in map (fun n -> Tdecl ("DK(" ^ (Number.to_konst n) ^ ", " ^ (Number.to_string n) ^ ");\n")) (unique_constants constlist) (****************************** Extracting operation counts ******************************) let count_stack_vars = let rec count_acode = function | ASeq (a, b) -> max (count_asched a) (count_asched b) | _ -> 0 and count_asched (Annotate (_, _, decl, _, code)) = (length decl) + (count_acode code) and count_ast = function | Asch a -> count_asched a | Block (d, a) -> (length d) + (Util.max_list (map count_ast a)) | For (_, _, _, body) -> count_ast body | If (_, body) -> count_ast body | _ -> 0 in function (Fcn (_, _, _, body)) -> count_ast body let count_memory_acc f = let rec count_var v = if (Variable.is_locative v) then 1 else 0 and count_acode = function | AInstr (Assign (v, _)) -> count_var v | ASeq (a, b) -> (count_asched a) + (count_asched b) | _ -> 0 and count_asched = function Annotate (_, _, _, _, code) -> count_acode code and count_ast = function | Asch a -> count_asched a | Block (_, a) -> (Util.sum_list (map count_ast a)) | Comma (a, b) -> (count_ast a) + (count_ast b) | For (_, _, _, body) -> count_ast body | If (_, body) -> count_ast body | _ -> 0 and count_acc_expr_func acc = function | Load v -> acc + (count_var v) | Plus a -> fold_left count_acc_expr_func acc a | Times (a, b) -> fold_left count_acc_expr_func acc [a; b] | Uminus a -> count_acc_expr_func acc a | _ -> acc in let (Fcn (typ, name, args, body)) = f in (count_ast body) + fold_left count_acc_expr_func 0 (fcn_to_expr_list f) let good_for_fma = To_alist.good_for_fma let build_fma = function | [a; Times (b, c)] when good_for_fma (b, c) -> Some (a, b, c) | [Times (b, c); a] when good_for_fma (b, c) -> Some (a, b, c) | [a; Uminus (Times (b, c))] when good_for_fma (b, c) -> Some (a, b, c) | [Uminus (Times (b, c)); a] when good_for_fma (b, c) -> Some (a, b, c) | _ -> None let rec count_flops_expr_func (adds, mults, fmas) = function | Plus [] -> (adds, mults, fmas) | Plus ([_; _] as a) -> begin match build_fma a with | None -> fold_left count_flops_expr_func (adds + (length a) - 1, mults, fmas) a | Some (a, b, c) -> fold_left count_flops_expr_func (adds, mults, fmas+1) [a; b; c] end | Plus (a :: b) -> count_flops_expr_func (adds, mults, fmas) (Plus [a; Plus b]) | Times (NaN MULTI_A,_) -> (adds, mults, fmas) | Times (NaN MULTI_B,_) -> (adds, mults, fmas) | Times (NaN I,b) -> count_flops_expr_func (adds, mults, fmas) b | Times (NaN CONJ,b) -> count_flops_expr_func (adds, mults, fmas) b | Times (a,b) -> fold_left count_flops_expr_func (adds, mults+1, fmas) [a; b] | CTimes (a,b) -> fold_left count_flops_expr_func (adds+1, mults+2, fmas) [a; b] | CTimesJ (a,b) -> fold_left count_flops_expr_func (adds+1, mults+2, fmas) [a; b] | Uminus a -> count_flops_expr_func (adds, mults, fmas) a | _ -> (adds, mults, fmas) let count_flops f = fold_left count_flops_expr_func (0, 0, 0) (fcn_to_expr_list f) let count_constants f = length (unique_constants (flatten (map expr_to_constants (fcn_to_expr_list f)))) let arith_complexity f = let (a, m, fmas) = count_flops f and v = count_stack_vars f and c = count_constants f and mem = count_memory_acc f in (a, m, fmas, v, c, mem) (* print the operation costs *) let print_cost f = let Fcn (_, _, _, _) = f and (a, m, fmas, v, c, mem) = arith_complexity f in "/*\n"^ " * This function contains " ^ (string_of_int (a + fmas)) ^ " FP additions, " ^ (string_of_int (m + fmas)) ^ " FP multiplications,\n" ^ " * (or, " ^ (string_of_int a) ^ " additions, " ^ (string_of_int m) ^ " multiplications, " ^ (string_of_int fmas) ^ " fused multiply/add),\n" ^ " * " ^ (string_of_int v) ^ " stack variables, " ^ (string_of_int c) ^ " constants, and " ^ (string_of_int mem) ^ " memory accesses\n" ^ " */\n" (***************************************** * functions that create C arrays *****************************************) type stride = | SVar of string | SConst of string | SInteger of int | SNeg of stride type sstride = | Simple of int | Constant of (string * int) | Composite of (string * int) | Negative of sstride let rec simplify_stride stride i = match (stride, i) with (_, 0) -> Simple 0 | (SInteger n, i) -> Simple (n * i) | (SConst s, i) -> Constant (s, i) | (SVar s, i) -> Composite (s, i) | (SNeg x, i) -> match (simplify_stride x i) with | Negative y -> y | y -> Negative y let rec cstride_to_string = function | Simple i -> string_of_int i | Constant (s, i) -> if !Magic.lisp_syntax then "(* " ^ s ^ " " ^ (string_of_int i) ^ ")" else s ^ " * " ^ (string_of_int i) | Composite (s, i) -> if !Magic.lisp_syntax then "(* " ^ s ^ " " ^ (string_of_int i) ^ ")" else "WS(" ^ s ^ ", " ^ (string_of_int i) ^ ")" | Negative x -> "-" ^ cstride_to_string x let aref name index = if !Magic.lisp_syntax then Printf.sprintf "(aref %s %s)" name index else Printf.sprintf "%s[%s]" name index let array_subscript name stride k = aref name (cstride_to_string (simplify_stride stride k)) let varray_subscript name vstride stride v i = let vindex = simplify_stride vstride v and iindex = simplify_stride stride i in let index = match (vindex, iindex) with (Simple vi, Simple ii) -> string_of_int (vi + ii) | (Simple 0, x) -> cstride_to_string x | (x, Simple 0) -> cstride_to_string x | _ -> (cstride_to_string vindex) ^ " + " ^ (cstride_to_string iindex) in aref name index let real_of s = "c_re(" ^ s ^ ")" let imag_of s = "c_im(" ^ s ^ ")" let flops_of f = let (add, mul, fma) = count_flops f in Printf.sprintf "{ %d, %d, %d, 0 }" add mul fma fftw-3.3.3/genfft/complex.ml0000644000175000017500000001131212054410050012646 00000000000000(* * Copyright (c) 1997-1999 Massachusetts Institute of Technology * Copyright (c) 2003, 2007-11 Matteo Frigo * Copyright (c) 2003, 2007-11 Massachusetts Institute of Technology * * This program is free software; you can redistribute it and/or modify * it under the terms of the GNU General Public License as published by * the Free Software Foundation; either version 2 of the License, or * (at your option) any later version. * * This program is distributed in the hope that it will be useful, * but WITHOUT ANY WARRANTY; without even the implied warranty of * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the * GNU General Public License for more details. * * You should have received a copy of the GNU General Public License * along with this program; if not, write to the Free Software * Foundation, Inc., 51 Franklin Street, Fifth Floor, Boston, MA 02110-1301 USA * *) (* abstraction layer for complex operations *) open Littlesimp open Expr (* type of complex expressions *) type expr = CE of Expr.expr * Expr.expr let two = CE (makeNum Number.two, makeNum Number.zero) let one = CE (makeNum Number.one, makeNum Number.zero) let i = CE (makeNum Number.zero, makeNum Number.one) let zero = CE (makeNum Number.zero, makeNum Number.zero) let make (r, i) = CE (r, i) let uminus (CE (a, b)) = CE (makeUminus a, makeUminus b) let inverse_int n = CE (makeNum (Number.div Number.one (Number.of_int n)), makeNum Number.zero) let inverse_int_sqrt n = CE (makeNum (Number.div Number.one (Number.sqrt (Number.of_int n))), makeNum Number.zero) let int_sqrt n = CE (makeNum (Number.sqrt (Number.of_int n)), makeNum Number.zero) let nan x = CE (NaN x, makeNum Number.zero) let half = inverse_int 2 let times3x3 (CE (a, b)) (CE (c, d)) = CE (makePlus [makeTimes (c, makePlus [a; makeUminus (b)]); makeTimes (b, makePlus [c; makeUminus (d)])], makePlus [makeTimes (a, makePlus [c; d]); makeUminus(makeTimes (c, makePlus [a; makeUminus (b)]))]) let times (CE (a, b)) (CE (c, d)) = if not !Magic.threemult then CE (makePlus [makeTimes (a, c); makeUminus (makeTimes (b, d))], makePlus [makeTimes (a, d); makeTimes (b, c)]) else if is_constant c && is_constant d then times3x3 (CE (a, b)) (CE (c, d)) else (* hope a and b are constant expressions *) times3x3 (CE (c, d)) (CE (a, b)) let ctimes (CE (a, _)) (CE (c, _)) = CE (CTimes (a, c), makeNum Number.zero) let ctimesj (CE (a, _)) (CE (c, _)) = CE (CTimesJ (a, c), makeNum Number.zero) (* complex exponential (of root of unity); returns exp(2*pi*i/n * m) *) let exp n i = let (c, s) = Number.cexp n i in CE (makeNum c, makeNum s) (* various trig functions evaluated at (2*pi*i/n * m) *) let sec n m = let (c, s) = Number.cexp n m in CE (makeNum (Number.div Number.one c), makeNum Number.zero) let csc n m = let (c, s) = Number.cexp n m in CE (makeNum (Number.div Number.one s), makeNum Number.zero) let tan n m = let (c, s) = Number.cexp n m in CE (makeNum (Number.div s c), makeNum Number.zero) let cot n m = let (c, s) = Number.cexp n m in CE (makeNum (Number.div c s), makeNum Number.zero) (* complex sum *) let plus a = let rec unzip_complex = function [] -> ([], []) | ((CE (a, b)) :: s) -> let (r,i) = unzip_complex s in (a::r), (b::i) in let (c, d) = unzip_complex a in CE (makePlus c, makePlus d) (* extract real/imaginary *) let real (CE (a, b)) = CE (a, makeNum Number.zero) let imag (CE (a, b)) = CE (b, makeNum Number.zero) let iimag (CE (a, b)) = CE (makeNum Number.zero, b) let conj (CE (a, b)) = CE (a, makeUminus b) (* abstraction of sum_{i=0}^{n-1} *) let sigma a b f = plus (List.map f (Util.interval a b)) (* store and assignment operations *) let store_real v (CE (a, b)) = Expr.Store (v, a) let store_imag v (CE (a, b)) = Expr.Store (v, b) let store (vr, vi) x = (store_real vr x, store_imag vi x) let assign_real v (CE (a, b)) = Expr.Assign (v, a) let assign_imag v (CE (a, b)) = Expr.Assign (v, b) let assign (vr, vi) x = (assign_real vr x, assign_imag vi x) (************************ shortcuts ************************) let (@*) = times let (@+) a b = plus [a; b] let (@-) a b = plus [a; uminus b] (* type of complex signals *) type signal = int -> expr (* make a finite signal infinite *) let infinite n signal i = if ((0 <= i) && (i < n)) then signal i else zero let hermitian n a = Util.array n (fun i -> if (i = 0) then real (a 0) else if (i < n - i) then (a i) else if (i > n - i) then conj (a (n - i)) else real (a i)) let antihermitian n a = Util.array n (fun i -> if (i = 0) then iimag (a 0) else if (i < n - i) then (a i) else if (i > n - i) then uminus (conj (a (n - i))) else iimag (a i)) fftw-3.3.3/genfft/conv.ml0000644000175000017500000001016512054410050012151 00000000000000(* * Copyright (c) 1997-1999 Massachusetts Institute of Technology * Copyright (c) 2003, 2007-11 Matteo Frigo * Copyright (c) 2003, 2007-11 Massachusetts Institute of Technology * * This program is free software; you can redistribute it and/or modify * it under the terms of the GNU General Public License as published by * the Free Software Foundation; either version 2 of the License, or * (at your option) any later version. * * This program is distributed in the hope that it will be useful, * but WITHOUT ANY WARRANTY; without even the implied warranty of * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the * GNU General Public License for more details. * * You should have received a copy of the GNU General Public License * along with this program; if not, write to the Free Software * Foundation, Inc., 51 Franklin Street, Fifth Floor, Boston, MA 02110-1301 USA * *) open Complex open Util let polyphase m a ph i = a (m * i + ph) let rec divmod n i = if (i < 0) then let (a, b) = divmod n (i + n) in (a - 1, b) else (i / n, i mod n) let unpolyphase m a i = let (x, y) = divmod m i in a y x let lift2 f a b i = f (a i) (b i) (* convolution of signals A and B *) let rec conv na a nb b = let rec naive na a nb b i = sigma 0 na (fun j -> (a j) @* (b (i - j))) and recur na a nb b = if (na <= 1 || nb <= 1) then naive na a nb b else let p = polyphase 2 in let ee = conv (na - na / 2) (p a 0) (nb - nb / 2) (p b 0) and eo = conv (na - na / 2) (p a 0) (nb / 2) (p b 1) and oe = conv (na / 2) (p a 1) (nb - nb / 2) (p b 0) and oo = conv (na / 2) (p a 1) (nb / 2) (p b 1) in unpolyphase 2 (function 0 -> fun i -> (ee i) @+ (oo (i - 1)) | 1 -> fun i -> (eo i) @+ (oe i) | _ -> failwith "recur") (* Karatsuba variant 1: (a+bx)(c+dx) = (ac+bdxx)+((a+b)(c+d)-ac-bd)x *) and karatsuba1 na a nb b = let p = polyphase 2 in let ae = p a 0 and nae = na - na / 2 and ao = p a 1 and nao = na / 2 and be = p b 0 and nbe = nb - nb / 2 and bo = p b 1 and nbo = nb / 2 in let ae = infinite nae ae and ao = infinite nao ao and be = infinite nbe be and bo = infinite nbo bo in let aeo = lift2 (@+) ae ao and naeo = nae and beo = lift2 (@+) be bo and nbeo = nbe in let ee = conv nae ae nbe be and oo = conv nao ao nbo bo and eoeo = conv naeo aeo nbeo beo in let q = function 0 -> fun i -> (ee i) @+ (oo (i - 1)) | 1 -> fun i -> (eoeo i) @- ((ee i) @+ (oo i)) | _ -> failwith "karatsuba1" in unpolyphase 2 q (* Karatsuba variant 2: (a+bx)(c+dx) = ((a+b)c-b(c-dxx))+x((a+b)c-a(c-d)) *) and karatsuba2 na a nb b = let p = polyphase 2 in let ae = p a 0 and nae = na - na / 2 and ao = p a 1 and nao = na / 2 and be = p b 0 and nbe = nb - nb / 2 and bo = p b 1 and nbo = nb / 2 in let ae = infinite nae ae and ao = infinite nao ao and be = infinite nbe be and bo = infinite nbo bo in let c1 = conv nae (lift2 (@+) ae ao) nbe be and c2 = conv nao ao (nbo + 1) (fun i -> be i @- bo (i - 1)) and c3 = conv nae ae nbe (lift2 (@-) be bo) in let q = function 0 -> lift2 (@-) c1 c2 | 1 -> lift2 (@-) c1 c3 | _ -> failwith "karatsuba2" in unpolyphase 2 q and karatsuba na a nb b = let m = na + nb - 1 in if (m < !Magic.karatsuba_min) then recur na a nb b else match !Magic.karatsuba_variant with 1 -> karatsuba1 na a nb b | 2 -> karatsuba2 na a nb b | _ -> failwith "unknown karatsuba variant" and via_circular na a nb b = let m = na + nb - 1 in if (m < !Magic.circular_min) then karatsuba na a nb b else let rec find_min n = if n >= m then n else find_min (2 * n) in circular (find_min 1) a b in let a = infinite na a and b = infinite nb b in let res = array (na + nb - 1) (via_circular na a nb b) in infinite (na + nb - 1) res and circular n a b = let via_dft n a b = let fa = Fft.dft (-1) n a and fb = Fft.dft (-1) n b and scale = inverse_int n in let fab i = ((fa i) @* (fb i)) @* scale in Fft.dft 1 n fab in via_dft n a b fftw-3.3.3/genfft/dag.ml0000644000175000017500000000650012054410050011735 00000000000000(* * Copyright (c) 1997-1999 Massachusetts Institute of Technology * Copyright (c) 2003, 2007-11 Matteo Frigo * Copyright (c) 2003, 2007-11 Massachusetts Institute of Technology * * This program is free software; you can redistribute it and/or modify * it under the terms of the GNU General Public License as published by * the Free Software Foundation; either version 2 of the License, or * (at your option) any later version. * * This program is distributed in the hope that it will be useful, * but WITHOUT ANY WARRANTY; without even the implied warranty of * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the * GNU General Public License for more details. * * You should have received a copy of the GNU General Public License * along with this program; if not, write to the Free Software * Foundation, Inc., 51 Franklin Street, Fifth Floor, Boston, MA 02110-1301 USA * *) open Util (* Here, we have functions to transform a sequence of assignments (variable = expression) into a DAG (a directed, acyclic graph). The nodes of the DAG are the assignments, and the edges indicate dependencies. (The DAG is analyzed in the scheduler to find an efficient ordering of the assignments.) This file also contains utilities to manipulate the DAG in various ways. *) (******************************************** * Dag structure ********************************************) type color = RED | BLUE | BLACK | YELLOW type dagnode = { assigned: Variable.variable; mutable expression: Expr.expr; input_variables: Variable.variable list; mutable successors: dagnode list; mutable predecessors: dagnode list; mutable label: int; mutable color: color} type dag = Dag of (dagnode list) (* true if node uses v *) let node_uses v node = List.exists (Variable.same v) node.input_variables (* true if assignment of v clobbers any input of node *) let node_clobbers node v = List.exists (Variable.same_location v) node.input_variables (* true if nodeb depends on nodea *) let depends_on nodea nodeb = node_uses nodea.assigned nodeb or node_clobbers nodea nodeb.assigned (* transform an assignment list into a dag *) let makedag alist = let dag = List.map (fun assignment -> let (v, x) = assignment in { assigned = v; expression = x; input_variables = Expr.find_vars x; successors = []; predecessors = []; label = 0; color = BLACK }) alist in begin for_list dag (fun i -> for_list dag (fun j -> if depends_on i j then begin i.successors <- j :: i.successors; j.predecessors <- i :: j.predecessors; end)); Dag dag; end let map f (Dag dag) = Dag (List.map f dag) let for_all (Dag dag) f = (* type system loophole *) let make_unit _ = () in make_unit (List.map f dag) let to_list (Dag dag) = dag let find_node f (Dag dag) = Util.find_elem f dag (* breadth-first search *) let rec bfs (Dag dag) node init_label = let _ = node.label <- init_label in let rec loop = function [] -> () | node :: rest -> let neighbors = node.predecessors @ node.successors in let m = min_list (List.map (fun node -> node.label) neighbors) in if (node.label > m + 1) then begin node.label <- m + 1; loop (rest @ neighbors); end else loop rest in let neighbors = node.predecessors @ node.successors in loop neighbors fftw-3.3.3/genfft/expr.ml0000644000175000017500000001243212054410050012161 00000000000000(* * Copyright (c) 1997-1999 Massachusetts Institute of Technology * Copyright (c) 2003, 2007-11 Matteo Frigo * Copyright (c) 2003, 2007-11 Massachusetts Institute of Technology * * This program is free software; you can redistribute it and/or modify * it under the terms of the GNU General Public License as published by * the Free Software Foundation; either version 2 of the License, or * (at your option) any later version. * * This program is distributed in the hope that it will be useful, * but WITHOUT ANY WARRANTY; without even the implied warranty of * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the * GNU General Public License for more details. * * You should have received a copy of the GNU General Public License * along with this program; if not, write to the Free Software * Foundation, Inc., 51 Franklin Street, Fifth Floor, Boston, MA 02110-1301 USA * *) (* Here, we define the data type encapsulating a symbolic arithmetic expression, and provide some routines for manipulating it. *) (* I will regret this hack : *) (* NEWS: I did *) type transcendent = I | MULTI_A | MULTI_B | CONJ type expr = | Num of Number.number | NaN of transcendent | Plus of expr list | Times of expr * expr | CTimes of expr * expr | CTimesJ of expr * expr (* CTimesJ (a, b) = conj(a) * b *) | Uminus of expr | Load of Variable.variable | Store of Variable.variable * expr type assignment = Assign of Variable.variable * expr (* various hash functions *) let hash_float x = let (mantissa, exponent) = frexp x in truncate (float_of_int(exponent) *. 1234.567 +. mantissa *. 10000.0) let sum_list l = List.fold_right (+) l 0 let transcendent_to_float = function | I -> 2.718281828459045235360287471 (* any transcendent number will do *) | MULTI_A -> 0.6931471805599453094172321214 | MULTI_B -> -0.3665129205816643270124391582 | CONJ -> 0.6019072301972345747375400015 let rec hash = function | Num x -> hash_float (Number.to_float x) | NaN x -> hash_float (transcendent_to_float x) | Load v -> 1 + 1237 * Variable.hash v | Store (v, x) -> 2 * Variable.hash v - 2345 * hash x | Plus l -> 5 + 23451 * sum_list (List.map Hashtbl.hash l) | Times (a, b) -> 41 + 31415 * (Hashtbl.hash a + Hashtbl.hash b) | CTimes (a, b) -> 49 + 3245 * (Hashtbl.hash a + Hashtbl.hash b) | CTimesJ (a, b) -> 31 + 3471 * (Hashtbl.hash a + Hashtbl.hash b) | Uminus x -> 42 + 12345 * (hash x) (* find all variables *) let rec find_vars x = match x with | Load y -> [y] | Plus l -> List.flatten (List.map find_vars l) | Times (a, b) -> (find_vars a) @ (find_vars b) | CTimes (a, b) -> (find_vars a) @ (find_vars b) | CTimesJ (a, b) -> (find_vars a) @ (find_vars b) | Uminus a -> find_vars a | _ -> [] (* TRUE if expression is a constant *) let is_constant = function | Num _ -> true | NaN _ -> true | Load v -> Variable.is_constant v | _ -> false let is_known_constant = function | Num _ -> true | NaN _ -> true | _ -> false (* expr to string, used for debugging *) let rec foldr_string_concat l = match l with [] -> "" | [a] -> a | a :: b -> a ^ " " ^ (foldr_string_concat b) let string_of_transcendent = function | I -> "I" | MULTI_A -> "MULTI_A" | MULTI_B -> "MULTI_B" | CONJ -> "CONJ" let rec to_string = function | Load v -> Variable.unparse v | Num n -> string_of_float (Number.to_float n) | NaN n -> string_of_transcendent n | Plus x -> "(+ " ^ (foldr_string_concat (List.map to_string x)) ^ ")" | Times (a, b) -> "(* " ^ (to_string a) ^ " " ^ (to_string b) ^ ")" | CTimes (a, b) -> "(c* " ^ (to_string a) ^ " " ^ (to_string b) ^ ")" | CTimesJ (a, b) -> "(cj* " ^ (to_string a) ^ " " ^ (to_string b) ^ ")" | Uminus a -> "(- " ^ (to_string a) ^ ")" | Store (v, a) -> "(:= " ^ (Variable.unparse v) ^ " " ^ (to_string a) ^ ")" let rec to_string_a d x = if (d = 0) then "..." else match x with | Load v -> Variable.unparse v | Num n -> Number.to_konst n | NaN n -> string_of_transcendent n | Plus x -> "(+ " ^ (foldr_string_concat (List.map (to_string_a (d - 1)) x)) ^ ")" | Times (a, b) -> "(* " ^ (to_string_a (d - 1) a) ^ " " ^ (to_string_a (d - 1) b) ^ ")" | CTimes (a, b) -> "(c* " ^ (to_string_a (d - 1) a) ^ " " ^ (to_string_a (d - 1) b) ^ ")" | CTimesJ (a, b) -> "(cj* " ^ (to_string_a (d - 1) a) ^ " " ^ (to_string_a (d - 1) b) ^ ")" | Uminus a -> "(- " ^ (to_string_a (d-1) a) ^ ")" | Store (v, a) -> "(:= " ^ (Variable.unparse v) ^ " " ^ (to_string_a (d-1) a) ^ ")" let to_string = to_string_a 10 let assignment_to_string = function | Assign (v, a) -> "(:= " ^ (Variable.unparse v) ^ " " ^ (to_string a) ^ ")" let dump print = List.iter (fun x -> print ((assignment_to_string x) ^ "\n")) (* find all constants in a given expression *) let rec expr_to_constants = function | Num n -> [n] | Plus a -> List.flatten (List.map expr_to_constants a) | Times (a, b) -> (expr_to_constants a) @ (expr_to_constants b) | CTimes (a, b) -> (expr_to_constants a) @ (expr_to_constants b) | CTimesJ (a, b) -> (expr_to_constants a) @ (expr_to_constants b) | Uminus a -> expr_to_constants a | _ -> [] let add_float_key_value list_so_far k = if List.exists (fun k2 -> Number.equal k k2) list_so_far then list_so_far else k :: list_so_far let unique_constants = List.fold_left add_float_key_value [] fftw-3.3.3/genfft/fft.ml0000644000175000017500000002544312054410050011770 00000000000000(* * Copyright (c) 1997-1999 Massachusetts Institute of Technology * Copyright (c) 2003, 2007-11 Matteo Frigo * Copyright (c) 2003, 2007-11 Massachusetts Institute of Technology * * This program is free software; you can redistribute it and/or modify * it under the terms of the GNU General Public License as published by * the Free Software Foundation; either version 2 of the License, or * (at your option) any later version. * * This program is distributed in the hope that it will be useful, * but WITHOUT ANY WARRANTY; without even the implied warranty of * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the * GNU General Public License for more details. * * You should have received a copy of the GNU General Public License * along with this program; if not, write to the Free Software * Foundation, Inc., 51 Franklin Street, Fifth Floor, Boston, MA 02110-1301 USA * *) (* This is the part of the generator that actually computes the FFT in symbolic form *) open Complex open Util (* choose a suitable factor of n *) let choose_factor n = (* first choice: i such that gcd(i, n / i) = 1, i as big as possible *) let choose1 n = let rec loop i f = if (i * i > n) then f else if ((n mod i) == 0 && gcd i (n / i) == 1) then loop (i + 1) i else loop (i + 1) f in loop 1 1 (* second choice: the biggest factor i of n, where i < sqrt(n), if any *) and choose2 n = let rec loop i f = if (i * i > n) then f else if ((n mod i) == 0) then loop (i + 1) i else loop (i + 1) f in loop 1 1 in let i = choose1 n in if (i > 1) then i else choose2 n let is_power_of_two n = (n > 0) && ((n - 1) land n == 0) let rec dft_prime sign n input = let sum filter i = sigma 0 n (fun j -> let coeff = filter (exp n (sign * i * j)) in coeff @* (input j)) in let computation_even = array n (sum identity) and computation_odd = let sumr = array n (sum real) and sumi = array n (sum ((times Complex.i) @@ imag)) in array n (fun i -> if (i = 0) then (* expose some common subexpressions *) input 0 @+ sigma 1 ((n + 1) / 2) (fun j -> input j @+ input (n - j)) else let i' = min i (n - i) in if (i < n - i) then sumr i' @+ sumi i' else sumr i' @- sumi i') in if (n >= !Magic.rader_min) then dft_rader sign n input else if (n == 2) then computation_even else computation_odd and dft_rader sign p input = let half = let one_half = inverse_int 2 in times one_half and make_product n a b = let scale_factor = inverse_int n in array n (fun i -> a i @* (scale_factor @* b i)) in (* generates a convolution using ffts. (all arguments are the same as to gen_convolution, below) *) let gen_convolution_by_fft n a b addtoall = let fft_a = dft 1 n a and fft_b = dft 1 n b in let fft_ab = make_product n fft_a fft_b and dc_term i = if (i == 0) then addtoall else zero in let fft_ab1 = array n (fun i -> fft_ab i @+ dc_term i) and sum = fft_a 0 in let conv = dft (-1) n fft_ab1 in (sum, conv) (* alternate routine for convolution. Seems to work better for small sizes. I have no idea why. *) and gen_convolution_by_fft_alt n a b addtoall = let ap = array n (fun i -> half (a i @+ a ((n - i) mod n))) and am = array n (fun i -> half (a i @- a ((n - i) mod n))) and bp = array n (fun i -> half (b i @+ b ((n - i) mod n))) and bm = array n (fun i -> half (b i @- b ((n - i) mod n))) in let fft_ap = dft 1 n ap and fft_am = dft 1 n am and fft_bp = dft 1 n bp and fft_bm = dft 1 n bm in let fft_abpp = make_product n fft_ap fft_bp and fft_abpm = make_product n fft_ap fft_bm and fft_abmp = make_product n fft_am fft_bp and fft_abmm = make_product n fft_am fft_bm and sum = fft_ap 0 @+ fft_am 0 and dc_term i = if (i == 0) then addtoall else zero in let fft_ab1 = array n (fun i -> (fft_abpp i @+ fft_abmm i) @+ dc_term i) and fft_ab2 = array n (fun i -> fft_abpm i @+ fft_abmp i) in let conv1 = dft (-1) n fft_ab1 and conv2 = dft (-1) n fft_ab2 in let conv = array n (fun i -> conv1 i @+ conv2 i) in (sum, conv) (* generator of assignment list assigning conv to the convolution of a and b, all of which are of length n. addtoall is added to all of the elements of the result. Returns (sum, convolution) pair where sum is the sum of the elements of a. *) in let gen_convolution = if (p <= !Magic.alternate_convolution) then gen_convolution_by_fft_alt else gen_convolution_by_fft (* fft generator for prime n = p using Rader's algorithm for turning the fft into a convolution, which then can be performed in a variety of ways *) in let g = find_generator p in let ginv = pow_mod g (p - 2) p in let input_perm = array p (fun i -> input (pow_mod g i p)) and omega_perm = array p (fun i -> exp p (sign * (pow_mod ginv i p))) and output_perm = array p (fun i -> pow_mod ginv i p) in let (sum, conv) = (gen_convolution (p - 1) input_perm omega_perm (input 0)) in array p (fun i -> if (i = 0) then input 0 @+ sum else let i' = suchthat 0 (fun i' -> i = output_perm i') in conv i') (* our modified version of the conjugate-pair split-radix algorithm, which reduces the number of multiplications by rescaling the sub-transforms (power-of-two n's only) *) and newsplit sign n input = let rec s n k = (* recursive scale factor *) if n <= 4 then one else let k4 = (abs k) mod (n / 4) in let k4' = if k4 <= (n / 8) then k4 else (n/4 - k4) in (s (n / 4) k4') @* (real (exp n k4')) and sinv n k = (* 1 / s(n,k) *) if n <= 4 then one else let k4 = (abs k) mod (n / 4) in let k4' = if k4 <= (n / 8) then k4 else (n/4 - k4) in (sinv (n / 4) k4') @* (sec n k4') in let sdiv2 n k = (s n k) @* (sinv (2*n) k) (* s(n,k) / s(2*n,k) *) and sdiv4 n k = (* s(n,k) / s(4*n,k) *) let k4 = (abs k) mod n in sec (4*n) (if k4 <= (n / 2) then k4 else (n - k4)) in let t n k = (exp n k) @* (sdiv4 (n/4) k) and dft1 input = input and dft2 input = array 2 (fun k -> (input 0) @+ ((input 1) @* exp 2 k)) in let rec newsplit0 sign n input = if (n == 1) then dft1 input else if (n == 2) then dft2 input else let u = newsplit0 sign (n / 2) (fun i -> input (i*2)) and z = newsplitS sign (n / 4) (fun i -> input (i*4 + 1)) and z' = newsplitS sign (n / 4) (fun i -> input ((n + i*4 - 1) mod n)) and twid = array n (fun k -> s (n/4) k @* exp n (sign * k)) in let w = array n (fun k -> twid k @* z (k mod (n / 4))) and w' = array n (fun k -> conj (twid k) @* z' (k mod (n / 4))) in let ww = array n (fun k -> w k @+ w' k) in array n (fun k -> u (k mod (n / 2)) @+ ww k) and newsplitS sign n input = if (n == 1) then dft1 input else if (n == 2) then dft2 input else let u = newsplitS2 sign (n / 2) (fun i -> input (i*2)) and z = newsplitS sign (n / 4) (fun i -> input (i*4 + 1)) and z' = newsplitS sign (n / 4) (fun i -> input ((n + i*4 - 1) mod n)) in let w = array n (fun k -> t n (sign * k) @* z (k mod (n / 4))) and w' = array n (fun k -> conj (t n (sign * k)) @* z' (k mod (n / 4))) in let ww = array n (fun k -> w k @+ w' k) in array n (fun k -> u (k mod (n / 2)) @+ ww k) and newsplitS2 sign n input = if (n == 1) then dft1 input else if (n == 2) then dft2 input else let u = newsplitS4 sign (n / 2) (fun i -> input (i*2)) and z = newsplitS sign (n / 4) (fun i -> input (i*4 + 1)) and z' = newsplitS sign (n / 4) (fun i -> input ((n + i*4 - 1) mod n)) in let w = array n (fun k -> t n (sign * k) @* z (k mod (n / 4))) and w' = array n (fun k -> conj (t n (sign * k)) @* z' (k mod (n / 4))) in let ww = array n (fun k -> (w k @+ w' k) @* (sdiv2 n k)) in array n (fun k -> u (k mod (n / 2)) @+ ww k) and newsplitS4 sign n input = if (n == 1) then dft1 input else if (n == 2) then let f = dft2 input in array 2 (fun k -> (f k) @* (sinv 8 k)) else let u = newsplitS2 sign (n / 2) (fun i -> input (i*2)) and z = newsplitS sign (n / 4) (fun i -> input (i*4 + 1)) and z' = newsplitS sign (n / 4) (fun i -> input ((n + i*4 - 1) mod n)) in let w = array n (fun k -> t n (sign * k) @* z (k mod (n / 4))) and w' = array n (fun k -> conj (t n (sign * k)) @* z' (k mod (n / 4))) in let ww = array n (fun k -> w k @+ w' k) in array n (fun k -> (u (k mod (n / 2)) @+ ww k) @* (sdiv4 n k)) in newsplit0 sign n input and dft sign n input = let rec cooley_tukey sign n1 n2 input = let tmp1 = array n2 (fun i2 -> dft sign n1 (fun i1 -> input (i1 * n2 + i2))) in let tmp2 = array n1 (fun i1 -> array n2 (fun i2 -> exp n (sign * i1 * i2) @* tmp1 i2 i1)) in let tmp3 = array n1 (fun i1 -> dft sign n2 (tmp2 i1)) in (fun i -> tmp3 (i mod n1) (i / n1)) (* * This is "exponent -1" split-radix by Dan Bernstein. *) and split_radix_dit sign n input = let f0 = dft sign (n / 2) (fun i -> input (i * 2)) and f10 = dft sign (n / 4) (fun i -> input (i * 4 + 1)) and f11 = dft sign (n / 4) (fun i -> input ((n + i * 4 - 1) mod n)) in let g10 = array n (fun k -> exp n (sign * k) @* f10 (k mod (n / 4))) and g11 = array n (fun k -> exp n (- sign * k) @* f11 (k mod (n / 4))) in let g1 = array n (fun k -> g10 k @+ g11 k) in array n (fun k -> f0 (k mod (n / 2)) @+ g1 k) and split_radix_dif sign n input = let n2 = n / 2 and n4 = n / 4 in let x0 = array n2 (fun i -> input i @+ input (i + n2)) and x10 = array n4 (fun i -> input i @- input (i + n2)) and x11 = array n4 (fun i -> input (i + n4) @- input (i + n2 + n4)) in let x1 k i = exp n (k * i * sign) @* (x10 i @+ exp 4 (k * sign) @* x11 i) in let f0 = dft sign n2 x0 and f1 = array 4 (fun k -> dft sign n4 (x1 k)) in array n (fun k -> if k mod 2 = 0 then f0 (k / 2) else let k' = k mod 4 in f1 k' ((k - k') / 4)) and prime_factor sign n1 n2 input = let tmp1 = array n2 (fun i2 -> dft sign n1 (fun i1 -> input ((i1 * n2 + i2 * n1) mod n))) in let tmp2 = array n1 (fun i1 -> dft sign n2 (fun k2 -> tmp1 k2 i1)) in fun i -> tmp2 (i mod n1) (i mod n2) in let algorithm sign n = let r = choose_factor n in if List.mem n !Magic.rader_list then (* special cases *) dft_rader sign n else if (r == 1) then (* n is prime *) dft_prime sign n else if (gcd r (n / r)) == 1 then prime_factor sign r (n / r) else if (n mod 4 = 0 && n > 4) then if !Magic.newsplit && is_power_of_two n then newsplit sign n else if !Magic.dif_split_radix then split_radix_dif sign n else split_radix_dit sign n else cooley_tukey sign r (n / r) in array n (algorithm sign n input) fftw-3.3.3/genfft/gen_hc2c.ml0000644000175000017500000001216612054410050012657 00000000000000(* * Copyright (c) 1997-1999 Massachusetts Institute of Technology * Copyright (c) 2003, 2007-11 Matteo Frigo * Copyright (c) 2003, 2007-11 Massachusetts Institute of Technology * * This program is free software; you can redistribute it and/or modify * it under the terms of the GNU General Public License as published by * the Free Software Foundation; either version 2 of the License, or * (at your option) any later version. * * This program is distributed in the hope that it will be useful, * but WITHOUT ANY WARRANTY; without even the implied warranty of * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the * GNU General Public License for more details. * * You should have received a copy of the GNU General Public License * along with this program; if not, write to the Free Software * Foundation, Inc., 51 Franklin Street, Fifth Floor, Boston, MA 02110-1301 USA * *) open Util open Genutil open C type ditdif = DIT | DIF let ditdif = ref DIT let usage = "Usage: " ^ Sys.argv.(0) ^ " -n [ -dit | -dif ]" let urs = ref Stride_variable let speclist = [ "-dit", Arg.Unit(fun () -> ditdif := DIT), " generate a DIT codelet"; "-dif", Arg.Unit(fun () -> ditdif := DIF), " generate a DIF codelet"; "-with-rs", Arg.String(fun x -> urs := arg_to_stride x), " specialize for given R-stride"; ] let byi = Complex.times Complex.i let byui = Complex.times (Complex.uminus Complex.i) let sym n f i = if (i < n - i) then f i else Complex.conj (f i) let shuffle_eo fe fo i = if i mod 2 == 0 then fe (i/2) else fo ((i-1)/2) let generate n = let rs = "rs" and twarray = "W" and m = "m" and mb = "mb" and me = "me" and ms = "ms" (* the array names are from the point of view of the complex array (output in R2C, input in C2R) *) and arp = "Rp" (* real, positive *) and aip = "Ip" (* imag, positive *) and arm = "Rm" (* real, negative *) and aim = "Im" (* imag, negative *) in let sign = !Genutil.sign and name = !Magic.codelet_name and byvl x = choose_simd x (ctimes (CVar "VL", x)) in let (bytwiddle, num_twiddles, twdesc) = Twiddle.twiddle_policy 1 false in let nt = num_twiddles n in let byw = bytwiddle n sign (twiddle_array nt twarray) in let vrs = either_stride (!urs) (C.SVar rs) in (* assume a single location. No point in doing alias analysis *) let the_location = (Unique.make (), Unique.make ()) in let locations _ = the_location in let locr = (locative_array_c n (C.array_subscript arp vrs) (C.array_subscript arm vrs) locations "BUG") and loci = (locative_array_c n (C.array_subscript aip vrs) (C.array_subscript aim vrs) locations "BUG") and locp = (locative_array_c n (C.array_subscript arp vrs) (C.array_subscript aip vrs) locations "BUG") and locm = (locative_array_c n (C.array_subscript arm vrs) (C.array_subscript aim vrs) locations "BUG") in let locri i = if i mod 2 == 0 then locr (i/2) else loci ((i-1)/2) and locpm i = if i < n - i then locp i else locm (n-1-i) in let asch = match !ditdif with | DIT -> let output = Fft.dft sign n (byw (load_array_c n locri)) in let odag = store_array_c n locpm (sym n output) in standard_optimizer odag | DIF -> let output = byw (Fft.dft sign n (sym n (load_array_c n locpm))) in let odag = store_array_c n locri output in standard_optimizer odag in let vms = CVar "ms" and varp = CVar arp and vaip = CVar aip and varm = CVar arm and vaim = CVar aim and vm = CVar m and vmb = CVar mb and vme = CVar me in let body = Block ( [Decl ("INT", m)], [For (list_to_comma [Expr_assign (vm, vmb); Expr_assign (CVar twarray, CPlus [CVar twarray; ctimes (CPlus [vmb; CUminus (Integer 1)], Integer nt)])], Binop (" < ", vm, vme), list_to_comma [Expr_assign (vm, CPlus [vm; byvl (Integer 1)]); Expr_assign (varp, CPlus [varp; byvl vms]); Expr_assign (vaip, CPlus [vaip; byvl vms]); Expr_assign (varm, CPlus [varm; CUminus (byvl vms)]); Expr_assign (vaim, CPlus [vaim; CUminus (byvl vms)]); Expr_assign (CVar twarray, CPlus [CVar twarray; byvl (Integer nt)]); make_volatile_stride (4*n) (CVar rs) ], Asch asch)]) in let tree = Fcn ("static void", name, [Decl (C.realtypep, arp); Decl (C.realtypep, aip); Decl (C.realtypep, arm); Decl (C.realtypep, aim); Decl (C.constrealtypep, twarray); Decl (C.stridetype, rs); Decl ("INT", mb); Decl ("INT", me); Decl ("INT", ms)], finalize_fcn body) in let twinstr = Printf.sprintf "static const tw_instr twinstr[] = %s;\n\n" (twinstr_to_string "VL" (twdesc n)) and desc = Printf.sprintf "static const hc2c_desc desc = {%d, \"%s\", twinstr, &GENUS, %s};\n\n" n name (flops_of tree) and register = "X(khc2c_register)" in let init = "\n" ^ twinstr ^ desc ^ (declare_register_fcn name) ^ (Printf.sprintf "{\n%s(p, %s, &desc, HC2C_VIA_RDFT);\n}" register name) in (unparse tree) ^ "\n" ^ init let main () = begin parse (speclist @ Twiddle.speclist) usage; print_string (generate (check_size ())); end let _ = main() fftw-3.3.3/genfft/gen_hc2cdft.ml0000644000175000017500000001335412054410050013355 00000000000000(* * Copyright (c) 1997-1999 Massachusetts Institute of Technology * Copyright (c) 2003, 2007-11 Matteo Frigo * Copyright (c) 2003, 2007-11 Massachusetts Institute of Technology * * This program is free software; you can redistribute it and/or modify * it under the terms of the GNU General Public License as published by * the Free Software Foundation; either version 2 of the License, or * (at your option) any later version. * * This program is distributed in the hope that it will be useful, * but WITHOUT ANY WARRANTY; without even the implied warranty of * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the * GNU General Public License for more details. * * You should have received a copy of the GNU General Public License * along with this program; if not, write to the Free Software * Foundation, Inc., 51 Franklin Street, Fifth Floor, Boston, MA 02110-1301 USA * *) open Util open Genutil open C type ditdif = DIT | DIF let ditdif = ref DIT let usage = "Usage: " ^ Sys.argv.(0) ^ " -n [ -dit | -dif ]" let urs = ref Stride_variable let ums = ref Stride_variable let speclist = [ "-dit", Arg.Unit(fun () -> ditdif := DIT), " generate a DIT codelet"; "-dif", Arg.Unit(fun () -> ditdif := DIF), " generate a DIF codelet"; "-with-rs", Arg.String(fun x -> urs := arg_to_stride x), " specialize for given R-stride"; "-with-ms", Arg.String(fun x -> ums := arg_to_stride x), " specialize for given ms" ] let byi = Complex.times Complex.i let byui = Complex.times (Complex.uminus Complex.i) let shuffle_eo fe fo i = if i mod 2 == 0 then fe (i/2) else fo ((i-1)/2) let generate n = let rs = "rs" and twarray = "W" and m = "m" and mb = "mb" and me = "me" and ms = "ms" (* the array names are from the point of view of the complex array (output in R2C, input in C2R) *) and arp = "Rp" (* real, positive *) and aip = "Ip" (* imag, positive *) and arm = "Rm" (* real, negative *) and aim = "Im" (* imag, negative *) in let sign = !Genutil.sign and name = !Magic.codelet_name and byvl x = choose_simd x (ctimes (CVar "VL", x)) in let (bytwiddle, num_twiddles, twdesc) = Twiddle.twiddle_policy 1 false in let nt = num_twiddles n in let byw = bytwiddle n sign (twiddle_array nt twarray) in let vrs = either_stride (!urs) (C.SVar rs) in (* assume a single location. No point in doing alias analysis *) let the_location = (Unique.make (), Unique.make ()) in let locations _ = the_location in let rlocp = (locative_array_c n (C.array_subscript arp vrs) (C.array_subscript aip vrs) locations "BUG") and rlocm = (locative_array_c n (C.array_subscript arm vrs) (C.array_subscript aim vrs) locations "BUG") and clocp = (locative_array_c n (C.array_subscript arp vrs) (C.array_subscript aip vrs) locations "BUG") and clocm = (locative_array_c n (C.array_subscript arm vrs) (C.array_subscript aim vrs) locations "BUG") in let rloc i = if i mod 2 == 0 then rlocp (i/2) else rlocm ((i-1)/2) and cloc i = if i < n - i then clocp i else clocm (n-1-i) and sym n f i = if (i < n - i) then f i else Complex.conj (f i) and sym1 f i = if i mod 2 == 0 then Complex.plus [f i; Complex.conj (f (i+1))] else Complex.times (Complex.uminus Complex.i) (Complex.plus [f (i-1); Complex.uminus (Complex.conj (f i))]) and sym1i f i = if i mod 2 == 0 then Complex.plus [f i; Complex.times Complex.i (f (i+1))] else Complex.conj (Complex.plus [f (i-1); Complex.times (Complex.uminus Complex.i) (f i)]) in let asch = match !ditdif with | DIT -> let output = (Complex.times Complex.half) @@ (Fft.dft sign n (byw (sym1 (load_array_c n rloc)))) in let odag = store_array_c n cloc (sym n output) in standard_optimizer odag | DIF -> let output = byw (Fft.dft sign n (sym n (load_array_c n cloc))) in let odag = store_array_c n rloc (sym1i output) in standard_optimizer odag in let vms = CVar "ms" and varp = CVar arp and vaip = CVar aip and varm = CVar arm and vaim = CVar aim and vm = CVar m and vmb = CVar mb and vme = CVar me in let body = Block ( [Decl ("INT", m)], [For (list_to_comma [Expr_assign (vm, vmb); Expr_assign (CVar twarray, CPlus [CVar twarray; ctimes (CPlus [vmb; CUminus (Integer 1)], Integer nt)])], Binop (" < ", vm, vme), list_to_comma [Expr_assign (vm, CPlus [vm; byvl (Integer 1)]); Expr_assign (varp, CPlus [varp; byvl vms]); Expr_assign (vaip, CPlus [vaip; byvl vms]); Expr_assign (varm, CPlus [varm; CUminus (byvl vms)]); Expr_assign (vaim, CPlus [vaim; CUminus (byvl vms)]); Expr_assign (CVar twarray, CPlus [CVar twarray; byvl (Integer nt)]); make_volatile_stride (4*n) (CVar rs) ], Asch asch)] ) in let tree = Fcn ("static void", name, [Decl (C.realtypep, arp); Decl (C.realtypep, aip); Decl (C.realtypep, arm); Decl (C.realtypep, aim); Decl (C.constrealtypep, twarray); Decl (C.stridetype, rs); Decl ("INT", mb); Decl ("INT", me); Decl ("INT", ms)], finalize_fcn body) in let twinstr = Printf.sprintf "static const tw_instr twinstr[] = %s;\n\n" (twinstr_to_string "VL" (twdesc n)) and desc = Printf.sprintf "static const hc2c_desc desc = {%d, \"%s\", twinstr, &GENUS, %s};\n\n" n name (flops_of tree) and register = "X(khc2c_register)" in let init = "\n" ^ twinstr ^ desc ^ (declare_register_fcn name) ^ (Printf.sprintf "{\n%s(p, %s, &desc, HC2C_VIA_DFT);\n}" register name) in (unparse tree) ^ "\n" ^ init let main () = begin parse (speclist @ Twiddle.speclist) usage; print_string (generate (check_size ())); end let _ = main() fftw-3.3.3/genfft/gen_hc2cdft_c.ml0000644000175000017500000001422412054410050013654 00000000000000(* * Copyright (c) 1997-1999 Massachusetts Institute of Technology * Copyright (c) 2003, 2007-11 Matteo Frigo * Copyright (c) 2003, 2007-11 Massachusetts Institute of Technology * * This program is free software; you can redistribute it and/or modify * it under the terms of the GNU General Public License as published by * the Free Software Foundation; either version 2 of the License, or * (at your option) any later version. * * This program is distributed in the hope that it will be useful, * but WITHOUT ANY WARRANTY; without even the implied warranty of * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the * GNU General Public License for more details. * * You should have received a copy of the GNU General Public License * along with this program; if not, write to the Free Software * Foundation, Inc., 51 Franklin Street, Fifth Floor, Boston, MA 02110-1301 USA * *) open Util open Genutil open C type ditdif = DIT | DIF let ditdif = ref DIT let usage = "Usage: " ^ Sys.argv.(0) ^ " -n [ -dit | -dif ]" let urs = ref Stride_variable let ums = ref Stride_variable let speclist = [ "-dit", Arg.Unit(fun () -> ditdif := DIT), " generate a DIT codelet"; "-dif", Arg.Unit(fun () -> ditdif := DIF), " generate a DIF codelet"; "-with-rs", Arg.String(fun x -> urs := arg_to_stride x), " specialize for given R-stride"; "-with-ms", Arg.String(fun x -> ums := arg_to_stride x), " specialize for given ms" ] let byi = Complex.times Complex.i let byui = Complex.times (Complex.uminus Complex.i) let shuffle_eo fe fo i = if i mod 2 == 0 then fe (i/2) else fo ((i-1)/2) let generate n = let rs = "rs" and twarray = "W" and m = "m" and mb = "mb" and me = "me" and ms = "ms" (* the array names are from the point of view of the complex array (output in R2C, input in C2R) *) and arp = "Rp" (* real, positive *) and aip = "Ip" (* imag, positive *) and arm = "Rm" (* real, negative *) and aim = "Im" (* imag, negative *) in let sign = !Genutil.sign and name = !Magic.codelet_name and byvl x = choose_simd x (ctimes (CVar "VL", x)) and bytwvl x = choose_simd x (ctimes (CVar "TWVL", x)) and bytwvl_vl x = choose_simd x (ctimes (CVar "(TWVL/VL)", x)) in let (bytwiddle, num_twiddles, twdesc) = Twiddle.twiddle_policy 1 true in let nt = num_twiddles n in let byw = bytwiddle n sign (twiddle_array nt twarray) in let vrs = either_stride (!urs) (C.SVar rs) in let sms = stride_to_string "ms" !ums in let msms = "-" ^ sms in (* assume a single location. No point in doing alias analysis *) let the_location = (Unique.make (), Unique.make ()) in let locations _ = the_location in let rlocp = (locative_array_c n (C.array_subscript arp vrs) (C.array_subscript aip vrs) locations sms) and rlocm = (locative_array_c n (C.array_subscript arm vrs) (C.array_subscript aim vrs) locations msms) and clocp = (locative_array_c n (C.array_subscript arp vrs) (C.array_subscript aip vrs) locations sms) and clocm = (locative_array_c n (C.array_subscript arm vrs) (C.array_subscript aim vrs) locations msms) in let rloc i = if i mod 2 == 0 then rlocp (i/2) else rlocm ((i-1)/2) and cloc i = if i < n - i then clocp i else clocm (n-1-i) and sym n f i = if (i < n - i) then f i else Complex.times (Complex.nan Expr.CONJ) (f i) and sym1 f i = if i mod 2 == 0 then Complex.plus [f i; Complex.times (Complex.nan Expr.CONJ) (f (i+1))] else Complex.times (Complex.nan Expr.I) (Complex.plus [Complex.uminus (f (i-1)); Complex.times (Complex.nan Expr.CONJ) (f i)]) and sym1i f i = if i mod 2 == 0 then Complex.plus [f i; Complex.times (Complex.nan Expr.I) (f (i+1))] else Complex.times (Complex.nan Expr.CONJ) (Complex.plus [f (i-1); Complex.uminus (Complex.times (Complex.nan Expr.I) (f i))]) in let asch = match !ditdif with | DIT -> let output = (Complex.times Complex.half) @@ (Trig.dft_via_rdft sign n (byw (sym1 (load_array_r n rloc)))) in let odag = store_array_r n cloc (sym n output) in standard_optimizer odag | DIF -> let output = byw (Trig.dft_via_rdft sign n (sym n (load_array_r n cloc))) in let odag = store_array_r n rloc (sym1i output) in standard_optimizer odag in let vms = CVar sms and varp = CVar arp and vaip = CVar aip and varm = CVar arm and vaim = CVar aim and vm = CVar m and vmb = CVar mb and vme = CVar me in let body = Block ( [Decl ("INT", m)], [For (list_to_comma [Expr_assign (vm, vmb); Expr_assign (CVar twarray, CPlus [CVar twarray; ctimes (CPlus [vmb; CUminus (Integer 1)], bytwvl_vl (Integer nt))])], Binop (" < ", vm, vme), list_to_comma [Expr_assign (vm, CPlus [vm; byvl (Integer 1)]); Expr_assign (varp, CPlus [varp; byvl vms]); Expr_assign (vaip, CPlus [vaip; byvl vms]); Expr_assign (varm, CPlus [varm; CUminus (byvl vms)]); Expr_assign (vaim, CPlus [vaim; CUminus (byvl vms)]); Expr_assign (CVar twarray, CPlus [CVar twarray; bytwvl (Integer nt)]); make_volatile_stride (4*n) (CVar rs) ], Asch asch)] ) in let tree = Fcn ("static void", name, [Decl (C.realtypep, arp); Decl (C.realtypep, aip); Decl (C.realtypep, arm); Decl (C.realtypep, aim); Decl (C.constrealtypep, twarray); Decl (C.stridetype, rs); Decl ("INT", mb); Decl ("INT", me); Decl ("INT", ms)], finalize_fcn body) in let twinstr = Printf.sprintf "static const tw_instr twinstr[] = %s;\n\n" (twinstr_to_string "VL" (twdesc n)) and desc = Printf.sprintf "static const hc2c_desc desc = {%d, %s, twinstr, &GENUS, %s};\n\n" n (stringify name) (flops_of tree) and register = "X(khc2c_register)" in let init = "\n" ^ twinstr ^ desc ^ (declare_register_fcn name) ^ (Printf.sprintf "{\n%s(p, %s, &desc, HC2C_VIA_DFT);\n}" register name) in (unparse tree) ^ "\n" ^ init let main () = begin Simdmagic.simd_mode := true; parse (speclist @ Twiddle.speclist) usage; print_string (generate (check_size ())); end let _ = main() fftw-3.3.3/genfft/gen_hc2hc.ml0000644000175000017500000001106112054410050013020 00000000000000(* * Copyright (c) 1997-1999 Massachusetts Institute of Technology * Copyright (c) 2003, 2007-11 Matteo Frigo * Copyright (c) 2003, 2007-11 Massachusetts Institute of Technology * * This program is free software; you can redistribute it and/or modify * it under the terms of the GNU General Public License as published by * the Free Software Foundation; either version 2 of the License, or * (at your option) any later version. * * This program is distributed in the hope that it will be useful, * but WITHOUT ANY WARRANTY; without even the implied warranty of * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the * GNU General Public License for more details. * * You should have received a copy of the GNU General Public License * along with this program; if not, write to the Free Software * Foundation, Inc., 51 Franklin Street, Fifth Floor, Boston, MA 02110-1301 USA * *) open Util open Genutil open C type ditdif = DIT | DIF let ditdif = ref DIT let usage = "Usage: " ^ Sys.argv.(0) ^ " -n [ -dit | -dif ]" let urs = ref Stride_variable let speclist = [ "-dit", Arg.Unit(fun () -> ditdif := DIT), " generate a DIT codelet"; "-dif", Arg.Unit(fun () -> ditdif := DIF), " generate a DIF codelet"; "-with-rs", Arg.String(fun x -> urs := arg_to_stride x), " specialize for given R-stride"; ] let rioarray = "cr" and iioarray = "ci" let genone sign n transform load store vrs = let locations = unique_array_c n in let input = locative_array_c n (C.array_subscript rioarray vrs) (C.array_subscript iioarray vrs) locations "BUG" in let output = transform sign n (load n input) in let ioloc = locative_array_c n (C.array_subscript rioarray vrs) (C.array_subscript iioarray vrs) locations "BUG" in let odag = store n ioloc output in let annot = standard_optimizer odag in annot let byi = Complex.times Complex.i let byui = Complex.times (Complex.uminus Complex.i) let sym1 n f i = Complex.plus [Complex.real (f i); byi (Complex.imag (f (n - 1 - i)))] let sym2 n f i = if (i < n - i) then f i else byi (f i) let sym2i n f i = if (i < n - i) then f i else byui (f i) let generate n = let rs = "rs" and twarray = "W" and m = "m" and mb = "mb" and me = "me" and ms = "ms" in let sign = !Genutil.sign and name = !Magic.codelet_name and byvl x = choose_simd x (ctimes (CVar "VL", x)) in let (bytwiddle, num_twiddles, twdesc) = Twiddle.twiddle_policy 1 false in let nt = num_twiddles n in let byw = bytwiddle n sign (twiddle_array nt twarray) in let vrs = either_stride (!urs) (C.SVar rs) in let asch = match !ditdif with | DIT -> genone sign n (fun sign n input -> ((sym1 n) @@ (sym2 n)) (Fft.dft sign n (byw input))) load_array_c store_array_c vrs | DIF -> genone sign n (fun sign n input -> byw (Fft.dft sign n (((sym2i n) @@ (sym1 n)) input))) load_array_c store_array_c vrs in let vms = CVar "ms" and vrioarray = CVar rioarray and viioarray = CVar iioarray and vm = CVar m and vmb = CVar mb and vme = CVar me in let body = Block ( [Decl ("INT", m)], [For (list_to_comma [Expr_assign (vm, vmb); Expr_assign (CVar twarray, CPlus [CVar twarray; ctimes (CPlus [vmb; CUminus (Integer 1)], Integer nt)])], Binop (" < ", vm, vme), list_to_comma [Expr_assign (vm, CPlus [vm; byvl (Integer 1)]); Expr_assign (vrioarray, CPlus [vrioarray; byvl vms]); Expr_assign (viioarray, CPlus [viioarray; CUminus (byvl vms)]); Expr_assign (CVar twarray, CPlus [CVar twarray; byvl (Integer nt)]); make_volatile_stride (2*n) (CVar rs) ], Asch asch)]) in let tree = Fcn ("static void", name, [Decl (C.realtypep, rioarray); Decl (C.realtypep, iioarray); Decl (C.constrealtypep, twarray); Decl (C.stridetype, rs); Decl ("INT", mb); Decl ("INT", me); Decl ("INT", ms)], finalize_fcn body) in let twinstr = Printf.sprintf "static const tw_instr twinstr[] = %s;\n\n" (twinstr_to_string "VL" (twdesc n)) and desc = Printf.sprintf "static const hc2hc_desc desc = {%d, \"%s\", twinstr, &GENUS, %s};\n\n" n name (flops_of tree) and register = "X(khc2hc_register)" in let init = "\n" ^ twinstr ^ desc ^ (declare_register_fcn name) ^ (Printf.sprintf "{\n%s(p, %s, &desc);\n}" register name) in (unparse tree) ^ "\n" ^ init let main () = begin parse (speclist @ Twiddle.speclist) usage; print_string (generate (check_size ())); end let _ = main() fftw-3.3.3/genfft/gen_r2cb.ml0000644000175000017500000001131212054410050012660 00000000000000(* * Copyright (c) 1997-1999 Massachusetts Institute of Technology * Copyright (c) 2003, 2007-11 Matteo Frigo * Copyright (c) 2003, 2007-11 Massachusetts Institute of Technology * * This program is free software; you can redistribute it and/or modify * it under the terms of the GNU General Public License as published by * the Free Software Foundation; either version 2 of the License, or * (at your option) any later version. * * This program is distributed in the hope that it will be useful, * but WITHOUT ANY WARRANTY; without even the implied warranty of * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the * GNU General Public License for more details. * * You should have received a copy of the GNU General Public License * along with this program; if not, write to the Free Software * Foundation, Inc., 51 Franklin Street, Fifth Floor, Boston, MA 02110-1301 USA * *) open Util open Genutil open C let usage = "Usage: " ^ Sys.argv.(0) ^ " -n " let urs = ref Stride_variable let ucsr = ref Stride_variable let ucsi = ref Stride_variable let uivs = ref Stride_variable let uovs = ref Stride_variable let dftIII_flag = ref false let speclist = [ "-with-rs", Arg.String(fun x -> urs := arg_to_stride x), " specialize for given real-array stride"; "-with-csr", Arg.String(fun x -> ucsr := arg_to_stride x), " specialize for given complex-array real stride"; "-with-csi", Arg.String(fun x -> ucsi := arg_to_stride x), " specialize for given complex-array imaginary stride"; "-with-ivs", Arg.String(fun x -> uivs := arg_to_stride x), " specialize for given input vector stride"; "-with-ovs", Arg.String(fun x -> uovs := arg_to_stride x), " specialize for given output vector stride"; "-dft-III", Arg.Unit(fun () -> dftIII_flag := true), " produce shifted dftIII-style codelets" ] let hcdftIII sign n input = let input' i = if (i mod 2 == 0) then Complex.zero else let i' = (i - 1) / 2 in if (2 * i' < n - 1) then (input i') else if (2 * i' == n - 1) then Complex.real (input i') else Complex.conj (input (n - 1 - i')) in Fft.dft sign (2 * n) input' let generate n = let ar0 = "R0" and ar1 = "R1" and acr = "Cr" and aci = "Ci" and rs = "rs" and csr = "csr" and csi = "csi" and i = "i" and v = "v" and transform = if !dftIII_flag then hcdftIII else Trig.hdft in let sign = !Genutil.sign and name = !Magic.codelet_name in let vrs = either_stride (!urs) (C.SVar rs) and vcsr = either_stride (!ucsr) (C.SVar csr) and vcsi = either_stride (!ucsi) (C.SVar csi) in let sovs = stride_to_string "ovs" !uovs in let sivs = stride_to_string "ivs" !uivs in let locations = unique_array_c n in let input = locative_array_c n (C.array_subscript acr vcsr) (C.array_subscript aci vcsi) locations sivs in let output = transform sign n (load_array_hc n input) in let oloce = locative_array_c n (C.array_subscript ar0 vrs) (C.array_subscript "BUG" vrs) locations sovs and oloco = locative_array_c n (C.array_subscript ar1 vrs) (C.array_subscript "BUG" vrs) locations sovs in let oloc i = if i mod 2 == 0 then oloce (i/2) else oloco ((i-1)/2) in let odag = store_array_r n oloc output in let annot = standard_optimizer odag in let body = Block ( [Decl ("INT", i)], [For (Expr_assign (CVar i, CVar v), Binop (" > ", CVar i, Integer 0), list_to_comma [Expr_assign (CVar i, CPlus [CVar i; CUminus (Integer 1)]); Expr_assign (CVar ar0, CPlus [CVar ar0; CVar sovs]); Expr_assign (CVar ar1, CPlus [CVar ar1; CVar sovs]); Expr_assign (CVar acr, CPlus [CVar acr; CVar sivs]); Expr_assign (CVar aci, CPlus [CVar aci; CVar sivs]); make_volatile_stride (4*n) (CVar rs); make_volatile_stride (4*n) (CVar csr); make_volatile_stride (4*n) (CVar csi) ], Asch annot) ]) in let tree = Fcn ((if !Magic.standalone then "void" else "static void"), name, ([Decl (C.realtypep, ar0); Decl (C.realtypep, ar1); Decl (C.realtypep, acr); Decl (C.realtypep, aci); Decl (C.stridetype, rs); Decl (C.stridetype, csr); Decl (C.stridetype, csi); Decl ("INT", v); Decl ("INT", "ivs"); Decl ("INT", "ovs")]), finalize_fcn body) in let desc = Printf.sprintf "static const kr2c_desc desc = { %d, \"%s\", %s, &GENUS };\n\n" n name (flops_of tree) and init = (declare_register_fcn name) ^ "{" ^ " X(kr2c_register)(p, " ^ name ^ ", &desc);\n" ^ "}\n" in (unparse tree) ^ "\n" ^ (if !Magic.standalone then "" else desc ^ init) let main () = begin parse speclist usage; print_string (generate (check_size ())); end let _ = main() fftw-3.3.3/genfft/gen_mdct.ml0000644000175000017500000001672212054410050012771 00000000000000(* * Copyright (c) 1997-1999 Massachusetts Institute of Technology * Copyright (c) 2003, 2007-11 Matteo Frigo * Copyright (c) 2003, 2007-11 Massachusetts Institute of Technology * * This program is free software; you can redistribute it and/or modify * it under the terms of the GNU General Public License as published by * the Free Software Foundation; either version 2 of the License, or * (at your option) any later version. * * This program is distributed in the hope that it will be useful, * but WITHOUT ANY WARRANTY; without even the implied warranty of * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the * GNU General Public License for more details. * * You should have received a copy of the GNU General Public License * along with this program; if not, write to the Free Software * Foundation, Inc., 51 Franklin Street, Fifth Floor, Boston, MA 02110-1301 USA * *) (* generation of trigonometric transforms *) open Util open Genutil open C let usage = "Usage: " ^ Sys.argv.(0) ^ " -n " let uistride = ref Stride_variable let uostride = ref Stride_variable let uivstride = ref Stride_variable let uovstride = ref Stride_variable let normalization = ref 1 type mode = | MDCT | MDCT_MP3 | MDCT_VORBIS | MDCT_WINDOW | MDCT_WINDOW_SYM | IMDCT | IMDCT_MP3 | IMDCT_VORBIS | IMDCT_WINDOW | IMDCT_WINDOW_SYM | NONE let mode = ref NONE let speclist = [ "-with-istride", Arg.String(fun x -> uistride := arg_to_stride x), " specialize for given input stride"; "-with-ostride", Arg.String(fun x -> uostride := arg_to_stride x), " specialize for given output stride"; "-with-ivstride", Arg.String(fun x -> uivstride := arg_to_stride x), " specialize for given input vector stride"; "-with-ovstride", Arg.String(fun x -> uovstride := arg_to_stride x), " specialize for given output vector stride"; "-normalization", Arg.String(fun x -> normalization := int_of_string x), " normalization integer to divide by"; "-mdct", Arg.Unit(fun () -> mode := MDCT), " generate an MDCT codelet"; "-mdct-mp3", Arg.Unit(fun () -> mode := MDCT_MP3), " generate an MDCT codelet with MP3 windowing"; "-mdct-window", Arg.Unit(fun () -> mode := MDCT_WINDOW), " generate an MDCT codelet with window array"; "-mdct-window-sym", Arg.Unit(fun () -> mode := MDCT_WINDOW_SYM), " generate an MDCT codelet with symmetric window array"; "-imdct", Arg.Unit(fun () -> mode := IMDCT), " generate an IMDCT codelet"; "-imdct-mp3", Arg.Unit(fun () -> mode := IMDCT_MP3), " generate an IMDCT codelet with MP3 windowing"; "-imdct-window", Arg.Unit(fun () -> mode := IMDCT_WINDOW), " generate an IMDCT codelet with window array"; "-imdct-window-sym", Arg.Unit(fun () -> mode := IMDCT_WINDOW_SYM), " generate an IMDCT codelet with symmetric window array"; ] let unity_window n i = Complex.one (* MP3 window(k) = sin(pi/(2n) * (k + 1/2)) *) let mp3_window n k = Complex.imag (Complex.exp (8 * n) (2*k + 1)) (* Vorbis window(k) = sin(pi/2 * (mp3_window(k))^2) ... this is transcendental, though, so we can't do it with our current Complex.exp function *) let window_array n w = array n (fun i -> let stride = C.SInteger 1 and klass = Unique.make () in let refr = C.array_subscript w stride i in let kr = Variable.make_constant klass refr in load_r (kr, kr)) let load_window w n i = w i let load_window_sym w n i = w (if (i < n) then i else (2*n - 1 - i)) (* fixme: use same locations for input and output so that it works in-place? *) (* Note: only correct for even n! *) let load_array_mdct window n rarr iarr locations = let twon = 2 * n in let arr = load_array_c twon (locative_array_c twon rarr iarr locations "BUG") in let arrw = fun i -> Complex.times (window n i) (arr i) in array n ((Complex.times Complex.half) @@ (fun i -> if (i < n/2) then Complex.uminus (Complex.plus [arrw (i + n + n/2); arrw (n + n/2 - 1 - i)]) else Complex.plus [arrw (i - n/2); Complex.uminus (arrw (n + n/2 - 1 - i))])) let store_array_mdct window n rarr iarr locations arr = store_array_r n (locative_array_c n rarr iarr locations "BUG") arr let load_array_imdct window n rarr iarr locations = load_array_c n (locative_array_c n rarr iarr locations "BUG") let store_array_imdct window n rarr iarr locations arr = let n2 = n/2 in let threen2 = 3*n2 in let arr2 = fun i -> if (i < n2) then arr (i + n2) else if (i < threen2) then Complex.uminus (arr (threen2 - 1 - i)) else Complex.uminus (arr (i - threen2)) in let arr2w = fun i -> Complex.times (window n i) (arr2 i) in let twon = 2 * n in store_array_r twon (locative_array_c twon rarr iarr locations "BUG") arr2w let window_param = function MDCT_WINDOW -> true | MDCT_WINDOW_SYM -> true | IMDCT_WINDOW -> true | IMDCT_WINDOW_SYM -> true | _ -> false let generate n mode = let iarray = "I" and oarray = "O" and istride = "istride" and ostride = "ostride" and window = "W" and name = !Magic.codelet_name in let vistride = either_stride (!uistride) (C.SVar istride) and vostride = either_stride (!uostride) (C.SVar ostride) in let sivs = stride_to_string "ovs" !uovstride in let sovs = stride_to_string "ivs" !uivstride in let (transform, load_input, store_output) = match mode with | MDCT -> Trig.dctIV, load_array_mdct unity_window, store_array_mdct unity_window | MDCT_MP3 -> Trig.dctIV, load_array_mdct mp3_window, store_array_mdct unity_window | MDCT_WINDOW -> Trig.dctIV, load_array_mdct (load_window (window_array (2 * n) window)), store_array_mdct unity_window | MDCT_WINDOW_SYM -> Trig.dctIV, load_array_mdct (load_window_sym (window_array n window)), store_array_mdct unity_window | IMDCT -> Trig.dctIV, load_array_imdct unity_window, store_array_imdct unity_window | IMDCT_MP3 -> Trig.dctIV, load_array_imdct unity_window, store_array_imdct mp3_window | IMDCT_WINDOW -> Trig.dctIV, load_array_imdct unity_window, store_array_imdct (load_window (window_array (2 * n) window)) | IMDCT_WINDOW_SYM -> Trig.dctIV, load_array_imdct unity_window, store_array_imdct (load_window_sym (window_array n window)) | _ -> failwith "must specify transform kind" in let locations = unique_array_c (2*n) in let input = load_input n (C.array_subscript iarray vistride) (C.array_subscript "BUG" vistride) locations in let output = (Complex.times (Complex.inverse_int !normalization)) @@ (transform n input) in let odag = store_output n (C.array_subscript oarray vostride) (C.array_subscript "BUG" vostride) locations output in let annot = standard_optimizer odag in let tree = Fcn ("void", name, ([Decl (C.constrealtypep, iarray); Decl (C.realtypep, oarray)] @ (if stride_fixed !uistride then [] else [Decl (C.stridetype, istride)]) @ (if stride_fixed !uostride then [] else [Decl (C.stridetype, ostride)]) @ (choose_simd [] (if stride_fixed !uivstride then [] else [Decl ("int", sivs)])) @ (choose_simd [] (if stride_fixed !uovstride then [] else [Decl ("int", sovs)])) @ (if (not (window_param mode)) then [] else [Decl (C.constrealtypep, window)]) ), finalize_fcn (Asch annot)) in (unparse tree) ^ "\n" let main () = begin parse speclist usage; print_string (generate (check_size ()) !mode); end let _ = main() fftw-3.3.3/genfft/gen_notw.ml0000644000175000017500000001170112054410050013021 00000000000000(* * Copyright (c) 1997-1999 Massachusetts Institute of Technology * Copyright (c) 2003, 2007-11 Matteo Frigo * Copyright (c) 2003, 2007-11 Massachusetts Institute of Technology * * This program is free software; you can redistribute it and/or modify * it under the terms of the GNU General Public License as published by * the Free Software Foundation; either version 2 of the License, or * (at your option) any later version. * * This program is distributed in the hope that it will be useful, * but WITHOUT ANY WARRANTY; without even the implied warranty of * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the * GNU General Public License for more details. * * You should have received a copy of the GNU General Public License * along with this program; if not, write to the Free Software * Foundation, Inc., 51 Franklin Street, Fifth Floor, Boston, MA 02110-1301 USA * *) open Util open Genutil open C let usage = "Usage: " ^ Sys.argv.(0) ^ " -n " let uistride = ref Stride_variable let uostride = ref Stride_variable let uivstride = ref Stride_variable let uovstride = ref Stride_variable let speclist = [ "-with-istride", Arg.String(fun x -> uistride := arg_to_stride x), " specialize for given input stride"; "-with-ostride", Arg.String(fun x -> uostride := arg_to_stride x), " specialize for given output stride"; "-with-ivstride", Arg.String(fun x -> uivstride := arg_to_stride x), " specialize for given input vector stride"; "-with-ovstride", Arg.String(fun x -> uovstride := arg_to_stride x), " specialize for given output vector stride" ] let nonstandard_optimizer list_of_buddy_stores dag = let sched = standard_scheduler dag in let annot = Annotate.annotate list_of_buddy_stores sched in let _ = dump_asched annot in annot let generate n = let riarray = "ri" and iiarray = "ii" and roarray = "ro" and ioarray = "io" and istride = "is" and ostride = "os" and i = "i" and v = "v" in let sign = !Genutil.sign and name = !Magic.codelet_name and byvl x = choose_simd x (ctimes (CVar "(2 * VL)", x)) in let ename = expand_name name in let vistride = either_stride (!uistride) (C.SVar istride) and vostride = either_stride (!uostride) (C.SVar ostride) in let sovs = stride_to_string "ovs" !uovstride in let sivs = stride_to_string "ivs" !uivstride in let locations = unique_array_c n in let input = locative_array_c n (C.array_subscript riarray vistride) (C.array_subscript iiarray vistride) locations sivs in let output = Fft.dft sign n (load_array_c n input) in let oloc = locative_array_c n (C.array_subscript roarray vostride) (C.array_subscript ioarray vostride) locations sovs in let list_of_buddy_stores = let k = !Simdmagic.store_multiple in if (k > 1) then if (n mod k == 0) then List.append (List.map (fun i -> List.map (fun j -> (fst (oloc (k * i + j)))) (iota k)) (iota (n / k))) (List.map (fun i -> List.map (fun j -> (snd (oloc (k * i + j)))) (iota k)) (iota (n / k))) else failwith "invalid n for -store-multiple" else [] in let odag = store_array_c n oloc output in let annot = nonstandard_optimizer list_of_buddy_stores odag in let body = Block ( [Decl ("INT", i)], [For (Expr_assign (CVar i, CVar v), Binop (" > ", CVar i, Integer 0), list_to_comma [Expr_assign (CVar i, CPlus [CVar i; CUminus (byvl (Integer 1))]); Expr_assign (CVar riarray, CPlus [CVar riarray; byvl (CVar sivs)]); Expr_assign (CVar iiarray, CPlus [CVar iiarray; byvl (CVar sivs)]); Expr_assign (CVar roarray, CPlus [CVar roarray; byvl (CVar sovs)]); Expr_assign (CVar ioarray, CPlus [CVar ioarray; byvl (CVar sovs)]); make_volatile_stride (4*n) (CVar istride); make_volatile_stride (4*n) (CVar ostride) ], Asch annot) ]) in let tree = Fcn ((if !Magic.standalone then "void" else "static void"), ename, ([Decl (C.constrealtypep, riarray); Decl (C.constrealtypep, iiarray); Decl (C.realtypep, roarray); Decl (C.realtypep, ioarray); Decl (C.stridetype, istride); Decl (C.stridetype, ostride); Decl ("INT", v); Decl ("INT", "ivs"); Decl ("INT", "ovs")]), finalize_fcn body) in let desc = Printf.sprintf "static const kdft_desc desc = { %d, %s, %s, &GENUS, %s, %s, %s, %s };\n" n (stringify name) (flops_of tree) (stride_to_solverparm !uistride) (stride_to_solverparm !uostride) (choose_simd "0" (stride_to_solverparm !uivstride)) (choose_simd "0" (stride_to_solverparm !uovstride)) and init = (declare_register_fcn name) ^ "{" ^ " X(kdft_register)(p, " ^ ename ^ ", &desc);\n" ^ "}\n" in ((unparse tree) ^ "\n" ^ (if !Magic.standalone then "" else desc ^ init)) let main () = begin parse speclist usage; print_string (generate (check_size ())); end let _ = main() fftw-3.3.3/genfft/gen_notw_c.ml0000644000175000017500000001156312054410050013331 00000000000000(* * Copyright (c) 1997-1999 Massachusetts Institute of Technology * Copyright (c) 2003, 2007-11 Matteo Frigo * Copyright (c) 2003, 2007-11 Massachusetts Institute of Technology * * This program is free software; you can redistribute it and/or modify * it under the terms of the GNU General Public License as published by * the Free Software Foundation; either version 2 of the License, or * (at your option) any later version. * * This program is distributed in the hope that it will be useful, * but WITHOUT ANY WARRANTY; without even the implied warranty of * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the * GNU General Public License for more details. * * You should have received a copy of the GNU General Public License * along with this program; if not, write to the Free Software * Foundation, Inc., 51 Franklin Street, Fifth Floor, Boston, MA 02110-1301 USA * *) open Util open Genutil open C let usage = "Usage: " ^ Sys.argv.(0) ^ " -n " let uistride = ref Stride_variable let uostride = ref Stride_variable let uivstride = ref Stride_variable let uovstride = ref Stride_variable let speclist = [ "-with-istride", Arg.String(fun x -> uistride := arg_to_stride x), " specialize for given input stride"; "-with-ostride", Arg.String(fun x -> uostride := arg_to_stride x), " specialize for given output stride"; "-with-ivstride", Arg.String(fun x -> uivstride := arg_to_stride x), " specialize for given input vector stride"; "-with-ovstride", Arg.String(fun x -> uovstride := arg_to_stride x), " specialize for given output vector stride" ] let nonstandard_optimizer list_of_buddy_stores dag = let sched = standard_scheduler dag in let annot = Annotate.annotate list_of_buddy_stores sched in let _ = dump_asched annot in annot let generate n = let riarray = "xi" and roarray = "xo" and istride = "is" and ostride = "os" and i = "i" and v = "v" in let sign = !Genutil.sign and name = !Magic.codelet_name and byvl x = choose_simd x (ctimes (CVar "VL", x)) in let ename = expand_name name in let vistride = either_stride (!uistride) (C.SVar istride) and vostride = either_stride (!uostride) (C.SVar ostride) in let sivs = stride_to_string "ivs" !uivstride in let sovs = stride_to_string "ovs" !uovstride in let fft = Trig.dft_via_rdft in let locations = unique_array_c n in let input = locative_array_c n (C.array_subscript riarray vistride) (C.array_subscript "BUG" vistride) locations sivs in let output = fft sign n (load_array_r n input) in let oloc = locative_array_c n (C.array_subscript roarray vostride) (C.array_subscript "BUG" vostride) locations sovs in let list_of_buddy_stores = let k = !Simdmagic.store_multiple in if (k > 1) then if (n mod k == 0) then List.map (fun i -> List.map (fun j -> (fst (oloc (k * i + j)))) (iota k)) (iota (n / k)) else failwith "invalid n for -store-multiple" else [] in let odag = store_array_r n oloc output in let annot = nonstandard_optimizer list_of_buddy_stores odag in let body = Block ( [Decl ("INT", i); Decl (C.constrealtypep, riarray); Decl (C.realtypep, roarray)], [Stmt_assign (CVar riarray, CVar (if (sign < 0) then "ri" else "ii")); Stmt_assign (CVar roarray, CVar (if (sign < 0) then "ro" else "io")); For (Expr_assign (CVar i, CVar v), Binop (" > ", CVar i, Integer 0), list_to_comma [Expr_assign (CVar i, CPlus [CVar i; CUminus (byvl (Integer 1))]); Expr_assign (CVar riarray, CPlus [CVar riarray; byvl (CVar sivs)]); Expr_assign (CVar roarray, CPlus [CVar roarray; byvl (CVar sovs)]); make_volatile_stride (2*n) (CVar istride); make_volatile_stride (2*n) (CVar ostride) ], Asch annot); ]) in let tree = Fcn ((if !Magic.standalone then "void" else "static void"), ename, ([Decl (C.constrealtypep, "ri"); Decl (C.constrealtypep, "ii"); Decl (C.realtypep, "ro"); Decl (C.realtypep, "io"); Decl (C.stridetype, istride); Decl (C.stridetype, ostride); Decl ("INT", v); Decl ("INT", "ivs"); Decl ("INT", "ovs")]), finalize_fcn body) in let desc = Printf.sprintf "static const kdft_desc desc = { %d, %s, %s, &GENUS, %s, %s, %s, %s };\n" n (stringify name) (flops_of tree) (stride_to_solverparm !uistride) (stride_to_solverparm !uostride) (choose_simd "0" (stride_to_solverparm !uivstride)) (choose_simd "0" (stride_to_solverparm !uovstride)) and init = (declare_register_fcn name) ^ "{" ^ " X(kdft_register)(p, " ^ ename ^ ", &desc);\n" ^ "}\n" in ((unparse tree) ^ "\n" ^ (if !Magic.standalone then "" else desc ^ init)) let main () = begin Simdmagic.simd_mode := true; parse speclist usage; print_string (generate (check_size ())); end let _ = main() fftw-3.3.3/genfft/gen_r2cf.ml0000644000175000017500000001123312054410050012666 00000000000000(* * Copyright (c) 1997-1999 Massachusetts Institute of Technology * Copyright (c) 2003, 2007-11 Matteo Frigo * Copyright (c) 2003, 2007-11 Massachusetts Institute of Technology * * This program is free software; you can redistribute it and/or modify * it under the terms of the GNU General Public License as published by * the Free Software Foundation; either version 2 of the License, or * (at your option) any later version. * * This program is distributed in the hope that it will be useful, * but WITHOUT ANY WARRANTY; without even the implied warranty of * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the * GNU General Public License for more details. * * You should have received a copy of the GNU General Public License * along with this program; if not, write to the Free Software * Foundation, Inc., 51 Franklin Street, Fifth Floor, Boston, MA 02110-1301 USA * *) open Util open Genutil open C let usage = "Usage: " ^ Sys.argv.(0) ^ " -n " let urs = ref Stride_variable let ucsr = ref Stride_variable let ucsi = ref Stride_variable let uivs = ref Stride_variable let uovs = ref Stride_variable let dftII_flag = ref false let speclist = [ "-with-rs", Arg.String(fun x -> urs := arg_to_stride x), " specialize for given real-array stride"; "-with-csr", Arg.String(fun x -> ucsr := arg_to_stride x), " specialize for given complex-array real stride"; "-with-csi", Arg.String(fun x -> ucsi := arg_to_stride x), " specialize for given complex-array imaginary stride"; "-with-ivs", Arg.String(fun x -> uivs := arg_to_stride x), " specialize for given input vector stride"; "-with-ovs", Arg.String(fun x -> uovs := arg_to_stride x), " specialize for given output vector stride"; "-dft-II", Arg.Unit(fun () -> dftII_flag := true), " produce shifted dftII-style codelets" ] let rdftII sign n input = let input' i = if i < n then input i else Complex.zero in let f = Fft.dft sign (2 * n) input' in let g i = f (2 * i + 1) in fun i -> if (i < n - i) then g i else if (2 * i + 1 == n) then Complex.real (g i) else Complex.zero let generate n = let ar0 = "R0" and ar1 = "R1" and acr = "Cr" and aci = "Ci" and rs = "rs" and csr = "csr" and csi = "csi" and i = "i" and v = "v" and transform = if !dftII_flag then rdftII else Trig.rdft in let sign = !Genutil.sign and name = !Magic.codelet_name in let vrs = either_stride (!urs) (C.SVar rs) and vcsr = either_stride (!ucsr) (C.SVar csr) and vcsi = either_stride (!ucsi) (C.SVar csi) in let sovs = stride_to_string "ovs" !uovs in let sivs = stride_to_string "ivs" !uivs in let locations = unique_array_c n in let inpute = locative_array_c n (C.array_subscript ar0 vrs) (C.array_subscript "BUG" vrs) locations sivs and inputo = locative_array_c n (C.array_subscript ar1 vrs) (C.array_subscript "BUG" vrs) locations sivs in let input i = if i mod 2 == 0 then inpute (i/2) else inputo ((i-1)/2) in let output = transform sign n (load_array_r n input) in let oloc = locative_array_c n (C.array_subscript acr vcsr) (C.array_subscript aci vcsi) locations sovs in let odag = store_array_hc n oloc output in let annot = standard_optimizer odag in let body = Block ( [Decl ("INT", i)], [For (Expr_assign (CVar i, CVar v), Binop (" > ", CVar i, Integer 0), list_to_comma [Expr_assign (CVar i, CPlus [CVar i; CUminus (Integer 1)]); Expr_assign (CVar ar0, CPlus [CVar ar0; CVar sivs]); Expr_assign (CVar ar1, CPlus [CVar ar1; CVar sivs]); Expr_assign (CVar acr, CPlus [CVar acr; CVar sovs]); Expr_assign (CVar aci, CPlus [CVar aci; CVar sovs]); make_volatile_stride (4*n) (CVar rs); make_volatile_stride (4*n) (CVar csr); make_volatile_stride (4*n) (CVar csi) ], Asch annot) ]) in let tree = Fcn ((if !Magic.standalone then "void" else "static void"), name, ([Decl (C.realtypep, ar0); Decl (C.realtypep, ar1); Decl (C.realtypep, acr); Decl (C.realtypep, aci); Decl (C.stridetype, rs); Decl (C.stridetype, csr); Decl (C.stridetype, csi); Decl ("INT", v); Decl ("INT", "ivs"); Decl ("INT", "ovs")]), finalize_fcn body) in let desc = Printf.sprintf "static const kr2c_desc desc = { %d, \"%s\", %s, &GENUS };\n\n" n name (flops_of tree) and init = (declare_register_fcn name) ^ "{" ^ " X(kr2c_register)(p, " ^ name ^ ", &desc);\n" ^ "}\n" in (unparse tree) ^ "\n" ^ (if !Magic.standalone then "" else desc ^ init) let main () = begin parse speclist usage; print_string (generate (check_size ())); end let _ = main() fftw-3.3.3/genfft/gen_r2r.ml0000644000175000017500000001627212054410050012547 00000000000000(* * Copyright (c) 1997-1999 Massachusetts Institute of Technology * Copyright (c) 2003, 2007-11 Matteo Frigo * Copyright (c) 2003, 2007-11 Massachusetts Institute of Technology * * This program is free software; you can redistribute it and/or modify * it under the terms of the GNU General Public License as published by * the Free Software Foundation; either version 2 of the License, or * (at your option) any later version. * * This program is distributed in the hope that it will be useful, * but WITHOUT ANY WARRANTY; without even the implied warranty of * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the * GNU General Public License for more details. * * You should have received a copy of the GNU General Public License * along with this program; if not, write to the Free Software * Foundation, Inc., 51 Franklin Street, Fifth Floor, Boston, MA 02110-1301 USA * *) (* generation of trigonometric transforms *) open Util open Genutil open C let usage = "Usage: " ^ Sys.argv.(0) ^ " -n " let uistride = ref Stride_variable let uostride = ref Stride_variable let uivstride = ref Stride_variable let uovstride = ref Stride_variable type mode = | RDFT | HDFT | DHT | REDFT00 | REDFT10 | REDFT01 | REDFT11 | RODFT00 | RODFT10 | RODFT01 | RODFT11 | NONE let mode = ref NONE let normsqr = ref 1 let unitary = ref false let noloop = ref false let speclist = [ "-with-istride", Arg.String(fun x -> uistride := arg_to_stride x), " specialize for given input stride"; "-with-ostride", Arg.String(fun x -> uostride := arg_to_stride x), " specialize for given output stride"; "-with-ivstride", Arg.String(fun x -> uivstride := arg_to_stride x), " specialize for given input vector stride"; "-with-ovstride", Arg.String(fun x -> uovstride := arg_to_stride x), " specialize for given output vector stride"; "-rdft", Arg.Unit(fun () -> mode := RDFT), " generate a real DFT codelet"; "-hdft", Arg.Unit(fun () -> mode := HDFT), " generate a Hermitian DFT codelet"; "-dht", Arg.Unit(fun () -> mode := DHT), " generate a DHT codelet"; "-redft00", Arg.Unit(fun () -> mode := REDFT00), " generate a DCT-I codelet"; "-redft10", Arg.Unit(fun () -> mode := REDFT10), " generate a DCT-II codelet"; "-redft01", Arg.Unit(fun () -> mode := REDFT01), " generate a DCT-III codelet"; "-redft11", Arg.Unit(fun () -> mode := REDFT11), " generate a DCT-IV codelet"; "-rodft00", Arg.Unit(fun () -> mode := RODFT00), " generate a DST-I codelet"; "-rodft10", Arg.Unit(fun () -> mode := RODFT10), " generate a DST-II codelet"; "-rodft01", Arg.Unit(fun () -> mode := RODFT01), " generate a DST-III codelet"; "-rodft11", Arg.Unit(fun () -> mode := RODFT11), " generate a DST-IV codelet"; "-normalization", Arg.String(fun x -> let ix = int_of_string x in normsqr := ix * ix), " normalization integer to divide by"; "-normsqr", Arg.String(fun x -> normsqr := int_of_string x), " integer square of normalization to divide by"; "-unitary", Arg.Unit(fun () -> unitary := true), " unitary normalization (up overall scale factor)"; "-noloop", Arg.Unit(fun () -> noloop := true), " no vector loop"; ] let sqrt_half = Complex.inverse_int_sqrt 2 let sqrt_two = Complex.int_sqrt 2 let rescale sc s1 s2 input i = if ((i == s1 || i == s2) && !unitary) then Complex.times (input i) sc else input i let generate n mode = let iarray = "I" and oarray = "O" and istride = "is" and ostride = "os" and i = "i" and v = "v" in let sign = !Genutil.sign and name = !Magic.codelet_name in let vistride = either_stride (!uistride) (C.SVar istride) and vostride = either_stride (!uostride) (C.SVar ostride) in let sovs = stride_to_string "ovs" !uovstride in let sivs = stride_to_string "ivs" !uivstride in let (transform, load_input, store_output, si1,si2,so1,so2) = match mode with | RDFT -> Trig.rdft sign, load_array_r, store_array_hc, -1,-1,-1,-1 | HDFT -> Trig.hdft sign, load_array_c, store_array_r, -1,-1,-1,-1 (* TODO *) | DHT -> Trig.dht 1, load_array_r, store_array_r, -1,-1,-1,-1 | REDFT00 -> Trig.dctI, load_array_r, store_array_r, 0,n-1,0,n-1 | REDFT10 -> Trig.dctII, load_array_r, store_array_r, -1,-1,0,-1 | REDFT01 -> Trig.dctIII, load_array_r, store_array_r, 0,-1,-1,-1 | REDFT11 -> Trig.dctIV, load_array_r, store_array_r, -1,-1,-1,-1 | RODFT00 -> Trig.dstI, load_array_r, store_array_r, -1,-1,-1,-1 | RODFT10 -> Trig.dstII, load_array_r, store_array_r, -1,-1,n-1,-1 | RODFT01 -> Trig.dstIII, load_array_r, store_array_r, n-1,-1,-1,-1 | RODFT11 -> Trig.dstIV, load_array_r, store_array_r, -1,-1,-1,-1 | _ -> failwith "must specify transform kind" in let locations = unique_array_c n in let input = locative_array_c n (C.array_subscript iarray vistride) (C.array_subscript "BUG" vistride) locations sivs in let output = rescale sqrt_half so1 so2 ((Complex.times (Complex.inverse_int_sqrt !normsqr)) @@ (transform n (rescale sqrt_two si1 si2 (load_array_c n input)))) in let oloc = locative_array_c n (C.array_subscript oarray vostride) (C.array_subscript "BUG" vostride) locations sovs in let odag = store_output n oloc output in let annot = standard_optimizer odag in let body = if !noloop then Block([], [Asch annot]) else Block ( [Decl ("INT", i)], [For (Expr_assign (CVar i, CVar v), Binop (" > ", CVar i, Integer 0), list_to_comma [Expr_assign (CVar i, CPlus [CVar i; CUminus (Integer 1)]); Expr_assign (CVar iarray, CPlus [CVar iarray; CVar sivs]); Expr_assign (CVar oarray, CPlus [CVar oarray; CVar sovs]); make_volatile_stride (2*n) (CVar istride); make_volatile_stride (2*n) (CVar ostride) ], Asch annot) ]) in let tree = Fcn ((if !Magic.standalone then "void" else "static void"), name, ([Decl (C.constrealtypep, iarray); Decl (C.realtypep, oarray)] @ (if stride_fixed !uistride then [] else [Decl (C.stridetype, istride)]) @ (if stride_fixed !uostride then [] else [Decl (C.stridetype, ostride)]) @ (if !noloop then [] else [Decl ("INT", v)] @ (if stride_fixed !uivstride then [] else [Decl ("INT", "ivs")]) @ (if stride_fixed !uovstride then [] else [Decl ("INT", "ovs")]))), finalize_fcn body) in let desc = Printf.sprintf "static const kr2r_desc desc = { %d, \"%s\", %s, &GENUS, %s };\n\n" n name (flops_of tree) (match mode with | RDFT -> "RDFT00" | HDFT -> "HDFT00" | DHT -> "DHT" | REDFT00 -> "REDFT00" | REDFT10 -> "REDFT10" | REDFT01 -> "REDFT01" | REDFT11 -> "REDFT11" | RODFT00 -> "RODFT00" | RODFT10 -> "RODFT10" | RODFT01 -> "RODFT01" | RODFT11 -> "RODFT11" | _ -> failwith "must specify a transform kind") and init = (declare_register_fcn name) ^ "{" ^ " X(kr2r_register)(p, " ^ name ^ ", &desc);\n" ^ "}\n" in (unparse tree) ^ "\n" ^ (if !Magic.standalone then "" else desc ^ init) let main () = begin parse speclist usage; print_string (generate (check_size ()) !mode); end let _ = main() fftw-3.3.3/genfft/gen_twiddle.ml0000644000175000017500000001056712054410050013477 00000000000000(* * Copyright (c) 1997-1999 Massachusetts Institute of Technology * Copyright (c) 2003, 2007-11 Matteo Frigo * Copyright (c) 2003, 2007-11 Massachusetts Institute of Technology * * This program is free software; you can redistribute it and/or modify * it under the terms of the GNU General Public License as published by * the Free Software Foundation; either version 2 of the License, or * (at your option) any later version. * * This program is distributed in the hope that it will be useful, * but WITHOUT ANY WARRANTY; without even the implied warranty of * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the * GNU General Public License for more details. * * You should have received a copy of the GNU General Public License * along with this program; if not, write to the Free Software * Foundation, Inc., 51 Franklin Street, Fifth Floor, Boston, MA 02110-1301 USA * *) open Util open Genutil open C type ditdif = DIT | DIF let ditdif = ref DIT let usage = "Usage: " ^ Sys.argv.(0) ^ " -n [ -dit | -dif ]" let urs = ref Stride_variable let ums = ref Stride_variable let speclist = [ "-dit", Arg.Unit(fun () -> ditdif := DIT), " generate a DIT codelet"; "-dif", Arg.Unit(fun () -> ditdif := DIF), " generate a DIF codelet"; "-with-rs", Arg.String(fun x -> urs := arg_to_stride x), " specialize for given i/o stride"; "-with-ms", Arg.String(fun x -> ums := arg_to_stride x), " specialize for given ms" ] let generate n = let rioarray = "ri" and iioarray = "ii" and rs = "rs" and twarray = "W" and m = "m" and mb = "mb" and me = "me" and ms = "ms" in let sign = !Genutil.sign and name = !Magic.codelet_name and byvl x = choose_simd x (ctimes (CVar "(2 * VL)", x)) in let ename = expand_name name in let (bytwiddle, num_twiddles, twdesc) = Twiddle.twiddle_policy 0 false in let nt = num_twiddles n in let byw = bytwiddle n sign (twiddle_array nt twarray) in let vrs = either_stride (!urs) (C.SVar rs) in let sms = stride_to_string "ms" !ums in let locations = unique_array_c n in let iloc = locative_array_c n (C.array_subscript rioarray vrs) (C.array_subscript iioarray vrs) locations sms and oloc = locative_array_c n (C.array_subscript rioarray vrs) (C.array_subscript iioarray vrs) locations sms in let liloc = load_array_c n iloc in let output = match !ditdif with | DIT -> array n (Fft.dft sign n (byw liloc)) | DIF -> array n (byw (Fft.dft sign n liloc)) in let odag = store_array_c n oloc output in let annot = standard_optimizer odag in let vm = CVar m and vmb = CVar mb and vme = CVar me in let body = Block ( [Decl ("INT", m)], [For (list_to_comma [Expr_assign (vm, vmb); Expr_assign (CVar twarray, CPlus [CVar twarray; ctimes (vmb, Integer nt)])], Binop (" < ", vm, vme), list_to_comma [Expr_assign (vm, CPlus [vm; byvl (Integer 1)]); Expr_assign (CVar rioarray, CPlus [CVar rioarray; byvl (CVar sms)]); Expr_assign (CVar iioarray, CPlus [CVar iioarray; byvl (CVar sms)]); Expr_assign (CVar twarray, CPlus [CVar twarray; byvl (Integer nt)]); make_volatile_stride (2*n) (CVar rs) ], Asch annot)]) in let tree = Fcn (((if !Magic.standalone then "" else "static ") ^ "void"), ename, [Decl (C.realtypep, rioarray); Decl (C.realtypep, iioarray); Decl (C.constrealtypep, twarray); Decl (C.stridetype, rs); Decl ("INT", mb); Decl ("INT", me); Decl ("INT", ms)], finalize_fcn body) in let twinstr = Printf.sprintf "static const tw_instr twinstr[] = %s;\n\n" (twinstr_to_string "(2 * VL)" (twdesc n)) and desc = Printf.sprintf "static const ct_desc desc = {%d, %s, twinstr, &GENUS, %s, %s, %s, %s};\n\n" n (stringify name) (flops_of tree) (stride_to_solverparm !urs) "0" (stride_to_solverparm !ums) and register = match !ditdif with | DIT -> "X(kdft_dit_register)" | DIF -> "X(kdft_dif_register)" in let init = "\n" ^ twinstr ^ desc ^ (declare_register_fcn name) ^ (Printf.sprintf "{\n%s(p, %s, &desc);\n}" register ename) in (unparse tree) ^ "\n" ^ (if !Magic.standalone then "" else init) let main () = begin parse (speclist @ Twiddle.speclist) usage; print_string (generate (check_size ())); end let _ = main() fftw-3.3.3/genfft/gen_twiddle_c.ml0000644000175000017500000001105112054410050013766 00000000000000(* * Copyright (c) 1997-1999 Massachusetts Institute of Technology * Copyright (c) 2003, 2007-11 Matteo Frigo * Copyright (c) 2003, 2007-11 Massachusetts Institute of Technology * * This program is free software; you can redistribute it and/or modify * it under the terms of the GNU General Public License as published by * the Free Software Foundation; either version 2 of the License, or * (at your option) any later version. * * This program is distributed in the hope that it will be useful, * but WITHOUT ANY WARRANTY; without even the implied warranty of * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the * GNU General Public License for more details. * * You should have received a copy of the GNU General Public License * along with this program; if not, write to the Free Software * Foundation, Inc., 51 Franklin Street, Fifth Floor, Boston, MA 02110-1301 USA * *) open Util open Genutil open C type ditdif = DIT | DIF let ditdif = ref DIT let usage = "Usage: " ^ Sys.argv.(0) ^ " -n [ -dit | -dif ]" let urs = ref Stride_variable let ums = ref Stride_variable let speclist = [ "-dit", Arg.Unit(fun () -> ditdif := DIT), " generate a DIT codelet"; "-dif", Arg.Unit(fun () -> ditdif := DIF), " generate a DIF codelet"; "-with-rs", Arg.String(fun x -> urs := arg_to_stride x), " specialize for given i/o stride"; "-with-ms", Arg.String(fun x -> ums := arg_to_stride x), " specialize for given ms" ] let generate n = let rioarray = "x" and rs = "rs" and twarray = "W" and m = "m" and mb = "mb" and me = "me" and ms = "ms" in let sign = !Genutil.sign and name = !Magic.codelet_name and byvl x = choose_simd x (ctimes (CVar "VL", x)) and bytwvl x = choose_simd x (ctimes (CVar "TWVL", x)) and bytwvl_vl x = choose_simd x (ctimes (CVar "(TWVL/VL)", x)) in let ename = expand_name name in let (bytwiddle, num_twiddles, twdesc) = Twiddle.twiddle_policy 0 true in let nt = num_twiddles n in let byw = bytwiddle n sign (twiddle_array nt twarray) in let vrs = either_stride (!urs) (C.SVar rs) in let sms = stride_to_string "ms" !ums in let locations = unique_array_c n in let iloc = locative_array_c n (C.array_subscript rioarray vrs) (C.array_subscript "BUG" vrs) locations sms and oloc = locative_array_c n (C.array_subscript rioarray vrs) (C.array_subscript "BUG" vrs) locations sms in let liloc = load_array_r n iloc in let fft = Trig.dft_via_rdft in let output = match !ditdif with | DIT -> array n (fft sign n (byw liloc)) | DIF -> array n (byw (fft sign n liloc)) in let odag = store_array_r n oloc output in let annot = standard_optimizer odag in let vm = CVar m and vmb = CVar mb and vme = CVar me in let body = Block ( [Decl ("INT", m); Decl (C.realtypep, rioarray)], [Stmt_assign (CVar rioarray, CVar (if (sign < 0) then "ri" else "ii")); For (list_to_comma [Expr_assign (vm, vmb); Expr_assign (CVar twarray, CPlus [CVar twarray; ctimes (vmb, bytwvl_vl (Integer nt))])], Binop (" < ", vm, vme), list_to_comma [Expr_assign (vm, CPlus [vm; byvl (Integer 1)]); Expr_assign (CVar rioarray, CPlus [CVar rioarray; byvl (CVar sms)]); Expr_assign (CVar twarray, CPlus [CVar twarray; bytwvl (Integer nt)]); make_volatile_stride n (CVar rs) ], Asch annot)]) in let tree = Fcn (((if !Magic.standalone then "" else "static ") ^ "void"), ename, [Decl (C.realtypep, "ri"); Decl (C.realtypep, "ii"); Decl (C.constrealtypep, twarray); Decl (C.stridetype, rs); Decl ("INT", mb); Decl ("INT", me); Decl ("INT", ms)], finalize_fcn body) in let twinstr = Printf.sprintf "static const tw_instr twinstr[] = %s;\n\n" (twinstr_to_string "VL" (twdesc n)) and desc = Printf.sprintf "static const ct_desc desc = {%d, %s, twinstr, &GENUS, %s, %s, %s, %s};\n\n" n (stringify name) (flops_of tree) (stride_to_solverparm !urs) "0" (stride_to_solverparm !ums) and register = match !ditdif with | DIT -> "X(kdft_dit_register)" | DIF -> "X(kdft_dif_register)" in let init = "\n" ^ twinstr ^ desc ^ (declare_register_fcn name) ^ (Printf.sprintf "{\n%s(p, %s, &desc);\n}" register ename) in (unparse tree) ^ "\n" ^ (if !Magic.standalone then "" else init) let main () = begin Simdmagic.simd_mode := true; parse (speclist @ Twiddle.speclist) usage; print_string (generate (check_size ())); end let _ = main() fftw-3.3.3/genfft/gen_twidsq.ml0000644000175000017500000001136212054410050013350 00000000000000(* * Copyright (c) 1997-1999 Massachusetts Institute of Technology * Copyright (c) 2003, 2007-11 Matteo Frigo * Copyright (c) 2003, 2007-11 Massachusetts Institute of Technology * * This program is free software; you can redistribute it and/or modify * it under the terms of the GNU General Public License as published by * the Free Software Foundation; either version 2 of the License, or * (at your option) any later version. * * This program is distributed in the hope that it will be useful, * but WITHOUT ANY WARRANTY; without even the implied warranty of * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the * GNU General Public License for more details. * * You should have received a copy of the GNU General Public License * along with this program; if not, write to the Free Software * Foundation, Inc., 51 Franklin Street, Fifth Floor, Boston, MA 02110-1301 USA * *) open Util open Genutil open C type ditdif = DIT | DIF let ditdif = ref DIT let usage = "Usage: " ^ Sys.argv.(0) ^ " -n [ -dit | -dif ]" let reload_twiddle = ref false let urs = ref Stride_variable let uvs = ref Stride_variable let ums = ref Stride_variable let speclist = [ "-dit", Arg.Unit(fun () -> ditdif := DIT), " generate a DIT codelet"; "-dif", Arg.Unit(fun () -> ditdif := DIF), " generate a DIF codelet"; "-reload-twiddle", Arg.Unit(fun () -> reload_twiddle := true), " do not collect common twiddle factors"; "-with-rs", Arg.String(fun x -> urs := arg_to_stride x), " specialize for given input stride"; "-with-vs", Arg.String(fun x -> uvs := arg_to_stride x), " specialize for given vector stride"; "-with-ms", Arg.String(fun x -> ums := arg_to_stride x), " specialize for given ms" ] let generate n = let rioarray = "rio" and iioarray = "iio" and rs = "rs" and vs = "vs" and twarray = "W" and m = "m" and mb = "mb" and me = "me" and ms = "ms" in let sign = !Genutil.sign and name = !Magic.codelet_name in let (bytwiddle, num_twiddles, twdesc) = Twiddle.twiddle_policy 0 false in let nt = num_twiddles n in let svs = either_stride (!uvs) (C.SVar vs) and srs = either_stride (!urs) (C.SVar rs) in let byw = if !reload_twiddle then array n (fun v -> bytwiddle n sign (twiddle_array nt twarray)) else let a = bytwiddle n sign (twiddle_array nt twarray) in fun v -> a in let locations = unique_v_array_c n n in let ioi = locative_v_array_c n n (C.varray_subscript rioarray svs srs) (C.varray_subscript iioarray svs srs) locations "BUG" and ioo = locative_v_array_c n n (C.varray_subscript rioarray svs srs) (C.varray_subscript iioarray svs srs) locations "BUG" in let lioi = load_v_array_c n n ioi in let output = match !ditdif with | DIT -> array n (fun v -> Fft.dft sign n (byw v (lioi v))) | DIF -> array n (fun v -> byw v (Fft.dft sign n (lioi v))) in let odag = store_v_array_c n n ioo (transpose output) in let annot = standard_optimizer odag in let vm = CVar m and vmb = CVar mb and vme = CVar me in let body = Block ( [Decl ("INT", m)], [For (list_to_comma [Expr_assign (vm, vmb); Expr_assign (CVar twarray, CPlus [CVar twarray; ctimes (vmb, Integer nt)])], Binop (" < ", vm, vme), list_to_comma [Expr_assign (vm, CPlus [vm; Integer 1]); Expr_assign (CVar rioarray, CPlus [CVar rioarray; CVar ms]); Expr_assign (CVar iioarray, CPlus [CVar iioarray; CVar ms]); Expr_assign (CVar twarray, CPlus [CVar twarray; Integer nt]); make_volatile_stride (2*n) (CVar rs); make_volatile_stride (2*0) (CVar vs) ], Asch annot)]) in let tree = Fcn (("static void"), name, [Decl (C.realtypep, rioarray); Decl (C.realtypep, iioarray); Decl (C.constrealtypep, twarray); Decl (C.stridetype, rs); Decl (C.stridetype, vs); Decl ("INT", mb); Decl ("INT", me); Decl ("INT", ms)], finalize_fcn body) in let twinstr = Printf.sprintf "static const tw_instr twinstr[] = %s;\n\n" (Twiddle.twinstr_to_c_string (twdesc n)) and desc = Printf.sprintf "static const ct_desc desc = {%d, \"%s\", twinstr, &GENUS, %s, %s, %s, %s};\n\n" n name (flops_of tree) (stride_to_solverparm !urs) (stride_to_solverparm !uvs) (stride_to_solverparm !ums) and register = match !ditdif with | DIT -> "X(kdft_ditsq_register)" | DIF -> "X(kdft_difsq_register)" in let init = "\n" ^ twinstr ^ desc ^ (declare_register_fcn name) ^ (Printf.sprintf "{\n%s(p, %s, &desc);\n}" register name) in (unparse tree) ^ "\n" ^ init let main () = begin parse (speclist @ Twiddle.speclist) usage; print_string (generate (check_size ())); end let _ = main() fftw-3.3.3/genfft/gen_twidsq_c.ml0000644000175000017500000001213212054410050013646 00000000000000(* * Copyright (c) 1997-1999 Massachusetts Institute of Technology * Copyright (c) 2003, 2007-11 Matteo Frigo * Copyright (c) 2003, 2007-11 Massachusetts Institute of Technology * * This program is free software; you can redistribute it and/or modify * it under the terms of the GNU General Public License as published by * the Free Software Foundation; either version 2 of the License, or * (at your option) any later version. * * This program is distributed in the hope that it will be useful, * but WITHOUT ANY WARRANTY; without even the implied warranty of * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the * GNU General Public License for more details. * * You should have received a copy of the GNU General Public License * along with this program; if not, write to the Free Software * Foundation, Inc., 51 Franklin Street, Fifth Floor, Boston, MA 02110-1301 USA * *) open Util open Genutil open C type ditdif = DIT | DIF let ditdif = ref DIT let usage = "Usage: " ^ Sys.argv.(0) ^ " -n [ -dit | -dif ]" let reload_twiddle = ref false let urs = ref Stride_variable let uvs = ref Stride_variable let ums = ref Stride_variable let speclist = [ "-dit", Arg.Unit(fun () -> ditdif := DIT), " generate a DIT codelet"; "-dif", Arg.Unit(fun () -> ditdif := DIF), " generate a DIF codelet"; "-reload-twiddle", Arg.Unit(fun () -> reload_twiddle := true), " do not collect common twiddle factors"; "-with-rs", Arg.String(fun x -> urs := arg_to_stride x), " specialize for given input stride"; "-with-vs", Arg.String(fun x -> uvs := arg_to_stride x), " specialize for given vector stride"; "-with-ms", Arg.String(fun x -> ums := arg_to_stride x), " specialize for given ms" ] let generate n = let rioarray = "x" and rs = "rs" and vs = "vs" and twarray = "W" and m = "m" and mb = "mb" and me = "me" and ms = "ms" in let sign = !Genutil.sign and name = !Magic.codelet_name and byvl x = choose_simd x (ctimes (CVar "VL", x)) and bytwvl x = choose_simd x (ctimes (CVar "TWVL", x)) and bytwvl_vl x = choose_simd x (ctimes (CVar "(TWVL/VL)", x)) in let ename = expand_name name in let (bytwiddle, num_twiddles, twdesc) = Twiddle.twiddle_policy 0 true in let nt = num_twiddles n in let svs = either_stride (!uvs) (C.SVar vs) and srs = either_stride (!urs) (C.SVar rs) in let sms = stride_to_string "ms" !ums in let byw = if !reload_twiddle then array n (fun v -> bytwiddle n sign (twiddle_array nt twarray)) else let a = bytwiddle n sign (twiddle_array nt twarray) in fun v -> a in let locations = unique_v_array_c n n in let ioi = locative_v_array_c n n (C.varray_subscript rioarray svs srs) (C.varray_subscript "BUG" svs srs) locations sms and ioo = locative_v_array_c n n (C.varray_subscript rioarray svs srs) (C.varray_subscript "BUG" svs srs) locations sms in let lioi = load_v_array_c n n ioi in let fft = Trig.dft_via_rdft in let output = match !ditdif with | DIT -> array n (fun v -> fft sign n (byw v (lioi v))) | DIF -> array n (fun v -> byw v (fft sign n (lioi v))) in let odag = store_v_array_c n n ioo (transpose output) in let annot = standard_optimizer odag in let vm = CVar m and vmb = CVar mb and vme = CVar me in let body = Block ( [Decl ("INT", m); Decl (C.realtypep, rioarray)], [Stmt_assign (CVar rioarray, CVar (if (sign < 0) then "ri" else "ii")); For (list_to_comma [Expr_assign (vm, vmb); Expr_assign (CVar twarray, CPlus [CVar twarray; ctimes (vmb, bytwvl_vl (Integer nt))])], Binop (" < ", vm, vme), list_to_comma [Expr_assign (vm, CPlus [vm; byvl (Integer 1)]); Expr_assign (CVar rioarray, CPlus [CVar rioarray; byvl (CVar sms)]); Expr_assign (CVar twarray, CPlus [CVar twarray; bytwvl (Integer nt)]); make_volatile_stride (2*n) (CVar rs); make_volatile_stride (2*n) (CVar vs) ], Asch annot)]) in let tree = Fcn (("static void"), ename, [Decl (C.realtypep, "ri"); Decl (C.realtypep, "ii"); Decl (C.constrealtypep, twarray); Decl (C.stridetype, rs); Decl (C.stridetype, vs); Decl ("INT", mb); Decl ("INT", me); Decl ("INT", ms)], finalize_fcn body) in let twinstr = Printf.sprintf "static const tw_instr twinstr[] = %s;\n\n" (twinstr_to_string "VL" (twdesc n)) and desc = Printf.sprintf "static const ct_desc desc = {%d, %s, twinstr, &GENUS, %s, %s, %s, %s};\n\n" n (stringify name) (flops_of tree) (stride_to_solverparm !urs) (stride_to_solverparm !uvs) (stride_to_solverparm !ums) and register = match !ditdif with | DIT -> "X(kdft_ditsq_register)" | DIF -> "X(kdft_difsq_register)" in let init = "\n" ^ twinstr ^ desc ^ (declare_register_fcn name) ^ (Printf.sprintf "{\n%s(p, %s, &desc);\n}" register ename) in (unparse tree) ^ "\n" ^ init let main () = begin parse (speclist @ Twiddle.speclist) usage; print_string (generate (check_size ())); end let _ = main() fftw-3.3.3/genfft/genutil.ml0000644000175000017500000002126412054410050012655 00000000000000(* * Copyright (c) 1997-1999 Massachusetts Institute of Technology * Copyright (c) 2003, 2007-11 Matteo Frigo * Copyright (c) 2003, 2007-11 Massachusetts Institute of Technology * * This program is free software; you can redistribute it and/or modify * it under the terms of the GNU General Public License as published by * the Free Software Foundation; either version 2 of the License, or * (at your option) any later version. * * This program is distributed in the hope that it will be useful, * but WITHOUT ANY WARRANTY; without even the implied warranty of * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the * GNU General Public License for more details. * * You should have received a copy of the GNU General Public License * along with this program; if not, write to the Free Software * Foundation, Inc., 51 Franklin Street, Fifth Floor, Boston, MA 02110-1301 USA * *) (* utilities common to all generators *) open Util let choose_simd a b = if !Simdmagic.simd_mode then b else a let unique_array n = array n (fun _ -> Unique.make ()) let unique_array_c n = array n (fun _ -> (Unique.make (), Unique.make ())) let unique_v_array_c veclen n = array veclen (fun _ -> unique_array_c n) let locative_array_c n rarr iarr loc vs = array n (fun i -> let klass = Unique.make () in let (rloc, iloc) = loc i in (Variable.make_locative rloc klass rarr i vs, Variable.make_locative iloc klass iarr i vs)) let locative_v_array_c veclen n rarr iarr loc vs = array veclen (fun v -> array n (fun i -> let klass = Unique.make () in let (rloc, iloc) = loc v i in (Variable.make_locative rloc klass (rarr v) i vs, Variable.make_locative iloc klass (iarr v) i vs))) let temporary_array n = array n (fun i -> Variable.make_temporary ()) let temporary_array_c n = let tmpr = temporary_array n and tmpi = temporary_array n in array n (fun i -> (tmpr i, tmpi i)) let temporary_v_array_c veclen n = array veclen (fun v -> temporary_array_c n) let temporary_array_c n = let tmpr = temporary_array n and tmpi = temporary_array n in array n (fun i -> (tmpr i, tmpi i)) let load_c (vr, vi) = Complex.make (Expr.Load vr, Expr.Load vi) let load_r (vr, vi) = Complex.make (Expr.Load vr, Expr.Num (Number.zero)) let twiddle_array nt w = array (nt/2) (fun i -> let stride = choose_simd (C.SInteger 1) (C.SConst "TWVL") and klass = Unique.make () in let (refr, refi) = (C.array_subscript w stride (2 * i), C.array_subscript w stride (2 * i + 1)) in let (kr, ki) = (Variable.make_constant klass refr, Variable.make_constant klass refi) in load_c (kr, ki)) let load_array_c n var = array n (fun i -> load_c (var i)) let load_array_r n var = array n (fun i -> load_r (var i)) let load_array_hc n var = array n (fun i -> if (i < n - i) then load_c (var i) else if (i > n - i) then Complex.times Complex.i (load_c (var (n - i))) else load_r (var i)) let load_v_array_c veclen n var = array veclen (fun v -> load_array_c n (var v)) let store_c (vr, vi) x = [Complex.store_real vr x; Complex.store_imag vi x] let store_r (vr, vi) x = Complex.store_real vr x let store_i (vr, vi) x = Complex.store_imag vi x let assign_array_c n dst src = List.flatten (rmap (iota n) (fun i -> let (ar, ai) = Complex.assign (dst i) (src i) in [ar; ai])) let assign_v_array_c veclen n dst src = List.flatten (rmap (iota veclen) (fun v -> assign_array_c n (dst v) (src v))) let vassign_v_array_c veclen n dst src = List.flatten (rmap (iota n) (fun i -> List.flatten (rmap (iota veclen) (fun v -> let (ar, ai) = Complex.assign (dst v i) (src v i) in [ar; ai])))) let store_array_r n dst src = rmap (iota n) (fun i -> store_r (dst i) (src i)) let store_array_c n dst src = List.flatten (rmap (iota n) (fun i -> store_c (dst i) (src i))) let store_array_hc n dst src = List.flatten (rmap (iota n) (fun i -> if (i < n - i) then store_c (dst i) (src i) else if (i > n - i) then [] else [store_r (dst i) (Complex.real (src i))])) let store_v_array_c veclen n dst src = List.flatten (rmap (iota veclen) (fun v -> store_array_c n (dst v) (src v))) let elementwise f n a = array n (fun i -> f (a i)) let conj_array_c = elementwise Complex.conj let real_array_c = elementwise Complex.real let imag_array_c = elementwise Complex.imag let elementwise_v f veclen n a = array veclen (fun v -> array n (fun i -> f (a v i))) let conj_v_array_c = elementwise_v Complex.conj let real_v_array_c = elementwise_v Complex.real let imag_v_array_c = elementwise_v Complex.imag let transpose f i j = f j i let symmetrize f i j = if i <= j then f i j else f j i (* utilities for command-line parsing *) let standard_arg_parse_fail _ = failwith "too many arguments" let dump_dag alist = let fnam = !Magic.dag_dump_file in if (String.length fnam > 0) then let ochan = open_out fnam in begin To_alist.dump (output_string ochan) alist; close_out ochan; end let dump_alist alist = let fnam = !Magic.alist_dump_file in if (String.length fnam > 0) then let ochan = open_out fnam in begin Expr.dump (output_string ochan) alist; close_out ochan; end let dump_asched asched = let fnam = !Magic.asched_dump_file in if (String.length fnam > 0) then let ochan = open_out fnam in begin Annotate.dump (output_string ochan) asched; close_out ochan; end (* utilities for optimization *) let standard_scheduler dag = let optim = Algsimp.algsimp dag in let alist = To_alist.to_assignments optim in let _ = dump_alist alist in let _ = dump_dag alist in if !Magic.precompute_twiddles then Schedule.isolate_precomputations_and_schedule alist else Schedule.schedule alist let standard_optimizer dag = let sched = standard_scheduler dag in let annot = Annotate.annotate [] sched in let _ = dump_asched annot in annot let size = ref None let sign = ref (-1) let speclist = [ "-n", Arg.Int(fun i -> size := Some i), " generate a codelet of size "; "-sign", Arg.Int(fun i -> if (i > 0) then sign := 1 else sign := (-1)), " sign of transform"; ] let check_size () = match !size with | Some i -> i | None -> failwith "must specify -n" let expand_name name = if name = "" then "noname" else name let declare_register_fcn name = if name = "" then "void NAME(planner *p)\n" else "void " ^ (choose_simd "X" "XSIMD") ^ "(codelet_" ^ name ^ ")(planner *p)\n" let stringify name = if name = "" then "STRINGIZE(NAME)" else choose_simd ("\"" ^ name ^ "\"") ("XSIMD_STRING(\"" ^ name ^ "\")") let parse user_speclist usage = Arg.parse (user_speclist @ speclist @ Magic.speclist @ Simdmagic.speclist) standard_arg_parse_fail usage let rec list_to_c = function [] -> "" | [a] -> (string_of_int a) | a :: b -> (string_of_int a) ^ ", " ^ (list_to_c b) let rec list_to_comma = function | [a; b] -> C.Comma (a, b) | a :: b -> C.Comma (a, list_to_comma b) | _ -> failwith "list_to_comma" type stride = Stride_variable | Fixed_int of int | Fixed_string of string let either_stride a b = match a with Fixed_int x -> C.SInteger x | Fixed_string x -> C.SConst x | _ -> b let stride_fixed = function Stride_variable -> false | _ -> true let arg_to_stride s = try Fixed_int (int_of_string s) with Failure "int_of_string" -> Fixed_string s let stride_to_solverparm = function Stride_variable -> "0" | Fixed_int x -> string_of_int x | Fixed_string x -> x let stride_to_string s = function Stride_variable -> s | Fixed_int x -> string_of_int x | Fixed_string x -> x (* output the command line *) let cmdline () = List.fold_right (fun a b -> a ^ " " ^ b) (Array.to_list Sys.argv) "" let unparse tree = "/* Generated by: " ^ (cmdline ()) ^ "*/\n\n" ^ (C.print_cost tree) ^ (if String.length !Magic.inklude > 0 then (Printf.sprintf "#include \"%s\"\n\n" !Magic.inklude) else "") ^ (if !Simdmagic.simd_mode then Simd.unparse_function tree else C.unparse_function tree) let finalize_fcn ast = let mergedecls = function C.Block (d1, [C.Block (d2, s)]) -> C.Block (d1 @ d2, s) | x -> x and extract_constants = if !Simdmagic.simd_mode then Simd.extract_constants else C.extract_constants in mergedecls (C.Block (extract_constants ast, [ast; C.Simd_leavefun])) let twinstr_to_string vl x = if !Simdmagic.simd_mode then Twiddle.twinstr_to_simd_string vl x else Twiddle.twinstr_to_c_string x let make_volatile_stride n x = C.CCall ("MAKE_VOLATILE_STRIDE", C.Comma((C.Integer n), x)) fftw-3.3.3/genfft/littlesimp.ml0000644000175000017500000000510512054410050013370 00000000000000(* * Copyright (c) 1997-1999 Massachusetts Institute of Technology * Copyright (c) 2003, 2007-11 Matteo Frigo * Copyright (c) 2003, 2007-11 Massachusetts Institute of Technology * * This program is free software; you can redistribute it and/or modify * it under the terms of the GNU General Public License as published by * the Free Software Foundation; either version 2 of the License, or * (at your option) any later version. * * This program is distributed in the hope that it will be useful, * but WITHOUT ANY WARRANTY; without even the implied warranty of * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the * GNU General Public License for more details. * * You should have received a copy of the GNU General Public License * along with this program; if not, write to the Free Software * Foundation, Inc., 51 Franklin Street, Fifth Floor, Boston, MA 02110-1301 USA * *) (* * The LittleSimplifier module implements a subset of the simplifications * of the AlgSimp module. These simplifications can be executed * quickly here, while they would take a long time using the heavy * machinery of AlgSimp. * * For example, 0 * x is simplified to 0 tout court by the LittleSimplifier. * On the other hand, AlgSimp would first simplify x, generating lots * of common subexpressions, storing them in a table etc, just to * discard all the work later. Similarly, the LittleSimplifier * reduces the constant FFT in Rader's algorithm to a constant sequence. *) open Expr let rec makeNum = function | n -> Num n and makeUminus = function | Uminus a -> a | Num a -> makeNum (Number.negate a) | a -> Uminus a and makeTimes = function | (Num a, Num b) -> makeNum (Number.mul a b) | (Num a, Times (Num b, c)) -> makeTimes (makeNum (Number.mul a b), c) | (Num a, b) when Number.is_zero a -> makeNum (Number.zero) | (Num a, b) when Number.is_one a -> b | (Num a, b) when Number.is_mone a -> makeUminus b | (Num a, Uminus b) -> Times (makeUminus (Num a), b) | (a, (Num b as b')) -> makeTimes (b', a) | (a, b) -> Times (a, b) and makePlus l = let rec reduceSum x = match x with [] -> [] | [Num a] -> if Number.is_zero a then [] else x | (Num a) :: (Num b) :: c -> reduceSum ((makeNum (Number.add a b)) :: c) | ((Num _) as a') :: b :: c -> b :: reduceSum (a' :: c) | a :: s -> a :: reduceSum s in match reduceSum l with [] -> makeNum (Number.zero) | [a] -> a | [a; b] when a == b -> makeTimes (Num Number.two, a) | [Times (Num a, b); Times (Num c, d)] when b == d -> makeTimes (makePlus [Num a; Num c], b) | a -> Plus a fftw-3.3.3/genfft/magic.ml0000644000175000017500000001400112054410050012255 00000000000000(* * Copyright (c) 1997-1999 Massachusetts Institute of Technology * Copyright (c) 2003, 2007-11 Matteo Frigo * Copyright (c) 2003, 2007-11 Massachusetts Institute of Technology * * This program is free software; you can redistribute it and/or modify * it under the terms of the GNU General Public License as published by * the Free Software Foundation; either version 2 of the License, or * (at your option) any later version. * * This program is distributed in the hope that it will be useful, * but WITHOUT ANY WARRANTY; without even the implied warranty of * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the * GNU General Public License for more details. * * You should have received a copy of the GNU General Public License * along with this program; if not, write to the Free Software * Foundation, Inc., 51 Franklin Street, Fifth Floor, Boston, MA 02110-1301 USA * *) (* magic parameters *) let verbose = ref false let vneg = ref false let karatsuba_min = ref 15 let karatsuba_variant = ref 2 let circular_min = ref 64 let rader_min = ref 13 let rader_list = ref [5] let alternate_convolution = ref 17 let threemult = ref false let inline_single = ref true let inline_loads = ref false let inline_loads_constants = ref false let inline_constants = ref true let trivial_stores = ref false let locations_are_special = ref false let strength_reduce_mul = ref false let number_of_variables = ref 4 let codelet_name = ref "unnamed" let randomized_cse = ref true let dif_split_radix = ref false let enable_fma = ref false let deep_collect_depth = ref 1 let schedule_type = ref 0 let compact = ref false let dag_dump_file = ref "" let alist_dump_file = ref "" let asched_dump_file = ref "" let lisp_syntax = ref false let network_transposition = ref true let inklude = ref "" let generic_arith = ref false let reorder_insns = ref false let reorder_loads = ref false let reorder_stores = ref false let precompute_twiddles = ref false let newsplit = ref false let standalone = ref false let pipeline_latency = ref 0 let schedule_for_pipeline = ref false let generate_bytw = ref true (* command-line parser for magic parameters *) let undocumented = " Undocumented voodoo parameter" let set_bool var = Arg.Unit (fun () -> var := true) let unset_bool var = Arg.Unit (fun () -> var := false) let set_int var = Arg.Int(fun i -> var := i) let set_string var = Arg.String(fun s -> var := s) let speclist = [ "-name", set_string codelet_name, " set codelet name"; "-standalone", set_bool standalone, " standalone codelet (no desc)"; "-include", set_string inklude, undocumented; "-verbose", set_bool verbose, " Enable verbose logging messages to stderr"; "-rader-min", set_int rader_min, " : Use Rader's algorithm for prime sizes >= "; "-threemult", set_bool threemult, " Use 3-multiply complex multiplications"; "-karatsuba-min", set_int karatsuba_min, undocumented; "-karatsuba-variant", set_int karatsuba_variant, undocumented; "-circular-min", set_int circular_min, undocumented; "-compact", set_bool compact, " Mangle variable names to reduce size of source code"; "-no-compact", unset_bool compact, " Disable -compact"; "-dump-dag", set_string dag_dump_file, undocumented; "-dump-alist", set_string alist_dump_file, undocumented; "-dump-asched", set_string asched_dump_file, undocumented; "-lisp-syntax", set_bool lisp_syntax, undocumented; "-alternate-convolution", set_int alternate_convolution, undocumented; "-deep-collect-depth", set_int deep_collect_depth, undocumented; "-schedule-type", set_int schedule_type, undocumented; "-pipeline-latency", set_int pipeline_latency, undocumented; "-schedule-for-pipeline", set_bool schedule_for_pipeline, undocumented; "-dif-split-radix", set_bool dif_split_radix, undocumented; "-dit-split-radix", unset_bool dif_split_radix, undocumented; "-generic-arith", set_bool generic_arith, undocumented; "-no-generic-arith", unset_bool generic_arith, undocumented; "-precompute-twiddles", set_bool precompute_twiddles, undocumented; "-no-precompute-twiddles", unset_bool precompute_twiddles, undocumented; "-inline-single", set_bool inline_single, undocumented; "-no-inline-single", unset_bool inline_single, undocumented; "-inline-loads", set_bool inline_loads, undocumented; "-no-inline-loads", unset_bool inline_loads, undocumented; "-inline-loads-constants", set_bool inline_loads_constants, undocumented; "-no-inline-loads-constants", unset_bool inline_loads_constants, undocumented; "-inline-constants", set_bool inline_constants, undocumented; "-no-inline-constants", unset_bool inline_constants, undocumented; "-trivial-stores", set_bool trivial_stores, undocumented; "-no-trivial-stores", unset_bool trivial_stores, undocumented; "-locations-are-special", set_bool locations_are_special, undocumented; "-no-locations-are-special", unset_bool locations_are_special, undocumented; "-randomized-cse", set_bool randomized_cse, undocumented; "-no-randomized-cse", unset_bool randomized_cse, undocumented; "-network-transposition", set_bool network_transposition, undocumented; "-no-network-transposition", unset_bool network_transposition, undocumented; "-reorder-insns", set_bool reorder_insns, undocumented; "-no-reorder-insns", unset_bool reorder_insns, undocumented; "-reorder-loads", set_bool reorder_loads, undocumented; "-no-reorder-loads", unset_bool reorder_loads, undocumented; "-reorder-stores", set_bool reorder_stores, undocumented; "-no-reorder-stores", unset_bool reorder_stores, undocumented; "-newsplit", set_bool newsplit, undocumented; "-vneg", set_bool vneg, undocumented; "-fma", set_bool enable_fma, undocumented; "-no-fma", unset_bool enable_fma, undocumented; "-variables", set_int number_of_variables, undocumented; "-strength-reduce-mul", set_bool strength_reduce_mul, undocumented; "-no-strength-reduce-mul", unset_bool strength_reduce_mul, undocumented; "-generate-bytw", set_bool generate_bytw, undocumented; "-no-generate-bytw", unset_bool generate_bytw, undocumented; ] fftw-3.3.3/genfft/monads.ml0000644000175000017500000000406512054410050012467 00000000000000(* * Copyright (c) 1997-1999 Massachusetts Institute of Technology * Copyright (c) 2003, 2007-11 Matteo Frigo * Copyright (c) 2003, 2007-11 Massachusetts Institute of Technology * * This program is free software; you can redistribute it and/or modify * it under the terms of the GNU General Public License as published by * the Free Software Foundation; either version 2 of the License, or * (at your option) any later version. * * This program is distributed in the hope that it will be useful, * but WITHOUT ANY WARRANTY; without even the implied warranty of * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the * GNU General Public License for more details. * * You should have received a copy of the GNU General Public License * along with this program; if not, write to the Free Software * Foundation, Inc., 51 Franklin Street, Fifth Floor, Boston, MA 02110-1301 USA * *) (************************************************************* * Monads *************************************************************) (* * Phil Wadler has many well written papers about monads. See * http://cm.bell-labs.com/cm/cs/who/wadler/ *) (* vanilla state monad *) module StateMonad = struct let returnM x = fun s -> (x, s) let (>>=) = fun m k -> fun s -> let (a', s') = m s in let (a'', s'') = k a' s' in (a'', s'') let (>>) = fun m k -> m >>= fun _ -> k let rec mapM f = function [] -> returnM [] | a :: b -> f a >>= fun a' -> mapM f b >>= fun b' -> returnM (a' :: b') let runM m x initial_state = let (a, _) = m x initial_state in a let fetchState = fun s -> s, s let storeState newState = fun _ -> (), newState end (* monad with built-in memoizing capabilities *) module MemoMonad = struct open StateMonad let memoizing lookupM insertM f k = lookupM k >>= fun vMaybe -> match vMaybe with Some value -> returnM value | None -> f k >>= fun value -> insertM k value >> returnM value let runM initial_state m x = StateMonad.runM m x initial_state end fftw-3.3.3/genfft/number.ml0000644000175000017500000001316212054410050012474 00000000000000(* * Copyright (c) 1997-1999 Massachusetts Institute of Technology * Copyright (c) 2003, 2007-11 Matteo Frigo * Copyright (c) 2003, 2007-11 Massachusetts Institute of Technology * * This program is free software; you can redistribute it and/or modify * it under the terms of the GNU General Public License as published by * the Free Software Foundation; either version 2 of the License, or * (at your option) any later version. * * This program is distributed in the hope that it will be useful, * but WITHOUT ANY WARRANTY; without even the implied warranty of * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the * GNU General Public License for more details. * * You should have received a copy of the GNU General Public License * along with this program; if not, write to the Free Software * Foundation, Inc., 51 Franklin Street, Fifth Floor, Boston, MA 02110-1301 USA * *) (* The generator keeps track of numeric constants in symbolic expressions using the abstract number type, defined in this file. Our implementation of the number type uses arbitrary-precision arithmetic from the built-in Num package in order to maintain an accurate representation of constants. This allows us to output constants with many decimal places in the generated C code, ensuring that we will take advantage of the full precision available on current and future machines. Note that we have to write our own routine to compute roots of unity, since the Num package only supplies simple arithmetic. The arbitrary-precision operations in Num look like the normal operations except that they have an appended slash (e.g. +/ -/ */ // etcetera). *) open Num type number = N of num let makeNum n = N n (* decimal digits of precision to maintain internally, and to print out: *) let precision = 50 let print_precision = 45 let inveps = (Int 10) **/ (Int precision) let epsilon = (Int 1) // inveps let pinveps = (Int 10) **/ (Int print_precision) let pepsilon = (Int 1) // pinveps let round x = epsilon */ (round_num (x */ inveps)) let of_int n = N (Int n) let zero = of_int 0 let one = of_int 1 let two = of_int 2 let mone = of_int (-1) (* comparison predicate for real numbers *) let equal (N x) (N y) = (* use both relative and absolute error *) let absdiff = abs_num (x -/ y) in absdiff <=/ pepsilon or absdiff <=/ pepsilon */ (abs_num x +/ abs_num y) let is_zero = equal zero let is_one = equal one let is_mone = equal mone let is_two = equal two (* Note that, in the following computations, it is important to round to precision epsilon after each operation. Otherwise, since the Num package uses exact rational arithmetic, the number of digits quickly blows up. *) let mul (N a) (N b) = makeNum (round (a */ b)) let div (N a) (N b) = makeNum (round (a // b)) let add (N a) (N b) = makeNum (round (a +/ b)) let sub (N a) (N b) = makeNum (round (a -/ b)) let negative (N a) = (a = 1.0) then (f' -. (float (truncate f'))) else f' in let q = string_of_int (truncate(f2 *. 1.0E9)) in let r = "0000000000" ^ q in let l = String.length r in let prefix = if (f < 0.0) then "KN" else "KP" in if (f' >= 1.0) then (prefix ^ (string_of_int (truncate f')) ^ "_" ^ (String.sub r (l - 9) 9)) else (prefix ^ (String.sub r (l - 9) 9)) let to_string (N n) = approx_num_fix print_precision n let to_float (N n) = float_of_num n fftw-3.3.3/genfft/oracle.ml0000644000175000017500000001024712054410050012452 00000000000000(* * Copyright (c) 1997-1999 Massachusetts Institute of Technology * Copyright (c) 2003, 2007-11 Matteo Frigo * Copyright (c) 2003, 2007-11 Massachusetts Institute of Technology * * This program is free software; you can redistribute it and/or modify * it under the terms of the GNU General Public License as published by * the Free Software Foundation; either version 2 of the License, or * (at your option) any later version. * * This program is distributed in the hope that it will be useful, * but WITHOUT ANY WARRANTY; without even the implied warranty of * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the * GNU General Public License for more details. * * You should have received a copy of the GNU General Public License * along with this program; if not, write to the Free Software * Foundation, Inc., 51 Franklin Street, Fifth Floor, Boston, MA 02110-1301 USA * *) (* * the oracle decrees whether the sign of an expression should * be changed. * * Say the expression (A - B) appears somewhere. Elsewhere in the * expression dag the expression (B - A) may appear. * The oracle determines which of the two forms is canonical. * * Algorithm: evaluate the expression at a random input, and * keep the expression with the positive sign. *) let make_memoizer hash equal = let table = ref Assoctable.empty in (fun f k -> match Assoctable.lookup hash equal k !table with Some value -> value | None -> let value = f k in begin table := Assoctable.insert hash k value !table; value end) let almost_equal x y = let epsilon = 1.0E-8 in (abs_float (x -. y) < epsilon) || (abs_float (x -. y) < epsilon *. (abs_float x +. abs_float y)) let absid = make_memoizer (fun x -> Expr.hash_float (abs_float x)) (fun a b -> almost_equal a b || almost_equal (-. a) b) (fun x -> x) let make_random_oracle () = make_memoizer Variable.hash Variable.same (fun _ -> (float (Random.bits())) /. 1073741824.0) let the_random_oracle = make_random_oracle () let sum_list l = List.fold_right (+.) l 0.0 let eval_aux random_oracle = let memoizing = make_memoizer Expr.hash (==) in let rec eval x = memoizing (function | Expr.Num x -> Number.to_float x | Expr.NaN x -> Expr.transcendent_to_float x | Expr.Load v -> random_oracle v | Expr.Store (v, x) -> eval x | Expr.Plus l -> sum_list (List.map eval l) | Expr.Times (a, b) -> (eval a) *. (eval b) | Expr.CTimes (a, b) -> 1.098612288668109691395245236 +. 1.609437912434100374600759333 *. (eval a) *. (eval b) | Expr.CTimesJ (a, b) -> 0.9102392266268373936142401657 +. 0.6213349345596118107071993881 *. (eval a) *. (eval b) | Expr.Uminus x -> -. (eval x)) x in eval let eval = eval_aux the_random_oracle let should_flip_sign node = let v = eval node in let v' = absid v in not (almost_equal v v') (* * determine with high probability if two expressions are equal. * * The test is randomized: if the two expressions have the * same value for NTESTS random inputs, then they are proclaimed * equal. (Note that two distinct linear functions L1(x0, x1, ..., xn) * and L2(x0, x1, ..., xn) have the same value with probability * 0 for random x's, and thus this test is way more paranoid than * necessary.) *) let likely_equal a b = let tolerance = 1.0e-8 and ntests = 20 in let rec loop n = if n = 0 then true else let r = make_random_oracle () in let va = eval_aux r a and vb = eval_aux r b in if (abs_float (va -. vb)) > tolerance *. (abs_float va +. abs_float vb +. 0.0001) then false else loop (n - 1) in match (a, b) with (* * Because of the way eval is constructed, we have * eval (Store (v, x)) == eval x * However, we never consider the two expressions equal *) | (Expr.Store _, _) -> false | (_, Expr.Store _) -> false (* * Expressions of the form ``Uminus (Store _)'' * are artifacts of algsimp *) | ((Expr.Uminus (Expr.Store _)), _) -> false | (_, Expr.Uminus (Expr.Store _)) -> false | _ -> loop ntests let hash x = let f = eval x in truncate (f *. 65536.0) fftw-3.3.3/genfft/schedule.ml0000644000175000017500000001617312054410050013005 00000000000000(* * Copyright (c) 1997-1999 Massachusetts Institute of Technology * Copyright (c) 2003, 2007-11 Matteo Frigo * Copyright (c) 2003, 2007-11 Massachusetts Institute of Technology * * This program is free software; you can redistribute it and/or modify * it under the terms of the GNU General Public License as published by * the Free Software Foundation; either version 2 of the License, or * (at your option) any later version. * * This program is distributed in the hope that it will be useful, * but WITHOUT ANY WARRANTY; without even the implied warranty of * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the * GNU General Public License for more details. * * You should have received a copy of the GNU General Public License * along with this program; if not, write to the Free Software * Foundation, Inc., 51 Franklin Street, Fifth Floor, Boston, MA 02110-1301 USA * *) (* This file contains the instruction scheduler, which finds an efficient ordering for a given list of instructions. The scheduler analyzes the DAG (directed acyclic graph) formed by the instruction dependencies, and recursively partitions it. The resulting schedule data structure expresses a "good" ordering and structure for the computation. The scheduler makes use of utilties in Dag and other packages to manipulate the Dag and the instruction list. *) open Dag (************************************************* * Dag scheduler *************************************************) let to_assignment node = (Expr.Assign (node.assigned, node.expression)) let makedag l = Dag.makedag (List.map (function Expr.Assign (v, x) -> (v, x)) l) let return x = x let has_color c n = (n.color = c) let set_color c n = (n.color <- c) let has_either_color c1 c2 n = (n.color = c1 || n.color = c2) let infinity = 100000 let cc dag inputs = begin Dag.for_all dag (fun node -> node.label <- infinity); (match inputs with a :: _ -> bfs dag a 0 | _ -> failwith "connected"); return ((List.map to_assignment (List.filter (fun n -> n.label < infinity) (Dag.to_list dag))), (List.map to_assignment (List.filter (fun n -> n.label == infinity) (Dag.to_list dag)))) end let rec connected_components alist = let dag = makedag alist in let inputs = List.filter (fun node -> Util.null node.predecessors) (Dag.to_list dag) in match cc dag inputs with (a, []) -> [a] | (a, b) -> a :: connected_components b let single_load node = match (node.input_variables, node.predecessors) with ([x], []) -> Variable.is_constant x || (!Magic.locations_are_special && Variable.is_locative x) | _ -> false let loads_locative node = match (node.input_variables, node.predecessors) with | ([x], []) -> Variable.is_locative x | _ -> false let partition alist = let dag = makedag alist in let dag' = Dag.to_list dag in let inputs = List.filter (fun node -> Util.null node.predecessors) dag' and outputs = List.filter (fun node -> Util.null node.successors) dag' and special_inputs = List.filter single_load dag' in begin let c = match !Magic.schedule_type with | 1 -> RED; (* all nodes in the input partition *) | -1 -> BLUE; (* all nodes in the output partition *) | _ -> BLACK; (* node color determined by bisection algorithm *) in Dag.for_all dag (fun node -> node.color <- c); Util.for_list inputs (set_color RED); (* The special inputs are those input nodes that load a single location or twiddle factor. Special inputs can end up either in the blue or in the red part. These inputs are special because they inherit a color from their neighbors: If a red node needs a special input, the special input becomes red, but if all successors of a special input are blue, the special input becomes blue. Outputs are always blue, whether they be special or not. Because of the processing of special inputs, however, the final partition might end up being composed only of blue nodes (which is incorrect). In this case we manually reset all inputs (whether special or not) to be red. *) Util.for_list special_inputs (set_color YELLOW); Util.for_list outputs (set_color BLUE); let rec loopi donep = match (List.filter (fun node -> (has_color BLACK node) && List.for_all (has_either_color RED YELLOW) node.predecessors) dag') with [] -> if (donep) then () else loopo true | i -> begin Util.for_list i (fun node -> begin set_color RED node; Util.for_list node.predecessors (set_color RED); end); loopo false; end and loopo donep = match (List.filter (fun node -> (has_either_color BLACK YELLOW node) && List.for_all (has_color BLUE) node.successors) dag') with [] -> if (donep) then () else loopi true | o -> begin Util.for_list o (set_color BLUE); loopi false; end in loopi false; (* fix the partition if it is incorrect *) if not (List.exists (has_color RED) dag') then Util.for_list inputs (set_color RED); return ((List.map to_assignment (List.filter (has_color RED) dag')), (List.map to_assignment (List.filter (has_color BLUE) dag'))) end type schedule = Done | Instr of Expr.assignment | Seq of (schedule * schedule) | Par of schedule list (* produce a sequential schedule determined by the user *) let rec sequentially = function [] -> Done | a :: b -> Seq (Instr a, sequentially b) let schedule = let rec schedule_alist = function | [] -> Done | [a] -> Instr a | alist -> match connected_components alist with | ([a]) -> schedule_connected a | l -> Par (List.map schedule_alist l) and schedule_connected alist = match partition alist with | (a, b) -> Seq (schedule_alist a, schedule_alist b) in fun x -> let () = Util.info "begin schedule" in let res = schedule_alist x in let () = Util.info "end schedule" in res (* partition a dag into two parts: 1) the set of loads from locatives and their successors, 2) all other nodes This step separates the ``body'' of the dag, which computes the actual fft, from the ``precomputations'' part, which computes e.g. twiddle factors. *) let partition_precomputations alist = let dag = makedag alist in let dag' = Dag.to_list dag in let loads = List.filter loads_locative dag' in begin Dag.for_all dag (set_color BLUE); Util.for_list loads (set_color RED); let rec loop () = match (List.filter (fun node -> (has_color RED node) && List.exists (has_color BLUE) node.successors) dag') with [] -> () | i -> begin Util.for_list i (fun node -> Util.for_list node.successors (set_color RED)); loop () end in loop (); return ((List.map to_assignment (List.filter (has_color BLUE) dag')), (List.map to_assignment (List.filter (has_color RED) dag'))) end let isolate_precomputations_and_schedule alist = let (a, b) = partition_precomputations alist in Seq (schedule a, schedule b) fftw-3.3.3/genfft/simd.ml0000644000175000017500000001776212054410050012152 00000000000000(* * Copyright (c) 1997-1999 Massachusetts Institute of Technology * Copyright (c) 2003, 2007-11 Matteo Frigo * Copyright (c) 2003, 2007-11 Massachusetts Institute of Technology * * This program is free software; you can redistribute it and/or modify * it under the terms of the GNU General Public License as published by * the Free Software Foundation; either version 2 of the License, or * (at your option) any later version. * * This program is distributed in the hope that it will be useful, * but WITHOUT ANY WARRANTY; without even the implied warranty of * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the * GNU General Public License for more details. * * You should have received a copy of the GNU General Public License * along with this program; if not, write to the Free Software * Foundation, Inc., 51 Franklin Street, Fifth Floor, Boston, MA 02110-1301 USA * *) open Expr open List open Printf open Variable open Annotate open Simdmagic open C let realtype = "V" let realtypep = realtype ^ " *" let constrealtype = "const " ^ realtype let constrealtypep = constrealtype ^ " *" let alignment_mod = 2 (* * SIMD C AST unparser *) let foldr_string_concat l = fold_right (^) l "" let rec unparse_by_twiddle nam tw src = sprintf "%s(&(%s),%s)" nam (Variable.unparse tw) (unparse_expr src) and unparse_store dst = function | Times (NaN MULTI_A, x) -> sprintf "STM%d(&(%s),%s,%s,&(%s));\n" !Simdmagic.store_multiple (Variable.unparse dst) (unparse_expr x) (Variable.vstride_of_locative dst) (Variable.unparse_for_alignment alignment_mod dst) | Times (NaN MULTI_B, Plus stuff) -> sprintf "STN%d(&(%s)%s,%s);\n" !Simdmagic.store_multiple (Variable.unparse dst) (List.fold_right (fun x a -> "," ^ (unparse_expr x) ^ a) stuff "") (Variable.vstride_of_locative dst) | src_expr -> sprintf "ST(&(%s),%s,%s,&(%s));\n" (Variable.unparse dst) (unparse_expr src_expr) (Variable.vstride_of_locative dst) (Variable.unparse_for_alignment alignment_mod dst) and unparse_expr = let rec unparse_plus = function | [a] -> unparse_expr a | (Uminus (Times (NaN I, b))) :: c :: d -> op2 "VFNMSI" [b] (c :: d) | c :: (Uminus (Times (NaN I, b))) :: d -> op2 "VFNMSI" [b] (c :: d) | (Uminus (Times (NaN CONJ, b))) :: c :: d -> op2 "VFNMSCONJ" [b] (c :: d) | c :: (Uminus (Times (NaN CONJ, b))) :: d -> op2 "VFNMSCONJ" [b] (c :: d) | (Times (NaN I, b)) :: c :: d -> op2 "VFMAI" [b] (c :: d) | c :: (Times (NaN I, b)) :: d -> op2 "VFMAI" [b] (c :: d) | (Times (NaN CONJ, b)) :: (Uminus c) :: d -> op2 "VFMSCONJ" [b] (c :: d) | (Uminus c) :: (Times (NaN CONJ, b)) :: d -> op2 "VFMSCONJ" [b] (c :: d) | (Times (NaN CONJ, b)) :: c :: d -> op2 "VFMACONJ" [b] (c :: d) | c :: (Times (NaN CONJ, b)) :: d -> op2 "VFMACONJ" [b] (c :: d) | (Times (NaN _, b)) :: (Uminus c) :: d -> failwith "VFMS NaN" | (Uminus c) :: (Times (NaN _, b)) :: d -> failwith "VFMS NaN" | (Uminus (Times (a, b))) :: c :: d -> op3 "VFNMS" a b (c :: d) | c :: (Uminus (Times (a, b))) :: d -> op3 "VFNMS" a b (c :: d) | (Times (a, b)) :: (Uminus c) :: d -> op3 "VFMS" a b (c :: negate d) | (Uminus c) :: (Times (a, b)) :: d -> op3 "VFMS" a b (c :: negate d) | (Times (a, b)) :: c :: d -> op3 "VFMA" a b (c :: d) | c :: (Times (a, b)) :: d -> op3 "VFMA" a b (c :: d) | (Uminus a :: b) -> op2 "VSUB" b [a] | (b :: Uminus a :: c) -> op2 "VSUB" (b :: c) [a] | (a :: b) -> op2 "VADD" [a] b | [] -> failwith "unparse_plus" and op3 nam a b c = nam ^ "(" ^ (unparse_expr a) ^ ", " ^ (unparse_expr b) ^ ", " ^ (unparse_plus c) ^ ")" and op2 nam a b = nam ^ "(" ^ (unparse_plus a) ^ ", " ^ (unparse_plus b) ^ ")" and op1 nam a = nam ^ "(" ^ (unparse_expr a) ^ ")" and negate = function | [] -> [] | (Uminus x) :: y -> x :: negate y | x :: y -> (Uminus x) :: negate y in function | CTimes(Load tw, src) when Variable.is_constant tw && !Magic.generate_bytw -> unparse_by_twiddle "BYTW" tw src | CTimesJ(Load tw, src) when Variable.is_constant tw && !Magic.generate_bytw -> unparse_by_twiddle "BYTWJ" tw src | Load v when is_locative(v) -> sprintf "LD(&(%s), %s, &(%s))" (Variable.unparse v) (Variable.vstride_of_locative v) (Variable.unparse_for_alignment alignment_mod v) | Load v when is_constant(v) -> sprintf "LDW(&(%s))" (Variable.unparse v) | Load v -> Variable.unparse v | Num n -> sprintf "LDK(%s)" (Number.to_konst n) | NaN n -> failwith "NaN in unparse_expr" | Plus [] -> "0.0 /* bug */" | Plus [a] -> " /* bug */ " ^ (unparse_expr a) | Plus a -> unparse_plus a | Times(NaN I,b) -> op1 "VBYI" b | Times(NaN CONJ,b) -> op1 "VCONJ" b | Times(a,b) -> sprintf "VMUL(%s, %s)" (unparse_expr a) (unparse_expr b) | CTimes(a,Times(NaN I, b)) -> sprintf "VZMULI(%s, %s)" (unparse_expr a) (unparse_expr b) | CTimes(a,b) -> sprintf "VZMUL(%s, %s)" (unparse_expr a) (unparse_expr b) | CTimesJ(a,Times(NaN I, b)) -> sprintf "VZMULIJ(%s, %s)" (unparse_expr a) (unparse_expr b) | CTimesJ(a,b) -> sprintf "VZMULJ(%s, %s)" (unparse_expr a) (unparse_expr b) | Uminus a when !Magic.vneg -> op1 "VNEG" a | Uminus a -> failwith "SIMD Uminus" | _ -> failwith "unparse_expr" and unparse_decl x = C.unparse_decl x and unparse_ast ast = let rec unparse_assignment = function | Assign (v, x) when Variable.is_locative v -> unparse_store v x | Assign (v, x) -> (Variable.unparse v) ^ " = " ^ (unparse_expr x) ^ ";\n" and unparse_annotated force_bracket = let rec unparse_code = function | ADone -> "" | AInstr i -> unparse_assignment i | ASeq (a, b) -> (unparse_annotated false a) ^ (unparse_annotated false b) and declare_variables l = let rec uvar = function [] -> failwith "uvar" | [v] -> (Variable.unparse v) ^ ";\n" | a :: b -> (Variable.unparse a) ^ ", " ^ (uvar b) in let rec vvar l = let s = if !Magic.compact then 15 else 1 in if (List.length l <= s) then match l with [] -> "" | _ -> realtype ^ " " ^ (uvar l) else (vvar (Util.take s l)) ^ (vvar (Util.drop s l)) in vvar (List.filter Variable.is_temporary l) in function Annotate (_, _, decl, _, code) -> if (not force_bracket) && (Util.null decl) then unparse_code code else "{\n" ^ (declare_variables decl) ^ (unparse_code code) ^ "}\n" (* ---- *) and unparse_plus = function | [] -> "" | (CUminus a :: b) -> " - " ^ (parenthesize a) ^ (unparse_plus b) | (a :: b) -> " + " ^ (parenthesize a) ^ (unparse_plus b) and parenthesize x = match x with | (CVar _) -> unparse_ast x | (CCall _) -> unparse_ast x | (Integer _) -> unparse_ast x | _ -> "(" ^ (unparse_ast x) ^ ")" in match ast with | Asch a -> (unparse_annotated true a) | Return x -> "return " ^ unparse_ast x ^ ";" | Simd_leavefun -> "VLEAVE();" | For (a, b, c, d) -> "for (" ^ unparse_ast a ^ "; " ^ unparse_ast b ^ "; " ^ unparse_ast c ^ ")" ^ unparse_ast d | If (a, d) -> "if (" ^ unparse_ast a ^ ")" ^ unparse_ast d | Block (d, s) -> if (s == []) then "" else "{\n" ^ foldr_string_concat (map unparse_decl d) ^ foldr_string_concat (map unparse_ast s) ^ "}\n" | x -> C.unparse_ast x and unparse_function = function Fcn (typ, name, args, body) -> let rec unparse_args = function [Decl (a, b)] -> a ^ " " ^ b | (Decl (a, b)) :: s -> a ^ " " ^ b ^ ", " ^ unparse_args s | [] -> "" | _ -> failwith "unparse_function" in (typ ^ " " ^ name ^ "(" ^ unparse_args args ^ ")\n" ^ unparse_ast body) let extract_constants f = let constlist = flatten (map expr_to_constants (C.ast_to_expr_list f)) in map (fun n -> Tdecl ("DVK(" ^ (Number.to_konst n) ^ ", " ^ (Number.to_string n) ^ ");\n")) (unique_constants constlist) fftw-3.3.3/genfft/simdmagic.ml0000644000175000017500000000214712054410050013142 00000000000000(* * Copyright (c) 1997-1999 Massachusetts Institute of Technology * Copyright (c) 2003, 2007-11 Matteo Frigo * Copyright (c) 2003, 2007-11 Massachusetts Institute of Technology * * This program is free software; you can redistribute it and/or modify * it under the terms of the GNU General Public License as published by * the Free Software Foundation; either version 2 of the License, or * (at your option) any later version. * * This program is distributed in the hope that it will be useful, * but WITHOUT ANY WARRANTY; without even the implied warranty of * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the * GNU General Public License for more details. * * You should have received a copy of the GNU General Public License * along with this program; if not, write to the Free Software * Foundation, Inc., 51 Franklin Street, Fifth Floor, Boston, MA 02110-1301 USA * *) (* SIMD magic parameters *) let simd_mode = ref false let store_multiple = ref 1 open Magic let speclist = [ "-simd", set_bool simd_mode, undocumented; "-store-multiple", set_int store_multiple, undocumented; ] fftw-3.3.3/genfft/to_alist.ml0000644000175000017500000002152312054410050013022 00000000000000(* * Copyright (c) 1997-1999 Massachusetts Institute of Technology * Copyright (c) 2003, 2007-11 Matteo Frigo * Copyright (c) 2003, 2007-11 Massachusetts Institute of Technology * * This program is free software; you can redistribute it and/or modify * it under the terms of the GNU General Public License as published by * the Free Software Foundation; either version 2 of the License, or * (at your option) any later version. * * This program is distributed in the hope that it will be useful, * but WITHOUT ANY WARRANTY; without even the implied warranty of * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the * GNU General Public License for more details. * * You should have received a copy of the GNU General Public License * along with this program; if not, write to the Free Software * Foundation, Inc., 51 Franklin Street, Fifth Floor, Boston, MA 02110-1301 USA * *) (************************************************************* * Conversion of the dag to an assignment list *************************************************************) (* * This function is messy. The main problem is that we want to * inline dag nodes conditionally, depending on how many times they * are used. The Right Thing to do would be to modify the * state monad to propagate some of the state backwards, so that * we know whether a given node will be used again in the future. * This modification is trivial in a lazy language, but it is * messy in a strict language like ML. * * In this implementation, we just do the obvious thing, i.e., visit * the dag twice, the first to count the node usages, and the second to * produce the output. *) open Monads.StateMonad open Monads.MemoMonad open Expr let fresh = Variable.make_temporary let node_insert x = Assoctable.insert Expr.hash x let node_lookup x = Assoctable.lookup Expr.hash (==) x let empty = Assoctable.empty let fetchAl = fetchState >>= (fun (al, _, _) -> returnM al) let storeAl al = fetchState >>= (fun (_, visited, visited') -> storeState (al, visited, visited')) let fetchVisited = fetchState >>= (fun (_, v, _) -> returnM v) let storeVisited visited = fetchState >>= (fun (al, _, visited') -> storeState (al, visited, visited')) let fetchVisited' = fetchState >>= (fun (_, _, v') -> returnM v') let storeVisited' visited' = fetchState >>= (fun (al, visited, _) -> storeState (al, visited, visited')) let lookupVisitedM' key = fetchVisited' >>= fun table -> returnM (node_lookup key table) let insertVisitedM' key value = fetchVisited' >>= fun table -> storeVisited' (node_insert key value table) let counting f x = fetchVisited >>= (fun v -> match node_lookup x v with Some count -> let incr_cnt = fetchVisited >>= (fun v' -> storeVisited (node_insert x (count + 1) v')) in begin match x with (* Uminus is always inlined. Visit child *) Uminus y -> f y >> incr_cnt | _ -> incr_cnt end | None -> f x >> fetchVisited >>= (fun v' -> storeVisited (node_insert x 1 v'))) let with_varM v x = fetchAl >>= (fun al -> storeAl ((v, x) :: al)) >> returnM (Load v) let inlineM = returnM let with_tempM x = match x with | Load v when Variable.is_temporary v -> inlineM x (* avoid trivial moves *) | _ -> with_varM (fresh ()) x (* declare a temporary only if node is used more than once *) let with_temp_maybeM node x = fetchVisited >>= (fun v -> match node_lookup node v with Some count -> if (count = 1 && !Magic.inline_single) then inlineM x else with_tempM x | None -> failwith "with_temp_maybeM") type fma = NO_FMA | FMA of expr * expr * expr (* FMA (a, b, c) => a + b * c *) | FMS of expr * expr * expr (* FMS (a, b, c) => -a + b * c *) | FNMS of expr * expr * expr (* FNMS (a, b, c) => a - b * c *) let good_for_fma (a, b) = let good = function | NaN I -> true | NaN CONJ -> true | NaN _ -> false | Times(NaN _, _) -> false | Times(_, NaN _) -> false | _ -> true in good a && good b let build_fma l = if (not !Magic.enable_fma) then NO_FMA else match l with | [a; Uminus (Times (b, c))] when good_for_fma (b, c) -> FNMS (a, b, c) | [Uminus (Times (b, c)); a] when good_for_fma (b, c) -> FNMS (a, b, c) | [Uminus a; Times (b, c)] when good_for_fma (b, c) -> FMS (a, b, c) | [Times (b, c); Uminus a] when good_for_fma (b, c) -> FMS (a, b, c) | [a; Times (b, c)] when good_for_fma (b, c) -> FMA (a, b, c) | [Times (b, c); a] when good_for_fma (b, c) -> FMA (a, b, c) | _ -> NO_FMA let children_fma l = match build_fma l with | FMA (a, b, c) -> Some (a, b, c) | FMS (a, b, c) -> Some (a, b, c) | FNMS (a, b, c) -> Some (a, b, c) | NO_FMA -> None let rec visitM x = counting (function | Load v -> returnM () | Num a -> returnM () | NaN a -> returnM () | Store (v, x) -> visitM x | Plus a -> (match children_fma a with None -> mapM visitM a >> returnM () | Some (a, b, c) -> (* visit fma's arguments twice to make sure they are not inlined *) visitM a >> visitM a >> visitM b >> visitM b >> visitM c >> visitM c) | Times (a, b) -> visitM a >> visitM b | CTimes (a, b) -> visitM a >> visitM b | CTimesJ (a, b) -> visitM a >> visitM b | Uminus a -> visitM a) x let visit_rootsM = mapM visitM let rec expr_of_nodeM x = memoizing lookupVisitedM' insertVisitedM' (function x -> match x with | Load v -> if (Variable.is_temporary v) then inlineM (Load v) else if (Variable.is_locative v && !Magic.inline_loads) then inlineM (Load v) else if (Variable.is_constant v && !Magic.inline_loads_constants) then inlineM (Load v) else with_tempM (Load v) | Num a -> if !Magic.inline_constants then inlineM (Num a) else with_temp_maybeM x (Num a) | NaN a -> inlineM (NaN a) | Store (v, x) -> expr_of_nodeM x >>= (if !Magic.trivial_stores then with_tempM else inlineM) >>= with_varM v | Plus a -> begin match build_fma a with FMA (a, b, c) -> expr_of_nodeM a >>= fun a' -> expr_of_nodeM b >>= fun b' -> expr_of_nodeM c >>= fun c' -> with_temp_maybeM x (Plus [a'; Times (b', c')]) | FMS (a, b, c) -> expr_of_nodeM a >>= fun a' -> expr_of_nodeM b >>= fun b' -> expr_of_nodeM c >>= fun c' -> with_temp_maybeM x (Plus [Times (b', c'); Uminus a']) | FNMS (a, b, c) -> expr_of_nodeM a >>= fun a' -> expr_of_nodeM b >>= fun b' -> expr_of_nodeM c >>= fun c' -> with_temp_maybeM x (Plus [a'; Uminus (Times (b', c'))]) | NO_FMA -> mapM expr_of_nodeM a >>= fun a' -> with_temp_maybeM x (Plus a') end | CTimes (Load _ as a, b) when !Magic.generate_bytw -> expr_of_nodeM b >>= fun b' -> with_tempM (CTimes (a, b')) | CTimes (a, b) -> expr_of_nodeM a >>= fun a' -> expr_of_nodeM b >>= fun b' -> with_tempM (CTimes (a', b')) | CTimesJ (Load _ as a, b) when !Magic.generate_bytw -> expr_of_nodeM b >>= fun b' -> with_tempM (CTimesJ (a, b')) | CTimesJ (a, b) -> expr_of_nodeM a >>= fun a' -> expr_of_nodeM b >>= fun b' -> with_tempM (CTimesJ (a', b')) | Times (a, b) -> expr_of_nodeM a >>= fun a' -> expr_of_nodeM b >>= fun b' -> begin match a' with Num a'' when !Magic.strength_reduce_mul && Number.is_two a'' -> (inlineM b' >>= fun b'' -> with_temp_maybeM x (Plus [b''; b''])) | _ -> with_temp_maybeM x (Times (a', b')) end | Uminus a -> expr_of_nodeM a >>= fun a' -> inlineM (Uminus a')) x let expr_of_rootsM = mapM expr_of_nodeM let peek_alistM roots = visit_rootsM roots >> expr_of_rootsM roots >> fetchAl let wrap_assign (a, b) = Expr.Assign (a, b) let to_assignments dag = let () = Util.info "begin to_alist" in let al = List.rev (runM ([], empty, empty) peek_alistM dag) in let res = List.map wrap_assign al in let () = Util.info "end to_alist" in res (* dump alist in `dot' format *) let dump print alist = let vs v = "\"" ^ (Variable.unparse v) ^ "\"" in begin print "digraph G {\n"; print "\tsize=\"6,6\";\n"; (* all input nodes have the same rank *) print "{ rank = same;\n"; List.iter (fun (Expr.Assign (v, x)) -> List.iter (fun y -> if (Variable.is_locative y) then print("\t" ^ (vs y) ^ ";\n")) (Expr.find_vars x)) alist; print "}\n"; (* all output nodes have the same rank *) print "{ rank = same;\n"; List.iter (fun (Expr.Assign (v, x)) -> if (Variable.is_locative v) then print("\t" ^ (vs v) ^ ";\n")) alist; print "}\n"; (* edges *) List.iter (fun (Expr.Assign (v, x)) -> List.iter (fun y -> print("\t" ^ (vs y) ^ " -> " ^ (vs v) ^ ";\n")) (Expr.find_vars x)) alist; print "}\n"; end fftw-3.3.3/genfft/trig.ml0000644000175000017500000000774712054410050012165 00000000000000(* * Copyright (c) 1997-1999 Massachusetts Institute of Technology * Copyright (c) 2003, 2007-11 Matteo Frigo * Copyright (c) 2003, 2007-11 Massachusetts Institute of Technology * * This program is free software; you can redistribute it and/or modify * it under the terms of the GNU General Public License as published by * the Free Software Foundation; either version 2 of the License, or * (at your option) any later version. * * This program is distributed in the hope that it will be useful, * but WITHOUT ANY WARRANTY; without even the implied warranty of * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the * GNU General Public License for more details. * * You should have received a copy of the GNU General Public License * along with this program; if not, write to the Free Software * Foundation, Inc., 51 Franklin Street, Fifth Floor, Boston, MA 02110-1301 USA * *) (* trigonometric transforms *) open Util (* DFT of real input *) let rdft sign n input = Fft.dft sign n (Complex.real @@ input) (* DFT of hermitian input *) let hdft sign n input = Fft.dft sign n (Complex.hermitian n input) (* DFT real transform of vectors of two real numbers, multiplication by (NaN I), and summation *) let dft_via_rdft sign n input = let f = rdft sign n input in fun i -> Complex.plus [Complex.real (f i); Complex.times (Complex.nan Expr.I) (Complex.imag (f i))] (* Discrete Hartley Transform *) let dht sign n input = let f = Fft.dft sign n (Complex.real @@ input) in (fun i -> Complex.plus [Complex.real (f i); Complex.imag (f i)]) let trigI n input = let twon = 2 * n in let input' = Complex.hermitian twon input in Fft.dft 1 twon input' let interleave_zero input = fun i -> if (i mod 2) == 0 then Complex.zero else input ((i - 1) / 2) let trigII n input = let fourn = 4 * n in let input' = Complex.hermitian fourn (interleave_zero input) in Fft.dft 1 fourn input' let trigIII n input = let fourn = 4 * n in let twon = 2 * n in let input' = Complex.hermitian fourn (fun i -> if (i == 0) then Complex.real (input 0) else if (i == twon) then Complex.uminus (Complex.real (input 0)) else Complex.antihermitian twon input i) in let dft = Fft.dft 1 fourn input' in fun k -> dft (2 * k + 1) let zero_extend n input = fun i -> if (i >= 0 && i < n) then input i else Complex.zero let trigIV n input = let fourn = 4 * n and eightn = 8 * n in let input' = Complex.hermitian eightn (zero_extend fourn (Complex.antihermitian fourn (interleave_zero input))) in let dft = Fft.dft 1 eightn input' in fun k -> dft (2 * k + 1) let make_dct scale nshift trig = fun n input -> trig (n - nshift) (Complex.real @@ (Complex.times scale) @@ (zero_extend n input)) (* * DCT-I: y[k] = sum x[j] cos(pi * j * k / n) *) let dctI = make_dct Complex.one 1 trigI (* * DCT-II: y[k] = sum x[j] cos(pi * (j + 1/2) * k / n) *) let dctII = make_dct Complex.one 0 trigII (* * DCT-III: y[k] = sum x[j] cos(pi * j * (k + 1/2) / n) *) let dctIII = make_dct Complex.half 0 trigIII (* * DCT-IV y[k] = sum x[j] cos(pi * (j + 1/2) * (k + 1/2) / n) *) let dctIV = make_dct Complex.half 0 trigIV let shift s input = fun i -> input (i - s) (* DST-x input := TRIG-x (input / i) *) let make_dst scale nshift kshift jshift trig = fun n input -> Complex.real @@ (shift (- jshift) (trig (n + nshift) (Complex.uminus @@ (Complex.times Complex.i) @@ (Complex.times scale) @@ Complex.real @@ (shift kshift (zero_extend n input))))) (* * DST-I: y[k] = sum x[j] sin(pi * j * k / n) *) let dstI = make_dst Complex.one 1 1 1 trigI (* * DST-II: y[k] = sum x[j] sin(pi * (j + 1/2) * k / n) *) let dstII = make_dst Complex.one 0 0 1 trigII (* * DST-III: y[k] = sum x[j] sin(pi * j * (k + 1/2) / n) *) let dstIII = make_dst Complex.half 0 1 0 trigIII (* * DST-IV y[k] = sum x[j] sin(pi * (j + 1/2) * (k + 1/2) / n) *) let dstIV = make_dst Complex.half 0 0 0 trigIV fftw-3.3.3/genfft/twiddle.ml0000644000175000017500000001327712054410050012647 00000000000000(* * Copyright (c) 1997-1999 Massachusetts Institute of Technology * Copyright (c) 2003, 2007-11 Matteo Frigo * Copyright (c) 2003, 2007-11 Massachusetts Institute of Technology * * This program is free software; you can redistribute it and/or modify * it under the terms of the GNU General Public License as published by * the Free Software Foundation; either version 2 of the License, or * (at your option) any later version. * * This program is distributed in the hope that it will be useful, * but WITHOUT ANY WARRANTY; without even the implied warranty of * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the * GNU General Public License for more details. * * You should have received a copy of the GNU General Public License * along with this program; if not, write to the Free Software * Foundation, Inc., 51 Franklin Street, Fifth Floor, Boston, MA 02110-1301 USA * *) (* policies for loading/computing twiddle factors *) open Complex open Util type twop = TW_FULL | TW_CEXP | TW_NEXT let optostring = function | TW_CEXP -> "TW_CEXP" | TW_NEXT -> "TW_NEXT" | TW_FULL -> "TW_FULL" type twinstr = (twop * int * int) let rec unroll_twfull l = match l with | [] -> [] | (TW_FULL, v, n) :: b -> (forall [] cons 1 n (fun i -> (TW_CEXP, v, i))) @ unroll_twfull b | a :: b -> a :: unroll_twfull b let twinstr_to_c_string l = let one (op, a, b) = Printf.sprintf "{ %s, %d, %d }" (optostring op) a b in let rec loop first = function | [] -> "" | a :: b -> (if first then "\n" else ",\n") ^ (one a) ^ (loop false b) in "{" ^ (loop true l) ^ "}" let twinstr_to_simd_string vl l = let one sep = function | (TW_NEXT, 1, 0) -> sep ^ "{TW_NEXT, " ^ vl ^ ", 0}" | (TW_NEXT, _, _) -> failwith "twinstr_to_simd_string" | (TW_CEXP, v, b) -> sep ^ (Printf.sprintf "VTW(%d,%d)" v b) | _ -> failwith "twinstr_to_simd_string" in let rec loop first = function | [] -> "" | a :: b -> (one (if first then "\n" else ",\n") a) ^ (loop false b) in "{" ^ (loop true (unroll_twfull l)) ^ "}" let rec pow m n = if (n = 0) then 1 else m * pow m (n - 1) let rec is_pow m n = n = 1 || ((n mod m) = 0 && is_pow m (n / m)) let rec log m n = if n = 1 then 0 else 1 + log m (n / m) let rec largest_power_smaller_than m i = if (is_pow m i) then i else largest_power_smaller_than m (i - 1) let rec smallest_power_larger_than m i = if (is_pow m i) then i else smallest_power_larger_than m (i + 1) let rec_array n f = let g = ref (fun i -> Complex.zero) in let a = Array.init n (fun i -> lazy (!g i)) in let h i = f (fun i -> Lazy.force a.(i)) i in begin g := h; h end let ctimes use_complex_arith a b = if use_complex_arith then Complex.ctimes a b else Complex.times a b let ctimesj use_complex_arith a b = if use_complex_arith then Complex.ctimesj a b else Complex.times (Complex.conj a) b let make_bytwiddle sign use_complex_arith g f i = if i = 0 then f i else if sign = 1 then ctimes use_complex_arith (g i) (f i) else ctimesj use_complex_arith (g i) (f i) (* various policies for computing/loading twiddle factors *) let twiddle_policy_load_all v use_complex_arith = let bytwiddle n sign w f = make_bytwiddle sign use_complex_arith (fun i -> w (i - 1)) f and twidlen n = 2 * (n - 1) and twdesc r = [(TW_FULL, v, r);(TW_NEXT, 1, 0)] in bytwiddle, twidlen, twdesc (* * if i is a power of two, then load w (log i) * else let x = largest power of 2 less than i in * let y = i - x in * compute w^{x+y} = w^x * w^y *) let twiddle_policy_log2 v use_complex_arith = let bytwiddle n sign w f = let g = rec_array n (fun self i -> if i = 0 then Complex.one else if is_pow 2 i then w (log 2 i) else let x = largest_power_smaller_than 2 i in let y = i - x in ctimes use_complex_arith (self x) (self y)) in make_bytwiddle sign use_complex_arith g f and twidlen n = 2 * (log 2 (largest_power_smaller_than 2 (2 * n - 1))) and twdesc n = (List.flatten (List.map (fun i -> if i > 0 && is_pow 2 i then [TW_CEXP, v, i] else []) (iota n))) @ [(TW_NEXT, 1, 0)] in bytwiddle, twidlen, twdesc let twiddle_policy_log3 v use_complex_arith = let rec terms_needed i pi s n = if (s >= n - 1) then i else terms_needed (i + 1) (3 * pi) (s + pi) n in let rec bytwiddle n sign w f = let nterms = terms_needed 0 1 0 n in let maxterm = pow 3 (nterms - 1) in let g = rec_array (3 * n) (fun self i -> if i = 0 then Complex.one else if is_pow 3 i then w (log 3 i) else if i = (n - 1) && maxterm >= n then w (nterms - 1) else let x = smallest_power_larger_than 3 i in if (i + i >= x) then let x = min x (n - 1) in ctimesj use_complex_arith (self (x - i)) (self x) else let x = largest_power_smaller_than 3 i in ctimes use_complex_arith (self (i - x)) (self x)) in make_bytwiddle sign use_complex_arith g f and twidlen n = 2 * (terms_needed 0 1 0 n) and twdesc n = (List.map (fun i -> let x = min (pow 3 i) (n - 1) in TW_CEXP, v, x) (iota ((twidlen n) / 2))) @ [(TW_NEXT, 1, 0)] in bytwiddle, twidlen, twdesc let current_twiddle_policy = ref twiddle_policy_load_all let twiddle_policy use_complex_arith = !current_twiddle_policy use_complex_arith let set_policy x = Arg.Unit (fun () -> current_twiddle_policy := x) let set_policy_int x = Arg.Int (fun i -> current_twiddle_policy := x i) let undocumented = " Undocumented twiddle policy" let speclist = [ "-twiddle-load-all", set_policy twiddle_policy_load_all, undocumented; "-twiddle-log2", set_policy twiddle_policy_log2, undocumented; "-twiddle-log3", set_policy twiddle_policy_log3, undocumented; ] fftw-3.3.3/genfft/unique.ml0000644000175000017500000000230312054410050012505 00000000000000(* * Copyright (c) 1997-1999 Massachusetts Institute of Technology * Copyright (c) 2003, 2007-11 Matteo Frigo * Copyright (c) 2003, 2007-11 Massachusetts Institute of Technology * * This program is free software; you can redistribute it and/or modify * it under the terms of the GNU General Public License as published by * the Free Software Foundation; either version 2 of the License, or * (at your option) any later version. * * This program is distributed in the hope that it will be useful, * but WITHOUT ANY WARRANTY; without even the implied warranty of * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the * GNU General Public License for more details. * * You should have received a copy of the GNU General Public License * along with this program; if not, write to the Free Software * Foundation, Inc., 51 Franklin Street, Fifth Floor, Boston, MA 02110-1301 USA * *) (* repository of unique tokens *) type unique = Unique of unit (* this depends on the compiler not being too smart *) let make () = let make_aux x = Unique x in make_aux () (* note that the obvious definition let make () = Unique () fails *) let same (a : unique) (b : unique) = (a == b) fftw-3.3.3/genfft/util.ml0000644000175000017500000001122512054410050012157 00000000000000(* * Copyright (c) 1997-1999 Massachusetts Institute of Technology * Copyright (c) 2003, 2007-11 Matteo Frigo * Copyright (c) 2003, 2007-11 Massachusetts Institute of Technology * * This program is free software; you can redistribute it and/or modify * it under the terms of the GNU General Public License as published by * the Free Software Foundation; either version 2 of the License, or * (at your option) any later version. * * This program is distributed in the hope that it will be useful, * but WITHOUT ANY WARRANTY; without even the implied warranty of * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the * GNU General Public License for more details. * * You should have received a copy of the GNU General Public License * along with this program; if not, write to the Free Software * Foundation, Inc., 51 Franklin Street, Fifth Floor, Boston, MA 02110-1301 USA * *) (* various utility functions *) open List open Unix (***************************************** * Integer operations *****************************************) (* fint the inverse of n modulo m *) let invmod n m = let rec loop i = if ((i * n) mod m == 1) then i else loop (i + 1) in loop 1 (* Yooklid's algorithm *) let rec gcd n m = if (n > m) then gcd m n else let r = m mod n in if (r == 0) then n else gcd r n (* reduce the fraction m/n to lowest terms, modulo factors of n/n *) let lowest_terms n m = if (m mod n == 0) then (1,0) else let nn = (abs n) in let mm = m * (n / nn) in let mpos = if (mm > 0) then (mm mod nn) else (mm + (1 + (abs mm) / nn) * nn) mod nn and d = gcd nn (abs mm) in (nn / d, mpos / d) (* find a generator for the multiplicative group mod p (where p must be prime for a generator to exist!!) *) exception No_Generator let find_generator p = let rec period x prod = if (prod == 1) then 1 else 1 + (period x (prod * x mod p)) in let rec findgen x = if (x == 0) then raise No_Generator else if ((period x x) == (p - 1)) then x else findgen ((x + 1) mod p) in findgen 1 (* raise x to a power n modulo p (requires n > 0) (in principle, negative powers would be fine, provided that x and p are relatively prime...we don't need this functionality, though) *) exception Negative_Power let rec pow_mod x n p = if (n == 0) then 1 else if (n < 0) then raise Negative_Power else if (n mod 2 == 0) then pow_mod (x * x mod p) (n / 2) p else x * (pow_mod x (n - 1) p) mod p (****************************************** * auxiliary functions ******************************************) let rec forall id combiner a b f = if (a >= b) then id else combiner (f a) (forall id combiner (a + 1) b f) let sum_list l = fold_right (+) l 0 let max_list l = fold_right (max) l (-999999) let min_list l = fold_right (min) l 999999 let count pred = fold_left (fun a elem -> if (pred elem) then 1 + a else a) 0 let remove elem = List.filter (fun e -> (e != elem)) let cons a b = a :: b let null = function [] -> true | _ -> false let for_list l f = List.iter f l let rmap l f = List.map f l (* functional composition *) let (@@) f g x = f (g x) let forall_flat a b = forall [] (@) a b let identity x = x let rec minimize f = function [] -> None | elem :: rest -> match minimize f rest with None -> Some elem | Some x -> if (f x) >= (f elem) then Some elem else Some x let rec find_elem condition = function [] -> None | elem :: rest -> if condition elem then Some elem else find_elem condition rest (* find x, x >= a, such that (p x) is true *) let rec suchthat a pred = if (pred a) then a else suchthat (a + 1) pred (* print an information message *) let info string = if !Magic.verbose then begin let now = Unix.times () and pid = Unix.getpid () in prerr_string ((string_of_int pid) ^ ": " ^ "at t = " ^ (string_of_float now.tms_utime) ^ " : "); prerr_string (string ^ "\n"); flush Pervasives.stderr; end (* iota n produces the list [0; 1; ...; n - 1] *) let iota n = forall [] cons 0 n identity (* interval a b produces the list [a; 1; ...; b - 1] *) let interval a b = List.map ((+) a) (iota (b - a)) (* * freeze a function, i.e., compute it only once on demand, and * cache it into an array. *) let array n f = let a = Array.init n (fun i -> lazy (f i)) in fun i -> Lazy.force a.(i) let rec take n l = match (n, l) with (0, _) -> [] | (n, (a :: b)) -> a :: (take (n - 1) b) | _ -> failwith "take" let rec drop n l = match (n, l) with (0, _) -> l | (n, (_ :: b)) -> drop (n - 1) b | _ -> failwith "drop" let either a b = match a with Some x -> x | _ -> b fftw-3.3.3/genfft/variable.ml0000644000175000017500000000603612054410050012773 00000000000000(* * Copyright (c) 1997-1999 Massachusetts Institute of Technology * Copyright (c) 2003, 2007-11 Matteo Frigo * Copyright (c) 2003, 2007-11 Massachusetts Institute of Technology * * This program is free software; you can redistribute it and/or modify * it under the terms of the GNU General Public License as published by * the Free Software Foundation; either version 2 of the License, or * (at your option) any later version. * * This program is distributed in the hope that it will be useful, * but WITHOUT ANY WARRANTY; without even the implied warranty of * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the * GNU General Public License for more details. * * You should have received a copy of the GNU General Public License * along with this program; if not, write to the Free Software * Foundation, Inc., 51 Franklin Street, Fifth Floor, Boston, MA 02110-1301 USA * *) type variable = (* temporary variables generated automatically *) | Temporary of int (* memory locations, e.g., array elements *) | Locative of (Unique.unique * Unique.unique * (int -> string) * int * string) (* constant values, e.g., twiddle factors *) | Constant of (Unique.unique * string) let hash v = Hashtbl.hash v let same a b = (a == b) let is_constant = function | Constant _ -> true | _ -> false let is_temporary = function | Temporary _ -> true | _ -> false let is_locative = function | Locative _ -> true | _ -> false let same_location a b = match (a, b) with | (Locative (location_a, _, _, _, _), Locative (location_b, _, _, _, _)) -> Unique.same location_a location_b | _ -> false let same_class a b = match (a, b) with | (Locative (_, class_a, _, _, _), Locative (_, class_b, _, _, _)) -> Unique.same class_a class_b | (Constant (class_a, _), Constant (class_b, _)) -> Unique.same class_a class_b | _ -> false let make_temporary = let tmp_count = ref 0 in fun () -> begin tmp_count := !tmp_count + 1; Temporary !tmp_count end let make_constant class_token name = Constant (class_token, name) let make_locative location_token class_token name i vs = Locative (location_token, class_token, name, i, vs) let vstride_of_locative = function | Locative (_, _, _, _, vs) -> vs | _ -> failwith "vstride_of_locative" (* special naming conventions for variables *) let rec base62_of_int k = let x = k mod 62 and y = k / 62 in let c = if x < 10 then Char.chr (x + Char.code '0') else if x < 36 then Char.chr (x + Char.code 'a' - 10) else Char.chr (x + Char.code 'A' - 36) in let s = String.make 1 c in let r = if y == 0 then "" else base62_of_int y in r ^ s let varname_of_int k = if !Magic.compact then base62_of_int k else string_of_int k let unparse = function | Temporary k -> "T" ^ (varname_of_int k) | Constant (_, name) -> name | Locative (_, _, name, i, _) -> name i let unparse_for_alignment m = function | Locative (_, _, name, i, _) -> name (i mod m) | _ -> failwith "unparse_for_alignment" fftw-3.3.3/genfft/algsimp.mli0000644000175000017500000000167512054410050013017 00000000000000(* * Copyright (c) 1997-1999 Massachusetts Institute of Technology * Copyright (c) 2003, 2007-11 Matteo Frigo * Copyright (c) 2003, 2007-11 Massachusetts Institute of Technology * * This program is free software; you can redistribute it and/or modify * it under the terms of the GNU General Public License as published by * the Free Software Foundation; either version 2 of the License, or * (at your option) any later version. * * This program is distributed in the hope that it will be useful, * but WITHOUT ANY WARRANTY; without even the implied warranty of * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the * GNU General Public License for more details. * * You should have received a copy of the GNU General Public License * along with this program; if not, write to the Free Software * Foundation, Inc., 51 Franklin Street, Fifth Floor, Boston, MA 02110-1301 USA * *) val algsimp : Expr.expr list -> Expr.expr list fftw-3.3.3/genfft/annotate.mli0000644000175000017500000000241012054410050013160 00000000000000(* * Copyright (c) 1997-1999 Massachusetts Institute of Technology * Copyright (c) 2003, 2007-11 Matteo Frigo * Copyright (c) 2003, 2007-11 Massachusetts Institute of Technology * * This program is free software; you can redistribute it and/or modify * it under the terms of the GNU General Public License as published by * the Free Software Foundation; either version 2 of the License, or * (at your option) any later version. * * This program is distributed in the hope that it will be useful, * but WITHOUT ANY WARRANTY; without even the implied warranty of * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the * GNU General Public License for more details. * * You should have received a copy of the GNU General Public License * along with this program; if not, write to the Free Software * Foundation, Inc., 51 Franklin Street, Fifth Floor, Boston, MA 02110-1301 USA * *) open Variable open Expr type annotated_schedule = Annotate of variable list * variable list * variable list * int * aschedule and aschedule = ADone | AInstr of assignment | ASeq of (annotated_schedule * annotated_schedule) val annotate : variable list list -> Schedule.schedule -> annotated_schedule val dump : (string -> unit) -> annotated_schedule -> unit fftw-3.3.3/genfft/assoctable.mli0000644000175000017500000000225112054410050013472 00000000000000(* * Copyright (c) 1997-1999 Massachusetts Institute of Technology * Copyright (c) 2003, 2007-11 Matteo Frigo * Copyright (c) 2003, 2007-11 Massachusetts Institute of Technology * * This program is free software; you can redistribute it and/or modify * it under the terms of the GNU General Public License as published by * the Free Software Foundation; either version 2 of the License, or * (at your option) any later version. * * This program is distributed in the hope that it will be useful, * but WITHOUT ANY WARRANTY; without even the implied warranty of * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the * GNU General Public License for more details. * * You should have received a copy of the GNU General Public License * along with this program; if not, write to the Free Software * Foundation, Inc., 51 Franklin Street, Fifth Floor, Boston, MA 02110-1301 USA * *) type ('a, 'b) elem = | Leaf | Node of int * ('a, 'b) elem * ('a, 'b) elem * ('a * 'b) list val empty : ('a, 'b) elem val lookup : ('a -> int) -> ('a -> 'b -> bool) -> 'a -> ('b, 'c) elem -> 'c option val insert : ('a -> int) -> 'a -> 'c -> ('a, 'c) elem -> ('a, 'c) elem fftw-3.3.3/genfft/c.mli0000644000175000017500000000457112054410050011603 00000000000000(* * Copyright (c) 1997-1999 Massachusetts Institute of Technology * Copyright (c) 2003, 2007-11 Matteo Frigo * Copyright (c) 2003, 2007-11 Massachusetts Institute of Technology * * This program is free software; you can redistribute it and/or modify * it under the terms of the GNU General Public License as published by * the Free Software Foundation; either version 2 of the License, or * (at your option) any later version. * * This program is distributed in the hope that it will be useful, * but WITHOUT ANY WARRANTY; without even the implied warranty of * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the * GNU General Public License for more details. * * You should have received a copy of the GNU General Public License * along with this program; if not, write to the Free Software * Foundation, Inc., 51 Franklin Street, Fifth Floor, Boston, MA 02110-1301 USA * *) type stride = | SVar of string | SConst of string | SInteger of int | SNeg of stride val array_subscript : string -> stride -> int -> string val varray_subscript : string -> stride -> stride -> int -> int -> string val real_of : string -> string val imag_of : string -> string val realtype : string val realtypep : string val constrealtype : string val constrealtypep : string val stridetype : string type c_decl = | Decl of string * string | Tdecl of string (* arbitrary text declaration *) and c_ast = | Asch of Annotate.annotated_schedule | Simd_leavefun | Return of c_ast | For of c_ast * c_ast * c_ast * c_ast | If of c_ast * c_ast | Block of (c_decl list) * (c_ast list) | Binop of string * c_ast * c_ast | Expr_assign of c_ast * c_ast | Stmt_assign of c_ast * c_ast | Comma of c_ast * c_ast | Integer of int | CVar of string | CCall of string * c_ast | CPlus of c_ast list | ITimes of c_ast * c_ast | CUminus of c_ast and c_fcn = | Fcn of string * string * c_decl list * c_ast val unparse_expr : Expr.expr -> string val unparse_assignment : Expr.assignment -> string val unparse_annotated : bool -> Annotate.annotated_schedule -> string val unparse_decl : c_decl -> string val unparse_ast : c_ast -> string val unparse_function : c_fcn -> string val flops_of : c_fcn -> string val print_cost : c_fcn -> string val ast_to_expr_list : c_ast -> Expr.expr list val extract_constants : c_ast -> c_decl list val ctimes : (c_ast * c_ast) -> c_ast fftw-3.3.3/genfft/complex.mli0000644000175000017500000000444212054410050013025 00000000000000(* * Copyright (c) 1997-1999 Massachusetts Institute of Technology * Copyright (c) 2003, 2007-11 Matteo Frigo * Copyright (c) 2003, 2007-11 Massachusetts Institute of Technology * * This program is free software; you can redistribute it and/or modify * it under the terms of the GNU General Public License as published by * the Free Software Foundation; either version 2 of the License, or * (at your option) any later version. * * This program is distributed in the hope that it will be useful, * but WITHOUT ANY WARRANTY; without even the implied warranty of * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the * GNU General Public License for more details. * * You should have received a copy of the GNU General Public License * along with this program; if not, write to the Free Software * Foundation, Inc., 51 Franklin Street, Fifth Floor, Boston, MA 02110-1301 USA * *) type expr val make : (Expr.expr * Expr.expr) -> expr val two : expr val one : expr val i : expr val zero : expr val half : expr val inverse_int : int -> expr val inverse_int_sqrt : int -> expr val int_sqrt : int -> expr val times : expr -> expr -> expr val ctimes : expr -> expr -> expr val ctimesj : expr -> expr -> expr val uminus : expr -> expr val exp : int -> int -> expr val sec : int -> int -> expr val csc : int -> int -> expr val tan : int -> int -> expr val cot : int -> int -> expr val plus : expr list -> expr val real : expr -> expr val imag : expr -> expr val conj : expr -> expr val nan : Expr.transcendent -> expr val sigma : int -> int -> (int -> expr) -> expr val (@*) : expr -> expr -> expr val (@+) : expr -> expr -> expr val (@-) : expr -> expr -> expr (* a signal is a map from integers to expressions *) type signal = int -> expr val infinite : int -> signal -> signal val store_real : Variable.variable -> expr -> Expr.expr val store_imag : Variable.variable -> expr -> Expr.expr val store : Variable.variable * Variable.variable -> expr -> Expr.expr * Expr.expr val assign_real : Variable.variable -> expr -> Expr.assignment val assign_imag : Variable.variable -> expr -> Expr.assignment val assign : Variable.variable * Variable.variable -> expr -> Expr.assignment * Expr.assignment val hermitian : int -> (int -> expr) -> int -> expr val antihermitian : int -> (int -> expr) -> int -> expr fftw-3.3.3/genfft/conv.mli0000644000175000017500000000173212054410050012322 00000000000000(* * Copyright (c) 1997-1999 Massachusetts Institute of Technology * Copyright (c) 2003, 2007-11 Matteo Frigo * Copyright (c) 2003, 2007-11 Massachusetts Institute of Technology * * This program is free software; you can redistribute it and/or modify * it under the terms of the GNU General Public License as published by * the Free Software Foundation; either version 2 of the License, or * (at your option) any later version. * * This program is distributed in the hope that it will be useful, * but WITHOUT ANY WARRANTY; without even the implied warranty of * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the * GNU General Public License for more details. * * You should have received a copy of the GNU General Public License * along with this program; if not, write to the Free Software * Foundation, Inc., 51 Franklin Street, Fifth Floor, Boston, MA 02110-1301 USA * *) val conv : int -> Complex.signal -> int -> Complex.signal -> Complex.signal fftw-3.3.3/genfft/dag.mli0000644000175000017500000000277512054410050012120 00000000000000(* * Copyright (c) 1997-1999 Massachusetts Institute of Technology * Copyright (c) 2003, 2007-11 Matteo Frigo * Copyright (c) 2003, 2007-11 Massachusetts Institute of Technology * * This program is free software; you can redistribute it and/or modify * it under the terms of the GNU General Public License as published by * the Free Software Foundation; either version 2 of the License, or * (at your option) any later version. * * This program is distributed in the hope that it will be useful, * but WITHOUT ANY WARRANTY; without even the implied warranty of * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the * GNU General Public License for more details. * * You should have received a copy of the GNU General Public License * along with this program; if not, write to the Free Software * Foundation, Inc., 51 Franklin Street, Fifth Floor, Boston, MA 02110-1301 USA * *) open Util type color = | RED | BLUE | BLACK | YELLOW type dagnode = { assigned: Variable.variable; mutable expression: Expr.expr; input_variables: Variable.variable list; mutable successors: dagnode list; mutable predecessors: dagnode list; mutable label: int; mutable color: color} type dag val makedag : (Variable.variable * Expr.expr) list -> dag val map : (dagnode -> dagnode) -> dag -> dag val for_all : dag -> (dagnode -> unit) -> unit val to_list : dag -> (dagnode list) val bfs : dag -> dagnode -> int -> unit val find_node : (dagnode -> bool) -> dag -> dagnode option fftw-3.3.3/genfft/expr.mli0000644000175000017500000000337112054410050012334 00000000000000(* * Copyright (c) 1997-1999 Massachusetts Institute of Technology * Copyright (c) 2003, 2007-11 Matteo Frigo * Copyright (c) 2003, 2007-11 Massachusetts Institute of Technology * * This program is free software; you can redistribute it and/or modify * it under the terms of the GNU General Public License as published by * the Free Software Foundation; either version 2 of the License, or * (at your option) any later version. * * This program is distributed in the hope that it will be useful, * but WITHOUT ANY WARRANTY; without even the implied warranty of * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the * GNU General Public License for more details. * * You should have received a copy of the GNU General Public License * along with this program; if not, write to the Free Software * Foundation, Inc., 51 Franklin Street, Fifth Floor, Boston, MA 02110-1301 USA * *) type transcendent = I | MULTI_A | MULTI_B | CONJ type expr = | Num of Number.number | NaN of transcendent | Plus of expr list | Times of expr * expr | CTimes of expr * expr | CTimesJ of expr * expr | Uminus of expr | Load of Variable.variable | Store of Variable.variable * expr type assignment = Assign of Variable.variable * expr val hash_float : float -> int val hash : expr -> int val to_string : expr -> string val assignment_to_string : assignment -> string val transcendent_to_float : transcendent -> float val string_of_transcendent : transcendent -> string val find_vars : expr -> Variable.variable list val is_constant : expr -> bool val is_known_constant : expr -> bool val dump : (string -> unit) -> assignment list -> unit val expr_to_constants : expr -> Number.number list val unique_constants : Number.number list -> Number.number list fftw-3.3.3/genfft/fft.mli0000644000175000017500000000170712054410050012136 00000000000000(* * Copyright (c) 1997-1999 Massachusetts Institute of Technology * Copyright (c) 2003, 2007-11 Matteo Frigo * Copyright (c) 2003, 2007-11 Massachusetts Institute of Technology * * This program is free software; you can redistribute it and/or modify * it under the terms of the GNU General Public License as published by * the Free Software Foundation; either version 2 of the License, or * (at your option) any later version. * * This program is distributed in the hope that it will be useful, * but WITHOUT ANY WARRANTY; without even the implied warranty of * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the * GNU General Public License for more details. * * You should have received a copy of the GNU General Public License * along with this program; if not, write to the Free Software * Foundation, Inc., 51 Franklin Street, Fifth Floor, Boston, MA 02110-1301 USA * *) val dft : int -> int -> Complex.signal -> Complex.signal fftw-3.3.3/genfft/littlesimp.mli0000644000175000017500000000207512054410050013544 00000000000000(* * Copyright (c) 1997-1999 Massachusetts Institute of Technology * Copyright (c) 2003, 2007-11 Matteo Frigo * Copyright (c) 2003, 2007-11 Massachusetts Institute of Technology * * This program is free software; you can redistribute it and/or modify * it under the terms of the GNU General Public License as published by * the Free Software Foundation; either version 2 of the License, or * (at your option) any later version. * * This program is distributed in the hope that it will be useful, * but WITHOUT ANY WARRANTY; without even the implied warranty of * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the * GNU General Public License for more details. * * You should have received a copy of the GNU General Public License * along with this program; if not, write to the Free Software * Foundation, Inc., 51 Franklin Street, Fifth Floor, Boston, MA 02110-1301 USA * *) val makeNum : Number.number -> Expr.expr val makeUminus : Expr.expr -> Expr.expr val makeTimes : Expr.expr * Expr.expr -> Expr.expr val makePlus : Expr.expr list -> Expr.expr fftw-3.3.3/genfft/number.mli0000644000175000017500000000316212054410050012644 00000000000000(* * Copyright (c) 1997-1999 Massachusetts Institute of Technology * Copyright (c) 2003, 2007-11 Matteo Frigo * Copyright (c) 2003, 2007-11 Massachusetts Institute of Technology * * This program is free software; you can redistribute it and/or modify * it under the terms of the GNU General Public License as published by * the Free Software Foundation; either version 2 of the License, or * (at your option) any later version. * * This program is distributed in the hope that it will be useful, * but WITHOUT ANY WARRANTY; without even the implied warranty of * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the * GNU General Public License for more details. * * You should have received a copy of the GNU General Public License * along with this program; if not, write to the Free Software * Foundation, Inc., 51 Franklin Street, Fifth Floor, Boston, MA 02110-1301 USA * *) type number val equal : number -> number -> bool val of_int : int -> number val zero : number val one : number val two : number val mone : number val is_zero : number -> bool val is_one : number -> bool val is_mone : number -> bool val is_two : number -> bool val mul : number -> number -> number val div : number -> number -> number val add : number -> number -> number val sub : number -> number -> number val negative : number -> bool val greater : number -> number -> bool val negate : number -> number val sqrt : number -> number (* cexp n i = (cos (2 * pi * i / n), sin (2 * pi * i / n)) *) val cexp : int -> int -> (number * number) val to_konst : number -> string val to_string : number -> string val to_float : number -> float fftw-3.3.3/genfft/oracle.mli0000644000175000017500000000200512054410050012614 00000000000000(* * Copyright (c) 1997-1999 Massachusetts Institute of Technology * Copyright (c) 2003, 2007-11 Matteo Frigo * Copyright (c) 2003, 2007-11 Massachusetts Institute of Technology * * This program is free software; you can redistribute it and/or modify * it under the terms of the GNU General Public License as published by * the Free Software Foundation; either version 2 of the License, or * (at your option) any later version. * * This program is distributed in the hope that it will be useful, * but WITHOUT ANY WARRANTY; without even the implied warranty of * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the * GNU General Public License for more details. * * You should have received a copy of the GNU General Public License * along with this program; if not, write to the Free Software * Foundation, Inc., 51 Franklin Street, Fifth Floor, Boston, MA 02110-1301 USA * *) val should_flip_sign : Expr.expr -> bool val likely_equal : Expr.expr -> Expr.expr -> bool val hash : Expr.expr -> int fftw-3.3.3/genfft/schedule.mli0000644000175000017500000000225712054410050013154 00000000000000(* * Copyright (c) 1997-1999 Massachusetts Institute of Technology * Copyright (c) 2003, 2007-11 Matteo Frigo * Copyright (c) 2003, 2007-11 Massachusetts Institute of Technology * * This program is free software; you can redistribute it and/or modify * it under the terms of the GNU General Public License as published by * the Free Software Foundation; either version 2 of the License, or * (at your option) any later version. * * This program is distributed in the hope that it will be useful, * but WITHOUT ANY WARRANTY; without even the implied warranty of * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the * GNU General Public License for more details. * * You should have received a copy of the GNU General Public License * along with this program; if not, write to the Free Software * Foundation, Inc., 51 Franklin Street, Fifth Floor, Boston, MA 02110-1301 USA * *) type schedule = | Done | Instr of Expr.assignment | Seq of (schedule * schedule) | Par of schedule list val schedule : Expr.assignment list -> schedule val sequentially : Expr.assignment list -> schedule val isolate_precomputations_and_schedule : Expr.assignment list -> schedule fftw-3.3.3/genfft/simd.mli0000644000175000017500000000211512054410050012305 00000000000000(* * Copyright (c) 1997-1999 Massachusetts Institute of Technology * Copyright (c) 2003, 2007-11 Matteo Frigo * Copyright (c) 2003, 2007-11 Massachusetts Institute of Technology * * This program is free software; you can redistribute it and/or modify * it under the terms of the GNU General Public License as published by * the Free Software Foundation; either version 2 of the License, or * (at your option) any later version. * * This program is distributed in the hope that it will be useful, * but WITHOUT ANY WARRANTY; without even the implied warranty of * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the * GNU General Public License for more details. * * You should have received a copy of the GNU General Public License * along with this program; if not, write to the Free Software * Foundation, Inc., 51 Franklin Street, Fifth Floor, Boston, MA 02110-1301 USA * *) val unparse_function : C.c_fcn -> string val extract_constants : C.c_ast -> C.c_decl list val realtype : string val realtypep : string val constrealtype : string val constrealtypep : string fftw-3.3.3/genfft/to_alist.mli0000644000175000017500000000206712054410050013175 00000000000000(* * Copyright (c) 1997-1999 Massachusetts Institute of Technology * Copyright (c) 2003, 2007-11 Matteo Frigo * Copyright (c) 2003, 2007-11 Massachusetts Institute of Technology * * This program is free software; you can redistribute it and/or modify * it under the terms of the GNU General Public License as published by * the Free Software Foundation; either version 2 of the License, or * (at your option) any later version. * * This program is distributed in the hope that it will be useful, * but WITHOUT ANY WARRANTY; without even the implied warranty of * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the * GNU General Public License for more details. * * You should have received a copy of the GNU General Public License * along with this program; if not, write to the Free Software * Foundation, Inc., 51 Franklin Street, Fifth Floor, Boston, MA 02110-1301 USA * *) val to_assignments : Expr.expr list -> Expr.assignment list val dump : (string -> unit) -> Expr.assignment list -> unit val good_for_fma : Expr.expr * Expr.expr -> bool fftw-3.3.3/genfft/trig.mli0000644000175000017500000000303712054410050012322 00000000000000(* * Copyright (c) 1997-1999 Massachusetts Institute of Technology * Copyright (c) 2003, 2007-11 Matteo Frigo * Copyright (c) 2003, 2007-11 Massachusetts Institute of Technology * * This program is free software; you can redistribute it and/or modify * it under the terms of the GNU General Public License as published by * the Free Software Foundation; either version 2 of the License, or * (at your option) any later version. * * This program is distributed in the hope that it will be useful, * but WITHOUT ANY WARRANTY; without even the implied warranty of * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the * GNU General Public License for more details. * * You should have received a copy of the GNU General Public License * along with this program; if not, write to the Free Software * Foundation, Inc., 51 Franklin Street, Fifth Floor, Boston, MA 02110-1301 USA * *) val rdft : int -> int -> Complex.signal -> Complex.signal val hdft : int -> int -> Complex.signal -> Complex.signal val dft_via_rdft : int -> int -> Complex.signal -> Complex.signal val dht : int -> int -> Complex.signal -> Complex.signal val dctI : int -> Complex.signal -> Complex.signal val dctII : int -> Complex.signal -> Complex.signal val dctIII : int -> Complex.signal -> Complex.signal val dctIV : int -> Complex.signal -> Complex.signal val dstI : int -> Complex.signal -> Complex.signal val dstII : int -> Complex.signal -> Complex.signal val dstIII : int -> Complex.signal -> Complex.signal val dstIV : int -> Complex.signal -> Complex.signal fftw-3.3.3/genfft/twiddle.mli0000644000175000017500000000234712054410050013014 00000000000000(* * Copyright (c) 1997-1999 Massachusetts Institute of Technology * Copyright (c) 2003, 2007-11 Matteo Frigo * Copyright (c) 2003, 2007-11 Massachusetts Institute of Technology * * This program is free software; you can redistribute it and/or modify * it under the terms of the GNU General Public License as published by * the Free Software Foundation; either version 2 of the License, or * (at your option) any later version. * * This program is distributed in the hope that it will be useful, * but WITHOUT ANY WARRANTY; without even the implied warranty of * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the * GNU General Public License for more details. * * You should have received a copy of the GNU General Public License * along with this program; if not, write to the Free Software * Foundation, Inc., 51 Franklin Street, Fifth Floor, Boston, MA 02110-1301 USA * *) val speclist : (string * Arg.spec * string) list type twinstr val twiddle_policy : int -> bool -> (int -> int -> (int -> Complex.expr) -> (int -> Complex.expr) -> int -> Complex.expr) *(int -> int) * (int -> twinstr list) val twinstr_to_c_string : twinstr list -> string val twinstr_to_simd_string : string -> twinstr list -> string fftw-3.3.3/genfft/unique.mli0000644000175000017500000000173012054410050012661 00000000000000(* * Copyright (c) 1997-1999 Massachusetts Institute of Technology * Copyright (c) 2003, 2007-11 Matteo Frigo * Copyright (c) 2003, 2007-11 Massachusetts Institute of Technology * * This program is free software; you can redistribute it and/or modify * it under the terms of the GNU General Public License as published by * the Free Software Foundation; either version 2 of the License, or * (at your option) any later version. * * This program is distributed in the hope that it will be useful, * but WITHOUT ANY WARRANTY; without even the implied warranty of * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the * GNU General Public License for more details. * * You should have received a copy of the GNU General Public License * along with this program; if not, write to the Free Software * Foundation, Inc., 51 Franklin Street, Fifth Floor, Boston, MA 02110-1301 USA * *) type unique val make : unit -> unique val same : unique -> unique -> bool fftw-3.3.3/genfft/util.mli0000644000175000017500000000372412054410050012335 00000000000000(* * Copyright (c) 1997-1999 Massachusetts Institute of Technology * Copyright (c) 2003, 2007-11 Matteo Frigo * Copyright (c) 2003, 2007-11 Massachusetts Institute of Technology * * This program is free software; you can redistribute it and/or modify * it under the terms of the GNU General Public License as published by * the Free Software Foundation; either version 2 of the License, or * (at your option) any later version. * * This program is distributed in the hope that it will be useful, * but WITHOUT ANY WARRANTY; without even the implied warranty of * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the * GNU General Public License for more details. * * You should have received a copy of the GNU General Public License * along with this program; if not, write to the Free Software * Foundation, Inc., 51 Franklin Street, Fifth Floor, Boston, MA 02110-1301 USA * *) val invmod : int -> int -> int val gcd : int -> int -> int val lowest_terms : int -> int -> int * int val find_generator : int -> int val pow_mod : int -> int -> int -> int val forall : 'a -> ('b -> 'a -> 'a) -> int -> int -> (int -> 'b) -> 'a val sum_list : int list -> int val max_list : int list -> int val min_list : int list -> int val count : ('a -> bool) -> 'a list -> int val remove : 'a -> 'a list -> 'a list val for_list : 'a list -> ('a -> unit) -> unit val rmap : 'a list -> ('a -> 'b) -> 'b list val cons : 'a -> 'a list -> 'a list val null : 'a list -> bool val (@@) : ('a -> 'b) -> ('c -> 'a) -> 'c -> 'b val forall_flat : int -> int -> (int -> 'a list) -> 'a list val identity : 'a -> 'a val minimize : ('a -> 'b) -> 'a list -> 'a option val find_elem : ('a -> bool) -> 'a list -> 'a option val suchthat : int -> (int -> bool) -> int val info : string -> unit val iota : int -> int list val interval : int -> int -> int list val array : int -> (int -> 'a) -> int -> 'a val take : int -> 'a list -> 'a list val drop : int -> 'a list -> 'a list val either : 'a option -> 'a -> 'a fftw-3.3.3/genfft/variable.mli0000644000175000017500000000276312054410050013147 00000000000000(* * Copyright (c) 1997-1999 Massachusetts Institute of Technology * Copyright (c) 2003, 2007-11 Matteo Frigo * Copyright (c) 2003, 2007-11 Massachusetts Institute of Technology * * This program is free software; you can redistribute it and/or modify * it under the terms of the GNU General Public License as published by * the Free Software Foundation; either version 2 of the License, or * (at your option) any later version. * * This program is distributed in the hope that it will be useful, * but WITHOUT ANY WARRANTY; without even the implied warranty of * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the * GNU General Public License for more details. * * You should have received a copy of the GNU General Public License * along with this program; if not, write to the Free Software * Foundation, Inc., 51 Franklin Street, Fifth Floor, Boston, MA 02110-1301 USA * *) type variable val hash : variable -> int val same : variable -> variable -> bool val is_constant : variable -> bool val is_temporary : variable -> bool val is_locative : variable -> bool val same_location : variable -> variable -> bool val same_class : variable -> variable -> bool val make_temporary : unit -> variable val make_constant : Unique.unique -> string -> variable val make_locative : Unique.unique -> Unique.unique -> (int -> string) -> int -> string -> variable val unparse : variable -> string val unparse_for_alignment : int -> variable -> string val vstride_of_locative : variable -> string fftw-3.3.3/kernel/0000755000175000017500000000000012054411106010741 500000000000000fftw-3.3.3/kernel/Makefile.am0000644000175000017500000000112512054410050012711 00000000000000AM_CPPFLAGS = -I$(top_srcdir)/simd noinst_LTLIBRARIES = libkernel.la # pkgincludedir = $(includedir)/fftw3@PREC_SUFFIX@ # pkginclude_HEADERS = ifftw.h cycle.h libkernel_la_SOURCES = align.c alloc.c assert.c awake.c buffered.c \ cpy1d.c cpy2d-pair.c cpy2d.c ct.c debug.c extract-reim.c hash.c iabs.c \ kalloc.c md5-1.c md5.c minmax.c ops.c pickdim.c plan.c planner.c \ primes.c print.c problem.c rader.c scan.c solver.c solvtab.c stride.c \ tensor.c tensor1.c tensor2.c tensor3.c tensor4.c tensor5.c tensor7.c \ tensor8.c tensor9.c tile2d.c timer.c transpose.c trig.c twiddle.c \ cycle.h ifftw.h fftw-3.3.3/kernel/Makefile.in0000644000175000017500000004702012054410200012723 00000000000000# Makefile.in generated by automake 1.11.6 from Makefile.am. # @configure_input@ # Copyright (C) 1994, 1995, 1996, 1997, 1998, 1999, 2000, 2001, 2002, # 2003, 2004, 2005, 2006, 2007, 2008, 2009, 2010, 2011 Free Software # Foundation, Inc. # This Makefile.in is free software; the Free Software Foundation # gives unlimited permission to copy and/or distribute it, # with or without modifications, as long as this notice is preserved. # This program is distributed in the hope that it will be useful, # but WITHOUT ANY WARRANTY, to the extent permitted by law; without # even the implied warranty of MERCHANTABILITY or FITNESS FOR A # PARTICULAR PURPOSE. @SET_MAKE@ VPATH = @srcdir@ am__make_dryrun = \ { \ am__dry=no; \ case $$MAKEFLAGS in \ *\\[\ \ ]*) \ echo 'am--echo: ; @echo "AM" OK' | $(MAKE) -f - 2>/dev/null \ | grep '^AM OK$$' >/dev/null || am__dry=yes;; \ *) \ for am__flg in $$MAKEFLAGS; do \ case $$am__flg in \ *=*|--*) ;; \ *n*) am__dry=yes; break;; \ esac; \ done;; \ esac; \ test $$am__dry = yes; \ } pkgdatadir = $(datadir)/@PACKAGE@ pkgincludedir = $(includedir)/@PACKAGE@ pkglibdir = $(libdir)/@PACKAGE@ pkglibexecdir = $(libexecdir)/@PACKAGE@ am__cd = CDPATH="$${ZSH_VERSION+.}$(PATH_SEPARATOR)" && cd install_sh_DATA = $(install_sh) -c -m 644 install_sh_PROGRAM = $(install_sh) -c install_sh_SCRIPT = $(install_sh) -c INSTALL_HEADER = $(INSTALL_DATA) transform = $(program_transform_name) NORMAL_INSTALL = : PRE_INSTALL = : POST_INSTALL = : NORMAL_UNINSTALL = : PRE_UNINSTALL = : POST_UNINSTALL = : build_triplet = @build@ host_triplet = @host@ subdir = kernel DIST_COMMON = $(srcdir)/Makefile.am $(srcdir)/Makefile.in ACLOCAL_M4 = $(top_srcdir)/aclocal.m4 am__aclocal_m4_deps = $(top_srcdir)/m4/acx_mpi.m4 \ $(top_srcdir)/m4/acx_pthread.m4 \ $(top_srcdir)/m4/ax_cc_maxopt.m4 \ $(top_srcdir)/m4/ax_check_compiler_flags.m4 \ $(top_srcdir)/m4/ax_compiler_vendor.m4 \ $(top_srcdir)/m4/ax_gcc_aligns_stack.m4 \ $(top_srcdir)/m4/ax_gcc_version.m4 \ $(top_srcdir)/m4/ax_openmp.m4 $(top_srcdir)/m4/libtool.m4 \ $(top_srcdir)/m4/ltoptions.m4 $(top_srcdir)/m4/ltsugar.m4 \ $(top_srcdir)/m4/ltversion.m4 $(top_srcdir)/m4/lt~obsolete.m4 \ $(top_srcdir)/configure.ac am__configure_deps = $(am__aclocal_m4_deps) $(CONFIGURE_DEPENDENCIES) \ $(ACLOCAL_M4) mkinstalldirs = $(install_sh) -d CONFIG_HEADER = $(top_builddir)/config.h CONFIG_CLEAN_FILES = CONFIG_CLEAN_VPATH_FILES = LTLIBRARIES = $(noinst_LTLIBRARIES) libkernel_la_LIBADD = am_libkernel_la_OBJECTS = align.lo alloc.lo assert.lo awake.lo \ buffered.lo cpy1d.lo cpy2d-pair.lo cpy2d.lo ct.lo debug.lo \ extract-reim.lo hash.lo iabs.lo kalloc.lo md5-1.lo md5.lo \ minmax.lo ops.lo pickdim.lo plan.lo planner.lo primes.lo \ print.lo problem.lo rader.lo scan.lo solver.lo solvtab.lo \ stride.lo tensor.lo tensor1.lo tensor2.lo tensor3.lo \ tensor4.lo tensor5.lo tensor7.lo tensor8.lo tensor9.lo \ tile2d.lo timer.lo transpose.lo trig.lo twiddle.lo libkernel_la_OBJECTS = $(am_libkernel_la_OBJECTS) DEFAULT_INCLUDES = -I.@am__isrc@ -I$(top_builddir) depcomp = $(SHELL) $(top_srcdir)/depcomp am__depfiles_maybe = depfiles am__mv = mv -f COMPILE = $(CC) $(DEFS) $(DEFAULT_INCLUDES) $(INCLUDES) $(AM_CPPFLAGS) \ $(CPPFLAGS) $(AM_CFLAGS) $(CFLAGS) LTCOMPILE = $(LIBTOOL) --tag=CC $(AM_LIBTOOLFLAGS) $(LIBTOOLFLAGS) \ --mode=compile $(CC) $(DEFS) $(DEFAULT_INCLUDES) $(INCLUDES) \ $(AM_CPPFLAGS) $(CPPFLAGS) $(AM_CFLAGS) $(CFLAGS) CCLD = $(CC) LINK = $(LIBTOOL) --tag=CC $(AM_LIBTOOLFLAGS) $(LIBTOOLFLAGS) \ --mode=link $(CCLD) $(AM_CFLAGS) $(CFLAGS) $(AM_LDFLAGS) \ $(LDFLAGS) -o $@ SOURCES = $(libkernel_la_SOURCES) DIST_SOURCES = $(libkernel_la_SOURCES) am__can_run_installinfo = \ case $$AM_UPDATE_INFO_DIR in \ n|no|NO) false;; \ *) (install-info --version) >/dev/null 2>&1;; \ esac ETAGS = etags CTAGS = ctags DISTFILES = $(DIST_COMMON) $(DIST_SOURCES) $(TEXINFOS) $(EXTRA_DIST) ACLOCAL = @ACLOCAL@ ALLOCA = @ALLOCA@ ALTIVEC_CFLAGS = @ALTIVEC_CFLAGS@ AMTAR = @AMTAR@ AR = @AR@ AS = @AS@ AUTOCONF = @AUTOCONF@ AUTOHEADER = @AUTOHEADER@ AUTOMAKE = @AUTOMAKE@ AVX_CFLAGS = @AVX_CFLAGS@ AWK = @AWK@ CC = @CC@ CCDEPMODE = @CCDEPMODE@ CFLAGS = @CFLAGS@ CHECK_PL_OPTS = @CHECK_PL_OPTS@ CPP = @CPP@ CPPFLAGS = @CPPFLAGS@ CYGPATH_W = @CYGPATH_W@ C_FFTW_R2R_KIND = @C_FFTW_R2R_KIND@ C_MPI_FINT = @C_MPI_FINT@ DEFS = @DEFS@ DEPDIR = @DEPDIR@ DLLTOOL = @DLLTOOL@ DSYMUTIL = @DSYMUTIL@ DUMPBIN = @DUMPBIN@ ECHO_C = @ECHO_C@ ECHO_N = @ECHO_N@ ECHO_T = @ECHO_T@ EGREP = @EGREP@ EXEEXT = @EXEEXT@ F77 = @F77@ FFLAGS = @FFLAGS@ FGREP = @FGREP@ FLIBS = @FLIBS@ GREP = @GREP@ INSTALL = @INSTALL@ INSTALL_DATA = @INSTALL_DATA@ INSTALL_PROGRAM = @INSTALL_PROGRAM@ INSTALL_SCRIPT = @INSTALL_SCRIPT@ INSTALL_STRIP_PROGRAM = @INSTALL_STRIP_PROGRAM@ LD = @LD@ LDFLAGS = @LDFLAGS@ LIBOBJS = @LIBOBJS@ LIBQUADMATH = @LIBQUADMATH@ LIBS = @LIBS@ LIBTOOL = @LIBTOOL@ LIPO = @LIPO@ LN_S = @LN_S@ LTLIBOBJS = @LTLIBOBJS@ MAINT = @MAINT@ MAKEINFO = @MAKEINFO@ MANIFEST_TOOL = @MANIFEST_TOOL@ MKDIR_P = @MKDIR_P@ MPICC = @MPICC@ MPILIBS = @MPILIBS@ MPIRUN = @MPIRUN@ NEON_CFLAGS = @NEON_CFLAGS@ NM = @NM@ NMEDIT = @NMEDIT@ OBJDUMP = @OBJDUMP@ OBJEXT = @OBJEXT@ OCAMLBUILD = @OCAMLBUILD@ OPENMP_CFLAGS = @OPENMP_CFLAGS@ OTOOL = @OTOOL@ OTOOL64 = @OTOOL64@ PACKAGE = @PACKAGE@ PACKAGE_BUGREPORT = @PACKAGE_BUGREPORT@ PACKAGE_NAME = @PACKAGE_NAME@ PACKAGE_STRING = @PACKAGE_STRING@ PACKAGE_TARNAME = @PACKAGE_TARNAME@ PACKAGE_URL = @PACKAGE_URL@ PACKAGE_VERSION = @PACKAGE_VERSION@ PATH_SEPARATOR = @PATH_SEPARATOR@ POW_LIB = @POW_LIB@ PRECISION = @PRECISION@ PREC_SUFFIX = @PREC_SUFFIX@ PTHREAD_CC = @PTHREAD_CC@ PTHREAD_CFLAGS = @PTHREAD_CFLAGS@ PTHREAD_LIBS = @PTHREAD_LIBS@ RANLIB = @RANLIB@ SED = @SED@ SET_MAKE = @SET_MAKE@ SHARED_VERSION_INFO = @SHARED_VERSION_INFO@ SHELL = @SHELL@ SSE2_CFLAGS = @SSE2_CFLAGS@ STACK_ALIGN_CFLAGS = @STACK_ALIGN_CFLAGS@ STRIP = @STRIP@ THREADLIBS = @THREADLIBS@ VERSION = @VERSION@ abs_builddir = @abs_builddir@ abs_srcdir = @abs_srcdir@ abs_top_builddir = @abs_top_builddir@ abs_top_srcdir = @abs_top_srcdir@ ac_ct_AR = @ac_ct_AR@ ac_ct_CC = @ac_ct_CC@ ac_ct_DUMPBIN = @ac_ct_DUMPBIN@ ac_ct_F77 = @ac_ct_F77@ acx_pthread_config = @acx_pthread_config@ am__include = @am__include@ am__leading_dot = @am__leading_dot@ am__quote = @am__quote@ am__tar = @am__tar@ am__untar = @am__untar@ bindir = @bindir@ build = @build@ build_alias = @build_alias@ build_cpu = @build_cpu@ build_os = @build_os@ build_vendor = @build_vendor@ builddir = @builddir@ datadir = @datadir@ datarootdir = @datarootdir@ docdir = @docdir@ dvidir = @dvidir@ exec_prefix = @exec_prefix@ host = @host@ host_alias = @host_alias@ host_cpu = @host_cpu@ host_os = @host_os@ host_vendor = @host_vendor@ htmldir = @htmldir@ includedir = @includedir@ infodir = @infodir@ install_sh = @install_sh@ libdir = @libdir@ libexecdir = @libexecdir@ localedir = @localedir@ localstatedir = @localstatedir@ mandir = @mandir@ mkdir_p = @mkdir_p@ oldincludedir = @oldincludedir@ pdfdir = @pdfdir@ prefix = @prefix@ program_transform_name = @program_transform_name@ psdir = @psdir@ sbindir = @sbindir@ sharedstatedir = @sharedstatedir@ srcdir = @srcdir@ sysconfdir = @sysconfdir@ target_alias = @target_alias@ top_build_prefix = @top_build_prefix@ top_builddir = @top_builddir@ top_srcdir = @top_srcdir@ AM_CPPFLAGS = -I$(top_srcdir)/simd noinst_LTLIBRARIES = libkernel.la # pkgincludedir = $(includedir)/fftw3@PREC_SUFFIX@ # pkginclude_HEADERS = ifftw.h cycle.h libkernel_la_SOURCES = align.c alloc.c assert.c awake.c buffered.c \ cpy1d.c cpy2d-pair.c cpy2d.c ct.c debug.c extract-reim.c hash.c iabs.c \ kalloc.c md5-1.c md5.c minmax.c ops.c pickdim.c plan.c planner.c \ primes.c print.c problem.c rader.c scan.c solver.c solvtab.c stride.c \ tensor.c tensor1.c tensor2.c tensor3.c tensor4.c tensor5.c tensor7.c \ tensor8.c tensor9.c tile2d.c timer.c transpose.c trig.c twiddle.c \ cycle.h ifftw.h all: all-am .SUFFIXES: .SUFFIXES: .c .lo .o .obj $(srcdir)/Makefile.in: @MAINTAINER_MODE_TRUE@ $(srcdir)/Makefile.am $(am__configure_deps) @for dep in $?; do \ case '$(am__configure_deps)' in \ *$$dep*) \ ( cd $(top_builddir) && $(MAKE) $(AM_MAKEFLAGS) am--refresh ) \ && { if test -f $@; then exit 0; else break; fi; }; \ exit 1;; \ esac; \ done; \ echo ' cd $(top_srcdir) && $(AUTOMAKE) --gnu kernel/Makefile'; \ $(am__cd) $(top_srcdir) && \ $(AUTOMAKE) --gnu kernel/Makefile .PRECIOUS: Makefile Makefile: $(srcdir)/Makefile.in $(top_builddir)/config.status @case '$?' in \ *config.status*) \ cd $(top_builddir) && $(MAKE) $(AM_MAKEFLAGS) am--refresh;; \ *) \ echo ' cd $(top_builddir) && $(SHELL) ./config.status $(subdir)/$@ $(am__depfiles_maybe)'; \ cd $(top_builddir) && $(SHELL) ./config.status $(subdir)/$@ $(am__depfiles_maybe);; \ esac; $(top_builddir)/config.status: $(top_srcdir)/configure $(CONFIG_STATUS_DEPENDENCIES) cd $(top_builddir) && $(MAKE) $(AM_MAKEFLAGS) am--refresh $(top_srcdir)/configure: @MAINTAINER_MODE_TRUE@ $(am__configure_deps) cd $(top_builddir) && $(MAKE) $(AM_MAKEFLAGS) am--refresh $(ACLOCAL_M4): @MAINTAINER_MODE_TRUE@ $(am__aclocal_m4_deps) cd $(top_builddir) && $(MAKE) $(AM_MAKEFLAGS) am--refresh $(am__aclocal_m4_deps): clean-noinstLTLIBRARIES: -test -z "$(noinst_LTLIBRARIES)" || rm -f $(noinst_LTLIBRARIES) @list='$(noinst_LTLIBRARIES)'; for p in $$list; do \ dir="`echo $$p | sed -e 's|/[^/]*$$||'`"; \ test "$$dir" != "$$p" || dir=.; \ echo "rm -f \"$${dir}/so_locations\""; \ rm -f "$${dir}/so_locations"; \ done libkernel.la: $(libkernel_la_OBJECTS) $(libkernel_la_DEPENDENCIES) $(EXTRA_libkernel_la_DEPENDENCIES) $(LINK) $(libkernel_la_OBJECTS) $(libkernel_la_LIBADD) $(LIBS) mostlyclean-compile: -rm -f *.$(OBJEXT) distclean-compile: -rm -f *.tab.c @AMDEP_TRUE@@am__include@ @am__quote@./$(DEPDIR)/align.Plo@am__quote@ @AMDEP_TRUE@@am__include@ @am__quote@./$(DEPDIR)/alloc.Plo@am__quote@ @AMDEP_TRUE@@am__include@ @am__quote@./$(DEPDIR)/assert.Plo@am__quote@ @AMDEP_TRUE@@am__include@ @am__quote@./$(DEPDIR)/awake.Plo@am__quote@ @AMDEP_TRUE@@am__include@ @am__quote@./$(DEPDIR)/buffered.Plo@am__quote@ @AMDEP_TRUE@@am__include@ @am__quote@./$(DEPDIR)/cpy1d.Plo@am__quote@ @AMDEP_TRUE@@am__include@ @am__quote@./$(DEPDIR)/cpy2d-pair.Plo@am__quote@ @AMDEP_TRUE@@am__include@ @am__quote@./$(DEPDIR)/cpy2d.Plo@am__quote@ @AMDEP_TRUE@@am__include@ @am__quote@./$(DEPDIR)/ct.Plo@am__quote@ @AMDEP_TRUE@@am__include@ @am__quote@./$(DEPDIR)/debug.Plo@am__quote@ @AMDEP_TRUE@@am__include@ @am__quote@./$(DEPDIR)/extract-reim.Plo@am__quote@ @AMDEP_TRUE@@am__include@ @am__quote@./$(DEPDIR)/hash.Plo@am__quote@ @AMDEP_TRUE@@am__include@ @am__quote@./$(DEPDIR)/iabs.Plo@am__quote@ @AMDEP_TRUE@@am__include@ @am__quote@./$(DEPDIR)/kalloc.Plo@am__quote@ @AMDEP_TRUE@@am__include@ @am__quote@./$(DEPDIR)/md5-1.Plo@am__quote@ @AMDEP_TRUE@@am__include@ @am__quote@./$(DEPDIR)/md5.Plo@am__quote@ @AMDEP_TRUE@@am__include@ @am__quote@./$(DEPDIR)/minmax.Plo@am__quote@ @AMDEP_TRUE@@am__include@ @am__quote@./$(DEPDIR)/ops.Plo@am__quote@ @AMDEP_TRUE@@am__include@ @am__quote@./$(DEPDIR)/pickdim.Plo@am__quote@ @AMDEP_TRUE@@am__include@ @am__quote@./$(DEPDIR)/plan.Plo@am__quote@ @AMDEP_TRUE@@am__include@ @am__quote@./$(DEPDIR)/planner.Plo@am__quote@ @AMDEP_TRUE@@am__include@ @am__quote@./$(DEPDIR)/primes.Plo@am__quote@ @AMDEP_TRUE@@am__include@ @am__quote@./$(DEPDIR)/print.Plo@am__quote@ @AMDEP_TRUE@@am__include@ @am__quote@./$(DEPDIR)/problem.Plo@am__quote@ @AMDEP_TRUE@@am__include@ @am__quote@./$(DEPDIR)/rader.Plo@am__quote@ @AMDEP_TRUE@@am__include@ @am__quote@./$(DEPDIR)/scan.Plo@am__quote@ @AMDEP_TRUE@@am__include@ @am__quote@./$(DEPDIR)/solver.Plo@am__quote@ @AMDEP_TRUE@@am__include@ @am__quote@./$(DEPDIR)/solvtab.Plo@am__quote@ @AMDEP_TRUE@@am__include@ @am__quote@./$(DEPDIR)/stride.Plo@am__quote@ @AMDEP_TRUE@@am__include@ @am__quote@./$(DEPDIR)/tensor.Plo@am__quote@ @AMDEP_TRUE@@am__include@ @am__quote@./$(DEPDIR)/tensor1.Plo@am__quote@ @AMDEP_TRUE@@am__include@ @am__quote@./$(DEPDIR)/tensor2.Plo@am__quote@ @AMDEP_TRUE@@am__include@ @am__quote@./$(DEPDIR)/tensor3.Plo@am__quote@ @AMDEP_TRUE@@am__include@ @am__quote@./$(DEPDIR)/tensor4.Plo@am__quote@ @AMDEP_TRUE@@am__include@ @am__quote@./$(DEPDIR)/tensor5.Plo@am__quote@ @AMDEP_TRUE@@am__include@ @am__quote@./$(DEPDIR)/tensor7.Plo@am__quote@ @AMDEP_TRUE@@am__include@ @am__quote@./$(DEPDIR)/tensor8.Plo@am__quote@ @AMDEP_TRUE@@am__include@ @am__quote@./$(DEPDIR)/tensor9.Plo@am__quote@ @AMDEP_TRUE@@am__include@ @am__quote@./$(DEPDIR)/tile2d.Plo@am__quote@ @AMDEP_TRUE@@am__include@ @am__quote@./$(DEPDIR)/timer.Plo@am__quote@ @AMDEP_TRUE@@am__include@ @am__quote@./$(DEPDIR)/transpose.Plo@am__quote@ @AMDEP_TRUE@@am__include@ @am__quote@./$(DEPDIR)/trig.Plo@am__quote@ @AMDEP_TRUE@@am__include@ @am__quote@./$(DEPDIR)/twiddle.Plo@am__quote@ .c.o: @am__fastdepCC_TRUE@ $(COMPILE) -MT $@ -MD -MP -MF $(DEPDIR)/$*.Tpo -c -o $@ $< @am__fastdepCC_TRUE@ $(am__mv) $(DEPDIR)/$*.Tpo $(DEPDIR)/$*.Po @AMDEP_TRUE@@am__fastdepCC_FALSE@ source='$<' object='$@' libtool=no @AMDEPBACKSLASH@ @AMDEP_TRUE@@am__fastdepCC_FALSE@ DEPDIR=$(DEPDIR) $(CCDEPMODE) $(depcomp) @AMDEPBACKSLASH@ @am__fastdepCC_FALSE@ $(COMPILE) -c $< .c.obj: @am__fastdepCC_TRUE@ $(COMPILE) -MT $@ -MD -MP -MF $(DEPDIR)/$*.Tpo -c -o $@ `$(CYGPATH_W) '$<'` @am__fastdepCC_TRUE@ $(am__mv) $(DEPDIR)/$*.Tpo $(DEPDIR)/$*.Po @AMDEP_TRUE@@am__fastdepCC_FALSE@ source='$<' object='$@' libtool=no @AMDEPBACKSLASH@ @AMDEP_TRUE@@am__fastdepCC_FALSE@ DEPDIR=$(DEPDIR) $(CCDEPMODE) $(depcomp) @AMDEPBACKSLASH@ @am__fastdepCC_FALSE@ $(COMPILE) -c `$(CYGPATH_W) '$<'` .c.lo: @am__fastdepCC_TRUE@ $(LTCOMPILE) -MT $@ -MD -MP -MF $(DEPDIR)/$*.Tpo -c -o $@ $< @am__fastdepCC_TRUE@ $(am__mv) $(DEPDIR)/$*.Tpo $(DEPDIR)/$*.Plo @AMDEP_TRUE@@am__fastdepCC_FALSE@ source='$<' object='$@' libtool=yes @AMDEPBACKSLASH@ @AMDEP_TRUE@@am__fastdepCC_FALSE@ DEPDIR=$(DEPDIR) $(CCDEPMODE) $(depcomp) @AMDEPBACKSLASH@ @am__fastdepCC_FALSE@ $(LTCOMPILE) -c -o $@ $< mostlyclean-libtool: -rm -f *.lo clean-libtool: -rm -rf .libs _libs ID: $(HEADERS) $(SOURCES) $(LISP) $(TAGS_FILES) list='$(SOURCES) $(HEADERS) $(LISP) $(TAGS_FILES)'; \ unique=`for i in $$list; do \ if test -f "$$i"; then echo $$i; else echo $(srcdir)/$$i; fi; \ done | \ $(AWK) '{ files[$$0] = 1; nonempty = 1; } \ END { if (nonempty) { for (i in files) print i; }; }'`; \ mkid -fID $$unique tags: TAGS TAGS: $(HEADERS) $(SOURCES) $(TAGS_DEPENDENCIES) \ $(TAGS_FILES) $(LISP) set x; \ here=`pwd`; \ list='$(SOURCES) $(HEADERS) $(LISP) $(TAGS_FILES)'; \ unique=`for i in $$list; do \ if test -f "$$i"; then echo $$i; else echo $(srcdir)/$$i; fi; \ done | \ $(AWK) '{ files[$$0] = 1; nonempty = 1; } \ END { if (nonempty) { for (i in files) print i; }; }'`; \ shift; \ if test -z "$(ETAGS_ARGS)$$*$$unique"; then :; else \ test -n "$$unique" || unique=$$empty_fix; \ if test $$# -gt 0; then \ $(ETAGS) $(ETAGSFLAGS) $(AM_ETAGSFLAGS) $(ETAGS_ARGS) \ "$$@" $$unique; \ else \ $(ETAGS) $(ETAGSFLAGS) $(AM_ETAGSFLAGS) $(ETAGS_ARGS) \ $$unique; \ fi; \ fi ctags: CTAGS CTAGS: $(HEADERS) $(SOURCES) $(TAGS_DEPENDENCIES) \ $(TAGS_FILES) $(LISP) list='$(SOURCES) $(HEADERS) $(LISP) $(TAGS_FILES)'; \ unique=`for i in $$list; do \ if test -f "$$i"; then echo $$i; else echo $(srcdir)/$$i; fi; \ done | \ $(AWK) '{ files[$$0] = 1; nonempty = 1; } \ END { if (nonempty) { for (i in files) print i; }; }'`; \ test -z "$(CTAGS_ARGS)$$unique" \ || $(CTAGS) $(CTAGSFLAGS) $(AM_CTAGSFLAGS) $(CTAGS_ARGS) \ $$unique GTAGS: here=`$(am__cd) $(top_builddir) && pwd` \ && $(am__cd) $(top_srcdir) \ && gtags -i $(GTAGS_ARGS) "$$here" distclean-tags: -rm -f TAGS ID GTAGS GRTAGS GSYMS GPATH tags distdir: $(DISTFILES) @srcdirstrip=`echo "$(srcdir)" | sed 's/[].[^$$\\*]/\\\\&/g'`; \ topsrcdirstrip=`echo "$(top_srcdir)" | sed 's/[].[^$$\\*]/\\\\&/g'`; \ list='$(DISTFILES)'; \ dist_files=`for file in $$list; do echo $$file; done | \ sed -e "s|^$$srcdirstrip/||;t" \ -e "s|^$$topsrcdirstrip/|$(top_builddir)/|;t"`; \ case $$dist_files in \ */*) $(MKDIR_P) `echo "$$dist_files" | \ sed '/\//!d;s|^|$(distdir)/|;s,/[^/]*$$,,' | \ sort -u` ;; \ esac; \ for file in $$dist_files; do \ if test -f $$file || test -d $$file; then d=.; else d=$(srcdir); fi; \ if test -d $$d/$$file; then \ dir=`echo "/$$file" | sed -e 's,/[^/]*$$,,'`; \ if test -d "$(distdir)/$$file"; then \ find "$(distdir)/$$file" -type d ! -perm -700 -exec chmod u+rwx {} \;; \ fi; \ if test -d $(srcdir)/$$file && test $$d != $(srcdir); then \ cp -fpR $(srcdir)/$$file "$(distdir)$$dir" || exit 1; \ find "$(distdir)/$$file" -type d ! -perm -700 -exec chmod u+rwx {} \;; \ fi; \ cp -fpR $$d/$$file "$(distdir)$$dir" || exit 1; \ else \ test -f "$(distdir)/$$file" \ || cp -p $$d/$$file "$(distdir)/$$file" \ || exit 1; \ fi; \ done check-am: all-am check: check-am all-am: Makefile $(LTLIBRARIES) installdirs: install: install-am install-exec: install-exec-am install-data: install-data-am uninstall: uninstall-am install-am: all-am @$(MAKE) $(AM_MAKEFLAGS) install-exec-am install-data-am installcheck: installcheck-am install-strip: if test -z '$(STRIP)'; then \ $(MAKE) $(AM_MAKEFLAGS) INSTALL_PROGRAM="$(INSTALL_STRIP_PROGRAM)" \ install_sh_PROGRAM="$(INSTALL_STRIP_PROGRAM)" INSTALL_STRIP_FLAG=-s \ install; \ else \ $(MAKE) $(AM_MAKEFLAGS) INSTALL_PROGRAM="$(INSTALL_STRIP_PROGRAM)" \ install_sh_PROGRAM="$(INSTALL_STRIP_PROGRAM)" INSTALL_STRIP_FLAG=-s \ "INSTALL_PROGRAM_ENV=STRIPPROG='$(STRIP)'" install; \ fi mostlyclean-generic: clean-generic: distclean-generic: -test -z "$(CONFIG_CLEAN_FILES)" || rm -f $(CONFIG_CLEAN_FILES) -test . = "$(srcdir)" || test -z "$(CONFIG_CLEAN_VPATH_FILES)" || rm -f $(CONFIG_CLEAN_VPATH_FILES) maintainer-clean-generic: @echo "This command is intended for maintainers to use" @echo "it deletes files that may require special tools to rebuild." clean: clean-am clean-am: clean-generic clean-libtool clean-noinstLTLIBRARIES \ mostlyclean-am distclean: distclean-am -rm -rf ./$(DEPDIR) -rm -f Makefile distclean-am: clean-am distclean-compile distclean-generic \ distclean-tags dvi: dvi-am dvi-am: html: html-am html-am: info: info-am info-am: install-data-am: install-dvi: install-dvi-am install-dvi-am: install-exec-am: install-html: install-html-am install-html-am: install-info: install-info-am install-info-am: install-man: install-pdf: install-pdf-am install-pdf-am: install-ps: install-ps-am install-ps-am: installcheck-am: maintainer-clean: maintainer-clean-am -rm -rf ./$(DEPDIR) -rm -f Makefile maintainer-clean-am: distclean-am maintainer-clean-generic mostlyclean: mostlyclean-am mostlyclean-am: mostlyclean-compile mostlyclean-generic \ mostlyclean-libtool pdf: pdf-am pdf-am: ps: ps-am ps-am: uninstall-am: .MAKE: install-am install-strip .PHONY: CTAGS GTAGS all all-am check check-am clean clean-generic \ clean-libtool clean-noinstLTLIBRARIES ctags distclean \ distclean-compile distclean-generic distclean-libtool \ distclean-tags distdir dvi dvi-am html html-am info info-am \ install install-am install-data install-data-am install-dvi \ install-dvi-am install-exec install-exec-am install-html \ install-html-am install-info install-info-am install-man \ install-pdf install-pdf-am install-ps install-ps-am \ install-strip installcheck installcheck-am installdirs \ maintainer-clean maintainer-clean-generic mostlyclean \ mostlyclean-compile mostlyclean-generic mostlyclean-libtool \ pdf pdf-am ps ps-am tags uninstall uninstall-am # Tell versions [3.59,3.63) of GNU make to not export all variables. # Otherwise a system limit (for SysV at least) may be exceeded. .NOEXPORT: fftw-3.3.3/kernel/align.c0000644000175000017500000000225612054410050012121 00000000000000/* * Copyright (c) 2003, 2007-11 Matteo Frigo * Copyright (c) 2003, 2007-11 Massachusetts Institute of Technology * * This program is free software; you can redistribute it and/or modify * it under the terms of the GNU General Public License as published by * the Free Software Foundation; either version 2 of the License, or * (at your option) any later version. * * This program is distributed in the hope that it will be useful, * but WITHOUT ANY WARRANTY; without even the implied warranty of * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the * GNU General Public License for more details. * * You should have received a copy of the GNU General Public License * along with this program; if not, write to the Free Software * Foundation, Inc., 51 Franklin Street, Fifth Floor, Boston, MA 02110-1301 USA * */ #include "ifftw.h" #if HAVE_SIMD # define ALGN 16 #else /* disable the alignment machinery, because it will break, e.g., if sizeof(R) == 12 (as in long-double/x86) */ # define ALGN 0 #endif /* NONPORTABLE */ int X(alignment_of)(R *p) { #if ALGN == 0 UNUSED(p); return 0; #else return (int)(((uintptr_t) p) % ALGN); #endif } fftw-3.3.3/kernel/alloc.c0000644000175000017500000001530412054410050012117 00000000000000/* * Copyright (c) 2003, 2007-11 Matteo Frigo * Copyright (c) 2003, 2007-11 Massachusetts Institute of Technology * * This program is free software; you can redistribute it and/or modify * it under the terms of the GNU General Public License as published by * the Free Software Foundation; either version 2 of the License, or * (at your option) any later version. * * This program is distributed in the hope that it will be useful, * but WITHOUT ANY WARRANTY; without even the implied warranty of * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the * GNU General Public License for more details. * * You should have received a copy of the GNU General Public License * along with this program; if not, write to the Free Software * Foundation, Inc., 51 Franklin Street, Fifth Floor, Boston, MA 02110-1301 USA * */ #include "ifftw.h" /********************************************************** * DEBUGGING CODE **********************************************************/ #if defined(FFTW_DEBUG_MALLOC) #include /* debugging malloc/free. 1) Initialize every malloced and freed area to random values, just to make sure we are not using uninitialized pointers. 2) check for blocks freed twice. 3) Check for writes past the ends of allocated blocks 4) destroy contents of freed blocks in order to detect incorrect reuse. 5) keep track of who allocates what and report memory leaks This code is a quick and dirty hack. May be nonportable. Use at your own risk. */ #define MAGIC ((size_t)0xABadCafe) #define PAD_FACTOR 2 #define SZ_HEADER (4 * sizeof(size_t)) #define HASHSZ 1031 static unsigned int hashaddr(void *p) { return ((unsigned long)p) % HASHSZ; } struct mstat { int siz; int maxsiz; int cnt; int maxcnt; }; static struct mstat mstat[MALLOC_WHAT_LAST]; struct minfo { const char *file; int line; size_t n; void *p; struct minfo *next; }; static struct minfo *minfo[HASHSZ] = {0}; #if defined(HAVE_THREADS) || defined(HAVE_OPENMP) int X(in_thread) = 0; #endif void *X(malloc_debug)(size_t n, enum malloc_tag what, const char *file, int line) { char *p; size_t i; struct minfo *info; struct mstat *stat = mstat + what; struct mstat *estat = mstat + EVERYTHING; if (n == 0) n = 1; if (!IN_THREAD) { stat->siz += n; if (stat->siz > stat->maxsiz) stat->maxsiz = stat->siz; estat->siz += n; if (estat->siz > estat->maxsiz) estat->maxsiz = estat->siz; } p = (char *) X(kernel_malloc)(PAD_FACTOR * n + SZ_HEADER); A(p); /* store the sz in a known position */ ((size_t *) p)[0] = n; ((size_t *) p)[1] = MAGIC; ((size_t *) p)[2] = what; /* fill with junk */ for (i = 0; i < PAD_FACTOR * n; i++) p[i + SZ_HEADER] = (char) (i ^ 0xEF); if (!IN_THREAD) { ++stat->cnt; ++estat->cnt; if (stat->cnt > stat->maxcnt) stat->maxcnt = stat->cnt; if (estat->cnt > estat->maxcnt) estat->maxcnt = estat->cnt; } /* skip the info we stored previously */ p = p + SZ_HEADER; if (!IN_THREAD) { unsigned int h = hashaddr(p); /* record allocation in allocation list */ info = (struct minfo *) malloc(sizeof(struct minfo)); info->n = n; info->file = file; info->line = line; info->p = p; info->next = minfo[h]; minfo[h] = info; } return (void *) p; } void X(ifree)(void *p) { char *q; A(p); q = ((char *) p) - SZ_HEADER; A(q); { size_t n = ((size_t *) q)[0]; size_t magic = ((size_t *) q)[1]; int what = ((size_t *) q)[2]; size_t i; struct mstat *stat = mstat + what; struct mstat *estat = mstat + EVERYTHING; /* set to zero to detect duplicate free's */ ((size_t *) q)[0] = 0; A(magic == MAGIC); ((size_t *) q)[1] = ~MAGIC; if (!IN_THREAD) { stat->siz -= n; A(stat->siz >= 0); estat->siz -= n; A(estat->siz >= 0); } /* check for writing past end of array: */ for (i = n; i < PAD_FACTOR * n; ++i) if (q[i + SZ_HEADER] != (char) (i ^ 0xEF)) { A(0 /* array bounds overwritten */ ); } for (i = 0; i < PAD_FACTOR * n; ++i) q[i + SZ_HEADER] = (char) (i ^ 0xAD); if (!IN_THREAD) { --stat->cnt; --estat->cnt; A(stat->cnt >= 0); A((stat->cnt == 0 && stat->siz == 0) || (stat->cnt > 0 && stat->siz > 0)); A(estat->cnt >= 0); A((estat->cnt == 0 && estat->siz == 0) || (estat->cnt > 0 && estat->siz > 0)); } X(kernel_free)(q); } if (!IN_THREAD) { /* delete minfo entry */ unsigned int h = hashaddr(p); struct minfo **i; for (i = minfo + h; *i; i = &((*i)->next)) { if ((*i)->p == p) { struct minfo *i0 = (*i)->next; free(*i); *i = i0; return; } } A(0 /* no entry in minfo list */ ); } } void X(malloc_print_minfo)(int verbose) { struct minfo *info; int what; unsigned int h; int leak = 0; if (verbose > 2) { static const char *names[MALLOC_WHAT_LAST] = { "EVERYTHING", "PLANS", "SOLVERS", "PROBLEMS", "BUFFERS", "HASHT", "TENSORS", "PLANNERS", "SLVDSC", "TWIDDLES", "STRIDES", "OTHER" }; printf("%12s %8s %8s %10s %10s\n", "what", "cnt", "maxcnt", "siz", "maxsiz"); for (what = 0; what < MALLOC_WHAT_LAST; ++what) { struct mstat *stat = mstat + what; printf("%12s %8d %8d %10d %10d\n", names[what], stat->cnt, stat->maxcnt, stat->siz, stat->maxsiz); } } for (h = 0; h < HASHSZ; ++h) if (minfo[h]) { printf("\nUnfreed allocations:\n"); break; } for (h = 0; h < HASHSZ; ++h) for (info = minfo[h]; info; info = info->next) { leak = 1; printf("%s:%d: %zd bytes at %p\n", info->file, info->line, info->n, info->p); } if (leak) abort(); } #else /********************************************************** * NON DEBUGGING CODE **********************************************************/ /* production version, no hacks */ void *X(malloc_plain)(size_t n) { void *p; if (n == 0) n = 1; p = X(kernel_malloc)(n); CK(p); #ifdef MIN_ALIGNMENT A((((uintptr_t)p) % MIN_ALIGNMENT) == 0); #endif return p; } void X(ifree)(void *p) { X(kernel_free)(p); } #endif void X(ifree0)(void *p) { /* common pattern */ if (p) X(ifree)(p); } fftw-3.3.3/kernel/assert.c0000644000175000017500000000216712054410050012331 00000000000000/* * Copyright (c) 2003, 2007-11 Matteo Frigo * Copyright (c) 2003, 2007-11 Massachusetts Institute of Technology * * This program is free software; you can redistribute it and/or modify * it under the terms of the GNU General Public License as published by * the Free Software Foundation; either version 2 of the License, or * (at your option) any later version. * * This program is distributed in the hope that it will be useful, * but WITHOUT ANY WARRANTY; without even the implied warranty of * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the * GNU General Public License for more details. * * You should have received a copy of the GNU General Public License * along with this program; if not, write to the Free Software * Foundation, Inc., 51 Franklin Street, Fifth Floor, Boston, MA 02110-1301 USA * */ #include "ifftw.h" #include #include void X(assertion_failed)(const char *s, int line, const char *file) { fflush(stdout); fprintf(stderr, "fftw: %s:%d: assertion failed: %s\n", file, line, s); #ifdef HAVE_ABORT abort(); #else exit(EXIT_FAILURE); #endif } fftw-3.3.3/kernel/awake.c0000644000175000017500000000174412054410050012120 00000000000000/* * Copyright (c) 2003, 2007-11 Matteo Frigo * Copyright (c) 2003, 2007-11 Massachusetts Institute of Technology * * This program is free software; you can redistribute it and/or modify * it under the terms of the GNU General Public License as published by * the Free Software Foundation; either version 2 of the License, or * (at your option) any later version. * * This program is distributed in the hope that it will be useful, * but WITHOUT ANY WARRANTY; without even the implied warranty of * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the * GNU General Public License for more details. * * You should have received a copy of the GNU General Public License * along with this program; if not, write to the Free Software * Foundation, Inc., 51 Franklin Street, Fifth Floor, Boston, MA 02110-1301 USA * */ #include "ifftw.h" void X(null_awake)(plan *ego, enum wakefulness wakefulness) { UNUSED(ego); UNUSED(wakefulness); /* do nothing */ } fftw-3.3.3/kernel/buffered.c0000644000175000017500000000440312054410050012605 00000000000000/* * Copyright (c) 2003, 2007-11 Matteo Frigo * Copyright (c) 2003, 2007-11 Massachusetts Institute of Technology * * This program is free software; you can redistribute it and/or modify * it under the terms of the GNU General Public License as published by * the Free Software Foundation; either version 2 of the License, or * (at your option) any later version. * * This program is distributed in the hope that it will be useful, * but WITHOUT ANY WARRANTY; without even the implied warranty of * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the * GNU General Public License for more details. * * You should have received a copy of the GNU General Public License * along with this program; if not, write to the Free Software * Foundation, Inc., 51 Franklin Street, Fifth Floor, Boston, MA 02110-1301 USA * */ /* routines shared by the various buffered solvers */ #include "ifftw.h" #define DEFAULT_MAXNBUF ((INT)256) /* approx. 512KB of buffers for complex data */ #define MAXBUFSZ (256 * 1024 / (INT)(sizeof(R))) INT X(nbuf)(INT n, INT vl, INT maxnbuf) { INT i, nbuf, lb; if (!maxnbuf) maxnbuf = DEFAULT_MAXNBUF; nbuf = X(imin)(maxnbuf, X(imin)(vl, X(imax)((INT)1, MAXBUFSZ / n))); /* * Look for a buffer number (not too small) that divides the * vector length, in order that we only need one child plan: */ lb = X(imax)(1, nbuf / 4); for (i = nbuf; i >= lb; --i) if (vl % i == 0) return i; /* whatever... */ return nbuf; } #define SKEW 6 /* need to be even for SIMD */ #define SKEWMOD 8 INT X(bufdist)(INT n, INT vl) { if (vl == 1) return n; else /* return smallest X such that X >= N and X == SKEW (mod SKEWMOD) */ return n + X(modulo)(SKEW - n, SKEWMOD); } int X(toobig)(INT n) { return n > MAXBUFSZ; } /* TRUE if there exists i < which such that maxnbuf[i] and maxnbuf[which] yield the same value, in which case we canonicalize on the minimum value */ int X(nbuf_redundant)(INT n, INT vl, int which, const INT *maxnbuf, int nmaxnbuf) { int i; (void)nmaxnbuf; /* UNUSED */ for (i = 0; i < which; ++i) if (X(nbuf)(n, vl, maxnbuf[i]) == X(nbuf)(n, vl, maxnbuf[which])) return 1; return 0; } fftw-3.3.3/kernel/cpy1d.c0000644000175000017500000000346212054410050012047 00000000000000/* * Copyright (c) 2003, 2007-11 Matteo Frigo * Copyright (c) 2003, 2007-11 Massachusetts Institute of Technology * * This program is free software; you can redistribute it and/or modify * it under the terms of the GNU General Public License as published by * the Free Software Foundation; either version 2 of the License, or * (at your option) any later version. * * This program is distributed in the hope that it will be useful, * but WITHOUT ANY WARRANTY; without even the implied warranty of * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the * GNU General Public License for more details. * * You should have received a copy of the GNU General Public License * along with this program; if not, write to the Free Software * Foundation, Inc., 51 Franklin Street, Fifth Floor, Boston, MA 02110-1301 USA * */ /* out of place 1D copy routine */ #include "ifftw.h" void X(cpy1d)(R *I, R *O, INT n0, INT is0, INT os0, INT vl) { INT i0, v; A(I != O); switch (vl) { case 1: if ((n0 & 1) || is0 != 1 || os0 != 1) { for (; n0 > 0; --n0, I += is0, O += os0) *O = *I; break; } n0 /= 2; is0 = 2; os0 = 2; /* fall through */ case 2: if ((n0 & 1) || is0 != 2 || os0 != 2) { for (; n0 > 0; --n0, I += is0, O += os0) { R x0 = I[0]; R x1 = I[1]; O[0] = x0; O[1] = x1; } break; } n0 /= 2; is0 = 4; os0 = 4; /* fall through */ case 4: for (; n0 > 0; --n0, I += is0, O += os0) { R x0 = I[0]; R x1 = I[1]; R x2 = I[2]; R x3 = I[3]; O[0] = x0; O[1] = x1; O[2] = x2; O[3] = x3; } break; default: for (i0 = 0; i0 < n0; ++i0) for (v = 0; v < vl; ++v) { R x0 = I[i0 * is0 + v]; O[i0 * os0 + v] = x0; } break; } } fftw-3.3.3/kernel/cpy2d-pair.c0000644000175000017500000000400612054410050012774 00000000000000/* * Copyright (c) 2003, 2007-11 Matteo Frigo * Copyright (c) 2003, 2007-11 Massachusetts Institute of Technology * * This program is free software; you can redistribute it and/or modify * it under the terms of the GNU General Public License as published by * the Free Software Foundation; either version 2 of the License, or * (at your option) any later version. * * This program is distributed in the hope that it will be useful, * but WITHOUT ANY WARRANTY; without even the implied warranty of * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the * GNU General Public License for more details. * * You should have received a copy of the GNU General Public License * along with this program; if not, write to the Free Software * Foundation, Inc., 51 Franklin Street, Fifth Floor, Boston, MA 02110-1301 USA * */ /* out of place copy routines for pairs of isomorphic 2D arrays */ #include "ifftw.h" void X(cpy2d_pair)(R *I0, R *I1, R *O0, R *O1, INT n0, INT is0, INT os0, INT n1, INT is1, INT os1) { INT i0, i1; for (i1 = 0; i1 < n1; ++i1) for (i0 = 0; i0 < n0; ++i0) { R x0 = I0[i0 * is0 + i1 * is1]; R x1 = I1[i0 * is0 + i1 * is1]; O0[i0 * os0 + i1 * os1] = x0; O1[i0 * os0 + i1 * os1] = x1; } } /* like cpy2d_pair, but read input contiguously if possible */ void X(cpy2d_pair_ci)(R *I0, R *I1, R *O0, R *O1, INT n0, INT is0, INT os0, INT n1, INT is1, INT os1) { if (IABS(is0) < IABS(is1)) /* inner loop is for n0 */ X(cpy2d_pair) (I0, I1, O0, O1, n0, is0, os0, n1, is1, os1); else X(cpy2d_pair) (I0, I1, O0, O1, n1, is1, os1, n0, is0, os0); } /* like cpy2d_pair, but write output contiguously if possible */ void X(cpy2d_pair_co)(R *I0, R *I1, R *O0, R *O1, INT n0, INT is0, INT os0, INT n1, INT is1, INT os1) { if (IABS(os0) < IABS(os1)) /* inner loop is for n0 */ X(cpy2d_pair) (I0, I1, O0, O1, n0, is0, os0, n1, is1, os1); else X(cpy2d_pair) (I0, I1, O0, O1, n1, is1, os1, n0, is0, os0); } fftw-3.3.3/kernel/cpy2d.c0000644000175000017500000001323012054410050012042 00000000000000/* * Copyright (c) 2003, 2007-11 Matteo Frigo * Copyright (c) 2003, 2007-11 Massachusetts Institute of Technology * * This program is free software; you can redistribute it and/or modify * it under the terms of the GNU General Public License as published by * the Free Software Foundation; either version 2 of the License, or * (at your option) any later version. * * This program is distributed in the hope that it will be useful, * but WITHOUT ANY WARRANTY; without even the implied warranty of * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the * GNU General Public License for more details. * * You should have received a copy of the GNU General Public License * along with this program; if not, write to the Free Software * Foundation, Inc., 51 Franklin Street, Fifth Floor, Boston, MA 02110-1301 USA * */ /* out of place 2D copy routines */ #include "ifftw.h" #if defined(__x86_64__) || defined(_M_X64) || defined(_M_AMD64) # ifdef HAVE_XMMINTRIN_H # include # define WIDE_TYPE __m128 # endif #endif #ifndef WIDE_TYPE /* fall back to double, which means that WIDE_TYPE will be unused */ # define WIDE_TYPE double #endif void X(cpy2d)(R *I, R *O, INT n0, INT is0, INT os0, INT n1, INT is1, INT os1, INT vl) { INT i0, i1, v; switch (vl) { case 1: for (i1 = 0; i1 < n1; ++i1) for (i0 = 0; i0 < n0; ++i0) { R x0 = I[i0 * is0 + i1 * is1]; O[i0 * os0 + i1 * os1] = x0; } break; case 2: if (1 && (2 * sizeof(R) == sizeof(WIDE_TYPE)) && (sizeof(WIDE_TYPE) > sizeof(double)) && (((size_t)I) % sizeof(WIDE_TYPE) == 0) && (((size_t)O) % sizeof(WIDE_TYPE) == 0) && ((is0 & 1) == 0) && ((is1 & 1) == 0) && ((os0 & 1) == 0) && ((os1 & 1) == 0)) { /* copy R[2] as WIDE_TYPE if WIDE_TYPE is large enough to hold R[2], and if the input is properly aligned. This is a win when R==double and WIDE_TYPE is 128 bits. */ for (i1 = 0; i1 < n1; ++i1) for (i0 = 0; i0 < n0; ++i0) { *(WIDE_TYPE *)&O[i0 * os0 + i1 * os1] = *(WIDE_TYPE *)&I[i0 * is0 + i1 * is1]; } } else if (1 && (2 * sizeof(R) == sizeof(double)) && (((size_t)I) % sizeof(double) == 0) && (((size_t)O) % sizeof(double) == 0) && ((is0 & 1) == 0) && ((is1 & 1) == 0) && ((os0 & 1) == 0) && ((os1 & 1) == 0)) { /* copy R[2] as double if double is large enough to hold R[2], and if the input is properly aligned. This case applies when R==float */ for (i1 = 0; i1 < n1; ++i1) for (i0 = 0; i0 < n0; ++i0) { *(double *)&O[i0 * os0 + i1 * os1] = *(double *)&I[i0 * is0 + i1 * is1]; } } else { for (i1 = 0; i1 < n1; ++i1) for (i0 = 0; i0 < n0; ++i0) { R x0 = I[i0 * is0 + i1 * is1]; R x1 = I[i0 * is0 + i1 * is1 + 1]; O[i0 * os0 + i1 * os1] = x0; O[i0 * os0 + i1 * os1 + 1] = x1; } } break; default: for (i1 = 0; i1 < n1; ++i1) for (i0 = 0; i0 < n0; ++i0) for (v = 0; v < vl; ++v) { R x0 = I[i0 * is0 + i1 * is1 + v]; O[i0 * os0 + i1 * os1 + v] = x0; } break; } } /* like cpy2d, but read input contiguously if possible */ void X(cpy2d_ci)(R *I, R *O, INT n0, INT is0, INT os0, INT n1, INT is1, INT os1, INT vl) { if (IABS(is0) < IABS(is1)) /* inner loop is for n0 */ X(cpy2d) (I, O, n0, is0, os0, n1, is1, os1, vl); else X(cpy2d) (I, O, n1, is1, os1, n0, is0, os0, vl); } /* like cpy2d, but write output contiguously if possible */ void X(cpy2d_co)(R *I, R *O, INT n0, INT is0, INT os0, INT n1, INT is1, INT os1, INT vl) { if (IABS(os0) < IABS(os1)) /* inner loop is for n0 */ X(cpy2d) (I, O, n0, is0, os0, n1, is1, os1, vl); else X(cpy2d) (I, O, n1, is1, os1, n0, is0, os0, vl); } /* tiled copy routines */ struct cpy2d_closure { R *I, *O; INT is0, os0, is1, os1, vl; R *buf; }; static void dotile(INT n0l, INT n0u, INT n1l, INT n1u, void *args) { struct cpy2d_closure *k = (struct cpy2d_closure *)args; X(cpy2d)(k->I + n0l * k->is0 + n1l * k->is1, k->O + n0l * k->os0 + n1l * k->os1, n0u - n0l, k->is0, k->os0, n1u - n1l, k->is1, k->os1, k->vl); } static void dotile_buf(INT n0l, INT n0u, INT n1l, INT n1u, void *args) { struct cpy2d_closure *k = (struct cpy2d_closure *)args; /* copy from I to buf */ X(cpy2d_ci)(k->I + n0l * k->is0 + n1l * k->is1, k->buf, n0u - n0l, k->is0, k->vl, n1u - n1l, k->is1, k->vl * (n0u - n0l), k->vl); /* copy from buf to O */ X(cpy2d_co)(k->buf, k->O + n0l * k->os0 + n1l * k->os1, n0u - n0l, k->vl, k->os0, n1u - n1l, k->vl * (n0u - n0l), k->os1, k->vl); } void X(cpy2d_tiled)(R *I, R *O, INT n0, INT is0, INT os0, INT n1, INT is1, INT os1, INT vl) { INT tilesz = X(compute_tilesz)(vl, 1 /* input array */ + 1 /* ouput array */); struct cpy2d_closure k; k.I = I; k.O = O; k.is0 = is0; k.os0 = os0; k.is1 = is1; k.os1 = os1; k.vl = vl; k.buf = 0; /* unused */ X(tile2d)(0, n0, 0, n1, tilesz, dotile, &k); } void X(cpy2d_tiledbuf)(R *I, R *O, INT n0, INT is0, INT os0, INT n1, INT is1, INT os1, INT vl) { R buf[CACHESIZE / (2 * sizeof(R))]; /* input and buffer in cache, or output and buffer in cache */ INT tilesz = X(compute_tilesz)(vl, 2); struct cpy2d_closure k; k.I = I; k.O = O; k.is0 = is0; k.os0 = os0; k.is1 = is1; k.os1 = os1; k.vl = vl; k.buf = buf; A(tilesz * tilesz * vl * sizeof(R) <= sizeof(buf)); X(tile2d)(0, n0, 0, n1, tilesz, dotile_buf, &k); } fftw-3.3.3/kernel/ct.c0000644000175000017500000000215112054410050011427 00000000000000/* * Copyright (c) 2003, 2007-11 Matteo Frigo * Copyright (c) 2003, 2007-11 Massachusetts Institute of Technology * * This program is free software; you can redistribute it and/or modify * it under the terms of the GNU General Public License as published by * the Free Software Foundation; either version 2 of the License, or * (at your option) any later version. * * This program is distributed in the hope that it will be useful, * but WITHOUT ANY WARRANTY; without even the implied warranty of * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the * GNU General Public License for more details. * * You should have received a copy of the GNU General Public License * along with this program; if not, write to the Free Software * Foundation, Inc., 51 Franklin Street, Fifth Floor, Boston, MA 02110-1301 USA * */ /* common routines for Cooley-Tukey algorithms */ #include "ifftw.h" #define POW2P(n) (((n) > 0) && (((n) & ((n) - 1)) == 0)) /* TRUE if radix-r is ugly for size n */ int X(ct_uglyp)(INT min_n, INT v, INT n, INT r) { return (n <= min_n) || (POW2P(n) && (v * (n / r)) <= 4); } fftw-3.3.3/kernel/debug.c0000644000175000017500000000263512054410050012116 00000000000000/* * Copyright (c) 2003, 2007-11 Matteo Frigo * Copyright (c) 2003, 2007-11 Massachusetts Institute of Technology * * This program is free software; you can redistribute it and/or modify * it under the terms of the GNU General Public License as published by * the Free Software Foundation; either version 2 of the License, or * (at your option) any later version. * * This program is distributed in the hope that it will be useful, * but WITHOUT ANY WARRANTY; without even the implied warranty of * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the * GNU General Public License for more details. * * You should have received a copy of the GNU General Public License * along with this program; if not, write to the Free Software * Foundation, Inc., 51 Franklin Street, Fifth Floor, Boston, MA 02110-1301 USA * */ #include "ifftw.h" #ifdef FFTW_DEBUG #include typedef struct { printer super; FILE *f; } P_file; static void putchr_file(printer *p_, char c) { P_file *p = (P_file *) p_; fputc(c, p->f); } static printer *mkprinter_file(FILE *f) { P_file *p = (P_file *) X(mkprinter)(sizeof(P_file), putchr_file, 0); p->f = f; return &p->super; } void X(debug)(const char *format, ...) { va_list ap; printer *p = mkprinter_file(stderr); va_start(ap, format); p->vprint(p, format, ap); va_end(ap); X(printer_destroy)(p); } #endif fftw-3.3.3/kernel/extract-reim.c0000644000175000017500000000231012054410050013422 00000000000000/* * Copyright (c) 2003, 2007-11 Matteo Frigo * Copyright (c) 2003, 2007-11 Massachusetts Institute of Technology * * This program is free software; you can redistribute it and/or modify * it under the terms of the GNU General Public License as published by * the Free Software Foundation; either version 2 of the License, or * (at your option) any later version. * * This program is distributed in the hope that it will be useful, * but WITHOUT ANY WARRANTY; without even the implied warranty of * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the * GNU General Public License for more details. * * You should have received a copy of the GNU General Public License * along with this program; if not, write to the Free Software * Foundation, Inc., 51 Franklin Street, Fifth Floor, Boston, MA 02110-1301 USA * */ #include "ifftw.h" /* decompose complex pointer into real and imaginary parts. Flip real and imaginary if there the sign does not match FFTW's idea of what the sign should be */ void X(extract_reim)(int sign, R *c, R **r, R **i) { if (sign == FFT_SIGN) { *r = c + 0; *i = c + 1; } else { *r = c + 1; *i = c + 0; } } fftw-3.3.3/kernel/hash.c0000644000175000017500000000175412054410050011754 00000000000000/* * Copyright (c) 2003, 2007-11 Matteo Frigo * Copyright (c) 2003, 2007-11 Massachusetts Institute of Technology * * This program is free software; you can redistribute it and/or modify * it under the terms of the GNU General Public License as published by * the Free Software Foundation; either version 2 of the License, or * (at your option) any later version. * * This program is distributed in the hope that it will be useful, * but WITHOUT ANY WARRANTY; without even the implied warranty of * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the * GNU General Public License for more details. * * You should have received a copy of the GNU General Public License * along with this program; if not, write to the Free Software * Foundation, Inc., 51 Franklin Street, Fifth Floor, Boston, MA 02110-1301 USA * */ #include "ifftw.h" unsigned X(hash)(const char *s) { unsigned h = 0xDEADBEEFu; do { h = h * 17 + (int)*s; } while (*s++); return h; } fftw-3.3.3/kernel/iabs.c0000644000175000017500000000163212054410050011742 00000000000000/* * Copyright (c) 2003, 2007-11 Matteo Frigo * Copyright (c) 2003, 2007-11 Massachusetts Institute of Technology * * This program is free software; you can redistribute it and/or modify * it under the terms of the GNU General Public License as published by * the Free Software Foundation; either version 2 of the License, or * (at your option) any later version. * * This program is distributed in the hope that it will be useful, * but WITHOUT ANY WARRANTY; without even the implied warranty of * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the * GNU General Public License for more details. * * You should have received a copy of the GNU General Public License * along with this program; if not, write to the Free Software * Foundation, Inc., 51 Franklin Street, Fifth Floor, Boston, MA 02110-1301 USA * */ #include "ifftw.h" INT X(iabs)(INT a) { return a < 0 ? (0 - a) : a; } fftw-3.3.3/kernel/kalloc.c0000644000175000017500000001112112054410050012263 00000000000000/* * Copyright (c) 2003, 2007-11 Matteo Frigo * Copyright (c) 2003, 2007-11 Massachusetts Institute of Technology * * This program is free software; you can redistribute it and/or modify * it under the terms of the GNU General Public License as published by * the Free Software Foundation; either version 2 of the License, or * (at your option) any later version. * * This program is distributed in the hope that it will be useful, * but WITHOUT ANY WARRANTY; without even the implied warranty of * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the * GNU General Public License for more details. * * You should have received a copy of the GNU General Public License * along with this program; if not, write to the Free Software * Foundation, Inc., 51 Franklin Street, Fifth Floor, Boston, MA 02110-1301 USA * */ #include "ifftw.h" /* ``kernel'' malloc(), with proper memory alignment */ #if defined(HAVE_DECL_MEMALIGN) && !HAVE_DECL_MEMALIGN # if defined(HAVE_MALLOC_H) # include # else extern void *memalign(size_t, size_t); # endif #endif #if defined(HAVE_DECL_POSIX_MEMALIGN) && !HAVE_DECL_POSIX_MEMALIGN extern int posix_memalign(void **, size_t, size_t); #endif #if defined(macintosh) /* MacOS 9 */ # include #endif #define real_free free /* memalign and malloc use ordinary free */ #define IS_POWER_OF_TWO(n) (((n) > 0) && (((n) & ((n) - 1)) == 0)) #if defined(WITH_OUR_MALLOC) && (MIN_ALIGNMENT >= 8) && IS_POWER_OF_TWO(MIN_ALIGNMENT) /* Our own MIN_ALIGNMENT-aligned malloc/free. Assumes sizeof(void*) is a power of two <= 8 and that malloc is at least sizeof(void*)-aligned. The main reason for this routine is that, as of this writing, Windows does not include any aligned allocation routines in its system libraries, and instead provides an implementation with a Visual C++ "Processor Pack" that you have to statically link into your program. We do not want to require users to have VC++ (e.g. gcc/MinGW should be fine). Our code should be at least as good as the MS _aligned_malloc, in any case, according to second-hand reports of the algorithm it employs (also based on plain malloc). */ static void *our_malloc(size_t n) { void *p0, *p; if (!(p0 = malloc(n + MIN_ALIGNMENT))) return (void *) 0; p = (void *) (((uintptr_t) p0 + MIN_ALIGNMENT) & (~((uintptr_t) (MIN_ALIGNMENT - 1)))); *((void **) p - 1) = p0; return p; } static void our_free(void *p) { if (p) free(*((void **) p - 1)); } #endif void *X(kernel_malloc)(size_t n) { void *p; #if defined(MIN_ALIGNMENT) # if defined(WITH_OUR_MALLOC) p = our_malloc(n); # undef real_free # define real_free our_free # elif defined(__FreeBSD__) && (MIN_ALIGNMENT <= 16) /* FreeBSD does not have memalign, but its malloc is 16-byte aligned. */ p = malloc(n); # elif (defined(__MACOSX__) || defined(__APPLE__)) && (MIN_ALIGNMENT <= 16) /* MacOS X malloc is already 16-byte aligned */ p = malloc(n); # elif defined(HAVE_MEMALIGN) p = memalign(MIN_ALIGNMENT, n); # elif defined(HAVE_POSIX_MEMALIGN) /* note: posix_memalign is broken in glibc 2.2.5: it constrains the size, not the alignment, to be (power of two) * sizeof(void*). The bug seems to have been fixed as of glibc 2.3.1. */ if (posix_memalign(&p, MIN_ALIGNMENT, n)) p = (void*) 0; # elif defined(__ICC) || defined(__INTEL_COMPILER) || defined(HAVE__MM_MALLOC) /* Intel's C compiler defines _mm_malloc and _mm_free intrinsics */ p = (void *) _mm_malloc(n, MIN_ALIGNMENT); # undef real_free # define real_free _mm_free # elif defined(_MSC_VER) /* MS Visual C++ 6.0 with a "Processor Pack" supports SIMD and _aligned_malloc/free (uses malloc.h) */ p = (void *) _aligned_malloc(n, MIN_ALIGNMENT); # undef real_free # define real_free _aligned_free # elif defined(macintosh) /* MacOS 9 */ p = (void *) MPAllocateAligned(n, # if MIN_ALIGNMENT == 8 kMPAllocate8ByteAligned, # elif MIN_ALIGNMENT == 16 kMPAllocate16ByteAligned, # elif MIN_ALIGNMENT == 32 kMPAllocate32ByteAligned, # else # error "Unknown alignment for MPAllocateAligned" # endif 0); # undef real_free # define real_free MPFree # else /* Add your machine here and send a patch to fftw@fftw.org or (e.g. for Windows) configure --with-our-malloc */ # error "Don't know how to malloc() aligned memory ... try configuring --with-our-malloc" # endif #else /* !defined(MIN_ALIGNMENT) */ p = malloc(n); #endif return p; } void X(kernel_free)(void *p) { real_free(p); } fftw-3.3.3/kernel/md5-1.c0000644000175000017500000000256612054410050011656 00000000000000/* * Copyright (c) 2003, 2007-11 Matteo Frigo * Copyright (c) 2003, 2007-11 Massachusetts Institute of Technology * * This program is free software; you can redistribute it and/or modify * it under the terms of the GNU General Public License as published by * the Free Software Foundation; either version 2 of the License, or * (at your option) any later version. * * This program is distributed in the hope that it will be useful, * but WITHOUT ANY WARRANTY; without even the implied warranty of * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the * GNU General Public License for more details. * * You should have received a copy of the GNU General Public License * along with this program; if not, write to the Free Software * Foundation, Inc., 51 Franklin Street, Fifth Floor, Boston, MA 02110-1301 USA * */ #include "ifftw.h" void X(md5putb)(md5 *p, const void *d_, size_t len) { size_t i; const unsigned char *d = (const unsigned char *)d_; for (i = 0; i < len; ++i) X(md5putc)(p, d[i]); } void X(md5puts)(md5 *p, const char *s) { /* also hash final '\0' */ do { X(md5putc)(p, *s); } while(*s++); } void X(md5int)(md5 *p, int i) { X(md5putb)(p, &i, sizeof(i)); } void X(md5INT)(md5 *p, INT i) { X(md5putb)(p, &i, sizeof(i)); } void X(md5unsigned)(md5 *p, unsigned i) { X(md5putb)(p, &i, sizeof(i)); } fftw-3.3.3/kernel/md5.c0000644000175000017500000001100612054410050011505 00000000000000/* * Copyright (c) 2003, 2007-11 Matteo Frigo * Copyright (c) 2003, 2007-11 Massachusetts Institute of Technology * * This program is free software; you can redistribute it and/or modify * it under the terms of the GNU General Public License as published by * the Free Software Foundation; either version 2 of the License, or * (at your option) any later version. * * This program is distributed in the hope that it will be useful, * but WITHOUT ANY WARRANTY; without even the implied warranty of * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the * GNU General Public License for more details. * * You should have received a copy of the GNU General Public License * along with this program; if not, write to the Free Software * Foundation, Inc., 51 Franklin Street, Fifth Floor, Boston, MA 02110-1301 USA * */ /* independent implementation of Ron Rivest's MD5 message-digest algorithm, based on rfc 1321. Optimized for small code size, not speed. Works as long as sizeof(md5uint) >= 4. */ #include "ifftw.h" /* sintab[i] = 4294967296.0 * abs(sin((double)(i + 1))) */ static const md5uint sintab[64] = { 0xd76aa478, 0xe8c7b756, 0x242070db, 0xc1bdceee, 0xf57c0faf, 0x4787c62a, 0xa8304613, 0xfd469501, 0x698098d8, 0x8b44f7af, 0xffff5bb1, 0x895cd7be, 0x6b901122, 0xfd987193, 0xa679438e, 0x49b40821, 0xf61e2562, 0xc040b340, 0x265e5a51, 0xe9b6c7aa, 0xd62f105d, 0x02441453, 0xd8a1e681, 0xe7d3fbc8, 0x21e1cde6, 0xc33707d6, 0xf4d50d87, 0x455a14ed, 0xa9e3e905, 0xfcefa3f8, 0x676f02d9, 0x8d2a4c8a, 0xfffa3942, 0x8771f681, 0x6d9d6122, 0xfde5380c, 0xa4beea44, 0x4bdecfa9, 0xf6bb4b60, 0xbebfbc70, 0x289b7ec6, 0xeaa127fa, 0xd4ef3085, 0x04881d05, 0xd9d4d039, 0xe6db99e5, 0x1fa27cf8, 0xc4ac5665, 0xf4292244, 0x432aff97, 0xab9423a7, 0xfc93a039, 0x655b59c3, 0x8f0ccc92, 0xffeff47d, 0x85845dd1, 0x6fa87e4f, 0xfe2ce6e0, 0xa3014314, 0x4e0811a1, 0xf7537e82, 0xbd3af235, 0x2ad7d2bb, 0xeb86d391 }; /* see rfc 1321 section 3.4 */ static const struct roundtab { char k; char s; } roundtab[64] = { { 0, 7}, { 1, 12}, { 2, 17}, { 3, 22}, { 4, 7}, { 5, 12}, { 6, 17}, { 7, 22}, { 8, 7}, { 9, 12}, { 10, 17}, { 11, 22}, { 12, 7}, { 13, 12}, { 14, 17}, { 15, 22}, { 1, 5}, { 6, 9}, { 11, 14}, { 0, 20}, { 5, 5}, { 10, 9}, { 15, 14}, { 4, 20}, { 9, 5}, { 14, 9}, { 3, 14}, { 8, 20}, { 13, 5}, { 2, 9}, { 7, 14}, { 12, 20}, { 5, 4}, { 8, 11}, { 11, 16}, { 14, 23}, { 1, 4}, { 4, 11}, { 7, 16}, { 10, 23}, { 13, 4}, { 0, 11}, { 3, 16}, { 6, 23}, { 9, 4}, { 12, 11}, { 15, 16}, { 2, 23}, { 0, 6}, { 7, 10}, { 14, 15}, { 5, 21}, { 12, 6}, { 3, 10}, { 10, 15}, { 1, 21}, { 8, 6}, { 15, 10}, { 6, 15}, { 13, 21}, { 4, 6}, { 11, 10}, { 2, 15}, { 9, 21} }; #define rol(a, s) ((a << (int)(s)) | (a >> (32 - (int)(s)))) static void doblock(md5sig state, const unsigned char *data) { md5uint a, b, c, d, t, x[16]; const md5uint msk = (md5uint)0xffffffffUL; int i; /* encode input bytes into md5uint */ for (i = 0; i < 16; ++i) { const unsigned char *p = data + 4 * i; x[i] = p[0] | (p[1] << 8) | (p[2] << 16) | (p[3] << 24); } a = state[0]; b = state[1]; c = state[2]; d = state[3]; for (i = 0; i < 64; ++i) { const struct roundtab *p = roundtab + i; switch (i >> 4) { case 0: a += (b & c) | (~b & d); break; case 1: a += (b & d) | (c & ~d); break; case 2: a += b ^ c ^ d; break; case 3: a += c ^ (b | ~d); break; } a += sintab[i]; a += x[(int)(p->k)]; a &= msk; t = b + rol(a, p->s); a = d; d = c; c = b; b = t; } state[0] = (state[0] + a) & msk; state[1] = (state[1] + b) & msk; state[2] = (state[2] + c) & msk; state[3] = (state[3] + d) & msk; } void X(md5begin)(md5 *p) { p->s[0] = 0x67452301; p->s[1] = 0xefcdab89; p->s[2] = 0x98badcfe; p->s[3] = 0x10325476; p->l = 0; } void X(md5putc)(md5 *p, unsigned char c) { p->c[p->l % 64] = c; if (((++p->l) % 64) == 0) doblock(p->s, p->c); } void X(md5end)(md5 *p) { unsigned l, i; l = 8 * p->l; /* length before padding, in bits */ /* rfc 1321 section 3.1: padding */ X(md5putc)(p, 0x80); while ((p->l % 64) != 56) X(md5putc)(p, 0x00); /* rfc 1321 section 3.2: length (little endian) */ for (i = 0; i < 8; ++i) { X(md5putc)(p, l & 0xFF); l = l >> 8; } /* Now p->l % 64 == 0 and signature is in p->s */ } fftw-3.3.3/kernel/minmax.c0000644000175000017500000000173112054410050012315 00000000000000/* * Copyright (c) 2003, 2007-11 Matteo Frigo * Copyright (c) 2003, 2007-11 Massachusetts Institute of Technology * * This program is free software; you can redistribute it and/or modify * it under the terms of the GNU General Public License as published by * the Free Software Foundation; either version 2 of the License, or * (at your option) any later version. * * This program is distributed in the hope that it will be useful, * but WITHOUT ANY WARRANTY; without even the implied warranty of * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the * GNU General Public License for more details. * * You should have received a copy of the GNU General Public License * along with this program; if not, write to the Free Software * Foundation, Inc., 51 Franklin Street, Fifth Floor, Boston, MA 02110-1301 USA * */ #include "ifftw.h" INT X(imax)(INT a, INT b) { return (a > b) ? a : b; } INT X(imin)(INT a, INT b) { return (a < b) ? a : b; } fftw-3.3.3/kernel/ops.c0000644000175000017500000000310512054410050011622 00000000000000/* * Copyright (c) 2003, 2007-11 Matteo Frigo * Copyright (c) 2003, 2007-11 Massachusetts Institute of Technology * * This program is free software; you can redistribute it and/or modify * it under the terms of the GNU General Public License as published by * the Free Software Foundation; either version 2 of the License, or * (at your option) any later version. * * This program is distributed in the hope that it will be useful, * but WITHOUT ANY WARRANTY; without even the implied warranty of * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the * GNU General Public License for more details. * * You should have received a copy of the GNU General Public License * along with this program; if not, write to the Free Software * Foundation, Inc., 51 Franklin Street, Fifth Floor, Boston, MA 02110-1301 USA * */ #include "ifftw.h" void X(ops_zero)(opcnt *dst) { dst->add = dst->mul = dst->fma = dst->other = 0; } void X(ops_cpy)(const opcnt *src, opcnt *dst) { *dst = *src; } void X(ops_other)(INT o, opcnt *dst) { X(ops_zero)(dst); dst->other = o; } void X(ops_madd)(INT m, const opcnt *a, const opcnt *b, opcnt *dst) { dst->add = m * a->add + b->add; dst->mul = m * a->mul + b->mul; dst->fma = m * a->fma + b->fma; dst->other = m * a->other + b->other; } void X(ops_add)(const opcnt *a, const opcnt *b, opcnt *dst) { X(ops_madd)(1, a, b, dst); } void X(ops_add2)(const opcnt *a, opcnt *dst) { X(ops_add)(a, dst, dst); } void X(ops_madd2)(INT m, const opcnt *a, opcnt *dst) { X(ops_madd)(m, a, dst, dst); } fftw-3.3.3/kernel/pickdim.c0000644000175000017500000000543512054410050012451 00000000000000/* * Copyright (c) 2003, 2007-11 Matteo Frigo * Copyright (c) 2003, 2007-11 Massachusetts Institute of Technology * * This program is free software; you can redistribute it and/or modify * it under the terms of the GNU General Public License as published by * the Free Software Foundation; either version 2 of the License, or * (at your option) any later version. * * This program is distributed in the hope that it will be useful, * but WITHOUT ANY WARRANTY; without even the implied warranty of * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the * GNU General Public License for more details. * * You should have received a copy of the GNU General Public License * along with this program; if not, write to the Free Software * Foundation, Inc., 51 Franklin Street, Fifth Floor, Boston, MA 02110-1301 USA * */ #include "ifftw.h" /* Given a solver which_dim, a vector sz, and whether or not the transform is out-of-place, return the actual dimension index that it corresponds to. The basic idea here is that we return the which_dim'th valid dimension, starting from the end if which_dim < 0. */ static int really_pickdim(int which_dim, const tensor *sz, int oop, int *dp) { int i; int count_ok = 0; if (which_dim > 0) { for (i = 0; i < sz->rnk; ++i) { if (oop || sz->dims[i].is == sz->dims[i].os) if (++count_ok == which_dim) { *dp = i; return 1; } } } else if (which_dim < 0) { for (i = sz->rnk - 1; i >= 0; --i) { if (oop || sz->dims[i].is == sz->dims[i].os) if (++count_ok == -which_dim) { *dp = i; return 1; } } } else { /* zero: pick the middle, if valid */ i = (sz->rnk - 1) / 2; if (i >= 0 && (oop || sz->dims[i].is == sz->dims[i].os)) { *dp = i; return 1; } } return 0; } /* Like really_pickdim, but only returns 1 if no previous "buddy" which_dim in the buddies list would give the same dim. */ int X(pickdim)(int which_dim, const int *buddies, int nbuddies, const tensor *sz, int oop, int *dp) { int i, d1; if (!really_pickdim(which_dim, sz, oop, dp)) return 0; /* check whether some buddy solver would produce the same dim. If so, consider this solver unapplicable and let the buddy take care of it. The smallest-indexed buddy is applicable. */ for (i = 0; i < nbuddies; ++i) { if (buddies[i] == which_dim) break; /* found self */ if (really_pickdim(buddies[i], sz, oop, &d1) && *dp == d1) return 0; /* found equivalent buddy */ } return 1; } fftw-3.3.3/kernel/plan.c0000644000175000017500000000344312054410050011760 00000000000000/* * Copyright (c) 2003, 2007-11 Matteo Frigo * Copyright (c) 2003, 2007-11 Massachusetts Institute of Technology * * This program is free software; you can redistribute it and/or modify * it under the terms of the GNU General Public License as published by * the Free Software Foundation; either version 2 of the License, or * (at your option) any later version. * * This program is distributed in the hope that it will be useful, * but WITHOUT ANY WARRANTY; without even the implied warranty of * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the * GNU General Public License for more details. * * You should have received a copy of the GNU General Public License * along with this program; if not, write to the Free Software * Foundation, Inc., 51 Franklin Street, Fifth Floor, Boston, MA 02110-1301 USA * */ #include "ifftw.h" /* "Plan: To bother about the best method of accomplishing an accidental result." (Ambrose Bierce, The Enlarged Devil's Dictionary). */ plan *X(mkplan)(size_t size, const plan_adt *adt) { plan *p = (plan *)MALLOC(size, PLANS); A(adt->destroy); p->adt = adt; X(ops_zero)(&p->ops); p->pcost = 0.0; p->wakefulness = SLEEPY; p->could_prune_now_p = 0; return p; } /* * destroy a plan */ void X(plan_destroy_internal)(plan *ego) { if (ego) { A(ego->wakefulness == SLEEPY); ego->adt->destroy(ego); X(ifree)(ego); } } /* dummy destroy routine for plans with no local state */ void X(plan_null_destroy)(plan *ego) { UNUSED(ego); /* nothing */ } void X(plan_awake)(plan *ego, enum wakefulness wakefulness) { if (ego) { A(((wakefulness == SLEEPY) ^ (ego->wakefulness == SLEEPY))); ego->adt->awake(ego, wakefulness); ego->wakefulness = wakefulness; } } fftw-3.3.3/kernel/planner.c0000644000175000017500000006236612054410050012476 00000000000000/* * Copyright (c) 2000 Matteo Frigo * Copyright (c) 2000 Massachusetts Institute of Technology * * This program is free software; you can redistribute it and/or modify * it under the terms of the GNU General Public License as published by * the Free Software Foundation; either version 2 of the License, or * (at your option) any later version. * * This program is distributed in the hope that it will be useful, * but WITHOUT ANY WARRANTY; without even the implied warranty of * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the * GNU General Public License for more details. * * You should have received a copy of the GNU General Public License * along with this program; if not, write to the Free Software * Foundation, Inc., 51 Franklin Street, Fifth Floor, Boston, MA 02110-1301 USA * */ #include "ifftw.h" #include /* GNU Coding Standards, Sec. 5.2: "Please write the comments in a GNU program in English, because English is the one language that nearly all programmers in all countries can read." ingemisco tanquam reus culpa rubet vultus meus supplicanti parce [rms] */ #define VALIDP(solution) ((solution)->flags.hash_info & H_VALID) #define LIVEP(solution) ((solution)->flags.hash_info & H_LIVE) #define SLVNDX(solution) ((solution)->flags.slvndx) #define BLISS(flags) (((flags).hash_info) & BLESSING) #define INFEASIBLE_SLVNDX ((1U<timelimit_impatience == 0); return (LEQ(a->u, b->u) && LEQ(b->l, a->l)); } else { return (LEQ(a->l, b->l) && a->timelimit_impatience <= b->timelimit_impatience); } } static unsigned addmod(unsigned a, unsigned b, unsigned p) { /* gcc-2.95/sparc produces incorrect code for the fast version below. */ #if defined(__sparc__) && defined(__GNUC__) /* slow version */ return (a + b) % p; #else /* faster version */ unsigned c = a + b; return c >= p ? c - p : c; #endif } /* slvdesc management: */ static void sgrow(planner *ego) { unsigned osiz = ego->slvdescsiz, nsiz = 1 + osiz + osiz / 4; slvdesc *ntab = (slvdesc *)MALLOC(nsiz * sizeof(slvdesc), SLVDESCS); slvdesc *otab = ego->slvdescs; unsigned i; ego->slvdescs = ntab; ego->slvdescsiz = nsiz; for (i = 0; i < osiz; ++i) ntab[i] = otab[i]; X(ifree0)(otab); } static void register_solver(planner *ego, solver *s) { slvdesc *n; int kind; if (s) { /* add s to solver list */ X(solver_use)(s); A(ego->nslvdesc < INFEASIBLE_SLVNDX); if (ego->nslvdesc >= ego->slvdescsiz) sgrow(ego); n = ego->slvdescs + ego->nslvdesc; n->slv = s; n->reg_nam = ego->cur_reg_nam; n->reg_id = ego->cur_reg_id++; A(strlen(n->reg_nam) < MAXNAM); n->nam_hash = X(hash)(n->reg_nam); kind = s->adt->problem_kind; n->next_for_same_problem_kind = ego->slvdescs_for_problem_kind[kind]; ego->slvdescs_for_problem_kind[kind] = ego->nslvdesc; ego->nslvdesc++; } } static unsigned slookup(planner *ego, char *nam, int id) { unsigned h = X(hash)(nam); /* used to avoid strcmp in the common case */ FORALL_SOLVERS(ego, s, sp, { UNUSED(s); if (sp->reg_id == id && sp->nam_hash == h && !strcmp(sp->reg_nam, nam)) return sp - ego->slvdescs; }); return INFEASIBLE_SLVNDX; } /* Compute a MD5 hash of the configuration of the planner. We store it into the wisdom file to make absolutely sure that we are reading wisdom that is applicable */ static void signature_of_configuration(md5 *m, planner *ego) { X(md5begin)(m); X(md5unsigned)(m, sizeof(R)); /* so we don't mix different precisions */ FORALL_SOLVERS(ego, s, sp, { UNUSED(s); X(md5int)(m, sp->reg_id); X(md5puts)(m, sp->reg_nam); }); X(md5end)(m); } /* md5-related stuff: */ /* first hash function */ static unsigned h1(const hashtab *ht, const md5sig s) { unsigned h = s[0] % ht->hashsiz; A(h == (s[0] % ht->hashsiz)); return h; } /* second hash function (for double hashing) */ static unsigned h2(const hashtab *ht, const md5sig s) { unsigned h = 1U + s[1] % (ht->hashsiz - 1); A(h == (1U + s[1] % (ht->hashsiz - 1))); return h; } static void md5hash(md5 *m, const problem *p, const planner *plnr) { X(md5begin)(m); X(md5unsigned)(m, sizeof(R)); /* so we don't mix different precisions */ X(md5int)(m, plnr->nthr); p->adt->hash(p, m); X(md5end)(m); } static int md5eq(const md5sig a, const md5sig b) { return a[0] == b[0] && a[1] == b[1] && a[2] == b[2] && a[3] == b[3]; } static void sigcpy(const md5sig a, md5sig b) { b[0] = a[0]; b[1] = a[1]; b[2] = a[2]; b[3] = a[3]; } /* memoization routines : */ /* liber scriptus proferetur in quo totum continetur unde mundus iudicetur */ struct solution_s { md5sig s; flags_t flags; }; static solution *htab_lookup(hashtab *ht, const md5sig s, const flags_t *flagsp) { unsigned g, h = h1(ht, s), d = h2(ht, s); solution *best = 0; ++ht->lookup; /* search all entries that match; select the one with the lowest flags.u */ /* This loop may potentially traverse the whole table, since at least one element is guaranteed to be !LIVEP, but all elements may be VALIDP. Hence, we stop after at the first invalid element or after traversing the whole table. */ g = h; do { solution *l = ht->solutions + g; ++ht->lookup_iter; if (VALIDP(l)) { if (LIVEP(l) && md5eq(s, l->s) && subsumes(&l->flags, SLVNDX(l), flagsp) ) { if (!best || LEQ(l->flags.u, best->flags.u)) best = l; } } else break; g = addmod(g, d, ht->hashsiz); } while (g != h); if (best) ++ht->succ_lookup; return best; } static solution *hlookup(planner *ego, const md5sig s, const flags_t *flagsp) { solution *sol = htab_lookup(&ego->htab_blessed, s, flagsp); if (!sol) sol = htab_lookup(&ego->htab_unblessed, s, flagsp); return sol; } static void fill_slot(hashtab *ht, const md5sig s, const flags_t *flagsp, unsigned slvndx, solution *slot) { ++ht->insert; ++ht->nelem; A(!LIVEP(slot)); slot->flags.u = flagsp->u; slot->flags.l = flagsp->l; slot->flags.timelimit_impatience = flagsp->timelimit_impatience; slot->flags.hash_info |= H_VALID | H_LIVE; SLVNDX(slot) = slvndx; /* keep this check enabled in case we add so many solvers that the bitfield overflows */ CK(SLVNDX(slot) == slvndx); sigcpy(s, slot->s); } static void kill_slot(hashtab *ht, solution *slot) { A(LIVEP(slot)); /* ==> */ A(VALIDP(slot)); --ht->nelem; slot->flags.hash_info = H_VALID; } static void hinsert0(hashtab *ht, const md5sig s, const flags_t *flagsp, unsigned slvndx) { solution *l; unsigned g, h = h1(ht, s), d = h2(ht, s); ++ht->insert_unknown; /* search for nonfull slot */ for (g = h; ; g = addmod(g, d, ht->hashsiz)) { ++ht->insert_iter; l = ht->solutions + g; if (!LIVEP(l)) break; A((g + d) % ht->hashsiz != h); } fill_slot(ht, s, flagsp, slvndx, l); } static void rehash(hashtab *ht, unsigned nsiz) { unsigned osiz = ht->hashsiz, h; solution *osol = ht->solutions, *nsol; nsiz = (unsigned)X(next_prime)((INT)nsiz); nsol = (solution *)MALLOC(nsiz * sizeof(solution), HASHT); ++ht->nrehash; /* init new table */ for (h = 0; h < nsiz; ++h) nsol[h].flags.hash_info = 0; /* install new table */ ht->hashsiz = nsiz; ht->solutions = nsol; ht->nelem = 0; /* copy table */ for (h = 0; h < osiz; ++h) { solution *l = osol + h; if (LIVEP(l)) hinsert0(ht, l->s, &l->flags, SLVNDX(l)); } X(ifree0)(osol); } static unsigned minsz(unsigned nelem) { return 1U + nelem + nelem / 8U; } static unsigned nextsz(unsigned nelem) { return minsz(minsz(nelem)); } static void hgrow(hashtab *ht) { unsigned nelem = ht->nelem; if (minsz(nelem) >= ht->hashsiz) rehash(ht, nextsz(nelem)); } #if 0 /* shrink the hash table, never used */ static void hshrink(hashtab *ht) { unsigned nelem = ht->nelem; /* always rehash after deletions */ rehash(ht, nextsz(nelem)); } #endif static void htab_insert(hashtab *ht, const md5sig s, const flags_t *flagsp, unsigned slvndx) { unsigned g, h = h1(ht, s), d = h2(ht, s); solution *first = 0; /* Remove all entries that are subsumed by the new one. */ /* This loop may potentially traverse the whole table, since at least one element is guaranteed to be !LIVEP, but all elements may be VALIDP. Hence, we stop after at the first invalid element or after traversing the whole table. */ g = h; do { solution *l = ht->solutions + g; ++ht->insert_iter; if (VALIDP(l)) { if (LIVEP(l) && md5eq(s, l->s)) { if (subsumes(flagsp, slvndx, &l->flags)) { if (!first) first = l; kill_slot(ht, l); } else { /* It is an error to insert an element that is subsumed by an existing entry. */ A(!subsumes(&l->flags, SLVNDX(l), flagsp)); } } } else break; g = addmod(g, d, ht->hashsiz); } while (g != h); if (first) { /* overwrite FIRST */ fill_slot(ht, s, flagsp, slvndx, first); } else { /* create a new entry */ hgrow(ht); hinsert0(ht, s, flagsp, slvndx); } } static void hinsert(planner *ego, const md5sig s, const flags_t *flagsp, unsigned slvndx) { htab_insert(BLISS(*flagsp) ? &ego->htab_blessed : &ego->htab_unblessed, s, flagsp, slvndx ); } static void invoke_hook(planner *ego, plan *pln, const problem *p, int optimalp) { if (ego->hook) ego->hook(ego, pln, p, optimalp); } #ifdef FFTW_RANDOM_ESTIMATOR /* a "random" estimate, used for debugging to generate "random" plans, albeit from a deterministic seed. */ unsigned X(random_estimate_seed) = 0; static double random_estimate(const planner *ego, const plan *pln, const problem *p) { md5 m; X(md5begin)(&m); X(md5unsigned)(&m, X(random_estimate_seed)); X(md5int)(&m, ego->nthr); p->adt->hash(p, &m); X(md5putb)(&m, &pln->ops, sizeof(pln->ops)); X(md5putb)(&m, &pln->adt, sizeof(pln->adt)); X(md5end)(&m); return ego->cost_hook ? ego->cost_hook(p, m.s[0], COST_MAX) : m.s[0]; } #endif double X(iestimate_cost)(const planner *ego, const plan *pln, const problem *p) { double cost = + pln->ops.add + pln->ops.mul #if HAVE_FMA + pln->ops.fma #else + 2 * pln->ops.fma #endif + pln->ops.other; if (ego->cost_hook) cost = ego->cost_hook(p, cost, COST_MAX); return cost; } static void evaluate_plan(planner *ego, plan *pln, const problem *p) { if (ESTIMATEP(ego) || !BELIEVE_PCOSTP(ego) || pln->pcost == 0.0) { ego->nplan++; if (ESTIMATEP(ego)) { estimate: /* heuristic */ #ifdef FFTW_RANDOM_ESTIMATOR pln->pcost = random_estimate(ego, pln, p); ego->epcost += X(iestimate_cost)(ego, pln, p); #else pln->pcost = X(iestimate_cost)(ego, pln, p); ego->epcost += pln->pcost; #endif } else { double t = X(measure_execution_time)(ego, pln, p); if (t < 0) { /* unavailable cycle counter */ /* Real programmers can write FORTRAN in any language */ goto estimate; } pln->pcost = t; ego->pcost += t; ego->need_timeout_check = 1; } } invoke_hook(ego, pln, p, 0); } /* maintain dynamic scoping of flags, nthr: */ static plan *invoke_solver(planner *ego, const problem *p, solver *s, const flags_t *nflags) { flags_t flags = ego->flags; int nthr = ego->nthr; plan *pln; ego->flags = *nflags; PLNR_TIMELIMIT_IMPATIENCE(ego) = 0; A(p->adt->problem_kind == s->adt->problem_kind); pln = s->adt->mkplan(s, p, ego); ego->nthr = nthr; ego->flags = flags; return pln; } /* maintain the invariant TIMED_OUT ==> NEED_TIMEOUT_CHECK */ static int timeout_p(planner *ego, const problem *p) { /* do not timeout when estimating. First, the estimator is the planner of last resort. Second, calling X(elapsed_since)() is slower than estimating */ if (!ESTIMATEP(ego)) { /* do not assume that X(elapsed_since)() is monotonic */ if (ego->timed_out) { A(ego->need_timeout_check); return 1; } if (ego->timelimit >= 0 && X(elapsed_since)(ego, p, ego->start_time) >= ego->timelimit) { ego->timed_out = 1; ego->need_timeout_check = 1; return 1; } } A(!ego->timed_out); ego->need_timeout_check = 0; return 0; } static plan *search0(planner *ego, const problem *p, unsigned *slvndx, const flags_t *flagsp) { plan *best = 0; int best_not_yet_timed = 1; /* Do not start a search if the planner timed out. This check is necessary, lest the relaxation mechanism kick in */ if (timeout_p(ego, p)) return 0; FORALL_SOLVERS_OF_KIND(p->adt->problem_kind, ego, s, sp, { plan *pln; pln = invoke_solver(ego, p, s, flagsp); if (ego->need_timeout_check) if (timeout_p(ego, p)) { X(plan_destroy_internal)(pln); X(plan_destroy_internal)(best); return 0; } if (pln) { /* read COULD_PRUNE_NOW_P because PLN may be destroyed before we use COULD_PRUNE_NOW_P */ int could_prune_now_p = pln->could_prune_now_p; if (best) { if (best_not_yet_timed) { evaluate_plan(ego, best, p); best_not_yet_timed = 0; } evaluate_plan(ego, pln, p); if (pln->pcost < best->pcost) { X(plan_destroy_internal)(best); best = pln; *slvndx = sp - ego->slvdescs; } else { X(plan_destroy_internal)(pln); } } else { best = pln; *slvndx = sp - ego->slvdescs; } if (ALLOW_PRUNINGP(ego) && could_prune_now_p) break; } }); return best; } static plan *search(planner *ego, const problem *p, unsigned *slvndx, flags_t *flagsp) { plan *pln = 0; unsigned i; /* relax impatience in this order: */ static const unsigned relax_tab[] = { 0, /* relax nothing */ NO_VRECURSE, NO_FIXED_RADIX_LARGE_N, NO_SLOW, NO_UGLY }; unsigned l_orig = flagsp->l; unsigned x = flagsp->u; /* guaranteed to be different from X */ unsigned last_x = ~x; for (i = 0; i < sizeof(relax_tab) / sizeof(relax_tab[0]); ++i) { if (LEQ(l_orig, x & ~relax_tab[i])) x = x & ~relax_tab[i]; if (x != last_x) { last_x = x; flagsp->l = x; pln = search0(ego, p, slvndx, flagsp); if (pln) break; } } if (!pln) { /* search [L_ORIG, U] */ if (l_orig != last_x) { last_x = l_orig; flagsp->l = l_orig; pln = search0(ego, p, slvndx, flagsp); } } return pln; } #define CHECK_FOR_BOGOSITY \ if ((ego->bogosity_hook ? \ (ego->wisdom_state = ego->bogosity_hook(ego->wisdom_state, p)) \ : ego->wisdom_state) == WISDOM_IS_BOGUS) \ goto wisdom_is_bogus; static plan *mkplan(planner *ego, const problem *p) { plan *pln; md5 m; unsigned slvndx; flags_t flags_of_solution; solution *sol; solver *s; ASSERT_ALIGNED_DOUBLE; A(LEQ(PLNR_L(ego), PLNR_U(ego))); if (ESTIMATEP(ego)) PLNR_TIMELIMIT_IMPATIENCE(ego) = 0; /* canonical form */ #ifdef FFTW_DEBUG check(&ego->htab_blessed); check(&ego->htab_unblessed); #endif pln = 0; CHECK_FOR_BOGOSITY; ego->timed_out = 0; ++ego->nprob; md5hash(&m, p, ego); flags_of_solution = ego->flags; if (ego->wisdom_state != WISDOM_IGNORE_ALL) { if ((sol = hlookup(ego, m.s, &flags_of_solution))) { /* wisdom is acceptable */ wisdom_state_t owisdom_state = ego->wisdom_state; /* this hook is mainly for MPI, to make sure that wisdom is in sync across all processes for MPI problems */ if (ego->wisdom_ok_hook && !ego->wisdom_ok_hook(p, sol->flags)) goto do_search; /* ignore not-ok wisdom */ slvndx = SLVNDX(sol); if (slvndx == INFEASIBLE_SLVNDX) { if (ego->wisdom_state == WISDOM_IGNORE_INFEASIBLE) goto do_search; else return 0; /* known to be infeasible */ } flags_of_solution = sol->flags; /* inherit blessing either from wisdom or from the planner */ flags_of_solution.hash_info |= BLISS(ego->flags); ego->wisdom_state = WISDOM_ONLY; s = ego->slvdescs[slvndx].slv; if (p->adt->problem_kind != s->adt->problem_kind) goto wisdom_is_bogus; pln = invoke_solver(ego, p, s, &flags_of_solution); CHECK_FOR_BOGOSITY; /* catch error in child solvers */ sol = 0; /* Paranoia: SOL may be dangling after invoke_solver(); make sure we don't accidentally reuse it. */ if (!pln) goto wisdom_is_bogus; ego->wisdom_state = owisdom_state; goto skip_search; } else if (ego->nowisdom_hook) /* for MPI, make sure lack of wisdom */ ego->nowisdom_hook(p); /* is in sync across all processes */ } do_search: /* cannot search in WISDOM_ONLY mode */ if (ego->wisdom_state == WISDOM_ONLY) goto wisdom_is_bogus; flags_of_solution = ego->flags; pln = search(ego, p, &slvndx, &flags_of_solution); CHECK_FOR_BOGOSITY; /* catch error in child solvers */ if (ego->timed_out) { A(!pln); if (PLNR_TIMELIMIT_IMPATIENCE(ego) != 0) { /* record (below) that this plan has failed because of timeout */ flags_of_solution.hash_info |= BLESSING; } else { /* this is not the top-level problem or timeout is not active: record no wisdom. */ return 0; } } else { /* canonicalize to infinite timeout */ flags_of_solution.timelimit_impatience = 0; } skip_search: if (ego->wisdom_state == WISDOM_NORMAL || ego->wisdom_state == WISDOM_ONLY) { if (pln) { hinsert(ego, m.s, &flags_of_solution, slvndx); invoke_hook(ego, pln, p, 1); } else { hinsert(ego, m.s, &flags_of_solution, INFEASIBLE_SLVNDX); } } return pln; wisdom_is_bogus: X(plan_destroy_internal)(pln); ego->wisdom_state = WISDOM_IS_BOGUS; return 0; } static void htab_destroy(hashtab *ht) { X(ifree)(ht->solutions); ht->solutions = 0; ht->nelem = 0U; } static void mkhashtab(hashtab *ht) { ht->nrehash = 0; ht->succ_lookup = ht->lookup = ht->lookup_iter = 0; ht->insert = ht->insert_iter = ht->insert_unknown = 0; ht->solutions = 0; ht->hashsiz = ht->nelem = 0U; hgrow(ht); /* so that hashsiz > 0 */ } /* destroy hash table entries. If FORGET_EVERYTHING, destroy the whole table. If FORGET_ACCURSED, then destroy entries that are not blessed. */ static void forget(planner *ego, amnesia a) { switch (a) { case FORGET_EVERYTHING: htab_destroy(&ego->htab_blessed); mkhashtab(&ego->htab_blessed); /* fall through */ case FORGET_ACCURSED: htab_destroy(&ego->htab_unblessed); mkhashtab(&ego->htab_unblessed); break; default: break; } } /* FIXME: what sort of version information should we write? */ #define WISDOM_PREAMBLE PACKAGE "-" VERSION " " STRINGIZE(X(wisdom)) static const char stimeout[] = "TIMEOUT"; /* tantus labor non sit cassus */ static void exprt(planner *ego, printer *p) { unsigned h; hashtab *ht = &ego->htab_blessed; md5 m; signature_of_configuration(&m, ego); p->print(p, "(" WISDOM_PREAMBLE " #x%M #x%M #x%M #x%M\n", m.s[0], m.s[1], m.s[2], m.s[3]); for (h = 0; h < ht->hashsiz; ++h) { solution *l = ht->solutions + h; if (LIVEP(l)) { const char *reg_nam; int reg_id; if (SLVNDX(l) == INFEASIBLE_SLVNDX) { reg_nam = stimeout; reg_id = 0; } else { slvdesc *sp = ego->slvdescs + SLVNDX(l); reg_nam = sp->reg_nam; reg_id = sp->reg_id; } /* qui salvandos salvas gratis salva me fons pietatis */ p->print(p, " (%s %d #x%x #x%x #x%x #x%M #x%M #x%M #x%M)\n", reg_nam, reg_id, l->flags.l, l->flags.u, l->flags.timelimit_impatience, l->s[0], l->s[1], l->s[2], l->s[3]); } } p->print(p, ")\n"); } /* mors stupebit et natura cum resurget creatura */ static int imprt(planner *ego, scanner *sc) { char buf[MAXNAM + 1]; md5uint sig[4]; unsigned l, u, timelimit_impatience; flags_t flags; int reg_id; unsigned slvndx; hashtab *ht = &ego->htab_blessed; hashtab old; md5 m; if (!sc->scan(sc, "(" WISDOM_PREAMBLE " #x%M #x%M #x%M #x%M\n", sig + 0, sig + 1, sig + 2, sig + 3)) return 0; /* don't need to restore hashtable */ signature_of_configuration(&m, ego); if (m.s[0] != sig[0] || m.s[1] != sig[1] || m.s[2] != sig[2] || m.s[3] != sig[3]) { /* invalid configuration */ return 0; } /* make a backup copy of the hash table (cache the hash) */ { unsigned h, hsiz = ht->hashsiz; old = *ht; old.solutions = (solution *)MALLOC(hsiz * sizeof(solution), HASHT); for (h = 0; h < hsiz; ++h) old.solutions[h] = ht->solutions[h]; } while (1) { if (sc->scan(sc, ")")) break; /* qua resurget ex favilla */ if (!sc->scan(sc, "(%*s %d #x%x #x%x #x%x #x%M #x%M #x%M #x%M)", MAXNAM, buf, ®_id, &l, &u, &timelimit_impatience, sig + 0, sig + 1, sig + 2, sig + 3)) goto bad; if (!strcmp(buf, stimeout) && reg_id == 0) { slvndx = INFEASIBLE_SLVNDX; } else { if (timelimit_impatience != 0) goto bad; slvndx = slookup(ego, buf, reg_id); if (slvndx == INFEASIBLE_SLVNDX) goto bad; } /* inter oves locum praesta */ flags.l = l; flags.u = u; flags.timelimit_impatience = timelimit_impatience; flags.hash_info = BLESSING; CK(flags.l == l); CK(flags.u == u); CK(flags.timelimit_impatience == timelimit_impatience); if (!hlookup(ego, sig, &flags)) hinsert(ego, sig, &flags, slvndx); } X(ifree0)(old.solutions); return 1; bad: /* ``The wisdom of FFTW must be above suspicion.'' */ X(ifree0)(ht->solutions); *ht = old; return 0; } /* * create a planner */ planner *X(mkplanner)(void) { int i; static const planner_adt padt = { register_solver, mkplan, forget, exprt, imprt }; planner *p = (planner *) MALLOC(sizeof(planner), PLANNERS); p->adt = &padt; p->nplan = p->nprob = 0; p->pcost = p->epcost = 0.0; p->hook = 0; p->cost_hook = 0; p->wisdom_ok_hook = 0; p->nowisdom_hook = 0; p->bogosity_hook = 0; p->cur_reg_nam = 0; p->wisdom_state = WISDOM_NORMAL; p->slvdescs = 0; p->nslvdesc = p->slvdescsiz = 0; p->flags.l = 0; p->flags.u = 0; p->flags.timelimit_impatience = 0; p->flags.hash_info = 0; p->nthr = 1; p->need_timeout_check = 1; p->timelimit = -1; mkhashtab(&p->htab_blessed); mkhashtab(&p->htab_unblessed); for (i = 0; i < PROBLEM_LAST; ++i) p->slvdescs_for_problem_kind[i] = -1; return p; } void X(planner_destroy)(planner *ego) { /* destroy hash table */ htab_destroy(&ego->htab_blessed); htab_destroy(&ego->htab_unblessed); /* destroy solvdesc table */ FORALL_SOLVERS(ego, s, sp, { UNUSED(sp); X(solver_destroy)(s); }); X(ifree0)(ego->slvdescs); X(ifree)(ego); /* dona eis requiem */ } plan *X(mkplan_d)(planner *ego, problem *p) { plan *pln = ego->adt->mkplan(ego, p); X(problem_destroy)(p); return pln; } /* like X(mkplan_d), but sets/resets flags as well */ plan *X(mkplan_f_d)(planner *ego, problem *p, unsigned l_set, unsigned u_set, unsigned u_reset) { flags_t oflags = ego->flags; plan *pln; PLNR_U(ego) &= ~u_reset; PLNR_L(ego) &= ~u_reset; PLNR_L(ego) |= l_set; PLNR_U(ego) |= u_set | l_set; pln = X(mkplan_d)(ego, p); ego->flags = oflags; return pln; } /* * Debugging code: */ #ifdef FFTW_DEBUG static void check(hashtab *ht) { unsigned live = 0; unsigned i; A(ht->nelem < ht->hashsiz); for (i = 0; i < ht->hashsiz; ++i) { solution *l = ht->solutions + i; if (LIVEP(l)) ++live; } A(ht->nelem == live); for (i = 0; i < ht->hashsiz; ++i) { solution *l1 = ht->solutions + i; int foundit = 0; if (LIVEP(l1)) { unsigned g, h = h1(ht, l1->s), d = h2(ht, l1->s); g = h; do { solution *l = ht->solutions + g; if (VALIDP(l)) { if (l1 == l) foundit = 1; else if (LIVEP(l) && md5eq(l1->s, l->s)) { A(!subsumes(&l->flags, SLVNDX(l), &l1->flags)); A(!subsumes(&l1->flags, SLVNDX(l1), &l->flags)); } } else break; g = addmod(g, d, ht->hashsiz); } while (g != h); A(foundit); } } } #endif fftw-3.3.3/kernel/primes.c0000644000175000017500000001157112054410050012326 00000000000000/* * Copyright (c) 2003, 2007-11 Matteo Frigo * Copyright (c) 2003, 2007-11 Massachusetts Institute of Technology * * This program is free software; you can redistribute it and/or modify * it under the terms of the GNU General Public License as published by * the Free Software Foundation; either version 2 of the License, or * (at your option) any later version. * * This program is distributed in the hope that it will be useful, * but WITHOUT ANY WARRANTY; without even the implied warranty of * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the * GNU General Public License for more details. * * You should have received a copy of the GNU General Public License * along with this program; if not, write to the Free Software * Foundation, Inc., 51 Franklin Street, Fifth Floor, Boston, MA 02110-1301 USA * */ #include "ifftw.h" /***************************************************************************/ /* Rader's algorithm requires lots of modular arithmetic, and if we aren't careful we can have errors due to integer overflows. */ /* Compute (x * y) mod p, but watch out for integer overflows; we must have 0 <= {x, y} < p. If overflow is common, this routine is somewhat slower than e.g. using 'long long' arithmetic. However, it has the advantage of working when INT is 64 bits, and is also faster when overflow is rare. FFTW calls this via the MULMOD macro, which further optimizes for the case of small integers. */ #define ADD_MOD(x, y, p) ((x) >= (p) - (y)) ? ((x) + ((y) - (p))) : ((x) + (y)) INT X(safe_mulmod)(INT x, INT y, INT p) { INT r; if (y > x) return X(safe_mulmod)(y, x, p); A(0 <= y && x < p); r = 0; while (y) { r = ADD_MOD(r, x*(y&1), p); y >>= 1; x = ADD_MOD(x, x, p); } return r; } /***************************************************************************/ /* Compute n^m mod p, where m >= 0 and p > 0. If we really cared, we could make this tail-recursive. */ INT X(power_mod)(INT n, INT m, INT p) { A(p > 0); if (m == 0) return 1; else if (m % 2 == 0) { INT x = X(power_mod)(n, m / 2, p); return MULMOD(x, x, p); } else return MULMOD(n, X(power_mod)(n, m - 1, p), p); } /* the following two routines were contributed by Greg Dionne. */ static INT get_prime_factors(INT n, INT *primef) { INT i; INT size = 0; A(n % 2 == 0); /* this routine is designed only for even n */ primef[size++] = (INT)2; do n >>= 1; while ((n & 1) == 0); if (n == 1) return size; for (i = 3; i * i <= n; i += 2) if (!(n % i)) { primef[size++] = i; do n /= i; while (!(n % i)); } if (n == 1) return size; primef[size++] = n; return size; } INT X(find_generator)(INT p) { INT n, i, size; INT primef[16]; /* smallest number = 32589158477190044730 > 2^64 */ INT pm1 = p - 1; if (p == 2) return 1; size = get_prime_factors(pm1, primef); n = 2; for (i = 0; i < size; i++) if (X(power_mod)(n, pm1 / primef[i], p) == 1) { i = -1; n++; } return n; } /* Return first prime divisor of n (It would be at best slightly faster to search a static table of primes; there are 6542 primes < 2^16.) */ INT X(first_divisor)(INT n) { INT i; if (n <= 1) return n; if (n % 2 == 0) return 2; for (i = 3; i*i <= n; i += 2) if (n % i == 0) return i; return n; } int X(is_prime)(INT n) { return(n > 1 && X(first_divisor)(n) == n); } INT X(next_prime)(INT n) { while (!X(is_prime)(n)) ++n; return n; } int X(factors_into)(INT n, const INT *primes) { for (; *primes != 0; ++primes) while ((n % *primes) == 0) n /= *primes; return (n == 1); } /* integer square root. Return floor(sqrt(N)) */ INT X(isqrt)(INT n) { INT guess, iguess; A(n >= 0); if (n == 0) return 0; guess = n; iguess = 1; do { guess = (guess + iguess) / 2; iguess = n / guess; } while (guess > iguess); return guess; } static INT isqrt_maybe(INT n) { INT guess = X(isqrt)(n); return guess * guess == n ? guess : 0; } #define divides(a, b) (((b) % (a)) == 0) INT X(choose_radix)(INT r, INT n) { if (r > 0) { if (divides(r, n)) return r; return 0; } else if (r == 0) { return X(first_divisor)(n); } else { /* r is negative. If n = (-r) * q^2, take q as the radix */ r = 0 - r; return (n > r && divides(r, n)) ? isqrt_maybe(n / r) : 0; } } /* return A mod N, works for all A including A < 0 */ INT X(modulo)(INT a, INT n) { A(n > 0); if (a >= 0) return a % n; else return (n - 1) - ((-(a + (INT)1)) % n); } /* TRUE if N factors into small primes */ int X(factors_into_small_primes)(INT n) { static const INT primes[] = { 2, 3, 5, 0 }; return X(factors_into)(n, primes); } fftw-3.3.3/kernel/print.c0000644000175000017500000001201612054410050012156 00000000000000/* * Copyright (c) 2003, 2007-11 Matteo Frigo * Copyright (c) 2003, 2007-11 Massachusetts Institute of Technology * * This program is free software; you can redistribute it and/or modify * it under the terms of the GNU General Public License as published by * the Free Software Foundation; either version 2 of the License, or * (at your option) any later version. * * This program is distributed in the hope that it will be useful, * but WITHOUT ANY WARRANTY; without even the implied warranty of * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the * GNU General Public License for more details. * * You should have received a copy of the GNU General Public License * along with this program; if not, write to the Free Software * Foundation, Inc., 51 Franklin Street, Fifth Floor, Boston, MA 02110-1301 USA * */ #include "ifftw.h" #include #include #include #define BSZ 64 static void myputs(printer *p, const char *s) { char c; while ((c = *s++)) p->putchr(p, c); } static void newline(printer *p) { int i; p->putchr(p, '\n'); for (i = 0; i < p->indent; ++i) p->putchr(p, ' '); } static const char *digits = "0123456789abcdef"; static void putint(printer *p, INT i) { char buf[BSZ]; char *f = buf; if (i < 0) { p->putchr(p, '-'); i = -i; } do { *f++ = digits[i % 10]; i /= 10; } while (i); do { p->putchr(p, *--f); } while (f != buf); } static void putulong(printer *p, unsigned long i, int base, int width) { char buf[BSZ]; char *f = buf; do { *f++ = digits[i % base]; i /= base; } while (i); while (width > f - buf) { p->putchr(p, '0'); --width; } do { p->putchr(p, *--f); } while (f != buf); } static void vprint(printer *p, const char *format, va_list ap) { const char *s = format; char c; INT ival; while ((c = *s++)) { switch (c) { case '%': switch ((c = *s++)) { case 'M': { /* md5 value */ md5uint x = va_arg(ap, md5uint); putulong(p, (unsigned long)(0xffffffffUL & x), 16, 8); break; } case 'c': { int x = va_arg(ap, int); p->putchr(p, x); break; } case 's': { char *x = va_arg(ap, char *); if (x) myputs(p, x); else goto putnull; break; } case 'd': { int x = va_arg(ap, int); ival = (INT)x; goto putival; } case 'D': { ival = va_arg(ap, INT); goto putival; } case 'v': { /* print optional vector length */ ival = va_arg(ap, INT); if (ival > 1) { myputs(p, "-x"); goto putival; } break; } case 'o': { /* integer option. Usage: %oNAME= */ ival = va_arg(ap, INT); if (ival) p->putchr(p, '/'); while ((c = *s++) != '=') if (ival) p->putchr(p, c); if (ival) { p->putchr(p, '='); goto putival; } break; } case 'u': { unsigned x = va_arg(ap, unsigned); putulong(p, (unsigned long)x, 10, 0); break; } case 'x': { unsigned x = va_arg(ap, unsigned); putulong(p, (unsigned long)x, 16, 0); break; } case '(': { /* newline, augment indent level */ p->indent += p->indent_incr; newline(p); break; } case ')': { /* decrement indent level */ p->indent -= p->indent_incr; break; } case 'p': { /* note difference from C's %p */ /* print plan */ plan *x = va_arg(ap, plan *); if (x) x->adt->print(x, p); else goto putnull; break; } case 'P': { /* print problem */ problem *x = va_arg(ap, problem *); if (x) x->adt->print(x, p); else goto putnull; break; } case 'T': { /* print tensor */ tensor *x = va_arg(ap, tensor *); if (x) X(tensor_print)(x, p); else goto putnull; break; } default: A(0 /* unknown format */); break; putnull: myputs(p, "(null)"); break; putival: putint(p, ival); break; } break; default: p->putchr(p, c); break; } } } static void print(printer *p, const char *format, ...) { va_list ap; va_start(ap, format); vprint(p, format, ap); va_end(ap); } printer *X(mkprinter)(size_t size, void (*putchr)(printer *p, char c), void (*cleanup)(printer *p)) { printer *s = (printer *)MALLOC(size, OTHER); s->print = print; s->vprint = vprint; s->putchr = putchr; s->cleanup = cleanup; s->indent = 0; s->indent_incr = 2; return s; } void X(printer_destroy)(printer *p) { if (p->cleanup) p->cleanup(p); X(ifree)(p); } fftw-3.3.3/kernel/problem.c0000644000175000017500000000352012054410050012462 00000000000000/* * Copyright (c) 2003, 2007-11 Matteo Frigo * Copyright (c) 2003, 2007-11 Massachusetts Institute of Technology * * This program is free software; you can redistribute it and/or modify * it under the terms of the GNU General Public License as published by * the Free Software Foundation; either version 2 of the License, or * (at your option) any later version. * * This program is distributed in the hope that it will be useful, * but WITHOUT ANY WARRANTY; without even the implied warranty of * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the * GNU General Public License for more details. * * You should have received a copy of the GNU General Public License * along with this program; if not, write to the Free Software * Foundation, Inc., 51 Franklin Street, Fifth Floor, Boston, MA 02110-1301 USA * */ #include "ifftw.h" /* constructor */ problem *X(mkproblem)(size_t sz, const problem_adt *adt) { problem *p = (problem *)MALLOC(sz, PROBLEMS); p->adt = adt; return p; } /* destructor */ void X(problem_destroy)(problem *ego) { if (ego) ego->adt->destroy(ego); } /* management of unsolvable problems */ static void unsolvable_destroy(problem *ego) { UNUSED(ego); } static void unsolvable_hash(const problem *p, md5 *m) { UNUSED(p); X(md5puts)(m, "unsolvable"); } static void unsolvable_print(const problem *ego, printer *p) { UNUSED(ego); p->print(p, "(unsolvable)"); } static void unsolvable_zero(const problem *ego) { UNUSED(ego); } static const problem_adt padt = { PROBLEM_UNSOLVABLE, unsolvable_hash, unsolvable_zero, unsolvable_print, unsolvable_destroy }; /* there is no point in malloc'ing this one */ static problem the_unsolvable_problem = { &padt }; problem *X(mkproblem_unsolvable)(void) { return &the_unsolvable_problem; } fftw-3.3.3/kernel/rader.c0000644000175000017500000000336412054410050012125 00000000000000/* * Copyright (c) 2003, 2007-11 Matteo Frigo * Copyright (c) 2003, 2007-11 Massachusetts Institute of Technology * * This program is free software; you can redistribute it and/or modify * it under the terms of the GNU General Public License as published by * the Free Software Foundation; either version 2 of the License, or * (at your option) any later version. * * This program is distributed in the hope that it will be useful, * but WITHOUT ANY WARRANTY; without even the implied warranty of * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the * GNU General Public License for more details. * * You should have received a copy of the GNU General Public License * along with this program; if not, write to the Free Software * Foundation, Inc., 51 Franklin Street, Fifth Floor, Boston, MA 02110-1301 USA * */ #include "ifftw.h" /* common routines for Rader solvers */ /* shared twiddle and omega lists, keyed by two/three integers. */ struct rader_tls { INT k1, k2, k3; R *W; int refcnt; rader_tl *cdr; }; void X(rader_tl_insert)(INT k1, INT k2, INT k3, R *W, rader_tl **tl) { rader_tl *t = (rader_tl *) MALLOC(sizeof(rader_tl), TWIDDLES); t->k1 = k1; t->k2 = k2; t->k3 = k3; t->W = W; t->refcnt = 1; t->cdr = *tl; *tl = t; } R *X(rader_tl_find)(INT k1, INT k2, INT k3, rader_tl *t) { while (t && (t->k1 != k1 || t->k2 != k2 || t->k3 != k3)) t = t->cdr; if (t) { ++t->refcnt; return t->W; } else return 0; } void X(rader_tl_delete)(R *W, rader_tl **tl) { if (W) { rader_tl **tp, *t; for (tp = tl; (t = *tp) && t->W != W; tp = &t->cdr) ; if (t && --t->refcnt <= 0) { *tp = t->cdr; X(ifree)(t->W); X(ifree)(t); } } } fftw-3.3.3/kernel/scan.c0000644000175000017500000001142612054410050011752 00000000000000/* * Copyright (c) 2003, 2007-11 Matteo Frigo * Copyright (c) 2003, 2007-11 Massachusetts Institute of Technology * * This program is free software; you can redistribute it and/or modify * it under the terms of the GNU General Public License as published by * the Free Software Foundation; either version 2 of the License, or * (at your option) any later version. * * This program is distributed in the hope that it will be useful, * but WITHOUT ANY WARRANTY; without even the implied warranty of * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the * GNU General Public License for more details. * * You should have received a copy of the GNU General Public License * along with this program; if not, write to the Free Software * Foundation, Inc., 51 Franklin Street, Fifth Floor, Boston, MA 02110-1301 USA * */ #include "ifftw.h" #include #include #include #include #ifdef USE_CTYPE #include #else /* Screw ctype. On linux, the is* functions call a routine that gets the ctype map in the current locale. Because this operation is expensive, the map is cached on a per-thread basis. I am not willing to link this crap with FFTW. Not over my dead body. Sic transit gloria mundi. */ #undef isspace #define isspace(x) ((x) >= 0 && (x) <= ' ') #undef isdigit #define isdigit(x) ((x) >= '0' && (x) <= '9') #undef isupper #define isupper(x) ((x) >= 'A' && (x) <= 'Z') #undef islower #define islower(x) ((x) >= 'a' && (x) <= 'z') #endif static int mygetc(scanner *sc) { if (sc->ungotc != EOF) { int c = sc->ungotc; sc->ungotc = EOF; return c; } return(sc->getchr(sc)); } #define GETCHR(sc) mygetc(sc) static void myungetc(scanner *sc, int c) { sc->ungotc = c; } #define UNGETCHR(sc, c) myungetc(sc, c) static void eat_blanks(scanner *sc) { int ch; while (ch = GETCHR(sc), isspace(ch)) ; UNGETCHR(sc, ch); } static void mygets(scanner *sc, char *s, size_t maxlen) { char *s0 = s; int ch; A(maxlen > 0); while ((ch = GETCHR(sc)) != EOF && !isspace(ch) && ch != ')' && ch != '(' && s < s0 + maxlen) *s++ = ch; *s = 0; UNGETCHR(sc, ch); } static long getlong(scanner *sc, int base, int *ret) { int sign = 1, ch, count; long x = 0; ch = GETCHR(sc); if (ch == '-' || ch == '+') { sign = ch == '-' ? -1 : 1; ch = GETCHR(sc); } for (count = 0; ; ++count) { if (isdigit(ch)) ch -= '0'; else if (isupper(ch)) ch -= 'A' - 10; else if (islower(ch)) ch -= 'a' - 10; else break; x = x * base + ch; ch = GETCHR(sc); } x *= sign; UNGETCHR(sc, ch); *ret = count > 0; return x; } /* vscan is mostly scanf-like, with our additional format specifiers, but with a few twists. It returns simply 0 or 1 indicating whether the match was successful. '(' and ')' in the format string match those characters preceded by any whitespace. Finally, if a character match fails, it will ungetchr() the last character back onto the stream. */ static int vscan(scanner *sc, const char *format, va_list ap) { const char *s = format; char c; int ch = 0; size_t fmt_len; while ((c = *s++)) { fmt_len = 0; switch (c) { case '%': getformat: switch ((c = *s++)) { case 's': { char *x = va_arg(ap, char *); mygets(sc, x, fmt_len); break; } case 'd': { int *x = va_arg(ap, int *); *x = (int) getlong(sc, 10, &ch); if (!ch) return 0; break; } case 'x': { int *x = va_arg(ap, int *); *x = (int) getlong(sc, 16, &ch); if (!ch) return 0; break; } case 'M': { md5uint *x = va_arg(ap, md5uint *); *x = (md5uint) (0xffffffffUL & getlong(sc, 16, &ch)); if (!ch) return 0; break; } case '*': { if ((fmt_len = va_arg(ap, int)) <= 0) return 0; goto getformat; } default: A(0 /* unknown format */); break; } break; default: if (isspace(c) || c == '(' || c == ')') eat_blanks(sc); if (!isspace(c) && (ch = GETCHR(sc)) != c) { UNGETCHR(sc, ch); return 0; } break; } } return 1; } static int scan(scanner *sc, const char *format, ...) { int ret; va_list ap; va_start(ap, format); ret = vscan(sc, format, ap); va_end(ap); return ret; } scanner *X(mkscanner)(size_t size, int (*getchr)(scanner *sc)) { scanner *s = (scanner *)MALLOC(size, OTHER); s->scan = scan; s->vscan = vscan; s->getchr = getchr; s->ungotc = EOF; return s; } void X(scanner_destroy)(scanner *sc) { X(ifree)(sc); } fftw-3.3.3/kernel/solver.c0000644000175000017500000000250512054410050012336 00000000000000/* * Copyright (c) 2003, 2007-11 Matteo Frigo * Copyright (c) 2003, 2007-11 Massachusetts Institute of Technology * * This program is free software; you can redistribute it and/or modify * it under the terms of the GNU General Public License as published by * the Free Software Foundation; either version 2 of the License, or * (at your option) any later version. * * This program is distributed in the hope that it will be useful, * but WITHOUT ANY WARRANTY; without even the implied warranty of * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the * GNU General Public License for more details. * * You should have received a copy of the GNU General Public License * along with this program; if not, write to the Free Software * Foundation, Inc., 51 Franklin Street, Fifth Floor, Boston, MA 02110-1301 USA * */ #include "ifftw.h" solver *X(mksolver)(size_t size, const solver_adt *adt) { solver *s = (solver *)MALLOC(size, SOLVERS); s->adt = adt; s->refcnt = 0; return s; } void X(solver_use)(solver *ego) { ++ego->refcnt; } void X(solver_destroy)(solver *ego) { if ((--ego->refcnt) == 0) { if (ego->adt->destroy) ego->adt->destroy(ego); X(ifree)(ego); } } void X(solver_register)(planner *plnr, solver *s) { plnr->adt->register_solver(plnr, s); } fftw-3.3.3/kernel/solvtab.c0000644000175000017500000000204612054410050012476 00000000000000/* * Copyright (c) 2003, 2007-11 Matteo Frigo * Copyright (c) 2003, 2007-11 Massachusetts Institute of Technology * * This program is free software; you can redistribute it and/or modify * it under the terms of the GNU General Public License as published by * the Free Software Foundation; either version 2 of the License, or * (at your option) any later version. * * This program is distributed in the hope that it will be useful, * but WITHOUT ANY WARRANTY; without even the implied warranty of * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the * GNU General Public License for more details. * * You should have received a copy of the GNU General Public License * along with this program; if not, write to the Free Software * Foundation, Inc., 51 Franklin Street, Fifth Floor, Boston, MA 02110-1301 USA * */ #include "ifftw.h" void X(solvtab_exec)(const solvtab tbl, planner *p) { for (; tbl->reg_nam; ++tbl) { p->cur_reg_nam = tbl->reg_nam; p->cur_reg_id = 0; tbl->reg(p); } p->cur_reg_nam = 0; } fftw-3.3.3/kernel/stride.c0000644000175000017500000000224112054410050012313 00000000000000/* * Copyright (c) 2003, 2007-11 Matteo Frigo * Copyright (c) 2003, 2007-11 Massachusetts Institute of Technology * * This program is free software; you can redistribute it and/or modify * it under the terms of the GNU General Public License as published by * the Free Software Foundation; either version 2 of the License, or * (at your option) any later version. * * This program is distributed in the hope that it will be useful, * but WITHOUT ANY WARRANTY; without even the implied warranty of * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the * GNU General Public License for more details. * * You should have received a copy of the GNU General Public License * along with this program; if not, write to the Free Software * Foundation, Inc., 51 Franklin Street, Fifth Floor, Boston, MA 02110-1301 USA * */ #include "ifftw.h" const INT X(an_INT_guaranteed_to_be_zero) = 0; #ifdef PRECOMPUTE_ARRAY_INDICES stride X(mkstride)(INT n, INT s) { int i; INT *p = (INT *) MALLOC(n * sizeof(INT), STRIDES); for (i = 0; i < n; ++i) p[i] = s * i; return p; } void X(stride_destroy)(stride p) { X(ifree0)(p); } #endif fftw-3.3.3/kernel/tensor.c0000644000175000017500000000570312054410050012341 00000000000000/* * Copyright (c) 2003, 2007-11 Matteo Frigo * Copyright (c) 2003, 2007-11 Massachusetts Institute of Technology * * This program is free software; you can redistribute it and/or modify * it under the terms of the GNU General Public License as published by * the Free Software Foundation; either version 2 of the License, or * (at your option) any later version. * * This program is distributed in the hope that it will be useful, * but WITHOUT ANY WARRANTY; without even the implied warranty of * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the * GNU General Public License for more details. * * You should have received a copy of the GNU General Public License * along with this program; if not, write to the Free Software * Foundation, Inc., 51 Franklin Street, Fifth Floor, Boston, MA 02110-1301 USA * */ #include "ifftw.h" tensor *X(mktensor)(int rnk) { tensor *x; A(rnk >= 0); #if defined(STRUCT_HACK_KR) if (FINITE_RNK(rnk) && rnk > 1) x = (tensor *)MALLOC(sizeof(tensor) + (rnk - 1) * sizeof(iodim), TENSORS); else x = (tensor *)MALLOC(sizeof(tensor), TENSORS); #elif defined(STRUCT_HACK_C99) if (FINITE_RNK(rnk)) x = (tensor *)MALLOC(sizeof(tensor) + rnk * sizeof(iodim), TENSORS); else x = (tensor *)MALLOC(sizeof(tensor), TENSORS); #else x = (tensor *)MALLOC(sizeof(tensor), TENSORS); if (FINITE_RNK(rnk) && rnk > 0) x->dims = (iodim *)MALLOC(sizeof(iodim) * rnk, TENSORS); else x->dims = 0; #endif x->rnk = rnk; return x; } void X(tensor_destroy)(tensor *sz) { #if !defined(STRUCT_HACK_C99) && !defined(STRUCT_HACK_KR) X(ifree0)(sz->dims); #endif X(ifree)(sz); } INT X(tensor_sz)(const tensor *sz) { int i; INT n = 1; if (!FINITE_RNK(sz->rnk)) return 0; for (i = 0; i < sz->rnk; ++i) n *= sz->dims[i].n; return n; } void X(tensor_md5)(md5 *p, const tensor *t) { int i; X(md5int)(p, t->rnk); if (FINITE_RNK(t->rnk)) { for (i = 0; i < t->rnk; ++i) { const iodim *q = t->dims + i; X(md5INT)(p, q->n); X(md5INT)(p, q->is); X(md5INT)(p, q->os); } } } /* treat a (rank <= 1)-tensor as a rank-1 tensor, extracting appropriate n, is, and os components */ int X(tensor_tornk1)(const tensor *t, INT *n, INT *is, INT *os) { A(t->rnk <= 1); if (t->rnk == 1) { const iodim *vd = t->dims; *n = vd[0].n; *is = vd[0].is; *os = vd[0].os; } else { *n = 1; *is = *os = 0; } return 1; } void X(tensor_print)(const tensor *x, printer *p) { if (FINITE_RNK(x->rnk)) { int i; int first = 1; p->print(p, "("); for (i = 0; i < x->rnk; ++i) { const iodim *d = x->dims + i; p->print(p, "%s(%D %D %D)", first ? "" : " ", d->n, d->is, d->os); first = 0; } p->print(p, ")"); } else { p->print(p, "rank-minfty"); } } fftw-3.3.3/kernel/tensor1.c0000644000175000017500000000211312054410050012412 00000000000000/* * Copyright (c) 2003, 2007-11 Matteo Frigo * Copyright (c) 2003, 2007-11 Massachusetts Institute of Technology * * This program is free software; you can redistribute it and/or modify * it under the terms of the GNU General Public License as published by * the Free Software Foundation; either version 2 of the License, or * (at your option) any later version. * * This program is distributed in the hope that it will be useful, * but WITHOUT ANY WARRANTY; without even the implied warranty of * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the * GNU General Public License for more details. * * You should have received a copy of the GNU General Public License * along with this program; if not, write to the Free Software * Foundation, Inc., 51 Franklin Street, Fifth Floor, Boston, MA 02110-1301 USA * */ #include "ifftw.h" tensor *X(mktensor_0d)(void) { return X(mktensor(0)); } tensor *X(mktensor_1d)(INT n, INT is, INT os) { tensor *x = X(mktensor)(1); x->dims[0].n = n; x->dims[0].is = is; x->dims[0].os = os; return x; } fftw-3.3.3/kernel/tensor2.c0000644000175000017500000000302312054410050012414 00000000000000/* * Copyright (c) 2003, 2007-11 Matteo Frigo * Copyright (c) 2003, 2007-11 Massachusetts Institute of Technology * * This program is free software; you can redistribute it and/or modify * it under the terms of the GNU General Public License as published by * the Free Software Foundation; either version 2 of the License, or * (at your option) any later version. * * This program is distributed in the hope that it will be useful, * but WITHOUT ANY WARRANTY; without even the implied warranty of * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the * GNU General Public License for more details. * * You should have received a copy of the GNU General Public License * along with this program; if not, write to the Free Software * Foundation, Inc., 51 Franklin Street, Fifth Floor, Boston, MA 02110-1301 USA * */ #include "ifftw.h" tensor *X(mktensor_2d)(INT n0, INT is0, INT os0, INT n1, INT is1, INT os1) { tensor *x = X(mktensor)(2); x->dims[0].n = n0; x->dims[0].is = is0; x->dims[0].os = os0; x->dims[1].n = n1; x->dims[1].is = is1; x->dims[1].os = os1; return x; } tensor *X(mktensor_3d)(INT n0, INT is0, INT os0, INT n1, INT is1, INT os1, INT n2, INT is2, INT os2) { tensor *x = X(mktensor)(3); x->dims[0].n = n0; x->dims[0].is = is0; x->dims[0].os = os0; x->dims[1].n = n1; x->dims[1].is = is1; x->dims[1].os = os1; x->dims[2].n = n2; x->dims[2].is = is2; x->dims[2].os = os2; return x; } fftw-3.3.3/kernel/tensor3.c0000644000175000017500000000421612054410050012422 00000000000000/* * Copyright (c) 2003, 2007-11 Matteo Frigo * Copyright (c) 2003, 2007-11 Massachusetts Institute of Technology * * This program is free software; you can redistribute it and/or modify * it under the terms of the GNU General Public License as published by * the Free Software Foundation; either version 2 of the License, or * (at your option) any later version. * * This program is distributed in the hope that it will be useful, * but WITHOUT ANY WARRANTY; without even the implied warranty of * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the * GNU General Public License for more details. * * You should have received a copy of the GNU General Public License * along with this program; if not, write to the Free Software * Foundation, Inc., 51 Franklin Street, Fifth Floor, Boston, MA 02110-1301 USA * */ #include "ifftw.h" /* Currently, mktensor_4d and mktensor_5d are only used in the MPI routines, where very complicated transpositions are required. Therefore we split them into a separate source file. */ tensor *X(mktensor_4d)(INT n0, INT is0, INT os0, INT n1, INT is1, INT os1, INT n2, INT is2, INT os2, INT n3, INT is3, INT os3) { tensor *x = X(mktensor)(4); x->dims[0].n = n0; x->dims[0].is = is0; x->dims[0].os = os0; x->dims[1].n = n1; x->dims[1].is = is1; x->dims[1].os = os1; x->dims[2].n = n2; x->dims[2].is = is2; x->dims[2].os = os2; x->dims[3].n = n3; x->dims[3].is = is3; x->dims[3].os = os3; return x; } tensor *X(mktensor_5d)(INT n0, INT is0, INT os0, INT n1, INT is1, INT os1, INT n2, INT is2, INT os2, INT n3, INT is3, INT os3, INT n4, INT is4, INT os4) { tensor *x = X(mktensor)(5); x->dims[0].n = n0; x->dims[0].is = is0; x->dims[0].os = os0; x->dims[1].n = n1; x->dims[1].is = is1; x->dims[1].os = os1; x->dims[2].n = n2; x->dims[2].is = is2; x->dims[2].os = os2; x->dims[3].n = n3; x->dims[3].is = is3; x->dims[3].os = os3; x->dims[4].n = n4; x->dims[4].is = is4; x->dims[4].os = os4; return x; } fftw-3.3.3/kernel/tensor4.c0000644000175000017500000000634012054410050012423 00000000000000/* * Copyright (c) 2003, 2007-11 Matteo Frigo * Copyright (c) 2003, 2007-11 Massachusetts Institute of Technology * * This program is free software; you can redistribute it and/or modify * it under the terms of the GNU General Public License as published by * the Free Software Foundation; either version 2 of the License, or * (at your option) any later version. * * This program is distributed in the hope that it will be useful, * but WITHOUT ANY WARRANTY; without even the implied warranty of * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the * GNU General Public License for more details. * * You should have received a copy of the GNU General Public License * along with this program; if not, write to the Free Software * Foundation, Inc., 51 Franklin Street, Fifth Floor, Boston, MA 02110-1301 USA * */ #include "ifftw.h" INT X(tensor_max_index)(const tensor *sz) { int i; INT ni = 0, no = 0; A(FINITE_RNK(sz->rnk)); for (i = 0; i < sz->rnk; ++i) { const iodim *p = sz->dims + i; ni += (p->n - 1) * X(iabs)(p->is); no += (p->n - 1) * X(iabs)(p->os); } return X(imax)(ni, no); } #define tensor_min_xstride(sz, xs) { \ A(FINITE_RNK(sz->rnk)); \ if (sz->rnk == 0) return 0; \ else { \ int i; \ INT s = X(iabs)(sz->dims[0].xs); \ for (i = 1; i < sz->rnk; ++i) \ s = X(imin)(s, X(iabs)(sz->dims[i].xs)); \ return s; \ } \ } INT X(tensor_min_istride)(const tensor *sz) tensor_min_xstride(sz, is) INT X(tensor_min_ostride)(const tensor *sz) tensor_min_xstride(sz, os) INT X(tensor_min_stride)(const tensor *sz) { return X(imin)(X(tensor_min_istride)(sz), X(tensor_min_ostride)(sz)); } int X(tensor_inplace_strides)(const tensor *sz) { int i; A(FINITE_RNK(sz->rnk)); for (i = 0; i < sz->rnk; ++i) { const iodim *p = sz->dims + i; if (p->is != p->os) return 0; } return 1; } int X(tensor_inplace_strides2)(const tensor *a, const tensor *b) { return X(tensor_inplace_strides(a)) && X(tensor_inplace_strides(b)); } /* return true (1) iff *any* strides of sz decrease when we tensor_inplace_copy(sz, k). */ static int tensor_strides_decrease(const tensor *sz, inplace_kind k) { if (FINITE_RNK(sz->rnk)) { int i; for (i = 0; i < sz->rnk; ++i) if ((sz->dims[i].os - sz->dims[i].is) * (k == INPLACE_OS ? (INT)1 : (INT)-1) < 0) return 1; } return 0; } /* Return true (1) iff *any* strides of sz decrease when we tensor_inplace_copy(k) *or* if *all* strides of sz are unchanged but *any* strides of vecsz decrease. This is used in indirect.c to determine whether to use INPLACE_IS or INPLACE_OS. Note: X(tensor_strides_decrease)(sz, vecsz, INPLACE_IS) || X(tensor_strides_decrease)(sz, vecsz, INPLACE_OS) || X(tensor_inplace_strides2)(p->sz, p->vecsz) must always be true. */ int X(tensor_strides_decrease)(const tensor *sz, const tensor *vecsz, inplace_kind k) { return(tensor_strides_decrease(sz, k) || (X(tensor_inplace_strides)(sz) && tensor_strides_decrease(vecsz, k))); } fftw-3.3.3/kernel/tensor5.c0000644000175000017500000000526312054410050012427 00000000000000/* * Copyright (c) 2003, 2007-11 Matteo Frigo * Copyright (c) 2003, 2007-11 Massachusetts Institute of Technology * * This program is free software; you can redistribute it and/or modify * it under the terms of the GNU General Public License as published by * the Free Software Foundation; either version 2 of the License, or * (at your option) any later version. * * This program is distributed in the hope that it will be useful, * but WITHOUT ANY WARRANTY; without even the implied warranty of * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the * GNU General Public License for more details. * * You should have received a copy of the GNU General Public License * along with this program; if not, write to the Free Software * Foundation, Inc., 51 Franklin Street, Fifth Floor, Boston, MA 02110-1301 USA * */ #include "ifftw.h" static void dimcpy(iodim *dst, const iodim *src, int rnk) { int i; if (FINITE_RNK(rnk)) for (i = 0; i < rnk; ++i) dst[i] = src[i]; } tensor *X(tensor_copy)(const tensor *sz) { tensor *x = X(mktensor)(sz->rnk); dimcpy(x->dims, sz->dims, sz->rnk); return x; } /* like X(tensor_copy), but makes strides in-place by setting os = is if k == INPLACE_IS or is = os if k == INPLACE_OS. */ tensor *X(tensor_copy_inplace)(const tensor *sz, inplace_kind k) { tensor *x = X(tensor_copy)(sz); if (FINITE_RNK(x->rnk)) { int i; if (k == INPLACE_OS) for (i = 0; i < x->rnk; ++i) x->dims[i].is = x->dims[i].os; else for (i = 0; i < x->rnk; ++i) x->dims[i].os = x->dims[i].is; } return x; } /* Like X(tensor_copy), but copy all of the dimensions *except* except_dim. */ tensor *X(tensor_copy_except)(const tensor *sz, int except_dim) { tensor *x; A(FINITE_RNK(sz->rnk) && sz->rnk >= 1 && except_dim < sz->rnk); x = X(mktensor)(sz->rnk - 1); dimcpy(x->dims, sz->dims, except_dim); dimcpy(x->dims + except_dim, sz->dims + except_dim + 1, x->rnk - except_dim); return x; } /* Like X(tensor_copy), but copy only rnk dimensions starting with start_dim. */ tensor *X(tensor_copy_sub)(const tensor *sz, int start_dim, int rnk) { tensor *x; A(FINITE_RNK(sz->rnk) && start_dim + rnk <= sz->rnk); x = X(mktensor)(rnk); dimcpy(x->dims, sz->dims + start_dim, rnk); return x; } tensor *X(tensor_append)(const tensor *a, const tensor *b) { if (!FINITE_RNK(a->rnk) || !FINITE_RNK(b->rnk)) { return X(mktensor)(RNK_MINFTY); } else { tensor *x = X(mktensor)(a->rnk + b->rnk); dimcpy(x->dims, a->dims, a->rnk); dimcpy(x->dims + a->rnk, b->dims, b->rnk); return x; } } fftw-3.3.3/kernel/tensor7.c0000644000175000017500000001445112054410050012430 00000000000000/* * Copyright (c) 2003, 2007-11 Matteo Frigo * Copyright (c) 2003, 2007-11 Massachusetts Institute of Technology * * This program is free software; you can redistribute it and/or modify * it under the terms of the GNU General Public License as published by * the Free Software Foundation; either version 2 of the License, or * (at your option) any later version. * * This program is distributed in the hope that it will be useful, * but WITHOUT ANY WARRANTY; without even the implied warranty of * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the * GNU General Public License for more details. * * You should have received a copy of the GNU General Public License * along with this program; if not, write to the Free Software * Foundation, Inc., 51 Franklin Street, Fifth Floor, Boston, MA 02110-1301 USA * */ #include "ifftw.h" static int signof(INT x) { if (x < 0) return -1; if (x == 0) return 0; /* if (x > 0) */ return 1; } /* total order among iodim's */ int X(dimcmp)(const iodim *a, const iodim *b) { INT sai = X(iabs)(a->is), sbi = X(iabs)(b->is); INT sao = X(iabs)(a->os), sbo = X(iabs)(b->os); INT sam = X(imin)(sai, sao), sbm = X(imin)(sbi, sbo); /* in descending order of min{istride, ostride} */ if (sam != sbm) return signof(sbm - sam); /* in case of a tie, in descending order of istride */ if (sbi != sai) return signof(sbi - sai); /* in case of a tie, in descending order of ostride */ if (sbo != sao) return signof(sbo - sao); /* in case of a tie, in ascending order of n */ return signof(a->n - b->n); } static void canonicalize(tensor *x) { if (x->rnk > 1) { qsort(x->dims, (size_t)x->rnk, sizeof(iodim), (int (*)(const void *, const void *))X(dimcmp)); } } static int compare_by_istride(const iodim *a, const iodim *b) { INT sai = X(iabs)(a->is), sbi = X(iabs)(b->is); /* in descending order of istride */ return signof(sbi - sai); } static tensor *really_compress(const tensor *sz) { int i, rnk; tensor *x; A(FINITE_RNK(sz->rnk)); for (i = rnk = 0; i < sz->rnk; ++i) { A(sz->dims[i].n > 0); if (sz->dims[i].n != 1) ++rnk; } x = X(mktensor)(rnk); for (i = rnk = 0; i < sz->rnk; ++i) { if (sz->dims[i].n != 1) x->dims[rnk++] = sz->dims[i]; } return x; } /* Like tensor_copy, but eliminate n == 1 dimensions, which never affect any transform or transform vector. Also, we sort the tensor into a canonical order of decreasing strides (see X(dimcmp) for an exact definition). In general, processing a loop/array in order of decreasing stride will improve locality. Both forward and backwards traversal of the tensor are considered e.g. by vrank-geq1, so sorting in increasing vs. decreasing order is not really important. */ tensor *X(tensor_compress)(const tensor *sz) { tensor *x = really_compress(sz); canonicalize(x); return x; } /* Return whether the strides of a and b are such that they form an effective contiguous 1d array. Assumes that a.is >= b.is. */ static int strides_contig(iodim *a, iodim *b) { return (a->is == b->is * b->n && a->os == b->os * b->n); } /* Like tensor_compress, but also compress into one dimension any group of dimensions that form a contiguous block of indices with some stride. (This can safely be done for transform vector sizes.) */ tensor *X(tensor_compress_contiguous)(const tensor *sz) { int i, rnk; tensor *sz2, *x; if (X(tensor_sz)(sz) == 0) return X(mktensor)(RNK_MINFTY); sz2 = really_compress(sz); A(FINITE_RNK(sz2->rnk)); if (sz2->rnk <= 1) { /* nothing to compress. */ if (0) { /* this call is redundant, because "sz->rnk <= 1" implies that the tensor is already canonical, but I am writing it explicitly because "logically" we need to canonicalize the tensor before returning. */ canonicalize(sz2); } return sz2; } /* sort in descending order of |istride|, so that compressible dimensions appear contigously */ qsort(sz2->dims, (size_t)sz2->rnk, sizeof(iodim), (int (*)(const void *, const void *))compare_by_istride); /* compute what the rank will be after compression */ for (i = rnk = 1; i < sz2->rnk; ++i) if (!strides_contig(sz2->dims + i - 1, sz2->dims + i)) ++rnk; /* merge adjacent dimensions whenever possible */ x = X(mktensor)(rnk); x->dims[0] = sz2->dims[0]; for (i = rnk = 1; i < sz2->rnk; ++i) { if (strides_contig(sz2->dims + i - 1, sz2->dims + i)) { x->dims[rnk - 1].n *= sz2->dims[i].n; x->dims[rnk - 1].is = sz2->dims[i].is; x->dims[rnk - 1].os = sz2->dims[i].os; } else { A(rnk < x->rnk); x->dims[rnk++] = sz2->dims[i]; } } X(tensor_destroy)(sz2); /* reduce to canonical form */ canonicalize(x); return x; } /* The inverse of X(tensor_append): splits the sz tensor into tensor a followed by tensor b, where a's rank is arnk. */ void X(tensor_split)(const tensor *sz, tensor **a, int arnk, tensor **b) { A(FINITE_RNK(sz->rnk) && FINITE_RNK(arnk)); *a = X(tensor_copy_sub)(sz, 0, arnk); *b = X(tensor_copy_sub)(sz, arnk, sz->rnk - arnk); } /* TRUE if the two tensors are equal */ int X(tensor_equal)(const tensor *a, const tensor *b) { if (a->rnk != b->rnk) return 0; if (FINITE_RNK(a->rnk)) { int i; for (i = 0; i < a->rnk; ++i) if (0 || a->dims[i].n != b->dims[i].n || a->dims[i].is != b->dims[i].is || a->dims[i].os != b->dims[i].os ) return 0; } return 1; } /* TRUE if the sets of input and output locations described by (append sz vecsz) are the same */ int X(tensor_inplace_locations)(const tensor *sz, const tensor *vecsz) { tensor *t = X(tensor_append)(sz, vecsz); tensor *ti = X(tensor_copy_inplace)(t, INPLACE_IS); tensor *to = X(tensor_copy_inplace)(t, INPLACE_OS); tensor *tic = X(tensor_compress_contiguous)(ti); tensor *toc = X(tensor_compress_contiguous)(to); int retval = X(tensor_equal)(tic, toc); X(tensor_destroy)(t); X(tensor_destroy4)(ti, to, tic, toc); return retval; } fftw-3.3.3/kernel/tensor8.c0000644000175000017500000000212112054410050012420 00000000000000/* * Copyright (c) 2003, 2007-11 Matteo Frigo * Copyright (c) 2003, 2007-11 Massachusetts Institute of Technology * * This program is free software; you can redistribute it and/or modify * it under the terms of the GNU General Public License as published by * the Free Software Foundation; either version 2 of the License, or * (at your option) any later version. * * This program is distributed in the hope that it will be useful, * but WITHOUT ANY WARRANTY; without even the implied warranty of * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the * GNU General Public License for more details. * * You should have received a copy of the GNU General Public License * along with this program; if not, write to the Free Software * Foundation, Inc., 51 Franklin Street, Fifth Floor, Boston, MA 02110-1301 USA * */ #include "ifftw.h" void X(tensor_destroy2)(tensor *a, tensor *b) { X(tensor_destroy)(a); X(tensor_destroy)(b); } void X(tensor_destroy4)(tensor *a, tensor *b, tensor *c, tensor *d) { X(tensor_destroy2)(a, b); X(tensor_destroy2)(c, d); } fftw-3.3.3/kernel/tensor9.c0000644000175000017500000000207512054410050012431 00000000000000/* * Copyright (c) 2003, 2007-11 Matteo Frigo * Copyright (c) 2003, 2007-11 Massachusetts Institute of Technology * * This program is free software; you can redistribute it and/or modify * it under the terms of the GNU General Public License as published by * the Free Software Foundation; either version 2 of the License, or * (at your option) any later version. * * This program is distributed in the hope that it will be useful, * but WITHOUT ANY WARRANTY; without even the implied warranty of * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the * GNU General Public License for more details. * * You should have received a copy of the GNU General Public License * along with this program; if not, write to the Free Software * Foundation, Inc., 51 Franklin Street, Fifth Floor, Boston, MA 02110-1301 USA * */ #include "ifftw.h" int X(tensor_kosherp)(const tensor *x) { int i; if (x->rnk < 0) return 0; if (FINITE_RNK(x->rnk)) { for (i = 0; i < x->rnk; ++i) if (x->dims[i].n < 0) return 0; } return 1; } fftw-3.3.3/kernel/tile2d.c0000644000175000017500000000323412054410050012207 00000000000000/* * Copyright (c) 2003, 2007-11 Matteo Frigo * Copyright (c) 2003, 2007-11 Massachusetts Institute of Technology * * This program is free software; you can redistribute it and/or modify * it under the terms of the GNU General Public License as published by * the Free Software Foundation; either version 2 of the License, or * (at your option) any later version. * * This program is distributed in the hope that it will be useful, * but WITHOUT ANY WARRANTY; without even the implied warranty of * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the * GNU General Public License for more details. * * You should have received a copy of the GNU General Public License * along with this program; if not, write to the Free Software * Foundation, Inc., 51 Franklin Street, Fifth Floor, Boston, MA 02110-1301 USA * */ /* out of place 2D copy routines */ #include "ifftw.h" void X(tile2d)(INT n0l, INT n0u, INT n1l, INT n1u, INT tilesz, void (*f)(INT n0l, INT n0u, INT n1l, INT n1u, void *args), void *args) { INT d0, d1; A(tilesz > 0); /* infinite loops otherwise */ tail: d0 = n0u - n0l; d1 = n1u - n1l; if (d0 >= d1 && d0 > tilesz) { INT n0m = (n0u + n0l) / 2; X(tile2d)(n0l, n0m, n1l, n1u, tilesz, f, args); n0l = n0m; goto tail; } else if (/* d1 >= d0 && */ d1 > tilesz) { INT n1m = (n1u + n1l) / 2; X(tile2d)(n0l, n0u, n1l, n1m, tilesz, f, args); n1l = n1m; goto tail; } else { f(n0l, n0u, n1l, n1u, args); } } INT X(compute_tilesz)(INT vl, int how_many_tiles_in_cache) { return X(isqrt)(CACHESIZE / (((INT)sizeof(R)) * vl * (INT)how_many_tiles_in_cache)); } fftw-3.3.3/kernel/timer.c0000644000175000017500000001166312054410050012151 00000000000000/* * Copyright (c) 2003, 2007-11 Matteo Frigo * Copyright (c) 2003, 2007-11 Massachusetts Institute of Technology * * This program is free software; you can redistribute it and/or modify * it under the terms of the GNU General Public License as published by * the Free Software Foundation; either version 2 of the License, or * (at your option) any later version. * * This program is distributed in the hope that it will be useful, * but WITHOUT ANY WARRANTY; without even the implied warranty of * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the * GNU General Public License for more details. * * You should have received a copy of the GNU General Public License * along with this program; if not, write to the Free Software * Foundation, Inc., 51 Franklin Street, Fifth Floor, Boston, MA 02110-1301 USA * */ #include "ifftw.h" #ifdef HAVE_UNISTD_H # include #endif #ifndef WITH_SLOW_TIMER # include "cycle.h" #endif #ifndef FFTW_TIME_LIMIT #define FFTW_TIME_LIMIT 2.0 /* don't run for more than two seconds */ #endif /* the following code is disabled for now, because it seems to require that we #include in ifftw.h to typedef LARGE_INTEGER crude_time, and this pulls in the whole Windows universe and leads to namespace conflicts (unless we did some hack like assuming sizeof(LARGE_INTEGER) == sizeof(long long). gettimeofday is provided by MinGW, which we use to cross-compile FFTW for Windows, and this seems to work well enough */ #if 0 && (defined(__WIN32__) || defined(_WIN32) || defined(_WIN64)) crude_time X(get_crude_time)(void) { crude_time tv; QueryPerformanceCounter(&tv); return tv; } static double elapsed_since(crude_time t0) { crude_time t1, freq; QueryPerformanceCounter(&t1); QueryPerformanceFrequency(&freq); return (((double) (t1.QuadPart - t0.QuadPart))) / ((double) freq.QuadPart); } # define TIME_MIN_SEC 1.0e-2 #elif defined(HAVE_GETTIMEOFDAY) crude_time X(get_crude_time)(void) { crude_time tv; gettimeofday(&tv, 0); return tv; } #define elapsed_sec(t1,t0) ((double)(t1.tv_sec - t0.tv_sec) + \ (double)(t1.tv_usec - t0.tv_usec) * 1.0E-6) static double elapsed_since(crude_time t0) { crude_time t1; gettimeofday(&t1, 0); return elapsed_sec(t1, t0); } # define TIME_MIN_SEC 1.0e-3 #else /* !HAVE_GETTIMEOFDAY */ /* Note that the only system where we are likely to need to fall back on the clock() function is Windows, for which CLOCKS_PER_SEC is 1000 and thus the clock wraps once every 50 days. This should hopefully be longer than the time required to create any single plan! */ crude_time X(get_crude_time)(void) { return clock(); } #define elapsed_sec(t1,t0) ((double) ((t1) - (t0)) / CLOCKS_PER_SEC) static double elapsed_since(crude_time t0) { return elapsed_sec(clock(), t0); } # define TIME_MIN_SEC 2.0e-1 /* from fftw2 */ #endif /* !HAVE_GETTIMEOFDAY */ double X(elapsed_since)(const planner *plnr, const problem *p, crude_time t0) { double t = elapsed_since(t0); if (plnr->cost_hook) t = plnr->cost_hook(p, t, COST_MAX); return t; } #ifdef WITH_SLOW_TIMER /* excruciatingly slow; only use this if there is no choice! */ typedef crude_time ticks; # define getticks X(get_crude_time) # define elapsed(t1,t0) elapsed_sec(t1,t0) # define TIME_MIN TIME_MIN_SEC # define TIME_REPEAT 4 /* from fftw2 */ # define HAVE_TICK_COUNTER #endif #ifdef HAVE_TICK_COUNTER # ifndef TIME_MIN # define TIME_MIN 100.0 # endif # ifndef TIME_REPEAT # define TIME_REPEAT 8 # endif static double measure(plan *pln, const problem *p, int iter) { ticks t0, t1; int i; t0 = getticks(); for (i = 0; i < iter; ++i) pln->adt->solve(pln, p); t1 = getticks(); return elapsed(t1, t0); } double X(measure_execution_time)(const planner *plnr, plan *pln, const problem *p) { int iter; int repeat; X(plan_awake)(pln, AWAKE_ZERO); p->adt->zero(p); start_over: for (iter = 1; iter; iter *= 2) { double tmin = 0; int first = 1; crude_time begin = X(get_crude_time)(); /* repeat the measurement TIME_REPEAT times */ for (repeat = 0; repeat < TIME_REPEAT; ++repeat) { double t = measure(pln, p, iter); if (plnr->cost_hook) t = plnr->cost_hook(p, t, COST_MAX); if (t < 0) goto start_over; if (first || t < tmin) tmin = t; first = 0; /* do not run for too long */ if (X(elapsed_since)(plnr, p, begin) > FFTW_TIME_LIMIT) break; } if (tmin >= TIME_MIN) { X(plan_awake)(pln, SLEEPY); return tmin / (double) iter; } } goto start_over; /* may happen if timer is screwed up */ } #else /* no cycle counter */ double X(measure_execution_time)(const planner *plnr, plan *pln, const problem *p) { UNUSED(plnr); UNUSED(p); UNUSED(pln); return -1.0; } #endif fftw-3.3.3/kernel/transpose.c0000644000175000017500000001226612054410050013047 00000000000000/* * Copyright (c) 2003, 2007-11 Matteo Frigo * Copyright (c) 2003, 2007-11 Massachusetts Institute of Technology * * This program is free software; you can redistribute it and/or modify * it under the terms of the GNU General Public License as published by * the Free Software Foundation; either version 2 of the License, or * (at your option) any later version. * * This program is distributed in the hope that it will be useful, * but WITHOUT ANY WARRANTY; without even the implied warranty of * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the * GNU General Public License for more details. * * You should have received a copy of the GNU General Public License * along with this program; if not, write to the Free Software * Foundation, Inc., 51 Franklin Street, Fifth Floor, Boston, MA 02110-1301 USA * */ #include "ifftw.h" /* in place square transposition, iterative */ void X(transpose)(R *I, INT n, INT s0, INT s1, INT vl) { INT i0, i1, v; switch (vl) { case 1: for (i1 = 1; i1 < n; ++i1) { for (i0 = 0; i0 < i1; ++i0) { R x0 = I[i1 * s0 + i0 * s1]; R y0 = I[i1 * s1 + i0 * s0]; I[i1 * s1 + i0 * s0] = x0; I[i1 * s0 + i0 * s1] = y0; } } break; case 2: for (i1 = 1; i1 < n; ++i1) { for (i0 = 0; i0 < i1; ++i0) { R x0 = I[i1 * s0 + i0 * s1]; R x1 = I[i1 * s0 + i0 * s1 + 1]; R y0 = I[i1 * s1 + i0 * s0]; R y1 = I[i1 * s1 + i0 * s0 + 1]; I[i1 * s1 + i0 * s0] = x0; I[i1 * s1 + i0 * s0 + 1] = x1; I[i1 * s0 + i0 * s1] = y0; I[i1 * s0 + i0 * s1 + 1] = y1; } } break; default: for (i1 = 1; i1 < n; ++i1) { for (i0 = 0; i0 < i1; ++i0) { for (v = 0; v < vl; ++v) { R x0 = I[i1 * s0 + i0 * s1 + v]; R y0 = I[i1 * s1 + i0 * s0 + v]; I[i1 * s1 + i0 * s0 + v] = x0; I[i1 * s0 + i0 * s1 + v] = y0; } } } break; } } struct transpose_closure { R *I; INT s0, s1, vl, tilesz; R *buf0, *buf1; }; static void dotile(INT n0l, INT n0u, INT n1l, INT n1u, void *args) { struct transpose_closure *k = (struct transpose_closure *)args; R *I = k->I; INT s0 = k->s0, s1 = k->s1, vl = k->vl; INT i0, i1, v; switch (vl) { case 1: for (i1 = n1l; i1 < n1u; ++i1) { for (i0 = n0l; i0 < n0u; ++i0) { R x0 = I[i1 * s0 + i0 * s1]; R y0 = I[i1 * s1 + i0 * s0]; I[i1 * s1 + i0 * s0] = x0; I[i1 * s0 + i0 * s1] = y0; } } break; case 2: for (i1 = n1l; i1 < n1u; ++i1) { for (i0 = n0l; i0 < n0u; ++i0) { R x0 = I[i1 * s0 + i0 * s1]; R x1 = I[i1 * s0 + i0 * s1 + 1]; R y0 = I[i1 * s1 + i0 * s0]; R y1 = I[i1 * s1 + i0 * s0 + 1]; I[i1 * s1 + i0 * s0] = x0; I[i1 * s1 + i0 * s0 + 1] = x1; I[i1 * s0 + i0 * s1] = y0; I[i1 * s0 + i0 * s1 + 1] = y1; } } break; default: for (i1 = n1l; i1 < n1u; ++i1) { for (i0 = n0l; i0 < n0u; ++i0) { for (v = 0; v < vl; ++v) { R x0 = I[i1 * s0 + i0 * s1 + v]; R y0 = I[i1 * s1 + i0 * s0 + v]; I[i1 * s1 + i0 * s0 + v] = x0; I[i1 * s0 + i0 * s1 + v] = y0; } } } } } static void dotile_buf(INT n0l, INT n0u, INT n1l, INT n1u, void *args) { struct transpose_closure *k = (struct transpose_closure *)args; X(cpy2d_ci)(k->I + n0l * k->s0 + n1l * k->s1, k->buf0, n0u - n0l, k->s0, k->vl, n1u - n1l, k->s1, k->vl * (n0u - n0l), k->vl); X(cpy2d_ci)(k->I + n0l * k->s1 + n1l * k->s0, k->buf1, n0u - n0l, k->s1, k->vl, n1u - n1l, k->s0, k->vl * (n0u - n0l), k->vl); X(cpy2d_co)(k->buf1, k->I + n0l * k->s0 + n1l * k->s1, n0u - n0l, k->vl, k->s0, n1u - n1l, k->vl * (n0u - n0l), k->s1, k->vl); X(cpy2d_co)(k->buf0, k->I + n0l * k->s1 + n1l * k->s0, n0u - n0l, k->vl, k->s1, n1u - n1l, k->vl * (n0u - n0l), k->s0, k->vl); } static void transpose_rec(R *I, INT n, void (*f)(INT n0l, INT n0u, INT n1l, INT n1u, void *args), struct transpose_closure *k) { tail: if (n > 1) { INT n2 = n / 2; k->I = I; X(tile2d)(0, n2, n2, n, k->tilesz, f, k); transpose_rec(I, n2, f, k); I += n2 * (k->s0 + k->s1); n -= n2; goto tail; } } void X(transpose_tiled)(R *I, INT n, INT s0, INT s1, INT vl) { struct transpose_closure k; k.s0 = s0; k.s1 = s1; k.vl = vl; /* two blocks must be in cache, to be swapped */ k.tilesz = X(compute_tilesz)(vl, 2); k.buf0 = k.buf1 = 0; /* unused */ transpose_rec(I, n, dotile, &k); } void X(transpose_tiledbuf)(R *I, INT n, INT s0, INT s1, INT vl) { struct transpose_closure k; /* Assume that the the rows of I conflict into the same cache lines, and therefore we don't need to reserve cache space for the input. If the rows don't conflict, there is no reason to use tiledbuf at all.*/ R buf0[CACHESIZE / (2 * sizeof(R))]; R buf1[CACHESIZE / (2 * sizeof(R))]; k.s0 = s0; k.s1 = s1; k.vl = vl; k.tilesz = X(compute_tilesz)(vl, 2); k.buf0 = buf0; k.buf1 = buf1; A(k.tilesz * k.tilesz * vl * sizeof(R) <= sizeof(buf0)); A(k.tilesz * k.tilesz * vl * sizeof(R) <= sizeof(buf1)); transpose_rec(I, n, dotile_buf, &k); } fftw-3.3.3/kernel/trig.c0000644000175000017500000001261112054410050011770 00000000000000/* * Copyright (c) 2003, 2007-11 Matteo Frigo * Copyright (c) 2003, 2007-11 Massachusetts Institute of Technology * * This program is free software; you can redistribute it and/or modify * it under the terms of the GNU General Public License as published by * the Free Software Foundation; either version 2 of the License, or * (at your option) any later version. * * This program is distributed in the hope that it will be useful, * but WITHOUT ANY WARRANTY; without even the implied warranty of * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the * GNU General Public License for more details. * * You should have received a copy of the GNU General Public License * along with this program; if not, write to the Free Software * Foundation, Inc., 51 Franklin Street, Fifth Floor, Boston, MA 02110-1301 USA * */ /* trigonometric functions */ #include "ifftw.h" #include #if defined(TRIGREAL_IS_LONG_DOUBLE) # define COS cosl # define SIN sinl # define KTRIG(x) (x##L) # ifndef HAVE_DECL_SINL extern long double sinl(long double x); # endif # ifndef HAVE_DECL_COSL extern long double cosl(long double x); # endif #elif defined(TRIGREAL_IS_QUAD) # define COS cosq # define SIN sinq # define KTRIG(x) (x##Q) extern __float128 sinq(__float128 x); extern __float128 cosq(__float128 x); #else # define COS cos # define SIN sin # define KTRIG(x) (x) #endif static const trigreal K2PI = KTRIG(6.2831853071795864769252867665590057683943388); #define by2pi(m, n) ((K2PI * (m)) / (n)) /* * Improve accuracy by reducing x to range [0..1/8] * before multiplication by 2 * PI. */ static void real_cexp(INT m, INT n, trigreal *out) { trigreal theta, c, s, t; unsigned octant = 0; INT quarter_n = n; n += n; n += n; m += m; m += m; if (m < 0) m += n; if (m > n - m) { m = n - m; octant |= 4; } if (m - quarter_n > 0) { m = m - quarter_n; octant |= 2; } if (m > quarter_n - m) { m = quarter_n - m; octant |= 1; } theta = by2pi(m, n); c = COS(theta); s = SIN(theta); if (octant & 1) { t = c; c = s; s = t; } if (octant & 2) { t = c; c = -s; s = t; } if (octant & 4) { s = -s; } out[0] = c; out[1] = s; } static INT choose_twshft(INT n) { INT log2r = 0; while (n > 0) { ++log2r; n /= 4; } return log2r; } static void cexpl_sqrtn_table(triggen *p, INT m, trigreal *res) { m += p->n * (m < 0); { INT m0 = m & p->twmsk; INT m1 = m >> p->twshft; trigreal wr0 = p->W0[2 * m0]; trigreal wi0 = p->W0[2 * m0 + 1]; trigreal wr1 = p->W1[2 * m1]; trigreal wi1 = p->W1[2 * m1 + 1]; res[0] = wr1 * wr0 - wi1 * wi0; res[1] = wi1 * wr0 + wr1 * wi0; } } /* multiply (xr, xi) by exp(FFT_SIGN * 2*pi*i*m/n) */ static void rotate_sqrtn_table(triggen *p, INT m, R xr, R xi, R *res) { m += p->n * (m < 0); { INT m0 = m & p->twmsk; INT m1 = m >> p->twshft; trigreal wr0 = p->W0[2 * m0]; trigreal wi0 = p->W0[2 * m0 + 1]; trigreal wr1 = p->W1[2 * m1]; trigreal wi1 = p->W1[2 * m1 + 1]; trigreal wr = wr1 * wr0 - wi1 * wi0; trigreal wi = wi1 * wr0 + wr1 * wi0; #if FFT_SIGN == -1 res[0] = xr * wr + xi * wi; res[1] = xi * wr - xr * wi; #else res[0] = xr * wr - xi * wi; res[1] = xi * wr + xr * wi; #endif } } static void cexpl_sincos(triggen *p, INT m, trigreal *res) { real_cexp(m, p->n, res); } static void cexp_zero(triggen *p, INT m, R *res) { UNUSED(p); UNUSED(m); res[0] = 0; res[1] = 0; } static void cexpl_zero(triggen *p, INT m, trigreal *res) { UNUSED(p); UNUSED(m); res[0] = 0; res[1] = 0; } static void cexp_generic(triggen *p, INT m, R *res) { trigreal resl[2]; p->cexpl(p, m, resl); res[0] = (R)resl[0]; res[1] = (R)resl[1]; } static void rotate_generic(triggen *p, INT m, R xr, R xi, R *res) { trigreal w[2]; p->cexpl(p, m, w); res[0] = xr * w[0] - xi * (FFT_SIGN * w[1]); res[1] = xi * w[0] + xr * (FFT_SIGN * w[1]); } triggen *X(mktriggen)(enum wakefulness wakefulness, INT n) { INT i, n0, n1; triggen *p = (triggen *)MALLOC(sizeof(*p), TWIDDLES); p->n = n; p->W0 = p->W1 = 0; p->cexp = 0; p->rotate = 0; switch (wakefulness) { case SLEEPY: A(0 /* can't happen */); break; case AWAKE_SQRTN_TABLE: { INT twshft = choose_twshft(n); p->twshft = twshft; p->twradix = ((INT)1) << twshft; p->twmsk = p->twradix - 1; n0 = p->twradix; n1 = (n + n0 - 1) / n0; p->W0 = (trigreal *)MALLOC(n0 * 2 * sizeof(trigreal), TWIDDLES); p->W1 = (trigreal *)MALLOC(n1 * 2 * sizeof(trigreal), TWIDDLES); for (i = 0; i < n0; ++i) real_cexp(i, n, p->W0 + 2 * i); for (i = 0; i < n1; ++i) real_cexp(i * p->twradix, n, p->W1 + 2 * i); p->cexpl = cexpl_sqrtn_table; p->rotate = rotate_sqrtn_table; break; } case AWAKE_SINCOS: p->cexpl = cexpl_sincos; break; case AWAKE_ZERO: p->cexp = cexp_zero; p->cexpl = cexpl_zero; break; } if (!p->cexp) { if (sizeof(trigreal) == sizeof(R)) p->cexp = (void (*)(triggen *, INT, R *))p->cexpl; else p->cexp = cexp_generic; } if (!p->rotate) p->rotate = rotate_generic; return p; } void X(triggen_destroy)(triggen *p) { X(ifree0)(p->W0); X(ifree0)(p->W1); X(ifree)(p); } fftw-3.3.3/kernel/twiddle.c0000644000175000017500000001241212054410050012456 00000000000000/* * Copyright (c) 2003, 2007-11 Matteo Frigo * Copyright (c) 2003, 2007-11 Massachusetts Institute of Technology * * This program is free software; you can redistribute it and/or modify * it under the terms of the GNU General Public License as published by * the Free Software Foundation; either version 2 of the License, or * (at your option) any later version. * * This program is distributed in the hope that it will be useful, * but WITHOUT ANY WARRANTY; without even the implied warranty of * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the * GNU General Public License for more details. * * You should have received a copy of the GNU General Public License * along with this program; if not, write to the Free Software * Foundation, Inc., 51 Franklin Street, Fifth Floor, Boston, MA 02110-1301 USA * */ /* Twiddle manipulation */ #include "ifftw.h" #include #define HASHSZ 109 /* hash table of known twiddle factors */ static twid *twlist[HASHSZ]; static INT hash(INT n, INT r) { INT h = n * 17 + r; if (h < 0) h = -h; return (h % HASHSZ); } static int equal_instr(const tw_instr *p, const tw_instr *q) { if (p == q) return 1; for (;; ++p, ++q) { if (p->op != q->op) return 0; switch (p->op) { case TW_NEXT: return (p->v == q->v); /* p->i is ignored */ case TW_FULL: case TW_HALF: if (p->v != q->v) return 0; /* p->i is ignored */ break; default: if (p->v != q->v || p->i != q->i) return 0; break; } } A(0 /* can't happen */); } static int ok_twid(const twid *t, enum wakefulness wakefulness, const tw_instr *q, INT n, INT r, INT m) { return (wakefulness == t->wakefulness && n == t->n && r == t->r && m <= t->m && equal_instr(t->instr, q)); } static twid *lookup(enum wakefulness wakefulness, const tw_instr *q, INT n, INT r, INT m) { twid *p; for (p = twlist[hash(n,r)]; p && !ok_twid(p, wakefulness, q, n, r, m); p = p->cdr) ; return p; } static INT twlen0(INT r, const tw_instr *p, INT *vl) { INT ntwiddle = 0; /* compute length of bytecode program */ A(r > 0); for ( ; p->op != TW_NEXT; ++p) { switch (p->op) { case TW_FULL: ntwiddle += (r - 1) * 2; break; case TW_HALF: ntwiddle += (r - 1); break; case TW_CEXP: ntwiddle += 2; break; case TW_COS: case TW_SIN: ntwiddle += 1; break; } } *vl = (INT)p->v; return ntwiddle; } INT X(twiddle_length)(INT r, const tw_instr *p) { INT vl; return twlen0(r, p, &vl); } static R *compute(enum wakefulness wakefulness, const tw_instr *instr, INT n, INT r, INT m) { INT ntwiddle, j, vl; R *W, *W0; const tw_instr *p; triggen *t = X(mktriggen)(wakefulness, n); p = instr; ntwiddle = twlen0(r, p, &vl); A(m % vl == 0); W0 = W = (R *)MALLOC((ntwiddle * (m / vl)) * sizeof(R), TWIDDLES); for (j = 0; j < m; j += vl) { for (p = instr; p->op != TW_NEXT; ++p) { switch (p->op) { case TW_FULL: { INT i; for (i = 1; i < r; ++i) { A((j + (INT)p->v) * i < n); A((j + (INT)p->v) * i > -n); t->cexp(t, (j + (INT)p->v) * i, W); W += 2; } break; } case TW_HALF: { INT i; A((r % 2) == 1); for (i = 1; i + i < r; ++i) { t->cexp(t, MULMOD(i, (j + (INT)p->v), n), W); W += 2; } break; } case TW_COS: { R d[2]; A((j + (INT)p->v) * p->i < n); A((j + (INT)p->v) * p->i > -n); t->cexp(t, (j + (INT)p->v) * (INT)p->i, d); *W++ = d[0]; break; } case TW_SIN: { R d[2]; A((j + (INT)p->v) * p->i < n); A((j + (INT)p->v) * p->i > -n); t->cexp(t, (j + (INT)p->v) * (INT)p->i, d); *W++ = d[1]; break; } case TW_CEXP: A((j + (INT)p->v) * p->i < n); A((j + (INT)p->v) * p->i > -n); t->cexp(t, (j + (INT)p->v) * (INT)p->i, W); W += 2; break; } } } X(triggen_destroy)(t); return W0; } static void mktwiddle(enum wakefulness wakefulness, twid **pp, const tw_instr *instr, INT n, INT r, INT m) { twid *p; INT h; if ((p = lookup(wakefulness, instr, n, r, m))) { ++p->refcnt; } else { p = (twid *) MALLOC(sizeof(twid), TWIDDLES); p->n = n; p->r = r; p->m = m; p->instr = instr; p->refcnt = 1; p->wakefulness = wakefulness; p->W = compute(wakefulness, instr, n, r, m); /* cons! onto twlist */ h = hash(n, r); p->cdr = twlist[h]; twlist[h] = p; } *pp = p; } static void twiddle_destroy(twid **pp) { twid *p = *pp; twid **q; if ((--p->refcnt) == 0) { /* remove p from twiddle list */ for (q = &twlist[hash(p->n, p->r)]; *q; q = &((*q)->cdr)) { if (*q == p) { *q = p->cdr; X(ifree)(p->W); X(ifree)(p); *pp = 0; return; } } A(0 /* can't happen */ ); } } void X(twiddle_awake)(enum wakefulness wakefulness, twid **pp, const tw_instr *instr, INT n, INT r, INT m) { switch (wakefulness) { case SLEEPY: twiddle_destroy(pp); break; default: mktwiddle(wakefulness, pp, instr, n, r, m); break; } } fftw-3.3.3/kernel/cycle.h0000644000175000017500000003302612054410050012132 00000000000000/* * Copyright (c) 2003, 2007-11 Matteo Frigo * Copyright (c) 2003, 2007-11 Massachusetts Institute of Technology * * Permission is hereby granted, free of charge, to any person obtaining * a copy of this software and associated documentation files (the * "Software"), to deal in the Software without restriction, including * without limitation the rights to use, copy, modify, merge, publish, * distribute, sublicense, and/or sell copies of the Software, and to * permit persons to whom the Software is furnished to do so, subject to * the following conditions: * * The above copyright notice and this permission notice shall be * included in all copies or substantial portions of the Software. * * THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, * EXPRESS OR IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF * MERCHANTABILITY, FITNESS FOR A PARTICULAR PURPOSE AND * NONINFRINGEMENT. IN NO EVENT SHALL THE AUTHORS OR COPYRIGHT HOLDERS BE * LIABLE FOR ANY CLAIM, DAMAGES OR OTHER LIABILITY, WHETHER IN AN ACTION * OF CONTRACT, TORT OR OTHERWISE, ARISING FROM, OUT OF OR IN CONNECTION * WITH THE SOFTWARE OR THE USE OR OTHER DEALINGS IN THE SOFTWARE. * */ /* machine-dependent cycle counters code. Needs to be inlined. */ /***************************************************************************/ /* To use the cycle counters in your code, simply #include "cycle.h" (this file), and then use the functions/macros: ticks getticks(void); ticks is an opaque typedef defined below, representing the current time. You extract the elapsed time between two calls to gettick() via: double elapsed(ticks t1, ticks t0); which returns a double-precision variable in arbitrary units. You are not expected to convert this into human units like seconds; it is intended only for *comparisons* of time intervals. (In order to use some of the OS-dependent timer routines like Solaris' gethrtime, you need to paste the autoconf snippet below into your configure.ac file and #include "config.h" before cycle.h, or define the relevant macros manually if you are not using autoconf.) */ /***************************************************************************/ /* This file uses macros like HAVE_GETHRTIME that are assumed to be defined according to whether the corresponding function/type/header is available on your system. The necessary macros are most conveniently defined if you are using GNU autoconf, via the tests: dnl --------------------------------------------------------------------- AC_C_INLINE AC_HEADER_TIME AC_CHECK_HEADERS([sys/time.h c_asm.h intrinsics.h mach/mach_time.h]) AC_CHECK_TYPE([hrtime_t],[AC_DEFINE(HAVE_HRTIME_T, 1, [Define to 1 if hrtime_t is defined in ])],,[#if HAVE_SYS_TIME_H #include #endif]) AC_CHECK_FUNCS([gethrtime read_real_time time_base_to_time clock_gettime mach_absolute_time]) dnl Cray UNICOS _rtc() (real-time clock) intrinsic AC_MSG_CHECKING([for _rtc intrinsic]) rtc_ok=yes AC_TRY_LINK([#ifdef HAVE_INTRINSICS_H #include #endif], [_rtc()], [AC_DEFINE(HAVE__RTC,1,[Define if you have the UNICOS _rtc() intrinsic.])], [rtc_ok=no]) AC_MSG_RESULT($rtc_ok) dnl --------------------------------------------------------------------- */ /***************************************************************************/ #if TIME_WITH_SYS_TIME # include # include #else # if HAVE_SYS_TIME_H # include # else # include # endif #endif #define INLINE_ELAPSED(INL) static INL double elapsed(ticks t1, ticks t0) \ { \ return (double)t1 - (double)t0; \ } /*----------------------------------------------------------------*/ /* Solaris */ #if defined(HAVE_GETHRTIME) && defined(HAVE_HRTIME_T) && !defined(HAVE_TICK_COUNTER) typedef hrtime_t ticks; #define getticks gethrtime INLINE_ELAPSED(inline) #define HAVE_TICK_COUNTER #endif /*----------------------------------------------------------------*/ /* AIX v. 4+ routines to read the real-time clock or time-base register */ #if defined(HAVE_READ_REAL_TIME) && defined(HAVE_TIME_BASE_TO_TIME) && !defined(HAVE_TICK_COUNTER) typedef timebasestruct_t ticks; static __inline ticks getticks(void) { ticks t; read_real_time(&t, TIMEBASE_SZ); return t; } static __inline double elapsed(ticks t1, ticks t0) /* time in nanoseconds */ { time_base_to_time(&t1, TIMEBASE_SZ); time_base_to_time(&t0, TIMEBASE_SZ); return (((double)t1.tb_high - (double)t0.tb_high) * 1.0e9 + ((double)t1.tb_low - (double)t0.tb_low)); } #define HAVE_TICK_COUNTER #endif /*----------------------------------------------------------------*/ /* * PowerPC ``cycle'' counter using the time base register. */ #if ((((defined(__GNUC__) && (defined(__powerpc__) || defined(__ppc__))) || (defined(__MWERKS__) && defined(macintosh)))) || (defined(__IBM_GCC_ASM) && (defined(__powerpc__) || defined(__ppc__)))) && !defined(HAVE_TICK_COUNTER) typedef unsigned long long ticks; static __inline__ ticks getticks(void) { unsigned int tbl, tbu0, tbu1; do { __asm__ __volatile__ ("mftbu %0" : "=r"(tbu0)); __asm__ __volatile__ ("mftb %0" : "=r"(tbl)); __asm__ __volatile__ ("mftbu %0" : "=r"(tbu1)); } while (tbu0 != tbu1); return (((unsigned long long)tbu0) << 32) | tbl; } INLINE_ELAPSED(__inline__) #define HAVE_TICK_COUNTER #endif /* MacOS/Mach (Darwin) time-base register interface (unlike UpTime, from Carbon, requires no additional libraries to be linked). */ #if defined(HAVE_MACH_ABSOLUTE_TIME) && defined(HAVE_MACH_MACH_TIME_H) && !defined(HAVE_TICK_COUNTER) #include typedef uint64_t ticks; #define getticks mach_absolute_time INLINE_ELAPSED(__inline__) #define HAVE_TICK_COUNTER #endif /*----------------------------------------------------------------*/ /* * Pentium cycle counter */ #if (defined(__GNUC__) || defined(__ICC)) && defined(__i386__) && !defined(HAVE_TICK_COUNTER) typedef unsigned long long ticks; static __inline__ ticks getticks(void) { ticks ret; __asm__ __volatile__("rdtsc": "=A" (ret)); /* no input, nothing else clobbered */ return ret; } INLINE_ELAPSED(__inline__) #define HAVE_TICK_COUNTER #define TIME_MIN 5000.0 /* unreliable pentium IV cycle counter */ #endif /* Visual C++ -- thanks to Morten Nissov for his help with this */ #if _MSC_VER >= 1200 && _M_IX86 >= 500 && !defined(HAVE_TICK_COUNTER) #include typedef LARGE_INTEGER ticks; #define RDTSC __asm __emit 0fh __asm __emit 031h /* hack for VC++ 5.0 */ static __inline ticks getticks(void) { ticks retval; __asm { RDTSC mov retval.HighPart, edx mov retval.LowPart, eax } return retval; } static __inline double elapsed(ticks t1, ticks t0) { return (double)t1.QuadPart - (double)t0.QuadPart; } #define HAVE_TICK_COUNTER #define TIME_MIN 5000.0 /* unreliable pentium IV cycle counter */ #endif /*----------------------------------------------------------------*/ /* * X86-64 cycle counter */ #if (defined(__GNUC__) || defined(__ICC) || defined(__SUNPRO_C)) && defined(__x86_64__) && !defined(HAVE_TICK_COUNTER) typedef unsigned long long ticks; static __inline__ ticks getticks(void) { unsigned a, d; asm volatile("rdtsc" : "=a" (a), "=d" (d)); return ((ticks)a) | (((ticks)d) << 32); } INLINE_ELAPSED(__inline__) #define HAVE_TICK_COUNTER #endif /* PGI compiler, courtesy Cristiano Calonaci, Andrea Tarsi, & Roberto Gori. NOTE: this code will fail to link unless you use the -Masmkeyword compiler option (grrr). */ #if defined(__PGI) && defined(__x86_64__) && !defined(HAVE_TICK_COUNTER) typedef unsigned long long ticks; static ticks getticks(void) { asm(" rdtsc; shl $0x20,%rdx; mov %eax,%eax; or %rdx,%rax; "); } INLINE_ELAPSED(__inline__) #define HAVE_TICK_COUNTER #endif /* Visual C++, courtesy of Dirk Michaelis */ #if _MSC_VER >= 1400 && (defined(_M_AMD64) || defined(_M_X64)) && !defined(HAVE_TICK_COUNTER) #include #pragma intrinsic(__rdtsc) typedef unsigned __int64 ticks; #define getticks __rdtsc INLINE_ELAPSED(__inline) #define HAVE_TICK_COUNTER #endif /*----------------------------------------------------------------*/ /* * IA64 cycle counter */ /* intel's icc/ecc compiler */ #if (defined(__EDG_VERSION) || defined(__ECC)) && defined(__ia64__) && !defined(HAVE_TICK_COUNTER) typedef unsigned long ticks; #include static __inline__ ticks getticks(void) { return __getReg(_IA64_REG_AR_ITC); } INLINE_ELAPSED(__inline__) #define HAVE_TICK_COUNTER #endif /* gcc */ #if defined(__GNUC__) && defined(__ia64__) && !defined(HAVE_TICK_COUNTER) typedef unsigned long ticks; static __inline__ ticks getticks(void) { ticks ret; __asm__ __volatile__ ("mov %0=ar.itc" : "=r"(ret)); return ret; } INLINE_ELAPSED(__inline__) #define HAVE_TICK_COUNTER #endif /* HP/UX IA64 compiler, courtesy Teresa L. Johnson: */ #if defined(__hpux) && defined(__ia64) && !defined(HAVE_TICK_COUNTER) #include typedef unsigned long ticks; static inline ticks getticks(void) { ticks ret; ret = _Asm_mov_from_ar (_AREG_ITC); return ret; } INLINE_ELAPSED(inline) #define HAVE_TICK_COUNTER #endif /* Microsoft Visual C++ */ #if defined(_MSC_VER) && defined(_M_IA64) && !defined(HAVE_TICK_COUNTER) typedef unsigned __int64 ticks; # ifdef __cplusplus extern "C" # endif ticks __getReg(int whichReg); #pragma intrinsic(__getReg) static __inline ticks getticks(void) { volatile ticks temp; temp = __getReg(3116); return temp; } INLINE_ELAPSED(inline) #define HAVE_TICK_COUNTER #endif /*----------------------------------------------------------------*/ /* * PA-RISC cycle counter */ #if defined(__hppa__) || defined(__hppa) && !defined(HAVE_TICK_COUNTER) typedef unsigned long ticks; # ifdef __GNUC__ static __inline__ ticks getticks(void) { ticks ret; __asm__ __volatile__("mfctl 16, %0": "=r" (ret)); /* no input, nothing else clobbered */ return ret; } # else # include static inline unsigned long getticks(void) { register ticks ret; _MFCTL(16, ret); return ret; } # endif INLINE_ELAPSED(inline) #define HAVE_TICK_COUNTER #endif /*----------------------------------------------------------------*/ /* S390, courtesy of James Treacy */ #if defined(__GNUC__) && defined(__s390__) && !defined(HAVE_TICK_COUNTER) typedef unsigned long long ticks; static __inline__ ticks getticks(void) { ticks cycles; __asm__("stck 0(%0)" : : "a" (&(cycles)) : "memory", "cc"); return cycles; } INLINE_ELAPSED(__inline__) #define HAVE_TICK_COUNTER #endif /*----------------------------------------------------------------*/ #if defined(__GNUC__) && defined(__alpha__) && !defined(HAVE_TICK_COUNTER) /* * The 32-bit cycle counter on alpha overflows pretty quickly, * unfortunately. A 1GHz machine overflows in 4 seconds. */ typedef unsigned int ticks; static __inline__ ticks getticks(void) { unsigned long cc; __asm__ __volatile__ ("rpcc %0" : "=r"(cc)); return (cc & 0xFFFFFFFF); } INLINE_ELAPSED(__inline__) #define HAVE_TICK_COUNTER #endif /*----------------------------------------------------------------*/ #if defined(__GNUC__) && defined(__sparc_v9__) && !defined(HAVE_TICK_COUNTER) typedef unsigned long ticks; static __inline__ ticks getticks(void) { ticks ret; __asm__ __volatile__("rd %%tick, %0" : "=r" (ret)); return ret; } INLINE_ELAPSED(__inline__) #define HAVE_TICK_COUNTER #endif /*----------------------------------------------------------------*/ #if (defined(__DECC) || defined(__DECCXX)) && defined(__alpha) && defined(HAVE_C_ASM_H) && !defined(HAVE_TICK_COUNTER) # include typedef unsigned int ticks; static __inline ticks getticks(void) { unsigned long cc; cc = asm("rpcc %v0"); return (cc & 0xFFFFFFFF); } INLINE_ELAPSED(__inline) #define HAVE_TICK_COUNTER #endif /*----------------------------------------------------------------*/ /* SGI/Irix */ #if defined(HAVE_CLOCK_GETTIME) && defined(CLOCK_SGI_CYCLE) && !defined(HAVE_TICK_COUNTER) typedef struct timespec ticks; static inline ticks getticks(void) { struct timespec t; clock_gettime(CLOCK_SGI_CYCLE, &t); return t; } static inline double elapsed(ticks t1, ticks t0) { return ((double)t1.tv_sec - (double)t0.tv_sec) * 1.0E9 + ((double)t1.tv_nsec - (double)t0.tv_nsec); } #define HAVE_TICK_COUNTER #endif /*----------------------------------------------------------------*/ /* Cray UNICOS _rtc() intrinsic function */ #if defined(HAVE__RTC) && !defined(HAVE_TICK_COUNTER) #ifdef HAVE_INTRINSICS_H # include #endif typedef long long ticks; #define getticks _rtc INLINE_ELAPSED(inline) #define HAVE_TICK_COUNTER #endif /*----------------------------------------------------------------*/ /* MIPS ZBus */ #if HAVE_MIPS_ZBUS_TIMER #if defined(__mips__) && !defined(HAVE_TICK_COUNTER) #include #include #include typedef uint64_t ticks; static inline ticks getticks(void) { static uint64_t* addr = 0; if (addr == 0) { uint32_t rq_addr = 0x10030000; int fd; int pgsize; pgsize = getpagesize(); fd = open ("/dev/mem", O_RDONLY | O_SYNC, 0); if (fd < 0) { perror("open"); return NULL; } addr = mmap(0, pgsize, PROT_READ, MAP_SHARED, fd, rq_addr); close(fd); if (addr == (uint64_t *)-1) { perror("mmap"); return NULL; } } return *addr; } INLINE_ELAPSED(inline) #define HAVE_TICK_COUNTER #endif #endif /* HAVE_MIPS_ZBUS_TIMER */ fftw-3.3.3/kernel/ifftw.h0000644000175000017500000010360612054410050012154 00000000000000/* * Copyright (c) 2003, 2007-11 Matteo Frigo * Copyright (c) 2003, 2007-11 Massachusetts Institute of Technology * * This program is free software; you can redistribute it and/or modify * it under the terms of the GNU General Public License as published by * the Free Software Foundation; either version 2 of the License, or * (at your option) any later version. * * This program is distributed in the hope that it will be useful, * but WITHOUT ANY WARRANTY; without even the implied warranty of * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the * GNU General Public License for more details. * * You should have received a copy of the GNU General Public License * along with this program; if not, write to the Free Software * Foundation, Inc., 51 Franklin Street, Fifth Floor, Boston, MA 02110-1301 USA * */ /* FFTW internal header file */ #ifndef __IFFTW_H__ #define __IFFTW_H__ #include "config.h" #include /* size_t */ #include /* va_list */ #include /* ptrdiff_t */ #if HAVE_SYS_TYPES_H # include #endif #if HAVE_STDINT_H # include /* uintptr_t, maybe */ #endif #if HAVE_INTTYPES_H # include /* uintptr_t, maybe */ #endif #ifdef __cplusplus extern "C" { #endif /* __cplusplus */ /* Windows annoyances -- since tests/hook.c uses some internal FFTW functions, we need to given them the dllexport attribute under Windows when compiling as a DLL (see api/fftw3.h). */ #if defined(FFTW_EXTERN) # define IFFTW_EXTERN FFTW_EXTERN #elif (defined(FFTW_DLL) || defined(DLL_EXPORT)) \ && (defined(_WIN32) || defined(__WIN32__)) # define IFFTW_EXTERN extern __declspec(dllexport) #else # define IFFTW_EXTERN extern #endif /* determine precision and name-mangling scheme */ #define CONCAT(prefix, name) prefix ## name #if defined(FFTW_SINGLE) typedef float R; # define X(name) CONCAT(fftwf_, name) #elif defined(FFTW_LDOUBLE) typedef long double R; # define X(name) CONCAT(fftwl_, name) # define TRIGREAL_IS_LONG_DOUBLE #elif defined(FFTW_QUAD) typedef __float128 R; # define X(name) CONCAT(fftwq_, name) # define TRIGREAL_IS_QUAD #else typedef double R; # define X(name) CONCAT(fftw_, name) #endif /* integral type large enough to contain a stride (what ``int'' should have been in the first place. */ typedef ptrdiff_t INT; /* dummy use of unused parameters to silence compiler warnings */ #define UNUSED(x) (void)x #define NELEM(array) ((int) (sizeof(array) / sizeof((array)[0]))) #define FFT_SIGN (-1) /* sign convention for forward transforms */ extern void X(extract_reim)(int sign, R *c, R **r, R **i); #define REGISTER_SOLVER(p, s) X(solver_register)(p, s) #define STRINGIZEx(x) #x #define STRINGIZE(x) STRINGIZEx(x) #define CIMPLIES(ante, post) (!(ante) || (post)) /* define HAVE_SIMD if any simd extensions are supported */ #if defined(HAVE_SSE) || defined(HAVE_SSE2) || defined(HAVE_ALTIVEC) || \ defined(HAVE_MIPS_PS) || defined(HAVE_AVX) #define HAVE_SIMD 1 #else #define HAVE_SIMD 0 #endif extern int X(have_simd_sse2)(void); extern int X(have_simd_avx)(void); extern int X(have_simd_altivec)(void); extern int X(have_simd_neon)(void); /* forward declarations */ typedef struct problem_s problem; typedef struct plan_s plan; typedef struct solver_s solver; typedef struct planner_s planner; typedef struct printer_s printer; typedef struct scanner_s scanner; /*-----------------------------------------------------------------------*/ /* alloca: */ #if HAVE_SIMD # ifdef HAVE_AVX # define MIN_ALIGNMENT 32 /* best alignment for AVX, conservative for * everything else */ # else /* Note that we cannot use 32-byte alignment for all SIMD. For example, MacOS X malloc is 16-byte aligned, but there was no posix_memalign in MacOS X until version 10.6. */ # define MIN_ALIGNMENT 16 # endif #endif #if defined(HAVE_ALLOCA) && defined(FFTW_ENABLE_ALLOCA) /* use alloca if available */ #ifndef alloca #ifdef __GNUC__ # define alloca __builtin_alloca #else # ifdef _MSC_VER # include # define alloca _alloca # else # if HAVE_ALLOCA_H # include # else # ifdef _AIX #pragma alloca # else # ifndef alloca /* predefined by HP cc +Olibcalls */ void *alloca(size_t); # endif # endif # endif # endif #endif #endif # ifdef MIN_ALIGNMENT # define STACK_MALLOC(T, p, n) \ { \ p = (T)alloca((n) + MIN_ALIGNMENT); \ p = (T)(((uintptr_t)p + (MIN_ALIGNMENT - 1)) & \ (~(uintptr_t)(MIN_ALIGNMENT - 1))); \ } # define STACK_FREE(n) # else /* HAVE_ALLOCA && !defined(MIN_ALIGNMENT) */ # define STACK_MALLOC(T, p, n) p = (T)alloca(n) # define STACK_FREE(n) # endif #else /* ! HAVE_ALLOCA */ /* use malloc instead of alloca */ # define STACK_MALLOC(T, p, n) p = (T)MALLOC(n, OTHER) # define STACK_FREE(n) X(ifree)(n) #endif /* ! HAVE_ALLOCA */ /* allocation of buffers. If these grow too large use malloc(), else use STACK_MALLOC (hopefully reducing to alloca()). */ /* 64KiB ought to be enough for anybody */ #define MAX_STACK_ALLOC ((size_t)64 * 1024) #define BUF_ALLOC(T, p, n) \ { \ if (n < MAX_STACK_ALLOC) { \ STACK_MALLOC(T, p, n); \ } else { \ p = (T)MALLOC(n, BUFFERS); \ } \ } #define BUF_FREE(p, n) \ { \ if (n < MAX_STACK_ALLOC) { \ STACK_FREE(p); \ } else { \ X(ifree)(p); \ } \ } /*-----------------------------------------------------------------------*/ /* define uintptr_t if it is not already defined */ #ifndef HAVE_UINTPTR_T # if SIZEOF_VOID_P == 0 # error sizeof void* is unknown! # elif SIZEOF_UNSIGNED_INT == SIZEOF_VOID_P typedef unsigned int uintptr_t; # elif SIZEOF_UNSIGNED_LONG == SIZEOF_VOID_P typedef unsigned long uintptr_t; # elif SIZEOF_UNSIGNED_LONG_LONG == SIZEOF_VOID_P typedef unsigned long long uintptr_t; # else # error no unsigned integer type matches void* sizeof! # endif #endif /*-----------------------------------------------------------------------*/ /* We can do an optimization for copying pairs of (aligned) floats when in single precision if 2*float = double. */ #define FFTW_2R_IS_DOUBLE (defined(FFTW_SINGLE) \ && SIZEOF_FLOAT != 0 \ && SIZEOF_DOUBLE == 2*SIZEOF_FLOAT) #define DOUBLE_ALIGNED(p) ((((uintptr_t)(p)) % sizeof(double)) == 0) /*-----------------------------------------------------------------------*/ /* assert.c: */ IFFTW_EXTERN void X(assertion_failed)(const char *s, int line, const char *file); /* always check */ #define CK(ex) \ (void)((ex) || (X(assertion_failed)(#ex, __LINE__, __FILE__), 0)) #ifdef FFTW_DEBUG /* check only if debug enabled */ #define A(ex) \ (void)((ex) || (X(assertion_failed)(#ex, __LINE__, __FILE__), 0)) #else #define A(ex) /* nothing */ #endif extern void X(debug)(const char *format, ...); #define D X(debug) /*-----------------------------------------------------------------------*/ /* kalloc.c: */ extern void *X(kernel_malloc)(size_t n); extern void X(kernel_free)(void *p); /*-----------------------------------------------------------------------*/ /* alloc.c: */ /* objects allocated by malloc, for statistical purposes */ enum malloc_tag { EVERYTHING, PLANS, SOLVERS, PROBLEMS, BUFFERS, HASHT, TENSORS, PLANNERS, SLVDESCS, TWIDDLES, STRIDES, OTHER, MALLOC_WHAT_LAST /* must be last */ }; IFFTW_EXTERN void X(ifree)(void *ptr); extern void X(ifree0)(void *ptr); #ifdef FFTW_DEBUG_MALLOC IFFTW_EXTERN void *X(malloc_debug)(size_t n, enum malloc_tag what, const char *file, int line); #define MALLOC(n, what) X(malloc_debug)(n, what, __FILE__, __LINE__) IFFTW_EXTERN void X(malloc_print_minfo)(int vrbose); #else /* ! FFTW_DEBUG_MALLOC */ IFFTW_EXTERN void *X(malloc_plain)(size_t sz); #define MALLOC(n, what) X(malloc_plain)(n) #endif #if defined(FFTW_DEBUG) && defined(FFTW_DEBUG_MALLOC) && (defined(HAVE_THREADS) || defined(HAVE_OPENMP)) extern int X(in_thread); # define IN_THREAD X(in_thread) # define THREAD_ON { int in_thread_save = X(in_thread); X(in_thread) = 1 # define THREAD_OFF X(in_thread) = in_thread_save; } #else # define IN_THREAD 0 # define THREAD_ON # define THREAD_OFF #endif /*-----------------------------------------------------------------------*/ /* low-resolution clock */ #ifdef FAKE_CRUDE_TIME typedef int crude_time; #else # if TIME_WITH_SYS_TIME # include # include # else # if HAVE_SYS_TIME_H # include # else # include # endif # endif # ifdef HAVE_BSDGETTIMEOFDAY # ifndef HAVE_GETTIMEOFDAY # define gettimeofday BSDgettimeofday # define HAVE_GETTIMEOFDAY 1 # endif # endif # if defined(HAVE_GETTIMEOFDAY) typedef struct timeval crude_time; # else typedef clock_t crude_time; # endif #endif /* else FAKE_CRUDE_TIME */ crude_time X(get_crude_time)(void); double X(elapsed_since)(const planner *plnr, const problem *p, crude_time t0); /* time in seconds since t0 */ /*-----------------------------------------------------------------------*/ /* ops.c: */ /* * ops counter. The total number of additions is add + fma * and the total number of multiplications is mul + fma. * Total flops = add + mul + 2 * fma */ typedef struct { double add; double mul; double fma; double other; } opcnt; void X(ops_zero)(opcnt *dst); void X(ops_other)(INT o, opcnt *dst); void X(ops_cpy)(const opcnt *src, opcnt *dst); void X(ops_add)(const opcnt *a, const opcnt *b, opcnt *dst); void X(ops_add2)(const opcnt *a, opcnt *dst); /* dst = m * a + b */ void X(ops_madd)(INT m, const opcnt *a, const opcnt *b, opcnt *dst); /* dst += m * a */ void X(ops_madd2)(INT m, const opcnt *a, opcnt *dst); /*-----------------------------------------------------------------------*/ /* minmax.c: */ INT X(imax)(INT a, INT b); INT X(imin)(INT a, INT b); /*-----------------------------------------------------------------------*/ /* iabs.c: */ INT X(iabs)(INT a); /* inline version */ #define IABS(x) (((x) < 0) ? (0 - (x)) : (x)) /*-----------------------------------------------------------------------*/ /* md5.c */ #if SIZEOF_UNSIGNED_INT >= 4 typedef unsigned int md5uint; #else typedef unsigned long md5uint; /* at least 32 bits as per C standard */ #endif typedef md5uint md5sig[4]; typedef struct { md5sig s; /* state and signature */ /* fields not meant to be used outside md5.c: */ unsigned char c[64]; /* stuff not yet processed */ unsigned l; /* total length. Should be 64 bits long, but this is good enough for us */ } md5; void X(md5begin)(md5 *p); void X(md5putb)(md5 *p, const void *d_, size_t len); void X(md5puts)(md5 *p, const char *s); void X(md5putc)(md5 *p, unsigned char c); void X(md5int)(md5 *p, int i); void X(md5INT)(md5 *p, INT i); void X(md5unsigned)(md5 *p, unsigned i); void X(md5end)(md5 *p); /*-----------------------------------------------------------------------*/ /* tensor.c: */ #define STRUCT_HACK_KR #undef STRUCT_HACK_C99 typedef struct { INT n; INT is; /* input stride */ INT os; /* output stride */ } iodim; typedef struct { int rnk; #if defined(STRUCT_HACK_KR) iodim dims[1]; #elif defined(STRUCT_HACK_C99) iodim dims[]; #else iodim *dims; #endif } tensor; /* Definition of rank -infinity. This definition has the property that if you want rank 0 or 1, you can simply test for rank <= 1. This is a common case. A tensor of rank -infinity has size 0. */ #define RNK_MINFTY ((int)(((unsigned) -1) >> 1)) #define FINITE_RNK(rnk) ((rnk) != RNK_MINFTY) typedef enum { INPLACE_IS, INPLACE_OS } inplace_kind; tensor *X(mktensor)(int rnk); tensor *X(mktensor_0d)(void); tensor *X(mktensor_1d)(INT n, INT is, INT os); tensor *X(mktensor_2d)(INT n0, INT is0, INT os0, INT n1, INT is1, INT os1); tensor *X(mktensor_3d)(INT n0, INT is0, INT os0, INT n1, INT is1, INT os1, INT n2, INT is2, INT os2); tensor *X(mktensor_4d)(INT n0, INT is0, INT os0, INT n1, INT is1, INT os1, INT n2, INT is2, INT os2, INT n3, INT is3, INT os3); tensor *X(mktensor_5d)(INT n0, INT is0, INT os0, INT n1, INT is1, INT os1, INT n2, INT is2, INT os2, INT n3, INT is3, INT os3, INT n4, INT is4, INT os4); INT X(tensor_sz)(const tensor *sz); void X(tensor_md5)(md5 *p, const tensor *t); INT X(tensor_max_index)(const tensor *sz); INT X(tensor_min_istride)(const tensor *sz); INT X(tensor_min_ostride)(const tensor *sz); INT X(tensor_min_stride)(const tensor *sz); int X(tensor_inplace_strides)(const tensor *sz); int X(tensor_inplace_strides2)(const tensor *a, const tensor *b); int X(tensor_strides_decrease)(const tensor *sz, const tensor *vecsz, inplace_kind k); tensor *X(tensor_copy)(const tensor *sz); int X(tensor_kosherp)(const tensor *x); tensor *X(tensor_copy_inplace)(const tensor *sz, inplace_kind k); tensor *X(tensor_copy_except)(const tensor *sz, int except_dim); tensor *X(tensor_copy_sub)(const tensor *sz, int start_dim, int rnk); tensor *X(tensor_compress)(const tensor *sz); tensor *X(tensor_compress_contiguous)(const tensor *sz); tensor *X(tensor_append)(const tensor *a, const tensor *b); void X(tensor_split)(const tensor *sz, tensor **a, int a_rnk, tensor **b); int X(tensor_tornk1)(const tensor *t, INT *n, INT *is, INT *os); void X(tensor_destroy)(tensor *sz); void X(tensor_destroy2)(tensor *a, tensor *b); void X(tensor_destroy4)(tensor *a, tensor *b, tensor *c, tensor *d); void X(tensor_print)(const tensor *sz, printer *p); int X(dimcmp)(const iodim *a, const iodim *b); int X(tensor_equal)(const tensor *a, const tensor *b); int X(tensor_inplace_locations)(const tensor *sz, const tensor *vecsz); /*-----------------------------------------------------------------------*/ /* problem.c: */ enum { /* a problem that cannot be solved */ PROBLEM_UNSOLVABLE, PROBLEM_DFT, PROBLEM_RDFT, PROBLEM_RDFT2, /* for mpi/ subdirectory */ PROBLEM_MPI_DFT, PROBLEM_MPI_RDFT, PROBLEM_MPI_RDFT2, PROBLEM_MPI_TRANSPOSE, PROBLEM_LAST }; typedef struct { int problem_kind; void (*hash) (const problem *ego, md5 *p); void (*zero) (const problem *ego); void (*print) (const problem *ego, printer *p); void (*destroy) (problem *ego); } problem_adt; struct problem_s { const problem_adt *adt; }; problem *X(mkproblem)(size_t sz, const problem_adt *adt); void X(problem_destroy)(problem *ego); problem *X(mkproblem_unsolvable)(void); /*-----------------------------------------------------------------------*/ /* print.c */ struct printer_s { void (*print)(printer *p, const char *format, ...); void (*vprint)(printer *p, const char *format, va_list ap); void (*putchr)(printer *p, char c); void (*cleanup)(printer *p); int indent; int indent_incr; }; printer *X(mkprinter)(size_t size, void (*putchr)(printer *p, char c), void (*cleanup)(printer *p)); IFFTW_EXTERN void X(printer_destroy)(printer *p); /*-----------------------------------------------------------------------*/ /* scan.c */ struct scanner_s { int (*scan)(scanner *sc, const char *format, ...); int (*vscan)(scanner *sc, const char *format, va_list ap); int (*getchr)(scanner *sc); int ungotc; }; scanner *X(mkscanner)(size_t size, int (*getchr)(scanner *sc)); void X(scanner_destroy)(scanner *sc); /*-----------------------------------------------------------------------*/ /* plan.c: */ enum wakefulness { SLEEPY, AWAKE_ZERO, AWAKE_SQRTN_TABLE, AWAKE_SINCOS }; typedef struct { void (*solve)(const plan *ego, const problem *p); void (*awake)(plan *ego, enum wakefulness wakefulness); void (*print)(const plan *ego, printer *p); void (*destroy)(plan *ego); } plan_adt; struct plan_s { const plan_adt *adt; opcnt ops; double pcost; enum wakefulness wakefulness; /* used for debugging only */ int could_prune_now_p; }; plan *X(mkplan)(size_t size, const plan_adt *adt); void X(plan_destroy_internal)(plan *ego); IFFTW_EXTERN void X(plan_awake)(plan *ego, enum wakefulness wakefulness); void X(plan_null_destroy)(plan *ego); /*-----------------------------------------------------------------------*/ /* solver.c: */ typedef struct { int problem_kind; plan *(*mkplan)(const solver *ego, const problem *p, planner *plnr); void (*destroy)(solver *ego); } solver_adt; struct solver_s { const solver_adt *adt; int refcnt; }; solver *X(mksolver)(size_t size, const solver_adt *adt); void X(solver_use)(solver *ego); void X(solver_destroy)(solver *ego); void X(solver_register)(planner *plnr, solver *s); /* shorthand */ #define MKSOLVER(type, adt) (type *)X(mksolver)(sizeof(type), adt) /*-----------------------------------------------------------------------*/ /* planner.c */ typedef struct slvdesc_s { solver *slv; const char *reg_nam; unsigned nam_hash; int reg_id; int next_for_same_problem_kind; } slvdesc; typedef struct solution_s solution; /* opaque */ /* interpretation of L and U: - if it returns a plan, the planner guarantees that all applicable plans at least as impatient as U have been tried, and that each plan in the solution is at least as impatient as L. - if it returns 0, the planner guarantees to have tried all solvers at least as impatient as L, and that none of them was applicable. The structure is packed to fit into 64 bits. */ typedef struct { unsigned l:20; unsigned hash_info:3; # define BITS_FOR_TIMELIMIT 9 unsigned timelimit_impatience:BITS_FOR_TIMELIMIT; unsigned u:20; /* abstraction break: we store the solver here to pad the structure to 64 bits. Otherwise, the struct is padded to 64 bits anyway, and another word is allocated for slvndx. */ # define BITS_FOR_SLVNDX 12 unsigned slvndx:BITS_FOR_SLVNDX; } flags_t; /* impatience flags */ enum { BELIEVE_PCOST = 0x0001, ESTIMATE = 0x0002, NO_DFT_R2HC = 0x0004, NO_SLOW = 0x0008, NO_VRECURSE = 0x0010, NO_INDIRECT_OP = 0x0020, NO_LARGE_GENERIC = 0x0040, NO_RANK_SPLITS = 0x0080, NO_VRANK_SPLITS = 0x0100, NO_NONTHREADED = 0x0200, NO_BUFFERING = 0x0400, NO_FIXED_RADIX_LARGE_N = 0x0800, NO_DESTROY_INPUT = 0x1000, NO_SIMD = 0x2000, CONSERVE_MEMORY = 0x4000, NO_DHT_R2HC = 0x8000, NO_UGLY = 0x10000, ALLOW_PRUNING = 0x20000 }; /* hashtable information */ enum { BLESSING = 0x1, /* save this entry */ H_VALID = 0x2, /* valid hastable entry */ H_LIVE = 0x4 /* entry is nonempty, implies H_VALID */ }; #define PLNR_L(plnr) ((plnr)->flags.l) #define PLNR_U(plnr) ((plnr)->flags.u) #define PLNR_TIMELIMIT_IMPATIENCE(plnr) ((plnr)->flags.timelimit_impatience) #define ESTIMATEP(plnr) (PLNR_U(plnr) & ESTIMATE) #define BELIEVE_PCOSTP(plnr) (PLNR_U(plnr) & BELIEVE_PCOST) #define ALLOW_PRUNINGP(plnr) (PLNR_U(plnr) & ALLOW_PRUNING) #define NO_INDIRECT_OP_P(plnr) (PLNR_L(plnr) & NO_INDIRECT_OP) #define NO_LARGE_GENERICP(plnr) (PLNR_L(plnr) & NO_LARGE_GENERIC) #define NO_RANK_SPLITSP(plnr) (PLNR_L(plnr) & NO_RANK_SPLITS) #define NO_VRANK_SPLITSP(plnr) (PLNR_L(plnr) & NO_VRANK_SPLITS) #define NO_VRECURSEP(plnr) (PLNR_L(plnr) & NO_VRECURSE) #define NO_DFT_R2HCP(plnr) (PLNR_L(plnr) & NO_DFT_R2HC) #define NO_SLOWP(plnr) (PLNR_L(plnr) & NO_SLOW) #define NO_UGLYP(plnr) (PLNR_L(plnr) & NO_UGLY) #define NO_FIXED_RADIX_LARGE_NP(plnr) \ (PLNR_L(plnr) & NO_FIXED_RADIX_LARGE_N) #define NO_NONTHREADEDP(plnr) \ ((PLNR_L(plnr) & NO_NONTHREADED) && (plnr)->nthr > 1) #define NO_DESTROY_INPUTP(plnr) (PLNR_L(plnr) & NO_DESTROY_INPUT) #define NO_SIMDP(plnr) (PLNR_L(plnr) & NO_SIMD) #define CONSERVE_MEMORYP(plnr) (PLNR_L(plnr) & CONSERVE_MEMORY) #define NO_DHT_R2HCP(plnr) (PLNR_L(plnr) & NO_DHT_R2HC) #define NO_BUFFERINGP(plnr) (PLNR_L(plnr) & NO_BUFFERING) typedef enum { FORGET_ACCURSED, FORGET_EVERYTHING } amnesia; typedef enum { /* WISDOM_NORMAL: planner may or may not use wisdom */ WISDOM_NORMAL, /* WISDOM_ONLY: planner must use wisdom and must avoid searching */ WISDOM_ONLY, /* WISDOM_IS_BOGUS: planner must return 0 as quickly as possible */ WISDOM_IS_BOGUS, /* WISDOM_IGNORE_INFEASIBLE: planner ignores infeasible wisdom */ WISDOM_IGNORE_INFEASIBLE, /* WISDOM_IGNORE_ALL: planner ignores all */ WISDOM_IGNORE_ALL } wisdom_state_t; typedef struct { void (*register_solver)(planner *ego, solver *s); plan *(*mkplan)(planner *ego, const problem *p); void (*forget)(planner *ego, amnesia a); void (*exprt)(planner *ego, printer *p); /* ``export'' is a reserved word in C++. */ int (*imprt)(planner *ego, scanner *sc); } planner_adt; /* hash table of solutions */ typedef struct { solution *solutions; unsigned hashsiz, nelem; /* statistics */ int lookup, succ_lookup, lookup_iter; int insert, insert_iter, insert_unknown; int nrehash; } hashtab; typedef enum { COST_SUM, COST_MAX } cost_kind; struct planner_s { const planner_adt *adt; void (*hook)(struct planner_s *plnr, plan *pln, const problem *p, int optimalp); double (*cost_hook)(const problem *p, double t, cost_kind k); int (*wisdom_ok_hook)(const problem *p, flags_t flags); void (*nowisdom_hook)(const problem *p); wisdom_state_t (*bogosity_hook)(wisdom_state_t state, const problem *p); /* solver descriptors */ slvdesc *slvdescs; unsigned nslvdesc, slvdescsiz; const char *cur_reg_nam; int cur_reg_id; int slvdescs_for_problem_kind[PROBLEM_LAST]; wisdom_state_t wisdom_state; hashtab htab_blessed; hashtab htab_unblessed; int nthr; flags_t flags; crude_time start_time; double timelimit; /* elapsed_since(start_time) at which to bail out */ int timed_out; /* whether most recent search timed out */ int need_timeout_check; /* various statistics */ int nplan; /* number of plans evaluated */ double pcost, epcost; /* total pcost of measured/estimated plans */ int nprob; /* number of problems evaluated */ }; planner *X(mkplanner)(void); void X(planner_destroy)(planner *ego); /* Iterate over all solvers. Read: @article{ baker93iterators, author = "Henry G. Baker, Jr.", title = "Iterators: Signs of Weakness in Object-Oriented Languages", journal = "{ACM} {OOPS} Messenger", volume = "4", number = "3", pages = "18--25" } */ #define FORALL_SOLVERS(ego, s, p, what) \ { \ unsigned _cnt; \ for (_cnt = 0; _cnt < ego->nslvdesc; ++_cnt) { \ slvdesc *p = ego->slvdescs + _cnt; \ solver *s = p->slv; \ what; \ } \ } #define FORALL_SOLVERS_OF_KIND(kind, ego, s, p, what) \ { \ int _cnt = ego->slvdescs_for_problem_kind[kind]; \ while (_cnt >= 0) { \ slvdesc *p = ego->slvdescs + _cnt; \ solver *s = p->slv; \ what; \ _cnt = p->next_for_same_problem_kind; \ } \ } /* make plan, destroy problem */ plan *X(mkplan_d)(planner *ego, problem *p); plan *X(mkplan_f_d)(planner *ego, problem *p, unsigned l_set, unsigned u_set, unsigned u_reset); /*-----------------------------------------------------------------------*/ /* stride.c: */ /* If PRECOMPUTE_ARRAY_INDICES is defined, precompute all strides. */ #if (defined(__i386__) || defined(__x86_64__) || _M_IX86 >= 500) && !defined(FFTW_LDOUBLE) #define PRECOMPUTE_ARRAY_INDICES #endif extern const INT X(an_INT_guaranteed_to_be_zero); #ifdef PRECOMPUTE_ARRAY_INDICES typedef INT *stride; #define WS(stride, i) (stride[i]) extern stride X(mkstride)(INT n, INT s); void X(stride_destroy)(stride p); /* hackery to prevent the compiler from copying the strides array onto the stack */ #define MAKE_VOLATILE_STRIDE(nptr, x) (x) = (x) + X(an_INT_guaranteed_to_be_zero) #else typedef INT stride; #define WS(stride, i) (stride * i) #define fftwf_mkstride(n, stride) stride #define fftw_mkstride(n, stride) stride #define fftwl_mkstride(n, stride) stride #define fftwf_stride_destroy(p) ((void) p) #define fftw_stride_destroy(p) ((void) p) #define fftwl_stride_destroy(p) ((void) p) /* hackery to prevent the compiler from ``optimizing'' induction variables in codelet loops. The problem is that for each K and for each expression of the form P[I + STRIDE * K] in a loop, most compilers will try to lift an induction variable PK := &P[I + STRIDE * K]. For large values of K this behavior overflows the register set, which is likely worse than doing the index computation in the first place. If we guess that there are more than ESTIMATED_AVAILABLE_INDEX_REGISTERS such pointers, we deliberately confuse the compiler by setting STRIDE ^= ZERO, where ZERO is a value guaranteed to be 0, but the compiler does not know this. 16 registers ought to be enough for anybody, or so the amd64 and ARM ISA's seem to imply. */ #define ESTIMATED_AVAILABLE_INDEX_REGISTERS 16 #define MAKE_VOLATILE_STRIDE(nptr, x) \ (nptr <= ESTIMATED_AVAILABLE_INDEX_REGISTERS ? \ 0 : \ ((x) = (x) ^ X(an_INT_guaranteed_to_be_zero))) #endif /* PRECOMPUTE_ARRAY_INDICES */ /*-----------------------------------------------------------------------*/ /* solvtab.c */ struct solvtab_s { void (*reg)(planner *); const char *reg_nam; }; typedef struct solvtab_s solvtab[]; void X(solvtab_exec)(const solvtab tbl, planner *p); #define SOLVTAB(s) { s, STRINGIZE(s) } #define SOLVTAB_END { 0, 0 } /*-----------------------------------------------------------------------*/ /* pickdim.c */ int X(pickdim)(int which_dim, const int *buddies, int nbuddies, const tensor *sz, int oop, int *dp); /*-----------------------------------------------------------------------*/ /* twiddle.c */ /* little language to express twiddle factors computation */ enum { TW_COS = 0, TW_SIN = 1, TW_CEXP = 2, TW_NEXT = 3, TW_FULL = 4, TW_HALF = 5 }; typedef struct { unsigned char op; signed char v; short i; } tw_instr; typedef struct twid_s { R *W; /* array of twiddle factors */ INT n, r, m; /* transform order, radix, # twiddle rows */ int refcnt; const tw_instr *instr; struct twid_s *cdr; enum wakefulness wakefulness; } twid; INT X(twiddle_length)(INT r, const tw_instr *p); void X(twiddle_awake)(enum wakefulness wakefulness, twid **pp, const tw_instr *instr, INT n, INT r, INT m); /*-----------------------------------------------------------------------*/ /* trig.c */ #if defined(TRIGREAL_IS_LONG_DOUBLE) typedef long double trigreal; #elif defined(TRIGREAL_IS_QUAD) typedef __float128 trigreal; #else typedef double trigreal; #endif typedef struct triggen_s triggen; struct triggen_s { void (*cexp)(triggen *t, INT m, R *result); void (*cexpl)(triggen *t, INT m, trigreal *result); void (*rotate)(triggen *p, INT m, R xr, R xi, R *res); INT twshft; INT twradix; INT twmsk; trigreal *W0, *W1; INT n; }; triggen *X(mktriggen)(enum wakefulness wakefulness, INT n); void X(triggen_destroy)(triggen *p); /*-----------------------------------------------------------------------*/ /* primes.c: */ #define MULMOD(x, y, p) \ (((x) <= 92681 - (y)) ? ((x) * (y)) % (p) : X(safe_mulmod)(x, y, p)) INT X(safe_mulmod)(INT x, INT y, INT p); INT X(power_mod)(INT n, INT m, INT p); INT X(find_generator)(INT p); INT X(first_divisor)(INT n); int X(is_prime)(INT n); INT X(next_prime)(INT n); int X(factors_into)(INT n, const INT *primes); int X(factors_into_small_primes)(INT n); INT X(choose_radix)(INT r, INT n); INT X(isqrt)(INT n); INT X(modulo)(INT a, INT n); #define GENERIC_MIN_BAD 173 /* min prime for which generic becomes bad */ /* thresholds below which certain solvers are considered SLOW. These are guesses believed to be conservative */ #define GENERIC_MAX_SLOW 16 #define RADER_MAX_SLOW 32 #define BLUESTEIN_MAX_SLOW 24 /*-----------------------------------------------------------------------*/ /* rader.c: */ typedef struct rader_tls rader_tl; void X(rader_tl_insert)(INT k1, INT k2, INT k3, R *W, rader_tl **tl); R *X(rader_tl_find)(INT k1, INT k2, INT k3, rader_tl *t); void X(rader_tl_delete)(R *W, rader_tl **tl); /*-----------------------------------------------------------------------*/ /* copy/transposition routines */ /* lower bound to the cache size, for tiled routines */ #define CACHESIZE 8192 INT X(compute_tilesz)(INT vl, int how_many_tiles_in_cache); void X(tile2d)(INT n0l, INT n0u, INT n1l, INT n1u, INT tilesz, void (*f)(INT n0l, INT n0u, INT n1l, INT n1u, void *args), void *args); void X(cpy1d)(R *I, R *O, INT n0, INT is0, INT os0, INT vl); void X(cpy2d)(R *I, R *O, INT n0, INT is0, INT os0, INT n1, INT is1, INT os1, INT vl); void X(cpy2d_ci)(R *I, R *O, INT n0, INT is0, INT os0, INT n1, INT is1, INT os1, INT vl); void X(cpy2d_co)(R *I, R *O, INT n0, INT is0, INT os0, INT n1, INT is1, INT os1, INT vl); void X(cpy2d_tiled)(R *I, R *O, INT n0, INT is0, INT os0, INT n1, INT is1, INT os1, INT vl); void X(cpy2d_tiledbuf)(R *I, R *O, INT n0, INT is0, INT os0, INT n1, INT is1, INT os1, INT vl); void X(cpy2d_pair)(R *I0, R *I1, R *O0, R *O1, INT n0, INT is0, INT os0, INT n1, INT is1, INT os1); void X(cpy2d_pair_ci)(R *I0, R *I1, R *O0, R *O1, INT n0, INT is0, INT os0, INT n1, INT is1, INT os1); void X(cpy2d_pair_co)(R *I0, R *I1, R *O0, R *O1, INT n0, INT is0, INT os0, INT n1, INT is1, INT os1); void X(transpose)(R *I, INT n, INT s0, INT s1, INT vl); void X(transpose_tiled)(R *I, INT n, INT s0, INT s1, INT vl); void X(transpose_tiledbuf)(R *I, INT n, INT s0, INT s1, INT vl); typedef void (*transpose_func)(R *I, INT n, INT s0, INT s1, INT vl); typedef void (*cpy2d_func)(R *I, R *O, INT n0, INT is0, INT os0, INT n1, INT is1, INT os1, INT vl); /*-----------------------------------------------------------------------*/ /* misc stuff */ void X(null_awake)(plan *ego, enum wakefulness wakefulness); double X(iestimate_cost)(const planner *, const plan *, const problem *); #ifdef FFTW_RANDOM_ESTIMATOR extern unsigned X(random_estimate_seed); #endif double X(measure_execution_time)(const planner *plnr, plan *pln, const problem *p); int X(alignment_of)(R *p); unsigned X(hash)(const char *s); INT X(nbuf)(INT n, INT vl, INT maxnbuf); int X(nbuf_redundant)(INT n, INT vl, int which, const INT *maxnbuf, int nmaxnbuf); INT X(bufdist)(INT n, INT vl); int X(toobig)(INT n); int X(ct_uglyp)(INT min_n, INT v, INT n, INT r); #if HAVE_SIMD R *X(taint)(R *p, INT s); R *X(join_taint)(R *p1, R *p2); #define TAINT(p, s) X(taint)(p, s) #define UNTAINT(p) ((R *) (((uintptr_t) (p)) & ~(uintptr_t)3)) #define TAINTOF(p) (((uintptr_t)(p)) & 3) #define JOIN_TAINT(p1, p2) X(join_taint)(p1, p2) #else #define TAINT(p, s) (p) #define UNTAINT(p) (p) #define TAINTOF(p) 0 #define JOIN_TAINT(p1, p2) p1 #endif #ifdef FFTW_DEBUG_ALIGNMENT # define ASSERT_ALIGNED_DOUBLE { \ double __foo; \ CK(!(((uintptr_t) &__foo) & 0x7)); \ } #else # define ASSERT_ALIGNED_DOUBLE #endif /* FFTW_DEBUG_ALIGNMENT */ /*-----------------------------------------------------------------------*/ /* macros used in codelets to reduce source code size */ typedef R E; /* internal precision of codelets. */ #if defined(FFTW_LDOUBLE) # define K(x) ((E) x##L) #elif defined(FFTW_QUAD) # define K(x) ((E) x##Q) #else # define K(x) ((E) x) #endif #define DK(name, value) const E name = K(value) /* FMA macros */ #if defined(__GNUC__) && (defined(__powerpc__) || defined(__ppc__) || defined(_POWER)) /* The obvious expression a * b + c does not work. If both x = a * b + c and y = a * b - c appear in the source, gcc computes t = a * b, x = t + c, y = t - c, thus destroying the fma. This peculiar coding seems to do the right thing on all of gcc-2.95, gcc-3.1, gcc-3.2, and gcc-3.3. It does the right thing on gcc-3.4 -fno-web (because the ``web'' pass splits the variable `x' for the single-assignment form). However, gcc-4.0 is a formidable adversary which succeeds in pessimizing two fma's into one multiplication and two additions. It does it very early in the game---before the optimization passes even start. The only real workaround seems to use fake inline asm such as asm ("# confuse gcc %0" : "=f"(a) : "0"(a)); return a * b + c; in each of the FMA, FMS, FNMA, and FNMS functions. However, this does not solve the problem either, because two equal asm statements count as a common subexpression! One must use *different* fake asm statements: in FMA: asm ("# confuse gcc for fma %0" : "=f"(a) : "0"(a)); in FMS: asm ("# confuse gcc for fms %0" : "=f"(a) : "0"(a)); etc. After these changes, gcc recalcitrantly generates the fma that was in the source to begin with. However, the extra asm() cruft confuses other passes of gcc, notably the instruction scheduler. (Of course, one could also generate the fma directly via inline asm, but this confuses the scheduler even more.) Steven and I have submitted more than one bug report to the gcc mailing list over the past few years, to no effect. Thus, I give up. gcc-4.0 can go to hell. I'll wait at least until gcc-4.3 is out before touching this crap again. */ static __inline__ E FMA(E a, E b, E c) { E x = a * b; x = x + c; return x; } static __inline__ E FMS(E a, E b, E c) { E x = a * b; x = x - c; return x; } static __inline__ E FNMA(E a, E b, E c) { E x = a * b; x = - (x + c); return x; } static __inline__ E FNMS(E a, E b, E c) { E x = a * b; x = - (x - c); return x; } #else #define FMA(a, b, c) (((a) * (b)) + (c)) #define FMS(a, b, c) (((a) * (b)) - (c)) #define FNMA(a, b, c) (- (((a) * (b)) + (c))) #define FNMS(a, b, c) ((c) - ((a) * (b))) #endif #ifdef __cplusplus } /* extern "C" */ #endif /* __cplusplus */ #endif /* __IFFTW_H__ */ fftw-3.3.3/simd-support/0000755000175000017500000000000012054411106012127 500000000000000fftw-3.3.3/simd-support/Makefile.am0000644000175000017500000000077112054410050014105 00000000000000AM_CPPFLAGS = -I$(top_srcdir)/kernel noinst_LTLIBRARIES = libsimd_support.la libsimd_sse2_nonportable.la libsimd_support_la_SOURCES = taint.c simd-common.h simd-sse2.h sse2.c \ x86-cpuid.h amd64-cpuid.h avx.c simd-avx.h altivec.c simd-altivec.h \ neon.c simd-neon.h # sse2-nonportable.c needs SSE2_CFLAGS, but Automake does not support # per-object CFLAGS. Thus we build a separate library. libsimd_sse2_nonportable_la_CFLAGS = $(SSE2_CFLAGS) libsimd_sse2_nonportable_la_SOURCES = sse2-nonportable.c fftw-3.3.3/simd-support/Makefile.in0000644000175000017500000004511612054410200014115 00000000000000# Makefile.in generated by automake 1.11.6 from Makefile.am. # @configure_input@ # Copyright (C) 1994, 1995, 1996, 1997, 1998, 1999, 2000, 2001, 2002, # 2003, 2004, 2005, 2006, 2007, 2008, 2009, 2010, 2011 Free Software # Foundation, Inc. # This Makefile.in is free software; the Free Software Foundation # gives unlimited permission to copy and/or distribute it, # with or without modifications, as long as this notice is preserved. # This program is distributed in the hope that it will be useful, # but WITHOUT ANY WARRANTY, to the extent permitted by law; without # even the implied warranty of MERCHANTABILITY or FITNESS FOR A # PARTICULAR PURPOSE. @SET_MAKE@ VPATH = @srcdir@ am__make_dryrun = \ { \ am__dry=no; \ case $$MAKEFLAGS in \ *\\[\ \ ]*) \ echo 'am--echo: ; @echo "AM" OK' | $(MAKE) -f - 2>/dev/null \ | grep '^AM OK$$' >/dev/null || am__dry=yes;; \ *) \ for am__flg in $$MAKEFLAGS; do \ case $$am__flg in \ *=*|--*) ;; \ *n*) am__dry=yes; break;; \ esac; \ done;; \ esac; \ test $$am__dry = yes; \ } pkgdatadir = $(datadir)/@PACKAGE@ pkgincludedir = $(includedir)/@PACKAGE@ pkglibdir = $(libdir)/@PACKAGE@ pkglibexecdir = $(libexecdir)/@PACKAGE@ am__cd = CDPATH="$${ZSH_VERSION+.}$(PATH_SEPARATOR)" && cd install_sh_DATA = $(install_sh) -c -m 644 install_sh_PROGRAM = $(install_sh) -c install_sh_SCRIPT = $(install_sh) -c INSTALL_HEADER = $(INSTALL_DATA) transform = $(program_transform_name) NORMAL_INSTALL = : PRE_INSTALL = : POST_INSTALL = : NORMAL_UNINSTALL = : PRE_UNINSTALL = : POST_UNINSTALL = : build_triplet = @build@ host_triplet = @host@ subdir = simd-support DIST_COMMON = $(srcdir)/Makefile.am $(srcdir)/Makefile.in ACLOCAL_M4 = $(top_srcdir)/aclocal.m4 am__aclocal_m4_deps = $(top_srcdir)/m4/acx_mpi.m4 \ $(top_srcdir)/m4/acx_pthread.m4 \ $(top_srcdir)/m4/ax_cc_maxopt.m4 \ $(top_srcdir)/m4/ax_check_compiler_flags.m4 \ $(top_srcdir)/m4/ax_compiler_vendor.m4 \ $(top_srcdir)/m4/ax_gcc_aligns_stack.m4 \ $(top_srcdir)/m4/ax_gcc_version.m4 \ $(top_srcdir)/m4/ax_openmp.m4 $(top_srcdir)/m4/libtool.m4 \ $(top_srcdir)/m4/ltoptions.m4 $(top_srcdir)/m4/ltsugar.m4 \ $(top_srcdir)/m4/ltversion.m4 $(top_srcdir)/m4/lt~obsolete.m4 \ $(top_srcdir)/configure.ac am__configure_deps = $(am__aclocal_m4_deps) $(CONFIGURE_DEPENDENCIES) \ $(ACLOCAL_M4) mkinstalldirs = $(install_sh) -d CONFIG_HEADER = $(top_builddir)/config.h CONFIG_CLEAN_FILES = CONFIG_CLEAN_VPATH_FILES = LTLIBRARIES = $(noinst_LTLIBRARIES) libsimd_sse2_nonportable_la_LIBADD = am_libsimd_sse2_nonportable_la_OBJECTS = \ libsimd_sse2_nonportable_la-sse2-nonportable.lo libsimd_sse2_nonportable_la_OBJECTS = \ $(am_libsimd_sse2_nonportable_la_OBJECTS) libsimd_sse2_nonportable_la_LINK = $(LIBTOOL) --tag=CC \ $(AM_LIBTOOLFLAGS) $(LIBTOOLFLAGS) --mode=link $(CCLD) \ $(libsimd_sse2_nonportable_la_CFLAGS) $(CFLAGS) $(AM_LDFLAGS) \ $(LDFLAGS) -o $@ libsimd_support_la_LIBADD = am_libsimd_support_la_OBJECTS = taint.lo sse2.lo avx.lo altivec.lo \ neon.lo libsimd_support_la_OBJECTS = $(am_libsimd_support_la_OBJECTS) DEFAULT_INCLUDES = -I.@am__isrc@ -I$(top_builddir) depcomp = $(SHELL) $(top_srcdir)/depcomp am__depfiles_maybe = depfiles am__mv = mv -f COMPILE = $(CC) $(DEFS) $(DEFAULT_INCLUDES) $(INCLUDES) $(AM_CPPFLAGS) \ $(CPPFLAGS) $(AM_CFLAGS) $(CFLAGS) LTCOMPILE = $(LIBTOOL) --tag=CC $(AM_LIBTOOLFLAGS) $(LIBTOOLFLAGS) \ --mode=compile $(CC) $(DEFS) $(DEFAULT_INCLUDES) $(INCLUDES) \ $(AM_CPPFLAGS) $(CPPFLAGS) $(AM_CFLAGS) $(CFLAGS) CCLD = $(CC) LINK = $(LIBTOOL) --tag=CC $(AM_LIBTOOLFLAGS) $(LIBTOOLFLAGS) \ --mode=link $(CCLD) $(AM_CFLAGS) $(CFLAGS) $(AM_LDFLAGS) \ $(LDFLAGS) -o $@ SOURCES = $(libsimd_sse2_nonportable_la_SOURCES) \ $(libsimd_support_la_SOURCES) DIST_SOURCES = $(libsimd_sse2_nonportable_la_SOURCES) \ $(libsimd_support_la_SOURCES) am__can_run_installinfo = \ case $$AM_UPDATE_INFO_DIR in \ n|no|NO) false;; \ *) (install-info --version) >/dev/null 2>&1;; \ esac ETAGS = etags CTAGS = ctags DISTFILES = $(DIST_COMMON) $(DIST_SOURCES) $(TEXINFOS) $(EXTRA_DIST) ACLOCAL = @ACLOCAL@ ALLOCA = @ALLOCA@ ALTIVEC_CFLAGS = @ALTIVEC_CFLAGS@ AMTAR = @AMTAR@ AR = @AR@ AS = @AS@ AUTOCONF = @AUTOCONF@ AUTOHEADER = @AUTOHEADER@ AUTOMAKE = @AUTOMAKE@ AVX_CFLAGS = @AVX_CFLAGS@ AWK = @AWK@ CC = @CC@ CCDEPMODE = @CCDEPMODE@ CFLAGS = @CFLAGS@ CHECK_PL_OPTS = @CHECK_PL_OPTS@ CPP = @CPP@ CPPFLAGS = @CPPFLAGS@ CYGPATH_W = @CYGPATH_W@ C_FFTW_R2R_KIND = @C_FFTW_R2R_KIND@ C_MPI_FINT = @C_MPI_FINT@ DEFS = @DEFS@ DEPDIR = @DEPDIR@ DLLTOOL = @DLLTOOL@ DSYMUTIL = @DSYMUTIL@ DUMPBIN = @DUMPBIN@ ECHO_C = @ECHO_C@ ECHO_N = @ECHO_N@ ECHO_T = @ECHO_T@ EGREP = @EGREP@ EXEEXT = @EXEEXT@ F77 = @F77@ FFLAGS = @FFLAGS@ FGREP = @FGREP@ FLIBS = @FLIBS@ GREP = @GREP@ INSTALL = @INSTALL@ INSTALL_DATA = @INSTALL_DATA@ INSTALL_PROGRAM = @INSTALL_PROGRAM@ INSTALL_SCRIPT = @INSTALL_SCRIPT@ INSTALL_STRIP_PROGRAM = @INSTALL_STRIP_PROGRAM@ LD = @LD@ LDFLAGS = @LDFLAGS@ LIBOBJS = @LIBOBJS@ LIBQUADMATH = @LIBQUADMATH@ LIBS = @LIBS@ LIBTOOL = @LIBTOOL@ LIPO = @LIPO@ LN_S = @LN_S@ LTLIBOBJS = @LTLIBOBJS@ MAINT = @MAINT@ MAKEINFO = @MAKEINFO@ MANIFEST_TOOL = @MANIFEST_TOOL@ MKDIR_P = @MKDIR_P@ MPICC = @MPICC@ MPILIBS = @MPILIBS@ MPIRUN = @MPIRUN@ NEON_CFLAGS = @NEON_CFLAGS@ NM = @NM@ NMEDIT = @NMEDIT@ OBJDUMP = @OBJDUMP@ OBJEXT = @OBJEXT@ OCAMLBUILD = @OCAMLBUILD@ OPENMP_CFLAGS = @OPENMP_CFLAGS@ OTOOL = @OTOOL@ OTOOL64 = @OTOOL64@ PACKAGE = @PACKAGE@ PACKAGE_BUGREPORT = @PACKAGE_BUGREPORT@ PACKAGE_NAME = @PACKAGE_NAME@ PACKAGE_STRING = @PACKAGE_STRING@ PACKAGE_TARNAME = @PACKAGE_TARNAME@ PACKAGE_URL = @PACKAGE_URL@ PACKAGE_VERSION = @PACKAGE_VERSION@ PATH_SEPARATOR = @PATH_SEPARATOR@ POW_LIB = @POW_LIB@ PRECISION = @PRECISION@ PREC_SUFFIX = @PREC_SUFFIX@ PTHREAD_CC = @PTHREAD_CC@ PTHREAD_CFLAGS = @PTHREAD_CFLAGS@ PTHREAD_LIBS = @PTHREAD_LIBS@ RANLIB = @RANLIB@ SED = @SED@ SET_MAKE = @SET_MAKE@ SHARED_VERSION_INFO = @SHARED_VERSION_INFO@ SHELL = @SHELL@ SSE2_CFLAGS = @SSE2_CFLAGS@ STACK_ALIGN_CFLAGS = @STACK_ALIGN_CFLAGS@ STRIP = @STRIP@ THREADLIBS = @THREADLIBS@ VERSION = @VERSION@ abs_builddir = @abs_builddir@ abs_srcdir = @abs_srcdir@ abs_top_builddir = @abs_top_builddir@ abs_top_srcdir = @abs_top_srcdir@ ac_ct_AR = @ac_ct_AR@ ac_ct_CC = @ac_ct_CC@ ac_ct_DUMPBIN = @ac_ct_DUMPBIN@ ac_ct_F77 = @ac_ct_F77@ acx_pthread_config = @acx_pthread_config@ am__include = @am__include@ am__leading_dot = @am__leading_dot@ am__quote = @am__quote@ am__tar = @am__tar@ am__untar = @am__untar@ bindir = @bindir@ build = @build@ build_alias = @build_alias@ build_cpu = @build_cpu@ build_os = @build_os@ build_vendor = @build_vendor@ builddir = @builddir@ datadir = @datadir@ datarootdir = @datarootdir@ docdir = @docdir@ dvidir = @dvidir@ exec_prefix = @exec_prefix@ host = @host@ host_alias = @host_alias@ host_cpu = @host_cpu@ host_os = @host_os@ host_vendor = @host_vendor@ htmldir = @htmldir@ includedir = @includedir@ infodir = @infodir@ install_sh = @install_sh@ libdir = @libdir@ libexecdir = @libexecdir@ localedir = @localedir@ localstatedir = @localstatedir@ mandir = @mandir@ mkdir_p = @mkdir_p@ oldincludedir = @oldincludedir@ pdfdir = @pdfdir@ prefix = @prefix@ program_transform_name = @program_transform_name@ psdir = @psdir@ sbindir = @sbindir@ sharedstatedir = @sharedstatedir@ srcdir = @srcdir@ sysconfdir = @sysconfdir@ target_alias = @target_alias@ top_build_prefix = @top_build_prefix@ top_builddir = @top_builddir@ top_srcdir = @top_srcdir@ AM_CPPFLAGS = -I$(top_srcdir)/kernel noinst_LTLIBRARIES = libsimd_support.la libsimd_sse2_nonportable.la libsimd_support_la_SOURCES = taint.c simd-common.h simd-sse2.h sse2.c \ x86-cpuid.h amd64-cpuid.h avx.c simd-avx.h altivec.c simd-altivec.h \ neon.c simd-neon.h # sse2-nonportable.c needs SSE2_CFLAGS, but Automake does not support # per-object CFLAGS. Thus we build a separate library. libsimd_sse2_nonportable_la_CFLAGS = $(SSE2_CFLAGS) libsimd_sse2_nonportable_la_SOURCES = sse2-nonportable.c all: all-am .SUFFIXES: .SUFFIXES: .c .lo .o .obj $(srcdir)/Makefile.in: @MAINTAINER_MODE_TRUE@ $(srcdir)/Makefile.am $(am__configure_deps) @for dep in $?; do \ case '$(am__configure_deps)' in \ *$$dep*) \ ( cd $(top_builddir) && $(MAKE) $(AM_MAKEFLAGS) am--refresh ) \ && { if test -f $@; then exit 0; else break; fi; }; \ exit 1;; \ esac; \ done; \ echo ' cd $(top_srcdir) && $(AUTOMAKE) --gnu simd-support/Makefile'; \ $(am__cd) $(top_srcdir) && \ $(AUTOMAKE) --gnu simd-support/Makefile .PRECIOUS: Makefile Makefile: $(srcdir)/Makefile.in $(top_builddir)/config.status @case '$?' in \ *config.status*) \ cd $(top_builddir) && $(MAKE) $(AM_MAKEFLAGS) am--refresh;; \ *) \ echo ' cd $(top_builddir) && $(SHELL) ./config.status $(subdir)/$@ $(am__depfiles_maybe)'; \ cd $(top_builddir) && $(SHELL) ./config.status $(subdir)/$@ $(am__depfiles_maybe);; \ esac; $(top_builddir)/config.status: $(top_srcdir)/configure $(CONFIG_STATUS_DEPENDENCIES) cd $(top_builddir) && $(MAKE) $(AM_MAKEFLAGS) am--refresh $(top_srcdir)/configure: @MAINTAINER_MODE_TRUE@ $(am__configure_deps) cd $(top_builddir) && $(MAKE) $(AM_MAKEFLAGS) am--refresh $(ACLOCAL_M4): @MAINTAINER_MODE_TRUE@ $(am__aclocal_m4_deps) cd $(top_builddir) && $(MAKE) $(AM_MAKEFLAGS) am--refresh $(am__aclocal_m4_deps): clean-noinstLTLIBRARIES: -test -z "$(noinst_LTLIBRARIES)" || rm -f $(noinst_LTLIBRARIES) @list='$(noinst_LTLIBRARIES)'; for p in $$list; do \ dir="`echo $$p | sed -e 's|/[^/]*$$||'`"; \ test "$$dir" != "$$p" || dir=.; \ echo "rm -f \"$${dir}/so_locations\""; \ rm -f "$${dir}/so_locations"; \ done libsimd_sse2_nonportable.la: $(libsimd_sse2_nonportable_la_OBJECTS) $(libsimd_sse2_nonportable_la_DEPENDENCIES) $(EXTRA_libsimd_sse2_nonportable_la_DEPENDENCIES) $(libsimd_sse2_nonportable_la_LINK) $(libsimd_sse2_nonportable_la_OBJECTS) $(libsimd_sse2_nonportable_la_LIBADD) $(LIBS) libsimd_support.la: $(libsimd_support_la_OBJECTS) $(libsimd_support_la_DEPENDENCIES) $(EXTRA_libsimd_support_la_DEPENDENCIES) $(LINK) $(libsimd_support_la_OBJECTS) $(libsimd_support_la_LIBADD) $(LIBS) mostlyclean-compile: -rm -f *.$(OBJEXT) distclean-compile: -rm -f *.tab.c @AMDEP_TRUE@@am__include@ @am__quote@./$(DEPDIR)/altivec.Plo@am__quote@ @AMDEP_TRUE@@am__include@ @am__quote@./$(DEPDIR)/avx.Plo@am__quote@ @AMDEP_TRUE@@am__include@ @am__quote@./$(DEPDIR)/libsimd_sse2_nonportable_la-sse2-nonportable.Plo@am__quote@ @AMDEP_TRUE@@am__include@ @am__quote@./$(DEPDIR)/neon.Plo@am__quote@ @AMDEP_TRUE@@am__include@ @am__quote@./$(DEPDIR)/sse2.Plo@am__quote@ @AMDEP_TRUE@@am__include@ @am__quote@./$(DEPDIR)/taint.Plo@am__quote@ .c.o: @am__fastdepCC_TRUE@ $(COMPILE) -MT $@ -MD -MP -MF $(DEPDIR)/$*.Tpo -c -o $@ $< @am__fastdepCC_TRUE@ $(am__mv) $(DEPDIR)/$*.Tpo $(DEPDIR)/$*.Po @AMDEP_TRUE@@am__fastdepCC_FALSE@ source='$<' object='$@' libtool=no @AMDEPBACKSLASH@ @AMDEP_TRUE@@am__fastdepCC_FALSE@ DEPDIR=$(DEPDIR) $(CCDEPMODE) $(depcomp) @AMDEPBACKSLASH@ @am__fastdepCC_FALSE@ $(COMPILE) -c $< .c.obj: @am__fastdepCC_TRUE@ $(COMPILE) -MT $@ -MD -MP -MF $(DEPDIR)/$*.Tpo -c -o $@ `$(CYGPATH_W) '$<'` @am__fastdepCC_TRUE@ $(am__mv) $(DEPDIR)/$*.Tpo $(DEPDIR)/$*.Po @AMDEP_TRUE@@am__fastdepCC_FALSE@ source='$<' object='$@' libtool=no @AMDEPBACKSLASH@ @AMDEP_TRUE@@am__fastdepCC_FALSE@ DEPDIR=$(DEPDIR) $(CCDEPMODE) $(depcomp) @AMDEPBACKSLASH@ @am__fastdepCC_FALSE@ $(COMPILE) -c `$(CYGPATH_W) '$<'` .c.lo: @am__fastdepCC_TRUE@ $(LTCOMPILE) -MT $@ -MD -MP -MF $(DEPDIR)/$*.Tpo -c -o $@ $< @am__fastdepCC_TRUE@ $(am__mv) $(DEPDIR)/$*.Tpo $(DEPDIR)/$*.Plo @AMDEP_TRUE@@am__fastdepCC_FALSE@ source='$<' object='$@' libtool=yes @AMDEPBACKSLASH@ @AMDEP_TRUE@@am__fastdepCC_FALSE@ DEPDIR=$(DEPDIR) $(CCDEPMODE) $(depcomp) @AMDEPBACKSLASH@ @am__fastdepCC_FALSE@ $(LTCOMPILE) -c -o $@ $< libsimd_sse2_nonportable_la-sse2-nonportable.lo: sse2-nonportable.c @am__fastdepCC_TRUE@ $(LIBTOOL) --tag=CC $(AM_LIBTOOLFLAGS) $(LIBTOOLFLAGS) --mode=compile $(CC) $(DEFS) $(DEFAULT_INCLUDES) $(INCLUDES) $(AM_CPPFLAGS) $(CPPFLAGS) $(libsimd_sse2_nonportable_la_CFLAGS) $(CFLAGS) -MT libsimd_sse2_nonportable_la-sse2-nonportable.lo -MD -MP -MF $(DEPDIR)/libsimd_sse2_nonportable_la-sse2-nonportable.Tpo -c -o libsimd_sse2_nonportable_la-sse2-nonportable.lo `test -f 'sse2-nonportable.c' || echo '$(srcdir)/'`sse2-nonportable.c @am__fastdepCC_TRUE@ $(am__mv) $(DEPDIR)/libsimd_sse2_nonportable_la-sse2-nonportable.Tpo $(DEPDIR)/libsimd_sse2_nonportable_la-sse2-nonportable.Plo @AMDEP_TRUE@@am__fastdepCC_FALSE@ source='sse2-nonportable.c' object='libsimd_sse2_nonportable_la-sse2-nonportable.lo' libtool=yes @AMDEPBACKSLASH@ @AMDEP_TRUE@@am__fastdepCC_FALSE@ DEPDIR=$(DEPDIR) $(CCDEPMODE) $(depcomp) @AMDEPBACKSLASH@ @am__fastdepCC_FALSE@ $(LIBTOOL) --tag=CC $(AM_LIBTOOLFLAGS) $(LIBTOOLFLAGS) --mode=compile $(CC) $(DEFS) $(DEFAULT_INCLUDES) $(INCLUDES) $(AM_CPPFLAGS) $(CPPFLAGS) $(libsimd_sse2_nonportable_la_CFLAGS) $(CFLAGS) -c -o libsimd_sse2_nonportable_la-sse2-nonportable.lo `test -f 'sse2-nonportable.c' || echo '$(srcdir)/'`sse2-nonportable.c mostlyclean-libtool: -rm -f *.lo clean-libtool: -rm -rf .libs _libs ID: $(HEADERS) $(SOURCES) $(LISP) $(TAGS_FILES) list='$(SOURCES) $(HEADERS) $(LISP) $(TAGS_FILES)'; \ unique=`for i in $$list; do \ if test -f "$$i"; then echo $$i; else echo $(srcdir)/$$i; fi; \ done | \ $(AWK) '{ files[$$0] = 1; nonempty = 1; } \ END { if (nonempty) { for (i in files) print i; }; }'`; \ mkid -fID $$unique tags: TAGS TAGS: $(HEADERS) $(SOURCES) $(TAGS_DEPENDENCIES) \ $(TAGS_FILES) $(LISP) set x; \ here=`pwd`; \ list='$(SOURCES) $(HEADERS) $(LISP) $(TAGS_FILES)'; \ unique=`for i in $$list; do \ if test -f "$$i"; then echo $$i; else echo $(srcdir)/$$i; fi; \ done | \ $(AWK) '{ files[$$0] = 1; nonempty = 1; } \ END { if (nonempty) { for (i in files) print i; }; }'`; \ shift; \ if test -z "$(ETAGS_ARGS)$$*$$unique"; then :; else \ test -n "$$unique" || unique=$$empty_fix; \ if test $$# -gt 0; then \ $(ETAGS) $(ETAGSFLAGS) $(AM_ETAGSFLAGS) $(ETAGS_ARGS) \ "$$@" $$unique; \ else \ $(ETAGS) $(ETAGSFLAGS) $(AM_ETAGSFLAGS) $(ETAGS_ARGS) \ $$unique; \ fi; \ fi ctags: CTAGS CTAGS: $(HEADERS) $(SOURCES) $(TAGS_DEPENDENCIES) \ $(TAGS_FILES) $(LISP) list='$(SOURCES) $(HEADERS) $(LISP) $(TAGS_FILES)'; \ unique=`for i in $$list; do \ if test -f "$$i"; then echo $$i; else echo $(srcdir)/$$i; fi; \ done | \ $(AWK) '{ files[$$0] = 1; nonempty = 1; } \ END { if (nonempty) { for (i in files) print i; }; }'`; \ test -z "$(CTAGS_ARGS)$$unique" \ || $(CTAGS) $(CTAGSFLAGS) $(AM_CTAGSFLAGS) $(CTAGS_ARGS) \ $$unique GTAGS: here=`$(am__cd) $(top_builddir) && pwd` \ && $(am__cd) $(top_srcdir) \ && gtags -i $(GTAGS_ARGS) "$$here" distclean-tags: -rm -f TAGS ID GTAGS GRTAGS GSYMS GPATH tags distdir: $(DISTFILES) @srcdirstrip=`echo "$(srcdir)" | sed 's/[].[^$$\\*]/\\\\&/g'`; \ topsrcdirstrip=`echo "$(top_srcdir)" | sed 's/[].[^$$\\*]/\\\\&/g'`; \ list='$(DISTFILES)'; \ dist_files=`for file in $$list; do echo $$file; done | \ sed -e "s|^$$srcdirstrip/||;t" \ -e "s|^$$topsrcdirstrip/|$(top_builddir)/|;t"`; \ case $$dist_files in \ */*) $(MKDIR_P) `echo "$$dist_files" | \ sed '/\//!d;s|^|$(distdir)/|;s,/[^/]*$$,,' | \ sort -u` ;; \ esac; \ for file in $$dist_files; do \ if test -f $$file || test -d $$file; then d=.; else d=$(srcdir); fi; \ if test -d $$d/$$file; then \ dir=`echo "/$$file" | sed -e 's,/[^/]*$$,,'`; \ if test -d "$(distdir)/$$file"; then \ find "$(distdir)/$$file" -type d ! -perm -700 -exec chmod u+rwx {} \;; \ fi; \ if test -d $(srcdir)/$$file && test $$d != $(srcdir); then \ cp -fpR $(srcdir)/$$file "$(distdir)$$dir" || exit 1; \ find "$(distdir)/$$file" -type d ! -perm -700 -exec chmod u+rwx {} \;; \ fi; \ cp -fpR $$d/$$file "$(distdir)$$dir" || exit 1; \ else \ test -f "$(distdir)/$$file" \ || cp -p $$d/$$file "$(distdir)/$$file" \ || exit 1; \ fi; \ done check-am: all-am check: check-am all-am: Makefile $(LTLIBRARIES) installdirs: install: install-am install-exec: install-exec-am install-data: install-data-am uninstall: uninstall-am install-am: all-am @$(MAKE) $(AM_MAKEFLAGS) install-exec-am install-data-am installcheck: installcheck-am install-strip: if test -z '$(STRIP)'; then \ $(MAKE) $(AM_MAKEFLAGS) INSTALL_PROGRAM="$(INSTALL_STRIP_PROGRAM)" \ install_sh_PROGRAM="$(INSTALL_STRIP_PROGRAM)" INSTALL_STRIP_FLAG=-s \ install; \ else \ $(MAKE) $(AM_MAKEFLAGS) INSTALL_PROGRAM="$(INSTALL_STRIP_PROGRAM)" \ install_sh_PROGRAM="$(INSTALL_STRIP_PROGRAM)" INSTALL_STRIP_FLAG=-s \ "INSTALL_PROGRAM_ENV=STRIPPROG='$(STRIP)'" install; \ fi mostlyclean-generic: clean-generic: distclean-generic: -test -z "$(CONFIG_CLEAN_FILES)" || rm -f $(CONFIG_CLEAN_FILES) -test . = "$(srcdir)" || test -z "$(CONFIG_CLEAN_VPATH_FILES)" || rm -f $(CONFIG_CLEAN_VPATH_FILES) maintainer-clean-generic: @echo "This command is intended for maintainers to use" @echo "it deletes files that may require special tools to rebuild." clean: clean-am clean-am: clean-generic clean-libtool clean-noinstLTLIBRARIES \ mostlyclean-am distclean: distclean-am -rm -rf ./$(DEPDIR) -rm -f Makefile distclean-am: clean-am distclean-compile distclean-generic \ distclean-tags dvi: dvi-am dvi-am: html: html-am html-am: info: info-am info-am: install-data-am: install-dvi: install-dvi-am install-dvi-am: install-exec-am: install-html: install-html-am install-html-am: install-info: install-info-am install-info-am: install-man: install-pdf: install-pdf-am install-pdf-am: install-ps: install-ps-am install-ps-am: installcheck-am: maintainer-clean: maintainer-clean-am -rm -rf ./$(DEPDIR) -rm -f Makefile maintainer-clean-am: distclean-am maintainer-clean-generic mostlyclean: mostlyclean-am mostlyclean-am: mostlyclean-compile mostlyclean-generic \ mostlyclean-libtool pdf: pdf-am pdf-am: ps: ps-am ps-am: uninstall-am: .MAKE: install-am install-strip .PHONY: CTAGS GTAGS all all-am check check-am clean clean-generic \ clean-libtool clean-noinstLTLIBRARIES ctags distclean \ distclean-compile distclean-generic distclean-libtool \ distclean-tags distdir dvi dvi-am html html-am info info-am \ install install-am install-data install-data-am install-dvi \ install-dvi-am install-exec install-exec-am install-html \ install-html-am install-info install-info-am install-man \ install-pdf install-pdf-am install-ps install-ps-am \ install-strip installcheck installcheck-am installdirs \ maintainer-clean maintainer-clean-generic mostlyclean \ mostlyclean-compile mostlyclean-generic mostlyclean-libtool \ pdf pdf-am ps ps-am tags uninstall uninstall-am # Tell versions [3.59,3.63) of GNU make to not export all variables. # Otherwise a system limit (for SysV at least) may be exceeded. .NOEXPORT: fftw-3.3.3/simd-support/sse2-nonportable.c0000644000175000017500000000272512054410050015413 00000000000000/* * Copyright (c) 2003, 2007-11 Matteo Frigo * Copyright (c) 2003, 2007-11 Massachusetts Institute of Technology * * This program is free software; you can redistribute it and/or modify * it under the terms of the GNU General Public License as published by * the Free Software Foundation; either version 2 of the License, or * (at your option) any later version. * * This program is distributed in the hope that it will be useful, * but WITHOUT ANY WARRANTY; without even the implied warranty of * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the * GNU General Public License for more details. * * You should have received a copy of the GNU General Public License * along with this program; if not, write to the Free Software * Foundation, Inc., 51 Franklin Street, Fifth Floor, Boston, MA 02110-1301 USA * */ #include "ifftw.h" #if HAVE_SSE2 /* this file must be compiled with -msse/-msse2 or equivalent, and it will fail at runtime on a machine that does not support sse/sse2 */ #include "simd-sse2.h" /* This will produce -0.0f (or -0.0d) even on broken compilers that do not distinguish +0.0 from -0.0. I bet some are still around. */ const union uvec X(sse2_pm) = { #ifdef FFTW_SINGLE { 0x00000000, 0x80000000, 0x00000000, 0x80000000 } #else { 0x00000000, 0x00000000, 0x00000000, 0x80000000 } #endif }; /* paranoia because of past compiler bugs */ void X(check_alignment_of_sse2_pm)(void) { CK(ALIGNED(&X(sse2_pm))); } #endif fftw-3.3.3/simd-support/taint.c0000644000175000017500000000247112054410050013333 00000000000000/* * Copyright (c) 2003, 2007-11 Matteo Frigo * Copyright (c) 2003, 2007-11 Massachusetts Institute of Technology * * This program is free software; you can redistribute it and/or modify * it under the terms of the GNU General Public License as published by * the Free Software Foundation; either version 2 of the License, or * (at your option) any later version. * * This program is distributed in the hope that it will be useful, * but WITHOUT ANY WARRANTY; without even the implied warranty of * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the * GNU General Public License for more details. * * You should have received a copy of the GNU General Public License * along with this program; if not, write to the Free Software * Foundation, Inc., 51 Franklin Street, Fifth Floor, Boston, MA 02110-1301 USA * */ #include "ifftw.h" #include "simd-common.h" #if HAVE_SIMD R *X(taint)(R *p, INT s) { if (((unsigned)s * sizeof(R)) % ALIGNMENT) p = (R *) (PTRINT(p) | TAINT_BIT); if (((unsigned)s * sizeof(R)) % ALIGNMENTA) p = (R *) (PTRINT(p) | TAINT_BITA); return p; } /* join the taint of two pointers that are supposed to be the same modulo the taint */ R *X(join_taint)(R *p1, R *p2) { A(UNTAINT(p1) == UNTAINT(p2)); return (R *)(PTRINT(p1) | PTRINT(p2)); } #endif fftw-3.3.3/simd-support/simd-common.h0000644000175000017500000000540412054410050014442 00000000000000/* * Copyright (c) 2003, 2007-11 Matteo Frigo * Copyright (c) 2003, 2007-11 Massachusetts Institute of Technology * * This program is free software; you can redistribute it and/or modify * it under the terms of the GNU General Public License as published by * the Free Software Foundation; either version 2 of the License, or * (at your option) any later version. * * This program is distributed in the hope that it will be useful, * but WITHOUT ANY WARRANTY; without even the implied warranty of * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the * GNU General Public License for more details. * * You should have received a copy of the GNU General Public License * along with this program; if not, write to the Free Software * Foundation, Inc., 51 Franklin Street, Fifth Floor, Boston, MA 02110-1301 USA * */ /* detection of alignment. This is complicated because a machine may support multiple SIMD extensions (e.g. SSE2 and AVX) but only one set of alignment contraints. So this alignment stuff cannot be defined in the SIMD header files. Rather than defining a separate set of "machine" header files, we just do this ugly ifdef here. */ #if defined(HAVE_SSE2) || defined(HAVE_AVX) # if defined(FFTW_SINGLE) # define ALIGNMENT 8 /* Alignment for the LD/ST macros */ # define ALIGNMENTA 16 /* Alignment for the LDA/STA macros */ # else # define ALIGNMENT 16 /* Alignment for the LD/ST macros */ # define ALIGNMENTA 16 /* Alignment for the LDA/STA macros */ # endif #elif defined(HAVE_ALTIVEC) # define ALIGNMENT 8 /* Alignment for the LD/ST macros */ # define ALIGNMENTA 16 /* Alignment for the LDA/STA macros */ #elif defined(HAVE_NEON) # define ALIGNMENT 8 /* Alignment for the LD/ST macros */ # define ALIGNMENTA 8 /* Alignment for the LDA/STA macros */ #endif #if HAVE_SIMD # ifndef ALIGNMENT # error "ALIGNMENT not defined" # endif # ifndef ALIGNMENTA # error "ALIGNMENTA not defined" # endif #endif /* rename for precision and for SIMD extensions */ #define XSIMD0(name, suffix) CONCAT(name, suffix) #define XSIMD(name) XSIMD0(X(name), SIMD_SUFFIX) #define XSIMD_STRING(x) x STRINGIZE(SIMD_SUFFIX) /* TAINT_BIT is set if pointers are not guaranteed to be multiples of ALIGNMENT */ #define TAINT_BIT 1 /* TAINT_BITA is set if pointers are not guaranteed to be multiples of ALIGNMENTA */ #define TAINT_BITA 2 #define PTRINT(p) ((uintptr_t)(p)) #define ALIGNED(p) \ (((PTRINT(UNTAINT(p)) % ALIGNMENT) == 0) && !(PTRINT(p) & TAINT_BIT)) #define ALIGNEDA(p) \ (((PTRINT(UNTAINT(p)) % ALIGNMENTA) == 0) && !(PTRINT(p) & TAINT_BITA)) #define SIMD_STRIDE_OK(x) (!(((x) * sizeof(R)) % ALIGNMENT)) #define SIMD_STRIDE_OKA(x) (!(((x) * sizeof(R)) % ALIGNMENTA)) #define SIMD_VSTRIDE_OK SIMD_STRIDE_OK fftw-3.3.3/simd-support/simd-sse2.h0000644000175000017500000002327712054410050014036 00000000000000/* * Copyright (c) 2003, 2007-11 Matteo Frigo * Copyright (c) 2003, 2007-11 Massachusetts Institute of Technology * * This program is free software; you can redistribute it and/or modify * it under the terms of the GNU General Public License as published by * the Free Software Foundation; either version 2 of the License, or * (at your option) any later version. * * This program is distributed in the hope that it will be useful, * but WITHOUT ANY WARRANTY; without even the implied warranty of * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the * GNU General Public License for more details. * * You should have received a copy of the GNU General Public License * along with this program; if not, write to the Free Software * Foundation, Inc., 51 Franklin Street, Fifth Floor, Boston, MA 02110-1301 USA * */ #if defined(FFTW_LDOUBLE) || defined(FFTW_QUAD) # error "SSE/SSE2 only works in single/double precision" #endif #ifdef FFTW_SINGLE # define DS(d,s) s /* single-precision option */ # define SUFF(name) name ## s #else # define DS(d,s) d /* double-precision option */ # define SUFF(name) name ## d #endif #define SIMD_SUFFIX _sse2 /* for renaming */ #define VL DS(1,2) /* SIMD vector length, in term of complex numbers */ #define SIMD_VSTRIDE_OKA(x) DS(1,((x) == 2)) #define SIMD_STRIDE_OKPAIR SIMD_STRIDE_OK #if defined(__GNUC__) && !defined(FFTW_SINGLE) && !defined(__SSE2__) # error "compiling simd-sse2.h in double precision without -msse2" #elif defined(__GNUC__) && defined(FFTW_SINGLE) && !defined(__SSE__) # error "compiling simd-sse2.h in single precision without -msse" #endif #ifdef _MSC_VER #ifndef inline #define inline __inline #endif #endif /* some versions of glibc's sys/cdefs.h define __inline to be empty, which is wrong because emmintrin.h defines several inline procedures */ #ifndef _MSC_VER #undef __inline #endif #ifdef FFTW_SINGLE # include #else # include #endif typedef DS(__m128d,__m128) V; #define VADD SUFF(_mm_add_p) #define VSUB SUFF(_mm_sub_p) #define VMUL SUFF(_mm_mul_p) #define VXOR SUFF(_mm_xor_p) #define SHUF SUFF(_mm_shuffle_p) #define UNPCKL SUFF(_mm_unpacklo_p) #define UNPCKH SUFF(_mm_unpackhi_p) #define SHUFVALS(fp0,fp1,fp2,fp3) \ (((fp3) << 6) | ((fp2) << 4) | ((fp1) << 2) | ((fp0))) #define VDUPL(x) DS(UNPCKL(x, x), SHUF(x, x, SHUFVALS(0, 0, 2, 2))) #define VDUPH(x) DS(UNPCKH(x, x), SHUF(x, x, SHUFVALS(1, 1, 3, 3))) #define STOREH(a, v) DS(_mm_storeh_pd(a, v), _mm_storeh_pi((__m64 *)(a), v)) #define STOREL(a, v) DS(_mm_storel_pd(a, v), _mm_storel_pi((__m64 *)(a), v)) #ifdef __GNUC__ /* * gcc-3.3 generates slow code for mm_set_ps (write all elements to * the stack and load __m128 from the stack). * * gcc-3.[34] generates slow code for mm_set_ps1 (load into low element * and shuffle). * * This hack forces gcc to generate a constant __m128 at compile time. */ union rvec { R r[DS(2,4)]; V v; }; # ifdef FFTW_SINGLE # define DVK(var, val) V var = __extension__ ({ \ static const union rvec _var = { {val,val,val,val} }; _var.v; }) # else # define DVK(var, val) V var = __extension__ ({ \ static const union rvec _var = { {val,val} }; _var.v; }) # endif # define LDK(x) x #else # define DVK(var, val) const R var = K(val) # define LDK(x) DS(_mm_set1_pd,_mm_set_ps1)(x) #endif union uvec { unsigned u[4]; V v; }; static inline V LDA(const R *x, INT ivs, const R *aligned_like) { (void)aligned_like; /* UNUSED */ (void)ivs; /* UNUSED */ return *(const V *)x; } static inline void STA(R *x, V v, INT ovs, const R *aligned_like) { (void)aligned_like; /* UNUSED */ (void)ovs; /* UNUSED */ *(V *)x = v; } #ifdef FFTW_SINGLE # ifdef _MSC_VER /* Temporarily disable the warning "uninitialized local variable 'name' used" and runtime checks for using a variable before it is defined which is erroneously triggered by the LOADL0 / LOADH macros as they only modify VAL partly each. */ # pragma warning(disable : 4700) # pragma runtime_checks("u", off) # endif static inline V LD(const R *x, INT ivs, const R *aligned_like) { V var; (void)aligned_like; /* UNUSED */ # ifdef __GNUC__ /* We use inline asm because gcc-3.x generates slow code for _mm_loadh_pi(). gcc-3.x insists upon having an existing variable for VAL, which is however never used. Thus, it generates code to move values in and out the variable. Worse still, gcc-4.0 stores VAL on the stack, causing valgrind to complain about uninitialized reads. */ __asm__("movlps %1, %0\n\tmovhps %2, %0" : "=x"(var) : "m"(x[0]), "m"(x[ivs])); # else # define LOADH(addr, val) _mm_loadh_pi(val, (const __m64 *)(addr)) # define LOADL0(addr, val) _mm_loadl_pi(val, (const __m64 *)(addr)) var = LOADL0(x, var); var = LOADH(x + ivs, var); # endif return var; } # ifdef _MSC_VER # pragma warning(default : 4700) # pragma runtime_checks("u", restore) # endif static inline void ST(R *x, V v, INT ovs, const R *aligned_like) { (void)aligned_like; /* UNUSED */ /* WARNING: the extra_iter hack depends upon STOREL occurring after STOREH */ STOREH(x + ovs, v); STOREL(x, v); } #else /* ! FFTW_SINGLE */ # define LD LDA # define ST STA #endif #define STM2 DS(STA,ST) #define STN2(x, v0, v1, ovs) /* nop */ #ifdef FFTW_SINGLE # define STM4(x, v, ovs, aligned_like) /* no-op */ /* STN4 is a macro, not a function, thanks to Visual C++ developers deciding "it would be infrequent that people would want to pass more than 3 [__m128 parameters] by value." 3 parameters ought to be enough for anybody. */ # define STN4(x, v0, v1, v2, v3, ovs) \ { \ V xxx0, xxx1, xxx2, xxx3; \ xxx0 = UNPCKL(v0, v2); \ xxx1 = UNPCKH(v0, v2); \ xxx2 = UNPCKL(v1, v3); \ xxx3 = UNPCKH(v1, v3); \ STA(x, UNPCKL(xxx0, xxx2), 0, 0); \ STA(x + ovs, UNPCKH(xxx0, xxx2), 0, 0); \ STA(x + 2 * ovs, UNPCKL(xxx1, xxx3), 0, 0); \ STA(x + 3 * ovs, UNPCKH(xxx1, xxx3), 0, 0); \ } #else /* !FFTW_SINGLE */ static inline void STM4(R *x, V v, INT ovs, const R *aligned_like) { (void)aligned_like; /* UNUSED */ STOREL(x, v); STOREH(x + ovs, v); } # define STN4(x, v0, v1, v2, v3, ovs) /* nothing */ #endif static inline V FLIP_RI(V x) { return SHUF(x, x, DS(1, SHUFVALS(1, 0, 3, 2))); } extern const union uvec X(sse2_pm); static inline V VCONJ(V x) { return VXOR(X(sse2_pm).v, x); } static inline V VBYI(V x) { x = VCONJ(x); x = FLIP_RI(x); return x; } /* FMA support */ #define VFMA(a, b, c) VADD(c, VMUL(a, b)) #define VFNMS(a, b, c) VSUB(c, VMUL(a, b)) #define VFMS(a, b, c) VSUB(VMUL(a, b), c) #define VFMAI(b, c) VADD(c, VBYI(b)) #define VFNMSI(b, c) VSUB(c, VBYI(b)) #define VFMACONJ(b,c) VADD(VCONJ(b),c) #define VFMSCONJ(b,c) VSUB(VCONJ(b),c) #define VFNMSCONJ(b,c) VSUB(c, VCONJ(b)) static inline V VZMUL(V tx, V sr) { V tr = VDUPL(tx); V ti = VDUPH(tx); tr = VMUL(sr, tr); sr = VBYI(sr); return VFMA(ti, sr, tr); } static inline V VZMULJ(V tx, V sr) { V tr = VDUPL(tx); V ti = VDUPH(tx); tr = VMUL(sr, tr); sr = VBYI(sr); return VFNMS(ti, sr, tr); } static inline V VZMULI(V tx, V sr) { V tr = VDUPL(tx); V ti = VDUPH(tx); ti = VMUL(ti, sr); sr = VBYI(sr); return VFMS(tr, sr, ti); } static inline V VZMULIJ(V tx, V sr) { V tr = VDUPL(tx); V ti = VDUPH(tx); ti = VMUL(ti, sr); sr = VBYI(sr); return VFMA(tr, sr, ti); } /* twiddle storage #1: compact, slower */ #ifdef FFTW_SINGLE # define VTW1(v,x) \ {TW_COS, v, x}, {TW_COS, v+1, x}, {TW_SIN, v, x}, {TW_SIN, v+1, x} static inline V BYTW1(const R *t, V sr) { const V *twp = (const V *)t; V tx = twp[0]; V tr = UNPCKL(tx, tx); V ti = UNPCKH(tx, tx); tr = VMUL(tr, sr); sr = VBYI(sr); return VFMA(ti, sr, tr); } static inline V BYTWJ1(const R *t, V sr) { const V *twp = (const V *)t; V tx = twp[0]; V tr = UNPCKL(tx, tx); V ti = UNPCKH(tx, tx); tr = VMUL(tr, sr); sr = VBYI(sr); return VFNMS(ti, sr, tr); } #else /* !FFTW_SINGLE */ # define VTW1(v,x) {TW_CEXP, v, x} static inline V BYTW1(const R *t, V sr) { V tx = LD(t, 1, t); return VZMUL(tx, sr); } static inline V BYTWJ1(const R *t, V sr) { V tx = LD(t, 1, t); return VZMULJ(tx, sr); } #endif #define TWVL1 (VL) /* twiddle storage #2: twice the space, faster (when in cache) */ #ifdef FFTW_SINGLE # define VTW2(v,x) \ {TW_COS, v, x}, {TW_COS, v, x}, {TW_COS, v+1, x}, {TW_COS, v+1, x}, \ {TW_SIN, v, -x}, {TW_SIN, v, x}, {TW_SIN, v+1, -x}, {TW_SIN, v+1, x} #else /* !FFTW_SINGLE */ # define VTW2(v,x) \ {TW_COS, v, x}, {TW_COS, v, x}, {TW_SIN, v, -x}, {TW_SIN, v, x} #endif #define TWVL2 (2 * VL) static inline V BYTW2(const R *t, V sr) { const V *twp = (const V *)t; V si = FLIP_RI(sr); V tr = twp[0], ti = twp[1]; return VFMA(tr, sr, VMUL(ti, si)); } static inline V BYTWJ2(const R *t, V sr) { const V *twp = (const V *)t; V si = FLIP_RI(sr); V tr = twp[0], ti = twp[1]; return VFNMS(ti, si, VMUL(tr, sr)); } /* twiddle storage #3 */ #ifdef FFTW_SINGLE # define VTW3(v,x) {TW_CEXP, v, x}, {TW_CEXP, v+1, x} # define TWVL3 (VL) #else # define VTW3(v,x) VTW1(v,x) # define TWVL3 TWVL1 #endif /* twiddle storage for split arrays */ #ifdef FFTW_SINGLE # define VTWS(v,x) \ {TW_COS, v, x}, {TW_COS, v+1, x}, {TW_COS, v+2, x}, {TW_COS, v+3, x}, \ {TW_SIN, v, x}, {TW_SIN, v+1, x}, {TW_SIN, v+2, x}, {TW_SIN, v+3, x} #else # define VTWS(v,x) \ {TW_COS, v, x}, {TW_COS, v+1, x}, {TW_SIN, v, x}, {TW_SIN, v+1, x} #endif #define TWVLS (2 * VL) #define VLEAVE() /* nothing */ #include "simd-common.h" fftw-3.3.3/simd-support/sse2.c0000644000175000017500000000422612054410050013070 00000000000000/* * Copyright (c) 2003, 2007-11 Matteo Frigo * Copyright (c) 2003, 2007-11 Massachusetts Institute of Technology * * This program is free software; you can redistribute it and/or modify * it under the terms of the GNU General Public License as published by * the Free Software Foundation; either version 2 of the License, or * (at your option) any later version. * * This program is distributed in the hope that it will be useful, * but WITHOUT ANY WARRANTY; without even the implied warranty of * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the * GNU General Public License for more details. * * You should have received a copy of the GNU General Public License * along with this program; if not, write to the Free Software * Foundation, Inc., 51 Franklin Street, Fifth Floor, Boston, MA 02110-1301 USA * */ #include "ifftw.h" #ifdef FFTW_SINGLE # define DS(d,s) s /* single-precision option */ #else # define DS(d,s) d /* double-precision option */ #endif #if HAVE_SSE2 # if defined(__x86_64__) || defined(_M_X64) || defined(_M_AMD64) int X(have_simd_sse2)(void) { return 1; } # else /* !x86_64 */ # include # include # include "x86-cpuid.h" static jmp_buf jb; static void sighandler(int x) { UNUSED(x); longjmp(jb, 1); } static int sse2_works(void) { void (*oldsig)(int); oldsig = signal(SIGILL, sighandler); if (setjmp(jb)) { signal(SIGILL, oldsig); return 0; } else { # ifdef _MSC_VER _asm { DS(xorpd,xorps) xmm0,xmm0 } # else /* asm volatile ("xorpd/s %xmm0, %xmm0"); */ asm volatile(DS(".byte 0x66; .byte 0x0f; .byte 0x57; .byte 0xc0", ".byte 0x0f; .byte 0x57; .byte 0xc0")); # endif signal(SIGILL, oldsig); return 1; } } extern void X(check_alignment_of_sse2_pm)(void); int X(have_simd_sse2)(void) { static int init = 0, res; if (!init) { res = !is_386() && has_cpuid() && (cpuid_edx(1) & (1 << DS(26,25))) && sse2_works(); init = 1; X(check_alignment_of_sse2_pm)(); } return res; } # endif #endif fftw-3.3.3/simd-support/x86-cpuid.h0000644000175000017500000000702012054410050013743 00000000000000/* * Copyright (c) 2003, 2007-11 Matteo Frigo * Copyright (c) 2003, 2007-11 Massachusetts Institute of Technology * * This program is free software; you can redistribute it and/or modify * it under the terms of the GNU General Public License as published by * the Free Software Foundation; either version 2 of the License, or * (at your option) any later version. * * This program is distributed in the hope that it will be useful, * but WITHOUT ANY WARRANTY; without even the implied warranty of * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the * GNU General Public License for more details. * * You should have received a copy of the GNU General Public License * along with this program; if not, write to the Free Software * Foundation, Inc., 51 Franklin Street, Fifth Floor, Boston, MA 02110-1301 USA * */ /* this code was kindly donated by Eric J. Korpela */ #ifdef _MSC_VER #ifndef inline #define inline __inline #endif #endif static inline int is_386() { #ifdef _MSC_VER unsigned int result,tst; _asm { pushfd pop eax mov edx,eax xor eax,40000h push eax popfd pushfd pop eax push edx popfd mov tst,edx mov result,eax } #else register unsigned int result,tst; __asm__ ( "pushfl\n\t" "popl %0\n\t" "movl %0,%1\n\t" "xorl $0x40000,%0\n\t" "pushl %0\n\t" "popfl\n\t" "pushfl\n\t" "popl %0\n\t" "pushl %1\n\t" "popfl" : "=r" (result), "=r" (tst) /* output */ : /* no inputs */ ); #endif return (result == tst); } static inline int has_cpuid() { #ifdef _MSC_VER unsigned int result,tst; _asm { pushfd pop eax mov edx,eax xor eax,200000h push eax popfd pushfd pop eax push edx popfd mov tst,edx mov result,eax } #else register unsigned int result,tst; __asm__ ( "pushfl\n\t" "pop %0\n\t" "movl %0,%1\n\t" "xorl $0x200000,%0\n\t" "pushl %0\n\t" "popfl\n\t" "pushfl\n\t" "popl %0\n\t" "pushl %1\n\t" "popfl" : "=r" (result), "=r" (tst) /* output */ : /* no inputs */ ); #endif return (result != tst); } static inline int cpuid_edx(int op) { # ifdef _MSC_VER int result; _asm { push ebx mov eax,op cpuid mov result,edx pop ebx } return result; # else int eax, ecx, edx; __asm__("push %%ebx\n\tcpuid\n\tpop %%ebx" : "=a" (eax), "=c" (ecx), "=d" (edx) : "a" (op)); return edx; # endif } static inline int cpuid_ecx(int op) { # ifdef _MSC_VER int result; _asm { push ebx mov eax,op cpuid mov result,ecx pop ebx } return result; # else int eax, ecx, edx; __asm__("push %%ebx\n\tcpuid\n\tpop %%ebx" : "=a" (eax), "=c" (ecx), "=d" (edx) : "a" (op)); return ecx; # endif } static inline int xgetbv_eax(int op) { # ifdef _MSC_VER int veax, vedx; _asm { mov ecx,op # if defined(__INTEL_COMPILER) || (_MSC_VER >= 1600) xgetbv # else __emit 15 __emit 1 __emit 208 # endif mov veax,eax mov vedx,edx } return veax; # else int eax, edx; __asm__ (".byte 0x0f, 0x01, 0xd0" : "=a"(eax), "=d"(edx) : "c" (op)); return eax; #endif } fftw-3.3.3/simd-support/amd64-cpuid.h0000644000175000017500000000414112054410050014232 00000000000000/* * Copyright (c) 2003, 2007-11 Matteo Frigo * Copyright (c) 2003, 2007-11 Massachusetts Institute of Technology * * This program is free software; you can redistribute it and/or modify * it under the terms of the GNU General Public License as published by * the Free Software Foundation; either version 2 of the License, or * (at your option) any later version. * * This program is distributed in the hope that it will be useful, * but WITHOUT ANY WARRANTY; without even the implied warranty of * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the * GNU General Public License for more details. * * You should have received a copy of the GNU General Public License * along with this program; if not, write to the Free Software * Foundation, Inc., 51 Franklin Street, Fifth Floor, Boston, MA 02110-1301 USA * */ #ifdef _MSC_VER #ifndef inline #define inline __inline #endif #endif #ifdef _MSC_VER #include #if (_MSC_VER >= 1600) && !defined(__INTEL_COMPILER) #include #endif #endif static inline int cpuid_ecx(int op) { # ifdef _MSC_VER # ifdef __INTEL_COMPILER int result; _asm { push rbx mov eax,op cpuid mov result,ecx pop rbx } return result; # else int cpu_info[4]; __cpuid(cpu_info,op); return cpu_info[2]; # endif # else int eax, ecx, edx; __asm__("pushq %%rbx\n\tcpuid\n\tpopq %%rbx" : "=a" (eax), "=c" (ecx), "=d" (edx) : "a" (op)); return ecx; # endif } static inline int xgetbv_eax(int op) { # ifdef _MSC_VER # ifdef __INTEL_COMPILER int veax, vedx; _asm { mov ecx,op xgetbv mov veax,eax mov vedx,edx } return veax; # else # if defined(_MSC_VER) && (_MSC_VER >= 1600) unsigned __int64 result; result = _xgetbv(op); return (int)result; # else # error "Need at least Visual Studio 10 SP1 for AVX support" # endif # endif # else int eax, edx; __asm__ (".byte 0x0f, 0x01, 0xd0" : "=a"(eax), "=d"(edx) : "c" (op)); return eax; #endif } fftw-3.3.3/simd-support/avx.c0000644000175000017500000000276412054410050013017 00000000000000/* * Copyright (c) 2003, 2007-11 Matteo Frigo * Copyright (c) 2003, 2007-11 Massachusetts Institute of Technology * * This program is free software; you can redistribute it and/or modify * it under the terms of the GNU General Public License as published by * the Free Software Foundation; either version 2 of the License, or * (at your option) any later version. * * This program is distributed in the hope that it will be useful, * but WITHOUT ANY WARRANTY; without even the implied warranty of * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the * GNU General Public License for more details. * * You should have received a copy of the GNU General Public License * along with this program; if not, write to the Free Software * Foundation, Inc., 51 Franklin Street, Fifth Floor, Boston, MA 02110-1301 USA * */ #include "ifftw.h" #if HAVE_AVX #if defined(__x86_64__) || defined(_M_X64) || defined(_M_AMD64) #include "amd64-cpuid.h" int X(have_simd_avx)(void) { static int init = 0, res; if (!init) { res = 1 && ((cpuid_ecx(1) & 0x18000000) == 0x18000000) && ((xgetbv_eax(0) & 0x6) == 0x6); init = 1; } return res; } #else /* 32-bit code */ #include "x86-cpuid.h" int X(have_simd_avx)(void) { static int init = 0, res; if (!init) { res = !is_386() && has_cpuid() && ((cpuid_ecx(1) & 0x18000000) == 0x18000000) && ((xgetbv_eax(0) & 0x6) == 0x6); init = 1; } return res; } #endif #endif fftw-3.3.3/simd-support/simd-avx.h0000644000175000017500000002544612054410050013760 00000000000000/* * Copyright (c) 2003, 2007-11 Matteo Frigo * Copyright (c) 2003, 2007-11 Massachusetts Institute of Technology * * This program is free software; you can redistribute it and/or modify * it under the terms of the GNU General Public License as published by * the Free Software Foundation; either version 2 of the License, or * (at your option) any later version. * * This program is distributed in the hope that it will be useful, * but WITHOUT ANY WARRANTY; without even the implied warranty of * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the * GNU General Public License for more details. * * You should have received a copy of the GNU General Public License * along with this program; if not, write to the Free Software * Foundation, Inc., 51 Franklin Street, Fifth Floor, Boston, MA 02110-1301 USA * */ #if defined(FFTW_LDOUBLE) || defined(FFTW_QUAD) #error "AVX only works in single or double precision" #endif #ifdef FFTW_SINGLE # define DS(d,s) s /* single-precision option */ # define SUFF(name) name ## s #else # define DS(d,s) d /* double-precision option */ # define SUFF(name) name ## d #endif #define SIMD_SUFFIX _avx /* for renaming */ #define VL DS(2, 4) /* SIMD complex vector length */ #define SIMD_VSTRIDE_OKA(x) ((x) == 2) #define SIMD_STRIDE_OKPAIR SIMD_STRIDE_OK #if defined(__GNUC__) && !defined(__AVX__) /* sanity check */ #error "compiling simd-avx.h without -mavx" #endif #ifdef _MSC_VER #ifndef inline #define inline __inline #endif #endif #include typedef DS(__m256d, __m256) V; #define VADD SUFF(_mm256_add_p) #define VSUB SUFF(_mm256_sub_p) #define VMUL SUFF(_mm256_mul_p) #define VXOR SUFF(_mm256_xor_p) #define VSHUF SUFF(_mm256_shuffle_p) #define SHUFVALD(fp0,fp1) \ (((fp1) << 3) | ((fp0) << 2) | ((fp1) << 1) | ((fp0))) #define SHUFVALS(fp0,fp1,fp2,fp3) \ (((fp3) << 6) | ((fp2) << 4) | ((fp1) << 2) | ((fp0))) #define VDUPL(x) DS(_mm256_unpacklo_pd(x, x), VSHUF(x, x, SHUFVALS(0, 0, 2, 2))) #define VDUPH(x) DS(_mm256_unpackhi_pd(x, x), VSHUF(x, x, SHUFVALS(1, 1, 3, 3))) #define VLIT(x0, x1) DS(_mm256_set_pd(x0, x1, x0, x1), _mm256_set_ps(x0, x1, x0, x1, x0, x1, x0, x1)) #define DVK(var, val) V var = VLIT(val, val) #define LDK(x) x static inline V LDA(const R *x, INT ivs, const R *aligned_like) { (void)aligned_like; /* UNUSED */ (void)ivs; /* UNUSED */ return SUFF(_mm256_loadu_p)(x); } static inline void STA(R *x, V v, INT ovs, const R *aligned_like) { (void)aligned_like; /* UNUSED */ (void)ovs; /* UNUSED */ SUFF(_mm256_storeu_p)(x, v); } #if FFTW_SINGLE #define LOADH(addr, val) _mm_loadh_pi(val, (const __m64 *)(addr)) #define LOADL(addr, val) _mm_loadl_pi(val, (const __m64 *)(addr)) #define STOREH(addr, val) _mm_storeh_pi((__m64 *)(addr), val) #define STOREL(addr, val) _mm_storel_pi((__m64 *)(addr), val) /* it seems like the only AVX way to store 4 complex floats is to extract two pairs of complex floats into two __m128 registers, and then use SSE-like half-stores. Similarly, to load 4 complex floats, we load two pairs of complex floats into two __m128 registers, and then pack the two __m128 registers into one __m256 value. */ static inline V LD(const R *x, INT ivs, const R *aligned_like) { __m128 l, h; V v; (void)aligned_like; /* UNUSED */ l = LOADL(x, l); l = LOADH(x + ivs, l); h = LOADL(x + 2*ivs, h); h = LOADH(x + 3*ivs, h); v = _mm256_castps128_ps256(l); v = _mm256_insertf128_ps(v, h, 1); return v; } static inline void ST(R *x, V v, INT ovs, const R *aligned_like) { __m128 h = _mm256_extractf128_ps(v, 1); __m128 l = _mm256_castps256_ps128(v); (void)aligned_like; /* UNUSED */ /* WARNING: the extra_iter hack depends upon STOREL occurring after STOREH */ STOREH(x + 3*ovs, h); STOREL(x + 2*ovs, h); STOREH(x + ovs, l); STOREL(x, l); } #define STM2(x, v, ovs, aligned_like) /* no-op */ static inline void STN2(R *x, V v0, V v1, INT ovs) { V x0 = VSHUF(v0, v1, SHUFVALS(0, 1, 0, 1)); V x1 = VSHUF(v0, v1, SHUFVALS(2, 3, 2, 3)); __m128 h0 = _mm256_extractf128_ps(x0, 1); __m128 l0 = _mm256_castps256_ps128(x0); __m128 h1 = _mm256_extractf128_ps(x1, 1); __m128 l1 = _mm256_castps256_ps128(x1); *(__m128 *)(x + 3*ovs) = h1; *(__m128 *)(x + 2*ovs) = h0; *(__m128 *)(x + 1*ovs) = l1; *(__m128 *)(x + 0*ovs) = l0; } #define STM4(x, v, ovs, aligned_like) /* no-op */ #define STN4(x, v0, v1, v2, v3, ovs) \ { \ V xxx0, xxx1, xxx2, xxx3; \ V yyy0, yyy1, yyy2, yyy3; \ xxx0 = _mm256_unpacklo_ps(v0, v2); \ xxx1 = _mm256_unpackhi_ps(v0, v2); \ xxx2 = _mm256_unpacklo_ps(v1, v3); \ xxx3 = _mm256_unpackhi_ps(v1, v3); \ yyy0 = _mm256_unpacklo_ps(xxx0, xxx2); \ yyy1 = _mm256_unpackhi_ps(xxx0, xxx2); \ yyy2 = _mm256_unpacklo_ps(xxx1, xxx3); \ yyy3 = _mm256_unpackhi_ps(xxx1, xxx3); \ *(__m128 *)(x + 0 * ovs) = _mm256_castps256_ps128(yyy0); \ *(__m128 *)(x + 4 * ovs) = _mm256_extractf128_ps(yyy0, 1); \ *(__m128 *)(x + 1 * ovs) = _mm256_castps256_ps128(yyy1); \ *(__m128 *)(x + 5 * ovs) = _mm256_extractf128_ps(yyy1, 1); \ *(__m128 *)(x + 2 * ovs) = _mm256_castps256_ps128(yyy2); \ *(__m128 *)(x + 6 * ovs) = _mm256_extractf128_ps(yyy2, 1); \ *(__m128 *)(x + 3 * ovs) = _mm256_castps256_ps128(yyy3); \ *(__m128 *)(x + 7 * ovs) = _mm256_extractf128_ps(yyy3, 1); \ } #else static inline __m128d VMOVAPD_LD(const R *x) { /* gcc-4.6 miscompiles the combination _mm256_castpd128_pd256(VMOVAPD_LD(x)) into a 256-bit vmovapd, which requires 32-byte aligment instead of 16-byte alignment. Force the use of vmovapd via asm until compilers stabilize. */ #if defined(__GNUC__) __m128d var; __asm__("vmovapd %1, %0\n" : "=x"(var) : "m"(x[0])); return var; #else return *(const __m128d *)x; #endif } static inline V LD(const R *x, INT ivs, const R *aligned_like) { V var; (void)aligned_like; /* UNUSED */ var = _mm256_castpd128_pd256(VMOVAPD_LD(x)); var = _mm256_insertf128_pd(var, *(const __m128d *)(x+ivs), 1); return var; } static inline void ST(R *x, V v, INT ovs, const R *aligned_like) { (void)aligned_like; /* UNUSED */ /* WARNING: the extra_iter hack depends upon the store of the low part occurring after the store of the high part */ *(__m128d *)(x + ovs) = _mm256_extractf128_pd(v, 1); *(__m128d *)x = _mm256_castpd256_pd128(v); } #define STM2 ST #define STN2(x, v0, v1, ovs) /* nop */ #define STM4(x, v, ovs, aligned_like) /* no-op */ /* STN4 is a macro, not a function, thanks to Visual C++ developers deciding "it would be infrequent that people would want to pass more than 3 [__m128 parameters] by value." Even though the comment was made about __m128 parameters, it appears to apply to __m256 parameters as well. */ #define STN4(x, v0, v1, v2, v3, ovs) \ { \ V xxx0, xxx1, xxx2, xxx3; \ xxx0 = _mm256_unpacklo_pd(v0, v1); \ xxx1 = _mm256_unpackhi_pd(v0, v1); \ xxx2 = _mm256_unpacklo_pd(v2, v3); \ xxx3 = _mm256_unpackhi_pd(v2, v3); \ STA(x, _mm256_permute2f128_pd(xxx0, xxx2, 0x20), 0, 0); \ STA(x + ovs, _mm256_permute2f128_pd(xxx1, xxx3, 0x20), 0, 0); \ STA(x + 2 * ovs, _mm256_permute2f128_pd(xxx0, xxx2, 0x31), 0, 0); \ STA(x + 3 * ovs, _mm256_permute2f128_pd(xxx1, xxx3, 0x31), 0, 0); \ } #endif static inline V FLIP_RI(V x) { return VSHUF(x, x, DS(SHUFVALD(1, 0), SHUFVALS(1, 0, 3, 2))); } static inline V VCONJ(V x) { V pmpm = VLIT(-0.0, 0.0); return VXOR(pmpm, x); } static inline V VBYI(V x) { return FLIP_RI(VCONJ(x)); } /* FMA support */ #define VFMA(a, b, c) VADD(c, VMUL(a, b)) #define VFNMS(a, b, c) VSUB(c, VMUL(a, b)) #define VFMS(a, b, c) VSUB(VMUL(a, b), c) #define VFMAI(b, c) VADD(c, VBYI(b)) #define VFNMSI(b, c) VSUB(c, VBYI(b)) #define VFMACONJ(b,c) VADD(VCONJ(b),c) #define VFMSCONJ(b,c) VSUB(VCONJ(b),c) #define VFNMSCONJ(b,c) VSUB(c, VCONJ(b)) static inline V VZMUL(V tx, V sr) { V tr = VDUPL(tx); V ti = VDUPH(tx); tr = VMUL(sr, tr); sr = VBYI(sr); return VFMA(ti, sr, tr); } static inline V VZMULJ(V tx, V sr) { V tr = VDUPL(tx); V ti = VDUPH(tx); tr = VMUL(sr, tr); sr = VBYI(sr); return VFNMS(ti, sr, tr); } static inline V VZMULI(V tx, V sr) { V tr = VDUPL(tx); V ti = VDUPH(tx); ti = VMUL(ti, sr); sr = VBYI(sr); return VFMS(tr, sr, ti); } static inline V VZMULIJ(V tx, V sr) { V tr = VDUPL(tx); V ti = VDUPH(tx); ti = VMUL(ti, sr); sr = VBYI(sr); return VFMA(tr, sr, ti); } /* twiddle storage #1: compact, slower */ #ifdef FFTW_SINGLE # define VTW1(v,x) {TW_CEXP, v, x}, {TW_CEXP, v+1, x}, {TW_CEXP, v+2, x}, {TW_CEXP, v+3, x} #else # define VTW1(v,x) {TW_CEXP, v, x}, {TW_CEXP, v+1, x} #endif #define TWVL1 (VL) static inline V BYTW1(const R *t, V sr) { return VZMUL(LDA(t, 2, t), sr); } static inline V BYTWJ1(const R *t, V sr) { return VZMULJ(LDA(t, 2, t), sr); } /* twiddle storage #2: twice the space, faster (when in cache) */ #ifdef FFTW_SINGLE # define VTW2(v,x) \ {TW_COS, v, x}, {TW_COS, v, x}, {TW_COS, v+1, x}, {TW_COS, v+1, x}, \ {TW_COS, v+2, x}, {TW_COS, v+2, x}, {TW_COS, v+3, x}, {TW_COS, v+3, x}, \ {TW_SIN, v, -x}, {TW_SIN, v, x}, {TW_SIN, v+1, -x}, {TW_SIN, v+1, x}, \ {TW_SIN, v+2, -x}, {TW_SIN, v+2, x}, {TW_SIN, v+3, -x}, {TW_SIN, v+3, x} #else # define VTW2(v,x) \ {TW_COS, v, x}, {TW_COS, v, x}, {TW_COS, v+1, x}, {TW_COS, v+1, x}, \ {TW_SIN, v, -x}, {TW_SIN, v, x}, {TW_SIN, v+1, -x}, {TW_SIN, v+1, x} #endif #define TWVL2 (2 * VL) static inline V BYTW2(const R *t, V sr) { const V *twp = (const V *)t; V si = FLIP_RI(sr); V tr = twp[0], ti = twp[1]; return VFMA(tr, sr, VMUL(ti, si)); } static inline V BYTWJ2(const R *t, V sr) { const V *twp = (const V *)t; V si = FLIP_RI(sr); V tr = twp[0], ti = twp[1]; return VFNMS(ti, si, VMUL(tr, sr)); } /* twiddle storage #3 */ #define VTW3 VTW1 #define TWVL3 TWVL1 /* twiddle storage for split arrays */ #ifdef FFTW_SINGLE # define VTWS(v,x) \ {TW_COS, v, x}, {TW_COS, v+1, x}, {TW_COS, v+2, x}, {TW_COS, v+3, x}, \ {TW_COS, v+4, x}, {TW_COS, v+5, x}, {TW_COS, v+6, x}, {TW_COS, v+7, x}, \ {TW_SIN, v, x}, {TW_SIN, v+1, x}, {TW_SIN, v+2, x}, {TW_SIN, v+3, x}, \ {TW_SIN, v+4, x}, {TW_SIN, v+5, x}, {TW_SIN, v+6, x}, {TW_SIN, v+7, x} #else # define VTWS(v,x) \ {TW_COS, v, x}, {TW_COS, v+1, x}, {TW_COS, v+2, x}, {TW_COS, v+3, x}, \ {TW_SIN, v, x}, {TW_SIN, v+1, x}, {TW_SIN, v+2, x}, {TW_SIN, v+3, x} #endif #define TWVLS (2 * VL) /* Use VZEROUPPER to avoid the penalty of switching from AVX to SSE. See Intel Optimization Manual (April 2011, version 248966), Section 11.3 */ #define VLEAVE _mm256_zeroupper #include "simd-common.h" fftw-3.3.3/simd-support/altivec.c0000644000175000017500000000364112054410050013643 00000000000000/* * Copyright (c) 2003, 2007-11 Matteo Frigo * Copyright (c) 2003, 2007-11 Massachusetts Institute of Technology * * This program is free software; you can redistribute it and/or modify * it under the terms of the GNU General Public License as published by * the Free Software Foundation; either version 2 of the License, or * (at your option) any later version. * * This program is distributed in the hope that it will be useful, * but WITHOUT ANY WARRANTY; without even the implied warranty of * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the * GNU General Public License for more details. * * You should have received a copy of the GNU General Public License * along with this program; if not, write to the Free Software * Foundation, Inc., 51 Franklin Street, Fifth Floor, Boston, MA 02110-1301 USA * */ #include "ifftw.h" #if HAVE_ALTIVEC #if HAVE_SYS_SYSCTL_H # include #endif #if HAVE_SYS_SYSCTL_H && HAVE_SYSCTL && defined(CTL_HW) && defined(HW_VECTORUNIT) /* code for darwin */ static int really_have_altivec(void) { int mib[2], altivecp; size_t len; mib[0] = CTL_HW; mib[1] = HW_VECTORUNIT; len = sizeof(altivecp); sysctl(mib, 2, &altivecp, &len, NULL, 0); return altivecp; } #else /* GNU/Linux and other non-Darwin systems (!HAVE_SYS_SYSCTL_H etc.) */ #include #include static jmp_buf jb; static void sighandler(int x) { longjmp(jb, 1); } static int really_have_altivec(void) { void (*oldsig)(int); oldsig = signal(SIGILL, sighandler); if (setjmp(jb)) { signal(SIGILL, oldsig); return 0; } else { __asm__ __volatile__ (".long 0x10000484"); /* vor 0,0,0 */ signal(SIGILL, oldsig); return 1; } return 0; } #endif int X(have_simd_altivec)(void) { static int init = 0, res; if (!init) { res = really_have_altivec(); init = 1; } return res; } #endif fftw-3.3.3/simd-support/simd-altivec.h0000644000175000017500000002060112054410050014575 00000000000000/* * Copyright (c) 2003, 2007-11 Matteo Frigo * Copyright (c) 2003, 2007-11 Massachusetts Institute of Technology * * This program is free software; you can redistribute it and/or modify * it under the terms of the GNU General Public License as published by * the Free Software Foundation; either version 2 of the License, or * (at your option) any later version. * * This program is distributed in the hope that it will be useful, * but WITHOUT ANY WARRANTY; without even the implied warranty of * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the * GNU General Public License for more details. * * You should have received a copy of the GNU General Public License * along with this program; if not, write to the Free Software * Foundation, Inc., 51 Franklin Street, Fifth Floor, Boston, MA 02110-1301 USA * */ #ifndef FFTW_SINGLE #error "ALTIVEC only works in single precision" #endif /* define these unconditionally, because they are used by taint.c which is compiled without altivec */ #define SIMD_SUFFIX _altivec /* for renaming */ #define VL 2 /* SIMD complex vector length */ #define SIMD_VSTRIDE_OKA(x) ((x) == 2) #define SIMD_STRIDE_OKPAIR SIMD_STRIDE_OKA #if !defined(__VEC__) && !defined(FAKE__VEC__) # error "compiling simd-altivec.h requires -maltivec or equivalent" #endif #ifdef HAVE_ALTIVEC_H # include #endif typedef vector float V; #define VLIT(x0, x1, x2, x3) {x0, x1, x2, x3} #define LDK(x) x #define DVK(var, val) const V var = VLIT(val, val, val, val) static inline V VADD(V a, V b) { return vec_add(a, b); } static inline V VSUB(V a, V b) { return vec_sub(a, b); } static inline V VFMA(V a, V b, V c) { return vec_madd(a, b, c); } static inline V VFNMS(V a, V b, V c) { return vec_nmsub(a, b, c); } static inline V VMUL(V a, V b) { DVK(zero, -0.0); return VFMA(a, b, zero); } static inline V VFMS(V a, V b, V c) { return VSUB(VMUL(a, b), c); } static inline V LDA(const R *x, INT ivs, const R *aligned_like) { UNUSED(ivs); UNUSED(aligned_like); return vec_ld(0, x); } static inline V LD(const R *x, INT ivs, const R *aligned_like) { /* common subexpressions */ const INT fivs = sizeof(R) * ivs; /* you are not expected to understand this: */ const vector unsigned int perm = VLIT(0, 0, 0xFFFFFFFF, 0xFFFFFFFF); vector unsigned char ml = vec_lvsr(fivs + 8, aligned_like); vector unsigned char mh = vec_lvsl(0, aligned_like); vector unsigned char msk = (vector unsigned char)vec_sel((V)mh, (V)ml, perm); /* end of common subexpressions */ return vec_perm(vec_ld(0, x), vec_ld(fivs, x), msk); } /* store lower half */ static inline void STH(R *x, V v, R *aligned_like) { v = vec_perm(v, v, vec_lvsr(0, aligned_like)); vec_ste(v, 0, x); vec_ste(v, sizeof(R), x); } static inline void STL(R *x, V v, INT ovs, R *aligned_like) { const INT fovs = sizeof(R) * ovs; v = vec_perm(v, v, vec_lvsr(fovs + 8, aligned_like)); vec_ste(v, fovs, x); vec_ste(v, sizeof(R) + fovs, x); } static inline void STA(R *x, V v, INT ovs, R *aligned_like) { UNUSED(ovs); UNUSED(aligned_like); vec_st(v, 0, x); } static inline void ST(R *x, V v, INT ovs, R *aligned_like) { /* WARNING: the extra_iter hack depends upon STH occurring after STL */ STL(x, v, ovs, aligned_like); STH(x, v, aligned_like); } #define STM2(x, v, ovs, aligned_like) /* no-op */ static inline void STN2(R *x, V v0, V v1, INT ovs) { const INT fovs = sizeof(R) * ovs; const vector unsigned int even = VLIT(0x00010203, 0x04050607, 0x10111213, 0x14151617); const vector unsigned int odd = VLIT(0x08090a0b, 0x0c0d0e0f, 0x18191a1b, 0x1c1d1e1f); vec_st(vec_perm(v0, v1, (vector unsigned char)even), 0, x); vec_st(vec_perm(v0, v1, (vector unsigned char)odd), fovs, x); } #define STM4(x, v, ovs, aligned_like) /* no-op */ static inline void STN4(R *x, V v0, V v1, V v2, V v3, INT ovs) { const INT fovs = sizeof(R) * ovs; V x0 = vec_mergeh(v0, v2); V x1 = vec_mergel(v0, v2); V x2 = vec_mergeh(v1, v3); V x3 = vec_mergel(v1, v3); V y0 = vec_mergeh(x0, x2); V y1 = vec_mergel(x0, x2); V y2 = vec_mergeh(x1, x3); V y3 = vec_mergel(x1, x3); vec_st(y0, 0, x); vec_st(y1, fovs, x); vec_st(y2, 2 * fovs, x); vec_st(y3, 3 * fovs, x); } static inline V FLIP_RI(V x) { const vector unsigned int perm = VLIT(0x04050607, 0x00010203, 0x0c0d0e0f, 0x08090a0b); return vec_perm(x, x, (vector unsigned char)perm); } static inline V VCONJ(V x) { const V pmpm = VLIT(0.0, -0.0, 0.0, -0.0); return vec_xor(x, pmpm); } static inline V VBYI(V x) { return FLIP_RI(VCONJ(x)); } static inline V VFMAI(V b, V c) { const V mpmp = VLIT(-1.0, 1.0, -1.0, 1.0); return VFMA(FLIP_RI(b), mpmp, c); } static inline V VFNMSI(V b, V c) { const V mpmp = VLIT(-1.0, 1.0, -1.0, 1.0); return VFNMS(FLIP_RI(b), mpmp, c); } static inline V VFMACONJ(V b, V c) { const V pmpm = VLIT(1.0, -1.0, 1.0, -1.0); return VFMA(b, pmpm, c); } static inline V VFNMSCONJ(V b, V c) { const V pmpm = VLIT(1.0, -1.0, 1.0, -1.0); return VFNMS(b, pmpm, c); } static inline V VFMSCONJ(V b, V c) { return VSUB(VCONJ(b), c); } static inline V VZMUL(V tx, V sr) { const vector unsigned int real = VLIT(0x00010203, 0x00010203, 0x08090a0b, 0x08090a0b); const vector unsigned int imag = VLIT(0x04050607, 0x04050607, 0x0c0d0e0f, 0x0c0d0e0f); V si = VBYI(sr); V tr = vec_perm(tx, tx, (vector unsigned char)real); V ti = vec_perm(tx, tx, (vector unsigned char)imag); return VFMA(ti, si, VMUL(tr, sr)); } static inline V VZMULJ(V tx, V sr) { const vector unsigned int real = VLIT(0x00010203, 0x00010203, 0x08090a0b, 0x08090a0b); const vector unsigned int imag = VLIT(0x04050607, 0x04050607, 0x0c0d0e0f, 0x0c0d0e0f); V si = VBYI(sr); V tr = vec_perm(tx, tx, (vector unsigned char)real); V ti = vec_perm(tx, tx, (vector unsigned char)imag); return VFNMS(ti, si, VMUL(tr, sr)); } static inline V VZMULI(V tx, V si) { const vector unsigned int real = VLIT(0x00010203, 0x00010203, 0x08090a0b, 0x08090a0b); const vector unsigned int imag = VLIT(0x04050607, 0x04050607, 0x0c0d0e0f, 0x0c0d0e0f); V sr = VBYI(si); V tr = vec_perm(tx, tx, (vector unsigned char)real); V ti = vec_perm(tx, tx, (vector unsigned char)imag); return VFNMS(ti, si, VMUL(tr, sr)); } static inline V VZMULIJ(V tx, V si) { const vector unsigned int real = VLIT(0x00010203, 0x00010203, 0x08090a0b, 0x08090a0b); const vector unsigned int imag = VLIT(0x04050607, 0x04050607, 0x0c0d0e0f, 0x0c0d0e0f); V sr = VBYI(si); V tr = vec_perm(tx, tx, (vector unsigned char)real); V ti = vec_perm(tx, tx, (vector unsigned char)imag); return VFMA(ti, si, VMUL(tr, sr)); } /* twiddle storage #1: compact, slower */ #define VTW1(v,x) \ {TW_COS, v, x}, {TW_COS, v+1, x}, {TW_SIN, v, x}, {TW_SIN, v+1, x} #define TWVL1 (VL) static inline V BYTW1(const R *t, V sr) { const V *twp = (const V *)t; V si = VBYI(sr); V tx = twp[0]; V tr = vec_mergeh(tx, tx); V ti = vec_mergel(tx, tx); return VFMA(ti, si, VMUL(tr, sr)); } static inline V BYTWJ1(const R *t, V sr) { const V *twp = (const V *)t; V si = VBYI(sr); V tx = twp[0]; V tr = vec_mergeh(tx, tx); V ti = vec_mergel(tx, tx); return VFNMS(ti, si, VMUL(tr, sr)); } /* twiddle storage #2: twice the space, faster (when in cache) */ #define VTW2(v,x) \ {TW_COS, v, x}, {TW_COS, v, x}, {TW_COS, v+1, x}, {TW_COS, v+1, x}, \ {TW_SIN, v, -x}, {TW_SIN, v, x}, {TW_SIN, v+1, -x}, {TW_SIN, v+1, x} #define TWVL2 (2 * VL) static inline V BYTW2(const R *t, V sr) { const V *twp = (const V *)t; V si = FLIP_RI(sr); V tr = twp[0], ti = twp[1]; return VFMA(ti, si, VMUL(tr, sr)); } static inline V BYTWJ2(const R *t, V sr) { const V *twp = (const V *)t; V si = FLIP_RI(sr); V tr = twp[0], ti = twp[1]; return VFNMS(ti, si, VMUL(tr, sr)); } /* twiddle storage #3 */ #define VTW3(v,x) {TW_CEXP, v, x}, {TW_CEXP, v+1, x} #define TWVL3 (VL) /* twiddle storage for split arrays */ #define VTWS(v,x) \ {TW_COS, v, x}, {TW_COS, v+1, x}, {TW_COS, v+2, x}, {TW_COS, v+3, x}, \ {TW_SIN, v, x}, {TW_SIN, v+1, x}, {TW_SIN, v+2, x}, {TW_SIN, v+3, x} #define TWVLS (2 * VL) #define VLEAVE() /* nothing */ #include "simd-common.h" fftw-3.3.3/simd-support/neon.c0000644000175000017500000000363712054410050013160 00000000000000/* * Copyright (c) 2003, 2007-11 Matteo Frigo * Copyright (c) 2003, 2007-11 Massachusetts Institute of Technology * * This program is free software; you can redistribute it and/or modify * it under the terms of the GNU General Public License as published by * the Free Software Foundation; either version 2 of the License, or * (at your option) any later version. * * This program is distributed in the hope that it will be useful, * but WITHOUT ANY WARRANTY; without even the implied warranty of * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the * GNU General Public License for more details. * * You should have received a copy of the GNU General Public License * along with this program; if not, write to the Free Software * Foundation, Inc., 51 Franklin Street, Fifth Floor, Boston, MA 02110-1301 USA * */ #include "ifftw.h" #if HAVE_NEON /* check for an environment where signals are known to work */ #if defined(unix) || defined(linux) # include # include static jmp_buf jb; static void sighandler(int x) { UNUSED(x); longjmp(jb, 1); } static int really_have_neon(void) { void (*oldsig)(int); oldsig = signal(SIGILL, sighandler); if (setjmp(jb)) { signal(SIGILL, oldsig); return 0; } else { /* paranoia: encode the instruction in binary because the assembler may not recognize it without -mfpu=neon */ /*asm volatile ("vand q0, q0, q0");*/ asm volatile (".long 0xf2000150"); signal(SIGILL, oldsig); return 1; } } extern void X(check_alignment_of_sse2_pm)(void); int X(have_simd_neon)(void) { static int init = 0, res; if (!init) { res = really_have_neon(); init = 1; } return res; } #else /* don't know how to autodetect NEON; assume it is present */ int X(have_simd_neon)(void) { return 1; } #endif #endif fftw-3.3.3/simd-support/simd-neon.h0000644000175000017500000001662112054410050014114 00000000000000/* * Copyright (c) 2003, 2007-11 Matteo Frigo * Copyright (c) 2003, 2007-11 Massachusetts Institute of Technology * * This program is free software; you can redistribute it and/or modify * it under the terms of the GNU General Public License as published by * the Free Software Foundation; either version 2 of the License, or * (at your option) any later version. * * This program is distributed in the hope that it will be useful, * but WITHOUT ANY WARRANTY; without even the implied warranty of * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the * GNU General Public License for more details. * * You should have received a copy of the GNU General Public License * along with this program; if not, write to the Free Software * Foundation, Inc., 51 Franklin Street, Fifth Floor, Boston, MA 02110-1301 USA * */ #ifndef FFTW_SINGLE #error "NEON only works in single precision" #endif /* define these unconditionally, because they are used by taint.c which is compiled without neon */ #define SIMD_SUFFIX _neon /* for renaming */ #define VL 2 /* SIMD complex vector length */ #define SIMD_VSTRIDE_OKA(x) ((x) == 2) #define SIMD_STRIDE_OKPAIR SIMD_STRIDE_OK #if defined(__GNUC__) && !defined(__ARM_NEON__) #error "compiling simd-neon.h requires -mfpu=neon or equivalent" #endif #include /* FIXME: I am not sure whether this code assumes little-endian ordering. VLIT may or may not be wrong for big-endian systems. */ typedef float32x4_t V; #define VLIT(x0, x1, x2, x3) {x0, x1, x2, x3} #define LDK(x) x #define DVK(var, val) const V var = VLIT(val, val, val, val) /* NEON has FMA, but a three-operand FMA is not too useful for FFT purposes. We normally compute t0=a+b*c t1=a-b*c In a three-operand instruction set this translates into t0=a t0+=b*c t1=a t1-=b*c At least one move must be implemented, negating the advantage of the FMA in the first place. At least some versions of gcc generate both moves. So we are better off generating t=b*c;t0=a+t;t1=a-t;*/ #if HAVE_FMA #warning "--enable-fma on NEON is probably a bad idea (see source code)" #endif #define VADD(a, b) vaddq_f32(a, b) #define VSUB(a, b) vsubq_f32(a, b) #define VMUL(a, b) vmulq_f32(a, b) #define VFMA(a, b, c) vmlaq_f32(c, a, b) /* a*b+c */ #define VFNMS(a, b, c) vmlsq_f32(c, a, b) /* FNMS=-(a*b-c) in powerpc terminology; MLS=c-a*b in ARM terminology */ #define VFMS(a, b, c) VSUB(VMUL(a, b), c) /* FMS=a*b-c in powerpc terminology; no equivalent arm instruction (?) */ static inline V LDA(const R *x, INT ivs, const R *aligned_like) { (void) aligned_like; /* UNUSED */ return vld1q_f32((const float32_t *)x); } static inline V LD(const R *x, INT ivs, const R *aligned_like) { (void) aligned_like; /* UNUSED */ return vcombine_f32(vld1_f32((float32_t *)x), vld1_f32((float32_t *)(x + ivs))); } static inline void STA(R *x, V v, INT ovs, const R *aligned_like) { (void) aligned_like; /* UNUSED */ vst1q_f32((float32_t *)x, v); } static inline void ST(R *x, V v, INT ovs, const R *aligned_like) { (void) aligned_like; /* UNUSED */ /* WARNING: the extra_iter hack depends upon store-low occurring after store-high */ vst1_f32((float32_t *)(x + ovs), vget_high_f32(v)); vst1_f32((float32_t *)x, vget_low_f32(v)); } /* 2x2 complex transpose and store */ #define STM2 ST #define STN2(x, v0, v1, ovs) /* nop */ /* store and 4x4 real transpose */ static inline void STM4(R *x, V v, INT ovs, const R *aligned_like) { (void) aligned_like; /* UNUSED */ vst1_lane_f32((float32_t *)(x) , vget_low_f32(v), 0); vst1_lane_f32((float32_t *)(x + ovs), vget_low_f32(v), 1); vst1_lane_f32((float32_t *)(x + 2 * ovs), vget_high_f32(v), 0); vst1_lane_f32((float32_t *)(x + 3 * ovs), vget_high_f32(v), 1); } #define STN4(x, v0, v1, v2, v3, ovs) /* use STM4 */ #define FLIP_RI(x) vrev64q_f32(x) static inline V VCONJ(V x) { #if 1 static const uint32x4_t pm = {0, 0x80000000u, 0, 0x80000000u}; return vreinterpretq_f32_u32(veorq_u32(vreinterpretq_u32_f32(x), pm)); #else const V pm = VLIT(1.0, -1.0, 1.0, -1.0); return VMUL(x, pm); #endif } static inline V VBYI(V x) { return FLIP_RI(VCONJ(x)); } static inline V VFMAI(V b, V c) { const V mp = VLIT(-1.0, 1.0, -1.0, 1.0); return VFMA(FLIP_RI(b), mp, c); } static inline V VFNMSI(V b, V c) { const V mp = VLIT(-1.0, 1.0, -1.0, 1.0); return VFNMS(FLIP_RI(b), mp, c); } static inline V VFMACONJ(V b, V c) { const V pm = VLIT(1.0, -1.0, 1.0, -1.0); return VFMA(b, pm, c); } static inline V VFNMSCONJ(V b, V c) { const V pm = VLIT(1.0, -1.0, 1.0, -1.0); return VFNMS(b, pm, c); } static inline V VFMSCONJ(V b, V c) { return VSUB(VCONJ(b), c); } #if 1 #define VEXTRACT_REIM(tr, ti, tx) \ { \ tr = vcombine_f32(vdup_lane_f32(vget_low_f32(tx), 0), \ vdup_lane_f32(vget_high_f32(tx), 0)); \ ti = vcombine_f32(vdup_lane_f32(vget_low_f32(tx), 1), \ vdup_lane_f32(vget_high_f32(tx), 1)); \ } #else /* this alternative might be faster in an ideal world, but gcc likes to spill VVV onto the stack */ #define VEXTRACT_REIM(tr, ti, tx) \ { \ float32x4x2_t vvv = vtrnq_f32(tx, tx); \ tr = vvv.val[0]; \ ti = vvv.val[1]; \ } #endif static inline V VZMUL(V tx, V sr) { V tr, ti; VEXTRACT_REIM(tr, ti, tx); tr = VMUL(sr, tr); sr = VBYI(sr); return VFMA(ti, sr, tr); } static inline V VZMULJ(V tx, V sr) { V tr, ti; VEXTRACT_REIM(tr, ti, tx); tr = VMUL(sr, tr); sr = VBYI(sr); return VFNMS(ti, sr, tr); } static inline V VZMULI(V tx, V sr) { V tr, ti; VEXTRACT_REIM(tr, ti, tx); ti = VMUL(ti, sr); sr = VBYI(sr); return VFMS(tr, sr, ti); } static inline V VZMULIJ(V tx, V sr) { V tr, ti; VEXTRACT_REIM(tr, ti, tx); ti = VMUL(ti, sr); sr = VBYI(sr); return VFMA(tr, sr, ti); } /* twiddle storage #1: compact, slower */ #define VTW1(v,x) {TW_CEXP, v, x}, {TW_CEXP, v+1, x} #define TWVL1 VL static inline V BYTW1(const R *t, V sr) { V tx = LDA(t, 2, 0); return VZMUL(tx, sr); } static inline V BYTWJ1(const R *t, V sr) { V tx = LDA(t, 2, 0); return VZMULJ(tx, sr); } /* twiddle storage #2: twice the space, faster (when in cache) */ # define VTW2(v,x) \ {TW_COS, v, x}, {TW_COS, v, x}, {TW_COS, v+1, x}, {TW_COS, v+1, x}, \ {TW_SIN, v, -x}, {TW_SIN, v, x}, {TW_SIN, v+1, -x}, {TW_SIN, v+1, x} #define TWVL2 (2 * VL) static inline V BYTW2(const R *t, V sr) { V si = FLIP_RI(sr); V tr = LDA(t, 2, 0), ti = LDA(t+2*VL, 2, 0); return VFMA(ti, si, VMUL(tr, sr)); } static inline V BYTWJ2(const R *t, V sr) { V si = FLIP_RI(sr); V tr = LDA(t, 2, 0), ti = LDA(t+2*VL, 2, 0); return VFNMS(ti, si, VMUL(tr, sr)); } /* twiddle storage #3 */ # define VTW3(v,x) {TW_CEXP, v, x}, {TW_CEXP, v+1, x} # define TWVL3 (VL) /* twiddle storage for split arrays */ # define VTWS(v,x) \ {TW_COS, v, x}, {TW_COS, v+1, x}, {TW_COS, v+2, x}, {TW_COS, v+3, x}, \ {TW_SIN, v, x}, {TW_SIN, v+1, x}, {TW_SIN, v+2, x}, {TW_SIN, v+3, x} #define TWVLS (2 * VL) #define VLEAVE() /* nothing */ #include "simd-common.h" fftw-3.3.3/dft/0000755000175000017500000000000012054411106010236 500000000000000fftw-3.3.3/dft/Makefile.am0000644000175000017500000000076612054410050012220 00000000000000AM_CPPFLAGS = -I$(top_srcdir)/kernel SUBDIRS = scalar simd noinst_LTLIBRARIES = libdft.la # pkgincludedir = $(includedir)/fftw3@PREC_SUFFIX@ # pkginclude_HEADERS = codelet-dft.h dft.h libdft_la_SOURCES = bluestein.c buffered.c conf.c ct.c dftw-direct.c \ dftw-directsq.c dftw-generic.c dftw-genericbuf.c direct.c generic.c \ indirect.c indirect-transpose.c kdft-dif.c kdft-difsq.c kdft-dit.c \ kdft.c nop.c plan.c problem.c rader.c rank-geq2.c solve.c vrank-geq1.c \ zero.c codelet-dft.h ct.h dft.h fftw-3.3.3/dft/Makefile.in0000644000175000017500000005634512054410177012247 00000000000000# Makefile.in generated by automake 1.11.6 from Makefile.am. # @configure_input@ # Copyright (C) 1994, 1995, 1996, 1997, 1998, 1999, 2000, 2001, 2002, # 2003, 2004, 2005, 2006, 2007, 2008, 2009, 2010, 2011 Free Software # Foundation, Inc. # This Makefile.in is free software; the Free Software Foundation # gives unlimited permission to copy and/or distribute it, # with or without modifications, as long as this notice is preserved. # This program is distributed in the hope that it will be useful, # but WITHOUT ANY WARRANTY, to the extent permitted by law; without # even the implied warranty of MERCHANTABILITY or FITNESS FOR A # PARTICULAR PURPOSE. @SET_MAKE@ VPATH = @srcdir@ am__make_dryrun = \ { \ am__dry=no; \ case $$MAKEFLAGS in \ *\\[\ \ ]*) \ echo 'am--echo: ; @echo "AM" OK' | $(MAKE) -f - 2>/dev/null \ | grep '^AM OK$$' >/dev/null || am__dry=yes;; \ *) \ for am__flg in $$MAKEFLAGS; do \ case $$am__flg in \ *=*|--*) ;; \ *n*) am__dry=yes; break;; \ esac; \ done;; \ esac; \ test $$am__dry = yes; \ } pkgdatadir = $(datadir)/@PACKAGE@ pkgincludedir = $(includedir)/@PACKAGE@ pkglibdir = $(libdir)/@PACKAGE@ pkglibexecdir = $(libexecdir)/@PACKAGE@ am__cd = CDPATH="$${ZSH_VERSION+.}$(PATH_SEPARATOR)" && cd install_sh_DATA = $(install_sh) -c -m 644 install_sh_PROGRAM = $(install_sh) -c install_sh_SCRIPT = $(install_sh) -c INSTALL_HEADER = $(INSTALL_DATA) transform = $(program_transform_name) NORMAL_INSTALL = : PRE_INSTALL = : POST_INSTALL = : NORMAL_UNINSTALL = : PRE_UNINSTALL = : POST_UNINSTALL = : build_triplet = @build@ host_triplet = @host@ subdir = dft DIST_COMMON = $(srcdir)/Makefile.am $(srcdir)/Makefile.in ACLOCAL_M4 = $(top_srcdir)/aclocal.m4 am__aclocal_m4_deps = $(top_srcdir)/m4/acx_mpi.m4 \ $(top_srcdir)/m4/acx_pthread.m4 \ $(top_srcdir)/m4/ax_cc_maxopt.m4 \ $(top_srcdir)/m4/ax_check_compiler_flags.m4 \ $(top_srcdir)/m4/ax_compiler_vendor.m4 \ $(top_srcdir)/m4/ax_gcc_aligns_stack.m4 \ $(top_srcdir)/m4/ax_gcc_version.m4 \ $(top_srcdir)/m4/ax_openmp.m4 $(top_srcdir)/m4/libtool.m4 \ $(top_srcdir)/m4/ltoptions.m4 $(top_srcdir)/m4/ltsugar.m4 \ $(top_srcdir)/m4/ltversion.m4 $(top_srcdir)/m4/lt~obsolete.m4 \ $(top_srcdir)/configure.ac am__configure_deps = $(am__aclocal_m4_deps) $(CONFIGURE_DEPENDENCIES) \ $(ACLOCAL_M4) mkinstalldirs = $(install_sh) -d CONFIG_HEADER = $(top_builddir)/config.h CONFIG_CLEAN_FILES = CONFIG_CLEAN_VPATH_FILES = LTLIBRARIES = $(noinst_LTLIBRARIES) libdft_la_LIBADD = am_libdft_la_OBJECTS = bluestein.lo buffered.lo conf.lo ct.lo \ dftw-direct.lo dftw-directsq.lo dftw-generic.lo \ dftw-genericbuf.lo direct.lo generic.lo indirect.lo \ indirect-transpose.lo kdft-dif.lo kdft-difsq.lo kdft-dit.lo \ kdft.lo nop.lo plan.lo problem.lo rader.lo rank-geq2.lo \ solve.lo vrank-geq1.lo zero.lo libdft_la_OBJECTS = $(am_libdft_la_OBJECTS) DEFAULT_INCLUDES = -I.@am__isrc@ -I$(top_builddir) depcomp = $(SHELL) $(top_srcdir)/depcomp am__depfiles_maybe = depfiles am__mv = mv -f COMPILE = $(CC) $(DEFS) $(DEFAULT_INCLUDES) $(INCLUDES) $(AM_CPPFLAGS) \ $(CPPFLAGS) $(AM_CFLAGS) $(CFLAGS) LTCOMPILE = $(LIBTOOL) --tag=CC $(AM_LIBTOOLFLAGS) $(LIBTOOLFLAGS) \ --mode=compile $(CC) $(DEFS) $(DEFAULT_INCLUDES) $(INCLUDES) \ $(AM_CPPFLAGS) $(CPPFLAGS) $(AM_CFLAGS) $(CFLAGS) CCLD = $(CC) LINK = $(LIBTOOL) --tag=CC $(AM_LIBTOOLFLAGS) $(LIBTOOLFLAGS) \ --mode=link $(CCLD) $(AM_CFLAGS) $(CFLAGS) $(AM_LDFLAGS) \ $(LDFLAGS) -o $@ SOURCES = $(libdft_la_SOURCES) DIST_SOURCES = $(libdft_la_SOURCES) RECURSIVE_TARGETS = all-recursive check-recursive dvi-recursive \ html-recursive info-recursive install-data-recursive \ install-dvi-recursive install-exec-recursive \ install-html-recursive install-info-recursive \ install-pdf-recursive install-ps-recursive install-recursive \ installcheck-recursive installdirs-recursive pdf-recursive \ ps-recursive uninstall-recursive am__can_run_installinfo = \ case $$AM_UPDATE_INFO_DIR in \ n|no|NO) false;; \ *) (install-info --version) >/dev/null 2>&1;; \ esac RECURSIVE_CLEAN_TARGETS = mostlyclean-recursive clean-recursive \ distclean-recursive maintainer-clean-recursive AM_RECURSIVE_TARGETS = $(RECURSIVE_TARGETS:-recursive=) \ $(RECURSIVE_CLEAN_TARGETS:-recursive=) tags TAGS ctags CTAGS \ distdir ETAGS = etags CTAGS = ctags DIST_SUBDIRS = $(SUBDIRS) DISTFILES = $(DIST_COMMON) $(DIST_SOURCES) $(TEXINFOS) $(EXTRA_DIST) am__relativize = \ dir0=`pwd`; \ sed_first='s,^\([^/]*\)/.*$$,\1,'; \ sed_rest='s,^[^/]*/*,,'; \ sed_last='s,^.*/\([^/]*\)$$,\1,'; \ sed_butlast='s,/*[^/]*$$,,'; \ while test -n "$$dir1"; do \ first=`echo "$$dir1" | sed -e "$$sed_first"`; \ if test "$$first" != "."; then \ if test "$$first" = ".."; then \ dir2=`echo "$$dir0" | sed -e "$$sed_last"`/"$$dir2"; \ dir0=`echo "$$dir0" | sed -e "$$sed_butlast"`; \ else \ first2=`echo "$$dir2" | sed -e "$$sed_first"`; \ if test "$$first2" = "$$first"; then \ dir2=`echo "$$dir2" | sed -e "$$sed_rest"`; \ else \ dir2="../$$dir2"; \ fi; \ dir0="$$dir0"/"$$first"; \ fi; \ fi; \ dir1=`echo "$$dir1" | sed -e "$$sed_rest"`; \ done; \ reldir="$$dir2" ACLOCAL = @ACLOCAL@ ALLOCA = @ALLOCA@ ALTIVEC_CFLAGS = @ALTIVEC_CFLAGS@ AMTAR = @AMTAR@ AR = @AR@ AS = @AS@ AUTOCONF = @AUTOCONF@ AUTOHEADER = @AUTOHEADER@ AUTOMAKE = @AUTOMAKE@ AVX_CFLAGS = @AVX_CFLAGS@ AWK = @AWK@ CC = @CC@ CCDEPMODE = @CCDEPMODE@ CFLAGS = @CFLAGS@ CHECK_PL_OPTS = @CHECK_PL_OPTS@ CPP = @CPP@ CPPFLAGS = @CPPFLAGS@ CYGPATH_W = @CYGPATH_W@ C_FFTW_R2R_KIND = @C_FFTW_R2R_KIND@ C_MPI_FINT = @C_MPI_FINT@ DEFS = @DEFS@ DEPDIR = @DEPDIR@ DLLTOOL = @DLLTOOL@ DSYMUTIL = @DSYMUTIL@ DUMPBIN = @DUMPBIN@ ECHO_C = @ECHO_C@ ECHO_N = @ECHO_N@ ECHO_T = @ECHO_T@ EGREP = @EGREP@ EXEEXT = @EXEEXT@ F77 = @F77@ FFLAGS = @FFLAGS@ FGREP = @FGREP@ FLIBS = @FLIBS@ GREP = @GREP@ INSTALL = @INSTALL@ INSTALL_DATA = @INSTALL_DATA@ INSTALL_PROGRAM = @INSTALL_PROGRAM@ INSTALL_SCRIPT = @INSTALL_SCRIPT@ INSTALL_STRIP_PROGRAM = @INSTALL_STRIP_PROGRAM@ LD = @LD@ LDFLAGS = @LDFLAGS@ LIBOBJS = @LIBOBJS@ LIBQUADMATH = @LIBQUADMATH@ LIBS = @LIBS@ LIBTOOL = @LIBTOOL@ LIPO = @LIPO@ LN_S = @LN_S@ LTLIBOBJS = @LTLIBOBJS@ MAINT = @MAINT@ MAKEINFO = @MAKEINFO@ MANIFEST_TOOL = @MANIFEST_TOOL@ MKDIR_P = @MKDIR_P@ MPICC = @MPICC@ MPILIBS = @MPILIBS@ MPIRUN = @MPIRUN@ NEON_CFLAGS = @NEON_CFLAGS@ NM = @NM@ NMEDIT = @NMEDIT@ OBJDUMP = @OBJDUMP@ OBJEXT = @OBJEXT@ OCAMLBUILD = @OCAMLBUILD@ OPENMP_CFLAGS = @OPENMP_CFLAGS@ OTOOL = @OTOOL@ OTOOL64 = @OTOOL64@ PACKAGE = @PACKAGE@ PACKAGE_BUGREPORT = @PACKAGE_BUGREPORT@ PACKAGE_NAME = @PACKAGE_NAME@ PACKAGE_STRING = @PACKAGE_STRING@ PACKAGE_TARNAME = @PACKAGE_TARNAME@ PACKAGE_URL = @PACKAGE_URL@ PACKAGE_VERSION = @PACKAGE_VERSION@ PATH_SEPARATOR = @PATH_SEPARATOR@ POW_LIB = @POW_LIB@ PRECISION = @PRECISION@ PREC_SUFFIX = @PREC_SUFFIX@ PTHREAD_CC = @PTHREAD_CC@ PTHREAD_CFLAGS = @PTHREAD_CFLAGS@ PTHREAD_LIBS = @PTHREAD_LIBS@ RANLIB = @RANLIB@ SED = @SED@ SET_MAKE = @SET_MAKE@ SHARED_VERSION_INFO = @SHARED_VERSION_INFO@ SHELL = @SHELL@ SSE2_CFLAGS = @SSE2_CFLAGS@ STACK_ALIGN_CFLAGS = @STACK_ALIGN_CFLAGS@ STRIP = @STRIP@ THREADLIBS = @THREADLIBS@ VERSION = @VERSION@ abs_builddir = @abs_builddir@ abs_srcdir = @abs_srcdir@ abs_top_builddir = @abs_top_builddir@ abs_top_srcdir = @abs_top_srcdir@ ac_ct_AR = @ac_ct_AR@ ac_ct_CC = @ac_ct_CC@ ac_ct_DUMPBIN = @ac_ct_DUMPBIN@ ac_ct_F77 = @ac_ct_F77@ acx_pthread_config = @acx_pthread_config@ am__include = @am__include@ am__leading_dot = @am__leading_dot@ am__quote = @am__quote@ am__tar = @am__tar@ am__untar = @am__untar@ bindir = @bindir@ build = @build@ build_alias = @build_alias@ build_cpu = @build_cpu@ build_os = @build_os@ build_vendor = @build_vendor@ builddir = @builddir@ datadir = @datadir@ datarootdir = @datarootdir@ docdir = @docdir@ dvidir = @dvidir@ exec_prefix = @exec_prefix@ host = @host@ host_alias = @host_alias@ host_cpu = @host_cpu@ host_os = @host_os@ host_vendor = @host_vendor@ htmldir = @htmldir@ includedir = @includedir@ infodir = @infodir@ install_sh = @install_sh@ libdir = @libdir@ libexecdir = @libexecdir@ localedir = @localedir@ localstatedir = @localstatedir@ mandir = @mandir@ mkdir_p = @mkdir_p@ oldincludedir = @oldincludedir@ pdfdir = @pdfdir@ prefix = @prefix@ program_transform_name = @program_transform_name@ psdir = @psdir@ sbindir = @sbindir@ sharedstatedir = @sharedstatedir@ srcdir = @srcdir@ sysconfdir = @sysconfdir@ target_alias = @target_alias@ top_build_prefix = @top_build_prefix@ top_builddir = @top_builddir@ top_srcdir = @top_srcdir@ AM_CPPFLAGS = -I$(top_srcdir)/kernel SUBDIRS = scalar simd noinst_LTLIBRARIES = libdft.la # pkgincludedir = $(includedir)/fftw3@PREC_SUFFIX@ # pkginclude_HEADERS = codelet-dft.h dft.h libdft_la_SOURCES = bluestein.c buffered.c conf.c ct.c dftw-direct.c \ dftw-directsq.c dftw-generic.c dftw-genericbuf.c direct.c generic.c \ indirect.c indirect-transpose.c kdft-dif.c kdft-difsq.c kdft-dit.c \ kdft.c nop.c plan.c problem.c rader.c rank-geq2.c solve.c vrank-geq1.c \ zero.c codelet-dft.h ct.h dft.h all: all-recursive .SUFFIXES: .SUFFIXES: .c .lo .o .obj $(srcdir)/Makefile.in: @MAINTAINER_MODE_TRUE@ $(srcdir)/Makefile.am $(am__configure_deps) @for dep in $?; do \ case '$(am__configure_deps)' in \ *$$dep*) \ ( cd $(top_builddir) && $(MAKE) $(AM_MAKEFLAGS) am--refresh ) \ && { if test -f $@; then exit 0; else break; fi; }; \ exit 1;; \ esac; \ done; \ echo ' cd $(top_srcdir) && $(AUTOMAKE) --gnu dft/Makefile'; \ $(am__cd) $(top_srcdir) && \ $(AUTOMAKE) --gnu dft/Makefile .PRECIOUS: Makefile Makefile: $(srcdir)/Makefile.in $(top_builddir)/config.status @case '$?' in \ *config.status*) \ cd $(top_builddir) && $(MAKE) $(AM_MAKEFLAGS) am--refresh;; \ *) \ echo ' cd $(top_builddir) && $(SHELL) ./config.status $(subdir)/$@ $(am__depfiles_maybe)'; \ cd $(top_builddir) && $(SHELL) ./config.status $(subdir)/$@ $(am__depfiles_maybe);; \ esac; $(top_builddir)/config.status: $(top_srcdir)/configure $(CONFIG_STATUS_DEPENDENCIES) cd $(top_builddir) && $(MAKE) $(AM_MAKEFLAGS) am--refresh $(top_srcdir)/configure: @MAINTAINER_MODE_TRUE@ $(am__configure_deps) cd $(top_builddir) && $(MAKE) $(AM_MAKEFLAGS) am--refresh $(ACLOCAL_M4): @MAINTAINER_MODE_TRUE@ $(am__aclocal_m4_deps) cd $(top_builddir) && $(MAKE) $(AM_MAKEFLAGS) am--refresh $(am__aclocal_m4_deps): clean-noinstLTLIBRARIES: -test -z "$(noinst_LTLIBRARIES)" || rm -f $(noinst_LTLIBRARIES) @list='$(noinst_LTLIBRARIES)'; for p in $$list; do \ dir="`echo $$p | sed -e 's|/[^/]*$$||'`"; \ test "$$dir" != "$$p" || dir=.; \ echo "rm -f \"$${dir}/so_locations\""; \ rm -f "$${dir}/so_locations"; \ done libdft.la: $(libdft_la_OBJECTS) $(libdft_la_DEPENDENCIES) $(EXTRA_libdft_la_DEPENDENCIES) $(LINK) $(libdft_la_OBJECTS) $(libdft_la_LIBADD) $(LIBS) mostlyclean-compile: -rm -f *.$(OBJEXT) distclean-compile: -rm -f *.tab.c @AMDEP_TRUE@@am__include@ @am__quote@./$(DEPDIR)/bluestein.Plo@am__quote@ @AMDEP_TRUE@@am__include@ @am__quote@./$(DEPDIR)/buffered.Plo@am__quote@ @AMDEP_TRUE@@am__include@ @am__quote@./$(DEPDIR)/conf.Plo@am__quote@ @AMDEP_TRUE@@am__include@ @am__quote@./$(DEPDIR)/ct.Plo@am__quote@ @AMDEP_TRUE@@am__include@ @am__quote@./$(DEPDIR)/dftw-direct.Plo@am__quote@ @AMDEP_TRUE@@am__include@ @am__quote@./$(DEPDIR)/dftw-directsq.Plo@am__quote@ @AMDEP_TRUE@@am__include@ @am__quote@./$(DEPDIR)/dftw-generic.Plo@am__quote@ @AMDEP_TRUE@@am__include@ @am__quote@./$(DEPDIR)/dftw-genericbuf.Plo@am__quote@ @AMDEP_TRUE@@am__include@ @am__quote@./$(DEPDIR)/direct.Plo@am__quote@ @AMDEP_TRUE@@am__include@ @am__quote@./$(DEPDIR)/generic.Plo@am__quote@ @AMDEP_TRUE@@am__include@ @am__quote@./$(DEPDIR)/indirect-transpose.Plo@am__quote@ @AMDEP_TRUE@@am__include@ @am__quote@./$(DEPDIR)/indirect.Plo@am__quote@ @AMDEP_TRUE@@am__include@ @am__quote@./$(DEPDIR)/kdft-dif.Plo@am__quote@ @AMDEP_TRUE@@am__include@ @am__quote@./$(DEPDIR)/kdft-difsq.Plo@am__quote@ @AMDEP_TRUE@@am__include@ @am__quote@./$(DEPDIR)/kdft-dit.Plo@am__quote@ @AMDEP_TRUE@@am__include@ @am__quote@./$(DEPDIR)/kdft.Plo@am__quote@ @AMDEP_TRUE@@am__include@ @am__quote@./$(DEPDIR)/nop.Plo@am__quote@ @AMDEP_TRUE@@am__include@ @am__quote@./$(DEPDIR)/plan.Plo@am__quote@ @AMDEP_TRUE@@am__include@ @am__quote@./$(DEPDIR)/problem.Plo@am__quote@ @AMDEP_TRUE@@am__include@ @am__quote@./$(DEPDIR)/rader.Plo@am__quote@ @AMDEP_TRUE@@am__include@ @am__quote@./$(DEPDIR)/rank-geq2.Plo@am__quote@ @AMDEP_TRUE@@am__include@ @am__quote@./$(DEPDIR)/solve.Plo@am__quote@ @AMDEP_TRUE@@am__include@ @am__quote@./$(DEPDIR)/vrank-geq1.Plo@am__quote@ @AMDEP_TRUE@@am__include@ @am__quote@./$(DEPDIR)/zero.Plo@am__quote@ .c.o: @am__fastdepCC_TRUE@ $(COMPILE) -MT $@ -MD -MP -MF $(DEPDIR)/$*.Tpo -c -o $@ $< @am__fastdepCC_TRUE@ $(am__mv) $(DEPDIR)/$*.Tpo $(DEPDIR)/$*.Po @AMDEP_TRUE@@am__fastdepCC_FALSE@ source='$<' object='$@' libtool=no @AMDEPBACKSLASH@ @AMDEP_TRUE@@am__fastdepCC_FALSE@ DEPDIR=$(DEPDIR) $(CCDEPMODE) $(depcomp) @AMDEPBACKSLASH@ @am__fastdepCC_FALSE@ $(COMPILE) -c $< .c.obj: @am__fastdepCC_TRUE@ $(COMPILE) -MT $@ -MD -MP -MF $(DEPDIR)/$*.Tpo -c -o $@ `$(CYGPATH_W) '$<'` @am__fastdepCC_TRUE@ $(am__mv) $(DEPDIR)/$*.Tpo $(DEPDIR)/$*.Po @AMDEP_TRUE@@am__fastdepCC_FALSE@ source='$<' object='$@' libtool=no @AMDEPBACKSLASH@ @AMDEP_TRUE@@am__fastdepCC_FALSE@ DEPDIR=$(DEPDIR) $(CCDEPMODE) $(depcomp) @AMDEPBACKSLASH@ @am__fastdepCC_FALSE@ $(COMPILE) -c `$(CYGPATH_W) '$<'` .c.lo: @am__fastdepCC_TRUE@ $(LTCOMPILE) -MT $@ -MD -MP -MF $(DEPDIR)/$*.Tpo -c -o $@ $< @am__fastdepCC_TRUE@ $(am__mv) $(DEPDIR)/$*.Tpo $(DEPDIR)/$*.Plo @AMDEP_TRUE@@am__fastdepCC_FALSE@ source='$<' object='$@' libtool=yes @AMDEPBACKSLASH@ @AMDEP_TRUE@@am__fastdepCC_FALSE@ DEPDIR=$(DEPDIR) $(CCDEPMODE) $(depcomp) @AMDEPBACKSLASH@ @am__fastdepCC_FALSE@ $(LTCOMPILE) -c -o $@ $< mostlyclean-libtool: -rm -f *.lo clean-libtool: -rm -rf .libs _libs # This directory's subdirectories are mostly independent; you can cd # into them and run `make' without going through this Makefile. # To change the values of `make' variables: instead of editing Makefiles, # (1) if the variable is set in `config.status', edit `config.status' # (which will cause the Makefiles to be regenerated when you run `make'); # (2) otherwise, pass the desired values on the `make' command line. $(RECURSIVE_TARGETS): @fail= failcom='exit 1'; \ for f in x $$MAKEFLAGS; do \ case $$f in \ *=* | --[!k]*);; \ *k*) failcom='fail=yes';; \ esac; \ done; \ dot_seen=no; \ target=`echo $@ | sed s/-recursive//`; \ list='$(SUBDIRS)'; for subdir in $$list; do \ echo "Making $$target in $$subdir"; \ if test "$$subdir" = "."; then \ dot_seen=yes; \ local_target="$$target-am"; \ else \ local_target="$$target"; \ fi; \ ($(am__cd) $$subdir && $(MAKE) $(AM_MAKEFLAGS) $$local_target) \ || eval $$failcom; \ done; \ if test "$$dot_seen" = "no"; then \ $(MAKE) $(AM_MAKEFLAGS) "$$target-am" || exit 1; \ fi; test -z "$$fail" $(RECURSIVE_CLEAN_TARGETS): @fail= failcom='exit 1'; \ for f in x $$MAKEFLAGS; do \ case $$f in \ *=* | --[!k]*);; \ *k*) failcom='fail=yes';; \ esac; \ done; \ dot_seen=no; \ case "$@" in \ distclean-* | maintainer-clean-*) list='$(DIST_SUBDIRS)' ;; \ *) list='$(SUBDIRS)' ;; \ esac; \ rev=''; for subdir in $$list; do \ if test "$$subdir" = "."; then :; else \ rev="$$subdir $$rev"; \ fi; \ done; \ rev="$$rev ."; \ target=`echo $@ | sed s/-recursive//`; \ for subdir in $$rev; do \ echo "Making $$target in $$subdir"; \ if test "$$subdir" = "."; then \ local_target="$$target-am"; \ else \ local_target="$$target"; \ fi; \ ($(am__cd) $$subdir && $(MAKE) $(AM_MAKEFLAGS) $$local_target) \ || eval $$failcom; \ done && test -z "$$fail" tags-recursive: list='$(SUBDIRS)'; for subdir in $$list; do \ test "$$subdir" = . || ($(am__cd) $$subdir && $(MAKE) $(AM_MAKEFLAGS) tags); \ done ctags-recursive: list='$(SUBDIRS)'; for subdir in $$list; do \ test "$$subdir" = . || ($(am__cd) $$subdir && $(MAKE) $(AM_MAKEFLAGS) ctags); \ done ID: $(HEADERS) $(SOURCES) $(LISP) $(TAGS_FILES) list='$(SOURCES) $(HEADERS) $(LISP) $(TAGS_FILES)'; \ unique=`for i in $$list; do \ if test -f "$$i"; then echo $$i; else echo $(srcdir)/$$i; fi; \ done | \ $(AWK) '{ files[$$0] = 1; nonempty = 1; } \ END { if (nonempty) { for (i in files) print i; }; }'`; \ mkid -fID $$unique tags: TAGS TAGS: tags-recursive $(HEADERS) $(SOURCES) $(TAGS_DEPENDENCIES) \ $(TAGS_FILES) $(LISP) set x; \ here=`pwd`; \ if ($(ETAGS) --etags-include --version) >/dev/null 2>&1; then \ include_option=--etags-include; \ empty_fix=.; \ else \ include_option=--include; \ empty_fix=; \ fi; \ list='$(SUBDIRS)'; for subdir in $$list; do \ if test "$$subdir" = .; then :; else \ test ! -f $$subdir/TAGS || \ set "$$@" "$$include_option=$$here/$$subdir/TAGS"; \ fi; \ done; \ list='$(SOURCES) $(HEADERS) $(LISP) $(TAGS_FILES)'; \ unique=`for i in $$list; do \ if test -f "$$i"; then echo $$i; else echo $(srcdir)/$$i; fi; \ done | \ $(AWK) '{ files[$$0] = 1; nonempty = 1; } \ END { if (nonempty) { for (i in files) print i; }; }'`; \ shift; \ if test -z "$(ETAGS_ARGS)$$*$$unique"; then :; else \ test -n "$$unique" || unique=$$empty_fix; \ if test $$# -gt 0; then \ $(ETAGS) $(ETAGSFLAGS) $(AM_ETAGSFLAGS) $(ETAGS_ARGS) \ "$$@" $$unique; \ else \ $(ETAGS) $(ETAGSFLAGS) $(AM_ETAGSFLAGS) $(ETAGS_ARGS) \ $$unique; \ fi; \ fi ctags: CTAGS CTAGS: ctags-recursive $(HEADERS) $(SOURCES) $(TAGS_DEPENDENCIES) \ $(TAGS_FILES) $(LISP) list='$(SOURCES) $(HEADERS) $(LISP) $(TAGS_FILES)'; \ unique=`for i in $$list; do \ if test -f "$$i"; then echo $$i; else echo $(srcdir)/$$i; fi; \ done | \ $(AWK) '{ files[$$0] = 1; nonempty = 1; } \ END { if (nonempty) { for (i in files) print i; }; }'`; \ test -z "$(CTAGS_ARGS)$$unique" \ || $(CTAGS) $(CTAGSFLAGS) $(AM_CTAGSFLAGS) $(CTAGS_ARGS) \ $$unique GTAGS: here=`$(am__cd) $(top_builddir) && pwd` \ && $(am__cd) $(top_srcdir) \ && gtags -i $(GTAGS_ARGS) "$$here" distclean-tags: -rm -f TAGS ID GTAGS GRTAGS GSYMS GPATH tags distdir: $(DISTFILES) @srcdirstrip=`echo "$(srcdir)" | sed 's/[].[^$$\\*]/\\\\&/g'`; \ topsrcdirstrip=`echo "$(top_srcdir)" | sed 's/[].[^$$\\*]/\\\\&/g'`; \ list='$(DISTFILES)'; \ dist_files=`for file in $$list; do echo $$file; done | \ sed -e "s|^$$srcdirstrip/||;t" \ -e "s|^$$topsrcdirstrip/|$(top_builddir)/|;t"`; \ case $$dist_files in \ */*) $(MKDIR_P) `echo "$$dist_files" | \ sed '/\//!d;s|^|$(distdir)/|;s,/[^/]*$$,,' | \ sort -u` ;; \ esac; \ for file in $$dist_files; do \ if test -f $$file || test -d $$file; then d=.; else d=$(srcdir); fi; \ if test -d $$d/$$file; then \ dir=`echo "/$$file" | sed -e 's,/[^/]*$$,,'`; \ if test -d "$(distdir)/$$file"; then \ find "$(distdir)/$$file" -type d ! -perm -700 -exec chmod u+rwx {} \;; \ fi; \ if test -d $(srcdir)/$$file && test $$d != $(srcdir); then \ cp -fpR $(srcdir)/$$file "$(distdir)$$dir" || exit 1; \ find "$(distdir)/$$file" -type d ! -perm -700 -exec chmod u+rwx {} \;; \ fi; \ cp -fpR $$d/$$file "$(distdir)$$dir" || exit 1; \ else \ test -f "$(distdir)/$$file" \ || cp -p $$d/$$file "$(distdir)/$$file" \ || exit 1; \ fi; \ done @list='$(DIST_SUBDIRS)'; for subdir in $$list; do \ if test "$$subdir" = .; then :; else \ $(am__make_dryrun) \ || test -d "$(distdir)/$$subdir" \ || $(MKDIR_P) "$(distdir)/$$subdir" \ || exit 1; \ dir1=$$subdir; dir2="$(distdir)/$$subdir"; \ $(am__relativize); \ new_distdir=$$reldir; \ dir1=$$subdir; dir2="$(top_distdir)"; \ $(am__relativize); \ new_top_distdir=$$reldir; \ echo " (cd $$subdir && $(MAKE) $(AM_MAKEFLAGS) top_distdir="$$new_top_distdir" distdir="$$new_distdir" \\"; \ echo " am__remove_distdir=: am__skip_length_check=: am__skip_mode_fix=: distdir)"; \ ($(am__cd) $$subdir && \ $(MAKE) $(AM_MAKEFLAGS) \ top_distdir="$$new_top_distdir" \ distdir="$$new_distdir" \ am__remove_distdir=: \ am__skip_length_check=: \ am__skip_mode_fix=: \ distdir) \ || exit 1; \ fi; \ done check-am: all-am check: check-recursive all-am: Makefile $(LTLIBRARIES) installdirs: installdirs-recursive installdirs-am: install: install-recursive install-exec: install-exec-recursive install-data: install-data-recursive uninstall: uninstall-recursive install-am: all-am @$(MAKE) $(AM_MAKEFLAGS) install-exec-am install-data-am installcheck: installcheck-recursive install-strip: if test -z '$(STRIP)'; then \ $(MAKE) $(AM_MAKEFLAGS) INSTALL_PROGRAM="$(INSTALL_STRIP_PROGRAM)" \ install_sh_PROGRAM="$(INSTALL_STRIP_PROGRAM)" INSTALL_STRIP_FLAG=-s \ install; \ else \ $(MAKE) $(AM_MAKEFLAGS) INSTALL_PROGRAM="$(INSTALL_STRIP_PROGRAM)" \ install_sh_PROGRAM="$(INSTALL_STRIP_PROGRAM)" INSTALL_STRIP_FLAG=-s \ "INSTALL_PROGRAM_ENV=STRIPPROG='$(STRIP)'" install; \ fi mostlyclean-generic: clean-generic: distclean-generic: -test -z "$(CONFIG_CLEAN_FILES)" || rm -f $(CONFIG_CLEAN_FILES) -test . = "$(srcdir)" || test -z "$(CONFIG_CLEAN_VPATH_FILES)" || rm -f $(CONFIG_CLEAN_VPATH_FILES) maintainer-clean-generic: @echo "This command is intended for maintainers to use" @echo "it deletes files that may require special tools to rebuild." clean: clean-recursive clean-am: clean-generic clean-libtool clean-noinstLTLIBRARIES \ mostlyclean-am distclean: distclean-recursive -rm -rf ./$(DEPDIR) -rm -f Makefile distclean-am: clean-am distclean-compile distclean-generic \ distclean-tags dvi: dvi-recursive dvi-am: html: html-recursive html-am: info: info-recursive info-am: install-data-am: install-dvi: install-dvi-recursive install-dvi-am: install-exec-am: install-html: install-html-recursive install-html-am: install-info: install-info-recursive install-info-am: install-man: install-pdf: install-pdf-recursive install-pdf-am: install-ps: install-ps-recursive install-ps-am: installcheck-am: maintainer-clean: maintainer-clean-recursive -rm -rf ./$(DEPDIR) -rm -f Makefile maintainer-clean-am: distclean-am maintainer-clean-generic mostlyclean: mostlyclean-recursive mostlyclean-am: mostlyclean-compile mostlyclean-generic \ mostlyclean-libtool pdf: pdf-recursive pdf-am: ps: ps-recursive ps-am: uninstall-am: .MAKE: $(RECURSIVE_CLEAN_TARGETS) $(RECURSIVE_TARGETS) ctags-recursive \ install-am install-strip tags-recursive .PHONY: $(RECURSIVE_CLEAN_TARGETS) $(RECURSIVE_TARGETS) CTAGS GTAGS \ all all-am check check-am clean clean-generic clean-libtool \ clean-noinstLTLIBRARIES ctags ctags-recursive distclean \ distclean-compile distclean-generic distclean-libtool \ distclean-tags distdir dvi dvi-am html html-am info info-am \ install install-am install-data install-data-am install-dvi \ install-dvi-am install-exec install-exec-am install-html \ install-html-am install-info install-info-am install-man \ install-pdf install-pdf-am install-ps install-ps-am \ install-strip installcheck installcheck-am installdirs \ installdirs-am maintainer-clean maintainer-clean-generic \ mostlyclean mostlyclean-compile mostlyclean-generic \ mostlyclean-libtool pdf pdf-am ps ps-am tags tags-recursive \ uninstall uninstall-am # Tell versions [3.59,3.63) of GNU make to not export all variables. # Otherwise a system limit (for SysV at least) may be exceeded. .NOEXPORT: fftw-3.3.3/dft/bluestein.c0000644000175000017500000001444112054410050012315 00000000000000/* * Copyright (c) 2003, 2007-11 Matteo Frigo * Copyright (c) 2003, 2007-11 Massachusetts Institute of Technology * * This program is free software; you can redistribute it and/or modify * it under the terms of the GNU General Public License as published by * the Free Software Foundation; either version 2 of the License, or * (at your option) any later version. * * This program is distributed in the hope that it will be useful, * but WITHOUT ANY WARRANTY; without even the implied warranty of * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the * GNU General Public License for more details. * * You should have received a copy of the GNU General Public License * along with this program; if not, write to the Free Software * Foundation, Inc., 51 Franklin Street, Fifth Floor, Boston, MA 02110-1301 USA * */ #include "dft.h" typedef struct { solver super; } S; typedef struct { plan_dft super; INT n; /* problem size */ INT nb; /* size of convolution */ R *w; /* lambda k . exp(2*pi*i*k^2/(2*n)) */ R *W; /* DFT(w) */ plan *cldf; INT is, os; } P; static void bluestein_sequence(enum wakefulness wakefulness, INT n, R *w) { INT k, ksq, n2 = 2 * n; triggen *t = X(mktriggen)(wakefulness, n2); ksq = 0; for (k = 0; k < n; ++k) { t->cexp(t, ksq, w+2*k); /* careful with overflow */ ksq += 2*k + 1; while (ksq > n2) ksq -= n2; } X(triggen_destroy)(t); } static void mktwiddle(enum wakefulness wakefulness, P *p) { INT i; INT n = p->n, nb = p->nb; R *w, *W; E nbf = (E)nb; p->w = w = (R *) MALLOC(2 * n * sizeof(R), TWIDDLES); p->W = W = (R *) MALLOC(2 * nb * sizeof(R), TWIDDLES); bluestein_sequence(wakefulness, n, w); for (i = 0; i < nb; ++i) W[2*i] = W[2*i+1] = K(0.0); W[0] = w[0] / nbf; W[1] = w[1] / nbf; for (i = 1; i < n; ++i) { W[2*i] = W[2*(nb-i)] = w[2*i] / nbf; W[2*i+1] = W[2*(nb-i)+1] = w[2*i+1] / nbf; } { plan_dft *cldf = (plan_dft *)p->cldf; /* cldf must be awake */ cldf->apply(p->cldf, W, W+1, W, W+1); } } static void apply(const plan *ego_, R *ri, R *ii, R *ro, R *io) { const P *ego = (const P *) ego_; INT i, n = ego->n, nb = ego->nb, is = ego->is, os = ego->os; R *w = ego->w, *W = ego->W; R *b = (R *) MALLOC(2 * nb * sizeof(R), BUFFERS); /* multiply input by conjugate bluestein sequence */ for (i = 0; i < n; ++i) { E xr = ri[i*is], xi = ii[i*is]; E wr = w[2*i], wi = w[2*i+1]; b[2*i] = xr * wr + xi * wi; b[2*i+1] = xi * wr - xr * wi; } for (; i < nb; ++i) b[2*i] = b[2*i+1] = K(0.0); /* convolution: FFT */ { plan_dft *cldf = (plan_dft *)ego->cldf; cldf->apply(ego->cldf, b, b+1, b, b+1); } /* convolution: pointwise multiplication */ for (i = 0; i < nb; ++i) { E xr = b[2*i], xi = b[2*i+1]; E wr = W[2*i], wi = W[2*i+1]; b[2*i] = xi * wr + xr * wi; b[2*i+1] = xr * wr - xi * wi; } /* convolution: IFFT by FFT with real/imag input/output swapped */ { plan_dft *cldf = (plan_dft *)ego->cldf; cldf->apply(ego->cldf, b, b+1, b, b+1); } /* multiply output by conjugate bluestein sequence */ for (i = 0; i < n; ++i) { E xi = b[2*i], xr = b[2*i+1]; E wr = w[2*i], wi = w[2*i+1]; ro[i*os] = xr * wr + xi * wi; io[i*os] = xi * wr - xr * wi; } X(ifree)(b); } static void awake(plan *ego_, enum wakefulness wakefulness) { P *ego = (P *) ego_; X(plan_awake)(ego->cldf, wakefulness); switch (wakefulness) { case SLEEPY: X(ifree0)(ego->w); ego->w = 0; X(ifree0)(ego->W); ego->W = 0; break; default: A(!ego->w); mktwiddle(wakefulness, ego); break; } } static int applicable(const solver *ego, const problem *p_, const planner *plnr) { const problem_dft *p = (const problem_dft *) p_; UNUSED(ego); return (1 && p->sz->rnk == 1 && p->vecsz->rnk == 0 /* FIXME: allow other sizes */ && X(is_prime)(p->sz->dims[0].n) /* FIXME: avoid infinite recursion of bluestein with itself. This works because all factors in child problems are 2, 3, 5 */ && p->sz->dims[0].n > 16 && CIMPLIES(NO_SLOWP(plnr), p->sz->dims[0].n > BLUESTEIN_MAX_SLOW) ); } static void destroy(plan *ego_) { P *ego = (P *) ego_; X(plan_destroy_internal)(ego->cldf); } static void print(const plan *ego_, printer *p) { const P *ego = (const P *)ego_; p->print(p, "(dft-bluestein-%D/%D%(%p%))", ego->n, ego->nb, ego->cldf); } static INT choose_transform_size(INT minsz) { while (!X(factors_into_small_primes)(minsz)) ++minsz; return minsz; } static plan *mkplan(const solver *ego, const problem *p_, planner *plnr) { const problem_dft *p = (const problem_dft *) p_; P *pln; INT n, nb; plan *cldf = 0; R *buf = (R *) 0; static const plan_adt padt = { X(dft_solve), awake, print, destroy }; if (!applicable(ego, p_, plnr)) return (plan *) 0; n = p->sz->dims[0].n; nb = choose_transform_size(2 * n - 1); buf = (R *) MALLOC(2 * nb * sizeof(R), BUFFERS); cldf = X(mkplan_f_d)(plnr, X(mkproblem_dft_d)(X(mktensor_1d)(nb, 2, 2), X(mktensor_1d)(1, 0, 0), buf, buf+1, buf, buf+1), NO_SLOW, 0, 0); if (!cldf) goto nada; X(ifree)(buf); pln = MKPLAN_DFT(P, &padt, apply); pln->n = n; pln->nb = nb; pln->w = 0; pln->W = 0; pln->cldf = cldf; pln->is = p->sz->dims[0].is; pln->os = p->sz->dims[0].os; X(ops_add)(&cldf->ops, &cldf->ops, &pln->super.super.ops); pln->super.super.ops.add += 4 * n + 2 * nb; pln->super.super.ops.mul += 8 * n + 4 * nb; pln->super.super.ops.other += 6 * (n + nb); return &(pln->super.super); nada: X(ifree0)(buf); X(plan_destroy_internal)(cldf); return (plan *)0; } static solver *mksolver(void) { static const solver_adt sadt = { PROBLEM_DFT, mkplan, 0 }; S *slv = MKSOLVER(S, &sadt); return &(slv->super); } void X(dft_bluestein_register)(planner *p) { REGISTER_SOLVER(p, mksolver()); } fftw-3.3.3/dft/buffered.c0000644000175000017500000001746712054410050012120 00000000000000/* * Copyright (c) 2003, 2007-11 Matteo Frigo * Copyright (c) 2003, 2007-11 Massachusetts Institute of Technology * * This program is free software; you can redistribute it and/or modify * it under the terms of the GNU General Public License as published by * the Free Software Foundation; either version 2 of the License, or * (at your option) any later version. * * This program is distributed in the hope that it will be useful, * but WITHOUT ANY WARRANTY; without even the implied warranty of * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the * GNU General Public License for more details. * * You should have received a copy of the GNU General Public License * along with this program; if not, write to the Free Software * Foundation, Inc., 51 Franklin Street, Fifth Floor, Boston, MA 02110-1301 USA * */ #include "dft.h" typedef struct { solver super; int maxnbuf_ndx; } S; static const INT maxnbufs[] = { 8, 256 }; typedef struct { plan_dft super; plan *cld, *cldcpy, *cldrest; INT n, vl, nbuf, bufdist; INT ivs_by_nbuf, ovs_by_nbuf; INT roffset, ioffset; } P; /* transform a vector input with the help of bufs */ static void apply(const plan *ego_, R *ri, R *ii, R *ro, R *io) { const P *ego = (const P *) ego_; INT nbuf = ego->nbuf; R *bufs = (R *)MALLOC(sizeof(R) * nbuf * ego->bufdist * 2, BUFFERS); plan_dft *cld = (plan_dft *) ego->cld; plan_dft *cldcpy = (plan_dft *) ego->cldcpy; plan_dft *cldrest; INT i, vl = ego->vl; INT ivs_by_nbuf = ego->ivs_by_nbuf, ovs_by_nbuf = ego->ovs_by_nbuf; INT roffset = ego->roffset, ioffset = ego->ioffset; for (i = nbuf; i <= vl; i += nbuf) { /* transform to bufs: */ cld->apply((plan *) cld, ri, ii, bufs + roffset, bufs + ioffset); ri += ivs_by_nbuf; ii += ivs_by_nbuf; /* copy back */ cldcpy->apply((plan *) cldcpy, bufs+roffset, bufs+ioffset, ro, io); ro += ovs_by_nbuf; io += ovs_by_nbuf; } X(ifree)(bufs); /* Do the remaining transforms, if any: */ cldrest = (plan_dft *) ego->cldrest; cldrest->apply((plan *) cldrest, ri, ii, ro, io); } static void awake(plan *ego_, enum wakefulness wakefulness) { P *ego = (P *) ego_; X(plan_awake)(ego->cld, wakefulness); X(plan_awake)(ego->cldcpy, wakefulness); X(plan_awake)(ego->cldrest, wakefulness); } static void destroy(plan *ego_) { P *ego = (P *) ego_; X(plan_destroy_internal)(ego->cldrest); X(plan_destroy_internal)(ego->cldcpy); X(plan_destroy_internal)(ego->cld); } static void print(const plan *ego_, printer *p) { const P *ego = (const P *) ego_; p->print(p, "(dft-buffered-%D%v/%D-%D%(%p%)%(%p%)%(%p%))", ego->n, ego->nbuf, ego->vl, ego->bufdist % ego->n, ego->cld, ego->cldcpy, ego->cldrest); } static int applicable0(const S *ego, const problem *p_, const planner *plnr) { const problem_dft *p = (const problem_dft *) p_; const iodim *d = p->sz->dims; if (1 && p->vecsz->rnk <= 1 && p->sz->rnk == 1 ) { INT vl, ivs, ovs; X(tensor_tornk1)(p->vecsz, &vl, &ivs, &ovs); if (X(toobig)(p->sz->dims[0].n) && CONSERVE_MEMORYP(plnr)) return 0; /* if this solver is redundant, in the sense that a solver of lower index generates the same plan, then prune this solver */ if (X(nbuf_redundant)(d[0].n, vl, ego->maxnbuf_ndx, maxnbufs, NELEM(maxnbufs))) return 0; /* In principle, the buffered transforms might be useful when working out of place. However, in order to prevent infinite loops in the planner, we require that the output stride of the buffered transforms be greater than 2. */ if (p->ri != p->ro) return (d[0].os > 2); /* * If the problem is in place, the input/output strides must * be the same or the whole thing must fit in the buffer. */ if (X(tensor_inplace_strides2)(p->sz, p->vecsz)) return 1; if (/* fits into buffer: */ ((p->vecsz->rnk == 0) || (X(nbuf)(d[0].n, p->vecsz->dims[0].n, maxnbufs[ego->maxnbuf_ndx]) == p->vecsz->dims[0].n))) return 1; } return 0; } static int applicable(const S *ego, const problem *p_, const planner *plnr) { if (NO_BUFFERINGP(plnr)) return 0; if (!applicable0(ego, p_, plnr)) return 0; if (NO_UGLYP(plnr)) { const problem_dft *p = (const problem_dft *) p_; if (p->ri != p->ro) return 0; if (X(toobig)(p->sz->dims[0].n)) return 0; } return 1; } static plan *mkplan(const solver *ego_, const problem *p_, planner *plnr) { P *pln; const S *ego = (const S *)ego_; plan *cld = (plan *) 0; plan *cldcpy = (plan *) 0; plan *cldrest = (plan *) 0; const problem_dft *p = (const problem_dft *) p_; R *bufs = (R *) 0; INT nbuf = 0, bufdist, n, vl; INT ivs, ovs, roffset, ioffset; static const plan_adt padt = { X(dft_solve), awake, print, destroy }; if (!applicable(ego, p_, plnr)) goto nada; n = X(tensor_sz)(p->sz); X(tensor_tornk1)(p->vecsz, &vl, &ivs, &ovs); nbuf = X(nbuf)(n, vl, maxnbufs[ego->maxnbuf_ndx]); bufdist = X(bufdist)(n, vl); A(nbuf > 0); /* attempt to keep real and imaginary part in the same order, so as to allow optimizations in the the copy plan */ roffset = (p->ri - p->ii > 0) ? (INT)1 : (INT)0; ioffset = 1 - roffset; /* initial allocation for the purpose of planning */ bufs = (R *) MALLOC(sizeof(R) * nbuf * bufdist * 2, BUFFERS); /* allow destruction of input if problem is in place */ cld = X(mkplan_f_d)(plnr, X(mkproblem_dft_d)( X(mktensor_1d)(n, p->sz->dims[0].is, 2), X(mktensor_1d)(nbuf, ivs, bufdist * 2), TAINT(p->ri, ivs * nbuf), TAINT(p->ii, ivs * nbuf), bufs + roffset, bufs + ioffset), 0, 0, (p->ri == p->ro) ? NO_DESTROY_INPUT : 0); if (!cld) goto nada; /* copying back from the buffer is a rank-0 transform: */ cldcpy = X(mkplan_d)(plnr, X(mkproblem_dft_d)( X(mktensor_0d)(), X(mktensor_2d)(nbuf, bufdist * 2, ovs, n, 2, p->sz->dims[0].os), bufs + roffset, bufs + ioffset, TAINT(p->ro, ovs * nbuf), TAINT(p->io, ovs * nbuf))); if (!cldcpy) goto nada; /* deallocate buffers, let apply() allocate them for real */ X(ifree)(bufs); bufs = 0; /* plan the leftover transforms (cldrest): */ { INT id = ivs * (nbuf * (vl / nbuf)); INT od = ovs * (nbuf * (vl / nbuf)); cldrest = X(mkplan_d)(plnr, X(mkproblem_dft_d)( X(tensor_copy)(p->sz), X(mktensor_1d)(vl % nbuf, ivs, ovs), p->ri+id, p->ii+id, p->ro+od, p->io+od)); } if (!cldrest) goto nada; pln = MKPLAN_DFT(P, &padt, apply); pln->cld = cld; pln->cldcpy = cldcpy; pln->cldrest = cldrest; pln->n = n; pln->vl = vl; pln->ivs_by_nbuf = ivs * nbuf; pln->ovs_by_nbuf = ovs * nbuf; pln->roffset = roffset; pln->ioffset = ioffset; pln->nbuf = nbuf; pln->bufdist = bufdist; { opcnt t; X(ops_add)(&cld->ops, &cldcpy->ops, &t); X(ops_madd)(vl / nbuf, &t, &cldrest->ops, &pln->super.super.ops); } return &(pln->super.super); nada: X(ifree0)(bufs); X(plan_destroy_internal)(cldrest); X(plan_destroy_internal)(cldcpy); X(plan_destroy_internal)(cld); return (plan *) 0; } static solver *mksolver(int maxnbuf_ndx) { static const solver_adt sadt = { PROBLEM_DFT, mkplan, 0 }; S *slv = MKSOLVER(S, &sadt); slv->maxnbuf_ndx = maxnbuf_ndx; return &(slv->super); } void X(dft_buffered_register)(planner *p) { size_t i; for (i = 0; i < NELEM(maxnbufs); ++i) REGISTER_SOLVER(p, mksolver(i)); } fftw-3.3.3/dft/conf.c0000644000175000017500000000350612054410050011250 00000000000000/* * Copyright (c) 2003, 2007-11 Matteo Frigo * Copyright (c) 2003, 2007-11 Massachusetts Institute of Technology * * This program is free software; you can redistribute it and/or modify * it under the terms of the GNU General Public License as published by * the Free Software Foundation; either version 2 of the License, or * (at your option) any later version. * * This program is distributed in the hope that it will be useful, * but WITHOUT ANY WARRANTY; without even the implied warranty of * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the * GNU General Public License for more details. * * You should have received a copy of the GNU General Public License * along with this program; if not, write to the Free Software * Foundation, Inc., 51 Franklin Street, Fifth Floor, Boston, MA 02110-1301 USA * */ #include "dft.h" static const solvtab s = { SOLVTAB(X(dft_indirect_register)), SOLVTAB(X(dft_indirect_transpose_register)), SOLVTAB(X(dft_rank_geq2_register)), SOLVTAB(X(dft_vrank_geq1_register)), SOLVTAB(X(dft_buffered_register)), SOLVTAB(X(dft_generic_register)), SOLVTAB(X(dft_rader_register)), SOLVTAB(X(dft_bluestein_register)), SOLVTAB(X(dft_nop_register)), SOLVTAB(X(ct_generic_register)), SOLVTAB(X(ct_genericbuf_register)), SOLVTAB_END }; void X(dft_conf_standard)(planner *p) { X(solvtab_exec)(s, p); X(solvtab_exec)(X(solvtab_dft_standard), p); #if HAVE_SSE2 if (X(have_simd_sse2)()) X(solvtab_exec)(X(solvtab_dft_sse2), p); #endif #if HAVE_AVX if (X(have_simd_avx)()) X(solvtab_exec)(X(solvtab_dft_avx), p); #endif #if HAVE_ALTIVEC if (X(have_simd_altivec)()) X(solvtab_exec)(X(solvtab_dft_altivec), p); #endif #if HAVE_NEON if (X(have_simd_neon)()) X(solvtab_exec)(X(solvtab_dft_neon), p); #endif } fftw-3.3.3/dft/ct.c0000644000175000017500000001422612054410050010732 00000000000000/* * Copyright (c) 2003, 2007-11 Matteo Frigo * Copyright (c) 2003, 2007-11 Massachusetts Institute of Technology * * This program is free software; you can redistribute it and/or modify * it under the terms of the GNU General Public License as published by * the Free Software Foundation; either version 2 of the License, or * (at your option) any later version. * * This program is distributed in the hope that it will be useful, * but WITHOUT ANY WARRANTY; without even the implied warranty of * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the * GNU General Public License for more details. * * You should have received a copy of the GNU General Public License * along with this program; if not, write to the Free Software * Foundation, Inc., 51 Franklin Street, Fifth Floor, Boston, MA 02110-1301 USA * */ #include "ct.h" ct_solver *(*X(mksolver_ct_hook))(size_t, INT, int, ct_mkinferior, ct_force_vrecursion) = 0; typedef struct { plan_dft super; plan *cld; plan *cldw; INT r; } P; static void apply_dit(const plan *ego_, R *ri, R *ii, R *ro, R *io) { const P *ego = (const P *) ego_; plan_dft *cld; plan_dftw *cldw; cld = (plan_dft *) ego->cld; cld->apply(ego->cld, ri, ii, ro, io); cldw = (plan_dftw *) ego->cldw; cldw->apply(ego->cldw, ro, io); } static void apply_dif(const plan *ego_, R *ri, R *ii, R *ro, R *io) { const P *ego = (const P *) ego_; plan_dft *cld; plan_dftw *cldw; cldw = (plan_dftw *) ego->cldw; cldw->apply(ego->cldw, ri, ii); cld = (plan_dft *) ego->cld; cld->apply(ego->cld, ri, ii, ro, io); } static void awake(plan *ego_, enum wakefulness wakefulness) { P *ego = (P *) ego_; X(plan_awake)(ego->cld, wakefulness); X(plan_awake)(ego->cldw, wakefulness); } static void destroy(plan *ego_) { P *ego = (P *) ego_; X(plan_destroy_internal)(ego->cldw); X(plan_destroy_internal)(ego->cld); } static void print(const plan *ego_, printer *p) { const P *ego = (const P *) ego_; p->print(p, "(dft-ct-%s/%D%(%p%)%(%p%))", ego->super.apply == apply_dit ? "dit" : "dif", ego->r, ego->cldw, ego->cld); } static int applicable0(const ct_solver *ego, const problem *p_, planner *plnr) { const problem_dft *p = (const problem_dft *) p_; INT r; return (1 && p->sz->rnk == 1 && p->vecsz->rnk <= 1 /* DIF destroys the input and we don't like it */ && (ego->dec == DECDIT || p->ri == p->ro || !NO_DESTROY_INPUTP(plnr)) && ((r = X(choose_radix)(ego->r, p->sz->dims[0].n)) > 1) && p->sz->dims[0].n > r); } int X(ct_applicable)(const ct_solver *ego, const problem *p_, planner *plnr) { const problem_dft *p; if (!applicable0(ego, p_, plnr)) return 0; p = (const problem_dft *) p_; return (0 || ego->dec == DECDIF+TRANSPOSE || p->vecsz->rnk == 0 || !NO_VRECURSEP(plnr) || (ego->force_vrecursionp && ego->force_vrecursionp(ego, p)) ); } static plan *mkplan(const solver *ego_, const problem *p_, planner *plnr) { const ct_solver *ego = (const ct_solver *) ego_; const problem_dft *p; P *pln = 0; plan *cld = 0, *cldw = 0; INT n, r, m, v, ivs, ovs; iodim *d; static const plan_adt padt = { X(dft_solve), awake, print, destroy }; if ((NO_NONTHREADEDP(plnr)) || !X(ct_applicable)(ego, p_, plnr)) return (plan *) 0; p = (const problem_dft *) p_; d = p->sz->dims; n = d[0].n; r = X(choose_radix)(ego->r, n); m = n / r; X(tensor_tornk1)(p->vecsz, &v, &ivs, &ovs); switch (ego->dec) { case DECDIT: { cldw = ego->mkcldw(ego, r, m * d[0].os, m * d[0].os, m, d[0].os, v, ovs, ovs, 0, m, p->ro, p->io, plnr); if (!cldw) goto nada; cld = X(mkplan_d)(plnr, X(mkproblem_dft_d)( X(mktensor_1d)(m, r * d[0].is, d[0].os), X(mktensor_2d)(r, d[0].is, m * d[0].os, v, ivs, ovs), p->ri, p->ii, p->ro, p->io) ); if (!cld) goto nada; pln = MKPLAN_DFT(P, &padt, apply_dit); break; } case DECDIF: case DECDIF+TRANSPOSE: { INT cors, covs; /* cldw ors, ovs */ if (ego->dec == DECDIF+TRANSPOSE) { cors = ivs; covs = m * d[0].is; /* ensure that we generate well-formed dftw subproblems */ /* FIXME: too conservative */ if (!(1 && r == v && d[0].is == r * cors)) goto nada; /* FIXME: allow in-place only for now, like in fftw-3.[01] */ if (!(1 && p->ri == p->ro && d[0].is == r * d[0].os && cors == d[0].os && covs == ovs )) goto nada; } else { cors = m * d[0].is; covs = ivs; } cldw = ego->mkcldw(ego, r, m * d[0].is, cors, m, d[0].is, v, ivs, covs, 0, m, p->ri, p->ii, plnr); if (!cldw) goto nada; cld = X(mkplan_d)(plnr, X(mkproblem_dft_d)( X(mktensor_1d)(m, d[0].is, r * d[0].os), X(mktensor_2d)(r, cors, d[0].os, v, covs, ovs), p->ri, p->ii, p->ro, p->io) ); if (!cld) goto nada; pln = MKPLAN_DFT(P, &padt, apply_dif); break; } default: A(0); } pln->cld = cld; pln->cldw = cldw; pln->r = r; X(ops_add)(&cld->ops, &cldw->ops, &pln->super.super.ops); /* inherit could_prune_now_p attribute from cldw */ pln->super.super.could_prune_now_p = cldw->could_prune_now_p; return &(pln->super.super); nada: X(plan_destroy_internal)(cldw); X(plan_destroy_internal)(cld); return (plan *) 0; } ct_solver *X(mksolver_ct)(size_t size, INT r, int dec, ct_mkinferior mkcldw, ct_force_vrecursion force_vrecursionp) { static const solver_adt sadt = { PROBLEM_DFT, mkplan, 0 }; ct_solver *slv = (ct_solver *)X(mksolver)(size, &sadt); slv->r = r; slv->dec = dec; slv->mkcldw = mkcldw; slv->force_vrecursionp = force_vrecursionp; return slv; } plan *X(mkplan_dftw)(size_t size, const plan_adt *adt, dftwapply apply) { plan_dftw *ego; ego = (plan_dftw *) X(mkplan)(size, adt); ego->apply = apply; return &(ego->super); } fftw-3.3.3/dft/dftw-direct.c0000644000175000017500000002172612054410050012543 00000000000000/* * Copyright (c) 2003, 2007-11 Matteo Frigo * Copyright (c) 2003, 2007-11 Massachusetts Institute of Technology * * This program is free software; you can redistribute it and/or modify * it under the terms of the GNU General Public License as published by * the Free Software Foundation; either version 2 of the License, or * (at your option) any later version. * * This program is distributed in the hope that it will be useful, * but WITHOUT ANY WARRANTY; without even the implied warranty of * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the * GNU General Public License for more details. * * You should have received a copy of the GNU General Public License * along with this program; if not, write to the Free Software * Foundation, Inc., 51 Franklin Street, Fifth Floor, Boston, MA 02110-1301 USA * */ #include "ct.h" typedef struct { ct_solver super; const ct_desc *desc; int bufferedp; kdftw k; } S; typedef struct { plan_dftw super; kdftw k; INT r; stride rs; INT m, ms, v, vs, mb, me, extra_iter; stride brs; twid *td; const S *slv; } P; /************************************************************* Nonbuffered code *************************************************************/ static void apply(const plan *ego_, R *rio, R *iio) { const P *ego = (const P *) ego_; INT i; ASSERT_ALIGNED_DOUBLE; for (i = 0; i < ego->v; ++i, rio += ego->vs, iio += ego->vs) { INT mb = ego->mb, ms = ego->ms; ego->k(rio + mb*ms, iio + mb*ms, ego->td->W, ego->rs, mb, ego->me, ms); } } static void apply_extra_iter(const plan *ego_, R *rio, R *iio) { const P *ego = (const P *) ego_; INT i, v = ego->v, vs = ego->vs; INT mb = ego->mb, me = ego->me, mm = me - 1, ms = ego->ms; ASSERT_ALIGNED_DOUBLE; for (i = 0; i < v; ++i, rio += vs, iio += vs) { ego->k(rio + mb*ms, iio + mb*ms, ego->td->W, ego->rs, mb, mm, ms); ego->k(rio + mm*ms, iio + mm*ms, ego->td->W, ego->rs, mm, mm+2, 0); } } /************************************************************* Buffered code *************************************************************/ static void dobatch(const P *ego, R *rA, R *iA, INT mb, INT me, R *buf) { INT brs = WS(ego->brs, 1); INT rs = WS(ego->rs, 1); INT ms = ego->ms; X(cpy2d_pair_ci)(rA + mb*ms, iA + mb*ms, buf, buf + 1, ego->r, rs, brs, me - mb, ms, 2); ego->k(buf, buf + 1, ego->td->W, ego->brs, mb, me, 2); X(cpy2d_pair_co)(buf, buf + 1, rA + mb*ms, iA + mb*ms, ego->r, brs, rs, me - mb, 2, ms); } /* must be even for SIMD alignment; should not be 2^k to avoid associativity conflicts */ static INT compute_batchsize(INT radix) { /* round up to multiple of 4 */ radix += 3; radix &= -4; return (radix + 2); } static void apply_buf(const plan *ego_, R *rio, R *iio) { const P *ego = (const P *) ego_; INT i, j, v = ego->v, r = ego->r; INT batchsz = compute_batchsize(r); R *buf; INT mb = ego->mb, me = ego->me; size_t bufsz = r * batchsz * 2 * sizeof(R); BUF_ALLOC(R *, buf, bufsz); for (i = 0; i < v; ++i, rio += ego->vs, iio += ego->vs) { for (j = mb; j + batchsz < me; j += batchsz) dobatch(ego, rio, iio, j, j + batchsz, buf); dobatch(ego, rio, iio, j, me, buf); } BUF_FREE(buf, bufsz); } /************************************************************* common code *************************************************************/ static void awake(plan *ego_, enum wakefulness wakefulness) { P *ego = (P *) ego_; X(twiddle_awake)(wakefulness, &ego->td, ego->slv->desc->tw, ego->r * ego->m, ego->r, ego->m + ego->extra_iter); } static void destroy(plan *ego_) { P *ego = (P *) ego_; X(stride_destroy)(ego->brs); X(stride_destroy)(ego->rs); } static void print(const plan *ego_, printer *p) { const P *ego = (const P *) ego_; const S *slv = ego->slv; const ct_desc *e = slv->desc; if (slv->bufferedp) p->print(p, "(dftw-directbuf/%D-%D/%D%v \"%s\")", compute_batchsize(ego->r), ego->r, X(twiddle_length)(ego->r, e->tw), ego->v, e->nam); else p->print(p, "(dftw-direct-%D/%D%v \"%s\")", ego->r, X(twiddle_length)(ego->r, e->tw), ego->v, e->nam); } static int applicable0(const S *ego, INT r, INT irs, INT ors, INT m, INT ms, INT v, INT ivs, INT ovs, INT mb, INT me, R *rio, R *iio, const planner *plnr, INT *extra_iter) { const ct_desc *e = ego->desc; UNUSED(v); return ( 1 && r == e->radix && irs == ors /* in-place along R */ && ivs == ovs /* in-place along V */ /* check for alignment/vector length restrictions */ && ((*extra_iter = 0, e->genus->okp(e, rio, iio, irs, ivs, m, mb, me, ms, plnr)) || (*extra_iter = 1, (1 /* FIXME: require full array, otherwise some threads may be extra_iter and other threads won't be. Generating the proper twiddle factors is a pain in this case */ && mb == 0 && me == m && e->genus->okp(e, rio, iio, irs, ivs, m, mb, me - 1, ms, plnr) && e->genus->okp(e, rio, iio, irs, ivs, m, me - 1, me + 1, ms, plnr)))) && (e->genus->okp(e, rio + ivs, iio + ivs, irs, ivs, m, mb, me - *extra_iter, ms, plnr)) ); } static int applicable0_buf(const S *ego, INT r, INT irs, INT ors, INT m, INT ms, INT v, INT ivs, INT ovs, INT mb, INT me, R *rio, R *iio, const planner *plnr) { const ct_desc *e = ego->desc; INT batchsz; UNUSED(v); UNUSED(ms); UNUSED(rio); UNUSED(iio); return ( 1 && r == e->radix && irs == ors /* in-place along R */ && ivs == ovs /* in-place along V */ /* check for alignment/vector length restrictions, both for batchsize and for the remainder */ && (batchsz = compute_batchsize(r), 1) && (e->genus->okp(e, 0, ((const R *)0) + 1, 2 * batchsz, 0, m, mb, mb + batchsz, 2, plnr)) && (e->genus->okp(e, 0, ((const R *)0) + 1, 2 * batchsz, 0, m, mb, me, 2, plnr)) ); } static int applicable(const S *ego, INT r, INT irs, INT ors, INT m, INT ms, INT v, INT ivs, INT ovs, INT mb, INT me, R *rio, R *iio, const planner *plnr, INT *extra_iter) { if (ego->bufferedp) { *extra_iter = 0; if (!applicable0_buf(ego, r, irs, ors, m, ms, v, ivs, ovs, mb, me, rio, iio, plnr)) return 0; } else { if (!applicable0(ego, r, irs, ors, m, ms, v, ivs, ovs, mb, me, rio, iio, plnr, extra_iter)) return 0; } if (NO_UGLYP(plnr) && X(ct_uglyp)((ego->bufferedp? (INT)512 : (INT)16), v, m * r, r)) return 0; if (m * r > 262144 && NO_FIXED_RADIX_LARGE_NP(plnr)) return 0; return 1; } static plan *mkcldw(const ct_solver *ego_, INT r, INT irs, INT ors, INT m, INT ms, INT v, INT ivs, INT ovs, INT mstart, INT mcount, R *rio, R *iio, planner *plnr) { const S *ego = (const S *) ego_; P *pln; const ct_desc *e = ego->desc; INT extra_iter; static const plan_adt padt = { 0, awake, print, destroy }; A(mstart >= 0 && mstart + mcount <= m); if (!applicable(ego, r, irs, ors, m, ms, v, ivs, ovs, mstart, mstart + mcount, rio, iio, plnr, &extra_iter)) return (plan *)0; if (ego->bufferedp) { pln = MKPLAN_DFTW(P, &padt, apply_buf); } else { pln = MKPLAN_DFTW(P, &padt, extra_iter ? apply_extra_iter : apply); } pln->k = ego->k; pln->rs = X(mkstride)(r, irs); pln->td = 0; pln->r = r; pln->m = m; pln->ms = ms; pln->v = v; pln->vs = ivs; pln->mb = mstart; pln->me = mstart + mcount; pln->slv = ego; pln->brs = X(mkstride)(r, 2 * compute_batchsize(r)); pln->extra_iter = extra_iter; X(ops_zero)(&pln->super.super.ops); X(ops_madd2)(v * (mcount/e->genus->vl), &e->ops, &pln->super.super.ops); if (ego->bufferedp) { /* 8 load/stores * N * V */ pln->super.super.ops.other += 8 * r * mcount * v; } pln->super.super.could_prune_now_p = (!ego->bufferedp && r >= 5 && r < 64 && m >= r); return &(pln->super.super); } static void regone(planner *plnr, kdftw codelet, const ct_desc *desc, int dec, int bufferedp) { S *slv = (S *)X(mksolver_ct)(sizeof(S), desc->radix, dec, mkcldw, 0); slv->k = codelet; slv->desc = desc; slv->bufferedp = bufferedp; REGISTER_SOLVER(plnr, &(slv->super.super)); if (X(mksolver_ct_hook)) { slv = (S *)X(mksolver_ct_hook)(sizeof(S), desc->radix, dec, mkcldw, 0); slv->k = codelet; slv->desc = desc; slv->bufferedp = bufferedp; REGISTER_SOLVER(plnr, &(slv->super.super)); } } void X(regsolver_ct_directw)(planner *plnr, kdftw codelet, const ct_desc *desc, int dec) { regone(plnr, codelet, desc, dec, /* bufferedp */ 0); regone(plnr, codelet, desc, dec, /* bufferedp */ 1); } fftw-3.3.3/dft/dftw-directsq.c0000644000175000017500000000772712054410050013114 00000000000000/* * Copyright (c) 2003, 2007-11 Matteo Frigo * Copyright (c) 2003, 2007-11 Massachusetts Institute of Technology * * This program is free software; you can redistribute it and/or modify * it under the terms of the GNU General Public License as published by * the Free Software Foundation; either version 2 of the License, or * (at your option) any later version. * * This program is distributed in the hope that it will be useful, * but WITHOUT ANY WARRANTY; without even the implied warranty of * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the * GNU General Public License for more details. * * You should have received a copy of the GNU General Public License * along with this program; if not, write to the Free Software * Foundation, Inc., 51 Franklin Street, Fifth Floor, Boston, MA 02110-1301 USA * */ #include "ct.h" typedef struct { ct_solver super; const ct_desc *desc; kdftwsq k; } S; typedef struct { plan_dftw super; kdftwsq k; INT r; stride rs, vs; INT m, ms, v, mb, me; twid *td; const S *slv; } P; static void apply(const plan *ego_, R *rio, R *iio) { const P *ego = (const P *) ego_; INT mb = ego->mb, ms = ego->ms; ego->k(rio + mb*ms, iio + mb*ms, ego->td->W, ego->rs, ego->vs, mb, ego->me, ms); } static void awake(plan *ego_, enum wakefulness wakefulness) { P *ego = (P *) ego_; X(twiddle_awake)(wakefulness, &ego->td, ego->slv->desc->tw, ego->r * ego->m, ego->r, ego->m); } static void destroy(plan *ego_) { P *ego = (P *) ego_; X(stride_destroy)(ego->rs); X(stride_destroy)(ego->vs); } static void print(const plan *ego_, printer *p) { const P *ego = (const P *) ego_; const S *slv = ego->slv; const ct_desc *e = slv->desc; p->print(p, "(dftw-directsq-%D/%D%v \"%s\")", ego->r, X(twiddle_length)(ego->r, e->tw), ego->v, e->nam); } static int applicable(const S *ego, INT r, INT irs, INT ors, INT m, INT ms, INT v, INT ivs, INT ovs, INT mb, INT me, R *rio, R *iio, const planner *plnr) { const ct_desc *e = ego->desc; UNUSED(v); return ( 1 && r == e->radix /* transpose r, v */ && r == v && irs == ovs && ivs == ors /* check for alignment/vector length restrictions */ && e->genus->okp(e, rio, iio, irs, ivs, m, mb, me, ms, plnr) ); } static plan *mkcldw(const ct_solver *ego_, INT r, INT irs, INT ors, INT m, INT ms, INT v, INT ivs, INT ovs, INT mstart, INT mcount, R *rio, R *iio, planner *plnr) { const S *ego = (const S *) ego_; P *pln; const ct_desc *e = ego->desc; static const plan_adt padt = { 0, awake, print, destroy }; A(mstart >= 0 && mstart + mcount <= m); if (!applicable(ego, r, irs, ors, m, ms, v, ivs, ovs, mstart, mstart + mcount, rio, iio, plnr)) return (plan *)0; pln = MKPLAN_DFTW(P, &padt, apply); pln->k = ego->k; pln->rs = X(mkstride)(r, irs); pln->vs = X(mkstride)(v, ivs); pln->td = 0; pln->r = r; pln->m = m; pln->ms = ms; pln->v = v; pln->mb = mstart; pln->me = mstart + mcount; pln->slv = ego; X(ops_zero)(&pln->super.super.ops); X(ops_madd2)(mcount/e->genus->vl, &e->ops, &pln->super.super.ops); return &(pln->super.super); } static void regone(planner *plnr, kdftwsq codelet, const ct_desc *desc, int dec) { S *slv = (S *)X(mksolver_ct)(sizeof(S), desc->radix, dec, mkcldw, 0); slv->k = codelet; slv->desc = desc; REGISTER_SOLVER(plnr, &(slv->super.super)); if (X(mksolver_ct_hook)) { slv = (S *)X(mksolver_ct_hook)(sizeof(S), desc->radix, dec, mkcldw, 0); slv->k = codelet; slv->desc = desc; REGISTER_SOLVER(plnr, &(slv->super.super)); } } void X(regsolver_ct_directwsq)(planner *plnr, kdftwsq codelet, const ct_desc *desc, int dec) { regone(plnr, codelet, desc, dec+TRANSPOSE); } fftw-3.3.3/dft/dftw-generic.c0000644000175000017500000001211612054410050012676 00000000000000/* * Copyright (c) 2003, 2007-11 Matteo Frigo * Copyright (c) 2003, 2007-11 Massachusetts Institute of Technology * * This program is free software; you can redistribute it and/or modify * it under the terms of the GNU General Public License as published by * the Free Software Foundation; either version 2 of the License, or * (at your option) any later version. * * This program is distributed in the hope that it will be useful, * but WITHOUT ANY WARRANTY; without even the implied warranty of * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the * GNU General Public License for more details. * * You should have received a copy of the GNU General Public License * along with this program; if not, write to the Free Software * Foundation, Inc., 51 Franklin Street, Fifth Floor, Boston, MA 02110-1301 USA * */ /* express a twiddle problem in terms of dft + multiplication by twiddle factors */ #include "ct.h" typedef ct_solver S; typedef struct { plan_dftw super; INT r, rs, m, mb, me, ms, v, vs; plan *cld; twid *td; const S *slv; int dec; } P; static void mktwiddle(P *ego, enum wakefulness wakefulness) { static const tw_instr tw[] = { { TW_FULL, 0, 0 }, { TW_NEXT, 1, 0 } }; /* note that R and M are swapped, to allow for sequential access both to data and twiddles */ X(twiddle_awake)(wakefulness, &ego->td, tw, ego->r * ego->m, ego->m, ego->r); } static void bytwiddle(const P *ego, R *rio, R *iio) { INT iv, ir, im; INT r = ego->r, rs = ego->rs; INT m = ego->m, mb = ego->mb, me = ego->me, ms = ego->ms; INT v = ego->v, vs = ego->vs; const R *W = ego->td->W; mb += (mb == 0); /* skip m=0 iteration */ for (iv = 0; iv < v; ++iv) { for (ir = 1; ir < r; ++ir) { for (im = mb; im < me; ++im) { R *pr = rio + ms * im + rs * ir; R *pi = iio + ms * im + rs * ir; E xr = *pr; E xi = *pi; E wr = W[2 * im + (2 * (m-1)) * ir - 2]; E wi = W[2 * im + (2 * (m-1)) * ir - 1]; *pr = xr * wr + xi * wi; *pi = xi * wr - xr * wi; } } rio += vs; iio += vs; } } static int applicable(INT irs, INT ors, INT ivs, INT ovs, const planner *plnr) { return (1 && irs == ors && ivs == ovs && !NO_SLOWP(plnr) ); } static void apply_dit(const plan *ego_, R *rio, R *iio) { const P *ego = (const P *) ego_; plan_dft *cld; INT dm = ego->ms * ego->mb; bytwiddle(ego, rio, iio); cld = (plan_dft *) ego->cld; cld->apply(ego->cld, rio + dm, iio + dm, rio + dm, iio + dm); } static void apply_dif(const plan *ego_, R *rio, R *iio) { const P *ego = (const P *) ego_; plan_dft *cld; INT dm = ego->ms * ego->mb; cld = (plan_dft *) ego->cld; cld->apply(ego->cld, rio + dm, iio + dm, rio + dm, iio + dm); bytwiddle(ego, rio, iio); } static void awake(plan *ego_, enum wakefulness wakefulness) { P *ego = (P *) ego_; X(plan_awake)(ego->cld, wakefulness); mktwiddle(ego, wakefulness); } static void destroy(plan *ego_) { P *ego = (P *) ego_; X(plan_destroy_internal)(ego->cld); } static void print(const plan *ego_, printer *p) { const P *ego = (const P *) ego_; p->print(p, "(dftw-generic-%s-%D-%D%v%(%p%))", ego->dec == DECDIT ? "dit" : "dif", ego->r, ego->m, ego->v, ego->cld); } static plan *mkcldw(const ct_solver *ego_, INT r, INT irs, INT ors, INT m, INT ms, INT v, INT ivs, INT ovs, INT mstart, INT mcount, R *rio, R *iio, planner *plnr) { const S *ego = (const S *)ego_; P *pln; plan *cld = 0; INT dm = ms * mstart; static const plan_adt padt = { 0, awake, print, destroy }; A(mstart >= 0 && mstart + mcount <= m); if (!applicable(irs, ors, ivs, ovs, plnr)) return (plan *)0; cld = X(mkplan_d)(plnr, X(mkproblem_dft_d)( X(mktensor_1d)(r, irs, irs), X(mktensor_2d)(mcount, ms, ms, v, ivs, ivs), rio + dm, iio + dm, rio + dm, iio + dm) ); if (!cld) goto nada; pln = MKPLAN_DFTW(P, &padt, ego->dec == DECDIT ? apply_dit : apply_dif); pln->slv = ego; pln->cld = cld; pln->r = r; pln->rs = irs; pln->m = m; pln->ms = ms; pln->v = v; pln->vs = ivs; pln->mb = mstart; pln->me = mstart + mcount; pln->dec = ego->dec; pln->td = 0; { double n0 = (r - 1) * (mcount - 1) * v; pln->super.super.ops = cld->ops; pln->super.super.ops.mul += 8 * n0; pln->super.super.ops.add += 4 * n0; pln->super.super.ops.other += 8 * n0; } return &(pln->super.super); nada: X(plan_destroy_internal)(cld); return (plan *) 0; } static void regsolver(planner *plnr, INT r, int dec) { S *slv = (S *)X(mksolver_ct)(sizeof(S), r, dec, mkcldw, 0); REGISTER_SOLVER(plnr, &(slv->super)); if (X(mksolver_ct_hook)) { slv = (S *)X(mksolver_ct_hook)(sizeof(S), r, dec, mkcldw, 0); REGISTER_SOLVER(plnr, &(slv->super)); } } void X(ct_generic_register)(planner *p) { regsolver(p, 0, DECDIT); regsolver(p, 0, DECDIF); } fftw-3.3.3/dft/dftw-genericbuf.c0000644000175000017500000001327712054410050013404 00000000000000/* * Copyright (c) 2003, 2007-11 Matteo Frigo * Copyright (c) 2003, 2007-11 Massachusetts Institute of Technology * * This program is free software; you can redistribute it and/or modify * it under the terms of the GNU General Public License as published by * the Free Software Foundation; either version 2 of the License, or * (at your option) any later version. * * This program is distributed in the hope that it will be useful, * but WITHOUT ANY WARRANTY; without even the implied warranty of * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the * GNU General Public License for more details. * * You should have received a copy of the GNU General Public License * along with this program; if not, write to the Free Software * Foundation, Inc., 51 Franklin Street, Fifth Floor, Boston, MA 02110-1301 USA * */ /* express a twiddle problem in terms of dft + multiplication by twiddle factors */ #include "ct.h" typedef struct { ct_solver super; INT batchsz; } S; typedef struct { plan_dftw super; INT r, rs, m, ms, v, vs, mb, me; INT batchsz; plan *cld; triggen *t; const S *slv; } P; #define BATCHDIST(r) ((r) + 16) /**************************************************************/ static void bytwiddle(const P *ego, INT mb, INT me, R *buf, R *rio, R *iio) { INT j, k; INT r = ego->r, rs = ego->rs, ms = ego->ms; triggen *t = ego->t; for (j = 0; j < r; ++j) { for (k = mb; k < me; ++k) t->rotate(t, j * k, rio[j * rs + k * ms], iio[j * rs + k * ms], &buf[j * 2 + 2 * BATCHDIST(r) * (k - mb) + 0]); } } static int applicable0(const S *ego, INT r, INT irs, INT ors, INT m, INT v, INT mcount) { return (1 && v == 1 && irs == ors && mcount >= ego->batchsz && mcount % ego->batchsz == 0 && r >= 64 && m >= r ); } static int applicable(const S *ego, INT r, INT irs, INT ors, INT m, INT v, INT mcount, const planner *plnr) { if (!applicable0(ego, r, irs, ors, m, v, mcount)) return 0; if (NO_UGLYP(plnr) && m * r < 65536) return 0; return 1; } static void dobatch(const P *ego, INT mb, INT me, R *buf, R *rio, R *iio) { plan_dft *cld; INT ms = ego->ms; bytwiddle(ego, mb, me, buf, rio, iio); cld = (plan_dft *) ego->cld; cld->apply(ego->cld, buf, buf + 1, buf, buf + 1); X(cpy2d_pair_co)(buf, buf + 1, rio + ms * mb, iio + ms * mb, me-mb, 2 * BATCHDIST(ego->r), ms, ego->r, 2, ego->rs); } static void apply(const plan *ego_, R *rio, R *iio) { const P *ego = (const P *) ego_; R *buf = (R *) MALLOC(sizeof(R) * 2 * BATCHDIST(ego->r) * ego->batchsz, BUFFERS); INT m; for (m = ego->mb; m < ego->me; m += ego->batchsz) dobatch(ego, m, m + ego->batchsz, buf, rio, iio); A(m == ego->me); X(ifree)(buf); } static void awake(plan *ego_, enum wakefulness wakefulness) { P *ego = (P *) ego_; X(plan_awake)(ego->cld, wakefulness); switch (wakefulness) { case SLEEPY: X(triggen_destroy)(ego->t); ego->t = 0; break; default: ego->t = X(mktriggen)(AWAKE_SQRTN_TABLE, ego->r * ego->m); break; } } static void destroy(plan *ego_) { P *ego = (P *) ego_; X(plan_destroy_internal)(ego->cld); } static void print(const plan *ego_, printer *p) { const P *ego = (const P *) ego_; p->print(p, "(dftw-genericbuf/%D-%D-%D%(%p%))", ego->batchsz, ego->r, ego->m, ego->cld); } static plan *mkcldw(const ct_solver *ego_, INT r, INT irs, INT ors, INT m, INT ms, INT v, INT ivs, INT ovs, INT mstart, INT mcount, R *rio, R *iio, planner *plnr) { const S *ego = (const S *)ego_; P *pln; plan *cld = 0; R *buf; static const plan_adt padt = { 0, awake, print, destroy }; UNUSED(ivs); UNUSED(ovs); UNUSED(rio); UNUSED(iio); A(mstart >= 0 && mstart + mcount <= m); if (!applicable(ego, r, irs, ors, m, v, mcount, plnr)) return (plan *)0; buf = (R *) MALLOC(sizeof(R) * 2 * BATCHDIST(r) * ego->batchsz, BUFFERS); cld = X(mkplan_d)(plnr, X(mkproblem_dft_d)( X(mktensor_1d)(r, 2, 2), X(mktensor_1d)(ego->batchsz, 2 * BATCHDIST(r), 2 * BATCHDIST(r)), buf, buf + 1, buf, buf + 1 ) ); X(ifree)(buf); if (!cld) goto nada; pln = MKPLAN_DFTW(P, &padt, apply); pln->slv = ego; pln->cld = cld; pln->r = r; pln->m = m; pln->ms = ms; pln->rs = irs; pln->batchsz = ego->batchsz; pln->mb = mstart; pln->me = mstart + mcount; { double n0 = (r - 1) * (mcount - 1); pln->super.super.ops = cld->ops; pln->super.super.ops.mul += 8 * n0; pln->super.super.ops.add += 4 * n0; pln->super.super.ops.other += 8 * n0; } return &(pln->super.super); nada: X(plan_destroy_internal)(cld); return (plan *) 0; } static void regsolver(planner *plnr, INT r, INT batchsz) { S *slv = (S *)X(mksolver_ct)(sizeof(S), r, DECDIT, mkcldw, 0); slv->batchsz = batchsz; REGISTER_SOLVER(plnr, &(slv->super.super)); if (X(mksolver_ct_hook)) { slv = (S *)X(mksolver_ct_hook)(sizeof(S), r, DECDIT, mkcldw, 0); slv->batchsz = batchsz; REGISTER_SOLVER(plnr, &(slv->super.super)); } } void X(ct_genericbuf_register)(planner *p) { static const INT radices[] = { -1, -2, -4, -8, -16, -32, -64 }; static const INT batchsizes[] = { 4, 8, 16, 32, 64 }; unsigned i, j; for (i = 0; i < sizeof(radices) / sizeof(radices[0]); ++i) for (j = 0; j < sizeof(batchsizes) / sizeof(batchsizes[0]); ++j) regsolver(p, radices[i], batchsizes[j]); } fftw-3.3.3/dft/direct.c0000644000175000017500000001746712054410050011610 00000000000000/* * Copyright (c) 2003, 2007-11 Matteo Frigo * Copyright (c) 2003, 2007-11 Massachusetts Institute of Technology * * This program is free software; you can redistribute it and/or modify * it under the terms of the GNU General Public License as published by * the Free Software Foundation; either version 2 of the License, or * (at your option) any later version. * * This program is distributed in the hope that it will be useful, * but WITHOUT ANY WARRANTY; without even the implied warranty of * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the * GNU General Public License for more details. * * You should have received a copy of the GNU General Public License * along with this program; if not, write to the Free Software * Foundation, Inc., 51 Franklin Street, Fifth Floor, Boston, MA 02110-1301 USA * */ /* direct DFT solver, if we have a codelet */ #include "dft.h" typedef struct { solver super; const kdft_desc *desc; kdft k; int bufferedp; } S; typedef struct { plan_dft super; stride is, os, bufstride; INT n, vl, ivs, ovs; kdft k; const S *slv; } P; static void dobatch(const P *ego, R *ri, R *ii, R *ro, R *io, R *buf, INT batchsz) { X(cpy2d_pair_ci)(ri, ii, buf, buf+1, ego->n, WS(ego->is, 1), WS(ego->bufstride, 1), batchsz, ego->ivs, 2); if (IABS(WS(ego->os, 1)) < IABS(ego->ovs)) { /* transform directly to output */ ego->k(buf, buf+1, ro, io, ego->bufstride, ego->os, batchsz, 2, ego->ovs); } else { /* transform to buffer and copy back */ ego->k(buf, buf+1, buf, buf+1, ego->bufstride, ego->bufstride, batchsz, 2, 2); X(cpy2d_pair_co)(buf, buf+1, ro, io, ego->n, WS(ego->bufstride, 1), WS(ego->os, 1), batchsz, 2, ego->ovs); } } static INT compute_batchsize(INT n) { /* round up to multiple of 4 */ n += 3; n &= -4; return (n + 2); } static void apply_buf(const plan *ego_, R *ri, R *ii, R *ro, R *io) { const P *ego = (const P *) ego_; R *buf; INT vl = ego->vl, n = ego->n, batchsz = compute_batchsize(n); INT i; size_t bufsz = n * batchsz * 2 * sizeof(R); BUF_ALLOC(R *, buf, bufsz); for (i = 0; i < vl - batchsz; i += batchsz) { dobatch(ego, ri, ii, ro, io, buf, batchsz); ri += batchsz * ego->ivs; ii += batchsz * ego->ivs; ro += batchsz * ego->ovs; io += batchsz * ego->ovs; } dobatch(ego, ri, ii, ro, io, buf, vl - i); BUF_FREE(buf, bufsz); } static void apply(const plan *ego_, R *ri, R *ii, R *ro, R *io) { const P *ego = (const P *) ego_; ASSERT_ALIGNED_DOUBLE; ego->k(ri, ii, ro, io, ego->is, ego->os, ego->vl, ego->ivs, ego->ovs); } static void apply_extra_iter(const plan *ego_, R *ri, R *ii, R *ro, R *io) { const P *ego = (const P *) ego_; INT vl = ego->vl; ASSERT_ALIGNED_DOUBLE; /* for 4-way SIMD when VL is odd: iterate over an even vector length VL, and then execute the last iteration as a 2-vector with vector stride 0. */ ego->k(ri, ii, ro, io, ego->is, ego->os, vl - 1, ego->ivs, ego->ovs); ego->k(ri + (vl - 1) * ego->ivs, ii + (vl - 1) * ego->ivs, ro + (vl - 1) * ego->ovs, io + (vl - 1) * ego->ovs, ego->is, ego->os, 1, 0, 0); } static void destroy(plan *ego_) { P *ego = (P *) ego_; X(stride_destroy)(ego->is); X(stride_destroy)(ego->os); X(stride_destroy)(ego->bufstride); } static void print(const plan *ego_, printer *p) { const P *ego = (const P *) ego_; const S *s = ego->slv; const kdft_desc *d = s->desc; if (ego->slv->bufferedp) p->print(p, "(dft-directbuf/%D-%D%v \"%s\")", compute_batchsize(d->sz), d->sz, ego->vl, d->nam); else p->print(p, "(dft-direct-%D%v \"%s\")", d->sz, ego->vl, d->nam); } static int applicable_buf(const solver *ego_, const problem *p_, const planner *plnr) { const S *ego = (const S *) ego_; const problem_dft *p = (const problem_dft *) p_; const kdft_desc *d = ego->desc; INT vl; INT ivs, ovs; INT batchsz; return ( 1 && p->sz->rnk == 1 && p->vecsz->rnk == 1 && p->sz->dims[0].n == d->sz /* check strides etc */ && X(tensor_tornk1)(p->vecsz, &vl, &ivs, &ovs) /* UGLY if IS <= IVS */ && !(NO_UGLYP(plnr) && X(iabs)(p->sz->dims[0].is) <= X(iabs)(ivs)) && (batchsz = compute_batchsize(d->sz), 1) && (d->genus->okp(d, 0, ((const R *)0) + 1, p->ro, p->io, 2 * batchsz, p->sz->dims[0].os, batchsz, 2, ovs, plnr)) && (d->genus->okp(d, 0, ((const R *)0) + 1, p->ro, p->io, 2 * batchsz, p->sz->dims[0].os, vl % batchsz, 2, ovs, plnr)) && (0 /* can operate out-of-place */ || p->ri != p->ro /* can operate in-place as long as strides are the same */ || X(tensor_inplace_strides2)(p->sz, p->vecsz) /* can do it if the problem fits in the buffer, no matter what the strides are */ || vl <= batchsz ) ); } static int applicable(const solver *ego_, const problem *p_, const planner *plnr, int *extra_iterp) { const S *ego = (const S *) ego_; const problem_dft *p = (const problem_dft *) p_; const kdft_desc *d = ego->desc; INT vl; INT ivs, ovs; return ( 1 && p->sz->rnk == 1 && p->vecsz->rnk <= 1 && p->sz->dims[0].n == d->sz /* check strides etc */ && X(tensor_tornk1)(p->vecsz, &vl, &ivs, &ovs) && ((*extra_iterp = 0, (d->genus->okp(d, p->ri, p->ii, p->ro, p->io, p->sz->dims[0].is, p->sz->dims[0].os, vl, ivs, ovs, plnr))) || (*extra_iterp = 1, ((d->genus->okp(d, p->ri, p->ii, p->ro, p->io, p->sz->dims[0].is, p->sz->dims[0].os, vl - 1, ivs, ovs, plnr)) && (d->genus->okp(d, p->ri, p->ii, p->ro, p->io, p->sz->dims[0].is, p->sz->dims[0].os, 2, 0, 0, plnr))))) && (0 /* can operate out-of-place */ || p->ri != p->ro /* can always compute one transform */ || vl == 1 /* can operate in-place as long as strides are the same */ || X(tensor_inplace_strides2)(p->sz, p->vecsz) ) ); } static plan *mkplan(const solver *ego_, const problem *p_, planner *plnr) { const S *ego = (const S *) ego_; P *pln; const problem_dft *p; iodim *d; const kdft_desc *e = ego->desc; static const plan_adt padt = { X(dft_solve), X(null_awake), print, destroy }; UNUSED(plnr); if (ego->bufferedp) { if (!applicable_buf(ego_, p_, plnr)) return (plan *)0; pln = MKPLAN_DFT(P, &padt, apply_buf); } else { int extra_iterp = 0; if (!applicable(ego_, p_, plnr, &extra_iterp)) return (plan *)0; pln = MKPLAN_DFT(P, &padt, extra_iterp ? apply_extra_iter : apply); } p = (const problem_dft *) p_; d = p->sz->dims; pln->k = ego->k; pln->n = d[0].n; pln->is = X(mkstride)(pln->n, d[0].is); pln->os = X(mkstride)(pln->n, d[0].os); pln->bufstride = X(mkstride)(pln->n, 2 * compute_batchsize(pln->n)); X(tensor_tornk1)(p->vecsz, &pln->vl, &pln->ivs, &pln->ovs); pln->slv = ego; X(ops_zero)(&pln->super.super.ops); X(ops_madd2)(pln->vl / e->genus->vl, &e->ops, &pln->super.super.ops); if (ego->bufferedp) pln->super.super.ops.other += 4 * pln->n * pln->vl; pln->super.super.could_prune_now_p = !ego->bufferedp; return &(pln->super.super); } static solver *mksolver(kdft k, const kdft_desc *desc, int bufferedp) { static const solver_adt sadt = { PROBLEM_DFT, mkplan, 0 }; S *slv = MKSOLVER(S, &sadt); slv->k = k; slv->desc = desc; slv->bufferedp = bufferedp; return &(slv->super); } solver *X(mksolver_dft_direct)(kdft k, const kdft_desc *desc) { return mksolver(k, desc, 0); } solver *X(mksolver_dft_directbuf)(kdft k, const kdft_desc *desc) { return mksolver(k, desc, 1); } fftw-3.3.3/dft/generic.c0000644000175000017500000001026612054410050011740 00000000000000/* * Copyright (c) 2003, 2007-11 Matteo Frigo * Copyright (c) 2003, 2007-11 Massachusetts Institute of Technology * * This program is free software; you can redistribute it and/or modify * it under the terms of the GNU General Public License as published by * the Free Software Foundation; either version 2 of the License, or * (at your option) any later version. * * This program is distributed in the hope that it will be useful, * but WITHOUT ANY WARRANTY; without even the implied warranty of * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the * GNU General Public License for more details. * * You should have received a copy of the GNU General Public License * along with this program; if not, write to the Free Software * Foundation, Inc., 51 Franklin Street, Fifth Floor, Boston, MA 02110-1301 USA * */ #include "dft.h" typedef struct { solver super; } S; typedef struct { plan_dft super; twid *td; INT n, is, os; } P; static void cdot(INT n, const E *x, const R *w, R *or0, R *oi0, R *or1, R *oi1) { INT i; E rr = x[0], ri = 0, ir = x[1], ii = 0; x += 2; for (i = 1; i + i < n; ++i) { rr += x[0] * w[0]; ir += x[1] * w[0]; ri += x[2] * w[1]; ii += x[3] * w[1]; x += 4; w += 2; } *or0 = rr + ii; *oi0 = ir - ri; *or1 = rr - ii; *oi1 = ir + ri; } static void hartley(INT n, const R *xr, const R *xi, INT xs, E *o, R *pr, R *pi) { INT i; E sr, si; o[0] = sr = xr[0]; o[1] = si = xi[0]; o += 2; for (i = 1; i + i < n; ++i) { sr += (o[0] = xr[i * xs] + xr[(n - i) * xs]); si += (o[1] = xi[i * xs] + xi[(n - i) * xs]); o[2] = xr[i * xs] - xr[(n - i) * xs]; o[3] = xi[i * xs] - xi[(n - i) * xs]; o += 4; } *pr = sr; *pi = si; } static void apply(const plan *ego_, R *ri, R *ii, R *ro, R *io) { const P *ego = (const P *) ego_; INT i; INT n = ego->n, is = ego->is, os = ego->os; const R *W = ego->td->W; E *buf; size_t bufsz = n * 2 * sizeof(E); BUF_ALLOC(E *, buf, bufsz); hartley(n, ri, ii, is, buf, ro, io); for (i = 1; i + i < n; ++i) { cdot(n, buf, W, ro + i * os, io + i * os, ro + (n - i) * os, io + (n - i) * os); W += n - 1; } BUF_FREE(buf, bufsz); } static void awake(plan *ego_, enum wakefulness wakefulness) { P *ego = (P *) ego_; static const tw_instr half_tw[] = { { TW_HALF, 1, 0 }, { TW_NEXT, 1, 0 } }; X(twiddle_awake)(wakefulness, &ego->td, half_tw, ego->n, ego->n, (ego->n - 1) / 2); } static void print(const plan *ego_, printer *p) { const P *ego = (const P *) ego_; p->print(p, "(dft-generic-%D)", ego->n); } static int applicable(const solver *ego, const problem *p_, const planner *plnr) { const problem_dft *p = (const problem_dft *) p_; UNUSED(ego); return (1 && p->sz->rnk == 1 && p->vecsz->rnk == 0 && (p->sz->dims[0].n % 2) == 1 && CIMPLIES(NO_LARGE_GENERICP(plnr), p->sz->dims[0].n < GENERIC_MIN_BAD) && CIMPLIES(NO_SLOWP(plnr), p->sz->dims[0].n > GENERIC_MAX_SLOW) && X(is_prime)(p->sz->dims[0].n) ); } static plan *mkplan(const solver *ego, const problem *p_, planner *plnr) { const problem_dft *p; P *pln; INT n; static const plan_adt padt = { X(dft_solve), awake, print, X(plan_null_destroy) }; if (!applicable(ego, p_, plnr)) return (plan *)0; pln = MKPLAN_DFT(P, &padt, apply); p = (const problem_dft *) p_; pln->n = n = p->sz->dims[0].n; pln->is = p->sz->dims[0].is; pln->os = p->sz->dims[0].os; pln->td = 0; pln->super.super.ops.add = (n-1) * 5; pln->super.super.ops.mul = 0; pln->super.super.ops.fma = (n-1) * (n-1) ; #if 0 /* these are nice pipelined sequential loads and should cost nothing */ pln->super.super.ops.other = (n-1)*(4 + 1 + 2 * (n-1)); /* approximate */ #endif return &(pln->super.super); } static solver *mksolver(void) { static const solver_adt sadt = { PROBLEM_DFT, mkplan, 0 }; S *slv = MKSOLVER(S, &sadt); return &(slv->super); } void X(dft_generic_register)(planner *p) { REGISTER_SOLVER(p, mksolver()); } fftw-3.3.3/dft/indirect.c0000644000175000017500000001527212054410050012127 00000000000000/* * Copyright (c) 2003, 2007-11 Matteo Frigo * Copyright (c) 2003, 2007-11 Massachusetts Institute of Technology * * This program is free software; you can redistribute it and/or modify * it under the terms of the GNU General Public License as published by * the Free Software Foundation; either version 2 of the License, or * (at your option) any later version. * * This program is distributed in the hope that it will be useful, * but WITHOUT ANY WARRANTY; without even the implied warranty of * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the * GNU General Public License for more details. * * You should have received a copy of the GNU General Public License * along with this program; if not, write to the Free Software * Foundation, Inc., 51 Franklin Street, Fifth Floor, Boston, MA 02110-1301 USA * */ /* solvers/plans for vectors of small DFT's that cannot be done in-place directly. Use a rank-0 plan to rearrange the data before or after the transform. Can also change an out-of-place plan into a copy + in-place (where the in-place transform is e.g. unit stride). */ /* FIXME: merge with rank-geq2.c(?), since this is just a special case of a rank split where the first/second transform has rank 0. */ #include "dft.h" typedef problem *(*mkcld_t) (const problem_dft *p); typedef struct { dftapply apply; problem *(*mkcld)(const problem_dft *p); const char *nam; } ndrct_adt; typedef struct { solver super; const ndrct_adt *adt; } S; typedef struct { plan_dft super; plan *cldcpy, *cld; const S *slv; } P; /*-----------------------------------------------------------------------*/ /* first rearrange, then transform */ static void apply_before(const plan *ego_, R *ri, R *ii, R *ro, R *io) { const P *ego = (const P *) ego_; { plan_dft *cldcpy = (plan_dft *) ego->cldcpy; cldcpy->apply(ego->cldcpy, ri, ii, ro, io); } { plan_dft *cld = (plan_dft *) ego->cld; cld->apply(ego->cld, ro, io, ro, io); } } static problem *mkcld_before(const problem_dft *p) { return X(mkproblem_dft_d)(X(tensor_copy_inplace)(p->sz, INPLACE_OS), X(tensor_copy_inplace)(p->vecsz, INPLACE_OS), p->ro, p->io, p->ro, p->io); } static const ndrct_adt adt_before = { apply_before, mkcld_before, "dft-indirect-before" }; /*-----------------------------------------------------------------------*/ /* first transform, then rearrange */ static void apply_after(const plan *ego_, R *ri, R *ii, R *ro, R *io) { const P *ego = (const P *) ego_; { plan_dft *cld = (plan_dft *) ego->cld; cld->apply(ego->cld, ri, ii, ri, ii); } { plan_dft *cldcpy = (plan_dft *) ego->cldcpy; cldcpy->apply(ego->cldcpy, ri, ii, ro, io); } } static problem *mkcld_after(const problem_dft *p) { return X(mkproblem_dft_d)(X(tensor_copy_inplace)(p->sz, INPLACE_IS), X(tensor_copy_inplace)(p->vecsz, INPLACE_IS), p->ri, p->ii, p->ri, p->ii); } static const ndrct_adt adt_after = { apply_after, mkcld_after, "dft-indirect-after" }; /*-----------------------------------------------------------------------*/ static void destroy(plan *ego_) { P *ego = (P *) ego_; X(plan_destroy_internal)(ego->cld); X(plan_destroy_internal)(ego->cldcpy); } static void awake(plan *ego_, enum wakefulness wakefulness) { P *ego = (P *) ego_; X(plan_awake)(ego->cldcpy, wakefulness); X(plan_awake)(ego->cld, wakefulness); } static void print(const plan *ego_, printer *p) { const P *ego = (const P *) ego_; const S *s = ego->slv; p->print(p, "(%s%(%p%)%(%p%))", s->adt->nam, ego->cld, ego->cldcpy); } static int applicable0(const solver *ego_, const problem *p_, const planner *plnr) { const S *ego = (const S *) ego_; const problem_dft *p = (const problem_dft *) p_; return (1 && FINITE_RNK(p->vecsz->rnk) /* problem must be a nontrivial transform, not just a copy */ && p->sz->rnk > 0 && (0 /* problem must be in-place & require some rearrangement of the data; to prevent infinite loops with indirect-transpose, we further require that at least some transform strides must decrease */ || (p->ri == p->ro && !X(tensor_inplace_strides2)(p->sz, p->vecsz) && X(tensor_strides_decrease)( p->sz, p->vecsz, ego->adt->apply == apply_after ? INPLACE_IS : INPLACE_OS)) /* or problem must be out of place, transforming from stride 1/2 to bigger stride, for apply_after */ || (p->ri != p->ro && ego->adt->apply == apply_after && !NO_DESTROY_INPUTP(plnr) && X(tensor_min_istride)(p->sz) <= 2 && X(tensor_min_ostride)(p->sz) > 2) /* or problem must be out of place, transforming to stride 1/2 from bigger stride, for apply_before */ || (p->ri != p->ro && ego->adt->apply == apply_before && X(tensor_min_ostride)(p->sz) <= 2 && X(tensor_min_istride)(p->sz) > 2) ) ); } static int applicable(const solver *ego_, const problem *p_, const planner *plnr) { if (!applicable0(ego_, p_, plnr)) return 0; { const problem_dft *p = (const problem_dft *) p_; if (NO_INDIRECT_OP_P(plnr) && p->ri != p->ro) return 0; } return 1; } static plan *mkplan(const solver *ego_, const problem *p_, planner *plnr) { const problem_dft *p = (const problem_dft *) p_; const S *ego = (const S *) ego_; P *pln; plan *cld = 0, *cldcpy = 0; static const plan_adt padt = { X(dft_solve), awake, print, destroy }; if (!applicable(ego_, p_, plnr)) return (plan *) 0; cldcpy = X(mkplan_d)(plnr, X(mkproblem_dft_d)(X(mktensor_0d)(), X(tensor_append)(p->vecsz, p->sz), p->ri, p->ii, p->ro, p->io)); if (!cldcpy) goto nada; cld = X(mkplan_f_d)(plnr, ego->adt->mkcld(p), NO_BUFFERING, 0, 0); if (!cld) goto nada; pln = MKPLAN_DFT(P, &padt, ego->adt->apply); pln->cld = cld; pln->cldcpy = cldcpy; pln->slv = ego; X(ops_add)(&cld->ops, &cldcpy->ops, &pln->super.super.ops); return &(pln->super.super); nada: X(plan_destroy_internal)(cld); X(plan_destroy_internal)(cldcpy); return (plan *)0; } static solver *mksolver(const ndrct_adt *adt) { static const solver_adt sadt = { PROBLEM_DFT, mkplan, 0 }; S *slv = MKSOLVER(S, &sadt); slv->adt = adt; return &(slv->super); } void X(dft_indirect_register)(planner *p) { unsigned i; static const ndrct_adt *const adts[] = { &adt_before, &adt_after }; for (i = 0; i < sizeof(adts) / sizeof(adts[0]); ++i) REGISTER_SOLVER(p, mksolver(adts[i])); } fftw-3.3.3/dft/indirect-transpose.c0000644000175000017500000001632312054410050014141 00000000000000/* * Copyright (c) 2003, 2007-11 Matteo Frigo * Copyright (c) 2003, 2007-11 Massachusetts Institute of Technology * * This program is free software; you can redistribute it and/or modify * it under the terms of the GNU General Public License as published by * the Free Software Foundation; either version 2 of the License, or * (at your option) any later version. * * This program is distributed in the hope that it will be useful, * but WITHOUT ANY WARRANTY; without even the implied warranty of * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the * GNU General Public License for more details. * * You should have received a copy of the GNU General Public License * along with this program; if not, write to the Free Software * Foundation, Inc., 51 Franklin Street, Fifth Floor, Boston, MA 02110-1301 USA * */ /* solvers/plans for vectors of DFTs corresponding to the columns of a matrix: first transpose the matrix so that the DFTs are contiguous, then do DFTs with transposed output. In particular, we restrict ourselves to the case of a square transpose (or a sequence thereof). */ #include "dft.h" typedef solver S; typedef struct { plan_dft super; INT vl, ivs, ovs; plan *cldtrans, *cld, *cldrest; } P; /* initial transpose is out-of-place from input to output */ static void apply_op(const plan *ego_, R *ri, R *ii, R *ro, R *io) { const P *ego = (const P *) ego_; INT vl = ego->vl, ivs = ego->ivs, ovs = ego->ovs, i; for (i = 0; i < vl; ++i) { { plan_dft *cldtrans = (plan_dft *) ego->cldtrans; cldtrans->apply(ego->cldtrans, ri, ii, ro, io); } { plan_dft *cld = (plan_dft *) ego->cld; cld->apply(ego->cld, ro, io, ro, io); } ri += ivs; ii += ivs; ro += ovs; io += ovs; } { plan_dft *cldrest = (plan_dft *) ego->cldrest; cldrest->apply(ego->cldrest, ri, ii, ro, io); } } static void destroy(plan *ego_) { P *ego = (P *) ego_; X(plan_destroy_internal)(ego->cldrest); X(plan_destroy_internal)(ego->cld); X(plan_destroy_internal)(ego->cldtrans); } static void awake(plan *ego_, enum wakefulness wakefulness) { P *ego = (P *) ego_; X(plan_awake)(ego->cldtrans, wakefulness); X(plan_awake)(ego->cld, wakefulness); X(plan_awake)(ego->cldrest, wakefulness); } static void print(const plan *ego_, printer *p) { const P *ego = (const P *) ego_; p->print(p, "(indirect-transpose%v%(%p%)%(%p%)%(%p%))", ego->vl, ego->cldtrans, ego->cld, ego->cldrest); } static int pickdim(const tensor *vs, const tensor *s, int *pdim0, int *pdim1) { int dim0, dim1; *pdim0 = *pdim1 = -1; for (dim0 = 0; dim0 < vs->rnk; ++dim0) for (dim1 = 0; dim1 < s->rnk; ++dim1) if (vs->dims[dim0].n * X(iabs)(vs->dims[dim0].is) <= X(iabs)(s->dims[dim1].is) && vs->dims[dim0].n >= s->dims[dim1].n && (*pdim0 == -1 || (X(iabs)(vs->dims[dim0].is) <= X(iabs)(vs->dims[*pdim0].is) && X(iabs)(s->dims[dim1].is) >= X(iabs)(s->dims[*pdim1].is)))) { *pdim0 = dim0; *pdim1 = dim1; } return (*pdim0 != -1 && *pdim1 != -1); } static int applicable0(const solver *ego_, const problem *p_, const planner *plnr, int *pdim0, int *pdim1) { const problem_dft *p = (const problem_dft *) p_; UNUSED(ego_); UNUSED(plnr); return (1 && FINITE_RNK(p->vecsz->rnk) && FINITE_RNK(p->sz->rnk) /* FIXME: can/should we relax this constraint? */ && X(tensor_inplace_strides2)(p->vecsz, p->sz) && pickdim(p->vecsz, p->sz, pdim0, pdim1) /* output should not *already* include the transpose (in which case we duplicate the regular indirect.c) */ && (p->sz->dims[*pdim1].os != p->vecsz->dims[*pdim0].is) ); } static int applicable(const solver *ego_, const problem *p_, const planner *plnr, int *pdim0, int *pdim1) { if (!applicable0(ego_, p_, plnr, pdim0, pdim1)) return 0; { const problem_dft *p = (const problem_dft *) p_; INT u = p->ri == p->ii + 1 || p->ii == p->ri + 1 ? (INT)2 : (INT)1; /* UGLY if does not result in contiguous transforms or transforms of contiguous vectors (since the latter at least have efficient transpositions) */ if (NO_UGLYP(plnr) && p->vecsz->dims[*pdim0].is != u && !(p->vecsz->rnk == 2 && p->vecsz->dims[1-*pdim0].is == u && p->vecsz->dims[*pdim0].is == u * p->vecsz->dims[1-*pdim0].n)) return 0; if (NO_INDIRECT_OP_P(plnr) && p->ri != p->ro) return 0; } return 1; } static plan *mkplan(const solver *ego_, const problem *p_, planner *plnr) { const problem_dft *p = (const problem_dft *) p_; P *pln; plan *cld = 0, *cldtrans = 0, *cldrest = 0; int pdim0, pdim1; tensor *ts, *tv; INT vl, ivs, ovs; R *rit, *iit, *rot, *iot; static const plan_adt padt = { X(dft_solve), awake, print, destroy }; if (!applicable(ego_, p_, plnr, &pdim0, &pdim1)) return (plan *) 0; vl = p->vecsz->dims[pdim0].n / p->sz->dims[pdim1].n; A(vl >= 1); ivs = p->sz->dims[pdim1].n * p->vecsz->dims[pdim0].is; ovs = p->sz->dims[pdim1].n * p->vecsz->dims[pdim0].os; rit = TAINT(p->ri, vl == 1 ? 0 : ivs); iit = TAINT(p->ii, vl == 1 ? 0 : ivs); rot = TAINT(p->ro, vl == 1 ? 0 : ovs); iot = TAINT(p->io, vl == 1 ? 0 : ovs); ts = X(tensor_copy_inplace)(p->sz, INPLACE_IS); ts->dims[pdim1].os = p->vecsz->dims[pdim0].is; tv = X(tensor_copy_inplace)(p->vecsz, INPLACE_IS); tv->dims[pdim0].os = p->sz->dims[pdim1].is; tv->dims[pdim0].n = p->sz->dims[pdim1].n; cldtrans = X(mkplan_d)(plnr, X(mkproblem_dft_d)(X(mktensor_0d)(), X(tensor_append)(tv, ts), rit, iit, rot, iot)); X(tensor_destroy2)(ts, tv); if (!cldtrans) goto nada; ts = X(tensor_copy)(p->sz); ts->dims[pdim1].is = p->vecsz->dims[pdim0].is; tv = X(tensor_copy)(p->vecsz); tv->dims[pdim0].is = p->sz->dims[pdim1].is; tv->dims[pdim0].n = p->sz->dims[pdim1].n; cld = X(mkplan_d)(plnr, X(mkproblem_dft_d)(ts, tv, rot, iot, rot, iot)); if (!cld) goto nada; tv = X(tensor_copy)(p->vecsz); tv->dims[pdim0].n -= vl * p->sz->dims[pdim1].n; cldrest = X(mkplan_d)(plnr, X(mkproblem_dft_d)(X(tensor_copy)(p->sz), tv, p->ri + ivs * vl, p->ii + ivs * vl, p->ro + ovs * vl, p->io + ovs * vl)); if (!cldrest) goto nada; pln = MKPLAN_DFT(P, &padt, apply_op); pln->cldtrans = cldtrans; pln->cld = cld; pln->cldrest = cldrest; pln->vl = vl; pln->ivs = ivs; pln->ovs = ovs; X(ops_cpy)(&cldrest->ops, &pln->super.super.ops); X(ops_madd2)(vl, &cld->ops, &pln->super.super.ops); X(ops_madd2)(vl, &cldtrans->ops, &pln->super.super.ops); return &(pln->super.super); nada: X(plan_destroy_internal)(cldrest); X(plan_destroy_internal)(cld); X(plan_destroy_internal)(cldtrans); return (plan *)0; } static solver *mksolver(void) { static const solver_adt sadt = { PROBLEM_DFT, mkplan, 0 }; S *slv = MKSOLVER(S, &sadt); return slv; } void X(dft_indirect_transpose_register)(planner *p) { REGISTER_SOLVER(p, mksolver()); } fftw-3.3.3/dft/kdft-dif.c0000644000175000017500000000174512054410050012016 00000000000000/* * Copyright (c) 2003, 2007-11 Matteo Frigo * Copyright (c) 2003, 2007-11 Massachusetts Institute of Technology * * This program is free software; you can redistribute it and/or modify * it under the terms of the GNU General Public License as published by * the Free Software Foundation; either version 2 of the License, or * (at your option) any later version. * * This program is distributed in the hope that it will be useful, * but WITHOUT ANY WARRANTY; without even the implied warranty of * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the * GNU General Public License for more details. * * You should have received a copy of the GNU General Public License * along with this program; if not, write to the Free Software * Foundation, Inc., 51 Franklin Street, Fifth Floor, Boston, MA 02110-1301 USA * */ #include "ct.h" void X(kdft_dif_register)(planner *p, kdftw codelet, const ct_desc *desc) { X(regsolver_ct_directw)(p, codelet, desc, DECDIF); } fftw-3.3.3/dft/kdft-difsq.c0000644000175000017500000000173712054410050012363 00000000000000/* * Copyright (c) 2003, 2007-11 Matteo Frigo * Copyright (c) 2003, 2007-11 Massachusetts Institute of Technology * * This program is free software; you can redistribute it and/or modify * it under the terms of the GNU General Public License as published by * the Free Software Foundation; either version 2 of the License, or * (at your option) any later version. * * This program is distributed in the hope that it will be useful, * but WITHOUT ANY WARRANTY; without even the implied warranty of * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the * GNU General Public License for more details. * * You should have received a copy of the GNU General Public License * along with this program; if not, write to the Free Software * Foundation, Inc., 51 Franklin Street, Fifth Floor, Boston, MA 02110-1301 USA * */ #include "ct.h" void X(kdft_difsq_register)(planner *p, kdftwsq k, const ct_desc *desc) { X(regsolver_ct_directwsq)(p, k, desc, DECDIF); } fftw-3.3.3/dft/kdft-dit.c0000644000175000017500000000174512054410050012034 00000000000000/* * Copyright (c) 2003, 2007-11 Matteo Frigo * Copyright (c) 2003, 2007-11 Massachusetts Institute of Technology * * This program is free software; you can redistribute it and/or modify * it under the terms of the GNU General Public License as published by * the Free Software Foundation; either version 2 of the License, or * (at your option) any later version. * * This program is distributed in the hope that it will be useful, * but WITHOUT ANY WARRANTY; without even the implied warranty of * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the * GNU General Public License for more details. * * You should have received a copy of the GNU General Public License * along with this program; if not, write to the Free Software * Foundation, Inc., 51 Franklin Street, Fifth Floor, Boston, MA 02110-1301 USA * */ #include "ct.h" void X(kdft_dit_register)(planner *p, kdftw codelet, const ct_desc *desc) { X(regsolver_ct_directw)(p, codelet, desc, DECDIT); } fftw-3.3.3/dft/kdft.c0000644000175000017500000000205612054410050011252 00000000000000/* * Copyright (c) 2003, 2007-11 Matteo Frigo * Copyright (c) 2003, 2007-11 Massachusetts Institute of Technology * * This program is free software; you can redistribute it and/or modify * it under the terms of the GNU General Public License as published by * the Free Software Foundation; either version 2 of the License, or * (at your option) any later version. * * This program is distributed in the hope that it will be useful, * but WITHOUT ANY WARRANTY; without even the implied warranty of * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the * GNU General Public License for more details. * * You should have received a copy of the GNU General Public License * along with this program; if not, write to the Free Software * Foundation, Inc., 51 Franklin Street, Fifth Floor, Boston, MA 02110-1301 USA * */ #include "dft.h" void X(kdft_register)(planner *p, kdft codelet, const kdft_desc *desc) { REGISTER_SOLVER(p, X(mksolver_dft_direct)(codelet, desc)); REGISTER_SOLVER(p, X(mksolver_dft_directbuf)(codelet, desc)); } fftw-3.3.3/dft/nop.c0000644000175000017500000000422412054410050011115 00000000000000/* * Copyright (c) 2003, 2007-11 Matteo Frigo * Copyright (c) 2003, 2007-11 Massachusetts Institute of Technology * * This program is free software; you can redistribute it and/or modify * it under the terms of the GNU General Public License as published by * the Free Software Foundation; either version 2 of the License, or * (at your option) any later version. * * This program is distributed in the hope that it will be useful, * but WITHOUT ANY WARRANTY; without even the implied warranty of * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the * GNU General Public License for more details. * * You should have received a copy of the GNU General Public License * along with this program; if not, write to the Free Software * Foundation, Inc., 51 Franklin Street, Fifth Floor, Boston, MA 02110-1301 USA * */ /* plans for vrank -infty DFTs (nothing to do) */ #include "dft.h" static void apply(const plan *ego_, R *ri, R *ii, R *ro, R *io) { UNUSED(ego_); UNUSED(ri); UNUSED(ii); UNUSED(ro); UNUSED(io); } static int applicable(const solver *ego_, const problem *p_) { const problem_dft *p = (const problem_dft *) p_; UNUSED(ego_); return 0 /* case 1 : -infty vector rank */ || (!FINITE_RNK(p->vecsz->rnk)) /* case 2 : rank-0 in-place dft */ || (1 && p->sz->rnk == 0 && FINITE_RNK(p->vecsz->rnk) && p->ro == p->ri && X(tensor_inplace_strides)(p->vecsz) ); } static void print(const plan *ego, printer *p) { UNUSED(ego); p->print(p, "(dft-nop)"); } static plan *mkplan(const solver *ego, const problem *p, planner *plnr) { static const plan_adt padt = { X(dft_solve), X(null_awake), print, X(plan_null_destroy) }; plan_dft *pln; UNUSED(plnr); if (!applicable(ego, p)) return (plan *) 0; pln = MKPLAN_DFT(plan_dft, &padt, apply); X(ops_zero)(&pln->super.ops); return &(pln->super); } static solver *mksolver(void) { static const solver_adt sadt = { PROBLEM_DFT, mkplan, 0 }; return MKSOLVER(solver, &sadt); } void X(dft_nop_register)(planner *p) { REGISTER_SOLVER(p, mksolver()); } fftw-3.3.3/dft/plan.c0000644000175000017500000000204212054410050011247 00000000000000/* * Copyright (c) 2003, 2007-11 Matteo Frigo * Copyright (c) 2003, 2007-11 Massachusetts Institute of Technology * * This program is free software; you can redistribute it and/or modify * it under the terms of the GNU General Public License as published by * the Free Software Foundation; either version 2 of the License, or * (at your option) any later version. * * This program is distributed in the hope that it will be useful, * but WITHOUT ANY WARRANTY; without even the implied warranty of * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the * GNU General Public License for more details. * * You should have received a copy of the GNU General Public License * along with this program; if not, write to the Free Software * Foundation, Inc., 51 Franklin Street, Fifth Floor, Boston, MA 02110-1301 USA * */ #include "dft.h" plan *X(mkplan_dft)(size_t size, const plan_adt *adt, dftapply apply) { plan_dft *ego; ego = (plan_dft *) X(mkplan)(size, adt); ego->apply = apply; return &(ego->super); } fftw-3.3.3/dft/problem.c0000644000175000017500000000670712054410050011771 00000000000000/* * Copyright (c) 2003, 2007-11 Matteo Frigo * Copyright (c) 2003, 2007-11 Massachusetts Institute of Technology * * This program is free software; you can redistribute it and/or modify * it under the terms of the GNU General Public License as published by * the Free Software Foundation; either version 2 of the License, or * (at your option) any later version. * * This program is distributed in the hope that it will be useful, * but WITHOUT ANY WARRANTY; without even the implied warranty of * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the * GNU General Public License for more details. * * You should have received a copy of the GNU General Public License * along with this program; if not, write to the Free Software * Foundation, Inc., 51 Franklin Street, Fifth Floor, Boston, MA 02110-1301 USA * */ #include "dft.h" #include static void destroy(problem *ego_) { problem_dft *ego = (problem_dft *) ego_; X(tensor_destroy2)(ego->vecsz, ego->sz); X(ifree)(ego_); } static void hash(const problem *p_, md5 *m) { const problem_dft *p = (const problem_dft *) p_; X(md5puts)(m, "dft"); X(md5int)(m, p->ri == p->ro); X(md5INT)(m, p->ii - p->ri); X(md5INT)(m, p->io - p->ro); X(md5int)(m, X(alignment_of)(p->ri)); X(md5int)(m, X(alignment_of)(p->ii)); X(md5int)(m, X(alignment_of)(p->ro)); X(md5int)(m, X(alignment_of)(p->io)); X(tensor_md5)(m, p->sz); X(tensor_md5)(m, p->vecsz); } static void print(const problem *ego_, printer *p) { const problem_dft *ego = (const problem_dft *) ego_; p->print(p, "(dft %d %d %d %D %D %T %T)", ego->ri == ego->ro, X(alignment_of)(ego->ri), X(alignment_of)(ego->ro), (INT)(ego->ii - ego->ri), (INT)(ego->io - ego->ro), ego->sz, ego->vecsz); } static void zero(const problem *ego_) { const problem_dft *ego = (const problem_dft *) ego_; tensor *sz = X(tensor_append)(ego->vecsz, ego->sz); X(dft_zerotens)(sz, UNTAINT(ego->ri), UNTAINT(ego->ii)); X(tensor_destroy)(sz); } static const problem_adt padt = { PROBLEM_DFT, hash, zero, print, destroy }; problem *X(mkproblem_dft)(const tensor *sz, const tensor *vecsz, R *ri, R *ii, R *ro, R *io) { problem_dft *ego; /* enforce pointer equality if untainted pointers are equal */ if (UNTAINT(ri) == UNTAINT(ro)) ri = ro = JOIN_TAINT(ri, ro); if (UNTAINT(ii) == UNTAINT(io)) ii = io = JOIN_TAINT(ii, io); /* more correctness conditions: */ A(TAINTOF(ri) == TAINTOF(ii)); A(TAINTOF(ro) == TAINTOF(io)); A(X(tensor_kosherp)(sz)); A(X(tensor_kosherp)(vecsz)); if (ri == ro || ii == io) { /* If either real or imag pointers are in place, both must be. */ if (ri != ro || ii != io || !X(tensor_inplace_locations)(sz, vecsz)) return X(mkproblem_unsolvable)(); } ego = (problem_dft *)X(mkproblem)(sizeof(problem_dft), &padt); ego->sz = X(tensor_compress)(sz); ego->vecsz = X(tensor_compress_contiguous)(vecsz); ego->ri = ri; ego->ii = ii; ego->ro = ro; ego->io = io; A(FINITE_RNK(ego->sz->rnk)); return &(ego->super); } /* Same as X(mkproblem_dft), but also destroy input tensors. */ problem *X(mkproblem_dft_d)(tensor *sz, tensor *vecsz, R *ri, R *ii, R *ro, R *io) { problem *p = X(mkproblem_dft)(sz, vecsz, ri, ii, ro, io); X(tensor_destroy2)(vecsz, sz); return p; } fftw-3.3.3/dft/rader.c0000644000175000017500000002127012054410050011416 00000000000000/* * Copyright (c) 2003, 2007-11 Matteo Frigo * Copyright (c) 2003, 2007-11 Massachusetts Institute of Technology * * This program is free software; you can redistribute it and/or modify * it under the terms of the GNU General Public License as published by * the Free Software Foundation; either version 2 of the License, or * (at your option) any later version. * * This program is distributed in the hope that it will be useful, * but WITHOUT ANY WARRANTY; without even the implied warranty of * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the * GNU General Public License for more details. * * You should have received a copy of the GNU General Public License * along with this program; if not, write to the Free Software * Foundation, Inc., 51 Franklin Street, Fifth Floor, Boston, MA 02110-1301 USA * */ #include "dft.h" /* * Compute transforms of prime sizes using Rader's trick: turn them * into convolutions of size n - 1, which you then perform via a pair * of FFTs. */ typedef struct { solver super; } S; typedef struct { plan_dft super; plan *cld1, *cld2; R *omega; INT n, g, ginv; INT is, os; plan *cld_omega; } P; static rader_tl *omegas = 0; static R *mkomega(enum wakefulness wakefulness, plan *p_, INT n, INT ginv) { plan_dft *p = (plan_dft *) p_; R *omega; INT i, gpower; trigreal scale; triggen *t; if ((omega = X(rader_tl_find)(n, n, ginv, omegas))) return omega; omega = (R *)MALLOC(sizeof(R) * (n - 1) * 2, TWIDDLES); scale = n - 1.0; /* normalization for convolution */ t = X(mktriggen)(wakefulness, n); for (i = 0, gpower = 1; i < n-1; ++i, gpower = MULMOD(gpower, ginv, n)) { trigreal w[2]; t->cexpl(t, gpower, w); omega[2*i] = w[0] / scale; omega[2*i+1] = FFT_SIGN * w[1] / scale; } X(triggen_destroy)(t); A(gpower == 1); p->apply(p_, omega, omega + 1, omega, omega + 1); X(rader_tl_insert)(n, n, ginv, omega, &omegas); return omega; } static void free_omega(R *omega) { X(rader_tl_delete)(omega, &omegas); } /***************************************************************************/ /* Below, we extensively use the identity that fft(x*)* = ifft(x) in order to share data between forward and backward transforms and to obviate the necessity of having separate forward and backward plans. (Although we often compute separate plans these days anyway due to the differing strides, etcetera.) Of course, since the new FFTW gives us separate pointers to the real and imaginary parts, we could have instead used the fft(r,i) = ifft(i,r) form of this identity, but it was easier to reuse the code from our old version. */ static void apply(const plan *ego_, R *ri, R *ii, R *ro, R *io) { const P *ego = (const P *) ego_; INT is, os; INT k, gpower, g, r; R *buf; R r0 = ri[0], i0 = ii[0]; r = ego->n; is = ego->is; os = ego->os; g = ego->g; buf = (R *) MALLOC(sizeof(R) * (r - 1) * 2, BUFFERS); /* First, permute the input, storing in buf: */ for (gpower = 1, k = 0; k < r - 1; ++k, gpower = MULMOD(gpower, g, r)) { R rA, iA; rA = ri[gpower * is]; iA = ii[gpower * is]; buf[2*k] = rA; buf[2*k + 1] = iA; } /* gpower == g^(r-1) mod r == 1 */; /* compute DFT of buf, storing in output (except DC): */ { plan_dft *cld = (plan_dft *) ego->cld1; cld->apply(ego->cld1, buf, buf+1, ro+os, io+os); } /* set output DC component: */ { ro[0] = r0 + ro[os]; io[0] = i0 + io[os]; } /* now, multiply by omega: */ { const R *omega = ego->omega; for (k = 0; k < r - 1; ++k) { E rB, iB, rW, iW; rW = omega[2*k]; iW = omega[2*k+1]; rB = ro[(k+1)*os]; iB = io[(k+1)*os]; ro[(k+1)*os] = rW * rB - iW * iB; io[(k+1)*os] = -(rW * iB + iW * rB); } } /* this will add input[0] to all of the outputs after the ifft */ ro[os] += r0; io[os] -= i0; /* inverse FFT: */ { plan_dft *cld = (plan_dft *) ego->cld2; cld->apply(ego->cld2, ro+os, io+os, buf, buf+1); } /* finally, do inverse permutation to unshuffle the output: */ { INT ginv = ego->ginv; gpower = 1; for (k = 0; k < r - 1; ++k, gpower = MULMOD(gpower, ginv, r)) { ro[gpower * os] = buf[2*k]; io[gpower * os] = -buf[2*k+1]; } A(gpower == 1); } X(ifree)(buf); } /***************************************************************************/ static void awake(plan *ego_, enum wakefulness wakefulness) { P *ego = (P *) ego_; X(plan_awake)(ego->cld1, wakefulness); X(plan_awake)(ego->cld2, wakefulness); X(plan_awake)(ego->cld_omega, wakefulness); switch (wakefulness) { case SLEEPY: free_omega(ego->omega); ego->omega = 0; break; default: ego->g = X(find_generator)(ego->n); ego->ginv = X(power_mod)(ego->g, ego->n - 2, ego->n); A(MULMOD(ego->g, ego->ginv, ego->n) == 1); ego->omega = mkomega(wakefulness, ego->cld_omega, ego->n, ego->ginv); break; } } static void destroy(plan *ego_) { P *ego = (P *) ego_; X(plan_destroy_internal)(ego->cld_omega); X(plan_destroy_internal)(ego->cld2); X(plan_destroy_internal)(ego->cld1); } static void print(const plan *ego_, printer *p) { const P *ego = (const P *)ego_; p->print(p, "(dft-rader-%D%ois=%oos=%(%p%)", ego->n, ego->is, ego->os, ego->cld1); if (ego->cld2 != ego->cld1) p->print(p, "%(%p%)", ego->cld2); if (ego->cld_omega != ego->cld1 && ego->cld_omega != ego->cld2) p->print(p, "%(%p%)", ego->cld_omega); p->putchr(p, ')'); } static int applicable(const solver *ego_, const problem *p_, const planner *plnr) { const problem_dft *p = (const problem_dft *) p_; UNUSED(ego_); return (1 && p->sz->rnk == 1 && p->vecsz->rnk == 0 && CIMPLIES(NO_SLOWP(plnr), p->sz->dims[0].n > RADER_MAX_SLOW) && X(is_prime)(p->sz->dims[0].n) /* proclaim the solver SLOW if p-1 is not easily factorizable. Bluestein should take care of this case. */ && CIMPLIES(NO_SLOWP(plnr), X(factors_into_small_primes)(p->sz->dims[0].n - 1)) ); } static int mkP(P *pln, INT n, INT is, INT os, R *ro, R *io, planner *plnr) { plan *cld1 = (plan *) 0; plan *cld2 = (plan *) 0; plan *cld_omega = (plan *) 0; R *buf = (R *) 0; /* initial allocation for the purpose of planning */ buf = (R *) MALLOC(sizeof(R) * (n - 1) * 2, BUFFERS); cld1 = X(mkplan_f_d)(plnr, X(mkproblem_dft_d)(X(mktensor_1d)(n - 1, 2, os), X(mktensor_1d)(1, 0, 0), buf, buf + 1, ro + os, io + os), NO_SLOW, 0, 0); if (!cld1) goto nada; cld2 = X(mkplan_f_d)(plnr, X(mkproblem_dft_d)(X(mktensor_1d)(n - 1, os, 2), X(mktensor_1d)(1, 0, 0), ro + os, io + os, buf, buf + 1), NO_SLOW, 0, 0); if (!cld2) goto nada; /* plan for omega array */ cld_omega = X(mkplan_f_d)(plnr, X(mkproblem_dft_d)(X(mktensor_1d)(n - 1, 2, 2), X(mktensor_1d)(1, 0, 0), buf, buf + 1, buf, buf + 1), NO_SLOW, ESTIMATE, 0); if (!cld_omega) goto nada; /* deallocate buffers; let awake() or apply() allocate them for real */ X(ifree)(buf); buf = 0; pln->cld1 = cld1; pln->cld2 = cld2; pln->cld_omega = cld_omega; pln->omega = 0; pln->n = n; pln->is = is; pln->os = os; X(ops_add)(&cld1->ops, &cld2->ops, &pln->super.super.ops); pln->super.super.ops.other += (n - 1) * (4 * 2 + 6) + 6; pln->super.super.ops.add += (n - 1) * 2 + 4; pln->super.super.ops.mul += (n - 1) * 4; return 1; nada: X(ifree0)(buf); X(plan_destroy_internal)(cld_omega); X(plan_destroy_internal)(cld2); X(plan_destroy_internal)(cld1); return 0; } static plan *mkplan(const solver *ego, const problem *p_, planner *plnr) { const problem_dft *p = (const problem_dft *) p_; P *pln; INT n; INT is, os; static const plan_adt padt = { X(dft_solve), awake, print, destroy }; if (!applicable(ego, p_, plnr)) return (plan *) 0; n = p->sz->dims[0].n; is = p->sz->dims[0].is; os = p->sz->dims[0].os; pln = MKPLAN_DFT(P, &padt, apply); if (!mkP(pln, n, is, os, p->ro, p->io, plnr)) { X(ifree)(pln); return (plan *) 0; } return &(pln->super.super); } static solver *mksolver(void) { static const solver_adt sadt = { PROBLEM_DFT, mkplan, 0 }; S *slv = MKSOLVER(S, &sadt); return &(slv->super); } void X(dft_rader_register)(planner *p) { REGISTER_SOLVER(p, mksolver()); } fftw-3.3.3/dft/rank-geq2.c0000644000175000017500000001311612054410050012110 00000000000000/* * Copyright (c) 2003, 2007-11 Matteo Frigo * Copyright (c) 2003, 2007-11 Massachusetts Institute of Technology * * This program is free software; you can redistribute it and/or modify * it under the terms of the GNU General Public License as published by * the Free Software Foundation; either version 2 of the License, or * (at your option) any later version. * * This program is distributed in the hope that it will be useful, * but WITHOUT ANY WARRANTY; without even the implied warranty of * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the * GNU General Public License for more details. * * You should have received a copy of the GNU General Public License * along with this program; if not, write to the Free Software * Foundation, Inc., 51 Franklin Street, Fifth Floor, Boston, MA 02110-1301 USA * */ /* plans for DFT of rank >= 2 (multidimensional) */ #include "dft.h" typedef struct { solver super; int spltrnk; const int *buddies; int nbuddies; } S; typedef struct { plan_dft super; plan *cld1, *cld2; const S *solver; } P; /* Compute multi-dimensional DFT by applying the two cld plans (lower-rnk DFTs). */ static void apply(const plan *ego_, R *ri, R *ii, R *ro, R *io) { const P *ego = (const P *) ego_; plan_dft *cld1, *cld2; cld1 = (plan_dft *) ego->cld1; cld1->apply(ego->cld1, ri, ii, ro, io); cld2 = (plan_dft *) ego->cld2; cld2->apply(ego->cld2, ro, io, ro, io); } static void awake(plan *ego_, enum wakefulness wakefulness) { P *ego = (P *) ego_; X(plan_awake)(ego->cld1, wakefulness); X(plan_awake)(ego->cld2, wakefulness); } static void destroy(plan *ego_) { P *ego = (P *) ego_; X(plan_destroy_internal)(ego->cld2); X(plan_destroy_internal)(ego->cld1); } static void print(const plan *ego_, printer *p) { const P *ego = (const P *) ego_; const S *s = ego->solver; p->print(p, "(dft-rank>=2/%d%(%p%)%(%p%))", s->spltrnk, ego->cld1, ego->cld2); } static int picksplit(const S *ego, const tensor *sz, int *rp) { A(sz->rnk > 1); /* cannot split rnk <= 1 */ if (!X(pickdim)(ego->spltrnk, ego->buddies, ego->nbuddies, sz, 1, rp)) return 0; *rp += 1; /* convert from dim. index to rank */ if (*rp >= sz->rnk) /* split must reduce rank */ return 0; return 1; } static int applicable0(const solver *ego_, const problem *p_, int *rp) { const problem_dft *p = (const problem_dft *) p_; const S *ego = (const S *)ego_; return (1 && FINITE_RNK(p->sz->rnk) && FINITE_RNK(p->vecsz->rnk) && p->sz->rnk >= 2 && picksplit(ego, p->sz, rp) ); } /* TODO: revise this. */ static int applicable(const solver *ego_, const problem *p_, const planner *plnr, int *rp) { const S *ego = (const S *)ego_; const problem_dft *p = (const problem_dft *) p_; if (!applicable0(ego_, p_, rp)) return 0; if (NO_RANK_SPLITSP(plnr) && (ego->spltrnk != ego->buddies[0])) return 0; /* Heuristic: if the vector stride is greater than the transform sz, don't use (prefer to do the vector loop first with a vrank-geq1 plan). */ if (NO_UGLYP(plnr)) if (p->vecsz->rnk > 0 && X(tensor_min_stride)(p->vecsz) > X(tensor_max_index)(p->sz)) return 0; return 1; } static plan *mkplan(const solver *ego_, const problem *p_, planner *plnr) { const S *ego = (const S *) ego_; const problem_dft *p; P *pln; plan *cld1 = 0, *cld2 = 0; tensor *sz1, *sz2, *vecszi, *sz2i; int spltrnk; static const plan_adt padt = { X(dft_solve), awake, print, destroy }; if (!applicable(ego_, p_, plnr, &spltrnk)) return (plan *) 0; p = (const problem_dft *) p_; X(tensor_split)(p->sz, &sz1, spltrnk, &sz2); vecszi = X(tensor_copy_inplace)(p->vecsz, INPLACE_OS); sz2i = X(tensor_copy_inplace)(sz2, INPLACE_OS); cld1 = X(mkplan_d)(plnr, X(mkproblem_dft_d)(X(tensor_copy)(sz2), X(tensor_append)(p->vecsz, sz1), p->ri, p->ii, p->ro, p->io)); if (!cld1) goto nada; cld2 = X(mkplan_d)(plnr, X(mkproblem_dft_d)( X(tensor_copy_inplace)(sz1, INPLACE_OS), X(tensor_append)(vecszi, sz2i), p->ro, p->io, p->ro, p->io)); if (!cld2) goto nada; pln = MKPLAN_DFT(P, &padt, apply); pln->cld1 = cld1; pln->cld2 = cld2; pln->solver = ego; X(ops_add)(&cld1->ops, &cld2->ops, &pln->super.super.ops); X(tensor_destroy4)(sz1, sz2, vecszi, sz2i); return &(pln->super.super); nada: X(plan_destroy_internal)(cld2); X(plan_destroy_internal)(cld1); X(tensor_destroy4)(sz1, sz2, vecszi, sz2i); return (plan *) 0; } static solver *mksolver(int spltrnk, const int *buddies, int nbuddies) { static const solver_adt sadt = { PROBLEM_DFT, mkplan, 0 }; S *slv = MKSOLVER(S, &sadt); slv->spltrnk = spltrnk; slv->buddies = buddies; slv->nbuddies = nbuddies; return &(slv->super); } void X(dft_rank_geq2_register)(planner *p) { int i; static const int buddies[] = { 1, 0, -2 }; const int nbuddies = (int)(sizeof(buddies) / sizeof(buddies[0])); for (i = 0; i < nbuddies; ++i) REGISTER_SOLVER(p, mksolver(buddies[i], buddies, nbuddies)); /* FIXME: Should we try more buddies? Another possible variant is to swap cld1 and cld2 (or rather, to swap their problems; they are not interchangeable because cld2 must be in-place). In past versions of FFTW, however, I seem to recall that such rearrangements have made little or no difference. */ } fftw-3.3.3/dft/solve.c0000644000175000017500000000222312054410050011446 00000000000000/* * Copyright (c) 2003, 2007-11 Matteo Frigo * Copyright (c) 2003, 2007-11 Massachusetts Institute of Technology * * This program is free software; you can redistribute it and/or modify * it under the terms of the GNU General Public License as published by * the Free Software Foundation; either version 2 of the License, or * (at your option) any later version. * * This program is distributed in the hope that it will be useful, * but WITHOUT ANY WARRANTY; without even the implied warranty of * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the * GNU General Public License for more details. * * You should have received a copy of the GNU General Public License * along with this program; if not, write to the Free Software * Foundation, Inc., 51 Franklin Street, Fifth Floor, Boston, MA 02110-1301 USA * */ #include "dft.h" /* use the apply() operation for DFT problems */ void X(dft_solve)(const plan *ego_, const problem *p_) { const plan_dft *ego = (const plan_dft *) ego_; const problem_dft *p = (const problem_dft *) p_; ego->apply(ego_, UNTAINT(p->ri), UNTAINT(p->ii), UNTAINT(p->ro), UNTAINT(p->io)); } fftw-3.3.3/dft/vrank-geq1.c0000644000175000017500000001333712054410050012302 00000000000000/* * Copyright (c) 2003, 2007-11 Matteo Frigo * Copyright (c) 2003, 2007-11 Massachusetts Institute of Technology * * This program is free software; you can redistribute it and/or modify * it under the terms of the GNU General Public License as published by * the Free Software Foundation; either version 2 of the License, or * (at your option) any later version. * * This program is distributed in the hope that it will be useful, * but WITHOUT ANY WARRANTY; without even the implied warranty of * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the * GNU General Public License for more details. * * You should have received a copy of the GNU General Public License * along with this program; if not, write to the Free Software * Foundation, Inc., 51 Franklin Street, Fifth Floor, Boston, MA 02110-1301 USA * */ /* Plans for handling vector transform loops. These are *just* the loops, and rely on child plans for the actual DFTs. They form a wrapper around solvers that don't have apply functions for non-null vectors. vrank-geq1 plans also recursively handle the case of multi-dimensional vectors, obviating the need for most solvers to deal with this. We can also play games here, such as reordering the vector loops. Each vrank-geq1 plan reduces the vector rank by 1, picking out a dimension determined by the vecloop_dim field of the solver. */ #include "dft.h" typedef struct { solver super; int vecloop_dim; const int *buddies; int nbuddies; } S; typedef struct { plan_dft super; plan *cld; INT vl; INT ivs, ovs; const S *solver; } P; static void apply(const plan *ego_, R *ri, R *ii, R *ro, R *io) { const P *ego = (const P *) ego_; INT i, vl = ego->vl; INT ivs = ego->ivs, ovs = ego->ovs; dftapply cldapply = ((plan_dft *) ego->cld)->apply; for (i = 0; i < vl; ++i) { cldapply(ego->cld, ri + i * ivs, ii + i * ivs, ro + i * ovs, io + i * ovs); } } static void awake(plan *ego_, enum wakefulness wakefulness) { P *ego = (P *) ego_; X(plan_awake)(ego->cld, wakefulness); } static void destroy(plan *ego_) { P *ego = (P *) ego_; X(plan_destroy_internal)(ego->cld); } static void print(const plan *ego_, printer *p) { const P *ego = (const P *) ego_; const S *s = ego->solver; p->print(p, "(dft-vrank>=1-x%D/%d%(%p%))", ego->vl, s->vecloop_dim, ego->cld); } static int pickdim(const S *ego, const tensor *vecsz, int oop, int *dp) { return X(pickdim)(ego->vecloop_dim, ego->buddies, ego->nbuddies, vecsz, oop, dp); } static int applicable0(const solver *ego_, const problem *p_, int *dp) { const S *ego = (const S *) ego_; const problem_dft *p = (const problem_dft *) p_; return (1 && FINITE_RNK(p->vecsz->rnk) && p->vecsz->rnk > 0 /* do not bother looping over rank-0 problems, since they are handled via rdft */ && p->sz->rnk > 0 && pickdim(ego, p->vecsz, p->ri != p->ro, dp) ); } static int applicable(const solver *ego_, const problem *p_, const planner *plnr, int *dp) { const S *ego = (const S *)ego_; const problem_dft *p; if (!applicable0(ego_, p_, dp)) return 0; /* fftw2 behavior */ if (NO_VRANK_SPLITSP(plnr) && (ego->vecloop_dim != ego->buddies[0])) return 0; p = (const problem_dft *) p_; if (NO_UGLYP(plnr)) { /* Heuristic: if the transform is multi-dimensional, and the vector stride is less than the transform size, then we probably want to use a rank>=2 plan first in order to combine this vector with the transform-dimension vectors. */ { iodim *d = p->vecsz->dims + *dp; if (1 && p->sz->rnk > 1 && X(imin)(X(iabs)(d->is), X(iabs)(d->os)) < X(tensor_max_index)(p->sz) ) return 0; } if (NO_NONTHREADEDP(plnr)) return 0; /* prefer threaded version */ } return 1; } static plan *mkplan(const solver *ego_, const problem *p_, planner *plnr) { const S *ego = (const S *) ego_; const problem_dft *p; P *pln; plan *cld; int vdim; iodim *d; static const plan_adt padt = { X(dft_solve), awake, print, destroy }; if (!applicable(ego_, p_, plnr, &vdim)) return (plan *) 0; p = (const problem_dft *) p_; d = p->vecsz->dims + vdim; A(d->n > 1); cld = X(mkplan_d)(plnr, X(mkproblem_dft_d)( X(tensor_copy)(p->sz), X(tensor_copy_except)(p->vecsz, vdim), TAINT(p->ri, d->is), TAINT(p->ii, d->is), TAINT(p->ro, d->os), TAINT(p->io, d->os))); if (!cld) return (plan *) 0; pln = MKPLAN_DFT(P, &padt, apply); pln->cld = cld; pln->vl = d->n; pln->ivs = d->is; pln->ovs = d->os; pln->solver = ego; X(ops_zero)(&pln->super.super.ops); pln->super.super.ops.other = 3.14159; /* magic to prefer codelet loops */ X(ops_madd2)(pln->vl, &cld->ops, &pln->super.super.ops); if (p->sz->rnk != 1 || (p->sz->dims[0].n > 64)) pln->super.super.pcost = pln->vl * cld->pcost; return &(pln->super.super); } static solver *mksolver(int vecloop_dim, const int *buddies, int nbuddies) { static const solver_adt sadt = { PROBLEM_DFT, mkplan, 0 }; S *slv = MKSOLVER(S, &sadt); slv->vecloop_dim = vecloop_dim; slv->buddies = buddies; slv->nbuddies = nbuddies; return &(slv->super); } void X(dft_vrank_geq1_register)(planner *p) { int i; /* FIXME: Should we try other vecloop_dim values? */ static const int buddies[] = { 1, -1 }; const int nbuddies = (int)(sizeof(buddies) / sizeof(buddies[0])); for (i = 0; i < nbuddies; ++i) REGISTER_SOLVER(p, mksolver(buddies[i], buddies, nbuddies)); } fftw-3.3.3/dft/zero.c0000644000175000017500000000274112054410050011302 00000000000000/* * Copyright (c) 2003, 2007-11 Matteo Frigo * Copyright (c) 2003, 2007-11 Massachusetts Institute of Technology * * This program is free software; you can redistribute it and/or modify * it under the terms of the GNU General Public License as published by * the Free Software Foundation; either version 2 of the License, or * (at your option) any later version. * * This program is distributed in the hope that it will be useful, * but WITHOUT ANY WARRANTY; without even the implied warranty of * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the * GNU General Public License for more details. * * You should have received a copy of the GNU General Public License * along with this program; if not, write to the Free Software * Foundation, Inc., 51 Franklin Street, Fifth Floor, Boston, MA 02110-1301 USA * */ #include "dft.h" /* fill a complex array with zeros. */ static void recur(const iodim *dims, int rnk, R *ri, R *ii) { if (rnk == RNK_MINFTY) return; else if (rnk == 0) ri[0] = ii[0] = K(0.0); else if (rnk > 0) { INT i, n = dims[0].n; INT is = dims[0].is; if (rnk == 1) { /* this case is redundant but faster */ for (i = 0; i < n; ++i) ri[i * is] = ii[i * is] = K(0.0); } else { for (i = 0; i < n; ++i) recur(dims + 1, rnk - 1, ri + i * is, ii + i * is); } } } void X(dft_zerotens)(tensor *sz, R *ri, R *ii) { recur(sz->dims, sz->rnk, ri, ii); } fftw-3.3.3/dft/codelet-dft.h0000644000175000017500000000564612054410050012531 00000000000000/* * Copyright (c) 2003, 2007-11 Matteo Frigo * Copyright (c) 2003, 2007-11 Massachusetts Institute of Technology * * This program is free software; you can redistribute it and/or modify * it under the terms of the GNU General Public License as published by * the Free Software Foundation; either version 2 of the License, or * (at your option) any later version. * * This program is distributed in the hope that it will be useful, * but WITHOUT ANY WARRANTY; without even the implied warranty of * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the * GNU General Public License for more details. * * You should have received a copy of the GNU General Public License * along with this program; if not, write to the Free Software * Foundation, Inc., 51 Franklin Street, Fifth Floor, Boston, MA 02110-1301 USA * */ /* * This header file must include every file or define every * type or macro which is required to compile a codelet. */ #ifndef __DFT_CODELET_H__ #define __DFT_CODELET_H__ #include "ifftw.h" /************************************************************** * types of codelets **************************************************************/ /* DFT codelets */ typedef struct kdft_desc_s kdft_desc; typedef struct { int (*okp)( const kdft_desc *desc, const R *ri, const R *ii, const R *ro, const R *io, INT is, INT os, INT vl, INT ivs, INT ovs, const planner *plnr); INT vl; } kdft_genus; struct kdft_desc_s { INT sz; /* size of transform computed */ const char *nam; opcnt ops; const kdft_genus *genus; INT is; INT os; INT ivs; INT ovs; }; typedef void (*kdft) (const R *ri, const R *ii, R *ro, R *io, stride is, stride os, INT vl, INT ivs, INT ovs); void X(kdft_register)(planner *p, kdft codelet, const kdft_desc *desc); typedef struct ct_desc_s ct_desc; typedef struct { int (*okp)( const struct ct_desc_s *desc, const R *rio, const R *iio, INT rs, INT vs, INT m, INT mb, INT me, INT ms, const planner *plnr); INT vl; } ct_genus; struct ct_desc_s { INT radix; const char *nam; const tw_instr *tw; const ct_genus *genus; opcnt ops; INT rs; INT vs; INT ms; }; typedef void (*kdftw) (R *rioarray, R *iioarray, const R *W, stride ios, INT mb, INT me, INT ms); void X(kdft_dit_register)(planner *p, kdftw codelet, const ct_desc *desc); void X(kdft_dif_register)(planner *p, kdftw codelet, const ct_desc *desc); typedef void (*kdftwsq) (R *rioarray, R *iioarray, const R *W, stride is, stride vs, INT mb, INT me, INT ms); void X(kdft_difsq_register)(planner *p, kdftwsq codelet, const ct_desc *desc); extern const solvtab X(solvtab_dft_standard); extern const solvtab X(solvtab_dft_sse2); extern const solvtab X(solvtab_dft_avx); extern const solvtab X(solvtab_dft_altivec); extern const solvtab X(solvtab_dft_neon); #endif /* __DFT_CODELET_H__ */ fftw-3.3.3/dft/ct.h0000644000175000017500000000453512054410050010741 00000000000000/* * Copyright (c) 2003, 2007-11 Matteo Frigo * Copyright (c) 2003, 2007-11 Massachusetts Institute of Technology * * This program is free software; you can redistribute it and/or modify * it under the terms of the GNU General Public License as published by * the Free Software Foundation; either version 2 of the License, or * (at your option) any later version. * * This program is distributed in the hope that it will be useful, * but WITHOUT ANY WARRANTY; without even the implied warranty of * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the * GNU General Public License for more details. * * You should have received a copy of the GNU General Public License * along with this program; if not, write to the Free Software * Foundation, Inc., 51 Franklin Street, Fifth Floor, Boston, MA 02110-1301 USA * */ #include "dft.h" typedef void (*dftwapply)(const plan *ego, R *rio, R *iio); typedef struct ct_solver_s ct_solver; typedef plan *(*ct_mkinferior)(const ct_solver *ego, INT r, INT irs, INT ors, INT m, INT ms, INT v, INT ivs, INT ovs, INT mstart, INT mcount, R *rio, R *iio, planner *plnr); typedef int (*ct_force_vrecursion)(const ct_solver *ego, const problem_dft *p); typedef struct { plan super; dftwapply apply; } plan_dftw; extern plan *X(mkplan_dftw)(size_t size, const plan_adt *adt, dftwapply apply); #define MKPLAN_DFTW(type, adt, apply) \ (type *)X(mkplan_dftw)(sizeof(type), adt, apply) struct ct_solver_s { solver super; INT r; int dec; # define DECDIF 0 # define DECDIT 1 # define TRANSPOSE 2 ct_mkinferior mkcldw; ct_force_vrecursion force_vrecursionp; }; int X(ct_applicable)(const ct_solver *, const problem *, planner *); ct_solver *X(mksolver_ct)(size_t size, INT r, int dec, ct_mkinferior mkcldw, ct_force_vrecursion force_vrecursionp); extern ct_solver *(*X(mksolver_ct_hook))(size_t, INT, int, ct_mkinferior, ct_force_vrecursion); void X(regsolver_ct_directw)(planner *plnr, kdftw codelet, const ct_desc *desc, int dec); void X(regsolver_ct_directwbuf)(planner *plnr, kdftw codelet, const ct_desc *desc, int dec); solver *X(mksolver_ctsq)(kdftwsq codelet, const ct_desc *desc, int dec); void X(regsolver_ct_directwsq)(planner *plnr, kdftwsq codelet, const ct_desc *desc, int dec); fftw-3.3.3/dft/dft.h0000644000175000017500000000514012054410050011101 00000000000000/* * Copyright (c) 2003, 2007-11 Matteo Frigo * Copyright (c) 2003, 2007-11 Massachusetts Institute of Technology * * This program is free software; you can redistribute it and/or modify * it under the terms of the GNU General Public License as published by * the Free Software Foundation; either version 2 of the License, or * (at your option) any later version. * * This program is distributed in the hope that it will be useful, * but WITHOUT ANY WARRANTY; without even the implied warranty of * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the * GNU General Public License for more details. * * You should have received a copy of the GNU General Public License * along with this program; if not, write to the Free Software * Foundation, Inc., 51 Franklin Street, Fifth Floor, Boston, MA 02110-1301 USA * */ #ifndef __DFT_H__ #define __DFT_H__ #include "ifftw.h" #include "codelet-dft.h" #ifdef __cplusplus extern "C" { #endif /* __cplusplus */ /* problem.c: */ typedef struct { problem super; tensor *sz, *vecsz; R *ri, *ii, *ro, *io; } problem_dft; void X(dft_zerotens)(tensor *sz, R *ri, R *ii); problem *X(mkproblem_dft)(const tensor *sz, const tensor *vecsz, R *ri, R *ii, R *ro, R *io); problem *X(mkproblem_dft_d)(tensor *sz, tensor *vecsz, R *ri, R *ii, R *ro, R *io); /* solve.c: */ void X(dft_solve)(const plan *ego_, const problem *p_); /* plan.c: */ typedef void (*dftapply) (const plan *ego, R *ri, R *ii, R *ro, R *io); typedef struct { plan super; dftapply apply; } plan_dft; plan *X(mkplan_dft)(size_t size, const plan_adt *adt, dftapply apply); #define MKPLAN_DFT(type, adt, apply) \ (type *)X(mkplan_dft)(sizeof(type), adt, apply) /* various solvers */ solver *X(mksolver_dft_direct)(kdft k, const kdft_desc *desc); solver *X(mksolver_dft_directbuf)(kdft k, const kdft_desc *desc); void X(dft_rank0_register)(planner *p); void X(dft_rank_geq2_register)(planner *p); void X(dft_indirect_register)(planner *p); void X(dft_indirect_transpose_register)(planner *p); void X(dft_vrank_geq1_register)(planner *p); void X(dft_vrank2_transpose_register)(planner *p); void X(dft_vrank3_transpose_register)(planner *p); void X(dft_buffered_register)(planner *p); void X(dft_generic_register)(planner *p); void X(dft_rader_register)(planner *p); void X(dft_bluestein_register)(planner *p); void X(dft_nop_register)(planner *p); void X(ct_generic_register)(planner *p); void X(ct_genericbuf_register)(planner *p); /* configurations */ void X(dft_conf_standard)(planner *p); #ifdef __cplusplus } /* extern "C" */ #endif /* __cplusplus */ #endif /* __DFT_H__ */ fftw-3.3.3/dft/scalar/0000755000175000017500000000000012054411106011503 500000000000000fftw-3.3.3/dft/scalar/Makefile.am0000644000175000017500000000024512054410050013455 00000000000000AM_CPPFLAGS = -I$(top_srcdir)/kernel -I$(top_srcdir)/dft SUBDIRS=codelets noinst_LTLIBRARIES = libdft_scalar.la libdft_scalar_la_SOURCES = n.c t.c f.h n.h q.h t.h fftw-3.3.3/dft/scalar/Makefile.in0000644000175000017500000005221112054410177013500 00000000000000# Makefile.in generated by automake 1.11.6 from Makefile.am. # @configure_input@ # Copyright (C) 1994, 1995, 1996, 1997, 1998, 1999, 2000, 2001, 2002, # 2003, 2004, 2005, 2006, 2007, 2008, 2009, 2010, 2011 Free Software # Foundation, Inc. # This Makefile.in is free software; the Free Software Foundation # gives unlimited permission to copy and/or distribute it, # with or without modifications, as long as this notice is preserved. # This program is distributed in the hope that it will be useful, # but WITHOUT ANY WARRANTY, to the extent permitted by law; without # even the implied warranty of MERCHANTABILITY or FITNESS FOR A # PARTICULAR PURPOSE. @SET_MAKE@ VPATH = @srcdir@ am__make_dryrun = \ { \ am__dry=no; \ case $$MAKEFLAGS in \ *\\[\ \ ]*) \ echo 'am--echo: ; @echo "AM" OK' | $(MAKE) -f - 2>/dev/null \ | grep '^AM OK$$' >/dev/null || am__dry=yes;; \ *) \ for am__flg in $$MAKEFLAGS; do \ case $$am__flg in \ *=*|--*) ;; \ *n*) am__dry=yes; break;; \ esac; \ done;; \ esac; \ test $$am__dry = yes; \ } pkgdatadir = $(datadir)/@PACKAGE@ pkgincludedir = $(includedir)/@PACKAGE@ pkglibdir = $(libdir)/@PACKAGE@ pkglibexecdir = $(libexecdir)/@PACKAGE@ am__cd = CDPATH="$${ZSH_VERSION+.}$(PATH_SEPARATOR)" && cd install_sh_DATA = $(install_sh) -c -m 644 install_sh_PROGRAM = $(install_sh) -c install_sh_SCRIPT = $(install_sh) -c INSTALL_HEADER = $(INSTALL_DATA) transform = $(program_transform_name) NORMAL_INSTALL = : PRE_INSTALL = : POST_INSTALL = : NORMAL_UNINSTALL = : PRE_UNINSTALL = : POST_UNINSTALL = : build_triplet = @build@ host_triplet = @host@ subdir = dft/scalar DIST_COMMON = $(srcdir)/Makefile.am $(srcdir)/Makefile.in ACLOCAL_M4 = $(top_srcdir)/aclocal.m4 am__aclocal_m4_deps = $(top_srcdir)/m4/acx_mpi.m4 \ $(top_srcdir)/m4/acx_pthread.m4 \ $(top_srcdir)/m4/ax_cc_maxopt.m4 \ $(top_srcdir)/m4/ax_check_compiler_flags.m4 \ $(top_srcdir)/m4/ax_compiler_vendor.m4 \ $(top_srcdir)/m4/ax_gcc_aligns_stack.m4 \ $(top_srcdir)/m4/ax_gcc_version.m4 \ $(top_srcdir)/m4/ax_openmp.m4 $(top_srcdir)/m4/libtool.m4 \ $(top_srcdir)/m4/ltoptions.m4 $(top_srcdir)/m4/ltsugar.m4 \ $(top_srcdir)/m4/ltversion.m4 $(top_srcdir)/m4/lt~obsolete.m4 \ $(top_srcdir)/configure.ac am__configure_deps = $(am__aclocal_m4_deps) $(CONFIGURE_DEPENDENCIES) \ $(ACLOCAL_M4) mkinstalldirs = $(install_sh) -d CONFIG_HEADER = $(top_builddir)/config.h CONFIG_CLEAN_FILES = CONFIG_CLEAN_VPATH_FILES = LTLIBRARIES = $(noinst_LTLIBRARIES) libdft_scalar_la_LIBADD = am_libdft_scalar_la_OBJECTS = n.lo t.lo libdft_scalar_la_OBJECTS = $(am_libdft_scalar_la_OBJECTS) DEFAULT_INCLUDES = -I.@am__isrc@ -I$(top_builddir) depcomp = $(SHELL) $(top_srcdir)/depcomp am__depfiles_maybe = depfiles am__mv = mv -f COMPILE = $(CC) $(DEFS) $(DEFAULT_INCLUDES) $(INCLUDES) $(AM_CPPFLAGS) \ $(CPPFLAGS) $(AM_CFLAGS) $(CFLAGS) LTCOMPILE = $(LIBTOOL) --tag=CC $(AM_LIBTOOLFLAGS) $(LIBTOOLFLAGS) \ --mode=compile $(CC) $(DEFS) $(DEFAULT_INCLUDES) $(INCLUDES) \ $(AM_CPPFLAGS) $(CPPFLAGS) $(AM_CFLAGS) $(CFLAGS) CCLD = $(CC) LINK = $(LIBTOOL) --tag=CC $(AM_LIBTOOLFLAGS) $(LIBTOOLFLAGS) \ --mode=link $(CCLD) $(AM_CFLAGS) $(CFLAGS) $(AM_LDFLAGS) \ $(LDFLAGS) -o $@ SOURCES = $(libdft_scalar_la_SOURCES) DIST_SOURCES = $(libdft_scalar_la_SOURCES) RECURSIVE_TARGETS = all-recursive check-recursive dvi-recursive \ html-recursive info-recursive install-data-recursive \ install-dvi-recursive install-exec-recursive \ install-html-recursive install-info-recursive \ install-pdf-recursive install-ps-recursive install-recursive \ installcheck-recursive installdirs-recursive pdf-recursive \ ps-recursive uninstall-recursive am__can_run_installinfo = \ case $$AM_UPDATE_INFO_DIR in \ n|no|NO) false;; \ *) (install-info --version) >/dev/null 2>&1;; \ esac RECURSIVE_CLEAN_TARGETS = mostlyclean-recursive clean-recursive \ distclean-recursive maintainer-clean-recursive AM_RECURSIVE_TARGETS = $(RECURSIVE_TARGETS:-recursive=) \ $(RECURSIVE_CLEAN_TARGETS:-recursive=) tags TAGS ctags CTAGS \ distdir ETAGS = etags CTAGS = ctags DIST_SUBDIRS = $(SUBDIRS) DISTFILES = $(DIST_COMMON) $(DIST_SOURCES) $(TEXINFOS) $(EXTRA_DIST) am__relativize = \ dir0=`pwd`; \ sed_first='s,^\([^/]*\)/.*$$,\1,'; \ sed_rest='s,^[^/]*/*,,'; \ sed_last='s,^.*/\([^/]*\)$$,\1,'; \ sed_butlast='s,/*[^/]*$$,,'; \ while test -n "$$dir1"; do \ first=`echo "$$dir1" | sed -e "$$sed_first"`; \ if test "$$first" != "."; then \ if test "$$first" = ".."; then \ dir2=`echo "$$dir0" | sed -e "$$sed_last"`/"$$dir2"; \ dir0=`echo "$$dir0" | sed -e "$$sed_butlast"`; \ else \ first2=`echo "$$dir2" | sed -e "$$sed_first"`; \ if test "$$first2" = "$$first"; then \ dir2=`echo "$$dir2" | sed -e "$$sed_rest"`; \ else \ dir2="../$$dir2"; \ fi; \ dir0="$$dir0"/"$$first"; \ fi; \ fi; \ dir1=`echo "$$dir1" | sed -e "$$sed_rest"`; \ done; \ reldir="$$dir2" ACLOCAL = @ACLOCAL@ ALLOCA = @ALLOCA@ ALTIVEC_CFLAGS = @ALTIVEC_CFLAGS@ AMTAR = @AMTAR@ AR = @AR@ AS = @AS@ AUTOCONF = @AUTOCONF@ AUTOHEADER = @AUTOHEADER@ AUTOMAKE = @AUTOMAKE@ AVX_CFLAGS = @AVX_CFLAGS@ AWK = @AWK@ CC = @CC@ CCDEPMODE = @CCDEPMODE@ CFLAGS = @CFLAGS@ CHECK_PL_OPTS = @CHECK_PL_OPTS@ CPP = @CPP@ CPPFLAGS = @CPPFLAGS@ CYGPATH_W = @CYGPATH_W@ C_FFTW_R2R_KIND = @C_FFTW_R2R_KIND@ C_MPI_FINT = @C_MPI_FINT@ DEFS = @DEFS@ DEPDIR = @DEPDIR@ DLLTOOL = @DLLTOOL@ DSYMUTIL = @DSYMUTIL@ DUMPBIN = @DUMPBIN@ ECHO_C = @ECHO_C@ ECHO_N = @ECHO_N@ ECHO_T = @ECHO_T@ EGREP = @EGREP@ EXEEXT = @EXEEXT@ F77 = @F77@ FFLAGS = @FFLAGS@ FGREP = @FGREP@ FLIBS = @FLIBS@ GREP = @GREP@ INSTALL = @INSTALL@ INSTALL_DATA = @INSTALL_DATA@ INSTALL_PROGRAM = @INSTALL_PROGRAM@ INSTALL_SCRIPT = @INSTALL_SCRIPT@ INSTALL_STRIP_PROGRAM = @INSTALL_STRIP_PROGRAM@ LD = @LD@ LDFLAGS = @LDFLAGS@ LIBOBJS = @LIBOBJS@ LIBQUADMATH = @LIBQUADMATH@ LIBS = @LIBS@ LIBTOOL = @LIBTOOL@ LIPO = @LIPO@ LN_S = @LN_S@ LTLIBOBJS = @LTLIBOBJS@ MAINT = @MAINT@ MAKEINFO = @MAKEINFO@ MANIFEST_TOOL = @MANIFEST_TOOL@ MKDIR_P = @MKDIR_P@ MPICC = @MPICC@ MPILIBS = @MPILIBS@ MPIRUN = @MPIRUN@ NEON_CFLAGS = @NEON_CFLAGS@ NM = @NM@ NMEDIT = @NMEDIT@ OBJDUMP = @OBJDUMP@ OBJEXT = @OBJEXT@ OCAMLBUILD = @OCAMLBUILD@ OPENMP_CFLAGS = @OPENMP_CFLAGS@ OTOOL = @OTOOL@ OTOOL64 = @OTOOL64@ PACKAGE = @PACKAGE@ PACKAGE_BUGREPORT = @PACKAGE_BUGREPORT@ PACKAGE_NAME = @PACKAGE_NAME@ PACKAGE_STRING = @PACKAGE_STRING@ PACKAGE_TARNAME = @PACKAGE_TARNAME@ PACKAGE_URL = @PACKAGE_URL@ PACKAGE_VERSION = @PACKAGE_VERSION@ PATH_SEPARATOR = @PATH_SEPARATOR@ POW_LIB = @POW_LIB@ PRECISION = @PRECISION@ PREC_SUFFIX = @PREC_SUFFIX@ PTHREAD_CC = @PTHREAD_CC@ PTHREAD_CFLAGS = @PTHREAD_CFLAGS@ PTHREAD_LIBS = @PTHREAD_LIBS@ RANLIB = @RANLIB@ SED = @SED@ SET_MAKE = @SET_MAKE@ SHARED_VERSION_INFO = @SHARED_VERSION_INFO@ SHELL = @SHELL@ SSE2_CFLAGS = @SSE2_CFLAGS@ STACK_ALIGN_CFLAGS = @STACK_ALIGN_CFLAGS@ STRIP = @STRIP@ THREADLIBS = @THREADLIBS@ VERSION = @VERSION@ abs_builddir = @abs_builddir@ abs_srcdir = @abs_srcdir@ abs_top_builddir = @abs_top_builddir@ abs_top_srcdir = @abs_top_srcdir@ ac_ct_AR = @ac_ct_AR@ ac_ct_CC = @ac_ct_CC@ ac_ct_DUMPBIN = @ac_ct_DUMPBIN@ ac_ct_F77 = @ac_ct_F77@ acx_pthread_config = @acx_pthread_config@ am__include = @am__include@ am__leading_dot = @am__leading_dot@ am__quote = @am__quote@ am__tar = @am__tar@ am__untar = @am__untar@ bindir = @bindir@ build = @build@ build_alias = @build_alias@ build_cpu = @build_cpu@ build_os = @build_os@ build_vendor = @build_vendor@ builddir = @builddir@ datadir = @datadir@ datarootdir = @datarootdir@ docdir = @docdir@ dvidir = @dvidir@ exec_prefix = @exec_prefix@ host = @host@ host_alias = @host_alias@ host_cpu = @host_cpu@ host_os = @host_os@ host_vendor = @host_vendor@ htmldir = @htmldir@ includedir = @includedir@ infodir = @infodir@ install_sh = @install_sh@ libdir = @libdir@ libexecdir = @libexecdir@ localedir = @localedir@ localstatedir = @localstatedir@ mandir = @mandir@ mkdir_p = @mkdir_p@ oldincludedir = @oldincludedir@ pdfdir = @pdfdir@ prefix = @prefix@ program_transform_name = @program_transform_name@ psdir = @psdir@ sbindir = @sbindir@ sharedstatedir = @sharedstatedir@ srcdir = @srcdir@ sysconfdir = @sysconfdir@ target_alias = @target_alias@ top_build_prefix = @top_build_prefix@ top_builddir = @top_builddir@ top_srcdir = @top_srcdir@ AM_CPPFLAGS = -I$(top_srcdir)/kernel -I$(top_srcdir)/dft SUBDIRS = codelets noinst_LTLIBRARIES = libdft_scalar.la libdft_scalar_la_SOURCES = n.c t.c f.h n.h q.h t.h all: all-recursive .SUFFIXES: .SUFFIXES: .c .lo .o .obj $(srcdir)/Makefile.in: @MAINTAINER_MODE_TRUE@ $(srcdir)/Makefile.am $(am__configure_deps) @for dep in $?; do \ case '$(am__configure_deps)' in \ *$$dep*) \ ( cd $(top_builddir) && $(MAKE) $(AM_MAKEFLAGS) am--refresh ) \ && { if test -f $@; then exit 0; else break; fi; }; \ exit 1;; \ esac; \ done; \ echo ' cd $(top_srcdir) && $(AUTOMAKE) --gnu dft/scalar/Makefile'; \ $(am__cd) $(top_srcdir) && \ $(AUTOMAKE) --gnu dft/scalar/Makefile .PRECIOUS: Makefile Makefile: $(srcdir)/Makefile.in $(top_builddir)/config.status @case '$?' in \ *config.status*) \ cd $(top_builddir) && $(MAKE) $(AM_MAKEFLAGS) am--refresh;; \ *) \ echo ' cd $(top_builddir) && $(SHELL) ./config.status $(subdir)/$@ $(am__depfiles_maybe)'; \ cd $(top_builddir) && $(SHELL) ./config.status $(subdir)/$@ $(am__depfiles_maybe);; \ esac; $(top_builddir)/config.status: $(top_srcdir)/configure $(CONFIG_STATUS_DEPENDENCIES) cd $(top_builddir) && $(MAKE) $(AM_MAKEFLAGS) am--refresh $(top_srcdir)/configure: @MAINTAINER_MODE_TRUE@ $(am__configure_deps) cd $(top_builddir) && $(MAKE) $(AM_MAKEFLAGS) am--refresh $(ACLOCAL_M4): @MAINTAINER_MODE_TRUE@ $(am__aclocal_m4_deps) cd $(top_builddir) && $(MAKE) $(AM_MAKEFLAGS) am--refresh $(am__aclocal_m4_deps): clean-noinstLTLIBRARIES: -test -z "$(noinst_LTLIBRARIES)" || rm -f $(noinst_LTLIBRARIES) @list='$(noinst_LTLIBRARIES)'; for p in $$list; do \ dir="`echo $$p | sed -e 's|/[^/]*$$||'`"; \ test "$$dir" != "$$p" || dir=.; \ echo "rm -f \"$${dir}/so_locations\""; \ rm -f "$${dir}/so_locations"; \ done libdft_scalar.la: $(libdft_scalar_la_OBJECTS) $(libdft_scalar_la_DEPENDENCIES) $(EXTRA_libdft_scalar_la_DEPENDENCIES) $(LINK) $(libdft_scalar_la_OBJECTS) $(libdft_scalar_la_LIBADD) $(LIBS) mostlyclean-compile: -rm -f *.$(OBJEXT) distclean-compile: -rm -f *.tab.c @AMDEP_TRUE@@am__include@ @am__quote@./$(DEPDIR)/n.Plo@am__quote@ @AMDEP_TRUE@@am__include@ @am__quote@./$(DEPDIR)/t.Plo@am__quote@ .c.o: @am__fastdepCC_TRUE@ $(COMPILE) -MT $@ -MD -MP -MF $(DEPDIR)/$*.Tpo -c -o $@ $< @am__fastdepCC_TRUE@ $(am__mv) $(DEPDIR)/$*.Tpo $(DEPDIR)/$*.Po @AMDEP_TRUE@@am__fastdepCC_FALSE@ source='$<' object='$@' libtool=no @AMDEPBACKSLASH@ @AMDEP_TRUE@@am__fastdepCC_FALSE@ DEPDIR=$(DEPDIR) $(CCDEPMODE) $(depcomp) @AMDEPBACKSLASH@ @am__fastdepCC_FALSE@ $(COMPILE) -c $< .c.obj: @am__fastdepCC_TRUE@ $(COMPILE) -MT $@ -MD -MP -MF $(DEPDIR)/$*.Tpo -c -o $@ `$(CYGPATH_W) '$<'` @am__fastdepCC_TRUE@ $(am__mv) $(DEPDIR)/$*.Tpo $(DEPDIR)/$*.Po @AMDEP_TRUE@@am__fastdepCC_FALSE@ source='$<' object='$@' libtool=no @AMDEPBACKSLASH@ @AMDEP_TRUE@@am__fastdepCC_FALSE@ DEPDIR=$(DEPDIR) $(CCDEPMODE) $(depcomp) @AMDEPBACKSLASH@ @am__fastdepCC_FALSE@ $(COMPILE) -c `$(CYGPATH_W) '$<'` .c.lo: @am__fastdepCC_TRUE@ $(LTCOMPILE) -MT $@ -MD -MP -MF $(DEPDIR)/$*.Tpo -c -o $@ $< @am__fastdepCC_TRUE@ $(am__mv) $(DEPDIR)/$*.Tpo $(DEPDIR)/$*.Plo @AMDEP_TRUE@@am__fastdepCC_FALSE@ source='$<' object='$@' libtool=yes @AMDEPBACKSLASH@ @AMDEP_TRUE@@am__fastdepCC_FALSE@ DEPDIR=$(DEPDIR) $(CCDEPMODE) $(depcomp) @AMDEPBACKSLASH@ @am__fastdepCC_FALSE@ $(LTCOMPILE) -c -o $@ $< mostlyclean-libtool: -rm -f *.lo clean-libtool: -rm -rf .libs _libs # This directory's subdirectories are mostly independent; you can cd # into them and run `make' without going through this Makefile. # To change the values of `make' variables: instead of editing Makefiles, # (1) if the variable is set in `config.status', edit `config.status' # (which will cause the Makefiles to be regenerated when you run `make'); # (2) otherwise, pass the desired values on the `make' command line. $(RECURSIVE_TARGETS): @fail= failcom='exit 1'; \ for f in x $$MAKEFLAGS; do \ case $$f in \ *=* | --[!k]*);; \ *k*) failcom='fail=yes';; \ esac; \ done; \ dot_seen=no; \ target=`echo $@ | sed s/-recursive//`; \ list='$(SUBDIRS)'; for subdir in $$list; do \ echo "Making $$target in $$subdir"; \ if test "$$subdir" = "."; then \ dot_seen=yes; \ local_target="$$target-am"; \ else \ local_target="$$target"; \ fi; \ ($(am__cd) $$subdir && $(MAKE) $(AM_MAKEFLAGS) $$local_target) \ || eval $$failcom; \ done; \ if test "$$dot_seen" = "no"; then \ $(MAKE) $(AM_MAKEFLAGS) "$$target-am" || exit 1; \ fi; test -z "$$fail" $(RECURSIVE_CLEAN_TARGETS): @fail= failcom='exit 1'; \ for f in x $$MAKEFLAGS; do \ case $$f in \ *=* | --[!k]*);; \ *k*) failcom='fail=yes';; \ esac; \ done; \ dot_seen=no; \ case "$@" in \ distclean-* | maintainer-clean-*) list='$(DIST_SUBDIRS)' ;; \ *) list='$(SUBDIRS)' ;; \ esac; \ rev=''; for subdir in $$list; do \ if test "$$subdir" = "."; then :; else \ rev="$$subdir $$rev"; \ fi; \ done; \ rev="$$rev ."; \ target=`echo $@ | sed s/-recursive//`; \ for subdir in $$rev; do \ echo "Making $$target in $$subdir"; \ if test "$$subdir" = "."; then \ local_target="$$target-am"; \ else \ local_target="$$target"; \ fi; \ ($(am__cd) $$subdir && $(MAKE) $(AM_MAKEFLAGS) $$local_target) \ || eval $$failcom; \ done && test -z "$$fail" tags-recursive: list='$(SUBDIRS)'; for subdir in $$list; do \ test "$$subdir" = . || ($(am__cd) $$subdir && $(MAKE) $(AM_MAKEFLAGS) tags); \ done ctags-recursive: list='$(SUBDIRS)'; for subdir in $$list; do \ test "$$subdir" = . || ($(am__cd) $$subdir && $(MAKE) $(AM_MAKEFLAGS) ctags); \ done ID: $(HEADERS) $(SOURCES) $(LISP) $(TAGS_FILES) list='$(SOURCES) $(HEADERS) $(LISP) $(TAGS_FILES)'; \ unique=`for i in $$list; do \ if test -f "$$i"; then echo $$i; else echo $(srcdir)/$$i; fi; \ done | \ $(AWK) '{ files[$$0] = 1; nonempty = 1; } \ END { if (nonempty) { for (i in files) print i; }; }'`; \ mkid -fID $$unique tags: TAGS TAGS: tags-recursive $(HEADERS) $(SOURCES) $(TAGS_DEPENDENCIES) \ $(TAGS_FILES) $(LISP) set x; \ here=`pwd`; \ if ($(ETAGS) --etags-include --version) >/dev/null 2>&1; then \ include_option=--etags-include; \ empty_fix=.; \ else \ include_option=--include; \ empty_fix=; \ fi; \ list='$(SUBDIRS)'; for subdir in $$list; do \ if test "$$subdir" = .; then :; else \ test ! -f $$subdir/TAGS || \ set "$$@" "$$include_option=$$here/$$subdir/TAGS"; \ fi; \ done; \ list='$(SOURCES) $(HEADERS) $(LISP) $(TAGS_FILES)'; \ unique=`for i in $$list; do \ if test -f "$$i"; then echo $$i; else echo $(srcdir)/$$i; fi; \ done | \ $(AWK) '{ files[$$0] = 1; nonempty = 1; } \ END { if (nonempty) { for (i in files) print i; }; }'`; \ shift; \ if test -z "$(ETAGS_ARGS)$$*$$unique"; then :; else \ test -n "$$unique" || unique=$$empty_fix; \ if test $$# -gt 0; then \ $(ETAGS) $(ETAGSFLAGS) $(AM_ETAGSFLAGS) $(ETAGS_ARGS) \ "$$@" $$unique; \ else \ $(ETAGS) $(ETAGSFLAGS) $(AM_ETAGSFLAGS) $(ETAGS_ARGS) \ $$unique; \ fi; \ fi ctags: CTAGS CTAGS: ctags-recursive $(HEADERS) $(SOURCES) $(TAGS_DEPENDENCIES) \ $(TAGS_FILES) $(LISP) list='$(SOURCES) $(HEADERS) $(LISP) $(TAGS_FILES)'; \ unique=`for i in $$list; do \ if test -f "$$i"; then echo $$i; else echo $(srcdir)/$$i; fi; \ done | \ $(AWK) '{ files[$$0] = 1; nonempty = 1; } \ END { if (nonempty) { for (i in files) print i; }; }'`; \ test -z "$(CTAGS_ARGS)$$unique" \ || $(CTAGS) $(CTAGSFLAGS) $(AM_CTAGSFLAGS) $(CTAGS_ARGS) \ $$unique GTAGS: here=`$(am__cd) $(top_builddir) && pwd` \ && $(am__cd) $(top_srcdir) \ && gtags -i $(GTAGS_ARGS) "$$here" distclean-tags: -rm -f TAGS ID GTAGS GRTAGS GSYMS GPATH tags distdir: $(DISTFILES) @srcdirstrip=`echo "$(srcdir)" | sed 's/[].[^$$\\*]/\\\\&/g'`; \ topsrcdirstrip=`echo "$(top_srcdir)" | sed 's/[].[^$$\\*]/\\\\&/g'`; \ list='$(DISTFILES)'; \ dist_files=`for file in $$list; do echo $$file; done | \ sed -e "s|^$$srcdirstrip/||;t" \ -e "s|^$$topsrcdirstrip/|$(top_builddir)/|;t"`; \ case $$dist_files in \ */*) $(MKDIR_P) `echo "$$dist_files" | \ sed '/\//!d;s|^|$(distdir)/|;s,/[^/]*$$,,' | \ sort -u` ;; \ esac; \ for file in $$dist_files; do \ if test -f $$file || test -d $$file; then d=.; else d=$(srcdir); fi; \ if test -d $$d/$$file; then \ dir=`echo "/$$file" | sed -e 's,/[^/]*$$,,'`; \ if test -d "$(distdir)/$$file"; then \ find "$(distdir)/$$file" -type d ! -perm -700 -exec chmod u+rwx {} \;; \ fi; \ if test -d $(srcdir)/$$file && test $$d != $(srcdir); then \ cp -fpR $(srcdir)/$$file "$(distdir)$$dir" || exit 1; \ find "$(distdir)/$$file" -type d ! -perm -700 -exec chmod u+rwx {} \;; \ fi; \ cp -fpR $$d/$$file "$(distdir)$$dir" || exit 1; \ else \ test -f "$(distdir)/$$file" \ || cp -p $$d/$$file "$(distdir)/$$file" \ || exit 1; \ fi; \ done @list='$(DIST_SUBDIRS)'; for subdir in $$list; do \ if test "$$subdir" = .; then :; else \ $(am__make_dryrun) \ || test -d "$(distdir)/$$subdir" \ || $(MKDIR_P) "$(distdir)/$$subdir" \ || exit 1; \ dir1=$$subdir; dir2="$(distdir)/$$subdir"; \ $(am__relativize); \ new_distdir=$$reldir; \ dir1=$$subdir; dir2="$(top_distdir)"; \ $(am__relativize); \ new_top_distdir=$$reldir; \ echo " (cd $$subdir && $(MAKE) $(AM_MAKEFLAGS) top_distdir="$$new_top_distdir" distdir="$$new_distdir" \\"; \ echo " am__remove_distdir=: am__skip_length_check=: am__skip_mode_fix=: distdir)"; \ ($(am__cd) $$subdir && \ $(MAKE) $(AM_MAKEFLAGS) \ top_distdir="$$new_top_distdir" \ distdir="$$new_distdir" \ am__remove_distdir=: \ am__skip_length_check=: \ am__skip_mode_fix=: \ distdir) \ || exit 1; \ fi; \ done check-am: all-am check: check-recursive all-am: Makefile $(LTLIBRARIES) installdirs: installdirs-recursive installdirs-am: install: install-recursive install-exec: install-exec-recursive install-data: install-data-recursive uninstall: uninstall-recursive install-am: all-am @$(MAKE) $(AM_MAKEFLAGS) install-exec-am install-data-am installcheck: installcheck-recursive install-strip: if test -z '$(STRIP)'; then \ $(MAKE) $(AM_MAKEFLAGS) INSTALL_PROGRAM="$(INSTALL_STRIP_PROGRAM)" \ install_sh_PROGRAM="$(INSTALL_STRIP_PROGRAM)" INSTALL_STRIP_FLAG=-s \ install; \ else \ $(MAKE) $(AM_MAKEFLAGS) INSTALL_PROGRAM="$(INSTALL_STRIP_PROGRAM)" \ install_sh_PROGRAM="$(INSTALL_STRIP_PROGRAM)" INSTALL_STRIP_FLAG=-s \ "INSTALL_PROGRAM_ENV=STRIPPROG='$(STRIP)'" install; \ fi mostlyclean-generic: clean-generic: distclean-generic: -test -z "$(CONFIG_CLEAN_FILES)" || rm -f $(CONFIG_CLEAN_FILES) -test . = "$(srcdir)" || test -z "$(CONFIG_CLEAN_VPATH_FILES)" || rm -f $(CONFIG_CLEAN_VPATH_FILES) maintainer-clean-generic: @echo "This command is intended for maintainers to use" @echo "it deletes files that may require special tools to rebuild." clean: clean-recursive clean-am: clean-generic clean-libtool clean-noinstLTLIBRARIES \ mostlyclean-am distclean: distclean-recursive -rm -rf ./$(DEPDIR) -rm -f Makefile distclean-am: clean-am distclean-compile distclean-generic \ distclean-tags dvi: dvi-recursive dvi-am: html: html-recursive html-am: info: info-recursive info-am: install-data-am: install-dvi: install-dvi-recursive install-dvi-am: install-exec-am: install-html: install-html-recursive install-html-am: install-info: install-info-recursive install-info-am: install-man: install-pdf: install-pdf-recursive install-pdf-am: install-ps: install-ps-recursive install-ps-am: installcheck-am: maintainer-clean: maintainer-clean-recursive -rm -rf ./$(DEPDIR) -rm -f Makefile maintainer-clean-am: distclean-am maintainer-clean-generic mostlyclean: mostlyclean-recursive mostlyclean-am: mostlyclean-compile mostlyclean-generic \ mostlyclean-libtool pdf: pdf-recursive pdf-am: ps: ps-recursive ps-am: uninstall-am: .MAKE: $(RECURSIVE_CLEAN_TARGETS) $(RECURSIVE_TARGETS) ctags-recursive \ install-am install-strip tags-recursive .PHONY: $(RECURSIVE_CLEAN_TARGETS) $(RECURSIVE_TARGETS) CTAGS GTAGS \ all all-am check check-am clean clean-generic clean-libtool \ clean-noinstLTLIBRARIES ctags ctags-recursive distclean \ distclean-compile distclean-generic distclean-libtool \ distclean-tags distdir dvi dvi-am html html-am info info-am \ install install-am install-data install-data-am install-dvi \ install-dvi-am install-exec install-exec-am install-html \ install-html-am install-info install-info-am install-man \ install-pdf install-pdf-am install-ps install-ps-am \ install-strip installcheck installcheck-am installdirs \ installdirs-am maintainer-clean maintainer-clean-generic \ mostlyclean mostlyclean-compile mostlyclean-generic \ mostlyclean-libtool pdf pdf-am ps ps-am tags tags-recursive \ uninstall uninstall-am # Tell versions [3.59,3.63) of GNU make to not export all variables. # Otherwise a system limit (for SysV at least) may be exceeded. .NOEXPORT: fftw-3.3.3/dft/scalar/n.c0000644000175000017500000000251512054410050012024 00000000000000/* * Copyright (c) 2003, 2007-11 Matteo Frigo * Copyright (c) 2003, 2007-11 Massachusetts Institute of Technology * * This program is free software; you can redistribute it and/or modify * it under the terms of the GNU General Public License as published by * the Free Software Foundation; either version 2 of the License, or * (at your option) any later version. * * This program is distributed in the hope that it will be useful, * but WITHOUT ANY WARRANTY; without even the implied warranty of * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the * GNU General Public License for more details. * * You should have received a copy of the GNU General Public License * along with this program; if not, write to the Free Software * Foundation, Inc., 51 Franklin Street, Fifth Floor, Boston, MA 02110-1301 USA * */ #include "codelet-dft.h" #include "n.h" static int okp(const kdft_desc *d, const R *ri, const R *ii, const R *ro, const R *io, INT is, INT os, INT vl, INT ivs, INT ovs, const planner *plnr) { UNUSED(ri); UNUSED(ii); UNUSED(ro); UNUSED(io); UNUSED(vl); UNUSED(plnr); return (1 && (!d->is || (d->is == is)) && (!d->os || (d->os == os)) && (!d->ivs || (d->ivs == ivs)) && (!d->ovs || (d->ovs == ovs)) ); } const kdft_genus GENUS = { okp, 1 }; fftw-3.3.3/dft/scalar/t.c0000644000175000017500000000240612054410050012031 00000000000000/* * Copyright (c) 2003, 2007-11 Matteo Frigo * Copyright (c) 2003, 2007-11 Massachusetts Institute of Technology * * This program is free software; you can redistribute it and/or modify * it under the terms of the GNU General Public License as published by * the Free Software Foundation; either version 2 of the License, or * (at your option) any later version. * * This program is distributed in the hope that it will be useful, * but WITHOUT ANY WARRANTY; without even the implied warranty of * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the * GNU General Public License for more details. * * You should have received a copy of the GNU General Public License * along with this program; if not, write to the Free Software * Foundation, Inc., 51 Franklin Street, Fifth Floor, Boston, MA 02110-1301 USA * */ #include "codelet-dft.h" #include "t.h" static int okp(const ct_desc *d, const R *rio, const R *iio, INT rs, INT vs, INT m, INT mb, INT me, INT ms, const planner *plnr) { UNUSED(rio); UNUSED(iio); UNUSED(m); UNUSED(mb); UNUSED(me); UNUSED(plnr); return (1 && (!d->rs || (d->rs == rs)) && (!d->vs || (d->vs == vs)) && (!d->ms || (d->ms == ms)) ); } const ct_genus GENUS = { okp, 1 }; fftw-3.3.3/dft/scalar/f.h0000644000175000017500000000006712054410050012021 00000000000000#include "t.h" /* same stuff, no need to duplicate */ fftw-3.3.3/dft/scalar/n.h0000644000175000017500000000161212054410050012026 00000000000000/* * Copyright (c) 2003, 2007-11 Matteo Frigo * Copyright (c) 2003, 2007-11 Massachusetts Institute of Technology * * This program is free software; you can redistribute it and/or modify * it under the terms of the GNU General Public License as published by * the Free Software Foundation; either version 2 of the License, or * (at your option) any later version. * * This program is distributed in the hope that it will be useful, * but WITHOUT ANY WARRANTY; without even the implied warranty of * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the * GNU General Public License for more details. * * You should have received a copy of the GNU General Public License * along with this program; if not, write to the Free Software * Foundation, Inc., 51 Franklin Street, Fifth Floor, Boston, MA 02110-1301 USA * */ #define GENUS X(dft_n_genus) extern const kdft_genus GENUS; fftw-3.3.3/dft/scalar/q.h0000644000175000017500000000006712054410050012034 00000000000000#include "t.h" /* same stuff, no need to duplicate */ fftw-3.3.3/dft/scalar/t.h0000644000175000017500000000161012054410050012032 00000000000000/* * Copyright (c) 2003, 2007-11 Matteo Frigo * Copyright (c) 2003, 2007-11 Massachusetts Institute of Technology * * This program is free software; you can redistribute it and/or modify * it under the terms of the GNU General Public License as published by * the Free Software Foundation; either version 2 of the License, or * (at your option) any later version. * * This program is distributed in the hope that it will be useful, * but WITHOUT ANY WARRANTY; without even the implied warranty of * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the * GNU General Public License for more details. * * You should have received a copy of the GNU General Public License * along with this program; if not, write to the Free Software * Foundation, Inc., 51 Franklin Street, Fifth Floor, Boston, MA 02110-1301 USA * */ #define GENUS X(dft_t_genus) extern const ct_genus GENUS; fftw-3.3.3/dft/scalar/codelets/0000755000175000017500000000000012054411106013305 500000000000000fftw-3.3.3/dft/scalar/codelets/Makefile.am0000644000175000017500000001107012054410050015255 00000000000000# This Makefile.am specifies a set of codelets, efficient transforms # of small sizes, that are used as building blocks (kernels) by FFTW # to build up large transforms, as well as the options for generating # and compiling them. # You can customize FFTW for special needs, e.g. to handle certain # sizes more efficiently, by adding new codelets to the lists of those # included by default. If you change the list of codelets, any new # ones you added will be automatically generated when you run the # bootstrap script (see "Generating your own code" in the FFTW # manual). ########################################################################### AM_CPPFLAGS = -I$(top_srcdir)/kernel -I$(top_srcdir)/dft \ -I$(top_srcdir)/dft/scalar noinst_LTLIBRARIES = libdft_scalar_codelets.la ########################################################################### # n1_ is a hard-coded FFT of size (base cases of FFT recursion) N1 = n1_2.c n1_3.c n1_4.c n1_5.c n1_6.c n1_7.c n1_8.c n1_9.c n1_10.c \ n1_11.c n1_12.c n1_13.c n1_14.c n1_15.c n1_16.c n1_32.c n1_64.c \ n1_20.c n1_25.c # n1_30.c n1_40.c n1_50.c ########################################################################### # t1_ is a "twiddle" FFT of size , implementing a radix-r DIT step T1 = t1_2.c t1_3.c t1_4.c t1_5.c t1_6.c t1_7.c t1_8.c t1_9.c \ t1_10.c t1_12.c t1_15.c t1_16.c t1_32.c t1_64.c \ t1_20.c t1_25.c # t1_30.c t1_40.c t1_50.c # t2_ is also a twiddle FFT, but instead of using a complete lookup table # of trig. functions, it partially generates the trig. values on the fly # (this is faster for large sizes). T2 = t2_4.c t2_8.c t2_16.c t2_32.c t2_64.c \ t2_5.c t2_10.c t2_20.c t2_25.c ########################################################################### # The F (DIF) codelets are used for a kind of in-place transform algorithm, # but the planner seems to never (or hardly ever) use them on the machines # we have access to, preferring the Q codelets and the use of buffers # for sub-transforms. So, we comment them out, at least for now. # f1_ is a "twiddle" FFT of size , implementing a radix-r DIF step F1 = # f1_2.c f1_3.c f1_4.c f1_5.c f1_6.c f1_7.c f1_8.c f1_9.c f1_10.c f1_12.c f1_15.c f1_16.c f1_32.c f1_64.c # like f1, but partially generates its trig. table on the fly F2 = # f2_4.c f2_8.c f2_16.c f2_32.c f2_64.c ########################################################################### # q1_ is twiddle FFTs of size (DIF step), where the output is # transposed. This is used for in-place transposes in sizes that are # divisible by ^2. These codelets have size ~ ^2, so you should # probably not use bigger than 8 or so. Q1 = q1_2.c q1_4.c q1_8.c q1_3.c q1_5.c q1_6.c ########################################################################### ALL_CODELETS = $(N1) $(T1) $(T2) $(F1) $(F2) $(Q1) BUILT_SOURCES= $(ALL_CODELETS) $(CODLIST) libdft_scalar_codelets_la_SOURCES = $(BUILT_SOURCES) SOLVTAB_NAME = X(solvtab_dft_standard) XRENAME=X # special rules for regenerating codelets. include $(top_srcdir)/support/Makefile.codelets if MAINTAINER_MODE FLAGS_N1=$(DFT_FLAGS_COMMON) FLAGS_T1=$(DFT_FLAGS_COMMON) FLAGS_T2=$(DFT_FLAGS_COMMON) -twiddle-log3 -precompute-twiddles FLAGS_F1=$(DFT_FLAGS_COMMON) FLAGS_F2=$(DFT_FLAGS_COMMON) -twiddle-log3 -precompute-twiddles FLAGS_Q1=$(DFT_FLAGS_COMMON) -reload-twiddle FLAGS_Q2=$(DFT_FLAGS_COMMON) -twiddle-log3 -precompute-twiddles n1_%.c: $(CODELET_DEPS) $(GEN_NOTW) ($(PRELUDE_COMMANDS_DFT); $(TWOVERS) $(GEN_NOTW) $(FLAGS_N1) -n $* -name n1_$* -include "n.h") | $(ADD_DATE) | $(INDENT) >$@ t1_%.c: $(CODELET_DEPS) $(GEN_TWIDDLE) ($(PRELUDE_COMMANDS_DFT); $(TWOVERS) $(GEN_TWIDDLE) $(FLAGS_T1) -n $* -name t1_$* -include "t.h") | $(ADD_DATE) | $(INDENT) >$@ t2_%.c: $(CODELET_DEPS) $(GEN_TWIDDLE) ($(PRELUDE_COMMANDS_DFT); $(TWOVERS) $(GEN_TWIDDLE) $(FLAGS_T2) -n $* -name t2_$* -include "t.h") | $(ADD_DATE) | $(INDENT) >$@ f1_%.c: $(CODELET_DEPS) $(GEN_TWIDDLE) ($(PRELUDE_COMMANDS_DFT); $(TWOVERS) $(GEN_TWIDDLE) $(FLAGS_F1) -dif -n $* -name f1_$* -include "f.h") | $(ADD_DATE) | $(INDENT) >$@ f2_%.c: $(CODELET_DEPS) $(GEN_TWIDDLE) ($(PRELUDE_COMMANDS_DFT); $(TWOVERS) $(GEN_TWIDDLE) $(FLAGS_F2) -dif -n $* -name f2_$* -include "f.h") | $(ADD_DATE) | $(INDENT) >$@ q1_%.c: $(CODELET_DEPS) $(GEN_TWIDSQ) ($(PRELUDE_COMMANDS_DFT); $(TWOVERS) $(GEN_TWIDSQ) $(FLAGS_Q1) -dif -n $* -name q1_$* -include "q.h") | $(ADD_DATE) | $(INDENT) >$@ q2_%.c: $(CODELET_DEPS) $(GEN_TWIDSQ) ($(PRELUDE_COMMANDS_DFT); $(TWOVERS) $(GEN_TWIDSQ) $(FLAGS_Q2) -dif -n $* -name q2_$* -include "q.h") | $(ADD_DATE) | $(INDENT) >$@ endif # MAINTAINER_MODE fftw-3.3.3/dft/scalar/codelets/Makefile.in0000644000175000017500000007042512054410177015311 00000000000000# Makefile.in generated by automake 1.11.6 from Makefile.am. # @configure_input@ # Copyright (C) 1994, 1995, 1996, 1997, 1998, 1999, 2000, 2001, 2002, # 2003, 2004, 2005, 2006, 2007, 2008, 2009, 2010, 2011 Free Software # Foundation, Inc. # This Makefile.in is free software; the Free Software Foundation # gives unlimited permission to copy and/or distribute it, # with or without modifications, as long as this notice is preserved. # This program is distributed in the hope that it will be useful, # but WITHOUT ANY WARRANTY, to the extent permitted by law; without # even the implied warranty of MERCHANTABILITY or FITNESS FOR A # PARTICULAR PURPOSE. @SET_MAKE@ # This Makefile.am specifies a set of codelets, efficient transforms # of small sizes, that are used as building blocks (kernels) by FFTW # to build up large transforms, as well as the options for generating # and compiling them. # You can customize FFTW for special needs, e.g. to handle certain # sizes more efficiently, by adding new codelets to the lists of those # included by default. If you change the list of codelets, any new # ones you added will be automatically generated when you run the # bootstrap script (see "Generating your own code" in the FFTW # manual). # -*- makefile -*- # This file contains special make rules to generate codelets. # Most of this file requires GNU make . VPATH = @srcdir@ am__make_dryrun = \ { \ am__dry=no; \ case $$MAKEFLAGS in \ *\\[\ \ ]*) \ echo 'am--echo: ; @echo "AM" OK' | $(MAKE) -f - 2>/dev/null \ | grep '^AM OK$$' >/dev/null || am__dry=yes;; \ *) \ for am__flg in $$MAKEFLAGS; do \ case $$am__flg in \ *=*|--*) ;; \ *n*) am__dry=yes; break;; \ esac; \ done;; \ esac; \ test $$am__dry = yes; \ } pkgdatadir = $(datadir)/@PACKAGE@ pkgincludedir = $(includedir)/@PACKAGE@ pkglibdir = $(libdir)/@PACKAGE@ pkglibexecdir = $(libexecdir)/@PACKAGE@ am__cd = CDPATH="$${ZSH_VERSION+.}$(PATH_SEPARATOR)" && cd install_sh_DATA = $(install_sh) -c -m 644 install_sh_PROGRAM = $(install_sh) -c install_sh_SCRIPT = $(install_sh) -c INSTALL_HEADER = $(INSTALL_DATA) transform = $(program_transform_name) NORMAL_INSTALL = : PRE_INSTALL = : POST_INSTALL = : NORMAL_UNINSTALL = : PRE_UNINSTALL = : POST_UNINSTALL = : build_triplet = @build@ host_triplet = @host@ DIST_COMMON = $(srcdir)/Makefile.am $(srcdir)/Makefile.in \ $(top_srcdir)/support/Makefile.codelets subdir = dft/scalar/codelets ACLOCAL_M4 = $(top_srcdir)/aclocal.m4 am__aclocal_m4_deps = $(top_srcdir)/m4/acx_mpi.m4 \ $(top_srcdir)/m4/acx_pthread.m4 \ $(top_srcdir)/m4/ax_cc_maxopt.m4 \ $(top_srcdir)/m4/ax_check_compiler_flags.m4 \ $(top_srcdir)/m4/ax_compiler_vendor.m4 \ $(top_srcdir)/m4/ax_gcc_aligns_stack.m4 \ $(top_srcdir)/m4/ax_gcc_version.m4 \ $(top_srcdir)/m4/ax_openmp.m4 $(top_srcdir)/m4/libtool.m4 \ $(top_srcdir)/m4/ltoptions.m4 $(top_srcdir)/m4/ltsugar.m4 \ $(top_srcdir)/m4/ltversion.m4 $(top_srcdir)/m4/lt~obsolete.m4 \ $(top_srcdir)/configure.ac am__configure_deps = $(am__aclocal_m4_deps) $(CONFIGURE_DEPENDENCIES) \ $(ACLOCAL_M4) mkinstalldirs = $(install_sh) -d CONFIG_HEADER = $(top_builddir)/config.h CONFIG_CLEAN_FILES = CONFIG_CLEAN_VPATH_FILES = LTLIBRARIES = $(noinst_LTLIBRARIES) libdft_scalar_codelets_la_LIBADD = am__objects_1 = n1_2.lo n1_3.lo n1_4.lo n1_5.lo n1_6.lo n1_7.lo \ n1_8.lo n1_9.lo n1_10.lo n1_11.lo n1_12.lo n1_13.lo n1_14.lo \ n1_15.lo n1_16.lo n1_32.lo n1_64.lo n1_20.lo n1_25.lo am__objects_2 = t1_2.lo t1_3.lo t1_4.lo t1_5.lo t1_6.lo t1_7.lo \ t1_8.lo t1_9.lo t1_10.lo t1_12.lo t1_15.lo t1_16.lo t1_32.lo \ t1_64.lo t1_20.lo t1_25.lo am__objects_3 = t2_4.lo t2_8.lo t2_16.lo t2_32.lo t2_64.lo t2_5.lo \ t2_10.lo t2_20.lo t2_25.lo am__objects_4 = am__objects_5 = q1_2.lo q1_4.lo q1_8.lo q1_3.lo q1_5.lo q1_6.lo am__objects_6 = $(am__objects_1) $(am__objects_2) $(am__objects_3) \ $(am__objects_4) $(am__objects_4) $(am__objects_5) am__objects_7 = codlist.lo am__objects_8 = $(am__objects_6) $(am__objects_7) am_libdft_scalar_codelets_la_OBJECTS = $(am__objects_8) libdft_scalar_codelets_la_OBJECTS = \ $(am_libdft_scalar_codelets_la_OBJECTS) DEFAULT_INCLUDES = -I.@am__isrc@ -I$(top_builddir) depcomp = $(SHELL) $(top_srcdir)/depcomp am__depfiles_maybe = depfiles am__mv = mv -f COMPILE = $(CC) $(DEFS) $(DEFAULT_INCLUDES) $(INCLUDES) $(AM_CPPFLAGS) \ $(CPPFLAGS) $(AM_CFLAGS) $(CFLAGS) LTCOMPILE = $(LIBTOOL) --tag=CC $(AM_LIBTOOLFLAGS) $(LIBTOOLFLAGS) \ --mode=compile $(CC) $(DEFS) $(DEFAULT_INCLUDES) $(INCLUDES) \ $(AM_CPPFLAGS) $(CPPFLAGS) $(AM_CFLAGS) $(CFLAGS) CCLD = $(CC) LINK = $(LIBTOOL) --tag=CC $(AM_LIBTOOLFLAGS) $(LIBTOOLFLAGS) \ --mode=link $(CCLD) $(AM_CFLAGS) $(CFLAGS) $(AM_LDFLAGS) \ $(LDFLAGS) -o $@ SOURCES = $(libdft_scalar_codelets_la_SOURCES) DIST_SOURCES = $(libdft_scalar_codelets_la_SOURCES) am__can_run_installinfo = \ case $$AM_UPDATE_INFO_DIR in \ n|no|NO) false;; \ *) (install-info --version) >/dev/null 2>&1;; \ esac ETAGS = etags CTAGS = ctags DISTFILES = $(DIST_COMMON) $(DIST_SOURCES) $(TEXINFOS) $(EXTRA_DIST) ACLOCAL = @ACLOCAL@ ALLOCA = @ALLOCA@ ALTIVEC_CFLAGS = @ALTIVEC_CFLAGS@ AMTAR = @AMTAR@ AR = @AR@ AS = @AS@ AUTOCONF = @AUTOCONF@ AUTOHEADER = @AUTOHEADER@ AUTOMAKE = @AUTOMAKE@ AVX_CFLAGS = @AVX_CFLAGS@ AWK = @AWK@ CC = @CC@ CCDEPMODE = @CCDEPMODE@ CFLAGS = @CFLAGS@ CHECK_PL_OPTS = @CHECK_PL_OPTS@ CPP = @CPP@ CPPFLAGS = @CPPFLAGS@ CYGPATH_W = @CYGPATH_W@ C_FFTW_R2R_KIND = @C_FFTW_R2R_KIND@ C_MPI_FINT = @C_MPI_FINT@ DEFS = @DEFS@ DEPDIR = @DEPDIR@ DLLTOOL = @DLLTOOL@ DSYMUTIL = @DSYMUTIL@ DUMPBIN = @DUMPBIN@ ECHO_C = @ECHO_C@ ECHO_N = @ECHO_N@ ECHO_T = @ECHO_T@ EGREP = @EGREP@ EXEEXT = @EXEEXT@ F77 = @F77@ FFLAGS = @FFLAGS@ FGREP = @FGREP@ FLIBS = @FLIBS@ GREP = @GREP@ INSTALL = @INSTALL@ INSTALL_DATA = @INSTALL_DATA@ INSTALL_PROGRAM = @INSTALL_PROGRAM@ INSTALL_SCRIPT = @INSTALL_SCRIPT@ INSTALL_STRIP_PROGRAM = @INSTALL_STRIP_PROGRAM@ LD = @LD@ LDFLAGS = @LDFLAGS@ LIBOBJS = @LIBOBJS@ LIBQUADMATH = @LIBQUADMATH@ LIBS = @LIBS@ LIBTOOL = @LIBTOOL@ LIPO = @LIPO@ LN_S = @LN_S@ LTLIBOBJS = @LTLIBOBJS@ MAINT = @MAINT@ MAKEINFO = @MAKEINFO@ MANIFEST_TOOL = @MANIFEST_TOOL@ MKDIR_P = @MKDIR_P@ MPICC = @MPICC@ MPILIBS = @MPILIBS@ MPIRUN = @MPIRUN@ NEON_CFLAGS = @NEON_CFLAGS@ NM = @NM@ NMEDIT = @NMEDIT@ OBJDUMP = @OBJDUMP@ OBJEXT = @OBJEXT@ OCAMLBUILD = @OCAMLBUILD@ OPENMP_CFLAGS = @OPENMP_CFLAGS@ OTOOL = @OTOOL@ OTOOL64 = @OTOOL64@ PACKAGE = @PACKAGE@ PACKAGE_BUGREPORT = @PACKAGE_BUGREPORT@ PACKAGE_NAME = @PACKAGE_NAME@ PACKAGE_STRING = @PACKAGE_STRING@ PACKAGE_TARNAME = @PACKAGE_TARNAME@ PACKAGE_URL = @PACKAGE_URL@ PACKAGE_VERSION = @PACKAGE_VERSION@ PATH_SEPARATOR = @PATH_SEPARATOR@ POW_LIB = @POW_LIB@ PRECISION = @PRECISION@ PREC_SUFFIX = @PREC_SUFFIX@ PTHREAD_CC = @PTHREAD_CC@ PTHREAD_CFLAGS = @PTHREAD_CFLAGS@ PTHREAD_LIBS = @PTHREAD_LIBS@ RANLIB = @RANLIB@ SED = @SED@ SET_MAKE = @SET_MAKE@ SHARED_VERSION_INFO = @SHARED_VERSION_INFO@ SHELL = @SHELL@ SSE2_CFLAGS = @SSE2_CFLAGS@ STACK_ALIGN_CFLAGS = @STACK_ALIGN_CFLAGS@ STRIP = @STRIP@ THREADLIBS = @THREADLIBS@ VERSION = @VERSION@ abs_builddir = @abs_builddir@ abs_srcdir = @abs_srcdir@ abs_top_builddir = @abs_top_builddir@ abs_top_srcdir = @abs_top_srcdir@ ac_ct_AR = @ac_ct_AR@ ac_ct_CC = @ac_ct_CC@ ac_ct_DUMPBIN = @ac_ct_DUMPBIN@ ac_ct_F77 = @ac_ct_F77@ acx_pthread_config = @acx_pthread_config@ am__include = @am__include@ am__leading_dot = @am__leading_dot@ am__quote = @am__quote@ am__tar = @am__tar@ am__untar = @am__untar@ bindir = @bindir@ build = @build@ build_alias = @build_alias@ build_cpu = @build_cpu@ build_os = @build_os@ build_vendor = @build_vendor@ builddir = @builddir@ datadir = @datadir@ datarootdir = @datarootdir@ docdir = @docdir@ dvidir = @dvidir@ exec_prefix = @exec_prefix@ host = @host@ host_alias = @host_alias@ host_cpu = @host_cpu@ host_os = @host_os@ host_vendor = @host_vendor@ htmldir = @htmldir@ includedir = @includedir@ infodir = @infodir@ install_sh = @install_sh@ libdir = @libdir@ libexecdir = @libexecdir@ localedir = @localedir@ localstatedir = @localstatedir@ mandir = @mandir@ mkdir_p = @mkdir_p@ oldincludedir = @oldincludedir@ pdfdir = @pdfdir@ prefix = @prefix@ program_transform_name = @program_transform_name@ psdir = @psdir@ sbindir = @sbindir@ sharedstatedir = @sharedstatedir@ srcdir = @srcdir@ sysconfdir = @sysconfdir@ target_alias = @target_alias@ top_build_prefix = @top_build_prefix@ top_builddir = @top_builddir@ top_srcdir = @top_srcdir@ ########################################################################### AM_CPPFLAGS = -I$(top_srcdir)/kernel -I$(top_srcdir)/dft \ -I$(top_srcdir)/dft/scalar noinst_LTLIBRARIES = libdft_scalar_codelets.la ########################################################################### # n1_ is a hard-coded FFT of size (base cases of FFT recursion) N1 = n1_2.c n1_3.c n1_4.c n1_5.c n1_6.c n1_7.c n1_8.c n1_9.c n1_10.c \ n1_11.c n1_12.c n1_13.c n1_14.c n1_15.c n1_16.c n1_32.c n1_64.c \ n1_20.c n1_25.c # n1_30.c n1_40.c n1_50.c ########################################################################### # t1_ is a "twiddle" FFT of size , implementing a radix-r DIT step T1 = t1_2.c t1_3.c t1_4.c t1_5.c t1_6.c t1_7.c t1_8.c t1_9.c \ t1_10.c t1_12.c t1_15.c t1_16.c t1_32.c t1_64.c \ t1_20.c t1_25.c # t1_30.c t1_40.c t1_50.c # t2_ is also a twiddle FFT, but instead of using a complete lookup table # of trig. functions, it partially generates the trig. values on the fly # (this is faster for large sizes). T2 = t2_4.c t2_8.c t2_16.c t2_32.c t2_64.c \ t2_5.c t2_10.c t2_20.c t2_25.c ########################################################################### # The F (DIF) codelets are used for a kind of in-place transform algorithm, # but the planner seems to never (or hardly ever) use them on the machines # we have access to, preferring the Q codelets and the use of buffers # for sub-transforms. So, we comment them out, at least for now. # f1_ is a "twiddle" FFT of size , implementing a radix-r DIF step F1 = # f1_2.c f1_3.c f1_4.c f1_5.c f1_6.c f1_7.c f1_8.c f1_9.c f1_10.c f1_12.c f1_15.c f1_16.c f1_32.c f1_64.c # like f1, but partially generates its trig. table on the fly F2 = # f2_4.c f2_8.c f2_16.c f2_32.c f2_64.c ########################################################################### # q1_ is twiddle FFTs of size (DIF step), where the output is # transposed. This is used for in-place transposes in sizes that are # divisible by ^2. These codelets have size ~ ^2, so you should # probably not use bigger than 8 or so. Q1 = q1_2.c q1_4.c q1_8.c q1_3.c q1_5.c q1_6.c ########################################################################### ALL_CODELETS = $(N1) $(T1) $(T2) $(F1) $(F2) $(Q1) BUILT_SOURCES = $(ALL_CODELETS) $(CODLIST) libdft_scalar_codelets_la_SOURCES = $(BUILT_SOURCES) SOLVTAB_NAME = X(solvtab_dft_standard) XRENAME = X CODLIST = codlist.c CODELET_NAME = codelet_ @MAINTAINER_MODE_TRUE@INDENT = indent -kr -cs -i5 -l800 -fca -nfc1 -sc -sob -cli4 -TR -Tplanner -TV @MAINTAINER_MODE_TRUE@TWOVERS = sh ${top_srcdir}/support/twovers.sh @MAINTAINER_MODE_TRUE@GENFFTDIR = ${top_builddir}/genfft @MAINTAINER_MODE_TRUE@GEN_NOTW = ${GENFFTDIR}/gen_notw.native @MAINTAINER_MODE_TRUE@GEN_NOTW_C = ${GENFFTDIR}/gen_notw_c.native @MAINTAINER_MODE_TRUE@GEN_TWIDDLE = ${GENFFTDIR}/gen_twiddle.native @MAINTAINER_MODE_TRUE@GEN_TWIDDLE_C = ${GENFFTDIR}/gen_twiddle_c.native @MAINTAINER_MODE_TRUE@GEN_TWIDSQ = ${GENFFTDIR}/gen_twidsq.native @MAINTAINER_MODE_TRUE@GEN_TWIDSQ_C = ${GENFFTDIR}/gen_twidsq_c.native @MAINTAINER_MODE_TRUE@GEN_R2CF = ${GENFFTDIR}/gen_r2cf.native @MAINTAINER_MODE_TRUE@GEN_R2CB = ${GENFFTDIR}/gen_r2cb.native @MAINTAINER_MODE_TRUE@GEN_HC2HC = ${GENFFTDIR}/gen_hc2hc.native @MAINTAINER_MODE_TRUE@GEN_HC2C = ${GENFFTDIR}/gen_hc2c.native @MAINTAINER_MODE_TRUE@GEN_HC2CDFT = ${GENFFTDIR}/gen_hc2cdft.native @MAINTAINER_MODE_TRUE@GEN_HC2CDFT_C = ${GENFFTDIR}/gen_hc2cdft_c.native @MAINTAINER_MODE_TRUE@GEN_R2R = ${GENFFTDIR}/gen_r2r.native @MAINTAINER_MODE_TRUE@PRELUDE_DFT = ${top_srcdir}/support/codelet_prelude.dft @MAINTAINER_MODE_TRUE@PRELUDE_RDFT = ${top_srcdir}/support/codelet_prelude.rdft @MAINTAINER_MODE_TRUE@ADD_DATE = sed -e s/@DATE@/"`date`"/ @MAINTAINER_MODE_TRUE@COPYRIGHT = ${top_srcdir}/COPYRIGHT @MAINTAINER_MODE_TRUE@CODELET_DEPS = $(COPYRIGHT) $(PRELUDE) @MAINTAINER_MODE_TRUE@PRELUDE_COMMANDS_DFT = cat $(COPYRIGHT) $(PRELUDE_DFT) @MAINTAINER_MODE_TRUE@PRELUDE_COMMANDS_RDFT = cat $(COPYRIGHT) $(PRELUDE_RDFT) @MAINTAINER_MODE_TRUE@FLAGS_COMMON = -compact -variables 4 @MAINTAINER_MODE_TRUE@DFT_FLAGS_COMMON = $(FLAGS_COMMON) -pipeline-latency 4 @MAINTAINER_MODE_TRUE@RDFT_FLAGS_COMMON = $(FLAGS_COMMON) -pipeline-latency 4 # special rules for regenerating codelets. @MAINTAINER_MODE_TRUE@FLAGS_N1 = $(DFT_FLAGS_COMMON) @MAINTAINER_MODE_TRUE@FLAGS_T1 = $(DFT_FLAGS_COMMON) @MAINTAINER_MODE_TRUE@FLAGS_T2 = $(DFT_FLAGS_COMMON) -twiddle-log3 -precompute-twiddles @MAINTAINER_MODE_TRUE@FLAGS_F1 = $(DFT_FLAGS_COMMON) @MAINTAINER_MODE_TRUE@FLAGS_F2 = $(DFT_FLAGS_COMMON) -twiddle-log3 -precompute-twiddles @MAINTAINER_MODE_TRUE@FLAGS_Q1 = $(DFT_FLAGS_COMMON) -reload-twiddle @MAINTAINER_MODE_TRUE@FLAGS_Q2 = $(DFT_FLAGS_COMMON) -twiddle-log3 -precompute-twiddles all: $(BUILT_SOURCES) $(MAKE) $(AM_MAKEFLAGS) all-am .SUFFIXES: .SUFFIXES: .c .lo .o .obj $(srcdir)/Makefile.in: @MAINTAINER_MODE_TRUE@ $(srcdir)/Makefile.am $(top_srcdir)/support/Makefile.codelets $(am__configure_deps) @for dep in $?; do \ case '$(am__configure_deps)' in \ *$$dep*) \ ( cd $(top_builddir) && $(MAKE) $(AM_MAKEFLAGS) am--refresh ) \ && { if test -f $@; then exit 0; else break; fi; }; \ exit 1;; \ esac; \ done; \ echo ' cd $(top_srcdir) && $(AUTOMAKE) --gnu dft/scalar/codelets/Makefile'; \ $(am__cd) $(top_srcdir) && \ $(AUTOMAKE) --gnu dft/scalar/codelets/Makefile .PRECIOUS: Makefile Makefile: $(srcdir)/Makefile.in $(top_builddir)/config.status @case '$?' in \ *config.status*) \ cd $(top_builddir) && $(MAKE) $(AM_MAKEFLAGS) am--refresh;; \ *) \ echo ' cd $(top_builddir) && $(SHELL) ./config.status $(subdir)/$@ $(am__depfiles_maybe)'; \ cd $(top_builddir) && $(SHELL) ./config.status $(subdir)/$@ $(am__depfiles_maybe);; \ esac; $(top_srcdir)/support/Makefile.codelets: $(top_builddir)/config.status: $(top_srcdir)/configure $(CONFIG_STATUS_DEPENDENCIES) cd $(top_builddir) && $(MAKE) $(AM_MAKEFLAGS) am--refresh $(top_srcdir)/configure: @MAINTAINER_MODE_TRUE@ $(am__configure_deps) cd $(top_builddir) && $(MAKE) $(AM_MAKEFLAGS) am--refresh $(ACLOCAL_M4): @MAINTAINER_MODE_TRUE@ $(am__aclocal_m4_deps) cd $(top_builddir) && $(MAKE) $(AM_MAKEFLAGS) am--refresh $(am__aclocal_m4_deps): clean-noinstLTLIBRARIES: -test -z "$(noinst_LTLIBRARIES)" || rm -f $(noinst_LTLIBRARIES) @list='$(noinst_LTLIBRARIES)'; for p in $$list; do \ dir="`echo $$p | sed -e 's|/[^/]*$$||'`"; \ test "$$dir" != "$$p" || dir=.; \ echo "rm -f \"$${dir}/so_locations\""; \ rm -f "$${dir}/so_locations"; \ done libdft_scalar_codelets.la: $(libdft_scalar_codelets_la_OBJECTS) $(libdft_scalar_codelets_la_DEPENDENCIES) $(EXTRA_libdft_scalar_codelets_la_DEPENDENCIES) $(LINK) $(libdft_scalar_codelets_la_OBJECTS) $(libdft_scalar_codelets_la_LIBADD) $(LIBS) mostlyclean-compile: -rm -f *.$(OBJEXT) distclean-compile: -rm -f *.tab.c @AMDEP_TRUE@@am__include@ @am__quote@./$(DEPDIR)/codlist.Plo@am__quote@ @AMDEP_TRUE@@am__include@ @am__quote@./$(DEPDIR)/n1_10.Plo@am__quote@ @AMDEP_TRUE@@am__include@ @am__quote@./$(DEPDIR)/n1_11.Plo@am__quote@ @AMDEP_TRUE@@am__include@ @am__quote@./$(DEPDIR)/n1_12.Plo@am__quote@ @AMDEP_TRUE@@am__include@ @am__quote@./$(DEPDIR)/n1_13.Plo@am__quote@ @AMDEP_TRUE@@am__include@ @am__quote@./$(DEPDIR)/n1_14.Plo@am__quote@ @AMDEP_TRUE@@am__include@ @am__quote@./$(DEPDIR)/n1_15.Plo@am__quote@ @AMDEP_TRUE@@am__include@ @am__quote@./$(DEPDIR)/n1_16.Plo@am__quote@ @AMDEP_TRUE@@am__include@ @am__quote@./$(DEPDIR)/n1_2.Plo@am__quote@ @AMDEP_TRUE@@am__include@ @am__quote@./$(DEPDIR)/n1_20.Plo@am__quote@ @AMDEP_TRUE@@am__include@ @am__quote@./$(DEPDIR)/n1_25.Plo@am__quote@ @AMDEP_TRUE@@am__include@ @am__quote@./$(DEPDIR)/n1_3.Plo@am__quote@ @AMDEP_TRUE@@am__include@ @am__quote@./$(DEPDIR)/n1_32.Plo@am__quote@ @AMDEP_TRUE@@am__include@ @am__quote@./$(DEPDIR)/n1_4.Plo@am__quote@ @AMDEP_TRUE@@am__include@ @am__quote@./$(DEPDIR)/n1_5.Plo@am__quote@ @AMDEP_TRUE@@am__include@ @am__quote@./$(DEPDIR)/n1_6.Plo@am__quote@ @AMDEP_TRUE@@am__include@ @am__quote@./$(DEPDIR)/n1_64.Plo@am__quote@ @AMDEP_TRUE@@am__include@ @am__quote@./$(DEPDIR)/n1_7.Plo@am__quote@ @AMDEP_TRUE@@am__include@ @am__quote@./$(DEPDIR)/n1_8.Plo@am__quote@ @AMDEP_TRUE@@am__include@ @am__quote@./$(DEPDIR)/n1_9.Plo@am__quote@ @AMDEP_TRUE@@am__include@ @am__quote@./$(DEPDIR)/q1_2.Plo@am__quote@ @AMDEP_TRUE@@am__include@ @am__quote@./$(DEPDIR)/q1_3.Plo@am__quote@ @AMDEP_TRUE@@am__include@ @am__quote@./$(DEPDIR)/q1_4.Plo@am__quote@ @AMDEP_TRUE@@am__include@ @am__quote@./$(DEPDIR)/q1_5.Plo@am__quote@ @AMDEP_TRUE@@am__include@ @am__quote@./$(DEPDIR)/q1_6.Plo@am__quote@ @AMDEP_TRUE@@am__include@ @am__quote@./$(DEPDIR)/q1_8.Plo@am__quote@ @AMDEP_TRUE@@am__include@ @am__quote@./$(DEPDIR)/t1_10.Plo@am__quote@ @AMDEP_TRUE@@am__include@ @am__quote@./$(DEPDIR)/t1_12.Plo@am__quote@ @AMDEP_TRUE@@am__include@ @am__quote@./$(DEPDIR)/t1_15.Plo@am__quote@ @AMDEP_TRUE@@am__include@ @am__quote@./$(DEPDIR)/t1_16.Plo@am__quote@ @AMDEP_TRUE@@am__include@ @am__quote@./$(DEPDIR)/t1_2.Plo@am__quote@ @AMDEP_TRUE@@am__include@ @am__quote@./$(DEPDIR)/t1_20.Plo@am__quote@ @AMDEP_TRUE@@am__include@ @am__quote@./$(DEPDIR)/t1_25.Plo@am__quote@ @AMDEP_TRUE@@am__include@ @am__quote@./$(DEPDIR)/t1_3.Plo@am__quote@ @AMDEP_TRUE@@am__include@ @am__quote@./$(DEPDIR)/t1_32.Plo@am__quote@ @AMDEP_TRUE@@am__include@ @am__quote@./$(DEPDIR)/t1_4.Plo@am__quote@ @AMDEP_TRUE@@am__include@ @am__quote@./$(DEPDIR)/t1_5.Plo@am__quote@ @AMDEP_TRUE@@am__include@ @am__quote@./$(DEPDIR)/t1_6.Plo@am__quote@ @AMDEP_TRUE@@am__include@ @am__quote@./$(DEPDIR)/t1_64.Plo@am__quote@ @AMDEP_TRUE@@am__include@ @am__quote@./$(DEPDIR)/t1_7.Plo@am__quote@ @AMDEP_TRUE@@am__include@ @am__quote@./$(DEPDIR)/t1_8.Plo@am__quote@ @AMDEP_TRUE@@am__include@ @am__quote@./$(DEPDIR)/t1_9.Plo@am__quote@ @AMDEP_TRUE@@am__include@ @am__quote@./$(DEPDIR)/t2_10.Plo@am__quote@ @AMDEP_TRUE@@am__include@ @am__quote@./$(DEPDIR)/t2_16.Plo@am__quote@ @AMDEP_TRUE@@am__include@ @am__quote@./$(DEPDIR)/t2_20.Plo@am__quote@ @AMDEP_TRUE@@am__include@ @am__quote@./$(DEPDIR)/t2_25.Plo@am__quote@ @AMDEP_TRUE@@am__include@ @am__quote@./$(DEPDIR)/t2_32.Plo@am__quote@ @AMDEP_TRUE@@am__include@ @am__quote@./$(DEPDIR)/t2_4.Plo@am__quote@ @AMDEP_TRUE@@am__include@ @am__quote@./$(DEPDIR)/t2_5.Plo@am__quote@ @AMDEP_TRUE@@am__include@ @am__quote@./$(DEPDIR)/t2_64.Plo@am__quote@ @AMDEP_TRUE@@am__include@ @am__quote@./$(DEPDIR)/t2_8.Plo@am__quote@ .c.o: @am__fastdepCC_TRUE@ $(COMPILE) -MT $@ -MD -MP -MF $(DEPDIR)/$*.Tpo -c -o $@ $< @am__fastdepCC_TRUE@ $(am__mv) $(DEPDIR)/$*.Tpo $(DEPDIR)/$*.Po @AMDEP_TRUE@@am__fastdepCC_FALSE@ source='$<' object='$@' libtool=no @AMDEPBACKSLASH@ @AMDEP_TRUE@@am__fastdepCC_FALSE@ DEPDIR=$(DEPDIR) $(CCDEPMODE) $(depcomp) @AMDEPBACKSLASH@ @am__fastdepCC_FALSE@ $(COMPILE) -c $< .c.obj: @am__fastdepCC_TRUE@ $(COMPILE) -MT $@ -MD -MP -MF $(DEPDIR)/$*.Tpo -c -o $@ `$(CYGPATH_W) '$<'` @am__fastdepCC_TRUE@ $(am__mv) $(DEPDIR)/$*.Tpo $(DEPDIR)/$*.Po @AMDEP_TRUE@@am__fastdepCC_FALSE@ source='$<' object='$@' libtool=no @AMDEPBACKSLASH@ @AMDEP_TRUE@@am__fastdepCC_FALSE@ DEPDIR=$(DEPDIR) $(CCDEPMODE) $(depcomp) @AMDEPBACKSLASH@ @am__fastdepCC_FALSE@ $(COMPILE) -c `$(CYGPATH_W) '$<'` .c.lo: @am__fastdepCC_TRUE@ $(LTCOMPILE) -MT $@ -MD -MP -MF $(DEPDIR)/$*.Tpo -c -o $@ $< @am__fastdepCC_TRUE@ $(am__mv) $(DEPDIR)/$*.Tpo $(DEPDIR)/$*.Plo @AMDEP_TRUE@@am__fastdepCC_FALSE@ source='$<' object='$@' libtool=yes @AMDEPBACKSLASH@ @AMDEP_TRUE@@am__fastdepCC_FALSE@ DEPDIR=$(DEPDIR) $(CCDEPMODE) $(depcomp) @AMDEPBACKSLASH@ @am__fastdepCC_FALSE@ $(LTCOMPILE) -c -o $@ $< mostlyclean-libtool: -rm -f *.lo clean-libtool: -rm -rf .libs _libs ID: $(HEADERS) $(SOURCES) $(LISP) $(TAGS_FILES) list='$(SOURCES) $(HEADERS) $(LISP) $(TAGS_FILES)'; \ unique=`for i in $$list; do \ if test -f "$$i"; then echo $$i; else echo $(srcdir)/$$i; fi; \ done | \ $(AWK) '{ files[$$0] = 1; nonempty = 1; } \ END { if (nonempty) { for (i in files) print i; }; }'`; \ mkid -fID $$unique tags: TAGS TAGS: $(HEADERS) $(SOURCES) $(TAGS_DEPENDENCIES) \ $(TAGS_FILES) $(LISP) set x; \ here=`pwd`; \ list='$(SOURCES) $(HEADERS) $(LISP) $(TAGS_FILES)'; \ unique=`for i in $$list; do \ if test -f "$$i"; then echo $$i; else echo $(srcdir)/$$i; fi; \ done | \ $(AWK) '{ files[$$0] = 1; nonempty = 1; } \ END { if (nonempty) { for (i in files) print i; }; }'`; \ shift; \ if test -z "$(ETAGS_ARGS)$$*$$unique"; then :; else \ test -n "$$unique" || unique=$$empty_fix; \ if test $$# -gt 0; then \ $(ETAGS) $(ETAGSFLAGS) $(AM_ETAGSFLAGS) $(ETAGS_ARGS) \ "$$@" $$unique; \ else \ $(ETAGS) $(ETAGSFLAGS) $(AM_ETAGSFLAGS) $(ETAGS_ARGS) \ $$unique; \ fi; \ fi ctags: CTAGS CTAGS: $(HEADERS) $(SOURCES) $(TAGS_DEPENDENCIES) \ $(TAGS_FILES) $(LISP) list='$(SOURCES) $(HEADERS) $(LISP) $(TAGS_FILES)'; \ unique=`for i in $$list; do \ if test -f "$$i"; then echo $$i; else echo $(srcdir)/$$i; fi; \ done | \ $(AWK) '{ files[$$0] = 1; nonempty = 1; } \ END { if (nonempty) { for (i in files) print i; }; }'`; \ test -z "$(CTAGS_ARGS)$$unique" \ || $(CTAGS) $(CTAGSFLAGS) $(AM_CTAGSFLAGS) $(CTAGS_ARGS) \ $$unique GTAGS: here=`$(am__cd) $(top_builddir) && pwd` \ && $(am__cd) $(top_srcdir) \ && gtags -i $(GTAGS_ARGS) "$$here" distclean-tags: -rm -f TAGS ID GTAGS GRTAGS GSYMS GPATH tags distdir: $(DISTFILES) @srcdirstrip=`echo "$(srcdir)" | sed 's/[].[^$$\\*]/\\\\&/g'`; \ topsrcdirstrip=`echo "$(top_srcdir)" | sed 's/[].[^$$\\*]/\\\\&/g'`; \ list='$(DISTFILES)'; \ dist_files=`for file in $$list; do echo $$file; done | \ sed -e "s|^$$srcdirstrip/||;t" \ -e "s|^$$topsrcdirstrip/|$(top_builddir)/|;t"`; \ case $$dist_files in \ */*) $(MKDIR_P) `echo "$$dist_files" | \ sed '/\//!d;s|^|$(distdir)/|;s,/[^/]*$$,,' | \ sort -u` ;; \ esac; \ for file in $$dist_files; do \ if test -f $$file || test -d $$file; then d=.; else d=$(srcdir); fi; \ if test -d $$d/$$file; then \ dir=`echo "/$$file" | sed -e 's,/[^/]*$$,,'`; \ if test -d "$(distdir)/$$file"; then \ find "$(distdir)/$$file" -type d ! -perm -700 -exec chmod u+rwx {} \;; \ fi; \ if test -d $(srcdir)/$$file && test $$d != $(srcdir); then \ cp -fpR $(srcdir)/$$file "$(distdir)$$dir" || exit 1; \ find "$(distdir)/$$file" -type d ! -perm -700 -exec chmod u+rwx {} \;; \ fi; \ cp -fpR $$d/$$file "$(distdir)$$dir" || exit 1; \ else \ test -f "$(distdir)/$$file" \ || cp -p $$d/$$file "$(distdir)/$$file" \ || exit 1; \ fi; \ done check-am: all-am check: $(BUILT_SOURCES) $(MAKE) $(AM_MAKEFLAGS) check-am all-am: Makefile $(LTLIBRARIES) installdirs: install: $(BUILT_SOURCES) $(MAKE) $(AM_MAKEFLAGS) install-am install-exec: install-exec-am install-data: install-data-am uninstall: uninstall-am install-am: all-am @$(MAKE) $(AM_MAKEFLAGS) install-exec-am install-data-am installcheck: installcheck-am install-strip: if test -z '$(STRIP)'; then \ $(MAKE) $(AM_MAKEFLAGS) INSTALL_PROGRAM="$(INSTALL_STRIP_PROGRAM)" \ install_sh_PROGRAM="$(INSTALL_STRIP_PROGRAM)" INSTALL_STRIP_FLAG=-s \ install; \ else \ $(MAKE) $(AM_MAKEFLAGS) INSTALL_PROGRAM="$(INSTALL_STRIP_PROGRAM)" \ install_sh_PROGRAM="$(INSTALL_STRIP_PROGRAM)" INSTALL_STRIP_FLAG=-s \ "INSTALL_PROGRAM_ENV=STRIPPROG='$(STRIP)'" install; \ fi mostlyclean-generic: clean-generic: distclean-generic: -test -z "$(CONFIG_CLEAN_FILES)" || rm -f $(CONFIG_CLEAN_FILES) -test . = "$(srcdir)" || test -z "$(CONFIG_CLEAN_VPATH_FILES)" || rm -f $(CONFIG_CLEAN_VPATH_FILES) maintainer-clean-generic: @echo "This command is intended for maintainers to use" @echo "it deletes files that may require special tools to rebuild." -test -z "$(BUILT_SOURCES)" || rm -f $(BUILT_SOURCES) clean: clean-am clean-am: clean-generic clean-libtool clean-noinstLTLIBRARIES \ mostlyclean-am distclean: distclean-am -rm -rf ./$(DEPDIR) -rm -f Makefile distclean-am: clean-am distclean-compile distclean-generic \ distclean-tags dvi: dvi-am dvi-am: html: html-am html-am: info: info-am info-am: install-data-am: install-dvi: install-dvi-am install-dvi-am: install-exec-am: install-html: install-html-am install-html-am: install-info: install-info-am install-info-am: install-man: install-pdf: install-pdf-am install-pdf-am: install-ps: install-ps-am install-ps-am: installcheck-am: maintainer-clean: maintainer-clean-am -rm -rf ./$(DEPDIR) -rm -f Makefile maintainer-clean-am: distclean-am maintainer-clean-generic \ maintainer-clean-local mostlyclean: mostlyclean-am mostlyclean-am: mostlyclean-compile mostlyclean-generic \ mostlyclean-libtool pdf: pdf-am pdf-am: ps: ps-am ps-am: uninstall-am: .MAKE: all check install install-am install-strip .PHONY: CTAGS GTAGS all all-am check check-am clean clean-generic \ clean-libtool clean-noinstLTLIBRARIES ctags distclean \ distclean-compile distclean-generic distclean-libtool \ distclean-tags distdir dvi dvi-am html html-am info info-am \ install install-am install-data install-data-am install-dvi \ install-dvi-am install-exec install-exec-am install-html \ install-html-am install-info install-info-am install-man \ install-pdf install-pdf-am install-ps install-ps-am \ install-strip installcheck installcheck-am installdirs \ maintainer-clean maintainer-clean-generic \ maintainer-clean-local mostlyclean mostlyclean-compile \ mostlyclean-generic mostlyclean-libtool pdf pdf-am ps ps-am \ tags uninstall uninstall-am # rule to build codlist $(CODLIST): Makefile ( \ echo "#include \"ifftw.h\""; \ echo $(INCLUDE_SIMD_HEADER); \ echo; \ for i in $(ALL_CODELETS) NIL; do \ if test "$$i" != NIL; then \ j=`basename $$i | sed -e 's/[.][cS]$$//g'`; \ echo "extern void $(XRENAME)($(CODELET_NAME)$$j)(planner *);"; \ fi \ done; \ echo; \ echo; \ echo "extern const solvtab $(SOLVTAB_NAME);"; \ echo "const solvtab $(SOLVTAB_NAME) = {"; \ for i in $(ALL_CODELETS) NIL; do \ if test "$$i" != NIL; then \ j=`basename $$i | sed -e 's/[.][cS]$$//g'`; \ echo " SOLVTAB($(XRENAME)($(CODELET_NAME)$$j)),"; \ fi \ done; \ echo " SOLVTAB_END"; \ echo "};"; \ ) >$@ # only delete codlist.c in maintainer-mode, since it is included in the dist # FIXME: is there a way to delete in 'make clean' only when builddir != srcdir? maintainer-clean-local: rm -f $(CODLIST) # cancel the hideous builtin rules that cause an infinite loop @MAINTAINER_MODE_TRUE@%: %.o @MAINTAINER_MODE_TRUE@%: %.s @MAINTAINER_MODE_TRUE@%: %.c @MAINTAINER_MODE_TRUE@%: %.S @MAINTAINER_MODE_TRUE@n1_%.c: $(CODELET_DEPS) $(GEN_NOTW) @MAINTAINER_MODE_TRUE@ ($(PRELUDE_COMMANDS_DFT); $(TWOVERS) $(GEN_NOTW) $(FLAGS_N1) -n $* -name n1_$* -include "n.h") | $(ADD_DATE) | $(INDENT) >$@ @MAINTAINER_MODE_TRUE@t1_%.c: $(CODELET_DEPS) $(GEN_TWIDDLE) @MAINTAINER_MODE_TRUE@ ($(PRELUDE_COMMANDS_DFT); $(TWOVERS) $(GEN_TWIDDLE) $(FLAGS_T1) -n $* -name t1_$* -include "t.h") | $(ADD_DATE) | $(INDENT) >$@ @MAINTAINER_MODE_TRUE@t2_%.c: $(CODELET_DEPS) $(GEN_TWIDDLE) @MAINTAINER_MODE_TRUE@ ($(PRELUDE_COMMANDS_DFT); $(TWOVERS) $(GEN_TWIDDLE) $(FLAGS_T2) -n $* -name t2_$* -include "t.h") | $(ADD_DATE) | $(INDENT) >$@ @MAINTAINER_MODE_TRUE@f1_%.c: $(CODELET_DEPS) $(GEN_TWIDDLE) @MAINTAINER_MODE_TRUE@ ($(PRELUDE_COMMANDS_DFT); $(TWOVERS) $(GEN_TWIDDLE) $(FLAGS_F1) -dif -n $* -name f1_$* -include "f.h") | $(ADD_DATE) | $(INDENT) >$@ @MAINTAINER_MODE_TRUE@f2_%.c: $(CODELET_DEPS) $(GEN_TWIDDLE) @MAINTAINER_MODE_TRUE@ ($(PRELUDE_COMMANDS_DFT); $(TWOVERS) $(GEN_TWIDDLE) $(FLAGS_F2) -dif -n $* -name f2_$* -include "f.h") | $(ADD_DATE) | $(INDENT) >$@ @MAINTAINER_MODE_TRUE@q1_%.c: $(CODELET_DEPS) $(GEN_TWIDSQ) @MAINTAINER_MODE_TRUE@ ($(PRELUDE_COMMANDS_DFT); $(TWOVERS) $(GEN_TWIDSQ) $(FLAGS_Q1) -dif -n $* -name q1_$* -include "q.h") | $(ADD_DATE) | $(INDENT) >$@ @MAINTAINER_MODE_TRUE@q2_%.c: $(CODELET_DEPS) $(GEN_TWIDSQ) @MAINTAINER_MODE_TRUE@ ($(PRELUDE_COMMANDS_DFT); $(TWOVERS) $(GEN_TWIDSQ) $(FLAGS_Q2) -dif -n $* -name q2_$* -include "q.h") | $(ADD_DATE) | $(INDENT) >$@ # Tell versions [3.59,3.63) of GNU make to not export all variables. # Otherwise a system limit (for SysV at least) may be exceeded. .NOEXPORT: fftw-3.3.3/dft/scalar/codelets/n1_2.c0000644000175000017500000000607012054410236014136 00000000000000/* * Copyright (c) 2003, 2007-11 Matteo Frigo * Copyright (c) 2003, 2007-11 Massachusetts Institute of Technology * * This program is free software; you can redistribute it and/or modify * it under the terms of the GNU General Public License as published by * the Free Software Foundation; either version 2 of the License, or * (at your option) any later version. * * This program is distributed in the hope that it will be useful, * but WITHOUT ANY WARRANTY; without even the implied warranty of * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the * GNU General Public License for more details. * * You should have received a copy of the GNU General Public License * along with this program; if not, write to the Free Software * Foundation, Inc., 51 Franklin Street, Fifth Floor, Boston, MA 02110-1301 USA * */ /* This file was automatically generated --- DO NOT EDIT */ /* Generated on Sun Nov 25 07:35:42 EST 2012 */ #include "codelet-dft.h" #ifdef HAVE_FMA /* Generated by: ../../../genfft/gen_notw.native -fma -reorder-insns -schedule-for-pipeline -compact -variables 4 -pipeline-latency 4 -n 2 -name n1_2 -include n.h */ /* * This function contains 4 FP additions, 0 FP multiplications, * (or, 4 additions, 0 multiplications, 0 fused multiply/add), * 5 stack variables, 0 constants, and 8 memory accesses */ #include "n.h" static void n1_2(const R *ri, const R *ii, R *ro, R *io, stride is, stride os, INT v, INT ivs, INT ovs) { { INT i; for (i = v; i > 0; i = i - 1, ri = ri + ivs, ii = ii + ivs, ro = ro + ovs, io = io + ovs, MAKE_VOLATILE_STRIDE(8, is), MAKE_VOLATILE_STRIDE(8, os)) { E T1, T2, T3, T4; T1 = ri[0]; T2 = ri[WS(is, 1)]; T3 = ii[0]; T4 = ii[WS(is, 1)]; ro[0] = T1 + T2; ro[WS(os, 1)] = T1 - T2; io[0] = T3 + T4; io[WS(os, 1)] = T3 - T4; } } } static const kdft_desc desc = { 2, "n1_2", {4, 0, 0, 0}, &GENUS, 0, 0, 0, 0 }; void X(codelet_n1_2) (planner *p) { X(kdft_register) (p, n1_2, &desc); } #else /* HAVE_FMA */ /* Generated by: ../../../genfft/gen_notw.native -compact -variables 4 -pipeline-latency 4 -n 2 -name n1_2 -include n.h */ /* * This function contains 4 FP additions, 0 FP multiplications, * (or, 4 additions, 0 multiplications, 0 fused multiply/add), * 5 stack variables, 0 constants, and 8 memory accesses */ #include "n.h" static void n1_2(const R *ri, const R *ii, R *ro, R *io, stride is, stride os, INT v, INT ivs, INT ovs) { { INT i; for (i = v; i > 0; i = i - 1, ri = ri + ivs, ii = ii + ivs, ro = ro + ovs, io = io + ovs, MAKE_VOLATILE_STRIDE(8, is), MAKE_VOLATILE_STRIDE(8, os)) { E T1, T2, T3, T4; T1 = ri[0]; T2 = ri[WS(is, 1)]; ro[WS(os, 1)] = T1 - T2; ro[0] = T1 + T2; T3 = ii[0]; T4 = ii[WS(is, 1)]; io[WS(os, 1)] = T3 - T4; io[0] = T3 + T4; } } } static const kdft_desc desc = { 2, "n1_2", {4, 0, 0, 0}, &GENUS, 0, 0, 0, 0 }; void X(codelet_n1_2) (planner *p) { X(kdft_register) (p, n1_2, &desc); } #endif /* HAVE_FMA */ fftw-3.3.3/dft/scalar/codelets/n1_3.c0000644000175000017500000001013612054410236014135 00000000000000/* * Copyright (c) 2003, 2007-11 Matteo Frigo * Copyright (c) 2003, 2007-11 Massachusetts Institute of Technology * * This program is free software; you can redistribute it and/or modify * it under the terms of the GNU General Public License as published by * the Free Software Foundation; either version 2 of the License, or * (at your option) any later version. * * This program is distributed in the hope that it will be useful, * but WITHOUT ANY WARRANTY; without even the implied warranty of * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the * GNU General Public License for more details. * * You should have received a copy of the GNU General Public License * along with this program; if not, write to the Free Software * Foundation, Inc., 51 Franklin Street, Fifth Floor, Boston, MA 02110-1301 USA * */ /* This file was automatically generated --- DO NOT EDIT */ /* Generated on Sun Nov 25 07:35:42 EST 2012 */ #include "codelet-dft.h" #ifdef HAVE_FMA /* Generated by: ../../../genfft/gen_notw.native -fma -reorder-insns -schedule-for-pipeline -compact -variables 4 -pipeline-latency 4 -n 3 -name n1_3 -include n.h */ /* * This function contains 12 FP additions, 6 FP multiplications, * (or, 6 additions, 0 multiplications, 6 fused multiply/add), * 15 stack variables, 2 constants, and 12 memory accesses */ #include "n.h" static void n1_3(const R *ri, const R *ii, R *ro, R *io, stride is, stride os, INT v, INT ivs, INT ovs) { DK(KP866025403, +0.866025403784438646763723170752936183471402627); DK(KP500000000, +0.500000000000000000000000000000000000000000000); { INT i; for (i = v; i > 0; i = i - 1, ri = ri + ivs, ii = ii + ivs, ro = ro + ovs, io = io + ovs, MAKE_VOLATILE_STRIDE(12, is), MAKE_VOLATILE_STRIDE(12, os)) { E T1, T9, T2, T3, T6, T7; T1 = ri[0]; T9 = ii[0]; T2 = ri[WS(is, 1)]; T3 = ri[WS(is, 2)]; T6 = ii[WS(is, 1)]; T7 = ii[WS(is, 2)]; { E T4, Tc, T8, Ta, T5, Tb; T4 = T2 + T3; Tc = T3 - T2; T8 = T6 - T7; Ta = T6 + T7; T5 = FNMS(KP500000000, T4, T1); ro[0] = T1 + T4; Tb = FNMS(KP500000000, Ta, T9); io[0] = T9 + Ta; ro[WS(os, 1)] = FMA(KP866025403, T8, T5); ro[WS(os, 2)] = FNMS(KP866025403, T8, T5); io[WS(os, 2)] = FNMS(KP866025403, Tc, Tb); io[WS(os, 1)] = FMA(KP866025403, Tc, Tb); } } } } static const kdft_desc desc = { 3, "n1_3", {6, 0, 6, 0}, &GENUS, 0, 0, 0, 0 }; void X(codelet_n1_3) (planner *p) { X(kdft_register) (p, n1_3, &desc); } #else /* HAVE_FMA */ /* Generated by: ../../../genfft/gen_notw.native -compact -variables 4 -pipeline-latency 4 -n 3 -name n1_3 -include n.h */ /* * This function contains 12 FP additions, 4 FP multiplications, * (or, 10 additions, 2 multiplications, 2 fused multiply/add), * 15 stack variables, 2 constants, and 12 memory accesses */ #include "n.h" static void n1_3(const R *ri, const R *ii, R *ro, R *io, stride is, stride os, INT v, INT ivs, INT ovs) { DK(KP500000000, +0.500000000000000000000000000000000000000000000); DK(KP866025403, +0.866025403784438646763723170752936183471402627); { INT i; for (i = v; i > 0; i = i - 1, ri = ri + ivs, ii = ii + ivs, ro = ro + ovs, io = io + ovs, MAKE_VOLATILE_STRIDE(12, is), MAKE_VOLATILE_STRIDE(12, os)) { E T1, Ta, T4, T9, T8, Tb, T5, Tc; T1 = ri[0]; Ta = ii[0]; { E T2, T3, T6, T7; T2 = ri[WS(is, 1)]; T3 = ri[WS(is, 2)]; T4 = T2 + T3; T9 = KP866025403 * (T3 - T2); T6 = ii[WS(is, 1)]; T7 = ii[WS(is, 2)]; T8 = KP866025403 * (T6 - T7); Tb = T6 + T7; } ro[0] = T1 + T4; io[0] = Ta + Tb; T5 = FNMS(KP500000000, T4, T1); ro[WS(os, 2)] = T5 - T8; ro[WS(os, 1)] = T5 + T8; Tc = FNMS(KP500000000, Tb, Ta); io[WS(os, 1)] = T9 + Tc; io[WS(os, 2)] = Tc - T9; } } } static const kdft_desc desc = { 3, "n1_3", {10, 2, 2, 0}, &GENUS, 0, 0, 0, 0 }; void X(codelet_n1_3) (planner *p) { X(kdft_register) (p, n1_3, &desc); } #endif /* HAVE_FMA */ fftw-3.3.3/dft/scalar/codelets/n1_4.c0000644000175000017500000001001612054410236014133 00000000000000/* * Copyright (c) 2003, 2007-11 Matteo Frigo * Copyright (c) 2003, 2007-11 Massachusetts Institute of Technology * * This program is free software; you can redistribute it and/or modify * it under the terms of the GNU General Public License as published by * the Free Software Foundation; either version 2 of the License, or * (at your option) any later version. * * This program is distributed in the hope that it will be useful, * but WITHOUT ANY WARRANTY; without even the implied warranty of * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the * GNU General Public License for more details. * * You should have received a copy of the GNU General Public License * along with this program; if not, write to the Free Software * Foundation, Inc., 51 Franklin Street, Fifth Floor, Boston, MA 02110-1301 USA * */ /* This file was automatically generated --- DO NOT EDIT */ /* Generated on Sun Nov 25 07:35:42 EST 2012 */ #include "codelet-dft.h" #ifdef HAVE_FMA /* Generated by: ../../../genfft/gen_notw.native -fma -reorder-insns -schedule-for-pipeline -compact -variables 4 -pipeline-latency 4 -n 4 -name n1_4 -include n.h */ /* * This function contains 16 FP additions, 0 FP multiplications, * (or, 16 additions, 0 multiplications, 0 fused multiply/add), * 13 stack variables, 0 constants, and 16 memory accesses */ #include "n.h" static void n1_4(const R *ri, const R *ii, R *ro, R *io, stride is, stride os, INT v, INT ivs, INT ovs) { { INT i; for (i = v; i > 0; i = i - 1, ri = ri + ivs, ii = ii + ivs, ro = ro + ovs, io = io + ovs, MAKE_VOLATILE_STRIDE(16, is), MAKE_VOLATILE_STRIDE(16, os)) { E T4, Tb, T3, Tf, T9, T5, Tc, Td; { E T1, T2, T7, T8; T1 = ri[0]; T2 = ri[WS(is, 2)]; T7 = ii[0]; T8 = ii[WS(is, 2)]; T4 = ri[WS(is, 1)]; Tb = T1 - T2; T3 = T1 + T2; Tf = T7 + T8; T9 = T7 - T8; T5 = ri[WS(is, 3)]; Tc = ii[WS(is, 1)]; Td = ii[WS(is, 3)]; } { E T6, Ta, Te, Tg; T6 = T4 + T5; Ta = T4 - T5; Te = Tc - Td; Tg = Tc + Td; io[WS(os, 3)] = Ta + T9; io[WS(os, 1)] = T9 - Ta; ro[0] = T3 + T6; ro[WS(os, 2)] = T3 - T6; io[0] = Tf + Tg; io[WS(os, 2)] = Tf - Tg; ro[WS(os, 3)] = Tb - Te; ro[WS(os, 1)] = Tb + Te; } } } } static const kdft_desc desc = { 4, "n1_4", {16, 0, 0, 0}, &GENUS, 0, 0, 0, 0 }; void X(codelet_n1_4) (planner *p) { X(kdft_register) (p, n1_4, &desc); } #else /* HAVE_FMA */ /* Generated by: ../../../genfft/gen_notw.native -compact -variables 4 -pipeline-latency 4 -n 4 -name n1_4 -include n.h */ /* * This function contains 16 FP additions, 0 FP multiplications, * (or, 16 additions, 0 multiplications, 0 fused multiply/add), * 13 stack variables, 0 constants, and 16 memory accesses */ #include "n.h" static void n1_4(const R *ri, const R *ii, R *ro, R *io, stride is, stride os, INT v, INT ivs, INT ovs) { { INT i; for (i = v; i > 0; i = i - 1, ri = ri + ivs, ii = ii + ivs, ro = ro + ovs, io = io + ovs, MAKE_VOLATILE_STRIDE(16, is), MAKE_VOLATILE_STRIDE(16, os)) { E T3, Tb, T9, Tf, T6, Ta, Te, Tg; { E T1, T2, T7, T8; T1 = ri[0]; T2 = ri[WS(is, 2)]; T3 = T1 + T2; Tb = T1 - T2; T7 = ii[0]; T8 = ii[WS(is, 2)]; T9 = T7 - T8; Tf = T7 + T8; } { E T4, T5, Tc, Td; T4 = ri[WS(is, 1)]; T5 = ri[WS(is, 3)]; T6 = T4 + T5; Ta = T4 - T5; Tc = ii[WS(is, 1)]; Td = ii[WS(is, 3)]; Te = Tc - Td; Tg = Tc + Td; } ro[WS(os, 2)] = T3 - T6; io[WS(os, 2)] = Tf - Tg; ro[0] = T3 + T6; io[0] = Tf + Tg; io[WS(os, 1)] = T9 - Ta; ro[WS(os, 1)] = Tb + Te; io[WS(os, 3)] = Ta + T9; ro[WS(os, 3)] = Tb - Te; } } } static const kdft_desc desc = { 4, "n1_4", {16, 0, 0, 0}, &GENUS, 0, 0, 0, 0 }; void X(codelet_n1_4) (planner *p) { X(kdft_register) (p, n1_4, &desc); } #endif /* HAVE_FMA */ fftw-3.3.3/dft/scalar/codelets/n1_5.c0000644000175000017500000001413312054410236014140 00000000000000/* * Copyright (c) 2003, 2007-11 Matteo Frigo * Copyright (c) 2003, 2007-11 Massachusetts Institute of Technology * * This program is free software; you can redistribute it and/or modify * it under the terms of the GNU General Public License as published by * the Free Software Foundation; either version 2 of the License, or * (at your option) any later version. * * This program is distributed in the hope that it will be useful, * but WITHOUT ANY WARRANTY; without even the implied warranty of * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the * GNU General Public License for more details. * * You should have received a copy of the GNU General Public License * along with this program; if not, write to the Free Software * Foundation, Inc., 51 Franklin Street, Fifth Floor, Boston, MA 02110-1301 USA * */ /* This file was automatically generated --- DO NOT EDIT */ /* Generated on Sun Nov 25 07:35:42 EST 2012 */ #include "codelet-dft.h" #ifdef HAVE_FMA /* Generated by: ../../../genfft/gen_notw.native -fma -reorder-insns -schedule-for-pipeline -compact -variables 4 -pipeline-latency 4 -n 5 -name n1_5 -include n.h */ /* * This function contains 32 FP additions, 18 FP multiplications, * (or, 14 additions, 0 multiplications, 18 fused multiply/add), * 37 stack variables, 4 constants, and 20 memory accesses */ #include "n.h" static void n1_5(const R *ri, const R *ii, R *ro, R *io, stride is, stride os, INT v, INT ivs, INT ovs) { DK(KP951056516, +0.951056516295153572116439333379382143405698634); DK(KP559016994, +0.559016994374947424102293417182819058860154590); DK(KP250000000, +0.250000000000000000000000000000000000000000000); DK(KP618033988, +0.618033988749894848204586834365638117720309180); { INT i; for (i = v; i > 0; i = i - 1, ri = ri + ivs, ii = ii + ivs, ro = ro + ovs, io = io + ovs, MAKE_VOLATILE_STRIDE(20, is), MAKE_VOLATILE_STRIDE(20, os)) { E Tq, Ti, Tk, Tu, Tw, Tp, Tb, Tj, Tr, Tv; { E T1, Tl, Ts, Tt, T8, Ta, Te, Tm, Tn, Th, To, T9; T1 = ri[0]; Tl = ii[0]; { E T2, T3, T5, T6; T2 = ri[WS(is, 1)]; T3 = ri[WS(is, 4)]; T5 = ri[WS(is, 2)]; T6 = ri[WS(is, 3)]; { E Tc, T4, T7, Td, Tf, Tg; Tc = ii[WS(is, 1)]; Ts = T2 - T3; T4 = T2 + T3; Tt = T5 - T6; T7 = T5 + T6; Td = ii[WS(is, 4)]; Tf = ii[WS(is, 2)]; Tg = ii[WS(is, 3)]; T8 = T4 + T7; Ta = T4 - T7; Te = Tc - Td; Tm = Tc + Td; Tn = Tf + Tg; Th = Tf - Tg; } } ro[0] = T1 + T8; To = Tm + Tn; Tq = Tm - Tn; Ti = FMA(KP618033988, Th, Te); Tk = FNMS(KP618033988, Te, Th); io[0] = Tl + To; T9 = FNMS(KP250000000, T8, T1); Tu = FMA(KP618033988, Tt, Ts); Tw = FNMS(KP618033988, Ts, Tt); Tp = FNMS(KP250000000, To, Tl); Tb = FMA(KP559016994, Ta, T9); Tj = FNMS(KP559016994, Ta, T9); } Tr = FMA(KP559016994, Tq, Tp); Tv = FNMS(KP559016994, Tq, Tp); ro[WS(os, 2)] = FNMS(KP951056516, Tk, Tj); ro[WS(os, 3)] = FMA(KP951056516, Tk, Tj); ro[WS(os, 1)] = FMA(KP951056516, Ti, Tb); ro[WS(os, 4)] = FNMS(KP951056516, Ti, Tb); io[WS(os, 2)] = FMA(KP951056516, Tw, Tv); io[WS(os, 3)] = FNMS(KP951056516, Tw, Tv); io[WS(os, 4)] = FMA(KP951056516, Tu, Tr); io[WS(os, 1)] = FNMS(KP951056516, Tu, Tr); } } } static const kdft_desc desc = { 5, "n1_5", {14, 0, 18, 0}, &GENUS, 0, 0, 0, 0 }; void X(codelet_n1_5) (planner *p) { X(kdft_register) (p, n1_5, &desc); } #else /* HAVE_FMA */ /* Generated by: ../../../genfft/gen_notw.native -compact -variables 4 -pipeline-latency 4 -n 5 -name n1_5 -include n.h */ /* * This function contains 32 FP additions, 12 FP multiplications, * (or, 26 additions, 6 multiplications, 6 fused multiply/add), * 21 stack variables, 4 constants, and 20 memory accesses */ #include "n.h" static void n1_5(const R *ri, const R *ii, R *ro, R *io, stride is, stride os, INT v, INT ivs, INT ovs) { DK(KP250000000, +0.250000000000000000000000000000000000000000000); DK(KP587785252, +0.587785252292473129168705954639072768597652438); DK(KP951056516, +0.951056516295153572116439333379382143405698634); DK(KP559016994, +0.559016994374947424102293417182819058860154590); { INT i; for (i = v; i > 0; i = i - 1, ri = ri + ivs, ii = ii + ivs, ro = ro + ovs, io = io + ovs, MAKE_VOLATILE_STRIDE(20, is), MAKE_VOLATILE_STRIDE(20, os)) { E T1, To, T8, Tt, T9, Ts, Te, Tp, Th, Tn; T1 = ri[0]; To = ii[0]; { E T2, T3, T4, T5, T6, T7; T2 = ri[WS(is, 1)]; T3 = ri[WS(is, 4)]; T4 = T2 + T3; T5 = ri[WS(is, 2)]; T6 = ri[WS(is, 3)]; T7 = T5 + T6; T8 = T4 + T7; Tt = T5 - T6; T9 = KP559016994 * (T4 - T7); Ts = T2 - T3; } { E Tc, Td, Tl, Tf, Tg, Tm; Tc = ii[WS(is, 1)]; Td = ii[WS(is, 4)]; Tl = Tc + Td; Tf = ii[WS(is, 2)]; Tg = ii[WS(is, 3)]; Tm = Tf + Tg; Te = Tc - Td; Tp = Tl + Tm; Th = Tf - Tg; Tn = KP559016994 * (Tl - Tm); } ro[0] = T1 + T8; io[0] = To + Tp; { E Ti, Tk, Tb, Tj, Ta; Ti = FMA(KP951056516, Te, KP587785252 * Th); Tk = FNMS(KP587785252, Te, KP951056516 * Th); Ta = FNMS(KP250000000, T8, T1); Tb = T9 + Ta; Tj = Ta - T9; ro[WS(os, 4)] = Tb - Ti; ro[WS(os, 3)] = Tj + Tk; ro[WS(os, 1)] = Tb + Ti; ro[WS(os, 2)] = Tj - Tk; } { E Tu, Tv, Tr, Tw, Tq; Tu = FMA(KP951056516, Ts, KP587785252 * Tt); Tv = FNMS(KP587785252, Ts, KP951056516 * Tt); Tq = FNMS(KP250000000, Tp, To); Tr = Tn + Tq; Tw = Tq - Tn; io[WS(os, 1)] = Tr - Tu; io[WS(os, 3)] = Tw - Tv; io[WS(os, 4)] = Tu + Tr; io[WS(os, 2)] = Tv + Tw; } } } } static const kdft_desc desc = { 5, "n1_5", {26, 6, 6, 0}, &GENUS, 0, 0, 0, 0 }; void X(codelet_n1_5) (planner *p) { X(kdft_register) (p, n1_5, &desc); } #endif /* HAVE_FMA */ fftw-3.3.3/dft/scalar/codelets/n1_6.c0000644000175000017500000001415212054410236014142 00000000000000/* * Copyright (c) 2003, 2007-11 Matteo Frigo * Copyright (c) 2003, 2007-11 Massachusetts Institute of Technology * * This program is free software; you can redistribute it and/or modify * it under the terms of the GNU General Public License as published by * the Free Software Foundation; either version 2 of the License, or * (at your option) any later version. * * This program is distributed in the hope that it will be useful, * but WITHOUT ANY WARRANTY; without even the implied warranty of * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the * GNU General Public License for more details. * * You should have received a copy of the GNU General Public License * along with this program; if not, write to the Free Software * Foundation, Inc., 51 Franklin Street, Fifth Floor, Boston, MA 02110-1301 USA * */ /* This file was automatically generated --- DO NOT EDIT */ /* Generated on Sun Nov 25 07:35:42 EST 2012 */ #include "codelet-dft.h" #ifdef HAVE_FMA /* Generated by: ../../../genfft/gen_notw.native -fma -reorder-insns -schedule-for-pipeline -compact -variables 4 -pipeline-latency 4 -n 6 -name n1_6 -include n.h */ /* * This function contains 36 FP additions, 12 FP multiplications, * (or, 24 additions, 0 multiplications, 12 fused multiply/add), * 30 stack variables, 2 constants, and 24 memory accesses */ #include "n.h" static void n1_6(const R *ri, const R *ii, R *ro, R *io, stride is, stride os, INT v, INT ivs, INT ovs) { DK(KP866025403, +0.866025403784438646763723170752936183471402627); DK(KP500000000, +0.500000000000000000000000000000000000000000000); { INT i; for (i = v; i > 0; i = i - 1, ri = ri + ivs, ii = ii + ivs, ro = ro + ovs, io = io + ovs, MAKE_VOLATILE_STRIDE(24, is), MAKE_VOLATILE_STRIDE(24, os)) { E TA, Tz; { E Tb, T3, Tx, Tp, Tj, Te, Ts, Ta, Tu, Ti, Tk; { E T1, T2, Tn, To; T1 = ri[0]; T2 = ri[WS(is, 3)]; Tn = ii[0]; To = ii[WS(is, 3)]; { E T4, T5, T7, T8; T4 = ri[WS(is, 2)]; Tb = T1 + T2; T3 = T1 - T2; Tx = Tn + To; Tp = Tn - To; T5 = ri[WS(is, 5)]; T7 = ri[WS(is, 4)]; T8 = ri[WS(is, 1)]; { E Tg, Tc, T6, Td, T9, Th; Tg = ii[WS(is, 2)]; Tc = T4 + T5; T6 = T4 - T5; Td = T7 + T8; T9 = T7 - T8; Th = ii[WS(is, 5)]; Tj = ii[WS(is, 4)]; Te = Tc + Td; TA = Td - Tc; Ts = T9 - T6; Ta = T6 + T9; Tu = Tg + Th; Ti = Tg - Th; Tk = ii[WS(is, 1)]; } } } ro[WS(os, 3)] = T3 + Ta; ro[0] = Tb + Te; { E Tf, Tv, Tl, Ty, Tr; Tf = FNMS(KP500000000, Ta, T3); Tv = Tj + Tk; Tl = Tj - Tk; { E Tt, Tw, Tq, Tm; Tt = FNMS(KP500000000, Te, Tb); Ty = Tu + Tv; Tw = Tu - Tv; Tq = Ti + Tl; Tm = Ti - Tl; io[0] = Tx + Ty; ro[WS(os, 1)] = FMA(KP866025403, Tm, Tf); ro[WS(os, 5)] = FNMS(KP866025403, Tm, Tf); Tr = FNMS(KP500000000, Tq, Tp); io[WS(os, 3)] = Tp + Tq; ro[WS(os, 2)] = FNMS(KP866025403, Tw, Tt); ro[WS(os, 4)] = FMA(KP866025403, Tw, Tt); } io[WS(os, 5)] = FNMS(KP866025403, Ts, Tr); io[WS(os, 1)] = FMA(KP866025403, Ts, Tr); Tz = FNMS(KP500000000, Ty, Tx); } } io[WS(os, 4)] = FMA(KP866025403, TA, Tz); io[WS(os, 2)] = FNMS(KP866025403, TA, Tz); } } } static const kdft_desc desc = { 6, "n1_6", {24, 0, 12, 0}, &GENUS, 0, 0, 0, 0 }; void X(codelet_n1_6) (planner *p) { X(kdft_register) (p, n1_6, &desc); } #else /* HAVE_FMA */ /* Generated by: ../../../genfft/gen_notw.native -compact -variables 4 -pipeline-latency 4 -n 6 -name n1_6 -include n.h */ /* * This function contains 36 FP additions, 8 FP multiplications, * (or, 32 additions, 4 multiplications, 4 fused multiply/add), * 23 stack variables, 2 constants, and 24 memory accesses */ #include "n.h" static void n1_6(const R *ri, const R *ii, R *ro, R *io, stride is, stride os, INT v, INT ivs, INT ovs) { DK(KP866025403, +0.866025403784438646763723170752936183471402627); DK(KP500000000, +0.500000000000000000000000000000000000000000000); { INT i; for (i = v; i > 0; i = i - 1, ri = ri + ivs, ii = ii + ivs, ro = ro + ovs, io = io + ovs, MAKE_VOLATILE_STRIDE(24, is), MAKE_VOLATILE_STRIDE(24, os)) { E T3, Tb, Tq, Tx, T6, Tc, T9, Td, Ta, Te, Ti, Tu, Tl, Tv, Tr; E Ty; { E T1, T2, To, Tp; T1 = ri[0]; T2 = ri[WS(is, 3)]; T3 = T1 - T2; Tb = T1 + T2; To = ii[0]; Tp = ii[WS(is, 3)]; Tq = To - Tp; Tx = To + Tp; } { E T4, T5, T7, T8; T4 = ri[WS(is, 2)]; T5 = ri[WS(is, 5)]; T6 = T4 - T5; Tc = T4 + T5; T7 = ri[WS(is, 4)]; T8 = ri[WS(is, 1)]; T9 = T7 - T8; Td = T7 + T8; } Ta = T6 + T9; Te = Tc + Td; { E Tg, Th, Tj, Tk; Tg = ii[WS(is, 2)]; Th = ii[WS(is, 5)]; Ti = Tg - Th; Tu = Tg + Th; Tj = ii[WS(is, 4)]; Tk = ii[WS(is, 1)]; Tl = Tj - Tk; Tv = Tj + Tk; } Tr = Ti + Tl; Ty = Tu + Tv; ro[WS(os, 3)] = T3 + Ta; io[WS(os, 3)] = Tq + Tr; ro[0] = Tb + Te; io[0] = Tx + Ty; { E Tf, Tm, Tn, Ts; Tf = FNMS(KP500000000, Ta, T3); Tm = KP866025403 * (Ti - Tl); ro[WS(os, 5)] = Tf - Tm; ro[WS(os, 1)] = Tf + Tm; Tn = KP866025403 * (T9 - T6); Ts = FNMS(KP500000000, Tr, Tq); io[WS(os, 1)] = Tn + Ts; io[WS(os, 5)] = Ts - Tn; } { E Tt, Tw, Tz, TA; Tt = FNMS(KP500000000, Te, Tb); Tw = KP866025403 * (Tu - Tv); ro[WS(os, 2)] = Tt - Tw; ro[WS(os, 4)] = Tt + Tw; Tz = FNMS(KP500000000, Ty, Tx); TA = KP866025403 * (Td - Tc); io[WS(os, 2)] = Tz - TA; io[WS(os, 4)] = TA + Tz; } } } } static const kdft_desc desc = { 6, "n1_6", {32, 4, 4, 0}, &GENUS, 0, 0, 0, 0 }; void X(codelet_n1_6) (planner *p) { X(kdft_register) (p, n1_6, &desc); } #endif /* HAVE_FMA */ fftw-3.3.3/dft/scalar/codelets/n1_7.c0000644000175000017500000002071212054410237014143 00000000000000/* * Copyright (c) 2003, 2007-11 Matteo Frigo * Copyright (c) 2003, 2007-11 Massachusetts Institute of Technology * * This program is free software; you can redistribute it and/or modify * it under the terms of the GNU General Public License as published by * the Free Software Foundation; either version 2 of the License, or * (at your option) any later version. * * This program is distributed in the hope that it will be useful, * but WITHOUT ANY WARRANTY; without even the implied warranty of * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the * GNU General Public License for more details. * * You should have received a copy of the GNU General Public License * along with this program; if not, write to the Free Software * Foundation, Inc., 51 Franklin Street, Fifth Floor, Boston, MA 02110-1301 USA * */ /* This file was automatically generated --- DO NOT EDIT */ /* Generated on Sun Nov 25 07:35:42 EST 2012 */ #include "codelet-dft.h" #ifdef HAVE_FMA /* Generated by: ../../../genfft/gen_notw.native -fma -reorder-insns -schedule-for-pipeline -compact -variables 4 -pipeline-latency 4 -n 7 -name n1_7 -include n.h */ /* * This function contains 60 FP additions, 42 FP multiplications, * (or, 18 additions, 0 multiplications, 42 fused multiply/add), * 51 stack variables, 6 constants, and 28 memory accesses */ #include "n.h" static void n1_7(const R *ri, const R *ii, R *ro, R *io, stride is, stride os, INT v, INT ivs, INT ovs) { DK(KP974927912, +0.974927912181823607018131682993931217232785801); DK(KP900968867, +0.900968867902419126236102319507445051165919162); DK(KP801937735, +0.801937735804838252472204639014890102331838324); DK(KP692021471, +0.692021471630095869627814897002069140197260599); DK(KP356895867, +0.356895867892209443894399510021300583399127187); DK(KP554958132, +0.554958132087371191422194871006410481067288862); { INT i; for (i = v; i > 0; i = i - 1, ri = ri + ivs, ii = ii + ivs, ro = ro + ovs, io = io + ovs, MAKE_VOLATILE_STRIDE(28, is), MAKE_VOLATILE_STRIDE(28, os)) { E Tz, TP, Ty, TK, TN, TE, Tw, TF; { E T1, TI, T4, TG, Ta, TT, Tp, TH, T7, Tk, TJ, TO, Tu, Tb, TB; E Tg, Tl, Th, Ti; T1 = ri[0]; Tz = ii[0]; { E T5, T6, Te, Tf; { E T2, T3, T8, T9; T2 = ri[WS(is, 1)]; T3 = ri[WS(is, 6)]; T8 = ri[WS(is, 3)]; T9 = ri[WS(is, 4)]; T5 = ri[WS(is, 2)]; TI = T3 - T2; T4 = T2 + T3; TG = T9 - T8; Ta = T8 + T9; T6 = ri[WS(is, 5)]; } Te = ii[WS(is, 2)]; TT = FMA(KP554958132, TG, TI); Tp = FNMS(KP356895867, T4, Ta); TH = T6 - T5; T7 = T5 + T6; Tf = ii[WS(is, 5)]; Tk = ii[WS(is, 3)]; TJ = FNMS(KP554958132, TI, TH); TO = FMA(KP554958132, TH, TG); Tu = FNMS(KP356895867, Ta, T7); Tb = FNMS(KP356895867, T7, T4); TB = Te + Tf; Tg = Te - Tf; Tl = ii[WS(is, 4)]; Th = ii[WS(is, 1)]; Ti = ii[WS(is, 6)]; } { E Tm, TA, Tj, TD, Ts, TL, Tx, TU, To, TR, Td, TM, Tv; { E TC, TQ, Tn, Tc; ro[0] = T1 + T4 + T7 + Ta; TC = Tk + Tl; Tm = Tk - Tl; TA = Th + Ti; Tj = Th - Ti; TD = FNMS(KP356895867, TC, TB); Ts = FMA(KP554958132, Tg, Tm); TL = FNMS(KP356895867, TA, TC); TQ = FNMS(KP356895867, TB, TA); Tx = FNMS(KP554958132, Tj, Tg); Tn = FMA(KP554958132, Tm, Tj); io[0] = Tz + TA + TB + TC; Tc = FNMS(KP692021471, Tb, Ta); TU = FMA(KP801937735, TT, TH); To = FMA(KP801937735, Tn, Tg); TR = FNMS(KP692021471, TQ, TC); Td = FNMS(KP900968867, Tc, T1); } { E Tt, Tr, TS, Tq; Tt = FNMS(KP801937735, Ts, Tj); Tq = FNMS(KP692021471, Tp, T7); TS = FNMS(KP900968867, TR, Tz); ro[WS(os, 1)] = FMA(KP974927912, To, Td); ro[WS(os, 6)] = FNMS(KP974927912, To, Td); Tr = FNMS(KP900968867, Tq, T1); io[WS(os, 6)] = FNMS(KP974927912, TU, TS); io[WS(os, 1)] = FMA(KP974927912, TU, TS); TP = FNMS(KP801937735, TO, TI); ro[WS(os, 2)] = FMA(KP974927912, Tt, Tr); ro[WS(os, 5)] = FNMS(KP974927912, Tt, Tr); TM = FNMS(KP692021471, TL, TB); } Ty = FNMS(KP801937735, Tx, Tm); Tv = FNMS(KP692021471, Tu, T4); TK = FNMS(KP801937735, TJ, TG); TN = FNMS(KP900968867, TM, Tz); TE = FNMS(KP692021471, TD, TA); Tw = FNMS(KP900968867, Tv, T1); } } io[WS(os, 5)] = FNMS(KP974927912, TP, TN); io[WS(os, 2)] = FMA(KP974927912, TP, TN); TF = FNMS(KP900968867, TE, Tz); ro[WS(os, 3)] = FMA(KP974927912, Ty, Tw); ro[WS(os, 4)] = FNMS(KP974927912, Ty, Tw); io[WS(os, 4)] = FNMS(KP974927912, TK, TF); io[WS(os, 3)] = FMA(KP974927912, TK, TF); } } } static const kdft_desc desc = { 7, "n1_7", {18, 0, 42, 0}, &GENUS, 0, 0, 0, 0 }; void X(codelet_n1_7) (planner *p) { X(kdft_register) (p, n1_7, &desc); } #else /* HAVE_FMA */ /* Generated by: ../../../genfft/gen_notw.native -compact -variables 4 -pipeline-latency 4 -n 7 -name n1_7 -include n.h */ /* * This function contains 60 FP additions, 36 FP multiplications, * (or, 36 additions, 12 multiplications, 24 fused multiply/add), * 25 stack variables, 6 constants, and 28 memory accesses */ #include "n.h" static void n1_7(const R *ri, const R *ii, R *ro, R *io, stride is, stride os, INT v, INT ivs, INT ovs) { DK(KP222520933, +0.222520933956314404288902564496794759466355569); DK(KP900968867, +0.900968867902419126236102319507445051165919162); DK(KP623489801, +0.623489801858733530525004884004239810632274731); DK(KP433883739, +0.433883739117558120475768332848358754609990728); DK(KP781831482, +0.781831482468029808708444526674057750232334519); DK(KP974927912, +0.974927912181823607018131682993931217232785801); { INT i; for (i = v; i > 0; i = i - 1, ri = ri + ivs, ii = ii + ivs, ro = ro + ovs, io = io + ovs, MAKE_VOLATILE_STRIDE(28, is), MAKE_VOLATILE_STRIDE(28, os)) { E T1, Tu, T4, Tq, Te, Tx, T7, Ts, Tk, Tv, Ta, Tr, Th, Tw; T1 = ri[0]; Tu = ii[0]; { E T2, T3, Tc, Td; T2 = ri[WS(is, 1)]; T3 = ri[WS(is, 6)]; T4 = T2 + T3; Tq = T3 - T2; Tc = ii[WS(is, 1)]; Td = ii[WS(is, 6)]; Te = Tc - Td; Tx = Tc + Td; } { E T5, T6, Ti, Tj; T5 = ri[WS(is, 2)]; T6 = ri[WS(is, 5)]; T7 = T5 + T6; Ts = T6 - T5; Ti = ii[WS(is, 2)]; Tj = ii[WS(is, 5)]; Tk = Ti - Tj; Tv = Ti + Tj; } { E T8, T9, Tf, Tg; T8 = ri[WS(is, 3)]; T9 = ri[WS(is, 4)]; Ta = T8 + T9; Tr = T9 - T8; Tf = ii[WS(is, 3)]; Tg = ii[WS(is, 4)]; Th = Tf - Tg; Tw = Tf + Tg; } ro[0] = T1 + T4 + T7 + Ta; io[0] = Tu + Tx + Tv + Tw; { E Tl, Tb, TB, TC; Tl = FNMS(KP781831482, Th, KP974927912 * Te) - (KP433883739 * Tk); Tb = FMA(KP623489801, Ta, T1) + FNMA(KP900968867, T7, KP222520933 * T4); ro[WS(os, 5)] = Tb - Tl; ro[WS(os, 2)] = Tb + Tl; TB = FNMS(KP781831482, Tr, KP974927912 * Tq) - (KP433883739 * Ts); TC = FMA(KP623489801, Tw, Tu) + FNMA(KP900968867, Tv, KP222520933 * Tx); io[WS(os, 2)] = TB + TC; io[WS(os, 5)] = TC - TB; } { E Tn, Tm, Tz, TA; Tn = FMA(KP781831482, Te, KP974927912 * Tk) + (KP433883739 * Th); Tm = FMA(KP623489801, T4, T1) + FNMA(KP900968867, Ta, KP222520933 * T7); ro[WS(os, 6)] = Tm - Tn; ro[WS(os, 1)] = Tm + Tn; Tz = FMA(KP781831482, Tq, KP974927912 * Ts) + (KP433883739 * Tr); TA = FMA(KP623489801, Tx, Tu) + FNMA(KP900968867, Tw, KP222520933 * Tv); io[WS(os, 1)] = Tz + TA; io[WS(os, 6)] = TA - Tz; } { E Tp, To, Tt, Ty; Tp = FMA(KP433883739, Te, KP974927912 * Th) - (KP781831482 * Tk); To = FMA(KP623489801, T7, T1) + FNMA(KP222520933, Ta, KP900968867 * T4); ro[WS(os, 4)] = To - Tp; ro[WS(os, 3)] = To + Tp; Tt = FMA(KP433883739, Tq, KP974927912 * Tr) - (KP781831482 * Ts); Ty = FMA(KP623489801, Tv, Tu) + FNMA(KP222520933, Tw, KP900968867 * Tx); io[WS(os, 3)] = Tt + Ty; io[WS(os, 4)] = Ty - Tt; } } } } static const kdft_desc desc = { 7, "n1_7", {36, 12, 24, 0}, &GENUS, 0, 0, 0, 0 }; void X(codelet_n1_7) (planner *p) { X(kdft_register) (p, n1_7, &desc); } #endif /* HAVE_FMA */ fftw-3.3.3/dft/scalar/codelets/n1_8.c0000644000175000017500000001600712054410237014146 00000000000000/* * Copyright (c) 2003, 2007-11 Matteo Frigo * Copyright (c) 2003, 2007-11 Massachusetts Institute of Technology * * This program is free software; you can redistribute it and/or modify * it under the terms of the GNU General Public License as published by * the Free Software Foundation; either version 2 of the License, or * (at your option) any later version. * * This program is distributed in the hope that it will be useful, * but WITHOUT ANY WARRANTY; without even the implied warranty of * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the * GNU General Public License for more details. * * You should have received a copy of the GNU General Public License * along with this program; if not, write to the Free Software * Foundation, Inc., 51 Franklin Street, Fifth Floor, Boston, MA 02110-1301 USA * */ /* This file was automatically generated --- DO NOT EDIT */ /* Generated on Sun Nov 25 07:35:42 EST 2012 */ #include "codelet-dft.h" #ifdef HAVE_FMA /* Generated by: ../../../genfft/gen_notw.native -fma -reorder-insns -schedule-for-pipeline -compact -variables 4 -pipeline-latency 4 -n 8 -name n1_8 -include n.h */ /* * This function contains 52 FP additions, 8 FP multiplications, * (or, 44 additions, 0 multiplications, 8 fused multiply/add), * 36 stack variables, 1 constants, and 32 memory accesses */ #include "n.h" static void n1_8(const R *ri, const R *ii, R *ro, R *io, stride is, stride os, INT v, INT ivs, INT ovs) { DK(KP707106781, +0.707106781186547524400844362104849039284835938); { INT i; for (i = v; i > 0; i = i - 1, ri = ri + ivs, ii = ii + ivs, ro = ro + ovs, io = io + ovs, MAKE_VOLATILE_STRIDE(32, is), MAKE_VOLATILE_STRIDE(32, os)) { E TF, TE, TD, TI; { E Tn, T3, TC, Ti, TB, T6, To, Tl, Td, TN, Tz, TH, Ta, Tq, Tt; E TM; { E T4, T5, Tj, Tk; { E T1, T2, Tg, Th; T1 = ri[0]; T2 = ri[WS(is, 4)]; Tg = ii[0]; Th = ii[WS(is, 4)]; T4 = ri[WS(is, 2)]; Tn = T1 - T2; T3 = T1 + T2; TC = Tg - Th; Ti = Tg + Th; T5 = ri[WS(is, 6)]; } Tj = ii[WS(is, 2)]; Tk = ii[WS(is, 6)]; { E Tb, Tc, Tw, Tx; Tb = ri[WS(is, 7)]; TB = T4 - T5; T6 = T4 + T5; To = Tj - Tk; Tl = Tj + Tk; Tc = ri[WS(is, 3)]; Tw = ii[WS(is, 7)]; Tx = ii[WS(is, 3)]; { E T8, Tv, Ty, T9, Tr, Ts; T8 = ri[WS(is, 1)]; Td = Tb + Tc; Tv = Tb - Tc; TN = Tw + Tx; Ty = Tw - Tx; T9 = ri[WS(is, 5)]; Tr = ii[WS(is, 1)]; Ts = ii[WS(is, 5)]; Tz = Tv - Ty; TH = Tv + Ty; Ta = T8 + T9; Tq = T8 - T9; Tt = Tr - Ts; TM = Tr + Ts; } } } { E TL, TG, Tu, Tf, Tm, TO; { E T7, Te, TP, TQ; TL = T3 - T6; T7 = T3 + T6; TG = Tt - Tq; Tu = Tq + Tt; Te = Ta + Td; Tf = Td - Ta; Tm = Ti - Tl; TP = Ti + Tl; TQ = TM + TN; TO = TM - TN; ro[0] = T7 + Te; ro[WS(os, 4)] = T7 - Te; io[0] = TP + TQ; io[WS(os, 4)] = TP - TQ; } { E Tp, TA, TJ, TK; TF = Tn - To; Tp = Tn + To; io[WS(os, 6)] = Tm - Tf; io[WS(os, 2)] = Tf + Tm; ro[WS(os, 2)] = TL + TO; ro[WS(os, 6)] = TL - TO; TA = Tu + Tz; TE = Tz - Tu; TD = TB + TC; TJ = TC - TB; TK = TG + TH; TI = TG - TH; ro[WS(os, 1)] = FMA(KP707106781, TA, Tp); ro[WS(os, 5)] = FNMS(KP707106781, TA, Tp); io[WS(os, 1)] = FMA(KP707106781, TK, TJ); io[WS(os, 5)] = FNMS(KP707106781, TK, TJ); } } } io[WS(os, 3)] = FMA(KP707106781, TE, TD); io[WS(os, 7)] = FNMS(KP707106781, TE, TD); ro[WS(os, 3)] = FMA(KP707106781, TI, TF); ro[WS(os, 7)] = FNMS(KP707106781, TI, TF); } } } static const kdft_desc desc = { 8, "n1_8", {44, 0, 8, 0}, &GENUS, 0, 0, 0, 0 }; void X(codelet_n1_8) (planner *p) { X(kdft_register) (p, n1_8, &desc); } #else /* HAVE_FMA */ /* Generated by: ../../../genfft/gen_notw.native -compact -variables 4 -pipeline-latency 4 -n 8 -name n1_8 -include n.h */ /* * This function contains 52 FP additions, 4 FP multiplications, * (or, 52 additions, 4 multiplications, 0 fused multiply/add), * 28 stack variables, 1 constants, and 32 memory accesses */ #include "n.h" static void n1_8(const R *ri, const R *ii, R *ro, R *io, stride is, stride os, INT v, INT ivs, INT ovs) { DK(KP707106781, +0.707106781186547524400844362104849039284835938); { INT i; for (i = v; i > 0; i = i - 1, ri = ri + ivs, ii = ii + ivs, ro = ro + ovs, io = io + ovs, MAKE_VOLATILE_STRIDE(32, is), MAKE_VOLATILE_STRIDE(32, os)) { E T3, Tn, Ti, TC, T6, TB, Tl, To, Td, TN, Tz, TH, Ta, TM, Tu; E TG; { E T1, T2, Tj, Tk; T1 = ri[0]; T2 = ri[WS(is, 4)]; T3 = T1 + T2; Tn = T1 - T2; { E Tg, Th, T4, T5; Tg = ii[0]; Th = ii[WS(is, 4)]; Ti = Tg + Th; TC = Tg - Th; T4 = ri[WS(is, 2)]; T5 = ri[WS(is, 6)]; T6 = T4 + T5; TB = T4 - T5; } Tj = ii[WS(is, 2)]; Tk = ii[WS(is, 6)]; Tl = Tj + Tk; To = Tj - Tk; { E Tb, Tc, Tv, Tw, Tx, Ty; Tb = ri[WS(is, 7)]; Tc = ri[WS(is, 3)]; Tv = Tb - Tc; Tw = ii[WS(is, 7)]; Tx = ii[WS(is, 3)]; Ty = Tw - Tx; Td = Tb + Tc; TN = Tw + Tx; Tz = Tv - Ty; TH = Tv + Ty; } { E T8, T9, Tq, Tr, Ts, Tt; T8 = ri[WS(is, 1)]; T9 = ri[WS(is, 5)]; Tq = T8 - T9; Tr = ii[WS(is, 1)]; Ts = ii[WS(is, 5)]; Tt = Tr - Ts; Ta = T8 + T9; TM = Tr + Ts; Tu = Tq + Tt; TG = Tt - Tq; } } { E T7, Te, TP, TQ; T7 = T3 + T6; Te = Ta + Td; ro[WS(os, 4)] = T7 - Te; ro[0] = T7 + Te; TP = Ti + Tl; TQ = TM + TN; io[WS(os, 4)] = TP - TQ; io[0] = TP + TQ; } { E Tf, Tm, TL, TO; Tf = Td - Ta; Tm = Ti - Tl; io[WS(os, 2)] = Tf + Tm; io[WS(os, 6)] = Tm - Tf; TL = T3 - T6; TO = TM - TN; ro[WS(os, 6)] = TL - TO; ro[WS(os, 2)] = TL + TO; } { E Tp, TA, TJ, TK; Tp = Tn + To; TA = KP707106781 * (Tu + Tz); ro[WS(os, 5)] = Tp - TA; ro[WS(os, 1)] = Tp + TA; TJ = TC - TB; TK = KP707106781 * (TG + TH); io[WS(os, 5)] = TJ - TK; io[WS(os, 1)] = TJ + TK; } { E TD, TE, TF, TI; TD = TB + TC; TE = KP707106781 * (Tz - Tu); io[WS(os, 7)] = TD - TE; io[WS(os, 3)] = TD + TE; TF = Tn - To; TI = KP707106781 * (TG - TH); ro[WS(os, 7)] = TF - TI; ro[WS(os, 3)] = TF + TI; } } } } static const kdft_desc desc = { 8, "n1_8", {52, 4, 0, 0}, &GENUS, 0, 0, 0, 0 }; void X(codelet_n1_8) (planner *p) { X(kdft_register) (p, n1_8, &desc); } #endif /* HAVE_FMA */ fftw-3.3.3/dft/scalar/codelets/n1_9.c0000644000175000017500000002647312054410237014157 00000000000000/* * Copyright (c) 2003, 2007-11 Matteo Frigo * Copyright (c) 2003, 2007-11 Massachusetts Institute of Technology * * This program is free software; you can redistribute it and/or modify * it under the terms of the GNU General Public License as published by * the Free Software Foundation; either version 2 of the License, or * (at your option) any later version. * * This program is distributed in the hope that it will be useful, * but WITHOUT ANY WARRANTY; without even the implied warranty of * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the * GNU General Public License for more details. * * You should have received a copy of the GNU General Public License * along with this program; if not, write to the Free Software * Foundation, Inc., 51 Franklin Street, Fifth Floor, Boston, MA 02110-1301 USA * */ /* This file was automatically generated --- DO NOT EDIT */ /* Generated on Sun Nov 25 07:35:42 EST 2012 */ #include "codelet-dft.h" #ifdef HAVE_FMA /* Generated by: ../../../genfft/gen_notw.native -fma -reorder-insns -schedule-for-pipeline -compact -variables 4 -pipeline-latency 4 -n 9 -name n1_9 -include n.h */ /* * This function contains 80 FP additions, 56 FP multiplications, * (or, 24 additions, 0 multiplications, 56 fused multiply/add), * 59 stack variables, 10 constants, and 36 memory accesses */ #include "n.h" static void n1_9(const R *ri, const R *ii, R *ro, R *io, stride is, stride os, INT v, INT ivs, INT ovs) { DK(KP954188894, +0.954188894138671133499268364187245676532219158); DK(KP363970234, +0.363970234266202361351047882776834043890471784); DK(KP852868531, +0.852868531952443209628250963940074071936020296); DK(KP984807753, +0.984807753012208059366743024589523013670643252); DK(KP492403876, +0.492403876506104029683371512294761506835321626); DK(KP777861913, +0.777861913430206160028177977318626690410586096); DK(KP839099631, +0.839099631177280011763127298123181364687434283); DK(KP176326980, +0.176326980708464973471090386868618986121633062); DK(KP866025403, +0.866025403784438646763723170752936183471402627); DK(KP500000000, +0.500000000000000000000000000000000000000000000); { INT i; for (i = v; i > 0; i = i - 1, ri = ri + ivs, ii = ii + ivs, ro = ro + ovs, io = io + ovs, MAKE_VOLATILE_STRIDE(36, is), MAKE_VOLATILE_STRIDE(36, os)) { E T17, TV, T14, TY, T11, T15; { E Tm, TM, TL, T5, Tl, T1f, Tb, Tt, Ta, T1c, TI, TX, TF, TW, Tc; E Td, Tp, Tq; { E T1, Th, Ti, Tj, T4, T2, T3; T1 = ri[0]; T2 = ri[WS(is, 3)]; T3 = ri[WS(is, 6)]; Th = ii[0]; Ti = ii[WS(is, 3)]; Tj = ii[WS(is, 6)]; T4 = T2 + T3; Tm = T3 - T2; { E T6, Tz, T7, T8, TA, TB, Tk; T6 = ri[WS(is, 1)]; TM = Ti - Tj; Tk = Ti + Tj; TL = FNMS(KP500000000, T4, T1); T5 = T1 + T4; Tz = ii[WS(is, 1)]; Tl = FNMS(KP500000000, Tk, Th); T1f = Th + Tk; T7 = ri[WS(is, 4)]; T8 = ri[WS(is, 7)]; TA = ii[WS(is, 4)]; TB = ii[WS(is, 7)]; { E TE, T9, TH, TC, TG, TD; Tb = ri[WS(is, 2)]; TE = T7 - T8; T9 = T7 + T8; TH = TB - TA; TC = TA + TB; Tt = ii[WS(is, 2)]; Ta = T6 + T9; TG = FNMS(KP500000000, T9, T6); T1c = Tz + TC; TD = FNMS(KP500000000, TC, Tz); TI = FNMS(KP866025403, TH, TG); TX = FMA(KP866025403, TH, TG); TF = FNMS(KP866025403, TE, TD); TW = FMA(KP866025403, TE, TD); Tc = ri[WS(is, 5)]; Td = ri[WS(is, 8)]; Tp = ii[WS(is, 5)]; Tq = ii[WS(is, 8)]; } } } { E Tn, TN, TZ, T10, TO, Ty, TJ, TP; { E Tw, Te, Tu, Tr; T17 = FNMS(KP866025403, Tm, Tl); Tn = FMA(KP866025403, Tm, Tl); Tw = Td - Tc; Te = Tc + Td; Tu = Tp + Tq; Tr = Tp - Tq; TN = FMA(KP866025403, TM, TL); TV = FNMS(KP866025403, TM, TL); { E Tf, To, T1d, Tv; Tf = Tb + Te; To = FNMS(KP500000000, Te, Tb); T1d = Tt + Tu; Tv = FNMS(KP500000000, Tu, Tt); { E Ts, Tg, T1i, Tx; Ts = FMA(KP866025403, Tr, To); TZ = FNMS(KP866025403, Tr, To); Tg = Ta + Tf; T1i = Tf - Ta; Tx = FMA(KP866025403, Tw, Tv); T10 = FNMS(KP866025403, Tw, Tv); { E T1e, T1g, T1b, T1h; T1e = T1c - T1d; T1g = T1c + T1d; ro[0] = T5 + Tg; T1b = FNMS(KP500000000, Tg, T5); io[0] = T1f + T1g; T1h = FNMS(KP500000000, T1g, T1f); TO = FMA(KP176326980, Ts, Tx); Ty = FNMS(KP176326980, Tx, Ts); ro[WS(os, 6)] = FNMS(KP866025403, T1e, T1b); ro[WS(os, 3)] = FMA(KP866025403, T1e, T1b); io[WS(os, 6)] = FNMS(KP866025403, T1i, T1h); io[WS(os, 3)] = FMA(KP866025403, T1i, T1h); TJ = FNMS(KP839099631, TI, TF); TP = FMA(KP839099631, TF, TI); } } } } { E TS, TK, TU, TQ, TT, TR; TS = FMA(KP777861913, TJ, Ty); TK = FNMS(KP777861913, TJ, Ty); TU = FNMS(KP777861913, TP, TO); TQ = FMA(KP777861913, TP, TO); TT = FMA(KP492403876, TK, Tn); io[WS(os, 1)] = FNMS(KP984807753, TK, Tn); TR = FNMS(KP492403876, TQ, TN); ro[WS(os, 1)] = FMA(KP984807753, TQ, TN); io[WS(os, 4)] = FMA(KP852868531, TU, TT); io[WS(os, 7)] = FNMS(KP852868531, TU, TT); ro[WS(os, 7)] = FNMS(KP852868531, TS, TR); ro[WS(os, 4)] = FMA(KP852868531, TS, TR); T14 = FNMS(KP176326980, TW, TX); TY = FMA(KP176326980, TX, TW); T11 = FNMS(KP363970234, T10, TZ); T15 = FMA(KP363970234, TZ, T10); } } } { E T12, T1a, T16, T18, T13, T19; T12 = FNMS(KP954188894, T11, TY); T1a = FMA(KP954188894, T11, TY); T16 = FNMS(KP954188894, T15, T14); T18 = FMA(KP954188894, T15, T14); T13 = FNMS(KP492403876, T12, TV); ro[WS(os, 2)] = FMA(KP984807753, T12, TV); T19 = FMA(KP492403876, T18, T17); io[WS(os, 2)] = FNMS(KP984807753, T18, T17); ro[WS(os, 8)] = FMA(KP852868531, T16, T13); ro[WS(os, 5)] = FNMS(KP852868531, T16, T13); io[WS(os, 8)] = FMA(KP852868531, T1a, T19); io[WS(os, 5)] = FNMS(KP852868531, T1a, T19); } } } } static const kdft_desc desc = { 9, "n1_9", {24, 0, 56, 0}, &GENUS, 0, 0, 0, 0 }; void X(codelet_n1_9) (planner *p) { X(kdft_register) (p, n1_9, &desc); } #else /* HAVE_FMA */ /* Generated by: ../../../genfft/gen_notw.native -compact -variables 4 -pipeline-latency 4 -n 9 -name n1_9 -include n.h */ /* * This function contains 80 FP additions, 40 FP multiplications, * (or, 60 additions, 20 multiplications, 20 fused multiply/add), * 39 stack variables, 8 constants, and 36 memory accesses */ #include "n.h" static void n1_9(const R *ri, const R *ii, R *ro, R *io, stride is, stride os, INT v, INT ivs, INT ovs) { DK(KP939692620, +0.939692620785908384054109277324731469936208134); DK(KP342020143, +0.342020143325668733044099614682259580763083368); DK(KP984807753, +0.984807753012208059366743024589523013670643252); DK(KP173648177, +0.173648177666930348851716626769314796000375677); DK(KP642787609, +0.642787609686539326322643409907263432907559884); DK(KP766044443, +0.766044443118978035202392650555416673935832457); DK(KP500000000, +0.500000000000000000000000000000000000000000000); DK(KP866025403, +0.866025403784438646763723170752936183471402627); { INT i; for (i = v; i > 0; i = i - 1, ri = ri + ivs, ii = ii + ivs, ro = ro + ovs, io = io + ovs, MAKE_VOLATILE_STRIDE(36, is), MAKE_VOLATILE_STRIDE(36, os)) { E T5, TO, Th, Tk, T1g, TR, Ta, T1c, Tq, TW, Tv, TX, Tf, T1d, TB; E T10, TG, TZ; { E T1, T2, T3, T4; T1 = ri[0]; T2 = ri[WS(is, 3)]; T3 = ri[WS(is, 6)]; T4 = T2 + T3; T5 = T1 + T4; TO = KP866025403 * (T3 - T2); Th = FNMS(KP500000000, T4, T1); } { E TP, Ti, Tj, TQ; TP = ii[0]; Ti = ii[WS(is, 3)]; Tj = ii[WS(is, 6)]; TQ = Ti + Tj; Tk = KP866025403 * (Ti - Tj); T1g = TP + TQ; TR = FNMS(KP500000000, TQ, TP); } { E T6, Ts, T9, Tr, Tp, Tt, Tm, Tu; T6 = ri[WS(is, 1)]; Ts = ii[WS(is, 1)]; { E T7, T8, Tn, To; T7 = ri[WS(is, 4)]; T8 = ri[WS(is, 7)]; T9 = T7 + T8; Tr = KP866025403 * (T8 - T7); Tn = ii[WS(is, 4)]; To = ii[WS(is, 7)]; Tp = KP866025403 * (Tn - To); Tt = Tn + To; } Ta = T6 + T9; T1c = Ts + Tt; Tm = FNMS(KP500000000, T9, T6); Tq = Tm + Tp; TW = Tm - Tp; Tu = FNMS(KP500000000, Tt, Ts); Tv = Tr + Tu; TX = Tu - Tr; } { E Tb, TD, Te, TC, TA, TE, Tx, TF; Tb = ri[WS(is, 2)]; TD = ii[WS(is, 2)]; { E Tc, Td, Ty, Tz; Tc = ri[WS(is, 5)]; Td = ri[WS(is, 8)]; Te = Tc + Td; TC = KP866025403 * (Td - Tc); Ty = ii[WS(is, 5)]; Tz = ii[WS(is, 8)]; TA = KP866025403 * (Ty - Tz); TE = Ty + Tz; } Tf = Tb + Te; T1d = TD + TE; Tx = FNMS(KP500000000, Te, Tb); TB = Tx + TA; T10 = Tx - TA; TF = FNMS(KP500000000, TE, TD); TG = TC + TF; TZ = TF - TC; } { E T1e, Tg, T1b, T1f, T1h, T1i; T1e = KP866025403 * (T1c - T1d); Tg = Ta + Tf; T1b = FNMS(KP500000000, Tg, T5); ro[0] = T5 + Tg; ro[WS(os, 3)] = T1b + T1e; ro[WS(os, 6)] = T1b - T1e; T1f = KP866025403 * (Tf - Ta); T1h = T1c + T1d; T1i = FNMS(KP500000000, T1h, T1g); io[WS(os, 3)] = T1f + T1i; io[0] = T1g + T1h; io[WS(os, 6)] = T1i - T1f; } { E Tl, TS, TI, TN, TM, TT, TJ, TU; Tl = Th + Tk; TS = TO + TR; { E Tw, TH, TK, TL; Tw = FMA(KP766044443, Tq, KP642787609 * Tv); TH = FMA(KP173648177, TB, KP984807753 * TG); TI = Tw + TH; TN = KP866025403 * (TH - Tw); TK = FNMS(KP642787609, Tq, KP766044443 * Tv); TL = FNMS(KP984807753, TB, KP173648177 * TG); TM = KP866025403 * (TK - TL); TT = TK + TL; } ro[WS(os, 1)] = Tl + TI; io[WS(os, 1)] = TS + TT; TJ = FNMS(KP500000000, TI, Tl); ro[WS(os, 7)] = TJ - TM; ro[WS(os, 4)] = TJ + TM; TU = FNMS(KP500000000, TT, TS); io[WS(os, 4)] = TN + TU; io[WS(os, 7)] = TU - TN; } { E TV, T14, T12, T13, T17, T1a, T18, T19; TV = Th - Tk; T14 = TR - TO; { E TY, T11, T15, T16; TY = FMA(KP173648177, TW, KP984807753 * TX); T11 = FNMS(KP939692620, T10, KP342020143 * TZ); T12 = TY + T11; T13 = KP866025403 * (T11 - TY); T15 = FNMS(KP984807753, TW, KP173648177 * TX); T16 = FMA(KP342020143, T10, KP939692620 * TZ); T17 = T15 - T16; T1a = KP866025403 * (T15 + T16); } ro[WS(os, 2)] = TV + T12; io[WS(os, 2)] = T14 + T17; T18 = FNMS(KP500000000, T17, T14); io[WS(os, 5)] = T13 + T18; io[WS(os, 8)] = T18 - T13; T19 = FNMS(KP500000000, T12, TV); ro[WS(os, 8)] = T19 - T1a; ro[WS(os, 5)] = T19 + T1a; } } } } static const kdft_desc desc = { 9, "n1_9", {60, 20, 20, 0}, &GENUS, 0, 0, 0, 0 }; void X(codelet_n1_9) (planner *p) { X(kdft_register) (p, n1_9, &desc); } #endif /* HAVE_FMA */ fftw-3.3.3/dft/scalar/codelets/n1_10.c0000644000175000017500000002465212054410237014224 00000000000000/* * Copyright (c) 2003, 2007-11 Matteo Frigo * Copyright (c) 2003, 2007-11 Massachusetts Institute of Technology * * This program is free software; you can redistribute it and/or modify * it under the terms of the GNU General Public License as published by * the Free Software Foundation; either version 2 of the License, or * (at your option) any later version. * * This program is distributed in the hope that it will be useful, * but WITHOUT ANY WARRANTY; without even the implied warranty of * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the * GNU General Public License for more details. * * You should have received a copy of the GNU General Public License * along with this program; if not, write to the Free Software * Foundation, Inc., 51 Franklin Street, Fifth Floor, Boston, MA 02110-1301 USA * */ /* This file was automatically generated --- DO NOT EDIT */ /* Generated on Sun Nov 25 07:35:43 EST 2012 */ #include "codelet-dft.h" #ifdef HAVE_FMA /* Generated by: ../../../genfft/gen_notw.native -fma -reorder-insns -schedule-for-pipeline -compact -variables 4 -pipeline-latency 4 -n 10 -name n1_10 -include n.h */ /* * This function contains 84 FP additions, 36 FP multiplications, * (or, 48 additions, 0 multiplications, 36 fused multiply/add), * 59 stack variables, 4 constants, and 40 memory accesses */ #include "n.h" static void n1_10(const R *ri, const R *ii, R *ro, R *io, stride is, stride os, INT v, INT ivs, INT ovs) { DK(KP951056516, +0.951056516295153572116439333379382143405698634); DK(KP559016994, +0.559016994374947424102293417182819058860154590); DK(KP250000000, +0.250000000000000000000000000000000000000000000); DK(KP618033988, +0.618033988749894848204586834365638117720309180); { INT i; for (i = v; i > 0; i = i - 1, ri = ri + ivs, ii = ii + ivs, ro = ro + ovs, io = io + ovs, MAKE_VOLATILE_STRIDE(40, is), MAKE_VOLATILE_STRIDE(40, os)) { E T1g, T1a, T18, T1m, T1k, T1f, T19, T11, T1h, T1l; { E Tj, T3, T1b, TN, T1j, TU, T1i, TV, Tq, T10, Ti, Ts, Tw, T15, Tx; E T13, TG, Ty, TB, TC; { E T1, T2, TL, TM; T1 = ri[0]; T2 = ri[WS(is, 5)]; TL = ii[0]; TM = ii[WS(is, 5)]; { E T7, Tk, T6, To, Tg, T8, Tb, Tc; { E T4, T5, Te, Tf; T4 = ri[WS(is, 2)]; Tj = T1 + T2; T3 = T1 - T2; T1b = TL + TM; TN = TL - TM; T5 = ri[WS(is, 7)]; Te = ri[WS(is, 6)]; Tf = ri[WS(is, 1)]; T7 = ri[WS(is, 8)]; Tk = T4 + T5; T6 = T4 - T5; To = Te + Tf; Tg = Te - Tf; T8 = ri[WS(is, 3)]; Tb = ri[WS(is, 4)]; Tc = ri[WS(is, 9)]; } { E TE, TF, Tu, Tv; { E Ta, Th, Tl, T9; Tu = ii[WS(is, 2)]; Tl = T7 + T8; T9 = T7 - T8; { E Tn, Td, Tm, Tp; Tn = Tb + Tc; Td = Tb - Tc; Tm = Tk + Tl; T1j = Tk - Tl; Ta = T6 + T9; TU = T6 - T9; Tp = Tn + To; T1i = Tn - To; Th = Td + Tg; TV = Td - Tg; Tq = Tm + Tp; T10 = Tm - Tp; Tv = ii[WS(is, 7)]; } Ti = Ta + Th; Ts = Ta - Th; } TE = ii[WS(is, 6)]; TF = ii[WS(is, 1)]; Tw = Tu - Tv; T15 = Tu + Tv; Tx = ii[WS(is, 8)]; T13 = TE + TF; TG = TE - TF; Ty = ii[WS(is, 3)]; TB = ii[WS(is, 4)]; TC = ii[WS(is, 9)]; } } } { E T17, TA, T14, TH, T1e, TQ, TS; { E TO, TP, T16, Tz; ro[WS(os, 5)] = T3 + Ti; T16 = Tx + Ty; Tz = Tx - Ty; { E T12, TD, T1c, T1d; T12 = TB + TC; TD = TB - TC; T1c = T15 + T16; T17 = T15 - T16; TO = Tw + Tz; TA = Tw - Tz; T1d = T12 + T13; T14 = T12 - T13; TP = TD + TG; TH = TD - TG; T1e = T1c + T1d; T1g = T1c - T1d; } ro[0] = Tj + Tq; TQ = TO + TP; TS = TO - TP; } { E TK, TI, TY, TW, TR, TJ, Tt, Tr, TZ, TX, TT; TK = FNMS(KP618033988, TA, TH); TI = FMA(KP618033988, TH, TA); io[0] = T1b + T1e; io[WS(os, 5)] = TN + TQ; Tr = FNMS(KP250000000, Ti, T3); TY = FNMS(KP618033988, TU, TV); TW = FMA(KP618033988, TV, TU); TR = FNMS(KP250000000, TQ, TN); TJ = FNMS(KP559016994, Ts, Tr); Tt = FMA(KP559016994, Ts, Tr); T1a = FMA(KP618033988, T14, T17); T18 = FNMS(KP618033988, T17, T14); ro[WS(os, 7)] = FNMS(KP951056516, TK, TJ); ro[WS(os, 3)] = FMA(KP951056516, TK, TJ); ro[WS(os, 1)] = FMA(KP951056516, TI, Tt); ro[WS(os, 9)] = FNMS(KP951056516, TI, Tt); TX = FNMS(KP559016994, TS, TR); TT = FMA(KP559016994, TS, TR); TZ = FNMS(KP250000000, Tq, Tj); io[WS(os, 3)] = FNMS(KP951056516, TY, TX); io[WS(os, 7)] = FMA(KP951056516, TY, TX); io[WS(os, 9)] = FMA(KP951056516, TW, TT); io[WS(os, 1)] = FNMS(KP951056516, TW, TT); T1m = FMA(KP618033988, T1i, T1j); T1k = FNMS(KP618033988, T1j, T1i); T1f = FNMS(KP250000000, T1e, T1b); T19 = FMA(KP559016994, T10, TZ); T11 = FNMS(KP559016994, T10, TZ); } } } ro[WS(os, 4)] = FNMS(KP951056516, T1a, T19); ro[WS(os, 6)] = FMA(KP951056516, T1a, T19); ro[WS(os, 8)] = FMA(KP951056516, T18, T11); ro[WS(os, 2)] = FNMS(KP951056516, T18, T11); T1h = FNMS(KP559016994, T1g, T1f); T1l = FMA(KP559016994, T1g, T1f); io[WS(os, 4)] = FMA(KP951056516, T1m, T1l); io[WS(os, 6)] = FNMS(KP951056516, T1m, T1l); io[WS(os, 8)] = FNMS(KP951056516, T1k, T1h); io[WS(os, 2)] = FMA(KP951056516, T1k, T1h); } } } static const kdft_desc desc = { 10, "n1_10", {48, 0, 36, 0}, &GENUS, 0, 0, 0, 0 }; void X(codelet_n1_10) (planner *p) { X(kdft_register) (p, n1_10, &desc); } #else /* HAVE_FMA */ /* Generated by: ../../../genfft/gen_notw.native -compact -variables 4 -pipeline-latency 4 -n 10 -name n1_10 -include n.h */ /* * This function contains 84 FP additions, 24 FP multiplications, * (or, 72 additions, 12 multiplications, 12 fused multiply/add), * 41 stack variables, 4 constants, and 40 memory accesses */ #include "n.h" static void n1_10(const R *ri, const R *ii, R *ro, R *io, stride is, stride os, INT v, INT ivs, INT ovs) { DK(KP250000000, +0.250000000000000000000000000000000000000000000); DK(KP559016994, +0.559016994374947424102293417182819058860154590); DK(KP587785252, +0.587785252292473129168705954639072768597652438); DK(KP951056516, +0.951056516295153572116439333379382143405698634); { INT i; for (i = v; i > 0; i = i - 1, ri = ri + ivs, ii = ii + ivs, ro = ro + ovs, io = io + ovs, MAKE_VOLATILE_STRIDE(40, is), MAKE_VOLATILE_STRIDE(40, os)) { E T3, Tj, TQ, T1e, TU, TV, T1c, T1b, Tm, Tp, Tq, Ta, Th, Ti, TA; E TH, T17, T14, T1f, T1g, T1h, TL, TM, TR; { E T1, T2, TO, TP; T1 = ri[0]; T2 = ri[WS(is, 5)]; T3 = T1 - T2; Tj = T1 + T2; TO = ii[0]; TP = ii[WS(is, 5)]; TQ = TO - TP; T1e = TO + TP; } { E T6, Tk, Tg, To, T9, Tl, Td, Tn; { E T4, T5, Te, Tf; T4 = ri[WS(is, 2)]; T5 = ri[WS(is, 7)]; T6 = T4 - T5; Tk = T4 + T5; Te = ri[WS(is, 6)]; Tf = ri[WS(is, 1)]; Tg = Te - Tf; To = Te + Tf; } { E T7, T8, Tb, Tc; T7 = ri[WS(is, 8)]; T8 = ri[WS(is, 3)]; T9 = T7 - T8; Tl = T7 + T8; Tb = ri[WS(is, 4)]; Tc = ri[WS(is, 9)]; Td = Tb - Tc; Tn = Tb + Tc; } TU = T6 - T9; TV = Td - Tg; T1c = Tk - Tl; T1b = Tn - To; Tm = Tk + Tl; Tp = Tn + To; Tq = Tm + Tp; Ta = T6 + T9; Th = Td + Tg; Ti = Ta + Th; } { E Tw, T15, TG, T13, Tz, T16, TD, T12; { E Tu, Tv, TE, TF; Tu = ii[WS(is, 2)]; Tv = ii[WS(is, 7)]; Tw = Tu - Tv; T15 = Tu + Tv; TE = ii[WS(is, 6)]; TF = ii[WS(is, 1)]; TG = TE - TF; T13 = TE + TF; } { E Tx, Ty, TB, TC; Tx = ii[WS(is, 8)]; Ty = ii[WS(is, 3)]; Tz = Tx - Ty; T16 = Tx + Ty; TB = ii[WS(is, 4)]; TC = ii[WS(is, 9)]; TD = TB - TC; T12 = TB + TC; } TA = Tw - Tz; TH = TD - TG; T17 = T15 - T16; T14 = T12 - T13; T1f = T15 + T16; T1g = T12 + T13; T1h = T1f + T1g; TL = Tw + Tz; TM = TD + TG; TR = TL + TM; } ro[WS(os, 5)] = T3 + Ti; io[WS(os, 5)] = TQ + TR; ro[0] = Tj + Tq; io[0] = T1e + T1h; { E TI, TK, Tt, TJ, Tr, Ts; TI = FMA(KP951056516, TA, KP587785252 * TH); TK = FNMS(KP587785252, TA, KP951056516 * TH); Tr = KP559016994 * (Ta - Th); Ts = FNMS(KP250000000, Ti, T3); Tt = Tr + Ts; TJ = Ts - Tr; ro[WS(os, 9)] = Tt - TI; ro[WS(os, 3)] = TJ + TK; ro[WS(os, 1)] = Tt + TI; ro[WS(os, 7)] = TJ - TK; } { E TW, TY, TT, TX, TN, TS; TW = FMA(KP951056516, TU, KP587785252 * TV); TY = FNMS(KP587785252, TU, KP951056516 * TV); TN = KP559016994 * (TL - TM); TS = FNMS(KP250000000, TR, TQ); TT = TN + TS; TX = TS - TN; io[WS(os, 1)] = TT - TW; io[WS(os, 7)] = TY + TX; io[WS(os, 9)] = TW + TT; io[WS(os, 3)] = TX - TY; } { E T18, T1a, T11, T19, TZ, T10; T18 = FNMS(KP587785252, T17, KP951056516 * T14); T1a = FMA(KP951056516, T17, KP587785252 * T14); TZ = FNMS(KP250000000, Tq, Tj); T10 = KP559016994 * (Tm - Tp); T11 = TZ - T10; T19 = T10 + TZ; ro[WS(os, 2)] = T11 - T18; ro[WS(os, 6)] = T19 + T1a; ro[WS(os, 8)] = T11 + T18; ro[WS(os, 4)] = T19 - T1a; } { E T1d, T1l, T1k, T1m, T1i, T1j; T1d = FNMS(KP587785252, T1c, KP951056516 * T1b); T1l = FMA(KP951056516, T1c, KP587785252 * T1b); T1i = FNMS(KP250000000, T1h, T1e); T1j = KP559016994 * (T1f - T1g); T1k = T1i - T1j; T1m = T1j + T1i; io[WS(os, 2)] = T1d + T1k; io[WS(os, 6)] = T1m - T1l; io[WS(os, 8)] = T1k - T1d; io[WS(os, 4)] = T1l + T1m; } } } } static const kdft_desc desc = { 10, "n1_10", {72, 12, 12, 0}, &GENUS, 0, 0, 0, 0 }; void X(codelet_n1_10) (planner *p) { X(kdft_register) (p, n1_10, &desc); } #endif /* HAVE_FMA */ fftw-3.3.3/dft/scalar/codelets/n1_11.c0000644000175000017500000003613212054410242014215 00000000000000/* * Copyright (c) 2003, 2007-11 Matteo Frigo * Copyright (c) 2003, 2007-11 Massachusetts Institute of Technology * * This program is free software; you can redistribute it and/or modify * it under the terms of the GNU General Public License as published by * the Free Software Foundation; either version 2 of the License, or * (at your option) any later version. * * This program is distributed in the hope that it will be useful, * but WITHOUT ANY WARRANTY; without even the implied warranty of * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the * GNU General Public License for more details. * * You should have received a copy of the GNU General Public License * along with this program; if not, write to the Free Software * Foundation, Inc., 51 Franklin Street, Fifth Floor, Boston, MA 02110-1301 USA * */ /* This file was automatically generated --- DO NOT EDIT */ /* Generated on Sun Nov 25 07:35:43 EST 2012 */ #include "codelet-dft.h" #ifdef HAVE_FMA /* Generated by: ../../../genfft/gen_notw.native -fma -reorder-insns -schedule-for-pipeline -compact -variables 4 -pipeline-latency 4 -n 11 -name n1_11 -include n.h */ /* * This function contains 140 FP additions, 110 FP multiplications, * (or, 30 additions, 0 multiplications, 110 fused multiply/add), * 84 stack variables, 10 constants, and 44 memory accesses */ #include "n.h" static void n1_11(const R *ri, const R *ii, R *ro, R *io, stride is, stride os, INT v, INT ivs, INT ovs) { DK(KP989821441, +0.989821441880932732376092037776718787376519372); DK(KP959492973, +0.959492973614497389890368057066327699062454848); DK(KP918985947, +0.918985947228994779780736114132655398124909697); DK(KP876768831, +0.876768831002589333891339807079336796764054852); DK(KP830830026, +0.830830026003772851058548298459246407048009821); DK(KP778434453, +0.778434453334651800608337670740821884709317477); DK(KP715370323, +0.715370323453429719112414662767260662417897278); DK(KP634356270, +0.634356270682424498893150776899916060542806975); DK(KP342584725, +0.342584725681637509502641509861112333758894680); DK(KP521108558, +0.521108558113202722944698153526659300680427422); { INT i; for (i = v; i > 0; i = i - 1, ri = ri + ivs, ii = ii + ivs, ro = ro + ovs, io = io + ovs, MAKE_VOLATILE_STRIDE(44, is), MAKE_VOLATILE_STRIDE(44, os)) { E T1, TA, T1p, T1y, T19, T1d, T1a, T1e; { E T1f, T1u, T4, T1q, Tg, T1t, T7, T1s, Ta, Td, T1r, TP, T1X, T26, Ti; E TG, T1O, T1w, TY, T1F, T17, To, T1i, T1k, T1h, Tr, T1j, Tu, T1g, Tx; E T21, TU, TL, TC, T1S, T1J, T1m, T12, T1z, T1b; T1 = ri[0]; T1f = ii[0]; { E T1E, T16, Tb, Tc, Tv, Tw; { E T2, T3, Te, Tf; T2 = ri[WS(is, 1)]; T3 = ri[WS(is, 10)]; Te = ri[WS(is, 5)]; Tf = ri[WS(is, 6)]; { E T5, T6, T8, T9; T5 = ri[WS(is, 2)]; T1u = T3 - T2; T4 = T2 + T3; T1q = Tf - Te; Tg = Te + Tf; T6 = ri[WS(is, 9)]; T8 = ri[WS(is, 3)]; T9 = ri[WS(is, 8)]; Tb = ri[WS(is, 4)]; T1t = T6 - T5; T7 = T5 + T6; T1s = T9 - T8; Ta = T8 + T9; Tc = ri[WS(is, 7)]; } } { E T25, Th, T1W, TO; T25 = FMA(KP521108558, T1q, T1u); T1W = FMA(KP521108558, T1s, T1q); TO = FNMS(KP342584725, T4, Ta); Th = FNMS(KP342584725, Ta, T7); Td = Tb + Tc; T1r = Tc - Tb; TP = FNMS(KP634356270, TO, Tg); T1X = FNMS(KP715370323, T1W, T1t); T26 = FMA(KP715370323, T25, T1r); { E TF, T1N, T1v, TX; TF = FNMS(KP342584725, Td, T4); Ti = FNMS(KP634356270, Th, Td); T1N = FNMS(KP521108558, T1t, T1r); T1v = FNMS(KP521108558, T1u, T1t); TG = FNMS(KP634356270, TF, T7); TX = FNMS(KP342584725, T7, Tg); T1O = FMA(KP715370323, T1N, T1q); T1w = FNMS(KP715370323, T1v, T1s); T1E = FMA(KP521108558, T1r, T1s); TY = FNMS(KP634356270, TX, T4); T16 = FNMS(KP342584725, Tg, Td); } } { E Ty, Tz, Tm, Tn; Tm = ii[WS(is, 3)]; T1F = FMA(KP715370323, T1E, T1u); Tn = ii[WS(is, 8)]; T17 = FNMS(KP634356270, T16, Ta); Ty = ii[WS(is, 5)]; Tz = ii[WS(is, 6)]; To = Tm - Tn; T1i = Tm + Tn; { E Tp, Tq, Ts, Tt; Tp = ii[WS(is, 2)]; T1k = Ty + Tz; TA = Ty - Tz; Tq = ii[WS(is, 9)]; Ts = ii[WS(is, 4)]; Tt = ii[WS(is, 7)]; Tv = ii[WS(is, 1)]; T1h = Tp + Tq; Tr = Tp - Tq; T1j = Ts + Tt; Tu = Ts - Tt; Tw = ii[WS(is, 10)]; } } { E TB, T1R, T20, TK, TT, T1I, T1l; T20 = FNMS(KP342584725, T1i, T1h); TK = FMA(KP521108558, To, TA); TT = FNMS(KP521108558, Tr, Tu); T1g = Tv + Tw; Tx = Tv - Tw; T21 = FNMS(KP634356270, T20, T1j); TU = FMA(KP715370323, TT, TA); TL = FNMS(KP715370323, TK, Tr); TB = FMA(KP521108558, TA, Tx); T1R = FNMS(KP342584725, T1j, T1g); T1I = FNMS(KP342584725, T1g, T1i); T1l = FNMS(KP342584725, T1k, T1j); TC = FMA(KP715370323, TB, Tu); T1S = FNMS(KP634356270, T1R, T1h); T1J = FNMS(KP634356270, T1I, T1k); T1m = FNMS(KP634356270, T1l, T1i); T12 = FMA(KP521108558, Tu, To); T1z = FNMS(KP342584725, T1h, T1k); T1b = FNMS(KP521108558, Tx, Tr); } } { E T13, T1A, T1c, T1Z, T1V, TH, TM, Tj, TD; ro[0] = T1 + T4 + T7 + Ta + Td + Tg; T13 = FMA(KP715370323, T12, Tx); T1A = FNMS(KP634356270, T1z, T1g); T1c = FNMS(KP715370323, T1b, To); io[0] = T1f + T1g + T1h + T1i + T1j + T1k; Tj = FNMS(KP778434453, Ti, T4); TD = FMA(KP830830026, TC, Tr); { E TE, T23, T28, Tl, Tk, T22, T27; T22 = FNMS(KP778434453, T21, T1g); T27 = FMA(KP830830026, T26, T1t); Tk = FNMS(KP876768831, Tj, Tg); TE = FMA(KP918985947, TD, To); T23 = FNMS(KP876768831, T22, T1k); T28 = FMA(KP918985947, T27, T1s); Tl = FNMS(KP959492973, Tk, T1); { E T1U, T1T, T24, T1Y; T1T = FNMS(KP778434453, T1S, T1k); T24 = FNMS(KP959492973, T23, T1f); T1Y = FMA(KP830830026, T1X, T1u); ro[WS(os, 1)] = FMA(KP989821441, TE, Tl); ro[WS(os, 10)] = FNMS(KP989821441, TE, Tl); T1U = FNMS(KP876768831, T1T, T1i); io[WS(os, 10)] = FNMS(KP989821441, T28, T24); io[WS(os, 1)] = FMA(KP989821441, T28, T24); T1Z = FNMS(KP918985947, T1Y, T1r); T1V = FNMS(KP959492973, T1U, T1f); } TH = FNMS(KP778434453, TG, Tg); TM = FMA(KP830830026, TL, Tx); } { E T1M, TZ, T14, T1Q; { E TN, TR, TV, TJ, TI, TQ, T1P; TQ = FNMS(KP778434453, TP, Td); io[WS(os, 9)] = FMA(KP989821441, T1Z, T1V); io[WS(os, 2)] = FNMS(KP989821441, T1Z, T1V); TI = FNMS(KP876768831, TH, Ta); TN = FNMS(KP918985947, TM, Tu); TR = FNMS(KP876768831, TQ, T7); TV = FNMS(KP830830026, TU, To); TJ = FNMS(KP959492973, TI, T1); { E T1L, TS, TW, T1K; T1K = FNMS(KP778434453, T1J, T1j); TS = FNMS(KP959492973, TR, T1); TW = FNMS(KP918985947, TV, Tx); ro[WS(os, 9)] = FMA(KP989821441, TN, TJ); ro[WS(os, 2)] = FNMS(KP989821441, TN, TJ); T1L = FNMS(KP876768831, T1K, T1h); ro[WS(os, 3)] = FMA(KP989821441, TW, TS); ro[WS(os, 8)] = FNMS(KP989821441, TW, TS); T1P = FNMS(KP830830026, T1O, T1s); T1M = FNMS(KP959492973, T1L, T1f); } TZ = FNMS(KP778434453, TY, Ta); T14 = FNMS(KP830830026, T13, TA); T1Q = FNMS(KP918985947, T1P, T1u); } { E T15, T11, T1C, T1G, T1B, T10; T1B = FNMS(KP778434453, T1A, T1i); T10 = FNMS(KP876768831, TZ, Td); T15 = FMA(KP918985947, T14, Tr); io[WS(os, 8)] = FNMS(KP989821441, T1Q, T1M); io[WS(os, 3)] = FMA(KP989821441, T1Q, T1M); T11 = FNMS(KP959492973, T10, T1); T1C = FNMS(KP876768831, T1B, T1j); T1G = FNMS(KP830830026, T1F, T1q); { E T1D, T1H, T1o, T1x, T1n, T18; T1n = FNMS(KP778434453, T1m, T1h); ro[WS(os, 7)] = FMA(KP989821441, T15, T11); ro[WS(os, 4)] = FNMS(KP989821441, T15, T11); T1D = FNMS(KP959492973, T1C, T1f); T1H = FMA(KP918985947, T1G, T1t); T1o = FNMS(KP876768831, T1n, T1g); T1x = FNMS(KP830830026, T1w, T1r); T18 = FNMS(KP778434453, T17, T7); io[WS(os, 7)] = FMA(KP989821441, T1H, T1D); io[WS(os, 4)] = FNMS(KP989821441, T1H, T1D); T1p = FNMS(KP959492973, T1o, T1f); T1y = FNMS(KP918985947, T1x, T1q); T19 = FNMS(KP876768831, T18, T4); T1d = FNMS(KP830830026, T1c, Tu); } } } } } io[WS(os, 6)] = FNMS(KP989821441, T1y, T1p); io[WS(os, 5)] = FMA(KP989821441, T1y, T1p); T1a = FNMS(KP959492973, T19, T1); T1e = FNMS(KP918985947, T1d, TA); ro[WS(os, 5)] = FMA(KP989821441, T1e, T1a); ro[WS(os, 6)] = FNMS(KP989821441, T1e, T1a); } } } static const kdft_desc desc = { 11, "n1_11", {30, 0, 110, 0}, &GENUS, 0, 0, 0, 0 }; void X(codelet_n1_11) (planner *p) { X(kdft_register) (p, n1_11, &desc); } #else /* HAVE_FMA */ /* Generated by: ../../../genfft/gen_notw.native -compact -variables 4 -pipeline-latency 4 -n 11 -name n1_11 -include n.h */ /* * This function contains 140 FP additions, 100 FP multiplications, * (or, 60 additions, 20 multiplications, 80 fused multiply/add), * 41 stack variables, 10 constants, and 44 memory accesses */ #include "n.h" static void n1_11(const R *ri, const R *ii, R *ro, R *io, stride is, stride os, INT v, INT ivs, INT ovs) { DK(KP654860733, +0.654860733945285064056925072466293553183791199); DK(KP142314838, +0.142314838273285140443792668616369668791051361); DK(KP959492973, +0.959492973614497389890368057066327699062454848); DK(KP415415013, +0.415415013001886425529274149229623203524004910); DK(KP841253532, +0.841253532831181168861811648919367717513292498); DK(KP989821441, +0.989821441880932732376092037776718787376519372); DK(KP909631995, +0.909631995354518371411715383079028460060241051); DK(KP281732556, +0.281732556841429697711417915346616899035777899); DK(KP540640817, +0.540640817455597582107635954318691695431770608); DK(KP755749574, +0.755749574354258283774035843972344420179717445); { INT i; for (i = v; i > 0; i = i - 1, ri = ri + ivs, ii = ii + ivs, ro = ro + ovs, io = io + ovs, MAKE_VOLATILE_STRIDE(44, is), MAKE_VOLATILE_STRIDE(44, os)) { E T1, TM, T4, TG, Tk, TR, Tw, TN, T7, TK, Ta, TH, Tn, TQ, Td; E TJ, Tq, TO, Tt, TP, Tg, TI; { E T2, T3, Ti, Tj; T1 = ri[0]; TM = ii[0]; T2 = ri[WS(is, 1)]; T3 = ri[WS(is, 10)]; T4 = T2 + T3; TG = T3 - T2; Ti = ii[WS(is, 1)]; Tj = ii[WS(is, 10)]; Tk = Ti - Tj; TR = Ti + Tj; { E Tu, Tv, T5, T6; Tu = ii[WS(is, 2)]; Tv = ii[WS(is, 9)]; Tw = Tu - Tv; TN = Tu + Tv; T5 = ri[WS(is, 2)]; T6 = ri[WS(is, 9)]; T7 = T5 + T6; TK = T6 - T5; } } { E T8, T9, To, Tp; T8 = ri[WS(is, 3)]; T9 = ri[WS(is, 8)]; Ta = T8 + T9; TH = T9 - T8; { E Tl, Tm, Tb, Tc; Tl = ii[WS(is, 3)]; Tm = ii[WS(is, 8)]; Tn = Tl - Tm; TQ = Tl + Tm; Tb = ri[WS(is, 4)]; Tc = ri[WS(is, 7)]; Td = Tb + Tc; TJ = Tc - Tb; } To = ii[WS(is, 4)]; Tp = ii[WS(is, 7)]; Tq = To - Tp; TO = To + Tp; { E Tr, Ts, Te, Tf; Tr = ii[WS(is, 5)]; Ts = ii[WS(is, 6)]; Tt = Tr - Ts; TP = Tr + Ts; Te = ri[WS(is, 5)]; Tf = ri[WS(is, 6)]; Tg = Te + Tf; TI = Tf - Te; } } { E Tx, Th, TZ, T10; ro[0] = T1 + T4 + T7 + Ta + Td + Tg; io[0] = TM + TR + TN + TQ + TO + TP; Tx = FMA(KP755749574, Tk, KP540640817 * Tn) + FNMS(KP909631995, Tt, KP281732556 * Tq) - (KP989821441 * Tw); Th = FMA(KP841253532, Ta, T1) + FNMS(KP959492973, Td, KP415415013 * Tg) + FNMA(KP142314838, T7, KP654860733 * T4); ro[WS(os, 7)] = Th - Tx; ro[WS(os, 4)] = Th + Tx; TZ = FMA(KP755749574, TG, KP540640817 * TH) + FNMS(KP909631995, TI, KP281732556 * TJ) - (KP989821441 * TK); T10 = FMA(KP841253532, TQ, TM) + FNMS(KP959492973, TO, KP415415013 * TP) + FNMA(KP142314838, TN, KP654860733 * TR); io[WS(os, 4)] = TZ + T10; io[WS(os, 7)] = T10 - TZ; { E TX, TY, Tz, Ty; TX = FMA(KP909631995, TG, KP755749574 * TK) + FNMA(KP540640817, TI, KP989821441 * TJ) - (KP281732556 * TH); TY = FMA(KP415415013, TR, TM) + FNMS(KP142314838, TO, KP841253532 * TP) + FNMA(KP959492973, TQ, KP654860733 * TN); io[WS(os, 2)] = TX + TY; io[WS(os, 9)] = TY - TX; Tz = FMA(KP909631995, Tk, KP755749574 * Tw) + FNMA(KP540640817, Tt, KP989821441 * Tq) - (KP281732556 * Tn); Ty = FMA(KP415415013, T4, T1) + FNMS(KP142314838, Td, KP841253532 * Tg) + FNMA(KP959492973, Ta, KP654860733 * T7); ro[WS(os, 9)] = Ty - Tz; ro[WS(os, 2)] = Ty + Tz; } } { E TB, TA, TT, TU; TB = FMA(KP540640817, Tk, KP909631995 * Tw) + FMA(KP989821441, Tn, KP755749574 * Tq) + (KP281732556 * Tt); TA = FMA(KP841253532, T4, T1) + FNMS(KP959492973, Tg, KP415415013 * T7) + FNMA(KP654860733, Td, KP142314838 * Ta); ro[WS(os, 10)] = TA - TB; ro[WS(os, 1)] = TA + TB; { E TV, TW, TD, TC; TV = FMA(KP540640817, TG, KP909631995 * TK) + FMA(KP989821441, TH, KP755749574 * TJ) + (KP281732556 * TI); TW = FMA(KP841253532, TR, TM) + FNMS(KP959492973, TP, KP415415013 * TN) + FNMA(KP654860733, TO, KP142314838 * TQ); io[WS(os, 1)] = TV + TW; io[WS(os, 10)] = TW - TV; TD = FMA(KP989821441, Tk, KP540640817 * Tq) + FNMS(KP909631995, Tn, KP755749574 * Tt) - (KP281732556 * Tw); TC = FMA(KP415415013, Ta, T1) + FNMS(KP654860733, Tg, KP841253532 * Td) + FNMA(KP959492973, T7, KP142314838 * T4); ro[WS(os, 8)] = TC - TD; ro[WS(os, 3)] = TC + TD; } TT = FMA(KP989821441, TG, KP540640817 * TJ) + FNMS(KP909631995, TH, KP755749574 * TI) - (KP281732556 * TK); TU = FMA(KP415415013, TQ, TM) + FNMS(KP654860733, TP, KP841253532 * TO) + FNMA(KP959492973, TN, KP142314838 * TR); io[WS(os, 3)] = TT + TU; io[WS(os, 8)] = TU - TT; { E TL, TS, TF, TE; TL = FMA(KP281732556, TG, KP755749574 * TH) + FNMS(KP909631995, TJ, KP989821441 * TI) - (KP540640817 * TK); TS = FMA(KP841253532, TN, TM) + FNMS(KP142314838, TP, KP415415013 * TO) + FNMA(KP654860733, TQ, KP959492973 * TR); io[WS(os, 5)] = TL + TS; io[WS(os, 6)] = TS - TL; TF = FMA(KP281732556, Tk, KP755749574 * Tn) + FNMS(KP909631995, Tq, KP989821441 * Tt) - (KP540640817 * Tw); TE = FMA(KP841253532, T7, T1) + FNMS(KP142314838, Tg, KP415415013 * Td) + FNMA(KP654860733, Ta, KP959492973 * T4); ro[WS(os, 6)] = TE - TF; ro[WS(os, 5)] = TE + TF; } } } } } static const kdft_desc desc = { 11, "n1_11", {60, 20, 80, 0}, &GENUS, 0, 0, 0, 0 }; void X(codelet_n1_11) (planner *p) { X(kdft_register) (p, n1_11, &desc); } #endif /* HAVE_FMA */ fftw-3.3.3/dft/scalar/codelets/n1_12.c0000644000175000017500000002576412054410237014233 00000000000000/* * Copyright (c) 2003, 2007-11 Matteo Frigo * Copyright (c) 2003, 2007-11 Massachusetts Institute of Technology * * This program is free software; you can redistribute it and/or modify * it under the terms of the GNU General Public License as published by * the Free Software Foundation; either version 2 of the License, or * (at your option) any later version. * * This program is distributed in the hope that it will be useful, * but WITHOUT ANY WARRANTY; without even the implied warranty of * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the * GNU General Public License for more details. * * You should have received a copy of the GNU General Public License * along with this program; if not, write to the Free Software * Foundation, Inc., 51 Franklin Street, Fifth Floor, Boston, MA 02110-1301 USA * */ /* This file was automatically generated --- DO NOT EDIT */ /* Generated on Sun Nov 25 07:35:43 EST 2012 */ #include "codelet-dft.h" #ifdef HAVE_FMA /* Generated by: ../../../genfft/gen_notw.native -fma -reorder-insns -schedule-for-pipeline -compact -variables 4 -pipeline-latency 4 -n 12 -name n1_12 -include n.h */ /* * This function contains 96 FP additions, 24 FP multiplications, * (or, 72 additions, 0 multiplications, 24 fused multiply/add), * 63 stack variables, 2 constants, and 48 memory accesses */ #include "n.h" static void n1_12(const R *ri, const R *ii, R *ro, R *io, stride is, stride os, INT v, INT ivs, INT ovs) { DK(KP866025403, +0.866025403784438646763723170752936183471402627); DK(KP500000000, +0.500000000000000000000000000000000000000000000); { INT i; for (i = v; i > 0; i = i - 1, ri = ri + ivs, ii = ii + ivs, ro = ro + ovs, io = io + ovs, MAKE_VOLATILE_STRIDE(48, is), MAKE_VOLATILE_STRIDE(48, os)) { E TT, TW, TF, T1q, TY, TQ, TX, T1n; { E TA, TS, TR, T5, Ts, Tz, TD, TV, TU, Ta, Tx, TC, T1d, Th, TJ; E TG, Tg, T1u, T1c, T1f, TM, TN, Tk, T1i; { E T6, Tt, Tu, Tv, T9; { E T1, To, Tp, Tq, T4, T2, T3, T7, T8, Tr; T1 = ri[0]; T2 = ri[WS(is, 4)]; T3 = ri[WS(is, 8)]; To = ii[0]; Tp = ii[WS(is, 4)]; Tq = ii[WS(is, 8)]; T4 = T2 + T3; TA = T3 - T2; T6 = ri[WS(is, 6)]; TS = Tp - Tq; Tr = Tp + Tq; TR = FNMS(KP500000000, T4, T1); T5 = T1 + T4; T7 = ri[WS(is, 10)]; Ts = To + Tr; Tz = FNMS(KP500000000, Tr, To); T8 = ri[WS(is, 2)]; Tt = ii[WS(is, 6)]; Tu = ii[WS(is, 10)]; Tv = ii[WS(is, 2)]; T9 = T7 + T8; TD = T8 - T7; } { E Tc, T1a, TH, TI, Tf, Td, Te, Tw, Ti, Tj, T1b; Tc = ri[WS(is, 3)]; TV = Tu - Tv; Tw = Tu + Tv; TU = FNMS(KP500000000, T9, T6); Ta = T6 + T9; Td = ri[WS(is, 7)]; Tx = Tt + Tw; TC = FNMS(KP500000000, Tw, Tt); Te = ri[WS(is, 11)]; T1a = ii[WS(is, 3)]; TH = ii[WS(is, 7)]; TI = ii[WS(is, 11)]; Tf = Td + Te; T1d = Te - Td; Th = ri[WS(is, 9)]; TJ = TH - TI; T1b = TH + TI; TG = FNMS(KP500000000, Tf, Tc); Tg = Tc + Tf; Ti = ri[WS(is, 1)]; T1u = T1a + T1b; T1c = FNMS(KP500000000, T1b, T1a); Tj = ri[WS(is, 5)]; T1f = ii[WS(is, 9)]; TM = ii[WS(is, 1)]; TN = ii[WS(is, 5)]; Tk = Ti + Tj; T1i = Tj - Ti; } } { E T1t, TO, TL, T1h, T1w, Tb, T1g, Tl; T1t = T5 - Ta; Tb = T5 + Ta; TO = TM - TN; T1g = TM + TN; TL = FNMS(KP500000000, Tk, Th); Tl = Th + Tk; { E T1x, Ty, T1v, Tn, Tm, T1y; T1x = Ts + Tx; Ty = Ts - Tx; T1v = T1f + T1g; T1h = FNMS(KP500000000, T1g, T1f); Tn = Tg - Tl; Tm = Tg + Tl; T1y = T1u + T1v; T1w = T1u - T1v; ro[0] = Tb + Tm; ro[WS(os, 6)] = Tb - Tm; io[WS(os, 3)] = Tn + Ty; io[0] = T1x + T1y; io[WS(os, 6)] = T1x - T1y; io[WS(os, 9)] = Ty - Tn; } { E TB, TE, T1o, T11, T1p, TK, TP, T15, T1k, T18, T14, T16, T1l, T1m; { E T1e, T1j, TZ, T10, T12, T13; TB = FNMS(KP866025403, TA, Tz); TZ = FMA(KP866025403, TA, Tz); T10 = FMA(KP866025403, TD, TC); TE = FNMS(KP866025403, TD, TC); T1o = FNMS(KP866025403, T1d, T1c); T1e = FMA(KP866025403, T1d, T1c); ro[WS(os, 9)] = T1t + T1w; ro[WS(os, 3)] = T1t - T1w; T1l = TZ + T10; T11 = TZ - T10; T1j = FMA(KP866025403, T1i, T1h); T1p = FNMS(KP866025403, T1i, T1h); TK = FNMS(KP866025403, TJ, TG); T12 = FMA(KP866025403, TJ, TG); T13 = FMA(KP866025403, TO, TL); TP = FNMS(KP866025403, TO, TL); TT = FNMS(KP866025403, TS, TR); T15 = FMA(KP866025403, TS, TR); T1m = T1e + T1j; T1k = T1e - T1j; T18 = T12 + T13; T14 = T12 - T13; T16 = FMA(KP866025403, TV, TU); TW = FNMS(KP866025403, TV, TU); } io[WS(os, 10)] = T1l - T1m; io[WS(os, 4)] = T1l + T1m; io[WS(os, 7)] = T11 + T14; io[WS(os, 1)] = T11 - T14; { E T17, T19, T1r, T1s; T17 = T15 + T16; T19 = T15 - T16; ro[WS(os, 7)] = T19 - T1k; ro[WS(os, 1)] = T19 + T1k; ro[WS(os, 4)] = T17 + T18; ro[WS(os, 10)] = T17 - T18; T1r = TB + TE; TF = TB - TE; T1s = T1o + T1p; T1q = T1o - T1p; TY = TK + TP; TQ = TK - TP; io[WS(os, 2)] = T1r - T1s; io[WS(os, 8)] = T1r + T1s; } } } } io[WS(os, 11)] = TF + TQ; io[WS(os, 5)] = TF - TQ; TX = TT + TW; T1n = TT - TW; ro[WS(os, 11)] = T1n - T1q; ro[WS(os, 5)] = T1n + T1q; ro[WS(os, 8)] = TX + TY; ro[WS(os, 2)] = TX - TY; } } } static const kdft_desc desc = { 12, "n1_12", {72, 0, 24, 0}, &GENUS, 0, 0, 0, 0 }; void X(codelet_n1_12) (planner *p) { X(kdft_register) (p, n1_12, &desc); } #else /* HAVE_FMA */ /* Generated by: ../../../genfft/gen_notw.native -compact -variables 4 -pipeline-latency 4 -n 12 -name n1_12 -include n.h */ /* * This function contains 96 FP additions, 16 FP multiplications, * (or, 88 additions, 8 multiplications, 8 fused multiply/add), * 43 stack variables, 2 constants, and 48 memory accesses */ #include "n.h" static void n1_12(const R *ri, const R *ii, R *ro, R *io, stride is, stride os, INT v, INT ivs, INT ovs) { DK(KP866025403, +0.866025403784438646763723170752936183471402627); DK(KP500000000, +0.500000000000000000000000000000000000000000000); { INT i; for (i = v; i > 0; i = i - 1, ri = ri + ivs, ii = ii + ivs, ro = ro + ovs, io = io + ovs, MAKE_VOLATILE_STRIDE(48, is), MAKE_VOLATILE_STRIDE(48, os)) { E T5, TR, TA, Ts, TS, Tz, Ta, TU, TD, Tx, TV, TC, Tg, T1a, TG; E TJ, T1u, T1d, Tl, T1f, TL, TO, T1v, T1i; { E T1, T2, T3, T4; T1 = ri[0]; T2 = ri[WS(is, 4)]; T3 = ri[WS(is, 8)]; T4 = T2 + T3; T5 = T1 + T4; TR = FNMS(KP500000000, T4, T1); TA = KP866025403 * (T3 - T2); } { E To, Tp, Tq, Tr; To = ii[0]; Tp = ii[WS(is, 4)]; Tq = ii[WS(is, 8)]; Tr = Tp + Tq; Ts = To + Tr; TS = KP866025403 * (Tp - Tq); Tz = FNMS(KP500000000, Tr, To); } { E T6, T7, T8, T9; T6 = ri[WS(is, 6)]; T7 = ri[WS(is, 10)]; T8 = ri[WS(is, 2)]; T9 = T7 + T8; Ta = T6 + T9; TU = FNMS(KP500000000, T9, T6); TD = KP866025403 * (T8 - T7); } { E Tt, Tu, Tv, Tw; Tt = ii[WS(is, 6)]; Tu = ii[WS(is, 10)]; Tv = ii[WS(is, 2)]; Tw = Tu + Tv; Tx = Tt + Tw; TV = KP866025403 * (Tu - Tv); TC = FNMS(KP500000000, Tw, Tt); } { E Tc, Td, Te, Tf; Tc = ri[WS(is, 3)]; Td = ri[WS(is, 7)]; Te = ri[WS(is, 11)]; Tf = Td + Te; Tg = Tc + Tf; T1a = KP866025403 * (Te - Td); TG = FNMS(KP500000000, Tf, Tc); } { E T1b, TH, TI, T1c; T1b = ii[WS(is, 3)]; TH = ii[WS(is, 7)]; TI = ii[WS(is, 11)]; T1c = TH + TI; TJ = KP866025403 * (TH - TI); T1u = T1b + T1c; T1d = FNMS(KP500000000, T1c, T1b); } { E Th, Ti, Tj, Tk; Th = ri[WS(is, 9)]; Ti = ri[WS(is, 1)]; Tj = ri[WS(is, 5)]; Tk = Ti + Tj; Tl = Th + Tk; T1f = KP866025403 * (Tj - Ti); TL = FNMS(KP500000000, Tk, Th); } { E T1g, TM, TN, T1h; T1g = ii[WS(is, 9)]; TM = ii[WS(is, 1)]; TN = ii[WS(is, 5)]; T1h = TM + TN; TO = KP866025403 * (TM - TN); T1v = T1g + T1h; T1i = FNMS(KP500000000, T1h, T1g); } { E Tb, Tm, T1t, T1w; Tb = T5 + Ta; Tm = Tg + Tl; ro[WS(os, 6)] = Tb - Tm; ro[0] = Tb + Tm; { E T1x, T1y, Tn, Ty; T1x = Ts + Tx; T1y = T1u + T1v; io[WS(os, 6)] = T1x - T1y; io[0] = T1x + T1y; Tn = Tg - Tl; Ty = Ts - Tx; io[WS(os, 3)] = Tn + Ty; io[WS(os, 9)] = Ty - Tn; } T1t = T5 - Ta; T1w = T1u - T1v; ro[WS(os, 3)] = T1t - T1w; ro[WS(os, 9)] = T1t + T1w; { E T11, T1l, T1k, T1m, T14, T18, T17, T19; { E TZ, T10, T1e, T1j; TZ = TA + Tz; T10 = TD + TC; T11 = TZ - T10; T1l = TZ + T10; T1e = T1a + T1d; T1j = T1f + T1i; T1k = T1e - T1j; T1m = T1e + T1j; } { E T12, T13, T15, T16; T12 = TG + TJ; T13 = TL + TO; T14 = T12 - T13; T18 = T12 + T13; T15 = TR + TS; T16 = TU + TV; T17 = T15 + T16; T19 = T15 - T16; } io[WS(os, 1)] = T11 - T14; ro[WS(os, 1)] = T19 + T1k; io[WS(os, 7)] = T11 + T14; ro[WS(os, 7)] = T19 - T1k; ro[WS(os, 10)] = T17 - T18; io[WS(os, 10)] = T1l - T1m; ro[WS(os, 4)] = T17 + T18; io[WS(os, 4)] = T1l + T1m; } { E TF, T1r, T1q, T1s, TQ, TY, TX, T1n; { E TB, TE, T1o, T1p; TB = Tz - TA; TE = TC - TD; TF = TB - TE; T1r = TB + TE; T1o = T1d - T1a; T1p = T1i - T1f; T1q = T1o - T1p; T1s = T1o + T1p; } { E TK, TP, TT, TW; TK = TG - TJ; TP = TL - TO; TQ = TK - TP; TY = TK + TP; TT = TR - TS; TW = TU - TV; TX = TT + TW; T1n = TT - TW; } io[WS(os, 5)] = TF - TQ; ro[WS(os, 5)] = T1n + T1q; io[WS(os, 11)] = TF + TQ; ro[WS(os, 11)] = T1n - T1q; ro[WS(os, 2)] = TX - TY; io[WS(os, 2)] = T1r - T1s; ro[WS(os, 8)] = TX + TY; io[WS(os, 8)] = T1r + T1s; } } } } } static const kdft_desc desc = { 12, "n1_12", {88, 8, 8, 0}, &GENUS, 0, 0, 0, 0 }; void X(codelet_n1_12) (planner *p) { X(kdft_register) (p, n1_12, &desc); } #endif /* HAVE_FMA */ fftw-3.3.3/dft/scalar/codelets/n1_13.c0000644000175000017500000005211012054410242014211 00000000000000/* * Copyright (c) 2003, 2007-11 Matteo Frigo * Copyright (c) 2003, 2007-11 Massachusetts Institute of Technology * * This program is free software; you can redistribute it and/or modify * it under the terms of the GNU General Public License as published by * the Free Software Foundation; either version 2 of the License, or * (at your option) any later version. * * This program is distributed in the hope that it will be useful, * but WITHOUT ANY WARRANTY; without even the implied warranty of * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the * GNU General Public License for more details. * * You should have received a copy of the GNU General Public License * along with this program; if not, write to the Free Software * Foundation, Inc., 51 Franklin Street, Fifth Floor, Boston, MA 02110-1301 USA * */ /* This file was automatically generated --- DO NOT EDIT */ /* Generated on Sun Nov 25 07:35:43 EST 2012 */ #include "codelet-dft.h" #ifdef HAVE_FMA /* Generated by: ../../../genfft/gen_notw.native -fma -reorder-insns -schedule-for-pipeline -compact -variables 4 -pipeline-latency 4 -n 13 -name n1_13 -include n.h */ /* * This function contains 176 FP additions, 114 FP multiplications, * (or, 62 additions, 0 multiplications, 114 fused multiply/add), * 87 stack variables, 25 constants, and 52 memory accesses */ #include "n.h" static void n1_13(const R *ri, const R *ii, R *ro, R *io, stride is, stride os, INT v, INT ivs, INT ovs) { DK(KP875502302, +0.875502302409147941146295545768755143177842006); DK(KP520028571, +0.520028571888864619117130500499232802493238139); DK(KP575140729, +0.575140729474003121368385547455453388461001608); DK(KP600477271, +0.600477271932665282925769253334763009352012849); DK(KP300462606, +0.300462606288665774426601772289207995520941381); DK(KP516520780, +0.516520780623489722840901288569017135705033622); DK(KP968287244, +0.968287244361984016049539446938120421179794516); DK(KP503537032, +0.503537032863766627246873853868466977093348562); DK(KP251768516, +0.251768516431883313623436926934233488546674281); DK(KP581704778, +0.581704778510515730456870384989698884939833902); DK(KP859542535, +0.859542535098774820163672132761689612766401925); DK(KP083333333, +0.083333333333333333333333333333333333333333333); DK(KP957805992, +0.957805992594665126462521754605754580515587217); DK(KP522026385, +0.522026385161275033714027226654165028300441940); DK(KP853480001, +0.853480001859823990758994934970528322872359049); DK(KP769338817, +0.769338817572980603471413688209101117038278899); DK(KP612264650, +0.612264650376756543746494474777125408779395514); DK(KP038632954, +0.038632954644348171955506895830342264440241080); DK(KP302775637, +0.302775637731994646559610633735247973125648287); DK(KP514918778, +0.514918778086315755491789696138117261566051239); DK(KP686558370, +0.686558370781754340655719594850823015421401653); DK(KP226109445, +0.226109445035782405468510155372505010481906348); DK(KP301479260, +0.301479260047709873958013540496673347309208464); DK(KP866025403, +0.866025403784438646763723170752936183471402627); DK(KP500000000, +0.500000000000000000000000000000000000000000000); { INT i; for (i = v; i > 0; i = i - 1, ri = ri + ivs, ii = ii + ivs, ro = ro + ovs, io = io + ovs, MAKE_VOLATILE_STRIDE(52, is), MAKE_VOLATILE_STRIDE(52, os)) { E T2B, T2H, T2I, T2G; { E T1, T1P, T2n, T2o, To, TH, T2h, T2k, TE, TB, TF, Tw, T2j, T2c, T1m; E T1W, T1X, T1c, T19, T1j, T12, T1f, T21, T24, T27, T1U; T1 = ri[0]; T1P = ii[0]; { E T2b, Tv, Ts, T2a; { E T2d, Tf, Tq, Ty, Tb, Tr, T6, Tx, Ti, Tt, Tu, Tl; { E T7, T8, T9, Td, Te; Td = ri[WS(is, 8)]; Te = ri[WS(is, 5)]; T7 = ri[WS(is, 12)]; T8 = ri[WS(is, 10)]; T9 = ri[WS(is, 4)]; T2d = Td - Te; Tf = Td + Te; { E T2, Ta, T3, T4; T2 = ri[WS(is, 1)]; Ta = T8 + T9; Tq = T8 - T9; T3 = ri[WS(is, 3)]; T4 = ri[WS(is, 9)]; { E Tg, T5, Th, Tj, Tk; Tg = ri[WS(is, 11)]; Ty = FMS(KP500000000, Ta, T7); Tb = T7 + Ta; Tr = T4 - T3; T5 = T3 + T4; Th = ri[WS(is, 6)]; Tj = ri[WS(is, 7)]; Tk = ri[WS(is, 2)]; T6 = T2 + T5; Tx = FNMS(KP500000000, T5, T2); Ti = Tg + Th; Tt = Tg - Th; Tu = Tj - Tk; Tl = Tj + Tk; } } } { E Tc, Tm, T2e, T2g; Tc = T6 + Tb; T2n = T6 - Tb; T2b = Ti - Tl; Tm = Ti + Tl; T2e = Tt + Tu; Tv = Tt - Tu; Ts = Tq - Tr; T2g = Tr + Tq; { E Tz, TA, Tn, T2f; Tz = Tx - Ty; T2a = Tx + Ty; TA = FNMS(KP500000000, Tm, Tf); Tn = Tf + Tm; T2f = FNMS(KP500000000, T2e, T2d); T2o = T2d + T2e; To = Tc + Tn; TH = Tc - Tn; T2h = FMA(KP866025403, T2g, T2f); T2k = FNMS(KP866025403, T2g, T2f); TE = Tz - TA; TB = Tz + TA; } } } { E T1R, TM, T10, T18, T1l, TX, T1k, T15, TP, T1a, T1b, TS; { E T16, TY, TZ, TK, TL; TK = ii[WS(is, 8)]; TF = Ts - Tv; Tw = Ts + Tv; T2j = FNMS(KP866025403, T2b, T2a); T2c = FMA(KP866025403, T2b, T2a); TL = ii[WS(is, 5)]; T16 = ii[WS(is, 12)]; TY = ii[WS(is, 10)]; TZ = ii[WS(is, 4)]; T1R = TK + TL; TM = TK - TL; { E T13, T17, TV, TW; T13 = ii[WS(is, 1)]; T17 = TY + TZ; T10 = TY - TZ; TV = ii[WS(is, 9)]; TW = ii[WS(is, 3)]; { E TN, T14, TO, TQ, TR; TN = ii[WS(is, 11)]; T18 = FMS(KP500000000, T17, T16); T1l = T16 + T17; TX = TV - TW; T14 = TW + TV; TO = ii[WS(is, 6)]; TQ = ii[WS(is, 7)]; TR = ii[WS(is, 2)]; T1k = T13 + T14; T15 = FNMS(KP500000000, T14, T13); TP = TN - TO; T1a = TN + TO; T1b = TQ + TR; TS = TQ - TR; } } } { E T1Q, T11, TT, T1S; T1Q = T1k + T1l; T1m = T1k - T1l; T11 = TX + T10; T1W = T10 - TX; T1X = TP - TS; TT = TP + TS; T1S = T1a + T1b; T1c = T1a - T1b; { E T1Z, TU, T1T, T20; T19 = T15 + T18; T1Z = T15 - T18; T1j = TM + TT; TU = FNMS(KP500000000, TT, TM); T1T = T1R + T1S; T20 = FNMS(KP500000000, T1S, T1R); T12 = FMA(KP866025403, T11, TU); T1f = FNMS(KP866025403, T11, TU); T21 = T1Z + T20; T24 = T1Z - T20; T27 = T1Q - T1T; T1U = T1Q + T1T; } } } } { E T1g, T1d, T25, T1Y; ro[0] = T1 + To; T1g = FNMS(KP866025403, T1c, T19); T1d = FMA(KP866025403, T1c, T19); T25 = T1W - T1X; T1Y = T1W + T1X; io[0] = T1P + T1U; { E T1C, T1B, T1F, T1K; { E TC, T1J, T1z, T1w, T1I, T1O, Tp, T1E, T1q, TI, T1o, T1s; { E TG, T1n, T1G, T1u, T1e, T1h, T1v, T1x, T1y, T1H, T1i; TC = FMA(KP301479260, TB, Tw); T1x = FNMS(KP226109445, Tw, TB); T1y = FMA(KP686558370, TE, TF); TG = FNMS(KP514918778, TF, TE); T1n = FNMS(KP302775637, T1m, T1j); T1G = FMA(KP302775637, T1j, T1m); T1u = FNMS(KP038632954, T12, T1d); T1e = FMA(KP038632954, T1d, T12); T1h = FMA(KP612264650, T1g, T1f); T1v = FNMS(KP612264650, T1f, T1g); T1J = FMA(KP769338817, T1y, T1x); T1z = FNMS(KP769338817, T1y, T1x); T1H = FNMS(KP853480001, T1v, T1u); T1w = FMA(KP853480001, T1v, T1u); T1I = FNMS(KP522026385, T1H, T1G); T1O = FMA(KP957805992, T1G, T1H); Tp = FNMS(KP083333333, To, T1); T1E = FMA(KP853480001, T1h, T1e); T1i = FNMS(KP853480001, T1h, T1e); T1q = FNMS(KP859542535, TG, TH); TI = FMA(KP581704778, TH, TG); T1o = FMA(KP957805992, T1n, T1i); T1s = FNMS(KP522026385, T1i, T1n); } { E T1A, T1D, T1t, T1L, T1M; { E T1p, TD, TJ, T1N, T1r; T1p = FNMS(KP251768516, TC, Tp); TD = FMA(KP503537032, TC, Tp); T1C = FNMS(KP968287244, T1z, T1w); T1A = FMA(KP968287244, T1z, T1w); TJ = FMA(KP516520780, TI, TD); T1N = FNMS(KP516520780, TI, TD); T1D = FNMS(KP300462606, T1q, T1p); T1r = FMA(KP300462606, T1q, T1p); ro[WS(os, 8)] = FNMS(KP600477271, T1O, T1N); ro[WS(os, 12)] = FMA(KP600477271, T1o, TJ); ro[WS(os, 1)] = FNMS(KP600477271, T1o, TJ); T1t = FNMS(KP575140729, T1s, T1r); T1B = FMA(KP575140729, T1s, T1r); ro[WS(os, 5)] = FMA(KP600477271, T1O, T1N); } T1L = FNMS(KP520028571, T1E, T1D); T1F = FMA(KP520028571, T1E, T1D); T1K = FMA(KP875502302, T1J, T1I); T1M = FNMS(KP875502302, T1J, T1I); ro[WS(os, 3)] = FMA(KP520028571, T1A, T1t); ro[WS(os, 9)] = FNMS(KP520028571, T1A, T1t); ro[WS(os, 6)] = FMA(KP575140729, T1M, T1L); ro[WS(os, 11)] = FNMS(KP575140729, T1M, T1L); } } { E T22, T2F, T2N, T2K, T2w, T2A, T1V, T2C, T28, T2y, T2M, T2q; { E T26, T2v, T2p, T2i, T2s, T2t, T2l, T2D, T2E, T2u, T2m; T2D = FNMS(KP226109445, T1Y, T21); T22 = FMA(KP301479260, T21, T1Y); ro[WS(os, 2)] = FMA(KP575140729, T1K, T1F); ro[WS(os, 7)] = FNMS(KP575140729, T1K, T1F); ro[WS(os, 4)] = FMA(KP520028571, T1C, T1B); ro[WS(os, 10)] = FNMS(KP520028571, T1C, T1B); T26 = FNMS(KP514918778, T25, T24); T2E = FMA(KP686558370, T24, T25); T2v = FNMS(KP302775637, T2n, T2o); T2p = FMA(KP302775637, T2o, T2n); T2i = FNMS(KP038632954, T2h, T2c); T2s = FMA(KP038632954, T2c, T2h); T2t = FMA(KP612264650, T2j, T2k); T2l = FNMS(KP612264650, T2k, T2j); T2F = FNMS(KP769338817, T2E, T2D); T2N = FMA(KP769338817, T2E, T2D); T2K = FMA(KP853480001, T2t, T2s); T2u = FNMS(KP853480001, T2t, T2s); T2w = FMA(KP957805992, T2v, T2u); T2A = FNMS(KP522026385, T2u, T2v); T1V = FNMS(KP083333333, T1U, T1P); T2m = FNMS(KP853480001, T2l, T2i); T2C = FMA(KP853480001, T2l, T2i); T28 = FMA(KP581704778, T27, T26); T2y = FNMS(KP859542535, T26, T27); T2M = FNMS(KP522026385, T2m, T2p); T2q = FMA(KP957805992, T2p, T2m); } { E T2O, T2Q, T2z, T2P, T2L; { E T23, T2x, T2r, T29, T2J; T23 = FMA(KP503537032, T22, T1V); T2x = FNMS(KP251768516, T22, T1V); T2O = FNMS(KP875502302, T2N, T2M); T2Q = FMA(KP875502302, T2N, T2M); T2r = FMA(KP516520780, T28, T23); T29 = FNMS(KP516520780, T28, T23); T2z = FMA(KP300462606, T2y, T2x); T2J = FNMS(KP300462606, T2y, T2x); io[WS(os, 12)] = FNMS(KP600477271, T2w, T2r); io[WS(os, 1)] = FMA(KP600477271, T2w, T2r); io[WS(os, 8)] = FMA(KP600477271, T2q, T29); io[WS(os, 5)] = FNMS(KP600477271, T2q, T29); T2P = FMA(KP520028571, T2K, T2J); T2L = FNMS(KP520028571, T2K, T2J); } T2B = FMA(KP575140729, T2A, T2z); T2H = FNMS(KP575140729, T2A, T2z); io[WS(os, 11)] = FMA(KP575140729, T2Q, T2P); io[WS(os, 6)] = FNMS(KP575140729, T2Q, T2P); io[WS(os, 7)] = FMA(KP575140729, T2O, T2L); io[WS(os, 2)] = FNMS(KP575140729, T2O, T2L); T2I = FMA(KP968287244, T2F, T2C); T2G = FNMS(KP968287244, T2F, T2C); } } } } } io[WS(os, 10)] = FMA(KP520028571, T2I, T2H); io[WS(os, 4)] = FNMS(KP520028571, T2I, T2H); io[WS(os, 9)] = FMA(KP520028571, T2G, T2B); io[WS(os, 3)] = FNMS(KP520028571, T2G, T2B); } } } static const kdft_desc desc = { 13, "n1_13", {62, 0, 114, 0}, &GENUS, 0, 0, 0, 0 }; void X(codelet_n1_13) (planner *p) { X(kdft_register) (p, n1_13, &desc); } #else /* HAVE_FMA */ /* Generated by: ../../../genfft/gen_notw.native -compact -variables 4 -pipeline-latency 4 -n 13 -name n1_13 -include n.h */ /* * This function contains 176 FP additions, 68 FP multiplications, * (or, 138 additions, 30 multiplications, 38 fused multiply/add), * 71 stack variables, 20 constants, and 52 memory accesses */ #include "n.h" static void n1_13(const R *ri, const R *ii, R *ro, R *io, stride is, stride os, INT v, INT ivs, INT ovs) { DK(KP2_000000000, +2.000000000000000000000000000000000000000000000); DK(KP083333333, +0.083333333333333333333333333333333333333333333); DK(KP251768516, +0.251768516431883313623436926934233488546674281); DK(KP075902986, +0.075902986037193865983102897245103540356428373); DK(KP132983124, +0.132983124607418643793760531921092974399165133); DK(KP258260390, +0.258260390311744861420450644284508567852516811); DK(KP1_732050807, +1.732050807568877293527446341505872366942805254); DK(KP300238635, +0.300238635966332641462884626667381504676006424); DK(KP011599105, +0.011599105605768290721655456654083252189827041); DK(KP156891391, +0.156891391051584611046832726756003269660212636); DK(KP256247671, +0.256247671582936600958684654061725059144125175); DK(KP174138601, +0.174138601152135905005660794929264742616964676); DK(KP575140729, +0.575140729474003121368385547455453388461001608); DK(KP503537032, +0.503537032863766627246873853868466977093348562); DK(KP113854479, +0.113854479055790798974654345867655310534642560); DK(KP265966249, +0.265966249214837287587521063842185948798330267); DK(KP387390585, +0.387390585467617292130675966426762851778775217); DK(KP866025403, +0.866025403784438646763723170752936183471402627); DK(KP300462606, +0.300462606288665774426601772289207995520941381); DK(KP500000000, +0.500000000000000000000000000000000000000000000); { INT i; for (i = v; i > 0; i = i - 1, ri = ri + ivs, ii = ii + ivs, ro = ro + ovs, io = io + ovs, MAKE_VOLATILE_STRIDE(52, is), MAKE_VOLATILE_STRIDE(52, os)) { E T1, T1q, Tt, Tu, To, T22, T20, T24, TF, TH, TA, TI, T1X, T25, T2a; E T2d, T18, T1n, T2k, T2n, T1l, T1r, T1f, T1o, T2h, T2m; T1 = ri[0]; T1q = ii[0]; { E Tf, Tp, Tb, TC, Tx, T6, TB, Tw, Ti, Tq, Tl, Tr, Tm, Ts, Td; E Te, Tc, Tn; Td = ri[WS(is, 8)]; Te = ri[WS(is, 5)]; Tf = Td + Te; Tp = Td - Te; { E T7, T8, T9, Ta; T7 = ri[WS(is, 12)]; T8 = ri[WS(is, 10)]; T9 = ri[WS(is, 4)]; Ta = T8 + T9; Tb = T7 + Ta; TC = T8 - T9; Tx = FNMS(KP500000000, Ta, T7); } { E T2, T3, T4, T5; T2 = ri[WS(is, 1)]; T3 = ri[WS(is, 3)]; T4 = ri[WS(is, 9)]; T5 = T3 + T4; T6 = T2 + T5; TB = T3 - T4; Tw = FNMS(KP500000000, T5, T2); } { E Tg, Th, Tj, Tk; Tg = ri[WS(is, 11)]; Th = ri[WS(is, 6)]; Ti = Tg + Th; Tq = Tg - Th; Tj = ri[WS(is, 7)]; Tk = ri[WS(is, 2)]; Tl = Tj + Tk; Tr = Tj - Tk; } Tm = Ti + Tl; Ts = Tq + Tr; Tt = Tp + Ts; Tu = T6 - Tb; Tc = T6 + Tb; Tn = Tf + Tm; To = Tc + Tn; T22 = KP300462606 * (Tc - Tn); { E T1Y, T1Z, TD, TE; T1Y = TB + TC; T1Z = Tq - Tr; T20 = T1Y - T1Z; T24 = T1Y + T1Z; TD = KP866025403 * (TB - TC); TE = FNMS(KP500000000, Ts, Tp); TF = TD - TE; TH = TD + TE; } { E Ty, Tz, T1V, T1W; Ty = Tw - Tx; Tz = KP866025403 * (Ti - Tl); TA = Ty + Tz; TI = Ty - Tz; T1V = Tw + Tx; T1W = FNMS(KP500000000, Tm, Tf); T1X = T1V - T1W; T25 = T1V + T1W; } } { E TZ, T2b, TV, T1i, T1a, TQ, T1h, T19, T12, T1d, T15, T1c, T16, T2c, TX; E TY, TW, T17; TX = ii[WS(is, 8)]; TY = ii[WS(is, 5)]; TZ = TX + TY; T2b = TX - TY; { E TR, TS, TT, TU; TR = ii[WS(is, 12)]; TS = ii[WS(is, 10)]; TT = ii[WS(is, 4)]; TU = TS + TT; TV = FNMS(KP500000000, TU, TR); T1i = TR + TU; T1a = TS - TT; } { E TM, TN, TO, TP; TM = ii[WS(is, 1)]; TN = ii[WS(is, 3)]; TO = ii[WS(is, 9)]; TP = TN + TO; TQ = FNMS(KP500000000, TP, TM); T1h = TM + TP; T19 = TN - TO; } { E T10, T11, T13, T14; T10 = ii[WS(is, 11)]; T11 = ii[WS(is, 6)]; T12 = T10 + T11; T1d = T10 - T11; T13 = ii[WS(is, 7)]; T14 = ii[WS(is, 2)]; T15 = T13 + T14; T1c = T13 - T14; } T16 = T12 + T15; T2c = T1d + T1c; T2a = T1h - T1i; T2d = T2b + T2c; TW = TQ + TV; T17 = FNMS(KP500000000, T16, TZ); T18 = TW - T17; T1n = TW + T17; { E T2i, T2j, T1j, T1k; T2i = TQ - TV; T2j = KP866025403 * (T15 - T12); T2k = T2i + T2j; T2n = T2i - T2j; T1j = T1h + T1i; T1k = TZ + T16; T1l = KP300462606 * (T1j - T1k); T1r = T1j + T1k; } { E T1b, T1e, T2f, T2g; T1b = T19 + T1a; T1e = T1c - T1d; T1f = T1b + T1e; T1o = T1e - T1b; T2f = FNMS(KP500000000, T2c, T2b); T2g = KP866025403 * (T1a - T19); T2h = T2f - T2g; T2m = T2g + T2f; } } ro[0] = T1 + To; io[0] = T1q + T1r; { E T1D, T1N, T1y, T1x, T1E, T1O, Tv, TK, T1J, T1Q, T1m, T1R, T1t, T1I, TG; E TJ; { E T1B, T1C, T1v, T1w; T1B = FMA(KP387390585, T1f, KP265966249 * T18); T1C = FMA(KP113854479, T1o, KP503537032 * T1n); T1D = T1B + T1C; T1N = T1C - T1B; T1y = FMA(KP575140729, Tu, KP174138601 * Tt); T1v = FNMS(KP156891391, TH, KP256247671 * TI); T1w = FMA(KP011599105, TF, KP300238635 * TA); T1x = T1v - T1w; T1E = T1y + T1x; T1O = KP1_732050807 * (T1v + T1w); } Tv = FNMS(KP174138601, Tu, KP575140729 * Tt); TG = FNMS(KP300238635, TF, KP011599105 * TA); TJ = FMA(KP256247671, TH, KP156891391 * TI); TK = TG - TJ; T1J = KP1_732050807 * (TJ + TG); T1Q = Tv - TK; { E T1g, T1H, T1p, T1s, T1G; T1g = FNMS(KP132983124, T1f, KP258260390 * T18); T1H = T1l - T1g; T1p = FNMS(KP251768516, T1o, KP075902986 * T1n); T1s = FNMS(KP083333333, T1r, T1q); T1G = T1s - T1p; T1m = FMA(KP2_000000000, T1g, T1l); T1R = T1H + T1G; T1t = FMA(KP2_000000000, T1p, T1s); T1I = T1G - T1H; } { E TL, T1u, T1P, T1S; TL = FMA(KP2_000000000, TK, Tv); T1u = T1m + T1t; io[WS(os, 1)] = TL + T1u; io[WS(os, 12)] = T1u - TL; { E T1z, T1A, T1T, T1U; T1z = FMS(KP2_000000000, T1x, T1y); T1A = T1t - T1m; io[WS(os, 5)] = T1z + T1A; io[WS(os, 8)] = T1A - T1z; T1T = T1R - T1Q; T1U = T1O + T1N; io[WS(os, 4)] = T1T - T1U; io[WS(os, 10)] = T1U + T1T; } T1P = T1N - T1O; T1S = T1Q + T1R; io[WS(os, 3)] = T1P + T1S; io[WS(os, 9)] = T1S - T1P; { E T1L, T1M, T1F, T1K; T1L = T1J + T1I; T1M = T1E + T1D; io[WS(os, 6)] = T1L - T1M; io[WS(os, 11)] = T1M + T1L; T1F = T1D - T1E; T1K = T1I - T1J; io[WS(os, 2)] = T1F + T1K; io[WS(os, 7)] = T1K - T1F; } } } { E T2y, T2I, T2J, T2K, T2B, T2L, T2e, T2p, T2u, T2G, T23, T2F, T28, T2t, T2l; E T2o; { E T2w, T2x, T2z, T2A; T2w = FMA(KP387390585, T20, KP265966249 * T1X); T2x = FNMS(KP503537032, T25, KP113854479 * T24); T2y = T2w + T2x; T2I = T2w - T2x; T2J = FMA(KP575140729, T2a, KP174138601 * T2d); T2z = FNMS(KP300238635, T2n, KP011599105 * T2m); T2A = FNMS(KP156891391, T2h, KP256247671 * T2k); T2K = T2z + T2A; T2B = KP1_732050807 * (T2z - T2A); T2L = T2J + T2K; } T2e = FNMS(KP575140729, T2d, KP174138601 * T2a); T2l = FMA(KP256247671, T2h, KP156891391 * T2k); T2o = FMA(KP300238635, T2m, KP011599105 * T2n); T2p = T2l - T2o; T2u = T2e - T2p; T2G = KP1_732050807 * (T2o + T2l); { E T21, T2r, T26, T27, T2s; T21 = FNMS(KP132983124, T20, KP258260390 * T1X); T2r = T22 - T21; T26 = FMA(KP251768516, T24, KP075902986 * T25); T27 = FNMS(KP083333333, To, T1); T2s = T27 - T26; T23 = FMA(KP2_000000000, T21, T22); T2F = T2s - T2r; T28 = FMA(KP2_000000000, T26, T27); T2t = T2r + T2s; } { E T29, T2q, T2N, T2O; T29 = T23 + T28; T2q = FMA(KP2_000000000, T2p, T2e); ro[WS(os, 12)] = T29 - T2q; ro[WS(os, 1)] = T29 + T2q; { E T2v, T2C, T2P, T2Q; T2v = T2t - T2u; T2C = T2y - T2B; ro[WS(os, 10)] = T2v - T2C; ro[WS(os, 4)] = T2v + T2C; T2P = T28 - T23; T2Q = FMS(KP2_000000000, T2K, T2J); ro[WS(os, 5)] = T2P - T2Q; ro[WS(os, 8)] = T2P + T2Q; } T2N = T2F - T2G; T2O = T2L - T2I; ro[WS(os, 11)] = T2N - T2O; ro[WS(os, 6)] = T2N + T2O; { E T2H, T2M, T2D, T2E; T2H = T2F + T2G; T2M = T2I + T2L; ro[WS(os, 7)] = T2H - T2M; ro[WS(os, 2)] = T2H + T2M; T2D = T2t + T2u; T2E = T2y + T2B; ro[WS(os, 3)] = T2D - T2E; ro[WS(os, 9)] = T2D + T2E; } } } } } } static const kdft_desc desc = { 13, "n1_13", {138, 30, 38, 0}, &GENUS, 0, 0, 0, 0 }; void X(codelet_n1_13) (planner *p) { X(kdft_register) (p, n1_13, &desc); } #endif /* HAVE_FMA */ fftw-3.3.3/dft/scalar/codelets/n1_14.c0000644000175000017500000003700712054410241014221 00000000000000/* * Copyright (c) 2003, 2007-11 Matteo Frigo * Copyright (c) 2003, 2007-11 Massachusetts Institute of Technology * * This program is free software; you can redistribute it and/or modify * it under the terms of the GNU General Public License as published by * the Free Software Foundation; either version 2 of the License, or * (at your option) any later version. * * This program is distributed in the hope that it will be useful, * but WITHOUT ANY WARRANTY; without even the implied warranty of * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the * GNU General Public License for more details. * * You should have received a copy of the GNU General Public License * along with this program; if not, write to the Free Software * Foundation, Inc., 51 Franklin Street, Fifth Floor, Boston, MA 02110-1301 USA * */ /* This file was automatically generated --- DO NOT EDIT */ /* Generated on Sun Nov 25 07:35:43 EST 2012 */ #include "codelet-dft.h" #ifdef HAVE_FMA /* Generated by: ../../../genfft/gen_notw.native -fma -reorder-insns -schedule-for-pipeline -compact -variables 4 -pipeline-latency 4 -n 14 -name n1_14 -include n.h */ /* * This function contains 148 FP additions, 84 FP multiplications, * (or, 64 additions, 0 multiplications, 84 fused multiply/add), * 80 stack variables, 6 constants, and 56 memory accesses */ #include "n.h" static void n1_14(const R *ri, const R *ii, R *ro, R *io, stride is, stride os, INT v, INT ivs, INT ovs) { DK(KP974927912, +0.974927912181823607018131682993931217232785801); DK(KP801937735, +0.801937735804838252472204639014890102331838324); DK(KP900968867, +0.900968867902419126236102319507445051165919162); DK(KP554958132, +0.554958132087371191422194871006410481067288862); DK(KP692021471, +0.692021471630095869627814897002069140197260599); DK(KP356895867, +0.356895867892209443894399510021300583399127187); { INT i; for (i = v; i > 0; i = i - 1, ri = ri + ivs, ii = ii + ivs, ro = ro + ovs, io = io + ovs, MAKE_VOLATILE_STRIDE(56, is), MAKE_VOLATILE_STRIDE(56, os)) { E Tp, T1L, T24, T1W, T1X, T28, T2a, T1Y, T29, T2b; { E T3, T1x, T1b, To, T1i, T1M, Ts, Ta, T1k, Tv, Th, T1j, T1K, Ty, TZ; E T14, Tz, T1Z, T27, T2c, T1d, TI, T23, T1G, T1D, TW, T1e, T22, T1A, TP; E T1c, T1n, T1s, T1f, T1P; { E T1, T2, T19, T1a; T1 = ri[0]; T2 = ri[WS(is, 7)]; T19 = ii[0]; T1a = ii[WS(is, 7)]; { E Tq, T6, Tr, T9, Te, Tx, Tn, Tw, Tk, Tf, Tb, Tc; { E Tl, Tm, Ti, Tj; { E T4, T5, T7, T8; T4 = ri[WS(is, 2)]; Tp = T1 + T2; T3 = T1 - T2; T1x = T19 + T1a; T1b = T19 - T1a; T5 = ri[WS(is, 9)]; T7 = ri[WS(is, 12)]; T8 = ri[WS(is, 5)]; Tl = ri[WS(is, 8)]; Tq = T4 + T5; T6 = T4 - T5; Tr = T7 + T8; T9 = T7 - T8; Tm = ri[WS(is, 1)]; } Ti = ri[WS(is, 6)]; Tj = ri[WS(is, 13)]; Te = ri[WS(is, 10)]; Tx = Tl + Tm; Tn = Tl - Tm; Tw = Ti + Tj; Tk = Ti - Tj; Tf = ri[WS(is, 3)]; Tb = ri[WS(is, 4)]; Tc = ri[WS(is, 11)]; } { E Tu, Tg, Tt, Td; To = Tk + Tn; T1i = Tn - Tk; Tu = Te + Tf; Tg = Te - Tf; Tt = Tb + Tc; Td = Tb - Tc; T1M = Tr - Tq; Ts = Tq + Tr; Ta = T6 + T9; T1k = T9 - T6; T1L = Tt - Tu; Tv = Tt + Tu; Th = Td + Tg; T1j = Tg - Td; T1K = Tw - Tx; Ty = Tw + Tx; TZ = FNMS(KP356895867, Ta, To); T14 = FNMS(KP356895867, To, Th); Tz = FNMS(KP356895867, Th, Ta); T1Z = FNMS(KP356895867, Ts, Ty); } } { E T1B, TE, T1C, TH, T1F, TV, TJ, T1E, TS, T1z, TO, TK, T1y, TL; { E TF, TG, TT, TU, TC, TD; TC = ii[WS(is, 4)]; TD = ii[WS(is, 11)]; T27 = FNMS(KP356895867, Tv, Ts); T2c = FNMS(KP356895867, Ty, Tv); TF = ii[WS(is, 10)]; T1B = TC + TD; TE = TC - TD; TG = ii[WS(is, 3)]; TT = ii[WS(is, 8)]; TU = ii[WS(is, 1)]; { E TQ, TR, TM, TN; TQ = ii[WS(is, 6)]; T1C = TF + TG; TH = TF - TG; T1F = TT + TU; TV = TT - TU; TR = ii[WS(is, 13)]; TM = ii[WS(is, 12)]; TN = ii[WS(is, 5)]; TJ = ii[WS(is, 2)]; T1E = TQ + TR; TS = TQ - TR; T1z = TM + TN; TO = TM - TN; TK = ii[WS(is, 9)]; } } T1d = TE + TH; TI = TE - TH; T23 = T1F - T1E; T1G = T1E + T1F; T1D = T1B + T1C; T24 = T1C - T1B; T1y = TJ + TK; TL = TJ - TK; TW = TS - TV; T1e = TS + TV; T22 = T1y - T1z; T1A = T1y + T1z; TP = TL - TO; T1c = TL + TO; T1n = FNMS(KP356895867, T1c, T1e); T1s = FNMS(KP356895867, T1d, T1c); T1f = FNMS(KP356895867, T1e, T1d); T1P = FNMS(KP356895867, T1A, T1G); } } { E T1U, T1H, T11, T12, T1o, T1q; ro[WS(os, 7)] = T3 + Ta + Th + To; io[WS(os, 7)] = T1b + T1c + T1d + T1e; T1U = FNMS(KP356895867, T1D, T1A); T1H = FNMS(KP356895867, T1G, T1D); ro[0] = Tp + Ts + Tv + Ty; io[0] = T1x + T1A + T1D + T1G; { E TB, TY, T1u, T1w, T10; { E TA, TX, T1t, T1v; TA = FNMS(KP692021471, Tz, To); TX = FMA(KP554958132, TW, TP); T1t = FNMS(KP692021471, T1s, T1e); T1v = FMA(KP554958132, T1i, T1k); TB = FNMS(KP900968867, TA, T3); TY = FMA(KP801937735, TX, TI); T1u = FNMS(KP900968867, T1t, T1b); T1w = FMA(KP801937735, T1v, T1j); } T10 = FNMS(KP692021471, TZ, Th); ro[WS(os, 1)] = FMA(KP974927912, TY, TB); ro[WS(os, 13)] = FNMS(KP974927912, TY, TB); io[WS(os, 13)] = FNMS(KP974927912, T1w, T1u); io[WS(os, 1)] = FMA(KP974927912, T1w, T1u); T11 = FNMS(KP900968867, T10, T3); T12 = FMA(KP554958132, TI, TW); T1o = FNMS(KP692021471, T1n, T1d); T1q = FMA(KP554958132, T1j, T1i); } { E T1J, T1N, T2d, T2f; { E T16, T17, T1g, T1l; { E T13, T1p, T1r, T15; T15 = FNMS(KP692021471, T14, Ta); T13 = FNMS(KP801937735, T12, TP); T1p = FNMS(KP900968867, T1o, T1b); T1r = FNMS(KP801937735, T1q, T1k); T16 = FNMS(KP900968867, T15, T3); ro[WS(os, 9)] = FMA(KP974927912, T13, T11); ro[WS(os, 5)] = FNMS(KP974927912, T13, T11); io[WS(os, 9)] = FMA(KP974927912, T1r, T1p); io[WS(os, 5)] = FNMS(KP974927912, T1r, T1p); T17 = FNMS(KP554958132, TP, TI); } T1g = FNMS(KP692021471, T1f, T1c); T1l = FNMS(KP554958132, T1k, T1j); { E T18, T1h, T1m, T1I; T1I = FNMS(KP692021471, T1H, T1A); T18 = FNMS(KP801937735, T17, TW); T1h = FNMS(KP900968867, T1g, T1b); T1m = FNMS(KP801937735, T1l, T1i); T1J = FNMS(KP900968867, T1I, T1x); ro[WS(os, 3)] = FMA(KP974927912, T18, T16); ro[WS(os, 11)] = FNMS(KP974927912, T18, T16); io[WS(os, 11)] = FNMS(KP974927912, T1m, T1h); io[WS(os, 3)] = FMA(KP974927912, T1m, T1h); T1N = FMA(KP554958132, T1M, T1L); } T2d = FNMS(KP692021471, T2c, Ts); T2f = FMA(KP554958132, T22, T24); } { E T1R, T1S, T20, T25; { E T1O, T2e, T2g, T1Q; T1Q = FNMS(KP692021471, T1P, T1D); T1O = FNMS(KP801937735, T1N, T1K); T2e = FNMS(KP900968867, T2d, Tp); T2g = FNMS(KP801937735, T2f, T23); T1R = FNMS(KP900968867, T1Q, T1x); io[WS(os, 10)] = FNMS(KP974927912, T1O, T1J); io[WS(os, 4)] = FMA(KP974927912, T1O, T1J); ro[WS(os, 4)] = FMA(KP974927912, T2g, T2e); ro[WS(os, 10)] = FNMS(KP974927912, T2g, T2e); T1S = FMA(KP554958132, T1L, T1K); } T20 = FNMS(KP692021471, T1Z, Tv); T25 = FMA(KP554958132, T24, T23); { E T1T, T21, T26, T1V; T1V = FNMS(KP692021471, T1U, T1G); T1T = FMA(KP801937735, T1S, T1M); T21 = FNMS(KP900968867, T20, Tp); T26 = FMA(KP801937735, T25, T22); T1W = FNMS(KP900968867, T1V, T1x); io[WS(os, 12)] = FNMS(KP974927912, T1T, T1R); io[WS(os, 2)] = FMA(KP974927912, T1T, T1R); ro[WS(os, 2)] = FMA(KP974927912, T26, T21); ro[WS(os, 12)] = FNMS(KP974927912, T26, T21); T1X = FNMS(KP554958132, T1K, T1M); } T28 = FNMS(KP692021471, T27, Ty); T2a = FNMS(KP554958132, T23, T22); } } } } T1Y = FNMS(KP801937735, T1X, T1L); T29 = FNMS(KP900968867, T28, Tp); T2b = FNMS(KP801937735, T2a, T24); io[WS(os, 8)] = FNMS(KP974927912, T1Y, T1W); io[WS(os, 6)] = FMA(KP974927912, T1Y, T1W); ro[WS(os, 6)] = FMA(KP974927912, T2b, T29); ro[WS(os, 8)] = FNMS(KP974927912, T2b, T29); } } } static const kdft_desc desc = { 14, "n1_14", {64, 0, 84, 0}, &GENUS, 0, 0, 0, 0 }; void X(codelet_n1_14) (planner *p) { X(kdft_register) (p, n1_14, &desc); } #else /* HAVE_FMA */ /* Generated by: ../../../genfft/gen_notw.native -compact -variables 4 -pipeline-latency 4 -n 14 -name n1_14 -include n.h */ /* * This function contains 148 FP additions, 72 FP multiplications, * (or, 100 additions, 24 multiplications, 48 fused multiply/add), * 43 stack variables, 6 constants, and 56 memory accesses */ #include "n.h" static void n1_14(const R *ri, const R *ii, R *ro, R *io, stride is, stride os, INT v, INT ivs, INT ovs) { DK(KP222520933, +0.222520933956314404288902564496794759466355569); DK(KP900968867, +0.900968867902419126236102319507445051165919162); DK(KP623489801, +0.623489801858733530525004884004239810632274731); DK(KP433883739, +0.433883739117558120475768332848358754609990728); DK(KP781831482, +0.781831482468029808708444526674057750232334519); DK(KP974927912, +0.974927912181823607018131682993931217232785801); { INT i; for (i = v; i > 0; i = i - 1, ri = ri + ivs, ii = ii + ivs, ro = ro + ovs, io = io + ovs, MAKE_VOLATILE_STRIDE(56, is), MAKE_VOLATILE_STRIDE(56, os)) { E T3, Tp, T16, T1f, Ta, T1q, Ts, T10, TG, T1z, T19, T1i, Th, T1s, Tv; E T12, TU, T1B, T17, T1o, To, T1r, Ty, T11, TN, T1A, T18, T1l; { E T1, T2, T14, T15; T1 = ri[0]; T2 = ri[WS(is, 7)]; T3 = T1 - T2; Tp = T1 + T2; T14 = ii[0]; T15 = ii[WS(is, 7)]; T16 = T14 - T15; T1f = T14 + T15; } { E T6, Tq, T9, Tr; { E T4, T5, T7, T8; T4 = ri[WS(is, 2)]; T5 = ri[WS(is, 9)]; T6 = T4 - T5; Tq = T4 + T5; T7 = ri[WS(is, 12)]; T8 = ri[WS(is, 5)]; T9 = T7 - T8; Tr = T7 + T8; } Ta = T6 + T9; T1q = Tr - Tq; Ts = Tq + Tr; T10 = T9 - T6; } { E TC, T1g, TF, T1h; { E TA, TB, TD, TE; TA = ii[WS(is, 2)]; TB = ii[WS(is, 9)]; TC = TA - TB; T1g = TA + TB; TD = ii[WS(is, 12)]; TE = ii[WS(is, 5)]; TF = TD - TE; T1h = TD + TE; } TG = TC - TF; T1z = T1g - T1h; T19 = TC + TF; T1i = T1g + T1h; } { E Td, Tt, Tg, Tu; { E Tb, Tc, Te, Tf; Tb = ri[WS(is, 4)]; Tc = ri[WS(is, 11)]; Td = Tb - Tc; Tt = Tb + Tc; Te = ri[WS(is, 10)]; Tf = ri[WS(is, 3)]; Tg = Te - Tf; Tu = Te + Tf; } Th = Td + Tg; T1s = Tt - Tu; Tv = Tt + Tu; T12 = Tg - Td; } { E TQ, T1m, TT, T1n; { E TO, TP, TR, TS; TO = ii[WS(is, 4)]; TP = ii[WS(is, 11)]; TQ = TO - TP; T1m = TO + TP; TR = ii[WS(is, 10)]; TS = ii[WS(is, 3)]; TT = TR - TS; T1n = TR + TS; } TU = TQ - TT; T1B = T1n - T1m; T17 = TQ + TT; T1o = T1m + T1n; } { E Tk, Tw, Tn, Tx; { E Ti, Tj, Tl, Tm; Ti = ri[WS(is, 6)]; Tj = ri[WS(is, 13)]; Tk = Ti - Tj; Tw = Ti + Tj; Tl = ri[WS(is, 8)]; Tm = ri[WS(is, 1)]; Tn = Tl - Tm; Tx = Tl + Tm; } To = Tk + Tn; T1r = Tw - Tx; Ty = Tw + Tx; T11 = Tn - Tk; } { E TJ, T1j, TM, T1k; { E TH, TI, TK, TL; TH = ii[WS(is, 6)]; TI = ii[WS(is, 13)]; TJ = TH - TI; T1j = TH + TI; TK = ii[WS(is, 8)]; TL = ii[WS(is, 1)]; TM = TK - TL; T1k = TK + TL; } TN = TJ - TM; T1A = T1k - T1j; T18 = TJ + TM; T1l = T1j + T1k; } ro[WS(os, 7)] = T3 + Ta + Th + To; io[WS(os, 7)] = T16 + T19 + T17 + T18; ro[0] = Tp + Ts + Tv + Ty; io[0] = T1f + T1i + T1o + T1l; { E TV, Tz, T1e, T1d; TV = FNMS(KP781831482, TN, KP974927912 * TG) - (KP433883739 * TU); Tz = FMA(KP623489801, To, T3) + FNMA(KP900968867, Th, KP222520933 * Ta); ro[WS(os, 5)] = Tz - TV; ro[WS(os, 9)] = Tz + TV; T1e = FNMS(KP781831482, T11, KP974927912 * T10) - (KP433883739 * T12); T1d = FMA(KP623489801, T18, T16) + FNMA(KP900968867, T17, KP222520933 * T19); io[WS(os, 5)] = T1d - T1e; io[WS(os, 9)] = T1e + T1d; } { E TX, TW, T1b, T1c; TX = FMA(KP781831482, TG, KP974927912 * TU) + (KP433883739 * TN); TW = FMA(KP623489801, Ta, T3) + FNMA(KP900968867, To, KP222520933 * Th); ro[WS(os, 13)] = TW - TX; ro[WS(os, 1)] = TW + TX; T1b = FMA(KP781831482, T10, KP974927912 * T12) + (KP433883739 * T11); T1c = FMA(KP623489801, T19, T16) + FNMA(KP900968867, T18, KP222520933 * T17); io[WS(os, 1)] = T1b + T1c; io[WS(os, 13)] = T1c - T1b; } { E TZ, TY, T13, T1a; TZ = FMA(KP433883739, TG, KP974927912 * TN) - (KP781831482 * TU); TY = FMA(KP623489801, Th, T3) + FNMA(KP222520933, To, KP900968867 * Ta); ro[WS(os, 11)] = TY - TZ; ro[WS(os, 3)] = TY + TZ; T13 = FMA(KP433883739, T10, KP974927912 * T11) - (KP781831482 * T12); T1a = FMA(KP623489801, T17, T16) + FNMA(KP222520933, T18, KP900968867 * T19); io[WS(os, 3)] = T13 + T1a; io[WS(os, 11)] = T1a - T13; } { E T1t, T1p, T1C, T1y; T1t = FNMS(KP433883739, T1r, KP781831482 * T1q) - (KP974927912 * T1s); T1p = FMA(KP623489801, T1i, T1f) + FNMA(KP900968867, T1l, KP222520933 * T1o); io[WS(os, 6)] = T1p - T1t; io[WS(os, 8)] = T1t + T1p; T1C = FNMS(KP433883739, T1A, KP781831482 * T1z) - (KP974927912 * T1B); T1y = FMA(KP623489801, Ts, Tp) + FNMA(KP900968867, Ty, KP222520933 * Tv); ro[WS(os, 6)] = T1y - T1C; ro[WS(os, 8)] = T1y + T1C; } { E T1v, T1u, T1E, T1D; T1v = FMA(KP433883739, T1q, KP781831482 * T1s) - (KP974927912 * T1r); T1u = FMA(KP623489801, T1o, T1f) + FNMA(KP222520933, T1l, KP900968867 * T1i); io[WS(os, 4)] = T1u - T1v; io[WS(os, 10)] = T1v + T1u; T1E = FMA(KP433883739, T1z, KP781831482 * T1B) - (KP974927912 * T1A); T1D = FMA(KP623489801, Tv, Tp) + FNMA(KP222520933, Ty, KP900968867 * Ts); ro[WS(os, 4)] = T1D - T1E; ro[WS(os, 10)] = T1D + T1E; } { E T1w, T1x, T1G, T1F; T1w = FMA(KP974927912, T1q, KP433883739 * T1s) + (KP781831482 * T1r); T1x = FMA(KP623489801, T1l, T1f) + FNMA(KP900968867, T1o, KP222520933 * T1i); io[WS(os, 2)] = T1w + T1x; io[WS(os, 12)] = T1x - T1w; T1G = FMA(KP974927912, T1z, KP433883739 * T1B) + (KP781831482 * T1A); T1F = FMA(KP623489801, Ty, Tp) + FNMA(KP900968867, Tv, KP222520933 * Ts); ro[WS(os, 12)] = T1F - T1G; ro[WS(os, 2)] = T1F + T1G; } } } } static const kdft_desc desc = { 14, "n1_14", {100, 24, 48, 0}, &GENUS, 0, 0, 0, 0 }; void X(codelet_n1_14) (planner *p) { X(kdft_register) (p, n1_14, &desc); } #endif /* HAVE_FMA */ fftw-3.3.3/dft/scalar/codelets/n1_15.c0000644000175000017500000004202312054410240014213 00000000000000/* * Copyright (c) 2003, 2007-11 Matteo Frigo * Copyright (c) 2003, 2007-11 Massachusetts Institute of Technology * * This program is free software; you can redistribute it and/or modify * it under the terms of the GNU General Public License as published by * the Free Software Foundation; either version 2 of the License, or * (at your option) any later version. * * This program is distributed in the hope that it will be useful, * but WITHOUT ANY WARRANTY; without even the implied warranty of * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the * GNU General Public License for more details. * * You should have received a copy of the GNU General Public License * along with this program; if not, write to the Free Software * Foundation, Inc., 51 Franklin Street, Fifth Floor, Boston, MA 02110-1301 USA * */ /* This file was automatically generated --- DO NOT EDIT */ /* Generated on Sun Nov 25 07:35:43 EST 2012 */ #include "codelet-dft.h" #ifdef HAVE_FMA /* Generated by: ../../../genfft/gen_notw.native -fma -reorder-insns -schedule-for-pipeline -compact -variables 4 -pipeline-latency 4 -n 15 -name n1_15 -include n.h */ /* * This function contains 156 FP additions, 84 FP multiplications, * (or, 72 additions, 0 multiplications, 84 fused multiply/add), * 75 stack variables, 6 constants, and 60 memory accesses */ #include "n.h" static void n1_15(const R *ri, const R *ii, R *ro, R *io, stride is, stride os, INT v, INT ivs, INT ovs) { DK(KP951056516, +0.951056516295153572116439333379382143405698634); DK(KP559016994, +0.559016994374947424102293417182819058860154590); DK(KP618033988, +0.618033988749894848204586834365638117720309180); DK(KP250000000, +0.250000000000000000000000000000000000000000000); DK(KP866025403, +0.866025403784438646763723170752936183471402627); DK(KP500000000, +0.500000000000000000000000000000000000000000000); { INT i; for (i = v; i > 0; i = i - 1, ri = ri + ivs, ii = ii + ivs, ro = ro + ovs, io = io + ovs, MAKE_VOLATILE_STRIDE(60, is), MAKE_VOLATILE_STRIDE(60, os)) { E T1r, T1g, T14, T13; { E T5, T2l, Tx, TV, T1z, T1X, T2s, Tr, T24, TT, T2e, T2n, T1Z, T1Q, T1B; E T11, T1H, TW, T2t, Tg, TX, T25, TI, T2h, T2m, T1Y, T1T, T1A; { E T1, T1v, T2, T3, Tu, Tv, TZ, T10; T1 = ri[0]; T1v = ii[0]; T2 = ri[WS(is, 5)]; T3 = ri[WS(is, 10)]; Tu = ii[WS(is, 5)]; Tv = ii[WS(is, 10)]; { E T1k, Tm, TM, TJ, Tl, T2c, T1j, T1m, TP, T1p, Tp, TQ; { E Th, T1h, TK, TL, Tk, Tn, To, T1i; { E Ti, Tj, T1y, T4; Th = ri[WS(is, 6)]; T1y = T3 - T2; T4 = T2 + T3; { E T1w, Tw, Tt, T1x; T1w = Tu + Tv; Tw = Tu - Tv; Ti = ri[WS(is, 11)]; T5 = T1 + T4; Tt = FNMS(KP500000000, T4, T1); T2l = T1v + T1w; T1x = FNMS(KP500000000, T1w, T1v); Tx = FNMS(KP866025403, Tw, Tt); TV = FMA(KP866025403, Tw, Tt); T1z = FMA(KP866025403, T1y, T1x); T1X = FNMS(KP866025403, T1y, T1x); Tj = ri[WS(is, 1)]; } T1h = ii[WS(is, 6)]; TK = ii[WS(is, 11)]; TL = ii[WS(is, 1)]; Tk = Ti + Tj; T1k = Tj - Ti; } Tm = ri[WS(is, 9)]; TM = TK - TL; T1i = TK + TL; TJ = FNMS(KP500000000, Tk, Th); Tl = Th + Tk; Tn = ri[WS(is, 14)]; To = ri[WS(is, 4)]; T2c = T1h + T1i; T1j = FNMS(KP500000000, T1i, T1h); T1m = ii[WS(is, 9)]; TP = ii[WS(is, 14)]; T1p = To - Tn; Tp = Tn + To; TQ = ii[WS(is, 4)]; } { E TN, TS, T1o, T2d; { E TO, T1n, TR, Tq; TN = FNMS(KP866025403, TM, TJ); TZ = FMA(KP866025403, TM, TJ); TO = FNMS(KP500000000, Tp, Tm); Tq = Tm + Tp; T1n = TP + TQ; TR = TP - TQ; T2s = Tl - Tq; Tr = Tl + Tq; T10 = FMA(KP866025403, TR, TO); TS = FNMS(KP866025403, TR, TO); T1o = FNMS(KP500000000, T1n, T1m); T2d = T1m + T1n; } { E T1O, T1l, T1P, T1q; T1O = FNMS(KP866025403, T1k, T1j); T1l = FMA(KP866025403, T1k, T1j); T24 = TN - TS; TT = TN + TS; T1P = FNMS(KP866025403, T1p, T1o); T1q = FMA(KP866025403, T1p, T1o); T2e = T2c - T2d; T2n = T2c + T2d; T1Z = T1O + T1P; T1Q = T1O - T1P; T1r = T1l - T1q; T1B = T1l + T1q; } } } { E T19, Tb, TB, Ty, Ta, T2f, T18, T1b, TE, T1e, Te, TF; { E T6, T16, Tz, TA, T9, T7, T8, Tc, Td, T17; T6 = ri[WS(is, 3)]; T7 = ri[WS(is, 8)]; T11 = TZ + T10; T1H = TZ - T10; T8 = ri[WS(is, 13)]; T16 = ii[WS(is, 3)]; Tz = ii[WS(is, 8)]; TA = ii[WS(is, 13)]; T9 = T7 + T8; T19 = T8 - T7; Tb = ri[WS(is, 12)]; TB = Tz - TA; T17 = Tz + TA; Ty = FNMS(KP500000000, T9, T6); Ta = T6 + T9; Tc = ri[WS(is, 2)]; Td = ri[WS(is, 7)]; T2f = T16 + T17; T18 = FNMS(KP500000000, T17, T16); T1b = ii[WS(is, 12)]; TE = ii[WS(is, 2)]; T1e = Td - Tc; Te = Tc + Td; TF = ii[WS(is, 7)]; } { E TC, TH, T1d, T2g; { E TD, T1c, TG, Tf; TC = FNMS(KP866025403, TB, Ty); TW = FMA(KP866025403, TB, Ty); TD = FNMS(KP500000000, Te, Tb); Tf = Tb + Te; T1c = TE + TF; TG = TE - TF; T2t = Ta - Tf; Tg = Ta + Tf; TX = FMA(KP866025403, TG, TD); TH = FNMS(KP866025403, TG, TD); T1d = FNMS(KP500000000, T1c, T1b); T2g = T1b + T1c; } { E T1R, T1a, T1S, T1f; T1R = FNMS(KP866025403, T19, T18); T1a = FMA(KP866025403, T19, T18); T25 = TC - TH; TI = TC + TH; T1S = FNMS(KP866025403, T1e, T1d); T1f = FMA(KP866025403, T1e, T1d); T2h = T2f - T2g; T2m = T2f + T2g; T1Y = T1R + T1S; T1T = T1R - T1S; T1g = T1a - T1f; T1A = T1a + T1f; } } } } { E TY, T1G, T1M, T1L, T2a, T29, Ts, T22, T21, T20; T2a = Tg - Tr; Ts = Tg + Tr; TY = TW + TX; T1G = TW - TX; T29 = FNMS(KP250000000, Ts, T5); ro[0] = T5 + Ts; { E T2q, T2p, T2o, TU; T2o = T2m + T2n; T2q = T2m - T2n; { E T2k, T2i, T2b, T2j; T2k = FMA(KP618033988, T2e, T2h); T2i = FNMS(KP618033988, T2h, T2e); T2b = FNMS(KP559016994, T2a, T29); T2j = FMA(KP559016994, T2a, T29); ro[WS(os, 3)] = FMA(KP951056516, T2i, T2b); ro[WS(os, 12)] = FNMS(KP951056516, T2i, T2b); ro[WS(os, 6)] = FMA(KP951056516, T2k, T2j); ro[WS(os, 9)] = FNMS(KP951056516, T2k, T2j); T2p = FNMS(KP250000000, T2o, T2l); } io[0] = T2l + T2o; TU = TI + TT; T1M = TI - TT; { E T2r, T2v, T2w, T2u; T2r = FNMS(KP559016994, T2q, T2p); T2v = FMA(KP559016994, T2q, T2p); T2w = FMA(KP618033988, T2s, T2t); T2u = FNMS(KP618033988, T2t, T2s); io[WS(os, 9)] = FMA(KP951056516, T2w, T2v); io[WS(os, 6)] = FNMS(KP951056516, T2w, T2v); io[WS(os, 12)] = FMA(KP951056516, T2u, T2r); io[WS(os, 3)] = FNMS(KP951056516, T2u, T2r); T1L = FNMS(KP250000000, TU, Tx); } ro[WS(os, 5)] = Tx + TU; } T20 = T1Y + T1Z; T22 = T1Y - T1Z; { E T1N, T1V, T1W, T1U; T1N = FNMS(KP559016994, T1M, T1L); T1V = FMA(KP559016994, T1M, T1L); T1W = FMA(KP618033988, T1Q, T1T); T1U = FNMS(KP618033988, T1T, T1Q); ro[WS(os, 11)] = FMA(KP951056516, T1W, T1V); ro[WS(os, 14)] = FNMS(KP951056516, T1W, T1V); ro[WS(os, 8)] = FMA(KP951056516, T1U, T1N); ro[WS(os, 2)] = FNMS(KP951056516, T1U, T1N); T21 = FNMS(KP250000000, T20, T1X); } io[WS(os, 5)] = T1X + T20; { E T1E, T1D, T1C, T12; T1C = T1A + T1B; T1E = T1A - T1B; { E T23, T27, T28, T26; T23 = FNMS(KP559016994, T22, T21); T27 = FMA(KP559016994, T22, T21); T28 = FMA(KP618033988, T24, T25); T26 = FNMS(KP618033988, T25, T24); io[WS(os, 14)] = FMA(KP951056516, T28, T27); io[WS(os, 11)] = FNMS(KP951056516, T28, T27); io[WS(os, 8)] = FNMS(KP951056516, T26, T23); io[WS(os, 2)] = FMA(KP951056516, T26, T23); T1D = FNMS(KP250000000, T1C, T1z); } io[WS(os, 10)] = T1z + T1C; T12 = TY + T11; T14 = TY - T11; { E T1F, T1J, T1K, T1I; T1F = FMA(KP559016994, T1E, T1D); T1J = FNMS(KP559016994, T1E, T1D); T1K = FNMS(KP618033988, T1G, T1H); T1I = FMA(KP618033988, T1H, T1G); io[WS(os, 13)] = FNMS(KP951056516, T1K, T1J); io[WS(os, 7)] = FMA(KP951056516, T1K, T1J); io[WS(os, 4)] = FMA(KP951056516, T1I, T1F); io[WS(os, 1)] = FNMS(KP951056516, T1I, T1F); T13 = FNMS(KP250000000, T12, TV); } ro[WS(os, 10)] = TV + T12; } } } { E T1t, T15, T1s, T1u; T1t = FNMS(KP559016994, T14, T13); T15 = FMA(KP559016994, T14, T13); T1s = FMA(KP618033988, T1r, T1g); T1u = FNMS(KP618033988, T1g, T1r); ro[WS(os, 13)] = FMA(KP951056516, T1u, T1t); ro[WS(os, 7)] = FNMS(KP951056516, T1u, T1t); ro[WS(os, 1)] = FMA(KP951056516, T1s, T15); ro[WS(os, 4)] = FNMS(KP951056516, T1s, T15); } } } } static const kdft_desc desc = { 15, "n1_15", {72, 0, 84, 0}, &GENUS, 0, 0, 0, 0 }; void X(codelet_n1_15) (planner *p) { X(kdft_register) (p, n1_15, &desc); } #else /* HAVE_FMA */ /* Generated by: ../../../genfft/gen_notw.native -compact -variables 4 -pipeline-latency 4 -n 15 -name n1_15 -include n.h */ /* * This function contains 156 FP additions, 56 FP multiplications, * (or, 128 additions, 28 multiplications, 28 fused multiply/add), * 69 stack variables, 6 constants, and 60 memory accesses */ #include "n.h" static void n1_15(const R *ri, const R *ii, R *ro, R *io, stride is, stride os, INT v, INT ivs, INT ovs) { DK(KP587785252, +0.587785252292473129168705954639072768597652438); DK(KP951056516, +0.951056516295153572116439333379382143405698634); DK(KP250000000, +0.250000000000000000000000000000000000000000000); DK(KP559016994, +0.559016994374947424102293417182819058860154590); DK(KP500000000, +0.500000000000000000000000000000000000000000000); DK(KP866025403, +0.866025403784438646763723170752936183471402627); { INT i; for (i = v; i > 0; i = i - 1, ri = ri + ivs, ii = ii + ivs, ro = ro + ovs, io = io + ovs, MAKE_VOLATILE_STRIDE(60, is), MAKE_VOLATILE_STRIDE(60, os)) { E T5, T2l, Tx, TV, T1C, T20, Tl, Tq, Tr, TN, TS, TT, T2c, T2d, T2n; E T1O, T1P, T22, T1l, T1q, T1w, TZ, T10, T11, Ta, Tf, Tg, TC, TH, TI; E T2f, T2g, T2m, T1R, T1S, T21, T1a, T1f, T1v, TW, TX, TY; { E T1, T1z, T4, T1y, Tw, T1A, Tt, T1B; T1 = ri[0]; T1z = ii[0]; { E T2, T3, Tu, Tv; T2 = ri[WS(is, 5)]; T3 = ri[WS(is, 10)]; T4 = T2 + T3; T1y = KP866025403 * (T3 - T2); Tu = ii[WS(is, 5)]; Tv = ii[WS(is, 10)]; Tw = KP866025403 * (Tu - Tv); T1A = Tu + Tv; } T5 = T1 + T4; T2l = T1z + T1A; Tt = FNMS(KP500000000, T4, T1); Tx = Tt - Tw; TV = Tt + Tw; T1B = FNMS(KP500000000, T1A, T1z); T1C = T1y + T1B; T20 = T1B - T1y; } { E Th, Tk, TJ, T1h, T1i, T1j, TM, T1k, Tm, Tp, TO, T1m, T1n, T1o, TR; E T1p; { E Ti, Tj, TK, TL; Th = ri[WS(is, 6)]; Ti = ri[WS(is, 11)]; Tj = ri[WS(is, 1)]; Tk = Ti + Tj; TJ = FNMS(KP500000000, Tk, Th); T1h = KP866025403 * (Tj - Ti); T1i = ii[WS(is, 6)]; TK = ii[WS(is, 11)]; TL = ii[WS(is, 1)]; T1j = TK + TL; TM = KP866025403 * (TK - TL); T1k = FNMS(KP500000000, T1j, T1i); } { E Tn, To, TP, TQ; Tm = ri[WS(is, 9)]; Tn = ri[WS(is, 14)]; To = ri[WS(is, 4)]; Tp = Tn + To; TO = FNMS(KP500000000, Tp, Tm); T1m = KP866025403 * (To - Tn); T1n = ii[WS(is, 9)]; TP = ii[WS(is, 14)]; TQ = ii[WS(is, 4)]; T1o = TP + TQ; TR = KP866025403 * (TP - TQ); T1p = FNMS(KP500000000, T1o, T1n); } Tl = Th + Tk; Tq = Tm + Tp; Tr = Tl + Tq; TN = TJ - TM; TS = TO - TR; TT = TN + TS; T2c = T1i + T1j; T2d = T1n + T1o; T2n = T2c + T2d; T1O = T1k - T1h; T1P = T1p - T1m; T22 = T1O + T1P; T1l = T1h + T1k; T1q = T1m + T1p; T1w = T1l + T1q; TZ = TJ + TM; T10 = TO + TR; T11 = TZ + T10; } { E T6, T9, Ty, T16, T17, T18, TB, T19, Tb, Te, TD, T1b, T1c, T1d, TG; E T1e; { E T7, T8, Tz, TA; T6 = ri[WS(is, 3)]; T7 = ri[WS(is, 8)]; T8 = ri[WS(is, 13)]; T9 = T7 + T8; Ty = FNMS(KP500000000, T9, T6); T16 = KP866025403 * (T8 - T7); T17 = ii[WS(is, 3)]; Tz = ii[WS(is, 8)]; TA = ii[WS(is, 13)]; T18 = Tz + TA; TB = KP866025403 * (Tz - TA); T19 = FNMS(KP500000000, T18, T17); } { E Tc, Td, TE, TF; Tb = ri[WS(is, 12)]; Tc = ri[WS(is, 2)]; Td = ri[WS(is, 7)]; Te = Tc + Td; TD = FNMS(KP500000000, Te, Tb); T1b = KP866025403 * (Td - Tc); T1c = ii[WS(is, 12)]; TE = ii[WS(is, 2)]; TF = ii[WS(is, 7)]; T1d = TE + TF; TG = KP866025403 * (TE - TF); T1e = FNMS(KP500000000, T1d, T1c); } Ta = T6 + T9; Tf = Tb + Te; Tg = Ta + Tf; TC = Ty - TB; TH = TD - TG; TI = TC + TH; T2f = T17 + T18; T2g = T1c + T1d; T2m = T2f + T2g; T1R = T19 - T16; T1S = T1e - T1b; T21 = T1R + T1S; T1a = T16 + T19; T1f = T1b + T1e; T1v = T1a + T1f; TW = Ty + TB; TX = TD + TG; TY = TW + TX; } { E T2a, Ts, T29, T2i, T2k, T2e, T2h, T2j, T2b; T2a = KP559016994 * (Tg - Tr); Ts = Tg + Tr; T29 = FNMS(KP250000000, Ts, T5); T2e = T2c - T2d; T2h = T2f - T2g; T2i = FNMS(KP587785252, T2h, KP951056516 * T2e); T2k = FMA(KP951056516, T2h, KP587785252 * T2e); ro[0] = T5 + Ts; T2j = T2a + T29; ro[WS(os, 9)] = T2j - T2k; ro[WS(os, 6)] = T2j + T2k; T2b = T29 - T2a; ro[WS(os, 12)] = T2b - T2i; ro[WS(os, 3)] = T2b + T2i; } { E T2q, T2o, T2p, T2u, T2w, T2s, T2t, T2v, T2r; T2q = KP559016994 * (T2m - T2n); T2o = T2m + T2n; T2p = FNMS(KP250000000, T2o, T2l); T2s = Tl - Tq; T2t = Ta - Tf; T2u = FNMS(KP587785252, T2t, KP951056516 * T2s); T2w = FMA(KP951056516, T2t, KP587785252 * T2s); io[0] = T2l + T2o; T2v = T2q + T2p; io[WS(os, 6)] = T2v - T2w; io[WS(os, 9)] = T2w + T2v; T2r = T2p - T2q; io[WS(os, 3)] = T2r - T2u; io[WS(os, 12)] = T2u + T2r; } { E T1M, TU, T1L, T1U, T1W, T1Q, T1T, T1V, T1N; T1M = KP559016994 * (TI - TT); TU = TI + TT; T1L = FNMS(KP250000000, TU, Tx); T1Q = T1O - T1P; T1T = T1R - T1S; T1U = FNMS(KP587785252, T1T, KP951056516 * T1Q); T1W = FMA(KP951056516, T1T, KP587785252 * T1Q); ro[WS(os, 5)] = Tx + TU; T1V = T1M + T1L; ro[WS(os, 14)] = T1V - T1W; ro[WS(os, 11)] = T1V + T1W; T1N = T1L - T1M; ro[WS(os, 2)] = T1N - T1U; ro[WS(os, 8)] = T1N + T1U; } { E T25, T23, T24, T1Z, T28, T1X, T1Y, T27, T26; T25 = KP559016994 * (T21 - T22); T23 = T21 + T22; T24 = FNMS(KP250000000, T23, T20); T1X = TN - TS; T1Y = TC - TH; T1Z = FNMS(KP587785252, T1Y, KP951056516 * T1X); T28 = FMA(KP951056516, T1Y, KP587785252 * T1X); io[WS(os, 5)] = T20 + T23; T27 = T25 + T24; io[WS(os, 11)] = T27 - T28; io[WS(os, 14)] = T28 + T27; T26 = T24 - T25; io[WS(os, 2)] = T1Z + T26; io[WS(os, 8)] = T26 - T1Z; } { E T1x, T1D, T1E, T1I, T1J, T1G, T1H, T1K, T1F; T1x = KP559016994 * (T1v - T1w); T1D = T1v + T1w; T1E = FNMS(KP250000000, T1D, T1C); T1G = TW - TX; T1H = TZ - T10; T1I = FMA(KP951056516, T1G, KP587785252 * T1H); T1J = FNMS(KP587785252, T1G, KP951056516 * T1H); io[WS(os, 10)] = T1C + T1D; T1K = T1E - T1x; io[WS(os, 7)] = T1J + T1K; io[WS(os, 13)] = T1K - T1J; T1F = T1x + T1E; io[WS(os, 1)] = T1F - T1I; io[WS(os, 4)] = T1I + T1F; } { E T13, T12, T14, T1s, T1u, T1g, T1r, T1t, T15; T13 = KP559016994 * (TY - T11); T12 = TY + T11; T14 = FNMS(KP250000000, T12, TV); T1g = T1a - T1f; T1r = T1l - T1q; T1s = FMA(KP951056516, T1g, KP587785252 * T1r); T1u = FNMS(KP587785252, T1g, KP951056516 * T1r); ro[WS(os, 10)] = TV + T12; T1t = T14 - T13; ro[WS(os, 7)] = T1t - T1u; ro[WS(os, 13)] = T1t + T1u; T15 = T13 + T14; ro[WS(os, 4)] = T15 - T1s; ro[WS(os, 1)] = T15 + T1s; } } } } static const kdft_desc desc = { 15, "n1_15", {128, 28, 28, 0}, &GENUS, 0, 0, 0, 0 }; void X(codelet_n1_15) (planner *p) { X(kdft_register) (p, n1_15, &desc); } #endif /* HAVE_FMA */ fftw-3.3.3/dft/scalar/codelets/n1_16.c0000644000175000017500000003541712054410242014227 00000000000000/* * Copyright (c) 2003, 2007-11 Matteo Frigo * Copyright (c) 2003, 2007-11 Massachusetts Institute of Technology * * This program is free software; you can redistribute it and/or modify * it under the terms of the GNU General Public License as published by * the Free Software Foundation; either version 2 of the License, or * (at your option) any later version. * * This program is distributed in the hope that it will be useful, * but WITHOUT ANY WARRANTY; without even the implied warranty of * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the * GNU General Public License for more details. * * You should have received a copy of the GNU General Public License * along with this program; if not, write to the Free Software * Foundation, Inc., 51 Franklin Street, Fifth Floor, Boston, MA 02110-1301 USA * */ /* This file was automatically generated --- DO NOT EDIT */ /* Generated on Sun Nov 25 07:35:44 EST 2012 */ #include "codelet-dft.h" #ifdef HAVE_FMA /* Generated by: ../../../genfft/gen_notw.native -fma -reorder-insns -schedule-for-pipeline -compact -variables 4 -pipeline-latency 4 -n 16 -name n1_16 -include n.h */ /* * This function contains 144 FP additions, 40 FP multiplications, * (or, 104 additions, 0 multiplications, 40 fused multiply/add), * 82 stack variables, 3 constants, and 64 memory accesses */ #include "n.h" static void n1_16(const R *ri, const R *ii, R *ro, R *io, stride is, stride os, INT v, INT ivs, INT ovs) { DK(KP923879532, +0.923879532511286756128183189396788286822416626); DK(KP414213562, +0.414213562373095048801688724209698078569671875); DK(KP707106781, +0.707106781186547524400844362104849039284835938); { INT i; for (i = v; i > 0; i = i - 1, ri = ri + ivs, ii = ii + ivs, ro = ro + ovs, io = io + ovs, MAKE_VOLATILE_STRIDE(64, is), MAKE_VOLATILE_STRIDE(64, os)) { E T1z, T1L, T1M, T1N, T1P, T1J, T1K, T1G, T1O, T1Q; { E T1l, T1H, T1R, T7, T1x, TN, TC, T25, T1E, T1b, T1Z, Tt, T2h, T22, T1D; E T1g, T1n, TQ, Te, T26, TT, T1m, TJ, T1S, Tj, T11, Ti, T1V, TZ, Tk; E T12, T13; { E Tq, T1c, Tp, T20, T1a, Tr, T1d, T1e; { E T4, TL, T3, T1k, Ty, T5, Tz, TA; { E T1, T2, Tw, Tx; T1 = ri[0]; T2 = ri[WS(is, 8)]; Tw = ii[0]; Tx = ii[WS(is, 8)]; T4 = ri[WS(is, 4)]; TL = T1 - T2; T3 = T1 + T2; T1k = Tw - Tx; Ty = Tw + Tx; T5 = ri[WS(is, 12)]; Tz = ii[WS(is, 4)]; TA = ii[WS(is, 12)]; } { E Tn, To, T18, T19; Tn = ri[WS(is, 15)]; { E T1j, T6, TM, TB; T1j = T4 - T5; T6 = T4 + T5; TM = Tz - TA; TB = Tz + TA; T1l = T1j + T1k; T1H = T1k - T1j; T1R = T3 - T6; T7 = T3 + T6; T1x = TL + TM; TN = TL - TM; TC = Ty + TB; T25 = Ty - TB; To = ri[WS(is, 7)]; } T18 = ii[WS(is, 15)]; T19 = ii[WS(is, 7)]; Tq = ri[WS(is, 3)]; T1c = Tn - To; Tp = Tn + To; T20 = T18 + T19; T1a = T18 - T19; Tr = ri[WS(is, 11)]; T1d = ii[WS(is, 3)]; T1e = ii[WS(is, 11)]; } } { E Tb, TP, Ta, TO, TF, Tc, TG, TH; { E T8, T9, TD, TE; T8 = ri[WS(is, 2)]; { E T17, Ts, T21, T1f; T17 = Tq - Tr; Ts = Tq + Tr; T21 = T1d + T1e; T1f = T1d - T1e; T1E = T1a - T17; T1b = T17 + T1a; T1Z = Tp - Ts; Tt = Tp + Ts; T2h = T20 + T21; T22 = T20 - T21; T1D = T1c + T1f; T1g = T1c - T1f; T9 = ri[WS(is, 10)]; } TD = ii[WS(is, 2)]; TE = ii[WS(is, 10)]; Tb = ri[WS(is, 14)]; TP = T8 - T9; Ta = T8 + T9; TO = TD - TE; TF = TD + TE; Tc = ri[WS(is, 6)]; TG = ii[WS(is, 14)]; TH = ii[WS(is, 6)]; } { E TR, Td, TS, TI; T1n = TP + TO; TQ = TO - TP; TR = Tb - Tc; Td = Tb + Tc; TS = TG - TH; TI = TG + TH; Te = Ta + Td; T26 = Td - Ta; TT = TR + TS; T1m = TR - TS; TJ = TF + TI; T1S = TF - TI; } } { E Tg, Th, TX, TY; Tg = ri[WS(is, 1)]; Th = ri[WS(is, 9)]; TX = ii[WS(is, 1)]; TY = ii[WS(is, 9)]; Tj = ri[WS(is, 5)]; T11 = Tg - Th; Ti = Tg + Th; T1V = TX + TY; TZ = TX - TY; Tk = ri[WS(is, 13)]; T12 = ii[WS(is, 5)]; T13 = ii[WS(is, 13)]; } } { E T2f, T1B, T10, T1U, T1X, T1A, T15, Tv, TK, T2i; { E Tf, Tu, T2j, T2k, T2g; T2f = T7 - Te; Tf = T7 + Te; { E TW, Tl, T1W, T14, Tm; TW = Tj - Tk; Tl = Tj + Tk; T1W = T12 + T13; T14 = T12 - T13; T1B = TZ - TW; T10 = TW + TZ; T1U = Ti - Tl; Tm = Ti + Tl; T2g = T1V + T1W; T1X = T1V - T1W; T1A = T11 + T14; T15 = T11 - T14; Tu = Tm + Tt; Tv = Tt - Tm; } TK = TC - TJ; T2j = TC + TJ; T2k = T2g + T2h; T2i = T2g - T2h; ro[0] = Tf + Tu; ro[WS(os, 8)] = Tf - Tu; io[0] = T2j + T2k; io[WS(os, 8)] = T2j - T2k; } { E T29, T1T, T27, T2d, T2a, T2b, T28, T24, T1Y, T23; T29 = T1R - T1S; T1T = T1R + T1S; io[WS(os, 12)] = TK - Tv; io[WS(os, 4)] = Tv + TK; ro[WS(os, 4)] = T2f + T2i; ro[WS(os, 12)] = T2f - T2i; T27 = T25 - T26; T2d = T26 + T25; T2a = T1X - T1U; T1Y = T1U + T1X; T23 = T1Z - T22; T2b = T1Z + T22; T28 = T23 - T1Y; T24 = T1Y + T23; { E T1I, TV, T1v, T1y, T1t, T1s, T1r, T1p, T1q, T1i; { E T1o, T2e, T2c, TU, T16, T1h; T1I = TQ + TT; TU = TQ - TT; io[WS(os, 14)] = FNMS(KP707106781, T28, T27); io[WS(os, 6)] = FMA(KP707106781, T28, T27); ro[WS(os, 2)] = FMA(KP707106781, T24, T1T); ro[WS(os, 10)] = FNMS(KP707106781, T24, T1T); T2e = T2a + T2b; T2c = T2a - T2b; TV = FMA(KP707106781, TU, TN); T1v = FNMS(KP707106781, TU, TN); io[WS(os, 10)] = FNMS(KP707106781, T2e, T2d); io[WS(os, 2)] = FMA(KP707106781, T2e, T2d); ro[WS(os, 6)] = FMA(KP707106781, T2c, T29); ro[WS(os, 14)] = FNMS(KP707106781, T2c, T29); T1o = T1m - T1n; T1y = T1n + T1m; T1t = FNMS(KP414213562, T10, T15); T16 = FMA(KP414213562, T15, T10); T1h = FNMS(KP414213562, T1g, T1b); T1s = FMA(KP414213562, T1b, T1g); T1r = FMA(KP707106781, T1o, T1l); T1p = FNMS(KP707106781, T1o, T1l); T1q = T16 + T1h; T1i = T16 - T1h; } { E T1w, T1u, T1C, T1F; io[WS(os, 15)] = FMA(KP923879532, T1q, T1p); io[WS(os, 7)] = FNMS(KP923879532, T1q, T1p); ro[WS(os, 3)] = FMA(KP923879532, T1i, TV); ro[WS(os, 11)] = FNMS(KP923879532, T1i, TV); T1w = T1t + T1s; T1u = T1s - T1t; T1z = FMA(KP707106781, T1y, T1x); T1L = FNMS(KP707106781, T1y, T1x); ro[WS(os, 15)] = FMA(KP923879532, T1w, T1v); ro[WS(os, 7)] = FNMS(KP923879532, T1w, T1v); io[WS(os, 3)] = FMA(KP923879532, T1u, T1r); io[WS(os, 11)] = FNMS(KP923879532, T1u, T1r); T1M = FNMS(KP414213562, T1A, T1B); T1C = FMA(KP414213562, T1B, T1A); T1F = FNMS(KP414213562, T1E, T1D); T1N = FMA(KP414213562, T1D, T1E); T1P = FMA(KP707106781, T1I, T1H); T1J = FNMS(KP707106781, T1I, T1H); T1K = T1F - T1C; T1G = T1C + T1F; } } } } } io[WS(os, 5)] = FMA(KP923879532, T1K, T1J); io[WS(os, 13)] = FNMS(KP923879532, T1K, T1J); ro[WS(os, 1)] = FMA(KP923879532, T1G, T1z); ro[WS(os, 9)] = FNMS(KP923879532, T1G, T1z); T1O = T1M - T1N; T1Q = T1M + T1N; io[WS(os, 1)] = FMA(KP923879532, T1Q, T1P); io[WS(os, 9)] = FNMS(KP923879532, T1Q, T1P); ro[WS(os, 5)] = FMA(KP923879532, T1O, T1L); ro[WS(os, 13)] = FNMS(KP923879532, T1O, T1L); } } } static const kdft_desc desc = { 16, "n1_16", {104, 0, 40, 0}, &GENUS, 0, 0, 0, 0 }; void X(codelet_n1_16) (planner *p) { X(kdft_register) (p, n1_16, &desc); } #else /* HAVE_FMA */ /* Generated by: ../../../genfft/gen_notw.native -compact -variables 4 -pipeline-latency 4 -n 16 -name n1_16 -include n.h */ /* * This function contains 144 FP additions, 24 FP multiplications, * (or, 136 additions, 16 multiplications, 8 fused multiply/add), * 50 stack variables, 3 constants, and 64 memory accesses */ #include "n.h" static void n1_16(const R *ri, const R *ii, R *ro, R *io, stride is, stride os, INT v, INT ivs, INT ovs) { DK(KP382683432, +0.382683432365089771728459984030398866761344562); DK(KP923879532, +0.923879532511286756128183189396788286822416626); DK(KP707106781, +0.707106781186547524400844362104849039284835938); { INT i; for (i = v; i > 0; i = i - 1, ri = ri + ivs, ii = ii + ivs, ro = ro + ovs, io = io + ovs, MAKE_VOLATILE_STRIDE(64, is), MAKE_VOLATILE_STRIDE(64, os)) { E T7, T1R, T25, TC, TN, T1x, T1H, T1l, Tt, T22, T2h, T1b, T1g, T1E, T1Z; E T1D, Te, T1S, T26, TJ, TQ, T1m, T1n, TT, Tm, T1X, T2g, T10, T15, T1B; E T1U, T1A; { E T3, TL, Ty, T1k, T6, T1j, TB, TM; { E T1, T2, Tw, Tx; T1 = ri[0]; T2 = ri[WS(is, 8)]; T3 = T1 + T2; TL = T1 - T2; Tw = ii[0]; Tx = ii[WS(is, 8)]; Ty = Tw + Tx; T1k = Tw - Tx; } { E T4, T5, Tz, TA; T4 = ri[WS(is, 4)]; T5 = ri[WS(is, 12)]; T6 = T4 + T5; T1j = T4 - T5; Tz = ii[WS(is, 4)]; TA = ii[WS(is, 12)]; TB = Tz + TA; TM = Tz - TA; } T7 = T3 + T6; T1R = T3 - T6; T25 = Ty - TB; TC = Ty + TB; TN = TL - TM; T1x = TL + TM; T1H = T1k - T1j; T1l = T1j + T1k; } { E Tp, T17, T1f, T20, Ts, T1c, T1a, T21; { E Tn, To, T1d, T1e; Tn = ri[WS(is, 15)]; To = ri[WS(is, 7)]; Tp = Tn + To; T17 = Tn - To; T1d = ii[WS(is, 15)]; T1e = ii[WS(is, 7)]; T1f = T1d - T1e; T20 = T1d + T1e; } { E Tq, Tr, T18, T19; Tq = ri[WS(is, 3)]; Tr = ri[WS(is, 11)]; Ts = Tq + Tr; T1c = Tq - Tr; T18 = ii[WS(is, 3)]; T19 = ii[WS(is, 11)]; T1a = T18 - T19; T21 = T18 + T19; } Tt = Tp + Ts; T22 = T20 - T21; T2h = T20 + T21; T1b = T17 - T1a; T1g = T1c + T1f; T1E = T1f - T1c; T1Z = Tp - Ts; T1D = T17 + T1a; } { E Ta, TP, TF, TO, Td, TR, TI, TS; { E T8, T9, TD, TE; T8 = ri[WS(is, 2)]; T9 = ri[WS(is, 10)]; Ta = T8 + T9; TP = T8 - T9; TD = ii[WS(is, 2)]; TE = ii[WS(is, 10)]; TF = TD + TE; TO = TD - TE; } { E Tb, Tc, TG, TH; Tb = ri[WS(is, 14)]; Tc = ri[WS(is, 6)]; Td = Tb + Tc; TR = Tb - Tc; TG = ii[WS(is, 14)]; TH = ii[WS(is, 6)]; TI = TG + TH; TS = TG - TH; } Te = Ta + Td; T1S = TF - TI; T26 = Td - Ta; TJ = TF + TI; TQ = TO - TP; T1m = TR - TS; T1n = TP + TO; TT = TR + TS; } { E Ti, T11, TZ, T1V, Tl, TW, T14, T1W; { E Tg, Th, TX, TY; Tg = ri[WS(is, 1)]; Th = ri[WS(is, 9)]; Ti = Tg + Th; T11 = Tg - Th; TX = ii[WS(is, 1)]; TY = ii[WS(is, 9)]; TZ = TX - TY; T1V = TX + TY; } { E Tj, Tk, T12, T13; Tj = ri[WS(is, 5)]; Tk = ri[WS(is, 13)]; Tl = Tj + Tk; TW = Tj - Tk; T12 = ii[WS(is, 5)]; T13 = ii[WS(is, 13)]; T14 = T12 - T13; T1W = T12 + T13; } Tm = Ti + Tl; T1X = T1V - T1W; T2g = T1V + T1W; T10 = TW + TZ; T15 = T11 - T14; T1B = T11 + T14; T1U = Ti - Tl; T1A = TZ - TW; } { E Tf, Tu, T2j, T2k; Tf = T7 + Te; Tu = Tm + Tt; ro[WS(os, 8)] = Tf - Tu; ro[0] = Tf + Tu; T2j = TC + TJ; T2k = T2g + T2h; io[WS(os, 8)] = T2j - T2k; io[0] = T2j + T2k; } { E Tv, TK, T2f, T2i; Tv = Tt - Tm; TK = TC - TJ; io[WS(os, 4)] = Tv + TK; io[WS(os, 12)] = TK - Tv; T2f = T7 - Te; T2i = T2g - T2h; ro[WS(os, 12)] = T2f - T2i; ro[WS(os, 4)] = T2f + T2i; } { E T1T, T27, T24, T28, T1Y, T23; T1T = T1R + T1S; T27 = T25 - T26; T1Y = T1U + T1X; T23 = T1Z - T22; T24 = KP707106781 * (T1Y + T23); T28 = KP707106781 * (T23 - T1Y); ro[WS(os, 10)] = T1T - T24; io[WS(os, 6)] = T27 + T28; ro[WS(os, 2)] = T1T + T24; io[WS(os, 14)] = T27 - T28; } { E T29, T2d, T2c, T2e, T2a, T2b; T29 = T1R - T1S; T2d = T26 + T25; T2a = T1X - T1U; T2b = T1Z + T22; T2c = KP707106781 * (T2a - T2b); T2e = KP707106781 * (T2a + T2b); ro[WS(os, 14)] = T29 - T2c; io[WS(os, 2)] = T2d + T2e; ro[WS(os, 6)] = T29 + T2c; io[WS(os, 10)] = T2d - T2e; } { E TV, T1r, T1p, T1v, T1i, T1q, T1u, T1w, TU, T1o; TU = KP707106781 * (TQ - TT); TV = TN + TU; T1r = TN - TU; T1o = KP707106781 * (T1m - T1n); T1p = T1l - T1o; T1v = T1l + T1o; { E T16, T1h, T1s, T1t; T16 = FMA(KP923879532, T10, KP382683432 * T15); T1h = FNMS(KP923879532, T1g, KP382683432 * T1b); T1i = T16 + T1h; T1q = T1h - T16; T1s = FNMS(KP923879532, T15, KP382683432 * T10); T1t = FMA(KP382683432, T1g, KP923879532 * T1b); T1u = T1s - T1t; T1w = T1s + T1t; } ro[WS(os, 11)] = TV - T1i; io[WS(os, 11)] = T1v - T1w; ro[WS(os, 3)] = TV + T1i; io[WS(os, 3)] = T1v + T1w; io[WS(os, 15)] = T1p - T1q; ro[WS(os, 15)] = T1r - T1u; io[WS(os, 7)] = T1p + T1q; ro[WS(os, 7)] = T1r + T1u; } { E T1z, T1L, T1J, T1P, T1G, T1K, T1O, T1Q, T1y, T1I; T1y = KP707106781 * (T1n + T1m); T1z = T1x + T1y; T1L = T1x - T1y; T1I = KP707106781 * (TQ + TT); T1J = T1H - T1I; T1P = T1H + T1I; { E T1C, T1F, T1M, T1N; T1C = FMA(KP382683432, T1A, KP923879532 * T1B); T1F = FNMS(KP382683432, T1E, KP923879532 * T1D); T1G = T1C + T1F; T1K = T1F - T1C; T1M = FNMS(KP382683432, T1B, KP923879532 * T1A); T1N = FMA(KP923879532, T1E, KP382683432 * T1D); T1O = T1M - T1N; T1Q = T1M + T1N; } ro[WS(os, 9)] = T1z - T1G; io[WS(os, 9)] = T1P - T1Q; ro[WS(os, 1)] = T1z + T1G; io[WS(os, 1)] = T1P + T1Q; io[WS(os, 13)] = T1J - T1K; ro[WS(os, 13)] = T1L - T1O; io[WS(os, 5)] = T1J + T1K; ro[WS(os, 5)] = T1L + T1O; } } } } static const kdft_desc desc = { 16, "n1_16", {136, 16, 8, 0}, &GENUS, 0, 0, 0, 0 }; void X(codelet_n1_16) (planner *p) { X(kdft_register) (p, n1_16, &desc); } #endif /* HAVE_FMA */ fftw-3.3.3/dft/scalar/codelets/n1_32.c0000644000175000017500000010514212054410250014215 00000000000000/* * Copyright (c) 2003, 2007-11 Matteo Frigo * Copyright (c) 2003, 2007-11 Massachusetts Institute of Technology * * This program is free software; you can redistribute it and/or modify * it under the terms of the GNU General Public License as published by * the Free Software Foundation; either version 2 of the License, or * (at your option) any later version. * * This program is distributed in the hope that it will be useful, * but WITHOUT ANY WARRANTY; without even the implied warranty of * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the * GNU General Public License for more details. * * You should have received a copy of the GNU General Public License * along with this program; if not, write to the Free Software * Foundation, Inc., 51 Franklin Street, Fifth Floor, Boston, MA 02110-1301 USA * */ /* This file was automatically generated --- DO NOT EDIT */ /* Generated on Sun Nov 25 07:35:45 EST 2012 */ #include "codelet-dft.h" #ifdef HAVE_FMA /* Generated by: ../../../genfft/gen_notw.native -fma -reorder-insns -schedule-for-pipeline -compact -variables 4 -pipeline-latency 4 -n 32 -name n1_32 -include n.h */ /* * This function contains 372 FP additions, 136 FP multiplications, * (or, 236 additions, 0 multiplications, 136 fused multiply/add), * 136 stack variables, 7 constants, and 128 memory accesses */ #include "n.h" static void n1_32(const R *ri, const R *ii, R *ro, R *io, stride is, stride os, INT v, INT ivs, INT ovs) { DK(KP980785280, +0.980785280403230449126182236134239036973933731); DK(KP198912367, +0.198912367379658006911597622644676228597850501); DK(KP831469612, +0.831469612302545237078788377617905756738560812); DK(KP668178637, +0.668178637919298919997757686523080761552472251); DK(KP923879532, +0.923879532511286756128183189396788286822416626); DK(KP707106781, +0.707106781186547524400844362104849039284835938); DK(KP414213562, +0.414213562373095048801688724209698078569671875); { INT i; for (i = v; i > 0; i = i - 1, ri = ri + ivs, ii = ii + ivs, ro = ro + ovs, io = io + ovs, MAKE_VOLATILE_STRIDE(128, is), MAKE_VOLATILE_STRIDE(128, os)) { E T3g, T3f, T3n, T3b, T3r, T3l, T3o, T3e, T3h, T3p; { E T2T, T3T, T4r, T7, T3t, T1z, T18, T4Z, Te, T50, T4s, T1f, T2W, T3u, T3U; E T1G, Tm, T1n, T3X, T3y, T2Z, T1O, T53, T4w, Tt, T1u, T3W, T3B, T2Y, T1V; E T52, T4z, T3O, T2t, T3L, T2K, T5F, TZ, T5I, T5X, T4R, T5k, T3M, T2E, T5j; E T4W, T3P, T2N, T3H, T22, T3E, T2j, T4H, T4K, T5A, TK, T5D, T5W, T2k, T2l; E T4G, T5h, T3F, T2d; { E Tj, T1L, Ti, T1I, T1j, Tk, T1k, T1l; { E T4, T1x, T3, T2R, T14, T5, T15, T16, T1C, T1F; { E T1, T2, T12, T13; T1 = ri[0]; T2 = ri[WS(is, 16)]; T12 = ii[0]; T13 = ii[WS(is, 16)]; T4 = ri[WS(is, 8)]; T1x = T1 - T2; T3 = T1 + T2; T2R = T12 - T13; T14 = T12 + T13; T5 = ri[WS(is, 24)]; T15 = ii[WS(is, 8)]; T16 = ii[WS(is, 24)]; } { E Tb, T1A, Ta, T1B, T1b, Tc, T1c, T1d; { E T8, T9, T19, T1a; T8 = ri[WS(is, 4)]; { E T2S, T6, T1y, T17; T2S = T4 - T5; T6 = T4 + T5; T1y = T15 - T16; T17 = T15 + T16; T2T = T2R - T2S; T3T = T2S + T2R; T4r = T3 - T6; T7 = T3 + T6; T3t = T1x - T1y; T1z = T1x + T1y; T18 = T14 + T17; T4Z = T14 - T17; T9 = ri[WS(is, 20)]; } T19 = ii[WS(is, 4)]; T1a = ii[WS(is, 20)]; Tb = ri[WS(is, 28)]; T1A = T8 - T9; Ta = T8 + T9; T1B = T19 - T1a; T1b = T19 + T1a; Tc = ri[WS(is, 12)]; T1c = ii[WS(is, 28)]; T1d = ii[WS(is, 12)]; } { E T2U, T1D, Td, T1E, T1e, T2V; T1C = T1A + T1B; T2U = T1B - T1A; T1D = Tb - Tc; Td = Tb + Tc; T1E = T1c - T1d; T1e = T1c + T1d; Te = Ta + Td; T50 = Td - Ta; T1F = T1D - T1E; T2V = T1D + T1E; T4s = T1b - T1e; T1f = T1b + T1e; T2W = T2U + T2V; T3u = T2U - T2V; } } { E Tg, Th, T1h, T1i; Tg = ri[WS(is, 2)]; T3U = T1F - T1C; T1G = T1C + T1F; Th = ri[WS(is, 18)]; T1h = ii[WS(is, 2)]; T1i = ii[WS(is, 18)]; Tj = ri[WS(is, 10)]; T1L = Tg - Th; Ti = Tg + Th; T1I = T1h - T1i; T1j = T1h + T1i; Tk = ri[WS(is, 26)]; T1k = ii[WS(is, 10)]; T1l = ii[WS(is, 26)]; } } { E Tq, T1S, Tp, T1P, T1q, Tr, T1r, T1s; { E Tn, To, T1o, T1p, T1J, Tl; Tn = ri[WS(is, 30)]; T1J = Tj - Tk; Tl = Tj + Tk; { E T1M, T1m, T3w, T1K; T1M = T1k - T1l; T1m = T1k + T1l; T3w = T1J + T1I; T1K = T1I - T1J; { E T4v, T3x, T1N, T4u; T4v = Ti - Tl; Tm = Ti + Tl; T3x = T1L - T1M; T1N = T1L + T1M; T4u = T1j - T1m; T1n = T1j + T1m; T3X = FNMS(KP414213562, T3w, T3x); T3y = FMA(KP414213562, T3x, T3w); T2Z = FMA(KP414213562, T1K, T1N); T1O = FNMS(KP414213562, T1N, T1K); T53 = T4v + T4u; T4w = T4u - T4v; To = ri[WS(is, 14)]; } } T1o = ii[WS(is, 30)]; T1p = ii[WS(is, 14)]; Tq = ri[WS(is, 6)]; T1S = Tn - To; Tp = Tn + To; T1P = T1o - T1p; T1q = T1o + T1p; Tr = ri[WS(is, 22)]; T1r = ii[WS(is, 6)]; T1s = ii[WS(is, 22)]; } { E T4S, T4V, T2L, T2M; { E T2G, TN, T4N, T2r, T2s, TQ, T4O, T2J, TV, T2x, TU, T4T, T2w, TW, T2A; E T2B; { E TO, TP, T2H, T2I; { E TL, TM, T2p, T2q, T1Q, Ts; TL = ri[WS(is, 31)]; T1Q = Tq - Tr; Ts = Tq + Tr; { E T1T, T1t, T3z, T1R; T1T = T1r - T1s; T1t = T1r + T1s; T3z = T1Q + T1P; T1R = T1P - T1Q; { E T4x, T3A, T1U, T4y; T4x = Tp - Ts; Tt = Tp + Ts; T3A = T1S - T1T; T1U = T1S + T1T; T4y = T1q - T1t; T1u = T1q + T1t; T3W = FMA(KP414213562, T3z, T3A); T3B = FNMS(KP414213562, T3A, T3z); T2Y = FNMS(KP414213562, T1R, T1U); T1V = FMA(KP414213562, T1U, T1R); T52 = T4x - T4y; T4z = T4x + T4y; TM = ri[WS(is, 15)]; } } T2p = ii[WS(is, 31)]; T2q = ii[WS(is, 15)]; TO = ri[WS(is, 7)]; T2G = TL - TM; TN = TL + TM; T4N = T2p + T2q; T2r = T2p - T2q; TP = ri[WS(is, 23)]; T2H = ii[WS(is, 7)]; T2I = ii[WS(is, 23)]; } { E TS, TT, T2u, T2v; TS = ri[WS(is, 3)]; T2s = TO - TP; TQ = TO + TP; T4O = T2H + T2I; T2J = T2H - T2I; TT = ri[WS(is, 19)]; T2u = ii[WS(is, 3)]; T2v = ii[WS(is, 19)]; TV = ri[WS(is, 27)]; T2x = TS - TT; TU = TS + TT; T4T = T2u + T2v; T2w = T2u - T2v; TW = ri[WS(is, 11)]; T2A = ii[WS(is, 27)]; T2B = ii[WS(is, 11)]; } } { E T2z, T4U, T2C, TR, TY, T4Q, TX; T3O = T2s + T2r; T2t = T2r - T2s; T2z = TV - TW; TX = TV + TW; T4U = T2A + T2B; T2C = T2A - T2B; T3L = T2G - T2J; T2K = T2G + T2J; T4S = TN - TQ; TR = TN + TQ; TY = TU + TX; T4Q = TX - TU; { E T4P, T5G, T5H, T2y, T2D; T4P = T4N - T4O; T5G = T4N + T4O; T5H = T4T + T4U; T4V = T4T - T4U; T5F = TR - TY; TZ = TR + TY; T5I = T5G - T5H; T5X = T5G + T5H; T2L = T2x + T2w; T2y = T2w - T2x; T2D = T2z + T2C; T2M = T2z - T2C; T4R = T4P - T4Q; T5k = T4Q + T4P; T3M = T2D - T2y; T2E = T2y + T2D; } } } { E T2f, Ty, T4C, T20, T21, TB, T4D, T2i, TG, T26, TF, T4I, T25, TH, T29; E T2a; { E Tz, TA, T2g, T2h; { E Tw, Tx, T1Y, T1Z; Tw = ri[WS(is, 1)]; T5j = T4S + T4V; T4W = T4S - T4V; T3P = T2L - T2M; T2N = T2L + T2M; Tx = ri[WS(is, 17)]; T1Y = ii[WS(is, 1)]; T1Z = ii[WS(is, 17)]; Tz = ri[WS(is, 9)]; T2f = Tw - Tx; Ty = Tw + Tx; T4C = T1Y + T1Z; T20 = T1Y - T1Z; TA = ri[WS(is, 25)]; T2g = ii[WS(is, 9)]; T2h = ii[WS(is, 25)]; } { E TD, TE, T23, T24; TD = ri[WS(is, 5)]; T21 = Tz - TA; TB = Tz + TA; T4D = T2g + T2h; T2i = T2g - T2h; TE = ri[WS(is, 21)]; T23 = ii[WS(is, 5)]; T24 = ii[WS(is, 21)]; TG = ri[WS(is, 29)]; T26 = TD - TE; TF = TD + TE; T4I = T23 + T24; T25 = T23 - T24; TH = ri[WS(is, 13)]; T29 = ii[WS(is, 29)]; T2a = ii[WS(is, 13)]; } } { E T28, T4J, T2b, TC, TJ, T4F, TI; T3H = T21 + T20; T22 = T20 - T21; T28 = TG - TH; TI = TG + TH; T4J = T29 + T2a; T2b = T29 - T2a; T3E = T2f - T2i; T2j = T2f + T2i; T4H = Ty - TB; TC = Ty + TB; TJ = TF + TI; T4F = TI - TF; { E T4E, T5B, T5C, T27, T2c; T4E = T4C - T4D; T5B = T4C + T4D; T5C = T4I + T4J; T4K = T4I - T4J; T5A = TC - TJ; TK = TC + TJ; T5D = T5B - T5C; T5W = T5B + T5C; T2k = T26 + T25; T27 = T25 - T26; T2c = T28 + T2b; T2l = T28 - T2b; T4G = T4E - T4F; T5h = T4F + T4E; T3F = T2c - T27; T2d = T27 + T2c; } } } } } } { E T3I, T2m, Tv, T60, T11, T10, T5Z, T1w; { E T5f, T5w, T5q, T5m, T5v, T5p; { E T5d, T5g, T5o, T4B, T5a, T5n, T5e, T56, T4Y, T57, T55; { E T4X, T4M, T5b, T5c, T51, T54; { E T4t, T4A, T58, T59, T4L; T5d = T4r + T4s; T4t = T4r - T4s; T5g = T4H + T4K; T4L = T4H - T4K; T3I = T2k - T2l; T2m = T2k + T2l; T4A = T4w - T4z; T5o = T4w + T4z; T4X = FNMS(KP414213562, T4W, T4R); T58 = FMA(KP414213562, T4R, T4W); T59 = FNMS(KP414213562, T4G, T4L); T4M = FMA(KP414213562, T4L, T4G); T5b = FNMS(KP707106781, T4A, T4t); T4B = FMA(KP707106781, T4A, T4t); T5c = T59 + T58; T5a = T58 - T59; T5n = T50 + T4Z; T51 = T4Z - T50; T54 = T52 - T53; T5e = T53 + T52; } ro[WS(os, 14)] = FNMS(KP923879532, T5c, T5b); T56 = T4M + T4X; T4Y = T4M - T4X; T57 = FMA(KP707106781, T54, T51); T55 = FNMS(KP707106781, T54, T51); ro[WS(os, 30)] = FMA(KP923879532, T5c, T5b); } ro[WS(os, 6)] = FMA(KP923879532, T4Y, T4B); ro[WS(os, 22)] = FNMS(KP923879532, T4Y, T4B); io[WS(os, 6)] = FMA(KP923879532, T5a, T57); io[WS(os, 22)] = FNMS(KP923879532, T5a, T57); io[WS(os, 30)] = FMA(KP923879532, T56, T55); io[WS(os, 14)] = FNMS(KP923879532, T56, T55); { E T5i, T5l, T5r, T5u, T5s, T5t; T5i = FMA(KP414213562, T5h, T5g); T5s = FNMS(KP414213562, T5g, T5h); T5t = FMA(KP414213562, T5j, T5k); T5l = FNMS(KP414213562, T5k, T5j); T5r = FNMS(KP707106781, T5e, T5d); T5f = FMA(KP707106781, T5e, T5d); T5w = T5s + T5t; T5u = T5s - T5t; ro[WS(os, 26)] = FNMS(KP923879532, T5u, T5r); T5q = T5l - T5i; T5m = T5i + T5l; T5v = FMA(KP707106781, T5o, T5n); T5p = FNMS(KP707106781, T5o, T5n); ro[WS(os, 10)] = FMA(KP923879532, T5u, T5r); } } ro[WS(os, 2)] = FMA(KP923879532, T5m, T5f); ro[WS(os, 18)] = FNMS(KP923879532, T5m, T5f); io[WS(os, 2)] = FMA(KP923879532, T5w, T5v); io[WS(os, 18)] = FNMS(KP923879532, T5w, T5v); io[WS(os, 10)] = FMA(KP923879532, T5q, T5p); io[WS(os, 26)] = FNMS(KP923879532, T5q, T5p); { E Tf, T1v, T5z, T5U, T1g, Tu, T5O, T5K, T5T, T5N, T5V, T5Y; { E T5E, T5J, T5P, T5S, T5L, T5M; { E T5x, T5y, T5Q, T5R; Tf = T7 + Te; T5x = T7 - Te; T5y = T1n - T1u; T1v = T1n + T1u; T5E = T5A + T5D; T5Q = T5D - T5A; T5R = T5F + T5I; T5J = T5F - T5I; T5P = T5x - T5y; T5z = T5x + T5y; T5U = T5Q + T5R; T5S = T5Q - T5R; T1g = T18 + T1f; T5L = T18 - T1f; T5M = Tt - Tm; Tu = Tm + Tt; } ro[WS(os, 28)] = FNMS(KP707106781, T5S, T5P); T5O = T5J - T5E; T5K = T5E + T5J; T5T = T5M + T5L; T5N = T5L - T5M; ro[WS(os, 12)] = FMA(KP707106781, T5S, T5P); } ro[WS(os, 4)] = FMA(KP707106781, T5K, T5z); ro[WS(os, 20)] = FNMS(KP707106781, T5K, T5z); io[WS(os, 4)] = FMA(KP707106781, T5U, T5T); io[WS(os, 20)] = FNMS(KP707106781, T5U, T5T); io[WS(os, 12)] = FMA(KP707106781, T5O, T5N); io[WS(os, 28)] = FNMS(KP707106781, T5O, T5N); T5V = Tf - Tu; Tv = Tf + Tu; T60 = T5W + T5X; T5Y = T5W - T5X; ro[WS(os, 8)] = T5V + T5Y; T11 = TZ - TK; T10 = TK + TZ; T5Z = T1g + T1v; T1w = T1g - T1v; ro[WS(os, 24)] = T5V - T5Y; } } ro[0] = Tv + T10; ro[WS(os, 16)] = Tv - T10; io[0] = T5Z + T60; io[WS(os, 16)] = T5Z - T60; io[WS(os, 24)] = T1w - T11; io[WS(os, 8)] = T11 + T1w; { E T39, T3k, T3j, T3a, T3d, T3c, T47, T4i, T4h, T41, T3D, T48, T4b, T4a, T4e; E T3N, T45, T3Z, T42, T3K, T3Q, T4d; { E T2e, T37, T1X, T33, T31, T2n, T2F, T2O; { E T1H, T1W, T2X, T30; T39 = FMA(KP707106781, T1G, T1z); T1H = FNMS(KP707106781, T1G, T1z); T1W = T1O - T1V; T3k = T1O + T1V; T3j = FMA(KP707106781, T2W, T2T); T2X = FNMS(KP707106781, T2W, T2T); T30 = T2Y - T2Z; T3a = T2Z + T2Y; T3d = FMA(KP707106781, T2d, T22); T2e = FNMS(KP707106781, T2d, T22); T37 = FNMS(KP923879532, T1W, T1H); T1X = FMA(KP923879532, T1W, T1H); T33 = FMA(KP923879532, T30, T2X); T31 = FNMS(KP923879532, T30, T2X); T2n = FNMS(KP707106781, T2m, T2j); T3c = FMA(KP707106781, T2m, T2j); T3g = FMA(KP707106781, T2E, T2t); T2F = FNMS(KP707106781, T2E, T2t); T2O = FNMS(KP707106781, T2N, T2K); T3f = FMA(KP707106781, T2N, T2K); } { E T3V, T3Y, T3G, T3J; { E T3v, T35, T2o, T34, T2P, T3C; T47 = FNMS(KP707106781, T3u, T3t); T3v = FMA(KP707106781, T3u, T3t); T35 = FNMS(KP668178637, T2e, T2n); T2o = FMA(KP668178637, T2n, T2e); T34 = FMA(KP668178637, T2F, T2O); T2P = FNMS(KP668178637, T2O, T2F); T3C = T3y - T3B; T4i = T3y + T3B; T4h = FNMS(KP707106781, T3U, T3T); T3V = FMA(KP707106781, T3U, T3T); { E T38, T36, T32, T2Q; T38 = T35 + T34; T36 = T34 - T35; T32 = T2o + T2P; T2Q = T2o - T2P; T41 = FNMS(KP923879532, T3C, T3v); T3D = FMA(KP923879532, T3C, T3v); ro[WS(os, 29)] = FMA(KP831469612, T38, T37); ro[WS(os, 13)] = FNMS(KP831469612, T38, T37); io[WS(os, 5)] = FMA(KP831469612, T36, T33); io[WS(os, 21)] = FNMS(KP831469612, T36, T33); io[WS(os, 29)] = FMA(KP831469612, T32, T31); io[WS(os, 13)] = FNMS(KP831469612, T32, T31); ro[WS(os, 5)] = FMA(KP831469612, T2Q, T1X); ro[WS(os, 21)] = FNMS(KP831469612, T2Q, T1X); T3Y = T3W - T3X; T48 = T3X + T3W; } } T4b = FMA(KP707106781, T3F, T3E); T3G = FNMS(KP707106781, T3F, T3E); T3J = FNMS(KP707106781, T3I, T3H); T4a = FMA(KP707106781, T3I, T3H); T4e = FMA(KP707106781, T3M, T3L); T3N = FNMS(KP707106781, T3M, T3L); T45 = FMA(KP923879532, T3Y, T3V); T3Z = FNMS(KP923879532, T3Y, T3V); T42 = FNMS(KP668178637, T3G, T3J); T3K = FMA(KP668178637, T3J, T3G); T3Q = FNMS(KP707106781, T3P, T3O); T4d = FMA(KP707106781, T3P, T3O); } } { E T4p, T49, T4l, T4j, T4n, T4c, T43, T3R, T4m, T4f; T43 = FMA(KP668178637, T3N, T3Q); T3R = FNMS(KP668178637, T3Q, T3N); T4p = FMA(KP923879532, T48, T47); T49 = FNMS(KP923879532, T48, T47); { E T44, T46, T40, T3S; T44 = T42 - T43; T46 = T42 + T43; T40 = T3R - T3K; T3S = T3K + T3R; ro[WS(os, 11)] = FMA(KP831469612, T44, T41); ro[WS(os, 27)] = FNMS(KP831469612, T44, T41); io[WS(os, 3)] = FMA(KP831469612, T46, T45); io[WS(os, 19)] = FNMS(KP831469612, T46, T45); io[WS(os, 11)] = FMA(KP831469612, T40, T3Z); io[WS(os, 27)] = FNMS(KP831469612, T40, T3Z); ro[WS(os, 3)] = FMA(KP831469612, T3S, T3D); ro[WS(os, 19)] = FNMS(KP831469612, T3S, T3D); } T4l = FNMS(KP923879532, T4i, T4h); T4j = FMA(KP923879532, T4i, T4h); T4n = FNMS(KP198912367, T4a, T4b); T4c = FMA(KP198912367, T4b, T4a); T4m = FMA(KP198912367, T4d, T4e); T4f = FNMS(KP198912367, T4e, T4d); T3n = FNMS(KP923879532, T3a, T39); T3b = FMA(KP923879532, T3a, T39); { E T4q, T4o, T4k, T4g; T4q = T4n + T4m; T4o = T4m - T4n; T4k = T4c + T4f; T4g = T4c - T4f; ro[WS(os, 31)] = FMA(KP980785280, T4q, T4p); ro[WS(os, 15)] = FNMS(KP980785280, T4q, T4p); io[WS(os, 7)] = FMA(KP980785280, T4o, T4l); io[WS(os, 23)] = FNMS(KP980785280, T4o, T4l); io[WS(os, 31)] = FMA(KP980785280, T4k, T4j); io[WS(os, 15)] = FNMS(KP980785280, T4k, T4j); ro[WS(os, 7)] = FMA(KP980785280, T4g, T49); ro[WS(os, 23)] = FNMS(KP980785280, T4g, T49); } T3r = FMA(KP923879532, T3k, T3j); T3l = FNMS(KP923879532, T3k, T3j); T3o = FNMS(KP198912367, T3c, T3d); T3e = FMA(KP198912367, T3d, T3c); } } } } T3h = FNMS(KP198912367, T3g, T3f); T3p = FMA(KP198912367, T3f, T3g); { E T3s, T3q, T3i, T3m; T3s = T3o + T3p; T3q = T3o - T3p; T3i = T3e + T3h; T3m = T3h - T3e; ro[WS(os, 9)] = FMA(KP980785280, T3q, T3n); ro[WS(os, 25)] = FNMS(KP980785280, T3q, T3n); io[WS(os, 1)] = FMA(KP980785280, T3s, T3r); io[WS(os, 17)] = FNMS(KP980785280, T3s, T3r); io[WS(os, 9)] = FMA(KP980785280, T3m, T3l); io[WS(os, 25)] = FNMS(KP980785280, T3m, T3l); ro[WS(os, 1)] = FMA(KP980785280, T3i, T3b); ro[WS(os, 17)] = FNMS(KP980785280, T3i, T3b); } } } } static const kdft_desc desc = { 32, "n1_32", {236, 0, 136, 0}, &GENUS, 0, 0, 0, 0 }; void X(codelet_n1_32) (planner *p) { X(kdft_register) (p, n1_32, &desc); } #else /* HAVE_FMA */ /* Generated by: ../../../genfft/gen_notw.native -compact -variables 4 -pipeline-latency 4 -n 32 -name n1_32 -include n.h */ /* * This function contains 372 FP additions, 84 FP multiplications, * (or, 340 additions, 52 multiplications, 32 fused multiply/add), * 100 stack variables, 7 constants, and 128 memory accesses */ #include "n.h" static void n1_32(const R *ri, const R *ii, R *ro, R *io, stride is, stride os, INT v, INT ivs, INT ovs) { DK(KP831469612, +0.831469612302545237078788377617905756738560812); DK(KP555570233, +0.555570233019602224742830813948532874374937191); DK(KP195090322, +0.195090322016128267848284868477022240927691618); DK(KP980785280, +0.980785280403230449126182236134239036973933731); DK(KP923879532, +0.923879532511286756128183189396788286822416626); DK(KP382683432, +0.382683432365089771728459984030398866761344562); DK(KP707106781, +0.707106781186547524400844362104849039284835938); { INT i; for (i = v; i > 0; i = i - 1, ri = ri + ivs, ii = ii + ivs, ro = ro + ovs, io = io + ovs, MAKE_VOLATILE_STRIDE(128, is), MAKE_VOLATILE_STRIDE(128, os)) { E T7, T4r, T4Z, T18, T1z, T3t, T3T, T2T, Te, T1f, T50, T4s, T2W, T3u, T1G; E T3U, Tm, T1n, T1O, T2Z, T3y, T3X, T4w, T53, Tt, T1u, T1V, T2Y, T3B, T3W; E T4z, T52, T2t, T3L, T3O, T2K, TR, TY, T5F, T5G, T5H, T5I, T4R, T5j, T2E; E T3P, T4W, T5k, T2N, T3M, T22, T3E, T3H, T2j, TC, TJ, T5A, T5B, T5C, T5D; E T4G, T5g, T2d, T3F, T4L, T5h, T2m, T3I; { E T3, T1x, T14, T2S, T6, T2R, T17, T1y; { E T1, T2, T12, T13; T1 = ri[0]; T2 = ri[WS(is, 16)]; T3 = T1 + T2; T1x = T1 - T2; T12 = ii[0]; T13 = ii[WS(is, 16)]; T14 = T12 + T13; T2S = T12 - T13; } { E T4, T5, T15, T16; T4 = ri[WS(is, 8)]; T5 = ri[WS(is, 24)]; T6 = T4 + T5; T2R = T4 - T5; T15 = ii[WS(is, 8)]; T16 = ii[WS(is, 24)]; T17 = T15 + T16; T1y = T15 - T16; } T7 = T3 + T6; T4r = T3 - T6; T4Z = T14 - T17; T18 = T14 + T17; T1z = T1x - T1y; T3t = T1x + T1y; T3T = T2S - T2R; T2T = T2R + T2S; } { E Ta, T1B, T1b, T1A, Td, T1D, T1e, T1E; { E T8, T9, T19, T1a; T8 = ri[WS(is, 4)]; T9 = ri[WS(is, 20)]; Ta = T8 + T9; T1B = T8 - T9; T19 = ii[WS(is, 4)]; T1a = ii[WS(is, 20)]; T1b = T19 + T1a; T1A = T19 - T1a; } { E Tb, Tc, T1c, T1d; Tb = ri[WS(is, 28)]; Tc = ri[WS(is, 12)]; Td = Tb + Tc; T1D = Tb - Tc; T1c = ii[WS(is, 28)]; T1d = ii[WS(is, 12)]; T1e = T1c + T1d; T1E = T1c - T1d; } Te = Ta + Td; T1f = T1b + T1e; T50 = Td - Ta; T4s = T1b - T1e; { E T2U, T2V, T1C, T1F; T2U = T1D - T1E; T2V = T1B + T1A; T2W = KP707106781 * (T2U - T2V); T3u = KP707106781 * (T2V + T2U); T1C = T1A - T1B; T1F = T1D + T1E; T1G = KP707106781 * (T1C - T1F); T3U = KP707106781 * (T1C + T1F); } } { E Ti, T1L, T1j, T1J, Tl, T1I, T1m, T1M, T1K, T1N; { E Tg, Th, T1h, T1i; Tg = ri[WS(is, 2)]; Th = ri[WS(is, 18)]; Ti = Tg + Th; T1L = Tg - Th; T1h = ii[WS(is, 2)]; T1i = ii[WS(is, 18)]; T1j = T1h + T1i; T1J = T1h - T1i; } { E Tj, Tk, T1k, T1l; Tj = ri[WS(is, 10)]; Tk = ri[WS(is, 26)]; Tl = Tj + Tk; T1I = Tj - Tk; T1k = ii[WS(is, 10)]; T1l = ii[WS(is, 26)]; T1m = T1k + T1l; T1M = T1k - T1l; } Tm = Ti + Tl; T1n = T1j + T1m; T1K = T1I + T1J; T1N = T1L - T1M; T1O = FNMS(KP923879532, T1N, KP382683432 * T1K); T2Z = FMA(KP923879532, T1K, KP382683432 * T1N); { E T3w, T3x, T4u, T4v; T3w = T1J - T1I; T3x = T1L + T1M; T3y = FNMS(KP382683432, T3x, KP923879532 * T3w); T3X = FMA(KP382683432, T3w, KP923879532 * T3x); T4u = T1j - T1m; T4v = Ti - Tl; T4w = T4u - T4v; T53 = T4v + T4u; } } { E Tp, T1S, T1q, T1Q, Ts, T1P, T1t, T1T, T1R, T1U; { E Tn, To, T1o, T1p; Tn = ri[WS(is, 30)]; To = ri[WS(is, 14)]; Tp = Tn + To; T1S = Tn - To; T1o = ii[WS(is, 30)]; T1p = ii[WS(is, 14)]; T1q = T1o + T1p; T1Q = T1o - T1p; } { E Tq, Tr, T1r, T1s; Tq = ri[WS(is, 6)]; Tr = ri[WS(is, 22)]; Ts = Tq + Tr; T1P = Tq - Tr; T1r = ii[WS(is, 6)]; T1s = ii[WS(is, 22)]; T1t = T1r + T1s; T1T = T1r - T1s; } Tt = Tp + Ts; T1u = T1q + T1t; T1R = T1P + T1Q; T1U = T1S - T1T; T1V = FMA(KP382683432, T1R, KP923879532 * T1U); T2Y = FNMS(KP923879532, T1R, KP382683432 * T1U); { E T3z, T3A, T4x, T4y; T3z = T1Q - T1P; T3A = T1S + T1T; T3B = FMA(KP923879532, T3z, KP382683432 * T3A); T3W = FNMS(KP382683432, T3z, KP923879532 * T3A); T4x = Tp - Ts; T4y = T1q - T1t; T4z = T4x + T4y; T52 = T4x - T4y; } } { E TN, T2p, T2J, T4S, TQ, T2G, T2s, T4T, TU, T2x, T2w, T4O, TX, T2z, T2C; E T4P; { E TL, TM, T2H, T2I; TL = ri[WS(is, 31)]; TM = ri[WS(is, 15)]; TN = TL + TM; T2p = TL - TM; T2H = ii[WS(is, 31)]; T2I = ii[WS(is, 15)]; T2J = T2H - T2I; T4S = T2H + T2I; } { E TO, TP, T2q, T2r; TO = ri[WS(is, 7)]; TP = ri[WS(is, 23)]; TQ = TO + TP; T2G = TO - TP; T2q = ii[WS(is, 7)]; T2r = ii[WS(is, 23)]; T2s = T2q - T2r; T4T = T2q + T2r; } { E TS, TT, T2u, T2v; TS = ri[WS(is, 3)]; TT = ri[WS(is, 19)]; TU = TS + TT; T2x = TS - TT; T2u = ii[WS(is, 3)]; T2v = ii[WS(is, 19)]; T2w = T2u - T2v; T4O = T2u + T2v; } { E TV, TW, T2A, T2B; TV = ri[WS(is, 27)]; TW = ri[WS(is, 11)]; TX = TV + TW; T2z = TV - TW; T2A = ii[WS(is, 27)]; T2B = ii[WS(is, 11)]; T2C = T2A - T2B; T4P = T2A + T2B; } T2t = T2p - T2s; T3L = T2p + T2s; T3O = T2J - T2G; T2K = T2G + T2J; TR = TN + TQ; TY = TU + TX; T5F = TR - TY; { E T4N, T4Q, T2y, T2D; T5G = T4S + T4T; T5H = T4O + T4P; T5I = T5G - T5H; T4N = TN - TQ; T4Q = T4O - T4P; T4R = T4N - T4Q; T5j = T4N + T4Q; T2y = T2w - T2x; T2D = T2z + T2C; T2E = KP707106781 * (T2y - T2D); T3P = KP707106781 * (T2y + T2D); { E T4U, T4V, T2L, T2M; T4U = T4S - T4T; T4V = TX - TU; T4W = T4U - T4V; T5k = T4V + T4U; T2L = T2z - T2C; T2M = T2x + T2w; T2N = KP707106781 * (T2L - T2M); T3M = KP707106781 * (T2M + T2L); } } } { E Ty, T2f, T21, T4C, TB, T1Y, T2i, T4D, TF, T28, T2b, T4I, TI, T23, T26; E T4J; { E Tw, Tx, T1Z, T20; Tw = ri[WS(is, 1)]; Tx = ri[WS(is, 17)]; Ty = Tw + Tx; T2f = Tw - Tx; T1Z = ii[WS(is, 1)]; T20 = ii[WS(is, 17)]; T21 = T1Z - T20; T4C = T1Z + T20; } { E Tz, TA, T2g, T2h; Tz = ri[WS(is, 9)]; TA = ri[WS(is, 25)]; TB = Tz + TA; T1Y = Tz - TA; T2g = ii[WS(is, 9)]; T2h = ii[WS(is, 25)]; T2i = T2g - T2h; T4D = T2g + T2h; } { E TD, TE, T29, T2a; TD = ri[WS(is, 5)]; TE = ri[WS(is, 21)]; TF = TD + TE; T28 = TD - TE; T29 = ii[WS(is, 5)]; T2a = ii[WS(is, 21)]; T2b = T29 - T2a; T4I = T29 + T2a; } { E TG, TH, T24, T25; TG = ri[WS(is, 29)]; TH = ri[WS(is, 13)]; TI = TG + TH; T23 = TG - TH; T24 = ii[WS(is, 29)]; T25 = ii[WS(is, 13)]; T26 = T24 - T25; T4J = T24 + T25; } T22 = T1Y + T21; T3E = T2f + T2i; T3H = T21 - T1Y; T2j = T2f - T2i; TC = Ty + TB; TJ = TF + TI; T5A = TC - TJ; { E T4E, T4F, T27, T2c; T5B = T4C + T4D; T5C = T4I + T4J; T5D = T5B - T5C; T4E = T4C - T4D; T4F = TI - TF; T4G = T4E - T4F; T5g = T4F + T4E; T27 = T23 - T26; T2c = T28 + T2b; T2d = KP707106781 * (T27 - T2c); T3F = KP707106781 * (T2c + T27); { E T4H, T4K, T2k, T2l; T4H = Ty - TB; T4K = T4I - T4J; T4L = T4H - T4K; T5h = T4H + T4K; T2k = T2b - T28; T2l = T23 + T26; T2m = KP707106781 * (T2k - T2l); T3I = KP707106781 * (T2k + T2l); } } } { E T4B, T57, T5a, T5c, T4Y, T56, T55, T5b; { E T4t, T4A, T58, T59; T4t = T4r - T4s; T4A = KP707106781 * (T4w - T4z); T4B = T4t + T4A; T57 = T4t - T4A; T58 = FNMS(KP923879532, T4L, KP382683432 * T4G); T59 = FMA(KP382683432, T4W, KP923879532 * T4R); T5a = T58 - T59; T5c = T58 + T59; } { E T4M, T4X, T51, T54; T4M = FMA(KP923879532, T4G, KP382683432 * T4L); T4X = FNMS(KP923879532, T4W, KP382683432 * T4R); T4Y = T4M + T4X; T56 = T4X - T4M; T51 = T4Z - T50; T54 = KP707106781 * (T52 - T53); T55 = T51 - T54; T5b = T51 + T54; } ro[WS(os, 22)] = T4B - T4Y; io[WS(os, 22)] = T5b - T5c; ro[WS(os, 6)] = T4B + T4Y; io[WS(os, 6)] = T5b + T5c; io[WS(os, 30)] = T55 - T56; ro[WS(os, 30)] = T57 - T5a; io[WS(os, 14)] = T55 + T56; ro[WS(os, 14)] = T57 + T5a; } { E T5f, T5r, T5u, T5w, T5m, T5q, T5p, T5v; { E T5d, T5e, T5s, T5t; T5d = T4r + T4s; T5e = KP707106781 * (T53 + T52); T5f = T5d + T5e; T5r = T5d - T5e; T5s = FNMS(KP382683432, T5h, KP923879532 * T5g); T5t = FMA(KP923879532, T5k, KP382683432 * T5j); T5u = T5s - T5t; T5w = T5s + T5t; } { E T5i, T5l, T5n, T5o; T5i = FMA(KP382683432, T5g, KP923879532 * T5h); T5l = FNMS(KP382683432, T5k, KP923879532 * T5j); T5m = T5i + T5l; T5q = T5l - T5i; T5n = T50 + T4Z; T5o = KP707106781 * (T4w + T4z); T5p = T5n - T5o; T5v = T5n + T5o; } ro[WS(os, 18)] = T5f - T5m; io[WS(os, 18)] = T5v - T5w; ro[WS(os, 2)] = T5f + T5m; io[WS(os, 2)] = T5v + T5w; io[WS(os, 26)] = T5p - T5q; ro[WS(os, 26)] = T5r - T5u; io[WS(os, 10)] = T5p + T5q; ro[WS(os, 10)] = T5r + T5u; } { E T5z, T5P, T5S, T5U, T5K, T5O, T5N, T5T; { E T5x, T5y, T5Q, T5R; T5x = T7 - Te; T5y = T1n - T1u; T5z = T5x + T5y; T5P = T5x - T5y; T5Q = T5D - T5A; T5R = T5F + T5I; T5S = KP707106781 * (T5Q - T5R); T5U = KP707106781 * (T5Q + T5R); } { E T5E, T5J, T5L, T5M; T5E = T5A + T5D; T5J = T5F - T5I; T5K = KP707106781 * (T5E + T5J); T5O = KP707106781 * (T5J - T5E); T5L = T18 - T1f; T5M = Tt - Tm; T5N = T5L - T5M; T5T = T5M + T5L; } ro[WS(os, 20)] = T5z - T5K; io[WS(os, 20)] = T5T - T5U; ro[WS(os, 4)] = T5z + T5K; io[WS(os, 4)] = T5T + T5U; io[WS(os, 28)] = T5N - T5O; ro[WS(os, 28)] = T5P - T5S; io[WS(os, 12)] = T5N + T5O; ro[WS(os, 12)] = T5P + T5S; } { E Tv, T5V, T5Y, T60, T10, T11, T1w, T5Z; { E Tf, Tu, T5W, T5X; Tf = T7 + Te; Tu = Tm + Tt; Tv = Tf + Tu; T5V = Tf - Tu; T5W = T5B + T5C; T5X = T5G + T5H; T5Y = T5W - T5X; T60 = T5W + T5X; } { E TK, TZ, T1g, T1v; TK = TC + TJ; TZ = TR + TY; T10 = TK + TZ; T11 = TZ - TK; T1g = T18 + T1f; T1v = T1n + T1u; T1w = T1g - T1v; T5Z = T1g + T1v; } ro[WS(os, 16)] = Tv - T10; io[WS(os, 16)] = T5Z - T60; ro[0] = Tv + T10; io[0] = T5Z + T60; io[WS(os, 8)] = T11 + T1w; ro[WS(os, 8)] = T5V + T5Y; io[WS(os, 24)] = T1w - T11; ro[WS(os, 24)] = T5V - T5Y; } { E T1X, T33, T31, T37, T2o, T34, T2P, T35; { E T1H, T1W, T2X, T30; T1H = T1z - T1G; T1W = T1O - T1V; T1X = T1H + T1W; T33 = T1H - T1W; T2X = T2T - T2W; T30 = T2Y - T2Z; T31 = T2X - T30; T37 = T2X + T30; } { E T2e, T2n, T2F, T2O; T2e = T22 - T2d; T2n = T2j - T2m; T2o = FMA(KP980785280, T2e, KP195090322 * T2n); T34 = FNMS(KP980785280, T2n, KP195090322 * T2e); T2F = T2t - T2E; T2O = T2K - T2N; T2P = FNMS(KP980785280, T2O, KP195090322 * T2F); T35 = FMA(KP195090322, T2O, KP980785280 * T2F); } { E T2Q, T38, T32, T36; T2Q = T2o + T2P; ro[WS(os, 23)] = T1X - T2Q; ro[WS(os, 7)] = T1X + T2Q; T38 = T34 + T35; io[WS(os, 23)] = T37 - T38; io[WS(os, 7)] = T37 + T38; T32 = T2P - T2o; io[WS(os, 31)] = T31 - T32; io[WS(os, 15)] = T31 + T32; T36 = T34 - T35; ro[WS(os, 31)] = T33 - T36; ro[WS(os, 15)] = T33 + T36; } } { E T3D, T41, T3Z, T45, T3K, T42, T3R, T43; { E T3v, T3C, T3V, T3Y; T3v = T3t - T3u; T3C = T3y - T3B; T3D = T3v + T3C; T41 = T3v - T3C; T3V = T3T - T3U; T3Y = T3W - T3X; T3Z = T3V - T3Y; T45 = T3V + T3Y; } { E T3G, T3J, T3N, T3Q; T3G = T3E - T3F; T3J = T3H - T3I; T3K = FMA(KP555570233, T3G, KP831469612 * T3J); T42 = FNMS(KP831469612, T3G, KP555570233 * T3J); T3N = T3L - T3M; T3Q = T3O - T3P; T3R = FNMS(KP831469612, T3Q, KP555570233 * T3N); T43 = FMA(KP831469612, T3N, KP555570233 * T3Q); } { E T3S, T46, T40, T44; T3S = T3K + T3R; ro[WS(os, 21)] = T3D - T3S; ro[WS(os, 5)] = T3D + T3S; T46 = T42 + T43; io[WS(os, 21)] = T45 - T46; io[WS(os, 5)] = T45 + T46; T40 = T3R - T3K; io[WS(os, 29)] = T3Z - T40; io[WS(os, 13)] = T3Z + T40; T44 = T42 - T43; ro[WS(os, 29)] = T41 - T44; ro[WS(os, 13)] = T41 + T44; } } { E T49, T4l, T4j, T4p, T4c, T4m, T4f, T4n; { E T47, T48, T4h, T4i; T47 = T3t + T3u; T48 = T3X + T3W; T49 = T47 + T48; T4l = T47 - T48; T4h = T3T + T3U; T4i = T3y + T3B; T4j = T4h - T4i; T4p = T4h + T4i; } { E T4a, T4b, T4d, T4e; T4a = T3E + T3F; T4b = T3H + T3I; T4c = FMA(KP980785280, T4a, KP195090322 * T4b); T4m = FNMS(KP195090322, T4a, KP980785280 * T4b); T4d = T3L + T3M; T4e = T3O + T3P; T4f = FNMS(KP195090322, T4e, KP980785280 * T4d); T4n = FMA(KP195090322, T4d, KP980785280 * T4e); } { E T4g, T4q, T4k, T4o; T4g = T4c + T4f; ro[WS(os, 17)] = T49 - T4g; ro[WS(os, 1)] = T49 + T4g; T4q = T4m + T4n; io[WS(os, 17)] = T4p - T4q; io[WS(os, 1)] = T4p + T4q; T4k = T4f - T4c; io[WS(os, 25)] = T4j - T4k; io[WS(os, 9)] = T4j + T4k; T4o = T4m - T4n; ro[WS(os, 25)] = T4l - T4o; ro[WS(os, 9)] = T4l + T4o; } } { E T3b, T3n, T3l, T3r, T3e, T3o, T3h, T3p; { E T39, T3a, T3j, T3k; T39 = T1z + T1G; T3a = T2Z + T2Y; T3b = T39 + T3a; T3n = T39 - T3a; T3j = T2T + T2W; T3k = T1O + T1V; T3l = T3j - T3k; T3r = T3j + T3k; } { E T3c, T3d, T3f, T3g; T3c = T22 + T2d; T3d = T2j + T2m; T3e = FMA(KP555570233, T3c, KP831469612 * T3d); T3o = FNMS(KP555570233, T3d, KP831469612 * T3c); T3f = T2t + T2E; T3g = T2K + T2N; T3h = FNMS(KP555570233, T3g, KP831469612 * T3f); T3p = FMA(KP831469612, T3g, KP555570233 * T3f); } { E T3i, T3s, T3m, T3q; T3i = T3e + T3h; ro[WS(os, 19)] = T3b - T3i; ro[WS(os, 3)] = T3b + T3i; T3s = T3o + T3p; io[WS(os, 19)] = T3r - T3s; io[WS(os, 3)] = T3r + T3s; T3m = T3h - T3e; io[WS(os, 27)] = T3l - T3m; io[WS(os, 11)] = T3l + T3m; T3q = T3o - T3p; ro[WS(os, 27)] = T3n - T3q; ro[WS(os, 11)] = T3n + T3q; } } } } } static const kdft_desc desc = { 32, "n1_32", {340, 52, 32, 0}, &GENUS, 0, 0, 0, 0 }; void X(codelet_n1_32) (planner *p) { X(kdft_register) (p, n1_32, &desc); } #endif /* HAVE_FMA */ fftw-3.3.3/dft/scalar/codelets/n1_64.c0000644000175000017500000024337212054410307014235 00000000000000/* * Copyright (c) 2003, 2007-11 Matteo Frigo * Copyright (c) 2003, 2007-11 Massachusetts Institute of Technology * * This program is free software; you can redistribute it and/or modify * it under the terms of the GNU General Public License as published by * the Free Software Foundation; either version 2 of the License, or * (at your option) any later version. * * This program is distributed in the hope that it will be useful, * but WITHOUT ANY WARRANTY; without even the implied warranty of * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the * GNU General Public License for more details. * * You should have received a copy of the GNU General Public License * along with this program; if not, write to the Free Software * Foundation, Inc., 51 Franklin Street, Fifth Floor, Boston, MA 02110-1301 USA * */ /* This file was automatically generated --- DO NOT EDIT */ /* Generated on Sun Nov 25 07:35:46 EST 2012 */ #include "codelet-dft.h" #ifdef HAVE_FMA /* Generated by: ../../../genfft/gen_notw.native -fma -reorder-insns -schedule-for-pipeline -compact -variables 4 -pipeline-latency 4 -n 64 -name n1_64 -include n.h */ /* * This function contains 912 FP additions, 392 FP multiplications, * (or, 520 additions, 0 multiplications, 392 fused multiply/add), * 202 stack variables, 15 constants, and 256 memory accesses */ #include "n.h" static void n1_64(const R *ri, const R *ii, R *ro, R *io, stride is, stride os, INT v, INT ivs, INT ovs) { DK(KP956940335, +0.956940335732208864935797886980269969482849206); DK(KP881921264, +0.881921264348355029712756863660388349508442621); DK(KP534511135, +0.534511135950791641089685961295362908582039528); DK(KP303346683, +0.303346683607342391675883946941299872384187453); DK(KP995184726, +0.995184726672196886244836953109479921575474869); DK(KP773010453, +0.773010453362736960810906609758469800971041293); DK(KP820678790, +0.820678790828660330972281985331011598767386482); DK(KP098491403, +0.098491403357164253077197521291327432293052451); DK(KP980785280, +0.980785280403230449126182236134239036973933731); DK(KP831469612, +0.831469612302545237078788377617905756738560812); DK(KP668178637, +0.668178637919298919997757686523080761552472251); DK(KP198912367, +0.198912367379658006911597622644676228597850501); DK(KP923879532, +0.923879532511286756128183189396788286822416626); DK(KP707106781, +0.707106781186547524400844362104849039284835938); DK(KP414213562, +0.414213562373095048801688724209698078569671875); { INT i; for (i = v; i > 0; i = i - 1, ri = ri + ivs, ii = ii + ivs, ro = ro + ovs, io = io + ovs, MAKE_VOLATILE_STRIDE(256, is), MAKE_VOLATILE_STRIDE(256, os)) { E T9b, T9e; { E T7B, T37, T5Z, T8F, Td9, Tf, TcB, TbB, T7C, T62, TdH, T2i, Tcb, Tah, T8G; E T3e, Tu, TdI, Tak, TbC, TbD, Tan, Tda, T2x, T65, T3m, T8I, T7G, T8J, T7J; E T64, T3t, Tdd, TK, Tce, Tas, Tcf, Tav, Tdc, T2N, T6G, T3G, T9k, T7O, T9l; E T7R, T6H, T3N, TdA, T1L, Tct, Tbs, Teo, Tdx, T6Y, T5j, T6V, T5Q, T9z, T8y; E Tcw, Tbb, T9C, T8n, Tdf, TZ, Tch, Taz, Tci, TaC, Tdg, T32, T6J, T3Z, T9n; E T7V, T9o, T7Y, T6K, T46, Tdp, T1g, Tcm, Tb1, Tej, Tdm, T6R, T4q, T6O, T4X; E T9s, T8f, Tcp, TaK, T9v, T84, Tdn, T1v, Tcq, Tb4, Tek, Tds, T6P, T4N, T6S; E T50, T9w, T8i, Tcn, TaV, T9t, T8b, Tdy, T20, Tcx, Tbv, Tep, TdD, T8q, T6W; E T5G, T6Z, T5T, T8t, T9D, T8B, Tcu, Tbm, T8l, T8m; { E T3s, T3p, T3M, T3J; { E Taf, T3d, T3a, Tag; { E T35, T3, T5Y, T26, T5X, T6, T36, T29, Tb, T39, Ta, T38, T2d, Tc, T2e; E T2f; { E T4, T5, T27, T28; { E T1, T2, T24, T25; T1 = ri[0]; T2 = ri[WS(is, 32)]; T24 = ii[0]; T25 = ii[WS(is, 32)]; T4 = ri[WS(is, 16)]; T35 = T1 - T2; T3 = T1 + T2; T5Y = T24 - T25; T26 = T24 + T25; T5 = ri[WS(is, 48)]; T27 = ii[WS(is, 16)]; T28 = ii[WS(is, 48)]; } { E T8, T9, T2b, T2c; T8 = ri[WS(is, 8)]; T5X = T4 - T5; T6 = T4 + T5; T36 = T27 - T28; T29 = T27 + T28; T9 = ri[WS(is, 40)]; T2b = ii[WS(is, 8)]; T2c = ii[WS(is, 40)]; Tb = ri[WS(is, 56)]; T39 = T8 - T9; Ta = T8 + T9; T38 = T2b - T2c; T2d = T2b + T2c; Tc = ri[WS(is, 24)]; T2e = ii[WS(is, 56)]; T2f = ii[WS(is, 24)]; } } { E T3b, T3c, T2g, T7, Te, Tbz, Td; T7B = T35 + T36; T37 = T35 - T36; T3b = Tb - Tc; Td = Tb + Tc; T3c = T2e - T2f; T2g = T2e + T2f; T5Z = T5X + T5Y; T8F = T5Y - T5X; Taf = T3 - T6; T7 = T3 + T6; Te = Ta + Td; Tbz = Td - Ta; { E T2a, T60, T61, TbA, T2h; TbA = T26 - T29; T2a = T26 + T29; T3d = T3b + T3c; T60 = T3b - T3c; Td9 = T7 - Te; Tf = T7 + Te; TcB = TbA - Tbz; TbB = Tbz + TbA; T61 = T39 + T38; T3a = T38 - T39; T2h = T2d + T2g; Tag = T2d - T2g; T7C = T61 + T60; T62 = T60 - T61; TdH = T2a - T2h; T2i = T2a + T2h; } } } { E T3j, Ti, T3h, T2l, T3g, Tl, T3k, T2o, Tq, T3q, Tp, T3o, T2s, Tr, T2t; E T2u; { E Tj, Tk, T2m, T2n; { E Tg, Th, T2j, T2k; Tg = ri[WS(is, 4)]; Tcb = Taf - Tag; Tah = Taf + Tag; T8G = T3a + T3d; T3e = T3a - T3d; Th = ri[WS(is, 36)]; T2j = ii[WS(is, 4)]; T2k = ii[WS(is, 36)]; Tj = ri[WS(is, 20)]; T3j = Tg - Th; Ti = Tg + Th; T3h = T2j - T2k; T2l = T2j + T2k; Tk = ri[WS(is, 52)]; T2m = ii[WS(is, 20)]; T2n = ii[WS(is, 52)]; } { E Tn, To, T2q, T2r; Tn = ri[WS(is, 60)]; T3g = Tj - Tk; Tl = Tj + Tk; T3k = T2m - T2n; T2o = T2m + T2n; To = ri[WS(is, 28)]; T2q = ii[WS(is, 60)]; T2r = ii[WS(is, 28)]; Tq = ri[WS(is, 12)]; T3q = Tn - To; Tp = Tn + To; T3o = T2q - T2r; T2s = T2q + T2r; Tr = ri[WS(is, 44)]; T2t = ii[WS(is, 12)]; T2u = ii[WS(is, 44)]; } } { E T3n, T3r, T2p, T2w; { E Tai, Tm, T2v, Tal, Tt, Taj, Ts, Tam; Tai = Ti - Tl; Tm = Ti + Tl; T3n = Tq - Tr; Ts = Tq + Tr; T3r = T2t - T2u; T2v = T2t + T2u; Tal = Tp - Ts; Tt = Tp + Ts; Taj = T2l - T2o; T2p = T2l + T2o; Tam = T2s - T2v; T2w = T2s + T2v; Tu = Tm + Tt; TdI = Tt - Tm; Tak = Tai + Taj; TbC = Taj - Tai; TbD = Tal + Tam; Tan = Tal - Tam; } { E T7F, T7E, T3i, T3l, T7H, T7I; T7F = T3h - T3g; T3i = T3g + T3h; T3l = T3j - T3k; T7E = T3j + T3k; Tda = T2p - T2w; T2x = T2p + T2w; T65 = FNMS(KP414213562, T3i, T3l); T3m = FMA(KP414213562, T3l, T3i); T3s = T3q - T3r; T7H = T3q + T3r; T7I = T3o - T3n; T3p = T3n + T3o; T8I = FNMS(KP414213562, T7E, T7F); T7G = FMA(KP414213562, T7F, T7E); T8J = FMA(KP414213562, T7H, T7I); T7J = FNMS(KP414213562, T7I, T7H); } } } } { E T3H, Ty, T3x, T2B, T3w, TB, T3I, T2E, TI, T2L, T3z, TF, T3E, T3K, T2I; E T3A; { E T2z, T2A, Tz, TA, Tw, Tx, T2C, T2D; Tw = ri[WS(is, 2)]; Tx = ri[WS(is, 34)]; T2z = ii[WS(is, 2)]; T64 = FMA(KP414213562, T3p, T3s); T3t = FNMS(KP414213562, T3s, T3p); T3H = Tw - Tx; Ty = Tw + Tx; T2A = ii[WS(is, 34)]; Tz = ri[WS(is, 18)]; TA = ri[WS(is, 50)]; T2C = ii[WS(is, 18)]; T3x = T2z - T2A; T2B = T2z + T2A; T3w = Tz - TA; TB = Tz + TA; T2D = ii[WS(is, 50)]; { E T2J, T3C, T2K, TG, TH; TG = ri[WS(is, 58)]; TH = ri[WS(is, 26)]; T2J = ii[WS(is, 58)]; T3I = T2C - T2D; T2E = T2C + T2D; T3C = TG - TH; TI = TG + TH; T2K = ii[WS(is, 26)]; { E T2G, T2H, TD, TE, T3D; TD = ri[WS(is, 10)]; TE = ri[WS(is, 42)]; T3D = T2J - T2K; T2L = T2J + T2K; T2G = ii[WS(is, 10)]; T3z = TD - TE; TF = TD + TE; T2H = ii[WS(is, 42)]; T3E = T3C - T3D; T3K = T3C + T3D; T2I = T2G + T2H; T3A = T2G - T2H; } } } { E T3L, T3B, T2F, T2M; { E Tat, Taq, Tar, TC, TJ, Tau; Tat = Ty - TB; TC = Ty + TB; TJ = TF + TI; Taq = TI - TF; T3L = T3A - T3z; T3B = T3z + T3A; Tdd = TC - TJ; TK = TC + TJ; Tar = T2B - T2E; T2F = T2B + T2E; Tau = T2I - T2L; T2M = T2I + T2L; Tce = Tar - Taq; Tas = Taq + Tar; Tcf = Tat - Tau; Tav = Tat + Tau; } { E T7M, T7Q, T7N, T3y, T3F, T7P; T7M = T3x - T3w; T3y = T3w + T3x; T3F = T3B - T3E; T7Q = T3B + T3E; Tdc = T2F - T2M; T2N = T2F + T2M; T6G = FMA(KP707106781, T3F, T3y); T3G = FNMS(KP707106781, T3F, T3y); T7N = T3L + T3K; T3M = T3K - T3L; T3J = T3H - T3I; T7P = T3H + T3I; T9k = FNMS(KP707106781, T7N, T7M); T7O = FMA(KP707106781, T7N, T7M); T9l = FNMS(KP707106781, T7Q, T7P); T7R = FMA(KP707106781, T7Q, T7P); } } } { E T5I, T1z, Tb8, T56, T53, T1C, Tb9, T5L, T1J, Tbq, T58, T1G, T5N, T5h, Tbp; E T5b; { E T54, T55, T1A, T1B, T1x, T1y, T5J, T5K; T1x = ri[WS(is, 63)]; T1y = ri[WS(is, 31)]; T54 = ii[WS(is, 63)]; T6H = FMA(KP707106781, T3M, T3J); T3N = FNMS(KP707106781, T3M, T3J); T5I = T1x - T1y; T1z = T1x + T1y; T55 = ii[WS(is, 31)]; T1A = ri[WS(is, 15)]; T1B = ri[WS(is, 47)]; T5J = ii[WS(is, 15)]; Tb8 = T54 + T55; T56 = T54 - T55; T53 = T1A - T1B; T1C = T1A + T1B; T5K = ii[WS(is, 47)]; { E T5e, T5d, T5f, T1H, T1I; T1H = ri[WS(is, 55)]; T1I = ri[WS(is, 23)]; T5e = ii[WS(is, 55)]; Tb9 = T5J + T5K; T5L = T5J - T5K; T5d = T1H - T1I; T1J = T1H + T1I; T5f = ii[WS(is, 23)]; { E T59, T5a, T1E, T1F, T5g; T1E = ri[WS(is, 7)]; T1F = ri[WS(is, 39)]; T5g = T5e - T5f; Tbq = T5e + T5f; T59 = ii[WS(is, 7)]; T58 = T1E - T1F; T1G = T1E + T1F; T5a = ii[WS(is, 39)]; T5N = T5d + T5g; T5h = T5d - T5g; Tbp = T59 + T5a; T5b = T59 - T5a; } } } { E Tb7, T5O, Tba, T57, T5i, T8x, T8w, T5M, T5P; { E Tbo, T5c, Tbr, Tdw, T1D, T1K, Tdv; Tbo = T1z - T1C; T1D = T1z + T1C; T1K = T1G + T1J; Tb7 = T1J - T1G; T5c = T58 + T5b; T5O = T5b - T58; TdA = T1D - T1K; T1L = T1D + T1K; Tbr = Tbp - Tbq; Tdw = Tbp + Tbq; Tba = Tb8 - Tb9; Tdv = Tb8 + Tb9; T8l = T56 - T53; T57 = T53 + T56; Tct = Tbo - Tbr; Tbs = Tbo + Tbr; Teo = Tdv + Tdw; Tdx = Tdv - Tdw; T5i = T5c - T5h; T8x = T5c + T5h; } T8w = T5I + T5L; T5M = T5I - T5L; T5P = T5N - T5O; T8m = T5O + T5N; T6Y = FMA(KP707106781, T5i, T57); T5j = FNMS(KP707106781, T5i, T57); T6V = FMA(KP707106781, T5P, T5M); T5Q = FNMS(KP707106781, T5P, T5M); T9z = FNMS(KP707106781, T8x, T8w); T8y = FMA(KP707106781, T8x, T8w); Tcw = Tba - Tb7; Tbb = Tb7 + Tba; } } } { E T82, T83, T45, T42, T87, T8a; { E T40, TN, T3Q, T2Q, T3P, TQ, T41, T2T, TX, T30, T3S, TU, T3X, T43, T2X; E T3T; { E T2O, T2P, TO, TP, TL, TM, T2R, T2S; TL = ri[WS(is, 62)]; TM = ri[WS(is, 30)]; T2O = ii[WS(is, 62)]; T9C = FNMS(KP707106781, T8m, T8l); T8n = FMA(KP707106781, T8m, T8l); T40 = TL - TM; TN = TL + TM; T2P = ii[WS(is, 30)]; TO = ri[WS(is, 14)]; TP = ri[WS(is, 46)]; T2R = ii[WS(is, 14)]; T3Q = T2O - T2P; T2Q = T2O + T2P; T3P = TO - TP; TQ = TO + TP; T2S = ii[WS(is, 46)]; { E T2Y, T3V, T2Z, TV, TW; TV = ri[WS(is, 54)]; TW = ri[WS(is, 22)]; T2Y = ii[WS(is, 54)]; T41 = T2R - T2S; T2T = T2R + T2S; T3V = TV - TW; TX = TV + TW; T2Z = ii[WS(is, 22)]; { E T2V, T2W, TS, TT, T3W; TS = ri[WS(is, 6)]; TT = ri[WS(is, 38)]; T3W = T2Y - T2Z; T30 = T2Y + T2Z; T2V = ii[WS(is, 6)]; T3S = TS - TT; TU = TS + TT; T2W = ii[WS(is, 38)]; T3X = T3V - T3W; T43 = T3V + T3W; T2X = T2V + T2W; T3T = T2V - T2W; } } } { E T44, T3U, T2U, T31; { E TaA, Tax, Tay, TR, TY, TaB; TaA = TN - TQ; TR = TN + TQ; TY = TU + TX; Tax = TX - TU; T44 = T3T - T3S; T3U = T3S + T3T; Tdf = TR - TY; TZ = TR + TY; Tay = T2Q - T2T; T2U = T2Q + T2T; TaB = T2X - T30; T31 = T2X + T30; Tch = Tay - Tax; Taz = Tax + Tay; Tci = TaA - TaB; TaC = TaA + TaB; } { E T7T, T7X, T7U, T3R, T3Y, T7W; T7T = T3Q - T3P; T3R = T3P + T3Q; T3Y = T3U - T3X; T7X = T3U + T3X; Tdg = T2U - T31; T32 = T2U + T31; T6J = FMA(KP707106781, T3Y, T3R); T3Z = FNMS(KP707106781, T3Y, T3R); T7U = T44 + T43; T45 = T43 - T44; T42 = T40 - T41; T7W = T40 + T41; T9n = FNMS(KP707106781, T7U, T7T); T7V = FMA(KP707106781, T7U, T7T); T9o = FNMS(KP707106781, T7X, T7W); T7Y = FMA(KP707106781, T7X, T7W); } } } { E T4P, T14, TaH, T4d, T4a, T17, TaI, T4S, T1e, TaZ, T4f, T1b, T4U, T4o, TaY; E T4i; { E T4b, T4c, T15, T16, T12, T13, T4Q, T4R; T12 = ri[WS(is, 1)]; T13 = ri[WS(is, 33)]; T4b = ii[WS(is, 1)]; T6K = FMA(KP707106781, T45, T42); T46 = FNMS(KP707106781, T45, T42); T4P = T12 - T13; T14 = T12 + T13; T4c = ii[WS(is, 33)]; T15 = ri[WS(is, 17)]; T16 = ri[WS(is, 49)]; T4Q = ii[WS(is, 17)]; TaH = T4b + T4c; T4d = T4b - T4c; T4a = T15 - T16; T17 = T15 + T16; T4R = ii[WS(is, 49)]; { E T4l, T4k, T4m, T1c, T1d; T1c = ri[WS(is, 57)]; T1d = ri[WS(is, 25)]; T4l = ii[WS(is, 57)]; TaI = T4Q + T4R; T4S = T4Q - T4R; T4k = T1c - T1d; T1e = T1c + T1d; T4m = ii[WS(is, 25)]; { E T4g, T4h, T19, T1a, T4n; T19 = ri[WS(is, 9)]; T1a = ri[WS(is, 41)]; T4n = T4l - T4m; TaZ = T4l + T4m; T4g = ii[WS(is, 9)]; T4f = T19 - T1a; T1b = T19 + T1a; T4h = ii[WS(is, 41)]; T4U = T4k + T4n; T4o = T4k - T4n; TaY = T4g + T4h; T4i = T4g - T4h; } } } { E TaG, T4V, TaJ, T4e, T4p, T8e, T8d, T4T, T4W; { E TaX, T4j, Tb0, Tdl, T18, T1f, Tdk; TaX = T14 - T17; T18 = T14 + T17; T1f = T1b + T1e; TaG = T1e - T1b; T4j = T4f + T4i; T4V = T4i - T4f; Tdp = T18 - T1f; T1g = T18 + T1f; Tb0 = TaY - TaZ; Tdl = TaY + TaZ; TaJ = TaH - TaI; Tdk = TaH + TaI; T82 = T4d - T4a; T4e = T4a + T4d; Tcm = TaX - Tb0; Tb1 = TaX + Tb0; Tej = Tdk + Tdl; Tdm = Tdk - Tdl; T4p = T4j - T4o; T8e = T4j + T4o; } T8d = T4P + T4S; T4T = T4P - T4S; T4W = T4U - T4V; T83 = T4V + T4U; T6R = FMA(KP707106781, T4p, T4e); T4q = FNMS(KP707106781, T4p, T4e); T6O = FMA(KP707106781, T4W, T4T); T4X = FNMS(KP707106781, T4W, T4T); T9s = FNMS(KP707106781, T8e, T8d); T8f = FMA(KP707106781, T8e, T8d); Tcp = TaJ - TaG; TaK = TaG + TaJ; } } { E T85, T4L, TaO, T1n, Tdq, TaN, T86, T4G, T4r, T1q, T4s, TaR, T4z, T4w, T1t; E T4t; { E T4C, T1j, T4D, TaL, T4K, T4H, T1m, T4E; { E T4I, T4J, T1h, T1i, T1k, T1l; T1h = ri[WS(is, 5)]; T1i = ri[WS(is, 37)]; T4I = ii[WS(is, 5)]; T9v = FNMS(KP707106781, T83, T82); T84 = FMA(KP707106781, T83, T82); T4C = T1h - T1i; T1j = T1h + T1i; T4J = ii[WS(is, 37)]; T1k = ri[WS(is, 21)]; T1l = ri[WS(is, 53)]; T4D = ii[WS(is, 21)]; TaL = T4I + T4J; T4K = T4I - T4J; T4H = T1k - T1l; T1m = T1k + T1l; T4E = ii[WS(is, 53)]; } { E T4x, T4y, T1r, T1s; { E T1o, T4F, TaM, T1p; T1o = ri[WS(is, 61)]; T85 = T4K - T4H; T4L = T4H + T4K; TaO = T1j - T1m; T1n = T1j + T1m; T4F = T4D - T4E; TaM = T4D + T4E; T1p = ri[WS(is, 29)]; T4x = ii[WS(is, 61)]; Tdq = TaL + TaM; TaN = TaL - TaM; T86 = T4C + T4F; T4G = T4C - T4F; T4r = T1o - T1p; T1q = T1o + T1p; T4y = ii[WS(is, 29)]; } T1r = ri[WS(is, 13)]; T1s = ri[WS(is, 45)]; T4s = ii[WS(is, 13)]; TaR = T4x + T4y; T4z = T4x - T4y; T4w = T1r - T1s; T1t = T1r + T1s; T4t = ii[WS(is, 45)]; } } { E T88, TaP, T89, TaU, T4Z, T4B, T4M, T4Y, T8g, T8h; { E T4A, Tb2, Tdr, T4v, Tb3; { E TaQ, T1u, T4u, TaS, TaT; T88 = T4z - T4w; T4A = T4w + T4z; TaQ = T1q - T1t; T1u = T1q + T1t; T4u = T4s - T4t; TaS = T4s + T4t; Tb2 = TaO + TaN; TaP = TaN - TaO; Tdr = TaR + TaS; TaT = TaR - TaS; T89 = T4r + T4u; T4v = T4r - T4u; Tdn = T1u - T1n; T1v = T1n + T1u; Tb3 = TaQ - TaT; TaU = TaQ + TaT; } T4Z = FNMS(KP414213562, T4v, T4A); T4B = FMA(KP414213562, T4A, T4v); Tcq = Tb2 - Tb3; Tb4 = Tb2 + Tb3; Tek = Tdq + Tdr; Tds = Tdq - Tdr; T4M = FNMS(KP414213562, T4L, T4G); T4Y = FMA(KP414213562, T4G, T4L); } T87 = FNMS(KP414213562, T86, T85); T8g = FMA(KP414213562, T85, T86); T6P = T4M + T4B; T4N = T4B - T4M; T6S = T4Y + T4Z; T50 = T4Y - T4Z; T8h = FNMS(KP414213562, T88, T89); T8a = FMA(KP414213562, T89, T88); T9w = T8g - T8h; T8i = T8g + T8h; Tcn = TaU - TaP; TaV = TaP + TaU; } } { E T8o, T5E, Tbf, T1S, TdB, Tbe, T8p, T5z, T5k, T1V, T5l, Tbi, T5s, T5p, T1Y; E T5m; { E T5v, T1O, T5w, Tbc, T5D, T5A, T1R, T5x; { E T5B, T5C, T1M, T1N, T1P, T1Q; T1M = ri[WS(is, 3)]; T1N = ri[WS(is, 35)]; T5B = ii[WS(is, 3)]; T9t = T8a - T87; T8b = T87 + T8a; T5v = T1M - T1N; T1O = T1M + T1N; T5C = ii[WS(is, 35)]; T1P = ri[WS(is, 19)]; T1Q = ri[WS(is, 51)]; T5w = ii[WS(is, 19)]; Tbc = T5B + T5C; T5D = T5B - T5C; T5A = T1P - T1Q; T1R = T1P + T1Q; T5x = ii[WS(is, 51)]; } { E T5q, T5r, T1W, T1X; { E T1T, T5y, Tbd, T1U; T1T = ri[WS(is, 59)]; T8o = T5D - T5A; T5E = T5A + T5D; Tbf = T1O - T1R; T1S = T1O + T1R; T5y = T5w - T5x; Tbd = T5w + T5x; T1U = ri[WS(is, 27)]; T5q = ii[WS(is, 59)]; TdB = Tbc + Tbd; Tbe = Tbc - Tbd; T8p = T5v + T5y; T5z = T5v - T5y; T5k = T1T - T1U; T1V = T1T + T1U; T5r = ii[WS(is, 27)]; } T1W = ri[WS(is, 11)]; T1X = ri[WS(is, 43)]; T5l = ii[WS(is, 11)]; Tbi = T5q + T5r; T5s = T5q - T5r; T5p = T1W - T1X; T1Y = T1W + T1X; T5m = ii[WS(is, 43)]; } } { E T8r, Tbg, T8s, Tbl, T5S, T5u, T5F, T5R, T8z, T8A; { E T5t, Tbt, TdC, T5o, Tbu; { E Tbh, T1Z, T5n, Tbj, Tbk; T8r = T5s - T5p; T5t = T5p + T5s; Tbh = T1V - T1Y; T1Z = T1V + T1Y; T5n = T5l - T5m; Tbj = T5l + T5m; Tbt = Tbf + Tbe; Tbg = Tbe - Tbf; TdC = Tbi + Tbj; Tbk = Tbi - Tbj; T8s = T5k + T5n; T5o = T5k - T5n; Tdy = T1Z - T1S; T20 = T1S + T1Z; Tbu = Tbh - Tbk; Tbl = Tbh + Tbk; } T5S = FNMS(KP414213562, T5o, T5t); T5u = FMA(KP414213562, T5t, T5o); Tcx = Tbt - Tbu; Tbv = Tbt + Tbu; Tep = TdB + TdC; TdD = TdB - TdC; T5F = FNMS(KP414213562, T5E, T5z); T5R = FMA(KP414213562, T5z, T5E); } T8q = FNMS(KP414213562, T8p, T8o); T8z = FMA(KP414213562, T8o, T8p); T6W = T5F + T5u; T5G = T5u - T5F; T6Z = T5R + T5S; T5T = T5R - T5S; T8A = FNMS(KP414213562, T8r, T8s); T8t = FMA(KP414213562, T8s, T8r); T9D = T8z - T8A; T8B = T8z + T8A; Tcu = Tbl - Tbg; Tbm = Tbg + Tbl; } } } { E T9A, T8u, TbE, Tao, Td7, Td8; { E Teq, Ten, Tex, Teh, TeB, Tev, Tey, Tem, Te9, Tec; { E Tef, Teu, Tel, T11, Tei, Tet, T2y, TeI, T23, T22, T33, Teg, TeD, TeG, T34; E TeH; { E TeE, TeF, Tv, T10, T1w, T21; Tef = Tf - Tu; Tv = Tf + Tu; T10 = TK + TZ; Teu = TZ - TK; Tel = Tej - Tek; TeE = Tej + Tek; T9A = T8t - T8q; T8u = T8q + T8t; TeD = Tv - T10; T11 = Tv + T10; TeF = Teo + Tep; Teq = Teo - Tep; Tei = T1g - T1v; T1w = T1g + T1v; T21 = T1L + T20; Ten = T1L - T20; Tet = T2i - T2x; T2y = T2i + T2x; TeI = TeE + TeF; TeG = TeE - TeF; T23 = T21 - T1w; T22 = T1w + T21; T33 = T2N + T32; Teg = T2N - T32; } ro[WS(os, 16)] = TeD + TeG; ro[WS(os, 48)] = TeD - TeG; ro[0] = T11 + T22; ro[WS(os, 32)] = T11 - T22; T34 = T2y - T33; TeH = T2y + T33; io[0] = TeH + TeI; io[WS(os, 32)] = TeH - TeI; io[WS(os, 48)] = T34 - T23; io[WS(os, 16)] = T23 + T34; Tex = Tef - Teg; Teh = Tef + Teg; TeB = Teu + Tet; Tev = Tet - Teu; Tey = Tel - Tei; Tem = Tei + Tel; } { E TdV, Tdb, TdJ, Te5, TdE, Tdz, Te6, Tdi, Teb, Te3, TdZ, TdY, TdW, TdM, TdR; E Tdu; { E TdL, Tde, Tdh, TdK, Tez, Ter; TdV = Td9 + Tda; Tdb = Td9 - Tda; TdJ = TdH - TdI; Te5 = TdI + TdH; Tez = Ten + Teq; Ter = Ten - Teq; TdL = Tdd + Tdc; Tde = Tdc - Tdd; { E TeA, TeC, Tew, Tes; TeA = Tey - Tez; TeC = Tey + Tez; Tew = Ter - Tem; Tes = Tem + Ter; ro[WS(os, 24)] = FMA(KP707106781, TeA, Tex); ro[WS(os, 56)] = FNMS(KP707106781, TeA, Tex); io[WS(os, 8)] = FMA(KP707106781, TeC, TeB); io[WS(os, 40)] = FNMS(KP707106781, TeC, TeB); io[WS(os, 24)] = FMA(KP707106781, Tew, Tev); io[WS(os, 56)] = FNMS(KP707106781, Tew, Tev); ro[WS(os, 8)] = FMA(KP707106781, Tes, Teh); ro[WS(os, 40)] = FNMS(KP707106781, Tes, Teh); Tdh = Tdf + Tdg; TdK = Tdf - Tdg; } { E Te1, Te2, Tdo, Tdt; TdE = TdA - TdD; Te1 = TdA + TdD; Te2 = Tdy + Tdx; Tdz = Tdx - Tdy; Te6 = Tde + Tdh; Tdi = Tde - Tdh; Teb = FMA(KP414213562, Te1, Te2); Te3 = FNMS(KP414213562, Te2, Te1); TdZ = Tdn + Tdm; Tdo = Tdm - Tdn; Tdt = Tdp - Tds; TdY = Tdp + Tds; TdW = TdL + TdK; TdM = TdK - TdL; TdR = FNMS(KP414213562, Tdo, Tdt); Tdu = FMA(KP414213562, Tdt, Tdo); } } { E TdT, Tea, Te0, TdU; { E Tdj, TdQ, TdF, TdP, TdN, TdS, TdO, TdG; TdT = FNMS(KP707106781, Tdi, Tdb); Tdj = FMA(KP707106781, Tdi, Tdb); Tea = FNMS(KP414213562, TdY, TdZ); Te0 = FMA(KP414213562, TdZ, TdY); TdQ = FMA(KP414213562, Tdz, TdE); TdF = FNMS(KP414213562, TdE, Tdz); TdP = FMA(KP707106781, TdM, TdJ); TdN = FNMS(KP707106781, TdM, TdJ); TdS = TdQ - TdR; TdU = TdR + TdQ; TdO = Tdu + TdF; TdG = Tdu - TdF; io[WS(os, 12)] = FMA(KP923879532, TdS, TdP); io[WS(os, 44)] = FNMS(KP923879532, TdS, TdP); ro[WS(os, 12)] = FMA(KP923879532, TdG, Tdj); ro[WS(os, 44)] = FNMS(KP923879532, TdG, Tdj); io[WS(os, 60)] = FMA(KP923879532, TdO, TdN); io[WS(os, 28)] = FNMS(KP923879532, TdO, TdN); } { E Te8, Te7, Ted, Tee, TdX, Te4; Te9 = FNMS(KP707106781, TdW, TdV); TdX = FMA(KP707106781, TdW, TdV); Te4 = Te0 + Te3; Te8 = Te3 - Te0; Te7 = FNMS(KP707106781, Te6, Te5); Ted = FMA(KP707106781, Te6, Te5); ro[WS(os, 60)] = FMA(KP923879532, TdU, TdT); ro[WS(os, 28)] = FNMS(KP923879532, TdU, TdT); ro[WS(os, 4)] = FMA(KP923879532, Te4, TdX); ro[WS(os, 36)] = FNMS(KP923879532, Te4, TdX); Tee = Tea + Teb; Tec = Tea - Teb; io[WS(os, 4)] = FMA(KP923879532, Tee, Ted); io[WS(os, 36)] = FNMS(KP923879532, Tee, Ted); io[WS(os, 20)] = FMA(KP923879532, Te8, Te7); io[WS(os, 52)] = FNMS(KP923879532, Te8, Te7); } } } { E TcP, Tcd, TcZ, TcD, Tcy, Tcv, TcT, Td0, Tck, Td4, TcX, TcS, TcK, Tcs, TcQ; E TcG; { E TcF, Tcg, Tcj, TcE, TcV, TcW, Tcc, TcC, Tco, Tcr; TbE = TbC + TbD; Tcc = TbC - TbD; TcC = Tan - Tak; Tao = Tak + Tan; TcF = FNMS(KP414213562, Tce, Tcf); Tcg = FMA(KP414213562, Tcf, Tce); ro[WS(os, 20)] = FMA(KP923879532, Tec, Te9); ro[WS(os, 52)] = FNMS(KP923879532, Tec, Te9); TcP = FNMS(KP707106781, Tcc, Tcb); Tcd = FMA(KP707106781, Tcc, Tcb); TcZ = FNMS(KP707106781, TcC, TcB); TcD = FMA(KP707106781, TcC, TcB); Tcj = FNMS(KP414213562, Tci, Tch); TcE = FMA(KP414213562, Tch, Tci); Tcy = FNMS(KP707106781, Tcx, Tcw); TcV = FMA(KP707106781, Tcx, Tcw); TcW = FMA(KP707106781, Tcu, Tct); Tcv = FNMS(KP707106781, Tcu, Tct); TcT = FMA(KP707106781, Tcn, Tcm); Tco = FNMS(KP707106781, Tcn, Tcm); Td0 = Tcg + Tcj; Tck = Tcg - Tcj; Td4 = FMA(KP198912367, TcV, TcW); TcX = FNMS(KP198912367, TcW, TcV); Tcr = FNMS(KP707106781, Tcq, Tcp); TcS = FMA(KP707106781, Tcq, Tcp); TcK = FNMS(KP668178637, Tco, Tcr); Tcs = FMA(KP668178637, Tcr, Tco); TcQ = TcF + TcE; TcG = TcE - TcF; } { E TcJ, Td5, TcU, TcM; { E Tcl, TcL, Tcz, TcN, TcH, TcO, TcI, TcA; TcJ = FNMS(KP923879532, Tck, Tcd); Tcl = FMA(KP923879532, Tck, Tcd); Td5 = FNMS(KP198912367, TcS, TcT); TcU = FMA(KP198912367, TcT, TcS); TcL = FMA(KP668178637, Tcv, Tcy); Tcz = FNMS(KP668178637, Tcy, Tcv); TcN = FMA(KP923879532, TcG, TcD); TcH = FNMS(KP923879532, TcG, TcD); TcO = TcK + TcL; TcM = TcK - TcL; TcI = Tcz - Tcs; TcA = Tcs + Tcz; io[WS(os, 6)] = FMA(KP831469612, TcO, TcN); io[WS(os, 38)] = FNMS(KP831469612, TcO, TcN); ro[WS(os, 6)] = FMA(KP831469612, TcA, Tcl); ro[WS(os, 38)] = FNMS(KP831469612, TcA, Tcl); io[WS(os, 22)] = FMA(KP831469612, TcI, TcH); io[WS(os, 54)] = FNMS(KP831469612, TcI, TcH); } { E Td2, Td1, Td3, Td6, TcR, TcY; Td7 = FMA(KP923879532, TcQ, TcP); TcR = FNMS(KP923879532, TcQ, TcP); TcY = TcU - TcX; Td2 = TcU + TcX; Td1 = FMA(KP923879532, Td0, TcZ); Td3 = FNMS(KP923879532, Td0, TcZ); ro[WS(os, 22)] = FMA(KP831469612, TcM, TcJ); ro[WS(os, 54)] = FNMS(KP831469612, TcM, TcJ); ro[WS(os, 14)] = FMA(KP980785280, TcY, TcR); ro[WS(os, 46)] = FNMS(KP980785280, TcY, TcR); Td6 = Td4 - Td5; Td8 = Td5 + Td4; io[WS(os, 14)] = FMA(KP980785280, Td6, Td3); io[WS(os, 46)] = FNMS(KP980785280, Td6, Td3); io[WS(os, 62)] = FMA(KP980785280, Td2, Td1); io[WS(os, 30)] = FNMS(KP980785280, Td2, Td1); } } } } { E T3f, T66, T63, T3u, T7z, T7A, Tc5, Tc8; { E TbR, Tap, Tc1, TbF, Tbw, Tbn, TbV, Tc2, TaE, Tc7, TbZ, TbU, TbN, Tb6, TbS; E TbI; { E TbH, Taw, TaD, TbG, TbX, TbY, TaW, Tb5; TbH = FMA(KP414213562, Tas, Tav); Taw = FNMS(KP414213562, Tav, Tas); ro[WS(os, 62)] = FMA(KP980785280, Td8, Td7); ro[WS(os, 30)] = FNMS(KP980785280, Td8, Td7); TbR = FMA(KP707106781, Tao, Tah); Tap = FNMS(KP707106781, Tao, Tah); Tc1 = FMA(KP707106781, TbE, TbB); TbF = FNMS(KP707106781, TbE, TbB); TaD = FMA(KP414213562, TaC, Taz); TbG = FNMS(KP414213562, Taz, TaC); Tbw = FNMS(KP707106781, Tbv, Tbs); TbX = FMA(KP707106781, Tbv, Tbs); TbY = FMA(KP707106781, Tbm, Tbb); Tbn = FNMS(KP707106781, Tbm, Tbb); TbV = FMA(KP707106781, TaV, TaK); TaW = FNMS(KP707106781, TaV, TaK); Tc2 = Taw + TaD; TaE = Taw - TaD; Tc7 = FMA(KP198912367, TbX, TbY); TbZ = FNMS(KP198912367, TbY, TbX); Tb5 = FNMS(KP707106781, Tb4, Tb1); TbU = FMA(KP707106781, Tb4, Tb1); TbN = FNMS(KP668178637, TaW, Tb5); Tb6 = FMA(KP668178637, Tb5, TaW); TbS = TbH + TbG; TbI = TbG - TbH; } { E TbP, Tc6, TbW, TbQ; { E TaF, TbM, Tbx, TbL, TbJ, TbO, TbK, Tby; TbP = FNMS(KP923879532, TaE, Tap); TaF = FMA(KP923879532, TaE, Tap); Tc6 = FNMS(KP198912367, TbU, TbV); TbW = FMA(KP198912367, TbV, TbU); TbM = FMA(KP668178637, Tbn, Tbw); Tbx = FNMS(KP668178637, Tbw, Tbn); TbL = FMA(KP923879532, TbI, TbF); TbJ = FNMS(KP923879532, TbI, TbF); TbO = TbM - TbN; TbQ = TbN + TbM; TbK = Tb6 + Tbx; Tby = Tb6 - Tbx; io[WS(os, 10)] = FMA(KP831469612, TbO, TbL); io[WS(os, 42)] = FNMS(KP831469612, TbO, TbL); ro[WS(os, 10)] = FMA(KP831469612, Tby, TaF); ro[WS(os, 42)] = FNMS(KP831469612, Tby, TaF); io[WS(os, 58)] = FMA(KP831469612, TbK, TbJ); io[WS(os, 26)] = FNMS(KP831469612, TbK, TbJ); } { E Tc4, Tc3, Tc9, Tca, TbT, Tc0; Tc5 = FNMS(KP923879532, TbS, TbR); TbT = FMA(KP923879532, TbS, TbR); Tc0 = TbW + TbZ; Tc4 = TbZ - TbW; Tc3 = FNMS(KP923879532, Tc2, Tc1); Tc9 = FMA(KP923879532, Tc2, Tc1); ro[WS(os, 58)] = FMA(KP831469612, TbQ, TbP); ro[WS(os, 26)] = FNMS(KP831469612, TbQ, TbP); ro[WS(os, 2)] = FMA(KP980785280, Tc0, TbT); ro[WS(os, 34)] = FNMS(KP980785280, Tc0, TbT); Tca = Tc6 + Tc7; Tc8 = Tc6 - Tc7; io[WS(os, 2)] = FMA(KP980785280, Tca, Tc9); io[WS(os, 34)] = FNMS(KP980785280, Tca, Tc9); io[WS(os, 18)] = FMA(KP980785280, Tc4, Tc3); io[WS(os, 50)] = FNMS(KP980785280, Tc4, Tc3); } } } { E T7h, T6F, T70, T6X, T7x, T7m, T7w, T7p, T7s, T6M, T7c, T6U, T7r, T75, T7i; E T78; { E T6T, T6Q, T77, T6I, T6L, T76, T73, T74; { E T7k, T7l, T6D, T6E, T7n, T7o; T3f = FMA(KP707106781, T3e, T37); T6D = FNMS(KP707106781, T3e, T37); T6E = T65 + T64; T66 = T64 - T65; T6T = FNMS(KP923879532, T6S, T6R); T7k = FMA(KP923879532, T6S, T6R); ro[WS(os, 18)] = FMA(KP980785280, Tc8, Tc5); ro[WS(os, 50)] = FNMS(KP980785280, Tc8, Tc5); T7h = FMA(KP923879532, T6E, T6D); T6F = FNMS(KP923879532, T6E, T6D); T7l = FMA(KP923879532, T6P, T6O); T6Q = FNMS(KP923879532, T6P, T6O); T70 = FNMS(KP923879532, T6Z, T6Y); T7n = FMA(KP923879532, T6Z, T6Y); T7o = FMA(KP923879532, T6W, T6V); T6X = FNMS(KP923879532, T6W, T6V); T77 = FNMS(KP198912367, T6G, T6H); T6I = FMA(KP198912367, T6H, T6G); T7x = FNMS(KP098491403, T7k, T7l); T7m = FMA(KP098491403, T7l, T7k); T7w = FMA(KP098491403, T7n, T7o); T7p = FNMS(KP098491403, T7o, T7n); T6L = FNMS(KP198912367, T6K, T6J); T76 = FMA(KP198912367, T6J, T6K); } T63 = FMA(KP707106781, T62, T5Z); T73 = FNMS(KP707106781, T62, T5Z); T7s = T6I + T6L; T6M = T6I - T6L; T7c = FNMS(KP820678790, T6Q, T6T); T6U = FMA(KP820678790, T6T, T6Q); T74 = T3m + T3t; T3u = T3m - T3t; T7r = FMA(KP923879532, T74, T73); T75 = FNMS(KP923879532, T74, T73); T7i = T77 + T76; T78 = T76 - T77; } { E T7b, T6N, T7f, T79, T71, T7d; T7b = FNMS(KP980785280, T6M, T6F); T6N = FMA(KP980785280, T6M, T6F); T7f = FMA(KP980785280, T78, T75); T79 = FNMS(KP980785280, T78, T75); T71 = FNMS(KP820678790, T70, T6X); T7d = FMA(KP820678790, T6X, T70); { E T7u, T7t, T7v, T7y, T7j, T7q; T7z = FMA(KP980785280, T7i, T7h); T7j = FNMS(KP980785280, T7i, T7h); T7q = T7m - T7p; T7u = T7m + T7p; { E T7g, T7e, T72, T7a; T7g = T7c + T7d; T7e = T7c - T7d; T72 = T6U + T71; T7a = T71 - T6U; ro[WS(os, 23)] = FMA(KP773010453, T7e, T7b); ro[WS(os, 55)] = FNMS(KP773010453, T7e, T7b); io[WS(os, 7)] = FMA(KP773010453, T7g, T7f); io[WS(os, 39)] = FNMS(KP773010453, T7g, T7f); io[WS(os, 23)] = FMA(KP773010453, T7a, T79); io[WS(os, 55)] = FNMS(KP773010453, T7a, T79); ro[WS(os, 7)] = FMA(KP773010453, T72, T6N); ro[WS(os, 39)] = FNMS(KP773010453, T72, T6N); ro[WS(os, 47)] = FNMS(KP995184726, T7q, T7j); ro[WS(os, 15)] = FMA(KP995184726, T7q, T7j); } T7t = FMA(KP980785280, T7s, T7r); T7v = FNMS(KP980785280, T7s, T7r); T7y = T7w - T7x; T7A = T7x + T7w; io[WS(os, 15)] = FMA(KP995184726, T7y, T7v); io[WS(os, 47)] = FNMS(KP995184726, T7y, T7v); io[WS(os, 63)] = FMA(KP995184726, T7u, T7t); io[WS(os, 31)] = FNMS(KP995184726, T7u, T7t); } } } { E T7D, T8K, T8H, T7K, Tad, Tae, T6x, T6A; { E T9V, T9j, T9E, T9B, Tab, Ta0, Taa, Ta3, Ta6, T9q, T9Q, T9y, Ta5, T9J, T9W; E T9M; { E T9x, T9u, T9L, T9m, T9p, T9K, T9H, T9I; { E T9Y, T9Z, T9h, T9i, Ta1, Ta2; T7D = FMA(KP707106781, T7C, T7B); T9h = FNMS(KP707106781, T7C, T7B); T9i = T8I - T8J; T8K = T8I + T8J; T9x = FNMS(KP923879532, T9w, T9v); T9Y = FMA(KP923879532, T9w, T9v); ro[WS(os, 63)] = FMA(KP995184726, T7A, T7z); ro[WS(os, 31)] = FNMS(KP995184726, T7A, T7z); T9V = FNMS(KP923879532, T9i, T9h); T9j = FMA(KP923879532, T9i, T9h); T9Z = FMA(KP923879532, T9t, T9s); T9u = FNMS(KP923879532, T9t, T9s); T9E = FNMS(KP923879532, T9D, T9C); Ta1 = FMA(KP923879532, T9D, T9C); Ta2 = FMA(KP923879532, T9A, T9z); T9B = FNMS(KP923879532, T9A, T9z); T9L = FNMS(KP668178637, T9k, T9l); T9m = FMA(KP668178637, T9l, T9k); Tab = FNMS(KP303346683, T9Y, T9Z); Ta0 = FMA(KP303346683, T9Z, T9Y); Taa = FMA(KP303346683, Ta1, Ta2); Ta3 = FNMS(KP303346683, Ta2, Ta1); T9p = FNMS(KP668178637, T9o, T9n); T9K = FMA(KP668178637, T9n, T9o); } T8H = FMA(KP707106781, T8G, T8F); T9H = FNMS(KP707106781, T8G, T8F); Ta6 = T9m + T9p; T9q = T9m - T9p; T9Q = FNMS(KP534511135, T9u, T9x); T9y = FMA(KP534511135, T9x, T9u); T9I = T7J - T7G; T7K = T7G + T7J; Ta5 = FNMS(KP923879532, T9I, T9H); T9J = FMA(KP923879532, T9I, T9H); T9W = T9L + T9K; T9M = T9K - T9L; } { E T9P, T9r, T9T, T9N, T9F, T9R; T9P = FNMS(KP831469612, T9q, T9j); T9r = FMA(KP831469612, T9q, T9j); T9T = FMA(KP831469612, T9M, T9J); T9N = FNMS(KP831469612, T9M, T9J); T9F = FNMS(KP534511135, T9E, T9B); T9R = FMA(KP534511135, T9B, T9E); { E Ta8, Ta7, Ta9, Tac, T9X, Ta4; Tad = FMA(KP831469612, T9W, T9V); T9X = FNMS(KP831469612, T9W, T9V); Ta4 = Ta0 - Ta3; Ta8 = Ta0 + Ta3; { E T9U, T9S, T9G, T9O; T9U = T9Q + T9R; T9S = T9Q - T9R; T9G = T9y + T9F; T9O = T9F - T9y; ro[WS(os, 21)] = FMA(KP881921264, T9S, T9P); ro[WS(os, 53)] = FNMS(KP881921264, T9S, T9P); io[WS(os, 5)] = FMA(KP881921264, T9U, T9T); io[WS(os, 37)] = FNMS(KP881921264, T9U, T9T); io[WS(os, 21)] = FMA(KP881921264, T9O, T9N); io[WS(os, 53)] = FNMS(KP881921264, T9O, T9N); ro[WS(os, 5)] = FMA(KP881921264, T9G, T9r); ro[WS(os, 37)] = FNMS(KP881921264, T9G, T9r); ro[WS(os, 45)] = FNMS(KP956940335, Ta4, T9X); ro[WS(os, 13)] = FMA(KP956940335, Ta4, T9X); } Ta7 = FMA(KP831469612, Ta6, Ta5); Ta9 = FNMS(KP831469612, Ta6, Ta5); Tac = Taa - Tab; Tae = Tab + Taa; io[WS(os, 13)] = FMA(KP956940335, Tac, Ta9); io[WS(os, 45)] = FNMS(KP956940335, Tac, Ta9); io[WS(os, 61)] = FMA(KP956940335, Ta8, Ta7); io[WS(os, 29)] = FNMS(KP956940335, Ta8, Ta7); } } } { E T6j, T3v, T5U, T5H, T6y, T6o, T6z, T6r, T6u, T48, T6f, T52, T6t, T67, T6k; E T6a; { E T51, T4O, T69, T3O, T47, T68; { E T6m, T6n, T6p, T6q; T51 = FNMS(KP923879532, T50, T4X); T6m = FMA(KP923879532, T50, T4X); ro[WS(os, 61)] = FMA(KP956940335, Tae, Tad); ro[WS(os, 29)] = FNMS(KP956940335, Tae, Tad); T6j = FMA(KP923879532, T3u, T3f); T3v = FNMS(KP923879532, T3u, T3f); T6n = FMA(KP923879532, T4N, T4q); T4O = FNMS(KP923879532, T4N, T4q); T5U = FNMS(KP923879532, T5T, T5Q); T6p = FMA(KP923879532, T5T, T5Q); T6q = FMA(KP923879532, T5G, T5j); T5H = FNMS(KP923879532, T5G, T5j); T69 = FMA(KP668178637, T3G, T3N); T3O = FNMS(KP668178637, T3N, T3G); T6y = FNMS(KP303346683, T6m, T6n); T6o = FMA(KP303346683, T6n, T6m); T6z = FMA(KP303346683, T6p, T6q); T6r = FNMS(KP303346683, T6q, T6p); T47 = FMA(KP668178637, T46, T3Z); T68 = FNMS(KP668178637, T3Z, T46); } T6u = T3O + T47; T48 = T3O - T47; T6f = FNMS(KP534511135, T4O, T51); T52 = FMA(KP534511135, T51, T4O); T6t = FMA(KP923879532, T66, T63); T67 = FNMS(KP923879532, T66, T63); T6k = T69 + T68; T6a = T68 - T69; } { E T6h, T49, T6d, T6b, T5V, T6e; T6h = FNMS(KP831469612, T48, T3v); T49 = FMA(KP831469612, T48, T3v); T6d = FMA(KP831469612, T6a, T67); T6b = FNMS(KP831469612, T6a, T67); T5V = FNMS(KP534511135, T5U, T5H); T6e = FMA(KP534511135, T5H, T5U); { E T6w, T6v, T6B, T6C, T6l, T6s; T6x = FNMS(KP831469612, T6k, T6j); T6l = FMA(KP831469612, T6k, T6j); T6s = T6o + T6r; T6w = T6r - T6o; { E T6g, T6i, T5W, T6c; T6g = T6e - T6f; T6i = T6f + T6e; T5W = T52 - T5V; T6c = T52 + T5V; ro[WS(os, 59)] = FMA(KP881921264, T6i, T6h); ro[WS(os, 27)] = FNMS(KP881921264, T6i, T6h); io[WS(os, 11)] = FMA(KP881921264, T6g, T6d); io[WS(os, 43)] = FNMS(KP881921264, T6g, T6d); io[WS(os, 59)] = FMA(KP881921264, T6c, T6b); io[WS(os, 27)] = FNMS(KP881921264, T6c, T6b); ro[WS(os, 11)] = FMA(KP881921264, T5W, T49); ro[WS(os, 43)] = FNMS(KP881921264, T5W, T49); ro[WS(os, 35)] = FNMS(KP956940335, T6s, T6l); ro[WS(os, 3)] = FMA(KP956940335, T6s, T6l); } T6v = FNMS(KP831469612, T6u, T6t); T6B = FMA(KP831469612, T6u, T6t); T6C = T6y + T6z; T6A = T6y - T6z; io[WS(os, 3)] = FMA(KP956940335, T6C, T6B); io[WS(os, 35)] = FNMS(KP956940335, T6C, T6B); io[WS(os, 19)] = FMA(KP956940335, T6w, T6v); io[WS(os, 51)] = FNMS(KP956940335, T6w, T6v); } } } { E T8X, T7L, T8C, T8v, T9c, T92, T9d, T95, T98, T80, T8T, T8k, T97, T8L, T8Y; E T8O; { E T8j, T8c, T8N, T7S, T7Z, T8M; { E T90, T91, T93, T94; T8j = FNMS(KP923879532, T8i, T8f); T90 = FMA(KP923879532, T8i, T8f); ro[WS(os, 19)] = FMA(KP956940335, T6A, T6x); ro[WS(os, 51)] = FNMS(KP956940335, T6A, T6x); T8X = FMA(KP923879532, T7K, T7D); T7L = FNMS(KP923879532, T7K, T7D); T91 = FMA(KP923879532, T8b, T84); T8c = FNMS(KP923879532, T8b, T84); T8C = FNMS(KP923879532, T8B, T8y); T93 = FMA(KP923879532, T8B, T8y); T94 = FMA(KP923879532, T8u, T8n); T8v = FNMS(KP923879532, T8u, T8n); T8N = FMA(KP198912367, T7O, T7R); T7S = FNMS(KP198912367, T7R, T7O); T9c = FNMS(KP098491403, T90, T91); T92 = FMA(KP098491403, T91, T90); T9d = FMA(KP098491403, T93, T94); T95 = FNMS(KP098491403, T94, T93); T7Z = FMA(KP198912367, T7Y, T7V); T8M = FNMS(KP198912367, T7V, T7Y); } T98 = T7S + T7Z; T80 = T7S - T7Z; T8T = FNMS(KP820678790, T8c, T8j); T8k = FMA(KP820678790, T8j, T8c); T97 = FMA(KP923879532, T8K, T8H); T8L = FNMS(KP923879532, T8K, T8H); T8Y = T8N + T8M; T8O = T8M - T8N; } { E T8V, T81, T8R, T8P, T8D, T8S; T8V = FNMS(KP980785280, T80, T7L); T81 = FMA(KP980785280, T80, T7L); T8R = FMA(KP980785280, T8O, T8L); T8P = FNMS(KP980785280, T8O, T8L); T8D = FNMS(KP820678790, T8C, T8v); T8S = FMA(KP820678790, T8v, T8C); { E T9a, T99, T9f, T9g, T8Z, T96; T9b = FNMS(KP980785280, T8Y, T8X); T8Z = FMA(KP980785280, T8Y, T8X); T96 = T92 + T95; T9a = T95 - T92; { E T8U, T8W, T8E, T8Q; T8U = T8S - T8T; T8W = T8T + T8S; T8E = T8k - T8D; T8Q = T8k + T8D; ro[WS(os, 57)] = FMA(KP773010453, T8W, T8V); ro[WS(os, 25)] = FNMS(KP773010453, T8W, T8V); io[WS(os, 9)] = FMA(KP773010453, T8U, T8R); io[WS(os, 41)] = FNMS(KP773010453, T8U, T8R); io[WS(os, 57)] = FMA(KP773010453, T8Q, T8P); io[WS(os, 25)] = FNMS(KP773010453, T8Q, T8P); ro[WS(os, 9)] = FMA(KP773010453, T8E, T81); ro[WS(os, 41)] = FNMS(KP773010453, T8E, T81); ro[WS(os, 33)] = FNMS(KP995184726, T96, T8Z); ro[WS(os, 1)] = FMA(KP995184726, T96, T8Z); } T99 = FNMS(KP980785280, T98, T97); T9f = FMA(KP980785280, T98, T97); T9g = T9c + T9d; T9e = T9c - T9d; io[WS(os, 1)] = FMA(KP995184726, T9g, T9f); io[WS(os, 33)] = FNMS(KP995184726, T9g, T9f); io[WS(os, 17)] = FMA(KP995184726, T9a, T99); io[WS(os, 49)] = FNMS(KP995184726, T9a, T99); } } } } } } } ro[WS(os, 17)] = FMA(KP995184726, T9e, T9b); ro[WS(os, 49)] = FNMS(KP995184726, T9e, T9b); } } } static const kdft_desc desc = { 64, "n1_64", {520, 0, 392, 0}, &GENUS, 0, 0, 0, 0 }; void X(codelet_n1_64) (planner *p) { X(kdft_register) (p, n1_64, &desc); } #else /* HAVE_FMA */ /* Generated by: ../../../genfft/gen_notw.native -compact -variables 4 -pipeline-latency 4 -n 64 -name n1_64 -include n.h */ /* * This function contains 912 FP additions, 248 FP multiplications, * (or, 808 additions, 144 multiplications, 104 fused multiply/add), * 172 stack variables, 15 constants, and 256 memory accesses */ #include "n.h" static void n1_64(const R *ri, const R *ii, R *ro, R *io, stride is, stride os, INT v, INT ivs, INT ovs) { DK(KP773010453, +0.773010453362736960810906609758469800971041293); DK(KP634393284, +0.634393284163645498215171613225493370675687095); DK(KP098017140, +0.098017140329560601994195563888641845861136673); DK(KP995184726, +0.995184726672196886244836953109479921575474869); DK(KP881921264, +0.881921264348355029712756863660388349508442621); DK(KP471396736, +0.471396736825997648556387625905254377657460319); DK(KP290284677, +0.290284677254462367636192375817395274691476278); DK(KP956940335, +0.956940335732208864935797886980269969482849206); DK(KP831469612, +0.831469612302545237078788377617905756738560812); DK(KP555570233, +0.555570233019602224742830813948532874374937191); DK(KP195090322, +0.195090322016128267848284868477022240927691618); DK(KP980785280, +0.980785280403230449126182236134239036973933731); DK(KP923879532, +0.923879532511286756128183189396788286822416626); DK(KP382683432, +0.382683432365089771728459984030398866761344562); DK(KP707106781, +0.707106781186547524400844362104849039284835938); { INT i; for (i = v; i > 0; i = i - 1, ri = ri + ivs, ii = ii + ivs, ro = ro + ovs, io = io + ovs, MAKE_VOLATILE_STRIDE(256, is), MAKE_VOLATILE_STRIDE(256, os)) { E T37, T7B, T8F, T5Z, Tf, Td9, TbB, TcB, T62, T7C, T2i, TdH, Tah, Tcb, T3e; E T8G, Tu, TdI, Tak, TbD, Tan, TbC, T2x, Tda, T3m, T65, T7G, T8J, T7J, T8I; E T3t, T64, TK, Tdd, Tas, Tce, Tav, Tcf, T2N, Tdc, T3G, T6G, T7O, T9k, T7R; E T9l, T3N, T6H, T1L, Tdv, Tbs, Tcw, TdC, Teo, T5j, T6V, T5Q, T6Y, T8y, T9C; E Tbb, Tct, T8n, T9z, TZ, Tdf, Taz, Tch, TaC, Tci, T32, Tdg, T3Z, T6J, T7V; E T9n, T7Y, T9o, T46, T6K, T1g, Tdp, Tb1, Tcm, Tdm, Tej, T4q, T6R, T4X, T6O; E T8f, T9s, TaK, Tcp, T84, T9v, T1v, Tdn, Tb4, Tcq, Tds, Tek, T4N, T6P, T50; E T6S, T8i, T9w, TaV, Tcn, T8b, T9t, T20, TdD, Tbv, Tcu, Tdy, Tep, T5G, T6Z; E T5T, T6W, T8B, T9A, Tbm, Tcx, T8u, T9D; { E T3, T35, T26, T5Y, T6, T5X, T29, T36, Ta, T39, T2d, T38, Td, T3b, T2g; E T3c; { E T1, T2, T24, T25; T1 = ri[0]; T2 = ri[WS(is, 32)]; T3 = T1 + T2; T35 = T1 - T2; T24 = ii[0]; T25 = ii[WS(is, 32)]; T26 = T24 + T25; T5Y = T24 - T25; } { E T4, T5, T27, T28; T4 = ri[WS(is, 16)]; T5 = ri[WS(is, 48)]; T6 = T4 + T5; T5X = T4 - T5; T27 = ii[WS(is, 16)]; T28 = ii[WS(is, 48)]; T29 = T27 + T28; T36 = T27 - T28; } { E T8, T9, T2b, T2c; T8 = ri[WS(is, 8)]; T9 = ri[WS(is, 40)]; Ta = T8 + T9; T39 = T8 - T9; T2b = ii[WS(is, 8)]; T2c = ii[WS(is, 40)]; T2d = T2b + T2c; T38 = T2b - T2c; } { E Tb, Tc, T2e, T2f; Tb = ri[WS(is, 56)]; Tc = ri[WS(is, 24)]; Td = Tb + Tc; T3b = Tb - Tc; T2e = ii[WS(is, 56)]; T2f = ii[WS(is, 24)]; T2g = T2e + T2f; T3c = T2e - T2f; } { E T7, Te, T2a, T2h; T37 = T35 - T36; T7B = T35 + T36; T8F = T5Y - T5X; T5Z = T5X + T5Y; T7 = T3 + T6; Te = Ta + Td; Tf = T7 + Te; Td9 = T7 - Te; { E Tbz, TbA, T60, T61; Tbz = T26 - T29; TbA = Td - Ta; TbB = Tbz - TbA; TcB = TbA + Tbz; T60 = T3b - T3c; T61 = T39 + T38; T62 = KP707106781 * (T60 - T61); T7C = KP707106781 * (T61 + T60); } T2a = T26 + T29; T2h = T2d + T2g; T2i = T2a + T2h; TdH = T2a - T2h; { E Taf, Tag, T3a, T3d; Taf = T3 - T6; Tag = T2d - T2g; Tah = Taf - Tag; Tcb = Taf + Tag; T3a = T38 - T39; T3d = T3b + T3c; T3e = KP707106781 * (T3a - T3d); T8G = KP707106781 * (T3a + T3d); } } } { E Ti, T3j, T2l, T3h, Tl, T3g, T2o, T3k, Tp, T3q, T2s, T3o, Ts, T3n, T2v; E T3r; { E Tg, Th, T2j, T2k; Tg = ri[WS(is, 4)]; Th = ri[WS(is, 36)]; Ti = Tg + Th; T3j = Tg - Th; T2j = ii[WS(is, 4)]; T2k = ii[WS(is, 36)]; T2l = T2j + T2k; T3h = T2j - T2k; } { E Tj, Tk, T2m, T2n; Tj = ri[WS(is, 20)]; Tk = ri[WS(is, 52)]; Tl = Tj + Tk; T3g = Tj - Tk; T2m = ii[WS(is, 20)]; T2n = ii[WS(is, 52)]; T2o = T2m + T2n; T3k = T2m - T2n; } { E Tn, To, T2q, T2r; Tn = ri[WS(is, 60)]; To = ri[WS(is, 28)]; Tp = Tn + To; T3q = Tn - To; T2q = ii[WS(is, 60)]; T2r = ii[WS(is, 28)]; T2s = T2q + T2r; T3o = T2q - T2r; } { E Tq, Tr, T2t, T2u; Tq = ri[WS(is, 12)]; Tr = ri[WS(is, 44)]; Ts = Tq + Tr; T3n = Tq - Tr; T2t = ii[WS(is, 12)]; T2u = ii[WS(is, 44)]; T2v = T2t + T2u; T3r = T2t - T2u; } { E Tm, Tt, Tai, Taj; Tm = Ti + Tl; Tt = Tp + Ts; Tu = Tm + Tt; TdI = Tt - Tm; Tai = T2l - T2o; Taj = Ti - Tl; Tak = Tai - Taj; TbD = Taj + Tai; } { E Tal, Tam, T2p, T2w; Tal = Tp - Ts; Tam = T2s - T2v; Tan = Tal + Tam; TbC = Tal - Tam; T2p = T2l + T2o; T2w = T2s + T2v; T2x = T2p + T2w; Tda = T2p - T2w; } { E T3i, T3l, T7E, T7F; T3i = T3g + T3h; T3l = T3j - T3k; T3m = FNMS(KP923879532, T3l, KP382683432 * T3i); T65 = FMA(KP923879532, T3i, KP382683432 * T3l); T7E = T3h - T3g; T7F = T3j + T3k; T7G = FNMS(KP382683432, T7F, KP923879532 * T7E); T8J = FMA(KP382683432, T7E, KP923879532 * T7F); } { E T7H, T7I, T3p, T3s; T7H = T3o - T3n; T7I = T3q + T3r; T7J = FMA(KP923879532, T7H, KP382683432 * T7I); T8I = FNMS(KP382683432, T7H, KP923879532 * T7I); T3p = T3n + T3o; T3s = T3q - T3r; T3t = FMA(KP382683432, T3p, KP923879532 * T3s); T64 = FNMS(KP923879532, T3p, KP382683432 * T3s); } } { E Ty, T3H, T2B, T3x, TB, T3w, T2E, T3I, TI, T3L, T2L, T3B, TF, T3K, T2I; E T3E; { E Tw, Tx, T2C, T2D; Tw = ri[WS(is, 2)]; Tx = ri[WS(is, 34)]; Ty = Tw + Tx; T3H = Tw - Tx; { E T2z, T2A, Tz, TA; T2z = ii[WS(is, 2)]; T2A = ii[WS(is, 34)]; T2B = T2z + T2A; T3x = T2z - T2A; Tz = ri[WS(is, 18)]; TA = ri[WS(is, 50)]; TB = Tz + TA; T3w = Tz - TA; } T2C = ii[WS(is, 18)]; T2D = ii[WS(is, 50)]; T2E = T2C + T2D; T3I = T2C - T2D; { E TG, TH, T3z, T2J, T2K, T3A; TG = ri[WS(is, 58)]; TH = ri[WS(is, 26)]; T3z = TG - TH; T2J = ii[WS(is, 58)]; T2K = ii[WS(is, 26)]; T3A = T2J - T2K; TI = TG + TH; T3L = T3z + T3A; T2L = T2J + T2K; T3B = T3z - T3A; } { E TD, TE, T3C, T2G, T2H, T3D; TD = ri[WS(is, 10)]; TE = ri[WS(is, 42)]; T3C = TD - TE; T2G = ii[WS(is, 10)]; T2H = ii[WS(is, 42)]; T3D = T2G - T2H; TF = TD + TE; T3K = T3D - T3C; T2I = T2G + T2H; T3E = T3C + T3D; } } { E TC, TJ, Taq, Tar; TC = Ty + TB; TJ = TF + TI; TK = TC + TJ; Tdd = TC - TJ; Taq = T2B - T2E; Tar = TI - TF; Tas = Taq - Tar; Tce = Tar + Taq; } { E Tat, Tau, T2F, T2M; Tat = Ty - TB; Tau = T2I - T2L; Tav = Tat - Tau; Tcf = Tat + Tau; T2F = T2B + T2E; T2M = T2I + T2L; T2N = T2F + T2M; Tdc = T2F - T2M; } { E T3y, T3F, T7M, T7N; T3y = T3w + T3x; T3F = KP707106781 * (T3B - T3E); T3G = T3y - T3F; T6G = T3y + T3F; T7M = T3x - T3w; T7N = KP707106781 * (T3K + T3L); T7O = T7M - T7N; T9k = T7M + T7N; } { E T7P, T7Q, T3J, T3M; T7P = T3H + T3I; T7Q = KP707106781 * (T3E + T3B); T7R = T7P - T7Q; T9l = T7P + T7Q; T3J = T3H - T3I; T3M = KP707106781 * (T3K - T3L); T3N = T3J - T3M; T6H = T3J + T3M; } } { E T1z, T53, T5L, Tbo, T1C, T5I, T56, Tbp, T1J, Tb9, T5h, T5N, T1G, Tb8, T5c; E T5O; { E T1x, T1y, T54, T55; T1x = ri[WS(is, 63)]; T1y = ri[WS(is, 31)]; T1z = T1x + T1y; T53 = T1x - T1y; { E T5J, T5K, T1A, T1B; T5J = ii[WS(is, 63)]; T5K = ii[WS(is, 31)]; T5L = T5J - T5K; Tbo = T5J + T5K; T1A = ri[WS(is, 15)]; T1B = ri[WS(is, 47)]; T1C = T1A + T1B; T5I = T1A - T1B; } T54 = ii[WS(is, 15)]; T55 = ii[WS(is, 47)]; T56 = T54 - T55; Tbp = T54 + T55; { E T1H, T1I, T5d, T5e, T5f, T5g; T1H = ri[WS(is, 55)]; T1I = ri[WS(is, 23)]; T5d = T1H - T1I; T5e = ii[WS(is, 55)]; T5f = ii[WS(is, 23)]; T5g = T5e - T5f; T1J = T1H + T1I; Tb9 = T5e + T5f; T5h = T5d + T5g; T5N = T5d - T5g; } { E T1E, T1F, T5b, T58, T59, T5a; T1E = ri[WS(is, 7)]; T1F = ri[WS(is, 39)]; T5b = T1E - T1F; T58 = ii[WS(is, 7)]; T59 = ii[WS(is, 39)]; T5a = T58 - T59; T1G = T1E + T1F; Tb8 = T58 + T59; T5c = T5a - T5b; T5O = T5b + T5a; } } { E T1D, T1K, Tbq, Tbr; T1D = T1z + T1C; T1K = T1G + T1J; T1L = T1D + T1K; Tdv = T1D - T1K; Tbq = Tbo - Tbp; Tbr = T1J - T1G; Tbs = Tbq - Tbr; Tcw = Tbr + Tbq; } { E TdA, TdB, T57, T5i; TdA = Tbo + Tbp; TdB = Tb8 + Tb9; TdC = TdA - TdB; Teo = TdA + TdB; T57 = T53 - T56; T5i = KP707106781 * (T5c - T5h); T5j = T57 - T5i; T6V = T57 + T5i; } { E T5M, T5P, T8w, T8x; T5M = T5I + T5L; T5P = KP707106781 * (T5N - T5O); T5Q = T5M - T5P; T6Y = T5M + T5P; T8w = T5L - T5I; T8x = KP707106781 * (T5c + T5h); T8y = T8w - T8x; T9C = T8w + T8x; } { E Tb7, Tba, T8l, T8m; Tb7 = T1z - T1C; Tba = Tb8 - Tb9; Tbb = Tb7 - Tba; Tct = Tb7 + Tba; T8l = T53 + T56; T8m = KP707106781 * (T5O + T5N); T8n = T8l - T8m; T9z = T8l + T8m; } } { E TN, T40, T2Q, T3Q, TQ, T3P, T2T, T41, TX, T44, T30, T3U, TU, T43, T2X; E T3X; { E TL, TM, T2R, T2S; TL = ri[WS(is, 62)]; TM = ri[WS(is, 30)]; TN = TL + TM; T40 = TL - TM; { E T2O, T2P, TO, TP; T2O = ii[WS(is, 62)]; T2P = ii[WS(is, 30)]; T2Q = T2O + T2P; T3Q = T2O - T2P; TO = ri[WS(is, 14)]; TP = ri[WS(is, 46)]; TQ = TO + TP; T3P = TO - TP; } T2R = ii[WS(is, 14)]; T2S = ii[WS(is, 46)]; T2T = T2R + T2S; T41 = T2R - T2S; { E TV, TW, T3S, T2Y, T2Z, T3T; TV = ri[WS(is, 54)]; TW = ri[WS(is, 22)]; T3S = TV - TW; T2Y = ii[WS(is, 54)]; T2Z = ii[WS(is, 22)]; T3T = T2Y - T2Z; TX = TV + TW; T44 = T3S + T3T; T30 = T2Y + T2Z; T3U = T3S - T3T; } { E TS, TT, T3V, T2V, T2W, T3W; TS = ri[WS(is, 6)]; TT = ri[WS(is, 38)]; T3V = TS - TT; T2V = ii[WS(is, 6)]; T2W = ii[WS(is, 38)]; T3W = T2V - T2W; TU = TS + TT; T43 = T3W - T3V; T2X = T2V + T2W; T3X = T3V + T3W; } } { E TR, TY, Tax, Tay; TR = TN + TQ; TY = TU + TX; TZ = TR + TY; Tdf = TR - TY; Tax = T2Q - T2T; Tay = TX - TU; Taz = Tax - Tay; Tch = Tay + Tax; } { E TaA, TaB, T2U, T31; TaA = TN - TQ; TaB = T2X - T30; TaC = TaA - TaB; Tci = TaA + TaB; T2U = T2Q + T2T; T31 = T2X + T30; T32 = T2U + T31; Tdg = T2U - T31; } { E T3R, T3Y, T7T, T7U; T3R = T3P + T3Q; T3Y = KP707106781 * (T3U - T3X); T3Z = T3R - T3Y; T6J = T3R + T3Y; T7T = T40 + T41; T7U = KP707106781 * (T3X + T3U); T7V = T7T - T7U; T9n = T7T + T7U; } { E T7W, T7X, T42, T45; T7W = T3Q - T3P; T7X = KP707106781 * (T43 + T44); T7Y = T7W - T7X; T9o = T7W + T7X; T42 = T40 - T41; T45 = KP707106781 * (T43 - T44); T46 = T42 - T45; T6K = T42 + T45; } } { E T14, T4P, T4d, TaG, T17, T4a, T4S, TaH, T1e, TaZ, T4j, T4V, T1b, TaY, T4o; E T4U; { E T12, T13, T4Q, T4R; T12 = ri[WS(is, 1)]; T13 = ri[WS(is, 33)]; T14 = T12 + T13; T4P = T12 - T13; { E T4b, T4c, T15, T16; T4b = ii[WS(is, 1)]; T4c = ii[WS(is, 33)]; T4d = T4b - T4c; TaG = T4b + T4c; T15 = ri[WS(is, 17)]; T16 = ri[WS(is, 49)]; T17 = T15 + T16; T4a = T15 - T16; } T4Q = ii[WS(is, 17)]; T4R = ii[WS(is, 49)]; T4S = T4Q - T4R; TaH = T4Q + T4R; { E T1c, T1d, T4f, T4g, T4h, T4i; T1c = ri[WS(is, 57)]; T1d = ri[WS(is, 25)]; T4f = T1c - T1d; T4g = ii[WS(is, 57)]; T4h = ii[WS(is, 25)]; T4i = T4g - T4h; T1e = T1c + T1d; TaZ = T4g + T4h; T4j = T4f - T4i; T4V = T4f + T4i; } { E T19, T1a, T4k, T4l, T4m, T4n; T19 = ri[WS(is, 9)]; T1a = ri[WS(is, 41)]; T4k = T19 - T1a; T4l = ii[WS(is, 9)]; T4m = ii[WS(is, 41)]; T4n = T4l - T4m; T1b = T19 + T1a; TaY = T4l + T4m; T4o = T4k + T4n; T4U = T4n - T4k; } } { E T18, T1f, TaX, Tb0; T18 = T14 + T17; T1f = T1b + T1e; T1g = T18 + T1f; Tdp = T18 - T1f; TaX = T14 - T17; Tb0 = TaY - TaZ; Tb1 = TaX - Tb0; Tcm = TaX + Tb0; } { E Tdk, Tdl, T4e, T4p; Tdk = TaG + TaH; Tdl = TaY + TaZ; Tdm = Tdk - Tdl; Tej = Tdk + Tdl; T4e = T4a + T4d; T4p = KP707106781 * (T4j - T4o); T4q = T4e - T4p; T6R = T4e + T4p; } { E T4T, T4W, T8d, T8e; T4T = T4P - T4S; T4W = KP707106781 * (T4U - T4V); T4X = T4T - T4W; T6O = T4T + T4W; T8d = T4P + T4S; T8e = KP707106781 * (T4o + T4j); T8f = T8d - T8e; T9s = T8d + T8e; } { E TaI, TaJ, T82, T83; TaI = TaG - TaH; TaJ = T1e - T1b; TaK = TaI - TaJ; Tcp = TaJ + TaI; T82 = T4d - T4a; T83 = KP707106781 * (T4U + T4V); T84 = T82 - T83; T9v = T82 + T83; } } { E T1j, TaR, T1m, TaS, T4G, T4L, TaT, TaQ, T89, T88, T1q, TaM, T1t, TaN, T4v; E T4A, TaO, TaL, T86, T85; { E T4H, T4F, T4C, T4K; { E T1h, T1i, T4D, T4E; T1h = ri[WS(is, 5)]; T1i = ri[WS(is, 37)]; T1j = T1h + T1i; T4H = T1h - T1i; T4D = ii[WS(is, 5)]; T4E = ii[WS(is, 37)]; T4F = T4D - T4E; TaR = T4D + T4E; } { E T1k, T1l, T4I, T4J; T1k = ri[WS(is, 21)]; T1l = ri[WS(is, 53)]; T1m = T1k + T1l; T4C = T1k - T1l; T4I = ii[WS(is, 21)]; T4J = ii[WS(is, 53)]; T4K = T4I - T4J; TaS = T4I + T4J; } T4G = T4C + T4F; T4L = T4H - T4K; TaT = TaR - TaS; TaQ = T1j - T1m; T89 = T4H + T4K; T88 = T4F - T4C; } { E T4r, T4z, T4w, T4u; { E T1o, T1p, T4x, T4y; T1o = ri[WS(is, 61)]; T1p = ri[WS(is, 29)]; T1q = T1o + T1p; T4r = T1o - T1p; T4x = ii[WS(is, 61)]; T4y = ii[WS(is, 29)]; T4z = T4x - T4y; TaM = T4x + T4y; } { E T1r, T1s, T4s, T4t; T1r = ri[WS(is, 13)]; T1s = ri[WS(is, 45)]; T1t = T1r + T1s; T4w = T1r - T1s; T4s = ii[WS(is, 13)]; T4t = ii[WS(is, 45)]; T4u = T4s - T4t; TaN = T4s + T4t; } T4v = T4r - T4u; T4A = T4w + T4z; TaO = TaM - TaN; TaL = T1q - T1t; T86 = T4z - T4w; T85 = T4r + T4u; } { E T1n, T1u, Tb2, Tb3; T1n = T1j + T1m; T1u = T1q + T1t; T1v = T1n + T1u; Tdn = T1u - T1n; Tb2 = TaT - TaQ; Tb3 = TaL + TaO; Tb4 = KP707106781 * (Tb2 - Tb3); Tcq = KP707106781 * (Tb2 + Tb3); } { E Tdq, Tdr, T4B, T4M; Tdq = TaR + TaS; Tdr = TaM + TaN; Tds = Tdq - Tdr; Tek = Tdq + Tdr; T4B = FNMS(KP923879532, T4A, KP382683432 * T4v); T4M = FMA(KP923879532, T4G, KP382683432 * T4L); T4N = T4B - T4M; T6P = T4M + T4B; } { E T4Y, T4Z, T8g, T8h; T4Y = FNMS(KP923879532, T4L, KP382683432 * T4G); T4Z = FMA(KP382683432, T4A, KP923879532 * T4v); T50 = T4Y - T4Z; T6S = T4Y + T4Z; T8g = FNMS(KP382683432, T89, KP923879532 * T88); T8h = FMA(KP923879532, T86, KP382683432 * T85); T8i = T8g - T8h; T9w = T8g + T8h; } { E TaP, TaU, T87, T8a; TaP = TaL - TaO; TaU = TaQ + TaT; TaV = KP707106781 * (TaP - TaU); Tcn = KP707106781 * (TaU + TaP); T87 = FNMS(KP382683432, T86, KP923879532 * T85); T8a = FMA(KP382683432, T88, KP923879532 * T89); T8b = T87 - T8a; T9t = T8a + T87; } } { E T1O, Tbc, T1R, Tbd, T5o, T5t, Tbf, Tbe, T8p, T8o, T1V, Tbi, T1Y, Tbj, T5z; E T5E, Tbk, Tbh, T8s, T8r; { E T5p, T5n, T5k, T5s; { E T1M, T1N, T5l, T5m; T1M = ri[WS(is, 3)]; T1N = ri[WS(is, 35)]; T1O = T1M + T1N; T5p = T1M - T1N; T5l = ii[WS(is, 3)]; T5m = ii[WS(is, 35)]; T5n = T5l - T5m; Tbc = T5l + T5m; } { E T1P, T1Q, T5q, T5r; T1P = ri[WS(is, 19)]; T1Q = ri[WS(is, 51)]; T1R = T1P + T1Q; T5k = T1P - T1Q; T5q = ii[WS(is, 19)]; T5r = ii[WS(is, 51)]; T5s = T5q - T5r; Tbd = T5q + T5r; } T5o = T5k + T5n; T5t = T5p - T5s; Tbf = T1O - T1R; Tbe = Tbc - Tbd; T8p = T5p + T5s; T8o = T5n - T5k; } { E T5A, T5y, T5v, T5D; { E T1T, T1U, T5w, T5x; T1T = ri[WS(is, 59)]; T1U = ri[WS(is, 27)]; T1V = T1T + T1U; T5A = T1T - T1U; T5w = ii[WS(is, 59)]; T5x = ii[WS(is, 27)]; T5y = T5w - T5x; Tbi = T5w + T5x; } { E T1W, T1X, T5B, T5C; T1W = ri[WS(is, 11)]; T1X = ri[WS(is, 43)]; T1Y = T1W + T1X; T5v = T1W - T1X; T5B = ii[WS(is, 11)]; T5C = ii[WS(is, 43)]; T5D = T5B - T5C; Tbj = T5B + T5C; } T5z = T5v + T5y; T5E = T5A - T5D; Tbk = Tbi - Tbj; Tbh = T1V - T1Y; T8s = T5A + T5D; T8r = T5y - T5v; } { E T1S, T1Z, Tbt, Tbu; T1S = T1O + T1R; T1Z = T1V + T1Y; T20 = T1S + T1Z; TdD = T1Z - T1S; Tbt = Tbh - Tbk; Tbu = Tbf + Tbe; Tbv = KP707106781 * (Tbt - Tbu); Tcu = KP707106781 * (Tbu + Tbt); } { E Tdw, Tdx, T5u, T5F; Tdw = Tbc + Tbd; Tdx = Tbi + Tbj; Tdy = Tdw - Tdx; Tep = Tdw + Tdx; T5u = FNMS(KP923879532, T5t, KP382683432 * T5o); T5F = FMA(KP382683432, T5z, KP923879532 * T5E); T5G = T5u - T5F; T6Z = T5u + T5F; } { E T5R, T5S, T8z, T8A; T5R = FNMS(KP923879532, T5z, KP382683432 * T5E); T5S = FMA(KP923879532, T5o, KP382683432 * T5t); T5T = T5R - T5S; T6W = T5S + T5R; T8z = FNMS(KP382683432, T8r, KP923879532 * T8s); T8A = FMA(KP382683432, T8o, KP923879532 * T8p); T8B = T8z - T8A; T9A = T8A + T8z; } { E Tbg, Tbl, T8q, T8t; Tbg = Tbe - Tbf; Tbl = Tbh + Tbk; Tbm = KP707106781 * (Tbg - Tbl); Tcx = KP707106781 * (Tbg + Tbl); T8q = FNMS(KP382683432, T8p, KP923879532 * T8o); T8t = FMA(KP923879532, T8r, KP382683432 * T8s); T8u = T8q - T8t; T9D = T8q + T8t; } } { E T11, TeD, TeG, TeI, T22, T23, T34, TeH; { E Tv, T10, TeE, TeF; Tv = Tf + Tu; T10 = TK + TZ; T11 = Tv + T10; TeD = Tv - T10; TeE = Tej + Tek; TeF = Teo + Tep; TeG = TeE - TeF; TeI = TeE + TeF; } { E T1w, T21, T2y, T33; T1w = T1g + T1v; T21 = T1L + T20; T22 = T1w + T21; T23 = T21 - T1w; T2y = T2i + T2x; T33 = T2N + T32; T34 = T2y - T33; TeH = T2y + T33; } ro[WS(os, 32)] = T11 - T22; io[WS(os, 32)] = TeH - TeI; ro[0] = T11 + T22; io[0] = TeH + TeI; io[WS(os, 16)] = T23 + T34; ro[WS(os, 16)] = TeD + TeG; io[WS(os, 48)] = T34 - T23; ro[WS(os, 48)] = TeD - TeG; } { E Teh, Tex, Tev, TeB, Tem, Tey, Ter, Tez; { E Tef, Teg, Tet, Teu; Tef = Tf - Tu; Teg = T2N - T32; Teh = Tef + Teg; Tex = Tef - Teg; Tet = T2i - T2x; Teu = TZ - TK; Tev = Tet - Teu; TeB = Teu + Tet; } { E Tei, Tel, Ten, Teq; Tei = T1g - T1v; Tel = Tej - Tek; Tem = Tei + Tel; Tey = Tel - Tei; Ten = T1L - T20; Teq = Teo - Tep; Ter = Ten - Teq; Tez = Ten + Teq; } { E Tes, TeC, Tew, TeA; Tes = KP707106781 * (Tem + Ter); ro[WS(os, 40)] = Teh - Tes; ro[WS(os, 8)] = Teh + Tes; TeC = KP707106781 * (Tey + Tez); io[WS(os, 40)] = TeB - TeC; io[WS(os, 8)] = TeB + TeC; Tew = KP707106781 * (Ter - Tem); io[WS(os, 56)] = Tev - Tew; io[WS(os, 24)] = Tev + Tew; TeA = KP707106781 * (Tey - Tez); ro[WS(os, 56)] = Tex - TeA; ro[WS(os, 24)] = Tex + TeA; } } { E Tdb, TdV, Te5, TdJ, Tdi, Te6, Te3, Teb, TdM, TdW, Tdu, TdQ, Te0, Tea, TdF; E TdR; { E Tde, Tdh, Tdo, Tdt; Tdb = Td9 - Tda; TdV = Td9 + Tda; Te5 = TdI + TdH; TdJ = TdH - TdI; Tde = Tdc - Tdd; Tdh = Tdf + Tdg; Tdi = KP707106781 * (Tde - Tdh); Te6 = KP707106781 * (Tde + Tdh); { E Te1, Te2, TdK, TdL; Te1 = Tdv + Tdy; Te2 = TdD + TdC; Te3 = FNMS(KP382683432, Te2, KP923879532 * Te1); Teb = FMA(KP923879532, Te2, KP382683432 * Te1); TdK = Tdf - Tdg; TdL = Tdd + Tdc; TdM = KP707106781 * (TdK - TdL); TdW = KP707106781 * (TdL + TdK); } Tdo = Tdm - Tdn; Tdt = Tdp - Tds; Tdu = FMA(KP923879532, Tdo, KP382683432 * Tdt); TdQ = FNMS(KP923879532, Tdt, KP382683432 * Tdo); { E TdY, TdZ, Tdz, TdE; TdY = Tdn + Tdm; TdZ = Tdp + Tds; Te0 = FMA(KP382683432, TdY, KP923879532 * TdZ); Tea = FNMS(KP382683432, TdZ, KP923879532 * TdY); Tdz = Tdv - Tdy; TdE = TdC - TdD; TdF = FNMS(KP923879532, TdE, KP382683432 * Tdz); TdR = FMA(KP382683432, TdE, KP923879532 * Tdz); } } { E Tdj, TdG, TdT, TdU; Tdj = Tdb + Tdi; TdG = Tdu + TdF; ro[WS(os, 44)] = Tdj - TdG; ro[WS(os, 12)] = Tdj + TdG; TdT = TdJ + TdM; TdU = TdQ + TdR; io[WS(os, 44)] = TdT - TdU; io[WS(os, 12)] = TdT + TdU; } { E TdN, TdO, TdP, TdS; TdN = TdJ - TdM; TdO = TdF - Tdu; io[WS(os, 60)] = TdN - TdO; io[WS(os, 28)] = TdN + TdO; TdP = Tdb - Tdi; TdS = TdQ - TdR; ro[WS(os, 60)] = TdP - TdS; ro[WS(os, 28)] = TdP + TdS; } { E TdX, Te4, Ted, Tee; TdX = TdV + TdW; Te4 = Te0 + Te3; ro[WS(os, 36)] = TdX - Te4; ro[WS(os, 4)] = TdX + Te4; Ted = Te5 + Te6; Tee = Tea + Teb; io[WS(os, 36)] = Ted - Tee; io[WS(os, 4)] = Ted + Tee; } { E Te7, Te8, Te9, Tec; Te7 = Te5 - Te6; Te8 = Te3 - Te0; io[WS(os, 52)] = Te7 - Te8; io[WS(os, 20)] = Te7 + Te8; Te9 = TdV - TdW; Tec = Tea - Teb; ro[WS(os, 52)] = Te9 - Tec; ro[WS(os, 20)] = Te9 + Tec; } } { E Tcd, TcP, TcD, TcZ, Tck, Td0, TcX, Td5, Tcs, TcK, TcG, TcQ, TcU, Td4, Tcz; E TcL, Tcc, TcC; Tcc = KP707106781 * (TbD + TbC); Tcd = Tcb - Tcc; TcP = Tcb + Tcc; TcC = KP707106781 * (Tak + Tan); TcD = TcB - TcC; TcZ = TcB + TcC; { E Tcg, Tcj, TcV, TcW; Tcg = FNMS(KP382683432, Tcf, KP923879532 * Tce); Tcj = FMA(KP923879532, Tch, KP382683432 * Tci); Tck = Tcg - Tcj; Td0 = Tcg + Tcj; TcV = Tct + Tcu; TcW = Tcw + Tcx; TcX = FNMS(KP195090322, TcW, KP980785280 * TcV); Td5 = FMA(KP195090322, TcV, KP980785280 * TcW); } { E Tco, Tcr, TcE, TcF; Tco = Tcm - Tcn; Tcr = Tcp - Tcq; Tcs = FMA(KP555570233, Tco, KP831469612 * Tcr); TcK = FNMS(KP831469612, Tco, KP555570233 * Tcr); TcE = FNMS(KP382683432, Tch, KP923879532 * Tci); TcF = FMA(KP382683432, Tce, KP923879532 * Tcf); TcG = TcE - TcF; TcQ = TcF + TcE; } { E TcS, TcT, Tcv, Tcy; TcS = Tcm + Tcn; TcT = Tcp + Tcq; TcU = FMA(KP980785280, TcS, KP195090322 * TcT); Td4 = FNMS(KP195090322, TcS, KP980785280 * TcT); Tcv = Tct - Tcu; Tcy = Tcw - Tcx; Tcz = FNMS(KP831469612, Tcy, KP555570233 * Tcv); TcL = FMA(KP831469612, Tcv, KP555570233 * Tcy); } { E Tcl, TcA, TcN, TcO; Tcl = Tcd + Tck; TcA = Tcs + Tcz; ro[WS(os, 42)] = Tcl - TcA; ro[WS(os, 10)] = Tcl + TcA; TcN = TcD + TcG; TcO = TcK + TcL; io[WS(os, 42)] = TcN - TcO; io[WS(os, 10)] = TcN + TcO; } { E TcH, TcI, TcJ, TcM; TcH = TcD - TcG; TcI = Tcz - Tcs; io[WS(os, 58)] = TcH - TcI; io[WS(os, 26)] = TcH + TcI; TcJ = Tcd - Tck; TcM = TcK - TcL; ro[WS(os, 58)] = TcJ - TcM; ro[WS(os, 26)] = TcJ + TcM; } { E TcR, TcY, Td7, Td8; TcR = TcP + TcQ; TcY = TcU + TcX; ro[WS(os, 34)] = TcR - TcY; ro[WS(os, 2)] = TcR + TcY; Td7 = TcZ + Td0; Td8 = Td4 + Td5; io[WS(os, 34)] = Td7 - Td8; io[WS(os, 2)] = Td7 + Td8; } { E Td1, Td2, Td3, Td6; Td1 = TcZ - Td0; Td2 = TcX - TcU; io[WS(os, 50)] = Td1 - Td2; io[WS(os, 18)] = Td1 + Td2; Td3 = TcP - TcQ; Td6 = Td4 - Td5; ro[WS(os, 50)] = Td3 - Td6; ro[WS(os, 18)] = Td3 + Td6; } } { E Tap, TbR, TbF, Tc1, TaE, Tc2, TbZ, Tc7, Tb6, TbM, TbI, TbS, TbW, Tc6, Tbx; E TbN, Tao, TbE; Tao = KP707106781 * (Tak - Tan); Tap = Tah - Tao; TbR = Tah + Tao; TbE = KP707106781 * (TbC - TbD); TbF = TbB - TbE; Tc1 = TbB + TbE; { E Taw, TaD, TbX, TbY; Taw = FNMS(KP923879532, Tav, KP382683432 * Tas); TaD = FMA(KP382683432, Taz, KP923879532 * TaC); TaE = Taw - TaD; Tc2 = Taw + TaD; TbX = Tbb + Tbm; TbY = Tbs + Tbv; TbZ = FNMS(KP555570233, TbY, KP831469612 * TbX); Tc7 = FMA(KP831469612, TbY, KP555570233 * TbX); } { E TaW, Tb5, TbG, TbH; TaW = TaK - TaV; Tb5 = Tb1 - Tb4; Tb6 = FMA(KP980785280, TaW, KP195090322 * Tb5); TbM = FNMS(KP980785280, Tb5, KP195090322 * TaW); TbG = FNMS(KP923879532, Taz, KP382683432 * TaC); TbH = FMA(KP923879532, Tas, KP382683432 * Tav); TbI = TbG - TbH; TbS = TbH + TbG; } { E TbU, TbV, Tbn, Tbw; TbU = TaK + TaV; TbV = Tb1 + Tb4; TbW = FMA(KP555570233, TbU, KP831469612 * TbV); Tc6 = FNMS(KP555570233, TbV, KP831469612 * TbU); Tbn = Tbb - Tbm; Tbw = Tbs - Tbv; Tbx = FNMS(KP980785280, Tbw, KP195090322 * Tbn); TbN = FMA(KP195090322, Tbw, KP980785280 * Tbn); } { E TaF, Tby, TbP, TbQ; TaF = Tap + TaE; Tby = Tb6 + Tbx; ro[WS(os, 46)] = TaF - Tby; ro[WS(os, 14)] = TaF + Tby; TbP = TbF + TbI; TbQ = TbM + TbN; io[WS(os, 46)] = TbP - TbQ; io[WS(os, 14)] = TbP + TbQ; } { E TbJ, TbK, TbL, TbO; TbJ = TbF - TbI; TbK = Tbx - Tb6; io[WS(os, 62)] = TbJ - TbK; io[WS(os, 30)] = TbJ + TbK; TbL = Tap - TaE; TbO = TbM - TbN; ro[WS(os, 62)] = TbL - TbO; ro[WS(os, 30)] = TbL + TbO; } { E TbT, Tc0, Tc9, Tca; TbT = TbR + TbS; Tc0 = TbW + TbZ; ro[WS(os, 38)] = TbT - Tc0; ro[WS(os, 6)] = TbT + Tc0; Tc9 = Tc1 + Tc2; Tca = Tc6 + Tc7; io[WS(os, 38)] = Tc9 - Tca; io[WS(os, 6)] = Tc9 + Tca; } { E Tc3, Tc4, Tc5, Tc8; Tc3 = Tc1 - Tc2; Tc4 = TbZ - TbW; io[WS(os, 54)] = Tc3 - Tc4; io[WS(os, 22)] = Tc3 + Tc4; Tc5 = TbR - TbS; Tc8 = Tc6 - Tc7; ro[WS(os, 54)] = Tc5 - Tc8; ro[WS(os, 22)] = Tc5 + Tc8; } } { E T6F, T7h, T7m, T7w, T7p, T7x, T6M, T7s, T6U, T7c, T75, T7r, T78, T7i, T71; E T7d; { E T6D, T6E, T7k, T7l; T6D = T37 + T3e; T6E = T65 + T64; T6F = T6D - T6E; T7h = T6D + T6E; T7k = T6O + T6P; T7l = T6R + T6S; T7m = FMA(KP956940335, T7k, KP290284677 * T7l); T7w = FNMS(KP290284677, T7k, KP956940335 * T7l); } { E T7n, T7o, T6I, T6L; T7n = T6V + T6W; T7o = T6Y + T6Z; T7p = FNMS(KP290284677, T7o, KP956940335 * T7n); T7x = FMA(KP290284677, T7n, KP956940335 * T7o); T6I = FNMS(KP555570233, T6H, KP831469612 * T6G); T6L = FMA(KP831469612, T6J, KP555570233 * T6K); T6M = T6I - T6L; T7s = T6I + T6L; } { E T6Q, T6T, T73, T74; T6Q = T6O - T6P; T6T = T6R - T6S; T6U = FMA(KP471396736, T6Q, KP881921264 * T6T); T7c = FNMS(KP881921264, T6Q, KP471396736 * T6T); T73 = T5Z + T62; T74 = T3m + T3t; T75 = T73 - T74; T7r = T73 + T74; } { E T76, T77, T6X, T70; T76 = FNMS(KP555570233, T6J, KP831469612 * T6K); T77 = FMA(KP555570233, T6G, KP831469612 * T6H); T78 = T76 - T77; T7i = T77 + T76; T6X = T6V - T6W; T70 = T6Y - T6Z; T71 = FNMS(KP881921264, T70, KP471396736 * T6X); T7d = FMA(KP881921264, T6X, KP471396736 * T70); } { E T6N, T72, T7f, T7g; T6N = T6F + T6M; T72 = T6U + T71; ro[WS(os, 43)] = T6N - T72; ro[WS(os, 11)] = T6N + T72; T7f = T75 + T78; T7g = T7c + T7d; io[WS(os, 43)] = T7f - T7g; io[WS(os, 11)] = T7f + T7g; } { E T79, T7a, T7b, T7e; T79 = T75 - T78; T7a = T71 - T6U; io[WS(os, 59)] = T79 - T7a; io[WS(os, 27)] = T79 + T7a; T7b = T6F - T6M; T7e = T7c - T7d; ro[WS(os, 59)] = T7b - T7e; ro[WS(os, 27)] = T7b + T7e; } { E T7j, T7q, T7z, T7A; T7j = T7h + T7i; T7q = T7m + T7p; ro[WS(os, 35)] = T7j - T7q; ro[WS(os, 3)] = T7j + T7q; T7z = T7r + T7s; T7A = T7w + T7x; io[WS(os, 35)] = T7z - T7A; io[WS(os, 3)] = T7z + T7A; } { E T7t, T7u, T7v, T7y; T7t = T7r - T7s; T7u = T7p - T7m; io[WS(os, 51)] = T7t - T7u; io[WS(os, 19)] = T7t + T7u; T7v = T7h - T7i; T7y = T7w - T7x; ro[WS(os, 51)] = T7v - T7y; ro[WS(os, 19)] = T7v + T7y; } } { E T9j, T9V, Ta0, Taa, Ta3, Tab, T9q, Ta6, T9y, T9Q, T9J, Ta5, T9M, T9W, T9F; E T9R; { E T9h, T9i, T9Y, T9Z; T9h = T7B + T7C; T9i = T8J + T8I; T9j = T9h - T9i; T9V = T9h + T9i; T9Y = T9s + T9t; T9Z = T9v + T9w; Ta0 = FMA(KP995184726, T9Y, KP098017140 * T9Z); Taa = FNMS(KP098017140, T9Y, KP995184726 * T9Z); } { E Ta1, Ta2, T9m, T9p; Ta1 = T9z + T9A; Ta2 = T9C + T9D; Ta3 = FNMS(KP098017140, Ta2, KP995184726 * Ta1); Tab = FMA(KP098017140, Ta1, KP995184726 * Ta2); T9m = FNMS(KP195090322, T9l, KP980785280 * T9k); T9p = FMA(KP195090322, T9n, KP980785280 * T9o); T9q = T9m - T9p; Ta6 = T9m + T9p; } { E T9u, T9x, T9H, T9I; T9u = T9s - T9t; T9x = T9v - T9w; T9y = FMA(KP634393284, T9u, KP773010453 * T9x); T9Q = FNMS(KP773010453, T9u, KP634393284 * T9x); T9H = T8F + T8G; T9I = T7G + T7J; T9J = T9H - T9I; Ta5 = T9H + T9I; } { E T9K, T9L, T9B, T9E; T9K = FNMS(KP195090322, T9o, KP980785280 * T9n); T9L = FMA(KP980785280, T9l, KP195090322 * T9k); T9M = T9K - T9L; T9W = T9L + T9K; T9B = T9z - T9A; T9E = T9C - T9D; T9F = FNMS(KP773010453, T9E, KP634393284 * T9B); T9R = FMA(KP773010453, T9B, KP634393284 * T9E); } { E T9r, T9G, T9T, T9U; T9r = T9j + T9q; T9G = T9y + T9F; ro[WS(os, 41)] = T9r - T9G; ro[WS(os, 9)] = T9r + T9G; T9T = T9J + T9M; T9U = T9Q + T9R; io[WS(os, 41)] = T9T - T9U; io[WS(os, 9)] = T9T + T9U; } { E T9N, T9O, T9P, T9S; T9N = T9J - T9M; T9O = T9F - T9y; io[WS(os, 57)] = T9N - T9O; io[WS(os, 25)] = T9N + T9O; T9P = T9j - T9q; T9S = T9Q - T9R; ro[WS(os, 57)] = T9P - T9S; ro[WS(os, 25)] = T9P + T9S; } { E T9X, Ta4, Tad, Tae; T9X = T9V + T9W; Ta4 = Ta0 + Ta3; ro[WS(os, 33)] = T9X - Ta4; ro[WS(os, 1)] = T9X + Ta4; Tad = Ta5 + Ta6; Tae = Taa + Tab; io[WS(os, 33)] = Tad - Tae; io[WS(os, 1)] = Tad + Tae; } { E Ta7, Ta8, Ta9, Tac; Ta7 = Ta5 - Ta6; Ta8 = Ta3 - Ta0; io[WS(os, 49)] = Ta7 - Ta8; io[WS(os, 17)] = Ta7 + Ta8; Ta9 = T9V - T9W; Tac = Taa - Tab; ro[WS(os, 49)] = Ta9 - Tac; ro[WS(os, 17)] = Ta9 + Tac; } } { E T3v, T6j, T6o, T6y, T6r, T6z, T48, T6u, T52, T6e, T67, T6t, T6a, T6k, T5V; E T6f; { E T3f, T3u, T6m, T6n; T3f = T37 - T3e; T3u = T3m - T3t; T3v = T3f - T3u; T6j = T3f + T3u; T6m = T4q + T4N; T6n = T4X + T50; T6o = FMA(KP634393284, T6m, KP773010453 * T6n); T6y = FNMS(KP634393284, T6n, KP773010453 * T6m); } { E T6p, T6q, T3O, T47; T6p = T5j + T5G; T6q = T5Q + T5T; T6r = FNMS(KP634393284, T6q, KP773010453 * T6p); T6z = FMA(KP773010453, T6q, KP634393284 * T6p); T3O = FNMS(KP980785280, T3N, KP195090322 * T3G); T47 = FMA(KP195090322, T3Z, KP980785280 * T46); T48 = T3O - T47; T6u = T3O + T47; } { E T4O, T51, T63, T66; T4O = T4q - T4N; T51 = T4X - T50; T52 = FMA(KP995184726, T4O, KP098017140 * T51); T6e = FNMS(KP995184726, T51, KP098017140 * T4O); T63 = T5Z - T62; T66 = T64 - T65; T67 = T63 - T66; T6t = T63 + T66; } { E T68, T69, T5H, T5U; T68 = FNMS(KP980785280, T3Z, KP195090322 * T46); T69 = FMA(KP980785280, T3G, KP195090322 * T3N); T6a = T68 - T69; T6k = T69 + T68; T5H = T5j - T5G; T5U = T5Q - T5T; T5V = FNMS(KP995184726, T5U, KP098017140 * T5H); T6f = FMA(KP098017140, T5U, KP995184726 * T5H); } { E T49, T5W, T6h, T6i; T49 = T3v + T48; T5W = T52 + T5V; ro[WS(os, 47)] = T49 - T5W; ro[WS(os, 15)] = T49 + T5W; T6h = T67 + T6a; T6i = T6e + T6f; io[WS(os, 47)] = T6h - T6i; io[WS(os, 15)] = T6h + T6i; } { E T6b, T6c, T6d, T6g; T6b = T67 - T6a; T6c = T5V - T52; io[WS(os, 63)] = T6b - T6c; io[WS(os, 31)] = T6b + T6c; T6d = T3v - T48; T6g = T6e - T6f; ro[WS(os, 63)] = T6d - T6g; ro[WS(os, 31)] = T6d + T6g; } { E T6l, T6s, T6B, T6C; T6l = T6j + T6k; T6s = T6o + T6r; ro[WS(os, 39)] = T6l - T6s; ro[WS(os, 7)] = T6l + T6s; T6B = T6t + T6u; T6C = T6y + T6z; io[WS(os, 39)] = T6B - T6C; io[WS(os, 7)] = T6B + T6C; } { E T6v, T6w, T6x, T6A; T6v = T6t - T6u; T6w = T6r - T6o; io[WS(os, 55)] = T6v - T6w; io[WS(os, 23)] = T6v + T6w; T6x = T6j - T6k; T6A = T6y - T6z; ro[WS(os, 55)] = T6x - T6A; ro[WS(os, 23)] = T6x + T6A; } } { E T7L, T8X, T92, T9c, T95, T9d, T80, T98, T8k, T8S, T8L, T97, T8O, T8Y, T8D; E T8T; { E T7D, T7K, T90, T91; T7D = T7B - T7C; T7K = T7G - T7J; T7L = T7D - T7K; T8X = T7D + T7K; T90 = T84 + T8b; T91 = T8f + T8i; T92 = FMA(KP471396736, T90, KP881921264 * T91); T9c = FNMS(KP471396736, T91, KP881921264 * T90); } { E T93, T94, T7S, T7Z; T93 = T8n + T8u; T94 = T8y + T8B; T95 = FNMS(KP471396736, T94, KP881921264 * T93); T9d = FMA(KP881921264, T94, KP471396736 * T93); T7S = FNMS(KP831469612, T7R, KP555570233 * T7O); T7Z = FMA(KP831469612, T7V, KP555570233 * T7Y); T80 = T7S - T7Z; T98 = T7S + T7Z; } { E T8c, T8j, T8H, T8K; T8c = T84 - T8b; T8j = T8f - T8i; T8k = FMA(KP956940335, T8c, KP290284677 * T8j); T8S = FNMS(KP956940335, T8j, KP290284677 * T8c); T8H = T8F - T8G; T8K = T8I - T8J; T8L = T8H - T8K; T97 = T8H + T8K; } { E T8M, T8N, T8v, T8C; T8M = FNMS(KP831469612, T7Y, KP555570233 * T7V); T8N = FMA(KP555570233, T7R, KP831469612 * T7O); T8O = T8M - T8N; T8Y = T8N + T8M; T8v = T8n - T8u; T8C = T8y - T8B; T8D = FNMS(KP956940335, T8C, KP290284677 * T8v); T8T = FMA(KP290284677, T8C, KP956940335 * T8v); } { E T81, T8E, T8V, T8W; T81 = T7L + T80; T8E = T8k + T8D; ro[WS(os, 45)] = T81 - T8E; ro[WS(os, 13)] = T81 + T8E; T8V = T8L + T8O; T8W = T8S + T8T; io[WS(os, 45)] = T8V - T8W; io[WS(os, 13)] = T8V + T8W; } { E T8P, T8Q, T8R, T8U; T8P = T8L - T8O; T8Q = T8D - T8k; io[WS(os, 61)] = T8P - T8Q; io[WS(os, 29)] = T8P + T8Q; T8R = T7L - T80; T8U = T8S - T8T; ro[WS(os, 61)] = T8R - T8U; ro[WS(os, 29)] = T8R + T8U; } { E T8Z, T96, T9f, T9g; T8Z = T8X + T8Y; T96 = T92 + T95; ro[WS(os, 37)] = T8Z - T96; ro[WS(os, 5)] = T8Z + T96; T9f = T97 + T98; T9g = T9c + T9d; io[WS(os, 37)] = T9f - T9g; io[WS(os, 5)] = T9f + T9g; } { E T99, T9a, T9b, T9e; T99 = T97 - T98; T9a = T95 - T92; io[WS(os, 53)] = T99 - T9a; io[WS(os, 21)] = T99 + T9a; T9b = T8X - T8Y; T9e = T9c - T9d; ro[WS(os, 53)] = T9b - T9e; ro[WS(os, 21)] = T9b + T9e; } } } } } static const kdft_desc desc = { 64, "n1_64", {808, 144, 104, 0}, &GENUS, 0, 0, 0, 0 }; void X(codelet_n1_64) (planner *p) { X(kdft_register) (p, n1_64, &desc); } #endif /* HAVE_FMA */ fftw-3.3.3/dft/scalar/codelets/n1_20.c0000644000175000017500000005056512054410243014224 00000000000000/* * Copyright (c) 2003, 2007-11 Matteo Frigo * Copyright (c) 2003, 2007-11 Massachusetts Institute of Technology * * This program is free software; you can redistribute it and/or modify * it under the terms of the GNU General Public License as published by * the Free Software Foundation; either version 2 of the License, or * (at your option) any later version. * * This program is distributed in the hope that it will be useful, * but WITHOUT ANY WARRANTY; without even the implied warranty of * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the * GNU General Public License for more details. * * You should have received a copy of the GNU General Public License * along with this program; if not, write to the Free Software * Foundation, Inc., 51 Franklin Street, Fifth Floor, Boston, MA 02110-1301 USA * */ /* This file was automatically generated --- DO NOT EDIT */ /* Generated on Sun Nov 25 07:35:46 EST 2012 */ #include "codelet-dft.h" #ifdef HAVE_FMA /* Generated by: ../../../genfft/gen_notw.native -fma -reorder-insns -schedule-for-pipeline -compact -variables 4 -pipeline-latency 4 -n 20 -name n1_20 -include n.h */ /* * This function contains 208 FP additions, 72 FP multiplications, * (or, 136 additions, 0 multiplications, 72 fused multiply/add), * 86 stack variables, 4 constants, and 80 memory accesses */ #include "n.h" static void n1_20(const R *ri, const R *ii, R *ro, R *io, stride is, stride os, INT v, INT ivs, INT ovs) { DK(KP951056516, +0.951056516295153572116439333379382143405698634); DK(KP559016994, +0.559016994374947424102293417182819058860154590); DK(KP618033988, +0.618033988749894848204586834365638117720309180); DK(KP250000000, +0.250000000000000000000000000000000000000000000); { INT i; for (i = v; i > 0; i = i - 1, ri = ri + ivs, ii = ii + ivs, ro = ro + ovs, io = io + ovs, MAKE_VOLATILE_STRIDE(80, is), MAKE_VOLATILE_STRIDE(80, os)) { E T1Y, T1Z, T1W, T1V; { E T1d, TP, TD, T7, T3b, T2N, T2f, T1R, T2U, TB, T2P, T2A, T3d, T37, T3j; E TJ, T2n, T1b, T1T, T1y, T2b, T2h, T1j, T2V, Tm, T2O, T2H, T3c, T34, T1e; E T1f, T3i, TG, T2m, T10, T1S, T1J, T28, T2g; { E T4, T1N, T3, T2L, TN, T5, T1O, T1P, T1h, T1i; { E T1, T2, TL, TM; T1 = ri[0]; T2 = ri[WS(is, 10)]; TL = ii[0]; TM = ii[WS(is, 10)]; T4 = ri[WS(is, 5)]; T1N = T1 - T2; T3 = T1 + T2; T2L = TL + TM; TN = TL - TM; T5 = ri[WS(is, 15)]; T1O = ii[WS(is, 5)]; T1P = ii[WS(is, 15)]; } { E T1o, Tp, T2u, T13, T14, Ts, T2v, T1r, Tx, T1t, Tw, T2x, T18, Ty, T1u; E T1v; { E Tq, Tr, T1p, T1q; { E Tn, To, T11, T12; Tn = ri[WS(is, 8)]; { E TO, T6, T2M, T1Q; TO = T4 - T5; T6 = T4 + T5; T2M = T1O + T1P; T1Q = T1O - T1P; T1d = TO + TN; TP = TN - TO; TD = T3 + T6; T7 = T3 - T6; T3b = T2L + T2M; T2N = T2L - T2M; T2f = T1N + T1Q; T1R = T1N - T1Q; To = ri[WS(is, 18)]; } T11 = ii[WS(is, 8)]; T12 = ii[WS(is, 18)]; Tq = ri[WS(is, 13)]; T1o = Tn - To; Tp = Tn + To; T2u = T11 + T12; T13 = T11 - T12; Tr = ri[WS(is, 3)]; T1p = ii[WS(is, 13)]; T1q = ii[WS(is, 3)]; } { E Tu, Tv, T16, T17; Tu = ri[WS(is, 12)]; T14 = Tq - Tr; Ts = Tq + Tr; T2v = T1p + T1q; T1r = T1p - T1q; Tv = ri[WS(is, 2)]; T16 = ii[WS(is, 12)]; T17 = ii[WS(is, 2)]; Tx = ri[WS(is, 17)]; T1t = Tu - Tv; Tw = Tu + Tv; T2x = T16 + T17; T18 = T16 - T17; Ty = ri[WS(is, 7)]; T1u = ii[WS(is, 17)]; T1v = ii[WS(is, 7)]; } } { E TH, T19, T1w, TI; { E Tt, T2w, T35, TA, T2z, T36, Tz, T2y; TH = Tp + Ts; Tt = Tp - Ts; T19 = Tx - Ty; Tz = Tx + Ty; T2y = T1u + T1v; T1w = T1u - T1v; T2w = T2u - T2v; T35 = T2u + T2v; TI = Tw + Tz; TA = Tw - Tz; T2z = T2x - T2y; T36 = T2x + T2y; T2U = Tt - TA; TB = Tt + TA; T2P = T2w + T2z; T2A = T2w - T2z; T3d = T35 + T36; T37 = T35 - T36; } { E T1s, T29, T1x, T2a, T15, T1a; T15 = T13 - T14; T1h = T14 + T13; T1i = T19 + T18; T1a = T18 - T19; T1s = T1o - T1r; T29 = T1o + T1r; T3j = TH - TI; TJ = TH + TI; T1x = T1t - T1w; T2a = T1t + T1w; T2n = T15 - T1a; T1b = T15 + T1a; T1T = T1s + T1x; T1y = T1s - T1x; T2b = T29 - T2a; T2h = T29 + T2a; } } } { E Ta, T1z, T2B, TS, TT, Td, T2C, T1C, Ti, T1E, Th, T2E, TX, Tj, T1F; E T1G; { E Tb, Tc, T1A, T1B; { E TQ, TR, T8, T9; T8 = ri[WS(is, 4)]; T9 = ri[WS(is, 14)]; T1j = T1h + T1i; T1Y = T1h - T1i; TQ = ii[WS(is, 4)]; TR = ii[WS(is, 14)]; Ta = T8 + T9; T1z = T8 - T9; Tb = ri[WS(is, 9)]; T2B = TQ + TR; TS = TQ - TR; Tc = ri[WS(is, 19)]; T1A = ii[WS(is, 9)]; T1B = ii[WS(is, 19)]; } { E Tf, Tg, TV, TW; Tf = ri[WS(is, 16)]; TT = Tb - Tc; Td = Tb + Tc; T2C = T1A + T1B; T1C = T1A - T1B; Tg = ri[WS(is, 6)]; TV = ii[WS(is, 16)]; TW = ii[WS(is, 6)]; Ti = ri[WS(is, 1)]; T1E = Tf - Tg; Th = Tf + Tg; T2E = TV + TW; TX = TV - TW; Tj = ri[WS(is, 11)]; T1F = ii[WS(is, 1)]; T1G = ii[WS(is, 11)]; } } { E TE, TY, T1H, TF; { E Te, T2D, T32, Tl, T2G, T33, Tk, T2F; TE = Ta + Td; Te = Ta - Td; TY = Ti - Tj; Tk = Ti + Tj; T2F = T1F + T1G; T1H = T1F - T1G; T2D = T2B - T2C; T32 = T2B + T2C; TF = Th + Tk; Tl = Th - Tk; T2G = T2E - T2F; T33 = T2E + T2F; T2V = Te - Tl; Tm = Te + Tl; T2O = T2D + T2G; T2H = T2D - T2G; T3c = T32 + T33; T34 = T32 - T33; } { E T1D, T26, T1I, T27, TU, TZ; TU = TS - TT; T1e = TT + TS; T1f = TY + TX; TZ = TX - TY; T1D = T1z - T1C; T26 = T1z + T1C; T3i = TE - TF; TG = TE + TF; T1I = T1E - T1H; T27 = T1E + T1H; T2m = TU - TZ; T10 = TU + TZ; T1S = T1D + T1I; T1J = T1D - T1I; T28 = T26 - T27; T2g = T26 + T27; } } } } { E T1g, T3g, T3f, T2S, T2R, T2k, T2j; { E T2s, T2r, TC, T2Q; T2s = Tm - TB; TC = Tm + TB; T1g = T1e + T1f; T1Z = T1e - T1f; T2r = FNMS(KP250000000, TC, T7); ro[WS(os, 10)] = T7 + TC; T2Q = T2O + T2P; T2S = T2O - T2P; { E T2K, T2I, T2t, T2J; T2K = FMA(KP618033988, T2A, T2H); T2I = FNMS(KP618033988, T2H, T2A); T2t = FNMS(KP559016994, T2s, T2r); T2J = FMA(KP559016994, T2s, T2r); ro[WS(os, 18)] = FMA(KP951056516, T2I, T2t); ro[WS(os, 2)] = FNMS(KP951056516, T2I, T2t); ro[WS(os, 6)] = FMA(KP951056516, T2K, T2J); ro[WS(os, 14)] = FNMS(KP951056516, T2K, T2J); T2R = FNMS(KP250000000, T2Q, T2N); } io[WS(os, 10)] = T2N + T2Q; } { E T30, T2Z, TK, T3e; TK = TG + TJ; T30 = TG - TJ; { E T2T, T2X, T2Y, T2W; T2T = FNMS(KP559016994, T2S, T2R); T2X = FMA(KP559016994, T2S, T2R); T2Y = FMA(KP618033988, T2U, T2V); T2W = FNMS(KP618033988, T2V, T2U); io[WS(os, 14)] = FMA(KP951056516, T2Y, T2X); io[WS(os, 6)] = FNMS(KP951056516, T2Y, T2X); io[WS(os, 18)] = FNMS(KP951056516, T2W, T2T); io[WS(os, 2)] = FMA(KP951056516, T2W, T2T); T2Z = FNMS(KP250000000, TK, TD); } ro[0] = TD + TK; T3e = T3c + T3d; T3g = T3c - T3d; { E T31, T39, T3a, T38; T31 = FMA(KP559016994, T30, T2Z); T39 = FNMS(KP559016994, T30, T2Z); T3a = FNMS(KP618033988, T34, T37); T38 = FMA(KP618033988, T37, T34); ro[WS(os, 8)] = FMA(KP951056516, T3a, T39); ro[WS(os, 12)] = FNMS(KP951056516, T3a, T39); ro[WS(os, 16)] = FMA(KP951056516, T38, T31); ro[WS(os, 4)] = FNMS(KP951056516, T38, T31); T3f = FNMS(KP250000000, T3e, T3b); } io[0] = T3b + T3e; } { E T24, T23, T1c, T2i; T1c = T10 + T1b; T24 = T10 - T1b; { E T3h, T3l, T3m, T3k; T3h = FMA(KP559016994, T3g, T3f); T3l = FNMS(KP559016994, T3g, T3f); T3m = FNMS(KP618033988, T3i, T3j); T3k = FMA(KP618033988, T3j, T3i); io[WS(os, 12)] = FMA(KP951056516, T3m, T3l); io[WS(os, 8)] = FNMS(KP951056516, T3m, T3l); io[WS(os, 16)] = FNMS(KP951056516, T3k, T3h); io[WS(os, 4)] = FMA(KP951056516, T3k, T3h); T23 = FNMS(KP250000000, T1c, TP); } io[WS(os, 5)] = TP + T1c; T2i = T2g + T2h; T2k = T2g - T2h; { E T25, T2d, T2e, T2c; T25 = FMA(KP559016994, T24, T23); T2d = FNMS(KP559016994, T24, T23); T2e = FNMS(KP618033988, T28, T2b); T2c = FMA(KP618033988, T2b, T28); io[WS(os, 17)] = FMA(KP951056516, T2e, T2d); io[WS(os, 13)] = FNMS(KP951056516, T2e, T2d); io[WS(os, 9)] = FMA(KP951056516, T2c, T25); io[WS(os, 1)] = FNMS(KP951056516, T2c, T25); T2j = FNMS(KP250000000, T2i, T2f); } ro[WS(os, 5)] = T2f + T2i; } { E T1m, T1l, T1k, T1U; T1k = T1g + T1j; T1m = T1g - T1j; { E T2l, T2p, T2q, T2o; T2l = FMA(KP559016994, T2k, T2j); T2p = FNMS(KP559016994, T2k, T2j); T2q = FNMS(KP618033988, T2m, T2n); T2o = FMA(KP618033988, T2n, T2m); ro[WS(os, 17)] = FNMS(KP951056516, T2q, T2p); ro[WS(os, 13)] = FMA(KP951056516, T2q, T2p); ro[WS(os, 9)] = FNMS(KP951056516, T2o, T2l); ro[WS(os, 1)] = FMA(KP951056516, T2o, T2l); T1l = FNMS(KP250000000, T1k, T1d); } io[WS(os, 15)] = T1d + T1k; T1U = T1S + T1T; T1W = T1S - T1T; { E T1n, T1L, T1M, T1K; T1n = FNMS(KP559016994, T1m, T1l); T1L = FMA(KP559016994, T1m, T1l); T1M = FMA(KP618033988, T1y, T1J); T1K = FNMS(KP618033988, T1J, T1y); io[WS(os, 19)] = FMA(KP951056516, T1M, T1L); io[WS(os, 11)] = FNMS(KP951056516, T1M, T1L); io[WS(os, 7)] = FMA(KP951056516, T1K, T1n); io[WS(os, 3)] = FNMS(KP951056516, T1K, T1n); T1V = FNMS(KP250000000, T1U, T1R); } ro[WS(os, 15)] = T1R + T1U; } } } { E T21, T1X, T20, T22; T21 = FMA(KP559016994, T1W, T1V); T1X = FNMS(KP559016994, T1W, T1V); T20 = FNMS(KP618033988, T1Z, T1Y); T22 = FMA(KP618033988, T1Y, T1Z); ro[WS(os, 19)] = FNMS(KP951056516, T22, T21); ro[WS(os, 11)] = FMA(KP951056516, T22, T21); ro[WS(os, 7)] = FNMS(KP951056516, T20, T1X); ro[WS(os, 3)] = FMA(KP951056516, T20, T1X); } } } } static const kdft_desc desc = { 20, "n1_20", {136, 0, 72, 0}, &GENUS, 0, 0, 0, 0 }; void X(codelet_n1_20) (planner *p) { X(kdft_register) (p, n1_20, &desc); } #else /* HAVE_FMA */ /* Generated by: ../../../genfft/gen_notw.native -compact -variables 4 -pipeline-latency 4 -n 20 -name n1_20 -include n.h */ /* * This function contains 208 FP additions, 48 FP multiplications, * (or, 184 additions, 24 multiplications, 24 fused multiply/add), * 81 stack variables, 4 constants, and 80 memory accesses */ #include "n.h" static void n1_20(const R *ri, const R *ii, R *ro, R *io, stride is, stride os, INT v, INT ivs, INT ovs) { DK(KP587785252, +0.587785252292473129168705954639072768597652438); DK(KP951056516, +0.951056516295153572116439333379382143405698634); DK(KP250000000, +0.250000000000000000000000000000000000000000000); DK(KP559016994, +0.559016994374947424102293417182819058860154590); { INT i; for (i = v; i > 0; i = i - 1, ri = ri + ivs, ii = ii + ivs, ro = ro + ovs, io = io + ovs, MAKE_VOLATILE_STRIDE(80, is), MAKE_VOLATILE_STRIDE(80, os)) { E T7, T2Q, T3h, TD, TP, T1U, T2l, T1d, Tt, TA, TB, T2w, T2z, T2S, T35; E T36, T3f, TH, TI, TJ, T15, T1a, T1b, T1s, T1x, T1W, T29, T2a, T2j, T1h; E T1i, T1j, Te, Tl, Tm, T2D, T2G, T2R, T32, T33, T3e, TE, TF, TG, TU; E TZ, T10, T1D, T1I, T1V, T26, T27, T2i, T1e, T1f, T1g; { E T3, T1Q, TN, T2O, T6, TO, T1T, T2P; { E T1, T2, TL, TM; T1 = ri[0]; T2 = ri[WS(is, 10)]; T3 = T1 + T2; T1Q = T1 - T2; TL = ii[0]; TM = ii[WS(is, 10)]; TN = TL - TM; T2O = TL + TM; } { E T4, T5, T1R, T1S; T4 = ri[WS(is, 5)]; T5 = ri[WS(is, 15)]; T6 = T4 + T5; TO = T4 - T5; T1R = ii[WS(is, 5)]; T1S = ii[WS(is, 15)]; T1T = T1R - T1S; T2P = T1R + T1S; } T7 = T3 - T6; T2Q = T2O - T2P; T3h = T2O + T2P; TD = T3 + T6; TP = TN - TO; T1U = T1Q - T1T; T2l = T1Q + T1T; T1d = TO + TN; } { E Tp, T1o, T13, T2u, Ts, T14, T1r, T2v, Tw, T1t, T18, T2x, Tz, T19, T1w; E T2y; { E Tn, To, T11, T12; Tn = ri[WS(is, 8)]; To = ri[WS(is, 18)]; Tp = Tn + To; T1o = Tn - To; T11 = ii[WS(is, 8)]; T12 = ii[WS(is, 18)]; T13 = T11 - T12; T2u = T11 + T12; } { E Tq, Tr, T1p, T1q; Tq = ri[WS(is, 13)]; Tr = ri[WS(is, 3)]; Ts = Tq + Tr; T14 = Tq - Tr; T1p = ii[WS(is, 13)]; T1q = ii[WS(is, 3)]; T1r = T1p - T1q; T2v = T1p + T1q; } { E Tu, Tv, T16, T17; Tu = ri[WS(is, 12)]; Tv = ri[WS(is, 2)]; Tw = Tu + Tv; T1t = Tu - Tv; T16 = ii[WS(is, 12)]; T17 = ii[WS(is, 2)]; T18 = T16 - T17; T2x = T16 + T17; } { E Tx, Ty, T1u, T1v; Tx = ri[WS(is, 17)]; Ty = ri[WS(is, 7)]; Tz = Tx + Ty; T19 = Tx - Ty; T1u = ii[WS(is, 17)]; T1v = ii[WS(is, 7)]; T1w = T1u - T1v; T2y = T1u + T1v; } Tt = Tp - Ts; TA = Tw - Tz; TB = Tt + TA; T2w = T2u - T2v; T2z = T2x - T2y; T2S = T2w + T2z; T35 = T2u + T2v; T36 = T2x + T2y; T3f = T35 + T36; TH = Tp + Ts; TI = Tw + Tz; TJ = TH + TI; T15 = T13 - T14; T1a = T18 - T19; T1b = T15 + T1a; T1s = T1o - T1r; T1x = T1t - T1w; T1W = T1s + T1x; T29 = T1o + T1r; T2a = T1t + T1w; T2j = T29 + T2a; T1h = T14 + T13; T1i = T19 + T18; T1j = T1h + T1i; } { E Ta, T1z, TS, T2B, Td, TT, T1C, T2C, Th, T1E, TX, T2E, Tk, TY, T1H; E T2F; { E T8, T9, TQ, TR; T8 = ri[WS(is, 4)]; T9 = ri[WS(is, 14)]; Ta = T8 + T9; T1z = T8 - T9; TQ = ii[WS(is, 4)]; TR = ii[WS(is, 14)]; TS = TQ - TR; T2B = TQ + TR; } { E Tb, Tc, T1A, T1B; Tb = ri[WS(is, 9)]; Tc = ri[WS(is, 19)]; Td = Tb + Tc; TT = Tb - Tc; T1A = ii[WS(is, 9)]; T1B = ii[WS(is, 19)]; T1C = T1A - T1B; T2C = T1A + T1B; } { E Tf, Tg, TV, TW; Tf = ri[WS(is, 16)]; Tg = ri[WS(is, 6)]; Th = Tf + Tg; T1E = Tf - Tg; TV = ii[WS(is, 16)]; TW = ii[WS(is, 6)]; TX = TV - TW; T2E = TV + TW; } { E Ti, Tj, T1F, T1G; Ti = ri[WS(is, 1)]; Tj = ri[WS(is, 11)]; Tk = Ti + Tj; TY = Ti - Tj; T1F = ii[WS(is, 1)]; T1G = ii[WS(is, 11)]; T1H = T1F - T1G; T2F = T1F + T1G; } Te = Ta - Td; Tl = Th - Tk; Tm = Te + Tl; T2D = T2B - T2C; T2G = T2E - T2F; T2R = T2D + T2G; T32 = T2B + T2C; T33 = T2E + T2F; T3e = T32 + T33; TE = Ta + Td; TF = Th + Tk; TG = TE + TF; TU = TS - TT; TZ = TX - TY; T10 = TU + TZ; T1D = T1z - T1C; T1I = T1E - T1H; T1V = T1D + T1I; T26 = T1z + T1C; T27 = T1E + T1H; T2i = T26 + T27; T1e = TT + TS; T1f = TY + TX; T1g = T1e + T1f; } { E T2s, TC, T2r, T2I, T2K, T2A, T2H, T2J, T2t; T2s = KP559016994 * (Tm - TB); TC = Tm + TB; T2r = FNMS(KP250000000, TC, T7); T2A = T2w - T2z; T2H = T2D - T2G; T2I = FNMS(KP587785252, T2H, KP951056516 * T2A); T2K = FMA(KP951056516, T2H, KP587785252 * T2A); ro[WS(os, 10)] = T7 + TC; T2J = T2s + T2r; ro[WS(os, 14)] = T2J - T2K; ro[WS(os, 6)] = T2J + T2K; T2t = T2r - T2s; ro[WS(os, 2)] = T2t - T2I; ro[WS(os, 18)] = T2t + T2I; } { E T2V, T2T, T2U, T2N, T2Y, T2L, T2M, T2X, T2W; T2V = KP559016994 * (T2R - T2S); T2T = T2R + T2S; T2U = FNMS(KP250000000, T2T, T2Q); T2L = Tt - TA; T2M = Te - Tl; T2N = FNMS(KP587785252, T2M, KP951056516 * T2L); T2Y = FMA(KP951056516, T2M, KP587785252 * T2L); io[WS(os, 10)] = T2Q + T2T; T2X = T2V + T2U; io[WS(os, 6)] = T2X - T2Y; io[WS(os, 14)] = T2Y + T2X; T2W = T2U - T2V; io[WS(os, 2)] = T2N + T2W; io[WS(os, 18)] = T2W - T2N; } { E T2Z, TK, T30, T38, T3a, T34, T37, T39, T31; T2Z = KP559016994 * (TG - TJ); TK = TG + TJ; T30 = FNMS(KP250000000, TK, TD); T34 = T32 - T33; T37 = T35 - T36; T38 = FMA(KP951056516, T34, KP587785252 * T37); T3a = FNMS(KP587785252, T34, KP951056516 * T37); ro[0] = TD + TK; T39 = T30 - T2Z; ro[WS(os, 12)] = T39 - T3a; ro[WS(os, 8)] = T39 + T3a; T31 = T2Z + T30; ro[WS(os, 4)] = T31 - T38; ro[WS(os, 16)] = T31 + T38; } { E T3g, T3i, T3j, T3d, T3m, T3b, T3c, T3l, T3k; T3g = KP559016994 * (T3e - T3f); T3i = T3e + T3f; T3j = FNMS(KP250000000, T3i, T3h); T3b = TE - TF; T3c = TH - TI; T3d = FMA(KP951056516, T3b, KP587785252 * T3c); T3m = FNMS(KP587785252, T3b, KP951056516 * T3c); io[0] = T3h + T3i; T3l = T3j - T3g; io[WS(os, 8)] = T3l - T3m; io[WS(os, 12)] = T3m + T3l; T3k = T3g + T3j; io[WS(os, 4)] = T3d + T3k; io[WS(os, 16)] = T3k - T3d; } { E T23, T1c, T24, T2c, T2e, T28, T2b, T2d, T25; T23 = KP559016994 * (T10 - T1b); T1c = T10 + T1b; T24 = FNMS(KP250000000, T1c, TP); T28 = T26 - T27; T2b = T29 - T2a; T2c = FMA(KP951056516, T28, KP587785252 * T2b); T2e = FNMS(KP587785252, T28, KP951056516 * T2b); io[WS(os, 5)] = TP + T1c; T2d = T24 - T23; io[WS(os, 13)] = T2d - T2e; io[WS(os, 17)] = T2d + T2e; T25 = T23 + T24; io[WS(os, 1)] = T25 - T2c; io[WS(os, 9)] = T25 + T2c; } { E T2k, T2m, T2n, T2h, T2p, T2f, T2g, T2q, T2o; T2k = KP559016994 * (T2i - T2j); T2m = T2i + T2j; T2n = FNMS(KP250000000, T2m, T2l); T2f = TU - TZ; T2g = T15 - T1a; T2h = FMA(KP951056516, T2f, KP587785252 * T2g); T2p = FNMS(KP587785252, T2f, KP951056516 * T2g); ro[WS(os, 5)] = T2l + T2m; T2q = T2n - T2k; ro[WS(os, 13)] = T2p + T2q; ro[WS(os, 17)] = T2q - T2p; T2o = T2k + T2n; ro[WS(os, 1)] = T2h + T2o; ro[WS(os, 9)] = T2o - T2h; } { E T1m, T1k, T1l, T1K, T1M, T1y, T1J, T1L, T1n; T1m = KP559016994 * (T1g - T1j); T1k = T1g + T1j; T1l = FNMS(KP250000000, T1k, T1d); T1y = T1s - T1x; T1J = T1D - T1I; T1K = FNMS(KP587785252, T1J, KP951056516 * T1y); T1M = FMA(KP951056516, T1J, KP587785252 * T1y); io[WS(os, 15)] = T1d + T1k; T1L = T1m + T1l; io[WS(os, 11)] = T1L - T1M; io[WS(os, 19)] = T1L + T1M; T1n = T1l - T1m; io[WS(os, 3)] = T1n - T1K; io[WS(os, 7)] = T1n + T1K; } { E T1Z, T1X, T1Y, T1P, T21, T1N, T1O, T22, T20; T1Z = KP559016994 * (T1V - T1W); T1X = T1V + T1W; T1Y = FNMS(KP250000000, T1X, T1U); T1N = T1h - T1i; T1O = T1e - T1f; T1P = FNMS(KP587785252, T1O, KP951056516 * T1N); T21 = FMA(KP951056516, T1O, KP587785252 * T1N); ro[WS(os, 15)] = T1U + T1X; T22 = T1Z + T1Y; ro[WS(os, 11)] = T21 + T22; ro[WS(os, 19)] = T22 - T21; T20 = T1Y - T1Z; ro[WS(os, 3)] = T1P + T20; ro[WS(os, 7)] = T20 - T1P; } } } } static const kdft_desc desc = { 20, "n1_20", {184, 24, 24, 0}, &GENUS, 0, 0, 0, 0 }; void X(codelet_n1_20) (planner *p) { X(kdft_register) (p, n1_20, &desc); } #endif /* HAVE_FMA */ fftw-3.3.3/dft/scalar/codelets/n1_25.c0000644000175000017500000011627312054410247014234 00000000000000/* * Copyright (c) 2003, 2007-11 Matteo Frigo * Copyright (c) 2003, 2007-11 Massachusetts Institute of Technology * * This program is free software; you can redistribute it and/or modify * it under the terms of the GNU General Public License as published by * the Free Software Foundation; either version 2 of the License, or * (at your option) any later version. * * This program is distributed in the hope that it will be useful, * but WITHOUT ANY WARRANTY; without even the implied warranty of * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the * GNU General Public License for more details. * * You should have received a copy of the GNU General Public License * along with this program; if not, write to the Free Software * Foundation, Inc., 51 Franklin Street, Fifth Floor, Boston, MA 02110-1301 USA * */ /* This file was automatically generated --- DO NOT EDIT */ /* Generated on Sun Nov 25 07:35:46 EST 2012 */ #include "codelet-dft.h" #ifdef HAVE_FMA /* Generated by: ../../../genfft/gen_notw.native -fma -reorder-insns -schedule-for-pipeline -compact -variables 4 -pipeline-latency 4 -n 25 -name n1_25 -include n.h */ /* * This function contains 352 FP additions, 268 FP multiplications, * (or, 84 additions, 0 multiplications, 268 fused multiply/add), * 164 stack variables, 47 constants, and 100 memory accesses */ #include "n.h" static void n1_25(const R *ri, const R *ii, R *ro, R *io, stride is, stride os, INT v, INT ivs, INT ovs) { DK(KP803003575, +0.803003575438660414833440593570376004635464850); DK(KP554608978, +0.554608978404018097464974850792216217022558774); DK(KP248028675, +0.248028675328619457762448260696444630363259177); DK(KP726211448, +0.726211448929902658173535992263577167607493062); DK(KP525970792, +0.525970792408939708442463226536226366643874659); DK(KP992114701, +0.992114701314477831049793042785778521453036709); DK(KP851038619, +0.851038619207379630836264138867114231259902550); DK(KP912575812, +0.912575812670962425556968549836277086778922727); DK(KP912018591, +0.912018591466481957908415381764119056233607330); DK(KP943557151, +0.943557151597354104399655195398983005179443399); DK(KP614372930, +0.614372930789563808870829930444362096004872855); DK(KP621716863, +0.621716863012209892444754556304102309693593202); DK(KP994076283, +0.994076283785401014123185814696322018529298887); DK(KP734762448, +0.734762448793050413546343770063151342619912334); DK(KP772036680, +0.772036680810363904029489473607579825330539880); DK(KP126329378, +0.126329378446108174786050455341811215027378105); DK(KP827271945, +0.827271945972475634034355757144307982555673741); DK(KP949179823, +0.949179823508441261575555465843363271711583843); DK(KP860541664, +0.860541664367944677098261680920518816412804187); DK(KP557913902, +0.557913902031834264187699648465567037992437152); DK(KP249506682, +0.249506682107067890488084201715862638334226305); DK(KP681693190, +0.681693190061530575150324149145440022633095390); DK(KP560319534, +0.560319534973832390111614715371676131169633784); DK(KP998026728, +0.998026728428271561952336806863450553336905220); DK(KP906616052, +0.906616052148196230441134447086066874408359177); DK(KP968479752, +0.968479752739016373193524836781420152702090879); DK(KP845997307, +0.845997307939530944175097360758058292389769300); DK(KP470564281, +0.470564281212251493087595091036643380879947982); DK(KP062914667, +0.062914667253649757225485955897349402364686947); DK(KP921177326, +0.921177326965143320250447435415066029359282231); DK(KP833417178, +0.833417178328688677408962550243238843138996060); DK(KP541454447, +0.541454447536312777046285590082819509052033189); DK(KP242145790, +0.242145790282157779872542093866183953459003101); DK(KP683113946, +0.683113946453479238701949862233725244439656928); DK(KP559154169, +0.559154169276087864842202529084232643714075927); DK(KP968583161, +0.968583161128631119490168375464735813836012403); DK(KP904730450, +0.904730450839922351881287709692877908104763647); DK(KP831864738, +0.831864738706457140726048799369896829771167132); DK(KP871714437, +0.871714437527667770979999223229522602943903653); DK(KP939062505, +0.939062505817492352556001843133229685779824606); DK(KP549754652, +0.549754652192770074288023275540779861653779767); DK(KP634619297, +0.634619297544148100711287640319130485732531031); DK(KP256756360, +0.256756360367726783319498520922669048172391148); DK(KP951056516, +0.951056516295153572116439333379382143405698634); DK(KP559016994, +0.559016994374947424102293417182819058860154590); DK(KP250000000, +0.250000000000000000000000000000000000000000000); DK(KP618033988, +0.618033988749894848204586834365638117720309180); { INT i; for (i = v; i > 0; i = i - 1, ri = ri + ivs, ii = ii + ivs, ro = ro + ovs, io = io + ovs, MAKE_VOLATILE_STRIDE(100, is), MAKE_VOLATILE_STRIDE(100, os)) { E T3Y, T3U, T3W, T42, T44, T3X, T3R, T3V, T3Z, T43; { E T4Q, T1U, T9, T3b, T45, T3e, T46, T1D, T4P, T1R, Ts, T1K, T18, T1E, T4z; E T5f, T3z, T22, T4s, T5b, T3C, T2o, T3D, T2h, T4p, T5c, T4w, T5e, T3A, T29; E T2z, T2y, TL, T1L, T1r, T1F, T4a, T57, T3v, T2x, T4k, T55, T3s, T2T, T2D; E T4c, T3t, T2M, T4h, T54, T1v, T1C, T1Q; { E T1, T2, T3, T5, T6; T1 = ri[0]; T2 = ri[WS(is, 5)]; T3 = ri[WS(is, 20)]; T5 = ri[WS(is, 10)]; T6 = ri[WS(is, 15)]; { E T3a, T3c, T1y, T1z, T1A, T39, T4, T1S, T1B, T3d; T1v = ii[0]; T4 = T2 + T3; T1S = T2 - T3; { E T7, T1T, T8, T1w, T1x; T7 = T5 + T6; T1T = T5 - T6; T1w = ii[WS(is, 5)]; T1x = ii[WS(is, 20)]; T4Q = FNMS(KP618033988, T1S, T1T); T1U = FMA(KP618033988, T1T, T1S); T8 = T4 + T7; T3a = T4 - T7; T3c = T1w - T1x; T1y = T1w + T1x; T1z = ii[WS(is, 10)]; T1A = ii[WS(is, 15)]; T39 = FNMS(KP250000000, T8, T1); T9 = T1 + T8; } T1B = T1z + T1A; T3d = T1z - T1A; T3b = FMA(KP559016994, T3a, T39); T45 = FNMS(KP559016994, T3a, T39); T3e = FMA(KP618033988, T3d, T3c); T46 = FNMS(KP618033988, T3c, T3d); T1C = T1y + T1B; T1Q = T1y - T1B; } } { E T24, T23, T28, T4v; { E Ta, TQ, Tj, TZ, T1Z, T20, Th, T26, T27, T1X, TX, T2l, T2m, Tq, T2c; E T2e, T12, T15, T2f, T1P, TT, TW; Ta = ri[WS(is, 1)]; T1P = FNMS(KP250000000, T1C, T1v); T1D = T1v + T1C; TQ = ii[WS(is, 1)]; Tj = ri[WS(is, 4)]; T4P = FNMS(KP559016994, T1Q, T1P); T1R = FMA(KP559016994, T1Q, T1P); TZ = ii[WS(is, 4)]; { E Tb, Tc, Te, Tf; Tb = ri[WS(is, 6)]; Tc = ri[WS(is, 21)]; Te = ri[WS(is, 11)]; Tf = ri[WS(is, 16)]; { E TR, Td, Tg, TS, TU, TV; TR = ii[WS(is, 6)]; T1Z = Tc - Tb; Td = Tb + Tc; T20 = Tf - Te; Tg = Te + Tf; TS = ii[WS(is, 21)]; TU = ii[WS(is, 11)]; TV = ii[WS(is, 16)]; Th = Td + Tg; T24 = Td - Tg; T26 = TR - TS; TT = TR + TS; TW = TU + TV; T27 = TV - TU; } } { E Tk, Tl, Tn, To; Tk = ri[WS(is, 9)]; T1X = TT - TW; TX = TT + TW; Tl = ri[WS(is, 24)]; Tn = ri[WS(is, 14)]; To = ri[WS(is, 19)]; { E T10, Tm, Tp, T11, T13, T14; T10 = ii[WS(is, 9)]; T2l = Tl - Tk; Tm = Tk + Tl; T2m = To - Tn; Tp = Tn + To; T11 = ii[WS(is, 24)]; T13 = ii[WS(is, 14)]; T14 = ii[WS(is, 19)]; Tq = Tm + Tp; T2c = Tm - Tp; T2e = T11 - T10; T12 = T10 + T11; T15 = T13 + T14; T2f = T14 - T13; } } { E T2j, T2b, T1W, T21, T4y, T2i; { E Ti, T16, Tr, TY, T17; T23 = FNMS(KP250000000, Th, Ta); Ti = Ta + Th; T2j = T15 - T12; T16 = T12 + T15; Tr = Tj + Tq; T2b = FMS(KP250000000, Tq, Tj); T1W = FNMS(KP250000000, TX, TQ); TY = TQ + TX; T21 = FMA(KP618033988, T20, T1Z); T4y = FNMS(KP618033988, T1Z, T20); T2i = FNMS(KP250000000, T16, TZ); T17 = TZ + T16; Ts = Ti + Tr; T1K = Ti - Tr; T18 = TY - T17; T1E = TY + T17; } { E T2n, T4r, T4x, T1Y; T2n = FMA(KP618033988, T2m, T2l); T4r = FNMS(KP618033988, T2l, T2m); T4x = FNMS(KP559016994, T1X, T1W); T1Y = FMA(KP559016994, T1X, T1W); { E T4o, T2g, T2d, T4n, T4q, T2k; T4o = FNMS(KP618033988, T2e, T2f); T2g = FMA(KP618033988, T2f, T2e); T4z = FMA(KP951056516, T4y, T4x); T5f = FNMS(KP951056516, T4y, T4x); T3z = FNMS(KP951056516, T21, T1Y); T22 = FMA(KP951056516, T21, T1Y); T4q = FMA(KP559016994, T2j, T2i); T2k = FNMS(KP559016994, T2j, T2i); T4s = FMA(KP951056516, T4r, T4q); T5b = FNMS(KP951056516, T4r, T4q); T3C = FNMS(KP951056516, T2n, T2k); T2o = FMA(KP951056516, T2n, T2k); T2d = FNMS(KP559016994, T2c, T2b); T4n = FMA(KP559016994, T2c, T2b); T28 = FNMS(KP618033988, T27, T26); T4v = FMA(KP618033988, T26, T27); T3D = FNMS(KP951056516, T2g, T2d); T2h = FMA(KP951056516, T2g, T2d); T4p = FMA(KP951056516, T4o, T4n); T5c = FNMS(KP951056516, T4o, T4n); } } } } { E Tt, T19, TC, T1i, T2u, T2v, TA, T2B, T2C, T2s, T1g, T2J, T2K, TJ, T2O; E T2Q, T1l, T1o, T2R; { E T4u, T25, T1c, T1f; Tt = ri[WS(is, 2)]; T19 = ii[WS(is, 2)]; TC = ri[WS(is, 3)]; T4u = FNMS(KP559016994, T24, T23); T25 = FMA(KP559016994, T24, T23); T1i = ii[WS(is, 3)]; { E Tu, Tv, Tx, Ty; Tu = ri[WS(is, 7)]; T4w = FNMS(KP951056516, T4v, T4u); T5e = FMA(KP951056516, T4v, T4u); T3A = FNMS(KP951056516, T28, T25); T29 = FMA(KP951056516, T28, T25); Tv = ri[WS(is, 22)]; Tx = ri[WS(is, 12)]; Ty = ri[WS(is, 17)]; { E T1a, Tw, Tz, T1b, T1d, T1e; T1a = ii[WS(is, 7)]; T2u = Tv - Tu; Tw = Tu + Tv; T2v = Ty - Tx; Tz = Tx + Ty; T1b = ii[WS(is, 22)]; T1d = ii[WS(is, 12)]; T1e = ii[WS(is, 17)]; TA = Tw + Tz; T2z = Tz - Tw; T2B = T1b - T1a; T1c = T1a + T1b; T1f = T1d + T1e; T2C = T1d - T1e; } } { E TD, TE, TG, TH; TD = ri[WS(is, 8)]; T2s = T1f - T1c; T1g = T1c + T1f; TE = ri[WS(is, 23)]; TG = ri[WS(is, 13)]; TH = ri[WS(is, 18)]; { E T1j, TF, TI, T1k, T1m, T1n; T1j = ii[WS(is, 8)]; T2J = TD - TE; TF = TD + TE; T2K = TG - TH; TI = TG + TH; T1k = ii[WS(is, 23)]; T1m = ii[WS(is, 13)]; T1n = ii[WS(is, 18)]; TJ = TF + TI; T2O = TI - TF; T2Q = T1k - T1j; T1l = T1j + T1k; T1o = T1m + T1n; T2R = T1n - T1m; } } } { E T2H, T2N, T2r, T2w, T49, T2G; { E TB, T1p, TK, T1h, T1q; T2y = FNMS(KP250000000, TA, Tt); TB = Tt + TA; T2H = T1o - T1l; T1p = T1l + T1o; TK = TC + TJ; T2N = FNMS(KP250000000, TJ, TC); T2r = FNMS(KP250000000, T1g, T19); T1h = T19 + T1g; T2w = FMA(KP618033988, T2v, T2u); T49 = FNMS(KP618033988, T2u, T2v); T2G = FNMS(KP250000000, T1p, T1i); T1q = T1i + T1p; TL = TB + TK; T1L = TB - TK; T1r = T1h - T1q; T1F = T1h + T1q; } { E T2S, T4j, T48, T2t; T2S = FMA(KP618033988, T2R, T2Q); T4j = FNMS(KP618033988, T2Q, T2R); T48 = FMA(KP559016994, T2s, T2r); T2t = FNMS(KP559016994, T2s, T2r); { E T4g, T2L, T2I, T4f, T4i, T2P; T4g = FNMS(KP618033988, T2J, T2K); T2L = FMA(KP618033988, T2K, T2J); T4a = FMA(KP951056516, T49, T48); T57 = FNMS(KP951056516, T49, T48); T3v = FNMS(KP951056516, T2w, T2t); T2x = FMA(KP951056516, T2w, T2t); T4i = FMA(KP559016994, T2O, T2N); T2P = FNMS(KP559016994, T2O, T2N); T4k = FNMS(KP951056516, T4j, T4i); T55 = FMA(KP951056516, T4j, T4i); T3s = FMA(KP951056516, T2S, T2P); T2T = FNMS(KP951056516, T2S, T2P); T2I = FNMS(KP559016994, T2H, T2G); T4f = FMA(KP559016994, T2H, T2G); T2D = FNMS(KP618033988, T2C, T2B); T4c = FMA(KP618033988, T2B, T2C); T3t = FMA(KP951056516, T2L, T2I); T2M = FNMS(KP951056516, T2L, T2I); T4h = FNMS(KP951056516, T4g, T4f); T54 = FMA(KP951056516, T4g, T4f); } } } } } { E T4d, T58, T3w, T3H, T3r, T3k, T36, T38, T3o, T3q, T3j, T2Z, T37; { E T2E, T1s, T1u, TP, T1t; { E TM, TO, TN, T4b, T2A; TM = Ts + TL; TO = Ts - TL; T4b = FMA(KP559016994, T2z, T2y); T2A = FNMS(KP559016994, T2z, T2y); TN = FNMS(KP250000000, TM, T9); T4d = FMA(KP951056516, T4c, T4b); T58 = FNMS(KP951056516, T4c, T4b); T3w = FMA(KP951056516, T2D, T2A); T2E = FNMS(KP951056516, T2D, T2A); T1s = FMA(KP618033988, T1r, T18); T1u = FNMS(KP618033988, T18, T1r); ro[0] = T9 + TM; TP = FMA(KP559016994, TO, TN); T1t = FNMS(KP559016994, TO, TN); } { E T1J, T1N, T1M, T1O, T1G, T1I, T1H; T1G = T1E + T1F; T1I = T1E - T1F; ro[WS(os, 15)] = FMA(KP951056516, T1u, T1t); ro[WS(os, 10)] = FNMS(KP951056516, T1u, T1t); ro[WS(os, 5)] = FMA(KP951056516, T1s, TP); ro[WS(os, 20)] = FNMS(KP951056516, T1s, TP); T1H = FNMS(KP250000000, T1G, T1D); io[0] = T1D + T1G; T1J = FMA(KP559016994, T1I, T1H); T1N = FNMS(KP559016994, T1I, T1H); T1M = FMA(KP618033988, T1L, T1K); T1O = FNMS(KP618033988, T1K, T1L); { E T1V, T3f, T3m, T3n, T2W, T2Y, T32, T3g, T3h, T35, T3i, T2X; T3H = FMA(KP951056516, T1U, T1R); T1V = FNMS(KP951056516, T1U, T1R); T3f = FMA(KP951056516, T3e, T3b); T3r = FNMS(KP951056516, T3e, T3b); io[WS(os, 15)] = FNMS(KP951056516, T1O, T1N); io[WS(os, 10)] = FMA(KP951056516, T1O, T1N); io[WS(os, 20)] = FMA(KP951056516, T1M, T1J); io[WS(os, 5)] = FNMS(KP951056516, T1M, T1J); { E T30, T2a, T2p, T31, T33, T2F, T2U, T34, T2q, T2V; T30 = FMA(KP256756360, T22, T29); T2a = FNMS(KP256756360, T29, T22); T2p = FMA(KP634619297, T2o, T2h); T31 = FNMS(KP634619297, T2h, T2o); T33 = FMA(KP549754652, T2x, T2E); T2F = FNMS(KP549754652, T2E, T2x); T2U = FNMS(KP939062505, T2T, T2M); T34 = FMA(KP939062505, T2M, T2T); T3m = FNMS(KP871714437, T2p, T2a); T2q = FMA(KP871714437, T2p, T2a); T3n = FNMS(KP831864738, T2U, T2F); T2V = FMA(KP831864738, T2U, T2F); T2W = FMA(KP904730450, T2V, T2q); T2Y = FNMS(KP904730450, T2V, T2q); T32 = FNMS(KP871714437, T31, T30); T3g = FMA(KP871714437, T31, T30); T3h = FMA(KP831864738, T34, T33); T35 = FNMS(KP831864738, T34, T33); } io[WS(os, 1)] = FMA(KP968583161, T2W, T1V); T3i = FMA(KP904730450, T3h, T3g); T3k = FNMS(KP904730450, T3h, T3g); T36 = FMA(KP559154169, T35, T32); T38 = FNMS(KP683113946, T32, T35); ro[WS(os, 1)] = FMA(KP968583161, T3i, T3f); T2X = FNMS(KP242145790, T2W, T1V); T3o = FMA(KP559154169, T3n, T3m); T3q = FNMS(KP683113946, T3m, T3n); T3j = FNMS(KP242145790, T3i, T3f); T2Z = FMA(KP541454447, T2Y, T2X); T37 = FNMS(KP541454447, T2Y, T2X); } } } { E T47, T4R, T5A, T5w, T5y, T5E, T5G, T5z, T5t, T5x; { E T53, T5j, T5u, T5v, T5i, T5D, T5m, T5p, T5C, T3p, T3l, T5s, T5q, T5r; T47 = FMA(KP951056516, T46, T45); T53 = FNMS(KP951056516, T46, T45); T3p = FNMS(KP541454447, T3k, T3j); T3l = FMA(KP541454447, T3k, T3j); io[WS(os, 16)] = FNMS(KP833417178, T38, T37); io[WS(os, 11)] = FMA(KP833417178, T38, T37); io[WS(os, 21)] = FMA(KP921177326, T36, T2Z); io[WS(os, 6)] = FNMS(KP921177326, T36, T2Z); ro[WS(os, 11)] = FNMS(KP833417178, T3q, T3p); ro[WS(os, 16)] = FMA(KP833417178, T3q, T3p); ro[WS(os, 21)] = FNMS(KP921177326, T3o, T3l); ro[WS(os, 6)] = FMA(KP921177326, T3o, T3l); T5j = FMA(KP951056516, T4Q, T4P); T4R = FNMS(KP951056516, T4Q, T4P); { E T5k, T56, T59, T5l, T5n, T5d, T5g, T5o, T5a, T5h; T5k = FNMS(KP062914667, T54, T55); T56 = FMA(KP062914667, T55, T54); T59 = FMA(KP634619297, T58, T57); T5l = FNMS(KP634619297, T57, T58); T5n = FNMS(KP470564281, T5b, T5c); T5d = FMA(KP470564281, T5c, T5b); T5g = FMA(KP549754652, T5f, T5e); T5o = FNMS(KP549754652, T5e, T5f); T5u = FNMS(KP845997307, T59, T56); T5a = FMA(KP845997307, T59, T56); T5v = FNMS(KP968479752, T5g, T5d); T5h = FMA(KP968479752, T5g, T5d); T5i = FMA(KP906616052, T5h, T5a); T5A = FNMS(KP906616052, T5h, T5a); T5D = FNMS(KP845997307, T5l, T5k); T5m = FMA(KP845997307, T5l, T5k); T5p = FMA(KP968479752, T5o, T5n); T5C = FNMS(KP968479752, T5o, T5n); } ro[WS(os, 2)] = FMA(KP998026728, T5i, T53); T5s = FMA(KP906616052, T5p, T5m); T5q = FNMS(KP906616052, T5p, T5m); T5w = FNMS(KP560319534, T5v, T5u); T5y = FMA(KP681693190, T5u, T5v); T5E = FNMS(KP681693190, T5D, T5C); T5G = FMA(KP560319534, T5C, T5D); T5r = FMA(KP249506682, T5q, T5j); io[WS(os, 2)] = FNMS(KP998026728, T5q, T5j); T5z = FNMS(KP249506682, T5i, T53); T5t = FNMS(KP557913902, T5s, T5r); T5x = FMA(KP557913902, T5s, T5r); } { E T4W, T4M, T4O, T50, T52, T4V, T4F, T4N; { E T4Y, T4Z, T4C, T4E, T4I, T4T, T4S, T4L, T5F, T5B, T4U, T4D; T5F = FMA(KP557913902, T5A, T5z); T5B = FNMS(KP557913902, T5A, T5z); io[WS(os, 7)] = FMA(KP860541664, T5y, T5x); io[WS(os, 22)] = FNMS(KP860541664, T5y, T5x); io[WS(os, 17)] = FMA(KP949179823, T5w, T5t); io[WS(os, 12)] = FNMS(KP949179823, T5w, T5t); ro[WS(os, 12)] = FNMS(KP949179823, T5G, T5F); ro[WS(os, 17)] = FMA(KP949179823, T5G, T5F); ro[WS(os, 7)] = FNMS(KP860541664, T5E, T5B); ro[WS(os, 22)] = FMA(KP860541664, T5E, T5B); { E T4J, T4e, T4l, T4K, T4G, T4t, T4A, T4H, T4m, T4B; T4J = FNMS(KP062914667, T4a, T4d); T4e = FMA(KP062914667, T4d, T4a); T4l = FNMS(KP827271945, T4k, T4h); T4K = FMA(KP827271945, T4h, T4k); T4G = FNMS(KP126329378, T4p, T4s); T4t = FMA(KP126329378, T4s, T4p); T4A = FMA(KP939062505, T4z, T4w); T4H = FNMS(KP939062505, T4w, T4z); T4Y = FNMS(KP772036680, T4l, T4e); T4m = FMA(KP772036680, T4l, T4e); T4Z = FNMS(KP734762448, T4A, T4t); T4B = FMA(KP734762448, T4A, T4t); T4C = FMA(KP994076283, T4B, T4m); T4E = FNMS(KP994076283, T4B, T4m); T4I = FMA(KP734762448, T4H, T4G); T4T = FNMS(KP734762448, T4H, T4G); T4S = FMA(KP772036680, T4K, T4J); T4L = FNMS(KP772036680, T4K, T4J); } ro[WS(os, 3)] = FMA(KP998026728, T4C, T47); T4U = FMA(KP994076283, T4T, T4S); T4W = FNMS(KP994076283, T4T, T4S); T4M = FNMS(KP621716863, T4L, T4I); T4O = FMA(KP614372930, T4I, T4L); io[WS(os, 3)] = FNMS(KP998026728, T4U, T4R); T4D = FNMS(KP249506682, T4C, T47); T50 = FMA(KP614372930, T4Z, T4Y); T52 = FNMS(KP621716863, T4Y, T4Z); T4V = FMA(KP249506682, T4U, T4R); T4F = FNMS(KP557913902, T4E, T4D); T4N = FMA(KP557913902, T4E, T4D); } { E T3S, T3T, T3G, T41, T3K, T3N, T40, T51, T4X, T3Q, T3O, T3P; T51 = FMA(KP557913902, T4W, T4V); T4X = FNMS(KP557913902, T4W, T4V); ro[WS(os, 18)] = FNMS(KP949179823, T4O, T4N); ro[WS(os, 13)] = FMA(KP949179823, T4O, T4N); ro[WS(os, 8)] = FMA(KP943557151, T4M, T4F); ro[WS(os, 23)] = FNMS(KP943557151, T4M, T4F); io[WS(os, 8)] = FMA(KP943557151, T52, T51); io[WS(os, 23)] = FNMS(KP943557151, T52, T51); io[WS(os, 18)] = FNMS(KP949179823, T50, T4X); io[WS(os, 13)] = FMA(KP949179823, T50, T4X); { E T3I, T3u, T3x, T3J, T3L, T3B, T3E, T3M, T3y, T3F; T3I = FMA(KP126329378, T3s, T3t); T3u = FNMS(KP126329378, T3t, T3s); T3x = FNMS(KP470564281, T3w, T3v); T3J = FMA(KP470564281, T3v, T3w); T3L = FNMS(KP634619297, T3z, T3A); T3B = FMA(KP634619297, T3A, T3z); T3E = FNMS(KP827271945, T3D, T3C); T3M = FMA(KP827271945, T3C, T3D); T3S = FMA(KP912018591, T3x, T3u); T3y = FNMS(KP912018591, T3x, T3u); T3T = FMA(KP912575812, T3E, T3B); T3F = FNMS(KP912575812, T3E, T3B); T3G = FNMS(KP851038619, T3F, T3y); T3Y = FMA(KP851038619, T3F, T3y); T41 = FNMS(KP912018591, T3J, T3I); T3K = FMA(KP912018591, T3J, T3I); T3N = FMA(KP912575812, T3M, T3L); T40 = FNMS(KP912575812, T3M, T3L); } ro[WS(os, 4)] = FNMS(KP992114701, T3G, T3r); T3Q = FNMS(KP851038619, T3N, T3K); T3O = FMA(KP851038619, T3N, T3K); T3U = FNMS(KP525970792, T3T, T3S); T3W = FMA(KP726211448, T3S, T3T); T42 = FNMS(KP726211448, T41, T40); T44 = FMA(KP525970792, T40, T41); T3P = FMA(KP248028675, T3O, T3H); io[WS(os, 4)] = FNMS(KP992114701, T3O, T3H); T3X = FMA(KP248028675, T3G, T3r); T3R = FNMS(KP554608978, T3Q, T3P); T3V = FMA(KP554608978, T3Q, T3P); } } } } } T3Z = FMA(KP554608978, T3Y, T3X); T43 = FNMS(KP554608978, T3Y, T3X); io[WS(os, 9)] = FNMS(KP803003575, T3W, T3V); io[WS(os, 24)] = FMA(KP803003575, T3W, T3V); io[WS(os, 19)] = FNMS(KP943557151, T3U, T3R); io[WS(os, 14)] = FMA(KP943557151, T3U, T3R); ro[WS(os, 14)] = FNMS(KP943557151, T44, T43); ro[WS(os, 19)] = FMA(KP943557151, T44, T43); ro[WS(os, 24)] = FMA(KP803003575, T42, T3Z); ro[WS(os, 9)] = FNMS(KP803003575, T42, T3Z); } } } static const kdft_desc desc = { 25, "n1_25", {84, 0, 268, 0}, &GENUS, 0, 0, 0, 0 }; void X(codelet_n1_25) (planner *p) { X(kdft_register) (p, n1_25, &desc); } #else /* HAVE_FMA */ /* Generated by: ../../../genfft/gen_notw.native -compact -variables 4 -pipeline-latency 4 -n 25 -name n1_25 -include n.h */ /* * This function contains 352 FP additions, 184 FP multiplications, * (or, 260 additions, 92 multiplications, 92 fused multiply/add), * 101 stack variables, 20 constants, and 100 memory accesses */ #include "n.h" static void n1_25(const R *ri, const R *ii, R *ro, R *io, stride is, stride os, INT v, INT ivs, INT ovs) { DK(KP425779291, +0.425779291565072648862502445744251703979973042); DK(KP904827052, +0.904827052466019527713668647932697593970413911); DK(KP637423989, +0.637423989748689710176712811676016195434917298); DK(KP770513242, +0.770513242775789230803009636396177847271667672); DK(KP998026728, +0.998026728428271561952336806863450553336905220); DK(KP062790519, +0.062790519529313376076178224565631133122484832); DK(KP992114701, +0.992114701314477831049793042785778521453036709); DK(KP125333233, +0.125333233564304245373118759816508793942918247); DK(KP684547105, +0.684547105928688673732283357621209269889519233); DK(KP728968627, +0.728968627421411523146730319055259111372571664); DK(KP481753674, +0.481753674101715274987191502872129653528542010); DK(KP876306680, +0.876306680043863587308115903922062583399064238); DK(KP844327925, +0.844327925502015078548558063966681505381659241); DK(KP535826794, +0.535826794978996618271308767867639978063575346); DK(KP248689887, +0.248689887164854788242283746006447968417567406); DK(KP968583161, +0.968583161128631119490168375464735813836012403); DK(KP250000000, +0.250000000000000000000000000000000000000000000); DK(KP559016994, +0.559016994374947424102293417182819058860154590); DK(KP587785252, +0.587785252292473129168705954639072768597652438); DK(KP951056516, +0.951056516295153572116439333379382143405698634); { INT i; for (i = v; i > 0; i = i - 1, ri = ri + ivs, ii = ii + ivs, ro = ro + ovs, io = io + ovs, MAKE_VOLATILE_STRIDE(100, is), MAKE_VOLATILE_STRIDE(100, os)) { E T9, T4u, T2T, TP, T3H, TW, T5y, T3I, T2Q, T4v, Ti, Tr, Ts, T5m, T5n; E T5v, T18, T4G, T34, T3M, T1G, T4J, T38, T3T, T1v, T4K, T37, T3W, T1j, T4H; E T35, T3P, TB, TK, TL, T5p, T5q, T5w, T1T, T4N, T3c, T41, T2r, T4Q, T3e; E T4b, T2g, T4R, T3f, T48, T24, T4O, T3b, T44; { E T1, T4, T7, T8, T2S, T2R, TN, TO; T1 = ri[0]; { E T2, T3, T5, T6; T2 = ri[WS(is, 5)]; T3 = ri[WS(is, 20)]; T4 = T2 + T3; T5 = ri[WS(is, 10)]; T6 = ri[WS(is, 15)]; T7 = T5 + T6; T8 = T4 + T7; T2S = T5 - T6; T2R = T2 - T3; } T9 = T1 + T8; T4u = FNMS(KP587785252, T2R, KP951056516 * T2S); T2T = FMA(KP951056516, T2R, KP587785252 * T2S); TN = KP559016994 * (T4 - T7); TO = FNMS(KP250000000, T8, T1); TP = TN + TO; T3H = TO - TN; } { E T2N, T2K, T2L, TS, T2O, TV, T2M, T2P; T2N = ii[0]; { E TQ, TR, TT, TU; TQ = ii[WS(is, 5)]; TR = ii[WS(is, 20)]; T2K = TQ + TR; TT = ii[WS(is, 10)]; TU = ii[WS(is, 15)]; T2L = TT + TU; TS = TQ - TR; T2O = T2K + T2L; TV = TT - TU; } TW = FMA(KP951056516, TS, KP587785252 * TV); T5y = T2N + T2O; T3I = FNMS(KP587785252, TS, KP951056516 * TV); T2M = KP559016994 * (T2K - T2L); T2P = FNMS(KP250000000, T2O, T2N); T2Q = T2M + T2P; T4v = T2P - T2M; } { E Ta, T1c, Tj, T1z, Th, T1h, TY, T1g, T13, T1d, T16, T1b, Tq, T1E, T1l; E T1D, T1q, T1A, T1t, T1y; Ta = ri[WS(is, 1)]; T1c = ii[WS(is, 1)]; Tj = ri[WS(is, 4)]; T1z = ii[WS(is, 4)]; { E Tb, Tc, Td, Te, Tf, Tg; Tb = ri[WS(is, 6)]; Tc = ri[WS(is, 21)]; Td = Tb + Tc; Te = ri[WS(is, 11)]; Tf = ri[WS(is, 16)]; Tg = Te + Tf; Th = Td + Tg; T1h = Te - Tf; TY = KP559016994 * (Td - Tg); T1g = Tb - Tc; } { E T11, T12, T19, T14, T15, T1a; T11 = ii[WS(is, 6)]; T12 = ii[WS(is, 21)]; T19 = T11 + T12; T14 = ii[WS(is, 11)]; T15 = ii[WS(is, 16)]; T1a = T14 + T15; T13 = T11 - T12; T1d = T19 + T1a; T16 = T14 - T15; T1b = KP559016994 * (T19 - T1a); } { E Tk, Tl, Tm, Tn, To, Tp; Tk = ri[WS(is, 9)]; Tl = ri[WS(is, 24)]; Tm = Tk + Tl; Tn = ri[WS(is, 14)]; To = ri[WS(is, 19)]; Tp = Tn + To; Tq = Tm + Tp; T1E = Tn - To; T1l = KP559016994 * (Tm - Tp); T1D = Tk - Tl; } { E T1o, T1p, T1w, T1r, T1s, T1x; T1o = ii[WS(is, 9)]; T1p = ii[WS(is, 24)]; T1w = T1o + T1p; T1r = ii[WS(is, 14)]; T1s = ii[WS(is, 19)]; T1x = T1r + T1s; T1q = T1o - T1p; T1A = T1w + T1x; T1t = T1r - T1s; T1y = KP559016994 * (T1w - T1x); } Ti = Ta + Th; Tr = Tj + Tq; Ts = Ti + Tr; T5m = T1c + T1d; T5n = T1z + T1A; T5v = T5m + T5n; { E T17, T3L, T10, T3K, TZ; T17 = FMA(KP951056516, T13, KP587785252 * T16); T3L = FNMS(KP587785252, T13, KP951056516 * T16); TZ = FNMS(KP250000000, Th, Ta); T10 = TY + TZ; T3K = TZ - TY; T18 = T10 + T17; T4G = T3K + T3L; T34 = T10 - T17; T3M = T3K - T3L; } { E T1F, T3R, T1C, T3S, T1B; T1F = FMA(KP951056516, T1D, KP587785252 * T1E); T3R = FNMS(KP587785252, T1D, KP951056516 * T1E); T1B = FNMS(KP250000000, T1A, T1z); T1C = T1y + T1B; T3S = T1B - T1y; T1G = T1C - T1F; T4J = T3S - T3R; T38 = T1F + T1C; T3T = T3R + T3S; } { E T1u, T3V, T1n, T3U, T1m; T1u = FMA(KP951056516, T1q, KP587785252 * T1t); T3V = FNMS(KP587785252, T1q, KP951056516 * T1t); T1m = FNMS(KP250000000, Tq, Tj); T1n = T1l + T1m; T3U = T1m - T1l; T1v = T1n + T1u; T4K = T3U + T3V; T37 = T1n - T1u; T3W = T3U - T3V; } { E T1i, T3N, T1f, T3O, T1e; T1i = FMA(KP951056516, T1g, KP587785252 * T1h); T3N = FNMS(KP587785252, T1g, KP951056516 * T1h); T1e = FNMS(KP250000000, T1d, T1c); T1f = T1b + T1e; T3O = T1e - T1b; T1j = T1f - T1i; T4H = T3O - T3N; T35 = T1i + T1f; T3P = T3N + T3O; } } { E Tt, T1X, TC, T2k, TA, T22, T1J, T21, T1O, T1Y, T1R, T1W, TJ, T2p, T26; E T2o, T2b, T2l, T2e, T2j; Tt = ri[WS(is, 2)]; T1X = ii[WS(is, 2)]; TC = ri[WS(is, 3)]; T2k = ii[WS(is, 3)]; { E Tu, Tv, Tw, Tx, Ty, Tz; Tu = ri[WS(is, 7)]; Tv = ri[WS(is, 22)]; Tw = Tu + Tv; Tx = ri[WS(is, 12)]; Ty = ri[WS(is, 17)]; Tz = Tx + Ty; TA = Tw + Tz; T22 = Tx - Ty; T1J = KP559016994 * (Tw - Tz); T21 = Tu - Tv; } { E T1M, T1N, T1U, T1P, T1Q, T1V; T1M = ii[WS(is, 7)]; T1N = ii[WS(is, 22)]; T1U = T1M + T1N; T1P = ii[WS(is, 12)]; T1Q = ii[WS(is, 17)]; T1V = T1P + T1Q; T1O = T1M - T1N; T1Y = T1U + T1V; T1R = T1P - T1Q; T1W = KP559016994 * (T1U - T1V); } { E TD, TE, TF, TG, TH, TI; TD = ri[WS(is, 8)]; TE = ri[WS(is, 23)]; TF = TD + TE; TG = ri[WS(is, 13)]; TH = ri[WS(is, 18)]; TI = TG + TH; TJ = TF + TI; T2p = TG - TH; T26 = KP559016994 * (TF - TI); T2o = TD - TE; } { E T29, T2a, T2h, T2c, T2d, T2i; T29 = ii[WS(is, 8)]; T2a = ii[WS(is, 23)]; T2h = T29 + T2a; T2c = ii[WS(is, 13)]; T2d = ii[WS(is, 18)]; T2i = T2c + T2d; T2b = T29 - T2a; T2l = T2h + T2i; T2e = T2c - T2d; T2j = KP559016994 * (T2h - T2i); } TB = Tt + TA; TK = TC + TJ; TL = TB + TK; T5p = T1X + T1Y; T5q = T2k + T2l; T5w = T5p + T5q; { E T1S, T40, T1L, T3Z, T1K; T1S = FMA(KP951056516, T1O, KP587785252 * T1R); T40 = FNMS(KP587785252, T1O, KP951056516 * T1R); T1K = FNMS(KP250000000, TA, Tt); T1L = T1J + T1K; T3Z = T1K - T1J; T1T = T1L + T1S; T4N = T3Z + T40; T3c = T1L - T1S; T41 = T3Z - T40; } { E T2q, T49, T2n, T4a, T2m; T2q = FMA(KP951056516, T2o, KP587785252 * T2p); T49 = FNMS(KP587785252, T2o, KP951056516 * T2p); T2m = FNMS(KP250000000, T2l, T2k); T2n = T2j + T2m; T4a = T2m - T2j; T2r = T2n - T2q; T4Q = T4a - T49; T3e = T2q + T2n; T4b = T49 + T4a; } { E T2f, T47, T28, T46, T27; T2f = FMA(KP951056516, T2b, KP587785252 * T2e); T47 = FNMS(KP587785252, T2b, KP951056516 * T2e); T27 = FNMS(KP250000000, TJ, TC); T28 = T26 + T27; T46 = T27 - T26; T2g = T28 + T2f; T4R = T46 + T47; T3f = T28 - T2f; T48 = T46 - T47; } { E T23, T42, T20, T43, T1Z; T23 = FMA(KP951056516, T21, KP587785252 * T22); T42 = FNMS(KP587785252, T21, KP951056516 * T22); T1Z = FNMS(KP250000000, T1Y, T1X); T20 = T1W + T1Z; T43 = T1Z - T1W; T24 = T20 - T23; T4O = T43 - T42; T3b = T23 + T20; T44 = T42 + T43; } } { E T5j, TM, T5k, T5s, T5u, T5o, T5r, T5t, T5l; T5j = KP559016994 * (Ts - TL); TM = Ts + TL; T5k = FNMS(KP250000000, TM, T9); T5o = T5m - T5n; T5r = T5p - T5q; T5s = FMA(KP951056516, T5o, KP587785252 * T5r); T5u = FNMS(KP587785252, T5o, KP951056516 * T5r); ro[0] = T9 + TM; T5t = T5k - T5j; ro[WS(os, 10)] = T5t - T5u; ro[WS(os, 15)] = T5t + T5u; T5l = T5j + T5k; ro[WS(os, 20)] = T5l - T5s; ro[WS(os, 5)] = T5l + T5s; } { E T5x, T5z, T5A, T5E, T5F, T5C, T5D, T5G, T5B; T5x = KP559016994 * (T5v - T5w); T5z = T5v + T5w; T5A = FNMS(KP250000000, T5z, T5y); T5C = Ti - Tr; T5D = TB - TK; T5E = FMA(KP951056516, T5C, KP587785252 * T5D); T5F = FNMS(KP587785252, T5C, KP951056516 * T5D); io[0] = T5y + T5z; T5G = T5A - T5x; io[WS(os, 10)] = T5F + T5G; io[WS(os, 15)] = T5G - T5F; T5B = T5x + T5A; io[WS(os, 5)] = T5B - T5E; io[WS(os, 20)] = T5E + T5B; } { E TX, T2U, T2u, T2Z, T2v, T2Y, T2A, T2V, T2D, T2J; TX = TP + TW; T2U = T2Q - T2T; { E T1k, T1H, T1I, T25, T2s, T2t; T1k = FMA(KP968583161, T18, KP248689887 * T1j); T1H = FMA(KP535826794, T1v, KP844327925 * T1G); T1I = T1k + T1H; T25 = FMA(KP876306680, T1T, KP481753674 * T24); T2s = FMA(KP728968627, T2g, KP684547105 * T2r); T2t = T25 + T2s; T2u = T1I + T2t; T2Z = T25 - T2s; T2v = KP559016994 * (T1I - T2t); T2Y = T1k - T1H; } { E T2y, T2z, T2H, T2B, T2C, T2I; T2y = FNMS(KP248689887, T18, KP968583161 * T1j); T2z = FNMS(KP844327925, T1v, KP535826794 * T1G); T2H = T2y + T2z; T2B = FNMS(KP481753674, T1T, KP876306680 * T24); T2C = FNMS(KP684547105, T2g, KP728968627 * T2r); T2I = T2B + T2C; T2A = T2y - T2z; T2V = T2H + T2I; T2D = T2B - T2C; T2J = KP559016994 * (T2H - T2I); } ro[WS(os, 1)] = TX + T2u; io[WS(os, 1)] = T2U + T2V; { E T2E, T2G, T2x, T2F, T2w; T2E = FMA(KP951056516, T2A, KP587785252 * T2D); T2G = FNMS(KP587785252, T2A, KP951056516 * T2D); T2w = FNMS(KP250000000, T2u, TX); T2x = T2v + T2w; T2F = T2w - T2v; ro[WS(os, 21)] = T2x - T2E; ro[WS(os, 16)] = T2F + T2G; ro[WS(os, 6)] = T2x + T2E; ro[WS(os, 11)] = T2F - T2G; } { E T30, T31, T2X, T32, T2W; T30 = FMA(KP951056516, T2Y, KP587785252 * T2Z); T31 = FNMS(KP587785252, T2Y, KP951056516 * T2Z); T2W = FNMS(KP250000000, T2V, T2U); T2X = T2J + T2W; T32 = T2W - T2J; io[WS(os, 6)] = T2X - T30; io[WS(os, 16)] = T32 - T31; io[WS(os, 21)] = T30 + T2X; io[WS(os, 11)] = T31 + T32; } } { E T4F, T52, T4U, T5b, T56, T57, T51, T5f, T53, T5e; T4F = T3H + T3I; T52 = T4v - T4u; { E T4I, T4L, T4M, T4P, T4S, T4T; T4I = FMA(KP728968627, T4G, KP684547105 * T4H); T4L = FNMS(KP992114701, T4K, KP125333233 * T4J); T4M = T4I + T4L; T4P = FMA(KP062790519, T4N, KP998026728 * T4O); T4S = FNMS(KP637423989, T4R, KP770513242 * T4Q); T4T = T4P + T4S; T4U = T4M + T4T; T5b = KP559016994 * (T4M - T4T); T56 = T4I - T4L; T57 = T4P - T4S; } { E T4V, T4W, T4X, T4Y, T4Z, T50; T4V = FNMS(KP684547105, T4G, KP728968627 * T4H); T4W = FMA(KP125333233, T4K, KP992114701 * T4J); T4X = T4V - T4W; T4Y = FNMS(KP998026728, T4N, KP062790519 * T4O); T4Z = FMA(KP770513242, T4R, KP637423989 * T4Q); T50 = T4Y - T4Z; T51 = KP559016994 * (T4X - T50); T5f = T4Y + T4Z; T53 = T4X + T50; T5e = T4V + T4W; } ro[WS(os, 3)] = T4F + T4U; io[WS(os, 3)] = T52 + T53; { E T58, T59, T55, T5a, T54; T58 = FMA(KP951056516, T56, KP587785252 * T57); T59 = FNMS(KP587785252, T56, KP951056516 * T57); T54 = FNMS(KP250000000, T53, T52); T55 = T51 + T54; T5a = T54 - T51; io[WS(os, 8)] = T55 - T58; io[WS(os, 18)] = T5a - T59; io[WS(os, 23)] = T58 + T55; io[WS(os, 13)] = T59 + T5a; } { E T5g, T5i, T5d, T5h, T5c; T5g = FMA(KP951056516, T5e, KP587785252 * T5f); T5i = FNMS(KP587785252, T5e, KP951056516 * T5f); T5c = FNMS(KP250000000, T4U, T4F); T5d = T5b + T5c; T5h = T5c - T5b; ro[WS(os, 23)] = T5d - T5g; ro[WS(os, 18)] = T5h + T5i; ro[WS(os, 8)] = T5d + T5g; ro[WS(os, 13)] = T5h - T5i; } } { E T3J, T4w, T4e, T4B, T4f, T4A, T4k, T4x, T4n, T4t; T3J = T3H - T3I; T4w = T4u + T4v; { E T3Q, T3X, T3Y, T45, T4c, T4d; T3Q = FMA(KP876306680, T3M, KP481753674 * T3P); T3X = FNMS(KP425779291, T3W, KP904827052 * T3T); T3Y = T3Q + T3X; T45 = FMA(KP535826794, T41, KP844327925 * T44); T4c = FMA(KP062790519, T48, KP998026728 * T4b); T4d = T45 + T4c; T4e = T3Y + T4d; T4B = T45 - T4c; T4f = KP559016994 * (T3Y - T4d); T4A = T3Q - T3X; } { E T4i, T4j, T4r, T4l, T4m, T4s; T4i = FNMS(KP481753674, T3M, KP876306680 * T3P); T4j = FMA(KP904827052, T3W, KP425779291 * T3T); T4r = T4i - T4j; T4l = FNMS(KP844327925, T41, KP535826794 * T44); T4m = FNMS(KP998026728, T48, KP062790519 * T4b); T4s = T4l + T4m; T4k = T4i + T4j; T4x = T4r + T4s; T4n = T4l - T4m; T4t = KP559016994 * (T4r - T4s); } ro[WS(os, 2)] = T3J + T4e; io[WS(os, 2)] = T4w + T4x; { E T4o, T4q, T4h, T4p, T4g; T4o = FMA(KP951056516, T4k, KP587785252 * T4n); T4q = FNMS(KP587785252, T4k, KP951056516 * T4n); T4g = FNMS(KP250000000, T4e, T3J); T4h = T4f + T4g; T4p = T4g - T4f; ro[WS(os, 22)] = T4h - T4o; ro[WS(os, 17)] = T4p + T4q; ro[WS(os, 7)] = T4h + T4o; ro[WS(os, 12)] = T4p - T4q; } { E T4C, T4D, T4z, T4E, T4y; T4C = FMA(KP951056516, T4A, KP587785252 * T4B); T4D = FNMS(KP587785252, T4A, KP951056516 * T4B); T4y = FNMS(KP250000000, T4x, T4w); T4z = T4t + T4y; T4E = T4y - T4t; io[WS(os, 7)] = T4z - T4C; io[WS(os, 17)] = T4E - T4D; io[WS(os, 22)] = T4C + T4z; io[WS(os, 12)] = T4D + T4E; } } { E T33, T3j, T3i, T3z, T3r, T3s, T3q, T3D, T3v, T3C; T33 = TP - TW; T3j = T2T + T2Q; { E T36, T39, T3a, T3d, T3g, T3h; T36 = FMA(KP535826794, T34, KP844327925 * T35); T39 = FMA(KP637423989, T37, KP770513242 * T38); T3a = T36 - T39; T3d = FNMS(KP425779291, T3c, KP904827052 * T3b); T3g = FNMS(KP992114701, T3f, KP125333233 * T3e); T3h = T3d + T3g; T3i = T3a + T3h; T3z = KP559016994 * (T3a - T3h); T3r = T3d - T3g; T3s = T36 + T39; } { E T3k, T3l, T3m, T3n, T3o, T3p; T3k = FNMS(KP844327925, T34, KP535826794 * T35); T3l = FNMS(KP637423989, T38, KP770513242 * T37); T3m = T3k + T3l; T3n = FMA(KP904827052, T3c, KP425779291 * T3b); T3o = FMA(KP125333233, T3f, KP992114701 * T3e); T3p = T3n + T3o; T3q = T3m - T3p; T3D = T3o - T3n; T3v = KP559016994 * (T3m + T3p); T3C = T3k - T3l; } ro[WS(os, 4)] = T33 + T3i; io[WS(os, 4)] = T3j + T3q; { E T3t, T3y, T3w, T3x, T3u; T3t = FNMS(KP587785252, T3s, KP951056516 * T3r); T3y = FMA(KP951056516, T3s, KP587785252 * T3r); T3u = FNMS(KP250000000, T3q, T3j); T3w = T3u - T3v; T3x = T3u + T3v; io[WS(os, 14)] = T3t + T3w; io[WS(os, 24)] = T3y + T3x; io[WS(os, 19)] = T3w - T3t; io[WS(os, 9)] = T3x - T3y; } { E T3E, T3G, T3B, T3F, T3A; T3E = FMA(KP951056516, T3C, KP587785252 * T3D); T3G = FNMS(KP587785252, T3C, KP951056516 * T3D); T3A = FNMS(KP250000000, T3i, T33); T3B = T3z + T3A; T3F = T3A - T3z; ro[WS(os, 24)] = T3B - T3E; ro[WS(os, 19)] = T3F + T3G; ro[WS(os, 9)] = T3B + T3E; ro[WS(os, 14)] = T3F - T3G; } } } } } static const kdft_desc desc = { 25, "n1_25", {260, 92, 92, 0}, &GENUS, 0, 0, 0, 0 }; void X(codelet_n1_25) (planner *p) { X(kdft_register) (p, n1_25, &desc); } #endif /* HAVE_FMA */ fftw-3.3.3/dft/scalar/codelets/t1_2.c0000644000175000017500000000670312054410243014145 00000000000000/* * Copyright (c) 2003, 2007-11 Matteo Frigo * Copyright (c) 2003, 2007-11 Massachusetts Institute of Technology * * This program is free software; you can redistribute it and/or modify * it under the terms of the GNU General Public License as published by * the Free Software Foundation; either version 2 of the License, or * (at your option) any later version. * * This program is distributed in the hope that it will be useful, * but WITHOUT ANY WARRANTY; without even the implied warranty of * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the * GNU General Public License for more details. * * You should have received a copy of the GNU General Public License * along with this program; if not, write to the Free Software * Foundation, Inc., 51 Franklin Street, Fifth Floor, Boston, MA 02110-1301 USA * */ /* This file was automatically generated --- DO NOT EDIT */ /* Generated on Sun Nov 25 07:35:47 EST 2012 */ #include "codelet-dft.h" #ifdef HAVE_FMA /* Generated by: ../../../genfft/gen_twiddle.native -fma -reorder-insns -schedule-for-pipeline -compact -variables 4 -pipeline-latency 4 -n 2 -name t1_2 -include t.h */ /* * This function contains 6 FP additions, 4 FP multiplications, * (or, 4 additions, 2 multiplications, 2 fused multiply/add), * 11 stack variables, 0 constants, and 8 memory accesses */ #include "t.h" static void t1_2(R *ri, R *ii, const R *W, stride rs, INT mb, INT me, INT ms) { { INT m; for (m = mb, W = W + (mb * 2); m < me; m = m + 1, ri = ri + ms, ii = ii + ms, W = W + 2, MAKE_VOLATILE_STRIDE(4, rs)) { E T1, Ta, T3, T6, T2, T5; T1 = ri[0]; Ta = ii[0]; T3 = ri[WS(rs, 1)]; T6 = ii[WS(rs, 1)]; T2 = W[0]; T5 = W[1]; { E T8, T4, T9, T7; T8 = T2 * T6; T4 = T2 * T3; T9 = FNMS(T5, T3, T8); T7 = FMA(T5, T6, T4); ii[0] = T9 + Ta; ii[WS(rs, 1)] = Ta - T9; ri[0] = T1 + T7; ri[WS(rs, 1)] = T1 - T7; } } } } static const tw_instr twinstr[] = { {TW_FULL, 0, 2}, {TW_NEXT, 1, 0} }; static const ct_desc desc = { 2, "t1_2", twinstr, &GENUS, {4, 2, 2, 0}, 0, 0, 0 }; void X(codelet_t1_2) (planner *p) { X(kdft_dit_register) (p, t1_2, &desc); } #else /* HAVE_FMA */ /* Generated by: ../../../genfft/gen_twiddle.native -compact -variables 4 -pipeline-latency 4 -n 2 -name t1_2 -include t.h */ /* * This function contains 6 FP additions, 4 FP multiplications, * (or, 4 additions, 2 multiplications, 2 fused multiply/add), * 9 stack variables, 0 constants, and 8 memory accesses */ #include "t.h" static void t1_2(R *ri, R *ii, const R *W, stride rs, INT mb, INT me, INT ms) { { INT m; for (m = mb, W = W + (mb * 2); m < me; m = m + 1, ri = ri + ms, ii = ii + ms, W = W + 2, MAKE_VOLATILE_STRIDE(4, rs)) { E T1, T8, T6, T7; T1 = ri[0]; T8 = ii[0]; { E T3, T5, T2, T4; T3 = ri[WS(rs, 1)]; T5 = ii[WS(rs, 1)]; T2 = W[0]; T4 = W[1]; T6 = FMA(T2, T3, T4 * T5); T7 = FNMS(T4, T3, T2 * T5); } ri[WS(rs, 1)] = T1 - T6; ii[WS(rs, 1)] = T8 - T7; ri[0] = T1 + T6; ii[0] = T7 + T8; } } } static const tw_instr twinstr[] = { {TW_FULL, 0, 2}, {TW_NEXT, 1, 0} }; static const ct_desc desc = { 2, "t1_2", twinstr, &GENUS, {4, 2, 2, 0}, 0, 0, 0 }; void X(codelet_t1_2) (planner *p) { X(kdft_dit_register) (p, t1_2, &desc); } #endif /* HAVE_FMA */ fftw-3.3.3/dft/scalar/codelets/t1_3.c0000644000175000017500000001142012054410243014136 00000000000000/* * Copyright (c) 2003, 2007-11 Matteo Frigo * Copyright (c) 2003, 2007-11 Massachusetts Institute of Technology * * This program is free software; you can redistribute it and/or modify * it under the terms of the GNU General Public License as published by * the Free Software Foundation; either version 2 of the License, or * (at your option) any later version. * * This program is distributed in the hope that it will be useful, * but WITHOUT ANY WARRANTY; without even the implied warranty of * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the * GNU General Public License for more details. * * You should have received a copy of the GNU General Public License * along with this program; if not, write to the Free Software * Foundation, Inc., 51 Franklin Street, Fifth Floor, Boston, MA 02110-1301 USA * */ /* This file was automatically generated --- DO NOT EDIT */ /* Generated on Sun Nov 25 07:35:47 EST 2012 */ #include "codelet-dft.h" #ifdef HAVE_FMA /* Generated by: ../../../genfft/gen_twiddle.native -fma -reorder-insns -schedule-for-pipeline -compact -variables 4 -pipeline-latency 4 -n 3 -name t1_3 -include t.h */ /* * This function contains 16 FP additions, 14 FP multiplications, * (or, 6 additions, 4 multiplications, 10 fused multiply/add), * 21 stack variables, 2 constants, and 12 memory accesses */ #include "t.h" static void t1_3(R *ri, R *ii, const R *W, stride rs, INT mb, INT me, INT ms) { DK(KP866025403, +0.866025403784438646763723170752936183471402627); DK(KP500000000, +0.500000000000000000000000000000000000000000000); { INT m; for (m = mb, W = W + (mb * 4); m < me; m = m + 1, ri = ri + ms, ii = ii + ms, W = W + 4, MAKE_VOLATILE_STRIDE(6, rs)) { E T1, Tm, T9, Tc, Tb, Th, T7, Ti, Ta, Tj, Td; T1 = ri[0]; Tm = ii[0]; { E T3, T6, T2, T5, Tg, T4, T8; T3 = ri[WS(rs, 1)]; T6 = ii[WS(rs, 1)]; T2 = W[0]; T5 = W[1]; T9 = ri[WS(rs, 2)]; Tc = ii[WS(rs, 2)]; Tg = T2 * T6; T4 = T2 * T3; T8 = W[2]; Tb = W[3]; Th = FNMS(T5, T3, Tg); T7 = FMA(T5, T6, T4); Ti = T8 * Tc; Ta = T8 * T9; } Tj = FNMS(Tb, T9, Ti); Td = FMA(Tb, Tc, Ta); { E Tk, Te, To, Tn, Tl, Tf; Tk = Th - Tj; Tl = Th + Tj; Te = T7 + Td; To = Td - T7; ii[0] = Tl + Tm; Tn = FNMS(KP500000000, Tl, Tm); ri[0] = T1 + Te; Tf = FNMS(KP500000000, Te, T1); ii[WS(rs, 1)] = FMA(KP866025403, To, Tn); ii[WS(rs, 2)] = FNMS(KP866025403, To, Tn); ri[WS(rs, 2)] = FNMS(KP866025403, Tk, Tf); ri[WS(rs, 1)] = FMA(KP866025403, Tk, Tf); } } } } static const tw_instr twinstr[] = { {TW_FULL, 0, 3}, {TW_NEXT, 1, 0} }; static const ct_desc desc = { 3, "t1_3", twinstr, &GENUS, {6, 4, 10, 0}, 0, 0, 0 }; void X(codelet_t1_3) (planner *p) { X(kdft_dit_register) (p, t1_3, &desc); } #else /* HAVE_FMA */ /* Generated by: ../../../genfft/gen_twiddle.native -compact -variables 4 -pipeline-latency 4 -n 3 -name t1_3 -include t.h */ /* * This function contains 16 FP additions, 12 FP multiplications, * (or, 10 additions, 6 multiplications, 6 fused multiply/add), * 15 stack variables, 2 constants, and 12 memory accesses */ #include "t.h" static void t1_3(R *ri, R *ii, const R *W, stride rs, INT mb, INT me, INT ms) { DK(KP866025403, +0.866025403784438646763723170752936183471402627); DK(KP500000000, +0.500000000000000000000000000000000000000000000); { INT m; for (m = mb, W = W + (mb * 4); m < me; m = m + 1, ri = ri + ms, ii = ii + ms, W = W + 4, MAKE_VOLATILE_STRIDE(6, rs)) { E T1, Ti, T6, Te, Tb, Tf, Tc, Th; T1 = ri[0]; Ti = ii[0]; { E T3, T5, T2, T4; T3 = ri[WS(rs, 1)]; T5 = ii[WS(rs, 1)]; T2 = W[0]; T4 = W[1]; T6 = FMA(T2, T3, T4 * T5); Te = FNMS(T4, T3, T2 * T5); } { E T8, Ta, T7, T9; T8 = ri[WS(rs, 2)]; Ta = ii[WS(rs, 2)]; T7 = W[2]; T9 = W[3]; Tb = FMA(T7, T8, T9 * Ta); Tf = FNMS(T9, T8, T7 * Ta); } Tc = T6 + Tb; Th = Te + Tf; ri[0] = T1 + Tc; ii[0] = Th + Ti; { E Td, Tg, Tj, Tk; Td = FNMS(KP500000000, Tc, T1); Tg = KP866025403 * (Te - Tf); ri[WS(rs, 2)] = Td - Tg; ri[WS(rs, 1)] = Td + Tg; Tj = KP866025403 * (Tb - T6); Tk = FNMS(KP500000000, Th, Ti); ii[WS(rs, 1)] = Tj + Tk; ii[WS(rs, 2)] = Tk - Tj; } } } } static const tw_instr twinstr[] = { {TW_FULL, 0, 3}, {TW_NEXT, 1, 0} }; static const ct_desc desc = { 3, "t1_3", twinstr, &GENUS, {10, 6, 6, 0}, 0, 0, 0 }; void X(codelet_t1_3) (planner *p) { X(kdft_dit_register) (p, t1_3, &desc); } #endif /* HAVE_FMA */ fftw-3.3.3/dft/scalar/codelets/t1_4.c0000644000175000017500000001204312054410243014141 00000000000000/* * Copyright (c) 2003, 2007-11 Matteo Frigo * Copyright (c) 2003, 2007-11 Massachusetts Institute of Technology * * This program is free software; you can redistribute it and/or modify * it under the terms of the GNU General Public License as published by * the Free Software Foundation; either version 2 of the License, or * (at your option) any later version. * * This program is distributed in the hope that it will be useful, * but WITHOUT ANY WARRANTY; without even the implied warranty of * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the * GNU General Public License for more details. * * You should have received a copy of the GNU General Public License * along with this program; if not, write to the Free Software * Foundation, Inc., 51 Franklin Street, Fifth Floor, Boston, MA 02110-1301 USA * */ /* This file was automatically generated --- DO NOT EDIT */ /* Generated on Sun Nov 25 07:35:47 EST 2012 */ #include "codelet-dft.h" #ifdef HAVE_FMA /* Generated by: ../../../genfft/gen_twiddle.native -fma -reorder-insns -schedule-for-pipeline -compact -variables 4 -pipeline-latency 4 -n 4 -name t1_4 -include t.h */ /* * This function contains 22 FP additions, 12 FP multiplications, * (or, 16 additions, 6 multiplications, 6 fused multiply/add), * 31 stack variables, 0 constants, and 16 memory accesses */ #include "t.h" static void t1_4(R *ri, R *ii, const R *W, stride rs, INT mb, INT me, INT ms) { { INT m; for (m = mb, W = W + (mb * 6); m < me; m = m + 1, ri = ri + ms, ii = ii + ms, W = W + 6, MAKE_VOLATILE_STRIDE(8, rs)) { E To, Te, Tm, T8, Tw, Tx, Tq, Tk; { E T1, Tv, Tu, T7, Tg, Tj, Tf, Ti, Tp, Th; T1 = ri[0]; Tv = ii[0]; { E T3, T6, T2, T5; T3 = ri[WS(rs, 2)]; T6 = ii[WS(rs, 2)]; T2 = W[2]; T5 = W[3]; { E Ta, Td, Tc, Tn, Tb, Tt, T4, T9; Ta = ri[WS(rs, 1)]; Td = ii[WS(rs, 1)]; Tt = T2 * T6; T4 = T2 * T3; T9 = W[0]; Tc = W[1]; Tu = FNMS(T5, T3, Tt); T7 = FMA(T5, T6, T4); Tn = T9 * Td; Tb = T9 * Ta; Tg = ri[WS(rs, 3)]; Tj = ii[WS(rs, 3)]; To = FNMS(Tc, Ta, Tn); Te = FMA(Tc, Td, Tb); Tf = W[4]; Ti = W[5]; } } Tm = T1 - T7; T8 = T1 + T7; Tw = Tu + Tv; Tx = Tv - Tu; Tp = Tf * Tj; Th = Tf * Tg; Tq = FNMS(Ti, Tg, Tp); Tk = FMA(Ti, Tj, Th); } { E Ts, Tr, Tl, Ty; Ts = To + Tq; Tr = To - Tq; Tl = Te + Tk; Ty = Te - Tk; ri[WS(rs, 1)] = Tm + Tr; ri[WS(rs, 3)] = Tm - Tr; ii[WS(rs, 2)] = Tw - Ts; ii[0] = Ts + Tw; ii[WS(rs, 3)] = Ty + Tx; ii[WS(rs, 1)] = Tx - Ty; ri[0] = T8 + Tl; ri[WS(rs, 2)] = T8 - Tl; } } } } static const tw_instr twinstr[] = { {TW_FULL, 0, 4}, {TW_NEXT, 1, 0} }; static const ct_desc desc = { 4, "t1_4", twinstr, &GENUS, {16, 6, 6, 0}, 0, 0, 0 }; void X(codelet_t1_4) (planner *p) { X(kdft_dit_register) (p, t1_4, &desc); } #else /* HAVE_FMA */ /* Generated by: ../../../genfft/gen_twiddle.native -compact -variables 4 -pipeline-latency 4 -n 4 -name t1_4 -include t.h */ /* * This function contains 22 FP additions, 12 FP multiplications, * (or, 16 additions, 6 multiplications, 6 fused multiply/add), * 13 stack variables, 0 constants, and 16 memory accesses */ #include "t.h" static void t1_4(R *ri, R *ii, const R *W, stride rs, INT mb, INT me, INT ms) { { INT m; for (m = mb, W = W + (mb * 6); m < me; m = m + 1, ri = ri + ms, ii = ii + ms, W = W + 6, MAKE_VOLATILE_STRIDE(8, rs)) { E T1, Tp, T6, To, Tc, Tk, Th, Tl; T1 = ri[0]; Tp = ii[0]; { E T3, T5, T2, T4; T3 = ri[WS(rs, 2)]; T5 = ii[WS(rs, 2)]; T2 = W[2]; T4 = W[3]; T6 = FMA(T2, T3, T4 * T5); To = FNMS(T4, T3, T2 * T5); } { E T9, Tb, T8, Ta; T9 = ri[WS(rs, 1)]; Tb = ii[WS(rs, 1)]; T8 = W[0]; Ta = W[1]; Tc = FMA(T8, T9, Ta * Tb); Tk = FNMS(Ta, T9, T8 * Tb); } { E Te, Tg, Td, Tf; Te = ri[WS(rs, 3)]; Tg = ii[WS(rs, 3)]; Td = W[4]; Tf = W[5]; Th = FMA(Td, Te, Tf * Tg); Tl = FNMS(Tf, Te, Td * Tg); } { E T7, Ti, Tn, Tq; T7 = T1 + T6; Ti = Tc + Th; ri[WS(rs, 2)] = T7 - Ti; ri[0] = T7 + Ti; Tn = Tk + Tl; Tq = To + Tp; ii[0] = Tn + Tq; ii[WS(rs, 2)] = Tq - Tn; } { E Tj, Tm, Tr, Ts; Tj = T1 - T6; Tm = Tk - Tl; ri[WS(rs, 3)] = Tj - Tm; ri[WS(rs, 1)] = Tj + Tm; Tr = Tp - To; Ts = Tc - Th; ii[WS(rs, 1)] = Tr - Ts; ii[WS(rs, 3)] = Ts + Tr; } } } } static const tw_instr twinstr[] = { {TW_FULL, 0, 4}, {TW_NEXT, 1, 0} }; static const ct_desc desc = { 4, "t1_4", twinstr, &GENUS, {16, 6, 6, 0}, 0, 0, 0 }; void X(codelet_t1_4) (planner *p) { X(kdft_dit_register) (p, t1_4, &desc); } #endif /* HAVE_FMA */ fftw-3.3.3/dft/scalar/codelets/t1_5.c0000644000175000017500000001633312054410244014151 00000000000000/* * Copyright (c) 2003, 2007-11 Matteo Frigo * Copyright (c) 2003, 2007-11 Massachusetts Institute of Technology * * This program is free software; you can redistribute it and/or modify * it under the terms of the GNU General Public License as published by * the Free Software Foundation; either version 2 of the License, or * (at your option) any later version. * * This program is distributed in the hope that it will be useful, * but WITHOUT ANY WARRANTY; without even the implied warranty of * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the * GNU General Public License for more details. * * You should have received a copy of the GNU General Public License * along with this program; if not, write to the Free Software * Foundation, Inc., 51 Franklin Street, Fifth Floor, Boston, MA 02110-1301 USA * */ /* This file was automatically generated --- DO NOT EDIT */ /* Generated on Sun Nov 25 07:35:47 EST 2012 */ #include "codelet-dft.h" #ifdef HAVE_FMA /* Generated by: ../../../genfft/gen_twiddle.native -fma -reorder-insns -schedule-for-pipeline -compact -variables 4 -pipeline-latency 4 -n 5 -name t1_5 -include t.h */ /* * This function contains 40 FP additions, 34 FP multiplications, * (or, 14 additions, 8 multiplications, 26 fused multiply/add), * 43 stack variables, 4 constants, and 20 memory accesses */ #include "t.h" static void t1_5(R *ri, R *ii, const R *W, stride rs, INT mb, INT me, INT ms) { DK(KP951056516, +0.951056516295153572116439333379382143405698634); DK(KP559016994, +0.559016994374947424102293417182819058860154590); DK(KP250000000, +0.250000000000000000000000000000000000000000000); DK(KP618033988, +0.618033988749894848204586834365638117720309180); { INT m; for (m = mb, W = W + (mb * 8); m < me; m = m + 1, ri = ri + ms, ii = ii + ms, W = W + 8, MAKE_VOLATILE_STRIDE(10, rs)) { E T1, TM, TJ, TA, TQ, Te, TC, Tk, TE, Tq; { E Tg, Tj, Tm, TB, Th, Tp, Tl, Ti, To, TD, Tn; T1 = ri[0]; TM = ii[0]; { E T9, Tc, Ty, Ta, Tb, Tx, T7, Tf, Tz, Td; { E T3, T6, T8, Tw, T4, T2, T5; T3 = ri[WS(rs, 1)]; T6 = ii[WS(rs, 1)]; T2 = W[0]; T9 = ri[WS(rs, 4)]; Tc = ii[WS(rs, 4)]; T8 = W[6]; Tw = T2 * T6; T4 = T2 * T3; T5 = W[1]; Ty = T8 * Tc; Ta = T8 * T9; Tb = W[7]; Tx = FNMS(T5, T3, Tw); T7 = FMA(T5, T6, T4); } Tg = ri[WS(rs, 2)]; Tz = FNMS(Tb, T9, Ty); Td = FMA(Tb, Tc, Ta); Tj = ii[WS(rs, 2)]; Tf = W[2]; TJ = Tx + Tz; TA = Tx - Tz; TQ = T7 - Td; Te = T7 + Td; Tm = ri[WS(rs, 3)]; TB = Tf * Tj; Th = Tf * Tg; Tp = ii[WS(rs, 3)]; Tl = W[4]; Ti = W[3]; To = W[5]; } TD = Tl * Tp; Tn = Tl * Tm; TC = FNMS(Ti, Tg, TB); Tk = FMA(Ti, Tj, Th); TE = FNMS(To, Tm, TD); Tq = FMA(To, Tp, Tn); } { E TG, TI, TO, TS, TU, Tu, TN, Tt, TK, TF; TK = TC + TE; TF = TC - TE; { E Tr, TR, TL, Ts; Tr = Tk + Tq; TR = Tk - Tq; TG = FMA(KP618033988, TF, TA); TI = FNMS(KP618033988, TA, TF); TO = TJ - TK; TL = TJ + TK; TS = FMA(KP618033988, TR, TQ); TU = FNMS(KP618033988, TQ, TR); Tu = Te - Tr; Ts = Te + Tr; ii[0] = TL + TM; TN = FNMS(KP250000000, TL, TM); ri[0] = T1 + Ts; Tt = FNMS(KP250000000, Ts, T1); } { E TT, TP, TH, Tv; TT = FNMS(KP559016994, TO, TN); TP = FMA(KP559016994, TO, TN); TH = FNMS(KP559016994, Tu, Tt); Tv = FMA(KP559016994, Tu, Tt); ii[WS(rs, 4)] = FMA(KP951056516, TS, TP); ii[WS(rs, 1)] = FNMS(KP951056516, TS, TP); ii[WS(rs, 3)] = FNMS(KP951056516, TU, TT); ii[WS(rs, 2)] = FMA(KP951056516, TU, TT); ri[WS(rs, 1)] = FMA(KP951056516, TG, Tv); ri[WS(rs, 4)] = FNMS(KP951056516, TG, Tv); ri[WS(rs, 3)] = FMA(KP951056516, TI, TH); ri[WS(rs, 2)] = FNMS(KP951056516, TI, TH); } } } } } static const tw_instr twinstr[] = { {TW_FULL, 0, 5}, {TW_NEXT, 1, 0} }; static const ct_desc desc = { 5, "t1_5", twinstr, &GENUS, {14, 8, 26, 0}, 0, 0, 0 }; void X(codelet_t1_5) (planner *p) { X(kdft_dit_register) (p, t1_5, &desc); } #else /* HAVE_FMA */ /* Generated by: ../../../genfft/gen_twiddle.native -compact -variables 4 -pipeline-latency 4 -n 5 -name t1_5 -include t.h */ /* * This function contains 40 FP additions, 28 FP multiplications, * (or, 26 additions, 14 multiplications, 14 fused multiply/add), * 29 stack variables, 4 constants, and 20 memory accesses */ #include "t.h" static void t1_5(R *ri, R *ii, const R *W, stride rs, INT mb, INT me, INT ms) { DK(KP250000000, +0.250000000000000000000000000000000000000000000); DK(KP559016994, +0.559016994374947424102293417182819058860154590); DK(KP587785252, +0.587785252292473129168705954639072768597652438); DK(KP951056516, +0.951056516295153572116439333379382143405698634); { INT m; for (m = mb, W = W + (mb * 8); m < me; m = m + 1, ri = ri + ms, ii = ii + ms, W = W + 8, MAKE_VOLATILE_STRIDE(10, rs)) { E T1, TE, Tu, Tx, TJ, TI, TB, TC, TD, Tc, Tn, To; T1 = ri[0]; TE = ii[0]; { E T6, Ts, Tm, Tw, Tb, Tt, Th, Tv; { E T3, T5, T2, T4; T3 = ri[WS(rs, 1)]; T5 = ii[WS(rs, 1)]; T2 = W[0]; T4 = W[1]; T6 = FMA(T2, T3, T4 * T5); Ts = FNMS(T4, T3, T2 * T5); } { E Tj, Tl, Ti, Tk; Tj = ri[WS(rs, 3)]; Tl = ii[WS(rs, 3)]; Ti = W[4]; Tk = W[5]; Tm = FMA(Ti, Tj, Tk * Tl); Tw = FNMS(Tk, Tj, Ti * Tl); } { E T8, Ta, T7, T9; T8 = ri[WS(rs, 4)]; Ta = ii[WS(rs, 4)]; T7 = W[6]; T9 = W[7]; Tb = FMA(T7, T8, T9 * Ta); Tt = FNMS(T9, T8, T7 * Ta); } { E Te, Tg, Td, Tf; Te = ri[WS(rs, 2)]; Tg = ii[WS(rs, 2)]; Td = W[2]; Tf = W[3]; Th = FMA(Td, Te, Tf * Tg); Tv = FNMS(Tf, Te, Td * Tg); } Tu = Ts - Tt; Tx = Tv - Tw; TJ = Th - Tm; TI = T6 - Tb; TB = Ts + Tt; TC = Tv + Tw; TD = TB + TC; Tc = T6 + Tb; Tn = Th + Tm; To = Tc + Tn; } ri[0] = T1 + To; ii[0] = TD + TE; { E Ty, TA, Tr, Tz, Tp, Tq; Ty = FMA(KP951056516, Tu, KP587785252 * Tx); TA = FNMS(KP587785252, Tu, KP951056516 * Tx); Tp = KP559016994 * (Tc - Tn); Tq = FNMS(KP250000000, To, T1); Tr = Tp + Tq; Tz = Tq - Tp; ri[WS(rs, 4)] = Tr - Ty; ri[WS(rs, 3)] = Tz + TA; ri[WS(rs, 1)] = Tr + Ty; ri[WS(rs, 2)] = Tz - TA; } { E TK, TL, TH, TM, TF, TG; TK = FMA(KP951056516, TI, KP587785252 * TJ); TL = FNMS(KP587785252, TI, KP951056516 * TJ); TF = KP559016994 * (TB - TC); TG = FNMS(KP250000000, TD, TE); TH = TF + TG; TM = TG - TF; ii[WS(rs, 1)] = TH - TK; ii[WS(rs, 3)] = TM - TL; ii[WS(rs, 4)] = TK + TH; ii[WS(rs, 2)] = TL + TM; } } } } static const tw_instr twinstr[] = { {TW_FULL, 0, 5}, {TW_NEXT, 1, 0} }; static const ct_desc desc = { 5, "t1_5", twinstr, &GENUS, {26, 14, 14, 0}, 0, 0, 0 }; void X(codelet_t1_5) (planner *p) { X(kdft_dit_register) (p, t1_5, &desc); } #endif /* HAVE_FMA */ fftw-3.3.3/dft/scalar/codelets/t1_6.c0000644000175000017500000001720612054410244014152 00000000000000/* * Copyright (c) 2003, 2007-11 Matteo Frigo * Copyright (c) 2003, 2007-11 Massachusetts Institute of Technology * * This program is free software; you can redistribute it and/or modify * it under the terms of the GNU General Public License as published by * the Free Software Foundation; either version 2 of the License, or * (at your option) any later version. * * This program is distributed in the hope that it will be useful, * but WITHOUT ANY WARRANTY; without even the implied warranty of * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the * GNU General Public License for more details. * * You should have received a copy of the GNU General Public License * along with this program; if not, write to the Free Software * Foundation, Inc., 51 Franklin Street, Fifth Floor, Boston, MA 02110-1301 USA * */ /* This file was automatically generated --- DO NOT EDIT */ /* Generated on Sun Nov 25 07:35:48 EST 2012 */ #include "codelet-dft.h" #ifdef HAVE_FMA /* Generated by: ../../../genfft/gen_twiddle.native -fma -reorder-insns -schedule-for-pipeline -compact -variables 4 -pipeline-latency 4 -n 6 -name t1_6 -include t.h */ /* * This function contains 46 FP additions, 32 FP multiplications, * (or, 24 additions, 10 multiplications, 22 fused multiply/add), * 47 stack variables, 2 constants, and 24 memory accesses */ #include "t.h" static void t1_6(R *ri, R *ii, const R *W, stride rs, INT mb, INT me, INT ms) { DK(KP866025403, +0.866025403784438646763723170752936183471402627); DK(KP500000000, +0.500000000000000000000000000000000000000000000); { INT m; for (m = mb, W = W + (mb * 10); m < me; m = m + 1, ri = ri + ms, ii = ii + ms, W = W + 10, MAKE_VOLATILE_STRIDE(12, rs)) { E TY, TU, T10, TZ; { E T1, TX, TW, T7, Tn, Tq, TJ, TR, TB, Tl, To, TK, Tt, Tw, Ts; E Tp, Tv; T1 = ri[0]; TX = ii[0]; { E T3, T6, T2, T5; T3 = ri[WS(rs, 3)]; T6 = ii[WS(rs, 3)]; T2 = W[4]; T5 = W[5]; { E Ta, Td, Tg, TF, Tb, Tj, Tf, Tc, Ti, TV, T4, T9; Ta = ri[WS(rs, 2)]; Td = ii[WS(rs, 2)]; TV = T2 * T6; T4 = T2 * T3; T9 = W[2]; Tg = ri[WS(rs, 5)]; TW = FNMS(T5, T3, TV); T7 = FMA(T5, T6, T4); TF = T9 * Td; Tb = T9 * Ta; Tj = ii[WS(rs, 5)]; Tf = W[8]; Tc = W[3]; Ti = W[9]; { E TG, Te, TI, Tk, TH, Th, Tm; Tn = ri[WS(rs, 4)]; TH = Tf * Tj; Th = Tf * Tg; TG = FNMS(Tc, Ta, TF); Te = FMA(Tc, Td, Tb); TI = FNMS(Ti, Tg, TH); Tk = FMA(Ti, Tj, Th); Tq = ii[WS(rs, 4)]; Tm = W[6]; TJ = TG - TI; TR = TG + TI; TB = Te + Tk; Tl = Te - Tk; To = Tm * Tn; TK = Tm * Tq; } Tt = ri[WS(rs, 1)]; Tw = ii[WS(rs, 1)]; Ts = W[0]; Tp = W[7]; Tv = W[1]; } } { E TA, T8, TL, Tr, TN, Tx, T11, TM, Tu; TA = T1 + T7; T8 = T1 - T7; TM = Ts * Tw; Tu = Ts * Tt; TL = FNMS(Tp, Tn, TK); Tr = FMA(Tp, Tq, To); TN = FNMS(Tv, Tt, TM); Tx = FMA(Tv, Tw, Tu); T11 = TX - TW; TY = TW + TX; { E TP, TT, TD, TE, TQ, Tz, T14, T13; { E TO, TS, TC, Ty, T12; TO = TL - TN; TS = TL + TN; TC = Tr + Tx; Ty = Tr - Tx; T12 = TJ + TO; TP = TJ - TO; TT = TR - TS; TU = TR + TS; Tz = Tl + Ty; T14 = Ty - Tl; ii[WS(rs, 3)] = T12 + T11; T13 = FNMS(KP500000000, T12, T11); T10 = TC - TB; TD = TB + TC; } ri[WS(rs, 3)] = T8 + Tz; TE = FNMS(KP500000000, Tz, T8); ii[WS(rs, 5)] = FNMS(KP866025403, T14, T13); ii[WS(rs, 1)] = FMA(KP866025403, T14, T13); TQ = FNMS(KP500000000, TD, TA); ri[WS(rs, 5)] = FNMS(KP866025403, TP, TE); ri[WS(rs, 1)] = FMA(KP866025403, TP, TE); ri[0] = TA + TD; ri[WS(rs, 4)] = FMA(KP866025403, TT, TQ); ri[WS(rs, 2)] = FNMS(KP866025403, TT, TQ); } } } ii[0] = TU + TY; TZ = FNMS(KP500000000, TU, TY); ii[WS(rs, 2)] = FNMS(KP866025403, T10, TZ); ii[WS(rs, 4)] = FMA(KP866025403, T10, TZ); } } } static const tw_instr twinstr[] = { {TW_FULL, 0, 6}, {TW_NEXT, 1, 0} }; static const ct_desc desc = { 6, "t1_6", twinstr, &GENUS, {24, 10, 22, 0}, 0, 0, 0 }; void X(codelet_t1_6) (planner *p) { X(kdft_dit_register) (p, t1_6, &desc); } #else /* HAVE_FMA */ /* Generated by: ../../../genfft/gen_twiddle.native -compact -variables 4 -pipeline-latency 4 -n 6 -name t1_6 -include t.h */ /* * This function contains 46 FP additions, 28 FP multiplications, * (or, 32 additions, 14 multiplications, 14 fused multiply/add), * 23 stack variables, 2 constants, and 24 memory accesses */ #include "t.h" static void t1_6(R *ri, R *ii, const R *W, stride rs, INT mb, INT me, INT ms) { DK(KP500000000, +0.500000000000000000000000000000000000000000000); DK(KP866025403, +0.866025403784438646763723170752936183471402627); { INT m; for (m = mb, W = W + (mb * 10); m < me; m = m + 1, ri = ri + ms, ii = ii + ms, W = W + 10, MAKE_VOLATILE_STRIDE(12, rs)) { E T7, TS, Tv, TO, Tt, TJ, Tx, TF, Ti, TI, Tw, TC; { E T1, TN, T6, TM; T1 = ri[0]; TN = ii[0]; { E T3, T5, T2, T4; T3 = ri[WS(rs, 3)]; T5 = ii[WS(rs, 3)]; T2 = W[4]; T4 = W[5]; T6 = FMA(T2, T3, T4 * T5); TM = FNMS(T4, T3, T2 * T5); } T7 = T1 - T6; TS = TN - TM; Tv = T1 + T6; TO = TM + TN; } { E Tn, TD, Ts, TE; { E Tk, Tm, Tj, Tl; Tk = ri[WS(rs, 4)]; Tm = ii[WS(rs, 4)]; Tj = W[6]; Tl = W[7]; Tn = FMA(Tj, Tk, Tl * Tm); TD = FNMS(Tl, Tk, Tj * Tm); } { E Tp, Tr, To, Tq; Tp = ri[WS(rs, 1)]; Tr = ii[WS(rs, 1)]; To = W[0]; Tq = W[1]; Ts = FMA(To, Tp, Tq * Tr); TE = FNMS(Tq, Tp, To * Tr); } Tt = Tn - Ts; TJ = TD + TE; Tx = Tn + Ts; TF = TD - TE; } { E Tc, TA, Th, TB; { E T9, Tb, T8, Ta; T9 = ri[WS(rs, 2)]; Tb = ii[WS(rs, 2)]; T8 = W[2]; Ta = W[3]; Tc = FMA(T8, T9, Ta * Tb); TA = FNMS(Ta, T9, T8 * Tb); } { E Te, Tg, Td, Tf; Te = ri[WS(rs, 5)]; Tg = ii[WS(rs, 5)]; Td = W[8]; Tf = W[9]; Th = FMA(Td, Te, Tf * Tg); TB = FNMS(Tf, Te, Td * Tg); } Ti = Tc - Th; TI = TA + TB; Tw = Tc + Th; TC = TA - TB; } { E TG, Tu, Tz, TR, TT, TU; TG = KP866025403 * (TC - TF); Tu = Ti + Tt; Tz = FNMS(KP500000000, Tu, T7); ri[WS(rs, 3)] = T7 + Tu; ri[WS(rs, 1)] = Tz + TG; ri[WS(rs, 5)] = Tz - TG; TR = KP866025403 * (Tt - Ti); TT = TC + TF; TU = FNMS(KP500000000, TT, TS); ii[WS(rs, 1)] = TR + TU; ii[WS(rs, 3)] = TT + TS; ii[WS(rs, 5)] = TU - TR; } { E TK, Ty, TH, TQ, TL, TP; TK = KP866025403 * (TI - TJ); Ty = Tw + Tx; TH = FNMS(KP500000000, Ty, Tv); ri[0] = Tv + Ty; ri[WS(rs, 4)] = TH + TK; ri[WS(rs, 2)] = TH - TK; TQ = KP866025403 * (Tx - Tw); TL = TI + TJ; TP = FNMS(KP500000000, TL, TO); ii[0] = TL + TO; ii[WS(rs, 4)] = TQ + TP; ii[WS(rs, 2)] = TP - TQ; } } } } static const tw_instr twinstr[] = { {TW_FULL, 0, 6}, {TW_NEXT, 1, 0} }; static const ct_desc desc = { 6, "t1_6", twinstr, &GENUS, {32, 14, 14, 0}, 0, 0, 0 }; void X(codelet_t1_6) (planner *p) { X(kdft_dit_register) (p, t1_6, &desc); } #endif /* HAVE_FMA */ fftw-3.3.3/dft/scalar/codelets/t1_7.c0000644000175000017500000002522712054410245014156 00000000000000/* * Copyright (c) 2003, 2007-11 Matteo Frigo * Copyright (c) 2003, 2007-11 Massachusetts Institute of Technology * * This program is free software; you can redistribute it and/or modify * it under the terms of the GNU General Public License as published by * the Free Software Foundation; either version 2 of the License, or * (at your option) any later version. * * This program is distributed in the hope that it will be useful, * but WITHOUT ANY WARRANTY; without even the implied warranty of * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the * GNU General Public License for more details. * * You should have received a copy of the GNU General Public License * along with this program; if not, write to the Free Software * Foundation, Inc., 51 Franklin Street, Fifth Floor, Boston, MA 02110-1301 USA * */ /* This file was automatically generated --- DO NOT EDIT */ /* Generated on Sun Nov 25 07:35:48 EST 2012 */ #include "codelet-dft.h" #ifdef HAVE_FMA /* Generated by: ../../../genfft/gen_twiddle.native -fma -reorder-insns -schedule-for-pipeline -compact -variables 4 -pipeline-latency 4 -n 7 -name t1_7 -include t.h */ /* * This function contains 72 FP additions, 66 FP multiplications, * (or, 18 additions, 12 multiplications, 54 fused multiply/add), * 66 stack variables, 6 constants, and 28 memory accesses */ #include "t.h" static void t1_7(R *ri, R *ii, const R *W, stride rs, INT mb, INT me, INT ms) { DK(KP974927912, +0.974927912181823607018131682993931217232785801); DK(KP801937735, +0.801937735804838252472204639014890102331838324); DK(KP900968867, +0.900968867902419126236102319507445051165919162); DK(KP692021471, +0.692021471630095869627814897002069140197260599); DK(KP554958132, +0.554958132087371191422194871006410481067288862); DK(KP356895867, +0.356895867892209443894399510021300583399127187); { INT m; for (m = mb, W = W + (mb * 12); m < me; m = m + 1, ri = ri + ms, ii = ii + ms, W = W + 12, MAKE_VOLATILE_STRIDE(14, rs)) { E T1c, T19, T1i, T18, T16, T1q, T1t, T1r, T1u, T1s; { E T1, TR, T1h, Te, Tt, Tw, T1a, TM, T1g, Tr, Tu, TS, Tz, TC, Ty; E Tv, TB; T1 = ri[0]; T1c = ii[0]; { E T9, Tc, TP, Ta, Tb, TO, T7; { E T3, T6, T8, TN, T4, T2, T5; T3 = ri[WS(rs, 1)]; T6 = ii[WS(rs, 1)]; T2 = W[0]; T9 = ri[WS(rs, 6)]; Tc = ii[WS(rs, 6)]; T8 = W[10]; TN = T2 * T6; T4 = T2 * T3; T5 = W[1]; TP = T8 * Tc; Ta = T8 * T9; Tb = W[11]; TO = FNMS(T5, T3, TN); T7 = FMA(T5, T6, T4); } { E Tg, Tj, Th, TI, Tm, Tp, Tl, Ti, To, TQ, Td, Tf; Tg = ri[WS(rs, 2)]; TQ = FNMS(Tb, T9, TP); Td = FMA(Tb, Tc, Ta); Tj = ii[WS(rs, 2)]; Tf = W[2]; T19 = TO + TQ; TR = TO - TQ; T1h = Td - T7; Te = T7 + Td; Th = Tf * Tg; TI = Tf * Tj; Tm = ri[WS(rs, 5)]; Tp = ii[WS(rs, 5)]; Tl = W[8]; Ti = W[3]; To = W[9]; { E TJ, Tk, TL, Tq, TK, Tn, Ts; Tt = ri[WS(rs, 3)]; TK = Tl * Tp; Tn = Tl * Tm; TJ = FNMS(Ti, Tg, TI); Tk = FMA(Ti, Tj, Th); TL = FNMS(To, Tm, TK); Tq = FMA(To, Tp, Tn); Tw = ii[WS(rs, 3)]; Ts = W[4]; T1a = TJ + TL; TM = TJ - TL; T1g = Tq - Tk; Tr = Tk + Tq; Tu = Ts * Tt; TS = Ts * Tw; } Tz = ri[WS(rs, 4)]; TC = ii[WS(rs, 4)]; Ty = W[6]; Tv = W[5]; TB = W[7]; } } { E TF, TT, Tx, TV, TD, T1d, TU, TA; TF = FNMS(KP356895867, Tr, Te); TU = Ty * TC; TA = Ty * Tz; TT = FNMS(Tv, Tt, TS); Tx = FMA(Tv, Tw, Tu); TV = FNMS(TB, Tz, TU); TD = FMA(TB, TC, TA); T1d = FNMS(KP356895867, T1a, T19); { E T1b, T15, T17, TW; T17 = FNMS(KP554958132, TR, TM); T1b = TT + TV; TW = TT - TV; { E TE, T1l, T1e, T12; T1i = TD - Tx; TE = Tx + TD; T1l = FNMS(KP356895867, T19, T1b); T1e = FNMS(KP692021471, T1d, T1b); ii[0] = T19 + T1a + T1b + T1c; T12 = FMA(KP554958132, TM, TW); { E TX, T1o, T1j, T14; TX = FMA(KP554958132, TW, TR); T1o = FMA(KP554958132, T1g, T1i); T1j = FMA(KP554958132, T1i, T1h); T14 = FNMS(KP356895867, TE, Tr); { E TZ, TG, T1m, T1f; TZ = FNMS(KP356895867, Te, TE); TG = FNMS(KP692021471, TF, TE); ri[0] = T1 + Te + Tr + TE; T1m = FNMS(KP692021471, T1l, T1a); T1f = FNMS(KP900968867, T1e, T1c); { E T13, TY, T1p, T1k; T13 = FNMS(KP801937735, T12, TR); TY = FMA(KP801937735, TX, TM); T1p = FNMS(KP801937735, T1o, T1h); T1k = FMA(KP801937735, T1j, T1g); T15 = FNMS(KP692021471, T14, Te); { E T10, TH, T1n, T11; T10 = FNMS(KP692021471, TZ, Tr); TH = FNMS(KP900968867, TG, T1); T1n = FNMS(KP900968867, T1m, T1c); ii[WS(rs, 6)] = FNMS(KP974927912, T1k, T1f); ii[WS(rs, 1)] = FMA(KP974927912, T1k, T1f); T11 = FNMS(KP900968867, T10, T1); ri[WS(rs, 1)] = FMA(KP974927912, TY, TH); ri[WS(rs, 6)] = FNMS(KP974927912, TY, TH); ii[WS(rs, 5)] = FNMS(KP974927912, T1p, T1n); ii[WS(rs, 2)] = FMA(KP974927912, T1p, T1n); ri[WS(rs, 2)] = FMA(KP974927912, T13, T11); ri[WS(rs, 5)] = FNMS(KP974927912, T13, T11); T18 = FNMS(KP801937735, T17, TW); } } } } } T16 = FNMS(KP900968867, T15, T1); T1q = FNMS(KP356895867, T1b, T1a); T1t = FNMS(KP554958132, T1h, T1g); } } } ri[WS(rs, 3)] = FMA(KP974927912, T18, T16); ri[WS(rs, 4)] = FNMS(KP974927912, T18, T16); T1r = FNMS(KP692021471, T1q, T19); T1u = FNMS(KP801937735, T1t, T1i); T1s = FNMS(KP900968867, T1r, T1c); ii[WS(rs, 4)] = FNMS(KP974927912, T1u, T1s); ii[WS(rs, 3)] = FMA(KP974927912, T1u, T1s); } } } static const tw_instr twinstr[] = { {TW_FULL, 0, 7}, {TW_NEXT, 1, 0} }; static const ct_desc desc = { 7, "t1_7", twinstr, &GENUS, {18, 12, 54, 0}, 0, 0, 0 }; void X(codelet_t1_7) (planner *p) { X(kdft_dit_register) (p, t1_7, &desc); } #else /* HAVE_FMA */ /* Generated by: ../../../genfft/gen_twiddle.native -compact -variables 4 -pipeline-latency 4 -n 7 -name t1_7 -include t.h */ /* * This function contains 72 FP additions, 60 FP multiplications, * (or, 36 additions, 24 multiplications, 36 fused multiply/add), * 29 stack variables, 6 constants, and 28 memory accesses */ #include "t.h" static void t1_7(R *ri, R *ii, const R *W, stride rs, INT mb, INT me, INT ms) { DK(KP222520933, +0.222520933956314404288902564496794759466355569); DK(KP900968867, +0.900968867902419126236102319507445051165919162); DK(KP623489801, +0.623489801858733530525004884004239810632274731); DK(KP433883739, +0.433883739117558120475768332848358754609990728); DK(KP781831482, +0.781831482468029808708444526674057750232334519); DK(KP974927912, +0.974927912181823607018131682993931217232785801); { INT m; for (m = mb, W = W + (mb * 12); m < me; m = m + 1, ri = ri + ms, ii = ii + ms, W = W + 12, MAKE_VOLATILE_STRIDE(14, rs)) { E T1, TR, Tc, TS, TC, TO, Tn, TT, TI, TP, Ty, TU, TF, TQ; T1 = ri[0]; TR = ii[0]; { E T6, TA, Tb, TB; { E T3, T5, T2, T4; T3 = ri[WS(rs, 1)]; T5 = ii[WS(rs, 1)]; T2 = W[0]; T4 = W[1]; T6 = FMA(T2, T3, T4 * T5); TA = FNMS(T4, T3, T2 * T5); } { E T8, Ta, T7, T9; T8 = ri[WS(rs, 6)]; Ta = ii[WS(rs, 6)]; T7 = W[10]; T9 = W[11]; Tb = FMA(T7, T8, T9 * Ta); TB = FNMS(T9, T8, T7 * Ta); } Tc = T6 + Tb; TS = Tb - T6; TC = TA - TB; TO = TA + TB; } { E Th, TG, Tm, TH; { E Te, Tg, Td, Tf; Te = ri[WS(rs, 2)]; Tg = ii[WS(rs, 2)]; Td = W[2]; Tf = W[3]; Th = FMA(Td, Te, Tf * Tg); TG = FNMS(Tf, Te, Td * Tg); } { E Tj, Tl, Ti, Tk; Tj = ri[WS(rs, 5)]; Tl = ii[WS(rs, 5)]; Ti = W[8]; Tk = W[9]; Tm = FMA(Ti, Tj, Tk * Tl); TH = FNMS(Tk, Tj, Ti * Tl); } Tn = Th + Tm; TT = Tm - Th; TI = TG - TH; TP = TG + TH; } { E Ts, TD, Tx, TE; { E Tp, Tr, To, Tq; Tp = ri[WS(rs, 3)]; Tr = ii[WS(rs, 3)]; To = W[4]; Tq = W[5]; Ts = FMA(To, Tp, Tq * Tr); TD = FNMS(Tq, Tp, To * Tr); } { E Tu, Tw, Tt, Tv; Tu = ri[WS(rs, 4)]; Tw = ii[WS(rs, 4)]; Tt = W[6]; Tv = W[7]; Tx = FMA(Tt, Tu, Tv * Tw); TE = FNMS(Tv, Tu, Tt * Tw); } Ty = Ts + Tx; TU = Tx - Ts; TF = TD - TE; TQ = TD + TE; } ri[0] = T1 + Tc + Tn + Ty; ii[0] = TO + TP + TQ + TR; { E TJ, Tz, TX, TY; TJ = FNMS(KP781831482, TF, KP974927912 * TC) - (KP433883739 * TI); Tz = FMA(KP623489801, Ty, T1) + FNMA(KP900968867, Tn, KP222520933 * Tc); ri[WS(rs, 5)] = Tz - TJ; ri[WS(rs, 2)] = Tz + TJ; TX = FNMS(KP781831482, TU, KP974927912 * TS) - (KP433883739 * TT); TY = FMA(KP623489801, TQ, TR) + FNMA(KP900968867, TP, KP222520933 * TO); ii[WS(rs, 2)] = TX + TY; ii[WS(rs, 5)] = TY - TX; } { E TL, TK, TV, TW; TL = FMA(KP781831482, TC, KP974927912 * TI) + (KP433883739 * TF); TK = FMA(KP623489801, Tc, T1) + FNMA(KP900968867, Ty, KP222520933 * Tn); ri[WS(rs, 6)] = TK - TL; ri[WS(rs, 1)] = TK + TL; TV = FMA(KP781831482, TS, KP974927912 * TT) + (KP433883739 * TU); TW = FMA(KP623489801, TO, TR) + FNMA(KP900968867, TQ, KP222520933 * TP); ii[WS(rs, 1)] = TV + TW; ii[WS(rs, 6)] = TW - TV; } { E TN, TM, TZ, T10; TN = FMA(KP433883739, TC, KP974927912 * TF) - (KP781831482 * TI); TM = FMA(KP623489801, Tn, T1) + FNMA(KP222520933, Ty, KP900968867 * Tc); ri[WS(rs, 4)] = TM - TN; ri[WS(rs, 3)] = TM + TN; TZ = FMA(KP433883739, TS, KP974927912 * TU) - (KP781831482 * TT); T10 = FMA(KP623489801, TP, TR) + FNMA(KP222520933, TQ, KP900968867 * TO); ii[WS(rs, 3)] = TZ + T10; ii[WS(rs, 4)] = T10 - TZ; } } } } static const tw_instr twinstr[] = { {TW_FULL, 0, 7}, {TW_NEXT, 1, 0} }; static const ct_desc desc = { 7, "t1_7", twinstr, &GENUS, {36, 24, 36, 0}, 0, 0, 0 }; void X(codelet_t1_7) (planner *p) { X(kdft_dit_register) (p, t1_7, &desc); } #endif /* HAVE_FMA */ fftw-3.3.3/dft/scalar/codelets/t1_8.c0000644000175000017500000002223312054410245014151 00000000000000/* * Copyright (c) 2003, 2007-11 Matteo Frigo * Copyright (c) 2003, 2007-11 Massachusetts Institute of Technology * * This program is free software; you can redistribute it and/or modify * it under the terms of the GNU General Public License as published by * the Free Software Foundation; either version 2 of the License, or * (at your option) any later version. * * This program is distributed in the hope that it will be useful, * but WITHOUT ANY WARRANTY; without even the implied warranty of * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the * GNU General Public License for more details. * * You should have received a copy of the GNU General Public License * along with this program; if not, write to the Free Software * Foundation, Inc., 51 Franklin Street, Fifth Floor, Boston, MA 02110-1301 USA * */ /* This file was automatically generated --- DO NOT EDIT */ /* Generated on Sun Nov 25 07:35:49 EST 2012 */ #include "codelet-dft.h" #ifdef HAVE_FMA /* Generated by: ../../../genfft/gen_twiddle.native -fma -reorder-insns -schedule-for-pipeline -compact -variables 4 -pipeline-latency 4 -n 8 -name t1_8 -include t.h */ /* * This function contains 66 FP additions, 36 FP multiplications, * (or, 44 additions, 14 multiplications, 22 fused multiply/add), * 61 stack variables, 1 constants, and 32 memory accesses */ #include "t.h" static void t1_8(R *ri, R *ii, const R *W, stride rs, INT mb, INT me, INT ms) { DK(KP707106781, +0.707106781186547524400844362104849039284835938); { INT m; for (m = mb, W = W + (mb * 14); m < me; m = m + 1, ri = ri + ms, ii = ii + ms, W = W + 14, MAKE_VOLATILE_STRIDE(16, rs)) { E T1g, T1f, T1e, Tm, T1q, T1o, T1p, TN, T1h, T1i; { E T1, T1m, T1l, T7, TS, Tk, TQ, Te, To, Tr, T17, TM, T12, Tu, TW; E Tp, Tx, Tt, Tq, Tw; { E T3, T6, T2, T5; T1 = ri[0]; T1m = ii[0]; T3 = ri[WS(rs, 4)]; T6 = ii[WS(rs, 4)]; T2 = W[6]; T5 = W[7]; { E Ta, Td, T9, Tc; { E Tg, Tj, Ti, TR, Th, T1k, T4, Tf; Tg = ri[WS(rs, 6)]; Tj = ii[WS(rs, 6)]; T1k = T2 * T6; T4 = T2 * T3; Tf = W[10]; Ti = W[11]; T1l = FNMS(T5, T3, T1k); T7 = FMA(T5, T6, T4); TR = Tf * Tj; Th = Tf * Tg; Ta = ri[WS(rs, 2)]; Td = ii[WS(rs, 2)]; TS = FNMS(Ti, Tg, TR); Tk = FMA(Ti, Tj, Th); T9 = W[2]; Tc = W[3]; } { E TB, TE, TH, T13, TC, TK, TG, TD, TJ, TP, Tb, TA, Tn; TB = ri[WS(rs, 7)]; TE = ii[WS(rs, 7)]; TP = T9 * Td; Tb = T9 * Ta; TA = W[12]; TH = ri[WS(rs, 3)]; TQ = FNMS(Tc, Ta, TP); Te = FMA(Tc, Td, Tb); T13 = TA * TE; TC = TA * TB; TK = ii[WS(rs, 3)]; TG = W[4]; TD = W[13]; TJ = W[5]; { E T14, TF, T16, TL, T15, TI; To = ri[WS(rs, 1)]; T15 = TG * TK; TI = TG * TH; T14 = FNMS(TD, TB, T13); TF = FMA(TD, TE, TC); T16 = FNMS(TJ, TH, T15); TL = FMA(TJ, TK, TI); Tr = ii[WS(rs, 1)]; Tn = W[0]; T17 = T14 - T16; T1g = T14 + T16; TM = TF + TL; T12 = TF - TL; } Tu = ri[WS(rs, 5)]; TW = Tn * Tr; Tp = Tn * To; Tx = ii[WS(rs, 5)]; Tt = W[8]; Tq = W[1]; Tw = W[9]; } } } { E T8, T1j, T1n, Tz, T1a, TU, Tl, T1b, T1c, T1v, T1t, T1w, T19, T1u, T1d; { E T1r, T10, TV, T1s, T11, T18; { E TO, TX, Ts, TZ, Ty, TT, TY, Tv; T8 = T1 + T7; TO = T1 - T7; TY = Tt * Tx; Tv = Tt * Tu; TX = FNMS(Tq, To, TW); Ts = FMA(Tq, Tr, Tp); TZ = FNMS(Tw, Tu, TY); Ty = FMA(Tw, Tx, Tv); TT = TQ - TS; T1j = TQ + TS; T1n = T1l + T1m; T1r = T1m - T1l; T10 = TX - TZ; T1f = TX + TZ; Tz = Ts + Ty; TV = Ts - Ty; T1a = TO - TT; TU = TO + TT; T1s = Te - Tk; Tl = Te + Tk; } T1b = T10 - TV; T11 = TV + T10; T18 = T12 - T17; T1c = T12 + T17; T1v = T1s + T1r; T1t = T1r - T1s; T1w = T18 - T11; T19 = T11 + T18; } ii[WS(rs, 3)] = FMA(KP707106781, T1w, T1v); ii[WS(rs, 7)] = FNMS(KP707106781, T1w, T1v); ri[WS(rs, 1)] = FMA(KP707106781, T19, TU); ri[WS(rs, 5)] = FNMS(KP707106781, T19, TU); T1u = T1b + T1c; T1d = T1b - T1c; ii[WS(rs, 1)] = FMA(KP707106781, T1u, T1t); ii[WS(rs, 5)] = FNMS(KP707106781, T1u, T1t); ri[WS(rs, 3)] = FMA(KP707106781, T1d, T1a); ri[WS(rs, 7)] = FNMS(KP707106781, T1d, T1a); T1e = T8 - Tl; Tm = T8 + Tl; T1q = T1n - T1j; T1o = T1j + T1n; T1p = TM - Tz; TN = Tz + TM; } } ii[WS(rs, 2)] = T1p + T1q; ii[WS(rs, 6)] = T1q - T1p; ri[0] = Tm + TN; ri[WS(rs, 4)] = Tm - TN; T1h = T1f - T1g; T1i = T1f + T1g; ii[0] = T1i + T1o; ii[WS(rs, 4)] = T1o - T1i; ri[WS(rs, 2)] = T1e + T1h; ri[WS(rs, 6)] = T1e - T1h; } } } static const tw_instr twinstr[] = { {TW_FULL, 0, 8}, {TW_NEXT, 1, 0} }; static const ct_desc desc = { 8, "t1_8", twinstr, &GENUS, {44, 14, 22, 0}, 0, 0, 0 }; void X(codelet_t1_8) (planner *p) { X(kdft_dit_register) (p, t1_8, &desc); } #else /* HAVE_FMA */ /* Generated by: ../../../genfft/gen_twiddle.native -compact -variables 4 -pipeline-latency 4 -n 8 -name t1_8 -include t.h */ /* * This function contains 66 FP additions, 32 FP multiplications, * (or, 52 additions, 18 multiplications, 14 fused multiply/add), * 28 stack variables, 1 constants, and 32 memory accesses */ #include "t.h" static void t1_8(R *ri, R *ii, const R *W, stride rs, INT mb, INT me, INT ms) { DK(KP707106781, +0.707106781186547524400844362104849039284835938); { INT m; for (m = mb, W = W + (mb * 14); m < me; m = m + 1, ri = ri + ms, ii = ii + ms, W = W + 14, MAKE_VOLATILE_STRIDE(16, rs)) { E T7, T1e, TH, T19, TF, T13, TR, TU, Ti, T1f, TK, T16, Tu, T12, TM; E TP; { E T1, T18, T6, T17; T1 = ri[0]; T18 = ii[0]; { E T3, T5, T2, T4; T3 = ri[WS(rs, 4)]; T5 = ii[WS(rs, 4)]; T2 = W[6]; T4 = W[7]; T6 = FMA(T2, T3, T4 * T5); T17 = FNMS(T4, T3, T2 * T5); } T7 = T1 + T6; T1e = T18 - T17; TH = T1 - T6; T19 = T17 + T18; } { E Tz, TS, TE, TT; { E Tw, Ty, Tv, Tx; Tw = ri[WS(rs, 7)]; Ty = ii[WS(rs, 7)]; Tv = W[12]; Tx = W[13]; Tz = FMA(Tv, Tw, Tx * Ty); TS = FNMS(Tx, Tw, Tv * Ty); } { E TB, TD, TA, TC; TB = ri[WS(rs, 3)]; TD = ii[WS(rs, 3)]; TA = W[4]; TC = W[5]; TE = FMA(TA, TB, TC * TD); TT = FNMS(TC, TB, TA * TD); } TF = Tz + TE; T13 = TS + TT; TR = Tz - TE; TU = TS - TT; } { E Tc, TI, Th, TJ; { E T9, Tb, T8, Ta; T9 = ri[WS(rs, 2)]; Tb = ii[WS(rs, 2)]; T8 = W[2]; Ta = W[3]; Tc = FMA(T8, T9, Ta * Tb); TI = FNMS(Ta, T9, T8 * Tb); } { E Te, Tg, Td, Tf; Te = ri[WS(rs, 6)]; Tg = ii[WS(rs, 6)]; Td = W[10]; Tf = W[11]; Th = FMA(Td, Te, Tf * Tg); TJ = FNMS(Tf, Te, Td * Tg); } Ti = Tc + Th; T1f = Tc - Th; TK = TI - TJ; T16 = TI + TJ; } { E To, TN, Tt, TO; { E Tl, Tn, Tk, Tm; Tl = ri[WS(rs, 1)]; Tn = ii[WS(rs, 1)]; Tk = W[0]; Tm = W[1]; To = FMA(Tk, Tl, Tm * Tn); TN = FNMS(Tm, Tl, Tk * Tn); } { E Tq, Ts, Tp, Tr; Tq = ri[WS(rs, 5)]; Ts = ii[WS(rs, 5)]; Tp = W[8]; Tr = W[9]; Tt = FMA(Tp, Tq, Tr * Ts); TO = FNMS(Tr, Tq, Tp * Ts); } Tu = To + Tt; T12 = TN + TO; TM = To - Tt; TP = TN - TO; } { E Tj, TG, T1b, T1c; Tj = T7 + Ti; TG = Tu + TF; ri[WS(rs, 4)] = Tj - TG; ri[0] = Tj + TG; { E T15, T1a, T11, T14; T15 = T12 + T13; T1a = T16 + T19; ii[0] = T15 + T1a; ii[WS(rs, 4)] = T1a - T15; T11 = T7 - Ti; T14 = T12 - T13; ri[WS(rs, 6)] = T11 - T14; ri[WS(rs, 2)] = T11 + T14; } T1b = TF - Tu; T1c = T19 - T16; ii[WS(rs, 2)] = T1b + T1c; ii[WS(rs, 6)] = T1c - T1b; { E TX, T1g, T10, T1d, TY, TZ; TX = TH - TK; T1g = T1e - T1f; TY = TP - TM; TZ = TR + TU; T10 = KP707106781 * (TY - TZ); T1d = KP707106781 * (TY + TZ); ri[WS(rs, 7)] = TX - T10; ii[WS(rs, 5)] = T1g - T1d; ri[WS(rs, 3)] = TX + T10; ii[WS(rs, 1)] = T1d + T1g; } { E TL, T1i, TW, T1h, TQ, TV; TL = TH + TK; T1i = T1f + T1e; TQ = TM + TP; TV = TR - TU; TW = KP707106781 * (TQ + TV); T1h = KP707106781 * (TV - TQ); ri[WS(rs, 5)] = TL - TW; ii[WS(rs, 7)] = T1i - T1h; ri[WS(rs, 1)] = TL + TW; ii[WS(rs, 3)] = T1h + T1i; } } } } } static const tw_instr twinstr[] = { {TW_FULL, 0, 8}, {TW_NEXT, 1, 0} }; static const ct_desc desc = { 8, "t1_8", twinstr, &GENUS, {52, 18, 14, 0}, 0, 0, 0 }; void X(codelet_t1_8) (planner *p) { X(kdft_dit_register) (p, t1_8, &desc); } #endif /* HAVE_FMA */ fftw-3.3.3/dft/scalar/codelets/t1_9.c0000644000175000017500000003410412054410245014152 00000000000000/* * Copyright (c) 2003, 2007-11 Matteo Frigo * Copyright (c) 2003, 2007-11 Massachusetts Institute of Technology * * This program is free software; you can redistribute it and/or modify * it under the terms of the GNU General Public License as published by * the Free Software Foundation; either version 2 of the License, or * (at your option) any later version. * * This program is distributed in the hope that it will be useful, * but WITHOUT ANY WARRANTY; without even the implied warranty of * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the * GNU General Public License for more details. * * You should have received a copy of the GNU General Public License * along with this program; if not, write to the Free Software * Foundation, Inc., 51 Franklin Street, Fifth Floor, Boston, MA 02110-1301 USA * */ /* This file was automatically generated --- DO NOT EDIT */ /* Generated on Sun Nov 25 07:35:49 EST 2012 */ #include "codelet-dft.h" #ifdef HAVE_FMA /* Generated by: ../../../genfft/gen_twiddle.native -fma -reorder-insns -schedule-for-pipeline -compact -variables 4 -pipeline-latency 4 -n 9 -name t1_9 -include t.h */ /* * This function contains 96 FP additions, 88 FP multiplications, * (or, 24 additions, 16 multiplications, 72 fused multiply/add), * 72 stack variables, 10 constants, and 36 memory accesses */ #include "t.h" static void t1_9(R *ri, R *ii, const R *W, stride rs, INT mb, INT me, INT ms) { DK(KP954188894, +0.954188894138671133499268364187245676532219158); DK(KP852868531, +0.852868531952443209628250963940074071936020296); DK(KP363970234, +0.363970234266202361351047882776834043890471784); DK(KP492403876, +0.492403876506104029683371512294761506835321626); DK(KP984807753, +0.984807753012208059366743024589523013670643252); DK(KP777861913, +0.777861913430206160028177977318626690410586096); DK(KP839099631, +0.839099631177280011763127298123181364687434283); DK(KP176326980, +0.176326980708464973471090386868618986121633062); DK(KP866025403, +0.866025403784438646763723170752936183471402627); DK(KP500000000, +0.500000000000000000000000000000000000000000000); { INT m; for (m = mb, W = W + (mb * 16); m < me; m = m + 1, ri = ri + ms, ii = ii + ms, W = W + 16, MAKE_VOLATILE_STRIDE(18, rs)) { E T1K, T24, T1H, T23; { E T1, T1R, T1Q, T10, T1W, Te, TB, T1l, T1r, T1q, T1M, TE, T1g, Tz, T12; E TC, TH, TK, T17, TR, TG, TJ, TD; T1 = ri[0]; T1R = ii[0]; { E T9, Tc, TY, Ta, Tb, TX, T7; { E T3, T6, T8, TW, T4, T2, T5; T3 = ri[WS(rs, 3)]; T6 = ii[WS(rs, 3)]; T2 = W[4]; T9 = ri[WS(rs, 6)]; Tc = ii[WS(rs, 6)]; T8 = W[10]; TW = T2 * T6; T4 = T2 * T3; T5 = W[5]; TY = T8 * Tc; Ta = T8 * T9; Tb = W[11]; TX = FNMS(T5, T3, TW); T7 = FMA(T5, T6, T4); } { E Th, Tk, Ti, T1n, Tn, Tq, Tp, T1i, Tx, T1j, To, Tj, TZ, Td, Tg; E TA, Tl, Ty; Th = ri[WS(rs, 1)]; TZ = FNMS(Tb, T9, TY); Td = FMA(Tb, Tc, Ta); Tk = ii[WS(rs, 1)]; Tg = W[0]; T1Q = TX + TZ; T10 = TX - TZ; T1W = Td - T7; Te = T7 + Td; Ti = Tg * Th; T1n = Tg * Tk; { E Tt, Tw, Ts, Tv, T1h, Tu, Tm; Tt = ri[WS(rs, 7)]; Tw = ii[WS(rs, 7)]; Ts = W[12]; Tv = W[13]; Tn = ri[WS(rs, 4)]; Tq = ii[WS(rs, 4)]; T1h = Ts * Tw; Tu = Ts * Tt; Tm = W[6]; Tp = W[7]; T1i = FNMS(Tv, Tt, T1h); Tx = FMA(Tv, Tw, Tu); T1j = Tm * Tq; To = Tm * Tn; } Tj = W[1]; TB = ri[WS(rs, 2)]; { E T1k, Tr, T1o, T1p; T1k = FNMS(Tp, Tn, T1j); Tr = FMA(Tp, Tq, To); T1o = FNMS(Tj, Th, T1n); Tl = FMA(Tj, Tk, Ti); T1p = T1k + T1i; T1l = T1i - T1k; Ty = Tr + Tx; T1r = Tr - Tx; T1q = FNMS(KP500000000, T1p, T1o); T1M = T1o + T1p; TE = ii[WS(rs, 2)]; } T1g = FNMS(KP500000000, Ty, Tl); Tz = Tl + Ty; TA = W[2]; { E TN, TQ, TP, T16, TO, TM; TN = ri[WS(rs, 8)]; TQ = ii[WS(rs, 8)]; TM = W[14]; T12 = TA * TE; TC = TA * TB; TP = W[15]; T16 = TM * TQ; TO = TM * TN; TH = ri[WS(rs, 5)]; TK = ii[WS(rs, 5)]; T17 = FNMS(TP, TN, T16); TR = FMA(TP, TQ, TO); TG = W[8]; TJ = W[9]; } TD = W[3]; } } { E TV, Tf, T1S, T1V, T1d, T1a, T19, T1N, TT, T1c; { E T13, TF, T15, TL, T14, TI, TS, T18; TV = FNMS(KP500000000, Te, T1); Tf = T1 + Te; T14 = TG * TK; TI = TG * TH; T13 = FNMS(TD, TB, T12); TF = FMA(TD, TE, TC); T15 = FNMS(TJ, TH, T14); TL = FMA(TJ, TK, TI); T1S = T1Q + T1R; T1V = FNMS(KP500000000, T1Q, T1R); T18 = T15 + T17; T1d = T15 - T17; TS = TL + TR; T1a = TR - TL; T19 = FNMS(KP500000000, T18, T13); T1N = T13 + T18; TT = TF + TS; T1c = FNMS(KP500000000, TS, TF); } { E T11, T1z, T1E, T1D, T21, T1X, T1I, T1C, T1Y, T1y, T20, T1u, T1U, TU; T1U = TT - Tz; TU = Tz + TT; { E T1P, T1O, T1L, T1T; T1P = T1M + T1N; T1O = T1M - T1N; T11 = FMA(KP866025403, T10, TV); T1z = FNMS(KP866025403, T10, TV); T1L = FNMS(KP500000000, TU, Tf); ri[0] = Tf + TU; T1T = FNMS(KP500000000, T1P, T1S); ii[0] = T1P + T1S; ri[WS(rs, 3)] = FMA(KP866025403, T1O, T1L); ri[WS(rs, 6)] = FNMS(KP866025403, T1O, T1L); ii[WS(rs, 6)] = FNMS(KP866025403, T1U, T1T); ii[WS(rs, 3)] = FMA(KP866025403, T1U, T1T); } { E T1B, T1m, T1w, T1f, T1s, T1A, T1b, T1e, T1x, T1t; T1E = FNMS(KP866025403, T1a, T19); T1b = FMA(KP866025403, T1a, T19); T1e = FMA(KP866025403, T1d, T1c); T1D = FNMS(KP866025403, T1d, T1c); T1B = FMA(KP866025403, T1l, T1g); T1m = FNMS(KP866025403, T1l, T1g); T21 = FNMS(KP866025403, T1W, T1V); T1X = FMA(KP866025403, T1W, T1V); T1w = FNMS(KP176326980, T1b, T1e); T1f = FMA(KP176326980, T1e, T1b); T1s = FNMS(KP866025403, T1r, T1q); T1A = FMA(KP866025403, T1r, T1q); T1x = FNMS(KP839099631, T1m, T1s); T1t = FMA(KP839099631, T1s, T1m); T1I = FNMS(KP176326980, T1A, T1B); T1C = FMA(KP176326980, T1B, T1A); T1Y = FNMS(KP777861913, T1x, T1w); T1y = FMA(KP777861913, T1x, T1w); T20 = FNMS(KP777861913, T1t, T1f); T1u = FMA(KP777861913, T1t, T1f); } { E T22, T1G, T1Z, T1F, T1J, T1v; ii[WS(rs, 1)] = FNMS(KP984807753, T1Y, T1X); T1v = FNMS(KP492403876, T1u, T11); ri[WS(rs, 1)] = FMA(KP984807753, T1u, T11); T1F = FNMS(KP363970234, T1E, T1D); T1J = FMA(KP363970234, T1D, T1E); ri[WS(rs, 4)] = FMA(KP852868531, T1y, T1v); ri[WS(rs, 7)] = FNMS(KP852868531, T1y, T1v); T1K = FNMS(KP954188894, T1J, T1I); T22 = FMA(KP954188894, T1J, T1I); T1G = FNMS(KP954188894, T1F, T1C); T24 = FMA(KP954188894, T1F, T1C); T1Z = FMA(KP492403876, T1Y, T1X); ii[WS(rs, 2)] = FNMS(KP984807753, T22, T21); ri[WS(rs, 2)] = FMA(KP984807753, T1G, T1z); T1H = FNMS(KP492403876, T1G, T1z); ii[WS(rs, 7)] = FNMS(KP852868531, T20, T1Z); ii[WS(rs, 4)] = FMA(KP852868531, T20, T1Z); T23 = FMA(KP492403876, T22, T21); } } } } ri[WS(rs, 8)] = FMA(KP852868531, T1K, T1H); ri[WS(rs, 5)] = FNMS(KP852868531, T1K, T1H); ii[WS(rs, 8)] = FMA(KP852868531, T24, T23); ii[WS(rs, 5)] = FNMS(KP852868531, T24, T23); } } } static const tw_instr twinstr[] = { {TW_FULL, 0, 9}, {TW_NEXT, 1, 0} }; static const ct_desc desc = { 9, "t1_9", twinstr, &GENUS, {24, 16, 72, 0}, 0, 0, 0 }; void X(codelet_t1_9) (planner *p) { X(kdft_dit_register) (p, t1_9, &desc); } #else /* HAVE_FMA */ /* Generated by: ../../../genfft/gen_twiddle.native -compact -variables 4 -pipeline-latency 4 -n 9 -name t1_9 -include t.h */ /* * This function contains 96 FP additions, 72 FP multiplications, * (or, 60 additions, 36 multiplications, 36 fused multiply/add), * 41 stack variables, 8 constants, and 36 memory accesses */ #include "t.h" static void t1_9(R *ri, R *ii, const R *W, stride rs, INT mb, INT me, INT ms) { DK(KP939692620, +0.939692620785908384054109277324731469936208134); DK(KP342020143, +0.342020143325668733044099614682259580763083368); DK(KP984807753, +0.984807753012208059366743024589523013670643252); DK(KP173648177, +0.173648177666930348851716626769314796000375677); DK(KP642787609, +0.642787609686539326322643409907263432907559884); DK(KP766044443, +0.766044443118978035202392650555416673935832457); DK(KP500000000, +0.500000000000000000000000000000000000000000000); DK(KP866025403, +0.866025403784438646763723170752936183471402627); { INT m; for (m = mb, W = W + (mb * 16); m < me; m = m + 1, ri = ri + ms, ii = ii + ms, W = W + 16, MAKE_VOLATILE_STRIDE(18, rs)) { E T1, T1B, TQ, T1G, Tc, TN, T1A, T1H, TL, T1x, T17, T1o, T1c, T1n, Tu; E T1w, TW, T1k, T11, T1l; { E T6, TO, Tb, TP; T1 = ri[0]; T1B = ii[0]; { E T3, T5, T2, T4; T3 = ri[WS(rs, 3)]; T5 = ii[WS(rs, 3)]; T2 = W[4]; T4 = W[5]; T6 = FMA(T2, T3, T4 * T5); TO = FNMS(T4, T3, T2 * T5); } { E T8, Ta, T7, T9; T8 = ri[WS(rs, 6)]; Ta = ii[WS(rs, 6)]; T7 = W[10]; T9 = W[11]; Tb = FMA(T7, T8, T9 * Ta); TP = FNMS(T9, T8, T7 * Ta); } TQ = KP866025403 * (TO - TP); T1G = KP866025403 * (Tb - T6); Tc = T6 + Tb; TN = FNMS(KP500000000, Tc, T1); T1A = TO + TP; T1H = FNMS(KP500000000, T1A, T1B); } { E Tz, T19, TE, T14, TJ, T15, TK, T1a; { E Tw, Ty, Tv, Tx; Tw = ri[WS(rs, 2)]; Ty = ii[WS(rs, 2)]; Tv = W[2]; Tx = W[3]; Tz = FMA(Tv, Tw, Tx * Ty); T19 = FNMS(Tx, Tw, Tv * Ty); } { E TB, TD, TA, TC; TB = ri[WS(rs, 5)]; TD = ii[WS(rs, 5)]; TA = W[8]; TC = W[9]; TE = FMA(TA, TB, TC * TD); T14 = FNMS(TC, TB, TA * TD); } { E TG, TI, TF, TH; TG = ri[WS(rs, 8)]; TI = ii[WS(rs, 8)]; TF = W[14]; TH = W[15]; TJ = FMA(TF, TG, TH * TI); T15 = FNMS(TH, TG, TF * TI); } TK = TE + TJ; T1a = T14 + T15; TL = Tz + TK; T1x = T19 + T1a; { E T13, T16, T18, T1b; T13 = FNMS(KP500000000, TK, Tz); T16 = KP866025403 * (T14 - T15); T17 = T13 + T16; T1o = T13 - T16; T18 = KP866025403 * (TJ - TE); T1b = FNMS(KP500000000, T1a, T19); T1c = T18 + T1b; T1n = T1b - T18; } } { E Ti, TY, Tn, TT, Ts, TU, Tt, TZ; { E Tf, Th, Te, Tg; Tf = ri[WS(rs, 1)]; Th = ii[WS(rs, 1)]; Te = W[0]; Tg = W[1]; Ti = FMA(Te, Tf, Tg * Th); TY = FNMS(Tg, Tf, Te * Th); } { E Tk, Tm, Tj, Tl; Tk = ri[WS(rs, 4)]; Tm = ii[WS(rs, 4)]; Tj = W[6]; Tl = W[7]; Tn = FMA(Tj, Tk, Tl * Tm); TT = FNMS(Tl, Tk, Tj * Tm); } { E Tp, Tr, To, Tq; Tp = ri[WS(rs, 7)]; Tr = ii[WS(rs, 7)]; To = W[12]; Tq = W[13]; Ts = FMA(To, Tp, Tq * Tr); TU = FNMS(Tq, Tp, To * Tr); } Tt = Tn + Ts; TZ = TT + TU; Tu = Ti + Tt; T1w = TY + TZ; { E TS, TV, TX, T10; TS = FNMS(KP500000000, Tt, Ti); TV = KP866025403 * (TT - TU); TW = TS + TV; T1k = TS - TV; TX = KP866025403 * (Ts - Tn); T10 = FNMS(KP500000000, TZ, TY); T11 = TX + T10; T1l = T10 - TX; } } { E T1y, Td, TM, T1v; T1y = KP866025403 * (T1w - T1x); Td = T1 + Tc; TM = Tu + TL; T1v = FNMS(KP500000000, TM, Td); ri[0] = Td + TM; ri[WS(rs, 3)] = T1v + T1y; ri[WS(rs, 6)] = T1v - T1y; } { E T1D, T1z, T1C, T1E; T1D = KP866025403 * (TL - Tu); T1z = T1w + T1x; T1C = T1A + T1B; T1E = FNMS(KP500000000, T1z, T1C); ii[0] = T1z + T1C; ii[WS(rs, 6)] = T1E - T1D; ii[WS(rs, 3)] = T1D + T1E; } { E TR, T1I, T1e, T1J, T1i, T1F, T1f, T1K; TR = TN + TQ; T1I = T1G + T1H; { E T12, T1d, T1g, T1h; T12 = FMA(KP766044443, TW, KP642787609 * T11); T1d = FMA(KP173648177, T17, KP984807753 * T1c); T1e = T12 + T1d; T1J = KP866025403 * (T1d - T12); T1g = FNMS(KP642787609, TW, KP766044443 * T11); T1h = FNMS(KP984807753, T17, KP173648177 * T1c); T1i = KP866025403 * (T1g - T1h); T1F = T1g + T1h; } ri[WS(rs, 1)] = TR + T1e; ii[WS(rs, 1)] = T1F + T1I; T1f = FNMS(KP500000000, T1e, TR); ri[WS(rs, 7)] = T1f - T1i; ri[WS(rs, 4)] = T1f + T1i; T1K = FNMS(KP500000000, T1F, T1I); ii[WS(rs, 4)] = T1J + T1K; ii[WS(rs, 7)] = T1K - T1J; } { E T1j, T1M, T1q, T1N, T1u, T1L, T1r, T1O; T1j = TN - TQ; T1M = T1H - T1G; { E T1m, T1p, T1s, T1t; T1m = FMA(KP173648177, T1k, KP984807753 * T1l); T1p = FNMS(KP939692620, T1o, KP342020143 * T1n); T1q = T1m + T1p; T1N = KP866025403 * (T1p - T1m); T1s = FNMS(KP984807753, T1k, KP173648177 * T1l); T1t = FMA(KP342020143, T1o, KP939692620 * T1n); T1u = KP866025403 * (T1s + T1t); T1L = T1s - T1t; } ri[WS(rs, 2)] = T1j + T1q; ii[WS(rs, 2)] = T1L + T1M; T1r = FNMS(KP500000000, T1q, T1j); ri[WS(rs, 8)] = T1r - T1u; ri[WS(rs, 5)] = T1r + T1u; T1O = FNMS(KP500000000, T1L, T1M); ii[WS(rs, 5)] = T1N + T1O; ii[WS(rs, 8)] = T1O - T1N; } } } } static const tw_instr twinstr[] = { {TW_FULL, 0, 9}, {TW_NEXT, 1, 0} }; static const ct_desc desc = { 9, "t1_9", twinstr, &GENUS, {60, 36, 36, 0}, 0, 0, 0 }; void X(codelet_t1_9) (planner *p) { X(kdft_dit_register) (p, t1_9, &desc); } #endif /* HAVE_FMA */ fftw-3.3.3/dft/scalar/codelets/t1_10.c0000644000175000017500000003333212054410246014225 00000000000000/* * Copyright (c) 2003, 2007-11 Matteo Frigo * Copyright (c) 2003, 2007-11 Massachusetts Institute of Technology * * This program is free software; you can redistribute it and/or modify * it under the terms of the GNU General Public License as published by * the Free Software Foundation; either version 2 of the License, or * (at your option) any later version. * * This program is distributed in the hope that it will be useful, * but WITHOUT ANY WARRANTY; without even the implied warranty of * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the * GNU General Public License for more details. * * You should have received a copy of the GNU General Public License * along with this program; if not, write to the Free Software * Foundation, Inc., 51 Franklin Street, Fifth Floor, Boston, MA 02110-1301 USA * */ /* This file was automatically generated --- DO NOT EDIT */ /* Generated on Sun Nov 25 07:35:49 EST 2012 */ #include "codelet-dft.h" #ifdef HAVE_FMA /* Generated by: ../../../genfft/gen_twiddle.native -fma -reorder-insns -schedule-for-pipeline -compact -variables 4 -pipeline-latency 4 -n 10 -name t1_10 -include t.h */ /* * This function contains 102 FP additions, 72 FP multiplications, * (or, 48 additions, 18 multiplications, 54 fused multiply/add), * 70 stack variables, 4 constants, and 40 memory accesses */ #include "t.h" static void t1_10(R *ri, R *ii, const R *W, stride rs, INT mb, INT me, INT ms) { DK(KP951056516, +0.951056516295153572116439333379382143405698634); DK(KP559016994, +0.559016994374947424102293417182819058860154590); DK(KP250000000, +0.250000000000000000000000000000000000000000000); DK(KP618033988, +0.618033988749894848204586834365638117720309180); { INT m; for (m = mb, W = W + (mb * 18); m < me; m = m + 1, ri = ri + ms, ii = ii + ms, W = W + 18, MAKE_VOLATILE_STRIDE(20, rs)) { E T1X, T21, T20, T22; { E T23, T1U, T8, T12, T1y, T25, T1P, T1H, T1Y, T18, T10, T2b, T1K, T1O, T15; E T1Z, T2a, Tz, T24, T1n; { E T1, T1T, T3, T6, T2, T5; T1 = ri[0]; T1T = ii[0]; T3 = ri[WS(rs, 5)]; T6 = ii[WS(rs, 5)]; T2 = W[8]; T5 = W[9]; { E T1w, TY, T1s, T1F, TM, T16, T1u, TS; { E TF, T1p, TO, TR, T1r, TL, TN, TQ, T1t, TP; { E TU, TX, TT, TW; { E TB, TE, T1R, T4, TA, TD; TB = ri[WS(rs, 4)]; TE = ii[WS(rs, 4)]; T1R = T2 * T6; T4 = T2 * T3; TA = W[6]; TD = W[7]; { E T1S, T7, T1o, TC; T1S = FNMS(T5, T3, T1R); T7 = FMA(T5, T6, T4); T1o = TA * TE; TC = TA * TB; T23 = T1T - T1S; T1U = T1S + T1T; T8 = T1 - T7; T12 = T1 + T7; TF = FMA(TD, TE, TC); T1p = FNMS(TD, TB, T1o); } } TU = ri[WS(rs, 1)]; TX = ii[WS(rs, 1)]; TT = W[0]; TW = W[1]; { E TH, TK, TJ, T1q, TI, T1v, TV, TG; TH = ri[WS(rs, 9)]; TK = ii[WS(rs, 9)]; T1v = TT * TX; TV = TT * TU; TG = W[16]; TJ = W[17]; T1w = FNMS(TW, TU, T1v); TY = FMA(TW, TX, TV); T1q = TG * TK; TI = TG * TH; TO = ri[WS(rs, 6)]; TR = ii[WS(rs, 6)]; T1r = FNMS(TJ, TH, T1q); TL = FMA(TJ, TK, TI); TN = W[10]; TQ = W[11]; } } T1s = T1p - T1r; T1F = T1p + T1r; TM = TF - TL; T16 = TF + TL; T1t = TN * TR; TP = TN * TO; T1u = FNMS(TQ, TO, T1t); TS = FMA(TQ, TR, TP); } { E T1e, Te, T1l, Tx, Tn, Tq, Tp, T1g, Tk, T1i, To; { E Tt, Tw, Tv, T1k, Tu; { E Ta, Td, T9, Tc, T1d, Tb, Ts; Ta = ri[WS(rs, 2)]; Td = ii[WS(rs, 2)]; { E T1G, T1x, TZ, T17; T1G = T1u + T1w; T1x = T1u - T1w; TZ = TS - TY; T17 = TS + TY; T1y = T1s - T1x; T25 = T1s + T1x; T1P = T1F + T1G; T1H = T1F - T1G; T1Y = T16 - T17; T18 = T16 + T17; T10 = TM + TZ; T2b = TM - TZ; T9 = W[2]; } Tc = W[3]; Tt = ri[WS(rs, 3)]; Tw = ii[WS(rs, 3)]; T1d = T9 * Td; Tb = T9 * Ta; Ts = W[4]; Tv = W[5]; T1e = FNMS(Tc, Ta, T1d); Te = FMA(Tc, Td, Tb); T1k = Ts * Tw; Tu = Ts * Tt; } { E Tg, Tj, Tf, Ti, T1f, Th, Tm; Tg = ri[WS(rs, 7)]; Tj = ii[WS(rs, 7)]; T1l = FNMS(Tv, Tt, T1k); Tx = FMA(Tv, Tw, Tu); Tf = W[12]; Ti = W[13]; Tn = ri[WS(rs, 8)]; Tq = ii[WS(rs, 8)]; T1f = Tf * Tj; Th = Tf * Tg; Tm = W[14]; Tp = W[15]; T1g = FNMS(Ti, Tg, T1f); Tk = FMA(Ti, Tj, Th); T1i = Tm * Tq; To = Tm * Tn; } } { E T1h, T1I, Tl, T13, T1j, Tr; T1h = T1e - T1g; T1I = T1e + T1g; Tl = Te - Tk; T13 = Te + Tk; T1j = FNMS(Tp, Tn, T1i); Tr = FMA(Tp, Tq, To); { E T1m, T1J, T14, Ty; T1m = T1j - T1l; T1J = T1j + T1l; T14 = Tr + Tx; Ty = Tr - Tx; T1K = T1I - T1J; T1O = T1I + T1J; T15 = T13 + T14; T1Z = T13 - T14; T2a = Tl - Ty; Tz = Tl + Ty; T24 = T1h + T1m; T1n = T1h - T1m; } } } } } { E T2c, T2e, T29, T2d; { E T1b, T11, T26, T28, T27; T1b = Tz - T10; T11 = Tz + T10; T26 = T24 + T25; T28 = T24 - T25; { E T1B, T1z, T1a, T1A, T1c; T1B = FNMS(KP618033988, T1n, T1y); T1z = FMA(KP618033988, T1y, T1n); ri[WS(rs, 5)] = T8 + T11; T1a = FNMS(KP250000000, T11, T8); T1A = FNMS(KP559016994, T1b, T1a); T1c = FMA(KP559016994, T1b, T1a); T27 = FNMS(KP250000000, T26, T23); T2c = FMA(KP618033988, T2b, T2a); T2e = FNMS(KP618033988, T2a, T2b); ri[WS(rs, 1)] = FMA(KP951056516, T1z, T1c); ri[WS(rs, 9)] = FNMS(KP951056516, T1z, T1c); ri[WS(rs, 3)] = FMA(KP951056516, T1B, T1A); ri[WS(rs, 7)] = FNMS(KP951056516, T1B, T1A); } ii[WS(rs, 5)] = T26 + T23; T29 = FMA(KP559016994, T28, T27); T2d = FNMS(KP559016994, T28, T27); } { E T1E, T1M, T1L, T1N, T19, T1D, T1C, T1Q, T1W, T1V; T19 = T15 + T18; T1D = T15 - T18; ii[WS(rs, 7)] = FMA(KP951056516, T2e, T2d); ii[WS(rs, 3)] = FNMS(KP951056516, T2e, T2d); ii[WS(rs, 9)] = FMA(KP951056516, T2c, T29); ii[WS(rs, 1)] = FNMS(KP951056516, T2c, T29); T1C = FNMS(KP250000000, T19, T12); ri[0] = T12 + T19; T1E = FNMS(KP559016994, T1D, T1C); T1M = FMA(KP559016994, T1D, T1C); T1L = FNMS(KP618033988, T1K, T1H); T1N = FMA(KP618033988, T1H, T1K); T1Q = T1O + T1P; T1W = T1O - T1P; ri[WS(rs, 6)] = FMA(KP951056516, T1N, T1M); ri[WS(rs, 4)] = FNMS(KP951056516, T1N, T1M); ri[WS(rs, 8)] = FMA(KP951056516, T1L, T1E); ri[WS(rs, 2)] = FNMS(KP951056516, T1L, T1E); T1V = FNMS(KP250000000, T1Q, T1U); ii[0] = T1Q + T1U; T1X = FNMS(KP559016994, T1W, T1V); T21 = FMA(KP559016994, T1W, T1V); T20 = FNMS(KP618033988, T1Z, T1Y); T22 = FMA(KP618033988, T1Y, T1Z); } } } ii[WS(rs, 6)] = FNMS(KP951056516, T22, T21); ii[WS(rs, 4)] = FMA(KP951056516, T22, T21); ii[WS(rs, 8)] = FNMS(KP951056516, T20, T1X); ii[WS(rs, 2)] = FMA(KP951056516, T20, T1X); } } } static const tw_instr twinstr[] = { {TW_FULL, 0, 10}, {TW_NEXT, 1, 0} }; static const ct_desc desc = { 10, "t1_10", twinstr, &GENUS, {48, 18, 54, 0}, 0, 0, 0 }; void X(codelet_t1_10) (planner *p) { X(kdft_dit_register) (p, t1_10, &desc); } #else /* HAVE_FMA */ /* Generated by: ../../../genfft/gen_twiddle.native -compact -variables 4 -pipeline-latency 4 -n 10 -name t1_10 -include t.h */ /* * This function contains 102 FP additions, 60 FP multiplications, * (or, 72 additions, 30 multiplications, 30 fused multiply/add), * 45 stack variables, 4 constants, and 40 memory accesses */ #include "t.h" static void t1_10(R *ri, R *ii, const R *W, stride rs, INT mb, INT me, INT ms) { DK(KP587785252, +0.587785252292473129168705954639072768597652438); DK(KP951056516, +0.951056516295153572116439333379382143405698634); DK(KP250000000, +0.250000000000000000000000000000000000000000000); DK(KP559016994, +0.559016994374947424102293417182819058860154590); { INT m; for (m = mb, W = W + (mb * 18); m < me; m = m + 1, ri = ri + ms, ii = ii + ms, W = W + 18, MAKE_VOLATILE_STRIDE(20, rs)) { E T7, T1O, TT, T1C, TF, TQ, TR, T1o, T1p, T1y, TX, TY, TZ, T1d, T1g; E T1M, Ti, Tt, Tu, T1r, T1s, T1x, TU, TV, TW, T16, T19, T1L; { E T1, T1B, T6, T1A; T1 = ri[0]; T1B = ii[0]; { E T3, T5, T2, T4; T3 = ri[WS(rs, 5)]; T5 = ii[WS(rs, 5)]; T2 = W[8]; T4 = W[9]; T6 = FMA(T2, T3, T4 * T5); T1A = FNMS(T4, T3, T2 * T5); } T7 = T1 - T6; T1O = T1B - T1A; TT = T1 + T6; T1C = T1A + T1B; } { E Tz, T1b, TP, T1f, TE, T1c, TK, T1e; { E Tw, Ty, Tv, Tx; Tw = ri[WS(rs, 4)]; Ty = ii[WS(rs, 4)]; Tv = W[6]; Tx = W[7]; Tz = FMA(Tv, Tw, Tx * Ty); T1b = FNMS(Tx, Tw, Tv * Ty); } { E TM, TO, TL, TN; TM = ri[WS(rs, 1)]; TO = ii[WS(rs, 1)]; TL = W[0]; TN = W[1]; TP = FMA(TL, TM, TN * TO); T1f = FNMS(TN, TM, TL * TO); } { E TB, TD, TA, TC; TB = ri[WS(rs, 9)]; TD = ii[WS(rs, 9)]; TA = W[16]; TC = W[17]; TE = FMA(TA, TB, TC * TD); T1c = FNMS(TC, TB, TA * TD); } { E TH, TJ, TG, TI; TH = ri[WS(rs, 6)]; TJ = ii[WS(rs, 6)]; TG = W[10]; TI = W[11]; TK = FMA(TG, TH, TI * TJ); T1e = FNMS(TI, TH, TG * TJ); } TF = Tz - TE; TQ = TK - TP; TR = TF + TQ; T1o = T1b + T1c; T1p = T1e + T1f; T1y = T1o + T1p; TX = Tz + TE; TY = TK + TP; TZ = TX + TY; T1d = T1b - T1c; T1g = T1e - T1f; T1M = T1d + T1g; } { E Tc, T14, Ts, T18, Th, T15, Tn, T17; { E T9, Tb, T8, Ta; T9 = ri[WS(rs, 2)]; Tb = ii[WS(rs, 2)]; T8 = W[2]; Ta = W[3]; Tc = FMA(T8, T9, Ta * Tb); T14 = FNMS(Ta, T9, T8 * Tb); } { E Tp, Tr, To, Tq; Tp = ri[WS(rs, 3)]; Tr = ii[WS(rs, 3)]; To = W[4]; Tq = W[5]; Ts = FMA(To, Tp, Tq * Tr); T18 = FNMS(Tq, Tp, To * Tr); } { E Te, Tg, Td, Tf; Te = ri[WS(rs, 7)]; Tg = ii[WS(rs, 7)]; Td = W[12]; Tf = W[13]; Th = FMA(Td, Te, Tf * Tg); T15 = FNMS(Tf, Te, Td * Tg); } { E Tk, Tm, Tj, Tl; Tk = ri[WS(rs, 8)]; Tm = ii[WS(rs, 8)]; Tj = W[14]; Tl = W[15]; Tn = FMA(Tj, Tk, Tl * Tm); T17 = FNMS(Tl, Tk, Tj * Tm); } Ti = Tc - Th; Tt = Tn - Ts; Tu = Ti + Tt; T1r = T14 + T15; T1s = T17 + T18; T1x = T1r + T1s; TU = Tc + Th; TV = Tn + Ts; TW = TU + TV; T16 = T14 - T15; T19 = T17 - T18; T1L = T16 + T19; } { E T11, TS, T12, T1i, T1k, T1a, T1h, T1j, T13; T11 = KP559016994 * (Tu - TR); TS = Tu + TR; T12 = FNMS(KP250000000, TS, T7); T1a = T16 - T19; T1h = T1d - T1g; T1i = FMA(KP951056516, T1a, KP587785252 * T1h); T1k = FNMS(KP587785252, T1a, KP951056516 * T1h); ri[WS(rs, 5)] = T7 + TS; T1j = T12 - T11; ri[WS(rs, 7)] = T1j - T1k; ri[WS(rs, 3)] = T1j + T1k; T13 = T11 + T12; ri[WS(rs, 9)] = T13 - T1i; ri[WS(rs, 1)] = T13 + T1i; } { E T1N, T1P, T1Q, T1U, T1W, T1S, T1T, T1V, T1R; T1N = KP559016994 * (T1L - T1M); T1P = T1L + T1M; T1Q = FNMS(KP250000000, T1P, T1O); T1S = Ti - Tt; T1T = TF - TQ; T1U = FMA(KP951056516, T1S, KP587785252 * T1T); T1W = FNMS(KP587785252, T1S, KP951056516 * T1T); ii[WS(rs, 5)] = T1P + T1O; T1V = T1Q - T1N; ii[WS(rs, 3)] = T1V - T1W; ii[WS(rs, 7)] = T1W + T1V; T1R = T1N + T1Q; ii[WS(rs, 1)] = T1R - T1U; ii[WS(rs, 9)] = T1U + T1R; } { E T1m, T10, T1l, T1u, T1w, T1q, T1t, T1v, T1n; T1m = KP559016994 * (TW - TZ); T10 = TW + TZ; T1l = FNMS(KP250000000, T10, TT); T1q = T1o - T1p; T1t = T1r - T1s; T1u = FNMS(KP587785252, T1t, KP951056516 * T1q); T1w = FMA(KP951056516, T1t, KP587785252 * T1q); ri[0] = TT + T10; T1v = T1m + T1l; ri[WS(rs, 4)] = T1v - T1w; ri[WS(rs, 6)] = T1v + T1w; T1n = T1l - T1m; ri[WS(rs, 2)] = T1n - T1u; ri[WS(rs, 8)] = T1n + T1u; } { E T1H, T1z, T1G, T1F, T1J, T1D, T1E, T1K, T1I; T1H = KP559016994 * (T1x - T1y); T1z = T1x + T1y; T1G = FNMS(KP250000000, T1z, T1C); T1D = TX - TY; T1E = TU - TV; T1F = FNMS(KP587785252, T1E, KP951056516 * T1D); T1J = FMA(KP951056516, T1E, KP587785252 * T1D); ii[0] = T1z + T1C; T1K = T1H + T1G; ii[WS(rs, 4)] = T1J + T1K; ii[WS(rs, 6)] = T1K - T1J; T1I = T1G - T1H; ii[WS(rs, 2)] = T1F + T1I; ii[WS(rs, 8)] = T1I - T1F; } } } } static const tw_instr twinstr[] = { {TW_FULL, 0, 10}, {TW_NEXT, 1, 0} }; static const ct_desc desc = { 10, "t1_10", twinstr, &GENUS, {72, 30, 30, 0}, 0, 0, 0 }; void X(codelet_t1_10) (planner *p) { X(kdft_dit_register) (p, t1_10, &desc); } #endif /* HAVE_FMA */ fftw-3.3.3/dft/scalar/codelets/t1_12.c0000644000175000017500000003504212054410246014227 00000000000000/* * Copyright (c) 2003, 2007-11 Matteo Frigo * Copyright (c) 2003, 2007-11 Massachusetts Institute of Technology * * This program is free software; you can redistribute it and/or modify * it under the terms of the GNU General Public License as published by * the Free Software Foundation; either version 2 of the License, or * (at your option) any later version. * * This program is distributed in the hope that it will be useful, * but WITHOUT ANY WARRANTY; without even the implied warranty of * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the * GNU General Public License for more details. * * You should have received a copy of the GNU General Public License * along with this program; if not, write to the Free Software * Foundation, Inc., 51 Franklin Street, Fifth Floor, Boston, MA 02110-1301 USA * */ /* This file was automatically generated --- DO NOT EDIT */ /* Generated on Sun Nov 25 07:35:50 EST 2012 */ #include "codelet-dft.h" #ifdef HAVE_FMA /* Generated by: ../../../genfft/gen_twiddle.native -fma -reorder-insns -schedule-for-pipeline -compact -variables 4 -pipeline-latency 4 -n 12 -name t1_12 -include t.h */ /* * This function contains 118 FP additions, 68 FP multiplications, * (or, 72 additions, 22 multiplications, 46 fused multiply/add), * 84 stack variables, 2 constants, and 48 memory accesses */ #include "t.h" static void t1_12(R *ri, R *ii, const R *W, stride rs, INT mb, INT me, INT ms) { DK(KP866025403, +0.866025403784438646763723170752936183471402627); DK(KP500000000, +0.500000000000000000000000000000000000000000000); { INT m; for (m = mb, W = W + (mb * 22); m < me; m = m + 1, ri = ri + ms, ii = ii + ms, W = W + 22, MAKE_VOLATILE_STRIDE(24, rs)) { E T2B, T2C; { E T1, T2i, T2e, Tl, T1Y, T10, T1S, TG, T2f, T1s, T2r, Ty, T1Z, T1H, T21; E T1d, TI, TL, T2h, T1l, T2o, Te, TJ, T1w, TO, TR, TN, TK, TQ; { E TW, TZ, TY, T1X, TX; T1 = ri[0]; T2i = ii[0]; { E Th, Tk, Tg, Tj, T2d, Ti, TV; Th = ri[WS(rs, 6)]; Tk = ii[WS(rs, 6)]; Tg = W[10]; Tj = W[11]; TW = ri[WS(rs, 9)]; TZ = ii[WS(rs, 9)]; T2d = Tg * Tk; Ti = Tg * Th; TV = W[16]; TY = W[17]; T2e = FNMS(Tj, Th, T2d); Tl = FMA(Tj, Tk, Ti); T1X = TV * TZ; TX = TV * TW; } { E Tn, Tq, Tt, T1o, To, Tw, Ts, Tp, Tv; { E TC, TF, TB, TE, T1R, TD, Tm; TC = ri[WS(rs, 3)]; TF = ii[WS(rs, 3)]; T1Y = FNMS(TY, TW, T1X); T10 = FMA(TY, TZ, TX); TB = W[4]; TE = W[5]; Tn = ri[WS(rs, 10)]; Tq = ii[WS(rs, 10)]; T1R = TB * TF; TD = TB * TC; Tm = W[18]; Tt = ri[WS(rs, 2)]; T1S = FNMS(TE, TC, T1R); TG = FMA(TE, TF, TD); T1o = Tm * Tq; To = Tm * Tn; Tw = ii[WS(rs, 2)]; Ts = W[2]; Tp = W[19]; Tv = W[3]; } { E T12, T15, T13, T1D, T18, T1b, T17, T14, T1a; { E T1p, Tr, T1r, Tx, T1q, Tu, T11; T12 = ri[WS(rs, 1)]; T1q = Ts * Tw; Tu = Ts * Tt; T1p = FNMS(Tp, Tn, T1o); Tr = FMA(Tp, Tq, To); T1r = FNMS(Tv, Tt, T1q); Tx = FMA(Tv, Tw, Tu); T15 = ii[WS(rs, 1)]; T11 = W[0]; T2f = T1p + T1r; T1s = T1p - T1r; T2r = Tx - Tr; Ty = Tr + Tx; T13 = T11 * T12; T1D = T11 * T15; } T18 = ri[WS(rs, 5)]; T1b = ii[WS(rs, 5)]; T17 = W[8]; T14 = W[1]; T1a = W[9]; { E T3, T6, T4, T1h, T9, Tc, T8, T5, Tb; { E T1E, T16, T1G, T1c, T1F, T19, T2; T3 = ri[WS(rs, 4)]; T1F = T17 * T1b; T19 = T17 * T18; T1E = FNMS(T14, T12, T1D); T16 = FMA(T14, T15, T13); T1G = FNMS(T1a, T18, T1F); T1c = FMA(T1a, T1b, T19); T6 = ii[WS(rs, 4)]; T2 = W[6]; T1Z = T1E + T1G; T1H = T1E - T1G; T21 = T1c - T16; T1d = T16 + T1c; T4 = T2 * T3; T1h = T2 * T6; } T9 = ri[WS(rs, 8)]; Tc = ii[WS(rs, 8)]; T8 = W[14]; T5 = W[7]; Tb = W[15]; { E T1i, T7, T1k, Td, T1j, Ta, TH; TI = ri[WS(rs, 7)]; T1j = T8 * Tc; Ta = T8 * T9; T1i = FNMS(T5, T3, T1h); T7 = FMA(T5, T6, T4); T1k = FNMS(Tb, T9, T1j); Td = FMA(Tb, Tc, Ta); TL = ii[WS(rs, 7)]; TH = W[12]; T2h = T1i + T1k; T1l = T1i - T1k; T2o = Td - T7; Te = T7 + Td; TJ = TH * TI; T1w = TH * TL; } TO = ri[WS(rs, 11)]; TR = ii[WS(rs, 11)]; TN = W[20]; TK = W[13]; TQ = W[21]; } } } } { E T1g, T1n, T2q, T1A, T1V, T28, TA, T2n, T1v, T1C, T1U, T29, T2m, T2k, T2l; E T1f, T2a, T20; { E T2g, T1T, TT, T2j, TU, T1e; { E Tf, T1x, TM, T1z, TS, Tz, T1y, TP; T1g = FNMS(KP500000000, Te, T1); Tf = T1 + Te; T1y = TN * TR; TP = TN * TO; T1x = FNMS(TK, TI, T1w); TM = FMA(TK, TL, TJ); T1z = FNMS(TQ, TO, T1y); TS = FMA(TQ, TR, TP); Tz = Tl + Ty; T1n = FNMS(KP500000000, Ty, Tl); T2q = FNMS(KP500000000, T2f, T2e); T2g = T2e + T2f; T1T = T1x + T1z; T1A = T1x - T1z; T1V = TS - TM; TT = TM + TS; T28 = Tf - Tz; TA = Tf + Tz; T2j = T2h + T2i; T2n = FNMS(KP500000000, T2h, T2i); } T1v = FNMS(KP500000000, TT, TG); TU = TG + TT; T1e = T10 + T1d; T1C = FNMS(KP500000000, T1d, T10); T1U = FNMS(KP500000000, T1T, T1S); T29 = T1S + T1T; T2m = T2j - T2g; T2k = T2g + T2j; T2l = TU - T1e; T1f = TU + T1e; T2a = T1Y + T1Z; T20 = FNMS(KP500000000, T1Z, T1Y); } { E T1m, T1K, T2y, T2p, T2x, T2s, T1L, T1t, T1B, T1N, T2c, T2b; ii[WS(rs, 9)] = T2m - T2l; ii[WS(rs, 3)] = T2l + T2m; ri[0] = TA + T1f; ri[WS(rs, 6)] = TA - T1f; T2c = T29 + T2a; T2b = T29 - T2a; T1m = FNMS(KP866025403, T1l, T1g); T1K = FMA(KP866025403, T1l, T1g); ii[0] = T2c + T2k; ii[WS(rs, 6)] = T2k - T2c; ri[WS(rs, 9)] = T28 + T2b; ri[WS(rs, 3)] = T28 - T2b; T2y = FNMS(KP866025403, T2o, T2n); T2p = FMA(KP866025403, T2o, T2n); T2x = FNMS(KP866025403, T2r, T2q); T2s = FMA(KP866025403, T2r, T2q); T1L = FMA(KP866025403, T1s, T1n); T1t = FNMS(KP866025403, T1s, T1n); T1B = FNMS(KP866025403, T1A, T1v); T1N = FMA(KP866025403, T1A, T1v); { E T24, T27, T1Q, T2u, T23, T2v, T2w, T2t; { E T1u, T1W, T22, T1O, T1I, T2z, T2A, T25, T26, T1M, T1J, T1P; T24 = T1m - T1t; T1u = T1m + T1t; T25 = FNMS(KP866025403, T1V, T1U); T1W = FMA(KP866025403, T1V, T1U); T26 = FNMS(KP866025403, T21, T20); T22 = FMA(KP866025403, T21, T20); T1O = FMA(KP866025403, T1H, T1C); T1I = FNMS(KP866025403, T1H, T1C); T2z = T2x + T2y; T2B = T2y - T2x; T27 = T25 - T26; T2A = T25 + T26; T1M = T1K + T1L; T1Q = T1K - T1L; T2C = T1B - T1I; T1J = T1B + T1I; T1P = T1N + T1O; T2u = T1N - T1O; ii[WS(rs, 8)] = T2A + T2z; ii[WS(rs, 2)] = T2z - T2A; ri[WS(rs, 8)] = T1u + T1J; ri[WS(rs, 2)] = T1u - T1J; ri[WS(rs, 10)] = T1M - T1P; ri[WS(rs, 4)] = T1M + T1P; T23 = T1W - T22; T2v = T1W + T22; T2w = T2s + T2p; T2t = T2p - T2s; } ii[WS(rs, 10)] = T2w - T2v; ii[WS(rs, 4)] = T2v + T2w; ri[WS(rs, 1)] = T1Q + T23; ri[WS(rs, 7)] = T1Q - T23; ii[WS(rs, 7)] = T2u + T2t; ii[WS(rs, 1)] = T2t - T2u; ri[WS(rs, 5)] = T24 + T27; ri[WS(rs, 11)] = T24 - T27; } } } } ii[WS(rs, 11)] = T2C + T2B; ii[WS(rs, 5)] = T2B - T2C; } } } static const tw_instr twinstr[] = { {TW_FULL, 0, 12}, {TW_NEXT, 1, 0} }; static const ct_desc desc = { 12, "t1_12", twinstr, &GENUS, {72, 22, 46, 0}, 0, 0, 0 }; void X(codelet_t1_12) (planner *p) { X(kdft_dit_register) (p, t1_12, &desc); } #else /* HAVE_FMA */ /* Generated by: ../../../genfft/gen_twiddle.native -compact -variables 4 -pipeline-latency 4 -n 12 -name t1_12 -include t.h */ /* * This function contains 118 FP additions, 60 FP multiplications, * (or, 88 additions, 30 multiplications, 30 fused multiply/add), * 47 stack variables, 2 constants, and 48 memory accesses */ #include "t.h" static void t1_12(R *ri, R *ii, const R *W, stride rs, INT mb, INT me, INT ms) { DK(KP500000000, +0.500000000000000000000000000000000000000000000); DK(KP866025403, +0.866025403784438646763723170752936183471402627); { INT m; for (m = mb, W = W + (mb * 22); m < me; m = m + 1, ri = ri + ms, ii = ii + ms, W = W + 22, MAKE_VOLATILE_STRIDE(24, rs)) { E T1, T1W, T18, T21, Tc, T15, T1V, T22, TR, T1E, T1o, T1D, T12, T1l, T1F; E T1G, Ti, T1S, T1d, T24, Tt, T1a, T1T, T25, TA, T1z, T1j, T1y, TL, T1g; E T1A, T1B; { E T6, T16, Tb, T17; T1 = ri[0]; T1W = ii[0]; { E T3, T5, T2, T4; T3 = ri[WS(rs, 4)]; T5 = ii[WS(rs, 4)]; T2 = W[6]; T4 = W[7]; T6 = FMA(T2, T3, T4 * T5); T16 = FNMS(T4, T3, T2 * T5); } { E T8, Ta, T7, T9; T8 = ri[WS(rs, 8)]; Ta = ii[WS(rs, 8)]; T7 = W[14]; T9 = W[15]; Tb = FMA(T7, T8, T9 * Ta); T17 = FNMS(T9, T8, T7 * Ta); } T18 = KP866025403 * (T16 - T17); T21 = KP866025403 * (Tb - T6); Tc = T6 + Tb; T15 = FNMS(KP500000000, Tc, T1); T1V = T16 + T17; T22 = FNMS(KP500000000, T1V, T1W); } { E T11, T1n, TW, T1m; { E TO, TQ, TN, TP; TO = ri[WS(rs, 9)]; TQ = ii[WS(rs, 9)]; TN = W[16]; TP = W[17]; TR = FMA(TN, TO, TP * TQ); T1E = FNMS(TP, TO, TN * TQ); } { E TY, T10, TX, TZ; TY = ri[WS(rs, 5)]; T10 = ii[WS(rs, 5)]; TX = W[8]; TZ = W[9]; T11 = FMA(TX, TY, TZ * T10); T1n = FNMS(TZ, TY, TX * T10); } { E TT, TV, TS, TU; TT = ri[WS(rs, 1)]; TV = ii[WS(rs, 1)]; TS = W[0]; TU = W[1]; TW = FMA(TS, TT, TU * TV); T1m = FNMS(TU, TT, TS * TV); } T1o = KP866025403 * (T1m - T1n); T1D = KP866025403 * (T11 - TW); T12 = TW + T11; T1l = FNMS(KP500000000, T12, TR); T1F = T1m + T1n; T1G = FNMS(KP500000000, T1F, T1E); } { E Ts, T1c, Tn, T1b; { E Tf, Th, Te, Tg; Tf = ri[WS(rs, 6)]; Th = ii[WS(rs, 6)]; Te = W[10]; Tg = W[11]; Ti = FMA(Te, Tf, Tg * Th); T1S = FNMS(Tg, Tf, Te * Th); } { E Tp, Tr, To, Tq; Tp = ri[WS(rs, 2)]; Tr = ii[WS(rs, 2)]; To = W[2]; Tq = W[3]; Ts = FMA(To, Tp, Tq * Tr); T1c = FNMS(Tq, Tp, To * Tr); } { E Tk, Tm, Tj, Tl; Tk = ri[WS(rs, 10)]; Tm = ii[WS(rs, 10)]; Tj = W[18]; Tl = W[19]; Tn = FMA(Tj, Tk, Tl * Tm); T1b = FNMS(Tl, Tk, Tj * Tm); } T1d = KP866025403 * (T1b - T1c); T24 = KP866025403 * (Ts - Tn); Tt = Tn + Ts; T1a = FNMS(KP500000000, Tt, Ti); T1T = T1b + T1c; T25 = FNMS(KP500000000, T1T, T1S); } { E TK, T1i, TF, T1h; { E Tx, Tz, Tw, Ty; Tx = ri[WS(rs, 3)]; Tz = ii[WS(rs, 3)]; Tw = W[4]; Ty = W[5]; TA = FMA(Tw, Tx, Ty * Tz); T1z = FNMS(Ty, Tx, Tw * Tz); } { E TH, TJ, TG, TI; TH = ri[WS(rs, 11)]; TJ = ii[WS(rs, 11)]; TG = W[20]; TI = W[21]; TK = FMA(TG, TH, TI * TJ); T1i = FNMS(TI, TH, TG * TJ); } { E TC, TE, TB, TD; TC = ri[WS(rs, 7)]; TE = ii[WS(rs, 7)]; TB = W[12]; TD = W[13]; TF = FMA(TB, TC, TD * TE); T1h = FNMS(TD, TC, TB * TE); } T1j = KP866025403 * (T1h - T1i); T1y = KP866025403 * (TK - TF); TL = TF + TK; T1g = FNMS(KP500000000, TL, TA); T1A = T1h + T1i; T1B = FNMS(KP500000000, T1A, T1z); } { E Tv, T1N, T1Y, T20, T14, T1Z, T1Q, T1R; { E Td, Tu, T1U, T1X; Td = T1 + Tc; Tu = Ti + Tt; Tv = Td + Tu; T1N = Td - Tu; T1U = T1S + T1T; T1X = T1V + T1W; T1Y = T1U + T1X; T20 = T1X - T1U; } { E TM, T13, T1O, T1P; TM = TA + TL; T13 = TR + T12; T14 = TM + T13; T1Z = TM - T13; T1O = T1z + T1A; T1P = T1E + T1F; T1Q = T1O - T1P; T1R = T1O + T1P; } ri[WS(rs, 6)] = Tv - T14; ii[WS(rs, 6)] = T1Y - T1R; ri[0] = Tv + T14; ii[0] = T1R + T1Y; ri[WS(rs, 3)] = T1N - T1Q; ii[WS(rs, 3)] = T1Z + T20; ri[WS(rs, 9)] = T1N + T1Q; ii[WS(rs, 9)] = T20 - T1Z; } { E T1t, T1x, T27, T2a, T1w, T28, T1I, T29; { E T1r, T1s, T23, T26; T1r = T15 + T18; T1s = T1a + T1d; T1t = T1r + T1s; T1x = T1r - T1s; T23 = T21 + T22; T26 = T24 + T25; T27 = T23 - T26; T2a = T26 + T23; } { E T1u, T1v, T1C, T1H; T1u = T1g + T1j; T1v = T1l + T1o; T1w = T1u + T1v; T28 = T1u - T1v; T1C = T1y + T1B; T1H = T1D + T1G; T1I = T1C - T1H; T29 = T1C + T1H; } ri[WS(rs, 10)] = T1t - T1w; ii[WS(rs, 10)] = T2a - T29; ri[WS(rs, 4)] = T1t + T1w; ii[WS(rs, 4)] = T29 + T2a; ri[WS(rs, 7)] = T1x - T1I; ii[WS(rs, 7)] = T28 + T27; ri[WS(rs, 1)] = T1x + T1I; ii[WS(rs, 1)] = T27 - T28; } { E T1f, T1J, T2d, T2f, T1q, T2g, T1M, T2e; { E T19, T1e, T2b, T2c; T19 = T15 - T18; T1e = T1a - T1d; T1f = T19 + T1e; T1J = T19 - T1e; T2b = T25 - T24; T2c = T22 - T21; T2d = T2b + T2c; T2f = T2c - T2b; } { E T1k, T1p, T1K, T1L; T1k = T1g - T1j; T1p = T1l - T1o; T1q = T1k + T1p; T2g = T1k - T1p; T1K = T1B - T1y; T1L = T1G - T1D; T1M = T1K - T1L; T2e = T1K + T1L; } ri[WS(rs, 2)] = T1f - T1q; ii[WS(rs, 2)] = T2d - T2e; ri[WS(rs, 8)] = T1f + T1q; ii[WS(rs, 8)] = T2e + T2d; ri[WS(rs, 11)] = T1J - T1M; ii[WS(rs, 11)] = T2g + T2f; ri[WS(rs, 5)] = T1J + T1M; ii[WS(rs, 5)] = T2f - T2g; } } } } static const tw_instr twinstr[] = { {TW_FULL, 0, 12}, {TW_NEXT, 1, 0} }; static const ct_desc desc = { 12, "t1_12", twinstr, &GENUS, {88, 30, 30, 0}, 0, 0, 0 }; void X(codelet_t1_12) (planner *p) { X(kdft_dit_register) (p, t1_12, &desc); } #endif /* HAVE_FMA */ fftw-3.3.3/dft/scalar/codelets/t1_15.c0000644000175000017500000005372012054410247014236 00000000000000/* * Copyright (c) 2003, 2007-11 Matteo Frigo * Copyright (c) 2003, 2007-11 Massachusetts Institute of Technology * * This program is free software; you can redistribute it and/or modify * it under the terms of the GNU General Public License as published by * the Free Software Foundation; either version 2 of the License, or * (at your option) any later version. * * This program is distributed in the hope that it will be useful, * but WITHOUT ANY WARRANTY; without even the implied warranty of * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the * GNU General Public License for more details. * * You should have received a copy of the GNU General Public License * along with this program; if not, write to the Free Software * Foundation, Inc., 51 Franklin Street, Fifth Floor, Boston, MA 02110-1301 USA * */ /* This file was automatically generated --- DO NOT EDIT */ /* Generated on Sun Nov 25 07:35:50 EST 2012 */ #include "codelet-dft.h" #ifdef HAVE_FMA /* Generated by: ../../../genfft/gen_twiddle.native -fma -reorder-insns -schedule-for-pipeline -compact -variables 4 -pipeline-latency 4 -n 15 -name t1_15 -include t.h */ /* * This function contains 184 FP additions, 140 FP multiplications, * (or, 72 additions, 28 multiplications, 112 fused multiply/add), * 89 stack variables, 6 constants, and 60 memory accesses */ #include "t.h" static void t1_15(R *ri, R *ii, const R *W, stride rs, INT mb, INT me, INT ms) { DK(KP951056516, +0.951056516295153572116439333379382143405698634); DK(KP559016994, +0.559016994374947424102293417182819058860154590); DK(KP250000000, +0.250000000000000000000000000000000000000000000); DK(KP618033988, +0.618033988749894848204586834365638117720309180); DK(KP866025403, +0.866025403784438646763723170752936183471402627); DK(KP500000000, +0.500000000000000000000000000000000000000000000); { INT m; for (m = mb, W = W + (mb * 28); m < me; m = m + 1, ri = ri + ms, ii = ii + ms, W = W + 28, MAKE_VOLATILE_STRIDE(30, rs)) { E T2d, T2O, T2Q, T2m, T2k, T2l, T2P, T2n; { E T1G, T3u, T3k, T3t, T1B, Tf, T37, T1y, T2V, T2M, T2a, T2i, T39, Tz, T2X; E T2t, T1O, T2e, T3a, TT, T10, T2Y, T2z, T1V, T2f, T2C, T12, T15, T14, T21; E T1c, T1Y, T13; { E T2I, T1k, T1m, T1p, T1o, T28, T1w, T25, T1n; { E T1, T3j, T9, Tc, Tb, T1D, T7, T1E, Ta, T1j, T1i, T1h; T1 = ri[0]; T3j = ii[0]; { E T3, T6, T2, T5, T1C, T4, T8; T3 = ri[WS(rs, 5)]; T6 = ii[WS(rs, 5)]; T2 = W[8]; T5 = W[9]; T9 = ri[WS(rs, 10)]; Tc = ii[WS(rs, 10)]; T1C = T2 * T6; T4 = T2 * T3; T8 = W[18]; Tb = W[19]; T1D = FNMS(T5, T3, T1C); T7 = FMA(T5, T6, T4); T1E = T8 * Tc; Ta = T8 * T9; } { E T1g, T1F, Td, T1f, T3i, Te, T2H; T1g = ri[WS(rs, 9)]; T1j = ii[WS(rs, 9)]; T1F = FNMS(Tb, T9, T1E); Td = FMA(Tb, Tc, Ta); T1f = W[16]; T1i = W[17]; T1G = T1D - T1F; T3i = T1D + T1F; T3u = Td - T7; Te = T7 + Td; T2H = T1f * T1j; T1h = T1f * T1g; T3k = T3i + T3j; T3t = FNMS(KP500000000, T3i, T3j); T1B = FNMS(KP500000000, Te, T1); Tf = T1 + Te; T2I = FNMS(T1i, T1g, T2H); } T1k = FMA(T1i, T1j, T1h); { E T1s, T1v, T1r, T1u, T27, T1t, T1l; T1s = ri[WS(rs, 4)]; T1v = ii[WS(rs, 4)]; T1r = W[6]; T1u = W[7]; T1m = ri[WS(rs, 14)]; T1p = ii[WS(rs, 14)]; T27 = T1r * T1v; T1t = T1r * T1s; T1l = W[26]; T1o = W[27]; T28 = FNMS(T1u, T1s, T27); T1w = FMA(T1u, T1v, T1t); T25 = T1l * T1p; T1n = T1l * T1m; } } { E Tl, T2p, Tn, Tq, Tp, T1M, Tx, T1J, To; { E Th, Tk, T26, T1q, Tg, Tj; Th = ri[WS(rs, 3)]; Tk = ii[WS(rs, 3)]; T26 = FNMS(T1o, T1m, T25); T1q = FMA(T1o, T1p, T1n); Tg = W[4]; Tj = W[5]; { E T29, T2J, T1x, T2L; T29 = T26 - T28; T2J = T26 + T28; T1x = T1q + T1w; T2L = T1w - T1q; { E T2o, Ti, T2K, T24; T2o = Tg * Tk; Ti = Tg * Th; T2K = FNMS(KP500000000, T2J, T2I); T37 = T2I + T2J; T24 = FNMS(KP500000000, T1x, T1k); T1y = T1k + T1x; Tl = FMA(Tj, Tk, Ti); T2V = FNMS(KP866025403, T2L, T2K); T2M = FMA(KP866025403, T2L, T2K); T2a = FNMS(KP866025403, T29, T24); T2i = FMA(KP866025403, T29, T24); T2p = FNMS(Tj, Th, T2o); } } } { E Tt, Tw, Ts, Tv, T1L, Tu, Tm; Tt = ri[WS(rs, 13)]; Tw = ii[WS(rs, 13)]; Ts = W[24]; Tv = W[25]; Tn = ri[WS(rs, 8)]; Tq = ii[WS(rs, 8)]; T1L = Ts * Tw; Tu = Ts * Tt; Tm = W[14]; Tp = W[15]; T1M = FNMS(Tv, Tt, T1L); Tx = FMA(Tv, Tw, Tu); T1J = Tm * Tq; To = Tm * Tn; } { E TF, T2v, TH, TK, TJ, T1T, TR, T1Q, TI; { E TB, TE, T1K, Tr, TA, TD; TB = ri[WS(rs, 12)]; TE = ii[WS(rs, 12)]; T1K = FNMS(Tp, Tn, T1J); Tr = FMA(Tp, Tq, To); TA = W[22]; TD = W[23]; { E T1N, T2q, Ty, T2s; T1N = T1K - T1M; T2q = T1K + T1M; Ty = Tr + Tx; T2s = Tx - Tr; { E T2u, TC, T2r, T1I; T2u = TA * TE; TC = TA * TB; T2r = FNMS(KP500000000, T2q, T2p); T39 = T2p + T2q; T1I = FNMS(KP500000000, Ty, Tl); Tz = Tl + Ty; TF = FMA(TD, TE, TC); T2X = FNMS(KP866025403, T2s, T2r); T2t = FMA(KP866025403, T2s, T2r); T1O = FNMS(KP866025403, T1N, T1I); T2e = FMA(KP866025403, T1N, T1I); T2v = FNMS(TD, TB, T2u); } } } { E TN, TQ, TM, TP, T1S, TO, TG; TN = ri[WS(rs, 7)]; TQ = ii[WS(rs, 7)]; TM = W[12]; TP = W[13]; TH = ri[WS(rs, 2)]; TK = ii[WS(rs, 2)]; T1S = TM * TQ; TO = TM * TN; TG = W[2]; TJ = W[3]; T1T = FNMS(TP, TN, T1S); TR = FMA(TP, TQ, TO); T1Q = TG * TK; TI = TG * TH; } { E TW, TZ, T1R, TL, TV, TY; TW = ri[WS(rs, 6)]; TZ = ii[WS(rs, 6)]; T1R = FNMS(TJ, TH, T1Q); TL = FMA(TJ, TK, TI); TV = W[10]; TY = W[11]; { E T1U, T2w, TS, T2y; T1U = T1R - T1T; T2w = T1R + T1T; TS = TL + TR; T2y = TR - TL; { E T2B, TX, T2x, T1P; T2B = TV * TZ; TX = TV * TW; T2x = FNMS(KP500000000, T2w, T2v); T3a = T2v + T2w; T1P = FNMS(KP500000000, TS, TF); TT = TF + TS; T10 = FMA(TY, TZ, TX); T2Y = FNMS(KP866025403, T2y, T2x); T2z = FMA(KP866025403, T2y, T2x); T1V = FNMS(KP866025403, T1U, T1P); T2f = FMA(KP866025403, T1U, T1P); T2C = FNMS(TY, TW, T2B); } } } { E T18, T1b, T17, T1a, T20, T19, T11; T18 = ri[WS(rs, 1)]; T1b = ii[WS(rs, 1)]; T17 = W[0]; T1a = W[1]; T12 = ri[WS(rs, 11)]; T15 = ii[WS(rs, 11)]; T20 = T17 * T1b; T19 = T17 * T18; T11 = W[20]; T14 = W[21]; T21 = FNMS(T1a, T18, T20); T1c = FMA(T1a, T1b, T19); T1Y = T11 * T15; T13 = T11 * T12; } } } } { E T2G, T2h, T3J, T3I, T32, T30, T1H, T1W, T3P, T3O, T2b; { E T3f, T3b, T1Z, T16, T3p, TU; T3f = T39 + T3a; T3b = T39 - T3a; T1Z = FNMS(T14, T12, T1Y); T16 = FMA(T14, T15, T13); T3p = Tz - TT; TU = Tz + TT; { E T3g, T2U, T23, T3c, T3e, T3q, T3s, T1A, T34, T3r, T3n; { E T22, T1d, T2F, T2E, T36, T2D; T22 = T1Z - T21; T2D = T1Z + T21; T1d = T16 + T1c; T2F = T1c - T16; T2E = FNMS(KP500000000, T2D, T2C); T36 = T2C + T2D; { E T1e, T1X, T38, T1z, T3o; T1e = T10 + T1d; T1X = FNMS(KP500000000, T1d, T10); T38 = T36 - T37; T3g = T36 + T37; T2G = FMA(KP866025403, T2F, T2E); T2U = FNMS(KP866025403, T2F, T2E); T1z = T1e + T1y; T3o = T1e - T1y; T2h = FMA(KP866025403, T22, T1X); T23 = FNMS(KP866025403, T22, T1X); T3c = FNMS(KP618033988, T3b, T38); T3e = FMA(KP618033988, T38, T3b); T3q = FNMS(KP618033988, T3p, T3o); T3s = FMA(KP618033988, T3o, T3p); T1A = TU + T1z; T34 = TU - T1z; } } { E T2W, T33, T3m, T3h, T2Z, T3d, T35, T3l; T3J = T2U + T2V; T2W = T2U - T2V; ri[0] = Tf + T1A; T33 = FNMS(KP250000000, T1A, Tf); T3m = T3f - T3g; T3h = T3f + T3g; T2Z = T2X - T2Y; T3I = T2X + T2Y; T3d = FMA(KP559016994, T34, T33); T35 = FNMS(KP559016994, T34, T33); ii[0] = T3h + T3k; T3l = FNMS(KP250000000, T3h, T3k); ri[WS(rs, 3)] = FMA(KP951056516, T3c, T35); ri[WS(rs, 12)] = FNMS(KP951056516, T3c, T35); ri[WS(rs, 6)] = FMA(KP951056516, T3e, T3d); ri[WS(rs, 9)] = FNMS(KP951056516, T3e, T3d); T3r = FMA(KP559016994, T3m, T3l); T3n = FNMS(KP559016994, T3m, T3l); T32 = FMA(KP618033988, T2W, T2Z); T30 = FNMS(KP618033988, T2Z, T2W); } ii[WS(rs, 12)] = FMA(KP951056516, T3q, T3n); ii[WS(rs, 3)] = FNMS(KP951056516, T3q, T3n); ii[WS(rs, 9)] = FMA(KP951056516, T3s, T3r); ii[WS(rs, 6)] = FNMS(KP951056516, T3s, T3r); T2d = FMA(KP866025403, T1G, T1B); T1H = FNMS(KP866025403, T1G, T1B); T1W = T1O + T1V; T3P = T1O - T1V; T3O = T23 - T2a; T2b = T23 + T2a; } } { E T3H, T3v, T2S, T3Q, T3S, T2R, T2c; T3H = FNMS(KP866025403, T3u, T3t); T3v = FMA(KP866025403, T3u, T3t); T2c = T1W + T2b; T2S = T1W - T2b; T3Q = FNMS(KP618033988, T3P, T3O); T3S = FMA(KP618033988, T3O, T3P); ri[WS(rs, 5)] = T1H + T2c; T2R = FNMS(KP250000000, T2c, T1H); { E T2g, T2j, T3G, T3E, T2A, T2N, T3y, T3A, T3M, T3L, T3z, T3F, T3B; { E T3C, T3D, T31, T2T, T3K; T2g = T2e + T2f; T3C = T2e - T2f; T3D = T2h - T2i; T2j = T2h + T2i; T31 = FMA(KP559016994, T2S, T2R); T2T = FNMS(KP559016994, T2S, T2R); T3K = T3I + T3J; T3M = T3I - T3J; ri[WS(rs, 8)] = FMA(KP951056516, T30, T2T); ri[WS(rs, 2)] = FNMS(KP951056516, T30, T2T); ri[WS(rs, 11)] = FMA(KP951056516, T32, T31); ri[WS(rs, 14)] = FNMS(KP951056516, T32, T31); ii[WS(rs, 5)] = T3K + T3H; T3L = FNMS(KP250000000, T3K, T3H); T3G = FNMS(KP618033988, T3C, T3D); T3E = FMA(KP618033988, T3D, T3C); } { E T3N, T3R, T3w, T3x; T3N = FNMS(KP559016994, T3M, T3L); T3R = FMA(KP559016994, T3M, T3L); T3w = T2t + T2z; T2A = T2t - T2z; T2N = T2G - T2M; T3x = T2G + T2M; ii[WS(rs, 8)] = FNMS(KP951056516, T3Q, T3N); ii[WS(rs, 2)] = FMA(KP951056516, T3Q, T3N); ii[WS(rs, 14)] = FMA(KP951056516, T3S, T3R); ii[WS(rs, 11)] = FNMS(KP951056516, T3S, T3R); T3y = T3w + T3x; T3A = T3w - T3x; } ii[WS(rs, 10)] = T3y + T3v; T3z = FNMS(KP250000000, T3y, T3v); T2O = FMA(KP618033988, T2N, T2A); T2Q = FNMS(KP618033988, T2A, T2N); T3F = FNMS(KP559016994, T3A, T3z); T3B = FMA(KP559016994, T3A, T3z); ii[WS(rs, 4)] = FMA(KP951056516, T3E, T3B); ii[WS(rs, 1)] = FNMS(KP951056516, T3E, T3B); ii[WS(rs, 13)] = FNMS(KP951056516, T3G, T3F); ii[WS(rs, 7)] = FMA(KP951056516, T3G, T3F); T2m = T2g - T2j; T2k = T2g + T2j; } } } } ri[WS(rs, 10)] = T2d + T2k; T2l = FNMS(KP250000000, T2k, T2d); T2P = FNMS(KP559016994, T2m, T2l); T2n = FMA(KP559016994, T2m, T2l); ri[WS(rs, 1)] = FMA(KP951056516, T2O, T2n); ri[WS(rs, 4)] = FNMS(KP951056516, T2O, T2n); ri[WS(rs, 13)] = FMA(KP951056516, T2Q, T2P); ri[WS(rs, 7)] = FNMS(KP951056516, T2Q, T2P); } } } static const tw_instr twinstr[] = { {TW_FULL, 0, 15}, {TW_NEXT, 1, 0} }; static const ct_desc desc = { 15, "t1_15", twinstr, &GENUS, {72, 28, 112, 0}, 0, 0, 0 }; void X(codelet_t1_15) (planner *p) { X(kdft_dit_register) (p, t1_15, &desc); } #else /* HAVE_FMA */ /* Generated by: ../../../genfft/gen_twiddle.native -compact -variables 4 -pipeline-latency 4 -n 15 -name t1_15 -include t.h */ /* * This function contains 184 FP additions, 112 FP multiplications, * (or, 128 additions, 56 multiplications, 56 fused multiply/add), * 65 stack variables, 6 constants, and 60 memory accesses */ #include "t.h" static void t1_15(R *ri, R *ii, const R *W, stride rs, INT mb, INT me, INT ms) { DK(KP587785252, +0.587785252292473129168705954639072768597652438); DK(KP951056516, +0.951056516295153572116439333379382143405698634); DK(KP250000000, +0.250000000000000000000000000000000000000000000); DK(KP559016994, +0.559016994374947424102293417182819058860154590); DK(KP500000000, +0.500000000000000000000000000000000000000000000); DK(KP866025403, +0.866025403784438646763723170752936183471402627); { INT m; for (m = mb, W = W + (mb * 28); m < me; m = m + 1, ri = ri + ms, ii = ii + ms, W = W + 28, MAKE_VOLATILE_STRIDE(30, rs)) { E T1q, T34, Td, T1n, T2S, T35, T13, T1k, T1l, T2E, T2F, T2O, T1H, T1T, T2k; E T2t, T2f, T2s, T1M, T1U, Tu, TL, TM, T2H, T2I, T2N, T1w, T1Q, T29, T2w; E T24, T2v, T1B, T1R; { E T1, T2R, T6, T1o, Tb, T1p, Tc, T2Q; T1 = ri[0]; T2R = ii[0]; { E T3, T5, T2, T4; T3 = ri[WS(rs, 5)]; T5 = ii[WS(rs, 5)]; T2 = W[8]; T4 = W[9]; T6 = FMA(T2, T3, T4 * T5); T1o = FNMS(T4, T3, T2 * T5); } { E T8, Ta, T7, T9; T8 = ri[WS(rs, 10)]; Ta = ii[WS(rs, 10)]; T7 = W[18]; T9 = W[19]; Tb = FMA(T7, T8, T9 * Ta); T1p = FNMS(T9, T8, T7 * Ta); } T1q = KP866025403 * (T1o - T1p); T34 = KP866025403 * (Tb - T6); Tc = T6 + Tb; Td = T1 + Tc; T1n = FNMS(KP500000000, Tc, T1); T2Q = T1o + T1p; T2S = T2Q + T2R; T35 = FNMS(KP500000000, T2Q, T2R); } { E TR, T2c, T18, T2h, TW, T1E, T11, T1F, T12, T2d, T1d, T1J, T1i, T1K, T1j; E T2i; { E TO, TQ, TN, TP; TO = ri[WS(rs, 6)]; TQ = ii[WS(rs, 6)]; TN = W[10]; TP = W[11]; TR = FMA(TN, TO, TP * TQ); T2c = FNMS(TP, TO, TN * TQ); } { E T15, T17, T14, T16; T15 = ri[WS(rs, 9)]; T17 = ii[WS(rs, 9)]; T14 = W[16]; T16 = W[17]; T18 = FMA(T14, T15, T16 * T17); T2h = FNMS(T16, T15, T14 * T17); } { E TT, TV, TS, TU; TT = ri[WS(rs, 11)]; TV = ii[WS(rs, 11)]; TS = W[20]; TU = W[21]; TW = FMA(TS, TT, TU * TV); T1E = FNMS(TU, TT, TS * TV); } { E TY, T10, TX, TZ; TY = ri[WS(rs, 1)]; T10 = ii[WS(rs, 1)]; TX = W[0]; TZ = W[1]; T11 = FMA(TX, TY, TZ * T10); T1F = FNMS(TZ, TY, TX * T10); } T12 = TW + T11; T2d = T1E + T1F; { E T1a, T1c, T19, T1b; T1a = ri[WS(rs, 14)]; T1c = ii[WS(rs, 14)]; T19 = W[26]; T1b = W[27]; T1d = FMA(T19, T1a, T1b * T1c); T1J = FNMS(T1b, T1a, T19 * T1c); } { E T1f, T1h, T1e, T1g; T1f = ri[WS(rs, 4)]; T1h = ii[WS(rs, 4)]; T1e = W[6]; T1g = W[7]; T1i = FMA(T1e, T1f, T1g * T1h); T1K = FNMS(T1g, T1f, T1e * T1h); } T1j = T1d + T1i; T2i = T1J + T1K; { E T1D, T1G, T2g, T2j; T13 = TR + T12; T1k = T18 + T1j; T1l = T13 + T1k; T2E = T2c + T2d; T2F = T2h + T2i; T2O = T2E + T2F; T1D = FNMS(KP500000000, T12, TR); T1G = KP866025403 * (T1E - T1F); T1H = T1D - T1G; T1T = T1D + T1G; T2g = KP866025403 * (T1i - T1d); T2j = FNMS(KP500000000, T2i, T2h); T2k = T2g + T2j; T2t = T2j - T2g; { E T2b, T2e, T1I, T1L; T2b = KP866025403 * (T11 - TW); T2e = FNMS(KP500000000, T2d, T2c); T2f = T2b + T2e; T2s = T2e - T2b; T1I = FNMS(KP500000000, T1j, T18); T1L = KP866025403 * (T1J - T1K); T1M = T1I - T1L; T1U = T1I + T1L; } } } { E Ti, T21, Tz, T26, Tn, T1t, Ts, T1u, Tt, T22, TE, T1y, TJ, T1z, TK; E T27; { E Tf, Th, Te, Tg; Tf = ri[WS(rs, 3)]; Th = ii[WS(rs, 3)]; Te = W[4]; Tg = W[5]; Ti = FMA(Te, Tf, Tg * Th); T21 = FNMS(Tg, Tf, Te * Th); } { E Tw, Ty, Tv, Tx; Tw = ri[WS(rs, 12)]; Ty = ii[WS(rs, 12)]; Tv = W[22]; Tx = W[23]; Tz = FMA(Tv, Tw, Tx * Ty); T26 = FNMS(Tx, Tw, Tv * Ty); } { E Tk, Tm, Tj, Tl; Tk = ri[WS(rs, 8)]; Tm = ii[WS(rs, 8)]; Tj = W[14]; Tl = W[15]; Tn = FMA(Tj, Tk, Tl * Tm); T1t = FNMS(Tl, Tk, Tj * Tm); } { E Tp, Tr, To, Tq; Tp = ri[WS(rs, 13)]; Tr = ii[WS(rs, 13)]; To = W[24]; Tq = W[25]; Ts = FMA(To, Tp, Tq * Tr); T1u = FNMS(Tq, Tp, To * Tr); } Tt = Tn + Ts; T22 = T1t + T1u; { E TB, TD, TA, TC; TB = ri[WS(rs, 2)]; TD = ii[WS(rs, 2)]; TA = W[2]; TC = W[3]; TE = FMA(TA, TB, TC * TD); T1y = FNMS(TC, TB, TA * TD); } { E TG, TI, TF, TH; TG = ri[WS(rs, 7)]; TI = ii[WS(rs, 7)]; TF = W[12]; TH = W[13]; TJ = FMA(TF, TG, TH * TI); T1z = FNMS(TH, TG, TF * TI); } TK = TE + TJ; T27 = T1y + T1z; { E T1s, T1v, T25, T28; Tu = Ti + Tt; TL = Tz + TK; TM = Tu + TL; T2H = T21 + T22; T2I = T26 + T27; T2N = T2H + T2I; T1s = FNMS(KP500000000, Tt, Ti); T1v = KP866025403 * (T1t - T1u); T1w = T1s - T1v; T1Q = T1s + T1v; T25 = KP866025403 * (TJ - TE); T28 = FNMS(KP500000000, T27, T26); T29 = T25 + T28; T2w = T28 - T25; { E T20, T23, T1x, T1A; T20 = KP866025403 * (Ts - Tn); T23 = FNMS(KP500000000, T22, T21); T24 = T20 + T23; T2v = T23 - T20; T1x = FNMS(KP500000000, TK, Tz); T1A = KP866025403 * (T1y - T1z); T1B = T1x - T1A; T1R = T1x + T1A; } } } { E T2C, T1m, T2B, T2K, T2M, T2G, T2J, T2L, T2D; T2C = KP559016994 * (TM - T1l); T1m = TM + T1l; T2B = FNMS(KP250000000, T1m, Td); T2G = T2E - T2F; T2J = T2H - T2I; T2K = FNMS(KP587785252, T2J, KP951056516 * T2G); T2M = FMA(KP951056516, T2J, KP587785252 * T2G); ri[0] = Td + T1m; T2L = T2C + T2B; ri[WS(rs, 9)] = T2L - T2M; ri[WS(rs, 6)] = T2L + T2M; T2D = T2B - T2C; ri[WS(rs, 12)] = T2D - T2K; ri[WS(rs, 3)] = T2D + T2K; } { E T2U, T2P, T2T, T2Y, T30, T2W, T2X, T2Z, T2V; T2U = KP559016994 * (T2N - T2O); T2P = T2N + T2O; T2T = FNMS(KP250000000, T2P, T2S); T2W = T13 - T1k; T2X = Tu - TL; T2Y = FNMS(KP587785252, T2X, KP951056516 * T2W); T30 = FMA(KP951056516, T2X, KP587785252 * T2W); ii[0] = T2P + T2S; T2Z = T2U + T2T; ii[WS(rs, 6)] = T2Z - T30; ii[WS(rs, 9)] = T30 + T2Z; T2V = T2T - T2U; ii[WS(rs, 3)] = T2V - T2Y; ii[WS(rs, 12)] = T2Y + T2V; } { E T2y, T2A, T1r, T1O, T2p, T2q, T2z, T2r; { E T2u, T2x, T1C, T1N; T2u = T2s - T2t; T2x = T2v - T2w; T2y = FNMS(KP587785252, T2x, KP951056516 * T2u); T2A = FMA(KP951056516, T2x, KP587785252 * T2u); T1r = T1n - T1q; T1C = T1w + T1B; T1N = T1H + T1M; T1O = T1C + T1N; T2p = FNMS(KP250000000, T1O, T1r); T2q = KP559016994 * (T1C - T1N); } ri[WS(rs, 5)] = T1r + T1O; T2z = T2q + T2p; ri[WS(rs, 14)] = T2z - T2A; ri[WS(rs, 11)] = T2z + T2A; T2r = T2p - T2q; ri[WS(rs, 2)] = T2r - T2y; ri[WS(rs, 8)] = T2r + T2y; } { E T3h, T3q, T3i, T3l, T3m, T3n, T3p, T3o; { E T3f, T3g, T3j, T3k; T3f = T1H - T1M; T3g = T1w - T1B; T3h = FNMS(KP587785252, T3g, KP951056516 * T3f); T3q = FMA(KP951056516, T3g, KP587785252 * T3f); T3i = T35 - T34; T3j = T2v + T2w; T3k = T2s + T2t; T3l = T3j + T3k; T3m = FNMS(KP250000000, T3l, T3i); T3n = KP559016994 * (T3j - T3k); } ii[WS(rs, 5)] = T3l + T3i; T3p = T3n + T3m; ii[WS(rs, 11)] = T3p - T3q; ii[WS(rs, 14)] = T3q + T3p; T3o = T3m - T3n; ii[WS(rs, 2)] = T3h + T3o; ii[WS(rs, 8)] = T3o - T3h; } { E T3c, T3d, T36, T37, T33, T38, T3e, T39; { E T3a, T3b, T31, T32; T3a = T1Q - T1R; T3b = T1T - T1U; T3c = FMA(KP951056516, T3a, KP587785252 * T3b); T3d = FNMS(KP587785252, T3a, KP951056516 * T3b); T36 = T34 + T35; T31 = T24 + T29; T32 = T2f + T2k; T37 = T31 + T32; T33 = KP559016994 * (T31 - T32); T38 = FNMS(KP250000000, T37, T36); } ii[WS(rs, 10)] = T37 + T36; T3e = T38 - T33; ii[WS(rs, 7)] = T3d + T3e; ii[WS(rs, 13)] = T3e - T3d; T39 = T33 + T38; ii[WS(rs, 1)] = T39 - T3c; ii[WS(rs, 4)] = T3c + T39; } { E T2m, T2o, T1P, T1W, T1X, T1Y, T2n, T1Z; { E T2a, T2l, T1S, T1V; T2a = T24 - T29; T2l = T2f - T2k; T2m = FMA(KP951056516, T2a, KP587785252 * T2l); T2o = FNMS(KP587785252, T2a, KP951056516 * T2l); T1P = T1n + T1q; T1S = T1Q + T1R; T1V = T1T + T1U; T1W = T1S + T1V; T1X = KP559016994 * (T1S - T1V); T1Y = FNMS(KP250000000, T1W, T1P); } ri[WS(rs, 10)] = T1P + T1W; T2n = T1Y - T1X; ri[WS(rs, 7)] = T2n - T2o; ri[WS(rs, 13)] = T2n + T2o; T1Z = T1X + T1Y; ri[WS(rs, 4)] = T1Z - T2m; ri[WS(rs, 1)] = T1Z + T2m; } } } } static const tw_instr twinstr[] = { {TW_FULL, 0, 15}, {TW_NEXT, 1, 0} }; static const ct_desc desc = { 15, "t1_15", twinstr, &GENUS, {128, 56, 56, 0}, 0, 0, 0 }; void X(codelet_t1_15) (planner *p) { X(kdft_dit_register) (p, t1_15, &desc); } #endif /* HAVE_FMA */ fftw-3.3.3/dft/scalar/codelets/t1_16.c0000644000175000017500000004767012054410251014241 00000000000000/* * Copyright (c) 2003, 2007-11 Matteo Frigo * Copyright (c) 2003, 2007-11 Massachusetts Institute of Technology * * This program is free software; you can redistribute it and/or modify * it under the terms of the GNU General Public License as published by * the Free Software Foundation; either version 2 of the License, or * (at your option) any later version. * * This program is distributed in the hope that it will be useful, * but WITHOUT ANY WARRANTY; without even the implied warranty of * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the * GNU General Public License for more details. * * You should have received a copy of the GNU General Public License * along with this program; if not, write to the Free Software * Foundation, Inc., 51 Franklin Street, Fifth Floor, Boston, MA 02110-1301 USA * */ /* This file was automatically generated --- DO NOT EDIT */ /* Generated on Sun Nov 25 07:35:51 EST 2012 */ #include "codelet-dft.h" #ifdef HAVE_FMA /* Generated by: ../../../genfft/gen_twiddle.native -fma -reorder-insns -schedule-for-pipeline -compact -variables 4 -pipeline-latency 4 -n 16 -name t1_16 -include t.h */ /* * This function contains 174 FP additions, 100 FP multiplications, * (or, 104 additions, 30 multiplications, 70 fused multiply/add), * 97 stack variables, 3 constants, and 64 memory accesses */ #include "t.h" static void t1_16(R *ri, R *ii, const R *W, stride rs, INT mb, INT me, INT ms) { DK(KP923879532, +0.923879532511286756128183189396788286822416626); DK(KP414213562, +0.414213562373095048801688724209698078569671875); DK(KP707106781, +0.707106781186547524400844362104849039284835938); { INT m; for (m = mb, W = W + (mb * 30); m < me; m = m + 1, ri = ri + ms, ii = ii + ms, W = W + 30, MAKE_VOLATILE_STRIDE(32, rs)) { E T3G, T3F; { E T3z, T3o, T8, T1I, T2o, T35, T2r, T1s, T2w, T36, T2p, T1F, T3k, T1N, T3A; E Tl, T1T, T2V, T1U, Tz, T29, T30, T2c, T11, TB, TE, T2h, T31, T2a, T1e; E TC, T1X, TH, TK, TG, TD, TJ; { E Ta, Td, Tb, T1J, Tg, Tj, Tf, Tc, Ti; { E T1h, T1k, T1n, T2k, T1i, T1q, T1m, T1j, T1p; { E T1, T3n, T3, T6, T2, T5; T1 = ri[0]; T3n = ii[0]; T3 = ri[WS(rs, 8)]; T6 = ii[WS(rs, 8)]; T2 = W[14]; T5 = W[15]; { E T3l, T4, T1g, T3m, T7; T1h = ri[WS(rs, 15)]; T1k = ii[WS(rs, 15)]; T3l = T2 * T6; T4 = T2 * T3; T1g = W[28]; T1n = ri[WS(rs, 7)]; T3m = FNMS(T5, T3, T3l); T7 = FMA(T5, T6, T4); T2k = T1g * T1k; T1i = T1g * T1h; T3z = T3n - T3m; T3o = T3m + T3n; T8 = T1 + T7; T1I = T1 - T7; T1q = ii[WS(rs, 7)]; T1m = W[12]; } T1j = W[29]; T1p = W[13]; } { E T1u, T1x, T1v, T2s, T1A, T1D, T1z, T1w, T1C; { E T2l, T1l, T2n, T1r, T2m, T1o, T1t; T1u = ri[WS(rs, 3)]; T2m = T1m * T1q; T1o = T1m * T1n; T2l = FNMS(T1j, T1h, T2k); T1l = FMA(T1j, T1k, T1i); T2n = FNMS(T1p, T1n, T2m); T1r = FMA(T1p, T1q, T1o); T1x = ii[WS(rs, 3)]; T1t = W[4]; T2o = T2l - T2n; T35 = T2l + T2n; T2r = T1l - T1r; T1s = T1l + T1r; T1v = T1t * T1u; T2s = T1t * T1x; } T1A = ri[WS(rs, 11)]; T1D = ii[WS(rs, 11)]; T1z = W[20]; T1w = W[5]; T1C = W[21]; { E T2t, T1y, T2v, T1E, T2u, T1B, T9; Ta = ri[WS(rs, 4)]; T2u = T1z * T1D; T1B = T1z * T1A; T2t = FNMS(T1w, T1u, T2s); T1y = FMA(T1w, T1x, T1v); T2v = FNMS(T1C, T1A, T2u); T1E = FMA(T1C, T1D, T1B); Td = ii[WS(rs, 4)]; T9 = W[6]; T2w = T2t - T2v; T36 = T2t + T2v; T2p = T1y - T1E; T1F = T1y + T1E; Tb = T9 * Ta; T1J = T9 * Td; } Tg = ri[WS(rs, 12)]; Tj = ii[WS(rs, 12)]; Tf = W[22]; Tc = W[7]; Ti = W[23]; } } { E TQ, TT, TR, T25, TW, TZ, TV, TS, TY; { E To, Tr, Tp, T1P, Tu, Tx, Tt, Tq, Tw; { E T1K, Te, T1M, Tk, T1L, Th, Tn; To = ri[WS(rs, 2)]; T1L = Tf * Tj; Th = Tf * Tg; T1K = FNMS(Tc, Ta, T1J); Te = FMA(Tc, Td, Tb); T1M = FNMS(Ti, Tg, T1L); Tk = FMA(Ti, Tj, Th); Tr = ii[WS(rs, 2)]; Tn = W[2]; T3k = T1K + T1M; T1N = T1K - T1M; T3A = Te - Tk; Tl = Te + Tk; Tp = Tn * To; T1P = Tn * Tr; } Tu = ri[WS(rs, 10)]; Tx = ii[WS(rs, 10)]; Tt = W[18]; Tq = W[3]; Tw = W[19]; { E T1Q, Ts, T1S, Ty, T1R, Tv, TP; TQ = ri[WS(rs, 1)]; T1R = Tt * Tx; Tv = Tt * Tu; T1Q = FNMS(Tq, To, T1P); Ts = FMA(Tq, Tr, Tp); T1S = FNMS(Tw, Tu, T1R); Ty = FMA(Tw, Tx, Tv); TT = ii[WS(rs, 1)]; TP = W[0]; T1T = T1Q - T1S; T2V = T1Q + T1S; T1U = Ts - Ty; Tz = Ts + Ty; TR = TP * TQ; T25 = TP * TT; } TW = ri[WS(rs, 9)]; TZ = ii[WS(rs, 9)]; TV = W[16]; TS = W[1]; TY = W[17]; } { E T13, T16, T14, T2d, T19, T1c, T18, T15, T1b; { E T26, TU, T28, T10, T27, TX, T12; T13 = ri[WS(rs, 5)]; T27 = TV * TZ; TX = TV * TW; T26 = FNMS(TS, TQ, T25); TU = FMA(TS, TT, TR); T28 = FNMS(TY, TW, T27); T10 = FMA(TY, TZ, TX); T16 = ii[WS(rs, 5)]; T12 = W[8]; T29 = T26 - T28; T30 = T26 + T28; T2c = TU - T10; T11 = TU + T10; T14 = T12 * T13; T2d = T12 * T16; } T19 = ri[WS(rs, 13)]; T1c = ii[WS(rs, 13)]; T18 = W[24]; T15 = W[9]; T1b = W[25]; { E T2e, T17, T2g, T1d, T2f, T1a, TA; TB = ri[WS(rs, 14)]; T2f = T18 * T1c; T1a = T18 * T19; T2e = FNMS(T15, T13, T2d); T17 = FMA(T15, T16, T14); T2g = FNMS(T1b, T19, T2f); T1d = FMA(T1b, T1c, T1a); TE = ii[WS(rs, 14)]; TA = W[26]; T2h = T2e - T2g; T31 = T2e + T2g; T2a = T17 - T1d; T1e = T17 + T1d; TC = TA * TB; T1X = TA * TE; } TH = ri[WS(rs, 6)]; TK = ii[WS(rs, 6)]; TG = W[10]; TD = W[27]; TJ = W[11]; } } } { E T2U, T3u, T2Z, T21, T1W, T34, T2X, T3f, T32, T3t, T1H, T3q, T3e, TO, T3g; E T37, T3r, T3s, T3h, T3i; { E Tm, T1Y, TF, T20, TL, T3p, T1Z, TI; T2U = T8 - Tl; Tm = T8 + Tl; T1Z = TG * TK; TI = TG * TH; T1Y = FNMS(TD, TB, T1X); TF = FMA(TD, TE, TC); T20 = FNMS(TJ, TH, T1Z); TL = FMA(TJ, TK, TI); T3p = T3k + T3o; T3u = T3o - T3k; { E T1f, TM, T1G, T3j, T2W, TN; T2Z = T11 - T1e; T1f = T11 + T1e; T21 = T1Y - T20; T2W = T1Y + T20; T1W = TF - TL; TM = TF + TL; T1G = T1s + T1F; T34 = T1s - T1F; T2X = T2V - T2W; T3j = T2V + T2W; T3f = T30 + T31; T32 = T30 - T31; T3t = TM - Tz; TN = Tz + TM; T3r = T1G - T1f; T1H = T1f + T1G; T3s = T3p - T3j; T3q = T3j + T3p; T3e = Tm - TN; TO = Tm + TN; T3g = T35 + T36; T37 = T35 - T36; } } ii[WS(rs, 12)] = T3s - T3r; ii[WS(rs, 4)] = T3r + T3s; ri[0] = TO + T1H; ri[WS(rs, 8)] = TO - T1H; T3h = T3f - T3g; T3i = T3f + T3g; { E T3a, T2Y, T3x, T3v, T3b, T33; ii[0] = T3i + T3q; ii[WS(rs, 8)] = T3q - T3i; ri[WS(rs, 4)] = T3e + T3h; ri[WS(rs, 12)] = T3e - T3h; T3a = T2U - T2X; T2Y = T2U + T2X; T3x = T3u - T3t; T3v = T3t + T3u; T3b = T32 - T2Z; T33 = T2Z + T32; { E T2E, T1O, T3B, T3H, T2x, T2q, T3C, T23, T2S, T2O, T2K, T2J, T3I, T2H, T2B; E T2j; { E T2F, T1V, T22, T2G, T3c, T38; T2E = T1I + T1N; T1O = T1I - T1N; T3B = T3z - T3A; T3H = T3A + T3z; T3c = T34 + T37; T38 = T34 - T37; T2F = T1U + T1T; T1V = T1T - T1U; { E T3d, T3w, T3y, T39; T3d = T3b - T3c; T3w = T3b + T3c; T3y = T38 - T33; T39 = T33 + T38; ri[WS(rs, 6)] = FMA(KP707106781, T3d, T3a); ri[WS(rs, 14)] = FNMS(KP707106781, T3d, T3a); ii[WS(rs, 10)] = FNMS(KP707106781, T3w, T3v); ii[WS(rs, 2)] = FMA(KP707106781, T3w, T3v); ii[WS(rs, 14)] = FNMS(KP707106781, T3y, T3x); ii[WS(rs, 6)] = FMA(KP707106781, T3y, T3x); ri[WS(rs, 2)] = FMA(KP707106781, T39, T2Y); ri[WS(rs, 10)] = FNMS(KP707106781, T39, T2Y); T22 = T1W + T21; T2G = T1W - T21; } { E T2M, T2N, T2b, T2i; T2x = T2r - T2w; T2M = T2r + T2w; T2N = T2o - T2p; T2q = T2o + T2p; T3C = T1V + T22; T23 = T1V - T22; T2S = FMA(KP414213562, T2M, T2N); T2O = FNMS(KP414213562, T2N, T2M); T2K = T29 - T2a; T2b = T29 + T2a; T2i = T2c - T2h; T2J = T2c + T2h; T3I = T2G - T2F; T2H = T2F + T2G; T2B = FNMS(KP414213562, T2b, T2i); T2j = FMA(KP414213562, T2i, T2b); } } { E T2R, T2L, T3L, T3M; { E T2A, T24, T2C, T2y, T3J, T3K, T2D, T2z; T2A = FNMS(KP707106781, T23, T1O); T24 = FMA(KP707106781, T23, T1O); T2R = FNMS(KP414213562, T2J, T2K); T2L = FMA(KP414213562, T2K, T2J); T2C = FMA(KP414213562, T2q, T2x); T2y = FNMS(KP414213562, T2x, T2q); T3J = FMA(KP707106781, T3I, T3H); T3L = FNMS(KP707106781, T3I, T3H); T3K = T2C - T2B; T2D = T2B + T2C; T3M = T2j + T2y; T2z = T2j - T2y; ii[WS(rs, 11)] = FNMS(KP923879532, T3K, T3J); ii[WS(rs, 3)] = FMA(KP923879532, T3K, T3J); ri[WS(rs, 3)] = FMA(KP923879532, T2z, T24); ri[WS(rs, 11)] = FNMS(KP923879532, T2z, T24); ri[WS(rs, 15)] = FMA(KP923879532, T2D, T2A); ri[WS(rs, 7)] = FNMS(KP923879532, T2D, T2A); } { E T2Q, T3D, T3E, T2T, T2I, T2P; T2Q = FNMS(KP707106781, T2H, T2E); T2I = FMA(KP707106781, T2H, T2E); T2P = T2L + T2O; T3G = T2O - T2L; T3F = FNMS(KP707106781, T3C, T3B); T3D = FMA(KP707106781, T3C, T3B); ii[WS(rs, 15)] = FMA(KP923879532, T3M, T3L); ii[WS(rs, 7)] = FNMS(KP923879532, T3M, T3L); ri[WS(rs, 1)] = FMA(KP923879532, T2P, T2I); ri[WS(rs, 9)] = FNMS(KP923879532, T2P, T2I); T3E = T2R + T2S; T2T = T2R - T2S; ii[WS(rs, 9)] = FNMS(KP923879532, T3E, T3D); ii[WS(rs, 1)] = FMA(KP923879532, T3E, T3D); ri[WS(rs, 5)] = FMA(KP923879532, T2T, T2Q); ri[WS(rs, 13)] = FNMS(KP923879532, T2T, T2Q); } } } } } } ii[WS(rs, 13)] = FNMS(KP923879532, T3G, T3F); ii[WS(rs, 5)] = FMA(KP923879532, T3G, T3F); } } } static const tw_instr twinstr[] = { {TW_FULL, 0, 16}, {TW_NEXT, 1, 0} }; static const ct_desc desc = { 16, "t1_16", twinstr, &GENUS, {104, 30, 70, 0}, 0, 0, 0 }; void X(codelet_t1_16) (planner *p) { X(kdft_dit_register) (p, t1_16, &desc); } #else /* HAVE_FMA */ /* Generated by: ../../../genfft/gen_twiddle.native -compact -variables 4 -pipeline-latency 4 -n 16 -name t1_16 -include t.h */ /* * This function contains 174 FP additions, 84 FP multiplications, * (or, 136 additions, 46 multiplications, 38 fused multiply/add), * 52 stack variables, 3 constants, and 64 memory accesses */ #include "t.h" static void t1_16(R *ri, R *ii, const R *W, stride rs, INT mb, INT me, INT ms) { DK(KP382683432, +0.382683432365089771728459984030398866761344562); DK(KP923879532, +0.923879532511286756128183189396788286822416626); DK(KP707106781, +0.707106781186547524400844362104849039284835938); { INT m; for (m = mb, W = W + (mb * 30); m < me; m = m + 1, ri = ri + ms, ii = ii + ms, W = W + 30, MAKE_VOLATILE_STRIDE(32, rs)) { E T7, T37, T1t, T2U, Ti, T38, T1w, T2R, Tu, T2s, T1C, T2c, TF, T2t, T1H; E T2d, T1f, T1q, T2B, T2C, T2D, T2E, T1Z, T2j, T24, T2k, TS, T13, T2w, T2x; E T2y, T2z, T1O, T2g, T1T, T2h; { E T1, T2T, T6, T2S; T1 = ri[0]; T2T = ii[0]; { E T3, T5, T2, T4; T3 = ri[WS(rs, 8)]; T5 = ii[WS(rs, 8)]; T2 = W[14]; T4 = W[15]; T6 = FMA(T2, T3, T4 * T5); T2S = FNMS(T4, T3, T2 * T5); } T7 = T1 + T6; T37 = T2T - T2S; T1t = T1 - T6; T2U = T2S + T2T; } { E Tc, T1u, Th, T1v; { E T9, Tb, T8, Ta; T9 = ri[WS(rs, 4)]; Tb = ii[WS(rs, 4)]; T8 = W[6]; Ta = W[7]; Tc = FMA(T8, T9, Ta * Tb); T1u = FNMS(Ta, T9, T8 * Tb); } { E Te, Tg, Td, Tf; Te = ri[WS(rs, 12)]; Tg = ii[WS(rs, 12)]; Td = W[22]; Tf = W[23]; Th = FMA(Td, Te, Tf * Tg); T1v = FNMS(Tf, Te, Td * Tg); } Ti = Tc + Th; T38 = Tc - Th; T1w = T1u - T1v; T2R = T1u + T1v; } { E To, T1y, Tt, T1z, T1A, T1B; { E Tl, Tn, Tk, Tm; Tl = ri[WS(rs, 2)]; Tn = ii[WS(rs, 2)]; Tk = W[2]; Tm = W[3]; To = FMA(Tk, Tl, Tm * Tn); T1y = FNMS(Tm, Tl, Tk * Tn); } { E Tq, Ts, Tp, Tr; Tq = ri[WS(rs, 10)]; Ts = ii[WS(rs, 10)]; Tp = W[18]; Tr = W[19]; Tt = FMA(Tp, Tq, Tr * Ts); T1z = FNMS(Tr, Tq, Tp * Ts); } Tu = To + Tt; T2s = T1y + T1z; T1A = T1y - T1z; T1B = To - Tt; T1C = T1A - T1B; T2c = T1B + T1A; } { E Tz, T1E, TE, T1F, T1D, T1G; { E Tw, Ty, Tv, Tx; Tw = ri[WS(rs, 14)]; Ty = ii[WS(rs, 14)]; Tv = W[26]; Tx = W[27]; Tz = FMA(Tv, Tw, Tx * Ty); T1E = FNMS(Tx, Tw, Tv * Ty); } { E TB, TD, TA, TC; TB = ri[WS(rs, 6)]; TD = ii[WS(rs, 6)]; TA = W[10]; TC = W[11]; TE = FMA(TA, TB, TC * TD); T1F = FNMS(TC, TB, TA * TD); } TF = Tz + TE; T2t = T1E + T1F; T1D = Tz - TE; T1G = T1E - T1F; T1H = T1D + T1G; T2d = T1D - T1G; } { E T19, T20, T1p, T1X, T1e, T21, T1k, T1W; { E T16, T18, T15, T17; T16 = ri[WS(rs, 15)]; T18 = ii[WS(rs, 15)]; T15 = W[28]; T17 = W[29]; T19 = FMA(T15, T16, T17 * T18); T20 = FNMS(T17, T16, T15 * T18); } { E T1m, T1o, T1l, T1n; T1m = ri[WS(rs, 11)]; T1o = ii[WS(rs, 11)]; T1l = W[20]; T1n = W[21]; T1p = FMA(T1l, T1m, T1n * T1o); T1X = FNMS(T1n, T1m, T1l * T1o); } { E T1b, T1d, T1a, T1c; T1b = ri[WS(rs, 7)]; T1d = ii[WS(rs, 7)]; T1a = W[12]; T1c = W[13]; T1e = FMA(T1a, T1b, T1c * T1d); T21 = FNMS(T1c, T1b, T1a * T1d); } { E T1h, T1j, T1g, T1i; T1h = ri[WS(rs, 3)]; T1j = ii[WS(rs, 3)]; T1g = W[4]; T1i = W[5]; T1k = FMA(T1g, T1h, T1i * T1j); T1W = FNMS(T1i, T1h, T1g * T1j); } T1f = T19 + T1e; T1q = T1k + T1p; T2B = T1f - T1q; T2C = T20 + T21; T2D = T1W + T1X; T2E = T2C - T2D; { E T1V, T1Y, T22, T23; T1V = T19 - T1e; T1Y = T1W - T1X; T1Z = T1V - T1Y; T2j = T1V + T1Y; T22 = T20 - T21; T23 = T1k - T1p; T24 = T22 + T23; T2k = T22 - T23; } } { E TM, T1K, T12, T1R, TR, T1L, TX, T1Q; { E TJ, TL, TI, TK; TJ = ri[WS(rs, 1)]; TL = ii[WS(rs, 1)]; TI = W[0]; TK = W[1]; TM = FMA(TI, TJ, TK * TL); T1K = FNMS(TK, TJ, TI * TL); } { E TZ, T11, TY, T10; TZ = ri[WS(rs, 13)]; T11 = ii[WS(rs, 13)]; TY = W[24]; T10 = W[25]; T12 = FMA(TY, TZ, T10 * T11); T1R = FNMS(T10, TZ, TY * T11); } { E TO, TQ, TN, TP; TO = ri[WS(rs, 9)]; TQ = ii[WS(rs, 9)]; TN = W[16]; TP = W[17]; TR = FMA(TN, TO, TP * TQ); T1L = FNMS(TP, TO, TN * TQ); } { E TU, TW, TT, TV; TU = ri[WS(rs, 5)]; TW = ii[WS(rs, 5)]; TT = W[8]; TV = W[9]; TX = FMA(TT, TU, TV * TW); T1Q = FNMS(TV, TU, TT * TW); } TS = TM + TR; T13 = TX + T12; T2w = TS - T13; T2x = T1K + T1L; T2y = T1Q + T1R; T2z = T2x - T2y; { E T1M, T1N, T1P, T1S; T1M = T1K - T1L; T1N = TX - T12; T1O = T1M + T1N; T2g = T1M - T1N; T1P = TM - TR; T1S = T1Q - T1R; T1T = T1P - T1S; T2h = T1P + T1S; } } { E T1J, T27, T3g, T3i, T26, T3h, T2a, T3d; { E T1x, T1I, T3e, T3f; T1x = T1t - T1w; T1I = KP707106781 * (T1C - T1H); T1J = T1x + T1I; T27 = T1x - T1I; T3e = KP707106781 * (T2d - T2c); T3f = T38 + T37; T3g = T3e + T3f; T3i = T3f - T3e; } { E T1U, T25, T28, T29; T1U = FMA(KP923879532, T1O, KP382683432 * T1T); T25 = FNMS(KP923879532, T24, KP382683432 * T1Z); T26 = T1U + T25; T3h = T25 - T1U; T28 = FNMS(KP923879532, T1T, KP382683432 * T1O); T29 = FMA(KP382683432, T24, KP923879532 * T1Z); T2a = T28 - T29; T3d = T28 + T29; } ri[WS(rs, 11)] = T1J - T26; ii[WS(rs, 11)] = T3g - T3d; ri[WS(rs, 3)] = T1J + T26; ii[WS(rs, 3)] = T3d + T3g; ri[WS(rs, 15)] = T27 - T2a; ii[WS(rs, 15)] = T3i - T3h; ri[WS(rs, 7)] = T27 + T2a; ii[WS(rs, 7)] = T3h + T3i; } { E T2v, T2H, T32, T34, T2G, T33, T2K, T2Z; { E T2r, T2u, T30, T31; T2r = T7 - Ti; T2u = T2s - T2t; T2v = T2r + T2u; T2H = T2r - T2u; T30 = TF - Tu; T31 = T2U - T2R; T32 = T30 + T31; T34 = T31 - T30; } { E T2A, T2F, T2I, T2J; T2A = T2w + T2z; T2F = T2B - T2E; T2G = KP707106781 * (T2A + T2F); T33 = KP707106781 * (T2F - T2A); T2I = T2z - T2w; T2J = T2B + T2E; T2K = KP707106781 * (T2I - T2J); T2Z = KP707106781 * (T2I + T2J); } ri[WS(rs, 10)] = T2v - T2G; ii[WS(rs, 10)] = T32 - T2Z; ri[WS(rs, 2)] = T2v + T2G; ii[WS(rs, 2)] = T2Z + T32; ri[WS(rs, 14)] = T2H - T2K; ii[WS(rs, 14)] = T34 - T33; ri[WS(rs, 6)] = T2H + T2K; ii[WS(rs, 6)] = T33 + T34; } { E T2f, T2n, T3a, T3c, T2m, T3b, T2q, T35; { E T2b, T2e, T36, T39; T2b = T1t + T1w; T2e = KP707106781 * (T2c + T2d); T2f = T2b + T2e; T2n = T2b - T2e; T36 = KP707106781 * (T1C + T1H); T39 = T37 - T38; T3a = T36 + T39; T3c = T39 - T36; } { E T2i, T2l, T2o, T2p; T2i = FMA(KP382683432, T2g, KP923879532 * T2h); T2l = FNMS(KP382683432, T2k, KP923879532 * T2j); T2m = T2i + T2l; T3b = T2l - T2i; T2o = FNMS(KP382683432, T2h, KP923879532 * T2g); T2p = FMA(KP923879532, T2k, KP382683432 * T2j); T2q = T2o - T2p; T35 = T2o + T2p; } ri[WS(rs, 9)] = T2f - T2m; ii[WS(rs, 9)] = T3a - T35; ri[WS(rs, 1)] = T2f + T2m; ii[WS(rs, 1)] = T35 + T3a; ri[WS(rs, 13)] = T2n - T2q; ii[WS(rs, 13)] = T3c - T3b; ri[WS(rs, 5)] = T2n + T2q; ii[WS(rs, 5)] = T3b + T3c; } { E TH, T2L, T2W, T2Y, T1s, T2X, T2O, T2P; { E Tj, TG, T2Q, T2V; Tj = T7 + Ti; TG = Tu + TF; TH = Tj + TG; T2L = Tj - TG; T2Q = T2s + T2t; T2V = T2R + T2U; T2W = T2Q + T2V; T2Y = T2V - T2Q; } { E T14, T1r, T2M, T2N; T14 = TS + T13; T1r = T1f + T1q; T1s = T14 + T1r; T2X = T1r - T14; T2M = T2x + T2y; T2N = T2C + T2D; T2O = T2M - T2N; T2P = T2M + T2N; } ri[WS(rs, 8)] = TH - T1s; ii[WS(rs, 8)] = T2W - T2P; ri[0] = TH + T1s; ii[0] = T2P + T2W; ri[WS(rs, 12)] = T2L - T2O; ii[WS(rs, 12)] = T2Y - T2X; ri[WS(rs, 4)] = T2L + T2O; ii[WS(rs, 4)] = T2X + T2Y; } } } } static const tw_instr twinstr[] = { {TW_FULL, 0, 16}, {TW_NEXT, 1, 0} }; static const ct_desc desc = { 16, "t1_16", twinstr, &GENUS, {136, 46, 38, 0}, 0, 0, 0 }; void X(codelet_t1_16) (planner *p) { X(kdft_dit_register) (p, t1_16, &desc); } #endif /* HAVE_FMA */ fftw-3.3.3/dft/scalar/codelets/t1_32.c0000644000175000017500000013250712054410257014237 00000000000000/* * Copyright (c) 2003, 2007-11 Matteo Frigo * Copyright (c) 2003, 2007-11 Massachusetts Institute of Technology * * This program is free software; you can redistribute it and/or modify * it under the terms of the GNU General Public License as published by * the Free Software Foundation; either version 2 of the License, or * (at your option) any later version. * * This program is distributed in the hope that it will be useful, * but WITHOUT ANY WARRANTY; without even the implied warranty of * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the * GNU General Public License for more details. * * You should have received a copy of the GNU General Public License * along with this program; if not, write to the Free Software * Foundation, Inc., 51 Franklin Street, Fifth Floor, Boston, MA 02110-1301 USA * */ /* This file was automatically generated --- DO NOT EDIT */ /* Generated on Sun Nov 25 07:35:51 EST 2012 */ #include "codelet-dft.h" #ifdef HAVE_FMA /* Generated by: ../../../genfft/gen_twiddle.native -fma -reorder-insns -schedule-for-pipeline -compact -variables 4 -pipeline-latency 4 -n 32 -name t1_32 -include t.h */ /* * This function contains 434 FP additions, 260 FP multiplications, * (or, 236 additions, 62 multiplications, 198 fused multiply/add), * 135 stack variables, 7 constants, and 128 memory accesses */ #include "t.h" static void t1_32(R *ri, R *ii, const R *W, stride rs, INT mb, INT me, INT ms) { DK(KP980785280, +0.980785280403230449126182236134239036973933731); DK(KP831469612, +0.831469612302545237078788377617905756738560812); DK(KP668178637, +0.668178637919298919997757686523080761552472251); DK(KP198912367, +0.198912367379658006911597622644676228597850501); DK(KP923879532, +0.923879532511286756128183189396788286822416626); DK(KP414213562, +0.414213562373095048801688724209698078569671875); DK(KP707106781, +0.707106781186547524400844362104849039284835938); { INT m; for (m = mb, W = W + (mb * 62); m < me; m = m + 1, ri = ri + ms, ii = ii + ms, W = W + 62, MAKE_VOLATILE_STRIDE(64, rs)) { E T90, T8Z; { E T8x, T87, T8, T3w, T83, T3B, T8y, Tl, T6F, Tz, T3J, T5T, T6G, TM, T3Q; E T5U, T46, T5Y, T7D, T6L, T5X, T3Z, T6M, T1f, T7E, T6R, T60, T4e, T6O, T1G; E T61, T4l, T78, T7N, T54, T6f, T32, T7b, T6c, T5r, T6X, T7I, T4v, T68, T29; E T70, T65, T4S, T5s, T5b, T7O, T7e, T79, T3t, T5t, T5i, T4H, T2y, T4A, T71; E T2m, T4B, T4F, T2s; { E T44, T1d, T3X, T6J, T11, T40, T42, T17, T5h, T5c; { E Ta, Td, Tg, T3x, Tb, Tj, Tf, Tc, Ti; { E T1, T86, T3, T6, T2, T5; T1 = ri[0]; T86 = ii[0]; T3 = ri[WS(rs, 16)]; T6 = ii[WS(rs, 16)]; T2 = W[30]; T5 = W[31]; { E T84, T4, T9, T85, T7; Ta = ri[WS(rs, 8)]; Td = ii[WS(rs, 8)]; T84 = T2 * T6; T4 = T2 * T3; T9 = W[14]; Tg = ri[WS(rs, 24)]; T85 = FNMS(T5, T3, T84); T7 = FMA(T5, T6, T4); T3x = T9 * Td; Tb = T9 * Ta; T8x = T86 - T85; T87 = T85 + T86; T8 = T1 + T7; T3w = T1 - T7; Tj = ii[WS(rs, 24)]; Tf = W[46]; } Tc = W[15]; Ti = W[47]; } { E Tu, Tx, T3F, Ts, Tw, T3G, Tv; { E To, Tr, Tp, T3E, Tq, Tt; { E T3y, Te, T3A, Tk, T3z, Th, Tn; To = ri[WS(rs, 4)]; T3z = Tf * Tj; Th = Tf * Tg; T3y = FNMS(Tc, Ta, T3x); Te = FMA(Tc, Td, Tb); T3A = FNMS(Ti, Tg, T3z); Tk = FMA(Ti, Tj, Th); Tr = ii[WS(rs, 4)]; Tn = W[6]; T83 = T3y + T3A; T3B = T3y - T3A; T8y = Te - Tk; Tl = Te + Tk; Tp = Tn * To; T3E = Tn * Tr; } Tq = W[7]; Tu = ri[WS(rs, 20)]; Tx = ii[WS(rs, 20)]; Tt = W[38]; T3F = FNMS(Tq, To, T3E); Ts = FMA(Tq, Tr, Tp); Tw = W[39]; T3G = Tt * Tx; Tv = Tt * Tu; } { E T3M, TF, TH, TK, TG, TJ, TE, TD, TC; { E TB, T3H, Ty, TA, T3I, T3D, T3L; TB = ri[WS(rs, 28)]; TE = ii[WS(rs, 28)]; T3H = FNMS(Tw, Tu, T3G); Ty = FMA(Tw, Tx, Tv); TA = W[54]; TD = W[55]; T6F = T3F + T3H; T3I = T3F - T3H; Tz = Ts + Ty; T3D = Ts - Ty; T3L = TA * TE; TC = TA * TB; T3J = T3D + T3I; T5T = T3I - T3D; T3M = FNMS(TD, TB, T3L); } TF = FMA(TD, TE, TC); TH = ri[WS(rs, 12)]; TK = ii[WS(rs, 12)]; TG = W[22]; TJ = W[23]; { E TU, T3U, T13, T16, T3W, T10, T12, T15, T41, T14; { E T19, T1c, T18, T1b, T3P, T3K; { E TQ, TT, T3N, TI, TP, TS; TQ = ri[WS(rs, 2)]; TT = ii[WS(rs, 2)]; T3N = TG * TK; TI = TG * TH; TP = W[2]; TS = W[3]; { E T3O, TL, T3T, TR; T3O = FNMS(TJ, TH, T3N); TL = FMA(TJ, TK, TI); T3T = TP * TT; TR = TP * TQ; T6G = T3M + T3O; T3P = T3M - T3O; TM = TF + TL; T3K = TF - TL; TU = FMA(TS, TT, TR); T3U = FNMS(TS, TQ, T3T); } } T3Q = T3K - T3P; T5U = T3K + T3P; T19 = ri[WS(rs, 26)]; T1c = ii[WS(rs, 26)]; T18 = W[50]; T1b = W[51]; { E TW, TZ, TY, T3V, TX, T43, T1a, TV; TW = ri[WS(rs, 18)]; TZ = ii[WS(rs, 18)]; T43 = T18 * T1c; T1a = T18 * T19; TV = W[34]; TY = W[35]; T44 = FNMS(T1b, T19, T43); T1d = FMA(T1b, T1c, T1a); T3V = TV * TZ; TX = TV * TW; T13 = ri[WS(rs, 10)]; T16 = ii[WS(rs, 10)]; T3W = FNMS(TY, TW, T3V); T10 = FMA(TY, TZ, TX); T12 = W[18]; T15 = W[19]; } } T3X = T3U - T3W; T6J = T3U + T3W; T11 = TU + T10; T40 = TU - T10; T41 = T12 * T16; T14 = T12 * T13; T42 = FNMS(T15, T13, T41); T17 = FMA(T15, T16, T14); } } } } { E T49, T1l, T4j, T1E, T1u, T1x, T1w, T4b, T1r, T4g, T1v; { E T1A, T1D, T1C, T4i, T1B; { E T1h, T1k, T1g, T1j, T48, T1i, T1z; T1h = ri[WS(rs, 30)]; T1k = ii[WS(rs, 30)]; { E T6K, T45, T1e, T3Y; T6K = T42 + T44; T45 = T42 - T44; T1e = T17 + T1d; T3Y = T17 - T1d; T46 = T40 + T45; T5Y = T40 - T45; T7D = T6J + T6K; T6L = T6J - T6K; T5X = T3X + T3Y; T3Z = T3X - T3Y; T6M = T11 - T1e; T1f = T11 + T1e; T1g = W[58]; } T1j = W[59]; T1A = ri[WS(rs, 22)]; T1D = ii[WS(rs, 22)]; T48 = T1g * T1k; T1i = T1g * T1h; T1z = W[42]; T1C = W[43]; T49 = FNMS(T1j, T1h, T48); T1l = FMA(T1j, T1k, T1i); T4i = T1z * T1D; T1B = T1z * T1A; } { E T1n, T1q, T1m, T1p, T4a, T1o, T1t; T1n = ri[WS(rs, 14)]; T1q = ii[WS(rs, 14)]; T4j = FNMS(T1C, T1A, T4i); T1E = FMA(T1C, T1D, T1B); T1m = W[26]; T1p = W[27]; T1u = ri[WS(rs, 6)]; T1x = ii[WS(rs, 6)]; T4a = T1m * T1q; T1o = T1m * T1n; T1t = W[10]; T1w = W[11]; T4b = FNMS(T1p, T1n, T4a); T1r = FMA(T1p, T1q, T1o); T4g = T1t * T1x; T1v = T1t * T1u; } } { E T4c, T6P, T1s, T4f, T4h, T1y; T4c = T49 - T4b; T6P = T49 + T4b; T1s = T1l + T1r; T4f = T1l - T1r; T4h = FNMS(T1w, T1u, T4g); T1y = FMA(T1w, T1x, T1v); { E T4k, T6Q, T4d, T1F; T4k = T4h - T4j; T6Q = T4h + T4j; T4d = T1y - T1E; T1F = T1y + T1E; T7E = T6P + T6Q; T6R = T6P - T6Q; T60 = T4c + T4d; T4e = T4c - T4d; T6O = T1s - T1F; T1G = T1s + T1F; T61 = T4f - T4k; T4l = T4f + T4k; } } } { E T4Z, T2H, T5p, T30, T2Q, T2T, T2S, T51, T2N, T5m, T2R; { E T2W, T2Z, T2Y, T5o, T2X; { E T2D, T2G, T2C, T2F, T4Y, T2E, T2V; T2D = ri[WS(rs, 31)]; T2G = ii[WS(rs, 31)]; T2C = W[60]; T2F = W[61]; T2W = ri[WS(rs, 23)]; T2Z = ii[WS(rs, 23)]; T4Y = T2C * T2G; T2E = T2C * T2D; T2V = W[44]; T2Y = W[45]; T4Z = FNMS(T2F, T2D, T4Y); T2H = FMA(T2F, T2G, T2E); T5o = T2V * T2Z; T2X = T2V * T2W; } { E T2J, T2M, T2I, T2L, T50, T2K, T2P; T2J = ri[WS(rs, 15)]; T2M = ii[WS(rs, 15)]; T5p = FNMS(T2Y, T2W, T5o); T30 = FMA(T2Y, T2Z, T2X); T2I = W[28]; T2L = W[29]; T2Q = ri[WS(rs, 7)]; T2T = ii[WS(rs, 7)]; T50 = T2I * T2M; T2K = T2I * T2J; T2P = W[12]; T2S = W[13]; T51 = FNMS(T2L, T2J, T50); T2N = FMA(T2L, T2M, T2K); T5m = T2P * T2T; T2R = T2P * T2Q; } } { E T52, T76, T2O, T5l, T5n, T2U; T52 = T4Z - T51; T76 = T4Z + T51; T2O = T2H + T2N; T5l = T2H - T2N; T5n = FNMS(T2S, T2Q, T5m); T2U = FMA(T2S, T2T, T2R); { E T5q, T77, T53, T31; T5q = T5n - T5p; T77 = T5n + T5p; T53 = T2U - T30; T31 = T2U + T30; T78 = T76 - T77; T7N = T76 + T77; T54 = T52 - T53; T6f = T52 + T53; T32 = T2O + T31; T7b = T2O - T31; T6c = T5l - T5q; T5r = T5l + T5q; } } } { E T4q, T1O, T4Q, T27, T1X, T20, T1Z, T4s, T1U, T4N, T1Y; { E T23, T26, T25, T4P, T24; { E T1K, T1N, T1J, T1M, T4p, T1L, T22; T1K = ri[WS(rs, 1)]; T1N = ii[WS(rs, 1)]; T1J = W[0]; T1M = W[1]; T23 = ri[WS(rs, 25)]; T26 = ii[WS(rs, 25)]; T4p = T1J * T1N; T1L = T1J * T1K; T22 = W[48]; T25 = W[49]; T4q = FNMS(T1M, T1K, T4p); T1O = FMA(T1M, T1N, T1L); T4P = T22 * T26; T24 = T22 * T23; } { E T1Q, T1T, T1P, T1S, T4r, T1R, T1W; T1Q = ri[WS(rs, 17)]; T1T = ii[WS(rs, 17)]; T4Q = FNMS(T25, T23, T4P); T27 = FMA(T25, T26, T24); T1P = W[32]; T1S = W[33]; T1X = ri[WS(rs, 9)]; T20 = ii[WS(rs, 9)]; T4r = T1P * T1T; T1R = T1P * T1Q; T1W = W[16]; T1Z = W[17]; T4s = FNMS(T1S, T1Q, T4r); T1U = FMA(T1S, T1T, T1R); T4N = T1W * T20; T1Y = T1W * T1X; } } { E T4t, T6V, T1V, T4M, T4O, T21; T4t = T4q - T4s; T6V = T4q + T4s; T1V = T1O + T1U; T4M = T1O - T1U; T4O = FNMS(T1Z, T1X, T4N); T21 = FMA(T1Z, T20, T1Y); { E T4R, T6W, T4u, T28; T4R = T4O - T4Q; T6W = T4O + T4Q; T4u = T21 - T27; T28 = T21 + T27; T6X = T6V - T6W; T7I = T6V + T6W; T4v = T4t - T4u; T68 = T4t + T4u; T29 = T1V + T28; T70 = T1V - T28; T65 = T4M - T4R; T4S = T4M + T4R; } } } { E T56, T38, T5g, T3r, T3h, T3k, T3j, T58, T3e, T5d, T3i; { E T3n, T3q, T3p, T5f, T3o; { E T34, T37, T33, T36, T55, T35, T3m; T34 = ri[WS(rs, 3)]; T37 = ii[WS(rs, 3)]; T33 = W[4]; T36 = W[5]; T3n = ri[WS(rs, 11)]; T3q = ii[WS(rs, 11)]; T55 = T33 * T37; T35 = T33 * T34; T3m = W[20]; T3p = W[21]; T56 = FNMS(T36, T34, T55); T38 = FMA(T36, T37, T35); T5f = T3m * T3q; T3o = T3m * T3n; } { E T3a, T3d, T39, T3c, T57, T3b, T3g; T3a = ri[WS(rs, 19)]; T3d = ii[WS(rs, 19)]; T5g = FNMS(T3p, T3n, T5f); T3r = FMA(T3p, T3q, T3o); T39 = W[36]; T3c = W[37]; T3h = ri[WS(rs, 27)]; T3k = ii[WS(rs, 27)]; T57 = T39 * T3d; T3b = T39 * T3a; T3g = W[52]; T3j = W[53]; T58 = FNMS(T3c, T3a, T57); T3e = FMA(T3c, T3d, T3b); T5d = T3g * T3k; T3i = T3g * T3h; } } { E T59, T7c, T3f, T5a, T5e, T3l, T7d, T3s; T59 = T56 - T58; T7c = T56 + T58; T3f = T38 + T3e; T5a = T38 - T3e; T5e = FNMS(T3j, T3h, T5d); T3l = FMA(T3j, T3k, T3i); T5h = T5e - T5g; T7d = T5e + T5g; T3s = T3l + T3r; T5c = T3l - T3r; T5s = T5a + T59; T5b = T59 - T5a; T7O = T7c + T7d; T7e = T7c - T7d; T79 = T3s - T3f; T3t = T3f + T3s; } } { E T4x, T2f, T2o, T2r, T4z, T2l, T2n, T2q, T4E, T2p; { E T2u, T2x, T2t, T2w; { E T2b, T2e, T2d, T4w, T2c, T2a; T2b = ri[WS(rs, 5)]; T2e = ii[WS(rs, 5)]; T2a = W[8]; T5t = T5c - T5h; T5i = T5c + T5h; T2d = W[9]; T4w = T2a * T2e; T2c = T2a * T2b; T2u = ri[WS(rs, 13)]; T2x = ii[WS(rs, 13)]; T4x = FNMS(T2d, T2b, T4w); T2f = FMA(T2d, T2e, T2c); T2t = W[24]; T2w = W[25]; } { E T2h, T2k, T2j, T4y, T2i, T4G, T2v, T2g; T2h = ri[WS(rs, 21)]; T2k = ii[WS(rs, 21)]; T4G = T2t * T2x; T2v = T2t * T2u; T2g = W[40]; T2j = W[41]; T4H = FNMS(T2w, T2u, T4G); T2y = FMA(T2w, T2x, T2v); T4y = T2g * T2k; T2i = T2g * T2h; T2o = ri[WS(rs, 29)]; T2r = ii[WS(rs, 29)]; T4z = FNMS(T2j, T2h, T4y); T2l = FMA(T2j, T2k, T2i); T2n = W[56]; T2q = W[57]; } } T4A = T4x - T4z; T71 = T4x + T4z; T2m = T2f + T2l; T4B = T2f - T2l; T4E = T2n * T2r; T2p = T2n * T2o; T4F = FNMS(T2q, T2o, T4E); T2s = FMA(T2q, T2r, T2p); } } { E T4T, T4C, T4J, T4U, T7y, T8q, T8p, T7B; { E T6E, T8j, T73, T6Y, T6H, T8k, T8i, T8h; { E T7C, TO, T80, T7Z, T8e, T89, T8d, T1H, T8b, T3v, T7T, T7L, T7U, T7Q, T2A; E T7K, T7P, T7W, T1I; { E T7X, T7Y, T7J, T82, T88; { E Tm, T4I, T72, T4D, T2z, TN; T6E = T8 - Tl; Tm = T8 + Tl; T4T = T4B + T4A; T4C = T4A - T4B; T4I = T4F - T4H; T72 = T4F + T4H; T4D = T2s - T2y; T2z = T2s + T2y; TN = Tz + TM; T8j = TM - Tz; T73 = T71 - T72; T7J = T71 + T72; T4J = T4D + T4I; T4U = T4D - T4I; T2A = T2m + T2z; T6Y = T2z - T2m; T7C = Tm - TN; TO = Tm + TN; } T7K = T7I - T7J; T7X = T7I + T7J; T7Y = T7N + T7O; T7P = T7N - T7O; T6H = T6F - T6G; T82 = T6F + T6G; T88 = T83 + T87; T8k = T87 - T83; T80 = T7X + T7Y; T7Z = T7X - T7Y; T8e = T88 - T82; T89 = T82 + T88; } { E T7H, T7M, T2B, T3u; T7H = T29 - T2A; T2B = T29 + T2A; T3u = T32 + T3t; T7M = T32 - T3t; T8d = T1G - T1f; T1H = T1f + T1G; T8b = T3u - T2B; T3v = T2B + T3u; T7T = T7K - T7H; T7L = T7H + T7K; T7U = T7M + T7P; T7Q = T7M - T7P; } T7W = TO - T1H; T1I = TO + T1H; { E T7S, T8f, T8g, T7V; { E T7R, T8c, T8a, T7G, T81, T7F; T8i = T7Q - T7L; T7R = T7L + T7Q; T81 = T7D + T7E; T7F = T7D - T7E; ri[0] = T1I + T3v; ri[WS(rs, 16)] = T1I - T3v; ri[WS(rs, 8)] = T7W + T7Z; ri[WS(rs, 24)] = T7W - T7Z; T8c = T89 - T81; T8a = T81 + T89; T7G = T7C + T7F; T7S = T7C - T7F; T8h = T8e - T8d; T8f = T8d + T8e; ii[WS(rs, 24)] = T8c - T8b; ii[WS(rs, 8)] = T8b + T8c; ii[WS(rs, 16)] = T8a - T80; ii[0] = T80 + T8a; ri[WS(rs, 4)] = FMA(KP707106781, T7R, T7G); ri[WS(rs, 20)] = FNMS(KP707106781, T7R, T7G); T8g = T7T + T7U; T7V = T7T - T7U; } ii[WS(rs, 20)] = FNMS(KP707106781, T8g, T8f); ii[WS(rs, 4)] = FMA(KP707106781, T8g, T8f); ri[WS(rs, 12)] = FMA(KP707106781, T7V, T7S); ri[WS(rs, 28)] = FNMS(KP707106781, T7V, T7S); } } { E T7f, T7m, T6I, T7a, T7A, T7w, T8r, T8l, T8m, T6T, T7j, T75, T8s, T7p, T7z; E T7t; { E T7n, T6N, T6S, T7o, T7u, T7v; T7f = T7b - T7e; T7u = T7b + T7e; ii[WS(rs, 28)] = FNMS(KP707106781, T8i, T8h); ii[WS(rs, 12)] = FMA(KP707106781, T8i, T8h); T7m = T6E + T6H; T6I = T6E - T6H; T7v = T78 + T79; T7a = T78 - T79; T7n = T6M + T6L; T6N = T6L - T6M; T7A = FMA(KP414213562, T7u, T7v); T7w = FNMS(KP414213562, T7v, T7u); T8r = T8k - T8j; T8l = T8j + T8k; T6S = T6O + T6R; T7o = T6O - T6R; { E T7s, T7r, T6Z, T74; T7s = T6X + T6Y; T6Z = T6X - T6Y; T74 = T70 - T73; T7r = T70 + T73; T8m = T6N + T6S; T6T = T6N - T6S; T7j = FNMS(KP414213562, T6Z, T74); T75 = FMA(KP414213562, T74, T6Z); T8s = T7o - T7n; T7p = T7n + T7o; T7z = FNMS(KP414213562, T7r, T7s); T7t = FMA(KP414213562, T7s, T7r); } } { E T7i, T6U, T8t, T8v, T7k, T7g; T7i = FNMS(KP707106781, T6T, T6I); T6U = FMA(KP707106781, T6T, T6I); T8t = FMA(KP707106781, T8s, T8r); T8v = FNMS(KP707106781, T8s, T8r); T7k = FMA(KP414213562, T7a, T7f); T7g = FNMS(KP414213562, T7f, T7a); { E T7q, T7x, T8n, T8o; T7y = FNMS(KP707106781, T7p, T7m); T7q = FMA(KP707106781, T7p, T7m); { E T7l, T8u, T8w, T7h; T7l = T7j + T7k; T8u = T7k - T7j; T8w = T75 + T7g; T7h = T75 - T7g; ri[WS(rs, 30)] = FMA(KP923879532, T7l, T7i); ri[WS(rs, 14)] = FNMS(KP923879532, T7l, T7i); ii[WS(rs, 22)] = FNMS(KP923879532, T8u, T8t); ii[WS(rs, 6)] = FMA(KP923879532, T8u, T8t); ii[WS(rs, 30)] = FMA(KP923879532, T8w, T8v); ii[WS(rs, 14)] = FNMS(KP923879532, T8w, T8v); ri[WS(rs, 6)] = FMA(KP923879532, T7h, T6U); ri[WS(rs, 22)] = FNMS(KP923879532, T7h, T6U); T7x = T7t + T7w; T8q = T7w - T7t; } T8p = FNMS(KP707106781, T8m, T8l); T8n = FMA(KP707106781, T8m, T8l); T8o = T7z + T7A; T7B = T7z - T7A; ri[WS(rs, 2)] = FMA(KP923879532, T7x, T7q); ri[WS(rs, 18)] = FNMS(KP923879532, T7x, T7q); ii[WS(rs, 18)] = FNMS(KP923879532, T8o, T8n); ii[WS(rs, 2)] = FMA(KP923879532, T8o, T8n); } } } } { E T5S, T8O, T8N, T5V, T6d, T6g, T66, T69, T8G, T8F; { E T5C, T3S, T8C, T4n, T8H, T8B, T8I, T5F, T5k, T5L, T5u, T4K, T4V; { E T5D, T5E, T8z, T8A, T5j; { E T3C, T3R, T47, T4m; T5S = T3w - T3B; T3C = T3w + T3B; ri[WS(rs, 10)] = FMA(KP923879532, T7B, T7y); ri[WS(rs, 26)] = FNMS(KP923879532, T7B, T7y); ii[WS(rs, 26)] = FNMS(KP923879532, T8q, T8p); ii[WS(rs, 10)] = FMA(KP923879532, T8q, T8p); T3R = T3J + T3Q; T8O = T3Q - T3J; T5D = FMA(KP414213562, T3Z, T46); T47 = FNMS(KP414213562, T46, T3Z); T4m = FMA(KP414213562, T4l, T4e); T5E = FNMS(KP414213562, T4e, T4l); T8N = T8y + T8x; T8z = T8x - T8y; T5C = FMA(KP707106781, T3R, T3C); T3S = FNMS(KP707106781, T3R, T3C); T8C = T47 + T4m; T4n = T47 - T4m; T8A = T5T + T5U; T5V = T5T - T5U; } T6d = T5i - T5b; T5j = T5b + T5i; T8H = FNMS(KP707106781, T8A, T8z); T8B = FMA(KP707106781, T8A, T8z); T8I = T5E - T5D; T5F = T5D + T5E; T5k = FNMS(KP707106781, T5j, T54); T5L = FMA(KP707106781, T5j, T54); T5u = T5s + T5t; T6g = T5s - T5t; T66 = T4J - T4C; T4K = T4C + T4J; T4V = T4T + T4U; T69 = T4T - T4U; } { E T5M, T5Q, T5J, T5P, T8L, T8M; { E T5y, T4o, T5A, T5w, T5z, T4X, T8J, T5K, T5v, T8K, T5B, T5x; T5y = FNMS(KP923879532, T4n, T3S); T4o = FMA(KP923879532, T4n, T3S); T5K = FMA(KP707106781, T5u, T5r); T5v = FNMS(KP707106781, T5u, T5r); { E T5I, T4L, T5H, T4W; T5I = FMA(KP707106781, T4K, T4v); T4L = FNMS(KP707106781, T4K, T4v); T5H = FMA(KP707106781, T4V, T4S); T4W = FNMS(KP707106781, T4V, T4S); T5M = FNMS(KP198912367, T5L, T5K); T5Q = FMA(KP198912367, T5K, T5L); T5A = FMA(KP668178637, T5k, T5v); T5w = FNMS(KP668178637, T5v, T5k); T5J = FMA(KP198912367, T5I, T5H); T5P = FNMS(KP198912367, T5H, T5I); T5z = FNMS(KP668178637, T4L, T4W); T4X = FMA(KP668178637, T4W, T4L); } T8J = FMA(KP923879532, T8I, T8H); T8L = FNMS(KP923879532, T8I, T8H); T8K = T5A - T5z; T5B = T5z + T5A; T8M = T4X + T5w; T5x = T4X - T5w; ii[WS(rs, 21)] = FNMS(KP831469612, T8K, T8J); ii[WS(rs, 5)] = FMA(KP831469612, T8K, T8J); ri[WS(rs, 5)] = FMA(KP831469612, T5x, T4o); ri[WS(rs, 21)] = FNMS(KP831469612, T5x, T4o); ri[WS(rs, 29)] = FMA(KP831469612, T5B, T5y); ri[WS(rs, 13)] = FNMS(KP831469612, T5B, T5y); } { E T5O, T8D, T8E, T5R, T5G, T5N; T5O = FNMS(KP923879532, T5F, T5C); T5G = FMA(KP923879532, T5F, T5C); T5N = T5J + T5M; T8G = T5M - T5J; T8F = FNMS(KP923879532, T8C, T8B); T8D = FMA(KP923879532, T8C, T8B); ii[WS(rs, 29)] = FMA(KP831469612, T8M, T8L); ii[WS(rs, 13)] = FNMS(KP831469612, T8M, T8L); ri[WS(rs, 1)] = FMA(KP980785280, T5N, T5G); ri[WS(rs, 17)] = FNMS(KP980785280, T5N, T5G); T8E = T5P + T5Q; T5R = T5P - T5Q; ii[WS(rs, 17)] = FNMS(KP980785280, T8E, T8D); ii[WS(rs, 1)] = FMA(KP980785280, T8E, T8D); ri[WS(rs, 9)] = FMA(KP980785280, T5R, T5O); ri[WS(rs, 25)] = FNMS(KP980785280, T5R, T5O); } } } { E T6o, T5W, T8W, T63, T8V, T8P, T8Q, T6r, T67, T6u, T6y, T6C, T6m, T6i; { E T6p, T5Z, T62, T6q; T6p = FNMS(KP414213562, T5X, T5Y); T5Z = FMA(KP414213562, T5Y, T5X); ii[WS(rs, 25)] = FNMS(KP980785280, T8G, T8F); ii[WS(rs, 9)] = FMA(KP980785280, T8G, T8F); T6o = FNMS(KP707106781, T5V, T5S); T5W = FMA(KP707106781, T5V, T5S); T62 = FNMS(KP414213562, T61, T60); T6q = FMA(KP414213562, T60, T61); T8W = T5Z + T62; T63 = T5Z - T62; T8V = FNMS(KP707106781, T8O, T8N); T8P = FMA(KP707106781, T8O, T8N); { E T6x, T6e, T6w, T6h; T8Q = T6q - T6p; T6r = T6p + T6q; T6x = FMA(KP707106781, T6d, T6c); T6e = FNMS(KP707106781, T6d, T6c); T6w = FMA(KP707106781, T6g, T6f); T6h = FNMS(KP707106781, T6g, T6f); T67 = FNMS(KP707106781, T66, T65); T6u = FMA(KP707106781, T66, T65); T6y = FNMS(KP198912367, T6x, T6w); T6C = FMA(KP198912367, T6w, T6x); T6m = FMA(KP668178637, T6e, T6h); T6i = FNMS(KP668178637, T6h, T6e); } } { E T6k, T64, T8R, T8T, T6t, T6a; T6k = FNMS(KP923879532, T63, T5W); T64 = FMA(KP923879532, T63, T5W); T8R = FMA(KP923879532, T8Q, T8P); T8T = FNMS(KP923879532, T8Q, T8P); T6t = FMA(KP707106781, T69, T68); T6a = FNMS(KP707106781, T69, T68); { E T6A, T8X, T8Y, T6D; { E T6s, T6B, T6l, T6b, T6z, T6v; T6A = FMA(KP923879532, T6r, T6o); T6s = FNMS(KP923879532, T6r, T6o); T6v = FMA(KP198912367, T6u, T6t); T6B = FNMS(KP198912367, T6t, T6u); T6l = FNMS(KP668178637, T67, T6a); T6b = FMA(KP668178637, T6a, T67); T6z = T6v - T6y; T90 = T6v + T6y; T8Z = FMA(KP923879532, T8W, T8V); T8X = FNMS(KP923879532, T8W, T8V); { E T6n, T8S, T8U, T6j; T6n = T6l - T6m; T8S = T6l + T6m; T8U = T6i - T6b; T6j = T6b + T6i; ri[WS(rs, 7)] = FMA(KP980785280, T6z, T6s); ri[WS(rs, 23)] = FNMS(KP980785280, T6z, T6s); ri[WS(rs, 11)] = FMA(KP831469612, T6n, T6k); ri[WS(rs, 27)] = FNMS(KP831469612, T6n, T6k); ii[WS(rs, 19)] = FNMS(KP831469612, T8S, T8R); ii[WS(rs, 3)] = FMA(KP831469612, T8S, T8R); ii[WS(rs, 27)] = FNMS(KP831469612, T8U, T8T); ii[WS(rs, 11)] = FMA(KP831469612, T8U, T8T); ri[WS(rs, 3)] = FMA(KP831469612, T6j, T64); ri[WS(rs, 19)] = FNMS(KP831469612, T6j, T64); T8Y = T6C - T6B; T6D = T6B + T6C; } } ii[WS(rs, 23)] = FNMS(KP980785280, T8Y, T8X); ii[WS(rs, 7)] = FMA(KP980785280, T8Y, T8X); ri[WS(rs, 31)] = FMA(KP980785280, T6D, T6A); ri[WS(rs, 15)] = FNMS(KP980785280, T6D, T6A); } } } } } } ii[WS(rs, 31)] = FMA(KP980785280, T90, T8Z); ii[WS(rs, 15)] = FNMS(KP980785280, T90, T8Z); } } } static const tw_instr twinstr[] = { {TW_FULL, 0, 32}, {TW_NEXT, 1, 0} }; static const ct_desc desc = { 32, "t1_32", twinstr, &GENUS, {236, 62, 198, 0}, 0, 0, 0 }; void X(codelet_t1_32) (planner *p) { X(kdft_dit_register) (p, t1_32, &desc); } #else /* HAVE_FMA */ /* Generated by: ../../../genfft/gen_twiddle.native -compact -variables 4 -pipeline-latency 4 -n 32 -name t1_32 -include t.h */ /* * This function contains 434 FP additions, 208 FP multiplications, * (or, 340 additions, 114 multiplications, 94 fused multiply/add), * 96 stack variables, 7 constants, and 128 memory accesses */ #include "t.h" static void t1_32(R *ri, R *ii, const R *W, stride rs, INT mb, INT me, INT ms) { DK(KP195090322, +0.195090322016128267848284868477022240927691618); DK(KP980785280, +0.980785280403230449126182236134239036973933731); DK(KP555570233, +0.555570233019602224742830813948532874374937191); DK(KP831469612, +0.831469612302545237078788377617905756738560812); DK(KP382683432, +0.382683432365089771728459984030398866761344562); DK(KP923879532, +0.923879532511286756128183189396788286822416626); DK(KP707106781, +0.707106781186547524400844362104849039284835938); { INT m; for (m = mb, W = W + (mb * 62); m < me; m = m + 1, ri = ri + ms, ii = ii + ms, W = W + 62, MAKE_VOLATILE_STRIDE(64, rs)) { E Tj, T5F, T7C, T7Q, T35, T4T, T78, T7m, T1Q, T61, T5Y, T6J, T3K, T59, T41; E T56, T2B, T67, T6e, T6O, T4b, T5d, T4s, T5g, TG, T7l, T5I, T73, T3a, T4U; E T3f, T4V, T14, T5N, T5M, T6E, T3m, T4Y, T3r, T4Z, T1r, T5P, T5S, T6F, T3x; E T51, T3C, T52, T2d, T5Z, T64, T6K, T3V, T57, T44, T5a, T2Y, T6f, T6a, T6P; E T4m, T5h, T4v, T5e; { E T1, T76, T6, T75, Tc, T32, Th, T33; T1 = ri[0]; T76 = ii[0]; { E T3, T5, T2, T4; T3 = ri[WS(rs, 16)]; T5 = ii[WS(rs, 16)]; T2 = W[30]; T4 = W[31]; T6 = FMA(T2, T3, T4 * T5); T75 = FNMS(T4, T3, T2 * T5); } { E T9, Tb, T8, Ta; T9 = ri[WS(rs, 8)]; Tb = ii[WS(rs, 8)]; T8 = W[14]; Ta = W[15]; Tc = FMA(T8, T9, Ta * Tb); T32 = FNMS(Ta, T9, T8 * Tb); } { E Te, Tg, Td, Tf; Te = ri[WS(rs, 24)]; Tg = ii[WS(rs, 24)]; Td = W[46]; Tf = W[47]; Th = FMA(Td, Te, Tf * Tg); T33 = FNMS(Tf, Te, Td * Tg); } { E T7, Ti, T7A, T7B; T7 = T1 + T6; Ti = Tc + Th; Tj = T7 + Ti; T5F = T7 - Ti; T7A = T76 - T75; T7B = Tc - Th; T7C = T7A - T7B; T7Q = T7B + T7A; } { E T31, T34, T74, T77; T31 = T1 - T6; T34 = T32 - T33; T35 = T31 - T34; T4T = T31 + T34; T74 = T32 + T33; T77 = T75 + T76; T78 = T74 + T77; T7m = T77 - T74; } } { E T1y, T3G, T1O, T3Z, T1D, T3H, T1J, T3Y; { E T1v, T1x, T1u, T1w; T1v = ri[WS(rs, 1)]; T1x = ii[WS(rs, 1)]; T1u = W[0]; T1w = W[1]; T1y = FMA(T1u, T1v, T1w * T1x); T3G = FNMS(T1w, T1v, T1u * T1x); } { E T1L, T1N, T1K, T1M; T1L = ri[WS(rs, 25)]; T1N = ii[WS(rs, 25)]; T1K = W[48]; T1M = W[49]; T1O = FMA(T1K, T1L, T1M * T1N); T3Z = FNMS(T1M, T1L, T1K * T1N); } { E T1A, T1C, T1z, T1B; T1A = ri[WS(rs, 17)]; T1C = ii[WS(rs, 17)]; T1z = W[32]; T1B = W[33]; T1D = FMA(T1z, T1A, T1B * T1C); T3H = FNMS(T1B, T1A, T1z * T1C); } { E T1G, T1I, T1F, T1H; T1G = ri[WS(rs, 9)]; T1I = ii[WS(rs, 9)]; T1F = W[16]; T1H = W[17]; T1J = FMA(T1F, T1G, T1H * T1I); T3Y = FNMS(T1H, T1G, T1F * T1I); } { E T1E, T1P, T5W, T5X; T1E = T1y + T1D; T1P = T1J + T1O; T1Q = T1E + T1P; T61 = T1E - T1P; T5W = T3G + T3H; T5X = T3Y + T3Z; T5Y = T5W - T5X; T6J = T5W + T5X; } { E T3I, T3J, T3X, T40; T3I = T3G - T3H; T3J = T1J - T1O; T3K = T3I + T3J; T59 = T3I - T3J; T3X = T1y - T1D; T40 = T3Y - T3Z; T41 = T3X - T40; T56 = T3X + T40; } } { E T2j, T4o, T2z, T49, T2o, T4p, T2u, T48; { E T2g, T2i, T2f, T2h; T2g = ri[WS(rs, 31)]; T2i = ii[WS(rs, 31)]; T2f = W[60]; T2h = W[61]; T2j = FMA(T2f, T2g, T2h * T2i); T4o = FNMS(T2h, T2g, T2f * T2i); } { E T2w, T2y, T2v, T2x; T2w = ri[WS(rs, 23)]; T2y = ii[WS(rs, 23)]; T2v = W[44]; T2x = W[45]; T2z = FMA(T2v, T2w, T2x * T2y); T49 = FNMS(T2x, T2w, T2v * T2y); } { E T2l, T2n, T2k, T2m; T2l = ri[WS(rs, 15)]; T2n = ii[WS(rs, 15)]; T2k = W[28]; T2m = W[29]; T2o = FMA(T2k, T2l, T2m * T2n); T4p = FNMS(T2m, T2l, T2k * T2n); } { E T2r, T2t, T2q, T2s; T2r = ri[WS(rs, 7)]; T2t = ii[WS(rs, 7)]; T2q = W[12]; T2s = W[13]; T2u = FMA(T2q, T2r, T2s * T2t); T48 = FNMS(T2s, T2r, T2q * T2t); } { E T2p, T2A, T6c, T6d; T2p = T2j + T2o; T2A = T2u + T2z; T2B = T2p + T2A; T67 = T2p - T2A; T6c = T4o + T4p; T6d = T48 + T49; T6e = T6c - T6d; T6O = T6c + T6d; } { E T47, T4a, T4q, T4r; T47 = T2j - T2o; T4a = T48 - T49; T4b = T47 - T4a; T5d = T47 + T4a; T4q = T4o - T4p; T4r = T2u - T2z; T4s = T4q + T4r; T5g = T4q - T4r; } } { E To, T36, TE, T3d, Tt, T37, Tz, T3c; { E Tl, Tn, Tk, Tm; Tl = ri[WS(rs, 4)]; Tn = ii[WS(rs, 4)]; Tk = W[6]; Tm = W[7]; To = FMA(Tk, Tl, Tm * Tn); T36 = FNMS(Tm, Tl, Tk * Tn); } { E TB, TD, TA, TC; TB = ri[WS(rs, 12)]; TD = ii[WS(rs, 12)]; TA = W[22]; TC = W[23]; TE = FMA(TA, TB, TC * TD); T3d = FNMS(TC, TB, TA * TD); } { E Tq, Ts, Tp, Tr; Tq = ri[WS(rs, 20)]; Ts = ii[WS(rs, 20)]; Tp = W[38]; Tr = W[39]; Tt = FMA(Tp, Tq, Tr * Ts); T37 = FNMS(Tr, Tq, Tp * Ts); } { E Tw, Ty, Tv, Tx; Tw = ri[WS(rs, 28)]; Ty = ii[WS(rs, 28)]; Tv = W[54]; Tx = W[55]; Tz = FMA(Tv, Tw, Tx * Ty); T3c = FNMS(Tx, Tw, Tv * Ty); } { E Tu, TF, T5G, T5H; Tu = To + Tt; TF = Tz + TE; TG = Tu + TF; T7l = TF - Tu; T5G = T36 + T37; T5H = T3c + T3d; T5I = T5G - T5H; T73 = T5G + T5H; } { E T38, T39, T3b, T3e; T38 = T36 - T37; T39 = To - Tt; T3a = T38 - T39; T4U = T39 + T38; T3b = Tz - TE; T3e = T3c - T3d; T3f = T3b + T3e; T4V = T3b - T3e; } } { E TM, T3i, T12, T3p, TR, T3j, TX, T3o; { E TJ, TL, TI, TK; TJ = ri[WS(rs, 2)]; TL = ii[WS(rs, 2)]; TI = W[2]; TK = W[3]; TM = FMA(TI, TJ, TK * TL); T3i = FNMS(TK, TJ, TI * TL); } { E TZ, T11, TY, T10; TZ = ri[WS(rs, 26)]; T11 = ii[WS(rs, 26)]; TY = W[50]; T10 = W[51]; T12 = FMA(TY, TZ, T10 * T11); T3p = FNMS(T10, TZ, TY * T11); } { E TO, TQ, TN, TP; TO = ri[WS(rs, 18)]; TQ = ii[WS(rs, 18)]; TN = W[34]; TP = W[35]; TR = FMA(TN, TO, TP * TQ); T3j = FNMS(TP, TO, TN * TQ); } { E TU, TW, TT, TV; TU = ri[WS(rs, 10)]; TW = ii[WS(rs, 10)]; TT = W[18]; TV = W[19]; TX = FMA(TT, TU, TV * TW); T3o = FNMS(TV, TU, TT * TW); } { E TS, T13, T5K, T5L; TS = TM + TR; T13 = TX + T12; T14 = TS + T13; T5N = TS - T13; T5K = T3i + T3j; T5L = T3o + T3p; T5M = T5K - T5L; T6E = T5K + T5L; } { E T3k, T3l, T3n, T3q; T3k = T3i - T3j; T3l = TX - T12; T3m = T3k + T3l; T4Y = T3k - T3l; T3n = TM - TR; T3q = T3o - T3p; T3r = T3n - T3q; T4Z = T3n + T3q; } } { E T19, T3t, T1p, T3A, T1e, T3u, T1k, T3z; { E T16, T18, T15, T17; T16 = ri[WS(rs, 30)]; T18 = ii[WS(rs, 30)]; T15 = W[58]; T17 = W[59]; T19 = FMA(T15, T16, T17 * T18); T3t = FNMS(T17, T16, T15 * T18); } { E T1m, T1o, T1l, T1n; T1m = ri[WS(rs, 22)]; T1o = ii[WS(rs, 22)]; T1l = W[42]; T1n = W[43]; T1p = FMA(T1l, T1m, T1n * T1o); T3A = FNMS(T1n, T1m, T1l * T1o); } { E T1b, T1d, T1a, T1c; T1b = ri[WS(rs, 14)]; T1d = ii[WS(rs, 14)]; T1a = W[26]; T1c = W[27]; T1e = FMA(T1a, T1b, T1c * T1d); T3u = FNMS(T1c, T1b, T1a * T1d); } { E T1h, T1j, T1g, T1i; T1h = ri[WS(rs, 6)]; T1j = ii[WS(rs, 6)]; T1g = W[10]; T1i = W[11]; T1k = FMA(T1g, T1h, T1i * T1j); T3z = FNMS(T1i, T1h, T1g * T1j); } { E T1f, T1q, T5Q, T5R; T1f = T19 + T1e; T1q = T1k + T1p; T1r = T1f + T1q; T5P = T1f - T1q; T5Q = T3t + T3u; T5R = T3z + T3A; T5S = T5Q - T5R; T6F = T5Q + T5R; } { E T3v, T3w, T3y, T3B; T3v = T3t - T3u; T3w = T1k - T1p; T3x = T3v + T3w; T51 = T3v - T3w; T3y = T19 - T1e; T3B = T3z - T3A; T3C = T3y - T3B; T52 = T3y + T3B; } } { E T1V, T3R, T20, T3S, T3Q, T3T, T26, T3M, T2b, T3N, T3L, T3O; { E T1S, T1U, T1R, T1T; T1S = ri[WS(rs, 5)]; T1U = ii[WS(rs, 5)]; T1R = W[8]; T1T = W[9]; T1V = FMA(T1R, T1S, T1T * T1U); T3R = FNMS(T1T, T1S, T1R * T1U); } { E T1X, T1Z, T1W, T1Y; T1X = ri[WS(rs, 21)]; T1Z = ii[WS(rs, 21)]; T1W = W[40]; T1Y = W[41]; T20 = FMA(T1W, T1X, T1Y * T1Z); T3S = FNMS(T1Y, T1X, T1W * T1Z); } T3Q = T1V - T20; T3T = T3R - T3S; { E T23, T25, T22, T24; T23 = ri[WS(rs, 29)]; T25 = ii[WS(rs, 29)]; T22 = W[56]; T24 = W[57]; T26 = FMA(T22, T23, T24 * T25); T3M = FNMS(T24, T23, T22 * T25); } { E T28, T2a, T27, T29; T28 = ri[WS(rs, 13)]; T2a = ii[WS(rs, 13)]; T27 = W[24]; T29 = W[25]; T2b = FMA(T27, T28, T29 * T2a); T3N = FNMS(T29, T28, T27 * T2a); } T3L = T26 - T2b; T3O = T3M - T3N; { E T21, T2c, T62, T63; T21 = T1V + T20; T2c = T26 + T2b; T2d = T21 + T2c; T5Z = T2c - T21; T62 = T3R + T3S; T63 = T3M + T3N; T64 = T62 - T63; T6K = T62 + T63; } { E T3P, T3U, T42, T43; T3P = T3L - T3O; T3U = T3Q + T3T; T3V = KP707106781 * (T3P - T3U); T57 = KP707106781 * (T3U + T3P); T42 = T3T - T3Q; T43 = T3L + T3O; T44 = KP707106781 * (T42 - T43); T5a = KP707106781 * (T42 + T43); } } { E T2G, T4c, T2L, T4d, T4e, T4f, T2R, T4i, T2W, T4j, T4h, T4k; { E T2D, T2F, T2C, T2E; T2D = ri[WS(rs, 3)]; T2F = ii[WS(rs, 3)]; T2C = W[4]; T2E = W[5]; T2G = FMA(T2C, T2D, T2E * T2F); T4c = FNMS(T2E, T2D, T2C * T2F); } { E T2I, T2K, T2H, T2J; T2I = ri[WS(rs, 19)]; T2K = ii[WS(rs, 19)]; T2H = W[36]; T2J = W[37]; T2L = FMA(T2H, T2I, T2J * T2K); T4d = FNMS(T2J, T2I, T2H * T2K); } T4e = T4c - T4d; T4f = T2G - T2L; { E T2O, T2Q, T2N, T2P; T2O = ri[WS(rs, 27)]; T2Q = ii[WS(rs, 27)]; T2N = W[52]; T2P = W[53]; T2R = FMA(T2N, T2O, T2P * T2Q); T4i = FNMS(T2P, T2O, T2N * T2Q); } { E T2T, T2V, T2S, T2U; T2T = ri[WS(rs, 11)]; T2V = ii[WS(rs, 11)]; T2S = W[20]; T2U = W[21]; T2W = FMA(T2S, T2T, T2U * T2V); T4j = FNMS(T2U, T2T, T2S * T2V); } T4h = T2R - T2W; T4k = T4i - T4j; { E T2M, T2X, T68, T69; T2M = T2G + T2L; T2X = T2R + T2W; T2Y = T2M + T2X; T6f = T2X - T2M; T68 = T4c + T4d; T69 = T4i + T4j; T6a = T68 - T69; T6P = T68 + T69; } { E T4g, T4l, T4t, T4u; T4g = T4e - T4f; T4l = T4h + T4k; T4m = KP707106781 * (T4g - T4l); T5h = KP707106781 * (T4g + T4l); T4t = T4h - T4k; T4u = T4f + T4e; T4v = KP707106781 * (T4t - T4u); T5e = KP707106781 * (T4u + T4t); } } { E T1t, T6X, T7a, T7c, T30, T7b, T70, T71; { E TH, T1s, T72, T79; TH = Tj + TG; T1s = T14 + T1r; T1t = TH + T1s; T6X = TH - T1s; T72 = T6E + T6F; T79 = T73 + T78; T7a = T72 + T79; T7c = T79 - T72; } { E T2e, T2Z, T6Y, T6Z; T2e = T1Q + T2d; T2Z = T2B + T2Y; T30 = T2e + T2Z; T7b = T2Z - T2e; T6Y = T6J + T6K; T6Z = T6O + T6P; T70 = T6Y - T6Z; T71 = T6Y + T6Z; } ri[WS(rs, 16)] = T1t - T30; ii[WS(rs, 16)] = T7a - T71; ri[0] = T1t + T30; ii[0] = T71 + T7a; ri[WS(rs, 24)] = T6X - T70; ii[WS(rs, 24)] = T7c - T7b; ri[WS(rs, 8)] = T6X + T70; ii[WS(rs, 8)] = T7b + T7c; } { E T6H, T6T, T7g, T7i, T6M, T6U, T6R, T6V; { E T6D, T6G, T7e, T7f; T6D = Tj - TG; T6G = T6E - T6F; T6H = T6D + T6G; T6T = T6D - T6G; T7e = T1r - T14; T7f = T78 - T73; T7g = T7e + T7f; T7i = T7f - T7e; } { E T6I, T6L, T6N, T6Q; T6I = T1Q - T2d; T6L = T6J - T6K; T6M = T6I + T6L; T6U = T6L - T6I; T6N = T2B - T2Y; T6Q = T6O - T6P; T6R = T6N - T6Q; T6V = T6N + T6Q; } { E T6S, T7d, T6W, T7h; T6S = KP707106781 * (T6M + T6R); ri[WS(rs, 20)] = T6H - T6S; ri[WS(rs, 4)] = T6H + T6S; T7d = KP707106781 * (T6U + T6V); ii[WS(rs, 4)] = T7d + T7g; ii[WS(rs, 20)] = T7g - T7d; T6W = KP707106781 * (T6U - T6V); ri[WS(rs, 28)] = T6T - T6W; ri[WS(rs, 12)] = T6T + T6W; T7h = KP707106781 * (T6R - T6M); ii[WS(rs, 12)] = T7h + T7i; ii[WS(rs, 28)] = T7i - T7h; } } { E T5J, T7n, T7t, T6n, T5U, T7k, T6x, T6B, T6q, T7s, T66, T6k, T6u, T6A, T6h; E T6l; { E T5O, T5T, T60, T65; T5J = T5F - T5I; T7n = T7l + T7m; T7t = T7m - T7l; T6n = T5F + T5I; T5O = T5M - T5N; T5T = T5P + T5S; T5U = KP707106781 * (T5O - T5T); T7k = KP707106781 * (T5O + T5T); { E T6v, T6w, T6o, T6p; T6v = T67 + T6a; T6w = T6e + T6f; T6x = FNMS(KP382683432, T6w, KP923879532 * T6v); T6B = FMA(KP923879532, T6w, KP382683432 * T6v); T6o = T5N + T5M; T6p = T5P - T5S; T6q = KP707106781 * (T6o + T6p); T7s = KP707106781 * (T6p - T6o); } T60 = T5Y - T5Z; T65 = T61 - T64; T66 = FMA(KP923879532, T60, KP382683432 * T65); T6k = FNMS(KP923879532, T65, KP382683432 * T60); { E T6s, T6t, T6b, T6g; T6s = T5Y + T5Z; T6t = T61 + T64; T6u = FMA(KP382683432, T6s, KP923879532 * T6t); T6A = FNMS(KP382683432, T6t, KP923879532 * T6s); T6b = T67 - T6a; T6g = T6e - T6f; T6h = FNMS(KP923879532, T6g, KP382683432 * T6b); T6l = FMA(KP382683432, T6g, KP923879532 * T6b); } } { E T5V, T6i, T7r, T7u; T5V = T5J + T5U; T6i = T66 + T6h; ri[WS(rs, 22)] = T5V - T6i; ri[WS(rs, 6)] = T5V + T6i; T7r = T6k + T6l; T7u = T7s + T7t; ii[WS(rs, 6)] = T7r + T7u; ii[WS(rs, 22)] = T7u - T7r; } { E T6j, T6m, T7v, T7w; T6j = T5J - T5U; T6m = T6k - T6l; ri[WS(rs, 30)] = T6j - T6m; ri[WS(rs, 14)] = T6j + T6m; T7v = T6h - T66; T7w = T7t - T7s; ii[WS(rs, 14)] = T7v + T7w; ii[WS(rs, 30)] = T7w - T7v; } { E T6r, T6y, T7j, T7o; T6r = T6n + T6q; T6y = T6u + T6x; ri[WS(rs, 18)] = T6r - T6y; ri[WS(rs, 2)] = T6r + T6y; T7j = T6A + T6B; T7o = T7k + T7n; ii[WS(rs, 2)] = T7j + T7o; ii[WS(rs, 18)] = T7o - T7j; } { E T6z, T6C, T7p, T7q; T6z = T6n - T6q; T6C = T6A - T6B; ri[WS(rs, 26)] = T6z - T6C; ri[WS(rs, 10)] = T6z + T6C; T7p = T6x - T6u; T7q = T7n - T7k; ii[WS(rs, 10)] = T7p + T7q; ii[WS(rs, 26)] = T7q - T7p; } } { E T3h, T4D, T7R, T7X, T3E, T7O, T4N, T4R, T46, T4A, T4G, T7W, T4K, T4Q, T4x; E T4B, T3g, T7P; T3g = KP707106781 * (T3a - T3f); T3h = T35 - T3g; T4D = T35 + T3g; T7P = KP707106781 * (T4V - T4U); T7R = T7P + T7Q; T7X = T7Q - T7P; { E T3s, T3D, T4L, T4M; T3s = FNMS(KP923879532, T3r, KP382683432 * T3m); T3D = FMA(KP382683432, T3x, KP923879532 * T3C); T3E = T3s - T3D; T7O = T3s + T3D; T4L = T4b + T4m; T4M = T4s + T4v; T4N = FNMS(KP555570233, T4M, KP831469612 * T4L); T4R = FMA(KP831469612, T4M, KP555570233 * T4L); } { E T3W, T45, T4E, T4F; T3W = T3K - T3V; T45 = T41 - T44; T46 = FMA(KP980785280, T3W, KP195090322 * T45); T4A = FNMS(KP980785280, T45, KP195090322 * T3W); T4E = FMA(KP923879532, T3m, KP382683432 * T3r); T4F = FNMS(KP923879532, T3x, KP382683432 * T3C); T4G = T4E + T4F; T7W = T4F - T4E; } { E T4I, T4J, T4n, T4w; T4I = T3K + T3V; T4J = T41 + T44; T4K = FMA(KP555570233, T4I, KP831469612 * T4J); T4Q = FNMS(KP555570233, T4J, KP831469612 * T4I); T4n = T4b - T4m; T4w = T4s - T4v; T4x = FNMS(KP980785280, T4w, KP195090322 * T4n); T4B = FMA(KP195090322, T4w, KP980785280 * T4n); } { E T3F, T4y, T7V, T7Y; T3F = T3h + T3E; T4y = T46 + T4x; ri[WS(rs, 23)] = T3F - T4y; ri[WS(rs, 7)] = T3F + T4y; T7V = T4A + T4B; T7Y = T7W + T7X; ii[WS(rs, 7)] = T7V + T7Y; ii[WS(rs, 23)] = T7Y - T7V; } { E T4z, T4C, T7Z, T80; T4z = T3h - T3E; T4C = T4A - T4B; ri[WS(rs, 31)] = T4z - T4C; ri[WS(rs, 15)] = T4z + T4C; T7Z = T4x - T46; T80 = T7X - T7W; ii[WS(rs, 15)] = T7Z + T80; ii[WS(rs, 31)] = T80 - T7Z; } { E T4H, T4O, T7N, T7S; T4H = T4D + T4G; T4O = T4K + T4N; ri[WS(rs, 19)] = T4H - T4O; ri[WS(rs, 3)] = T4H + T4O; T7N = T4Q + T4R; T7S = T7O + T7R; ii[WS(rs, 3)] = T7N + T7S; ii[WS(rs, 19)] = T7S - T7N; } { E T4P, T4S, T7T, T7U; T4P = T4D - T4G; T4S = T4Q - T4R; ri[WS(rs, 27)] = T4P - T4S; ri[WS(rs, 11)] = T4P + T4S; T7T = T4N - T4K; T7U = T7R - T7O; ii[WS(rs, 11)] = T7T + T7U; ii[WS(rs, 27)] = T7U - T7T; } } { E T4X, T5p, T7D, T7J, T54, T7y, T5z, T5D, T5c, T5m, T5s, T7I, T5w, T5C, T5j; E T5n, T4W, T7z; T4W = KP707106781 * (T4U + T4V); T4X = T4T - T4W; T5p = T4T + T4W; T7z = KP707106781 * (T3a + T3f); T7D = T7z + T7C; T7J = T7C - T7z; { E T50, T53, T5x, T5y; T50 = FNMS(KP382683432, T4Z, KP923879532 * T4Y); T53 = FMA(KP923879532, T51, KP382683432 * T52); T54 = T50 - T53; T7y = T50 + T53; T5x = T5d + T5e; T5y = T5g + T5h; T5z = FNMS(KP195090322, T5y, KP980785280 * T5x); T5D = FMA(KP195090322, T5x, KP980785280 * T5y); } { E T58, T5b, T5q, T5r; T58 = T56 - T57; T5b = T59 - T5a; T5c = FMA(KP555570233, T58, KP831469612 * T5b); T5m = FNMS(KP831469612, T58, KP555570233 * T5b); T5q = FMA(KP382683432, T4Y, KP923879532 * T4Z); T5r = FNMS(KP382683432, T51, KP923879532 * T52); T5s = T5q + T5r; T7I = T5r - T5q; } { E T5u, T5v, T5f, T5i; T5u = T56 + T57; T5v = T59 + T5a; T5w = FMA(KP980785280, T5u, KP195090322 * T5v); T5C = FNMS(KP195090322, T5u, KP980785280 * T5v); T5f = T5d - T5e; T5i = T5g - T5h; T5j = FNMS(KP831469612, T5i, KP555570233 * T5f); T5n = FMA(KP831469612, T5f, KP555570233 * T5i); } { E T55, T5k, T7H, T7K; T55 = T4X + T54; T5k = T5c + T5j; ri[WS(rs, 21)] = T55 - T5k; ri[WS(rs, 5)] = T55 + T5k; T7H = T5m + T5n; T7K = T7I + T7J; ii[WS(rs, 5)] = T7H + T7K; ii[WS(rs, 21)] = T7K - T7H; } { E T5l, T5o, T7L, T7M; T5l = T4X - T54; T5o = T5m - T5n; ri[WS(rs, 29)] = T5l - T5o; ri[WS(rs, 13)] = T5l + T5o; T7L = T5j - T5c; T7M = T7J - T7I; ii[WS(rs, 13)] = T7L + T7M; ii[WS(rs, 29)] = T7M - T7L; } { E T5t, T5A, T7x, T7E; T5t = T5p + T5s; T5A = T5w + T5z; ri[WS(rs, 17)] = T5t - T5A; ri[WS(rs, 1)] = T5t + T5A; T7x = T5C + T5D; T7E = T7y + T7D; ii[WS(rs, 1)] = T7x + T7E; ii[WS(rs, 17)] = T7E - T7x; } { E T5B, T5E, T7F, T7G; T5B = T5p - T5s; T5E = T5C - T5D; ri[WS(rs, 25)] = T5B - T5E; ri[WS(rs, 9)] = T5B + T5E; T7F = T5z - T5w; T7G = T7D - T7y; ii[WS(rs, 9)] = T7F + T7G; ii[WS(rs, 25)] = T7G - T7F; } } } } } static const tw_instr twinstr[] = { {TW_FULL, 0, 32}, {TW_NEXT, 1, 0} }; static const ct_desc desc = { 32, "t1_32", twinstr, &GENUS, {340, 114, 94, 0}, 0, 0, 0 }; void X(codelet_t1_32) (planner *p) { X(kdft_dit_register) (p, t1_32, &desc); } #endif /* HAVE_FMA */ fftw-3.3.3/dft/scalar/codelets/t1_64.c0000644000175000017500000032356312054410321014240 00000000000000/* * Copyright (c) 2003, 2007-11 Matteo Frigo * Copyright (c) 2003, 2007-11 Massachusetts Institute of Technology * * This program is free software; you can redistribute it and/or modify * it under the terms of the GNU General Public License as published by * the Free Software Foundation; either version 2 of the License, or * (at your option) any later version. * * This program is distributed in the hope that it will be useful, * but WITHOUT ANY WARRANTY; without even the implied warranty of * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the * GNU General Public License for more details. * * You should have received a copy of the GNU General Public License * along with this program; if not, write to the Free Software * Foundation, Inc., 51 Franklin Street, Fifth Floor, Boston, MA 02110-1301 USA * */ /* This file was automatically generated --- DO NOT EDIT */ /* Generated on Sun Nov 25 07:35:52 EST 2012 */ #include "codelet-dft.h" #ifdef HAVE_FMA /* Generated by: ../../../genfft/gen_twiddle.native -fma -reorder-insns -schedule-for-pipeline -compact -variables 4 -pipeline-latency 4 -n 64 -name t1_64 -include t.h */ /* * This function contains 1038 FP additions, 644 FP multiplications, * (or, 520 additions, 126 multiplications, 518 fused multiply/add), * 228 stack variables, 15 constants, and 256 memory accesses */ #include "t.h" static void t1_64(R *ri, R *ii, const R *W, stride rs, INT mb, INT me, INT ms) { DK(KP995184726, +0.995184726672196886244836953109479921575474869); DK(KP773010453, +0.773010453362736960810906609758469800971041293); DK(KP956940335, +0.956940335732208864935797886980269969482849206); DK(KP881921264, +0.881921264348355029712756863660388349508442621); DK(KP820678790, +0.820678790828660330972281985331011598767386482); DK(KP098491403, +0.098491403357164253077197521291327432293052451); DK(KP534511135, +0.534511135950791641089685961295362908582039528); DK(KP303346683, +0.303346683607342391675883946941299872384187453); DK(KP831469612, +0.831469612302545237078788377617905756738560812); DK(KP980785280, +0.980785280403230449126182236134239036973933731); DK(KP668178637, +0.668178637919298919997757686523080761552472251); DK(KP198912367, +0.198912367379658006911597622644676228597850501); DK(KP923879532, +0.923879532511286756128183189396788286822416626); DK(KP707106781, +0.707106781186547524400844362104849039284835938); DK(KP414213562, +0.414213562373095048801688724209698078569671875); { INT m; for (m = mb, W = W + (mb * 126); m < me; m = m + 1, ri = ri + ms, ii = ii + ms, W = W + 126, MAKE_VOLATILE_STRIDE(128, rs)) { E TeI, Tkk, Tkj, TeL; { E TiV, Tjm, T7e, TcA, TjR, Tkl, Tm, TeM, TeZ, Ths, T7Q, TcJ, T1G, TeW, TcI; E T7X, Tf5, Thv, T87, TcN, T29, Tf8, TcQ, T8u, TfU, ThS, Taq, Tdm, T5K, Tg9; E Tdx, Tbj, TcB, T7l, TiP, TeP, Tjl, TN, TcC, T7s, T7I, TcF, TeU, Thr, T7B; E TcG, T1f, TeR, Tfg, ThB, T8G, TcU, T32, Tfj, TcX, T93, Tft, ThH, T9h, Td3; E T3X, TfI, Tde, Taa, Thw, Tfb, Tf6, T2A, T8x, TcO, T8m, TcR, Tfm, ThC, T3t; E Tfh, T96, TcV, T8V, TcY, ThI, TfL, Tfu, T4o, Tad, Td4, T9w, Tdf, Tgc, ThT; E T6b, TfV, Tbm, Tdn, TaF, Tdy, ThN, T4Q, TfN, TfA, Taf, Ta1, Td8, Tdh, ThO; E T5h, TfO, TfF, Tag, T9M, Tdb, Tdi, ThY, T6D, Tge, Tg1, Tbo, Tba, Tdr, TdA; E TaN, Tdt, Tg5, ThZ, Tg2, T74, Tds, TaU; { E T7a, Te, T78, T8, TjP, TiU, T7c, Tk; { E T1, TiT, TiS, T7, Tg, Tj, Tf, Ti, T7b, Th; T1 = ri[0]; TiT = ii[0]; { E T3, T6, T2, T5; T3 = ri[WS(rs, 32)]; T6 = ii[WS(rs, 32)]; T2 = W[62]; T5 = W[63]; { E Ta, Td, Tc, T79, Tb, TiR, T4, T9; Ta = ri[WS(rs, 16)]; Td = ii[WS(rs, 16)]; TiR = T2 * T6; T4 = T2 * T3; T9 = W[30]; Tc = W[31]; TiS = FNMS(T5, T3, TiR); T7 = FMA(T5, T6, T4); T79 = T9 * Td; Tb = T9 * Ta; Tg = ri[WS(rs, 48)]; Tj = ii[WS(rs, 48)]; T7a = FNMS(Tc, Ta, T79); Te = FMA(Tc, Td, Tb); Tf = W[94]; Ti = W[95]; } } T78 = T1 - T7; T8 = T1 + T7; TjP = TiT - TiS; TiU = TiS + TiT; T7b = Tf * Tj; Th = Tf * Tg; T7c = FNMS(Ti, Tg, T7b); Tk = FMA(Ti, Tj, Th); } { E T7L, T1l, T7V, T1E, T1u, T1x, T1w, T7N, T1r, T7S, T1v; { E T1A, T1D, T1C, T7U, T1B; { E T1h, T1k, T1g, T1j, T7K, T1i, T1z; T1h = ri[WS(rs, 60)]; T1k = ii[WS(rs, 60)]; { E T7d, TiQ, Tl, TjQ; T7d = T7a - T7c; TiQ = T7a + T7c; Tl = Te + Tk; TjQ = Te - Tk; TiV = TiQ + TiU; Tjm = TiU - TiQ; T7e = T78 - T7d; TcA = T78 + T7d; TjR = TjP - TjQ; Tkl = TjQ + TjP; Tm = T8 + Tl; TeM = T8 - Tl; T1g = W[118]; } T1j = W[119]; T1A = ri[WS(rs, 44)]; T1D = ii[WS(rs, 44)]; T7K = T1g * T1k; T1i = T1g * T1h; T1z = W[86]; T1C = W[87]; T7L = FNMS(T1j, T1h, T7K); T1l = FMA(T1j, T1k, T1i); T7U = T1z * T1D; T1B = T1z * T1A; } { E T1n, T1q, T1m, T1p, T7M, T1o, T1t; T1n = ri[WS(rs, 28)]; T1q = ii[WS(rs, 28)]; T7V = FNMS(T1C, T1A, T7U); T1E = FMA(T1C, T1D, T1B); T1m = W[54]; T1p = W[55]; T1u = ri[WS(rs, 12)]; T1x = ii[WS(rs, 12)]; T7M = T1m * T1q; T1o = T1m * T1n; T1t = W[22]; T1w = W[23]; T7N = FNMS(T1p, T1n, T7M); T1r = FMA(T1p, T1q, T1o); T7S = T1t * T1x; T1v = T1t * T1u; } } { E T7O, TeX, T1s, T7R, T7T, T1y; T7O = T7L - T7N; TeX = T7L + T7N; T1s = T1l + T1r; T7R = T1l - T1r; T7T = FNMS(T1w, T1u, T7S); T1y = FMA(T1w, T1x, T1v); { E T7W, TeY, T7P, T1F; T7W = T7T - T7V; TeY = T7T + T7V; T7P = T1y - T1E; T1F = T1y + T1E; TeZ = TeX - TeY; Ths = TeX + TeY; T7Q = T7O + T7P; TcJ = T7O - T7P; T1G = T1s + T1F; TeW = T1s - T1F; TcI = T7R + T7W; T7X = T7R - T7W; } } } } { E T82, T1O, T8s, T27, T1X, T20, T1Z, T84, T1U, T8p, T1Y; { E T23, T26, T25, T8r, T24; { E T1K, T1N, T1J, T1M, T81, T1L, T22; T1K = ri[WS(rs, 2)]; T1N = ii[WS(rs, 2)]; T1J = W[2]; T1M = W[3]; T23 = ri[WS(rs, 50)]; T26 = ii[WS(rs, 50)]; T81 = T1J * T1N; T1L = T1J * T1K; T22 = W[98]; T25 = W[99]; T82 = FNMS(T1M, T1K, T81); T1O = FMA(T1M, T1N, T1L); T8r = T22 * T26; T24 = T22 * T23; } { E T1Q, T1T, T1P, T1S, T83, T1R, T1W; T1Q = ri[WS(rs, 34)]; T1T = ii[WS(rs, 34)]; T8s = FNMS(T25, T23, T8r); T27 = FMA(T25, T26, T24); T1P = W[66]; T1S = W[67]; T1X = ri[WS(rs, 18)]; T20 = ii[WS(rs, 18)]; T83 = T1P * T1T; T1R = T1P * T1Q; T1W = W[34]; T1Z = W[35]; T84 = FNMS(T1S, T1Q, T83); T1U = FMA(T1S, T1T, T1R); T8p = T1W * T20; T1Y = T1W * T1X; } } { E T85, Tf3, T1V, T8o, T8q, T21; T85 = T82 - T84; Tf3 = T82 + T84; T1V = T1O + T1U; T8o = T1O - T1U; T8q = FNMS(T1Z, T1X, T8p); T21 = FMA(T1Z, T20, T1Y); { E T8t, Tf4, T86, T28; T8t = T8q - T8s; Tf4 = T8q + T8s; T86 = T21 - T27; T28 = T21 + T27; Tf5 = Tf3 - Tf4; Thv = Tf3 + Tf4; T87 = T85 + T86; TcN = T85 - T86; T29 = T1V + T28; Tf8 = T1V - T28; TcQ = T8o + T8t; T8u = T8o - T8t; } } } { E Tal, T5p, Tbh, T5I, T5y, T5B, T5A, Tan, T5v, Tbe, T5z; { E T5E, T5H, T5G, Tbg, T5F; { E T5l, T5o, T5k, T5n, Tak, T5m, T5D; T5l = ri[WS(rs, 63)]; T5o = ii[WS(rs, 63)]; T5k = W[124]; T5n = W[125]; T5E = ri[WS(rs, 47)]; T5H = ii[WS(rs, 47)]; Tak = T5k * T5o; T5m = T5k * T5l; T5D = W[92]; T5G = W[93]; Tal = FNMS(T5n, T5l, Tak); T5p = FMA(T5n, T5o, T5m); Tbg = T5D * T5H; T5F = T5D * T5E; } { E T5r, T5u, T5q, T5t, Tam, T5s, T5x; T5r = ri[WS(rs, 31)]; T5u = ii[WS(rs, 31)]; Tbh = FNMS(T5G, T5E, Tbg); T5I = FMA(T5G, T5H, T5F); T5q = W[60]; T5t = W[61]; T5y = ri[WS(rs, 15)]; T5B = ii[WS(rs, 15)]; Tam = T5q * T5u; T5s = T5q * T5r; T5x = W[28]; T5A = W[29]; Tan = FNMS(T5t, T5r, Tam); T5v = FMA(T5t, T5u, T5s); Tbe = T5x * T5B; T5z = T5x * T5y; } } { E Tao, TfS, T5w, Tbd, Tbf, T5C; Tao = Tal - Tan; TfS = Tal + Tan; T5w = T5p + T5v; Tbd = T5p - T5v; Tbf = FNMS(T5A, T5y, Tbe); T5C = FMA(T5A, T5B, T5z); { E Tbi, TfT, Tap, T5J; Tbi = Tbf - Tbh; TfT = Tbf + Tbh; Tap = T5C - T5I; T5J = T5C + T5I; TfU = TfS - TfT; ThS = TfS + TfT; Taq = Tao + Tap; Tdm = Tao - Tap; T5K = T5w + T5J; Tg9 = T5w - T5J; Tdx = Tbd + Tbi; Tbj = Tbd - Tbi; } } } { E T7G, T1d, T7z, TeS, T11, T7C, T7E, T17, T7r, T7m; { E T7g, Ts, T7q, TL, TB, TE, TD, T7i, Ty, T7n, TC; { E TH, TK, TJ, T7p, TI; { E To, Tr, Tn, Tq, T7f, Tp, TG; To = ri[WS(rs, 8)]; Tr = ii[WS(rs, 8)]; Tn = W[14]; Tq = W[15]; TH = ri[WS(rs, 24)]; TK = ii[WS(rs, 24)]; T7f = Tn * Tr; Tp = Tn * To; TG = W[46]; TJ = W[47]; T7g = FNMS(Tq, To, T7f); Ts = FMA(Tq, Tr, Tp); T7p = TG * TK; TI = TG * TH; } { E Tu, Tx, Tt, Tw, T7h, Tv, TA; Tu = ri[WS(rs, 40)]; Tx = ii[WS(rs, 40)]; T7q = FNMS(TJ, TH, T7p); TL = FMA(TJ, TK, TI); Tt = W[78]; Tw = W[79]; TB = ri[WS(rs, 56)]; TE = ii[WS(rs, 56)]; T7h = Tt * Tx; Tv = Tt * Tu; TA = W[110]; TD = W[111]; T7i = FNMS(Tw, Tu, T7h); Ty = FMA(Tw, Tx, Tv); T7n = TA * TE; TC = TA * TB; } } { E T7j, TeN, Tz, T7k, T7o, TF, TeO, TM; T7j = T7g - T7i; TeN = T7g + T7i; Tz = Ts + Ty; T7k = Ts - Ty; T7o = FNMS(TD, TB, T7n); TF = FMA(TD, TE, TC); T7r = T7o - T7q; TeO = T7o + T7q; TM = TF + TL; T7m = TF - TL; TcB = T7k + T7j; T7l = T7j - T7k; TiP = TeN + TeO; TeP = TeN - TeO; Tjl = TM - Tz; TN = Tz + TM; } } { E T7w, TU, T13, T16, T7y, T10, T12, T15, T7D, T14; { E T19, T1c, T18, T1b; { E TQ, TT, TS, T7v, TR, TP; TQ = ri[WS(rs, 4)]; TT = ii[WS(rs, 4)]; TP = W[6]; TcC = T7m - T7r; T7s = T7m + T7r; TS = W[7]; T7v = TP * TT; TR = TP * TQ; T19 = ri[WS(rs, 52)]; T1c = ii[WS(rs, 52)]; T7w = FNMS(TS, TQ, T7v); TU = FMA(TS, TT, TR); T18 = W[102]; T1b = W[103]; } { E TW, TZ, TY, T7x, TX, T7F, T1a, TV; TW = ri[WS(rs, 36)]; TZ = ii[WS(rs, 36)]; T7F = T18 * T1c; T1a = T18 * T19; TV = W[70]; TY = W[71]; T7G = FNMS(T1b, T19, T7F); T1d = FMA(T1b, T1c, T1a); T7x = TV * TZ; TX = TV * TW; T13 = ri[WS(rs, 20)]; T16 = ii[WS(rs, 20)]; T7y = FNMS(TY, TW, T7x); T10 = FMA(TY, TZ, TX); T12 = W[38]; T15 = W[39]; } } T7z = T7w - T7y; TeS = T7w + T7y; T11 = TU + T10; T7C = TU - T10; T7D = T12 * T16; T14 = T12 * T13; T7E = FNMS(T15, T13, T7D); T17 = FMA(T15, T16, T14); } { E T8B, T2H, T91, T30, T2Q, T2T, T2S, T8D, T2N, T8Y, T2R; { E T2W, T2Z, T2Y, T90, T2X; { E T2D, T2G, T2C, T2F, T8A, T2E, T2V; T2D = ri[WS(rs, 62)]; T2G = ii[WS(rs, 62)]; { E TeT, T7H, T1e, T7A; TeT = T7E + T7G; T7H = T7E - T7G; T1e = T17 + T1d; T7A = T17 - T1d; T7I = T7C - T7H; TcF = T7C + T7H; TeU = TeS - TeT; Thr = TeS + TeT; T7B = T7z + T7A; TcG = T7z - T7A; T1f = T11 + T1e; TeR = T11 - T1e; T2C = W[122]; } T2F = W[123]; T2W = ri[WS(rs, 46)]; T2Z = ii[WS(rs, 46)]; T8A = T2C * T2G; T2E = T2C * T2D; T2V = W[90]; T2Y = W[91]; T8B = FNMS(T2F, T2D, T8A); T2H = FMA(T2F, T2G, T2E); T90 = T2V * T2Z; T2X = T2V * T2W; } { E T2J, T2M, T2I, T2L, T8C, T2K, T2P; T2J = ri[WS(rs, 30)]; T2M = ii[WS(rs, 30)]; T91 = FNMS(T2Y, T2W, T90); T30 = FMA(T2Y, T2Z, T2X); T2I = W[58]; T2L = W[59]; T2Q = ri[WS(rs, 14)]; T2T = ii[WS(rs, 14)]; T8C = T2I * T2M; T2K = T2I * T2J; T2P = W[26]; T2S = W[27]; T8D = FNMS(T2L, T2J, T8C); T2N = FMA(T2L, T2M, T2K); T8Y = T2P * T2T; T2R = T2P * T2Q; } } { E T8E, Tfe, T2O, T8X, T8Z, T2U; T8E = T8B - T8D; Tfe = T8B + T8D; T2O = T2H + T2N; T8X = T2H - T2N; T8Z = FNMS(T2S, T2Q, T8Y); T2U = FMA(T2S, T2T, T2R); { E T92, Tff, T8F, T31; T92 = T8Z - T91; Tff = T8Z + T91; T8F = T2U - T30; T31 = T2U + T30; Tfg = Tfe - Tff; ThB = Tfe + Tff; T8G = T8E + T8F; TcU = T8E - T8F; T32 = T2O + T31; Tfj = T2O - T31; TcX = T8X + T92; T93 = T8X - T92; } } } { E T9c, T3C, Ta8, T3V, T3L, T3O, T3N, T9e, T3I, Ta5, T3M; { E T3R, T3U, T3T, Ta7, T3S; { E T3y, T3B, T3x, T3A, T9b, T3z, T3Q; T3y = ri[WS(rs, 1)]; T3B = ii[WS(rs, 1)]; T3x = W[0]; T3A = W[1]; T3R = ri[WS(rs, 49)]; T3U = ii[WS(rs, 49)]; T9b = T3x * T3B; T3z = T3x * T3y; T3Q = W[96]; T3T = W[97]; T9c = FNMS(T3A, T3y, T9b); T3C = FMA(T3A, T3B, T3z); Ta7 = T3Q * T3U; T3S = T3Q * T3R; } { E T3E, T3H, T3D, T3G, T9d, T3F, T3K; T3E = ri[WS(rs, 33)]; T3H = ii[WS(rs, 33)]; Ta8 = FNMS(T3T, T3R, Ta7); T3V = FMA(T3T, T3U, T3S); T3D = W[64]; T3G = W[65]; T3L = ri[WS(rs, 17)]; T3O = ii[WS(rs, 17)]; T9d = T3D * T3H; T3F = T3D * T3E; T3K = W[32]; T3N = W[33]; T9e = FNMS(T3G, T3E, T9d); T3I = FMA(T3G, T3H, T3F); Ta5 = T3K * T3O; T3M = T3K * T3L; } } { E T9f, Tfr, T3J, Ta4, Ta6, T3P; T9f = T9c - T9e; Tfr = T9c + T9e; T3J = T3C + T3I; Ta4 = T3C - T3I; Ta6 = FNMS(T3N, T3L, Ta5); T3P = FMA(T3N, T3O, T3M); { E Ta9, Tfs, T9g, T3W; Ta9 = Ta6 - Ta8; Tfs = Ta6 + Ta8; T9g = T3P - T3V; T3W = T3P + T3V; Tft = Tfr - Tfs; ThH = Tfr + Tfs; T9h = T9f + T9g; Td3 = T9f - T9g; T3X = T3J + T3W; TfI = T3J - T3W; Tde = Ta4 + Ta9; Taa = Ta4 - Ta9; } } } } { E TaC, T69, Taw, Tga, T5X, Tar, TaA, T63; { E T8S, T3r, T8M, Tfk, T3f, T8H, T8Q, T3l; { E T8k, T8f, T8w, T8e; { E T8a, T2f, T8j, T2y, T2o, T2r, T2q, T8c, T2l, T8g, T2p; { E T2u, T2x, T2w, T8i, T2v; { E T2b, T2e, T2a, T2d, T89, T2c, T2t; T2b = ri[WS(rs, 10)]; T2e = ii[WS(rs, 10)]; T2a = W[18]; T2d = W[19]; T2u = ri[WS(rs, 26)]; T2x = ii[WS(rs, 26)]; T89 = T2a * T2e; T2c = T2a * T2b; T2t = W[50]; T2w = W[51]; T8a = FNMS(T2d, T2b, T89); T2f = FMA(T2d, T2e, T2c); T8i = T2t * T2x; T2v = T2t * T2u; } { E T2h, T2k, T2g, T2j, T8b, T2i, T2n; T2h = ri[WS(rs, 42)]; T2k = ii[WS(rs, 42)]; T8j = FNMS(T2w, T2u, T8i); T2y = FMA(T2w, T2x, T2v); T2g = W[82]; T2j = W[83]; T2o = ri[WS(rs, 58)]; T2r = ii[WS(rs, 58)]; T8b = T2g * T2k; T2i = T2g * T2h; T2n = W[114]; T2q = W[115]; T8c = FNMS(T2j, T2h, T8b); T2l = FMA(T2j, T2k, T2i); T8g = T2n * T2r; T2p = T2n * T2o; } } { E T8d, Tf9, T2m, T88, T8h, T2s, Tfa, T2z; T8d = T8a - T8c; Tf9 = T8a + T8c; T2m = T2f + T2l; T88 = T2f - T2l; T8h = FNMS(T2q, T2o, T8g); T2s = FMA(T2q, T2r, T2p); T8k = T8h - T8j; Tfa = T8h + T8j; T2z = T2s + T2y; T8f = T2s - T2y; T8w = T8d - T88; T8e = T88 + T8d; Thw = Tf9 + Tfa; Tfb = Tf9 - Tfa; Tf6 = T2z - T2m; T2A = T2m + T2z; } } { E T38, T8J, T3h, T3k, T8L, T3e, T3g, T3j, T8P, T3i; { E T3n, T3q, T3m, T3p; { E T34, T37, T33, T8v, T8l, T36, T8I, T35; T34 = ri[WS(rs, 6)]; T37 = ii[WS(rs, 6)]; T33 = W[10]; T8v = T8f + T8k; T8l = T8f - T8k; T36 = W[11]; T8I = T33 * T37; T35 = T33 * T34; T8x = T8v - T8w; TcO = T8w + T8v; T8m = T8e - T8l; TcR = T8e + T8l; T38 = FMA(T36, T37, T35); T8J = FNMS(T36, T34, T8I); } T3n = ri[WS(rs, 22)]; T3q = ii[WS(rs, 22)]; T3m = W[42]; T3p = W[43]; { E T3a, T3d, T3c, T8K, T3b, T8R, T3o, T39; T3a = ri[WS(rs, 38)]; T3d = ii[WS(rs, 38)]; T8R = T3m * T3q; T3o = T3m * T3n; T39 = W[74]; T3c = W[75]; T8S = FNMS(T3p, T3n, T8R); T3r = FMA(T3p, T3q, T3o); T8K = T39 * T3d; T3b = T39 * T3a; T3h = ri[WS(rs, 54)]; T3k = ii[WS(rs, 54)]; T8L = FNMS(T3c, T3a, T8K); T3e = FMA(T3c, T3d, T3b); T3g = W[106]; T3j = W[107]; } } T8M = T8J - T8L; Tfk = T8J + T8L; T3f = T38 + T3e; T8H = T38 - T3e; T8P = T3g * T3k; T3i = T3g * T3h; T8Q = FNMS(T3j, T3h, T8P); T3l = FMA(T3j, T3k, T3i); } } { E T9u, T9p, Tac, T9o; { E T9k, T43, T9t, T4m, T4c, T4f, T4e, T9m, T49, T9q, T4d; { E T4i, T4l, T4k, T9s, T4j; { E T3Z, T42, T3Y, T41, T9j, T40, T4h; { E T95, T8N, T8T, Tfl, T8O, T3s, T8U, T94; T3Z = ri[WS(rs, 9)]; T95 = T8M - T8H; T8N = T8H + T8M; T8T = T8Q - T8S; Tfl = T8Q + T8S; T8O = T3l - T3r; T3s = T3l + T3r; T42 = ii[WS(rs, 9)]; Tfm = Tfk - Tfl; ThC = Tfk + Tfl; T8U = T8O - T8T; T94 = T8O + T8T; T3t = T3f + T3s; Tfh = T3s - T3f; T96 = T94 - T95; TcV = T95 + T94; T8V = T8N - T8U; TcY = T8N + T8U; T3Y = W[16]; } T41 = W[17]; T4i = ri[WS(rs, 25)]; T4l = ii[WS(rs, 25)]; T9j = T3Y * T42; T40 = T3Y * T3Z; T4h = W[48]; T4k = W[49]; T9k = FNMS(T41, T3Z, T9j); T43 = FMA(T41, T42, T40); T9s = T4h * T4l; T4j = T4h * T4i; } { E T45, T48, T44, T47, T9l, T46, T4b; T45 = ri[WS(rs, 41)]; T48 = ii[WS(rs, 41)]; T9t = FNMS(T4k, T4i, T9s); T4m = FMA(T4k, T4l, T4j); T44 = W[80]; T47 = W[81]; T4c = ri[WS(rs, 57)]; T4f = ii[WS(rs, 57)]; T9l = T44 * T48; T46 = T44 * T45; T4b = W[112]; T4e = W[113]; T9m = FNMS(T47, T45, T9l); T49 = FMA(T47, T48, T46); T9q = T4b * T4f; T4d = T4b * T4c; } } { E T9n, TfJ, T4a, T9i, T9r, T4g, TfK, T4n; T9n = T9k - T9m; TfJ = T9k + T9m; T4a = T43 + T49; T9i = T43 - T49; T9r = FNMS(T4e, T4c, T9q); T4g = FMA(T4e, T4f, T4d); T9u = T9r - T9t; TfK = T9r + T9t; T4n = T4g + T4m; T9p = T4g - T4m; Tac = T9n - T9i; T9o = T9i + T9n; ThI = TfJ + TfK; TfL = TfJ - TfK; Tfu = T4n - T4a; T4o = T4a + T4n; } } { E T5Q, Tat, T5Z, T62, Tav, T5W, T5Y, T61, Taz, T60; { E T65, T68, T64, T67; { E T5M, T5P, T5L, Tab, T9v, T5O, Tas, T5N; T5M = ri[WS(rs, 7)]; T5P = ii[WS(rs, 7)]; T5L = W[12]; Tab = T9p + T9u; T9v = T9p - T9u; T5O = W[13]; Tas = T5L * T5P; T5N = T5L * T5M; Tad = Tab - Tac; Td4 = Tac + Tab; T9w = T9o - T9v; Tdf = T9o + T9v; T5Q = FMA(T5O, T5P, T5N); Tat = FNMS(T5O, T5M, Tas); } T65 = ri[WS(rs, 23)]; T68 = ii[WS(rs, 23)]; T64 = W[44]; T67 = W[45]; { E T5S, T5V, T5U, Tau, T5T, TaB, T66, T5R; T5S = ri[WS(rs, 39)]; T5V = ii[WS(rs, 39)]; TaB = T64 * T68; T66 = T64 * T65; T5R = W[76]; T5U = W[77]; TaC = FNMS(T67, T65, TaB); T69 = FMA(T67, T68, T66); Tau = T5R * T5V; T5T = T5R * T5S; T5Z = ri[WS(rs, 55)]; T62 = ii[WS(rs, 55)]; Tav = FNMS(T5U, T5S, Tau); T5W = FMA(T5U, T5V, T5T); T5Y = W[108]; T61 = W[109]; } } Taw = Tat - Tav; Tga = Tat + Tav; T5X = T5Q + T5W; Tar = T5Q - T5W; Taz = T5Y * T62; T60 = T5Y * T5Z; TaA = FNMS(T61, T5Z, Taz); T63 = FMA(T61, T62, T60); } } } { E T9E, Tda, TfE, TfB, Td9, T9L; { E T9T, Td7, Tfy, Tfz, Td6, Ta0; { E T9V, T4v, T9R, T4O, T4E, T4H, T4G, T9X, T4B, T9O, T4F; { E T4K, T4N, T4M, T9Q, T4L; { E T4r, T4u, T4q, T4t, T9U, T4s, T4J; { E Tbl, Tax, TaD, Tgb, Tay, T6a, TaE, Tbk; T4r = ri[WS(rs, 5)]; Tbl = Taw - Tar; Tax = Tar + Taw; TaD = TaA - TaC; Tgb = TaA + TaC; Tay = T63 - T69; T6a = T63 + T69; T4u = ii[WS(rs, 5)]; Tgc = Tga - Tgb; ThT = Tga + Tgb; TaE = Tay - TaD; Tbk = Tay + TaD; T6b = T5X + T6a; TfV = T6a - T5X; Tbm = Tbk - Tbl; Tdn = Tbl + Tbk; TaF = Tax - TaE; Tdy = Tax + TaE; T4q = W[8]; } T4t = W[9]; T4K = ri[WS(rs, 53)]; T4N = ii[WS(rs, 53)]; T9U = T4q * T4u; T4s = T4q * T4r; T4J = W[104]; T4M = W[105]; T9V = FNMS(T4t, T4r, T9U); T4v = FMA(T4t, T4u, T4s); T9Q = T4J * T4N; T4L = T4J * T4K; } { E T4x, T4A, T4w, T4z, T9W, T4y, T4D; T4x = ri[WS(rs, 37)]; T4A = ii[WS(rs, 37)]; T9R = FNMS(T4M, T4K, T9Q); T4O = FMA(T4M, T4N, T4L); T4w = W[72]; T4z = W[73]; T4E = ri[WS(rs, 21)]; T4H = ii[WS(rs, 21)]; T9W = T4w * T4A; T4y = T4w * T4x; T4D = W[40]; T4G = W[41]; T9X = FNMS(T4z, T4x, T9W); T4B = FMA(T4z, T4A, T4y); T9O = T4D * T4H; T4F = T4D * T4E; } } { E T9Y, Tfw, T4C, T9N, T9P, T4I; T9Y = T9V - T9X; Tfw = T9V + T9X; T4C = T4v + T4B; T9N = T4v - T4B; T9P = FNMS(T4G, T4E, T9O); T4I = FMA(T4G, T4H, T4F); { E Tfx, T9S, T9Z, T4P; Tfx = T9P + T9R; T9S = T9P - T9R; T9Z = T4I - T4O; T4P = T4I + T4O; T9T = T9N - T9S; Td7 = T9N + T9S; Tfy = Tfw - Tfx; ThN = Tfw + Tfx; Tfz = T4C - T4P; T4Q = T4C + T4P; Td6 = T9Y - T9Z; Ta0 = T9Y + T9Z; } } } { E T9G, T4W, T9C, T5f, T55, T58, T57, T9I, T52, T9z, T56; { E T5b, T5e, T5d, T9B, T5c; { E T4S, T4V, T4R, T4U, T9F, T4T, T5a; T4S = ri[WS(rs, 61)]; TfN = Tfz + Tfy; TfA = Tfy - Tfz; Taf = FMA(KP414213562, T9T, Ta0); Ta1 = FNMS(KP414213562, Ta0, T9T); Td8 = FNMS(KP414213562, Td7, Td6); Tdh = FMA(KP414213562, Td6, Td7); T4V = ii[WS(rs, 61)]; T4R = W[120]; T4U = W[121]; T5b = ri[WS(rs, 45)]; T5e = ii[WS(rs, 45)]; T9F = T4R * T4V; T4T = T4R * T4S; T5a = W[88]; T5d = W[89]; T9G = FNMS(T4U, T4S, T9F); T4W = FMA(T4U, T4V, T4T); T9B = T5a * T5e; T5c = T5a * T5b; } { E T4Y, T51, T4X, T50, T9H, T4Z, T54; T4Y = ri[WS(rs, 29)]; T51 = ii[WS(rs, 29)]; T9C = FNMS(T5d, T5b, T9B); T5f = FMA(T5d, T5e, T5c); T4X = W[56]; T50 = W[57]; T55 = ri[WS(rs, 13)]; T58 = ii[WS(rs, 13)]; T9H = T4X * T51; T4Z = T4X * T4Y; T54 = W[24]; T57 = W[25]; T9I = FNMS(T50, T4Y, T9H); T52 = FMA(T50, T51, T4Z); T9z = T54 * T58; T56 = T54 * T55; } } { E T9J, TfC, T53, T9y, T9A, T59; T9J = T9G - T9I; TfC = T9G + T9I; T53 = T4W + T52; T9y = T4W - T52; T9A = FNMS(T57, T55, T9z); T59 = FMA(T57, T58, T56); { E TfD, T9D, T9K, T5g; TfD = T9A + T9C; T9D = T9A - T9C; T9K = T59 - T5f; T5g = T59 + T5f; T9E = T9y - T9D; Tda = T9y + T9D; TfE = TfC - TfD; ThO = TfC + TfD; TfB = T53 - T5g; T5h = T53 + T5g; Td9 = T9J - T9K; T9L = T9J + T9K; } } } } { E Tb2, Tdq, TfZ, Tg0, Tdp, Tb9; { E Tb4, T6i, Tb0, T6B, T6r, T6u, T6t, Tb6, T6o, TaX, T6s; { E T6x, T6A, T6z, TaZ, T6y; { E T6e, T6h, T6d, T6g, Tb3, T6f, T6w; T6e = ri[WS(rs, 3)]; TfO = TfB - TfE; TfF = TfB + TfE; Tag = FNMS(KP414213562, T9E, T9L); T9M = FMA(KP414213562, T9L, T9E); Tdb = FMA(KP414213562, Tda, Td9); Tdi = FNMS(KP414213562, Td9, Tda); T6h = ii[WS(rs, 3)]; T6d = W[4]; T6g = W[5]; T6x = ri[WS(rs, 51)]; T6A = ii[WS(rs, 51)]; Tb3 = T6d * T6h; T6f = T6d * T6e; T6w = W[100]; T6z = W[101]; Tb4 = FNMS(T6g, T6e, Tb3); T6i = FMA(T6g, T6h, T6f); TaZ = T6w * T6A; T6y = T6w * T6x; } { E T6k, T6n, T6j, T6m, Tb5, T6l, T6q; T6k = ri[WS(rs, 35)]; T6n = ii[WS(rs, 35)]; Tb0 = FNMS(T6z, T6x, TaZ); T6B = FMA(T6z, T6A, T6y); T6j = W[68]; T6m = W[69]; T6r = ri[WS(rs, 19)]; T6u = ii[WS(rs, 19)]; Tb5 = T6j * T6n; T6l = T6j * T6k; T6q = W[36]; T6t = W[37]; Tb6 = FNMS(T6m, T6k, Tb5); T6o = FMA(T6m, T6n, T6l); TaX = T6q * T6u; T6s = T6q * T6r; } } { E Tb7, TfX, T6p, TaW, TaY, T6v; Tb7 = Tb4 - Tb6; TfX = Tb4 + Tb6; T6p = T6i + T6o; TaW = T6i - T6o; TaY = FNMS(T6t, T6r, TaX); T6v = FMA(T6t, T6u, T6s); { E TfY, Tb1, Tb8, T6C; TfY = TaY + Tb0; Tb1 = TaY - Tb0; Tb8 = T6v - T6B; T6C = T6v + T6B; Tb2 = TaW - Tb1; Tdq = TaW + Tb1; TfZ = TfX - TfY; ThY = TfX + TfY; Tg0 = T6p - T6C; T6D = T6p + T6C; Tdp = Tb7 - Tb8; Tb9 = Tb7 + Tb8; } } } { E TaP, T6J, TaL, T72, T6S, T6V, T6U, TaR, T6P, TaI, T6T; { E T6Y, T71, T70, TaK, T6Z; { E T6F, T6I, T6E, T6H, TaO, T6G, T6X; T6F = ri[WS(rs, 59)]; Tge = Tg0 + TfZ; Tg1 = TfZ - Tg0; Tbo = FMA(KP414213562, Tb2, Tb9); Tba = FNMS(KP414213562, Tb9, Tb2); Tdr = FNMS(KP414213562, Tdq, Tdp); TdA = FMA(KP414213562, Tdp, Tdq); T6I = ii[WS(rs, 59)]; T6E = W[116]; T6H = W[117]; T6Y = ri[WS(rs, 43)]; T71 = ii[WS(rs, 43)]; TaO = T6E * T6I; T6G = T6E * T6F; T6X = W[84]; T70 = W[85]; TaP = FNMS(T6H, T6F, TaO); T6J = FMA(T6H, T6I, T6G); TaK = T6X * T71; T6Z = T6X * T6Y; } { E T6L, T6O, T6K, T6N, TaQ, T6M, T6R; T6L = ri[WS(rs, 27)]; T6O = ii[WS(rs, 27)]; TaL = FNMS(T70, T6Y, TaK); T72 = FMA(T70, T71, T6Z); T6K = W[52]; T6N = W[53]; T6S = ri[WS(rs, 11)]; T6V = ii[WS(rs, 11)]; TaQ = T6K * T6O; T6M = T6K * T6L; T6R = W[20]; T6U = W[21]; TaR = FNMS(T6N, T6L, TaQ); T6P = FMA(T6N, T6O, T6M); TaI = T6R * T6V; T6T = T6R * T6S; } } { E TaS, Tg3, T6Q, TaH, TaJ, T6W; TaS = TaP - TaR; Tg3 = TaP + TaR; T6Q = T6J + T6P; TaH = T6J - T6P; TaJ = FNMS(T6U, T6S, TaI); T6W = FMA(T6U, T6V, T6T); { E Tg4, TaM, TaT, T73; Tg4 = TaJ + TaL; TaM = TaJ - TaL; TaT = T6W - T72; T73 = T6W + T72; TaN = TaH - TaM; Tdt = TaH + TaM; Tg5 = Tg3 - Tg4; ThZ = Tg3 + Tg4; Tg2 = T6Q - T73; T74 = T6Q + T73; Tds = TaS - TaT; TaU = TaS + TaT; } } } } } } { E Tgf, Tg6, Tbp, TaV, Tdu, TdB, Tje, Tjd, TjO, TjN; { E Thq, Tj7, Thy, ThA, Tht, Tj8, Thx, ThD, ThX, ThV, ThU, Ti0, ThM, ThK, ThJ; E ThP, TiI, TiZ, TiL, Tj0; { E Tio, T1I, Tj1, T3v, Tj2, TiX, TiN, Tir, T76, TiK, TiC, TiG, T5j, Tit, Tiw; E TiJ; { E TiO, TiW, Tip, Tiq; { E TO, T1H, T2B, T3u; Thq = Tm - TN; TO = Tm + TN; Tgf = Tg2 - Tg5; Tg6 = Tg2 + Tg5; Tbp = FNMS(KP414213562, TaN, TaU); TaV = FMA(KP414213562, TaU, TaN); Tdu = FMA(KP414213562, Tdt, Tds); TdB = FNMS(KP414213562, Tds, Tdt); T1H = T1f + T1G; Tj7 = T1G - T1f; Thy = T29 - T2A; T2B = T29 + T2A; T3u = T32 + T3t; ThA = T32 - T3t; Tht = Thr - Ths; TiO = Thr + Ths; Tio = TO - T1H; T1I = TO + T1H; Tj1 = T3u - T2B; T3v = T2B + T3u; TiW = TiP + TiV; Tj8 = TiV - TiP; } Thx = Thv - Thw; Tip = Thv + Thw; Tiq = ThB + ThC; ThD = ThB - ThC; { E T6c, T75, Tiz, TiA; ThX = T5K - T6b; T6c = T5K + T6b; Tj2 = TiW - TiO; TiX = TiO + TiW; TiN = Tip + Tiq; Tir = Tip - Tiq; T75 = T6D + T74; ThV = T74 - T6D; ThU = ThS - ThT; Tiz = ThS + ThT; TiA = ThY + ThZ; Ti0 = ThY - ThZ; { E T4p, Tiy, TiB, T5i, Tiu, Tiv; ThM = T3X - T4o; T4p = T3X + T4o; T76 = T6c + T75; Tiy = T6c - T75; TiK = Tiz + TiA; TiB = Tiz - TiA; T5i = T4Q + T5h; ThK = T5h - T4Q; ThJ = ThH - ThI; Tiu = ThH + ThI; Tiv = ThN + ThO; ThP = ThN - ThO; TiC = Tiy - TiB; TiG = Tiy + TiB; T5j = T4p + T5i; Tit = T4p - T5i; Tiw = Tiu - Tiv; TiJ = Tiu + Tiv; } } } { E TiE, Tis, TiD, Tj6, Tj5, Tj3, Tj4, TiH; { E T3w, TiF, Tix, T77, TiM, TiY; TiI = T1I - T3v; T3w = T1I + T3v; TiF = Tiw - Tit; Tix = Tit + Tiw; T77 = T5j + T76; TiZ = T76 - T5j; TiL = TiJ - TiK; TiM = TiJ + TiK; TiY = TiN + TiX; Tj0 = TiX - TiN; TiE = Tio - Tir; Tis = Tio + Tir; ri[0] = T3w + T77; ri[WS(rs, 32)] = T3w - T77; ii[WS(rs, 32)] = TiY - TiM; ii[0] = TiM + TiY; TiD = Tix + TiC; Tj6 = TiC - Tix; Tj5 = Tj2 - Tj1; Tj3 = Tj1 + Tj2; Tj4 = TiF + TiG; TiH = TiF - TiG; } ri[WS(rs, 8)] = FMA(KP707106781, TiD, Tis); ri[WS(rs, 40)] = FNMS(KP707106781, TiD, Tis); ii[WS(rs, 40)] = FNMS(KP707106781, Tj4, Tj3); ii[WS(rs, 8)] = FMA(KP707106781, Tj4, Tj3); ri[WS(rs, 24)] = FMA(KP707106781, TiH, TiE); ri[WS(rs, 56)] = FNMS(KP707106781, TiH, TiE); ii[WS(rs, 56)] = FNMS(KP707106781, Tj6, Tj5); ii[WS(rs, 24)] = FMA(KP707106781, Tj6, Tj5); } } { E Ti8, Thu, Tjf, Tj9, Tib, Tjg, Tja, ThF, Tih, ThW, Tif, Til, Ti5, ThR; ri[WS(rs, 16)] = TiI + TiL; ri[WS(rs, 48)] = TiI - TiL; ii[WS(rs, 48)] = Tj0 - TiZ; ii[WS(rs, 16)] = TiZ + Tj0; Ti8 = Thq + Tht; Thu = Thq - Tht; Tjf = Tj8 - Tj7; Tj9 = Tj7 + Tj8; { E Tie, ThL, Tid, ThQ; { E Ti9, Thz, Tia, ThE; Ti9 = Thy + Thx; Thz = Thx - Thy; Tia = ThA - ThD; ThE = ThA + ThD; Tib = Ti9 + Tia; Tjg = Tia - Ti9; Tja = Thz + ThE; ThF = Thz - ThE; Tie = ThJ + ThK; ThL = ThJ - ThK; } Tid = ThM + ThP; ThQ = ThM - ThP; Tih = ThU + ThV; ThW = ThU - ThV; Tif = FMA(KP414213562, Tie, Tid); Til = FNMS(KP414213562, Tid, Tie); Ti5 = FNMS(KP414213562, ThL, ThQ); ThR = FMA(KP414213562, ThQ, ThL); } { E Ti4, ThG, Tjh, Tjj, Tig, Ti1; Ti4 = FNMS(KP707106781, ThF, Thu); ThG = FMA(KP707106781, ThF, Thu); Tjh = FMA(KP707106781, Tjg, Tjf); Tjj = FNMS(KP707106781, Tjg, Tjf); Tig = ThX + Ti0; Ti1 = ThX - Ti0; { E Tik, Tjb, Tjc, Tin; { E Tic, Tim, Ti6, Ti2, Tij, Tii; Tik = FNMS(KP707106781, Tib, Ti8); Tic = FMA(KP707106781, Tib, Ti8); Tii = FNMS(KP414213562, Tih, Tig); Tim = FMA(KP414213562, Tig, Tih); Ti6 = FMA(KP414213562, ThW, Ti1); Ti2 = FNMS(KP414213562, Ti1, ThW); Tij = Tif + Tii; Tje = Tii - Tif; Tjd = FNMS(KP707106781, Tja, Tj9); Tjb = FMA(KP707106781, Tja, Tj9); { E Ti7, Tji, Tjk, Ti3; Ti7 = Ti5 + Ti6; Tji = Ti6 - Ti5; Tjk = ThR + Ti2; Ti3 = ThR - Ti2; ri[WS(rs, 4)] = FMA(KP923879532, Tij, Tic); ri[WS(rs, 36)] = FNMS(KP923879532, Tij, Tic); ri[WS(rs, 60)] = FMA(KP923879532, Ti7, Ti4); ri[WS(rs, 28)] = FNMS(KP923879532, Ti7, Ti4); ii[WS(rs, 44)] = FNMS(KP923879532, Tji, Tjh); ii[WS(rs, 12)] = FMA(KP923879532, Tji, Tjh); ii[WS(rs, 60)] = FMA(KP923879532, Tjk, Tjj); ii[WS(rs, 28)] = FNMS(KP923879532, Tjk, Tjj); ri[WS(rs, 12)] = FMA(KP923879532, Ti3, ThG); ri[WS(rs, 44)] = FNMS(KP923879532, Ti3, ThG); Tjc = Til + Tim; Tin = Til - Tim; } } ii[WS(rs, 36)] = FNMS(KP923879532, Tjc, Tjb); ii[WS(rs, 4)] = FMA(KP923879532, Tjc, Tjb); ri[WS(rs, 20)] = FMA(KP923879532, Tin, Tik); ri[WS(rs, 52)] = FNMS(KP923879532, Tin, Tik); } } } } { E TjD, TjJ, Tgo, Tf2, Tjp, Tjv, Tha, TgI, Tgd, Tgr, Tjw, Tjq, Tfp, Tgg, Thk; E Tho, Th8, Th4, Tgv, TgB, Tgl, TfR, TjE, Thd, TjK, TgP, Tgx, Tg8, Thh, Thn; E Th7, TgX; { E TgJ, TgK, TgM, TgN, Tg7, TfW, Th1, Thj, Th0, Th2; { E TgE, TeQ, TjB, Tjn, TgF, TgG, TjC, Tf1, TeV, Tf0; TgE = TeM - TeP; TeQ = TeM + TeP; TjB = Tjm - Tjl; Tjn = Tjl + Tjm; TgF = TeU - TeR; TeV = TeR + TeU; ii[WS(rs, 52)] = FNMS(KP923879532, Tje, Tjd); ii[WS(rs, 20)] = FMA(KP923879532, Tje, Tjd); Tf0 = TeW - TeZ; TgG = TeW + TeZ; TjC = Tf0 - TeV; Tf1 = TeV + Tf0; { E Tfi, Tgp, Tfd, Tfn; { E Tf7, Tjo, TgH, Tfc; TgJ = Tf5 - Tf6; Tf7 = Tf5 + Tf6; TjD = FMA(KP707106781, TjC, TjB); TjJ = FNMS(KP707106781, TjC, TjB); Tgo = FMA(KP707106781, Tf1, TeQ); Tf2 = FNMS(KP707106781, Tf1, TeQ); Tjo = TgF + TgG; TgH = TgF - TgG; Tfc = Tf8 + Tfb; TgK = Tf8 - Tfb; TgM = Tfg - Tfh; Tfi = Tfg + Tfh; Tjp = FMA(KP707106781, Tjo, Tjn); Tjv = FNMS(KP707106781, Tjo, Tjn); Tha = FNMS(KP707106781, TgH, TgE); TgI = FMA(KP707106781, TgH, TgE); Tgp = FMA(KP414213562, Tf7, Tfc); Tfd = FNMS(KP414213562, Tfc, Tf7); Tfn = Tfj + Tfm; TgN = Tfj - Tfm; } { E TgY, TgZ, Tgq, Tfo; Tgd = Tg9 + Tgc; TgY = Tg9 - Tgc; TgZ = Tg6 - Tg1; Tg7 = Tg1 + Tg6; TfW = TfU + TfV; Th1 = TfU - TfV; Tgq = FNMS(KP414213562, Tfi, Tfn); Tfo = FMA(KP414213562, Tfn, Tfi); Thj = FMA(KP707106781, TgZ, TgY); Th0 = FNMS(KP707106781, TgZ, TgY); Tgr = Tgp + Tgq; Tjw = Tgq - Tgp; Tjq = Tfd + Tfo; Tfp = Tfd - Tfo; Th2 = Tge - Tgf; Tgg = Tge + Tgf; } } } { E TgU, TgS, TgR, TgV, Thb, TgL; { E TfM, Tgu, TfH, TfP, Tgt, TfQ; { E Tfv, TfG, Thi, Th3; TgU = Tft - Tfu; Tfv = Tft + Tfu; TfG = TfA + TfF; TgS = TfF - TfA; TgR = TfI - TfL; TfM = TfI + TfL; Thi = FMA(KP707106781, Th2, Th1); Th3 = FNMS(KP707106781, Th2, Th1); Tgu = FMA(KP707106781, TfG, Tfv); TfH = FNMS(KP707106781, TfG, Tfv); Thk = FNMS(KP198912367, Thj, Thi); Tho = FMA(KP198912367, Thi, Thj); Th8 = FMA(KP668178637, Th0, Th3); Th4 = FNMS(KP668178637, Th3, Th0); TfP = TfN + TfO; TgV = TfN - TfO; } Tgt = FMA(KP707106781, TfP, TfM); TfQ = FNMS(KP707106781, TfP, TfM); Thb = FNMS(KP414213562, TgJ, TgK); TgL = FMA(KP414213562, TgK, TgJ); Tgv = FMA(KP198912367, Tgu, Tgt); TgB = FNMS(KP198912367, Tgt, Tgu); Tgl = FNMS(KP668178637, TfH, TfQ); TfR = FMA(KP668178637, TfQ, TfH); } { E Thg, TgT, Thc, TgO, Thf, TgW; Thc = FMA(KP414213562, TgM, TgN); TgO = FNMS(KP414213562, TgN, TgM); Thg = FMA(KP707106781, TgS, TgR); TgT = FNMS(KP707106781, TgS, TgR); TjE = Thc - Thb; Thd = Thb + Thc; TjK = TgL + TgO; TgP = TgL - TgO; Thf = FMA(KP707106781, TgV, TgU); TgW = FNMS(KP707106781, TgV, TgU); Tgx = FMA(KP707106781, Tg7, TfW); Tg8 = FNMS(KP707106781, Tg7, TfW); Thh = FMA(KP198912367, Thg, Thf); Thn = FNMS(KP198912367, Thf, Thg); Th7 = FNMS(KP668178637, TgT, TgW); TgX = FMA(KP668178637, TgW, TgT); } } } { E Tju, Tjt, TjI, TjH; { E Tgk, Tfq, Tjx, Tjz, Tgw, Tgh; Tgk = FNMS(KP923879532, Tfp, Tf2); Tfq = FMA(KP923879532, Tfp, Tf2); Tjx = FMA(KP923879532, Tjw, Tjv); Tjz = FNMS(KP923879532, Tjw, Tjv); Tgw = FMA(KP707106781, Tgg, Tgd); Tgh = FNMS(KP707106781, Tgg, Tgd); { E TgA, Tjr, Tjs, TgD; { E Tgs, TgC, Tgm, Tgi, Tgz, Tgy; TgA = FNMS(KP923879532, Tgr, Tgo); Tgs = FMA(KP923879532, Tgr, Tgo); Tgy = FNMS(KP198912367, Tgx, Tgw); TgC = FMA(KP198912367, Tgw, Tgx); Tgm = FMA(KP668178637, Tg8, Tgh); Tgi = FNMS(KP668178637, Tgh, Tg8); Tgz = Tgv + Tgy; Tju = Tgy - Tgv; Tjt = FNMS(KP923879532, Tjq, Tjp); Tjr = FMA(KP923879532, Tjq, Tjp); { E Tgn, Tjy, TjA, Tgj; Tgn = Tgl + Tgm; Tjy = Tgm - Tgl; TjA = TfR + Tgi; Tgj = TfR - Tgi; ri[WS(rs, 2)] = FMA(KP980785280, Tgz, Tgs); ri[WS(rs, 34)] = FNMS(KP980785280, Tgz, Tgs); ri[WS(rs, 58)] = FMA(KP831469612, Tgn, Tgk); ri[WS(rs, 26)] = FNMS(KP831469612, Tgn, Tgk); ii[WS(rs, 42)] = FNMS(KP831469612, Tjy, Tjx); ii[WS(rs, 10)] = FMA(KP831469612, Tjy, Tjx); ii[WS(rs, 58)] = FMA(KP831469612, TjA, Tjz); ii[WS(rs, 26)] = FNMS(KP831469612, TjA, Tjz); ri[WS(rs, 10)] = FMA(KP831469612, Tgj, Tfq); ri[WS(rs, 42)] = FNMS(KP831469612, Tgj, Tfq); Tjs = TgB + TgC; TgD = TgB - TgC; } } ii[WS(rs, 34)] = FNMS(KP980785280, Tjs, Tjr); ii[WS(rs, 2)] = FMA(KP980785280, Tjs, Tjr); ri[WS(rs, 18)] = FMA(KP980785280, TgD, TgA); ri[WS(rs, 50)] = FNMS(KP980785280, TgD, TgA); } } { E Th6, TjF, TjG, Th9, TgQ, Th5; Th6 = FNMS(KP923879532, TgP, TgI); TgQ = FMA(KP923879532, TgP, TgI); Th5 = TgX + Th4; TjI = Th4 - TgX; TjH = FNMS(KP923879532, TjE, TjD); TjF = FMA(KP923879532, TjE, TjD); ii[WS(rs, 50)] = FNMS(KP980785280, Tju, Tjt); ii[WS(rs, 18)] = FMA(KP980785280, Tju, Tjt); ri[WS(rs, 6)] = FMA(KP831469612, Th5, TgQ); ri[WS(rs, 38)] = FNMS(KP831469612, Th5, TgQ); TjG = Th7 + Th8; Th9 = Th7 - Th8; ii[WS(rs, 38)] = FNMS(KP831469612, TjG, TjF); ii[WS(rs, 6)] = FMA(KP831469612, TjG, TjF); ri[WS(rs, 22)] = FMA(KP831469612, Th9, Th6); ri[WS(rs, 54)] = FNMS(KP831469612, Th9, Th6); } { E Thm, TjL, TjM, Thp, The, Thl; Thm = FMA(KP923879532, Thd, Tha); The = FNMS(KP923879532, Thd, Tha); Thl = Thh - Thk; TjO = Thh + Thk; TjN = FMA(KP923879532, TjK, TjJ); TjL = FNMS(KP923879532, TjK, TjJ); ii[WS(rs, 54)] = FNMS(KP831469612, TjI, TjH); ii[WS(rs, 22)] = FMA(KP831469612, TjI, TjH); ri[WS(rs, 14)] = FMA(KP980785280, Thl, The); ri[WS(rs, 46)] = FNMS(KP980785280, Thl, The); TjM = Tho - Thn; Thp = Thn + Tho; ii[WS(rs, 46)] = FNMS(KP980785280, TjM, TjL); ii[WS(rs, 14)] = FMA(KP980785280, TjM, TjL); ri[WS(rs, 62)] = FMA(KP980785280, Thp, Thm); ri[WS(rs, 30)] = FNMS(KP980785280, Thp, Thm); } } } { E TjS, TcD, Tcw, TkO, TkN, Tcz; { E TbB, Tkw, Tkq, T99, TbF, TbL, Tbv, Taj, Tcu, Tcy, Tci, Tce, Tcr, Tcx, Tch; E Tc7, TkE, Tcn, TkK, TbZ, TbP, T7J, TbO, T7u, TkB, Tkn, TbI, TbM, Tbw, Tbs; E T7Y, TbQ; { E TbT, TbU, TbW, TbX, Tc4, Tc2, Tc1, Tc5, Tbn, Tbb, TaG, Tcb, Tct, Tca, Tcc; E Tbq, Tcl, TbV; { E T8W, Tbz, T8z, T97, T8n, T8y; TbT = FMA(KP707106781, T8m, T87); T8n = FNMS(KP707106781, T8m, T87); T8y = FNMS(KP707106781, T8x, T8u); TbU = FMA(KP707106781, T8x, T8u); TbW = FMA(KP707106781, T8V, T8G); T8W = FNMS(KP707106781, T8V, T8G); ii[WS(rs, 62)] = FMA(KP980785280, TjO, TjN); ii[WS(rs, 30)] = FNMS(KP980785280, TjO, TjN); Tbz = FMA(KP668178637, T8n, T8y); T8z = FNMS(KP668178637, T8y, T8n); T97 = FNMS(KP707106781, T96, T93); TbX = FMA(KP707106781, T96, T93); { E Tae, TbE, Ta3, Tah; { E T9x, Ta2, TbA, T98; Tc4 = FMA(KP707106781, T9w, T9h); T9x = FNMS(KP707106781, T9w, T9h); Ta2 = T9M - Ta1; Tc2 = Ta1 + T9M; Tc1 = FMA(KP707106781, Tad, Taa); Tae = FNMS(KP707106781, Tad, Taa); TbA = FNMS(KP668178637, T8W, T97); T98 = FMA(KP668178637, T97, T8W); TbE = FMA(KP923879532, Ta2, T9x); Ta3 = FNMS(KP923879532, Ta2, T9x); TbB = Tbz + TbA; Tkw = TbA - Tbz; Tkq = T8z + T98; T99 = T8z - T98; Tah = Taf - Tag; Tc5 = Taf + Tag; } { E Tc8, Tc9, TbD, Tai; Tbn = FNMS(KP707106781, Tbm, Tbj); Tc8 = FMA(KP707106781, Tbm, Tbj); Tc9 = Tba + TaV; Tbb = TaV - Tba; TaG = FNMS(KP707106781, TaF, Taq); Tcb = FMA(KP707106781, TaF, Taq); TbD = FMA(KP923879532, Tah, Tae); Tai = FNMS(KP923879532, Tah, Tae); Tct = FMA(KP923879532, Tc9, Tc8); Tca = FNMS(KP923879532, Tc9, Tc8); TbF = FMA(KP303346683, TbE, TbD); TbL = FNMS(KP303346683, TbD, TbE); Tbv = FNMS(KP534511135, Ta3, Tai); Taj = FMA(KP534511135, Tai, Ta3); Tcc = Tbo + Tbp; Tbq = Tbo - Tbp; } } } { E Tcq, Tc3, Tcs, Tcd, Tcp, Tc6; Tcs = FMA(KP923879532, Tcc, Tcb); Tcd = FNMS(KP923879532, Tcc, Tcb); Tcq = FMA(KP923879532, Tc2, Tc1); Tc3 = FNMS(KP923879532, Tc2, Tc1); Tcu = FNMS(KP098491403, Tct, Tcs); Tcy = FMA(KP098491403, Tcs, Tct); Tci = FMA(KP820678790, Tca, Tcd); Tce = FNMS(KP820678790, Tcd, Tca); Tcp = FMA(KP923879532, Tc5, Tc4); Tc6 = FNMS(KP923879532, Tc5, Tc4); Tcl = FNMS(KP198912367, TbT, TbU); TbV = FMA(KP198912367, TbU, TbT); Tcr = FMA(KP098491403, Tcq, Tcp); Tcx = FNMS(KP098491403, Tcp, Tcq); Tch = FNMS(KP820678790, Tc3, Tc6); Tc7 = FMA(KP820678790, Tc6, Tc3); } { E TbH, Tbc, Tcm, TbY; Tcm = FMA(KP198912367, TbW, TbX); TbY = FNMS(KP198912367, TbX, TbW); TbH = FMA(KP923879532, Tbb, TaG); Tbc = FNMS(KP923879532, Tbb, TaG); TkE = Tcm - Tcl; Tcn = Tcl + Tcm; TkK = TbV + TbY; TbZ = TbV - TbY; { E T7t, Tkm, TbG, Tbr; TjS = T7l + T7s; T7t = T7l - T7s; Tkm = TcC - TcB; TcD = TcB + TcC; TbP = FNMS(KP414213562, T7B, T7I); T7J = FMA(KP414213562, T7I, T7B); TbG = FMA(KP923879532, Tbq, Tbn); Tbr = FNMS(KP923879532, Tbq, Tbn); TbO = FNMS(KP707106781, T7t, T7e); T7u = FMA(KP707106781, T7t, T7e); TkB = FNMS(KP707106781, Tkm, Tkl); Tkn = FMA(KP707106781, Tkm, Tkl); TbI = FNMS(KP303346683, TbH, TbG); TbM = FMA(KP303346683, TbG, TbH); Tbw = FMA(KP534511135, Tbc, Tbr); Tbs = FNMS(KP534511135, Tbr, Tbc); T7Y = FNMS(KP414213562, T7X, T7Q); TbQ = FMA(KP414213562, T7Q, T7X); } } } { E TkJ, TkD, Tck, TbS, TbK, Tku, Tkt, TbN; { E TkA, Tby, Tkp, Tbu, Tkz, Tbx; { E Tbt, T9a, Tkx, Tky, Tkv; TkA = Taj + Tbs; Tbt = Taj - Tbs; { E TkC, T7Z, Tko, TbR, T80; TkC = T7J + T7Y; T7Z = T7J - T7Y; Tko = TbQ - TbP; TbR = TbP + TbQ; TkJ = FMA(KP923879532, TkC, TkB); TkD = FNMS(KP923879532, TkC, TkB); Tby = FMA(KP923879532, T7Z, T7u); T80 = FNMS(KP923879532, T7Z, T7u); Tkv = FNMS(KP923879532, Tko, Tkn); Tkp = FMA(KP923879532, Tko, Tkn); Tck = FMA(KP923879532, TbR, TbO); TbS = FNMS(KP923879532, TbR, TbO); T9a = FMA(KP831469612, T99, T80); Tbu = FNMS(KP831469612, T99, T80); } Tkz = FNMS(KP831469612, Tkw, Tkv); Tkx = FMA(KP831469612, Tkw, Tkv); Tky = Tbw - Tbv; Tbx = Tbv + Tbw; ri[WS(rs, 11)] = FMA(KP881921264, Tbt, T9a); ri[WS(rs, 43)] = FNMS(KP881921264, Tbt, T9a); ii[WS(rs, 43)] = FNMS(KP881921264, Tky, Tkx); ii[WS(rs, 11)] = FMA(KP881921264, Tky, Tkx); } { E TbC, TbJ, Tkr, Tks; TbK = FNMS(KP831469612, TbB, Tby); TbC = FMA(KP831469612, TbB, Tby); ri[WS(rs, 59)] = FMA(KP881921264, Tbx, Tbu); ri[WS(rs, 27)] = FNMS(KP881921264, Tbx, Tbu); ii[WS(rs, 59)] = FMA(KP881921264, TkA, Tkz); ii[WS(rs, 27)] = FNMS(KP881921264, TkA, Tkz); TbJ = TbF + TbI; Tku = TbI - TbF; Tkt = FNMS(KP831469612, Tkq, Tkp); Tkr = FMA(KP831469612, Tkq, Tkp); Tks = TbL + TbM; TbN = TbL - TbM; ri[WS(rs, 3)] = FMA(KP956940335, TbJ, TbC); ri[WS(rs, 35)] = FNMS(KP956940335, TbJ, TbC); ii[WS(rs, 35)] = FNMS(KP956940335, Tks, Tkr); ii[WS(rs, 3)] = FMA(KP956940335, Tks, Tkr); } } { E Tcg, TkI, TkH, Tcj; { E Tc0, Tcf, TkF, TkG; Tcg = FNMS(KP980785280, TbZ, TbS); Tc0 = FMA(KP980785280, TbZ, TbS); ri[WS(rs, 19)] = FMA(KP956940335, TbN, TbK); ri[WS(rs, 51)] = FNMS(KP956940335, TbN, TbK); ii[WS(rs, 51)] = FNMS(KP956940335, Tku, Tkt); ii[WS(rs, 19)] = FMA(KP956940335, Tku, Tkt); Tcf = Tc7 + Tce; TkI = Tce - Tc7; TkH = FNMS(KP980785280, TkE, TkD); TkF = FMA(KP980785280, TkE, TkD); TkG = Tch + Tci; Tcj = Tch - Tci; ri[WS(rs, 7)] = FMA(KP773010453, Tcf, Tc0); ri[WS(rs, 39)] = FNMS(KP773010453, Tcf, Tc0); ii[WS(rs, 39)] = FNMS(KP773010453, TkG, TkF); ii[WS(rs, 7)] = FMA(KP773010453, TkG, TkF); } { E Tco, Tcv, TkL, TkM; Tcw = FMA(KP980785280, Tcn, Tck); Tco = FNMS(KP980785280, Tcn, Tck); ri[WS(rs, 23)] = FMA(KP773010453, Tcj, Tcg); ri[WS(rs, 55)] = FNMS(KP773010453, Tcj, Tcg); ii[WS(rs, 55)] = FNMS(KP773010453, TkI, TkH); ii[WS(rs, 23)] = FMA(KP773010453, TkI, TkH); Tcv = Tcr - Tcu; TkO = Tcr + Tcu; TkN = FMA(KP980785280, TkK, TkJ); TkL = FNMS(KP980785280, TkK, TkJ); TkM = Tcy - Tcx; Tcz = Tcx + Tcy; ri[WS(rs, 15)] = FMA(KP995184726, Tcv, Tco); ri[WS(rs, 47)] = FNMS(KP995184726, Tcv, Tco); ii[WS(rs, 47)] = FNMS(KP995184726, TkM, TkL); ii[WS(rs, 15)] = FMA(KP995184726, TkM, TkL); } } } } { E TdN, Tk2, TjW, Td1, TdR, TdX, TdH, Tdl, TeG, TeK, Teu, Teq, TeD, TeJ, Tet; E Tej, Tka, Tez, Tkg, Teb, Te1, TcH, Te0, TcE, Tk7, TjT, TdU, TdY, TdI, TdE; E TcK, Te2; { E Te5, Te6, Te8, Te9, Teg, Tee, Ted, Teh, Tdz, Tdv, Tdo, Ten, TeF, Tem, Teo; E TdC, Tex, Te7; { E TcP, TcS, TcW, TcZ; Te5 = FNMS(KP707106781, TcO, TcN); TcP = FMA(KP707106781, TcO, TcN); ri[WS(rs, 63)] = FMA(KP995184726, Tcz, Tcw); ri[WS(rs, 31)] = FNMS(KP995184726, Tcz, Tcw); ii[WS(rs, 63)] = FMA(KP995184726, TkO, TkN); ii[WS(rs, 31)] = FNMS(KP995184726, TkO, TkN); TcS = FMA(KP707106781, TcR, TcQ); Te6 = FNMS(KP707106781, TcR, TcQ); Te8 = FNMS(KP707106781, TcV, TcU); TcW = FMA(KP707106781, TcV, TcU); TcZ = FMA(KP707106781, TcY, TcX); Te9 = FNMS(KP707106781, TcY, TcX); { E Tdg, TdQ, Tdd, Tdj; { E Td5, TdL, TcT, TdM, Td0, Tdc; Teg = FNMS(KP707106781, Td4, Td3); Td5 = FMA(KP707106781, Td4, Td3); TdL = FMA(KP198912367, TcP, TcS); TcT = FNMS(KP198912367, TcS, TcP); TdM = FNMS(KP198912367, TcW, TcZ); Td0 = FMA(KP198912367, TcZ, TcW); Tdc = Td8 + Tdb; Tee = Tdb - Td8; Ted = FNMS(KP707106781, Tdf, Tde); Tdg = FMA(KP707106781, Tdf, Tde); TdN = TdL + TdM; Tk2 = TdM - TdL; TjW = TcT + Td0; Td1 = TcT - Td0; TdQ = FMA(KP923879532, Tdc, Td5); Tdd = FNMS(KP923879532, Tdc, Td5); Tdj = Tdh + Tdi; Teh = Tdh - Tdi; } { E Tek, Tel, TdP, Tdk; Tdz = FMA(KP707106781, Tdy, Tdx); Tek = FNMS(KP707106781, Tdy, Tdx); Tel = Tdu - Tdr; Tdv = Tdr + Tdu; Tdo = FMA(KP707106781, Tdn, Tdm); Ten = FNMS(KP707106781, Tdn, Tdm); TdP = FMA(KP923879532, Tdj, Tdg); Tdk = FNMS(KP923879532, Tdj, Tdg); TeF = FMA(KP923879532, Tel, Tek); Tem = FNMS(KP923879532, Tel, Tek); TdR = FMA(KP098491403, TdQ, TdP); TdX = FNMS(KP098491403, TdP, TdQ); TdH = FNMS(KP820678790, Tdd, Tdk); Tdl = FMA(KP820678790, Tdk, Tdd); Teo = TdA - TdB; TdC = TdA + TdB; } } } { E TeC, Tef, TeE, Tep, TeB, Tei; TeE = FMA(KP923879532, Teo, Ten); Tep = FNMS(KP923879532, Teo, Ten); TeC = FMA(KP923879532, Tee, Ted); Tef = FNMS(KP923879532, Tee, Ted); TeG = FNMS(KP303346683, TeF, TeE); TeK = FMA(KP303346683, TeE, TeF); Teu = FMA(KP534511135, Tem, Tep); Teq = FNMS(KP534511135, Tep, Tem); TeB = FMA(KP923879532, Teh, Teg); Tei = FNMS(KP923879532, Teh, Teg); Tex = FNMS(KP668178637, Te5, Te6); Te7 = FMA(KP668178637, Te6, Te5); TeD = FMA(KP303346683, TeC, TeB); TeJ = FNMS(KP303346683, TeB, TeC); Tet = FNMS(KP534511135, Tef, Tei); Tej = FMA(KP534511135, Tei, Tef); } { E TdT, Tdw, Tey, Tea, TdS, TdD; Tey = FMA(KP668178637, Te8, Te9); Tea = FNMS(KP668178637, Te9, Te8); TdT = FMA(KP923879532, Tdv, Tdo); Tdw = FNMS(KP923879532, Tdv, Tdo); Tka = Tey - Tex; Tez = Tex + Tey; Tkg = Te7 + Tea; Teb = Te7 - Tea; Te1 = FNMS(KP414213562, TcF, TcG); TcH = FMA(KP414213562, TcG, TcF); TdS = FMA(KP923879532, TdC, Tdz); TdD = FNMS(KP923879532, TdC, Tdz); Te0 = FNMS(KP707106781, TcD, TcA); TcE = FMA(KP707106781, TcD, TcA); Tk7 = FNMS(KP707106781, TjS, TjR); TjT = FMA(KP707106781, TjS, TjR); TdU = FNMS(KP098491403, TdT, TdS); TdY = FMA(KP098491403, TdS, TdT); TdI = FMA(KP820678790, Tdw, TdD); TdE = FNMS(KP820678790, TdD, Tdw); TcK = FNMS(KP414213562, TcJ, TcI); Te2 = FMA(KP414213562, TcI, TcJ); } } { E Tkf, Tk9, Tew, Te4, TdW, Tk0, TjZ, TdZ; { E Tk6, TdK, TjV, TdG, Tk5, TdJ; { E TdF, Td2, Tk3, Tk4, Tk1; Tk6 = Tdl + TdE; TdF = Tdl - TdE; { E Tk8, TcL, TjU, Te3, TcM; Tk8 = TcK - TcH; TcL = TcH + TcK; TjU = Te1 + Te2; Te3 = Te1 - Te2; Tkf = FNMS(KP923879532, Tk8, Tk7); Tk9 = FMA(KP923879532, Tk8, Tk7); TdK = FMA(KP923879532, TcL, TcE); TcM = FNMS(KP923879532, TcL, TcE); Tk1 = FNMS(KP923879532, TjU, TjT); TjV = FMA(KP923879532, TjU, TjT); Tew = FNMS(KP923879532, Te3, Te0); Te4 = FMA(KP923879532, Te3, Te0); Td2 = FMA(KP980785280, Td1, TcM); TdG = FNMS(KP980785280, Td1, TcM); } Tk5 = FNMS(KP980785280, Tk2, Tk1); Tk3 = FMA(KP980785280, Tk2, Tk1); Tk4 = TdI - TdH; TdJ = TdH + TdI; ri[WS(rs, 9)] = FMA(KP773010453, TdF, Td2); ri[WS(rs, 41)] = FNMS(KP773010453, TdF, Td2); ii[WS(rs, 41)] = FNMS(KP773010453, Tk4, Tk3); ii[WS(rs, 9)] = FMA(KP773010453, Tk4, Tk3); } { E TdO, TdV, TjX, TjY; TdW = FNMS(KP980785280, TdN, TdK); TdO = FMA(KP980785280, TdN, TdK); ri[WS(rs, 57)] = FMA(KP773010453, TdJ, TdG); ri[WS(rs, 25)] = FNMS(KP773010453, TdJ, TdG); ii[WS(rs, 57)] = FMA(KP773010453, Tk6, Tk5); ii[WS(rs, 25)] = FNMS(KP773010453, Tk6, Tk5); TdV = TdR + TdU; Tk0 = TdU - TdR; TjZ = FNMS(KP980785280, TjW, TjV); TjX = FMA(KP980785280, TjW, TjV); TjY = TdX + TdY; TdZ = TdX - TdY; ri[WS(rs, 1)] = FMA(KP995184726, TdV, TdO); ri[WS(rs, 33)] = FNMS(KP995184726, TdV, TdO); ii[WS(rs, 33)] = FNMS(KP995184726, TjY, TjX); ii[WS(rs, 1)] = FMA(KP995184726, TjY, TjX); } } { E Tes, Tke, Tkd, Tev; { E Tec, Ter, Tkb, Tkc; Tes = FNMS(KP831469612, Teb, Te4); Tec = FMA(KP831469612, Teb, Te4); ri[WS(rs, 17)] = FMA(KP995184726, TdZ, TdW); ri[WS(rs, 49)] = FNMS(KP995184726, TdZ, TdW); ii[WS(rs, 49)] = FNMS(KP995184726, Tk0, TjZ); ii[WS(rs, 17)] = FMA(KP995184726, Tk0, TjZ); Ter = Tej + Teq; Tke = Teq - Tej; Tkd = FNMS(KP831469612, Tka, Tk9); Tkb = FMA(KP831469612, Tka, Tk9); Tkc = Tet + Teu; Tev = Tet - Teu; ri[WS(rs, 5)] = FMA(KP881921264, Ter, Tec); ri[WS(rs, 37)] = FNMS(KP881921264, Ter, Tec); ii[WS(rs, 37)] = FNMS(KP881921264, Tkc, Tkb); ii[WS(rs, 5)] = FMA(KP881921264, Tkc, Tkb); } { E TeA, TeH, Tkh, Tki; TeI = FMA(KP831469612, Tez, Tew); TeA = FNMS(KP831469612, Tez, Tew); ri[WS(rs, 21)] = FMA(KP881921264, Tev, Tes); ri[WS(rs, 53)] = FNMS(KP881921264, Tev, Tes); ii[WS(rs, 53)] = FNMS(KP881921264, Tke, Tkd); ii[WS(rs, 21)] = FMA(KP881921264, Tke, Tkd); TeH = TeD - TeG; Tkk = TeD + TeG; Tkj = FMA(KP831469612, Tkg, Tkf); Tkh = FNMS(KP831469612, Tkg, Tkf); Tki = TeK - TeJ; TeL = TeJ + TeK; ri[WS(rs, 13)] = FMA(KP956940335, TeH, TeA); ri[WS(rs, 45)] = FNMS(KP956940335, TeH, TeA); ii[WS(rs, 45)] = FNMS(KP956940335, Tki, Tkh); ii[WS(rs, 13)] = FMA(KP956940335, Tki, Tkh); } } } } } } } ri[WS(rs, 61)] = FMA(KP956940335, TeL, TeI); ri[WS(rs, 29)] = FNMS(KP956940335, TeL, TeI); ii[WS(rs, 61)] = FMA(KP956940335, Tkk, Tkj); ii[WS(rs, 29)] = FNMS(KP956940335, Tkk, Tkj); } } } static const tw_instr twinstr[] = { {TW_FULL, 0, 64}, {TW_NEXT, 1, 0} }; static const ct_desc desc = { 64, "t1_64", twinstr, &GENUS, {520, 126, 518, 0}, 0, 0, 0 }; void X(codelet_t1_64) (planner *p) { X(kdft_dit_register) (p, t1_64, &desc); } #else /* HAVE_FMA */ /* Generated by: ../../../genfft/gen_twiddle.native -compact -variables 4 -pipeline-latency 4 -n 64 -name t1_64 -include t.h */ /* * This function contains 1038 FP additions, 500 FP multiplications, * (or, 808 additions, 270 multiplications, 230 fused multiply/add), * 176 stack variables, 15 constants, and 256 memory accesses */ #include "t.h" static void t1_64(R *ri, R *ii, const R *W, stride rs, INT mb, INT me, INT ms) { DK(KP471396736, +0.471396736825997648556387625905254377657460319); DK(KP881921264, +0.881921264348355029712756863660388349508442621); DK(KP290284677, +0.290284677254462367636192375817395274691476278); DK(KP956940335, +0.956940335732208864935797886980269969482849206); DK(KP634393284, +0.634393284163645498215171613225493370675687095); DK(KP773010453, +0.773010453362736960810906609758469800971041293); DK(KP098017140, +0.098017140329560601994195563888641845861136673); DK(KP995184726, +0.995184726672196886244836953109479921575474869); DK(KP555570233, +0.555570233019602224742830813948532874374937191); DK(KP831469612, +0.831469612302545237078788377617905756738560812); DK(KP980785280, +0.980785280403230449126182236134239036973933731); DK(KP195090322, +0.195090322016128267848284868477022240927691618); DK(KP923879532, +0.923879532511286756128183189396788286822416626); DK(KP382683432, +0.382683432365089771728459984030398866761344562); DK(KP707106781, +0.707106781186547524400844362104849039284835938); { INT m; for (m = mb, W = W + (mb * 126); m < me; m = m + 1, ri = ri + ms, ii = ii + ms, W = W + 126, MAKE_VOLATILE_STRIDE(128, rs)) { E Tj, TcL, ThT, Tin, T6b, Taz, TgT, Thn, TG, Thm, TcO, TgO, T6m, ThQ, TaC; E Tim, T14, Tfq, T6y, T9O, TaG, Tc0, TcU, TeE, T1r, Tfr, T6J, T9P, TaJ, Tc1; E TcZ, TeF, T1Q, T2d, Tfx, Tfu, Tfv, Tfw, T6Q, TaM, Tdb, TeJ, T71, TaQ, T7a; E TaN, Td6, TeI, T77, TaP, T2B, T2Y, Tfz, TfA, TfB, TfC, T7h, TaW, Tdm, TeM; E T7s, TaU, T7B, TaX, Tdh, TeL, T7y, TaT, T5j, TfR, Tec, Tf0, TfY, Tgy, T8D; E Tbl, T8O, Tbx, T9l, Tbm, TdV, TeX, T9i, Tbw, T3M, TfL, TdL, TeQ, TfI, Tgt; E T7K, Tb2, T7V, Tbe, T8s, Tb3, Tdu, TeT, T8p, Tbd, T4x, TfJ, TdE, TdM, TfO; E Tgu, T87, T8v, T8i, T8u, Tba, Tbg, Tdz, TdN, Tb7, Tbh, T64, TfZ, Te5, Ted; E TfU, Tgz, T90, T9o, T9b, T9n, Tbt, Tbz, Te0, Tee, Tbq, TbA; { E T1, TgR, T6, TgQ, Tc, T68, Th, T69; T1 = ri[0]; TgR = ii[0]; { E T3, T5, T2, T4; T3 = ri[WS(rs, 32)]; T5 = ii[WS(rs, 32)]; T2 = W[62]; T4 = W[63]; T6 = FMA(T2, T3, T4 * T5); TgQ = FNMS(T4, T3, T2 * T5); } { E T9, Tb, T8, Ta; T9 = ri[WS(rs, 16)]; Tb = ii[WS(rs, 16)]; T8 = W[30]; Ta = W[31]; Tc = FMA(T8, T9, Ta * Tb); T68 = FNMS(Ta, T9, T8 * Tb); } { E Te, Tg, Td, Tf; Te = ri[WS(rs, 48)]; Tg = ii[WS(rs, 48)]; Td = W[94]; Tf = W[95]; Th = FMA(Td, Te, Tf * Tg); T69 = FNMS(Tf, Te, Td * Tg); } { E T7, Ti, ThR, ThS; T7 = T1 + T6; Ti = Tc + Th; Tj = T7 + Ti; TcL = T7 - Ti; ThR = TgR - TgQ; ThS = Tc - Th; ThT = ThR - ThS; Tin = ThS + ThR; } { E T67, T6a, TgP, TgS; T67 = T1 - T6; T6a = T68 - T69; T6b = T67 - T6a; Taz = T67 + T6a; TgP = T68 + T69; TgS = TgQ + TgR; TgT = TgP + TgS; Thn = TgS - TgP; } } { E To, T6c, Tt, T6d, T6e, T6f, Tz, T6i, TE, T6j, T6h, T6k; { E Tl, Tn, Tk, Tm; Tl = ri[WS(rs, 8)]; Tn = ii[WS(rs, 8)]; Tk = W[14]; Tm = W[15]; To = FMA(Tk, Tl, Tm * Tn); T6c = FNMS(Tm, Tl, Tk * Tn); } { E Tq, Ts, Tp, Tr; Tq = ri[WS(rs, 40)]; Ts = ii[WS(rs, 40)]; Tp = W[78]; Tr = W[79]; Tt = FMA(Tp, Tq, Tr * Ts); T6d = FNMS(Tr, Tq, Tp * Ts); } T6e = T6c - T6d; T6f = To - Tt; { E Tw, Ty, Tv, Tx; Tw = ri[WS(rs, 56)]; Ty = ii[WS(rs, 56)]; Tv = W[110]; Tx = W[111]; Tz = FMA(Tv, Tw, Tx * Ty); T6i = FNMS(Tx, Tw, Tv * Ty); } { E TB, TD, TA, TC; TB = ri[WS(rs, 24)]; TD = ii[WS(rs, 24)]; TA = W[46]; TC = W[47]; TE = FMA(TA, TB, TC * TD); T6j = FNMS(TC, TB, TA * TD); } T6h = Tz - TE; T6k = T6i - T6j; { E Tu, TF, TcM, TcN; Tu = To + Tt; TF = Tz + TE; TG = Tu + TF; Thm = TF - Tu; TcM = T6c + T6d; TcN = T6i + T6j; TcO = TcM - TcN; TgO = TcM + TcN; } { E T6g, T6l, TaA, TaB; T6g = T6e - T6f; T6l = T6h + T6k; T6m = KP707106781 * (T6g - T6l); ThQ = KP707106781 * (T6g + T6l); TaA = T6f + T6e; TaB = T6h - T6k; TaC = KP707106781 * (TaA + TaB); Tim = KP707106781 * (TaB - TaA); } } { E TS, TcQ, T6q, T6t, T13, TcR, T6r, T6w, T6s, T6x; { E TM, T6o, TR, T6p; { E TJ, TL, TI, TK; TJ = ri[WS(rs, 4)]; TL = ii[WS(rs, 4)]; TI = W[6]; TK = W[7]; TM = FMA(TI, TJ, TK * TL); T6o = FNMS(TK, TJ, TI * TL); } { E TO, TQ, TN, TP; TO = ri[WS(rs, 36)]; TQ = ii[WS(rs, 36)]; TN = W[70]; TP = W[71]; TR = FMA(TN, TO, TP * TQ); T6p = FNMS(TP, TO, TN * TQ); } TS = TM + TR; TcQ = T6o + T6p; T6q = T6o - T6p; T6t = TM - TR; } { E TX, T6u, T12, T6v; { E TU, TW, TT, TV; TU = ri[WS(rs, 20)]; TW = ii[WS(rs, 20)]; TT = W[38]; TV = W[39]; TX = FMA(TT, TU, TV * TW); T6u = FNMS(TV, TU, TT * TW); } { E TZ, T11, TY, T10; TZ = ri[WS(rs, 52)]; T11 = ii[WS(rs, 52)]; TY = W[102]; T10 = W[103]; T12 = FMA(TY, TZ, T10 * T11); T6v = FNMS(T10, TZ, TY * T11); } T13 = TX + T12; TcR = T6u + T6v; T6r = TX - T12; T6w = T6u - T6v; } T14 = TS + T13; Tfq = TcQ + TcR; T6s = T6q + T6r; T6x = T6t - T6w; T6y = FNMS(KP923879532, T6x, KP382683432 * T6s); T9O = FMA(KP923879532, T6s, KP382683432 * T6x); { E TaE, TaF, TcS, TcT; TaE = T6q - T6r; TaF = T6t + T6w; TaG = FNMS(KP382683432, TaF, KP923879532 * TaE); Tc0 = FMA(KP382683432, TaE, KP923879532 * TaF); TcS = TcQ - TcR; TcT = TS - T13; TcU = TcS - TcT; TeE = TcT + TcS; } } { E T1f, TcW, T6B, T6E, T1q, TcX, T6C, T6H, T6D, T6I; { E T19, T6z, T1e, T6A; { E T16, T18, T15, T17; T16 = ri[WS(rs, 60)]; T18 = ii[WS(rs, 60)]; T15 = W[118]; T17 = W[119]; T19 = FMA(T15, T16, T17 * T18); T6z = FNMS(T17, T16, T15 * T18); } { E T1b, T1d, T1a, T1c; T1b = ri[WS(rs, 28)]; T1d = ii[WS(rs, 28)]; T1a = W[54]; T1c = W[55]; T1e = FMA(T1a, T1b, T1c * T1d); T6A = FNMS(T1c, T1b, T1a * T1d); } T1f = T19 + T1e; TcW = T6z + T6A; T6B = T6z - T6A; T6E = T19 - T1e; } { E T1k, T6F, T1p, T6G; { E T1h, T1j, T1g, T1i; T1h = ri[WS(rs, 12)]; T1j = ii[WS(rs, 12)]; T1g = W[22]; T1i = W[23]; T1k = FMA(T1g, T1h, T1i * T1j); T6F = FNMS(T1i, T1h, T1g * T1j); } { E T1m, T1o, T1l, T1n; T1m = ri[WS(rs, 44)]; T1o = ii[WS(rs, 44)]; T1l = W[86]; T1n = W[87]; T1p = FMA(T1l, T1m, T1n * T1o); T6G = FNMS(T1n, T1m, T1l * T1o); } T1q = T1k + T1p; TcX = T6F + T6G; T6C = T1k - T1p; T6H = T6F - T6G; } T1r = T1f + T1q; Tfr = TcW + TcX; T6D = T6B + T6C; T6I = T6E - T6H; T6J = FMA(KP382683432, T6D, KP923879532 * T6I); T9P = FNMS(KP923879532, T6D, KP382683432 * T6I); { E TaH, TaI, TcV, TcY; TaH = T6B - T6C; TaI = T6E + T6H; TaJ = FMA(KP923879532, TaH, KP382683432 * TaI); Tc1 = FNMS(KP382683432, TaH, KP923879532 * TaI); TcV = T1f - T1q; TcY = TcW - TcX; TcZ = TcV + TcY; TeF = TcV - TcY; } } { E T1y, T6M, T1D, T6N, T1E, Td2, T1J, T74, T1O, T75, T1P, Td3, T21, Td8, T6W; E T6Z, T2c, Td9, T6R, T6U; { E T1v, T1x, T1u, T1w; T1v = ri[WS(rs, 2)]; T1x = ii[WS(rs, 2)]; T1u = W[2]; T1w = W[3]; T1y = FMA(T1u, T1v, T1w * T1x); T6M = FNMS(T1w, T1v, T1u * T1x); } { E T1A, T1C, T1z, T1B; T1A = ri[WS(rs, 34)]; T1C = ii[WS(rs, 34)]; T1z = W[66]; T1B = W[67]; T1D = FMA(T1z, T1A, T1B * T1C); T6N = FNMS(T1B, T1A, T1z * T1C); } T1E = T1y + T1D; Td2 = T6M + T6N; { E T1G, T1I, T1F, T1H; T1G = ri[WS(rs, 18)]; T1I = ii[WS(rs, 18)]; T1F = W[34]; T1H = W[35]; T1J = FMA(T1F, T1G, T1H * T1I); T74 = FNMS(T1H, T1G, T1F * T1I); } { E T1L, T1N, T1K, T1M; T1L = ri[WS(rs, 50)]; T1N = ii[WS(rs, 50)]; T1K = W[98]; T1M = W[99]; T1O = FMA(T1K, T1L, T1M * T1N); T75 = FNMS(T1M, T1L, T1K * T1N); } T1P = T1J + T1O; Td3 = T74 + T75; { E T1V, T6X, T20, T6Y; { E T1S, T1U, T1R, T1T; T1S = ri[WS(rs, 10)]; T1U = ii[WS(rs, 10)]; T1R = W[18]; T1T = W[19]; T1V = FMA(T1R, T1S, T1T * T1U); T6X = FNMS(T1T, T1S, T1R * T1U); } { E T1X, T1Z, T1W, T1Y; T1X = ri[WS(rs, 42)]; T1Z = ii[WS(rs, 42)]; T1W = W[82]; T1Y = W[83]; T20 = FMA(T1W, T1X, T1Y * T1Z); T6Y = FNMS(T1Y, T1X, T1W * T1Z); } T21 = T1V + T20; Td8 = T6X + T6Y; T6W = T1V - T20; T6Z = T6X - T6Y; } { E T26, T6S, T2b, T6T; { E T23, T25, T22, T24; T23 = ri[WS(rs, 58)]; T25 = ii[WS(rs, 58)]; T22 = W[114]; T24 = W[115]; T26 = FMA(T22, T23, T24 * T25); T6S = FNMS(T24, T23, T22 * T25); } { E T28, T2a, T27, T29; T28 = ri[WS(rs, 26)]; T2a = ii[WS(rs, 26)]; T27 = W[50]; T29 = W[51]; T2b = FMA(T27, T28, T29 * T2a); T6T = FNMS(T29, T28, T27 * T2a); } T2c = T26 + T2b; Td9 = T6S + T6T; T6R = T26 - T2b; T6U = T6S - T6T; } T1Q = T1E + T1P; T2d = T21 + T2c; Tfx = T1Q - T2d; Tfu = Td2 + Td3; Tfv = Td8 + Td9; Tfw = Tfu - Tfv; { E T6O, T6P, Td7, Tda; T6O = T6M - T6N; T6P = T1J - T1O; T6Q = T6O + T6P; TaM = T6O - T6P; Td7 = T1E - T1P; Tda = Td8 - Td9; Tdb = Td7 - Tda; TeJ = Td7 + Tda; } { E T6V, T70, T78, T79; T6V = T6R - T6U; T70 = T6W + T6Z; T71 = KP707106781 * (T6V - T70); TaQ = KP707106781 * (T70 + T6V); T78 = T6Z - T6W; T79 = T6R + T6U; T7a = KP707106781 * (T78 - T79); TaN = KP707106781 * (T78 + T79); } { E Td4, Td5, T73, T76; Td4 = Td2 - Td3; Td5 = T2c - T21; Td6 = Td4 - Td5; TeI = Td4 + Td5; T73 = T1y - T1D; T76 = T74 - T75; T77 = T73 - T76; TaP = T73 + T76; } } { E T2j, T7d, T2o, T7e, T2p, Tdd, T2u, T7v, T2z, T7w, T2A, Tde, T2M, Tdj, T7n; E T7q, T2X, Tdk, T7i, T7l; { E T2g, T2i, T2f, T2h; T2g = ri[WS(rs, 62)]; T2i = ii[WS(rs, 62)]; T2f = W[122]; T2h = W[123]; T2j = FMA(T2f, T2g, T2h * T2i); T7d = FNMS(T2h, T2g, T2f * T2i); } { E T2l, T2n, T2k, T2m; T2l = ri[WS(rs, 30)]; T2n = ii[WS(rs, 30)]; T2k = W[58]; T2m = W[59]; T2o = FMA(T2k, T2l, T2m * T2n); T7e = FNMS(T2m, T2l, T2k * T2n); } T2p = T2j + T2o; Tdd = T7d + T7e; { E T2r, T2t, T2q, T2s; T2r = ri[WS(rs, 14)]; T2t = ii[WS(rs, 14)]; T2q = W[26]; T2s = W[27]; T2u = FMA(T2q, T2r, T2s * T2t); T7v = FNMS(T2s, T2r, T2q * T2t); } { E T2w, T2y, T2v, T2x; T2w = ri[WS(rs, 46)]; T2y = ii[WS(rs, 46)]; T2v = W[90]; T2x = W[91]; T2z = FMA(T2v, T2w, T2x * T2y); T7w = FNMS(T2x, T2w, T2v * T2y); } T2A = T2u + T2z; Tde = T7v + T7w; { E T2G, T7o, T2L, T7p; { E T2D, T2F, T2C, T2E; T2D = ri[WS(rs, 6)]; T2F = ii[WS(rs, 6)]; T2C = W[10]; T2E = W[11]; T2G = FMA(T2C, T2D, T2E * T2F); T7o = FNMS(T2E, T2D, T2C * T2F); } { E T2I, T2K, T2H, T2J; T2I = ri[WS(rs, 38)]; T2K = ii[WS(rs, 38)]; T2H = W[74]; T2J = W[75]; T2L = FMA(T2H, T2I, T2J * T2K); T7p = FNMS(T2J, T2I, T2H * T2K); } T2M = T2G + T2L; Tdj = T7o + T7p; T7n = T2G - T2L; T7q = T7o - T7p; } { E T2R, T7j, T2W, T7k; { E T2O, T2Q, T2N, T2P; T2O = ri[WS(rs, 54)]; T2Q = ii[WS(rs, 54)]; T2N = W[106]; T2P = W[107]; T2R = FMA(T2N, T2O, T2P * T2Q); T7j = FNMS(T2P, T2O, T2N * T2Q); } { E T2T, T2V, T2S, T2U; T2T = ri[WS(rs, 22)]; T2V = ii[WS(rs, 22)]; T2S = W[42]; T2U = W[43]; T2W = FMA(T2S, T2T, T2U * T2V); T7k = FNMS(T2U, T2T, T2S * T2V); } T2X = T2R + T2W; Tdk = T7j + T7k; T7i = T2R - T2W; T7l = T7j - T7k; } T2B = T2p + T2A; T2Y = T2M + T2X; Tfz = T2B - T2Y; TfA = Tdd + Tde; TfB = Tdj + Tdk; TfC = TfA - TfB; { E T7f, T7g, Tdi, Tdl; T7f = T7d - T7e; T7g = T2u - T2z; T7h = T7f + T7g; TaW = T7f - T7g; Tdi = T2p - T2A; Tdl = Tdj - Tdk; Tdm = Tdi - Tdl; TeM = Tdi + Tdl; } { E T7m, T7r, T7z, T7A; T7m = T7i - T7l; T7r = T7n + T7q; T7s = KP707106781 * (T7m - T7r); TaU = KP707106781 * (T7r + T7m); T7z = T7q - T7n; T7A = T7i + T7l; T7B = KP707106781 * (T7z - T7A); TaX = KP707106781 * (T7z + T7A); } { E Tdf, Tdg, T7u, T7x; Tdf = Tdd - Tde; Tdg = T2X - T2M; Tdh = Tdf - Tdg; TeL = Tdf + Tdg; T7u = T2j - T2o; T7x = T7v - T7w; T7y = T7u - T7x; TaT = T7u + T7x; } } { E T4D, T9e, T4I, T9f, T4J, Te8, T4O, T8A, T4T, T8B, T4U, Te9, T56, TdS, T8G; E T8H, T5h, TdT, T8J, T8M; { E T4A, T4C, T4z, T4B; T4A = ri[WS(rs, 63)]; T4C = ii[WS(rs, 63)]; T4z = W[124]; T4B = W[125]; T4D = FMA(T4z, T4A, T4B * T4C); T9e = FNMS(T4B, T4A, T4z * T4C); } { E T4F, T4H, T4E, T4G; T4F = ri[WS(rs, 31)]; T4H = ii[WS(rs, 31)]; T4E = W[60]; T4G = W[61]; T4I = FMA(T4E, T4F, T4G * T4H); T9f = FNMS(T4G, T4F, T4E * T4H); } T4J = T4D + T4I; Te8 = T9e + T9f; { E T4L, T4N, T4K, T4M; T4L = ri[WS(rs, 15)]; T4N = ii[WS(rs, 15)]; T4K = W[28]; T4M = W[29]; T4O = FMA(T4K, T4L, T4M * T4N); T8A = FNMS(T4M, T4L, T4K * T4N); } { E T4Q, T4S, T4P, T4R; T4Q = ri[WS(rs, 47)]; T4S = ii[WS(rs, 47)]; T4P = W[92]; T4R = W[93]; T4T = FMA(T4P, T4Q, T4R * T4S); T8B = FNMS(T4R, T4Q, T4P * T4S); } T4U = T4O + T4T; Te9 = T8A + T8B; { E T50, T8E, T55, T8F; { E T4X, T4Z, T4W, T4Y; T4X = ri[WS(rs, 7)]; T4Z = ii[WS(rs, 7)]; T4W = W[12]; T4Y = W[13]; T50 = FMA(T4W, T4X, T4Y * T4Z); T8E = FNMS(T4Y, T4X, T4W * T4Z); } { E T52, T54, T51, T53; T52 = ri[WS(rs, 39)]; T54 = ii[WS(rs, 39)]; T51 = W[76]; T53 = W[77]; T55 = FMA(T51, T52, T53 * T54); T8F = FNMS(T53, T52, T51 * T54); } T56 = T50 + T55; TdS = T8E + T8F; T8G = T8E - T8F; T8H = T50 - T55; } { E T5b, T8K, T5g, T8L; { E T58, T5a, T57, T59; T58 = ri[WS(rs, 55)]; T5a = ii[WS(rs, 55)]; T57 = W[108]; T59 = W[109]; T5b = FMA(T57, T58, T59 * T5a); T8K = FNMS(T59, T58, T57 * T5a); } { E T5d, T5f, T5c, T5e; T5d = ri[WS(rs, 23)]; T5f = ii[WS(rs, 23)]; T5c = W[44]; T5e = W[45]; T5g = FMA(T5c, T5d, T5e * T5f); T8L = FNMS(T5e, T5d, T5c * T5f); } T5h = T5b + T5g; TdT = T8K + T8L; T8J = T5b - T5g; T8M = T8K - T8L; } { E T4V, T5i, Tea, Teb; T4V = T4J + T4U; T5i = T56 + T5h; T5j = T4V + T5i; TfR = T4V - T5i; Tea = Te8 - Te9; Teb = T5h - T56; Tec = Tea - Teb; Tf0 = Tea + Teb; } { E TfW, TfX, T8z, T8C; TfW = Te8 + Te9; TfX = TdS + TdT; TfY = TfW - TfX; Tgy = TfW + TfX; T8z = T4D - T4I; T8C = T8A - T8B; T8D = T8z - T8C; Tbl = T8z + T8C; } { E T8I, T8N, T9j, T9k; T8I = T8G - T8H; T8N = T8J + T8M; T8O = KP707106781 * (T8I - T8N); Tbx = KP707106781 * (T8I + T8N); T9j = T8J - T8M; T9k = T8H + T8G; T9l = KP707106781 * (T9j - T9k); Tbm = KP707106781 * (T9k + T9j); } { E TdR, TdU, T9g, T9h; TdR = T4J - T4U; TdU = TdS - TdT; TdV = TdR - TdU; TeX = TdR + TdU; T9g = T9e - T9f; T9h = T4O - T4T; T9i = T9g + T9h; Tbw = T9g - T9h; } } { E T36, T7G, T3b, T7H, T3c, Tdq, T3h, T8m, T3m, T8n, T3n, Tdr, T3z, TdI, T7Q; E T7T, T3K, TdJ, T7L, T7O; { E T33, T35, T32, T34; T33 = ri[WS(rs, 1)]; T35 = ii[WS(rs, 1)]; T32 = W[0]; T34 = W[1]; T36 = FMA(T32, T33, T34 * T35); T7G = FNMS(T34, T33, T32 * T35); } { E T38, T3a, T37, T39; T38 = ri[WS(rs, 33)]; T3a = ii[WS(rs, 33)]; T37 = W[64]; T39 = W[65]; T3b = FMA(T37, T38, T39 * T3a); T7H = FNMS(T39, T38, T37 * T3a); } T3c = T36 + T3b; Tdq = T7G + T7H; { E T3e, T3g, T3d, T3f; T3e = ri[WS(rs, 17)]; T3g = ii[WS(rs, 17)]; T3d = W[32]; T3f = W[33]; T3h = FMA(T3d, T3e, T3f * T3g); T8m = FNMS(T3f, T3e, T3d * T3g); } { E T3j, T3l, T3i, T3k; T3j = ri[WS(rs, 49)]; T3l = ii[WS(rs, 49)]; T3i = W[96]; T3k = W[97]; T3m = FMA(T3i, T3j, T3k * T3l); T8n = FNMS(T3k, T3j, T3i * T3l); } T3n = T3h + T3m; Tdr = T8m + T8n; { E T3t, T7R, T3y, T7S; { E T3q, T3s, T3p, T3r; T3q = ri[WS(rs, 9)]; T3s = ii[WS(rs, 9)]; T3p = W[16]; T3r = W[17]; T3t = FMA(T3p, T3q, T3r * T3s); T7R = FNMS(T3r, T3q, T3p * T3s); } { E T3v, T3x, T3u, T3w; T3v = ri[WS(rs, 41)]; T3x = ii[WS(rs, 41)]; T3u = W[80]; T3w = W[81]; T3y = FMA(T3u, T3v, T3w * T3x); T7S = FNMS(T3w, T3v, T3u * T3x); } T3z = T3t + T3y; TdI = T7R + T7S; T7Q = T3t - T3y; T7T = T7R - T7S; } { E T3E, T7M, T3J, T7N; { E T3B, T3D, T3A, T3C; T3B = ri[WS(rs, 57)]; T3D = ii[WS(rs, 57)]; T3A = W[112]; T3C = W[113]; T3E = FMA(T3A, T3B, T3C * T3D); T7M = FNMS(T3C, T3B, T3A * T3D); } { E T3G, T3I, T3F, T3H; T3G = ri[WS(rs, 25)]; T3I = ii[WS(rs, 25)]; T3F = W[48]; T3H = W[49]; T3J = FMA(T3F, T3G, T3H * T3I); T7N = FNMS(T3H, T3G, T3F * T3I); } T3K = T3E + T3J; TdJ = T7M + T7N; T7L = T3E - T3J; T7O = T7M - T7N; } { E T3o, T3L, TdH, TdK; T3o = T3c + T3n; T3L = T3z + T3K; T3M = T3o + T3L; TfL = T3o - T3L; TdH = T3c - T3n; TdK = TdI - TdJ; TdL = TdH - TdK; TeQ = TdH + TdK; } { E TfG, TfH, T7I, T7J; TfG = Tdq + Tdr; TfH = TdI + TdJ; TfI = TfG - TfH; Tgt = TfG + TfH; T7I = T7G - T7H; T7J = T3h - T3m; T7K = T7I + T7J; Tb2 = T7I - T7J; } { E T7P, T7U, T8q, T8r; T7P = T7L - T7O; T7U = T7Q + T7T; T7V = KP707106781 * (T7P - T7U); Tbe = KP707106781 * (T7U + T7P); T8q = T7T - T7Q; T8r = T7L + T7O; T8s = KP707106781 * (T8q - T8r); Tb3 = KP707106781 * (T8q + T8r); } { E Tds, Tdt, T8l, T8o; Tds = Tdq - Tdr; Tdt = T3K - T3z; Tdu = Tds - Tdt; TeT = Tds + Tdt; T8l = T36 - T3b; T8o = T8m - T8n; T8p = T8l - T8o; Tbd = T8l + T8o; } } { E T3X, TdB, T8a, T8d, T4v, Tdx, T80, T85, T48, TdC, T8b, T8g, T4k, Tdw, T7X; E T84; { E T3R, T88, T3W, T89; { E T3O, T3Q, T3N, T3P; T3O = ri[WS(rs, 5)]; T3Q = ii[WS(rs, 5)]; T3N = W[8]; T3P = W[9]; T3R = FMA(T3N, T3O, T3P * T3Q); T88 = FNMS(T3P, T3O, T3N * T3Q); } { E T3T, T3V, T3S, T3U; T3T = ri[WS(rs, 37)]; T3V = ii[WS(rs, 37)]; T3S = W[72]; T3U = W[73]; T3W = FMA(T3S, T3T, T3U * T3V); T89 = FNMS(T3U, T3T, T3S * T3V); } T3X = T3R + T3W; TdB = T88 + T89; T8a = T88 - T89; T8d = T3R - T3W; } { E T4p, T7Y, T4u, T7Z; { E T4m, T4o, T4l, T4n; T4m = ri[WS(rs, 13)]; T4o = ii[WS(rs, 13)]; T4l = W[24]; T4n = W[25]; T4p = FMA(T4l, T4m, T4n * T4o); T7Y = FNMS(T4n, T4m, T4l * T4o); } { E T4r, T4t, T4q, T4s; T4r = ri[WS(rs, 45)]; T4t = ii[WS(rs, 45)]; T4q = W[88]; T4s = W[89]; T4u = FMA(T4q, T4r, T4s * T4t); T7Z = FNMS(T4s, T4r, T4q * T4t); } T4v = T4p + T4u; Tdx = T7Y + T7Z; T80 = T7Y - T7Z; T85 = T4p - T4u; } { E T42, T8e, T47, T8f; { E T3Z, T41, T3Y, T40; T3Z = ri[WS(rs, 21)]; T41 = ii[WS(rs, 21)]; T3Y = W[40]; T40 = W[41]; T42 = FMA(T3Y, T3Z, T40 * T41); T8e = FNMS(T40, T3Z, T3Y * T41); } { E T44, T46, T43, T45; T44 = ri[WS(rs, 53)]; T46 = ii[WS(rs, 53)]; T43 = W[104]; T45 = W[105]; T47 = FMA(T43, T44, T45 * T46); T8f = FNMS(T45, T44, T43 * T46); } T48 = T42 + T47; TdC = T8e + T8f; T8b = T42 - T47; T8g = T8e - T8f; } { E T4e, T82, T4j, T83; { E T4b, T4d, T4a, T4c; T4b = ri[WS(rs, 61)]; T4d = ii[WS(rs, 61)]; T4a = W[120]; T4c = W[121]; T4e = FMA(T4a, T4b, T4c * T4d); T82 = FNMS(T4c, T4b, T4a * T4d); } { E T4g, T4i, T4f, T4h; T4g = ri[WS(rs, 29)]; T4i = ii[WS(rs, 29)]; T4f = W[56]; T4h = W[57]; T4j = FMA(T4f, T4g, T4h * T4i); T83 = FNMS(T4h, T4g, T4f * T4i); } T4k = T4e + T4j; Tdw = T82 + T83; T7X = T4e - T4j; T84 = T82 - T83; } { E T49, T4w, TdA, TdD; T49 = T3X + T48; T4w = T4k + T4v; T4x = T49 + T4w; TfJ = T4w - T49; TdA = T3X - T48; TdD = TdB - TdC; TdE = TdA + TdD; TdM = TdD - TdA; } { E TfM, TfN, T81, T86; TfM = TdB + TdC; TfN = Tdw + Tdx; TfO = TfM - TfN; Tgu = TfM + TfN; T81 = T7X - T80; T86 = T84 + T85; T87 = FNMS(KP923879532, T86, KP382683432 * T81); T8v = FMA(KP382683432, T86, KP923879532 * T81); } { E T8c, T8h, Tb8, Tb9; T8c = T8a + T8b; T8h = T8d - T8g; T8i = FMA(KP923879532, T8c, KP382683432 * T8h); T8u = FNMS(KP923879532, T8h, KP382683432 * T8c); Tb8 = T8a - T8b; Tb9 = T8d + T8g; Tba = FMA(KP382683432, Tb8, KP923879532 * Tb9); Tbg = FNMS(KP382683432, Tb9, KP923879532 * Tb8); } { E Tdv, Tdy, Tb5, Tb6; Tdv = T4k - T4v; Tdy = Tdw - Tdx; Tdz = Tdv - Tdy; TdN = Tdv + Tdy; Tb5 = T7X + T80; Tb6 = T84 - T85; Tb7 = FNMS(KP382683432, Tb6, KP923879532 * Tb5); Tbh = FMA(KP923879532, Tb6, KP382683432 * Tb5); } } { E T5u, TdW, T8S, T8V, T62, Te3, T94, T99, T5F, TdX, T8T, T8Y, T5R, Te2, T93; E T96; { E T5o, T8Q, T5t, T8R; { E T5l, T5n, T5k, T5m; T5l = ri[WS(rs, 3)]; T5n = ii[WS(rs, 3)]; T5k = W[4]; T5m = W[5]; T5o = FMA(T5k, T5l, T5m * T5n); T8Q = FNMS(T5m, T5l, T5k * T5n); } { E T5q, T5s, T5p, T5r; T5q = ri[WS(rs, 35)]; T5s = ii[WS(rs, 35)]; T5p = W[68]; T5r = W[69]; T5t = FMA(T5p, T5q, T5r * T5s); T8R = FNMS(T5r, T5q, T5p * T5s); } T5u = T5o + T5t; TdW = T8Q + T8R; T8S = T8Q - T8R; T8V = T5o - T5t; } { E T5W, T97, T61, T98; { E T5T, T5V, T5S, T5U; T5T = ri[WS(rs, 11)]; T5V = ii[WS(rs, 11)]; T5S = W[20]; T5U = W[21]; T5W = FMA(T5S, T5T, T5U * T5V); T97 = FNMS(T5U, T5T, T5S * T5V); } { E T5Y, T60, T5X, T5Z; T5Y = ri[WS(rs, 43)]; T60 = ii[WS(rs, 43)]; T5X = W[84]; T5Z = W[85]; T61 = FMA(T5X, T5Y, T5Z * T60); T98 = FNMS(T5Z, T5Y, T5X * T60); } T62 = T5W + T61; Te3 = T97 + T98; T94 = T5W - T61; T99 = T97 - T98; } { E T5z, T8W, T5E, T8X; { E T5w, T5y, T5v, T5x; T5w = ri[WS(rs, 19)]; T5y = ii[WS(rs, 19)]; T5v = W[36]; T5x = W[37]; T5z = FMA(T5v, T5w, T5x * T5y); T8W = FNMS(T5x, T5w, T5v * T5y); } { E T5B, T5D, T5A, T5C; T5B = ri[WS(rs, 51)]; T5D = ii[WS(rs, 51)]; T5A = W[100]; T5C = W[101]; T5E = FMA(T5A, T5B, T5C * T5D); T8X = FNMS(T5C, T5B, T5A * T5D); } T5F = T5z + T5E; TdX = T8W + T8X; T8T = T5z - T5E; T8Y = T8W - T8X; } { E T5L, T91, T5Q, T92; { E T5I, T5K, T5H, T5J; T5I = ri[WS(rs, 59)]; T5K = ii[WS(rs, 59)]; T5H = W[116]; T5J = W[117]; T5L = FMA(T5H, T5I, T5J * T5K); T91 = FNMS(T5J, T5I, T5H * T5K); } { E T5N, T5P, T5M, T5O; T5N = ri[WS(rs, 27)]; T5P = ii[WS(rs, 27)]; T5M = W[52]; T5O = W[53]; T5Q = FMA(T5M, T5N, T5O * T5P); T92 = FNMS(T5O, T5N, T5M * T5P); } T5R = T5L + T5Q; Te2 = T91 + T92; T93 = T91 - T92; T96 = T5L - T5Q; } { E T5G, T63, Te1, Te4; T5G = T5u + T5F; T63 = T5R + T62; T64 = T5G + T63; TfZ = T63 - T5G; Te1 = T5R - T62; Te4 = Te2 - Te3; Te5 = Te1 + Te4; Ted = Te1 - Te4; } { E TfS, TfT, T8U, T8Z; TfS = TdW + TdX; TfT = Te2 + Te3; TfU = TfS - TfT; Tgz = TfS + TfT; T8U = T8S + T8T; T8Z = T8V - T8Y; T90 = FNMS(KP923879532, T8Z, KP382683432 * T8U); T9o = FMA(KP923879532, T8U, KP382683432 * T8Z); } { E T95, T9a, Tbr, Tbs; T95 = T93 + T94; T9a = T96 - T99; T9b = FMA(KP382683432, T95, KP923879532 * T9a); T9n = FNMS(KP923879532, T95, KP382683432 * T9a); Tbr = T93 - T94; Tbs = T96 + T99; Tbt = FMA(KP923879532, Tbr, KP382683432 * Tbs); Tbz = FNMS(KP382683432, Tbr, KP923879532 * Tbs); } { E TdY, TdZ, Tbo, Tbp; TdY = TdW - TdX; TdZ = T5u - T5F; Te0 = TdY - TdZ; Tee = TdZ + TdY; Tbo = T8S - T8T; Tbp = T8V + T8Y; Tbq = FNMS(KP382683432, Tbp, KP923879532 * Tbo); TbA = FMA(KP382683432, Tbo, KP923879532 * Tbp); } } { E T1t, Tgn, TgK, TgL, TgV, Th1, T30, Th0, T66, TgX, Tgw, TgE, TgB, TgF, Tgq; E TgM; { E TH, T1s, TgI, TgJ; TH = Tj + TG; T1s = T14 + T1r; T1t = TH + T1s; Tgn = TH - T1s; TgI = Tgt + Tgu; TgJ = Tgy + Tgz; TgK = TgI - TgJ; TgL = TgI + TgJ; } { E TgN, TgU, T2e, T2Z; TgN = Tfq + Tfr; TgU = TgO + TgT; TgV = TgN + TgU; Th1 = TgU - TgN; T2e = T1Q + T2d; T2Z = T2B + T2Y; T30 = T2e + T2Z; Th0 = T2Z - T2e; } { E T4y, T65, Tgs, Tgv; T4y = T3M + T4x; T65 = T5j + T64; T66 = T4y + T65; TgX = T65 - T4y; Tgs = T3M - T4x; Tgv = Tgt - Tgu; Tgw = Tgs + Tgv; TgE = Tgv - Tgs; } { E Tgx, TgA, Tgo, Tgp; Tgx = T5j - T64; TgA = Tgy - Tgz; TgB = Tgx - TgA; TgF = Tgx + TgA; Tgo = Tfu + Tfv; Tgp = TfA + TfB; Tgq = Tgo - Tgp; TgM = Tgo + Tgp; } { E T31, TgW, TgH, TgY; T31 = T1t + T30; ri[WS(rs, 32)] = T31 - T66; ri[0] = T31 + T66; TgW = TgM + TgV; ii[0] = TgL + TgW; ii[WS(rs, 32)] = TgW - TgL; TgH = T1t - T30; ri[WS(rs, 48)] = TgH - TgK; ri[WS(rs, 16)] = TgH + TgK; TgY = TgV - TgM; ii[WS(rs, 16)] = TgX + TgY; ii[WS(rs, 48)] = TgY - TgX; } { E Tgr, TgC, TgZ, Th2; Tgr = Tgn + Tgq; TgC = KP707106781 * (Tgw + TgB); ri[WS(rs, 40)] = Tgr - TgC; ri[WS(rs, 8)] = Tgr + TgC; TgZ = KP707106781 * (TgE + TgF); Th2 = Th0 + Th1; ii[WS(rs, 8)] = TgZ + Th2; ii[WS(rs, 40)] = Th2 - TgZ; } { E TgD, TgG, Th3, Th4; TgD = Tgn - Tgq; TgG = KP707106781 * (TgE - TgF); ri[WS(rs, 56)] = TgD - TgG; ri[WS(rs, 24)] = TgD + TgG; Th3 = KP707106781 * (TgB - Tgw); Th4 = Th1 - Th0; ii[WS(rs, 24)] = Th3 + Th4; ii[WS(rs, 56)] = Th4 - Th3; } } { E Tft, Tg7, Tgh, Tgl, Th9, Thf, TfE, Th6, TfQ, Tg4, Tga, The, Tge, Tgk, Tg1; E Tg5; { E Tfp, Tfs, Tgf, Tgg; Tfp = Tj - TG; Tfs = Tfq - Tfr; Tft = Tfp - Tfs; Tg7 = Tfp + Tfs; Tgf = TfR + TfU; Tgg = TfY + TfZ; Tgh = FNMS(KP382683432, Tgg, KP923879532 * Tgf); Tgl = FMA(KP923879532, Tgg, KP382683432 * Tgf); } { E Th7, Th8, Tfy, TfD; Th7 = T1r - T14; Th8 = TgT - TgO; Th9 = Th7 + Th8; Thf = Th8 - Th7; Tfy = Tfw - Tfx; TfD = Tfz + TfC; TfE = KP707106781 * (Tfy - TfD); Th6 = KP707106781 * (Tfy + TfD); } { E TfK, TfP, Tg8, Tg9; TfK = TfI - TfJ; TfP = TfL - TfO; TfQ = FMA(KP923879532, TfK, KP382683432 * TfP); Tg4 = FNMS(KP923879532, TfP, KP382683432 * TfK); Tg8 = Tfx + Tfw; Tg9 = Tfz - TfC; Tga = KP707106781 * (Tg8 + Tg9); The = KP707106781 * (Tg9 - Tg8); } { E Tgc, Tgd, TfV, Tg0; Tgc = TfI + TfJ; Tgd = TfL + TfO; Tge = FMA(KP382683432, Tgc, KP923879532 * Tgd); Tgk = FNMS(KP382683432, Tgd, KP923879532 * Tgc); TfV = TfR - TfU; Tg0 = TfY - TfZ; Tg1 = FNMS(KP923879532, Tg0, KP382683432 * TfV); Tg5 = FMA(KP382683432, Tg0, KP923879532 * TfV); } { E TfF, Tg2, Thd, Thg; TfF = Tft + TfE; Tg2 = TfQ + Tg1; ri[WS(rs, 44)] = TfF - Tg2; ri[WS(rs, 12)] = TfF + Tg2; Thd = Tg4 + Tg5; Thg = The + Thf; ii[WS(rs, 12)] = Thd + Thg; ii[WS(rs, 44)] = Thg - Thd; } { E Tg3, Tg6, Thh, Thi; Tg3 = Tft - TfE; Tg6 = Tg4 - Tg5; ri[WS(rs, 60)] = Tg3 - Tg6; ri[WS(rs, 28)] = Tg3 + Tg6; Thh = Tg1 - TfQ; Thi = Thf - The; ii[WS(rs, 28)] = Thh + Thi; ii[WS(rs, 60)] = Thi - Thh; } { E Tgb, Tgi, Th5, Tha; Tgb = Tg7 + Tga; Tgi = Tge + Tgh; ri[WS(rs, 36)] = Tgb - Tgi; ri[WS(rs, 4)] = Tgb + Tgi; Th5 = Tgk + Tgl; Tha = Th6 + Th9; ii[WS(rs, 4)] = Th5 + Tha; ii[WS(rs, 36)] = Tha - Th5; } { E Tgj, Tgm, Thb, Thc; Tgj = Tg7 - Tga; Tgm = Tgk - Tgl; ri[WS(rs, 52)] = Tgj - Tgm; ri[WS(rs, 20)] = Tgj + Tgm; Thb = Tgh - Tge; Thc = Th9 - Th6; ii[WS(rs, 20)] = Thb + Thc; ii[WS(rs, 52)] = Thc - Thb; } } { E Td1, Ten, Tdo, ThA, ThD, ThJ, Teq, ThI, Teh, TeB, Tel, Tex, TdQ, TeA, Tek; E Teu; { E TcP, Td0, Teo, Tep; TcP = TcL - TcO; Td0 = KP707106781 * (TcU - TcZ); Td1 = TcP - Td0; Ten = TcP + Td0; { E Tdc, Tdn, ThB, ThC; Tdc = FNMS(KP923879532, Tdb, KP382683432 * Td6); Tdn = FMA(KP382683432, Tdh, KP923879532 * Tdm); Tdo = Tdc - Tdn; ThA = Tdc + Tdn; ThB = KP707106781 * (TeF - TeE); ThC = Thn - Thm; ThD = ThB + ThC; ThJ = ThC - ThB; } Teo = FMA(KP923879532, Td6, KP382683432 * Tdb); Tep = FNMS(KP923879532, Tdh, KP382683432 * Tdm); Teq = Teo + Tep; ThI = Tep - Teo; { E Te7, Tev, Teg, Tew, Te6, Tef; Te6 = KP707106781 * (Te0 - Te5); Te7 = TdV - Te6; Tev = TdV + Te6; Tef = KP707106781 * (Ted - Tee); Teg = Tec - Tef; Tew = Tec + Tef; Teh = FNMS(KP980785280, Teg, KP195090322 * Te7); TeB = FMA(KP831469612, Tew, KP555570233 * Tev); Tel = FMA(KP195090322, Teg, KP980785280 * Te7); Tex = FNMS(KP555570233, Tew, KP831469612 * Tev); } { E TdG, Tes, TdP, Tet, TdF, TdO; TdF = KP707106781 * (Tdz - TdE); TdG = Tdu - TdF; Tes = Tdu + TdF; TdO = KP707106781 * (TdM - TdN); TdP = TdL - TdO; Tet = TdL + TdO; TdQ = FMA(KP980785280, TdG, KP195090322 * TdP); TeA = FNMS(KP555570233, Tet, KP831469612 * Tes); Tek = FNMS(KP980785280, TdP, KP195090322 * TdG); Teu = FMA(KP555570233, Tes, KP831469612 * Tet); } } { E Tdp, Tei, ThH, ThK; Tdp = Td1 + Tdo; Tei = TdQ + Teh; ri[WS(rs, 46)] = Tdp - Tei; ri[WS(rs, 14)] = Tdp + Tei; ThH = Tek + Tel; ThK = ThI + ThJ; ii[WS(rs, 14)] = ThH + ThK; ii[WS(rs, 46)] = ThK - ThH; } { E Tej, Tem, ThL, ThM; Tej = Td1 - Tdo; Tem = Tek - Tel; ri[WS(rs, 62)] = Tej - Tem; ri[WS(rs, 30)] = Tej + Tem; ThL = Teh - TdQ; ThM = ThJ - ThI; ii[WS(rs, 30)] = ThL + ThM; ii[WS(rs, 62)] = ThM - ThL; } { E Ter, Tey, Thz, ThE; Ter = Ten + Teq; Tey = Teu + Tex; ri[WS(rs, 38)] = Ter - Tey; ri[WS(rs, 6)] = Ter + Tey; Thz = TeA + TeB; ThE = ThA + ThD; ii[WS(rs, 6)] = Thz + ThE; ii[WS(rs, 38)] = ThE - Thz; } { E Tez, TeC, ThF, ThG; Tez = Ten - Teq; TeC = TeA - TeB; ri[WS(rs, 54)] = Tez - TeC; ri[WS(rs, 22)] = Tez + TeC; ThF = Tex - Teu; ThG = ThD - ThA; ii[WS(rs, 22)] = ThF + ThG; ii[WS(rs, 54)] = ThG - ThF; } } { E TeH, Tf9, TeO, Thk, Thp, Thv, Tfc, Thu, Tf3, Tfn, Tf7, Tfj, TeW, Tfm, Tf6; E Tfg; { E TeD, TeG, Tfa, Tfb; TeD = TcL + TcO; TeG = KP707106781 * (TeE + TeF); TeH = TeD - TeG; Tf9 = TeD + TeG; { E TeK, TeN, Thl, Tho; TeK = FNMS(KP382683432, TeJ, KP923879532 * TeI); TeN = FMA(KP923879532, TeL, KP382683432 * TeM); TeO = TeK - TeN; Thk = TeK + TeN; Thl = KP707106781 * (TcU + TcZ); Tho = Thm + Thn; Thp = Thl + Tho; Thv = Tho - Thl; } Tfa = FMA(KP382683432, TeI, KP923879532 * TeJ); Tfb = FNMS(KP382683432, TeL, KP923879532 * TeM); Tfc = Tfa + Tfb; Thu = Tfb - Tfa; { E TeZ, Tfh, Tf2, Tfi, TeY, Tf1; TeY = KP707106781 * (Tee + Ted); TeZ = TeX - TeY; Tfh = TeX + TeY; Tf1 = KP707106781 * (Te0 + Te5); Tf2 = Tf0 - Tf1; Tfi = Tf0 + Tf1; Tf3 = FNMS(KP831469612, Tf2, KP555570233 * TeZ); Tfn = FMA(KP195090322, Tfh, KP980785280 * Tfi); Tf7 = FMA(KP831469612, TeZ, KP555570233 * Tf2); Tfj = FNMS(KP195090322, Tfi, KP980785280 * Tfh); } { E TeS, Tfe, TeV, Tff, TeR, TeU; TeR = KP707106781 * (TdE + Tdz); TeS = TeQ - TeR; Tfe = TeQ + TeR; TeU = KP707106781 * (TdM + TdN); TeV = TeT - TeU; Tff = TeT + TeU; TeW = FMA(KP555570233, TeS, KP831469612 * TeV); Tfm = FNMS(KP195090322, Tfe, KP980785280 * Tff); Tf6 = FNMS(KP831469612, TeS, KP555570233 * TeV); Tfg = FMA(KP980785280, Tfe, KP195090322 * Tff); } } { E TeP, Tf4, Tht, Thw; TeP = TeH + TeO; Tf4 = TeW + Tf3; ri[WS(rs, 42)] = TeP - Tf4; ri[WS(rs, 10)] = TeP + Tf4; Tht = Tf6 + Tf7; Thw = Thu + Thv; ii[WS(rs, 10)] = Tht + Thw; ii[WS(rs, 42)] = Thw - Tht; } { E Tf5, Tf8, Thx, Thy; Tf5 = TeH - TeO; Tf8 = Tf6 - Tf7; ri[WS(rs, 58)] = Tf5 - Tf8; ri[WS(rs, 26)] = Tf5 + Tf8; Thx = Tf3 - TeW; Thy = Thv - Thu; ii[WS(rs, 26)] = Thx + Thy; ii[WS(rs, 58)] = Thy - Thx; } { E Tfd, Tfk, Thj, Thq; Tfd = Tf9 + Tfc; Tfk = Tfg + Tfj; ri[WS(rs, 34)] = Tfd - Tfk; ri[WS(rs, 2)] = Tfd + Tfk; Thj = Tfm + Tfn; Thq = Thk + Thp; ii[WS(rs, 2)] = Thj + Thq; ii[WS(rs, 34)] = Thq - Thj; } { E Tfl, Tfo, Thr, Ths; Tfl = Tf9 - Tfc; Tfo = Tfm - Tfn; ri[WS(rs, 50)] = Tfl - Tfo; ri[WS(rs, 18)] = Tfl + Tfo; Thr = Tfj - Tfg; Ths = Thp - Thk; ii[WS(rs, 18)] = Thr + Ths; ii[WS(rs, 50)] = Ths - Thr; } } { E T6L, T9x, TiD, TiJ, T7E, TiI, T9A, TiA, T8y, T9K, T9u, T9E, T9r, T9L, T9v; E T9H; { E T6n, T6K, TiB, TiC; T6n = T6b - T6m; T6K = T6y - T6J; T6L = T6n - T6K; T9x = T6n + T6K; TiB = T9P - T9O; TiC = Tin - Tim; TiD = TiB + TiC; TiJ = TiC - TiB; } { E T7c, T9y, T7D, T9z; { E T72, T7b, T7t, T7C; T72 = T6Q - T71; T7b = T77 - T7a; T7c = FNMS(KP980785280, T7b, KP195090322 * T72); T9y = FMA(KP980785280, T72, KP195090322 * T7b); T7t = T7h - T7s; T7C = T7y - T7B; T7D = FMA(KP195090322, T7t, KP980785280 * T7C); T9z = FNMS(KP980785280, T7t, KP195090322 * T7C); } T7E = T7c - T7D; TiI = T9z - T9y; T9A = T9y + T9z; TiA = T7c + T7D; } { E T8k, T9C, T8x, T9D; { E T7W, T8j, T8t, T8w; T7W = T7K - T7V; T8j = T87 - T8i; T8k = T7W - T8j; T9C = T7W + T8j; T8t = T8p - T8s; T8w = T8u - T8v; T8x = T8t - T8w; T9D = T8t + T8w; } T8y = FMA(KP995184726, T8k, KP098017140 * T8x); T9K = FNMS(KP634393284, T9D, KP773010453 * T9C); T9u = FNMS(KP995184726, T8x, KP098017140 * T8k); T9E = FMA(KP634393284, T9C, KP773010453 * T9D); } { E T9d, T9F, T9q, T9G; { E T8P, T9c, T9m, T9p; T8P = T8D - T8O; T9c = T90 - T9b; T9d = T8P - T9c; T9F = T8P + T9c; T9m = T9i - T9l; T9p = T9n - T9o; T9q = T9m - T9p; T9G = T9m + T9p; } T9r = FNMS(KP995184726, T9q, KP098017140 * T9d); T9L = FMA(KP773010453, T9G, KP634393284 * T9F); T9v = FMA(KP098017140, T9q, KP995184726 * T9d); T9H = FNMS(KP634393284, T9G, KP773010453 * T9F); } { E T7F, T9s, TiH, TiK; T7F = T6L + T7E; T9s = T8y + T9r; ri[WS(rs, 47)] = T7F - T9s; ri[WS(rs, 15)] = T7F + T9s; TiH = T9u + T9v; TiK = TiI + TiJ; ii[WS(rs, 15)] = TiH + TiK; ii[WS(rs, 47)] = TiK - TiH; } { E T9t, T9w, TiL, TiM; T9t = T6L - T7E; T9w = T9u - T9v; ri[WS(rs, 63)] = T9t - T9w; ri[WS(rs, 31)] = T9t + T9w; TiL = T9r - T8y; TiM = TiJ - TiI; ii[WS(rs, 31)] = TiL + TiM; ii[WS(rs, 63)] = TiM - TiL; } { E T9B, T9I, Tiz, TiE; T9B = T9x + T9A; T9I = T9E + T9H; ri[WS(rs, 39)] = T9B - T9I; ri[WS(rs, 7)] = T9B + T9I; Tiz = T9K + T9L; TiE = TiA + TiD; ii[WS(rs, 7)] = Tiz + TiE; ii[WS(rs, 39)] = TiE - Tiz; } { E T9J, T9M, TiF, TiG; T9J = T9x - T9A; T9M = T9K - T9L; ri[WS(rs, 55)] = T9J - T9M; ri[WS(rs, 23)] = T9J + T9M; TiF = T9H - T9E; TiG = TiD - TiA; ii[WS(rs, 23)] = TiF + TiG; ii[WS(rs, 55)] = TiG - TiF; } } { E TaL, TbJ, Ti9, Tif, Tb0, Tie, TbM, Ti6, Tbk, TbW, TbG, TbQ, TbD, TbX, TbH; E TbT; { E TaD, TaK, Ti7, Ti8; TaD = Taz - TaC; TaK = TaG - TaJ; TaL = TaD - TaK; TbJ = TaD + TaK; Ti7 = Tc1 - Tc0; Ti8 = ThT - ThQ; Ti9 = Ti7 + Ti8; Tif = Ti8 - Ti7; } { E TaS, TbK, TaZ, TbL; { E TaO, TaR, TaV, TaY; TaO = TaM - TaN; TaR = TaP - TaQ; TaS = FNMS(KP831469612, TaR, KP555570233 * TaO); TbK = FMA(KP555570233, TaR, KP831469612 * TaO); TaV = TaT - TaU; TaY = TaW - TaX; TaZ = FMA(KP831469612, TaV, KP555570233 * TaY); TbL = FNMS(KP831469612, TaY, KP555570233 * TaV); } Tb0 = TaS - TaZ; Tie = TbL - TbK; TbM = TbK + TbL; Ti6 = TaS + TaZ; } { E Tbc, TbO, Tbj, TbP; { E Tb4, Tbb, Tbf, Tbi; Tb4 = Tb2 - Tb3; Tbb = Tb7 - Tba; Tbc = Tb4 - Tbb; TbO = Tb4 + Tbb; Tbf = Tbd - Tbe; Tbi = Tbg - Tbh; Tbj = Tbf - Tbi; TbP = Tbf + Tbi; } Tbk = FMA(KP956940335, Tbc, KP290284677 * Tbj); TbW = FNMS(KP471396736, TbP, KP881921264 * TbO); TbG = FNMS(KP956940335, Tbj, KP290284677 * Tbc); TbQ = FMA(KP471396736, TbO, KP881921264 * TbP); } { E Tbv, TbR, TbC, TbS; { E Tbn, Tbu, Tby, TbB; Tbn = Tbl - Tbm; Tbu = Tbq - Tbt; Tbv = Tbn - Tbu; TbR = Tbn + Tbu; Tby = Tbw - Tbx; TbB = Tbz - TbA; TbC = Tby - TbB; TbS = Tby + TbB; } TbD = FNMS(KP956940335, TbC, KP290284677 * Tbv); TbX = FMA(KP881921264, TbS, KP471396736 * TbR); TbH = FMA(KP290284677, TbC, KP956940335 * Tbv); TbT = FNMS(KP471396736, TbS, KP881921264 * TbR); } { E Tb1, TbE, Tid, Tig; Tb1 = TaL + Tb0; TbE = Tbk + TbD; ri[WS(rs, 45)] = Tb1 - TbE; ri[WS(rs, 13)] = Tb1 + TbE; Tid = TbG + TbH; Tig = Tie + Tif; ii[WS(rs, 13)] = Tid + Tig; ii[WS(rs, 45)] = Tig - Tid; } { E TbF, TbI, Tih, Tii; TbF = TaL - Tb0; TbI = TbG - TbH; ri[WS(rs, 61)] = TbF - TbI; ri[WS(rs, 29)] = TbF + TbI; Tih = TbD - Tbk; Tii = Tif - Tie; ii[WS(rs, 29)] = Tih + Tii; ii[WS(rs, 61)] = Tii - Tih; } { E TbN, TbU, Ti5, Tia; TbN = TbJ + TbM; TbU = TbQ + TbT; ri[WS(rs, 37)] = TbN - TbU; ri[WS(rs, 5)] = TbN + TbU; Ti5 = TbW + TbX; Tia = Ti6 + Ti9; ii[WS(rs, 5)] = Ti5 + Tia; ii[WS(rs, 37)] = Tia - Ti5; } { E TbV, TbY, Tib, Tic; TbV = TbJ - TbM; TbY = TbW - TbX; ri[WS(rs, 53)] = TbV - TbY; ri[WS(rs, 21)] = TbV + TbY; Tib = TbT - TbQ; Tic = Ti9 - Ti6; ii[WS(rs, 21)] = Tib + Tic; ii[WS(rs, 53)] = Tic - Tib; } } { E Tc3, Tcv, ThV, Ti1, Tca, Ti0, Tcy, ThO, Tci, TcI, Tcs, TcC, Tcp, TcJ, Tct; E TcF; { E TbZ, Tc2, ThP, ThU; TbZ = Taz + TaC; Tc2 = Tc0 + Tc1; Tc3 = TbZ - Tc2; Tcv = TbZ + Tc2; ThP = TaG + TaJ; ThU = ThQ + ThT; ThV = ThP + ThU; Ti1 = ThU - ThP; } { E Tc6, Tcw, Tc9, Tcx; { E Tc4, Tc5, Tc7, Tc8; Tc4 = TaM + TaN; Tc5 = TaP + TaQ; Tc6 = FNMS(KP195090322, Tc5, KP980785280 * Tc4); Tcw = FMA(KP980785280, Tc5, KP195090322 * Tc4); Tc7 = TaT + TaU; Tc8 = TaW + TaX; Tc9 = FMA(KP195090322, Tc7, KP980785280 * Tc8); Tcx = FNMS(KP195090322, Tc8, KP980785280 * Tc7); } Tca = Tc6 - Tc9; Ti0 = Tcx - Tcw; Tcy = Tcw + Tcx; ThO = Tc6 + Tc9; } { E Tce, TcA, Tch, TcB; { E Tcc, Tcd, Tcf, Tcg; Tcc = Tbd + Tbe; Tcd = Tba + Tb7; Tce = Tcc - Tcd; TcA = Tcc + Tcd; Tcf = Tb2 + Tb3; Tcg = Tbg + Tbh; Tch = Tcf - Tcg; TcB = Tcf + Tcg; } Tci = FMA(KP634393284, Tce, KP773010453 * Tch); TcI = FNMS(KP098017140, TcA, KP995184726 * TcB); Tcs = FNMS(KP773010453, Tce, KP634393284 * Tch); TcC = FMA(KP995184726, TcA, KP098017140 * TcB); } { E Tcl, TcD, Tco, TcE; { E Tcj, Tck, Tcm, Tcn; Tcj = Tbl + Tbm; Tck = TbA + Tbz; Tcl = Tcj - Tck; TcD = Tcj + Tck; Tcm = Tbw + Tbx; Tcn = Tbq + Tbt; Tco = Tcm - Tcn; TcE = Tcm + Tcn; } Tcp = FNMS(KP773010453, Tco, KP634393284 * Tcl); TcJ = FMA(KP098017140, TcD, KP995184726 * TcE); Tct = FMA(KP773010453, Tcl, KP634393284 * Tco); TcF = FNMS(KP098017140, TcE, KP995184726 * TcD); } { E Tcb, Tcq, ThZ, Ti2; Tcb = Tc3 + Tca; Tcq = Tci + Tcp; ri[WS(rs, 41)] = Tcb - Tcq; ri[WS(rs, 9)] = Tcb + Tcq; ThZ = Tcs + Tct; Ti2 = Ti0 + Ti1; ii[WS(rs, 9)] = ThZ + Ti2; ii[WS(rs, 41)] = Ti2 - ThZ; } { E Tcr, Tcu, Ti3, Ti4; Tcr = Tc3 - Tca; Tcu = Tcs - Tct; ri[WS(rs, 57)] = Tcr - Tcu; ri[WS(rs, 25)] = Tcr + Tcu; Ti3 = Tcp - Tci; Ti4 = Ti1 - Ti0; ii[WS(rs, 25)] = Ti3 + Ti4; ii[WS(rs, 57)] = Ti4 - Ti3; } { E Tcz, TcG, ThN, ThW; Tcz = Tcv + Tcy; TcG = TcC + TcF; ri[WS(rs, 33)] = Tcz - TcG; ri[WS(rs, 1)] = Tcz + TcG; ThN = TcI + TcJ; ThW = ThO + ThV; ii[WS(rs, 1)] = ThN + ThW; ii[WS(rs, 33)] = ThW - ThN; } { E TcH, TcK, ThX, ThY; TcH = Tcv - Tcy; TcK = TcI - TcJ; ri[WS(rs, 49)] = TcH - TcK; ri[WS(rs, 17)] = TcH + TcK; ThX = TcF - TcC; ThY = ThV - ThO; ii[WS(rs, 17)] = ThX + ThY; ii[WS(rs, 49)] = ThY - ThX; } } { E T9R, Taj, Tip, Tiv, T9Y, Tiu, Tam, Tik, Ta6, Taw, Tag, Taq, Tad, Tax, Tah; E Tat; { E T9N, T9Q, Til, Tio; T9N = T6b + T6m; T9Q = T9O + T9P; T9R = T9N - T9Q; Taj = T9N + T9Q; Til = T6y + T6J; Tio = Tim + Tin; Tip = Til + Tio; Tiv = Tio - Til; } { E T9U, Tak, T9X, Tal; { E T9S, T9T, T9V, T9W; T9S = T6Q + T71; T9T = T77 + T7a; T9U = FNMS(KP555570233, T9T, KP831469612 * T9S); Tak = FMA(KP555570233, T9S, KP831469612 * T9T); T9V = T7h + T7s; T9W = T7y + T7B; T9X = FMA(KP831469612, T9V, KP555570233 * T9W); Tal = FNMS(KP555570233, T9V, KP831469612 * T9W); } T9Y = T9U - T9X; Tiu = Tal - Tak; Tam = Tak + Tal; Tik = T9U + T9X; } { E Ta2, Tao, Ta5, Tap; { E Ta0, Ta1, Ta3, Ta4; Ta0 = T8p + T8s; Ta1 = T8i + T87; Ta2 = Ta0 - Ta1; Tao = Ta0 + Ta1; Ta3 = T7K + T7V; Ta4 = T8u + T8v; Ta5 = Ta3 - Ta4; Tap = Ta3 + Ta4; } Ta6 = FMA(KP471396736, Ta2, KP881921264 * Ta5); Taw = FNMS(KP290284677, Tao, KP956940335 * Tap); Tag = FNMS(KP881921264, Ta2, KP471396736 * Ta5); Taq = FMA(KP956940335, Tao, KP290284677 * Tap); } { E Ta9, Tar, Tac, Tas; { E Ta7, Ta8, Taa, Tab; Ta7 = T8D + T8O; Ta8 = T9o + T9n; Ta9 = Ta7 - Ta8; Tar = Ta7 + Ta8; Taa = T9i + T9l; Tab = T90 + T9b; Tac = Taa - Tab; Tas = Taa + Tab; } Tad = FNMS(KP881921264, Tac, KP471396736 * Ta9); Tax = FMA(KP290284677, Tar, KP956940335 * Tas); Tah = FMA(KP881921264, Ta9, KP471396736 * Tac); Tat = FNMS(KP290284677, Tas, KP956940335 * Tar); } { E T9Z, Tae, Tit, Tiw; T9Z = T9R + T9Y; Tae = Ta6 + Tad; ri[WS(rs, 43)] = T9Z - Tae; ri[WS(rs, 11)] = T9Z + Tae; Tit = Tag + Tah; Tiw = Tiu + Tiv; ii[WS(rs, 11)] = Tit + Tiw; ii[WS(rs, 43)] = Tiw - Tit; } { E Taf, Tai, Tix, Tiy; Taf = T9R - T9Y; Tai = Tag - Tah; ri[WS(rs, 59)] = Taf - Tai; ri[WS(rs, 27)] = Taf + Tai; Tix = Tad - Ta6; Tiy = Tiv - Tiu; ii[WS(rs, 27)] = Tix + Tiy; ii[WS(rs, 59)] = Tiy - Tix; } { E Tan, Tau, Tij, Tiq; Tan = Taj + Tam; Tau = Taq + Tat; ri[WS(rs, 35)] = Tan - Tau; ri[WS(rs, 3)] = Tan + Tau; Tij = Taw + Tax; Tiq = Tik + Tip; ii[WS(rs, 3)] = Tij + Tiq; ii[WS(rs, 35)] = Tiq - Tij; } { E Tav, Tay, Tir, Tis; Tav = Taj - Tam; Tay = Taw - Tax; ri[WS(rs, 51)] = Tav - Tay; ri[WS(rs, 19)] = Tav + Tay; Tir = Tat - Taq; Tis = Tip - Tik; ii[WS(rs, 19)] = Tir + Tis; ii[WS(rs, 51)] = Tis - Tir; } } } } } static const tw_instr twinstr[] = { {TW_FULL, 0, 64}, {TW_NEXT, 1, 0} }; static const ct_desc desc = { 64, "t1_64", twinstr, &GENUS, {808, 270, 230, 0}, 0, 0, 0 }; void X(codelet_t1_64) (planner *p) { X(kdft_dit_register) (p, t1_64, &desc); } #endif /* HAVE_FMA */ fftw-3.3.3/dft/scalar/codelets/t1_20.c0000644000175000017500000006674712054410252014243 00000000000000/* * Copyright (c) 2003, 2007-11 Matteo Frigo * Copyright (c) 2003, 2007-11 Massachusetts Institute of Technology * * This program is free software; you can redistribute it and/or modify * it under the terms of the GNU General Public License as published by * the Free Software Foundation; either version 2 of the License, or * (at your option) any later version. * * This program is distributed in the hope that it will be useful, * but WITHOUT ANY WARRANTY; without even the implied warranty of * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the * GNU General Public License for more details. * * You should have received a copy of the GNU General Public License * along with this program; if not, write to the Free Software * Foundation, Inc., 51 Franklin Street, Fifth Floor, Boston, MA 02110-1301 USA * */ /* This file was automatically generated --- DO NOT EDIT */ /* Generated on Sun Nov 25 07:35:53 EST 2012 */ #include "codelet-dft.h" #ifdef HAVE_FMA /* Generated by: ../../../genfft/gen_twiddle.native -fma -reorder-insns -schedule-for-pipeline -compact -variables 4 -pipeline-latency 4 -n 20 -name t1_20 -include t.h */ /* * This function contains 246 FP additions, 148 FP multiplications, * (or, 136 additions, 38 multiplications, 110 fused multiply/add), * 97 stack variables, 4 constants, and 80 memory accesses */ #include "t.h" static void t1_20(R *ri, R *ii, const R *W, stride rs, INT mb, INT me, INT ms) { DK(KP951056516, +0.951056516295153572116439333379382143405698634); DK(KP559016994, +0.559016994374947424102293417182819058860154590); DK(KP250000000, +0.250000000000000000000000000000000000000000000); DK(KP618033988, +0.618033988749894848204586834365638117720309180); { INT m; for (m = mb, W = W + (mb * 38); m < me; m = m + 1, ri = ri + ms, ii = ii + ms, W = W + 38, MAKE_VOLATILE_STRIDE(40, rs)) { E T4P, T4Y, T50, T4U, T4S, T4T, T4Z, T4V; { E T4N, T4r, T8, T2i, T4n, T2n, T4O, Tl, T2v, T3v, T40, T4b, TN, T2b, T3F; E T3i, T2R, T3z, T3W, T4f, T27, T2f, T3J, T3a, T2K, T3y, T3T, T4e, T1G, T2e; E T3I, T33, T2C, T3w, T43, T4c, T1e, T2c, T3G, T3p; { E T1, T4q, T3, T6, T2, T5; T1 = ri[0]; T4q = ii[0]; T3 = ri[WS(rs, 10)]; T6 = ii[WS(rs, 10)]; T2 = W[18]; T5 = W[19]; { E Ta, Td, Tg, T2j, Tb, Tj, Tf, Tc, Ti; { E T4o, T4, T9, T4p, T7; Ta = ri[WS(rs, 5)]; Td = ii[WS(rs, 5)]; T4o = T2 * T6; T4 = T2 * T3; T9 = W[8]; Tg = ri[WS(rs, 15)]; T4p = FNMS(T5, T3, T4o); T7 = FMA(T5, T6, T4); T2j = T9 * Td; Tb = T9 * Ta; T4N = T4q - T4p; T4r = T4p + T4q; T8 = T1 + T7; T2i = T1 - T7; Tj = ii[WS(rs, 15)]; Tf = W[28]; } Tc = W[9]; Ti = W[29]; { E T3d, Ts, T2t, TL, TB, TE, TD, T3f, Ty, T2q, TC; { E TH, TK, TJ, T2s, TI; { E To, Tr, Tp, T3c, Tq, TG; { E T2k, Te, T2m, Tk, T2l, Th, Tn; To = ri[WS(rs, 4)]; T2l = Tf * Tj; Th = Tf * Tg; T2k = FNMS(Tc, Ta, T2j); Te = FMA(Tc, Td, Tb); T2m = FNMS(Ti, Tg, T2l); Tk = FMA(Ti, Tj, Th); Tr = ii[WS(rs, 4)]; Tn = W[6]; T4n = T2k + T2m; T2n = T2k - T2m; T4O = Te - Tk; Tl = Te + Tk; Tp = Tn * To; T3c = Tn * Tr; } Tq = W[7]; TH = ri[WS(rs, 19)]; TK = ii[WS(rs, 19)]; TG = W[36]; T3d = FNMS(Tq, To, T3c); Ts = FMA(Tq, Tr, Tp); TJ = W[37]; T2s = TG * TK; TI = TG * TH; } { E Tu, Tx, Tt, Tw, T3e, Tv, TA; Tu = ri[WS(rs, 14)]; Tx = ii[WS(rs, 14)]; T2t = FNMS(TJ, TH, T2s); TL = FMA(TJ, TK, TI); Tt = W[26]; Tw = W[27]; TB = ri[WS(rs, 9)]; TE = ii[WS(rs, 9)]; T3e = Tt * Tx; Tv = Tt * Tu; TA = W[16]; TD = W[17]; T3f = FNMS(Tw, Tu, T3e); Ty = FMA(Tw, Tx, Tv); T2q = TA * TE; TC = TA * TB; } } { E T3g, T3Y, Tz, T2p, T2r, TF; T3g = T3d - T3f; T3Y = T3d + T3f; Tz = Ts + Ty; T2p = Ts - Ty; T2r = FNMS(TD, TB, T2q); TF = FMA(TD, TE, TC); { E T3Z, T2u, T3h, TM; T3Z = T2r + T2t; T2u = T2r - T2t; T3h = TF - TL; TM = TF + TL; T2v = T2p - T2u; T3v = T2p + T2u; T40 = T3Y - T3Z; T4b = T3Y + T3Z; TN = Tz - TM; T2b = Tz + TM; T3F = T3g - T3h; T3i = T3g + T3h; } } } } } { E T35, T1M, T2P, T25, T1V, T1Y, T1X, T37, T1S, T2M, T1W; { E T21, T24, T23, T2O, T22; { E T1I, T1L, T1H, T1K, T34, T1J, T20; T1I = ri[WS(rs, 12)]; T1L = ii[WS(rs, 12)]; T1H = W[22]; T1K = W[23]; T21 = ri[WS(rs, 7)]; T24 = ii[WS(rs, 7)]; T34 = T1H * T1L; T1J = T1H * T1I; T20 = W[12]; T23 = W[13]; T35 = FNMS(T1K, T1I, T34); T1M = FMA(T1K, T1L, T1J); T2O = T20 * T24; T22 = T20 * T21; } { E T1O, T1R, T1N, T1Q, T36, T1P, T1U; T1O = ri[WS(rs, 2)]; T1R = ii[WS(rs, 2)]; T2P = FNMS(T23, T21, T2O); T25 = FMA(T23, T24, T22); T1N = W[2]; T1Q = W[3]; T1V = ri[WS(rs, 17)]; T1Y = ii[WS(rs, 17)]; T36 = T1N * T1R; T1P = T1N * T1O; T1U = W[32]; T1X = W[33]; T37 = FNMS(T1Q, T1O, T36); T1S = FMA(T1Q, T1R, T1P); T2M = T1U * T1Y; T1W = T1U * T1V; } } { E T38, T3U, T1T, T2L, T2N, T1Z; T38 = T35 - T37; T3U = T35 + T37; T1T = T1M + T1S; T2L = T1M - T1S; T2N = FNMS(T1X, T1V, T2M); T1Z = FMA(T1X, T1Y, T1W); { E T3V, T2Q, T39, T26; T3V = T2N + T2P; T2Q = T2N - T2P; T39 = T1Z - T25; T26 = T1Z + T25; T2R = T2L - T2Q; T3z = T2L + T2Q; T3W = T3U - T3V; T4f = T3U + T3V; T27 = T1T - T26; T2f = T1T + T26; T3J = T38 - T39; T3a = T38 + T39; } } } { E T2Y, T1l, T2I, T1E, T1u, T1x, T1w, T30, T1r, T2F, T1v; { E T1A, T1D, T1C, T2H, T1B; { E T1h, T1k, T1g, T1j, T2X, T1i, T1z; T1h = ri[WS(rs, 8)]; T1k = ii[WS(rs, 8)]; T1g = W[14]; T1j = W[15]; T1A = ri[WS(rs, 3)]; T1D = ii[WS(rs, 3)]; T2X = T1g * T1k; T1i = T1g * T1h; T1z = W[4]; T1C = W[5]; T2Y = FNMS(T1j, T1h, T2X); T1l = FMA(T1j, T1k, T1i); T2H = T1z * T1D; T1B = T1z * T1A; } { E T1n, T1q, T1m, T1p, T2Z, T1o, T1t; T1n = ri[WS(rs, 18)]; T1q = ii[WS(rs, 18)]; T2I = FNMS(T1C, T1A, T2H); T1E = FMA(T1C, T1D, T1B); T1m = W[34]; T1p = W[35]; T1u = ri[WS(rs, 13)]; T1x = ii[WS(rs, 13)]; T2Z = T1m * T1q; T1o = T1m * T1n; T1t = W[24]; T1w = W[25]; T30 = FNMS(T1p, T1n, T2Z); T1r = FMA(T1p, T1q, T1o); T2F = T1t * T1x; T1v = T1t * T1u; } } { E T31, T3R, T1s, T2E, T2G, T1y; T31 = T2Y - T30; T3R = T2Y + T30; T1s = T1l + T1r; T2E = T1l - T1r; T2G = FNMS(T1w, T1u, T2F); T1y = FMA(T1w, T1x, T1v); { E T3S, T2J, T32, T1F; T3S = T2G + T2I; T2J = T2G - T2I; T32 = T1y - T1E; T1F = T1y + T1E; T2K = T2E - T2J; T3y = T2E + T2J; T3T = T3R - T3S; T4e = T3R + T3S; T1G = T1s - T1F; T2e = T1s + T1F; T3I = T31 - T32; T33 = T31 + T32; } } } { E T3k, TT, T2A, T1c, T12, T15, T14, T3m, TZ, T2x, T13; { E T18, T1b, T1a, T2z, T19; { E TP, TS, TO, TR, T3j, TQ, T17; TP = ri[WS(rs, 16)]; TS = ii[WS(rs, 16)]; TO = W[30]; TR = W[31]; T18 = ri[WS(rs, 11)]; T1b = ii[WS(rs, 11)]; T3j = TO * TS; TQ = TO * TP; T17 = W[20]; T1a = W[21]; T3k = FNMS(TR, TP, T3j); TT = FMA(TR, TS, TQ); T2z = T17 * T1b; T19 = T17 * T18; } { E TV, TY, TU, TX, T3l, TW, T11; TV = ri[WS(rs, 6)]; TY = ii[WS(rs, 6)]; T2A = FNMS(T1a, T18, T2z); T1c = FMA(T1a, T1b, T19); TU = W[10]; TX = W[11]; T12 = ri[WS(rs, 1)]; T15 = ii[WS(rs, 1)]; T3l = TU * TY; TW = TU * TV; T11 = W[0]; T14 = W[1]; T3m = FNMS(TX, TV, T3l); TZ = FMA(TX, TY, TW); T2x = T11 * T15; T13 = T11 * T12; } } { E T3n, T41, T10, T2w, T2y, T16; T3n = T3k - T3m; T41 = T3k + T3m; T10 = TT + TZ; T2w = TT - TZ; T2y = FNMS(T14, T12, T2x); T16 = FMA(T14, T15, T13); { E T42, T2B, T3o, T1d; T42 = T2y + T2A; T2B = T2y - T2A; T3o = T16 - T1c; T1d = T16 + T1c; T2C = T2w - T2B; T3w = T2w + T2B; T43 = T41 - T42; T4c = T41 + T42; T1e = T10 - T1d; T2c = T10 + T1d; T3G = T3n - T3o; T3p = T3n + T3o; } } } { E T4s, T4k, T4l, T4h, T4j, T49, T4y, T4A, T48; { E T4D, T4C, T2a, T47, T45, T4B, T4M, T4K, T46, T3Q; { E Tm, T1f, T4J, T4I, T28, T3X, T44, T29, T3P, T3O; T4D = T3T + T3W; T3X = T3T - T3W; T44 = T40 - T43; T4C = T40 + T43; T2a = T8 + Tl; Tm = T8 - Tl; T1f = TN + T1e; T4J = TN - T1e; T4I = T1G - T27; T28 = T1G + T27; T47 = FMA(KP618033988, T3X, T44); T45 = FNMS(KP618033988, T44, T3X); T29 = T1f + T28; T3P = T1f - T28; T4B = T4r - T4n; T4s = T4n + T4r; ri[WS(rs, 10)] = Tm + T29; T3O = FNMS(KP250000000, T29, Tm); T4M = FMA(KP618033988, T4I, T4J); T4K = FNMS(KP618033988, T4J, T4I); T46 = FMA(KP559016994, T3P, T3O); T3Q = FNMS(KP559016994, T3P, T3O); } { E T2d, T4w, T4x, T2g, T2h; { E T4d, T4G, T4F, T4g, T4E, T4L, T4H; T4k = T4b + T4c; T4d = T4b - T4c; T4G = T4C - T4D; T4E = T4C + T4D; ri[WS(rs, 18)] = FMA(KP951056516, T45, T3Q); ri[WS(rs, 2)] = FNMS(KP951056516, T45, T3Q); ri[WS(rs, 6)] = FMA(KP951056516, T47, T46); ri[WS(rs, 14)] = FNMS(KP951056516, T47, T46); ii[WS(rs, 10)] = T4E + T4B; T4F = FNMS(KP250000000, T4E, T4B); T4g = T4e - T4f; T4l = T4e + T4f; T2d = T2b + T2c; T4w = T2b - T2c; T4L = FMA(KP559016994, T4G, T4F); T4H = FNMS(KP559016994, T4G, T4F); T4h = FMA(KP618033988, T4g, T4d); T4j = FNMS(KP618033988, T4d, T4g); ii[WS(rs, 18)] = FNMS(KP951056516, T4K, T4H); ii[WS(rs, 2)] = FMA(KP951056516, T4K, T4H); ii[WS(rs, 14)] = FMA(KP951056516, T4M, T4L); ii[WS(rs, 6)] = FNMS(KP951056516, T4M, T4L); T4x = T2e - T2f; T2g = T2e + T2f; } T2h = T2d + T2g; T49 = T2d - T2g; T4y = FMA(KP618033988, T4x, T4w); T4A = FNMS(KP618033988, T4w, T4x); ri[0] = T2a + T2h; T48 = FNMS(KP250000000, T2h, T2a); } } { E T3u, T51, T5a, T5c, T56, T54; { E T53, T52, T3t, T3r, T2o, T59, T58, T2T, T2V, T4u, T4t, T2U, T3s, T2W; { E T3b, T3q, T4i, T4a, T4m; T53 = T33 + T3a; T3b = T33 - T3a; T3q = T3i - T3p; T52 = T3i + T3p; T4i = FNMS(KP559016994, T49, T48); T4a = FMA(KP559016994, T49, T48); T4m = T4k + T4l; T4u = T4k - T4l; ri[WS(rs, 16)] = FMA(KP951056516, T4h, T4a); ri[WS(rs, 4)] = FNMS(KP951056516, T4h, T4a); ri[WS(rs, 8)] = FMA(KP951056516, T4j, T4i); ri[WS(rs, 12)] = FNMS(KP951056516, T4j, T4i); ii[0] = T4m + T4s; T4t = FNMS(KP250000000, T4m, T4s); T3t = FMA(KP618033988, T3b, T3q); T3r = FNMS(KP618033988, T3q, T3b); } T3u = T2i + T2n; T2o = T2i - T2n; { E T4v, T4z, T2D, T2S; T4v = FMA(KP559016994, T4u, T4t); T4z = FNMS(KP559016994, T4u, T4t); T2D = T2v + T2C; T59 = T2v - T2C; T58 = T2K - T2R; T2S = T2K + T2R; ii[WS(rs, 16)] = FNMS(KP951056516, T4y, T4v); ii[WS(rs, 4)] = FMA(KP951056516, T4y, T4v); ii[WS(rs, 12)] = FMA(KP951056516, T4A, T4z); ii[WS(rs, 8)] = FNMS(KP951056516, T4A, T4z); T2T = T2D + T2S; T2V = T2D - T2S; } ri[WS(rs, 15)] = T2o + T2T; T2U = FNMS(KP250000000, T2T, T2o); T51 = T4O + T4N; T4P = T4N - T4O; T5a = FNMS(KP618033988, T59, T58); T5c = FMA(KP618033988, T58, T59); T3s = FMA(KP559016994, T2V, T2U); T2W = FNMS(KP559016994, T2V, T2U); ri[WS(rs, 7)] = FNMS(KP951056516, T3r, T2W); ri[WS(rs, 3)] = FMA(KP951056516, T3r, T2W); ri[WS(rs, 19)] = FNMS(KP951056516, T3t, T3s); ri[WS(rs, 11)] = FMA(KP951056516, T3t, T3s); T56 = T52 - T53; T54 = T52 + T53; } { E T4Q, T4R, T3N, T3L, T4W, T4X, T3B, T3D, T3H, T3K, T55, T3C, T3M, T3E; T4Q = T3F + T3G; T3H = T3F - T3G; T3K = T3I - T3J; T4R = T3I + T3J; ii[WS(rs, 15)] = T54 + T51; T55 = FNMS(KP250000000, T54, T51); T3N = FNMS(KP618033988, T3H, T3K); T3L = FMA(KP618033988, T3K, T3H); { E T57, T5b, T3x, T3A; T57 = FNMS(KP559016994, T56, T55); T5b = FMA(KP559016994, T56, T55); T3x = T3v + T3w; T4W = T3v - T3w; T4X = T3y - T3z; T3A = T3y + T3z; ii[WS(rs, 7)] = FMA(KP951056516, T5a, T57); ii[WS(rs, 3)] = FNMS(KP951056516, T5a, T57); ii[WS(rs, 19)] = FMA(KP951056516, T5c, T5b); ii[WS(rs, 11)] = FNMS(KP951056516, T5c, T5b); T3B = T3x + T3A; T3D = T3x - T3A; } ri[WS(rs, 5)] = T3u + T3B; T3C = FNMS(KP250000000, T3B, T3u); T4Y = FMA(KP618033988, T4X, T4W); T50 = FNMS(KP618033988, T4W, T4X); T3M = FNMS(KP559016994, T3D, T3C); T3E = FMA(KP559016994, T3D, T3C); ri[WS(rs, 9)] = FNMS(KP951056516, T3L, T3E); ri[WS(rs, 1)] = FMA(KP951056516, T3L, T3E); ri[WS(rs, 17)] = FNMS(KP951056516, T3N, T3M); ri[WS(rs, 13)] = FMA(KP951056516, T3N, T3M); T4U = T4Q - T4R; T4S = T4Q + T4R; } } } } ii[WS(rs, 5)] = T4S + T4P; T4T = FNMS(KP250000000, T4S, T4P); T4Z = FNMS(KP559016994, T4U, T4T); T4V = FMA(KP559016994, T4U, T4T); ii[WS(rs, 9)] = FMA(KP951056516, T4Y, T4V); ii[WS(rs, 1)] = FNMS(KP951056516, T4Y, T4V); ii[WS(rs, 17)] = FMA(KP951056516, T50, T4Z); ii[WS(rs, 13)] = FNMS(KP951056516, T50, T4Z); } } } static const tw_instr twinstr[] = { {TW_FULL, 0, 20}, {TW_NEXT, 1, 0} }; static const ct_desc desc = { 20, "t1_20", twinstr, &GENUS, {136, 38, 110, 0}, 0, 0, 0 }; void X(codelet_t1_20) (planner *p) { X(kdft_dit_register) (p, t1_20, &desc); } #else /* HAVE_FMA */ /* Generated by: ../../../genfft/gen_twiddle.native -compact -variables 4 -pipeline-latency 4 -n 20 -name t1_20 -include t.h */ /* * This function contains 246 FP additions, 124 FP multiplications, * (or, 184 additions, 62 multiplications, 62 fused multiply/add), * 85 stack variables, 4 constants, and 80 memory accesses */ #include "t.h" static void t1_20(R *ri, R *ii, const R *W, stride rs, INT mb, INT me, INT ms) { DK(KP587785252, +0.587785252292473129168705954639072768597652438); DK(KP951056516, +0.951056516295153572116439333379382143405698634); DK(KP250000000, +0.250000000000000000000000000000000000000000000); DK(KP559016994, +0.559016994374947424102293417182819058860154590); { INT m; for (m = mb, W = W + (mb * 38); m < me; m = m + 1, ri = ri + ms, ii = ii + ms, W = W + 38, MAKE_VOLATILE_STRIDE(40, rs)) { E Tj, T1R, T4g, T4p, T2q, T37, T3Q, T42, T1r, T1O, T1P, T3i, T3l, T44, T3D; E T3E, T3K, T1V, T1W, T1X, T23, T28, T4r, T2W, T2X, T4c, T33, T34, T35, T2G; E T2L, T2M, TG, T13, T14, T3p, T3s, T43, T3A, T3B, T3J, T1S, T1T, T1U, T2e; E T2j, T4q, T2T, T2U, T4b, T30, T31, T32, T2v, T2A, T2B; { E T1, T3O, T6, T3N, Tc, T2n, Th, T2o; T1 = ri[0]; T3O = ii[0]; { E T3, T5, T2, T4; T3 = ri[WS(rs, 10)]; T5 = ii[WS(rs, 10)]; T2 = W[18]; T4 = W[19]; T6 = FMA(T2, T3, T4 * T5); T3N = FNMS(T4, T3, T2 * T5); } { E T9, Tb, T8, Ta; T9 = ri[WS(rs, 5)]; Tb = ii[WS(rs, 5)]; T8 = W[8]; Ta = W[9]; Tc = FMA(T8, T9, Ta * Tb); T2n = FNMS(Ta, T9, T8 * Tb); } { E Te, Tg, Td, Tf; Te = ri[WS(rs, 15)]; Tg = ii[WS(rs, 15)]; Td = W[28]; Tf = W[29]; Th = FMA(Td, Te, Tf * Tg); T2o = FNMS(Tf, Te, Td * Tg); } { E T7, Ti, T4e, T4f; T7 = T1 + T6; Ti = Tc + Th; Tj = T7 - Ti; T1R = T7 + Ti; T4e = T3O - T3N; T4f = Tc - Th; T4g = T4e - T4f; T4p = T4f + T4e; } { E T2m, T2p, T3M, T3P; T2m = T1 - T6; T2p = T2n - T2o; T2q = T2m - T2p; T37 = T2m + T2p; T3M = T2n + T2o; T3P = T3N + T3O; T3Q = T3M + T3P; T42 = T3P - T3M; } } { E T1f, T3g, T21, T2C, T1N, T3k, T27, T2K, T1q, T3h, T22, T2F, T1C, T3j, T26; E T2H; { E T19, T1Z, T1e, T20; { E T16, T18, T15, T17; T16 = ri[WS(rs, 8)]; T18 = ii[WS(rs, 8)]; T15 = W[14]; T17 = W[15]; T19 = FMA(T15, T16, T17 * T18); T1Z = FNMS(T17, T16, T15 * T18); } { E T1b, T1d, T1a, T1c; T1b = ri[WS(rs, 18)]; T1d = ii[WS(rs, 18)]; T1a = W[34]; T1c = W[35]; T1e = FMA(T1a, T1b, T1c * T1d); T20 = FNMS(T1c, T1b, T1a * T1d); } T1f = T19 + T1e; T3g = T1Z + T20; T21 = T1Z - T20; T2C = T19 - T1e; } { E T1H, T2I, T1M, T2J; { E T1E, T1G, T1D, T1F; T1E = ri[WS(rs, 17)]; T1G = ii[WS(rs, 17)]; T1D = W[32]; T1F = W[33]; T1H = FMA(T1D, T1E, T1F * T1G); T2I = FNMS(T1F, T1E, T1D * T1G); } { E T1J, T1L, T1I, T1K; T1J = ri[WS(rs, 7)]; T1L = ii[WS(rs, 7)]; T1I = W[12]; T1K = W[13]; T1M = FMA(T1I, T1J, T1K * T1L); T2J = FNMS(T1K, T1J, T1I * T1L); } T1N = T1H + T1M; T3k = T2I + T2J; T27 = T1H - T1M; T2K = T2I - T2J; } { E T1k, T2D, T1p, T2E; { E T1h, T1j, T1g, T1i; T1h = ri[WS(rs, 13)]; T1j = ii[WS(rs, 13)]; T1g = W[24]; T1i = W[25]; T1k = FMA(T1g, T1h, T1i * T1j); T2D = FNMS(T1i, T1h, T1g * T1j); } { E T1m, T1o, T1l, T1n; T1m = ri[WS(rs, 3)]; T1o = ii[WS(rs, 3)]; T1l = W[4]; T1n = W[5]; T1p = FMA(T1l, T1m, T1n * T1o); T2E = FNMS(T1n, T1m, T1l * T1o); } T1q = T1k + T1p; T3h = T2D + T2E; T22 = T1k - T1p; T2F = T2D - T2E; } { E T1w, T24, T1B, T25; { E T1t, T1v, T1s, T1u; T1t = ri[WS(rs, 12)]; T1v = ii[WS(rs, 12)]; T1s = W[22]; T1u = W[23]; T1w = FMA(T1s, T1t, T1u * T1v); T24 = FNMS(T1u, T1t, T1s * T1v); } { E T1y, T1A, T1x, T1z; T1y = ri[WS(rs, 2)]; T1A = ii[WS(rs, 2)]; T1x = W[2]; T1z = W[3]; T1B = FMA(T1x, T1y, T1z * T1A); T25 = FNMS(T1z, T1y, T1x * T1A); } T1C = T1w + T1B; T3j = T24 + T25; T26 = T24 - T25; T2H = T1w - T1B; } T1r = T1f - T1q; T1O = T1C - T1N; T1P = T1r + T1O; T3i = T3g - T3h; T3l = T3j - T3k; T44 = T3i + T3l; T3D = T3g + T3h; T3E = T3j + T3k; T3K = T3D + T3E; T1V = T1f + T1q; T1W = T1C + T1N; T1X = T1V + T1W; T23 = T21 + T22; T28 = T26 + T27; T4r = T23 + T28; T2W = T21 - T22; T2X = T26 - T27; T4c = T2W + T2X; T33 = T2C + T2F; T34 = T2H + T2K; T35 = T33 + T34; T2G = T2C - T2F; T2L = T2H - T2K; T2M = T2G + T2L; } { E Tu, T3n, T2c, T2r, T12, T3r, T2i, T2z, TF, T3o, T2d, T2u, TR, T3q, T2h; E T2w; { E To, T2a, Tt, T2b; { E Tl, Tn, Tk, Tm; Tl = ri[WS(rs, 4)]; Tn = ii[WS(rs, 4)]; Tk = W[6]; Tm = W[7]; To = FMA(Tk, Tl, Tm * Tn); T2a = FNMS(Tm, Tl, Tk * Tn); } { E Tq, Ts, Tp, Tr; Tq = ri[WS(rs, 14)]; Ts = ii[WS(rs, 14)]; Tp = W[26]; Tr = W[27]; Tt = FMA(Tp, Tq, Tr * Ts); T2b = FNMS(Tr, Tq, Tp * Ts); } Tu = To + Tt; T3n = T2a + T2b; T2c = T2a - T2b; T2r = To - Tt; } { E TW, T2x, T11, T2y; { E TT, TV, TS, TU; TT = ri[WS(rs, 1)]; TV = ii[WS(rs, 1)]; TS = W[0]; TU = W[1]; TW = FMA(TS, TT, TU * TV); T2x = FNMS(TU, TT, TS * TV); } { E TY, T10, TX, TZ; TY = ri[WS(rs, 11)]; T10 = ii[WS(rs, 11)]; TX = W[20]; TZ = W[21]; T11 = FMA(TX, TY, TZ * T10); T2y = FNMS(TZ, TY, TX * T10); } T12 = TW + T11; T3r = T2x + T2y; T2i = TW - T11; T2z = T2x - T2y; } { E Tz, T2s, TE, T2t; { E Tw, Ty, Tv, Tx; Tw = ri[WS(rs, 9)]; Ty = ii[WS(rs, 9)]; Tv = W[16]; Tx = W[17]; Tz = FMA(Tv, Tw, Tx * Ty); T2s = FNMS(Tx, Tw, Tv * Ty); } { E TB, TD, TA, TC; TB = ri[WS(rs, 19)]; TD = ii[WS(rs, 19)]; TA = W[36]; TC = W[37]; TE = FMA(TA, TB, TC * TD); T2t = FNMS(TC, TB, TA * TD); } TF = Tz + TE; T3o = T2s + T2t; T2d = Tz - TE; T2u = T2s - T2t; } { E TL, T2f, TQ, T2g; { E TI, TK, TH, TJ; TI = ri[WS(rs, 16)]; TK = ii[WS(rs, 16)]; TH = W[30]; TJ = W[31]; TL = FMA(TH, TI, TJ * TK); T2f = FNMS(TJ, TI, TH * TK); } { E TN, TP, TM, TO; TN = ri[WS(rs, 6)]; TP = ii[WS(rs, 6)]; TM = W[10]; TO = W[11]; TQ = FMA(TM, TN, TO * TP); T2g = FNMS(TO, TN, TM * TP); } TR = TL + TQ; T3q = T2f + T2g; T2h = T2f - T2g; T2w = TL - TQ; } TG = Tu - TF; T13 = TR - T12; T14 = TG + T13; T3p = T3n - T3o; T3s = T3q - T3r; T43 = T3p + T3s; T3A = T3n + T3o; T3B = T3q + T3r; T3J = T3A + T3B; T1S = Tu + TF; T1T = TR + T12; T1U = T1S + T1T; T2e = T2c + T2d; T2j = T2h + T2i; T4q = T2e + T2j; T2T = T2c - T2d; T2U = T2h - T2i; T4b = T2T + T2U; T30 = T2r + T2u; T31 = T2w + T2z; T32 = T30 + T31; T2v = T2r - T2u; T2A = T2w - T2z; T2B = T2v + T2A; } { E T3e, T1Q, T3d, T3u, T3w, T3m, T3t, T3v, T3f; T3e = KP559016994 * (T14 - T1P); T1Q = T14 + T1P; T3d = FNMS(KP250000000, T1Q, Tj); T3m = T3i - T3l; T3t = T3p - T3s; T3u = FNMS(KP587785252, T3t, KP951056516 * T3m); T3w = FMA(KP951056516, T3t, KP587785252 * T3m); ri[WS(rs, 10)] = Tj + T1Q; T3v = T3e + T3d; ri[WS(rs, 14)] = T3v - T3w; ri[WS(rs, 6)] = T3v + T3w; T3f = T3d - T3e; ri[WS(rs, 2)] = T3f - T3u; ri[WS(rs, 18)] = T3f + T3u; } { E T47, T45, T46, T41, T4a, T3Z, T40, T49, T48; T47 = KP559016994 * (T43 - T44); T45 = T43 + T44; T46 = FNMS(KP250000000, T45, T42); T3Z = T1r - T1O; T40 = TG - T13; T41 = FNMS(KP587785252, T40, KP951056516 * T3Z); T4a = FMA(KP951056516, T40, KP587785252 * T3Z); ii[WS(rs, 10)] = T45 + T42; T49 = T47 + T46; ii[WS(rs, 6)] = T49 - T4a; ii[WS(rs, 14)] = T4a + T49; T48 = T46 - T47; ii[WS(rs, 2)] = T41 + T48; ii[WS(rs, 18)] = T48 - T41; } { E T3x, T1Y, T3y, T3G, T3I, T3C, T3F, T3H, T3z; T3x = KP559016994 * (T1U - T1X); T1Y = T1U + T1X; T3y = FNMS(KP250000000, T1Y, T1R); T3C = T3A - T3B; T3F = T3D - T3E; T3G = FMA(KP951056516, T3C, KP587785252 * T3F); T3I = FNMS(KP587785252, T3C, KP951056516 * T3F); ri[0] = T1R + T1Y; T3H = T3y - T3x; ri[WS(rs, 12)] = T3H - T3I; ri[WS(rs, 8)] = T3H + T3I; T3z = T3x + T3y; ri[WS(rs, 4)] = T3z - T3G; ri[WS(rs, 16)] = T3z + T3G; } { E T3U, T3L, T3V, T3T, T3Y, T3R, T3S, T3X, T3W; T3U = KP559016994 * (T3J - T3K); T3L = T3J + T3K; T3V = FNMS(KP250000000, T3L, T3Q); T3R = T1S - T1T; T3S = T1V - T1W; T3T = FMA(KP951056516, T3R, KP587785252 * T3S); T3Y = FNMS(KP587785252, T3R, KP951056516 * T3S); ii[0] = T3L + T3Q; T3X = T3V - T3U; ii[WS(rs, 8)] = T3X - T3Y; ii[WS(rs, 12)] = T3Y + T3X; T3W = T3U + T3V; ii[WS(rs, 4)] = T3T + T3W; ii[WS(rs, 16)] = T3W - T3T; } { E T2P, T2N, T2O, T2l, T2R, T29, T2k, T2S, T2Q; T2P = KP559016994 * (T2B - T2M); T2N = T2B + T2M; T2O = FNMS(KP250000000, T2N, T2q); T29 = T23 - T28; T2k = T2e - T2j; T2l = FNMS(KP587785252, T2k, KP951056516 * T29); T2R = FMA(KP951056516, T2k, KP587785252 * T29); ri[WS(rs, 15)] = T2q + T2N; T2S = T2P + T2O; ri[WS(rs, 11)] = T2R + T2S; ri[WS(rs, 19)] = T2S - T2R; T2Q = T2O - T2P; ri[WS(rs, 3)] = T2l + T2Q; ri[WS(rs, 7)] = T2Q - T2l; } { E T4u, T4s, T4t, T4y, T4A, T4w, T4x, T4z, T4v; T4u = KP559016994 * (T4q - T4r); T4s = T4q + T4r; T4t = FNMS(KP250000000, T4s, T4p); T4w = T2G - T2L; T4x = T2v - T2A; T4y = FNMS(KP587785252, T4x, KP951056516 * T4w); T4A = FMA(KP951056516, T4x, KP587785252 * T4w); ii[WS(rs, 15)] = T4s + T4p; T4z = T4u + T4t; ii[WS(rs, 11)] = T4z - T4A; ii[WS(rs, 19)] = T4A + T4z; T4v = T4t - T4u; ii[WS(rs, 3)] = T4v - T4y; ii[WS(rs, 7)] = T4y + T4v; } { E T36, T38, T39, T2Z, T3b, T2V, T2Y, T3c, T3a; T36 = KP559016994 * (T32 - T35); T38 = T32 + T35; T39 = FNMS(KP250000000, T38, T37); T2V = T2T - T2U; T2Y = T2W - T2X; T2Z = FMA(KP951056516, T2V, KP587785252 * T2Y); T3b = FNMS(KP587785252, T2V, KP951056516 * T2Y); ri[WS(rs, 5)] = T37 + T38; T3c = T39 - T36; ri[WS(rs, 13)] = T3b + T3c; ri[WS(rs, 17)] = T3c - T3b; T3a = T36 + T39; ri[WS(rs, 1)] = T2Z + T3a; ri[WS(rs, 9)] = T3a - T2Z; } { E T4d, T4h, T4i, T4m, T4o, T4k, T4l, T4n, T4j; T4d = KP559016994 * (T4b - T4c); T4h = T4b + T4c; T4i = FNMS(KP250000000, T4h, T4g); T4k = T30 - T31; T4l = T33 - T34; T4m = FMA(KP951056516, T4k, KP587785252 * T4l); T4o = FNMS(KP587785252, T4k, KP951056516 * T4l); ii[WS(rs, 5)] = T4h + T4g; T4n = T4i - T4d; ii[WS(rs, 13)] = T4n - T4o; ii[WS(rs, 17)] = T4o + T4n; T4j = T4d + T4i; ii[WS(rs, 1)] = T4j - T4m; ii[WS(rs, 9)] = T4m + T4j; } } } } static const tw_instr twinstr[] = { {TW_FULL, 0, 20}, {TW_NEXT, 1, 0} }; static const ct_desc desc = { 20, "t1_20", twinstr, &GENUS, {184, 62, 62, 0}, 0, 0, 0 }; void X(codelet_t1_20) (planner *p) { X(kdft_dit_register) (p, t1_20, &desc); } #endif /* HAVE_FMA */ fftw-3.3.3/dft/scalar/codelets/t1_25.c0000644000175000017500000014110612054410260014226 00000000000000/* * Copyright (c) 2003, 2007-11 Matteo Frigo * Copyright (c) 2003, 2007-11 Massachusetts Institute of Technology * * This program is free software; you can redistribute it and/or modify * it under the terms of the GNU General Public License as published by * the Free Software Foundation; either version 2 of the License, or * (at your option) any later version. * * This program is distributed in the hope that it will be useful, * but WITHOUT ANY WARRANTY; without even the implied warranty of * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the * GNU General Public License for more details. * * You should have received a copy of the GNU General Public License * along with this program; if not, write to the Free Software * Foundation, Inc., 51 Franklin Street, Fifth Floor, Boston, MA 02110-1301 USA * */ /* This file was automatically generated --- DO NOT EDIT */ /* Generated on Sun Nov 25 07:35:54 EST 2012 */ #include "codelet-dft.h" #ifdef HAVE_FMA /* Generated by: ../../../genfft/gen_twiddle.native -fma -reorder-insns -schedule-for-pipeline -compact -variables 4 -pipeline-latency 4 -n 25 -name t1_25 -include t.h */ /* * This function contains 400 FP additions, 364 FP multiplications, * (or, 84 additions, 48 multiplications, 316 fused multiply/add), * 181 stack variables, 47 constants, and 100 memory accesses */ #include "t.h" static void t1_25(R *ri, R *ii, const R *W, stride rs, INT mb, INT me, INT ms) { DK(KP860541664, +0.860541664367944677098261680920518816412804187); DK(KP681693190, +0.681693190061530575150324149145440022633095390); DK(KP560319534, +0.560319534973832390111614715371676131169633784); DK(KP949179823, +0.949179823508441261575555465843363271711583843); DK(KP557913902, +0.557913902031834264187699648465567037992437152); DK(KP249506682, +0.249506682107067890488084201715862638334226305); DK(KP906616052, +0.906616052148196230441134447086066874408359177); DK(KP968479752, +0.968479752739016373193524836781420152702090879); DK(KP621716863, +0.621716863012209892444754556304102309693593202); DK(KP614372930, +0.614372930789563808870829930444362096004872855); DK(KP845997307, +0.845997307939530944175097360758058292389769300); DK(KP998026728, +0.998026728428271561952336806863450553336905220); DK(KP994076283, +0.994076283785401014123185814696322018529298887); DK(KP734762448, +0.734762448793050413546343770063151342619912334); DK(KP772036680, +0.772036680810363904029489473607579825330539880); DK(KP062914667, +0.062914667253649757225485955897349402364686947); DK(KP803003575, +0.803003575438660414833440593570376004635464850); DK(KP943557151, +0.943557151597354104399655195398983005179443399); DK(KP554608978, +0.554608978404018097464974850792216217022558774); DK(KP248028675, +0.248028675328619457762448260696444630363259177); DK(KP726211448, +0.726211448929902658173535992263577167607493062); DK(KP525970792, +0.525970792408939708442463226536226366643874659); DK(KP921177326, +0.921177326965143320250447435415066029359282231); DK(KP833417178, +0.833417178328688677408962550243238843138996060); DK(KP541454447, +0.541454447536312777046285590082819509052033189); DK(KP992114701, +0.992114701314477831049793042785778521453036709); DK(KP242145790, +0.242145790282157779872542093866183953459003101); DK(KP851038619, +0.851038619207379630836264138867114231259902550); DK(KP912575812, +0.912575812670962425556968549836277086778922727); DK(KP559154169, +0.559154169276087864842202529084232643714075927); DK(KP683113946, +0.683113946453479238701949862233725244439656928); DK(KP912018591, +0.912018591466481957908415381764119056233607330); DK(KP968583161, +0.968583161128631119490168375464735813836012403); DK(KP470564281, +0.470564281212251493087595091036643380879947982); DK(KP827271945, +0.827271945972475634034355757144307982555673741); DK(KP904730450, +0.904730450839922351881287709692877908104763647); DK(KP126329378, +0.126329378446108174786050455341811215027378105); DK(KP831864738, +0.831864738706457140726048799369896829771167132); DK(KP549754652, +0.549754652192770074288023275540779861653779767); DK(KP871714437, +0.871714437527667770979999223229522602943903653); DK(KP634619297, +0.634619297544148100711287640319130485732531031); DK(KP939062505, +0.939062505817492352556001843133229685779824606); DK(KP256756360, +0.256756360367726783319498520922669048172391148); DK(KP951056516, +0.951056516295153572116439333379382143405698634); DK(KP559016994, +0.559016994374947424102293417182819058860154590); DK(KP250000000, +0.250000000000000000000000000000000000000000000); DK(KP618033988, +0.618033988749894848204586834365638117720309180); { INT m; for (m = mb, W = W + (mb * 48); m < me; m = m + 1, ri = ri + ms, ii = ii + ms, W = W + 48, MAKE_VOLATILE_STRIDE(50, rs)) { E T7I, T6Q, T6O, T7O, T7M, T7H, T6P, T6H, T7J, T7N; { E T78, T5G, T3Y, T3M, T7C, T7c, T77, T6Y, Tt, T3L, T5T, T4P, T5Q, T4W, T3G; E T2G, T5P, T4T, T5S, T4M, T65, T45, T68, T4c, T2Z, T11, T67, T49, T64, T42; E T5Y, T4r, T61, T4k, T3d, T1z, T60, T4h, T5X, T4o, T3g, T1G, T3q, T4z, T4G; E T26, T3i, T1M, T3k, T1S; { E T3u, T2e, T3E, T4O, T4V, T2E, T3w, T2k, T3y, T2q; { E T1, T6X, T3P, T7, T3W, Tq, T9, Tc, Tb, T3U, Tk, T3Q, Ta; { E T3, T6, T2, T5; T1 = ri[0]; T6X = ii[0]; T3 = ri[WS(rs, 5)]; T6 = ii[WS(rs, 5)]; T2 = W[8]; T5 = W[9]; { E Tm, Tp, To, T3V, Tn, T3O, T4, Tl; Tm = ri[WS(rs, 15)]; Tp = ii[WS(rs, 15)]; T3O = T2 * T6; T4 = T2 * T3; Tl = W[28]; To = W[29]; T3P = FNMS(T5, T3, T3O); T7 = FMA(T5, T6, T4); T3V = Tl * Tp; Tn = Tl * Tm; { E Tg, Tj, Tf, Ti, T3T, Th, T8; Tg = ri[WS(rs, 10)]; Tj = ii[WS(rs, 10)]; T3W = FNMS(To, Tm, T3V); Tq = FMA(To, Tp, Tn); Tf = W[18]; Ti = W[19]; T9 = ri[WS(rs, 20)]; Tc = ii[WS(rs, 20)]; T3T = Tf * Tj; Th = Tf * Tg; T8 = W[38]; Tb = W[39]; T3U = FNMS(Ti, Tg, T3T); Tk = FMA(Ti, Tj, Th); T3Q = T8 * Tc; Ta = T8 * T9; } } } { E T6V, T3X, T7b, Tr, T3R, Td; T6V = T3U + T3W; T3X = T3U - T3W; T7b = Tk - Tq; Tr = Tk + Tq; T3R = FNMS(Tb, T9, T3Q); Td = FMA(Tb, Tc, Ta); { E T3S, T7a, Te, T6W, T6U, Ts; T3S = T3P - T3R; T6U = T3P + T3R; T7a = T7 - Td; Te = T7 + Td; T78 = T6U - T6V; T6W = T6U + T6V; T5G = FNMS(KP618033988, T3S, T3X); T3Y = FMA(KP618033988, T3X, T3S); T3M = Te - Tr; Ts = Te + Tr; T7C = FNMS(KP618033988, T7a, T7b); T7c = FMA(KP618033988, T7b, T7a); T77 = FNMS(KP250000000, T6W, T6X); T6Y = T6W + T6X; Tt = T1 + Ts; T3L = FNMS(KP250000000, Ts, T1); } } } { E T2g, T2j, T2m, T3v, T2h, T2p, T2l, T2i, T2o, T3x, T2n; { E T2a, T2d, T29, T2c; T2a = ri[WS(rs, 3)]; T2d = ii[WS(rs, 3)]; T29 = W[4]; T2c = W[5]; { E T2t, T2w, T2z, T3A, T2u, T2C, T2y, T2v, T2B, T3t, T2b, T2s, T2f; T2t = ri[WS(rs, 13)]; T2w = ii[WS(rs, 13)]; T3t = T29 * T2d; T2b = T29 * T2a; T2s = W[24]; T2z = ri[WS(rs, 18)]; T3u = FNMS(T2c, T2a, T3t); T2e = FMA(T2c, T2d, T2b); T3A = T2s * T2w; T2u = T2s * T2t; T2C = ii[WS(rs, 18)]; T2y = W[34]; T2v = W[25]; T2B = W[35]; { E T3B, T2x, T3D, T2D, T3C, T2A; T2g = ri[WS(rs, 8)]; T3C = T2y * T2C; T2A = T2y * T2z; T3B = FNMS(T2v, T2t, T3A); T2x = FMA(T2v, T2w, T2u); T3D = FNMS(T2B, T2z, T3C); T2D = FMA(T2B, T2C, T2A); T2j = ii[WS(rs, 8)]; T2f = W[14]; T3E = T3B + T3D; T4O = T3D - T3B; T4V = T2x - T2D; T2E = T2x + T2D; } T2m = ri[WS(rs, 23)]; T3v = T2f * T2j; T2h = T2f * T2g; T2p = ii[WS(rs, 23)]; T2l = W[44]; T2i = W[15]; T2o = W[45]; } } T3x = T2l * T2p; T2n = T2l * T2m; T3w = FNMS(T2i, T2g, T3v); T2k = FMA(T2i, T2j, T2h); T3y = FNMS(T2o, T2m, T3x); T2q = FMA(T2o, T2p, T2n); } { E T2N, Tz, T2X, T44, T4b, TZ, T2P, TF, T2R, TL; { E TB, TE, TH, T2O, TC, TK, TG, TD, TJ, T2Q, TI; { E Tv, Ty, Tu, Tx; { E T4S, T4L, T4R, T4K, T4N, T3z; Tv = ri[WS(rs, 1)]; T4N = T3y - T3w; T3z = T3w + T3y; { E T4U, T2r, T3F, T2F; T4U = T2k - T2q; T2r = T2k + T2q; T5T = FNMS(KP618033988, T4N, T4O); T4P = FMA(KP618033988, T4O, T4N); T3F = T3z + T3E; T4S = T3E - T3z; T5Q = FNMS(KP618033988, T4U, T4V); T4W = FMA(KP618033988, T4V, T4U); T2F = T2r + T2E; T4L = T2E - T2r; T3G = T3u + T3F; T4R = FNMS(KP250000000, T3F, T3u); T2G = T2e + T2F; T4K = FNMS(KP250000000, T2F, T2e); Ty = ii[WS(rs, 1)]; } T5P = FMA(KP559016994, T4S, T4R); T4T = FNMS(KP559016994, T4S, T4R); T5S = FMA(KP559016994, T4L, T4K); T4M = FNMS(KP559016994, T4L, T4K); Tu = W[0]; } Tx = W[1]; { E TO, TR, TU, T2T, TP, TX, TT, TQ, TW, T2M, Tw, TN, TA; TO = ri[WS(rs, 11)]; TR = ii[WS(rs, 11)]; T2M = Tu * Ty; Tw = Tu * Tv; TN = W[20]; TU = ri[WS(rs, 16)]; T2N = FNMS(Tx, Tv, T2M); Tz = FMA(Tx, Ty, Tw); T2T = TN * TR; TP = TN * TO; TX = ii[WS(rs, 16)]; TT = W[30]; TQ = W[21]; TW = W[31]; { E T2U, TS, T2W, TY, T2V, TV; TB = ri[WS(rs, 6)]; T2V = TT * TX; TV = TT * TU; T2U = FNMS(TQ, TO, T2T); TS = FMA(TQ, TR, TP); T2W = FNMS(TW, TU, T2V); TY = FMA(TW, TX, TV); TE = ii[WS(rs, 6)]; TA = W[10]; T2X = T2U + T2W; T44 = T2W - T2U; T4b = TY - TS; TZ = TS + TY; } TH = ri[WS(rs, 21)]; T2O = TA * TE; TC = TA * TB; TK = ii[WS(rs, 21)]; TG = W[40]; TD = W[11]; TJ = W[41]; } } T2Q = TG * TK; TI = TG * TH; T2P = FNMS(TD, TB, T2O); TF = FMA(TD, TE, TC); T2R = FNMS(TJ, TH, T2Q); TL = FMA(TJ, TK, TI); } { E T31, T17, T3b, T4q, T4j, T1x, T33, T1d, T35, T1j; { E T19, T1c, T1f, T32, T1a, T1i, T1e, T1b, T1h, T34, T1g; { E T13, T16, T12, T15; { E T48, T41, T47, T40, T43, T2S; T13 = ri[WS(rs, 4)]; T43 = T2P - T2R; T2S = T2P + T2R; { E T4a, TM, T2Y, T10; T4a = TL - TF; TM = TF + TL; T65 = FMA(KP618033988, T43, T44); T45 = FNMS(KP618033988, T44, T43); T2Y = T2S + T2X; T48 = T2S - T2X; T68 = FNMS(KP618033988, T4a, T4b); T4c = FMA(KP618033988, T4b, T4a); T10 = TM + TZ; T41 = TM - TZ; T2Z = T2N + T2Y; T47 = FNMS(KP250000000, T2Y, T2N); T11 = Tz + T10; T40 = FNMS(KP250000000, T10, Tz); T16 = ii[WS(rs, 4)]; } T67 = FNMS(KP559016994, T48, T47); T49 = FMA(KP559016994, T48, T47); T64 = FNMS(KP559016994, T41, T40); T42 = FMA(KP559016994, T41, T40); T12 = W[6]; } T15 = W[7]; { E T1m, T1p, T1s, T37, T1n, T1v, T1r, T1o, T1u, T30, T14, T1l, T18; T1m = ri[WS(rs, 14)]; T1p = ii[WS(rs, 14)]; T30 = T12 * T16; T14 = T12 * T13; T1l = W[26]; T1s = ri[WS(rs, 19)]; T31 = FNMS(T15, T13, T30); T17 = FMA(T15, T16, T14); T37 = T1l * T1p; T1n = T1l * T1m; T1v = ii[WS(rs, 19)]; T1r = W[36]; T1o = W[27]; T1u = W[37]; { E T38, T1q, T3a, T1w, T39, T1t; T19 = ri[WS(rs, 9)]; T39 = T1r * T1v; T1t = T1r * T1s; T38 = FNMS(T1o, T1m, T37); T1q = FMA(T1o, T1p, T1n); T3a = FNMS(T1u, T1s, T39); T1w = FMA(T1u, T1v, T1t); T1c = ii[WS(rs, 9)]; T18 = W[16]; T3b = T38 + T3a; T4q = T3a - T38; T4j = T1w - T1q; T1x = T1q + T1w; } T1f = ri[WS(rs, 24)]; T32 = T18 * T1c; T1a = T18 * T19; T1i = ii[WS(rs, 24)]; T1e = W[46]; T1b = W[17]; T1h = W[47]; } } T34 = T1e * T1i; T1g = T1e * T1f; T33 = FNMS(T1b, T19, T32); T1d = FMA(T1b, T1c, T1a); T35 = FNMS(T1h, T1f, T34); T1j = FMA(T1h, T1i, T1g); } { E T1I, T1L, T1O, T3h, T1J, T1R, T1N, T1K, T1Q, T3j, T1P; { E T1C, T1F, T1B, T1E; { E T4g, T4n, T4f, T4m, T4p, T36; T1C = ri[WS(rs, 2)]; T4p = T35 - T33; T36 = T33 + T35; { E T4i, T1k, T3c, T1y; T4i = T1j - T1d; T1k = T1d + T1j; T5Y = FNMS(KP618033988, T4p, T4q); T4r = FMA(KP618033988, T4q, T4p); T3c = T36 + T3b; T4g = T3b - T36; T61 = FNMS(KP618033988, T4i, T4j); T4k = FMA(KP618033988, T4j, T4i); T1y = T1k + T1x; T4n = T1k - T1x; T3d = T31 + T3c; T4f = FNMS(KP250000000, T3c, T31); T1z = T17 + T1y; T4m = FNMS(KP250000000, T1y, T17); T1F = ii[WS(rs, 2)]; } T60 = FMA(KP559016994, T4g, T4f); T4h = FNMS(KP559016994, T4g, T4f); T5X = FNMS(KP559016994, T4n, T4m); T4o = FMA(KP559016994, T4n, T4m); T1B = W[2]; } T1E = W[3]; { E T1V, T1Y, T21, T3m, T1W, T24, T20, T1X, T23, T3f, T1D, T1U, T1H; T1V = ri[WS(rs, 12)]; T1Y = ii[WS(rs, 12)]; T3f = T1B * T1F; T1D = T1B * T1C; T1U = W[22]; T21 = ri[WS(rs, 17)]; T3g = FNMS(T1E, T1C, T3f); T1G = FMA(T1E, T1F, T1D); T3m = T1U * T1Y; T1W = T1U * T1V; T24 = ii[WS(rs, 17)]; T20 = W[32]; T1X = W[23]; T23 = W[33]; { E T3n, T1Z, T3p, T25, T3o, T22; T1I = ri[WS(rs, 7)]; T3o = T20 * T24; T22 = T20 * T21; T3n = FNMS(T1X, T1V, T3m); T1Z = FMA(T1X, T1Y, T1W); T3p = FNMS(T23, T21, T3o); T25 = FMA(T23, T24, T22); T1L = ii[WS(rs, 7)]; T1H = W[12]; T3q = T3n + T3p; T4z = T3n - T3p; T4G = T25 - T1Z; T26 = T1Z + T25; } T1O = ri[WS(rs, 22)]; T3h = T1H * T1L; T1J = T1H * T1I; T1R = ii[WS(rs, 22)]; T1N = W[42]; T1K = W[13]; T1Q = W[43]; } } T3j = T1N * T1R; T1P = T1N * T1O; T3i = FNMS(T1K, T1I, T3h); T1M = FMA(T1K, T1L, T1J); T3k = FNMS(T1Q, T1O, T3j); T1S = FMA(T1Q, T1R, T1P); } } } } { E T6R, T5M, T4A, T5J, T4H, T6S, T5I, T4E, T5L, T4x, T3K, T3I, T2K, T74, T76; E T2J; { E T1A, T72, T73, T2H, T28, T2I; { E T3e, T4D, T4w, T4C, T4v, T3H, T4y, T3l; T6R = T2Z + T3d; T3e = T2Z - T3d; T4y = T3k - T3i; T3l = T3i + T3k; { E T4F, T1T, T3r, T27, T3s; T4F = T1S - T1M; T1T = T1M + T1S; T5M = FMA(KP618033988, T4y, T4z); T4A = FNMS(KP618033988, T4z, T4y); T3r = T3l + T3q; T4D = T3q - T3l; T5J = FNMS(KP618033988, T4F, T4G); T4H = FMA(KP618033988, T4G, T4F); T27 = T1T + T26; T4w = T26 - T1T; T3s = T3g + T3r; T4C = FNMS(KP250000000, T3r, T3g); T28 = T1G + T27; T4v = FNMS(KP250000000, T27, T1G); T3H = T3s - T3G; T6S = T3s + T3G; } T5I = FMA(KP559016994, T4D, T4C); T4E = FNMS(KP559016994, T4D, T4C); T5L = FMA(KP559016994, T4w, T4v); T4x = FNMS(KP559016994, T4w, T4v); T3K = FNMS(KP618033988, T3e, T3H); T3I = FMA(KP618033988, T3H, T3e); } T1A = T11 + T1z; T72 = T11 - T1z; T73 = T28 - T2G; T2H = T28 + T2G; T2I = T1A + T2H; T2K = T1A - T2H; T74 = FMA(KP618033988, T73, T72); T76 = FNMS(KP618033988, T72, T73); ri[0] = Tt + T2I; T2J = FNMS(KP250000000, T2I, Tt); } { E T5F, T7B, T7u, T5E, T5C, T7A, T7y, T7t, T5D, T5v; { E T3Z, T5d, T7p, T7d, T5m, T5l, T56, T7k, T59, T7l, T5z, T5g, T7g, T7i, T52; E T50, T5x, T5q, T5A, T5j, T70, T6Z, T3N; T5F = FNMS(KP559016994, T3M, T3L); T3N = FMA(KP559016994, T3M, T3L); { E T79, T3J, T2L, T6T; T79 = FMA(KP559016994, T78, T77); T7B = FNMS(KP559016994, T78, T77); T3J = FNMS(KP559016994, T2K, T2J); T2L = FMA(KP559016994, T2K, T2J); T6T = T6R + T6S; T70 = T6R - T6S; T3Z = FMA(KP951056516, T3Y, T3N); T5d = FNMS(KP951056516, T3Y, T3N); ri[WS(rs, 5)] = FMA(KP951056516, T3I, T2L); ri[WS(rs, 20)] = FNMS(KP951056516, T3I, T2L); ri[WS(rs, 15)] = FMA(KP951056516, T3K, T3J); ri[WS(rs, 10)] = FNMS(KP951056516, T3K, T3J); ii[0] = T6T + T6Y; T6Z = FNMS(KP250000000, T6T, T6Y); T7p = FMA(KP951056516, T7c, T79); T7d = FNMS(KP951056516, T7c, T79); } { E T5e, T54, T4e, T5f, T5o, T5p, T5i, T4B, T58, T4Y, T55, T4t, T4I, T5h; { E T4Q, T4X, T4l, T4s; { E T46, T71, T75, T4d; T5m = FNMS(KP951056516, T45, T42); T46 = FMA(KP951056516, T45, T42); T71 = FMA(KP559016994, T70, T6Z); T75 = FNMS(KP559016994, T70, T6Z); T4d = FMA(KP951056516, T4c, T49); T5l = FNMS(KP951056516, T4c, T49); T5e = FMA(KP951056516, T4P, T4M); T4Q = FNMS(KP951056516, T4P, T4M); ii[WS(rs, 20)] = FMA(KP951056516, T74, T71); ii[WS(rs, 5)] = FNMS(KP951056516, T74, T71); ii[WS(rs, 15)] = FNMS(KP951056516, T76, T75); ii[WS(rs, 10)] = FMA(KP951056516, T76, T75); T54 = FNMS(KP256756360, T46, T4d); T4e = FMA(KP256756360, T4d, T46); T4X = FNMS(KP951056516, T4W, T4T); T5f = FMA(KP951056516, T4W, T4T); } T5o = FNMS(KP951056516, T4k, T4h); T4l = FMA(KP951056516, T4k, T4h); T4s = FNMS(KP951056516, T4r, T4o); T5p = FMA(KP951056516, T4r, T4o); T5i = FMA(KP951056516, T4A, T4x); T4B = FNMS(KP951056516, T4A, T4x); T58 = FNMS(KP939062505, T4Q, T4X); T4Y = FMA(KP939062505, T4X, T4Q); T55 = FNMS(KP634619297, T4l, T4s); T4t = FMA(KP634619297, T4s, T4l); T4I = FMA(KP951056516, T4H, T4E); T5h = FNMS(KP951056516, T4H, T4E); } { E T7e, T4u, T57, T4J, T7f, T4Z; T7e = FNMS(KP871714437, T55, T54); T56 = FMA(KP871714437, T55, T54); T4u = FMA(KP871714437, T4t, T4e); T7k = FNMS(KP871714437, T4t, T4e); T57 = FNMS(KP549754652, T4B, T4I); T4J = FMA(KP549754652, T4I, T4B); T7f = FMA(KP831864738, T58, T57); T59 = FNMS(KP831864738, T58, T57); T4Z = FMA(KP831864738, T4Y, T4J); T7l = FNMS(KP831864738, T4Y, T4J); T5z = FMA(KP126329378, T5e, T5f); T5g = FNMS(KP126329378, T5f, T5e); T7g = FMA(KP904730450, T7f, T7e); T7i = FNMS(KP904730450, T7f, T7e); T52 = FNMS(KP904730450, T4Z, T4u); T50 = FMA(KP904730450, T4Z, T4u); } T5x = FNMS(KP827271945, T5o, T5p); T5q = FMA(KP827271945, T5p, T5o); T5A = FMA(KP470564281, T5h, T5i); T5j = FNMS(KP470564281, T5i, T5h); } { E T7q, T5B, T5k, T7x, T5w, T5n; ri[WS(rs, 1)] = FMA(KP968583161, T50, T3Z); T7q = FMA(KP912018591, T5A, T5z); T5B = FNMS(KP912018591, T5A, T5z); T5k = FNMS(KP912018591, T5j, T5g); T7x = FMA(KP912018591, T5j, T5g); T5w = FNMS(KP634619297, T5l, T5m); T5n = FMA(KP634619297, T5m, T5l); ii[WS(rs, 1)] = FMA(KP968583161, T7g, T7d); { E T5y, T7w, T7s, T5s, T5u, T7o, T7m, T7n, T7j, T5t; { E T5c, T5a, T51, T7r, T5r, T53, T5b, T7h; T5c = FNMS(KP683113946, T56, T59); T5a = FMA(KP559154169, T59, T56); T7r = FNMS(KP912575812, T5x, T5w); T5y = FMA(KP912575812, T5x, T5w); T5r = FNMS(KP912575812, T5q, T5n); T7w = FMA(KP912575812, T5q, T5n); T7s = FMA(KP851038619, T7r, T7q); T7u = FNMS(KP851038619, T7r, T7q); T5s = FNMS(KP851038619, T5r, T5k); T5u = FMA(KP851038619, T5r, T5k); T51 = FNMS(KP242145790, T50, T3Z); ii[WS(rs, 4)] = FNMS(KP992114701, T7s, T7p); ri[WS(rs, 4)] = FNMS(KP992114701, T5s, T5d); T7o = FNMS(KP683113946, T7k, T7l); T7m = FMA(KP559154169, T7l, T7k); T53 = FMA(KP541454447, T52, T51); T5b = FNMS(KP541454447, T52, T51); T7h = FNMS(KP242145790, T7g, T7d); ri[WS(rs, 11)] = FNMS(KP833417178, T5c, T5b); ri[WS(rs, 16)] = FMA(KP833417178, T5c, T5b); ri[WS(rs, 21)] = FNMS(KP921177326, T5a, T53); ri[WS(rs, 6)] = FMA(KP921177326, T5a, T53); T7n = FNMS(KP541454447, T7i, T7h); T7j = FMA(KP541454447, T7i, T7h); } T5E = FMA(KP525970792, T5y, T5B); T5C = FNMS(KP726211448, T5B, T5y); ii[WS(rs, 21)] = FMA(KP921177326, T7m, T7j); ii[WS(rs, 6)] = FNMS(KP921177326, T7m, T7j); ii[WS(rs, 11)] = FMA(KP833417178, T7o, T7n); ii[WS(rs, 16)] = FNMS(KP833417178, T7o, T7n); T5t = FMA(KP248028675, T5s, T5d); T7A = FNMS(KP525970792, T7w, T7x); T7y = FMA(KP726211448, T7x, T7w); T7t = FMA(KP248028675, T7s, T7p); T5D = FNMS(KP554608978, T5u, T5t); T5v = FMA(KP554608978, T5u, T5t); } } } { E T5H, T6p, T7P, T7D, T6y, T6x, T6l, T7X, T6i, T7W, T6L, T6s, T7S, T7U, T6e; E T6c, T6J, T6C, T6M, T6v, T7z, T7v; ri[WS(rs, 14)] = FNMS(KP943557151, T5E, T5D); ri[WS(rs, 19)] = FMA(KP943557151, T5E, T5D); ri[WS(rs, 24)] = FMA(KP803003575, T5C, T5v); ri[WS(rs, 9)] = FNMS(KP803003575, T5C, T5v); T7z = FNMS(KP554608978, T7u, T7t); T7v = FMA(KP554608978, T7u, T7t); T5H = FMA(KP951056516, T5G, T5F); T6p = FNMS(KP951056516, T5G, T5F); ii[WS(rs, 14)] = FMA(KP943557151, T7A, T7z); ii[WS(rs, 19)] = FNMS(KP943557151, T7A, T7z); ii[WS(rs, 24)] = FMA(KP803003575, T7y, T7v); ii[WS(rs, 9)] = FNMS(KP803003575, T7y, T7v); { E T6t, T6u, T6A, T6j, T5O, T6B, T6q, T6r, T5Z, T6h, T6a, T6k, T5V, T62; { E T66, T69, T5K, T5N, T5R, T5U; T6t = FNMS(KP951056516, T5J, T5I); T5K = FMA(KP951056516, T5J, T5I); T5N = FMA(KP951056516, T5M, T5L); T6u = FNMS(KP951056516, T5M, T5L); T6A = FMA(KP951056516, T65, T64); T66 = FNMS(KP951056516, T65, T64); T7P = FNMS(KP951056516, T7C, T7B); T7D = FMA(KP951056516, T7C, T7B); T6j = FNMS(KP062914667, T5K, T5N); T5O = FMA(KP062914667, T5N, T5K); T69 = FMA(KP951056516, T68, T67); T6B = FNMS(KP951056516, T68, T67); T6q = FMA(KP951056516, T5Q, T5P); T5R = FNMS(KP951056516, T5Q, T5P); T5U = FNMS(KP951056516, T5T, T5S); T6r = FMA(KP951056516, T5T, T5S); T6y = FMA(KP951056516, T5Y, T5X); T5Z = FNMS(KP951056516, T5Y, T5X); T6h = FNMS(KP939062505, T66, T69); T6a = FMA(KP939062505, T69, T66); T6k = FMA(KP827271945, T5R, T5U); T5V = FNMS(KP827271945, T5U, T5R); T62 = FMA(KP951056516, T61, T60); T6x = FNMS(KP951056516, T61, T60); } { E T7Q, T5W, T6g, T63, T7R, T6b; T7Q = FMA(KP772036680, T6k, T6j); T6l = FNMS(KP772036680, T6k, T6j); T5W = FMA(KP772036680, T5V, T5O); T7X = FNMS(KP772036680, T5V, T5O); T6g = FMA(KP126329378, T5Z, T62); T63 = FNMS(KP126329378, T62, T5Z); T7R = FNMS(KP734762448, T6h, T6g); T6i = FMA(KP734762448, T6h, T6g); T6b = FNMS(KP734762448, T6a, T63); T7W = FMA(KP734762448, T6a, T63); T6L = FNMS(KP062914667, T6q, T6r); T6s = FMA(KP062914667, T6r, T6q); T7S = FMA(KP994076283, T7R, T7Q); T7U = FNMS(KP994076283, T7R, T7Q); T6e = FMA(KP994076283, T6b, T5W); T6c = FNMS(KP994076283, T6b, T5W); } T6J = FNMS(KP549754652, T6A, T6B); T6C = FMA(KP549754652, T6B, T6A); T6M = FNMS(KP634619297, T6t, T6u); T6v = FMA(KP634619297, T6u, T6t); } { E T7E, T6N, T6w, T7L, T6I, T6z; ri[WS(rs, 3)] = FMA(KP998026728, T6c, T5H); T7E = FMA(KP845997307, T6M, T6L); T6N = FNMS(KP845997307, T6M, T6L); T6w = FMA(KP845997307, T6v, T6s); T7L = FNMS(KP845997307, T6v, T6s); T6I = FMA(KP470564281, T6x, T6y); T6z = FNMS(KP470564281, T6y, T6x); ii[WS(rs, 3)] = FNMS(KP998026728, T7S, T7P); { E T6K, T7K, T7G, T6E, T6G, T80, T7Y, T7Z, T7V, T6F; { E T6o, T6m, T6d, T7F, T6D, T6f, T6n, T7T; T6o = FMA(KP614372930, T6i, T6l); T6m = FNMS(KP621716863, T6l, T6i); T7F = FNMS(KP968479752, T6J, T6I); T6K = FMA(KP968479752, T6J, T6I); T6D = FMA(KP968479752, T6C, T6z); T7K = FNMS(KP968479752, T6C, T6z); T7G = FMA(KP906616052, T7F, T7E); T7I = FNMS(KP906616052, T7F, T7E); T6E = FMA(KP906616052, T6D, T6w); T6G = FNMS(KP906616052, T6D, T6w); T6d = FNMS(KP249506682, T6c, T5H); ii[WS(rs, 2)] = FNMS(KP998026728, T7G, T7D); ri[WS(rs, 2)] = FMA(KP998026728, T6E, T6p); T80 = FNMS(KP614372930, T7W, T7X); T7Y = FMA(KP621716863, T7X, T7W); T6f = FNMS(KP557913902, T6e, T6d); T6n = FMA(KP557913902, T6e, T6d); T7T = FMA(KP249506682, T7S, T7P); ri[WS(rs, 18)] = FNMS(KP949179823, T6o, T6n); ri[WS(rs, 13)] = FMA(KP949179823, T6o, T6n); ri[WS(rs, 8)] = FMA(KP943557151, T6m, T6f); ri[WS(rs, 23)] = FNMS(KP943557151, T6m, T6f); T7Z = FNMS(KP557913902, T7U, T7T); T7V = FMA(KP557913902, T7U, T7T); } T6Q = FNMS(KP560319534, T6K, T6N); T6O = FMA(KP681693190, T6N, T6K); ii[WS(rs, 23)] = FMA(KP943557151, T7Y, T7V); ii[WS(rs, 8)] = FNMS(KP943557151, T7Y, T7V); ii[WS(rs, 13)] = FMA(KP949179823, T80, T7Z); ii[WS(rs, 18)] = FNMS(KP949179823, T80, T7Z); T6F = FNMS(KP249506682, T6E, T6p); T7O = FNMS(KP560319534, T7K, T7L); T7M = FMA(KP681693190, T7L, T7K); T7H = FMA(KP249506682, T7G, T7D); T6P = FMA(KP557913902, T6G, T6F); T6H = FNMS(KP557913902, T6G, T6F); } } } } } } ri[WS(rs, 12)] = FNMS(KP949179823, T6Q, T6P); ri[WS(rs, 17)] = FMA(KP949179823, T6Q, T6P); ri[WS(rs, 7)] = FMA(KP860541664, T6O, T6H); ri[WS(rs, 22)] = FNMS(KP860541664, T6O, T6H); T7J = FMA(KP557913902, T7I, T7H); T7N = FNMS(KP557913902, T7I, T7H); ii[WS(rs, 12)] = FNMS(KP949179823, T7O, T7N); ii[WS(rs, 17)] = FMA(KP949179823, T7O, T7N); ii[WS(rs, 22)] = FNMS(KP860541664, T7M, T7J); ii[WS(rs, 7)] = FMA(KP860541664, T7M, T7J); } } } static const tw_instr twinstr[] = { {TW_FULL, 0, 25}, {TW_NEXT, 1, 0} }; static const ct_desc desc = { 25, "t1_25", twinstr, &GENUS, {84, 48, 316, 0}, 0, 0, 0 }; void X(codelet_t1_25) (planner *p) { X(kdft_dit_register) (p, t1_25, &desc); } #else /* HAVE_FMA */ /* Generated by: ../../../genfft/gen_twiddle.native -compact -variables 4 -pipeline-latency 4 -n 25 -name t1_25 -include t.h */ /* * This function contains 400 FP additions, 280 FP multiplications, * (or, 260 additions, 140 multiplications, 140 fused multiply/add), * 101 stack variables, 20 constants, and 100 memory accesses */ #include "t.h" static void t1_25(R *ri, R *ii, const R *W, stride rs, INT mb, INT me, INT ms) { DK(KP998026728, +0.998026728428271561952336806863450553336905220); DK(KP062790519, +0.062790519529313376076178224565631133122484832); DK(KP425779291, +0.425779291565072648862502445744251703979973042); DK(KP904827052, +0.904827052466019527713668647932697593970413911); DK(KP992114701, +0.992114701314477831049793042785778521453036709); DK(KP125333233, +0.125333233564304245373118759816508793942918247); DK(KP637423989, +0.637423989748689710176712811676016195434917298); DK(KP770513242, +0.770513242775789230803009636396177847271667672); DK(KP684547105, +0.684547105928688673732283357621209269889519233); DK(KP728968627, +0.728968627421411523146730319055259111372571664); DK(KP481753674, +0.481753674101715274987191502872129653528542010); DK(KP876306680, +0.876306680043863587308115903922062583399064238); DK(KP844327925, +0.844327925502015078548558063966681505381659241); DK(KP535826794, +0.535826794978996618271308767867639978063575346); DK(KP248689887, +0.248689887164854788242283746006447968417567406); DK(KP968583161, +0.968583161128631119490168375464735813836012403); DK(KP587785252, +0.587785252292473129168705954639072768597652438); DK(KP951056516, +0.951056516295153572116439333379382143405698634); DK(KP250000000, +0.250000000000000000000000000000000000000000000); DK(KP559016994, +0.559016994374947424102293417182819058860154590); { INT m; for (m = mb, W = W + (mb * 48); m < me; m = m + 1, ri = ri + ms, ii = ii + ms, W = W + 48, MAKE_VOLATILE_STRIDE(50, rs)) { E T1, T6b, T2l, T6o, To, T2m, T6a, T6p, T6t, T6S, T2u, T4I, T2i, T60, T3O; E T5D, T4r, T58, T3Z, T5C, T4q, T5b, TS, T5W, T2G, T5s, T4g, T4M, T2R, T5t; E T4h, T4P, T1l, T5X, T33, T5w, T4j, T4W, T3e, T5v, T4k, T4T, T1P, T5Z, T3r; E T5z, T4o, T51, T3C, T5A, T4n, T54; { E T6, T2o, Tb, T2p, Tc, T68, Th, T2r, Tm, T2s, Tn, T69; T1 = ri[0]; T6b = ii[0]; { E T3, T5, T2, T4; T3 = ri[WS(rs, 5)]; T5 = ii[WS(rs, 5)]; T2 = W[8]; T4 = W[9]; T6 = FMA(T2, T3, T4 * T5); T2o = FNMS(T4, T3, T2 * T5); } { E T8, Ta, T7, T9; T8 = ri[WS(rs, 20)]; Ta = ii[WS(rs, 20)]; T7 = W[38]; T9 = W[39]; Tb = FMA(T7, T8, T9 * Ta); T2p = FNMS(T9, T8, T7 * Ta); } Tc = T6 + Tb; T68 = T2o + T2p; { E Te, Tg, Td, Tf; Te = ri[WS(rs, 10)]; Tg = ii[WS(rs, 10)]; Td = W[18]; Tf = W[19]; Th = FMA(Td, Te, Tf * Tg); T2r = FNMS(Tf, Te, Td * Tg); } { E Tj, Tl, Ti, Tk; Tj = ri[WS(rs, 15)]; Tl = ii[WS(rs, 15)]; Ti = W[28]; Tk = W[29]; Tm = FMA(Ti, Tj, Tk * Tl); T2s = FNMS(Tk, Tj, Ti * Tl); } Tn = Th + Tm; T69 = T2r + T2s; T2l = KP559016994 * (Tc - Tn); T6o = KP559016994 * (T68 - T69); To = Tc + Tn; T2m = FNMS(KP250000000, To, T1); T6a = T68 + T69; T6p = FNMS(KP250000000, T6a, T6b); { E T6r, T6s, T2q, T2t; T6r = T6 - Tb; T6s = Th - Tm; T6t = FMA(KP951056516, T6r, KP587785252 * T6s); T6S = FNMS(KP587785252, T6r, KP951056516 * T6s); T2q = T2o - T2p; T2t = T2r - T2s; T2u = FMA(KP951056516, T2q, KP587785252 * T2t); T4I = FNMS(KP587785252, T2q, KP951056516 * T2t); } } { E T1U, T3S, T3J, T3M, T3X, T3W, T3P, T3Q, T3T, T25, T2g, T2h; { E T1R, T1T, T1Q, T1S; T1R = ri[WS(rs, 3)]; T1T = ii[WS(rs, 3)]; T1Q = W[4]; T1S = W[5]; T1U = FMA(T1Q, T1R, T1S * T1T); T3S = FNMS(T1S, T1R, T1Q * T1T); } { E T1Z, T3H, T2f, T3L, T24, T3I, T2a, T3K; { E T1W, T1Y, T1V, T1X; T1W = ri[WS(rs, 8)]; T1Y = ii[WS(rs, 8)]; T1V = W[14]; T1X = W[15]; T1Z = FMA(T1V, T1W, T1X * T1Y); T3H = FNMS(T1X, T1W, T1V * T1Y); } { E T2c, T2e, T2b, T2d; T2c = ri[WS(rs, 18)]; T2e = ii[WS(rs, 18)]; T2b = W[34]; T2d = W[35]; T2f = FMA(T2b, T2c, T2d * T2e); T3L = FNMS(T2d, T2c, T2b * T2e); } { E T21, T23, T20, T22; T21 = ri[WS(rs, 23)]; T23 = ii[WS(rs, 23)]; T20 = W[44]; T22 = W[45]; T24 = FMA(T20, T21, T22 * T23); T3I = FNMS(T22, T21, T20 * T23); } { E T27, T29, T26, T28; T27 = ri[WS(rs, 13)]; T29 = ii[WS(rs, 13)]; T26 = W[24]; T28 = W[25]; T2a = FMA(T26, T27, T28 * T29); T3K = FNMS(T28, T27, T26 * T29); } T3J = T3H - T3I; T3M = T3K - T3L; T3X = T2a - T2f; T3W = T1Z - T24; T3P = T3H + T3I; T3Q = T3K + T3L; T3T = T3P + T3Q; T25 = T1Z + T24; T2g = T2a + T2f; T2h = T25 + T2g; } T2i = T1U + T2h; T60 = T3S + T3T; { E T3N, T57, T3G, T56, T3E, T3F; T3N = FMA(KP951056516, T3J, KP587785252 * T3M); T57 = FNMS(KP587785252, T3J, KP951056516 * T3M); T3E = KP559016994 * (T25 - T2g); T3F = FNMS(KP250000000, T2h, T1U); T3G = T3E + T3F; T56 = T3F - T3E; T3O = T3G + T3N; T5D = T56 + T57; T4r = T3G - T3N; T58 = T56 - T57; } { E T3Y, T59, T3V, T5a, T3R, T3U; T3Y = FMA(KP951056516, T3W, KP587785252 * T3X); T59 = FNMS(KP587785252, T3W, KP951056516 * T3X); T3R = KP559016994 * (T3P - T3Q); T3U = FNMS(KP250000000, T3T, T3S); T3V = T3R + T3U; T5a = T3U - T3R; T3Z = T3V - T3Y; T5C = T5a - T59; T4q = T3Y + T3V; T5b = T59 + T5a; } } { E Tu, T2K, T2B, T2E, T2P, T2O, T2H, T2I, T2L, TF, TQ, TR; { E Tr, Tt, Tq, Ts; Tr = ri[WS(rs, 1)]; Tt = ii[WS(rs, 1)]; Tq = W[0]; Ts = W[1]; Tu = FMA(Tq, Tr, Ts * Tt); T2K = FNMS(Ts, Tr, Tq * Tt); } { E Tz, T2z, TP, T2D, TE, T2A, TK, T2C; { E Tw, Ty, Tv, Tx; Tw = ri[WS(rs, 6)]; Ty = ii[WS(rs, 6)]; Tv = W[10]; Tx = W[11]; Tz = FMA(Tv, Tw, Tx * Ty); T2z = FNMS(Tx, Tw, Tv * Ty); } { E TM, TO, TL, TN; TM = ri[WS(rs, 16)]; TO = ii[WS(rs, 16)]; TL = W[30]; TN = W[31]; TP = FMA(TL, TM, TN * TO); T2D = FNMS(TN, TM, TL * TO); } { E TB, TD, TA, TC; TB = ri[WS(rs, 21)]; TD = ii[WS(rs, 21)]; TA = W[40]; TC = W[41]; TE = FMA(TA, TB, TC * TD); T2A = FNMS(TC, TB, TA * TD); } { E TH, TJ, TG, TI; TH = ri[WS(rs, 11)]; TJ = ii[WS(rs, 11)]; TG = W[20]; TI = W[21]; TK = FMA(TG, TH, TI * TJ); T2C = FNMS(TI, TH, TG * TJ); } T2B = T2z - T2A; T2E = T2C - T2D; T2P = TK - TP; T2O = Tz - TE; T2H = T2z + T2A; T2I = T2C + T2D; T2L = T2H + T2I; TF = Tz + TE; TQ = TK + TP; TR = TF + TQ; } TS = Tu + TR; T5W = T2K + T2L; { E T2F, T4L, T2y, T4K, T2w, T2x; T2F = FMA(KP951056516, T2B, KP587785252 * T2E); T4L = FNMS(KP587785252, T2B, KP951056516 * T2E); T2w = KP559016994 * (TF - TQ); T2x = FNMS(KP250000000, TR, Tu); T2y = T2w + T2x; T4K = T2x - T2w; T2G = T2y + T2F; T5s = T4K + T4L; T4g = T2y - T2F; T4M = T4K - T4L; } { E T2Q, T4N, T2N, T4O, T2J, T2M; T2Q = FMA(KP951056516, T2O, KP587785252 * T2P); T4N = FNMS(KP587785252, T2O, KP951056516 * T2P); T2J = KP559016994 * (T2H - T2I); T2M = FNMS(KP250000000, T2L, T2K); T2N = T2J + T2M; T4O = T2M - T2J; T2R = T2N - T2Q; T5t = T4O - T4N; T4h = T2Q + T2N; T4P = T4N + T4O; } } { E TX, T37, T2Y, T31, T3c, T3b, T34, T35, T38, T18, T1j, T1k; { E TU, TW, TT, TV; TU = ri[WS(rs, 4)]; TW = ii[WS(rs, 4)]; TT = W[6]; TV = W[7]; TX = FMA(TT, TU, TV * TW); T37 = FNMS(TV, TU, TT * TW); } { E T12, T2W, T1i, T30, T17, T2X, T1d, T2Z; { E TZ, T11, TY, T10; TZ = ri[WS(rs, 9)]; T11 = ii[WS(rs, 9)]; TY = W[16]; T10 = W[17]; T12 = FMA(TY, TZ, T10 * T11); T2W = FNMS(T10, TZ, TY * T11); } { E T1f, T1h, T1e, T1g; T1f = ri[WS(rs, 19)]; T1h = ii[WS(rs, 19)]; T1e = W[36]; T1g = W[37]; T1i = FMA(T1e, T1f, T1g * T1h); T30 = FNMS(T1g, T1f, T1e * T1h); } { E T14, T16, T13, T15; T14 = ri[WS(rs, 24)]; T16 = ii[WS(rs, 24)]; T13 = W[46]; T15 = W[47]; T17 = FMA(T13, T14, T15 * T16); T2X = FNMS(T15, T14, T13 * T16); } { E T1a, T1c, T19, T1b; T1a = ri[WS(rs, 14)]; T1c = ii[WS(rs, 14)]; T19 = W[26]; T1b = W[27]; T1d = FMA(T19, T1a, T1b * T1c); T2Z = FNMS(T1b, T1a, T19 * T1c); } T2Y = T2W - T2X; T31 = T2Z - T30; T3c = T1d - T1i; T3b = T12 - T17; T34 = T2W + T2X; T35 = T2Z + T30; T38 = T34 + T35; T18 = T12 + T17; T1j = T1d + T1i; T1k = T18 + T1j; } T1l = TX + T1k; T5X = T37 + T38; { E T32, T4V, T2V, T4U, T2T, T2U; T32 = FMA(KP951056516, T2Y, KP587785252 * T31); T4V = FNMS(KP587785252, T2Y, KP951056516 * T31); T2T = KP559016994 * (T18 - T1j); T2U = FNMS(KP250000000, T1k, TX); T2V = T2T + T2U; T4U = T2U - T2T; T33 = T2V + T32; T5w = T4U + T4V; T4j = T2V - T32; T4W = T4U - T4V; } { E T3d, T4R, T3a, T4S, T36, T39; T3d = FMA(KP951056516, T3b, KP587785252 * T3c); T4R = FNMS(KP587785252, T3b, KP951056516 * T3c); T36 = KP559016994 * (T34 - T35); T39 = FNMS(KP250000000, T38, T37); T3a = T36 + T39; T4S = T39 - T36; T3e = T3a - T3d; T5v = T4S - T4R; T4k = T3d + T3a; T4T = T4R + T4S; } } { E T1r, T3v, T3m, T3p, T3A, T3z, T3s, T3t, T3w, T1C, T1N, T1O; { E T1o, T1q, T1n, T1p; T1o = ri[WS(rs, 2)]; T1q = ii[WS(rs, 2)]; T1n = W[2]; T1p = W[3]; T1r = FMA(T1n, T1o, T1p * T1q); T3v = FNMS(T1p, T1o, T1n * T1q); } { E T1w, T3k, T1M, T3o, T1B, T3l, T1H, T3n; { E T1t, T1v, T1s, T1u; T1t = ri[WS(rs, 7)]; T1v = ii[WS(rs, 7)]; T1s = W[12]; T1u = W[13]; T1w = FMA(T1s, T1t, T1u * T1v); T3k = FNMS(T1u, T1t, T1s * T1v); } { E T1J, T1L, T1I, T1K; T1J = ri[WS(rs, 17)]; T1L = ii[WS(rs, 17)]; T1I = W[32]; T1K = W[33]; T1M = FMA(T1I, T1J, T1K * T1L); T3o = FNMS(T1K, T1J, T1I * T1L); } { E T1y, T1A, T1x, T1z; T1y = ri[WS(rs, 22)]; T1A = ii[WS(rs, 22)]; T1x = W[42]; T1z = W[43]; T1B = FMA(T1x, T1y, T1z * T1A); T3l = FNMS(T1z, T1y, T1x * T1A); } { E T1E, T1G, T1D, T1F; T1E = ri[WS(rs, 12)]; T1G = ii[WS(rs, 12)]; T1D = W[22]; T1F = W[23]; T1H = FMA(T1D, T1E, T1F * T1G); T3n = FNMS(T1F, T1E, T1D * T1G); } T3m = T3k - T3l; T3p = T3n - T3o; T3A = T1H - T1M; T3z = T1w - T1B; T3s = T3k + T3l; T3t = T3n + T3o; T3w = T3s + T3t; T1C = T1w + T1B; T1N = T1H + T1M; T1O = T1C + T1N; } T1P = T1r + T1O; T5Z = T3v + T3w; { E T3q, T50, T3j, T4Z, T3h, T3i; T3q = FMA(KP951056516, T3m, KP587785252 * T3p); T50 = FNMS(KP587785252, T3m, KP951056516 * T3p); T3h = KP559016994 * (T1C - T1N); T3i = FNMS(KP250000000, T1O, T1r); T3j = T3h + T3i; T4Z = T3i - T3h; T3r = T3j + T3q; T5z = T4Z + T50; T4o = T3j - T3q; T51 = T4Z - T50; } { E T3B, T52, T3y, T53, T3u, T3x; T3B = FMA(KP951056516, T3z, KP587785252 * T3A); T52 = FNMS(KP587785252, T3z, KP951056516 * T3A); T3u = KP559016994 * (T3s - T3t); T3x = FNMS(KP250000000, T3w, T3v); T3y = T3u + T3x; T53 = T3x - T3u; T3C = T3y - T3B; T5A = T53 - T52; T4n = T3B + T3y; T54 = T52 + T53; } } { E T62, T64, Tp, T2k, T5T, T5U, T63, T5V; { E T5Y, T61, T1m, T2j; T5Y = T5W - T5X; T61 = T5Z - T60; T62 = FMA(KP951056516, T5Y, KP587785252 * T61); T64 = FNMS(KP587785252, T5Y, KP951056516 * T61); Tp = T1 + To; T1m = TS + T1l; T2j = T1P + T2i; T2k = T1m + T2j; T5T = KP559016994 * (T1m - T2j); T5U = FNMS(KP250000000, T2k, Tp); } ri[0] = Tp + T2k; T63 = T5U - T5T; ri[WS(rs, 10)] = T63 - T64; ri[WS(rs, 15)] = T63 + T64; T5V = T5T + T5U; ri[WS(rs, 20)] = T5V - T62; ri[WS(rs, 5)] = T5V + T62; } { E T6i, T6j, T6c, T67, T6d, T6e, T6k, T6f; { E T6g, T6h, T65, T66; T6g = TS - T1l; T6h = T1P - T2i; T6i = FMA(KP951056516, T6g, KP587785252 * T6h); T6j = FNMS(KP587785252, T6g, KP951056516 * T6h); T6c = T6a + T6b; T65 = T5W + T5X; T66 = T5Z + T60; T67 = T65 + T66; T6d = KP559016994 * (T65 - T66); T6e = FNMS(KP250000000, T67, T6c); } ii[0] = T67 + T6c; T6k = T6e - T6d; ii[WS(rs, 10)] = T6j + T6k; ii[WS(rs, 15)] = T6k - T6j; T6f = T6d + T6e; ii[WS(rs, 5)] = T6f - T6i; ii[WS(rs, 20)] = T6i + T6f; } { E T2v, T4f, T6u, T6G, T42, T6z, T43, T6y, T4A, T6H, T4D, T6F, T4u, T6L, T4v; E T6K, T48, T6v, T4b, T6n, T2n, T6q; T2n = T2l + T2m; T2v = T2n + T2u; T4f = T2n - T2u; T6q = T6o + T6p; T6u = T6q - T6t; T6G = T6t + T6q; { E T2S, T3f, T3g, T3D, T40, T41; T2S = FMA(KP968583161, T2G, KP248689887 * T2R); T3f = FMA(KP535826794, T33, KP844327925 * T3e); T3g = T2S + T3f; T3D = FMA(KP876306680, T3r, KP481753674 * T3C); T40 = FMA(KP728968627, T3O, KP684547105 * T3Z); T41 = T3D + T40; T42 = T3g + T41; T6z = T3D - T40; T43 = KP559016994 * (T3g - T41); T6y = T2S - T3f; } { E T4y, T4z, T6D, T4B, T4C, T6E; T4y = FNMS(KP844327925, T4g, KP535826794 * T4h); T4z = FNMS(KP637423989, T4k, KP770513242 * T4j); T6D = T4y + T4z; T4B = FMA(KP125333233, T4r, KP992114701 * T4q); T4C = FMA(KP904827052, T4o, KP425779291 * T4n); T6E = T4C + T4B; T4A = T4y - T4z; T6H = KP559016994 * (T6D + T6E); T4D = T4B - T4C; T6F = T6D - T6E; } { E T4i, T4l, T4m, T4p, T4s, T4t; T4i = FMA(KP535826794, T4g, KP844327925 * T4h); T4l = FMA(KP637423989, T4j, KP770513242 * T4k); T4m = T4i - T4l; T4p = FNMS(KP425779291, T4o, KP904827052 * T4n); T4s = FNMS(KP992114701, T4r, KP125333233 * T4q); T4t = T4p + T4s; T4u = T4m + T4t; T6L = T4p - T4s; T4v = KP559016994 * (T4m - T4t); T6K = T4i + T4l; } { E T46, T47, T6l, T49, T4a, T6m; T46 = FNMS(KP248689887, T2G, KP968583161 * T2R); T47 = FNMS(KP844327925, T33, KP535826794 * T3e); T6l = T46 + T47; T49 = FNMS(KP481753674, T3r, KP876306680 * T3C); T4a = FNMS(KP684547105, T3O, KP728968627 * T3Z); T6m = T49 + T4a; T48 = T46 - T47; T6v = KP559016994 * (T6l - T6m); T4b = T49 - T4a; T6n = T6l + T6m; } ri[WS(rs, 1)] = T2v + T42; ii[WS(rs, 1)] = T6n + T6u; ri[WS(rs, 4)] = T4f + T4u; ii[WS(rs, 4)] = T6F + T6G; { E T4c, T4e, T45, T4d, T44; T4c = FMA(KP951056516, T48, KP587785252 * T4b); T4e = FNMS(KP587785252, T48, KP951056516 * T4b); T44 = FNMS(KP250000000, T42, T2v); T45 = T43 + T44; T4d = T44 - T43; ri[WS(rs, 21)] = T45 - T4c; ri[WS(rs, 16)] = T4d + T4e; ri[WS(rs, 6)] = T45 + T4c; ri[WS(rs, 11)] = T4d - T4e; } { E T6A, T6B, T6x, T6C, T6w; T6A = FMA(KP951056516, T6y, KP587785252 * T6z); T6B = FNMS(KP587785252, T6y, KP951056516 * T6z); T6w = FNMS(KP250000000, T6n, T6u); T6x = T6v + T6w; T6C = T6w - T6v; ii[WS(rs, 6)] = T6x - T6A; ii[WS(rs, 16)] = T6C - T6B; ii[WS(rs, 21)] = T6A + T6x; ii[WS(rs, 11)] = T6B + T6C; } { E T4E, T4G, T4x, T4F, T4w; T4E = FMA(KP951056516, T4A, KP587785252 * T4D); T4G = FNMS(KP587785252, T4A, KP951056516 * T4D); T4w = FNMS(KP250000000, T4u, T4f); T4x = T4v + T4w; T4F = T4w - T4v; ri[WS(rs, 24)] = T4x - T4E; ri[WS(rs, 19)] = T4F + T4G; ri[WS(rs, 9)] = T4x + T4E; ri[WS(rs, 14)] = T4F - T4G; } { E T6M, T6N, T6J, T6O, T6I; T6M = FMA(KP951056516, T6K, KP587785252 * T6L); T6N = FNMS(KP587785252, T6K, KP951056516 * T6L); T6I = FNMS(KP250000000, T6F, T6G); T6J = T6H + T6I; T6O = T6I - T6H; ii[WS(rs, 9)] = T6J - T6M; ii[WS(rs, 19)] = T6O - T6N; ii[WS(rs, 24)] = T6M + T6J; ii[WS(rs, 14)] = T6N + T6O; } } { E T4J, T5r, T6U, T76, T5e, T6Z, T5f, T6Y, T5M, T77, T5P, T75, T5G, T7b, T5H; E T7a, T5k, T6V, T5n, T6R, T4H, T6T; T4H = T2m - T2l; T4J = T4H - T4I; T5r = T4H + T4I; T6T = T6p - T6o; T6U = T6S + T6T; T76 = T6T - T6S; { E T4Q, T4X, T4Y, T55, T5c, T5d; T4Q = FMA(KP876306680, T4M, KP481753674 * T4P); T4X = FNMS(KP425779291, T4W, KP904827052 * T4T); T4Y = T4Q + T4X; T55 = FMA(KP535826794, T51, KP844327925 * T54); T5c = FMA(KP062790519, T58, KP998026728 * T5b); T5d = T55 + T5c; T5e = T4Y + T5d; T6Z = T55 - T5c; T5f = KP559016994 * (T4Y - T5d); T6Y = T4Q - T4X; } { E T5K, T5L, T73, T5N, T5O, T74; T5K = FNMS(KP684547105, T5s, KP728968627 * T5t); T5L = FMA(KP125333233, T5w, KP992114701 * T5v); T73 = T5K - T5L; T5N = FNMS(KP998026728, T5z, KP062790519 * T5A); T5O = FMA(KP770513242, T5D, KP637423989 * T5C); T74 = T5N - T5O; T5M = T5K + T5L; T77 = KP559016994 * (T73 - T74); T5P = T5N + T5O; T75 = T73 + T74; } { E T5u, T5x, T5y, T5B, T5E, T5F; T5u = FMA(KP728968627, T5s, KP684547105 * T5t); T5x = FNMS(KP992114701, T5w, KP125333233 * T5v); T5y = T5u + T5x; T5B = FMA(KP062790519, T5z, KP998026728 * T5A); T5E = FNMS(KP637423989, T5D, KP770513242 * T5C); T5F = T5B + T5E; T5G = T5y + T5F; T7b = T5B - T5E; T5H = KP559016994 * (T5y - T5F); T7a = T5u - T5x; } { E T5i, T5j, T6P, T5l, T5m, T6Q; T5i = FNMS(KP481753674, T4M, KP876306680 * T4P); T5j = FMA(KP904827052, T4W, KP425779291 * T4T); T6P = T5i - T5j; T5l = FNMS(KP844327925, T51, KP535826794 * T54); T5m = FNMS(KP998026728, T58, KP062790519 * T5b); T6Q = T5l + T5m; T5k = T5i + T5j; T6V = KP559016994 * (T6P - T6Q); T5n = T5l - T5m; T6R = T6P + T6Q; } ri[WS(rs, 2)] = T4J + T5e; ii[WS(rs, 2)] = T6R + T6U; ri[WS(rs, 3)] = T5r + T5G; ii[WS(rs, 3)] = T75 + T76; { E T5o, T5q, T5h, T5p, T5g; T5o = FMA(KP951056516, T5k, KP587785252 * T5n); T5q = FNMS(KP587785252, T5k, KP951056516 * T5n); T5g = FNMS(KP250000000, T5e, T4J); T5h = T5f + T5g; T5p = T5g - T5f; ri[WS(rs, 22)] = T5h - T5o; ri[WS(rs, 17)] = T5p + T5q; ri[WS(rs, 7)] = T5h + T5o; ri[WS(rs, 12)] = T5p - T5q; } { E T70, T71, T6X, T72, T6W; T70 = FMA(KP951056516, T6Y, KP587785252 * T6Z); T71 = FNMS(KP587785252, T6Y, KP951056516 * T6Z); T6W = FNMS(KP250000000, T6R, T6U); T6X = T6V + T6W; T72 = T6W - T6V; ii[WS(rs, 7)] = T6X - T70; ii[WS(rs, 17)] = T72 - T71; ii[WS(rs, 22)] = T70 + T6X; ii[WS(rs, 12)] = T71 + T72; } { E T5Q, T5S, T5J, T5R, T5I; T5Q = FMA(KP951056516, T5M, KP587785252 * T5P); T5S = FNMS(KP587785252, T5M, KP951056516 * T5P); T5I = FNMS(KP250000000, T5G, T5r); T5J = T5H + T5I; T5R = T5I - T5H; ri[WS(rs, 23)] = T5J - T5Q; ri[WS(rs, 18)] = T5R + T5S; ri[WS(rs, 8)] = T5J + T5Q; ri[WS(rs, 13)] = T5R - T5S; } { E T7c, T7d, T79, T7e, T78; T7c = FMA(KP951056516, T7a, KP587785252 * T7b); T7d = FNMS(KP587785252, T7a, KP951056516 * T7b); T78 = FNMS(KP250000000, T75, T76); T79 = T77 + T78; T7e = T78 - T77; ii[WS(rs, 8)] = T79 - T7c; ii[WS(rs, 18)] = T7e - T7d; ii[WS(rs, 23)] = T7c + T79; ii[WS(rs, 13)] = T7d + T7e; } } } } } static const tw_instr twinstr[] = { {TW_FULL, 0, 25}, {TW_NEXT, 1, 0} }; static const ct_desc desc = { 25, "t1_25", twinstr, &GENUS, {260, 140, 140, 0}, 0, 0, 0 }; void X(codelet_t1_25) (planner *p) { X(kdft_dit_register) (p, t1_25, &desc); } #endif /* HAVE_FMA */ fftw-3.3.3/dft/scalar/codelets/t2_4.c0000644000175000017500000001224212054410257014150 00000000000000/* * Copyright (c) 2003, 2007-11 Matteo Frigo * Copyright (c) 2003, 2007-11 Massachusetts Institute of Technology * * This program is free software; you can redistribute it and/or modify * it under the terms of the GNU General Public License as published by * the Free Software Foundation; either version 2 of the License, or * (at your option) any later version. * * This program is distributed in the hope that it will be useful, * but WITHOUT ANY WARRANTY; without even the implied warranty of * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the * GNU General Public License for more details. * * You should have received a copy of the GNU General Public License * along with this program; if not, write to the Free Software * Foundation, Inc., 51 Franklin Street, Fifth Floor, Boston, MA 02110-1301 USA * */ /* This file was automatically generated --- DO NOT EDIT */ /* Generated on Sun Nov 25 07:35:59 EST 2012 */ #include "codelet-dft.h" #ifdef HAVE_FMA /* Generated by: ../../../genfft/gen_twiddle.native -fma -reorder-insns -schedule-for-pipeline -compact -variables 4 -pipeline-latency 4 -twiddle-log3 -precompute-twiddles -n 4 -name t2_4 -include t.h */ /* * This function contains 24 FP additions, 16 FP multiplications, * (or, 16 additions, 8 multiplications, 8 fused multiply/add), * 33 stack variables, 0 constants, and 16 memory accesses */ #include "t.h" static void t2_4(R *ri, R *ii, const R *W, stride rs, INT mb, INT me, INT ms) { { INT m; for (m = mb, W = W + (mb * 4); m < me; m = m + 1, ri = ri + ms, ii = ii + ms, W = W + 4, MAKE_VOLATILE_STRIDE(8, rs)) { E Ti, Tq, To, Te, Ty, Tz, Tm, Ts; { E T2, T6, T3, T5; T2 = W[0]; T6 = W[3]; T3 = W[2]; T5 = W[1]; { E T1, Tx, Td, Tw, Tj, Tl, Ta, T4, Tk, Tr; T1 = ri[0]; Ta = T2 * T6; T4 = T2 * T3; Tx = ii[0]; { E T8, Tb, T7, Tc; T8 = ri[WS(rs, 2)]; Tb = FNMS(T5, T3, Ta); T7 = FMA(T5, T6, T4); Tc = ii[WS(rs, 2)]; { E Tf, Th, T9, Tv, Tg, Tp; Tf = ri[WS(rs, 1)]; Th = ii[WS(rs, 1)]; T9 = T7 * T8; Tv = T7 * Tc; Tg = T2 * Tf; Tp = T2 * Th; Td = FMA(Tb, Tc, T9); Tw = FNMS(Tb, T8, Tv); Ti = FMA(T5, Th, Tg); Tq = FNMS(T5, Tf, Tp); } Tj = ri[WS(rs, 3)]; Tl = ii[WS(rs, 3)]; } To = T1 - Td; Te = T1 + Td; Ty = Tw + Tx; Tz = Tx - Tw; Tk = T3 * Tj; Tr = T3 * Tl; Tm = FMA(T6, Tl, Tk); Ts = FNMS(T6, Tj, Tr); } } { E Tn, TA, Tu, Tt; Tn = Ti + Tm; TA = Ti - Tm; Tu = Tq + Ts; Tt = Tq - Ts; ii[WS(rs, 3)] = TA + Tz; ii[WS(rs, 1)] = Tz - TA; ri[0] = Te + Tn; ri[WS(rs, 2)] = Te - Tn; ri[WS(rs, 1)] = To + Tt; ri[WS(rs, 3)] = To - Tt; ii[WS(rs, 2)] = Ty - Tu; ii[0] = Tu + Ty; } } } } static const tw_instr twinstr[] = { {TW_CEXP, 0, 1}, {TW_CEXP, 0, 3}, {TW_NEXT, 1, 0} }; static const ct_desc desc = { 4, "t2_4", twinstr, &GENUS, {16, 8, 8, 0}, 0, 0, 0 }; void X(codelet_t2_4) (planner *p) { X(kdft_dit_register) (p, t2_4, &desc); } #else /* HAVE_FMA */ /* Generated by: ../../../genfft/gen_twiddle.native -compact -variables 4 -pipeline-latency 4 -twiddle-log3 -precompute-twiddles -n 4 -name t2_4 -include t.h */ /* * This function contains 24 FP additions, 16 FP multiplications, * (or, 16 additions, 8 multiplications, 8 fused multiply/add), * 21 stack variables, 0 constants, and 16 memory accesses */ #include "t.h" static void t2_4(R *ri, R *ii, const R *W, stride rs, INT mb, INT me, INT ms) { { INT m; for (m = mb, W = W + (mb * 4); m < me; m = m + 1, ri = ri + ms, ii = ii + ms, W = W + 4, MAKE_VOLATILE_STRIDE(8, rs)) { E T2, T4, T3, T5, T6, T8; T2 = W[0]; T4 = W[1]; T3 = W[2]; T5 = W[3]; T6 = FMA(T2, T3, T4 * T5); T8 = FNMS(T4, T3, T2 * T5); { E T1, Tp, Ta, To, Te, Tk, Th, Tl, T7, T9; T1 = ri[0]; Tp = ii[0]; T7 = ri[WS(rs, 2)]; T9 = ii[WS(rs, 2)]; Ta = FMA(T6, T7, T8 * T9); To = FNMS(T8, T7, T6 * T9); { E Tc, Td, Tf, Tg; Tc = ri[WS(rs, 1)]; Td = ii[WS(rs, 1)]; Te = FMA(T2, Tc, T4 * Td); Tk = FNMS(T4, Tc, T2 * Td); Tf = ri[WS(rs, 3)]; Tg = ii[WS(rs, 3)]; Th = FMA(T3, Tf, T5 * Tg); Tl = FNMS(T5, Tf, T3 * Tg); } { E Tb, Ti, Tn, Tq; Tb = T1 + Ta; Ti = Te + Th; ri[WS(rs, 2)] = Tb - Ti; ri[0] = Tb + Ti; Tn = Tk + Tl; Tq = To + Tp; ii[0] = Tn + Tq; ii[WS(rs, 2)] = Tq - Tn; } { E Tj, Tm, Tr, Ts; Tj = T1 - Ta; Tm = Tk - Tl; ri[WS(rs, 3)] = Tj - Tm; ri[WS(rs, 1)] = Tj + Tm; Tr = Tp - To; Ts = Te - Th; ii[WS(rs, 1)] = Tr - Ts; ii[WS(rs, 3)] = Ts + Tr; } } } } } static const tw_instr twinstr[] = { {TW_CEXP, 0, 1}, {TW_CEXP, 0, 3}, {TW_NEXT, 1, 0} }; static const ct_desc desc = { 4, "t2_4", twinstr, &GENUS, {16, 8, 8, 0}, 0, 0, 0 }; void X(codelet_t2_4) (planner *p) { X(kdft_dit_register) (p, t2_4, &desc); } #endif /* HAVE_FMA */ fftw-3.3.3/dft/scalar/codelets/t2_8.c0000644000175000017500000002420312054410260014146 00000000000000/* * Copyright (c) 2003, 2007-11 Matteo Frigo * Copyright (c) 2003, 2007-11 Massachusetts Institute of Technology * * This program is free software; you can redistribute it and/or modify * it under the terms of the GNU General Public License as published by * the Free Software Foundation; either version 2 of the License, or * (at your option) any later version. * * This program is distributed in the hope that it will be useful, * but WITHOUT ANY WARRANTY; without even the implied warranty of * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the * GNU General Public License for more details. * * You should have received a copy of the GNU General Public License * along with this program; if not, write to the Free Software * Foundation, Inc., 51 Franklin Street, Fifth Floor, Boston, MA 02110-1301 USA * */ /* This file was automatically generated --- DO NOT EDIT */ /* Generated on Sun Nov 25 07:35:59 EST 2012 */ #include "codelet-dft.h" #ifdef HAVE_FMA /* Generated by: ../../../genfft/gen_twiddle.native -fma -reorder-insns -schedule-for-pipeline -compact -variables 4 -pipeline-latency 4 -twiddle-log3 -precompute-twiddles -n 8 -name t2_8 -include t.h */ /* * This function contains 74 FP additions, 50 FP multiplications, * (or, 44 additions, 20 multiplications, 30 fused multiply/add), * 64 stack variables, 1 constants, and 32 memory accesses */ #include "t.h" static void t2_8(R *ri, R *ii, const R *W, stride rs, INT mb, INT me, INT ms) { DK(KP707106781, +0.707106781186547524400844362104849039284835938); { INT m; for (m = mb, W = W + (mb * 6); m < me; m = m + 1, ri = ri + ms, ii = ii + ms, W = W + 6, MAKE_VOLATILE_STRIDE(16, rs)) { E TS, T1m, TJ, T1l, T1k, Tw, T1w, T1u; { E T2, T3, Tl, Tn, T5, T4, Tm, Tr, T6; T2 = W[0]; T3 = W[2]; Tl = W[4]; Tn = W[5]; T5 = W[1]; T4 = T2 * T3; Tm = T2 * Tl; Tr = T2 * Tn; T6 = W[3]; { E T1, T1s, TG, Td, T1r, Tu, TY, Tk, TW, T18, T1d, TD, TH, TA, T13; E TE, T14; { E To, Ts, Tf, T7, T8, Ti, Tb, T9, Tc, TC, Ta, TF, TB, Tg, Th; E Tj; T1 = ri[0]; To = FMA(T5, Tn, Tm); Ts = FNMS(T5, Tl, Tr); Tf = FMA(T5, T6, T4); T7 = FNMS(T5, T6, T4); Ta = T2 * T6; T1s = ii[0]; T8 = ri[WS(rs, 4)]; TF = Tf * Tn; TB = Tf * Tl; Ti = FNMS(T5, T3, Ta); Tb = FMA(T5, T3, Ta); T9 = T7 * T8; Tc = ii[WS(rs, 4)]; TG = FNMS(Ti, Tl, TF); TC = FMA(Ti, Tn, TB); { E Tp, T1q, Tt, Tq, TX; Tp = ri[WS(rs, 6)]; Td = FMA(Tb, Tc, T9); T1q = T7 * Tc; Tt = ii[WS(rs, 6)]; Tq = To * Tp; Tg = ri[WS(rs, 2)]; T1r = FNMS(Tb, T8, T1q); TX = To * Tt; Tu = FMA(Ts, Tt, Tq); Th = Tf * Tg; Tj = ii[WS(rs, 2)]; TY = FNMS(Ts, Tp, TX); } { E TO, TQ, TN, TP, T1a, T1b; { E TK, TM, TL, T19, TV; TK = ri[WS(rs, 7)]; TM = ii[WS(rs, 7)]; Tk = FMA(Ti, Tj, Th); TV = Tf * Tj; TL = Tl * TK; T19 = Tl * TM; TO = ri[WS(rs, 3)]; TW = FNMS(Ti, Tg, TV); TQ = ii[WS(rs, 3)]; TN = FMA(Tn, TM, TL); TP = T3 * TO; T1a = FNMS(Tn, TK, T19); T1b = T3 * TQ; } { E Tx, Tz, Ty, T12, T1c, TR; Tx = ri[WS(rs, 1)]; TR = FMA(T6, TQ, TP); Tz = ii[WS(rs, 1)]; T1c = FNMS(T6, TO, T1b); Ty = T2 * Tx; T18 = TN - TR; TS = TN + TR; T12 = T2 * Tz; T1d = T1a - T1c; T1m = T1a + T1c; TD = ri[WS(rs, 5)]; TH = ii[WS(rs, 5)]; TA = FMA(T5, Tz, Ty); T13 = FNMS(T5, Tx, T12); TE = TC * TD; T14 = TC * TH; } } } { E Te, T1p, T1t, Tv; { E T1g, T10, T1z, T1B, T1A, T1j, T1C, T1f; { E T1x, T11, T16, T1y; { E TU, TZ, TI, T15; Te = T1 + Td; TU = T1 - Td; TZ = TW - TY; T1p = TW + TY; TI = FMA(TG, TH, TE); T15 = FNMS(TG, TD, T14); T1t = T1r + T1s; T1x = T1s - T1r; T1g = TU - TZ; T10 = TU + TZ; T11 = TA - TI; TJ = TA + TI; T1l = T13 + T15; T16 = T13 - T15; T1y = Tk - Tu; Tv = Tk + Tu; } { E T1i, T1e, T17, T1h; T1i = T18 + T1d; T1e = T18 - T1d; T17 = T11 + T16; T1h = T16 - T11; T1z = T1x - T1y; T1B = T1y + T1x; T1A = T1h + T1i; T1j = T1h - T1i; T1C = T1e - T17; T1f = T17 + T1e; } } ri[WS(rs, 7)] = FNMS(KP707106781, T1j, T1g); ii[WS(rs, 7)] = FNMS(KP707106781, T1C, T1B); ri[WS(rs, 1)] = FMA(KP707106781, T1f, T10); ri[WS(rs, 5)] = FNMS(KP707106781, T1f, T10); ii[WS(rs, 1)] = FMA(KP707106781, T1A, T1z); ii[WS(rs, 5)] = FNMS(KP707106781, T1A, T1z); ri[WS(rs, 3)] = FMA(KP707106781, T1j, T1g); ii[WS(rs, 3)] = FMA(KP707106781, T1C, T1B); } T1k = Te - Tv; Tw = Te + Tv; T1w = T1t - T1p; T1u = T1p + T1t; } } } { E TT, T1v, T1n, T1o; TT = TJ + TS; T1v = TS - TJ; T1n = T1l - T1m; T1o = T1l + T1m; ii[WS(rs, 2)] = T1v + T1w; ii[WS(rs, 6)] = T1w - T1v; ri[0] = Tw + TT; ri[WS(rs, 4)] = Tw - TT; ii[0] = T1o + T1u; ii[WS(rs, 4)] = T1u - T1o; ri[WS(rs, 2)] = T1k + T1n; ri[WS(rs, 6)] = T1k - T1n; } } } } static const tw_instr twinstr[] = { {TW_CEXP, 0, 1}, {TW_CEXP, 0, 3}, {TW_CEXP, 0, 7}, {TW_NEXT, 1, 0} }; static const ct_desc desc = { 8, "t2_8", twinstr, &GENUS, {44, 20, 30, 0}, 0, 0, 0 }; void X(codelet_t2_8) (planner *p) { X(kdft_dit_register) (p, t2_8, &desc); } #else /* HAVE_FMA */ /* Generated by: ../../../genfft/gen_twiddle.native -compact -variables 4 -pipeline-latency 4 -twiddle-log3 -precompute-twiddles -n 8 -name t2_8 -include t.h */ /* * This function contains 74 FP additions, 44 FP multiplications, * (or, 56 additions, 26 multiplications, 18 fused multiply/add), * 42 stack variables, 1 constants, and 32 memory accesses */ #include "t.h" static void t2_8(R *ri, R *ii, const R *W, stride rs, INT mb, INT me, INT ms) { DK(KP707106781, +0.707106781186547524400844362104849039284835938); { INT m; for (m = mb, W = W + (mb * 6); m < me; m = m + 1, ri = ri + ms, ii = ii + ms, W = W + 6, MAKE_VOLATILE_STRIDE(16, rs)) { E T2, T5, T3, T6, T8, Tc, Tg, Ti, Tl, Tm, Tn, Tz, Tp, Tx; { E T4, Tb, T7, Ta; T2 = W[0]; T5 = W[1]; T3 = W[2]; T6 = W[3]; T4 = T2 * T3; Tb = T5 * T3; T7 = T5 * T6; Ta = T2 * T6; T8 = T4 - T7; Tc = Ta + Tb; Tg = T4 + T7; Ti = Ta - Tb; Tl = W[4]; Tm = W[5]; Tn = FMA(T2, Tl, T5 * Tm); Tz = FNMS(Ti, Tl, Tg * Tm); Tp = FNMS(T5, Tl, T2 * Tm); Tx = FMA(Tg, Tl, Ti * Tm); } { E Tf, T1i, TL, T1d, TJ, T17, TV, TY, Ts, T1j, TO, T1a, TC, T16, TQ; E TT; { E T1, T1c, Te, T1b, T9, Td; T1 = ri[0]; T1c = ii[0]; T9 = ri[WS(rs, 4)]; Td = ii[WS(rs, 4)]; Te = FMA(T8, T9, Tc * Td); T1b = FNMS(Tc, T9, T8 * Td); Tf = T1 + Te; T1i = T1c - T1b; TL = T1 - Te; T1d = T1b + T1c; } { E TF, TW, TI, TX; { E TD, TE, TG, TH; TD = ri[WS(rs, 7)]; TE = ii[WS(rs, 7)]; TF = FMA(Tl, TD, Tm * TE); TW = FNMS(Tm, TD, Tl * TE); TG = ri[WS(rs, 3)]; TH = ii[WS(rs, 3)]; TI = FMA(T3, TG, T6 * TH); TX = FNMS(T6, TG, T3 * TH); } TJ = TF + TI; T17 = TW + TX; TV = TF - TI; TY = TW - TX; } { E Tk, TM, Tr, TN; { E Th, Tj, To, Tq; Th = ri[WS(rs, 2)]; Tj = ii[WS(rs, 2)]; Tk = FMA(Tg, Th, Ti * Tj); TM = FNMS(Ti, Th, Tg * Tj); To = ri[WS(rs, 6)]; Tq = ii[WS(rs, 6)]; Tr = FMA(Tn, To, Tp * Tq); TN = FNMS(Tp, To, Tn * Tq); } Ts = Tk + Tr; T1j = Tk - Tr; TO = TM - TN; T1a = TM + TN; } { E Tw, TR, TB, TS; { E Tu, Tv, Ty, TA; Tu = ri[WS(rs, 1)]; Tv = ii[WS(rs, 1)]; Tw = FMA(T2, Tu, T5 * Tv); TR = FNMS(T5, Tu, T2 * Tv); Ty = ri[WS(rs, 5)]; TA = ii[WS(rs, 5)]; TB = FMA(Tx, Ty, Tz * TA); TS = FNMS(Tz, Ty, Tx * TA); } TC = Tw + TB; T16 = TR + TS; TQ = Tw - TB; TT = TR - TS; } { E Tt, TK, T1f, T1g; Tt = Tf + Ts; TK = TC + TJ; ri[WS(rs, 4)] = Tt - TK; ri[0] = Tt + TK; { E T19, T1e, T15, T18; T19 = T16 + T17; T1e = T1a + T1d; ii[0] = T19 + T1e; ii[WS(rs, 4)] = T1e - T19; T15 = Tf - Ts; T18 = T16 - T17; ri[WS(rs, 6)] = T15 - T18; ri[WS(rs, 2)] = T15 + T18; } T1f = TJ - TC; T1g = T1d - T1a; ii[WS(rs, 2)] = T1f + T1g; ii[WS(rs, 6)] = T1g - T1f; { E T11, T1k, T14, T1h, T12, T13; T11 = TL - TO; T1k = T1i - T1j; T12 = TT - TQ; T13 = TV + TY; T14 = KP707106781 * (T12 - T13); T1h = KP707106781 * (T12 + T13); ri[WS(rs, 7)] = T11 - T14; ii[WS(rs, 5)] = T1k - T1h; ri[WS(rs, 3)] = T11 + T14; ii[WS(rs, 1)] = T1h + T1k; } { E TP, T1m, T10, T1l, TU, TZ; TP = TL + TO; T1m = T1j + T1i; TU = TQ + TT; TZ = TV - TY; T10 = KP707106781 * (TU + TZ); T1l = KP707106781 * (TZ - TU); ri[WS(rs, 5)] = TP - T10; ii[WS(rs, 7)] = T1m - T1l; ri[WS(rs, 1)] = TP + T10; ii[WS(rs, 3)] = T1l + T1m; } } } } } } static const tw_instr twinstr[] = { {TW_CEXP, 0, 1}, {TW_CEXP, 0, 3}, {TW_CEXP, 0, 7}, {TW_NEXT, 1, 0} }; static const ct_desc desc = { 8, "t2_8", twinstr, &GENUS, {56, 26, 18, 0}, 0, 0, 0 }; void X(codelet_t2_8) (planner *p) { X(kdft_dit_register) (p, t2_8, &desc); } #endif /* HAVE_FMA */ fftw-3.3.3/dft/scalar/codelets/t2_16.c0000644000175000017500000005463412054410261014241 00000000000000/* * Copyright (c) 2003, 2007-11 Matteo Frigo * Copyright (c) 2003, 2007-11 Massachusetts Institute of Technology * * This program is free software; you can redistribute it and/or modify * it under the terms of the GNU General Public License as published by * the Free Software Foundation; either version 2 of the License, or * (at your option) any later version. * * This program is distributed in the hope that it will be useful, * but WITHOUT ANY WARRANTY; without even the implied warranty of * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the * GNU General Public License for more details. * * You should have received a copy of the GNU General Public License * along with this program; if not, write to the Free Software * Foundation, Inc., 51 Franklin Street, Fifth Floor, Boston, MA 02110-1301 USA * */ /* This file was automatically generated --- DO NOT EDIT */ /* Generated on Sun Nov 25 07:36:00 EST 2012 */ #include "codelet-dft.h" #ifdef HAVE_FMA /* Generated by: ../../../genfft/gen_twiddle.native -fma -reorder-insns -schedule-for-pipeline -compact -variables 4 -pipeline-latency 4 -twiddle-log3 -precompute-twiddles -n 16 -name t2_16 -include t.h */ /* * This function contains 196 FP additions, 134 FP multiplications, * (or, 104 additions, 42 multiplications, 92 fused multiply/add), * 100 stack variables, 3 constants, and 64 memory accesses */ #include "t.h" static void t2_16(R *ri, R *ii, const R *W, stride rs, INT mb, INT me, INT ms) { DK(KP923879532, +0.923879532511286756128183189396788286822416626); DK(KP414213562, +0.414213562373095048801688724209698078569671875); DK(KP707106781, +0.707106781186547524400844362104849039284835938); { INT m; for (m = mb, W = W + (mb * 8); m < me; m = m + 1, ri = ri + ms, ii = ii + ms, W = W + 8, MAKE_VOLATILE_STRIDE(32, rs)) { E T3S, T3R; { E T2, Tf, TM, TO, T3, Tg, TN, TS, T4, Tp, T6, T5, Th; T2 = W[0]; Tf = W[2]; TM = W[6]; TO = W[7]; T3 = W[4]; Tg = T2 * Tf; TN = T2 * TM; TS = T2 * TO; T4 = T2 * T3; Tp = Tf * T3; T6 = W[5]; T5 = W[1]; Th = W[3]; { E TZ, Te, T1U, T3A, T3L, T2D, T1G, T2A, T3h, T1R, T2B, T2I, T3i, Tx, T3M; E T1Z, T3w, TL, T26, T25, T37, T1d, T2o, T2l, T3c, T1s, T2m, T2t, T3d, TX; E T10, TV, T2a, TY, T2b; { E TF, TP, TT, Tq, TW, Tz, Tu, TI, TC, T1m, T1f, T1p, T1j, Tr, Ts; E Tv, To, T1W; { E Ti, Tm, T1L, T1O, T1D, T1A, T1x, T2y, T1F, T2x; { E T1, T7, Tb, T3z, T8, T1z, T9, Tc; { E T1i, T1e, T1C, T1y, Tt, Ta, Tl; T1 = ri[0]; Tt = Tf * T6; Ta = T2 * T6; T7 = FMA(T5, T6, T4); TF = FNMS(T5, T6, T4); TP = FMA(T5, TO, TN); TT = FNMS(T5, TM, TS); Tq = FNMS(Th, T6, Tp); TW = FMA(Th, T6, Tp); Tz = FMA(T5, Th, Tg); Ti = FNMS(T5, Th, Tg); Tl = T2 * Th; Tu = FMA(Th, T3, Tt); TZ = FNMS(Th, T3, Tt); TI = FMA(T5, T3, Ta); Tb = FNMS(T5, T3, Ta); T1i = Ti * T6; T1e = Ti * T3; T1C = Tz * T6; T1y = Tz * T3; Tm = FMA(T5, Tf, Tl); TC = FNMS(T5, Tf, Tl); T3z = ii[0]; T8 = ri[WS(rs, 8)]; T1m = FNMS(Tm, T6, T1e); T1f = FMA(Tm, T6, T1e); T1p = FMA(Tm, T3, T1i); T1j = FNMS(Tm, T3, T1i); T1L = FNMS(TC, T6, T1y); T1z = FMA(TC, T6, T1y); T1O = FMA(TC, T3, T1C); T1D = FNMS(TC, T3, T1C); T9 = T7 * T8; Tc = ii[WS(rs, 8)]; } { E T1u, T1w, T1v, T2w, T3y, T1B, T1E, Td, T3x; T1u = ri[WS(rs, 15)]; T1w = ii[WS(rs, 15)]; T1A = ri[WS(rs, 7)]; Td = FMA(Tb, Tc, T9); T3x = T7 * Tc; T1v = TM * T1u; T2w = TM * T1w; Te = T1 + Td; T1U = T1 - Td; T3y = FNMS(Tb, T8, T3x); T1B = T1z * T1A; T1E = ii[WS(rs, 7)]; T1x = FMA(TO, T1w, T1v); T3A = T3y + T3z; T3L = T3z - T3y; T2y = T1z * T1E; T1F = FMA(T1D, T1E, T1B); T2x = FNMS(TO, T1u, T2w); } } { E T1H, T1I, T1J, T1M, T1P, T2z; T1H = ri[WS(rs, 3)]; T2z = FNMS(T1D, T1A, T2y); T2D = T1x - T1F; T1G = T1x + T1F; T1I = Tf * T1H; T2A = T2x - T2z; T3h = T2x + T2z; T1J = ii[WS(rs, 3)]; T1M = ri[WS(rs, 11)]; T1P = ii[WS(rs, 11)]; { E Tj, Tk, Tn, T1V; { E T1K, T2F, T1Q, T2H, T2E, T1N, T2G; Tj = ri[WS(rs, 4)]; T1K = FMA(Th, T1J, T1I); T2E = Tf * T1J; T1N = T1L * T1M; T2G = T1L * T1P; Tk = Ti * Tj; T2F = FNMS(Th, T1H, T2E); T1Q = FMA(T1O, T1P, T1N); T2H = FNMS(T1O, T1M, T2G); Tn = ii[WS(rs, 4)]; Tr = ri[WS(rs, 12)]; T1R = T1K + T1Q; T2B = T1K - T1Q; T2I = T2F - T2H; T3i = T2F + T2H; T1V = Ti * Tn; Ts = Tq * Tr; Tv = ii[WS(rs, 12)]; } To = FMA(Tm, Tn, Tk); T1W = FNMS(Tm, Tj, T1V); } } } { E T19, T1b, T18, T2i, T1a, T2j; { E TE, T22, TK, T24; { E TA, TD, TB, T21, TG, TJ, TH, T23, T1Y, Tw, T1X; TA = ri[WS(rs, 2)]; Tw = FMA(Tu, Tv, Ts); T1X = Tq * Tv; TD = ii[WS(rs, 2)]; TB = Tz * TA; Tx = To + Tw; T3M = To - Tw; T1Y = FNMS(Tu, Tr, T1X); T21 = Tz * TD; TG = ri[WS(rs, 10)]; TJ = ii[WS(rs, 10)]; T1Z = T1W - T1Y; T3w = T1W + T1Y; TH = TF * TG; T23 = TF * TJ; TE = FMA(TC, TD, TB); T22 = FNMS(TC, TA, T21); TK = FMA(TI, TJ, TH); T24 = FNMS(TI, TG, T23); } { E T15, T17, T16, T2h; T15 = ri[WS(rs, 1)]; T17 = ii[WS(rs, 1)]; TL = TE + TK; T26 = TE - TK; T25 = T22 - T24; T37 = T22 + T24; T16 = T2 * T15; T2h = T2 * T17; T19 = ri[WS(rs, 9)]; T1b = ii[WS(rs, 9)]; T18 = FMA(T5, T17, T16); T2i = FNMS(T5, T15, T2h); T1a = T3 * T19; T2j = T3 * T1b; } } { E T1n, T1q, T1l, T2q, T1o, T2r; { E T1g, T1k, T1h, T2p, T1c, T2k; T1g = ri[WS(rs, 5)]; T1k = ii[WS(rs, 5)]; T1c = FMA(T6, T1b, T1a); T2k = FNMS(T6, T19, T2j); T1h = T1f * T1g; T2p = T1f * T1k; T1d = T18 + T1c; T2o = T18 - T1c; T2l = T2i - T2k; T3c = T2i + T2k; T1n = ri[WS(rs, 13)]; T1q = ii[WS(rs, 13)]; T1l = FMA(T1j, T1k, T1h); T2q = FNMS(T1j, T1g, T2p); T1o = T1m * T1n; T2r = T1m * T1q; } { E TQ, TU, TR, T29, T1r, T2s; TQ = ri[WS(rs, 14)]; TU = ii[WS(rs, 14)]; T1r = FMA(T1p, T1q, T1o); T2s = FNMS(T1p, T1n, T2r); TR = TP * TQ; T29 = TP * TU; T1s = T1l + T1r; T2m = T1l - T1r; T2t = T2q - T2s; T3d = T2q + T2s; TX = ri[WS(rs, 6)]; T10 = ii[WS(rs, 6)]; TV = FMA(TT, TU, TR); T2a = FNMS(TT, TQ, T29); TY = TW * TX; T2b = TW * T10; } } } } { E T36, T3G, T3b, T3g, T28, T2d, T3F, T39, T3e, T3q, T3C, T3j, T3u, T3t; { E T3D, T1T, T3r, T14, T3E, T3s; { E Ty, T3B, T11, T2c, T13, T3v; T36 = Te - Tx; Ty = Te + Tx; T3B = T3w + T3A; T3G = T3A - T3w; T11 = FMA(TZ, T10, TY); T2c = FNMS(TZ, TX, T2b); { E T1t, T1S, T12, T38; T3b = T1d - T1s; T1t = T1d + T1s; T1S = T1G + T1R; T3g = T1G - T1R; T12 = TV + T11; T28 = TV - T11; T2d = T2a - T2c; T38 = T2a + T2c; T3D = T1S - T1t; T1T = T1t + T1S; T13 = TL + T12; T3F = T12 - TL; T39 = T37 - T38; T3v = T37 + T38; } T3e = T3c - T3d; T3r = T3c + T3d; T3q = Ty - T13; T14 = Ty + T13; T3E = T3B - T3v; T3C = T3v + T3B; T3s = T3h + T3i; T3j = T3h - T3i; } ri[WS(rs, 8)] = T14 - T1T; ri[0] = T14 + T1T; ii[WS(rs, 12)] = T3E - T3D; T3u = T3r + T3s; T3t = T3r - T3s; ii[WS(rs, 4)] = T3D + T3E; } { E T3m, T3a, T3J, T3H; ii[0] = T3u + T3C; ii[WS(rs, 8)] = T3C - T3u; ri[WS(rs, 4)] = T3q + T3t; ri[WS(rs, 12)] = T3q - T3t; T3m = T36 - T39; T3a = T36 + T39; T3J = T3G - T3F; T3H = T3F + T3G; { E T2Q, T20, T3N, T3T, T2J, T2C, T3O, T2f, T34, T30, T2W, T2V, T3U, T2T, T2N; E T2v; { E T2R, T27, T2e, T2S; { E T3n, T3f, T3o, T3k; T2Q = T1U + T1Z; T20 = T1U - T1Z; T3n = T3e - T3b; T3f = T3b + T3e; T3o = T3g + T3j; T3k = T3g - T3j; T3N = T3L - T3M; T3T = T3M + T3L; { E T3p, T3I, T3K, T3l; T3p = T3n - T3o; T3I = T3n + T3o; T3K = T3k - T3f; T3l = T3f + T3k; ri[WS(rs, 6)] = FMA(KP707106781, T3p, T3m); ri[WS(rs, 14)] = FNMS(KP707106781, T3p, T3m); ii[WS(rs, 10)] = FNMS(KP707106781, T3I, T3H); ii[WS(rs, 2)] = FMA(KP707106781, T3I, T3H); ii[WS(rs, 14)] = FNMS(KP707106781, T3K, T3J); ii[WS(rs, 6)] = FMA(KP707106781, T3K, T3J); ri[WS(rs, 2)] = FMA(KP707106781, T3l, T3a); ri[WS(rs, 10)] = FNMS(KP707106781, T3l, T3a); T2R = T26 + T25; T27 = T25 - T26; T2e = T28 + T2d; T2S = T28 - T2d; } } { E T2Y, T2Z, T2n, T2u; T2J = T2D - T2I; T2Y = T2D + T2I; T2Z = T2A - T2B; T2C = T2A + T2B; T3O = T27 + T2e; T2f = T27 - T2e; T34 = FMA(KP414213562, T2Y, T2Z); T30 = FNMS(KP414213562, T2Z, T2Y); T2W = T2l - T2m; T2n = T2l + T2m; T2u = T2o - T2t; T2V = T2o + T2t; T3U = T2S - T2R; T2T = T2R + T2S; T2N = FNMS(KP414213562, T2n, T2u); T2v = FMA(KP414213562, T2u, T2n); } } { E T33, T2X, T3X, T3Y; { E T2M, T2g, T2O, T2K, T3V, T3W, T2P, T2L; T2M = FNMS(KP707106781, T2f, T20); T2g = FMA(KP707106781, T2f, T20); T33 = FNMS(KP414213562, T2V, T2W); T2X = FMA(KP414213562, T2W, T2V); T2O = FMA(KP414213562, T2C, T2J); T2K = FNMS(KP414213562, T2J, T2C); T3V = FMA(KP707106781, T3U, T3T); T3X = FNMS(KP707106781, T3U, T3T); T3W = T2O - T2N; T2P = T2N + T2O; T3Y = T2v + T2K; T2L = T2v - T2K; ii[WS(rs, 11)] = FNMS(KP923879532, T3W, T3V); ii[WS(rs, 3)] = FMA(KP923879532, T3W, T3V); ri[WS(rs, 3)] = FMA(KP923879532, T2L, T2g); ri[WS(rs, 11)] = FNMS(KP923879532, T2L, T2g); ri[WS(rs, 15)] = FMA(KP923879532, T2P, T2M); ri[WS(rs, 7)] = FNMS(KP923879532, T2P, T2M); } { E T32, T3P, T3Q, T35, T2U, T31; T32 = FNMS(KP707106781, T2T, T2Q); T2U = FMA(KP707106781, T2T, T2Q); T31 = T2X + T30; T3S = T30 - T2X; T3R = FNMS(KP707106781, T3O, T3N); T3P = FMA(KP707106781, T3O, T3N); ii[WS(rs, 15)] = FMA(KP923879532, T3Y, T3X); ii[WS(rs, 7)] = FNMS(KP923879532, T3Y, T3X); ri[WS(rs, 1)] = FMA(KP923879532, T31, T2U); ri[WS(rs, 9)] = FNMS(KP923879532, T31, T2U); T3Q = T33 + T34; T35 = T33 - T34; ii[WS(rs, 9)] = FNMS(KP923879532, T3Q, T3P); ii[WS(rs, 1)] = FMA(KP923879532, T3Q, T3P); ri[WS(rs, 5)] = FMA(KP923879532, T35, T32); ri[WS(rs, 13)] = FNMS(KP923879532, T35, T32); } } } } } } } ii[WS(rs, 13)] = FNMS(KP923879532, T3S, T3R); ii[WS(rs, 5)] = FMA(KP923879532, T3S, T3R); } } } static const tw_instr twinstr[] = { {TW_CEXP, 0, 1}, {TW_CEXP, 0, 3}, {TW_CEXP, 0, 9}, {TW_CEXP, 0, 15}, {TW_NEXT, 1, 0} }; static const ct_desc desc = { 16, "t2_16", twinstr, &GENUS, {104, 42, 92, 0}, 0, 0, 0 }; void X(codelet_t2_16) (planner *p) { X(kdft_dit_register) (p, t2_16, &desc); } #else /* HAVE_FMA */ /* Generated by: ../../../genfft/gen_twiddle.native -compact -variables 4 -pipeline-latency 4 -twiddle-log3 -precompute-twiddles -n 16 -name t2_16 -include t.h */ /* * This function contains 196 FP additions, 108 FP multiplications, * (or, 156 additions, 68 multiplications, 40 fused multiply/add), * 82 stack variables, 3 constants, and 64 memory accesses */ #include "t.h" static void t2_16(R *ri, R *ii, const R *W, stride rs, INT mb, INT me, INT ms) { DK(KP382683432, +0.382683432365089771728459984030398866761344562); DK(KP923879532, +0.923879532511286756128183189396788286822416626); DK(KP707106781, +0.707106781186547524400844362104849039284835938); { INT m; for (m = mb, W = W + (mb * 8); m < me; m = m + 1, ri = ri + ms, ii = ii + ms, W = W + 8, MAKE_VOLATILE_STRIDE(32, rs)) { E T2, T5, Tg, Ti, Tk, To, TE, TC, T6, T3, T8, TW, TJ, Tt, TU; E Tc, Tx, TH, TN, TO, TP, TR, T1f, T1k, T1b, T1i, T1y, T1H, T1u, T1F; { E T7, Tv, Ta, Ts, T4, Tw, Tb, Tr; { E Th, Tn, Tj, Tm; T2 = W[0]; T5 = W[1]; Tg = W[2]; Ti = W[3]; Th = T2 * Tg; Tn = T5 * Tg; Tj = T5 * Ti; Tm = T2 * Ti; Tk = Th - Tj; To = Tm + Tn; TE = Tm - Tn; TC = Th + Tj; T6 = W[5]; T7 = T5 * T6; Tv = Tg * T6; Ta = T2 * T6; Ts = Ti * T6; T3 = W[4]; T4 = T2 * T3; Tw = Ti * T3; Tb = T5 * T3; Tr = Tg * T3; } T8 = T4 + T7; TW = Tv - Tw; TJ = Ta + Tb; Tt = Tr - Ts; TU = Tr + Ts; Tc = Ta - Tb; Tx = Tv + Tw; TH = T4 - T7; TN = W[6]; TO = W[7]; TP = FMA(T2, TN, T5 * TO); TR = FNMS(T5, TN, T2 * TO); { E T1d, T1e, T19, T1a; T1d = Tk * T6; T1e = To * T3; T1f = T1d - T1e; T1k = T1d + T1e; T19 = Tk * T3; T1a = To * T6; T1b = T19 + T1a; T1i = T19 - T1a; } { E T1w, T1x, T1s, T1t; T1w = TC * T6; T1x = TE * T3; T1y = T1w - T1x; T1H = T1w + T1x; T1s = TC * T3; T1t = TE * T6; T1u = T1s + T1t; T1F = T1s - T1t; } } { E Tf, T3r, T1N, T3e, TA, T3s, T1Q, T3b, TM, T2M, T1W, T2w, TZ, T2N, T21; E T2x, T1B, T1K, T2V, T2W, T2X, T2Y, T2j, T2D, T2o, T2E, T18, T1n, T2Q, T2R; E T2S, T2T, T28, T2A, T2d, T2B; { E T1, T3d, Te, T3c, T9, Td; T1 = ri[0]; T3d = ii[0]; T9 = ri[WS(rs, 8)]; Td = ii[WS(rs, 8)]; Te = FMA(T8, T9, Tc * Td); T3c = FNMS(Tc, T9, T8 * Td); Tf = T1 + Te; T3r = T3d - T3c; T1N = T1 - Te; T3e = T3c + T3d; } { E Tq, T1O, Tz, T1P; { E Tl, Tp, Tu, Ty; Tl = ri[WS(rs, 4)]; Tp = ii[WS(rs, 4)]; Tq = FMA(Tk, Tl, To * Tp); T1O = FNMS(To, Tl, Tk * Tp); Tu = ri[WS(rs, 12)]; Ty = ii[WS(rs, 12)]; Tz = FMA(Tt, Tu, Tx * Ty); T1P = FNMS(Tx, Tu, Tt * Ty); } TA = Tq + Tz; T3s = Tq - Tz; T1Q = T1O - T1P; T3b = T1O + T1P; } { E TG, T1S, TL, T1T, T1U, T1V; { E TD, TF, TI, TK; TD = ri[WS(rs, 2)]; TF = ii[WS(rs, 2)]; TG = FMA(TC, TD, TE * TF); T1S = FNMS(TE, TD, TC * TF); TI = ri[WS(rs, 10)]; TK = ii[WS(rs, 10)]; TL = FMA(TH, TI, TJ * TK); T1T = FNMS(TJ, TI, TH * TK); } TM = TG + TL; T2M = T1S + T1T; T1U = T1S - T1T; T1V = TG - TL; T1W = T1U - T1V; T2w = T1V + T1U; } { E TT, T1Y, TY, T1Z, T1X, T20; { E TQ, TS, TV, TX; TQ = ri[WS(rs, 14)]; TS = ii[WS(rs, 14)]; TT = FMA(TP, TQ, TR * TS); T1Y = FNMS(TR, TQ, TP * TS); TV = ri[WS(rs, 6)]; TX = ii[WS(rs, 6)]; TY = FMA(TU, TV, TW * TX); T1Z = FNMS(TW, TV, TU * TX); } TZ = TT + TY; T2N = T1Y + T1Z; T1X = TT - TY; T20 = T1Y - T1Z; T21 = T1X + T20; T2x = T1X - T20; } { E T1r, T2k, T1J, T2h, T1A, T2l, T1E, T2g; { E T1p, T1q, T1G, T1I; T1p = ri[WS(rs, 15)]; T1q = ii[WS(rs, 15)]; T1r = FMA(TN, T1p, TO * T1q); T2k = FNMS(TO, T1p, TN * T1q); T1G = ri[WS(rs, 11)]; T1I = ii[WS(rs, 11)]; T1J = FMA(T1F, T1G, T1H * T1I); T2h = FNMS(T1H, T1G, T1F * T1I); } { E T1v, T1z, T1C, T1D; T1v = ri[WS(rs, 7)]; T1z = ii[WS(rs, 7)]; T1A = FMA(T1u, T1v, T1y * T1z); T2l = FNMS(T1y, T1v, T1u * T1z); T1C = ri[WS(rs, 3)]; T1D = ii[WS(rs, 3)]; T1E = FMA(Tg, T1C, Ti * T1D); T2g = FNMS(Ti, T1C, Tg * T1D); } T1B = T1r + T1A; T1K = T1E + T1J; T2V = T1B - T1K; T2W = T2k + T2l; T2X = T2g + T2h; T2Y = T2W - T2X; { E T2f, T2i, T2m, T2n; T2f = T1r - T1A; T2i = T2g - T2h; T2j = T2f - T2i; T2D = T2f + T2i; T2m = T2k - T2l; T2n = T1E - T1J; T2o = T2m + T2n; T2E = T2m - T2n; } } { E T14, T24, T1m, T2b, T17, T25, T1h, T2a; { E T12, T13, T1j, T1l; T12 = ri[WS(rs, 1)]; T13 = ii[WS(rs, 1)]; T14 = FMA(T2, T12, T5 * T13); T24 = FNMS(T5, T12, T2 * T13); T1j = ri[WS(rs, 13)]; T1l = ii[WS(rs, 13)]; T1m = FMA(T1i, T1j, T1k * T1l); T2b = FNMS(T1k, T1j, T1i * T1l); } { E T15, T16, T1c, T1g; T15 = ri[WS(rs, 9)]; T16 = ii[WS(rs, 9)]; T17 = FMA(T3, T15, T6 * T16); T25 = FNMS(T6, T15, T3 * T16); T1c = ri[WS(rs, 5)]; T1g = ii[WS(rs, 5)]; T1h = FMA(T1b, T1c, T1f * T1g); T2a = FNMS(T1f, T1c, T1b * T1g); } T18 = T14 + T17; T1n = T1h + T1m; T2Q = T18 - T1n; T2R = T24 + T25; T2S = T2a + T2b; T2T = T2R - T2S; { E T26, T27, T29, T2c; T26 = T24 - T25; T27 = T1h - T1m; T28 = T26 + T27; T2A = T26 - T27; T29 = T14 - T17; T2c = T2a - T2b; T2d = T29 - T2c; T2B = T29 + T2c; } } { E T23, T2r, T3A, T3C, T2q, T3B, T2u, T3x; { E T1R, T22, T3y, T3z; T1R = T1N - T1Q; T22 = KP707106781 * (T1W - T21); T23 = T1R + T22; T2r = T1R - T22; T3y = KP707106781 * (T2x - T2w); T3z = T3s + T3r; T3A = T3y + T3z; T3C = T3z - T3y; } { E T2e, T2p, T2s, T2t; T2e = FMA(KP923879532, T28, KP382683432 * T2d); T2p = FNMS(KP923879532, T2o, KP382683432 * T2j); T2q = T2e + T2p; T3B = T2p - T2e; T2s = FNMS(KP923879532, T2d, KP382683432 * T28); T2t = FMA(KP382683432, T2o, KP923879532 * T2j); T2u = T2s - T2t; T3x = T2s + T2t; } ri[WS(rs, 11)] = T23 - T2q; ii[WS(rs, 11)] = T3A - T3x; ri[WS(rs, 3)] = T23 + T2q; ii[WS(rs, 3)] = T3x + T3A; ri[WS(rs, 15)] = T2r - T2u; ii[WS(rs, 15)] = T3C - T3B; ri[WS(rs, 7)] = T2r + T2u; ii[WS(rs, 7)] = T3B + T3C; } { E T2P, T31, T3m, T3o, T30, T3n, T34, T3j; { E T2L, T2O, T3k, T3l; T2L = Tf - TA; T2O = T2M - T2N; T2P = T2L + T2O; T31 = T2L - T2O; T3k = TZ - TM; T3l = T3e - T3b; T3m = T3k + T3l; T3o = T3l - T3k; } { E T2U, T2Z, T32, T33; T2U = T2Q + T2T; T2Z = T2V - T2Y; T30 = KP707106781 * (T2U + T2Z); T3n = KP707106781 * (T2Z - T2U); T32 = T2T - T2Q; T33 = T2V + T2Y; T34 = KP707106781 * (T32 - T33); T3j = KP707106781 * (T32 + T33); } ri[WS(rs, 10)] = T2P - T30; ii[WS(rs, 10)] = T3m - T3j; ri[WS(rs, 2)] = T2P + T30; ii[WS(rs, 2)] = T3j + T3m; ri[WS(rs, 14)] = T31 - T34; ii[WS(rs, 14)] = T3o - T3n; ri[WS(rs, 6)] = T31 + T34; ii[WS(rs, 6)] = T3n + T3o; } { E T2z, T2H, T3u, T3w, T2G, T3v, T2K, T3p; { E T2v, T2y, T3q, T3t; T2v = T1N + T1Q; T2y = KP707106781 * (T2w + T2x); T2z = T2v + T2y; T2H = T2v - T2y; T3q = KP707106781 * (T1W + T21); T3t = T3r - T3s; T3u = T3q + T3t; T3w = T3t - T3q; } { E T2C, T2F, T2I, T2J; T2C = FMA(KP382683432, T2A, KP923879532 * T2B); T2F = FNMS(KP382683432, T2E, KP923879532 * T2D); T2G = T2C + T2F; T3v = T2F - T2C; T2I = FNMS(KP382683432, T2B, KP923879532 * T2A); T2J = FMA(KP923879532, T2E, KP382683432 * T2D); T2K = T2I - T2J; T3p = T2I + T2J; } ri[WS(rs, 9)] = T2z - T2G; ii[WS(rs, 9)] = T3u - T3p; ri[WS(rs, 1)] = T2z + T2G; ii[WS(rs, 1)] = T3p + T3u; ri[WS(rs, 13)] = T2H - T2K; ii[WS(rs, 13)] = T3w - T3v; ri[WS(rs, 5)] = T2H + T2K; ii[WS(rs, 5)] = T3v + T3w; } { E T11, T35, T3g, T3i, T1M, T3h, T38, T39; { E TB, T10, T3a, T3f; TB = Tf + TA; T10 = TM + TZ; T11 = TB + T10; T35 = TB - T10; T3a = T2M + T2N; T3f = T3b + T3e; T3g = T3a + T3f; T3i = T3f - T3a; } { E T1o, T1L, T36, T37; T1o = T18 + T1n; T1L = T1B + T1K; T1M = T1o + T1L; T3h = T1L - T1o; T36 = T2R + T2S; T37 = T2W + T2X; T38 = T36 - T37; T39 = T36 + T37; } ri[WS(rs, 8)] = T11 - T1M; ii[WS(rs, 8)] = T3g - T39; ri[0] = T11 + T1M; ii[0] = T39 + T3g; ri[WS(rs, 12)] = T35 - T38; ii[WS(rs, 12)] = T3i - T3h; ri[WS(rs, 4)] = T35 + T38; ii[WS(rs, 4)] = T3h + T3i; } } } } } static const tw_instr twinstr[] = { {TW_CEXP, 0, 1}, {TW_CEXP, 0, 3}, {TW_CEXP, 0, 9}, {TW_CEXP, 0, 15}, {TW_NEXT, 1, 0} }; static const ct_desc desc = { 16, "t2_16", twinstr, &GENUS, {156, 68, 40, 0}, 0, 0, 0 }; void X(codelet_t2_16) (planner *p) { X(kdft_dit_register) (p, t2_16, &desc); } #endif /* HAVE_FMA */ fftw-3.3.3/dft/scalar/codelets/t2_32.c0000644000175000017500000015054112054410271014232 00000000000000/* * Copyright (c) 2003, 2007-11 Matteo Frigo * Copyright (c) 2003, 2007-11 Massachusetts Institute of Technology * * This program is free software; you can redistribute it and/or modify * it under the terms of the GNU General Public License as published by * the Free Software Foundation; either version 2 of the License, or * (at your option) any later version. * * This program is distributed in the hope that it will be useful, * but WITHOUT ANY WARRANTY; without even the implied warranty of * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the * GNU General Public License for more details. * * You should have received a copy of the GNU General Public License * along with this program; if not, write to the Free Software * Foundation, Inc., 51 Franklin Street, Fifth Floor, Boston, MA 02110-1301 USA * */ /* This file was automatically generated --- DO NOT EDIT */ /* Generated on Sun Nov 25 07:36:00 EST 2012 */ #include "codelet-dft.h" #ifdef HAVE_FMA /* Generated by: ../../../genfft/gen_twiddle.native -fma -reorder-insns -schedule-for-pipeline -compact -variables 4 -pipeline-latency 4 -twiddle-log3 -precompute-twiddles -n 32 -name t2_32 -include t.h */ /* * This function contains 488 FP additions, 350 FP multiplications, * (or, 236 additions, 98 multiplications, 252 fused multiply/add), * 181 stack variables, 7 constants, and 128 memory accesses */ #include "t.h" static void t2_32(R *ri, R *ii, const R *W, stride rs, INT mb, INT me, INT ms) { DK(KP980785280, +0.980785280403230449126182236134239036973933731); DK(KP831469612, +0.831469612302545237078788377617905756738560812); DK(KP668178637, +0.668178637919298919997757686523080761552472251); DK(KP198912367, +0.198912367379658006911597622644676228597850501); DK(KP923879532, +0.923879532511286756128183189396788286822416626); DK(KP414213562, +0.414213562373095048801688724209698078569671875); DK(KP707106781, +0.707106781186547524400844362104849039284835938); { INT m; for (m = mb, W = W + (mb * 8); m < me; m = m + 1, ri = ri + ms, ii = ii + ms, W = W + 8, MAKE_VOLATILE_STRIDE(64, rs)) { E T9A, T9z; { E T2, T8, T3, T6, Te, Tr, T18, T4, Ta, Tz, T1n, T10, Ti, T5, Tc; T2 = W[0]; T8 = W[4]; T3 = W[2]; T6 = W[3]; Te = W[6]; Tr = T2 * T8; T18 = T3 * T8; T4 = T2 * T3; Ta = T2 * T6; Tz = T3 * Te; T1n = T8 * Te; T10 = T2 * Te; Ti = W[7]; T5 = W[1]; Tc = W[5]; { E T34, T31, T2X, T2T, Tq, T46, T8H, T97, TH, T98, T4b, T8D, TZ, T7f, T4j; E T6t, T1g, T7g, T4q, T6u, T4z, T6x, T1J, T7m, T7l, T8d, T6y, T4G, T2k, T7o; E T7r, T8e, T6A, T4O, T6B, T4V, T6P, T5E, T7L, T3G, T6M, T61, T8n, T7I, T6I; E T55, T7A, T2N, T6F, T5s, T8i, T7x, T5L, T62, T43, T7J, T5S, T63, T7O, T8o; E T2U, T2R, T2V, T57, T3a, T5h, T2Y, T32, T35; { E T1K, T23, T1N, T26, T2b, T1U, T3C, T3j, T3z, T3f, T1R, T29, TR, Th, T2J; E T2F, Td, TP, T3r, T3n, T2w, T2s, T3Q, T3M, T1Z, T1V, T2g, T2c; { E T11, T1C, TM, Tb, TJ, T7, T1o, T19, T1w, T1F, T15, T1s, T1d, T1z, TW; E TS, Ty, T48, TG, T4a; { E T1, TA, Ts, TE, Tw, Tn, Tj, T8G, Tk, To, T14; T1 = ri[0]; TA = FMA(T6, Ti, Tz); T1K = FNMS(T6, Ti, Tz); T14 = T2 * Ti; { E T1r, TD, T1c, Tv; T1r = T8 * Ti; TD = T3 * Ti; T11 = FNMS(T5, Ti, T10); T1C = FMA(T5, Ti, T10); TM = FMA(T5, T3, Ta); Tb = FNMS(T5, T3, Ta); TJ = FNMS(T5, T6, T4); T7 = FMA(T5, T6, T4); T1o = FMA(Tc, Ti, T1n); T23 = FMA(T6, Tc, T18); T19 = FNMS(T6, Tc, T18); T1w = FNMS(T5, Tc, Tr); Ts = FMA(T5, Tc, Tr); T1c = T3 * Tc; Tv = T2 * Tc; T1F = FNMS(T5, Te, T14); T15 = FMA(T5, Te, T14); T1s = FNMS(Tc, Te, T1r); T1N = FMA(T6, Te, TD); TE = FNMS(T6, Te, TD); { E T1T, T3i, T3e, T1Q; T1T = TJ * Tc; T3i = TJ * Ti; T3e = TJ * Te; T1Q = TJ * T8; { E Tg, T2I, T2E, T9; Tg = T7 * Tc; T2I = T7 * Ti; T2E = T7 * Te; T9 = T7 * T8; { E T3q, T3m, T2v, T2r; T3q = T19 * Ti; T3m = T19 * Te; T2v = T1w * Ti; T2r = T1w * Te; { E T2W, T2S, T3P, T3L; T2W = T23 * Ti; T2S = T23 * Te; T3P = Ts * Ti; T3L = Ts * Te; T26 = FNMS(T6, T8, T1c); T1d = FMA(T6, T8, T1c); T1z = FMA(T5, T8, Tv); Tw = FNMS(T5, T8, Tv); T2b = FNMS(TM, T8, T1T); T1U = FMA(TM, T8, T1T); T3C = FNMS(TM, Te, T3i); T3j = FMA(TM, Te, T3i); T3z = FMA(TM, Ti, T3e); T3f = FNMS(TM, Ti, T3e); T1R = FNMS(TM, Tc, T1Q); T29 = FMA(TM, Tc, T1Q); TR = FNMS(Tb, T8, Tg); Th = FMA(Tb, T8, Tg); T34 = FMA(Tb, Te, T2I); T2J = FNMS(Tb, Te, T2I); T31 = FNMS(Tb, Ti, T2E); T2F = FMA(Tb, Ti, T2E); Td = FNMS(Tb, Tc, T9); TP = FMA(Tb, Tc, T9); T2X = FNMS(T26, Te, T2W); T2T = FMA(T26, Ti, T2S); T3r = FNMS(T1d, Te, T3q); T3n = FMA(T1d, Ti, T3m); T2w = FNMS(T1z, Te, T2v); T2s = FMA(T1z, Ti, T2r); T3Q = FNMS(Tw, Te, T3P); T3M = FMA(Tw, Ti, T3L); { E T1Y, T1S, T2f, T2a; T1Y = T1R * Ti; T1S = T1R * Te; T2f = T29 * Ti; T2a = T29 * Te; { E Tm, Tf, TV, TQ; Tm = Td * Ti; Tf = Td * Te; TV = TP * Ti; TQ = TP * Te; T1Z = FNMS(T1U, Te, T1Y); T1V = FMA(T1U, Ti, T1S); T2g = FNMS(T2b, Te, T2f); T2c = FMA(T2b, Ti, T2a); Tn = FNMS(Th, Te, Tm); Tj = FMA(Th, Ti, Tf); TW = FNMS(TR, Te, TV); TS = FMA(TR, Ti, TQ); T8G = ii[0]; } } } } } } } Tk = ri[WS(rs, 16)]; To = ii[WS(rs, 16)]; { E Tt, Tx, Tu, T47, TB, TF, TC, T49; { E Tl, T8E, Tp, T8F; Tt = ri[WS(rs, 8)]; Tx = ii[WS(rs, 8)]; Tl = Tj * Tk; T8E = Tj * To; Tu = Ts * Tt; T47 = Ts * Tx; Tp = FMA(Tn, To, Tl); T8F = FNMS(Tn, Tk, T8E); TB = ri[WS(rs, 24)]; TF = ii[WS(rs, 24)]; Tq = T1 + Tp; T46 = T1 - Tp; T8H = T8F + T8G; T97 = T8G - T8F; TC = TA * TB; T49 = TA * TF; } Ty = FMA(Tw, Tx, Tu); T48 = FNMS(Tw, Tt, T47); TG = FMA(TE, TF, TC); T4a = FNMS(TE, TB, T49); } } { E TT, TX, TO, T4f, TU, T4g; { E TK, TN, TL, T4e; TK = ri[WS(rs, 4)]; TN = ii[WS(rs, 4)]; TH = Ty + TG; T98 = Ty - TG; T4b = T48 - T4a; T8D = T48 + T4a; TL = TJ * TK; T4e = TJ * TN; TT = ri[WS(rs, 20)]; TX = ii[WS(rs, 20)]; TO = FMA(TM, TN, TL); T4f = FNMS(TM, TK, T4e); TU = TS * TT; T4g = TS * TX; } { E T17, T4m, T1a, T1e, T4d, T4i; { E T12, T16, TY, T4h, T13, T4l; T12 = ri[WS(rs, 28)]; T16 = ii[WS(rs, 28)]; TY = FMA(TW, TX, TU); T4h = FNMS(TW, TT, T4g); T13 = T11 * T12; T4l = T11 * T16; TZ = TO + TY; T4d = TO - TY; T7f = T4f + T4h; T4i = T4f - T4h; T17 = FMA(T15, T16, T13); T4m = FNMS(T15, T12, T4l); } T4j = T4d + T4i; T6t = T4i - T4d; T1a = ri[WS(rs, 12)]; T1e = ii[WS(rs, 12)]; { E T1m, T4u, T1H, T4E, T1x, T1A, T1u, T4w, T1y, T4B; { E T1D, T1G, T1E, T4D; { E T1f, T4o, T4k, T4p; { E T1j, T1l, T1b, T4n, T1k, T4t; T1j = ri[WS(rs, 2)]; T1l = ii[WS(rs, 2)]; T1b = T19 * T1a; T4n = T19 * T1e; T1k = T7 * T1j; T4t = T7 * T1l; T1f = FMA(T1d, T1e, T1b); T4o = FNMS(T1d, T1a, T4n); T1m = FMA(Tb, T1l, T1k); T4u = FNMS(Tb, T1j, T4t); } T1g = T17 + T1f; T4k = T17 - T1f; T7g = T4m + T4o; T4p = T4m - T4o; T1D = ri[WS(rs, 26)]; T1G = ii[WS(rs, 26)]; T4q = T4k - T4p; T6u = T4k + T4p; T1E = T1C * T1D; T4D = T1C * T1G; } { E T1p, T1t, T1q, T4v; T1p = ri[WS(rs, 18)]; T1t = ii[WS(rs, 18)]; T1H = FMA(T1F, T1G, T1E); T4E = FNMS(T1F, T1D, T4D); T1q = T1o * T1p; T4v = T1o * T1t; T1x = ri[WS(rs, 10)]; T1A = ii[WS(rs, 10)]; T1u = FMA(T1s, T1t, T1q); T4w = FNMS(T1s, T1p, T4v); T1y = T1w * T1x; T4B = T1w * T1A; } } { E T4A, T1v, T7j, T4x, T1B, T4C; T4A = T1m - T1u; T1v = T1m + T1u; T7j = T4u + T4w; T4x = T4u - T4w; T1B = FMA(T1z, T1A, T1y); T4C = FNMS(T1z, T1x, T4B); { E T1I, T4y, T4F, T7k; T1I = T1B + T1H; T4y = T1B - T1H; T4F = T4C - T4E; T7k = T4C + T4E; T4z = T4x - T4y; T6x = T4x + T4y; T1J = T1v + T1I; T7m = T1v - T1I; T7l = T7j - T7k; T8d = T7j + T7k; T6y = T4A - T4F; T4G = T4A + T4F; } } } } } } { E T5Z, T3u, T5V, T5C, T7G, T5D, T3F, T5X, T4P, T4U; { E T1P, T4J, T2i, T4T, T21, T4L, T28, T4R; { E T1L, T1O, T1W, T20; T1L = ri[WS(rs, 30)]; T1O = ii[WS(rs, 30)]; { E T2d, T2h, T1M, T4I, T2e, T4S; T2d = ri[WS(rs, 22)]; T2h = ii[WS(rs, 22)]; T1M = T1K * T1L; T4I = T1K * T1O; T2e = T2c * T2d; T4S = T2c * T2h; T1P = FMA(T1N, T1O, T1M); T4J = FNMS(T1N, T1L, T4I); T2i = FMA(T2g, T2h, T2e); T4T = FNMS(T2g, T2d, T4S); } T1W = ri[WS(rs, 14)]; T20 = ii[WS(rs, 14)]; { E T24, T27, T1X, T4K, T25, T4Q; T24 = ri[WS(rs, 6)]; T27 = ii[WS(rs, 6)]; T1X = T1V * T1W; T4K = T1V * T20; T25 = T23 * T24; T4Q = T23 * T27; T21 = FMA(T1Z, T20, T1X); T4L = FNMS(T1Z, T1W, T4K); T28 = FMA(T26, T27, T25); T4R = FNMS(T26, T24, T4Q); } } { E T22, T7p, T4M, T4N, T2j, T7q; T4P = T1P - T21; T22 = T1P + T21; T7p = T4J + T4L; T4M = T4J - T4L; T4N = T28 - T2i; T2j = T28 + T2i; T7q = T4R + T4T; T4U = T4R - T4T; T2k = T22 + T2j; T7o = T22 - T2j; T7r = T7p - T7q; T8e = T7p + T7q; T6A = T4M + T4N; T4O = T4M - T4N; } } { E T3l, T5z, T3E, T3v, T3t, T3w, T3x, T5B, T3A, T3B, T3D, T3y, T5W; { E T3g, T3k, T3h, T5y; T3g = ri[WS(rs, 31)]; T3k = ii[WS(rs, 31)]; T3A = ri[WS(rs, 23)]; T6B = T4P - T4U; T4V = T4P + T4U; T3h = T3f * T3g; T5y = T3f * T3k; T3B = T3z * T3A; T3D = ii[WS(rs, 23)]; T3l = FMA(T3j, T3k, T3h); T5z = FNMS(T3j, T3g, T5y); } { E T3o, T5Y, T3s, T3p, T5A; T3o = ri[WS(rs, 15)]; T3E = FMA(T3C, T3D, T3B); T5Y = T3z * T3D; T3s = ii[WS(rs, 15)]; T3p = T3n * T3o; T3v = ri[WS(rs, 7)]; T5Z = FNMS(T3C, T3A, T5Y); T5A = T3n * T3s; T3t = FMA(T3r, T3s, T3p); T3w = TP * T3v; T3x = ii[WS(rs, 7)]; T5B = FNMS(T3r, T3o, T5A); } T3u = T3l + T3t; T5V = T3l - T3t; T3y = FMA(TR, T3x, T3w); T5W = TP * T3x; T5C = T5z - T5B; T7G = T5z + T5B; T5D = T3y - T3E; T3F = T3y + T3E; T5X = FNMS(TR, T3v, T5W); } { E T2L, T5q, T5m, T2z, T7v, T53, T2D, T5o; { E T2q, T50, T2y, T2A, T2C, T52, T2B, T5n; { E T2G, T2K, T2n, T4Z, T2t, T51; { E T2o, T2p, T60, T7H; T2n = ri[WS(rs, 1)]; T6P = T5C + T5D; T5E = T5C - T5D; T7L = T3u - T3F; T3G = T3u + T3F; T60 = T5X - T5Z; T7H = T5X + T5Z; T2o = T2 * T2n; T2p = ii[WS(rs, 1)]; T6M = T5V - T60; T61 = T5V + T60; T8n = T7G + T7H; T7I = T7G - T7H; T4Z = T2 * T2p; T2q = FMA(T5, T2p, T2o); } T2G = ri[WS(rs, 25)]; T2K = ii[WS(rs, 25)]; T50 = FNMS(T5, T2n, T4Z); { E T2x, T2u, T2H, T5p; T2t = ri[WS(rs, 17)]; T2H = T2F * T2G; T5p = T2F * T2K; T2x = ii[WS(rs, 17)]; T2u = T2s * T2t; T2L = FMA(T2J, T2K, T2H); T5q = FNMS(T2J, T2G, T5p); T51 = T2s * T2x; T2y = FMA(T2w, T2x, T2u); } T2A = ri[WS(rs, 9)]; T2C = ii[WS(rs, 9)]; T52 = FNMS(T2w, T2t, T51); } T5m = T2q - T2y; T2z = T2q + T2y; T2B = T8 * T2A; T5n = T8 * T2C; T7v = T50 + T52; T53 = T50 - T52; T2D = FMA(Tc, T2C, T2B); T5o = FNMS(Tc, T2A, T5n); } { E T3N, T3K, T3O, T5G, T41, T5Q, T3R, T3U, T3W; { E T3H, T3I, T3J, T3Y, T40, T5F, T3Z, T5P; T3H = ri[WS(rs, 3)]; { E T54, T2M, T5r, T7w; T54 = T2D - T2L; T2M = T2D + T2L; T5r = T5o - T5q; T7w = T5o + T5q; T6I = T53 + T54; T55 = T53 - T54; T7A = T2z - T2M; T2N = T2z + T2M; T6F = T5m - T5r; T5s = T5m + T5r; T8i = T7v + T7w; T7x = T7v - T7w; T3I = T3 * T3H; } T3J = ii[WS(rs, 3)]; T3Y = ri[WS(rs, 11)]; T40 = ii[WS(rs, 11)]; T3N = ri[WS(rs, 19)]; T3K = FMA(T6, T3J, T3I); T5F = T3 * T3J; T3Z = Td * T3Y; T5P = Td * T40; T3O = T3M * T3N; T5G = FNMS(T6, T3H, T5F); T41 = FMA(Th, T40, T3Z); T5Q = FNMS(Th, T3Y, T5P); T3R = ii[WS(rs, 19)]; T3U = ri[WS(rs, 27)]; T3W = ii[WS(rs, 27)]; } { E T2O, T2P, T2Q, T37, T39, T56, T38, T5g; { E T3T, T5K, T5I, T3X, T5O, T7M, T5J; T2O = ri[WS(rs, 5)]; { E T3S, T5H, T3V, T5N; T3S = FMA(T3Q, T3R, T3O); T5H = T3M * T3R; T3V = Te * T3U; T5N = Te * T3W; T3T = T3K + T3S; T5K = T3K - T3S; T5I = FNMS(T3Q, T3N, T5H); T3X = FMA(Ti, T3W, T3V); T5O = FNMS(Ti, T3U, T5N); T2P = T29 * T2O; } T7M = T5G + T5I; T5J = T5G - T5I; { E T42, T5M, T7N, T5R; T42 = T3X + T41; T5M = T3X - T41; T7N = T5O + T5Q; T5R = T5O - T5Q; T5L = T5J - T5K; T62 = T5K + T5J; T43 = T3T + T42; T7J = T42 - T3T; T5S = T5M + T5R; T63 = T5M - T5R; T7O = T7M - T7N; T8o = T7M + T7N; T2Q = ii[WS(rs, 5)]; } } T37 = ri[WS(rs, 13)]; T39 = ii[WS(rs, 13)]; T2U = ri[WS(rs, 21)]; T2R = FMA(T2b, T2Q, T2P); T56 = T29 * T2Q; T38 = T1R * T37; T5g = T1R * T39; T2V = T2T * T2U; T57 = FNMS(T2b, T2O, T56); T3a = FMA(T1U, T39, T38); T5h = FNMS(T1U, T37, T5g); T2Y = ii[WS(rs, 21)]; T32 = ri[WS(rs, 29)]; T35 = ii[WS(rs, 29)]; } } } } } { E T5c, T5t, T5j, T5u, T88, T90, T8Z, T8b; { E T7e, T8T, T7y, T7D, T7h, T8U, T8S, T8R; { E T8c, T1i, T8A, T8z, T8O, T8J, T8N, T2l, T8L, T45, T8t, T8l, T8u, T8q, T3c; E T8k, T8p, T8w, T2m; { E T8x, T8y, T8j, T8C, T8I; { E TI, T30, T5b, T59, T36, T5f, T1h, T7B, T5a; TI = Tq + TH; T7e = Tq - TH; { E T2Z, T58, T33, T5e; T2Z = FMA(T2X, T2Y, T2V); T58 = T2T * T2Y; T33 = T31 * T32; T5e = T31 * T35; T30 = T2R + T2Z; T5b = T2R - T2Z; T59 = FNMS(T2X, T2U, T58); T36 = FMA(T34, T35, T33); T5f = FNMS(T34, T32, T5e); T1h = TZ + T1g; T8T = T1g - TZ; } T7B = T57 + T59; T5a = T57 - T59; { E T3b, T5d, T7C, T5i; T3b = T36 + T3a; T5d = T36 - T3a; T7C = T5f + T5h; T5i = T5f - T5h; T5c = T5a - T5b; T5t = T5b + T5a; T3c = T30 + T3b; T7y = T3b - T30; T5j = T5d + T5i; T5u = T5d - T5i; T7D = T7B - T7C; T8j = T7B + T7C; T8c = TI - T1h; T1i = TI + T1h; } } T8k = T8i - T8j; T8x = T8i + T8j; T8y = T8n + T8o; T8p = T8n - T8o; T7h = T7f - T7g; T8C = T7f + T7g; T8I = T8D + T8H; T8U = T8H - T8D; T8A = T8x + T8y; T8z = T8x - T8y; T8O = T8I - T8C; T8J = T8C + T8I; } { E T8h, T8m, T3d, T44; T8h = T2N - T3c; T3d = T2N + T3c; T44 = T3G + T43; T8m = T3G - T43; T8N = T2k - T1J; T2l = T1J + T2k; T8L = T44 - T3d; T45 = T3d + T44; T8t = T8k - T8h; T8l = T8h + T8k; T8u = T8m + T8p; T8q = T8m - T8p; } T8w = T1i - T2l; T2m = T1i + T2l; { E T8s, T8P, T8Q, T8v; { E T8r, T8M, T8K, T8g, T8B, T8f; T8S = T8q - T8l; T8r = T8l + T8q; T8B = T8d + T8e; T8f = T8d - T8e; ri[0] = T2m + T45; ri[WS(rs, 16)] = T2m - T45; ri[WS(rs, 8)] = T8w + T8z; ri[WS(rs, 24)] = T8w - T8z; T8M = T8J - T8B; T8K = T8B + T8J; T8g = T8c + T8f; T8s = T8c - T8f; T8R = T8O - T8N; T8P = T8N + T8O; ii[WS(rs, 24)] = T8M - T8L; ii[WS(rs, 8)] = T8L + T8M; ii[WS(rs, 16)] = T8K - T8A; ii[0] = T8A + T8K; ri[WS(rs, 4)] = FMA(KP707106781, T8r, T8g); ri[WS(rs, 20)] = FNMS(KP707106781, T8r, T8g); T8Q = T8t + T8u; T8v = T8t - T8u; } ii[WS(rs, 20)] = FNMS(KP707106781, T8Q, T8P); ii[WS(rs, 4)] = FMA(KP707106781, T8Q, T8P); ri[WS(rs, 12)] = FMA(KP707106781, T8v, T8s); ri[WS(rs, 28)] = FNMS(KP707106781, T8v, T8s); } } { E T7P, T7W, T7i, T7K, T8a, T86, T91, T8V, T8W, T7t, T7T, T7F, T92, T7Z, T89; E T83; { E T7X, T7n, T7s, T7Y, T84, T85; T7P = T7L - T7O; T84 = T7L + T7O; ii[WS(rs, 28)] = FNMS(KP707106781, T8S, T8R); ii[WS(rs, 12)] = FMA(KP707106781, T8S, T8R); T7W = T7e + T7h; T7i = T7e - T7h; T85 = T7I + T7J; T7K = T7I - T7J; T7X = T7m + T7l; T7n = T7l - T7m; T8a = FMA(KP414213562, T84, T85); T86 = FNMS(KP414213562, T85, T84); T91 = T8U - T8T; T8V = T8T + T8U; T7s = T7o + T7r; T7Y = T7o - T7r; { E T82, T81, T7z, T7E; T82 = T7x + T7y; T7z = T7x - T7y; T7E = T7A - T7D; T81 = T7A + T7D; T8W = T7n + T7s; T7t = T7n - T7s; T7T = FNMS(KP414213562, T7z, T7E); T7F = FMA(KP414213562, T7E, T7z); T92 = T7Y - T7X; T7Z = T7X + T7Y; T89 = FNMS(KP414213562, T81, T82); T83 = FMA(KP414213562, T82, T81); } } { E T7S, T7u, T93, T95, T7U, T7Q; T7S = FNMS(KP707106781, T7t, T7i); T7u = FMA(KP707106781, T7t, T7i); T93 = FMA(KP707106781, T92, T91); T95 = FNMS(KP707106781, T92, T91); T7U = FMA(KP414213562, T7K, T7P); T7Q = FNMS(KP414213562, T7P, T7K); { E T80, T87, T8X, T8Y; T88 = FNMS(KP707106781, T7Z, T7W); T80 = FMA(KP707106781, T7Z, T7W); { E T7V, T94, T96, T7R; T7V = T7T + T7U; T94 = T7U - T7T; T96 = T7F + T7Q; T7R = T7F - T7Q; ri[WS(rs, 30)] = FMA(KP923879532, T7V, T7S); ri[WS(rs, 14)] = FNMS(KP923879532, T7V, T7S); ii[WS(rs, 22)] = FNMS(KP923879532, T94, T93); ii[WS(rs, 6)] = FMA(KP923879532, T94, T93); ii[WS(rs, 30)] = FMA(KP923879532, T96, T95); ii[WS(rs, 14)] = FNMS(KP923879532, T96, T95); ri[WS(rs, 6)] = FMA(KP923879532, T7R, T7u); ri[WS(rs, 22)] = FNMS(KP923879532, T7R, T7u); T87 = T83 + T86; T90 = T86 - T83; } T8Z = FNMS(KP707106781, T8W, T8V); T8X = FMA(KP707106781, T8W, T8V); T8Y = T89 + T8a; T8b = T89 - T8a; ri[WS(rs, 2)] = FMA(KP923879532, T87, T80); ri[WS(rs, 18)] = FNMS(KP923879532, T87, T80); ii[WS(rs, 18)] = FNMS(KP923879532, T8Y, T8X); ii[WS(rs, 2)] = FMA(KP923879532, T8Y, T8X); } } } } { E T6s, T9o, T9n, T6v, T6N, T6Q, T6G, T6J, T9g, T9f; { E T6c, T4s, T9c, T4X, T9h, T9b, T9i, T6f, T5U, T6l, T64, T5k, T5v; { E T6d, T6e, T99, T9a, T5T; { E T4c, T4r, T4H, T4W; T6s = T46 - T4b; T4c = T46 + T4b; ri[WS(rs, 10)] = FMA(KP923879532, T8b, T88); ri[WS(rs, 26)] = FNMS(KP923879532, T8b, T88); ii[WS(rs, 26)] = FNMS(KP923879532, T90, T8Z); ii[WS(rs, 10)] = FMA(KP923879532, T90, T8Z); T4r = T4j + T4q; T9o = T4q - T4j; T6d = FMA(KP414213562, T4z, T4G); T4H = FNMS(KP414213562, T4G, T4z); T4W = FMA(KP414213562, T4V, T4O); T6e = FNMS(KP414213562, T4O, T4V); T9n = T98 + T97; T99 = T97 - T98; T6c = FMA(KP707106781, T4r, T4c); T4s = FNMS(KP707106781, T4r, T4c); T9c = T4H + T4W; T4X = T4H - T4W; T9a = T6t + T6u; T6v = T6t - T6u; } T6N = T5S - T5L; T5T = T5L + T5S; T9h = FNMS(KP707106781, T9a, T99); T9b = FMA(KP707106781, T9a, T99); T9i = T6e - T6d; T6f = T6d + T6e; T5U = FNMS(KP707106781, T5T, T5E); T6l = FMA(KP707106781, T5T, T5E); T64 = T62 + T63; T6Q = T62 - T63; T6G = T5j - T5c; T5k = T5c + T5j; T5v = T5t + T5u; T6J = T5t - T5u; } { E T6m, T6q, T6j, T6p, T9l, T9m; { E T68, T4Y, T6a, T66, T69, T5x, T9j, T6k, T65, T9k, T6b, T67; T68 = FNMS(KP923879532, T4X, T4s); T4Y = FMA(KP923879532, T4X, T4s); T6k = FMA(KP707106781, T64, T61); T65 = FNMS(KP707106781, T64, T61); { E T6i, T5l, T6h, T5w; T6i = FMA(KP707106781, T5k, T55); T5l = FNMS(KP707106781, T5k, T55); T6h = FMA(KP707106781, T5v, T5s); T5w = FNMS(KP707106781, T5v, T5s); T6m = FNMS(KP198912367, T6l, T6k); T6q = FMA(KP198912367, T6k, T6l); T6a = FMA(KP668178637, T5U, T65); T66 = FNMS(KP668178637, T65, T5U); T6j = FMA(KP198912367, T6i, T6h); T6p = FNMS(KP198912367, T6h, T6i); T69 = FNMS(KP668178637, T5l, T5w); T5x = FMA(KP668178637, T5w, T5l); } T9j = FMA(KP923879532, T9i, T9h); T9l = FNMS(KP923879532, T9i, T9h); T9k = T6a - T69; T6b = T69 + T6a; T9m = T5x + T66; T67 = T5x - T66; ii[WS(rs, 21)] = FNMS(KP831469612, T9k, T9j); ii[WS(rs, 5)] = FMA(KP831469612, T9k, T9j); ri[WS(rs, 5)] = FMA(KP831469612, T67, T4Y); ri[WS(rs, 21)] = FNMS(KP831469612, T67, T4Y); ri[WS(rs, 29)] = FMA(KP831469612, T6b, T68); ri[WS(rs, 13)] = FNMS(KP831469612, T6b, T68); } { E T6o, T9d, T9e, T6r, T6g, T6n; T6o = FNMS(KP923879532, T6f, T6c); T6g = FMA(KP923879532, T6f, T6c); T6n = T6j + T6m; T9g = T6m - T6j; T9f = FNMS(KP923879532, T9c, T9b); T9d = FMA(KP923879532, T9c, T9b); ii[WS(rs, 29)] = FMA(KP831469612, T9m, T9l); ii[WS(rs, 13)] = FNMS(KP831469612, T9m, T9l); ri[WS(rs, 1)] = FMA(KP980785280, T6n, T6g); ri[WS(rs, 17)] = FNMS(KP980785280, T6n, T6g); T9e = T6p + T6q; T6r = T6p - T6q; ii[WS(rs, 17)] = FNMS(KP980785280, T9e, T9d); ii[WS(rs, 1)] = FMA(KP980785280, T9e, T9d); ri[WS(rs, 9)] = FMA(KP980785280, T6r, T6o); ri[WS(rs, 25)] = FNMS(KP980785280, T6r, T6o); } } } { E T6Y, T6w, T9w, T6D, T9v, T9p, T9q, T71, T6H, T74, T78, T7c, T6W, T6S; { E T6Z, T6z, T6C, T70; T6Z = FNMS(KP414213562, T6x, T6y); T6z = FMA(KP414213562, T6y, T6x); ii[WS(rs, 25)] = FNMS(KP980785280, T9g, T9f); ii[WS(rs, 9)] = FMA(KP980785280, T9g, T9f); T6Y = FNMS(KP707106781, T6v, T6s); T6w = FMA(KP707106781, T6v, T6s); T6C = FNMS(KP414213562, T6B, T6A); T70 = FMA(KP414213562, T6A, T6B); T9w = T6z + T6C; T6D = T6z - T6C; T9v = FNMS(KP707106781, T9o, T9n); T9p = FMA(KP707106781, T9o, T9n); { E T77, T6O, T76, T6R; T9q = T70 - T6Z; T71 = T6Z + T70; T77 = FMA(KP707106781, T6N, T6M); T6O = FNMS(KP707106781, T6N, T6M); T76 = FMA(KP707106781, T6Q, T6P); T6R = FNMS(KP707106781, T6Q, T6P); T6H = FNMS(KP707106781, T6G, T6F); T74 = FMA(KP707106781, T6G, T6F); T78 = FNMS(KP198912367, T77, T76); T7c = FMA(KP198912367, T76, T77); T6W = FMA(KP668178637, T6O, T6R); T6S = FNMS(KP668178637, T6R, T6O); } } { E T6U, T6E, T9r, T9t, T73, T6K; T6U = FNMS(KP923879532, T6D, T6w); T6E = FMA(KP923879532, T6D, T6w); T9r = FMA(KP923879532, T9q, T9p); T9t = FNMS(KP923879532, T9q, T9p); T73 = FMA(KP707106781, T6J, T6I); T6K = FNMS(KP707106781, T6J, T6I); { E T7a, T9x, T9y, T7d; { E T72, T7b, T6V, T6L, T79, T75; T7a = FMA(KP923879532, T71, T6Y); T72 = FNMS(KP923879532, T71, T6Y); T75 = FMA(KP198912367, T74, T73); T7b = FNMS(KP198912367, T73, T74); T6V = FNMS(KP668178637, T6H, T6K); T6L = FMA(KP668178637, T6K, T6H); T79 = T75 - T78; T9A = T75 + T78; T9z = FMA(KP923879532, T9w, T9v); T9x = FNMS(KP923879532, T9w, T9v); { E T6X, T9s, T9u, T6T; T6X = T6V - T6W; T9s = T6V + T6W; T9u = T6S - T6L; T6T = T6L + T6S; ri[WS(rs, 7)] = FMA(KP980785280, T79, T72); ri[WS(rs, 23)] = FNMS(KP980785280, T79, T72); ri[WS(rs, 11)] = FMA(KP831469612, T6X, T6U); ri[WS(rs, 27)] = FNMS(KP831469612, T6X, T6U); ii[WS(rs, 19)] = FNMS(KP831469612, T9s, T9r); ii[WS(rs, 3)] = FMA(KP831469612, T9s, T9r); ii[WS(rs, 27)] = FNMS(KP831469612, T9u, T9t); ii[WS(rs, 11)] = FMA(KP831469612, T9u, T9t); ri[WS(rs, 3)] = FMA(KP831469612, T6T, T6E); ri[WS(rs, 19)] = FNMS(KP831469612, T6T, T6E); T9y = T7c - T7b; T7d = T7b + T7c; } } ii[WS(rs, 23)] = FNMS(KP980785280, T9y, T9x); ii[WS(rs, 7)] = FMA(KP980785280, T9y, T9x); ri[WS(rs, 31)] = FMA(KP980785280, T7d, T7a); ri[WS(rs, 15)] = FNMS(KP980785280, T7d, T7a); } } } } } } } ii[WS(rs, 31)] = FMA(KP980785280, T9A, T9z); ii[WS(rs, 15)] = FNMS(KP980785280, T9A, T9z); } } } static const tw_instr twinstr[] = { {TW_CEXP, 0, 1}, {TW_CEXP, 0, 3}, {TW_CEXP, 0, 9}, {TW_CEXP, 0, 27}, {TW_NEXT, 1, 0} }; static const ct_desc desc = { 32, "t2_32", twinstr, &GENUS, {236, 98, 252, 0}, 0, 0, 0 }; void X(codelet_t2_32) (planner *p) { X(kdft_dit_register) (p, t2_32, &desc); } #else /* HAVE_FMA */ /* Generated by: ../../../genfft/gen_twiddle.native -compact -variables 4 -pipeline-latency 4 -twiddle-log3 -precompute-twiddles -n 32 -name t2_32 -include t.h */ /* * This function contains 488 FP additions, 280 FP multiplications, * (or, 376 additions, 168 multiplications, 112 fused multiply/add), * 158 stack variables, 7 constants, and 128 memory accesses */ #include "t.h" static void t2_32(R *ri, R *ii, const R *W, stride rs, INT mb, INT me, INT ms) { DK(KP195090322, +0.195090322016128267848284868477022240927691618); DK(KP980785280, +0.980785280403230449126182236134239036973933731); DK(KP555570233, +0.555570233019602224742830813948532874374937191); DK(KP831469612, +0.831469612302545237078788377617905756738560812); DK(KP382683432, +0.382683432365089771728459984030398866761344562); DK(KP923879532, +0.923879532511286756128183189396788286822416626); DK(KP707106781, +0.707106781186547524400844362104849039284835938); { INT m; for (m = mb, W = W + (mb * 8); m < me; m = m + 1, ri = ri + ms, ii = ii + ms, W = W + 8, MAKE_VOLATILE_STRIDE(64, rs)) { E T2, T5, T3, T6, T8, TM, TO, Td, T9, Te, Th, Tl, TD, TH, T1y; E T1H, T15, T1A, T11, T1F, T1n, T1p, T2q, T2I, T2u, T2K, T2V, T3b, T2Z, T3d; E Tu, Ty, T3l, T3n, T1t, T1v, T2f, T2h, T1a, T1e, T32, T34, T1W, T1Y, T2C; E T2E, Tg, TR, Tk, TS, Tm, TV, To, TT, T1M, T21, T1P, T22, T1Q, T25; E T1S, T23; { E Ts, T1d, Tx, T18, Tt, T1c, Tw, T19, TB, T14, TG, TZ, TC, T13, TF; E T10; { E T4, Tc, T7, Tb; T2 = W[0]; T5 = W[1]; T3 = W[2]; T6 = W[3]; T4 = T2 * T3; Tc = T5 * T3; T7 = T5 * T6; Tb = T2 * T6; T8 = T4 + T7; TM = T4 - T7; TO = Tb + Tc; Td = Tb - Tc; T9 = W[4]; Ts = T2 * T9; T1d = T6 * T9; Tx = T5 * T9; T18 = T3 * T9; Te = W[5]; Tt = T5 * Te; T1c = T3 * Te; Tw = T2 * Te; T19 = T6 * Te; Th = W[6]; TB = T3 * Th; T14 = T5 * Th; TG = T6 * Th; TZ = T2 * Th; Tl = W[7]; TC = T6 * Tl; T13 = T2 * Tl; TF = T3 * Tl; T10 = T5 * Tl; } TD = TB + TC; TH = TF - TG; T1y = TZ + T10; T1H = TF + TG; T15 = T13 + T14; T1A = T13 - T14; T11 = TZ - T10; T1F = TB - TC; T1n = FMA(T9, Th, Te * Tl); T1p = FNMS(Te, Th, T9 * Tl); { E T2o, T2p, T2s, T2t; T2o = T8 * Th; T2p = Td * Tl; T2q = T2o + T2p; T2I = T2o - T2p; T2s = T8 * Tl; T2t = Td * Th; T2u = T2s - T2t; T2K = T2s + T2t; } { E T2T, T2U, T2X, T2Y; T2T = TM * Th; T2U = TO * Tl; T2V = T2T - T2U; T3b = T2T + T2U; T2X = TM * Tl; T2Y = TO * Th; T2Z = T2X + T2Y; T3d = T2X - T2Y; Tu = Ts + Tt; Ty = Tw - Tx; T3l = FMA(Tu, Th, Ty * Tl); T3n = FNMS(Ty, Th, Tu * Tl); } T1t = Ts - Tt; T1v = Tw + Tx; T2f = FMA(T1t, Th, T1v * Tl); T2h = FNMS(T1v, Th, T1t * Tl); T1a = T18 - T19; T1e = T1c + T1d; T32 = FMA(T1a, Th, T1e * Tl); T34 = FNMS(T1e, Th, T1a * Tl); T1W = T18 + T19; T1Y = T1c - T1d; T2C = FMA(T1W, Th, T1Y * Tl); T2E = FNMS(T1Y, Th, T1W * Tl); { E Ta, Tf, Ti, Tj; Ta = T8 * T9; Tf = Td * Te; Tg = Ta - Tf; TR = Ta + Tf; Ti = T8 * Te; Tj = Td * T9; Tk = Ti + Tj; TS = Ti - Tj; } Tm = FMA(Tg, Th, Tk * Tl); TV = FNMS(TS, Th, TR * Tl); To = FNMS(Tk, Th, Tg * Tl); TT = FMA(TR, Th, TS * Tl); { E T1K, T1L, T1N, T1O; T1K = TM * T9; T1L = TO * Te; T1M = T1K - T1L; T21 = T1K + T1L; T1N = TM * Te; T1O = TO * T9; T1P = T1N + T1O; T22 = T1N - T1O; } T1Q = FMA(T1M, Th, T1P * Tl); T25 = FNMS(T22, Th, T21 * Tl); T1S = FNMS(T1P, Th, T1M * Tl); T23 = FMA(T21, Th, T22 * Tl); } { E TL, T6f, T8c, T8q, T3F, T5t, T7I, T7W, T2y, T6B, T6y, T7j, T4k, T5J, T4B; E T5G, T3h, T6H, T6O, T7o, T4L, T5N, T52, T5Q, T1i, T7V, T6i, T7D, T3K, T5u; E T3P, T5v, T1E, T6n, T6m, T7e, T3W, T5y, T41, T5z, T29, T6p, T6s, T7f, T47; E T5B, T4c, T5C, T2R, T6z, T6E, T7k, T4v, T5H, T4E, T5K, T3y, T6P, T6K, T7p; E T4W, T5R, T55, T5O; { E T1, T7G, Tq, T7F, TA, T3C, TJ, T3D, Tn, Tp; T1 = ri[0]; T7G = ii[0]; Tn = ri[WS(rs, 16)]; Tp = ii[WS(rs, 16)]; Tq = FMA(Tm, Tn, To * Tp); T7F = FNMS(To, Tn, Tm * Tp); { E Tv, Tz, TE, TI; Tv = ri[WS(rs, 8)]; Tz = ii[WS(rs, 8)]; TA = FMA(Tu, Tv, Ty * Tz); T3C = FNMS(Ty, Tv, Tu * Tz); TE = ri[WS(rs, 24)]; TI = ii[WS(rs, 24)]; TJ = FMA(TD, TE, TH * TI); T3D = FNMS(TH, TE, TD * TI); } { E Tr, TK, T8a, T8b; Tr = T1 + Tq; TK = TA + TJ; TL = Tr + TK; T6f = Tr - TK; T8a = T7G - T7F; T8b = TA - TJ; T8c = T8a - T8b; T8q = T8b + T8a; } { E T3B, T3E, T7E, T7H; T3B = T1 - Tq; T3E = T3C - T3D; T3F = T3B - T3E; T5t = T3B + T3E; T7E = T3C + T3D; T7H = T7F + T7G; T7I = T7E + T7H; T7W = T7H - T7E; } } { E T2e, T4g, T2w, T4z, T2j, T4h, T2n, T4y; { E T2c, T2d, T2r, T2v; T2c = ri[WS(rs, 1)]; T2d = ii[WS(rs, 1)]; T2e = FMA(T2, T2c, T5 * T2d); T4g = FNMS(T5, T2c, T2 * T2d); T2r = ri[WS(rs, 25)]; T2v = ii[WS(rs, 25)]; T2w = FMA(T2q, T2r, T2u * T2v); T4z = FNMS(T2u, T2r, T2q * T2v); } { E T2g, T2i, T2l, T2m; T2g = ri[WS(rs, 17)]; T2i = ii[WS(rs, 17)]; T2j = FMA(T2f, T2g, T2h * T2i); T4h = FNMS(T2h, T2g, T2f * T2i); T2l = ri[WS(rs, 9)]; T2m = ii[WS(rs, 9)]; T2n = FMA(T9, T2l, Te * T2m); T4y = FNMS(Te, T2l, T9 * T2m); } { E T2k, T2x, T6w, T6x; T2k = T2e + T2j; T2x = T2n + T2w; T2y = T2k + T2x; T6B = T2k - T2x; T6w = T4g + T4h; T6x = T4y + T4z; T6y = T6w - T6x; T7j = T6w + T6x; } { E T4i, T4j, T4x, T4A; T4i = T4g - T4h; T4j = T2n - T2w; T4k = T4i + T4j; T5J = T4i - T4j; T4x = T2e - T2j; T4A = T4y - T4z; T4B = T4x - T4A; T5G = T4x + T4A; } } { E T31, T4Y, T3f, T4J, T36, T4Z, T3a, T4I; { E T2W, T30, T3c, T3e; T2W = ri[WS(rs, 31)]; T30 = ii[WS(rs, 31)]; T31 = FMA(T2V, T2W, T2Z * T30); T4Y = FNMS(T2Z, T2W, T2V * T30); T3c = ri[WS(rs, 23)]; T3e = ii[WS(rs, 23)]; T3f = FMA(T3b, T3c, T3d * T3e); T4J = FNMS(T3d, T3c, T3b * T3e); } { E T33, T35, T38, T39; T33 = ri[WS(rs, 15)]; T35 = ii[WS(rs, 15)]; T36 = FMA(T32, T33, T34 * T35); T4Z = FNMS(T34, T33, T32 * T35); T38 = ri[WS(rs, 7)]; T39 = ii[WS(rs, 7)]; T3a = FMA(TR, T38, TS * T39); T4I = FNMS(TS, T38, TR * T39); } { E T37, T3g, T6M, T6N; T37 = T31 + T36; T3g = T3a + T3f; T3h = T37 + T3g; T6H = T37 - T3g; T6M = T4Y + T4Z; T6N = T4I + T4J; T6O = T6M - T6N; T7o = T6M + T6N; } { E T4H, T4K, T50, T51; T4H = T31 - T36; T4K = T4I - T4J; T4L = T4H - T4K; T5N = T4H + T4K; T50 = T4Y - T4Z; T51 = T3a - T3f; T52 = T50 + T51; T5Q = T50 - T51; } } { E TQ, T3G, T1g, T3N, TX, T3H, T17, T3M; { E TN, TP, T1b, T1f; TN = ri[WS(rs, 4)]; TP = ii[WS(rs, 4)]; TQ = FMA(TM, TN, TO * TP); T3G = FNMS(TO, TN, TM * TP); T1b = ri[WS(rs, 12)]; T1f = ii[WS(rs, 12)]; T1g = FMA(T1a, T1b, T1e * T1f); T3N = FNMS(T1e, T1b, T1a * T1f); } { E TU, TW, T12, T16; TU = ri[WS(rs, 20)]; TW = ii[WS(rs, 20)]; TX = FMA(TT, TU, TV * TW); T3H = FNMS(TV, TU, TT * TW); T12 = ri[WS(rs, 28)]; T16 = ii[WS(rs, 28)]; T17 = FMA(T11, T12, T15 * T16); T3M = FNMS(T15, T12, T11 * T16); } { E TY, T1h, T6g, T6h; TY = TQ + TX; T1h = T17 + T1g; T1i = TY + T1h; T7V = T1h - TY; T6g = T3G + T3H; T6h = T3M + T3N; T6i = T6g - T6h; T7D = T6g + T6h; } { E T3I, T3J, T3L, T3O; T3I = T3G - T3H; T3J = TQ - TX; T3K = T3I - T3J; T5u = T3J + T3I; T3L = T17 - T1g; T3O = T3M - T3N; T3P = T3L + T3O; T5v = T3L - T3O; } } { E T1m, T3S, T1C, T3Z, T1r, T3T, T1x, T3Y; { E T1k, T1l, T1z, T1B; T1k = ri[WS(rs, 2)]; T1l = ii[WS(rs, 2)]; T1m = FMA(T8, T1k, Td * T1l); T3S = FNMS(Td, T1k, T8 * T1l); T1z = ri[WS(rs, 26)]; T1B = ii[WS(rs, 26)]; T1C = FMA(T1y, T1z, T1A * T1B); T3Z = FNMS(T1A, T1z, T1y * T1B); } { E T1o, T1q, T1u, T1w; T1o = ri[WS(rs, 18)]; T1q = ii[WS(rs, 18)]; T1r = FMA(T1n, T1o, T1p * T1q); T3T = FNMS(T1p, T1o, T1n * T1q); T1u = ri[WS(rs, 10)]; T1w = ii[WS(rs, 10)]; T1x = FMA(T1t, T1u, T1v * T1w); T3Y = FNMS(T1v, T1u, T1t * T1w); } { E T1s, T1D, T6k, T6l; T1s = T1m + T1r; T1D = T1x + T1C; T1E = T1s + T1D; T6n = T1s - T1D; T6k = T3S + T3T; T6l = T3Y + T3Z; T6m = T6k - T6l; T7e = T6k + T6l; } { E T3U, T3V, T3X, T40; T3U = T3S - T3T; T3V = T1x - T1C; T3W = T3U + T3V; T5y = T3U - T3V; T3X = T1m - T1r; T40 = T3Y - T3Z; T41 = T3X - T40; T5z = T3X + T40; } } { E T1J, T43, T27, T4a, T1U, T44, T20, T49; { E T1G, T1I, T24, T26; T1G = ri[WS(rs, 30)]; T1I = ii[WS(rs, 30)]; T1J = FMA(T1F, T1G, T1H * T1I); T43 = FNMS(T1H, T1G, T1F * T1I); T24 = ri[WS(rs, 22)]; T26 = ii[WS(rs, 22)]; T27 = FMA(T23, T24, T25 * T26); T4a = FNMS(T25, T24, T23 * T26); } { E T1R, T1T, T1X, T1Z; T1R = ri[WS(rs, 14)]; T1T = ii[WS(rs, 14)]; T1U = FMA(T1Q, T1R, T1S * T1T); T44 = FNMS(T1S, T1R, T1Q * T1T); T1X = ri[WS(rs, 6)]; T1Z = ii[WS(rs, 6)]; T20 = FMA(T1W, T1X, T1Y * T1Z); T49 = FNMS(T1Y, T1X, T1W * T1Z); } { E T1V, T28, T6q, T6r; T1V = T1J + T1U; T28 = T20 + T27; T29 = T1V + T28; T6p = T1V - T28; T6q = T43 + T44; T6r = T49 + T4a; T6s = T6q - T6r; T7f = T6q + T6r; } { E T45, T46, T48, T4b; T45 = T43 - T44; T46 = T20 - T27; T47 = T45 + T46; T5B = T45 - T46; T48 = T1J - T1U; T4b = T49 - T4a; T4c = T48 - T4b; T5C = T48 + T4b; } } { E T2B, T4r, T2G, T4s, T4q, T4t, T2M, T4m, T2P, T4n, T4l, T4o; { E T2z, T2A, T2D, T2F; T2z = ri[WS(rs, 5)]; T2A = ii[WS(rs, 5)]; T2B = FMA(T21, T2z, T22 * T2A); T4r = FNMS(T22, T2z, T21 * T2A); T2D = ri[WS(rs, 21)]; T2F = ii[WS(rs, 21)]; T2G = FMA(T2C, T2D, T2E * T2F); T4s = FNMS(T2E, T2D, T2C * T2F); } T4q = T2B - T2G; T4t = T4r - T4s; { E T2J, T2L, T2N, T2O; T2J = ri[WS(rs, 29)]; T2L = ii[WS(rs, 29)]; T2M = FMA(T2I, T2J, T2K * T2L); T4m = FNMS(T2K, T2J, T2I * T2L); T2N = ri[WS(rs, 13)]; T2O = ii[WS(rs, 13)]; T2P = FMA(T1M, T2N, T1P * T2O); T4n = FNMS(T1P, T2N, T1M * T2O); } T4l = T2M - T2P; T4o = T4m - T4n; { E T2H, T2Q, T6C, T6D; T2H = T2B + T2G; T2Q = T2M + T2P; T2R = T2H + T2Q; T6z = T2Q - T2H; T6C = T4r + T4s; T6D = T4m + T4n; T6E = T6C - T6D; T7k = T6C + T6D; } { E T4p, T4u, T4C, T4D; T4p = T4l - T4o; T4u = T4q + T4t; T4v = KP707106781 * (T4p - T4u); T5H = KP707106781 * (T4u + T4p); T4C = T4t - T4q; T4D = T4l + T4o; T4E = KP707106781 * (T4C - T4D); T5K = KP707106781 * (T4C + T4D); } } { E T3k, T4M, T3p, T4N, T4O, T4P, T3t, T4S, T3w, T4T, T4R, T4U; { E T3i, T3j, T3m, T3o; T3i = ri[WS(rs, 3)]; T3j = ii[WS(rs, 3)]; T3k = FMA(T3, T3i, T6 * T3j); T4M = FNMS(T6, T3i, T3 * T3j); T3m = ri[WS(rs, 19)]; T3o = ii[WS(rs, 19)]; T3p = FMA(T3l, T3m, T3n * T3o); T4N = FNMS(T3n, T3m, T3l * T3o); } T4O = T4M - T4N; T4P = T3k - T3p; { E T3r, T3s, T3u, T3v; T3r = ri[WS(rs, 27)]; T3s = ii[WS(rs, 27)]; T3t = FMA(Th, T3r, Tl * T3s); T4S = FNMS(Tl, T3r, Th * T3s); T3u = ri[WS(rs, 11)]; T3v = ii[WS(rs, 11)]; T3w = FMA(Tg, T3u, Tk * T3v); T4T = FNMS(Tk, T3u, Tg * T3v); } T4R = T3t - T3w; T4U = T4S - T4T; { E T3q, T3x, T6I, T6J; T3q = T3k + T3p; T3x = T3t + T3w; T3y = T3q + T3x; T6P = T3x - T3q; T6I = T4M + T4N; T6J = T4S + T4T; T6K = T6I - T6J; T7p = T6I + T6J; } { E T4Q, T4V, T53, T54; T4Q = T4O - T4P; T4V = T4R + T4U; T4W = KP707106781 * (T4Q - T4V); T5R = KP707106781 * (T4Q + T4V); T53 = T4R - T4U; T54 = T4P + T4O; T55 = KP707106781 * (T53 - T54); T5O = KP707106781 * (T54 + T53); } } { E T2b, T7x, T7K, T7M, T3A, T7L, T7A, T7B; { E T1j, T2a, T7C, T7J; T1j = TL + T1i; T2a = T1E + T29; T2b = T1j + T2a; T7x = T1j - T2a; T7C = T7e + T7f; T7J = T7D + T7I; T7K = T7C + T7J; T7M = T7J - T7C; } { E T2S, T3z, T7y, T7z; T2S = T2y + T2R; T3z = T3h + T3y; T3A = T2S + T3z; T7L = T3z - T2S; T7y = T7j + T7k; T7z = T7o + T7p; T7A = T7y - T7z; T7B = T7y + T7z; } ri[WS(rs, 16)] = T2b - T3A; ii[WS(rs, 16)] = T7K - T7B; ri[0] = T2b + T3A; ii[0] = T7B + T7K; ri[WS(rs, 24)] = T7x - T7A; ii[WS(rs, 24)] = T7M - T7L; ri[WS(rs, 8)] = T7x + T7A; ii[WS(rs, 8)] = T7L + T7M; } { E T7h, T7t, T7Q, T7S, T7m, T7u, T7r, T7v; { E T7d, T7g, T7O, T7P; T7d = TL - T1i; T7g = T7e - T7f; T7h = T7d + T7g; T7t = T7d - T7g; T7O = T29 - T1E; T7P = T7I - T7D; T7Q = T7O + T7P; T7S = T7P - T7O; } { E T7i, T7l, T7n, T7q; T7i = T2y - T2R; T7l = T7j - T7k; T7m = T7i + T7l; T7u = T7l - T7i; T7n = T3h - T3y; T7q = T7o - T7p; T7r = T7n - T7q; T7v = T7n + T7q; } { E T7s, T7N, T7w, T7R; T7s = KP707106781 * (T7m + T7r); ri[WS(rs, 20)] = T7h - T7s; ri[WS(rs, 4)] = T7h + T7s; T7N = KP707106781 * (T7u + T7v); ii[WS(rs, 4)] = T7N + T7Q; ii[WS(rs, 20)] = T7Q - T7N; T7w = KP707106781 * (T7u - T7v); ri[WS(rs, 28)] = T7t - T7w; ri[WS(rs, 12)] = T7t + T7w; T7R = KP707106781 * (T7r - T7m); ii[WS(rs, 12)] = T7R + T7S; ii[WS(rs, 28)] = T7S - T7R; } } { E T6j, T7X, T83, T6X, T6u, T7U, T77, T7b, T70, T82, T6G, T6U, T74, T7a, T6R; E T6V; { E T6o, T6t, T6A, T6F; T6j = T6f - T6i; T7X = T7V + T7W; T83 = T7W - T7V; T6X = T6f + T6i; T6o = T6m - T6n; T6t = T6p + T6s; T6u = KP707106781 * (T6o - T6t); T7U = KP707106781 * (T6o + T6t); { E T75, T76, T6Y, T6Z; T75 = T6H + T6K; T76 = T6O + T6P; T77 = FNMS(KP382683432, T76, KP923879532 * T75); T7b = FMA(KP923879532, T76, KP382683432 * T75); T6Y = T6n + T6m; T6Z = T6p - T6s; T70 = KP707106781 * (T6Y + T6Z); T82 = KP707106781 * (T6Z - T6Y); } T6A = T6y - T6z; T6F = T6B - T6E; T6G = FMA(KP923879532, T6A, KP382683432 * T6F); T6U = FNMS(KP923879532, T6F, KP382683432 * T6A); { E T72, T73, T6L, T6Q; T72 = T6y + T6z; T73 = T6B + T6E; T74 = FMA(KP382683432, T72, KP923879532 * T73); T7a = FNMS(KP382683432, T73, KP923879532 * T72); T6L = T6H - T6K; T6Q = T6O - T6P; T6R = FNMS(KP923879532, T6Q, KP382683432 * T6L); T6V = FMA(KP382683432, T6Q, KP923879532 * T6L); } } { E T6v, T6S, T81, T84; T6v = T6j + T6u; T6S = T6G + T6R; ri[WS(rs, 22)] = T6v - T6S; ri[WS(rs, 6)] = T6v + T6S; T81 = T6U + T6V; T84 = T82 + T83; ii[WS(rs, 6)] = T81 + T84; ii[WS(rs, 22)] = T84 - T81; } { E T6T, T6W, T85, T86; T6T = T6j - T6u; T6W = T6U - T6V; ri[WS(rs, 30)] = T6T - T6W; ri[WS(rs, 14)] = T6T + T6W; T85 = T6R - T6G; T86 = T83 - T82; ii[WS(rs, 14)] = T85 + T86; ii[WS(rs, 30)] = T86 - T85; } { E T71, T78, T7T, T7Y; T71 = T6X + T70; T78 = T74 + T77; ri[WS(rs, 18)] = T71 - T78; ri[WS(rs, 2)] = T71 + T78; T7T = T7a + T7b; T7Y = T7U + T7X; ii[WS(rs, 2)] = T7T + T7Y; ii[WS(rs, 18)] = T7Y - T7T; } { E T79, T7c, T7Z, T80; T79 = T6X - T70; T7c = T7a - T7b; ri[WS(rs, 26)] = T79 - T7c; ri[WS(rs, 10)] = T79 + T7c; T7Z = T77 - T74; T80 = T7X - T7U; ii[WS(rs, 10)] = T7Z + T80; ii[WS(rs, 26)] = T80 - T7Z; } } { E T3R, T5d, T8r, T8x, T4e, T8o, T5n, T5r, T4G, T5a, T5g, T8w, T5k, T5q, T57; E T5b, T3Q, T8p; T3Q = KP707106781 * (T3K - T3P); T3R = T3F - T3Q; T5d = T3F + T3Q; T8p = KP707106781 * (T5v - T5u); T8r = T8p + T8q; T8x = T8q - T8p; { E T42, T4d, T5l, T5m; T42 = FNMS(KP923879532, T41, KP382683432 * T3W); T4d = FMA(KP382683432, T47, KP923879532 * T4c); T4e = T42 - T4d; T8o = T42 + T4d; T5l = T4L + T4W; T5m = T52 + T55; T5n = FNMS(KP555570233, T5m, KP831469612 * T5l); T5r = FMA(KP831469612, T5m, KP555570233 * T5l); } { E T4w, T4F, T5e, T5f; T4w = T4k - T4v; T4F = T4B - T4E; T4G = FMA(KP980785280, T4w, KP195090322 * T4F); T5a = FNMS(KP980785280, T4F, KP195090322 * T4w); T5e = FMA(KP923879532, T3W, KP382683432 * T41); T5f = FNMS(KP923879532, T47, KP382683432 * T4c); T5g = T5e + T5f; T8w = T5f - T5e; } { E T5i, T5j, T4X, T56; T5i = T4k + T4v; T5j = T4B + T4E; T5k = FMA(KP555570233, T5i, KP831469612 * T5j); T5q = FNMS(KP555570233, T5j, KP831469612 * T5i); T4X = T4L - T4W; T56 = T52 - T55; T57 = FNMS(KP980785280, T56, KP195090322 * T4X); T5b = FMA(KP195090322, T56, KP980785280 * T4X); } { E T4f, T58, T8v, T8y; T4f = T3R + T4e; T58 = T4G + T57; ri[WS(rs, 23)] = T4f - T58; ri[WS(rs, 7)] = T4f + T58; T8v = T5a + T5b; T8y = T8w + T8x; ii[WS(rs, 7)] = T8v + T8y; ii[WS(rs, 23)] = T8y - T8v; } { E T59, T5c, T8z, T8A; T59 = T3R - T4e; T5c = T5a - T5b; ri[WS(rs, 31)] = T59 - T5c; ri[WS(rs, 15)] = T59 + T5c; T8z = T57 - T4G; T8A = T8x - T8w; ii[WS(rs, 15)] = T8z + T8A; ii[WS(rs, 31)] = T8A - T8z; } { E T5h, T5o, T8n, T8s; T5h = T5d + T5g; T5o = T5k + T5n; ri[WS(rs, 19)] = T5h - T5o; ri[WS(rs, 3)] = T5h + T5o; T8n = T5q + T5r; T8s = T8o + T8r; ii[WS(rs, 3)] = T8n + T8s; ii[WS(rs, 19)] = T8s - T8n; } { E T5p, T5s, T8t, T8u; T5p = T5d - T5g; T5s = T5q - T5r; ri[WS(rs, 27)] = T5p - T5s; ri[WS(rs, 11)] = T5p + T5s; T8t = T5n - T5k; T8u = T8r - T8o; ii[WS(rs, 11)] = T8t + T8u; ii[WS(rs, 27)] = T8u - T8t; } } { E T5x, T5Z, T8d, T8j, T5E, T88, T69, T6d, T5M, T5W, T62, T8i, T66, T6c, T5T; E T5X, T5w, T89; T5w = KP707106781 * (T5u + T5v); T5x = T5t - T5w; T5Z = T5t + T5w; T89 = KP707106781 * (T3K + T3P); T8d = T89 + T8c; T8j = T8c - T89; { E T5A, T5D, T67, T68; T5A = FNMS(KP382683432, T5z, KP923879532 * T5y); T5D = FMA(KP923879532, T5B, KP382683432 * T5C); T5E = T5A - T5D; T88 = T5A + T5D; T67 = T5N + T5O; T68 = T5Q + T5R; T69 = FNMS(KP195090322, T68, KP980785280 * T67); T6d = FMA(KP195090322, T67, KP980785280 * T68); } { E T5I, T5L, T60, T61; T5I = T5G - T5H; T5L = T5J - T5K; T5M = FMA(KP555570233, T5I, KP831469612 * T5L); T5W = FNMS(KP831469612, T5I, KP555570233 * T5L); T60 = FMA(KP382683432, T5y, KP923879532 * T5z); T61 = FNMS(KP382683432, T5B, KP923879532 * T5C); T62 = T60 + T61; T8i = T61 - T60; } { E T64, T65, T5P, T5S; T64 = T5G + T5H; T65 = T5J + T5K; T66 = FMA(KP980785280, T64, KP195090322 * T65); T6c = FNMS(KP195090322, T64, KP980785280 * T65); T5P = T5N - T5O; T5S = T5Q - T5R; T5T = FNMS(KP831469612, T5S, KP555570233 * T5P); T5X = FMA(KP831469612, T5P, KP555570233 * T5S); } { E T5F, T5U, T8h, T8k; T5F = T5x + T5E; T5U = T5M + T5T; ri[WS(rs, 21)] = T5F - T5U; ri[WS(rs, 5)] = T5F + T5U; T8h = T5W + T5X; T8k = T8i + T8j; ii[WS(rs, 5)] = T8h + T8k; ii[WS(rs, 21)] = T8k - T8h; } { E T5V, T5Y, T8l, T8m; T5V = T5x - T5E; T5Y = T5W - T5X; ri[WS(rs, 29)] = T5V - T5Y; ri[WS(rs, 13)] = T5V + T5Y; T8l = T5T - T5M; T8m = T8j - T8i; ii[WS(rs, 13)] = T8l + T8m; ii[WS(rs, 29)] = T8m - T8l; } { E T63, T6a, T87, T8e; T63 = T5Z + T62; T6a = T66 + T69; ri[WS(rs, 17)] = T63 - T6a; ri[WS(rs, 1)] = T63 + T6a; T87 = T6c + T6d; T8e = T88 + T8d; ii[WS(rs, 1)] = T87 + T8e; ii[WS(rs, 17)] = T8e - T87; } { E T6b, T6e, T8f, T8g; T6b = T5Z - T62; T6e = T6c - T6d; ri[WS(rs, 25)] = T6b - T6e; ri[WS(rs, 9)] = T6b + T6e; T8f = T69 - T66; T8g = T8d - T88; ii[WS(rs, 9)] = T8f + T8g; ii[WS(rs, 25)] = T8g - T8f; } } } } } } static const tw_instr twinstr[] = { {TW_CEXP, 0, 1}, {TW_CEXP, 0, 3}, {TW_CEXP, 0, 9}, {TW_CEXP, 0, 27}, {TW_NEXT, 1, 0} }; static const ct_desc desc = { 32, "t2_32", twinstr, &GENUS, {376, 168, 112, 0}, 0, 0, 0 }; void X(codelet_t2_32) (planner *p) { X(kdft_dit_register) (p, t2_32, &desc); } #endif /* HAVE_FMA */ fftw-3.3.3/dft/scalar/codelets/t2_64.c0000644000175000017500000035635712054410341014252 00000000000000/* * Copyright (c) 2003, 2007-11 Matteo Frigo * Copyright (c) 2003, 2007-11 Massachusetts Institute of Technology * * This program is free software; you can redistribute it and/or modify * it under the terms of the GNU General Public License as published by * the Free Software Foundation; either version 2 of the License, or * (at your option) any later version. * * This program is distributed in the hope that it will be useful, * but WITHOUT ANY WARRANTY; without even the implied warranty of * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the * GNU General Public License for more details. * * You should have received a copy of the GNU General Public License * along with this program; if not, write to the Free Software * Foundation, Inc., 51 Franklin Street, Fifth Floor, Boston, MA 02110-1301 USA * */ /* This file was automatically generated --- DO NOT EDIT */ /* Generated on Sun Nov 25 07:36:01 EST 2012 */ #include "codelet-dft.h" #ifdef HAVE_FMA /* Generated by: ../../../genfft/gen_twiddle.native -fma -reorder-insns -schedule-for-pipeline -compact -variables 4 -pipeline-latency 4 -twiddle-log3 -precompute-twiddles -n 64 -name t2_64 -include t.h */ /* * This function contains 1154 FP additions, 840 FP multiplications, * (or, 520 additions, 206 multiplications, 634 fused multiply/add), * 349 stack variables, 15 constants, and 256 memory accesses */ #include "t.h" static void t2_64(R *ri, R *ii, const R *W, stride rs, INT mb, INT me, INT ms) { DK(KP995184726, +0.995184726672196886244836953109479921575474869); DK(KP773010453, +0.773010453362736960810906609758469800971041293); DK(KP956940335, +0.956940335732208864935797886980269969482849206); DK(KP881921264, +0.881921264348355029712756863660388349508442621); DK(KP820678790, +0.820678790828660330972281985331011598767386482); DK(KP098491403, +0.098491403357164253077197521291327432293052451); DK(KP534511135, +0.534511135950791641089685961295362908582039528); DK(KP303346683, +0.303346683607342391675883946941299872384187453); DK(KP831469612, +0.831469612302545237078788377617905756738560812); DK(KP980785280, +0.980785280403230449126182236134239036973933731); DK(KP668178637, +0.668178637919298919997757686523080761552472251); DK(KP198912367, +0.198912367379658006911597622644676228597850501); DK(KP923879532, +0.923879532511286756128183189396788286822416626); DK(KP707106781, +0.707106781186547524400844362104849039284835938); DK(KP414213562, +0.414213562373095048801688724209698078569671875); { INT m; for (m = mb, W = W + (mb * 10); m < me; m = m + 1, ri = ri + ms, ii = ii + ms, W = W + 10, MAKE_VOLATILE_STRIDE(128, rs)) { E Tg0, TlC, TlB, Tg3; { E T2, T3, Tc, T8, Te, T5, T6, T14, T3d, T3i, TJ, T7, Tr, T3g, TG; E T10, T3a, TL, TP, Tb, Tt, T17, Td, Ti, T3N, T3R, T1i, Tu, T1I, T2U; E T1t, T3U, T5O, T48, T2u, T7B, TK, T79, T3D, T2h, T2l, T3G, T1x, T3X, T2d; E T1M, T2X, T4B, T4x, T3j, T4T, T29, T5s, T81, T5w, T7X, T7N, T7h, T64, T6a; E T6e, T7l, T60, T7R, T6h, T5A, T7o, T6J, T6k, T5E, T6N, T7r, T6x, T6t, T7c; E TO, T2x, T7E, TU, TQ, T2C, T2y, T5R, T4b, T4c, T4g, T4W, T3m, T3r, T3n; E T1k, Tx, Ty, T4p, T4s, TC, T23, T1Z, T19, Th, T31, T35, T1e, T44, T41; E T1a, T6W, T70, T55, T59, T3v, T3z, Tf, T1R, T2N, T2Q, T1V, T1p, T1l, Tm; { E T1H, T1s, T2g, Tg, Tw, TH, T2t, T47, T3h, T3M, T4w, T28, T3Q, T4A, T2c; E Ts; { E T4, T13, TI, TF, TZ, Ta, T9; T2 = W[0]; T3 = W[2]; Tc = W[5]; T8 = W[4]; Te = W[6]; T4 = T2 * T3; T13 = T2 * Tc; TI = T3 * Tc; TF = T3 * T8; T1H = T8 * Te; TZ = T2 * T8; T5 = W[1]; T6 = W[3]; T1s = T3 * Te; T2g = T2 * Te; T14 = FNMS(T5, T8, T13); T3d = FMA(T5, T8, T13); T3i = FNMS(T6, T8, TI); TJ = FMA(T6, T8, TI); T7 = FNMS(T5, T6, T4); Tr = FMA(T5, T6, T4); Ta = T2 * T6; Tg = T7 * Tc; Tw = Tr * Tc; T3g = FMA(T6, Tc, TF); TG = FNMS(T6, Tc, TF); T10 = FMA(T5, Tc, TZ); T3a = FNMS(T5, Tc, TZ); TH = TG * Te; T2t = T10 * Te; T47 = T3a * Te; T3h = T3g * Te; TL = W[8]; TP = W[9]; T9 = T7 * T8; Tb = FMA(T5, T3, Ta); Tt = FNMS(T5, T3, Ta); T3M = T2 * TL; T4w = T8 * TL; T28 = T3 * TL; T3Q = T2 * TP; T4A = T8 * TP; T2c = T3 * TP; T17 = FNMS(Tb, Tc, T9); Td = FMA(Tb, Tc, T9); Ts = Tr * T8; Ti = W[7]; } { E T5r, T80, T1L, T2k, T1w, T5z, T2B, T2v; T3N = FMA(T5, TP, T3M); T3R = FNMS(T5, TL, T3Q); T1i = FMA(Tt, Tc, Ts); Tu = FNMS(Tt, Tc, Ts); T1I = FNMS(Tc, Ti, T1H); T2U = FMA(Tc, Ti, T1H); T1t = FMA(T6, Ti, T1s); T3U = FNMS(T6, Ti, T1s); T5O = FNMS(T3d, Ti, T47); T48 = FMA(T3d, Ti, T47); T2u = FMA(T14, Ti, T2t); T7B = FNMS(T14, Ti, T2t); T1L = T8 * Ti; T2k = T2 * Ti; T1w = T3 * Ti; TK = FMA(TJ, Ti, TH); T79 = FNMS(TJ, Ti, TH); T3D = FMA(T5, Ti, T2g); T2h = FNMS(T5, Ti, T2g); T2l = FMA(T5, Te, T2k); T3G = FNMS(T5, Te, T2k); T1x = FNMS(T6, Te, T1w); T3X = FMA(T6, Te, T1w); T2d = FNMS(T6, TL, T2c); T1M = FMA(Tc, Te, T1L); T2X = FNMS(Tc, Te, T1L); T4B = FNMS(Tc, TL, T4A); T4x = FMA(Tc, TP, T4w); T3j = FMA(T3i, Ti, T3h); T4T = FNMS(T3i, Ti, T3h); T29 = FMA(T6, TP, T28); T5r = T3g * TL; T80 = T7 * TP; { E T7M, T7g, T63, T5v, T7W; T5v = T3g * TP; T7W = T7 * TL; T5s = FMA(T3i, TP, T5r); T81 = FNMS(Tb, TL, T80); T5w = FNMS(T3i, TL, T5v); T7X = FMA(Tb, TP, T7W); T7M = TG * TL; T7g = T10 * TL; T63 = T3a * TP; { E T6d, T7k, T69, T5Z, T7Q; T69 = Tr * TL; T7N = FMA(TJ, TP, T7M); T7h = FMA(T14, TP, T7g); T64 = FNMS(T3d, TL, T63); T6a = FMA(Tt, TP, T69); T6d = Tr * TP; T7k = T10 * TP; T5Z = T3a * TL; T7Q = TG * TP; T6e = FNMS(Tt, TL, T6d); T7l = FNMS(T14, TL, T7k); T60 = FMA(T3d, TP, T5Z); T7R = FNMS(TJ, TL, T7Q); T5z = Tr * Te; } } { E T6I, T5D, T6M, T6s, T6w; T6I = T7 * Te; T5D = Tr * Ti; T6M = T7 * Ti; T6h = FNMS(Tt, Ti, T5z); T5A = FMA(Tt, Ti, T5z); T7o = FMA(Tb, Ti, T6I); T6J = FNMS(Tb, Ti, T6I); T6k = FMA(Tt, Te, T5D); T5E = FNMS(Tt, Te, T5D); T6N = FMA(Tb, Te, T6M); T7r = FNMS(Tb, Te, T6M); T6s = T2U * TL; T6w = T2U * TP; { E TN, TT, TM, T2w; TN = TG * Ti; T2w = T10 * Ti; T6x = FNMS(T2X, TL, T6w); T6t = FMA(T2X, TP, T6s); T7c = FMA(TJ, Te, TN); TO = FNMS(TJ, Te, TN); TT = TK * TP; TM = TK * TL; T2x = FNMS(T14, Te, T2w); T7E = FMA(T14, Te, T2w); TU = FNMS(TO, TL, TT); TQ = FMA(TO, TP, TM); T2B = T2u * TP; T2v = T2u * TL; } } { E T1Y, T22, Tv, TB; { E T49, T4f, T4a, T3l, T3q, T3k; T4a = T3a * Ti; T2C = FNMS(T2x, TL, T2B); T2y = FMA(T2x, TP, T2v); T5R = FMA(T3d, Te, T4a); T4b = FNMS(T3d, Te, T4a); T49 = T48 * TL; T4f = T48 * TP; T3l = T3g * Ti; T4c = FMA(T4b, TP, T49); T4g = FNMS(T4b, TL, T4f); T4W = FMA(T3i, Te, T3l); T3m = FNMS(T3i, Te, T3l); T1Y = Tu * TL; T3q = T3j * TP; T3k = T3j * TL; T22 = Tu * TP; Tv = Tu * Te; T3r = FNMS(T3m, TL, T3q); T3n = FMA(T3m, TP, T3k); TB = Tu * Ti; T1k = FNMS(Tt, T8, Tw); Tx = FMA(Tt, T8, Tw); } { E T30, T34, T18, T1d; T30 = T17 * TL; T34 = T17 * TP; T18 = T17 * Te; Ty = FMA(Tx, Ti, Tv); T4p = FNMS(Tx, Ti, Tv); T4s = FMA(Tx, Te, TB); TC = FNMS(Tx, Te, TB); T23 = FNMS(Tx, TL, T22); T1Z = FMA(Tx, TP, T1Y); T1d = T17 * Ti; T19 = FMA(Tb, T8, Tg); Th = FNMS(Tb, T8, Tg); { E T1j, T1o, T1Q, T1U; T1j = T1i * TL; { E T6V, T6Z, T54, T58; T6V = Ty * TL; T6Z = Ty * TP; T31 = FMA(T19, TP, T30); T35 = FNMS(T19, TL, T34); T1e = FMA(T19, Te, T1d); T44 = FNMS(T19, Te, T1d); T41 = FMA(T19, Ti, T18); T1a = FNMS(T19, Ti, T18); T6W = FMA(TC, TP, T6V); T70 = FNMS(TC, TL, T6Z); T1o = T1i * TP; T54 = T41 * TL; T58 = T41 * TP; T1Q = T1i * Te; T1U = T1i * Ti; T55 = FMA(T44, TP, T54); T59 = FNMS(T44, TL, T58); } T3v = Td * TL; T3z = Td * TP; Tf = Td * Te; T1R = FMA(T1k, Ti, T1Q); T2N = FNMS(T1k, Ti, T1Q); T2Q = FMA(T1k, Te, T1U); T1V = FNMS(T1k, Te, T1U); T1p = FNMS(T1k, TL, T1o); T1l = FMA(T1k, TP, T1j); Tm = Td * Ti; } } } } } { E Tl9, TlD, TY, Tg4, T8w, TdS, TkE, Tkd, T2G, Tge, Tgh, TiK, Te1, T98, Te0; E T9f, Te5, T9p, Tgq, T39, Te8, T9M, TiN, Tgn, TeE, TbI, Thr, T74, TeP, TcB; E Tja, Thc, T8D, TdT, T1B, TkD, T8K, TdU, Tg7, Tk7, T8T, TdY, T27, Tg9, T90; E TdX, Tgc, TiJ, T9Y, Tec, T4k, TgB, Tal, Tef, Tgy, TiT, Taz, Tel, T5d, Th0; E Tbs, Tew, TgL, TiZ, T3K, Tgo, Tgt, TiO, T9P, Te6, T9E, Te9, T4L, Tgz, TgE; E TiU, Tao, Ted, Tad, Teg, T5I, TgM, Th3, Tj0, Tbv, Tem, TaO, Tex, T7v, Thd; E Thu, Tjb, TcE, TeF, TbX, TeQ, T68, Tj5, Tez, Teq, Tbj, Tbx, TgS, Th5, T6B; E Tj6, TeA, Tet, Tb4, Tby, TgX, Th6, T7V, Tjg, TeS, TeJ, Tcs, TcG, Thj, Thw; E T84, T83, T85, Tc7, T8k, Tc3, T86, T89, T8b; { E T3w, T3A, T4H, T4E, T8e, T8i, T5j, T5n, T4U, T4S, T4V, Tau, T5b, Tbq, T4X; E T50, T52; { E T72, Tcz, Tcv, T6Q, Tha, TbG, T6U, Tcx, T99, T9e; { E T1, Tkb, Tp, Tka, TR, TV, TE, T8s, TS, T8t; { E Tn, Tj, T8d, T8h, T5i, T5m; T1 = ri[0]; T8d = T1R * TL; T8h = T1R * TP; T3w = FMA(Th, TP, T3v); T3A = FNMS(Th, TL, T3z); Tn = FMA(Th, Te, Tm); T4H = FNMS(Th, Te, Tm); T4E = FMA(Th, Ti, Tf); Tj = FNMS(Th, Ti, Tf); T8e = FMA(T1V, TP, T8d); T8i = FNMS(T1V, TL, T8h); Tkb = ii[0]; T5i = T4E * TL; T5m = T4E * TP; { E Tk, To, Tl, Tk9; Tk = ri[WS(rs, 32)]; To = ii[WS(rs, 32)]; T5j = FMA(T4H, TP, T5i); T5n = FNMS(T4H, TL, T5m); Tl = Tj * Tk; Tk9 = Tj * To; { E Tz, TD, TA, T8r; Tz = ri[WS(rs, 16)]; TD = ii[WS(rs, 16)]; Tp = FMA(Tn, To, Tl); Tka = FNMS(Tn, Tk, Tk9); TA = Ty * Tz; T8r = Ty * TD; TR = ri[WS(rs, 48)]; TV = ii[WS(rs, 48)]; TE = FMA(TC, TD, TA); T8s = FNMS(TC, Tz, T8r); TS = TQ * TR; T8t = TQ * TV; } } } { E T8q, Tq, Tl7, Tkc, TW, T8u; T8q = T1 - Tp; Tq = T1 + Tp; Tl7 = Tkb - Tka; Tkc = Tka + Tkb; TW = FMA(TU, TV, TS); T8u = FNMS(TU, TR, T8t); { E TX, Tl8, T8v, Tk8; TX = TE + TW; Tl8 = TE - TW; T8v = T8s - T8u; Tk8 = T8s + T8u; Tl9 = Tl7 - Tl8; TlD = Tl8 + Tl7; TY = Tq + TX; Tg4 = Tq - TX; T8w = T8q - T8v; TdS = T8q + T8v; TkE = Tkc - Tk8; Tkd = Tk8 + Tkc; } } } { E T2f, T93, T2E, T9d, T2n, T95, T2s, T9b; { E T2a, T2e, T2i, T2m; T2a = ri[WS(rs, 60)]; T2e = ii[WS(rs, 60)]; { E T2z, T2D, T2b, T92, T2A, T9c; T2z = ri[WS(rs, 44)]; T2D = ii[WS(rs, 44)]; T2b = T29 * T2a; T92 = T29 * T2e; T2A = T2y * T2z; T9c = T2y * T2D; T2f = FMA(T2d, T2e, T2b); T93 = FNMS(T2d, T2a, T92); T2E = FMA(T2C, T2D, T2A); T9d = FNMS(T2C, T2z, T9c); } T2i = ri[WS(rs, 28)]; T2m = ii[WS(rs, 28)]; { E T2p, T2r, T2j, T94, T2q, T9a; T2p = ri[WS(rs, 12)]; T2r = ii[WS(rs, 12)]; T2j = T2h * T2i; T94 = T2h * T2m; T2q = TG * T2p; T9a = TG * T2r; T2n = FMA(T2l, T2m, T2j); T95 = FNMS(T2l, T2i, T94); T2s = FMA(TJ, T2r, T2q); T9b = FNMS(TJ, T2p, T9a); } } { E T2o, Tgf, T96, T97, T2F, Tgg; T99 = T2f - T2n; T2o = T2f + T2n; Tgf = T93 + T95; T96 = T93 - T95; T97 = T2s - T2E; T2F = T2s + T2E; Tgg = T9b + T9d; T9e = T9b - T9d; T2G = T2o + T2F; Tge = T2o - T2F; Tgh = Tgf - Tgg; TiK = Tgf + Tgg; Te1 = T96 - T97; T98 = T96 + T97; } } { E T9K, T2T, T9G, T9n, Tgl, T9o, T38, T9I; { E T2M, T9k, T37, T2V, T2S, T2W, T2Y, T9m, T32, T33, T36, T2Z, T9H; { E T2J, T2L, T2K, T9j; T2J = ri[WS(rs, 2)]; T2L = ii[WS(rs, 2)]; T32 = ri[WS(rs, 50)]; Te0 = T99 + T9e; T9f = T99 - T9e; T2K = Tr * T2J; T9j = Tr * T2L; T33 = T31 * T32; T36 = ii[WS(rs, 50)]; T2M = FMA(Tt, T2L, T2K); T9k = FNMS(Tt, T2J, T9j); } { E T2O, T9J, T2R, T2P, T9l; T2O = ri[WS(rs, 34)]; T37 = FMA(T35, T36, T33); T9J = T31 * T36; T2R = ii[WS(rs, 34)]; T2P = T2N * T2O; T2V = ri[WS(rs, 18)]; T9K = FNMS(T35, T32, T9J); T9l = T2N * T2R; T2S = FMA(T2Q, T2R, T2P); T2W = T2U * T2V; T2Y = ii[WS(rs, 18)]; T9m = FNMS(T2Q, T2O, T9l); } T2T = T2M + T2S; T9G = T2M - T2S; T2Z = FMA(T2X, T2Y, T2W); T9H = T2U * T2Y; T9n = T9k - T9m; Tgl = T9k + T9m; T9o = T2Z - T37; T38 = T2Z + T37; T9I = FNMS(T2X, T2V, T9H); } { E T6H, TbD, T6P, T6R, T6T, TbF, T6S, Tcw; { E T6X, T71, T6E, TbC, T6K, TbE; { E T6F, T6G, T9L, Tgm; T6E = ri[WS(rs, 63)]; Te5 = T9n - T9o; T9p = T9n + T9o; Tgq = T2T - T38; T39 = T2T + T38; T9L = T9I - T9K; Tgm = T9I + T9K; T6F = TL * T6E; T6G = ii[WS(rs, 63)]; Te8 = T9G + T9L; T9M = T9G - T9L; TiN = Tgl + Tgm; Tgn = Tgl - Tgm; TbC = TL * T6G; T6H = FMA(TP, T6G, T6F); } T6X = ri[WS(rs, 47)]; T71 = ii[WS(rs, 47)]; TbD = FNMS(TP, T6E, TbC); { E T6O, T6L, T6Y, Tcy; T6K = ri[WS(rs, 31)]; T6Y = T6W * T6X; Tcy = T6W * T71; T6O = ii[WS(rs, 31)]; T6L = T6J * T6K; T72 = FMA(T70, T71, T6Y); Tcz = FNMS(T70, T6X, Tcy); TbE = T6J * T6O; T6P = FMA(T6N, T6O, T6L); } T6R = ri[WS(rs, 15)]; T6T = ii[WS(rs, 15)]; TbF = FNMS(T6N, T6K, TbE); } Tcv = T6H - T6P; T6Q = T6H + T6P; T6S = TK * T6R; Tcw = TK * T6T; Tha = TbD + TbF; TbG = TbD - TbF; T6U = FMA(TO, T6T, T6S); Tcx = FNMS(TO, T6R, Tcw); } } { E T1J, T1G, T1K, T8O, T25, T8Y, T1N, T1S, T1W; { E T1b, T16, T1c, T8y, T1z, T8I, T1f, T1m, T1q; { E T11, T12, T15, T1u, T1y, T8x, T1v, T8H; T11 = ri[WS(rs, 8)]; { E TbH, T73, TcA, Thb; TbH = T6U - T72; T73 = T6U + T72; TcA = Tcx - Tcz; Thb = Tcx + Tcz; TeE = TbG - TbH; TbI = TbG + TbH; Thr = T6Q - T73; T74 = T6Q + T73; TeP = Tcv + TcA; TcB = Tcv - TcA; Tja = Tha + Thb; Thc = Tha - Thb; T12 = T10 * T11; } T15 = ii[WS(rs, 8)]; T1u = ri[WS(rs, 24)]; T1y = ii[WS(rs, 24)]; T1b = ri[WS(rs, 40)]; T16 = FMA(T14, T15, T12); T8x = T10 * T15; T1v = T1t * T1u; T8H = T1t * T1y; T1c = T1a * T1b; T8y = FNMS(T14, T11, T8x); T1z = FMA(T1x, T1y, T1v); T8I = FNMS(T1x, T1u, T8H); T1f = ii[WS(rs, 40)]; T1m = ri[WS(rs, 56)]; T1q = ii[WS(rs, 56)]; } { E T1D, T1E, T1F, T20, T24, T8N, T21, T8X; { E T1h, T8C, T8A, T1r, T8G, Tg5, T8B; T1D = ri[WS(rs, 4)]; { E T1g, T8z, T1n, T8F; T1g = FMA(T1e, T1f, T1c); T8z = T1a * T1f; T1n = T1l * T1m; T8F = T1l * T1q; T1h = T16 + T1g; T8C = T16 - T1g; T8A = FNMS(T1e, T1b, T8z); T1r = FMA(T1p, T1q, T1n); T8G = FNMS(T1p, T1m, T8F); T1E = T7 * T1D; } Tg5 = T8y + T8A; T8B = T8y - T8A; { E T1A, T8E, Tg6, T8J; T1A = T1r + T1z; T8E = T1r - T1z; Tg6 = T8G + T8I; T8J = T8G - T8I; T8D = T8B - T8C; TdT = T8C + T8B; T1B = T1h + T1A; TkD = T1A - T1h; T8K = T8E + T8J; TdU = T8E - T8J; Tg7 = Tg5 - Tg6; Tk7 = Tg5 + Tg6; T1F = ii[WS(rs, 4)]; } } T20 = ri[WS(rs, 52)]; T24 = ii[WS(rs, 52)]; T1J = ri[WS(rs, 36)]; T1G = FMA(Tb, T1F, T1E); T8N = T7 * T1F; T21 = T1Z * T20; T8X = T1Z * T24; T1K = T1I * T1J; T8O = FNMS(Tb, T1D, T8N); T25 = FMA(T23, T24, T21); T8Y = FNMS(T23, T20, T8X); T1N = ii[WS(rs, 36)]; T1S = ri[WS(rs, 20)]; T1W = ii[WS(rs, 20)]; } } { E T3V, T3T, T3W, T9T, T4i, Taj, T3Y, T42, T45; { E T3O, T3P, T3S, T4d, T4h, T9S, T4e, Tai; { E T1P, T8U, T8Q, T1X, T8W, Tga, T8R; T3O = ri[WS(rs, 62)]; { E T1O, T8P, T1T, T8V; T1O = FMA(T1M, T1N, T1K); T8P = T1I * T1N; T1T = T1R * T1S; T8V = T1R * T1W; T1P = T1G + T1O; T8U = T1G - T1O; T8Q = FNMS(T1M, T1J, T8P); T1X = FMA(T1V, T1W, T1T); T8W = FNMS(T1V, T1S, T8V); T3P = T3N * T3O; } Tga = T8O + T8Q; T8R = T8O - T8Q; { E T26, T8S, Tgb, T8Z; T26 = T1X + T25; T8S = T1X - T25; Tgb = T8W + T8Y; T8Z = T8W - T8Y; T8T = T8R + T8S; TdY = T8R - T8S; T27 = T1P + T26; Tg9 = T1P - T26; T90 = T8U - T8Z; TdX = T8U + T8Z; Tgc = Tga - Tgb; TiJ = Tga + Tgb; T3S = ii[WS(rs, 62)]; } } T4d = ri[WS(rs, 46)]; T4h = ii[WS(rs, 46)]; T3V = ri[WS(rs, 30)]; T3T = FMA(T3R, T3S, T3P); T9S = T3N * T3S; T4e = T4c * T4d; Tai = T4c * T4h; T3W = T3U * T3V; T9T = FNMS(T3R, T3O, T9S); T4i = FMA(T4g, T4h, T4e); Taj = FNMS(T4g, T4d, Tai); T3Y = ii[WS(rs, 30)]; T42 = ri[WS(rs, 14)]; T45 = ii[WS(rs, 14)]; } { E T4P, T4Q, T4R, T56, T5a, Tat, T57, Tbp; { E T40, Taf, T9V, T46, Tah, Tgw, T9W; T4P = ri[WS(rs, 1)]; { E T3Z, T9U, T43, Tag; T3Z = FMA(T3X, T3Y, T3W); T9U = T3U * T3Y; T43 = T41 * T42; Tag = T41 * T45; T40 = T3T + T3Z; Taf = T3T - T3Z; T9V = FNMS(T3X, T3V, T9U); T46 = FMA(T44, T45, T43); Tah = FNMS(T44, T42, Tag); T4Q = T2 * T4P; } Tgw = T9T + T9V; T9W = T9T - T9V; { E T4j, T9X, Tgx, Tak; T4j = T46 + T4i; T9X = T46 - T4i; Tgx = Tah + Taj; Tak = Tah - Taj; T9Y = T9W + T9X; Tec = T9W - T9X; T4k = T40 + T4j; TgB = T40 - T4j; Tal = Taf - Tak; Tef = Taf + Tak; Tgy = Tgw - Tgx; TiT = Tgw + Tgx; T4R = ii[WS(rs, 1)]; } } T56 = ri[WS(rs, 49)]; T5a = ii[WS(rs, 49)]; T4U = ri[WS(rs, 33)]; T4S = FMA(T5, T4R, T4Q); Tat = T2 * T4R; T57 = T55 * T56; Tbp = T55 * T5a; T4V = T4T * T4U; Tau = FNMS(T5, T4P, Tat); T5b = FMA(T59, T5a, T57); Tbq = FNMS(T59, T56, Tbp); T4X = ii[WS(rs, 33)]; T50 = ri[WS(rs, 17)]; T52 = ii[WS(rs, 17)]; } } } } { E T7a, T78, T7b, TbL, T7t, TbU, T7d, T7i, T7m; { E T4q, T4o, T4r, Ta1, T4J, Taa, T4t, T4y, T4C; { E T3o, T3f, T3p, T9s, T3I, T9B, T3s, T3x, T3B; { E T3b, T3c, T3e, T3E, T3H, T9r, T3F, T9A; { E T4Z, Tbm, Taw, T53, Tbo, TgJ, Tax; T3b = ri[WS(rs, 10)]; { E T4Y, Tav, T51, Tbn; T4Y = FMA(T4W, T4X, T4V); Tav = T4T * T4X; T51 = T48 * T50; Tbn = T48 * T52; T4Z = T4S + T4Y; Tbm = T4S - T4Y; Taw = FNMS(T4W, T4U, Tav); T53 = FMA(T4b, T52, T51); Tbo = FNMS(T4b, T50, Tbn); T3c = T3a * T3b; } TgJ = Tau + Taw; Tax = Tau - Taw; { E T5c, Tay, TgK, Tbr; T5c = T53 + T5b; Tay = T53 - T5b; TgK = Tbo + Tbq; Tbr = Tbo - Tbq; Taz = Tax + Tay; Tel = Tax - Tay; T5d = T4Z + T5c; Th0 = T4Z - T5c; Tbs = Tbm - Tbr; Tew = Tbm + Tbr; TgL = TgJ - TgK; TiZ = TgJ + TgK; T3e = ii[WS(rs, 10)]; } } T3E = ri[WS(rs, 26)]; T3H = ii[WS(rs, 26)]; T3o = ri[WS(rs, 42)]; T3f = FMA(T3d, T3e, T3c); T9r = T3a * T3e; T3F = T3D * T3E; T9A = T3D * T3H; T3p = T3n * T3o; T9s = FNMS(T3d, T3b, T9r); T3I = FMA(T3G, T3H, T3F); T9B = FNMS(T3G, T3E, T9A); T3s = ii[WS(rs, 42)]; T3x = ri[WS(rs, 58)]; T3B = ii[WS(rs, 58)]; } { E T4l, T4m, T4n, T4F, T4I, Ta0, T4G, Ta9; { E T3u, T9q, T9u, T3C, T9z, Tgr, T9v; T4l = ri[WS(rs, 6)]; { E T3t, T9t, T3y, T9y; T3t = FMA(T3r, T3s, T3p); T9t = T3n * T3s; T3y = T3w * T3x; T9y = T3w * T3B; T3u = T3f + T3t; T9q = T3f - T3t; T9u = FNMS(T3r, T3o, T9t); T3C = FMA(T3A, T3B, T3y); T9z = FNMS(T3A, T3x, T9y); T4m = T3g * T4l; } Tgr = T9s + T9u; T9v = T9s - T9u; { E T3J, T9x, Tgs, T9C; T3J = T3C + T3I; T9x = T3C - T3I; Tgs = T9z + T9B; T9C = T9z - T9B; { E T9w, T9O, T9D, T9N; T9w = T9q + T9v; T9O = T9v - T9q; T3K = T3u + T3J; Tgo = T3J - T3u; T9D = T9x - T9C; T9N = T9x + T9C; Tgt = Tgr - Tgs; TiO = Tgr + Tgs; T9P = T9N - T9O; Te6 = T9O + T9N; T9E = T9w - T9D; Te9 = T9w + T9D; T4n = ii[WS(rs, 6)]; } } } T4F = ri[WS(rs, 22)]; T4I = ii[WS(rs, 22)]; T4q = ri[WS(rs, 38)]; T4o = FMA(T3i, T4n, T4m); Ta0 = T3g * T4n; T4G = T4E * T4F; Ta9 = T4E * T4I; T4r = T4p * T4q; Ta1 = FNMS(T3i, T4l, Ta0); T4J = FMA(T4H, T4I, T4G); Taa = FNMS(T4H, T4F, Ta9); T4t = ii[WS(rs, 38)]; T4y = ri[WS(rs, 54)]; T4C = ii[WS(rs, 54)]; } } { E T5k, T5h, T5l, TaC, T5G, TaL, T5o, T5t, T5x; { E T5e, T5f, T5g, T5B, T5F, TaB, T5C, TaK; { E T4v, T9Z, Ta3, T4D, Ta8, TgC, Ta4; T5e = ri[WS(rs, 9)]; { E T4u, Ta2, T4z, Ta7; T4u = FMA(T4s, T4t, T4r); Ta2 = T4p * T4t; T4z = T4x * T4y; Ta7 = T4x * T4C; T4v = T4o + T4u; T9Z = T4o - T4u; Ta3 = FNMS(T4s, T4q, Ta2); T4D = FMA(T4B, T4C, T4z); Ta8 = FNMS(T4B, T4y, Ta7); T5f = T8 * T5e; } TgC = Ta1 + Ta3; Ta4 = Ta1 - Ta3; { E T4K, Ta6, TgD, Tab; T4K = T4D + T4J; Ta6 = T4D - T4J; TgD = Ta8 + Taa; Tab = Ta8 - Taa; { E Ta5, Tan, Tac, Tam; Ta5 = T9Z + Ta4; Tan = Ta4 - T9Z; T4L = T4v + T4K; Tgz = T4K - T4v; Tac = Ta6 - Tab; Tam = Ta6 + Tab; TgE = TgC - TgD; TiU = TgC + TgD; Tao = Tam - Tan; Ted = Tan + Tam; Tad = Ta5 - Tac; Teg = Ta5 + Tac; T5g = ii[WS(rs, 9)]; } } } T5B = ri[WS(rs, 25)]; T5F = ii[WS(rs, 25)]; T5k = ri[WS(rs, 41)]; T5h = FMA(Tc, T5g, T5f); TaB = T8 * T5g; T5C = T5A * T5B; TaK = T5A * T5F; T5l = T5j * T5k; TaC = FNMS(Tc, T5e, TaB); T5G = FMA(T5E, T5F, T5C); TaL = FNMS(T5E, T5B, TaK); T5o = ii[WS(rs, 41)]; T5t = ri[WS(rs, 57)]; T5x = ii[WS(rs, 57)]; } { E T75, T76, T77, T7p, T7s, TbK, T7q, TbT; { E T5q, TaA, TaE, T5y, TaJ, Th1, TaF; T75 = ri[WS(rs, 7)]; { E T5p, TaD, T5u, TaI; T5p = FMA(T5n, T5o, T5l); TaD = T5j * T5o; T5u = T5s * T5t; TaI = T5s * T5x; T5q = T5h + T5p; TaA = T5h - T5p; TaE = FNMS(T5n, T5k, TaD); T5y = FMA(T5w, T5x, T5u); TaJ = FNMS(T5w, T5t, TaI); T76 = T1i * T75; } Th1 = TaC + TaE; TaF = TaC - TaE; { E T5H, TaH, Th2, TaM; T5H = T5y + T5G; TaH = T5y - T5G; Th2 = TaJ + TaL; TaM = TaJ - TaL; { E TaG, Tbu, TaN, Tbt; TaG = TaA + TaF; Tbu = TaF - TaA; T5I = T5q + T5H; TgM = T5H - T5q; TaN = TaH - TaM; Tbt = TaH + TaM; Th3 = Th1 - Th2; Tj0 = Th1 + Th2; Tbv = Tbt - Tbu; Tem = Tbu + Tbt; TaO = TaG - TaN; Tex = TaG + TaN; T77 = ii[WS(rs, 7)]; } } } T7p = ri[WS(rs, 23)]; T7s = ii[WS(rs, 23)]; T7a = ri[WS(rs, 39)]; T78 = FMA(T1k, T77, T76); TbK = T1i * T77; T7q = T7o * T7p; TbT = T7o * T7s; T7b = T79 * T7a; TbL = FNMS(T1k, T75, TbK); T7t = FMA(T7r, T7s, T7q); TbU = FNMS(T7r, T7p, TbT); T7d = ii[WS(rs, 39)]; T7i = ri[WS(rs, 55)]; T7m = ii[WS(rs, 55)]; } } } { E T6i, T6g, T6j, TaY, T6z, TaU, T6l, T6o, T6q; { E T5P, T5N, T5Q, Tbd, T66, Tb9, T5S, T5V, T5X; { E T5K, T5L, T5M, T61, T65, Tbc, T62, Tb8; { E T7f, TbJ, TbN, T7n, TbS, Ths, TbO; T5K = ri[WS(rs, 5)]; { E T7e, TbM, T7j, TbR; T7e = FMA(T7c, T7d, T7b); TbM = T79 * T7d; T7j = T7h * T7i; TbR = T7h * T7m; T7f = T78 + T7e; TbJ = T78 - T7e; TbN = FNMS(T7c, T7a, TbM); T7n = FMA(T7l, T7m, T7j); TbS = FNMS(T7l, T7i, TbR); T5L = Td * T5K; } Ths = TbL + TbN; TbO = TbL - TbN; { E T7u, TbQ, Tht, TbV; T7u = T7n + T7t; TbQ = T7n - T7t; Tht = TbS + TbU; TbV = TbS - TbU; { E TbP, TcD, TbW, TcC; TbP = TbJ + TbO; TcD = TbO - TbJ; T7v = T7f + T7u; Thd = T7u - T7f; TbW = TbQ - TbV; TcC = TbQ + TbV; Thu = Ths - Tht; Tjb = Ths + Tht; TcE = TcC - TcD; TeF = TcD + TcC; TbX = TbP - TbW; TeQ = TbP + TbW; T5M = ii[WS(rs, 5)]; } } } T61 = ri[WS(rs, 53)]; T65 = ii[WS(rs, 53)]; T5P = ri[WS(rs, 37)]; T5N = FMA(Th, T5M, T5L); Tbc = Td * T5M; T62 = T60 * T61; Tb8 = T60 * T65; T5Q = T5O * T5P; Tbd = FNMS(Th, T5K, Tbc); T66 = FMA(T64, T65, T62); Tb9 = FNMS(T64, T61, Tb8); T5S = ii[WS(rs, 37)]; T5V = ri[WS(rs, 21)]; T5X = ii[WS(rs, 21)]; } { E T6b, T6c, T6f, T6u, T6y, TaX, T6v, TaT; { E T5U, Tb5, Tbf, T5Y, Tb7; T6b = ri[WS(rs, 61)]; { E T5T, Tbe, T5W, Tb6; T5T = FMA(T5R, T5S, T5Q); Tbe = T5O * T5S; T5W = T3j * T5V; Tb6 = T3j * T5X; T5U = T5N + T5T; Tb5 = T5N - T5T; Tbf = FNMS(T5R, T5P, Tbe); T5Y = FMA(T3m, T5X, T5W); Tb7 = FNMS(T3m, T5V, Tb6); T6c = T6a * T6b; } { E TgO, Tbg, T67, Tbh; TgO = Tbd + Tbf; Tbg = Tbd - Tbf; T67 = T5Y + T66; Tbh = T5Y - T66; { E TgP, Tba, Tbi, Teo; TgP = Tb7 + Tb9; Tba = Tb7 - Tb9; Tbi = Tbg + Tbh; Teo = Tbg - Tbh; { E TgR, Tbb, Tep, TgQ; TgR = T5U - T67; T68 = T5U + T67; Tbb = Tb5 - Tba; Tep = Tb5 + Tba; TgQ = TgO - TgP; Tj5 = TgO + TgP; Tez = FMA(KP414213562, Teo, Tep); Teq = FNMS(KP414213562, Tep, Teo); Tbj = FNMS(KP414213562, Tbi, Tbb); Tbx = FMA(KP414213562, Tbb, Tbi); TgS = TgQ - TgR; Th5 = TgR + TgQ; T6f = ii[WS(rs, 61)]; } } } } T6u = ri[WS(rs, 45)]; T6y = ii[WS(rs, 45)]; T6i = ri[WS(rs, 29)]; T6g = FMA(T6e, T6f, T6c); TaX = T6a * T6f; T6v = T6t * T6u; TaT = T6t * T6y; T6j = T6h * T6i; TaY = FNMS(T6e, T6b, TaX); T6z = FMA(T6x, T6y, T6v); TaU = FNMS(T6x, T6u, TaT); T6l = ii[WS(rs, 29)]; T6o = ri[WS(rs, 13)]; T6q = ii[WS(rs, 13)]; } } { E T7C, T7A, T7D, Tcm, T7T, Tci, T7F, T7I, T7K; { E T7x, T7y, T7z, T7O, T7S, Tcl, T7P, Tch; { E T6n, TaQ, Tb0, T6r, TaS; T7x = ri[WS(rs, 3)]; { E T6m, TaZ, T6p, TaR; T6m = FMA(T6k, T6l, T6j); TaZ = T6h * T6l; T6p = T17 * T6o; TaR = T17 * T6q; T6n = T6g + T6m; TaQ = T6g - T6m; Tb0 = FNMS(T6k, T6i, TaZ); T6r = FMA(T19, T6q, T6p); TaS = FNMS(T19, T6o, TaR); T7y = T3 * T7x; } { E TgU, Tb1, T6A, Tb2; TgU = TaY + Tb0; Tb1 = TaY - Tb0; T6A = T6r + T6z; Tb2 = T6r - T6z; { E TgV, TaV, Tb3, Ter; TgV = TaS + TaU; TaV = TaS - TaU; Tb3 = Tb1 + Tb2; Ter = Tb1 - Tb2; { E TgT, TaW, Tes, TgW; TgT = T6n - T6A; T6B = T6n + T6A; TaW = TaQ - TaV; Tes = TaQ + TaV; TgW = TgU - TgV; Tj6 = TgU + TgV; TeA = FNMS(KP414213562, Ter, Tes); Tet = FMA(KP414213562, Tes, Ter); Tb4 = FMA(KP414213562, Tb3, TaW); Tby = FNMS(KP414213562, TaW, Tb3); TgX = TgT + TgW; Th6 = TgT - TgW; T7z = ii[WS(rs, 3)]; } } } } T7O = ri[WS(rs, 51)]; T7S = ii[WS(rs, 51)]; T7C = ri[WS(rs, 35)]; T7A = FMA(T6, T7z, T7y); Tcl = T3 * T7z; T7P = T7N * T7O; Tch = T7N * T7S; T7D = T7B * T7C; Tcm = FNMS(T6, T7x, Tcl); T7T = FMA(T7R, T7S, T7P); Tci = FNMS(T7R, T7O, Tch); T7F = ii[WS(rs, 35)]; T7I = ri[WS(rs, 19)]; T7K = ii[WS(rs, 19)]; } { E T7Y, T7Z, T82, T8f, T8j, Tc6, T8g, Tc2; { E T7H, Tce, Tco, T7L, Tcg; T7Y = ri[WS(rs, 59)]; { E T7G, Tcn, T7J, Tcf; T7G = FMA(T7E, T7F, T7D); Tcn = T7B * T7F; T7J = T2u * T7I; Tcf = T2u * T7K; T7H = T7A + T7G; Tce = T7A - T7G; Tco = FNMS(T7E, T7C, Tcn); T7L = FMA(T2x, T7K, T7J); Tcg = FNMS(T2x, T7I, Tcf); T7Z = T7X * T7Y; } { E Thf, Tcp, T7U, Tcq; Thf = Tcm + Tco; Tcp = Tcm - Tco; T7U = T7L + T7T; Tcq = T7L - T7T; { E Thg, Tcj, Tcr, TeH; Thg = Tcg + Tci; Tcj = Tcg - Tci; Tcr = Tcp + Tcq; TeH = Tcp - Tcq; { E Thi, Tck, TeI, Thh; Thi = T7H - T7U; T7V = T7H + T7U; Tck = Tce - Tcj; TeI = Tce + Tcj; Thh = Thf - Thg; Tjg = Thf + Thg; TeS = FMA(KP414213562, TeH, TeI); TeJ = FNMS(KP414213562, TeI, TeH); Tcs = FNMS(KP414213562, Tcr, Tck); TcG = FMA(KP414213562, Tck, Tcr); Thj = Thh - Thi; Thw = Thi + Thh; T82 = ii[WS(rs, 59)]; } } } } T8f = ri[WS(rs, 43)]; T8j = ii[WS(rs, 43)]; T84 = ri[WS(rs, 27)]; T83 = FMA(T81, T82, T7Z); Tc6 = T7X * T82; T8g = T8e * T8f; Tc2 = T8e * T8j; T85 = Te * T84; Tc7 = FNMS(T81, T7Y, Tc6); T8k = FMA(T8i, T8j, T8g); Tc3 = FNMS(T8i, T8f, Tc2); T86 = ii[WS(rs, 27)]; T89 = ri[WS(rs, 11)]; T8b = ii[WS(rs, 11)]; } } } } } { E TeT, TeM, Tcd, TcH, Tho, Thx, Tkw, Tkv, Tl6, Tl5; { E TiI, Tkp, TiQ, TiS, TiL, Tkq, TiP, TiV, Tjf, Tjd, Tjc, Tji, Tj4, Tj2, Tj1; E Tj7, Tkh, Tki; { E TjG, T2I, Tkj, T4N, Tkk, Tkf, Tk5, TjJ, T8o, Tk2, TjL, T6D, TjY, TjU, Tk1; E TjO; { E T8m, Tjh, T3L, T4M, Tk6, Tke, TjH, TjI; { E T1C, T88, TbZ, Tc9, T8c, Tc1, T2H; T1C = TY + T1B; TiI = TY - T1B; { E T87, Tc8, T8a, Tc0; T87 = FMA(Ti, T86, T85); Tc8 = Te * T86; T8a = Tu * T89; Tc0 = Tu * T8b; T88 = T83 + T87; TbZ = T83 - T87; Tc9 = FNMS(Ti, T84, Tc8); T8c = FMA(Tx, T8b, T8a); Tc1 = FNMS(Tx, T89, Tc0); T2H = T27 + T2G; Tkp = T2G - T27; } { E Thl, Tca, T8l, Tcb; Thl = Tc7 + Tc9; Tca = Tc7 - Tc9; T8l = T8c + T8k; Tcb = T8c - T8k; { E Thm, Tc4, Tcc, TeK; Thm = Tc1 + Tc3; Tc4 = Tc1 - Tc3; Tcc = Tca + Tcb; TeK = Tca - Tcb; { E Thk, Tc5, TeL, Thn; Thk = T88 - T8l; T8m = T88 + T8l; Tc5 = TbZ - Tc4; TeL = TbZ + Tc4; Thn = Thl - Thm; Tjh = Thl + Thm; TeT = FNMS(KP414213562, TeK, TeL); TeM = FMA(KP414213562, TeL, TeK); Tcd = FMA(KP414213562, Tcc, Tc5); TcH = FNMS(KP414213562, Tc5, Tcc); Tho = Thk + Thn; Thx = Thk - Thn; TjG = T1C - T2H; T2I = T1C + T2H; } } } } TiQ = T39 - T3K; T3L = T39 + T3K; T4M = T4k + T4L; TiS = T4k - T4L; TiL = TiJ - TiK; Tk6 = TiJ + TiK; Tke = Tk7 + Tkd; Tkq = Tkd - Tk7; TiP = TiN - TiO; TjH = TiN + TiO; Tkj = T4M - T3L; T4N = T3L + T4M; Tkk = Tke - Tk6; Tkf = Tk6 + Tke; TjI = TiT + TiU; TiV = TiT - TiU; { E TjR, TjQ, TjS, T7w, T8n; Tjf = T74 - T7v; T7w = T74 + T7v; T8n = T7V + T8m; Tjd = T8m - T7V; Tjc = Tja - Tjb; TjR = Tja + Tjb; Tk5 = TjH + TjI; TjJ = TjH - TjI; TjQ = T7w - T8n; T8o = T7w + T8n; Tji = Tjg - Tjh; TjS = Tjg + Tjh; { E TjM, TjN, T5J, T6C, TjT; Tj4 = T5d - T5I; T5J = T5d + T5I; T6C = T68 + T6B; Tj2 = T6B - T68; TjT = TjR - TjS; Tk2 = TjR + TjS; Tj1 = TiZ - Tj0; TjM = TiZ + Tj0; TjL = T5J - T6C; T6D = T5J + T6C; Tj7 = Tj5 - Tj6; TjN = Tj5 + Tj6; TjY = TjQ + TjT; TjU = TjQ - TjT; Tk1 = TjM + TjN; TjO = TjM - TjN; } } } { E Tk0, Tk3, TjW, Tko, Tkn, Tkl, Tkm, TjZ; { E TjP, TjX, Tk4, Tkg, T4O, T8p, TjK, TjV; Tk0 = T2I - T4N; T4O = T2I + T4N; T8p = T6D + T8o; Tkh = T8o - T6D; TjP = TjL + TjO; TjX = TjO - TjL; Tk3 = Tk1 - Tk2; Tk4 = Tk1 + Tk2; ri[0] = T4O + T8p; ri[WS(rs, 32)] = T4O - T8p; Tkg = Tk5 + Tkf; Tki = Tkf - Tk5; TjW = TjG - TjJ; TjK = TjG + TjJ; TjV = TjP + TjU; Tko = TjU - TjP; Tkn = Tkk - Tkj; Tkl = Tkj + Tkk; ii[WS(rs, 32)] = Tkg - Tk4; ii[0] = Tk4 + Tkg; ri[WS(rs, 8)] = FMA(KP707106781, TjV, TjK); ri[WS(rs, 40)] = FNMS(KP707106781, TjV, TjK); Tkm = TjX + TjY; TjZ = TjX - TjY; } ii[WS(rs, 40)] = FNMS(KP707106781, Tkm, Tkl); ii[WS(rs, 8)] = FMA(KP707106781, Tkm, Tkl); ri[WS(rs, 24)] = FMA(KP707106781, TjZ, TjW); ri[WS(rs, 56)] = FNMS(KP707106781, TjZ, TjW); ii[WS(rs, 56)] = FNMS(KP707106781, Tko, Tkn); ii[WS(rs, 24)] = FMA(KP707106781, Tko, Tkn); ri[WS(rs, 16)] = Tk0 + Tk3; ri[WS(rs, 48)] = Tk0 - Tk3; } } { E Tjq, TiM, Tkx, Tkr, Tjt, Tky, Tks, TiX, Tjz, Tje, Tjx, TjD, Tjn, Tj9, Tjr; E TiR; ii[WS(rs, 48)] = Tki - Tkh; ii[WS(rs, 16)] = Tkh + Tki; Tjq = TiI + TiL; TiM = TiI - TiL; Tkx = Tkq - Tkp; Tkr = Tkp + Tkq; Tjr = TiQ + TiP; TiR = TiP - TiQ; { E Tjw, Tj3, Tjs, TiW, Tjv, Tj8; Tjs = TiS - TiV; TiW = TiS + TiV; Tjw = Tj1 + Tj2; Tj3 = Tj1 - Tj2; Tjt = Tjr + Tjs; Tky = Tjs - Tjr; Tks = TiR + TiW; TiX = TiR - TiW; Tjv = Tj4 + Tj7; Tj8 = Tj4 - Tj7; Tjz = Tjc + Tjd; Tje = Tjc - Tjd; Tjx = FMA(KP414213562, Tjw, Tjv); TjD = FNMS(KP414213562, Tjv, Tjw); Tjn = FNMS(KP414213562, Tj3, Tj8); Tj9 = FMA(KP414213562, Tj8, Tj3); } { E Tjm, TiY, Tkz, TkB, Tjy, Tjj; Tjm = FNMS(KP707106781, TiX, TiM); TiY = FMA(KP707106781, TiX, TiM); Tkz = FMA(KP707106781, Tky, Tkx); TkB = FNMS(KP707106781, Tky, Tkx); Tjy = Tjf + Tji; Tjj = Tjf - Tji; { E TjC, Tkt, Tku, TjF; { E Tju, TjE, Tjo, Tjk, TjB, TjA; TjC = FNMS(KP707106781, Tjt, Tjq); Tju = FMA(KP707106781, Tjt, Tjq); TjA = FNMS(KP414213562, Tjz, Tjy); TjE = FMA(KP414213562, Tjy, Tjz); Tjo = FMA(KP414213562, Tje, Tjj); Tjk = FNMS(KP414213562, Tjj, Tje); TjB = Tjx + TjA; Tkw = TjA - Tjx; Tkv = FNMS(KP707106781, Tks, Tkr); Tkt = FMA(KP707106781, Tks, Tkr); { E Tjp, TkA, TkC, Tjl; Tjp = Tjn + Tjo; TkA = Tjo - Tjn; TkC = Tj9 + Tjk; Tjl = Tj9 - Tjk; ri[WS(rs, 4)] = FMA(KP923879532, TjB, Tju); ri[WS(rs, 36)] = FNMS(KP923879532, TjB, Tju); ri[WS(rs, 60)] = FMA(KP923879532, Tjp, Tjm); ri[WS(rs, 28)] = FNMS(KP923879532, Tjp, Tjm); ii[WS(rs, 44)] = FNMS(KP923879532, TkA, Tkz); ii[WS(rs, 12)] = FMA(KP923879532, TkA, Tkz); ii[WS(rs, 60)] = FMA(KP923879532, TkC, TkB); ii[WS(rs, 28)] = FNMS(KP923879532, TkC, TkB); ri[WS(rs, 12)] = FMA(KP923879532, Tjl, TiY); ri[WS(rs, 44)] = FNMS(KP923879532, Tjl, TiY); Tku = TjD + TjE; TjF = TjD - TjE; } } ii[WS(rs, 36)] = FNMS(KP923879532, Tku, Tkt); ii[WS(rs, 4)] = FMA(KP923879532, Tku, Tkt); ri[WS(rs, 20)] = FMA(KP923879532, TjF, TjC); ri[WS(rs, 52)] = FNMS(KP923879532, TjF, TjC); } } } } { E TkV, Tl1, ThG, Tgk, TkH, TkN, Tis, Ti0, Thv, ThJ, TkO, TkI, TgH, Thy, TiC; E TiG, Tiq, Tim, ThN, ThT, ThD, Th9, TkW, Tiv, Tl2, Ti7, ThP, Thq, Tiz, TiF; E Tip, Tif; { E Ti1, Ti2, Ti4, Ti5, Thp, The, Tij, TiB, Tii, Tik; { E ThW, Tg8, TkT, TkF, ThX, ThY, TkU, Tgj, Tgd, Tgi; ThW = Tg4 - Tg7; Tg8 = Tg4 + Tg7; TkT = TkE - TkD; TkF = TkD + TkE; ThX = Tgc - Tg9; Tgd = Tg9 + Tgc; ii[WS(rs, 52)] = FNMS(KP923879532, Tkw, Tkv); ii[WS(rs, 20)] = FMA(KP923879532, Tkw, Tkv); Tgi = Tge - Tgh; ThY = Tge + Tgh; TkU = Tgi - Tgd; Tgj = Tgd + Tgi; { E TgA, ThH, Tgv, TgF; { E Tgp, TkG, ThZ, Tgu; Ti1 = Tgn - Tgo; Tgp = Tgn + Tgo; TkV = FMA(KP707106781, TkU, TkT); Tl1 = FNMS(KP707106781, TkU, TkT); ThG = FMA(KP707106781, Tgj, Tg8); Tgk = FNMS(KP707106781, Tgj, Tg8); TkG = ThX + ThY; ThZ = ThX - ThY; Tgu = Tgq + Tgt; Ti2 = Tgq - Tgt; Ti4 = Tgy - Tgz; TgA = Tgy + Tgz; TkH = FMA(KP707106781, TkG, TkF); TkN = FNMS(KP707106781, TkG, TkF); Tis = FNMS(KP707106781, ThZ, ThW); Ti0 = FMA(KP707106781, ThZ, ThW); ThH = FMA(KP414213562, Tgp, Tgu); Tgv = FNMS(KP414213562, Tgu, Tgp); TgF = TgB + TgE; Ti5 = TgB - TgE; } { E Tig, Tih, ThI, TgG; Thv = Thr + Thu; Tig = Thr - Thu; Tih = Tho - Thj; Thp = Thj + Tho; The = Thc + Thd; Tij = Thc - Thd; ThI = FNMS(KP414213562, TgA, TgF); TgG = FMA(KP414213562, TgF, TgA); TiB = FMA(KP707106781, Tih, Tig); Tii = FNMS(KP707106781, Tih, Tig); ThJ = ThH + ThI; TkO = ThI - ThH; TkI = Tgv + TgG; TgH = Tgv - TgG; Tik = Thw - Thx; Thy = Thw + Thx; } } } { E Tic, Tia, Ti9, Tid, Tit, Ti3; { E Th4, ThM, TgZ, Th7, ThL, Th8; { E TgN, TgY, TiA, Til; Tic = TgL - TgM; TgN = TgL + TgM; TgY = TgS + TgX; Tia = TgX - TgS; Ti9 = Th0 - Th3; Th4 = Th0 + Th3; TiA = FMA(KP707106781, Tik, Tij); Til = FNMS(KP707106781, Tik, Tij); ThM = FMA(KP707106781, TgY, TgN); TgZ = FNMS(KP707106781, TgY, TgN); TiC = FNMS(KP198912367, TiB, TiA); TiG = FMA(KP198912367, TiA, TiB); Tiq = FMA(KP668178637, Tii, Til); Tim = FNMS(KP668178637, Til, Tii); Th7 = Th5 + Th6; Tid = Th5 - Th6; } ThL = FMA(KP707106781, Th7, Th4); Th8 = FNMS(KP707106781, Th7, Th4); Tit = FNMS(KP414213562, Ti1, Ti2); Ti3 = FMA(KP414213562, Ti2, Ti1); ThN = FMA(KP198912367, ThM, ThL); ThT = FNMS(KP198912367, ThL, ThM); ThD = FNMS(KP668178637, TgZ, Th8); Th9 = FMA(KP668178637, Th8, TgZ); } { E Tiy, Tib, Tiu, Ti6, Tix, Tie; Tiu = FMA(KP414213562, Ti4, Ti5); Ti6 = FNMS(KP414213562, Ti5, Ti4); Tiy = FMA(KP707106781, Tia, Ti9); Tib = FNMS(KP707106781, Tia, Ti9); TkW = Tiu - Tit; Tiv = Tit + Tiu; Tl2 = Ti3 + Ti6; Ti7 = Ti3 - Ti6; Tix = FMA(KP707106781, Tid, Tic); Tie = FNMS(KP707106781, Tid, Tic); ThP = FMA(KP707106781, Thp, The); Thq = FNMS(KP707106781, Thp, The); Tiz = FMA(KP198912367, Tiy, Tix); TiF = FNMS(KP198912367, Tix, Tiy); Tip = FNMS(KP668178637, Tib, Tie); Tif = FMA(KP668178637, Tie, Tib); } } } { E TkM, TkL, Tl0, TkZ; { E ThC, TgI, TkP, TkR, ThO, Thz; ThC = FNMS(KP923879532, TgH, Tgk); TgI = FMA(KP923879532, TgH, Tgk); TkP = FMA(KP923879532, TkO, TkN); TkR = FNMS(KP923879532, TkO, TkN); ThO = FMA(KP707106781, Thy, Thv); Thz = FNMS(KP707106781, Thy, Thv); { E ThS, TkJ, TkK, ThV; { E ThK, ThU, ThE, ThA, ThR, ThQ; ThS = FNMS(KP923879532, ThJ, ThG); ThK = FMA(KP923879532, ThJ, ThG); ThQ = FNMS(KP198912367, ThP, ThO); ThU = FMA(KP198912367, ThO, ThP); ThE = FMA(KP668178637, Thq, Thz); ThA = FNMS(KP668178637, Thz, Thq); ThR = ThN + ThQ; TkM = ThQ - ThN; TkL = FNMS(KP923879532, TkI, TkH); TkJ = FMA(KP923879532, TkI, TkH); { E ThF, TkQ, TkS, ThB; ThF = ThD + ThE; TkQ = ThE - ThD; TkS = Th9 + ThA; ThB = Th9 - ThA; ri[WS(rs, 2)] = FMA(KP980785280, ThR, ThK); ri[WS(rs, 34)] = FNMS(KP980785280, ThR, ThK); ri[WS(rs, 58)] = FMA(KP831469612, ThF, ThC); ri[WS(rs, 26)] = FNMS(KP831469612, ThF, ThC); ii[WS(rs, 42)] = FNMS(KP831469612, TkQ, TkP); ii[WS(rs, 10)] = FMA(KP831469612, TkQ, TkP); ii[WS(rs, 58)] = FMA(KP831469612, TkS, TkR); ii[WS(rs, 26)] = FNMS(KP831469612, TkS, TkR); ri[WS(rs, 10)] = FMA(KP831469612, ThB, TgI); ri[WS(rs, 42)] = FNMS(KP831469612, ThB, TgI); TkK = ThT + ThU; ThV = ThT - ThU; } } ii[WS(rs, 34)] = FNMS(KP980785280, TkK, TkJ); ii[WS(rs, 2)] = FMA(KP980785280, TkK, TkJ); ri[WS(rs, 18)] = FMA(KP980785280, ThV, ThS); ri[WS(rs, 50)] = FNMS(KP980785280, ThV, ThS); } } { E Tio, TkX, TkY, Tir, Ti8, Tin; Tio = FNMS(KP923879532, Ti7, Ti0); Ti8 = FMA(KP923879532, Ti7, Ti0); Tin = Tif + Tim; Tl0 = Tim - Tif; TkZ = FNMS(KP923879532, TkW, TkV); TkX = FMA(KP923879532, TkW, TkV); ii[WS(rs, 50)] = FNMS(KP980785280, TkM, TkL); ii[WS(rs, 18)] = FMA(KP980785280, TkM, TkL); ri[WS(rs, 6)] = FMA(KP831469612, Tin, Ti8); ri[WS(rs, 38)] = FNMS(KP831469612, Tin, Ti8); TkY = Tip + Tiq; Tir = Tip - Tiq; ii[WS(rs, 38)] = FNMS(KP831469612, TkY, TkX); ii[WS(rs, 6)] = FMA(KP831469612, TkY, TkX); ri[WS(rs, 22)] = FMA(KP831469612, Tir, Tio); ri[WS(rs, 54)] = FNMS(KP831469612, Tir, Tio); } { E TiE, Tl3, Tl4, TiH, Tiw, TiD; TiE = FMA(KP923879532, Tiv, Tis); Tiw = FNMS(KP923879532, Tiv, Tis); TiD = Tiz - TiC; Tl6 = Tiz + TiC; Tl5 = FMA(KP923879532, Tl2, Tl1); Tl3 = FNMS(KP923879532, Tl2, Tl1); ii[WS(rs, 54)] = FNMS(KP831469612, Tl0, TkZ); ii[WS(rs, 22)] = FMA(KP831469612, Tl0, TkZ); ri[WS(rs, 14)] = FMA(KP980785280, TiD, Tiw); ri[WS(rs, 46)] = FNMS(KP980785280, TiD, Tiw); Tl4 = TiG - TiF; TiH = TiF + TiG; ii[WS(rs, 46)] = FNMS(KP980785280, Tl4, Tl3); ii[WS(rs, 14)] = FMA(KP980785280, Tl4, Tl3); ri[WS(rs, 62)] = FMA(KP980785280, TiH, TiE); ri[WS(rs, 30)] = FNMS(KP980785280, TiH, TiE); } } } { E Tla, TdV, TdO, Tm6, Tm5, TdR; { E TcT, TlO, TlI, Tar, TcX, Td3, TcN, TbB, TdM, TdQ, TdA, Tdw, TdJ, TdP, Tdz; E Tdp, TlW, TdF, Tm2, Tdh, Td7, T91, Td6, T8M, TlT, TlF, Td0, Td4, TcO, TcK; E T9g, Td8; { E Tdb, Tdc, Tde, Tdf, Tdm, Tdk, Tdj, Tdn, TcF, Tct, TbY, Tdt, TdL, Tds, Tdu; E TcI, TdD, Tdd; { E Tae, TcR, T9R, Tap, T9F, T9Q; Tdb = FMA(KP707106781, T9E, T9p); T9F = FNMS(KP707106781, T9E, T9p); T9Q = FNMS(KP707106781, T9P, T9M); Tdc = FMA(KP707106781, T9P, T9M); Tde = FMA(KP707106781, Tad, T9Y); Tae = FNMS(KP707106781, Tad, T9Y); ii[WS(rs, 62)] = FMA(KP980785280, Tl6, Tl5); ii[WS(rs, 30)] = FNMS(KP980785280, Tl6, Tl5); TcR = FMA(KP668178637, T9F, T9Q); T9R = FNMS(KP668178637, T9Q, T9F); Tap = FNMS(KP707106781, Tao, Tal); Tdf = FMA(KP707106781, Tao, Tal); { E Tbw, TcW, Tbl, Tbz; { E TaP, Tbk, TcS, Taq; Tdm = FMA(KP707106781, TaO, Taz); TaP = FNMS(KP707106781, TaO, Taz); Tbk = Tb4 - Tbj; Tdk = Tbj + Tb4; Tdj = FMA(KP707106781, Tbv, Tbs); Tbw = FNMS(KP707106781, Tbv, Tbs); TcS = FNMS(KP668178637, Tae, Tap); Taq = FMA(KP668178637, Tap, Tae); TcW = FMA(KP923879532, Tbk, TaP); Tbl = FNMS(KP923879532, Tbk, TaP); TcT = TcR + TcS; TlO = TcS - TcR; TlI = T9R + Taq; Tar = T9R - Taq; Tbz = Tbx - Tby; Tdn = Tbx + Tby; } { E Tdq, Tdr, TcV, TbA; TcF = FNMS(KP707106781, TcE, TcB); Tdq = FMA(KP707106781, TcE, TcB); Tdr = Tcs + Tcd; Tct = Tcd - Tcs; TbY = FNMS(KP707106781, TbX, TbI); Tdt = FMA(KP707106781, TbX, TbI); TcV = FMA(KP923879532, Tbz, Tbw); TbA = FNMS(KP923879532, Tbz, Tbw); TdL = FMA(KP923879532, Tdr, Tdq); Tds = FNMS(KP923879532, Tdr, Tdq); TcX = FMA(KP303346683, TcW, TcV); Td3 = FNMS(KP303346683, TcV, TcW); TcN = FNMS(KP534511135, Tbl, TbA); TbB = FMA(KP534511135, TbA, Tbl); Tdu = TcG + TcH; TcI = TcG - TcH; } } } { E TdI, Tdl, TdK, Tdv, TdH, Tdo; TdK = FMA(KP923879532, Tdu, Tdt); Tdv = FNMS(KP923879532, Tdu, Tdt); TdI = FMA(KP923879532, Tdk, Tdj); Tdl = FNMS(KP923879532, Tdk, Tdj); TdM = FNMS(KP098491403, TdL, TdK); TdQ = FMA(KP098491403, TdK, TdL); TdA = FMA(KP820678790, Tds, Tdv); Tdw = FNMS(KP820678790, Tdv, Tds); TdH = FMA(KP923879532, Tdn, Tdm); Tdo = FNMS(KP923879532, Tdn, Tdm); TdD = FNMS(KP198912367, Tdb, Tdc); Tdd = FMA(KP198912367, Tdc, Tdb); TdJ = FMA(KP098491403, TdI, TdH); TdP = FNMS(KP098491403, TdH, TdI); Tdz = FNMS(KP820678790, Tdl, Tdo); Tdp = FMA(KP820678790, Tdo, Tdl); } { E TcZ, Tcu, TdE, Tdg; TdE = FMA(KP198912367, Tde, Tdf); Tdg = FNMS(KP198912367, Tdf, Tde); TcZ = FMA(KP923879532, Tct, TbY); Tcu = FNMS(KP923879532, Tct, TbY); TlW = TdE - TdD; TdF = TdD + TdE; Tm2 = Tdd + Tdg; Tdh = Tdd - Tdg; { E T8L, TlE, TcY, TcJ; Tla = T8D + T8K; T8L = T8D - T8K; TlE = TdU - TdT; TdV = TdT + TdU; Td7 = FNMS(KP414213562, T8T, T90); T91 = FMA(KP414213562, T90, T8T); TcY = FMA(KP923879532, TcI, TcF); TcJ = FNMS(KP923879532, TcI, TcF); Td6 = FNMS(KP707106781, T8L, T8w); T8M = FMA(KP707106781, T8L, T8w); TlT = FNMS(KP707106781, TlE, TlD); TlF = FMA(KP707106781, TlE, TlD); Td0 = FNMS(KP303346683, TcZ, TcY); Td4 = FMA(KP303346683, TcY, TcZ); TcO = FMA(KP534511135, Tcu, TcJ); TcK = FNMS(KP534511135, TcJ, Tcu); T9g = FNMS(KP414213562, T9f, T98); Td8 = FMA(KP414213562, T98, T9f); } } } { E Tm1, TlV, TdC, Tda, Td2, TlM, TlL, Td5; { E TlS, TcQ, TlH, TcM, TlR, TcP; { E TcL, Tas, TlP, TlQ, TlN; TlS = TbB + TcK; TcL = TbB - TcK; { E TlU, T9h, TlG, Td9, T9i; TlU = T91 + T9g; T9h = T91 - T9g; TlG = Td8 - Td7; Td9 = Td7 + Td8; Tm1 = FMA(KP923879532, TlU, TlT); TlV = FNMS(KP923879532, TlU, TlT); TcQ = FMA(KP923879532, T9h, T8M); T9i = FNMS(KP923879532, T9h, T8M); TlN = FNMS(KP923879532, TlG, TlF); TlH = FMA(KP923879532, TlG, TlF); TdC = FMA(KP923879532, Td9, Td6); Tda = FNMS(KP923879532, Td9, Td6); Tas = FMA(KP831469612, Tar, T9i); TcM = FNMS(KP831469612, Tar, T9i); } TlR = FNMS(KP831469612, TlO, TlN); TlP = FMA(KP831469612, TlO, TlN); TlQ = TcO - TcN; TcP = TcN + TcO; ri[WS(rs, 11)] = FMA(KP881921264, TcL, Tas); ri[WS(rs, 43)] = FNMS(KP881921264, TcL, Tas); ii[WS(rs, 43)] = FNMS(KP881921264, TlQ, TlP); ii[WS(rs, 11)] = FMA(KP881921264, TlQ, TlP); } { E TcU, Td1, TlJ, TlK; Td2 = FNMS(KP831469612, TcT, TcQ); TcU = FMA(KP831469612, TcT, TcQ); ri[WS(rs, 59)] = FMA(KP881921264, TcP, TcM); ri[WS(rs, 27)] = FNMS(KP881921264, TcP, TcM); ii[WS(rs, 59)] = FMA(KP881921264, TlS, TlR); ii[WS(rs, 27)] = FNMS(KP881921264, TlS, TlR); Td1 = TcX + Td0; TlM = Td0 - TcX; TlL = FNMS(KP831469612, TlI, TlH); TlJ = FMA(KP831469612, TlI, TlH); TlK = Td3 + Td4; Td5 = Td3 - Td4; ri[WS(rs, 3)] = FMA(KP956940335, Td1, TcU); ri[WS(rs, 35)] = FNMS(KP956940335, Td1, TcU); ii[WS(rs, 35)] = FNMS(KP956940335, TlK, TlJ); ii[WS(rs, 3)] = FMA(KP956940335, TlK, TlJ); } } { E Tdy, Tm0, TlZ, TdB; { E Tdi, Tdx, TlX, TlY; Tdy = FNMS(KP980785280, Tdh, Tda); Tdi = FMA(KP980785280, Tdh, Tda); ri[WS(rs, 19)] = FMA(KP956940335, Td5, Td2); ri[WS(rs, 51)] = FNMS(KP956940335, Td5, Td2); ii[WS(rs, 51)] = FNMS(KP956940335, TlM, TlL); ii[WS(rs, 19)] = FMA(KP956940335, TlM, TlL); Tdx = Tdp + Tdw; Tm0 = Tdw - Tdp; TlZ = FNMS(KP980785280, TlW, TlV); TlX = FMA(KP980785280, TlW, TlV); TlY = Tdz + TdA; TdB = Tdz - TdA; ri[WS(rs, 7)] = FMA(KP773010453, Tdx, Tdi); ri[WS(rs, 39)] = FNMS(KP773010453, Tdx, Tdi); ii[WS(rs, 39)] = FNMS(KP773010453, TlY, TlX); ii[WS(rs, 7)] = FMA(KP773010453, TlY, TlX); } { E TdG, TdN, Tm3, Tm4; TdO = FMA(KP980785280, TdF, TdC); TdG = FNMS(KP980785280, TdF, TdC); ri[WS(rs, 23)] = FMA(KP773010453, TdB, Tdy); ri[WS(rs, 55)] = FNMS(KP773010453, TdB, Tdy); ii[WS(rs, 55)] = FNMS(KP773010453, Tm0, TlZ); ii[WS(rs, 23)] = FMA(KP773010453, Tm0, TlZ); TdN = TdJ - TdM; Tm6 = TdJ + TdM; Tm5 = FMA(KP980785280, Tm2, Tm1); Tm3 = FNMS(KP980785280, Tm2, Tm1); Tm4 = TdQ - TdP; TdR = TdP + TdQ; ri[WS(rs, 15)] = FMA(KP995184726, TdN, TdG); ri[WS(rs, 47)] = FNMS(KP995184726, TdN, TdG); ii[WS(rs, 47)] = FNMS(KP995184726, Tm4, Tm3); ii[WS(rs, 15)] = FMA(KP995184726, Tm4, Tm3); } } } } { E Tf5, Tlk, Tle, Tej, Tf9, Tff, TeZ, TeD, TfY, Tg2, TfM, TfI, TfV, Tg1, TfL; E TfB, Tls, TfR, Tly, Tft, Tfj, TdZ, Tfi, TdW, Tlp, Tlb, Tfc, Tfg, Tf0, TeW; E Te2, Tfk; { E Tfn, Tfo, Tfq, Tfr, Tfy, Tfw, Tfv, Tfz, TeR, TeN, TeG, TfF, TfX, TfE, TfG; E TeU, TfP, Tfp; { E Te7, Tea, Tee, Teh; Tfn = FNMS(KP707106781, Te6, Te5); Te7 = FMA(KP707106781, Te6, Te5); ri[WS(rs, 63)] = FMA(KP995184726, TdR, TdO); ri[WS(rs, 31)] = FNMS(KP995184726, TdR, TdO); ii[WS(rs, 63)] = FMA(KP995184726, Tm6, Tm5); ii[WS(rs, 31)] = FNMS(KP995184726, Tm6, Tm5); Tea = FMA(KP707106781, Te9, Te8); Tfo = FNMS(KP707106781, Te9, Te8); Tfq = FNMS(KP707106781, Ted, Tec); Tee = FMA(KP707106781, Ted, Tec); Teh = FMA(KP707106781, Teg, Tef); Tfr = FNMS(KP707106781, Teg, Tef); { E Tey, Tf8, Tev, TeB; { E Ten, Tf3, Teb, Tf4, Tei, Teu; Tfy = FNMS(KP707106781, Tem, Tel); Ten = FMA(KP707106781, Tem, Tel); Tf3 = FMA(KP198912367, Te7, Tea); Teb = FNMS(KP198912367, Tea, Te7); Tf4 = FNMS(KP198912367, Tee, Teh); Tei = FMA(KP198912367, Teh, Tee); Teu = Teq + Tet; Tfw = Tet - Teq; Tfv = FNMS(KP707106781, Tex, Tew); Tey = FMA(KP707106781, Tex, Tew); Tf5 = Tf3 + Tf4; Tlk = Tf4 - Tf3; Tle = Teb + Tei; Tej = Teb - Tei; Tf8 = FMA(KP923879532, Teu, Ten); Tev = FNMS(KP923879532, Teu, Ten); TeB = Tez + TeA; Tfz = Tez - TeA; } { E TfC, TfD, Tf7, TeC; TeR = FMA(KP707106781, TeQ, TeP); TfC = FNMS(KP707106781, TeQ, TeP); TfD = TeM - TeJ; TeN = TeJ + TeM; TeG = FMA(KP707106781, TeF, TeE); TfF = FNMS(KP707106781, TeF, TeE); Tf7 = FMA(KP923879532, TeB, Tey); TeC = FNMS(KP923879532, TeB, Tey); TfX = FMA(KP923879532, TfD, TfC); TfE = FNMS(KP923879532, TfD, TfC); Tf9 = FMA(KP098491403, Tf8, Tf7); Tff = FNMS(KP098491403, Tf7, Tf8); TeZ = FNMS(KP820678790, Tev, TeC); TeD = FMA(KP820678790, TeC, Tev); TfG = TeS - TeT; TeU = TeS + TeT; } } } { E TfU, Tfx, TfW, TfH, TfT, TfA; TfW = FMA(KP923879532, TfG, TfF); TfH = FNMS(KP923879532, TfG, TfF); TfU = FMA(KP923879532, Tfw, Tfv); Tfx = FNMS(KP923879532, Tfw, Tfv); TfY = FNMS(KP303346683, TfX, TfW); Tg2 = FMA(KP303346683, TfW, TfX); TfM = FMA(KP534511135, TfE, TfH); TfI = FNMS(KP534511135, TfH, TfE); TfT = FMA(KP923879532, Tfz, Tfy); TfA = FNMS(KP923879532, Tfz, Tfy); TfP = FNMS(KP668178637, Tfn, Tfo); Tfp = FMA(KP668178637, Tfo, Tfn); TfV = FMA(KP303346683, TfU, TfT); Tg1 = FNMS(KP303346683, TfT, TfU); TfL = FNMS(KP534511135, Tfx, TfA); TfB = FMA(KP534511135, TfA, Tfx); } { E Tfb, TeO, TfQ, Tfs, Tfa, TeV; TfQ = FMA(KP668178637, Tfq, Tfr); Tfs = FNMS(KP668178637, Tfr, Tfq); Tfb = FMA(KP923879532, TeN, TeG); TeO = FNMS(KP923879532, TeN, TeG); Tls = TfQ - TfP; TfR = TfP + TfQ; Tly = Tfp + Tfs; Tft = Tfp - Tfs; Tfj = FNMS(KP414213562, TdX, TdY); TdZ = FMA(KP414213562, TdY, TdX); Tfa = FMA(KP923879532, TeU, TeR); TeV = FNMS(KP923879532, TeU, TeR); Tfi = FNMS(KP707106781, TdV, TdS); TdW = FMA(KP707106781, TdV, TdS); Tlp = FNMS(KP707106781, Tla, Tl9); Tlb = FMA(KP707106781, Tla, Tl9); Tfc = FNMS(KP098491403, Tfb, Tfa); Tfg = FMA(KP098491403, Tfa, Tfb); Tf0 = FMA(KP820678790, TeO, TeV); TeW = FNMS(KP820678790, TeV, TeO); Te2 = FNMS(KP414213562, Te1, Te0); Tfk = FMA(KP414213562, Te0, Te1); } } { E Tlx, Tlr, TfO, Tfm, Tfe, Tli, Tlh, Tfh; { E Tlo, Tf2, Tld, TeY, Tln, Tf1; { E TeX, Tek, Tll, Tlm, Tlj; Tlo = TeD + TeW; TeX = TeD - TeW; { E Tlq, Te3, Tlc, Tfl, Te4; Tlq = Te2 - TdZ; Te3 = TdZ + Te2; Tlc = Tfj + Tfk; Tfl = Tfj - Tfk; Tlx = FNMS(KP923879532, Tlq, Tlp); Tlr = FMA(KP923879532, Tlq, Tlp); Tf2 = FMA(KP923879532, Te3, TdW); Te4 = FNMS(KP923879532, Te3, TdW); Tlj = FNMS(KP923879532, Tlc, Tlb); Tld = FMA(KP923879532, Tlc, Tlb); TfO = FNMS(KP923879532, Tfl, Tfi); Tfm = FMA(KP923879532, Tfl, Tfi); Tek = FMA(KP980785280, Tej, Te4); TeY = FNMS(KP980785280, Tej, Te4); } Tln = FNMS(KP980785280, Tlk, Tlj); Tll = FMA(KP980785280, Tlk, Tlj); Tlm = Tf0 - TeZ; Tf1 = TeZ + Tf0; ri[WS(rs, 9)] = FMA(KP773010453, TeX, Tek); ri[WS(rs, 41)] = FNMS(KP773010453, TeX, Tek); ii[WS(rs, 41)] = FNMS(KP773010453, Tlm, Tll); ii[WS(rs, 9)] = FMA(KP773010453, Tlm, Tll); } { E Tf6, Tfd, Tlf, Tlg; Tfe = FNMS(KP980785280, Tf5, Tf2); Tf6 = FMA(KP980785280, Tf5, Tf2); ri[WS(rs, 57)] = FMA(KP773010453, Tf1, TeY); ri[WS(rs, 25)] = FNMS(KP773010453, Tf1, TeY); ii[WS(rs, 57)] = FMA(KP773010453, Tlo, Tln); ii[WS(rs, 25)] = FNMS(KP773010453, Tlo, Tln); Tfd = Tf9 + Tfc; Tli = Tfc - Tf9; Tlh = FNMS(KP980785280, Tle, Tld); Tlf = FMA(KP980785280, Tle, Tld); Tlg = Tff + Tfg; Tfh = Tff - Tfg; ri[WS(rs, 1)] = FMA(KP995184726, Tfd, Tf6); ri[WS(rs, 33)] = FNMS(KP995184726, Tfd, Tf6); ii[WS(rs, 33)] = FNMS(KP995184726, Tlg, Tlf); ii[WS(rs, 1)] = FMA(KP995184726, Tlg, Tlf); } } { E TfK, Tlw, Tlv, TfN; { E Tfu, TfJ, Tlt, Tlu; TfK = FNMS(KP831469612, Tft, Tfm); Tfu = FMA(KP831469612, Tft, Tfm); ri[WS(rs, 17)] = FMA(KP995184726, Tfh, Tfe); ri[WS(rs, 49)] = FNMS(KP995184726, Tfh, Tfe); ii[WS(rs, 49)] = FNMS(KP995184726, Tli, Tlh); ii[WS(rs, 17)] = FMA(KP995184726, Tli, Tlh); TfJ = TfB + TfI; Tlw = TfI - TfB; Tlv = FNMS(KP831469612, Tls, Tlr); Tlt = FMA(KP831469612, Tls, Tlr); Tlu = TfL + TfM; TfN = TfL - TfM; ri[WS(rs, 5)] = FMA(KP881921264, TfJ, Tfu); ri[WS(rs, 37)] = FNMS(KP881921264, TfJ, Tfu); ii[WS(rs, 37)] = FNMS(KP881921264, Tlu, Tlt); ii[WS(rs, 5)] = FMA(KP881921264, Tlu, Tlt); } { E TfS, TfZ, Tlz, TlA; Tg0 = FMA(KP831469612, TfR, TfO); TfS = FNMS(KP831469612, TfR, TfO); ri[WS(rs, 21)] = FMA(KP881921264, TfN, TfK); ri[WS(rs, 53)] = FNMS(KP881921264, TfN, TfK); ii[WS(rs, 53)] = FNMS(KP881921264, Tlw, Tlv); ii[WS(rs, 21)] = FMA(KP881921264, Tlw, Tlv); TfZ = TfV - TfY; TlC = TfV + TfY; TlB = FMA(KP831469612, Tly, Tlx); Tlz = FNMS(KP831469612, Tly, Tlx); TlA = Tg2 - Tg1; Tg3 = Tg1 + Tg2; ri[WS(rs, 13)] = FMA(KP956940335, TfZ, TfS); ri[WS(rs, 45)] = FNMS(KP956940335, TfZ, TfS); ii[WS(rs, 45)] = FNMS(KP956940335, TlA, Tlz); ii[WS(rs, 13)] = FMA(KP956940335, TlA, Tlz); } } } } } } } } ri[WS(rs, 61)] = FMA(KP956940335, Tg3, Tg0); ri[WS(rs, 29)] = FNMS(KP956940335, Tg3, Tg0); ii[WS(rs, 61)] = FMA(KP956940335, TlC, TlB); ii[WS(rs, 29)] = FNMS(KP956940335, TlC, TlB); } } } static const tw_instr twinstr[] = { {TW_CEXP, 0, 1}, {TW_CEXP, 0, 3}, {TW_CEXP, 0, 9}, {TW_CEXP, 0, 27}, {TW_CEXP, 0, 63}, {TW_NEXT, 1, 0} }; static const ct_desc desc = { 64, "t2_64", twinstr, &GENUS, {520, 206, 634, 0}, 0, 0, 0 }; void X(codelet_t2_64) (planner *p) { X(kdft_dit_register) (p, t2_64, &desc); } #else /* HAVE_FMA */ /* Generated by: ../../../genfft/gen_twiddle.native -compact -variables 4 -pipeline-latency 4 -twiddle-log3 -precompute-twiddles -n 64 -name t2_64 -include t.h */ /* * This function contains 1154 FP additions, 660 FP multiplications, * (or, 880 additions, 386 multiplications, 274 fused multiply/add), * 302 stack variables, 15 constants, and 256 memory accesses */ #include "t.h" static void t2_64(R *ri, R *ii, const R *W, stride rs, INT mb, INT me, INT ms) { DK(KP471396736, +0.471396736825997648556387625905254377657460319); DK(KP881921264, +0.881921264348355029712756863660388349508442621); DK(KP290284677, +0.290284677254462367636192375817395274691476278); DK(KP956940335, +0.956940335732208864935797886980269969482849206); DK(KP634393284, +0.634393284163645498215171613225493370675687095); DK(KP773010453, +0.773010453362736960810906609758469800971041293); DK(KP098017140, +0.098017140329560601994195563888641845861136673); DK(KP995184726, +0.995184726672196886244836953109479921575474869); DK(KP555570233, +0.555570233019602224742830813948532874374937191); DK(KP831469612, +0.831469612302545237078788377617905756738560812); DK(KP980785280, +0.980785280403230449126182236134239036973933731); DK(KP195090322, +0.195090322016128267848284868477022240927691618); DK(KP923879532, +0.923879532511286756128183189396788286822416626); DK(KP382683432, +0.382683432365089771728459984030398866761344562); DK(KP707106781, +0.707106781186547524400844362104849039284835938); { INT m; for (m = mb, W = W + (mb * 10); m < me; m = m + 1, ri = ri + ms, ii = ii + ms, W = W + 10, MAKE_VOLATILE_STRIDE(128, rs)) { E T2, T5, T3, T6, Te, T9, TP, T3e, T1e, T39, T3c, TT, T1a, T37, T8; E Tw, Td, Ty, Tm, Th, T1C, T3K, T1V, T3x, T3I, T1G, T1R, T3v, T2m, T2q; E T5Y, T6u, T53, T5B, T62, T6w, T57, T5D, T2V, T2X, Tg, TE, T3Y, T3V, T3j; E Tl, TA, T3g, T1j, T1t, TV, T2C, T2z, T1u, TZ, T1h, To, T1p, T6j, T6H; E Ts, T1l, T6l, T6F, T2P, T4b, T4x, T5i, T2R, T49, T4z, T5g, TG, T4k, T4m; E TK, T21, T3O, T3Q, T25, TW, T10, T11, T79, T6X, T5M, T6b, T1v, T30, T69; E T77, T13, T2F, T2D, T6p, T6O, T1x, T2a, T2f, T6V, T28, T6r, T2h, T6Q, T32; E T5K, T5w, T4G, T4Q, T3m, T4h, T4I, T5y, T3k, T4f, T41, T4S, T4Y, T3q, T3D; E T3F, T5r, T3s, T4W, T3Z, T5p; { E Ta, Tj, Tx, TC, Tf, Tk, Tz, TD, T1B, T1E, T2o, T2l, T1T, T1Q, T1A; E T1F, T2p, T2k, T1U, T1P; { E T4, T1d, T19, Tb, T1c, T7, Tc, T18, TR, TO, TS, TN; T2 = W[0]; T5 = W[1]; T3 = W[2]; T6 = W[3]; Te = W[5]; T9 = W[4]; T4 = T2 * T3; T1d = T5 * T9; T19 = T5 * Te; Tb = T2 * T6; T1c = T2 * Te; T7 = T5 * T6; Tc = T5 * T3; T18 = T2 * T9; TR = T3 * Te; TO = T6 * Te; TS = T6 * T9; TN = T3 * T9; TP = TN - TO; T3e = TR - TS; T1e = T1c - T1d; T39 = T1c + T1d; T3c = TN + TO; TT = TR + TS; T1a = T18 + T19; T37 = T18 - T19; T8 = T4 - T7; Ta = T8 * T9; Tj = T8 * Te; Tw = T4 + T7; Tx = Tw * T9; TC = Tw * Te; Td = Tb + Tc; Tf = Td * Te; Tk = Td * T9; Ty = Tb - Tc; Tz = Ty * Te; TD = Ty * T9; Tm = W[7]; T1B = T6 * Tm; T1E = T3 * Tm; T2o = T2 * Tm; T2l = T5 * Tm; T1T = T9 * Tm; T1Q = Te * Tm; Th = W[6]; T1A = T3 * Th; T1F = T6 * Th; T2p = T5 * Th; T2k = T2 * Th; T1U = Te * Th; T1P = T9 * Th; } T1C = T1A + T1B; T3K = T1E + T1F; T1V = T1T + T1U; T3x = T2o - T2p; T3I = T1A - T1B; T1G = T1E - T1F; T1R = T1P - T1Q; { E T5W, T5X, T55, T56; T3v = T2k + T2l; T2m = T2k - T2l; T2q = T2o + T2p; T5W = T8 * Th; T5X = Td * Tm; T5Y = T5W - T5X; T6u = T5W + T5X; { E T51, T52, T60, T61; T51 = Tw * Th; T52 = Ty * Tm; T53 = T51 + T52; T5B = T51 - T52; T60 = T8 * Tm; T61 = Td * Th; T62 = T60 + T61; T6w = T60 - T61; } T55 = Tw * Tm; T56 = Ty * Th; T57 = T55 - T56; T5D = T55 + T56; { E Ti, Tq, TF, TJ, T3W, T3X, T3T, T3U, T3h, T3i, Tn, Tr, TB, TI, T3d; E T3f, T1k, T1o, T1Z, T23, TQ, TU, T2A, T2B, T2x, T2y, T20, T24, TX, TY; E T1i, T1n; T2V = T1P + T1Q; T2X = T1T - T1U; Tg = Ta + Tf; Ti = Tg * Th; Tq = Tg * Tm; TE = TC + TD; TF = TE * Tm; TJ = TE * Th; T3W = T37 * Tm; T3X = T39 * Th; T3Y = T3W - T3X; T3T = T37 * Th; T3U = T39 * Tm; T3V = T3T + T3U; T3h = T3c * Tm; T3i = T3e * Th; T3j = T3h - T3i; Tl = Tj - Tk; Tn = Tl * Tm; Tr = Tl * Th; TA = Tx - Tz; TB = TA * Th; TI = TA * Tm; T3d = T3c * Th; T3f = T3e * Tm; T3g = T3d + T3f; T1j = Tj + Tk; T1k = T1j * Tm; T1o = T1j * Th; T1t = Tx + Tz; T1Z = T1t * Th; T23 = T1t * Tm; TQ = TP * Th; TU = TT * Tm; TV = TQ + TU; T2A = T1a * Tm; T2B = T1e * Th; T2C = T2A - T2B; T2x = T1a * Th; T2y = T1e * Tm; T2z = T2x + T2y; T1u = TC - TD; T20 = T1u * Tm; T24 = T1u * Th; TX = TP * Tm; TY = TT * Th; TZ = TX - TY; T1h = Ta - Tf; T1i = T1h * Th; T1n = T1h * Tm; To = Ti - Tn; T1p = T1n + T1o; T6j = TQ - TU; T6H = T2A + T2B; Ts = Tq + Tr; T1l = T1i - T1k; T6l = TX + TY; T6F = T2x - T2y; T2P = T1Z - T20; T4b = TI + TJ; T4x = T3d - T3f; T5i = T3W + T3X; T2R = T23 + T24; T49 = TB - TF; T4z = T3h + T3i; T5g = T3T - T3U; TG = TB + TF; T4k = Ti + Tn; T4m = Tq - Tr; TK = TI - TJ; T21 = T1Z + T20; T3O = T1i + T1k; T3Q = T1n - T1o; T25 = T23 - T24; TW = W[8]; T10 = W[9]; T11 = FMA(TV, TW, TZ * T10); T79 = FNMS(T25, TW, T21 * T10); T6X = FNMS(Td, TW, T8 * T10); T5M = FNMS(T2X, TW, T2V * T10); T6b = FNMS(TK, TW, TG * T10); T1v = FMA(T1t, TW, T1u * T10); T30 = FMA(T1h, TW, T1j * T10); T69 = FMA(TG, TW, TK * T10); T77 = FMA(T21, TW, T25 * T10); T13 = FNMS(TZ, TW, TV * T10); T2F = FNMS(T2C, TW, T2z * T10); T2D = FMA(T2z, TW, T2C * T10); T6p = FMA(T1a, TW, T1e * T10); T6O = FMA(TP, TW, TT * T10); T1x = FNMS(T1u, TW, T1t * T10); T2a = FNMS(TE, TW, TA * T10); T2f = FMA(T3, TW, T6 * T10); T6V = FMA(T8, TW, Td * T10); T28 = FMA(TA, TW, TE * T10); T6r = FNMS(T1e, TW, T1a * T10); T2h = FNMS(T6, TW, T3 * T10); T6Q = FNMS(TT, TW, TP * T10); T32 = FNMS(T1j, TW, T1h * T10); T5K = FMA(T2V, TW, T2X * T10); T5w = FMA(Tw, TW, Ty * T10); T4G = FMA(T3O, TW, T3Q * T10); T4Q = FMA(T4k, TW, T4m * T10); T3m = FNMS(T3j, TW, T3g * T10); T4h = FNMS(Te, TW, T9 * T10); T4I = FNMS(T3Q, TW, T3O * T10); T5y = FNMS(Ty, TW, Tw * T10); T3k = FMA(T3g, TW, T3j * T10); T4f = FMA(T9, TW, Te * T10); T41 = FNMS(T3Y, TW, T3V * T10); T4S = FNMS(T4m, TW, T4k * T10); T4Y = FNMS(T3e, TW, T3c * T10); T3q = FMA(Tg, TW, Tl * T10); T3D = FMA(T2, TW, T5 * T10); T3F = FNMS(T5, TW, T2 * T10); T5r = FNMS(T39, TW, T37 * T10); T3s = FNMS(Tl, TW, Tg * T10); T4W = FMA(T3c, TW, T3e * T10); T3Z = FMA(T3V, TW, T3Y * T10); T5p = FMA(T37, TW, T39 * T10); } } } { E T17, TdV, Tj3, Tjx, T7l, TbJ, Ti3, Tix, T1K, Tiw, TdY, ThY, T7w, Tj0, TbM; E Tjw, T2e, TgA, T7I, TaY, TbQ, Tda, Te4, TfO, T2J, TgB, T7T, TaZ, TbT, Tdb; E Te9, TfP, T36, T3B, TgH, TgE, TgF, TgG, T80, TbW, Tel, TfT, T8b, Tc0, T8k; E TbX, Teg, TfS, T8h, TbZ, T45, T4q, TgJ, TgK, TgL, TgM, T8r, Tc6, Tew, TfW; E T8C, Tc4, T8L, Tc7, Ter, TfV, T8I, Tc3, T6B, Th1, Tfm, Tga, Th8, ThI, T9N; E Tcv, T9Y, TcH, Tav, Tcw, Tf5, Tg7, Tas, TcG, T5c, TgV, TeV, Tg0, TgS, ThD; E T8U, Tcc, T95, Tco, T9C, Tcd, TeE, Tg3, T9z, Tcn, T5R, TgT, TeO, TeW, TgY; E ThE, T9h, T9F, T9s, T9E, Tck, Tcq, TeJ, TeX, Tch, Tcr, T7e, Th9, Tff, Tfn; E Th4, ThJ, Taa, Tay, Tal, Tax, TcD, TcJ, Tfa, Tfo, TcA, TcK; { E T1, Ti1, Tu, Ti0, TM, T7i, T15, T7j, Tp, Tt; T1 = ri[0]; Ti1 = ii[0]; Tp = ri[WS(rs, 32)]; Tt = ii[WS(rs, 32)]; Tu = FMA(To, Tp, Ts * Tt); Ti0 = FNMS(Ts, Tp, To * Tt); { E TH, TL, T12, T14; TH = ri[WS(rs, 16)]; TL = ii[WS(rs, 16)]; TM = FMA(TG, TH, TK * TL); T7i = FNMS(TK, TH, TG * TL); T12 = ri[WS(rs, 48)]; T14 = ii[WS(rs, 48)]; T15 = FMA(T11, T12, T13 * T14); T7j = FNMS(T13, T12, T11 * T14); } { E Tv, T16, Tj1, Tj2; Tv = T1 + Tu; T16 = TM + T15; T17 = Tv + T16; TdV = Tv - T16; Tj1 = Ti1 - Ti0; Tj2 = TM - T15; Tj3 = Tj1 - Tj2; Tjx = Tj2 + Tj1; } { E T7h, T7k, ThZ, Ti2; T7h = T1 - Tu; T7k = T7i - T7j; T7l = T7h - T7k; TbJ = T7h + T7k; ThZ = T7i + T7j; Ti2 = Ti0 + Ti1; Ti3 = ThZ + Ti2; Tix = Ti2 - ThZ; } } { E T1g, T7m, T1r, T7n, T7o, T7p, T1z, T7s, T1I, T7t, T7r, T7u; { E T1b, T1f, T1m, T1q; T1b = ri[WS(rs, 8)]; T1f = ii[WS(rs, 8)]; T1g = FMA(T1a, T1b, T1e * T1f); T7m = FNMS(T1e, T1b, T1a * T1f); T1m = ri[WS(rs, 40)]; T1q = ii[WS(rs, 40)]; T1r = FMA(T1l, T1m, T1p * T1q); T7n = FNMS(T1p, T1m, T1l * T1q); } T7o = T7m - T7n; T7p = T1g - T1r; { E T1w, T1y, T1D, T1H; T1w = ri[WS(rs, 56)]; T1y = ii[WS(rs, 56)]; T1z = FMA(T1v, T1w, T1x * T1y); T7s = FNMS(T1x, T1w, T1v * T1y); T1D = ri[WS(rs, 24)]; T1H = ii[WS(rs, 24)]; T1I = FMA(T1C, T1D, T1G * T1H); T7t = FNMS(T1G, T1D, T1C * T1H); } T7r = T1z - T1I; T7u = T7s - T7t; { E T1s, T1J, TdW, TdX; T1s = T1g + T1r; T1J = T1z + T1I; T1K = T1s + T1J; Tiw = T1J - T1s; TdW = T7m + T7n; TdX = T7s + T7t; TdY = TdW - TdX; ThY = TdW + TdX; } { E T7q, T7v, TbK, TbL; T7q = T7o - T7p; T7v = T7r + T7u; T7w = KP707106781 * (T7q - T7v); Tj0 = KP707106781 * (T7q + T7v); TbK = T7p + T7o; TbL = T7r - T7u; TbM = KP707106781 * (TbK + TbL); Tjw = KP707106781 * (TbL - TbK); } } { E T1Y, Te0, T7A, T7D, T2d, Te1, T7B, T7G, T7C, T7H; { E T1O, T7y, T1X, T7z; { E T1M, T1N, T1S, T1W; T1M = ri[WS(rs, 4)]; T1N = ii[WS(rs, 4)]; T1O = FMA(T8, T1M, Td * T1N); T7y = FNMS(Td, T1M, T8 * T1N); T1S = ri[WS(rs, 36)]; T1W = ii[WS(rs, 36)]; T1X = FMA(T1R, T1S, T1V * T1W); T7z = FNMS(T1V, T1S, T1R * T1W); } T1Y = T1O + T1X; Te0 = T7y + T7z; T7A = T7y - T7z; T7D = T1O - T1X; } { E T27, T7E, T2c, T7F; { E T22, T26, T29, T2b; T22 = ri[WS(rs, 20)]; T26 = ii[WS(rs, 20)]; T27 = FMA(T21, T22, T25 * T26); T7E = FNMS(T25, T22, T21 * T26); T29 = ri[WS(rs, 52)]; T2b = ii[WS(rs, 52)]; T2c = FMA(T28, T29, T2a * T2b); T7F = FNMS(T2a, T29, T28 * T2b); } T2d = T27 + T2c; Te1 = T7E + T7F; T7B = T27 - T2c; T7G = T7E - T7F; } T2e = T1Y + T2d; TgA = Te0 + Te1; T7C = T7A + T7B; T7H = T7D - T7G; T7I = FNMS(KP923879532, T7H, KP382683432 * T7C); TaY = FMA(KP923879532, T7C, KP382683432 * T7H); { E TbO, TbP, Te2, Te3; TbO = T7A - T7B; TbP = T7D + T7G; TbQ = FNMS(KP382683432, TbP, KP923879532 * TbO); Tda = FMA(KP382683432, TbO, KP923879532 * TbP); Te2 = Te0 - Te1; Te3 = T1Y - T2d; Te4 = Te2 - Te3; TfO = Te3 + Te2; } } { E T2t, Te6, T7L, T7O, T2I, Te7, T7M, T7R, T7N, T7S; { E T2j, T7J, T2s, T7K; { E T2g, T2i, T2n, T2r; T2g = ri[WS(rs, 60)]; T2i = ii[WS(rs, 60)]; T2j = FMA(T2f, T2g, T2h * T2i); T7J = FNMS(T2h, T2g, T2f * T2i); T2n = ri[WS(rs, 28)]; T2r = ii[WS(rs, 28)]; T2s = FMA(T2m, T2n, T2q * T2r); T7K = FNMS(T2q, T2n, T2m * T2r); } T2t = T2j + T2s; Te6 = T7J + T7K; T7L = T7J - T7K; T7O = T2j - T2s; } { E T2w, T7P, T2H, T7Q; { E T2u, T2v, T2E, T2G; T2u = ri[WS(rs, 12)]; T2v = ii[WS(rs, 12)]; T2w = FMA(TP, T2u, TT * T2v); T7P = FNMS(TT, T2u, TP * T2v); T2E = ri[WS(rs, 44)]; T2G = ii[WS(rs, 44)]; T2H = FMA(T2D, T2E, T2F * T2G); T7Q = FNMS(T2F, T2E, T2D * T2G); } T2I = T2w + T2H; Te7 = T7P + T7Q; T7M = T2w - T2H; T7R = T7P - T7Q; } T2J = T2t + T2I; TgB = Te6 + Te7; T7N = T7L + T7M; T7S = T7O - T7R; T7T = FMA(KP382683432, T7N, KP923879532 * T7S); TaZ = FNMS(KP923879532, T7N, KP382683432 * T7S); { E TbR, TbS, Te5, Te8; TbR = T7L - T7M; TbS = T7O + T7R; TbT = FMA(KP923879532, TbR, KP382683432 * TbS); Tdb = FNMS(KP382683432, TbR, KP923879532 * TbS); Te5 = T2t - T2I; Te8 = Te6 - Te7; Te9 = Te5 + Te8; TfP = Te5 - Te8; } } { E T2O, T7W, T2T, T7X, T2U, Tec, T2Z, T8e, T34, T8f, T35, Ted, T3p, Tei, T86; E T89, T3A, Tej, T81, T84; { E T2M, T2N, T2Q, T2S; T2M = ri[WS(rs, 2)]; T2N = ii[WS(rs, 2)]; T2O = FMA(Tw, T2M, Ty * T2N); T7W = FNMS(Ty, T2M, Tw * T2N); T2Q = ri[WS(rs, 34)]; T2S = ii[WS(rs, 34)]; T2T = FMA(T2P, T2Q, T2R * T2S); T7X = FNMS(T2R, T2Q, T2P * T2S); } T2U = T2O + T2T; Tec = T7W + T7X; { E T2W, T2Y, T31, T33; T2W = ri[WS(rs, 18)]; T2Y = ii[WS(rs, 18)]; T2Z = FMA(T2V, T2W, T2X * T2Y); T8e = FNMS(T2X, T2W, T2V * T2Y); T31 = ri[WS(rs, 50)]; T33 = ii[WS(rs, 50)]; T34 = FMA(T30, T31, T32 * T33); T8f = FNMS(T32, T31, T30 * T33); } T35 = T2Z + T34; Ted = T8e + T8f; { E T3b, T87, T3o, T88; { E T38, T3a, T3l, T3n; T38 = ri[WS(rs, 10)]; T3a = ii[WS(rs, 10)]; T3b = FMA(T37, T38, T39 * T3a); T87 = FNMS(T39, T38, T37 * T3a); T3l = ri[WS(rs, 42)]; T3n = ii[WS(rs, 42)]; T3o = FMA(T3k, T3l, T3m * T3n); T88 = FNMS(T3m, T3l, T3k * T3n); } T3p = T3b + T3o; Tei = T87 + T88; T86 = T3b - T3o; T89 = T87 - T88; } { E T3u, T82, T3z, T83; { E T3r, T3t, T3w, T3y; T3r = ri[WS(rs, 58)]; T3t = ii[WS(rs, 58)]; T3u = FMA(T3q, T3r, T3s * T3t); T82 = FNMS(T3s, T3r, T3q * T3t); T3w = ri[WS(rs, 26)]; T3y = ii[WS(rs, 26)]; T3z = FMA(T3v, T3w, T3x * T3y); T83 = FNMS(T3x, T3w, T3v * T3y); } T3A = T3u + T3z; Tej = T82 + T83; T81 = T3u - T3z; T84 = T82 - T83; } T36 = T2U + T35; T3B = T3p + T3A; TgH = T36 - T3B; TgE = Tec + Ted; TgF = Tei + Tej; TgG = TgE - TgF; { E T7Y, T7Z, Teh, Tek; T7Y = T7W - T7X; T7Z = T2Z - T34; T80 = T7Y + T7Z; TbW = T7Y - T7Z; Teh = T2U - T35; Tek = Tei - Tej; Tel = Teh - Tek; TfT = Teh + Tek; } { E T85, T8a, T8i, T8j; T85 = T81 - T84; T8a = T86 + T89; T8b = KP707106781 * (T85 - T8a); Tc0 = KP707106781 * (T8a + T85); T8i = T89 - T86; T8j = T81 + T84; T8k = KP707106781 * (T8i - T8j); TbX = KP707106781 * (T8i + T8j); } { E Tee, Tef, T8d, T8g; Tee = Tec - Ted; Tef = T3A - T3p; Teg = Tee - Tef; TfS = Tee + Tef; T8d = T2O - T2T; T8g = T8e - T8f; T8h = T8d - T8g; TbZ = T8d + T8g; } } { E T3H, T8n, T3M, T8o, T3N, Ten, T3S, T8F, T43, T8G, T44, Teo, T4e, Tet, T8x; E T8A, T4p, Teu, T8s, T8v; { E T3E, T3G, T3J, T3L; T3E = ri[WS(rs, 62)]; T3G = ii[WS(rs, 62)]; T3H = FMA(T3D, T3E, T3F * T3G); T8n = FNMS(T3F, T3E, T3D * T3G); T3J = ri[WS(rs, 30)]; T3L = ii[WS(rs, 30)]; T3M = FMA(T3I, T3J, T3K * T3L); T8o = FNMS(T3K, T3J, T3I * T3L); } T3N = T3H + T3M; Ten = T8n + T8o; { E T3P, T3R, T40, T42; T3P = ri[WS(rs, 14)]; T3R = ii[WS(rs, 14)]; T3S = FMA(T3O, T3P, T3Q * T3R); T8F = FNMS(T3Q, T3P, T3O * T3R); T40 = ri[WS(rs, 46)]; T42 = ii[WS(rs, 46)]; T43 = FMA(T3Z, T40, T41 * T42); T8G = FNMS(T41, T40, T3Z * T42); } T44 = T3S + T43; Teo = T8F + T8G; { E T48, T8y, T4d, T8z; { E T46, T47, T4a, T4c; T46 = ri[WS(rs, 6)]; T47 = ii[WS(rs, 6)]; T48 = FMA(T3c, T46, T3e * T47); T8y = FNMS(T3e, T46, T3c * T47); T4a = ri[WS(rs, 38)]; T4c = ii[WS(rs, 38)]; T4d = FMA(T49, T4a, T4b * T4c); T8z = FNMS(T4b, T4a, T49 * T4c); } T4e = T48 + T4d; Tet = T8y + T8z; T8x = T48 - T4d; T8A = T8y - T8z; } { E T4j, T8t, T4o, T8u; { E T4g, T4i, T4l, T4n; T4g = ri[WS(rs, 54)]; T4i = ii[WS(rs, 54)]; T4j = FMA(T4f, T4g, T4h * T4i); T8t = FNMS(T4h, T4g, T4f * T4i); T4l = ri[WS(rs, 22)]; T4n = ii[WS(rs, 22)]; T4o = FMA(T4k, T4l, T4m * T4n); T8u = FNMS(T4m, T4l, T4k * T4n); } T4p = T4j + T4o; Teu = T8t + T8u; T8s = T4j - T4o; T8v = T8t - T8u; } T45 = T3N + T44; T4q = T4e + T4p; TgJ = T45 - T4q; TgK = Ten + Teo; TgL = Tet + Teu; TgM = TgK - TgL; { E T8p, T8q, Tes, Tev; T8p = T8n - T8o; T8q = T3S - T43; T8r = T8p + T8q; Tc6 = T8p - T8q; Tes = T3N - T44; Tev = Tet - Teu; Tew = Tes - Tev; TfW = Tes + Tev; } { E T8w, T8B, T8J, T8K; T8w = T8s - T8v; T8B = T8x + T8A; T8C = KP707106781 * (T8w - T8B); Tc4 = KP707106781 * (T8B + T8w); T8J = T8A - T8x; T8K = T8s + T8v; T8L = KP707106781 * (T8J - T8K); Tc7 = KP707106781 * (T8J + T8K); } { E Tep, Teq, T8E, T8H; Tep = Ten - Teo; Teq = T4p - T4e; Ter = Tep - Teq; TfV = Tep + Teq; T8E = T3H - T3M; T8H = T8F - T8G; T8I = T8E - T8H; Tc3 = T8E + T8H; } } { E T5V, Tao, T64, Tap, T65, Tfi, T68, T9K, T6d, T9L, T6e, Tfj, T6o, Tf2, T9Q; E T9R, T6z, Tf3, T9T, T9W; { E T5T, T5U, T5Z, T63; T5T = ri[WS(rs, 63)]; T5U = ii[WS(rs, 63)]; T5V = FMA(TW, T5T, T10 * T5U); Tao = FNMS(T10, T5T, TW * T5U); T5Z = ri[WS(rs, 31)]; T63 = ii[WS(rs, 31)]; T64 = FMA(T5Y, T5Z, T62 * T63); Tap = FNMS(T62, T5Z, T5Y * T63); } T65 = T5V + T64; Tfi = Tao + Tap; { E T66, T67, T6a, T6c; T66 = ri[WS(rs, 15)]; T67 = ii[WS(rs, 15)]; T68 = FMA(TV, T66, TZ * T67); T9K = FNMS(TZ, T66, TV * T67); T6a = ri[WS(rs, 47)]; T6c = ii[WS(rs, 47)]; T6d = FMA(T69, T6a, T6b * T6c); T9L = FNMS(T6b, T6a, T69 * T6c); } T6e = T68 + T6d; Tfj = T9K + T9L; { E T6i, T9O, T6n, T9P; { E T6g, T6h, T6k, T6m; T6g = ri[WS(rs, 7)]; T6h = ii[WS(rs, 7)]; T6i = FMA(T1t, T6g, T1u * T6h); T9O = FNMS(T1u, T6g, T1t * T6h); T6k = ri[WS(rs, 39)]; T6m = ii[WS(rs, 39)]; T6n = FMA(T6j, T6k, T6l * T6m); T9P = FNMS(T6l, T6k, T6j * T6m); } T6o = T6i + T6n; Tf2 = T9O + T9P; T9Q = T9O - T9P; T9R = T6i - T6n; } { E T6t, T9U, T6y, T9V; { E T6q, T6s, T6v, T6x; T6q = ri[WS(rs, 55)]; T6s = ii[WS(rs, 55)]; T6t = FMA(T6p, T6q, T6r * T6s); T9U = FNMS(T6r, T6q, T6p * T6s); T6v = ri[WS(rs, 23)]; T6x = ii[WS(rs, 23)]; T6y = FMA(T6u, T6v, T6w * T6x); T9V = FNMS(T6w, T6v, T6u * T6x); } T6z = T6t + T6y; Tf3 = T9U + T9V; T9T = T6t - T6y; T9W = T9U - T9V; } { E T6f, T6A, Tfk, Tfl; T6f = T65 + T6e; T6A = T6o + T6z; T6B = T6f + T6A; Th1 = T6f - T6A; Tfk = Tfi - Tfj; Tfl = T6z - T6o; Tfm = Tfk - Tfl; Tga = Tfk + Tfl; } { E Th6, Th7, T9J, T9M; Th6 = Tfi + Tfj; Th7 = Tf2 + Tf3; Th8 = Th6 - Th7; ThI = Th6 + Th7; T9J = T5V - T64; T9M = T9K - T9L; T9N = T9J - T9M; Tcv = T9J + T9M; } { E T9S, T9X, Tat, Tau; T9S = T9Q - T9R; T9X = T9T + T9W; T9Y = KP707106781 * (T9S - T9X); TcH = KP707106781 * (T9S + T9X); Tat = T9T - T9W; Tau = T9R + T9Q; Tav = KP707106781 * (Tat - Tau); Tcw = KP707106781 * (Tau + Tat); } { E Tf1, Tf4, Taq, Tar; Tf1 = T65 - T6e; Tf4 = Tf2 - Tf3; Tf5 = Tf1 - Tf4; Tg7 = Tf1 + Tf4; Taq = Tao - Tap; Tar = T68 - T6d; Tas = Taq + Tar; TcG = Taq - Tar; } } { E T4w, T8Q, T4B, T8R, T4C, TeA, T4F, T9w, T4K, T9x, T4L, TeB, T4V, TeS, T90; E T93, T5a, TeT, T8V, T8Y; { E T4u, T4v, T4y, T4A; T4u = ri[WS(rs, 1)]; T4v = ii[WS(rs, 1)]; T4w = FMA(T2, T4u, T5 * T4v); T8Q = FNMS(T5, T4u, T2 * T4v); T4y = ri[WS(rs, 33)]; T4A = ii[WS(rs, 33)]; T4B = FMA(T4x, T4y, T4z * T4A); T8R = FNMS(T4z, T4y, T4x * T4A); } T4C = T4w + T4B; TeA = T8Q + T8R; { E T4D, T4E, T4H, T4J; T4D = ri[WS(rs, 17)]; T4E = ii[WS(rs, 17)]; T4F = FMA(T3V, T4D, T3Y * T4E); T9w = FNMS(T3Y, T4D, T3V * T4E); T4H = ri[WS(rs, 49)]; T4J = ii[WS(rs, 49)]; T4K = FMA(T4G, T4H, T4I * T4J); T9x = FNMS(T4I, T4H, T4G * T4J); } T4L = T4F + T4K; TeB = T9w + T9x; { E T4P, T91, T4U, T92; { E T4N, T4O, T4R, T4T; T4N = ri[WS(rs, 9)]; T4O = ii[WS(rs, 9)]; T4P = FMA(T9, T4N, Te * T4O); T91 = FNMS(Te, T4N, T9 * T4O); T4R = ri[WS(rs, 41)]; T4T = ii[WS(rs, 41)]; T4U = FMA(T4Q, T4R, T4S * T4T); T92 = FNMS(T4S, T4R, T4Q * T4T); } T4V = T4P + T4U; TeS = T91 + T92; T90 = T4P - T4U; T93 = T91 - T92; } { E T50, T8W, T59, T8X; { E T4X, T4Z, T54, T58; T4X = ri[WS(rs, 57)]; T4Z = ii[WS(rs, 57)]; T50 = FMA(T4W, T4X, T4Y * T4Z); T8W = FNMS(T4Y, T4X, T4W * T4Z); T54 = ri[WS(rs, 25)]; T58 = ii[WS(rs, 25)]; T59 = FMA(T53, T54, T57 * T58); T8X = FNMS(T57, T54, T53 * T58); } T5a = T50 + T59; TeT = T8W + T8X; T8V = T50 - T59; T8Y = T8W - T8X; } { E T4M, T5b, TeR, TeU; T4M = T4C + T4L; T5b = T4V + T5a; T5c = T4M + T5b; TgV = T4M - T5b; TeR = T4C - T4L; TeU = TeS - TeT; TeV = TeR - TeU; Tg0 = TeR + TeU; } { E TgQ, TgR, T8S, T8T; TgQ = TeA + TeB; TgR = TeS + TeT; TgS = TgQ - TgR; ThD = TgQ + TgR; T8S = T8Q - T8R; T8T = T4F - T4K; T8U = T8S + T8T; Tcc = T8S - T8T; } { E T8Z, T94, T9A, T9B; T8Z = T8V - T8Y; T94 = T90 + T93; T95 = KP707106781 * (T8Z - T94); Tco = KP707106781 * (T94 + T8Z); T9A = T93 - T90; T9B = T8V + T8Y; T9C = KP707106781 * (T9A - T9B); Tcd = KP707106781 * (T9A + T9B); } { E TeC, TeD, T9v, T9y; TeC = TeA - TeB; TeD = T5a - T4V; TeE = TeC - TeD; Tg3 = TeC + TeD; T9v = T4w - T4B; T9y = T9w - T9x; T9z = T9v - T9y; Tcn = T9v + T9y; } } { E T5l, TeL, T9k, T9n, T5P, TeH, T9a, T9f, T5u, TeM, T9l, T9q, T5G, TeG, T97; E T9e; { E T5f, T9i, T5k, T9j; { E T5d, T5e, T5h, T5j; T5d = ri[WS(rs, 5)]; T5e = ii[WS(rs, 5)]; T5f = FMA(Tg, T5d, Tl * T5e); T9i = FNMS(Tl, T5d, Tg * T5e); T5h = ri[WS(rs, 37)]; T5j = ii[WS(rs, 37)]; T5k = FMA(T5g, T5h, T5i * T5j); T9j = FNMS(T5i, T5h, T5g * T5j); } T5l = T5f + T5k; TeL = T9i + T9j; T9k = T9i - T9j; T9n = T5f - T5k; } { E T5J, T98, T5O, T99; { E T5H, T5I, T5L, T5N; T5H = ri[WS(rs, 13)]; T5I = ii[WS(rs, 13)]; T5J = FMA(T1h, T5H, T1j * T5I); T98 = FNMS(T1j, T5H, T1h * T5I); T5L = ri[WS(rs, 45)]; T5N = ii[WS(rs, 45)]; T5O = FMA(T5K, T5L, T5M * T5N); T99 = FNMS(T5M, T5L, T5K * T5N); } T5P = T5J + T5O; TeH = T98 + T99; T9a = T98 - T99; T9f = T5J - T5O; } { E T5o, T9o, T5t, T9p; { E T5m, T5n, T5q, T5s; T5m = ri[WS(rs, 21)]; T5n = ii[WS(rs, 21)]; T5o = FMA(T3g, T5m, T3j * T5n); T9o = FNMS(T3j, T5m, T3g * T5n); T5q = ri[WS(rs, 53)]; T5s = ii[WS(rs, 53)]; T5t = FMA(T5p, T5q, T5r * T5s); T9p = FNMS(T5r, T5q, T5p * T5s); } T5u = T5o + T5t; TeM = T9o + T9p; T9l = T5o - T5t; T9q = T9o - T9p; } { E T5A, T9c, T5F, T9d; { E T5x, T5z, T5C, T5E; T5x = ri[WS(rs, 61)]; T5z = ii[WS(rs, 61)]; T5A = FMA(T5w, T5x, T5y * T5z); T9c = FNMS(T5y, T5x, T5w * T5z); T5C = ri[WS(rs, 29)]; T5E = ii[WS(rs, 29)]; T5F = FMA(T5B, T5C, T5D * T5E); T9d = FNMS(T5D, T5C, T5B * T5E); } T5G = T5A + T5F; TeG = T9c + T9d; T97 = T5A - T5F; T9e = T9c - T9d; } { E T5v, T5Q, TeK, TeN; T5v = T5l + T5u; T5Q = T5G + T5P; T5R = T5v + T5Q; TgT = T5Q - T5v; TeK = T5l - T5u; TeN = TeL - TeM; TeO = TeK + TeN; TeW = TeN - TeK; } { E TgW, TgX, T9b, T9g; TgW = TeL + TeM; TgX = TeG + TeH; TgY = TgW - TgX; ThE = TgW + TgX; T9b = T97 - T9a; T9g = T9e + T9f; T9h = FNMS(KP923879532, T9g, KP382683432 * T9b); T9F = FMA(KP382683432, T9g, KP923879532 * T9b); } { E T9m, T9r, Tci, Tcj; T9m = T9k + T9l; T9r = T9n - T9q; T9s = FMA(KP923879532, T9m, KP382683432 * T9r); T9E = FNMS(KP923879532, T9r, KP382683432 * T9m); Tci = T9k - T9l; Tcj = T9n + T9q; Tck = FMA(KP382683432, Tci, KP923879532 * Tcj); Tcq = FNMS(KP382683432, Tcj, KP923879532 * Tci); } { E TeF, TeI, Tcf, Tcg; TeF = T5G - T5P; TeI = TeG - TeH; TeJ = TeF - TeI; TeX = TeF + TeI; Tcf = T97 + T9a; Tcg = T9e - T9f; Tch = FNMS(KP382683432, Tcg, KP923879532 * Tcf); Tcr = FMA(KP923879532, Tcg, KP382683432 * Tcf); } } { E T6K, Tf6, Ta2, Ta5, T7c, Tfd, Tae, Taj, T6T, Tf7, Ta3, Ta8, T73, Tfc, Tad; E Tag; { E T6E, Ta0, T6J, Ta1; { E T6C, T6D, T6G, T6I; T6C = ri[WS(rs, 3)]; T6D = ii[WS(rs, 3)]; T6E = FMA(T3, T6C, T6 * T6D); Ta0 = FNMS(T6, T6C, T3 * T6D); T6G = ri[WS(rs, 35)]; T6I = ii[WS(rs, 35)]; T6J = FMA(T6F, T6G, T6H * T6I); Ta1 = FNMS(T6H, T6G, T6F * T6I); } T6K = T6E + T6J; Tf6 = Ta0 + Ta1; Ta2 = Ta0 - Ta1; Ta5 = T6E - T6J; } { E T76, Tah, T7b, Tai; { E T74, T75, T78, T7a; T74 = ri[WS(rs, 11)]; T75 = ii[WS(rs, 11)]; T76 = FMA(TA, T74, TE * T75); Tah = FNMS(TE, T74, TA * T75); T78 = ri[WS(rs, 43)]; T7a = ii[WS(rs, 43)]; T7b = FMA(T77, T78, T79 * T7a); Tai = FNMS(T79, T78, T77 * T7a); } T7c = T76 + T7b; Tfd = Tah + Tai; Tae = T76 - T7b; Taj = Tah - Tai; } { E T6N, Ta6, T6S, Ta7; { E T6L, T6M, T6P, T6R; T6L = ri[WS(rs, 19)]; T6M = ii[WS(rs, 19)]; T6N = FMA(T2z, T6L, T2C * T6M); Ta6 = FNMS(T2C, T6L, T2z * T6M); T6P = ri[WS(rs, 51)]; T6R = ii[WS(rs, 51)]; T6S = FMA(T6O, T6P, T6Q * T6R); Ta7 = FNMS(T6Q, T6P, T6O * T6R); } T6T = T6N + T6S; Tf7 = Ta6 + Ta7; Ta3 = T6N - T6S; Ta8 = Ta6 - Ta7; } { E T6Z, Tab, T72, Tac; { E T6W, T6Y, T70, T71; T6W = ri[WS(rs, 59)]; T6Y = ii[WS(rs, 59)]; T6Z = FMA(T6V, T6W, T6X * T6Y); Tab = FNMS(T6X, T6W, T6V * T6Y); T70 = ri[WS(rs, 27)]; T71 = ii[WS(rs, 27)]; T72 = FMA(Th, T70, Tm * T71); Tac = FNMS(Tm, T70, Th * T71); } T73 = T6Z + T72; Tfc = Tab + Tac; Tad = Tab - Tac; Tag = T6Z - T72; } { E T6U, T7d, Tfb, Tfe; T6U = T6K + T6T; T7d = T73 + T7c; T7e = T6U + T7d; Th9 = T7d - T6U; Tfb = T73 - T7c; Tfe = Tfc - Tfd; Tff = Tfb + Tfe; Tfn = Tfb - Tfe; } { E Th2, Th3, Ta4, Ta9; Th2 = Tf6 + Tf7; Th3 = Tfc + Tfd; Th4 = Th2 - Th3; ThJ = Th2 + Th3; Ta4 = Ta2 + Ta3; Ta9 = Ta5 - Ta8; Taa = FNMS(KP923879532, Ta9, KP382683432 * Ta4); Tay = FMA(KP923879532, Ta4, KP382683432 * Ta9); } { E Taf, Tak, TcB, TcC; Taf = Tad + Tae; Tak = Tag - Taj; Tal = FMA(KP382683432, Taf, KP923879532 * Tak); Tax = FNMS(KP923879532, Taf, KP382683432 * Tak); TcB = Tad - Tae; TcC = Tag + Taj; TcD = FMA(KP923879532, TcB, KP382683432 * TcC); TcJ = FNMS(KP382683432, TcB, KP923879532 * TcC); } { E Tf8, Tf9, Tcy, Tcz; Tf8 = Tf6 - Tf7; Tf9 = T6K - T6T; Tfa = Tf8 - Tf9; Tfo = Tf9 + Tf8; Tcy = Ta2 - Ta3; Tcz = Ta5 + Ta8; TcA = FNMS(KP382683432, Tcz, KP923879532 * Tcy); TcK = FMA(KP382683432, Tcy, KP923879532 * Tcz); } } { E T2L, Thx, ThU, ThV, Ti5, Tib, T4s, Tia, T7g, Ti7, ThG, ThO, ThL, ThP, ThA; E ThW; { E T1L, T2K, ThS, ThT; T1L = T17 + T1K; T2K = T2e + T2J; T2L = T1L + T2K; Thx = T1L - T2K; ThS = ThD + ThE; ThT = ThI + ThJ; ThU = ThS - ThT; ThV = ThS + ThT; } { E ThX, Ti4, T3C, T4r; ThX = TgA + TgB; Ti4 = ThY + Ti3; Ti5 = ThX + Ti4; Tib = Ti4 - ThX; T3C = T36 + T3B; T4r = T45 + T4q; T4s = T3C + T4r; Tia = T4r - T3C; } { E T5S, T7f, ThC, ThF; T5S = T5c + T5R; T7f = T6B + T7e; T7g = T5S + T7f; Ti7 = T7f - T5S; ThC = T5c - T5R; ThF = ThD - ThE; ThG = ThC + ThF; ThO = ThF - ThC; } { E ThH, ThK, Thy, Thz; ThH = T6B - T7e; ThK = ThI - ThJ; ThL = ThH - ThK; ThP = ThH + ThK; Thy = TgE + TgF; Thz = TgK + TgL; ThA = Thy - Thz; ThW = Thy + Thz; } { E T4t, Ti6, ThR, Ti8; T4t = T2L + T4s; ri[WS(rs, 32)] = T4t - T7g; ri[0] = T4t + T7g; Ti6 = ThW + Ti5; ii[0] = ThV + Ti6; ii[WS(rs, 32)] = Ti6 - ThV; ThR = T2L - T4s; ri[WS(rs, 48)] = ThR - ThU; ri[WS(rs, 16)] = ThR + ThU; Ti8 = Ti5 - ThW; ii[WS(rs, 16)] = Ti7 + Ti8; ii[WS(rs, 48)] = Ti8 - Ti7; } { E ThB, ThM, Ti9, Tic; ThB = Thx + ThA; ThM = KP707106781 * (ThG + ThL); ri[WS(rs, 40)] = ThB - ThM; ri[WS(rs, 8)] = ThB + ThM; Ti9 = KP707106781 * (ThO + ThP); Tic = Tia + Tib; ii[WS(rs, 8)] = Ti9 + Tic; ii[WS(rs, 40)] = Tic - Ti9; } { E ThN, ThQ, Tid, Tie; ThN = Thx - ThA; ThQ = KP707106781 * (ThO - ThP); ri[WS(rs, 56)] = ThN - ThQ; ri[WS(rs, 24)] = ThN + ThQ; Tid = KP707106781 * (ThL - ThG); Tie = Tib - Tia; ii[WS(rs, 24)] = Tid + Tie; ii[WS(rs, 56)] = Tie - Tid; } } { E TgD, Thh, Thr, Thv, Tij, Tip, TgO, Tig, Th0, The, Thk, Tio, Tho, Thu, Thb; E Thf; { E Tgz, TgC, Thp, Thq; Tgz = T17 - T1K; TgC = TgA - TgB; TgD = Tgz - TgC; Thh = Tgz + TgC; Thp = Th1 + Th4; Thq = Th8 + Th9; Thr = FNMS(KP382683432, Thq, KP923879532 * Thp); Thv = FMA(KP923879532, Thq, KP382683432 * Thp); } { E Tih, Tii, TgI, TgN; Tih = T2J - T2e; Tii = Ti3 - ThY; Tij = Tih + Tii; Tip = Tii - Tih; TgI = TgG - TgH; TgN = TgJ + TgM; TgO = KP707106781 * (TgI - TgN); Tig = KP707106781 * (TgI + TgN); } { E TgU, TgZ, Thi, Thj; TgU = TgS - TgT; TgZ = TgV - TgY; Th0 = FMA(KP923879532, TgU, KP382683432 * TgZ); The = FNMS(KP923879532, TgZ, KP382683432 * TgU); Thi = TgH + TgG; Thj = TgJ - TgM; Thk = KP707106781 * (Thi + Thj); Tio = KP707106781 * (Thj - Thi); } { E Thm, Thn, Th5, Tha; Thm = TgS + TgT; Thn = TgV + TgY; Tho = FMA(KP382683432, Thm, KP923879532 * Thn); Thu = FNMS(KP382683432, Thn, KP923879532 * Thm); Th5 = Th1 - Th4; Tha = Th8 - Th9; Thb = FNMS(KP923879532, Tha, KP382683432 * Th5); Thf = FMA(KP382683432, Tha, KP923879532 * Th5); } { E TgP, Thc, Tin, Tiq; TgP = TgD + TgO; Thc = Th0 + Thb; ri[WS(rs, 44)] = TgP - Thc; ri[WS(rs, 12)] = TgP + Thc; Tin = The + Thf; Tiq = Tio + Tip; ii[WS(rs, 12)] = Tin + Tiq; ii[WS(rs, 44)] = Tiq - Tin; } { E Thd, Thg, Tir, Tis; Thd = TgD - TgO; Thg = The - Thf; ri[WS(rs, 60)] = Thd - Thg; ri[WS(rs, 28)] = Thd + Thg; Tir = Thb - Th0; Tis = Tip - Tio; ii[WS(rs, 28)] = Tir + Tis; ii[WS(rs, 60)] = Tis - Tir; } { E Thl, Ths, Tif, Tik; Thl = Thh + Thk; Ths = Tho + Thr; ri[WS(rs, 36)] = Thl - Ths; ri[WS(rs, 4)] = Thl + Ths; Tif = Thu + Thv; Tik = Tig + Tij; ii[WS(rs, 4)] = Tif + Tik; ii[WS(rs, 36)] = Tik - Tif; } { E Tht, Thw, Til, Tim; Tht = Thh - Thk; Thw = Thu - Thv; ri[WS(rs, 52)] = Tht - Thw; ri[WS(rs, 20)] = Tht + Thw; Til = Thr - Tho; Tim = Tij - Tig; ii[WS(rs, 20)] = Til + Tim; ii[WS(rs, 52)] = Tim - Til; } } { E Teb, Tfx, Tey, TiK, TiN, TiT, TfA, TiS, Tfr, TfL, Tfv, TfH, Tf0, TfK, Tfu; E TfE; { E TdZ, Tea, Tfy, Tfz; TdZ = TdV - TdY; Tea = KP707106781 * (Te4 - Te9); Teb = TdZ - Tea; Tfx = TdZ + Tea; { E Tem, Tex, TiL, TiM; Tem = FNMS(KP923879532, Tel, KP382683432 * Teg); Tex = FMA(KP382683432, Ter, KP923879532 * Tew); Tey = Tem - Tex; TiK = Tem + Tex; TiL = KP707106781 * (TfP - TfO); TiM = Tix - Tiw; TiN = TiL + TiM; TiT = TiM - TiL; } Tfy = FMA(KP923879532, Teg, KP382683432 * Tel); Tfz = FNMS(KP923879532, Ter, KP382683432 * Tew); TfA = Tfy + Tfz; TiS = Tfz - Tfy; { E Tfh, TfF, Tfq, TfG, Tfg, Tfp; Tfg = KP707106781 * (Tfa - Tff); Tfh = Tf5 - Tfg; TfF = Tf5 + Tfg; Tfp = KP707106781 * (Tfn - Tfo); Tfq = Tfm - Tfp; TfG = Tfm + Tfp; Tfr = FNMS(KP980785280, Tfq, KP195090322 * Tfh); TfL = FMA(KP831469612, TfG, KP555570233 * TfF); Tfv = FMA(KP195090322, Tfq, KP980785280 * Tfh); TfH = FNMS(KP555570233, TfG, KP831469612 * TfF); } { E TeQ, TfC, TeZ, TfD, TeP, TeY; TeP = KP707106781 * (TeJ - TeO); TeQ = TeE - TeP; TfC = TeE + TeP; TeY = KP707106781 * (TeW - TeX); TeZ = TeV - TeY; TfD = TeV + TeY; Tf0 = FMA(KP980785280, TeQ, KP195090322 * TeZ); TfK = FNMS(KP555570233, TfD, KP831469612 * TfC); Tfu = FNMS(KP980785280, TeZ, KP195090322 * TeQ); TfE = FMA(KP555570233, TfC, KP831469612 * TfD); } } { E Tez, Tfs, TiR, TiU; Tez = Teb + Tey; Tfs = Tf0 + Tfr; ri[WS(rs, 46)] = Tez - Tfs; ri[WS(rs, 14)] = Tez + Tfs; TiR = Tfu + Tfv; TiU = TiS + TiT; ii[WS(rs, 14)] = TiR + TiU; ii[WS(rs, 46)] = TiU - TiR; } { E Tft, Tfw, TiV, TiW; Tft = Teb - Tey; Tfw = Tfu - Tfv; ri[WS(rs, 62)] = Tft - Tfw; ri[WS(rs, 30)] = Tft + Tfw; TiV = Tfr - Tf0; TiW = TiT - TiS; ii[WS(rs, 30)] = TiV + TiW; ii[WS(rs, 62)] = TiW - TiV; } { E TfB, TfI, TiJ, TiO; TfB = Tfx + TfA; TfI = TfE + TfH; ri[WS(rs, 38)] = TfB - TfI; ri[WS(rs, 6)] = TfB + TfI; TiJ = TfK + TfL; TiO = TiK + TiN; ii[WS(rs, 6)] = TiJ + TiO; ii[WS(rs, 38)] = TiO - TiJ; } { E TfJ, TfM, TiP, TiQ; TfJ = Tfx - TfA; TfM = TfK - TfL; ri[WS(rs, 54)] = TfJ - TfM; ri[WS(rs, 22)] = TfJ + TfM; TiP = TfH - TfE; TiQ = TiN - TiK; ii[WS(rs, 22)] = TiP + TiQ; ii[WS(rs, 54)] = TiQ - TiP; } } { E TfR, Tgj, TfY, Tiu, Tiz, TiF, Tgm, TiE, Tgd, Tgx, Tgh, Tgt, Tg6, Tgw, Tgg; E Tgq; { E TfN, TfQ, Tgk, Tgl; TfN = TdV + TdY; TfQ = KP707106781 * (TfO + TfP); TfR = TfN - TfQ; Tgj = TfN + TfQ; { E TfU, TfX, Tiv, Tiy; TfU = FNMS(KP382683432, TfT, KP923879532 * TfS); TfX = FMA(KP923879532, TfV, KP382683432 * TfW); TfY = TfU - TfX; Tiu = TfU + TfX; Tiv = KP707106781 * (Te4 + Te9); Tiy = Tiw + Tix; Tiz = Tiv + Tiy; TiF = Tiy - Tiv; } Tgk = FMA(KP382683432, TfS, KP923879532 * TfT); Tgl = FNMS(KP382683432, TfV, KP923879532 * TfW); Tgm = Tgk + Tgl; TiE = Tgl - Tgk; { E Tg9, Tgr, Tgc, Tgs, Tg8, Tgb; Tg8 = KP707106781 * (Tfo + Tfn); Tg9 = Tg7 - Tg8; Tgr = Tg7 + Tg8; Tgb = KP707106781 * (Tfa + Tff); Tgc = Tga - Tgb; Tgs = Tga + Tgb; Tgd = FNMS(KP831469612, Tgc, KP555570233 * Tg9); Tgx = FMA(KP195090322, Tgr, KP980785280 * Tgs); Tgh = FMA(KP831469612, Tg9, KP555570233 * Tgc); Tgt = FNMS(KP195090322, Tgs, KP980785280 * Tgr); } { E Tg2, Tgo, Tg5, Tgp, Tg1, Tg4; Tg1 = KP707106781 * (TeO + TeJ); Tg2 = Tg0 - Tg1; Tgo = Tg0 + Tg1; Tg4 = KP707106781 * (TeW + TeX); Tg5 = Tg3 - Tg4; Tgp = Tg3 + Tg4; Tg6 = FMA(KP555570233, Tg2, KP831469612 * Tg5); Tgw = FNMS(KP195090322, Tgo, KP980785280 * Tgp); Tgg = FNMS(KP831469612, Tg2, KP555570233 * Tg5); Tgq = FMA(KP980785280, Tgo, KP195090322 * Tgp); } } { E TfZ, Tge, TiD, TiG; TfZ = TfR + TfY; Tge = Tg6 + Tgd; ri[WS(rs, 42)] = TfZ - Tge; ri[WS(rs, 10)] = TfZ + Tge; TiD = Tgg + Tgh; TiG = TiE + TiF; ii[WS(rs, 10)] = TiD + TiG; ii[WS(rs, 42)] = TiG - TiD; } { E Tgf, Tgi, TiH, TiI; Tgf = TfR - TfY; Tgi = Tgg - Tgh; ri[WS(rs, 58)] = Tgf - Tgi; ri[WS(rs, 26)] = Tgf + Tgi; TiH = Tgd - Tg6; TiI = TiF - TiE; ii[WS(rs, 26)] = TiH + TiI; ii[WS(rs, 58)] = TiI - TiH; } { E Tgn, Tgu, Tit, TiA; Tgn = Tgj + Tgm; Tgu = Tgq + Tgt; ri[WS(rs, 34)] = Tgn - Tgu; ri[WS(rs, 2)] = Tgn + Tgu; Tit = Tgw + Tgx; TiA = Tiu + Tiz; ii[WS(rs, 2)] = Tit + TiA; ii[WS(rs, 34)] = TiA - Tit; } { E Tgv, Tgy, TiB, TiC; Tgv = Tgj - Tgm; Tgy = Tgw - Tgx; ri[WS(rs, 50)] = Tgv - Tgy; ri[WS(rs, 18)] = Tgv + Tgy; TiB = Tgt - Tgq; TiC = Tiz - Tiu; ii[WS(rs, 18)] = TiB + TiC; ii[WS(rs, 50)] = TiC - TiB; } } { E T7V, TaH, TjN, TjT, T8O, TjS, TaK, TjK, T9I, TaU, TaE, TaO, TaB, TaV, TaF; E TaR; { E T7x, T7U, TjL, TjM; T7x = T7l - T7w; T7U = T7I - T7T; T7V = T7x - T7U; TaH = T7x + T7U; TjL = TaZ - TaY; TjM = Tjx - Tjw; TjN = TjL + TjM; TjT = TjM - TjL; } { E T8m, TaI, T8N, TaJ; { E T8c, T8l, T8D, T8M; T8c = T80 - T8b; T8l = T8h - T8k; T8m = FNMS(KP980785280, T8l, KP195090322 * T8c); TaI = FMA(KP980785280, T8c, KP195090322 * T8l); T8D = T8r - T8C; T8M = T8I - T8L; T8N = FMA(KP195090322, T8D, KP980785280 * T8M); TaJ = FNMS(KP980785280, T8D, KP195090322 * T8M); } T8O = T8m - T8N; TjS = TaJ - TaI; TaK = TaI + TaJ; TjK = T8m + T8N; } { E T9u, TaM, T9H, TaN; { E T96, T9t, T9D, T9G; T96 = T8U - T95; T9t = T9h - T9s; T9u = T96 - T9t; TaM = T96 + T9t; T9D = T9z - T9C; T9G = T9E - T9F; T9H = T9D - T9G; TaN = T9D + T9G; } T9I = FMA(KP995184726, T9u, KP098017140 * T9H); TaU = FNMS(KP634393284, TaN, KP773010453 * TaM); TaE = FNMS(KP995184726, T9H, KP098017140 * T9u); TaO = FMA(KP634393284, TaM, KP773010453 * TaN); } { E Tan, TaP, TaA, TaQ; { E T9Z, Tam, Taw, Taz; T9Z = T9N - T9Y; Tam = Taa - Tal; Tan = T9Z - Tam; TaP = T9Z + Tam; Taw = Tas - Tav; Taz = Tax - Tay; TaA = Taw - Taz; TaQ = Taw + Taz; } TaB = FNMS(KP995184726, TaA, KP098017140 * Tan); TaV = FMA(KP773010453, TaQ, KP634393284 * TaP); TaF = FMA(KP098017140, TaA, KP995184726 * Tan); TaR = FNMS(KP634393284, TaQ, KP773010453 * TaP); } { E T8P, TaC, TjR, TjU; T8P = T7V + T8O; TaC = T9I + TaB; ri[WS(rs, 47)] = T8P - TaC; ri[WS(rs, 15)] = T8P + TaC; TjR = TaE + TaF; TjU = TjS + TjT; ii[WS(rs, 15)] = TjR + TjU; ii[WS(rs, 47)] = TjU - TjR; } { E TaD, TaG, TjV, TjW; TaD = T7V - T8O; TaG = TaE - TaF; ri[WS(rs, 63)] = TaD - TaG; ri[WS(rs, 31)] = TaD + TaG; TjV = TaB - T9I; TjW = TjT - TjS; ii[WS(rs, 31)] = TjV + TjW; ii[WS(rs, 63)] = TjW - TjV; } { E TaL, TaS, TjJ, TjO; TaL = TaH + TaK; TaS = TaO + TaR; ri[WS(rs, 39)] = TaL - TaS; ri[WS(rs, 7)] = TaL + TaS; TjJ = TaU + TaV; TjO = TjK + TjN; ii[WS(rs, 7)] = TjJ + TjO; ii[WS(rs, 39)] = TjO - TjJ; } { E TaT, TaW, TjP, TjQ; TaT = TaH - TaK; TaW = TaU - TaV; ri[WS(rs, 55)] = TaT - TaW; ri[WS(rs, 23)] = TaT + TaW; TjP = TaR - TaO; TjQ = TjN - TjK; ii[WS(rs, 23)] = TjP + TjQ; ii[WS(rs, 55)] = TjQ - TjP; } } { E TbV, TcT, Tjj, Tjp, Tca, Tjo, TcW, Tjg, Tcu, Td6, TcQ, Td0, TcN, Td7, TcR; E Td3; { E TbN, TbU, Tjh, Tji; TbN = TbJ - TbM; TbU = TbQ - TbT; TbV = TbN - TbU; TcT = TbN + TbU; Tjh = Tdb - Tda; Tji = Tj3 - Tj0; Tjj = Tjh + Tji; Tjp = Tji - Tjh; } { E Tc2, TcU, Tc9, TcV; { E TbY, Tc1, Tc5, Tc8; TbY = TbW - TbX; Tc1 = TbZ - Tc0; Tc2 = FNMS(KP831469612, Tc1, KP555570233 * TbY); TcU = FMA(KP555570233, Tc1, KP831469612 * TbY); Tc5 = Tc3 - Tc4; Tc8 = Tc6 - Tc7; Tc9 = FMA(KP831469612, Tc5, KP555570233 * Tc8); TcV = FNMS(KP831469612, Tc8, KP555570233 * Tc5); } Tca = Tc2 - Tc9; Tjo = TcV - TcU; TcW = TcU + TcV; Tjg = Tc2 + Tc9; } { E Tcm, TcY, Tct, TcZ; { E Tce, Tcl, Tcp, Tcs; Tce = Tcc - Tcd; Tcl = Tch - Tck; Tcm = Tce - Tcl; TcY = Tce + Tcl; Tcp = Tcn - Tco; Tcs = Tcq - Tcr; Tct = Tcp - Tcs; TcZ = Tcp + Tcs; } Tcu = FMA(KP956940335, Tcm, KP290284677 * Tct); Td6 = FNMS(KP471396736, TcZ, KP881921264 * TcY); TcQ = FNMS(KP956940335, Tct, KP290284677 * Tcm); Td0 = FMA(KP471396736, TcY, KP881921264 * TcZ); } { E TcF, Td1, TcM, Td2; { E Tcx, TcE, TcI, TcL; Tcx = Tcv - Tcw; TcE = TcA - TcD; TcF = Tcx - TcE; Td1 = Tcx + TcE; TcI = TcG - TcH; TcL = TcJ - TcK; TcM = TcI - TcL; Td2 = TcI + TcL; } TcN = FNMS(KP956940335, TcM, KP290284677 * TcF); Td7 = FMA(KP881921264, Td2, KP471396736 * Td1); TcR = FMA(KP290284677, TcM, KP956940335 * TcF); Td3 = FNMS(KP471396736, Td2, KP881921264 * Td1); } { E Tcb, TcO, Tjn, Tjq; Tcb = TbV + Tca; TcO = Tcu + TcN; ri[WS(rs, 45)] = Tcb - TcO; ri[WS(rs, 13)] = Tcb + TcO; Tjn = TcQ + TcR; Tjq = Tjo + Tjp; ii[WS(rs, 13)] = Tjn + Tjq; ii[WS(rs, 45)] = Tjq - Tjn; } { E TcP, TcS, Tjr, Tjs; TcP = TbV - Tca; TcS = TcQ - TcR; ri[WS(rs, 61)] = TcP - TcS; ri[WS(rs, 29)] = TcP + TcS; Tjr = TcN - Tcu; Tjs = Tjp - Tjo; ii[WS(rs, 29)] = Tjr + Tjs; ii[WS(rs, 61)] = Tjs - Tjr; } { E TcX, Td4, Tjf, Tjk; TcX = TcT + TcW; Td4 = Td0 + Td3; ri[WS(rs, 37)] = TcX - Td4; ri[WS(rs, 5)] = TcX + Td4; Tjf = Td6 + Td7; Tjk = Tjg + Tjj; ii[WS(rs, 5)] = Tjf + Tjk; ii[WS(rs, 37)] = Tjk - Tjf; } { E Td5, Td8, Tjl, Tjm; Td5 = TcT - TcW; Td8 = Td6 - Td7; ri[WS(rs, 53)] = Td5 - Td8; ri[WS(rs, 21)] = Td5 + Td8; Tjl = Td3 - Td0; Tjm = Tjj - Tjg; ii[WS(rs, 21)] = Tjl + Tjm; ii[WS(rs, 53)] = Tjm - Tjl; } } { E Tdd, TdF, Tj5, Tjb, Tdk, Tja, TdI, TiY, Tds, TdS, TdC, TdM, Tdz, TdT, TdD; E TdP; { E Td9, Tdc, TiZ, Tj4; Td9 = TbJ + TbM; Tdc = Tda + Tdb; Tdd = Td9 - Tdc; TdF = Td9 + Tdc; TiZ = TbQ + TbT; Tj4 = Tj0 + Tj3; Tj5 = TiZ + Tj4; Tjb = Tj4 - TiZ; } { E Tdg, TdG, Tdj, TdH; { E Tde, Tdf, Tdh, Tdi; Tde = TbW + TbX; Tdf = TbZ + Tc0; Tdg = FNMS(KP195090322, Tdf, KP980785280 * Tde); TdG = FMA(KP980785280, Tdf, KP195090322 * Tde); Tdh = Tc3 + Tc4; Tdi = Tc6 + Tc7; Tdj = FMA(KP195090322, Tdh, KP980785280 * Tdi); TdH = FNMS(KP195090322, Tdi, KP980785280 * Tdh); } Tdk = Tdg - Tdj; Tja = TdH - TdG; TdI = TdG + TdH; TiY = Tdg + Tdj; } { E Tdo, TdK, Tdr, TdL; { E Tdm, Tdn, Tdp, Tdq; Tdm = Tcn + Tco; Tdn = Tck + Tch; Tdo = Tdm - Tdn; TdK = Tdm + Tdn; Tdp = Tcc + Tcd; Tdq = Tcq + Tcr; Tdr = Tdp - Tdq; TdL = Tdp + Tdq; } Tds = FMA(KP634393284, Tdo, KP773010453 * Tdr); TdS = FNMS(KP098017140, TdK, KP995184726 * TdL); TdC = FNMS(KP773010453, Tdo, KP634393284 * Tdr); TdM = FMA(KP995184726, TdK, KP098017140 * TdL); } { E Tdv, TdN, Tdy, TdO; { E Tdt, Tdu, Tdw, Tdx; Tdt = Tcv + Tcw; Tdu = TcK + TcJ; Tdv = Tdt - Tdu; TdN = Tdt + Tdu; Tdw = TcG + TcH; Tdx = TcA + TcD; Tdy = Tdw - Tdx; TdO = Tdw + Tdx; } Tdz = FNMS(KP773010453, Tdy, KP634393284 * Tdv); TdT = FMA(KP098017140, TdN, KP995184726 * TdO); TdD = FMA(KP773010453, Tdv, KP634393284 * Tdy); TdP = FNMS(KP098017140, TdO, KP995184726 * TdN); } { E Tdl, TdA, Tj9, Tjc; Tdl = Tdd + Tdk; TdA = Tds + Tdz; ri[WS(rs, 41)] = Tdl - TdA; ri[WS(rs, 9)] = Tdl + TdA; Tj9 = TdC + TdD; Tjc = Tja + Tjb; ii[WS(rs, 9)] = Tj9 + Tjc; ii[WS(rs, 41)] = Tjc - Tj9; } { E TdB, TdE, Tjd, Tje; TdB = Tdd - Tdk; TdE = TdC - TdD; ri[WS(rs, 57)] = TdB - TdE; ri[WS(rs, 25)] = TdB + TdE; Tjd = Tdz - Tds; Tje = Tjb - Tja; ii[WS(rs, 25)] = Tjd + Tje; ii[WS(rs, 57)] = Tje - Tjd; } { E TdJ, TdQ, TiX, Tj6; TdJ = TdF + TdI; TdQ = TdM + TdP; ri[WS(rs, 33)] = TdJ - TdQ; ri[WS(rs, 1)] = TdJ + TdQ; TiX = TdS + TdT; Tj6 = TiY + Tj5; ii[WS(rs, 1)] = TiX + Tj6; ii[WS(rs, 33)] = Tj6 - TiX; } { E TdR, TdU, Tj7, Tj8; TdR = TdF - TdI; TdU = TdS - TdT; ri[WS(rs, 49)] = TdR - TdU; ri[WS(rs, 17)] = TdR + TdU; Tj7 = TdP - TdM; Tj8 = Tj5 - TiY; ii[WS(rs, 17)] = Tj7 + Tj8; ii[WS(rs, 49)] = Tj8 - Tj7; } } { E Tb1, Tbt, Tjz, TjF, Tb8, TjE, Tbw, Tju, Tbg, TbG, Tbq, TbA, Tbn, TbH, Tbr; E TbD; { E TaX, Tb0, Tjv, Tjy; TaX = T7l + T7w; Tb0 = TaY + TaZ; Tb1 = TaX - Tb0; Tbt = TaX + Tb0; Tjv = T7I + T7T; Tjy = Tjw + Tjx; Tjz = Tjv + Tjy; TjF = Tjy - Tjv; } { E Tb4, Tbu, Tb7, Tbv; { E Tb2, Tb3, Tb5, Tb6; Tb2 = T80 + T8b; Tb3 = T8h + T8k; Tb4 = FNMS(KP555570233, Tb3, KP831469612 * Tb2); Tbu = FMA(KP555570233, Tb2, KP831469612 * Tb3); Tb5 = T8r + T8C; Tb6 = T8I + T8L; Tb7 = FMA(KP831469612, Tb5, KP555570233 * Tb6); Tbv = FNMS(KP555570233, Tb5, KP831469612 * Tb6); } Tb8 = Tb4 - Tb7; TjE = Tbv - Tbu; Tbw = Tbu + Tbv; Tju = Tb4 + Tb7; } { E Tbc, Tby, Tbf, Tbz; { E Tba, Tbb, Tbd, Tbe; Tba = T9z + T9C; Tbb = T9s + T9h; Tbc = Tba - Tbb; Tby = Tba + Tbb; Tbd = T8U + T95; Tbe = T9E + T9F; Tbf = Tbd - Tbe; Tbz = Tbd + Tbe; } Tbg = FMA(KP471396736, Tbc, KP881921264 * Tbf); TbG = FNMS(KP290284677, Tby, KP956940335 * Tbz); Tbq = FNMS(KP881921264, Tbc, KP471396736 * Tbf); TbA = FMA(KP956940335, Tby, KP290284677 * Tbz); } { E Tbj, TbB, Tbm, TbC; { E Tbh, Tbi, Tbk, Tbl; Tbh = T9N + T9Y; Tbi = Tay + Tax; Tbj = Tbh - Tbi; TbB = Tbh + Tbi; Tbk = Tas + Tav; Tbl = Taa + Tal; Tbm = Tbk - Tbl; TbC = Tbk + Tbl; } Tbn = FNMS(KP881921264, Tbm, KP471396736 * Tbj); TbH = FMA(KP290284677, TbB, KP956940335 * TbC); Tbr = FMA(KP881921264, Tbj, KP471396736 * Tbm); TbD = FNMS(KP290284677, TbC, KP956940335 * TbB); } { E Tb9, Tbo, TjD, TjG; Tb9 = Tb1 + Tb8; Tbo = Tbg + Tbn; ri[WS(rs, 43)] = Tb9 - Tbo; ri[WS(rs, 11)] = Tb9 + Tbo; TjD = Tbq + Tbr; TjG = TjE + TjF; ii[WS(rs, 11)] = TjD + TjG; ii[WS(rs, 43)] = TjG - TjD; } { E Tbp, Tbs, TjH, TjI; Tbp = Tb1 - Tb8; Tbs = Tbq - Tbr; ri[WS(rs, 59)] = Tbp - Tbs; ri[WS(rs, 27)] = Tbp + Tbs; TjH = Tbn - Tbg; TjI = TjF - TjE; ii[WS(rs, 27)] = TjH + TjI; ii[WS(rs, 59)] = TjI - TjH; } { E Tbx, TbE, Tjt, TjA; Tbx = Tbt + Tbw; TbE = TbA + TbD; ri[WS(rs, 35)] = Tbx - TbE; ri[WS(rs, 3)] = Tbx + TbE; Tjt = TbG + TbH; TjA = Tju + Tjz; ii[WS(rs, 3)] = Tjt + TjA; ii[WS(rs, 35)] = TjA - Tjt; } { E TbF, TbI, TjB, TjC; TbF = Tbt - Tbw; TbI = TbG - TbH; ri[WS(rs, 51)] = TbF - TbI; ri[WS(rs, 19)] = TbF + TbI; TjB = TbD - TbA; TjC = Tjz - Tju; ii[WS(rs, 19)] = TjB + TjC; ii[WS(rs, 51)] = TjC - TjB; } } } } } } static const tw_instr twinstr[] = { {TW_CEXP, 0, 1}, {TW_CEXP, 0, 3}, {TW_CEXP, 0, 9}, {TW_CEXP, 0, 27}, {TW_CEXP, 0, 63}, {TW_NEXT, 1, 0} }; static const ct_desc desc = { 64, "t2_64", twinstr, &GENUS, {880, 386, 274, 0}, 0, 0, 0 }; void X(codelet_t2_64) (planner *p) { X(kdft_dit_register) (p, t2_64, &desc); } #endif /* HAVE_FMA */ fftw-3.3.3/dft/scalar/codelets/t2_5.c0000644000175000017500000001726112054410271014153 00000000000000/* * Copyright (c) 2003, 2007-11 Matteo Frigo * Copyright (c) 2003, 2007-11 Massachusetts Institute of Technology * * This program is free software; you can redistribute it and/or modify * it under the terms of the GNU General Public License as published by * the Free Software Foundation; either version 2 of the License, or * (at your option) any later version. * * This program is distributed in the hope that it will be useful, * but WITHOUT ANY WARRANTY; without even the implied warranty of * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the * GNU General Public License for more details. * * You should have received a copy of the GNU General Public License * along with this program; if not, write to the Free Software * Foundation, Inc., 51 Franklin Street, Fifth Floor, Boston, MA 02110-1301 USA * */ /* This file was automatically generated --- DO NOT EDIT */ /* Generated on Sun Nov 25 07:36:09 EST 2012 */ #include "codelet-dft.h" #ifdef HAVE_FMA /* Generated by: ../../../genfft/gen_twiddle.native -fma -reorder-insns -schedule-for-pipeline -compact -variables 4 -pipeline-latency 4 -twiddle-log3 -precompute-twiddles -n 5 -name t2_5 -include t.h */ /* * This function contains 44 FP additions, 40 FP multiplications, * (or, 14 additions, 10 multiplications, 30 fused multiply/add), * 47 stack variables, 4 constants, and 20 memory accesses */ #include "t.h" static void t2_5(R *ri, R *ii, const R *W, stride rs, INT mb, INT me, INT ms) { DK(KP951056516, +0.951056516295153572116439333379382143405698634); DK(KP559016994, +0.559016994374947424102293417182819058860154590); DK(KP250000000, +0.250000000000000000000000000000000000000000000); DK(KP618033988, +0.618033988749894848204586834365638117720309180); { INT m; for (m = mb, W = W + (mb * 4); m < me; m = m + 1, ri = ri + ms, ii = ii + ms, W = W + 4, MAKE_VOLATILE_STRIDE(10, rs)) { E Ta, T1, TO, Tp, TS, Ti, TL, TC, To, TE, Ts, TF, T2, T8, T5; E TT, Tt, TG; T2 = W[0]; Ta = W[3]; T8 = W[2]; T5 = W[1]; { E Tq, Tr, Te, T9; T1 = ri[0]; Te = T2 * Ta; T9 = T2 * T8; TO = ii[0]; { E T3, Tf, Tm, Tj, Tb, T4, T6, Tc, Tg; T3 = ri[WS(rs, 1)]; Tf = FMA(T5, T8, Te); Tm = FNMS(T5, T8, Te); Tj = FMA(T5, Ta, T9); Tb = FNMS(T5, Ta, T9); T4 = T2 * T3; T6 = ii[WS(rs, 1)]; Tc = ri[WS(rs, 4)]; Tg = ii[WS(rs, 4)]; { E Tk, Tl, Tn, TD; { E T7, Tz, Th, TB, Ty, Td, TA; Tk = ri[WS(rs, 2)]; T7 = FMA(T5, T6, T4); Ty = T2 * T6; Td = Tb * Tc; TA = Tb * Tg; Tl = Tj * Tk; Tz = FNMS(T5, T3, Ty); Th = FMA(Tf, Tg, Td); TB = FNMS(Tf, Tc, TA); Tn = ii[WS(rs, 2)]; Tp = ri[WS(rs, 3)]; TS = T7 - Th; Ti = T7 + Th; TL = Tz + TB; TC = Tz - TB; TD = Tj * Tn; Tq = T8 * Tp; Tr = ii[WS(rs, 3)]; } To = FMA(Tm, Tn, Tl); TE = FNMS(Tm, Tk, TD); } } Ts = FMA(Ta, Tr, Tq); TF = T8 * Tr; } TT = To - Ts; Tt = To + Ts; TG = FNMS(Ta, Tp, TF); { E TU, TW, TV, TR, Tw, Tu; TU = FMA(KP618033988, TT, TS); TW = FNMS(KP618033988, TS, TT); Tw = Ti - Tt; Tu = Ti + Tt; { E TM, TH, Tv, TI, TK; TM = TE + TG; TH = TE - TG; ri[0] = T1 + Tu; Tv = FNMS(KP250000000, Tu, T1); TI = FMA(KP618033988, TH, TC); TK = FNMS(KP618033988, TC, TH); { E TQ, TN, TJ, Tx, TP; TQ = TL - TM; TN = TL + TM; TJ = FNMS(KP559016994, Tw, Tv); Tx = FMA(KP559016994, Tw, Tv); ii[0] = TN + TO; TP = FNMS(KP250000000, TN, TO); ri[WS(rs, 1)] = FMA(KP951056516, TI, Tx); ri[WS(rs, 4)] = FNMS(KP951056516, TI, Tx); ri[WS(rs, 3)] = FMA(KP951056516, TK, TJ); ri[WS(rs, 2)] = FNMS(KP951056516, TK, TJ); TV = FNMS(KP559016994, TQ, TP); TR = FMA(KP559016994, TQ, TP); } } ii[WS(rs, 4)] = FMA(KP951056516, TU, TR); ii[WS(rs, 1)] = FNMS(KP951056516, TU, TR); ii[WS(rs, 3)] = FNMS(KP951056516, TW, TV); ii[WS(rs, 2)] = FMA(KP951056516, TW, TV); } } } } static const tw_instr twinstr[] = { {TW_CEXP, 0, 1}, {TW_CEXP, 0, 3}, {TW_NEXT, 1, 0} }; static const ct_desc desc = { 5, "t2_5", twinstr, &GENUS, {14, 10, 30, 0}, 0, 0, 0 }; void X(codelet_t2_5) (planner *p) { X(kdft_dit_register) (p, t2_5, &desc); } #else /* HAVE_FMA */ /* Generated by: ../../../genfft/gen_twiddle.native -compact -variables 4 -pipeline-latency 4 -twiddle-log3 -precompute-twiddles -n 5 -name t2_5 -include t.h */ /* * This function contains 44 FP additions, 32 FP multiplications, * (or, 30 additions, 18 multiplications, 14 fused multiply/add), * 37 stack variables, 4 constants, and 20 memory accesses */ #include "t.h" static void t2_5(R *ri, R *ii, const R *W, stride rs, INT mb, INT me, INT ms) { DK(KP250000000, +0.250000000000000000000000000000000000000000000); DK(KP559016994, +0.559016994374947424102293417182819058860154590); DK(KP587785252, +0.587785252292473129168705954639072768597652438); DK(KP951056516, +0.951056516295153572116439333379382143405698634); { INT m; for (m = mb, W = W + (mb * 4); m < me; m = m + 1, ri = ri + ms, ii = ii + ms, W = W + 4, MAKE_VOLATILE_STRIDE(10, rs)) { E T2, T4, T7, T9, Tb, Tl, Tf, Tj; { E T8, Te, Ta, Td; T2 = W[0]; T4 = W[1]; T7 = W[2]; T9 = W[3]; T8 = T2 * T7; Te = T4 * T7; Ta = T4 * T9; Td = T2 * T9; Tb = T8 - Ta; Tl = Td - Te; Tf = Td + Te; Tj = T8 + Ta; } { E T1, TI, Ty, TB, TN, TM, TF, TG, TH, Ti, Tr, Ts; T1 = ri[0]; TI = ii[0]; { E T6, Tw, Tq, TA, Th, Tx, Tn, Tz; { E T3, T5, To, Tp; T3 = ri[WS(rs, 1)]; T5 = ii[WS(rs, 1)]; T6 = FMA(T2, T3, T4 * T5); Tw = FNMS(T4, T3, T2 * T5); To = ri[WS(rs, 3)]; Tp = ii[WS(rs, 3)]; Tq = FMA(T7, To, T9 * Tp); TA = FNMS(T9, To, T7 * Tp); } { E Tc, Tg, Tk, Tm; Tc = ri[WS(rs, 4)]; Tg = ii[WS(rs, 4)]; Th = FMA(Tb, Tc, Tf * Tg); Tx = FNMS(Tf, Tc, Tb * Tg); Tk = ri[WS(rs, 2)]; Tm = ii[WS(rs, 2)]; Tn = FMA(Tj, Tk, Tl * Tm); Tz = FNMS(Tl, Tk, Tj * Tm); } Ty = Tw - Tx; TB = Tz - TA; TN = Tn - Tq; TM = T6 - Th; TF = Tw + Tx; TG = Tz + TA; TH = TF + TG; Ti = T6 + Th; Tr = Tn + Tq; Ts = Ti + Tr; } ri[0] = T1 + Ts; ii[0] = TH + TI; { E TC, TE, Tv, TD, Tt, Tu; TC = FMA(KP951056516, Ty, KP587785252 * TB); TE = FNMS(KP587785252, Ty, KP951056516 * TB); Tt = KP559016994 * (Ti - Tr); Tu = FNMS(KP250000000, Ts, T1); Tv = Tt + Tu; TD = Tu - Tt; ri[WS(rs, 4)] = Tv - TC; ri[WS(rs, 3)] = TD + TE; ri[WS(rs, 1)] = Tv + TC; ri[WS(rs, 2)] = TD - TE; } { E TO, TP, TL, TQ, TJ, TK; TO = FMA(KP951056516, TM, KP587785252 * TN); TP = FNMS(KP587785252, TM, KP951056516 * TN); TJ = KP559016994 * (TF - TG); TK = FNMS(KP250000000, TH, TI); TL = TJ + TK; TQ = TK - TJ; ii[WS(rs, 1)] = TL - TO; ii[WS(rs, 3)] = TQ - TP; ii[WS(rs, 4)] = TO + TL; ii[WS(rs, 2)] = TP + TQ; } } } } } static const tw_instr twinstr[] = { {TW_CEXP, 0, 1}, {TW_CEXP, 0, 3}, {TW_NEXT, 1, 0} }; static const ct_desc desc = { 5, "t2_5", twinstr, &GENUS, {30, 18, 14, 0}, 0, 0, 0 }; void X(codelet_t2_5) (planner *p) { X(kdft_dit_register) (p, t2_5, &desc); } #endif /* HAVE_FMA */ fftw-3.3.3/dft/scalar/codelets/t2_10.c0000644000175000017500000003447612054410271014236 00000000000000/* * Copyright (c) 2003, 2007-11 Matteo Frigo * Copyright (c) 2003, 2007-11 Massachusetts Institute of Technology * * This program is free software; you can redistribute it and/or modify * it under the terms of the GNU General Public License as published by * the Free Software Foundation; either version 2 of the License, or * (at your option) any later version. * * This program is distributed in the hope that it will be useful, * but WITHOUT ANY WARRANTY; without even the implied warranty of * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the * GNU General Public License for more details. * * You should have received a copy of the GNU General Public License * along with this program; if not, write to the Free Software * Foundation, Inc., 51 Franklin Street, Fifth Floor, Boston, MA 02110-1301 USA * */ /* This file was automatically generated --- DO NOT EDIT */ /* Generated on Sun Nov 25 07:36:09 EST 2012 */ #include "codelet-dft.h" #ifdef HAVE_FMA /* Generated by: ../../../genfft/gen_twiddle.native -fma -reorder-insns -schedule-for-pipeline -compact -variables 4 -pipeline-latency 4 -twiddle-log3 -precompute-twiddles -n 10 -name t2_10 -include t.h */ /* * This function contains 114 FP additions, 94 FP multiplications, * (or, 48 additions, 28 multiplications, 66 fused multiply/add), * 85 stack variables, 4 constants, and 40 memory accesses */ #include "t.h" static void t2_10(R *ri, R *ii, const R *W, stride rs, INT mb, INT me, INT ms) { DK(KP951056516, +0.951056516295153572116439333379382143405698634); DK(KP559016994, +0.559016994374947424102293417182819058860154590); DK(KP250000000, +0.250000000000000000000000000000000000000000000); DK(KP618033988, +0.618033988749894848204586834365638117720309180); { INT m; for (m = mb, W = W + (mb * 6); m < me; m = m + 1, ri = ri + ms, ii = ii + ms, W = W + 6, MAKE_VOLATILE_STRIDE(20, rs)) { E T27, T2b, T2a, T2c; { E T2, T3, T8, Tc, T5, T4, TX, T11, TE, T6, TB, TA; T2 = W[0]; T3 = W[2]; T8 = W[4]; Tc = W[5]; T5 = W[1]; T4 = T2 * T3; TX = T3 * T8; TA = T2 * T8; T11 = T3 * Tc; TE = T2 * Tc; T6 = W[3]; TB = FMA(T5, Tc, TA); { E T2d, T24, T1c, Tk, T1i, T28, T2l, T1a, T2f, T1I, T1R, T1Z, TL, T1v, T1d; E Tz, T1S, T1r, TH, T1t; { E T1, TF, TY, T12, Tl, T7, T23, To, Tb, Te, Ti, Th, Td, Tw, Ts; E Ta; T1 = ri[0]; TF = FNMS(T5, T8, TE); TY = FMA(T6, Tc, TX); T12 = FNMS(T6, T8, T11); Tl = FMA(T5, T6, T4); T7 = FNMS(T5, T6, T4); Ta = T2 * T6; T23 = ii[0]; { E Tg, T9, Tv, Tr; Tg = T7 * Tc; T9 = T7 * T8; Tv = Tl * Tc; Tr = Tl * T8; To = FNMS(T5, T3, Ta); Tb = FMA(T5, T3, Ta); Te = ri[WS(rs, 5)]; Ti = ii[WS(rs, 5)]; Th = FNMS(Tb, T8, Tg); Td = FMA(Tb, Tc, T9); Tw = FNMS(To, T8, Tv); Ts = FMA(To, Tc, Tr); } { E T18, T1G, T1g, TW, T1P, T1C, T14, T1E; { E TR, T1z, TV, T1B, TZ, T13, T15, T17, T10, T1D; { E TO, TQ, TP, T22, Tj, T1y, T21, Tf; TO = ri[WS(rs, 4)]; T21 = Td * Ti; Tf = Td * Te; TQ = ii[WS(rs, 4)]; TP = T7 * TO; T22 = FNMS(Th, Te, T21); Tj = FMA(Th, Ti, Tf); T1y = T7 * TQ; TR = FMA(Tb, TQ, TP); T2d = T23 - T22; T24 = T22 + T23; T1c = T1 + Tj; Tk = T1 - Tj; T1z = FNMS(Tb, TO, T1y); } T15 = ri[WS(rs, 1)]; T17 = ii[WS(rs, 1)]; { E TS, TU, T16, T1F, TT, T1A; TS = ri[WS(rs, 9)]; TU = ii[WS(rs, 9)]; T16 = T2 * T15; T1F = T2 * T17; TT = T8 * TS; T1A = T8 * TU; T18 = FMA(T5, T17, T16); T1G = FNMS(T5, T15, T1F); TV = FMA(Tc, TU, TT); T1B = FNMS(Tc, TS, T1A); } TZ = ri[WS(rs, 6)]; T13 = ii[WS(rs, 6)]; T1g = TR + TV; TW = TR - TV; T1P = T1z + T1B; T1C = T1z - T1B; T10 = TY * TZ; T1D = TY * T13; T14 = FMA(T12, T13, T10); T1E = FNMS(T12, TZ, T1D); } { E Tq, T1o, Ty, TC, TG, T1q, TD, T1s; { E TI, TK, Tt, T1p; { E Tm, T1n, Tp, Tn; Tm = ri[WS(rs, 2)]; Tp = ii[WS(rs, 2)]; { E T19, T1h, T1Q, T1H; T19 = T14 - T18; T1h = T14 + T18; T1Q = T1E + T1G; T1H = T1E - T1G; Tn = Tl * Tm; T1i = T1g + T1h; T28 = T1g - T1h; T2l = TW - T19; T1a = TW + T19; T2f = T1C + T1H; T1I = T1C - T1H; T1R = T1P - T1Q; T1Z = T1P + T1Q; T1n = Tl * Tp; } Tq = FMA(To, Tp, Tn); TI = ri[WS(rs, 3)]; TK = ii[WS(rs, 3)]; T1o = FNMS(To, Tm, T1n); } { E Tx, Tu, TJ, T1u; Tt = ri[WS(rs, 7)]; TJ = T3 * TI; T1u = T3 * TK; Tx = ii[WS(rs, 7)]; Tu = Ts * Tt; TL = FMA(T6, TK, TJ); T1v = FNMS(T6, TI, T1u); T1p = Ts * Tx; Ty = FMA(Tw, Tx, Tu); } TC = ri[WS(rs, 8)]; TG = ii[WS(rs, 8)]; T1q = FNMS(Tw, Tt, T1p); } T1d = Tq + Ty; Tz = Tq - Ty; TD = TB * TC; T1s = TB * TG; T1S = T1o + T1q; T1r = T1o - T1q; TH = FMA(TF, TG, TD); T1t = FNMS(TF, TC, T1s); } } } { E T1f, T29, T1Y, T1U, T2j, T2n, T2m, T2o; { E T2k, T2e, T1l, T1L, T1J, T1k, T1b, T1e, TM; T1e = TH + TL; TM = TH - TL; { E T1w, T1T, TN, T1x; T1w = T1t - T1v; T1T = T1t + T1v; T1f = T1d + T1e; T29 = T1d - T1e; T2k = Tz - TM; TN = Tz + TM; T1x = T1r - T1w; T2e = T1r + T1w; T1Y = T1S + T1T; T1U = T1S - T1T; T1l = TN - T1a; T1b = TN + T1a; T1L = FNMS(KP618033988, T1x, T1I); T1J = FMA(KP618033988, T1I, T1x); } T1k = FNMS(KP250000000, T1b, Tk); ri[WS(rs, 5)] = Tk + T1b; { E T2g, T2i, T2h, T1K, T1m; T2g = T2e + T2f; T2i = T2e - T2f; T1K = FNMS(KP559016994, T1l, T1k); T1m = FMA(KP559016994, T1l, T1k); T2h = FNMS(KP250000000, T2g, T2d); ri[WS(rs, 1)] = FMA(KP951056516, T1J, T1m); ri[WS(rs, 9)] = FNMS(KP951056516, T1J, T1m); ri[WS(rs, 3)] = FMA(KP951056516, T1L, T1K); ri[WS(rs, 7)] = FNMS(KP951056516, T1L, T1K); ii[WS(rs, 5)] = T2g + T2d; T2j = FMA(KP559016994, T2i, T2h); T2n = FNMS(KP559016994, T2i, T2h); T2m = FMA(KP618033988, T2l, T2k); T2o = FNMS(KP618033988, T2k, T2l); } } { E T1O, T1W, T1V, T1X, T1j, T1N, T1M, T20, T26, T25; T1j = T1f + T1i; T1N = T1f - T1i; ii[WS(rs, 7)] = FMA(KP951056516, T2o, T2n); ii[WS(rs, 3)] = FNMS(KP951056516, T2o, T2n); ii[WS(rs, 9)] = FMA(KP951056516, T2m, T2j); ii[WS(rs, 1)] = FNMS(KP951056516, T2m, T2j); T1M = FNMS(KP250000000, T1j, T1c); ri[0] = T1c + T1j; T1O = FNMS(KP559016994, T1N, T1M); T1W = FMA(KP559016994, T1N, T1M); T1V = FNMS(KP618033988, T1U, T1R); T1X = FMA(KP618033988, T1R, T1U); T20 = T1Y + T1Z; T26 = T1Y - T1Z; ri[WS(rs, 6)] = FMA(KP951056516, T1X, T1W); ri[WS(rs, 4)] = FNMS(KP951056516, T1X, T1W); ri[WS(rs, 8)] = FMA(KP951056516, T1V, T1O); ri[WS(rs, 2)] = FNMS(KP951056516, T1V, T1O); T25 = FNMS(KP250000000, T20, T24); ii[0] = T20 + T24; T27 = FNMS(KP559016994, T26, T25); T2b = FMA(KP559016994, T26, T25); T2a = FNMS(KP618033988, T29, T28); T2c = FMA(KP618033988, T28, T29); } } } } ii[WS(rs, 6)] = FNMS(KP951056516, T2c, T2b); ii[WS(rs, 4)] = FMA(KP951056516, T2c, T2b); ii[WS(rs, 8)] = FNMS(KP951056516, T2a, T27); ii[WS(rs, 2)] = FMA(KP951056516, T2a, T27); } } } static const tw_instr twinstr[] = { {TW_CEXP, 0, 1}, {TW_CEXP, 0, 3}, {TW_CEXP, 0, 9}, {TW_NEXT, 1, 0} }; static const ct_desc desc = { 10, "t2_10", twinstr, &GENUS, {48, 28, 66, 0}, 0, 0, 0 }; void X(codelet_t2_10) (planner *p) { X(kdft_dit_register) (p, t2_10, &desc); } #else /* HAVE_FMA */ /* Generated by: ../../../genfft/gen_twiddle.native -compact -variables 4 -pipeline-latency 4 -twiddle-log3 -precompute-twiddles -n 10 -name t2_10 -include t.h */ /* * This function contains 114 FP additions, 80 FP multiplications, * (or, 76 additions, 42 multiplications, 38 fused multiply/add), * 63 stack variables, 4 constants, and 40 memory accesses */ #include "t.h" static void t2_10(R *ri, R *ii, const R *W, stride rs, INT mb, INT me, INT ms) { DK(KP587785252, +0.587785252292473129168705954639072768597652438); DK(KP951056516, +0.951056516295153572116439333379382143405698634); DK(KP250000000, +0.250000000000000000000000000000000000000000000); DK(KP559016994, +0.559016994374947424102293417182819058860154590); { INT m; for (m = mb, W = W + (mb * 6); m < me; m = m + 1, ri = ri + ms, ii = ii + ms, W = W + 6, MAKE_VOLATILE_STRIDE(20, rs)) { E T2, T5, T3, T6, T8, Tm, Tc, Tk, T9, Td, Te, TM, TO, Tg, Tp; E Tv, Tx, Tr; { E T4, Tb, T7, Ta; T2 = W[0]; T5 = W[1]; T3 = W[2]; T6 = W[3]; T4 = T2 * T3; Tb = T5 * T3; T7 = T5 * T6; Ta = T2 * T6; T8 = T4 - T7; Tm = Ta - Tb; Tc = Ta + Tb; Tk = T4 + T7; T9 = W[4]; Td = W[5]; Te = FMA(T8, T9, Tc * Td); TM = FMA(T3, T9, T6 * Td); TO = FNMS(T6, T9, T3 * Td); Tg = FNMS(Tc, T9, T8 * Td); Tp = FMA(Tk, T9, Tm * Td); Tv = FMA(T2, T9, T5 * Td); Tx = FNMS(T5, T9, T2 * Td); Tr = FNMS(Tm, T9, Tk * Td); } { E Tj, T1S, TX, T1G, TL, TU, TV, T1s, T1t, T1C, T11, T12, T13, T1h, T1k; E T1Q, Tu, TD, TE, T1v, T1w, T1B, TY, TZ, T10, T1a, T1d, T1P; { E T1, T1F, Ti, T1E, Tf, Th; T1 = ri[0]; T1F = ii[0]; Tf = ri[WS(rs, 5)]; Th = ii[WS(rs, 5)]; Ti = FMA(Te, Tf, Tg * Th); T1E = FNMS(Tg, Tf, Te * Th); Tj = T1 - Ti; T1S = T1F - T1E; TX = T1 + Ti; T1G = T1E + T1F; } { E TH, T1f, TT, T1j, TK, T1g, TQ, T1i; { E TF, TG, TR, TS; TF = ri[WS(rs, 4)]; TG = ii[WS(rs, 4)]; TH = FMA(T8, TF, Tc * TG); T1f = FNMS(Tc, TF, T8 * TG); TR = ri[WS(rs, 1)]; TS = ii[WS(rs, 1)]; TT = FMA(T2, TR, T5 * TS); T1j = FNMS(T5, TR, T2 * TS); } { E TI, TJ, TN, TP; TI = ri[WS(rs, 9)]; TJ = ii[WS(rs, 9)]; TK = FMA(T9, TI, Td * TJ); T1g = FNMS(Td, TI, T9 * TJ); TN = ri[WS(rs, 6)]; TP = ii[WS(rs, 6)]; TQ = FMA(TM, TN, TO * TP); T1i = FNMS(TO, TN, TM * TP); } TL = TH - TK; TU = TQ - TT; TV = TL + TU; T1s = T1f + T1g; T1t = T1i + T1j; T1C = T1s + T1t; T11 = TH + TK; T12 = TQ + TT; T13 = T11 + T12; T1h = T1f - T1g; T1k = T1i - T1j; T1Q = T1h + T1k; } { E To, T18, TC, T1c, Tt, T19, Tz, T1b; { E Tl, Tn, TA, TB; Tl = ri[WS(rs, 2)]; Tn = ii[WS(rs, 2)]; To = FMA(Tk, Tl, Tm * Tn); T18 = FNMS(Tm, Tl, Tk * Tn); TA = ri[WS(rs, 3)]; TB = ii[WS(rs, 3)]; TC = FMA(T3, TA, T6 * TB); T1c = FNMS(T6, TA, T3 * TB); } { E Tq, Ts, Tw, Ty; Tq = ri[WS(rs, 7)]; Ts = ii[WS(rs, 7)]; Tt = FMA(Tp, Tq, Tr * Ts); T19 = FNMS(Tr, Tq, Tp * Ts); Tw = ri[WS(rs, 8)]; Ty = ii[WS(rs, 8)]; Tz = FMA(Tv, Tw, Tx * Ty); T1b = FNMS(Tx, Tw, Tv * Ty); } Tu = To - Tt; TD = Tz - TC; TE = Tu + TD; T1v = T18 + T19; T1w = T1b + T1c; T1B = T1v + T1w; TY = To + Tt; TZ = Tz + TC; T10 = TY + TZ; T1a = T18 - T19; T1d = T1b - T1c; T1P = T1a + T1d; } { E T15, TW, T16, T1m, T1o, T1e, T1l, T1n, T17; T15 = KP559016994 * (TE - TV); TW = TE + TV; T16 = FNMS(KP250000000, TW, Tj); T1e = T1a - T1d; T1l = T1h - T1k; T1m = FMA(KP951056516, T1e, KP587785252 * T1l); T1o = FNMS(KP587785252, T1e, KP951056516 * T1l); ri[WS(rs, 5)] = Tj + TW; T1n = T16 - T15; ri[WS(rs, 7)] = T1n - T1o; ri[WS(rs, 3)] = T1n + T1o; T17 = T15 + T16; ri[WS(rs, 9)] = T17 - T1m; ri[WS(rs, 1)] = T17 + T1m; } { E T1R, T1T, T1U, T1Y, T20, T1W, T1X, T1Z, T1V; T1R = KP559016994 * (T1P - T1Q); T1T = T1P + T1Q; T1U = FNMS(KP250000000, T1T, T1S); T1W = Tu - TD; T1X = TL - TU; T1Y = FMA(KP951056516, T1W, KP587785252 * T1X); T20 = FNMS(KP587785252, T1W, KP951056516 * T1X); ii[WS(rs, 5)] = T1T + T1S; T1Z = T1U - T1R; ii[WS(rs, 3)] = T1Z - T20; ii[WS(rs, 7)] = T20 + T1Z; T1V = T1R + T1U; ii[WS(rs, 1)] = T1V - T1Y; ii[WS(rs, 9)] = T1Y + T1V; } { E T1q, T14, T1p, T1y, T1A, T1u, T1x, T1z, T1r; T1q = KP559016994 * (T10 - T13); T14 = T10 + T13; T1p = FNMS(KP250000000, T14, TX); T1u = T1s - T1t; T1x = T1v - T1w; T1y = FNMS(KP587785252, T1x, KP951056516 * T1u); T1A = FMA(KP951056516, T1x, KP587785252 * T1u); ri[0] = TX + T14; T1z = T1q + T1p; ri[WS(rs, 4)] = T1z - T1A; ri[WS(rs, 6)] = T1z + T1A; T1r = T1p - T1q; ri[WS(rs, 2)] = T1r - T1y; ri[WS(rs, 8)] = T1r + T1y; } { E T1L, T1D, T1K, T1J, T1N, T1H, T1I, T1O, T1M; T1L = KP559016994 * (T1B - T1C); T1D = T1B + T1C; T1K = FNMS(KP250000000, T1D, T1G); T1H = T11 - T12; T1I = TY - TZ; T1J = FNMS(KP587785252, T1I, KP951056516 * T1H); T1N = FMA(KP951056516, T1I, KP587785252 * T1H); ii[0] = T1D + T1G; T1O = T1L + T1K; ii[WS(rs, 4)] = T1N + T1O; ii[WS(rs, 6)] = T1O - T1N; T1M = T1K - T1L; ii[WS(rs, 2)] = T1J + T1M; ii[WS(rs, 8)] = T1M - T1J; } } } } } static const tw_instr twinstr[] = { {TW_CEXP, 0, 1}, {TW_CEXP, 0, 3}, {TW_CEXP, 0, 9}, {TW_NEXT, 1, 0} }; static const ct_desc desc = { 10, "t2_10", twinstr, &GENUS, {76, 42, 38, 0}, 0, 0, 0 }; void X(codelet_t2_10) (planner *p) { X(kdft_dit_register) (p, t2_10, &desc); } #endif /* HAVE_FMA */ fftw-3.3.3/dft/scalar/codelets/t2_20.c0000644000175000017500000007324412054410273014235 00000000000000/* * Copyright (c) 2003, 2007-11 Matteo Frigo * Copyright (c) 2003, 2007-11 Massachusetts Institute of Technology * * This program is free software; you can redistribute it and/or modify * it under the terms of the GNU General Public License as published by * the Free Software Foundation; either version 2 of the License, or * (at your option) any later version. * * This program is distributed in the hope that it will be useful, * but WITHOUT ANY WARRANTY; without even the implied warranty of * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the * GNU General Public License for more details. * * You should have received a copy of the GNU General Public License * along with this program; if not, write to the Free Software * Foundation, Inc., 51 Franklin Street, Fifth Floor, Boston, MA 02110-1301 USA * */ /* This file was automatically generated --- DO NOT EDIT */ /* Generated on Sun Nov 25 07:36:09 EST 2012 */ #include "codelet-dft.h" #ifdef HAVE_FMA /* Generated by: ../../../genfft/gen_twiddle.native -fma -reorder-insns -schedule-for-pipeline -compact -variables 4 -pipeline-latency 4 -twiddle-log3 -precompute-twiddles -n 20 -name t2_20 -include t.h */ /* * This function contains 276 FP additions, 198 FP multiplications, * (or, 136 additions, 58 multiplications, 140 fused multiply/add), * 142 stack variables, 4 constants, and 80 memory accesses */ #include "t.h" static void t2_20(R *ri, R *ii, const R *W, stride rs, INT mb, INT me, INT ms) { DK(KP951056516, +0.951056516295153572116439333379382143405698634); DK(KP559016994, +0.559016994374947424102293417182819058860154590); DK(KP250000000, +0.250000000000000000000000000000000000000000000); DK(KP618033988, +0.618033988749894848204586834365638117720309180); { INT m; for (m = mb, W = W + (mb * 8); m < me; m = m + 1, ri = ri + ms, ii = ii + ms, W = W + 8, MAKE_VOLATILE_STRIDE(40, rs)) { E T59, T5i, T5k, T5e, T5c, T5d, T5j, T5f; { E T2, Th, Tf, T6, T5, Tl, T1p, T1n, Ti, T3, Tt, Tv, T24, T1f, T1D; E Tb, T1P, Tm, T21, T1b, T7, T1A, Tw, T1H, T13, TA, T1L, T17, T1S, Tq; E T1o, T2g, T1t, T2c, TO, TK; { E T1e, Ta, Tk, Tg; T2 = W[0]; Th = W[3]; Tf = W[2]; T6 = W[5]; T5 = W[1]; Tk = T2 * Th; Tg = T2 * Tf; T1e = Tf * T6; Ta = T2 * T6; Tl = FMA(T5, Tf, Tk); T1p = FNMS(T5, Tf, Tk); T1n = FMA(T5, Th, Tg); Ti = FNMS(T5, Th, Tg); T3 = W[4]; Tt = W[6]; Tv = W[7]; { E Tp, Tj, TN, TJ; Tp = Ti * T6; T24 = FMA(Th, T3, T1e); T1f = FNMS(Th, T3, T1e); T1D = FNMS(T5, T3, Ta); Tb = FMA(T5, T3, Ta); Tj = Ti * T3; { E T1a, T4, Tu, T1G; T1a = Tf * T3; T4 = T2 * T3; Tu = Ti * Tt; T1G = T2 * Tt; { E T12, Tz, T1K, T16; T12 = Tf * Tt; Tz = Ti * Tv; T1K = T2 * Tv; T16 = Tf * Tv; T1P = FNMS(Tl, T6, Tj); Tm = FMA(Tl, T6, Tj); T21 = FNMS(Th, T6, T1a); T1b = FMA(Th, T6, T1a); T7 = FNMS(T5, T6, T4); T1A = FMA(T5, T6, T4); Tw = FMA(Tl, Tv, Tu); T1H = FMA(T5, Tv, T1G); T13 = FMA(Th, Tv, T12); TA = FNMS(Tl, Tt, Tz); T1L = FNMS(T5, Tt, T1K); T17 = FNMS(Th, Tt, T16); T1S = FMA(Tl, T3, Tp); Tq = FNMS(Tl, T3, Tp); } } T1o = T1n * T3; T2g = T1n * Tv; TN = Tm * Tv; TJ = Tm * Tt; T1t = T1n * T6; T2c = T1n * Tt; TO = FNMS(Tq, Tt, TN); TK = FMA(Tq, Tv, TJ); } } { E Te, T2C, T4L, T57, T58, TD, T2H, T4H, T3C, T3Z, T11, T2v, T2P, T3P, T4k; E T4v, T3u, T43, T2r, T2z, T3b, T3T, T4g, T4z, T3n, T42, T20, T2y, T34, T3S; E T4d, T4y, T1c, T19, T1d, T3E, T1w, T2U, T1g, T1j, T1l; { E T2d, T2h, T2k, T1q, T1u, T2n, TL, TI, TM, T3x, TZ, T2N, TP, TS, TU; { E T1, T4K, T8, T9, Tc; T1 = ri[0]; T4K = ii[0]; T8 = ri[WS(rs, 10)]; T2d = FMA(T1p, Tv, T2c); T2h = FNMS(T1p, Tt, T2g); T2k = FMA(T1p, T6, T1o); T1q = FNMS(T1p, T6, T1o); T1u = FMA(T1p, T3, T1t); T2n = FNMS(T1p, T3, T1t); T9 = T7 * T8; Tc = ii[WS(rs, 10)]; { E Tx, Ts, T2F, TC, T2E; { E Tn, Tr, To, T2D, T4J, Ty, TB, Td, T4I; Tn = ri[WS(rs, 5)]; Tr = ii[WS(rs, 5)]; Tx = ri[WS(rs, 15)]; Td = FMA(Tb, Tc, T9); T4I = T7 * Tc; To = Tm * Tn; T2D = Tm * Tr; Te = T1 + Td; T2C = T1 - Td; T4J = FNMS(Tb, T8, T4I); Ty = Tw * Tx; TB = ii[WS(rs, 15)]; Ts = FMA(Tq, Tr, To); T4L = T4J + T4K; T57 = T4K - T4J; T2F = Tw * TB; TC = FMA(TA, TB, Ty); T2E = FNMS(Tq, Tn, T2D); } { E TF, TG, TH, TW, TY, T2G, T3w, TX, T2M; TF = ri[WS(rs, 4)]; T2G = FNMS(TA, Tx, T2F); T58 = Ts - TC; TD = Ts + TC; TG = Ti * TF; T2H = T2E - T2G; T4H = T2E + T2G; TH = ii[WS(rs, 4)]; TW = ri[WS(rs, 19)]; TY = ii[WS(rs, 19)]; TL = ri[WS(rs, 14)]; TI = FMA(Tl, TH, TG); T3w = Ti * TH; TX = Tt * TW; T2M = Tt * TY; TM = TK * TL; T3x = FNMS(Tl, TF, T3w); TZ = FMA(Tv, TY, TX); T2N = FNMS(Tv, TW, T2M); TP = ii[WS(rs, 14)]; TS = ri[WS(rs, 9)]; TU = ii[WS(rs, 9)]; } } } { E T27, T26, T28, T3p, T2p, T39, T29, T2e, T2i; { E T22, T23, T25, T2l, T2o, T3o, T2m, T38; { E TR, T2J, T3z, TV, T2L, T4i, T3A; T22 = ri[WS(rs, 12)]; { E TQ, T3y, TT, T2K; TQ = FMA(TO, TP, TM); T3y = TK * TP; TT = T3 * TS; T2K = T3 * TU; TR = TI + TQ; T2J = TI - TQ; T3z = FNMS(TO, TL, T3y); TV = FMA(T6, TU, TT); T2L = FNMS(T6, TS, T2K); T23 = T21 * T22; } T4i = T3x + T3z; T3A = T3x - T3z; { E T10, T3B, T4j, T2O; T10 = TV + TZ; T3B = TV - TZ; T4j = T2L + T2N; T2O = T2L - T2N; T3C = T3A + T3B; T3Z = T3A - T3B; T11 = TR - T10; T2v = TR + T10; T2P = T2J - T2O; T3P = T2J + T2O; T4k = T4i - T4j; T4v = T4i + T4j; T25 = ii[WS(rs, 12)]; } } T2l = ri[WS(rs, 7)]; T2o = ii[WS(rs, 7)]; T27 = ri[WS(rs, 2)]; T26 = FMA(T24, T25, T23); T3o = T21 * T25; T2m = T2k * T2l; T38 = T2k * T2o; T28 = T1n * T27; T3p = FNMS(T24, T22, T3o); T2p = FMA(T2n, T2o, T2m); T39 = FNMS(T2n, T2l, T38); T29 = ii[WS(rs, 2)]; T2e = ri[WS(rs, 17)]; T2i = ii[WS(rs, 17)]; } { E T1I, T1F, T1J, T3i, T1Y, T32, T1M, T1Q, T1T; { E T1B, T1C, T1E, T1V, T1X, T3h, T1W, T31; { E T2b, T35, T3r, T2j, T37, T4e, T3s; T1B = ri[WS(rs, 8)]; { E T2a, T3q, T2f, T36; T2a = FMA(T1p, T29, T28); T3q = T1n * T29; T2f = T2d * T2e; T36 = T2d * T2i; T2b = T26 + T2a; T35 = T26 - T2a; T3r = FNMS(T1p, T27, T3q); T2j = FMA(T2h, T2i, T2f); T37 = FNMS(T2h, T2e, T36); T1C = T1A * T1B; } T4e = T3p + T3r; T3s = T3p - T3r; { E T2q, T3t, T4f, T3a; T2q = T2j + T2p; T3t = T2j - T2p; T4f = T37 + T39; T3a = T37 - T39; T3u = T3s + T3t; T43 = T3s - T3t; T2r = T2b - T2q; T2z = T2b + T2q; T3b = T35 - T3a; T3T = T35 + T3a; T4g = T4e - T4f; T4z = T4e + T4f; T1E = ii[WS(rs, 8)]; } } T1V = ri[WS(rs, 3)]; T1X = ii[WS(rs, 3)]; T1I = ri[WS(rs, 18)]; T1F = FMA(T1D, T1E, T1C); T3h = T1A * T1E; T1W = Tf * T1V; T31 = Tf * T1X; T1J = T1H * T1I; T3i = FNMS(T1D, T1B, T3h); T1Y = FMA(Th, T1X, T1W); T32 = FNMS(Th, T1V, T31); T1M = ii[WS(rs, 18)]; T1Q = ri[WS(rs, 13)]; T1T = ii[WS(rs, 13)]; } { E T14, T15, T18, T1r, T1v, T3D, T1s, T2T; { E T1O, T2Y, T3k, T1U, T30, T4b, T3l; T14 = ri[WS(rs, 16)]; { E T1N, T3j, T1R, T2Z; T1N = FMA(T1L, T1M, T1J); T3j = T1H * T1M; T1R = T1P * T1Q; T2Z = T1P * T1T; T1O = T1F + T1N; T2Y = T1F - T1N; T3k = FNMS(T1L, T1I, T3j); T1U = FMA(T1S, T1T, T1R); T30 = FNMS(T1S, T1Q, T2Z); T15 = T13 * T14; } T4b = T3i + T3k; T3l = T3i - T3k; { E T1Z, T3m, T4c, T33; T1Z = T1U + T1Y; T3m = T1U - T1Y; T4c = T30 + T32; T33 = T30 - T32; T3n = T3l + T3m; T42 = T3l - T3m; T20 = T1O - T1Z; T2y = T1O + T1Z; T34 = T2Y - T33; T3S = T2Y + T33; T4d = T4b - T4c; T4y = T4b + T4c; T18 = ii[WS(rs, 16)]; } } T1r = ri[WS(rs, 11)]; T1v = ii[WS(rs, 11)]; T1c = ri[WS(rs, 6)]; T19 = FMA(T17, T18, T15); T3D = T13 * T18; T1s = T1q * T1r; T2T = T1q * T1v; T1d = T1b * T1c; T3E = FNMS(T17, T14, T3D); T1w = FMA(T1u, T1v, T1s); T2U = FNMS(T1u, T1r, T2T); T1g = ii[WS(rs, 6)]; T1j = ri[WS(rs, 1)]; T1l = ii[WS(rs, 1)]; } } } } { E T3J, T40, T2W, T3Q, T4M, T4E, T4F, T4U, T4S; { E T4X, T2u, T2w, T4w, T4W, T4r, T4p, T54, T56, T4V, T4a, T4q; { E T4h, TE, T4n, T53, T1z, T2s, T52; { E T1i, T2Q, T3G, T1m, T2S, T4l, T3H; T4h = T4d - T4g; T4X = T4d + T4g; { E T1h, T3F, T1k, T2R; T1h = FMA(T1f, T1g, T1d); T3F = T1b * T1g; T1k = T2 * T1j; T2R = T2 * T1l; T1i = T19 + T1h; T2Q = T19 - T1h; T3G = FNMS(T1f, T1c, T3F); T1m = FMA(T5, T1l, T1k); T2S = FNMS(T5, T1j, T2R); } TE = Te - TD; T2u = Te + TD; T4l = T3E + T3G; T3H = T3E - T3G; { E T1x, T3I, T4m, T2V, T1y; T1x = T1m + T1w; T3I = T1m - T1w; T4m = T2S + T2U; T2V = T2S - T2U; T3J = T3H + T3I; T40 = T3H - T3I; T1y = T1i - T1x; T2w = T1i + T1x; T2W = T2Q - T2V; T3Q = T2Q + T2V; T4n = T4l - T4m; T4w = T4l + T4m; T53 = T11 - T1y; T1z = T11 + T1y; T2s = T20 + T2r; T52 = T20 - T2r; } } { E T49, T48, T4o, T2t; T4o = T4k - T4n; T4W = T4k + T4n; T49 = T1z - T2s; T2t = T1z + T2s; T4r = FMA(KP618033988, T4h, T4o); T4p = FNMS(KP618033988, T4o, T4h); T54 = FNMS(KP618033988, T53, T52); T56 = FMA(KP618033988, T52, T53); ri[WS(rs, 10)] = TE + T2t; T48 = FNMS(KP250000000, T2t, TE); T4V = T4L - T4H; T4M = T4H + T4L; T4a = FNMS(KP559016994, T49, T48); T4q = FMA(KP559016994, T49, T48); } } { E T2x, T4Q, T4B, T4D, T4R, T2A, T51, T55; { E T4x, T50, T4Y, T4A, T4Z; T4E = T4v + T4w; T4x = T4v - T4w; ri[WS(rs, 18)] = FMA(KP951056516, T4p, T4a); ri[WS(rs, 2)] = FNMS(KP951056516, T4p, T4a); ri[WS(rs, 6)] = FMA(KP951056516, T4r, T4q); ri[WS(rs, 14)] = FNMS(KP951056516, T4r, T4q); T50 = T4W - T4X; T4Y = T4W + T4X; T4A = T4y - T4z; T4F = T4y + T4z; T2x = T2v + T2w; T4Q = T2v - T2w; ii[WS(rs, 10)] = T4Y + T4V; T4Z = FNMS(KP250000000, T4Y, T4V); T4B = FMA(KP618033988, T4A, T4x); T4D = FNMS(KP618033988, T4x, T4A); T4R = T2y - T2z; T2A = T2y + T2z; T51 = FNMS(KP559016994, T50, T4Z); T55 = FMA(KP559016994, T50, T4Z); } { E T4t, T4s, T2B, T4u, T4C; T2B = T2x + T2A; T4t = T2x - T2A; ii[WS(rs, 18)] = FNMS(KP951056516, T54, T51); ii[WS(rs, 2)] = FMA(KP951056516, T54, T51); ii[WS(rs, 14)] = FMA(KP951056516, T56, T55); ii[WS(rs, 6)] = FNMS(KP951056516, T56, T55); ri[0] = T2u + T2B; T4s = FNMS(KP250000000, T2B, T2u); T4u = FMA(KP559016994, T4t, T4s); T4C = FNMS(KP559016994, T4t, T4s); T4U = FNMS(KP618033988, T4Q, T4R); T4S = FMA(KP618033988, T4R, T4Q); ri[WS(rs, 16)] = FMA(KP951056516, T4B, T4u); ri[WS(rs, 4)] = FNMS(KP951056516, T4B, T4u); ri[WS(rs, 8)] = FMA(KP951056516, T4D, T4C); ri[WS(rs, 12)] = FNMS(KP951056516, T4D, T4C); } } } { E T3O, T5u, T5w, T5l, T5q, T5o; { E T5n, T5m, T2I, T4O, T3N, T3L, T2X, T5t, T4N, T5s, T3c, T3v, T3K, T4G; T5n = T3n + T3u; T3v = T3n - T3u; T3K = T3C - T3J; T5m = T3C + T3J; T3O = T2C + T2H; T2I = T2C - T2H; T4O = T4E - T4F; T4G = T4E + T4F; T3N = FMA(KP618033988, T3v, T3K); T3L = FNMS(KP618033988, T3K, T3v); T2X = T2P + T2W; T5t = T2P - T2W; ii[0] = T4G + T4M; T4N = FNMS(KP250000000, T4G, T4M); T5s = T34 - T3b; T3c = T34 + T3b; { E T3f, T3e, T4P, T4T, T3d, T3M, T3g; T4P = FMA(KP559016994, T4O, T4N); T4T = FNMS(KP559016994, T4O, T4N); T3f = T2X - T3c; T3d = T2X + T3c; ii[WS(rs, 16)] = FNMS(KP951056516, T4S, T4P); ii[WS(rs, 4)] = FMA(KP951056516, T4S, T4P); ii[WS(rs, 12)] = FMA(KP951056516, T4U, T4T); ii[WS(rs, 8)] = FNMS(KP951056516, T4U, T4T); ri[WS(rs, 15)] = T2I + T3d; T3e = FNMS(KP250000000, T3d, T2I); T5u = FNMS(KP618033988, T5t, T5s); T5w = FMA(KP618033988, T5s, T5t); T5l = T58 + T57; T59 = T57 - T58; T3M = FMA(KP559016994, T3f, T3e); T3g = FNMS(KP559016994, T3f, T3e); ri[WS(rs, 7)] = FNMS(KP951056516, T3L, T3g); ri[WS(rs, 3)] = FMA(KP951056516, T3L, T3g); ri[WS(rs, 19)] = FNMS(KP951056516, T3N, T3M); ri[WS(rs, 11)] = FMA(KP951056516, T3N, T3M); T5q = T5m - T5n; T5o = T5m + T5n; } } { E T5a, T5b, T47, T45, T5g, T5h, T3V, T3X, T41, T44, T5p, T3W, T46, T3Y; T5a = T3Z + T40; T41 = T3Z - T40; T44 = T42 - T43; T5b = T42 + T43; ii[WS(rs, 15)] = T5o + T5l; T5p = FNMS(KP250000000, T5o, T5l); T47 = FNMS(KP618033988, T41, T44); T45 = FMA(KP618033988, T44, T41); { E T5r, T5v, T3R, T3U; T5r = FNMS(KP559016994, T5q, T5p); T5v = FMA(KP559016994, T5q, T5p); T3R = T3P + T3Q; T5g = T3P - T3Q; T5h = T3S - T3T; T3U = T3S + T3T; ii[WS(rs, 7)] = FMA(KP951056516, T5u, T5r); ii[WS(rs, 3)] = FNMS(KP951056516, T5u, T5r); ii[WS(rs, 19)] = FMA(KP951056516, T5w, T5v); ii[WS(rs, 11)] = FNMS(KP951056516, T5w, T5v); T3V = T3R + T3U; T3X = T3R - T3U; } ri[WS(rs, 5)] = T3O + T3V; T3W = FNMS(KP250000000, T3V, T3O); T5i = FMA(KP618033988, T5h, T5g); T5k = FNMS(KP618033988, T5g, T5h); T46 = FNMS(KP559016994, T3X, T3W); T3Y = FMA(KP559016994, T3X, T3W); ri[WS(rs, 9)] = FNMS(KP951056516, T45, T3Y); ri[WS(rs, 1)] = FMA(KP951056516, T45, T3Y); ri[WS(rs, 17)] = FNMS(KP951056516, T47, T46); ri[WS(rs, 13)] = FMA(KP951056516, T47, T46); T5e = T5a - T5b; T5c = T5a + T5b; } } } } } ii[WS(rs, 5)] = T5c + T59; T5d = FNMS(KP250000000, T5c, T59); T5j = FNMS(KP559016994, T5e, T5d); T5f = FMA(KP559016994, T5e, T5d); ii[WS(rs, 9)] = FMA(KP951056516, T5i, T5f); ii[WS(rs, 1)] = FNMS(KP951056516, T5i, T5f); ii[WS(rs, 17)] = FMA(KP951056516, T5k, T5j); ii[WS(rs, 13)] = FNMS(KP951056516, T5k, T5j); } } } static const tw_instr twinstr[] = { {TW_CEXP, 0, 1}, {TW_CEXP, 0, 3}, {TW_CEXP, 0, 9}, {TW_CEXP, 0, 19}, {TW_NEXT, 1, 0} }; static const ct_desc desc = { 20, "t2_20", twinstr, &GENUS, {136, 58, 140, 0}, 0, 0, 0 }; void X(codelet_t2_20) (planner *p) { X(kdft_dit_register) (p, t2_20, &desc); } #else /* HAVE_FMA */ /* Generated by: ../../../genfft/gen_twiddle.native -compact -variables 4 -pipeline-latency 4 -twiddle-log3 -precompute-twiddles -n 20 -name t2_20 -include t.h */ /* * This function contains 276 FP additions, 164 FP multiplications, * (or, 204 additions, 92 multiplications, 72 fused multiply/add), * 123 stack variables, 4 constants, and 80 memory accesses */ #include "t.h" static void t2_20(R *ri, R *ii, const R *W, stride rs, INT mb, INT me, INT ms) { DK(KP587785252, +0.587785252292473129168705954639072768597652438); DK(KP951056516, +0.951056516295153572116439333379382143405698634); DK(KP250000000, +0.250000000000000000000000000000000000000000000); DK(KP559016994, +0.559016994374947424102293417182819058860154590); { INT m; for (m = mb, W = W + (mb * 8); m < me; m = m + 1, ri = ri + ms, ii = ii + ms, W = W + 8, MAKE_VOLATILE_STRIDE(40, rs)) { E T2, T5, Tg, Ti, Tk, To, T1h, T1f, T6, T3, T8, T14, T1Q, Tc, T1O; E T1v, T18, T1t, T1n, T24, T1j, T22, Tq, Tu, T1E, T1G, Tx, Ty, Tz, TJ; E T1Z, TB, T1X, T1A, TZ, TL, T1y, TX; { E T7, T16, Ta, T13, T4, T17, Tb, T12; { E Th, Tn, Tj, Tm; T2 = W[0]; T5 = W[1]; Tg = W[2]; Ti = W[3]; Th = T2 * Tg; Tn = T5 * Tg; Tj = T5 * Ti; Tm = T2 * Ti; Tk = Th - Tj; To = Tm + Tn; T1h = Tm - Tn; T1f = Th + Tj; T6 = W[5]; T7 = T5 * T6; T16 = Tg * T6; Ta = T2 * T6; T13 = Ti * T6; T3 = W[4]; T4 = T2 * T3; T17 = Ti * T3; Tb = T5 * T3; T12 = Tg * T3; } T8 = T4 - T7; T14 = T12 + T13; T1Q = T16 + T17; Tc = Ta + Tb; T1O = T12 - T13; T1v = Ta - Tb; T18 = T16 - T17; T1t = T4 + T7; { E T1l, T1m, T1g, T1i; T1l = T1f * T6; T1m = T1h * T3; T1n = T1l + T1m; T24 = T1l - T1m; T1g = T1f * T3; T1i = T1h * T6; T1j = T1g - T1i; T22 = T1g + T1i; { E Tl, Tp, Ts, Tt; Tl = Tk * T3; Tp = To * T6; Tq = Tl + Tp; Ts = Tk * T6; Tt = To * T3; Tu = Ts - Tt; T1E = Tl - Tp; T1G = Ts + Tt; Tx = W[6]; Ty = W[7]; Tz = FMA(Tk, Tx, To * Ty); TJ = FMA(Tq, Tx, Tu * Ty); T1Z = FNMS(T1h, Tx, T1f * Ty); TB = FNMS(To, Tx, Tk * Ty); T1X = FMA(T1f, Tx, T1h * Ty); T1A = FNMS(T5, Tx, T2 * Ty); TZ = FNMS(Ti, Tx, Tg * Ty); TL = FNMS(Tu, Tx, Tq * Ty); T1y = FMA(T2, Tx, T5 * Ty); TX = FMA(Tg, Tx, Ti * Ty); } } } { E TF, T2b, T4A, T4J, T2K, T3r, T4a, T4m, T1N, T28, T29, T3C, T3F, T4o, T3X; E T3Y, T44, T2f, T2g, T2h, T2n, T2s, T4L, T3g, T3h, T4w, T3n, T3o, T3p, T30; E T35, T36, TW, T1r, T1s, T3J, T3M, T4n, T3U, T3V, T43, T2c, T2d, T2e, T2y; E T2D, T4K, T3d, T3e, T4v, T3k, T3l, T3m, T2P, T2U, T2V; { E T1, T48, Te, T47, Tw, T2H, TD, T2I, T9, Td; T1 = ri[0]; T48 = ii[0]; T9 = ri[WS(rs, 10)]; Td = ii[WS(rs, 10)]; Te = FMA(T8, T9, Tc * Td); T47 = FNMS(Tc, T9, T8 * Td); { E Tr, Tv, TA, TC; Tr = ri[WS(rs, 5)]; Tv = ii[WS(rs, 5)]; Tw = FMA(Tq, Tr, Tu * Tv); T2H = FNMS(Tu, Tr, Tq * Tv); TA = ri[WS(rs, 15)]; TC = ii[WS(rs, 15)]; TD = FMA(Tz, TA, TB * TC); T2I = FNMS(TB, TA, Tz * TC); } { E Tf, TE, T4y, T4z; Tf = T1 + Te; TE = Tw + TD; TF = Tf - TE; T2b = Tf + TE; T4y = T48 - T47; T4z = Tw - TD; T4A = T4y - T4z; T4J = T4z + T4y; } { E T2G, T2J, T46, T49; T2G = T1 - Te; T2J = T2H - T2I; T2K = T2G - T2J; T3r = T2G + T2J; T46 = T2H + T2I; T49 = T47 + T48; T4a = T46 + T49; T4m = T49 - T46; } } { E T1D, T3A, T2l, T2W, T27, T3E, T2r, T34, T1M, T3B, T2m, T2Z, T1W, T3D, T2q; E T31; { E T1x, T2j, T1C, T2k; { E T1u, T1w, T1z, T1B; T1u = ri[WS(rs, 8)]; T1w = ii[WS(rs, 8)]; T1x = FMA(T1t, T1u, T1v * T1w); T2j = FNMS(T1v, T1u, T1t * T1w); T1z = ri[WS(rs, 18)]; T1B = ii[WS(rs, 18)]; T1C = FMA(T1y, T1z, T1A * T1B); T2k = FNMS(T1A, T1z, T1y * T1B); } T1D = T1x + T1C; T3A = T2j + T2k; T2l = T2j - T2k; T2W = T1x - T1C; } { E T21, T32, T26, T33; { E T1Y, T20, T23, T25; T1Y = ri[WS(rs, 17)]; T20 = ii[WS(rs, 17)]; T21 = FMA(T1X, T1Y, T1Z * T20); T32 = FNMS(T1Z, T1Y, T1X * T20); T23 = ri[WS(rs, 7)]; T25 = ii[WS(rs, 7)]; T26 = FMA(T22, T23, T24 * T25); T33 = FNMS(T24, T23, T22 * T25); } T27 = T21 + T26; T3E = T32 + T33; T2r = T21 - T26; T34 = T32 - T33; } { E T1I, T2X, T1L, T2Y; { E T1F, T1H, T1J, T1K; T1F = ri[WS(rs, 13)]; T1H = ii[WS(rs, 13)]; T1I = FMA(T1E, T1F, T1G * T1H); T2X = FNMS(T1G, T1F, T1E * T1H); T1J = ri[WS(rs, 3)]; T1K = ii[WS(rs, 3)]; T1L = FMA(Tg, T1J, Ti * T1K); T2Y = FNMS(Ti, T1J, Tg * T1K); } T1M = T1I + T1L; T3B = T2X + T2Y; T2m = T1I - T1L; T2Z = T2X - T2Y; } { E T1S, T2o, T1V, T2p; { E T1P, T1R, T1T, T1U; T1P = ri[WS(rs, 12)]; T1R = ii[WS(rs, 12)]; T1S = FMA(T1O, T1P, T1Q * T1R); T2o = FNMS(T1Q, T1P, T1O * T1R); T1T = ri[WS(rs, 2)]; T1U = ii[WS(rs, 2)]; T1V = FMA(T1f, T1T, T1h * T1U); T2p = FNMS(T1h, T1T, T1f * T1U); } T1W = T1S + T1V; T3D = T2o + T2p; T2q = T2o - T2p; T31 = T1S - T1V; } T1N = T1D - T1M; T28 = T1W - T27; T29 = T1N + T28; T3C = T3A - T3B; T3F = T3D - T3E; T4o = T3C + T3F; T3X = T3A + T3B; T3Y = T3D + T3E; T44 = T3X + T3Y; T2f = T1D + T1M; T2g = T1W + T27; T2h = T2f + T2g; T2n = T2l + T2m; T2s = T2q + T2r; T4L = T2n + T2s; T3g = T2l - T2m; T3h = T2q - T2r; T4w = T3g + T3h; T3n = T2W + T2Z; T3o = T31 + T34; T3p = T3n + T3o; T30 = T2W - T2Z; T35 = T31 - T34; T36 = T30 + T35; } { E TO, T3H, T2w, T2L, T1q, T3L, T2C, T2T, TV, T3I, T2x, T2O, T1b, T3K, T2B; E T2Q; { E TI, T2u, TN, T2v; { E TG, TH, TK, TM; TG = ri[WS(rs, 4)]; TH = ii[WS(rs, 4)]; TI = FMA(Tk, TG, To * TH); T2u = FNMS(To, TG, Tk * TH); TK = ri[WS(rs, 14)]; TM = ii[WS(rs, 14)]; TN = FMA(TJ, TK, TL * TM); T2v = FNMS(TL, TK, TJ * TM); } TO = TI + TN; T3H = T2u + T2v; T2w = T2u - T2v; T2L = TI - TN; } { E T1e, T2R, T1p, T2S; { E T1c, T1d, T1k, T1o; T1c = ri[WS(rs, 1)]; T1d = ii[WS(rs, 1)]; T1e = FMA(T2, T1c, T5 * T1d); T2R = FNMS(T5, T1c, T2 * T1d); T1k = ri[WS(rs, 11)]; T1o = ii[WS(rs, 11)]; T1p = FMA(T1j, T1k, T1n * T1o); T2S = FNMS(T1n, T1k, T1j * T1o); } T1q = T1e + T1p; T3L = T2R + T2S; T2C = T1e - T1p; T2T = T2R - T2S; } { E TR, T2M, TU, T2N; { E TP, TQ, TS, TT; TP = ri[WS(rs, 9)]; TQ = ii[WS(rs, 9)]; TR = FMA(T3, TP, T6 * TQ); T2M = FNMS(T6, TP, T3 * TQ); TS = ri[WS(rs, 19)]; TT = ii[WS(rs, 19)]; TU = FMA(Tx, TS, Ty * TT); T2N = FNMS(Ty, TS, Tx * TT); } TV = TR + TU; T3I = T2M + T2N; T2x = TR - TU; T2O = T2M - T2N; } { E T11, T2z, T1a, T2A; { E TY, T10, T15, T19; TY = ri[WS(rs, 16)]; T10 = ii[WS(rs, 16)]; T11 = FMA(TX, TY, TZ * T10); T2z = FNMS(TZ, TY, TX * T10); T15 = ri[WS(rs, 6)]; T19 = ii[WS(rs, 6)]; T1a = FMA(T14, T15, T18 * T19); T2A = FNMS(T18, T15, T14 * T19); } T1b = T11 + T1a; T3K = T2z + T2A; T2B = T2z - T2A; T2Q = T11 - T1a; } TW = TO - TV; T1r = T1b - T1q; T1s = TW + T1r; T3J = T3H - T3I; T3M = T3K - T3L; T4n = T3J + T3M; T3U = T3H + T3I; T3V = T3K + T3L; T43 = T3U + T3V; T2c = TO + TV; T2d = T1b + T1q; T2e = T2c + T2d; T2y = T2w + T2x; T2D = T2B + T2C; T4K = T2y + T2D; T3d = T2w - T2x; T3e = T2B - T2C; T4v = T3d + T3e; T3k = T2L + T2O; T3l = T2Q + T2T; T3m = T3k + T3l; T2P = T2L - T2O; T2U = T2Q - T2T; T2V = T2P + T2U; } { E T3y, T2a, T3x, T3O, T3Q, T3G, T3N, T3P, T3z; T3y = KP559016994 * (T1s - T29); T2a = T1s + T29; T3x = FNMS(KP250000000, T2a, TF); T3G = T3C - T3F; T3N = T3J - T3M; T3O = FNMS(KP587785252, T3N, KP951056516 * T3G); T3Q = FMA(KP951056516, T3N, KP587785252 * T3G); ri[WS(rs, 10)] = TF + T2a; T3P = T3y + T3x; ri[WS(rs, 14)] = T3P - T3Q; ri[WS(rs, 6)] = T3P + T3Q; T3z = T3x - T3y; ri[WS(rs, 2)] = T3z - T3O; ri[WS(rs, 18)] = T3z + T3O; } { E T4r, T4p, T4q, T4l, T4u, T4j, T4k, T4t, T4s; T4r = KP559016994 * (T4n - T4o); T4p = T4n + T4o; T4q = FNMS(KP250000000, T4p, T4m); T4j = T1N - T28; T4k = TW - T1r; T4l = FNMS(KP587785252, T4k, KP951056516 * T4j); T4u = FMA(KP951056516, T4k, KP587785252 * T4j); ii[WS(rs, 10)] = T4p + T4m; T4t = T4r + T4q; ii[WS(rs, 6)] = T4t - T4u; ii[WS(rs, 14)] = T4u + T4t; T4s = T4q - T4r; ii[WS(rs, 2)] = T4l + T4s; ii[WS(rs, 18)] = T4s - T4l; } { E T3R, T2i, T3S, T40, T42, T3W, T3Z, T41, T3T; T3R = KP559016994 * (T2e - T2h); T2i = T2e + T2h; T3S = FNMS(KP250000000, T2i, T2b); T3W = T3U - T3V; T3Z = T3X - T3Y; T40 = FMA(KP951056516, T3W, KP587785252 * T3Z); T42 = FNMS(KP587785252, T3W, KP951056516 * T3Z); ri[0] = T2b + T2i; T41 = T3S - T3R; ri[WS(rs, 12)] = T41 - T42; ri[WS(rs, 8)] = T41 + T42; T3T = T3R + T3S; ri[WS(rs, 4)] = T3T - T40; ri[WS(rs, 16)] = T3T + T40; } { E T4e, T45, T4f, T4d, T4i, T4b, T4c, T4h, T4g; T4e = KP559016994 * (T43 - T44); T45 = T43 + T44; T4f = FNMS(KP250000000, T45, T4a); T4b = T2c - T2d; T4c = T2f - T2g; T4d = FMA(KP951056516, T4b, KP587785252 * T4c); T4i = FNMS(KP587785252, T4b, KP951056516 * T4c); ii[0] = T45 + T4a; T4h = T4f - T4e; ii[WS(rs, 8)] = T4h - T4i; ii[WS(rs, 12)] = T4i + T4h; T4g = T4e + T4f; ii[WS(rs, 4)] = T4d + T4g; ii[WS(rs, 16)] = T4g - T4d; } { E T39, T37, T38, T2F, T3b, T2t, T2E, T3c, T3a; T39 = KP559016994 * (T2V - T36); T37 = T2V + T36; T38 = FNMS(KP250000000, T37, T2K); T2t = T2n - T2s; T2E = T2y - T2D; T2F = FNMS(KP587785252, T2E, KP951056516 * T2t); T3b = FMA(KP951056516, T2E, KP587785252 * T2t); ri[WS(rs, 15)] = T2K + T37; T3c = T39 + T38; ri[WS(rs, 11)] = T3b + T3c; ri[WS(rs, 19)] = T3c - T3b; T3a = T38 - T39; ri[WS(rs, 3)] = T2F + T3a; ri[WS(rs, 7)] = T3a - T2F; } { E T4O, T4M, T4N, T4S, T4U, T4Q, T4R, T4T, T4P; T4O = KP559016994 * (T4K - T4L); T4M = T4K + T4L; T4N = FNMS(KP250000000, T4M, T4J); T4Q = T30 - T35; T4R = T2P - T2U; T4S = FNMS(KP587785252, T4R, KP951056516 * T4Q); T4U = FMA(KP951056516, T4R, KP587785252 * T4Q); ii[WS(rs, 15)] = T4M + T4J; T4T = T4O + T4N; ii[WS(rs, 11)] = T4T - T4U; ii[WS(rs, 19)] = T4U + T4T; T4P = T4N - T4O; ii[WS(rs, 3)] = T4P - T4S; ii[WS(rs, 7)] = T4S + T4P; } { E T3q, T3s, T3t, T3j, T3v, T3f, T3i, T3w, T3u; T3q = KP559016994 * (T3m - T3p); T3s = T3m + T3p; T3t = FNMS(KP250000000, T3s, T3r); T3f = T3d - T3e; T3i = T3g - T3h; T3j = FMA(KP951056516, T3f, KP587785252 * T3i); T3v = FNMS(KP587785252, T3f, KP951056516 * T3i); ri[WS(rs, 5)] = T3r + T3s; T3w = T3t - T3q; ri[WS(rs, 13)] = T3v + T3w; ri[WS(rs, 17)] = T3w - T3v; T3u = T3q + T3t; ri[WS(rs, 1)] = T3j + T3u; ri[WS(rs, 9)] = T3u - T3j; } { E T4x, T4B, T4C, T4G, T4I, T4E, T4F, T4H, T4D; T4x = KP559016994 * (T4v - T4w); T4B = T4v + T4w; T4C = FNMS(KP250000000, T4B, T4A); T4E = T3k - T3l; T4F = T3n - T3o; T4G = FMA(KP951056516, T4E, KP587785252 * T4F); T4I = FNMS(KP587785252, T4E, KP951056516 * T4F); ii[WS(rs, 5)] = T4B + T4A; T4H = T4C - T4x; ii[WS(rs, 13)] = T4H - T4I; ii[WS(rs, 17)] = T4I + T4H; T4D = T4x + T4C; ii[WS(rs, 1)] = T4D - T4G; ii[WS(rs, 9)] = T4G + T4D; } } } } } static const tw_instr twinstr[] = { {TW_CEXP, 0, 1}, {TW_CEXP, 0, 3}, {TW_CEXP, 0, 9}, {TW_CEXP, 0, 19}, {TW_NEXT, 1, 0} }; static const ct_desc desc = { 20, "t2_20", twinstr, &GENUS, {204, 92, 72, 0}, 0, 0, 0 }; void X(codelet_t2_20) (planner *p) { X(kdft_dit_register) (p, t2_20, &desc); } #endif /* HAVE_FMA */ fftw-3.3.3/dft/scalar/codelets/t2_25.c0000644000175000017500000015245012054410301014227 00000000000000/* * Copyright (c) 2003, 2007-11 Matteo Frigo * Copyright (c) 2003, 2007-11 Massachusetts Institute of Technology * * This program is free software; you can redistribute it and/or modify * it under the terms of the GNU General Public License as published by * the Free Software Foundation; either version 2 of the License, or * (at your option) any later version. * * This program is distributed in the hope that it will be useful, * but WITHOUT ANY WARRANTY; without even the implied warranty of * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the * GNU General Public License for more details. * * You should have received a copy of the GNU General Public License * along with this program; if not, write to the Free Software * Foundation, Inc., 51 Franklin Street, Fifth Floor, Boston, MA 02110-1301 USA * */ /* This file was automatically generated --- DO NOT EDIT */ /* Generated on Sun Nov 25 07:36:11 EST 2012 */ #include "codelet-dft.h" #ifdef HAVE_FMA /* Generated by: ../../../genfft/gen_twiddle.native -fma -reorder-insns -schedule-for-pipeline -compact -variables 4 -pipeline-latency 4 -twiddle-log3 -precompute-twiddles -n 25 -name t2_25 -include t.h */ /* * This function contains 440 FP additions, 434 FP multiplications, * (or, 84 additions, 78 multiplications, 356 fused multiply/add), * 215 stack variables, 47 constants, and 100 memory accesses */ #include "t.h" static void t2_25(R *ri, R *ii, const R *W, stride rs, INT mb, INT me, INT ms) { DK(KP860541664, +0.860541664367944677098261680920518816412804187); DK(KP681693190, +0.681693190061530575150324149145440022633095390); DK(KP560319534, +0.560319534973832390111614715371676131169633784); DK(KP949179823, +0.949179823508441261575555465843363271711583843); DK(KP557913902, +0.557913902031834264187699648465567037992437152); DK(KP249506682, +0.249506682107067890488084201715862638334226305); DK(KP906616052, +0.906616052148196230441134447086066874408359177); DK(KP968479752, +0.968479752739016373193524836781420152702090879); DK(KP621716863, +0.621716863012209892444754556304102309693593202); DK(KP614372930, +0.614372930789563808870829930444362096004872855); DK(KP845997307, +0.845997307939530944175097360758058292389769300); DK(KP998026728, +0.998026728428271561952336806863450553336905220); DK(KP994076283, +0.994076283785401014123185814696322018529298887); DK(KP734762448, +0.734762448793050413546343770063151342619912334); DK(KP772036680, +0.772036680810363904029489473607579825330539880); DK(KP062914667, +0.062914667253649757225485955897349402364686947); DK(KP803003575, +0.803003575438660414833440593570376004635464850); DK(KP943557151, +0.943557151597354104399655195398983005179443399); DK(KP554608978, +0.554608978404018097464974850792216217022558774); DK(KP248028675, +0.248028675328619457762448260696444630363259177); DK(KP921177326, +0.921177326965143320250447435415066029359282231); DK(KP833417178, +0.833417178328688677408962550243238843138996060); DK(KP726211448, +0.726211448929902658173535992263577167607493062); DK(KP525970792, +0.525970792408939708442463226536226366643874659); DK(KP541454447, +0.541454447536312777046285590082819509052033189); DK(KP242145790, +0.242145790282157779872542093866183953459003101); DK(KP992114701, +0.992114701314477831049793042785778521453036709); DK(KP559154169, +0.559154169276087864842202529084232643714075927); DK(KP683113946, +0.683113946453479238701949862233725244439656928); DK(KP851038619, +0.851038619207379630836264138867114231259902550); DK(KP912575812, +0.912575812670962425556968549836277086778922727); DK(KP912018591, +0.912018591466481957908415381764119056233607330); DK(KP470564281, +0.470564281212251493087595091036643380879947982); DK(KP968583161, +0.968583161128631119490168375464735813836012403); DK(KP827271945, +0.827271945972475634034355757144307982555673741); DK(KP126329378, +0.126329378446108174786050455341811215027378105); DK(KP904730450, +0.904730450839922351881287709692877908104763647); DK(KP831864738, +0.831864738706457140726048799369896829771167132); DK(KP871714437, +0.871714437527667770979999223229522602943903653); DK(KP549754652, +0.549754652192770074288023275540779861653779767); DK(KP634619297, +0.634619297544148100711287640319130485732531031); DK(KP939062505, +0.939062505817492352556001843133229685779824606); DK(KP256756360, +0.256756360367726783319498520922669048172391148); DK(KP951056516, +0.951056516295153572116439333379382143405698634); DK(KP559016994, +0.559016994374947424102293417182819058860154590); DK(KP250000000, +0.250000000000000000000000000000000000000000000); DK(KP618033988, +0.618033988749894848204586834365638117720309180); { INT m; for (m = mb, W = W + (mb * 8); m < me; m = m + 1, ri = ri + ms, ii = ii + ms, W = W + 8, MAKE_VOLATILE_STRIDE(50, rs)) { E T8c, T7k, T7i, T8i, T8g, T8b, T7j, T7b, T8d, T8h; { E T2, T8, T3, T6, Tk, Tv, TS, T4, Ta, TD, T2L, T10, Tm, T5, Tc; T2 = W[0]; T8 = W[4]; T3 = W[2]; T6 = W[3]; Tk = W[6]; Tv = T2 * T8; TS = T3 * T8; T4 = T2 * T3; Ta = T2 * T6; TD = T8 * Tk; T2L = T2 * Tk; T10 = T3 * Tk; Tm = W[7]; T5 = W[1]; Tc = W[5]; { E T7G, T86, T4s, T6a, T4g, TN, T4f, T7C, T7s, T7B, T5q, T6k, T3a, T5j, T6n; E T6m, T5g, T4a, T5n, T6j, T6C, T4G, T6z, T4z, T1v, T3t, T6y, T4w, T6B, T4D; E T6v, T4O, T6s, T4V, T21, T3H, T6r, T4S, T6u, T4L, T26, T3K, T5a, T2A, T3U; E T53, T2c, T3M, T2k, T3O; { E T11, T1b, Tb, T19, T7, T2m, TT, T15, T2Q, TX, T2p, T1g, T2a, T2e, T2i; E T27, T1c, T1O, T1K, T1q, T1m, T2x, T2t, T1W, T1S, T2G, T3Y, T2N, T5p, T38; E T48, T5i, T2K, T40, T2S, T41; { E T2M, T1j, T1l, T2X, T2U, T35, T31, T7r, T7p, T7o, T2O, T2R; { E T1, Tj, T4j, TK, T4q, TC, T4o, Tt, T4l; { E TE, Tw, TI, TA, Th, Tr, Tn, Td, Te, Ti, T14, T2P, TH, Tx, TB; T1 = ri[0]; T11 = FMA(T6, Tm, T10); T14 = T3 * Tm; T2P = T2 * Tm; TH = T8 * Tm; T2M = FMA(T5, Tm, T2L); T1b = FNMS(T5, T3, Ta); Tb = FMA(T5, T3, Ta); T19 = FMA(T5, T6, T4); T7 = FNMS(T5, T6, T4); T2m = FNMS(T6, Tc, TS); TT = FMA(T6, Tc, TS); TE = FMA(Tc, Tm, TD); T1j = FMA(T5, Tc, Tv); Tw = FNMS(T5, Tc, Tv); { E TW, Tz, T1f, T2d; TW = T3 * Tc; Tz = T2 * Tc; T15 = FNMS(T6, Tk, T14); T2Q = FNMS(T5, Tk, T2P); TI = FNMS(Tc, Tk, TH); T1f = T19 * Tc; T2d = T19 * Tk; { E T2h, T1a, Tg, Tq; T2h = T19 * Tm; T1a = T19 * T8; Tg = T7 * Tc; Tq = T7 * Tm; { E Tl, T9, T1p, T1k; Tl = T7 * Tk; T9 = T7 * T8; T1p = T1j * Tm; T1k = T1j * Tk; { E T34, T30, T1N, T1J; T34 = TT * Tm; T30 = TT * Tk; T1N = Tw * Tm; T1J = Tw * Tk; TX = FNMS(T6, T8, TW); T2p = FMA(T6, T8, TW); TA = FMA(T5, T8, Tz); T1l = FNMS(T5, T8, Tz); T1g = FMA(T1b, T8, T1f); T2a = FNMS(T1b, T8, T1f); T2e = FMA(T1b, Tm, T2d); T2i = FNMS(T1b, Tk, T2h); T27 = FMA(T1b, Tc, T1a); T1c = FNMS(T1b, Tc, T1a); T2X = FMA(Tb, T8, Tg); Th = FNMS(Tb, T8, Tg); Tr = FNMS(Tb, Tk, Tq); Tn = FMA(Tb, Tm, Tl); Td = FMA(Tb, Tc, T9); T2U = FNMS(Tb, Tc, T9); T35 = FNMS(TX, Tk, T34); T31 = FMA(TX, Tm, T30); T1O = FNMS(TA, Tk, T1N); T1K = FMA(TA, Tm, T1J); T1q = FNMS(T1l, Tk, T1p); T1m = FMA(T1l, Tm, T1k); { E T2w, T2s, T1V, T1R; T2w = T27 * Tm; T2s = T27 * Tk; T1V = Td * Tm; T1R = Td * Tk; T2x = FNMS(T2a, Tk, T2w); T2t = FMA(T2a, Tm, T2s); T1W = FNMS(Th, Tk, T1V); T1S = FMA(Th, Tm, T1R); T7r = ii[0]; Te = ri[WS(rs, 5)]; Ti = ii[WS(rs, 5)]; } } } } } { E TF, TJ, Tf, T4i, TG, T4p; TF = ri[WS(rs, 15)]; TJ = ii[WS(rs, 15)]; Tf = Td * Te; T4i = Td * Ti; TG = TE * TF; T4p = TE * TJ; Tj = FMA(Th, Ti, Tf); T4j = FNMS(Th, Te, T4i); TK = FMA(TI, TJ, TG); T4q = FNMS(TI, TF, T4p); } Tx = ri[WS(rs, 10)]; TB = ii[WS(rs, 10)]; { E To, Ts, Ty, T4n, Tp, T4k; To = ri[WS(rs, 20)]; Ts = ii[WS(rs, 20)]; Ty = Tw * Tx; T4n = Tw * TB; Tp = Tn * To; T4k = Tn * Ts; TC = FMA(TA, TB, Ty); T4o = FNMS(TA, Tx, T4n); Tt = FMA(Tr, Ts, Tp); T4l = FNMS(Tr, To, T4k); } } { E TL, T7F, T4r, Tu, T7E, T4m, TM; TL = TC + TK; T7F = TC - TK; T4r = T4o - T4q; T7p = T4o + T4q; Tu = Tj + Tt; T7E = Tj - Tt; T4m = T4j - T4l; T7o = T4j + T4l; T7G = FMA(KP618033988, T7F, T7E); T86 = FNMS(KP618033988, T7E, T7F); T4s = FMA(KP618033988, T4r, T4m); T6a = FNMS(KP618033988, T4m, T4r); T4g = Tu - TL; TM = Tu + TL; TN = T1 + TM; T4f = FNMS(KP250000000, TM, T1); } } { E T2D, T2F, T7q, T2E, T3X; T2D = ri[WS(rs, 3)]; T2F = ii[WS(rs, 3)]; T7C = T7o - T7p; T7q = T7o + T7p; T2E = T3 * T2D; T3X = T3 * T2F; { E T2V, T2W, T2Y, T32, T36; T2V = ri[WS(rs, 13)]; T7s = T7q + T7r; T7B = FNMS(KP250000000, T7q, T7r); T2G = FMA(T6, T2F, T2E); T3Y = FNMS(T6, T2D, T3X); T2W = T2U * T2V; T2Y = ii[WS(rs, 13)]; T32 = ri[WS(rs, 18)]; T36 = ii[WS(rs, 18)]; { E T2H, T2I, T2J, T3Z; { E T2Z, T45, T37, T47, T44, T33, T46; T2H = ri[WS(rs, 8)]; T2Z = FMA(T2X, T2Y, T2W); T44 = T2U * T2Y; T33 = T31 * T32; T46 = T31 * T36; T2I = T1j * T2H; T45 = FNMS(T2X, T2V, T44); T37 = FMA(T35, T36, T33); T47 = FNMS(T35, T32, T46); T2J = ii[WS(rs, 8)]; T2N = ri[WS(rs, 23)]; T5p = T2Z - T37; T38 = T2Z + T37; T48 = T45 + T47; T5i = T47 - T45; T3Z = T1j * T2J; T2O = T2M * T2N; T2R = ii[WS(rs, 23)]; } T2K = FMA(T1l, T2J, T2I); T40 = FNMS(T1l, T2H, T3Z); } } } T2S = FMA(T2Q, T2R, T2O); T41 = T2M * T2R; } { E TR, T3h, T1t, T4F, T3r, T4y, TZ, T3j, T17, T3l; { E T12, T16, T13, T3k; { E TO, TP, T5m, T5l, TQ; { E T2T, T5o, T42, T5f, T39; TO = ri[WS(rs, 1)]; T2T = T2K + T2S; T5o = T2K - T2S; T42 = FNMS(T2Q, T2N, T41); TP = T2 * TO; T5q = FMA(KP618033988, T5p, T5o); T6k = FNMS(KP618033988, T5o, T5p); T5f = T38 - T2T; T39 = T2T + T38; { E T43, T5h, T5e, T49; T43 = T40 + T42; T5h = T42 - T40; T5e = FNMS(KP250000000, T39, T2G); T3a = T2G + T39; T5j = FMA(KP618033988, T5i, T5h); T6n = FNMS(KP618033988, T5h, T5i); T5m = T48 - T43; T49 = T43 + T48; T6m = FMA(KP559016994, T5f, T5e); T5g = FNMS(KP559016994, T5f, T5e); T5l = FNMS(KP250000000, T49, T3Y); T4a = T3Y + T49; TQ = ii[WS(rs, 1)]; } } { E T1n, T1r, T1i, T1o, T3o, T3p; { E T1d, T1h, T1e, T3n, T3g; T1d = ri[WS(rs, 11)]; T1h = ii[WS(rs, 11)]; T5n = FNMS(KP559016994, T5m, T5l); T6j = FMA(KP559016994, T5m, T5l); TR = FMA(T5, TQ, TP); T3g = T2 * TQ; T1e = T1c * T1d; T3n = T1c * T1h; T1n = ri[WS(rs, 16)]; T3h = FNMS(T5, TO, T3g); T1r = ii[WS(rs, 16)]; T1i = FMA(T1g, T1h, T1e); T1o = T1m * T1n; T3o = FNMS(T1g, T1d, T3n); T3p = T1m * T1r; } { E TU, TY, TV, T3i, T3q, T1s; TU = ri[WS(rs, 6)]; T1s = FMA(T1q, T1r, T1o); TY = ii[WS(rs, 6)]; T3q = FNMS(T1q, T1n, T3p); TV = TT * TU; T1t = T1i + T1s; T4F = T1s - T1i; T3i = TT * TY; T3r = T3o + T3q; T4y = T3q - T3o; T12 = ri[WS(rs, 21)]; T16 = ii[WS(rs, 21)]; TZ = FMA(TX, TY, TV); T3j = FNMS(TX, TU, T3i); T13 = T11 * T12; T3k = T11 * T16; } } } T17 = FMA(T15, T16, T13); T3l = FNMS(T15, T12, T3k); } { E T1z, T3v, T4N, T1Z, T3F, T4U, T1D, T3x, T1H, T3z; { E T1E, T1G, T1F, T3y; { E T1w, T1y, T1x, T4v, T4C, T4u, T4B, T3u, T18, T4E; T1w = ri[WS(rs, 4)]; T1y = ii[WS(rs, 4)]; T18 = TZ + T17; T4E = T17 - TZ; { E T3m, T4x, T1u, T3s; T3m = T3j + T3l; T4x = T3j - T3l; T1x = T7 * T1w; T6C = FNMS(KP618033988, T4E, T4F); T4G = FMA(KP618033988, T4F, T4E); T1u = T18 + T1t; T4v = T18 - T1t; T6z = FMA(KP618033988, T4x, T4y); T4z = FNMS(KP618033988, T4y, T4x); T3s = T3m + T3r; T4C = T3m - T3r; T1v = TR + T1u; T4u = FNMS(KP250000000, T1u, TR); T3t = T3h + T3s; T4B = FNMS(KP250000000, T3s, T3h); T3u = T7 * T1y; } T6y = FNMS(KP559016994, T4v, T4u); T4w = FMA(KP559016994, T4v, T4u); T6B = FNMS(KP559016994, T4C, T4B); T4D = FMA(KP559016994, T4C, T4B); T1z = FMA(Tb, T1y, T1x); T3v = FNMS(Tb, T1w, T3u); } { E T1Q, T3C, T1Y, T3E; { E T1L, T1P, T1T, T1X, T1M, T3B, T1U, T3D; T1L = ri[WS(rs, 14)]; T1P = ii[WS(rs, 14)]; T1T = ri[WS(rs, 19)]; T1X = ii[WS(rs, 19)]; T1M = T1K * T1L; T3B = T1K * T1P; T1U = T1S * T1T; T3D = T1S * T1X; T1Q = FMA(T1O, T1P, T1M); T3C = FNMS(T1O, T1L, T3B); T1Y = FMA(T1W, T1X, T1U); T3E = FNMS(T1W, T1T, T3D); } { E T1A, T1C, T1B, T3w; T1A = ri[WS(rs, 9)]; T1C = ii[WS(rs, 9)]; T4N = T1Y - T1Q; T1Z = T1Q + T1Y; T3F = T3C + T3E; T4U = T3E - T3C; T1B = T8 * T1A; T3w = T8 * T1C; T1E = ri[WS(rs, 24)]; T1G = ii[WS(rs, 24)]; T1D = FMA(Tc, T1C, T1B); T3x = FNMS(Tc, T1A, T3w); T1F = Tk * T1E; T3y = Tk * T1G; } } T1H = FMA(Tm, T1G, T1F); T3z = FNMS(Tm, T1E, T3y); } { E T2f, T2j, T2g, T3N; { E T23, T25, T24, T4R, T4K, T4Q, T4J, T3J, T1I, T4M; T23 = ri[WS(rs, 2)]; T25 = ii[WS(rs, 2)]; T1I = T1D + T1H; T4M = T1H - T1D; { E T3A, T4T, T20, T3G; T3A = T3x + T3z; T4T = T3z - T3x; T24 = T19 * T23; T6v = FNMS(KP618033988, T4M, T4N); T4O = FMA(KP618033988, T4N, T4M); T20 = T1I + T1Z; T4R = T1I - T1Z; T6s = FNMS(KP618033988, T4T, T4U); T4V = FMA(KP618033988, T4U, T4T); T3G = T3A + T3F; T4K = T3F - T3A; T21 = T1z + T20; T4Q = FNMS(KP250000000, T20, T1z); T3H = T3v + T3G; T4J = FNMS(KP250000000, T3G, T3v); T3J = T19 * T25; } T6r = FNMS(KP559016994, T4R, T4Q); T4S = FMA(KP559016994, T4R, T4Q); T6u = FMA(KP559016994, T4K, T4J); T4L = FNMS(KP559016994, T4K, T4J); T26 = FMA(T1b, T25, T24); T3K = FNMS(T1b, T23, T3J); } { E T2r, T3R, T2z, T3T; { E T2n, T2q, T2u, T2y, T2o, T3Q, T2v, T3S; T2n = ri[WS(rs, 12)]; T2q = ii[WS(rs, 12)]; T2u = ri[WS(rs, 17)]; T2y = ii[WS(rs, 17)]; T2o = T2m * T2n; T3Q = T2m * T2q; T2v = T2t * T2u; T3S = T2t * T2y; T2r = FMA(T2p, T2q, T2o); T3R = FNMS(T2p, T2n, T3Q); T2z = FMA(T2x, T2y, T2v); T3T = FNMS(T2x, T2u, T3S); } { E T28, T2b, T29, T3L; T28 = ri[WS(rs, 7)]; T2b = ii[WS(rs, 7)]; T5a = T2z - T2r; T2A = T2r + T2z; T3U = T3R + T3T; T53 = T3R - T3T; T29 = T27 * T28; T3L = T27 * T2b; T2f = ri[WS(rs, 22)]; T2j = ii[WS(rs, 22)]; T2c = FMA(T2a, T2b, T29); T3M = FNMS(T2a, T28, T3L); T2g = T2e * T2f; T3N = T2e * T2j; } } T2k = FMA(T2i, T2j, T2g); T3O = FNMS(T2i, T2f, T3N); } } } } { E T7l, T5b, T6d, T54, T6g, T51, T6f, T7m, T6c, T58, T4e, T4c, T7A, T7y, T4d; E T3f; { E T7w, T22, T7x, T3b, T3I, T3c, T3e, T3d; T7l = T3t + T3H; T3I = T3t - T3H; { E T2l, T59, T3P, T52; T2l = T2c + T2k; T59 = T2k - T2c; T3P = T3M + T3O; T52 = T3O - T3M; T5b = FMA(KP618033988, T5a, T59); T6d = FNMS(KP618033988, T59, T5a); { E T50, T2B, T57, T3V; T50 = T2A - T2l; T2B = T2l + T2A; T54 = FNMS(KP618033988, T53, T52); T6g = FMA(KP618033988, T52, T53); T57 = T3U - T3P; T3V = T3P + T3U; { E T4Z, T2C, T56, T3W, T4b; T4Z = FNMS(KP250000000, T2B, T26); T2C = T26 + T2B; T56 = FNMS(KP250000000, T3V, T3K); T3W = T3K + T3V; T7w = T1v - T21; T22 = T1v + T21; T51 = FNMS(KP559016994, T50, T4Z); T6f = FMA(KP559016994, T50, T4Z); T4b = T3W - T4a; T7m = T3W + T4a; T6c = FMA(KP559016994, T57, T56); T58 = FNMS(KP559016994, T57, T56); T7x = T2C - T3a; T3b = T2C + T3a; T4e = FNMS(KP618033988, T3I, T4b); T4c = FMA(KP618033988, T4b, T3I); } } } T3c = T22 + T3b; T3e = T22 - T3b; ri[0] = TN + T3c; T3d = FNMS(KP250000000, T3c, TN); T7A = FNMS(KP618033988, T7w, T7x); T7y = FMA(KP618033988, T7x, T7w); T4d = FNMS(KP559016994, T3e, T3d); T3f = FMA(KP559016994, T3e, T3d); } { E T69, T85, T7Y, T68, T66, T84, T82, T7X, T67, T5Z; { E T4t, T5H, T5Q, T7T, T7H, T5P, T5M, T5L, T5A, T7O, T5D, T7P, T7K, T7M, T5u; E T5w, T5K, T63, T61, T5U, T7D, T7z, T7v; { E T7u, T7t, T4h, T7n; T69 = FNMS(KP559016994, T4g, T4f); T4h = FMA(KP559016994, T4g, T4f); T7u = T7l - T7m; T7n = T7l + T7m; ri[WS(rs, 5)] = FMA(KP951056516, T4c, T3f); ri[WS(rs, 20)] = FNMS(KP951056516, T4c, T3f); ri[WS(rs, 15)] = FMA(KP951056516, T4e, T4d); ri[WS(rs, 10)] = FNMS(KP951056516, T4e, T4d); ii[0] = T7n + T7s; T7t = FNMS(KP250000000, T7n, T7s); T4t = FMA(KP951056516, T4s, T4h); T5H = FNMS(KP951056516, T4s, T4h); T7D = FMA(KP559016994, T7C, T7B); T85 = FNMS(KP559016994, T7C, T7B); T7z = FNMS(KP559016994, T7u, T7t); T7v = FMA(KP559016994, T7u, T7t); } { E T5I, T5J, T5S, T4P, T5y, T4I, T5C, T5s, T4W, T5T, T55, T5c; { E T4A, T4H, T5k, T5r; T5Q = FNMS(KP951056516, T4z, T4w); T4A = FMA(KP951056516, T4z, T4w); T7T = FMA(KP951056516, T7G, T7D); T7H = FNMS(KP951056516, T7G, T7D); ii[WS(rs, 20)] = FMA(KP951056516, T7y, T7v); ii[WS(rs, 5)] = FNMS(KP951056516, T7y, T7v); ii[WS(rs, 15)] = FNMS(KP951056516, T7A, T7z); ii[WS(rs, 10)] = FMA(KP951056516, T7A, T7z); T4H = FMA(KP951056516, T4G, T4D); T5P = FNMS(KP951056516, T4G, T4D); T5I = FMA(KP951056516, T5j, T5g); T5k = FNMS(KP951056516, T5j, T5g); T5r = FNMS(KP951056516, T5q, T5n); T5J = FMA(KP951056516, T5q, T5n); T5S = FNMS(KP951056516, T4O, T4L); T4P = FMA(KP951056516, T4O, T4L); T5y = FNMS(KP256756360, T4A, T4H); T4I = FMA(KP256756360, T4H, T4A); T5C = FNMS(KP939062505, T5k, T5r); T5s = FMA(KP939062505, T5r, T5k); T4W = FNMS(KP951056516, T4V, T4S); T5T = FMA(KP951056516, T4V, T4S); T5M = FMA(KP951056516, T54, T51); T55 = FNMS(KP951056516, T54, T51); T5c = FMA(KP951056516, T5b, T58); T5L = FNMS(KP951056516, T5b, T58); } { E T4Y, T5t, T5z, T4X; T5z = FNMS(KP634619297, T4P, T4W); T4X = FMA(KP634619297, T4W, T4P); { E T5B, T5d, T7I, T7J; T5B = FNMS(KP549754652, T55, T5c); T5d = FMA(KP549754652, T5c, T55); T7I = FNMS(KP871714437, T5z, T5y); T5A = FMA(KP871714437, T5z, T5y); T4Y = FMA(KP871714437, T4X, T4I); T7O = FNMS(KP871714437, T4X, T4I); T7J = FMA(KP831864738, T5C, T5B); T5D = FNMS(KP831864738, T5C, T5B); T5t = FMA(KP831864738, T5s, T5d); T7P = FNMS(KP831864738, T5s, T5d); T7K = FMA(KP904730450, T7J, T7I); T7M = FNMS(KP904730450, T7J, T7I); } T5u = FMA(KP904730450, T5t, T4Y); T5w = FNMS(KP904730450, T5t, T4Y); } T5K = FNMS(KP126329378, T5J, T5I); T63 = FMA(KP126329378, T5I, T5J); T61 = FNMS(KP827271945, T5S, T5T); T5U = FMA(KP827271945, T5T, T5S); } { E T65, T81, T62, T80, T7W, T5W, T5Y; { E T5O, T5V, T64, T5N; ri[WS(rs, 1)] = FMA(KP968583161, T5u, T4t); T64 = FMA(KP470564281, T5L, T5M); T5N = FNMS(KP470564281, T5M, T5L); { E T60, T5R, T7U, T7V; T60 = FNMS(KP634619297, T5P, T5Q); T5R = FMA(KP634619297, T5Q, T5P); T7U = FMA(KP912018591, T64, T63); T65 = FNMS(KP912018591, T64, T63); T5O = FNMS(KP912018591, T5N, T5K); T81 = FMA(KP912018591, T5N, T5K); T7V = FNMS(KP912575812, T61, T60); T62 = FMA(KP912575812, T61, T60); T5V = FNMS(KP912575812, T5U, T5R); T80 = FMA(KP912575812, T5U, T5R); T7W = FMA(KP851038619, T7V, T7U); T7Y = FNMS(KP851038619, T7V, T7U); ii[WS(rs, 1)] = FMA(KP968583161, T7K, T7H); } T5W = FNMS(KP851038619, T5V, T5O); T5Y = FMA(KP851038619, T5V, T5O); } { E T5G, T5E, T7S, T7Q, T7L, T5F, T5x, T5v, T5X, T7R, T7N; T5G = FNMS(KP683113946, T5A, T5D); T5E = FMA(KP559154169, T5D, T5A); ii[WS(rs, 4)] = FNMS(KP992114701, T7W, T7T); ri[WS(rs, 4)] = FNMS(KP992114701, T5W, T5H); T5v = FNMS(KP242145790, T5u, T4t); T7S = FNMS(KP683113946, T7O, T7P); T7Q = FMA(KP559154169, T7P, T7O); T7L = FNMS(KP242145790, T7K, T7H); T5F = FNMS(KP541454447, T5w, T5v); T5x = FMA(KP541454447, T5w, T5v); T68 = FMA(KP525970792, T62, T65); T66 = FNMS(KP726211448, T65, T62); ri[WS(rs, 11)] = FNMS(KP833417178, T5G, T5F); ri[WS(rs, 16)] = FMA(KP833417178, T5G, T5F); ri[WS(rs, 21)] = FNMS(KP921177326, T5E, T5x); ri[WS(rs, 6)] = FMA(KP921177326, T5E, T5x); T7R = FNMS(KP541454447, T7M, T7L); T7N = FMA(KP541454447, T7M, T7L); T5X = FMA(KP248028675, T5W, T5H); ii[WS(rs, 11)] = FMA(KP833417178, T7S, T7R); ii[WS(rs, 16)] = FNMS(KP833417178, T7S, T7R); ii[WS(rs, 21)] = FMA(KP921177326, T7Q, T7N); ii[WS(rs, 6)] = FNMS(KP921177326, T7Q, T7N); T84 = FNMS(KP525970792, T80, T81); T82 = FMA(KP726211448, T81, T80); T7X = FMA(KP248028675, T7W, T7T); T67 = FNMS(KP554608978, T5Y, T5X); T5Z = FMA(KP554608978, T5Y, T5X); } } } { E T6b, T6T, T8j, T87, T72, T71, T6P, T8r, T6M, T8q, T7f, T6W, T8m, T8o, T6I; E T6G, T7d, T76, T7g, T6Z, T83, T7Z; ri[WS(rs, 14)] = FNMS(KP943557151, T68, T67); ri[WS(rs, 19)] = FMA(KP943557151, T68, T67); ri[WS(rs, 24)] = FMA(KP803003575, T66, T5Z); ri[WS(rs, 9)] = FNMS(KP803003575, T66, T5Z); T83 = FNMS(KP554608978, T7Y, T7X); T7Z = FMA(KP554608978, T7Y, T7X); T6b = FMA(KP951056516, T6a, T69); T6T = FNMS(KP951056516, T6a, T69); ii[WS(rs, 14)] = FMA(KP943557151, T84, T83); ii[WS(rs, 19)] = FNMS(KP943557151, T84, T83); ii[WS(rs, 24)] = FMA(KP803003575, T82, T7Z); ii[WS(rs, 9)] = FNMS(KP803003575, T82, T7Z); { E T6X, T6Y, T74, T6N, T6i, T75, T6U, T6V, T6t, T6L, T6E, T6O, T6p, T6w; { E T6A, T6D, T6e, T6h, T6l, T6o; T6X = FNMS(KP951056516, T6d, T6c); T6e = FMA(KP951056516, T6d, T6c); T6h = FMA(KP951056516, T6g, T6f); T6Y = FNMS(KP951056516, T6g, T6f); T74 = FMA(KP951056516, T6z, T6y); T6A = FNMS(KP951056516, T6z, T6y); T8j = FNMS(KP951056516, T86, T85); T87 = FMA(KP951056516, T86, T85); T6N = FNMS(KP062914667, T6e, T6h); T6i = FMA(KP062914667, T6h, T6e); T6D = FMA(KP951056516, T6C, T6B); T75 = FNMS(KP951056516, T6C, T6B); T6U = FMA(KP951056516, T6k, T6j); T6l = FNMS(KP951056516, T6k, T6j); T6o = FNMS(KP951056516, T6n, T6m); T6V = FMA(KP951056516, T6n, T6m); T72 = FMA(KP951056516, T6s, T6r); T6t = FNMS(KP951056516, T6s, T6r); T6L = FNMS(KP939062505, T6A, T6D); T6E = FMA(KP939062505, T6D, T6A); T6O = FMA(KP827271945, T6l, T6o); T6p = FNMS(KP827271945, T6o, T6l); T6w = FMA(KP951056516, T6v, T6u); T71 = FNMS(KP951056516, T6v, T6u); } { E T8k, T6q, T6K, T6x, T8l, T6F; T8k = FMA(KP772036680, T6O, T6N); T6P = FNMS(KP772036680, T6O, T6N); T6q = FMA(KP772036680, T6p, T6i); T8r = FNMS(KP772036680, T6p, T6i); T6K = FMA(KP126329378, T6t, T6w); T6x = FNMS(KP126329378, T6w, T6t); T8l = FNMS(KP734762448, T6L, T6K); T6M = FMA(KP734762448, T6L, T6K); T6F = FNMS(KP734762448, T6E, T6x); T8q = FMA(KP734762448, T6E, T6x); T7f = FNMS(KP062914667, T6U, T6V); T6W = FMA(KP062914667, T6V, T6U); T8m = FMA(KP994076283, T8l, T8k); T8o = FNMS(KP994076283, T8l, T8k); T6I = FMA(KP994076283, T6F, T6q); T6G = FNMS(KP994076283, T6F, T6q); } T7d = FNMS(KP549754652, T74, T75); T76 = FMA(KP549754652, T75, T74); T7g = FNMS(KP634619297, T6X, T6Y); T6Z = FMA(KP634619297, T6Y, T6X); } { E T88, T7h, T70, T8f, T7c, T73; ri[WS(rs, 3)] = FMA(KP998026728, T6G, T6b); T88 = FMA(KP845997307, T7g, T7f); T7h = FNMS(KP845997307, T7g, T7f); T70 = FMA(KP845997307, T6Z, T6W); T8f = FNMS(KP845997307, T6Z, T6W); T7c = FMA(KP470564281, T71, T72); T73 = FNMS(KP470564281, T72, T71); ii[WS(rs, 3)] = FNMS(KP998026728, T8m, T8j); { E T7e, T8e, T8a, T78, T7a, T8u, T8s, T8t, T8p, T79; { E T6S, T6Q, T6H, T89, T77, T6J, T6R, T8n; T6S = FMA(KP614372930, T6M, T6P); T6Q = FNMS(KP621716863, T6P, T6M); T89 = FNMS(KP968479752, T7d, T7c); T7e = FMA(KP968479752, T7d, T7c); T77 = FMA(KP968479752, T76, T73); T8e = FNMS(KP968479752, T76, T73); T8a = FMA(KP906616052, T89, T88); T8c = FNMS(KP906616052, T89, T88); T78 = FMA(KP906616052, T77, T70); T7a = FNMS(KP906616052, T77, T70); T6H = FNMS(KP249506682, T6G, T6b); ii[WS(rs, 2)] = FNMS(KP998026728, T8a, T87); ri[WS(rs, 2)] = FMA(KP998026728, T78, T6T); T8u = FNMS(KP614372930, T8q, T8r); T8s = FMA(KP621716863, T8r, T8q); T6J = FNMS(KP557913902, T6I, T6H); T6R = FMA(KP557913902, T6I, T6H); T8n = FMA(KP249506682, T8m, T8j); ri[WS(rs, 18)] = FNMS(KP949179823, T6S, T6R); ri[WS(rs, 13)] = FMA(KP949179823, T6S, T6R); ri[WS(rs, 8)] = FMA(KP943557151, T6Q, T6J); ri[WS(rs, 23)] = FNMS(KP943557151, T6Q, T6J); T8t = FNMS(KP557913902, T8o, T8n); T8p = FMA(KP557913902, T8o, T8n); } T7k = FNMS(KP560319534, T7e, T7h); T7i = FMA(KP681693190, T7h, T7e); ii[WS(rs, 23)] = FMA(KP943557151, T8s, T8p); ii[WS(rs, 8)] = FNMS(KP943557151, T8s, T8p); ii[WS(rs, 13)] = FMA(KP949179823, T8u, T8t); ii[WS(rs, 18)] = FNMS(KP949179823, T8u, T8t); T79 = FNMS(KP249506682, T78, T6T); T8i = FNMS(KP560319534, T8e, T8f); T8g = FMA(KP681693190, T8f, T8e); T8b = FMA(KP249506682, T8a, T87); T7j = FMA(KP557913902, T7a, T79); T7b = FNMS(KP557913902, T7a, T79); } } } } } } } ri[WS(rs, 12)] = FNMS(KP949179823, T7k, T7j); ri[WS(rs, 17)] = FMA(KP949179823, T7k, T7j); ri[WS(rs, 7)] = FMA(KP860541664, T7i, T7b); ri[WS(rs, 22)] = FNMS(KP860541664, T7i, T7b); T8d = FMA(KP557913902, T8c, T8b); T8h = FNMS(KP557913902, T8c, T8b); ii[WS(rs, 12)] = FNMS(KP949179823, T8i, T8h); ii[WS(rs, 17)] = FMA(KP949179823, T8i, T8h); ii[WS(rs, 22)] = FNMS(KP860541664, T8g, T8d); ii[WS(rs, 7)] = FMA(KP860541664, T8g, T8d); } } } static const tw_instr twinstr[] = { {TW_CEXP, 0, 1}, {TW_CEXP, 0, 3}, {TW_CEXP, 0, 9}, {TW_CEXP, 0, 24}, {TW_NEXT, 1, 0} }; static const ct_desc desc = { 25, "t2_25", twinstr, &GENUS, {84, 78, 356, 0}, 0, 0, 0 }; void X(codelet_t2_25) (planner *p) { X(kdft_dit_register) (p, t2_25, &desc); } #else /* HAVE_FMA */ /* Generated by: ../../../genfft/gen_twiddle.native -compact -variables 4 -pipeline-latency 4 -twiddle-log3 -precompute-twiddles -n 25 -name t2_25 -include t.h */ /* * This function contains 440 FP additions, 340 FP multiplications, * (or, 280 additions, 180 multiplications, 160 fused multiply/add), * 149 stack variables, 20 constants, and 100 memory accesses */ #include "t.h" static void t2_25(R *ri, R *ii, const R *W, stride rs, INT mb, INT me, INT ms) { DK(KP998026728, +0.998026728428271561952336806863450553336905220); DK(KP062790519, +0.062790519529313376076178224565631133122484832); DK(KP425779291, +0.425779291565072648862502445744251703979973042); DK(KP904827052, +0.904827052466019527713668647932697593970413911); DK(KP992114701, +0.992114701314477831049793042785778521453036709); DK(KP125333233, +0.125333233564304245373118759816508793942918247); DK(KP637423989, +0.637423989748689710176712811676016195434917298); DK(KP770513242, +0.770513242775789230803009636396177847271667672); DK(KP684547105, +0.684547105928688673732283357621209269889519233); DK(KP728968627, +0.728968627421411523146730319055259111372571664); DK(KP481753674, +0.481753674101715274987191502872129653528542010); DK(KP876306680, +0.876306680043863587308115903922062583399064238); DK(KP844327925, +0.844327925502015078548558063966681505381659241); DK(KP535826794, +0.535826794978996618271308767867639978063575346); DK(KP248689887, +0.248689887164854788242283746006447968417567406); DK(KP968583161, +0.968583161128631119490168375464735813836012403); DK(KP587785252, +0.587785252292473129168705954639072768597652438); DK(KP951056516, +0.951056516295153572116439333379382143405698634); DK(KP250000000, +0.250000000000000000000000000000000000000000000); DK(KP559016994, +0.559016994374947424102293417182819058860154590); { INT m; for (m = mb, W = W + (mb * 8); m < me; m = m + 1, ri = ri + ms, ii = ii + ms, W = W + 8, MAKE_VOLATILE_STRIDE(50, rs)) { E T2, T5, T3, T6, T8, Td, T16, T14, Te, T9, T21, T23, Tx, TR, T1g; E TB, T1f, TV, T1Q, Tg, T1S, Tk, T18, T2s, T1c, T2q, Tn, To, Tp, Tr; E T28, T2x, TY, T2k, T2m, T2v, TG, TE, T10, T1h, T1E, T26, T1B, T1G, T1V; E T1X, T1z, T1j; { E Tw, TT, Tz, TQ, Tv, TU, TA, TP; { E T4, Tc, T7, Tb; T2 = W[0]; T5 = W[1]; T3 = W[2]; T6 = W[3]; T4 = T2 * T3; Tc = T5 * T3; T7 = T5 * T6; Tb = T2 * T6; T8 = T4 - T7; Td = Tb + Tc; T16 = Tb - Tc; T14 = T4 + T7; Te = W[5]; Tw = T5 * Te; TT = T3 * Te; Tz = T2 * Te; TQ = T6 * Te; T9 = W[4]; Tv = T2 * T9; TU = T6 * T9; TA = T5 * T9; TP = T3 * T9; } T21 = TP - TQ; T23 = TT + TU; { E T15, T17, Ta, Tf, T1a, T1b, Ti, Tj; Tx = Tv - Tw; TR = TP + TQ; T1g = Tz - TA; TB = Tz + TA; T1f = Tv + Tw; TV = TT - TU; T15 = T14 * T9; T17 = T16 * Te; T1Q = T15 + T17; Ta = T8 * T9; Tf = Td * Te; Tg = Ta + Tf; T1a = T14 * Te; T1b = T16 * T9; T1S = T1a - T1b; Ti = T8 * Te; Tj = Td * T9; Tk = Ti - Tj; T18 = T15 - T17; T2s = Ti + Tj; T1c = T1a + T1b; T2q = Ta - Tf; Tn = W[6]; To = W[7]; Tp = FMA(T8, Tn, Td * To); Tr = FNMS(Td, Tn, T8 * To); T28 = FNMS(T1S, Tn, T1Q * To); T2x = FNMS(TV, Tn, TR * To); TY = FMA(T3, Tn, T6 * To); T2k = FMA(T2, Tn, T5 * To); T2m = FNMS(T5, Tn, T2 * To); T2v = FMA(TR, Tn, TV * To); TG = FNMS(Te, Tn, T9 * To); TE = FMA(T9, Tn, Te * To); T10 = FNMS(T6, Tn, T3 * To); T1h = FMA(T1f, Tn, T1g * To); T1E = FMA(Tg, Tn, Tk * To); T26 = FMA(T1Q, Tn, T1S * To); T1B = FNMS(TB, Tn, Tx * To); T1G = FNMS(Tk, Tn, Tg * To); T1V = FMA(T14, Tn, T16 * To); T1X = FNMS(T16, Tn, T14 * To); T1z = FMA(Tx, Tn, TB * To); T1j = FNMS(T1g, Tn, T1f * To); } } { E T1, T6v, T2F, T6I, TK, T2G, T6u, T6J, T6N, T7c, T2O, T52, T2C, T6k, T48; E T5X, T4L, T5s, T4j, T5W, T4K, T5v, T1o, T6g, T30, T5M, T4A, T56, T3b, T5N; E T4B, T59, T1L, T6h, T3n, T5Q, T4D, T5g, T3y, T5P, T4E, T5d, T2d, T6j, T3L; E T5T, T4I, T5l, T3W, T5U, T4H, T5o; { E Tm, T2I, Tt, T2J, Tu, T6s, TD, T2L, TI, T2M, TJ, T6t; T1 = ri[0]; T6v = ii[0]; { E Th, Tl, Tq, Ts; Th = ri[WS(rs, 5)]; Tl = ii[WS(rs, 5)]; Tm = FMA(Tg, Th, Tk * Tl); T2I = FNMS(Tk, Th, Tg * Tl); Tq = ri[WS(rs, 20)]; Ts = ii[WS(rs, 20)]; Tt = FMA(Tp, Tq, Tr * Ts); T2J = FNMS(Tr, Tq, Tp * Ts); } Tu = Tm + Tt; T6s = T2I + T2J; { E Ty, TC, TF, TH; Ty = ri[WS(rs, 10)]; TC = ii[WS(rs, 10)]; TD = FMA(Tx, Ty, TB * TC); T2L = FNMS(TB, Ty, Tx * TC); TF = ri[WS(rs, 15)]; TH = ii[WS(rs, 15)]; TI = FMA(TE, TF, TG * TH); T2M = FNMS(TG, TF, TE * TH); } TJ = TD + TI; T6t = T2L + T2M; T2F = KP559016994 * (Tu - TJ); T6I = KP559016994 * (T6s - T6t); TK = Tu + TJ; T2G = FNMS(KP250000000, TK, T1); T6u = T6s + T6t; T6J = FNMS(KP250000000, T6u, T6v); { E T6L, T6M, T2K, T2N; T6L = Tm - Tt; T6M = TD - TI; T6N = FMA(KP951056516, T6L, KP587785252 * T6M); T7c = FNMS(KP587785252, T6L, KP951056516 * T6M); T2K = T2I - T2J; T2N = T2L - T2M; T2O = FMA(KP951056516, T2K, KP587785252 * T2N); T52 = FNMS(KP587785252, T2K, KP951056516 * T2N); } } { E T2g, T4c, T43, T46, T4h, T4g, T49, T4a, T4d, T2p, T2A, T2B, T2e, T2f; T2e = ri[WS(rs, 3)]; T2f = ii[WS(rs, 3)]; T2g = FMA(T3, T2e, T6 * T2f); T4c = FNMS(T6, T2e, T3 * T2f); { E T2j, T41, T2z, T45, T2o, T42, T2u, T44; { E T2h, T2i, T2w, T2y; T2h = ri[WS(rs, 8)]; T2i = ii[WS(rs, 8)]; T2j = FMA(T1f, T2h, T1g * T2i); T41 = FNMS(T1g, T2h, T1f * T2i); T2w = ri[WS(rs, 18)]; T2y = ii[WS(rs, 18)]; T2z = FMA(T2v, T2w, T2x * T2y); T45 = FNMS(T2x, T2w, T2v * T2y); } { E T2l, T2n, T2r, T2t; T2l = ri[WS(rs, 23)]; T2n = ii[WS(rs, 23)]; T2o = FMA(T2k, T2l, T2m * T2n); T42 = FNMS(T2m, T2l, T2k * T2n); T2r = ri[WS(rs, 13)]; T2t = ii[WS(rs, 13)]; T2u = FMA(T2q, T2r, T2s * T2t); T44 = FNMS(T2s, T2r, T2q * T2t); } T43 = T41 - T42; T46 = T44 - T45; T4h = T2u - T2z; T4g = T2j - T2o; T49 = T41 + T42; T4a = T44 + T45; T4d = T49 + T4a; T2p = T2j + T2o; T2A = T2u + T2z; T2B = T2p + T2A; } T2C = T2g + T2B; T6k = T4c + T4d; { E T47, T5r, T40, T5q, T3Y, T3Z; T47 = FMA(KP951056516, T43, KP587785252 * T46); T5r = FNMS(KP587785252, T43, KP951056516 * T46); T3Y = KP559016994 * (T2p - T2A); T3Z = FNMS(KP250000000, T2B, T2g); T40 = T3Y + T3Z; T5q = T3Z - T3Y; T48 = T40 + T47; T5X = T5q + T5r; T4L = T40 - T47; T5s = T5q - T5r; } { E T4i, T5t, T4f, T5u, T4b, T4e; T4i = FMA(KP951056516, T4g, KP587785252 * T4h); T5t = FNMS(KP587785252, T4g, KP951056516 * T4h); T4b = KP559016994 * (T49 - T4a); T4e = FNMS(KP250000000, T4d, T4c); T4f = T4b + T4e; T5u = T4e - T4b; T4j = T4f - T4i; T5W = T5u - T5t; T4K = T4i + T4f; T5v = T5t + T5u; } } { E TO, T34, T2V, T2Y, T39, T38, T31, T32, T35, T13, T1m, T1n, TM, TN; TM = ri[WS(rs, 1)]; TN = ii[WS(rs, 1)]; TO = FMA(T2, TM, T5 * TN); T34 = FNMS(T5, TM, T2 * TN); { E TX, T2T, T1l, T2X, T12, T2U, T1e, T2W; { E TS, TW, T1i, T1k; TS = ri[WS(rs, 6)]; TW = ii[WS(rs, 6)]; TX = FMA(TR, TS, TV * TW); T2T = FNMS(TV, TS, TR * TW); T1i = ri[WS(rs, 16)]; T1k = ii[WS(rs, 16)]; T1l = FMA(T1h, T1i, T1j * T1k); T2X = FNMS(T1j, T1i, T1h * T1k); } { E TZ, T11, T19, T1d; TZ = ri[WS(rs, 21)]; T11 = ii[WS(rs, 21)]; T12 = FMA(TY, TZ, T10 * T11); T2U = FNMS(T10, TZ, TY * T11); T19 = ri[WS(rs, 11)]; T1d = ii[WS(rs, 11)]; T1e = FMA(T18, T19, T1c * T1d); T2W = FNMS(T1c, T19, T18 * T1d); } T2V = T2T - T2U; T2Y = T2W - T2X; T39 = T1e - T1l; T38 = TX - T12; T31 = T2T + T2U; T32 = T2W + T2X; T35 = T31 + T32; T13 = TX + T12; T1m = T1e + T1l; T1n = T13 + T1m; } T1o = TO + T1n; T6g = T34 + T35; { E T2Z, T55, T2S, T54, T2Q, T2R; T2Z = FMA(KP951056516, T2V, KP587785252 * T2Y); T55 = FNMS(KP587785252, T2V, KP951056516 * T2Y); T2Q = KP559016994 * (T13 - T1m); T2R = FNMS(KP250000000, T1n, TO); T2S = T2Q + T2R; T54 = T2R - T2Q; T30 = T2S + T2Z; T5M = T54 + T55; T4A = T2S - T2Z; T56 = T54 - T55; } { E T3a, T57, T37, T58, T33, T36; T3a = FMA(KP951056516, T38, KP587785252 * T39); T57 = FNMS(KP587785252, T38, KP951056516 * T39); T33 = KP559016994 * (T31 - T32); T36 = FNMS(KP250000000, T35, T34); T37 = T33 + T36; T58 = T36 - T33; T3b = T37 - T3a; T5N = T58 - T57; T4B = T3a + T37; T59 = T57 + T58; } } { E T1r, T3r, T3i, T3l, T3w, T3v, T3o, T3p, T3s, T1y, T1J, T1K, T1p, T1q; T1p = ri[WS(rs, 4)]; T1q = ii[WS(rs, 4)]; T1r = FMA(T8, T1p, Td * T1q); T3r = FNMS(Td, T1p, T8 * T1q); { E T1u, T3g, T1I, T3k, T1x, T3h, T1D, T3j; { E T1s, T1t, T1F, T1H; T1s = ri[WS(rs, 9)]; T1t = ii[WS(rs, 9)]; T1u = FMA(T9, T1s, Te * T1t); T3g = FNMS(Te, T1s, T9 * T1t); T1F = ri[WS(rs, 19)]; T1H = ii[WS(rs, 19)]; T1I = FMA(T1E, T1F, T1G * T1H); T3k = FNMS(T1G, T1F, T1E * T1H); } { E T1v, T1w, T1A, T1C; T1v = ri[WS(rs, 24)]; T1w = ii[WS(rs, 24)]; T1x = FMA(Tn, T1v, To * T1w); T3h = FNMS(To, T1v, Tn * T1w); T1A = ri[WS(rs, 14)]; T1C = ii[WS(rs, 14)]; T1D = FMA(T1z, T1A, T1B * T1C); T3j = FNMS(T1B, T1A, T1z * T1C); } T3i = T3g - T3h; T3l = T3j - T3k; T3w = T1D - T1I; T3v = T1u - T1x; T3o = T3g + T3h; T3p = T3j + T3k; T3s = T3o + T3p; T1y = T1u + T1x; T1J = T1D + T1I; T1K = T1y + T1J; } T1L = T1r + T1K; T6h = T3r + T3s; { E T3m, T5f, T3f, T5e, T3d, T3e; T3m = FMA(KP951056516, T3i, KP587785252 * T3l); T5f = FNMS(KP587785252, T3i, KP951056516 * T3l); T3d = KP559016994 * (T1y - T1J); T3e = FNMS(KP250000000, T1K, T1r); T3f = T3d + T3e; T5e = T3e - T3d; T3n = T3f + T3m; T5Q = T5e + T5f; T4D = T3f - T3m; T5g = T5e - T5f; } { E T3x, T5b, T3u, T5c, T3q, T3t; T3x = FMA(KP951056516, T3v, KP587785252 * T3w); T5b = FNMS(KP587785252, T3v, KP951056516 * T3w); T3q = KP559016994 * (T3o - T3p); T3t = FNMS(KP250000000, T3s, T3r); T3u = T3q + T3t; T5c = T3t - T3q; T3y = T3u - T3x; T5P = T5c - T5b; T4E = T3x + T3u; T5d = T5b + T5c; } } { E T1P, T3P, T3G, T3J, T3U, T3T, T3M, T3N, T3Q, T20, T2b, T2c, T1N, T1O; T1N = ri[WS(rs, 2)]; T1O = ii[WS(rs, 2)]; T1P = FMA(T14, T1N, T16 * T1O); T3P = FNMS(T16, T1N, T14 * T1O); { E T1U, T3E, T2a, T3I, T1Z, T3F, T25, T3H; { E T1R, T1T, T27, T29; T1R = ri[WS(rs, 7)]; T1T = ii[WS(rs, 7)]; T1U = FMA(T1Q, T1R, T1S * T1T); T3E = FNMS(T1S, T1R, T1Q * T1T); T27 = ri[WS(rs, 17)]; T29 = ii[WS(rs, 17)]; T2a = FMA(T26, T27, T28 * T29); T3I = FNMS(T28, T27, T26 * T29); } { E T1W, T1Y, T22, T24; T1W = ri[WS(rs, 22)]; T1Y = ii[WS(rs, 22)]; T1Z = FMA(T1V, T1W, T1X * T1Y); T3F = FNMS(T1X, T1W, T1V * T1Y); T22 = ri[WS(rs, 12)]; T24 = ii[WS(rs, 12)]; T25 = FMA(T21, T22, T23 * T24); T3H = FNMS(T23, T22, T21 * T24); } T3G = T3E - T3F; T3J = T3H - T3I; T3U = T25 - T2a; T3T = T1U - T1Z; T3M = T3E + T3F; T3N = T3H + T3I; T3Q = T3M + T3N; T20 = T1U + T1Z; T2b = T25 + T2a; T2c = T20 + T2b; } T2d = T1P + T2c; T6j = T3P + T3Q; { E T3K, T5k, T3D, T5j, T3B, T3C; T3K = FMA(KP951056516, T3G, KP587785252 * T3J); T5k = FNMS(KP587785252, T3G, KP951056516 * T3J); T3B = KP559016994 * (T20 - T2b); T3C = FNMS(KP250000000, T2c, T1P); T3D = T3B + T3C; T5j = T3C - T3B; T3L = T3D + T3K; T5T = T5j + T5k; T4I = T3D - T3K; T5l = T5j - T5k; } { E T3V, T5m, T3S, T5n, T3O, T3R; T3V = FMA(KP951056516, T3T, KP587785252 * T3U); T5m = FNMS(KP587785252, T3T, KP951056516 * T3U); T3O = KP559016994 * (T3M - T3N); T3R = FNMS(KP250000000, T3Q, T3P); T3S = T3O + T3R; T5n = T3R - T3O; T3W = T3S - T3V; T5U = T5n - T5m; T4H = T3V + T3S; T5o = T5m + T5n; } } { E T6m, T6o, TL, T2E, T6d, T6e, T6n, T6f; { E T6i, T6l, T1M, T2D; T6i = T6g - T6h; T6l = T6j - T6k; T6m = FMA(KP951056516, T6i, KP587785252 * T6l); T6o = FNMS(KP587785252, T6i, KP951056516 * T6l); TL = T1 + TK; T1M = T1o + T1L; T2D = T2d + T2C; T2E = T1M + T2D; T6d = KP559016994 * (T1M - T2D); T6e = FNMS(KP250000000, T2E, TL); } ri[0] = TL + T2E; T6n = T6e - T6d; ri[WS(rs, 10)] = T6n - T6o; ri[WS(rs, 15)] = T6n + T6o; T6f = T6d + T6e; ri[WS(rs, 20)] = T6f - T6m; ri[WS(rs, 5)] = T6f + T6m; } { E T6C, T6D, T6w, T6r, T6x, T6y, T6E, T6z; { E T6A, T6B, T6p, T6q; T6A = T1o - T1L; T6B = T2d - T2C; T6C = FMA(KP951056516, T6A, KP587785252 * T6B); T6D = FNMS(KP587785252, T6A, KP951056516 * T6B); T6w = T6u + T6v; T6p = T6g + T6h; T6q = T6j + T6k; T6r = T6p + T6q; T6x = KP559016994 * (T6p - T6q); T6y = FNMS(KP250000000, T6r, T6w); } ii[0] = T6r + T6w; T6E = T6y - T6x; ii[WS(rs, 10)] = T6D + T6E; ii[WS(rs, 15)] = T6E - T6D; T6z = T6x + T6y; ii[WS(rs, 5)] = T6z - T6C; ii[WS(rs, 20)] = T6C + T6z; } { E T2P, T4z, T6O, T70, T4m, T6T, T4n, T6S, T4U, T71, T4X, T6Z, T4O, T75, T4P; E T74, T4s, T6P, T4v, T6H, T2H, T6K; T2H = T2F + T2G; T2P = T2H + T2O; T4z = T2H - T2O; T6K = T6I + T6J; T6O = T6K - T6N; T70 = T6N + T6K; { E T3c, T3z, T3A, T3X, T4k, T4l; T3c = FMA(KP968583161, T30, KP248689887 * T3b); T3z = FMA(KP535826794, T3n, KP844327925 * T3y); T3A = T3c + T3z; T3X = FMA(KP876306680, T3L, KP481753674 * T3W); T4k = FMA(KP728968627, T48, KP684547105 * T4j); T4l = T3X + T4k; T4m = T3A + T4l; T6T = T3X - T4k; T4n = KP559016994 * (T3A - T4l); T6S = T3c - T3z; } { E T4S, T4T, T6X, T4V, T4W, T6Y; T4S = FNMS(KP844327925, T4A, KP535826794 * T4B); T4T = FNMS(KP637423989, T4E, KP770513242 * T4D); T6X = T4S + T4T; T4V = FMA(KP125333233, T4L, KP992114701 * T4K); T4W = FMA(KP904827052, T4I, KP425779291 * T4H); T6Y = T4W + T4V; T4U = T4S - T4T; T71 = KP559016994 * (T6X + T6Y); T4X = T4V - T4W; T6Z = T6X - T6Y; } { E T4C, T4F, T4G, T4J, T4M, T4N; T4C = FMA(KP535826794, T4A, KP844327925 * T4B); T4F = FMA(KP637423989, T4D, KP770513242 * T4E); T4G = T4C - T4F; T4J = FNMS(KP425779291, T4I, KP904827052 * T4H); T4M = FNMS(KP992114701, T4L, KP125333233 * T4K); T4N = T4J + T4M; T4O = T4G + T4N; T75 = T4J - T4M; T4P = KP559016994 * (T4G - T4N); T74 = T4C + T4F; } { E T4q, T4r, T6F, T4t, T4u, T6G; T4q = FNMS(KP248689887, T30, KP968583161 * T3b); T4r = FNMS(KP844327925, T3n, KP535826794 * T3y); T6F = T4q + T4r; T4t = FNMS(KP481753674, T3L, KP876306680 * T3W); T4u = FNMS(KP684547105, T48, KP728968627 * T4j); T6G = T4t + T4u; T4s = T4q - T4r; T6P = KP559016994 * (T6F - T6G); T4v = T4t - T4u; T6H = T6F + T6G; } ri[WS(rs, 1)] = T2P + T4m; ii[WS(rs, 1)] = T6H + T6O; ri[WS(rs, 4)] = T4z + T4O; ii[WS(rs, 4)] = T6Z + T70; { E T4w, T4y, T4p, T4x, T4o; T4w = FMA(KP951056516, T4s, KP587785252 * T4v); T4y = FNMS(KP587785252, T4s, KP951056516 * T4v); T4o = FNMS(KP250000000, T4m, T2P); T4p = T4n + T4o; T4x = T4o - T4n; ri[WS(rs, 21)] = T4p - T4w; ri[WS(rs, 16)] = T4x + T4y; ri[WS(rs, 6)] = T4p + T4w; ri[WS(rs, 11)] = T4x - T4y; } { E T6U, T6V, T6R, T6W, T6Q; T6U = FMA(KP951056516, T6S, KP587785252 * T6T); T6V = FNMS(KP587785252, T6S, KP951056516 * T6T); T6Q = FNMS(KP250000000, T6H, T6O); T6R = T6P + T6Q; T6W = T6Q - T6P; ii[WS(rs, 6)] = T6R - T6U; ii[WS(rs, 16)] = T6W - T6V; ii[WS(rs, 21)] = T6U + T6R; ii[WS(rs, 11)] = T6V + T6W; } { E T4Y, T50, T4R, T4Z, T4Q; T4Y = FMA(KP951056516, T4U, KP587785252 * T4X); T50 = FNMS(KP587785252, T4U, KP951056516 * T4X); T4Q = FNMS(KP250000000, T4O, T4z); T4R = T4P + T4Q; T4Z = T4Q - T4P; ri[WS(rs, 24)] = T4R - T4Y; ri[WS(rs, 19)] = T4Z + T50; ri[WS(rs, 9)] = T4R + T4Y; ri[WS(rs, 14)] = T4Z - T50; } { E T76, T77, T73, T78, T72; T76 = FMA(KP951056516, T74, KP587785252 * T75); T77 = FNMS(KP587785252, T74, KP951056516 * T75); T72 = FNMS(KP250000000, T6Z, T70); T73 = T71 + T72; T78 = T72 - T71; ii[WS(rs, 9)] = T73 - T76; ii[WS(rs, 19)] = T78 - T77; ii[WS(rs, 24)] = T76 + T73; ii[WS(rs, 14)] = T77 + T78; } } { E T53, T5L, T7e, T7q, T5y, T7j, T5z, T7i, T66, T7r, T69, T7p, T60, T7v, T61; E T7u, T5E, T7f, T5H, T7b, T51, T7d; T51 = T2G - T2F; T53 = T51 - T52; T5L = T51 + T52; T7d = T6J - T6I; T7e = T7c + T7d; T7q = T7d - T7c; { E T5a, T5h, T5i, T5p, T5w, T5x; T5a = FMA(KP876306680, T56, KP481753674 * T59); T5h = FNMS(KP425779291, T5g, KP904827052 * T5d); T5i = T5a + T5h; T5p = FMA(KP535826794, T5l, KP844327925 * T5o); T5w = FMA(KP062790519, T5s, KP998026728 * T5v); T5x = T5p + T5w; T5y = T5i + T5x; T7j = T5p - T5w; T5z = KP559016994 * (T5i - T5x); T7i = T5a - T5h; } { E T64, T65, T7n, T67, T68, T7o; T64 = FNMS(KP684547105, T5M, KP728968627 * T5N); T65 = FMA(KP125333233, T5Q, KP992114701 * T5P); T7n = T64 - T65; T67 = FNMS(KP998026728, T5T, KP062790519 * T5U); T68 = FMA(KP770513242, T5X, KP637423989 * T5W); T7o = T67 - T68; T66 = T64 + T65; T7r = KP559016994 * (T7n - T7o); T69 = T67 + T68; T7p = T7n + T7o; } { E T5O, T5R, T5S, T5V, T5Y, T5Z; T5O = FMA(KP728968627, T5M, KP684547105 * T5N); T5R = FNMS(KP992114701, T5Q, KP125333233 * T5P); T5S = T5O + T5R; T5V = FMA(KP062790519, T5T, KP998026728 * T5U); T5Y = FNMS(KP637423989, T5X, KP770513242 * T5W); T5Z = T5V + T5Y; T60 = T5S + T5Z; T7v = T5V - T5Y; T61 = KP559016994 * (T5S - T5Z); T7u = T5O - T5R; } { E T5C, T5D, T79, T5F, T5G, T7a; T5C = FNMS(KP481753674, T56, KP876306680 * T59); T5D = FMA(KP904827052, T5g, KP425779291 * T5d); T79 = T5C - T5D; T5F = FNMS(KP844327925, T5l, KP535826794 * T5o); T5G = FNMS(KP998026728, T5s, KP062790519 * T5v); T7a = T5F + T5G; T5E = T5C + T5D; T7f = KP559016994 * (T79 - T7a); T5H = T5F - T5G; T7b = T79 + T7a; } ri[WS(rs, 2)] = T53 + T5y; ii[WS(rs, 2)] = T7b + T7e; ri[WS(rs, 3)] = T5L + T60; ii[WS(rs, 3)] = T7p + T7q; { E T5I, T5K, T5B, T5J, T5A; T5I = FMA(KP951056516, T5E, KP587785252 * T5H); T5K = FNMS(KP587785252, T5E, KP951056516 * T5H); T5A = FNMS(KP250000000, T5y, T53); T5B = T5z + T5A; T5J = T5A - T5z; ri[WS(rs, 22)] = T5B - T5I; ri[WS(rs, 17)] = T5J + T5K; ri[WS(rs, 7)] = T5B + T5I; ri[WS(rs, 12)] = T5J - T5K; } { E T7k, T7l, T7h, T7m, T7g; T7k = FMA(KP951056516, T7i, KP587785252 * T7j); T7l = FNMS(KP587785252, T7i, KP951056516 * T7j); T7g = FNMS(KP250000000, T7b, T7e); T7h = T7f + T7g; T7m = T7g - T7f; ii[WS(rs, 7)] = T7h - T7k; ii[WS(rs, 17)] = T7m - T7l; ii[WS(rs, 22)] = T7k + T7h; ii[WS(rs, 12)] = T7l + T7m; } { E T6a, T6c, T63, T6b, T62; T6a = FMA(KP951056516, T66, KP587785252 * T69); T6c = FNMS(KP587785252, T66, KP951056516 * T69); T62 = FNMS(KP250000000, T60, T5L); T63 = T61 + T62; T6b = T62 - T61; ri[WS(rs, 23)] = T63 - T6a; ri[WS(rs, 18)] = T6b + T6c; ri[WS(rs, 8)] = T63 + T6a; ri[WS(rs, 13)] = T6b - T6c; } { E T7w, T7x, T7t, T7y, T7s; T7w = FMA(KP951056516, T7u, KP587785252 * T7v); T7x = FNMS(KP587785252, T7u, KP951056516 * T7v); T7s = FNMS(KP250000000, T7p, T7q); T7t = T7r + T7s; T7y = T7s - T7r; ii[WS(rs, 8)] = T7t - T7w; ii[WS(rs, 18)] = T7y - T7x; ii[WS(rs, 23)] = T7w + T7t; ii[WS(rs, 13)] = T7x + T7y; } } } } } } static const tw_instr twinstr[] = { {TW_CEXP, 0, 1}, {TW_CEXP, 0, 3}, {TW_CEXP, 0, 9}, {TW_CEXP, 0, 24}, {TW_NEXT, 1, 0} }; static const ct_desc desc = { 25, "t2_25", twinstr, &GENUS, {280, 180, 160, 0}, 0, 0, 0 }; void X(codelet_t2_25) (planner *p) { X(kdft_dit_register) (p, t2_25, &desc); } #endif /* HAVE_FMA */ fftw-3.3.3/dft/scalar/codelets/q1_2.c0000644000175000017500000001072612054410301014135 00000000000000/* * Copyright (c) 2003, 2007-11 Matteo Frigo * Copyright (c) 2003, 2007-11 Massachusetts Institute of Technology * * This program is free software; you can redistribute it and/or modify * it under the terms of the GNU General Public License as published by * the Free Software Foundation; either version 2 of the License, or * (at your option) any later version. * * This program is distributed in the hope that it will be useful, * but WITHOUT ANY WARRANTY; without even the implied warranty of * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the * GNU General Public License for more details. * * You should have received a copy of the GNU General Public License * along with this program; if not, write to the Free Software * Foundation, Inc., 51 Franklin Street, Fifth Floor, Boston, MA 02110-1301 USA * */ /* This file was automatically generated --- DO NOT EDIT */ /* Generated on Sun Nov 25 07:36:17 EST 2012 */ #include "codelet-dft.h" #ifdef HAVE_FMA /* Generated by: ../../../genfft/gen_twidsq.native -fma -reorder-insns -schedule-for-pipeline -compact -variables 4 -pipeline-latency 4 -reload-twiddle -dif -n 2 -name q1_2 -include q.h */ /* * This function contains 12 FP additions, 8 FP multiplications, * (or, 8 additions, 4 multiplications, 4 fused multiply/add), * 21 stack variables, 0 constants, and 16 memory accesses */ #include "q.h" static void q1_2(R *rio, R *iio, const R *W, stride rs, stride vs, INT mb, INT me, INT ms) { { INT m; for (m = mb, W = W + (mb * 2); m < me; m = m + 1, rio = rio + ms, iio = iio + ms, W = W + 2, MAKE_VOLATILE_STRIDE(4, rs), MAKE_VOLATILE_STRIDE(0, vs)) { E T9, T6, T5; { E T1, T2, T7, T8, Tb, T4, Tc, Th, Ti, Te, Tj, Td, Tg; T1 = rio[0]; T2 = rio[WS(rs, 1)]; T7 = iio[0]; T8 = iio[WS(rs, 1)]; Tb = rio[WS(vs, 1)]; T4 = T1 - T2; Tc = rio[WS(vs, 1) + WS(rs, 1)]; T9 = T7 - T8; Th = iio[WS(vs, 1)]; Ti = iio[WS(vs, 1) + WS(rs, 1)]; Te = Tb - Tc; rio[0] = T1 + T2; iio[0] = T7 + T8; Tj = Th - Ti; rio[WS(rs, 1)] = Tb + Tc; iio[WS(rs, 1)] = Th + Ti; Td = W[0]; Tg = W[1]; { E T3, Tk, Tf, Ta; T3 = W[0]; T6 = W[1]; Tk = Td * Tj; Tf = Td * Te; Ta = T3 * T9; T5 = T3 * T4; iio[WS(vs, 1) + WS(rs, 1)] = FNMS(Tg, Te, Tk); rio[WS(vs, 1) + WS(rs, 1)] = FMA(Tg, Tj, Tf); iio[WS(vs, 1)] = FNMS(T6, T4, Ta); } } rio[WS(vs, 1)] = FMA(T6, T9, T5); } } } static const tw_instr twinstr[] = { {TW_FULL, 0, 2}, {TW_NEXT, 1, 0} }; static const ct_desc desc = { 2, "q1_2", twinstr, &GENUS, {8, 4, 4, 0}, 0, 0, 0 }; void X(codelet_q1_2) (planner *p) { X(kdft_difsq_register) (p, q1_2, &desc); } #else /* HAVE_FMA */ /* Generated by: ../../../genfft/gen_twidsq.native -compact -variables 4 -pipeline-latency 4 -reload-twiddle -dif -n 2 -name q1_2 -include q.h */ /* * This function contains 12 FP additions, 8 FP multiplications, * (or, 8 additions, 4 multiplications, 4 fused multiply/add), * 17 stack variables, 0 constants, and 16 memory accesses */ #include "q.h" static void q1_2(R *rio, R *iio, const R *W, stride rs, stride vs, INT mb, INT me, INT ms) { { INT m; for (m = mb, W = W + (mb * 2); m < me; m = m + 1, rio = rio + ms, iio = iio + ms, W = W + 2, MAKE_VOLATILE_STRIDE(4, rs), MAKE_VOLATILE_STRIDE(0, vs)) { E T1, T2, T4, T6, T7, T8, T9, Ta, Tc, Te, Tf, Tg; T1 = rio[0]; T2 = rio[WS(rs, 1)]; T4 = T1 - T2; T6 = iio[0]; T7 = iio[WS(rs, 1)]; T8 = T6 - T7; T9 = rio[WS(vs, 1)]; Ta = rio[WS(vs, 1) + WS(rs, 1)]; Tc = T9 - Ta; Te = iio[WS(vs, 1)]; Tf = iio[WS(vs, 1) + WS(rs, 1)]; Tg = Te - Tf; rio[0] = T1 + T2; iio[0] = T6 + T7; rio[WS(rs, 1)] = T9 + Ta; iio[WS(rs, 1)] = Te + Tf; { E Tb, Td, T3, T5; Tb = W[0]; Td = W[1]; rio[WS(vs, 1) + WS(rs, 1)] = FMA(Tb, Tc, Td * Tg); iio[WS(vs, 1) + WS(rs, 1)] = FNMS(Td, Tc, Tb * Tg); T3 = W[0]; T5 = W[1]; rio[WS(vs, 1)] = FMA(T3, T4, T5 * T8); iio[WS(vs, 1)] = FNMS(T5, T4, T3 * T8); } } } } static const tw_instr twinstr[] = { {TW_FULL, 0, 2}, {TW_NEXT, 1, 0} }; static const ct_desc desc = { 2, "q1_2", twinstr, &GENUS, {8, 4, 4, 0}, 0, 0, 0 }; void X(codelet_q1_2) (planner *p) { X(kdft_difsq_register) (p, q1_2, &desc); } #endif /* HAVE_FMA */ fftw-3.3.3/dft/scalar/codelets/q1_4.c0000644000175000017500000003363512054410301014143 00000000000000/* * Copyright (c) 2003, 2007-11 Matteo Frigo * Copyright (c) 2003, 2007-11 Massachusetts Institute of Technology * * This program is free software; you can redistribute it and/or modify * it under the terms of the GNU General Public License as published by * the Free Software Foundation; either version 2 of the License, or * (at your option) any later version. * * This program is distributed in the hope that it will be useful, * but WITHOUT ANY WARRANTY; without even the implied warranty of * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the * GNU General Public License for more details. * * You should have received a copy of the GNU General Public License * along with this program; if not, write to the Free Software * Foundation, Inc., 51 Franklin Street, Fifth Floor, Boston, MA 02110-1301 USA * */ /* This file was automatically generated --- DO NOT EDIT */ /* Generated on Sun Nov 25 07:36:17 EST 2012 */ #include "codelet-dft.h" #ifdef HAVE_FMA /* Generated by: ../../../genfft/gen_twidsq.native -fma -reorder-insns -schedule-for-pipeline -compact -variables 4 -pipeline-latency 4 -reload-twiddle -dif -n 4 -name q1_4 -include q.h */ /* * This function contains 88 FP additions, 48 FP multiplications, * (or, 64 additions, 24 multiplications, 24 fused multiply/add), * 76 stack variables, 0 constants, and 64 memory accesses */ #include "q.h" static void q1_4(R *rio, R *iio, const R *W, stride rs, stride vs, INT mb, INT me, INT ms) { { INT m; for (m = mb, W = W + (mb * 6); m < me; m = m + 1, rio = rio + ms, iio = iio + ms, W = W + 6, MAKE_VOLATILE_STRIDE(8, rs), MAKE_VOLATILE_STRIDE(0, vs)) { E T1X, T1S, T1L, T1Y, T1R; { E T3, Tf, Tv, Ti, Tw, Tx, T6, Tm, Tc, Ts, T1T, T1H, T29, T1W, T2a; E T2b, T1K, T20, T1Q, T26, TN, TB, T13, TQ, T14, T15, TE, TU, TK, T10; E T1l, T19, T1a, T1h, T1B, T1o, T1C, T1b, T1D, T1e, T1c; { E T1I, T1P, T1J, T1M; { E Tb, T4, T5, T8; { E T1, T2, T9, Ta, Tg, Th; T1 = rio[0]; T2 = rio[WS(rs, 2)]; T9 = iio[0]; Ta = iio[WS(rs, 2)]; Tg = iio[WS(rs, 1)]; T3 = T1 + T2; Tf = T1 - T2; Th = iio[WS(rs, 3)]; Tv = T9 + Ta; Tb = T9 - Ta; T4 = rio[WS(rs, 1)]; Ti = Tg - Th; Tw = Tg + Th; T5 = rio[WS(rs, 3)]; } Tx = Tv - Tw; T8 = T4 - T5; T6 = T4 + T5; { E T1N, T1O, T1F, T1G, T1U, T1V; T1F = rio[WS(vs, 3)]; T1G = rio[WS(vs, 3) + WS(rs, 2)]; Tm = Tb - T8; Tc = T8 + Tb; Ts = T3 - T6; T1T = T1F - T1G; T1H = T1F + T1G; T1N = iio[WS(vs, 3)]; T1O = iio[WS(vs, 3) + WS(rs, 2)]; T1U = iio[WS(vs, 3) + WS(rs, 1)]; T1V = iio[WS(vs, 3) + WS(rs, 3)]; T1I = rio[WS(vs, 3) + WS(rs, 1)]; T1P = T1N - T1O; T29 = T1N + T1O; T1W = T1U - T1V; T2a = T1U + T1V; T1J = rio[WS(vs, 3) + WS(rs, 3)]; } } T2b = T29 - T2a; T1M = T1I - T1J; T1K = T1I + T1J; { E TC, TJ, TD, TG; { E TH, TI, Tz, TA, TO, TP; Tz = rio[WS(vs, 1)]; TA = rio[WS(vs, 1) + WS(rs, 2)]; T20 = T1P - T1M; T1Q = T1M + T1P; T26 = T1H - T1K; TN = Tz - TA; TB = Tz + TA; TH = iio[WS(vs, 1)]; TI = iio[WS(vs, 1) + WS(rs, 2)]; TO = iio[WS(vs, 1) + WS(rs, 1)]; TP = iio[WS(vs, 1) + WS(rs, 3)]; TC = rio[WS(vs, 1) + WS(rs, 1)]; TJ = TH - TI; T13 = TH + TI; TQ = TO - TP; T14 = TO + TP; TD = rio[WS(vs, 1) + WS(rs, 3)]; } T15 = T13 - T14; TG = TC - TD; TE = TC + TD; { E T1f, T1g, T17, T18, T1m, T1n; T17 = rio[WS(vs, 2)]; T18 = rio[WS(vs, 2) + WS(rs, 2)]; TU = TJ - TG; TK = TG + TJ; T10 = TB - TE; T1l = T17 - T18; T19 = T17 + T18; T1f = iio[WS(vs, 2)]; T1g = iio[WS(vs, 2) + WS(rs, 2)]; T1m = iio[WS(vs, 2) + WS(rs, 1)]; T1n = iio[WS(vs, 2) + WS(rs, 3)]; T1a = rio[WS(vs, 2) + WS(rs, 1)]; T1h = T1f - T1g; T1B = T1f + T1g; T1o = T1m - T1n; T1C = T1m + T1n; T1b = rio[WS(vs, 2) + WS(rs, 3)]; } } } T1D = T1B - T1C; T1e = T1a - T1b; T1c = T1a + T1b; { E T1s, T1i, T1y, T28, T27, Tr, Tu; rio[0] = T3 + T6; iio[0] = Tv + Tw; T1s = T1h - T1e; T1i = T1e + T1h; T1y = T19 - T1c; rio[WS(rs, 1)] = TB + TE; iio[WS(rs, 1)] = T13 + T14; rio[WS(rs, 2)] = T19 + T1c; iio[WS(rs, 2)] = T1B + T1C; iio[WS(rs, 3)] = T29 + T2a; rio[WS(rs, 3)] = T1H + T1K; Tr = W[2]; Tu = W[3]; { E T25, Ty, Tt, T2c; T25 = W[2]; T28 = W[3]; Ty = Tr * Tx; Tt = Tr * Ts; T2c = T25 * T2b; T27 = T25 * T26; iio[WS(vs, 2)] = FNMS(Tu, Ts, Ty); rio[WS(vs, 2)] = FMA(Tu, Tx, Tt); iio[WS(vs, 2) + WS(rs, 3)] = FNMS(T28, T26, T2c); } rio[WS(vs, 2) + WS(rs, 3)] = FMA(T28, T2b, T27); { E Tp, T1v, T23, T22, T1Z, TR, TM, TF; { E T1A, T1z, TZ, T12; TZ = W[2]; T12 = W[3]; { E T1x, T16, T11, T1E; T1x = W[2]; T1A = W[3]; T16 = TZ * T15; T11 = TZ * T10; T1E = T1x * T1D; T1z = T1x * T1y; iio[WS(vs, 2) + WS(rs, 1)] = FNMS(T12, T10, T16); rio[WS(vs, 2) + WS(rs, 1)] = FMA(T12, T15, T11); iio[WS(vs, 2) + WS(rs, 2)] = FNMS(T1A, T1y, T1E); } rio[WS(vs, 2) + WS(rs, 2)] = FMA(T1A, T1D, T1z); { E Tj, Te, T7, T1p, T1k, T1j; Tp = Tf + Ti; Tj = Tf - Ti; Te = W[5]; T7 = W[4]; { E T1d, T1q, Tk, Td; T1p = T1l - T1o; T1v = T1l + T1o; T1k = W[5]; Tk = Te * Tc; Td = T7 * Tc; T1d = W[4]; T1q = T1k * T1i; rio[WS(vs, 3)] = FMA(T7, Tj, Tk); iio[WS(vs, 3)] = FNMS(Te, Tj, Td); T1j = T1d * T1i; rio[WS(vs, 3) + WS(rs, 2)] = FMA(T1d, T1p, T1q); } T23 = T1T + T1W; T1X = T1T - T1W; T22 = W[1]; iio[WS(vs, 3) + WS(rs, 2)] = FNMS(T1k, T1p, T1j); T1Z = W[0]; } } { E TX, TW, TT, TY, TV, T24, T21; TX = TN + TQ; TR = TN - TQ; T24 = T22 * T20; TW = W[1]; T21 = T1Z * T20; TT = W[0]; rio[WS(vs, 1) + WS(rs, 3)] = FMA(T1Z, T23, T24); TY = TW * TU; iio[WS(vs, 1) + WS(rs, 3)] = FNMS(T22, T23, T21); TV = TT * TU; rio[WS(vs, 1) + WS(rs, 1)] = FMA(TT, TX, TY); TM = W[5]; iio[WS(vs, 1) + WS(rs, 1)] = FNMS(TW, TX, TV); TF = W[4]; } { E To, Tl, Tq, Tn, TS, TL; TS = TM * TK; To = W[1]; TL = TF * TK; Tl = W[0]; rio[WS(vs, 3) + WS(rs, 1)] = FMA(TF, TR, TS); Tq = To * Tm; iio[WS(vs, 3) + WS(rs, 1)] = FNMS(TM, TR, TL); Tn = Tl * Tm; { E T1u, T1r, T1w, T1t; rio[WS(vs, 1)] = FMA(Tl, Tp, Tq); T1u = W[1]; iio[WS(vs, 1)] = FNMS(To, Tp, Tn); T1r = W[0]; T1w = T1u * T1s; T1S = W[5]; T1t = T1r * T1s; T1L = W[4]; rio[WS(vs, 1) + WS(rs, 2)] = FMA(T1r, T1v, T1w); T1Y = T1S * T1Q; iio[WS(vs, 1) + WS(rs, 2)] = FNMS(T1u, T1v, T1t); T1R = T1L * T1Q; } } } } } rio[WS(vs, 3) + WS(rs, 3)] = FMA(T1L, T1X, T1Y); iio[WS(vs, 3) + WS(rs, 3)] = FNMS(T1S, T1X, T1R); } } } static const tw_instr twinstr[] = { {TW_FULL, 0, 4}, {TW_NEXT, 1, 0} }; static const ct_desc desc = { 4, "q1_4", twinstr, &GENUS, {64, 24, 24, 0}, 0, 0, 0 }; void X(codelet_q1_4) (planner *p) { X(kdft_difsq_register) (p, q1_4, &desc); } #else /* HAVE_FMA */ /* Generated by: ../../../genfft/gen_twidsq.native -compact -variables 4 -pipeline-latency 4 -reload-twiddle -dif -n 4 -name q1_4 -include q.h */ /* * This function contains 88 FP additions, 48 FP multiplications, * (or, 64 additions, 24 multiplications, 24 fused multiply/add), * 37 stack variables, 0 constants, and 64 memory accesses */ #include "q.h" static void q1_4(R *rio, R *iio, const R *W, stride rs, stride vs, INT mb, INT me, INT ms) { { INT m; for (m = mb, W = W + (mb * 6); m < me; m = m + 1, rio = rio + ms, iio = iio + ms, W = W + 6, MAKE_VOLATILE_STRIDE(8, rs), MAKE_VOLATILE_STRIDE(0, vs)) { E T3, Te, Tb, Tq, T6, T8, Th, Tr, Tv, TG, TD, TS, Ty, TA, TJ; E TT, TX, T18, T15, T1k, T10, T12, T1b, T1l, T1p, T1A, T1x, T1M, T1s, T1u; E T1D, T1N; { E T1, T2, T9, Ta; T1 = rio[0]; T2 = rio[WS(rs, 2)]; T3 = T1 + T2; Te = T1 - T2; T9 = iio[0]; Ta = iio[WS(rs, 2)]; Tb = T9 - Ta; Tq = T9 + Ta; } { E T4, T5, Tf, Tg; T4 = rio[WS(rs, 1)]; T5 = rio[WS(rs, 3)]; T6 = T4 + T5; T8 = T4 - T5; Tf = iio[WS(rs, 1)]; Tg = iio[WS(rs, 3)]; Th = Tf - Tg; Tr = Tf + Tg; } { E Tt, Tu, TB, TC; Tt = rio[WS(vs, 1)]; Tu = rio[WS(vs, 1) + WS(rs, 2)]; Tv = Tt + Tu; TG = Tt - Tu; TB = iio[WS(vs, 1)]; TC = iio[WS(vs, 1) + WS(rs, 2)]; TD = TB - TC; TS = TB + TC; } { E Tw, Tx, TH, TI; Tw = rio[WS(vs, 1) + WS(rs, 1)]; Tx = rio[WS(vs, 1) + WS(rs, 3)]; Ty = Tw + Tx; TA = Tw - Tx; TH = iio[WS(vs, 1) + WS(rs, 1)]; TI = iio[WS(vs, 1) + WS(rs, 3)]; TJ = TH - TI; TT = TH + TI; } { E TV, TW, T13, T14; TV = rio[WS(vs, 2)]; TW = rio[WS(vs, 2) + WS(rs, 2)]; TX = TV + TW; T18 = TV - TW; T13 = iio[WS(vs, 2)]; T14 = iio[WS(vs, 2) + WS(rs, 2)]; T15 = T13 - T14; T1k = T13 + T14; } { E TY, TZ, T19, T1a; TY = rio[WS(vs, 2) + WS(rs, 1)]; TZ = rio[WS(vs, 2) + WS(rs, 3)]; T10 = TY + TZ; T12 = TY - TZ; T19 = iio[WS(vs, 2) + WS(rs, 1)]; T1a = iio[WS(vs, 2) + WS(rs, 3)]; T1b = T19 - T1a; T1l = T19 + T1a; } { E T1n, T1o, T1v, T1w; T1n = rio[WS(vs, 3)]; T1o = rio[WS(vs, 3) + WS(rs, 2)]; T1p = T1n + T1o; T1A = T1n - T1o; T1v = iio[WS(vs, 3)]; T1w = iio[WS(vs, 3) + WS(rs, 2)]; T1x = T1v - T1w; T1M = T1v + T1w; } { E T1q, T1r, T1B, T1C; T1q = rio[WS(vs, 3) + WS(rs, 1)]; T1r = rio[WS(vs, 3) + WS(rs, 3)]; T1s = T1q + T1r; T1u = T1q - T1r; T1B = iio[WS(vs, 3) + WS(rs, 1)]; T1C = iio[WS(vs, 3) + WS(rs, 3)]; T1D = T1B - T1C; T1N = T1B + T1C; } rio[0] = T3 + T6; iio[0] = Tq + Tr; rio[WS(rs, 1)] = Tv + Ty; iio[WS(rs, 1)] = TS + TT; rio[WS(rs, 2)] = TX + T10; iio[WS(rs, 2)] = T1k + T1l; iio[WS(rs, 3)] = T1M + T1N; rio[WS(rs, 3)] = T1p + T1s; { E Tc, Ti, T7, Td; Tc = T8 + Tb; Ti = Te - Th; T7 = W[4]; Td = W[5]; iio[WS(vs, 3)] = FNMS(Td, Ti, T7 * Tc); rio[WS(vs, 3)] = FMA(Td, Tc, T7 * Ti); } { E T1K, T1O, T1J, T1L; T1K = T1p - T1s; T1O = T1M - T1N; T1J = W[2]; T1L = W[3]; rio[WS(vs, 2) + WS(rs, 3)] = FMA(T1J, T1K, T1L * T1O); iio[WS(vs, 2) + WS(rs, 3)] = FNMS(T1L, T1K, T1J * T1O); } { E Tk, Tm, Tj, Tl; Tk = Tb - T8; Tm = Te + Th; Tj = W[0]; Tl = W[1]; iio[WS(vs, 1)] = FNMS(Tl, Tm, Tj * Tk); rio[WS(vs, 1)] = FMA(Tl, Tk, Tj * Tm); } { E To, Ts, Tn, Tp; To = T3 - T6; Ts = Tq - Tr; Tn = W[2]; Tp = W[3]; rio[WS(vs, 2)] = FMA(Tn, To, Tp * Ts); iio[WS(vs, 2)] = FNMS(Tp, To, Tn * Ts); } { E T16, T1c, T11, T17; T16 = T12 + T15; T1c = T18 - T1b; T11 = W[4]; T17 = W[5]; iio[WS(vs, 3) + WS(rs, 2)] = FNMS(T17, T1c, T11 * T16); rio[WS(vs, 3) + WS(rs, 2)] = FMA(T17, T16, T11 * T1c); } { E T1G, T1I, T1F, T1H; T1G = T1x - T1u; T1I = T1A + T1D; T1F = W[0]; T1H = W[1]; iio[WS(vs, 1) + WS(rs, 3)] = FNMS(T1H, T1I, T1F * T1G); rio[WS(vs, 1) + WS(rs, 3)] = FMA(T1H, T1G, T1F * T1I); } { E TQ, TU, TP, TR; TQ = Tv - Ty; TU = TS - TT; TP = W[2]; TR = W[3]; rio[WS(vs, 2) + WS(rs, 1)] = FMA(TP, TQ, TR * TU); iio[WS(vs, 2) + WS(rs, 1)] = FNMS(TR, TQ, TP * TU); } { E T1e, T1g, T1d, T1f; T1e = T15 - T12; T1g = T18 + T1b; T1d = W[0]; T1f = W[1]; iio[WS(vs, 1) + WS(rs, 2)] = FNMS(T1f, T1g, T1d * T1e); rio[WS(vs, 1) + WS(rs, 2)] = FMA(T1f, T1e, T1d * T1g); } { E T1i, T1m, T1h, T1j; T1i = TX - T10; T1m = T1k - T1l; T1h = W[2]; T1j = W[3]; rio[WS(vs, 2) + WS(rs, 2)] = FMA(T1h, T1i, T1j * T1m); iio[WS(vs, 2) + WS(rs, 2)] = FNMS(T1j, T1i, T1h * T1m); } { E T1y, T1E, T1t, T1z; T1y = T1u + T1x; T1E = T1A - T1D; T1t = W[4]; T1z = W[5]; iio[WS(vs, 3) + WS(rs, 3)] = FNMS(T1z, T1E, T1t * T1y); rio[WS(vs, 3) + WS(rs, 3)] = FMA(T1z, T1y, T1t * T1E); } { E TM, TO, TL, TN; TM = TD - TA; TO = TG + TJ; TL = W[0]; TN = W[1]; iio[WS(vs, 1) + WS(rs, 1)] = FNMS(TN, TO, TL * TM); rio[WS(vs, 1) + WS(rs, 1)] = FMA(TN, TM, TL * TO); } { E TE, TK, Tz, TF; TE = TA + TD; TK = TG - TJ; Tz = W[4]; TF = W[5]; iio[WS(vs, 3) + WS(rs, 1)] = FNMS(TF, TK, Tz * TE); rio[WS(vs, 3) + WS(rs, 1)] = FMA(TF, TE, Tz * TK); } } } } static const tw_instr twinstr[] = { {TW_FULL, 0, 4}, {TW_NEXT, 1, 0} }; static const ct_desc desc = { 4, "q1_4", twinstr, &GENUS, {64, 24, 24, 0}, 0, 0, 0 }; void X(codelet_q1_4) (planner *p) { X(kdft_difsq_register) (p, q1_4, &desc); } #endif /* HAVE_FMA */ fftw-3.3.3/dft/scalar/codelets/q1_8.c0000644000175000017500000020474012054410310014144 00000000000000/* * Copyright (c) 2003, 2007-11 Matteo Frigo * Copyright (c) 2003, 2007-11 Massachusetts Institute of Technology * * This program is free software; you can redistribute it and/or modify * it under the terms of the GNU General Public License as published by * the Free Software Foundation; either version 2 of the License, or * (at your option) any later version. * * This program is distributed in the hope that it will be useful, * but WITHOUT ANY WARRANTY; without even the implied warranty of * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the * GNU General Public License for more details. * * You should have received a copy of the GNU General Public License * along with this program; if not, write to the Free Software * Foundation, Inc., 51 Franklin Street, Fifth Floor, Boston, MA 02110-1301 USA * */ /* This file was automatically generated --- DO NOT EDIT */ /* Generated on Sun Nov 25 07:36:17 EST 2012 */ #include "codelet-dft.h" #ifdef HAVE_FMA /* Generated by: ../../../genfft/gen_twidsq.native -fma -reorder-insns -schedule-for-pipeline -compact -variables 4 -pipeline-latency 4 -reload-twiddle -dif -n 8 -name q1_8 -include q.h */ /* * This function contains 528 FP additions, 288 FP multiplications, * (or, 352 additions, 112 multiplications, 176 fused multiply/add), * 190 stack variables, 1 constants, and 256 memory accesses */ #include "q.h" static void q1_8(R *rio, R *iio, const R *W, stride rs, stride vs, INT mb, INT me, INT ms) { DK(KP707106781, +0.707106781186547524400844362104849039284835938); { INT m; for (m = mb, W = W + (mb * 14); m < me; m = m + 1, rio = rio + ms, iio = iio + ms, W = W + 14, MAKE_VOLATILE_STRIDE(16, rs), MAKE_VOLATILE_STRIDE(0, vs)) { E T9C, T9N, T9l, T9E, T9D, T9O; { E TV, Tk, T1d, T7, T18, T1t, TQ, TD, T5t, T4S, T5L, T4F, T5G, T61, T5o; E T5b, T6Z, T6o, T7h, T6b, T7c, T7x, T6U, T6H, Tbx, TaW, TbP, TaJ, TbK, Tc5; E Tbs, Tbf, T2r, T1Q, T2J, T1D, T2E, T2Z, T2m, T29, T3X, T3m, T4f, T39, T4a; E T4v, T3S, T3F, T8v, T7U, T8N, T7H, T8I, T93, T8q, T8d, Ta1, T9q, Taj, T9d; E Tae, Taz, T9W, T9J, Te, T19, T1u, T1g, Tv, TR, TG, TW, T5H, T4M, T5O; E T62, T5p, T53, T5u, T5e, T6i, T7d, T7y, T7k, T6z, T6V, T6K, T70, TbL, TaQ; E TbS, Tc6, Tbt, Tb7, Tby, Tbi, T1K, T2F, T30, T2M, T21, T2n, T2c, T2s, T4b; E T3g, T4i, T4w, T3T, T3x, T3Y, T3I, T7O, T8J, T94, T8Q, T85, T8r, T8g, T8w; E Tak, T9r, T9K, T9A, Taf, T9k, Tal, T9u; { E T9a, T9F, T99, Tac, T9p, T9b, T9G, T9H; { E TaG, Tbb, TaF, TbI, TaV, TaH, Tbc, Tbd; { E T4C, T57, T4B, T5E, T4R, T4D, T58, T59; { E T4, Tz, T3, T16, Tj, T5, TA, TB; { E T1, T2, Th, Ti; T1 = rio[0]; T2 = rio[WS(rs, 4)]; Th = iio[0]; Ti = iio[WS(rs, 4)]; T4 = rio[WS(rs, 2)]; Tz = T1 - T2; T3 = T1 + T2; T16 = Th + Ti; Tj = Th - Ti; T5 = rio[WS(rs, 6)]; TA = iio[WS(rs, 2)]; TB = iio[WS(rs, 6)]; } { E T4z, T4A, T4P, T4Q; T4z = rio[WS(vs, 3)]; { E Tg, T6, T17, TC; Tg = T4 - T5; T6 = T4 + T5; T17 = TA + TB; TC = TA - TB; TV = Tj - Tg; Tk = Tg + Tj; T1d = T3 - T6; T7 = T3 + T6; T18 = T16 - T17; T1t = T16 + T17; TQ = Tz + TC; TD = Tz - TC; T4A = rio[WS(vs, 3) + WS(rs, 4)]; } T4P = iio[WS(vs, 3)]; T4Q = iio[WS(vs, 3) + WS(rs, 4)]; T4C = rio[WS(vs, 3) + WS(rs, 2)]; T57 = T4z - T4A; T4B = T4z + T4A; T5E = T4P + T4Q; T4R = T4P - T4Q; T4D = rio[WS(vs, 3) + WS(rs, 6)]; T58 = iio[WS(vs, 3) + WS(rs, 2)]; T59 = iio[WS(vs, 3) + WS(rs, 6)]; } } { E T68, T6D, T67, T7a, T6n, T69, T6E, T6F; { E T65, T66, T6l, T6m; T65 = rio[WS(vs, 4)]; { E T4O, T4E, T5F, T5a; T4O = T4C - T4D; T4E = T4C + T4D; T5F = T58 + T59; T5a = T58 - T59; T5t = T4R - T4O; T4S = T4O + T4R; T5L = T4B - T4E; T4F = T4B + T4E; T5G = T5E - T5F; T61 = T5E + T5F; T5o = T57 + T5a; T5b = T57 - T5a; T66 = rio[WS(vs, 4) + WS(rs, 4)]; } T6l = iio[WS(vs, 4)]; T6m = iio[WS(vs, 4) + WS(rs, 4)]; T68 = rio[WS(vs, 4) + WS(rs, 2)]; T6D = T65 - T66; T67 = T65 + T66; T7a = T6l + T6m; T6n = T6l - T6m; T69 = rio[WS(vs, 4) + WS(rs, 6)]; T6E = iio[WS(vs, 4) + WS(rs, 2)]; T6F = iio[WS(vs, 4) + WS(rs, 6)]; } { E TaD, TaE, TaT, TaU; TaD = rio[WS(vs, 7)]; { E T6k, T6a, T7b, T6G; T6k = T68 - T69; T6a = T68 + T69; T7b = T6E + T6F; T6G = T6E - T6F; T6Z = T6n - T6k; T6o = T6k + T6n; T7h = T67 - T6a; T6b = T67 + T6a; T7c = T7a - T7b; T7x = T7a + T7b; T6U = T6D + T6G; T6H = T6D - T6G; TaE = rio[WS(vs, 7) + WS(rs, 4)]; } TaT = iio[WS(vs, 7)]; TaU = iio[WS(vs, 7) + WS(rs, 4)]; TaG = rio[WS(vs, 7) + WS(rs, 2)]; Tbb = TaD - TaE; TaF = TaD + TaE; TbI = TaT + TaU; TaV = TaT - TaU; TaH = rio[WS(vs, 7) + WS(rs, 6)]; Tbc = iio[WS(vs, 7) + WS(rs, 2)]; Tbd = iio[WS(vs, 7) + WS(rs, 6)]; } } } { E T36, T3B, T35, T48, T3l, T37, T3C, T3D; { E T1A, T25, T1z, T2C, T1P, T1B, T26, T27; { E T1x, T1y, T1N, T1O; T1x = rio[WS(vs, 1)]; { E TaS, TaI, TbJ, Tbe; TaS = TaG - TaH; TaI = TaG + TaH; TbJ = Tbc + Tbd; Tbe = Tbc - Tbd; Tbx = TaV - TaS; TaW = TaS + TaV; TbP = TaF - TaI; TaJ = TaF + TaI; TbK = TbI - TbJ; Tc5 = TbI + TbJ; Tbs = Tbb + Tbe; Tbf = Tbb - Tbe; T1y = rio[WS(vs, 1) + WS(rs, 4)]; } T1N = iio[WS(vs, 1)]; T1O = iio[WS(vs, 1) + WS(rs, 4)]; T1A = rio[WS(vs, 1) + WS(rs, 2)]; T25 = T1x - T1y; T1z = T1x + T1y; T2C = T1N + T1O; T1P = T1N - T1O; T1B = rio[WS(vs, 1) + WS(rs, 6)]; T26 = iio[WS(vs, 1) + WS(rs, 2)]; T27 = iio[WS(vs, 1) + WS(rs, 6)]; } { E T33, T34, T3j, T3k; T33 = rio[WS(vs, 2)]; { E T1M, T1C, T2D, T28; T1M = T1A - T1B; T1C = T1A + T1B; T2D = T26 + T27; T28 = T26 - T27; T2r = T1P - T1M; T1Q = T1M + T1P; T2J = T1z - T1C; T1D = T1z + T1C; T2E = T2C - T2D; T2Z = T2C + T2D; T2m = T25 + T28; T29 = T25 - T28; T34 = rio[WS(vs, 2) + WS(rs, 4)]; } T3j = iio[WS(vs, 2)]; T3k = iio[WS(vs, 2) + WS(rs, 4)]; T36 = rio[WS(vs, 2) + WS(rs, 2)]; T3B = T33 - T34; T35 = T33 + T34; T48 = T3j + T3k; T3l = T3j - T3k; T37 = rio[WS(vs, 2) + WS(rs, 6)]; T3C = iio[WS(vs, 2) + WS(rs, 2)]; T3D = iio[WS(vs, 2) + WS(rs, 6)]; } } { E T7E, T89, T7D, T8G, T7T, T7F, T8a, T8b; { E T7B, T7C, T7R, T7S; T7B = rio[WS(vs, 5)]; { E T3i, T38, T49, T3E; T3i = T36 - T37; T38 = T36 + T37; T49 = T3C + T3D; T3E = T3C - T3D; T3X = T3l - T3i; T3m = T3i + T3l; T4f = T35 - T38; T39 = T35 + T38; T4a = T48 - T49; T4v = T48 + T49; T3S = T3B + T3E; T3F = T3B - T3E; T7C = rio[WS(vs, 5) + WS(rs, 4)]; } T7R = iio[WS(vs, 5)]; T7S = iio[WS(vs, 5) + WS(rs, 4)]; T7E = rio[WS(vs, 5) + WS(rs, 2)]; T89 = T7B - T7C; T7D = T7B + T7C; T8G = T7R + T7S; T7T = T7R - T7S; T7F = rio[WS(vs, 5) + WS(rs, 6)]; T8a = iio[WS(vs, 5) + WS(rs, 2)]; T8b = iio[WS(vs, 5) + WS(rs, 6)]; } { E T97, T98, T9n, T9o; T97 = rio[WS(vs, 6)]; { E T7Q, T7G, T8H, T8c; T7Q = T7E - T7F; T7G = T7E + T7F; T8H = T8a + T8b; T8c = T8a - T8b; T8v = T7T - T7Q; T7U = T7Q + T7T; T8N = T7D - T7G; T7H = T7D + T7G; T8I = T8G - T8H; T93 = T8G + T8H; T8q = T89 + T8c; T8d = T89 - T8c; T98 = rio[WS(vs, 6) + WS(rs, 4)]; } T9n = iio[WS(vs, 6)]; T9o = iio[WS(vs, 6) + WS(rs, 4)]; T9a = rio[WS(vs, 6) + WS(rs, 2)]; T9F = T97 - T98; T99 = T97 + T98; Tac = T9n + T9o; T9p = T9n - T9o; T9b = rio[WS(vs, 6) + WS(rs, 6)]; T9G = iio[WS(vs, 6) + WS(rs, 2)]; T9H = iio[WS(vs, 6) + WS(rs, 6)]; } } } } { E TbQ, TaX, Tbg, Tb6, TbR, Tb0; { E T5M, T4T, T5c, T52, T5N, T4W; { E Tu, TE, TF, Tp; { E Tb, Tq, Ta, T1e, Tt, Tc, Tm, Tn; { E T8, T9, Tr, Ts; T8 = rio[WS(rs, 1)]; { E T9m, T9c, Tad, T9I; T9m = T9a - T9b; T9c = T9a + T9b; Tad = T9G + T9H; T9I = T9G - T9H; Ta1 = T9p - T9m; T9q = T9m + T9p; Taj = T99 - T9c; T9d = T99 + T9c; Tae = Tac - Tad; Taz = Tac + Tad; T9W = T9F + T9I; T9J = T9F - T9I; T9 = rio[WS(rs, 5)]; } Tr = iio[WS(rs, 1)]; Ts = iio[WS(rs, 5)]; Tb = rio[WS(rs, 7)]; Tq = T8 - T9; Ta = T8 + T9; T1e = Tr + Ts; Tt = Tr - Ts; Tc = rio[WS(rs, 3)]; Tm = iio[WS(rs, 7)]; Tn = iio[WS(rs, 3)]; } { E Tl, Td, T1f, To; Tu = Tq + Tt; TE = Tt - Tq; Tl = Tb - Tc; Td = Tb + Tc; T1f = Tm + Tn; To = Tm - Tn; Te = Ta + Td; T19 = Td - Ta; T1u = T1e + T1f; T1g = T1e - T1f; TF = Tl + To; Tp = Tl - To; } } { E T4I, T4Y, T4U, T51, T4L, T4V; { E T4Z, T50, T4G, T4H, T4J, T4K; T4G = rio[WS(vs, 3) + WS(rs, 1)]; T4H = rio[WS(vs, 3) + WS(rs, 5)]; Tv = Tp - Tu; TR = Tu + Tp; TG = TE - TF; TW = TE + TF; T4I = T4G + T4H; T4Y = T4G - T4H; T4Z = iio[WS(vs, 3) + WS(rs, 1)]; T50 = iio[WS(vs, 3) + WS(rs, 5)]; T4J = rio[WS(vs, 3) + WS(rs, 7)]; T4K = rio[WS(vs, 3) + WS(rs, 3)]; T4U = iio[WS(vs, 3) + WS(rs, 7)]; T51 = T4Z - T50; T5M = T4Z + T50; T4L = T4J + T4K; T4T = T4J - T4K; T4V = iio[WS(vs, 3) + WS(rs, 3)]; } T5c = T51 - T4Y; T52 = T4Y + T51; T5H = T4L - T4I; T4M = T4I + T4L; T5N = T4U + T4V; T4W = T4U - T4V; } } { E T7i, T6p, T6y, T6I, T6s, T7j; { E T6e, T6u, T6q, T6x, T6h, T6r; { E T6v, T6w, T6f, T6g; { E T4X, T5d, T6c, T6d; T6c = rio[WS(vs, 4) + WS(rs, 1)]; T6d = rio[WS(vs, 4) + WS(rs, 5)]; T5O = T5M - T5N; T62 = T5M + T5N; T4X = T4T - T4W; T5d = T4T + T4W; T6e = T6c + T6d; T6u = T6c - T6d; T5p = T52 + T4X; T53 = T4X - T52; T5u = T5c + T5d; T5e = T5c - T5d; T6v = iio[WS(vs, 4) + WS(rs, 1)]; T6w = iio[WS(vs, 4) + WS(rs, 5)]; } T6f = rio[WS(vs, 4) + WS(rs, 7)]; T6g = rio[WS(vs, 4) + WS(rs, 3)]; T6q = iio[WS(vs, 4) + WS(rs, 7)]; T7i = T6v + T6w; T6x = T6v - T6w; T6p = T6f - T6g; T6h = T6f + T6g; T6r = iio[WS(vs, 4) + WS(rs, 3)]; } T6y = T6u + T6x; T6I = T6x - T6u; T6i = T6e + T6h; T7d = T6h - T6e; T6s = T6q - T6r; T7j = T6q + T6r; } { E Tb2, TaM, TaY, Tb5, TaP, TaZ; { E Tb3, Tb4, TaN, TaO; { E T6J, T6t, TaK, TaL; TaK = rio[WS(vs, 7) + WS(rs, 1)]; TaL = rio[WS(vs, 7) + WS(rs, 5)]; T7y = T7i + T7j; T7k = T7i - T7j; T6J = T6p + T6s; T6t = T6p - T6s; Tb2 = TaK - TaL; TaM = TaK + TaL; T6z = T6t - T6y; T6V = T6y + T6t; T6K = T6I - T6J; T70 = T6I + T6J; Tb3 = iio[WS(vs, 7) + WS(rs, 1)]; Tb4 = iio[WS(vs, 7) + WS(rs, 5)]; } TaN = rio[WS(vs, 7) + WS(rs, 7)]; TaO = rio[WS(vs, 7) + WS(rs, 3)]; TaY = iio[WS(vs, 7) + WS(rs, 7)]; Tb5 = Tb3 - Tb4; TbQ = Tb3 + Tb4; TaP = TaN + TaO; TaX = TaN - TaO; TaZ = iio[WS(vs, 7) + WS(rs, 3)]; } Tbg = Tb5 - Tb2; Tb6 = Tb2 + Tb5; TbL = TaP - TaM; TaQ = TaM + TaP; TbR = TaY + TaZ; Tb0 = TaY - TaZ; } } } { E T4g, T3n, T3G, T3w, T4h, T3q; { E T2K, T1R, T20, T2a, T1U, T2L; { E T1G, T1W, T1S, T1Z, T1J, T1T; { E T1X, T1Y, T1H, T1I; { E Tb1, Tbh, T1E, T1F; T1E = rio[WS(vs, 1) + WS(rs, 1)]; T1F = rio[WS(vs, 1) + WS(rs, 5)]; TbS = TbQ - TbR; Tc6 = TbQ + TbR; Tb1 = TaX - Tb0; Tbh = TaX + Tb0; T1G = T1E + T1F; T1W = T1E - T1F; Tbt = Tb6 + Tb1; Tb7 = Tb1 - Tb6; Tby = Tbg + Tbh; Tbi = Tbg - Tbh; T1X = iio[WS(vs, 1) + WS(rs, 1)]; T1Y = iio[WS(vs, 1) + WS(rs, 5)]; } T1H = rio[WS(vs, 1) + WS(rs, 7)]; T1I = rio[WS(vs, 1) + WS(rs, 3)]; T1S = iio[WS(vs, 1) + WS(rs, 7)]; T2K = T1X + T1Y; T1Z = T1X - T1Y; T1R = T1H - T1I; T1J = T1H + T1I; T1T = iio[WS(vs, 1) + WS(rs, 3)]; } T20 = T1W + T1Z; T2a = T1Z - T1W; T1K = T1G + T1J; T2F = T1J - T1G; T1U = T1S - T1T; T2L = T1S + T1T; } { E T3s, T3c, T3o, T3v, T3f, T3p; { E T3t, T3u, T3d, T3e; { E T2b, T1V, T3a, T3b; T3a = rio[WS(vs, 2) + WS(rs, 1)]; T3b = rio[WS(vs, 2) + WS(rs, 5)]; T30 = T2K + T2L; T2M = T2K - T2L; T2b = T1R + T1U; T1V = T1R - T1U; T3s = T3a - T3b; T3c = T3a + T3b; T21 = T1V - T20; T2n = T20 + T1V; T2c = T2a - T2b; T2s = T2a + T2b; T3t = iio[WS(vs, 2) + WS(rs, 1)]; T3u = iio[WS(vs, 2) + WS(rs, 5)]; } T3d = rio[WS(vs, 2) + WS(rs, 7)]; T3e = rio[WS(vs, 2) + WS(rs, 3)]; T3o = iio[WS(vs, 2) + WS(rs, 7)]; T3v = T3t - T3u; T4g = T3t + T3u; T3f = T3d + T3e; T3n = T3d - T3e; T3p = iio[WS(vs, 2) + WS(rs, 3)]; } T3G = T3v - T3s; T3w = T3s + T3v; T4b = T3f - T3c; T3g = T3c + T3f; T4h = T3o + T3p; T3q = T3o - T3p; } } { E T8O, T7V, T84, T8e, T7Y, T8P; { E T7K, T80, T7W, T83, T7N, T7X; { E T81, T82, T7L, T7M; { E T3r, T3H, T7I, T7J; T7I = rio[WS(vs, 5) + WS(rs, 1)]; T7J = rio[WS(vs, 5) + WS(rs, 5)]; T4i = T4g - T4h; T4w = T4g + T4h; T3r = T3n - T3q; T3H = T3n + T3q; T7K = T7I + T7J; T80 = T7I - T7J; T3T = T3w + T3r; T3x = T3r - T3w; T3Y = T3G + T3H; T3I = T3G - T3H; T81 = iio[WS(vs, 5) + WS(rs, 1)]; T82 = iio[WS(vs, 5) + WS(rs, 5)]; } T7L = rio[WS(vs, 5) + WS(rs, 7)]; T7M = rio[WS(vs, 5) + WS(rs, 3)]; T7W = iio[WS(vs, 5) + WS(rs, 7)]; T8O = T81 + T82; T83 = T81 - T82; T7V = T7L - T7M; T7N = T7L + T7M; T7X = iio[WS(vs, 5) + WS(rs, 3)]; } T84 = T80 + T83; T8e = T83 - T80; T7O = T7K + T7N; T8J = T7N - T7K; T7Y = T7W - T7X; T8P = T7W + T7X; } { E T9w, T9g, T9s, T9z, T9j, T9t; { E T9x, T9y, T9h, T9i; { E T8f, T7Z, T9e, T9f; T9e = rio[WS(vs, 6) + WS(rs, 1)]; T9f = rio[WS(vs, 6) + WS(rs, 5)]; T94 = T8O + T8P; T8Q = T8O - T8P; T8f = T7V + T7Y; T7Z = T7V - T7Y; T9w = T9e - T9f; T9g = T9e + T9f; T85 = T7Z - T84; T8r = T84 + T7Z; T8g = T8e - T8f; T8w = T8e + T8f; T9x = iio[WS(vs, 6) + WS(rs, 1)]; T9y = iio[WS(vs, 6) + WS(rs, 5)]; } T9h = rio[WS(vs, 6) + WS(rs, 7)]; T9i = rio[WS(vs, 6) + WS(rs, 3)]; T9s = iio[WS(vs, 6) + WS(rs, 7)]; T9z = T9x - T9y; Tak = T9x + T9y; T9j = T9h + T9i; T9r = T9h - T9i; T9t = iio[WS(vs, 6) + WS(rs, 3)]; } T9K = T9z - T9w; T9A = T9w + T9z; Taf = T9j - T9g; T9k = T9g + T9j; Tal = T9s + T9t; T9u = T9s - T9t; } } } } } { E T9X, T9B, Ta2, T9M, T2T, T2Q, TbT, TbH, TbO, TbN, TbU; { E Tam, TaA, T9v, T9L; rio[0] = T7 + Te; iio[0] = T1t + T1u; Tam = Tak - Tal; TaA = Tak + Tal; T9v = T9r - T9u; T9L = T9r + T9u; rio[WS(rs, 1)] = T1D + T1K; iio[WS(rs, 1)] = T2Z + T30; T9X = T9A + T9v; T9B = T9v - T9A; Ta2 = T9K + T9L; T9M = T9K - T9L; rio[WS(rs, 2)] = T39 + T3g; iio[WS(rs, 2)] = T4v + T4w; rio[WS(rs, 3)] = T4F + T4M; iio[WS(rs, 3)] = T61 + T62; rio[WS(rs, 4)] = T6b + T6i; iio[WS(rs, 4)] = T7x + T7y; rio[WS(rs, 5)] = T7H + T7O; iio[WS(rs, 5)] = T93 + T94; rio[WS(rs, 6)] = T9d + T9k; iio[WS(rs, 6)] = Taz + TaA; rio[WS(rs, 7)] = TaJ + TaQ; iio[WS(rs, 7)] = Tc5 + Tc6; { E T10, T13, T1h, T1a, Tat, Taq, TbC, TbF, TbE, TbG, TbD; { E T1q, T1v, T1s, T1w, T1r; { E T2N, T2B, T2I, T2H, T2O; { E TS, TX, TP, TU, T2G, TY, TT; T10 = FMA(KP707106781, TR, TQ); TS = FNMS(KP707106781, TR, TQ); TX = FNMS(KP707106781, TW, TV); T13 = FMA(KP707106781, TW, TV); TP = W[8]; TU = W[9]; T2T = T2J + T2M; T2N = T2J - T2M; T2G = T2E - T2F; T2Q = T2F + T2E; TY = TP * TX; TT = TP * TS; T2B = W[10]; T2I = W[11]; iio[WS(vs, 5)] = FNMS(TU, TS, TY); rio[WS(vs, 5)] = FMA(TU, TX, TT); T2H = T2B * T2G; T2O = T2I * T2G; } { E T1n, T1k, T1j, T1m, T1l, T1o, T1p; T1h = T1d - T1g; T1n = T1d + T1g; T1k = T19 + T18; T1a = T18 - T19; iio[WS(vs, 6) + WS(rs, 1)] = FNMS(T2I, T2N, T2H); rio[WS(vs, 6) + WS(rs, 1)] = FMA(T2B, T2N, T2O); T1j = W[2]; T1m = W[3]; T1q = T7 - Te; T1v = T1t - T1u; T1l = T1j * T1k; T1o = T1m * T1k; T1p = W[6]; T1s = W[7]; iio[WS(vs, 2)] = FNMS(T1m, T1n, T1l); rio[WS(vs, 2)] = FMA(T1j, T1n, T1o); T1w = T1p * T1v; T1r = T1p * T1q; } } { E Tc2, Tc7, Tc4, Tc8, Tc3; { E Tan, Tag, Tab, Tai, Tah, Tao, Tc1; Tat = Taj + Tam; Tan = Taj - Tam; Tag = Tae - Taf; Taq = Taf + Tae; iio[WS(vs, 4)] = FNMS(T1s, T1q, T1w); rio[WS(vs, 4)] = FMA(T1s, T1v, T1r); Tab = W[10]; Tai = W[11]; Tc2 = TaJ - TaQ; Tc7 = Tc5 - Tc6; Tah = Tab * Tag; Tao = Tai * Tag; Tc1 = W[6]; Tc4 = W[7]; iio[WS(vs, 6) + WS(rs, 6)] = FNMS(Tai, Tan, Tah); rio[WS(vs, 6) + WS(rs, 6)] = FMA(Tab, Tan, Tao); Tc8 = Tc1 * Tc7; Tc3 = Tc1 * Tc2; } { E Tbu, Tbz, Tbr, Tbw, TbA, Tbv, TbB; TbC = FMA(KP707106781, Tbt, Tbs); Tbu = FNMS(KP707106781, Tbt, Tbs); Tbz = FNMS(KP707106781, Tby, Tbx); TbF = FMA(KP707106781, Tby, Tbx); iio[WS(vs, 4) + WS(rs, 7)] = FNMS(Tc4, Tc2, Tc8); rio[WS(vs, 4) + WS(rs, 7)] = FMA(Tc4, Tc7, Tc3); Tbr = W[8]; Tbw = W[9]; TbA = Tbr * Tbz; Tbv = Tbr * Tbu; TbB = W[0]; TbE = W[1]; iio[WS(vs, 5) + WS(rs, 7)] = FNMS(Tbw, Tbu, TbA); rio[WS(vs, 5) + WS(rs, 7)] = FMA(Tbw, Tbz, Tbv); TbG = TbB * TbF; TbD = TbB * TbC; } } } { E T2o, T2t, T2q, T2u, T2p; { E T2w, T2z, T2y, T2A, T2x; { E TZ, T12, T14, T11, T2v; iio[WS(vs, 1) + WS(rs, 7)] = FNMS(TbE, TbC, TbG); rio[WS(vs, 1) + WS(rs, 7)] = FMA(TbE, TbF, TbD); TZ = W[0]; T12 = W[1]; T2o = FNMS(KP707106781, T2n, T2m); T2w = FMA(KP707106781, T2n, T2m); T2z = FMA(KP707106781, T2s, T2r); T2t = FNMS(KP707106781, T2s, T2r); T14 = TZ * T13; T11 = TZ * T10; T2v = W[0]; T2y = W[1]; iio[WS(vs, 1)] = FNMS(T12, T10, T14); rio[WS(vs, 1)] = FMA(T12, T13, T11); T2A = T2v * T2z; T2x = T2v * T2w; } { E T15, T1c, T1b, T1i, T2l; iio[WS(vs, 1) + WS(rs, 1)] = FNMS(T2y, T2w, T2A); rio[WS(vs, 1) + WS(rs, 1)] = FMA(T2y, T2z, T2x); T15 = W[10]; T1c = W[11]; T1b = T15 * T1a; T1i = T1c * T1a; T2l = W[8]; T2q = W[9]; iio[WS(vs, 6)] = FNMS(T1c, T1h, T1b); rio[WS(vs, 6)] = FMA(T15, T1h, T1i); T2u = T2l * T2t; T2p = T2l * T2o; } } { E TbZ, TbM, TbV, TbY, TbX, Tc0; { E Tap, Tas, TbW, Tar, Tau; iio[WS(vs, 5) + WS(rs, 1)] = FNMS(T2q, T2o, T2u); rio[WS(vs, 5) + WS(rs, 1)] = FMA(T2q, T2t, T2p); Tap = W[2]; Tas = W[3]; TbT = TbP - TbS; TbZ = TbP + TbS; TbW = TbL + TbK; TbM = TbK - TbL; Tar = Tap * Taq; Tau = Tas * Taq; TbV = W[2]; TbY = W[3]; iio[WS(vs, 2) + WS(rs, 6)] = FNMS(Tas, Tat, Tar); rio[WS(vs, 2) + WS(rs, 6)] = FMA(Tap, Tat, Tau); TbX = TbV * TbW; Tc0 = TbY * TbW; } { E Taw, TaB, Tav, Tay, TaC, Tax; Taw = T9d - T9k; TaB = Taz - TaA; iio[WS(vs, 2) + WS(rs, 7)] = FNMS(TbY, TbZ, TbX); rio[WS(vs, 2) + WS(rs, 7)] = FMA(TbV, TbZ, Tc0); Tav = W[6]; Tay = W[7]; TaC = Tav * TaB; Tax = Tav * Taw; TbH = W[10]; TbO = W[11]; iio[WS(vs, 4) + WS(rs, 6)] = FNMS(Tay, Taw, TaC); rio[WS(vs, 4) + WS(rs, 6)] = FMA(Tay, TaB, Tax); TbN = TbH * TbM; TbU = TbO * TbM; } } } } } { E T5q, T5v, T8R, T8K, T90, T95, T92, T96, T91; { E T3U, T3Z, T74, T77, T9Y, Ta3, T7l, T7e, T8X, T8T, T8W, T8V, T8Y; { E T5y, T5B, T5A, T5C, T5z; { E T5Y, T63, T60, T64, T5Z; { E T2P, T2S, T2R, T2U, T5X; iio[WS(vs, 6) + WS(rs, 7)] = FNMS(TbO, TbT, TbN); rio[WS(vs, 6) + WS(rs, 7)] = FMA(TbH, TbT, TbU); T2P = W[2]; T2S = W[3]; T5Y = T4F - T4M; T63 = T61 - T62; T2R = T2P * T2Q; T2U = T2S * T2Q; T5X = W[6]; T60 = W[7]; iio[WS(vs, 2) + WS(rs, 1)] = FNMS(T2S, T2T, T2R); rio[WS(vs, 2) + WS(rs, 1)] = FMA(T2P, T2T, T2U); T64 = T5X * T63; T5Z = T5X * T5Y; } { E T42, T45, T41, T44, T46, T43, T5x; T3U = FNMS(KP707106781, T3T, T3S); T42 = FMA(KP707106781, T3T, T3S); T45 = FMA(KP707106781, T3Y, T3X); T3Z = FNMS(KP707106781, T3Y, T3X); iio[WS(vs, 4) + WS(rs, 3)] = FNMS(T60, T5Y, T64); rio[WS(vs, 4) + WS(rs, 3)] = FMA(T60, T63, T5Z); T41 = W[0]; T44 = W[1]; T5q = FNMS(KP707106781, T5p, T5o); T5y = FMA(KP707106781, T5p, T5o); T5B = FMA(KP707106781, T5u, T5t); T5v = FNMS(KP707106781, T5u, T5t); T46 = T41 * T45; T43 = T41 * T42; T5x = W[0]; T5A = W[1]; iio[WS(vs, 1) + WS(rs, 2)] = FNMS(T44, T42, T46); rio[WS(vs, 1) + WS(rs, 2)] = FMA(T44, T45, T43); T5C = T5x * T5B; T5z = T5x * T5y; } } { E Ta6, Ta9, Ta8, Taa, Ta7; { E T6W, T71, T6T, T6Y, T72, T6X, Ta5; T74 = FMA(KP707106781, T6V, T6U); T6W = FNMS(KP707106781, T6V, T6U); T71 = FNMS(KP707106781, T70, T6Z); T77 = FMA(KP707106781, T70, T6Z); iio[WS(vs, 1) + WS(rs, 3)] = FNMS(T5A, T5y, T5C); rio[WS(vs, 1) + WS(rs, 3)] = FMA(T5A, T5B, T5z); T6T = W[8]; T6Y = W[9]; T9Y = FNMS(KP707106781, T9X, T9W); Ta6 = FMA(KP707106781, T9X, T9W); Ta9 = FMA(KP707106781, Ta2, Ta1); Ta3 = FNMS(KP707106781, Ta2, Ta1); T72 = T6T * T71; T6X = T6T * T6W; Ta5 = W[0]; Ta8 = W[1]; iio[WS(vs, 5) + WS(rs, 4)] = FNMS(T6Y, T6W, T72); rio[WS(vs, 5) + WS(rs, 4)] = FMA(T6Y, T71, T6X); Taa = Ta5 * Ta9; Ta7 = Ta5 * Ta6; } { E T7r, T7o, T7n, T7q, T8U, T7p, T7s; T7l = T7h - T7k; T7r = T7h + T7k; T7o = T7d + T7c; T7e = T7c - T7d; iio[WS(vs, 1) + WS(rs, 6)] = FNMS(Ta8, Ta6, Taa); rio[WS(vs, 1) + WS(rs, 6)] = FMA(Ta8, Ta9, Ta7); T7n = W[2]; T7q = W[3]; T8R = T8N - T8Q; T8X = T8N + T8Q; T8U = T8J + T8I; T8K = T8I - T8J; T7p = T7n * T7o; T7s = T7q * T7o; T8T = W[2]; T8W = W[3]; iio[WS(vs, 2) + WS(rs, 4)] = FNMS(T7q, T7r, T7p); rio[WS(vs, 2) + WS(rs, 4)] = FMA(T7n, T7r, T7s); T8V = T8T * T8U; T8Y = T8W * T8U; } } } { E T5P, T5D, T5K, T5J, T5Q, Ta0, Ta4, T9Z; { E T5V, T5I, T5R, T5U, T5T, T5W; { E T2W, T31, T2V, T2Y, T5S, T32, T2X; T2W = T1D - T1K; T31 = T2Z - T30; iio[WS(vs, 2) + WS(rs, 5)] = FNMS(T8W, T8X, T8V); rio[WS(vs, 2) + WS(rs, 5)] = FMA(T8T, T8X, T8Y); T2V = W[6]; T2Y = W[7]; T5P = T5L - T5O; T5V = T5L + T5O; T5S = T5H + T5G; T5I = T5G - T5H; T32 = T2V * T31; T2X = T2V * T2W; T5R = W[2]; T5U = W[3]; iio[WS(vs, 4) + WS(rs, 1)] = FNMS(T2Y, T2W, T32); rio[WS(vs, 4) + WS(rs, 1)] = FMA(T2Y, T31, T2X); T5T = T5R * T5S; T5W = T5U * T5S; } { E T3R, T3W, T40, T3V; iio[WS(vs, 2) + WS(rs, 3)] = FNMS(T5U, T5V, T5T); rio[WS(vs, 2) + WS(rs, 3)] = FMA(T5R, T5V, T5W); T3R = W[8]; T3W = W[9]; T40 = T3R * T3Z; T3V = T3R * T3U; T5D = W[10]; T5K = W[11]; iio[WS(vs, 5) + WS(rs, 2)] = FNMS(T3W, T3U, T40); rio[WS(vs, 5) + WS(rs, 2)] = FMA(T3W, T3Z, T3V); T5J = T5D * T5I; T5Q = T5K * T5I; } } { E T73, T76, T78, T75, T9V; iio[WS(vs, 6) + WS(rs, 3)] = FNMS(T5K, T5P, T5J); rio[WS(vs, 6) + WS(rs, 3)] = FMA(T5D, T5P, T5Q); T73 = W[0]; T76 = W[1]; T78 = T73 * T77; T75 = T73 * T74; T9V = W[8]; Ta0 = W[9]; iio[WS(vs, 1) + WS(rs, 4)] = FNMS(T76, T74, T78); rio[WS(vs, 1) + WS(rs, 4)] = FMA(T76, T77, T75); Ta4 = T9V * Ta3; T9Z = T9V * T9Y; } { E T79, T7g, T7f, T7m, T8Z; iio[WS(vs, 5) + WS(rs, 6)] = FNMS(Ta0, T9Y, Ta4); rio[WS(vs, 5) + WS(rs, 6)] = FMA(Ta0, Ta3, T9Z); T79 = W[10]; T7g = W[11]; T90 = T7H - T7O; T95 = T93 - T94; T7f = T79 * T7e; T7m = T7g * T7e; T8Z = W[6]; T92 = W[7]; iio[WS(vs, 6) + WS(rs, 4)] = FNMS(T7g, T7l, T7f); rio[WS(vs, 6) + WS(rs, 4)] = FMA(T79, T7l, T7m); T96 = T8Z * T95; T91 = T8Z * T90; } } } { E T8A, T8D, T8C, T8E, T8B; { E T4s, T4x, T4u, T4y, T4t; { E T4p, T4m, T5s, T5w, T5r; { E T4j, T4c, T47, T4e, T4d, T4k, T5n; T4p = T4f + T4i; T4j = T4f - T4i; T4c = T4a - T4b; T4m = T4b + T4a; iio[WS(vs, 4) + WS(rs, 5)] = FNMS(T92, T90, T96); rio[WS(vs, 4) + WS(rs, 5)] = FMA(T92, T95, T91); T47 = W[10]; T4e = W[11]; T4d = T47 * T4c; T4k = T4e * T4c; T5n = W[8]; T5s = W[9]; iio[WS(vs, 6) + WS(rs, 2)] = FNMS(T4e, T4j, T4d); rio[WS(vs, 6) + WS(rs, 2)] = FMA(T47, T4j, T4k); T5w = T5n * T5v; T5r = T5n * T5q; } { E T4l, T4o, T4n, T4q, T4r; iio[WS(vs, 5) + WS(rs, 3)] = FNMS(T5s, T5q, T5w); rio[WS(vs, 5) + WS(rs, 3)] = FMA(T5s, T5v, T5r); T4l = W[2]; T4o = W[3]; T4s = T39 - T3g; T4x = T4v - T4w; T4n = T4l * T4m; T4q = T4o * T4m; T4r = W[6]; T4u = W[7]; iio[WS(vs, 2) + WS(rs, 2)] = FNMS(T4o, T4p, T4n); rio[WS(vs, 2) + WS(rs, 2)] = FMA(T4l, T4p, T4q); T4y = T4r * T4x; T4t = T4r * T4s; } } { E T8F, T8M, T8L, T8S; { E T7u, T7z, T7t, T7w, T7A, T7v; T7u = T6b - T6i; T7z = T7x - T7y; iio[WS(vs, 4) + WS(rs, 2)] = FNMS(T4u, T4s, T4y); rio[WS(vs, 4) + WS(rs, 2)] = FMA(T4u, T4x, T4t); T7t = W[6]; T7w = W[7]; T7A = T7t * T7z; T7v = T7t * T7u; T8F = W[10]; T8M = W[11]; iio[WS(vs, 4) + WS(rs, 4)] = FNMS(T7w, T7u, T7A); rio[WS(vs, 4) + WS(rs, 4)] = FMA(T7w, T7z, T7v); T8L = T8F * T8K; T8S = T8M * T8K; } { E T8s, T8x, T8p, T8u, T8y, T8t, T8z; T8A = FMA(KP707106781, T8r, T8q); T8s = FNMS(KP707106781, T8r, T8q); T8x = FNMS(KP707106781, T8w, T8v); T8D = FMA(KP707106781, T8w, T8v); iio[WS(vs, 6) + WS(rs, 5)] = FNMS(T8M, T8R, T8L); rio[WS(vs, 6) + WS(rs, 5)] = FMA(T8F, T8R, T8S); T8p = W[8]; T8u = W[9]; T8y = T8p * T8x; T8t = T8p * T8s; T8z = W[0]; T8C = W[1]; iio[WS(vs, 5) + WS(rs, 5)] = FNMS(T8u, T8s, T8y); rio[WS(vs, 5) + WS(rs, 5)] = FMA(T8u, T8x, T8t); T8E = T8z * T8D; T8B = T8z * T8A; } } } { E T3y, T3J, T3h, T3A, T3z, T3K; { E T54, T5f, T4N, T56, T55, T5g; { E Tw, TH, Tf, Ty, Tx, TI; { E TN, TJ, TM, TL, TO, TK; TK = FMA(KP707106781, Tv, Tk); Tw = FNMS(KP707106781, Tv, Tk); iio[WS(vs, 1) + WS(rs, 5)] = FNMS(T8C, T8A, T8E); rio[WS(vs, 1) + WS(rs, 5)] = FMA(T8C, T8D, T8B); TH = FNMS(KP707106781, TG, TD); TN = FMA(KP707106781, TG, TD); TJ = W[4]; TM = W[5]; Tf = W[12]; TL = TJ * TK; TO = TM * TK; Ty = W[13]; Tx = Tf * Tw; iio[WS(vs, 3)] = FNMS(TM, TN, TL); rio[WS(vs, 3)] = FMA(TJ, TN, TO); } TI = Ty * Tw; iio[WS(vs, 7)] = FNMS(Ty, TH, Tx); { E T5h, T5l, T5k, T5j, T5m, T5i; T5i = FMA(KP707106781, T53, T4S); T54 = FNMS(KP707106781, T53, T4S); rio[WS(vs, 7)] = FMA(Tf, TH, TI); T5h = W[4]; T5f = FNMS(KP707106781, T5e, T5b); T5l = FMA(KP707106781, T5e, T5b); T5k = W[5]; T5j = T5h * T5i; T4N = W[12]; T5m = T5k * T5i; T56 = W[13]; iio[WS(vs, 3) + WS(rs, 3)] = FNMS(T5k, T5l, T5j); T55 = T4N * T54; rio[WS(vs, 3) + WS(rs, 3)] = FMA(T5h, T5l, T5m); } } T5g = T56 * T54; { E T22, T2d, T1L, T24, T23, T2e; { E T2j, T2f, T2i, T2h, T2k, T2g; iio[WS(vs, 7) + WS(rs, 3)] = FNMS(T56, T5f, T55); T22 = FNMS(KP707106781, T21, T1Q); T2g = FMA(KP707106781, T21, T1Q); rio[WS(vs, 7) + WS(rs, 3)] = FMA(T4N, T5f, T5g); T2d = FNMS(KP707106781, T2c, T29); T2j = FMA(KP707106781, T2c, T29); T2f = W[4]; T2i = W[5]; T1L = W[12]; T2h = T2f * T2g; T2k = T2i * T2g; T24 = W[13]; T23 = T1L * T22; iio[WS(vs, 3) + WS(rs, 1)] = FNMS(T2i, T2j, T2h); rio[WS(vs, 3) + WS(rs, 1)] = FMA(T2f, T2j, T2k); } T2e = T24 * T22; iio[WS(vs, 7) + WS(rs, 1)] = FNMS(T24, T2d, T23); { E T3L, T3P, T3O, T3N, T3Q, T3M; T3M = FMA(KP707106781, T3x, T3m); T3y = FNMS(KP707106781, T3x, T3m); rio[WS(vs, 7) + WS(rs, 1)] = FMA(T1L, T2d, T2e); T3L = W[4]; T3J = FNMS(KP707106781, T3I, T3F); T3P = FMA(KP707106781, T3I, T3F); T3O = W[5]; T3N = T3L * T3M; T3h = W[12]; T3Q = T3O * T3M; T3A = W[13]; iio[WS(vs, 3) + WS(rs, 2)] = FNMS(T3O, T3P, T3N); T3z = T3h * T3y; rio[WS(vs, 3) + WS(rs, 2)] = FMA(T3L, T3P, T3Q); } } } T3K = T3A * T3y; { E Tb8, Tbj, TaR, Tba, Tb9, Tbk; { E T6A, T6L, T6j, T6C, T6B, T6M; { E T6R, T6N, T6Q, T6P, T6S, T6O; iio[WS(vs, 7) + WS(rs, 2)] = FNMS(T3A, T3J, T3z); T6A = FNMS(KP707106781, T6z, T6o); T6O = FMA(KP707106781, T6z, T6o); rio[WS(vs, 7) + WS(rs, 2)] = FMA(T3h, T3J, T3K); T6L = FNMS(KP707106781, T6K, T6H); T6R = FMA(KP707106781, T6K, T6H); T6N = W[4]; T6Q = W[5]; T6j = W[12]; T6P = T6N * T6O; T6S = T6Q * T6O; T6C = W[13]; T6B = T6j * T6A; iio[WS(vs, 3) + WS(rs, 4)] = FNMS(T6Q, T6R, T6P); rio[WS(vs, 3) + WS(rs, 4)] = FMA(T6N, T6R, T6S); } T6M = T6C * T6A; iio[WS(vs, 7) + WS(rs, 4)] = FNMS(T6C, T6L, T6B); { E Tbl, Tbp, Tbo, Tbn, Tbq, Tbm; Tbm = FMA(KP707106781, Tb7, TaW); Tb8 = FNMS(KP707106781, Tb7, TaW); rio[WS(vs, 7) + WS(rs, 4)] = FMA(T6j, T6L, T6M); Tbl = W[4]; Tbj = FNMS(KP707106781, Tbi, Tbf); Tbp = FMA(KP707106781, Tbi, Tbf); Tbo = W[5]; Tbn = Tbl * Tbm; TaR = W[12]; Tbq = Tbo * Tbm; Tba = W[13]; iio[WS(vs, 3) + WS(rs, 7)] = FNMS(Tbo, Tbp, Tbn); Tb9 = TaR * Tb8; rio[WS(vs, 3) + WS(rs, 7)] = FMA(Tbl, Tbp, Tbq); } } Tbk = Tba * Tb8; { E T86, T8h, T7P, T88, T87, T8i; { E T8n, T8j, T8m, T8l, T8o, T8k; iio[WS(vs, 7) + WS(rs, 7)] = FNMS(Tba, Tbj, Tb9); T86 = FNMS(KP707106781, T85, T7U); T8k = FMA(KP707106781, T85, T7U); rio[WS(vs, 7) + WS(rs, 7)] = FMA(TaR, Tbj, Tbk); T8h = FNMS(KP707106781, T8g, T8d); T8n = FMA(KP707106781, T8g, T8d); T8j = W[4]; T8m = W[5]; T7P = W[12]; T8l = T8j * T8k; T8o = T8m * T8k; T88 = W[13]; T87 = T7P * T86; iio[WS(vs, 3) + WS(rs, 5)] = FNMS(T8m, T8n, T8l); rio[WS(vs, 3) + WS(rs, 5)] = FMA(T8j, T8n, T8o); } T8i = T88 * T86; iio[WS(vs, 7) + WS(rs, 5)] = FNMS(T88, T8h, T87); { E T9P, T9T, T9S, T9R, T9U, T9Q; T9Q = FMA(KP707106781, T9B, T9q); T9C = FNMS(KP707106781, T9B, T9q); rio[WS(vs, 7) + WS(rs, 5)] = FMA(T7P, T8h, T8i); T9P = W[4]; T9N = FNMS(KP707106781, T9M, T9J); T9T = FMA(KP707106781, T9M, T9J); T9S = W[5]; T9R = T9P * T9Q; T9l = W[12]; T9U = T9S * T9Q; T9E = W[13]; iio[WS(vs, 3) + WS(rs, 6)] = FNMS(T9S, T9T, T9R); T9D = T9l * T9C; rio[WS(vs, 3) + WS(rs, 6)] = FMA(T9P, T9T, T9U); } } } } } } } } T9O = T9E * T9C; iio[WS(vs, 7) + WS(rs, 6)] = FNMS(T9E, T9N, T9D); rio[WS(vs, 7) + WS(rs, 6)] = FMA(T9l, T9N, T9O); } } } static const tw_instr twinstr[] = { {TW_FULL, 0, 8}, {TW_NEXT, 1, 0} }; static const ct_desc desc = { 8, "q1_8", twinstr, &GENUS, {352, 112, 176, 0}, 0, 0, 0 }; void X(codelet_q1_8) (planner *p) { X(kdft_difsq_register) (p, q1_8, &desc); } #else /* HAVE_FMA */ /* Generated by: ../../../genfft/gen_twidsq.native -compact -variables 4 -pipeline-latency 4 -reload-twiddle -dif -n 8 -name q1_8 -include q.h */ /* * This function contains 528 FP additions, 256 FP multiplications, * (or, 416 additions, 144 multiplications, 112 fused multiply/add), * 142 stack variables, 1 constants, and 256 memory accesses */ #include "q.h" static void q1_8(R *rio, R *iio, const R *W, stride rs, stride vs, INT mb, INT me, INT ms) { DK(KP707106781, +0.707106781186547524400844362104849039284835938); { INT m; for (m = mb, W = W + (mb * 14); m < me; m = m + 1, rio = rio + ms, iio = iio + ms, W = W + 14, MAKE_VOLATILE_STRIDE(16, rs), MAKE_VOLATILE_STRIDE(0, vs)) { E T7, T14, T1g, Tk, TC, TQ, T10, TM, T1w, T2p, T2z, T1H, T1M, T1W, T2j; E T1V, T7R, T8O, T90, T84, T8m, T8A, T8K, T8w, T9g, Ta9, Taj, T9r, T9w, T9G; E Ta3, T9F, Te, T17, T1h, Tp, Tu, TE, T11, TD, T1p, T2m, T2y, T1C, T1U; E T28, T2i, T24, T7Y, T8R, T91, T89, T8e, T8o, T8L, T8n, T99, Ta6, Tai, T9m; E T9E, T9S, Ta2, T9O, T2H, T3E, T3Q, T2U, T3c, T3q, T3A, T3m, T46, T4Z, T59; E T4h, T4m, T4w, T4T, T4v, T5h, T6e, T6q, T5u, T5M, T60, T6a, T5W, T6G, T7z; E T7J, T6R, T6W, T76, T7t, T75, T2O, T3H, T3R, T2Z, T34, T3e, T3B, T3d, T3Z; E T4W, T58, T4c, T4u, T4I, T4S, T4E, T5o, T6h, T6r, T5z, T5E, T5O, T6b, T5N; E T6z, T7w, T7I, T6M, T74, T7i, T7s, T7e; { E T3, Ty, Tj, TY, T6, Tg, TB, TZ; { E T1, T2, Th, Ti; T1 = rio[0]; T2 = rio[WS(rs, 4)]; T3 = T1 + T2; Ty = T1 - T2; Th = iio[0]; Ti = iio[WS(rs, 4)]; Tj = Th - Ti; TY = Th + Ti; } { E T4, T5, Tz, TA; T4 = rio[WS(rs, 2)]; T5 = rio[WS(rs, 6)]; T6 = T4 + T5; Tg = T4 - T5; Tz = iio[WS(rs, 2)]; TA = iio[WS(rs, 6)]; TB = Tz - TA; TZ = Tz + TA; } T7 = T3 + T6; T14 = T3 - T6; T1g = TY + TZ; Tk = Tg + Tj; TC = Ty - TB; TQ = Tj - Tg; T10 = TY - TZ; TM = Ty + TB; } { E T1s, T1I, T1L, T2n, T1v, T1D, T1G, T2o; { E T1q, T1r, T1J, T1K; T1q = rio[WS(vs, 1) + WS(rs, 1)]; T1r = rio[WS(vs, 1) + WS(rs, 5)]; T1s = T1q + T1r; T1I = T1q - T1r; T1J = iio[WS(vs, 1) + WS(rs, 1)]; T1K = iio[WS(vs, 1) + WS(rs, 5)]; T1L = T1J - T1K; T2n = T1J + T1K; } { E T1t, T1u, T1E, T1F; T1t = rio[WS(vs, 1) + WS(rs, 7)]; T1u = rio[WS(vs, 1) + WS(rs, 3)]; T1v = T1t + T1u; T1D = T1t - T1u; T1E = iio[WS(vs, 1) + WS(rs, 7)]; T1F = iio[WS(vs, 1) + WS(rs, 3)]; T1G = T1E - T1F; T2o = T1E + T1F; } T1w = T1s + T1v; T2p = T2n - T2o; T2z = T2n + T2o; T1H = T1D - T1G; T1M = T1I + T1L; T1W = T1D + T1G; T2j = T1v - T1s; T1V = T1L - T1I; } { E T7N, T8i, T83, T8I, T7Q, T80, T8l, T8J; { E T7L, T7M, T81, T82; T7L = rio[WS(vs, 6)]; T7M = rio[WS(vs, 6) + WS(rs, 4)]; T7N = T7L + T7M; T8i = T7L - T7M; T81 = iio[WS(vs, 6)]; T82 = iio[WS(vs, 6) + WS(rs, 4)]; T83 = T81 - T82; T8I = T81 + T82; } { E T7O, T7P, T8j, T8k; T7O = rio[WS(vs, 6) + WS(rs, 2)]; T7P = rio[WS(vs, 6) + WS(rs, 6)]; T7Q = T7O + T7P; T80 = T7O - T7P; T8j = iio[WS(vs, 6) + WS(rs, 2)]; T8k = iio[WS(vs, 6) + WS(rs, 6)]; T8l = T8j - T8k; T8J = T8j + T8k; } T7R = T7N + T7Q; T8O = T7N - T7Q; T90 = T8I + T8J; T84 = T80 + T83; T8m = T8i - T8l; T8A = T83 - T80; T8K = T8I - T8J; T8w = T8i + T8l; } { E T9c, T9s, T9v, Ta7, T9f, T9n, T9q, Ta8; { E T9a, T9b, T9t, T9u; T9a = rio[WS(vs, 7) + WS(rs, 1)]; T9b = rio[WS(vs, 7) + WS(rs, 5)]; T9c = T9a + T9b; T9s = T9a - T9b; T9t = iio[WS(vs, 7) + WS(rs, 1)]; T9u = iio[WS(vs, 7) + WS(rs, 5)]; T9v = T9t - T9u; Ta7 = T9t + T9u; } { E T9d, T9e, T9o, T9p; T9d = rio[WS(vs, 7) + WS(rs, 7)]; T9e = rio[WS(vs, 7) + WS(rs, 3)]; T9f = T9d + T9e; T9n = T9d - T9e; T9o = iio[WS(vs, 7) + WS(rs, 7)]; T9p = iio[WS(vs, 7) + WS(rs, 3)]; T9q = T9o - T9p; Ta8 = T9o + T9p; } T9g = T9c + T9f; Ta9 = Ta7 - Ta8; Taj = Ta7 + Ta8; T9r = T9n - T9q; T9w = T9s + T9v; T9G = T9n + T9q; Ta3 = T9f - T9c; T9F = T9v - T9s; } { E Ta, Tq, Tt, T15, Td, Tl, To, T16; { E T8, T9, Tr, Ts; T8 = rio[WS(rs, 1)]; T9 = rio[WS(rs, 5)]; Ta = T8 + T9; Tq = T8 - T9; Tr = iio[WS(rs, 1)]; Ts = iio[WS(rs, 5)]; Tt = Tr - Ts; T15 = Tr + Ts; } { E Tb, Tc, Tm, Tn; Tb = rio[WS(rs, 7)]; Tc = rio[WS(rs, 3)]; Td = Tb + Tc; Tl = Tb - Tc; Tm = iio[WS(rs, 7)]; Tn = iio[WS(rs, 3)]; To = Tm - Tn; T16 = Tm + Tn; } Te = Ta + Td; T17 = T15 - T16; T1h = T15 + T16; Tp = Tl - To; Tu = Tq + Tt; TE = Tl + To; T11 = Td - Ta; TD = Tt - Tq; } { E T1l, T1Q, T1B, T2g, T1o, T1y, T1T, T2h; { E T1j, T1k, T1z, T1A; T1j = rio[WS(vs, 1)]; T1k = rio[WS(vs, 1) + WS(rs, 4)]; T1l = T1j + T1k; T1Q = T1j - T1k; T1z = iio[WS(vs, 1)]; T1A = iio[WS(vs, 1) + WS(rs, 4)]; T1B = T1z - T1A; T2g = T1z + T1A; } { E T1m, T1n, T1R, T1S; T1m = rio[WS(vs, 1) + WS(rs, 2)]; T1n = rio[WS(vs, 1) + WS(rs, 6)]; T1o = T1m + T1n; T1y = T1m - T1n; T1R = iio[WS(vs, 1) + WS(rs, 2)]; T1S = iio[WS(vs, 1) + WS(rs, 6)]; T1T = T1R - T1S; T2h = T1R + T1S; } T1p = T1l + T1o; T2m = T1l - T1o; T2y = T2g + T2h; T1C = T1y + T1B; T1U = T1Q - T1T; T28 = T1B - T1y; T2i = T2g - T2h; T24 = T1Q + T1T; } { E T7U, T8a, T8d, T8P, T7X, T85, T88, T8Q; { E T7S, T7T, T8b, T8c; T7S = rio[WS(vs, 6) + WS(rs, 1)]; T7T = rio[WS(vs, 6) + WS(rs, 5)]; T7U = T7S + T7T; T8a = T7S - T7T; T8b = iio[WS(vs, 6) + WS(rs, 1)]; T8c = iio[WS(vs, 6) + WS(rs, 5)]; T8d = T8b - T8c; T8P = T8b + T8c; } { E T7V, T7W, T86, T87; T7V = rio[WS(vs, 6) + WS(rs, 7)]; T7W = rio[WS(vs, 6) + WS(rs, 3)]; T7X = T7V + T7W; T85 = T7V - T7W; T86 = iio[WS(vs, 6) + WS(rs, 7)]; T87 = iio[WS(vs, 6) + WS(rs, 3)]; T88 = T86 - T87; T8Q = T86 + T87; } T7Y = T7U + T7X; T8R = T8P - T8Q; T91 = T8P + T8Q; T89 = T85 - T88; T8e = T8a + T8d; T8o = T85 + T88; T8L = T7X - T7U; T8n = T8d - T8a; } { E T95, T9A, T9l, Ta0, T98, T9i, T9D, Ta1; { E T93, T94, T9j, T9k; T93 = rio[WS(vs, 7)]; T94 = rio[WS(vs, 7) + WS(rs, 4)]; T95 = T93 + T94; T9A = T93 - T94; T9j = iio[WS(vs, 7)]; T9k = iio[WS(vs, 7) + WS(rs, 4)]; T9l = T9j - T9k; Ta0 = T9j + T9k; } { E T96, T97, T9B, T9C; T96 = rio[WS(vs, 7) + WS(rs, 2)]; T97 = rio[WS(vs, 7) + WS(rs, 6)]; T98 = T96 + T97; T9i = T96 - T97; T9B = iio[WS(vs, 7) + WS(rs, 2)]; T9C = iio[WS(vs, 7) + WS(rs, 6)]; T9D = T9B - T9C; Ta1 = T9B + T9C; } T99 = T95 + T98; Ta6 = T95 - T98; Tai = Ta0 + Ta1; T9m = T9i + T9l; T9E = T9A - T9D; T9S = T9l - T9i; Ta2 = Ta0 - Ta1; T9O = T9A + T9D; } { E T2D, T38, T2T, T3y, T2G, T2Q, T3b, T3z; { E T2B, T2C, T2R, T2S; T2B = rio[WS(vs, 2)]; T2C = rio[WS(vs, 2) + WS(rs, 4)]; T2D = T2B + T2C; T38 = T2B - T2C; T2R = iio[WS(vs, 2)]; T2S = iio[WS(vs, 2) + WS(rs, 4)]; T2T = T2R - T2S; T3y = T2R + T2S; } { E T2E, T2F, T39, T3a; T2E = rio[WS(vs, 2) + WS(rs, 2)]; T2F = rio[WS(vs, 2) + WS(rs, 6)]; T2G = T2E + T2F; T2Q = T2E - T2F; T39 = iio[WS(vs, 2) + WS(rs, 2)]; T3a = iio[WS(vs, 2) + WS(rs, 6)]; T3b = T39 - T3a; T3z = T39 + T3a; } T2H = T2D + T2G; T3E = T2D - T2G; T3Q = T3y + T3z; T2U = T2Q + T2T; T3c = T38 - T3b; T3q = T2T - T2Q; T3A = T3y - T3z; T3m = T38 + T3b; } { E T42, T4i, T4l, T4X, T45, T4d, T4g, T4Y; { E T40, T41, T4j, T4k; T40 = rio[WS(vs, 3) + WS(rs, 1)]; T41 = rio[WS(vs, 3) + WS(rs, 5)]; T42 = T40 + T41; T4i = T40 - T41; T4j = iio[WS(vs, 3) + WS(rs, 1)]; T4k = iio[WS(vs, 3) + WS(rs, 5)]; T4l = T4j - T4k; T4X = T4j + T4k; } { E T43, T44, T4e, T4f; T43 = rio[WS(vs, 3) + WS(rs, 7)]; T44 = rio[WS(vs, 3) + WS(rs, 3)]; T45 = T43 + T44; T4d = T43 - T44; T4e = iio[WS(vs, 3) + WS(rs, 7)]; T4f = iio[WS(vs, 3) + WS(rs, 3)]; T4g = T4e - T4f; T4Y = T4e + T4f; } T46 = T42 + T45; T4Z = T4X - T4Y; T59 = T4X + T4Y; T4h = T4d - T4g; T4m = T4i + T4l; T4w = T4d + T4g; T4T = T45 - T42; T4v = T4l - T4i; } { E T5d, T5I, T5t, T68, T5g, T5q, T5L, T69; { E T5b, T5c, T5r, T5s; T5b = rio[WS(vs, 4)]; T5c = rio[WS(vs, 4) + WS(rs, 4)]; T5d = T5b + T5c; T5I = T5b - T5c; T5r = iio[WS(vs, 4)]; T5s = iio[WS(vs, 4) + WS(rs, 4)]; T5t = T5r - T5s; T68 = T5r + T5s; } { E T5e, T5f, T5J, T5K; T5e = rio[WS(vs, 4) + WS(rs, 2)]; T5f = rio[WS(vs, 4) + WS(rs, 6)]; T5g = T5e + T5f; T5q = T5e - T5f; T5J = iio[WS(vs, 4) + WS(rs, 2)]; T5K = iio[WS(vs, 4) + WS(rs, 6)]; T5L = T5J - T5K; T69 = T5J + T5K; } T5h = T5d + T5g; T6e = T5d - T5g; T6q = T68 + T69; T5u = T5q + T5t; T5M = T5I - T5L; T60 = T5t - T5q; T6a = T68 - T69; T5W = T5I + T5L; } { E T6C, T6S, T6V, T7x, T6F, T6N, T6Q, T7y; { E T6A, T6B, T6T, T6U; T6A = rio[WS(vs, 5) + WS(rs, 1)]; T6B = rio[WS(vs, 5) + WS(rs, 5)]; T6C = T6A + T6B; T6S = T6A - T6B; T6T = iio[WS(vs, 5) + WS(rs, 1)]; T6U = iio[WS(vs, 5) + WS(rs, 5)]; T6V = T6T - T6U; T7x = T6T + T6U; } { E T6D, T6E, T6O, T6P; T6D = rio[WS(vs, 5) + WS(rs, 7)]; T6E = rio[WS(vs, 5) + WS(rs, 3)]; T6F = T6D + T6E; T6N = T6D - T6E; T6O = iio[WS(vs, 5) + WS(rs, 7)]; T6P = iio[WS(vs, 5) + WS(rs, 3)]; T6Q = T6O - T6P; T7y = T6O + T6P; } T6G = T6C + T6F; T7z = T7x - T7y; T7J = T7x + T7y; T6R = T6N - T6Q; T6W = T6S + T6V; T76 = T6N + T6Q; T7t = T6F - T6C; T75 = T6V - T6S; } { E T2K, T30, T33, T3F, T2N, T2V, T2Y, T3G; { E T2I, T2J, T31, T32; T2I = rio[WS(vs, 2) + WS(rs, 1)]; T2J = rio[WS(vs, 2) + WS(rs, 5)]; T2K = T2I + T2J; T30 = T2I - T2J; T31 = iio[WS(vs, 2) + WS(rs, 1)]; T32 = iio[WS(vs, 2) + WS(rs, 5)]; T33 = T31 - T32; T3F = T31 + T32; } { E T2L, T2M, T2W, T2X; T2L = rio[WS(vs, 2) + WS(rs, 7)]; T2M = rio[WS(vs, 2) + WS(rs, 3)]; T2N = T2L + T2M; T2V = T2L - T2M; T2W = iio[WS(vs, 2) + WS(rs, 7)]; T2X = iio[WS(vs, 2) + WS(rs, 3)]; T2Y = T2W - T2X; T3G = T2W + T2X; } T2O = T2K + T2N; T3H = T3F - T3G; T3R = T3F + T3G; T2Z = T2V - T2Y; T34 = T30 + T33; T3e = T2V + T2Y; T3B = T2N - T2K; T3d = T33 - T30; } { E T3V, T4q, T4b, T4Q, T3Y, T48, T4t, T4R; { E T3T, T3U, T49, T4a; T3T = rio[WS(vs, 3)]; T3U = rio[WS(vs, 3) + WS(rs, 4)]; T3V = T3T + T3U; T4q = T3T - T3U; T49 = iio[WS(vs, 3)]; T4a = iio[WS(vs, 3) + WS(rs, 4)]; T4b = T49 - T4a; T4Q = T49 + T4a; } { E T3W, T3X, T4r, T4s; T3W = rio[WS(vs, 3) + WS(rs, 2)]; T3X = rio[WS(vs, 3) + WS(rs, 6)]; T3Y = T3W + T3X; T48 = T3W - T3X; T4r = iio[WS(vs, 3) + WS(rs, 2)]; T4s = iio[WS(vs, 3) + WS(rs, 6)]; T4t = T4r - T4s; T4R = T4r + T4s; } T3Z = T3V + T3Y; T4W = T3V - T3Y; T58 = T4Q + T4R; T4c = T48 + T4b; T4u = T4q - T4t; T4I = T4b - T48; T4S = T4Q - T4R; T4E = T4q + T4t; } { E T5k, T5A, T5D, T6f, T5n, T5v, T5y, T6g; { E T5i, T5j, T5B, T5C; T5i = rio[WS(vs, 4) + WS(rs, 1)]; T5j = rio[WS(vs, 4) + WS(rs, 5)]; T5k = T5i + T5j; T5A = T5i - T5j; T5B = iio[WS(vs, 4) + WS(rs, 1)]; T5C = iio[WS(vs, 4) + WS(rs, 5)]; T5D = T5B - T5C; T6f = T5B + T5C; } { E T5l, T5m, T5w, T5x; T5l = rio[WS(vs, 4) + WS(rs, 7)]; T5m = rio[WS(vs, 4) + WS(rs, 3)]; T5n = T5l + T5m; T5v = T5l - T5m; T5w = iio[WS(vs, 4) + WS(rs, 7)]; T5x = iio[WS(vs, 4) + WS(rs, 3)]; T5y = T5w - T5x; T6g = T5w + T5x; } T5o = T5k + T5n; T6h = T6f - T6g; T6r = T6f + T6g; T5z = T5v - T5y; T5E = T5A + T5D; T5O = T5v + T5y; T6b = T5n - T5k; T5N = T5D - T5A; } { E T6v, T70, T6L, T7q, T6y, T6I, T73, T7r; { E T6t, T6u, T6J, T6K; T6t = rio[WS(vs, 5)]; T6u = rio[WS(vs, 5) + WS(rs, 4)]; T6v = T6t + T6u; T70 = T6t - T6u; T6J = iio[WS(vs, 5)]; T6K = iio[WS(vs, 5) + WS(rs, 4)]; T6L = T6J - T6K; T7q = T6J + T6K; } { E T6w, T6x, T71, T72; T6w = rio[WS(vs, 5) + WS(rs, 2)]; T6x = rio[WS(vs, 5) + WS(rs, 6)]; T6y = T6w + T6x; T6I = T6w - T6x; T71 = iio[WS(vs, 5) + WS(rs, 2)]; T72 = iio[WS(vs, 5) + WS(rs, 6)]; T73 = T71 - T72; T7r = T71 + T72; } T6z = T6v + T6y; T7w = T6v - T6y; T7I = T7q + T7r; T6M = T6I + T6L; T74 = T70 - T73; T7i = T6L - T6I; T7s = T7q - T7r; T7e = T70 + T73; } rio[0] = T7 + Te; iio[0] = T1g + T1h; rio[WS(rs, 1)] = T1p + T1w; iio[WS(rs, 1)] = T2y + T2z; rio[WS(rs, 3)] = T3Z + T46; rio[WS(rs, 2)] = T2H + T2O; iio[WS(rs, 2)] = T3Q + T3R; iio[WS(rs, 3)] = T58 + T59; rio[WS(rs, 6)] = T7R + T7Y; iio[WS(rs, 6)] = T90 + T91; iio[WS(rs, 5)] = T7I + T7J; rio[WS(rs, 5)] = T6z + T6G; iio[WS(rs, 4)] = T6q + T6r; rio[WS(rs, 4)] = T5h + T5o; rio[WS(rs, 7)] = T99 + T9g; iio[WS(rs, 7)] = Tai + Taj; { E T12, T18, TX, T13; T12 = T10 - T11; T18 = T14 - T17; TX = W[10]; T13 = W[11]; iio[WS(vs, 6)] = FNMS(T13, T18, TX * T12); rio[WS(vs, 6)] = FMA(T13, T12, TX * T18); } { E Tag, Tak, Taf, Tah; Tag = T99 - T9g; Tak = Tai - Taj; Taf = W[6]; Tah = W[7]; rio[WS(vs, 4) + WS(rs, 7)] = FMA(Taf, Tag, Tah * Tak); iio[WS(vs, 4) + WS(rs, 7)] = FNMS(Tah, Tag, Taf * Tak); } { E T8M, T8S, T8H, T8N; T8M = T8K - T8L; T8S = T8O - T8R; T8H = W[10]; T8N = W[11]; iio[WS(vs, 6) + WS(rs, 6)] = FNMS(T8N, T8S, T8H * T8M); rio[WS(vs, 6) + WS(rs, 6)] = FMA(T8N, T8M, T8H * T8S); } { E T2k, T2q, T2f, T2l; T2k = T2i - T2j; T2q = T2m - T2p; T2f = W[10]; T2l = W[11]; iio[WS(vs, 6) + WS(rs, 1)] = FNMS(T2l, T2q, T2f * T2k); rio[WS(vs, 6) + WS(rs, 1)] = FMA(T2l, T2k, T2f * T2q); } { E Ta4, Taa, T9Z, Ta5; Ta4 = Ta2 - Ta3; Taa = Ta6 - Ta9; T9Z = W[10]; Ta5 = W[11]; iio[WS(vs, 6) + WS(rs, 7)] = FNMS(Ta5, Taa, T9Z * Ta4); rio[WS(vs, 6) + WS(rs, 7)] = FMA(Ta5, Ta4, T9Z * Taa); } { E T8Y, T92, T8X, T8Z; T8Y = T7R - T7Y; T92 = T90 - T91; T8X = W[6]; T8Z = W[7]; rio[WS(vs, 4) + WS(rs, 6)] = FMA(T8X, T8Y, T8Z * T92); iio[WS(vs, 4) + WS(rs, 6)] = FNMS(T8Z, T8Y, T8X * T92); } { E T2w, T2A, T2v, T2x; T2w = T1p - T1w; T2A = T2y - T2z; T2v = W[6]; T2x = W[7]; rio[WS(vs, 4) + WS(rs, 1)] = FMA(T2v, T2w, T2x * T2A); iio[WS(vs, 4) + WS(rs, 1)] = FNMS(T2x, T2w, T2v * T2A); } { E Tac, Tae, Tab, Tad; Tac = Ta3 + Ta2; Tae = Ta6 + Ta9; Tab = W[2]; Tad = W[3]; iio[WS(vs, 2) + WS(rs, 7)] = FNMS(Tad, Tae, Tab * Tac); rio[WS(vs, 2) + WS(rs, 7)] = FMA(Tad, Tac, Tab * Tae); } { E T8U, T8W, T8T, T8V; T8U = T8L + T8K; T8W = T8O + T8R; T8T = W[2]; T8V = W[3]; iio[WS(vs, 2) + WS(rs, 6)] = FNMS(T8V, T8W, T8T * T8U); rio[WS(vs, 2) + WS(rs, 6)] = FMA(T8V, T8U, T8T * T8W); } { E T1a, T1c, T19, T1b; T1a = T11 + T10; T1c = T14 + T17; T19 = W[2]; T1b = W[3]; iio[WS(vs, 2)] = FNMS(T1b, T1c, T19 * T1a); rio[WS(vs, 2)] = FMA(T1b, T1a, T19 * T1c); } { E T1e, T1i, T1d, T1f; T1e = T7 - Te; T1i = T1g - T1h; T1d = W[6]; T1f = W[7]; rio[WS(vs, 4)] = FMA(T1d, T1e, T1f * T1i); iio[WS(vs, 4)] = FNMS(T1f, T1e, T1d * T1i); } { E T2s, T2u, T2r, T2t; T2s = T2j + T2i; T2u = T2m + T2p; T2r = W[2]; T2t = W[3]; iio[WS(vs, 2) + WS(rs, 1)] = FNMS(T2t, T2u, T2r * T2s); rio[WS(vs, 2) + WS(rs, 1)] = FMA(T2t, T2s, T2r * T2u); } { E T3C, T3I, T3x, T3D; T3C = T3A - T3B; T3I = T3E - T3H; T3x = W[10]; T3D = W[11]; iio[WS(vs, 6) + WS(rs, 2)] = FNMS(T3D, T3I, T3x * T3C); rio[WS(vs, 6) + WS(rs, 2)] = FMA(T3D, T3C, T3x * T3I); } { E T4U, T50, T4P, T4V; T4U = T4S - T4T; T50 = T4W - T4Z; T4P = W[10]; T4V = W[11]; iio[WS(vs, 6) + WS(rs, 3)] = FNMS(T4V, T50, T4P * T4U); rio[WS(vs, 6) + WS(rs, 3)] = FMA(T4V, T4U, T4P * T50); } { E T56, T5a, T55, T57; T56 = T3Z - T46; T5a = T58 - T59; T55 = W[6]; T57 = W[7]; rio[WS(vs, 4) + WS(rs, 3)] = FMA(T55, T56, T57 * T5a); iio[WS(vs, 4) + WS(rs, 3)] = FNMS(T57, T56, T55 * T5a); } { E T6o, T6s, T6n, T6p; T6o = T5h - T5o; T6s = T6q - T6r; T6n = W[6]; T6p = W[7]; rio[WS(vs, 4) + WS(rs, 4)] = FMA(T6n, T6o, T6p * T6s); iio[WS(vs, 4) + WS(rs, 4)] = FNMS(T6p, T6o, T6n * T6s); } { E T7u, T7A, T7p, T7v; T7u = T7s - T7t; T7A = T7w - T7z; T7p = W[10]; T7v = W[11]; iio[WS(vs, 6) + WS(rs, 5)] = FNMS(T7v, T7A, T7p * T7u); rio[WS(vs, 6) + WS(rs, 5)] = FMA(T7v, T7u, T7p * T7A); } { E T6c, T6i, T67, T6d; T6c = T6a - T6b; T6i = T6e - T6h; T67 = W[10]; T6d = W[11]; iio[WS(vs, 6) + WS(rs, 4)] = FNMS(T6d, T6i, T67 * T6c); rio[WS(vs, 6) + WS(rs, 4)] = FMA(T6d, T6c, T67 * T6i); } { E T7G, T7K, T7F, T7H; T7G = T6z - T6G; T7K = T7I - T7J; T7F = W[6]; T7H = W[7]; rio[WS(vs, 4) + WS(rs, 5)] = FMA(T7F, T7G, T7H * T7K); iio[WS(vs, 4) + WS(rs, 5)] = FNMS(T7H, T7G, T7F * T7K); } { E T3O, T3S, T3N, T3P; T3O = T2H - T2O; T3S = T3Q - T3R; T3N = W[6]; T3P = W[7]; rio[WS(vs, 4) + WS(rs, 2)] = FMA(T3N, T3O, T3P * T3S); iio[WS(vs, 4) + WS(rs, 2)] = FNMS(T3P, T3O, T3N * T3S); } { E T3K, T3M, T3J, T3L; T3K = T3B + T3A; T3M = T3E + T3H; T3J = W[2]; T3L = W[3]; iio[WS(vs, 2) + WS(rs, 2)] = FNMS(T3L, T3M, T3J * T3K); rio[WS(vs, 2) + WS(rs, 2)] = FMA(T3L, T3K, T3J * T3M); } { E T7C, T7E, T7B, T7D; T7C = T7t + T7s; T7E = T7w + T7z; T7B = W[2]; T7D = W[3]; iio[WS(vs, 2) + WS(rs, 5)] = FNMS(T7D, T7E, T7B * T7C); rio[WS(vs, 2) + WS(rs, 5)] = FMA(T7D, T7C, T7B * T7E); } { E T6k, T6m, T6j, T6l; T6k = T6b + T6a; T6m = T6e + T6h; T6j = W[2]; T6l = W[3]; iio[WS(vs, 2) + WS(rs, 4)] = FNMS(T6l, T6m, T6j * T6k); rio[WS(vs, 2) + WS(rs, 4)] = FMA(T6l, T6k, T6j * T6m); } { E T52, T54, T51, T53; T52 = T4T + T4S; T54 = T4W + T4Z; T51 = W[2]; T53 = W[3]; iio[WS(vs, 2) + WS(rs, 3)] = FNMS(T53, T54, T51 * T52); rio[WS(vs, 2) + WS(rs, 3)] = FMA(T53, T52, T51 * T54); } { E T5G, T5S, T5Q, T5U, T5F, T5P; T5F = KP707106781 * (T5z - T5E); T5G = T5u - T5F; T5S = T5u + T5F; T5P = KP707106781 * (T5N - T5O); T5Q = T5M - T5P; T5U = T5M + T5P; { E T5p, T5H, T5R, T5T; T5p = W[12]; T5H = W[13]; iio[WS(vs, 7) + WS(rs, 4)] = FNMS(T5H, T5Q, T5p * T5G); rio[WS(vs, 7) + WS(rs, 4)] = FMA(T5H, T5G, T5p * T5Q); T5R = W[4]; T5T = W[5]; iio[WS(vs, 3) + WS(rs, 4)] = FNMS(T5T, T5U, T5R * T5S); rio[WS(vs, 3) + WS(rs, 4)] = FMA(T5T, T5S, T5R * T5U); } } { E Tw, TI, TG, TK, Tv, TF; Tv = KP707106781 * (Tp - Tu); Tw = Tk - Tv; TI = Tk + Tv; TF = KP707106781 * (TD - TE); TG = TC - TF; TK = TC + TF; { E Tf, Tx, TH, TJ; Tf = W[12]; Tx = W[13]; iio[WS(vs, 7)] = FNMS(Tx, TG, Tf * Tw); rio[WS(vs, 7)] = FMA(Tx, Tw, Tf * TG); TH = W[4]; TJ = W[5]; iio[WS(vs, 3)] = FNMS(TJ, TK, TH * TI); rio[WS(vs, 3)] = FMA(TJ, TI, TH * TK); } } { E T9Q, T9W, T9U, T9Y, T9P, T9T; T9P = KP707106781 * (T9w + T9r); T9Q = T9O - T9P; T9W = T9O + T9P; T9T = KP707106781 * (T9F + T9G); T9U = T9S - T9T; T9Y = T9S + T9T; { E T9N, T9R, T9V, T9X; T9N = W[8]; T9R = W[9]; rio[WS(vs, 5) + WS(rs, 7)] = FMA(T9N, T9Q, T9R * T9U); iio[WS(vs, 5) + WS(rs, 7)] = FNMS(T9R, T9Q, T9N * T9U); T9V = W[0]; T9X = W[1]; rio[WS(vs, 1) + WS(rs, 7)] = FMA(T9V, T9W, T9X * T9Y); iio[WS(vs, 1) + WS(rs, 7)] = FNMS(T9X, T9W, T9V * T9Y); } } { E T36, T3i, T3g, T3k, T35, T3f; T35 = KP707106781 * (T2Z - T34); T36 = T2U - T35; T3i = T2U + T35; T3f = KP707106781 * (T3d - T3e); T3g = T3c - T3f; T3k = T3c + T3f; { E T2P, T37, T3h, T3j; T2P = W[12]; T37 = W[13]; iio[WS(vs, 7) + WS(rs, 2)] = FNMS(T37, T3g, T2P * T36); rio[WS(vs, 7) + WS(rs, 2)] = FMA(T37, T36, T2P * T3g); T3h = W[4]; T3j = W[5]; iio[WS(vs, 3) + WS(rs, 2)] = FNMS(T3j, T3k, T3h * T3i); rio[WS(vs, 3) + WS(rs, 2)] = FMA(T3j, T3i, T3h * T3k); } } { E T5Y, T64, T62, T66, T5X, T61; T5X = KP707106781 * (T5E + T5z); T5Y = T5W - T5X; T64 = T5W + T5X; T61 = KP707106781 * (T5N + T5O); T62 = T60 - T61; T66 = T60 + T61; { E T5V, T5Z, T63, T65; T5V = W[8]; T5Z = W[9]; rio[WS(vs, 5) + WS(rs, 4)] = FMA(T5V, T5Y, T5Z * T62); iio[WS(vs, 5) + WS(rs, 4)] = FNMS(T5Z, T5Y, T5V * T62); T63 = W[0]; T65 = W[1]; rio[WS(vs, 1) + WS(rs, 4)] = FMA(T63, T64, T65 * T66); iio[WS(vs, 1) + WS(rs, 4)] = FNMS(T65, T64, T63 * T66); } } { E T7g, T7m, T7k, T7o, T7f, T7j; T7f = KP707106781 * (T6W + T6R); T7g = T7e - T7f; T7m = T7e + T7f; T7j = KP707106781 * (T75 + T76); T7k = T7i - T7j; T7o = T7i + T7j; { E T7d, T7h, T7l, T7n; T7d = W[8]; T7h = W[9]; rio[WS(vs, 5) + WS(rs, 5)] = FMA(T7d, T7g, T7h * T7k); iio[WS(vs, 5) + WS(rs, 5)] = FNMS(T7h, T7g, T7d * T7k); T7l = W[0]; T7n = W[1]; rio[WS(vs, 1) + WS(rs, 5)] = FMA(T7l, T7m, T7n * T7o); iio[WS(vs, 1) + WS(rs, 5)] = FNMS(T7n, T7m, T7l * T7o); } } { E T8g, T8s, T8q, T8u, T8f, T8p; T8f = KP707106781 * (T89 - T8e); T8g = T84 - T8f; T8s = T84 + T8f; T8p = KP707106781 * (T8n - T8o); T8q = T8m - T8p; T8u = T8m + T8p; { E T7Z, T8h, T8r, T8t; T7Z = W[12]; T8h = W[13]; iio[WS(vs, 7) + WS(rs, 6)] = FNMS(T8h, T8q, T7Z * T8g); rio[WS(vs, 7) + WS(rs, 6)] = FMA(T8h, T8g, T7Z * T8q); T8r = W[4]; T8t = W[5]; iio[WS(vs, 3) + WS(rs, 6)] = FNMS(T8t, T8u, T8r * T8s); rio[WS(vs, 3) + WS(rs, 6)] = FMA(T8t, T8s, T8r * T8u); } } { E T4G, T4M, T4K, T4O, T4F, T4J; T4F = KP707106781 * (T4m + T4h); T4G = T4E - T4F; T4M = T4E + T4F; T4J = KP707106781 * (T4v + T4w); T4K = T4I - T4J; T4O = T4I + T4J; { E T4D, T4H, T4L, T4N; T4D = W[8]; T4H = W[9]; rio[WS(vs, 5) + WS(rs, 3)] = FMA(T4D, T4G, T4H * T4K); iio[WS(vs, 5) + WS(rs, 3)] = FNMS(T4H, T4G, T4D * T4K); T4L = W[0]; T4N = W[1]; rio[WS(vs, 1) + WS(rs, 3)] = FMA(T4L, T4M, T4N * T4O); iio[WS(vs, 1) + WS(rs, 3)] = FNMS(T4N, T4M, T4L * T4O); } } { E TO, TU, TS, TW, TN, TR; TN = KP707106781 * (Tu + Tp); TO = TM - TN; TU = TM + TN; TR = KP707106781 * (TD + TE); TS = TQ - TR; TW = TQ + TR; { E TL, TP, TT, TV; TL = W[8]; TP = W[9]; rio[WS(vs, 5)] = FMA(TL, TO, TP * TS); iio[WS(vs, 5)] = FNMS(TP, TO, TL * TS); TT = W[0]; TV = W[1]; rio[WS(vs, 1)] = FMA(TT, TU, TV * TW); iio[WS(vs, 1)] = FNMS(TV, TU, TT * TW); } } { E T26, T2c, T2a, T2e, T25, T29; T25 = KP707106781 * (T1M + T1H); T26 = T24 - T25; T2c = T24 + T25; T29 = KP707106781 * (T1V + T1W); T2a = T28 - T29; T2e = T28 + T29; { E T23, T27, T2b, T2d; T23 = W[8]; T27 = W[9]; rio[WS(vs, 5) + WS(rs, 1)] = FMA(T23, T26, T27 * T2a); iio[WS(vs, 5) + WS(rs, 1)] = FNMS(T27, T26, T23 * T2a); T2b = W[0]; T2d = W[1]; rio[WS(vs, 1) + WS(rs, 1)] = FMA(T2b, T2c, T2d * T2e); iio[WS(vs, 1) + WS(rs, 1)] = FNMS(T2d, T2c, T2b * T2e); } } { E T9y, T9K, T9I, T9M, T9x, T9H; T9x = KP707106781 * (T9r - T9w); T9y = T9m - T9x; T9K = T9m + T9x; T9H = KP707106781 * (T9F - T9G); T9I = T9E - T9H; T9M = T9E + T9H; { E T9h, T9z, T9J, T9L; T9h = W[12]; T9z = W[13]; iio[WS(vs, 7) + WS(rs, 7)] = FNMS(T9z, T9I, T9h * T9y); rio[WS(vs, 7) + WS(rs, 7)] = FMA(T9z, T9y, T9h * T9I); T9J = W[4]; T9L = W[5]; iio[WS(vs, 3) + WS(rs, 7)] = FNMS(T9L, T9M, T9J * T9K); rio[WS(vs, 3) + WS(rs, 7)] = FMA(T9L, T9K, T9J * T9M); } } { E T6Y, T7a, T78, T7c, T6X, T77; T6X = KP707106781 * (T6R - T6W); T6Y = T6M - T6X; T7a = T6M + T6X; T77 = KP707106781 * (T75 - T76); T78 = T74 - T77; T7c = T74 + T77; { E T6H, T6Z, T79, T7b; T6H = W[12]; T6Z = W[13]; iio[WS(vs, 7) + WS(rs, 5)] = FNMS(T6Z, T78, T6H * T6Y); rio[WS(vs, 7) + WS(rs, 5)] = FMA(T6Z, T6Y, T6H * T78); T79 = W[4]; T7b = W[5]; iio[WS(vs, 3) + WS(rs, 5)] = FNMS(T7b, T7c, T79 * T7a); rio[WS(vs, 3) + WS(rs, 5)] = FMA(T7b, T7a, T79 * T7c); } } { E T1O, T20, T1Y, T22, T1N, T1X; T1N = KP707106781 * (T1H - T1M); T1O = T1C - T1N; T20 = T1C + T1N; T1X = KP707106781 * (T1V - T1W); T1Y = T1U - T1X; T22 = T1U + T1X; { E T1x, T1P, T1Z, T21; T1x = W[12]; T1P = W[13]; iio[WS(vs, 7) + WS(rs, 1)] = FNMS(T1P, T1Y, T1x * T1O); rio[WS(vs, 7) + WS(rs, 1)] = FMA(T1P, T1O, T1x * T1Y); T1Z = W[4]; T21 = W[5]; iio[WS(vs, 3) + WS(rs, 1)] = FNMS(T21, T22, T1Z * T20); rio[WS(vs, 3) + WS(rs, 1)] = FMA(T21, T20, T1Z * T22); } } { E T4o, T4A, T4y, T4C, T4n, T4x; T4n = KP707106781 * (T4h - T4m); T4o = T4c - T4n; T4A = T4c + T4n; T4x = KP707106781 * (T4v - T4w); T4y = T4u - T4x; T4C = T4u + T4x; { E T47, T4p, T4z, T4B; T47 = W[12]; T4p = W[13]; iio[WS(vs, 7) + WS(rs, 3)] = FNMS(T4p, T4y, T47 * T4o); rio[WS(vs, 7) + WS(rs, 3)] = FMA(T4p, T4o, T47 * T4y); T4z = W[4]; T4B = W[5]; iio[WS(vs, 3) + WS(rs, 3)] = FNMS(T4B, T4C, T4z * T4A); rio[WS(vs, 3) + WS(rs, 3)] = FMA(T4B, T4A, T4z * T4C); } } { E T3o, T3u, T3s, T3w, T3n, T3r; T3n = KP707106781 * (T34 + T2Z); T3o = T3m - T3n; T3u = T3m + T3n; T3r = KP707106781 * (T3d + T3e); T3s = T3q - T3r; T3w = T3q + T3r; { E T3l, T3p, T3t, T3v; T3l = W[8]; T3p = W[9]; rio[WS(vs, 5) + WS(rs, 2)] = FMA(T3l, T3o, T3p * T3s); iio[WS(vs, 5) + WS(rs, 2)] = FNMS(T3p, T3o, T3l * T3s); T3t = W[0]; T3v = W[1]; rio[WS(vs, 1) + WS(rs, 2)] = FMA(T3t, T3u, T3v * T3w); iio[WS(vs, 1) + WS(rs, 2)] = FNMS(T3v, T3u, T3t * T3w); } } { E T8y, T8E, T8C, T8G, T8x, T8B; T8x = KP707106781 * (T8e + T89); T8y = T8w - T8x; T8E = T8w + T8x; T8B = KP707106781 * (T8n + T8o); T8C = T8A - T8B; T8G = T8A + T8B; { E T8v, T8z, T8D, T8F; T8v = W[8]; T8z = W[9]; rio[WS(vs, 5) + WS(rs, 6)] = FMA(T8v, T8y, T8z * T8C); iio[WS(vs, 5) + WS(rs, 6)] = FNMS(T8z, T8y, T8v * T8C); T8D = W[0]; T8F = W[1]; rio[WS(vs, 1) + WS(rs, 6)] = FMA(T8D, T8E, T8F * T8G); iio[WS(vs, 1) + WS(rs, 6)] = FNMS(T8F, T8E, T8D * T8G); } } } } } static const tw_instr twinstr[] = { {TW_FULL, 0, 8}, {TW_NEXT, 1, 0} }; static const ct_desc desc = { 8, "q1_8", twinstr, &GENUS, {416, 144, 112, 0}, 0, 0, 0 }; void X(codelet_q1_8) (planner *p) { X(kdft_difsq_register) (p, q1_8, &desc); } #endif /* HAVE_FMA */ fftw-3.3.3/dft/scalar/codelets/q1_3.c0000644000175000017500000002174612054410307014150 00000000000000/* * Copyright (c) 2003, 2007-11 Matteo Frigo * Copyright (c) 2003, 2007-11 Massachusetts Institute of Technology * * This program is free software; you can redistribute it and/or modify * it under the terms of the GNU General Public License as published by * the Free Software Foundation; either version 2 of the License, or * (at your option) any later version. * * This program is distributed in the hope that it will be useful, * but WITHOUT ANY WARRANTY; without even the implied warranty of * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the * GNU General Public License for more details. * * You should have received a copy of the GNU General Public License * along with this program; if not, write to the Free Software * Foundation, Inc., 51 Franklin Street, Fifth Floor, Boston, MA 02110-1301 USA * */ /* This file was automatically generated --- DO NOT EDIT */ /* Generated on Sun Nov 25 07:36:23 EST 2012 */ #include "codelet-dft.h" #ifdef HAVE_FMA /* Generated by: ../../../genfft/gen_twidsq.native -fma -reorder-insns -schedule-for-pipeline -compact -variables 4 -pipeline-latency 4 -reload-twiddle -dif -n 3 -name q1_3 -include q.h */ /* * This function contains 48 FP additions, 42 FP multiplications, * (or, 18 additions, 12 multiplications, 30 fused multiply/add), * 56 stack variables, 2 constants, and 36 memory accesses */ #include "q.h" static void q1_3(R *rio, R *iio, const R *W, stride rs, stride vs, INT mb, INT me, INT ms) { DK(KP866025403, +0.866025403784438646763723170752936183471402627); DK(KP500000000, +0.500000000000000000000000000000000000000000000); { INT m; for (m = mb, W = W + (mb * 4); m < me; m = m + 1, rio = rio + ms, iio = iio + ms, W = W + 4, MAKE_VOLATILE_STRIDE(6, rs), MAKE_VOLATILE_STRIDE(0, vs)) { E Tk, Tn, Tm, To, Tl; { E T1, Td, T4, Tg, Tp, T9, Te, T6, Tf, TB, TE, Ts, TZ, Tu, Tx; E TC, TN, TO, TD, TV, T10, TP, Tq, Tr; { E T2, T3, T7, T8; T1 = rio[0]; T2 = rio[WS(rs, 1)]; T3 = rio[WS(rs, 2)]; Td = iio[0]; T7 = iio[WS(rs, 1)]; T8 = iio[WS(rs, 2)]; T4 = T2 + T3; Tg = T3 - T2; Tp = rio[WS(vs, 1)]; T9 = T7 - T8; Te = T7 + T8; T6 = FNMS(KP500000000, T4, T1); Tq = rio[WS(vs, 1) + WS(rs, 1)]; Tr = rio[WS(vs, 1) + WS(rs, 2)]; Tf = FNMS(KP500000000, Te, Td); } { E Tv, Tw, TT, TU; TB = iio[WS(vs, 1)]; Tv = iio[WS(vs, 1) + WS(rs, 1)]; TE = Tr - Tq; Ts = Tq + Tr; Tw = iio[WS(vs, 1) + WS(rs, 2)]; TZ = iio[WS(vs, 2)]; TT = iio[WS(vs, 2) + WS(rs, 1)]; Tu = FNMS(KP500000000, Ts, Tp); Tx = Tv - Tw; TC = Tv + Tw; TU = iio[WS(vs, 2) + WS(rs, 2)]; TN = rio[WS(vs, 2)]; TO = rio[WS(vs, 2) + WS(rs, 1)]; TD = FNMS(KP500000000, TC, TB); TV = TT - TU; T10 = TT + TU; TP = rio[WS(vs, 2) + WS(rs, 2)]; } { E T11, T12, TS, TQ; rio[0] = T1 + T4; iio[0] = Td + Te; T11 = FNMS(KP500000000, T10, TZ); T12 = TP - TO; TQ = TO + TP; rio[WS(rs, 1)] = Tp + Ts; iio[WS(rs, 1)] = TB + TC; iio[WS(rs, 2)] = TZ + T10; TS = FNMS(KP500000000, TQ, TN); rio[WS(rs, 2)] = TN + TQ; { E TW, T13, Ty, TI, TL, TF, TH, TK; { E Ta, Th, T5, Tc; Tk = FNMS(KP866025403, T9, T6); Ta = FMA(KP866025403, T9, T6); Th = FMA(KP866025403, Tg, Tf); Tn = FNMS(KP866025403, Tg, Tf); T5 = W[0]; Tc = W[1]; { E T16, T19, T18, T1a, T17, Ti, Tb, T15; TW = FMA(KP866025403, TV, TS); T16 = FNMS(KP866025403, TV, TS); T19 = FNMS(KP866025403, T12, T11); T13 = FMA(KP866025403, T12, T11); Ti = T5 * Th; Tb = T5 * Ta; T15 = W[2]; T18 = W[3]; iio[WS(vs, 1)] = FNMS(Tc, Ta, Ti); rio[WS(vs, 1)] = FMA(Tc, Th, Tb); T1a = T15 * T19; T17 = T15 * T16; Ty = FMA(KP866025403, Tx, Tu); TI = FNMS(KP866025403, Tx, Tu); TL = FNMS(KP866025403, TE, TD); TF = FMA(KP866025403, TE, TD); iio[WS(vs, 2) + WS(rs, 2)] = FNMS(T18, T16, T1a); rio[WS(vs, 2) + WS(rs, 2)] = FMA(T18, T19, T17); TH = W[2]; TK = W[3]; } } { E TA, TG, Tz, TM, TJ, Tt; TM = TH * TL; TJ = TH * TI; Tt = W[0]; TA = W[1]; iio[WS(vs, 2) + WS(rs, 1)] = FNMS(TK, TI, TM); rio[WS(vs, 2) + WS(rs, 1)] = FMA(TK, TL, TJ); TG = Tt * TF; Tz = Tt * Ty; { E TR, TY, T14, TX, Tj; iio[WS(vs, 1) + WS(rs, 1)] = FNMS(TA, Ty, TG); rio[WS(vs, 1) + WS(rs, 1)] = FMA(TA, TF, Tz); TR = W[0]; TY = W[1]; T14 = TR * T13; TX = TR * TW; Tj = W[2]; Tm = W[3]; iio[WS(vs, 1) + WS(rs, 2)] = FNMS(TY, TW, T14); rio[WS(vs, 1) + WS(rs, 2)] = FMA(TY, T13, TX); To = Tj * Tn; Tl = Tj * Tk; } } } } } iio[WS(vs, 2)] = FNMS(Tm, Tk, To); rio[WS(vs, 2)] = FMA(Tm, Tn, Tl); } } } static const tw_instr twinstr[] = { {TW_FULL, 0, 3}, {TW_NEXT, 1, 0} }; static const ct_desc desc = { 3, "q1_3", twinstr, &GENUS, {18, 12, 30, 0}, 0, 0, 0 }; void X(codelet_q1_3) (planner *p) { X(kdft_difsq_register) (p, q1_3, &desc); } #else /* HAVE_FMA */ /* Generated by: ../../../genfft/gen_twidsq.native -compact -variables 4 -pipeline-latency 4 -reload-twiddle -dif -n 3 -name q1_3 -include q.h */ /* * This function contains 48 FP additions, 36 FP multiplications, * (or, 30 additions, 18 multiplications, 18 fused multiply/add), * 35 stack variables, 2 constants, and 36 memory accesses */ #include "q.h" static void q1_3(R *rio, R *iio, const R *W, stride rs, stride vs, INT mb, INT me, INT ms) { DK(KP866025403, +0.866025403784438646763723170752936183471402627); DK(KP500000000, +0.500000000000000000000000000000000000000000000); { INT m; for (m = mb, W = W + (mb * 4); m < me; m = m + 1, rio = rio + ms, iio = iio + ms, W = W + 4, MAKE_VOLATILE_STRIDE(6, rs), MAKE_VOLATILE_STRIDE(0, vs)) { E T1, T4, T6, Tc, Td, Te, T9, Tf, Tl, To, Tq, Tw, Tx, Ty, Tt; E Tz, TR, TS, TN, TT, TF, TI, TK, TQ; { E T2, T3, Tr, Ts; T1 = rio[0]; T2 = rio[WS(rs, 1)]; T3 = rio[WS(rs, 2)]; T4 = T2 + T3; T6 = FNMS(KP500000000, T4, T1); Tc = KP866025403 * (T3 - T2); { E T7, T8, Tm, Tn; Td = iio[0]; T7 = iio[WS(rs, 1)]; T8 = iio[WS(rs, 2)]; Te = T7 + T8; T9 = KP866025403 * (T7 - T8); Tf = FNMS(KP500000000, Te, Td); Tl = rio[WS(vs, 1)]; Tm = rio[WS(vs, 1) + WS(rs, 1)]; Tn = rio[WS(vs, 1) + WS(rs, 2)]; To = Tm + Tn; Tq = FNMS(KP500000000, To, Tl); Tw = KP866025403 * (Tn - Tm); } Tx = iio[WS(vs, 1)]; Tr = iio[WS(vs, 1) + WS(rs, 1)]; Ts = iio[WS(vs, 1) + WS(rs, 2)]; Ty = Tr + Ts; Tt = KP866025403 * (Tr - Ts); Tz = FNMS(KP500000000, Ty, Tx); { E TL, TM, TG, TH; TR = iio[WS(vs, 2)]; TL = iio[WS(vs, 2) + WS(rs, 1)]; TM = iio[WS(vs, 2) + WS(rs, 2)]; TS = TL + TM; TN = KP866025403 * (TL - TM); TT = FNMS(KP500000000, TS, TR); TF = rio[WS(vs, 2)]; TG = rio[WS(vs, 2) + WS(rs, 1)]; TH = rio[WS(vs, 2) + WS(rs, 2)]; TI = TG + TH; TK = FNMS(KP500000000, TI, TF); TQ = KP866025403 * (TH - TG); } } rio[0] = T1 + T4; iio[0] = Td + Te; rio[WS(rs, 1)] = Tl + To; iio[WS(rs, 1)] = Tx + Ty; iio[WS(rs, 2)] = TR + TS; rio[WS(rs, 2)] = TF + TI; { E Ta, Tg, T5, Tb; Ta = T6 + T9; Tg = Tc + Tf; T5 = W[0]; Tb = W[1]; rio[WS(vs, 1)] = FMA(T5, Ta, Tb * Tg); iio[WS(vs, 1)] = FNMS(Tb, Ta, T5 * Tg); } { E TW, TY, TV, TX; TW = TK - TN; TY = TT - TQ; TV = W[2]; TX = W[3]; rio[WS(vs, 2) + WS(rs, 2)] = FMA(TV, TW, TX * TY); iio[WS(vs, 2) + WS(rs, 2)] = FNMS(TX, TW, TV * TY); } { E TC, TE, TB, TD; TC = Tq - Tt; TE = Tz - Tw; TB = W[2]; TD = W[3]; rio[WS(vs, 2) + WS(rs, 1)] = FMA(TB, TC, TD * TE); iio[WS(vs, 2) + WS(rs, 1)] = FNMS(TD, TC, TB * TE); } { E Tu, TA, Tp, Tv; Tu = Tq + Tt; TA = Tw + Tz; Tp = W[0]; Tv = W[1]; rio[WS(vs, 1) + WS(rs, 1)] = FMA(Tp, Tu, Tv * TA); iio[WS(vs, 1) + WS(rs, 1)] = FNMS(Tv, Tu, Tp * TA); } { E TO, TU, TJ, TP; TO = TK + TN; TU = TQ + TT; TJ = W[0]; TP = W[1]; rio[WS(vs, 1) + WS(rs, 2)] = FMA(TJ, TO, TP * TU); iio[WS(vs, 1) + WS(rs, 2)] = FNMS(TP, TO, TJ * TU); } { E Ti, Tk, Th, Tj; Ti = T6 - T9; Tk = Tf - Tc; Th = W[2]; Tj = W[3]; rio[WS(vs, 2)] = FMA(Th, Ti, Tj * Tk); iio[WS(vs, 2)] = FNMS(Tj, Ti, Th * Tk); } } } } static const tw_instr twinstr[] = { {TW_FULL, 0, 3}, {TW_NEXT, 1, 0} }; static const ct_desc desc = { 3, "q1_3", twinstr, &GENUS, {30, 18, 18, 0}, 0, 0, 0 }; void X(codelet_q1_3) (planner *p) { X(kdft_difsq_register) (p, q1_3, &desc); } #endif /* HAVE_FMA */ fftw-3.3.3/dft/scalar/codelets/q1_5.c0000644000175000017500000006770112054410310014145 00000000000000/* * Copyright (c) 2003, 2007-11 Matteo Frigo * Copyright (c) 2003, 2007-11 Massachusetts Institute of Technology * * This program is free software; you can redistribute it and/or modify * it under the terms of the GNU General Public License as published by * the Free Software Foundation; either version 2 of the License, or * (at your option) any later version. * * This program is distributed in the hope that it will be useful, * but WITHOUT ANY WARRANTY; without even the implied warranty of * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the * GNU General Public License for more details. * * You should have received a copy of the GNU General Public License * along with this program; if not, write to the Free Software * Foundation, Inc., 51 Franklin Street, Fifth Floor, Boston, MA 02110-1301 USA * */ /* This file was automatically generated --- DO NOT EDIT */ /* Generated on Sun Nov 25 07:36:23 EST 2012 */ #include "codelet-dft.h" #ifdef HAVE_FMA /* Generated by: ../../../genfft/gen_twidsq.native -fma -reorder-insns -schedule-for-pipeline -compact -variables 4 -pipeline-latency 4 -reload-twiddle -dif -n 5 -name q1_5 -include q.h */ /* * This function contains 200 FP additions, 170 FP multiplications, * (or, 70 additions, 40 multiplications, 130 fused multiply/add), * 104 stack variables, 4 constants, and 100 memory accesses */ #include "q.h" static void q1_5(R *rio, R *iio, const R *W, stride rs, stride vs, INT mb, INT me, INT ms) { DK(KP951056516, +0.951056516295153572116439333379382143405698634); DK(KP559016994, +0.559016994374947424102293417182819058860154590); DK(KP250000000, +0.250000000000000000000000000000000000000000000); DK(KP618033988, +0.618033988749894848204586834365638117720309180); { INT m; for (m = mb, W = W + (mb * 8); m < me; m = m + 1, rio = rio + ms, iio = iio + ms, W = W + 8, MAKE_VOLATILE_STRIDE(10, rs), MAKE_VOLATILE_STRIDE(0, vs)) { E T1x, T1w, T1v; { E T1, Tn, TM, Tw, Tb, T8, Ta, TV, Tq, Ts, TH, Tj, Tr, T1h, T1q; E T1G, T12, T15, T1P, T14, T1k, T1m, T1B, T1d, T1l, T2b, T2k, T2A, T1W, T1Z; E T3Z, T1Y, T2e, T2g, T2v, T27, T2f, T3D, T42, T44, T4j, T3V, T43, T2J, T48; E T4o, T3K, T3N, T35, T3M, T2V, T3e, T3u, T2Q, T2T, T37, T30, T2S, T2W; { E T1Q, T2j, T1V, T1R; { E Tp, Ti, Td, Te; { E T5, T6, T2, T3, T7, Tv; T1 = rio[0]; T5 = rio[WS(rs, 2)]; T6 = rio[WS(rs, 3)]; T2 = rio[WS(rs, 1)]; T3 = rio[WS(rs, 4)]; Tn = iio[0]; T7 = T5 + T6; Tv = T5 - T6; { E T4, Tu, Tg, Th; T4 = T2 + T3; Tu = T2 - T3; Tg = iio[WS(rs, 2)]; Th = iio[WS(rs, 3)]; TM = FNMS(KP618033988, Tu, Tv); Tw = FMA(KP618033988, Tv, Tu); Tb = T4 - T7; T8 = T4 + T7; Tp = Tg + Th; Ti = Tg - Th; Ta = FNMS(KP250000000, T8, T1); Td = iio[WS(rs, 1)]; Te = iio[WS(rs, 4)]; } } { E TW, T1p, T11, TX; TV = rio[WS(vs, 1)]; { E TZ, T10, Tf, To; TZ = rio[WS(vs, 1) + WS(rs, 2)]; T10 = rio[WS(vs, 1) + WS(rs, 3)]; Tf = Td - Te; To = Td + Te; TW = rio[WS(vs, 1) + WS(rs, 1)]; T1p = TZ - T10; T11 = TZ + T10; Tq = To + Tp; Ts = To - Tp; TH = FNMS(KP618033988, Tf, Ti); Tj = FMA(KP618033988, Ti, Tf); Tr = FNMS(KP250000000, Tq, Tn); TX = rio[WS(vs, 1) + WS(rs, 4)]; } { E T17, T1j, T1c, T18; T1h = iio[WS(vs, 1)]; { E T1a, T1b, TY, T1o; T1a = iio[WS(vs, 1) + WS(rs, 2)]; T1b = iio[WS(vs, 1) + WS(rs, 3)]; TY = TW + TX; T1o = TW - TX; T17 = iio[WS(vs, 1) + WS(rs, 1)]; T1j = T1a + T1b; T1c = T1a - T1b; T1q = FMA(KP618033988, T1p, T1o); T1G = FNMS(KP618033988, T1o, T1p); T12 = TY + T11; T15 = TY - T11; T18 = iio[WS(vs, 1) + WS(rs, 4)]; } T1P = rio[WS(vs, 2)]; T14 = FNMS(KP250000000, T12, TV); { E T1T, T1i, T19, T1U; T1T = rio[WS(vs, 2) + WS(rs, 2)]; T1i = T17 + T18; T19 = T17 - T18; T1U = rio[WS(vs, 2) + WS(rs, 3)]; T1Q = rio[WS(vs, 2) + WS(rs, 1)]; T1k = T1i + T1j; T1m = T1i - T1j; T1B = FNMS(KP618033988, T19, T1c); T1d = FMA(KP618033988, T1c, T19); T2j = T1T - T1U; T1V = T1T + T1U; T1l = FNMS(KP250000000, T1k, T1h); T1R = rio[WS(vs, 2) + WS(rs, 4)]; } } } } { E T3P, T41, T3U, T3Q; { E T21, T2d, T26, T22; T2b = iio[WS(vs, 2)]; { E T24, T25, T1S, T2i; T24 = iio[WS(vs, 2) + WS(rs, 2)]; T25 = iio[WS(vs, 2) + WS(rs, 3)]; T1S = T1Q + T1R; T2i = T1Q - T1R; T21 = iio[WS(vs, 2) + WS(rs, 1)]; T2d = T24 + T25; T26 = T24 - T25; T2k = FMA(KP618033988, T2j, T2i); T2A = FNMS(KP618033988, T2i, T2j); T1W = T1S + T1V; T1Z = T1S - T1V; T22 = iio[WS(vs, 2) + WS(rs, 4)]; } T3Z = iio[WS(vs, 4)]; T1Y = FNMS(KP250000000, T1W, T1P); { E T3S, T2c, T23, T3T; T3S = iio[WS(vs, 4) + WS(rs, 2)]; T2c = T21 + T22; T23 = T21 - T22; T3T = iio[WS(vs, 4) + WS(rs, 3)]; T3P = iio[WS(vs, 4) + WS(rs, 1)]; T2e = T2c + T2d; T2g = T2c - T2d; T2v = FNMS(KP618033988, T23, T26); T27 = FMA(KP618033988, T26, T23); T41 = T3S + T3T; T3U = T3S - T3T; T2f = FNMS(KP250000000, T2e, T2b); T3Q = iio[WS(vs, 4) + WS(rs, 4)]; } } { E T3E, T47, T3J, T3F; T3D = rio[WS(vs, 4)]; { E T3H, T3I, T3R, T40; T3H = rio[WS(vs, 4) + WS(rs, 2)]; T3I = rio[WS(vs, 4) + WS(rs, 3)]; T3R = T3P - T3Q; T40 = T3P + T3Q; T3E = rio[WS(vs, 4) + WS(rs, 1)]; T47 = T3H - T3I; T3J = T3H + T3I; T42 = T40 + T41; T44 = T40 - T41; T4j = FNMS(KP618033988, T3R, T3U); T3V = FMA(KP618033988, T3U, T3R); T43 = FNMS(KP250000000, T42, T3Z); T3F = rio[WS(vs, 4) + WS(rs, 4)]; } { E T2K, T3d, T2P, T2L; T2J = rio[WS(vs, 3)]; { E T2N, T2O, T3G, T46; T2N = rio[WS(vs, 3) + WS(rs, 2)]; T2O = rio[WS(vs, 3) + WS(rs, 3)]; T3G = T3E + T3F; T46 = T3E - T3F; T2K = rio[WS(vs, 3) + WS(rs, 1)]; T3d = T2N - T2O; T2P = T2N + T2O; T48 = FMA(KP618033988, T47, T46); T4o = FNMS(KP618033988, T46, T47); T3K = T3G + T3J; T3N = T3G - T3J; T2L = rio[WS(vs, 3) + WS(rs, 4)]; } T35 = iio[WS(vs, 3)]; T3M = FNMS(KP250000000, T3K, T3D); { E T2Y, T3c, T2M, T2Z; T2Y = iio[WS(vs, 3) + WS(rs, 2)]; T3c = T2K - T2L; T2M = T2K + T2L; T2Z = iio[WS(vs, 3) + WS(rs, 3)]; T2V = iio[WS(vs, 3) + WS(rs, 1)]; T3e = FMA(KP618033988, T3d, T3c); T3u = FNMS(KP618033988, T3c, T3d); T2Q = T2M + T2P; T2T = T2M - T2P; T37 = T2Y + T2Z; T30 = T2Y - T2Z; T2S = FNMS(KP250000000, T2Q, T2J); T2W = iio[WS(vs, 3) + WS(rs, 4)]; } } } } } { E T3a, T31, T3p, T39, T2X, T36, T38; rio[0] = T1 + T8; iio[0] = Tn + Tq; rio[WS(rs, 1)] = TV + T12; T2X = T2V - T2W; T36 = T2V + T2W; iio[WS(rs, 1)] = T1h + T1k; rio[WS(rs, 2)] = T1P + T1W; T3a = T36 - T37; T38 = T36 + T37; T31 = FMA(KP618033988, T30, T2X); T3p = FNMS(KP618033988, T2X, T30); T39 = FNMS(KP250000000, T38, T35); iio[WS(rs, 2)] = T2b + T2e; iio[WS(rs, 4)] = T3Z + T42; rio[WS(rs, 4)] = T3D + T3K; rio[WS(rs, 3)] = T2J + T2Q; iio[WS(rs, 3)] = T35 + T38; { E T3O, T45, T2r, T2q, T2p, TT, TS, TR; { E TG, TL, TD, TC, TB, Tc, Tt; TG = FNMS(KP559016994, Tb, Ta); Tc = FMA(KP559016994, Tb, Ta); Tt = FMA(KP559016994, Ts, Tr); TL = FNMS(KP559016994, Ts, Tr); { E T9, Tm, Tk, TA, Tx; T9 = W[0]; Tm = W[1]; Tk = FMA(KP951056516, Tj, Tc); TA = FNMS(KP951056516, Tj, Tc); Tx = FNMS(KP951056516, Tw, Tt); TD = FMA(KP951056516, Tw, Tt); { E Tz, Tl, Ty, TE; Tz = W[6]; Tl = T9 * Tk; TC = W[7]; Ty = T9 * Tx; TE = Tz * TD; TB = Tz * TA; rio[WS(vs, 1)] = FMA(Tm, Tx, Tl); iio[WS(vs, 1)] = FNMS(Tm, Tk, Ty); iio[WS(vs, 4)] = FNMS(TC, TA, TE); } } rio[WS(vs, 4)] = FMA(TC, TD, TB); { E TF, TK, TI, TQ, TN; TF = W[2]; TK = W[3]; TI = FNMS(KP951056516, TH, TG); TQ = FMA(KP951056516, TH, TG); TN = FMA(KP951056516, TM, TL); TT = FNMS(KP951056516, TM, TL); { E TP, TJ, TO, TU; TP = W[4]; TJ = TF * TI; TS = W[5]; TO = TF * TN; TU = TP * TT; TR = TP * TQ; rio[WS(vs, 2)] = FMA(TK, TN, TJ); iio[WS(vs, 2)] = FNMS(TK, TI, TO); iio[WS(vs, 3)] = FNMS(TS, TQ, TU); } } } rio[WS(vs, 3)] = FMA(TS, TT, TR); { E T20, T2h, T2H, T2G, T2F, T2u, T2z; T20 = FMA(KP559016994, T1Z, T1Y); T2u = FNMS(KP559016994, T1Z, T1Y); T2z = FNMS(KP559016994, T2g, T2f); T2h = FMA(KP559016994, T2g, T2f); { E T2t, T2y, T2w, T2E, T2B; T2t = W[2]; T2y = W[3]; T2w = FNMS(KP951056516, T2v, T2u); T2E = FMA(KP951056516, T2v, T2u); T2B = FMA(KP951056516, T2A, T2z); T2H = FNMS(KP951056516, T2A, T2z); { E T2D, T2x, T2C, T2I; T2D = W[4]; T2x = T2t * T2w; T2G = W[5]; T2C = T2t * T2B; T2I = T2D * T2H; T2F = T2D * T2E; rio[WS(vs, 2) + WS(rs, 2)] = FMA(T2y, T2B, T2x); iio[WS(vs, 2) + WS(rs, 2)] = FNMS(T2y, T2w, T2C); iio[WS(vs, 3) + WS(rs, 2)] = FNMS(T2G, T2E, T2I); } } rio[WS(vs, 3) + WS(rs, 2)] = FMA(T2G, T2H, T2F); { E T4v, T4u, T4t, T4i, T4n; T3O = FMA(KP559016994, T3N, T3M); T4i = FNMS(KP559016994, T3N, T3M); T4n = FNMS(KP559016994, T44, T43); T45 = FMA(KP559016994, T44, T43); { E T4h, T4m, T4k, T4s, T4p; T4h = W[2]; T4m = W[3]; T4k = FNMS(KP951056516, T4j, T4i); T4s = FMA(KP951056516, T4j, T4i); T4p = FMA(KP951056516, T4o, T4n); T4v = FNMS(KP951056516, T4o, T4n); { E T4r, T4l, T4q, T4w; T4r = W[4]; T4l = T4h * T4k; T4u = W[5]; T4q = T4h * T4p; T4w = T4r * T4v; T4t = T4r * T4s; rio[WS(vs, 2) + WS(rs, 4)] = FMA(T4m, T4p, T4l); iio[WS(vs, 2) + WS(rs, 4)] = FNMS(T4m, T4k, T4q); iio[WS(vs, 3) + WS(rs, 4)] = FNMS(T4u, T4s, T4w); } } rio[WS(vs, 3) + WS(rs, 4)] = FMA(T4u, T4v, T4t); { E T1X, T2a, T28, T2o, T2l; T1X = W[0]; T2a = W[1]; T28 = FMA(KP951056516, T27, T20); T2o = FNMS(KP951056516, T27, T20); T2l = FNMS(KP951056516, T2k, T2h); T2r = FMA(KP951056516, T2k, T2h); { E T2n, T29, T2m, T2s; T2n = W[6]; T29 = T1X * T28; T2q = W[7]; T2m = T1X * T2l; T2s = T2n * T2r; T2p = T2n * T2o; rio[WS(vs, 1) + WS(rs, 2)] = FMA(T2a, T2l, T29); iio[WS(vs, 1) + WS(rs, 2)] = FNMS(T2a, T28, T2m); iio[WS(vs, 4) + WS(rs, 2)] = FNMS(T2q, T2o, T2s); } } } } rio[WS(vs, 4) + WS(rs, 2)] = FMA(T2q, T2r, T2p); { E T3B, T3A, T3z, T4f, T4e, T4d; { E T3o, T3t, T3l, T3k, T3j, T2U, T3b; T3o = FNMS(KP559016994, T2T, T2S); T2U = FMA(KP559016994, T2T, T2S); T3b = FMA(KP559016994, T3a, T39); T3t = FNMS(KP559016994, T3a, T39); { E T2R, T34, T32, T3i, T3f; T2R = W[0]; T34 = W[1]; T32 = FMA(KP951056516, T31, T2U); T3i = FNMS(KP951056516, T31, T2U); T3f = FNMS(KP951056516, T3e, T3b); T3l = FMA(KP951056516, T3e, T3b); { E T3h, T33, T3g, T3m; T3h = W[6]; T33 = T2R * T32; T3k = W[7]; T3g = T2R * T3f; T3m = T3h * T3l; T3j = T3h * T3i; rio[WS(vs, 1) + WS(rs, 3)] = FMA(T34, T3f, T33); iio[WS(vs, 1) + WS(rs, 3)] = FNMS(T34, T32, T3g); iio[WS(vs, 4) + WS(rs, 3)] = FNMS(T3k, T3i, T3m); } } rio[WS(vs, 4) + WS(rs, 3)] = FMA(T3k, T3l, T3j); { E T3n, T3s, T3q, T3y, T3v; T3n = W[2]; T3s = W[3]; T3q = FNMS(KP951056516, T3p, T3o); T3y = FMA(KP951056516, T3p, T3o); T3v = FMA(KP951056516, T3u, T3t); T3B = FNMS(KP951056516, T3u, T3t); { E T3x, T3r, T3w, T3C; T3x = W[4]; T3r = T3n * T3q; T3A = W[5]; T3w = T3n * T3v; T3C = T3x * T3B; T3z = T3x * T3y; rio[WS(vs, 2) + WS(rs, 3)] = FMA(T3s, T3v, T3r); iio[WS(vs, 2) + WS(rs, 3)] = FNMS(T3s, T3q, T3w); iio[WS(vs, 3) + WS(rs, 3)] = FNMS(T3A, T3y, T3C); } } } rio[WS(vs, 3) + WS(rs, 3)] = FMA(T3A, T3B, T3z); { E T3L, T3Y, T3W, T4c, T49; T3L = W[0]; T3Y = W[1]; T3W = FMA(KP951056516, T3V, T3O); T4c = FNMS(KP951056516, T3V, T3O); T49 = FNMS(KP951056516, T48, T45); T4f = FMA(KP951056516, T48, T45); { E T4b, T3X, T4a, T4g; T4b = W[6]; T3X = T3L * T3W; T4e = W[7]; T4a = T3L * T49; T4g = T4b * T4f; T4d = T4b * T4c; rio[WS(vs, 1) + WS(rs, 4)] = FMA(T3Y, T49, T3X); iio[WS(vs, 1) + WS(rs, 4)] = FNMS(T3Y, T3W, T4a); iio[WS(vs, 4) + WS(rs, 4)] = FNMS(T4e, T4c, T4g); } } rio[WS(vs, 4) + WS(rs, 4)] = FMA(T4e, T4f, T4d); { E T16, T1n, T1N, T1M, T1L, T1A, T1F; T16 = FMA(KP559016994, T15, T14); T1A = FNMS(KP559016994, T15, T14); T1F = FNMS(KP559016994, T1m, T1l); T1n = FMA(KP559016994, T1m, T1l); { E T1z, T1E, T1C, T1K, T1H; T1z = W[2]; T1E = W[3]; T1C = FNMS(KP951056516, T1B, T1A); T1K = FMA(KP951056516, T1B, T1A); T1H = FMA(KP951056516, T1G, T1F); T1N = FNMS(KP951056516, T1G, T1F); { E T1J, T1D, T1I, T1O; T1J = W[4]; T1D = T1z * T1C; T1M = W[5]; T1I = T1z * T1H; T1O = T1J * T1N; T1L = T1J * T1K; rio[WS(vs, 2) + WS(rs, 1)] = FMA(T1E, T1H, T1D); iio[WS(vs, 2) + WS(rs, 1)] = FNMS(T1E, T1C, T1I); iio[WS(vs, 3) + WS(rs, 1)] = FNMS(T1M, T1K, T1O); } } rio[WS(vs, 3) + WS(rs, 1)] = FMA(T1M, T1N, T1L); { E T13, T1g, T1e, T1u, T1r; T13 = W[0]; T1g = W[1]; T1e = FMA(KP951056516, T1d, T16); T1u = FNMS(KP951056516, T1d, T16); T1r = FNMS(KP951056516, T1q, T1n); T1x = FMA(KP951056516, T1q, T1n); { E T1t, T1f, T1s, T1y; T1t = W[6]; T1f = T13 * T1e; T1w = W[7]; T1s = T13 * T1r; T1y = T1t * T1x; T1v = T1t * T1u; rio[WS(vs, 1) + WS(rs, 1)] = FMA(T1g, T1r, T1f); iio[WS(vs, 1) + WS(rs, 1)] = FNMS(T1g, T1e, T1s); iio[WS(vs, 4) + WS(rs, 1)] = FNMS(T1w, T1u, T1y); } } } } } } } rio[WS(vs, 4) + WS(rs, 1)] = FMA(T1w, T1x, T1v); } } } static const tw_instr twinstr[] = { {TW_FULL, 0, 5}, {TW_NEXT, 1, 0} }; static const ct_desc desc = { 5, "q1_5", twinstr, &GENUS, {70, 40, 130, 0}, 0, 0, 0 }; void X(codelet_q1_5) (planner *p) { X(kdft_difsq_register) (p, q1_5, &desc); } #else /* HAVE_FMA */ /* Generated by: ../../../genfft/gen_twidsq.native -compact -variables 4 -pipeline-latency 4 -reload-twiddle -dif -n 5 -name q1_5 -include q.h */ /* * This function contains 200 FP additions, 140 FP multiplications, * (or, 130 additions, 70 multiplications, 70 fused multiply/add), * 75 stack variables, 4 constants, and 100 memory accesses */ #include "q.h" static void q1_5(R *rio, R *iio, const R *W, stride rs, stride vs, INT mb, INT me, INT ms) { DK(KP250000000, +0.250000000000000000000000000000000000000000000); DK(KP587785252, +0.587785252292473129168705954639072768597652438); DK(KP951056516, +0.951056516295153572116439333379382143405698634); DK(KP559016994, +0.559016994374947424102293417182819058860154590); { INT m; for (m = mb, W = W + (mb * 8); m < me; m = m + 1, rio = rio + ms, iio = iio + ms, W = W + 8, MAKE_VOLATILE_STRIDE(10, rs), MAKE_VOLATILE_STRIDE(0, vs)) { E T1, Ta, TG, Tv, T8, Tb, Tp, Tj, TD, To, Tq, Tr, TN, TW, T1s; E T1h, TU, TX, T1b, T15, T1p, T1a, T1c, T1d, T1z, T1I, T2e, T23, T1G, T1J; E T1X, T1R, T2b, T1W, T1Y, T1Z, T3v, T3p, T3J, T3u, T3w, T3x, T37, T3g, T3M; E T3B, T3e, T3h, T2l, T2u, T30, T2P, T2s, T2v, T2J, T2D, T2X, T2I, T2K, T2L; { E T7, Tu, T4, Tt; T1 = rio[0]; { E T5, T6, T2, T3; T5 = rio[WS(rs, 2)]; T6 = rio[WS(rs, 3)]; T7 = T5 + T6; Tu = T5 - T6; T2 = rio[WS(rs, 1)]; T3 = rio[WS(rs, 4)]; T4 = T2 + T3; Tt = T2 - T3; } Ta = KP559016994 * (T4 - T7); TG = FNMS(KP587785252, Tt, KP951056516 * Tu); Tv = FMA(KP951056516, Tt, KP587785252 * Tu); T8 = T4 + T7; Tb = FNMS(KP250000000, T8, T1); } { E Ti, Tn, Tf, Tm; Tp = iio[0]; { E Tg, Th, Td, Te; Tg = iio[WS(rs, 2)]; Th = iio[WS(rs, 3)]; Ti = Tg - Th; Tn = Tg + Th; Td = iio[WS(rs, 1)]; Te = iio[WS(rs, 4)]; Tf = Td - Te; Tm = Td + Te; } Tj = FMA(KP951056516, Tf, KP587785252 * Ti); TD = FNMS(KP587785252, Tf, KP951056516 * Ti); To = KP559016994 * (Tm - Tn); Tq = Tm + Tn; Tr = FNMS(KP250000000, Tq, Tp); } { E TT, T1g, TQ, T1f; TN = rio[WS(vs, 1)]; { E TR, TS, TO, TP; TR = rio[WS(vs, 1) + WS(rs, 2)]; TS = rio[WS(vs, 1) + WS(rs, 3)]; TT = TR + TS; T1g = TR - TS; TO = rio[WS(vs, 1) + WS(rs, 1)]; TP = rio[WS(vs, 1) + WS(rs, 4)]; TQ = TO + TP; T1f = TO - TP; } TW = KP559016994 * (TQ - TT); T1s = FNMS(KP587785252, T1f, KP951056516 * T1g); T1h = FMA(KP951056516, T1f, KP587785252 * T1g); TU = TQ + TT; TX = FNMS(KP250000000, TU, TN); } { E T14, T19, T11, T18; T1b = iio[WS(vs, 1)]; { E T12, T13, TZ, T10; T12 = iio[WS(vs, 1) + WS(rs, 2)]; T13 = iio[WS(vs, 1) + WS(rs, 3)]; T14 = T12 - T13; T19 = T12 + T13; TZ = iio[WS(vs, 1) + WS(rs, 1)]; T10 = iio[WS(vs, 1) + WS(rs, 4)]; T11 = TZ - T10; T18 = TZ + T10; } T15 = FMA(KP951056516, T11, KP587785252 * T14); T1p = FNMS(KP587785252, T11, KP951056516 * T14); T1a = KP559016994 * (T18 - T19); T1c = T18 + T19; T1d = FNMS(KP250000000, T1c, T1b); } { E T1F, T22, T1C, T21; T1z = rio[WS(vs, 2)]; { E T1D, T1E, T1A, T1B; T1D = rio[WS(vs, 2) + WS(rs, 2)]; T1E = rio[WS(vs, 2) + WS(rs, 3)]; T1F = T1D + T1E; T22 = T1D - T1E; T1A = rio[WS(vs, 2) + WS(rs, 1)]; T1B = rio[WS(vs, 2) + WS(rs, 4)]; T1C = T1A + T1B; T21 = T1A - T1B; } T1I = KP559016994 * (T1C - T1F); T2e = FNMS(KP587785252, T21, KP951056516 * T22); T23 = FMA(KP951056516, T21, KP587785252 * T22); T1G = T1C + T1F; T1J = FNMS(KP250000000, T1G, T1z); } { E T1Q, T1V, T1N, T1U; T1X = iio[WS(vs, 2)]; { E T1O, T1P, T1L, T1M; T1O = iio[WS(vs, 2) + WS(rs, 2)]; T1P = iio[WS(vs, 2) + WS(rs, 3)]; T1Q = T1O - T1P; T1V = T1O + T1P; T1L = iio[WS(vs, 2) + WS(rs, 1)]; T1M = iio[WS(vs, 2) + WS(rs, 4)]; T1N = T1L - T1M; T1U = T1L + T1M; } T1R = FMA(KP951056516, T1N, KP587785252 * T1Q); T2b = FNMS(KP587785252, T1N, KP951056516 * T1Q); T1W = KP559016994 * (T1U - T1V); T1Y = T1U + T1V; T1Z = FNMS(KP250000000, T1Y, T1X); } { E T3o, T3t, T3l, T3s; T3v = iio[WS(vs, 4)]; { E T3m, T3n, T3j, T3k; T3m = iio[WS(vs, 4) + WS(rs, 2)]; T3n = iio[WS(vs, 4) + WS(rs, 3)]; T3o = T3m - T3n; T3t = T3m + T3n; T3j = iio[WS(vs, 4) + WS(rs, 1)]; T3k = iio[WS(vs, 4) + WS(rs, 4)]; T3l = T3j - T3k; T3s = T3j + T3k; } T3p = FMA(KP951056516, T3l, KP587785252 * T3o); T3J = FNMS(KP587785252, T3l, KP951056516 * T3o); T3u = KP559016994 * (T3s - T3t); T3w = T3s + T3t; T3x = FNMS(KP250000000, T3w, T3v); } { E T3d, T3A, T3a, T3z; T37 = rio[WS(vs, 4)]; { E T3b, T3c, T38, T39; T3b = rio[WS(vs, 4) + WS(rs, 2)]; T3c = rio[WS(vs, 4) + WS(rs, 3)]; T3d = T3b + T3c; T3A = T3b - T3c; T38 = rio[WS(vs, 4) + WS(rs, 1)]; T39 = rio[WS(vs, 4) + WS(rs, 4)]; T3a = T38 + T39; T3z = T38 - T39; } T3g = KP559016994 * (T3a - T3d); T3M = FNMS(KP587785252, T3z, KP951056516 * T3A); T3B = FMA(KP951056516, T3z, KP587785252 * T3A); T3e = T3a + T3d; T3h = FNMS(KP250000000, T3e, T37); } { E T2r, T2O, T2o, T2N; T2l = rio[WS(vs, 3)]; { E T2p, T2q, T2m, T2n; T2p = rio[WS(vs, 3) + WS(rs, 2)]; T2q = rio[WS(vs, 3) + WS(rs, 3)]; T2r = T2p + T2q; T2O = T2p - T2q; T2m = rio[WS(vs, 3) + WS(rs, 1)]; T2n = rio[WS(vs, 3) + WS(rs, 4)]; T2o = T2m + T2n; T2N = T2m - T2n; } T2u = KP559016994 * (T2o - T2r); T30 = FNMS(KP587785252, T2N, KP951056516 * T2O); T2P = FMA(KP951056516, T2N, KP587785252 * T2O); T2s = T2o + T2r; T2v = FNMS(KP250000000, T2s, T2l); } { E T2C, T2H, T2z, T2G; T2J = iio[WS(vs, 3)]; { E T2A, T2B, T2x, T2y; T2A = iio[WS(vs, 3) + WS(rs, 2)]; T2B = iio[WS(vs, 3) + WS(rs, 3)]; T2C = T2A - T2B; T2H = T2A + T2B; T2x = iio[WS(vs, 3) + WS(rs, 1)]; T2y = iio[WS(vs, 3) + WS(rs, 4)]; T2z = T2x - T2y; T2G = T2x + T2y; } T2D = FMA(KP951056516, T2z, KP587785252 * T2C); T2X = FNMS(KP587785252, T2z, KP951056516 * T2C); T2I = KP559016994 * (T2G - T2H); T2K = T2G + T2H; T2L = FNMS(KP250000000, T2K, T2J); } rio[0] = T1 + T8; iio[0] = Tp + Tq; rio[WS(rs, 1)] = TN + TU; iio[WS(rs, 1)] = T1b + T1c; rio[WS(rs, 2)] = T1z + T1G; iio[WS(rs, 2)] = T1X + T1Y; iio[WS(rs, 4)] = T3v + T3w; rio[WS(rs, 4)] = T37 + T3e; rio[WS(rs, 3)] = T2l + T2s; iio[WS(rs, 3)] = T2J + T2K; { E Tk, Ty, Tw, TA, Tc, Ts; Tc = Ta + Tb; Tk = Tc + Tj; Ty = Tc - Tj; Ts = To + Tr; Tw = Ts - Tv; TA = Tv + Ts; { E T9, Tl, Tx, Tz; T9 = W[0]; Tl = W[1]; rio[WS(vs, 1)] = FMA(T9, Tk, Tl * Tw); iio[WS(vs, 1)] = FNMS(Tl, Tk, T9 * Tw); Tx = W[6]; Tz = W[7]; rio[WS(vs, 4)] = FMA(Tx, Ty, Tz * TA); iio[WS(vs, 4)] = FNMS(Tz, Ty, Tx * TA); } } { E TE, TK, TI, TM, TC, TH; TC = Tb - Ta; TE = TC - TD; TK = TC + TD; TH = Tr - To; TI = TG + TH; TM = TH - TG; { E TB, TF, TJ, TL; TB = W[2]; TF = W[3]; rio[WS(vs, 2)] = FMA(TB, TE, TF * TI); iio[WS(vs, 2)] = FNMS(TF, TE, TB * TI); TJ = W[4]; TL = W[5]; rio[WS(vs, 3)] = FMA(TJ, TK, TL * TM); iio[WS(vs, 3)] = FNMS(TL, TK, TJ * TM); } } { E T2c, T2i, T2g, T2k, T2a, T2f; T2a = T1J - T1I; T2c = T2a - T2b; T2i = T2a + T2b; T2f = T1Z - T1W; T2g = T2e + T2f; T2k = T2f - T2e; { E T29, T2d, T2h, T2j; T29 = W[2]; T2d = W[3]; rio[WS(vs, 2) + WS(rs, 2)] = FMA(T29, T2c, T2d * T2g); iio[WS(vs, 2) + WS(rs, 2)] = FNMS(T2d, T2c, T29 * T2g); T2h = W[4]; T2j = W[5]; rio[WS(vs, 3) + WS(rs, 2)] = FMA(T2h, T2i, T2j * T2k); iio[WS(vs, 3) + WS(rs, 2)] = FNMS(T2j, T2i, T2h * T2k); } } { E T3K, T3Q, T3O, T3S, T3I, T3N; T3I = T3h - T3g; T3K = T3I - T3J; T3Q = T3I + T3J; T3N = T3x - T3u; T3O = T3M + T3N; T3S = T3N - T3M; { E T3H, T3L, T3P, T3R; T3H = W[2]; T3L = W[3]; rio[WS(vs, 2) + WS(rs, 4)] = FMA(T3H, T3K, T3L * T3O); iio[WS(vs, 2) + WS(rs, 4)] = FNMS(T3L, T3K, T3H * T3O); T3P = W[4]; T3R = W[5]; rio[WS(vs, 3) + WS(rs, 4)] = FMA(T3P, T3Q, T3R * T3S); iio[WS(vs, 3) + WS(rs, 4)] = FNMS(T3R, T3Q, T3P * T3S); } } { E T1S, T26, T24, T28, T1K, T20; T1K = T1I + T1J; T1S = T1K + T1R; T26 = T1K - T1R; T20 = T1W + T1Z; T24 = T20 - T23; T28 = T23 + T20; { E T1H, T1T, T25, T27; T1H = W[0]; T1T = W[1]; rio[WS(vs, 1) + WS(rs, 2)] = FMA(T1H, T1S, T1T * T24); iio[WS(vs, 1) + WS(rs, 2)] = FNMS(T1T, T1S, T1H * T24); T25 = W[6]; T27 = W[7]; rio[WS(vs, 4) + WS(rs, 2)] = FMA(T25, T26, T27 * T28); iio[WS(vs, 4) + WS(rs, 2)] = FNMS(T27, T26, T25 * T28); } } { E T2E, T2S, T2Q, T2U, T2w, T2M; T2w = T2u + T2v; T2E = T2w + T2D; T2S = T2w - T2D; T2M = T2I + T2L; T2Q = T2M - T2P; T2U = T2P + T2M; { E T2t, T2F, T2R, T2T; T2t = W[0]; T2F = W[1]; rio[WS(vs, 1) + WS(rs, 3)] = FMA(T2t, T2E, T2F * T2Q); iio[WS(vs, 1) + WS(rs, 3)] = FNMS(T2F, T2E, T2t * T2Q); T2R = W[6]; T2T = W[7]; rio[WS(vs, 4) + WS(rs, 3)] = FMA(T2R, T2S, T2T * T2U); iio[WS(vs, 4) + WS(rs, 3)] = FNMS(T2T, T2S, T2R * T2U); } } { E T2Y, T34, T32, T36, T2W, T31; T2W = T2v - T2u; T2Y = T2W - T2X; T34 = T2W + T2X; T31 = T2L - T2I; T32 = T30 + T31; T36 = T31 - T30; { E T2V, T2Z, T33, T35; T2V = W[2]; T2Z = W[3]; rio[WS(vs, 2) + WS(rs, 3)] = FMA(T2V, T2Y, T2Z * T32); iio[WS(vs, 2) + WS(rs, 3)] = FNMS(T2Z, T2Y, T2V * T32); T33 = W[4]; T35 = W[5]; rio[WS(vs, 3) + WS(rs, 3)] = FMA(T33, T34, T35 * T36); iio[WS(vs, 3) + WS(rs, 3)] = FNMS(T35, T34, T33 * T36); } } { E T3q, T3E, T3C, T3G, T3i, T3y; T3i = T3g + T3h; T3q = T3i + T3p; T3E = T3i - T3p; T3y = T3u + T3x; T3C = T3y - T3B; T3G = T3B + T3y; { E T3f, T3r, T3D, T3F; T3f = W[0]; T3r = W[1]; rio[WS(vs, 1) + WS(rs, 4)] = FMA(T3f, T3q, T3r * T3C); iio[WS(vs, 1) + WS(rs, 4)] = FNMS(T3r, T3q, T3f * T3C); T3D = W[6]; T3F = W[7]; rio[WS(vs, 4) + WS(rs, 4)] = FMA(T3D, T3E, T3F * T3G); iio[WS(vs, 4) + WS(rs, 4)] = FNMS(T3F, T3E, T3D * T3G); } } { E T1q, T1w, T1u, T1y, T1o, T1t; T1o = TX - TW; T1q = T1o - T1p; T1w = T1o + T1p; T1t = T1d - T1a; T1u = T1s + T1t; T1y = T1t - T1s; { E T1n, T1r, T1v, T1x; T1n = W[2]; T1r = W[3]; rio[WS(vs, 2) + WS(rs, 1)] = FMA(T1n, T1q, T1r * T1u); iio[WS(vs, 2) + WS(rs, 1)] = FNMS(T1r, T1q, T1n * T1u); T1v = W[4]; T1x = W[5]; rio[WS(vs, 3) + WS(rs, 1)] = FMA(T1v, T1w, T1x * T1y); iio[WS(vs, 3) + WS(rs, 1)] = FNMS(T1x, T1w, T1v * T1y); } } { E T16, T1k, T1i, T1m, TY, T1e; TY = TW + TX; T16 = TY + T15; T1k = TY - T15; T1e = T1a + T1d; T1i = T1e - T1h; T1m = T1h + T1e; { E TV, T17, T1j, T1l; TV = W[0]; T17 = W[1]; rio[WS(vs, 1) + WS(rs, 1)] = FMA(TV, T16, T17 * T1i); iio[WS(vs, 1) + WS(rs, 1)] = FNMS(T17, T16, TV * T1i); T1j = W[6]; T1l = W[7]; rio[WS(vs, 4) + WS(rs, 1)] = FMA(T1j, T1k, T1l * T1m); iio[WS(vs, 4) + WS(rs, 1)] = FNMS(T1l, T1k, T1j * T1m); } } } } } static const tw_instr twinstr[] = { {TW_FULL, 0, 5}, {TW_NEXT, 1, 0} }; static const ct_desc desc = { 5, "q1_5", twinstr, &GENUS, {130, 70, 70, 0}, 0, 0, 0 }; void X(codelet_q1_5) (planner *p) { X(kdft_difsq_register) (p, q1_5, &desc); } #endif /* HAVE_FMA */ fftw-3.3.3/dft/scalar/codelets/q1_6.c0000644000175000017500000011175612054410312014150 00000000000000/* * Copyright (c) 2003, 2007-11 Matteo Frigo * Copyright (c) 2003, 2007-11 Massachusetts Institute of Technology * * This program is free software; you can redistribute it and/or modify * it under the terms of the GNU General Public License as published by * the Free Software Foundation; either version 2 of the License, or * (at your option) any later version. * * This program is distributed in the hope that it will be useful, * but WITHOUT ANY WARRANTY; without even the implied warranty of * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the * GNU General Public License for more details. * * You should have received a copy of the GNU General Public License * along with this program; if not, write to the Free Software * Foundation, Inc., 51 Franklin Street, Fifth Floor, Boston, MA 02110-1301 USA * */ /* This file was automatically generated --- DO NOT EDIT */ /* Generated on Sun Nov 25 07:36:24 EST 2012 */ #include "codelet-dft.h" #ifdef HAVE_FMA /* Generated by: ../../../genfft/gen_twidsq.native -fma -reorder-insns -schedule-for-pipeline -compact -variables 4 -pipeline-latency 4 -reload-twiddle -dif -n 6 -name q1_6 -include q.h */ /* * This function contains 276 FP additions, 192 FP multiplications, * (or, 144 additions, 60 multiplications, 132 fused multiply/add), * 129 stack variables, 2 constants, and 144 memory accesses */ #include "q.h" static void q1_6(R *rio, R *iio, const R *W, stride rs, stride vs, INT mb, INT me, INT ms) { DK(KP866025403, +0.866025403784438646763723170752936183471402627); DK(KP500000000, +0.500000000000000000000000000000000000000000000); { INT m; for (m = mb, W = W + (mb * 10); m < me; m = m + 1, rio = rio + ms, iio = iio + ms, W = W + 10, MAKE_VOLATILE_STRIDE(12, rs), MAKE_VOLATILE_STRIDE(0, vs)) { E T4c, T4f, T4e, T4g, T4d; { E T3, Tw, Ta, TW, Tg, TG, TM, TT, TU, TP, Tn, T17, TV, TJ, Tv; E T1A, T1e, T20, T1k, T1K, T1Q, T1X, T1Y, T1T, T1r, T1Z, T1N, T1z, T31, T32; E T2X, T2v, T2b, T33, T2R, T2D, T2E, T2i, T34, T3f, T2o, T2O, T2U, T3I, T3m; E T48, T3s, T3S, T3Y, T45, T46, T41, T3z, T4j, T47, T3V, T3H, T4M, T4q, T5c; E T4w, T4W, T52, T59, T5a, T55, T4D, T5b, T4Z, T4L, T6d, T5r, T6e, T69, T5H; E T5w, T5n, T6f, T63, T5P, T5s, T5o, T5p; { E T2f, T2k, T2g, T2c, T2d; { E T1b, T1g, T1c, T18, T19; { E T4, Tc, Te, T9, T5; { E T1, T2, T7, T8; T1 = rio[0]; T2 = rio[WS(rs, 3)]; T7 = rio[WS(rs, 4)]; T8 = rio[WS(rs, 1)]; T4 = rio[WS(rs, 2)]; Tc = T1 - T2; T3 = T1 + T2; Te = T7 - T8; T9 = T7 + T8; T5 = rio[WS(rs, 5)]; } { E TN, Tj, Tk, Tl, Tt, Th, Ti; Th = iio[WS(rs, 2)]; Ti = iio[WS(rs, 5)]; { E Tr, Ts, Td, T6, Tf; Tr = iio[0]; Td = T4 - T5; T6 = T4 + T5; TN = Th + Ti; Tj = Th - Ti; Tf = Td + Te; Tw = Te - Td; Ta = T6 + T9; TW = T9 - T6; Tg = FNMS(KP500000000, Tf, Tc); TG = Tc + Tf; Ts = iio[WS(rs, 3)]; TM = FNMS(KP500000000, Ta, T3); Tk = iio[WS(rs, 4)]; Tl = iio[WS(rs, 1)]; Tt = Tr - Ts; TT = Tr + Ts; } { E T15, TO, Tm, T16, Tu; T15 = rio[WS(vs, 1)]; TO = Tk + Tl; Tm = Tk - Tl; T16 = rio[WS(vs, 1) + WS(rs, 3)]; T1b = rio[WS(vs, 1) + WS(rs, 4)]; TU = TN + TO; TP = TN - TO; Tu = Tj + Tm; Tn = Tj - Tm; T1g = T15 - T16; T17 = T15 + T16; TV = FNMS(KP500000000, TU, TT); TJ = Tt + Tu; Tv = FNMS(KP500000000, Tu, Tt); T1c = rio[WS(vs, 1) + WS(rs, 1)]; T18 = rio[WS(vs, 1) + WS(rs, 2)]; T19 = rio[WS(vs, 1) + WS(rs, 5)]; } } } { E T1v, T1R, T1n, T1w, T1o, T1p; { E T1l, T1i, T1d, T1h, T1a, T1m, T1j; T1l = iio[WS(vs, 1) + WS(rs, 2)]; T1i = T1b - T1c; T1d = T1b + T1c; T1h = T18 - T19; T1a = T18 + T19; T1m = iio[WS(vs, 1) + WS(rs, 5)]; T1v = iio[WS(vs, 1)]; T1j = T1h + T1i; T1A = T1i - T1h; T1e = T1a + T1d; T20 = T1d - T1a; T1R = T1l + T1m; T1n = T1l - T1m; T1k = FNMS(KP500000000, T1j, T1g); T1K = T1g + T1j; T1Q = FNMS(KP500000000, T1e, T17); T1w = iio[WS(vs, 1) + WS(rs, 3)]; T1o = iio[WS(vs, 1) + WS(rs, 4)]; T1p = iio[WS(vs, 1) + WS(rs, 1)]; } { E T2z, T2V, T2r, T2A, T2s, T2t; { E T2p, T1x, T1S, T1q, T2q, T1y; T2p = iio[WS(vs, 2) + WS(rs, 2)]; T1X = T1v + T1w; T1x = T1v - T1w; T1S = T1o + T1p; T1q = T1o - T1p; T2q = iio[WS(vs, 2) + WS(rs, 5)]; T2z = iio[WS(vs, 2)]; T1Y = T1R + T1S; T1T = T1R - T1S; T1y = T1n + T1q; T1r = T1n - T1q; T2V = T2p + T2q; T2r = T2p - T2q; T1Z = FNMS(KP500000000, T1Y, T1X); T1N = T1x + T1y; T1z = FNMS(KP500000000, T1y, T1x); T2A = iio[WS(vs, 2) + WS(rs, 3)]; T2s = iio[WS(vs, 2) + WS(rs, 4)]; T2t = iio[WS(vs, 2) + WS(rs, 1)]; } { E T29, T2B, T2W, T2u, T2a, T2C; T29 = rio[WS(vs, 2)]; T31 = T2z + T2A; T2B = T2z - T2A; T2W = T2s + T2t; T2u = T2s - T2t; T2a = rio[WS(vs, 2) + WS(rs, 3)]; T2f = rio[WS(vs, 2) + WS(rs, 4)]; T32 = T2V + T2W; T2X = T2V - T2W; T2C = T2r + T2u; T2v = T2r - T2u; T2k = T29 - T2a; T2b = T29 + T2a; T33 = FNMS(KP500000000, T32, T31); T2R = T2B + T2C; T2D = FNMS(KP500000000, T2C, T2B); T2g = rio[WS(vs, 2) + WS(rs, 1)]; T2c = rio[WS(vs, 2) + WS(rs, 2)]; T2d = rio[WS(vs, 2) + WS(rs, 5)]; } } } } { E T4n, T4s, T4o, T4k, T4l; { E T3j, T3o, T3k, T3g, T3h; { E T3d, T2m, T2h, T2l, T2e, T3e, T2n; T3d = rio[WS(vs, 3)]; T2m = T2f - T2g; T2h = T2f + T2g; T2l = T2c - T2d; T2e = T2c + T2d; T3e = rio[WS(vs, 3) + WS(rs, 3)]; T3j = rio[WS(vs, 3) + WS(rs, 4)]; T2n = T2l + T2m; T2E = T2m - T2l; T2i = T2e + T2h; T34 = T2h - T2e; T3o = T3d - T3e; T3f = T3d + T3e; T2o = FNMS(KP500000000, T2n, T2k); T2O = T2k + T2n; T2U = FNMS(KP500000000, T2i, T2b); T3k = rio[WS(vs, 3) + WS(rs, 1)]; T3g = rio[WS(vs, 3) + WS(rs, 2)]; T3h = rio[WS(vs, 3) + WS(rs, 5)]; } { E T3D, T3Z, T3v, T3E, T3w, T3x; { E T3t, T3q, T3l, T3p, T3i, T3u, T3r; T3t = iio[WS(vs, 3) + WS(rs, 2)]; T3q = T3j - T3k; T3l = T3j + T3k; T3p = T3g - T3h; T3i = T3g + T3h; T3u = iio[WS(vs, 3) + WS(rs, 5)]; T3D = iio[WS(vs, 3)]; T3r = T3p + T3q; T3I = T3q - T3p; T3m = T3i + T3l; T48 = T3l - T3i; T3Z = T3t + T3u; T3v = T3t - T3u; T3s = FNMS(KP500000000, T3r, T3o); T3S = T3o + T3r; T3Y = FNMS(KP500000000, T3m, T3f); T3E = iio[WS(vs, 3) + WS(rs, 3)]; T3w = iio[WS(vs, 3) + WS(rs, 4)]; T3x = iio[WS(vs, 3) + WS(rs, 1)]; } { E T4h, T3F, T40, T3y, T4i, T3G; T4h = rio[WS(vs, 4)]; T45 = T3D + T3E; T3F = T3D - T3E; T40 = T3w + T3x; T3y = T3w - T3x; T4i = rio[WS(vs, 4) + WS(rs, 3)]; T4n = rio[WS(vs, 4) + WS(rs, 4)]; T46 = T3Z + T40; T41 = T3Z - T40; T3G = T3v + T3y; T3z = T3v - T3y; T4s = T4h - T4i; T4j = T4h + T4i; T47 = FNMS(KP500000000, T46, T45); T3V = T3F + T3G; T3H = FNMS(KP500000000, T3G, T3F); T4o = rio[WS(vs, 4) + WS(rs, 1)]; T4k = rio[WS(vs, 4) + WS(rs, 2)]; T4l = rio[WS(vs, 4) + WS(rs, 5)]; } } } { E T4H, T53, T4z, T4I, T4A, T4B; { E T4x, T4u, T4p, T4t, T4m, T4y, T4v; T4x = iio[WS(vs, 4) + WS(rs, 2)]; T4u = T4n - T4o; T4p = T4n + T4o; T4t = T4k - T4l; T4m = T4k + T4l; T4y = iio[WS(vs, 4) + WS(rs, 5)]; T4H = iio[WS(vs, 4)]; T4v = T4t + T4u; T4M = T4u - T4t; T4q = T4m + T4p; T5c = T4p - T4m; T53 = T4x + T4y; T4z = T4x - T4y; T4w = FNMS(KP500000000, T4v, T4s); T4W = T4s + T4v; T52 = FNMS(KP500000000, T4q, T4j); T4I = iio[WS(vs, 4) + WS(rs, 3)]; T4A = iio[WS(vs, 4) + WS(rs, 4)]; T4B = iio[WS(vs, 4) + WS(rs, 1)]; } { E T5L, T67, T5D, T5M, T5E, T5F; { E T5B, T4J, T54, T4C, T5C, T4K; T5B = iio[WS(vs, 5) + WS(rs, 2)]; T59 = T4H + T4I; T4J = T4H - T4I; T54 = T4A + T4B; T4C = T4A - T4B; T5C = iio[WS(vs, 5) + WS(rs, 5)]; T5L = iio[WS(vs, 5)]; T5a = T53 + T54; T55 = T53 - T54; T4K = T4z + T4C; T4D = T4z - T4C; T67 = T5B + T5C; T5D = T5B - T5C; T5b = FNMS(KP500000000, T5a, T59); T4Z = T4J + T4K; T4L = FNMS(KP500000000, T4K, T4J); T5M = iio[WS(vs, 5) + WS(rs, 3)]; T5E = iio[WS(vs, 5) + WS(rs, 4)]; T5F = iio[WS(vs, 5) + WS(rs, 1)]; } { E T5l, T5N, T68, T5G, T5m, T5O; T5l = rio[WS(vs, 5)]; T6d = T5L + T5M; T5N = T5L - T5M; T68 = T5E + T5F; T5G = T5E - T5F; T5m = rio[WS(vs, 5) + WS(rs, 3)]; T5r = rio[WS(vs, 5) + WS(rs, 4)]; T6e = T67 + T68; T69 = T67 - T68; T5O = T5D + T5G; T5H = T5D - T5G; T5w = T5l - T5m; T5n = T5l + T5m; T6f = FNMS(KP500000000, T6e, T6d); T63 = T5N + T5O; T5P = FNMS(KP500000000, T5O, T5N); T5s = rio[WS(vs, 5) + WS(rs, 1)]; T5o = rio[WS(vs, 5) + WS(rs, 2)]; T5p = rio[WS(vs, 5) + WS(rs, 5)]; } } } } } { E T6a, T6h, T5I, T5R, T65, T6c; { E T5Q, T5u, T6g, T5A, T60, T66; { E T5y, T5t, T5x, T5q, T5z; rio[0] = T3 + Ta; T5y = T5r - T5s; T5t = T5r + T5s; T5x = T5o - T5p; T5q = T5o + T5p; iio[0] = TT + TU; rio[WS(rs, 1)] = T17 + T1e; T5z = T5x + T5y; T5Q = T5y - T5x; T5u = T5q + T5t; T6g = T5t - T5q; T5A = FNMS(KP500000000, T5z, T5w); T60 = T5w + T5z; iio[WS(rs, 1)] = T1X + T1Y; T66 = FNMS(KP500000000, T5u, T5n); rio[WS(rs, 2)] = T2b + T2i; } iio[WS(rs, 2)] = T31 + T32; iio[WS(rs, 4)] = T59 + T5a; rio[WS(rs, 4)] = T4j + T4q; rio[WS(rs, 3)] = T3f + T3m; iio[WS(rs, 3)] = T45 + T46; { E TA, TD, TQ, T10, T13, TX, TZ, T12; rio[WS(rs, 5)] = T5n + T5u; iio[WS(rs, 5)] = T6d + T6e; { E To, Tx, Tb, Tq; TA = FNMS(KP866025403, Tn, Tg); To = FMA(KP866025403, Tn, Tg); Tx = FMA(KP866025403, Tw, Tv); TD = FNMS(KP866025403, Tw, Tv); Tb = W[0]; Tq = W[1]; { E TI, TK, TH, Ty, Tp, TF; Ty = Tb * Tx; Tp = Tb * To; TF = W[4]; TI = W[5]; iio[WS(vs, 1)] = FNMS(Tq, To, Ty); rio[WS(vs, 1)] = FMA(Tq, Tx, Tp); TK = TF * TJ; TH = TF * TG; TQ = FNMS(KP866025403, TP, TM); T10 = FMA(KP866025403, TP, TM); T13 = FMA(KP866025403, TW, TV); TX = FNMS(KP866025403, TW, TV); iio[WS(vs, 3)] = FNMS(TI, TG, TK); rio[WS(vs, 3)] = FMA(TI, TJ, TH); TZ = W[6]; T12 = W[7]; } } { E TC, TE, TB, TL, TS; { E T62, T64, T61, T14, T11, T5Z; T14 = TZ * T13; T11 = TZ * T10; T5Z = W[4]; T62 = W[5]; iio[WS(vs, 4)] = FNMS(T12, T10, T14); rio[WS(vs, 4)] = FMA(T12, T13, T11); T64 = T5Z * T63; T61 = T5Z * T60; { E T6k, T6n, T6j, T6m, T6o, T6l, Tz; T6a = FNMS(KP866025403, T69, T66); T6k = FMA(KP866025403, T69, T66); T6n = FMA(KP866025403, T6g, T6f); T6h = FNMS(KP866025403, T6g, T6f); iio[WS(vs, 3) + WS(rs, 5)] = FNMS(T62, T60, T64); rio[WS(vs, 3) + WS(rs, 5)] = FMA(T62, T63, T61); T6j = W[6]; T6m = W[7]; T6o = T6j * T6n; T6l = T6j * T6k; Tz = W[8]; TC = W[9]; iio[WS(vs, 4) + WS(rs, 5)] = FNMS(T6m, T6k, T6o); rio[WS(vs, 4) + WS(rs, 5)] = FMA(T6m, T6n, T6l); TE = Tz * TD; TB = Tz * TA; } } iio[WS(vs, 5)] = FNMS(TC, TA, TE); rio[WS(vs, 5)] = FMA(TC, TD, TB); TL = W[2]; TS = W[3]; { E T5U, T5X, T5W, T5Y, T5V, TY, TR, T5T; T5I = FMA(KP866025403, T5H, T5A); T5U = FNMS(KP866025403, T5H, T5A); T5X = FNMS(KP866025403, T5Q, T5P); T5R = FMA(KP866025403, T5Q, T5P); TY = TL * TX; TR = TL * TQ; T5T = W[8]; T5W = W[9]; iio[WS(vs, 2)] = FNMS(TS, TQ, TY); rio[WS(vs, 2)] = FMA(TS, TX, TR); T5Y = T5T * T5X; T5V = T5T * T5U; iio[WS(vs, 5) + WS(rs, 5)] = FNMS(T5W, T5U, T5Y); rio[WS(vs, 5) + WS(rs, 5)] = FMA(T5W, T5X, T5V); T65 = W[2]; T6c = W[3]; } } } } { E T5g, T5j, T5f, T5i; { E T1E, T1H, T3M, T3P, T56, T5d, T58, T5e, T57; { E T1s, T1B, T1f, T1u; { E T5K, T5S, T5J, T6i, T6b, T5v; T6i = T65 * T6h; T6b = T65 * T6a; T5v = W[0]; T5K = W[1]; iio[WS(vs, 2) + WS(rs, 5)] = FNMS(T6c, T6a, T6i); rio[WS(vs, 2) + WS(rs, 5)] = FMA(T6c, T6h, T6b); T5S = T5v * T5R; T5J = T5v * T5I; T1E = FNMS(KP866025403, T1r, T1k); T1s = FMA(KP866025403, T1r, T1k); T1B = FMA(KP866025403, T1A, T1z); T1H = FNMS(KP866025403, T1A, T1z); iio[WS(vs, 1) + WS(rs, 5)] = FNMS(T5K, T5I, T5S); rio[WS(vs, 1) + WS(rs, 5)] = FMA(T5K, T5R, T5J); T1f = W[0]; T1u = W[1]; } { E T3U, T3W, T3T, T1C, T1t, T3R; T1C = T1f * T1B; T1t = T1f * T1s; T3R = W[4]; T3U = W[5]; iio[WS(vs, 1) + WS(rs, 1)] = FNMS(T1u, T1s, T1C); rio[WS(vs, 1) + WS(rs, 1)] = FMA(T1u, T1B, T1t); T3W = T3R * T3V; T3T = T3R * T3S; { E T3A, T3J, T3n, T3C, T3K, T3B, T51; T3M = FNMS(KP866025403, T3z, T3s); T3A = FMA(KP866025403, T3z, T3s); T3J = FMA(KP866025403, T3I, T3H); T3P = FNMS(KP866025403, T3I, T3H); iio[WS(vs, 3) + WS(rs, 3)] = FNMS(T3U, T3S, T3W); rio[WS(vs, 3) + WS(rs, 3)] = FMA(T3U, T3V, T3T); T3n = W[0]; T3C = W[1]; T5g = FMA(KP866025403, T55, T52); T56 = FNMS(KP866025403, T55, T52); T5d = FNMS(KP866025403, T5c, T5b); T5j = FMA(KP866025403, T5c, T5b); T3K = T3n * T3J; T3B = T3n * T3A; T51 = W[2]; T58 = W[3]; iio[WS(vs, 1) + WS(rs, 3)] = FNMS(T3C, T3A, T3K); rio[WS(vs, 1) + WS(rs, 3)] = FMA(T3C, T3J, T3B); T5e = T51 * T5d; T57 = T51 * T56; } } } { E T38, T3b, T3O, T3Q, T3N, T37, T3a; { E T2Y, T35, T2T, T30, T36, T2Z, T3L; T38 = FMA(KP866025403, T2X, T2U); T2Y = FNMS(KP866025403, T2X, T2U); T35 = FNMS(KP866025403, T34, T33); T3b = FMA(KP866025403, T34, T33); iio[WS(vs, 2) + WS(rs, 4)] = FNMS(T58, T56, T5e); rio[WS(vs, 2) + WS(rs, 4)] = FMA(T58, T5d, T57); T2T = W[2]; T30 = W[3]; T36 = T2T * T35; T2Z = T2T * T2Y; T3L = W[8]; T3O = W[9]; iio[WS(vs, 2) + WS(rs, 2)] = FNMS(T30, T2Y, T36); rio[WS(vs, 2) + WS(rs, 2)] = FMA(T30, T35, T2Z); T3Q = T3L * T3P; T3N = T3L * T3M; } iio[WS(vs, 5) + WS(rs, 3)] = FNMS(T3O, T3M, T3Q); rio[WS(vs, 5) + WS(rs, 3)] = FMA(T3O, T3P, T3N); T37 = W[6]; T3a = W[7]; { E T1G, T1I, T1F, T3c, T39, T1D; T3c = T37 * T3b; T39 = T37 * T38; T1D = W[8]; T1G = W[9]; iio[WS(vs, 4) + WS(rs, 2)] = FNMS(T3a, T38, T3c); rio[WS(vs, 4) + WS(rs, 2)] = FMA(T3a, T3b, T39); T1I = T1D * T1H; T1F = T1D * T1E; iio[WS(vs, 5) + WS(rs, 1)] = FNMS(T1G, T1E, T1I); rio[WS(vs, 5) + WS(rs, 1)] = FMA(T1G, T1H, T1F); T5f = W[6]; T5i = W[7]; } } } { E T4Q, T4T, T2I, T2w, T2F, T2L, T2y, T2G, T2x, T4V, T4Y; { E T1M, T1O, T1L, T5k, T5h, T1J; T5k = T5f * T5j; T5h = T5f * T5g; T1J = W[4]; T1M = W[5]; iio[WS(vs, 4) + WS(rs, 4)] = FNMS(T5i, T5g, T5k); rio[WS(vs, 4) + WS(rs, 4)] = FMA(T5i, T5j, T5h); T1O = T1J * T1N; T1L = T1J * T1K; iio[WS(vs, 3) + WS(rs, 1)] = FNMS(T1M, T1K, T1O); rio[WS(vs, 3) + WS(rs, 1)] = FMA(T1M, T1N, T1L); T4V = W[4]; T4Y = W[5]; } { E T4E, T4N, T4G, T4O, T4F, T50, T4X, T4r; T4Q = FNMS(KP866025403, T4D, T4w); T4E = FMA(KP866025403, T4D, T4w); T4N = FMA(KP866025403, T4M, T4L); T4T = FNMS(KP866025403, T4M, T4L); T50 = T4V * T4Z; T4X = T4V * T4W; T4r = W[0]; T4G = W[1]; iio[WS(vs, 3) + WS(rs, 4)] = FNMS(T4Y, T4W, T50); rio[WS(vs, 3) + WS(rs, 4)] = FMA(T4Y, T4Z, T4X); T4O = T4r * T4N; T4F = T4r * T4E; { E T2N, T2Q, T2S, T2P, T2j; iio[WS(vs, 1) + WS(rs, 4)] = FNMS(T4G, T4E, T4O); rio[WS(vs, 1) + WS(rs, 4)] = FMA(T4G, T4N, T4F); T2N = W[4]; T2Q = W[5]; T2I = FNMS(KP866025403, T2v, T2o); T2w = FMA(KP866025403, T2v, T2o); T2F = FMA(KP866025403, T2E, T2D); T2L = FNMS(KP866025403, T2E, T2D); T2S = T2N * T2R; T2P = T2N * T2O; T2j = W[0]; T2y = W[1]; iio[WS(vs, 3) + WS(rs, 2)] = FNMS(T2Q, T2O, T2S); rio[WS(vs, 3) + WS(rs, 2)] = FMA(T2Q, T2R, T2P); T2G = T2j * T2F; T2x = T2j * T2w; } } { E T1U, T21, T2H, T2K; { E T24, T27, T23, T26; T1U = FNMS(KP866025403, T1T, T1Q); T24 = FMA(KP866025403, T1T, T1Q); T27 = FMA(KP866025403, T20, T1Z); T21 = FNMS(KP866025403, T20, T1Z); iio[WS(vs, 1) + WS(rs, 2)] = FNMS(T2y, T2w, T2G); rio[WS(vs, 1) + WS(rs, 2)] = FMA(T2y, T2F, T2x); T23 = W[6]; T26 = W[7]; { E T42, T49, T44, T4a, T43, T28, T25, T3X; T4c = FMA(KP866025403, T41, T3Y); T42 = FNMS(KP866025403, T41, T3Y); T49 = FNMS(KP866025403, T48, T47); T4f = FMA(KP866025403, T48, T47); T28 = T23 * T27; T25 = T23 * T24; T3X = W[2]; T44 = W[3]; iio[WS(vs, 4) + WS(rs, 1)] = FNMS(T26, T24, T28); rio[WS(vs, 4) + WS(rs, 1)] = FMA(T26, T27, T25); T4a = T3X * T49; T43 = T3X * T42; iio[WS(vs, 2) + WS(rs, 3)] = FNMS(T44, T42, T4a); rio[WS(vs, 2) + WS(rs, 3)] = FMA(T44, T49, T43); T2H = W[8]; T2K = W[9]; } } { E T4S, T4U, T4R, T2M, T2J, T4P; T2M = T2H * T2L; T2J = T2H * T2I; T4P = W[8]; T4S = W[9]; iio[WS(vs, 5) + WS(rs, 2)] = FNMS(T2K, T2I, T2M); rio[WS(vs, 5) + WS(rs, 2)] = FMA(T2K, T2L, T2J); T4U = T4P * T4T; T4R = T4P * T4Q; { E T1P, T1W, T22, T1V, T4b; iio[WS(vs, 5) + WS(rs, 4)] = FNMS(T4S, T4Q, T4U); rio[WS(vs, 5) + WS(rs, 4)] = FMA(T4S, T4T, T4R); T1P = W[2]; T1W = W[3]; T22 = T1P * T21; T1V = T1P * T1U; T4b = W[6]; T4e = W[7]; iio[WS(vs, 2) + WS(rs, 1)] = FNMS(T1W, T1U, T22); rio[WS(vs, 2) + WS(rs, 1)] = FMA(T1W, T21, T1V); T4g = T4b * T4f; T4d = T4b * T4c; } } } } } } } iio[WS(vs, 4) + WS(rs, 3)] = FNMS(T4e, T4c, T4g); rio[WS(vs, 4) + WS(rs, 3)] = FMA(T4e, T4f, T4d); } } } static const tw_instr twinstr[] = { {TW_FULL, 0, 6}, {TW_NEXT, 1, 0} }; static const ct_desc desc = { 6, "q1_6", twinstr, &GENUS, {144, 60, 132, 0}, 0, 0, 0 }; void X(codelet_q1_6) (planner *p) { X(kdft_difsq_register) (p, q1_6, &desc); } #else /* HAVE_FMA */ /* Generated by: ../../../genfft/gen_twidsq.native -compact -variables 4 -pipeline-latency 4 -reload-twiddle -dif -n 6 -name q1_6 -include q.h */ /* * This function contains 276 FP additions, 168 FP multiplications, * (or, 192 additions, 84 multiplications, 84 fused multiply/add), * 85 stack variables, 2 constants, and 144 memory accesses */ #include "q.h" static void q1_6(R *rio, R *iio, const R *W, stride rs, stride vs, INT mb, INT me, INT ms) { DK(KP500000000, +0.500000000000000000000000000000000000000000000); DK(KP866025403, +0.866025403784438646763723170752936183471402627); { INT m; for (m = mb, W = W + (mb * 10); m < me; m = m + 1, rio = rio + ms, iio = iio + ms, W = W + 10, MAKE_VOLATILE_STRIDE(12, rs), MAKE_VOLATILE_STRIDE(0, vs)) { E T3, Tc, Tt, TM, TX, T16, T1n, T1G, T2h, T2A, T1R, T20, T2L, T2U, T3b; E T3u, T3F, T3O, T45, T4o, T4Z, T5i, T4z, T4I, Ta, TP, Tf, Tq, Tn, TN; E Tu, TJ, T14, T1J, T19, T1k, T1h, T1H, T1o, T1D, T2b, T2B, T2i, T2x, T1Y; E T2D, T23, T2e, T2S, T3x, T2X, T38, T35, T3v, T3c, T3r, T3M, T4r, T3R, T42; E T3Z, T4p, T46, T4l, T4T, T5j, T50, T5f, T4G, T5l, T4L, T4W; { E T1, T2, T1l, T1m; T1 = rio[0]; T2 = rio[WS(rs, 3)]; T3 = T1 + T2; Tc = T1 - T2; { E Tr, Ts, TV, TW; Tr = iio[0]; Ts = iio[WS(rs, 3)]; Tt = Tr - Ts; TM = Tr + Ts; TV = rio[WS(vs, 1)]; TW = rio[WS(vs, 1) + WS(rs, 3)]; TX = TV + TW; T16 = TV - TW; } T1l = iio[WS(vs, 1)]; T1m = iio[WS(vs, 1) + WS(rs, 3)]; T1n = T1l - T1m; T1G = T1l + T1m; { E T2f, T2g, T1P, T1Q; T2f = iio[WS(vs, 2)]; T2g = iio[WS(vs, 2) + WS(rs, 3)]; T2h = T2f - T2g; T2A = T2f + T2g; T1P = rio[WS(vs, 2)]; T1Q = rio[WS(vs, 2) + WS(rs, 3)]; T1R = T1P + T1Q; T20 = T1P - T1Q; } } { E T2J, T2K, T43, T44; T2J = rio[WS(vs, 3)]; T2K = rio[WS(vs, 3) + WS(rs, 3)]; T2L = T2J + T2K; T2U = T2J - T2K; { E T39, T3a, T3D, T3E; T39 = iio[WS(vs, 3)]; T3a = iio[WS(vs, 3) + WS(rs, 3)]; T3b = T39 - T3a; T3u = T39 + T3a; T3D = rio[WS(vs, 4)]; T3E = rio[WS(vs, 4) + WS(rs, 3)]; T3F = T3D + T3E; T3O = T3D - T3E; } T43 = iio[WS(vs, 4)]; T44 = iio[WS(vs, 4) + WS(rs, 3)]; T45 = T43 - T44; T4o = T43 + T44; { E T4X, T4Y, T4x, T4y; T4X = iio[WS(vs, 5)]; T4Y = iio[WS(vs, 5) + WS(rs, 3)]; T4Z = T4X - T4Y; T5i = T4X + T4Y; T4x = rio[WS(vs, 5)]; T4y = rio[WS(vs, 5) + WS(rs, 3)]; T4z = T4x + T4y; T4I = T4x - T4y; } } { E T6, Td, T9, Te; { E T4, T5, T7, T8; T4 = rio[WS(rs, 2)]; T5 = rio[WS(rs, 5)]; T6 = T4 + T5; Td = T4 - T5; T7 = rio[WS(rs, 4)]; T8 = rio[WS(rs, 1)]; T9 = T7 + T8; Te = T7 - T8; } Ta = T6 + T9; TP = KP866025403 * (T9 - T6); Tf = Td + Te; Tq = KP866025403 * (Te - Td); } { E Tj, TH, Tm, TI; { E Th, Ti, Tk, Tl; Th = iio[WS(rs, 2)]; Ti = iio[WS(rs, 5)]; Tj = Th - Ti; TH = Th + Ti; Tk = iio[WS(rs, 4)]; Tl = iio[WS(rs, 1)]; Tm = Tk - Tl; TI = Tk + Tl; } Tn = KP866025403 * (Tj - Tm); TN = TH + TI; Tu = Tj + Tm; TJ = KP866025403 * (TH - TI); } { E T10, T17, T13, T18; { E TY, TZ, T11, T12; TY = rio[WS(vs, 1) + WS(rs, 2)]; TZ = rio[WS(vs, 1) + WS(rs, 5)]; T10 = TY + TZ; T17 = TY - TZ; T11 = rio[WS(vs, 1) + WS(rs, 4)]; T12 = rio[WS(vs, 1) + WS(rs, 1)]; T13 = T11 + T12; T18 = T11 - T12; } T14 = T10 + T13; T1J = KP866025403 * (T13 - T10); T19 = T17 + T18; T1k = KP866025403 * (T18 - T17); } { E T1d, T1B, T1g, T1C; { E T1b, T1c, T1e, T1f; T1b = iio[WS(vs, 1) + WS(rs, 2)]; T1c = iio[WS(vs, 1) + WS(rs, 5)]; T1d = T1b - T1c; T1B = T1b + T1c; T1e = iio[WS(vs, 1) + WS(rs, 4)]; T1f = iio[WS(vs, 1) + WS(rs, 1)]; T1g = T1e - T1f; T1C = T1e + T1f; } T1h = KP866025403 * (T1d - T1g); T1H = T1B + T1C; T1o = T1d + T1g; T1D = KP866025403 * (T1B - T1C); } { E T27, T2v, T2a, T2w; { E T25, T26, T28, T29; T25 = iio[WS(vs, 2) + WS(rs, 2)]; T26 = iio[WS(vs, 2) + WS(rs, 5)]; T27 = T25 - T26; T2v = T25 + T26; T28 = iio[WS(vs, 2) + WS(rs, 4)]; T29 = iio[WS(vs, 2) + WS(rs, 1)]; T2a = T28 - T29; T2w = T28 + T29; } T2b = KP866025403 * (T27 - T2a); T2B = T2v + T2w; T2i = T27 + T2a; T2x = KP866025403 * (T2v - T2w); } { E T1U, T21, T1X, T22; { E T1S, T1T, T1V, T1W; T1S = rio[WS(vs, 2) + WS(rs, 2)]; T1T = rio[WS(vs, 2) + WS(rs, 5)]; T1U = T1S + T1T; T21 = T1S - T1T; T1V = rio[WS(vs, 2) + WS(rs, 4)]; T1W = rio[WS(vs, 2) + WS(rs, 1)]; T1X = T1V + T1W; T22 = T1V - T1W; } T1Y = T1U + T1X; T2D = KP866025403 * (T1X - T1U); T23 = T21 + T22; T2e = KP866025403 * (T22 - T21); } { E T2O, T2V, T2R, T2W; { E T2M, T2N, T2P, T2Q; T2M = rio[WS(vs, 3) + WS(rs, 2)]; T2N = rio[WS(vs, 3) + WS(rs, 5)]; T2O = T2M + T2N; T2V = T2M - T2N; T2P = rio[WS(vs, 3) + WS(rs, 4)]; T2Q = rio[WS(vs, 3) + WS(rs, 1)]; T2R = T2P + T2Q; T2W = T2P - T2Q; } T2S = T2O + T2R; T3x = KP866025403 * (T2R - T2O); T2X = T2V + T2W; T38 = KP866025403 * (T2W - T2V); } { E T31, T3p, T34, T3q; { E T2Z, T30, T32, T33; T2Z = iio[WS(vs, 3) + WS(rs, 2)]; T30 = iio[WS(vs, 3) + WS(rs, 5)]; T31 = T2Z - T30; T3p = T2Z + T30; T32 = iio[WS(vs, 3) + WS(rs, 4)]; T33 = iio[WS(vs, 3) + WS(rs, 1)]; T34 = T32 - T33; T3q = T32 + T33; } T35 = KP866025403 * (T31 - T34); T3v = T3p + T3q; T3c = T31 + T34; T3r = KP866025403 * (T3p - T3q); } { E T3I, T3P, T3L, T3Q; { E T3G, T3H, T3J, T3K; T3G = rio[WS(vs, 4) + WS(rs, 2)]; T3H = rio[WS(vs, 4) + WS(rs, 5)]; T3I = T3G + T3H; T3P = T3G - T3H; T3J = rio[WS(vs, 4) + WS(rs, 4)]; T3K = rio[WS(vs, 4) + WS(rs, 1)]; T3L = T3J + T3K; T3Q = T3J - T3K; } T3M = T3I + T3L; T4r = KP866025403 * (T3L - T3I); T3R = T3P + T3Q; T42 = KP866025403 * (T3Q - T3P); } { E T3V, T4j, T3Y, T4k; { E T3T, T3U, T3W, T3X; T3T = iio[WS(vs, 4) + WS(rs, 2)]; T3U = iio[WS(vs, 4) + WS(rs, 5)]; T3V = T3T - T3U; T4j = T3T + T3U; T3W = iio[WS(vs, 4) + WS(rs, 4)]; T3X = iio[WS(vs, 4) + WS(rs, 1)]; T3Y = T3W - T3X; T4k = T3W + T3X; } T3Z = KP866025403 * (T3V - T3Y); T4p = T4j + T4k; T46 = T3V + T3Y; T4l = KP866025403 * (T4j - T4k); } { E T4P, T5d, T4S, T5e; { E T4N, T4O, T4Q, T4R; T4N = iio[WS(vs, 5) + WS(rs, 2)]; T4O = iio[WS(vs, 5) + WS(rs, 5)]; T4P = T4N - T4O; T5d = T4N + T4O; T4Q = iio[WS(vs, 5) + WS(rs, 4)]; T4R = iio[WS(vs, 5) + WS(rs, 1)]; T4S = T4Q - T4R; T5e = T4Q + T4R; } T4T = KP866025403 * (T4P - T4S); T5j = T5d + T5e; T50 = T4P + T4S; T5f = KP866025403 * (T5d - T5e); } { E T4C, T4J, T4F, T4K; { E T4A, T4B, T4D, T4E; T4A = rio[WS(vs, 5) + WS(rs, 2)]; T4B = rio[WS(vs, 5) + WS(rs, 5)]; T4C = T4A + T4B; T4J = T4A - T4B; T4D = rio[WS(vs, 5) + WS(rs, 4)]; T4E = rio[WS(vs, 5) + WS(rs, 1)]; T4F = T4D + T4E; T4K = T4D - T4E; } T4G = T4C + T4F; T5l = KP866025403 * (T4F - T4C); T4L = T4J + T4K; T4W = KP866025403 * (T4K - T4J); } rio[0] = T3 + Ta; iio[0] = TM + TN; rio[WS(rs, 1)] = TX + T14; iio[WS(rs, 1)] = T1G + T1H; rio[WS(rs, 3)] = T2L + T2S; rio[WS(rs, 2)] = T1R + T1Y; iio[WS(rs, 2)] = T2A + T2B; iio[WS(rs, 3)] = T3u + T3v; iio[WS(rs, 4)] = T4o + T4p; iio[WS(rs, 5)] = T5i + T5j; rio[WS(rs, 5)] = T4z + T4G; rio[WS(rs, 4)] = T3F + T3M; { E T1w, T1y, T1v, T1x; T1w = T16 + T19; T1y = T1n + T1o; T1v = W[4]; T1x = W[5]; rio[WS(vs, 3) + WS(rs, 1)] = FMA(T1v, T1w, T1x * T1y); iio[WS(vs, 3) + WS(rs, 1)] = FNMS(T1x, T1w, T1v * T1y); } { E T58, T5a, T57, T59; T58 = T4I + T4L; T5a = T4Z + T50; T57 = W[4]; T59 = W[5]; rio[WS(vs, 3) + WS(rs, 5)] = FMA(T57, T58, T59 * T5a); iio[WS(vs, 3) + WS(rs, 5)] = FNMS(T59, T58, T57 * T5a); } { E TC, TE, TB, TD; TC = Tc + Tf; TE = Tt + Tu; TB = W[4]; TD = W[5]; rio[WS(vs, 3)] = FMA(TB, TC, TD * TE); iio[WS(vs, 3)] = FNMS(TD, TC, TB * TE); } { E T4e, T4g, T4d, T4f; T4e = T3O + T3R; T4g = T45 + T46; T4d = W[4]; T4f = W[5]; rio[WS(vs, 3) + WS(rs, 4)] = FMA(T4d, T4e, T4f * T4g); iio[WS(vs, 3) + WS(rs, 4)] = FNMS(T4f, T4e, T4d * T4g); } { E T3k, T3m, T3j, T3l; T3k = T2U + T2X; T3m = T3b + T3c; T3j = W[4]; T3l = W[5]; rio[WS(vs, 3) + WS(rs, 3)] = FMA(T3j, T3k, T3l * T3m); iio[WS(vs, 3) + WS(rs, 3)] = FNMS(T3l, T3k, T3j * T3m); } { E T2q, T2s, T2p, T2r; T2q = T20 + T23; T2s = T2h + T2i; T2p = W[4]; T2r = W[5]; rio[WS(vs, 3) + WS(rs, 2)] = FMA(T2p, T2q, T2r * T2s); iio[WS(vs, 3) + WS(rs, 2)] = FNMS(T2r, T2q, T2p * T2s); } { E T5g, T5o, T5m, T5q, T5c, T5k; T5c = FNMS(KP500000000, T4G, T4z); T5g = T5c - T5f; T5o = T5c + T5f; T5k = FNMS(KP500000000, T5j, T5i); T5m = T5k - T5l; T5q = T5l + T5k; { E T5b, T5h, T5n, T5p; T5b = W[2]; T5h = W[3]; rio[WS(vs, 2) + WS(rs, 5)] = FMA(T5b, T5g, T5h * T5m); iio[WS(vs, 2) + WS(rs, 5)] = FNMS(T5h, T5g, T5b * T5m); T5n = W[6]; T5p = W[7]; rio[WS(vs, 4) + WS(rs, 5)] = FMA(T5n, T5o, T5p * T5q); iio[WS(vs, 4) + WS(rs, 5)] = FNMS(T5p, T5o, T5n * T5q); } } { E To, Ty, Tw, TA, Tg, Tv; Tg = FNMS(KP500000000, Tf, Tc); To = Tg + Tn; Ty = Tg - Tn; Tv = FNMS(KP500000000, Tu, Tt); Tw = Tq + Tv; TA = Tv - Tq; { E Tb, Tp, Tx, Tz; Tb = W[0]; Tp = W[1]; rio[WS(vs, 1)] = FMA(Tb, To, Tp * Tw); iio[WS(vs, 1)] = FNMS(Tp, To, Tb * Tw); Tx = W[8]; Tz = W[9]; rio[WS(vs, 5)] = FMA(Tx, Ty, Tz * TA); iio[WS(vs, 5)] = FNMS(Tz, Ty, Tx * TA); } } { E T36, T3g, T3e, T3i, T2Y, T3d; T2Y = FNMS(KP500000000, T2X, T2U); T36 = T2Y + T35; T3g = T2Y - T35; T3d = FNMS(KP500000000, T3c, T3b); T3e = T38 + T3d; T3i = T3d - T38; { E T2T, T37, T3f, T3h; T2T = W[0]; T37 = W[1]; rio[WS(vs, 1) + WS(rs, 3)] = FMA(T2T, T36, T37 * T3e); iio[WS(vs, 1) + WS(rs, 3)] = FNMS(T37, T36, T2T * T3e); T3f = W[8]; T3h = W[9]; rio[WS(vs, 5) + WS(rs, 3)] = FMA(T3f, T3g, T3h * T3i); iio[WS(vs, 5) + WS(rs, 3)] = FNMS(T3h, T3g, T3f * T3i); } } { E T2y, T2G, T2E, T2I, T2u, T2C; T2u = FNMS(KP500000000, T1Y, T1R); T2y = T2u - T2x; T2G = T2u + T2x; T2C = FNMS(KP500000000, T2B, T2A); T2E = T2C - T2D; T2I = T2D + T2C; { E T2t, T2z, T2F, T2H; T2t = W[2]; T2z = W[3]; rio[WS(vs, 2) + WS(rs, 2)] = FMA(T2t, T2y, T2z * T2E); iio[WS(vs, 2) + WS(rs, 2)] = FNMS(T2z, T2y, T2t * T2E); T2F = W[6]; T2H = W[7]; rio[WS(vs, 4) + WS(rs, 2)] = FMA(T2F, T2G, T2H * T2I); iio[WS(vs, 4) + WS(rs, 2)] = FNMS(T2H, T2G, T2F * T2I); } } { E T3s, T3A, T3y, T3C, T3o, T3w; T3o = FNMS(KP500000000, T2S, T2L); T3s = T3o - T3r; T3A = T3o + T3r; T3w = FNMS(KP500000000, T3v, T3u); T3y = T3w - T3x; T3C = T3x + T3w; { E T3n, T3t, T3z, T3B; T3n = W[2]; T3t = W[3]; rio[WS(vs, 2) + WS(rs, 3)] = FMA(T3n, T3s, T3t * T3y); iio[WS(vs, 2) + WS(rs, 3)] = FNMS(T3t, T3s, T3n * T3y); T3z = W[6]; T3B = W[7]; rio[WS(vs, 4) + WS(rs, 3)] = FMA(T3z, T3A, T3B * T3C); iio[WS(vs, 4) + WS(rs, 3)] = FNMS(T3B, T3A, T3z * T3C); } } { E T1E, T1M, T1K, T1O, T1A, T1I; T1A = FNMS(KP500000000, T14, TX); T1E = T1A - T1D; T1M = T1A + T1D; T1I = FNMS(KP500000000, T1H, T1G); T1K = T1I - T1J; T1O = T1J + T1I; { E T1z, T1F, T1L, T1N; T1z = W[2]; T1F = W[3]; rio[WS(vs, 2) + WS(rs, 1)] = FMA(T1z, T1E, T1F * T1K); iio[WS(vs, 2) + WS(rs, 1)] = FNMS(T1F, T1E, T1z * T1K); T1L = W[6]; T1N = W[7]; rio[WS(vs, 4) + WS(rs, 1)] = FMA(T1L, T1M, T1N * T1O); iio[WS(vs, 4) + WS(rs, 1)] = FNMS(T1N, T1M, T1L * T1O); } } { E T4m, T4u, T4s, T4w, T4i, T4q; T4i = FNMS(KP500000000, T3M, T3F); T4m = T4i - T4l; T4u = T4i + T4l; T4q = FNMS(KP500000000, T4p, T4o); T4s = T4q - T4r; T4w = T4r + T4q; { E T4h, T4n, T4t, T4v; T4h = W[2]; T4n = W[3]; rio[WS(vs, 2) + WS(rs, 4)] = FMA(T4h, T4m, T4n * T4s); iio[WS(vs, 2) + WS(rs, 4)] = FNMS(T4n, T4m, T4h * T4s); T4t = W[6]; T4v = W[7]; rio[WS(vs, 4) + WS(rs, 4)] = FMA(T4t, T4u, T4v * T4w); iio[WS(vs, 4) + WS(rs, 4)] = FNMS(T4v, T4u, T4t * T4w); } } { E TK, TS, TQ, TU, TG, TO; TG = FNMS(KP500000000, Ta, T3); TK = TG - TJ; TS = TG + TJ; TO = FNMS(KP500000000, TN, TM); TQ = TO - TP; TU = TP + TO; { E TF, TL, TR, TT; TF = W[2]; TL = W[3]; rio[WS(vs, 2)] = FMA(TF, TK, TL * TQ); iio[WS(vs, 2)] = FNMS(TL, TK, TF * TQ); TR = W[6]; TT = W[7]; rio[WS(vs, 4)] = FMA(TR, TS, TT * TU); iio[WS(vs, 4)] = FNMS(TT, TS, TR * TU); } } { E T2c, T2m, T2k, T2o, T24, T2j; T24 = FNMS(KP500000000, T23, T20); T2c = T24 + T2b; T2m = T24 - T2b; T2j = FNMS(KP500000000, T2i, T2h); T2k = T2e + T2j; T2o = T2j - T2e; { E T1Z, T2d, T2l, T2n; T1Z = W[0]; T2d = W[1]; rio[WS(vs, 1) + WS(rs, 2)] = FMA(T1Z, T2c, T2d * T2k); iio[WS(vs, 1) + WS(rs, 2)] = FNMS(T2d, T2c, T1Z * T2k); T2l = W[8]; T2n = W[9]; rio[WS(vs, 5) + WS(rs, 2)] = FMA(T2l, T2m, T2n * T2o); iio[WS(vs, 5) + WS(rs, 2)] = FNMS(T2n, T2m, T2l * T2o); } } { E T40, T4a, T48, T4c, T3S, T47; T3S = FNMS(KP500000000, T3R, T3O); T40 = T3S + T3Z; T4a = T3S - T3Z; T47 = FNMS(KP500000000, T46, T45); T48 = T42 + T47; T4c = T47 - T42; { E T3N, T41, T49, T4b; T3N = W[0]; T41 = W[1]; rio[WS(vs, 1) + WS(rs, 4)] = FMA(T3N, T40, T41 * T48); iio[WS(vs, 1) + WS(rs, 4)] = FNMS(T41, T40, T3N * T48); T49 = W[8]; T4b = W[9]; rio[WS(vs, 5) + WS(rs, 4)] = FMA(T49, T4a, T4b * T4c); iio[WS(vs, 5) + WS(rs, 4)] = FNMS(T4b, T4a, T49 * T4c); } } { E T1i, T1s, T1q, T1u, T1a, T1p; T1a = FNMS(KP500000000, T19, T16); T1i = T1a + T1h; T1s = T1a - T1h; T1p = FNMS(KP500000000, T1o, T1n); T1q = T1k + T1p; T1u = T1p - T1k; { E T15, T1j, T1r, T1t; T15 = W[0]; T1j = W[1]; rio[WS(vs, 1) + WS(rs, 1)] = FMA(T15, T1i, T1j * T1q); iio[WS(vs, 1) + WS(rs, 1)] = FNMS(T1j, T1i, T15 * T1q); T1r = W[8]; T1t = W[9]; rio[WS(vs, 5) + WS(rs, 1)] = FMA(T1r, T1s, T1t * T1u); iio[WS(vs, 5) + WS(rs, 1)] = FNMS(T1t, T1s, T1r * T1u); } } { E T4U, T54, T52, T56, T4M, T51; T4M = FNMS(KP500000000, T4L, T4I); T4U = T4M + T4T; T54 = T4M - T4T; T51 = FNMS(KP500000000, T50, T4Z); T52 = T4W + T51; T56 = T51 - T4W; { E T4H, T4V, T53, T55; T4H = W[0]; T4V = W[1]; rio[WS(vs, 1) + WS(rs, 5)] = FMA(T4H, T4U, T4V * T52); iio[WS(vs, 1) + WS(rs, 5)] = FNMS(T4V, T4U, T4H * T52); T53 = W[8]; T55 = W[9]; rio[WS(vs, 5) + WS(rs, 5)] = FMA(T53, T54, T55 * T56); iio[WS(vs, 5) + WS(rs, 5)] = FNMS(T55, T54, T53 * T56); } } } } } static const tw_instr twinstr[] = { {TW_FULL, 0, 6}, {TW_NEXT, 1, 0} }; static const ct_desc desc = { 6, "q1_6", twinstr, &GENUS, {192, 84, 84, 0}, 0, 0, 0 }; void X(codelet_q1_6) (planner *p) { X(kdft_difsq_register) (p, q1_6, &desc); } #endif /* HAVE_FMA */ fftw-3.3.3/dft/scalar/codelets/codlist.c0000644000175000017500000000705512054410311015036 00000000000000#include "ifftw.h" extern void X(codelet_n1_2)(planner *); extern void X(codelet_n1_3)(planner *); extern void X(codelet_n1_4)(planner *); extern void X(codelet_n1_5)(planner *); extern void X(codelet_n1_6)(planner *); extern void X(codelet_n1_7)(planner *); extern void X(codelet_n1_8)(planner *); extern void X(codelet_n1_9)(planner *); extern void X(codelet_n1_10)(planner *); extern void X(codelet_n1_11)(planner *); extern void X(codelet_n1_12)(planner *); extern void X(codelet_n1_13)(planner *); extern void X(codelet_n1_14)(planner *); extern void X(codelet_n1_15)(planner *); extern void X(codelet_n1_16)(planner *); extern void X(codelet_n1_32)(planner *); extern void X(codelet_n1_64)(planner *); extern void X(codelet_n1_20)(planner *); extern void X(codelet_n1_25)(planner *); extern void X(codelet_t1_2)(planner *); extern void X(codelet_t1_3)(planner *); extern void X(codelet_t1_4)(planner *); extern void X(codelet_t1_5)(planner *); extern void X(codelet_t1_6)(planner *); extern void X(codelet_t1_7)(planner *); extern void X(codelet_t1_8)(planner *); extern void X(codelet_t1_9)(planner *); extern void X(codelet_t1_10)(planner *); extern void X(codelet_t1_12)(planner *); extern void X(codelet_t1_15)(planner *); extern void X(codelet_t1_16)(planner *); extern void X(codelet_t1_32)(planner *); extern void X(codelet_t1_64)(planner *); extern void X(codelet_t1_20)(planner *); extern void X(codelet_t1_25)(planner *); extern void X(codelet_t2_4)(planner *); extern void X(codelet_t2_8)(planner *); extern void X(codelet_t2_16)(planner *); extern void X(codelet_t2_32)(planner *); extern void X(codelet_t2_64)(planner *); extern void X(codelet_t2_5)(planner *); extern void X(codelet_t2_10)(planner *); extern void X(codelet_t2_20)(planner *); extern void X(codelet_t2_25)(planner *); extern void X(codelet_q1_2)(planner *); extern void X(codelet_q1_4)(planner *); extern void X(codelet_q1_8)(planner *); extern void X(codelet_q1_3)(planner *); extern void X(codelet_q1_5)(planner *); extern void X(codelet_q1_6)(planner *); extern const solvtab X(solvtab_dft_standard); const solvtab X(solvtab_dft_standard) = { SOLVTAB(X(codelet_n1_2)), SOLVTAB(X(codelet_n1_3)), SOLVTAB(X(codelet_n1_4)), SOLVTAB(X(codelet_n1_5)), SOLVTAB(X(codelet_n1_6)), SOLVTAB(X(codelet_n1_7)), SOLVTAB(X(codelet_n1_8)), SOLVTAB(X(codelet_n1_9)), SOLVTAB(X(codelet_n1_10)), SOLVTAB(X(codelet_n1_11)), SOLVTAB(X(codelet_n1_12)), SOLVTAB(X(codelet_n1_13)), SOLVTAB(X(codelet_n1_14)), SOLVTAB(X(codelet_n1_15)), SOLVTAB(X(codelet_n1_16)), SOLVTAB(X(codelet_n1_32)), SOLVTAB(X(codelet_n1_64)), SOLVTAB(X(codelet_n1_20)), SOLVTAB(X(codelet_n1_25)), SOLVTAB(X(codelet_t1_2)), SOLVTAB(X(codelet_t1_3)), SOLVTAB(X(codelet_t1_4)), SOLVTAB(X(codelet_t1_5)), SOLVTAB(X(codelet_t1_6)), SOLVTAB(X(codelet_t1_7)), SOLVTAB(X(codelet_t1_8)), SOLVTAB(X(codelet_t1_9)), SOLVTAB(X(codelet_t1_10)), SOLVTAB(X(codelet_t1_12)), SOLVTAB(X(codelet_t1_15)), SOLVTAB(X(codelet_t1_16)), SOLVTAB(X(codelet_t1_32)), SOLVTAB(X(codelet_t1_64)), SOLVTAB(X(codelet_t1_20)), SOLVTAB(X(codelet_t1_25)), SOLVTAB(X(codelet_t2_4)), SOLVTAB(X(codelet_t2_8)), SOLVTAB(X(codelet_t2_16)), SOLVTAB(X(codelet_t2_32)), SOLVTAB(X(codelet_t2_64)), SOLVTAB(X(codelet_t2_5)), SOLVTAB(X(codelet_t2_10)), SOLVTAB(X(codelet_t2_20)), SOLVTAB(X(codelet_t2_25)), SOLVTAB(X(codelet_q1_2)), SOLVTAB(X(codelet_q1_4)), SOLVTAB(X(codelet_q1_8)), SOLVTAB(X(codelet_q1_3)), SOLVTAB(X(codelet_q1_5)), SOLVTAB(X(codelet_q1_6)), SOLVTAB_END }; fftw-3.3.3/dft/simd/0000755000175000017500000000000012054411107011173 500000000000000fftw-3.3.3/dft/simd/Makefile.am0000644000175000017500000000025212054410050013142 00000000000000SUBDIRS = common sse2 avx altivec neon EXTRA_DIST = n1b.h n1f.h n2b.h n2f.h n2s.h q1b.h q1f.h t1b.h t1bu.h \ t1f.h t1fu.h t2b.h t2f.h t3b.h t3f.h ts.h codlist.mk simd.mk fftw-3.3.3/dft/simd/Makefile.in0000644000175000017500000004447612054410177013205 00000000000000# Makefile.in generated by automake 1.11.6 from Makefile.am. # @configure_input@ # Copyright (C) 1994, 1995, 1996, 1997, 1998, 1999, 2000, 2001, 2002, # 2003, 2004, 2005, 2006, 2007, 2008, 2009, 2010, 2011 Free Software # Foundation, Inc. # This Makefile.in is free software; the Free Software Foundation # gives unlimited permission to copy and/or distribute it, # with or without modifications, as long as this notice is preserved. # This program is distributed in the hope that it will be useful, # but WITHOUT ANY WARRANTY, to the extent permitted by law; without # even the implied warranty of MERCHANTABILITY or FITNESS FOR A # PARTICULAR PURPOSE. @SET_MAKE@ VPATH = @srcdir@ am__make_dryrun = \ { \ am__dry=no; \ case $$MAKEFLAGS in \ *\\[\ \ ]*) \ echo 'am--echo: ; @echo "AM" OK' | $(MAKE) -f - 2>/dev/null \ | grep '^AM OK$$' >/dev/null || am__dry=yes;; \ *) \ for am__flg in $$MAKEFLAGS; do \ case $$am__flg in \ *=*|--*) ;; \ *n*) am__dry=yes; break;; \ esac; \ done;; \ esac; \ test $$am__dry = yes; \ } pkgdatadir = $(datadir)/@PACKAGE@ pkgincludedir = $(includedir)/@PACKAGE@ pkglibdir = $(libdir)/@PACKAGE@ pkglibexecdir = $(libexecdir)/@PACKAGE@ am__cd = CDPATH="$${ZSH_VERSION+.}$(PATH_SEPARATOR)" && cd install_sh_DATA = $(install_sh) -c -m 644 install_sh_PROGRAM = $(install_sh) -c install_sh_SCRIPT = $(install_sh) -c INSTALL_HEADER = $(INSTALL_DATA) transform = $(program_transform_name) NORMAL_INSTALL = : PRE_INSTALL = : POST_INSTALL = : NORMAL_UNINSTALL = : PRE_UNINSTALL = : POST_UNINSTALL = : build_triplet = @build@ host_triplet = @host@ subdir = dft/simd DIST_COMMON = $(srcdir)/Makefile.am $(srcdir)/Makefile.in ACLOCAL_M4 = $(top_srcdir)/aclocal.m4 am__aclocal_m4_deps = $(top_srcdir)/m4/acx_mpi.m4 \ $(top_srcdir)/m4/acx_pthread.m4 \ $(top_srcdir)/m4/ax_cc_maxopt.m4 \ $(top_srcdir)/m4/ax_check_compiler_flags.m4 \ $(top_srcdir)/m4/ax_compiler_vendor.m4 \ $(top_srcdir)/m4/ax_gcc_aligns_stack.m4 \ $(top_srcdir)/m4/ax_gcc_version.m4 \ $(top_srcdir)/m4/ax_openmp.m4 $(top_srcdir)/m4/libtool.m4 \ $(top_srcdir)/m4/ltoptions.m4 $(top_srcdir)/m4/ltsugar.m4 \ $(top_srcdir)/m4/ltversion.m4 $(top_srcdir)/m4/lt~obsolete.m4 \ $(top_srcdir)/configure.ac am__configure_deps = $(am__aclocal_m4_deps) $(CONFIGURE_DEPENDENCIES) \ $(ACLOCAL_M4) mkinstalldirs = $(install_sh) -d CONFIG_HEADER = $(top_builddir)/config.h CONFIG_CLEAN_FILES = CONFIG_CLEAN_VPATH_FILES = SOURCES = DIST_SOURCES = RECURSIVE_TARGETS = all-recursive check-recursive dvi-recursive \ html-recursive info-recursive install-data-recursive \ install-dvi-recursive install-exec-recursive \ install-html-recursive install-info-recursive \ install-pdf-recursive install-ps-recursive install-recursive \ installcheck-recursive installdirs-recursive pdf-recursive \ ps-recursive uninstall-recursive am__can_run_installinfo = \ case $$AM_UPDATE_INFO_DIR in \ n|no|NO) false;; \ *) (install-info --version) >/dev/null 2>&1;; \ esac RECURSIVE_CLEAN_TARGETS = mostlyclean-recursive clean-recursive \ distclean-recursive maintainer-clean-recursive AM_RECURSIVE_TARGETS = $(RECURSIVE_TARGETS:-recursive=) \ $(RECURSIVE_CLEAN_TARGETS:-recursive=) tags TAGS ctags CTAGS \ distdir ETAGS = etags CTAGS = ctags DIST_SUBDIRS = $(SUBDIRS) DISTFILES = $(DIST_COMMON) $(DIST_SOURCES) $(TEXINFOS) $(EXTRA_DIST) am__relativize = \ dir0=`pwd`; \ sed_first='s,^\([^/]*\)/.*$$,\1,'; \ sed_rest='s,^[^/]*/*,,'; \ sed_last='s,^.*/\([^/]*\)$$,\1,'; \ sed_butlast='s,/*[^/]*$$,,'; \ while test -n "$$dir1"; do \ first=`echo "$$dir1" | sed -e "$$sed_first"`; \ if test "$$first" != "."; then \ if test "$$first" = ".."; then \ dir2=`echo "$$dir0" | sed -e "$$sed_last"`/"$$dir2"; \ dir0=`echo "$$dir0" | sed -e "$$sed_butlast"`; \ else \ first2=`echo "$$dir2" | sed -e "$$sed_first"`; \ if test "$$first2" = "$$first"; then \ dir2=`echo "$$dir2" | sed -e "$$sed_rest"`; \ else \ dir2="../$$dir2"; \ fi; \ dir0="$$dir0"/"$$first"; \ fi; \ fi; \ dir1=`echo "$$dir1" | sed -e "$$sed_rest"`; \ done; \ reldir="$$dir2" ACLOCAL = @ACLOCAL@ ALLOCA = @ALLOCA@ ALTIVEC_CFLAGS = @ALTIVEC_CFLAGS@ AMTAR = @AMTAR@ AR = @AR@ AS = @AS@ AUTOCONF = @AUTOCONF@ AUTOHEADER = @AUTOHEADER@ AUTOMAKE = @AUTOMAKE@ AVX_CFLAGS = @AVX_CFLAGS@ AWK = @AWK@ CC = @CC@ CCDEPMODE = @CCDEPMODE@ CFLAGS = @CFLAGS@ CHECK_PL_OPTS = @CHECK_PL_OPTS@ CPP = @CPP@ CPPFLAGS = @CPPFLAGS@ CYGPATH_W = @CYGPATH_W@ C_FFTW_R2R_KIND = @C_FFTW_R2R_KIND@ C_MPI_FINT = @C_MPI_FINT@ DEFS = @DEFS@ DEPDIR = @DEPDIR@ DLLTOOL = @DLLTOOL@ DSYMUTIL = @DSYMUTIL@ DUMPBIN = @DUMPBIN@ ECHO_C = @ECHO_C@ ECHO_N = @ECHO_N@ ECHO_T = @ECHO_T@ EGREP = @EGREP@ EXEEXT = @EXEEXT@ F77 = @F77@ FFLAGS = @FFLAGS@ FGREP = @FGREP@ FLIBS = @FLIBS@ GREP = @GREP@ INSTALL = @INSTALL@ INSTALL_DATA = @INSTALL_DATA@ INSTALL_PROGRAM = @INSTALL_PROGRAM@ INSTALL_SCRIPT = @INSTALL_SCRIPT@ INSTALL_STRIP_PROGRAM = @INSTALL_STRIP_PROGRAM@ LD = @LD@ LDFLAGS = @LDFLAGS@ LIBOBJS = @LIBOBJS@ LIBQUADMATH = @LIBQUADMATH@ LIBS = @LIBS@ LIBTOOL = @LIBTOOL@ LIPO = @LIPO@ LN_S = @LN_S@ LTLIBOBJS = @LTLIBOBJS@ MAINT = @MAINT@ MAKEINFO = @MAKEINFO@ MANIFEST_TOOL = @MANIFEST_TOOL@ MKDIR_P = @MKDIR_P@ MPICC = @MPICC@ MPILIBS = @MPILIBS@ MPIRUN = @MPIRUN@ NEON_CFLAGS = @NEON_CFLAGS@ NM = @NM@ NMEDIT = @NMEDIT@ OBJDUMP = @OBJDUMP@ OBJEXT = @OBJEXT@ OCAMLBUILD = @OCAMLBUILD@ OPENMP_CFLAGS = @OPENMP_CFLAGS@ OTOOL = @OTOOL@ OTOOL64 = @OTOOL64@ PACKAGE = @PACKAGE@ PACKAGE_BUGREPORT = @PACKAGE_BUGREPORT@ PACKAGE_NAME = @PACKAGE_NAME@ PACKAGE_STRING = @PACKAGE_STRING@ PACKAGE_TARNAME = @PACKAGE_TARNAME@ PACKAGE_URL = @PACKAGE_URL@ PACKAGE_VERSION = @PACKAGE_VERSION@ PATH_SEPARATOR = @PATH_SEPARATOR@ POW_LIB = @POW_LIB@ PRECISION = @PRECISION@ PREC_SUFFIX = @PREC_SUFFIX@ PTHREAD_CC = @PTHREAD_CC@ PTHREAD_CFLAGS = @PTHREAD_CFLAGS@ PTHREAD_LIBS = @PTHREAD_LIBS@ RANLIB = @RANLIB@ SED = @SED@ SET_MAKE = @SET_MAKE@ SHARED_VERSION_INFO = @SHARED_VERSION_INFO@ SHELL = @SHELL@ SSE2_CFLAGS = @SSE2_CFLAGS@ STACK_ALIGN_CFLAGS = @STACK_ALIGN_CFLAGS@ STRIP = @STRIP@ THREADLIBS = @THREADLIBS@ VERSION = @VERSION@ abs_builddir = @abs_builddir@ abs_srcdir = @abs_srcdir@ abs_top_builddir = @abs_top_builddir@ abs_top_srcdir = @abs_top_srcdir@ ac_ct_AR = @ac_ct_AR@ ac_ct_CC = @ac_ct_CC@ ac_ct_DUMPBIN = @ac_ct_DUMPBIN@ ac_ct_F77 = @ac_ct_F77@ acx_pthread_config = @acx_pthread_config@ am__include = @am__include@ am__leading_dot = @am__leading_dot@ am__quote = @am__quote@ am__tar = @am__tar@ am__untar = @am__untar@ bindir = @bindir@ build = @build@ build_alias = @build_alias@ build_cpu = @build_cpu@ build_os = @build_os@ build_vendor = @build_vendor@ builddir = @builddir@ datadir = @datadir@ datarootdir = @datarootdir@ docdir = @docdir@ dvidir = @dvidir@ exec_prefix = @exec_prefix@ host = @host@ host_alias = @host_alias@ host_cpu = @host_cpu@ host_os = @host_os@ host_vendor = @host_vendor@ htmldir = @htmldir@ includedir = @includedir@ infodir = @infodir@ install_sh = @install_sh@ libdir = @libdir@ libexecdir = @libexecdir@ localedir = @localedir@ localstatedir = @localstatedir@ mandir = @mandir@ mkdir_p = @mkdir_p@ oldincludedir = @oldincludedir@ pdfdir = @pdfdir@ prefix = @prefix@ program_transform_name = @program_transform_name@ psdir = @psdir@ sbindir = @sbindir@ sharedstatedir = @sharedstatedir@ srcdir = @srcdir@ sysconfdir = @sysconfdir@ target_alias = @target_alias@ top_build_prefix = @top_build_prefix@ top_builddir = @top_builddir@ top_srcdir = @top_srcdir@ SUBDIRS = common sse2 avx altivec neon EXTRA_DIST = n1b.h n1f.h n2b.h n2f.h n2s.h q1b.h q1f.h t1b.h t1bu.h \ t1f.h t1fu.h t2b.h t2f.h t3b.h t3f.h ts.h codlist.mk simd.mk all: all-recursive .SUFFIXES: $(srcdir)/Makefile.in: @MAINTAINER_MODE_TRUE@ $(srcdir)/Makefile.am $(am__configure_deps) @for dep in $?; do \ case '$(am__configure_deps)' in \ *$$dep*) \ ( cd $(top_builddir) && $(MAKE) $(AM_MAKEFLAGS) am--refresh ) \ && { if test -f $@; then exit 0; else break; fi; }; \ exit 1;; \ esac; \ done; \ echo ' cd $(top_srcdir) && $(AUTOMAKE) --gnu dft/simd/Makefile'; \ $(am__cd) $(top_srcdir) && \ $(AUTOMAKE) --gnu dft/simd/Makefile .PRECIOUS: Makefile Makefile: $(srcdir)/Makefile.in $(top_builddir)/config.status @case '$?' in \ *config.status*) \ cd $(top_builddir) && $(MAKE) $(AM_MAKEFLAGS) am--refresh;; \ *) \ echo ' cd $(top_builddir) && $(SHELL) ./config.status $(subdir)/$@ $(am__depfiles_maybe)'; \ cd $(top_builddir) && $(SHELL) ./config.status $(subdir)/$@ $(am__depfiles_maybe);; \ esac; $(top_builddir)/config.status: $(top_srcdir)/configure $(CONFIG_STATUS_DEPENDENCIES) cd $(top_builddir) && $(MAKE) $(AM_MAKEFLAGS) am--refresh $(top_srcdir)/configure: @MAINTAINER_MODE_TRUE@ $(am__configure_deps) cd $(top_builddir) && $(MAKE) $(AM_MAKEFLAGS) am--refresh $(ACLOCAL_M4): @MAINTAINER_MODE_TRUE@ $(am__aclocal_m4_deps) cd $(top_builddir) && $(MAKE) $(AM_MAKEFLAGS) am--refresh $(am__aclocal_m4_deps): mostlyclean-libtool: -rm -f *.lo clean-libtool: -rm -rf .libs _libs # This directory's subdirectories are mostly independent; you can cd # into them and run `make' without going through this Makefile. # To change the values of `make' variables: instead of editing Makefiles, # (1) if the variable is set in `config.status', edit `config.status' # (which will cause the Makefiles to be regenerated when you run `make'); # (2) otherwise, pass the desired values on the `make' command line. $(RECURSIVE_TARGETS): @fail= failcom='exit 1'; \ for f in x $$MAKEFLAGS; do \ case $$f in \ *=* | --[!k]*);; \ *k*) failcom='fail=yes';; \ esac; \ done; \ dot_seen=no; \ target=`echo $@ | sed s/-recursive//`; \ list='$(SUBDIRS)'; for subdir in $$list; do \ echo "Making $$target in $$subdir"; \ if test "$$subdir" = "."; then \ dot_seen=yes; \ local_target="$$target-am"; \ else \ local_target="$$target"; \ fi; \ ($(am__cd) $$subdir && $(MAKE) $(AM_MAKEFLAGS) $$local_target) \ || eval $$failcom; \ done; \ if test "$$dot_seen" = "no"; then \ $(MAKE) $(AM_MAKEFLAGS) "$$target-am" || exit 1; \ fi; test -z "$$fail" $(RECURSIVE_CLEAN_TARGETS): @fail= failcom='exit 1'; \ for f in x $$MAKEFLAGS; do \ case $$f in \ *=* | --[!k]*);; \ *k*) failcom='fail=yes';; \ esac; \ done; \ dot_seen=no; \ case "$@" in \ distclean-* | maintainer-clean-*) list='$(DIST_SUBDIRS)' ;; \ *) list='$(SUBDIRS)' ;; \ esac; \ rev=''; for subdir in $$list; do \ if test "$$subdir" = "."; then :; else \ rev="$$subdir $$rev"; \ fi; \ done; \ rev="$$rev ."; \ target=`echo $@ | sed s/-recursive//`; \ for subdir in $$rev; do \ echo "Making $$target in $$subdir"; \ if test "$$subdir" = "."; then \ local_target="$$target-am"; \ else \ local_target="$$target"; \ fi; \ ($(am__cd) $$subdir && $(MAKE) $(AM_MAKEFLAGS) $$local_target) \ || eval $$failcom; \ done && test -z "$$fail" tags-recursive: list='$(SUBDIRS)'; for subdir in $$list; do \ test "$$subdir" = . || ($(am__cd) $$subdir && $(MAKE) $(AM_MAKEFLAGS) tags); \ done ctags-recursive: list='$(SUBDIRS)'; for subdir in $$list; do \ test "$$subdir" = . || ($(am__cd) $$subdir && $(MAKE) $(AM_MAKEFLAGS) ctags); \ done ID: $(HEADERS) $(SOURCES) $(LISP) $(TAGS_FILES) list='$(SOURCES) $(HEADERS) $(LISP) $(TAGS_FILES)'; \ unique=`for i in $$list; do \ if test -f "$$i"; then echo $$i; else echo $(srcdir)/$$i; fi; \ done | \ $(AWK) '{ files[$$0] = 1; nonempty = 1; } \ END { if (nonempty) { for (i in files) print i; }; }'`; \ mkid -fID $$unique tags: TAGS TAGS: tags-recursive $(HEADERS) $(SOURCES) $(TAGS_DEPENDENCIES) \ $(TAGS_FILES) $(LISP) set x; \ here=`pwd`; \ if ($(ETAGS) --etags-include --version) >/dev/null 2>&1; then \ include_option=--etags-include; \ empty_fix=.; \ else \ include_option=--include; \ empty_fix=; \ fi; \ list='$(SUBDIRS)'; for subdir in $$list; do \ if test "$$subdir" = .; then :; else \ test ! -f $$subdir/TAGS || \ set "$$@" "$$include_option=$$here/$$subdir/TAGS"; \ fi; \ done; \ list='$(SOURCES) $(HEADERS) $(LISP) $(TAGS_FILES)'; \ unique=`for i in $$list; do \ if test -f "$$i"; then echo $$i; else echo $(srcdir)/$$i; fi; \ done | \ $(AWK) '{ files[$$0] = 1; nonempty = 1; } \ END { if (nonempty) { for (i in files) print i; }; }'`; \ shift; \ if test -z "$(ETAGS_ARGS)$$*$$unique"; then :; else \ test -n "$$unique" || unique=$$empty_fix; \ if test $$# -gt 0; then \ $(ETAGS) $(ETAGSFLAGS) $(AM_ETAGSFLAGS) $(ETAGS_ARGS) \ "$$@" $$unique; \ else \ $(ETAGS) $(ETAGSFLAGS) $(AM_ETAGSFLAGS) $(ETAGS_ARGS) \ $$unique; \ fi; \ fi ctags: CTAGS CTAGS: ctags-recursive $(HEADERS) $(SOURCES) $(TAGS_DEPENDENCIES) \ $(TAGS_FILES) $(LISP) list='$(SOURCES) $(HEADERS) $(LISP) $(TAGS_FILES)'; \ unique=`for i in $$list; do \ if test -f "$$i"; then echo $$i; else echo $(srcdir)/$$i; fi; \ done | \ $(AWK) '{ files[$$0] = 1; nonempty = 1; } \ END { if (nonempty) { for (i in files) print i; }; }'`; \ test -z "$(CTAGS_ARGS)$$unique" \ || $(CTAGS) $(CTAGSFLAGS) $(AM_CTAGSFLAGS) $(CTAGS_ARGS) \ $$unique GTAGS: here=`$(am__cd) $(top_builddir) && pwd` \ && $(am__cd) $(top_srcdir) \ && gtags -i $(GTAGS_ARGS) "$$here" distclean-tags: -rm -f TAGS ID GTAGS GRTAGS GSYMS GPATH tags distdir: $(DISTFILES) @srcdirstrip=`echo "$(srcdir)" | sed 's/[].[^$$\\*]/\\\\&/g'`; \ topsrcdirstrip=`echo "$(top_srcdir)" | sed 's/[].[^$$\\*]/\\\\&/g'`; \ list='$(DISTFILES)'; \ dist_files=`for file in $$list; do echo $$file; done | \ sed -e "s|^$$srcdirstrip/||;t" \ -e "s|^$$topsrcdirstrip/|$(top_builddir)/|;t"`; \ case $$dist_files in \ */*) $(MKDIR_P) `echo "$$dist_files" | \ sed '/\//!d;s|^|$(distdir)/|;s,/[^/]*$$,,' | \ sort -u` ;; \ esac; \ for file in $$dist_files; do \ if test -f $$file || test -d $$file; then d=.; else d=$(srcdir); fi; \ if test -d $$d/$$file; then \ dir=`echo "/$$file" | sed -e 's,/[^/]*$$,,'`; \ if test -d "$(distdir)/$$file"; then \ find "$(distdir)/$$file" -type d ! -perm -700 -exec chmod u+rwx {} \;; \ fi; \ if test -d $(srcdir)/$$file && test $$d != $(srcdir); then \ cp -fpR $(srcdir)/$$file "$(distdir)$$dir" || exit 1; \ find "$(distdir)/$$file" -type d ! -perm -700 -exec chmod u+rwx {} \;; \ fi; \ cp -fpR $$d/$$file "$(distdir)$$dir" || exit 1; \ else \ test -f "$(distdir)/$$file" \ || cp -p $$d/$$file "$(distdir)/$$file" \ || exit 1; \ fi; \ done @list='$(DIST_SUBDIRS)'; for subdir in $$list; do \ if test "$$subdir" = .; then :; else \ $(am__make_dryrun) \ || test -d "$(distdir)/$$subdir" \ || $(MKDIR_P) "$(distdir)/$$subdir" \ || exit 1; \ dir1=$$subdir; dir2="$(distdir)/$$subdir"; \ $(am__relativize); \ new_distdir=$$reldir; \ dir1=$$subdir; dir2="$(top_distdir)"; \ $(am__relativize); \ new_top_distdir=$$reldir; \ echo " (cd $$subdir && $(MAKE) $(AM_MAKEFLAGS) top_distdir="$$new_top_distdir" distdir="$$new_distdir" \\"; \ echo " am__remove_distdir=: am__skip_length_check=: am__skip_mode_fix=: distdir)"; \ ($(am__cd) $$subdir && \ $(MAKE) $(AM_MAKEFLAGS) \ top_distdir="$$new_top_distdir" \ distdir="$$new_distdir" \ am__remove_distdir=: \ am__skip_length_check=: \ am__skip_mode_fix=: \ distdir) \ || exit 1; \ fi; \ done check-am: all-am check: check-recursive all-am: Makefile installdirs: installdirs-recursive installdirs-am: install: install-recursive install-exec: install-exec-recursive install-data: install-data-recursive uninstall: uninstall-recursive install-am: all-am @$(MAKE) $(AM_MAKEFLAGS) install-exec-am install-data-am installcheck: installcheck-recursive install-strip: if test -z '$(STRIP)'; then \ $(MAKE) $(AM_MAKEFLAGS) INSTALL_PROGRAM="$(INSTALL_STRIP_PROGRAM)" \ install_sh_PROGRAM="$(INSTALL_STRIP_PROGRAM)" INSTALL_STRIP_FLAG=-s \ install; \ else \ $(MAKE) $(AM_MAKEFLAGS) INSTALL_PROGRAM="$(INSTALL_STRIP_PROGRAM)" \ install_sh_PROGRAM="$(INSTALL_STRIP_PROGRAM)" INSTALL_STRIP_FLAG=-s \ "INSTALL_PROGRAM_ENV=STRIPPROG='$(STRIP)'" install; \ fi mostlyclean-generic: clean-generic: distclean-generic: -test -z "$(CONFIG_CLEAN_FILES)" || rm -f $(CONFIG_CLEAN_FILES) -test . = "$(srcdir)" || test -z "$(CONFIG_CLEAN_VPATH_FILES)" || rm -f $(CONFIG_CLEAN_VPATH_FILES) maintainer-clean-generic: @echo "This command is intended for maintainers to use" @echo "it deletes files that may require special tools to rebuild." clean: clean-recursive clean-am: clean-generic clean-libtool mostlyclean-am distclean: distclean-recursive -rm -f Makefile distclean-am: clean-am distclean-generic distclean-tags dvi: dvi-recursive dvi-am: html: html-recursive html-am: info: info-recursive info-am: install-data-am: install-dvi: install-dvi-recursive install-dvi-am: install-exec-am: install-html: install-html-recursive install-html-am: install-info: install-info-recursive install-info-am: install-man: install-pdf: install-pdf-recursive install-pdf-am: install-ps: install-ps-recursive install-ps-am: installcheck-am: maintainer-clean: maintainer-clean-recursive -rm -f Makefile maintainer-clean-am: distclean-am maintainer-clean-generic mostlyclean: mostlyclean-recursive mostlyclean-am: mostlyclean-generic mostlyclean-libtool pdf: pdf-recursive pdf-am: ps: ps-recursive ps-am: uninstall-am: .MAKE: $(RECURSIVE_CLEAN_TARGETS) $(RECURSIVE_TARGETS) ctags-recursive \ install-am install-strip tags-recursive .PHONY: $(RECURSIVE_CLEAN_TARGETS) $(RECURSIVE_TARGETS) CTAGS GTAGS \ all all-am check check-am clean clean-generic clean-libtool \ ctags ctags-recursive distclean distclean-generic \ distclean-libtool distclean-tags distdir dvi dvi-am html \ html-am info info-am install install-am install-data \ install-data-am install-dvi install-dvi-am install-exec \ install-exec-am install-html install-html-am install-info \ install-info-am install-man install-pdf install-pdf-am \ install-ps install-ps-am install-strip installcheck \ installcheck-am installdirs installdirs-am maintainer-clean \ maintainer-clean-generic mostlyclean mostlyclean-generic \ mostlyclean-libtool pdf pdf-am ps ps-am tags tags-recursive \ uninstall uninstall-am # Tell versions [3.59,3.63) of GNU make to not export all variables. # Otherwise a system limit (for SysV at least) may be exceeded. .NOEXPORT: fftw-3.3.3/dft/simd/n1b.h0000644000175000017500000000165112054410050011743 00000000000000/* * Copyright (c) 2003, 2007-11 Matteo Frigo * Copyright (c) 2003, 2007-11 Massachusetts Institute of Technology * * This program is free software; you can redistribute it and/or modify * it under the terms of the GNU General Public License as published by * the Free Software Foundation; either version 2 of the License, or * (at your option) any later version. * * This program is distributed in the hope that it will be useful, * but WITHOUT ANY WARRANTY; without even the implied warranty of * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the * GNU General Public License for more details. * * You should have received a copy of the GNU General Public License * along with this program; if not, write to the Free Software * Foundation, Inc., 51 Franklin Street, Fifth Floor, Boston, MA 02110-1301 USA * */ #include SIMD_HEADER #define GENUS XSIMD(dft_n1bsimd_genus) extern const kdft_genus GENUS; fftw-3.3.3/dft/simd/n1f.h0000644000175000017500000000165112054410050011747 00000000000000/* * Copyright (c) 2003, 2007-11 Matteo Frigo * Copyright (c) 2003, 2007-11 Massachusetts Institute of Technology * * This program is free software; you can redistribute it and/or modify * it under the terms of the GNU General Public License as published by * the Free Software Foundation; either version 2 of the License, or * (at your option) any later version. * * This program is distributed in the hope that it will be useful, * but WITHOUT ANY WARRANTY; without even the implied warranty of * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the * GNU General Public License for more details. * * You should have received a copy of the GNU General Public License * along with this program; if not, write to the Free Software * Foundation, Inc., 51 Franklin Street, Fifth Floor, Boston, MA 02110-1301 USA * */ #include SIMD_HEADER #define GENUS XSIMD(dft_n1fsimd_genus) extern const kdft_genus GENUS; fftw-3.3.3/dft/simd/n2b.h0000644000175000017500000000170312054410050011742 00000000000000/* * Copyright (c) 2003, 2007-11 Matteo Frigo * Copyright (c) 2003, 2007-11 Massachusetts Institute of Technology * * This program is free software; you can redistribute it and/or modify * it under the terms of the GNU General Public License as published by * the Free Software Foundation; either version 2 of the License, or * (at your option) any later version. * * This program is distributed in the hope that it will be useful, * but WITHOUT ANY WARRANTY; without even the implied warranty of * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the * GNU General Public License for more details. * * You should have received a copy of the GNU General Public License * along with this program; if not, write to the Free Software * Foundation, Inc., 51 Franklin Street, Fifth Floor, Boston, MA 02110-1301 USA * */ #include SIMD_HEADER #undef LD #define LD LDA #define GENUS XSIMD(dft_n2bsimd_genus) extern const kdft_genus GENUS; fftw-3.3.3/dft/simd/n2f.h0000644000175000017500000000170312054410050011746 00000000000000/* * Copyright (c) 2003, 2007-11 Matteo Frigo * Copyright (c) 2003, 2007-11 Massachusetts Institute of Technology * * This program is free software; you can redistribute it and/or modify * it under the terms of the GNU General Public License as published by * the Free Software Foundation; either version 2 of the License, or * (at your option) any later version. * * This program is distributed in the hope that it will be useful, * but WITHOUT ANY WARRANTY; without even the implied warranty of * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the * GNU General Public License for more details. * * You should have received a copy of the GNU General Public License * along with this program; if not, write to the Free Software * Foundation, Inc., 51 Franklin Street, Fifth Floor, Boston, MA 02110-1301 USA * */ #include SIMD_HEADER #undef LD #define LD LDA #define GENUS XSIMD(dft_n2fsimd_genus) extern const kdft_genus GENUS; fftw-3.3.3/dft/simd/n2s.h0000644000175000017500000000170312054410050011763 00000000000000/* * Copyright (c) 2003, 2007-11 Matteo Frigo * Copyright (c) 2003, 2007-11 Massachusetts Institute of Technology * * This program is free software; you can redistribute it and/or modify * it under the terms of the GNU General Public License as published by * the Free Software Foundation; either version 2 of the License, or * (at your option) any later version. * * This program is distributed in the hope that it will be useful, * but WITHOUT ANY WARRANTY; without even the implied warranty of * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the * GNU General Public License for more details. * * You should have received a copy of the GNU General Public License * along with this program; if not, write to the Free Software * Foundation, Inc., 51 Franklin Street, Fifth Floor, Boston, MA 02110-1301 USA * */ #include SIMD_HEADER #undef LD #define LD LDA #define GENUS XSIMD(dft_n2ssimd_genus) extern const kdft_genus GENUS; fftw-3.3.3/dft/simd/q1b.h0000644000175000017500000000176512054410050011754 00000000000000/* * Copyright (c) 2003, 2007-11 Matteo Frigo * Copyright (c) 2003, 2007-11 Massachusetts Institute of Technology * * This program is free software; you can redistribute it and/or modify * it under the terms of the GNU General Public License as published by * the Free Software Foundation; either version 2 of the License, or * (at your option) any later version. * * This program is distributed in the hope that it will be useful, * but WITHOUT ANY WARRANTY; without even the implied warranty of * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the * GNU General Public License for more details. * * You should have received a copy of the GNU General Public License * along with this program; if not, write to the Free Software * Foundation, Inc., 51 Franklin Street, Fifth Floor, Boston, MA 02110-1301 USA * */ #include SIMD_HEADER #define VTW VTW1 #define TWVL TWVL1 #define BYTW BYTW1 #define BYTWJ BYTWJ1 #define GENUS XSIMD(dft_q1bsimd_genus) extern const ct_genus GENUS; fftw-3.3.3/dft/simd/q1f.h0000644000175000017500000000176512054410050011760 00000000000000/* * Copyright (c) 2003, 2007-11 Matteo Frigo * Copyright (c) 2003, 2007-11 Massachusetts Institute of Technology * * This program is free software; you can redistribute it and/or modify * it under the terms of the GNU General Public License as published by * the Free Software Foundation; either version 2 of the License, or * (at your option) any later version. * * This program is distributed in the hope that it will be useful, * but WITHOUT ANY WARRANTY; without even the implied warranty of * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the * GNU General Public License for more details. * * You should have received a copy of the GNU General Public License * along with this program; if not, write to the Free Software * Foundation, Inc., 51 Franklin Street, Fifth Floor, Boston, MA 02110-1301 USA * */ #include SIMD_HEADER #define VTW VTW1 #define TWVL TWVL1 #define BYTW BYTW1 #define BYTWJ BYTWJ1 #define GENUS XSIMD(dft_q1fsimd_genus) extern const ct_genus GENUS; fftw-3.3.3/dft/simd/t1b.h0000644000175000017500000000205012054410050011743 00000000000000/* * Copyright (c) 2003, 2007-11 Matteo Frigo * Copyright (c) 2003, 2007-11 Massachusetts Institute of Technology * * This program is free software; you can redistribute it and/or modify * it under the terms of the GNU General Public License as published by * the Free Software Foundation; either version 2 of the License, or * (at your option) any later version. * * This program is distributed in the hope that it will be useful, * but WITHOUT ANY WARRANTY; without even the implied warranty of * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the * GNU General Public License for more details. * * You should have received a copy of the GNU General Public License * along with this program; if not, write to the Free Software * Foundation, Inc., 51 Franklin Street, Fifth Floor, Boston, MA 02110-1301 USA * */ #include SIMD_HEADER #undef LD #define LD LDA #undef ST #define ST STA #define VTW VTW1 #define TWVL TWVL1 #define BYTW BYTW1 #define BYTWJ BYTWJ1 #define GENUS XSIMD(dft_t1bsimd_genus) extern const ct_genus GENUS; fftw-3.3.3/dft/simd/t1bu.h0000644000175000017500000000176612054410050012145 00000000000000/* * Copyright (c) 2003, 2007-11 Matteo Frigo * Copyright (c) 2003, 2007-11 Massachusetts Institute of Technology * * This program is free software; you can redistribute it and/or modify * it under the terms of the GNU General Public License as published by * the Free Software Foundation; either version 2 of the License, or * (at your option) any later version. * * This program is distributed in the hope that it will be useful, * but WITHOUT ANY WARRANTY; without even the implied warranty of * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the * GNU General Public License for more details. * * You should have received a copy of the GNU General Public License * along with this program; if not, write to the Free Software * Foundation, Inc., 51 Franklin Street, Fifth Floor, Boston, MA 02110-1301 USA * */ #include SIMD_HEADER #define VTW VTW1 #define TWVL TWVL1 #define BYTW BYTW1 #define BYTWJ BYTWJ1 #define GENUS XSIMD(dft_t1busimd_genus) extern const ct_genus GENUS; fftw-3.3.3/dft/simd/t1f.h0000644000175000017500000000205012054410050011747 00000000000000/* * Copyright (c) 2003, 2007-11 Matteo Frigo * Copyright (c) 2003, 2007-11 Massachusetts Institute of Technology * * This program is free software; you can redistribute it and/or modify * it under the terms of the GNU General Public License as published by * the Free Software Foundation; either version 2 of the License, or * (at your option) any later version. * * This program is distributed in the hope that it will be useful, * but WITHOUT ANY WARRANTY; without even the implied warranty of * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the * GNU General Public License for more details. * * You should have received a copy of the GNU General Public License * along with this program; if not, write to the Free Software * Foundation, Inc., 51 Franklin Street, Fifth Floor, Boston, MA 02110-1301 USA * */ #include SIMD_HEADER #undef LD #define LD LDA #undef ST #define ST STA #define VTW VTW1 #define TWVL TWVL1 #define BYTW BYTW1 #define BYTWJ BYTWJ1 #define GENUS XSIMD(dft_t1fsimd_genus) extern const ct_genus GENUS; fftw-3.3.3/dft/simd/t1fu.h0000644000175000017500000000176612054410050012151 00000000000000/* * Copyright (c) 2003, 2007-11 Matteo Frigo * Copyright (c) 2003, 2007-11 Massachusetts Institute of Technology * * This program is free software; you can redistribute it and/or modify * it under the terms of the GNU General Public License as published by * the Free Software Foundation; either version 2 of the License, or * (at your option) any later version. * * This program is distributed in the hope that it will be useful, * but WITHOUT ANY WARRANTY; without even the implied warranty of * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the * GNU General Public License for more details. * * You should have received a copy of the GNU General Public License * along with this program; if not, write to the Free Software * Foundation, Inc., 51 Franklin Street, Fifth Floor, Boston, MA 02110-1301 USA * */ #include SIMD_HEADER #define VTW VTW1 #define TWVL TWVL1 #define BYTW BYTW1 #define BYTWJ BYTWJ1 #define GENUS XSIMD(dft_t1fusimd_genus) extern const ct_genus GENUS; fftw-3.3.3/dft/simd/t2b.h0000644000175000017500000000205012054410050011744 00000000000000/* * Copyright (c) 2003, 2007-11 Matteo Frigo * Copyright (c) 2003, 2007-11 Massachusetts Institute of Technology * * This program is free software; you can redistribute it and/or modify * it under the terms of the GNU General Public License as published by * the Free Software Foundation; either version 2 of the License, or * (at your option) any later version. * * This program is distributed in the hope that it will be useful, * but WITHOUT ANY WARRANTY; without even the implied warranty of * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the * GNU General Public License for more details. * * You should have received a copy of the GNU General Public License * along with this program; if not, write to the Free Software * Foundation, Inc., 51 Franklin Street, Fifth Floor, Boston, MA 02110-1301 USA * */ #include SIMD_HEADER #undef LD #define LD LDA #undef ST #define ST STA #define VTW VTW2 #define TWVL TWVL2 #define BYTW BYTW2 #define BYTWJ BYTWJ2 #define GENUS XSIMD(dft_t2bsimd_genus) extern const ct_genus GENUS; fftw-3.3.3/dft/simd/t2f.h0000644000175000017500000000205012054410050011750 00000000000000/* * Copyright (c) 2003, 2007-11 Matteo Frigo * Copyright (c) 2003, 2007-11 Massachusetts Institute of Technology * * This program is free software; you can redistribute it and/or modify * it under the terms of the GNU General Public License as published by * the Free Software Foundation; either version 2 of the License, or * (at your option) any later version. * * This program is distributed in the hope that it will be useful, * but WITHOUT ANY WARRANTY; without even the implied warranty of * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the * GNU General Public License for more details. * * You should have received a copy of the GNU General Public License * along with this program; if not, write to the Free Software * Foundation, Inc., 51 Franklin Street, Fifth Floor, Boston, MA 02110-1301 USA * */ #include SIMD_HEADER #undef LD #define LD LDA #undef ST #define ST STA #define VTW VTW2 #define TWVL TWVL2 #define BYTW BYTW2 #define BYTWJ BYTWJ2 #define GENUS XSIMD(dft_t2fsimd_genus) extern const ct_genus GENUS; fftw-3.3.3/dft/simd/t3b.h0000644000175000017500000000212212054410050011745 00000000000000/* * Copyright (c) 2003, 2007-11 Matteo Frigo * Copyright (c) 2003, 2007-11 Massachusetts Institute of Technology * * This program is free software; you can redistribute it and/or modify * it under the terms of the GNU General Public License as published by * the Free Software Foundation; either version 2 of the License, or * (at your option) any later version. * * This program is distributed in the hope that it will be useful, * but WITHOUT ANY WARRANTY; without even the implied warranty of * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the * GNU General Public License for more details. * * You should have received a copy of the GNU General Public License * along with this program; if not, write to the Free Software * Foundation, Inc., 51 Franklin Street, Fifth Floor, Boston, MA 02110-1301 USA * */ #include SIMD_HEADER #undef LD #define LD LDA #undef ST #define ST STA #define VTW VTW3 #define TWVL TWVL3 #define LDW(x) LDA(x, 0, 0) /* load twiddle factor */ /* same as t1b otherwise */ #define GENUS XSIMD(dft_t1bsimd_genus) extern const ct_genus GENUS; fftw-3.3.3/dft/simd/t3f.h0000644000175000017500000000212212054410050011751 00000000000000/* * Copyright (c) 2003, 2007-11 Matteo Frigo * Copyright (c) 2003, 2007-11 Massachusetts Institute of Technology * * This program is free software; you can redistribute it and/or modify * it under the terms of the GNU General Public License as published by * the Free Software Foundation; either version 2 of the License, or * (at your option) any later version. * * This program is distributed in the hope that it will be useful, * but WITHOUT ANY WARRANTY; without even the implied warranty of * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the * GNU General Public License for more details. * * You should have received a copy of the GNU General Public License * along with this program; if not, write to the Free Software * Foundation, Inc., 51 Franklin Street, Fifth Floor, Boston, MA 02110-1301 USA * */ #include SIMD_HEADER #undef LD #define LD LDA #undef ST #define ST STA #define VTW VTW3 #define TWVL TWVL3 #define LDW(x) LDA(x, 0, 0) /* load twiddle factor */ /* same as t1f otherwise */ #define GENUS XSIMD(dft_t1fsimd_genus) extern const ct_genus GENUS; fftw-3.3.3/dft/simd/ts.h0000644000175000017500000000206512054410050011711 00000000000000/* * Copyright (c) 2003, 2007-11 Matteo Frigo * Copyright (c) 2003, 2007-11 Massachusetts Institute of Technology * * This program is free software; you can redistribute it and/or modify * it under the terms of the GNU General Public License as published by * the Free Software Foundation; either version 2 of the License, or * (at your option) any later version. * * This program is distributed in the hope that it will be useful, * but WITHOUT ANY WARRANTY; without even the implied warranty of * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the * GNU General Public License for more details. * * You should have received a copy of the GNU General Public License * along with this program; if not, write to the Free Software * Foundation, Inc., 51 Franklin Street, Fifth Floor, Boston, MA 02110-1301 USA * */ #include SIMD_HEADER #undef LD #define LD LDA #undef ST #define ST STA #define VTW VTWS #define TWVL TWVLS #define LDW(x) LDA(x, 0, 0) /* load twiddle factor */ #define GENUS XSIMD(dft_tssimd_genus) extern const ct_genus GENUS; fftw-3.3.3/dft/simd/codlist.mk0000644000175000017500000000704112054410050013103 00000000000000# This file contains a standard list of DFT SIMD codelets. It is # included by common/Makefile to generate the C files with the actual # codelets in them. It is included by {sse,sse2,...}/Makefile to # generate and compile stub files that include common/*.c # You can customize FFTW for special needs, e.g. to handle certain # sizes more efficiently, by adding new codelets to the lists of those # included by default. If you change the list of codelets, any new # ones you added will be automatically generated when you run the # bootstrap script (see "Generating your own code" in the FFTW # manual). ########################################################################### # n1fv_ is a hard-coded FFTW_FORWARD FFT of size , using SIMD N1F = n1fv_2.c n1fv_3.c n1fv_4.c n1fv_5.c n1fv_6.c n1fv_7.c n1fv_8.c \ n1fv_9.c n1fv_10.c n1fv_11.c n1fv_12.c n1fv_13.c n1fv_14.c n1fv_15.c \ n1fv_16.c n1fv_32.c n1fv_64.c n1fv_128.c n1fv_20.c n1fv_25.c # as above, with restricted input vector stride N2F = n2fv_2.c n2fv_4.c n2fv_6.c n2fv_8.c n2fv_10.c n2fv_12.c \ n2fv_14.c n2fv_16.c n2fv_32.c n2fv_64.c n2fv_20.c # as above, but FFTW_BACKWARD N1B = n1bv_2.c n1bv_3.c n1bv_4.c n1bv_5.c n1bv_6.c n1bv_7.c n1bv_8.c \ n1bv_9.c n1bv_10.c n1bv_11.c n1bv_12.c n1bv_13.c n1bv_14.c n1bv_15.c \ n1bv_16.c n1bv_32.c n1bv_64.c n1bv_128.c n1bv_20.c n1bv_25.c N2B = n2bv_2.c n2bv_4.c n2bv_6.c n2bv_8.c n2bv_10.c n2bv_12.c \ n2bv_14.c n2bv_16.c n2bv_32.c n2bv_64.c n2bv_20.c # split-complex codelets N2S = n2sv_4.c n2sv_8.c n2sv_16.c n2sv_32.c n2sv_64.c ########################################################################### # t1fv_ is a "twiddle" FFT of size , implementing a radix-r DIT step # for an FFTW_FORWARD transform, using SIMD T1F = t1fv_2.c t1fv_3.c t1fv_4.c t1fv_5.c t1fv_6.c t1fv_7.c t1fv_8.c \ t1fv_9.c t1fv_10.c t1fv_12.c t1fv_15.c t1fv_16.c t1fv_32.c t1fv_64.c \ t1fv_20.c t1fv_25.c # same as t1fv_*, but with different twiddle storage scheme T2F = t2fv_2.c t2fv_4.c t2fv_8.c t2fv_16.c t2fv_32.c t2fv_64.c \ t2fv_5.c t2fv_10.c t2fv_20.c t2fv_25.c T3F = t3fv_4.c t3fv_8.c t3fv_16.c t3fv_32.c t3fv_5.c t3fv_10.c \ t3fv_20.c t3fv_25.c T1FU = t1fuv_2.c t1fuv_3.c t1fuv_4.c t1fuv_5.c t1fuv_6.c t1fuv_7.c \ t1fuv_8.c t1fuv_9.c t1fuv_10.c # as above, but FFTW_BACKWARD T1B = t1bv_2.c t1bv_3.c t1bv_4.c t1bv_5.c t1bv_6.c t1bv_7.c t1bv_8.c \ t1bv_9.c t1bv_10.c t1bv_12.c t1bv_15.c t1bv_16.c t1bv_32.c t1bv_64.c \ t1bv_20.c t1bv_25.c # same as t1bv_*, but with different twiddle storage scheme T2B = t2bv_2.c t2bv_4.c t2bv_8.c t2bv_16.c t2bv_32.c t2bv_64.c \ t2bv_5.c t2bv_10.c t2bv_20.c t2bv_25.c T3B = t3bv_4.c t3bv_8.c t3bv_16.c t3bv_32.c t3bv_5.c t3bv_10.c \ t3bv_20.c t3bv_25.c T1BU = t1buv_2.c t1buv_3.c t1buv_4.c t1buv_5.c t1buv_6.c t1buv_7.c \ t1buv_8.c t1buv_9.c t1buv_10.c # split-complex codelets T1S = t1sv_2.c t1sv_4.c t1sv_8.c t1sv_16.c t1sv_32.c T2S = t2sv_4.c t2sv_8.c t2sv_16.c t2sv_32.c ########################################################################### # q1fv_ is twiddle FFTW_FORWARD FFTs of size (DIF step), # where the output is transposed, using SIMD. This is used for # in-place transposes in sizes that are divisible by ^2. These # codelets have size ~ ^2, so you should probably not use # bigger than 8 or so. Q1F = q1fv_2.c q1fv_4.c q1fv_5.c q1fv_8.c # as above, but FFTW_BACKWARD Q1B = q1bv_2.c q1bv_4.c q1bv_5.c q1bv_8.c ########################################################################### SIMD_CODELETS = $(N1F) $(N1B) $(N2F) $(N2B) $(N2S) $(T1FU) $(T1F) \ $(T2F) $(T3F) $(T1BU) $(T1B) $(T2B) $(T3B) $(T1S) $(T2S) $(Q1F) $(Q1B) fftw-3.3.3/dft/simd/simd.mk0000644000175000017500000000054412054410050012377 00000000000000AM_CPPFLAGS = -I$(top_srcdir)/kernel -I$(top_srcdir)/dft \ -I$(top_srcdir)/dft/simd -I$(top_srcdir)/simd-support EXTRA_DIST = $(SIMD_CODELETS) genus.c codlist.c $(EXTRA_DIST): Makefile ( \ echo "/* Generated automatically. DO NOT EDIT! */"; \ echo "#define SIMD_HEADER \"$(SIMD_HEADER)\""; \ echo "#include \"../common/"$*".c\""; \ ) >$@ fftw-3.3.3/dft/simd/common/0000755000175000017500000000000012054411106012462 500000000000000fftw-3.3.3/dft/simd/common/Makefile.am0000644000175000017500000000727712054410050014450 00000000000000# include the list of codelets include $(top_srcdir)/dft/simd/codlist.mk ALL_CODELETS = $(SIMD_CODELETS) BUILT_SOURCES= $(SIMD_CODELETS) $(CODLIST) EXTRA_DIST = $(BUILT_SOURCES) genus.c INCLUDE_SIMD_HEADER="\#include SIMD_HEADER" XRENAME=XSIMD SOLVTAB_NAME = XSIMD(solvtab_dft) # include special rules for regenerating codelets. include $(top_srcdir)/support/Makefile.codelets if MAINTAINER_MODE GFLAGS = -simd $(FLAGS_COMMON) -pipeline-latency 8 FLAGS_T2S=-twiddle-log3 -precompute-twiddles FLAGS_T3=-twiddle-log3 -precompute-twiddles -no-generate-bytw n1fv_%.c: $(CODELET_DEPS) $(GEN_NOTW_C) ($(PRELUDE_COMMANDS_DFT); $(TWOVERS) $(GEN_NOTW_C) $(GFLAGS) -n $* -name n1fv_$* -include "n1f.h") | $(ADD_DATE) | $(INDENT) >$@ n2fv_%.c: $(CODELET_DEPS) $(GEN_NOTW_C) ($(PRELUDE_COMMANDS_DFT); $(TWOVERS) $(GEN_NOTW_C) $(GFLAGS) -n $* -name n2fv_$* -with-ostride 2 -include "n2f.h" -store-multiple 2) | $(ADD_DATE) | $(INDENT) >$@ n1bv_%.c: $(CODELET_DEPS) $(GEN_NOTW_C) ($(PRELUDE_COMMANDS_DFT); $(TWOVERS) $(GEN_NOTW_C) $(GFLAGS) -sign 1 -n $* -name n1bv_$* -include "n1b.h") | $(ADD_DATE) | $(INDENT) >$@ n2bv_%.c: $(CODELET_DEPS) $(GEN_NOTW_C) ($(PRELUDE_COMMANDS_DFT); $(TWOVERS) $(GEN_NOTW_C) $(GFLAGS) -sign 1 -n $* -name n2bv_$* -with-ostride 2 -include "n2b.h" -store-multiple 2) | $(ADD_DATE) | $(INDENT) >$@ n2sv_%.c: $(CODELET_DEPS) $(GEN_NOTW) ($(PRELUDE_COMMANDS_DFT); $(TWOVERS) $(GEN_NOTW) $(GFLAGS) -n $* -name n2sv_$* -with-ostride 1 -include "n2s.h" -store-multiple 4) | $(ADD_DATE) | $(INDENT) >$@ t1fv_%.c: $(CODELET_DEPS) $(GEN_TWIDDLE_C) ($(PRELUDE_COMMANDS_DFT); $(TWOVERS) $(GEN_TWIDDLE_C) $(GFLAGS) -n $* -name t1fv_$* -include "t1f.h") | $(ADD_DATE) | $(INDENT) >$@ t1fuv_%.c: $(CODELET_DEPS) $(GEN_TWIDDLE_C) ($(PRELUDE_COMMANDS_DFT); $(TWOVERS) $(GEN_TWIDDLE_C) $(GFLAGS) -n $* -name t1fuv_$* -include "t1fu.h") | $(ADD_DATE) | $(INDENT) >$@ t2fv_%.c: $(CODELET_DEPS) $(GEN_TWIDDLE_C) ($(PRELUDE_COMMANDS_DFT); $(TWOVERS) $(GEN_TWIDDLE_C) $(GFLAGS) -n $* -name t2fv_$* -include "t2f.h") | $(ADD_DATE) | $(INDENT) >$@ t3fv_%.c: $(CODELET_DEPS) $(GEN_TWIDDLE_C) ($(PRELUDE_COMMANDS_DFT); $(TWOVERS) $(GEN_TWIDDLE_C) $(GFLAGS) $(FLAGS_T3) -n $* -name t3fv_$* -include "t3f.h") | $(ADD_DATE) | $(INDENT) >$@ t1bv_%.c: $(CODELET_DEPS) $(GEN_TWIDDLE_C) ($(PRELUDE_COMMANDS_DFT); $(TWOVERS) $(GEN_TWIDDLE_C) $(GFLAGS) -n $* -name t1bv_$* -include "t1b.h" -sign 1) | $(ADD_DATE) | $(INDENT) >$@ t1buv_%.c: $(CODELET_DEPS) $(GEN_TWIDDLE_C) ($(PRELUDE_COMMANDS_DFT); $(TWOVERS) $(GEN_TWIDDLE_C) $(GFLAGS) -n $* -name t1buv_$* -include "t1bu.h" -sign 1) | $(ADD_DATE) | $(INDENT) >$@ t2bv_%.c: $(CODELET_DEPS) $(GEN_TWIDDLE_C) ($(PRELUDE_COMMANDS_DFT); $(TWOVERS) $(GEN_TWIDDLE_C) $(GFLAGS) -n $* -name t2bv_$* -include "t2b.h" -sign 1) | $(ADD_DATE) | $(INDENT) >$@ t3bv_%.c: $(CODELET_DEPS) $(GEN_TWIDDLE_C) ($(PRELUDE_COMMANDS_DFT); $(TWOVERS) $(GEN_TWIDDLE_C) $(GFLAGS) $(FLAGS_T3) -n $* -name t3bv_$* -include "t3b.h" -sign 1) | $(ADD_DATE) | $(INDENT) >$@ t1sv_%.c: $(CODELET_DEPS) $(GEN_TWIDDLE) ($(PRELUDE_COMMANDS_DFT); $(TWOVERS) $(GEN_TWIDDLE) $(GFLAGS) -n $* -name t1sv_$* -include "ts.h") | $(ADD_DATE) | $(INDENT) >$@ t2sv_%.c: $(CODELET_DEPS) $(GEN_TWIDDLE) ($(PRELUDE_COMMANDS_DFT); $(TWOVERS) $(GEN_TWIDDLE) $(GFLAGS) $(FLAGS_T2S) -n $* -name t2sv_$* -include "ts.h") | $(ADD_DATE) | $(INDENT) >$@ q1fv_%.c: $(CODELET_DEPS) $(GEN_TWIDSQ_C) ($(PRELUDE_COMMANDS_DFT); $(TWOVERS) $(GEN_TWIDSQ_C) $(GFLAGS) -n $* -dif -name q1fv_$* -include "q1f.h") | $(ADD_DATE) | $(INDENT) >$@ q1bv_%.c: $(CODELET_DEPS) $(GEN_TWIDSQ_C) ($(PRELUDE_COMMANDS_DFT); $(TWOVERS) $(GEN_TWIDSQ_C) $(GFLAGS) -n $* -dif -name q1bv_$* -include "q1b.h" -sign 1) | $(ADD_DATE) | $(INDENT) >$@ endif # MAINTAINER_MODE fftw-3.3.3/dft/simd/common/Makefile.in0000644000175000017500000005516012054410177014465 00000000000000# Makefile.in generated by automake 1.11.6 from Makefile.am. # @configure_input@ # Copyright (C) 1994, 1995, 1996, 1997, 1998, 1999, 2000, 2001, 2002, # 2003, 2004, 2005, 2006, 2007, 2008, 2009, 2010, 2011 Free Software # Foundation, Inc. # This Makefile.in is free software; the Free Software Foundation # gives unlimited permission to copy and/or distribute it, # with or without modifications, as long as this notice is preserved. # This program is distributed in the hope that it will be useful, # but WITHOUT ANY WARRANTY, to the extent permitted by law; without # even the implied warranty of MERCHANTABILITY or FITNESS FOR A # PARTICULAR PURPOSE. @SET_MAKE@ # include the list of codelets # This file contains a standard list of DFT SIMD codelets. It is # included by common/Makefile to generate the C files with the actual # codelets in them. It is included by {sse,sse2,...}/Makefile to # generate and compile stub files that include common/*.c # You can customize FFTW for special needs, e.g. to handle certain # sizes more efficiently, by adding new codelets to the lists of those # included by default. If you change the list of codelets, any new # ones you added will be automatically generated when you run the # bootstrap script (see "Generating your own code" in the FFTW # manual). # -*- makefile -*- # This file contains special make rules to generate codelets. # Most of this file requires GNU make . VPATH = @srcdir@ am__make_dryrun = \ { \ am__dry=no; \ case $$MAKEFLAGS in \ *\\[\ \ ]*) \ echo 'am--echo: ; @echo "AM" OK' | $(MAKE) -f - 2>/dev/null \ | grep '^AM OK$$' >/dev/null || am__dry=yes;; \ *) \ for am__flg in $$MAKEFLAGS; do \ case $$am__flg in \ *=*|--*) ;; \ *n*) am__dry=yes; break;; \ esac; \ done;; \ esac; \ test $$am__dry = yes; \ } pkgdatadir = $(datadir)/@PACKAGE@ pkgincludedir = $(includedir)/@PACKAGE@ pkglibdir = $(libdir)/@PACKAGE@ pkglibexecdir = $(libexecdir)/@PACKAGE@ am__cd = CDPATH="$${ZSH_VERSION+.}$(PATH_SEPARATOR)" && cd install_sh_DATA = $(install_sh) -c -m 644 install_sh_PROGRAM = $(install_sh) -c install_sh_SCRIPT = $(install_sh) -c INSTALL_HEADER = $(INSTALL_DATA) transform = $(program_transform_name) NORMAL_INSTALL = : PRE_INSTALL = : POST_INSTALL = : NORMAL_UNINSTALL = : PRE_UNINSTALL = : POST_UNINSTALL = : build_triplet = @build@ host_triplet = @host@ DIST_COMMON = $(srcdir)/Makefile.am $(srcdir)/Makefile.in \ $(top_srcdir)/dft/simd/codlist.mk \ $(top_srcdir)/support/Makefile.codelets subdir = dft/simd/common ACLOCAL_M4 = $(top_srcdir)/aclocal.m4 am__aclocal_m4_deps = $(top_srcdir)/m4/acx_mpi.m4 \ $(top_srcdir)/m4/acx_pthread.m4 \ $(top_srcdir)/m4/ax_cc_maxopt.m4 \ $(top_srcdir)/m4/ax_check_compiler_flags.m4 \ $(top_srcdir)/m4/ax_compiler_vendor.m4 \ $(top_srcdir)/m4/ax_gcc_aligns_stack.m4 \ $(top_srcdir)/m4/ax_gcc_version.m4 \ $(top_srcdir)/m4/ax_openmp.m4 $(top_srcdir)/m4/libtool.m4 \ $(top_srcdir)/m4/ltoptions.m4 $(top_srcdir)/m4/ltsugar.m4 \ $(top_srcdir)/m4/ltversion.m4 $(top_srcdir)/m4/lt~obsolete.m4 \ $(top_srcdir)/configure.ac am__configure_deps = $(am__aclocal_m4_deps) $(CONFIGURE_DEPENDENCIES) \ $(ACLOCAL_M4) mkinstalldirs = $(install_sh) -d CONFIG_HEADER = $(top_builddir)/config.h CONFIG_CLEAN_FILES = CONFIG_CLEAN_VPATH_FILES = SOURCES = DIST_SOURCES = am__can_run_installinfo = \ case $$AM_UPDATE_INFO_DIR in \ n|no|NO) false;; \ *) (install-info --version) >/dev/null 2>&1;; \ esac DISTFILES = $(DIST_COMMON) $(DIST_SOURCES) $(TEXINFOS) $(EXTRA_DIST) ACLOCAL = @ACLOCAL@ ALLOCA = @ALLOCA@ ALTIVEC_CFLAGS = @ALTIVEC_CFLAGS@ AMTAR = @AMTAR@ AR = @AR@ AS = @AS@ AUTOCONF = @AUTOCONF@ AUTOHEADER = @AUTOHEADER@ AUTOMAKE = @AUTOMAKE@ AVX_CFLAGS = @AVX_CFLAGS@ AWK = @AWK@ CC = @CC@ CCDEPMODE = @CCDEPMODE@ CFLAGS = @CFLAGS@ CHECK_PL_OPTS = @CHECK_PL_OPTS@ CPP = @CPP@ CPPFLAGS = @CPPFLAGS@ CYGPATH_W = @CYGPATH_W@ C_FFTW_R2R_KIND = @C_FFTW_R2R_KIND@ C_MPI_FINT = @C_MPI_FINT@ DEFS = @DEFS@ DEPDIR = @DEPDIR@ DLLTOOL = @DLLTOOL@ DSYMUTIL = @DSYMUTIL@ DUMPBIN = @DUMPBIN@ ECHO_C = @ECHO_C@ ECHO_N = @ECHO_N@ ECHO_T = @ECHO_T@ EGREP = @EGREP@ EXEEXT = @EXEEXT@ F77 = @F77@ FFLAGS = @FFLAGS@ FGREP = @FGREP@ FLIBS = @FLIBS@ GREP = @GREP@ INSTALL = @INSTALL@ INSTALL_DATA = @INSTALL_DATA@ INSTALL_PROGRAM = @INSTALL_PROGRAM@ INSTALL_SCRIPT = @INSTALL_SCRIPT@ INSTALL_STRIP_PROGRAM = @INSTALL_STRIP_PROGRAM@ LD = @LD@ LDFLAGS = @LDFLAGS@ LIBOBJS = @LIBOBJS@ LIBQUADMATH = @LIBQUADMATH@ LIBS = @LIBS@ LIBTOOL = @LIBTOOL@ LIPO = @LIPO@ LN_S = @LN_S@ LTLIBOBJS = @LTLIBOBJS@ MAINT = @MAINT@ MAKEINFO = @MAKEINFO@ MANIFEST_TOOL = @MANIFEST_TOOL@ MKDIR_P = @MKDIR_P@ MPICC = @MPICC@ MPILIBS = @MPILIBS@ MPIRUN = @MPIRUN@ NEON_CFLAGS = @NEON_CFLAGS@ NM = @NM@ NMEDIT = @NMEDIT@ OBJDUMP = @OBJDUMP@ OBJEXT = @OBJEXT@ OCAMLBUILD = @OCAMLBUILD@ OPENMP_CFLAGS = @OPENMP_CFLAGS@ OTOOL = @OTOOL@ OTOOL64 = @OTOOL64@ PACKAGE = @PACKAGE@ PACKAGE_BUGREPORT = @PACKAGE_BUGREPORT@ PACKAGE_NAME = @PACKAGE_NAME@ PACKAGE_STRING = @PACKAGE_STRING@ PACKAGE_TARNAME = @PACKAGE_TARNAME@ PACKAGE_URL = @PACKAGE_URL@ PACKAGE_VERSION = @PACKAGE_VERSION@ PATH_SEPARATOR = @PATH_SEPARATOR@ POW_LIB = @POW_LIB@ PRECISION = @PRECISION@ PREC_SUFFIX = @PREC_SUFFIX@ PTHREAD_CC = @PTHREAD_CC@ PTHREAD_CFLAGS = @PTHREAD_CFLAGS@ PTHREAD_LIBS = @PTHREAD_LIBS@ RANLIB = @RANLIB@ SED = @SED@ SET_MAKE = @SET_MAKE@ SHARED_VERSION_INFO = @SHARED_VERSION_INFO@ SHELL = @SHELL@ SSE2_CFLAGS = @SSE2_CFLAGS@ STACK_ALIGN_CFLAGS = @STACK_ALIGN_CFLAGS@ STRIP = @STRIP@ THREADLIBS = @THREADLIBS@ VERSION = @VERSION@ abs_builddir = @abs_builddir@ abs_srcdir = @abs_srcdir@ abs_top_builddir = @abs_top_builddir@ abs_top_srcdir = @abs_top_srcdir@ ac_ct_AR = @ac_ct_AR@ ac_ct_CC = @ac_ct_CC@ ac_ct_DUMPBIN = @ac_ct_DUMPBIN@ ac_ct_F77 = @ac_ct_F77@ acx_pthread_config = @acx_pthread_config@ am__include = @am__include@ am__leading_dot = @am__leading_dot@ am__quote = @am__quote@ am__tar = @am__tar@ am__untar = @am__untar@ bindir = @bindir@ build = @build@ build_alias = @build_alias@ build_cpu = @build_cpu@ build_os = @build_os@ build_vendor = @build_vendor@ builddir = @builddir@ datadir = @datadir@ datarootdir = @datarootdir@ docdir = @docdir@ dvidir = @dvidir@ exec_prefix = @exec_prefix@ host = @host@ host_alias = @host_alias@ host_cpu = @host_cpu@ host_os = @host_os@ host_vendor = @host_vendor@ htmldir = @htmldir@ includedir = @includedir@ infodir = @infodir@ install_sh = @install_sh@ libdir = @libdir@ libexecdir = @libexecdir@ localedir = @localedir@ localstatedir = @localstatedir@ mandir = @mandir@ mkdir_p = @mkdir_p@ oldincludedir = @oldincludedir@ pdfdir = @pdfdir@ prefix = @prefix@ program_transform_name = @program_transform_name@ psdir = @psdir@ sbindir = @sbindir@ sharedstatedir = @sharedstatedir@ srcdir = @srcdir@ sysconfdir = @sysconfdir@ target_alias = @target_alias@ top_build_prefix = @top_build_prefix@ top_builddir = @top_builddir@ top_srcdir = @top_srcdir@ ########################################################################### # n1fv_ is a hard-coded FFTW_FORWARD FFT of size , using SIMD N1F = n1fv_2.c n1fv_3.c n1fv_4.c n1fv_5.c n1fv_6.c n1fv_7.c n1fv_8.c \ n1fv_9.c n1fv_10.c n1fv_11.c n1fv_12.c n1fv_13.c n1fv_14.c n1fv_15.c \ n1fv_16.c n1fv_32.c n1fv_64.c n1fv_128.c n1fv_20.c n1fv_25.c # as above, with restricted input vector stride N2F = n2fv_2.c n2fv_4.c n2fv_6.c n2fv_8.c n2fv_10.c n2fv_12.c \ n2fv_14.c n2fv_16.c n2fv_32.c n2fv_64.c n2fv_20.c # as above, but FFTW_BACKWARD N1B = n1bv_2.c n1bv_3.c n1bv_4.c n1bv_5.c n1bv_6.c n1bv_7.c n1bv_8.c \ n1bv_9.c n1bv_10.c n1bv_11.c n1bv_12.c n1bv_13.c n1bv_14.c n1bv_15.c \ n1bv_16.c n1bv_32.c n1bv_64.c n1bv_128.c n1bv_20.c n1bv_25.c N2B = n2bv_2.c n2bv_4.c n2bv_6.c n2bv_8.c n2bv_10.c n2bv_12.c \ n2bv_14.c n2bv_16.c n2bv_32.c n2bv_64.c n2bv_20.c # split-complex codelets N2S = n2sv_4.c n2sv_8.c n2sv_16.c n2sv_32.c n2sv_64.c ########################################################################### # t1fv_ is a "twiddle" FFT of size , implementing a radix-r DIT step # for an FFTW_FORWARD transform, using SIMD T1F = t1fv_2.c t1fv_3.c t1fv_4.c t1fv_5.c t1fv_6.c t1fv_7.c t1fv_8.c \ t1fv_9.c t1fv_10.c t1fv_12.c t1fv_15.c t1fv_16.c t1fv_32.c t1fv_64.c \ t1fv_20.c t1fv_25.c # same as t1fv_*, but with different twiddle storage scheme T2F = t2fv_2.c t2fv_4.c t2fv_8.c t2fv_16.c t2fv_32.c t2fv_64.c \ t2fv_5.c t2fv_10.c t2fv_20.c t2fv_25.c T3F = t3fv_4.c t3fv_8.c t3fv_16.c t3fv_32.c t3fv_5.c t3fv_10.c \ t3fv_20.c t3fv_25.c T1FU = t1fuv_2.c t1fuv_3.c t1fuv_4.c t1fuv_5.c t1fuv_6.c t1fuv_7.c \ t1fuv_8.c t1fuv_9.c t1fuv_10.c # as above, but FFTW_BACKWARD T1B = t1bv_2.c t1bv_3.c t1bv_4.c t1bv_5.c t1bv_6.c t1bv_7.c t1bv_8.c \ t1bv_9.c t1bv_10.c t1bv_12.c t1bv_15.c t1bv_16.c t1bv_32.c t1bv_64.c \ t1bv_20.c t1bv_25.c # same as t1bv_*, but with different twiddle storage scheme T2B = t2bv_2.c t2bv_4.c t2bv_8.c t2bv_16.c t2bv_32.c t2bv_64.c \ t2bv_5.c t2bv_10.c t2bv_20.c t2bv_25.c T3B = t3bv_4.c t3bv_8.c t3bv_16.c t3bv_32.c t3bv_5.c t3bv_10.c \ t3bv_20.c t3bv_25.c T1BU = t1buv_2.c t1buv_3.c t1buv_4.c t1buv_5.c t1buv_6.c t1buv_7.c \ t1buv_8.c t1buv_9.c t1buv_10.c # split-complex codelets T1S = t1sv_2.c t1sv_4.c t1sv_8.c t1sv_16.c t1sv_32.c T2S = t2sv_4.c t2sv_8.c t2sv_16.c t2sv_32.c ########################################################################### # q1fv_ is twiddle FFTW_FORWARD FFTs of size (DIF step), # where the output is transposed, using SIMD. This is used for # in-place transposes in sizes that are divisible by ^2. These # codelets have size ~ ^2, so you should probably not use # bigger than 8 or so. Q1F = q1fv_2.c q1fv_4.c q1fv_5.c q1fv_8.c # as above, but FFTW_BACKWARD Q1B = q1bv_2.c q1bv_4.c q1bv_5.c q1bv_8.c ########################################################################### SIMD_CODELETS = $(N1F) $(N1B) $(N2F) $(N2B) $(N2S) $(T1FU) $(T1F) \ $(T2F) $(T3F) $(T1BU) $(T1B) $(T2B) $(T3B) $(T1S) $(T2S) $(Q1F) $(Q1B) ALL_CODELETS = $(SIMD_CODELETS) BUILT_SOURCES = $(SIMD_CODELETS) $(CODLIST) EXTRA_DIST = $(BUILT_SOURCES) genus.c INCLUDE_SIMD_HEADER = "\#include SIMD_HEADER" XRENAME = XSIMD SOLVTAB_NAME = XSIMD(solvtab_dft) CODLIST = codlist.c CODELET_NAME = codelet_ @MAINTAINER_MODE_TRUE@INDENT = indent -kr -cs -i5 -l800 -fca -nfc1 -sc -sob -cli4 -TR -Tplanner -TV @MAINTAINER_MODE_TRUE@TWOVERS = sh ${top_srcdir}/support/twovers.sh @MAINTAINER_MODE_TRUE@GENFFTDIR = ${top_builddir}/genfft @MAINTAINER_MODE_TRUE@GEN_NOTW = ${GENFFTDIR}/gen_notw.native @MAINTAINER_MODE_TRUE@GEN_NOTW_C = ${GENFFTDIR}/gen_notw_c.native @MAINTAINER_MODE_TRUE@GEN_TWIDDLE = ${GENFFTDIR}/gen_twiddle.native @MAINTAINER_MODE_TRUE@GEN_TWIDDLE_C = ${GENFFTDIR}/gen_twiddle_c.native @MAINTAINER_MODE_TRUE@GEN_TWIDSQ = ${GENFFTDIR}/gen_twidsq.native @MAINTAINER_MODE_TRUE@GEN_TWIDSQ_C = ${GENFFTDIR}/gen_twidsq_c.native @MAINTAINER_MODE_TRUE@GEN_R2CF = ${GENFFTDIR}/gen_r2cf.native @MAINTAINER_MODE_TRUE@GEN_R2CB = ${GENFFTDIR}/gen_r2cb.native @MAINTAINER_MODE_TRUE@GEN_HC2HC = ${GENFFTDIR}/gen_hc2hc.native @MAINTAINER_MODE_TRUE@GEN_HC2C = ${GENFFTDIR}/gen_hc2c.native @MAINTAINER_MODE_TRUE@GEN_HC2CDFT = ${GENFFTDIR}/gen_hc2cdft.native @MAINTAINER_MODE_TRUE@GEN_HC2CDFT_C = ${GENFFTDIR}/gen_hc2cdft_c.native @MAINTAINER_MODE_TRUE@GEN_R2R = ${GENFFTDIR}/gen_r2r.native @MAINTAINER_MODE_TRUE@PRELUDE_DFT = ${top_srcdir}/support/codelet_prelude.dft @MAINTAINER_MODE_TRUE@PRELUDE_RDFT = ${top_srcdir}/support/codelet_prelude.rdft @MAINTAINER_MODE_TRUE@ADD_DATE = sed -e s/@DATE@/"`date`"/ @MAINTAINER_MODE_TRUE@COPYRIGHT = ${top_srcdir}/COPYRIGHT @MAINTAINER_MODE_TRUE@CODELET_DEPS = $(COPYRIGHT) $(PRELUDE) @MAINTAINER_MODE_TRUE@PRELUDE_COMMANDS_DFT = cat $(COPYRIGHT) $(PRELUDE_DFT) @MAINTAINER_MODE_TRUE@PRELUDE_COMMANDS_RDFT = cat $(COPYRIGHT) $(PRELUDE_RDFT) @MAINTAINER_MODE_TRUE@FLAGS_COMMON = -compact -variables 4 @MAINTAINER_MODE_TRUE@DFT_FLAGS_COMMON = $(FLAGS_COMMON) -pipeline-latency 4 @MAINTAINER_MODE_TRUE@RDFT_FLAGS_COMMON = $(FLAGS_COMMON) -pipeline-latency 4 # include special rules for regenerating codelets. @MAINTAINER_MODE_TRUE@GFLAGS = -simd $(FLAGS_COMMON) -pipeline-latency 8 @MAINTAINER_MODE_TRUE@FLAGS_T2S = -twiddle-log3 -precompute-twiddles @MAINTAINER_MODE_TRUE@FLAGS_T3 = -twiddle-log3 -precompute-twiddles -no-generate-bytw all: $(BUILT_SOURCES) $(MAKE) $(AM_MAKEFLAGS) all-am .SUFFIXES: $(srcdir)/Makefile.in: @MAINTAINER_MODE_TRUE@ $(srcdir)/Makefile.am $(top_srcdir)/dft/simd/codlist.mk $(top_srcdir)/support/Makefile.codelets $(am__configure_deps) @for dep in $?; do \ case '$(am__configure_deps)' in \ *$$dep*) \ ( cd $(top_builddir) && $(MAKE) $(AM_MAKEFLAGS) am--refresh ) \ && { if test -f $@; then exit 0; else break; fi; }; \ exit 1;; \ esac; \ done; \ echo ' cd $(top_srcdir) && $(AUTOMAKE) --gnu dft/simd/common/Makefile'; \ $(am__cd) $(top_srcdir) && \ $(AUTOMAKE) --gnu dft/simd/common/Makefile .PRECIOUS: Makefile Makefile: $(srcdir)/Makefile.in $(top_builddir)/config.status @case '$?' in \ *config.status*) \ cd $(top_builddir) && $(MAKE) $(AM_MAKEFLAGS) am--refresh;; \ *) \ echo ' cd $(top_builddir) && $(SHELL) ./config.status $(subdir)/$@ $(am__depfiles_maybe)'; \ cd $(top_builddir) && $(SHELL) ./config.status $(subdir)/$@ $(am__depfiles_maybe);; \ esac; $(top_srcdir)/dft/simd/codlist.mk $(top_srcdir)/support/Makefile.codelets: $(top_builddir)/config.status: $(top_srcdir)/configure $(CONFIG_STATUS_DEPENDENCIES) cd $(top_builddir) && $(MAKE) $(AM_MAKEFLAGS) am--refresh $(top_srcdir)/configure: @MAINTAINER_MODE_TRUE@ $(am__configure_deps) cd $(top_builddir) && $(MAKE) $(AM_MAKEFLAGS) am--refresh $(ACLOCAL_M4): @MAINTAINER_MODE_TRUE@ $(am__aclocal_m4_deps) cd $(top_builddir) && $(MAKE) $(AM_MAKEFLAGS) am--refresh $(am__aclocal_m4_deps): mostlyclean-libtool: -rm -f *.lo clean-libtool: -rm -rf .libs _libs tags: TAGS TAGS: ctags: CTAGS CTAGS: distdir: $(DISTFILES) @srcdirstrip=`echo "$(srcdir)" | sed 's/[].[^$$\\*]/\\\\&/g'`; \ topsrcdirstrip=`echo "$(top_srcdir)" | sed 's/[].[^$$\\*]/\\\\&/g'`; \ list='$(DISTFILES)'; \ dist_files=`for file in $$list; do echo $$file; done | \ sed -e "s|^$$srcdirstrip/||;t" \ -e "s|^$$topsrcdirstrip/|$(top_builddir)/|;t"`; \ case $$dist_files in \ */*) $(MKDIR_P) `echo "$$dist_files" | \ sed '/\//!d;s|^|$(distdir)/|;s,/[^/]*$$,,' | \ sort -u` ;; \ esac; \ for file in $$dist_files; do \ if test -f $$file || test -d $$file; then d=.; else d=$(srcdir); fi; \ if test -d $$d/$$file; then \ dir=`echo "/$$file" | sed -e 's,/[^/]*$$,,'`; \ if test -d "$(distdir)/$$file"; then \ find "$(distdir)/$$file" -type d ! -perm -700 -exec chmod u+rwx {} \;; \ fi; \ if test -d $(srcdir)/$$file && test $$d != $(srcdir); then \ cp -fpR $(srcdir)/$$file "$(distdir)$$dir" || exit 1; \ find "$(distdir)/$$file" -type d ! -perm -700 -exec chmod u+rwx {} \;; \ fi; \ cp -fpR $$d/$$file "$(distdir)$$dir" || exit 1; \ else \ test -f "$(distdir)/$$file" \ || cp -p $$d/$$file "$(distdir)/$$file" \ || exit 1; \ fi; \ done check-am: all-am check: $(BUILT_SOURCES) $(MAKE) $(AM_MAKEFLAGS) check-am all-am: Makefile installdirs: install: $(BUILT_SOURCES) $(MAKE) $(AM_MAKEFLAGS) install-am install-exec: install-exec-am install-data: install-data-am uninstall: uninstall-am install-am: all-am @$(MAKE) $(AM_MAKEFLAGS) install-exec-am install-data-am installcheck: installcheck-am install-strip: if test -z '$(STRIP)'; then \ $(MAKE) $(AM_MAKEFLAGS) INSTALL_PROGRAM="$(INSTALL_STRIP_PROGRAM)" \ install_sh_PROGRAM="$(INSTALL_STRIP_PROGRAM)" INSTALL_STRIP_FLAG=-s \ install; \ else \ $(MAKE) $(AM_MAKEFLAGS) INSTALL_PROGRAM="$(INSTALL_STRIP_PROGRAM)" \ install_sh_PROGRAM="$(INSTALL_STRIP_PROGRAM)" INSTALL_STRIP_FLAG=-s \ "INSTALL_PROGRAM_ENV=STRIPPROG='$(STRIP)'" install; \ fi mostlyclean-generic: clean-generic: distclean-generic: -test -z "$(CONFIG_CLEAN_FILES)" || rm -f $(CONFIG_CLEAN_FILES) -test . = "$(srcdir)" || test -z "$(CONFIG_CLEAN_VPATH_FILES)" || rm -f $(CONFIG_CLEAN_VPATH_FILES) maintainer-clean-generic: @echo "This command is intended for maintainers to use" @echo "it deletes files that may require special tools to rebuild." -test -z "$(BUILT_SOURCES)" || rm -f $(BUILT_SOURCES) clean: clean-am clean-am: clean-generic clean-libtool mostlyclean-am distclean: distclean-am -rm -f Makefile distclean-am: clean-am distclean-generic dvi: dvi-am dvi-am: html: html-am html-am: info: info-am info-am: install-data-am: install-dvi: install-dvi-am install-dvi-am: install-exec-am: install-html: install-html-am install-html-am: install-info: install-info-am install-info-am: install-man: install-pdf: install-pdf-am install-pdf-am: install-ps: install-ps-am install-ps-am: installcheck-am: maintainer-clean: maintainer-clean-am -rm -f Makefile maintainer-clean-am: distclean-am maintainer-clean-generic \ maintainer-clean-local mostlyclean: mostlyclean-am mostlyclean-am: mostlyclean-generic mostlyclean-libtool pdf: pdf-am pdf-am: ps: ps-am ps-am: uninstall-am: .MAKE: all check install install-am install-strip .PHONY: all all-am check check-am clean clean-generic clean-libtool \ distclean distclean-generic distclean-libtool distdir dvi \ dvi-am html html-am info info-am install install-am \ install-data install-data-am install-dvi install-dvi-am \ install-exec install-exec-am install-html install-html-am \ install-info install-info-am install-man install-pdf \ install-pdf-am install-ps install-ps-am install-strip \ installcheck installcheck-am installdirs maintainer-clean \ maintainer-clean-generic maintainer-clean-local mostlyclean \ mostlyclean-generic mostlyclean-libtool pdf pdf-am ps ps-am \ uninstall uninstall-am # rule to build codlist $(CODLIST): Makefile ( \ echo "#include \"ifftw.h\""; \ echo $(INCLUDE_SIMD_HEADER); \ echo; \ for i in $(ALL_CODELETS) NIL; do \ if test "$$i" != NIL; then \ j=`basename $$i | sed -e 's/[.][cS]$$//g'`; \ echo "extern void $(XRENAME)($(CODELET_NAME)$$j)(planner *);"; \ fi \ done; \ echo; \ echo; \ echo "extern const solvtab $(SOLVTAB_NAME);"; \ echo "const solvtab $(SOLVTAB_NAME) = {"; \ for i in $(ALL_CODELETS) NIL; do \ if test "$$i" != NIL; then \ j=`basename $$i | sed -e 's/[.][cS]$$//g'`; \ echo " SOLVTAB($(XRENAME)($(CODELET_NAME)$$j)),"; \ fi \ done; \ echo " SOLVTAB_END"; \ echo "};"; \ ) >$@ # only delete codlist.c in maintainer-mode, since it is included in the dist # FIXME: is there a way to delete in 'make clean' only when builddir != srcdir? maintainer-clean-local: rm -f $(CODLIST) # cancel the hideous builtin rules that cause an infinite loop @MAINTAINER_MODE_TRUE@%: %.o @MAINTAINER_MODE_TRUE@%: %.s @MAINTAINER_MODE_TRUE@%: %.c @MAINTAINER_MODE_TRUE@%: %.S @MAINTAINER_MODE_TRUE@n1fv_%.c: $(CODELET_DEPS) $(GEN_NOTW_C) @MAINTAINER_MODE_TRUE@ ($(PRELUDE_COMMANDS_DFT); $(TWOVERS) $(GEN_NOTW_C) $(GFLAGS) -n $* -name n1fv_$* -include "n1f.h") | $(ADD_DATE) | $(INDENT) >$@ @MAINTAINER_MODE_TRUE@n2fv_%.c: $(CODELET_DEPS) $(GEN_NOTW_C) @MAINTAINER_MODE_TRUE@ ($(PRELUDE_COMMANDS_DFT); $(TWOVERS) $(GEN_NOTW_C) $(GFLAGS) -n $* -name n2fv_$* -with-ostride 2 -include "n2f.h" -store-multiple 2) | $(ADD_DATE) | $(INDENT) >$@ @MAINTAINER_MODE_TRUE@n1bv_%.c: $(CODELET_DEPS) $(GEN_NOTW_C) @MAINTAINER_MODE_TRUE@ ($(PRELUDE_COMMANDS_DFT); $(TWOVERS) $(GEN_NOTW_C) $(GFLAGS) -sign 1 -n $* -name n1bv_$* -include "n1b.h") | $(ADD_DATE) | $(INDENT) >$@ @MAINTAINER_MODE_TRUE@n2bv_%.c: $(CODELET_DEPS) $(GEN_NOTW_C) @MAINTAINER_MODE_TRUE@ ($(PRELUDE_COMMANDS_DFT); $(TWOVERS) $(GEN_NOTW_C) $(GFLAGS) -sign 1 -n $* -name n2bv_$* -with-ostride 2 -include "n2b.h" -store-multiple 2) | $(ADD_DATE) | $(INDENT) >$@ @MAINTAINER_MODE_TRUE@n2sv_%.c: $(CODELET_DEPS) $(GEN_NOTW) @MAINTAINER_MODE_TRUE@ ($(PRELUDE_COMMANDS_DFT); $(TWOVERS) $(GEN_NOTW) $(GFLAGS) -n $* -name n2sv_$* -with-ostride 1 -include "n2s.h" -store-multiple 4) | $(ADD_DATE) | $(INDENT) >$@ @MAINTAINER_MODE_TRUE@t1fv_%.c: $(CODELET_DEPS) $(GEN_TWIDDLE_C) @MAINTAINER_MODE_TRUE@ ($(PRELUDE_COMMANDS_DFT); $(TWOVERS) $(GEN_TWIDDLE_C) $(GFLAGS) -n $* -name t1fv_$* -include "t1f.h") | $(ADD_DATE) | $(INDENT) >$@ @MAINTAINER_MODE_TRUE@t1fuv_%.c: $(CODELET_DEPS) $(GEN_TWIDDLE_C) @MAINTAINER_MODE_TRUE@ ($(PRELUDE_COMMANDS_DFT); $(TWOVERS) $(GEN_TWIDDLE_C) $(GFLAGS) -n $* -name t1fuv_$* -include "t1fu.h") | $(ADD_DATE) | $(INDENT) >$@ @MAINTAINER_MODE_TRUE@t2fv_%.c: $(CODELET_DEPS) $(GEN_TWIDDLE_C) @MAINTAINER_MODE_TRUE@ ($(PRELUDE_COMMANDS_DFT); $(TWOVERS) $(GEN_TWIDDLE_C) $(GFLAGS) -n $* -name t2fv_$* -include "t2f.h") | $(ADD_DATE) | $(INDENT) >$@ @MAINTAINER_MODE_TRUE@t3fv_%.c: $(CODELET_DEPS) $(GEN_TWIDDLE_C) @MAINTAINER_MODE_TRUE@ ($(PRELUDE_COMMANDS_DFT); $(TWOVERS) $(GEN_TWIDDLE_C) $(GFLAGS) $(FLAGS_T3) -n $* -name t3fv_$* -include "t3f.h") | $(ADD_DATE) | $(INDENT) >$@ @MAINTAINER_MODE_TRUE@t1bv_%.c: $(CODELET_DEPS) $(GEN_TWIDDLE_C) @MAINTAINER_MODE_TRUE@ ($(PRELUDE_COMMANDS_DFT); $(TWOVERS) $(GEN_TWIDDLE_C) $(GFLAGS) -n $* -name t1bv_$* -include "t1b.h" -sign 1) | $(ADD_DATE) | $(INDENT) >$@ @MAINTAINER_MODE_TRUE@t1buv_%.c: $(CODELET_DEPS) $(GEN_TWIDDLE_C) @MAINTAINER_MODE_TRUE@ ($(PRELUDE_COMMANDS_DFT); $(TWOVERS) $(GEN_TWIDDLE_C) $(GFLAGS) -n $* -name t1buv_$* -include "t1bu.h" -sign 1) | $(ADD_DATE) | $(INDENT) >$@ @MAINTAINER_MODE_TRUE@t2bv_%.c: $(CODELET_DEPS) $(GEN_TWIDDLE_C) @MAINTAINER_MODE_TRUE@ ($(PRELUDE_COMMANDS_DFT); $(TWOVERS) $(GEN_TWIDDLE_C) $(GFLAGS) -n $* -name t2bv_$* -include "t2b.h" -sign 1) | $(ADD_DATE) | $(INDENT) >$@ @MAINTAINER_MODE_TRUE@t3bv_%.c: $(CODELET_DEPS) $(GEN_TWIDDLE_C) @MAINTAINER_MODE_TRUE@ ($(PRELUDE_COMMANDS_DFT); $(TWOVERS) $(GEN_TWIDDLE_C) $(GFLAGS) $(FLAGS_T3) -n $* -name t3bv_$* -include "t3b.h" -sign 1) | $(ADD_DATE) | $(INDENT) >$@ @MAINTAINER_MODE_TRUE@t1sv_%.c: $(CODELET_DEPS) $(GEN_TWIDDLE) @MAINTAINER_MODE_TRUE@ ($(PRELUDE_COMMANDS_DFT); $(TWOVERS) $(GEN_TWIDDLE) $(GFLAGS) -n $* -name t1sv_$* -include "ts.h") | $(ADD_DATE) | $(INDENT) >$@ @MAINTAINER_MODE_TRUE@t2sv_%.c: $(CODELET_DEPS) $(GEN_TWIDDLE) @MAINTAINER_MODE_TRUE@ ($(PRELUDE_COMMANDS_DFT); $(TWOVERS) $(GEN_TWIDDLE) $(GFLAGS) $(FLAGS_T2S) -n $* -name t2sv_$* -include "ts.h") | $(ADD_DATE) | $(INDENT) >$@ @MAINTAINER_MODE_TRUE@q1fv_%.c: $(CODELET_DEPS) $(GEN_TWIDSQ_C) @MAINTAINER_MODE_TRUE@ ($(PRELUDE_COMMANDS_DFT); $(TWOVERS) $(GEN_TWIDSQ_C) $(GFLAGS) -n $* -dif -name q1fv_$* -include "q1f.h") | $(ADD_DATE) | $(INDENT) >$@ @MAINTAINER_MODE_TRUE@q1bv_%.c: $(CODELET_DEPS) $(GEN_TWIDSQ_C) @MAINTAINER_MODE_TRUE@ ($(PRELUDE_COMMANDS_DFT); $(TWOVERS) $(GEN_TWIDSQ_C) $(GFLAGS) -n $* -dif -name q1bv_$* -include "q1b.h" -sign 1) | $(ADD_DATE) | $(INDENT) >$@ # Tell versions [3.59,3.63) of GNU make to not export all variables. # Otherwise a system limit (for SysV at least) may be exceeded. .NOEXPORT: fftw-3.3.3/dft/simd/common/n1fv_2.c0000644000175000017500000000635012054410344013650 00000000000000/* * Copyright (c) 2003, 2007-11 Matteo Frigo * Copyright (c) 2003, 2007-11 Massachusetts Institute of Technology * * This program is free software; you can redistribute it and/or modify * it under the terms of the GNU General Public License as published by * the Free Software Foundation; either version 2 of the License, or * (at your option) any later version. * * This program is distributed in the hope that it will be useful, * but WITHOUT ANY WARRANTY; without even the implied warranty of * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the * GNU General Public License for more details. * * You should have received a copy of the GNU General Public License * along with this program; if not, write to the Free Software * Foundation, Inc., 51 Franklin Street, Fifth Floor, Boston, MA 02110-1301 USA * */ /* This file was automatically generated --- DO NOT EDIT */ /* Generated on Sun Nov 25 07:36:52 EST 2012 */ #include "codelet-dft.h" #ifdef HAVE_FMA /* Generated by: ../../../genfft/gen_notw_c.native -fma -reorder-insns -schedule-for-pipeline -simd -compact -variables 4 -pipeline-latency 8 -n 2 -name n1fv_2 -include n1f.h */ /* * This function contains 2 FP additions, 0 FP multiplications, * (or, 2 additions, 0 multiplications, 0 fused multiply/add), * 5 stack variables, 0 constants, and 4 memory accesses */ #include "n1f.h" static void n1fv_2(const R *ri, const R *ii, R *ro, R *io, stride is, stride os, INT v, INT ivs, INT ovs) { { INT i; const R *xi; R *xo; xi = ri; xo = ro; for (i = v; i > 0; i = i - VL, xi = xi + (VL * ivs), xo = xo + (VL * ovs), MAKE_VOLATILE_STRIDE(4, is), MAKE_VOLATILE_STRIDE(4, os)) { V T1, T2; T1 = LD(&(xi[0]), ivs, &(xi[0])); T2 = LD(&(xi[WS(is, 1)]), ivs, &(xi[WS(is, 1)])); ST(&(xo[0]), VADD(T1, T2), ovs, &(xo[0])); ST(&(xo[WS(os, 1)]), VSUB(T1, T2), ovs, &(xo[WS(os, 1)])); } } VLEAVE(); } static const kdft_desc desc = { 2, XSIMD_STRING("n1fv_2"), {2, 0, 0, 0}, &GENUS, 0, 0, 0, 0 }; void XSIMD(codelet_n1fv_2) (planner *p) { X(kdft_register) (p, n1fv_2, &desc); } #else /* HAVE_FMA */ /* Generated by: ../../../genfft/gen_notw_c.native -simd -compact -variables 4 -pipeline-latency 8 -n 2 -name n1fv_2 -include n1f.h */ /* * This function contains 2 FP additions, 0 FP multiplications, * (or, 2 additions, 0 multiplications, 0 fused multiply/add), * 5 stack variables, 0 constants, and 4 memory accesses */ #include "n1f.h" static void n1fv_2(const R *ri, const R *ii, R *ro, R *io, stride is, stride os, INT v, INT ivs, INT ovs) { { INT i; const R *xi; R *xo; xi = ri; xo = ro; for (i = v; i > 0; i = i - VL, xi = xi + (VL * ivs), xo = xo + (VL * ovs), MAKE_VOLATILE_STRIDE(4, is), MAKE_VOLATILE_STRIDE(4, os)) { V T1, T2; T1 = LD(&(xi[0]), ivs, &(xi[0])); T2 = LD(&(xi[WS(is, 1)]), ivs, &(xi[WS(is, 1)])); ST(&(xo[WS(os, 1)]), VSUB(T1, T2), ovs, &(xo[WS(os, 1)])); ST(&(xo[0]), VADD(T1, T2), ovs, &(xo[0])); } } VLEAVE(); } static const kdft_desc desc = { 2, XSIMD_STRING("n1fv_2"), {2, 0, 0, 0}, &GENUS, 0, 0, 0, 0 }; void XSIMD(codelet_n1fv_2) (planner *p) { X(kdft_register) (p, n1fv_2, &desc); } #endif /* HAVE_FMA */ fftw-3.3.3/dft/simd/common/n1fv_3.c0000644000175000017500000001001112054410344013636 00000000000000/* * Copyright (c) 2003, 2007-11 Matteo Frigo * Copyright (c) 2003, 2007-11 Massachusetts Institute of Technology * * This program is free software; you can redistribute it and/or modify * it under the terms of the GNU General Public License as published by * the Free Software Foundation; either version 2 of the License, or * (at your option) any later version. * * This program is distributed in the hope that it will be useful, * but WITHOUT ANY WARRANTY; without even the implied warranty of * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the * GNU General Public License for more details. * * You should have received a copy of the GNU General Public License * along with this program; if not, write to the Free Software * Foundation, Inc., 51 Franklin Street, Fifth Floor, Boston, MA 02110-1301 USA * */ /* This file was automatically generated --- DO NOT EDIT */ /* Generated on Sun Nov 25 07:36:52 EST 2012 */ #include "codelet-dft.h" #ifdef HAVE_FMA /* Generated by: ../../../genfft/gen_notw_c.native -fma -reorder-insns -schedule-for-pipeline -simd -compact -variables 4 -pipeline-latency 8 -n 3 -name n1fv_3 -include n1f.h */ /* * This function contains 6 FP additions, 4 FP multiplications, * (or, 3 additions, 1 multiplications, 3 fused multiply/add), * 11 stack variables, 2 constants, and 6 memory accesses */ #include "n1f.h" static void n1fv_3(const R *ri, const R *ii, R *ro, R *io, stride is, stride os, INT v, INT ivs, INT ovs) { DVK(KP500000000, +0.500000000000000000000000000000000000000000000); DVK(KP866025403, +0.866025403784438646763723170752936183471402627); { INT i; const R *xi; R *xo; xi = ri; xo = ro; for (i = v; i > 0; i = i - VL, xi = xi + (VL * ivs), xo = xo + (VL * ovs), MAKE_VOLATILE_STRIDE(6, is), MAKE_VOLATILE_STRIDE(6, os)) { V T1, T2, T3, T6, T4, T5; T1 = LD(&(xi[0]), ivs, &(xi[0])); T2 = LD(&(xi[WS(is, 1)]), ivs, &(xi[WS(is, 1)])); T3 = LD(&(xi[WS(is, 2)]), ivs, &(xi[0])); T6 = VMUL(LDK(KP866025403), VSUB(T3, T2)); T4 = VADD(T2, T3); T5 = VFNMS(LDK(KP500000000), T4, T1); ST(&(xo[0]), VADD(T1, T4), ovs, &(xo[0])); ST(&(xo[WS(os, 1)]), VFMAI(T6, T5), ovs, &(xo[WS(os, 1)])); ST(&(xo[WS(os, 2)]), VFNMSI(T6, T5), ovs, &(xo[0])); } } VLEAVE(); } static const kdft_desc desc = { 3, XSIMD_STRING("n1fv_3"), {3, 1, 3, 0}, &GENUS, 0, 0, 0, 0 }; void XSIMD(codelet_n1fv_3) (planner *p) { X(kdft_register) (p, n1fv_3, &desc); } #else /* HAVE_FMA */ /* Generated by: ../../../genfft/gen_notw_c.native -simd -compact -variables 4 -pipeline-latency 8 -n 3 -name n1fv_3 -include n1f.h */ /* * This function contains 6 FP additions, 2 FP multiplications, * (or, 5 additions, 1 multiplications, 1 fused multiply/add), * 11 stack variables, 2 constants, and 6 memory accesses */ #include "n1f.h" static void n1fv_3(const R *ri, const R *ii, R *ro, R *io, stride is, stride os, INT v, INT ivs, INT ovs) { DVK(KP500000000, +0.500000000000000000000000000000000000000000000); DVK(KP866025403, +0.866025403784438646763723170752936183471402627); { INT i; const R *xi; R *xo; xi = ri; xo = ro; for (i = v; i > 0; i = i - VL, xi = xi + (VL * ivs), xo = xo + (VL * ovs), MAKE_VOLATILE_STRIDE(6, is), MAKE_VOLATILE_STRIDE(6, os)) { V T1, T4, T6, T2, T3, T5; T1 = LD(&(xi[0]), ivs, &(xi[0])); T2 = LD(&(xi[WS(is, 1)]), ivs, &(xi[WS(is, 1)])); T3 = LD(&(xi[WS(is, 2)]), ivs, &(xi[0])); T4 = VADD(T2, T3); T6 = VBYI(VMUL(LDK(KP866025403), VSUB(T3, T2))); ST(&(xo[0]), VADD(T1, T4), ovs, &(xo[0])); T5 = VFNMS(LDK(KP500000000), T4, T1); ST(&(xo[WS(os, 2)]), VSUB(T5, T6), ovs, &(xo[0])); ST(&(xo[WS(os, 1)]), VADD(T5, T6), ovs, &(xo[WS(os, 1)])); } } VLEAVE(); } static const kdft_desc desc = { 3, XSIMD_STRING("n1fv_3"), {5, 1, 1, 0}, &GENUS, 0, 0, 0, 0 }; void XSIMD(codelet_n1fv_3) (planner *p) { X(kdft_register) (p, n1fv_3, &desc); } #endif /* HAVE_FMA */ fftw-3.3.3/dft/simd/common/n1fv_4.c0000644000175000017500000000774712054410344013665 00000000000000/* * Copyright (c) 2003, 2007-11 Matteo Frigo * Copyright (c) 2003, 2007-11 Massachusetts Institute of Technology * * This program is free software; you can redistribute it and/or modify * it under the terms of the GNU General Public License as published by * the Free Software Foundation; either version 2 of the License, or * (at your option) any later version. * * This program is distributed in the hope that it will be useful, * but WITHOUT ANY WARRANTY; without even the implied warranty of * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the * GNU General Public License for more details. * * You should have received a copy of the GNU General Public License * along with this program; if not, write to the Free Software * Foundation, Inc., 51 Franklin Street, Fifth Floor, Boston, MA 02110-1301 USA * */ /* This file was automatically generated --- DO NOT EDIT */ /* Generated on Sun Nov 25 07:36:52 EST 2012 */ #include "codelet-dft.h" #ifdef HAVE_FMA /* Generated by: ../../../genfft/gen_notw_c.native -fma -reorder-insns -schedule-for-pipeline -simd -compact -variables 4 -pipeline-latency 8 -n 4 -name n1fv_4 -include n1f.h */ /* * This function contains 8 FP additions, 2 FP multiplications, * (or, 6 additions, 0 multiplications, 2 fused multiply/add), * 11 stack variables, 0 constants, and 8 memory accesses */ #include "n1f.h" static void n1fv_4(const R *ri, const R *ii, R *ro, R *io, stride is, stride os, INT v, INT ivs, INT ovs) { { INT i; const R *xi; R *xo; xi = ri; xo = ro; for (i = v; i > 0; i = i - VL, xi = xi + (VL * ivs), xo = xo + (VL * ovs), MAKE_VOLATILE_STRIDE(8, is), MAKE_VOLATILE_STRIDE(8, os)) { V T1, T2, T4, T5; T1 = LD(&(xi[0]), ivs, &(xi[0])); T2 = LD(&(xi[WS(is, 2)]), ivs, &(xi[0])); T4 = LD(&(xi[WS(is, 1)]), ivs, &(xi[WS(is, 1)])); T5 = LD(&(xi[WS(is, 3)]), ivs, &(xi[WS(is, 1)])); { V T3, T7, T6, T8; T3 = VSUB(T1, T2); T7 = VADD(T1, T2); T6 = VSUB(T4, T5); T8 = VADD(T4, T5); ST(&(xo[WS(os, 2)]), VSUB(T7, T8), ovs, &(xo[0])); ST(&(xo[0]), VADD(T7, T8), ovs, &(xo[0])); ST(&(xo[WS(os, 3)]), VFMAI(T6, T3), ovs, &(xo[WS(os, 1)])); ST(&(xo[WS(os, 1)]), VFNMSI(T6, T3), ovs, &(xo[WS(os, 1)])); } } } VLEAVE(); } static const kdft_desc desc = { 4, XSIMD_STRING("n1fv_4"), {6, 0, 2, 0}, &GENUS, 0, 0, 0, 0 }; void XSIMD(codelet_n1fv_4) (planner *p) { X(kdft_register) (p, n1fv_4, &desc); } #else /* HAVE_FMA */ /* Generated by: ../../../genfft/gen_notw_c.native -simd -compact -variables 4 -pipeline-latency 8 -n 4 -name n1fv_4 -include n1f.h */ /* * This function contains 8 FP additions, 0 FP multiplications, * (or, 8 additions, 0 multiplications, 0 fused multiply/add), * 11 stack variables, 0 constants, and 8 memory accesses */ #include "n1f.h" static void n1fv_4(const R *ri, const R *ii, R *ro, R *io, stride is, stride os, INT v, INT ivs, INT ovs) { { INT i; const R *xi; R *xo; xi = ri; xo = ro; for (i = v; i > 0; i = i - VL, xi = xi + (VL * ivs), xo = xo + (VL * ovs), MAKE_VOLATILE_STRIDE(8, is), MAKE_VOLATILE_STRIDE(8, os)) { V T3, T7, T6, T8; { V T1, T2, T4, T5; T1 = LD(&(xi[0]), ivs, &(xi[0])); T2 = LD(&(xi[WS(is, 2)]), ivs, &(xi[0])); T3 = VSUB(T1, T2); T7 = VADD(T1, T2); T4 = LD(&(xi[WS(is, 1)]), ivs, &(xi[WS(is, 1)])); T5 = LD(&(xi[WS(is, 3)]), ivs, &(xi[WS(is, 1)])); T6 = VBYI(VSUB(T4, T5)); T8 = VADD(T4, T5); } ST(&(xo[WS(os, 1)]), VSUB(T3, T6), ovs, &(xo[WS(os, 1)])); ST(&(xo[0]), VADD(T7, T8), ovs, &(xo[0])); ST(&(xo[WS(os, 3)]), VADD(T3, T6), ovs, &(xo[WS(os, 1)])); ST(&(xo[WS(os, 2)]), VSUB(T7, T8), ovs, &(xo[0])); } } VLEAVE(); } static const kdft_desc desc = { 4, XSIMD_STRING("n1fv_4"), {8, 0, 0, 0}, &GENUS, 0, 0, 0, 0 }; void XSIMD(codelet_n1fv_4) (planner *p) { X(kdft_register) (p, n1fv_4, &desc); } #endif /* HAVE_FMA */ fftw-3.3.3/dft/simd/common/n1fv_5.c0000644000175000017500000001273012054410344013652 00000000000000/* * Copyright (c) 2003, 2007-11 Matteo Frigo * Copyright (c) 2003, 2007-11 Massachusetts Institute of Technology * * This program is free software; you can redistribute it and/or modify * it under the terms of the GNU General Public License as published by * the Free Software Foundation; either version 2 of the License, or * (at your option) any later version. * * This program is distributed in the hope that it will be useful, * but WITHOUT ANY WARRANTY; without even the implied warranty of * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the * GNU General Public License for more details. * * You should have received a copy of the GNU General Public License * along with this program; if not, write to the Free Software * Foundation, Inc., 51 Franklin Street, Fifth Floor, Boston, MA 02110-1301 USA * */ /* This file was automatically generated --- DO NOT EDIT */ /* Generated on Sun Nov 25 07:36:52 EST 2012 */ #include "codelet-dft.h" #ifdef HAVE_FMA /* Generated by: ../../../genfft/gen_notw_c.native -fma -reorder-insns -schedule-for-pipeline -simd -compact -variables 4 -pipeline-latency 8 -n 5 -name n1fv_5 -include n1f.h */ /* * This function contains 16 FP additions, 11 FP multiplications, * (or, 7 additions, 2 multiplications, 9 fused multiply/add), * 23 stack variables, 4 constants, and 10 memory accesses */ #include "n1f.h" static void n1fv_5(const R *ri, const R *ii, R *ro, R *io, stride is, stride os, INT v, INT ivs, INT ovs) { DVK(KP559016994, +0.559016994374947424102293417182819058860154590); DVK(KP250000000, +0.250000000000000000000000000000000000000000000); DVK(KP618033988, +0.618033988749894848204586834365638117720309180); DVK(KP951056516, +0.951056516295153572116439333379382143405698634); { INT i; const R *xi; R *xo; xi = ri; xo = ro; for (i = v; i > 0; i = i - VL, xi = xi + (VL * ivs), xo = xo + (VL * ovs), MAKE_VOLATILE_STRIDE(10, is), MAKE_VOLATILE_STRIDE(10, os)) { V T1, T2, T3, T5, T6; T1 = LD(&(xi[0]), ivs, &(xi[0])); T2 = LD(&(xi[WS(is, 1)]), ivs, &(xi[WS(is, 1)])); T3 = LD(&(xi[WS(is, 4)]), ivs, &(xi[0])); T5 = LD(&(xi[WS(is, 2)]), ivs, &(xi[0])); T6 = LD(&(xi[WS(is, 3)]), ivs, &(xi[WS(is, 1)])); { V Tc, T4, Td, T7; Tc = VSUB(T2, T3); T4 = VADD(T2, T3); Td = VSUB(T5, T6); T7 = VADD(T5, T6); { V Tg, Te, Ta, T8, T9, Tf, Tb; Tg = VMUL(LDK(KP951056516), VFNMS(LDK(KP618033988), Tc, Td)); Te = VMUL(LDK(KP951056516), VFMA(LDK(KP618033988), Td, Tc)); Ta = VSUB(T4, T7); T8 = VADD(T4, T7); T9 = VFNMS(LDK(KP250000000), T8, T1); ST(&(xo[0]), VADD(T1, T8), ovs, &(xo[0])); Tf = VFNMS(LDK(KP559016994), Ta, T9); Tb = VFMA(LDK(KP559016994), Ta, T9); ST(&(xo[WS(os, 2)]), VFMAI(Tg, Tf), ovs, &(xo[0])); ST(&(xo[WS(os, 3)]), VFNMSI(Tg, Tf), ovs, &(xo[WS(os, 1)])); ST(&(xo[WS(os, 4)]), VFMAI(Te, Tb), ovs, &(xo[0])); ST(&(xo[WS(os, 1)]), VFNMSI(Te, Tb), ovs, &(xo[WS(os, 1)])); } } } } VLEAVE(); } static const kdft_desc desc = { 5, XSIMD_STRING("n1fv_5"), {7, 2, 9, 0}, &GENUS, 0, 0, 0, 0 }; void XSIMD(codelet_n1fv_5) (planner *p) { X(kdft_register) (p, n1fv_5, &desc); } #else /* HAVE_FMA */ /* Generated by: ../../../genfft/gen_notw_c.native -simd -compact -variables 4 -pipeline-latency 8 -n 5 -name n1fv_5 -include n1f.h */ /* * This function contains 16 FP additions, 6 FP multiplications, * (or, 13 additions, 3 multiplications, 3 fused multiply/add), * 18 stack variables, 4 constants, and 10 memory accesses */ #include "n1f.h" static void n1fv_5(const R *ri, const R *ii, R *ro, R *io, stride is, stride os, INT v, INT ivs, INT ovs) { DVK(KP250000000, +0.250000000000000000000000000000000000000000000); DVK(KP587785252, +0.587785252292473129168705954639072768597652438); DVK(KP951056516, +0.951056516295153572116439333379382143405698634); DVK(KP559016994, +0.559016994374947424102293417182819058860154590); { INT i; const R *xi; R *xo; xi = ri; xo = ro; for (i = v; i > 0; i = i - VL, xi = xi + (VL * ivs), xo = xo + (VL * ovs), MAKE_VOLATILE_STRIDE(10, is), MAKE_VOLATILE_STRIDE(10, os)) { V T8, T7, Td, T9, Tc; T8 = LD(&(xi[0]), ivs, &(xi[0])); { V T1, T2, T3, T4, T5, T6; T1 = LD(&(xi[WS(is, 1)]), ivs, &(xi[WS(is, 1)])); T2 = LD(&(xi[WS(is, 4)]), ivs, &(xi[0])); T3 = VADD(T1, T2); T4 = LD(&(xi[WS(is, 2)]), ivs, &(xi[0])); T5 = LD(&(xi[WS(is, 3)]), ivs, &(xi[WS(is, 1)])); T6 = VADD(T4, T5); T7 = VMUL(LDK(KP559016994), VSUB(T3, T6)); Td = VSUB(T4, T5); T9 = VADD(T3, T6); Tc = VSUB(T1, T2); } ST(&(xo[0]), VADD(T8, T9), ovs, &(xo[0])); { V Te, Tf, Tb, Tg, Ta; Te = VBYI(VFMA(LDK(KP951056516), Tc, VMUL(LDK(KP587785252), Td))); Tf = VBYI(VFNMS(LDK(KP587785252), Tc, VMUL(LDK(KP951056516), Td))); Ta = VFNMS(LDK(KP250000000), T9, T8); Tb = VADD(T7, Ta); Tg = VSUB(Ta, T7); ST(&(xo[WS(os, 1)]), VSUB(Tb, Te), ovs, &(xo[WS(os, 1)])); ST(&(xo[WS(os, 3)]), VSUB(Tg, Tf), ovs, &(xo[WS(os, 1)])); ST(&(xo[WS(os, 4)]), VADD(Te, Tb), ovs, &(xo[0])); ST(&(xo[WS(os, 2)]), VADD(Tf, Tg), ovs, &(xo[0])); } } } VLEAVE(); } static const kdft_desc desc = { 5, XSIMD_STRING("n1fv_5"), {13, 3, 3, 0}, &GENUS, 0, 0, 0, 0 }; void XSIMD(codelet_n1fv_5) (planner *p) { X(kdft_register) (p, n1fv_5, &desc); } #endif /* HAVE_FMA */ fftw-3.3.3/dft/simd/common/n1fv_6.c0000644000175000017500000001262612054410344013657 00000000000000/* * Copyright (c) 2003, 2007-11 Matteo Frigo * Copyright (c) 2003, 2007-11 Massachusetts Institute of Technology * * This program is free software; you can redistribute it and/or modify * it under the terms of the GNU General Public License as published by * the Free Software Foundation; either version 2 of the License, or * (at your option) any later version. * * This program is distributed in the hope that it will be useful, * but WITHOUT ANY WARRANTY; without even the implied warranty of * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the * GNU General Public License for more details. * * You should have received a copy of the GNU General Public License * along with this program; if not, write to the Free Software * Foundation, Inc., 51 Franklin Street, Fifth Floor, Boston, MA 02110-1301 USA * */ /* This file was automatically generated --- DO NOT EDIT */ /* Generated on Sun Nov 25 07:36:52 EST 2012 */ #include "codelet-dft.h" #ifdef HAVE_FMA /* Generated by: ../../../genfft/gen_notw_c.native -fma -reorder-insns -schedule-for-pipeline -simd -compact -variables 4 -pipeline-latency 8 -n 6 -name n1fv_6 -include n1f.h */ /* * This function contains 18 FP additions, 8 FP multiplications, * (or, 12 additions, 2 multiplications, 6 fused multiply/add), * 23 stack variables, 2 constants, and 12 memory accesses */ #include "n1f.h" static void n1fv_6(const R *ri, const R *ii, R *ro, R *io, stride is, stride os, INT v, INT ivs, INT ovs) { DVK(KP500000000, +0.500000000000000000000000000000000000000000000); DVK(KP866025403, +0.866025403784438646763723170752936183471402627); { INT i; const R *xi; R *xo; xi = ri; xo = ro; for (i = v; i > 0; i = i - VL, xi = xi + (VL * ivs), xo = xo + (VL * ovs), MAKE_VOLATILE_STRIDE(12, is), MAKE_VOLATILE_STRIDE(12, os)) { V T1, T2, T4, T5, T7, T8; T1 = LD(&(xi[0]), ivs, &(xi[0])); T2 = LD(&(xi[WS(is, 3)]), ivs, &(xi[WS(is, 1)])); T4 = LD(&(xi[WS(is, 2)]), ivs, &(xi[0])); T5 = LD(&(xi[WS(is, 5)]), ivs, &(xi[WS(is, 1)])); T7 = LD(&(xi[WS(is, 4)]), ivs, &(xi[0])); T8 = LD(&(xi[WS(is, 1)]), ivs, &(xi[WS(is, 1)])); { V T3, Td, T6, Te, T9, Tf; T3 = VSUB(T1, T2); Td = VADD(T1, T2); T6 = VSUB(T4, T5); Te = VADD(T4, T5); T9 = VSUB(T7, T8); Tf = VADD(T7, T8); { V Tg, Ti, Ta, Tc, Th, Tb; Tg = VADD(Te, Tf); Ti = VMUL(LDK(KP866025403), VSUB(Tf, Te)); Ta = VADD(T6, T9); Tc = VMUL(LDK(KP866025403), VSUB(T9, T6)); Th = VFNMS(LDK(KP500000000), Tg, Td); ST(&(xo[0]), VADD(Td, Tg), ovs, &(xo[0])); Tb = VFNMS(LDK(KP500000000), Ta, T3); ST(&(xo[WS(os, 3)]), VADD(T3, Ta), ovs, &(xo[WS(os, 1)])); ST(&(xo[WS(os, 4)]), VFMAI(Ti, Th), ovs, &(xo[0])); ST(&(xo[WS(os, 2)]), VFNMSI(Ti, Th), ovs, &(xo[0])); ST(&(xo[WS(os, 1)]), VFMAI(Tc, Tb), ovs, &(xo[WS(os, 1)])); ST(&(xo[WS(os, 5)]), VFNMSI(Tc, Tb), ovs, &(xo[WS(os, 1)])); } } } } VLEAVE(); } static const kdft_desc desc = { 6, XSIMD_STRING("n1fv_6"), {12, 2, 6, 0}, &GENUS, 0, 0, 0, 0 }; void XSIMD(codelet_n1fv_6) (planner *p) { X(kdft_register) (p, n1fv_6, &desc); } #else /* HAVE_FMA */ /* Generated by: ../../../genfft/gen_notw_c.native -simd -compact -variables 4 -pipeline-latency 8 -n 6 -name n1fv_6 -include n1f.h */ /* * This function contains 18 FP additions, 4 FP multiplications, * (or, 16 additions, 2 multiplications, 2 fused multiply/add), * 19 stack variables, 2 constants, and 12 memory accesses */ #include "n1f.h" static void n1fv_6(const R *ri, const R *ii, R *ro, R *io, stride is, stride os, INT v, INT ivs, INT ovs) { DVK(KP866025403, +0.866025403784438646763723170752936183471402627); DVK(KP500000000, +0.500000000000000000000000000000000000000000000); { INT i; const R *xi; R *xo; xi = ri; xo = ro; for (i = v; i > 0; i = i - VL, xi = xi + (VL * ivs), xo = xo + (VL * ovs), MAKE_VOLATILE_STRIDE(12, is), MAKE_VOLATILE_STRIDE(12, os)) { V T3, Td, T6, Te, T9, Tf, Ta, Tg, T1, T2; T1 = LD(&(xi[0]), ivs, &(xi[0])); T2 = LD(&(xi[WS(is, 3)]), ivs, &(xi[WS(is, 1)])); T3 = VSUB(T1, T2); Td = VADD(T1, T2); { V T4, T5, T7, T8; T4 = LD(&(xi[WS(is, 2)]), ivs, &(xi[0])); T5 = LD(&(xi[WS(is, 5)]), ivs, &(xi[WS(is, 1)])); T6 = VSUB(T4, T5); Te = VADD(T4, T5); T7 = LD(&(xi[WS(is, 4)]), ivs, &(xi[0])); T8 = LD(&(xi[WS(is, 1)]), ivs, &(xi[WS(is, 1)])); T9 = VSUB(T7, T8); Tf = VADD(T7, T8); } Ta = VADD(T6, T9); Tg = VADD(Te, Tf); ST(&(xo[WS(os, 3)]), VADD(T3, Ta), ovs, &(xo[WS(os, 1)])); ST(&(xo[0]), VADD(Td, Tg), ovs, &(xo[0])); { V Tb, Tc, Th, Ti; Tb = VFNMS(LDK(KP500000000), Ta, T3); Tc = VBYI(VMUL(LDK(KP866025403), VSUB(T9, T6))); ST(&(xo[WS(os, 5)]), VSUB(Tb, Tc), ovs, &(xo[WS(os, 1)])); ST(&(xo[WS(os, 1)]), VADD(Tb, Tc), ovs, &(xo[WS(os, 1)])); Th = VFNMS(LDK(KP500000000), Tg, Td); Ti = VBYI(VMUL(LDK(KP866025403), VSUB(Tf, Te))); ST(&(xo[WS(os, 2)]), VSUB(Th, Ti), ovs, &(xo[0])); ST(&(xo[WS(os, 4)]), VADD(Th, Ti), ovs, &(xo[0])); } } } VLEAVE(); } static const kdft_desc desc = { 6, XSIMD_STRING("n1fv_6"), {16, 2, 2, 0}, &GENUS, 0, 0, 0, 0 }; void XSIMD(codelet_n1fv_6) (planner *p) { X(kdft_register) (p, n1fv_6, &desc); } #endif /* HAVE_FMA */ fftw-3.3.3/dft/simd/common/n1fv_7.c0000644000175000017500000001627612054410344013665 00000000000000/* * Copyright (c) 2003, 2007-11 Matteo Frigo * Copyright (c) 2003, 2007-11 Massachusetts Institute of Technology * * This program is free software; you can redistribute it and/or modify * it under the terms of the GNU General Public License as published by * the Free Software Foundation; either version 2 of the License, or * (at your option) any later version. * * This program is distributed in the hope that it will be useful, * but WITHOUT ANY WARRANTY; without even the implied warranty of * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the * GNU General Public License for more details. * * You should have received a copy of the GNU General Public License * along with this program; if not, write to the Free Software * Foundation, Inc., 51 Franklin Street, Fifth Floor, Boston, MA 02110-1301 USA * */ /* This file was automatically generated --- DO NOT EDIT */ /* Generated on Sun Nov 25 07:36:52 EST 2012 */ #include "codelet-dft.h" #ifdef HAVE_FMA /* Generated by: ../../../genfft/gen_notw_c.native -fma -reorder-insns -schedule-for-pipeline -simd -compact -variables 4 -pipeline-latency 8 -n 7 -name n1fv_7 -include n1f.h */ /* * This function contains 30 FP additions, 24 FP multiplications, * (or, 9 additions, 3 multiplications, 21 fused multiply/add), * 37 stack variables, 6 constants, and 14 memory accesses */ #include "n1f.h" static void n1fv_7(const R *ri, const R *ii, R *ro, R *io, stride is, stride os, INT v, INT ivs, INT ovs) { DVK(KP900968867, +0.900968867902419126236102319507445051165919162); DVK(KP692021471, +0.692021471630095869627814897002069140197260599); DVK(KP801937735, +0.801937735804838252472204639014890102331838324); DVK(KP974927912, +0.974927912181823607018131682993931217232785801); DVK(KP356895867, +0.356895867892209443894399510021300583399127187); DVK(KP554958132, +0.554958132087371191422194871006410481067288862); { INT i; const R *xi; R *xo; xi = ri; xo = ro; for (i = v; i > 0; i = i - VL, xi = xi + (VL * ivs), xo = xo + (VL * ovs), MAKE_VOLATILE_STRIDE(14, is), MAKE_VOLATILE_STRIDE(14, os)) { V T1, T2, T3, T8, T9, T5, T6; T1 = LD(&(xi[0]), ivs, &(xi[0])); T2 = LD(&(xi[WS(is, 1)]), ivs, &(xi[WS(is, 1)])); T3 = LD(&(xi[WS(is, 6)]), ivs, &(xi[0])); T8 = LD(&(xi[WS(is, 3)]), ivs, &(xi[WS(is, 1)])); T9 = LD(&(xi[WS(is, 4)]), ivs, &(xi[0])); T5 = LD(&(xi[WS(is, 2)]), ivs, &(xi[0])); T6 = LD(&(xi[WS(is, 5)]), ivs, &(xi[WS(is, 1)])); { V Te, T4, Tf, Ta, Tg, T7; Te = VSUB(T3, T2); T4 = VADD(T2, T3); Tf = VSUB(T9, T8); Ta = VADD(T8, T9); Tg = VSUB(T6, T5); T7 = VADD(T5, T6); { V Tm, Tb, Tr, Th, Tj, To; Tm = VFMA(LDK(KP554958132), Tf, Te); Tb = VFNMS(LDK(KP356895867), T4, Ta); Tr = VFNMS(LDK(KP554958132), Te, Tg); Th = VFMA(LDK(KP554958132), Tg, Tf); ST(&(xo[0]), VADD(T1, VADD(T4, VADD(T7, Ta))), ovs, &(xo[0])); Tj = VFNMS(LDK(KP356895867), T7, T4); To = VFNMS(LDK(KP356895867), Ta, T7); { V Tn, Tc, Ts, Ti; Tn = VMUL(LDK(KP974927912), VFMA(LDK(KP801937735), Tm, Tg)); Tc = VFNMS(LDK(KP692021471), Tb, T7); Ts = VMUL(LDK(KP974927912), VFNMS(LDK(KP801937735), Tr, Tf)); Ti = VMUL(LDK(KP974927912), VFNMS(LDK(KP801937735), Th, Te)); { V Tk, Tp, Td, Tl, Tq; Tk = VFNMS(LDK(KP692021471), Tj, Ta); Tp = VFNMS(LDK(KP692021471), To, T4); Td = VFNMS(LDK(KP900968867), Tc, T1); Tl = VFNMS(LDK(KP900968867), Tk, T1); Tq = VFNMS(LDK(KP900968867), Tp, T1); ST(&(xo[WS(os, 2)]), VFMAI(Ti, Td), ovs, &(xo[0])); ST(&(xo[WS(os, 5)]), VFNMSI(Ti, Td), ovs, &(xo[WS(os, 1)])); ST(&(xo[WS(os, 1)]), VFMAI(Tn, Tl), ovs, &(xo[WS(os, 1)])); ST(&(xo[WS(os, 6)]), VFNMSI(Tn, Tl), ovs, &(xo[0])); ST(&(xo[WS(os, 3)]), VFMAI(Ts, Tq), ovs, &(xo[WS(os, 1)])); ST(&(xo[WS(os, 4)]), VFNMSI(Ts, Tq), ovs, &(xo[0])); } } } } } } VLEAVE(); } static const kdft_desc desc = { 7, XSIMD_STRING("n1fv_7"), {9, 3, 21, 0}, &GENUS, 0, 0, 0, 0 }; void XSIMD(codelet_n1fv_7) (planner *p) { X(kdft_register) (p, n1fv_7, &desc); } #else /* HAVE_FMA */ /* Generated by: ../../../genfft/gen_notw_c.native -simd -compact -variables 4 -pipeline-latency 8 -n 7 -name n1fv_7 -include n1f.h */ /* * This function contains 30 FP additions, 18 FP multiplications, * (or, 18 additions, 6 multiplications, 12 fused multiply/add), * 24 stack variables, 6 constants, and 14 memory accesses */ #include "n1f.h" static void n1fv_7(const R *ri, const R *ii, R *ro, R *io, stride is, stride os, INT v, INT ivs, INT ovs) { DVK(KP900968867, +0.900968867902419126236102319507445051165919162); DVK(KP222520933, +0.222520933956314404288902564496794759466355569); DVK(KP623489801, +0.623489801858733530525004884004239810632274731); DVK(KP781831482, +0.781831482468029808708444526674057750232334519); DVK(KP974927912, +0.974927912181823607018131682993931217232785801); DVK(KP433883739, +0.433883739117558120475768332848358754609990728); { INT i; const R *xi; R *xo; xi = ri; xo = ro; for (i = v; i > 0; i = i - VL, xi = xi + (VL * ivs), xo = xo + (VL * ovs), MAKE_VOLATILE_STRIDE(14, is), MAKE_VOLATILE_STRIDE(14, os)) { V T1, Ta, Td, T4, Tc, T7, Te, T8, T9, Tj, Ti; T1 = LD(&(xi[0]), ivs, &(xi[0])); T8 = LD(&(xi[WS(is, 3)]), ivs, &(xi[WS(is, 1)])); T9 = LD(&(xi[WS(is, 4)]), ivs, &(xi[0])); Ta = VADD(T8, T9); Td = VSUB(T9, T8); { V T2, T3, T5, T6; T2 = LD(&(xi[WS(is, 1)]), ivs, &(xi[WS(is, 1)])); T3 = LD(&(xi[WS(is, 6)]), ivs, &(xi[0])); T4 = VADD(T2, T3); Tc = VSUB(T3, T2); T5 = LD(&(xi[WS(is, 2)]), ivs, &(xi[0])); T6 = LD(&(xi[WS(is, 5)]), ivs, &(xi[WS(is, 1)])); T7 = VADD(T5, T6); Te = VSUB(T6, T5); } ST(&(xo[0]), VADD(T1, VADD(T4, VADD(T7, Ta))), ovs, &(xo[0])); Tj = VBYI(VFMA(LDK(KP433883739), Tc, VFNMS(LDK(KP781831482), Te, VMUL(LDK(KP974927912), Td)))); Ti = VFMA(LDK(KP623489801), T7, VFNMS(LDK(KP222520933), Ta, VFNMS(LDK(KP900968867), T4, T1))); ST(&(xo[WS(os, 4)]), VSUB(Ti, Tj), ovs, &(xo[0])); ST(&(xo[WS(os, 3)]), VADD(Ti, Tj), ovs, &(xo[WS(os, 1)])); { V Tf, Tb, Th, Tg; Tf = VBYI(VFNMS(LDK(KP781831482), Td, VFNMS(LDK(KP433883739), Te, VMUL(LDK(KP974927912), Tc)))); Tb = VFMA(LDK(KP623489801), Ta, VFNMS(LDK(KP900968867), T7, VFNMS(LDK(KP222520933), T4, T1))); ST(&(xo[WS(os, 5)]), VSUB(Tb, Tf), ovs, &(xo[WS(os, 1)])); ST(&(xo[WS(os, 2)]), VADD(Tb, Tf), ovs, &(xo[0])); Th = VBYI(VFMA(LDK(KP781831482), Tc, VFMA(LDK(KP974927912), Te, VMUL(LDK(KP433883739), Td)))); Tg = VFMA(LDK(KP623489801), T4, VFNMS(LDK(KP900968867), Ta, VFNMS(LDK(KP222520933), T7, T1))); ST(&(xo[WS(os, 6)]), VSUB(Tg, Th), ovs, &(xo[0])); ST(&(xo[WS(os, 1)]), VADD(Tg, Th), ovs, &(xo[WS(os, 1)])); } } } VLEAVE(); } static const kdft_desc desc = { 7, XSIMD_STRING("n1fv_7"), {18, 6, 12, 0}, &GENUS, 0, 0, 0, 0 }; void XSIMD(codelet_n1fv_7) (planner *p) { X(kdft_register) (p, n1fv_7, &desc); } #endif /* HAVE_FMA */ fftw-3.3.3/dft/simd/common/n1fv_8.c0000644000175000017500000001415312054410344013656 00000000000000/* * Copyright (c) 2003, 2007-11 Matteo Frigo * Copyright (c) 2003, 2007-11 Massachusetts Institute of Technology * * This program is free software; you can redistribute it and/or modify * it under the terms of the GNU General Public License as published by * the Free Software Foundation; either version 2 of the License, or * (at your option) any later version. * * This program is distributed in the hope that it will be useful, * but WITHOUT ANY WARRANTY; without even the implied warranty of * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the * GNU General Public License for more details. * * You should have received a copy of the GNU General Public License * along with this program; if not, write to the Free Software * Foundation, Inc., 51 Franklin Street, Fifth Floor, Boston, MA 02110-1301 USA * */ /* This file was automatically generated --- DO NOT EDIT */ /* Generated on Sun Nov 25 07:36:52 EST 2012 */ #include "codelet-dft.h" #ifdef HAVE_FMA /* Generated by: ../../../genfft/gen_notw_c.native -fma -reorder-insns -schedule-for-pipeline -simd -compact -variables 4 -pipeline-latency 8 -n 8 -name n1fv_8 -include n1f.h */ /* * This function contains 26 FP additions, 10 FP multiplications, * (or, 16 additions, 0 multiplications, 10 fused multiply/add), * 30 stack variables, 1 constants, and 16 memory accesses */ #include "n1f.h" static void n1fv_8(const R *ri, const R *ii, R *ro, R *io, stride is, stride os, INT v, INT ivs, INT ovs) { DVK(KP707106781, +0.707106781186547524400844362104849039284835938); { INT i; const R *xi; R *xo; xi = ri; xo = ro; for (i = v; i > 0; i = i - VL, xi = xi + (VL * ivs), xo = xo + (VL * ovs), MAKE_VOLATILE_STRIDE(16, is), MAKE_VOLATILE_STRIDE(16, os)) { V T1, T2, Tc, Td, T4, T5, T7, T8; T1 = LD(&(xi[0]), ivs, &(xi[0])); T2 = LD(&(xi[WS(is, 4)]), ivs, &(xi[0])); Tc = LD(&(xi[WS(is, 2)]), ivs, &(xi[0])); Td = LD(&(xi[WS(is, 6)]), ivs, &(xi[0])); T4 = LD(&(xi[WS(is, 1)]), ivs, &(xi[WS(is, 1)])); T5 = LD(&(xi[WS(is, 5)]), ivs, &(xi[WS(is, 1)])); T7 = LD(&(xi[WS(is, 7)]), ivs, &(xi[WS(is, 1)])); T8 = LD(&(xi[WS(is, 3)]), ivs, &(xi[WS(is, 1)])); { V T3, Tj, Te, Tk, T6, Tm, T9, Tn, Tp, Tl; T3 = VSUB(T1, T2); Tj = VADD(T1, T2); Te = VSUB(Tc, Td); Tk = VADD(Tc, Td); T6 = VSUB(T4, T5); Tm = VADD(T4, T5); T9 = VSUB(T7, T8); Tn = VADD(T7, T8); Tp = VSUB(Tj, Tk); Tl = VADD(Tj, Tk); { V Tq, To, Ta, Tf; Tq = VSUB(Tn, Tm); To = VADD(Tm, Tn); Ta = VADD(T6, T9); Tf = VSUB(T9, T6); { V Tg, Ti, Tb, Th; ST(&(xo[0]), VADD(Tl, To), ovs, &(xo[0])); ST(&(xo[WS(os, 4)]), VSUB(Tl, To), ovs, &(xo[0])); ST(&(xo[WS(os, 2)]), VFMAI(Tq, Tp), ovs, &(xo[0])); ST(&(xo[WS(os, 6)]), VFNMSI(Tq, Tp), ovs, &(xo[0])); Tg = VFNMS(LDK(KP707106781), Tf, Te); Ti = VFMA(LDK(KP707106781), Tf, Te); Tb = VFMA(LDK(KP707106781), Ta, T3); Th = VFNMS(LDK(KP707106781), Ta, T3); ST(&(xo[WS(os, 3)]), VFMAI(Ti, Th), ovs, &(xo[WS(os, 1)])); ST(&(xo[WS(os, 5)]), VFNMSI(Ti, Th), ovs, &(xo[WS(os, 1)])); ST(&(xo[WS(os, 7)]), VFMAI(Tg, Tb), ovs, &(xo[WS(os, 1)])); ST(&(xo[WS(os, 1)]), VFNMSI(Tg, Tb), ovs, &(xo[WS(os, 1)])); } } } } } VLEAVE(); } static const kdft_desc desc = { 8, XSIMD_STRING("n1fv_8"), {16, 0, 10, 0}, &GENUS, 0, 0, 0, 0 }; void XSIMD(codelet_n1fv_8) (planner *p) { X(kdft_register) (p, n1fv_8, &desc); } #else /* HAVE_FMA */ /* Generated by: ../../../genfft/gen_notw_c.native -simd -compact -variables 4 -pipeline-latency 8 -n 8 -name n1fv_8 -include n1f.h */ /* * This function contains 26 FP additions, 2 FP multiplications, * (or, 26 additions, 2 multiplications, 0 fused multiply/add), * 22 stack variables, 1 constants, and 16 memory accesses */ #include "n1f.h" static void n1fv_8(const R *ri, const R *ii, R *ro, R *io, stride is, stride os, INT v, INT ivs, INT ovs) { DVK(KP707106781, +0.707106781186547524400844362104849039284835938); { INT i; const R *xi; R *xo; xi = ri; xo = ro; for (i = v; i > 0; i = i - VL, xi = xi + (VL * ivs), xo = xo + (VL * ovs), MAKE_VOLATILE_STRIDE(16, is), MAKE_VOLATILE_STRIDE(16, os)) { V T3, Tj, Tf, Tk, Ta, Tn, Tc, Tm; { V T1, T2, Td, Te; T1 = LD(&(xi[0]), ivs, &(xi[0])); T2 = LD(&(xi[WS(is, 4)]), ivs, &(xi[0])); T3 = VSUB(T1, T2); Tj = VADD(T1, T2); Td = LD(&(xi[WS(is, 2)]), ivs, &(xi[0])); Te = LD(&(xi[WS(is, 6)]), ivs, &(xi[0])); Tf = VSUB(Td, Te); Tk = VADD(Td, Te); { V T4, T5, T6, T7, T8, T9; T4 = LD(&(xi[WS(is, 1)]), ivs, &(xi[WS(is, 1)])); T5 = LD(&(xi[WS(is, 5)]), ivs, &(xi[WS(is, 1)])); T6 = VSUB(T4, T5); T7 = LD(&(xi[WS(is, 7)]), ivs, &(xi[WS(is, 1)])); T8 = LD(&(xi[WS(is, 3)]), ivs, &(xi[WS(is, 1)])); T9 = VSUB(T7, T8); Ta = VMUL(LDK(KP707106781), VADD(T6, T9)); Tn = VADD(T7, T8); Tc = VMUL(LDK(KP707106781), VSUB(T9, T6)); Tm = VADD(T4, T5); } } { V Tb, Tg, Tp, Tq; Tb = VADD(T3, Ta); Tg = VBYI(VSUB(Tc, Tf)); ST(&(xo[WS(os, 7)]), VSUB(Tb, Tg), ovs, &(xo[WS(os, 1)])); ST(&(xo[WS(os, 1)]), VADD(Tb, Tg), ovs, &(xo[WS(os, 1)])); Tp = VSUB(Tj, Tk); Tq = VBYI(VSUB(Tn, Tm)); ST(&(xo[WS(os, 6)]), VSUB(Tp, Tq), ovs, &(xo[0])); ST(&(xo[WS(os, 2)]), VADD(Tp, Tq), ovs, &(xo[0])); } { V Th, Ti, Tl, To; Th = VSUB(T3, Ta); Ti = VBYI(VADD(Tf, Tc)); ST(&(xo[WS(os, 5)]), VSUB(Th, Ti), ovs, &(xo[WS(os, 1)])); ST(&(xo[WS(os, 3)]), VADD(Th, Ti), ovs, &(xo[WS(os, 1)])); Tl = VADD(Tj, Tk); To = VADD(Tm, Tn); ST(&(xo[WS(os, 4)]), VSUB(Tl, To), ovs, &(xo[0])); ST(&(xo[0]), VADD(Tl, To), ovs, &(xo[0])); } } } VLEAVE(); } static const kdft_desc desc = { 8, XSIMD_STRING("n1fv_8"), {26, 2, 0, 0}, &GENUS, 0, 0, 0, 0 }; void XSIMD(codelet_n1fv_8) (planner *p) { X(kdft_register) (p, n1fv_8, &desc); } #endif /* HAVE_FMA */ fftw-3.3.3/dft/simd/common/n1fv_9.c0000644000175000017500000002473712054410344013670 00000000000000/* * Copyright (c) 2003, 2007-11 Matteo Frigo * Copyright (c) 2003, 2007-11 Massachusetts Institute of Technology * * This program is free software; you can redistribute it and/or modify * it under the terms of the GNU General Public License as published by * the Free Software Foundation; either version 2 of the License, or * (at your option) any later version. * * This program is distributed in the hope that it will be useful, * but WITHOUT ANY WARRANTY; without even the implied warranty of * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the * GNU General Public License for more details. * * You should have received a copy of the GNU General Public License * along with this program; if not, write to the Free Software * Foundation, Inc., 51 Franklin Street, Fifth Floor, Boston, MA 02110-1301 USA * */ /* This file was automatically generated --- DO NOT EDIT */ /* Generated on Sun Nov 25 07:36:52 EST 2012 */ #include "codelet-dft.h" #ifdef HAVE_FMA /* Generated by: ../../../genfft/gen_notw_c.native -fma -reorder-insns -schedule-for-pipeline -simd -compact -variables 4 -pipeline-latency 8 -n 9 -name n1fv_9 -include n1f.h */ /* * This function contains 46 FP additions, 38 FP multiplications, * (or, 12 additions, 4 multiplications, 34 fused multiply/add), * 68 stack variables, 19 constants, and 18 memory accesses */ #include "n1f.h" static void n1fv_9(const R *ri, const R *ii, R *ro, R *io, stride is, stride os, INT v, INT ivs, INT ovs) { DVK(KP939692620, +0.939692620785908384054109277324731469936208134); DVK(KP826351822, +0.826351822333069651148283373230685203999624323); DVK(KP879385241, +0.879385241571816768108218554649462939872416269); DVK(KP984807753, +0.984807753012208059366743024589523013670643252); DVK(KP666666666, +0.666666666666666666666666666666666666666666667); DVK(KP852868531, +0.852868531952443209628250963940074071936020296); DVK(KP907603734, +0.907603734547952313649323976213898122064543220); DVK(KP420276625, +0.420276625461206169731530603237061658838781920); DVK(KP673648177, +0.673648177666930348851716626769314796000375677); DVK(KP898197570, +0.898197570222573798468955502359086394667167570); DVK(KP347296355, +0.347296355333860697703433253538629592000751354); DVK(KP866025403, +0.866025403784438646763723170752936183471402627); DVK(KP439692620, +0.439692620785908384054109277324731469936208134); DVK(KP203604859, +0.203604859554852403062088995281827210665664861); DVK(KP152703644, +0.152703644666139302296566746461370407999248646); DVK(KP586256827, +0.586256827714544512072145703099641959914944179); DVK(KP968908795, +0.968908795874236621082202410917456709164223497); DVK(KP726681596, +0.726681596905677465811651808188092531873167623); DVK(KP500000000, +0.500000000000000000000000000000000000000000000); { INT i; const R *xi; R *xo; xi = ri; xo = ro; for (i = v; i > 0; i = i - VL, xi = xi + (VL * ivs), xo = xo + (VL * ovs), MAKE_VOLATILE_STRIDE(18, is), MAKE_VOLATILE_STRIDE(18, os)) { V T1, T2, T3, T6, Tb, T7, T8, Tc, Td, Tv, T4; T1 = LD(&(xi[0]), ivs, &(xi[0])); T2 = LD(&(xi[WS(is, 3)]), ivs, &(xi[WS(is, 1)])); T3 = LD(&(xi[WS(is, 6)]), ivs, &(xi[0])); T6 = LD(&(xi[WS(is, 1)]), ivs, &(xi[WS(is, 1)])); Tb = LD(&(xi[WS(is, 2)]), ivs, &(xi[0])); T7 = LD(&(xi[WS(is, 4)]), ivs, &(xi[0])); T8 = LD(&(xi[WS(is, 7)]), ivs, &(xi[WS(is, 1)])); Tc = LD(&(xi[WS(is, 5)]), ivs, &(xi[WS(is, 1)])); Td = LD(&(xi[WS(is, 8)]), ivs, &(xi[0])); Tv = VSUB(T3, T2); T4 = VADD(T2, T3); { V Tl, T9, Tm, Te, Tj, T5; Tl = VSUB(T7, T8); T9 = VADD(T7, T8); Tm = VSUB(Td, Tc); Te = VADD(Tc, Td); Tj = VFNMS(LDK(KP500000000), T4, T1); T5 = VADD(T1, T4); { V Tn, Ta, Tk, Tf; Tn = VFNMS(LDK(KP500000000), T9, T6); Ta = VADD(T6, T9); Tk = VFNMS(LDK(KP500000000), Te, Tb); Tf = VADD(Tb, Te); { V Ty, TC, To, TB, Tx, Ts, Tg, Ti; Ty = VFNMS(LDK(KP726681596), Tl, Tn); TC = VFMA(LDK(KP968908795), Tn, Tl); To = VFNMS(LDK(KP586256827), Tn, Tm); TB = VFNMS(LDK(KP152703644), Tm, Tk); Tx = VFMA(LDK(KP203604859), Tk, Tm); Ts = VFNMS(LDK(KP439692620), Tl, Tk); Tg = VADD(Ta, Tf); Ti = VMUL(LDK(KP866025403), VSUB(Tf, Ta)); { V Tz, TI, TF, TD, Tt, Th, Tq, Tp; Tp = VFNMS(LDK(KP347296355), To, Tl); Tz = VFMA(LDK(KP898197570), Ty, Tx); TI = VFNMS(LDK(KP898197570), Ty, Tx); TF = VFNMS(LDK(KP673648177), TC, TB); TD = VFMA(LDK(KP673648177), TC, TB); Tt = VFNMS(LDK(KP420276625), Ts, Tm); ST(&(xo[0]), VADD(T5, Tg), ovs, &(xo[0])); Th = VFNMS(LDK(KP500000000), Tg, T5); Tq = VFNMS(LDK(KP907603734), Tp, Tk); { V TA, TJ, TE, TG, Tu, Tr, TK, TH, Tw; TA = VFMA(LDK(KP852868531), Tz, Tj); TJ = VFMA(LDK(KP666666666), TD, TI); TE = VMUL(LDK(KP984807753), VFNMS(LDK(KP879385241), Tv, TD)); TG = VFNMS(LDK(KP500000000), Tz, TF); Tu = VFNMS(LDK(KP826351822), Tt, Tn); ST(&(xo[WS(os, 6)]), VFNMSI(Ti, Th), ovs, &(xo[0])); ST(&(xo[WS(os, 3)]), VFMAI(Ti, Th), ovs, &(xo[WS(os, 1)])); Tr = VFNMS(LDK(KP939692620), Tq, Tj); TK = VMUL(LDK(KP866025403), VFMA(LDK(KP852868531), TJ, Tv)); ST(&(xo[WS(os, 8)]), VFMAI(TE, TA), ovs, &(xo[0])); ST(&(xo[WS(os, 1)]), VFNMSI(TE, TA), ovs, &(xo[WS(os, 1)])); TH = VFMA(LDK(KP852868531), TG, Tj); Tw = VMUL(LDK(KP984807753), VFMA(LDK(KP879385241), Tv, Tu)); ST(&(xo[WS(os, 4)]), VFMAI(TK, TH), ovs, &(xo[0])); ST(&(xo[WS(os, 5)]), VFNMSI(TK, TH), ovs, &(xo[WS(os, 1)])); ST(&(xo[WS(os, 7)]), VFMAI(Tw, Tr), ovs, &(xo[WS(os, 1)])); ST(&(xo[WS(os, 2)]), VFNMSI(Tw, Tr), ovs, &(xo[0])); } } } } } } } VLEAVE(); } static const kdft_desc desc = { 9, XSIMD_STRING("n1fv_9"), {12, 4, 34, 0}, &GENUS, 0, 0, 0, 0 }; void XSIMD(codelet_n1fv_9) (planner *p) { X(kdft_register) (p, n1fv_9, &desc); } #else /* HAVE_FMA */ /* Generated by: ../../../genfft/gen_notw_c.native -simd -compact -variables 4 -pipeline-latency 8 -n 9 -name n1fv_9 -include n1f.h */ /* * This function contains 46 FP additions, 26 FP multiplications, * (or, 30 additions, 10 multiplications, 16 fused multiply/add), * 41 stack variables, 14 constants, and 18 memory accesses */ #include "n1f.h" static void n1fv_9(const R *ri, const R *ii, R *ro, R *io, stride is, stride os, INT v, INT ivs, INT ovs) { DVK(KP342020143, +0.342020143325668733044099614682259580763083368); DVK(KP813797681, +0.813797681349373692844693217248393223289101568); DVK(KP939692620, +0.939692620785908384054109277324731469936208134); DVK(KP296198132, +0.296198132726023843175338011893050938967728390); DVK(KP642787609, +0.642787609686539326322643409907263432907559884); DVK(KP663413948, +0.663413948168938396205421319635891297216863310); DVK(KP556670399, +0.556670399226419366452912952047023132968291906); DVK(KP766044443, +0.766044443118978035202392650555416673935832457); DVK(KP984807753, +0.984807753012208059366743024589523013670643252); DVK(KP150383733, +0.150383733180435296639271897612501926072238258); DVK(KP852868531, +0.852868531952443209628250963940074071936020296); DVK(KP173648177, +0.173648177666930348851716626769314796000375677); DVK(KP500000000, +0.500000000000000000000000000000000000000000000); DVK(KP866025403, +0.866025403784438646763723170752936183471402627); { INT i; const R *xi; R *xo; xi = ri; xo = ro; for (i = v; i > 0; i = i - VL, xi = xi + (VL * ivs), xo = xo + (VL * ovs), MAKE_VOLATILE_STRIDE(18, is), MAKE_VOLATILE_STRIDE(18, os)) { V T5, Ts, Tj, To, Tf, Tn, Tp, Tu, Tl, Ta, Tk, Tm, Tt; { V T1, T2, T3, T4; T1 = LD(&(xi[0]), ivs, &(xi[0])); T2 = LD(&(xi[WS(is, 3)]), ivs, &(xi[WS(is, 1)])); T3 = LD(&(xi[WS(is, 6)]), ivs, &(xi[0])); T4 = VADD(T2, T3); T5 = VADD(T1, T4); Ts = VMUL(LDK(KP866025403), VSUB(T3, T2)); Tj = VFNMS(LDK(KP500000000), T4, T1); } { V Tb, Te, Tc, Td; Tb = LD(&(xi[WS(is, 2)]), ivs, &(xi[0])); Tc = LD(&(xi[WS(is, 5)]), ivs, &(xi[WS(is, 1)])); Td = LD(&(xi[WS(is, 8)]), ivs, &(xi[0])); Te = VADD(Tc, Td); To = VSUB(Td, Tc); Tf = VADD(Tb, Te); Tn = VFNMS(LDK(KP500000000), Te, Tb); Tp = VFMA(LDK(KP173648177), Tn, VMUL(LDK(KP852868531), To)); Tu = VFNMS(LDK(KP984807753), Tn, VMUL(LDK(KP150383733), To)); } { V T6, T9, T7, T8; T6 = LD(&(xi[WS(is, 1)]), ivs, &(xi[WS(is, 1)])); T7 = LD(&(xi[WS(is, 4)]), ivs, &(xi[0])); T8 = LD(&(xi[WS(is, 7)]), ivs, &(xi[WS(is, 1)])); T9 = VADD(T7, T8); Tl = VSUB(T8, T7); Ta = VADD(T6, T9); Tk = VFNMS(LDK(KP500000000), T9, T6); Tm = VFMA(LDK(KP766044443), Tk, VMUL(LDK(KP556670399), Tl)); Tt = VFNMS(LDK(KP642787609), Tk, VMUL(LDK(KP663413948), Tl)); } { V Ti, Tg, Th, Tz, TA; Ti = VBYI(VMUL(LDK(KP866025403), VSUB(Tf, Ta))); Tg = VADD(Ta, Tf); Th = VFNMS(LDK(KP500000000), Tg, T5); ST(&(xo[0]), VADD(T5, Tg), ovs, &(xo[0])); ST(&(xo[WS(os, 3)]), VADD(Th, Ti), ovs, &(xo[WS(os, 1)])); ST(&(xo[WS(os, 6)]), VSUB(Th, Ti), ovs, &(xo[0])); Tz = VFMA(LDK(KP173648177), Tk, VFNMS(LDK(KP296198132), To, VFNMS(LDK(KP939692620), Tn, VFNMS(LDK(KP852868531), Tl, Tj)))); TA = VBYI(VSUB(VFNMS(LDK(KP342020143), Tn, VFNMS(LDK(KP150383733), Tl, VFNMS(LDK(KP984807753), Tk, VMUL(LDK(KP813797681), To)))), Ts)); ST(&(xo[WS(os, 7)]), VSUB(Tz, TA), ovs, &(xo[WS(os, 1)])); ST(&(xo[WS(os, 2)]), VADD(Tz, TA), ovs, &(xo[0])); { V Tr, Tx, Tw, Ty, Tq, Tv; Tq = VADD(Tm, Tp); Tr = VADD(Tj, Tq); Tx = VFMA(LDK(KP866025403), VSUB(Tt, Tu), VFNMS(LDK(KP500000000), Tq, Tj)); Tv = VADD(Tt, Tu); Tw = VBYI(VADD(Ts, Tv)); Ty = VBYI(VADD(Ts, VFNMS(LDK(KP500000000), Tv, VMUL(LDK(KP866025403), VSUB(Tp, Tm))))); ST(&(xo[WS(os, 8)]), VSUB(Tr, Tw), ovs, &(xo[0])); ST(&(xo[WS(os, 4)]), VADD(Tx, Ty), ovs, &(xo[0])); ST(&(xo[WS(os, 1)]), VADD(Tw, Tr), ovs, &(xo[WS(os, 1)])); ST(&(xo[WS(os, 5)]), VSUB(Tx, Ty), ovs, &(xo[WS(os, 1)])); } } } } VLEAVE(); } static const kdft_desc desc = { 9, XSIMD_STRING("n1fv_9"), {30, 10, 16, 0}, &GENUS, 0, 0, 0, 0 }; void XSIMD(codelet_n1fv_9) (planner *p) { X(kdft_register) (p, n1fv_9, &desc); } #endif /* HAVE_FMA */ fftw-3.3.3/dft/simd/common/n1fv_10.c0000644000175000017500000002045212054410344013726 00000000000000/* * Copyright (c) 2003, 2007-11 Matteo Frigo * Copyright (c) 2003, 2007-11 Massachusetts Institute of Technology * * This program is free software; you can redistribute it and/or modify * it under the terms of the GNU General Public License as published by * the Free Software Foundation; either version 2 of the License, or * (at your option) any later version. * * This program is distributed in the hope that it will be useful, * but WITHOUT ANY WARRANTY; without even the implied warranty of * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the * GNU General Public License for more details. * * You should have received a copy of the GNU General Public License * along with this program; if not, write to the Free Software * Foundation, Inc., 51 Franklin Street, Fifth Floor, Boston, MA 02110-1301 USA * */ /* This file was automatically generated --- DO NOT EDIT */ /* Generated on Sun Nov 25 07:36:52 EST 2012 */ #include "codelet-dft.h" #ifdef HAVE_FMA /* Generated by: ../../../genfft/gen_notw_c.native -fma -reorder-insns -schedule-for-pipeline -simd -compact -variables 4 -pipeline-latency 8 -n 10 -name n1fv_10 -include n1f.h */ /* * This function contains 42 FP additions, 22 FP multiplications, * (or, 24 additions, 4 multiplications, 18 fused multiply/add), * 43 stack variables, 4 constants, and 20 memory accesses */ #include "n1f.h" static void n1fv_10(const R *ri, const R *ii, R *ro, R *io, stride is, stride os, INT v, INT ivs, INT ovs) { DVK(KP559016994, +0.559016994374947424102293417182819058860154590); DVK(KP250000000, +0.250000000000000000000000000000000000000000000); DVK(KP618033988, +0.618033988749894848204586834365638117720309180); DVK(KP951056516, +0.951056516295153572116439333379382143405698634); { INT i; const R *xi; R *xo; xi = ri; xo = ro; for (i = v; i > 0; i = i - VL, xi = xi + (VL * ivs), xo = xo + (VL * ovs), MAKE_VOLATILE_STRIDE(20, is), MAKE_VOLATILE_STRIDE(20, os)) { V Tb, Tr, T3, Ts, T6, Tw, Tg, Tt, T9, Tc, T1, T2; T1 = LD(&(xi[0]), ivs, &(xi[0])); T2 = LD(&(xi[WS(is, 5)]), ivs, &(xi[WS(is, 1)])); { V T4, T5, Te, Tf, T7, T8; T4 = LD(&(xi[WS(is, 2)]), ivs, &(xi[0])); T5 = LD(&(xi[WS(is, 7)]), ivs, &(xi[WS(is, 1)])); Te = LD(&(xi[WS(is, 6)]), ivs, &(xi[0])); Tf = LD(&(xi[WS(is, 1)]), ivs, &(xi[WS(is, 1)])); T7 = LD(&(xi[WS(is, 8)]), ivs, &(xi[0])); T8 = LD(&(xi[WS(is, 3)]), ivs, &(xi[WS(is, 1)])); Tb = LD(&(xi[WS(is, 4)]), ivs, &(xi[0])); Tr = VADD(T1, T2); T3 = VSUB(T1, T2); Ts = VADD(T4, T5); T6 = VSUB(T4, T5); Tw = VADD(Te, Tf); Tg = VSUB(Te, Tf); Tt = VADD(T7, T8); T9 = VSUB(T7, T8); Tc = LD(&(xi[WS(is, 9)]), ivs, &(xi[WS(is, 1)])); } { V TD, Tu, Tm, Ta, Td, Tv; TD = VSUB(Ts, Tt); Tu = VADD(Ts, Tt); Tm = VSUB(T6, T9); Ta = VADD(T6, T9); Td = VSUB(Tb, Tc); Tv = VADD(Tb, Tc); { V TC, Tx, Tn, Th; TC = VSUB(Tv, Tw); Tx = VADD(Tv, Tw); Tn = VSUB(Td, Tg); Th = VADD(Td, Tg); { V Ty, TA, TE, TG, Ti, Tk, To, Tq, Tz, Tj; Ty = VADD(Tu, Tx); TA = VSUB(Tu, Tx); TE = VMUL(LDK(KP951056516), VFNMS(LDK(KP618033988), TD, TC)); TG = VMUL(LDK(KP951056516), VFMA(LDK(KP618033988), TC, TD)); Ti = VADD(Ta, Th); Tk = VSUB(Ta, Th); To = VMUL(LDK(KP951056516), VFMA(LDK(KP618033988), Tn, Tm)); Tq = VMUL(LDK(KP951056516), VFNMS(LDK(KP618033988), Tm, Tn)); Tz = VFNMS(LDK(KP250000000), Ty, Tr); ST(&(xo[0]), VADD(Tr, Ty), ovs, &(xo[0])); Tj = VFNMS(LDK(KP250000000), Ti, T3); ST(&(xo[WS(os, 5)]), VADD(T3, Ti), ovs, &(xo[WS(os, 1)])); { V TB, TF, Tl, Tp; TB = VFNMS(LDK(KP559016994), TA, Tz); TF = VFMA(LDK(KP559016994), TA, Tz); Tl = VFMA(LDK(KP559016994), Tk, Tj); Tp = VFNMS(LDK(KP559016994), Tk, Tj); ST(&(xo[WS(os, 4)]), VFMAI(TG, TF), ovs, &(xo[0])); ST(&(xo[WS(os, 6)]), VFNMSI(TG, TF), ovs, &(xo[0])); ST(&(xo[WS(os, 8)]), VFNMSI(TE, TB), ovs, &(xo[0])); ST(&(xo[WS(os, 2)]), VFMAI(TE, TB), ovs, &(xo[0])); ST(&(xo[WS(os, 3)]), VFNMSI(Tq, Tp), ovs, &(xo[WS(os, 1)])); ST(&(xo[WS(os, 7)]), VFMAI(Tq, Tp), ovs, &(xo[WS(os, 1)])); ST(&(xo[WS(os, 9)]), VFMAI(To, Tl), ovs, &(xo[WS(os, 1)])); ST(&(xo[WS(os, 1)]), VFNMSI(To, Tl), ovs, &(xo[WS(os, 1)])); } } } } } } VLEAVE(); } static const kdft_desc desc = { 10, XSIMD_STRING("n1fv_10"), {24, 4, 18, 0}, &GENUS, 0, 0, 0, 0 }; void XSIMD(codelet_n1fv_10) (planner *p) { X(kdft_register) (p, n1fv_10, &desc); } #else /* HAVE_FMA */ /* Generated by: ../../../genfft/gen_notw_c.native -simd -compact -variables 4 -pipeline-latency 8 -n 10 -name n1fv_10 -include n1f.h */ /* * This function contains 42 FP additions, 12 FP multiplications, * (or, 36 additions, 6 multiplications, 6 fused multiply/add), * 33 stack variables, 4 constants, and 20 memory accesses */ #include "n1f.h" static void n1fv_10(const R *ri, const R *ii, R *ro, R *io, stride is, stride os, INT v, INT ivs, INT ovs) { DVK(KP250000000, +0.250000000000000000000000000000000000000000000); DVK(KP559016994, +0.559016994374947424102293417182819058860154590); DVK(KP587785252, +0.587785252292473129168705954639072768597652438); DVK(KP951056516, +0.951056516295153572116439333379382143405698634); { INT i; const R *xi; R *xo; xi = ri; xo = ro; for (i = v; i > 0; i = i - VL, xi = xi + (VL * ivs), xo = xo + (VL * ovs), MAKE_VOLATILE_STRIDE(20, is), MAKE_VOLATILE_STRIDE(20, os)) { V Ti, Ty, Tm, Tn, Tw, Tt, Tz, TA, TB, T7, Te, Tj, Tg, Th; Tg = LD(&(xi[0]), ivs, &(xi[0])); Th = LD(&(xi[WS(is, 5)]), ivs, &(xi[WS(is, 1)])); Ti = VSUB(Tg, Th); Ty = VADD(Tg, Th); { V T3, Tu, Td, Ts, T6, Tv, Ta, Tr; { V T1, T2, Tb, Tc; T1 = LD(&(xi[WS(is, 2)]), ivs, &(xi[0])); T2 = LD(&(xi[WS(is, 7)]), ivs, &(xi[WS(is, 1)])); T3 = VSUB(T1, T2); Tu = VADD(T1, T2); Tb = LD(&(xi[WS(is, 6)]), ivs, &(xi[0])); Tc = LD(&(xi[WS(is, 1)]), ivs, &(xi[WS(is, 1)])); Td = VSUB(Tb, Tc); Ts = VADD(Tb, Tc); } { V T4, T5, T8, T9; T4 = LD(&(xi[WS(is, 8)]), ivs, &(xi[0])); T5 = LD(&(xi[WS(is, 3)]), ivs, &(xi[WS(is, 1)])); T6 = VSUB(T4, T5); Tv = VADD(T4, T5); T8 = LD(&(xi[WS(is, 4)]), ivs, &(xi[0])); T9 = LD(&(xi[WS(is, 9)]), ivs, &(xi[WS(is, 1)])); Ta = VSUB(T8, T9); Tr = VADD(T8, T9); } Tm = VSUB(T3, T6); Tn = VSUB(Ta, Td); Tw = VSUB(Tu, Tv); Tt = VSUB(Tr, Ts); Tz = VADD(Tu, Tv); TA = VADD(Tr, Ts); TB = VADD(Tz, TA); T7 = VADD(T3, T6); Te = VADD(Ta, Td); Tj = VADD(T7, Te); } ST(&(xo[WS(os, 5)]), VADD(Ti, Tj), ovs, &(xo[WS(os, 1)])); ST(&(xo[0]), VADD(Ty, TB), ovs, &(xo[0])); { V To, Tq, Tl, Tp, Tf, Tk; To = VBYI(VFMA(LDK(KP951056516), Tm, VMUL(LDK(KP587785252), Tn))); Tq = VBYI(VFNMS(LDK(KP587785252), Tm, VMUL(LDK(KP951056516), Tn))); Tf = VMUL(LDK(KP559016994), VSUB(T7, Te)); Tk = VFNMS(LDK(KP250000000), Tj, Ti); Tl = VADD(Tf, Tk); Tp = VSUB(Tk, Tf); ST(&(xo[WS(os, 1)]), VSUB(Tl, To), ovs, &(xo[WS(os, 1)])); ST(&(xo[WS(os, 7)]), VADD(Tq, Tp), ovs, &(xo[WS(os, 1)])); ST(&(xo[WS(os, 9)]), VADD(To, Tl), ovs, &(xo[WS(os, 1)])); ST(&(xo[WS(os, 3)]), VSUB(Tp, Tq), ovs, &(xo[WS(os, 1)])); } { V Tx, TF, TE, TG, TC, TD; Tx = VBYI(VFNMS(LDK(KP587785252), Tw, VMUL(LDK(KP951056516), Tt))); TF = VBYI(VFMA(LDK(KP951056516), Tw, VMUL(LDK(KP587785252), Tt))); TC = VFNMS(LDK(KP250000000), TB, Ty); TD = VMUL(LDK(KP559016994), VSUB(Tz, TA)); TE = VSUB(TC, TD); TG = VADD(TD, TC); ST(&(xo[WS(os, 2)]), VADD(Tx, TE), ovs, &(xo[0])); ST(&(xo[WS(os, 6)]), VSUB(TG, TF), ovs, &(xo[0])); ST(&(xo[WS(os, 8)]), VSUB(TE, Tx), ovs, &(xo[0])); ST(&(xo[WS(os, 4)]), VADD(TF, TG), ovs, &(xo[0])); } } } VLEAVE(); } static const kdft_desc desc = { 10, XSIMD_STRING("n1fv_10"), {36, 6, 6, 0}, &GENUS, 0, 0, 0, 0 }; void XSIMD(codelet_n1fv_10) (planner *p) { X(kdft_register) (p, n1fv_10, &desc); } #endif /* HAVE_FMA */ fftw-3.3.3/dft/simd/common/n1fv_11.c0000644000175000017500000002731412054410346013735 00000000000000/* * Copyright (c) 2003, 2007-11 Matteo Frigo * Copyright (c) 2003, 2007-11 Massachusetts Institute of Technology * * This program is free software; you can redistribute it and/or modify * it under the terms of the GNU General Public License as published by * the Free Software Foundation; either version 2 of the License, or * (at your option) any later version. * * This program is distributed in the hope that it will be useful, * but WITHOUT ANY WARRANTY; without even the implied warranty of * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the * GNU General Public License for more details. * * You should have received a copy of the GNU General Public License * along with this program; if not, write to the Free Software * Foundation, Inc., 51 Franklin Street, Fifth Floor, Boston, MA 02110-1301 USA * */ /* This file was automatically generated --- DO NOT EDIT */ /* Generated on Sun Nov 25 07:36:52 EST 2012 */ #include "codelet-dft.h" #ifdef HAVE_FMA /* Generated by: ../../../genfft/gen_notw_c.native -fma -reorder-insns -schedule-for-pipeline -simd -compact -variables 4 -pipeline-latency 8 -n 11 -name n1fv_11 -include n1f.h */ /* * This function contains 70 FP additions, 60 FP multiplications, * (or, 15 additions, 5 multiplications, 55 fused multiply/add), * 67 stack variables, 11 constants, and 22 memory accesses */ #include "n1f.h" static void n1fv_11(const R *ri, const R *ii, R *ro, R *io, stride is, stride os, INT v, INT ivs, INT ovs) { DVK(KP959492973, +0.959492973614497389890368057066327699062454848); DVK(KP876768831, +0.876768831002589333891339807079336796764054852); DVK(KP918985947, +0.918985947228994779780736114132655398124909697); DVK(KP989821441, +0.989821441880932732376092037776718787376519372); DVK(KP778434453, +0.778434453334651800608337670740821884709317477); DVK(KP830830026, +0.830830026003772851058548298459246407048009821); DVK(KP372785597, +0.372785597771792209609773152906148328659002598); DVK(KP634356270, +0.634356270682424498893150776899916060542806975); DVK(KP715370323, +0.715370323453429719112414662767260662417897278); DVK(KP342584725, +0.342584725681637509502641509861112333758894680); DVK(KP521108558, +0.521108558113202722944698153526659300680427422); { INT i; const R *xi; R *xo; xi = ri; xo = ro; for (i = v; i > 0; i = i - VL, xi = xi + (VL * ivs), xo = xo + (VL * ovs), MAKE_VOLATILE_STRIDE(22, is), MAKE_VOLATILE_STRIDE(22, os)) { V T1, Tb, T4, Tp, Tg, Tq, T7, Tn, Ta, Tm, Tc, Tr; T1 = LD(&(xi[0]), ivs, &(xi[0])); { V T2, T3, Te, Tf; T2 = LD(&(xi[WS(is, 1)]), ivs, &(xi[WS(is, 1)])); T3 = LD(&(xi[WS(is, 10)]), ivs, &(xi[0])); Te = LD(&(xi[WS(is, 5)]), ivs, &(xi[WS(is, 1)])); Tf = LD(&(xi[WS(is, 6)]), ivs, &(xi[0])); { V T5, T6, T8, T9; T5 = LD(&(xi[WS(is, 2)]), ivs, &(xi[0])); T6 = LD(&(xi[WS(is, 9)]), ivs, &(xi[WS(is, 1)])); T8 = LD(&(xi[WS(is, 3)]), ivs, &(xi[WS(is, 1)])); T9 = LD(&(xi[WS(is, 8)]), ivs, &(xi[0])); Tb = LD(&(xi[WS(is, 4)]), ivs, &(xi[0])); T4 = VADD(T2, T3); Tp = VSUB(T3, T2); Tg = VADD(Te, Tf); Tq = VSUB(Tf, Te); T7 = VADD(T5, T6); Tn = VSUB(T6, T5); Ta = VADD(T8, T9); Tm = VSUB(T9, T8); Tc = LD(&(xi[WS(is, 7)]), ivs, &(xi[WS(is, 1)])); } } Tr = VFMA(LDK(KP521108558), Tq, Tp); { V TS, TE, Th, Td, To, T12, TO, TB, T11, TN, TA, TF; T11 = VFNMS(LDK(KP521108558), Tp, Tn); TN = VFNMS(LDK(KP342584725), T7, Tg); TA = VFMA(LDK(KP521108558), Tm, Tq); TS = VFMA(LDK(KP715370323), Tm, Tp); TE = VFNMS(LDK(KP342584725), T4, Ta); Th = VFNMS(LDK(KP342584725), Ta, T7); Td = VADD(Tb, Tc); To = VSUB(Tc, Tb); T12 = VFNMS(LDK(KP715370323), T11, Tm); TO = VFNMS(LDK(KP634356270), TN, T4); TB = VFNMS(LDK(KP715370323), TA, Tn); TF = VFNMS(LDK(KP634356270), TE, Tg); { V T14, TD, TV, Tu, TY, Tx, Tk, TR, TI, TM, TJ, TT, Ts; TJ = VFNMS(LDK(KP521108558), Tn, To); TT = VFMA(LDK(KP372785597), To, TS); Ts = VFMA(LDK(KP715370323), Tr, To); ST(&(xo[0]), VADD(T1, VADD(T4, VADD(T7, VADD(Ta, VADD(Td, Tg))))), ovs, &(xo[0])); { V TW, Tv, Ti, T13; TW = VFNMS(LDK(KP342584725), Tg, Td); Tv = VFNMS(LDK(KP342584725), Td, T4); Ti = VFNMS(LDK(KP634356270), Th, Td); T13 = VFNMS(LDK(KP830830026), T12, To); { V TP, TC, TG, TK; TP = VFNMS(LDK(KP778434453), TO, Ta); TC = VFMA(LDK(KP830830026), TB, Tp); TG = VFNMS(LDK(KP778434453), TF, Td); TK = VFMA(LDK(KP715370323), TJ, Tq); { V TU, Tt, TX, Tw; TU = VFNMS(LDK(KP830830026), TT, Tq); Tt = VFMA(LDK(KP830830026), Ts, Tn); TX = VFNMS(LDK(KP634356270), TW, Ta); Tw = VFNMS(LDK(KP634356270), Tv, T7); { V Tj, TQ, TH, TL; Tj = VFNMS(LDK(KP778434453), Ti, T4); T14 = VMUL(LDK(KP989821441), VFNMS(LDK(KP918985947), T13, Tq)); TQ = VFNMS(LDK(KP876768831), TP, Td); TD = VMUL(LDK(KP989821441), VFNMS(LDK(KP918985947), TC, To)); TH = VFNMS(LDK(KP876768831), TG, T7); TL = VFNMS(LDK(KP830830026), TK, Tm); TV = VMUL(LDK(KP989821441), VFMA(LDK(KP918985947), TU, Tn)); Tu = VMUL(LDK(KP989821441), VFMA(LDK(KP918985947), Tt, Tm)); TY = VFNMS(LDK(KP778434453), TX, T7); Tx = VFNMS(LDK(KP778434453), Tw, Tg); Tk = VFNMS(LDK(KP876768831), Tj, Tg); TR = VFNMS(LDK(KP959492973), TQ, T1); TI = VFNMS(LDK(KP959492973), TH, T1); TM = VMUL(LDK(KP989821441), VFNMS(LDK(KP918985947), TL, Tp)); } } } } { V TZ, Ty, Tl, T10, Tz; TZ = VFNMS(LDK(KP876768831), TY, T4); Ty = VFNMS(LDK(KP876768831), Tx, Ta); Tl = VFNMS(LDK(KP959492973), Tk, T1); ST(&(xo[WS(os, 7)]), VFMAI(TV, TR), ovs, &(xo[WS(os, 1)])); ST(&(xo[WS(os, 4)]), VFNMSI(TV, TR), ovs, &(xo[0])); ST(&(xo[WS(os, 3)]), VFMAI(TM, TI), ovs, &(xo[WS(os, 1)])); ST(&(xo[WS(os, 8)]), VFNMSI(TM, TI), ovs, &(xo[0])); T10 = VFNMS(LDK(KP959492973), TZ, T1); Tz = VFNMS(LDK(KP959492973), Ty, T1); ST(&(xo[WS(os, 1)]), VFMAI(Tu, Tl), ovs, &(xo[WS(os, 1)])); ST(&(xo[WS(os, 10)]), VFNMSI(Tu, Tl), ovs, &(xo[0])); ST(&(xo[WS(os, 5)]), VFMAI(T14, T10), ovs, &(xo[WS(os, 1)])); ST(&(xo[WS(os, 6)]), VFNMSI(T14, T10), ovs, &(xo[0])); ST(&(xo[WS(os, 9)]), VFMAI(TD, Tz), ovs, &(xo[WS(os, 1)])); ST(&(xo[WS(os, 2)]), VFNMSI(TD, Tz), ovs, &(xo[0])); } } } } } VLEAVE(); } static const kdft_desc desc = { 11, XSIMD_STRING("n1fv_11"), {15, 5, 55, 0}, &GENUS, 0, 0, 0, 0 }; void XSIMD(codelet_n1fv_11) (planner *p) { X(kdft_register) (p, n1fv_11, &desc); } #else /* HAVE_FMA */ /* Generated by: ../../../genfft/gen_notw_c.native -simd -compact -variables 4 -pipeline-latency 8 -n 11 -name n1fv_11 -include n1f.h */ /* * This function contains 70 FP additions, 50 FP multiplications, * (or, 30 additions, 10 multiplications, 40 fused multiply/add), * 32 stack variables, 10 constants, and 22 memory accesses */ #include "n1f.h" static void n1fv_11(const R *ri, const R *ii, R *ro, R *io, stride is, stride os, INT v, INT ivs, INT ovs) { DVK(KP654860733, +0.654860733945285064056925072466293553183791199); DVK(KP142314838, +0.142314838273285140443792668616369668791051361); DVK(KP959492973, +0.959492973614497389890368057066327699062454848); DVK(KP415415013, +0.415415013001886425529274149229623203524004910); DVK(KP841253532, +0.841253532831181168861811648919367717513292498); DVK(KP989821441, +0.989821441880932732376092037776718787376519372); DVK(KP909631995, +0.909631995354518371411715383079028460060241051); DVK(KP281732556, +0.281732556841429697711417915346616899035777899); DVK(KP540640817, +0.540640817455597582107635954318691695431770608); DVK(KP755749574, +0.755749574354258283774035843972344420179717445); { INT i; const R *xi; R *xo; xi = ri; xo = ro; for (i = v; i > 0; i = i - VL, xi = xi + (VL * ivs), xo = xo + (VL * ovs), MAKE_VOLATILE_STRIDE(22, is), MAKE_VOLATILE_STRIDE(22, os)) { V T1, T4, Ti, Tg, Tl, Td, Tk, Ta, Tj, T7, Tm, Tb, Tc, Tt, Ts; T1 = LD(&(xi[0]), ivs, &(xi[0])); { V T2, T3, Te, Tf; T2 = LD(&(xi[WS(is, 1)]), ivs, &(xi[WS(is, 1)])); T3 = LD(&(xi[WS(is, 10)]), ivs, &(xi[0])); T4 = VADD(T2, T3); Ti = VSUB(T3, T2); Te = LD(&(xi[WS(is, 5)]), ivs, &(xi[WS(is, 1)])); Tf = LD(&(xi[WS(is, 6)]), ivs, &(xi[0])); Tg = VADD(Te, Tf); Tl = VSUB(Tf, Te); } Tb = LD(&(xi[WS(is, 4)]), ivs, &(xi[0])); Tc = LD(&(xi[WS(is, 7)]), ivs, &(xi[WS(is, 1)])); Td = VADD(Tb, Tc); Tk = VSUB(Tc, Tb); { V T8, T9, T5, T6; T8 = LD(&(xi[WS(is, 3)]), ivs, &(xi[WS(is, 1)])); T9 = LD(&(xi[WS(is, 8)]), ivs, &(xi[0])); Ta = VADD(T8, T9); Tj = VSUB(T9, T8); T5 = LD(&(xi[WS(is, 2)]), ivs, &(xi[0])); T6 = LD(&(xi[WS(is, 9)]), ivs, &(xi[WS(is, 1)])); T7 = VADD(T5, T6); Tm = VSUB(T6, T5); } ST(&(xo[0]), VADD(T1, VADD(T4, VADD(T7, VADD(Ta, VADD(Td, Tg))))), ovs, &(xo[0])); { V Tn, Th, Tv, Tu; Tn = VBYI(VFMA(LDK(KP755749574), Ti, VFMA(LDK(KP540640817), Tj, VFNMS(LDK(KP909631995), Tl, VFNMS(LDK(KP989821441), Tm, VMUL(LDK(KP281732556), Tk)))))); Th = VFMA(LDK(KP841253532), Ta, VFMA(LDK(KP415415013), Tg, VFNMS(LDK(KP959492973), Td, VFNMS(LDK(KP142314838), T7, VFNMS(LDK(KP654860733), T4, T1))))); ST(&(xo[WS(os, 7)]), VSUB(Th, Tn), ovs, &(xo[WS(os, 1)])); ST(&(xo[WS(os, 4)]), VADD(Th, Tn), ovs, &(xo[0])); Tv = VBYI(VFMA(LDK(KP281732556), Ti, VFMA(LDK(KP755749574), Tj, VFNMS(LDK(KP909631995), Tk, VFNMS(LDK(KP540640817), Tm, VMUL(LDK(KP989821441), Tl)))))); Tu = VFMA(LDK(KP841253532), T7, VFMA(LDK(KP415415013), Td, VFNMS(LDK(KP142314838), Tg, VFNMS(LDK(KP654860733), Ta, VFNMS(LDK(KP959492973), T4, T1))))); ST(&(xo[WS(os, 6)]), VSUB(Tu, Tv), ovs, &(xo[0])); ST(&(xo[WS(os, 5)]), VADD(Tu, Tv), ovs, &(xo[WS(os, 1)])); } Tt = VBYI(VFMA(LDK(KP989821441), Ti, VFMA(LDK(KP540640817), Tk, VFNMS(LDK(KP909631995), Tj, VFNMS(LDK(KP281732556), Tm, VMUL(LDK(KP755749574), Tl)))))); Ts = VFMA(LDK(KP415415013), Ta, VFMA(LDK(KP841253532), Td, VFNMS(LDK(KP654860733), Tg, VFNMS(LDK(KP959492973), T7, VFNMS(LDK(KP142314838), T4, T1))))); ST(&(xo[WS(os, 8)]), VSUB(Ts, Tt), ovs, &(xo[0])); ST(&(xo[WS(os, 3)]), VADD(Ts, Tt), ovs, &(xo[WS(os, 1)])); { V Tr, Tq, Tp, To; Tr = VBYI(VFMA(LDK(KP540640817), Ti, VFMA(LDK(KP909631995), Tm, VFMA(LDK(KP989821441), Tj, VFMA(LDK(KP755749574), Tk, VMUL(LDK(KP281732556), Tl)))))); Tq = VFMA(LDK(KP841253532), T4, VFMA(LDK(KP415415013), T7, VFNMS(LDK(KP959492973), Tg, VFNMS(LDK(KP654860733), Td, VFNMS(LDK(KP142314838), Ta, T1))))); ST(&(xo[WS(os, 10)]), VSUB(Tq, Tr), ovs, &(xo[0])); ST(&(xo[WS(os, 1)]), VADD(Tq, Tr), ovs, &(xo[WS(os, 1)])); Tp = VBYI(VFMA(LDK(KP909631995), Ti, VFNMS(LDK(KP540640817), Tl, VFNMS(LDK(KP989821441), Tk, VFNMS(LDK(KP281732556), Tj, VMUL(LDK(KP755749574), Tm)))))); To = VFMA(LDK(KP415415013), T4, VFMA(LDK(KP841253532), Tg, VFNMS(LDK(KP142314838), Td, VFNMS(LDK(KP959492973), Ta, VFNMS(LDK(KP654860733), T7, T1))))); ST(&(xo[WS(os, 9)]), VSUB(To, Tp), ovs, &(xo[WS(os, 1)])); ST(&(xo[WS(os, 2)]), VADD(To, Tp), ovs, &(xo[0])); } } } VLEAVE(); } static const kdft_desc desc = { 11, XSIMD_STRING("n1fv_11"), {30, 10, 40, 0}, &GENUS, 0, 0, 0, 0 }; void XSIMD(codelet_n1fv_11) (planner *p) { X(kdft_register) (p, n1fv_11, &desc); } #endif /* HAVE_FMA */ fftw-3.3.3/dft/simd/common/n1fv_12.c0000644000175000017500000002106712054410344013733 00000000000000/* * Copyright (c) 2003, 2007-11 Matteo Frigo * Copyright (c) 2003, 2007-11 Massachusetts Institute of Technology * * This program is free software; you can redistribute it and/or modify * it under the terms of the GNU General Public License as published by * the Free Software Foundation; either version 2 of the License, or * (at your option) any later version. * * This program is distributed in the hope that it will be useful, * but WITHOUT ANY WARRANTY; without even the implied warranty of * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the * GNU General Public License for more details. * * You should have received a copy of the GNU General Public License * along with this program; if not, write to the Free Software * Foundation, Inc., 51 Franklin Street, Fifth Floor, Boston, MA 02110-1301 USA * */ /* This file was automatically generated --- DO NOT EDIT */ /* Generated on Sun Nov 25 07:36:52 EST 2012 */ #include "codelet-dft.h" #ifdef HAVE_FMA /* Generated by: ../../../genfft/gen_notw_c.native -fma -reorder-insns -schedule-for-pipeline -simd -compact -variables 4 -pipeline-latency 8 -n 12 -name n1fv_12 -include n1f.h */ /* * This function contains 48 FP additions, 20 FP multiplications, * (or, 30 additions, 2 multiplications, 18 fused multiply/add), * 49 stack variables, 2 constants, and 24 memory accesses */ #include "n1f.h" static void n1fv_12(const R *ri, const R *ii, R *ro, R *io, stride is, stride os, INT v, INT ivs, INT ovs) { DVK(KP866025403, +0.866025403784438646763723170752936183471402627); DVK(KP500000000, +0.500000000000000000000000000000000000000000000); { INT i; const R *xi; R *xo; xi = ri; xo = ro; for (i = v; i > 0; i = i - VL, xi = xi + (VL * ivs), xo = xo + (VL * ovs), MAKE_VOLATILE_STRIDE(24, is), MAKE_VOLATILE_STRIDE(24, os)) { V T1, T6, Tk, Tn, Tc, Td, Tf, Tr, T4, Ts, T9, Tg, Te, Tl; { V T2, T3, T7, T8; T1 = LD(&(xi[0]), ivs, &(xi[0])); T6 = LD(&(xi[WS(is, 6)]), ivs, &(xi[0])); T2 = LD(&(xi[WS(is, 4)]), ivs, &(xi[0])); T3 = LD(&(xi[WS(is, 8)]), ivs, &(xi[0])); T7 = LD(&(xi[WS(is, 10)]), ivs, &(xi[0])); T8 = LD(&(xi[WS(is, 2)]), ivs, &(xi[0])); Tk = LD(&(xi[WS(is, 3)]), ivs, &(xi[WS(is, 1)])); Tn = LD(&(xi[WS(is, 9)]), ivs, &(xi[WS(is, 1)])); Tc = LD(&(xi[WS(is, 11)]), ivs, &(xi[WS(is, 1)])); Td = LD(&(xi[WS(is, 7)]), ivs, &(xi[WS(is, 1)])); Tf = LD(&(xi[WS(is, 1)]), ivs, &(xi[WS(is, 1)])); Tr = VSUB(T3, T2); T4 = VADD(T2, T3); Ts = VSUB(T8, T7); T9 = VADD(T7, T8); Tg = LD(&(xi[WS(is, 5)]), ivs, &(xi[WS(is, 1)])); } Te = VSUB(Tc, Td); Tl = VADD(Td, Tc); { V T5, TF, TB, Tt, Ta, TG, Th, To, Tm, TI; T5 = VFNMS(LDK(KP500000000), T4, T1); TF = VADD(T1, T4); TB = VADD(Tr, Ts); Tt = VSUB(Tr, Ts); Ta = VFNMS(LDK(KP500000000), T9, T6); TG = VADD(T6, T9); Th = VSUB(Tf, Tg); To = VADD(Tf, Tg); Tm = VFNMS(LDK(KP500000000), Tl, Tk); TI = VADD(Tk, Tl); { V TH, TL, Tb, Tx, TJ, Tp, Ti, TA; TH = VSUB(TF, TG); TL = VADD(TF, TG); Tb = VSUB(T5, Ta); Tx = VADD(T5, Ta); TJ = VADD(Tn, To); Tp = VFNMS(LDK(KP500000000), To, Tn); Ti = VADD(Te, Th); TA = VSUB(Te, Th); { V Tq, Ty, TK, TM; Tq = VSUB(Tm, Tp); Ty = VADD(Tm, Tp); TK = VSUB(TI, TJ); TM = VADD(TI, TJ); { V TC, TE, Tj, Tv; TC = VMUL(LDK(KP866025403), VSUB(TA, TB)); TE = VMUL(LDK(KP866025403), VADD(TB, TA)); Tj = VFMA(LDK(KP866025403), Ti, Tb); Tv = VFNMS(LDK(KP866025403), Ti, Tb); { V Tz, TD, Tu, Tw; Tz = VSUB(Tx, Ty); TD = VADD(Tx, Ty); Tu = VFNMS(LDK(KP866025403), Tt, Tq); Tw = VFMA(LDK(KP866025403), Tt, Tq); ST(&(xo[0]), VADD(TL, TM), ovs, &(xo[0])); ST(&(xo[WS(os, 6)]), VSUB(TL, TM), ovs, &(xo[0])); ST(&(xo[WS(os, 3)]), VFMAI(TK, TH), ovs, &(xo[WS(os, 1)])); ST(&(xo[WS(os, 9)]), VFNMSI(TK, TH), ovs, &(xo[WS(os, 1)])); ST(&(xo[WS(os, 4)]), VFMAI(TE, TD), ovs, &(xo[0])); ST(&(xo[WS(os, 8)]), VFNMSI(TE, TD), ovs, &(xo[0])); ST(&(xo[WS(os, 10)]), VFNMSI(TC, Tz), ovs, &(xo[0])); ST(&(xo[WS(os, 2)]), VFMAI(TC, Tz), ovs, &(xo[0])); ST(&(xo[WS(os, 5)]), VFNMSI(Tw, Tv), ovs, &(xo[WS(os, 1)])); ST(&(xo[WS(os, 7)]), VFMAI(Tw, Tv), ovs, &(xo[WS(os, 1)])); ST(&(xo[WS(os, 11)]), VFMAI(Tu, Tj), ovs, &(xo[WS(os, 1)])); ST(&(xo[WS(os, 1)]), VFNMSI(Tu, Tj), ovs, &(xo[WS(os, 1)])); } } } } } } } VLEAVE(); } static const kdft_desc desc = { 12, XSIMD_STRING("n1fv_12"), {30, 2, 18, 0}, &GENUS, 0, 0, 0, 0 }; void XSIMD(codelet_n1fv_12) (planner *p) { X(kdft_register) (p, n1fv_12, &desc); } #else /* HAVE_FMA */ /* Generated by: ../../../genfft/gen_notw_c.native -simd -compact -variables 4 -pipeline-latency 8 -n 12 -name n1fv_12 -include n1f.h */ /* * This function contains 48 FP additions, 8 FP multiplications, * (or, 44 additions, 4 multiplications, 4 fused multiply/add), * 27 stack variables, 2 constants, and 24 memory accesses */ #include "n1f.h" static void n1fv_12(const R *ri, const R *ii, R *ro, R *io, stride is, stride os, INT v, INT ivs, INT ovs) { DVK(KP500000000, +0.500000000000000000000000000000000000000000000); DVK(KP866025403, +0.866025403784438646763723170752936183471402627); { INT i; const R *xi; R *xo; xi = ri; xo = ro; for (i = v; i > 0; i = i - VL, xi = xi + (VL * ivs), xo = xo + (VL * ovs), MAKE_VOLATILE_STRIDE(24, is), MAKE_VOLATILE_STRIDE(24, os)) { V T5, Ta, TJ, Ty, Tq, Tp, Tg, Tl, TI, TA, Tz, Tu; { V T1, T6, T4, Tw, T9, Tx; T1 = LD(&(xi[0]), ivs, &(xi[0])); T6 = LD(&(xi[WS(is, 6)]), ivs, &(xi[0])); { V T2, T3, T7, T8; T2 = LD(&(xi[WS(is, 4)]), ivs, &(xi[0])); T3 = LD(&(xi[WS(is, 8)]), ivs, &(xi[0])); T4 = VADD(T2, T3); Tw = VSUB(T3, T2); T7 = LD(&(xi[WS(is, 10)]), ivs, &(xi[0])); T8 = LD(&(xi[WS(is, 2)]), ivs, &(xi[0])); T9 = VADD(T7, T8); Tx = VSUB(T8, T7); } T5 = VADD(T1, T4); Ta = VADD(T6, T9); TJ = VADD(Tw, Tx); Ty = VMUL(LDK(KP866025403), VSUB(Tw, Tx)); Tq = VFNMS(LDK(KP500000000), T9, T6); Tp = VFNMS(LDK(KP500000000), T4, T1); } { V Tc, Th, Tf, Ts, Tk, Tt; Tc = LD(&(xi[WS(is, 3)]), ivs, &(xi[WS(is, 1)])); Th = LD(&(xi[WS(is, 9)]), ivs, &(xi[WS(is, 1)])); { V Td, Te, Ti, Tj; Td = LD(&(xi[WS(is, 7)]), ivs, &(xi[WS(is, 1)])); Te = LD(&(xi[WS(is, 11)]), ivs, &(xi[WS(is, 1)])); Tf = VADD(Td, Te); Ts = VSUB(Te, Td); Ti = LD(&(xi[WS(is, 1)]), ivs, &(xi[WS(is, 1)])); Tj = LD(&(xi[WS(is, 5)]), ivs, &(xi[WS(is, 1)])); Tk = VADD(Ti, Tj); Tt = VSUB(Tj, Ti); } Tg = VADD(Tc, Tf); Tl = VADD(Th, Tk); TI = VADD(Ts, Tt); TA = VFNMS(LDK(KP500000000), Tk, Th); Tz = VFNMS(LDK(KP500000000), Tf, Tc); Tu = VMUL(LDK(KP866025403), VSUB(Ts, Tt)); } { V Tb, Tm, Tn, To; Tb = VSUB(T5, Ta); Tm = VBYI(VSUB(Tg, Tl)); ST(&(xo[WS(os, 9)]), VSUB(Tb, Tm), ovs, &(xo[WS(os, 1)])); ST(&(xo[WS(os, 3)]), VADD(Tb, Tm), ovs, &(xo[WS(os, 1)])); Tn = VADD(T5, Ta); To = VADD(Tg, Tl); ST(&(xo[WS(os, 6)]), VSUB(Tn, To), ovs, &(xo[0])); ST(&(xo[0]), VADD(Tn, To), ovs, &(xo[0])); } { V Tv, TE, TC, TD, Tr, TB; Tr = VSUB(Tp, Tq); Tv = VSUB(Tr, Tu); TE = VADD(Tr, Tu); TB = VSUB(Tz, TA); TC = VBYI(VADD(Ty, TB)); TD = VBYI(VSUB(Ty, TB)); ST(&(xo[WS(os, 5)]), VSUB(Tv, TC), ovs, &(xo[WS(os, 1)])); ST(&(xo[WS(os, 11)]), VSUB(TE, TD), ovs, &(xo[WS(os, 1)])); ST(&(xo[WS(os, 7)]), VADD(TC, Tv), ovs, &(xo[WS(os, 1)])); ST(&(xo[WS(os, 1)]), VADD(TD, TE), ovs, &(xo[WS(os, 1)])); } { V TK, TM, TH, TL, TF, TG; TK = VBYI(VMUL(LDK(KP866025403), VSUB(TI, TJ))); TM = VBYI(VMUL(LDK(KP866025403), VADD(TJ, TI))); TF = VADD(Tp, Tq); TG = VADD(Tz, TA); TH = VSUB(TF, TG); TL = VADD(TF, TG); ST(&(xo[WS(os, 10)]), VSUB(TH, TK), ovs, &(xo[0])); ST(&(xo[WS(os, 4)]), VADD(TL, TM), ovs, &(xo[0])); ST(&(xo[WS(os, 2)]), VADD(TH, TK), ovs, &(xo[0])); ST(&(xo[WS(os, 8)]), VSUB(TL, TM), ovs, &(xo[0])); } } } VLEAVE(); } static const kdft_desc desc = { 12, XSIMD_STRING("n1fv_12"), {44, 4, 4, 0}, &GENUS, 0, 0, 0, 0 }; void XSIMD(codelet_n1fv_12) (planner *p) { X(kdft_register) (p, n1fv_12, &desc); } #endif /* HAVE_FMA */ fftw-3.3.3/dft/simd/common/n1fv_13.c0000644000175000017500000003733612054410346013744 00000000000000/* * Copyright (c) 2003, 2007-11 Matteo Frigo * Copyright (c) 2003, 2007-11 Massachusetts Institute of Technology * * This program is free software; you can redistribute it and/or modify * it under the terms of the GNU General Public License as published by * the Free Software Foundation; either version 2 of the License, or * (at your option) any later version. * * This program is distributed in the hope that it will be useful, * but WITHOUT ANY WARRANTY; without even the implied warranty of * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the * GNU General Public License for more details. * * You should have received a copy of the GNU General Public License * along with this program; if not, write to the Free Software * Foundation, Inc., 51 Franklin Street, Fifth Floor, Boston, MA 02110-1301 USA * */ /* This file was automatically generated --- DO NOT EDIT */ /* Generated on Sun Nov 25 07:36:52 EST 2012 */ #include "codelet-dft.h" #ifdef HAVE_FMA /* Generated by: ../../../genfft/gen_notw_c.native -fma -reorder-insns -schedule-for-pipeline -simd -compact -variables 4 -pipeline-latency 8 -n 13 -name n1fv_13 -include n1f.h */ /* * This function contains 88 FP additions, 63 FP multiplications, * (or, 31 additions, 6 multiplications, 57 fused multiply/add), * 96 stack variables, 23 constants, and 26 memory accesses */ #include "n1f.h" static void n1fv_13(const R *ri, const R *ii, R *ro, R *io, stride is, stride os, INT v, INT ivs, INT ovs) { DVK(KP904176221, +0.904176221990848204433795481776887926501523162); DVK(KP575140729, +0.575140729474003121368385547455453388461001608); DVK(KP300462606, +0.300462606288665774426601772289207995520941381); DVK(KP516520780, +0.516520780623489722840901288569017135705033622); DVK(KP522026385, +0.522026385161275033714027226654165028300441940); DVK(KP957805992, +0.957805992594665126462521754605754580515587217); DVK(KP600477271, +0.600477271932665282925769253334763009352012849); DVK(KP251768516, +0.251768516431883313623436926934233488546674281); DVK(KP503537032, +0.503537032863766627246873853868466977093348562); DVK(KP769338817, +0.769338817572980603471413688209101117038278899); DVK(KP859542535, +0.859542535098774820163672132761689612766401925); DVK(KP581704778, +0.581704778510515730456870384989698884939833902); DVK(KP853480001, +0.853480001859823990758994934970528322872359049); DVK(KP083333333, +0.083333333333333333333333333333333333333333333); DVK(KP226109445, +0.226109445035782405468510155372505010481906348); DVK(KP301479260, +0.301479260047709873958013540496673347309208464); DVK(KP686558370, +0.686558370781754340655719594850823015421401653); DVK(KP514918778, +0.514918778086315755491789696138117261566051239); DVK(KP038632954, +0.038632954644348171955506895830342264440241080); DVK(KP612264650, +0.612264650376756543746494474777125408779395514); DVK(KP302775637, +0.302775637731994646559610633735247973125648287); DVK(KP866025403, +0.866025403784438646763723170752936183471402627); DVK(KP500000000, +0.500000000000000000000000000000000000000000000); { INT i; const R *xi; R *xo; xi = ri; xo = ro; for (i = v; i > 0; i = i - VL, xi = xi + (VL * ivs), xo = xo + (VL * ovs), MAKE_VOLATILE_STRIDE(26, is), MAKE_VOLATILE_STRIDE(26, os)) { V T1, T7, T2, Tg, Tf, TN, Th, Tq, Ta, Tj, T5, Tr, Tk; T1 = LD(&(xi[0]), ivs, &(xi[0])); { V Td, Te, T8, T9, T3, T4; Td = LD(&(xi[WS(is, 8)]), ivs, &(xi[0])); Te = LD(&(xi[WS(is, 5)]), ivs, &(xi[WS(is, 1)])); T7 = LD(&(xi[WS(is, 12)]), ivs, &(xi[0])); T8 = LD(&(xi[WS(is, 10)]), ivs, &(xi[0])); T9 = LD(&(xi[WS(is, 4)]), ivs, &(xi[0])); T2 = LD(&(xi[WS(is, 1)]), ivs, &(xi[WS(is, 1)])); T3 = LD(&(xi[WS(is, 3)]), ivs, &(xi[WS(is, 1)])); T4 = LD(&(xi[WS(is, 9)]), ivs, &(xi[WS(is, 1)])); Tg = LD(&(xi[WS(is, 11)]), ivs, &(xi[WS(is, 1)])); Tf = VADD(Td, Te); TN = VSUB(Td, Te); Th = LD(&(xi[WS(is, 6)]), ivs, &(xi[0])); Tq = VSUB(T8, T9); Ta = VADD(T8, T9); Tj = LD(&(xi[WS(is, 7)]), ivs, &(xi[WS(is, 1)])); T5 = VADD(T3, T4); Tr = VSUB(T4, T3); Tk = LD(&(xi[WS(is, 2)]), ivs, &(xi[0])); } { V Tt, Ti, Ty, Tb, Ts, TQ, Tx, T6, Tu, Tl; Tt = VSUB(Tg, Th); Ti = VADD(Tg, Th); Ty = VFMS(LDK(KP500000000), Ta, T7); Tb = VADD(T7, Ta); Ts = VSUB(Tq, Tr); TQ = VADD(Tr, Tq); Tx = VFNMS(LDK(KP500000000), T5, T2); T6 = VADD(T2, T5); Tu = VSUB(Tj, Tk); Tl = VADD(Tj, Tk); { V TK, Tz, Tc, TX, Tv, TO, TL, Tm; TK = VADD(Tx, Ty); Tz = VSUB(Tx, Ty); Tc = VADD(T6, Tb); TX = VSUB(T6, Tb); Tv = VSUB(Tt, Tu); TO = VADD(Tt, Tu); TL = VSUB(Ti, Tl); Tm = VADD(Ti, Tl); { V TF, Tw, TP, TY, TT, TM, TA, Tn; TF = VSUB(Ts, Tv); Tw = VADD(Ts, Tv); TP = VFNMS(LDK(KP500000000), TO, TN); TY = VADD(TN, TO); TT = VFNMS(LDK(KP866025403), TL, TK); TM = VFMA(LDK(KP866025403), TL, TK); TA = VFNMS(LDK(KP500000000), Tm, Tf); Tn = VADD(Tf, Tm); { V T1f, T1n, TI, T18, T1k, T1c, TD, T17, T10, T1m, T16, T1e, TU, TR; TU = VFNMS(LDK(KP866025403), TQ, TP); TR = VFMA(LDK(KP866025403), TQ, TP); { V TZ, T15, TE, TB; TZ = VFMA(LDK(KP302775637), TY, TX); T15 = VFNMS(LDK(KP302775637), TX, TY); TE = VSUB(Tz, TA); TB = VADD(Tz, TA); { V TH, To, TV, T13; TH = VSUB(Tc, Tn); To = VADD(Tc, Tn); TV = VFNMS(LDK(KP612264650), TU, TT); T13 = VFMA(LDK(KP612264650), TT, TU); { V TS, T12, TG, T1b; TS = VFNMS(LDK(KP038632954), TR, TM); T12 = VFMA(LDK(KP038632954), TM, TR); TG = VFNMS(LDK(KP514918778), TF, TE); T1b = VFMA(LDK(KP686558370), TE, TF); { V TC, T1a, Tp, TW, T14; TC = VFMA(LDK(KP301479260), TB, Tw); T1a = VFNMS(LDK(KP226109445), Tw, TB); Tp = VFNMS(LDK(KP083333333), To, T1); ST(&(xo[0]), VADD(T1, To), ovs, &(xo[0])); T1f = VFMA(LDK(KP853480001), TV, TS); TW = VFNMS(LDK(KP853480001), TV, TS); T1n = VFMA(LDK(KP853480001), T13, T12); T14 = VFNMS(LDK(KP853480001), T13, T12); TI = VFMA(LDK(KP581704778), TH, TG); T18 = VFNMS(LDK(KP859542535), TG, TH); T1k = VFMA(LDK(KP769338817), T1b, T1a); T1c = VFNMS(LDK(KP769338817), T1b, T1a); TD = VFMA(LDK(KP503537032), TC, Tp); T17 = VFNMS(LDK(KP251768516), TC, Tp); T10 = VMUL(LDK(KP600477271), VFMA(LDK(KP957805992), TZ, TW)); T1m = VFNMS(LDK(KP522026385), TW, TZ); T16 = VMUL(LDK(KP600477271), VFMA(LDK(KP957805992), T15, T14)); T1e = VFNMS(LDK(KP522026385), T14, T15); } } } } { V T1o, T1q, T1g, T1i, T1d, T1h, T1l, T1p; { V T11, TJ, T19, T1j; T11 = VFMA(LDK(KP516520780), TI, TD); TJ = VFNMS(LDK(KP516520780), TI, TD); T19 = VFMA(LDK(KP300462606), T18, T17); T1j = VFNMS(LDK(KP300462606), T18, T17); T1o = VMUL(LDK(KP575140729), VFNMS(LDK(KP904176221), T1n, T1m)); T1q = VMUL(LDK(KP575140729), VFMA(LDK(KP904176221), T1n, T1m)); T1g = VMUL(LDK(KP575140729), VFMA(LDK(KP904176221), T1f, T1e)); T1i = VMUL(LDK(KP575140729), VFNMS(LDK(KP904176221), T1f, T1e)); ST(&(xo[WS(os, 12)]), VFNMSI(T16, T11), ovs, &(xo[0])); ST(&(xo[WS(os, 1)]), VFMAI(T16, T11), ovs, &(xo[WS(os, 1)])); ST(&(xo[WS(os, 8)]), VFMAI(T10, TJ), ovs, &(xo[0])); ST(&(xo[WS(os, 5)]), VFNMSI(T10, TJ), ovs, &(xo[WS(os, 1)])); T1d = VFNMS(LDK(KP503537032), T1c, T19); T1h = VFMA(LDK(KP503537032), T1c, T19); T1l = VFNMS(LDK(KP503537032), T1k, T1j); T1p = VFMA(LDK(KP503537032), T1k, T1j); } ST(&(xo[WS(os, 9)]), VFMAI(T1g, T1d), ovs, &(xo[WS(os, 1)])); ST(&(xo[WS(os, 4)]), VFNMSI(T1g, T1d), ovs, &(xo[0])); ST(&(xo[WS(os, 10)]), VFNMSI(T1i, T1h), ovs, &(xo[0])); ST(&(xo[WS(os, 3)]), VFMAI(T1i, T1h), ovs, &(xo[WS(os, 1)])); ST(&(xo[WS(os, 7)]), VFMAI(T1o, T1l), ovs, &(xo[WS(os, 1)])); ST(&(xo[WS(os, 6)]), VFNMSI(T1o, T1l), ovs, &(xo[0])); ST(&(xo[WS(os, 11)]), VFMAI(T1q, T1p), ovs, &(xo[WS(os, 1)])); ST(&(xo[WS(os, 2)]), VFNMSI(T1q, T1p), ovs, &(xo[0])); } } } } } } } VLEAVE(); } static const kdft_desc desc = { 13, XSIMD_STRING("n1fv_13"), {31, 6, 57, 0}, &GENUS, 0, 0, 0, 0 }; void XSIMD(codelet_n1fv_13) (planner *p) { X(kdft_register) (p, n1fv_13, &desc); } #else /* HAVE_FMA */ /* Generated by: ../../../genfft/gen_notw_c.native -simd -compact -variables 4 -pipeline-latency 8 -n 13 -name n1fv_13 -include n1f.h */ /* * This function contains 88 FP additions, 34 FP multiplications, * (or, 69 additions, 15 multiplications, 19 fused multiply/add), * 60 stack variables, 20 constants, and 26 memory accesses */ #include "n1f.h" static void n1fv_13(const R *ri, const R *ii, R *ro, R *io, stride is, stride os, INT v, INT ivs, INT ovs) { DVK(KP2_000000000, +2.000000000000000000000000000000000000000000000); DVK(KP083333333, +0.083333333333333333333333333333333333333333333); DVK(KP075902986, +0.075902986037193865983102897245103540356428373); DVK(KP251768516, +0.251768516431883313623436926934233488546674281); DVK(KP132983124, +0.132983124607418643793760531921092974399165133); DVK(KP258260390, +0.258260390311744861420450644284508567852516811); DVK(KP1_732050807, +1.732050807568877293527446341505872366942805254); DVK(KP300238635, +0.300238635966332641462884626667381504676006424); DVK(KP011599105, +0.011599105605768290721655456654083252189827041); DVK(KP156891391, +0.156891391051584611046832726756003269660212636); DVK(KP256247671, +0.256247671582936600958684654061725059144125175); DVK(KP174138601, +0.174138601152135905005660794929264742616964676); DVK(KP575140729, +0.575140729474003121368385547455453388461001608); DVK(KP503537032, +0.503537032863766627246873853868466977093348562); DVK(KP113854479, +0.113854479055790798974654345867655310534642560); DVK(KP265966249, +0.265966249214837287587521063842185948798330267); DVK(KP387390585, +0.387390585467617292130675966426762851778775217); DVK(KP300462606, +0.300462606288665774426601772289207995520941381); DVK(KP866025403, +0.866025403784438646763723170752936183471402627); DVK(KP500000000, +0.500000000000000000000000000000000000000000000); { INT i; const R *xi; R *xo; xi = ri; xo = ro; for (i = v; i > 0; i = i - VL, xi = xi + (VL * ivs), xo = xo + (VL * ovs), MAKE_VOLATILE_STRIDE(26, is), MAKE_VOLATILE_STRIDE(26, os)) { V TW, Tb, Tm, Tu, TC, TR, TX, TK, TU, Tz, TB, TN, TT; TW = LD(&(xi[0]), ivs, &(xi[0])); { V T3, TH, Tl, Tw, Tp, Tg, Tv, To, T6, Tr, T9, Ts, Ta, TI, T1; V T2, Tq, Tt; T1 = LD(&(xi[WS(is, 8)]), ivs, &(xi[0])); T2 = LD(&(xi[WS(is, 5)]), ivs, &(xi[WS(is, 1)])); T3 = VSUB(T1, T2); TH = VADD(T1, T2); { V Th, Ti, Tj, Tk; Th = LD(&(xi[WS(is, 12)]), ivs, &(xi[0])); Ti = LD(&(xi[WS(is, 10)]), ivs, &(xi[0])); Tj = LD(&(xi[WS(is, 4)]), ivs, &(xi[0])); Tk = VADD(Ti, Tj); Tl = VADD(Th, Tk); Tw = VSUB(Ti, Tj); Tp = VFNMS(LDK(KP500000000), Tk, Th); } { V Tc, Td, Te, Tf; Tc = LD(&(xi[WS(is, 1)]), ivs, &(xi[WS(is, 1)])); Td = LD(&(xi[WS(is, 3)]), ivs, &(xi[WS(is, 1)])); Te = LD(&(xi[WS(is, 9)]), ivs, &(xi[WS(is, 1)])); Tf = VADD(Td, Te); Tg = VADD(Tc, Tf); Tv = VSUB(Td, Te); To = VFNMS(LDK(KP500000000), Tf, Tc); } { V T4, T5, T7, T8; T4 = LD(&(xi[WS(is, 11)]), ivs, &(xi[WS(is, 1)])); T5 = LD(&(xi[WS(is, 6)]), ivs, &(xi[0])); T6 = VSUB(T4, T5); Tr = VADD(T4, T5); T7 = LD(&(xi[WS(is, 7)]), ivs, &(xi[WS(is, 1)])); T8 = LD(&(xi[WS(is, 2)]), ivs, &(xi[0])); T9 = VSUB(T7, T8); Ts = VADD(T7, T8); } Ta = VADD(T6, T9); TI = VADD(Tr, Ts); Tb = VADD(T3, Ta); Tm = VSUB(Tg, Tl); Tq = VSUB(To, Tp); Tt = VMUL(LDK(KP866025403), VSUB(Tr, Ts)); Tu = VADD(Tq, Tt); TC = VSUB(Tq, Tt); { V TP, TQ, TG, TJ; TP = VADD(Tg, Tl); TQ = VADD(TH, TI); TR = VMUL(LDK(KP300462606), VSUB(TP, TQ)); TX = VADD(TP, TQ); TG = VADD(To, Tp); TJ = VFNMS(LDK(KP500000000), TI, TH); TK = VSUB(TG, TJ); TU = VADD(TG, TJ); } { V Tx, Ty, TL, TM; Tx = VMUL(LDK(KP866025403), VSUB(Tv, Tw)); Ty = VFNMS(LDK(KP500000000), Ta, T3); Tz = VSUB(Tx, Ty); TB = VADD(Tx, Ty); TL = VADD(Tv, Tw); TM = VSUB(T6, T9); TN = VSUB(TL, TM); TT = VADD(TL, TM); } } ST(&(xo[0]), VADD(TW, TX), ovs, &(xo[0])); { V T19, T1n, T14, T13, T1f, T1k, Tn, TE, T1e, T1j, TS, T1m, TZ, T1c, TA; V TD; { V T17, T18, T11, T12; T17 = VFMA(LDK(KP387390585), TN, VMUL(LDK(KP265966249), TK)); T18 = VFNMS(LDK(KP503537032), TU, VMUL(LDK(KP113854479), TT)); T19 = VSUB(T17, T18); T1n = VADD(T17, T18); T14 = VFMA(LDK(KP575140729), Tm, VMUL(LDK(KP174138601), Tb)); T11 = VFNMS(LDK(KP156891391), TB, VMUL(LDK(KP256247671), TC)); T12 = VFMA(LDK(KP011599105), Tz, VMUL(LDK(KP300238635), Tu)); T13 = VSUB(T11, T12); T1f = VADD(T14, T13); T1k = VMUL(LDK(KP1_732050807), VADD(T11, T12)); } Tn = VFNMS(LDK(KP174138601), Tm, VMUL(LDK(KP575140729), Tb)); TA = VFNMS(LDK(KP300238635), Tz, VMUL(LDK(KP011599105), Tu)); TD = VFMA(LDK(KP256247671), TB, VMUL(LDK(KP156891391), TC)); TE = VSUB(TA, TD); T1e = VMUL(LDK(KP1_732050807), VADD(TD, TA)); T1j = VSUB(Tn, TE); { V TO, T1b, TV, TY, T1a; TO = VFNMS(LDK(KP132983124), TN, VMUL(LDK(KP258260390), TK)); T1b = VSUB(TR, TO); TV = VFMA(LDK(KP251768516), TT, VMUL(LDK(KP075902986), TU)); TY = VFNMS(LDK(KP083333333), TX, TW); T1a = VSUB(TY, TV); TS = VFMA(LDK(KP2_000000000), TO, TR); T1m = VADD(T1b, T1a); TZ = VFMA(LDK(KP2_000000000), TV, TY); T1c = VSUB(T1a, T1b); } { V TF, T10, T1l, T1o; TF = VBYI(VFMA(LDK(KP2_000000000), TE, Tn)); T10 = VADD(TS, TZ); ST(&(xo[WS(os, 1)]), VADD(TF, T10), ovs, &(xo[WS(os, 1)])); ST(&(xo[WS(os, 12)]), VSUB(T10, TF), ovs, &(xo[0])); { V T15, T16, T1p, T1q; T15 = VBYI(VFMS(LDK(KP2_000000000), T13, T14)); T16 = VSUB(TZ, TS); ST(&(xo[WS(os, 5)]), VADD(T15, T16), ovs, &(xo[WS(os, 1)])); ST(&(xo[WS(os, 8)]), VSUB(T16, T15), ovs, &(xo[0])); T1p = VADD(T1n, T1m); T1q = VBYI(VADD(T1j, T1k)); ST(&(xo[WS(os, 4)]), VSUB(T1p, T1q), ovs, &(xo[0])); ST(&(xo[WS(os, 9)]), VADD(T1q, T1p), ovs, &(xo[WS(os, 1)])); } T1l = VBYI(VSUB(T1j, T1k)); T1o = VSUB(T1m, T1n); ST(&(xo[WS(os, 3)]), VADD(T1l, T1o), ovs, &(xo[WS(os, 1)])); ST(&(xo[WS(os, 10)]), VSUB(T1o, T1l), ovs, &(xo[0])); { V T1h, T1i, T1d, T1g; T1h = VBYI(VSUB(T1e, T1f)); T1i = VSUB(T1c, T19); ST(&(xo[WS(os, 6)]), VADD(T1h, T1i), ovs, &(xo[0])); ST(&(xo[WS(os, 7)]), VSUB(T1i, T1h), ovs, &(xo[WS(os, 1)])); T1d = VADD(T19, T1c); T1g = VBYI(VADD(T1e, T1f)); ST(&(xo[WS(os, 2)]), VSUB(T1d, T1g), ovs, &(xo[0])); ST(&(xo[WS(os, 11)]), VADD(T1g, T1d), ovs, &(xo[WS(os, 1)])); } } } } } VLEAVE(); } static const kdft_desc desc = { 13, XSIMD_STRING("n1fv_13"), {69, 15, 19, 0}, &GENUS, 0, 0, 0, 0 }; void XSIMD(codelet_n1fv_13) (planner *p) { X(kdft_register) (p, n1fv_13, &desc); } #endif /* HAVE_FMA */ fftw-3.3.3/dft/simd/common/n1fv_14.c0000644000175000017500000002731612054410346013742 00000000000000/* * Copyright (c) 2003, 2007-11 Matteo Frigo * Copyright (c) 2003, 2007-11 Massachusetts Institute of Technology * * This program is free software; you can redistribute it and/or modify * it under the terms of the GNU General Public License as published by * the Free Software Foundation; either version 2 of the License, or * (at your option) any later version. * * This program is distributed in the hope that it will be useful, * but WITHOUT ANY WARRANTY; without even the implied warranty of * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the * GNU General Public License for more details. * * You should have received a copy of the GNU General Public License * along with this program; if not, write to the Free Software * Foundation, Inc., 51 Franklin Street, Fifth Floor, Boston, MA 02110-1301 USA * */ /* This file was automatically generated --- DO NOT EDIT */ /* Generated on Sun Nov 25 07:36:52 EST 2012 */ #include "codelet-dft.h" #ifdef HAVE_FMA /* Generated by: ../../../genfft/gen_notw_c.native -fma -reorder-insns -schedule-for-pipeline -simd -compact -variables 4 -pipeline-latency 8 -n 14 -name n1fv_14 -include n1f.h */ /* * This function contains 74 FP additions, 48 FP multiplications, * (or, 32 additions, 6 multiplications, 42 fused multiply/add), * 63 stack variables, 6 constants, and 28 memory accesses */ #include "n1f.h" static void n1fv_14(const R *ri, const R *ii, R *ro, R *io, stride is, stride os, INT v, INT ivs, INT ovs) { DVK(KP900968867, +0.900968867902419126236102319507445051165919162); DVK(KP801937735, +0.801937735804838252472204639014890102331838324); DVK(KP974927912, +0.974927912181823607018131682993931217232785801); DVK(KP692021471, +0.692021471630095869627814897002069140197260599); DVK(KP554958132, +0.554958132087371191422194871006410481067288862); DVK(KP356895867, +0.356895867892209443894399510021300583399127187); { INT i; const R *xi; R *xo; xi = ri; xo = ro; for (i = v; i > 0; i = i - VL, xi = xi + (VL * ivs), xo = xo + (VL * ovs), MAKE_VOLATILE_STRIDE(28, is), MAKE_VOLATILE_STRIDE(28, os)) { V TH, T3, TP, Tn, Ta, Ts, TW, TK, TO, Tk, TM, Tg, TL, Td, T1; V T2; T1 = LD(&(xi[0]), ivs, &(xi[0])); T2 = LD(&(xi[WS(is, 7)]), ivs, &(xi[WS(is, 1)])); { V Ti, TI, T6, TJ, T9, Tj, Te, Tf, Tb, Tc; { V T4, T5, T7, T8, Tl, Tm; T4 = LD(&(xi[WS(is, 2)]), ivs, &(xi[0])); T5 = LD(&(xi[WS(is, 9)]), ivs, &(xi[WS(is, 1)])); T7 = LD(&(xi[WS(is, 12)]), ivs, &(xi[0])); T8 = LD(&(xi[WS(is, 5)]), ivs, &(xi[WS(is, 1)])); Tl = LD(&(xi[WS(is, 8)]), ivs, &(xi[0])); Tm = LD(&(xi[WS(is, 1)]), ivs, &(xi[WS(is, 1)])); Ti = LD(&(xi[WS(is, 6)]), ivs, &(xi[0])); TH = VADD(T1, T2); T3 = VSUB(T1, T2); TI = VADD(T4, T5); T6 = VSUB(T4, T5); TJ = VADD(T7, T8); T9 = VSUB(T7, T8); TP = VADD(Tl, Tm); Tn = VSUB(Tl, Tm); Tj = LD(&(xi[WS(is, 13)]), ivs, &(xi[WS(is, 1)])); Te = LD(&(xi[WS(is, 10)]), ivs, &(xi[0])); Tf = LD(&(xi[WS(is, 3)]), ivs, &(xi[WS(is, 1)])); Tb = LD(&(xi[WS(is, 4)]), ivs, &(xi[0])); Tc = LD(&(xi[WS(is, 11)]), ivs, &(xi[WS(is, 1)])); } Ta = VADD(T6, T9); Ts = VSUB(T9, T6); TW = VSUB(TJ, TI); TK = VADD(TI, TJ); TO = VADD(Ti, Tj); Tk = VSUB(Ti, Tj); TM = VADD(Te, Tf); Tg = VSUB(Te, Tf); TL = VADD(Tb, Tc); Td = VSUB(Tb, Tc); } { V T18, TB, T13, TY, TG, Tw, T11, Tr, T16, TT, Tz, TE, TU, TQ; TU = VSUB(TO, TP); TQ = VADD(TO, TP); { V Tt, To, TV, TN; Tt = VSUB(Tn, Tk); To = VADD(Tk, Tn); TV = VSUB(TL, TM); TN = VADD(TL, TM); { V Tu, Th, TZ, T17; Tu = VSUB(Tg, Td); Th = VADD(Td, Tg); TZ = VFNMS(LDK(KP356895867), TK, TQ); T17 = VFNMS(LDK(KP554958132), TU, TW); { V Tp, TA, T14, TR; Tp = VFNMS(LDK(KP356895867), Ta, To); TA = VFMA(LDK(KP554958132), Tt, Ts); ST(&(xo[0]), VADD(TH, VADD(TK, VADD(TN, TQ))), ovs, &(xo[0])); T14 = VFNMS(LDK(KP356895867), TN, TK); TR = VFNMS(LDK(KP356895867), TQ, TN); { V T12, TX, Tx, TC; T12 = VFMA(LDK(KP554958132), TV, TU); TX = VFMA(LDK(KP554958132), TW, TV); ST(&(xo[WS(os, 7)]), VADD(T3, VADD(Ta, VADD(Th, To))), ovs, &(xo[WS(os, 1)])); Tx = VFNMS(LDK(KP356895867), Th, Ta); TC = VFNMS(LDK(KP356895867), To, Th); { V TF, Tv, T10, Tq; TF = VFNMS(LDK(KP554958132), Ts, Tu); Tv = VFMA(LDK(KP554958132), Tu, Tt); T10 = VFNMS(LDK(KP692021471), TZ, TN); T18 = VMUL(LDK(KP974927912), VFNMS(LDK(KP801937735), T17, TV)); Tq = VFNMS(LDK(KP692021471), Tp, Th); TB = VMUL(LDK(KP974927912), VFMA(LDK(KP801937735), TA, Tu)); { V T15, TS, Ty, TD; T15 = VFNMS(LDK(KP692021471), T14, TQ); TS = VFNMS(LDK(KP692021471), TR, TK); T13 = VMUL(LDK(KP974927912), VFMA(LDK(KP801937735), T12, TW)); TY = VMUL(LDK(KP974927912), VFNMS(LDK(KP801937735), TX, TU)); Ty = VFNMS(LDK(KP692021471), Tx, To); TD = VFNMS(LDK(KP692021471), TC, Ta); TG = VMUL(LDK(KP974927912), VFNMS(LDK(KP801937735), TF, Tt)); Tw = VMUL(LDK(KP974927912), VFNMS(LDK(KP801937735), Tv, Ts)); T11 = VFNMS(LDK(KP900968867), T10, TH); Tr = VFNMS(LDK(KP900968867), Tq, T3); T16 = VFNMS(LDK(KP900968867), T15, TH); TT = VFNMS(LDK(KP900968867), TS, TH); Tz = VFNMS(LDK(KP900968867), Ty, T3); TE = VFNMS(LDK(KP900968867), TD, T3); } } } } } } ST(&(xo[WS(os, 12)]), VFNMSI(T13, T11), ovs, &(xo[0])); ST(&(xo[WS(os, 2)]), VFMAI(T13, T11), ovs, &(xo[0])); ST(&(xo[WS(os, 9)]), VFMAI(Tw, Tr), ovs, &(xo[WS(os, 1)])); ST(&(xo[WS(os, 5)]), VFNMSI(Tw, Tr), ovs, &(xo[WS(os, 1)])); ST(&(xo[WS(os, 8)]), VFNMSI(T18, T16), ovs, &(xo[0])); ST(&(xo[WS(os, 6)]), VFMAI(T18, T16), ovs, &(xo[0])); ST(&(xo[WS(os, 10)]), VFNMSI(TY, TT), ovs, &(xo[0])); ST(&(xo[WS(os, 4)]), VFMAI(TY, TT), ovs, &(xo[0])); ST(&(xo[WS(os, 1)]), VFMAI(TB, Tz), ovs, &(xo[WS(os, 1)])); ST(&(xo[WS(os, 13)]), VFNMSI(TB, Tz), ovs, &(xo[WS(os, 1)])); ST(&(xo[WS(os, 3)]), VFMAI(TG, TE), ovs, &(xo[WS(os, 1)])); ST(&(xo[WS(os, 11)]), VFNMSI(TG, TE), ovs, &(xo[WS(os, 1)])); } } } VLEAVE(); } static const kdft_desc desc = { 14, XSIMD_STRING("n1fv_14"), {32, 6, 42, 0}, &GENUS, 0, 0, 0, 0 }; void XSIMD(codelet_n1fv_14) (planner *p) { X(kdft_register) (p, n1fv_14, &desc); } #else /* HAVE_FMA */ /* Generated by: ../../../genfft/gen_notw_c.native -simd -compact -variables 4 -pipeline-latency 8 -n 14 -name n1fv_14 -include n1f.h */ /* * This function contains 74 FP additions, 36 FP multiplications, * (or, 50 additions, 12 multiplications, 24 fused multiply/add), * 33 stack variables, 6 constants, and 28 memory accesses */ #include "n1f.h" static void n1fv_14(const R *ri, const R *ii, R *ro, R *io, stride is, stride os, INT v, INT ivs, INT ovs) { DVK(KP222520933, +0.222520933956314404288902564496794759466355569); DVK(KP900968867, +0.900968867902419126236102319507445051165919162); DVK(KP623489801, +0.623489801858733530525004884004239810632274731); DVK(KP433883739, +0.433883739117558120475768332848358754609990728); DVK(KP781831482, +0.781831482468029808708444526674057750232334519); DVK(KP974927912, +0.974927912181823607018131682993931217232785801); { INT i; const R *xi; R *xo; xi = ri; xo = ro; for (i = v; i > 0; i = i - VL, xi = xi + (VL * ivs), xo = xo + (VL * ovs), MAKE_VOLATILE_STRIDE(28, is), MAKE_VOLATILE_STRIDE(28, os)) { V T3, Ty, To, TK, Tr, TE, Ta, TJ, Tq, TB, Th, TL, Ts, TH, T1; V T2; T1 = LD(&(xi[0]), ivs, &(xi[0])); T2 = LD(&(xi[WS(is, 7)]), ivs, &(xi[WS(is, 1)])); T3 = VSUB(T1, T2); Ty = VADD(T1, T2); { V Tk, TC, Tn, TD; { V Ti, Tj, Tl, Tm; Ti = LD(&(xi[WS(is, 6)]), ivs, &(xi[0])); Tj = LD(&(xi[WS(is, 13)]), ivs, &(xi[WS(is, 1)])); Tk = VSUB(Ti, Tj); TC = VADD(Ti, Tj); Tl = LD(&(xi[WS(is, 8)]), ivs, &(xi[0])); Tm = LD(&(xi[WS(is, 1)]), ivs, &(xi[WS(is, 1)])); Tn = VSUB(Tl, Tm); TD = VADD(Tl, Tm); } To = VADD(Tk, Tn); TK = VSUB(TC, TD); Tr = VSUB(Tn, Tk); TE = VADD(TC, TD); } { V T6, Tz, T9, TA; { V T4, T5, T7, T8; T4 = LD(&(xi[WS(is, 2)]), ivs, &(xi[0])); T5 = LD(&(xi[WS(is, 9)]), ivs, &(xi[WS(is, 1)])); T6 = VSUB(T4, T5); Tz = VADD(T4, T5); T7 = LD(&(xi[WS(is, 12)]), ivs, &(xi[0])); T8 = LD(&(xi[WS(is, 5)]), ivs, &(xi[WS(is, 1)])); T9 = VSUB(T7, T8); TA = VADD(T7, T8); } Ta = VADD(T6, T9); TJ = VSUB(TA, Tz); Tq = VSUB(T9, T6); TB = VADD(Tz, TA); } { V Td, TF, Tg, TG; { V Tb, Tc, Te, Tf; Tb = LD(&(xi[WS(is, 4)]), ivs, &(xi[0])); Tc = LD(&(xi[WS(is, 11)]), ivs, &(xi[WS(is, 1)])); Td = VSUB(Tb, Tc); TF = VADD(Tb, Tc); Te = LD(&(xi[WS(is, 10)]), ivs, &(xi[0])); Tf = LD(&(xi[WS(is, 3)]), ivs, &(xi[WS(is, 1)])); Tg = VSUB(Te, Tf); TG = VADD(Te, Tf); } Th = VADD(Td, Tg); TL = VSUB(TF, TG); Ts = VSUB(Tg, Td); TH = VADD(TF, TG); } ST(&(xo[WS(os, 7)]), VADD(T3, VADD(Ta, VADD(Th, To))), ovs, &(xo[WS(os, 1)])); ST(&(xo[0]), VADD(Ty, VADD(TB, VADD(TH, TE))), ovs, &(xo[0])); { V Tt, Tp, TP, TQ; Tt = VBYI(VFNMS(LDK(KP781831482), Tr, VFNMS(LDK(KP433883739), Ts, VMUL(LDK(KP974927912), Tq)))); Tp = VFMA(LDK(KP623489801), To, VFNMS(LDK(KP900968867), Th, VFNMS(LDK(KP222520933), Ta, T3))); ST(&(xo[WS(os, 5)]), VSUB(Tp, Tt), ovs, &(xo[WS(os, 1)])); ST(&(xo[WS(os, 9)]), VADD(Tp, Tt), ovs, &(xo[WS(os, 1)])); TP = VBYI(VFMA(LDK(KP974927912), TJ, VFMA(LDK(KP433883739), TL, VMUL(LDK(KP781831482), TK)))); TQ = VFMA(LDK(KP623489801), TE, VFNMS(LDK(KP900968867), TH, VFNMS(LDK(KP222520933), TB, Ty))); ST(&(xo[WS(os, 2)]), VADD(TP, TQ), ovs, &(xo[0])); ST(&(xo[WS(os, 12)]), VSUB(TQ, TP), ovs, &(xo[0])); } { V Tv, Tu, TM, TI; Tv = VBYI(VFMA(LDK(KP781831482), Tq, VFMA(LDK(KP974927912), Ts, VMUL(LDK(KP433883739), Tr)))); Tu = VFMA(LDK(KP623489801), Ta, VFNMS(LDK(KP900968867), To, VFNMS(LDK(KP222520933), Th, T3))); ST(&(xo[WS(os, 13)]), VSUB(Tu, Tv), ovs, &(xo[WS(os, 1)])); ST(&(xo[WS(os, 1)]), VADD(Tu, Tv), ovs, &(xo[WS(os, 1)])); TM = VBYI(VFNMS(LDK(KP433883739), TK, VFNMS(LDK(KP974927912), TL, VMUL(LDK(KP781831482), TJ)))); TI = VFMA(LDK(KP623489801), TB, VFNMS(LDK(KP900968867), TE, VFNMS(LDK(KP222520933), TH, Ty))); ST(&(xo[WS(os, 6)]), VSUB(TI, TM), ovs, &(xo[0])); ST(&(xo[WS(os, 8)]), VADD(TM, TI), ovs, &(xo[0])); } { V TO, TN, Tx, Tw; TO = VBYI(VFMA(LDK(KP433883739), TJ, VFNMS(LDK(KP974927912), TK, VMUL(LDK(KP781831482), TL)))); TN = VFMA(LDK(KP623489801), TH, VFNMS(LDK(KP222520933), TE, VFNMS(LDK(KP900968867), TB, Ty))); ST(&(xo[WS(os, 4)]), VSUB(TN, TO), ovs, &(xo[0])); ST(&(xo[WS(os, 10)]), VADD(TO, TN), ovs, &(xo[0])); Tx = VBYI(VFMA(LDK(KP433883739), Tq, VFNMS(LDK(KP781831482), Ts, VMUL(LDK(KP974927912), Tr)))); Tw = VFMA(LDK(KP623489801), Th, VFNMS(LDK(KP222520933), To, VFNMS(LDK(KP900968867), Ta, T3))); ST(&(xo[WS(os, 11)]), VSUB(Tw, Tx), ovs, &(xo[WS(os, 1)])); ST(&(xo[WS(os, 3)]), VADD(Tw, Tx), ovs, &(xo[WS(os, 1)])); } } } VLEAVE(); } static const kdft_desc desc = { 14, XSIMD_STRING("n1fv_14"), {50, 12, 24, 0}, &GENUS, 0, 0, 0, 0 }; void XSIMD(codelet_n1fv_14) (planner *p) { X(kdft_register) (p, n1fv_14, &desc); } #endif /* HAVE_FMA */ fftw-3.3.3/dft/simd/common/n1fv_15.c0000644000175000017500000003155412054410345013741 00000000000000/* * Copyright (c) 2003, 2007-11 Matteo Frigo * Copyright (c) 2003, 2007-11 Massachusetts Institute of Technology * * This program is free software; you can redistribute it and/or modify * it under the terms of the GNU General Public License as published by * the Free Software Foundation; either version 2 of the License, or * (at your option) any later version. * * This program is distributed in the hope that it will be useful, * but WITHOUT ANY WARRANTY; without even the implied warranty of * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the * GNU General Public License for more details. * * You should have received a copy of the GNU General Public License * along with this program; if not, write to the Free Software * Foundation, Inc., 51 Franklin Street, Fifth Floor, Boston, MA 02110-1301 USA * */ /* This file was automatically generated --- DO NOT EDIT */ /* Generated on Sun Nov 25 07:36:52 EST 2012 */ #include "codelet-dft.h" #ifdef HAVE_FMA /* Generated by: ../../../genfft/gen_notw_c.native -fma -reorder-insns -schedule-for-pipeline -simd -compact -variables 4 -pipeline-latency 8 -n 15 -name n1fv_15 -include n1f.h */ /* * This function contains 78 FP additions, 49 FP multiplications, * (or, 36 additions, 7 multiplications, 42 fused multiply/add), * 78 stack variables, 8 constants, and 30 memory accesses */ #include "n1f.h" static void n1fv_15(const R *ri, const R *ii, R *ro, R *io, stride is, stride os, INT v, INT ivs, INT ovs) { DVK(KP823639103, +0.823639103546331925877420039278190003029660514); DVK(KP910592997, +0.910592997310029334643087372129977886038870291); DVK(KP559016994, +0.559016994374947424102293417182819058860154590); DVK(KP951056516, +0.951056516295153572116439333379382143405698634); DVK(KP866025403, +0.866025403784438646763723170752936183471402627); DVK(KP250000000, +0.250000000000000000000000000000000000000000000); DVK(KP618033988, +0.618033988749894848204586834365638117720309180); DVK(KP500000000, +0.500000000000000000000000000000000000000000000); { INT i; const R *xi; R *xo; xi = ri; xo = ro; for (i = v; i > 0; i = i - VL, xi = xi + (VL * ivs), xo = xo + (VL * ovs), MAKE_VOLATILE_STRIDE(30, is), MAKE_VOLATILE_STRIDE(30, os)) { V Tb, TX, TM, TQ, Th, TB, T5, Ti, Ta, TC, TN, Te, TG, Tq, Tj; V T1, T2, T3; T1 = LD(&(xi[0]), ivs, &(xi[0])); T2 = LD(&(xi[WS(is, 5)]), ivs, &(xi[WS(is, 1)])); T3 = LD(&(xi[WS(is, 10)]), ivs, &(xi[0])); { V T6, T7, T8, Tm, Tn, To; T6 = LD(&(xi[WS(is, 3)]), ivs, &(xi[WS(is, 1)])); T7 = LD(&(xi[WS(is, 8)]), ivs, &(xi[0])); T8 = LD(&(xi[WS(is, 13)]), ivs, &(xi[WS(is, 1)])); Tm = LD(&(xi[WS(is, 9)]), ivs, &(xi[WS(is, 1)])); Tn = LD(&(xi[WS(is, 14)]), ivs, &(xi[0])); To = LD(&(xi[WS(is, 4)]), ivs, &(xi[0])); { V T4, Tc, T9, Td, Tp; Tb = LD(&(xi[WS(is, 12)]), ivs, &(xi[0])); T4 = VADD(T2, T3); TX = VSUB(T3, T2); Tc = LD(&(xi[WS(is, 2)]), ivs, &(xi[0])); TM = VSUB(T8, T7); T9 = VADD(T7, T8); Td = LD(&(xi[WS(is, 7)]), ivs, &(xi[WS(is, 1)])); Tp = VADD(Tn, To); TQ = VSUB(To, Tn); Th = LD(&(xi[WS(is, 6)]), ivs, &(xi[0])); TB = VFNMS(LDK(KP500000000), T4, T1); T5 = VADD(T1, T4); Ti = LD(&(xi[WS(is, 11)]), ivs, &(xi[WS(is, 1)])); Ta = VADD(T6, T9); TC = VFNMS(LDK(KP500000000), T9, T6); TN = VSUB(Td, Tc); Te = VADD(Tc, Td); TG = VFNMS(LDK(KP500000000), Tp, Tm); Tq = VADD(Tm, Tp); Tj = LD(&(xi[WS(is, 1)]), ivs, &(xi[WS(is, 1)])); } } { V TY, TO, Tf, TD, TP, Tk; TY = VADD(TM, TN); TO = VSUB(TM, TN); Tf = VADD(Tb, Te); TD = VFNMS(LDK(KP500000000), Te, Tb); TP = VSUB(Tj, Ti); Tk = VADD(Ti, Tj); { V Tx, Tg, TE, TU, TZ, TR, Tl, TF; Tx = VSUB(Ta, Tf); Tg = VADD(Ta, Tf); TE = VADD(TC, TD); TU = VSUB(TC, TD); TZ = VADD(TP, TQ); TR = VSUB(TP, TQ); Tl = VADD(Th, Tk); TF = VFNMS(LDK(KP500000000), Tk, Th); { V T12, T10, T18, TS, Tw, Tr, TH, TV, T11, T1g; T12 = VSUB(TY, TZ); T10 = VADD(TY, TZ); T18 = VFNMS(LDK(KP618033988), TO, TR); TS = VFMA(LDK(KP618033988), TR, TO); Tw = VSUB(Tl, Tq); Tr = VADD(Tl, Tq); TH = VADD(TF, TG); TV = VSUB(TF, TG); T11 = VFNMS(LDK(KP250000000), T10, TX); T1g = VMUL(LDK(KP866025403), VADD(TX, T10)); { V TA, Ty, Tu, TK, TI, T1a, TW, T1b, T13, Tt, Ts, TJ, T1f; TA = VMUL(LDK(KP951056516), VFMA(LDK(KP618033988), Tw, Tx)); Ty = VMUL(LDK(KP951056516), VFNMS(LDK(KP618033988), Tx, Tw)); Ts = VADD(Tg, Tr); Tu = VSUB(Tg, Tr); TK = VSUB(TE, TH); TI = VADD(TE, TH); T1a = VFNMS(LDK(KP618033988), TU, TV); TW = VFMA(LDK(KP618033988), TV, TU); T1b = VFNMS(LDK(KP559016994), T12, T11); T13 = VFMA(LDK(KP559016994), T12, T11); ST(&(xo[0]), VADD(T5, Ts), ovs, &(xo[0])); Tt = VFNMS(LDK(KP250000000), Ts, T5); TJ = VFNMS(LDK(KP250000000), TI, TB); T1f = VADD(TB, TI); { V T1c, T1e, T16, T14, Tv, Tz, T17, TL; T1c = VMUL(LDK(KP951056516), VFNMS(LDK(KP910592997), T1b, T1a)); T1e = VMUL(LDK(KP951056516), VFMA(LDK(KP910592997), T1b, T1a)); T16 = VMUL(LDK(KP951056516), VFMA(LDK(KP910592997), T13, TW)); T14 = VMUL(LDK(KP951056516), VFNMS(LDK(KP910592997), T13, TW)); Tv = VFNMS(LDK(KP559016994), Tu, Tt); Tz = VFMA(LDK(KP559016994), Tu, Tt); T17 = VFNMS(LDK(KP559016994), TK, TJ); TL = VFMA(LDK(KP559016994), TK, TJ); ST(&(xo[WS(os, 10)]), VFMAI(T1g, T1f), ovs, &(xo[0])); ST(&(xo[WS(os, 5)]), VFNMSI(T1g, T1f), ovs, &(xo[WS(os, 1)])); { V T19, T1d, T15, TT; ST(&(xo[WS(os, 12)]), VFMAI(Ty, Tv), ovs, &(xo[0])); ST(&(xo[WS(os, 3)]), VFNMSI(Ty, Tv), ovs, &(xo[WS(os, 1)])); ST(&(xo[WS(os, 9)]), VFMAI(TA, Tz), ovs, &(xo[WS(os, 1)])); ST(&(xo[WS(os, 6)]), VFNMSI(TA, Tz), ovs, &(xo[0])); T19 = VFMA(LDK(KP823639103), T18, T17); T1d = VFNMS(LDK(KP823639103), T18, T17); T15 = VFNMS(LDK(KP823639103), TS, TL); TT = VFMA(LDK(KP823639103), TS, TL); ST(&(xo[WS(os, 2)]), VFMAI(T1c, T19), ovs, &(xo[0])); ST(&(xo[WS(os, 13)]), VFNMSI(T1c, T19), ovs, &(xo[WS(os, 1)])); ST(&(xo[WS(os, 7)]), VFMAI(T1e, T1d), ovs, &(xo[WS(os, 1)])); ST(&(xo[WS(os, 8)]), VFNMSI(T1e, T1d), ovs, &(xo[0])); ST(&(xo[WS(os, 4)]), VFMAI(T16, T15), ovs, &(xo[0])); ST(&(xo[WS(os, 11)]), VFNMSI(T16, T15), ovs, &(xo[WS(os, 1)])); ST(&(xo[WS(os, 14)]), VFMAI(T14, TT), ovs, &(xo[0])); ST(&(xo[WS(os, 1)]), VFNMSI(T14, TT), ovs, &(xo[WS(os, 1)])); } } } } } } } } VLEAVE(); } static const kdft_desc desc = { 15, XSIMD_STRING("n1fv_15"), {36, 7, 42, 0}, &GENUS, 0, 0, 0, 0 }; void XSIMD(codelet_n1fv_15) (planner *p) { X(kdft_register) (p, n1fv_15, &desc); } #else /* HAVE_FMA */ /* Generated by: ../../../genfft/gen_notw_c.native -simd -compact -variables 4 -pipeline-latency 8 -n 15 -name n1fv_15 -include n1f.h */ /* * This function contains 78 FP additions, 25 FP multiplications, * (or, 64 additions, 11 multiplications, 14 fused multiply/add), * 55 stack variables, 10 constants, and 30 memory accesses */ #include "n1f.h" static void n1fv_15(const R *ri, const R *ii, R *ro, R *io, stride is, stride os, INT v, INT ivs, INT ovs) { DVK(KP216506350, +0.216506350946109661690930792688234045867850657); DVK(KP509036960, +0.509036960455127183450980863393907648510733164); DVK(KP823639103, +0.823639103546331925877420039278190003029660514); DVK(KP587785252, +0.587785252292473129168705954639072768597652438); DVK(KP951056516, +0.951056516295153572116439333379382143405698634); DVK(KP250000000, +0.250000000000000000000000000000000000000000000); DVK(KP559016994, +0.559016994374947424102293417182819058860154590); DVK(KP866025403, +0.866025403784438646763723170752936183471402627); DVK(KP484122918, +0.484122918275927110647408174972799951354115213); DVK(KP500000000, +0.500000000000000000000000000000000000000000000); { INT i; const R *xi; R *xo; xi = ri; xo = ro; for (i = v; i > 0; i = i - VL, xi = xi + (VL * ivs), xo = xo + (VL * ovs), MAKE_VOLATILE_STRIDE(30, is), MAKE_VOLATILE_STRIDE(30, os)) { V T5, T10, TB, TO, TU, TV, TR, Ta, Tf, Tg, Tl, Tq, Tr, TE, TH; V TI, TZ, T11, T1f, T1g; { V T1, T2, T3, T4; T1 = LD(&(xi[0]), ivs, &(xi[0])); T2 = LD(&(xi[WS(is, 5)]), ivs, &(xi[WS(is, 1)])); T3 = LD(&(xi[WS(is, 10)]), ivs, &(xi[0])); T4 = VADD(T2, T3); T5 = VADD(T1, T4); T10 = VSUB(T3, T2); TB = VFNMS(LDK(KP500000000), T4, T1); } { V T6, T9, TC, TP, Tm, Tp, TG, TN, Tb, Te, TD, TQ, Th, Tk, TF; V TM, TX, TY; { V T7, T8, Tn, To; T6 = LD(&(xi[WS(is, 3)]), ivs, &(xi[WS(is, 1)])); T7 = LD(&(xi[WS(is, 8)]), ivs, &(xi[0])); T8 = LD(&(xi[WS(is, 13)]), ivs, &(xi[WS(is, 1)])); T9 = VADD(T7, T8); TC = VFNMS(LDK(KP500000000), T9, T6); TP = VSUB(T8, T7); Tm = LD(&(xi[WS(is, 9)]), ivs, &(xi[WS(is, 1)])); Tn = LD(&(xi[WS(is, 14)]), ivs, &(xi[0])); To = LD(&(xi[WS(is, 4)]), ivs, &(xi[0])); Tp = VADD(Tn, To); TG = VFNMS(LDK(KP500000000), Tp, Tm); TN = VSUB(To, Tn); } { V Tc, Td, Ti, Tj; Tb = LD(&(xi[WS(is, 12)]), ivs, &(xi[0])); Tc = LD(&(xi[WS(is, 2)]), ivs, &(xi[0])); Td = LD(&(xi[WS(is, 7)]), ivs, &(xi[WS(is, 1)])); Te = VADD(Tc, Td); TD = VFNMS(LDK(KP500000000), Te, Tb); TQ = VSUB(Td, Tc); Th = LD(&(xi[WS(is, 6)]), ivs, &(xi[0])); Ti = LD(&(xi[WS(is, 11)]), ivs, &(xi[WS(is, 1)])); Tj = LD(&(xi[WS(is, 1)]), ivs, &(xi[WS(is, 1)])); Tk = VADD(Ti, Tj); TF = VFNMS(LDK(KP500000000), Tk, Th); TM = VSUB(Tj, Ti); } TO = VSUB(TM, TN); TU = VSUB(TF, TG); TV = VSUB(TC, TD); TR = VSUB(TP, TQ); Ta = VADD(T6, T9); Tf = VADD(Tb, Te); Tg = VADD(Ta, Tf); Tl = VADD(Th, Tk); Tq = VADD(Tm, Tp); Tr = VADD(Tl, Tq); TE = VADD(TC, TD); TH = VADD(TF, TG); TI = VADD(TE, TH); TX = VADD(TP, TQ); TY = VADD(TM, TN); TZ = VMUL(LDK(KP484122918), VSUB(TX, TY)); T11 = VADD(TX, TY); } T1f = VADD(TB, TI); T1g = VBYI(VMUL(LDK(KP866025403), VADD(T10, T11))); ST(&(xo[WS(os, 5)]), VSUB(T1f, T1g), ovs, &(xo[WS(os, 1)])); ST(&(xo[WS(os, 10)]), VADD(T1f, T1g), ovs, &(xo[0])); { V Tu, Ts, Tt, Ty, TA, Tw, Tx, Tz, Tv; Tu = VMUL(LDK(KP559016994), VSUB(Tg, Tr)); Ts = VADD(Tg, Tr); Tt = VFNMS(LDK(KP250000000), Ts, T5); Tw = VSUB(Tl, Tq); Tx = VSUB(Ta, Tf); Ty = VBYI(VFNMS(LDK(KP587785252), Tx, VMUL(LDK(KP951056516), Tw))); TA = VBYI(VFMA(LDK(KP951056516), Tx, VMUL(LDK(KP587785252), Tw))); ST(&(xo[0]), VADD(T5, Ts), ovs, &(xo[0])); Tz = VADD(Tu, Tt); ST(&(xo[WS(os, 6)]), VSUB(Tz, TA), ovs, &(xo[0])); ST(&(xo[WS(os, 9)]), VADD(TA, Tz), ovs, &(xo[WS(os, 1)])); Tv = VSUB(Tt, Tu); ST(&(xo[WS(os, 3)]), VSUB(Tv, Ty), ovs, &(xo[WS(os, 1)])); ST(&(xo[WS(os, 12)]), VADD(Ty, Tv), ovs, &(xo[0])); } { V TS, TW, T1b, T18, T13, T1a, TL, T17, T12, TJ, TK; TS = VFNMS(LDK(KP509036960), TR, VMUL(LDK(KP823639103), TO)); TW = VFNMS(LDK(KP587785252), TV, VMUL(LDK(KP951056516), TU)); T1b = VFMA(LDK(KP951056516), TV, VMUL(LDK(KP587785252), TU)); T18 = VFMA(LDK(KP823639103), TR, VMUL(LDK(KP509036960), TO)); T12 = VFNMS(LDK(KP216506350), T11, VMUL(LDK(KP866025403), T10)); T13 = VSUB(TZ, T12); T1a = VADD(TZ, T12); TJ = VFNMS(LDK(KP250000000), TI, TB); TK = VMUL(LDK(KP559016994), VSUB(TE, TH)); TL = VSUB(TJ, TK); T17 = VADD(TK, TJ); { V TT, T14, T1d, T1e; TT = VSUB(TL, TS); T14 = VBYI(VSUB(TW, T13)); ST(&(xo[WS(os, 8)]), VSUB(TT, T14), ovs, &(xo[0])); ST(&(xo[WS(os, 7)]), VADD(TT, T14), ovs, &(xo[WS(os, 1)])); T1d = VSUB(T17, T18); T1e = VBYI(VADD(T1b, T1a)); ST(&(xo[WS(os, 11)]), VSUB(T1d, T1e), ovs, &(xo[WS(os, 1)])); ST(&(xo[WS(os, 4)]), VADD(T1d, T1e), ovs, &(xo[0])); } { V T15, T16, T19, T1c; T15 = VADD(TL, TS); T16 = VBYI(VADD(TW, T13)); ST(&(xo[WS(os, 13)]), VSUB(T15, T16), ovs, &(xo[WS(os, 1)])); ST(&(xo[WS(os, 2)]), VADD(T15, T16), ovs, &(xo[0])); T19 = VADD(T17, T18); T1c = VBYI(VSUB(T1a, T1b)); ST(&(xo[WS(os, 14)]), VSUB(T19, T1c), ovs, &(xo[0])); ST(&(xo[WS(os, 1)]), VADD(T19, T1c), ovs, &(xo[WS(os, 1)])); } } } } VLEAVE(); } static const kdft_desc desc = { 15, XSIMD_STRING("n1fv_15"), {64, 11, 14, 0}, &GENUS, 0, 0, 0, 0 }; void XSIMD(codelet_n1fv_15) (planner *p) { X(kdft_register) (p, n1fv_15, &desc); } #endif /* HAVE_FMA */ fftw-3.3.3/dft/simd/common/n1fv_16.c0000644000175000017500000002671112054410346013742 00000000000000/* * Copyright (c) 2003, 2007-11 Matteo Frigo * Copyright (c) 2003, 2007-11 Massachusetts Institute of Technology * * This program is free software; you can redistribute it and/or modify * it under the terms of the GNU General Public License as published by * the Free Software Foundation; either version 2 of the License, or * (at your option) any later version. * * This program is distributed in the hope that it will be useful, * but WITHOUT ANY WARRANTY; without even the implied warranty of * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the * GNU General Public License for more details. * * You should have received a copy of the GNU General Public License * along with this program; if not, write to the Free Software * Foundation, Inc., 51 Franklin Street, Fifth Floor, Boston, MA 02110-1301 USA * */ /* This file was automatically generated --- DO NOT EDIT */ /* Generated on Sun Nov 25 07:36:53 EST 2012 */ #include "codelet-dft.h" #ifdef HAVE_FMA /* Generated by: ../../../genfft/gen_notw_c.native -fma -reorder-insns -schedule-for-pipeline -simd -compact -variables 4 -pipeline-latency 8 -n 16 -name n1fv_16 -include n1f.h */ /* * This function contains 72 FP additions, 34 FP multiplications, * (or, 38 additions, 0 multiplications, 34 fused multiply/add), * 54 stack variables, 3 constants, and 32 memory accesses */ #include "n1f.h" static void n1fv_16(const R *ri, const R *ii, R *ro, R *io, stride is, stride os, INT v, INT ivs, INT ovs) { DVK(KP923879532, +0.923879532511286756128183189396788286822416626); DVK(KP414213562, +0.414213562373095048801688724209698078569671875); DVK(KP707106781, +0.707106781186547524400844362104849039284835938); { INT i; const R *xi; R *xo; xi = ri; xo = ro; for (i = v; i > 0; i = i - VL, xi = xi + (VL * ivs), xo = xo + (VL * ovs), MAKE_VOLATILE_STRIDE(32, is), MAKE_VOLATILE_STRIDE(32, os)) { V T7, Tu, TF, TB, T13, TL, TO, TX, TC, Te, TP, Th, TQ, Tk, TW; V T16; { V TH, TU, Tz, Tf, TK, TV, TA, TM, Ta, TN, Td, Tg, Ti, Tj; { V T1, T2, T4, T5, To, Tp, Tr, Ts; T1 = LD(&(xi[0]), ivs, &(xi[0])); T2 = LD(&(xi[WS(is, 8)]), ivs, &(xi[0])); T4 = LD(&(xi[WS(is, 4)]), ivs, &(xi[0])); T5 = LD(&(xi[WS(is, 12)]), ivs, &(xi[0])); To = LD(&(xi[WS(is, 14)]), ivs, &(xi[0])); Tp = LD(&(xi[WS(is, 6)]), ivs, &(xi[0])); Tr = LD(&(xi[WS(is, 2)]), ivs, &(xi[0])); Ts = LD(&(xi[WS(is, 10)]), ivs, &(xi[0])); { V T8, TJ, Tq, TI, Tt, T9, Tb, Tc, T3, T6; T8 = LD(&(xi[WS(is, 1)]), ivs, &(xi[WS(is, 1)])); TH = VSUB(T1, T2); T3 = VADD(T1, T2); TU = VSUB(T4, T5); T6 = VADD(T4, T5); TJ = VSUB(To, Tp); Tq = VADD(To, Tp); TI = VSUB(Tr, Ts); Tt = VADD(Tr, Ts); T9 = LD(&(xi[WS(is, 9)]), ivs, &(xi[WS(is, 1)])); Tb = LD(&(xi[WS(is, 5)]), ivs, &(xi[WS(is, 1)])); Tc = LD(&(xi[WS(is, 13)]), ivs, &(xi[WS(is, 1)])); T7 = VSUB(T3, T6); Tz = VADD(T3, T6); Tf = LD(&(xi[WS(is, 15)]), ivs, &(xi[WS(is, 1)])); TK = VADD(TI, TJ); TV = VSUB(TJ, TI); TA = VADD(Tt, Tq); Tu = VSUB(Tq, Tt); TM = VSUB(T8, T9); Ta = VADD(T8, T9); TN = VSUB(Tb, Tc); Td = VADD(Tb, Tc); Tg = LD(&(xi[WS(is, 7)]), ivs, &(xi[WS(is, 1)])); Ti = LD(&(xi[WS(is, 3)]), ivs, &(xi[WS(is, 1)])); Tj = LD(&(xi[WS(is, 11)]), ivs, &(xi[WS(is, 1)])); } } TF = VSUB(Tz, TA); TB = VADD(Tz, TA); T13 = VFNMS(LDK(KP707106781), TK, TH); TL = VFMA(LDK(KP707106781), TK, TH); TO = VFNMS(LDK(KP414213562), TN, TM); TX = VFMA(LDK(KP414213562), TM, TN); TC = VADD(Ta, Td); Te = VSUB(Ta, Td); TP = VSUB(Tf, Tg); Th = VADD(Tf, Tg); TQ = VSUB(Tj, Ti); Tk = VADD(Ti, Tj); TW = VFNMS(LDK(KP707106781), TV, TU); T16 = VFMA(LDK(KP707106781), TV, TU); } { V TY, TR, Tl, TD; TY = VFMA(LDK(KP414213562), TP, TQ); TR = VFNMS(LDK(KP414213562), TQ, TP); Tl = VSUB(Th, Tk); TD = VADD(Th, Tk); { V TS, T17, TZ, T14; TS = VADD(TO, TR); T17 = VSUB(TR, TO); TZ = VSUB(TX, TY); T14 = VADD(TX, TY); { V TE, TG, Tm, Tv; TE = VADD(TC, TD); TG = VSUB(TD, TC); Tm = VADD(Te, Tl); Tv = VSUB(Tl, Te); { V T18, T1a, TT, T11; T18 = VFNMS(LDK(KP923879532), T17, T16); T1a = VFMA(LDK(KP923879532), T17, T16); TT = VFNMS(LDK(KP923879532), TS, TL); T11 = VFMA(LDK(KP923879532), TS, TL); { V T15, T19, T10, T12; T15 = VFNMS(LDK(KP923879532), T14, T13); T19 = VFMA(LDK(KP923879532), T14, T13); T10 = VFNMS(LDK(KP923879532), TZ, TW); T12 = VFMA(LDK(KP923879532), TZ, TW); ST(&(xo[WS(os, 4)]), VFMAI(TG, TF), ovs, &(xo[0])); ST(&(xo[WS(os, 12)]), VFNMSI(TG, TF), ovs, &(xo[0])); ST(&(xo[0]), VADD(TB, TE), ovs, &(xo[0])); ST(&(xo[WS(os, 8)]), VSUB(TB, TE), ovs, &(xo[0])); { V Tw, Ty, Tn, Tx; Tw = VFNMS(LDK(KP707106781), Tv, Tu); Ty = VFMA(LDK(KP707106781), Tv, Tu); Tn = VFNMS(LDK(KP707106781), Tm, T7); Tx = VFMA(LDK(KP707106781), Tm, T7); ST(&(xo[WS(os, 3)]), VFMAI(T1a, T19), ovs, &(xo[WS(os, 1)])); ST(&(xo[WS(os, 13)]), VFNMSI(T1a, T19), ovs, &(xo[WS(os, 1)])); ST(&(xo[WS(os, 11)]), VFMAI(T18, T15), ovs, &(xo[WS(os, 1)])); ST(&(xo[WS(os, 5)]), VFNMSI(T18, T15), ovs, &(xo[WS(os, 1)])); ST(&(xo[WS(os, 1)]), VFNMSI(T12, T11), ovs, &(xo[WS(os, 1)])); ST(&(xo[WS(os, 15)]), VFMAI(T12, T11), ovs, &(xo[WS(os, 1)])); ST(&(xo[WS(os, 7)]), VFMAI(T10, TT), ovs, &(xo[WS(os, 1)])); ST(&(xo[WS(os, 9)]), VFNMSI(T10, TT), ovs, &(xo[WS(os, 1)])); ST(&(xo[WS(os, 14)]), VFNMSI(Ty, Tx), ovs, &(xo[0])); ST(&(xo[WS(os, 2)]), VFMAI(Ty, Tx), ovs, &(xo[0])); ST(&(xo[WS(os, 10)]), VFMAI(Tw, Tn), ovs, &(xo[0])); ST(&(xo[WS(os, 6)]), VFNMSI(Tw, Tn), ovs, &(xo[0])); } } } } } } } } VLEAVE(); } static const kdft_desc desc = { 16, XSIMD_STRING("n1fv_16"), {38, 0, 34, 0}, &GENUS, 0, 0, 0, 0 }; void XSIMD(codelet_n1fv_16) (planner *p) { X(kdft_register) (p, n1fv_16, &desc); } #else /* HAVE_FMA */ /* Generated by: ../../../genfft/gen_notw_c.native -simd -compact -variables 4 -pipeline-latency 8 -n 16 -name n1fv_16 -include n1f.h */ /* * This function contains 72 FP additions, 12 FP multiplications, * (or, 68 additions, 8 multiplications, 4 fused multiply/add), * 30 stack variables, 3 constants, and 32 memory accesses */ #include "n1f.h" static void n1fv_16(const R *ri, const R *ii, R *ro, R *io, stride is, stride os, INT v, INT ivs, INT ovs) { DVK(KP923879532, +0.923879532511286756128183189396788286822416626); DVK(KP382683432, +0.382683432365089771728459984030398866761344562); DVK(KP707106781, +0.707106781186547524400844362104849039284835938); { INT i; const R *xi; R *xo; xi = ri; xo = ro; for (i = v; i > 0; i = i - VL, xi = xi + (VL * ivs), xo = xo + (VL * ovs), MAKE_VOLATILE_STRIDE(32, is), MAKE_VOLATILE_STRIDE(32, os)) { V Tp, T13, Tu, TN, Tm, T14, Tv, TY, T7, T17, Ty, TT, Te, T16, Tx; V TQ; { V Tn, To, TM, Ts, Tt, TL; Tn = LD(&(xi[WS(is, 4)]), ivs, &(xi[0])); To = LD(&(xi[WS(is, 12)]), ivs, &(xi[0])); TM = VADD(Tn, To); Ts = LD(&(xi[0]), ivs, &(xi[0])); Tt = LD(&(xi[WS(is, 8)]), ivs, &(xi[0])); TL = VADD(Ts, Tt); Tp = VSUB(Tn, To); T13 = VADD(TL, TM); Tu = VSUB(Ts, Tt); TN = VSUB(TL, TM); } { V Ti, TW, Tl, TX; { V Tg, Th, Tj, Tk; Tg = LD(&(xi[WS(is, 14)]), ivs, &(xi[0])); Th = LD(&(xi[WS(is, 6)]), ivs, &(xi[0])); Ti = VSUB(Tg, Th); TW = VADD(Tg, Th); Tj = LD(&(xi[WS(is, 2)]), ivs, &(xi[0])); Tk = LD(&(xi[WS(is, 10)]), ivs, &(xi[0])); Tl = VSUB(Tj, Tk); TX = VADD(Tj, Tk); } Tm = VMUL(LDK(KP707106781), VSUB(Ti, Tl)); T14 = VADD(TX, TW); Tv = VMUL(LDK(KP707106781), VADD(Tl, Ti)); TY = VSUB(TW, TX); } { V T3, TR, T6, TS; { V T1, T2, T4, T5; T1 = LD(&(xi[WS(is, 15)]), ivs, &(xi[WS(is, 1)])); T2 = LD(&(xi[WS(is, 7)]), ivs, &(xi[WS(is, 1)])); T3 = VSUB(T1, T2); TR = VADD(T1, T2); T4 = LD(&(xi[WS(is, 3)]), ivs, &(xi[WS(is, 1)])); T5 = LD(&(xi[WS(is, 11)]), ivs, &(xi[WS(is, 1)])); T6 = VSUB(T4, T5); TS = VADD(T4, T5); } T7 = VFNMS(LDK(KP923879532), T6, VMUL(LDK(KP382683432), T3)); T17 = VADD(TR, TS); Ty = VFMA(LDK(KP923879532), T3, VMUL(LDK(KP382683432), T6)); TT = VSUB(TR, TS); } { V Ta, TO, Td, TP; { V T8, T9, Tb, Tc; T8 = LD(&(xi[WS(is, 1)]), ivs, &(xi[WS(is, 1)])); T9 = LD(&(xi[WS(is, 9)]), ivs, &(xi[WS(is, 1)])); Ta = VSUB(T8, T9); TO = VADD(T8, T9); Tb = LD(&(xi[WS(is, 5)]), ivs, &(xi[WS(is, 1)])); Tc = LD(&(xi[WS(is, 13)]), ivs, &(xi[WS(is, 1)])); Td = VSUB(Tb, Tc); TP = VADD(Tb, Tc); } Te = VFMA(LDK(KP382683432), Ta, VMUL(LDK(KP923879532), Td)); T16 = VADD(TO, TP); Tx = VFNMS(LDK(KP382683432), Td, VMUL(LDK(KP923879532), Ta)); TQ = VSUB(TO, TP); } { V T15, T18, T19, T1a; T15 = VADD(T13, T14); T18 = VADD(T16, T17); ST(&(xo[WS(os, 8)]), VSUB(T15, T18), ovs, &(xo[0])); ST(&(xo[0]), VADD(T15, T18), ovs, &(xo[0])); T19 = VSUB(T13, T14); T1a = VBYI(VSUB(T17, T16)); ST(&(xo[WS(os, 12)]), VSUB(T19, T1a), ovs, &(xo[0])); ST(&(xo[WS(os, 4)]), VADD(T19, T1a), ovs, &(xo[0])); } { V TV, T11, T10, T12, TU, TZ; TU = VMUL(LDK(KP707106781), VADD(TQ, TT)); TV = VADD(TN, TU); T11 = VSUB(TN, TU); TZ = VMUL(LDK(KP707106781), VSUB(TT, TQ)); T10 = VBYI(VADD(TY, TZ)); T12 = VBYI(VSUB(TZ, TY)); ST(&(xo[WS(os, 14)]), VSUB(TV, T10), ovs, &(xo[0])); ST(&(xo[WS(os, 6)]), VADD(T11, T12), ovs, &(xo[0])); ST(&(xo[WS(os, 2)]), VADD(TV, T10), ovs, &(xo[0])); ST(&(xo[WS(os, 10)]), VSUB(T11, T12), ovs, &(xo[0])); } { V Tr, TB, TA, TC; { V Tf, Tq, Tw, Tz; Tf = VSUB(T7, Te); Tq = VSUB(Tm, Tp); Tr = VBYI(VSUB(Tf, Tq)); TB = VBYI(VADD(Tq, Tf)); Tw = VADD(Tu, Tv); Tz = VADD(Tx, Ty); TA = VSUB(Tw, Tz); TC = VADD(Tw, Tz); } ST(&(xo[WS(os, 7)]), VADD(Tr, TA), ovs, &(xo[WS(os, 1)])); ST(&(xo[WS(os, 15)]), VSUB(TC, TB), ovs, &(xo[WS(os, 1)])); ST(&(xo[WS(os, 9)]), VSUB(TA, Tr), ovs, &(xo[WS(os, 1)])); ST(&(xo[WS(os, 1)]), VADD(TB, TC), ovs, &(xo[WS(os, 1)])); } { V TF, TJ, TI, TK; { V TD, TE, TG, TH; TD = VSUB(Tu, Tv); TE = VADD(Te, T7); TF = VADD(TD, TE); TJ = VSUB(TD, TE); TG = VADD(Tp, Tm); TH = VSUB(Ty, Tx); TI = VBYI(VADD(TG, TH)); TK = VBYI(VSUB(TH, TG)); } ST(&(xo[WS(os, 13)]), VSUB(TF, TI), ovs, &(xo[WS(os, 1)])); ST(&(xo[WS(os, 5)]), VADD(TJ, TK), ovs, &(xo[WS(os, 1)])); ST(&(xo[WS(os, 3)]), VADD(TF, TI), ovs, &(xo[WS(os, 1)])); ST(&(xo[WS(os, 11)]), VSUB(TJ, TK), ovs, &(xo[WS(os, 1)])); } } } VLEAVE(); } static const kdft_desc desc = { 16, XSIMD_STRING("n1fv_16"), {68, 8, 4, 0}, &GENUS, 0, 0, 0, 0 }; void XSIMD(codelet_n1fv_16) (planner *p) { X(kdft_register) (p, n1fv_16, &desc); } #endif /* HAVE_FMA */ fftw-3.3.3/dft/simd/common/n1fv_32.c0000644000175000017500000006014012054410352013727 00000000000000/* * Copyright (c) 2003, 2007-11 Matteo Frigo * Copyright (c) 2003, 2007-11 Massachusetts Institute of Technology * * This program is free software; you can redistribute it and/or modify * it under the terms of the GNU General Public License as published by * the Free Software Foundation; either version 2 of the License, or * (at your option) any later version. * * This program is distributed in the hope that it will be useful, * but WITHOUT ANY WARRANTY; without even the implied warranty of * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the * GNU General Public License for more details. * * You should have received a copy of the GNU General Public License * along with this program; if not, write to the Free Software * Foundation, Inc., 51 Franklin Street, Fifth Floor, Boston, MA 02110-1301 USA * */ /* This file was automatically generated --- DO NOT EDIT */ /* Generated on Sun Nov 25 07:36:54 EST 2012 */ #include "codelet-dft.h" #ifdef HAVE_FMA /* Generated by: ../../../genfft/gen_notw_c.native -fma -reorder-insns -schedule-for-pipeline -simd -compact -variables 4 -pipeline-latency 8 -n 32 -name n1fv_32 -include n1f.h */ /* * This function contains 186 FP additions, 98 FP multiplications, * (or, 88 additions, 0 multiplications, 98 fused multiply/add), * 104 stack variables, 7 constants, and 64 memory accesses */ #include "n1f.h" static void n1fv_32(const R *ri, const R *ii, R *ro, R *io, stride is, stride os, INT v, INT ivs, INT ovs) { DVK(KP980785280, +0.980785280403230449126182236134239036973933731); DVK(KP831469612, +0.831469612302545237078788377617905756738560812); DVK(KP198912367, +0.198912367379658006911597622644676228597850501); DVK(KP668178637, +0.668178637919298919997757686523080761552472251); DVK(KP923879532, +0.923879532511286756128183189396788286822416626); DVK(KP414213562, +0.414213562373095048801688724209698078569671875); DVK(KP707106781, +0.707106781186547524400844362104849039284835938); { INT i; const R *xi; R *xo; xi = ri; xo = ro; for (i = v; i > 0; i = i - VL, xi = xi + (VL * ivs), xo = xo + (VL * ovs), MAKE_VOLATILE_STRIDE(64, is), MAKE_VOLATILE_STRIDE(64, os)) { V T1h, Tr, T1a, T1k, TI, T1b, T1L, T1P, T1I, T1G, T1O, T1Q, T1H, T1z, T1c; V TZ; { V T2x, T1T, T2K, T1W, T1p, Tb, T1A, T16, Tu, TF, T2N, T2H, T2b, T2t, TY; V T1w, TT, T1v, T20, T2C, Tj, Te, T2h, To, T2f, T23, T2D, TB, TG, Th; V T2i, Tk; { V TL, TW, TP, TQ, T2F, T27, T28, TO; { V T1, T2, T12, T13, T4, T5, T7, T8; T1 = LD(&(xi[0]), ivs, &(xi[0])); T2 = LD(&(xi[WS(is, 16)]), ivs, &(xi[0])); T12 = LD(&(xi[WS(is, 8)]), ivs, &(xi[0])); T13 = LD(&(xi[WS(is, 24)]), ivs, &(xi[0])); T4 = LD(&(xi[WS(is, 4)]), ivs, &(xi[0])); T5 = LD(&(xi[WS(is, 20)]), ivs, &(xi[0])); T7 = LD(&(xi[WS(is, 28)]), ivs, &(xi[0])); T8 = LD(&(xi[WS(is, 12)]), ivs, &(xi[0])); { V TM, T25, T26, TN; { V TJ, T3, T14, T1U, T6, T1V, T9, TK, TU, TV, T1R, T1S, Ta, T15; TJ = LD(&(xi[WS(is, 31)]), ivs, &(xi[WS(is, 1)])); T1R = VADD(T1, T2); T3 = VSUB(T1, T2); T1S = VADD(T12, T13); T14 = VSUB(T12, T13); T1U = VADD(T4, T5); T6 = VSUB(T4, T5); T1V = VADD(T7, T8); T9 = VSUB(T7, T8); TK = LD(&(xi[WS(is, 15)]), ivs, &(xi[WS(is, 1)])); TU = LD(&(xi[WS(is, 23)]), ivs, &(xi[WS(is, 1)])); T2x = VSUB(T1R, T1S); T1T = VADD(T1R, T1S); TV = LD(&(xi[WS(is, 7)]), ivs, &(xi[WS(is, 1)])); TM = LD(&(xi[WS(is, 3)]), ivs, &(xi[WS(is, 1)])); T2K = VSUB(T1V, T1U); T1W = VADD(T1U, T1V); Ta = VADD(T6, T9); T15 = VSUB(T9, T6); T25 = VADD(TJ, TK); TL = VSUB(TJ, TK); T26 = VADD(TV, TU); TW = VSUB(TU, TV); TN = LD(&(xi[WS(is, 19)]), ivs, &(xi[WS(is, 1)])); TP = LD(&(xi[WS(is, 27)]), ivs, &(xi[WS(is, 1)])); T1p = VFNMS(LDK(KP707106781), Ta, T3); Tb = VFMA(LDK(KP707106781), Ta, T3); T1A = VFMA(LDK(KP707106781), T15, T14); T16 = VFNMS(LDK(KP707106781), T15, T14); TQ = LD(&(xi[WS(is, 11)]), ivs, &(xi[WS(is, 1)])); } T2F = VSUB(T25, T26); T27 = VADD(T25, T26); T28 = VADD(TM, TN); TO = VSUB(TM, TN); } } { V Ty, T21, Tx, Tz, T1Y, T1Z; { V Ts, Tt, TD, T29, TR, TE, Tv, Tw; Ts = LD(&(xi[WS(is, 1)]), ivs, &(xi[WS(is, 1)])); Tt = LD(&(xi[WS(is, 17)]), ivs, &(xi[WS(is, 1)])); TD = LD(&(xi[WS(is, 9)]), ivs, &(xi[WS(is, 1)])); T29 = VADD(TP, TQ); TR = VSUB(TP, TQ); TE = LD(&(xi[WS(is, 25)]), ivs, &(xi[WS(is, 1)])); Tv = LD(&(xi[WS(is, 5)]), ivs, &(xi[WS(is, 1)])); Tw = LD(&(xi[WS(is, 21)]), ivs, &(xi[WS(is, 1)])); Ty = LD(&(xi[WS(is, 29)]), ivs, &(xi[WS(is, 1)])); T1Y = VADD(Ts, Tt); Tu = VSUB(Ts, Tt); { V T2G, T2a, TX, TS; T2G = VSUB(T29, T28); T2a = VADD(T28, T29); TX = VSUB(TR, TO); TS = VADD(TO, TR); T1Z = VADD(TD, TE); TF = VSUB(TD, TE); T21 = VADD(Tv, Tw); Tx = VSUB(Tv, Tw); T2N = VFMA(LDK(KP414213562), T2F, T2G); T2H = VFNMS(LDK(KP414213562), T2G, T2F); T2b = VSUB(T27, T2a); T2t = VADD(T27, T2a); TY = VFMA(LDK(KP707106781), TX, TW); T1w = VFNMS(LDK(KP707106781), TX, TW); TT = VFMA(LDK(KP707106781), TS, TL); T1v = VFNMS(LDK(KP707106781), TS, TL); Tz = LD(&(xi[WS(is, 13)]), ivs, &(xi[WS(is, 1)])); } } T20 = VADD(T1Y, T1Z); T2C = VSUB(T1Y, T1Z); { V Tc, Td, Tm, Tn; Tc = LD(&(xi[WS(is, 2)]), ivs, &(xi[0])); Td = LD(&(xi[WS(is, 18)]), ivs, &(xi[0])); Tm = LD(&(xi[WS(is, 22)]), ivs, &(xi[0])); Tn = LD(&(xi[WS(is, 6)]), ivs, &(xi[0])); { V Tf, TA, T22, Tg; Tf = LD(&(xi[WS(is, 10)]), ivs, &(xi[0])); TA = VSUB(Ty, Tz); T22 = VADD(Ty, Tz); Tg = LD(&(xi[WS(is, 26)]), ivs, &(xi[0])); Tj = LD(&(xi[WS(is, 30)]), ivs, &(xi[0])); Te = VSUB(Tc, Td); T2h = VADD(Tc, Td); To = VSUB(Tm, Tn); T2f = VADD(Tn, Tm); T23 = VADD(T21, T22); T2D = VSUB(T21, T22); TB = VADD(Tx, TA); TG = VSUB(Tx, TA); Th = VSUB(Tf, Tg); T2i = VADD(Tf, Tg); Tk = LD(&(xi[WS(is, 14)]), ivs, &(xi[0])); } } } } { V T1t, TH, T1s, TC, T2P, T2U, T2n, T2d, T2w, T2u, T1q, T19, T1B, Tq, T2W; V T2M, T2B, T2T, T2v, T2r, T2o, T2m, T2X, T2I; { V T1X, T2p, T2E, T2O, T2s, T2y, T2j, T17, Ti, T2e, Tl, T2c, T2l, T24; T1X = VSUB(T1T, T1W); T2p = VADD(T1T, T1W); T2E = VFNMS(LDK(KP414213562), T2D, T2C); T2O = VFMA(LDK(KP414213562), T2C, T2D); T2s = VADD(T20, T23); T24 = VSUB(T20, T23); T1t = VFNMS(LDK(KP707106781), TG, TF); TH = VFMA(LDK(KP707106781), TG, TF); T1s = VFNMS(LDK(KP707106781), TB, Tu); TC = VFMA(LDK(KP707106781), TB, Tu); T2y = VSUB(T2h, T2i); T2j = VADD(T2h, T2i); T17 = VFMA(LDK(KP414213562), Te, Th); Ti = VFNMS(LDK(KP414213562), Th, Te); T2e = VADD(Tj, Tk); Tl = VSUB(Tj, Tk); T2c = VADD(T24, T2b); T2l = VSUB(T2b, T24); { V T2L, T2A, T2q, T2k; T2P = VSUB(T2N, T2O); T2U = VADD(T2O, T2N); { V T2z, T2g, T18, Tp; T2z = VSUB(T2e, T2f); T2g = VADD(T2e, T2f); T18 = VFMA(LDK(KP414213562), Tl, To); Tp = VFNMS(LDK(KP414213562), To, Tl); T2n = VFMA(LDK(KP707106781), T2c, T1X); T2d = VFNMS(LDK(KP707106781), T2c, T1X); T2w = VSUB(T2t, T2s); T2u = VADD(T2s, T2t); T2L = VSUB(T2z, T2y); T2A = VADD(T2y, T2z); T2q = VADD(T2j, T2g); T2k = VSUB(T2g, T2j); T1q = VADD(T17, T18); T19 = VSUB(T17, T18); T1B = VSUB(Tp, Ti); Tq = VADD(Ti, Tp); } T2W = VFNMS(LDK(KP707106781), T2L, T2K); T2M = VFMA(LDK(KP707106781), T2L, T2K); T2B = VFMA(LDK(KP707106781), T2A, T2x); T2T = VFNMS(LDK(KP707106781), T2A, T2x); T2v = VSUB(T2p, T2q); T2r = VADD(T2p, T2q); T2o = VFMA(LDK(KP707106781), T2l, T2k); T2m = VFNMS(LDK(KP707106781), T2l, T2k); T2X = VSUB(T2H, T2E); T2I = VADD(T2E, T2H); } } { V T2V, T2Z, T2Y, T30, T2R, T2J; T2V = VFNMS(LDK(KP923879532), T2U, T2T); T2Z = VFMA(LDK(KP923879532), T2U, T2T); ST(&(xo[WS(os, 24)]), VFNMSI(T2w, T2v), ovs, &(xo[0])); ST(&(xo[WS(os, 8)]), VFMAI(T2w, T2v), ovs, &(xo[0])); ST(&(xo[0]), VADD(T2r, T2u), ovs, &(xo[0])); ST(&(xo[WS(os, 16)]), VSUB(T2r, T2u), ovs, &(xo[0])); ST(&(xo[WS(os, 28)]), VFNMSI(T2o, T2n), ovs, &(xo[0])); ST(&(xo[WS(os, 4)]), VFMAI(T2o, T2n), ovs, &(xo[0])); ST(&(xo[WS(os, 20)]), VFMAI(T2m, T2d), ovs, &(xo[0])); ST(&(xo[WS(os, 12)]), VFNMSI(T2m, T2d), ovs, &(xo[0])); T2Y = VFMA(LDK(KP923879532), T2X, T2W); T30 = VFNMS(LDK(KP923879532), T2X, T2W); T2R = VFMA(LDK(KP923879532), T2I, T2B); T2J = VFNMS(LDK(KP923879532), T2I, T2B); { V T1J, T1r, T1C, T1M, T2S, T2Q, T1u, T1D, T1E, T1x; T1J = VFNMS(LDK(KP923879532), T1q, T1p); T1r = VFMA(LDK(KP923879532), T1q, T1p); T1C = VFMA(LDK(KP923879532), T1B, T1A); T1M = VFNMS(LDK(KP923879532), T1B, T1A); ST(&(xo[WS(os, 6)]), VFNMSI(T30, T2Z), ovs, &(xo[0])); ST(&(xo[WS(os, 26)]), VFMAI(T30, T2Z), ovs, &(xo[0])); ST(&(xo[WS(os, 22)]), VFNMSI(T2Y, T2V), ovs, &(xo[0])); ST(&(xo[WS(os, 10)]), VFMAI(T2Y, T2V), ovs, &(xo[0])); T2S = VFMA(LDK(KP923879532), T2P, T2M); T2Q = VFNMS(LDK(KP923879532), T2P, T2M); T1u = VFMA(LDK(KP668178637), T1t, T1s); T1D = VFNMS(LDK(KP668178637), T1s, T1t); T1E = VFNMS(LDK(KP668178637), T1v, T1w); T1x = VFMA(LDK(KP668178637), T1w, T1v); { V T1K, T1F, T1N, T1y; T1h = VFNMS(LDK(KP923879532), Tq, Tb); Tr = VFMA(LDK(KP923879532), Tq, Tb); ST(&(xo[WS(os, 30)]), VFNMSI(T2S, T2R), ovs, &(xo[0])); ST(&(xo[WS(os, 2)]), VFMAI(T2S, T2R), ovs, &(xo[0])); ST(&(xo[WS(os, 18)]), VFMAI(T2Q, T2J), ovs, &(xo[0])); ST(&(xo[WS(os, 14)]), VFNMSI(T2Q, T2J), ovs, &(xo[0])); T1K = VADD(T1D, T1E); T1F = VSUB(T1D, T1E); T1N = VSUB(T1x, T1u); T1y = VADD(T1u, T1x); T1a = VFMA(LDK(KP923879532), T19, T16); T1k = VFNMS(LDK(KP923879532), T19, T16); TI = VFNMS(LDK(KP198912367), TH, TC); T1b = VFMA(LDK(KP198912367), TC, TH); T1L = VFMA(LDK(KP831469612), T1K, T1J); T1P = VFNMS(LDK(KP831469612), T1K, T1J); T1I = VFMA(LDK(KP831469612), T1F, T1C); T1G = VFNMS(LDK(KP831469612), T1F, T1C); T1O = VFMA(LDK(KP831469612), T1N, T1M); T1Q = VFNMS(LDK(KP831469612), T1N, T1M); T1H = VFMA(LDK(KP831469612), T1y, T1r); T1z = VFNMS(LDK(KP831469612), T1y, T1r); T1c = VFMA(LDK(KP198912367), TT, TY); TZ = VFNMS(LDK(KP198912367), TY, TT); } } } } } { V T1d, T1i, T10, T1l; ST(&(xo[WS(os, 21)]), VFNMSI(T1O, T1L), ovs, &(xo[WS(os, 1)])); ST(&(xo[WS(os, 11)]), VFMAI(T1O, T1L), ovs, &(xo[WS(os, 1)])); ST(&(xo[WS(os, 27)]), VFMAI(T1Q, T1P), ovs, &(xo[WS(os, 1)])); ST(&(xo[WS(os, 5)]), VFNMSI(T1Q, T1P), ovs, &(xo[WS(os, 1)])); ST(&(xo[WS(os, 3)]), VFMAI(T1I, T1H), ovs, &(xo[WS(os, 1)])); ST(&(xo[WS(os, 29)]), VFNMSI(T1I, T1H), ovs, &(xo[WS(os, 1)])); ST(&(xo[WS(os, 19)]), VFMAI(T1G, T1z), ovs, &(xo[WS(os, 1)])); ST(&(xo[WS(os, 13)]), VFNMSI(T1G, T1z), ovs, &(xo[WS(os, 1)])); T1d = VSUB(T1b, T1c); T1i = VADD(T1b, T1c); T10 = VADD(TI, TZ); T1l = VSUB(TZ, TI); { V T1n, T1j, T1e, T1g, T1o, T1m, T11, T1f; T1n = VFMA(LDK(KP980785280), T1i, T1h); T1j = VFNMS(LDK(KP980785280), T1i, T1h); T1e = VFNMS(LDK(KP980785280), T1d, T1a); T1g = VFMA(LDK(KP980785280), T1d, T1a); T1o = VFMA(LDK(KP980785280), T1l, T1k); T1m = VFNMS(LDK(KP980785280), T1l, T1k); T11 = VFNMS(LDK(KP980785280), T10, Tr); T1f = VFMA(LDK(KP980785280), T10, Tr); ST(&(xo[WS(os, 23)]), VFMAI(T1m, T1j), ovs, &(xo[WS(os, 1)])); ST(&(xo[WS(os, 9)]), VFNMSI(T1m, T1j), ovs, &(xo[WS(os, 1)])); ST(&(xo[WS(os, 25)]), VFNMSI(T1o, T1n), ovs, &(xo[WS(os, 1)])); ST(&(xo[WS(os, 7)]), VFMAI(T1o, T1n), ovs, &(xo[WS(os, 1)])); ST(&(xo[WS(os, 31)]), VFMAI(T1g, T1f), ovs, &(xo[WS(os, 1)])); ST(&(xo[WS(os, 1)]), VFNMSI(T1g, T1f), ovs, &(xo[WS(os, 1)])); ST(&(xo[WS(os, 15)]), VFMAI(T1e, T11), ovs, &(xo[WS(os, 1)])); ST(&(xo[WS(os, 17)]), VFNMSI(T1e, T11), ovs, &(xo[WS(os, 1)])); } } } } VLEAVE(); } static const kdft_desc desc = { 32, XSIMD_STRING("n1fv_32"), {88, 0, 98, 0}, &GENUS, 0, 0, 0, 0 }; void XSIMD(codelet_n1fv_32) (planner *p) { X(kdft_register) (p, n1fv_32, &desc); } #else /* HAVE_FMA */ /* Generated by: ../../../genfft/gen_notw_c.native -simd -compact -variables 4 -pipeline-latency 8 -n 32 -name n1fv_32 -include n1f.h */ /* * This function contains 186 FP additions, 42 FP multiplications, * (or, 170 additions, 26 multiplications, 16 fused multiply/add), * 58 stack variables, 7 constants, and 64 memory accesses */ #include "n1f.h" static void n1fv_32(const R *ri, const R *ii, R *ro, R *io, stride is, stride os, INT v, INT ivs, INT ovs) { DVK(KP195090322, +0.195090322016128267848284868477022240927691618); DVK(KP980785280, +0.980785280403230449126182236134239036973933731); DVK(KP555570233, +0.555570233019602224742830813948532874374937191); DVK(KP831469612, +0.831469612302545237078788377617905756738560812); DVK(KP382683432, +0.382683432365089771728459984030398866761344562); DVK(KP923879532, +0.923879532511286756128183189396788286822416626); DVK(KP707106781, +0.707106781186547524400844362104849039284835938); { INT i; const R *xi; R *xo; xi = ri; xo = ro; for (i = v; i > 0; i = i - VL, xi = xi + (VL * ivs), xo = xo + (VL * ovs), MAKE_VOLATILE_STRIDE(64, is), MAKE_VOLATILE_STRIDE(64, os)) { V T1T, T1W, T2K, T2x, T16, T1A, Tb, T1p, TT, T1v, TY, T1w, T27, T2a, T2b; V T2H, T2O, TC, T1s, TH, T1t, T20, T23, T24, T2E, T2N, T2g, T2j, Tq, T1B; V T19, T1q, T2A, T2L; { V T3, T1R, T15, T1S, T6, T1U, T9, T1V, T12, Ta; { V T1, T2, T13, T14; T1 = LD(&(xi[0]), ivs, &(xi[0])); T2 = LD(&(xi[WS(is, 16)]), ivs, &(xi[0])); T3 = VSUB(T1, T2); T1R = VADD(T1, T2); T13 = LD(&(xi[WS(is, 8)]), ivs, &(xi[0])); T14 = LD(&(xi[WS(is, 24)]), ivs, &(xi[0])); T15 = VSUB(T13, T14); T1S = VADD(T13, T14); } { V T4, T5, T7, T8; T4 = LD(&(xi[WS(is, 4)]), ivs, &(xi[0])); T5 = LD(&(xi[WS(is, 20)]), ivs, &(xi[0])); T6 = VSUB(T4, T5); T1U = VADD(T4, T5); T7 = LD(&(xi[WS(is, 28)]), ivs, &(xi[0])); T8 = LD(&(xi[WS(is, 12)]), ivs, &(xi[0])); T9 = VSUB(T7, T8); T1V = VADD(T7, T8); } T1T = VADD(T1R, T1S); T1W = VADD(T1U, T1V); T2K = VSUB(T1V, T1U); T2x = VSUB(T1R, T1S); T12 = VMUL(LDK(KP707106781), VSUB(T9, T6)); T16 = VSUB(T12, T15); T1A = VADD(T15, T12); Ta = VMUL(LDK(KP707106781), VADD(T6, T9)); Tb = VADD(T3, Ta); T1p = VSUB(T3, Ta); } { V TL, T25, TX, T26, TO, T28, TR, T29; { V TJ, TK, TV, TW; TJ = LD(&(xi[WS(is, 31)]), ivs, &(xi[WS(is, 1)])); TK = LD(&(xi[WS(is, 15)]), ivs, &(xi[WS(is, 1)])); TL = VSUB(TJ, TK); T25 = VADD(TJ, TK); TV = LD(&(xi[WS(is, 7)]), ivs, &(xi[WS(is, 1)])); TW = LD(&(xi[WS(is, 23)]), ivs, &(xi[WS(is, 1)])); TX = VSUB(TV, TW); T26 = VADD(TV, TW); } { V TM, TN, TP, TQ; TM = LD(&(xi[WS(is, 3)]), ivs, &(xi[WS(is, 1)])); TN = LD(&(xi[WS(is, 19)]), ivs, &(xi[WS(is, 1)])); TO = VSUB(TM, TN); T28 = VADD(TM, TN); TP = LD(&(xi[WS(is, 27)]), ivs, &(xi[WS(is, 1)])); TQ = LD(&(xi[WS(is, 11)]), ivs, &(xi[WS(is, 1)])); TR = VSUB(TP, TQ); T29 = VADD(TP, TQ); } { V TS, TU, T2F, T2G; TS = VMUL(LDK(KP707106781), VADD(TO, TR)); TT = VADD(TL, TS); T1v = VSUB(TL, TS); TU = VMUL(LDK(KP707106781), VSUB(TR, TO)); TY = VSUB(TU, TX); T1w = VADD(TX, TU); T27 = VADD(T25, T26); T2a = VADD(T28, T29); T2b = VSUB(T27, T2a); T2F = VSUB(T25, T26); T2G = VSUB(T29, T28); T2H = VFNMS(LDK(KP382683432), T2G, VMUL(LDK(KP923879532), T2F)); T2O = VFMA(LDK(KP382683432), T2F, VMUL(LDK(KP923879532), T2G)); } } { V Tu, T1Y, TG, T1Z, Tx, T21, TA, T22; { V Ts, Tt, TE, TF; Ts = LD(&(xi[WS(is, 1)]), ivs, &(xi[WS(is, 1)])); Tt = LD(&(xi[WS(is, 17)]), ivs, &(xi[WS(is, 1)])); Tu = VSUB(Ts, Tt); T1Y = VADD(Ts, Tt); TE = LD(&(xi[WS(is, 9)]), ivs, &(xi[WS(is, 1)])); TF = LD(&(xi[WS(is, 25)]), ivs, &(xi[WS(is, 1)])); TG = VSUB(TE, TF); T1Z = VADD(TE, TF); } { V Tv, Tw, Ty, Tz; Tv = LD(&(xi[WS(is, 5)]), ivs, &(xi[WS(is, 1)])); Tw = LD(&(xi[WS(is, 21)]), ivs, &(xi[WS(is, 1)])); Tx = VSUB(Tv, Tw); T21 = VADD(Tv, Tw); Ty = LD(&(xi[WS(is, 29)]), ivs, &(xi[WS(is, 1)])); Tz = LD(&(xi[WS(is, 13)]), ivs, &(xi[WS(is, 1)])); TA = VSUB(Ty, Tz); T22 = VADD(Ty, Tz); } { V TB, TD, T2C, T2D; TB = VMUL(LDK(KP707106781), VADD(Tx, TA)); TC = VADD(Tu, TB); T1s = VSUB(Tu, TB); TD = VMUL(LDK(KP707106781), VSUB(TA, Tx)); TH = VSUB(TD, TG); T1t = VADD(TG, TD); T20 = VADD(T1Y, T1Z); T23 = VADD(T21, T22); T24 = VSUB(T20, T23); T2C = VSUB(T1Y, T1Z); T2D = VSUB(T22, T21); T2E = VFMA(LDK(KP923879532), T2C, VMUL(LDK(KP382683432), T2D)); T2N = VFNMS(LDK(KP382683432), T2C, VMUL(LDK(KP923879532), T2D)); } } { V Te, T2h, To, T2f, Th, T2i, Tl, T2e, Ti, Tp; { V Tc, Td, Tm, Tn; Tc = LD(&(xi[WS(is, 2)]), ivs, &(xi[0])); Td = LD(&(xi[WS(is, 18)]), ivs, &(xi[0])); Te = VSUB(Tc, Td); T2h = VADD(Tc, Td); Tm = LD(&(xi[WS(is, 6)]), ivs, &(xi[0])); Tn = LD(&(xi[WS(is, 22)]), ivs, &(xi[0])); To = VSUB(Tm, Tn); T2f = VADD(Tm, Tn); } { V Tf, Tg, Tj, Tk; Tf = LD(&(xi[WS(is, 10)]), ivs, &(xi[0])); Tg = LD(&(xi[WS(is, 26)]), ivs, &(xi[0])); Th = VSUB(Tf, Tg); T2i = VADD(Tf, Tg); Tj = LD(&(xi[WS(is, 30)]), ivs, &(xi[0])); Tk = LD(&(xi[WS(is, 14)]), ivs, &(xi[0])); Tl = VSUB(Tj, Tk); T2e = VADD(Tj, Tk); } T2g = VADD(T2e, T2f); T2j = VADD(T2h, T2i); Ti = VFNMS(LDK(KP382683432), Th, VMUL(LDK(KP923879532), Te)); Tp = VFMA(LDK(KP923879532), Tl, VMUL(LDK(KP382683432), To)); Tq = VADD(Ti, Tp); T1B = VSUB(Tp, Ti); { V T17, T18, T2y, T2z; T17 = VFNMS(LDK(KP923879532), To, VMUL(LDK(KP382683432), Tl)); T18 = VFMA(LDK(KP382683432), Te, VMUL(LDK(KP923879532), Th)); T19 = VSUB(T17, T18); T1q = VADD(T18, T17); T2y = VSUB(T2h, T2i); T2z = VSUB(T2e, T2f); T2A = VMUL(LDK(KP707106781), VADD(T2y, T2z)); T2L = VMUL(LDK(KP707106781), VSUB(T2z, T2y)); } } { V T2d, T2n, T2m, T2o; { V T1X, T2c, T2k, T2l; T1X = VSUB(T1T, T1W); T2c = VMUL(LDK(KP707106781), VADD(T24, T2b)); T2d = VADD(T1X, T2c); T2n = VSUB(T1X, T2c); T2k = VSUB(T2g, T2j); T2l = VMUL(LDK(KP707106781), VSUB(T2b, T24)); T2m = VBYI(VADD(T2k, T2l)); T2o = VBYI(VSUB(T2l, T2k)); } ST(&(xo[WS(os, 28)]), VSUB(T2d, T2m), ovs, &(xo[0])); ST(&(xo[WS(os, 12)]), VADD(T2n, T2o), ovs, &(xo[0])); ST(&(xo[WS(os, 4)]), VADD(T2d, T2m), ovs, &(xo[0])); ST(&(xo[WS(os, 20)]), VSUB(T2n, T2o), ovs, &(xo[0])); } { V T2r, T2v, T2u, T2w; { V T2p, T2q, T2s, T2t; T2p = VADD(T1T, T1W); T2q = VADD(T2j, T2g); T2r = VADD(T2p, T2q); T2v = VSUB(T2p, T2q); T2s = VADD(T20, T23); T2t = VADD(T27, T2a); T2u = VADD(T2s, T2t); T2w = VBYI(VSUB(T2t, T2s)); } ST(&(xo[WS(os, 16)]), VSUB(T2r, T2u), ovs, &(xo[0])); ST(&(xo[WS(os, 8)]), VADD(T2v, T2w), ovs, &(xo[0])); ST(&(xo[0]), VADD(T2r, T2u), ovs, &(xo[0])); ST(&(xo[WS(os, 24)]), VSUB(T2v, T2w), ovs, &(xo[0])); } { V T2V, T2Z, T2Y, T30; { V T2T, T2U, T2W, T2X; T2T = VSUB(T2H, T2E); T2U = VSUB(T2L, T2K); T2V = VBYI(VSUB(T2T, T2U)); T2Z = VBYI(VADD(T2U, T2T)); T2W = VSUB(T2x, T2A); T2X = VSUB(T2O, T2N); T2Y = VSUB(T2W, T2X); T30 = VADD(T2W, T2X); } ST(&(xo[WS(os, 10)]), VADD(T2V, T2Y), ovs, &(xo[0])); ST(&(xo[WS(os, 26)]), VSUB(T30, T2Z), ovs, &(xo[0])); ST(&(xo[WS(os, 22)]), VSUB(T2Y, T2V), ovs, &(xo[0])); ST(&(xo[WS(os, 6)]), VADD(T2Z, T30), ovs, &(xo[0])); } { V T2J, T2R, T2Q, T2S; { V T2B, T2I, T2M, T2P; T2B = VADD(T2x, T2A); T2I = VADD(T2E, T2H); T2J = VADD(T2B, T2I); T2R = VSUB(T2B, T2I); T2M = VADD(T2K, T2L); T2P = VADD(T2N, T2O); T2Q = VBYI(VADD(T2M, T2P)); T2S = VBYI(VSUB(T2P, T2M)); } ST(&(xo[WS(os, 30)]), VSUB(T2J, T2Q), ovs, &(xo[0])); ST(&(xo[WS(os, 14)]), VADD(T2R, T2S), ovs, &(xo[0])); ST(&(xo[WS(os, 2)]), VADD(T2J, T2Q), ovs, &(xo[0])); ST(&(xo[WS(os, 18)]), VSUB(T2R, T2S), ovs, &(xo[0])); } { V T1r, T1C, T1M, T1K, T1F, T1N, T1y, T1J; T1r = VADD(T1p, T1q); T1C = VADD(T1A, T1B); T1M = VSUB(T1p, T1q); T1K = VSUB(T1B, T1A); { V T1D, T1E, T1u, T1x; T1D = VFNMS(LDK(KP555570233), T1s, VMUL(LDK(KP831469612), T1t)); T1E = VFMA(LDK(KP555570233), T1v, VMUL(LDK(KP831469612), T1w)); T1F = VADD(T1D, T1E); T1N = VSUB(T1E, T1D); T1u = VFMA(LDK(KP831469612), T1s, VMUL(LDK(KP555570233), T1t)); T1x = VFNMS(LDK(KP555570233), T1w, VMUL(LDK(KP831469612), T1v)); T1y = VADD(T1u, T1x); T1J = VSUB(T1x, T1u); } { V T1z, T1G, T1P, T1Q; T1z = VADD(T1r, T1y); T1G = VBYI(VADD(T1C, T1F)); ST(&(xo[WS(os, 29)]), VSUB(T1z, T1G), ovs, &(xo[WS(os, 1)])); ST(&(xo[WS(os, 3)]), VADD(T1z, T1G), ovs, &(xo[WS(os, 1)])); T1P = VBYI(VADD(T1K, T1J)); T1Q = VADD(T1M, T1N); ST(&(xo[WS(os, 5)]), VADD(T1P, T1Q), ovs, &(xo[WS(os, 1)])); ST(&(xo[WS(os, 27)]), VSUB(T1Q, T1P), ovs, &(xo[WS(os, 1)])); } { V T1H, T1I, T1L, T1O; T1H = VSUB(T1r, T1y); T1I = VBYI(VSUB(T1F, T1C)); ST(&(xo[WS(os, 19)]), VSUB(T1H, T1I), ovs, &(xo[WS(os, 1)])); ST(&(xo[WS(os, 13)]), VADD(T1H, T1I), ovs, &(xo[WS(os, 1)])); T1L = VBYI(VSUB(T1J, T1K)); T1O = VSUB(T1M, T1N); ST(&(xo[WS(os, 11)]), VADD(T1L, T1O), ovs, &(xo[WS(os, 1)])); ST(&(xo[WS(os, 21)]), VSUB(T1O, T1L), ovs, &(xo[WS(os, 1)])); } } { V Tr, T1a, T1k, T1i, T1d, T1l, T10, T1h; Tr = VADD(Tb, Tq); T1a = VADD(T16, T19); T1k = VSUB(Tb, Tq); T1i = VSUB(T19, T16); { V T1b, T1c, TI, TZ; T1b = VFNMS(LDK(KP195090322), TC, VMUL(LDK(KP980785280), TH)); T1c = VFMA(LDK(KP195090322), TT, VMUL(LDK(KP980785280), TY)); T1d = VADD(T1b, T1c); T1l = VSUB(T1c, T1b); TI = VFMA(LDK(KP980785280), TC, VMUL(LDK(KP195090322), TH)); TZ = VFNMS(LDK(KP195090322), TY, VMUL(LDK(KP980785280), TT)); T10 = VADD(TI, TZ); T1h = VSUB(TZ, TI); } { V T11, T1e, T1n, T1o; T11 = VADD(Tr, T10); T1e = VBYI(VADD(T1a, T1d)); ST(&(xo[WS(os, 31)]), VSUB(T11, T1e), ovs, &(xo[WS(os, 1)])); ST(&(xo[WS(os, 1)]), VADD(T11, T1e), ovs, &(xo[WS(os, 1)])); T1n = VBYI(VADD(T1i, T1h)); T1o = VADD(T1k, T1l); ST(&(xo[WS(os, 7)]), VADD(T1n, T1o), ovs, &(xo[WS(os, 1)])); ST(&(xo[WS(os, 25)]), VSUB(T1o, T1n), ovs, &(xo[WS(os, 1)])); } { V T1f, T1g, T1j, T1m; T1f = VSUB(Tr, T10); T1g = VBYI(VSUB(T1d, T1a)); ST(&(xo[WS(os, 17)]), VSUB(T1f, T1g), ovs, &(xo[WS(os, 1)])); ST(&(xo[WS(os, 15)]), VADD(T1f, T1g), ovs, &(xo[WS(os, 1)])); T1j = VBYI(VSUB(T1h, T1i)); T1m = VSUB(T1k, T1l); ST(&(xo[WS(os, 9)]), VADD(T1j, T1m), ovs, &(xo[WS(os, 1)])); ST(&(xo[WS(os, 23)]), VSUB(T1m, T1j), ovs, &(xo[WS(os, 1)])); } } } } VLEAVE(); } static const kdft_desc desc = { 32, XSIMD_STRING("n1fv_32"), {170, 26, 16, 0}, &GENUS, 0, 0, 0, 0 }; void XSIMD(codelet_n1fv_32) (planner *p) { X(kdft_register) (p, n1fv_32, &desc); } #endif /* HAVE_FMA */ fftw-3.3.3/dft/simd/common/n1fv_64.c0000644000175000017500000015735212054410372013752 00000000000000/* * Copyright (c) 2003, 2007-11 Matteo Frigo * Copyright (c) 2003, 2007-11 Massachusetts Institute of Technology * * This program is free software; you can redistribute it and/or modify * it under the terms of the GNU General Public License as published by * the Free Software Foundation; either version 2 of the License, or * (at your option) any later version. * * This program is distributed in the hope that it will be useful, * but WITHOUT ANY WARRANTY; without even the implied warranty of * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the * GNU General Public License for more details. * * You should have received a copy of the GNU General Public License * along with this program; if not, write to the Free Software * Foundation, Inc., 51 Franklin Street, Fifth Floor, Boston, MA 02110-1301 USA * */ /* This file was automatically generated --- DO NOT EDIT */ /* Generated on Sun Nov 25 07:36:54 EST 2012 */ #include "codelet-dft.h" #ifdef HAVE_FMA /* Generated by: ../../../genfft/gen_notw_c.native -fma -reorder-insns -schedule-for-pipeline -simd -compact -variables 4 -pipeline-latency 8 -n 64 -name n1fv_64 -include n1f.h */ /* * This function contains 456 FP additions, 258 FP multiplications, * (or, 198 additions, 0 multiplications, 258 fused multiply/add), * 168 stack variables, 15 constants, and 128 memory accesses */ #include "n1f.h" static void n1fv_64(const R *ri, const R *ii, R *ro, R *io, stride is, stride os, INT v, INT ivs, INT ovs) { DVK(KP773010453, +0.773010453362736960810906609758469800971041293); DVK(KP820678790, +0.820678790828660330972281985331011598767386482); DVK(KP881921264, +0.881921264348355029712756863660388349508442621); DVK(KP534511135, +0.534511135950791641089685961295362908582039528); DVK(KP995184726, +0.995184726672196886244836953109479921575474869); DVK(KP098491403, +0.098491403357164253077197521291327432293052451); DVK(KP980785280, +0.980785280403230449126182236134239036973933731); DVK(KP956940335, +0.956940335732208864935797886980269969482849206); DVK(KP303346683, +0.303346683607342391675883946941299872384187453); DVK(KP831469612, +0.831469612302545237078788377617905756738560812); DVK(KP923879532, +0.923879532511286756128183189396788286822416626); DVK(KP668178637, +0.668178637919298919997757686523080761552472251); DVK(KP198912367, +0.198912367379658006911597622644676228597850501); DVK(KP414213562, +0.414213562373095048801688724209698078569671875); DVK(KP707106781, +0.707106781186547524400844362104849039284835938); { INT i; const R *xi; R *xo; xi = ri; xo = ro; for (i = v; i > 0; i = i - VL, xi = xi + (VL * ivs), xo = xo + (VL * ovs), MAKE_VOLATILE_STRIDE(128, is), MAKE_VOLATILE_STRIDE(128, os)) { V T5T, T5S, T5X, T65, T5Z, T5R, T67, T63, T5U, T64; { V T7, T26, T5k, T6A, T47, T69, T2V, T3z, T6B, T4e, T6a, T5n, T3M, T2Y, T27; V Tm, T3A, T3l, T2a, TC, T5p, T4o, T6E, T6e, T3i, T3B, TR, T29, T4x, T5q; V T6h, T6D, T39, T3H, T3I, T3c, T5N, T57, T72, T6w, T5O, T5e, T71, T6t, T2y; V T1W, T2x, T1N, T33, T34, T3E, T32, T1p, T2v, T1g, T2u, T4M, T5K, T6p, T6Z; V T6m, T6Y, T5L, T4T; { V T4g, T4l, T3j, Tu, Tx, T4h, TA, T4i; { V T1, T2, T23, T24, T4, T5, T20, T21; T1 = LD(&(xi[0]), ivs, &(xi[0])); T2 = LD(&(xi[WS(is, 32)]), ivs, &(xi[0])); T23 = LD(&(xi[WS(is, 8)]), ivs, &(xi[0])); T24 = LD(&(xi[WS(is, 40)]), ivs, &(xi[0])); T4 = LD(&(xi[WS(is, 16)]), ivs, &(xi[0])); T5 = LD(&(xi[WS(is, 48)]), ivs, &(xi[0])); T20 = LD(&(xi[WS(is, 56)]), ivs, &(xi[0])); T21 = LD(&(xi[WS(is, 24)]), ivs, &(xi[0])); { V Ta, T48, Tk, T4c, T49, Td, Tf, Tg; { V T8, T43, T3, T44, T25, T5i, T6, T45, T22, T9, Ti, Tj, Tb, Tc; T8 = LD(&(xi[WS(is, 4)]), ivs, &(xi[0])); T43 = VSUB(T1, T2); T3 = VADD(T1, T2); T44 = VSUB(T23, T24); T25 = VADD(T23, T24); T5i = VSUB(T4, T5); T6 = VADD(T4, T5); T45 = VSUB(T20, T21); T22 = VADD(T20, T21); T9 = LD(&(xi[WS(is, 36)]), ivs, &(xi[0])); Ti = LD(&(xi[WS(is, 12)]), ivs, &(xi[0])); Tj = LD(&(xi[WS(is, 44)]), ivs, &(xi[0])); Tb = LD(&(xi[WS(is, 20)]), ivs, &(xi[0])); Tc = LD(&(xi[WS(is, 52)]), ivs, &(xi[0])); { V T2T, T46, T5j, T2U; T7 = VSUB(T3, T6); T2T = VADD(T3, T6); T46 = VADD(T44, T45); T5j = VSUB(T45, T44); T26 = VSUB(T22, T25); T2U = VADD(T25, T22); Ta = VADD(T8, T9); T48 = VSUB(T8, T9); Tk = VADD(Ti, Tj); T4c = VSUB(Tj, Ti); T5k = VFNMS(LDK(KP707106781), T5j, T5i); T6A = VFMA(LDK(KP707106781), T5j, T5i); T47 = VFMA(LDK(KP707106781), T46, T43); T69 = VFNMS(LDK(KP707106781), T46, T43); T2V = VADD(T2T, T2U); T3z = VSUB(T2T, T2U); T49 = VSUB(Tb, Tc); Td = VADD(Tb, Tc); } Tf = LD(&(xi[WS(is, 60)]), ivs, &(xi[0])); Tg = LD(&(xi[WS(is, 28)]), ivs, &(xi[0])); } { V Te, T2W, T5l, T4a, Tq, Tt, Tv, Tw, T5m, T4d, Tl, T2X, Ty, Tz, To; V Tp; To = LD(&(xi[WS(is, 2)]), ivs, &(xi[0])); Tp = LD(&(xi[WS(is, 34)]), ivs, &(xi[0])); { V Th, T4b, Tr, Ts; Tr = LD(&(xi[WS(is, 18)]), ivs, &(xi[0])); Ts = LD(&(xi[WS(is, 50)]), ivs, &(xi[0])); Te = VSUB(Ta, Td); T2W = VADD(Ta, Td); T5l = VFMA(LDK(KP414213562), T48, T49); T4a = VFNMS(LDK(KP414213562), T49, T48); Th = VADD(Tf, Tg); T4b = VSUB(Tf, Tg); Tq = VADD(To, Tp); T4g = VSUB(To, Tp); T4l = VSUB(Tr, Ts); Tt = VADD(Tr, Ts); Tv = LD(&(xi[WS(is, 10)]), ivs, &(xi[0])); Tw = LD(&(xi[WS(is, 42)]), ivs, &(xi[0])); T5m = VFMA(LDK(KP414213562), T4b, T4c); T4d = VFNMS(LDK(KP414213562), T4c, T4b); Tl = VSUB(Th, Tk); T2X = VADD(Th, Tk); Ty = LD(&(xi[WS(is, 58)]), ivs, &(xi[0])); Tz = LD(&(xi[WS(is, 26)]), ivs, &(xi[0])); } T3j = VADD(Tq, Tt); Tu = VSUB(Tq, Tt); Tx = VADD(Tv, Tw); T4h = VSUB(Tv, Tw); T6B = VSUB(T4d, T4a); T4e = VADD(T4a, T4d); T6a = VADD(T5l, T5m); T5n = VSUB(T5l, T5m); T3M = VSUB(T2X, T2W); T2Y = VADD(T2W, T2X); T27 = VSUB(Tl, Te); Tm = VADD(Te, Tl); TA = VADD(Ty, Tz); T4i = VSUB(Ty, Tz); } } } { V TK, T4p, T4u, T4k, T6d, T4n, T6c, TL, TN, TO, T3g, TJ, TF, TI; { V TD, TE, TG, TH; TD = LD(&(xi[WS(is, 62)]), ivs, &(xi[0])); TE = LD(&(xi[WS(is, 30)]), ivs, &(xi[0])); TG = LD(&(xi[WS(is, 14)]), ivs, &(xi[0])); TH = LD(&(xi[WS(is, 46)]), ivs, &(xi[0])); TK = LD(&(xi[WS(is, 54)]), ivs, &(xi[0])); { V T3k, TB, T4j, T4m; T3k = VADD(Tx, TA); TB = VSUB(Tx, TA); T4j = VADD(T4h, T4i); T4m = VSUB(T4h, T4i); T4p = VSUB(TD, TE); TF = VADD(TD, TE); T4u = VSUB(TH, TG); TI = VADD(TG, TH); T3A = VSUB(T3j, T3k); T3l = VADD(T3j, T3k); T2a = VFMA(LDK(KP414213562), Tu, TB); TC = VFNMS(LDK(KP414213562), TB, Tu); T4k = VFMA(LDK(KP707106781), T4j, T4g); T6d = VFNMS(LDK(KP707106781), T4j, T4g); T4n = VFMA(LDK(KP707106781), T4m, T4l); T6c = VFNMS(LDK(KP707106781), T4m, T4l); TL = LD(&(xi[WS(is, 22)]), ivs, &(xi[0])); } TN = LD(&(xi[WS(is, 6)]), ivs, &(xi[0])); TO = LD(&(xi[WS(is, 38)]), ivs, &(xi[0])); } T3g = VADD(TF, TI); TJ = VSUB(TF, TI); { V T3a, T1E, T52, T5b, T1x, T4Z, T6r, T6u, T5a, T1U, T55, T5c, T1L, T3b; { V T4V, T1t, T58, T1w, T1Q, T1T, T1I, T4Y, T59, T1J, T53, T1H; { V T1r, TM, T4r, TP, T4q, T1s, T1u, T1v; T1r = LD(&(xi[WS(is, 63)]), ivs, &(xi[WS(is, 1)])); T5p = VFMA(LDK(KP198912367), T4k, T4n); T4o = VFNMS(LDK(KP198912367), T4n, T4k); T6E = VFMA(LDK(KP668178637), T6c, T6d); T6e = VFNMS(LDK(KP668178637), T6d, T6c); TM = VADD(TK, TL); T4r = VSUB(TK, TL); TP = VADD(TN, TO); T4q = VSUB(TN, TO); T1s = LD(&(xi[WS(is, 31)]), ivs, &(xi[WS(is, 1)])); T1u = LD(&(xi[WS(is, 15)]), ivs, &(xi[WS(is, 1)])); T1v = LD(&(xi[WS(is, 47)]), ivs, &(xi[WS(is, 1)])); { V T1R, T4X, T6g, T4t, T6f, T4w, T1S, T1O, T1P; T1O = LD(&(xi[WS(is, 55)]), ivs, &(xi[WS(is, 1)])); T1P = LD(&(xi[WS(is, 23)]), ivs, &(xi[WS(is, 1)])); T1R = LD(&(xi[WS(is, 7)]), ivs, &(xi[WS(is, 1)])); { V T3h, TQ, T4s, T4v; T3h = VADD(TP, TM); TQ = VSUB(TM, TP); T4s = VADD(T4q, T4r); T4v = VSUB(T4r, T4q); T4V = VSUB(T1r, T1s); T1t = VADD(T1r, T1s); T58 = VSUB(T1v, T1u); T1w = VADD(T1u, T1v); T4X = VSUB(T1O, T1P); T1Q = VADD(T1O, T1P); T3i = VADD(T3g, T3h); T3B = VSUB(T3g, T3h); TR = VFNMS(LDK(KP414213562), TQ, TJ); T29 = VFMA(LDK(KP414213562), TJ, TQ); T6g = VFNMS(LDK(KP707106781), T4s, T4p); T4t = VFMA(LDK(KP707106781), T4s, T4p); T6f = VFNMS(LDK(KP707106781), T4v, T4u); T4w = VFMA(LDK(KP707106781), T4v, T4u); T1S = LD(&(xi[WS(is, 39)]), ivs, &(xi[WS(is, 1)])); } { V T4W, T1A, T50, T51, T1D, T1F, T1G; { V T1y, T1z, T1B, T1C; T1y = LD(&(xi[WS(is, 3)]), ivs, &(xi[WS(is, 1)])); T1z = LD(&(xi[WS(is, 35)]), ivs, &(xi[WS(is, 1)])); T1B = LD(&(xi[WS(is, 19)]), ivs, &(xi[WS(is, 1)])); T1C = LD(&(xi[WS(is, 51)]), ivs, &(xi[WS(is, 1)])); T4x = VFNMS(LDK(KP198912367), T4w, T4t); T5q = VFMA(LDK(KP198912367), T4t, T4w); T6h = VFNMS(LDK(KP668178637), T6g, T6f); T6D = VFMA(LDK(KP668178637), T6f, T6g); T4W = VSUB(T1R, T1S); T1T = VADD(T1R, T1S); T1A = VADD(T1y, T1z); T50 = VSUB(T1y, T1z); T51 = VSUB(T1C, T1B); T1D = VADD(T1B, T1C); } T1F = LD(&(xi[WS(is, 59)]), ivs, &(xi[WS(is, 1)])); T1G = LD(&(xi[WS(is, 27)]), ivs, &(xi[WS(is, 1)])); T1I = LD(&(xi[WS(is, 11)]), ivs, &(xi[WS(is, 1)])); T4Y = VADD(T4W, T4X); T59 = VSUB(T4X, T4W); T1J = LD(&(xi[WS(is, 43)]), ivs, &(xi[WS(is, 1)])); T3a = VADD(T1A, T1D); T1E = VSUB(T1A, T1D); T52 = VFMA(LDK(KP414213562), T51, T50); T5b = VFNMS(LDK(KP414213562), T50, T51); T53 = VSUB(T1F, T1G); T1H = VADD(T1F, T1G); } } } { V T37, T54, T1K, T38; T1x = VSUB(T1t, T1w); T37 = VADD(T1t, T1w); T4Z = VFMA(LDK(KP707106781), T4Y, T4V); T6r = VFNMS(LDK(KP707106781), T4Y, T4V); T54 = VSUB(T1J, T1I); T1K = VADD(T1I, T1J); T6u = VFNMS(LDK(KP707106781), T59, T58); T5a = VFMA(LDK(KP707106781), T59, T58); T38 = VADD(T1T, T1Q); T1U = VSUB(T1Q, T1T); T55 = VFNMS(LDK(KP414213562), T54, T53); T5c = VFMA(LDK(KP414213562), T53, T54); T1L = VSUB(T1H, T1K); T3b = VADD(T1H, T1K); T39 = VADD(T37, T38); T3H = VSUB(T37, T38); } } { V T4A, TW, T4N, TZ, T1j, T1m, T4O, T4D, T13, T4F, T16, T4G, T1a, T4I, T4J; V T1d; { V TU, TV, TX, TY, T56, T6v; TU = LD(&(xi[WS(is, 1)]), ivs, &(xi[WS(is, 1)])); T56 = VADD(T52, T55); T6v = VSUB(T55, T52); { V T5d, T6s, T1V, T1M; T5d = VADD(T5b, T5c); T6s = VSUB(T5c, T5b); T1V = VSUB(T1L, T1E); T1M = VADD(T1E, T1L); T3I = VSUB(T3b, T3a); T3c = VADD(T3a, T3b); T5N = VFNMS(LDK(KP923879532), T56, T4Z); T57 = VFMA(LDK(KP923879532), T56, T4Z); T72 = VFNMS(LDK(KP923879532), T6v, T6u); T6w = VFMA(LDK(KP923879532), T6v, T6u); T5O = VFNMS(LDK(KP923879532), T5d, T5a); T5e = VFMA(LDK(KP923879532), T5d, T5a); T71 = VFMA(LDK(KP923879532), T6s, T6r); T6t = VFNMS(LDK(KP923879532), T6s, T6r); T2y = VFNMS(LDK(KP707106781), T1V, T1U); T1W = VFMA(LDK(KP707106781), T1V, T1U); T2x = VFNMS(LDK(KP707106781), T1M, T1x); T1N = VFMA(LDK(KP707106781), T1M, T1x); TV = LD(&(xi[WS(is, 33)]), ivs, &(xi[WS(is, 1)])); } TX = LD(&(xi[WS(is, 17)]), ivs, &(xi[WS(is, 1)])); TY = LD(&(xi[WS(is, 49)]), ivs, &(xi[WS(is, 1)])); { V T1h, T1i, T1k, T1l; T1h = LD(&(xi[WS(is, 9)]), ivs, &(xi[WS(is, 1)])); T1i = LD(&(xi[WS(is, 41)]), ivs, &(xi[WS(is, 1)])); T1k = LD(&(xi[WS(is, 57)]), ivs, &(xi[WS(is, 1)])); T1l = LD(&(xi[WS(is, 25)]), ivs, &(xi[WS(is, 1)])); { V T11, T4B, T4C, T12, T14, T15; T11 = LD(&(xi[WS(is, 5)]), ivs, &(xi[WS(is, 1)])); T4A = VSUB(TU, TV); TW = VADD(TU, TV); T4N = VSUB(TX, TY); TZ = VADD(TX, TY); T1j = VADD(T1h, T1i); T4B = VSUB(T1h, T1i); T1m = VADD(T1k, T1l); T4C = VSUB(T1k, T1l); T12 = LD(&(xi[WS(is, 37)]), ivs, &(xi[WS(is, 1)])); T14 = LD(&(xi[WS(is, 21)]), ivs, &(xi[WS(is, 1)])); T15 = LD(&(xi[WS(is, 53)]), ivs, &(xi[WS(is, 1)])); { V T18, T19, T1b, T1c; T18 = LD(&(xi[WS(is, 61)]), ivs, &(xi[WS(is, 1)])); T19 = LD(&(xi[WS(is, 29)]), ivs, &(xi[WS(is, 1)])); T1b = LD(&(xi[WS(is, 13)]), ivs, &(xi[WS(is, 1)])); T1c = LD(&(xi[WS(is, 45)]), ivs, &(xi[WS(is, 1)])); T4O = VSUB(T4B, T4C); T4D = VADD(T4B, T4C); T13 = VADD(T11, T12); T4F = VSUB(T11, T12); T16 = VADD(T14, T15); T4G = VSUB(T14, T15); T1a = VADD(T18, T19); T4I = VSUB(T18, T19); T4J = VSUB(T1b, T1c); T1d = VADD(T1b, T1c); } } } } { V T30, T10, T6k, T4E, T4Q, T4H, T17, T6n, T4P, T1e, T4K, T4R, T1n, T31; T30 = VADD(TW, TZ); T10 = VSUB(TW, TZ); T6k = VFNMS(LDK(KP707106781), T4D, T4A); T4E = VFMA(LDK(KP707106781), T4D, T4A); T4Q = VFMA(LDK(KP414213562), T4F, T4G); T4H = VFNMS(LDK(KP414213562), T4G, T4F); T33 = VADD(T13, T16); T17 = VSUB(T13, T16); T6n = VFNMS(LDK(KP707106781), T4O, T4N); T4P = VFMA(LDK(KP707106781), T4O, T4N); T34 = VADD(T1a, T1d); T1e = VSUB(T1a, T1d); T4K = VFMA(LDK(KP414213562), T4J, T4I); T4R = VFNMS(LDK(KP414213562), T4I, T4J); T1n = VSUB(T1j, T1m); T31 = VADD(T1j, T1m); { V T1f, T1o, T6o, T4L, T4S, T6l; T1f = VADD(T17, T1e); T1o = VSUB(T17, T1e); T6o = VSUB(T4H, T4K); T4L = VADD(T4H, T4K); T4S = VADD(T4Q, T4R); T6l = VSUB(T4Q, T4R); T3E = VSUB(T30, T31); T32 = VADD(T30, T31); T1p = VFMA(LDK(KP707106781), T1o, T1n); T2v = VFNMS(LDK(KP707106781), T1o, T1n); T1g = VFMA(LDK(KP707106781), T1f, T10); T2u = VFNMS(LDK(KP707106781), T1f, T10); T4M = VFMA(LDK(KP923879532), T4L, T4E); T5K = VFNMS(LDK(KP923879532), T4L, T4E); T6p = VFMA(LDK(KP923879532), T6o, T6n); T6Z = VFNMS(LDK(KP923879532), T6o, T6n); T6m = VFNMS(LDK(KP923879532), T6l, T6k); T6Y = VFMA(LDK(KP923879532), T6l, T6k); T5L = VFNMS(LDK(KP923879532), T4S, T4P); T4T = VFMA(LDK(KP923879532), T4S, T4P); } } } } } } { V T6b, T6F, T7f, T6X, T70, T79, T7a, T73, T6C, T76, T77, T6i; { V T2Z, T3r, T3s, T3m, T3d, T3v; T2Z = VSUB(T2V, T2Y); T3r = VADD(T2V, T2Y); T3s = VADD(T3l, T3i); T3m = VSUB(T3i, T3l); T3d = VSUB(T39, T3c); T3v = VADD(T39, T3c); { V T3x, T3t, T3P, T3J, T3D, T3V, T3Q, T3G, T36, T3u, T3Y, T3O, T6V, T6W; { V T3N, T3C, T3F, T35; T3N = VSUB(T3B, T3A); T3C = VADD(T3A, T3B); T3F = VSUB(T33, T34); T35 = VADD(T33, T34); T3x = VSUB(T3r, T3s); T3t = VADD(T3r, T3s); T3P = VFMA(LDK(KP414213562), T3H, T3I); T3J = VFNMS(LDK(KP414213562), T3I, T3H); T3D = VFMA(LDK(KP707106781), T3C, T3z); T3V = VFNMS(LDK(KP707106781), T3C, T3z); T3Q = VFMA(LDK(KP414213562), T3E, T3F); T3G = VFNMS(LDK(KP414213562), T3F, T3E); T36 = VSUB(T32, T35); T3u = VADD(T32, T35); T3Y = VFNMS(LDK(KP707106781), T3N, T3M); T3O = VFMA(LDK(KP707106781), T3N, T3M); } T6b = VFNMS(LDK(KP923879532), T6a, T69); T6V = VFMA(LDK(KP923879532), T6a, T69); T6W = VADD(T6E, T6D); T6F = VSUB(T6D, T6E); { V T3K, T3Z, T3e, T3n; T3K = VADD(T3G, T3J); T3Z = VSUB(T3J, T3G); T3e = VADD(T36, T3d); T3n = VSUB(T3d, T36); { V T3w, T3y, T3R, T3W; T3w = VADD(T3u, T3v); T3y = VSUB(T3v, T3u); T3R = VSUB(T3P, T3Q); T3W = VADD(T3Q, T3P); { V T42, T40, T3L, T3T; T42 = VFNMS(LDK(KP923879532), T3Z, T3Y); T40 = VFMA(LDK(KP923879532), T3Z, T3Y); T3L = VFNMS(LDK(KP923879532), T3K, T3D); T3T = VFMA(LDK(KP923879532), T3K, T3D); { V T3o, T3q, T3f, T3p; T3o = VFNMS(LDK(KP707106781), T3n, T3m); T3q = VFMA(LDK(KP707106781), T3n, T3m); T3f = VFNMS(LDK(KP707106781), T3e, T2Z); T3p = VFMA(LDK(KP707106781), T3e, T2Z); ST(&(xo[WS(os, 48)]), VFNMSI(T3y, T3x), ovs, &(xo[0])); ST(&(xo[WS(os, 16)]), VFMAI(T3y, T3x), ovs, &(xo[0])); ST(&(xo[0]), VADD(T3t, T3w), ovs, &(xo[0])); ST(&(xo[WS(os, 32)]), VSUB(T3t, T3w), ovs, &(xo[0])); { V T41, T3X, T3S, T3U; T41 = VFMA(LDK(KP923879532), T3W, T3V); T3X = VFNMS(LDK(KP923879532), T3W, T3V); T3S = VFNMS(LDK(KP923879532), T3R, T3O); T3U = VFMA(LDK(KP923879532), T3R, T3O); ST(&(xo[WS(os, 8)]), VFMAI(T3q, T3p), ovs, &(xo[0])); ST(&(xo[WS(os, 56)]), VFNMSI(T3q, T3p), ovs, &(xo[0])); ST(&(xo[WS(os, 40)]), VFMAI(T3o, T3f), ovs, &(xo[0])); ST(&(xo[WS(os, 24)]), VFNMSI(T3o, T3f), ovs, &(xo[0])); ST(&(xo[WS(os, 44)]), VFNMSI(T40, T3X), ovs, &(xo[0])); ST(&(xo[WS(os, 20)]), VFMAI(T40, T3X), ovs, &(xo[0])); ST(&(xo[WS(os, 52)]), VFMAI(T42, T41), ovs, &(xo[0])); ST(&(xo[WS(os, 12)]), VFNMSI(T42, T41), ovs, &(xo[0])); ST(&(xo[WS(os, 4)]), VFMAI(T3U, T3T), ovs, &(xo[0])); ST(&(xo[WS(os, 60)]), VFNMSI(T3U, T3T), ovs, &(xo[0])); ST(&(xo[WS(os, 36)]), VFMAI(T3S, T3L), ovs, &(xo[0])); ST(&(xo[WS(os, 28)]), VFNMSI(T3S, T3L), ovs, &(xo[0])); T7f = VFNMS(LDK(KP831469612), T6W, T6V); T6X = VFMA(LDK(KP831469612), T6W, T6V); } } } } } T70 = VFMA(LDK(KP303346683), T6Z, T6Y); T79 = VFNMS(LDK(KP303346683), T6Y, T6Z); T7a = VFNMS(LDK(KP303346683), T71, T72); T73 = VFMA(LDK(KP303346683), T72, T71); T6C = VFNMS(LDK(KP923879532), T6B, T6A); T76 = VFMA(LDK(KP923879532), T6B, T6A); T77 = VSUB(T6e, T6h); T6i = VADD(T6e, T6h); } } { V T2r, T2D, T2C, T2s, T5H, T5o, T5v, T5D, T5r, T5I, T5x, T5h, T5F, T5B; { V TT, T2f, T2n, T1Y, T28, T2b, T2l, T2p, T2j, T2k; { V T1X, T2d, T7h, T7l, T2e, T1q, T75, T7d, T7m, T7k, T7c, T7e, Tn, TS; T2r = VFNMS(LDK(KP707106781), Tm, T7); Tn = VFMA(LDK(KP707106781), Tm, T7); TS = VADD(TC, TR); T2D = VSUB(TR, TC); { V T7b, T7j, T74, T7i, T78, T7g; T1X = VFNMS(LDK(KP198912367), T1W, T1N); T2d = VFMA(LDK(KP198912367), T1N, T1W); T7g = VADD(T79, T7a); T7b = VSUB(T79, T7a); T7j = VSUB(T73, T70); T74 = VADD(T70, T73); T7i = VFNMS(LDK(KP831469612), T77, T76); T78 = VFMA(LDK(KP831469612), T77, T76); T2j = VFNMS(LDK(KP923879532), TS, Tn); TT = VFMA(LDK(KP923879532), TS, Tn); T7h = VFMA(LDK(KP956940335), T7g, T7f); T7l = VFNMS(LDK(KP956940335), T7g, T7f); T2e = VFMA(LDK(KP198912367), T1g, T1p); T1q = VFNMS(LDK(KP198912367), T1p, T1g); T75 = VFNMS(LDK(KP956940335), T74, T6X); T7d = VFMA(LDK(KP956940335), T74, T6X); T7m = VFNMS(LDK(KP956940335), T7j, T7i); T7k = VFMA(LDK(KP956940335), T7j, T7i); T7c = VFNMS(LDK(KP956940335), T7b, T78); T7e = VFMA(LDK(KP956940335), T7b, T78); } T2k = VADD(T2e, T2d); T2f = VSUB(T2d, T2e); ST(&(xo[WS(os, 45)]), VFNMSI(T7k, T7h), ovs, &(xo[WS(os, 1)])); ST(&(xo[WS(os, 19)]), VFMAI(T7k, T7h), ovs, &(xo[WS(os, 1)])); ST(&(xo[WS(os, 51)]), VFMAI(T7m, T7l), ovs, &(xo[WS(os, 1)])); ST(&(xo[WS(os, 13)]), VFNMSI(T7m, T7l), ovs, &(xo[WS(os, 1)])); ST(&(xo[WS(os, 3)]), VFMAI(T7e, T7d), ovs, &(xo[WS(os, 1)])); ST(&(xo[WS(os, 61)]), VFNMSI(T7e, T7d), ovs, &(xo[WS(os, 1)])); ST(&(xo[WS(os, 35)]), VFMAI(T7c, T75), ovs, &(xo[WS(os, 1)])); ST(&(xo[WS(os, 29)]), VFNMSI(T7c, T75), ovs, &(xo[WS(os, 1)])); T2n = VSUB(T1X, T1q); T1Y = VADD(T1q, T1X); T2C = VFNMS(LDK(KP707106781), T27, T26); T28 = VFMA(LDK(KP707106781), T27, T26); T2b = VSUB(T29, T2a); T2s = VADD(T2a, T29); } T2l = VFNMS(LDK(KP980785280), T2k, T2j); T2p = VFMA(LDK(KP980785280), T2k, T2j); { V T5z, T4z, T5A, T5g; { V T4f, T4y, T1Z, T2h, T4U, T5t, T2m, T2c, T5u, T5f; T5H = VFNMS(LDK(KP923879532), T4e, T47); T4f = VFMA(LDK(KP923879532), T4e, T47); T4y = VADD(T4o, T4x); T5T = VSUB(T4x, T4o); T1Z = VFNMS(LDK(KP980785280), T1Y, TT); T2h = VFMA(LDK(KP980785280), T1Y, TT); T4U = VFNMS(LDK(KP098491403), T4T, T4M); T5t = VFMA(LDK(KP098491403), T4M, T4T); T2m = VFNMS(LDK(KP923879532), T2b, T28); T2c = VFMA(LDK(KP923879532), T2b, T28); T5u = VFMA(LDK(KP098491403), T57, T5e); T5f = VFNMS(LDK(KP098491403), T5e, T57); T5z = VFNMS(LDK(KP980785280), T4y, T4f); T4z = VFMA(LDK(KP980785280), T4y, T4f); T5S = VFNMS(LDK(KP923879532), T5n, T5k); T5o = VFMA(LDK(KP923879532), T5n, T5k); { V T2o, T2q, T2i, T2g; T2o = VFMA(LDK(KP980785280), T2n, T2m); T2q = VFNMS(LDK(KP980785280), T2n, T2m); T2i = VFMA(LDK(KP980785280), T2f, T2c); T2g = VFNMS(LDK(KP980785280), T2f, T2c); T5A = VADD(T5t, T5u); T5v = VSUB(T5t, T5u); T5D = VSUB(T5f, T4U); T5g = VADD(T4U, T5f); ST(&(xo[WS(os, 46)]), VFNMSI(T2o, T2l), ovs, &(xo[0])); ST(&(xo[WS(os, 18)]), VFMAI(T2o, T2l), ovs, &(xo[0])); ST(&(xo[WS(os, 50)]), VFMAI(T2q, T2p), ovs, &(xo[0])); ST(&(xo[WS(os, 14)]), VFNMSI(T2q, T2p), ovs, &(xo[0])); ST(&(xo[WS(os, 2)]), VFMAI(T2i, T2h), ovs, &(xo[0])); ST(&(xo[WS(os, 62)]), VFNMSI(T2i, T2h), ovs, &(xo[0])); ST(&(xo[WS(os, 34)]), VFMAI(T2g, T1Z), ovs, &(xo[0])); ST(&(xo[WS(os, 30)]), VFNMSI(T2g, T1Z), ovs, &(xo[0])); T5r = VSUB(T5p, T5q); T5I = VADD(T5p, T5q); } } T5x = VFMA(LDK(KP995184726), T5g, T4z); T5h = VFNMS(LDK(KP995184726), T5g, T4z); T5F = VFMA(LDK(KP995184726), T5A, T5z); T5B = VFNMS(LDK(KP995184726), T5A, T5z); } } { V T6J, T6R, T6L, T6z, T6T, T6P; { V T6N, T6j, T6O, T6y; { V T6q, T6H, T5C, T5s, T6I, T6x; T6q = VFNMS(LDK(KP534511135), T6p, T6m); T6H = VFMA(LDK(KP534511135), T6m, T6p); T5C = VFNMS(LDK(KP980785280), T5r, T5o); T5s = VFMA(LDK(KP980785280), T5r, T5o); T6I = VFMA(LDK(KP534511135), T6t, T6w); T6x = VFNMS(LDK(KP534511135), T6w, T6t); T6N = VFMA(LDK(KP831469612), T6i, T6b); T6j = VFNMS(LDK(KP831469612), T6i, T6b); { V T5E, T5G, T5y, T5w; T5E = VFNMS(LDK(KP995184726), T5D, T5C); T5G = VFMA(LDK(KP995184726), T5D, T5C); T5y = VFMA(LDK(KP995184726), T5v, T5s); T5w = VFNMS(LDK(KP995184726), T5v, T5s); T6O = VADD(T6H, T6I); T6J = VSUB(T6H, T6I); T6R = VSUB(T6x, T6q); T6y = VADD(T6q, T6x); ST(&(xo[WS(os, 47)]), VFMAI(T5E, T5B), ovs, &(xo[WS(os, 1)])); ST(&(xo[WS(os, 17)]), VFNMSI(T5E, T5B), ovs, &(xo[WS(os, 1)])); ST(&(xo[WS(os, 49)]), VFNMSI(T5G, T5F), ovs, &(xo[WS(os, 1)])); ST(&(xo[WS(os, 15)]), VFMAI(T5G, T5F), ovs, &(xo[WS(os, 1)])); ST(&(xo[WS(os, 63)]), VFMAI(T5y, T5x), ovs, &(xo[WS(os, 1)])); ST(&(xo[WS(os, 1)]), VFNMSI(T5y, T5x), ovs, &(xo[WS(os, 1)])); ST(&(xo[WS(os, 31)]), VFMAI(T5w, T5h), ovs, &(xo[WS(os, 1)])); ST(&(xo[WS(os, 33)]), VFNMSI(T5w, T5h), ovs, &(xo[WS(os, 1)])); } } T6L = VFMA(LDK(KP881921264), T6y, T6j); T6z = VFNMS(LDK(KP881921264), T6y, T6j); T6T = VFMA(LDK(KP881921264), T6O, T6N); T6P = VFNMS(LDK(KP881921264), T6O, T6N); } { V T2H, T2P, T2J, T2B, T2R, T2N; { V T2L, T2t, T2M, T2A; { V T2z, T2F, T6Q, T6G, T2G, T2w; T2z = VFMA(LDK(KP668178637), T2y, T2x); T2F = VFNMS(LDK(KP668178637), T2x, T2y); T6Q = VFMA(LDK(KP831469612), T6F, T6C); T6G = VFNMS(LDK(KP831469612), T6F, T6C); T2G = VFNMS(LDK(KP668178637), T2u, T2v); T2w = VFMA(LDK(KP668178637), T2v, T2u); T2L = VFNMS(LDK(KP923879532), T2s, T2r); T2t = VFMA(LDK(KP923879532), T2s, T2r); { V T6S, T6U, T6M, T6K; T6S = VFNMS(LDK(KP881921264), T6R, T6Q); T6U = VFMA(LDK(KP881921264), T6R, T6Q); T6M = VFMA(LDK(KP881921264), T6J, T6G); T6K = VFNMS(LDK(KP881921264), T6J, T6G); T2M = VADD(T2G, T2F); T2H = VSUB(T2F, T2G); T2P = VSUB(T2z, T2w); T2A = VADD(T2w, T2z); ST(&(xo[WS(os, 43)]), VFMAI(T6S, T6P), ovs, &(xo[WS(os, 1)])); ST(&(xo[WS(os, 21)]), VFNMSI(T6S, T6P), ovs, &(xo[WS(os, 1)])); ST(&(xo[WS(os, 53)]), VFNMSI(T6U, T6T), ovs, &(xo[WS(os, 1)])); ST(&(xo[WS(os, 11)]), VFMAI(T6U, T6T), ovs, &(xo[WS(os, 1)])); ST(&(xo[WS(os, 59)]), VFMAI(T6M, T6L), ovs, &(xo[WS(os, 1)])); ST(&(xo[WS(os, 5)]), VFNMSI(T6M, T6L), ovs, &(xo[WS(os, 1)])); ST(&(xo[WS(os, 27)]), VFMAI(T6K, T6z), ovs, &(xo[WS(os, 1)])); ST(&(xo[WS(os, 37)]), VFNMSI(T6K, T6z), ovs, &(xo[WS(os, 1)])); } } T2J = VFMA(LDK(KP831469612), T2A, T2t); T2B = VFNMS(LDK(KP831469612), T2A, T2t); T2R = VFNMS(LDK(KP831469612), T2M, T2L); T2N = VFMA(LDK(KP831469612), T2M, T2L); } { V T61, T5J, T62, T5Q; { V T5M, T5V, T2O, T2E, T5W, T5P; T5M = VFMA(LDK(KP820678790), T5L, T5K); T5V = VFNMS(LDK(KP820678790), T5K, T5L); T2O = VFMA(LDK(KP923879532), T2D, T2C); T2E = VFNMS(LDK(KP923879532), T2D, T2C); T5W = VFNMS(LDK(KP820678790), T5N, T5O); T5P = VFMA(LDK(KP820678790), T5O, T5N); T61 = VFNMS(LDK(KP980785280), T5I, T5H); T5J = VFMA(LDK(KP980785280), T5I, T5H); { V T2Q, T2S, T2K, T2I; T2Q = VFNMS(LDK(KP831469612), T2P, T2O); T2S = VFMA(LDK(KP831469612), T2P, T2O); T2K = VFMA(LDK(KP831469612), T2H, T2E); T2I = VFNMS(LDK(KP831469612), T2H, T2E); T62 = VADD(T5V, T5W); T5X = VSUB(T5V, T5W); T65 = VSUB(T5P, T5M); T5Q = VADD(T5M, T5P); ST(&(xo[WS(os, 42)]), VFMAI(T2Q, T2N), ovs, &(xo[0])); ST(&(xo[WS(os, 22)]), VFNMSI(T2Q, T2N), ovs, &(xo[0])); ST(&(xo[WS(os, 54)]), VFNMSI(T2S, T2R), ovs, &(xo[0])); ST(&(xo[WS(os, 10)]), VFMAI(T2S, T2R), ovs, &(xo[0])); ST(&(xo[WS(os, 58)]), VFMAI(T2K, T2J), ovs, &(xo[0])); ST(&(xo[WS(os, 6)]), VFNMSI(T2K, T2J), ovs, &(xo[0])); ST(&(xo[WS(os, 26)]), VFMAI(T2I, T2B), ovs, &(xo[0])); ST(&(xo[WS(os, 38)]), VFNMSI(T2I, T2B), ovs, &(xo[0])); } } T5Z = VFMA(LDK(KP773010453), T5Q, T5J); T5R = VFNMS(LDK(KP773010453), T5Q, T5J); T67 = VFNMS(LDK(KP773010453), T62, T61); T63 = VFMA(LDK(KP773010453), T62, T61); } } } } } } T5U = VFMA(LDK(KP980785280), T5T, T5S); T64 = VFNMS(LDK(KP980785280), T5T, T5S); { V T68, T66, T5Y, T60; T68 = VFNMS(LDK(KP773010453), T65, T64); T66 = VFMA(LDK(KP773010453), T65, T64); T5Y = VFNMS(LDK(KP773010453), T5X, T5U); T60 = VFMA(LDK(KP773010453), T5X, T5U); ST(&(xo[WS(os, 41)]), VFNMSI(T66, T63), ovs, &(xo[WS(os, 1)])); ST(&(xo[WS(os, 23)]), VFMAI(T66, T63), ovs, &(xo[WS(os, 1)])); ST(&(xo[WS(os, 55)]), VFMAI(T68, T67), ovs, &(xo[WS(os, 1)])); ST(&(xo[WS(os, 9)]), VFNMSI(T68, T67), ovs, &(xo[WS(os, 1)])); ST(&(xo[WS(os, 7)]), VFMAI(T60, T5Z), ovs, &(xo[WS(os, 1)])); ST(&(xo[WS(os, 57)]), VFNMSI(T60, T5Z), ovs, &(xo[WS(os, 1)])); ST(&(xo[WS(os, 39)]), VFMAI(T5Y, T5R), ovs, &(xo[WS(os, 1)])); ST(&(xo[WS(os, 25)]), VFNMSI(T5Y, T5R), ovs, &(xo[WS(os, 1)])); } } } VLEAVE(); } static const kdft_desc desc = { 64, XSIMD_STRING("n1fv_64"), {198, 0, 258, 0}, &GENUS, 0, 0, 0, 0 }; void XSIMD(codelet_n1fv_64) (planner *p) { X(kdft_register) (p, n1fv_64, &desc); } #else /* HAVE_FMA */ /* Generated by: ../../../genfft/gen_notw_c.native -simd -compact -variables 4 -pipeline-latency 8 -n 64 -name n1fv_64 -include n1f.h */ /* * This function contains 456 FP additions, 124 FP multiplications, * (or, 404 additions, 72 multiplications, 52 fused multiply/add), * 108 stack variables, 15 constants, and 128 memory accesses */ #include "n1f.h" static void n1fv_64(const R *ri, const R *ii, R *ro, R *io, stride is, stride os, INT v, INT ivs, INT ovs) { DVK(KP471396736, +0.471396736825997648556387625905254377657460319); DVK(KP881921264, +0.881921264348355029712756863660388349508442621); DVK(KP098017140, +0.098017140329560601994195563888641845861136673); DVK(KP995184726, +0.995184726672196886244836953109479921575474869); DVK(KP290284677, +0.290284677254462367636192375817395274691476278); DVK(KP956940335, +0.956940335732208864935797886980269969482849206); DVK(KP634393284, +0.634393284163645498215171613225493370675687095); DVK(KP773010453, +0.773010453362736960810906609758469800971041293); DVK(KP555570233, +0.555570233019602224742830813948532874374937191); DVK(KP831469612, +0.831469612302545237078788377617905756738560812); DVK(KP980785280, +0.980785280403230449126182236134239036973933731); DVK(KP195090322, +0.195090322016128267848284868477022240927691618); DVK(KP382683432, +0.382683432365089771728459984030398866761344562); DVK(KP923879532, +0.923879532511286756128183189396788286822416626); DVK(KP707106781, +0.707106781186547524400844362104849039284835938); { INT i; const R *xi; R *xo; xi = ri; xo = ro; for (i = v; i > 0; i = i - VL, xi = xi + (VL * ivs), xo = xo + (VL * ovs), MAKE_VOLATILE_STRIDE(128, is), MAKE_VOLATILE_STRIDE(128, os)) { V T4p, T5q, Tb, T39, T2n, T3A, T6f, T6T, Tq, T3B, T6i, T76, T2i, T3a, T4w; V T5r, TI, T2p, T6C, T6V, T3h, T3E, T4L, T5u, TZ, T2q, T6F, T6U, T3e, T3D; V T4E, T5t, T23, T2N, T6t, T71, T6w, T72, T2c, T2O, T3t, T41, T5f, T5R, T5k; V T5S, T3w, T42, T1s, T2K, T6m, T6Y, T6p, T6Z, T1B, T2L, T3m, T3Y, T4Y, T5O; V T53, T5P, T3p, T3Z; { V T3, T4n, T2m, T4o, T6, T5p, T9, T5o; { V T1, T2, T2k, T2l; T1 = LD(&(xi[0]), ivs, &(xi[0])); T2 = LD(&(xi[WS(is, 32)]), ivs, &(xi[0])); T3 = VSUB(T1, T2); T4n = VADD(T1, T2); T2k = LD(&(xi[WS(is, 16)]), ivs, &(xi[0])); T2l = LD(&(xi[WS(is, 48)]), ivs, &(xi[0])); T2m = VSUB(T2k, T2l); T4o = VADD(T2k, T2l); } { V T4, T5, T7, T8; T4 = LD(&(xi[WS(is, 8)]), ivs, &(xi[0])); T5 = LD(&(xi[WS(is, 40)]), ivs, &(xi[0])); T6 = VSUB(T4, T5); T5p = VADD(T4, T5); T7 = LD(&(xi[WS(is, 56)]), ivs, &(xi[0])); T8 = LD(&(xi[WS(is, 24)]), ivs, &(xi[0])); T9 = VSUB(T7, T8); T5o = VADD(T7, T8); } T4p = VSUB(T4n, T4o); T5q = VSUB(T5o, T5p); { V Ta, T2j, T6d, T6e; Ta = VMUL(LDK(KP707106781), VADD(T6, T9)); Tb = VADD(T3, Ta); T39 = VSUB(T3, Ta); T2j = VMUL(LDK(KP707106781), VSUB(T9, T6)); T2n = VSUB(T2j, T2m); T3A = VADD(T2m, T2j); T6d = VADD(T4n, T4o); T6e = VADD(T5p, T5o); T6f = VADD(T6d, T6e); T6T = VSUB(T6d, T6e); } } { V Te, T4q, To, T4u, Th, T4r, Tl, T4t; { V Tc, Td, Tm, Tn; Tc = LD(&(xi[WS(is, 4)]), ivs, &(xi[0])); Td = LD(&(xi[WS(is, 36)]), ivs, &(xi[0])); Te = VSUB(Tc, Td); T4q = VADD(Tc, Td); Tm = LD(&(xi[WS(is, 12)]), ivs, &(xi[0])); Tn = LD(&(xi[WS(is, 44)]), ivs, &(xi[0])); To = VSUB(Tm, Tn); T4u = VADD(Tm, Tn); } { V Tf, Tg, Tj, Tk; Tf = LD(&(xi[WS(is, 20)]), ivs, &(xi[0])); Tg = LD(&(xi[WS(is, 52)]), ivs, &(xi[0])); Th = VSUB(Tf, Tg); T4r = VADD(Tf, Tg); Tj = LD(&(xi[WS(is, 60)]), ivs, &(xi[0])); Tk = LD(&(xi[WS(is, 28)]), ivs, &(xi[0])); Tl = VSUB(Tj, Tk); T4t = VADD(Tj, Tk); } { V Ti, Tp, T6g, T6h; Ti = VFNMS(LDK(KP382683432), Th, VMUL(LDK(KP923879532), Te)); Tp = VFMA(LDK(KP923879532), Tl, VMUL(LDK(KP382683432), To)); Tq = VADD(Ti, Tp); T3B = VSUB(Tp, Ti); T6g = VADD(T4q, T4r); T6h = VADD(T4t, T4u); T6i = VADD(T6g, T6h); T76 = VSUB(T6h, T6g); } { V T2g, T2h, T4s, T4v; T2g = VFNMS(LDK(KP923879532), To, VMUL(LDK(KP382683432), Tl)); T2h = VFMA(LDK(KP382683432), Te, VMUL(LDK(KP923879532), Th)); T2i = VSUB(T2g, T2h); T3a = VADD(T2h, T2g); T4s = VSUB(T4q, T4r); T4v = VSUB(T4t, T4u); T4w = VMUL(LDK(KP707106781), VADD(T4s, T4v)); T5r = VMUL(LDK(KP707106781), VSUB(T4v, T4s)); } } { V Tu, T4F, TG, T4G, TB, T4J, TD, T4I; { V Ts, Tt, TE, TF; Ts = LD(&(xi[WS(is, 62)]), ivs, &(xi[0])); Tt = LD(&(xi[WS(is, 30)]), ivs, &(xi[0])); Tu = VSUB(Ts, Tt); T4F = VADD(Ts, Tt); TE = LD(&(xi[WS(is, 14)]), ivs, &(xi[0])); TF = LD(&(xi[WS(is, 46)]), ivs, &(xi[0])); TG = VSUB(TE, TF); T4G = VADD(TE, TF); { V Tv, Tw, Tx, Ty, Tz, TA; Tv = LD(&(xi[WS(is, 6)]), ivs, &(xi[0])); Tw = LD(&(xi[WS(is, 38)]), ivs, &(xi[0])); Tx = VSUB(Tv, Tw); Ty = LD(&(xi[WS(is, 54)]), ivs, &(xi[0])); Tz = LD(&(xi[WS(is, 22)]), ivs, &(xi[0])); TA = VSUB(Ty, Tz); TB = VMUL(LDK(KP707106781), VADD(Tx, TA)); T4J = VADD(Tv, Tw); TD = VMUL(LDK(KP707106781), VSUB(TA, Tx)); T4I = VADD(Ty, Tz); } } { V TC, TH, T6A, T6B; TC = VADD(Tu, TB); TH = VSUB(TD, TG); TI = VFMA(LDK(KP195090322), TC, VMUL(LDK(KP980785280), TH)); T2p = VFNMS(LDK(KP195090322), TH, VMUL(LDK(KP980785280), TC)); T6A = VADD(T4F, T4G); T6B = VADD(T4J, T4I); T6C = VADD(T6A, T6B); T6V = VSUB(T6A, T6B); } { V T3f, T3g, T4H, T4K; T3f = VSUB(Tu, TB); T3g = VADD(TG, TD); T3h = VFNMS(LDK(KP555570233), T3g, VMUL(LDK(KP831469612), T3f)); T3E = VFMA(LDK(KP555570233), T3f, VMUL(LDK(KP831469612), T3g)); T4H = VSUB(T4F, T4G); T4K = VSUB(T4I, T4J); T4L = VFNMS(LDK(KP382683432), T4K, VMUL(LDK(KP923879532), T4H)); T5u = VFMA(LDK(KP382683432), T4H, VMUL(LDK(KP923879532), T4K)); } } { V TS, T4z, TW, T4y, TP, T4C, TX, T4B; { V TQ, TR, TU, TV; TQ = LD(&(xi[WS(is, 18)]), ivs, &(xi[0])); TR = LD(&(xi[WS(is, 50)]), ivs, &(xi[0])); TS = VSUB(TQ, TR); T4z = VADD(TQ, TR); TU = LD(&(xi[WS(is, 2)]), ivs, &(xi[0])); TV = LD(&(xi[WS(is, 34)]), ivs, &(xi[0])); TW = VSUB(TU, TV); T4y = VADD(TU, TV); { V TJ, TK, TL, TM, TN, TO; TJ = LD(&(xi[WS(is, 58)]), ivs, &(xi[0])); TK = LD(&(xi[WS(is, 26)]), ivs, &(xi[0])); TL = VSUB(TJ, TK); TM = LD(&(xi[WS(is, 10)]), ivs, &(xi[0])); TN = LD(&(xi[WS(is, 42)]), ivs, &(xi[0])); TO = VSUB(TM, TN); TP = VMUL(LDK(KP707106781), VSUB(TL, TO)); T4C = VADD(TM, TN); TX = VMUL(LDK(KP707106781), VADD(TO, TL)); T4B = VADD(TJ, TK); } } { V TT, TY, T6D, T6E; TT = VSUB(TP, TS); TY = VADD(TW, TX); TZ = VFNMS(LDK(KP195090322), TY, VMUL(LDK(KP980785280), TT)); T2q = VFMA(LDK(KP980785280), TY, VMUL(LDK(KP195090322), TT)); T6D = VADD(T4y, T4z); T6E = VADD(T4C, T4B); T6F = VADD(T6D, T6E); T6U = VSUB(T6D, T6E); } { V T3c, T3d, T4A, T4D; T3c = VSUB(TW, TX); T3d = VADD(TS, TP); T3e = VFMA(LDK(KP831469612), T3c, VMUL(LDK(KP555570233), T3d)); T3D = VFNMS(LDK(KP555570233), T3c, VMUL(LDK(KP831469612), T3d)); T4A = VSUB(T4y, T4z); T4D = VSUB(T4B, T4C); T4E = VFMA(LDK(KP923879532), T4A, VMUL(LDK(KP382683432), T4D)); T5t = VFNMS(LDK(KP382683432), T4A, VMUL(LDK(KP923879532), T4D)); } } { V T1F, T55, T2a, T56, T1M, T5h, T27, T5g, T58, T59, T1U, T5a, T25, T5b, T5c; V T21, T5d, T24; { V T1D, T1E, T28, T29; T1D = LD(&(xi[WS(is, 63)]), ivs, &(xi[WS(is, 1)])); T1E = LD(&(xi[WS(is, 31)]), ivs, &(xi[WS(is, 1)])); T1F = VSUB(T1D, T1E); T55 = VADD(T1D, T1E); T28 = LD(&(xi[WS(is, 15)]), ivs, &(xi[WS(is, 1)])); T29 = LD(&(xi[WS(is, 47)]), ivs, &(xi[WS(is, 1)])); T2a = VSUB(T28, T29); T56 = VADD(T28, T29); } { V T1G, T1H, T1I, T1J, T1K, T1L; T1G = LD(&(xi[WS(is, 7)]), ivs, &(xi[WS(is, 1)])); T1H = LD(&(xi[WS(is, 39)]), ivs, &(xi[WS(is, 1)])); T1I = VSUB(T1G, T1H); T1J = LD(&(xi[WS(is, 55)]), ivs, &(xi[WS(is, 1)])); T1K = LD(&(xi[WS(is, 23)]), ivs, &(xi[WS(is, 1)])); T1L = VSUB(T1J, T1K); T1M = VMUL(LDK(KP707106781), VADD(T1I, T1L)); T5h = VADD(T1G, T1H); T27 = VMUL(LDK(KP707106781), VSUB(T1L, T1I)); T5g = VADD(T1J, T1K); } { V T1Q, T1T, T1X, T20; { V T1O, T1P, T1R, T1S; T1O = LD(&(xi[WS(is, 3)]), ivs, &(xi[WS(is, 1)])); T1P = LD(&(xi[WS(is, 35)]), ivs, &(xi[WS(is, 1)])); T1Q = VSUB(T1O, T1P); T58 = VADD(T1O, T1P); T1R = LD(&(xi[WS(is, 19)]), ivs, &(xi[WS(is, 1)])); T1S = LD(&(xi[WS(is, 51)]), ivs, &(xi[WS(is, 1)])); T1T = VSUB(T1R, T1S); T59 = VADD(T1R, T1S); } T1U = VFNMS(LDK(KP382683432), T1T, VMUL(LDK(KP923879532), T1Q)); T5a = VSUB(T58, T59); T25 = VFMA(LDK(KP382683432), T1Q, VMUL(LDK(KP923879532), T1T)); { V T1V, T1W, T1Y, T1Z; T1V = LD(&(xi[WS(is, 59)]), ivs, &(xi[WS(is, 1)])); T1W = LD(&(xi[WS(is, 27)]), ivs, &(xi[WS(is, 1)])); T1X = VSUB(T1V, T1W); T5b = VADD(T1V, T1W); T1Y = LD(&(xi[WS(is, 11)]), ivs, &(xi[WS(is, 1)])); T1Z = LD(&(xi[WS(is, 43)]), ivs, &(xi[WS(is, 1)])); T20 = VSUB(T1Y, T1Z); T5c = VADD(T1Y, T1Z); } T21 = VFMA(LDK(KP923879532), T1X, VMUL(LDK(KP382683432), T20)); T5d = VSUB(T5b, T5c); T24 = VFNMS(LDK(KP923879532), T20, VMUL(LDK(KP382683432), T1X)); } { V T1N, T22, T6r, T6s; T1N = VADD(T1F, T1M); T22 = VADD(T1U, T21); T23 = VSUB(T1N, T22); T2N = VADD(T1N, T22); T6r = VADD(T55, T56); T6s = VADD(T5h, T5g); T6t = VADD(T6r, T6s); T71 = VSUB(T6r, T6s); } { V T6u, T6v, T26, T2b; T6u = VADD(T58, T59); T6v = VADD(T5b, T5c); T6w = VADD(T6u, T6v); T72 = VSUB(T6v, T6u); T26 = VSUB(T24, T25); T2b = VSUB(T27, T2a); T2c = VSUB(T26, T2b); T2O = VADD(T2b, T26); } { V T3r, T3s, T57, T5e; T3r = VSUB(T1F, T1M); T3s = VADD(T25, T24); T3t = VADD(T3r, T3s); T41 = VSUB(T3r, T3s); T57 = VSUB(T55, T56); T5e = VMUL(LDK(KP707106781), VADD(T5a, T5d)); T5f = VADD(T57, T5e); T5R = VSUB(T57, T5e); } { V T5i, T5j, T3u, T3v; T5i = VSUB(T5g, T5h); T5j = VMUL(LDK(KP707106781), VSUB(T5d, T5a)); T5k = VADD(T5i, T5j); T5S = VSUB(T5j, T5i); T3u = VADD(T2a, T27); T3v = VSUB(T21, T1U); T3w = VADD(T3u, T3v); T42 = VSUB(T3v, T3u); } } { V T1q, T4P, T1v, T4O, T1n, T50, T1w, T4Z, T4U, T4V, T18, T4W, T1z, T4R, T4S; V T1f, T4T, T1y; { V T1o, T1p, T1t, T1u; T1o = LD(&(xi[WS(is, 17)]), ivs, &(xi[WS(is, 1)])); T1p = LD(&(xi[WS(is, 49)]), ivs, &(xi[WS(is, 1)])); T1q = VSUB(T1o, T1p); T4P = VADD(T1o, T1p); T1t = LD(&(xi[WS(is, 1)]), ivs, &(xi[WS(is, 1)])); T1u = LD(&(xi[WS(is, 33)]), ivs, &(xi[WS(is, 1)])); T1v = VSUB(T1t, T1u); T4O = VADD(T1t, T1u); } { V T1h, T1i, T1j, T1k, T1l, T1m; T1h = LD(&(xi[WS(is, 57)]), ivs, &(xi[WS(is, 1)])); T1i = LD(&(xi[WS(is, 25)]), ivs, &(xi[WS(is, 1)])); T1j = VSUB(T1h, T1i); T1k = LD(&(xi[WS(is, 9)]), ivs, &(xi[WS(is, 1)])); T1l = LD(&(xi[WS(is, 41)]), ivs, &(xi[WS(is, 1)])); T1m = VSUB(T1k, T1l); T1n = VMUL(LDK(KP707106781), VSUB(T1j, T1m)); T50 = VADD(T1k, T1l); T1w = VMUL(LDK(KP707106781), VADD(T1m, T1j)); T4Z = VADD(T1h, T1i); } { V T14, T17, T1b, T1e; { V T12, T13, T15, T16; T12 = LD(&(xi[WS(is, 61)]), ivs, &(xi[WS(is, 1)])); T13 = LD(&(xi[WS(is, 29)]), ivs, &(xi[WS(is, 1)])); T14 = VSUB(T12, T13); T4U = VADD(T12, T13); T15 = LD(&(xi[WS(is, 13)]), ivs, &(xi[WS(is, 1)])); T16 = LD(&(xi[WS(is, 45)]), ivs, &(xi[WS(is, 1)])); T17 = VSUB(T15, T16); T4V = VADD(T15, T16); } T18 = VFNMS(LDK(KP923879532), T17, VMUL(LDK(KP382683432), T14)); T4W = VSUB(T4U, T4V); T1z = VFMA(LDK(KP923879532), T14, VMUL(LDK(KP382683432), T17)); { V T19, T1a, T1c, T1d; T19 = LD(&(xi[WS(is, 5)]), ivs, &(xi[WS(is, 1)])); T1a = LD(&(xi[WS(is, 37)]), ivs, &(xi[WS(is, 1)])); T1b = VSUB(T19, T1a); T4R = VADD(T19, T1a); T1c = LD(&(xi[WS(is, 21)]), ivs, &(xi[WS(is, 1)])); T1d = LD(&(xi[WS(is, 53)]), ivs, &(xi[WS(is, 1)])); T1e = VSUB(T1c, T1d); T4S = VADD(T1c, T1d); } T1f = VFMA(LDK(KP382683432), T1b, VMUL(LDK(KP923879532), T1e)); T4T = VSUB(T4R, T4S); T1y = VFNMS(LDK(KP382683432), T1e, VMUL(LDK(KP923879532), T1b)); } { V T1g, T1r, T6k, T6l; T1g = VSUB(T18, T1f); T1r = VSUB(T1n, T1q); T1s = VSUB(T1g, T1r); T2K = VADD(T1r, T1g); T6k = VADD(T4O, T4P); T6l = VADD(T50, T4Z); T6m = VADD(T6k, T6l); T6Y = VSUB(T6k, T6l); } { V T6n, T6o, T1x, T1A; T6n = VADD(T4R, T4S); T6o = VADD(T4U, T4V); T6p = VADD(T6n, T6o); T6Z = VSUB(T6o, T6n); T1x = VADD(T1v, T1w); T1A = VADD(T1y, T1z); T1B = VSUB(T1x, T1A); T2L = VADD(T1x, T1A); } { V T3k, T3l, T4Q, T4X; T3k = VSUB(T1v, T1w); T3l = VADD(T1f, T18); T3m = VADD(T3k, T3l); T3Y = VSUB(T3k, T3l); T4Q = VSUB(T4O, T4P); T4X = VMUL(LDK(KP707106781), VADD(T4T, T4W)); T4Y = VADD(T4Q, T4X); T5O = VSUB(T4Q, T4X); } { V T51, T52, T3n, T3o; T51 = VSUB(T4Z, T50); T52 = VMUL(LDK(KP707106781), VSUB(T4W, T4T)); T53 = VADD(T51, T52); T5P = VSUB(T52, T51); T3n = VADD(T1q, T1n); T3o = VSUB(T1z, T1y); T3p = VADD(T3n, T3o); T3Z = VSUB(T3o, T3n); } } { V T6N, T6R, T6Q, T6S; { V T6L, T6M, T6O, T6P; T6L = VADD(T6f, T6i); T6M = VADD(T6F, T6C); T6N = VADD(T6L, T6M); T6R = VSUB(T6L, T6M); T6O = VADD(T6m, T6p); T6P = VADD(T6t, T6w); T6Q = VADD(T6O, T6P); T6S = VBYI(VSUB(T6P, T6O)); } ST(&(xo[WS(os, 32)]), VSUB(T6N, T6Q), ovs, &(xo[0])); ST(&(xo[WS(os, 16)]), VADD(T6R, T6S), ovs, &(xo[0])); ST(&(xo[0]), VADD(T6N, T6Q), ovs, &(xo[0])); ST(&(xo[WS(os, 48)]), VSUB(T6R, T6S), ovs, &(xo[0])); } { V T6j, T6G, T6y, T6H, T6q, T6x; T6j = VSUB(T6f, T6i); T6G = VSUB(T6C, T6F); T6q = VSUB(T6m, T6p); T6x = VSUB(T6t, T6w); T6y = VMUL(LDK(KP707106781), VADD(T6q, T6x)); T6H = VMUL(LDK(KP707106781), VSUB(T6x, T6q)); { V T6z, T6I, T6J, T6K; T6z = VADD(T6j, T6y); T6I = VBYI(VADD(T6G, T6H)); ST(&(xo[WS(os, 56)]), VSUB(T6z, T6I), ovs, &(xo[0])); ST(&(xo[WS(os, 8)]), VADD(T6z, T6I), ovs, &(xo[0])); T6J = VSUB(T6j, T6y); T6K = VBYI(VSUB(T6H, T6G)); ST(&(xo[WS(os, 40)]), VSUB(T6J, T6K), ovs, &(xo[0])); ST(&(xo[WS(os, 24)]), VADD(T6J, T6K), ovs, &(xo[0])); } } { V T6X, T7i, T78, T7g, T74, T7f, T7b, T7j, T6W, T77; T6W = VMUL(LDK(KP707106781), VADD(T6U, T6V)); T6X = VADD(T6T, T6W); T7i = VSUB(T6T, T6W); T77 = VMUL(LDK(KP707106781), VSUB(T6V, T6U)); T78 = VADD(T76, T77); T7g = VSUB(T77, T76); { V T70, T73, T79, T7a; T70 = VFMA(LDK(KP923879532), T6Y, VMUL(LDK(KP382683432), T6Z)); T73 = VFNMS(LDK(KP382683432), T72, VMUL(LDK(KP923879532), T71)); T74 = VADD(T70, T73); T7f = VSUB(T73, T70); T79 = VFNMS(LDK(KP382683432), T6Y, VMUL(LDK(KP923879532), T6Z)); T7a = VFMA(LDK(KP382683432), T71, VMUL(LDK(KP923879532), T72)); T7b = VADD(T79, T7a); T7j = VSUB(T7a, T79); } { V T75, T7c, T7l, T7m; T75 = VADD(T6X, T74); T7c = VBYI(VADD(T78, T7b)); ST(&(xo[WS(os, 60)]), VSUB(T75, T7c), ovs, &(xo[0])); ST(&(xo[WS(os, 4)]), VADD(T75, T7c), ovs, &(xo[0])); T7l = VBYI(VADD(T7g, T7f)); T7m = VADD(T7i, T7j); ST(&(xo[WS(os, 12)]), VADD(T7l, T7m), ovs, &(xo[0])); ST(&(xo[WS(os, 52)]), VSUB(T7m, T7l), ovs, &(xo[0])); } { V T7d, T7e, T7h, T7k; T7d = VSUB(T6X, T74); T7e = VBYI(VSUB(T7b, T78)); ST(&(xo[WS(os, 36)]), VSUB(T7d, T7e), ovs, &(xo[0])); ST(&(xo[WS(os, 28)]), VADD(T7d, T7e), ovs, &(xo[0])); T7h = VBYI(VSUB(T7f, T7g)); T7k = VSUB(T7i, T7j); ST(&(xo[WS(os, 20)]), VADD(T7h, T7k), ovs, &(xo[0])); ST(&(xo[WS(os, 44)]), VSUB(T7k, T7h), ovs, &(xo[0])); } } { V T5N, T68, T61, T69, T5U, T65, T5Y, T66; { V T5L, T5M, T5Z, T60; T5L = VSUB(T4p, T4w); T5M = VSUB(T5u, T5t); T5N = VADD(T5L, T5M); T68 = VSUB(T5L, T5M); T5Z = VFNMS(LDK(KP555570233), T5O, VMUL(LDK(KP831469612), T5P)); T60 = VFMA(LDK(KP555570233), T5R, VMUL(LDK(KP831469612), T5S)); T61 = VADD(T5Z, T60); T69 = VSUB(T60, T5Z); } { V T5Q, T5T, T5W, T5X; T5Q = VFMA(LDK(KP831469612), T5O, VMUL(LDK(KP555570233), T5P)); T5T = VFNMS(LDK(KP555570233), T5S, VMUL(LDK(KP831469612), T5R)); T5U = VADD(T5Q, T5T); T65 = VSUB(T5T, T5Q); T5W = VSUB(T5r, T5q); T5X = VSUB(T4L, T4E); T5Y = VADD(T5W, T5X); T66 = VSUB(T5X, T5W); } { V T5V, T62, T6b, T6c; T5V = VADD(T5N, T5U); T62 = VBYI(VADD(T5Y, T61)); ST(&(xo[WS(os, 58)]), VSUB(T5V, T62), ovs, &(xo[0])); ST(&(xo[WS(os, 6)]), VADD(T5V, T62), ovs, &(xo[0])); T6b = VBYI(VADD(T66, T65)); T6c = VADD(T68, T69); ST(&(xo[WS(os, 10)]), VADD(T6b, T6c), ovs, &(xo[0])); ST(&(xo[WS(os, 54)]), VSUB(T6c, T6b), ovs, &(xo[0])); } { V T63, T64, T67, T6a; T63 = VSUB(T5N, T5U); T64 = VBYI(VSUB(T61, T5Y)); ST(&(xo[WS(os, 38)]), VSUB(T63, T64), ovs, &(xo[0])); ST(&(xo[WS(os, 26)]), VADD(T63, T64), ovs, &(xo[0])); T67 = VBYI(VSUB(T65, T66)); T6a = VSUB(T68, T69); ST(&(xo[WS(os, 22)]), VADD(T67, T6a), ovs, &(xo[0])); ST(&(xo[WS(os, 42)]), VSUB(T6a, T67), ovs, &(xo[0])); } } { V T11, T2C, T2v, T2D, T2e, T2z, T2s, T2A; { V Tr, T10, T2t, T2u; Tr = VSUB(Tb, Tq); T10 = VSUB(TI, TZ); T11 = VADD(Tr, T10); T2C = VSUB(Tr, T10); T2t = VFNMS(LDK(KP634393284), T1B, VMUL(LDK(KP773010453), T1s)); T2u = VFMA(LDK(KP773010453), T2c, VMUL(LDK(KP634393284), T23)); T2v = VADD(T2t, T2u); T2D = VSUB(T2u, T2t); } { V T1C, T2d, T2o, T2r; T1C = VFMA(LDK(KP634393284), T1s, VMUL(LDK(KP773010453), T1B)); T2d = VFNMS(LDK(KP634393284), T2c, VMUL(LDK(KP773010453), T23)); T2e = VADD(T1C, T2d); T2z = VSUB(T2d, T1C); T2o = VSUB(T2i, T2n); T2r = VSUB(T2p, T2q); T2s = VADD(T2o, T2r); T2A = VSUB(T2r, T2o); } { V T2f, T2w, T2F, T2G; T2f = VADD(T11, T2e); T2w = VBYI(VADD(T2s, T2v)); ST(&(xo[WS(os, 57)]), VSUB(T2f, T2w), ovs, &(xo[WS(os, 1)])); ST(&(xo[WS(os, 7)]), VADD(T2f, T2w), ovs, &(xo[WS(os, 1)])); T2F = VBYI(VADD(T2A, T2z)); T2G = VADD(T2C, T2D); ST(&(xo[WS(os, 9)]), VADD(T2F, T2G), ovs, &(xo[WS(os, 1)])); ST(&(xo[WS(os, 55)]), VSUB(T2G, T2F), ovs, &(xo[WS(os, 1)])); } { V T2x, T2y, T2B, T2E; T2x = VSUB(T11, T2e); T2y = VBYI(VSUB(T2v, T2s)); ST(&(xo[WS(os, 39)]), VSUB(T2x, T2y), ovs, &(xo[WS(os, 1)])); ST(&(xo[WS(os, 25)]), VADD(T2x, T2y), ovs, &(xo[WS(os, 1)])); T2B = VBYI(VSUB(T2z, T2A)); T2E = VSUB(T2C, T2D); ST(&(xo[WS(os, 23)]), VADD(T2B, T2E), ovs, &(xo[WS(os, 1)])); ST(&(xo[WS(os, 41)]), VSUB(T2E, T2B), ovs, &(xo[WS(os, 1)])); } } { V T3j, T3Q, T3J, T3R, T3y, T3N, T3G, T3O; { V T3b, T3i, T3H, T3I; T3b = VADD(T39, T3a); T3i = VADD(T3e, T3h); T3j = VADD(T3b, T3i); T3Q = VSUB(T3b, T3i); T3H = VFNMS(LDK(KP290284677), T3m, VMUL(LDK(KP956940335), T3p)); T3I = VFMA(LDK(KP290284677), T3t, VMUL(LDK(KP956940335), T3w)); T3J = VADD(T3H, T3I); T3R = VSUB(T3I, T3H); } { V T3q, T3x, T3C, T3F; T3q = VFMA(LDK(KP956940335), T3m, VMUL(LDK(KP290284677), T3p)); T3x = VFNMS(LDK(KP290284677), T3w, VMUL(LDK(KP956940335), T3t)); T3y = VADD(T3q, T3x); T3N = VSUB(T3x, T3q); T3C = VADD(T3A, T3B); T3F = VADD(T3D, T3E); T3G = VADD(T3C, T3F); T3O = VSUB(T3F, T3C); } { V T3z, T3K, T3T, T3U; T3z = VADD(T3j, T3y); T3K = VBYI(VADD(T3G, T3J)); ST(&(xo[WS(os, 61)]), VSUB(T3z, T3K), ovs, &(xo[WS(os, 1)])); ST(&(xo[WS(os, 3)]), VADD(T3z, T3K), ovs, &(xo[WS(os, 1)])); T3T = VBYI(VADD(T3O, T3N)); T3U = VADD(T3Q, T3R); ST(&(xo[WS(os, 13)]), VADD(T3T, T3U), ovs, &(xo[WS(os, 1)])); ST(&(xo[WS(os, 51)]), VSUB(T3U, T3T), ovs, &(xo[WS(os, 1)])); } { V T3L, T3M, T3P, T3S; T3L = VSUB(T3j, T3y); T3M = VBYI(VSUB(T3J, T3G)); ST(&(xo[WS(os, 35)]), VSUB(T3L, T3M), ovs, &(xo[WS(os, 1)])); ST(&(xo[WS(os, 29)]), VADD(T3L, T3M), ovs, &(xo[WS(os, 1)])); T3P = VBYI(VSUB(T3N, T3O)); T3S = VSUB(T3Q, T3R); ST(&(xo[WS(os, 19)]), VADD(T3P, T3S), ovs, &(xo[WS(os, 1)])); ST(&(xo[WS(os, 45)]), VSUB(T3S, T3P), ovs, &(xo[WS(os, 1)])); } } { V T4N, T5G, T5z, T5H, T5m, T5D, T5w, T5E; { V T4x, T4M, T5x, T5y; T4x = VADD(T4p, T4w); T4M = VADD(T4E, T4L); T4N = VADD(T4x, T4M); T5G = VSUB(T4x, T4M); T5x = VFNMS(LDK(KP195090322), T4Y, VMUL(LDK(KP980785280), T53)); T5y = VFMA(LDK(KP195090322), T5f, VMUL(LDK(KP980785280), T5k)); T5z = VADD(T5x, T5y); T5H = VSUB(T5y, T5x); } { V T54, T5l, T5s, T5v; T54 = VFMA(LDK(KP980785280), T4Y, VMUL(LDK(KP195090322), T53)); T5l = VFNMS(LDK(KP195090322), T5k, VMUL(LDK(KP980785280), T5f)); T5m = VADD(T54, T5l); T5D = VSUB(T5l, T54); T5s = VADD(T5q, T5r); T5v = VADD(T5t, T5u); T5w = VADD(T5s, T5v); T5E = VSUB(T5v, T5s); } { V T5n, T5A, T5J, T5K; T5n = VADD(T4N, T5m); T5A = VBYI(VADD(T5w, T5z)); ST(&(xo[WS(os, 62)]), VSUB(T5n, T5A), ovs, &(xo[0])); ST(&(xo[WS(os, 2)]), VADD(T5n, T5A), ovs, &(xo[0])); T5J = VBYI(VADD(T5E, T5D)); T5K = VADD(T5G, T5H); ST(&(xo[WS(os, 14)]), VADD(T5J, T5K), ovs, &(xo[0])); ST(&(xo[WS(os, 50)]), VSUB(T5K, T5J), ovs, &(xo[0])); } { V T5B, T5C, T5F, T5I; T5B = VSUB(T4N, T5m); T5C = VBYI(VSUB(T5z, T5w)); ST(&(xo[WS(os, 34)]), VSUB(T5B, T5C), ovs, &(xo[0])); ST(&(xo[WS(os, 30)]), VADD(T5B, T5C), ovs, &(xo[0])); T5F = VBYI(VSUB(T5D, T5E)); T5I = VSUB(T5G, T5H); ST(&(xo[WS(os, 18)]), VADD(T5F, T5I), ovs, &(xo[0])); ST(&(xo[WS(os, 46)]), VSUB(T5I, T5F), ovs, &(xo[0])); } } { V T2J, T34, T2X, T35, T2Q, T31, T2U, T32; { V T2H, T2I, T2V, T2W; T2H = VADD(Tb, Tq); T2I = VADD(T2q, T2p); T2J = VADD(T2H, T2I); T34 = VSUB(T2H, T2I); T2V = VFNMS(LDK(KP098017140), T2L, VMUL(LDK(KP995184726), T2K)); T2W = VFMA(LDK(KP995184726), T2O, VMUL(LDK(KP098017140), T2N)); T2X = VADD(T2V, T2W); T35 = VSUB(T2W, T2V); } { V T2M, T2P, T2S, T2T; T2M = VFMA(LDK(KP098017140), T2K, VMUL(LDK(KP995184726), T2L)); T2P = VFNMS(LDK(KP098017140), T2O, VMUL(LDK(KP995184726), T2N)); T2Q = VADD(T2M, T2P); T31 = VSUB(T2P, T2M); T2S = VADD(T2n, T2i); T2T = VADD(TZ, TI); T2U = VADD(T2S, T2T); T32 = VSUB(T2T, T2S); } { V T2R, T2Y, T37, T38; T2R = VADD(T2J, T2Q); T2Y = VBYI(VADD(T2U, T2X)); ST(&(xo[WS(os, 63)]), VSUB(T2R, T2Y), ovs, &(xo[WS(os, 1)])); ST(&(xo[WS(os, 1)]), VADD(T2R, T2Y), ovs, &(xo[WS(os, 1)])); T37 = VBYI(VADD(T32, T31)); T38 = VADD(T34, T35); ST(&(xo[WS(os, 15)]), VADD(T37, T38), ovs, &(xo[WS(os, 1)])); ST(&(xo[WS(os, 49)]), VSUB(T38, T37), ovs, &(xo[WS(os, 1)])); } { V T2Z, T30, T33, T36; T2Z = VSUB(T2J, T2Q); T30 = VBYI(VSUB(T2X, T2U)); ST(&(xo[WS(os, 33)]), VSUB(T2Z, T30), ovs, &(xo[WS(os, 1)])); ST(&(xo[WS(os, 31)]), VADD(T2Z, T30), ovs, &(xo[WS(os, 1)])); T33 = VBYI(VSUB(T31, T32)); T36 = VSUB(T34, T35); ST(&(xo[WS(os, 17)]), VADD(T33, T36), ovs, &(xo[WS(os, 1)])); ST(&(xo[WS(os, 47)]), VSUB(T36, T33), ovs, &(xo[WS(os, 1)])); } } { V T3X, T4i, T4b, T4j, T44, T4f, T48, T4g; { V T3V, T3W, T49, T4a; T3V = VSUB(T39, T3a); T3W = VSUB(T3E, T3D); T3X = VADD(T3V, T3W); T4i = VSUB(T3V, T3W); T49 = VFNMS(LDK(KP471396736), T3Y, VMUL(LDK(KP881921264), T3Z)); T4a = VFMA(LDK(KP471396736), T41, VMUL(LDK(KP881921264), T42)); T4b = VADD(T49, T4a); T4j = VSUB(T4a, T49); } { V T40, T43, T46, T47; T40 = VFMA(LDK(KP881921264), T3Y, VMUL(LDK(KP471396736), T3Z)); T43 = VFNMS(LDK(KP471396736), T42, VMUL(LDK(KP881921264), T41)); T44 = VADD(T40, T43); T4f = VSUB(T43, T40); T46 = VSUB(T3B, T3A); T47 = VSUB(T3h, T3e); T48 = VADD(T46, T47); T4g = VSUB(T47, T46); } { V T45, T4c, T4l, T4m; T45 = VADD(T3X, T44); T4c = VBYI(VADD(T48, T4b)); ST(&(xo[WS(os, 59)]), VSUB(T45, T4c), ovs, &(xo[WS(os, 1)])); ST(&(xo[WS(os, 5)]), VADD(T45, T4c), ovs, &(xo[WS(os, 1)])); T4l = VBYI(VADD(T4g, T4f)); T4m = VADD(T4i, T4j); ST(&(xo[WS(os, 11)]), VADD(T4l, T4m), ovs, &(xo[WS(os, 1)])); ST(&(xo[WS(os, 53)]), VSUB(T4m, T4l), ovs, &(xo[WS(os, 1)])); } { V T4d, T4e, T4h, T4k; T4d = VSUB(T3X, T44); T4e = VBYI(VSUB(T4b, T48)); ST(&(xo[WS(os, 37)]), VSUB(T4d, T4e), ovs, &(xo[WS(os, 1)])); ST(&(xo[WS(os, 27)]), VADD(T4d, T4e), ovs, &(xo[WS(os, 1)])); T4h = VBYI(VSUB(T4f, T4g)); T4k = VSUB(T4i, T4j); ST(&(xo[WS(os, 21)]), VADD(T4h, T4k), ovs, &(xo[WS(os, 1)])); ST(&(xo[WS(os, 43)]), VSUB(T4k, T4h), ovs, &(xo[WS(os, 1)])); } } } } VLEAVE(); } static const kdft_desc desc = { 64, XSIMD_STRING("n1fv_64"), {404, 72, 52, 0}, &GENUS, 0, 0, 0, 0 }; void XSIMD(codelet_n1fv_64) (planner *p) { X(kdft_register) (p, n1fv_64, &desc); } #endif /* HAVE_FMA */ fftw-3.3.3/dft/simd/common/n1fv_128.c0000644000175000017500000037764012054410547014043 00000000000000/* * Copyright (c) 2003, 2007-11 Matteo Frigo * Copyright (c) 2003, 2007-11 Massachusetts Institute of Technology * * This program is free software; you can redistribute it and/or modify * it under the terms of the GNU General Public License as published by * the Free Software Foundation; either version 2 of the License, or * (at your option) any later version. * * This program is distributed in the hope that it will be useful, * but WITHOUT ANY WARRANTY; without even the implied warranty of * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the * GNU General Public License for more details. * * You should have received a copy of the GNU General Public License * along with this program; if not, write to the Free Software * Foundation, Inc., 51 Franklin Street, Fifth Floor, Boston, MA 02110-1301 USA * */ /* This file was automatically generated --- DO NOT EDIT */ /* Generated on Sun Nov 25 07:36:54 EST 2012 */ #include "codelet-dft.h" #ifdef HAVE_FMA /* Generated by: ../../../genfft/gen_notw_c.native -fma -reorder-insns -schedule-for-pipeline -simd -compact -variables 4 -pipeline-latency 8 -n 128 -name n1fv_128 -include n1f.h */ /* * This function contains 1082 FP additions, 642 FP multiplications, * (or, 440 additions, 0 multiplications, 642 fused multiply/add), * 295 stack variables, 31 constants, and 256 memory accesses */ #include "n1f.h" static void n1fv_128(const R *ri, const R *ii, R *ro, R *io, stride is, stride os, INT v, INT ivs, INT ovs) { DVK(KP903989293, +0.903989293123443331586200297230537048710132025); DVK(KP941544065, +0.941544065183020778412509402599502357185589796); DVK(KP357805721, +0.357805721314524104672487743774474392487532769); DVK(KP472964775, +0.472964775891319928124438237972992463904131113); DVK(KP857728610, +0.857728610000272069902269984284770137042490799); DVK(KP970031253, +0.970031253194543992603984207286100251456865962); DVK(KP250486960, +0.250486960191305461595702160124721208578685568); DVK(KP998795456, +0.998795456205172392714771604759100694443203615); DVK(KP740951125, +0.740951125354959091175616897495162729728955309); DVK(KP599376933, +0.599376933681923766271389869014404232837890546); DVK(KP906347169, +0.906347169019147157946142717268914412664134293); DVK(KP049126849, +0.049126849769467254105343321271313617079695752); DVK(KP989176509, +0.989176509964780973451673738016243063983689533); DVK(KP803207531, +0.803207531480644909806676512963141923879569427); DVK(KP741650546, +0.741650546272035369581266691172079863842265220); DVK(KP956940335, +0.956940335732208864935797886980269969482849206); DVK(KP148335987, +0.148335987538347428753676511486911367000625355); DVK(KP881921264, +0.881921264348355029712756863660388349508442621); DVK(KP303346683, +0.303346683607342391675883946941299872384187453); DVK(KP534511135, +0.534511135950791641089685961295362908582039528); DVK(KP773010453, +0.773010453362736960810906609758469800971041293); DVK(KP995184726, +0.995184726672196886244836953109479921575474869); DVK(KP820678790, +0.820678790828660330972281985331011598767386482); DVK(KP098491403, +0.098491403357164253077197521291327432293052451); DVK(KP831469612, +0.831469612302545237078788377617905756738560812); DVK(KP980785280, +0.980785280403230449126182236134239036973933731); DVK(KP198912367, +0.198912367379658006911597622644676228597850501); DVK(KP668178637, +0.668178637919298919997757686523080761552472251); DVK(KP923879532, +0.923879532511286756128183189396788286822416626); DVK(KP414213562, +0.414213562373095048801688724209698078569671875); DVK(KP707106781, +0.707106781186547524400844362104849039284835938); { INT i; const R *xi; R *xo; xi = ri; xo = ro; for (i = v; i > 0; i = i - VL, xi = xi + (VL * ivs), xo = xo + (VL * ovs), MAKE_VOLATILE_STRIDE(256, is), MAKE_VOLATILE_STRIDE(256, os)) { V T6a, T5J, T6b, T5K, T6B, T6C, T6J, T6A, T6o, T6j, T6r, T68, T6e, T5O, T5R; V T6d, T6D, T6K; { V Tad, TcZ, T6Z, T8T, T4U, Tr, Tfq, TgG, Ted, Tgf, Td0, Tcc, T9k, T84, Tb6; V Tbt, Td8, TdK, TeK, Tgq, TeV, Tgt, T7q, T94, T3p, T5X, T7B, T97, T2G, T5U; V TbD, Tc0, Tdf, TdN, Tf5, Tgx, Tfg, TgA, T7J, T9b, T4E, T64, T7U, T9e, T3V; V T61, Td2, Td3, T85, T72, T4V, TI, Tcd, Tas, TgH, Tek, Tgg, Tft, T86, T75; V T4W, TZ, TaI, Tcg, Tdr, TdG, Tgi, Tet, Tgj, Teq, T8X, T7a, T5M, T1B, T8W; V T7d, T5N, T1s, TaX, Tcf, Tdo, TdH, Tgl, TeC, Tgm, Tez, T90, T7h, T5P, T2c; V T8Z, T7k, T5Q, T23, T3Y, T49, TdL, Tdb, Tbu, Tbl, Tgu, TeR, Tgr, TeY, Tf6; V TbG, T5V, T3s, T5Y, T3f, T95, T7E, T98, T7x, T4g, T4f, T4q, TbH, T41, TbI; V T44, T4h, T4j, T4k, Tf9, TbN; { V Tu, TF, Ty, TL, TW, Tah, Tx, Tag, Tee, Tz, TM, TN, Teh, Tan, TP; V TQ; { V TeG, T2A, Tbq, TeT, Tbp, TeH, T3m, T2x, Td6, T7o, T2q, T3l, T7z, Tbr, T2D; V T82, T83; { V Ta7, T3, Ta8, T4O, Taa, Tab, Ta, T4P, Te, Tc9, Th, Tca, Tl, Tc6, Tc7; V To; { V T1, T2, T4M, T4N; T1 = LD(&(xi[0]), ivs, &(xi[0])); T2 = LD(&(xi[WS(is, 64)]), ivs, &(xi[0])); T4M = LD(&(xi[WS(is, 32)]), ivs, &(xi[0])); T4N = LD(&(xi[WS(is, 96)]), ivs, &(xi[0])); { V T4, T5, T7, T8; T4 = LD(&(xi[WS(is, 16)]), ivs, &(xi[0])); T5 = LD(&(xi[WS(is, 80)]), ivs, &(xi[0])); T7 = LD(&(xi[WS(is, 112)]), ivs, &(xi[0])); T8 = LD(&(xi[WS(is, 48)]), ivs, &(xi[0])); { V Tc, T6, T9, Td, Tf, Tg; Tc = LD(&(xi[WS(is, 8)]), ivs, &(xi[0])); Ta7 = VADD(T1, T2); T3 = VSUB(T1, T2); Ta8 = VADD(T4M, T4N); T4O = VSUB(T4M, T4N); Taa = VADD(T4, T5); T6 = VSUB(T4, T5); Tab = VADD(T7, T8); T9 = VSUB(T7, T8); Td = LD(&(xi[WS(is, 72)]), ivs, &(xi[0])); Tf = LD(&(xi[WS(is, 40)]), ivs, &(xi[0])); Tg = LD(&(xi[WS(is, 104)]), ivs, &(xi[0])); { V Tj, Tk, Tm, Tn; Tj = LD(&(xi[WS(is, 120)]), ivs, &(xi[0])); Tk = LD(&(xi[WS(is, 56)]), ivs, &(xi[0])); Tm = LD(&(xi[WS(is, 88)]), ivs, &(xi[0])); Tn = LD(&(xi[WS(is, 24)]), ivs, &(xi[0])); Ta = VADD(T6, T9); T4P = VSUB(T9, T6); Te = VSUB(Tc, Td); Tc9 = VADD(Tc, Td); Th = VSUB(Tf, Tg); Tca = VADD(Tf, Tg); Tl = VSUB(Tj, Tk); Tc6 = VADD(Tj, Tk); Tc7 = VADD(Tn, Tm); To = VSUB(Tm, Tn); } } } } { V T6X, Tb, Te9, Ta9, Tcb, Tea, T4R, Ti, Tfo, Tac, Tp, T4S, Tc8, Teb, T4Q; T6X = VFNMS(LDK(KP707106781), Ta, T3); Tb = VFMA(LDK(KP707106781), Ta, T3); Te9 = VSUB(Ta7, Ta8); Ta9 = VADD(Ta7, Ta8); Tcb = VADD(Tc9, Tca); Tea = VSUB(Tc9, Tca); T4R = VFMA(LDK(KP414213562), Te, Th); Ti = VFNMS(LDK(KP414213562), Th, Te); Tfo = VSUB(Tab, Taa); Tac = VADD(Taa, Tab); Tp = VFNMS(LDK(KP414213562), To, Tl); T4S = VFMA(LDK(KP414213562), Tl, To); Tc8 = VADD(Tc6, Tc7); Teb = VSUB(Tc6, Tc7); T4Q = VFNMS(LDK(KP707106781), T4P, T4O); T82 = VFMA(LDK(KP707106781), T4P, T4O); { V T4T, T6Y, Tq, Tfp, Tec; T4T = VSUB(T4R, T4S); T6Y = VADD(T4R, T4S); T83 = VSUB(Tp, Ti); Tq = VADD(Ti, Tp); Tfp = VSUB(Teb, Tea); Tec = VADD(Tea, Teb); Tad = VSUB(Ta9, Tac); TcZ = VADD(Ta9, Tac); T6Z = VFMA(LDK(KP923879532), T6Y, T6X); T8T = VFNMS(LDK(KP923879532), T6Y, T6X); T4U = VFMA(LDK(KP923879532), T4T, T4Q); T6a = VFNMS(LDK(KP923879532), T4T, T4Q); Tr = VFMA(LDK(KP923879532), Tq, Tb); T5J = VFNMS(LDK(KP923879532), Tq, Tb); Tfq = VFMA(LDK(KP707106781), Tfp, Tfo); TgG = VFNMS(LDK(KP707106781), Tfp, Tfo); Ted = VFMA(LDK(KP707106781), Tec, Te9); Tgf = VFNMS(LDK(KP707106781), Tec, Te9); Td0 = VADD(Tcb, Tc8); Tcc = VSUB(Tc8, Tcb); } } } { V T2i, T3j, Tb2, T2B, Tb5, T3k, T2p, T2C; { V T2m, Tb0, Tb1, Tb3, T2l, T2n; { V T2g, T2h, T3h, T3i, T2j, T2k; T2g = LD(&(xi[WS(is, 1)]), ivs, &(xi[WS(is, 1)])); T2h = LD(&(xi[WS(is, 65)]), ivs, &(xi[WS(is, 1)])); T3h = LD(&(xi[WS(is, 33)]), ivs, &(xi[WS(is, 1)])); T3i = LD(&(xi[WS(is, 97)]), ivs, &(xi[WS(is, 1)])); T2j = LD(&(xi[WS(is, 17)]), ivs, &(xi[WS(is, 1)])); T2k = LD(&(xi[WS(is, 81)]), ivs, &(xi[WS(is, 1)])); T2m = LD(&(xi[WS(is, 113)]), ivs, &(xi[WS(is, 1)])); T9k = VFNMS(LDK(KP923879532), T83, T82); T84 = VFMA(LDK(KP923879532), T83, T82); T2i = VSUB(T2g, T2h); Tb0 = VADD(T2g, T2h); T3j = VSUB(T3h, T3i); Tb1 = VADD(T3h, T3i); Tb3 = VADD(T2j, T2k); T2l = VSUB(T2j, T2k); T2n = LD(&(xi[WS(is, 49)]), ivs, &(xi[WS(is, 1)])); } { V T2r, T2s, T2u, T2v; T2r = LD(&(xi[WS(is, 9)]), ivs, &(xi[WS(is, 1)])); T2s = LD(&(xi[WS(is, 73)]), ivs, &(xi[WS(is, 1)])); T2u = LD(&(xi[WS(is, 41)]), ivs, &(xi[WS(is, 1)])); T2v = LD(&(xi[WS(is, 105)]), ivs, &(xi[WS(is, 1)])); TeG = VSUB(Tb0, Tb1); Tb2 = VADD(Tb0, Tb1); { V T2y, T2z, Tb4, T2o, Tbn, T2t, Tbo, T2w; T2y = LD(&(xi[WS(is, 121)]), ivs, &(xi[WS(is, 1)])); T2z = LD(&(xi[WS(is, 57)]), ivs, &(xi[WS(is, 1)])); Tb4 = VADD(T2m, T2n); T2o = VSUB(T2m, T2n); Tbn = VADD(T2r, T2s); T2t = VSUB(T2r, T2s); Tbo = VADD(T2u, T2v); T2w = VSUB(T2u, T2v); T2B = LD(&(xi[WS(is, 25)]), ivs, &(xi[WS(is, 1)])); T2A = VSUB(T2y, T2z); Tbq = VADD(T2y, T2z); TeT = VSUB(Tb3, Tb4); Tb5 = VADD(Tb3, Tb4); T3k = VSUB(T2l, T2o); T2p = VADD(T2l, T2o); Tbp = VADD(Tbn, Tbo); TeH = VSUB(Tbn, Tbo); T3m = VFMA(LDK(KP414213562), T2t, T2w); T2x = VFNMS(LDK(KP414213562), T2w, T2t); T2C = LD(&(xi[WS(is, 89)]), ivs, &(xi[WS(is, 1)])); } } } Td6 = VADD(Tb2, Tb5); Tb6 = VSUB(Tb2, Tb5); T7o = VFNMS(LDK(KP707106781), T2p, T2i); T2q = VFMA(LDK(KP707106781), T2p, T2i); T3l = VFMA(LDK(KP707106781), T3k, T3j); T7z = VFNMS(LDK(KP707106781), T3k, T3j); Tbr = VADD(T2B, T2C); T2D = VSUB(T2B, T2C); } { V Tf1, Tfe, Tf2, TbZ, T3M, T4B, Tdd, T3F, T7H, T4A, T7S, TbW, Tf3, T4C, T3T; { V T3x, T4y, Tbz, T3Q, TbC, T4z, T3E, T3R, T3P, TbU, TbV, T3S; { V T3y, T3z, T3B, T3C; { V T3v, T3w, T4w, T4x; T3v = LD(&(xi[WS(is, 127)]), ivs, &(xi[WS(is, 1)])); T3w = LD(&(xi[WS(is, 63)]), ivs, &(xi[WS(is, 1)])); T4w = LD(&(xi[WS(is, 95)]), ivs, &(xi[WS(is, 1)])); T4x = LD(&(xi[WS(is, 31)]), ivs, &(xi[WS(is, 1)])); T3y = LD(&(xi[WS(is, 15)]), ivs, &(xi[WS(is, 1)])); { V Tbs, TeI, T3n, T2E, Tbx; Tbs = VADD(Tbq, Tbr); TeI = VSUB(Tbq, Tbr); T3n = VFNMS(LDK(KP414213562), T2A, T2D); T2E = VFMA(LDK(KP414213562), T2D, T2A); T3x = VSUB(T3v, T3w); Tbx = VADD(T3v, T3w); { V Tby, Td7, TeJ, TeU; T4y = VSUB(T4w, T4x); Tby = VADD(T4x, T4w); Td7 = VADD(Tbp, Tbs); Tbt = VSUB(Tbp, Tbs); TeJ = VADD(TeH, TeI); TeU = VSUB(TeH, TeI); { V T7p, T3o, T7A, T2F; T7p = VSUB(T3m, T3n); T3o = VADD(T3m, T3n); T7A = VSUB(T2x, T2E); T2F = VADD(T2x, T2E); Tbz = VADD(Tbx, Tby); Tf1 = VSUB(Tbx, Tby); Td8 = VADD(Td6, Td7); TdK = VSUB(Td6, Td7); TeK = VFMA(LDK(KP707106781), TeJ, TeG); Tgq = VFNMS(LDK(KP707106781), TeJ, TeG); TeV = VFMA(LDK(KP707106781), TeU, TeT); Tgt = VFNMS(LDK(KP707106781), TeU, TeT); T7q = VFMA(LDK(KP923879532), T7p, T7o); T94 = VFNMS(LDK(KP923879532), T7p, T7o); T3p = VFMA(LDK(KP923879532), T3o, T3l); T5X = VFNMS(LDK(KP923879532), T3o, T3l); T7B = VFNMS(LDK(KP923879532), T7A, T7z); T97 = VFMA(LDK(KP923879532), T7A, T7z); T2G = VFMA(LDK(KP923879532), T2F, T2q); T5U = VFNMS(LDK(KP923879532), T2F, T2q); T3z = LD(&(xi[WS(is, 79)]), ivs, &(xi[WS(is, 1)])); } } } T3B = LD(&(xi[WS(is, 111)]), ivs, &(xi[WS(is, 1)])); T3C = LD(&(xi[WS(is, 47)]), ivs, &(xi[WS(is, 1)])); } { V T3G, T3H, T3J, T3K; T3G = LD(&(xi[WS(is, 7)]), ivs, &(xi[WS(is, 1)])); T3H = LD(&(xi[WS(is, 71)]), ivs, &(xi[WS(is, 1)])); T3J = LD(&(xi[WS(is, 103)]), ivs, &(xi[WS(is, 1)])); T3K = LD(&(xi[WS(is, 39)]), ivs, &(xi[WS(is, 1)])); { V T3N, T3A, TbA, T3D, TbB, T3I, TbX, T3L, TbY, T3O; T3N = LD(&(xi[WS(is, 119)]), ivs, &(xi[WS(is, 1)])); T3A = VSUB(T3y, T3z); TbA = VADD(T3y, T3z); T3D = VSUB(T3B, T3C); TbB = VADD(T3B, T3C); T3I = VSUB(T3G, T3H); TbX = VADD(T3G, T3H); T3L = VSUB(T3J, T3K); TbY = VADD(T3K, T3J); T3O = LD(&(xi[WS(is, 55)]), ivs, &(xi[WS(is, 1)])); T3Q = LD(&(xi[WS(is, 87)]), ivs, &(xi[WS(is, 1)])); Tfe = VSUB(TbB, TbA); TbC = VADD(TbA, TbB); T4z = VSUB(T3D, T3A); T3E = VADD(T3A, T3D); T3R = LD(&(xi[WS(is, 23)]), ivs, &(xi[WS(is, 1)])); Tf2 = VSUB(TbX, TbY); TbZ = VADD(TbX, TbY); T3M = VFMA(LDK(KP414213562), T3L, T3I); T4B = VFNMS(LDK(KP414213562), T3I, T3L); T3P = VSUB(T3N, T3O); TbU = VADD(T3N, T3O); } } } Tdd = VADD(Tbz, TbC); TbD = VSUB(Tbz, TbC); TbV = VADD(T3R, T3Q); T3S = VSUB(T3Q, T3R); T3F = VFMA(LDK(KP707106781), T3E, T3x); T7H = VFNMS(LDK(KP707106781), T3E, T3x); T4A = VFMA(LDK(KP707106781), T4z, T4y); T7S = VFNMS(LDK(KP707106781), T4z, T4y); TbW = VADD(TbU, TbV); Tf3 = VSUB(TbU, TbV); T4C = VFMA(LDK(KP414213562), T3P, T3S); T3T = VFNMS(LDK(KP414213562), T3S, T3P); } { V TD, Tae, TE, TJ, TK, TU, TV; { V Ts, Tt, Tde, Tf4, Tff; Ts = LD(&(xi[WS(is, 4)]), ivs, &(xi[0])); Tt = LD(&(xi[WS(is, 68)]), ivs, &(xi[0])); TD = LD(&(xi[WS(is, 36)]), ivs, &(xi[0])); Tde = VADD(TbZ, TbW); Tc0 = VSUB(TbW, TbZ); Tf4 = VADD(Tf2, Tf3); Tff = VSUB(Tf3, Tf2); { V T7I, T4D, T7T, T3U; T7I = VSUB(T4C, T4B); T4D = VADD(T4B, T4C); T7T = VSUB(T3T, T3M); T3U = VADD(T3M, T3T); Tae = VADD(Ts, Tt); Tu = VSUB(Ts, Tt); Tdf = VADD(Tdd, Tde); TdN = VSUB(Tdd, Tde); Tf5 = VFMA(LDK(KP707106781), Tf4, Tf1); Tgx = VFNMS(LDK(KP707106781), Tf4, Tf1); Tfg = VFMA(LDK(KP707106781), Tff, Tfe); TgA = VFNMS(LDK(KP707106781), Tff, Tfe); T7J = VFMA(LDK(KP923879532), T7I, T7H); T9b = VFNMS(LDK(KP923879532), T7I, T7H); T4E = VFMA(LDK(KP923879532), T4D, T4A); T64 = VFNMS(LDK(KP923879532), T4D, T4A); T7U = VFNMS(LDK(KP923879532), T7T, T7S); T9e = VFMA(LDK(KP923879532), T7T, T7S); T3V = VFMA(LDK(KP923879532), T3U, T3F); T61 = VFNMS(LDK(KP923879532), T3U, T3F); TE = LD(&(xi[WS(is, 100)]), ivs, &(xi[0])); } } TJ = LD(&(xi[WS(is, 124)]), ivs, &(xi[0])); TK = LD(&(xi[WS(is, 60)]), ivs, &(xi[0])); TU = LD(&(xi[WS(is, 92)]), ivs, &(xi[0])); TV = LD(&(xi[WS(is, 28)]), ivs, &(xi[0])); { V Tal, Tam, Tv, Tw, Taf; Tv = LD(&(xi[WS(is, 20)]), ivs, &(xi[0])); Tw = LD(&(xi[WS(is, 84)]), ivs, &(xi[0])); Taf = VADD(TD, TE); TF = VSUB(TD, TE); Ty = LD(&(xi[WS(is, 116)]), ivs, &(xi[0])); TL = VSUB(TJ, TK); Tal = VADD(TJ, TK); TW = VSUB(TU, TV); Tam = VADD(TV, TU); Tah = VADD(Tv, Tw); Tx = VSUB(Tv, Tw); Tag = VADD(Tae, Taf); Tee = VSUB(Tae, Taf); Tz = LD(&(xi[WS(is, 52)]), ivs, &(xi[0])); TM = LD(&(xi[WS(is, 12)]), ivs, &(xi[0])); TN = LD(&(xi[WS(is, 76)]), ivs, &(xi[0])); Teh = VSUB(Tal, Tam); Tan = VADD(Tal, Tam); TP = LD(&(xi[WS(is, 108)]), ivs, &(xi[0])); TQ = LD(&(xi[WS(is, 44)]), ivs, &(xi[0])); } } } } { V Tev, TeA, Tdm, TaP, Tew, TaV, T1U, T29, T7f, T1N, T28, T7i, Tex, TaS, T21; V T2a; { V Tem, Ter, Ten, TaD, T1j, T1y, TaA, Tdp, T1c, T78, T7b, T1x, TaG, Teo, T1z; V T1q; { V T14, T1v, Taw, Taz, T1b, T1w, T1n, T1o, T1m, TaE, TaF, T1p; { V Tau, Tav, T15, T16, T18, T19; { V T12, Tai, TA, Tao, TO, T13; T12 = LD(&(xi[WS(is, 2)]), ivs, &(xi[0])); Tai = VADD(Ty, Tz); TA = VSUB(Ty, Tz); Tao = VADD(TM, TN); TO = VSUB(TM, TN); T13 = LD(&(xi[WS(is, 66)]), ivs, &(xi[0])); { V T1t, Tap, TR, Taj, Tef, TG, TB, T1u; T1t = LD(&(xi[WS(is, 34)]), ivs, &(xi[0])); Tap = VADD(TP, TQ); TR = VSUB(TP, TQ); Taj = VADD(Tah, Tai); Tef = VSUB(Tah, Tai); TG = VSUB(Tx, TA); TB = VADD(Tx, TA); Tau = VADD(T12, T13); T14 = VSUB(T12, T13); T1u = LD(&(xi[WS(is, 98)]), ivs, &(xi[0])); { V Taq, Tei, TX, TS, Tak; Taq = VADD(Tao, Tap); Tei = VSUB(Tap, Tao); TX = VSUB(TR, TO); TS = VADD(TO, TR); Tak = VSUB(Tag, Taj); Td2 = VADD(Tag, Taj); { V Teg, Tfs, T71, TH; Teg = VFNMS(LDK(KP414213562), Tef, Tee); Tfs = VFMA(LDK(KP414213562), Tee, Tef); T71 = VFNMS(LDK(KP707106781), TG, TF); TH = VFMA(LDK(KP707106781), TG, TF); { V T70, TC, Tar, Tej, Tfr; T70 = VFNMS(LDK(KP707106781), TB, Tu); TC = VFMA(LDK(KP707106781), TB, Tu); Tar = VSUB(Tan, Taq); Td3 = VADD(Tan, Taq); Tej = VFNMS(LDK(KP414213562), Tei, Teh); Tfr = VFMA(LDK(KP414213562), Teh, Tei); { V T74, TY, T73, TT; T74 = VFNMS(LDK(KP707106781), TX, TW); TY = VFMA(LDK(KP707106781), TX, TW); T73 = VFNMS(LDK(KP707106781), TS, TL); TT = VFMA(LDK(KP707106781), TS, TL); T85 = VFNMS(LDK(KP668178637), T70, T71); T72 = VFMA(LDK(KP668178637), T71, T70); T4V = VFMA(LDK(KP198912367), TC, TH); TI = VFNMS(LDK(KP198912367), TH, TC); Tcd = VSUB(Tar, Tak); Tas = VADD(Tak, Tar); TgH = VSUB(Tej, Teg); Tek = VADD(Teg, Tej); Tgg = VADD(Tfs, Tfr); Tft = VSUB(Tfr, Tfs); T86 = VFNMS(LDK(KP668178637), T73, T74); T75 = VFMA(LDK(KP668178637), T74, T73); T4W = VFMA(LDK(KP198912367), TT, TY); TZ = VFNMS(LDK(KP198912367), TY, TT); Tav = VADD(T1t, T1u); T1v = VSUB(T1t, T1u); } } } } } } T15 = LD(&(xi[WS(is, 18)]), ivs, &(xi[0])); T16 = LD(&(xi[WS(is, 82)]), ivs, &(xi[0])); T18 = LD(&(xi[WS(is, 114)]), ivs, &(xi[0])); T19 = LD(&(xi[WS(is, 50)]), ivs, &(xi[0])); { V T1d, T1e, T1g, T1h, Tax, T17, Tay, T1a; T1d = LD(&(xi[WS(is, 10)]), ivs, &(xi[0])); Taw = VADD(Tau, Tav); Tem = VSUB(Tau, Tav); T1e = LD(&(xi[WS(is, 74)]), ivs, &(xi[0])); T1g = LD(&(xi[WS(is, 42)]), ivs, &(xi[0])); T1h = LD(&(xi[WS(is, 106)]), ivs, &(xi[0])); Tax = VADD(T15, T16); T17 = VSUB(T15, T16); Tay = VADD(T18, T19); T1a = VSUB(T18, T19); { V T1k, T1f, TaB, T1i, TaC, T1l; T1k = LD(&(xi[WS(is, 122)]), ivs, &(xi[0])); T1f = VSUB(T1d, T1e); TaB = VADD(T1d, T1e); T1i = VSUB(T1g, T1h); TaC = VADD(T1g, T1h); T1l = LD(&(xi[WS(is, 58)]), ivs, &(xi[0])); Taz = VADD(Tax, Tay); Ter = VSUB(Tax, Tay); T1b = VADD(T17, T1a); T1w = VSUB(T17, T1a); T1n = LD(&(xi[WS(is, 26)]), ivs, &(xi[0])); T1o = LD(&(xi[WS(is, 90)]), ivs, &(xi[0])); Ten = VSUB(TaB, TaC); TaD = VADD(TaB, TaC); T1j = VFNMS(LDK(KP414213562), T1i, T1f); T1y = VFMA(LDK(KP414213562), T1f, T1i); T1m = VSUB(T1k, T1l); TaE = VADD(T1k, T1l); } } } TaA = VSUB(Taw, Taz); Tdp = VADD(Taw, Taz); TaF = VADD(T1n, T1o); T1p = VSUB(T1n, T1o); T1c = VFMA(LDK(KP707106781), T1b, T14); T78 = VFNMS(LDK(KP707106781), T1b, T14); T7b = VFNMS(LDK(KP707106781), T1w, T1v); T1x = VFMA(LDK(KP707106781), T1w, T1v); TaG = VADD(TaE, TaF); Teo = VSUB(TaE, TaF); T1z = VFNMS(LDK(KP414213562), T1m, T1p); T1q = VFMA(LDK(KP414213562), T1p, T1m); } { V T1F, T26, T1Q, TaT, TaL, TaO, T27, T1M, T1Y, T1Z, TaU, T1T, TaQ, T1X, T20; V TaR; { V T24, TaJ, T25, T1G, T1H, T1J, T1K, T1D, T1E; T1D = LD(&(xi[WS(is, 126)]), ivs, &(xi[0])); T1E = LD(&(xi[WS(is, 62)]), ivs, &(xi[0])); T24 = LD(&(xi[WS(is, 94)]), ivs, &(xi[0])); { V TaH, Tdq, Tes, Tep; TaH = VSUB(TaD, TaG); Tdq = VADD(TaD, TaG); Tes = VSUB(Ten, Teo); Tep = VADD(Ten, Teo); { V T79, T1A, T7c, T1r; T79 = VSUB(T1y, T1z); T1A = VADD(T1y, T1z); T7c = VSUB(T1j, T1q); T1r = VADD(T1j, T1q); TaJ = VADD(T1D, T1E); T1F = VSUB(T1D, T1E); TaI = VFNMS(LDK(KP414213562), TaH, TaA); Tcg = VFMA(LDK(KP414213562), TaA, TaH); Tdr = VADD(Tdp, Tdq); TdG = VSUB(Tdp, Tdq); Tgi = VFNMS(LDK(KP707106781), Tes, Ter); Tet = VFMA(LDK(KP707106781), Tes, Ter); Tgj = VFNMS(LDK(KP707106781), Tep, Tem); Teq = VFMA(LDK(KP707106781), Tep, Tem); T8X = VFNMS(LDK(KP923879532), T79, T78); T7a = VFMA(LDK(KP923879532), T79, T78); T5M = VFNMS(LDK(KP923879532), T1A, T1x); T1B = VFMA(LDK(KP923879532), T1A, T1x); T8W = VFMA(LDK(KP923879532), T7c, T7b); T7d = VFNMS(LDK(KP923879532), T7c, T7b); T5N = VFNMS(LDK(KP923879532), T1r, T1c); T1s = VFMA(LDK(KP923879532), T1r, T1c); T25 = LD(&(xi[WS(is, 30)]), ivs, &(xi[0])); } } T1G = LD(&(xi[WS(is, 14)]), ivs, &(xi[0])); T1H = LD(&(xi[WS(is, 78)]), ivs, &(xi[0])); T1J = LD(&(xi[WS(is, 110)]), ivs, &(xi[0])); T1K = LD(&(xi[WS(is, 46)]), ivs, &(xi[0])); { V T1R, T1I, TaM, T1L, TaN, T1S, T1O, T1P, TaK, T1V, T1W; T1O = LD(&(xi[WS(is, 6)]), ivs, &(xi[0])); T1P = LD(&(xi[WS(is, 70)]), ivs, &(xi[0])); T26 = VSUB(T24, T25); TaK = VADD(T25, T24); T1R = LD(&(xi[WS(is, 102)]), ivs, &(xi[0])); T1I = VSUB(T1G, T1H); TaM = VADD(T1G, T1H); T1L = VSUB(T1J, T1K); TaN = VADD(T1J, T1K); T1Q = VSUB(T1O, T1P); TaT = VADD(T1O, T1P); Tev = VSUB(TaJ, TaK); TaL = VADD(TaJ, TaK); T1S = LD(&(xi[WS(is, 38)]), ivs, &(xi[0])); T1V = LD(&(xi[WS(is, 118)]), ivs, &(xi[0])); T1W = LD(&(xi[WS(is, 54)]), ivs, &(xi[0])); TeA = VSUB(TaN, TaM); TaO = VADD(TaM, TaN); T27 = VSUB(T1L, T1I); T1M = VADD(T1I, T1L); T1Y = LD(&(xi[WS(is, 86)]), ivs, &(xi[0])); T1Z = LD(&(xi[WS(is, 22)]), ivs, &(xi[0])); TaU = VADD(T1S, T1R); T1T = VSUB(T1R, T1S); TaQ = VADD(T1V, T1W); T1X = VSUB(T1V, T1W); } } Tdm = VADD(TaL, TaO); TaP = VSUB(TaL, TaO); T20 = VSUB(T1Y, T1Z); TaR = VADD(T1Z, T1Y); Tew = VSUB(TaT, TaU); TaV = VADD(TaT, TaU); T1U = VFMA(LDK(KP414213562), T1T, T1Q); T29 = VFNMS(LDK(KP414213562), T1Q, T1T); T7f = VFNMS(LDK(KP707106781), T1M, T1F); T1N = VFMA(LDK(KP707106781), T1M, T1F); T28 = VFMA(LDK(KP707106781), T27, T26); T7i = VFNMS(LDK(KP707106781), T27, T26); Tex = VSUB(TaQ, TaR); TaS = VADD(TaQ, TaR); T21 = VFNMS(LDK(KP414213562), T20, T1X); T2a = VFMA(LDK(KP414213562), T1X, T20); } } { V T2J, T2U, T30, T3b, TeL, Tb9, TeO, Tbg, T2M, Tba, T2P, Tbb, T34, Tbh, T33; V T35; { V T2H, T2I, T2S, T2T, T2Y, T2Z, T39, T3a; T2H = LD(&(xi[WS(is, 5)]), ivs, &(xi[WS(is, 1)])); { V Tdn, TaW, Tey, TeB; Tdn = VADD(TaV, TaS); TaW = VSUB(TaS, TaV); Tey = VADD(Tew, Tex); TeB = VSUB(Tex, Tew); { V T2b, T7g, T22, T7j; T2b = VADD(T29, T2a); T7g = VSUB(T2a, T29); T22 = VADD(T1U, T21); T7j = VSUB(T21, T1U); TaX = VFNMS(LDK(KP414213562), TaW, TaP); Tcf = VFMA(LDK(KP414213562), TaP, TaW); Tdo = VADD(Tdm, Tdn); TdH = VSUB(Tdm, Tdn); Tgl = VFNMS(LDK(KP707106781), TeB, TeA); TeC = VFMA(LDK(KP707106781), TeB, TeA); Tgm = VFNMS(LDK(KP707106781), Tey, Tev); Tez = VFMA(LDK(KP707106781), Tey, Tev); T90 = VFNMS(LDK(KP923879532), T7g, T7f); T7h = VFMA(LDK(KP923879532), T7g, T7f); T5P = VFNMS(LDK(KP923879532), T2b, T28); T2c = VFMA(LDK(KP923879532), T2b, T28); T8Z = VFMA(LDK(KP923879532), T7j, T7i); T7k = VFNMS(LDK(KP923879532), T7j, T7i); T5Q = VFNMS(LDK(KP923879532), T22, T1N); T23 = VFMA(LDK(KP923879532), T22, T1N); T2I = LD(&(xi[WS(is, 69)]), ivs, &(xi[WS(is, 1)])); } } T2S = LD(&(xi[WS(is, 37)]), ivs, &(xi[WS(is, 1)])); T2T = LD(&(xi[WS(is, 101)]), ivs, &(xi[WS(is, 1)])); T2Y = LD(&(xi[WS(is, 125)]), ivs, &(xi[WS(is, 1)])); T2Z = LD(&(xi[WS(is, 61)]), ivs, &(xi[WS(is, 1)])); T39 = LD(&(xi[WS(is, 29)]), ivs, &(xi[WS(is, 1)])); T3a = LD(&(xi[WS(is, 93)]), ivs, &(xi[WS(is, 1)])); { V T2K, Tbe, Tbf, T2L, T2N, T2O, Tb7, Tb8, T31, T32; T2K = LD(&(xi[WS(is, 21)]), ivs, &(xi[WS(is, 1)])); T2J = VSUB(T2H, T2I); Tb7 = VADD(T2H, T2I); T2U = VSUB(T2S, T2T); Tb8 = VADD(T2S, T2T); T30 = VSUB(T2Y, T2Z); Tbe = VADD(T2Y, T2Z); T3b = VSUB(T39, T3a); Tbf = VADD(T39, T3a); T2L = LD(&(xi[WS(is, 85)]), ivs, &(xi[WS(is, 1)])); T2N = LD(&(xi[WS(is, 117)]), ivs, &(xi[WS(is, 1)])); T2O = LD(&(xi[WS(is, 53)]), ivs, &(xi[WS(is, 1)])); TeL = VSUB(Tb7, Tb8); Tb9 = VADD(Tb7, Tb8); T31 = LD(&(xi[WS(is, 13)]), ivs, &(xi[WS(is, 1)])); T32 = LD(&(xi[WS(is, 77)]), ivs, &(xi[WS(is, 1)])); TeO = VSUB(Tbe, Tbf); Tbg = VADD(Tbe, Tbf); T2M = VSUB(T2K, T2L); Tba = VADD(T2K, T2L); T2P = VSUB(T2N, T2O); Tbb = VADD(T2N, T2O); T34 = LD(&(xi[WS(is, 109)]), ivs, &(xi[WS(is, 1)])); Tbh = VADD(T31, T32); T33 = VSUB(T31, T32); T35 = LD(&(xi[WS(is, 45)]), ivs, &(xi[WS(is, 1)])); } } { V T4d, T4e, T4o, T4p; { V T2X, T3q, T7t, T7C, T3r, T3e, T7D, T7w; { V T47, TbE, Tbd, Td9, TeW, TeN, T7s, T2W, T7r, T2R, TeP, Tbj, T37, T3c, T48; { V T3W, T3X, TeM, Tbc, T2Q, T2V, Tbi, T36; T3W = LD(&(xi[WS(is, 3)]), ivs, &(xi[WS(is, 1)])); T3X = LD(&(xi[WS(is, 67)]), ivs, &(xi[WS(is, 1)])); TeM = VSUB(Tba, Tbb); Tbc = VADD(Tba, Tbb); T2Q = VADD(T2M, T2P); T2V = VSUB(T2M, T2P); T47 = LD(&(xi[WS(is, 99)]), ivs, &(xi[WS(is, 1)])); Tbi = VADD(T34, T35); T36 = VSUB(T34, T35); TbE = VADD(T3W, T3X); T3Y = VSUB(T3W, T3X); Tbd = VSUB(Tb9, Tbc); Td9 = VADD(Tb9, Tbc); TeW = VFMA(LDK(KP414213562), TeL, TeM); TeN = VFNMS(LDK(KP414213562), TeM, TeL); T7s = VFNMS(LDK(KP707106781), T2V, T2U); T2W = VFMA(LDK(KP707106781), T2V, T2U); T7r = VFNMS(LDK(KP707106781), T2Q, T2J); T2R = VFMA(LDK(KP707106781), T2Q, T2J); TeP = VSUB(Tbh, Tbi); Tbj = VADD(Tbh, Tbi); T37 = VADD(T33, T36); T3c = VSUB(T33, T36); T48 = LD(&(xi[WS(is, 35)]), ivs, &(xi[WS(is, 1)])); } T2X = VFNMS(LDK(KP198912367), T2W, T2R); T3q = VFMA(LDK(KP198912367), T2R, T2W); T7t = VFMA(LDK(KP668178637), T7s, T7r); T7C = VFNMS(LDK(KP668178637), T7r, T7s); { V Tbk, Tda, TeX, TeQ; Tbk = VSUB(Tbg, Tbj); Tda = VADD(Tbg, Tbj); TeX = VFNMS(LDK(KP414213562), TeO, TeP); TeQ = VFMA(LDK(KP414213562), TeP, TeO); { V T7v, T3d, T7u, T38, TbF; T7v = VFNMS(LDK(KP707106781), T3c, T3b); T3d = VFMA(LDK(KP707106781), T3c, T3b); T7u = VFNMS(LDK(KP707106781), T37, T30); T38 = VFMA(LDK(KP707106781), T37, T30); T49 = VSUB(T47, T48); TbF = VADD(T48, T47); TdL = VSUB(Td9, Tda); Tdb = VADD(Td9, Tda); Tbu = VSUB(Tbd, Tbk); Tbl = VADD(Tbd, Tbk); Tgu = VSUB(TeN, TeQ); TeR = VADD(TeN, TeQ); Tgr = VSUB(TeW, TeX); TeY = VADD(TeW, TeX); T3r = VFNMS(LDK(KP198912367), T38, T3d); T3e = VFMA(LDK(KP198912367), T3d, T38); T7D = VFMA(LDK(KP668178637), T7u, T7v); T7w = VFNMS(LDK(KP668178637), T7v, T7u); Tf6 = VSUB(TbE, TbF); TbG = VADD(TbE, TbF); } } } T4d = LD(&(xi[WS(is, 123)]), ivs, &(xi[WS(is, 1)])); T5V = VSUB(T3q, T3r); T3s = VADD(T3q, T3r); T5Y = VSUB(T2X, T3e); T3f = VADD(T2X, T3e); T95 = VSUB(T7D, T7C); T7E = VADD(T7C, T7D); T98 = VSUB(T7t, T7w); T7x = VADD(T7t, T7w); T4e = LD(&(xi[WS(is, 59)]), ivs, &(xi[WS(is, 1)])); T4o = LD(&(xi[WS(is, 91)]), ivs, &(xi[WS(is, 1)])); T4p = LD(&(xi[WS(is, 27)]), ivs, &(xi[WS(is, 1)])); } { V T3Z, T40, T42, T43, TbL, TbM; T3Z = LD(&(xi[WS(is, 19)]), ivs, &(xi[WS(is, 1)])); T40 = LD(&(xi[WS(is, 83)]), ivs, &(xi[WS(is, 1)])); T42 = LD(&(xi[WS(is, 115)]), ivs, &(xi[WS(is, 1)])); T43 = LD(&(xi[WS(is, 51)]), ivs, &(xi[WS(is, 1)])); T4g = LD(&(xi[WS(is, 11)]), ivs, &(xi[WS(is, 1)])); T4f = VSUB(T4d, T4e); TbL = VADD(T4d, T4e); T4q = VSUB(T4o, T4p); TbM = VADD(T4p, T4o); TbH = VADD(T3Z, T40); T41 = VSUB(T3Z, T40); TbI = VADD(T42, T43); T44 = VSUB(T42, T43); T4h = LD(&(xi[WS(is, 75)]), ivs, &(xi[WS(is, 1)])); T4j = LD(&(xi[WS(is, 107)]), ivs, &(xi[WS(is, 1)])); T4k = LD(&(xi[WS(is, 43)]), ivs, &(xi[WS(is, 1)])); Tf9 = VSUB(TbL, TbM); TbN = VADD(TbL, TbM); } } } } } { V TgB, Tgy, T62, T4H, T65, T4u, T9c, T7X, T9f, T7Q, Tg0, Tga, TfF, TeF, TfT; V TfU, TfP, Tg7, TfI, Tfy, TfA, Tf0, Tfz, Tfl, Tg2, TfS; { V Tc1, TbS, Tfc, Tfj, TdX, Te5, TdZ, TdR, Te7, Te3, TdU, Te4; { V TdF, TdS, Tdx, Td5, TdO, TdE, TdC, Tdt, Tdk; { V Tdc, TdA, T4F, T4c, T7V, T7M, T4G, T4t, T7W, T7P, TdB, Tdj; { V Td1, Tdg, TbK, Tf8, Tfh, T4b, T7L, T46, T7K, TbQ, Tfa, T4r, T4m, Td4; TdF = VSUB(TcZ, Td0); Td1 = VADD(TcZ, Td0); { V TbJ, Tf7, T4a, T45; TbJ = VADD(TbH, TbI); Tf7 = VSUB(TbI, TbH); T4a = VSUB(T44, T41); T45 = VADD(T41, T44); { V TbO, T4i, TbP, T4l; TbO = VADD(T4g, T4h); T4i = VSUB(T4g, T4h); TbP = VADD(T4j, T4k); T4l = VSUB(T4j, T4k); Tdg = VADD(TbG, TbJ); TbK = VSUB(TbG, TbJ); Tf8 = VFMA(LDK(KP414213562), Tf7, Tf6); Tfh = VFNMS(LDK(KP414213562), Tf6, Tf7); T4b = VFMA(LDK(KP707106781), T4a, T49); T7L = VFNMS(LDK(KP707106781), T4a, T49); T46 = VFMA(LDK(KP707106781), T45, T3Y); T7K = VFNMS(LDK(KP707106781), T45, T3Y); TbQ = VADD(TbO, TbP); Tfa = VSUB(TbP, TbO); T4r = VSUB(T4l, T4i); T4m = VADD(T4i, T4l); Td4 = VADD(Td2, Td3); TdS = VSUB(Td3, Td2); } } Tdc = VSUB(Td8, Tdb); TdA = VADD(Td8, Tdb); T4F = VFNMS(LDK(KP198912367), T46, T4b); T4c = VFMA(LDK(KP198912367), T4b, T46); T7V = VFMA(LDK(KP668178637), T7K, T7L); T7M = VFNMS(LDK(KP668178637), T7L, T7K); { V Tdh, TbR, Tfb, Tfi; Tdh = VADD(TbN, TbQ); TbR = VSUB(TbN, TbQ); Tfb = VFNMS(LDK(KP414213562), Tfa, Tf9); Tfi = VFMA(LDK(KP414213562), Tf9, Tfa); { V T4s, T7O, T4n, T7N, Tdi; T4s = VFMA(LDK(KP707106781), T4r, T4q); T7O = VFNMS(LDK(KP707106781), T4r, T4q); T4n = VFMA(LDK(KP707106781), T4m, T4f); T7N = VFNMS(LDK(KP707106781), T4m, T4f); Tdx = VADD(Td1, Td4); Td5 = VSUB(Td1, Td4); TdO = VSUB(Tdh, Tdg); Tdi = VADD(Tdg, Tdh); Tc1 = VSUB(TbR, TbK); TbS = VADD(TbK, TbR); TgB = VSUB(Tfb, Tf8); Tfc = VADD(Tf8, Tfb); Tgy = VSUB(Tfi, Tfh); Tfj = VADD(Tfh, Tfi); T4G = VFMA(LDK(KP198912367), T4n, T4s); T4t = VFNMS(LDK(KP198912367), T4s, T4n); T7W = VFNMS(LDK(KP668178637), T7N, T7O); T7P = VFMA(LDK(KP668178637), T7O, T7N); TdB = VADD(Tdf, Tdi); Tdj = VSUB(Tdf, Tdi); } } } T62 = VSUB(T4G, T4F); T4H = VADD(T4F, T4G); T65 = VSUB(T4t, T4c); T4u = VADD(T4c, T4t); T9c = VSUB(T7V, T7W); T7X = VADD(T7V, T7W); T9f = VSUB(T7P, T7M); T7Q = VADD(T7M, T7P); TdE = VSUB(TdB, TdA); TdC = VADD(TdA, TdB); Tdt = VSUB(Tdj, Tdc); Tdk = VADD(Tdc, Tdj); } { V TdT, Tdl, Tdv, TdJ, Te1, Te2, TdQ, Tdz, TdD, Tdu, Tdw; { V TdI, TdP, TdV, TdW, TdM, Tds, Tdy; TdI = VADD(TdG, TdH); TdT = VSUB(TdH, TdG); TdP = VFNMS(LDK(KP414213562), TdO, TdN); TdV = VFMA(LDK(KP414213562), TdN, TdO); TdW = VFMA(LDK(KP414213562), TdK, TdL); TdM = VFNMS(LDK(KP414213562), TdL, TdK); Tdl = VFNMS(LDK(KP707106781), Tdk, Td5); Tdv = VFMA(LDK(KP707106781), Tdk, Td5); Tds = VSUB(Tdo, Tdr); Tdy = VADD(Tdr, Tdo); TdJ = VFMA(LDK(KP707106781), TdI, TdF); Te1 = VFNMS(LDK(KP707106781), TdI, TdF); TdX = VSUB(TdV, TdW); Te2 = VADD(TdW, TdV); Te5 = VSUB(TdP, TdM); TdQ = VADD(TdM, TdP); Tdz = VADD(Tdx, Tdy); TdD = VSUB(Tdx, Tdy); Tdu = VFNMS(LDK(KP707106781), Tdt, Tds); Tdw = VFMA(LDK(KP707106781), Tdt, Tds); } TdZ = VFMA(LDK(KP923879532), TdQ, TdJ); TdR = VFNMS(LDK(KP923879532), TdQ, TdJ); Te7 = VFMA(LDK(KP923879532), Te2, Te1); Te3 = VFNMS(LDK(KP923879532), Te2, Te1); ST(&(xo[WS(os, 32)]), VFMAI(TdE, TdD), ovs, &(xo[0])); ST(&(xo[WS(os, 96)]), VFNMSI(TdE, TdD), ovs, &(xo[0])); ST(&(xo[0]), VADD(Tdz, TdC), ovs, &(xo[0])); ST(&(xo[WS(os, 64)]), VSUB(Tdz, TdC), ovs, &(xo[0])); ST(&(xo[WS(os, 16)]), VFMAI(Tdw, Tdv), ovs, &(xo[0])); ST(&(xo[WS(os, 112)]), VFNMSI(Tdw, Tdv), ovs, &(xo[0])); ST(&(xo[WS(os, 80)]), VFMAI(Tdu, Tdl), ovs, &(xo[0])); ST(&(xo[WS(os, 48)]), VFNMSI(Tdu, Tdl), ovs, &(xo[0])); TdU = VFMA(LDK(KP707106781), TdT, TdS); Te4 = VFNMS(LDK(KP707106781), TdT, TdS); } } { V Tcx, TcJ, TcI, Tcy, TcA, Tbm, Tcp, TaZ, Tcs, Tci, Tbv, TcB, TcD, TbT, Tc2; V TcE, Tat, TaY; Tcx = VFNMS(LDK(KP707106781), Tas, Tad); Tat = VFMA(LDK(KP707106781), Tas, Tad); TaY = VADD(TaI, TaX); TcJ = VSUB(TaX, TaI); { V Tce, Tch, Te8, Te6, TdY, Te0; TcI = VFNMS(LDK(KP707106781), Tcd, Tcc); Tce = VFMA(LDK(KP707106781), Tcd, Tcc); Tch = VSUB(Tcf, Tcg); Tcy = VADD(Tcg, Tcf); Te8 = VFNMS(LDK(KP923879532), Te5, Te4); Te6 = VFMA(LDK(KP923879532), Te5, Te4); TdY = VFNMS(LDK(KP923879532), TdX, TdU); Te0 = VFMA(LDK(KP923879532), TdX, TdU); TcA = VFNMS(LDK(KP707106781), Tbl, Tb6); Tbm = VFMA(LDK(KP707106781), Tbl, Tb6); Tcp = VFNMS(LDK(KP923879532), TaY, Tat); TaZ = VFMA(LDK(KP923879532), TaY, Tat); Tcs = VFNMS(LDK(KP923879532), Tch, Tce); Tci = VFMA(LDK(KP923879532), Tch, Tce); ST(&(xo[WS(os, 88)]), VFNMSI(Te6, Te3), ovs, &(xo[0])); ST(&(xo[WS(os, 40)]), VFMAI(Te6, Te3), ovs, &(xo[0])); ST(&(xo[WS(os, 104)]), VFMAI(Te8, Te7), ovs, &(xo[0])); ST(&(xo[WS(os, 24)]), VFNMSI(Te8, Te7), ovs, &(xo[0])); ST(&(xo[WS(os, 8)]), VFMAI(Te0, TdZ), ovs, &(xo[0])); ST(&(xo[WS(os, 120)]), VFNMSI(Te0, TdZ), ovs, &(xo[0])); ST(&(xo[WS(os, 72)]), VFMAI(TdY, TdR), ovs, &(xo[0])); ST(&(xo[WS(os, 56)]), VFNMSI(TdY, TdR), ovs, &(xo[0])); Tbv = VFMA(LDK(KP707106781), Tbu, Tbt); TcB = VFNMS(LDK(KP707106781), Tbu, Tbt); TcD = VFNMS(LDK(KP707106781), TbS, TbD); TbT = VFMA(LDK(KP707106781), TbS, TbD); Tc2 = VFMA(LDK(KP707106781), Tc1, Tc0); TcE = VFNMS(LDK(KP707106781), Tc1, Tc0); } { V TcR, Tcz, TcU, TcK, Tcq, Tcl, Tct, Tc4; { V Tck, Tbw, Tcj, Tc3; Tck = VFMA(LDK(KP198912367), Tbm, Tbv); Tbw = VFNMS(LDK(KP198912367), Tbv, Tbm); Tcj = VFMA(LDK(KP198912367), TbT, Tc2); Tc3 = VFNMS(LDK(KP198912367), Tc2, TbT); TcR = VFNMS(LDK(KP923879532), Tcy, Tcx); Tcz = VFMA(LDK(KP923879532), Tcy, Tcx); TcU = VFMA(LDK(KP923879532), TcJ, TcI); TcK = VFNMS(LDK(KP923879532), TcJ, TcI); Tcq = VADD(Tck, Tcj); Tcl = VSUB(Tcj, Tck); Tct = VSUB(Tc3, Tbw); Tc4 = VADD(Tbw, Tc3); } { V TfN, Tel, TfY, Tfu, Tfw, Tfv, TcT, TcX, TcQ, TcO, TcW, TcY, TcP, TcH, TfZ; V TeE; { V Teu, TcS, TcN, TcV, TcG, TeD; TfN = VFNMS(LDK(KP923879532), Tek, Ted); Tel = VFMA(LDK(KP923879532), Tek, Ted); { V TcM, TcC, Tcr, Tcv; TcM = VFNMS(LDK(KP668178637), TcA, TcB); TcC = VFMA(LDK(KP668178637), TcB, TcA); Tcr = VFNMS(LDK(KP980785280), Tcq, Tcp); Tcv = VFMA(LDK(KP980785280), Tcq, Tcp); { V Tco, Tcm, Tcu, Tcw; Tco = VFMA(LDK(KP980785280), Tcl, Tci); Tcm = VFNMS(LDK(KP980785280), Tcl, Tci); Tcu = VFMA(LDK(KP980785280), Tct, Tcs); Tcw = VFNMS(LDK(KP980785280), Tct, Tcs); { V Tcn, Tc5, TcL, TcF; Tcn = VFMA(LDK(KP980785280), Tc4, TaZ); Tc5 = VFNMS(LDK(KP980785280), Tc4, TaZ); TcL = VFNMS(LDK(KP668178637), TcD, TcE); TcF = VFMA(LDK(KP668178637), TcE, TcD); TfY = VFNMS(LDK(KP923879532), Tft, Tfq); Tfu = VFMA(LDK(KP923879532), Tft, Tfq); Tfw = VFMA(LDK(KP198912367), Teq, Tet); Teu = VFNMS(LDK(KP198912367), Tet, Teq); ST(&(xo[WS(os, 92)]), VFNMSI(Tcu, Tcr), ovs, &(xo[0])); ST(&(xo[WS(os, 36)]), VFMAI(Tcu, Tcr), ovs, &(xo[0])); ST(&(xo[WS(os, 100)]), VFMAI(Tcw, Tcv), ovs, &(xo[0])); ST(&(xo[WS(os, 28)]), VFNMSI(Tcw, Tcv), ovs, &(xo[0])); ST(&(xo[WS(os, 4)]), VFMAI(Tco, Tcn), ovs, &(xo[0])); ST(&(xo[WS(os, 124)]), VFNMSI(Tco, Tcn), ovs, &(xo[0])); ST(&(xo[WS(os, 68)]), VFMAI(Tcm, Tc5), ovs, &(xo[0])); ST(&(xo[WS(os, 60)]), VFNMSI(Tcm, Tc5), ovs, &(xo[0])); TcS = VADD(TcM, TcL); TcN = VSUB(TcL, TcM); TcV = VSUB(TcF, TcC); TcG = VADD(TcC, TcF); TeD = VFNMS(LDK(KP198912367), TeC, Tez); Tfv = VFMA(LDK(KP198912367), Tez, TeC); } } } TcT = VFMA(LDK(KP831469612), TcS, TcR); TcX = VFNMS(LDK(KP831469612), TcS, TcR); TcQ = VFMA(LDK(KP831469612), TcN, TcK); TcO = VFNMS(LDK(KP831469612), TcN, TcK); TcW = VFNMS(LDK(KP831469612), TcV, TcU); TcY = VFMA(LDK(KP831469612), TcV, TcU); TcP = VFMA(LDK(KP831469612), TcG, Tcz); TcH = VFNMS(LDK(KP831469612), TcG, Tcz); TfZ = VSUB(TeD, Teu); TeE = VADD(Teu, TeD); } { V TfQ, TeS, TfO, Tfx, TeZ, TfR, Tfd, Tfk; TfQ = VFNMS(LDK(KP923879532), TeR, TeK); TeS = VFMA(LDK(KP923879532), TeR, TeK); ST(&(xo[WS(os, 84)]), VFMAI(TcW, TcT), ovs, &(xo[0])); ST(&(xo[WS(os, 44)]), VFNMSI(TcW, TcT), ovs, &(xo[0])); ST(&(xo[WS(os, 108)]), VFNMSI(TcY, TcX), ovs, &(xo[0])); ST(&(xo[WS(os, 20)]), VFMAI(TcY, TcX), ovs, &(xo[0])); ST(&(xo[WS(os, 116)]), VFMAI(TcQ, TcP), ovs, &(xo[0])); ST(&(xo[WS(os, 12)]), VFNMSI(TcQ, TcP), ovs, &(xo[0])); ST(&(xo[WS(os, 52)]), VFMAI(TcO, TcH), ovs, &(xo[0])); ST(&(xo[WS(os, 76)]), VFNMSI(TcO, TcH), ovs, &(xo[0])); Tg0 = VFNMS(LDK(KP980785280), TfZ, TfY); Tga = VFMA(LDK(KP980785280), TfZ, TfY); TfF = VFNMS(LDK(KP980785280), TeE, Tel); TeF = VFMA(LDK(KP980785280), TeE, Tel); TfO = VADD(Tfw, Tfv); Tfx = VSUB(Tfv, Tfw); TeZ = VFMA(LDK(KP923879532), TeY, TeV); TfR = VFNMS(LDK(KP923879532), TeY, TeV); TfT = VFNMS(LDK(KP923879532), Tfc, Tf5); Tfd = VFMA(LDK(KP923879532), Tfc, Tf5); Tfk = VFMA(LDK(KP923879532), Tfj, Tfg); TfU = VFNMS(LDK(KP923879532), Tfj, Tfg); TfP = VFMA(LDK(KP980785280), TfO, TfN); Tg7 = VFNMS(LDK(KP980785280), TfO, TfN); TfI = VFNMS(LDK(KP980785280), Tfx, Tfu); Tfy = VFMA(LDK(KP980785280), Tfx, Tfu); TfA = VFMA(LDK(KP098491403), TeS, TeZ); Tf0 = VFNMS(LDK(KP098491403), TeZ, TeS); Tfz = VFMA(LDK(KP098491403), Tfd, Tfk); Tfl = VFNMS(LDK(KP098491403), Tfk, Tfd); Tg2 = VFNMS(LDK(KP820678790), TfQ, TfR); TfS = VFMA(LDK(KP820678790), TfR, TfQ); } } } } } { V T8x, T8y, T8F, T8w, T8k, T8f, T8n, T80, T9l, T76, T87, T8U, T89, T7e, T7l; V T8a; { V The, Tho, TgT, Tgp, Th7, Th8, Thg, Th6, Th3, Thl, TgW, TgM, TgU, TgP, TgX; V TgE; { V Th1, TgI, TgK, TgJ; { V Tgh, Thc, Tgk, TfG, TfB, TfJ, Tfm, Tg1, TfV, Tgn, TfL, TfH; Th1 = VFMA(LDK(KP923879532), Tgg, Tgf); Tgh = VFNMS(LDK(KP923879532), Tgg, Tgf); Thc = VFNMS(LDK(KP923879532), TgH, TgG); TgI = VFMA(LDK(KP923879532), TgH, TgG); TgK = VFMA(LDK(KP668178637), Tgi, Tgj); Tgk = VFNMS(LDK(KP668178637), Tgj, Tgi); TfG = VADD(TfA, Tfz); TfB = VSUB(Tfz, TfA); TfJ = VSUB(Tfl, Tf0); Tfm = VADD(Tf0, Tfl); Tg1 = VFNMS(LDK(KP820678790), TfT, TfU); TfV = VFMA(LDK(KP820678790), TfU, TfT); Tgn = VFNMS(LDK(KP668178637), Tgm, Tgl); TgJ = VFMA(LDK(KP668178637), Tgl, Tgm); TfL = VFMA(LDK(KP995184726), TfG, TfF); TfH = VFNMS(LDK(KP995184726), TfG, TfF); { V TfE, TfC, TfM, TfK; TfE = VFMA(LDK(KP995184726), TfB, Tfy); TfC = VFNMS(LDK(KP995184726), TfB, Tfy); TfM = VFNMS(LDK(KP995184726), TfJ, TfI); TfK = VFMA(LDK(KP995184726), TfJ, TfI); { V TfD, Tfn, Tg8, Tg3; TfD = VFMA(LDK(KP995184726), Tfm, TeF); Tfn = VFNMS(LDK(KP995184726), Tfm, TeF); Tg8 = VADD(Tg2, Tg1); Tg3 = VSUB(Tg1, Tg2); { V Tgb, TfW, Thd, Tgo; Tgb = VSUB(TfV, TfS); TfW = VADD(TfS, TfV); Thd = VSUB(Tgn, Tgk); Tgo = VADD(Tgk, Tgn); ST(&(xo[WS(os, 98)]), VFMAI(TfM, TfL), ovs, &(xo[0])); ST(&(xo[WS(os, 30)]), VFNMSI(TfM, TfL), ovs, &(xo[0])); ST(&(xo[WS(os, 94)]), VFNMSI(TfK, TfH), ovs, &(xo[0])); ST(&(xo[WS(os, 34)]), VFMAI(TfK, TfH), ovs, &(xo[0])); ST(&(xo[WS(os, 2)]), VFMAI(TfE, TfD), ovs, &(xo[0])); ST(&(xo[WS(os, 126)]), VFNMSI(TfE, TfD), ovs, &(xo[0])); ST(&(xo[WS(os, 66)]), VFMAI(TfC, Tfn), ovs, &(xo[0])); ST(&(xo[WS(os, 62)]), VFNMSI(TfC, Tfn), ovs, &(xo[0])); { V Tgd, Tg9, Tg6, Tg4; Tgd = VFNMS(LDK(KP773010453), Tg8, Tg7); Tg9 = VFMA(LDK(KP773010453), Tg8, Tg7); Tg6 = VFMA(LDK(KP773010453), Tg3, Tg0); Tg4 = VFNMS(LDK(KP773010453), Tg3, Tg0); { V Tge, Tgc, Tg5, TfX; Tge = VFMA(LDK(KP773010453), Tgb, Tga); Tgc = VFNMS(LDK(KP773010453), Tgb, Tga); Tg5 = VFMA(LDK(KP773010453), TfW, TfP); TfX = VFNMS(LDK(KP773010453), TfW, TfP); The = VFMA(LDK(KP831469612), Thd, Thc); Tho = VFNMS(LDK(KP831469612), Thd, Thc); TgT = VFMA(LDK(KP831469612), Tgo, Tgh); Tgp = VFNMS(LDK(KP831469612), Tgo, Tgh); ST(&(xo[WS(os, 110)]), VFNMSI(Tge, Tgd), ovs, &(xo[0])); ST(&(xo[WS(os, 18)]), VFMAI(Tge, Tgd), ovs, &(xo[0])); ST(&(xo[WS(os, 82)]), VFMAI(Tgc, Tg9), ovs, &(xo[0])); ST(&(xo[WS(os, 46)]), VFNMSI(Tgc, Tg9), ovs, &(xo[0])); ST(&(xo[WS(os, 114)]), VFMAI(Tg6, Tg5), ovs, &(xo[0])); ST(&(xo[WS(os, 14)]), VFNMSI(Tg6, Tg5), ovs, &(xo[0])); ST(&(xo[WS(os, 50)]), VFMAI(Tg4, TfX), ovs, &(xo[0])); ST(&(xo[WS(os, 78)]), VFNMSI(Tg4, TfX), ovs, &(xo[0])); } } } } } } { V Th4, Tgs, Tgv, Th5, Tgz, TgC, Th2, TgL; Th4 = VFMA(LDK(KP923879532), Tgr, Tgq); Tgs = VFNMS(LDK(KP923879532), Tgr, Tgq); Tgv = VFMA(LDK(KP923879532), Tgu, Tgt); Th5 = VFNMS(LDK(KP923879532), Tgu, Tgt); Th7 = VFMA(LDK(KP923879532), Tgy, Tgx); Tgz = VFNMS(LDK(KP923879532), Tgy, Tgx); TgC = VFMA(LDK(KP923879532), TgB, TgA); Th8 = VFNMS(LDK(KP923879532), TgB, TgA); Th2 = VADD(TgK, TgJ); TgL = VSUB(TgJ, TgK); { V TgO, Tgw, TgN, TgD; TgO = VFMA(LDK(KP534511135), Tgs, Tgv); Tgw = VFNMS(LDK(KP534511135), Tgv, Tgs); TgN = VFMA(LDK(KP534511135), Tgz, TgC); TgD = VFNMS(LDK(KP534511135), TgC, Tgz); Thg = VFNMS(LDK(KP303346683), Th4, Th5); Th6 = VFMA(LDK(KP303346683), Th5, Th4); Th3 = VFMA(LDK(KP831469612), Th2, Th1); Thl = VFNMS(LDK(KP831469612), Th2, Th1); TgW = VFNMS(LDK(KP831469612), TgL, TgI); TgM = VFMA(LDK(KP831469612), TgL, TgI); TgU = VADD(TgO, TgN); TgP = VSUB(TgN, TgO); TgX = VSUB(TgD, Tgw); TgE = VADD(Tgw, TgD); } } } { V T8u, T8v, T7R, T8d, T7G, Thm, Thh, Thp, Tha, T7Y, Thr, Thn; { V T7y, T7F, TgZ, TgV; T8u = VFNMS(LDK(KP831469612), T7x, T7q); T7y = VFMA(LDK(KP831469612), T7x, T7q); T7F = VFMA(LDK(KP831469612), T7E, T7B); T8v = VFNMS(LDK(KP831469612), T7E, T7B); T8x = VFNMS(LDK(KP831469612), T7Q, T7J); T7R = VFMA(LDK(KP831469612), T7Q, T7J); TgZ = VFMA(LDK(KP881921264), TgU, TgT); TgV = VFNMS(LDK(KP881921264), TgU, TgT); { V TgS, TgQ, Th0, TgY; TgS = VFMA(LDK(KP881921264), TgP, TgM); TgQ = VFNMS(LDK(KP881921264), TgP, TgM); Th0 = VFNMS(LDK(KP881921264), TgX, TgW); TgY = VFMA(LDK(KP881921264), TgX, TgW); { V TgR, TgF, Thf, Th9; TgR = VFMA(LDK(KP881921264), TgE, Tgp); TgF = VFNMS(LDK(KP881921264), TgE, Tgp); Thf = VFNMS(LDK(KP303346683), Th7, Th8); Th9 = VFMA(LDK(KP303346683), Th8, Th7); T8d = VFNMS(LDK(KP148335987), T7y, T7F); T7G = VFMA(LDK(KP148335987), T7F, T7y); ST(&(xo[WS(os, 106)]), VFMAI(Th0, TgZ), ovs, &(xo[0])); ST(&(xo[WS(os, 22)]), VFNMSI(Th0, TgZ), ovs, &(xo[0])); ST(&(xo[WS(os, 86)]), VFNMSI(TgY, TgV), ovs, &(xo[0])); ST(&(xo[WS(os, 42)]), VFMAI(TgY, TgV), ovs, &(xo[0])); ST(&(xo[WS(os, 10)]), VFMAI(TgS, TgR), ovs, &(xo[0])); ST(&(xo[WS(os, 118)]), VFNMSI(TgS, TgR), ovs, &(xo[0])); ST(&(xo[WS(os, 74)]), VFMAI(TgQ, TgF), ovs, &(xo[0])); ST(&(xo[WS(os, 54)]), VFNMSI(TgQ, TgF), ovs, &(xo[0])); Thm = VADD(Thg, Thf); Thh = VSUB(Thf, Thg); Thp = VSUB(Th9, Th6); Tha = VADD(Th6, Th9); T7Y = VFMA(LDK(KP831469612), T7X, T7U); T8y = VFNMS(LDK(KP831469612), T7X, T7U); } } } Thr = VFNMS(LDK(KP956940335), Thm, Thl); Thn = VFMA(LDK(KP956940335), Thm, Thl); { V Thk, Thi, Ths, Thq; Thk = VFMA(LDK(KP956940335), Thh, The); Thi = VFNMS(LDK(KP956940335), Thh, The); Ths = VFMA(LDK(KP956940335), Thp, Tho); Thq = VFNMS(LDK(KP956940335), Thp, Tho); { V Thj, Thb, T8e, T7Z; Thj = VFMA(LDK(KP956940335), Tha, Th3); Thb = VFNMS(LDK(KP956940335), Tha, Th3); T8e = VFNMS(LDK(KP148335987), T7R, T7Y); T7Z = VFMA(LDK(KP148335987), T7Y, T7R); T8F = VFMA(LDK(KP741650546), T8u, T8v); T8w = VFNMS(LDK(KP741650546), T8v, T8u); ST(&(xo[WS(os, 102)]), VFNMSI(Ths, Thr), ovs, &(xo[0])); ST(&(xo[WS(os, 26)]), VFMAI(Ths, Thr), ovs, &(xo[0])); ST(&(xo[WS(os, 90)]), VFMAI(Thq, Thn), ovs, &(xo[0])); ST(&(xo[WS(os, 38)]), VFNMSI(Thq, Thn), ovs, &(xo[0])); ST(&(xo[WS(os, 122)]), VFMAI(Thk, Thj), ovs, &(xo[0])); ST(&(xo[WS(os, 6)]), VFNMSI(Thk, Thj), ovs, &(xo[0])); ST(&(xo[WS(os, 58)]), VFMAI(Thi, Thb), ovs, &(xo[0])); ST(&(xo[WS(os, 70)]), VFNMSI(Thi, Thb), ovs, &(xo[0])); T8k = VADD(T8d, T8e); T8f = VSUB(T8d, T8e); T8n = VSUB(T7Z, T7G); T80 = VADD(T7G, T7Z); } } T9l = VSUB(T75, T72); T76 = VADD(T72, T75); T87 = VSUB(T85, T86); T8U = VADD(T85, T86); T89 = VFNMS(LDK(KP303346683), T7a, T7d); T7e = VFMA(LDK(KP303346683), T7d, T7a); T7l = VFMA(LDK(KP303346683), T7k, T7h); T8a = VFNMS(LDK(KP303346683), T7h, T7k); } } { V T11, T5h, T5a, T55, T5d, T4K, T5C, T5x, T5F, T5q, T4X, T4Z, T1C, T2d, T50; { V T5k, T3g, T3t, T5l, T5n, T4v, T4I, T5o, T8G, T8z; T5k = VFNMS(LDK(KP980785280), T3f, T2G); T3g = VFMA(LDK(KP980785280), T3f, T2G); T8G = VFMA(LDK(KP741650546), T8x, T8y); T8z = VFNMS(LDK(KP741650546), T8y, T8x); { V T8r, T77, T8C, T88; T8r = VFNMS(LDK(KP831469612), T76, T6Z); T77 = VFMA(LDK(KP831469612), T76, T6Z); T8C = VFNMS(LDK(KP831469612), T87, T84); T88 = VFMA(LDK(KP831469612), T87, T84); { V T8D, T7m, T8s, T8b; T8D = VSUB(T7l, T7e); T7m = VADD(T7e, T7l); T8s = VADD(T89, T8a); T8b = VSUB(T89, T8a); { V T8M, T8H, T8P, T8A; T8M = VADD(T8F, T8G); T8H = VSUB(T8F, T8G); T8P = VSUB(T8z, T8w); T8A = VADD(T8w, T8z); { V T8E, T8O, T8j, T7n; T8E = VFNMS(LDK(KP956940335), T8D, T8C); T8O = VFMA(LDK(KP956940335), T8D, T8C); T8j = VFNMS(LDK(KP956940335), T7m, T77); T7n = VFMA(LDK(KP956940335), T7m, T77); { V T8t, T8L, T8m, T8c; T8t = VFNMS(LDK(KP956940335), T8s, T8r); T8L = VFMA(LDK(KP956940335), T8s, T8r); T8m = VFNMS(LDK(KP956940335), T8b, T88); T8c = VFMA(LDK(KP956940335), T8b, T88); { V T8K, T8I, T8S, T8Q; T8K = VFMA(LDK(KP803207531), T8H, T8E); T8I = VFNMS(LDK(KP803207531), T8H, T8E); T8S = VFMA(LDK(KP803207531), T8P, T8O); T8Q = VFNMS(LDK(KP803207531), T8P, T8O); { V T8p, T8l, T8h, T81; T8p = VFNMS(LDK(KP989176509), T8k, T8j); T8l = VFMA(LDK(KP989176509), T8k, T8j); T8h = VFMA(LDK(KP989176509), T80, T7n); T81 = VFNMS(LDK(KP989176509), T80, T7n); { V T8J, T8B, T8R, T8N; T8J = VFMA(LDK(KP803207531), T8A, T8t); T8B = VFNMS(LDK(KP803207531), T8A, T8t); T8R = VFMA(LDK(KP803207531), T8M, T8L); T8N = VFNMS(LDK(KP803207531), T8M, T8L); { V T8q, T8o, T8i, T8g; T8q = VFNMS(LDK(KP989176509), T8n, T8m); T8o = VFMA(LDK(KP989176509), T8n, T8m); T8i = VFMA(LDK(KP989176509), T8f, T8c); T8g = VFNMS(LDK(KP989176509), T8f, T8c); ST(&(xo[WS(os, 115)]), VFMAI(T8K, T8J), ovs, &(xo[WS(os, 1)])); ST(&(xo[WS(os, 13)]), VFNMSI(T8K, T8J), ovs, &(xo[WS(os, 1)])); ST(&(xo[WS(os, 51)]), VFMAI(T8I, T8B), ovs, &(xo[WS(os, 1)])); ST(&(xo[WS(os, 77)]), VFNMSI(T8I, T8B), ovs, &(xo[WS(os, 1)])); ST(&(xo[WS(os, 109)]), VFNMSI(T8S, T8R), ovs, &(xo[WS(os, 1)])); ST(&(xo[WS(os, 19)]), VFMAI(T8S, T8R), ovs, &(xo[WS(os, 1)])); ST(&(xo[WS(os, 83)]), VFMAI(T8Q, T8N), ovs, &(xo[WS(os, 1)])); ST(&(xo[WS(os, 45)]), VFNMSI(T8Q, T8N), ovs, &(xo[WS(os, 1)])); ST(&(xo[WS(os, 99)]), VFMAI(T8q, T8p), ovs, &(xo[WS(os, 1)])); ST(&(xo[WS(os, 29)]), VFNMSI(T8q, T8p), ovs, &(xo[WS(os, 1)])); ST(&(xo[WS(os, 93)]), VFNMSI(T8o, T8l), ovs, &(xo[WS(os, 1)])); ST(&(xo[WS(os, 35)]), VFMAI(T8o, T8l), ovs, &(xo[WS(os, 1)])); ST(&(xo[WS(os, 3)]), VFMAI(T8i, T8h), ovs, &(xo[WS(os, 1)])); ST(&(xo[WS(os, 125)]), VFNMSI(T8i, T8h), ovs, &(xo[WS(os, 1)])); ST(&(xo[WS(os, 67)]), VFMAI(T8g, T81), ovs, &(xo[WS(os, 1)])); ST(&(xo[WS(os, 61)]), VFNMSI(T8g, T81), ovs, &(xo[WS(os, 1)])); T3t = VFMA(LDK(KP980785280), T3s, T3p); T5l = VFNMS(LDK(KP980785280), T3s, T3p); } } } } } } } } } T5n = VFNMS(LDK(KP980785280), T4u, T3V); T4v = VFMA(LDK(KP980785280), T4u, T3V); T4I = VFMA(LDK(KP980785280), T4H, T4E); T5o = VFNMS(LDK(KP980785280), T4H, T4E); { V T53, T3u, T54, T4J, T5v, T5m, T5w, T5p, T10; T6b = VSUB(TZ, TI); T10 = VADD(TI, TZ); T53 = VFMA(LDK(KP049126849), T3g, T3t); T3u = VFNMS(LDK(KP049126849), T3t, T3g); T54 = VFMA(LDK(KP049126849), T4v, T4I); T4J = VFNMS(LDK(KP049126849), T4I, T4v); T5v = VFNMS(LDK(KP906347169), T5k, T5l); T5m = VFMA(LDK(KP906347169), T5l, T5k); T5w = VFNMS(LDK(KP906347169), T5n, T5o); T5p = VFMA(LDK(KP906347169), T5o, T5n); T11 = VFMA(LDK(KP980785280), T10, Tr); T5h = VFNMS(LDK(KP980785280), T10, Tr); T5a = VADD(T53, T54); T55 = VSUB(T53, T54); T5d = VSUB(T4J, T3u); T4K = VADD(T3u, T4J); T5C = VADD(T5v, T5w); T5x = VSUB(T5v, T5w); T5F = VSUB(T5p, T5m); T5q = VADD(T5m, T5p); T4X = VSUB(T4V, T4W); T5K = VADD(T4V, T4W); } T4Z = VFMA(LDK(KP098491403), T1s, T1B); T1C = VFNMS(LDK(KP098491403), T1B, T1s); T2d = VFNMS(LDK(KP098491403), T2c, T23); T50 = VFMA(LDK(KP098491403), T23, T2c); } { V T9y, T9t, T9B, T9i, T9o, T9n, T9F, T8V, T9Q, T9m, T9R, T92, Ta0, T9V, Ta3; V T9O; { V T9I, T9J, T9L, T9d, T5s, T4Y, T5t, T2e, T5i, T51, T9r, T9a, T9g, T9M, T96; V T99; T9I = VFMA(LDK(KP831469612), T95, T94); T96 = VFNMS(LDK(KP831469612), T95, T94); T99 = VFNMS(LDK(KP831469612), T98, T97); T9J = VFMA(LDK(KP831469612), T98, T97); T9L = VFMA(LDK(KP831469612), T9c, T9b); T9d = VFNMS(LDK(KP831469612), T9c, T9b); T5s = VFNMS(LDK(KP980785280), T4X, T4U); T4Y = VFMA(LDK(KP980785280), T4X, T4U); T5t = VSUB(T2d, T1C); T2e = VADD(T1C, T2d); T5i = VADD(T4Z, T50); T51 = VSUB(T4Z, T50); T9r = VFNMS(LDK(KP599376933), T96, T99); T9a = VFMA(LDK(KP599376933), T99, T96); T9g = VFNMS(LDK(KP831469612), T9f, T9e); T9M = VFMA(LDK(KP831469612), T9f, T9e); { V T5u, T5E, T8Y, T91; T5u = VFMA(LDK(KP995184726), T5t, T5s); T5E = VFNMS(LDK(KP995184726), T5t, T5s); { V T59, T2f, T5j, T5B; T59 = VFNMS(LDK(KP995184726), T2e, T11); T2f = VFMA(LDK(KP995184726), T2e, T11); T5j = VFMA(LDK(KP995184726), T5i, T5h); T5B = VFNMS(LDK(KP995184726), T5i, T5h); { V T5c, T52, T9s, T9h; T5c = VFNMS(LDK(KP995184726), T51, T4Y); T52 = VFMA(LDK(KP995184726), T51, T4Y); T9s = VFNMS(LDK(KP599376933), T9d, T9g); T9h = VFMA(LDK(KP599376933), T9g, T9d); { V T5A, T5y, T5I, T5G; T5A = VFMA(LDK(KP740951125), T5x, T5u); T5y = VFNMS(LDK(KP740951125), T5x, T5u); T5I = VFNMS(LDK(KP740951125), T5F, T5E); T5G = VFMA(LDK(KP740951125), T5F, T5E); { V T5f, T5b, T57, T4L; T5f = VFMA(LDK(KP998795456), T5a, T59); T5b = VFNMS(LDK(KP998795456), T5a, T59); T57 = VFMA(LDK(KP998795456), T4K, T2f); T4L = VFNMS(LDK(KP998795456), T4K, T2f); { V T5z, T5r, T5H, T5D; T5z = VFMA(LDK(KP740951125), T5q, T5j); T5r = VFNMS(LDK(KP740951125), T5q, T5j); T5H = VFNMS(LDK(KP740951125), T5C, T5B); T5D = VFMA(LDK(KP740951125), T5C, T5B); { V T5g, T5e, T58, T56; T5g = VFMA(LDK(KP998795456), T5d, T5c); T5e = VFNMS(LDK(KP998795456), T5d, T5c); T58 = VFMA(LDK(KP998795456), T55, T52); T56 = VFNMS(LDK(KP998795456), T55, T52); T9y = VADD(T9r, T9s); T9t = VSUB(T9r, T9s); T9B = VSUB(T9h, T9a); T9i = VADD(T9a, T9h); ST(&(xo[WS(os, 15)]), VFMAI(T5A, T5z), ovs, &(xo[WS(os, 1)])); ST(&(xo[WS(os, 113)]), VFNMSI(T5A, T5z), ovs, &(xo[WS(os, 1)])); ST(&(xo[WS(os, 79)]), VFMAI(T5y, T5r), ovs, &(xo[WS(os, 1)])); ST(&(xo[WS(os, 49)]), VFNMSI(T5y, T5r), ovs, &(xo[WS(os, 1)])); ST(&(xo[WS(os, 111)]), VFMAI(T5I, T5H), ovs, &(xo[WS(os, 1)])); ST(&(xo[WS(os, 17)]), VFNMSI(T5I, T5H), ovs, &(xo[WS(os, 1)])); ST(&(xo[WS(os, 81)]), VFNMSI(T5G, T5D), ovs, &(xo[WS(os, 1)])); ST(&(xo[WS(os, 47)]), VFMAI(T5G, T5D), ovs, &(xo[WS(os, 1)])); ST(&(xo[WS(os, 97)]), VFNMSI(T5g, T5f), ovs, &(xo[WS(os, 1)])); ST(&(xo[WS(os, 31)]), VFMAI(T5g, T5f), ovs, &(xo[WS(os, 1)])); ST(&(xo[WS(os, 95)]), VFMAI(T5e, T5b), ovs, &(xo[WS(os, 1)])); ST(&(xo[WS(os, 33)]), VFNMSI(T5e, T5b), ovs, &(xo[WS(os, 1)])); ST(&(xo[WS(os, 127)]), VFMAI(T58, T57), ovs, &(xo[WS(os, 1)])); ST(&(xo[WS(os, 1)]), VFNMSI(T58, T57), ovs, &(xo[WS(os, 1)])); ST(&(xo[WS(os, 63)]), VFMAI(T56, T4L), ovs, &(xo[WS(os, 1)])); ST(&(xo[WS(os, 65)]), VFNMSI(T56, T4L), ovs, &(xo[WS(os, 1)])); } } } } } } T9o = VFNMS(LDK(KP534511135), T8W, T8X); T8Y = VFMA(LDK(KP534511135), T8X, T8W); T91 = VFMA(LDK(KP534511135), T90, T8Z); T9n = VFNMS(LDK(KP534511135), T8Z, T90); { V T9T, T9K, T9U, T9N; T9T = VFMA(LDK(KP250486960), T9I, T9J); T9K = VFNMS(LDK(KP250486960), T9J, T9I); T9U = VFMA(LDK(KP250486960), T9L, T9M); T9N = VFNMS(LDK(KP250486960), T9M, T9L); T9F = VFNMS(LDK(KP831469612), T8U, T8T); T8V = VFMA(LDK(KP831469612), T8U, T8T); T9Q = VFNMS(LDK(KP831469612), T9l, T9k); T9m = VFMA(LDK(KP831469612), T9l, T9k); T9R = VSUB(T8Y, T91); T92 = VADD(T8Y, T91); Ta0 = VADD(T9T, T9U); T9V = VSUB(T9T, T9U); Ta3 = VSUB(T9N, T9K); T9O = VADD(T9K, T9N); } } } { V T6y, T6z, T63, T9Y, T9W, Ta6, Ta4, T9D, T9z, T9v, T9j, T6h, T60, T9H, T9Z; V T9A, T9q, T66, T9X, T9P; { V T5W, T9S, Ta2, T9x, T93, T5Z, T9G, T9p; T6y = VFMA(LDK(KP980785280), T5V, T5U); T5W = VFNMS(LDK(KP980785280), T5V, T5U); T9S = VFMA(LDK(KP881921264), T9R, T9Q); Ta2 = VFNMS(LDK(KP881921264), T9R, T9Q); T9x = VFNMS(LDK(KP881921264), T92, T8V); T93 = VFMA(LDK(KP881921264), T92, T8V); T5Z = VFMA(LDK(KP980785280), T5Y, T5X); T6z = VFNMS(LDK(KP980785280), T5Y, T5X); T6B = VFMA(LDK(KP980785280), T62, T61); T63 = VFNMS(LDK(KP980785280), T62, T61); T9G = VADD(T9o, T9n); T9p = VSUB(T9n, T9o); T9Y = VFMA(LDK(KP970031253), T9V, T9S); T9W = VFNMS(LDK(KP970031253), T9V, T9S); Ta6 = VFMA(LDK(KP970031253), Ta3, Ta2); Ta4 = VFNMS(LDK(KP970031253), Ta3, Ta2); T9D = VFNMS(LDK(KP857728610), T9y, T9x); T9z = VFMA(LDK(KP857728610), T9y, T9x); T9v = VFMA(LDK(KP857728610), T9i, T93); T9j = VFNMS(LDK(KP857728610), T9i, T93); T6h = VFMA(LDK(KP472964775), T5W, T5Z); T60 = VFNMS(LDK(KP472964775), T5Z, T5W); T9H = VFMA(LDK(KP881921264), T9G, T9F); T9Z = VFNMS(LDK(KP881921264), T9G, T9F); T9A = VFNMS(LDK(KP881921264), T9p, T9m); T9q = VFMA(LDK(KP881921264), T9p, T9m); T66 = VFMA(LDK(KP980785280), T65, T64); T6C = VFNMS(LDK(KP980785280), T65, T64); } T9X = VFMA(LDK(KP970031253), T9O, T9H); T9P = VFNMS(LDK(KP970031253), T9O, T9H); { V Ta5, Ta1, T9E, T9C; Ta5 = VFMA(LDK(KP970031253), Ta0, T9Z); Ta1 = VFNMS(LDK(KP970031253), Ta0, T9Z); T9E = VFNMS(LDK(KP857728610), T9B, T9A); T9C = VFMA(LDK(KP857728610), T9B, T9A); { V T9w, T9u, T6i, T67; T9w = VFMA(LDK(KP857728610), T9t, T9q); T9u = VFNMS(LDK(KP857728610), T9t, T9q); T6i = VFMA(LDK(KP472964775), T63, T66); T67 = VFNMS(LDK(KP472964775), T66, T63); T6J = VFNMS(LDK(KP357805721), T6y, T6z); T6A = VFMA(LDK(KP357805721), T6z, T6y); ST(&(xo[WS(os, 123)]), VFMAI(T9Y, T9X), ovs, &(xo[WS(os, 1)])); ST(&(xo[WS(os, 5)]), VFNMSI(T9Y, T9X), ovs, &(xo[WS(os, 1)])); ST(&(xo[WS(os, 59)]), VFMAI(T9W, T9P), ovs, &(xo[WS(os, 1)])); ST(&(xo[WS(os, 69)]), VFNMSI(T9W, T9P), ovs, &(xo[WS(os, 1)])); ST(&(xo[WS(os, 101)]), VFNMSI(Ta6, Ta5), ovs, &(xo[WS(os, 1)])); ST(&(xo[WS(os, 27)]), VFMAI(Ta6, Ta5), ovs, &(xo[WS(os, 1)])); ST(&(xo[WS(os, 91)]), VFMAI(Ta4, Ta1), ovs, &(xo[WS(os, 1)])); ST(&(xo[WS(os, 37)]), VFNMSI(Ta4, Ta1), ovs, &(xo[WS(os, 1)])); ST(&(xo[WS(os, 107)]), VFMAI(T9E, T9D), ovs, &(xo[WS(os, 1)])); ST(&(xo[WS(os, 21)]), VFNMSI(T9E, T9D), ovs, &(xo[WS(os, 1)])); ST(&(xo[WS(os, 85)]), VFNMSI(T9C, T9z), ovs, &(xo[WS(os, 1)])); ST(&(xo[WS(os, 43)]), VFMAI(T9C, T9z), ovs, &(xo[WS(os, 1)])); ST(&(xo[WS(os, 11)]), VFMAI(T9w, T9v), ovs, &(xo[WS(os, 1)])); ST(&(xo[WS(os, 117)]), VFNMSI(T9w, T9v), ovs, &(xo[WS(os, 1)])); ST(&(xo[WS(os, 75)]), VFMAI(T9u, T9j), ovs, &(xo[WS(os, 1)])); ST(&(xo[WS(os, 53)]), VFNMSI(T9u, T9j), ovs, &(xo[WS(os, 1)])); T6o = VADD(T6h, T6i); T6j = VSUB(T6h, T6i); T6r = VSUB(T67, T60); T68 = VADD(T60, T67); } } T6e = VFMA(LDK(KP820678790), T5M, T5N); T5O = VFNMS(LDK(KP820678790), T5N, T5M); T5R = VFNMS(LDK(KP820678790), T5Q, T5P); T6d = VFMA(LDK(KP820678790), T5P, T5Q); } } } } } } T6D = VFMA(LDK(KP357805721), T6C, T6B); T6K = VFNMS(LDK(KP357805721), T6B, T6C); { V T5L, T6v, T6c, T6G; T5L = VFNMS(LDK(KP980785280), T5K, T5J); T6v = VFMA(LDK(KP980785280), T5K, T5J); T6c = VFNMS(LDK(KP980785280), T6b, T6a); T6G = VFMA(LDK(KP980785280), T6b, T6a); { V T5S, T6H, T6f, T6w; T5S = VADD(T5O, T5R); T6H = VSUB(T5O, T5R); T6f = VSUB(T6d, T6e); T6w = VADD(T6e, T6d); { V T6L, T6Q, T6E, T6T; T6L = VSUB(T6J, T6K); T6Q = VADD(T6J, T6K); T6E = VADD(T6A, T6D); T6T = VSUB(T6D, T6A); { V T6S, T6I, T5T, T6n; T6S = VFNMS(LDK(KP773010453), T6H, T6G); T6I = VFMA(LDK(KP773010453), T6H, T6G); T5T = VFNMS(LDK(KP773010453), T5S, T5L); T6n = VFMA(LDK(KP773010453), T5S, T5L); { V T6P, T6x, T6g, T6q; T6P = VFNMS(LDK(KP773010453), T6w, T6v); T6x = VFMA(LDK(KP773010453), T6w, T6v); T6g = VFNMS(LDK(KP773010453), T6f, T6c); T6q = VFMA(LDK(KP773010453), T6f, T6c); { V T6M, T6O, T6U, T6W; T6M = VFNMS(LDK(KP941544065), T6L, T6I); T6O = VFMA(LDK(KP941544065), T6L, T6I); T6U = VFMA(LDK(KP941544065), T6T, T6S); T6W = VFNMS(LDK(KP941544065), T6T, T6S); { V T6p, T6t, T69, T6l; T6p = VFNMS(LDK(KP903989293), T6o, T6n); T6t = VFMA(LDK(KP903989293), T6o, T6n); T69 = VFNMS(LDK(KP903989293), T68, T5T); T6l = VFMA(LDK(KP903989293), T68, T5T); { V T6F, T6N, T6R, T6V; T6F = VFNMS(LDK(KP941544065), T6E, T6x); T6N = VFMA(LDK(KP941544065), T6E, T6x); T6R = VFMA(LDK(KP941544065), T6Q, T6P); T6V = VFNMS(LDK(KP941544065), T6Q, T6P); { V T6s, T6u, T6k, T6m; T6s = VFNMS(LDK(KP903989293), T6r, T6q); T6u = VFMA(LDK(KP903989293), T6r, T6q); T6k = VFNMS(LDK(KP903989293), T6j, T6g); T6m = VFMA(LDK(KP903989293), T6j, T6g); ST(&(xo[WS(os, 7)]), VFMAI(T6O, T6N), ovs, &(xo[WS(os, 1)])); ST(&(xo[WS(os, 121)]), VFNMSI(T6O, T6N), ovs, &(xo[WS(os, 1)])); ST(&(xo[WS(os, 71)]), VFMAI(T6M, T6F), ovs, &(xo[WS(os, 1)])); ST(&(xo[WS(os, 57)]), VFNMSI(T6M, T6F), ovs, &(xo[WS(os, 1)])); ST(&(xo[WS(os, 103)]), VFMAI(T6W, T6V), ovs, &(xo[WS(os, 1)])); ST(&(xo[WS(os, 25)]), VFNMSI(T6W, T6V), ovs, &(xo[WS(os, 1)])); ST(&(xo[WS(os, 89)]), VFNMSI(T6U, T6R), ovs, &(xo[WS(os, 1)])); ST(&(xo[WS(os, 39)]), VFMAI(T6U, T6R), ovs, &(xo[WS(os, 1)])); ST(&(xo[WS(os, 105)]), VFNMSI(T6u, T6t), ovs, &(xo[WS(os, 1)])); ST(&(xo[WS(os, 23)]), VFMAI(T6u, T6t), ovs, &(xo[WS(os, 1)])); ST(&(xo[WS(os, 87)]), VFMAI(T6s, T6p), ovs, &(xo[WS(os, 1)])); ST(&(xo[WS(os, 41)]), VFNMSI(T6s, T6p), ovs, &(xo[WS(os, 1)])); ST(&(xo[WS(os, 119)]), VFMAI(T6m, T6l), ovs, &(xo[WS(os, 1)])); ST(&(xo[WS(os, 9)]), VFNMSI(T6m, T6l), ovs, &(xo[WS(os, 1)])); ST(&(xo[WS(os, 55)]), VFMAI(T6k, T69), ovs, &(xo[WS(os, 1)])); ST(&(xo[WS(os, 73)]), VFNMSI(T6k, T69), ovs, &(xo[WS(os, 1)])); } } } } } } } } } } } VLEAVE(); } static const kdft_desc desc = { 128, XSIMD_STRING("n1fv_128"), {440, 0, 642, 0}, &GENUS, 0, 0, 0, 0 }; void XSIMD(codelet_n1fv_128) (planner *p) { X(kdft_register) (p, n1fv_128, &desc); } #else /* HAVE_FMA */ /* Generated by: ../../../genfft/gen_notw_c.native -simd -compact -variables 4 -pipeline-latency 8 -n 128 -name n1fv_128 -include n1f.h */ /* * This function contains 1082 FP additions, 330 FP multiplications, * (or, 938 additions, 186 multiplications, 144 fused multiply/add), * 194 stack variables, 31 constants, and 256 memory accesses */ #include "n1f.h" static void n1fv_128(const R *ri, const R *ii, R *ro, R *io, stride is, stride os, INT v, INT ivs, INT ovs) { DVK(KP941544065, +0.941544065183020778412509402599502357185589796); DVK(KP336889853, +0.336889853392220050689253212619147570477766780); DVK(KP903989293, +0.903989293123443331586200297230537048710132025); DVK(KP427555093, +0.427555093430282094320966856888798534304578629); DVK(KP970031253, +0.970031253194543992603984207286100251456865962); DVK(KP242980179, +0.242980179903263889948274162077471118320990783); DVK(KP857728610, +0.857728610000272069902269984284770137042490799); DVK(KP514102744, +0.514102744193221726593693838968815772608049120); DVK(KP671558954, +0.671558954847018400625376850427421803228750632); DVK(KP740951125, +0.740951125354959091175616897495162729728955309); DVK(KP049067674, +0.049067674327418014254954976942682658314745363); DVK(KP998795456, +0.998795456205172392714771604759100694443203615); DVK(KP595699304, +0.595699304492433343467036528829969889511926338); DVK(KP803207531, +0.803207531480644909806676512963141923879569427); DVK(KP146730474, +0.146730474455361751658850129646717819706215317); DVK(KP989176509, +0.989176509964780973451673738016243063983689533); DVK(KP956940335, +0.956940335732208864935797886980269969482849206); DVK(KP290284677, +0.290284677254462367636192375817395274691476278); DVK(KP881921264, +0.881921264348355029712756863660388349508442621); DVK(KP471396736, +0.471396736825997648556387625905254377657460319); DVK(KP634393284, +0.634393284163645498215171613225493370675687095); DVK(KP773010453, +0.773010453362736960810906609758469800971041293); DVK(KP098017140, +0.098017140329560601994195563888641845861136673); DVK(KP995184726, +0.995184726672196886244836953109479921575474869); DVK(KP555570233, +0.555570233019602224742830813948532874374937191); DVK(KP831469612, +0.831469612302545237078788377617905756738560812); DVK(KP195090322, +0.195090322016128267848284868477022240927691618); DVK(KP980785280, +0.980785280403230449126182236134239036973933731); DVK(KP382683432, +0.382683432365089771728459984030398866761344562); DVK(KP923879532, +0.923879532511286756128183189396788286822416626); DVK(KP707106781, +0.707106781186547524400844362104849039284835938); { INT i; const R *xi; R *xo; xi = ri; xo = ro; for (i = v; i > 0; i = i - VL, xi = xi + (VL * ivs), xo = xo + (VL * ovs), MAKE_VOLATILE_STRIDE(256, is), MAKE_VOLATILE_STRIDE(256, os)) { V Tr, T5J, Ted, Tgf, Tfq, TgH, T4U, T6b, T6Z, T8T, Tad, TcZ, Tcc, Td0, T84; V T9l, Tb6, Tbt, T2G, T5X, TeV, Tgr, T3p, T5V, T7B, T95, TeK, Tgt, T7q, T97; V Td8, TdK, TbD, Tc0, T3V, T61, Tfg, TgB, T4E, T65, T7U, T9f, Tf5, Tgx, T7J; V T9b, Tdf, TdN, Td2, Td3, TI, T4V, Tft, Tgg, TZ, T4W, T75, T86, Tek, TgG; V T72, T85, Tas, Tcd, Tdp, Tdq, TdG, Teq, Tgm, Tet, Tgl, T1s, T5P, T1B, T5Q; V T7d, T8Z, TaI, Tcf, T7a, T90, Tdm, Tdn, TdH, Tez, Tgi, TeC, Tgj, T23, T5N; V T2c, T5M, T7k, T8X, TaX, Tcg, T7h, T8W, Tbl, Tbu, Tdb, TdL, TeY, Tgu, TeR; V Tgq, T7x, T98, T7E, T94, T3f, T5Y, T3s, T5U, TbS, Tc1, Tdi, TdO, Tfj, Tgy; V Tfc, TgA, T7Q, T9e, T7X, T9c, T4u, T64, T4H, T62; { V T3, Ta7, T4P, Ta8, Ta, Tab, T4M, Taa, Tc9, Tca, Ti, Tea, T4S, Tc6, Tc7; V Tp, Teb, T4R; { V T1, T2, T4N, T4O; T1 = LD(&(xi[0]), ivs, &(xi[0])); T2 = LD(&(xi[WS(is, 64)]), ivs, &(xi[0])); T3 = VSUB(T1, T2); Ta7 = VADD(T1, T2); T4N = LD(&(xi[WS(is, 32)]), ivs, &(xi[0])); T4O = LD(&(xi[WS(is, 96)]), ivs, &(xi[0])); T4P = VSUB(T4N, T4O); Ta8 = VADD(T4N, T4O); } { V T4, T5, T6, T7, T8, T9; T4 = LD(&(xi[WS(is, 16)]), ivs, &(xi[0])); T5 = LD(&(xi[WS(is, 80)]), ivs, &(xi[0])); T6 = VSUB(T4, T5); T7 = LD(&(xi[WS(is, 112)]), ivs, &(xi[0])); T8 = LD(&(xi[WS(is, 48)]), ivs, &(xi[0])); T9 = VSUB(T7, T8); Ta = VMUL(LDK(KP707106781), VADD(T6, T9)); Tab = VADD(T7, T8); T4M = VMUL(LDK(KP707106781), VSUB(T9, T6)); Taa = VADD(T4, T5); } { V Te, Th, Tl, To; { V Tc, Td, Tf, Tg; Tc = LD(&(xi[WS(is, 8)]), ivs, &(xi[0])); Td = LD(&(xi[WS(is, 72)]), ivs, &(xi[0])); Te = VSUB(Tc, Td); Tc9 = VADD(Tc, Td); Tf = LD(&(xi[WS(is, 40)]), ivs, &(xi[0])); Tg = LD(&(xi[WS(is, 104)]), ivs, &(xi[0])); Th = VSUB(Tf, Tg); Tca = VADD(Tf, Tg); } Ti = VFNMS(LDK(KP382683432), Th, VMUL(LDK(KP923879532), Te)); Tea = VSUB(Tc9, Tca); T4S = VFMA(LDK(KP382683432), Te, VMUL(LDK(KP923879532), Th)); { V Tj, Tk, Tm, Tn; Tj = LD(&(xi[WS(is, 120)]), ivs, &(xi[0])); Tk = LD(&(xi[WS(is, 56)]), ivs, &(xi[0])); Tl = VSUB(Tj, Tk); Tc6 = VADD(Tj, Tk); Tm = LD(&(xi[WS(is, 24)]), ivs, &(xi[0])); Tn = LD(&(xi[WS(is, 88)]), ivs, &(xi[0])); To = VSUB(Tm, Tn); Tc7 = VADD(Tm, Tn); } Tp = VFMA(LDK(KP923879532), Tl, VMUL(LDK(KP382683432), To)); Teb = VSUB(Tc6, Tc7); T4R = VFNMS(LDK(KP923879532), To, VMUL(LDK(KP382683432), Tl)); } { V Tb, Tq, Te9, Tec; Tb = VADD(T3, Ta); Tq = VADD(Ti, Tp); Tr = VADD(Tb, Tq); T5J = VSUB(Tb, Tq); Te9 = VSUB(Ta7, Ta8); Tec = VMUL(LDK(KP707106781), VADD(Tea, Teb)); Ted = VADD(Te9, Tec); Tgf = VSUB(Te9, Tec); } { V Tfo, Tfp, T4Q, T4T; Tfo = VSUB(Tab, Taa); Tfp = VMUL(LDK(KP707106781), VSUB(Teb, Tea)); Tfq = VADD(Tfo, Tfp); TgH = VSUB(Tfp, Tfo); T4Q = VSUB(T4M, T4P); T4T = VSUB(T4R, T4S); T4U = VADD(T4Q, T4T); T6b = VSUB(T4T, T4Q); } { V T6X, T6Y, Ta9, Tac; T6X = VSUB(T3, Ta); T6Y = VADD(T4S, T4R); T6Z = VADD(T6X, T6Y); T8T = VSUB(T6X, T6Y); Ta9 = VADD(Ta7, Ta8); Tac = VADD(Taa, Tab); Tad = VSUB(Ta9, Tac); TcZ = VADD(Ta9, Tac); } { V Tc8, Tcb, T82, T83; Tc8 = VADD(Tc6, Tc7); Tcb = VADD(Tc9, Tca); Tcc = VSUB(Tc8, Tcb); Td0 = VADD(Tcb, Tc8); T82 = VADD(T4P, T4M); T83 = VSUB(Tp, Ti); T84 = VADD(T82, T83); T9l = VSUB(T83, T82); } } { V Tb0, Tb1, T2i, Tb2, T3k, Tb3, Tb4, T2p, Tb5, T3h, T2x, TeH, T3n, Tbs, T2E; V TeI, T3m, Tbp, T2l, T2o, TeG, TeJ; { V T2g, T2h, T3i, T3j; T2g = LD(&(xi[WS(is, 1)]), ivs, &(xi[WS(is, 1)])); T2h = LD(&(xi[WS(is, 65)]), ivs, &(xi[WS(is, 1)])); Tb0 = VADD(T2g, T2h); T3i = LD(&(xi[WS(is, 33)]), ivs, &(xi[WS(is, 1)])); T3j = LD(&(xi[WS(is, 97)]), ivs, &(xi[WS(is, 1)])); Tb1 = VADD(T3i, T3j); T2i = VSUB(T2g, T2h); Tb2 = VADD(Tb0, Tb1); T3k = VSUB(T3i, T3j); } { V T2j, T2k, T2m, T2n; T2j = LD(&(xi[WS(is, 17)]), ivs, &(xi[WS(is, 1)])); T2k = LD(&(xi[WS(is, 81)]), ivs, &(xi[WS(is, 1)])); T2l = VSUB(T2j, T2k); Tb3 = VADD(T2j, T2k); T2m = LD(&(xi[WS(is, 113)]), ivs, &(xi[WS(is, 1)])); T2n = LD(&(xi[WS(is, 49)]), ivs, &(xi[WS(is, 1)])); T2o = VSUB(T2m, T2n); Tb4 = VADD(T2m, T2n); } T2p = VMUL(LDK(KP707106781), VADD(T2l, T2o)); Tb5 = VADD(Tb3, Tb4); T3h = VMUL(LDK(KP707106781), VSUB(T2o, T2l)); { V T2t, Tbq, T2w, Tbr; { V T2r, T2s, T2u, T2v; T2r = LD(&(xi[WS(is, 9)]), ivs, &(xi[WS(is, 1)])); T2s = LD(&(xi[WS(is, 73)]), ivs, &(xi[WS(is, 1)])); T2t = VSUB(T2r, T2s); Tbq = VADD(T2r, T2s); T2u = LD(&(xi[WS(is, 41)]), ivs, &(xi[WS(is, 1)])); T2v = LD(&(xi[WS(is, 105)]), ivs, &(xi[WS(is, 1)])); T2w = VSUB(T2u, T2v); Tbr = VADD(T2u, T2v); } T2x = VFNMS(LDK(KP382683432), T2w, VMUL(LDK(KP923879532), T2t)); TeH = VSUB(Tbq, Tbr); T3n = VFMA(LDK(KP382683432), T2t, VMUL(LDK(KP923879532), T2w)); Tbs = VADD(Tbq, Tbr); } { V T2A, Tbn, T2D, Tbo; { V T2y, T2z, T2B, T2C; T2y = LD(&(xi[WS(is, 121)]), ivs, &(xi[WS(is, 1)])); T2z = LD(&(xi[WS(is, 57)]), ivs, &(xi[WS(is, 1)])); T2A = VSUB(T2y, T2z); Tbn = VADD(T2y, T2z); T2B = LD(&(xi[WS(is, 25)]), ivs, &(xi[WS(is, 1)])); T2C = LD(&(xi[WS(is, 89)]), ivs, &(xi[WS(is, 1)])); T2D = VSUB(T2B, T2C); Tbo = VADD(T2B, T2C); } T2E = VFMA(LDK(KP923879532), T2A, VMUL(LDK(KP382683432), T2D)); TeI = VSUB(Tbn, Tbo); T3m = VFNMS(LDK(KP923879532), T2D, VMUL(LDK(KP382683432), T2A)); Tbp = VADD(Tbn, Tbo); } Tb6 = VSUB(Tb2, Tb5); Tbt = VSUB(Tbp, Tbs); { V T2q, T2F, TeT, TeU; T2q = VADD(T2i, T2p); T2F = VADD(T2x, T2E); T2G = VADD(T2q, T2F); T5X = VSUB(T2q, T2F); TeT = VSUB(Tb4, Tb3); TeU = VMUL(LDK(KP707106781), VSUB(TeI, TeH)); TeV = VADD(TeT, TeU); Tgr = VSUB(TeU, TeT); } { V T3l, T3o, T7z, T7A; T3l = VSUB(T3h, T3k); T3o = VSUB(T3m, T3n); T3p = VADD(T3l, T3o); T5V = VSUB(T3o, T3l); T7z = VADD(T3k, T3h); T7A = VSUB(T2E, T2x); T7B = VADD(T7z, T7A); T95 = VSUB(T7A, T7z); } TeG = VSUB(Tb0, Tb1); TeJ = VMUL(LDK(KP707106781), VADD(TeH, TeI)); TeK = VADD(TeG, TeJ); Tgt = VSUB(TeG, TeJ); { V T7o, T7p, Td6, Td7; T7o = VSUB(T2i, T2p); T7p = VADD(T3n, T3m); T7q = VADD(T7o, T7p); T97 = VSUB(T7o, T7p); Td6 = VADD(Tb2, Tb5); Td7 = VADD(Tbs, Tbp); Td8 = VADD(Td6, Td7); TdK = VSUB(Td6, Td7); } } { V Tbx, Tby, T3x, Tbz, T4z, TbA, TbB, T3E, TbC, T4w, T3M, Tf2, T4C, TbZ, T3T; V Tf3, T4B, TbW, T3A, T3D, Tf1, Tf4; { V T3v, T3w, T4x, T4y; T3v = LD(&(xi[WS(is, 127)]), ivs, &(xi[WS(is, 1)])); T3w = LD(&(xi[WS(is, 63)]), ivs, &(xi[WS(is, 1)])); Tbx = VADD(T3v, T3w); T4x = LD(&(xi[WS(is, 31)]), ivs, &(xi[WS(is, 1)])); T4y = LD(&(xi[WS(is, 95)]), ivs, &(xi[WS(is, 1)])); Tby = VADD(T4x, T4y); T3x = VSUB(T3v, T3w); Tbz = VADD(Tbx, Tby); T4z = VSUB(T4x, T4y); } { V T3y, T3z, T3B, T3C; T3y = LD(&(xi[WS(is, 15)]), ivs, &(xi[WS(is, 1)])); T3z = LD(&(xi[WS(is, 79)]), ivs, &(xi[WS(is, 1)])); T3A = VSUB(T3y, T3z); TbA = VADD(T3y, T3z); T3B = LD(&(xi[WS(is, 111)]), ivs, &(xi[WS(is, 1)])); T3C = LD(&(xi[WS(is, 47)]), ivs, &(xi[WS(is, 1)])); T3D = VSUB(T3B, T3C); TbB = VADD(T3B, T3C); } T3E = VMUL(LDK(KP707106781), VADD(T3A, T3D)); TbC = VADD(TbA, TbB); T4w = VMUL(LDK(KP707106781), VSUB(T3D, T3A)); { V T3I, TbX, T3L, TbY; { V T3G, T3H, T3J, T3K; T3G = LD(&(xi[WS(is, 7)]), ivs, &(xi[WS(is, 1)])); T3H = LD(&(xi[WS(is, 71)]), ivs, &(xi[WS(is, 1)])); T3I = VSUB(T3G, T3H); TbX = VADD(T3G, T3H); T3J = LD(&(xi[WS(is, 39)]), ivs, &(xi[WS(is, 1)])); T3K = LD(&(xi[WS(is, 103)]), ivs, &(xi[WS(is, 1)])); T3L = VSUB(T3J, T3K); TbY = VADD(T3J, T3K); } T3M = VFNMS(LDK(KP382683432), T3L, VMUL(LDK(KP923879532), T3I)); Tf2 = VSUB(TbX, TbY); T4C = VFMA(LDK(KP382683432), T3I, VMUL(LDK(KP923879532), T3L)); TbZ = VADD(TbX, TbY); } { V T3P, TbU, T3S, TbV; { V T3N, T3O, T3Q, T3R; T3N = LD(&(xi[WS(is, 119)]), ivs, &(xi[WS(is, 1)])); T3O = LD(&(xi[WS(is, 55)]), ivs, &(xi[WS(is, 1)])); T3P = VSUB(T3N, T3O); TbU = VADD(T3N, T3O); T3Q = LD(&(xi[WS(is, 23)]), ivs, &(xi[WS(is, 1)])); T3R = LD(&(xi[WS(is, 87)]), ivs, &(xi[WS(is, 1)])); T3S = VSUB(T3Q, T3R); TbV = VADD(T3Q, T3R); } T3T = VFMA(LDK(KP923879532), T3P, VMUL(LDK(KP382683432), T3S)); Tf3 = VSUB(TbU, TbV); T4B = VFNMS(LDK(KP923879532), T3S, VMUL(LDK(KP382683432), T3P)); TbW = VADD(TbU, TbV); } TbD = VSUB(Tbz, TbC); Tc0 = VSUB(TbW, TbZ); { V T3F, T3U, Tfe, Tff; T3F = VADD(T3x, T3E); T3U = VADD(T3M, T3T); T3V = VADD(T3F, T3U); T61 = VSUB(T3F, T3U); Tfe = VSUB(TbB, TbA); Tff = VMUL(LDK(KP707106781), VSUB(Tf3, Tf2)); Tfg = VADD(Tfe, Tff); TgB = VSUB(Tff, Tfe); } { V T4A, T4D, T7S, T7T; T4A = VSUB(T4w, T4z); T4D = VSUB(T4B, T4C); T4E = VADD(T4A, T4D); T65 = VSUB(T4D, T4A); T7S = VADD(T4z, T4w); T7T = VSUB(T3T, T3M); T7U = VADD(T7S, T7T); T9f = VSUB(T7T, T7S); } Tf1 = VSUB(Tbx, Tby); Tf4 = VMUL(LDK(KP707106781), VADD(Tf2, Tf3)); Tf5 = VADD(Tf1, Tf4); Tgx = VSUB(Tf1, Tf4); { V T7H, T7I, Tdd, Tde; T7H = VSUB(T3x, T3E); T7I = VADD(T4C, T4B); T7J = VADD(T7H, T7I); T9b = VSUB(T7H, T7I); Tdd = VADD(Tbz, TbC); Tde = VADD(TbZ, TbW); Tdf = VADD(Tdd, Tde); TdN = VSUB(Tdd, Tde); } } { V Tu, Tee, TG, Tag, TL, Teh, TX, Tan, TB, Tef, TD, Taj, TS, Tei, TU; V Taq, Teg, Tej; { V Ts, Tt, Tae, TE, TF, Taf; Ts = LD(&(xi[WS(is, 4)]), ivs, &(xi[0])); Tt = LD(&(xi[WS(is, 68)]), ivs, &(xi[0])); Tae = VADD(Ts, Tt); TE = LD(&(xi[WS(is, 36)]), ivs, &(xi[0])); TF = LD(&(xi[WS(is, 100)]), ivs, &(xi[0])); Taf = VADD(TE, TF); Tu = VSUB(Ts, Tt); Tee = VSUB(Tae, Taf); TG = VSUB(TE, TF); Tag = VADD(Tae, Taf); } { V TJ, TK, Tal, TV, TW, Tam; TJ = LD(&(xi[WS(is, 124)]), ivs, &(xi[0])); TK = LD(&(xi[WS(is, 60)]), ivs, &(xi[0])); Tal = VADD(TJ, TK); TV = LD(&(xi[WS(is, 28)]), ivs, &(xi[0])); TW = LD(&(xi[WS(is, 92)]), ivs, &(xi[0])); Tam = VADD(TV, TW); TL = VSUB(TJ, TK); Teh = VSUB(Tal, Tam); TX = VSUB(TV, TW); Tan = VADD(Tal, Tam); } { V Tx, Tah, TA, Tai; { V Tv, Tw, Ty, Tz; Tv = LD(&(xi[WS(is, 20)]), ivs, &(xi[0])); Tw = LD(&(xi[WS(is, 84)]), ivs, &(xi[0])); Tx = VSUB(Tv, Tw); Tah = VADD(Tv, Tw); Ty = LD(&(xi[WS(is, 116)]), ivs, &(xi[0])); Tz = LD(&(xi[WS(is, 52)]), ivs, &(xi[0])); TA = VSUB(Ty, Tz); Tai = VADD(Ty, Tz); } TB = VMUL(LDK(KP707106781), VADD(Tx, TA)); Tef = VSUB(Tai, Tah); TD = VMUL(LDK(KP707106781), VSUB(TA, Tx)); Taj = VADD(Tah, Tai); } { V TO, Tao, TR, Tap; { V TM, TN, TP, TQ; TM = LD(&(xi[WS(is, 12)]), ivs, &(xi[0])); TN = LD(&(xi[WS(is, 76)]), ivs, &(xi[0])); TO = VSUB(TM, TN); Tao = VADD(TM, TN); TP = LD(&(xi[WS(is, 108)]), ivs, &(xi[0])); TQ = LD(&(xi[WS(is, 44)]), ivs, &(xi[0])); TR = VSUB(TP, TQ); Tap = VADD(TP, TQ); } TS = VMUL(LDK(KP707106781), VADD(TO, TR)); Tei = VSUB(Tap, Tao); TU = VMUL(LDK(KP707106781), VSUB(TR, TO)); Taq = VADD(Tao, Tap); } Td2 = VADD(Tag, Taj); Td3 = VADD(Tan, Taq); { V TC, TH, Tfr, Tfs; TC = VADD(Tu, TB); TH = VSUB(TD, TG); TI = VFMA(LDK(KP980785280), TC, VMUL(LDK(KP195090322), TH)); T4V = VFNMS(LDK(KP195090322), TC, VMUL(LDK(KP980785280), TH)); Tfr = VFNMS(LDK(KP382683432), Tee, VMUL(LDK(KP923879532), Tef)); Tfs = VFMA(LDK(KP382683432), Teh, VMUL(LDK(KP923879532), Tei)); Tft = VADD(Tfr, Tfs); Tgg = VSUB(Tfs, Tfr); } { V TT, TY, T73, T74; TT = VADD(TL, TS); TY = VSUB(TU, TX); TZ = VFNMS(LDK(KP195090322), TY, VMUL(LDK(KP980785280), TT)); T4W = VFMA(LDK(KP195090322), TT, VMUL(LDK(KP980785280), TY)); T73 = VSUB(TL, TS); T74 = VADD(TX, TU); T75 = VFNMS(LDK(KP555570233), T74, VMUL(LDK(KP831469612), T73)); T86 = VFMA(LDK(KP555570233), T73, VMUL(LDK(KP831469612), T74)); } Teg = VFMA(LDK(KP923879532), Tee, VMUL(LDK(KP382683432), Tef)); Tej = VFNMS(LDK(KP382683432), Tei, VMUL(LDK(KP923879532), Teh)); Tek = VADD(Teg, Tej); TgG = VSUB(Tej, Teg); { V T70, T71, Tak, Tar; T70 = VSUB(Tu, TB); T71 = VADD(TG, TD); T72 = VFMA(LDK(KP831469612), T70, VMUL(LDK(KP555570233), T71)); T85 = VFNMS(LDK(KP555570233), T70, VMUL(LDK(KP831469612), T71)); Tak = VSUB(Tag, Taj); Tar = VSUB(Tan, Taq); Tas = VMUL(LDK(KP707106781), VADD(Tak, Tar)); Tcd = VMUL(LDK(KP707106781), VSUB(Tar, Tak)); } } { V Tav, Tau, T1b, Taw, T1v, Tay, Tax, T18, Taz, T1w, T1j, Teo, T1z, TaD, T1q; V Ten, T1y, TaG, T14, T17, Tem, Tep; { V T19, T1a, T1t, T1u; T19 = LD(&(xi[WS(is, 34)]), ivs, &(xi[0])); T1a = LD(&(xi[WS(is, 98)]), ivs, &(xi[0])); Tav = VADD(T19, T1a); T1t = LD(&(xi[WS(is, 2)]), ivs, &(xi[0])); T1u = LD(&(xi[WS(is, 66)]), ivs, &(xi[0])); Tau = VADD(T1t, T1u); T1b = VSUB(T19, T1a); Taw = VADD(Tau, Tav); T1v = VSUB(T1t, T1u); } { V T12, T13, T15, T16; T12 = LD(&(xi[WS(is, 114)]), ivs, &(xi[0])); T13 = LD(&(xi[WS(is, 50)]), ivs, &(xi[0])); T14 = VSUB(T12, T13); Tay = VADD(T12, T13); T15 = LD(&(xi[WS(is, 18)]), ivs, &(xi[0])); T16 = LD(&(xi[WS(is, 82)]), ivs, &(xi[0])); T17 = VSUB(T15, T16); Tax = VADD(T15, T16); } T18 = VMUL(LDK(KP707106781), VSUB(T14, T17)); Taz = VADD(Tax, Tay); T1w = VMUL(LDK(KP707106781), VADD(T17, T14)); { V T1f, TaB, T1i, TaC; { V T1d, T1e, T1g, T1h; T1d = LD(&(xi[WS(is, 122)]), ivs, &(xi[0])); T1e = LD(&(xi[WS(is, 58)]), ivs, &(xi[0])); T1f = VSUB(T1d, T1e); TaB = VADD(T1d, T1e); T1g = LD(&(xi[WS(is, 26)]), ivs, &(xi[0])); T1h = LD(&(xi[WS(is, 90)]), ivs, &(xi[0])); T1i = VSUB(T1g, T1h); TaC = VADD(T1g, T1h); } T1j = VFNMS(LDK(KP923879532), T1i, VMUL(LDK(KP382683432), T1f)); Teo = VSUB(TaB, TaC); T1z = VFMA(LDK(KP923879532), T1f, VMUL(LDK(KP382683432), T1i)); TaD = VADD(TaB, TaC); } { V T1m, TaE, T1p, TaF; { V T1k, T1l, T1n, T1o; T1k = LD(&(xi[WS(is, 10)]), ivs, &(xi[0])); T1l = LD(&(xi[WS(is, 74)]), ivs, &(xi[0])); T1m = VSUB(T1k, T1l); TaE = VADD(T1k, T1l); T1n = LD(&(xi[WS(is, 42)]), ivs, &(xi[0])); T1o = LD(&(xi[WS(is, 106)]), ivs, &(xi[0])); T1p = VSUB(T1n, T1o); TaF = VADD(T1n, T1o); } T1q = VFMA(LDK(KP382683432), T1m, VMUL(LDK(KP923879532), T1p)); Ten = VSUB(TaE, TaF); T1y = VFNMS(LDK(KP382683432), T1p, VMUL(LDK(KP923879532), T1m)); TaG = VADD(TaE, TaF); } Tdp = VADD(Taw, Taz); Tdq = VADD(TaG, TaD); TdG = VSUB(Tdp, Tdq); Tem = VSUB(Tau, Tav); Tep = VMUL(LDK(KP707106781), VADD(Ten, Teo)); Teq = VADD(Tem, Tep); Tgm = VSUB(Tem, Tep); { V Ter, Tes, T1c, T1r; Ter = VSUB(Tay, Tax); Tes = VMUL(LDK(KP707106781), VSUB(Teo, Ten)); Tet = VADD(Ter, Tes); Tgl = VSUB(Tes, Ter); T1c = VSUB(T18, T1b); T1r = VSUB(T1j, T1q); T1s = VADD(T1c, T1r); T5P = VSUB(T1r, T1c); } { V T1x, T1A, T7b, T7c; T1x = VADD(T1v, T1w); T1A = VADD(T1y, T1z); T1B = VADD(T1x, T1A); T5Q = VSUB(T1x, T1A); T7b = VADD(T1b, T18); T7c = VSUB(T1z, T1y); T7d = VADD(T7b, T7c); T8Z = VSUB(T7c, T7b); } { V TaA, TaH, T78, T79; TaA = VSUB(Taw, Taz); TaH = VSUB(TaD, TaG); TaI = VFMA(LDK(KP923879532), TaA, VMUL(LDK(KP382683432), TaH)); Tcf = VFNMS(LDK(KP382683432), TaA, VMUL(LDK(KP923879532), TaH)); T78 = VSUB(T1v, T1w); T79 = VADD(T1q, T1j); T7a = VADD(T78, T79); T90 = VSUB(T78, T79); } } { V TaJ, TaK, T1F, TaL, T27, TaM, TaN, T1M, TaO, T24, T1U, Tew, T2a, TaV, T21; V Tex, T29, TaS, T1I, T1L, Tev, Tey; { V T1D, T1E, T25, T26; T1D = LD(&(xi[WS(is, 126)]), ivs, &(xi[0])); T1E = LD(&(xi[WS(is, 62)]), ivs, &(xi[0])); TaJ = VADD(T1D, T1E); T25 = LD(&(xi[WS(is, 30)]), ivs, &(xi[0])); T26 = LD(&(xi[WS(is, 94)]), ivs, &(xi[0])); TaK = VADD(T25, T26); T1F = VSUB(T1D, T1E); TaL = VADD(TaJ, TaK); T27 = VSUB(T25, T26); } { V T1G, T1H, T1J, T1K; T1G = LD(&(xi[WS(is, 14)]), ivs, &(xi[0])); T1H = LD(&(xi[WS(is, 78)]), ivs, &(xi[0])); T1I = VSUB(T1G, T1H); TaM = VADD(T1G, T1H); T1J = LD(&(xi[WS(is, 110)]), ivs, &(xi[0])); T1K = LD(&(xi[WS(is, 46)]), ivs, &(xi[0])); T1L = VSUB(T1J, T1K); TaN = VADD(T1J, T1K); } T1M = VMUL(LDK(KP707106781), VADD(T1I, T1L)); TaO = VADD(TaM, TaN); T24 = VMUL(LDK(KP707106781), VSUB(T1L, T1I)); { V T1Q, TaT, T1T, TaU; { V T1O, T1P, T1R, T1S; T1O = LD(&(xi[WS(is, 6)]), ivs, &(xi[0])); T1P = LD(&(xi[WS(is, 70)]), ivs, &(xi[0])); T1Q = VSUB(T1O, T1P); TaT = VADD(T1O, T1P); T1R = LD(&(xi[WS(is, 38)]), ivs, &(xi[0])); T1S = LD(&(xi[WS(is, 102)]), ivs, &(xi[0])); T1T = VSUB(T1R, T1S); TaU = VADD(T1R, T1S); } T1U = VFNMS(LDK(KP382683432), T1T, VMUL(LDK(KP923879532), T1Q)); Tew = VSUB(TaT, TaU); T2a = VFMA(LDK(KP382683432), T1Q, VMUL(LDK(KP923879532), T1T)); TaV = VADD(TaT, TaU); } { V T1X, TaQ, T20, TaR; { V T1V, T1W, T1Y, T1Z; T1V = LD(&(xi[WS(is, 118)]), ivs, &(xi[0])); T1W = LD(&(xi[WS(is, 54)]), ivs, &(xi[0])); T1X = VSUB(T1V, T1W); TaQ = VADD(T1V, T1W); T1Y = LD(&(xi[WS(is, 22)]), ivs, &(xi[0])); T1Z = LD(&(xi[WS(is, 86)]), ivs, &(xi[0])); T20 = VSUB(T1Y, T1Z); TaR = VADD(T1Y, T1Z); } T21 = VFMA(LDK(KP923879532), T1X, VMUL(LDK(KP382683432), T20)); Tex = VSUB(TaQ, TaR); T29 = VFNMS(LDK(KP923879532), T20, VMUL(LDK(KP382683432), T1X)); TaS = VADD(TaQ, TaR); } Tdm = VADD(TaL, TaO); Tdn = VADD(TaV, TaS); TdH = VSUB(Tdm, Tdn); Tev = VSUB(TaJ, TaK); Tey = VMUL(LDK(KP707106781), VADD(Tew, Tex)); Tez = VADD(Tev, Tey); Tgi = VSUB(Tev, Tey); { V TeA, TeB, T1N, T22; TeA = VSUB(TaN, TaM); TeB = VMUL(LDK(KP707106781), VSUB(Tex, Tew)); TeC = VADD(TeA, TeB); Tgj = VSUB(TeB, TeA); T1N = VADD(T1F, T1M); T22 = VADD(T1U, T21); T23 = VADD(T1N, T22); T5N = VSUB(T1N, T22); } { V T28, T2b, T7i, T7j; T28 = VSUB(T24, T27); T2b = VSUB(T29, T2a); T2c = VADD(T28, T2b); T5M = VSUB(T2b, T28); T7i = VADD(T27, T24); T7j = VSUB(T21, T1U); T7k = VADD(T7i, T7j); T8X = VSUB(T7j, T7i); } { V TaP, TaW, T7f, T7g; TaP = VSUB(TaL, TaO); TaW = VSUB(TaS, TaV); TaX = VFNMS(LDK(KP382683432), TaW, VMUL(LDK(KP923879532), TaP)); Tcg = VFMA(LDK(KP382683432), TaP, VMUL(LDK(KP923879532), TaW)); T7f = VSUB(T1F, T1M); T7g = VADD(T2a, T29); T7h = VADD(T7f, T7g); T8W = VSUB(T7f, T7g); } } { V T2J, TeL, T2V, Tb9, T30, TeO, T3c, Tbg, T2Q, TeM, T2S, Tbc, T37, TeP, T39; V Tbj; { V T2H, T2I, Tb7, T2T, T2U, Tb8; T2H = LD(&(xi[WS(is, 5)]), ivs, &(xi[WS(is, 1)])); T2I = LD(&(xi[WS(is, 69)]), ivs, &(xi[WS(is, 1)])); Tb7 = VADD(T2H, T2I); T2T = LD(&(xi[WS(is, 37)]), ivs, &(xi[WS(is, 1)])); T2U = LD(&(xi[WS(is, 101)]), ivs, &(xi[WS(is, 1)])); Tb8 = VADD(T2T, T2U); T2J = VSUB(T2H, T2I); TeL = VSUB(Tb7, Tb8); T2V = VSUB(T2T, T2U); Tb9 = VADD(Tb7, Tb8); } { V T2Y, T2Z, Tbe, T3a, T3b, Tbf; T2Y = LD(&(xi[WS(is, 125)]), ivs, &(xi[WS(is, 1)])); T2Z = LD(&(xi[WS(is, 61)]), ivs, &(xi[WS(is, 1)])); Tbe = VADD(T2Y, T2Z); T3a = LD(&(xi[WS(is, 29)]), ivs, &(xi[WS(is, 1)])); T3b = LD(&(xi[WS(is, 93)]), ivs, &(xi[WS(is, 1)])); Tbf = VADD(T3a, T3b); T30 = VSUB(T2Y, T2Z); TeO = VSUB(Tbe, Tbf); T3c = VSUB(T3a, T3b); Tbg = VADD(Tbe, Tbf); } { V T2M, Tba, T2P, Tbb; { V T2K, T2L, T2N, T2O; T2K = LD(&(xi[WS(is, 21)]), ivs, &(xi[WS(is, 1)])); T2L = LD(&(xi[WS(is, 85)]), ivs, &(xi[WS(is, 1)])); T2M = VSUB(T2K, T2L); Tba = VADD(T2K, T2L); T2N = LD(&(xi[WS(is, 117)]), ivs, &(xi[WS(is, 1)])); T2O = LD(&(xi[WS(is, 53)]), ivs, &(xi[WS(is, 1)])); T2P = VSUB(T2N, T2O); Tbb = VADD(T2N, T2O); } T2Q = VMUL(LDK(KP707106781), VADD(T2M, T2P)); TeM = VSUB(Tbb, Tba); T2S = VMUL(LDK(KP707106781), VSUB(T2P, T2M)); Tbc = VADD(Tba, Tbb); } { V T33, Tbh, T36, Tbi; { V T31, T32, T34, T35; T31 = LD(&(xi[WS(is, 13)]), ivs, &(xi[WS(is, 1)])); T32 = LD(&(xi[WS(is, 77)]), ivs, &(xi[WS(is, 1)])); T33 = VSUB(T31, T32); Tbh = VADD(T31, T32); T34 = LD(&(xi[WS(is, 109)]), ivs, &(xi[WS(is, 1)])); T35 = LD(&(xi[WS(is, 45)]), ivs, &(xi[WS(is, 1)])); T36 = VSUB(T34, T35); Tbi = VADD(T34, T35); } T37 = VMUL(LDK(KP707106781), VADD(T33, T36)); TeP = VSUB(Tbi, Tbh); T39 = VMUL(LDK(KP707106781), VSUB(T36, T33)); Tbj = VADD(Tbh, Tbi); } { V Tbd, Tbk, TeN, TeQ; Tbd = VSUB(Tb9, Tbc); Tbk = VSUB(Tbg, Tbj); Tbl = VMUL(LDK(KP707106781), VADD(Tbd, Tbk)); Tbu = VMUL(LDK(KP707106781), VSUB(Tbk, Tbd)); { V Td9, Tda, TeW, TeX; Td9 = VADD(Tb9, Tbc); Tda = VADD(Tbg, Tbj); Tdb = VADD(Td9, Tda); TdL = VSUB(Tda, Td9); TeW = VFNMS(LDK(KP382683432), TeL, VMUL(LDK(KP923879532), TeM)); TeX = VFMA(LDK(KP382683432), TeO, VMUL(LDK(KP923879532), TeP)); TeY = VADD(TeW, TeX); Tgu = VSUB(TeX, TeW); } TeN = VFMA(LDK(KP923879532), TeL, VMUL(LDK(KP382683432), TeM)); TeQ = VFNMS(LDK(KP382683432), TeP, VMUL(LDK(KP923879532), TeO)); TeR = VADD(TeN, TeQ); Tgq = VSUB(TeQ, TeN); { V T7t, T7C, T7w, T7D; { V T7r, T7s, T7u, T7v; T7r = VSUB(T2J, T2Q); T7s = VADD(T2V, T2S); T7t = VFMA(LDK(KP831469612), T7r, VMUL(LDK(KP555570233), T7s)); T7C = VFNMS(LDK(KP555570233), T7r, VMUL(LDK(KP831469612), T7s)); T7u = VSUB(T30, T37); T7v = VADD(T3c, T39); T7w = VFNMS(LDK(KP555570233), T7v, VMUL(LDK(KP831469612), T7u)); T7D = VFMA(LDK(KP555570233), T7u, VMUL(LDK(KP831469612), T7v)); } T7x = VADD(T7t, T7w); T98 = VSUB(T7D, T7C); T7E = VADD(T7C, T7D); T94 = VSUB(T7w, T7t); } { V T2X, T3q, T3e, T3r; { V T2R, T2W, T38, T3d; T2R = VADD(T2J, T2Q); T2W = VSUB(T2S, T2V); T2X = VFMA(LDK(KP980785280), T2R, VMUL(LDK(KP195090322), T2W)); T3q = VFNMS(LDK(KP195090322), T2R, VMUL(LDK(KP980785280), T2W)); T38 = VADD(T30, T37); T3d = VSUB(T39, T3c); T3e = VFNMS(LDK(KP195090322), T3d, VMUL(LDK(KP980785280), T38)); T3r = VFMA(LDK(KP195090322), T38, VMUL(LDK(KP980785280), T3d)); } T3f = VADD(T2X, T3e); T5Y = VSUB(T3r, T3q); T3s = VADD(T3q, T3r); T5U = VSUB(T3e, T2X); } } } { V T3Y, Tf6, T4a, TbG, T4f, Tf9, T4r, TbN, T45, Tf7, T47, TbJ, T4m, Tfa, T4o; V TbQ; { V T3W, T3X, TbE, T48, T49, TbF; T3W = LD(&(xi[WS(is, 3)]), ivs, &(xi[WS(is, 1)])); T3X = LD(&(xi[WS(is, 67)]), ivs, &(xi[WS(is, 1)])); TbE = VADD(T3W, T3X); T48 = LD(&(xi[WS(is, 35)]), ivs, &(xi[WS(is, 1)])); T49 = LD(&(xi[WS(is, 99)]), ivs, &(xi[WS(is, 1)])); TbF = VADD(T48, T49); T3Y = VSUB(T3W, T3X); Tf6 = VSUB(TbE, TbF); T4a = VSUB(T48, T49); TbG = VADD(TbE, TbF); } { V T4d, T4e, TbL, T4p, T4q, TbM; T4d = LD(&(xi[WS(is, 123)]), ivs, &(xi[WS(is, 1)])); T4e = LD(&(xi[WS(is, 59)]), ivs, &(xi[WS(is, 1)])); TbL = VADD(T4d, T4e); T4p = LD(&(xi[WS(is, 27)]), ivs, &(xi[WS(is, 1)])); T4q = LD(&(xi[WS(is, 91)]), ivs, &(xi[WS(is, 1)])); TbM = VADD(T4p, T4q); T4f = VSUB(T4d, T4e); Tf9 = VSUB(TbL, TbM); T4r = VSUB(T4p, T4q); TbN = VADD(TbL, TbM); } { V T41, TbH, T44, TbI; { V T3Z, T40, T42, T43; T3Z = LD(&(xi[WS(is, 19)]), ivs, &(xi[WS(is, 1)])); T40 = LD(&(xi[WS(is, 83)]), ivs, &(xi[WS(is, 1)])); T41 = VSUB(T3Z, T40); TbH = VADD(T3Z, T40); T42 = LD(&(xi[WS(is, 115)]), ivs, &(xi[WS(is, 1)])); T43 = LD(&(xi[WS(is, 51)]), ivs, &(xi[WS(is, 1)])); T44 = VSUB(T42, T43); TbI = VADD(T42, T43); } T45 = VMUL(LDK(KP707106781), VADD(T41, T44)); Tf7 = VSUB(TbI, TbH); T47 = VMUL(LDK(KP707106781), VSUB(T44, T41)); TbJ = VADD(TbH, TbI); } { V T4i, TbO, T4l, TbP; { V T4g, T4h, T4j, T4k; T4g = LD(&(xi[WS(is, 11)]), ivs, &(xi[WS(is, 1)])); T4h = LD(&(xi[WS(is, 75)]), ivs, &(xi[WS(is, 1)])); T4i = VSUB(T4g, T4h); TbO = VADD(T4g, T4h); T4j = LD(&(xi[WS(is, 107)]), ivs, &(xi[WS(is, 1)])); T4k = LD(&(xi[WS(is, 43)]), ivs, &(xi[WS(is, 1)])); T4l = VSUB(T4j, T4k); TbP = VADD(T4j, T4k); } T4m = VMUL(LDK(KP707106781), VADD(T4i, T4l)); Tfa = VSUB(TbP, TbO); T4o = VMUL(LDK(KP707106781), VSUB(T4l, T4i)); TbQ = VADD(TbO, TbP); } { V TbK, TbR, Tf8, Tfb; TbK = VSUB(TbG, TbJ); TbR = VSUB(TbN, TbQ); TbS = VMUL(LDK(KP707106781), VADD(TbK, TbR)); Tc1 = VMUL(LDK(KP707106781), VSUB(TbR, TbK)); { V Tdg, Tdh, Tfh, Tfi; Tdg = VADD(TbG, TbJ); Tdh = VADD(TbN, TbQ); Tdi = VADD(Tdg, Tdh); TdO = VSUB(Tdh, Tdg); Tfh = VFNMS(LDK(KP382683432), Tf6, VMUL(LDK(KP923879532), Tf7)); Tfi = VFMA(LDK(KP382683432), Tf9, VMUL(LDK(KP923879532), Tfa)); Tfj = VADD(Tfh, Tfi); Tgy = VSUB(Tfi, Tfh); } Tf8 = VFMA(LDK(KP923879532), Tf6, VMUL(LDK(KP382683432), Tf7)); Tfb = VFNMS(LDK(KP382683432), Tfa, VMUL(LDK(KP923879532), Tf9)); Tfc = VADD(Tf8, Tfb); TgA = VSUB(Tfb, Tf8); { V T7M, T7V, T7P, T7W; { V T7K, T7L, T7N, T7O; T7K = VSUB(T3Y, T45); T7L = VADD(T4a, T47); T7M = VFMA(LDK(KP831469612), T7K, VMUL(LDK(KP555570233), T7L)); T7V = VFNMS(LDK(KP555570233), T7K, VMUL(LDK(KP831469612), T7L)); T7N = VSUB(T4f, T4m); T7O = VADD(T4r, T4o); T7P = VFNMS(LDK(KP555570233), T7O, VMUL(LDK(KP831469612), T7N)); T7W = VFMA(LDK(KP555570233), T7N, VMUL(LDK(KP831469612), T7O)); } T7Q = VADD(T7M, T7P); T9e = VSUB(T7P, T7M); T7X = VADD(T7V, T7W); T9c = VSUB(T7W, T7V); } { V T4c, T4F, T4t, T4G; { V T46, T4b, T4n, T4s; T46 = VADD(T3Y, T45); T4b = VSUB(T47, T4a); T4c = VFMA(LDK(KP980785280), T46, VMUL(LDK(KP195090322), T4b)); T4F = VFNMS(LDK(KP195090322), T46, VMUL(LDK(KP980785280), T4b)); T4n = VADD(T4f, T4m); T4s = VSUB(T4o, T4r); T4t = VFNMS(LDK(KP195090322), T4s, VMUL(LDK(KP980785280), T4n)); T4G = VFMA(LDK(KP195090322), T4n, VMUL(LDK(KP980785280), T4s)); } T4u = VADD(T4c, T4t); T64 = VSUB(T4t, T4c); T4H = VADD(T4F, T4G); T62 = VSUB(T4G, T4F); } } } { V Td5, Tdx, TdC, TdE, Tdk, Tdt, Tds, Tdy, Tdz, TdD; { V Td1, Td4, TdA, TdB; Td1 = VADD(TcZ, Td0); Td4 = VADD(Td2, Td3); Td5 = VSUB(Td1, Td4); Tdx = VADD(Td1, Td4); TdA = VADD(Td8, Tdb); TdB = VADD(Tdf, Tdi); TdC = VADD(TdA, TdB); TdE = VBYI(VSUB(TdB, TdA)); } { V Tdc, Tdj, Tdo, Tdr; Tdc = VSUB(Td8, Tdb); Tdj = VSUB(Tdf, Tdi); Tdk = VMUL(LDK(KP707106781), VADD(Tdc, Tdj)); Tdt = VMUL(LDK(KP707106781), VSUB(Tdj, Tdc)); Tdo = VADD(Tdm, Tdn); Tdr = VADD(Tdp, Tdq); Tds = VSUB(Tdo, Tdr); Tdy = VADD(Tdr, Tdo); } Tdz = VADD(Tdx, Tdy); ST(&(xo[WS(os, 64)]), VSUB(Tdz, TdC), ovs, &(xo[0])); ST(&(xo[0]), VADD(Tdz, TdC), ovs, &(xo[0])); TdD = VSUB(Tdx, Tdy); ST(&(xo[WS(os, 96)]), VSUB(TdD, TdE), ovs, &(xo[0])); ST(&(xo[WS(os, 32)]), VADD(TdD, TdE), ovs, &(xo[0])); { V Tdl, Tdu, Tdv, Tdw; Tdl = VADD(Td5, Tdk); Tdu = VBYI(VADD(Tds, Tdt)); ST(&(xo[WS(os, 112)]), VSUB(Tdl, Tdu), ovs, &(xo[0])); ST(&(xo[WS(os, 16)]), VADD(Tdl, Tdu), ovs, &(xo[0])); Tdv = VSUB(Td5, Tdk); Tdw = VBYI(VSUB(Tdt, Tds)); ST(&(xo[WS(os, 80)]), VSUB(Tdv, Tdw), ovs, &(xo[0])); ST(&(xo[WS(os, 48)]), VADD(Tdv, Tdw), ovs, &(xo[0])); } } { V TdJ, Te4, TdX, Te5, TdQ, Te1, TdU, Te2; { V TdF, TdI, TdV, TdW; TdF = VSUB(TcZ, Td0); TdI = VMUL(LDK(KP707106781), VADD(TdG, TdH)); TdJ = VADD(TdF, TdI); Te4 = VSUB(TdF, TdI); TdV = VFNMS(LDK(KP382683432), TdK, VMUL(LDK(KP923879532), TdL)); TdW = VFMA(LDK(KP382683432), TdN, VMUL(LDK(KP923879532), TdO)); TdX = VADD(TdV, TdW); Te5 = VSUB(TdW, TdV); } { V TdM, TdP, TdS, TdT; TdM = VFMA(LDK(KP923879532), TdK, VMUL(LDK(KP382683432), TdL)); TdP = VFNMS(LDK(KP382683432), TdO, VMUL(LDK(KP923879532), TdN)); TdQ = VADD(TdM, TdP); Te1 = VSUB(TdP, TdM); TdS = VSUB(Td3, Td2); TdT = VMUL(LDK(KP707106781), VSUB(TdH, TdG)); TdU = VADD(TdS, TdT); Te2 = VSUB(TdT, TdS); } { V TdR, TdY, Te7, Te8; TdR = VADD(TdJ, TdQ); TdY = VBYI(VADD(TdU, TdX)); ST(&(xo[WS(os, 120)]), VSUB(TdR, TdY), ovs, &(xo[0])); ST(&(xo[WS(os, 8)]), VADD(TdR, TdY), ovs, &(xo[0])); Te7 = VBYI(VADD(Te2, Te1)); Te8 = VADD(Te4, Te5); ST(&(xo[WS(os, 24)]), VADD(Te7, Te8), ovs, &(xo[0])); ST(&(xo[WS(os, 104)]), VSUB(Te8, Te7), ovs, &(xo[0])); } { V TdZ, Te0, Te3, Te6; TdZ = VSUB(TdJ, TdQ); Te0 = VBYI(VSUB(TdX, TdU)); ST(&(xo[WS(os, 72)]), VSUB(TdZ, Te0), ovs, &(xo[0])); ST(&(xo[WS(os, 56)]), VADD(TdZ, Te0), ovs, &(xo[0])); Te3 = VBYI(VSUB(Te1, Te2)); Te6 = VSUB(Te4, Te5); ST(&(xo[WS(os, 40)]), VADD(Te3, Te6), ovs, &(xo[0])); ST(&(xo[WS(os, 88)]), VSUB(Te6, Te3), ovs, &(xo[0])); } } { V TaZ, Tcs, Tci, Tcq, Tc4, Tct, Tcl, Tcp; { V Tat, TaY, Tce, Tch; Tat = VADD(Tad, Tas); TaY = VADD(TaI, TaX); TaZ = VADD(Tat, TaY); Tcs = VSUB(Tat, TaY); Tce = VADD(Tcc, Tcd); Tch = VADD(Tcf, Tcg); Tci = VADD(Tce, Tch); Tcq = VSUB(Tch, Tce); { V Tbw, Tcj, Tc3, Tck; { V Tbm, Tbv, TbT, Tc2; Tbm = VADD(Tb6, Tbl); Tbv = VADD(Tbt, Tbu); Tbw = VFMA(LDK(KP980785280), Tbm, VMUL(LDK(KP195090322), Tbv)); Tcj = VFNMS(LDK(KP195090322), Tbm, VMUL(LDK(KP980785280), Tbv)); TbT = VADD(TbD, TbS); Tc2 = VADD(Tc0, Tc1); Tc3 = VFNMS(LDK(KP195090322), Tc2, VMUL(LDK(KP980785280), TbT)); Tck = VFMA(LDK(KP195090322), TbT, VMUL(LDK(KP980785280), Tc2)); } Tc4 = VADD(Tbw, Tc3); Tct = VSUB(Tck, Tcj); Tcl = VADD(Tcj, Tck); Tcp = VSUB(Tc3, Tbw); } } { V Tc5, Tcm, Tcv, Tcw; Tc5 = VADD(TaZ, Tc4); Tcm = VBYI(VADD(Tci, Tcl)); ST(&(xo[WS(os, 124)]), VSUB(Tc5, Tcm), ovs, &(xo[0])); ST(&(xo[WS(os, 4)]), VADD(Tc5, Tcm), ovs, &(xo[0])); Tcv = VBYI(VADD(Tcq, Tcp)); Tcw = VADD(Tcs, Tct); ST(&(xo[WS(os, 28)]), VADD(Tcv, Tcw), ovs, &(xo[0])); ST(&(xo[WS(os, 100)]), VSUB(Tcw, Tcv), ovs, &(xo[0])); } { V Tcn, Tco, Tcr, Tcu; Tcn = VSUB(TaZ, Tc4); Tco = VBYI(VSUB(Tcl, Tci)); ST(&(xo[WS(os, 68)]), VSUB(Tcn, Tco), ovs, &(xo[0])); ST(&(xo[WS(os, 60)]), VADD(Tcn, Tco), ovs, &(xo[0])); Tcr = VBYI(VSUB(Tcp, Tcq)); Tcu = VSUB(Tcs, Tct); ST(&(xo[WS(os, 36)]), VADD(Tcr, Tcu), ovs, &(xo[0])); ST(&(xo[WS(os, 92)]), VSUB(Tcu, Tcr), ovs, &(xo[0])); } } { V Tcz, TcU, TcK, TcS, TcG, TcV, TcN, TcR; { V Tcx, Tcy, TcI, TcJ; Tcx = VSUB(Tad, Tas); Tcy = VSUB(Tcg, Tcf); Tcz = VADD(Tcx, Tcy); TcU = VSUB(Tcx, Tcy); TcI = VSUB(Tcd, Tcc); TcJ = VSUB(TaX, TaI); TcK = VADD(TcI, TcJ); TcS = VSUB(TcJ, TcI); { V TcC, TcL, TcF, TcM; { V TcA, TcB, TcD, TcE; TcA = VSUB(Tb6, Tbl); TcB = VSUB(Tbu, Tbt); TcC = VFMA(LDK(KP831469612), TcA, VMUL(LDK(KP555570233), TcB)); TcL = VFNMS(LDK(KP555570233), TcA, VMUL(LDK(KP831469612), TcB)); TcD = VSUB(TbD, TbS); TcE = VSUB(Tc1, Tc0); TcF = VFNMS(LDK(KP555570233), TcE, VMUL(LDK(KP831469612), TcD)); TcM = VFMA(LDK(KP555570233), TcD, VMUL(LDK(KP831469612), TcE)); } TcG = VADD(TcC, TcF); TcV = VSUB(TcM, TcL); TcN = VADD(TcL, TcM); TcR = VSUB(TcF, TcC); } } { V TcH, TcO, TcX, TcY; TcH = VADD(Tcz, TcG); TcO = VBYI(VADD(TcK, TcN)); ST(&(xo[WS(os, 116)]), VSUB(TcH, TcO), ovs, &(xo[0])); ST(&(xo[WS(os, 12)]), VADD(TcH, TcO), ovs, &(xo[0])); TcX = VBYI(VADD(TcS, TcR)); TcY = VADD(TcU, TcV); ST(&(xo[WS(os, 20)]), VADD(TcX, TcY), ovs, &(xo[0])); ST(&(xo[WS(os, 108)]), VSUB(TcY, TcX), ovs, &(xo[0])); } { V TcP, TcQ, TcT, TcW; TcP = VSUB(Tcz, TcG); TcQ = VBYI(VSUB(TcN, TcK)); ST(&(xo[WS(os, 76)]), VSUB(TcP, TcQ), ovs, &(xo[0])); ST(&(xo[WS(os, 52)]), VADD(TcP, TcQ), ovs, &(xo[0])); TcT = VBYI(VSUB(TcR, TcS)); TcW = VSUB(TcU, TcV); ST(&(xo[WS(os, 44)]), VADD(TcT, TcW), ovs, &(xo[0])); ST(&(xo[WS(os, 84)]), VSUB(TcW, TcT), ovs, &(xo[0])); } } { V TeF, Tg8, TfI, Tg0, Tfy, Tga, TfG, TfP, Tfm, TfJ, TfB, TfF, TfW, Tgb, Tg3; V Tg7; { V Tel, TfY, TeE, TfZ, Teu, TeD; Tel = VADD(Ted, Tek); TfY = VSUB(Tft, Tfq); Teu = VFMA(LDK(KP980785280), Teq, VMUL(LDK(KP195090322), Tet)); TeD = VFNMS(LDK(KP195090322), TeC, VMUL(LDK(KP980785280), Tez)); TeE = VADD(Teu, TeD); TfZ = VSUB(TeD, Teu); TeF = VADD(Tel, TeE); Tg8 = VSUB(TfZ, TfY); TfI = VSUB(Tel, TeE); Tg0 = VADD(TfY, TfZ); } { V Tfu, TfN, Tfx, TfO, Tfv, Tfw; Tfu = VADD(Tfq, Tft); TfN = VSUB(Ted, Tek); Tfv = VFNMS(LDK(KP195090322), Teq, VMUL(LDK(KP980785280), Tet)); Tfw = VFMA(LDK(KP195090322), Tez, VMUL(LDK(KP980785280), TeC)); Tfx = VADD(Tfv, Tfw); TfO = VSUB(Tfw, Tfv); Tfy = VADD(Tfu, Tfx); Tga = VSUB(TfN, TfO); TfG = VSUB(Tfx, Tfu); TfP = VADD(TfN, TfO); } { V Tf0, Tfz, Tfl, TfA; { V TeS, TeZ, Tfd, Tfk; TeS = VADD(TeK, TeR); TeZ = VADD(TeV, TeY); Tf0 = VFMA(LDK(KP995184726), TeS, VMUL(LDK(KP098017140), TeZ)); Tfz = VFNMS(LDK(KP098017140), TeS, VMUL(LDK(KP995184726), TeZ)); Tfd = VADD(Tf5, Tfc); Tfk = VADD(Tfg, Tfj); Tfl = VFNMS(LDK(KP098017140), Tfk, VMUL(LDK(KP995184726), Tfd)); TfA = VFMA(LDK(KP098017140), Tfd, VMUL(LDK(KP995184726), Tfk)); } Tfm = VADD(Tf0, Tfl); TfJ = VSUB(TfA, Tfz); TfB = VADD(Tfz, TfA); TfF = VSUB(Tfl, Tf0); } { V TfS, Tg1, TfV, Tg2; { V TfQ, TfR, TfT, TfU; TfQ = VSUB(TeK, TeR); TfR = VSUB(TeY, TeV); TfS = VFMA(LDK(KP773010453), TfQ, VMUL(LDK(KP634393284), TfR)); Tg1 = VFNMS(LDK(KP634393284), TfQ, VMUL(LDK(KP773010453), TfR)); TfT = VSUB(Tf5, Tfc); TfU = VSUB(Tfj, Tfg); TfV = VFNMS(LDK(KP634393284), TfU, VMUL(LDK(KP773010453), TfT)); Tg2 = VFMA(LDK(KP634393284), TfT, VMUL(LDK(KP773010453), TfU)); } TfW = VADD(TfS, TfV); Tgb = VSUB(Tg2, Tg1); Tg3 = VADD(Tg1, Tg2); Tg7 = VSUB(TfV, TfS); } { V Tfn, TfC, Tg9, Tgc; Tfn = VADD(TeF, Tfm); TfC = VBYI(VADD(Tfy, TfB)); ST(&(xo[WS(os, 126)]), VSUB(Tfn, TfC), ovs, &(xo[0])); ST(&(xo[WS(os, 2)]), VADD(Tfn, TfC), ovs, &(xo[0])); Tg9 = VBYI(VSUB(Tg7, Tg8)); Tgc = VSUB(Tga, Tgb); ST(&(xo[WS(os, 46)]), VADD(Tg9, Tgc), ovs, &(xo[0])); ST(&(xo[WS(os, 82)]), VSUB(Tgc, Tg9), ovs, &(xo[0])); } { V Tgd, Tge, TfD, TfE; Tgd = VBYI(VADD(Tg8, Tg7)); Tge = VADD(Tga, Tgb); ST(&(xo[WS(os, 18)]), VADD(Tgd, Tge), ovs, &(xo[0])); ST(&(xo[WS(os, 110)]), VSUB(Tge, Tgd), ovs, &(xo[0])); TfD = VSUB(TeF, Tfm); TfE = VBYI(VSUB(TfB, Tfy)); ST(&(xo[WS(os, 66)]), VSUB(TfD, TfE), ovs, &(xo[0])); ST(&(xo[WS(os, 62)]), VADD(TfD, TfE), ovs, &(xo[0])); } { V TfH, TfK, TfX, Tg4; TfH = VBYI(VSUB(TfF, TfG)); TfK = VSUB(TfI, TfJ); ST(&(xo[WS(os, 34)]), VADD(TfH, TfK), ovs, &(xo[0])); ST(&(xo[WS(os, 94)]), VSUB(TfK, TfH), ovs, &(xo[0])); TfX = VADD(TfP, TfW); Tg4 = VBYI(VADD(Tg0, Tg3)); ST(&(xo[WS(os, 114)]), VSUB(TfX, Tg4), ovs, &(xo[0])); ST(&(xo[WS(os, 14)]), VADD(TfX, Tg4), ovs, &(xo[0])); } { V Tg5, Tg6, TfL, TfM; Tg5 = VSUB(TfP, TfW); Tg6 = VBYI(VSUB(Tg3, Tg0)); ST(&(xo[WS(os, 78)]), VSUB(Tg5, Tg6), ovs, &(xo[0])); ST(&(xo[WS(os, 50)]), VADD(Tg5, Tg6), ovs, &(xo[0])); TfL = VBYI(VADD(TfG, TfF)); TfM = VADD(TfI, TfJ); ST(&(xo[WS(os, 30)]), VADD(TfL, TfM), ovs, &(xo[0])); ST(&(xo[WS(os, 98)]), VSUB(TfM, TfL), ovs, &(xo[0])); } } { V Tgp, Thm, TgW, The, TgM, Tho, TgU, Th3, TgE, TgX, TgP, TgT, Tha, Thp, Thh; V Thl; { V Tgh, Thc, Tgo, Thd, Tgk, Tgn; Tgh = VSUB(Tgf, Tgg); Thc = VADD(TgH, TgG); Tgk = VFMA(LDK(KP555570233), Tgi, VMUL(LDK(KP831469612), Tgj)); Tgn = VFNMS(LDK(KP555570233), Tgm, VMUL(LDK(KP831469612), Tgl)); Tgo = VSUB(Tgk, Tgn); Thd = VADD(Tgn, Tgk); Tgp = VADD(Tgh, Tgo); Thm = VSUB(Thd, Thc); TgW = VSUB(Tgh, Tgo); The = VADD(Thc, Thd); } { V TgI, Th1, TgL, Th2, TgJ, TgK; TgI = VSUB(TgG, TgH); Th1 = VADD(Tgf, Tgg); TgJ = VFNMS(LDK(KP555570233), Tgj, VMUL(LDK(KP831469612), Tgi)); TgK = VFMA(LDK(KP831469612), Tgm, VMUL(LDK(KP555570233), Tgl)); TgL = VSUB(TgJ, TgK); Th2 = VADD(TgK, TgJ); TgM = VADD(TgI, TgL); Tho = VSUB(Th1, Th2); TgU = VSUB(TgL, TgI); Th3 = VADD(Th1, Th2); } { V Tgw, TgN, TgD, TgO; { V Tgs, Tgv, Tgz, TgC; Tgs = VSUB(Tgq, Tgr); Tgv = VSUB(Tgt, Tgu); Tgw = VFMA(LDK(KP471396736), Tgs, VMUL(LDK(KP881921264), Tgv)); TgN = VFNMS(LDK(KP471396736), Tgv, VMUL(LDK(KP881921264), Tgs)); Tgz = VSUB(Tgx, Tgy); TgC = VSUB(TgA, TgB); TgD = VFNMS(LDK(KP471396736), TgC, VMUL(LDK(KP881921264), Tgz)); TgO = VFMA(LDK(KP881921264), TgC, VMUL(LDK(KP471396736), Tgz)); } TgE = VADD(Tgw, TgD); TgX = VSUB(TgO, TgN); TgP = VADD(TgN, TgO); TgT = VSUB(TgD, Tgw); } { V Th6, Thf, Th9, Thg; { V Th4, Th5, Th7, Th8; Th4 = VADD(Tgr, Tgq); Th5 = VADD(Tgt, Tgu); Th6 = VFMA(LDK(KP290284677), Th4, VMUL(LDK(KP956940335), Th5)); Thf = VFNMS(LDK(KP290284677), Th5, VMUL(LDK(KP956940335), Th4)); Th7 = VADD(Tgx, Tgy); Th8 = VADD(TgB, TgA); Th9 = VFNMS(LDK(KP290284677), Th8, VMUL(LDK(KP956940335), Th7)); Thg = VFMA(LDK(KP956940335), Th8, VMUL(LDK(KP290284677), Th7)); } Tha = VADD(Th6, Th9); Thp = VSUB(Thg, Thf); Thh = VADD(Thf, Thg); Thl = VSUB(Th9, Th6); } { V TgF, TgQ, Thn, Thq; TgF = VADD(Tgp, TgE); TgQ = VBYI(VADD(TgM, TgP)); ST(&(xo[WS(os, 118)]), VSUB(TgF, TgQ), ovs, &(xo[0])); ST(&(xo[WS(os, 10)]), VADD(TgF, TgQ), ovs, &(xo[0])); Thn = VBYI(VSUB(Thl, Thm)); Thq = VSUB(Tho, Thp); ST(&(xo[WS(os, 38)]), VADD(Thn, Thq), ovs, &(xo[0])); ST(&(xo[WS(os, 90)]), VSUB(Thq, Thn), ovs, &(xo[0])); } { V Thr, Ths, TgR, TgS; Thr = VBYI(VADD(Thm, Thl)); Ths = VADD(Tho, Thp); ST(&(xo[WS(os, 26)]), VADD(Thr, Ths), ovs, &(xo[0])); ST(&(xo[WS(os, 102)]), VSUB(Ths, Thr), ovs, &(xo[0])); TgR = VSUB(Tgp, TgE); TgS = VBYI(VSUB(TgP, TgM)); ST(&(xo[WS(os, 74)]), VSUB(TgR, TgS), ovs, &(xo[0])); ST(&(xo[WS(os, 54)]), VADD(TgR, TgS), ovs, &(xo[0])); } { V TgV, TgY, Thb, Thi; TgV = VBYI(VSUB(TgT, TgU)); TgY = VSUB(TgW, TgX); ST(&(xo[WS(os, 42)]), VADD(TgV, TgY), ovs, &(xo[0])); ST(&(xo[WS(os, 86)]), VSUB(TgY, TgV), ovs, &(xo[0])); Thb = VADD(Th3, Tha); Thi = VBYI(VADD(The, Thh)); ST(&(xo[WS(os, 122)]), VSUB(Thb, Thi), ovs, &(xo[0])); ST(&(xo[WS(os, 6)]), VADD(Thb, Thi), ovs, &(xo[0])); } { V Thj, Thk, TgZ, Th0; Thj = VSUB(Th3, Tha); Thk = VBYI(VSUB(Thh, The)); ST(&(xo[WS(os, 70)]), VSUB(Thj, Thk), ovs, &(xo[0])); ST(&(xo[WS(os, 58)]), VADD(Thj, Thk), ovs, &(xo[0])); TgZ = VBYI(VADD(TgU, TgT)); Th0 = VADD(TgW, TgX); ST(&(xo[WS(os, 22)]), VADD(TgZ, Th0), ovs, &(xo[0])); ST(&(xo[WS(os, 106)]), VSUB(Th0, TgZ), ovs, &(xo[0])); } } { V T80, T8n, T8f, T8j, T8A, T8P, T8H, T8L, T7n, T8M, T8O, T8c, T8k, T8t, T8E; V T8m; { V T7G, T8d, T7Z, T8e; { V T7y, T7F, T7R, T7Y; T7y = VADD(T7q, T7x); T7F = VADD(T7B, T7E); T7G = VFMA(LDK(KP989176509), T7y, VMUL(LDK(KP146730474), T7F)); T8d = VFNMS(LDK(KP146730474), T7y, VMUL(LDK(KP989176509), T7F)); T7R = VADD(T7J, T7Q); T7Y = VADD(T7U, T7X); T7Z = VFNMS(LDK(KP146730474), T7Y, VMUL(LDK(KP989176509), T7R)); T8e = VFMA(LDK(KP146730474), T7R, VMUL(LDK(KP989176509), T7Y)); } T80 = VADD(T7G, T7Z); T8n = VSUB(T8e, T8d); T8f = VADD(T8d, T8e); T8j = VSUB(T7Z, T7G); } { V T8w, T8F, T8z, T8G; { V T8u, T8v, T8x, T8y; T8u = VSUB(T7q, T7x); T8v = VSUB(T7E, T7B); T8w = VFMA(LDK(KP803207531), T8u, VMUL(LDK(KP595699304), T8v)); T8F = VFNMS(LDK(KP595699304), T8u, VMUL(LDK(KP803207531), T8v)); T8x = VSUB(T7J, T7Q); T8y = VSUB(T7X, T7U); T8z = VFNMS(LDK(KP595699304), T8y, VMUL(LDK(KP803207531), T8x)); T8G = VFMA(LDK(KP595699304), T8x, VMUL(LDK(KP803207531), T8y)); } T8A = VADD(T8w, T8z); T8P = VSUB(T8G, T8F); T8H = VADD(T8F, T8G); T8L = VSUB(T8z, T8w); } { V T77, T8r, T88, T8C, T7m, T8D, T8b, T8s, T76, T87; T76 = VADD(T72, T75); T77 = VADD(T6Z, T76); T8r = VSUB(T6Z, T76); T87 = VADD(T85, T86); T88 = VADD(T84, T87); T8C = VSUB(T87, T84); { V T7e, T7l, T89, T8a; T7e = VFMA(LDK(KP956940335), T7a, VMUL(LDK(KP290284677), T7d)); T7l = VFNMS(LDK(KP290284677), T7k, VMUL(LDK(KP956940335), T7h)); T7m = VADD(T7e, T7l); T8D = VSUB(T7l, T7e); T89 = VFNMS(LDK(KP290284677), T7a, VMUL(LDK(KP956940335), T7d)); T8a = VFMA(LDK(KP290284677), T7h, VMUL(LDK(KP956940335), T7k)); T8b = VADD(T89, T8a); T8s = VSUB(T8a, T89); } T7n = VADD(T77, T7m); T8M = VSUB(T8D, T8C); T8O = VSUB(T8r, T8s); T8c = VADD(T88, T8b); T8k = VSUB(T8b, T88); T8t = VADD(T8r, T8s); T8E = VADD(T8C, T8D); T8m = VSUB(T77, T7m); } { V T81, T8g, T8N, T8Q; T81 = VADD(T7n, T80); T8g = VBYI(VADD(T8c, T8f)); ST(&(xo[WS(os, 125)]), VSUB(T81, T8g), ovs, &(xo[WS(os, 1)])); ST(&(xo[WS(os, 3)]), VADD(T81, T8g), ovs, &(xo[WS(os, 1)])); T8N = VBYI(VSUB(T8L, T8M)); T8Q = VSUB(T8O, T8P); ST(&(xo[WS(os, 45)]), VADD(T8N, T8Q), ovs, &(xo[WS(os, 1)])); ST(&(xo[WS(os, 83)]), VSUB(T8Q, T8N), ovs, &(xo[WS(os, 1)])); } { V T8R, T8S, T8h, T8i; T8R = VBYI(VADD(T8M, T8L)); T8S = VADD(T8O, T8P); ST(&(xo[WS(os, 19)]), VADD(T8R, T8S), ovs, &(xo[WS(os, 1)])); ST(&(xo[WS(os, 109)]), VSUB(T8S, T8R), ovs, &(xo[WS(os, 1)])); T8h = VSUB(T7n, T80); T8i = VBYI(VSUB(T8f, T8c)); ST(&(xo[WS(os, 67)]), VSUB(T8h, T8i), ovs, &(xo[WS(os, 1)])); ST(&(xo[WS(os, 61)]), VADD(T8h, T8i), ovs, &(xo[WS(os, 1)])); } { V T8l, T8o, T8B, T8I; T8l = VBYI(VSUB(T8j, T8k)); T8o = VSUB(T8m, T8n); ST(&(xo[WS(os, 35)]), VADD(T8l, T8o), ovs, &(xo[WS(os, 1)])); ST(&(xo[WS(os, 93)]), VSUB(T8o, T8l), ovs, &(xo[WS(os, 1)])); T8B = VADD(T8t, T8A); T8I = VBYI(VADD(T8E, T8H)); ST(&(xo[WS(os, 115)]), VSUB(T8B, T8I), ovs, &(xo[WS(os, 1)])); ST(&(xo[WS(os, 13)]), VADD(T8B, T8I), ovs, &(xo[WS(os, 1)])); } { V T8J, T8K, T8p, T8q; T8J = VSUB(T8t, T8A); T8K = VBYI(VSUB(T8H, T8E)); ST(&(xo[WS(os, 77)]), VSUB(T8J, T8K), ovs, &(xo[WS(os, 1)])); ST(&(xo[WS(os, 51)]), VADD(T8J, T8K), ovs, &(xo[WS(os, 1)])); T8p = VBYI(VADD(T8k, T8j)); T8q = VADD(T8m, T8n); ST(&(xo[WS(os, 29)]), VADD(T8p, T8q), ovs, &(xo[WS(os, 1)])); ST(&(xo[WS(os, 99)]), VSUB(T8q, T8p), ovs, &(xo[WS(os, 1)])); } } { V T4K, T5d, T55, T59, T5q, T5F, T5x, T5B, T2f, T5C, T5E, T52, T5a, T5j, T5u; V T5c; { V T3u, T53, T4J, T54; { V T3g, T3t, T4v, T4I; T3g = VADD(T2G, T3f); T3t = VADD(T3p, T3s); T3u = VFMA(LDK(KP998795456), T3g, VMUL(LDK(KP049067674), T3t)); T53 = VFNMS(LDK(KP049067674), T3g, VMUL(LDK(KP998795456), T3t)); T4v = VADD(T3V, T4u); T4I = VADD(T4E, T4H); T4J = VFNMS(LDK(KP049067674), T4I, VMUL(LDK(KP998795456), T4v)); T54 = VFMA(LDK(KP049067674), T4v, VMUL(LDK(KP998795456), T4I)); } T4K = VADD(T3u, T4J); T5d = VSUB(T54, T53); T55 = VADD(T53, T54); T59 = VSUB(T4J, T3u); } { V T5m, T5v, T5p, T5w; { V T5k, T5l, T5n, T5o; T5k = VSUB(T2G, T3f); T5l = VSUB(T3s, T3p); T5m = VFMA(LDK(KP740951125), T5k, VMUL(LDK(KP671558954), T5l)); T5v = VFNMS(LDK(KP671558954), T5k, VMUL(LDK(KP740951125), T5l)); T5n = VSUB(T3V, T4u); T5o = VSUB(T4H, T4E); T5p = VFNMS(LDK(KP671558954), T5o, VMUL(LDK(KP740951125), T5n)); T5w = VFMA(LDK(KP671558954), T5n, VMUL(LDK(KP740951125), T5o)); } T5q = VADD(T5m, T5p); T5F = VSUB(T5w, T5v); T5x = VADD(T5v, T5w); T5B = VSUB(T5p, T5m); } { V T11, T5h, T4Y, T5s, T2e, T5t, T51, T5i, T10, T4X; T10 = VADD(TI, TZ); T11 = VADD(Tr, T10); T5h = VSUB(Tr, T10); T4X = VADD(T4V, T4W); T4Y = VADD(T4U, T4X); T5s = VSUB(T4X, T4U); { V T1C, T2d, T4Z, T50; T1C = VFMA(LDK(KP098017140), T1s, VMUL(LDK(KP995184726), T1B)); T2d = VFNMS(LDK(KP098017140), T2c, VMUL(LDK(KP995184726), T23)); T2e = VADD(T1C, T2d); T5t = VSUB(T2d, T1C); T4Z = VFNMS(LDK(KP098017140), T1B, VMUL(LDK(KP995184726), T1s)); T50 = VFMA(LDK(KP995184726), T2c, VMUL(LDK(KP098017140), T23)); T51 = VADD(T4Z, T50); T5i = VSUB(T50, T4Z); } T2f = VADD(T11, T2e); T5C = VSUB(T5t, T5s); T5E = VSUB(T5h, T5i); T52 = VADD(T4Y, T51); T5a = VSUB(T51, T4Y); T5j = VADD(T5h, T5i); T5u = VADD(T5s, T5t); T5c = VSUB(T11, T2e); } { V T4L, T56, T5D, T5G; T4L = VADD(T2f, T4K); T56 = VBYI(VADD(T52, T55)); ST(&(xo[WS(os, 127)]), VSUB(T4L, T56), ovs, &(xo[WS(os, 1)])); ST(&(xo[WS(os, 1)]), VADD(T4L, T56), ovs, &(xo[WS(os, 1)])); T5D = VBYI(VSUB(T5B, T5C)); T5G = VSUB(T5E, T5F); ST(&(xo[WS(os, 47)]), VADD(T5D, T5G), ovs, &(xo[WS(os, 1)])); ST(&(xo[WS(os, 81)]), VSUB(T5G, T5D), ovs, &(xo[WS(os, 1)])); } { V T5H, T5I, T57, T58; T5H = VBYI(VADD(T5C, T5B)); T5I = VADD(T5E, T5F); ST(&(xo[WS(os, 17)]), VADD(T5H, T5I), ovs, &(xo[WS(os, 1)])); ST(&(xo[WS(os, 111)]), VSUB(T5I, T5H), ovs, &(xo[WS(os, 1)])); T57 = VSUB(T2f, T4K); T58 = VBYI(VSUB(T55, T52)); ST(&(xo[WS(os, 65)]), VSUB(T57, T58), ovs, &(xo[WS(os, 1)])); ST(&(xo[WS(os, 63)]), VADD(T57, T58), ovs, &(xo[WS(os, 1)])); } { V T5b, T5e, T5r, T5y; T5b = VBYI(VSUB(T59, T5a)); T5e = VSUB(T5c, T5d); ST(&(xo[WS(os, 33)]), VADD(T5b, T5e), ovs, &(xo[WS(os, 1)])); ST(&(xo[WS(os, 95)]), VSUB(T5e, T5b), ovs, &(xo[WS(os, 1)])); T5r = VADD(T5j, T5q); T5y = VBYI(VADD(T5u, T5x)); ST(&(xo[WS(os, 113)]), VSUB(T5r, T5y), ovs, &(xo[WS(os, 1)])); ST(&(xo[WS(os, 15)]), VADD(T5r, T5y), ovs, &(xo[WS(os, 1)])); } { V T5z, T5A, T5f, T5g; T5z = VSUB(T5j, T5q); T5A = VBYI(VSUB(T5x, T5u)); ST(&(xo[WS(os, 79)]), VSUB(T5z, T5A), ovs, &(xo[WS(os, 1)])); ST(&(xo[WS(os, 49)]), VADD(T5z, T5A), ovs, &(xo[WS(os, 1)])); T5f = VBYI(VADD(T5a, T59)); T5g = VADD(T5c, T5d); ST(&(xo[WS(os, 31)]), VADD(T5f, T5g), ovs, &(xo[WS(os, 1)])); ST(&(xo[WS(os, 97)]), VSUB(T5g, T5f), ovs, &(xo[WS(os, 1)])); } } { V T9i, T9B, T9t, T9x, T9O, Ta3, T9V, T9Z, T93, Ta0, Ta2, T9q, T9y, T9H, T9S; V T9A; { V T9a, T9r, T9h, T9s; { V T96, T99, T9d, T9g; T96 = VSUB(T94, T95); T99 = VSUB(T97, T98); T9a = VFMA(LDK(KP514102744), T96, VMUL(LDK(KP857728610), T99)); T9r = VFNMS(LDK(KP514102744), T99, VMUL(LDK(KP857728610), T96)); T9d = VSUB(T9b, T9c); T9g = VSUB(T9e, T9f); T9h = VFNMS(LDK(KP514102744), T9g, VMUL(LDK(KP857728610), T9d)); T9s = VFMA(LDK(KP857728610), T9g, VMUL(LDK(KP514102744), T9d)); } T9i = VADD(T9a, T9h); T9B = VSUB(T9s, T9r); T9t = VADD(T9r, T9s); T9x = VSUB(T9h, T9a); } { V T9K, T9T, T9N, T9U; { V T9I, T9J, T9L, T9M; T9I = VADD(T95, T94); T9J = VADD(T97, T98); T9K = VFMA(LDK(KP242980179), T9I, VMUL(LDK(KP970031253), T9J)); T9T = VFNMS(LDK(KP242980179), T9J, VMUL(LDK(KP970031253), T9I)); T9L = VADD(T9b, T9c); T9M = VADD(T9f, T9e); T9N = VFNMS(LDK(KP242980179), T9M, VMUL(LDK(KP970031253), T9L)); T9U = VFMA(LDK(KP970031253), T9M, VMUL(LDK(KP242980179), T9L)); } T9O = VADD(T9K, T9N); Ta3 = VSUB(T9U, T9T); T9V = VADD(T9T, T9U); T9Z = VSUB(T9N, T9K); } { V T8V, T9F, T9m, T9Q, T92, T9R, T9p, T9G, T8U, T9k; T8U = VSUB(T86, T85); T8V = VSUB(T8T, T8U); T9F = VADD(T8T, T8U); T9k = VSUB(T75, T72); T9m = VSUB(T9k, T9l); T9Q = VADD(T9l, T9k); { V T8Y, T91, T9n, T9o; T8Y = VFMA(LDK(KP471396736), T8W, VMUL(LDK(KP881921264), T8X)); T91 = VFNMS(LDK(KP471396736), T90, VMUL(LDK(KP881921264), T8Z)); T92 = VSUB(T8Y, T91); T9R = VADD(T91, T8Y); T9n = VFNMS(LDK(KP471396736), T8X, VMUL(LDK(KP881921264), T8W)); T9o = VFMA(LDK(KP881921264), T90, VMUL(LDK(KP471396736), T8Z)); T9p = VSUB(T9n, T9o); T9G = VADD(T9o, T9n); } T93 = VADD(T8V, T92); Ta0 = VSUB(T9R, T9Q); Ta2 = VSUB(T9F, T9G); T9q = VADD(T9m, T9p); T9y = VSUB(T9p, T9m); T9H = VADD(T9F, T9G); T9S = VADD(T9Q, T9R); T9A = VSUB(T8V, T92); } { V T9j, T9u, Ta1, Ta4; T9j = VADD(T93, T9i); T9u = VBYI(VADD(T9q, T9t)); ST(&(xo[WS(os, 117)]), VSUB(T9j, T9u), ovs, &(xo[WS(os, 1)])); ST(&(xo[WS(os, 11)]), VADD(T9j, T9u), ovs, &(xo[WS(os, 1)])); Ta1 = VBYI(VSUB(T9Z, Ta0)); Ta4 = VSUB(Ta2, Ta3); ST(&(xo[WS(os, 37)]), VADD(Ta1, Ta4), ovs, &(xo[WS(os, 1)])); ST(&(xo[WS(os, 91)]), VSUB(Ta4, Ta1), ovs, &(xo[WS(os, 1)])); } { V Ta5, Ta6, T9v, T9w; Ta5 = VBYI(VADD(Ta0, T9Z)); Ta6 = VADD(Ta2, Ta3); ST(&(xo[WS(os, 27)]), VADD(Ta5, Ta6), ovs, &(xo[WS(os, 1)])); ST(&(xo[WS(os, 101)]), VSUB(Ta6, Ta5), ovs, &(xo[WS(os, 1)])); T9v = VSUB(T93, T9i); T9w = VBYI(VSUB(T9t, T9q)); ST(&(xo[WS(os, 75)]), VSUB(T9v, T9w), ovs, &(xo[WS(os, 1)])); ST(&(xo[WS(os, 53)]), VADD(T9v, T9w), ovs, &(xo[WS(os, 1)])); } { V T9z, T9C, T9P, T9W; T9z = VBYI(VSUB(T9x, T9y)); T9C = VSUB(T9A, T9B); ST(&(xo[WS(os, 43)]), VADD(T9z, T9C), ovs, &(xo[WS(os, 1)])); ST(&(xo[WS(os, 85)]), VSUB(T9C, T9z), ovs, &(xo[WS(os, 1)])); T9P = VADD(T9H, T9O); T9W = VBYI(VADD(T9S, T9V)); ST(&(xo[WS(os, 123)]), VSUB(T9P, T9W), ovs, &(xo[WS(os, 1)])); ST(&(xo[WS(os, 5)]), VADD(T9P, T9W), ovs, &(xo[WS(os, 1)])); } { V T9X, T9Y, T9D, T9E; T9X = VSUB(T9H, T9O); T9Y = VBYI(VSUB(T9V, T9S)); ST(&(xo[WS(os, 69)]), VSUB(T9X, T9Y), ovs, &(xo[WS(os, 1)])); ST(&(xo[WS(os, 59)]), VADD(T9X, T9Y), ovs, &(xo[WS(os, 1)])); T9D = VBYI(VADD(T9y, T9x)); T9E = VADD(T9A, T9B); ST(&(xo[WS(os, 21)]), VADD(T9D, T9E), ovs, &(xo[WS(os, 1)])); ST(&(xo[WS(os, 107)]), VSUB(T9E, T9D), ovs, &(xo[WS(os, 1)])); } } { V T68, T6r, T6j, T6n, T6E, T6T, T6L, T6P, T5T, T6Q, T6S, T6g, T6o, T6x, T6I; V T6q; { V T60, T6h, T67, T6i; { V T5W, T5Z, T63, T66; T5W = VSUB(T5U, T5V); T5Z = VSUB(T5X, T5Y); T60 = VFMA(LDK(KP427555093), T5W, VMUL(LDK(KP903989293), T5Z)); T6h = VFNMS(LDK(KP427555093), T5Z, VMUL(LDK(KP903989293), T5W)); T63 = VSUB(T61, T62); T66 = VSUB(T64, T65); T67 = VFNMS(LDK(KP427555093), T66, VMUL(LDK(KP903989293), T63)); T6i = VFMA(LDK(KP903989293), T66, VMUL(LDK(KP427555093), T63)); } T68 = VADD(T60, T67); T6r = VSUB(T6i, T6h); T6j = VADD(T6h, T6i); T6n = VSUB(T67, T60); } { V T6A, T6J, T6D, T6K; { V T6y, T6z, T6B, T6C; T6y = VADD(T5V, T5U); T6z = VADD(T5X, T5Y); T6A = VFMA(LDK(KP336889853), T6y, VMUL(LDK(KP941544065), T6z)); T6J = VFNMS(LDK(KP336889853), T6z, VMUL(LDK(KP941544065), T6y)); T6B = VADD(T61, T62); T6C = VADD(T65, T64); T6D = VFNMS(LDK(KP336889853), T6C, VMUL(LDK(KP941544065), T6B)); T6K = VFMA(LDK(KP941544065), T6C, VMUL(LDK(KP336889853), T6B)); } T6E = VADD(T6A, T6D); T6T = VSUB(T6K, T6J); T6L = VADD(T6J, T6K); T6P = VSUB(T6D, T6A); } { V T5L, T6v, T6c, T6G, T5S, T6H, T6f, T6w, T5K, T6a; T5K = VSUB(T4W, T4V); T5L = VSUB(T5J, T5K); T6v = VADD(T5J, T5K); T6a = VSUB(TZ, TI); T6c = VSUB(T6a, T6b); T6G = VADD(T6b, T6a); { V T5O, T5R, T6d, T6e; T5O = VFMA(LDK(KP773010453), T5M, VMUL(LDK(KP634393284), T5N)); T5R = VFNMS(LDK(KP634393284), T5Q, VMUL(LDK(KP773010453), T5P)); T5S = VSUB(T5O, T5R); T6H = VADD(T5R, T5O); T6d = VFNMS(LDK(KP634393284), T5M, VMUL(LDK(KP773010453), T5N)); T6e = VFMA(LDK(KP634393284), T5P, VMUL(LDK(KP773010453), T5Q)); T6f = VSUB(T6d, T6e); T6w = VADD(T6e, T6d); } T5T = VADD(T5L, T5S); T6Q = VSUB(T6H, T6G); T6S = VSUB(T6v, T6w); T6g = VADD(T6c, T6f); T6o = VSUB(T6f, T6c); T6x = VADD(T6v, T6w); T6I = VADD(T6G, T6H); T6q = VSUB(T5L, T5S); } { V T69, T6k, T6R, T6U; T69 = VADD(T5T, T68); T6k = VBYI(VADD(T6g, T6j)); ST(&(xo[WS(os, 119)]), VSUB(T69, T6k), ovs, &(xo[WS(os, 1)])); ST(&(xo[WS(os, 9)]), VADD(T69, T6k), ovs, &(xo[WS(os, 1)])); T6R = VBYI(VSUB(T6P, T6Q)); T6U = VSUB(T6S, T6T); ST(&(xo[WS(os, 39)]), VADD(T6R, T6U), ovs, &(xo[WS(os, 1)])); ST(&(xo[WS(os, 89)]), VSUB(T6U, T6R), ovs, &(xo[WS(os, 1)])); } { V T6V, T6W, T6l, T6m; T6V = VBYI(VADD(T6Q, T6P)); T6W = VADD(T6S, T6T); ST(&(xo[WS(os, 25)]), VADD(T6V, T6W), ovs, &(xo[WS(os, 1)])); ST(&(xo[WS(os, 103)]), VSUB(T6W, T6V), ovs, &(xo[WS(os, 1)])); T6l = VSUB(T5T, T68); T6m = VBYI(VSUB(T6j, T6g)); ST(&(xo[WS(os, 73)]), VSUB(T6l, T6m), ovs, &(xo[WS(os, 1)])); ST(&(xo[WS(os, 55)]), VADD(T6l, T6m), ovs, &(xo[WS(os, 1)])); } { V T6p, T6s, T6F, T6M; T6p = VBYI(VSUB(T6n, T6o)); T6s = VSUB(T6q, T6r); ST(&(xo[WS(os, 41)]), VADD(T6p, T6s), ovs, &(xo[WS(os, 1)])); ST(&(xo[WS(os, 87)]), VSUB(T6s, T6p), ovs, &(xo[WS(os, 1)])); T6F = VADD(T6x, T6E); T6M = VBYI(VADD(T6I, T6L)); ST(&(xo[WS(os, 121)]), VSUB(T6F, T6M), ovs, &(xo[WS(os, 1)])); ST(&(xo[WS(os, 7)]), VADD(T6F, T6M), ovs, &(xo[WS(os, 1)])); } { V T6N, T6O, T6t, T6u; T6N = VSUB(T6x, T6E); T6O = VBYI(VSUB(T6L, T6I)); ST(&(xo[WS(os, 71)]), VSUB(T6N, T6O), ovs, &(xo[WS(os, 1)])); ST(&(xo[WS(os, 57)]), VADD(T6N, T6O), ovs, &(xo[WS(os, 1)])); T6t = VBYI(VADD(T6o, T6n)); T6u = VADD(T6q, T6r); ST(&(xo[WS(os, 23)]), VADD(T6t, T6u), ovs, &(xo[WS(os, 1)])); ST(&(xo[WS(os, 105)]), VSUB(T6u, T6t), ovs, &(xo[WS(os, 1)])); } } } } VLEAVE(); } static const kdft_desc desc = { 128, XSIMD_STRING("n1fv_128"), {938, 186, 144, 0}, &GENUS, 0, 0, 0, 0 }; void XSIMD(codelet_n1fv_128) (planner *p) { X(kdft_register) (p, n1fv_128, &desc); } #endif /* HAVE_FMA */ fftw-3.3.3/dft/simd/common/n1fv_20.c0000644000175000017500000003540512054410347013736 00000000000000/* * Copyright (c) 2003, 2007-11 Matteo Frigo * Copyright (c) 2003, 2007-11 Massachusetts Institute of Technology * * This program is free software; you can redistribute it and/or modify * it under the terms of the GNU General Public License as published by * the Free Software Foundation; either version 2 of the License, or * (at your option) any later version. * * This program is distributed in the hope that it will be useful, * but WITHOUT ANY WARRANTY; without even the implied warranty of * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the * GNU General Public License for more details. * * You should have received a copy of the GNU General Public License * along with this program; if not, write to the Free Software * Foundation, Inc., 51 Franklin Street, Fifth Floor, Boston, MA 02110-1301 USA * */ /* This file was automatically generated --- DO NOT EDIT */ /* Generated on Sun Nov 25 07:36:54 EST 2012 */ #include "codelet-dft.h" #ifdef HAVE_FMA /* Generated by: ../../../genfft/gen_notw_c.native -fma -reorder-insns -schedule-for-pipeline -simd -compact -variables 4 -pipeline-latency 8 -n 20 -name n1fv_20 -include n1f.h */ /* * This function contains 104 FP additions, 50 FP multiplications, * (or, 58 additions, 4 multiplications, 46 fused multiply/add), * 71 stack variables, 4 constants, and 40 memory accesses */ #include "n1f.h" static void n1fv_20(const R *ri, const R *ii, R *ro, R *io, stride is, stride os, INT v, INT ivs, INT ovs) { DVK(KP559016994, +0.559016994374947424102293417182819058860154590); DVK(KP618033988, +0.618033988749894848204586834365638117720309180); DVK(KP951056516, +0.951056516295153572116439333379382143405698634); DVK(KP250000000, +0.250000000000000000000000000000000000000000000); { INT i; const R *xi; R *xo; xi = ri; xo = ro; for (i = v; i > 0; i = i - VL, xi = xi + (VL * ivs), xo = xo + (VL * ovs), MAKE_VOLATILE_STRIDE(40, is), MAKE_VOLATILE_STRIDE(40, os)) { V TU, TI, TP, TX, TM, TW, TT, TF; { V T3, Tm, T1r, T13, Ta, TN, TH, TA, TG, Tt, Th, TO, T1u, T1C, T1n; V T1a, T1m, T1h, T1x, T1D, TE, Ti; { V T1, T2, Tk, Tl; T1 = LD(&(xi[0]), ivs, &(xi[0])); T2 = LD(&(xi[WS(is, 10)]), ivs, &(xi[0])); Tk = LD(&(xi[WS(is, 5)]), ivs, &(xi[WS(is, 1)])); Tl = LD(&(xi[WS(is, 15)]), ivs, &(xi[WS(is, 1)])); { V T14, T6, T1c, Tw, Tn, T1f, Tz, T17, T9, To, Tq, T1b, Td, Tr, Te; V Tf, T15, Tp; { V Tx, Ty, T7, T8, Tb, Tc; { V T4, T5, Tu, Tv, T11, T12; T4 = LD(&(xi[WS(is, 4)]), ivs, &(xi[0])); T5 = LD(&(xi[WS(is, 14)]), ivs, &(xi[0])); Tu = LD(&(xi[WS(is, 13)]), ivs, &(xi[WS(is, 1)])); Tv = LD(&(xi[WS(is, 3)]), ivs, &(xi[WS(is, 1)])); Tx = LD(&(xi[WS(is, 17)]), ivs, &(xi[WS(is, 1)])); T3 = VSUB(T1, T2); T11 = VADD(T1, T2); Tm = VSUB(Tk, Tl); T12 = VADD(Tk, Tl); T14 = VADD(T4, T5); T6 = VSUB(T4, T5); T1c = VADD(Tu, Tv); Tw = VSUB(Tu, Tv); Ty = LD(&(xi[WS(is, 7)]), ivs, &(xi[WS(is, 1)])); T7 = LD(&(xi[WS(is, 16)]), ivs, &(xi[0])); T8 = LD(&(xi[WS(is, 6)]), ivs, &(xi[0])); T1r = VADD(T11, T12); T13 = VSUB(T11, T12); } Tb = LD(&(xi[WS(is, 8)]), ivs, &(xi[0])); Tc = LD(&(xi[WS(is, 18)]), ivs, &(xi[0])); Tn = LD(&(xi[WS(is, 9)]), ivs, &(xi[WS(is, 1)])); T1f = VADD(Tx, Ty); Tz = VSUB(Tx, Ty); T17 = VADD(T7, T8); T9 = VSUB(T7, T8); To = LD(&(xi[WS(is, 19)]), ivs, &(xi[WS(is, 1)])); Tq = LD(&(xi[WS(is, 1)]), ivs, &(xi[WS(is, 1)])); T1b = VADD(Tb, Tc); Td = VSUB(Tb, Tc); Tr = LD(&(xi[WS(is, 11)]), ivs, &(xi[WS(is, 1)])); Te = LD(&(xi[WS(is, 12)]), ivs, &(xi[0])); Tf = LD(&(xi[WS(is, 2)]), ivs, &(xi[0])); } Ta = VADD(T6, T9); TN = VSUB(T6, T9); T15 = VADD(Tn, To); Tp = VSUB(Tn, To); TH = VSUB(Tz, Tw); TA = VADD(Tw, Tz); { V T1d, T1v, T18, Ts, T1e, Tg, T16, T1s; T1d = VSUB(T1b, T1c); T1v = VADD(T1b, T1c); T18 = VADD(Tq, Tr); Ts = VSUB(Tq, Tr); T1e = VADD(Te, Tf); Tg = VSUB(Te, Tf); T16 = VSUB(T14, T15); T1s = VADD(T14, T15); { V T1t, T19, T1w, T1g; T1t = VADD(T17, T18); T19 = VSUB(T17, T18); TG = VSUB(Ts, Tp); Tt = VADD(Tp, Ts); T1w = VADD(T1e, T1f); T1g = VSUB(T1e, T1f); Th = VADD(Td, Tg); TO = VSUB(Td, Tg); T1u = VADD(T1s, T1t); T1C = VSUB(T1s, T1t); T1n = VSUB(T16, T19); T1a = VADD(T16, T19); T1m = VSUB(T1d, T1g); T1h = VADD(T1d, T1g); T1x = VADD(T1v, T1w); T1D = VSUB(T1v, T1w); } } } } TE = VSUB(Ta, Th); Ti = VADD(Ta, Th); { V TL, T1k, T1A, Tj, TD, T1E, T1G, TK, TC, T1j, T1z, T1i, T1y, TB; TL = VSUB(TA, Tt); TB = VADD(Tt, TA); T1i = VADD(T1a, T1h); T1k = VSUB(T1a, T1h); T1y = VADD(T1u, T1x); T1A = VSUB(T1u, T1x); Tj = VADD(T3, Ti); TD = VFNMS(LDK(KP250000000), Ti, T3); T1E = VMUL(LDK(KP951056516), VFMA(LDK(KP618033988), T1D, T1C)); T1G = VMUL(LDK(KP951056516), VFNMS(LDK(KP618033988), T1C, T1D)); TK = VFNMS(LDK(KP250000000), TB, Tm); TC = VADD(Tm, TB); T1j = VFNMS(LDK(KP250000000), T1i, T13); ST(&(xo[0]), VADD(T1r, T1y), ovs, &(xo[0])); T1z = VFNMS(LDK(KP250000000), T1y, T1r); ST(&(xo[WS(os, 10)]), VADD(T13, T1i), ovs, &(xo[0])); { V T1p, T1l, T1o, T1q, T1F, T1B; TU = VFNMS(LDK(KP618033988), TG, TH); TI = VFMA(LDK(KP618033988), TH, TG); TP = VFMA(LDK(KP618033988), TO, TN); TX = VFNMS(LDK(KP618033988), TN, TO); ST(&(xo[WS(os, 15)]), VFMAI(TC, Tj), ovs, &(xo[WS(os, 1)])); ST(&(xo[WS(os, 5)]), VFNMSI(TC, Tj), ovs, &(xo[WS(os, 1)])); T1p = VFMA(LDK(KP559016994), T1k, T1j); T1l = VFNMS(LDK(KP559016994), T1k, T1j); T1o = VMUL(LDK(KP951056516), VFNMS(LDK(KP618033988), T1n, T1m)); T1q = VMUL(LDK(KP951056516), VFMA(LDK(KP618033988), T1m, T1n)); T1F = VFNMS(LDK(KP559016994), T1A, T1z); T1B = VFMA(LDK(KP559016994), T1A, T1z); ST(&(xo[WS(os, 14)]), VFMAI(T1q, T1p), ovs, &(xo[0])); ST(&(xo[WS(os, 6)]), VFNMSI(T1q, T1p), ovs, &(xo[0])); ST(&(xo[WS(os, 18)]), VFNMSI(T1o, T1l), ovs, &(xo[0])); ST(&(xo[WS(os, 2)]), VFMAI(T1o, T1l), ovs, &(xo[0])); ST(&(xo[WS(os, 16)]), VFNMSI(T1E, T1B), ovs, &(xo[0])); ST(&(xo[WS(os, 4)]), VFMAI(T1E, T1B), ovs, &(xo[0])); ST(&(xo[WS(os, 12)]), VFMAI(T1G, T1F), ovs, &(xo[0])); ST(&(xo[WS(os, 8)]), VFNMSI(T1G, T1F), ovs, &(xo[0])); TM = VFNMS(LDK(KP559016994), TL, TK); TW = VFMA(LDK(KP559016994), TL, TK); TT = VFNMS(LDK(KP559016994), TE, TD); TF = VFMA(LDK(KP559016994), TE, TD); } } } { V T10, TY, TQ, TS, TJ, TR, TZ, TV; T10 = VFMA(LDK(KP951056516), TX, TW); TY = VFNMS(LDK(KP951056516), TX, TW); TQ = VFMA(LDK(KP951056516), TP, TM); TS = VFNMS(LDK(KP951056516), TP, TM); TJ = VFMA(LDK(KP951056516), TI, TF); TR = VFNMS(LDK(KP951056516), TI, TF); TZ = VFMA(LDK(KP951056516), TU, TT); TV = VFNMS(LDK(KP951056516), TU, TT); ST(&(xo[WS(os, 11)]), VFMAI(TS, TR), ovs, &(xo[WS(os, 1)])); ST(&(xo[WS(os, 9)]), VFNMSI(TS, TR), ovs, &(xo[WS(os, 1)])); ST(&(xo[WS(os, 19)]), VFMAI(TQ, TJ), ovs, &(xo[WS(os, 1)])); ST(&(xo[WS(os, 1)]), VFNMSI(TQ, TJ), ovs, &(xo[WS(os, 1)])); ST(&(xo[WS(os, 3)]), VFMAI(TY, TV), ovs, &(xo[WS(os, 1)])); ST(&(xo[WS(os, 17)]), VFNMSI(TY, TV), ovs, &(xo[WS(os, 1)])); ST(&(xo[WS(os, 7)]), VFMAI(T10, TZ), ovs, &(xo[WS(os, 1)])); ST(&(xo[WS(os, 13)]), VFNMSI(T10, TZ), ovs, &(xo[WS(os, 1)])); } } } VLEAVE(); } static const kdft_desc desc = { 20, XSIMD_STRING("n1fv_20"), {58, 4, 46, 0}, &GENUS, 0, 0, 0, 0 }; void XSIMD(codelet_n1fv_20) (planner *p) { X(kdft_register) (p, n1fv_20, &desc); } #else /* HAVE_FMA */ /* Generated by: ../../../genfft/gen_notw_c.native -simd -compact -variables 4 -pipeline-latency 8 -n 20 -name n1fv_20 -include n1f.h */ /* * This function contains 104 FP additions, 24 FP multiplications, * (or, 92 additions, 12 multiplications, 12 fused multiply/add), * 53 stack variables, 4 constants, and 40 memory accesses */ #include "n1f.h" static void n1fv_20(const R *ri, const R *ii, R *ro, R *io, stride is, stride os, INT v, INT ivs, INT ovs) { DVK(KP587785252, +0.587785252292473129168705954639072768597652438); DVK(KP951056516, +0.951056516295153572116439333379382143405698634); DVK(KP250000000, +0.250000000000000000000000000000000000000000000); DVK(KP559016994, +0.559016994374947424102293417182819058860154590); { INT i; const R *xi; R *xo; xi = ri; xo = ro; for (i = v; i > 0; i = i - VL, xi = xi + (VL * ivs), xo = xo + (VL * ovs), MAKE_VOLATILE_STRIDE(40, is), MAKE_VOLATILE_STRIDE(40, os)) { V T3, T1B, Tm, T1i, TG, TN, TO, TH, T13, T16, T1k, T1u, T1v, T1z, T1r; V T1s, T1y, T1a, T1d, T1j, Ti, TD, TB, TL, Tj, TC; { V T1, T2, T1g, Tk, Tl, T1h; T1 = LD(&(xi[0]), ivs, &(xi[0])); T2 = LD(&(xi[WS(is, 10)]), ivs, &(xi[0])); T1g = VADD(T1, T2); Tk = LD(&(xi[WS(is, 5)]), ivs, &(xi[WS(is, 1)])); Tl = LD(&(xi[WS(is, 15)]), ivs, &(xi[WS(is, 1)])); T1h = VADD(Tk, Tl); T3 = VSUB(T1, T2); T1B = VADD(T1g, T1h); Tm = VSUB(Tk, Tl); T1i = VSUB(T1g, T1h); } { V T6, T18, Tw, T12, Tz, T15, T9, T1b, Td, T11, Tp, T19, Ts, T1c, Tg; V T14; { V T4, T5, Tu, Tv; T4 = LD(&(xi[WS(is, 4)]), ivs, &(xi[0])); T5 = LD(&(xi[WS(is, 14)]), ivs, &(xi[0])); T6 = VSUB(T4, T5); T18 = VADD(T4, T5); Tu = LD(&(xi[WS(is, 13)]), ivs, &(xi[WS(is, 1)])); Tv = LD(&(xi[WS(is, 3)]), ivs, &(xi[WS(is, 1)])); Tw = VSUB(Tu, Tv); T12 = VADD(Tu, Tv); } { V Tx, Ty, T7, T8; Tx = LD(&(xi[WS(is, 17)]), ivs, &(xi[WS(is, 1)])); Ty = LD(&(xi[WS(is, 7)]), ivs, &(xi[WS(is, 1)])); Tz = VSUB(Tx, Ty); T15 = VADD(Tx, Ty); T7 = LD(&(xi[WS(is, 16)]), ivs, &(xi[0])); T8 = LD(&(xi[WS(is, 6)]), ivs, &(xi[0])); T9 = VSUB(T7, T8); T1b = VADD(T7, T8); } { V Tb, Tc, Tn, To; Tb = LD(&(xi[WS(is, 8)]), ivs, &(xi[0])); Tc = LD(&(xi[WS(is, 18)]), ivs, &(xi[0])); Td = VSUB(Tb, Tc); T11 = VADD(Tb, Tc); Tn = LD(&(xi[WS(is, 9)]), ivs, &(xi[WS(is, 1)])); To = LD(&(xi[WS(is, 19)]), ivs, &(xi[WS(is, 1)])); Tp = VSUB(Tn, To); T19 = VADD(Tn, To); } { V Tq, Tr, Te, Tf; Tq = LD(&(xi[WS(is, 1)]), ivs, &(xi[WS(is, 1)])); Tr = LD(&(xi[WS(is, 11)]), ivs, &(xi[WS(is, 1)])); Ts = VSUB(Tq, Tr); T1c = VADD(Tq, Tr); Te = LD(&(xi[WS(is, 12)]), ivs, &(xi[0])); Tf = LD(&(xi[WS(is, 2)]), ivs, &(xi[0])); Tg = VSUB(Te, Tf); T14 = VADD(Te, Tf); } TG = VSUB(Ts, Tp); TN = VSUB(T6, T9); TO = VSUB(Td, Tg); TH = VSUB(Tz, Tw); T13 = VSUB(T11, T12); T16 = VSUB(T14, T15); T1k = VADD(T13, T16); T1u = VADD(T11, T12); T1v = VADD(T14, T15); T1z = VADD(T1u, T1v); T1r = VADD(T18, T19); T1s = VADD(T1b, T1c); T1y = VADD(T1r, T1s); T1a = VSUB(T18, T19); T1d = VSUB(T1b, T1c); T1j = VADD(T1a, T1d); { V Ta, Th, Tt, TA; Ta = VADD(T6, T9); Th = VADD(Td, Tg); Ti = VADD(Ta, Th); TD = VMUL(LDK(KP559016994), VSUB(Ta, Th)); Tt = VADD(Tp, Ts); TA = VADD(Tw, Tz); TB = VADD(Tt, TA); TL = VMUL(LDK(KP559016994), VSUB(TA, Tt)); } } Tj = VADD(T3, Ti); TC = VBYI(VADD(Tm, TB)); ST(&(xo[WS(os, 5)]), VSUB(Tj, TC), ovs, &(xo[WS(os, 1)])); ST(&(xo[WS(os, 15)]), VADD(Tj, TC), ovs, &(xo[WS(os, 1)])); { V T1A, T1C, T1D, T1x, T1G, T1t, T1w, T1F, T1E; T1A = VMUL(LDK(KP559016994), VSUB(T1y, T1z)); T1C = VADD(T1y, T1z); T1D = VFNMS(LDK(KP250000000), T1C, T1B); T1t = VSUB(T1r, T1s); T1w = VSUB(T1u, T1v); T1x = VBYI(VFMA(LDK(KP951056516), T1t, VMUL(LDK(KP587785252), T1w))); T1G = VBYI(VFNMS(LDK(KP587785252), T1t, VMUL(LDK(KP951056516), T1w))); ST(&(xo[0]), VADD(T1B, T1C), ovs, &(xo[0])); T1F = VSUB(T1D, T1A); ST(&(xo[WS(os, 8)]), VSUB(T1F, T1G), ovs, &(xo[0])); ST(&(xo[WS(os, 12)]), VADD(T1G, T1F), ovs, &(xo[0])); T1E = VADD(T1A, T1D); ST(&(xo[WS(os, 4)]), VADD(T1x, T1E), ovs, &(xo[0])); ST(&(xo[WS(os, 16)]), VSUB(T1E, T1x), ovs, &(xo[0])); } { V T1n, T1l, T1m, T1f, T1q, T17, T1e, T1p, T1o; T1n = VMUL(LDK(KP559016994), VSUB(T1j, T1k)); T1l = VADD(T1j, T1k); T1m = VFNMS(LDK(KP250000000), T1l, T1i); T17 = VSUB(T13, T16); T1e = VSUB(T1a, T1d); T1f = VBYI(VFNMS(LDK(KP587785252), T1e, VMUL(LDK(KP951056516), T17))); T1q = VBYI(VFMA(LDK(KP951056516), T1e, VMUL(LDK(KP587785252), T17))); ST(&(xo[WS(os, 10)]), VADD(T1i, T1l), ovs, &(xo[0])); T1p = VADD(T1n, T1m); ST(&(xo[WS(os, 6)]), VSUB(T1p, T1q), ovs, &(xo[0])); ST(&(xo[WS(os, 14)]), VADD(T1q, T1p), ovs, &(xo[0])); T1o = VSUB(T1m, T1n); ST(&(xo[WS(os, 2)]), VADD(T1f, T1o), ovs, &(xo[0])); ST(&(xo[WS(os, 18)]), VSUB(T1o, T1f), ovs, &(xo[0])); } { V TI, TP, TX, TU, TM, TW, TF, TT, TK, TE; TI = VFMA(LDK(KP951056516), TG, VMUL(LDK(KP587785252), TH)); TP = VFMA(LDK(KP951056516), TN, VMUL(LDK(KP587785252), TO)); TX = VFNMS(LDK(KP587785252), TN, VMUL(LDK(KP951056516), TO)); TU = VFNMS(LDK(KP587785252), TG, VMUL(LDK(KP951056516), TH)); TK = VFMS(LDK(KP250000000), TB, Tm); TM = VADD(TK, TL); TW = VSUB(TL, TK); TE = VFNMS(LDK(KP250000000), Ti, T3); TF = VADD(TD, TE); TT = VSUB(TE, TD); { V TJ, TQ, TZ, T10; TJ = VADD(TF, TI); TQ = VBYI(VSUB(TM, TP)); ST(&(xo[WS(os, 19)]), VSUB(TJ, TQ), ovs, &(xo[WS(os, 1)])); ST(&(xo[WS(os, 1)]), VADD(TJ, TQ), ovs, &(xo[WS(os, 1)])); TZ = VADD(TT, TU); T10 = VBYI(VADD(TX, TW)); ST(&(xo[WS(os, 13)]), VSUB(TZ, T10), ovs, &(xo[WS(os, 1)])); ST(&(xo[WS(os, 7)]), VADD(TZ, T10), ovs, &(xo[WS(os, 1)])); } { V TR, TS, TV, TY; TR = VSUB(TF, TI); TS = VBYI(VADD(TP, TM)); ST(&(xo[WS(os, 11)]), VSUB(TR, TS), ovs, &(xo[WS(os, 1)])); ST(&(xo[WS(os, 9)]), VADD(TR, TS), ovs, &(xo[WS(os, 1)])); TV = VSUB(TT, TU); TY = VBYI(VSUB(TW, TX)); ST(&(xo[WS(os, 17)]), VSUB(TV, TY), ovs, &(xo[WS(os, 1)])); ST(&(xo[WS(os, 3)]), VADD(TV, TY), ovs, &(xo[WS(os, 1)])); } } } } VLEAVE(); } static const kdft_desc desc = { 20, XSIMD_STRING("n1fv_20"), {92, 12, 12, 0}, &GENUS, 0, 0, 0, 0 }; void XSIMD(codelet_n1fv_20) (planner *p) { X(kdft_register) (p, n1fv_20, &desc); } #endif /* HAVE_FMA */ fftw-3.3.3/dft/simd/common/n1fv_25.c0000644000175000017500000010726012054410356013742 00000000000000/* * Copyright (c) 2003, 2007-11 Matteo Frigo * Copyright (c) 2003, 2007-11 Massachusetts Institute of Technology * * This program is free software; you can redistribute it and/or modify * it under the terms of the GNU General Public License as published by * the Free Software Foundation; either version 2 of the License, or * (at your option) any later version. * * This program is distributed in the hope that it will be useful, * but WITHOUT ANY WARRANTY; without even the implied warranty of * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the * GNU General Public License for more details. * * You should have received a copy of the GNU General Public License * along with this program; if not, write to the Free Software * Foundation, Inc., 51 Franklin Street, Fifth Floor, Boston, MA 02110-1301 USA * */ /* This file was automatically generated --- DO NOT EDIT */ /* Generated on Sun Nov 25 07:36:55 EST 2012 */ #include "codelet-dft.h" #ifdef HAVE_FMA /* Generated by: ../../../genfft/gen_notw_c.native -fma -reorder-insns -schedule-for-pipeline -simd -compact -variables 4 -pipeline-latency 8 -n 25 -name n1fv_25 -include n1f.h */ /* * This function contains 224 FP additions, 193 FP multiplications, * (or, 43 additions, 12 multiplications, 181 fused multiply/add), * 215 stack variables, 67 constants, and 50 memory accesses */ #include "n1f.h" static void n1fv_25(const R *ri, const R *ii, R *ro, R *io, stride is, stride os, INT v, INT ivs, INT ovs) { DVK(KP792626838, +0.792626838241819413632131824093538848057784557); DVK(KP876091699, +0.876091699473550838204498029706869638173524346); DVK(KP803003575, +0.803003575438660414833440593570376004635464850); DVK(KP617882369, +0.617882369114440893914546919006756321695042882); DVK(KP242145790, +0.242145790282157779872542093866183953459003101); DVK(KP968583161, +0.968583161128631119490168375464735813836012403); DVK(KP999544308, +0.999544308746292983948881682379742149196758193); DVK(KP683113946, +0.683113946453479238701949862233725244439656928); DVK(KP559154169, +0.559154169276087864842202529084232643714075927); DVK(KP904730450, +0.904730450839922351881287709692877908104763647); DVK(KP829049696, +0.829049696159252993975487806364305442437946767); DVK(KP831864738, +0.831864738706457140726048799369896829771167132); DVK(KP916574801, +0.916574801383451584742370439148878693530976769); DVK(KP894834959, +0.894834959464455102997960030820114611498661386); DVK(KP809385824, +0.809385824416008241660603814668679683846476688); DVK(KP447417479, +0.447417479732227551498980015410057305749330693); DVK(KP860541664, +0.860541664367944677098261680920518816412804187); DVK(KP897376177, +0.897376177523557693138608077137219684419427330); DVK(KP876306680, +0.876306680043863587308115903922062583399064238); DVK(KP681693190, +0.681693190061530575150324149145440022633095390); DVK(KP560319534, +0.560319534973832390111614715371676131169633784); DVK(KP855719849, +0.855719849902058969314654733608091555096772472); DVK(KP237294955, +0.237294955877110315393888866460840817927895961); DVK(KP949179823, +0.949179823508441261575555465843363271711583843); DVK(KP904508497, +0.904508497187473712051146708591409529430077295); DVK(KP997675361, +0.997675361079556513670859573984492383596555031); DVK(KP262346850, +0.262346850930607871785420028382979691334784273); DVK(KP763932022, +0.763932022500210303590826331268723764559381640); DVK(KP992114701, +0.992114701314477831049793042785778521453036709); DVK(KP690983005, +0.690983005625052575897706582817180941139845410); DVK(KP952936919, +0.952936919628306576880750665357914584765951388); DVK(KP998026728, +0.998026728428271561952336806863450553336905220); DVK(KP570584518, +0.570584518783621657366766175430996792655723863); DVK(KP669429328, +0.669429328479476605641803240971985825917022098); DVK(KP923225144, +0.923225144846402650453449441572664695995209956); DVK(KP906616052, +0.906616052148196230441134447086066874408359177); DVK(KP956723877, +0.956723877038460305821989399535483155872969262); DVK(KP522616830, +0.522616830205754336872861364785224694908468440); DVK(KP945422727, +0.945422727388575946270360266328811958657216298); DVK(KP912575812, +0.912575812670962425556968549836277086778922727); DVK(KP982009705, +0.982009705009746369461829878184175962711969869); DVK(KP921078979, +0.921078979742360627699756128143719920817673854); DVK(KP734762448, +0.734762448793050413546343770063151342619912334); DVK(KP951056516, +0.951056516295153572116439333379382143405698634); DVK(KP958953096, +0.958953096729998668045963838399037225970891871); DVK(KP867381224, +0.867381224396525206773171885031575671309956167); DVK(KP269969613, +0.269969613759572083574752974412347470060951301); DVK(KP244189809, +0.244189809627953270309879511234821255780225091); DVK(KP845997307, +0.845997307939530944175097360758058292389769300); DVK(KP772036680, +0.772036680810363904029489473607579825330539880); DVK(KP132830569, +0.132830569247582714407653942074819768844536507); DVK(KP120146378, +0.120146378570687701782758537356596213647956445); DVK(KP987388751, +0.987388751065621252324603216482382109400433949); DVK(KP893101515, +0.893101515366181661711202267938416198338079437); DVK(KP786782374, +0.786782374965295178365099601674911834788448471); DVK(KP869845200, +0.869845200362138853122720822420327157933056305); DVK(KP447533225, +0.447533225982656890041886979663652563063114397); DVK(KP494780565, +0.494780565770515410344588413655324772219443730); DVK(KP578046249, +0.578046249379945007321754579646815604023525655); DVK(KP522847744, +0.522847744331509716623755382187077770911012542); DVK(KP059835404, +0.059835404262124915169548397419498386427871950); DVK(KP066152395, +0.066152395967733048213034281011006031460903353); DVK(KP603558818, +0.603558818296015001454675132653458027918768137); DVK(KP667278218, +0.667278218140296670899089292254759909713898805); DVK(KP559016994, +0.559016994374947424102293417182819058860154590); DVK(KP250000000, +0.250000000000000000000000000000000000000000000); DVK(KP618033988, +0.618033988749894848204586834365638117720309180); { INT i; const R *xi; R *xo; xi = ri; xo = ro; for (i = v; i > 0; i = i - VL, xi = xi + (VL * ivs), xo = xo + (VL * ovs), MAKE_VOLATILE_STRIDE(50, is), MAKE_VOLATILE_STRIDE(50, os)) { V T1g, T1k, T1I, T24, T2a, T1G, T1A, T1l, T1B, T1H, T1d; { V T2z, T1q, Ta, T9, T3n, Ty, Tl, T2O, T2W, T2l, T2s, TV, T1i, T1K, T1S; V T3z, T3t, Tk, T3o, Tp, T2g, T2N, T2V, T2o, T2t, T1a, T1j, T1J, T1R, Tz; V Tt, TA, Tw; { V T1, T5, T6, T2, T3; T1 = LD(&(xi[0]), ivs, &(xi[0])); T5 = LD(&(xi[WS(is, 10)]), ivs, &(xi[0])); T6 = LD(&(xi[WS(is, 15)]), ivs, &(xi[WS(is, 1)])); T2 = LD(&(xi[WS(is, 5)]), ivs, &(xi[WS(is, 1)])); T3 = LD(&(xi[WS(is, 20)]), ivs, &(xi[0])); { V TH, TW, TK, TS, T10, T8, TN, TT, T17, TZ, T11; TH = LD(&(xi[WS(is, 2)]), ivs, &(xi[0])); TW = LD(&(xi[WS(is, 3)]), ivs, &(xi[WS(is, 1)])); { V TI, TJ, TL, T7, T1p, T4, T1o, TM, TX, TY; TI = LD(&(xi[WS(is, 7)]), ivs, &(xi[WS(is, 1)])); TJ = LD(&(xi[WS(is, 22)]), ivs, &(xi[0])); TL = LD(&(xi[WS(is, 12)]), ivs, &(xi[0])); T7 = VADD(T5, T6); T1p = VSUB(T5, T6); T4 = VADD(T2, T3); T1o = VSUB(T2, T3); TM = LD(&(xi[WS(is, 17)]), ivs, &(xi[WS(is, 1)])); TX = LD(&(xi[WS(is, 8)]), ivs, &(xi[0])); TK = VADD(TI, TJ); TS = VSUB(TI, TJ); TY = LD(&(xi[WS(is, 23)]), ivs, &(xi[WS(is, 1)])); T10 = LD(&(xi[WS(is, 13)]), ivs, &(xi[WS(is, 1)])); T2z = VFNMS(LDK(KP618033988), T1o, T1p); T1q = VFMA(LDK(KP618033988), T1p, T1o); Ta = VSUB(T4, T7); T8 = VADD(T4, T7); TN = VADD(TL, TM); TT = VSUB(TM, TL); T17 = VSUB(TX, TY); TZ = VADD(TX, TY); T11 = LD(&(xi[WS(is, 18)]), ivs, &(xi[0])); } { V Tc, T2m, T19, Tn, To, Tr, Tj, T16, T2n, Ts, Tu, Tv; { V TU, T2j, TO, TQ, T12, T18; Tc = LD(&(xi[WS(is, 1)]), ivs, &(xi[WS(is, 1)])); T9 = VFNMS(LDK(KP250000000), T8, T1); T3n = VADD(T1, T8); TU = VFNMS(LDK(KP618033988), TT, TS); T2j = VFMA(LDK(KP618033988), TS, TT); TO = VADD(TK, TN); TQ = VSUB(TN, TK); T12 = VADD(T10, T11); T18 = VSUB(T10, T11); Ty = LD(&(xi[WS(is, 4)]), ivs, &(xi[0])); { V T3r, T15, T13, Tf, Ti, T2k, TR, TP, T3s, T14; { V Td, Te, Tg, Th; Td = LD(&(xi[WS(is, 6)]), ivs, &(xi[0])); Te = LD(&(xi[WS(is, 21)]), ivs, &(xi[WS(is, 1)])); Tg = LD(&(xi[WS(is, 11)]), ivs, &(xi[WS(is, 1)])); Th = LD(&(xi[WS(is, 16)]), ivs, &(xi[0])); TP = VFNMS(LDK(KP250000000), TO, TH); T3r = VADD(TH, TO); T2m = VFNMS(LDK(KP618033988), T17, T18); T19 = VFMA(LDK(KP618033988), T18, T17); T15 = VSUB(T12, TZ); T13 = VADD(TZ, T12); Tf = VADD(Td, Te); Tn = VSUB(Td, Te); To = VSUB(Th, Tg); Ti = VADD(Tg, Th); } T2k = VFMA(LDK(KP559016994), TQ, TP); TR = VFNMS(LDK(KP559016994), TQ, TP); Tr = LD(&(xi[WS(is, 24)]), ivs, &(xi[0])); T3s = VADD(TW, T13); T14 = VFNMS(LDK(KP250000000), T13, TW); Tj = VADD(Tf, Ti); Tl = VSUB(Tf, Ti); T2O = VFNMS(LDK(KP667278218), T2k, T2j); T2W = VFMA(LDK(KP603558818), T2j, T2k); T2l = VFMA(LDK(KP066152395), T2k, T2j); T2s = VFNMS(LDK(KP059835404), T2j, T2k); TV = VFNMS(LDK(KP522847744), TU, TR); T1i = VFMA(LDK(KP578046249), TR, TU); T1K = VFNMS(LDK(KP494780565), TR, TU); T1S = VFMA(LDK(KP447533225), TU, TR); T16 = VFNMS(LDK(KP559016994), T15, T14); T2n = VFMA(LDK(KP559016994), T15, T14); T3z = VSUB(T3r, T3s); T3t = VADD(T3r, T3s); Ts = LD(&(xi[WS(is, 9)]), ivs, &(xi[WS(is, 1)])); Tu = LD(&(xi[WS(is, 19)]), ivs, &(xi[WS(is, 1)])); Tv = LD(&(xi[WS(is, 14)]), ivs, &(xi[0])); } } Tk = VFNMS(LDK(KP250000000), Tj, Tc); T3o = VADD(Tc, Tj); Tp = VFNMS(LDK(KP618033988), To, Tn); T2g = VFMA(LDK(KP618033988), Tn, To); T2N = VFMA(LDK(KP066152395), T2n, T2m); T2V = VFNMS(LDK(KP059835404), T2m, T2n); T2o = VFMA(LDK(KP869845200), T2n, T2m); T2t = VFNMS(LDK(KP786782374), T2m, T2n); T1a = VFNMS(LDK(KP893101515), T19, T16); T1j = VFMA(LDK(KP987388751), T16, T19); T1J = VFNMS(LDK(KP120146378), T19, T16); T1R = VFMA(LDK(KP132830569), T16, T19); Tz = VADD(Ts, Tr); Tt = VSUB(Tr, Ts); TA = VADD(Tv, Tu); Tw = VSUB(Tu, Tv); } } } { V T2p, T2I, T2u, T2C, Tx, T2d, T2X, T34, T2P, T3b, T2b, Tb, T2Q, T2Z, T2h; V T2w, Tq, T1e, T1M, T1U, TE, T2c, T3q, T3y; T2p = VFNMS(LDK(KP772036680), T2o, T2l); T2I = VFMA(LDK(KP772036680), T2o, T2l); T2u = VFMA(LDK(KP772036680), T2t, T2s); T2C = VFNMS(LDK(KP772036680), T2t, T2s); { V TD, TB, Tm, T2f, T3p, TC; Tx = VFMA(LDK(KP618033988), Tw, Tt); T2d = VFNMS(LDK(KP618033988), Tt, Tw); TD = VSUB(Tz, TA); TB = VADD(Tz, TA); Tm = VFMA(LDK(KP559016994), Tl, Tk); T2f = VFNMS(LDK(KP559016994), Tl, Tk); T2X = VFMA(LDK(KP845997307), T2W, T2V); T34 = VFNMS(LDK(KP845997307), T2W, T2V); T2P = VFNMS(LDK(KP845997307), T2O, T2N); T3b = VFMA(LDK(KP845997307), T2O, T2N); T2b = VFNMS(LDK(KP559016994), Ta, T9); Tb = VFMA(LDK(KP559016994), Ta, T9); T3p = VADD(Ty, TB); TC = VFMS(LDK(KP250000000), TB, Ty); T2Q = VFNMS(LDK(KP522847744), T2g, T2f); T2Z = VFMA(LDK(KP578046249), T2f, T2g); T2h = VFMA(LDK(KP893101515), T2g, T2f); T2w = VFNMS(LDK(KP987388751), T2f, T2g); Tq = VFNMS(LDK(KP244189809), Tp, Tm); T1e = VFMA(LDK(KP269969613), Tm, Tp); T1M = VFMA(LDK(KP667278218), Tm, Tp); T1U = VFNMS(LDK(KP603558818), Tp, Tm); TE = VFNMS(LDK(KP559016994), TD, TC); T2c = VFMA(LDK(KP559016994), TD, TC); T3q = VADD(T3o, T3p); T3y = VSUB(T3o, T3p); } { V T1Z, T25, T1P, T22, T1X, TG, T1b, T28, T1t, T1y, T1x, T1E, T1Q, T1Y; { V T26, T1L, T1T, TF, T1f, T1W, T3m, T3g, T2M, T2G, T39, T3j, T21, T1O, T20; V T27; T26 = VFMA(LDK(KP867381224), T1K, T1J); T1L = VFNMS(LDK(KP867381224), T1K, T1J); T20 = VFNMS(LDK(KP958953096), T1S, T1R); T1T = VFMA(LDK(KP958953096), T1S, T1R); { V T2R, T2Y, T2e, T2v, T1N, T1V; T2R = VFNMS(LDK(KP494780565), T2c, T2d); T2Y = VFMA(LDK(KP447533225), T2d, T2c); T2e = VFMA(LDK(KP120146378), T2d, T2c); T2v = VFNMS(LDK(KP132830569), T2c, T2d); TF = VFNMS(LDK(KP667278218), TE, Tx); T1f = VFMA(LDK(KP603558818), Tx, TE); T1N = VFMA(LDK(KP869845200), TE, Tx); T1V = VFNMS(LDK(KP786782374), Tx, TE); { V T3A, T3C, T3w, T3u; T3A = VMUL(LDK(KP951056516), VFMA(LDK(KP618033988), T3z, T3y)); T3C = VMUL(LDK(KP951056516), VFNMS(LDK(KP618033988), T3y, T3z)); T3w = VSUB(T3q, T3t); T3u = VADD(T3q, T3t); { V T2B, T2x, T2H, T2i; T2B = VFMA(LDK(KP734762448), T2w, T2v); T2x = VFNMS(LDK(KP734762448), T2w, T2v); T2H = VFNMS(LDK(KP734762448), T2h, T2e); T2i = VFMA(LDK(KP734762448), T2h, T2e); { V T30, T35, T3c, T2S, T3v; T30 = VFNMS(LDK(KP921078979), T2Z, T2Y); T35 = VFMA(LDK(KP921078979), T2Z, T2Y); T3c = VFMA(LDK(KP982009705), T2R, T2Q); T2S = VFNMS(LDK(KP982009705), T2R, T2Q); T1W = VFMA(LDK(KP912575812), T1V, T1U); T1Z = VFNMS(LDK(KP912575812), T1V, T1U); T1O = VFMA(LDK(KP912575812), T1N, T1M); T25 = VFNMS(LDK(KP912575812), T1N, T1M); ST(&(xo[0]), VADD(T3u, T3n), ovs, &(xo[0])); T3v = VFNMS(LDK(KP250000000), T3u, T3n); { V T2y, T2J, T2q, T2D; T2y = VFMA(LDK(KP945422727), T2x, T2u); T2J = VFMA(LDK(KP522616830), T2x, T2I); T2q = VFMA(LDK(KP956723877), T2p, T2i); T2D = VFNMS(LDK(KP522616830), T2i, T2C); { V T3e, T31, T36, T2T; T3e = VFMA(LDK(KP906616052), T30, T2X); T31 = VFNMS(LDK(KP906616052), T30, T2X); T36 = VFNMS(LDK(KP923225144), T2S, T2P); T2T = VFMA(LDK(KP923225144), T2S, T2P); { V T3k, T3d, T3x, T3B; T3k = VFNMS(LDK(KP669429328), T3b, T3c); T3d = VFMA(LDK(KP570584518), T3c, T3b); T3x = VFMA(LDK(KP559016994), T3w, T3v); T3B = VFNMS(LDK(KP559016994), T3w, T3v); { V T2A, T2K, T2r, T2E; T2A = VMUL(LDK(KP998026728), VFMA(LDK(KP952936919), T2z, T2y)); T2K = VFNMS(LDK(KP690983005), T2J, T2u); T2r = VFMA(LDK(KP992114701), T2q, T2b); T2E = VFMA(LDK(KP763932022), T2D, T2p); { V T32, T3a, T37, T3h; T32 = VMUL(LDK(KP998026728), VFNMS(LDK(KP952936919), T2z, T31)); T3a = VFMA(LDK(KP262346850), T31, T2z); T37 = VFNMS(LDK(KP997675361), T36, T35); T3h = VFNMS(LDK(KP904508497), T36, T34); { V T2U, T33, T3l, T3f; T2U = VFMA(LDK(KP949179823), T2T, T2b); T33 = VFNMS(LDK(KP237294955), T2T, T2b); T3l = VFNMS(LDK(KP669429328), T3e, T3k); T3f = VFMA(LDK(KP618033988), T3e, T3d); ST(&(xo[WS(os, 20)]), VFMAI(T3A, T3x), ovs, &(xo[0])); ST(&(xo[WS(os, 5)]), VFNMSI(T3A, T3x), ovs, &(xo[WS(os, 1)])); ST(&(xo[WS(os, 15)]), VFNMSI(T3C, T3B), ovs, &(xo[WS(os, 1)])); ST(&(xo[WS(os, 10)]), VFMAI(T3C, T3B), ovs, &(xo[0])); { V T2L, T2F, T38, T3i; T2L = VFMA(LDK(KP855719849), T2K, T2H); ST(&(xo[WS(os, 22)]), VFMAI(T2A, T2r), ovs, &(xo[0])); ST(&(xo[WS(os, 3)]), VFNMSI(T2A, T2r), ovs, &(xo[WS(os, 1)])); T2F = VFNMS(LDK(KP855719849), T2E, T2B); T38 = VFMA(LDK(KP560319534), T37, T34); T3i = VFNMS(LDK(KP681693190), T3h, T35); ST(&(xo[WS(os, 23)]), VFMAI(T32, T2U), ovs, &(xo[WS(os, 1)])); ST(&(xo[WS(os, 2)]), VFNMSI(T32, T2U), ovs, &(xo[0])); T3m = VMUL(LDK(KP951056516), VFNMS(LDK(KP876306680), T3l, T3a)); T3g = VMUL(LDK(KP951056516), VFNMS(LDK(KP949179823), T3f, T3a)); T2M = VMUL(LDK(KP951056516), VFNMS(LDK(KP992114701), T2L, T2z)); T2G = VFMA(LDK(KP897376177), T2F, T2b); T39 = VFNMS(LDK(KP949179823), T38, T33); T3j = VFNMS(LDK(KP860541664), T3i, T33); T21 = VFMA(LDK(KP447417479), T1O, T20); } } } } } } } } } } } T1P = VFNMS(LDK(KP809385824), T1O, T1L); ST(&(xo[WS(os, 17)]), VFMAI(T2M, T2G), ovs, &(xo[WS(os, 1)])); ST(&(xo[WS(os, 8)]), VFNMSI(T2M, T2G), ovs, &(xo[0])); ST(&(xo[WS(os, 12)]), VFMAI(T3g, T39), ovs, &(xo[0])); ST(&(xo[WS(os, 13)]), VFNMSI(T3g, T39), ovs, &(xo[WS(os, 1)])); ST(&(xo[WS(os, 7)]), VFMAI(T3m, T3j), ovs, &(xo[WS(os, 1)])); ST(&(xo[WS(os, 18)]), VFNMSI(T3m, T3j), ovs, &(xo[0])); T22 = VFMA(LDK(KP690983005), T21, T1L); T27 = VFMA(LDK(KP447417479), T1W, T26); T1X = VFMA(LDK(KP894834959), T1W, T1T); { V T1r, T1s, T1v, T1w; T1r = VFNMS(LDK(KP916574801), T1f, T1e); T1g = VFMA(LDK(KP916574801), T1f, T1e); T1k = VFNMS(LDK(KP831864738), T1j, T1i); T1s = VFMA(LDK(KP831864738), T1j, T1i); T1v = VFNMS(LDK(KP829049696), TF, Tq); TG = VFMA(LDK(KP829049696), TF, Tq); T1b = VFMA(LDK(KP831864738), T1a, TV); T1w = VFNMS(LDK(KP831864738), T1a, TV); T28 = VFNMS(LDK(KP763932022), T27, T1T); T1t = VFMA(LDK(KP904730450), T1s, T1r); T1y = VFNMS(LDK(KP904730450), T1s, T1r); T1x = VFMA(LDK(KP559154169), T1w, T1v); T1E = VFNMS(LDK(KP683113946), T1v, T1w); } } T1Q = VFNMS(LDK(KP992114701), T1P, Tb); T1Y = VMUL(LDK(KP951056516), VFNMS(LDK(KP992114701), T1X, T1q)); { V T1u, T1F, T1z, T1h, T1c, T23, T29; T23 = VFNMS(LDK(KP999544308), T22, T1Z); T29 = VFNMS(LDK(KP999544308), T28, T25); T1I = VMUL(LDK(KP951056516), VFMA(LDK(KP968583161), T1t, T1q)); T1u = VFNMS(LDK(KP242145790), T1t, T1q); T1F = VFMA(LDK(KP617882369), T1y, T1E); T1z = VFMA(LDK(KP559016994), T1y, T1x); T1h = VFNMS(LDK(KP904730450), T1b, TG); T1c = VFMA(LDK(KP904730450), T1b, TG); ST(&(xo[WS(os, 21)]), VFNMSI(T1Y, T1Q), ovs, &(xo[WS(os, 1)])); ST(&(xo[WS(os, 4)]), VFMAI(T1Y, T1Q), ovs, &(xo[0])); T24 = VFNMS(LDK(KP803003575), T23, Tb); T2a = VMUL(LDK(KP951056516), VFNMS(LDK(KP803003575), T29, T1q)); T1G = VMUL(LDK(KP951056516), VFNMS(LDK(KP876306680), T1F, T1u)); T1A = VMUL(LDK(KP951056516), VFMA(LDK(KP968583161), T1z, T1u)); T1l = VFNMS(LDK(KP904730450), T1k, T1h); T1B = VADD(T1g, T1h); T1H = VFMA(LDK(KP968583161), T1c, Tb); T1d = VFNMS(LDK(KP242145790), T1c, Tb); } } } } ST(&(xo[WS(os, 9)]), VFMAI(T2a, T24), ovs, &(xo[WS(os, 1)])); ST(&(xo[WS(os, 16)]), VFNMSI(T2a, T24), ovs, &(xo[0])); { V T1m, T1C, T1n, T1D; T1m = VFNMS(LDK(KP618033988), T1l, T1g); T1C = VFNMS(LDK(KP683113946), T1B, T1k); ST(&(xo[WS(os, 24)]), VFMAI(T1I, T1H), ovs, &(xo[0])); ST(&(xo[WS(os, 1)]), VFNMSI(T1I, T1H), ovs, &(xo[WS(os, 1)])); T1n = VFNMS(LDK(KP876091699), T1m, T1d); T1D = VFMA(LDK(KP792626838), T1C, T1d); ST(&(xo[WS(os, 19)]), VFMAI(T1A, T1n), ovs, &(xo[WS(os, 1)])); ST(&(xo[WS(os, 6)]), VFNMSI(T1A, T1n), ovs, &(xo[0])); ST(&(xo[WS(os, 14)]), VFMAI(T1G, T1D), ovs, &(xo[0])); ST(&(xo[WS(os, 11)]), VFNMSI(T1G, T1D), ovs, &(xo[WS(os, 1)])); } } } VLEAVE(); } static const kdft_desc desc = { 25, XSIMD_STRING("n1fv_25"), {43, 12, 181, 0}, &GENUS, 0, 0, 0, 0 }; void XSIMD(codelet_n1fv_25) (planner *p) { X(kdft_register) (p, n1fv_25, &desc); } #else /* HAVE_FMA */ /* Generated by: ../../../genfft/gen_notw_c.native -simd -compact -variables 4 -pipeline-latency 8 -n 25 -name n1fv_25 -include n1f.h */ /* * This function contains 224 FP additions, 140 FP multiplications, * (or, 146 additions, 62 multiplications, 78 fused multiply/add), * 115 stack variables, 40 constants, and 50 memory accesses */ #include "n1f.h" static void n1fv_25(const R *ri, const R *ii, R *ro, R *io, stride is, stride os, INT v, INT ivs, INT ovs) { DVK(KP809016994, +0.809016994374947424102293417182819058860154590); DVK(KP309016994, +0.309016994374947424102293417182819058860154590); DVK(KP770513242, +0.770513242775789230803009636396177847271667672); DVK(KP1_274847979, +1.274847979497379420353425623352032390869834596); DVK(KP992114701, +0.992114701314477831049793042785778521453036709); DVK(KP250666467, +0.250666467128608490746237519633017587885836494); DVK(KP637423989, +0.637423989748689710176712811676016195434917298); DVK(KP1_541026485, +1.541026485551578461606019272792355694543335344); DVK(KP125333233, +0.125333233564304245373118759816508793942918247); DVK(KP1_984229402, +1.984229402628955662099586085571557042906073418); DVK(KP248689887, +0.248689887164854788242283746006447968417567406); DVK(KP1_937166322, +1.937166322257262238980336750929471627672024806); DVK(KP497379774, +0.497379774329709576484567492012895936835134813); DVK(KP968583161, +0.968583161128631119490168375464735813836012403); DVK(KP904827052, +0.904827052466019527713668647932697593970413911); DVK(KP851558583, +0.851558583130145297725004891488503407959946084); DVK(KP425779291, +0.425779291565072648862502445744251703979973042); DVK(KP1_809654104, +1.809654104932039055427337295865395187940827822); DVK(KP844327925, +0.844327925502015078548558063966681505381659241); DVK(KP1_071653589, +1.071653589957993236542617535735279956127150691); DVK(KP481753674, +0.481753674101715274987191502872129653528542010); DVK(KP1_752613360, +1.752613360087727174616231807844125166798128477); DVK(KP535826794, +0.535826794978996618271308767867639978063575346); DVK(KP1_688655851, +1.688655851004030157097116127933363010763318483); DVK(KP963507348, +0.963507348203430549974383005744259307057084020); DVK(KP876306680, +0.876306680043863587308115903922062583399064238); DVK(KP998026728, +0.998026728428271561952336806863450553336905220); DVK(KP125581039, +0.125581039058626752152356449131262266244969664); DVK(KP684547105, +0.684547105928688673732283357621209269889519233); DVK(KP1_457937254, +1.457937254842823046293460638110518222745143328); DVK(KP062790519, +0.062790519529313376076178224565631133122484832); DVK(KP1_996053456, +1.996053456856543123904673613726901106673810439); DVK(KP1_369094211, +1.369094211857377347464566715242418539779038465); DVK(KP728968627, +0.728968627421411523146730319055259111372571664); DVK(KP293892626, +0.293892626146236564584352977319536384298826219); DVK(KP475528258, +0.475528258147576786058219666689691071702849317); DVK(KP250000000, +0.250000000000000000000000000000000000000000000); DVK(KP587785252, +0.587785252292473129168705954639072768597652438); DVK(KP951056516, +0.951056516295153572116439333379382143405698634); DVK(KP559016994, +0.559016994374947424102293417182819058860154590); { INT i; const R *xi; R *xo; xi = ri; xo = ro; for (i = v; i > 0; i = i - VL, xi = xi + (VL * ivs), xo = xo + (VL * ovs), MAKE_VOLATILE_STRIDE(50, is), MAKE_VOLATILE_STRIDE(50, os)) { V T7, T1g, T26, Ta, T2R, T2N, T2O, T2P, T19, T1Y, T16, T1Z, T1a, T2v, T1l; V T2m, TU, T21, TR, T22, TV, T2u, T1k, T2l, T2K, T2L, T2M, TE, T1R, TB; V T1S, TF, T2r, T1i, T2j, Tp, T1U, Tm, T1V, Tq, T2s, T1h, T2i; { V T8, T6, T1f, T3, T1e, T25, T9; T8 = LD(&(xi[0]), ivs, &(xi[0])); { V T4, T5, T1, T2; T4 = LD(&(xi[WS(is, 10)]), ivs, &(xi[0])); T5 = LD(&(xi[WS(is, 15)]), ivs, &(xi[WS(is, 1)])); T6 = VADD(T4, T5); T1f = VSUB(T4, T5); T1 = LD(&(xi[WS(is, 5)]), ivs, &(xi[WS(is, 1)])); T2 = LD(&(xi[WS(is, 20)]), ivs, &(xi[0])); T3 = VADD(T1, T2); T1e = VSUB(T1, T2); } T7 = VMUL(LDK(KP559016994), VSUB(T3, T6)); T1g = VFMA(LDK(KP951056516), T1e, VMUL(LDK(KP587785252), T1f)); T25 = VMUL(LDK(KP951056516), T1f); T26 = VFNMS(LDK(KP587785252), T1e, T25); T9 = VADD(T3, T6); Ta = VFNMS(LDK(KP250000000), T9, T8); T2R = VADD(T8, T9); } { V TO, T13, TN, TT, TP, TS, T12, T18, T14, T17, T15, TQ; TO = LD(&(xi[WS(is, 2)]), ivs, &(xi[0])); T13 = LD(&(xi[WS(is, 3)]), ivs, &(xi[WS(is, 1)])); { V TH, TI, TJ, TK, TL, TM; TH = LD(&(xi[WS(is, 7)]), ivs, &(xi[WS(is, 1)])); TI = LD(&(xi[WS(is, 22)]), ivs, &(xi[0])); TJ = VADD(TH, TI); TK = LD(&(xi[WS(is, 12)]), ivs, &(xi[0])); TL = LD(&(xi[WS(is, 17)]), ivs, &(xi[WS(is, 1)])); TM = VADD(TK, TL); TN = VMUL(LDK(KP559016994), VSUB(TJ, TM)); TT = VSUB(TK, TL); TP = VADD(TJ, TM); TS = VSUB(TH, TI); } { V TW, TX, TY, TZ, T10, T11; TW = LD(&(xi[WS(is, 8)]), ivs, &(xi[0])); TX = LD(&(xi[WS(is, 23)]), ivs, &(xi[WS(is, 1)])); TY = VADD(TW, TX); TZ = LD(&(xi[WS(is, 13)]), ivs, &(xi[WS(is, 1)])); T10 = LD(&(xi[WS(is, 18)]), ivs, &(xi[0])); T11 = VADD(TZ, T10); T12 = VMUL(LDK(KP559016994), VSUB(TY, T11)); T18 = VSUB(TZ, T10); T14 = VADD(TY, T11); T17 = VSUB(TW, TX); } T2N = VADD(TO, TP); T2O = VADD(T13, T14); T2P = VADD(T2N, T2O); T19 = VFMA(LDK(KP475528258), T17, VMUL(LDK(KP293892626), T18)); T1Y = VFNMS(LDK(KP293892626), T17, VMUL(LDK(KP475528258), T18)); T15 = VFNMS(LDK(KP250000000), T14, T13); T16 = VADD(T12, T15); T1Z = VSUB(T15, T12); T1a = VFNMS(LDK(KP1_369094211), T19, VMUL(LDK(KP728968627), T16)); T2v = VFMA(LDK(KP1_996053456), T1Y, VMUL(LDK(KP062790519), T1Z)); T1l = VFMA(LDK(KP1_457937254), T19, VMUL(LDK(KP684547105), T16)); T2m = VFNMS(LDK(KP998026728), T1Z, VMUL(LDK(KP125581039), T1Y)); TU = VFMA(LDK(KP475528258), TS, VMUL(LDK(KP293892626), TT)); T21 = VFNMS(LDK(KP293892626), TS, VMUL(LDK(KP475528258), TT)); TQ = VFNMS(LDK(KP250000000), TP, TO); TR = VADD(TN, TQ); T22 = VSUB(TQ, TN); TV = VFNMS(LDK(KP963507348), TU, VMUL(LDK(KP876306680), TR)); T2u = VFMA(LDK(KP1_688655851), T21, VMUL(LDK(KP535826794), T22)); T1k = VFMA(LDK(KP1_752613360), TU, VMUL(LDK(KP481753674), TR)); T2l = VFNMS(LDK(KP844327925), T22, VMUL(LDK(KP1_071653589), T21)); } { V Tj, Ty, Ti, To, Tk, Tn, Tx, TD, Tz, TC, TA, Tl; Tj = LD(&(xi[WS(is, 1)]), ivs, &(xi[WS(is, 1)])); Ty = LD(&(xi[WS(is, 4)]), ivs, &(xi[0])); { V Tc, Td, Te, Tf, Tg, Th; Tc = LD(&(xi[WS(is, 6)]), ivs, &(xi[0])); Td = LD(&(xi[WS(is, 21)]), ivs, &(xi[WS(is, 1)])); Te = VADD(Tc, Td); Tf = LD(&(xi[WS(is, 11)]), ivs, &(xi[WS(is, 1)])); Tg = LD(&(xi[WS(is, 16)]), ivs, &(xi[0])); Th = VADD(Tf, Tg); Ti = VMUL(LDK(KP559016994), VSUB(Te, Th)); To = VSUB(Tf, Tg); Tk = VADD(Te, Th); Tn = VSUB(Tc, Td); } { V Tr, Ts, Tt, Tu, Tv, Tw; Tr = LD(&(xi[WS(is, 9)]), ivs, &(xi[WS(is, 1)])); Ts = LD(&(xi[WS(is, 24)]), ivs, &(xi[0])); Tt = VADD(Tr, Ts); Tu = LD(&(xi[WS(is, 14)]), ivs, &(xi[0])); Tv = LD(&(xi[WS(is, 19)]), ivs, &(xi[WS(is, 1)])); Tw = VADD(Tu, Tv); Tx = VMUL(LDK(KP559016994), VSUB(Tt, Tw)); TD = VSUB(Tu, Tv); Tz = VADD(Tt, Tw); TC = VSUB(Tr, Ts); } T2K = VADD(Tj, Tk); T2L = VADD(Ty, Tz); T2M = VADD(T2K, T2L); TE = VFMA(LDK(KP475528258), TC, VMUL(LDK(KP293892626), TD)); T1R = VFNMS(LDK(KP293892626), TC, VMUL(LDK(KP475528258), TD)); TA = VFNMS(LDK(KP250000000), Tz, Ty); TB = VADD(Tx, TA); T1S = VSUB(TA, Tx); TF = VFNMS(LDK(KP1_688655851), TE, VMUL(LDK(KP535826794), TB)); T2r = VFNMS(LDK(KP425779291), T1S, VMUL(LDK(KP1_809654104), T1R)); T1i = VFMA(LDK(KP1_071653589), TE, VMUL(LDK(KP844327925), TB)); T2j = VFMA(LDK(KP851558583), T1R, VMUL(LDK(KP904827052), T1S)); Tp = VFMA(LDK(KP475528258), Tn, VMUL(LDK(KP293892626), To)); T1U = VFNMS(LDK(KP293892626), Tn, VMUL(LDK(KP475528258), To)); Tl = VFNMS(LDK(KP250000000), Tk, Tj); Tm = VADD(Ti, Tl); T1V = VSUB(Tl, Ti); Tq = VFNMS(LDK(KP497379774), Tp, VMUL(LDK(KP968583161), Tm)); T2s = VFMA(LDK(KP963507348), T1U, VMUL(LDK(KP876306680), T1V)); T1h = VFMA(LDK(KP1_937166322), Tp, VMUL(LDK(KP248689887), Tm)); T2i = VFNMS(LDK(KP481753674), T1V, VMUL(LDK(KP1_752613360), T1U)); } { V T2Q, T2S, T2T, T2X, T2Y, T2V, T2W, T2Z, T2U; T2Q = VMUL(LDK(KP559016994), VSUB(T2M, T2P)); T2S = VADD(T2M, T2P); T2T = VFNMS(LDK(KP250000000), T2S, T2R); T2V = VSUB(T2K, T2L); T2W = VSUB(T2N, T2O); T2X = VBYI(VFMA(LDK(KP951056516), T2V, VMUL(LDK(KP587785252), T2W))); T2Y = VBYI(VFNMS(LDK(KP587785252), T2V, VMUL(LDK(KP951056516), T2W))); ST(&(xo[0]), VADD(T2R, T2S), ovs, &(xo[0])); T2Z = VSUB(T2T, T2Q); ST(&(xo[WS(os, 10)]), VADD(T2Y, T2Z), ovs, &(xo[0])); ST(&(xo[WS(os, 15)]), VSUB(T2Z, T2Y), ovs, &(xo[WS(os, 1)])); T2U = VADD(T2Q, T2T); ST(&(xo[WS(os, 5)]), VSUB(T2U, T2X), ovs, &(xo[WS(os, 1)])); ST(&(xo[WS(os, 20)]), VADD(T2X, T2U), ovs, &(xo[0])); } { V T2t, T2y, T2z, T2w, T1T, T1W, T1X, T2c, T2d, T2e, T29, T2a, T2b, T20, T23; V T24, T2p, T2o, T2q, T28, T2D, T2C, T2E, T2x, T2F; T2t = VSUB(T2r, T2s); T2y = VADD(T2i, T2j); T2z = VSUB(T2l, T2m); T2w = VSUB(T2u, T2v); T1T = VFNMS(LDK(KP125333233), T1S, VMUL(LDK(KP1_984229402), T1R)); T1W = VFMA(LDK(KP1_457937254), T1U, VMUL(LDK(KP684547105), T1V)); T1X = VSUB(T1T, T1W); T2c = VFNMS(LDK(KP1_996053456), T21, VMUL(LDK(KP062790519), T22)); T2d = VFMA(LDK(KP1_541026485), T1Y, VMUL(LDK(KP637423989), T1Z)); T2e = VSUB(T2c, T2d); T29 = VFNMS(LDK(KP1_369094211), T1U, VMUL(LDK(KP728968627), T1V)); T2a = VFMA(LDK(KP250666467), T1R, VMUL(LDK(KP992114701), T1S)); T2b = VSUB(T29, T2a); T20 = VFNMS(LDK(KP770513242), T1Z, VMUL(LDK(KP1_274847979), T1Y)); T23 = VFMA(LDK(KP125581039), T21, VMUL(LDK(KP998026728), T22)); T24 = VSUB(T20, T23); { V T2k, T2n, T2A, T2B; T2k = VSUB(T2i, T2j); T2n = VADD(T2l, T2m); T2p = VADD(T2k, T2n); T2o = VMUL(LDK(KP559016994), VSUB(T2k, T2n)); T2q = VFNMS(LDK(KP250000000), T2p, T26); T28 = VSUB(Ta, T7); T2A = VADD(T2s, T2r); T2B = VADD(T2u, T2v); T2D = VADD(T2A, T2B); T2C = VMUL(LDK(KP559016994), VSUB(T2A, T2B)); T2E = VFNMS(LDK(KP250000000), T2D, T28); } { V T2I, T2J, T27, T2f; T2I = VBYI(VADD(T26, T2p)); T2J = VADD(T28, T2D); ST(&(xo[WS(os, 2)]), VADD(T2I, T2J), ovs, &(xo[0])); ST(&(xo[WS(os, 23)]), VSUB(T2J, T2I), ovs, &(xo[WS(os, 1)])); T27 = VBYI(VSUB(VADD(T1X, T24), T26)); T2f = VADD(T28, VADD(T2b, T2e)); ST(&(xo[WS(os, 3)]), VADD(T27, T2f), ovs, &(xo[WS(os, 1)])); ST(&(xo[WS(os, 22)]), VSUB(T2f, T27), ovs, &(xo[0])); } T2x = VBYI(VADD(T2o, VADD(T2q, VFNMS(LDK(KP587785252), T2w, VMUL(LDK(KP951056516), T2t))))); T2F = VFMA(LDK(KP951056516), T2y, VFMA(LDK(KP587785252), T2z, VADD(T2C, T2E))); ST(&(xo[WS(os, 7)]), VADD(T2x, T2F), ovs, &(xo[WS(os, 1)])); ST(&(xo[WS(os, 18)]), VSUB(T2F, T2x), ovs, &(xo[0])); { V T2G, T2H, T2g, T2h; T2G = VBYI(VADD(T2q, VSUB(VFMA(LDK(KP587785252), T2t, VMUL(LDK(KP951056516), T2w)), T2o))); T2H = VFMA(LDK(KP587785252), T2y, VSUB(VFNMS(LDK(KP951056516), T2z, T2E), T2C)); ST(&(xo[WS(os, 12)]), VADD(T2G, T2H), ovs, &(xo[0])); ST(&(xo[WS(os, 13)]), VSUB(T2H, T2G), ovs, &(xo[WS(os, 1)])); T2g = VFMA(LDK(KP309016994), T2b, VFNMS(LDK(KP809016994), T2e, VFNMS(LDK(KP587785252), VADD(T23, T20), VFNMS(LDK(KP951056516), VADD(T1W, T1T), T28)))); T2h = VBYI(VSUB(VFNMS(LDK(KP587785252), VADD(T2c, T2d), VFNMS(LDK(KP809016994), T24, VFNMS(LDK(KP951056516), VADD(T29, T2a), VMUL(LDK(KP309016994), T1X)))), T26)); ST(&(xo[WS(os, 17)]), VSUB(T2g, T2h), ovs, &(xo[WS(os, 1)])); ST(&(xo[WS(os, 8)]), VADD(T2g, T2h), ovs, &(xo[0])); } } { V T1p, T1u, T1w, T1q, T1B, T1C, T1D, T1L, T1M, T1N, T1I, T1J, T1K, T1E, T1F; V T1G, T1n, T1r, T1s, Tb, T1c, T1v, T1x, T1t, T1y; T1p = VSUB(TF, Tq); T1u = VSUB(T1i, T1h); T1w = VSUB(T1l, T1k); T1q = VSUB(TV, T1a); T1B = VFMA(LDK(KP1_688655851), Tp, VMUL(LDK(KP535826794), Tm)); T1C = VFMA(LDK(KP1_541026485), TE, VMUL(LDK(KP637423989), TB)); T1D = VSUB(T1B, T1C); T1L = VFMA(LDK(KP851558583), TU, VMUL(LDK(KP904827052), TR)); T1M = VFMA(LDK(KP1_984229402), T19, VMUL(LDK(KP125333233), T16)); T1N = VADD(T1L, T1M); T1I = VFNMS(LDK(KP844327925), Tm, VMUL(LDK(KP1_071653589), Tp)); T1J = VFNMS(LDK(KP1_274847979), TE, VMUL(LDK(KP770513242), TB)); T1K = VADD(T1I, T1J); T1E = VFNMS(LDK(KP425779291), TR, VMUL(LDK(KP1_809654104), TU)); T1F = VFNMS(LDK(KP992114701), T16, VMUL(LDK(KP250666467), T19)); T1G = VADD(T1E, T1F); { V T1j, T1m, TG, T1b; T1j = VADD(T1h, T1i); T1m = VADD(T1k, T1l); T1n = VADD(T1j, T1m); T1r = VFMS(LDK(KP250000000), T1n, T1g); T1s = VMUL(LDK(KP559016994), VSUB(T1m, T1j)); Tb = VADD(T7, Ta); TG = VADD(Tq, TF); T1b = VADD(TV, T1a); T1c = VADD(TG, T1b); T1v = VFNMS(LDK(KP250000000), T1c, Tb); T1x = VMUL(LDK(KP559016994), VSUB(TG, T1b)); } { V T1d, T1o, T1H, T1O; T1d = VADD(Tb, T1c); T1o = VBYI(VADD(T1g, T1n)); ST(&(xo[WS(os, 1)]), VSUB(T1d, T1o), ovs, &(xo[WS(os, 1)])); ST(&(xo[WS(os, 24)]), VADD(T1d, T1o), ovs, &(xo[0])); T1H = VADD(Tb, VADD(T1D, T1G)); T1O = VBYI(VADD(T1g, VSUB(T1K, T1N))); ST(&(xo[WS(os, 21)]), VSUB(T1H, T1O), ovs, &(xo[WS(os, 1)])); ST(&(xo[WS(os, 4)]), VADD(T1H, T1O), ovs, &(xo[0])); } T1t = VBYI(VADD(VFMA(LDK(KP587785252), T1p, VMUL(LDK(KP951056516), T1q)), VSUB(T1r, T1s))); T1y = VFMA(LDK(KP587785252), T1u, VFNMS(LDK(KP951056516), T1w, VSUB(T1v, T1x))); ST(&(xo[WS(os, 11)]), VADD(T1t, T1y), ovs, &(xo[WS(os, 1)])); ST(&(xo[WS(os, 14)]), VSUB(T1y, T1t), ovs, &(xo[0])); { V T1z, T1A, T1P, T1Q; T1z = VBYI(VADD(VFNMS(LDK(KP587785252), T1q, VMUL(LDK(KP951056516), T1p)), VADD(T1r, T1s))); T1A = VFMA(LDK(KP951056516), T1u, VADD(T1x, VFMA(LDK(KP587785252), T1w, T1v))); ST(&(xo[WS(os, 6)]), VADD(T1z, T1A), ovs, &(xo[0])); ST(&(xo[WS(os, 19)]), VSUB(T1A, T1z), ovs, &(xo[WS(os, 1)])); T1P = VBYI(VADD(T1g, VFMA(LDK(KP309016994), T1K, VFMA(LDK(KP587785252), VSUB(T1F, T1E), VFNMS(LDK(KP951056516), VADD(T1B, T1C), VMUL(LDK(KP809016994), T1N)))))); T1Q = VFMA(LDK(KP309016994), T1D, VFMA(LDK(KP951056516), VSUB(T1I, T1J), VFMA(LDK(KP587785252), VSUB(T1M, T1L), VFNMS(LDK(KP809016994), T1G, Tb)))); ST(&(xo[WS(os, 9)]), VADD(T1P, T1Q), ovs, &(xo[WS(os, 1)])); ST(&(xo[WS(os, 16)]), VSUB(T1Q, T1P), ovs, &(xo[0])); } } } } VLEAVE(); } static const kdft_desc desc = { 25, XSIMD_STRING("n1fv_25"), {146, 62, 78, 0}, &GENUS, 0, 0, 0, 0 }; void XSIMD(codelet_n1fv_25) (planner *p) { X(kdft_register) (p, n1fv_25, &desc); } #endif /* HAVE_FMA */ fftw-3.3.3/dft/simd/common/n1bv_2.c0000644000175000017500000000637012054410352013645 00000000000000/* * Copyright (c) 2003, 2007-11 Matteo Frigo * Copyright (c) 2003, 2007-11 Massachusetts Institute of Technology * * This program is free software; you can redistribute it and/or modify * it under the terms of the GNU General Public License as published by * the Free Software Foundation; either version 2 of the License, or * (at your option) any later version. * * This program is distributed in the hope that it will be useful, * but WITHOUT ANY WARRANTY; without even the implied warranty of * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the * GNU General Public License for more details. * * You should have received a copy of the GNU General Public License * along with this program; if not, write to the Free Software * Foundation, Inc., 51 Franklin Street, Fifth Floor, Boston, MA 02110-1301 USA * */ /* This file was automatically generated --- DO NOT EDIT */ /* Generated on Sun Nov 25 07:36:58 EST 2012 */ #include "codelet-dft.h" #ifdef HAVE_FMA /* Generated by: ../../../genfft/gen_notw_c.native -fma -reorder-insns -schedule-for-pipeline -simd -compact -variables 4 -pipeline-latency 8 -sign 1 -n 2 -name n1bv_2 -include n1b.h */ /* * This function contains 2 FP additions, 0 FP multiplications, * (or, 2 additions, 0 multiplications, 0 fused multiply/add), * 5 stack variables, 0 constants, and 4 memory accesses */ #include "n1b.h" static void n1bv_2(const R *ri, const R *ii, R *ro, R *io, stride is, stride os, INT v, INT ivs, INT ovs) { { INT i; const R *xi; R *xo; xi = ii; xo = io; for (i = v; i > 0; i = i - VL, xi = xi + (VL * ivs), xo = xo + (VL * ovs), MAKE_VOLATILE_STRIDE(4, is), MAKE_VOLATILE_STRIDE(4, os)) { V T1, T2; T1 = LD(&(xi[0]), ivs, &(xi[0])); T2 = LD(&(xi[WS(is, 1)]), ivs, &(xi[WS(is, 1)])); ST(&(xo[0]), VADD(T1, T2), ovs, &(xo[0])); ST(&(xo[WS(os, 1)]), VSUB(T1, T2), ovs, &(xo[WS(os, 1)])); } } VLEAVE(); } static const kdft_desc desc = { 2, XSIMD_STRING("n1bv_2"), {2, 0, 0, 0}, &GENUS, 0, 0, 0, 0 }; void XSIMD(codelet_n1bv_2) (planner *p) { X(kdft_register) (p, n1bv_2, &desc); } #else /* HAVE_FMA */ /* Generated by: ../../../genfft/gen_notw_c.native -simd -compact -variables 4 -pipeline-latency 8 -sign 1 -n 2 -name n1bv_2 -include n1b.h */ /* * This function contains 2 FP additions, 0 FP multiplications, * (or, 2 additions, 0 multiplications, 0 fused multiply/add), * 5 stack variables, 0 constants, and 4 memory accesses */ #include "n1b.h" static void n1bv_2(const R *ri, const R *ii, R *ro, R *io, stride is, stride os, INT v, INT ivs, INT ovs) { { INT i; const R *xi; R *xo; xi = ii; xo = io; for (i = v; i > 0; i = i - VL, xi = xi + (VL * ivs), xo = xo + (VL * ovs), MAKE_VOLATILE_STRIDE(4, is), MAKE_VOLATILE_STRIDE(4, os)) { V T1, T2; T1 = LD(&(xi[0]), ivs, &(xi[0])); T2 = LD(&(xi[WS(is, 1)]), ivs, &(xi[WS(is, 1)])); ST(&(xo[WS(os, 1)]), VSUB(T1, T2), ovs, &(xo[WS(os, 1)])); ST(&(xo[0]), VADD(T1, T2), ovs, &(xo[0])); } } VLEAVE(); } static const kdft_desc desc = { 2, XSIMD_STRING("n1bv_2"), {2, 0, 0, 0}, &GENUS, 0, 0, 0, 0 }; void XSIMD(codelet_n1bv_2) (planner *p) { X(kdft_register) (p, n1bv_2, &desc); } #endif /* HAVE_FMA */ fftw-3.3.3/dft/simd/common/n1bv_3.c0000644000175000017500000001003112054410352013633 00000000000000/* * Copyright (c) 2003, 2007-11 Matteo Frigo * Copyright (c) 2003, 2007-11 Massachusetts Institute of Technology * * This program is free software; you can redistribute it and/or modify * it under the terms of the GNU General Public License as published by * the Free Software Foundation; either version 2 of the License, or * (at your option) any later version. * * This program is distributed in the hope that it will be useful, * but WITHOUT ANY WARRANTY; without even the implied warranty of * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the * GNU General Public License for more details. * * You should have received a copy of the GNU General Public License * along with this program; if not, write to the Free Software * Foundation, Inc., 51 Franklin Street, Fifth Floor, Boston, MA 02110-1301 USA * */ /* This file was automatically generated --- DO NOT EDIT */ /* Generated on Sun Nov 25 07:36:58 EST 2012 */ #include "codelet-dft.h" #ifdef HAVE_FMA /* Generated by: ../../../genfft/gen_notw_c.native -fma -reorder-insns -schedule-for-pipeline -simd -compact -variables 4 -pipeline-latency 8 -sign 1 -n 3 -name n1bv_3 -include n1b.h */ /* * This function contains 6 FP additions, 4 FP multiplications, * (or, 3 additions, 1 multiplications, 3 fused multiply/add), * 11 stack variables, 2 constants, and 6 memory accesses */ #include "n1b.h" static void n1bv_3(const R *ri, const R *ii, R *ro, R *io, stride is, stride os, INT v, INT ivs, INT ovs) { DVK(KP500000000, +0.500000000000000000000000000000000000000000000); DVK(KP866025403, +0.866025403784438646763723170752936183471402627); { INT i; const R *xi; R *xo; xi = ii; xo = io; for (i = v; i > 0; i = i - VL, xi = xi + (VL * ivs), xo = xo + (VL * ovs), MAKE_VOLATILE_STRIDE(6, is), MAKE_VOLATILE_STRIDE(6, os)) { V T1, T2, T3, T6, T4, T5; T1 = LD(&(xi[0]), ivs, &(xi[0])); T2 = LD(&(xi[WS(is, 1)]), ivs, &(xi[WS(is, 1)])); T3 = LD(&(xi[WS(is, 2)]), ivs, &(xi[0])); T6 = VMUL(LDK(KP866025403), VSUB(T2, T3)); T4 = VADD(T2, T3); T5 = VFNMS(LDK(KP500000000), T4, T1); ST(&(xo[0]), VADD(T1, T4), ovs, &(xo[0])); ST(&(xo[WS(os, 2)]), VFNMSI(T6, T5), ovs, &(xo[0])); ST(&(xo[WS(os, 1)]), VFMAI(T6, T5), ovs, &(xo[WS(os, 1)])); } } VLEAVE(); } static const kdft_desc desc = { 3, XSIMD_STRING("n1bv_3"), {3, 1, 3, 0}, &GENUS, 0, 0, 0, 0 }; void XSIMD(codelet_n1bv_3) (planner *p) { X(kdft_register) (p, n1bv_3, &desc); } #else /* HAVE_FMA */ /* Generated by: ../../../genfft/gen_notw_c.native -simd -compact -variables 4 -pipeline-latency 8 -sign 1 -n 3 -name n1bv_3 -include n1b.h */ /* * This function contains 6 FP additions, 2 FP multiplications, * (or, 5 additions, 1 multiplications, 1 fused multiply/add), * 11 stack variables, 2 constants, and 6 memory accesses */ #include "n1b.h" static void n1bv_3(const R *ri, const R *ii, R *ro, R *io, stride is, stride os, INT v, INT ivs, INT ovs) { DVK(KP500000000, +0.500000000000000000000000000000000000000000000); DVK(KP866025403, +0.866025403784438646763723170752936183471402627); { INT i; const R *xi; R *xo; xi = ii; xo = io; for (i = v; i > 0; i = i - VL, xi = xi + (VL * ivs), xo = xo + (VL * ovs), MAKE_VOLATILE_STRIDE(6, is), MAKE_VOLATILE_STRIDE(6, os)) { V T4, T3, T5, T1, T2, T6; T4 = LD(&(xi[0]), ivs, &(xi[0])); T1 = LD(&(xi[WS(is, 1)]), ivs, &(xi[WS(is, 1)])); T2 = LD(&(xi[WS(is, 2)]), ivs, &(xi[0])); T3 = VBYI(VMUL(LDK(KP866025403), VSUB(T1, T2))); T5 = VADD(T1, T2); ST(&(xo[0]), VADD(T4, T5), ovs, &(xo[0])); T6 = VFNMS(LDK(KP500000000), T5, T4); ST(&(xo[WS(os, 1)]), VADD(T3, T6), ovs, &(xo[WS(os, 1)])); ST(&(xo[WS(os, 2)]), VSUB(T6, T3), ovs, &(xo[0])); } } VLEAVE(); } static const kdft_desc desc = { 3, XSIMD_STRING("n1bv_3"), {5, 1, 1, 0}, &GENUS, 0, 0, 0, 0 }; void XSIMD(codelet_n1bv_3) (planner *p) { X(kdft_register) (p, n1bv_3, &desc); } #endif /* HAVE_FMA */ fftw-3.3.3/dft/simd/common/n1bv_4.c0000644000175000017500000000776712054410352013662 00000000000000/* * Copyright (c) 2003, 2007-11 Matteo Frigo * Copyright (c) 2003, 2007-11 Massachusetts Institute of Technology * * This program is free software; you can redistribute it and/or modify * it under the terms of the GNU General Public License as published by * the Free Software Foundation; either version 2 of the License, or * (at your option) any later version. * * This program is distributed in the hope that it will be useful, * but WITHOUT ANY WARRANTY; without even the implied warranty of * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the * GNU General Public License for more details. * * You should have received a copy of the GNU General Public License * along with this program; if not, write to the Free Software * Foundation, Inc., 51 Franklin Street, Fifth Floor, Boston, MA 02110-1301 USA * */ /* This file was automatically generated --- DO NOT EDIT */ /* Generated on Sun Nov 25 07:36:58 EST 2012 */ #include "codelet-dft.h" #ifdef HAVE_FMA /* Generated by: ../../../genfft/gen_notw_c.native -fma -reorder-insns -schedule-for-pipeline -simd -compact -variables 4 -pipeline-latency 8 -sign 1 -n 4 -name n1bv_4 -include n1b.h */ /* * This function contains 8 FP additions, 2 FP multiplications, * (or, 6 additions, 0 multiplications, 2 fused multiply/add), * 11 stack variables, 0 constants, and 8 memory accesses */ #include "n1b.h" static void n1bv_4(const R *ri, const R *ii, R *ro, R *io, stride is, stride os, INT v, INT ivs, INT ovs) { { INT i; const R *xi; R *xo; xi = ii; xo = io; for (i = v; i > 0; i = i - VL, xi = xi + (VL * ivs), xo = xo + (VL * ovs), MAKE_VOLATILE_STRIDE(8, is), MAKE_VOLATILE_STRIDE(8, os)) { V T1, T2, T4, T5; T1 = LD(&(xi[0]), ivs, &(xi[0])); T2 = LD(&(xi[WS(is, 2)]), ivs, &(xi[0])); T4 = LD(&(xi[WS(is, 1)]), ivs, &(xi[WS(is, 1)])); T5 = LD(&(xi[WS(is, 3)]), ivs, &(xi[WS(is, 1)])); { V T3, T7, T6, T8; T3 = VSUB(T1, T2); T7 = VADD(T1, T2); T6 = VSUB(T4, T5); T8 = VADD(T4, T5); ST(&(xo[WS(os, 2)]), VSUB(T7, T8), ovs, &(xo[0])); ST(&(xo[0]), VADD(T7, T8), ovs, &(xo[0])); ST(&(xo[WS(os, 1)]), VFMAI(T6, T3), ovs, &(xo[WS(os, 1)])); ST(&(xo[WS(os, 3)]), VFNMSI(T6, T3), ovs, &(xo[WS(os, 1)])); } } } VLEAVE(); } static const kdft_desc desc = { 4, XSIMD_STRING("n1bv_4"), {6, 0, 2, 0}, &GENUS, 0, 0, 0, 0 }; void XSIMD(codelet_n1bv_4) (planner *p) { X(kdft_register) (p, n1bv_4, &desc); } #else /* HAVE_FMA */ /* Generated by: ../../../genfft/gen_notw_c.native -simd -compact -variables 4 -pipeline-latency 8 -sign 1 -n 4 -name n1bv_4 -include n1b.h */ /* * This function contains 8 FP additions, 0 FP multiplications, * (or, 8 additions, 0 multiplications, 0 fused multiply/add), * 11 stack variables, 0 constants, and 8 memory accesses */ #include "n1b.h" static void n1bv_4(const R *ri, const R *ii, R *ro, R *io, stride is, stride os, INT v, INT ivs, INT ovs) { { INT i; const R *xi; R *xo; xi = ii; xo = io; for (i = v; i > 0; i = i - VL, xi = xi + (VL * ivs), xo = xo + (VL * ovs), MAKE_VOLATILE_STRIDE(8, is), MAKE_VOLATILE_STRIDE(8, os)) { V T3, T7, T6, T8; { V T1, T2, T4, T5; T1 = LD(&(xi[0]), ivs, &(xi[0])); T2 = LD(&(xi[WS(is, 2)]), ivs, &(xi[0])); T3 = VSUB(T1, T2); T7 = VADD(T1, T2); T4 = LD(&(xi[WS(is, 1)]), ivs, &(xi[WS(is, 1)])); T5 = LD(&(xi[WS(is, 3)]), ivs, &(xi[WS(is, 1)])); T6 = VBYI(VSUB(T4, T5)); T8 = VADD(T4, T5); } ST(&(xo[WS(os, 3)]), VSUB(T3, T6), ovs, &(xo[WS(os, 1)])); ST(&(xo[0]), VADD(T7, T8), ovs, &(xo[0])); ST(&(xo[WS(os, 1)]), VADD(T3, T6), ovs, &(xo[WS(os, 1)])); ST(&(xo[WS(os, 2)]), VSUB(T7, T8), ovs, &(xo[0])); } } VLEAVE(); } static const kdft_desc desc = { 4, XSIMD_STRING("n1bv_4"), {8, 0, 0, 0}, &GENUS, 0, 0, 0, 0 }; void XSIMD(codelet_n1bv_4) (planner *p) { X(kdft_register) (p, n1bv_4, &desc); } #endif /* HAVE_FMA */ fftw-3.3.3/dft/simd/common/n1bv_5.c0000644000175000017500000001275012054410352013647 00000000000000/* * Copyright (c) 2003, 2007-11 Matteo Frigo * Copyright (c) 2003, 2007-11 Massachusetts Institute of Technology * * This program is free software; you can redistribute it and/or modify * it under the terms of the GNU General Public License as published by * the Free Software Foundation; either version 2 of the License, or * (at your option) any later version. * * This program is distributed in the hope that it will be useful, * but WITHOUT ANY WARRANTY; without even the implied warranty of * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the * GNU General Public License for more details. * * You should have received a copy of the GNU General Public License * along with this program; if not, write to the Free Software * Foundation, Inc., 51 Franklin Street, Fifth Floor, Boston, MA 02110-1301 USA * */ /* This file was automatically generated --- DO NOT EDIT */ /* Generated on Sun Nov 25 07:36:58 EST 2012 */ #include "codelet-dft.h" #ifdef HAVE_FMA /* Generated by: ../../../genfft/gen_notw_c.native -fma -reorder-insns -schedule-for-pipeline -simd -compact -variables 4 -pipeline-latency 8 -sign 1 -n 5 -name n1bv_5 -include n1b.h */ /* * This function contains 16 FP additions, 11 FP multiplications, * (or, 7 additions, 2 multiplications, 9 fused multiply/add), * 23 stack variables, 4 constants, and 10 memory accesses */ #include "n1b.h" static void n1bv_5(const R *ri, const R *ii, R *ro, R *io, stride is, stride os, INT v, INT ivs, INT ovs) { DVK(KP559016994, +0.559016994374947424102293417182819058860154590); DVK(KP250000000, +0.250000000000000000000000000000000000000000000); DVK(KP618033988, +0.618033988749894848204586834365638117720309180); DVK(KP951056516, +0.951056516295153572116439333379382143405698634); { INT i; const R *xi; R *xo; xi = ii; xo = io; for (i = v; i > 0; i = i - VL, xi = xi + (VL * ivs), xo = xo + (VL * ovs), MAKE_VOLATILE_STRIDE(10, is), MAKE_VOLATILE_STRIDE(10, os)) { V T1, T2, T3, T5, T6; T1 = LD(&(xi[0]), ivs, &(xi[0])); T2 = LD(&(xi[WS(is, 1)]), ivs, &(xi[WS(is, 1)])); T3 = LD(&(xi[WS(is, 4)]), ivs, &(xi[0])); T5 = LD(&(xi[WS(is, 2)]), ivs, &(xi[0])); T6 = LD(&(xi[WS(is, 3)]), ivs, &(xi[WS(is, 1)])); { V Tc, T4, Td, T7; Tc = VSUB(T2, T3); T4 = VADD(T2, T3); Td = VSUB(T5, T6); T7 = VADD(T5, T6); { V Tg, Te, Ta, T8, T9, Tf, Tb; Tg = VMUL(LDK(KP951056516), VFNMS(LDK(KP618033988), Tc, Td)); Te = VMUL(LDK(KP951056516), VFMA(LDK(KP618033988), Td, Tc)); Ta = VSUB(T4, T7); T8 = VADD(T4, T7); T9 = VFNMS(LDK(KP250000000), T8, T1); ST(&(xo[0]), VADD(T1, T8), ovs, &(xo[0])); Tf = VFNMS(LDK(KP559016994), Ta, T9); Tb = VFMA(LDK(KP559016994), Ta, T9); ST(&(xo[WS(os, 2)]), VFNMSI(Tg, Tf), ovs, &(xo[0])); ST(&(xo[WS(os, 3)]), VFMAI(Tg, Tf), ovs, &(xo[WS(os, 1)])); ST(&(xo[WS(os, 4)]), VFNMSI(Te, Tb), ovs, &(xo[0])); ST(&(xo[WS(os, 1)]), VFMAI(Te, Tb), ovs, &(xo[WS(os, 1)])); } } } } VLEAVE(); } static const kdft_desc desc = { 5, XSIMD_STRING("n1bv_5"), {7, 2, 9, 0}, &GENUS, 0, 0, 0, 0 }; void XSIMD(codelet_n1bv_5) (planner *p) { X(kdft_register) (p, n1bv_5, &desc); } #else /* HAVE_FMA */ /* Generated by: ../../../genfft/gen_notw_c.native -simd -compact -variables 4 -pipeline-latency 8 -sign 1 -n 5 -name n1bv_5 -include n1b.h */ /* * This function contains 16 FP additions, 6 FP multiplications, * (or, 13 additions, 3 multiplications, 3 fused multiply/add), * 18 stack variables, 4 constants, and 10 memory accesses */ #include "n1b.h" static void n1bv_5(const R *ri, const R *ii, R *ro, R *io, stride is, stride os, INT v, INT ivs, INT ovs) { DVK(KP250000000, +0.250000000000000000000000000000000000000000000); DVK(KP587785252, +0.587785252292473129168705954639072768597652438); DVK(KP951056516, +0.951056516295153572116439333379382143405698634); DVK(KP559016994, +0.559016994374947424102293417182819058860154590); { INT i; const R *xi; R *xo; xi = ii; xo = io; for (i = v; i > 0; i = i - VL, xi = xi + (VL * ivs), xo = xo + (VL * ovs), MAKE_VOLATILE_STRIDE(10, is), MAKE_VOLATILE_STRIDE(10, os)) { V Tb, T3, Tc, T6, Ta; Tb = LD(&(xi[0]), ivs, &(xi[0])); { V T1, T2, T8, T4, T5, T9; T1 = LD(&(xi[WS(is, 1)]), ivs, &(xi[WS(is, 1)])); T2 = LD(&(xi[WS(is, 4)]), ivs, &(xi[0])); T8 = VADD(T1, T2); T4 = LD(&(xi[WS(is, 2)]), ivs, &(xi[0])); T5 = LD(&(xi[WS(is, 3)]), ivs, &(xi[WS(is, 1)])); T9 = VADD(T4, T5); T3 = VSUB(T1, T2); Tc = VADD(T8, T9); T6 = VSUB(T4, T5); Ta = VMUL(LDK(KP559016994), VSUB(T8, T9)); } ST(&(xo[0]), VADD(Tb, Tc), ovs, &(xo[0])); { V T7, Tf, Te, Tg, Td; T7 = VBYI(VFMA(LDK(KP951056516), T3, VMUL(LDK(KP587785252), T6))); Tf = VBYI(VFNMS(LDK(KP951056516), T6, VMUL(LDK(KP587785252), T3))); Td = VFNMS(LDK(KP250000000), Tc, Tb); Te = VADD(Ta, Td); Tg = VSUB(Td, Ta); ST(&(xo[WS(os, 1)]), VADD(T7, Te), ovs, &(xo[WS(os, 1)])); ST(&(xo[WS(os, 3)]), VSUB(Tg, Tf), ovs, &(xo[WS(os, 1)])); ST(&(xo[WS(os, 4)]), VSUB(Te, T7), ovs, &(xo[0])); ST(&(xo[WS(os, 2)]), VADD(Tf, Tg), ovs, &(xo[0])); } } } VLEAVE(); } static const kdft_desc desc = { 5, XSIMD_STRING("n1bv_5"), {13, 3, 3, 0}, &GENUS, 0, 0, 0, 0 }; void XSIMD(codelet_n1bv_5) (planner *p) { X(kdft_register) (p, n1bv_5, &desc); } #endif /* HAVE_FMA */ fftw-3.3.3/dft/simd/common/n1bv_6.c0000644000175000017500000001264612054410352013654 00000000000000/* * Copyright (c) 2003, 2007-11 Matteo Frigo * Copyright (c) 2003, 2007-11 Massachusetts Institute of Technology * * This program is free software; you can redistribute it and/or modify * it under the terms of the GNU General Public License as published by * the Free Software Foundation; either version 2 of the License, or * (at your option) any later version. * * This program is distributed in the hope that it will be useful, * but WITHOUT ANY WARRANTY; without even the implied warranty of * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the * GNU General Public License for more details. * * You should have received a copy of the GNU General Public License * along with this program; if not, write to the Free Software * Foundation, Inc., 51 Franklin Street, Fifth Floor, Boston, MA 02110-1301 USA * */ /* This file was automatically generated --- DO NOT EDIT */ /* Generated on Sun Nov 25 07:36:58 EST 2012 */ #include "codelet-dft.h" #ifdef HAVE_FMA /* Generated by: ../../../genfft/gen_notw_c.native -fma -reorder-insns -schedule-for-pipeline -simd -compact -variables 4 -pipeline-latency 8 -sign 1 -n 6 -name n1bv_6 -include n1b.h */ /* * This function contains 18 FP additions, 8 FP multiplications, * (or, 12 additions, 2 multiplications, 6 fused multiply/add), * 23 stack variables, 2 constants, and 12 memory accesses */ #include "n1b.h" static void n1bv_6(const R *ri, const R *ii, R *ro, R *io, stride is, stride os, INT v, INT ivs, INT ovs) { DVK(KP500000000, +0.500000000000000000000000000000000000000000000); DVK(KP866025403, +0.866025403784438646763723170752936183471402627); { INT i; const R *xi; R *xo; xi = ii; xo = io; for (i = v; i > 0; i = i - VL, xi = xi + (VL * ivs), xo = xo + (VL * ovs), MAKE_VOLATILE_STRIDE(12, is), MAKE_VOLATILE_STRIDE(12, os)) { V T1, T2, T4, T5, T7, T8; T1 = LD(&(xi[0]), ivs, &(xi[0])); T2 = LD(&(xi[WS(is, 3)]), ivs, &(xi[WS(is, 1)])); T4 = LD(&(xi[WS(is, 2)]), ivs, &(xi[0])); T5 = LD(&(xi[WS(is, 5)]), ivs, &(xi[WS(is, 1)])); T7 = LD(&(xi[WS(is, 4)]), ivs, &(xi[0])); T8 = LD(&(xi[WS(is, 1)]), ivs, &(xi[WS(is, 1)])); { V T3, Td, T6, Te, T9, Tf; T3 = VSUB(T1, T2); Td = VADD(T1, T2); T6 = VSUB(T4, T5); Te = VADD(T4, T5); T9 = VSUB(T7, T8); Tf = VADD(T7, T8); { V Tg, Ti, Ta, Tc, Th, Tb; Tg = VADD(Te, Tf); Ti = VMUL(LDK(KP866025403), VSUB(Te, Tf)); Ta = VADD(T6, T9); Tc = VMUL(LDK(KP866025403), VSUB(T6, T9)); Th = VFNMS(LDK(KP500000000), Tg, Td); ST(&(xo[0]), VADD(Td, Tg), ovs, &(xo[0])); Tb = VFNMS(LDK(KP500000000), Ta, T3); ST(&(xo[WS(os, 3)]), VADD(T3, Ta), ovs, &(xo[WS(os, 1)])); ST(&(xo[WS(os, 4)]), VFMAI(Ti, Th), ovs, &(xo[0])); ST(&(xo[WS(os, 2)]), VFNMSI(Ti, Th), ovs, &(xo[0])); ST(&(xo[WS(os, 5)]), VFNMSI(Tc, Tb), ovs, &(xo[WS(os, 1)])); ST(&(xo[WS(os, 1)]), VFMAI(Tc, Tb), ovs, &(xo[WS(os, 1)])); } } } } VLEAVE(); } static const kdft_desc desc = { 6, XSIMD_STRING("n1bv_6"), {12, 2, 6, 0}, &GENUS, 0, 0, 0, 0 }; void XSIMD(codelet_n1bv_6) (planner *p) { X(kdft_register) (p, n1bv_6, &desc); } #else /* HAVE_FMA */ /* Generated by: ../../../genfft/gen_notw_c.native -simd -compact -variables 4 -pipeline-latency 8 -sign 1 -n 6 -name n1bv_6 -include n1b.h */ /* * This function contains 18 FP additions, 4 FP multiplications, * (or, 16 additions, 2 multiplications, 2 fused multiply/add), * 19 stack variables, 2 constants, and 12 memory accesses */ #include "n1b.h" static void n1bv_6(const R *ri, const R *ii, R *ro, R *io, stride is, stride os, INT v, INT ivs, INT ovs) { DVK(KP500000000, +0.500000000000000000000000000000000000000000000); DVK(KP866025403, +0.866025403784438646763723170752936183471402627); { INT i; const R *xi; R *xo; xi = ii; xo = io; for (i = v; i > 0; i = i - VL, xi = xi + (VL * ivs), xo = xo + (VL * ovs), MAKE_VOLATILE_STRIDE(12, is), MAKE_VOLATILE_STRIDE(12, os)) { V Ta, Td, T3, Te, T6, Tf, Tb, Tg, T8, T9; T8 = LD(&(xi[0]), ivs, &(xi[0])); T9 = LD(&(xi[WS(is, 3)]), ivs, &(xi[WS(is, 1)])); Ta = VSUB(T8, T9); Td = VADD(T8, T9); { V T1, T2, T4, T5; T1 = LD(&(xi[WS(is, 2)]), ivs, &(xi[0])); T2 = LD(&(xi[WS(is, 5)]), ivs, &(xi[WS(is, 1)])); T3 = VSUB(T1, T2); Te = VADD(T1, T2); T4 = LD(&(xi[WS(is, 4)]), ivs, &(xi[0])); T5 = LD(&(xi[WS(is, 1)]), ivs, &(xi[WS(is, 1)])); T6 = VSUB(T4, T5); Tf = VADD(T4, T5); } Tb = VADD(T3, T6); Tg = VADD(Te, Tf); ST(&(xo[WS(os, 3)]), VADD(Ta, Tb), ovs, &(xo[WS(os, 1)])); ST(&(xo[0]), VADD(Td, Tg), ovs, &(xo[0])); { V T7, Tc, Th, Ti; T7 = VBYI(VMUL(LDK(KP866025403), VSUB(T3, T6))); Tc = VFNMS(LDK(KP500000000), Tb, Ta); ST(&(xo[WS(os, 1)]), VADD(T7, Tc), ovs, &(xo[WS(os, 1)])); ST(&(xo[WS(os, 5)]), VSUB(Tc, T7), ovs, &(xo[WS(os, 1)])); Th = VFNMS(LDK(KP500000000), Tg, Td); Ti = VBYI(VMUL(LDK(KP866025403), VSUB(Te, Tf))); ST(&(xo[WS(os, 2)]), VSUB(Th, Ti), ovs, &(xo[0])); ST(&(xo[WS(os, 4)]), VADD(Ti, Th), ovs, &(xo[0])); } } } VLEAVE(); } static const kdft_desc desc = { 6, XSIMD_STRING("n1bv_6"), {16, 2, 2, 0}, &GENUS, 0, 0, 0, 0 }; void XSIMD(codelet_n1bv_6) (planner *p) { X(kdft_register) (p, n1bv_6, &desc); } #endif /* HAVE_FMA */ fftw-3.3.3/dft/simd/common/n1bv_7.c0000644000175000017500000001631612054410352013653 00000000000000/* * Copyright (c) 2003, 2007-11 Matteo Frigo * Copyright (c) 2003, 2007-11 Massachusetts Institute of Technology * * This program is free software; you can redistribute it and/or modify * it under the terms of the GNU General Public License as published by * the Free Software Foundation; either version 2 of the License, or * (at your option) any later version. * * This program is distributed in the hope that it will be useful, * but WITHOUT ANY WARRANTY; without even the implied warranty of * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the * GNU General Public License for more details. * * You should have received a copy of the GNU General Public License * along with this program; if not, write to the Free Software * Foundation, Inc., 51 Franklin Street, Fifth Floor, Boston, MA 02110-1301 USA * */ /* This file was automatically generated --- DO NOT EDIT */ /* Generated on Sun Nov 25 07:36:58 EST 2012 */ #include "codelet-dft.h" #ifdef HAVE_FMA /* Generated by: ../../../genfft/gen_notw_c.native -fma -reorder-insns -schedule-for-pipeline -simd -compact -variables 4 -pipeline-latency 8 -sign 1 -n 7 -name n1bv_7 -include n1b.h */ /* * This function contains 30 FP additions, 24 FP multiplications, * (or, 9 additions, 3 multiplications, 21 fused multiply/add), * 37 stack variables, 6 constants, and 14 memory accesses */ #include "n1b.h" static void n1bv_7(const R *ri, const R *ii, R *ro, R *io, stride is, stride os, INT v, INT ivs, INT ovs) { DVK(KP900968867, +0.900968867902419126236102319507445051165919162); DVK(KP692021471, +0.692021471630095869627814897002069140197260599); DVK(KP801937735, +0.801937735804838252472204639014890102331838324); DVK(KP974927912, +0.974927912181823607018131682993931217232785801); DVK(KP356895867, +0.356895867892209443894399510021300583399127187); DVK(KP554958132, +0.554958132087371191422194871006410481067288862); { INT i; const R *xi; R *xo; xi = ii; xo = io; for (i = v; i > 0; i = i - VL, xi = xi + (VL * ivs), xo = xo + (VL * ovs), MAKE_VOLATILE_STRIDE(14, is), MAKE_VOLATILE_STRIDE(14, os)) { V T1, T2, T3, T8, T9, T5, T6; T1 = LD(&(xi[0]), ivs, &(xi[0])); T2 = LD(&(xi[WS(is, 1)]), ivs, &(xi[WS(is, 1)])); T3 = LD(&(xi[WS(is, 6)]), ivs, &(xi[0])); T8 = LD(&(xi[WS(is, 3)]), ivs, &(xi[WS(is, 1)])); T9 = LD(&(xi[WS(is, 4)]), ivs, &(xi[0])); T5 = LD(&(xi[WS(is, 2)]), ivs, &(xi[0])); T6 = LD(&(xi[WS(is, 5)]), ivs, &(xi[WS(is, 1)])); { V Tg, T4, Te, Ta, Tf, T7; Tg = VSUB(T2, T3); T4 = VADD(T2, T3); Te = VSUB(T8, T9); Ta = VADD(T8, T9); Tf = VSUB(T5, T6); T7 = VADD(T5, T6); { V Tr, Tj, Tm, Th, To, Tb; Tr = VFMA(LDK(KP554958132), Te, Tg); Tj = VFNMS(LDK(KP356895867), T4, Ta); Tm = VFMA(LDK(KP554958132), Tf, Te); Th = VFNMS(LDK(KP554958132), Tg, Tf); ST(&(xo[0]), VADD(T1, VADD(T4, VADD(T7, Ta))), ovs, &(xo[0])); To = VFNMS(LDK(KP356895867), T7, T4); Tb = VFNMS(LDK(KP356895867), Ta, T7); { V Ts, Tk, Tn, Ti; Ts = VMUL(LDK(KP974927912), VFMA(LDK(KP801937735), Tr, Tf)); Tk = VFNMS(LDK(KP692021471), Tj, T7); Tn = VMUL(LDK(KP974927912), VFNMS(LDK(KP801937735), Tm, Tg)); Ti = VMUL(LDK(KP974927912), VFNMS(LDK(KP801937735), Th, Te)); { V Tp, Tc, Tl, Tq, Td; Tp = VFNMS(LDK(KP692021471), To, Ta); Tc = VFNMS(LDK(KP692021471), Tb, T4); Tl = VFNMS(LDK(KP900968867), Tk, T1); Tq = VFNMS(LDK(KP900968867), Tp, T1); Td = VFNMS(LDK(KP900968867), Tc, T1); ST(&(xo[WS(os, 5)]), VFNMSI(Tn, Tl), ovs, &(xo[WS(os, 1)])); ST(&(xo[WS(os, 2)]), VFMAI(Tn, Tl), ovs, &(xo[0])); ST(&(xo[WS(os, 6)]), VFNMSI(Ts, Tq), ovs, &(xo[0])); ST(&(xo[WS(os, 1)]), VFMAI(Ts, Tq), ovs, &(xo[WS(os, 1)])); ST(&(xo[WS(os, 4)]), VFNMSI(Ti, Td), ovs, &(xo[0])); ST(&(xo[WS(os, 3)]), VFMAI(Ti, Td), ovs, &(xo[WS(os, 1)])); } } } } } } VLEAVE(); } static const kdft_desc desc = { 7, XSIMD_STRING("n1bv_7"), {9, 3, 21, 0}, &GENUS, 0, 0, 0, 0 }; void XSIMD(codelet_n1bv_7) (planner *p) { X(kdft_register) (p, n1bv_7, &desc); } #else /* HAVE_FMA */ /* Generated by: ../../../genfft/gen_notw_c.native -simd -compact -variables 4 -pipeline-latency 8 -sign 1 -n 7 -name n1bv_7 -include n1b.h */ /* * This function contains 30 FP additions, 18 FP multiplications, * (or, 18 additions, 6 multiplications, 12 fused multiply/add), * 24 stack variables, 6 constants, and 14 memory accesses */ #include "n1b.h" static void n1bv_7(const R *ri, const R *ii, R *ro, R *io, stride is, stride os, INT v, INT ivs, INT ovs) { DVK(KP222520933, +0.222520933956314404288902564496794759466355569); DVK(KP900968867, +0.900968867902419126236102319507445051165919162); DVK(KP623489801, +0.623489801858733530525004884004239810632274731); DVK(KP433883739, +0.433883739117558120475768332848358754609990728); DVK(KP781831482, +0.781831482468029808708444526674057750232334519); DVK(KP974927912, +0.974927912181823607018131682993931217232785801); { INT i; const R *xi; R *xo; xi = ii; xo = io; for (i = v; i > 0; i = i - VL, xi = xi + (VL * ivs), xo = xo + (VL * ovs), MAKE_VOLATILE_STRIDE(14, is), MAKE_VOLATILE_STRIDE(14, os)) { V Tb, T9, Tc, T3, Te, T6, Td, T7, T8, Ti, Tj; Tb = LD(&(xi[0]), ivs, &(xi[0])); T7 = LD(&(xi[WS(is, 2)]), ivs, &(xi[0])); T8 = LD(&(xi[WS(is, 5)]), ivs, &(xi[WS(is, 1)])); T9 = VSUB(T7, T8); Tc = VADD(T7, T8); { V T1, T2, T4, T5; T1 = LD(&(xi[WS(is, 1)]), ivs, &(xi[WS(is, 1)])); T2 = LD(&(xi[WS(is, 6)]), ivs, &(xi[0])); T3 = VSUB(T1, T2); Te = VADD(T1, T2); T4 = LD(&(xi[WS(is, 3)]), ivs, &(xi[WS(is, 1)])); T5 = LD(&(xi[WS(is, 4)]), ivs, &(xi[0])); T6 = VSUB(T4, T5); Td = VADD(T4, T5); } ST(&(xo[0]), VADD(Tb, VADD(Te, VADD(Tc, Td))), ovs, &(xo[0])); Ti = VBYI(VFNMS(LDK(KP781831482), T6, VFNMS(LDK(KP433883739), T9, VMUL(LDK(KP974927912), T3)))); Tj = VFMA(LDK(KP623489801), Td, VFNMS(LDK(KP900968867), Tc, VFNMS(LDK(KP222520933), Te, Tb))); ST(&(xo[WS(os, 2)]), VADD(Ti, Tj), ovs, &(xo[0])); ST(&(xo[WS(os, 5)]), VSUB(Tj, Ti), ovs, &(xo[WS(os, 1)])); { V Ta, Tf, Tg, Th; Ta = VBYI(VFMA(LDK(KP433883739), T3, VFNMS(LDK(KP781831482), T9, VMUL(LDK(KP974927912), T6)))); Tf = VFMA(LDK(KP623489801), Tc, VFNMS(LDK(KP222520933), Td, VFNMS(LDK(KP900968867), Te, Tb))); ST(&(xo[WS(os, 3)]), VADD(Ta, Tf), ovs, &(xo[WS(os, 1)])); ST(&(xo[WS(os, 4)]), VSUB(Tf, Ta), ovs, &(xo[0])); Tg = VBYI(VFMA(LDK(KP781831482), T3, VFMA(LDK(KP974927912), T9, VMUL(LDK(KP433883739), T6)))); Th = VFMA(LDK(KP623489801), Te, VFNMS(LDK(KP900968867), Td, VFNMS(LDK(KP222520933), Tc, Tb))); ST(&(xo[WS(os, 1)]), VADD(Tg, Th), ovs, &(xo[WS(os, 1)])); ST(&(xo[WS(os, 6)]), VSUB(Th, Tg), ovs, &(xo[0])); } } } VLEAVE(); } static const kdft_desc desc = { 7, XSIMD_STRING("n1bv_7"), {18, 6, 12, 0}, &GENUS, 0, 0, 0, 0 }; void XSIMD(codelet_n1bv_7) (planner *p) { X(kdft_register) (p, n1bv_7, &desc); } #endif /* HAVE_FMA */ fftw-3.3.3/dft/simd/common/n1bv_8.c0000644000175000017500000001417312054410353013654 00000000000000/* * Copyright (c) 2003, 2007-11 Matteo Frigo * Copyright (c) 2003, 2007-11 Massachusetts Institute of Technology * * This program is free software; you can redistribute it and/or modify * it under the terms of the GNU General Public License as published by * the Free Software Foundation; either version 2 of the License, or * (at your option) any later version. * * This program is distributed in the hope that it will be useful, * but WITHOUT ANY WARRANTY; without even the implied warranty of * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the * GNU General Public License for more details. * * You should have received a copy of the GNU General Public License * along with this program; if not, write to the Free Software * Foundation, Inc., 51 Franklin Street, Fifth Floor, Boston, MA 02110-1301 USA * */ /* This file was automatically generated --- DO NOT EDIT */ /* Generated on Sun Nov 25 07:36:59 EST 2012 */ #include "codelet-dft.h" #ifdef HAVE_FMA /* Generated by: ../../../genfft/gen_notw_c.native -fma -reorder-insns -schedule-for-pipeline -simd -compact -variables 4 -pipeline-latency 8 -sign 1 -n 8 -name n1bv_8 -include n1b.h */ /* * This function contains 26 FP additions, 10 FP multiplications, * (or, 16 additions, 0 multiplications, 10 fused multiply/add), * 30 stack variables, 1 constants, and 16 memory accesses */ #include "n1b.h" static void n1bv_8(const R *ri, const R *ii, R *ro, R *io, stride is, stride os, INT v, INT ivs, INT ovs) { DVK(KP707106781, +0.707106781186547524400844362104849039284835938); { INT i; const R *xi; R *xo; xi = ii; xo = io; for (i = v; i > 0; i = i - VL, xi = xi + (VL * ivs), xo = xo + (VL * ovs), MAKE_VOLATILE_STRIDE(16, is), MAKE_VOLATILE_STRIDE(16, os)) { V T1, T2, Tc, Td, T4, T5, T7, T8; T1 = LD(&(xi[0]), ivs, &(xi[0])); T2 = LD(&(xi[WS(is, 4)]), ivs, &(xi[0])); Tc = LD(&(xi[WS(is, 2)]), ivs, &(xi[0])); Td = LD(&(xi[WS(is, 6)]), ivs, &(xi[0])); T4 = LD(&(xi[WS(is, 1)]), ivs, &(xi[WS(is, 1)])); T5 = LD(&(xi[WS(is, 5)]), ivs, &(xi[WS(is, 1)])); T7 = LD(&(xi[WS(is, 7)]), ivs, &(xi[WS(is, 1)])); T8 = LD(&(xi[WS(is, 3)]), ivs, &(xi[WS(is, 1)])); { V T3, Tj, Te, Tk, T6, Tm, T9, Tn, Tp, Tl; T3 = VSUB(T1, T2); Tj = VADD(T1, T2); Te = VSUB(Tc, Td); Tk = VADD(Tc, Td); T6 = VSUB(T4, T5); Tm = VADD(T4, T5); T9 = VSUB(T7, T8); Tn = VADD(T7, T8); Tp = VADD(Tj, Tk); Tl = VSUB(Tj, Tk); { V Tq, To, Ta, Tf; Tq = VADD(Tm, Tn); To = VSUB(Tm, Tn); Ta = VADD(T6, T9); Tf = VSUB(T6, T9); { V Tg, Ti, Tb, Th; ST(&(xo[WS(os, 2)]), VFMAI(To, Tl), ovs, &(xo[0])); ST(&(xo[WS(os, 6)]), VFNMSI(To, Tl), ovs, &(xo[0])); ST(&(xo[0]), VADD(Tp, Tq), ovs, &(xo[0])); ST(&(xo[WS(os, 4)]), VSUB(Tp, Tq), ovs, &(xo[0])); Tg = VFNMS(LDK(KP707106781), Tf, Te); Ti = VFMA(LDK(KP707106781), Tf, Te); Tb = VFNMS(LDK(KP707106781), Ta, T3); Th = VFMA(LDK(KP707106781), Ta, T3); ST(&(xo[WS(os, 7)]), VFNMSI(Ti, Th), ovs, &(xo[WS(os, 1)])); ST(&(xo[WS(os, 1)]), VFMAI(Ti, Th), ovs, &(xo[WS(os, 1)])); ST(&(xo[WS(os, 5)]), VFMAI(Tg, Tb), ovs, &(xo[WS(os, 1)])); ST(&(xo[WS(os, 3)]), VFNMSI(Tg, Tb), ovs, &(xo[WS(os, 1)])); } } } } } VLEAVE(); } static const kdft_desc desc = { 8, XSIMD_STRING("n1bv_8"), {16, 0, 10, 0}, &GENUS, 0, 0, 0, 0 }; void XSIMD(codelet_n1bv_8) (planner *p) { X(kdft_register) (p, n1bv_8, &desc); } #else /* HAVE_FMA */ /* Generated by: ../../../genfft/gen_notw_c.native -simd -compact -variables 4 -pipeline-latency 8 -sign 1 -n 8 -name n1bv_8 -include n1b.h */ /* * This function contains 26 FP additions, 2 FP multiplications, * (or, 26 additions, 2 multiplications, 0 fused multiply/add), * 22 stack variables, 1 constants, and 16 memory accesses */ #include "n1b.h" static void n1bv_8(const R *ri, const R *ii, R *ro, R *io, stride is, stride os, INT v, INT ivs, INT ovs) { DVK(KP707106781, +0.707106781186547524400844362104849039284835938); { INT i; const R *xi; R *xo; xi = ii; xo = io; for (i = v; i > 0; i = i - VL, xi = xi + (VL * ivs), xo = xo + (VL * ovs), MAKE_VOLATILE_STRIDE(16, is), MAKE_VOLATILE_STRIDE(16, os)) { V Ta, Tk, Te, Tj, T7, Tn, Tf, Tm; { V T8, T9, Tc, Td; T8 = LD(&(xi[WS(is, 2)]), ivs, &(xi[0])); T9 = LD(&(xi[WS(is, 6)]), ivs, &(xi[0])); Ta = VSUB(T8, T9); Tk = VADD(T8, T9); Tc = LD(&(xi[0]), ivs, &(xi[0])); Td = LD(&(xi[WS(is, 4)]), ivs, &(xi[0])); Te = VSUB(Tc, Td); Tj = VADD(Tc, Td); { V T1, T2, T3, T4, T5, T6; T1 = LD(&(xi[WS(is, 1)]), ivs, &(xi[WS(is, 1)])); T2 = LD(&(xi[WS(is, 5)]), ivs, &(xi[WS(is, 1)])); T3 = VSUB(T1, T2); T4 = LD(&(xi[WS(is, 7)]), ivs, &(xi[WS(is, 1)])); T5 = LD(&(xi[WS(is, 3)]), ivs, &(xi[WS(is, 1)])); T6 = VSUB(T4, T5); T7 = VMUL(LDK(KP707106781), VSUB(T3, T6)); Tn = VADD(T4, T5); Tf = VMUL(LDK(KP707106781), VADD(T3, T6)); Tm = VADD(T1, T2); } } { V Tb, Tg, Tp, Tq; Tb = VBYI(VSUB(T7, Ta)); Tg = VSUB(Te, Tf); ST(&(xo[WS(os, 3)]), VADD(Tb, Tg), ovs, &(xo[WS(os, 1)])); ST(&(xo[WS(os, 5)]), VSUB(Tg, Tb), ovs, &(xo[WS(os, 1)])); Tp = VADD(Tj, Tk); Tq = VADD(Tm, Tn); ST(&(xo[WS(os, 4)]), VSUB(Tp, Tq), ovs, &(xo[0])); ST(&(xo[0]), VADD(Tp, Tq), ovs, &(xo[0])); } { V Th, Ti, Tl, To; Th = VBYI(VADD(Ta, T7)); Ti = VADD(Te, Tf); ST(&(xo[WS(os, 1)]), VADD(Th, Ti), ovs, &(xo[WS(os, 1)])); ST(&(xo[WS(os, 7)]), VSUB(Ti, Th), ovs, &(xo[WS(os, 1)])); Tl = VSUB(Tj, Tk); To = VBYI(VSUB(Tm, Tn)); ST(&(xo[WS(os, 6)]), VSUB(Tl, To), ovs, &(xo[0])); ST(&(xo[WS(os, 2)]), VADD(Tl, To), ovs, &(xo[0])); } } } VLEAVE(); } static const kdft_desc desc = { 8, XSIMD_STRING("n1bv_8"), {26, 2, 0, 0}, &GENUS, 0, 0, 0, 0 }; void XSIMD(codelet_n1bv_8) (planner *p) { X(kdft_register) (p, n1bv_8, &desc); } #endif /* HAVE_FMA */ fftw-3.3.3/dft/simd/common/n1bv_9.c0000644000175000017500000002475612054410353013665 00000000000000/* * Copyright (c) 2003, 2007-11 Matteo Frigo * Copyright (c) 2003, 2007-11 Massachusetts Institute of Technology * * This program is free software; you can redistribute it and/or modify * it under the terms of the GNU General Public License as published by * the Free Software Foundation; either version 2 of the License, or * (at your option) any later version. * * This program is distributed in the hope that it will be useful, * but WITHOUT ANY WARRANTY; without even the implied warranty of * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the * GNU General Public License for more details. * * You should have received a copy of the GNU General Public License * along with this program; if not, write to the Free Software * Foundation, Inc., 51 Franklin Street, Fifth Floor, Boston, MA 02110-1301 USA * */ /* This file was automatically generated --- DO NOT EDIT */ /* Generated on Sun Nov 25 07:36:59 EST 2012 */ #include "codelet-dft.h" #ifdef HAVE_FMA /* Generated by: ../../../genfft/gen_notw_c.native -fma -reorder-insns -schedule-for-pipeline -simd -compact -variables 4 -pipeline-latency 8 -sign 1 -n 9 -name n1bv_9 -include n1b.h */ /* * This function contains 46 FP additions, 38 FP multiplications, * (or, 12 additions, 4 multiplications, 34 fused multiply/add), * 68 stack variables, 19 constants, and 18 memory accesses */ #include "n1b.h" static void n1bv_9(const R *ri, const R *ii, R *ro, R *io, stride is, stride os, INT v, INT ivs, INT ovs) { DVK(KP939692620, +0.939692620785908384054109277324731469936208134); DVK(KP907603734, +0.907603734547952313649323976213898122064543220); DVK(KP852868531, +0.852868531952443209628250963940074071936020296); DVK(KP666666666, +0.666666666666666666666666666666666666666666667); DVK(KP879385241, +0.879385241571816768108218554649462939872416269); DVK(KP984807753, +0.984807753012208059366743024589523013670643252); DVK(KP826351822, +0.826351822333069651148283373230685203999624323); DVK(KP347296355, +0.347296355333860697703433253538629592000751354); DVK(KP898197570, +0.898197570222573798468955502359086394667167570); DVK(KP673648177, +0.673648177666930348851716626769314796000375677); DVK(KP420276625, +0.420276625461206169731530603237061658838781920); DVK(KP866025403, +0.866025403784438646763723170752936183471402627); DVK(KP586256827, +0.586256827714544512072145703099641959914944179); DVK(KP968908795, +0.968908795874236621082202410917456709164223497); DVK(KP726681596, +0.726681596905677465811651808188092531873167623); DVK(KP439692620, +0.439692620785908384054109277324731469936208134); DVK(KP203604859, +0.203604859554852403062088995281827210665664861); DVK(KP152703644, +0.152703644666139302296566746461370407999248646); DVK(KP500000000, +0.500000000000000000000000000000000000000000000); { INT i; const R *xi; R *xo; xi = ii; xo = io; for (i = v; i > 0; i = i - VL, xi = xi + (VL * ivs), xo = xo + (VL * ovs), MAKE_VOLATILE_STRIDE(18, is), MAKE_VOLATILE_STRIDE(18, os)) { V T1, T2, T3, T6, Tf, T7, T8, Tb, Tc, Tp, T4; T1 = LD(&(xi[0]), ivs, &(xi[0])); T2 = LD(&(xi[WS(is, 3)]), ivs, &(xi[WS(is, 1)])); T3 = LD(&(xi[WS(is, 6)]), ivs, &(xi[0])); T6 = LD(&(xi[WS(is, 2)]), ivs, &(xi[0])); Tf = LD(&(xi[WS(is, 1)]), ivs, &(xi[WS(is, 1)])); T7 = LD(&(xi[WS(is, 5)]), ivs, &(xi[WS(is, 1)])); T8 = LD(&(xi[WS(is, 8)]), ivs, &(xi[0])); Tb = LD(&(xi[WS(is, 4)]), ivs, &(xi[0])); Tc = LD(&(xi[WS(is, 7)]), ivs, &(xi[WS(is, 1)])); Tp = VSUB(T2, T3); T4 = VADD(T2, T3); { V Te, T9, Tg, Td, TF, T5; Te = VSUB(T8, T7); T9 = VADD(T7, T8); Tg = VADD(Tb, Tc); Td = VSUB(Tb, Tc); TF = VADD(T1, T4); T5 = VFNMS(LDK(KP500000000), T4, T1); { V Ta, TH, Th, TG; Ta = VFNMS(LDK(KP500000000), T9, T6); TH = VADD(T6, T9); Th = VFNMS(LDK(KP500000000), Tg, Tf); TG = VADD(Tf, Tg); { V Tr, Tu, Tm, Tv, Ts, Ti, TI, TK; Tr = VFNMS(LDK(KP152703644), Te, Ta); Tu = VFMA(LDK(KP203604859), Ta, Te); Tm = VFNMS(LDK(KP439692620), Td, Ta); Tv = VFNMS(LDK(KP726681596), Td, Th); Ts = VFMA(LDK(KP968908795), Th, Td); Ti = VFNMS(LDK(KP586256827), Th, Te); TI = VADD(TG, TH); TK = VMUL(LDK(KP866025403), VSUB(TG, TH)); { V Tt, TA, Tw, Tz, Tj, TJ, To, TE, Tn; Tn = VFNMS(LDK(KP420276625), Tm, Te); Tt = VFNMS(LDK(KP673648177), Ts, Tr); TA = VFMA(LDK(KP673648177), Ts, Tr); Tw = VFMA(LDK(KP898197570), Tv, Tu); Tz = VFNMS(LDK(KP898197570), Tv, Tu); Tj = VFNMS(LDK(KP347296355), Ti, Td); ST(&(xo[0]), VADD(TI, TF), ovs, &(xo[0])); TJ = VFNMS(LDK(KP500000000), TI, TF); To = VFNMS(LDK(KP826351822), Tn, Th); TE = VMUL(LDK(KP984807753), VFMA(LDK(KP879385241), Tp, TA)); { V TB, TD, Tx, Tk, Tq, TC, Ty, Tl; TB = VFMA(LDK(KP666666666), TA, Tz); TD = VFMA(LDK(KP852868531), Tw, T5); Tx = VFNMS(LDK(KP500000000), Tw, Tt); Tk = VFNMS(LDK(KP907603734), Tj, Ta); ST(&(xo[WS(os, 6)]), VFNMSI(TK, TJ), ovs, &(xo[0])); ST(&(xo[WS(os, 3)]), VFMAI(TK, TJ), ovs, &(xo[WS(os, 1)])); Tq = VMUL(LDK(KP984807753), VFNMS(LDK(KP879385241), Tp, To)); TC = VMUL(LDK(KP866025403), VFNMS(LDK(KP852868531), TB, Tp)); ST(&(xo[WS(os, 8)]), VFNMSI(TE, TD), ovs, &(xo[0])); ST(&(xo[WS(os, 1)]), VFMAI(TE, TD), ovs, &(xo[WS(os, 1)])); Ty = VFMA(LDK(KP852868531), Tx, T5); Tl = VFNMS(LDK(KP939692620), Tk, T5); ST(&(xo[WS(os, 5)]), VFNMSI(TC, Ty), ovs, &(xo[WS(os, 1)])); ST(&(xo[WS(os, 4)]), VFMAI(TC, Ty), ovs, &(xo[0])); ST(&(xo[WS(os, 2)]), VFMAI(Tq, Tl), ovs, &(xo[0])); ST(&(xo[WS(os, 7)]), VFNMSI(Tq, Tl), ovs, &(xo[WS(os, 1)])); } } } } } } } VLEAVE(); } static const kdft_desc desc = { 9, XSIMD_STRING("n1bv_9"), {12, 4, 34, 0}, &GENUS, 0, 0, 0, 0 }; void XSIMD(codelet_n1bv_9) (planner *p) { X(kdft_register) (p, n1bv_9, &desc); } #else /* HAVE_FMA */ /* Generated by: ../../../genfft/gen_notw_c.native -simd -compact -variables 4 -pipeline-latency 8 -sign 1 -n 9 -name n1bv_9 -include n1b.h */ /* * This function contains 46 FP additions, 26 FP multiplications, * (or, 30 additions, 10 multiplications, 16 fused multiply/add), * 41 stack variables, 14 constants, and 18 memory accesses */ #include "n1b.h" static void n1bv_9(const R *ri, const R *ii, R *ro, R *io, stride is, stride os, INT v, INT ivs, INT ovs) { DVK(KP342020143, +0.342020143325668733044099614682259580763083368); DVK(KP813797681, +0.813797681349373692844693217248393223289101568); DVK(KP939692620, +0.939692620785908384054109277324731469936208134); DVK(KP296198132, +0.296198132726023843175338011893050938967728390); DVK(KP642787609, +0.642787609686539326322643409907263432907559884); DVK(KP663413948, +0.663413948168938396205421319635891297216863310); DVK(KP556670399, +0.556670399226419366452912952047023132968291906); DVK(KP766044443, +0.766044443118978035202392650555416673935832457); DVK(KP984807753, +0.984807753012208059366743024589523013670643252); DVK(KP150383733, +0.150383733180435296639271897612501926072238258); DVK(KP852868531, +0.852868531952443209628250963940074071936020296); DVK(KP173648177, +0.173648177666930348851716626769314796000375677); DVK(KP866025403, +0.866025403784438646763723170752936183471402627); DVK(KP500000000, +0.500000000000000000000000000000000000000000000); { INT i; const R *xi; R *xo; xi = ii; xo = io; for (i = v; i > 0; i = i - VL, xi = xi + (VL * ivs), xo = xo + (VL * ovs), MAKE_VOLATILE_STRIDE(18, is), MAKE_VOLATILE_STRIDE(18, os)) { V T5, Ty, Tm, Ti, Tw, Th, Tj, To, Tb, Tv, Ta, Tc, Tn; { V T1, T2, T3, T4; T1 = LD(&(xi[0]), ivs, &(xi[0])); T2 = LD(&(xi[WS(is, 3)]), ivs, &(xi[WS(is, 1)])); T3 = LD(&(xi[WS(is, 6)]), ivs, &(xi[0])); T4 = VADD(T2, T3); T5 = VFNMS(LDK(KP500000000), T4, T1); Ty = VADD(T1, T4); Tm = VMUL(LDK(KP866025403), VSUB(T2, T3)); } { V Td, Tg, Te, Tf; Td = LD(&(xi[WS(is, 2)]), ivs, &(xi[0])); Te = LD(&(xi[WS(is, 5)]), ivs, &(xi[WS(is, 1)])); Tf = LD(&(xi[WS(is, 8)]), ivs, &(xi[0])); Tg = VADD(Te, Tf); Ti = VSUB(Te, Tf); Tw = VADD(Td, Tg); Th = VFNMS(LDK(KP500000000), Tg, Td); Tj = VFNMS(LDK(KP852868531), Ti, VMUL(LDK(KP173648177), Th)); To = VFMA(LDK(KP150383733), Ti, VMUL(LDK(KP984807753), Th)); } { V T6, T9, T7, T8; T6 = LD(&(xi[WS(is, 1)]), ivs, &(xi[WS(is, 1)])); T7 = LD(&(xi[WS(is, 4)]), ivs, &(xi[0])); T8 = LD(&(xi[WS(is, 7)]), ivs, &(xi[WS(is, 1)])); T9 = VADD(T7, T8); Tb = VSUB(T7, T8); Tv = VADD(T6, T9); Ta = VFNMS(LDK(KP500000000), T9, T6); Tc = VFNMS(LDK(KP556670399), Tb, VMUL(LDK(KP766044443), Ta)); Tn = VFMA(LDK(KP663413948), Tb, VMUL(LDK(KP642787609), Ta)); } { V Tx, Tz, TA, Tt, Tu; Tx = VBYI(VMUL(LDK(KP866025403), VSUB(Tv, Tw))); Tz = VADD(Tv, Tw); TA = VFNMS(LDK(KP500000000), Tz, Ty); ST(&(xo[WS(os, 3)]), VADD(Tx, TA), ovs, &(xo[WS(os, 1)])); ST(&(xo[0]), VADD(Ty, Tz), ovs, &(xo[0])); ST(&(xo[WS(os, 6)]), VSUB(TA, Tx), ovs, &(xo[0])); Tt = VFMA(LDK(KP852868531), Tb, VFMA(LDK(KP173648177), Ta, VFMA(LDK(KP296198132), Ti, VFNMS(LDK(KP939692620), Th, T5)))); Tu = VBYI(VSUB(VFMA(LDK(KP984807753), Ta, VFMA(LDK(KP813797681), Ti, VFNMS(LDK(KP150383733), Tb, VMUL(LDK(KP342020143), Th)))), Tm)); ST(&(xo[WS(os, 7)]), VSUB(Tt, Tu), ovs, &(xo[WS(os, 1)])); ST(&(xo[WS(os, 2)]), VADD(Tt, Tu), ovs, &(xo[0])); { V Tl, Ts, Tq, Tr, Tk, Tp; Tk = VADD(Tc, Tj); Tl = VADD(T5, Tk); Ts = VFMA(LDK(KP866025403), VSUB(To, Tn), VFNMS(LDK(KP500000000), Tk, T5)); Tp = VADD(Tn, To); Tq = VBYI(VADD(Tm, Tp)); Tr = VBYI(VADD(Tm, VFNMS(LDK(KP500000000), Tp, VMUL(LDK(KP866025403), VSUB(Tc, Tj))))); ST(&(xo[WS(os, 8)]), VSUB(Tl, Tq), ovs, &(xo[0])); ST(&(xo[WS(os, 5)]), VSUB(Ts, Tr), ovs, &(xo[WS(os, 1)])); ST(&(xo[WS(os, 1)]), VADD(Tl, Tq), ovs, &(xo[WS(os, 1)])); ST(&(xo[WS(os, 4)]), VADD(Tr, Ts), ovs, &(xo[0])); } } } } VLEAVE(); } static const kdft_desc desc = { 9, XSIMD_STRING("n1bv_9"), {30, 10, 16, 0}, &GENUS, 0, 0, 0, 0 }; void XSIMD(codelet_n1bv_9) (planner *p) { X(kdft_register) (p, n1bv_9, &desc); } #endif /* HAVE_FMA */ fftw-3.3.3/dft/simd/common/n1bv_10.c0000644000175000017500000002047212054410354013725 00000000000000/* * Copyright (c) 2003, 2007-11 Matteo Frigo * Copyright (c) 2003, 2007-11 Massachusetts Institute of Technology * * This program is free software; you can redistribute it and/or modify * it under the terms of the GNU General Public License as published by * the Free Software Foundation; either version 2 of the License, or * (at your option) any later version. * * This program is distributed in the hope that it will be useful, * but WITHOUT ANY WARRANTY; without even the implied warranty of * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the * GNU General Public License for more details. * * You should have received a copy of the GNU General Public License * along with this program; if not, write to the Free Software * Foundation, Inc., 51 Franklin Street, Fifth Floor, Boston, MA 02110-1301 USA * */ /* This file was automatically generated --- DO NOT EDIT */ /* Generated on Sun Nov 25 07:36:59 EST 2012 */ #include "codelet-dft.h" #ifdef HAVE_FMA /* Generated by: ../../../genfft/gen_notw_c.native -fma -reorder-insns -schedule-for-pipeline -simd -compact -variables 4 -pipeline-latency 8 -sign 1 -n 10 -name n1bv_10 -include n1b.h */ /* * This function contains 42 FP additions, 22 FP multiplications, * (or, 24 additions, 4 multiplications, 18 fused multiply/add), * 43 stack variables, 4 constants, and 20 memory accesses */ #include "n1b.h" static void n1bv_10(const R *ri, const R *ii, R *ro, R *io, stride is, stride os, INT v, INT ivs, INT ovs) { DVK(KP559016994, +0.559016994374947424102293417182819058860154590); DVK(KP250000000, +0.250000000000000000000000000000000000000000000); DVK(KP618033988, +0.618033988749894848204586834365638117720309180); DVK(KP951056516, +0.951056516295153572116439333379382143405698634); { INT i; const R *xi; R *xo; xi = ii; xo = io; for (i = v; i > 0; i = i - VL, xi = xi + (VL * ivs), xo = xo + (VL * ovs), MAKE_VOLATILE_STRIDE(20, is), MAKE_VOLATILE_STRIDE(20, os)) { V Tb, Tr, T3, Ts, T6, Tw, Tg, Tt, T9, Tc, T1, T2; T1 = LD(&(xi[0]), ivs, &(xi[0])); T2 = LD(&(xi[WS(is, 5)]), ivs, &(xi[WS(is, 1)])); { V T4, T5, Te, Tf, T7, T8; T4 = LD(&(xi[WS(is, 2)]), ivs, &(xi[0])); T5 = LD(&(xi[WS(is, 7)]), ivs, &(xi[WS(is, 1)])); Te = LD(&(xi[WS(is, 6)]), ivs, &(xi[0])); Tf = LD(&(xi[WS(is, 1)]), ivs, &(xi[WS(is, 1)])); T7 = LD(&(xi[WS(is, 8)]), ivs, &(xi[0])); T8 = LD(&(xi[WS(is, 3)]), ivs, &(xi[WS(is, 1)])); Tb = LD(&(xi[WS(is, 4)]), ivs, &(xi[0])); Tr = VADD(T1, T2); T3 = VSUB(T1, T2); Ts = VADD(T4, T5); T6 = VSUB(T4, T5); Tw = VADD(Te, Tf); Tg = VSUB(Te, Tf); Tt = VADD(T7, T8); T9 = VSUB(T7, T8); Tc = LD(&(xi[WS(is, 9)]), ivs, &(xi[WS(is, 1)])); } { V TD, Tu, Tm, Ta, Td, Tv; TD = VSUB(Ts, Tt); Tu = VADD(Ts, Tt); Tm = VSUB(T6, T9); Ta = VADD(T6, T9); Td = VSUB(Tb, Tc); Tv = VADD(Tb, Tc); { V TC, Tx, Tn, Th; TC = VSUB(Tv, Tw); Tx = VADD(Tv, Tw); Tn = VSUB(Td, Tg); Th = VADD(Td, Tg); { V Ty, TA, TE, TG, Ti, Tk, To, Tq, Tz, Tj; Ty = VADD(Tu, Tx); TA = VSUB(Tu, Tx); TE = VMUL(LDK(KP951056516), VFNMS(LDK(KP618033988), TD, TC)); TG = VMUL(LDK(KP951056516), VFMA(LDK(KP618033988), TC, TD)); Ti = VADD(Ta, Th); Tk = VSUB(Ta, Th); To = VMUL(LDK(KP951056516), VFMA(LDK(KP618033988), Tn, Tm)); Tq = VMUL(LDK(KP951056516), VFNMS(LDK(KP618033988), Tm, Tn)); Tz = VFNMS(LDK(KP250000000), Ty, Tr); ST(&(xo[0]), VADD(Tr, Ty), ovs, &(xo[0])); Tj = VFNMS(LDK(KP250000000), Ti, T3); ST(&(xo[WS(os, 5)]), VADD(T3, Ti), ovs, &(xo[WS(os, 1)])); { V TB, TF, Tl, Tp; TB = VFNMS(LDK(KP559016994), TA, Tz); TF = VFMA(LDK(KP559016994), TA, Tz); Tl = VFMA(LDK(KP559016994), Tk, Tj); Tp = VFNMS(LDK(KP559016994), Tk, Tj); ST(&(xo[WS(os, 4)]), VFNMSI(TG, TF), ovs, &(xo[0])); ST(&(xo[WS(os, 6)]), VFMAI(TG, TF), ovs, &(xo[0])); ST(&(xo[WS(os, 8)]), VFMAI(TE, TB), ovs, &(xo[0])); ST(&(xo[WS(os, 2)]), VFNMSI(TE, TB), ovs, &(xo[0])); ST(&(xo[WS(os, 3)]), VFMAI(Tq, Tp), ovs, &(xo[WS(os, 1)])); ST(&(xo[WS(os, 7)]), VFNMSI(Tq, Tp), ovs, &(xo[WS(os, 1)])); ST(&(xo[WS(os, 9)]), VFNMSI(To, Tl), ovs, &(xo[WS(os, 1)])); ST(&(xo[WS(os, 1)]), VFMAI(To, Tl), ovs, &(xo[WS(os, 1)])); } } } } } } VLEAVE(); } static const kdft_desc desc = { 10, XSIMD_STRING("n1bv_10"), {24, 4, 18, 0}, &GENUS, 0, 0, 0, 0 }; void XSIMD(codelet_n1bv_10) (planner *p) { X(kdft_register) (p, n1bv_10, &desc); } #else /* HAVE_FMA */ /* Generated by: ../../../genfft/gen_notw_c.native -simd -compact -variables 4 -pipeline-latency 8 -sign 1 -n 10 -name n1bv_10 -include n1b.h */ /* * This function contains 42 FP additions, 12 FP multiplications, * (or, 36 additions, 6 multiplications, 6 fused multiply/add), * 33 stack variables, 4 constants, and 20 memory accesses */ #include "n1b.h" static void n1bv_10(const R *ri, const R *ii, R *ro, R *io, stride is, stride os, INT v, INT ivs, INT ovs) { DVK(KP250000000, +0.250000000000000000000000000000000000000000000); DVK(KP559016994, +0.559016994374947424102293417182819058860154590); DVK(KP587785252, +0.587785252292473129168705954639072768597652438); DVK(KP951056516, +0.951056516295153572116439333379382143405698634); { INT i; const R *xi; R *xo; xi = ii; xo = io; for (i = v; i > 0; i = i - VL, xi = xi + (VL * ivs), xo = xo + (VL * ovs), MAKE_VOLATILE_STRIDE(20, is), MAKE_VOLATILE_STRIDE(20, os)) { V Tl, Ty, T7, Te, Tw, Tt, Tz, TA, TB, Tg, Th, Tm, Tj, Tk; Tj = LD(&(xi[0]), ivs, &(xi[0])); Tk = LD(&(xi[WS(is, 5)]), ivs, &(xi[WS(is, 1)])); Tl = VSUB(Tj, Tk); Ty = VADD(Tj, Tk); { V T3, Tr, Td, Tv, T6, Ts, Ta, Tu; { V T1, T2, Tb, Tc; T1 = LD(&(xi[WS(is, 2)]), ivs, &(xi[0])); T2 = LD(&(xi[WS(is, 7)]), ivs, &(xi[WS(is, 1)])); T3 = VSUB(T1, T2); Tr = VADD(T1, T2); Tb = LD(&(xi[WS(is, 6)]), ivs, &(xi[0])); Tc = LD(&(xi[WS(is, 1)]), ivs, &(xi[WS(is, 1)])); Td = VSUB(Tb, Tc); Tv = VADD(Tb, Tc); } { V T4, T5, T8, T9; T4 = LD(&(xi[WS(is, 8)]), ivs, &(xi[0])); T5 = LD(&(xi[WS(is, 3)]), ivs, &(xi[WS(is, 1)])); T6 = VSUB(T4, T5); Ts = VADD(T4, T5); T8 = LD(&(xi[WS(is, 4)]), ivs, &(xi[0])); T9 = LD(&(xi[WS(is, 9)]), ivs, &(xi[WS(is, 1)])); Ta = VSUB(T8, T9); Tu = VADD(T8, T9); } T7 = VSUB(T3, T6); Te = VSUB(Ta, Td); Tw = VSUB(Tu, Tv); Tt = VSUB(Tr, Ts); Tz = VADD(Tr, Ts); TA = VADD(Tu, Tv); TB = VADD(Tz, TA); Tg = VADD(T3, T6); Th = VADD(Ta, Td); Tm = VADD(Tg, Th); } ST(&(xo[WS(os, 5)]), VADD(Tl, Tm), ovs, &(xo[WS(os, 1)])); ST(&(xo[0]), VADD(Ty, TB), ovs, &(xo[0])); { V Tf, Tq, To, Tp, Ti, Tn; Tf = VBYI(VFMA(LDK(KP951056516), T7, VMUL(LDK(KP587785252), Te))); Tq = VBYI(VFNMS(LDK(KP951056516), Te, VMUL(LDK(KP587785252), T7))); Ti = VMUL(LDK(KP559016994), VSUB(Tg, Th)); Tn = VFNMS(LDK(KP250000000), Tm, Tl); To = VADD(Ti, Tn); Tp = VSUB(Tn, Ti); ST(&(xo[WS(os, 1)]), VADD(Tf, To), ovs, &(xo[WS(os, 1)])); ST(&(xo[WS(os, 7)]), VADD(Tq, Tp), ovs, &(xo[WS(os, 1)])); ST(&(xo[WS(os, 9)]), VSUB(To, Tf), ovs, &(xo[WS(os, 1)])); ST(&(xo[WS(os, 3)]), VSUB(Tp, Tq), ovs, &(xo[WS(os, 1)])); } { V Tx, TG, TE, TF, TC, TD; Tx = VBYI(VFNMS(LDK(KP951056516), Tw, VMUL(LDK(KP587785252), Tt))); TG = VBYI(VFMA(LDK(KP951056516), Tt, VMUL(LDK(KP587785252), Tw))); TC = VFNMS(LDK(KP250000000), TB, Ty); TD = VMUL(LDK(KP559016994), VSUB(Tz, TA)); TE = VSUB(TC, TD); TF = VADD(TD, TC); ST(&(xo[WS(os, 2)]), VADD(Tx, TE), ovs, &(xo[0])); ST(&(xo[WS(os, 6)]), VADD(TG, TF), ovs, &(xo[0])); ST(&(xo[WS(os, 8)]), VSUB(TE, Tx), ovs, &(xo[0])); ST(&(xo[WS(os, 4)]), VSUB(TF, TG), ovs, &(xo[0])); } } } VLEAVE(); } static const kdft_desc desc = { 10, XSIMD_STRING("n1bv_10"), {36, 6, 6, 0}, &GENUS, 0, 0, 0, 0 }; void XSIMD(codelet_n1bv_10) (planner *p) { X(kdft_register) (p, n1bv_10, &desc); } #endif /* HAVE_FMA */ fftw-3.3.3/dft/simd/common/n1bv_11.c0000644000175000017500000002734112054410356013732 00000000000000/* * Copyright (c) 2003, 2007-11 Matteo Frigo * Copyright (c) 2003, 2007-11 Massachusetts Institute of Technology * * This program is free software; you can redistribute it and/or modify * it under the terms of the GNU General Public License as published by * the Free Software Foundation; either version 2 of the License, or * (at your option) any later version. * * This program is distributed in the hope that it will be useful, * but WITHOUT ANY WARRANTY; without even the implied warranty of * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the * GNU General Public License for more details. * * You should have received a copy of the GNU General Public License * along with this program; if not, write to the Free Software * Foundation, Inc., 51 Franklin Street, Fifth Floor, Boston, MA 02110-1301 USA * */ /* This file was automatically generated --- DO NOT EDIT */ /* Generated on Sun Nov 25 07:37:00 EST 2012 */ #include "codelet-dft.h" #ifdef HAVE_FMA /* Generated by: ../../../genfft/gen_notw_c.native -fma -reorder-insns -schedule-for-pipeline -simd -compact -variables 4 -pipeline-latency 8 -sign 1 -n 11 -name n1bv_11 -include n1b.h */ /* * This function contains 70 FP additions, 60 FP multiplications, * (or, 15 additions, 5 multiplications, 55 fused multiply/add), * 67 stack variables, 11 constants, and 22 memory accesses */ #include "n1b.h" static void n1bv_11(const R *ri, const R *ii, R *ro, R *io, stride is, stride os, INT v, INT ivs, INT ovs) { DVK(KP959492973, +0.959492973614497389890368057066327699062454848); DVK(KP876768831, +0.876768831002589333891339807079336796764054852); DVK(KP918985947, +0.918985947228994779780736114132655398124909697); DVK(KP989821441, +0.989821441880932732376092037776718787376519372); DVK(KP778434453, +0.778434453334651800608337670740821884709317477); DVK(KP830830026, +0.830830026003772851058548298459246407048009821); DVK(KP372785597, +0.372785597771792209609773152906148328659002598); DVK(KP634356270, +0.634356270682424498893150776899916060542806975); DVK(KP715370323, +0.715370323453429719112414662767260662417897278); DVK(KP342584725, +0.342584725681637509502641509861112333758894680); DVK(KP521108558, +0.521108558113202722944698153526659300680427422); { INT i; const R *xi; R *xo; xi = ii; xo = io; for (i = v; i > 0; i = i - VL, xi = xi + (VL * ivs), xo = xo + (VL * ovs), MAKE_VOLATILE_STRIDE(22, is), MAKE_VOLATILE_STRIDE(22, os)) { V T1, Tb, T4, Tq, Tg, Tm, T7, Tp, Ta, To, Tc, T11; T1 = LD(&(xi[0]), ivs, &(xi[0])); { V T2, T3, Te, Tf; T2 = LD(&(xi[WS(is, 1)]), ivs, &(xi[WS(is, 1)])); T3 = LD(&(xi[WS(is, 10)]), ivs, &(xi[0])); Te = LD(&(xi[WS(is, 5)]), ivs, &(xi[WS(is, 1)])); Tf = LD(&(xi[WS(is, 6)]), ivs, &(xi[0])); { V T5, T6, T8, T9; T5 = LD(&(xi[WS(is, 2)]), ivs, &(xi[0])); T6 = LD(&(xi[WS(is, 9)]), ivs, &(xi[WS(is, 1)])); T8 = LD(&(xi[WS(is, 3)]), ivs, &(xi[WS(is, 1)])); T9 = LD(&(xi[WS(is, 8)]), ivs, &(xi[0])); Tb = LD(&(xi[WS(is, 4)]), ivs, &(xi[0])); T4 = VADD(T2, T3); Tq = VSUB(T2, T3); Tg = VADD(Te, Tf); Tm = VSUB(Te, Tf); T7 = VADD(T5, T6); Tp = VSUB(T5, T6); Ta = VADD(T8, T9); To = VSUB(T8, T9); Tc = LD(&(xi[WS(is, 7)]), ivs, &(xi[WS(is, 1)])); } } T11 = VFMA(LDK(KP521108558), Tm, Tq); { V TA, TS, TE, TW, Td, Tn, Ts, Tw, Tr, Tv, TT, TF; Tr = VFNMS(LDK(KP521108558), Tq, Tp); Tv = VFNMS(LDK(KP342584725), T7, Tg); TA = VFMA(LDK(KP715370323), To, Tq); TS = VFMA(LDK(KP521108558), To, Tm); TE = VFNMS(LDK(KP342584725), T4, Ta); TW = VFNMS(LDK(KP342584725), Ta, T7); Td = VADD(Tb, Tc); Tn = VSUB(Tb, Tc); Ts = VFNMS(LDK(KP715370323), Tr, To); Tw = VFNMS(LDK(KP634356270), Tv, T4); TT = VFNMS(LDK(KP715370323), TS, Tp); TF = VFNMS(LDK(KP634356270), TE, Tg); { V Tu, TV, TD, TL, T14, TP, TZ, Tj, Tz, TI, TB, TJ, TM; TB = VFMA(LDK(KP372785597), Tn, TA); TJ = VFNMS(LDK(KP521108558), Tp, Tn); { V T12, TN, TX, Th; T12 = VFMA(LDK(KP715370323), T11, Tn); ST(&(xo[0]), VADD(Tg, VADD(Td, VADD(Ta, VADD(T7, VADD(T4, T1))))), ovs, &(xo[0])); TN = VFNMS(LDK(KP342584725), Td, T4); TX = VFNMS(LDK(KP634356270), TW, Td); Th = VFNMS(LDK(KP342584725), Tg, Td); { V Tt, Tx, TU, TG; Tt = VFNMS(LDK(KP830830026), Ts, Tn); Tx = VFNMS(LDK(KP778434453), Tw, Ta); TU = VFMA(LDK(KP830830026), TT, Tq); TG = VFNMS(LDK(KP778434453), TF, Td); { V TC, TK, T13, TO; TC = VFNMS(LDK(KP830830026), TB, Tm); TK = VFMA(LDK(KP715370323), TJ, Tm); T13 = VFMA(LDK(KP830830026), T12, Tp); TO = VFNMS(LDK(KP634356270), TN, T7); { V TY, Ti, Ty, TH; TY = VFNMS(LDK(KP778434453), TX, T4); Ti = VFNMS(LDK(KP634356270), Th, Ta); Tu = VMUL(LDK(KP989821441), VFNMS(LDK(KP918985947), Tt, Tm)); Ty = VFNMS(LDK(KP876768831), Tx, Td); TV = VMUL(LDK(KP989821441), VFNMS(LDK(KP918985947), TU, Tn)); TH = VFNMS(LDK(KP876768831), TG, T7); TD = VMUL(LDK(KP989821441), VFMA(LDK(KP918985947), TC, Tp)); TL = VFNMS(LDK(KP830830026), TK, To); T14 = VMUL(LDK(KP989821441), VFMA(LDK(KP918985947), T13, To)); TP = VFNMS(LDK(KP778434453), TO, Tg); TZ = VFNMS(LDK(KP876768831), TY, Tg); Tj = VFNMS(LDK(KP778434453), Ti, T7); Tz = VFNMS(LDK(KP959492973), Ty, T1); TI = VFNMS(LDK(KP959492973), TH, T1); } } } } TM = VMUL(LDK(KP989821441), VFNMS(LDK(KP918985947), TL, Tq)); { V TQ, T10, Tk, TR, Tl; TQ = VFNMS(LDK(KP876768831), TP, Ta); T10 = VFNMS(LDK(KP959492973), TZ, T1); Tk = VFNMS(LDK(KP876768831), Tj, T4); ST(&(xo[WS(os, 7)]), VFMAI(TD, Tz), ovs, &(xo[WS(os, 1)])); ST(&(xo[WS(os, 4)]), VFNMSI(TD, Tz), ovs, &(xo[0])); ST(&(xo[WS(os, 8)]), VFNMSI(TM, TI), ovs, &(xo[0])); ST(&(xo[WS(os, 3)]), VFMAI(TM, TI), ovs, &(xo[WS(os, 1)])); TR = VFNMS(LDK(KP959492973), TQ, T1); ST(&(xo[WS(os, 10)]), VFNMSI(T14, T10), ovs, &(xo[0])); ST(&(xo[WS(os, 1)]), VFMAI(T14, T10), ovs, &(xo[WS(os, 1)])); Tl = VFNMS(LDK(KP959492973), Tk, T1); ST(&(xo[WS(os, 9)]), VFMAI(TV, TR), ovs, &(xo[WS(os, 1)])); ST(&(xo[WS(os, 2)]), VFNMSI(TV, TR), ovs, &(xo[0])); ST(&(xo[WS(os, 6)]), VFNMSI(Tu, Tl), ovs, &(xo[0])); ST(&(xo[WS(os, 5)]), VFMAI(Tu, Tl), ovs, &(xo[WS(os, 1)])); } } } } } VLEAVE(); } static const kdft_desc desc = { 11, XSIMD_STRING("n1bv_11"), {15, 5, 55, 0}, &GENUS, 0, 0, 0, 0 }; void XSIMD(codelet_n1bv_11) (planner *p) { X(kdft_register) (p, n1bv_11, &desc); } #else /* HAVE_FMA */ /* Generated by: ../../../genfft/gen_notw_c.native -simd -compact -variables 4 -pipeline-latency 8 -sign 1 -n 11 -name n1bv_11 -include n1b.h */ /* * This function contains 70 FP additions, 50 FP multiplications, * (or, 30 additions, 10 multiplications, 40 fused multiply/add), * 32 stack variables, 10 constants, and 22 memory accesses */ #include "n1b.h" static void n1bv_11(const R *ri, const R *ii, R *ro, R *io, stride is, stride os, INT v, INT ivs, INT ovs) { DVK(KP959492973, +0.959492973614497389890368057066327699062454848); DVK(KP654860733, +0.654860733945285064056925072466293553183791199); DVK(KP142314838, +0.142314838273285140443792668616369668791051361); DVK(KP415415013, +0.415415013001886425529274149229623203524004910); DVK(KP841253532, +0.841253532831181168861811648919367717513292498); DVK(KP540640817, +0.540640817455597582107635954318691695431770608); DVK(KP909631995, +0.909631995354518371411715383079028460060241051); DVK(KP989821441, +0.989821441880932732376092037776718787376519372); DVK(KP755749574, +0.755749574354258283774035843972344420179717445); DVK(KP281732556, +0.281732556841429697711417915346616899035777899); { INT i; const R *xi; R *xo; xi = ii; xo = io; for (i = v; i > 0; i = i - VL, xi = xi + (VL * ivs), xo = xo + (VL * ovs), MAKE_VOLATILE_STRIDE(22, is), MAKE_VOLATILE_STRIDE(22, os)) { V Th, T3, Tm, Tf, Ti, Tc, Tj, T9, Tk, T6, Tl, Ta, Tb, Ts, Tt; Th = LD(&(xi[0]), ivs, &(xi[0])); { V T1, T2, Td, Te; T1 = LD(&(xi[WS(is, 1)]), ivs, &(xi[WS(is, 1)])); T2 = LD(&(xi[WS(is, 10)]), ivs, &(xi[0])); T3 = VSUB(T1, T2); Tm = VADD(T1, T2); Td = LD(&(xi[WS(is, 2)]), ivs, &(xi[0])); Te = LD(&(xi[WS(is, 9)]), ivs, &(xi[WS(is, 1)])); Tf = VSUB(Td, Te); Ti = VADD(Td, Te); } Ta = LD(&(xi[WS(is, 4)]), ivs, &(xi[0])); Tb = LD(&(xi[WS(is, 7)]), ivs, &(xi[WS(is, 1)])); Tc = VSUB(Ta, Tb); Tj = VADD(Ta, Tb); { V T7, T8, T4, T5; T7 = LD(&(xi[WS(is, 5)]), ivs, &(xi[WS(is, 1)])); T8 = LD(&(xi[WS(is, 6)]), ivs, &(xi[0])); T9 = VSUB(T7, T8); Tk = VADD(T7, T8); T4 = LD(&(xi[WS(is, 3)]), ivs, &(xi[WS(is, 1)])); T5 = LD(&(xi[WS(is, 8)]), ivs, &(xi[0])); T6 = VSUB(T4, T5); Tl = VADD(T4, T5); } ST(&(xo[0]), VADD(Th, VADD(Tm, VADD(Ti, VADD(Tl, VADD(Tj, Tk))))), ovs, &(xo[0])); { V Tg, Tn, Tu, Tv; Tg = VBYI(VFMA(LDK(KP281732556), T3, VFMA(LDK(KP755749574), T6, VFNMS(LDK(KP909631995), Tc, VFNMS(LDK(KP540640817), Tf, VMUL(LDK(KP989821441), T9)))))); Tn = VFMA(LDK(KP841253532), Ti, VFMA(LDK(KP415415013), Tj, VFNMS(LDK(KP142314838), Tk, VFNMS(LDK(KP654860733), Tl, VFNMS(LDK(KP959492973), Tm, Th))))); ST(&(xo[WS(os, 5)]), VADD(Tg, Tn), ovs, &(xo[WS(os, 1)])); ST(&(xo[WS(os, 6)]), VSUB(Tn, Tg), ovs, &(xo[0])); Tu = VBYI(VFMA(LDK(KP755749574), T3, VFMA(LDK(KP540640817), T6, VFNMS(LDK(KP909631995), T9, VFNMS(LDK(KP989821441), Tf, VMUL(LDK(KP281732556), Tc)))))); Tv = VFMA(LDK(KP841253532), Tl, VFMA(LDK(KP415415013), Tk, VFNMS(LDK(KP959492973), Tj, VFNMS(LDK(KP142314838), Ti, VFNMS(LDK(KP654860733), Tm, Th))))); ST(&(xo[WS(os, 4)]), VADD(Tu, Tv), ovs, &(xo[0])); ST(&(xo[WS(os, 7)]), VSUB(Tv, Tu), ovs, &(xo[WS(os, 1)])); } Ts = VBYI(VFMA(LDK(KP909631995), T3, VFNMS(LDK(KP540640817), T9, VFNMS(LDK(KP989821441), Tc, VFNMS(LDK(KP281732556), T6, VMUL(LDK(KP755749574), Tf)))))); Tt = VFMA(LDK(KP415415013), Tm, VFMA(LDK(KP841253532), Tk, VFNMS(LDK(KP142314838), Tj, VFNMS(LDK(KP959492973), Tl, VFNMS(LDK(KP654860733), Ti, Th))))); ST(&(xo[WS(os, 2)]), VADD(Ts, Tt), ovs, &(xo[0])); ST(&(xo[WS(os, 9)]), VSUB(Tt, Ts), ovs, &(xo[WS(os, 1)])); { V Tq, Tr, To, Tp; Tq = VBYI(VFMA(LDK(KP540640817), T3, VFMA(LDK(KP909631995), Tf, VFMA(LDK(KP989821441), T6, VFMA(LDK(KP755749574), Tc, VMUL(LDK(KP281732556), T9)))))); Tr = VFMA(LDK(KP841253532), Tm, VFMA(LDK(KP415415013), Ti, VFNMS(LDK(KP959492973), Tk, VFNMS(LDK(KP654860733), Tj, VFNMS(LDK(KP142314838), Tl, Th))))); ST(&(xo[WS(os, 1)]), VADD(Tq, Tr), ovs, &(xo[WS(os, 1)])); ST(&(xo[WS(os, 10)]), VSUB(Tr, Tq), ovs, &(xo[0])); To = VBYI(VFMA(LDK(KP989821441), T3, VFMA(LDK(KP540640817), Tc, VFNMS(LDK(KP909631995), T6, VFNMS(LDK(KP281732556), Tf, VMUL(LDK(KP755749574), T9)))))); Tp = VFMA(LDK(KP415415013), Tl, VFMA(LDK(KP841253532), Tj, VFNMS(LDK(KP654860733), Tk, VFNMS(LDK(KP959492973), Ti, VFNMS(LDK(KP142314838), Tm, Th))))); ST(&(xo[WS(os, 3)]), VADD(To, Tp), ovs, &(xo[WS(os, 1)])); ST(&(xo[WS(os, 8)]), VSUB(Tp, To), ovs, &(xo[0])); } } } VLEAVE(); } static const kdft_desc desc = { 11, XSIMD_STRING("n1bv_11"), {30, 10, 40, 0}, &GENUS, 0, 0, 0, 0 }; void XSIMD(codelet_n1bv_11) (planner *p) { X(kdft_register) (p, n1bv_11, &desc); } #endif /* HAVE_FMA */ fftw-3.3.3/dft/simd/common/n1bv_12.c0000644000175000017500000002112612054410356013726 00000000000000/* * Copyright (c) 2003, 2007-11 Matteo Frigo * Copyright (c) 2003, 2007-11 Massachusetts Institute of Technology * * This program is free software; you can redistribute it and/or modify * it under the terms of the GNU General Public License as published by * the Free Software Foundation; either version 2 of the License, or * (at your option) any later version. * * This program is distributed in the hope that it will be useful, * but WITHOUT ANY WARRANTY; without even the implied warranty of * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the * GNU General Public License for more details. * * You should have received a copy of the GNU General Public License * along with this program; if not, write to the Free Software * Foundation, Inc., 51 Franklin Street, Fifth Floor, Boston, MA 02110-1301 USA * */ /* This file was automatically generated --- DO NOT EDIT */ /* Generated on Sun Nov 25 07:37:02 EST 2012 */ #include "codelet-dft.h" #ifdef HAVE_FMA /* Generated by: ../../../genfft/gen_notw_c.native -fma -reorder-insns -schedule-for-pipeline -simd -compact -variables 4 -pipeline-latency 8 -sign 1 -n 12 -name n1bv_12 -include n1b.h */ /* * This function contains 48 FP additions, 20 FP multiplications, * (or, 30 additions, 2 multiplications, 18 fused multiply/add), * 49 stack variables, 2 constants, and 24 memory accesses */ #include "n1b.h" static void n1bv_12(const R *ri, const R *ii, R *ro, R *io, stride is, stride os, INT v, INT ivs, INT ovs) { DVK(KP866025403, +0.866025403784438646763723170752936183471402627); DVK(KP500000000, +0.500000000000000000000000000000000000000000000); { INT i; const R *xi; R *xo; xi = ii; xo = io; for (i = v; i > 0; i = i - VL, xi = xi + (VL * ivs), xo = xo + (VL * ovs), MAKE_VOLATILE_STRIDE(24, is), MAKE_VOLATILE_STRIDE(24, os)) { V T1, T6, Tc, Th, Td, Te, Ti, Tz, T4, TA, T9, Tj, Tf, Tw; { V T2, T3, T7, T8; T1 = LD(&(xi[0]), ivs, &(xi[0])); T6 = LD(&(xi[WS(is, 6)]), ivs, &(xi[0])); T2 = LD(&(xi[WS(is, 4)]), ivs, &(xi[0])); T3 = LD(&(xi[WS(is, 8)]), ivs, &(xi[0])); T7 = LD(&(xi[WS(is, 10)]), ivs, &(xi[0])); T8 = LD(&(xi[WS(is, 2)]), ivs, &(xi[0])); Tc = LD(&(xi[WS(is, 3)]), ivs, &(xi[WS(is, 1)])); Th = LD(&(xi[WS(is, 9)]), ivs, &(xi[WS(is, 1)])); Td = LD(&(xi[WS(is, 7)]), ivs, &(xi[WS(is, 1)])); Te = LD(&(xi[WS(is, 11)]), ivs, &(xi[WS(is, 1)])); Ti = LD(&(xi[WS(is, 1)]), ivs, &(xi[WS(is, 1)])); Tz = VSUB(T2, T3); T4 = VADD(T2, T3); TA = VSUB(T7, T8); T9 = VADD(T7, T8); Tj = LD(&(xi[WS(is, 5)]), ivs, &(xi[WS(is, 1)])); } Tf = VADD(Td, Te); Tw = VSUB(Td, Te); { V T5, Tp, TJ, TB, Ta, Tq, Tk, Tx, Tg, Ts; T5 = VADD(T1, T4); Tp = VFNMS(LDK(KP500000000), T4, T1); TJ = VSUB(Tz, TA); TB = VADD(Tz, TA); Ta = VADD(T6, T9); Tq = VFNMS(LDK(KP500000000), T9, T6); Tk = VADD(Ti, Tj); Tx = VSUB(Tj, Ti); Tg = VADD(Tc, Tf); Ts = VFNMS(LDK(KP500000000), Tf, Tc); { V Tr, TF, Tb, Tn, TG, Ty, Tl, Tt; Tr = VADD(Tp, Tq); TF = VSUB(Tp, Tq); Tb = VSUB(T5, Ta); Tn = VADD(T5, Ta); TG = VADD(Tw, Tx); Ty = VSUB(Tw, Tx); Tl = VADD(Th, Tk); Tt = VFNMS(LDK(KP500000000), Tk, Th); { V TC, TE, TH, TL, Tu, TI, Tm, To; TC = VMUL(LDK(KP866025403), VSUB(Ty, TB)); TE = VMUL(LDK(KP866025403), VADD(TB, Ty)); TH = VFNMS(LDK(KP866025403), TG, TF); TL = VFMA(LDK(KP866025403), TG, TF); Tu = VADD(Ts, Tt); TI = VSUB(Ts, Tt); Tm = VSUB(Tg, Tl); To = VADD(Tg, Tl); { V TK, TM, Tv, TD; TK = VFMA(LDK(KP866025403), TJ, TI); TM = VFNMS(LDK(KP866025403), TJ, TI); Tv = VSUB(Tr, Tu); TD = VADD(Tr, Tu); ST(&(xo[0]), VADD(Tn, To), ovs, &(xo[0])); ST(&(xo[WS(os, 6)]), VSUB(Tn, To), ovs, &(xo[0])); ST(&(xo[WS(os, 9)]), VFMAI(Tm, Tb), ovs, &(xo[WS(os, 1)])); ST(&(xo[WS(os, 3)]), VFNMSI(Tm, Tb), ovs, &(xo[WS(os, 1)])); ST(&(xo[WS(os, 5)]), VFMAI(TM, TL), ovs, &(xo[WS(os, 1)])); ST(&(xo[WS(os, 7)]), VFNMSI(TM, TL), ovs, &(xo[WS(os, 1)])); ST(&(xo[WS(os, 11)]), VFNMSI(TK, TH), ovs, &(xo[WS(os, 1)])); ST(&(xo[WS(os, 1)]), VFMAI(TK, TH), ovs, &(xo[WS(os, 1)])); ST(&(xo[WS(os, 8)]), VFNMSI(TE, TD), ovs, &(xo[0])); ST(&(xo[WS(os, 4)]), VFMAI(TE, TD), ovs, &(xo[0])); ST(&(xo[WS(os, 2)]), VFMAI(TC, Tv), ovs, &(xo[0])); ST(&(xo[WS(os, 10)]), VFNMSI(TC, Tv), ovs, &(xo[0])); } } } } } } VLEAVE(); } static const kdft_desc desc = { 12, XSIMD_STRING("n1bv_12"), {30, 2, 18, 0}, &GENUS, 0, 0, 0, 0 }; void XSIMD(codelet_n1bv_12) (planner *p) { X(kdft_register) (p, n1bv_12, &desc); } #else /* HAVE_FMA */ /* Generated by: ../../../genfft/gen_notw_c.native -simd -compact -variables 4 -pipeline-latency 8 -sign 1 -n 12 -name n1bv_12 -include n1b.h */ /* * This function contains 48 FP additions, 8 FP multiplications, * (or, 44 additions, 4 multiplications, 4 fused multiply/add), * 27 stack variables, 2 constants, and 24 memory accesses */ #include "n1b.h" static void n1bv_12(const R *ri, const R *ii, R *ro, R *io, stride is, stride os, INT v, INT ivs, INT ovs) { DVK(KP866025403, +0.866025403784438646763723170752936183471402627); DVK(KP500000000, +0.500000000000000000000000000000000000000000000); { INT i; const R *xi; R *xo; xi = ii; xo = io; for (i = v; i > 0; i = i - VL, xi = xi + (VL * ivs), xo = xo + (VL * ovs), MAKE_VOLATILE_STRIDE(24, is), MAKE_VOLATILE_STRIDE(24, os)) { V T5, Ta, TG, TF, Ty, Tm, Ti, Tp, TJ, TI, Tx, Ts; { V T1, T6, T4, Tk, T9, Tl; T1 = LD(&(xi[0]), ivs, &(xi[0])); T6 = LD(&(xi[WS(is, 6)]), ivs, &(xi[0])); { V T2, T3, T7, T8; T2 = LD(&(xi[WS(is, 4)]), ivs, &(xi[0])); T3 = LD(&(xi[WS(is, 8)]), ivs, &(xi[0])); T4 = VADD(T2, T3); Tk = VSUB(T2, T3); T7 = LD(&(xi[WS(is, 10)]), ivs, &(xi[0])); T8 = LD(&(xi[WS(is, 2)]), ivs, &(xi[0])); T9 = VADD(T7, T8); Tl = VSUB(T7, T8); } T5 = VFNMS(LDK(KP500000000), T4, T1); Ta = VFNMS(LDK(KP500000000), T9, T6); TG = VADD(T6, T9); TF = VADD(T1, T4); Ty = VADD(Tk, Tl); Tm = VMUL(LDK(KP866025403), VSUB(Tk, Tl)); } { V Tn, Tq, Te, To, Th, Tr; Tn = LD(&(xi[WS(is, 3)]), ivs, &(xi[WS(is, 1)])); Tq = LD(&(xi[WS(is, 9)]), ivs, &(xi[WS(is, 1)])); { V Tc, Td, Tf, Tg; Tc = LD(&(xi[WS(is, 7)]), ivs, &(xi[WS(is, 1)])); Td = LD(&(xi[WS(is, 11)]), ivs, &(xi[WS(is, 1)])); Te = VSUB(Tc, Td); To = VADD(Tc, Td); Tf = LD(&(xi[WS(is, 1)]), ivs, &(xi[WS(is, 1)])); Tg = LD(&(xi[WS(is, 5)]), ivs, &(xi[WS(is, 1)])); Th = VSUB(Tf, Tg); Tr = VADD(Tf, Tg); } Ti = VMUL(LDK(KP866025403), VSUB(Te, Th)); Tp = VFNMS(LDK(KP500000000), To, Tn); TJ = VADD(Tq, Tr); TI = VADD(Tn, To); Tx = VADD(Te, Th); Ts = VFNMS(LDK(KP500000000), Tr, Tq); } { V TH, TK, TL, TM; TH = VSUB(TF, TG); TK = VBYI(VSUB(TI, TJ)); ST(&(xo[WS(os, 3)]), VSUB(TH, TK), ovs, &(xo[WS(os, 1)])); ST(&(xo[WS(os, 9)]), VADD(TH, TK), ovs, &(xo[WS(os, 1)])); TL = VADD(TF, TG); TM = VADD(TI, TJ); ST(&(xo[WS(os, 6)]), VSUB(TL, TM), ovs, &(xo[0])); ST(&(xo[0]), VADD(TL, TM), ovs, &(xo[0])); } { V Tj, Tv, Tu, Tw, Tb, Tt; Tb = VSUB(T5, Ta); Tj = VSUB(Tb, Ti); Tv = VADD(Tb, Ti); Tt = VSUB(Tp, Ts); Tu = VBYI(VADD(Tm, Tt)); Tw = VBYI(VSUB(Tt, Tm)); ST(&(xo[WS(os, 11)]), VSUB(Tj, Tu), ovs, &(xo[WS(os, 1)])); ST(&(xo[WS(os, 5)]), VADD(Tv, Tw), ovs, &(xo[WS(os, 1)])); ST(&(xo[WS(os, 1)]), VADD(Tj, Tu), ovs, &(xo[WS(os, 1)])); ST(&(xo[WS(os, 7)]), VSUB(Tv, Tw), ovs, &(xo[WS(os, 1)])); } { V Tz, TD, TC, TE, TA, TB; Tz = VBYI(VMUL(LDK(KP866025403), VSUB(Tx, Ty))); TD = VBYI(VMUL(LDK(KP866025403), VADD(Ty, Tx))); TA = VADD(T5, Ta); TB = VADD(Tp, Ts); TC = VSUB(TA, TB); TE = VADD(TA, TB); ST(&(xo[WS(os, 2)]), VADD(Tz, TC), ovs, &(xo[0])); ST(&(xo[WS(os, 8)]), VSUB(TE, TD), ovs, &(xo[0])); ST(&(xo[WS(os, 10)]), VSUB(TC, Tz), ovs, &(xo[0])); ST(&(xo[WS(os, 4)]), VADD(TD, TE), ovs, &(xo[0])); } } } VLEAVE(); } static const kdft_desc desc = { 12, XSIMD_STRING("n1bv_12"), {44, 4, 4, 0}, &GENUS, 0, 0, 0, 0 }; void XSIMD(codelet_n1bv_12) (planner *p) { X(kdft_register) (p, n1bv_12, &desc); } #endif /* HAVE_FMA */ fftw-3.3.3/dft/simd/common/n1bv_13.c0000644000175000017500000003735612054410360013736 00000000000000/* * Copyright (c) 2003, 2007-11 Matteo Frigo * Copyright (c) 2003, 2007-11 Massachusetts Institute of Technology * * This program is free software; you can redistribute it and/or modify * it under the terms of the GNU General Public License as published by * the Free Software Foundation; either version 2 of the License, or * (at your option) any later version. * * This program is distributed in the hope that it will be useful, * but WITHOUT ANY WARRANTY; without even the implied warranty of * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the * GNU General Public License for more details. * * You should have received a copy of the GNU General Public License * along with this program; if not, write to the Free Software * Foundation, Inc., 51 Franklin Street, Fifth Floor, Boston, MA 02110-1301 USA * */ /* This file was automatically generated --- DO NOT EDIT */ /* Generated on Sun Nov 25 07:37:02 EST 2012 */ #include "codelet-dft.h" #ifdef HAVE_FMA /* Generated by: ../../../genfft/gen_notw_c.native -fma -reorder-insns -schedule-for-pipeline -simd -compact -variables 4 -pipeline-latency 8 -sign 1 -n 13 -name n1bv_13 -include n1b.h */ /* * This function contains 88 FP additions, 63 FP multiplications, * (or, 31 additions, 6 multiplications, 57 fused multiply/add), * 96 stack variables, 23 constants, and 26 memory accesses */ #include "n1b.h" static void n1bv_13(const R *ri, const R *ii, R *ro, R *io, stride is, stride os, INT v, INT ivs, INT ovs) { DVK(KP904176221, +0.904176221990848204433795481776887926501523162); DVK(KP575140729, +0.575140729474003121368385547455453388461001608); DVK(KP300462606, +0.300462606288665774426601772289207995520941381); DVK(KP516520780, +0.516520780623489722840901288569017135705033622); DVK(KP522026385, +0.522026385161275033714027226654165028300441940); DVK(KP957805992, +0.957805992594665126462521754605754580515587217); DVK(KP600477271, +0.600477271932665282925769253334763009352012849); DVK(KP251768516, +0.251768516431883313623436926934233488546674281); DVK(KP503537032, +0.503537032863766627246873853868466977093348562); DVK(KP769338817, +0.769338817572980603471413688209101117038278899); DVK(KP859542535, +0.859542535098774820163672132761689612766401925); DVK(KP581704778, +0.581704778510515730456870384989698884939833902); DVK(KP853480001, +0.853480001859823990758994934970528322872359049); DVK(KP083333333, +0.083333333333333333333333333333333333333333333); DVK(KP226109445, +0.226109445035782405468510155372505010481906348); DVK(KP301479260, +0.301479260047709873958013540496673347309208464); DVK(KP686558370, +0.686558370781754340655719594850823015421401653); DVK(KP514918778, +0.514918778086315755491789696138117261566051239); DVK(KP038632954, +0.038632954644348171955506895830342264440241080); DVK(KP612264650, +0.612264650376756543746494474777125408779395514); DVK(KP302775637, +0.302775637731994646559610633735247973125648287); DVK(KP866025403, +0.866025403784438646763723170752936183471402627); DVK(KP500000000, +0.500000000000000000000000000000000000000000000); { INT i; const R *xi; R *xo; xi = ii; xo = io; for (i = v; i > 0; i = i - VL, xi = xi + (VL * ivs), xo = xo + (VL * ovs), MAKE_VOLATILE_STRIDE(26, is), MAKE_VOLATILE_STRIDE(26, os)) { V T1, T7, T2, Tg, Tf, TN, Th, Tq, Ta, Tj, T5, Tr, Tk; T1 = LD(&(xi[0]), ivs, &(xi[0])); { V Td, Te, T8, T9, T3, T4; Td = LD(&(xi[WS(is, 8)]), ivs, &(xi[0])); Te = LD(&(xi[WS(is, 5)]), ivs, &(xi[WS(is, 1)])); T7 = LD(&(xi[WS(is, 12)]), ivs, &(xi[0])); T8 = LD(&(xi[WS(is, 10)]), ivs, &(xi[0])); T9 = LD(&(xi[WS(is, 4)]), ivs, &(xi[0])); T2 = LD(&(xi[WS(is, 1)]), ivs, &(xi[WS(is, 1)])); T3 = LD(&(xi[WS(is, 3)]), ivs, &(xi[WS(is, 1)])); T4 = LD(&(xi[WS(is, 9)]), ivs, &(xi[WS(is, 1)])); Tg = LD(&(xi[WS(is, 11)]), ivs, &(xi[WS(is, 1)])); Tf = VADD(Td, Te); TN = VSUB(Td, Te); Th = LD(&(xi[WS(is, 6)]), ivs, &(xi[0])); Tq = VSUB(T8, T9); Ta = VADD(T8, T9); Tj = LD(&(xi[WS(is, 7)]), ivs, &(xi[WS(is, 1)])); T5 = VADD(T3, T4); Tr = VSUB(T4, T3); Tk = LD(&(xi[WS(is, 2)]), ivs, &(xi[0])); } { V Tt, Ti, Ty, Tb, Ts, TQ, Tx, T6, Tu, Tl; Tt = VSUB(Tg, Th); Ti = VADD(Tg, Th); Ty = VFMS(LDK(KP500000000), Ta, T7); Tb = VADD(T7, Ta); Ts = VSUB(Tq, Tr); TQ = VADD(Tr, Tq); Tx = VFNMS(LDK(KP500000000), T5, T2); T6 = VADD(T2, T5); Tu = VSUB(Tj, Tk); Tl = VADD(Tj, Tk); { V TK, Tz, Tc, TX, Tv, TO, TL, Tm; TK = VADD(Tx, Ty); Tz = VSUB(Tx, Ty); Tc = VADD(T6, Tb); TX = VSUB(T6, Tb); Tv = VSUB(Tt, Tu); TO = VADD(Tt, Tu); TL = VSUB(Ti, Tl); Tm = VADD(Ti, Tl); { V TF, Tw, TP, TY, TT, TM, TA, Tn; TF = VSUB(Ts, Tv); Tw = VADD(Ts, Tv); TP = VFNMS(LDK(KP500000000), TO, TN); TY = VADD(TN, TO); TT = VFNMS(LDK(KP866025403), TL, TK); TM = VFMA(LDK(KP866025403), TL, TK); TA = VFNMS(LDK(KP500000000), Tm, Tf); Tn = VADD(Tf, Tm); { V T1f, T1n, TI, T18, T1k, T1c, TD, T17, T10, T1m, T16, T1e, TU, TR; TU = VFNMS(LDK(KP866025403), TQ, TP); TR = VFMA(LDK(KP866025403), TQ, TP); { V TZ, T15, TE, TB; TZ = VFMA(LDK(KP302775637), TY, TX); T15 = VFNMS(LDK(KP302775637), TX, TY); TE = VSUB(Tz, TA); TB = VADD(Tz, TA); { V TH, To, TV, T13; TH = VSUB(Tc, Tn); To = VADD(Tc, Tn); TV = VFNMS(LDK(KP612264650), TU, TT); T13 = VFMA(LDK(KP612264650), TT, TU); { V TS, T12, TG, T1b; TS = VFNMS(LDK(KP038632954), TR, TM); T12 = VFMA(LDK(KP038632954), TM, TR); TG = VFNMS(LDK(KP514918778), TF, TE); T1b = VFMA(LDK(KP686558370), TE, TF); { V TC, T1a, Tp, TW, T14; TC = VFMA(LDK(KP301479260), TB, Tw); T1a = VFNMS(LDK(KP226109445), Tw, TB); Tp = VFNMS(LDK(KP083333333), To, T1); ST(&(xo[0]), VADD(T1, To), ovs, &(xo[0])); T1f = VFMA(LDK(KP853480001), TV, TS); TW = VFNMS(LDK(KP853480001), TV, TS); T1n = VFMA(LDK(KP853480001), T13, T12); T14 = VFNMS(LDK(KP853480001), T13, T12); TI = VFMA(LDK(KP581704778), TH, TG); T18 = VFNMS(LDK(KP859542535), TG, TH); T1k = VFMA(LDK(KP769338817), T1b, T1a); T1c = VFNMS(LDK(KP769338817), T1b, T1a); TD = VFMA(LDK(KP503537032), TC, Tp); T17 = VFNMS(LDK(KP251768516), TC, Tp); T10 = VMUL(LDK(KP600477271), VFMA(LDK(KP957805992), TZ, TW)); T1m = VFNMS(LDK(KP522026385), TW, TZ); T16 = VMUL(LDK(KP600477271), VFMA(LDK(KP957805992), T15, T14)); T1e = VFNMS(LDK(KP522026385), T14, T15); } } } } { V T1o, T1q, T1g, T1i, T1d, T1h, T1l, T1p; { V T11, TJ, T19, T1j; T11 = VFMA(LDK(KP516520780), TI, TD); TJ = VFNMS(LDK(KP516520780), TI, TD); T19 = VFMA(LDK(KP300462606), T18, T17); T1j = VFNMS(LDK(KP300462606), T18, T17); T1o = VMUL(LDK(KP575140729), VFNMS(LDK(KP904176221), T1n, T1m)); T1q = VMUL(LDK(KP575140729), VFMA(LDK(KP904176221), T1n, T1m)); T1g = VMUL(LDK(KP575140729), VFMA(LDK(KP904176221), T1f, T1e)); T1i = VMUL(LDK(KP575140729), VFNMS(LDK(KP904176221), T1f, T1e)); ST(&(xo[WS(os, 12)]), VFMAI(T16, T11), ovs, &(xo[0])); ST(&(xo[WS(os, 1)]), VFNMSI(T16, T11), ovs, &(xo[WS(os, 1)])); ST(&(xo[WS(os, 8)]), VFNMSI(T10, TJ), ovs, &(xo[0])); ST(&(xo[WS(os, 5)]), VFMAI(T10, TJ), ovs, &(xo[WS(os, 1)])); T1d = VFNMS(LDK(KP503537032), T1c, T19); T1h = VFMA(LDK(KP503537032), T1c, T19); T1l = VFNMS(LDK(KP503537032), T1k, T1j); T1p = VFMA(LDK(KP503537032), T1k, T1j); } ST(&(xo[WS(os, 9)]), VFNMSI(T1g, T1d), ovs, &(xo[WS(os, 1)])); ST(&(xo[WS(os, 4)]), VFMAI(T1g, T1d), ovs, &(xo[0])); ST(&(xo[WS(os, 10)]), VFMAI(T1i, T1h), ovs, &(xo[0])); ST(&(xo[WS(os, 3)]), VFNMSI(T1i, T1h), ovs, &(xo[WS(os, 1)])); ST(&(xo[WS(os, 7)]), VFNMSI(T1o, T1l), ovs, &(xo[WS(os, 1)])); ST(&(xo[WS(os, 6)]), VFMAI(T1o, T1l), ovs, &(xo[0])); ST(&(xo[WS(os, 11)]), VFNMSI(T1q, T1p), ovs, &(xo[WS(os, 1)])); ST(&(xo[WS(os, 2)]), VFMAI(T1q, T1p), ovs, &(xo[0])); } } } } } } } VLEAVE(); } static const kdft_desc desc = { 13, XSIMD_STRING("n1bv_13"), {31, 6, 57, 0}, &GENUS, 0, 0, 0, 0 }; void XSIMD(codelet_n1bv_13) (planner *p) { X(kdft_register) (p, n1bv_13, &desc); } #else /* HAVE_FMA */ /* Generated by: ../../../genfft/gen_notw_c.native -simd -compact -variables 4 -pipeline-latency 8 -sign 1 -n 13 -name n1bv_13 -include n1b.h */ /* * This function contains 88 FP additions, 34 FP multiplications, * (or, 69 additions, 15 multiplications, 19 fused multiply/add), * 60 stack variables, 20 constants, and 26 memory accesses */ #include "n1b.h" static void n1bv_13(const R *ri, const R *ii, R *ro, R *io, stride is, stride os, INT v, INT ivs, INT ovs) { DVK(KP2_000000000, +2.000000000000000000000000000000000000000000000); DVK(KP083333333, +0.083333333333333333333333333333333333333333333); DVK(KP075902986, +0.075902986037193865983102897245103540356428373); DVK(KP251768516, +0.251768516431883313623436926934233488546674281); DVK(KP132983124, +0.132983124607418643793760531921092974399165133); DVK(KP258260390, +0.258260390311744861420450644284508567852516811); DVK(KP1_732050807, +1.732050807568877293527446341505872366942805254); DVK(KP300238635, +0.300238635966332641462884626667381504676006424); DVK(KP011599105, +0.011599105605768290721655456654083252189827041); DVK(KP256247671, +0.256247671582936600958684654061725059144125175); DVK(KP156891391, +0.156891391051584611046832726756003269660212636); DVK(KP174138601, +0.174138601152135905005660794929264742616964676); DVK(KP575140729, +0.575140729474003121368385547455453388461001608); DVK(KP503537032, +0.503537032863766627246873853868466977093348562); DVK(KP113854479, +0.113854479055790798974654345867655310534642560); DVK(KP265966249, +0.265966249214837287587521063842185948798330267); DVK(KP387390585, +0.387390585467617292130675966426762851778775217); DVK(KP300462606, +0.300462606288665774426601772289207995520941381); DVK(KP866025403, +0.866025403784438646763723170752936183471402627); DVK(KP500000000, +0.500000000000000000000000000000000000000000000); { INT i; const R *xi; R *xo; xi = ii; xo = io; for (i = v; i > 0; i = i - VL, xi = xi + (VL * ivs), xo = xo + (VL * ovs), MAKE_VOLATILE_STRIDE(26, is), MAKE_VOLATILE_STRIDE(26, os)) { V TW, Tb, Tm, Ts, TB, TR, TX, TK, TU, Tz, TC, TN, TT; TW = LD(&(xi[0]), ivs, &(xi[0])); { V Te, TH, Ta, Tu, Tp, T5, Tt, To, Th, Tw, Tk, Tx, Tl, TI, Tc; V Td, Tq, Tr; Tc = LD(&(xi[WS(is, 8)]), ivs, &(xi[0])); Td = LD(&(xi[WS(is, 5)]), ivs, &(xi[WS(is, 1)])); Te = VSUB(Tc, Td); TH = VADD(Tc, Td); { V T6, T7, T8, T9; T6 = LD(&(xi[WS(is, 12)]), ivs, &(xi[0])); T7 = LD(&(xi[WS(is, 10)]), ivs, &(xi[0])); T8 = LD(&(xi[WS(is, 4)]), ivs, &(xi[0])); T9 = VADD(T7, T8); Ta = VADD(T6, T9); Tu = VFNMS(LDK(KP500000000), T9, T6); Tp = VSUB(T7, T8); } { V T1, T2, T3, T4; T1 = LD(&(xi[WS(is, 1)]), ivs, &(xi[WS(is, 1)])); T2 = LD(&(xi[WS(is, 3)]), ivs, &(xi[WS(is, 1)])); T3 = LD(&(xi[WS(is, 9)]), ivs, &(xi[WS(is, 1)])); T4 = VADD(T2, T3); T5 = VADD(T1, T4); Tt = VFNMS(LDK(KP500000000), T4, T1); To = VSUB(T2, T3); } { V Tf, Tg, Ti, Tj; Tf = LD(&(xi[WS(is, 11)]), ivs, &(xi[WS(is, 1)])); Tg = LD(&(xi[WS(is, 6)]), ivs, &(xi[0])); Th = VSUB(Tf, Tg); Tw = VADD(Tf, Tg); Ti = LD(&(xi[WS(is, 7)]), ivs, &(xi[WS(is, 1)])); Tj = LD(&(xi[WS(is, 2)]), ivs, &(xi[0])); Tk = VSUB(Ti, Tj); Tx = VADD(Ti, Tj); } Tl = VADD(Th, Tk); TI = VADD(Tw, Tx); Tb = VSUB(T5, Ta); Tm = VADD(Te, Tl); Tq = VMUL(LDK(KP866025403), VSUB(To, Tp)); Tr = VFNMS(LDK(KP500000000), Tl, Te); Ts = VADD(Tq, Tr); TB = VSUB(Tq, Tr); { V TP, TQ, TG, TJ; TP = VADD(T5, Ta); TQ = VADD(TH, TI); TR = VMUL(LDK(KP300462606), VSUB(TP, TQ)); TX = VADD(TP, TQ); TG = VADD(Tt, Tu); TJ = VFNMS(LDK(KP500000000), TI, TH); TK = VSUB(TG, TJ); TU = VADD(TG, TJ); } { V Tv, Ty, TL, TM; Tv = VSUB(Tt, Tu); Ty = VMUL(LDK(KP866025403), VSUB(Tw, Tx)); Tz = VSUB(Tv, Ty); TC = VADD(Tv, Ty); TL = VADD(To, Tp); TM = VSUB(Th, Tk); TN = VSUB(TL, TM); TT = VADD(TL, TM); } } ST(&(xo[0]), VADD(TW, TX), ovs, &(xo[0])); { V T1c, T1n, T11, T14, T17, T1k, Tn, TE, T18, T1j, TS, T1m, TZ, T1f, TA; V TD; { V T1a, T1b, T12, T13; T1a = VFMA(LDK(KP387390585), TN, VMUL(LDK(KP265966249), TK)); T1b = VFNMS(LDK(KP503537032), TU, VMUL(LDK(KP113854479), TT)); T1c = VSUB(T1a, T1b); T1n = VADD(T1a, T1b); T11 = VFMA(LDK(KP575140729), Tb, VMUL(LDK(KP174138601), Tm)); T12 = VFNMS(LDK(KP256247671), Tz, VMUL(LDK(KP156891391), Ts)); T13 = VFMA(LDK(KP011599105), TB, VMUL(LDK(KP300238635), TC)); T14 = VADD(T12, T13); T17 = VSUB(T11, T14); T1k = VMUL(LDK(KP1_732050807), VSUB(T12, T13)); } Tn = VFNMS(LDK(KP575140729), Tm, VMUL(LDK(KP174138601), Tb)); TA = VFMA(LDK(KP256247671), Ts, VMUL(LDK(KP156891391), Tz)); TD = VFNMS(LDK(KP011599105), TC, VMUL(LDK(KP300238635), TB)); TE = VADD(TA, TD); T18 = VMUL(LDK(KP1_732050807), VSUB(TD, TA)); T1j = VSUB(Tn, TE); { V TO, T1e, TV, TY, T1d; TO = VFNMS(LDK(KP132983124), TN, VMUL(LDK(KP258260390), TK)); T1e = VSUB(TR, TO); TV = VFMA(LDK(KP251768516), TT, VMUL(LDK(KP075902986), TU)); TY = VFNMS(LDK(KP083333333), TX, TW); T1d = VSUB(TY, TV); TS = VFMA(LDK(KP2_000000000), TO, TR); T1m = VADD(T1e, T1d); TZ = VFMA(LDK(KP2_000000000), TV, TY); T1f = VSUB(T1d, T1e); } { V TF, T10, T1l, T1o; TF = VBYI(VFMA(LDK(KP2_000000000), TE, Tn)); T10 = VADD(TS, TZ); ST(&(xo[WS(os, 1)]), VADD(TF, T10), ovs, &(xo[WS(os, 1)])); ST(&(xo[WS(os, 12)]), VSUB(T10, TF), ovs, &(xo[0])); { V T15, T16, T1p, T1q; T15 = VBYI(VFMA(LDK(KP2_000000000), T14, T11)); T16 = VSUB(TZ, TS); ST(&(xo[WS(os, 5)]), VADD(T15, T16), ovs, &(xo[WS(os, 1)])); ST(&(xo[WS(os, 8)]), VSUB(T16, T15), ovs, &(xo[0])); T1p = VADD(T1n, T1m); T1q = VBYI(VADD(T1j, T1k)); ST(&(xo[WS(os, 4)]), VSUB(T1p, T1q), ovs, &(xo[0])); ST(&(xo[WS(os, 9)]), VADD(T1q, T1p), ovs, &(xo[WS(os, 1)])); } T1l = VBYI(VSUB(T1j, T1k)); T1o = VSUB(T1m, T1n); ST(&(xo[WS(os, 3)]), VADD(T1l, T1o), ovs, &(xo[WS(os, 1)])); ST(&(xo[WS(os, 10)]), VSUB(T1o, T1l), ovs, &(xo[0])); { V T1h, T1i, T19, T1g; T1h = VBYI(VADD(T18, T17)); T1i = VSUB(T1f, T1c); ST(&(xo[WS(os, 6)]), VADD(T1h, T1i), ovs, &(xo[0])); ST(&(xo[WS(os, 7)]), VSUB(T1i, T1h), ovs, &(xo[WS(os, 1)])); T19 = VBYI(VSUB(T17, T18)); T1g = VADD(T1c, T1f); ST(&(xo[WS(os, 2)]), VADD(T19, T1g), ovs, &(xo[0])); ST(&(xo[WS(os, 11)]), VSUB(T1g, T19), ovs, &(xo[WS(os, 1)])); } } } } } VLEAVE(); } static const kdft_desc desc = { 13, XSIMD_STRING("n1bv_13"), {69, 15, 19, 0}, &GENUS, 0, 0, 0, 0 }; void XSIMD(codelet_n1bv_13) (planner *p) { X(kdft_register) (p, n1bv_13, &desc); } #endif /* HAVE_FMA */ fftw-3.3.3/dft/simd/common/n1bv_14.c0000644000175000017500000002733612054410360013734 00000000000000/* * Copyright (c) 2003, 2007-11 Matteo Frigo * Copyright (c) 2003, 2007-11 Massachusetts Institute of Technology * * This program is free software; you can redistribute it and/or modify * it under the terms of the GNU General Public License as published by * the Free Software Foundation; either version 2 of the License, or * (at your option) any later version. * * This program is distributed in the hope that it will be useful, * but WITHOUT ANY WARRANTY; without even the implied warranty of * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the * GNU General Public License for more details. * * You should have received a copy of the GNU General Public License * along with this program; if not, write to the Free Software * Foundation, Inc., 51 Franklin Street, Fifth Floor, Boston, MA 02110-1301 USA * */ /* This file was automatically generated --- DO NOT EDIT */ /* Generated on Sun Nov 25 07:37:02 EST 2012 */ #include "codelet-dft.h" #ifdef HAVE_FMA /* Generated by: ../../../genfft/gen_notw_c.native -fma -reorder-insns -schedule-for-pipeline -simd -compact -variables 4 -pipeline-latency 8 -sign 1 -n 14 -name n1bv_14 -include n1b.h */ /* * This function contains 74 FP additions, 48 FP multiplications, * (or, 32 additions, 6 multiplications, 42 fused multiply/add), * 63 stack variables, 6 constants, and 28 memory accesses */ #include "n1b.h" static void n1bv_14(const R *ri, const R *ii, R *ro, R *io, stride is, stride os, INT v, INT ivs, INT ovs) { DVK(KP900968867, +0.900968867902419126236102319507445051165919162); DVK(KP801937735, +0.801937735804838252472204639014890102331838324); DVK(KP974927912, +0.974927912181823607018131682993931217232785801); DVK(KP692021471, +0.692021471630095869627814897002069140197260599); DVK(KP554958132, +0.554958132087371191422194871006410481067288862); DVK(KP356895867, +0.356895867892209443894399510021300583399127187); { INT i; const R *xi; R *xo; xi = ii; xo = io; for (i = v; i > 0; i = i - VL, xi = xi + (VL * ivs), xo = xo + (VL * ovs), MAKE_VOLATILE_STRIDE(28, is), MAKE_VOLATILE_STRIDE(28, os)) { V TH, T3, TP, Tn, Ta, Tu, TU, TK, TO, Tk, TM, Tg, TL, Td, T1; V T2; T1 = LD(&(xi[0]), ivs, &(xi[0])); T2 = LD(&(xi[WS(is, 7)]), ivs, &(xi[WS(is, 1)])); { V Ti, TI, T6, TJ, T9, Tj, Te, Tf, Tb, Tc; { V T4, T5, T7, T8, Tl, Tm; T4 = LD(&(xi[WS(is, 2)]), ivs, &(xi[0])); T5 = LD(&(xi[WS(is, 9)]), ivs, &(xi[WS(is, 1)])); T7 = LD(&(xi[WS(is, 12)]), ivs, &(xi[0])); T8 = LD(&(xi[WS(is, 5)]), ivs, &(xi[WS(is, 1)])); Tl = LD(&(xi[WS(is, 8)]), ivs, &(xi[0])); Tm = LD(&(xi[WS(is, 1)]), ivs, &(xi[WS(is, 1)])); Ti = LD(&(xi[WS(is, 6)]), ivs, &(xi[0])); TH = VADD(T1, T2); T3 = VSUB(T1, T2); TI = VADD(T4, T5); T6 = VSUB(T4, T5); TJ = VADD(T7, T8); T9 = VSUB(T7, T8); TP = VADD(Tl, Tm); Tn = VSUB(Tl, Tm); Tj = LD(&(xi[WS(is, 13)]), ivs, &(xi[WS(is, 1)])); Te = LD(&(xi[WS(is, 10)]), ivs, &(xi[0])); Tf = LD(&(xi[WS(is, 3)]), ivs, &(xi[WS(is, 1)])); Tb = LD(&(xi[WS(is, 4)]), ivs, &(xi[0])); Tc = LD(&(xi[WS(is, 11)]), ivs, &(xi[WS(is, 1)])); } Ta = VADD(T6, T9); Tu = VSUB(T6, T9); TU = VSUB(TI, TJ); TK = VADD(TI, TJ); TO = VADD(Ti, Tj); Tk = VSUB(Ti, Tj); TM = VADD(Te, Tf); Tg = VSUB(Te, Tf); TL = VADD(Tb, Tc); Td = VSUB(Tb, Tc); } { V T13, TG, TY, T18, TB, Tw, TT, Tz, T11, T16, TE, Tr, TV, TQ; TV = VSUB(TP, TO); TQ = VADD(TO, TP); { V Ts, To, TW, TN; Ts = VSUB(Tk, Tn); To = VADD(Tk, Tn); TW = VSUB(TM, TL); TN = VADD(TL, TM); { V Tt, Th, TR, T12; Tt = VSUB(Td, Tg); Th = VADD(Td, Tg); TR = VFNMS(LDK(KP356895867), TK, TQ); T12 = VFNMS(LDK(KP554958132), TV, TU); { V Tx, TF, TZ, T14; Tx = VFNMS(LDK(KP356895867), Ta, To); TF = VFMA(LDK(KP554958132), Ts, Tu); ST(&(xo[0]), VADD(TH, VADD(TK, VADD(TN, TQ))), ovs, &(xo[0])); TZ = VFNMS(LDK(KP356895867), TN, TK); T14 = VFNMS(LDK(KP356895867), TQ, TN); { V TX, T17, TC, Tp; TX = VFMA(LDK(KP554958132), TW, TV); T17 = VFMA(LDK(KP554958132), TU, TW); ST(&(xo[WS(os, 7)]), VADD(T3, VADD(Ta, VADD(Th, To))), ovs, &(xo[WS(os, 1)])); TC = VFNMS(LDK(KP356895867), Th, Ta); Tp = VFNMS(LDK(KP356895867), To, Th); { V TA, Tv, TS, Ty; TA = VFMA(LDK(KP554958132), Tt, Ts); Tv = VFNMS(LDK(KP554958132), Tu, Tt); TS = VFNMS(LDK(KP692021471), TR, TN); T13 = VMUL(LDK(KP974927912), VFNMS(LDK(KP801937735), T12, TW)); Ty = VFNMS(LDK(KP692021471), Tx, Th); TG = VMUL(LDK(KP974927912), VFMA(LDK(KP801937735), TF, Tt)); { V T10, T15, TD, Tq; T10 = VFNMS(LDK(KP692021471), TZ, TQ); T15 = VFNMS(LDK(KP692021471), T14, TK); TY = VMUL(LDK(KP974927912), VFMA(LDK(KP801937735), TX, TU)); T18 = VMUL(LDK(KP974927912), VFNMS(LDK(KP801937735), T17, TV)); TD = VFNMS(LDK(KP692021471), TC, To); Tq = VFNMS(LDK(KP692021471), Tp, Ta); TB = VMUL(LDK(KP974927912), VFNMS(LDK(KP801937735), TA, Tu)); Tw = VMUL(LDK(KP974927912), VFNMS(LDK(KP801937735), Tv, Ts)); TT = VFNMS(LDK(KP900968867), TS, TH); Tz = VFNMS(LDK(KP900968867), Ty, T3); T11 = VFNMS(LDK(KP900968867), T10, TH); T16 = VFNMS(LDK(KP900968867), T15, TH); TE = VFNMS(LDK(KP900968867), TD, T3); Tr = VFNMS(LDK(KP900968867), Tq, T3); } } } } } } ST(&(xo[WS(os, 2)]), VFMAI(TY, TT), ovs, &(xo[0])); ST(&(xo[WS(os, 12)]), VFNMSI(TY, TT), ovs, &(xo[0])); ST(&(xo[WS(os, 9)]), VFMAI(TB, Tz), ovs, &(xo[WS(os, 1)])); ST(&(xo[WS(os, 5)]), VFNMSI(TB, Tz), ovs, &(xo[WS(os, 1)])); ST(&(xo[WS(os, 6)]), VFMAI(T13, T11), ovs, &(xo[0])); ST(&(xo[WS(os, 8)]), VFNMSI(T13, T11), ovs, &(xo[0])); ST(&(xo[WS(os, 4)]), VFMAI(T18, T16), ovs, &(xo[0])); ST(&(xo[WS(os, 10)]), VFNMSI(T18, T16), ovs, &(xo[0])); ST(&(xo[WS(os, 13)]), VFNMSI(TG, TE), ovs, &(xo[WS(os, 1)])); ST(&(xo[WS(os, 1)]), VFMAI(TG, TE), ovs, &(xo[WS(os, 1)])); ST(&(xo[WS(os, 11)]), VFNMSI(Tw, Tr), ovs, &(xo[WS(os, 1)])); ST(&(xo[WS(os, 3)]), VFMAI(Tw, Tr), ovs, &(xo[WS(os, 1)])); } } } VLEAVE(); } static const kdft_desc desc = { 14, XSIMD_STRING("n1bv_14"), {32, 6, 42, 0}, &GENUS, 0, 0, 0, 0 }; void XSIMD(codelet_n1bv_14) (planner *p) { X(kdft_register) (p, n1bv_14, &desc); } #else /* HAVE_FMA */ /* Generated by: ../../../genfft/gen_notw_c.native -simd -compact -variables 4 -pipeline-latency 8 -sign 1 -n 14 -name n1bv_14 -include n1b.h */ /* * This function contains 74 FP additions, 36 FP multiplications, * (or, 50 additions, 12 multiplications, 24 fused multiply/add), * 33 stack variables, 6 constants, and 28 memory accesses */ #include "n1b.h" static void n1bv_14(const R *ri, const R *ii, R *ro, R *io, stride is, stride os, INT v, INT ivs, INT ovs) { DVK(KP900968867, +0.900968867902419126236102319507445051165919162); DVK(KP222520933, +0.222520933956314404288902564496794759466355569); DVK(KP623489801, +0.623489801858733530525004884004239810632274731); DVK(KP781831482, +0.781831482468029808708444526674057750232334519); DVK(KP974927912, +0.974927912181823607018131682993931217232785801); DVK(KP433883739, +0.433883739117558120475768332848358754609990728); { INT i; const R *xi; R *xo; xi = ii; xo = io; for (i = v; i > 0; i = i - VL, xi = xi + (VL * ivs), xo = xo + (VL * ovs), MAKE_VOLATILE_STRIDE(28, is), MAKE_VOLATILE_STRIDE(28, os)) { V Tp, Ty, Tl, TL, Tq, TE, T7, TJ, Ts, TB, Te, TK, Tr, TH, Tn; V To; Tn = LD(&(xi[0]), ivs, &(xi[0])); To = LD(&(xi[WS(is, 7)]), ivs, &(xi[WS(is, 1)])); Tp = VSUB(Tn, To); Ty = VADD(Tn, To); { V Th, TC, Tk, TD; { V Tf, Tg, Ti, Tj; Tf = LD(&(xi[WS(is, 4)]), ivs, &(xi[0])); Tg = LD(&(xi[WS(is, 11)]), ivs, &(xi[WS(is, 1)])); Th = VSUB(Tf, Tg); TC = VADD(Tf, Tg); Ti = LD(&(xi[WS(is, 10)]), ivs, &(xi[0])); Tj = LD(&(xi[WS(is, 3)]), ivs, &(xi[WS(is, 1)])); Tk = VSUB(Ti, Tj); TD = VADD(Ti, Tj); } Tl = VSUB(Th, Tk); TL = VSUB(TD, TC); Tq = VADD(Th, Tk); TE = VADD(TC, TD); } { V T3, Tz, T6, TA; { V T1, T2, T4, T5; T1 = LD(&(xi[WS(is, 2)]), ivs, &(xi[0])); T2 = LD(&(xi[WS(is, 9)]), ivs, &(xi[WS(is, 1)])); T3 = VSUB(T1, T2); Tz = VADD(T1, T2); T4 = LD(&(xi[WS(is, 12)]), ivs, &(xi[0])); T5 = LD(&(xi[WS(is, 5)]), ivs, &(xi[WS(is, 1)])); T6 = VSUB(T4, T5); TA = VADD(T4, T5); } T7 = VSUB(T3, T6); TJ = VSUB(Tz, TA); Ts = VADD(T3, T6); TB = VADD(Tz, TA); } { V Ta, TF, Td, TG; { V T8, T9, Tb, Tc; T8 = LD(&(xi[WS(is, 6)]), ivs, &(xi[0])); T9 = LD(&(xi[WS(is, 13)]), ivs, &(xi[WS(is, 1)])); Ta = VSUB(T8, T9); TF = VADD(T8, T9); Tb = LD(&(xi[WS(is, 8)]), ivs, &(xi[0])); Tc = LD(&(xi[WS(is, 1)]), ivs, &(xi[WS(is, 1)])); Td = VSUB(Tb, Tc); TG = VADD(Tb, Tc); } Te = VSUB(Ta, Td); TK = VSUB(TG, TF); Tr = VADD(Ta, Td); TH = VADD(TF, TG); } ST(&(xo[WS(os, 7)]), VADD(Tp, VADD(Ts, VADD(Tq, Tr))), ovs, &(xo[WS(os, 1)])); ST(&(xo[0]), VADD(Ty, VADD(TB, VADD(TE, TH))), ovs, &(xo[0])); { V Tm, Tt, TQ, TP; Tm = VBYI(VFMA(LDK(KP433883739), T7, VFNMS(LDK(KP781831482), Tl, VMUL(LDK(KP974927912), Te)))); Tt = VFMA(LDK(KP623489801), Tq, VFNMS(LDK(KP222520933), Tr, VFNMS(LDK(KP900968867), Ts, Tp))); ST(&(xo[WS(os, 3)]), VADD(Tm, Tt), ovs, &(xo[WS(os, 1)])); ST(&(xo[WS(os, 11)]), VSUB(Tt, Tm), ovs, &(xo[WS(os, 1)])); TQ = VBYI(VFMA(LDK(KP974927912), TJ, VFMA(LDK(KP433883739), TL, VMUL(LDK(KP781831482), TK)))); TP = VFMA(LDK(KP623489801), TH, VFNMS(LDK(KP900968867), TE, VFNMS(LDK(KP222520933), TB, Ty))); ST(&(xo[WS(os, 12)]), VSUB(TP, TQ), ovs, &(xo[0])); ST(&(xo[WS(os, 2)]), VADD(TP, TQ), ovs, &(xo[0])); } { V Tu, Tv, TM, TI; Tu = VBYI(VFMA(LDK(KP781831482), T7, VFMA(LDK(KP974927912), Tl, VMUL(LDK(KP433883739), Te)))); Tv = VFMA(LDK(KP623489801), Ts, VFNMS(LDK(KP900968867), Tr, VFNMS(LDK(KP222520933), Tq, Tp))); ST(&(xo[WS(os, 1)]), VADD(Tu, Tv), ovs, &(xo[WS(os, 1)])); ST(&(xo[WS(os, 13)]), VSUB(Tv, Tu), ovs, &(xo[WS(os, 1)])); TM = VBYI(VFNMS(LDK(KP433883739), TK, VFNMS(LDK(KP974927912), TL, VMUL(LDK(KP781831482), TJ)))); TI = VFMA(LDK(KP623489801), TB, VFNMS(LDK(KP900968867), TH, VFNMS(LDK(KP222520933), TE, Ty))); ST(&(xo[WS(os, 6)]), VSUB(TI, TM), ovs, &(xo[0])); ST(&(xo[WS(os, 8)]), VADD(TI, TM), ovs, &(xo[0])); } { V TO, TN, Tx, Tw; TO = VBYI(VFMA(LDK(KP433883739), TJ, VFNMS(LDK(KP974927912), TK, VMUL(LDK(KP781831482), TL)))); TN = VFMA(LDK(KP623489801), TE, VFNMS(LDK(KP222520933), TH, VFNMS(LDK(KP900968867), TB, Ty))); ST(&(xo[WS(os, 4)]), VSUB(TN, TO), ovs, &(xo[0])); ST(&(xo[WS(os, 10)]), VADD(TN, TO), ovs, &(xo[0])); Tx = VBYI(VFNMS(LDK(KP781831482), Te, VFNMS(LDK(KP433883739), Tl, VMUL(LDK(KP974927912), T7)))); Tw = VFMA(LDK(KP623489801), Tr, VFNMS(LDK(KP900968867), Tq, VFNMS(LDK(KP222520933), Ts, Tp))); ST(&(xo[WS(os, 5)]), VSUB(Tw, Tx), ovs, &(xo[WS(os, 1)])); ST(&(xo[WS(os, 9)]), VADD(Tx, Tw), ovs, &(xo[WS(os, 1)])); } } } VLEAVE(); } static const kdft_desc desc = { 14, XSIMD_STRING("n1bv_14"), {50, 12, 24, 0}, &GENUS, 0, 0, 0, 0 }; void XSIMD(codelet_n1bv_14) (planner *p) { X(kdft_register) (p, n1bv_14, &desc); } #endif /* HAVE_FMA */ fftw-3.3.3/dft/simd/common/n1bv_15.c0000644000175000017500000003157412054410360013734 00000000000000/* * Copyright (c) 2003, 2007-11 Matteo Frigo * Copyright (c) 2003, 2007-11 Massachusetts Institute of Technology * * This program is free software; you can redistribute it and/or modify * it under the terms of the GNU General Public License as published by * the Free Software Foundation; either version 2 of the License, or * (at your option) any later version. * * This program is distributed in the hope that it will be useful, * but WITHOUT ANY WARRANTY; without even the implied warranty of * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the * GNU General Public License for more details. * * You should have received a copy of the GNU General Public License * along with this program; if not, write to the Free Software * Foundation, Inc., 51 Franklin Street, Fifth Floor, Boston, MA 02110-1301 USA * */ /* This file was automatically generated --- DO NOT EDIT */ /* Generated on Sun Nov 25 07:37:04 EST 2012 */ #include "codelet-dft.h" #ifdef HAVE_FMA /* Generated by: ../../../genfft/gen_notw_c.native -fma -reorder-insns -schedule-for-pipeline -simd -compact -variables 4 -pipeline-latency 8 -sign 1 -n 15 -name n1bv_15 -include n1b.h */ /* * This function contains 78 FP additions, 49 FP multiplications, * (or, 36 additions, 7 multiplications, 42 fused multiply/add), * 78 stack variables, 8 constants, and 30 memory accesses */ #include "n1b.h" static void n1bv_15(const R *ri, const R *ii, R *ro, R *io, stride is, stride os, INT v, INT ivs, INT ovs) { DVK(KP823639103, +0.823639103546331925877420039278190003029660514); DVK(KP910592997, +0.910592997310029334643087372129977886038870291); DVK(KP559016994, +0.559016994374947424102293417182819058860154590); DVK(KP951056516, +0.951056516295153572116439333379382143405698634); DVK(KP866025403, +0.866025403784438646763723170752936183471402627); DVK(KP250000000, +0.250000000000000000000000000000000000000000000); DVK(KP618033988, +0.618033988749894848204586834365638117720309180); DVK(KP500000000, +0.500000000000000000000000000000000000000000000); { INT i; const R *xi; R *xo; xi = ii; xo = io; for (i = v; i > 0; i = i - VL, xi = xi + (VL * ivs), xo = xo + (VL * ovs), MAKE_VOLATILE_STRIDE(30, is), MAKE_VOLATILE_STRIDE(30, os)) { V Tb, TH, Tw, TA, Th, T11, T5, Ti, T12, Ta, Tx, Te, Tq, T16, Tj; V T1, T2, T3; T1 = LD(&(xi[0]), ivs, &(xi[0])); T2 = LD(&(xi[WS(is, 5)]), ivs, &(xi[WS(is, 1)])); T3 = LD(&(xi[WS(is, 10)]), ivs, &(xi[0])); { V T6, T7, T8, Tm, Tn, To; T6 = LD(&(xi[WS(is, 3)]), ivs, &(xi[WS(is, 1)])); T7 = LD(&(xi[WS(is, 8)]), ivs, &(xi[0])); T8 = LD(&(xi[WS(is, 13)]), ivs, &(xi[WS(is, 1)])); Tm = LD(&(xi[WS(is, 9)]), ivs, &(xi[WS(is, 1)])); Tn = LD(&(xi[WS(is, 14)]), ivs, &(xi[0])); To = LD(&(xi[WS(is, 4)]), ivs, &(xi[0])); { V T4, Tc, T9, Td, Tp; Tb = LD(&(xi[WS(is, 12)]), ivs, &(xi[0])); T4 = VADD(T2, T3); TH = VSUB(T2, T3); Tc = LD(&(xi[WS(is, 2)]), ivs, &(xi[0])); Tw = VSUB(T7, T8); T9 = VADD(T7, T8); Td = LD(&(xi[WS(is, 7)]), ivs, &(xi[WS(is, 1)])); Tp = VADD(Tn, To); TA = VSUB(Tn, To); Th = LD(&(xi[WS(is, 6)]), ivs, &(xi[0])); T11 = VADD(T1, T4); T5 = VFNMS(LDK(KP500000000), T4, T1); Ti = LD(&(xi[WS(is, 11)]), ivs, &(xi[WS(is, 1)])); T12 = VADD(T6, T9); Ta = VFNMS(LDK(KP500000000), T9, T6); Tx = VSUB(Tc, Td); Te = VADD(Tc, Td); Tq = VFNMS(LDK(KP500000000), Tp, Tm); T16 = VADD(Tm, Tp); Tj = LD(&(xi[WS(is, 1)]), ivs, &(xi[WS(is, 1)])); } } { V TI, Ty, T13, Tf, Tz, Tk; TI = VADD(Tw, Tx); Ty = VSUB(Tw, Tx); T13 = VADD(Tb, Te); Tf = VFNMS(LDK(KP500000000), Te, Tb); Tz = VSUB(Ti, Tj); Tk = VADD(Ti, Tj); { V T1d, T14, Tg, TE, TJ, TB, T15, Tl; T1d = VSUB(T12, T13); T14 = VADD(T12, T13); Tg = VADD(Ta, Tf); TE = VSUB(Ta, Tf); TJ = VADD(Tz, TA); TB = VSUB(Tz, TA); T15 = VADD(Th, Tk); Tl = VFNMS(LDK(KP500000000), Tk, Th); { V TM, TK, TS, TC, T1c, T17, Tr, TF, TL, T10; TM = VSUB(TI, TJ); TK = VADD(TI, TJ); TS = VFNMS(LDK(KP618033988), Ty, TB); TC = VFMA(LDK(KP618033988), TB, Ty); T1c = VSUB(T15, T16); T17 = VADD(T15, T16); Tr = VADD(Tl, Tq); TF = VSUB(Tl, Tq); TL = VFNMS(LDK(KP250000000), TK, TH); T10 = VMUL(LDK(KP866025403), VADD(TH, TK)); { V T1g, T1e, T1a, Tu, Ts, TU, TG, TV, TN, T19, T18, Tt, TZ; T1g = VMUL(LDK(KP951056516), VFMA(LDK(KP618033988), T1c, T1d)); T1e = VMUL(LDK(KP951056516), VFNMS(LDK(KP618033988), T1d, T1c)); T18 = VADD(T14, T17); T1a = VSUB(T14, T17); Tu = VSUB(Tg, Tr); Ts = VADD(Tg, Tr); TU = VFNMS(LDK(KP618033988), TE, TF); TG = VFMA(LDK(KP618033988), TF, TE); TV = VFNMS(LDK(KP559016994), TM, TL); TN = VFMA(LDK(KP559016994), TM, TL); ST(&(xo[0]), VADD(T11, T18), ovs, &(xo[0])); T19 = VFNMS(LDK(KP250000000), T18, T11); Tt = VFNMS(LDK(KP250000000), Ts, T5); TZ = VADD(T5, Ts); { V TW, TY, TQ, TO, T1b, T1f, TR, Tv; TW = VMUL(LDK(KP951056516), VFMA(LDK(KP910592997), TV, TU)); TY = VMUL(LDK(KP951056516), VFNMS(LDK(KP910592997), TV, TU)); TQ = VMUL(LDK(KP951056516), VFNMS(LDK(KP910592997), TN, TG)); TO = VMUL(LDK(KP951056516), VFMA(LDK(KP910592997), TN, TG)); T1b = VFNMS(LDK(KP559016994), T1a, T19); T1f = VFMA(LDK(KP559016994), T1a, T19); TR = VFNMS(LDK(KP559016994), Tu, Tt); Tv = VFMA(LDK(KP559016994), Tu, Tt); ST(&(xo[WS(os, 10)]), VFMAI(T10, TZ), ovs, &(xo[0])); ST(&(xo[WS(os, 5)]), VFNMSI(T10, TZ), ovs, &(xo[WS(os, 1)])); { V TT, TX, TP, TD; ST(&(xo[WS(os, 12)]), VFNMSI(T1e, T1b), ovs, &(xo[0])); ST(&(xo[WS(os, 3)]), VFMAI(T1e, T1b), ovs, &(xo[WS(os, 1)])); ST(&(xo[WS(os, 9)]), VFNMSI(T1g, T1f), ovs, &(xo[WS(os, 1)])); ST(&(xo[WS(os, 6)]), VFMAI(T1g, T1f), ovs, &(xo[0])); TT = VFNMS(LDK(KP823639103), TS, TR); TX = VFMA(LDK(KP823639103), TS, TR); TP = VFMA(LDK(KP823639103), TC, Tv); TD = VFNMS(LDK(KP823639103), TC, Tv); ST(&(xo[WS(os, 13)]), VFMAI(TW, TT), ovs, &(xo[WS(os, 1)])); ST(&(xo[WS(os, 2)]), VFNMSI(TW, TT), ovs, &(xo[0])); ST(&(xo[WS(os, 8)]), VFMAI(TY, TX), ovs, &(xo[0])); ST(&(xo[WS(os, 7)]), VFNMSI(TY, TX), ovs, &(xo[WS(os, 1)])); ST(&(xo[WS(os, 11)]), VFMAI(TQ, TP), ovs, &(xo[WS(os, 1)])); ST(&(xo[WS(os, 4)]), VFNMSI(TQ, TP), ovs, &(xo[0])); ST(&(xo[WS(os, 14)]), VFNMSI(TO, TD), ovs, &(xo[0])); ST(&(xo[WS(os, 1)]), VFMAI(TO, TD), ovs, &(xo[WS(os, 1)])); } } } } } } } } VLEAVE(); } static const kdft_desc desc = { 15, XSIMD_STRING("n1bv_15"), {36, 7, 42, 0}, &GENUS, 0, 0, 0, 0 }; void XSIMD(codelet_n1bv_15) (planner *p) { X(kdft_register) (p, n1bv_15, &desc); } #else /* HAVE_FMA */ /* Generated by: ../../../genfft/gen_notw_c.native -simd -compact -variables 4 -pipeline-latency 8 -sign 1 -n 15 -name n1bv_15 -include n1b.h */ /* * This function contains 78 FP additions, 25 FP multiplications, * (or, 64 additions, 11 multiplications, 14 fused multiply/add), * 55 stack variables, 10 constants, and 30 memory accesses */ #include "n1b.h" static void n1bv_15(const R *ri, const R *ii, R *ro, R *io, stride is, stride os, INT v, INT ivs, INT ovs) { DVK(KP216506350, +0.216506350946109661690930792688234045867850657); DVK(KP509036960, +0.509036960455127183450980863393907648510733164); DVK(KP823639103, +0.823639103546331925877420039278190003029660514); DVK(KP951056516, +0.951056516295153572116439333379382143405698634); DVK(KP587785252, +0.587785252292473129168705954639072768597652438); DVK(KP250000000, +0.250000000000000000000000000000000000000000000); DVK(KP559016994, +0.559016994374947424102293417182819058860154590); DVK(KP866025403, +0.866025403784438646763723170752936183471402627); DVK(KP484122918, +0.484122918275927110647408174972799951354115213); DVK(KP500000000, +0.500000000000000000000000000000000000000000000); { INT i; const R *xi; R *xo; xi = ii; xo = io; for (i = v; i > 0; i = i - VL, xi = xi + (VL * ivs), xo = xo + (VL * ovs), MAKE_VOLATILE_STRIDE(30, is), MAKE_VOLATILE_STRIDE(30, os)) { V Ti, T11, TH, Ts, TL, TM, Tz, TC, TD, TI, T12, T13, T14, T15, T16; V T17, Tf, Tj, TZ, T10; { V TF, Tg, Th, TG; TF = LD(&(xi[0]), ivs, &(xi[0])); Tg = LD(&(xi[WS(is, 5)]), ivs, &(xi[WS(is, 1)])); Th = LD(&(xi[WS(is, 10)]), ivs, &(xi[0])); TG = VADD(Tg, Th); Ti = VSUB(Tg, Th); T11 = VADD(TF, TG); TH = VFNMS(LDK(KP500000000), TG, TF); } { V Tm, Tn, T3, To, Tw, Tx, Td, Ty, Tp, Tq, T6, Tr, Tt, Tu, Ta; V Tv, T7, Te; { V T1, T2, Tb, Tc; Tm = LD(&(xi[WS(is, 3)]), ivs, &(xi[WS(is, 1)])); T1 = LD(&(xi[WS(is, 8)]), ivs, &(xi[0])); T2 = LD(&(xi[WS(is, 13)]), ivs, &(xi[WS(is, 1)])); Tn = VADD(T1, T2); T3 = VSUB(T1, T2); To = VFNMS(LDK(KP500000000), Tn, Tm); Tw = LD(&(xi[WS(is, 9)]), ivs, &(xi[WS(is, 1)])); Tb = LD(&(xi[WS(is, 14)]), ivs, &(xi[0])); Tc = LD(&(xi[WS(is, 4)]), ivs, &(xi[0])); Tx = VADD(Tb, Tc); Td = VSUB(Tb, Tc); Ty = VFNMS(LDK(KP500000000), Tx, Tw); } { V T4, T5, T8, T9; Tp = LD(&(xi[WS(is, 12)]), ivs, &(xi[0])); T4 = LD(&(xi[WS(is, 2)]), ivs, &(xi[0])); T5 = LD(&(xi[WS(is, 7)]), ivs, &(xi[WS(is, 1)])); Tq = VADD(T4, T5); T6 = VSUB(T4, T5); Tr = VFNMS(LDK(KP500000000), Tq, Tp); Tt = LD(&(xi[WS(is, 6)]), ivs, &(xi[0])); T8 = LD(&(xi[WS(is, 11)]), ivs, &(xi[WS(is, 1)])); T9 = LD(&(xi[WS(is, 1)]), ivs, &(xi[WS(is, 1)])); Tu = VADD(T8, T9); Ta = VSUB(T8, T9); Tv = VFNMS(LDK(KP500000000), Tu, Tt); } Ts = VSUB(To, Tr); TL = VSUB(T3, T6); TM = VSUB(Ta, Td); Tz = VSUB(Tv, Ty); TC = VADD(To, Tr); TD = VADD(Tv, Ty); TI = VADD(TC, TD); T12 = VADD(Tm, Tn); T13 = VADD(Tp, Tq); T14 = VADD(T12, T13); T15 = VADD(Tt, Tu); T16 = VADD(Tw, Tx); T17 = VADD(T15, T16); T7 = VADD(T3, T6); Te = VADD(Ta, Td); Tf = VMUL(LDK(KP484122918), VSUB(T7, Te)); Tj = VADD(T7, Te); } TZ = VADD(TH, TI); T10 = VBYI(VMUL(LDK(KP866025403), VADD(Ti, Tj))); ST(&(xo[WS(os, 5)]), VSUB(TZ, T10), ovs, &(xo[WS(os, 1)])); ST(&(xo[WS(os, 10)]), VADD(T10, TZ), ovs, &(xo[0])); { V T1a, T18, T19, T1e, T1f, T1c, T1d, T1g, T1b; T1a = VMUL(LDK(KP559016994), VSUB(T14, T17)); T18 = VADD(T14, T17); T19 = VFNMS(LDK(KP250000000), T18, T11); T1c = VSUB(T12, T13); T1d = VSUB(T15, T16); T1e = VBYI(VFNMS(LDK(KP951056516), T1d, VMUL(LDK(KP587785252), T1c))); T1f = VBYI(VFMA(LDK(KP951056516), T1c, VMUL(LDK(KP587785252), T1d))); ST(&(xo[0]), VADD(T11, T18), ovs, &(xo[0])); T1g = VADD(T1a, T19); ST(&(xo[WS(os, 6)]), VADD(T1f, T1g), ovs, &(xo[0])); ST(&(xo[WS(os, 9)]), VSUB(T1g, T1f), ovs, &(xo[WS(os, 1)])); T1b = VSUB(T19, T1a); ST(&(xo[WS(os, 3)]), VSUB(T1b, T1e), ovs, &(xo[WS(os, 1)])); ST(&(xo[WS(os, 12)]), VADD(T1e, T1b), ovs, &(xo[0])); } { V TA, TN, TU, TS, Tl, TR, TK, TV, Tk, TE, TJ; TA = VFMA(LDK(KP951056516), Ts, VMUL(LDK(KP587785252), Tz)); TN = VFMA(LDK(KP823639103), TL, VMUL(LDK(KP509036960), TM)); TU = VFNMS(LDK(KP823639103), TM, VMUL(LDK(KP509036960), TL)); TS = VFNMS(LDK(KP951056516), Tz, VMUL(LDK(KP587785252), Ts)); Tk = VFNMS(LDK(KP216506350), Tj, VMUL(LDK(KP866025403), Ti)); Tl = VADD(Tf, Tk); TR = VSUB(Tf, Tk); TE = VMUL(LDK(KP559016994), VSUB(TC, TD)); TJ = VFNMS(LDK(KP250000000), TI, TH); TK = VADD(TE, TJ); TV = VSUB(TJ, TE); { V TB, TO, TX, TY; TB = VBYI(VADD(Tl, TA)); TO = VSUB(TK, TN); ST(&(xo[WS(os, 1)]), VADD(TB, TO), ovs, &(xo[WS(os, 1)])); ST(&(xo[WS(os, 14)]), VSUB(TO, TB), ovs, &(xo[0])); TX = VBYI(VSUB(TS, TR)); TY = VSUB(TV, TU); ST(&(xo[WS(os, 7)]), VADD(TX, TY), ovs, &(xo[WS(os, 1)])); ST(&(xo[WS(os, 8)]), VSUB(TY, TX), ovs, &(xo[0])); } { V TP, TQ, TT, TW; TP = VBYI(VSUB(Tl, TA)); TQ = VADD(TN, TK); ST(&(xo[WS(os, 4)]), VADD(TP, TQ), ovs, &(xo[0])); ST(&(xo[WS(os, 11)]), VSUB(TQ, TP), ovs, &(xo[WS(os, 1)])); TT = VBYI(VADD(TR, TS)); TW = VADD(TU, TV); ST(&(xo[WS(os, 2)]), VADD(TT, TW), ovs, &(xo[0])); ST(&(xo[WS(os, 13)]), VSUB(TW, TT), ovs, &(xo[WS(os, 1)])); } } } } VLEAVE(); } static const kdft_desc desc = { 15, XSIMD_STRING("n1bv_15"), {64, 11, 14, 0}, &GENUS, 0, 0, 0, 0 }; void XSIMD(codelet_n1bv_15) (planner *p) { X(kdft_register) (p, n1bv_15, &desc); } #endif /* HAVE_FMA */ fftw-3.3.3/dft/simd/common/n1bv_16.c0000644000175000017500000002673112054410361013735 00000000000000/* * Copyright (c) 2003, 2007-11 Matteo Frigo * Copyright (c) 2003, 2007-11 Massachusetts Institute of Technology * * This program is free software; you can redistribute it and/or modify * it under the terms of the GNU General Public License as published by * the Free Software Foundation; either version 2 of the License, or * (at your option) any later version. * * This program is distributed in the hope that it will be useful, * but WITHOUT ANY WARRANTY; without even the implied warranty of * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the * GNU General Public License for more details. * * You should have received a copy of the GNU General Public License * along with this program; if not, write to the Free Software * Foundation, Inc., 51 Franklin Street, Fifth Floor, Boston, MA 02110-1301 USA * */ /* This file was automatically generated --- DO NOT EDIT */ /* Generated on Sun Nov 25 07:37:04 EST 2012 */ #include "codelet-dft.h" #ifdef HAVE_FMA /* Generated by: ../../../genfft/gen_notw_c.native -fma -reorder-insns -schedule-for-pipeline -simd -compact -variables 4 -pipeline-latency 8 -sign 1 -n 16 -name n1bv_16 -include n1b.h */ /* * This function contains 72 FP additions, 34 FP multiplications, * (or, 38 additions, 0 multiplications, 34 fused multiply/add), * 54 stack variables, 3 constants, and 32 memory accesses */ #include "n1b.h" static void n1bv_16(const R *ri, const R *ii, R *ro, R *io, stride is, stride os, INT v, INT ivs, INT ovs) { DVK(KP923879532, +0.923879532511286756128183189396788286822416626); DVK(KP414213562, +0.414213562373095048801688724209698078569671875); DVK(KP707106781, +0.707106781186547524400844362104849039284835938); { INT i; const R *xi; R *xo; xi = ii; xo = io; for (i = v; i > 0; i = i - VL, xi = xi + (VL * ivs), xo = xo + (VL * ovs), MAKE_VOLATILE_STRIDE(32, is), MAKE_VOLATILE_STRIDE(32, os)) { V T7, Tu, TF, TB, T13, TL, TO, TX, TC, Te, TP, Th, TQ, Tk, TW; V T16; { V TH, TU, Tz, Tf, TK, TV, TA, TM, Ta, TN, Td, Tg, Ti, Tj; { V T1, T2, T4, T5, To, Tp, Tr, Ts; T1 = LD(&(xi[0]), ivs, &(xi[0])); T2 = LD(&(xi[WS(is, 8)]), ivs, &(xi[0])); T4 = LD(&(xi[WS(is, 4)]), ivs, &(xi[0])); T5 = LD(&(xi[WS(is, 12)]), ivs, &(xi[0])); To = LD(&(xi[WS(is, 2)]), ivs, &(xi[0])); Tp = LD(&(xi[WS(is, 10)]), ivs, &(xi[0])); Tr = LD(&(xi[WS(is, 14)]), ivs, &(xi[0])); Ts = LD(&(xi[WS(is, 6)]), ivs, &(xi[0])); { V T8, TI, Tq, TJ, Tt, T9, Tb, Tc, T3, T6; T8 = LD(&(xi[WS(is, 1)]), ivs, &(xi[WS(is, 1)])); TH = VSUB(T1, T2); T3 = VADD(T1, T2); TU = VSUB(T4, T5); T6 = VADD(T4, T5); TI = VSUB(To, Tp); Tq = VADD(To, Tp); TJ = VSUB(Tr, Ts); Tt = VADD(Tr, Ts); T9 = LD(&(xi[WS(is, 9)]), ivs, &(xi[WS(is, 1)])); Tb = LD(&(xi[WS(is, 5)]), ivs, &(xi[WS(is, 1)])); Tc = LD(&(xi[WS(is, 13)]), ivs, &(xi[WS(is, 1)])); T7 = VSUB(T3, T6); Tz = VADD(T3, T6); Tf = LD(&(xi[WS(is, 15)]), ivs, &(xi[WS(is, 1)])); TK = VADD(TI, TJ); TV = VSUB(TI, TJ); TA = VADD(Tq, Tt); Tu = VSUB(Tq, Tt); TM = VSUB(T8, T9); Ta = VADD(T8, T9); TN = VSUB(Tb, Tc); Td = VADD(Tb, Tc); Tg = LD(&(xi[WS(is, 7)]), ivs, &(xi[WS(is, 1)])); Ti = LD(&(xi[WS(is, 3)]), ivs, &(xi[WS(is, 1)])); Tj = LD(&(xi[WS(is, 11)]), ivs, &(xi[WS(is, 1)])); } } TF = VADD(Tz, TA); TB = VSUB(Tz, TA); T13 = VFNMS(LDK(KP707106781), TK, TH); TL = VFMA(LDK(KP707106781), TK, TH); TO = VFNMS(LDK(KP414213562), TN, TM); TX = VFMA(LDK(KP414213562), TM, TN); TC = VADD(Ta, Td); Te = VSUB(Ta, Td); TP = VSUB(Tf, Tg); Th = VADD(Tf, Tg); TQ = VSUB(Tj, Ti); Tk = VADD(Ti, Tj); TW = VFMA(LDK(KP707106781), TV, TU); T16 = VFNMS(LDK(KP707106781), TV, TU); } { V TY, TR, Tl, TD; TY = VFMA(LDK(KP414213562), TP, TQ); TR = VFNMS(LDK(KP414213562), TQ, TP); Tl = VSUB(Th, Tk); TD = VADD(Th, Tk); { V TS, T17, TZ, T14; TS = VADD(TO, TR); T17 = VSUB(TO, TR); TZ = VSUB(TX, TY); T14 = VADD(TX, TY); { V TE, TG, Tm, Tv; TE = VSUB(TC, TD); TG = VADD(TC, TD); Tm = VADD(Te, Tl); Tv = VSUB(Te, Tl); { V T18, T1a, TT, T11; T18 = VFMA(LDK(KP923879532), T17, T16); T1a = VFNMS(LDK(KP923879532), T17, T16); TT = VFNMS(LDK(KP923879532), TS, TL); T11 = VFMA(LDK(KP923879532), TS, TL); { V T15, T19, T10, T12; T15 = VFNMS(LDK(KP923879532), T14, T13); T19 = VFMA(LDK(KP923879532), T14, T13); T10 = VFNMS(LDK(KP923879532), TZ, TW); T12 = VFMA(LDK(KP923879532), TZ, TW); ST(&(xo[0]), VADD(TF, TG), ovs, &(xo[0])); ST(&(xo[WS(os, 8)]), VSUB(TF, TG), ovs, &(xo[0])); ST(&(xo[WS(os, 4)]), VFMAI(TE, TB), ovs, &(xo[0])); ST(&(xo[WS(os, 12)]), VFNMSI(TE, TB), ovs, &(xo[0])); { V Tw, Ty, Tn, Tx; Tw = VFNMS(LDK(KP707106781), Tv, Tu); Ty = VFMA(LDK(KP707106781), Tv, Tu); Tn = VFNMS(LDK(KP707106781), Tm, T7); Tx = VFMA(LDK(KP707106781), Tm, T7); ST(&(xo[WS(os, 3)]), VFNMSI(T1a, T19), ovs, &(xo[WS(os, 1)])); ST(&(xo[WS(os, 13)]), VFMAI(T1a, T19), ovs, &(xo[WS(os, 1)])); ST(&(xo[WS(os, 11)]), VFNMSI(T18, T15), ovs, &(xo[WS(os, 1)])); ST(&(xo[WS(os, 5)]), VFMAI(T18, T15), ovs, &(xo[WS(os, 1)])); ST(&(xo[WS(os, 15)]), VFNMSI(T12, T11), ovs, &(xo[WS(os, 1)])); ST(&(xo[WS(os, 1)]), VFMAI(T12, T11), ovs, &(xo[WS(os, 1)])); ST(&(xo[WS(os, 9)]), VFMAI(T10, TT), ovs, &(xo[WS(os, 1)])); ST(&(xo[WS(os, 7)]), VFNMSI(T10, TT), ovs, &(xo[WS(os, 1)])); ST(&(xo[WS(os, 2)]), VFMAI(Ty, Tx), ovs, &(xo[0])); ST(&(xo[WS(os, 14)]), VFNMSI(Ty, Tx), ovs, &(xo[0])); ST(&(xo[WS(os, 10)]), VFMAI(Tw, Tn), ovs, &(xo[0])); ST(&(xo[WS(os, 6)]), VFNMSI(Tw, Tn), ovs, &(xo[0])); } } } } } } } } VLEAVE(); } static const kdft_desc desc = { 16, XSIMD_STRING("n1bv_16"), {38, 0, 34, 0}, &GENUS, 0, 0, 0, 0 }; void XSIMD(codelet_n1bv_16) (planner *p) { X(kdft_register) (p, n1bv_16, &desc); } #else /* HAVE_FMA */ /* Generated by: ../../../genfft/gen_notw_c.native -simd -compact -variables 4 -pipeline-latency 8 -sign 1 -n 16 -name n1bv_16 -include n1b.h */ /* * This function contains 72 FP additions, 12 FP multiplications, * (or, 68 additions, 8 multiplications, 4 fused multiply/add), * 30 stack variables, 3 constants, and 32 memory accesses */ #include "n1b.h" static void n1bv_16(const R *ri, const R *ii, R *ro, R *io, stride is, stride os, INT v, INT ivs, INT ovs) { DVK(KP382683432, +0.382683432365089771728459984030398866761344562); DVK(KP923879532, +0.923879532511286756128183189396788286822416626); DVK(KP707106781, +0.707106781186547524400844362104849039284835938); { INT i; const R *xi; R *xo; xi = ii; xo = io; for (i = v; i > 0; i = i - VL, xi = xi + (VL * ivs), xo = xo + (VL * ovs), MAKE_VOLATILE_STRIDE(32, is), MAKE_VOLATILE_STRIDE(32, os)) { V Tp, T13, Tu, TY, Tm, T14, Tv, TU, T7, T16, Tx, TN, Te, T17, Ty; V TQ; { V Tn, To, TX, Ts, Tt, TW; Tn = LD(&(xi[WS(is, 4)]), ivs, &(xi[0])); To = LD(&(xi[WS(is, 12)]), ivs, &(xi[0])); TX = VADD(Tn, To); Ts = LD(&(xi[0]), ivs, &(xi[0])); Tt = LD(&(xi[WS(is, 8)]), ivs, &(xi[0])); TW = VADD(Ts, Tt); Tp = VSUB(Tn, To); T13 = VADD(TW, TX); Tu = VSUB(Ts, Tt); TY = VSUB(TW, TX); } { V Ti, TS, Tl, TT; { V Tg, Th, Tj, Tk; Tg = LD(&(xi[WS(is, 2)]), ivs, &(xi[0])); Th = LD(&(xi[WS(is, 10)]), ivs, &(xi[0])); Ti = VSUB(Tg, Th); TS = VADD(Tg, Th); Tj = LD(&(xi[WS(is, 14)]), ivs, &(xi[0])); Tk = LD(&(xi[WS(is, 6)]), ivs, &(xi[0])); Tl = VSUB(Tj, Tk); TT = VADD(Tj, Tk); } Tm = VMUL(LDK(KP707106781), VSUB(Ti, Tl)); T14 = VADD(TS, TT); Tv = VMUL(LDK(KP707106781), VADD(Ti, Tl)); TU = VSUB(TS, TT); } { V T3, TL, T6, TM; { V T1, T2, T4, T5; T1 = LD(&(xi[WS(is, 1)]), ivs, &(xi[WS(is, 1)])); T2 = LD(&(xi[WS(is, 9)]), ivs, &(xi[WS(is, 1)])); T3 = VSUB(T1, T2); TL = VADD(T1, T2); T4 = LD(&(xi[WS(is, 5)]), ivs, &(xi[WS(is, 1)])); T5 = LD(&(xi[WS(is, 13)]), ivs, &(xi[WS(is, 1)])); T6 = VSUB(T4, T5); TM = VADD(T4, T5); } T7 = VFNMS(LDK(KP382683432), T6, VMUL(LDK(KP923879532), T3)); T16 = VADD(TL, TM); Tx = VFMA(LDK(KP382683432), T3, VMUL(LDK(KP923879532), T6)); TN = VSUB(TL, TM); } { V Ta, TO, Td, TP; { V T8, T9, Tb, Tc; T8 = LD(&(xi[WS(is, 15)]), ivs, &(xi[WS(is, 1)])); T9 = LD(&(xi[WS(is, 7)]), ivs, &(xi[WS(is, 1)])); Ta = VSUB(T8, T9); TO = VADD(T8, T9); Tb = LD(&(xi[WS(is, 3)]), ivs, &(xi[WS(is, 1)])); Tc = LD(&(xi[WS(is, 11)]), ivs, &(xi[WS(is, 1)])); Td = VSUB(Tb, Tc); TP = VADD(Tb, Tc); } Te = VFMA(LDK(KP923879532), Ta, VMUL(LDK(KP382683432), Td)); T17 = VADD(TO, TP); Ty = VFNMS(LDK(KP382683432), Ta, VMUL(LDK(KP923879532), Td)); TQ = VSUB(TO, TP); } { V T15, T18, T19, T1a; T15 = VSUB(T13, T14); T18 = VBYI(VSUB(T16, T17)); ST(&(xo[WS(os, 12)]), VSUB(T15, T18), ovs, &(xo[0])); ST(&(xo[WS(os, 4)]), VADD(T15, T18), ovs, &(xo[0])); T19 = VADD(T13, T14); T1a = VADD(T16, T17); ST(&(xo[WS(os, 8)]), VSUB(T19, T1a), ovs, &(xo[0])); ST(&(xo[0]), VADD(T19, T1a), ovs, &(xo[0])); } { V TV, T11, T10, T12, TR, TZ; TR = VMUL(LDK(KP707106781), VSUB(TN, TQ)); TV = VBYI(VSUB(TR, TU)); T11 = VBYI(VADD(TU, TR)); TZ = VMUL(LDK(KP707106781), VADD(TN, TQ)); T10 = VSUB(TY, TZ); T12 = VADD(TY, TZ); ST(&(xo[WS(os, 6)]), VADD(TV, T10), ovs, &(xo[0])); ST(&(xo[WS(os, 14)]), VSUB(T12, T11), ovs, &(xo[0])); ST(&(xo[WS(os, 10)]), VSUB(T10, TV), ovs, &(xo[0])); ST(&(xo[WS(os, 2)]), VADD(T11, T12), ovs, &(xo[0])); } { V Tr, TB, TA, TC; { V Tf, Tq, Tw, Tz; Tf = VSUB(T7, Te); Tq = VSUB(Tm, Tp); Tr = VBYI(VSUB(Tf, Tq)); TB = VBYI(VADD(Tq, Tf)); Tw = VSUB(Tu, Tv); Tz = VSUB(Tx, Ty); TA = VSUB(Tw, Tz); TC = VADD(Tw, Tz); } ST(&(xo[WS(os, 5)]), VADD(Tr, TA), ovs, &(xo[WS(os, 1)])); ST(&(xo[WS(os, 13)]), VSUB(TC, TB), ovs, &(xo[WS(os, 1)])); ST(&(xo[WS(os, 11)]), VSUB(TA, Tr), ovs, &(xo[WS(os, 1)])); ST(&(xo[WS(os, 3)]), VADD(TB, TC), ovs, &(xo[WS(os, 1)])); } { V TF, TJ, TI, TK; { V TD, TE, TG, TH; TD = VADD(Tu, Tv); TE = VADD(T7, Te); TF = VADD(TD, TE); TJ = VSUB(TD, TE); TG = VADD(Tp, Tm); TH = VADD(Tx, Ty); TI = VBYI(VADD(TG, TH)); TK = VBYI(VSUB(TH, TG)); } ST(&(xo[WS(os, 15)]), VSUB(TF, TI), ovs, &(xo[WS(os, 1)])); ST(&(xo[WS(os, 7)]), VADD(TJ, TK), ovs, &(xo[WS(os, 1)])); ST(&(xo[WS(os, 1)]), VADD(TF, TI), ovs, &(xo[WS(os, 1)])); ST(&(xo[WS(os, 9)]), VSUB(TJ, TK), ovs, &(xo[WS(os, 1)])); } } } VLEAVE(); } static const kdft_desc desc = { 16, XSIMD_STRING("n1bv_16"), {68, 8, 4, 0}, &GENUS, 0, 0, 0, 0 }; void XSIMD(codelet_n1bv_16) (planner *p) { X(kdft_register) (p, n1bv_16, &desc); } #endif /* HAVE_FMA */ fftw-3.3.3/dft/simd/common/n1bv_32.c0000644000175000017500000006016012054410364013730 00000000000000/* * Copyright (c) 2003, 2007-11 Matteo Frigo * Copyright (c) 2003, 2007-11 Massachusetts Institute of Technology * * This program is free software; you can redistribute it and/or modify * it under the terms of the GNU General Public License as published by * the Free Software Foundation; either version 2 of the License, or * (at your option) any later version. * * This program is distributed in the hope that it will be useful, * but WITHOUT ANY WARRANTY; without even the implied warranty of * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the * GNU General Public License for more details. * * You should have received a copy of the GNU General Public License * along with this program; if not, write to the Free Software * Foundation, Inc., 51 Franklin Street, Fifth Floor, Boston, MA 02110-1301 USA * */ /* This file was automatically generated --- DO NOT EDIT */ /* Generated on Sun Nov 25 07:37:04 EST 2012 */ #include "codelet-dft.h" #ifdef HAVE_FMA /* Generated by: ../../../genfft/gen_notw_c.native -fma -reorder-insns -schedule-for-pipeline -simd -compact -variables 4 -pipeline-latency 8 -sign 1 -n 32 -name n1bv_32 -include n1b.h */ /* * This function contains 186 FP additions, 98 FP multiplications, * (or, 88 additions, 0 multiplications, 98 fused multiply/add), * 104 stack variables, 7 constants, and 64 memory accesses */ #include "n1b.h" static void n1bv_32(const R *ri, const R *ii, R *ro, R *io, stride is, stride os, INT v, INT ivs, INT ovs) { DVK(KP980785280, +0.980785280403230449126182236134239036973933731); DVK(KP831469612, +0.831469612302545237078788377617905756738560812); DVK(KP198912367, +0.198912367379658006911597622644676228597850501); DVK(KP668178637, +0.668178637919298919997757686523080761552472251); DVK(KP923879532, +0.923879532511286756128183189396788286822416626); DVK(KP414213562, +0.414213562373095048801688724209698078569671875); DVK(KP707106781, +0.707106781186547524400844362104849039284835938); { INT i; const R *xi; R *xo; xi = ii; xo = io; for (i = v; i > 0; i = i - VL, xi = xi + (VL * ivs), xo = xo + (VL * ovs), MAKE_VOLATILE_STRIDE(64, is), MAKE_VOLATILE_STRIDE(64, os)) { V T1h, Tr, T1a, T1k, TI, T1b, T1L, T1P, T1I, T1G, T1O, T1Q, T1H, T1z, T1c; V TZ; { V T2x, T1T, T2K, T1W, T1p, Tb, T1A, T16, Tu, TF, T2O, T2H, T2b, T2t, TY; V T1w, TT, T1v, T20, T2C, Tj, Te, T2e, To, T2i, T23, T2D, TB, TG, Th; V T2f, Tk; { V TL, TW, TP, TQ, T2F, T27, T28, TO; { V T1, T2, T12, T13, T4, T5, T7, T8; T1 = LD(&(xi[0]), ivs, &(xi[0])); T2 = LD(&(xi[WS(is, 16)]), ivs, &(xi[0])); T12 = LD(&(xi[WS(is, 8)]), ivs, &(xi[0])); T13 = LD(&(xi[WS(is, 24)]), ivs, &(xi[0])); T4 = LD(&(xi[WS(is, 4)]), ivs, &(xi[0])); T5 = LD(&(xi[WS(is, 20)]), ivs, &(xi[0])); T7 = LD(&(xi[WS(is, 28)]), ivs, &(xi[0])); T8 = LD(&(xi[WS(is, 12)]), ivs, &(xi[0])); { V TM, T25, T26, TN; { V TJ, T3, T14, T1U, T6, T1V, T9, TK, TU, TV, T1R, T1S, Ta, T15; TJ = LD(&(xi[WS(is, 31)]), ivs, &(xi[WS(is, 1)])); T1R = VADD(T1, T2); T3 = VSUB(T1, T2); T1S = VADD(T12, T13); T14 = VSUB(T12, T13); T1U = VADD(T4, T5); T6 = VSUB(T4, T5); T1V = VADD(T7, T8); T9 = VSUB(T7, T8); TK = LD(&(xi[WS(is, 15)]), ivs, &(xi[WS(is, 1)])); TU = LD(&(xi[WS(is, 23)]), ivs, &(xi[WS(is, 1)])); T2x = VSUB(T1R, T1S); T1T = VADD(T1R, T1S); TV = LD(&(xi[WS(is, 7)]), ivs, &(xi[WS(is, 1)])); TM = LD(&(xi[WS(is, 3)]), ivs, &(xi[WS(is, 1)])); T2K = VSUB(T1U, T1V); T1W = VADD(T1U, T1V); Ta = VADD(T6, T9); T15 = VSUB(T6, T9); T25 = VADD(TJ, TK); TL = VSUB(TJ, TK); T26 = VADD(TV, TU); TW = VSUB(TU, TV); TN = LD(&(xi[WS(is, 19)]), ivs, &(xi[WS(is, 1)])); TP = LD(&(xi[WS(is, 27)]), ivs, &(xi[WS(is, 1)])); T1p = VFNMS(LDK(KP707106781), Ta, T3); Tb = VFMA(LDK(KP707106781), Ta, T3); T1A = VFNMS(LDK(KP707106781), T15, T14); T16 = VFMA(LDK(KP707106781), T15, T14); TQ = LD(&(xi[WS(is, 11)]), ivs, &(xi[WS(is, 1)])); } T2F = VSUB(T25, T26); T27 = VADD(T25, T26); T28 = VADD(TM, TN); TO = VSUB(TM, TN); } } { V Ty, T21, Tx, Tz, T1Y, T1Z; { V Ts, Tt, TD, T29, TR, TE, Tv, Tw; Ts = LD(&(xi[WS(is, 1)]), ivs, &(xi[WS(is, 1)])); Tt = LD(&(xi[WS(is, 17)]), ivs, &(xi[WS(is, 1)])); TD = LD(&(xi[WS(is, 9)]), ivs, &(xi[WS(is, 1)])); T29 = VADD(TP, TQ); TR = VSUB(TP, TQ); TE = LD(&(xi[WS(is, 25)]), ivs, &(xi[WS(is, 1)])); Tv = LD(&(xi[WS(is, 5)]), ivs, &(xi[WS(is, 1)])); Tw = LD(&(xi[WS(is, 21)]), ivs, &(xi[WS(is, 1)])); Ty = LD(&(xi[WS(is, 29)]), ivs, &(xi[WS(is, 1)])); T1Y = VADD(Ts, Tt); Tu = VSUB(Ts, Tt); { V T2G, T2a, TX, TS; T2G = VSUB(T29, T28); T2a = VADD(T28, T29); TX = VSUB(TR, TO); TS = VADD(TO, TR); T1Z = VADD(TD, TE); TF = VSUB(TD, TE); T21 = VADD(Tv, Tw); Tx = VSUB(Tv, Tw); T2O = VFMA(LDK(KP414213562), T2F, T2G); T2H = VFNMS(LDK(KP414213562), T2G, T2F); T2b = VSUB(T27, T2a); T2t = VADD(T27, T2a); TY = VFMA(LDK(KP707106781), TX, TW); T1w = VFNMS(LDK(KP707106781), TX, TW); TT = VFMA(LDK(KP707106781), TS, TL); T1v = VFNMS(LDK(KP707106781), TS, TL); Tz = LD(&(xi[WS(is, 13)]), ivs, &(xi[WS(is, 1)])); } } T20 = VADD(T1Y, T1Z); T2C = VSUB(T1Y, T1Z); { V Tc, Td, Tm, Tn; Tc = LD(&(xi[WS(is, 2)]), ivs, &(xi[0])); Td = LD(&(xi[WS(is, 18)]), ivs, &(xi[0])); Tm = LD(&(xi[WS(is, 22)]), ivs, &(xi[0])); Tn = LD(&(xi[WS(is, 6)]), ivs, &(xi[0])); { V Tf, TA, T22, Tg; Tf = LD(&(xi[WS(is, 10)]), ivs, &(xi[0])); TA = VSUB(Ty, Tz); T22 = VADD(Ty, Tz); Tg = LD(&(xi[WS(is, 26)]), ivs, &(xi[0])); Tj = LD(&(xi[WS(is, 30)]), ivs, &(xi[0])); Te = VSUB(Tc, Td); T2e = VADD(Tc, Td); To = VSUB(Tm, Tn); T2i = VADD(Tn, Tm); T23 = VADD(T21, T22); T2D = VSUB(T21, T22); TB = VADD(Tx, TA); TG = VSUB(Tx, TA); Th = VSUB(Tf, Tg); T2f = VADD(Tf, Tg); Tk = LD(&(xi[WS(is, 14)]), ivs, &(xi[0])); } } } } { V T1t, TH, T1s, TC, T2P, T2U, T2n, T2d, T2w, T2u, T1q, T19, T1B, Tq, T2W; V T2M, T2B, T2T, T2v, T2r, T2o, T2m, T2X, T2I; { V T1X, T2p, T2E, T2N, T2s, T2y, T2g, T17, Ti, T2h, Tl, T2c, T2l, T24; T1X = VSUB(T1T, T1W); T2p = VADD(T1T, T1W); T2E = VFNMS(LDK(KP414213562), T2D, T2C); T2N = VFMA(LDK(KP414213562), T2C, T2D); T2s = VADD(T20, T23); T24 = VSUB(T20, T23); T1t = VFNMS(LDK(KP707106781), TG, TF); TH = VFMA(LDK(KP707106781), TG, TF); T1s = VFNMS(LDK(KP707106781), TB, Tu); TC = VFMA(LDK(KP707106781), TB, Tu); T2y = VSUB(T2e, T2f); T2g = VADD(T2e, T2f); T17 = VFMA(LDK(KP414213562), Te, Th); Ti = VFNMS(LDK(KP414213562), Th, Te); T2h = VADD(Tj, Tk); Tl = VSUB(Tj, Tk); T2c = VADD(T24, T2b); T2l = VSUB(T24, T2b); { V T2L, T2A, T2q, T2k; T2P = VSUB(T2N, T2O); T2U = VADD(T2N, T2O); { V T2z, T2j, T18, Tp; T2z = VSUB(T2h, T2i); T2j = VADD(T2h, T2i); T18 = VFMA(LDK(KP414213562), Tl, To); Tp = VFNMS(LDK(KP414213562), To, Tl); T2n = VFMA(LDK(KP707106781), T2c, T1X); T2d = VFNMS(LDK(KP707106781), T2c, T1X); T2w = VADD(T2s, T2t); T2u = VSUB(T2s, T2t); T2L = VSUB(T2y, T2z); T2A = VADD(T2y, T2z); T2q = VADD(T2g, T2j); T2k = VSUB(T2g, T2j); T1q = VADD(T17, T18); T19 = VSUB(T17, T18); T1B = VSUB(Ti, Tp); Tq = VADD(Ti, Tp); } T2W = VFNMS(LDK(KP707106781), T2L, T2K); T2M = VFMA(LDK(KP707106781), T2L, T2K); T2B = VFMA(LDK(KP707106781), T2A, T2x); T2T = VFNMS(LDK(KP707106781), T2A, T2x); T2v = VADD(T2p, T2q); T2r = VSUB(T2p, T2q); T2o = VFMA(LDK(KP707106781), T2l, T2k); T2m = VFNMS(LDK(KP707106781), T2l, T2k); T2X = VSUB(T2E, T2H); T2I = VADD(T2E, T2H); } } { V T2V, T2Z, T2Y, T30, T2R, T2J; T2V = VFNMS(LDK(KP923879532), T2U, T2T); T2Z = VFMA(LDK(KP923879532), T2U, T2T); ST(&(xo[WS(os, 16)]), VSUB(T2v, T2w), ovs, &(xo[0])); ST(&(xo[0]), VADD(T2v, T2w), ovs, &(xo[0])); ST(&(xo[WS(os, 8)]), VFMAI(T2u, T2r), ovs, &(xo[0])); ST(&(xo[WS(os, 24)]), VFNMSI(T2u, T2r), ovs, &(xo[0])); ST(&(xo[WS(os, 4)]), VFMAI(T2o, T2n), ovs, &(xo[0])); ST(&(xo[WS(os, 28)]), VFNMSI(T2o, T2n), ovs, &(xo[0])); ST(&(xo[WS(os, 20)]), VFMAI(T2m, T2d), ovs, &(xo[0])); ST(&(xo[WS(os, 12)]), VFNMSI(T2m, T2d), ovs, &(xo[0])); T2Y = VFMA(LDK(KP923879532), T2X, T2W); T30 = VFNMS(LDK(KP923879532), T2X, T2W); T2R = VFMA(LDK(KP923879532), T2I, T2B); T2J = VFNMS(LDK(KP923879532), T2I, T2B); { V T1J, T1r, T1C, T1M, T2S, T2Q, T1u, T1D, T1E, T1x; T1J = VFNMS(LDK(KP923879532), T1q, T1p); T1r = VFMA(LDK(KP923879532), T1q, T1p); T1C = VFNMS(LDK(KP923879532), T1B, T1A); T1M = VFMA(LDK(KP923879532), T1B, T1A); ST(&(xo[WS(os, 6)]), VFNMSI(T30, T2Z), ovs, &(xo[0])); ST(&(xo[WS(os, 26)]), VFMAI(T30, T2Z), ovs, &(xo[0])); ST(&(xo[WS(os, 22)]), VFNMSI(T2Y, T2V), ovs, &(xo[0])); ST(&(xo[WS(os, 10)]), VFMAI(T2Y, T2V), ovs, &(xo[0])); T2S = VFMA(LDK(KP923879532), T2P, T2M); T2Q = VFNMS(LDK(KP923879532), T2P, T2M); T1u = VFMA(LDK(KP668178637), T1t, T1s); T1D = VFNMS(LDK(KP668178637), T1s, T1t); T1E = VFNMS(LDK(KP668178637), T1v, T1w); T1x = VFMA(LDK(KP668178637), T1w, T1v); { V T1K, T1F, T1N, T1y; T1h = VFNMS(LDK(KP923879532), Tq, Tb); Tr = VFMA(LDK(KP923879532), Tq, Tb); ST(&(xo[WS(os, 30)]), VFNMSI(T2S, T2R), ovs, &(xo[0])); ST(&(xo[WS(os, 2)]), VFMAI(T2S, T2R), ovs, &(xo[0])); ST(&(xo[WS(os, 18)]), VFMAI(T2Q, T2J), ovs, &(xo[0])); ST(&(xo[WS(os, 14)]), VFNMSI(T2Q, T2J), ovs, &(xo[0])); T1K = VADD(T1D, T1E); T1F = VSUB(T1D, T1E); T1N = VSUB(T1u, T1x); T1y = VADD(T1u, T1x); T1a = VFMA(LDK(KP923879532), T19, T16); T1k = VFNMS(LDK(KP923879532), T19, T16); TI = VFNMS(LDK(KP198912367), TH, TC); T1b = VFMA(LDK(KP198912367), TC, TH); T1L = VFMA(LDK(KP831469612), T1K, T1J); T1P = VFNMS(LDK(KP831469612), T1K, T1J); T1I = VFMA(LDK(KP831469612), T1F, T1C); T1G = VFNMS(LDK(KP831469612), T1F, T1C); T1O = VFNMS(LDK(KP831469612), T1N, T1M); T1Q = VFMA(LDK(KP831469612), T1N, T1M); T1H = VFMA(LDK(KP831469612), T1y, T1r); T1z = VFNMS(LDK(KP831469612), T1y, T1r); T1c = VFMA(LDK(KP198912367), TT, TY); TZ = VFNMS(LDK(KP198912367), TY, TT); } } } } } { V T1d, T1i, T10, T1l; ST(&(xo[WS(os, 21)]), VFMAI(T1O, T1L), ovs, &(xo[WS(os, 1)])); ST(&(xo[WS(os, 11)]), VFNMSI(T1O, T1L), ovs, &(xo[WS(os, 1)])); ST(&(xo[WS(os, 27)]), VFNMSI(T1Q, T1P), ovs, &(xo[WS(os, 1)])); ST(&(xo[WS(os, 5)]), VFMAI(T1Q, T1P), ovs, &(xo[WS(os, 1)])); ST(&(xo[WS(os, 29)]), VFMAI(T1I, T1H), ovs, &(xo[WS(os, 1)])); ST(&(xo[WS(os, 3)]), VFNMSI(T1I, T1H), ovs, &(xo[WS(os, 1)])); ST(&(xo[WS(os, 13)]), VFMAI(T1G, T1z), ovs, &(xo[WS(os, 1)])); ST(&(xo[WS(os, 19)]), VFNMSI(T1G, T1z), ovs, &(xo[WS(os, 1)])); T1d = VSUB(T1b, T1c); T1i = VADD(T1b, T1c); T10 = VADD(TI, TZ); T1l = VSUB(TI, TZ); { V T1n, T1j, T1e, T1g, T1o, T1m, T11, T1f; T1n = VFMA(LDK(KP980785280), T1i, T1h); T1j = VFNMS(LDK(KP980785280), T1i, T1h); T1e = VFNMS(LDK(KP980785280), T1d, T1a); T1g = VFMA(LDK(KP980785280), T1d, T1a); T1o = VFNMS(LDK(KP980785280), T1l, T1k); T1m = VFMA(LDK(KP980785280), T1l, T1k); T11 = VFNMS(LDK(KP980785280), T10, Tr); T1f = VFMA(LDK(KP980785280), T10, Tr); ST(&(xo[WS(os, 23)]), VFNMSI(T1m, T1j), ovs, &(xo[WS(os, 1)])); ST(&(xo[WS(os, 9)]), VFMAI(T1m, T1j), ovs, &(xo[WS(os, 1)])); ST(&(xo[WS(os, 25)]), VFMAI(T1o, T1n), ovs, &(xo[WS(os, 1)])); ST(&(xo[WS(os, 7)]), VFNMSI(T1o, T1n), ovs, &(xo[WS(os, 1)])); ST(&(xo[WS(os, 1)]), VFMAI(T1g, T1f), ovs, &(xo[WS(os, 1)])); ST(&(xo[WS(os, 31)]), VFNMSI(T1g, T1f), ovs, &(xo[WS(os, 1)])); ST(&(xo[WS(os, 17)]), VFMAI(T1e, T11), ovs, &(xo[WS(os, 1)])); ST(&(xo[WS(os, 15)]), VFNMSI(T1e, T11), ovs, &(xo[WS(os, 1)])); } } } } VLEAVE(); } static const kdft_desc desc = { 32, XSIMD_STRING("n1bv_32"), {88, 0, 98, 0}, &GENUS, 0, 0, 0, 0 }; void XSIMD(codelet_n1bv_32) (planner *p) { X(kdft_register) (p, n1bv_32, &desc); } #else /* HAVE_FMA */ /* Generated by: ../../../genfft/gen_notw_c.native -simd -compact -variables 4 -pipeline-latency 8 -sign 1 -n 32 -name n1bv_32 -include n1b.h */ /* * This function contains 186 FP additions, 42 FP multiplications, * (or, 170 additions, 26 multiplications, 16 fused multiply/add), * 58 stack variables, 7 constants, and 64 memory accesses */ #include "n1b.h" static void n1bv_32(const R *ri, const R *ii, R *ro, R *io, stride is, stride os, INT v, INT ivs, INT ovs) { DVK(KP555570233, +0.555570233019602224742830813948532874374937191); DVK(KP831469612, +0.831469612302545237078788377617905756738560812); DVK(KP195090322, +0.195090322016128267848284868477022240927691618); DVK(KP980785280, +0.980785280403230449126182236134239036973933731); DVK(KP382683432, +0.382683432365089771728459984030398866761344562); DVK(KP923879532, +0.923879532511286756128183189396788286822416626); DVK(KP707106781, +0.707106781186547524400844362104849039284835938); { INT i; const R *xi; R *xo; xi = ii; xo = io; for (i = v; i > 0; i = i - VL, xi = xi + (VL * ivs), xo = xo + (VL * ovs), MAKE_VOLATILE_STRIDE(64, is), MAKE_VOLATILE_STRIDE(64, os)) { V T2f, T2k, T2N, T2M, T19, T1B, Tb, T1p, TT, T1v, TY, T1w, T2E, T2F, T2G; V T24, T2o, TC, T1s, TH, T1t, T2B, T2C, T2D, T1X, T2n, T2I, T2J, Tq, T1A; V T14, T1q, T2c, T2l; { V T3, T2i, T18, T2j, T6, T2d, T9, T2e, T15, Ta; { V T1, T2, T16, T17; T1 = LD(&(xi[0]), ivs, &(xi[0])); T2 = LD(&(xi[WS(is, 16)]), ivs, &(xi[0])); T3 = VSUB(T1, T2); T2i = VADD(T1, T2); T16 = LD(&(xi[WS(is, 8)]), ivs, &(xi[0])); T17 = LD(&(xi[WS(is, 24)]), ivs, &(xi[0])); T18 = VSUB(T16, T17); T2j = VADD(T16, T17); } { V T4, T5, T7, T8; T4 = LD(&(xi[WS(is, 4)]), ivs, &(xi[0])); T5 = LD(&(xi[WS(is, 20)]), ivs, &(xi[0])); T6 = VSUB(T4, T5); T2d = VADD(T4, T5); T7 = LD(&(xi[WS(is, 28)]), ivs, &(xi[0])); T8 = LD(&(xi[WS(is, 12)]), ivs, &(xi[0])); T9 = VSUB(T7, T8); T2e = VADD(T7, T8); } T2f = VSUB(T2d, T2e); T2k = VSUB(T2i, T2j); T2N = VADD(T2d, T2e); T2M = VADD(T2i, T2j); T15 = VMUL(LDK(KP707106781), VSUB(T6, T9)); T19 = VSUB(T15, T18); T1B = VADD(T18, T15); Ta = VMUL(LDK(KP707106781), VADD(T6, T9)); Tb = VSUB(T3, Ta); T1p = VADD(T3, Ta); } { V TL, T21, TW, T1Y, TO, T22, TS, T1Z; { V TJ, TK, TU, TV; TJ = LD(&(xi[WS(is, 3)]), ivs, &(xi[WS(is, 1)])); TK = LD(&(xi[WS(is, 19)]), ivs, &(xi[WS(is, 1)])); TL = VSUB(TJ, TK); T21 = VADD(TJ, TK); TU = LD(&(xi[WS(is, 31)]), ivs, &(xi[WS(is, 1)])); TV = LD(&(xi[WS(is, 15)]), ivs, &(xi[WS(is, 1)])); TW = VSUB(TU, TV); T1Y = VADD(TU, TV); } { V TM, TN, TQ, TR; TM = LD(&(xi[WS(is, 27)]), ivs, &(xi[WS(is, 1)])); TN = LD(&(xi[WS(is, 11)]), ivs, &(xi[WS(is, 1)])); TO = VSUB(TM, TN); T22 = VADD(TM, TN); TQ = LD(&(xi[WS(is, 7)]), ivs, &(xi[WS(is, 1)])); TR = LD(&(xi[WS(is, 23)]), ivs, &(xi[WS(is, 1)])); TS = VSUB(TQ, TR); T1Z = VADD(TQ, TR); } { V TP, TX, T20, T23; TP = VMUL(LDK(KP707106781), VSUB(TL, TO)); TT = VSUB(TP, TS); T1v = VADD(TS, TP); TX = VMUL(LDK(KP707106781), VADD(TL, TO)); TY = VSUB(TW, TX); T1w = VADD(TW, TX); T2E = VADD(T1Y, T1Z); T2F = VADD(T21, T22); T2G = VSUB(T2E, T2F); T20 = VSUB(T1Y, T1Z); T23 = VSUB(T21, T22); T24 = VFMA(LDK(KP923879532), T20, VMUL(LDK(KP382683432), T23)); T2o = VFNMS(LDK(KP382683432), T20, VMUL(LDK(KP923879532), T23)); } } { V Tu, T1U, TF, T1R, Tx, T1V, TB, T1S; { V Ts, Tt, TD, TE; Ts = LD(&(xi[WS(is, 5)]), ivs, &(xi[WS(is, 1)])); Tt = LD(&(xi[WS(is, 21)]), ivs, &(xi[WS(is, 1)])); Tu = VSUB(Ts, Tt); T1U = VADD(Ts, Tt); TD = LD(&(xi[WS(is, 1)]), ivs, &(xi[WS(is, 1)])); TE = LD(&(xi[WS(is, 17)]), ivs, &(xi[WS(is, 1)])); TF = VSUB(TD, TE); T1R = VADD(TD, TE); } { V Tv, Tw, Tz, TA; Tv = LD(&(xi[WS(is, 29)]), ivs, &(xi[WS(is, 1)])); Tw = LD(&(xi[WS(is, 13)]), ivs, &(xi[WS(is, 1)])); Tx = VSUB(Tv, Tw); T1V = VADD(Tv, Tw); Tz = LD(&(xi[WS(is, 9)]), ivs, &(xi[WS(is, 1)])); TA = LD(&(xi[WS(is, 25)]), ivs, &(xi[WS(is, 1)])); TB = VSUB(Tz, TA); T1S = VADD(Tz, TA); } { V Ty, TG, T1T, T1W; Ty = VMUL(LDK(KP707106781), VSUB(Tu, Tx)); TC = VSUB(Ty, TB); T1s = VADD(TB, Ty); TG = VMUL(LDK(KP707106781), VADD(Tu, Tx)); TH = VSUB(TF, TG); T1t = VADD(TF, TG); T2B = VADD(T1R, T1S); T2C = VADD(T1U, T1V); T2D = VSUB(T2B, T2C); T1T = VSUB(T1R, T1S); T1W = VSUB(T1U, T1V); T1X = VFNMS(LDK(KP382683432), T1W, VMUL(LDK(KP923879532), T1T)); T2n = VFMA(LDK(KP382683432), T1T, VMUL(LDK(KP923879532), T1W)); } } { V Te, T26, To, T29, Th, T27, Tl, T2a, Ti, Tp; { V Tc, Td, Tm, Tn; Tc = LD(&(xi[WS(is, 2)]), ivs, &(xi[0])); Td = LD(&(xi[WS(is, 18)]), ivs, &(xi[0])); Te = VSUB(Tc, Td); T26 = VADD(Tc, Td); Tm = LD(&(xi[WS(is, 30)]), ivs, &(xi[0])); Tn = LD(&(xi[WS(is, 14)]), ivs, &(xi[0])); To = VSUB(Tm, Tn); T29 = VADD(Tm, Tn); } { V Tf, Tg, Tj, Tk; Tf = LD(&(xi[WS(is, 10)]), ivs, &(xi[0])); Tg = LD(&(xi[WS(is, 26)]), ivs, &(xi[0])); Th = VSUB(Tf, Tg); T27 = VADD(Tf, Tg); Tj = LD(&(xi[WS(is, 6)]), ivs, &(xi[0])); Tk = LD(&(xi[WS(is, 22)]), ivs, &(xi[0])); Tl = VSUB(Tj, Tk); T2a = VADD(Tj, Tk); } T2I = VADD(T26, T27); T2J = VADD(T29, T2a); Ti = VFMA(LDK(KP382683432), Te, VMUL(LDK(KP923879532), Th)); Tp = VFNMS(LDK(KP382683432), To, VMUL(LDK(KP923879532), Tl)); Tq = VSUB(Ti, Tp); T1A = VADD(Ti, Tp); { V T12, T13, T28, T2b; T12 = VFNMS(LDK(KP382683432), Th, VMUL(LDK(KP923879532), Te)); T13 = VFMA(LDK(KP923879532), To, VMUL(LDK(KP382683432), Tl)); T14 = VSUB(T12, T13); T1q = VADD(T12, T13); T28 = VSUB(T26, T27); T2b = VSUB(T29, T2a); T2c = VMUL(LDK(KP707106781), VSUB(T28, T2b)); T2l = VMUL(LDK(KP707106781), VADD(T28, T2b)); } } { V T2L, T2R, T2Q, T2S; { V T2H, T2K, T2O, T2P; T2H = VMUL(LDK(KP707106781), VSUB(T2D, T2G)); T2K = VSUB(T2I, T2J); T2L = VBYI(VSUB(T2H, T2K)); T2R = VBYI(VADD(T2K, T2H)); T2O = VSUB(T2M, T2N); T2P = VMUL(LDK(KP707106781), VADD(T2D, T2G)); T2Q = VSUB(T2O, T2P); T2S = VADD(T2O, T2P); } ST(&(xo[WS(os, 12)]), VADD(T2L, T2Q), ovs, &(xo[0])); ST(&(xo[WS(os, 28)]), VSUB(T2S, T2R), ovs, &(xo[0])); ST(&(xo[WS(os, 20)]), VSUB(T2Q, T2L), ovs, &(xo[0])); ST(&(xo[WS(os, 4)]), VADD(T2R, T2S), ovs, &(xo[0])); } { V T2h, T2r, T2q, T2s; { V T25, T2g, T2m, T2p; T25 = VSUB(T1X, T24); T2g = VSUB(T2c, T2f); T2h = VBYI(VSUB(T25, T2g)); T2r = VBYI(VADD(T2g, T25)); T2m = VSUB(T2k, T2l); T2p = VSUB(T2n, T2o); T2q = VSUB(T2m, T2p); T2s = VADD(T2m, T2p); } ST(&(xo[WS(os, 10)]), VADD(T2h, T2q), ovs, &(xo[0])); ST(&(xo[WS(os, 26)]), VSUB(T2s, T2r), ovs, &(xo[0])); ST(&(xo[WS(os, 22)]), VSUB(T2q, T2h), ovs, &(xo[0])); ST(&(xo[WS(os, 6)]), VADD(T2r, T2s), ovs, &(xo[0])); } { V T2V, T2Z, T2Y, T30; { V T2T, T2U, T2W, T2X; T2T = VADD(T2M, T2N); T2U = VADD(T2I, T2J); T2V = VSUB(T2T, T2U); T2Z = VADD(T2T, T2U); T2W = VADD(T2B, T2C); T2X = VADD(T2E, T2F); T2Y = VBYI(VSUB(T2W, T2X)); T30 = VADD(T2W, T2X); } ST(&(xo[WS(os, 24)]), VSUB(T2V, T2Y), ovs, &(xo[0])); ST(&(xo[0]), VADD(T2Z, T30), ovs, &(xo[0])); ST(&(xo[WS(os, 8)]), VADD(T2V, T2Y), ovs, &(xo[0])); ST(&(xo[WS(os, 16)]), VSUB(T2Z, T30), ovs, &(xo[0])); } { V T2v, T2z, T2y, T2A; { V T2t, T2u, T2w, T2x; T2t = VADD(T2k, T2l); T2u = VADD(T1X, T24); T2v = VADD(T2t, T2u); T2z = VSUB(T2t, T2u); T2w = VADD(T2f, T2c); T2x = VADD(T2n, T2o); T2y = VBYI(VADD(T2w, T2x)); T2A = VBYI(VSUB(T2x, T2w)); } ST(&(xo[WS(os, 30)]), VSUB(T2v, T2y), ovs, &(xo[0])); ST(&(xo[WS(os, 14)]), VADD(T2z, T2A), ovs, &(xo[0])); ST(&(xo[WS(os, 2)]), VADD(T2v, T2y), ovs, &(xo[0])); ST(&(xo[WS(os, 18)]), VSUB(T2z, T2A), ovs, &(xo[0])); } { V T1r, T1C, T1M, T1K, T1F, T1N, T1y, T1J; T1r = VSUB(T1p, T1q); T1C = VSUB(T1A, T1B); T1M = VADD(T1p, T1q); T1K = VADD(T1B, T1A); { V T1D, T1E, T1u, T1x; T1D = VFNMS(LDK(KP195090322), T1s, VMUL(LDK(KP980785280), T1t)); T1E = VFMA(LDK(KP195090322), T1v, VMUL(LDK(KP980785280), T1w)); T1F = VSUB(T1D, T1E); T1N = VADD(T1D, T1E); T1u = VFMA(LDK(KP980785280), T1s, VMUL(LDK(KP195090322), T1t)); T1x = VFNMS(LDK(KP195090322), T1w, VMUL(LDK(KP980785280), T1v)); T1y = VSUB(T1u, T1x); T1J = VADD(T1u, T1x); } { V T1z, T1G, T1P, T1Q; T1z = VADD(T1r, T1y); T1G = VBYI(VADD(T1C, T1F)); ST(&(xo[WS(os, 25)]), VSUB(T1z, T1G), ovs, &(xo[WS(os, 1)])); ST(&(xo[WS(os, 7)]), VADD(T1z, T1G), ovs, &(xo[WS(os, 1)])); T1P = VBYI(VADD(T1K, T1J)); T1Q = VADD(T1M, T1N); ST(&(xo[WS(os, 1)]), VADD(T1P, T1Q), ovs, &(xo[WS(os, 1)])); ST(&(xo[WS(os, 31)]), VSUB(T1Q, T1P), ovs, &(xo[WS(os, 1)])); } { V T1H, T1I, T1L, T1O; T1H = VSUB(T1r, T1y); T1I = VBYI(VSUB(T1F, T1C)); ST(&(xo[WS(os, 23)]), VSUB(T1H, T1I), ovs, &(xo[WS(os, 1)])); ST(&(xo[WS(os, 9)]), VADD(T1H, T1I), ovs, &(xo[WS(os, 1)])); T1L = VBYI(VSUB(T1J, T1K)); T1O = VSUB(T1M, T1N); ST(&(xo[WS(os, 15)]), VADD(T1L, T1O), ovs, &(xo[WS(os, 1)])); ST(&(xo[WS(os, 17)]), VSUB(T1O, T1L), ovs, &(xo[WS(os, 1)])); } } { V Tr, T1a, T1k, T1i, T1d, T1l, T10, T1h; Tr = VSUB(Tb, Tq); T1a = VSUB(T14, T19); T1k = VADD(Tb, Tq); T1i = VADD(T19, T14); { V T1b, T1c, TI, TZ; T1b = VFNMS(LDK(KP555570233), TC, VMUL(LDK(KP831469612), TH)); T1c = VFMA(LDK(KP555570233), TT, VMUL(LDK(KP831469612), TY)); T1d = VSUB(T1b, T1c); T1l = VADD(T1b, T1c); TI = VFMA(LDK(KP831469612), TC, VMUL(LDK(KP555570233), TH)); TZ = VFNMS(LDK(KP555570233), TY, VMUL(LDK(KP831469612), TT)); T10 = VSUB(TI, TZ); T1h = VADD(TI, TZ); } { V T11, T1e, T1n, T1o; T11 = VADD(Tr, T10); T1e = VBYI(VADD(T1a, T1d)); ST(&(xo[WS(os, 27)]), VSUB(T11, T1e), ovs, &(xo[WS(os, 1)])); ST(&(xo[WS(os, 5)]), VADD(T11, T1e), ovs, &(xo[WS(os, 1)])); T1n = VBYI(VADD(T1i, T1h)); T1o = VADD(T1k, T1l); ST(&(xo[WS(os, 3)]), VADD(T1n, T1o), ovs, &(xo[WS(os, 1)])); ST(&(xo[WS(os, 29)]), VSUB(T1o, T1n), ovs, &(xo[WS(os, 1)])); } { V T1f, T1g, T1j, T1m; T1f = VSUB(Tr, T10); T1g = VBYI(VSUB(T1d, T1a)); ST(&(xo[WS(os, 21)]), VSUB(T1f, T1g), ovs, &(xo[WS(os, 1)])); ST(&(xo[WS(os, 11)]), VADD(T1f, T1g), ovs, &(xo[WS(os, 1)])); T1j = VBYI(VSUB(T1h, T1i)); T1m = VSUB(T1k, T1l); ST(&(xo[WS(os, 13)]), VADD(T1j, T1m), ovs, &(xo[WS(os, 1)])); ST(&(xo[WS(os, 19)]), VSUB(T1m, T1j), ovs, &(xo[WS(os, 1)])); } } } } VLEAVE(); } static const kdft_desc desc = { 32, XSIMD_STRING("n1bv_32"), {170, 26, 16, 0}, &GENUS, 0, 0, 0, 0 }; void XSIMD(codelet_n1bv_32) (planner *p) { X(kdft_register) (p, n1bv_32, &desc); } #endif /* HAVE_FMA */ fftw-3.3.3/dft/simd/common/n1bv_64.c0000644000175000017500000015736212054410405013744 00000000000000/* * Copyright (c) 2003, 2007-11 Matteo Frigo * Copyright (c) 2003, 2007-11 Massachusetts Institute of Technology * * This program is free software; you can redistribute it and/or modify * it under the terms of the GNU General Public License as published by * the Free Software Foundation; either version 2 of the License, or * (at your option) any later version. * * This program is distributed in the hope that it will be useful, * but WITHOUT ANY WARRANTY; without even the implied warranty of * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the * GNU General Public License for more details. * * You should have received a copy of the GNU General Public License * along with this program; if not, write to the Free Software * Foundation, Inc., 51 Franklin Street, Fifth Floor, Boston, MA 02110-1301 USA * */ /* This file was automatically generated --- DO NOT EDIT */ /* Generated on Sun Nov 25 07:37:05 EST 2012 */ #include "codelet-dft.h" #ifdef HAVE_FMA /* Generated by: ../../../genfft/gen_notw_c.native -fma -reorder-insns -schedule-for-pipeline -simd -compact -variables 4 -pipeline-latency 8 -sign 1 -n 64 -name n1bv_64 -include n1b.h */ /* * This function contains 456 FP additions, 258 FP multiplications, * (or, 198 additions, 0 multiplications, 258 fused multiply/add), * 168 stack variables, 15 constants, and 128 memory accesses */ #include "n1b.h" static void n1bv_64(const R *ri, const R *ii, R *ro, R *io, stride is, stride os, INT v, INT ivs, INT ovs) { DVK(KP773010453, +0.773010453362736960810906609758469800971041293); DVK(KP820678790, +0.820678790828660330972281985331011598767386482); DVK(KP881921264, +0.881921264348355029712756863660388349508442621); DVK(KP534511135, +0.534511135950791641089685961295362908582039528); DVK(KP995184726, +0.995184726672196886244836953109479921575474869); DVK(KP098491403, +0.098491403357164253077197521291327432293052451); DVK(KP980785280, +0.980785280403230449126182236134239036973933731); DVK(KP956940335, +0.956940335732208864935797886980269969482849206); DVK(KP303346683, +0.303346683607342391675883946941299872384187453); DVK(KP831469612, +0.831469612302545237078788377617905756738560812); DVK(KP923879532, +0.923879532511286756128183189396788286822416626); DVK(KP668178637, +0.668178637919298919997757686523080761552472251); DVK(KP198912367, +0.198912367379658006911597622644676228597850501); DVK(KP414213562, +0.414213562373095048801688724209698078569671875); DVK(KP707106781, +0.707106781186547524400844362104849039284835938); { INT i; const R *xi; R *xo; xi = ii; xo = io; for (i = v; i > 0; i = i - VL, xi = xi + (VL * ivs), xo = xo + (VL * ovs), MAKE_VOLATILE_STRIDE(128, is), MAKE_VOLATILE_STRIDE(128, os)) { V T5T, T5S, T5X, T65, T5Z, T5R, T67, T63, T5U, T64; { V T7, T26, T5k, T6A, T47, T69, T2V, T3z, T6B, T4e, T6a, T5n, T3M, T2Y, T27; V Tm, T3A, T3i, T29, TC, T5p, T4o, T6D, T6e, T3l, T3B, TR, T2a, T4x, T5q; V T6h, T6E, T39, T3H, T3I, T3c, T5N, T57, T72, T6w, T5O, T5e, T71, T6t, T2y; V T1W, T2x, T1N, T33, T34, T3E, T32, T1p, T2v, T1g, T2u, T4M, T5K, T6p, T6Z; V T6m, T6Y, T5L, T4T; { V T4g, T4l, T3g, Tu, Tx, T4h, TA, T4i; { V T1, T2, T23, T24, T4, T5, T20, T21; T1 = LD(&(xi[0]), ivs, &(xi[0])); T2 = LD(&(xi[WS(is, 32)]), ivs, &(xi[0])); T23 = LD(&(xi[WS(is, 56)]), ivs, &(xi[0])); T24 = LD(&(xi[WS(is, 24)]), ivs, &(xi[0])); T4 = LD(&(xi[WS(is, 16)]), ivs, &(xi[0])); T5 = LD(&(xi[WS(is, 48)]), ivs, &(xi[0])); T20 = LD(&(xi[WS(is, 8)]), ivs, &(xi[0])); T21 = LD(&(xi[WS(is, 40)]), ivs, &(xi[0])); { V Ta, T48, Tk, T4c, T49, Td, Tf, Tg; { V T8, T43, T3, T45, T25, T5i, T6, T44, T22, T9, Ti, Tj, Tb, Tc; T8 = LD(&(xi[WS(is, 4)]), ivs, &(xi[0])); T43 = VSUB(T1, T2); T3 = VADD(T1, T2); T45 = VSUB(T23, T24); T25 = VADD(T23, T24); T5i = VSUB(T4, T5); T6 = VADD(T4, T5); T44 = VSUB(T20, T21); T22 = VADD(T20, T21); T9 = LD(&(xi[WS(is, 36)]), ivs, &(xi[0])); Ti = LD(&(xi[WS(is, 12)]), ivs, &(xi[0])); Tj = LD(&(xi[WS(is, 44)]), ivs, &(xi[0])); Tb = LD(&(xi[WS(is, 20)]), ivs, &(xi[0])); Tc = LD(&(xi[WS(is, 52)]), ivs, &(xi[0])); { V T2T, T46, T5j, T2U; T7 = VSUB(T3, T6); T2T = VADD(T3, T6); T46 = VADD(T44, T45); T5j = VSUB(T44, T45); T26 = VSUB(T22, T25); T2U = VADD(T22, T25); Ta = VADD(T8, T9); T48 = VSUB(T8, T9); Tk = VADD(Ti, Tj); T4c = VSUB(Tj, Ti); T5k = VFMA(LDK(KP707106781), T5j, T5i); T6A = VFNMS(LDK(KP707106781), T5j, T5i); T47 = VFMA(LDK(KP707106781), T46, T43); T69 = VFNMS(LDK(KP707106781), T46, T43); T2V = VADD(T2T, T2U); T3z = VSUB(T2T, T2U); T49 = VSUB(Tb, Tc); Td = VADD(Tb, Tc); } Tf = LD(&(xi[WS(is, 60)]), ivs, &(xi[0])); Tg = LD(&(xi[WS(is, 28)]), ivs, &(xi[0])); } { V Te, T2W, T5l, T4a, Tq, Tt, Tv, Tw, T5m, T4d, Tl, T2X, Ty, Tz, To; V Tp; To = LD(&(xi[WS(is, 2)]), ivs, &(xi[0])); Tp = LD(&(xi[WS(is, 34)]), ivs, &(xi[0])); { V Th, T4b, Tr, Ts; Tr = LD(&(xi[WS(is, 18)]), ivs, &(xi[0])); Ts = LD(&(xi[WS(is, 50)]), ivs, &(xi[0])); Te = VSUB(Ta, Td); T2W = VADD(Ta, Td); T5l = VFMA(LDK(KP414213562), T48, T49); T4a = VFNMS(LDK(KP414213562), T49, T48); Th = VADD(Tf, Tg); T4b = VSUB(Tf, Tg); Tq = VADD(To, Tp); T4g = VSUB(To, Tp); T4l = VSUB(Tr, Ts); Tt = VADD(Tr, Ts); Tv = LD(&(xi[WS(is, 10)]), ivs, &(xi[0])); Tw = LD(&(xi[WS(is, 42)]), ivs, &(xi[0])); T5m = VFMA(LDK(KP414213562), T4b, T4c); T4d = VFNMS(LDK(KP414213562), T4c, T4b); Tl = VSUB(Th, Tk); T2X = VADD(Th, Tk); Ty = LD(&(xi[WS(is, 58)]), ivs, &(xi[0])); Tz = LD(&(xi[WS(is, 26)]), ivs, &(xi[0])); } T3g = VADD(Tq, Tt); Tu = VSUB(Tq, Tt); Tx = VADD(Tv, Tw); T4h = VSUB(Tv, Tw); T6B = VSUB(T4a, T4d); T4e = VADD(T4a, T4d); T6a = VADD(T5l, T5m); T5n = VSUB(T5l, T5m); T3M = VSUB(T2W, T2X); T2Y = VADD(T2W, T2X); T27 = VSUB(Te, Tl); Tm = VADD(Te, Tl); TA = VADD(Ty, Tz); T4i = VSUB(Ty, Tz); } } } { V TK, T4p, T4u, T4k, T6d, T4n, T6c, TL, TN, TO, T3j, TJ, TF, TI; { V TD, TE, TG, TH; TD = LD(&(xi[WS(is, 62)]), ivs, &(xi[0])); TE = LD(&(xi[WS(is, 30)]), ivs, &(xi[0])); TG = LD(&(xi[WS(is, 14)]), ivs, &(xi[0])); TH = LD(&(xi[WS(is, 46)]), ivs, &(xi[0])); TK = LD(&(xi[WS(is, 54)]), ivs, &(xi[0])); { V T3h, TB, T4j, T4m; T3h = VADD(Tx, TA); TB = VSUB(Tx, TA); T4j = VADD(T4h, T4i); T4m = VSUB(T4h, T4i); T4p = VSUB(TD, TE); TF = VADD(TD, TE); T4u = VSUB(TH, TG); TI = VADD(TG, TH); T3A = VSUB(T3g, T3h); T3i = VADD(T3g, T3h); T29 = VFMA(LDK(KP414213562), Tu, TB); TC = VFNMS(LDK(KP414213562), TB, Tu); T4k = VFMA(LDK(KP707106781), T4j, T4g); T6d = VFNMS(LDK(KP707106781), T4j, T4g); T4n = VFMA(LDK(KP707106781), T4m, T4l); T6c = VFNMS(LDK(KP707106781), T4m, T4l); TL = LD(&(xi[WS(is, 22)]), ivs, &(xi[0])); } TN = LD(&(xi[WS(is, 6)]), ivs, &(xi[0])); TO = LD(&(xi[WS(is, 38)]), ivs, &(xi[0])); } T3j = VADD(TF, TI); TJ = VSUB(TF, TI); { V T3a, T1E, T52, T5b, T1x, T4Z, T6r, T6u, T5a, T1U, T55, T5c, T1L, T3b; { V T4V, T1t, T58, T1w, T1Q, T1T, T1I, T4Y, T59, T1J, T53, T1H; { V T1r, TM, T4r, TP, T4q, T1s, T1u, T1v; T1r = LD(&(xi[WS(is, 63)]), ivs, &(xi[WS(is, 1)])); T5p = VFMA(LDK(KP198912367), T4k, T4n); T4o = VFNMS(LDK(KP198912367), T4n, T4k); T6D = VFMA(LDK(KP668178637), T6c, T6d); T6e = VFNMS(LDK(KP668178637), T6d, T6c); TM = VADD(TK, TL); T4r = VSUB(TK, TL); TP = VADD(TN, TO); T4q = VSUB(TN, TO); T1s = LD(&(xi[WS(is, 31)]), ivs, &(xi[WS(is, 1)])); T1u = LD(&(xi[WS(is, 15)]), ivs, &(xi[WS(is, 1)])); T1v = LD(&(xi[WS(is, 47)]), ivs, &(xi[WS(is, 1)])); { V T1R, T4X, T6g, T4t, T6f, T4w, T1S, T1O, T1P; T1O = LD(&(xi[WS(is, 55)]), ivs, &(xi[WS(is, 1)])); T1P = LD(&(xi[WS(is, 23)]), ivs, &(xi[WS(is, 1)])); T1R = LD(&(xi[WS(is, 7)]), ivs, &(xi[WS(is, 1)])); { V T3k, TQ, T4s, T4v; T3k = VADD(TP, TM); TQ = VSUB(TM, TP); T4s = VADD(T4q, T4r); T4v = VSUB(T4r, T4q); T4V = VSUB(T1r, T1s); T1t = VADD(T1r, T1s); T58 = VSUB(T1v, T1u); T1w = VADD(T1u, T1v); T4X = VSUB(T1O, T1P); T1Q = VADD(T1O, T1P); T3l = VADD(T3j, T3k); T3B = VSUB(T3j, T3k); TR = VFNMS(LDK(KP414213562), TQ, TJ); T2a = VFMA(LDK(KP414213562), TJ, TQ); T6g = VFNMS(LDK(KP707106781), T4s, T4p); T4t = VFMA(LDK(KP707106781), T4s, T4p); T6f = VFNMS(LDK(KP707106781), T4v, T4u); T4w = VFMA(LDK(KP707106781), T4v, T4u); T1S = LD(&(xi[WS(is, 39)]), ivs, &(xi[WS(is, 1)])); } { V T4W, T1A, T50, T51, T1D, T1F, T1G; { V T1y, T1z, T1B, T1C; T1y = LD(&(xi[WS(is, 3)]), ivs, &(xi[WS(is, 1)])); T1z = LD(&(xi[WS(is, 35)]), ivs, &(xi[WS(is, 1)])); T1B = LD(&(xi[WS(is, 19)]), ivs, &(xi[WS(is, 1)])); T1C = LD(&(xi[WS(is, 51)]), ivs, &(xi[WS(is, 1)])); T4x = VFNMS(LDK(KP198912367), T4w, T4t); T5q = VFMA(LDK(KP198912367), T4t, T4w); T6h = VFNMS(LDK(KP668178637), T6g, T6f); T6E = VFMA(LDK(KP668178637), T6f, T6g); T4W = VSUB(T1R, T1S); T1T = VADD(T1R, T1S); T1A = VADD(T1y, T1z); T50 = VSUB(T1y, T1z); T51 = VSUB(T1C, T1B); T1D = VADD(T1B, T1C); } T1F = LD(&(xi[WS(is, 59)]), ivs, &(xi[WS(is, 1)])); T1G = LD(&(xi[WS(is, 27)]), ivs, &(xi[WS(is, 1)])); T1I = LD(&(xi[WS(is, 11)]), ivs, &(xi[WS(is, 1)])); T4Y = VADD(T4W, T4X); T59 = VSUB(T4X, T4W); T1J = LD(&(xi[WS(is, 43)]), ivs, &(xi[WS(is, 1)])); T3a = VADD(T1A, T1D); T1E = VSUB(T1A, T1D); T52 = VFMA(LDK(KP414213562), T51, T50); T5b = VFNMS(LDK(KP414213562), T50, T51); T53 = VSUB(T1F, T1G); T1H = VADD(T1F, T1G); } } } { V T37, T54, T1K, T38; T1x = VSUB(T1t, T1w); T37 = VADD(T1t, T1w); T4Z = VFMA(LDK(KP707106781), T4Y, T4V); T6r = VFNMS(LDK(KP707106781), T4Y, T4V); T54 = VSUB(T1J, T1I); T1K = VADD(T1I, T1J); T6u = VFNMS(LDK(KP707106781), T59, T58); T5a = VFMA(LDK(KP707106781), T59, T58); T38 = VADD(T1T, T1Q); T1U = VSUB(T1Q, T1T); T55 = VFNMS(LDK(KP414213562), T54, T53); T5c = VFMA(LDK(KP414213562), T53, T54); T1L = VSUB(T1H, T1K); T3b = VADD(T1H, T1K); T39 = VADD(T37, T38); T3H = VSUB(T37, T38); } } { V T4A, TW, T4N, TZ, T1j, T1m, T4O, T4D, T13, T4F, T16, T4G, T1a, T4I, T4J; V T1d; { V TU, TV, TX, TY, T56, T6v; TU = LD(&(xi[WS(is, 1)]), ivs, &(xi[WS(is, 1)])); T56 = VADD(T52, T55); T6v = VSUB(T55, T52); { V T5d, T6s, T1V, T1M; T5d = VADD(T5b, T5c); T6s = VSUB(T5c, T5b); T1V = VSUB(T1L, T1E); T1M = VADD(T1E, T1L); T3I = VSUB(T3b, T3a); T3c = VADD(T3a, T3b); T5N = VFNMS(LDK(KP923879532), T56, T4Z); T57 = VFMA(LDK(KP923879532), T56, T4Z); T72 = VFNMS(LDK(KP923879532), T6v, T6u); T6w = VFMA(LDK(KP923879532), T6v, T6u); T5O = VFNMS(LDK(KP923879532), T5d, T5a); T5e = VFMA(LDK(KP923879532), T5d, T5a); T71 = VFMA(LDK(KP923879532), T6s, T6r); T6t = VFNMS(LDK(KP923879532), T6s, T6r); T2y = VFNMS(LDK(KP707106781), T1V, T1U); T1W = VFMA(LDK(KP707106781), T1V, T1U); T2x = VFNMS(LDK(KP707106781), T1M, T1x); T1N = VFMA(LDK(KP707106781), T1M, T1x); TV = LD(&(xi[WS(is, 33)]), ivs, &(xi[WS(is, 1)])); } TX = LD(&(xi[WS(is, 17)]), ivs, &(xi[WS(is, 1)])); TY = LD(&(xi[WS(is, 49)]), ivs, &(xi[WS(is, 1)])); { V T1h, T1i, T1k, T1l; T1h = LD(&(xi[WS(is, 9)]), ivs, &(xi[WS(is, 1)])); T1i = LD(&(xi[WS(is, 41)]), ivs, &(xi[WS(is, 1)])); T1k = LD(&(xi[WS(is, 57)]), ivs, &(xi[WS(is, 1)])); T1l = LD(&(xi[WS(is, 25)]), ivs, &(xi[WS(is, 1)])); { V T11, T4B, T4C, T12, T14, T15; T11 = LD(&(xi[WS(is, 5)]), ivs, &(xi[WS(is, 1)])); T4A = VSUB(TU, TV); TW = VADD(TU, TV); T4N = VSUB(TX, TY); TZ = VADD(TX, TY); T1j = VADD(T1h, T1i); T4B = VSUB(T1h, T1i); T1m = VADD(T1k, T1l); T4C = VSUB(T1k, T1l); T12 = LD(&(xi[WS(is, 37)]), ivs, &(xi[WS(is, 1)])); T14 = LD(&(xi[WS(is, 21)]), ivs, &(xi[WS(is, 1)])); T15 = LD(&(xi[WS(is, 53)]), ivs, &(xi[WS(is, 1)])); { V T18, T19, T1b, T1c; T18 = LD(&(xi[WS(is, 61)]), ivs, &(xi[WS(is, 1)])); T19 = LD(&(xi[WS(is, 29)]), ivs, &(xi[WS(is, 1)])); T1b = LD(&(xi[WS(is, 13)]), ivs, &(xi[WS(is, 1)])); T1c = LD(&(xi[WS(is, 45)]), ivs, &(xi[WS(is, 1)])); T4O = VSUB(T4B, T4C); T4D = VADD(T4B, T4C); T13 = VADD(T11, T12); T4F = VSUB(T11, T12); T16 = VADD(T14, T15); T4G = VSUB(T14, T15); T1a = VADD(T18, T19); T4I = VSUB(T18, T19); T4J = VSUB(T1b, T1c); T1d = VADD(T1b, T1c); } } } } { V T30, T10, T6k, T4E, T4Q, T4H, T17, T6n, T4P, T1e, T4K, T4R, T1n, T31; T30 = VADD(TW, TZ); T10 = VSUB(TW, TZ); T6k = VFNMS(LDK(KP707106781), T4D, T4A); T4E = VFMA(LDK(KP707106781), T4D, T4A); T4Q = VFMA(LDK(KP414213562), T4F, T4G); T4H = VFNMS(LDK(KP414213562), T4G, T4F); T33 = VADD(T13, T16); T17 = VSUB(T13, T16); T6n = VFNMS(LDK(KP707106781), T4O, T4N); T4P = VFMA(LDK(KP707106781), T4O, T4N); T34 = VADD(T1a, T1d); T1e = VSUB(T1a, T1d); T4K = VFMA(LDK(KP414213562), T4J, T4I); T4R = VFNMS(LDK(KP414213562), T4I, T4J); T1n = VSUB(T1j, T1m); T31 = VADD(T1j, T1m); { V T1f, T1o, T6o, T4L, T4S, T6l; T1f = VADD(T17, T1e); T1o = VSUB(T17, T1e); T6o = VSUB(T4H, T4K); T4L = VADD(T4H, T4K); T4S = VADD(T4Q, T4R); T6l = VSUB(T4Q, T4R); T3E = VSUB(T30, T31); T32 = VADD(T30, T31); T1p = VFMA(LDK(KP707106781), T1o, T1n); T2v = VFNMS(LDK(KP707106781), T1o, T1n); T1g = VFMA(LDK(KP707106781), T1f, T10); T2u = VFNMS(LDK(KP707106781), T1f, T10); T4M = VFMA(LDK(KP923879532), T4L, T4E); T5K = VFNMS(LDK(KP923879532), T4L, T4E); T6p = VFMA(LDK(KP923879532), T6o, T6n); T6Z = VFNMS(LDK(KP923879532), T6o, T6n); T6m = VFNMS(LDK(KP923879532), T6l, T6k); T6Y = VFMA(LDK(KP923879532), T6l, T6k); T5L = VFNMS(LDK(KP923879532), T4S, T4P); T4T = VFMA(LDK(KP923879532), T4S, T4P); } } } } } } { V T6b, T6F, T7f, T6X, T70, T79, T7a, T73, T6C, T76, T77, T6i; { V T2Z, T3r, T3s, T3m, T3d, T3v; T2Z = VSUB(T2V, T2Y); T3r = VADD(T2V, T2Y); T3s = VADD(T3i, T3l); T3m = VSUB(T3i, T3l); T3d = VSUB(T39, T3c); T3v = VADD(T39, T3c); { V T3x, T3t, T3Q, T3J, T3D, T3V, T3G, T3P, T3u, T36, T3O, T3Y, T6V, T6W; { V T3N, T3C, T3F, T35; T3N = VSUB(T3A, T3B); T3C = VADD(T3A, T3B); T3F = VSUB(T33, T34); T35 = VADD(T33, T34); T3x = VADD(T3r, T3s); T3t = VSUB(T3r, T3s); T3Q = VFMA(LDK(KP414213562), T3H, T3I); T3J = VFNMS(LDK(KP414213562), T3I, T3H); T3D = VFMA(LDK(KP707106781), T3C, T3z); T3V = VFNMS(LDK(KP707106781), T3C, T3z); T3G = VFNMS(LDK(KP414213562), T3F, T3E); T3P = VFMA(LDK(KP414213562), T3E, T3F); T3u = VADD(T32, T35); T36 = VSUB(T32, T35); T3O = VFMA(LDK(KP707106781), T3N, T3M); T3Y = VFNMS(LDK(KP707106781), T3N, T3M); } T6b = VFNMS(LDK(KP923879532), T6a, T69); T6V = VFMA(LDK(KP923879532), T6a, T69); T6W = VADD(T6D, T6E); T6F = VSUB(T6D, T6E); { V T3R, T3W, T3K, T3Z; T3R = VSUB(T3P, T3Q); T3W = VADD(T3P, T3Q); T3K = VADD(T3G, T3J); T3Z = VSUB(T3G, T3J); { V T3e, T3n, T3w, T3y; T3e = VADD(T36, T3d); T3n = VSUB(T36, T3d); T3w = VSUB(T3u, T3v); T3y = VADD(T3u, T3v); { V T41, T3X, T3S, T3U; T41 = VFMA(LDK(KP923879532), T3W, T3V); T3X = VFNMS(LDK(KP923879532), T3W, T3V); T3S = VFNMS(LDK(KP923879532), T3R, T3O); T3U = VFMA(LDK(KP923879532), T3R, T3O); { V T42, T40, T3L, T3T; T42 = VFNMS(LDK(KP923879532), T3Z, T3Y); T40 = VFMA(LDK(KP923879532), T3Z, T3Y); T3L = VFNMS(LDK(KP923879532), T3K, T3D); T3T = VFMA(LDK(KP923879532), T3K, T3D); { V T3o, T3q, T3f, T3p; T3o = VFNMS(LDK(KP707106781), T3n, T3m); T3q = VFMA(LDK(KP707106781), T3n, T3m); T3f = VFNMS(LDK(KP707106781), T3e, T2Z); T3p = VFMA(LDK(KP707106781), T3e, T2Z); ST(&(xo[WS(os, 32)]), VSUB(T3x, T3y), ovs, &(xo[0])); ST(&(xo[0]), VADD(T3x, T3y), ovs, &(xo[0])); ST(&(xo[WS(os, 16)]), VFMAI(T3w, T3t), ovs, &(xo[0])); ST(&(xo[WS(os, 48)]), VFNMSI(T3w, T3t), ovs, &(xo[0])); ST(&(xo[WS(os, 44)]), VFNMSI(T40, T3X), ovs, &(xo[0])); ST(&(xo[WS(os, 20)]), VFMAI(T40, T3X), ovs, &(xo[0])); ST(&(xo[WS(os, 52)]), VFMAI(T42, T41), ovs, &(xo[0])); ST(&(xo[WS(os, 12)]), VFNMSI(T42, T41), ovs, &(xo[0])); ST(&(xo[WS(os, 4)]), VFMAI(T3U, T3T), ovs, &(xo[0])); ST(&(xo[WS(os, 60)]), VFNMSI(T3U, T3T), ovs, &(xo[0])); ST(&(xo[WS(os, 36)]), VFMAI(T3S, T3L), ovs, &(xo[0])); ST(&(xo[WS(os, 28)]), VFNMSI(T3S, T3L), ovs, &(xo[0])); ST(&(xo[WS(os, 56)]), VFNMSI(T3q, T3p), ovs, &(xo[0])); ST(&(xo[WS(os, 8)]), VFMAI(T3q, T3p), ovs, &(xo[0])); ST(&(xo[WS(os, 40)]), VFMAI(T3o, T3f), ovs, &(xo[0])); ST(&(xo[WS(os, 24)]), VFNMSI(T3o, T3f), ovs, &(xo[0])); T7f = VFNMS(LDK(KP831469612), T6W, T6V); T6X = VFMA(LDK(KP831469612), T6W, T6V); } } } } } T70 = VFMA(LDK(KP303346683), T6Z, T6Y); T79 = VFNMS(LDK(KP303346683), T6Y, T6Z); T7a = VFNMS(LDK(KP303346683), T71, T72); T73 = VFMA(LDK(KP303346683), T72, T71); T6C = VFMA(LDK(KP923879532), T6B, T6A); T76 = VFNMS(LDK(KP923879532), T6B, T6A); T77 = VSUB(T6e, T6h); T6i = VADD(T6e, T6h); } } { V T2r, T2D, T2C, T2s, T5H, T5o, T5v, T5D, T5r, T5I, T5x, T5h, T5F, T5B; { V TT, T2f, T2n, T1Y, T28, T2b, T2l, T2p, T2j, T2k; { V T1q, T2d, T7h, T7l, T2e, T1X, T75, T7d, T7m, T7k, T7c, T7e, Tn, TS; T2r = VFNMS(LDK(KP707106781), Tm, T7); Tn = VFMA(LDK(KP707106781), Tm, T7); TS = VADD(TC, TR); T2D = VSUB(TC, TR); { V T7b, T7j, T74, T7i, T78, T7g; T1q = VFNMS(LDK(KP198912367), T1p, T1g); T2d = VFMA(LDK(KP198912367), T1g, T1p); T7g = VADD(T79, T7a); T7b = VSUB(T79, T7a); T7j = VSUB(T70, T73); T74 = VADD(T70, T73); T7i = VFNMS(LDK(KP831469612), T77, T76); T78 = VFMA(LDK(KP831469612), T77, T76); T2j = VFNMS(LDK(KP923879532), TS, Tn); TT = VFMA(LDK(KP923879532), TS, Tn); T7h = VFMA(LDK(KP956940335), T7g, T7f); T7l = VFNMS(LDK(KP956940335), T7g, T7f); T2e = VFMA(LDK(KP198912367), T1N, T1W); T1X = VFNMS(LDK(KP198912367), T1W, T1N); T75 = VFNMS(LDK(KP956940335), T74, T6X); T7d = VFMA(LDK(KP956940335), T74, T6X); T7m = VFMA(LDK(KP956940335), T7j, T7i); T7k = VFNMS(LDK(KP956940335), T7j, T7i); T7c = VFNMS(LDK(KP956940335), T7b, T78); T7e = VFMA(LDK(KP956940335), T7b, T78); } T2k = VADD(T2d, T2e); T2f = VSUB(T2d, T2e); ST(&(xo[WS(os, 45)]), VFMAI(T7k, T7h), ovs, &(xo[WS(os, 1)])); ST(&(xo[WS(os, 19)]), VFNMSI(T7k, T7h), ovs, &(xo[WS(os, 1)])); ST(&(xo[WS(os, 51)]), VFNMSI(T7m, T7l), ovs, &(xo[WS(os, 1)])); ST(&(xo[WS(os, 13)]), VFMAI(T7m, T7l), ovs, &(xo[WS(os, 1)])); ST(&(xo[WS(os, 61)]), VFMAI(T7e, T7d), ovs, &(xo[WS(os, 1)])); ST(&(xo[WS(os, 3)]), VFNMSI(T7e, T7d), ovs, &(xo[WS(os, 1)])); ST(&(xo[WS(os, 29)]), VFMAI(T7c, T75), ovs, &(xo[WS(os, 1)])); ST(&(xo[WS(os, 35)]), VFNMSI(T7c, T75), ovs, &(xo[WS(os, 1)])); T2n = VSUB(T1q, T1X); T1Y = VADD(T1q, T1X); T2C = VFNMS(LDK(KP707106781), T27, T26); T28 = VFMA(LDK(KP707106781), T27, T26); T2b = VSUB(T29, T2a); T2s = VADD(T29, T2a); } T2l = VFNMS(LDK(KP980785280), T2k, T2j); T2p = VFMA(LDK(KP980785280), T2k, T2j); { V T5z, T4z, T5A, T5g; { V T4f, T4y, T1Z, T2h, T4U, T5t, T2m, T2c, T5u, T5f; T5H = VFNMS(LDK(KP923879532), T4e, T47); T4f = VFMA(LDK(KP923879532), T4e, T47); T4y = VADD(T4o, T4x); T5T = VSUB(T4o, T4x); T1Z = VFNMS(LDK(KP980785280), T1Y, TT); T2h = VFMA(LDK(KP980785280), T1Y, TT); T4U = VFNMS(LDK(KP098491403), T4T, T4M); T5t = VFMA(LDK(KP098491403), T4M, T4T); T2m = VFNMS(LDK(KP923879532), T2b, T28); T2c = VFMA(LDK(KP923879532), T2b, T28); T5u = VFMA(LDK(KP098491403), T57, T5e); T5f = VFNMS(LDK(KP098491403), T5e, T57); T5z = VFNMS(LDK(KP980785280), T4y, T4f); T4z = VFMA(LDK(KP980785280), T4y, T4f); T5S = VFNMS(LDK(KP923879532), T5n, T5k); T5o = VFMA(LDK(KP923879532), T5n, T5k); { V T2o, T2q, T2i, T2g; T2o = VFMA(LDK(KP980785280), T2n, T2m); T2q = VFNMS(LDK(KP980785280), T2n, T2m); T2i = VFMA(LDK(KP980785280), T2f, T2c); T2g = VFNMS(LDK(KP980785280), T2f, T2c); T5A = VADD(T5t, T5u); T5v = VSUB(T5t, T5u); T5D = VSUB(T4U, T5f); T5g = VADD(T4U, T5f); ST(&(xo[WS(os, 46)]), VFNMSI(T2o, T2l), ovs, &(xo[0])); ST(&(xo[WS(os, 18)]), VFMAI(T2o, T2l), ovs, &(xo[0])); ST(&(xo[WS(os, 50)]), VFMAI(T2q, T2p), ovs, &(xo[0])); ST(&(xo[WS(os, 14)]), VFNMSI(T2q, T2p), ovs, &(xo[0])); ST(&(xo[WS(os, 2)]), VFMAI(T2i, T2h), ovs, &(xo[0])); ST(&(xo[WS(os, 62)]), VFNMSI(T2i, T2h), ovs, &(xo[0])); ST(&(xo[WS(os, 34)]), VFMAI(T2g, T1Z), ovs, &(xo[0])); ST(&(xo[WS(os, 30)]), VFNMSI(T2g, T1Z), ovs, &(xo[0])); T5r = VSUB(T5p, T5q); T5I = VADD(T5p, T5q); } } T5x = VFMA(LDK(KP995184726), T5g, T4z); T5h = VFNMS(LDK(KP995184726), T5g, T4z); T5F = VFMA(LDK(KP995184726), T5A, T5z); T5B = VFNMS(LDK(KP995184726), T5A, T5z); } } { V T6J, T6R, T6L, T6z, T6T, T6P; { V T6N, T6j, T6O, T6y; { V T6q, T6H, T5C, T5s, T6I, T6x; T6q = VFNMS(LDK(KP534511135), T6p, T6m); T6H = VFMA(LDK(KP534511135), T6m, T6p); T5C = VFNMS(LDK(KP980785280), T5r, T5o); T5s = VFMA(LDK(KP980785280), T5r, T5o); T6I = VFMA(LDK(KP534511135), T6t, T6w); T6x = VFNMS(LDK(KP534511135), T6w, T6t); T6N = VFMA(LDK(KP831469612), T6i, T6b); T6j = VFNMS(LDK(KP831469612), T6i, T6b); { V T5E, T5G, T5y, T5w; T5E = VFMA(LDK(KP995184726), T5D, T5C); T5G = VFNMS(LDK(KP995184726), T5D, T5C); T5y = VFMA(LDK(KP995184726), T5v, T5s); T5w = VFNMS(LDK(KP995184726), T5v, T5s); T6O = VADD(T6H, T6I); T6J = VSUB(T6H, T6I); T6R = VSUB(T6q, T6x); T6y = VADD(T6q, T6x); ST(&(xo[WS(os, 47)]), VFNMSI(T5E, T5B), ovs, &(xo[WS(os, 1)])); ST(&(xo[WS(os, 17)]), VFMAI(T5E, T5B), ovs, &(xo[WS(os, 1)])); ST(&(xo[WS(os, 49)]), VFMAI(T5G, T5F), ovs, &(xo[WS(os, 1)])); ST(&(xo[WS(os, 15)]), VFNMSI(T5G, T5F), ovs, &(xo[WS(os, 1)])); ST(&(xo[WS(os, 1)]), VFMAI(T5y, T5x), ovs, &(xo[WS(os, 1)])); ST(&(xo[WS(os, 63)]), VFNMSI(T5y, T5x), ovs, &(xo[WS(os, 1)])); ST(&(xo[WS(os, 33)]), VFMAI(T5w, T5h), ovs, &(xo[WS(os, 1)])); ST(&(xo[WS(os, 31)]), VFNMSI(T5w, T5h), ovs, &(xo[WS(os, 1)])); } } T6L = VFMA(LDK(KP881921264), T6y, T6j); T6z = VFNMS(LDK(KP881921264), T6y, T6j); T6T = VFMA(LDK(KP881921264), T6O, T6N); T6P = VFNMS(LDK(KP881921264), T6O, T6N); } { V T2H, T2P, T2J, T2B, T2R, T2N; { V T2L, T2t, T2M, T2A; { V T2w, T2F, T6Q, T6G, T2G, T2z; T2w = VFMA(LDK(KP668178637), T2v, T2u); T2F = VFNMS(LDK(KP668178637), T2u, T2v); T6Q = VFNMS(LDK(KP831469612), T6F, T6C); T6G = VFMA(LDK(KP831469612), T6F, T6C); T2G = VFNMS(LDK(KP668178637), T2x, T2y); T2z = VFMA(LDK(KP668178637), T2y, T2x); T2L = VFNMS(LDK(KP923879532), T2s, T2r); T2t = VFMA(LDK(KP923879532), T2s, T2r); { V T6S, T6U, T6M, T6K; T6S = VFMA(LDK(KP881921264), T6R, T6Q); T6U = VFNMS(LDK(KP881921264), T6R, T6Q); T6M = VFMA(LDK(KP881921264), T6J, T6G); T6K = VFNMS(LDK(KP881921264), T6J, T6G); T2M = VADD(T2F, T2G); T2H = VSUB(T2F, T2G); T2P = VSUB(T2w, T2z); T2A = VADD(T2w, T2z); ST(&(xo[WS(os, 43)]), VFNMSI(T6S, T6P), ovs, &(xo[WS(os, 1)])); ST(&(xo[WS(os, 21)]), VFMAI(T6S, T6P), ovs, &(xo[WS(os, 1)])); ST(&(xo[WS(os, 53)]), VFMAI(T6U, T6T), ovs, &(xo[WS(os, 1)])); ST(&(xo[WS(os, 11)]), VFNMSI(T6U, T6T), ovs, &(xo[WS(os, 1)])); ST(&(xo[WS(os, 5)]), VFMAI(T6M, T6L), ovs, &(xo[WS(os, 1)])); ST(&(xo[WS(os, 59)]), VFNMSI(T6M, T6L), ovs, &(xo[WS(os, 1)])); ST(&(xo[WS(os, 37)]), VFMAI(T6K, T6z), ovs, &(xo[WS(os, 1)])); ST(&(xo[WS(os, 27)]), VFNMSI(T6K, T6z), ovs, &(xo[WS(os, 1)])); } } T2J = VFMA(LDK(KP831469612), T2A, T2t); T2B = VFNMS(LDK(KP831469612), T2A, T2t); T2R = VFNMS(LDK(KP831469612), T2M, T2L); T2N = VFMA(LDK(KP831469612), T2M, T2L); } { V T61, T5J, T62, T5Q; { V T5M, T5V, T2O, T2E, T5W, T5P; T5M = VFMA(LDK(KP820678790), T5L, T5K); T5V = VFNMS(LDK(KP820678790), T5K, T5L); T2O = VFMA(LDK(KP923879532), T2D, T2C); T2E = VFNMS(LDK(KP923879532), T2D, T2C); T5W = VFNMS(LDK(KP820678790), T5N, T5O); T5P = VFMA(LDK(KP820678790), T5O, T5N); T61 = VFNMS(LDK(KP980785280), T5I, T5H); T5J = VFMA(LDK(KP980785280), T5I, T5H); { V T2Q, T2S, T2K, T2I; T2Q = VFNMS(LDK(KP831469612), T2P, T2O); T2S = VFMA(LDK(KP831469612), T2P, T2O); T2K = VFMA(LDK(KP831469612), T2H, T2E); T2I = VFNMS(LDK(KP831469612), T2H, T2E); T62 = VADD(T5V, T5W); T5X = VSUB(T5V, T5W); T65 = VSUB(T5M, T5P); T5Q = VADD(T5M, T5P); ST(&(xo[WS(os, 42)]), VFMAI(T2Q, T2N), ovs, &(xo[0])); ST(&(xo[WS(os, 22)]), VFNMSI(T2Q, T2N), ovs, &(xo[0])); ST(&(xo[WS(os, 54)]), VFNMSI(T2S, T2R), ovs, &(xo[0])); ST(&(xo[WS(os, 10)]), VFMAI(T2S, T2R), ovs, &(xo[0])); ST(&(xo[WS(os, 58)]), VFMAI(T2K, T2J), ovs, &(xo[0])); ST(&(xo[WS(os, 6)]), VFNMSI(T2K, T2J), ovs, &(xo[0])); ST(&(xo[WS(os, 26)]), VFMAI(T2I, T2B), ovs, &(xo[0])); ST(&(xo[WS(os, 38)]), VFNMSI(T2I, T2B), ovs, &(xo[0])); } } T5Z = VFMA(LDK(KP773010453), T5Q, T5J); T5R = VFNMS(LDK(KP773010453), T5Q, T5J); T67 = VFNMS(LDK(KP773010453), T62, T61); T63 = VFMA(LDK(KP773010453), T62, T61); } } } } } } T5U = VFNMS(LDK(KP980785280), T5T, T5S); T64 = VFMA(LDK(KP980785280), T5T, T5S); { V T68, T66, T5Y, T60; T68 = VFMA(LDK(KP773010453), T65, T64); T66 = VFNMS(LDK(KP773010453), T65, T64); T5Y = VFNMS(LDK(KP773010453), T5X, T5U); T60 = VFMA(LDK(KP773010453), T5X, T5U); ST(&(xo[WS(os, 41)]), VFMAI(T66, T63), ovs, &(xo[WS(os, 1)])); ST(&(xo[WS(os, 23)]), VFNMSI(T66, T63), ovs, &(xo[WS(os, 1)])); ST(&(xo[WS(os, 55)]), VFNMSI(T68, T67), ovs, &(xo[WS(os, 1)])); ST(&(xo[WS(os, 9)]), VFMAI(T68, T67), ovs, &(xo[WS(os, 1)])); ST(&(xo[WS(os, 57)]), VFMAI(T60, T5Z), ovs, &(xo[WS(os, 1)])); ST(&(xo[WS(os, 7)]), VFNMSI(T60, T5Z), ovs, &(xo[WS(os, 1)])); ST(&(xo[WS(os, 25)]), VFMAI(T5Y, T5R), ovs, &(xo[WS(os, 1)])); ST(&(xo[WS(os, 39)]), VFNMSI(T5Y, T5R), ovs, &(xo[WS(os, 1)])); } } } VLEAVE(); } static const kdft_desc desc = { 64, XSIMD_STRING("n1bv_64"), {198, 0, 258, 0}, &GENUS, 0, 0, 0, 0 }; void XSIMD(codelet_n1bv_64) (planner *p) { X(kdft_register) (p, n1bv_64, &desc); } #else /* HAVE_FMA */ /* Generated by: ../../../genfft/gen_notw_c.native -simd -compact -variables 4 -pipeline-latency 8 -sign 1 -n 64 -name n1bv_64 -include n1b.h */ /* * This function contains 456 FP additions, 124 FP multiplications, * (or, 404 additions, 72 multiplications, 52 fused multiply/add), * 108 stack variables, 15 constants, and 128 memory accesses */ #include "n1b.h" static void n1bv_64(const R *ri, const R *ii, R *ro, R *io, stride is, stride os, INT v, INT ivs, INT ovs) { DVK(KP634393284, +0.634393284163645498215171613225493370675687095); DVK(KP773010453, +0.773010453362736960810906609758469800971041293); DVK(KP290284677, +0.290284677254462367636192375817395274691476278); DVK(KP956940335, +0.956940335732208864935797886980269969482849206); DVK(KP098017140, +0.098017140329560601994195563888641845861136673); DVK(KP995184726, +0.995184726672196886244836953109479921575474869); DVK(KP471396736, +0.471396736825997648556387625905254377657460319); DVK(KP881921264, +0.881921264348355029712756863660388349508442621); DVK(KP195090322, +0.195090322016128267848284868477022240927691618); DVK(KP980785280, +0.980785280403230449126182236134239036973933731); DVK(KP555570233, +0.555570233019602224742830813948532874374937191); DVK(KP831469612, +0.831469612302545237078788377617905756738560812); DVK(KP923879532, +0.923879532511286756128183189396788286822416626); DVK(KP382683432, +0.382683432365089771728459984030398866761344562); DVK(KP707106781, +0.707106781186547524400844362104849039284835938); { INT i; const R *xi; R *xo; xi = ii; xo = io; for (i = v; i > 0; i = i - VL, xi = xi + (VL * ivs), xo = xo + (VL * ovs), MAKE_VOLATILE_STRIDE(128, is), MAKE_VOLATILE_STRIDE(128, os)) { V T4p, T5u, Tb, T3A, T2q, T3v, T6G, T78, Tq, T3w, T6B, T79, T2l, T3B, T4w; V T5r, TI, T2g, T6u, T74, T3q, T3D, T4E, T5o, TZ, T2h, T6x, T75, T3t, T3E; V T4L, T5p, T23, T2N, T6m, T70, T6p, T71, T2c, T2O, T3i, T3Y, T5f, T5R, T5k; V T5S, T3l, T3Z, T1s, T2K, T6f, T6X, T6i, T6Y, T1B, T2L, T3b, T3V, T4Y, T5O; V T53, T5P, T3e, T3W; { V T3, T4n, T2p, T4o, T6, T5s, T9, T5t; { V T1, T2, T2n, T2o; T1 = LD(&(xi[0]), ivs, &(xi[0])); T2 = LD(&(xi[WS(is, 32)]), ivs, &(xi[0])); T3 = VSUB(T1, T2); T4n = VADD(T1, T2); T2n = LD(&(xi[WS(is, 16)]), ivs, &(xi[0])); T2o = LD(&(xi[WS(is, 48)]), ivs, &(xi[0])); T2p = VSUB(T2n, T2o); T4o = VADD(T2n, T2o); } { V T4, T5, T7, T8; T4 = LD(&(xi[WS(is, 8)]), ivs, &(xi[0])); T5 = LD(&(xi[WS(is, 40)]), ivs, &(xi[0])); T6 = VSUB(T4, T5); T5s = VADD(T4, T5); T7 = LD(&(xi[WS(is, 56)]), ivs, &(xi[0])); T8 = LD(&(xi[WS(is, 24)]), ivs, &(xi[0])); T9 = VSUB(T7, T8); T5t = VADD(T7, T8); } T4p = VSUB(T4n, T4o); T5u = VSUB(T5s, T5t); { V Ta, T2m, T6E, T6F; Ta = VMUL(LDK(KP707106781), VADD(T6, T9)); Tb = VSUB(T3, Ta); T3A = VADD(T3, Ta); T2m = VMUL(LDK(KP707106781), VSUB(T6, T9)); T2q = VSUB(T2m, T2p); T3v = VADD(T2p, T2m); T6E = VADD(T4n, T4o); T6F = VADD(T5s, T5t); T6G = VSUB(T6E, T6F); T78 = VADD(T6E, T6F); } } { V Te, T4q, To, T4t, Th, T4r, Tl, T4u; { V Tc, Td, Tm, Tn; Tc = LD(&(xi[WS(is, 4)]), ivs, &(xi[0])); Td = LD(&(xi[WS(is, 36)]), ivs, &(xi[0])); Te = VSUB(Tc, Td); T4q = VADD(Tc, Td); Tm = LD(&(xi[WS(is, 60)]), ivs, &(xi[0])); Tn = LD(&(xi[WS(is, 28)]), ivs, &(xi[0])); To = VSUB(Tm, Tn); T4t = VADD(Tm, Tn); } { V Tf, Tg, Tj, Tk; Tf = LD(&(xi[WS(is, 20)]), ivs, &(xi[0])); Tg = LD(&(xi[WS(is, 52)]), ivs, &(xi[0])); Th = VSUB(Tf, Tg); T4r = VADD(Tf, Tg); Tj = LD(&(xi[WS(is, 12)]), ivs, &(xi[0])); Tk = LD(&(xi[WS(is, 44)]), ivs, &(xi[0])); Tl = VSUB(Tj, Tk); T4u = VADD(Tj, Tk); } { V Ti, Tp, T6z, T6A; Ti = VFMA(LDK(KP382683432), Te, VMUL(LDK(KP923879532), Th)); Tp = VFNMS(LDK(KP382683432), To, VMUL(LDK(KP923879532), Tl)); Tq = VSUB(Ti, Tp); T3w = VADD(Ti, Tp); T6z = VADD(T4q, T4r); T6A = VADD(T4t, T4u); T6B = VSUB(T6z, T6A); T79 = VADD(T6z, T6A); } { V T2j, T2k, T4s, T4v; T2j = VFNMS(LDK(KP382683432), Th, VMUL(LDK(KP923879532), Te)); T2k = VFMA(LDK(KP923879532), To, VMUL(LDK(KP382683432), Tl)); T2l = VSUB(T2j, T2k); T3B = VADD(T2j, T2k); T4s = VSUB(T4q, T4r); T4v = VSUB(T4t, T4u); T4w = VMUL(LDK(KP707106781), VADD(T4s, T4v)); T5r = VMUL(LDK(KP707106781), VSUB(T4s, T4v)); } } { V TB, T4z, TF, T4y, Ty, T4C, TG, T4B; { V Tz, TA, TD, TE; Tz = LD(&(xi[WS(is, 18)]), ivs, &(xi[0])); TA = LD(&(xi[WS(is, 50)]), ivs, &(xi[0])); TB = VSUB(Tz, TA); T4z = VADD(Tz, TA); TD = LD(&(xi[WS(is, 2)]), ivs, &(xi[0])); TE = LD(&(xi[WS(is, 34)]), ivs, &(xi[0])); TF = VSUB(TD, TE); T4y = VADD(TD, TE); { V Ts, Tt, Tu, Tv, Tw, Tx; Ts = LD(&(xi[WS(is, 10)]), ivs, &(xi[0])); Tt = LD(&(xi[WS(is, 42)]), ivs, &(xi[0])); Tu = VSUB(Ts, Tt); Tv = LD(&(xi[WS(is, 58)]), ivs, &(xi[0])); Tw = LD(&(xi[WS(is, 26)]), ivs, &(xi[0])); Tx = VSUB(Tv, Tw); Ty = VMUL(LDK(KP707106781), VSUB(Tu, Tx)); T4C = VADD(Tv, Tw); TG = VMUL(LDK(KP707106781), VADD(Tu, Tx)); T4B = VADD(Ts, Tt); } } { V TC, TH, T6s, T6t; TC = VSUB(Ty, TB); TH = VSUB(TF, TG); TI = VFMA(LDK(KP831469612), TC, VMUL(LDK(KP555570233), TH)); T2g = VFNMS(LDK(KP555570233), TC, VMUL(LDK(KP831469612), TH)); T6s = VADD(T4y, T4z); T6t = VADD(T4B, T4C); T6u = VSUB(T6s, T6t); T74 = VADD(T6s, T6t); } { V T3o, T3p, T4A, T4D; T3o = VADD(TB, Ty); T3p = VADD(TF, TG); T3q = VFMA(LDK(KP980785280), T3o, VMUL(LDK(KP195090322), T3p)); T3D = VFNMS(LDK(KP195090322), T3o, VMUL(LDK(KP980785280), T3p)); T4A = VSUB(T4y, T4z); T4D = VSUB(T4B, T4C); T4E = VFMA(LDK(KP382683432), T4A, VMUL(LDK(KP923879532), T4D)); T5o = VFNMS(LDK(KP382683432), T4D, VMUL(LDK(KP923879532), T4A)); } } { V TS, T4J, TW, T4I, TP, T4G, TX, T4F; { V TQ, TR, TU, TV; TQ = LD(&(xi[WS(is, 14)]), ivs, &(xi[0])); TR = LD(&(xi[WS(is, 46)]), ivs, &(xi[0])); TS = VSUB(TQ, TR); T4J = VADD(TQ, TR); TU = LD(&(xi[WS(is, 62)]), ivs, &(xi[0])); TV = LD(&(xi[WS(is, 30)]), ivs, &(xi[0])); TW = VSUB(TU, TV); T4I = VADD(TU, TV); { V TJ, TK, TL, TM, TN, TO; TJ = LD(&(xi[WS(is, 6)]), ivs, &(xi[0])); TK = LD(&(xi[WS(is, 38)]), ivs, &(xi[0])); TL = VSUB(TJ, TK); TM = LD(&(xi[WS(is, 54)]), ivs, &(xi[0])); TN = LD(&(xi[WS(is, 22)]), ivs, &(xi[0])); TO = VSUB(TM, TN); TP = VMUL(LDK(KP707106781), VSUB(TL, TO)); T4G = VADD(TM, TN); TX = VMUL(LDK(KP707106781), VADD(TL, TO)); T4F = VADD(TJ, TK); } } { V TT, TY, T6v, T6w; TT = VSUB(TP, TS); TY = VSUB(TW, TX); TZ = VFNMS(LDK(KP555570233), TY, VMUL(LDK(KP831469612), TT)); T2h = VFMA(LDK(KP555570233), TT, VMUL(LDK(KP831469612), TY)); T6v = VADD(T4I, T4J); T6w = VADD(T4F, T4G); T6x = VSUB(T6v, T6w); T75 = VADD(T6v, T6w); } { V T3r, T3s, T4H, T4K; T3r = VADD(TS, TP); T3s = VADD(TW, TX); T3t = VFNMS(LDK(KP195090322), T3s, VMUL(LDK(KP980785280), T3r)); T3E = VFMA(LDK(KP195090322), T3r, VMUL(LDK(KP980785280), T3s)); T4H = VSUB(T4F, T4G); T4K = VSUB(T4I, T4J); T4L = VFNMS(LDK(KP382683432), T4K, VMUL(LDK(KP923879532), T4H)); T5p = VFMA(LDK(KP923879532), T4K, VMUL(LDK(KP382683432), T4H)); } } { V T21, T5h, T26, T5g, T1Y, T5d, T27, T5c, T55, T56, T1J, T57, T29, T58, T59; V T1Q, T5a, T2a; { V T1Z, T20, T24, T25; T1Z = LD(&(xi[WS(is, 15)]), ivs, &(xi[WS(is, 1)])); T20 = LD(&(xi[WS(is, 47)]), ivs, &(xi[WS(is, 1)])); T21 = VSUB(T1Z, T20); T5h = VADD(T1Z, T20); T24 = LD(&(xi[WS(is, 63)]), ivs, &(xi[WS(is, 1)])); T25 = LD(&(xi[WS(is, 31)]), ivs, &(xi[WS(is, 1)])); T26 = VSUB(T24, T25); T5g = VADD(T24, T25); } { V T1S, T1T, T1U, T1V, T1W, T1X; T1S = LD(&(xi[WS(is, 7)]), ivs, &(xi[WS(is, 1)])); T1T = LD(&(xi[WS(is, 39)]), ivs, &(xi[WS(is, 1)])); T1U = VSUB(T1S, T1T); T1V = LD(&(xi[WS(is, 55)]), ivs, &(xi[WS(is, 1)])); T1W = LD(&(xi[WS(is, 23)]), ivs, &(xi[WS(is, 1)])); T1X = VSUB(T1V, T1W); T1Y = VMUL(LDK(KP707106781), VSUB(T1U, T1X)); T5d = VADD(T1V, T1W); T27 = VMUL(LDK(KP707106781), VADD(T1U, T1X)); T5c = VADD(T1S, T1T); } { V T1F, T1I, T1M, T1P; { V T1D, T1E, T1G, T1H; T1D = LD(&(xi[WS(is, 3)]), ivs, &(xi[WS(is, 1)])); T1E = LD(&(xi[WS(is, 35)]), ivs, &(xi[WS(is, 1)])); T1F = VSUB(T1D, T1E); T55 = VADD(T1D, T1E); T1G = LD(&(xi[WS(is, 19)]), ivs, &(xi[WS(is, 1)])); T1H = LD(&(xi[WS(is, 51)]), ivs, &(xi[WS(is, 1)])); T1I = VSUB(T1G, T1H); T56 = VADD(T1G, T1H); } T1J = VFNMS(LDK(KP382683432), T1I, VMUL(LDK(KP923879532), T1F)); T57 = VSUB(T55, T56); T29 = VFMA(LDK(KP382683432), T1F, VMUL(LDK(KP923879532), T1I)); { V T1K, T1L, T1N, T1O; T1K = LD(&(xi[WS(is, 59)]), ivs, &(xi[WS(is, 1)])); T1L = LD(&(xi[WS(is, 27)]), ivs, &(xi[WS(is, 1)])); T1M = VSUB(T1K, T1L); T58 = VADD(T1K, T1L); T1N = LD(&(xi[WS(is, 11)]), ivs, &(xi[WS(is, 1)])); T1O = LD(&(xi[WS(is, 43)]), ivs, &(xi[WS(is, 1)])); T1P = VSUB(T1N, T1O); T59 = VADD(T1N, T1O); } T1Q = VFMA(LDK(KP923879532), T1M, VMUL(LDK(KP382683432), T1P)); T5a = VSUB(T58, T59); T2a = VFNMS(LDK(KP382683432), T1M, VMUL(LDK(KP923879532), T1P)); } { V T1R, T22, T6k, T6l; T1R = VSUB(T1J, T1Q); T22 = VSUB(T1Y, T21); T23 = VSUB(T1R, T22); T2N = VADD(T22, T1R); T6k = VADD(T5g, T5h); T6l = VADD(T5c, T5d); T6m = VSUB(T6k, T6l); T70 = VADD(T6k, T6l); } { V T6n, T6o, T28, T2b; T6n = VADD(T55, T56); T6o = VADD(T58, T59); T6p = VSUB(T6n, T6o); T71 = VADD(T6n, T6o); T28 = VSUB(T26, T27); T2b = VSUB(T29, T2a); T2c = VSUB(T28, T2b); T2O = VADD(T28, T2b); } { V T3g, T3h, T5b, T5e; T3g = VADD(T26, T27); T3h = VADD(T1J, T1Q); T3i = VADD(T3g, T3h); T3Y = VSUB(T3g, T3h); T5b = VMUL(LDK(KP707106781), VSUB(T57, T5a)); T5e = VSUB(T5c, T5d); T5f = VSUB(T5b, T5e); T5R = VADD(T5e, T5b); } { V T5i, T5j, T3j, T3k; T5i = VSUB(T5g, T5h); T5j = VMUL(LDK(KP707106781), VADD(T57, T5a)); T5k = VSUB(T5i, T5j); T5S = VADD(T5i, T5j); T3j = VADD(T21, T1Y); T3k = VADD(T29, T2a); T3l = VADD(T3j, T3k); T3Z = VSUB(T3k, T3j); } } { V T1q, T50, T1v, T4Z, T1n, T4W, T1w, T4V, T4O, T4P, T18, T4Q, T1y, T4R, T4S; V T1f, T4T, T1z; { V T1o, T1p, T1t, T1u; T1o = LD(&(xi[WS(is, 17)]), ivs, &(xi[WS(is, 1)])); T1p = LD(&(xi[WS(is, 49)]), ivs, &(xi[WS(is, 1)])); T1q = VSUB(T1o, T1p); T50 = VADD(T1o, T1p); T1t = LD(&(xi[WS(is, 1)]), ivs, &(xi[WS(is, 1)])); T1u = LD(&(xi[WS(is, 33)]), ivs, &(xi[WS(is, 1)])); T1v = VSUB(T1t, T1u); T4Z = VADD(T1t, T1u); } { V T1h, T1i, T1j, T1k, T1l, T1m; T1h = LD(&(xi[WS(is, 9)]), ivs, &(xi[WS(is, 1)])); T1i = LD(&(xi[WS(is, 41)]), ivs, &(xi[WS(is, 1)])); T1j = VSUB(T1h, T1i); T1k = LD(&(xi[WS(is, 57)]), ivs, &(xi[WS(is, 1)])); T1l = LD(&(xi[WS(is, 25)]), ivs, &(xi[WS(is, 1)])); T1m = VSUB(T1k, T1l); T1n = VMUL(LDK(KP707106781), VSUB(T1j, T1m)); T4W = VADD(T1k, T1l); T1w = VMUL(LDK(KP707106781), VADD(T1j, T1m)); T4V = VADD(T1h, T1i); } { V T14, T17, T1b, T1e; { V T12, T13, T15, T16; T12 = LD(&(xi[WS(is, 5)]), ivs, &(xi[WS(is, 1)])); T13 = LD(&(xi[WS(is, 37)]), ivs, &(xi[WS(is, 1)])); T14 = VSUB(T12, T13); T4O = VADD(T12, T13); T15 = LD(&(xi[WS(is, 21)]), ivs, &(xi[WS(is, 1)])); T16 = LD(&(xi[WS(is, 53)]), ivs, &(xi[WS(is, 1)])); T17 = VSUB(T15, T16); T4P = VADD(T15, T16); } T18 = VFNMS(LDK(KP382683432), T17, VMUL(LDK(KP923879532), T14)); T4Q = VSUB(T4O, T4P); T1y = VFMA(LDK(KP382683432), T14, VMUL(LDK(KP923879532), T17)); { V T19, T1a, T1c, T1d; T19 = LD(&(xi[WS(is, 61)]), ivs, &(xi[WS(is, 1)])); T1a = LD(&(xi[WS(is, 29)]), ivs, &(xi[WS(is, 1)])); T1b = VSUB(T19, T1a); T4R = VADD(T19, T1a); T1c = LD(&(xi[WS(is, 13)]), ivs, &(xi[WS(is, 1)])); T1d = LD(&(xi[WS(is, 45)]), ivs, &(xi[WS(is, 1)])); T1e = VSUB(T1c, T1d); T4S = VADD(T1c, T1d); } T1f = VFMA(LDK(KP923879532), T1b, VMUL(LDK(KP382683432), T1e)); T4T = VSUB(T4R, T4S); T1z = VFNMS(LDK(KP382683432), T1b, VMUL(LDK(KP923879532), T1e)); } { V T1g, T1r, T6d, T6e; T1g = VSUB(T18, T1f); T1r = VSUB(T1n, T1q); T1s = VSUB(T1g, T1r); T2K = VADD(T1r, T1g); T6d = VADD(T4Z, T50); T6e = VADD(T4V, T4W); T6f = VSUB(T6d, T6e); T6X = VADD(T6d, T6e); } { V T6g, T6h, T1x, T1A; T6g = VADD(T4O, T4P); T6h = VADD(T4R, T4S); T6i = VSUB(T6g, T6h); T6Y = VADD(T6g, T6h); T1x = VSUB(T1v, T1w); T1A = VSUB(T1y, T1z); T1B = VSUB(T1x, T1A); T2L = VADD(T1x, T1A); } { V T39, T3a, T4U, T4X; T39 = VADD(T1v, T1w); T3a = VADD(T18, T1f); T3b = VADD(T39, T3a); T3V = VSUB(T39, T3a); T4U = VMUL(LDK(KP707106781), VSUB(T4Q, T4T)); T4X = VSUB(T4V, T4W); T4Y = VSUB(T4U, T4X); T5O = VADD(T4X, T4U); } { V T51, T52, T3c, T3d; T51 = VSUB(T4Z, T50); T52 = VMUL(LDK(KP707106781), VADD(T4Q, T4T)); T53 = VSUB(T51, T52); T5P = VADD(T51, T52); T3c = VADD(T1q, T1n); T3d = VADD(T1y, T1z); T3e = VADD(T3c, T3d); T3W = VSUB(T3d, T3c); } } { V T7h, T7l, T7k, T7m; { V T7f, T7g, T7i, T7j; T7f = VADD(T78, T79); T7g = VADD(T74, T75); T7h = VSUB(T7f, T7g); T7l = VADD(T7f, T7g); T7i = VADD(T6X, T6Y); T7j = VADD(T70, T71); T7k = VBYI(VSUB(T7i, T7j)); T7m = VADD(T7i, T7j); } ST(&(xo[WS(os, 48)]), VSUB(T7h, T7k), ovs, &(xo[0])); ST(&(xo[0]), VADD(T7l, T7m), ovs, &(xo[0])); ST(&(xo[WS(os, 16)]), VADD(T7h, T7k), ovs, &(xo[0])); ST(&(xo[WS(os, 32)]), VSUB(T7l, T7m), ovs, &(xo[0])); } { V T76, T7a, T73, T7b, T6Z, T72; T76 = VSUB(T74, T75); T7a = VSUB(T78, T79); T6Z = VSUB(T6X, T6Y); T72 = VSUB(T70, T71); T73 = VMUL(LDK(KP707106781), VSUB(T6Z, T72)); T7b = VMUL(LDK(KP707106781), VADD(T6Z, T72)); { V T77, T7c, T7d, T7e; T77 = VBYI(VSUB(T73, T76)); T7c = VSUB(T7a, T7b); ST(&(xo[WS(os, 24)]), VADD(T77, T7c), ovs, &(xo[0])); ST(&(xo[WS(os, 40)]), VSUB(T7c, T77), ovs, &(xo[0])); T7d = VBYI(VADD(T76, T73)); T7e = VADD(T7a, T7b); ST(&(xo[WS(os, 8)]), VADD(T7d, T7e), ovs, &(xo[0])); ST(&(xo[WS(os, 56)]), VSUB(T7e, T7d), ovs, &(xo[0])); } } { V T6C, T6S, T6I, T6P, T6r, T6Q, T6L, T6T, T6y, T6H; T6y = VMUL(LDK(KP707106781), VSUB(T6u, T6x)); T6C = VSUB(T6y, T6B); T6S = VADD(T6B, T6y); T6H = VMUL(LDK(KP707106781), VADD(T6u, T6x)); T6I = VSUB(T6G, T6H); T6P = VADD(T6G, T6H); { V T6j, T6q, T6J, T6K; T6j = VFNMS(LDK(KP382683432), T6i, VMUL(LDK(KP923879532), T6f)); T6q = VFMA(LDK(KP923879532), T6m, VMUL(LDK(KP382683432), T6p)); T6r = VSUB(T6j, T6q); T6Q = VADD(T6j, T6q); T6J = VFMA(LDK(KP382683432), T6f, VMUL(LDK(KP923879532), T6i)); T6K = VFNMS(LDK(KP382683432), T6m, VMUL(LDK(KP923879532), T6p)); T6L = VSUB(T6J, T6K); T6T = VADD(T6J, T6K); } { V T6D, T6M, T6V, T6W; T6D = VBYI(VSUB(T6r, T6C)); T6M = VSUB(T6I, T6L); ST(&(xo[WS(os, 20)]), VADD(T6D, T6M), ovs, &(xo[0])); ST(&(xo[WS(os, 44)]), VSUB(T6M, T6D), ovs, &(xo[0])); T6V = VSUB(T6P, T6Q); T6W = VBYI(VSUB(T6T, T6S)); ST(&(xo[WS(os, 36)]), VSUB(T6V, T6W), ovs, &(xo[0])); ST(&(xo[WS(os, 28)]), VADD(T6V, T6W), ovs, &(xo[0])); } { V T6N, T6O, T6R, T6U; T6N = VBYI(VADD(T6C, T6r)); T6O = VADD(T6I, T6L); ST(&(xo[WS(os, 12)]), VADD(T6N, T6O), ovs, &(xo[0])); ST(&(xo[WS(os, 52)]), VSUB(T6O, T6N), ovs, &(xo[0])); T6R = VADD(T6P, T6Q); T6U = VBYI(VADD(T6S, T6T)); ST(&(xo[WS(os, 60)]), VSUB(T6R, T6U), ovs, &(xo[0])); ST(&(xo[WS(os, 4)]), VADD(T6R, T6U), ovs, &(xo[0])); } } { V T5N, T68, T61, T69, T5U, T65, T5Y, T66; { V T5L, T5M, T5Z, T60; T5L = VADD(T4p, T4w); T5M = VADD(T5o, T5p); T5N = VSUB(T5L, T5M); T68 = VADD(T5L, T5M); T5Z = VFNMS(LDK(KP195090322), T5O, VMUL(LDK(KP980785280), T5P)); T60 = VFMA(LDK(KP195090322), T5R, VMUL(LDK(KP980785280), T5S)); T61 = VSUB(T5Z, T60); T69 = VADD(T5Z, T60); } { V T5Q, T5T, T5W, T5X; T5Q = VFMA(LDK(KP980785280), T5O, VMUL(LDK(KP195090322), T5P)); T5T = VFNMS(LDK(KP195090322), T5S, VMUL(LDK(KP980785280), T5R)); T5U = VSUB(T5Q, T5T); T65 = VADD(T5Q, T5T); T5W = VADD(T4E, T4L); T5X = VADD(T5u, T5r); T5Y = VSUB(T5W, T5X); T66 = VADD(T5X, T5W); } { V T5V, T62, T6b, T6c; T5V = VADD(T5N, T5U); T62 = VBYI(VADD(T5Y, T61)); ST(&(xo[WS(os, 50)]), VSUB(T5V, T62), ovs, &(xo[0])); ST(&(xo[WS(os, 14)]), VADD(T5V, T62), ovs, &(xo[0])); T6b = VBYI(VADD(T66, T65)); T6c = VADD(T68, T69); ST(&(xo[WS(os, 2)]), VADD(T6b, T6c), ovs, &(xo[0])); ST(&(xo[WS(os, 62)]), VSUB(T6c, T6b), ovs, &(xo[0])); } { V T63, T64, T67, T6a; T63 = VSUB(T5N, T5U); T64 = VBYI(VSUB(T61, T5Y)); ST(&(xo[WS(os, 46)]), VSUB(T63, T64), ovs, &(xo[0])); ST(&(xo[WS(os, 18)]), VADD(T63, T64), ovs, &(xo[0])); T67 = VBYI(VSUB(T65, T66)); T6a = VSUB(T68, T69); ST(&(xo[WS(os, 30)]), VADD(T67, T6a), ovs, &(xo[0])); ST(&(xo[WS(os, 34)]), VSUB(T6a, T67), ovs, &(xo[0])); } } { V T11, T2C, T2v, T2D, T2e, T2z, T2s, T2A; { V Tr, T10, T2t, T2u; Tr = VSUB(Tb, Tq); T10 = VSUB(TI, TZ); T11 = VSUB(Tr, T10); T2C = VADD(Tr, T10); T2t = VFNMS(LDK(KP471396736), T1s, VMUL(LDK(KP881921264), T1B)); T2u = VFMA(LDK(KP471396736), T23, VMUL(LDK(KP881921264), T2c)); T2v = VSUB(T2t, T2u); T2D = VADD(T2t, T2u); } { V T1C, T2d, T2i, T2r; T1C = VFMA(LDK(KP881921264), T1s, VMUL(LDK(KP471396736), T1B)); T2d = VFNMS(LDK(KP471396736), T2c, VMUL(LDK(KP881921264), T23)); T2e = VSUB(T1C, T2d); T2z = VADD(T1C, T2d); T2i = VSUB(T2g, T2h); T2r = VSUB(T2l, T2q); T2s = VSUB(T2i, T2r); T2A = VADD(T2r, T2i); } { V T2f, T2w, T2F, T2G; T2f = VADD(T11, T2e); T2w = VBYI(VADD(T2s, T2v)); ST(&(xo[WS(os, 53)]), VSUB(T2f, T2w), ovs, &(xo[WS(os, 1)])); ST(&(xo[WS(os, 11)]), VADD(T2f, T2w), ovs, &(xo[WS(os, 1)])); T2F = VBYI(VADD(T2A, T2z)); T2G = VADD(T2C, T2D); ST(&(xo[WS(os, 5)]), VADD(T2F, T2G), ovs, &(xo[WS(os, 1)])); ST(&(xo[WS(os, 59)]), VSUB(T2G, T2F), ovs, &(xo[WS(os, 1)])); } { V T2x, T2y, T2B, T2E; T2x = VSUB(T11, T2e); T2y = VBYI(VSUB(T2v, T2s)); ST(&(xo[WS(os, 43)]), VSUB(T2x, T2y), ovs, &(xo[WS(os, 1)])); ST(&(xo[WS(os, 21)]), VADD(T2x, T2y), ovs, &(xo[WS(os, 1)])); T2B = VBYI(VSUB(T2z, T2A)); T2E = VSUB(T2C, T2D); ST(&(xo[WS(os, 27)]), VADD(T2B, T2E), ovs, &(xo[WS(os, 1)])); ST(&(xo[WS(os, 37)]), VSUB(T2E, T2B), ovs, &(xo[WS(os, 1)])); } } { V T3n, T3O, T3J, T3R, T3y, T3Q, T3G, T3N; { V T3f, T3m, T3H, T3I; T3f = VFNMS(LDK(KP098017140), T3e, VMUL(LDK(KP995184726), T3b)); T3m = VFMA(LDK(KP995184726), T3i, VMUL(LDK(KP098017140), T3l)); T3n = VSUB(T3f, T3m); T3O = VADD(T3f, T3m); T3H = VFMA(LDK(KP098017140), T3b, VMUL(LDK(KP995184726), T3e)); T3I = VFNMS(LDK(KP098017140), T3i, VMUL(LDK(KP995184726), T3l)); T3J = VSUB(T3H, T3I); T3R = VADD(T3H, T3I); } { V T3u, T3x, T3C, T3F; T3u = VADD(T3q, T3t); T3x = VADD(T3v, T3w); T3y = VSUB(T3u, T3x); T3Q = VADD(T3x, T3u); T3C = VADD(T3A, T3B); T3F = VADD(T3D, T3E); T3G = VSUB(T3C, T3F); T3N = VADD(T3C, T3F); } { V T3z, T3K, T3T, T3U; T3z = VBYI(VSUB(T3n, T3y)); T3K = VSUB(T3G, T3J); ST(&(xo[WS(os, 17)]), VADD(T3z, T3K), ovs, &(xo[WS(os, 1)])); ST(&(xo[WS(os, 47)]), VSUB(T3K, T3z), ovs, &(xo[WS(os, 1)])); T3T = VSUB(T3N, T3O); T3U = VBYI(VSUB(T3R, T3Q)); ST(&(xo[WS(os, 33)]), VSUB(T3T, T3U), ovs, &(xo[WS(os, 1)])); ST(&(xo[WS(os, 31)]), VADD(T3T, T3U), ovs, &(xo[WS(os, 1)])); } { V T3L, T3M, T3P, T3S; T3L = VBYI(VADD(T3y, T3n)); T3M = VADD(T3G, T3J); ST(&(xo[WS(os, 15)]), VADD(T3L, T3M), ovs, &(xo[WS(os, 1)])); ST(&(xo[WS(os, 49)]), VSUB(T3M, T3L), ovs, &(xo[WS(os, 1)])); T3P = VADD(T3N, T3O); T3S = VBYI(VADD(T3Q, T3R)); ST(&(xo[WS(os, 63)]), VSUB(T3P, T3S), ovs, &(xo[WS(os, 1)])); ST(&(xo[WS(os, 1)]), VADD(T3P, T3S), ovs, &(xo[WS(os, 1)])); } } { V T4N, T5G, T5z, T5H, T5m, T5D, T5w, T5E; { V T4x, T4M, T5x, T5y; T4x = VSUB(T4p, T4w); T4M = VSUB(T4E, T4L); T4N = VSUB(T4x, T4M); T5G = VADD(T4x, T4M); T5x = VFNMS(LDK(KP555570233), T4Y, VMUL(LDK(KP831469612), T53)); T5y = VFMA(LDK(KP555570233), T5f, VMUL(LDK(KP831469612), T5k)); T5z = VSUB(T5x, T5y); T5H = VADD(T5x, T5y); } { V T54, T5l, T5q, T5v; T54 = VFMA(LDK(KP831469612), T4Y, VMUL(LDK(KP555570233), T53)); T5l = VFNMS(LDK(KP555570233), T5k, VMUL(LDK(KP831469612), T5f)); T5m = VSUB(T54, T5l); T5D = VADD(T54, T5l); T5q = VSUB(T5o, T5p); T5v = VSUB(T5r, T5u); T5w = VSUB(T5q, T5v); T5E = VADD(T5v, T5q); } { V T5n, T5A, T5J, T5K; T5n = VADD(T4N, T5m); T5A = VBYI(VADD(T5w, T5z)); ST(&(xo[WS(os, 54)]), VSUB(T5n, T5A), ovs, &(xo[0])); ST(&(xo[WS(os, 10)]), VADD(T5n, T5A), ovs, &(xo[0])); T5J = VBYI(VADD(T5E, T5D)); T5K = VADD(T5G, T5H); ST(&(xo[WS(os, 6)]), VADD(T5J, T5K), ovs, &(xo[0])); ST(&(xo[WS(os, 58)]), VSUB(T5K, T5J), ovs, &(xo[0])); } { V T5B, T5C, T5F, T5I; T5B = VSUB(T4N, T5m); T5C = VBYI(VSUB(T5z, T5w)); ST(&(xo[WS(os, 42)]), VSUB(T5B, T5C), ovs, &(xo[0])); ST(&(xo[WS(os, 22)]), VADD(T5B, T5C), ovs, &(xo[0])); T5F = VBYI(VSUB(T5D, T5E)); T5I = VSUB(T5G, T5H); ST(&(xo[WS(os, 26)]), VADD(T5F, T5I), ovs, &(xo[0])); ST(&(xo[WS(os, 38)]), VSUB(T5I, T5F), ovs, &(xo[0])); } } { V T2J, T34, T2X, T35, T2Q, T31, T2U, T32; { V T2H, T2I, T2V, T2W; T2H = VADD(Tb, Tq); T2I = VADD(T2g, T2h); T2J = VSUB(T2H, T2I); T34 = VADD(T2H, T2I); T2V = VFNMS(LDK(KP290284677), T2K, VMUL(LDK(KP956940335), T2L)); T2W = VFMA(LDK(KP290284677), T2N, VMUL(LDK(KP956940335), T2O)); T2X = VSUB(T2V, T2W); T35 = VADD(T2V, T2W); } { V T2M, T2P, T2S, T2T; T2M = VFMA(LDK(KP956940335), T2K, VMUL(LDK(KP290284677), T2L)); T2P = VFNMS(LDK(KP290284677), T2O, VMUL(LDK(KP956940335), T2N)); T2Q = VSUB(T2M, T2P); T31 = VADD(T2M, T2P); T2S = VADD(TI, TZ); T2T = VADD(T2q, T2l); T2U = VSUB(T2S, T2T); T32 = VADD(T2T, T2S); } { V T2R, T2Y, T37, T38; T2R = VADD(T2J, T2Q); T2Y = VBYI(VADD(T2U, T2X)); ST(&(xo[WS(os, 51)]), VSUB(T2R, T2Y), ovs, &(xo[WS(os, 1)])); ST(&(xo[WS(os, 13)]), VADD(T2R, T2Y), ovs, &(xo[WS(os, 1)])); T37 = VBYI(VADD(T32, T31)); T38 = VADD(T34, T35); ST(&(xo[WS(os, 3)]), VADD(T37, T38), ovs, &(xo[WS(os, 1)])); ST(&(xo[WS(os, 61)]), VSUB(T38, T37), ovs, &(xo[WS(os, 1)])); } { V T2Z, T30, T33, T36; T2Z = VSUB(T2J, T2Q); T30 = VBYI(VSUB(T2X, T2U)); ST(&(xo[WS(os, 45)]), VSUB(T2Z, T30), ovs, &(xo[WS(os, 1)])); ST(&(xo[WS(os, 19)]), VADD(T2Z, T30), ovs, &(xo[WS(os, 1)])); T33 = VBYI(VSUB(T31, T32)); T36 = VSUB(T34, T35); ST(&(xo[WS(os, 29)]), VADD(T33, T36), ovs, &(xo[WS(os, 1)])); ST(&(xo[WS(os, 35)]), VSUB(T36, T33), ovs, &(xo[WS(os, 1)])); } } { V T41, T4g, T4b, T4j, T44, T4i, T48, T4f; { V T3X, T40, T49, T4a; T3X = VFNMS(LDK(KP634393284), T3W, VMUL(LDK(KP773010453), T3V)); T40 = VFMA(LDK(KP773010453), T3Y, VMUL(LDK(KP634393284), T3Z)); T41 = VSUB(T3X, T40); T4g = VADD(T3X, T40); T49 = VFMA(LDK(KP634393284), T3V, VMUL(LDK(KP773010453), T3W)); T4a = VFNMS(LDK(KP634393284), T3Y, VMUL(LDK(KP773010453), T3Z)); T4b = VSUB(T49, T4a); T4j = VADD(T49, T4a); } { V T42, T43, T46, T47; T42 = VSUB(T3D, T3E); T43 = VSUB(T3w, T3v); T44 = VSUB(T42, T43); T4i = VADD(T43, T42); T46 = VSUB(T3A, T3B); T47 = VSUB(T3q, T3t); T48 = VSUB(T46, T47); T4f = VADD(T46, T47); } { V T45, T4c, T4l, T4m; T45 = VBYI(VSUB(T41, T44)); T4c = VSUB(T48, T4b); ST(&(xo[WS(os, 23)]), VADD(T45, T4c), ovs, &(xo[WS(os, 1)])); ST(&(xo[WS(os, 41)]), VSUB(T4c, T45), ovs, &(xo[WS(os, 1)])); T4l = VSUB(T4f, T4g); T4m = VBYI(VSUB(T4j, T4i)); ST(&(xo[WS(os, 39)]), VSUB(T4l, T4m), ovs, &(xo[WS(os, 1)])); ST(&(xo[WS(os, 25)]), VADD(T4l, T4m), ovs, &(xo[WS(os, 1)])); } { V T4d, T4e, T4h, T4k; T4d = VBYI(VADD(T44, T41)); T4e = VADD(T48, T4b); ST(&(xo[WS(os, 9)]), VADD(T4d, T4e), ovs, &(xo[WS(os, 1)])); ST(&(xo[WS(os, 55)]), VSUB(T4e, T4d), ovs, &(xo[WS(os, 1)])); T4h = VADD(T4f, T4g); T4k = VBYI(VADD(T4i, T4j)); ST(&(xo[WS(os, 57)]), VSUB(T4h, T4k), ovs, &(xo[WS(os, 1)])); ST(&(xo[WS(os, 7)]), VADD(T4h, T4k), ovs, &(xo[WS(os, 1)])); } } } } VLEAVE(); } static const kdft_desc desc = { 64, XSIMD_STRING("n1bv_64"), {404, 72, 52, 0}, &GENUS, 0, 0, 0, 0 }; void XSIMD(codelet_n1bv_64) (planner *p) { X(kdft_register) (p, n1bv_64, &desc); } #endif /* HAVE_FMA */ fftw-3.3.3/dft/simd/common/n1bv_128.c0000644000175000017500000037766012054410566014042 00000000000000/* * Copyright (c) 2003, 2007-11 Matteo Frigo * Copyright (c) 2003, 2007-11 Massachusetts Institute of Technology * * This program is free software; you can redistribute it and/or modify * it under the terms of the GNU General Public License as published by * the Free Software Foundation; either version 2 of the License, or * (at your option) any later version. * * This program is distributed in the hope that it will be useful, * but WITHOUT ANY WARRANTY; without even the implied warranty of * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the * GNU General Public License for more details. * * You should have received a copy of the GNU General Public License * along with this program; if not, write to the Free Software * Foundation, Inc., 51 Franklin Street, Fifth Floor, Boston, MA 02110-1301 USA * */ /* This file was automatically generated --- DO NOT EDIT */ /* Generated on Sun Nov 25 07:37:08 EST 2012 */ #include "codelet-dft.h" #ifdef HAVE_FMA /* Generated by: ../../../genfft/gen_notw_c.native -fma -reorder-insns -schedule-for-pipeline -simd -compact -variables 4 -pipeline-latency 8 -sign 1 -n 128 -name n1bv_128 -include n1b.h */ /* * This function contains 1082 FP additions, 642 FP multiplications, * (or, 440 additions, 0 multiplications, 642 fused multiply/add), * 295 stack variables, 31 constants, and 256 memory accesses */ #include "n1b.h" static void n1bv_128(const R *ri, const R *ii, R *ro, R *io, stride is, stride os, INT v, INT ivs, INT ovs) { DVK(KP903989293, +0.903989293123443331586200297230537048710132025); DVK(KP941544065, +0.941544065183020778412509402599502357185589796); DVK(KP357805721, +0.357805721314524104672487743774474392487532769); DVK(KP472964775, +0.472964775891319928124438237972992463904131113); DVK(KP857728610, +0.857728610000272069902269984284770137042490799); DVK(KP970031253, +0.970031253194543992603984207286100251456865962); DVK(KP250486960, +0.250486960191305461595702160124721208578685568); DVK(KP998795456, +0.998795456205172392714771604759100694443203615); DVK(KP740951125, +0.740951125354959091175616897495162729728955309); DVK(KP599376933, +0.599376933681923766271389869014404232837890546); DVK(KP906347169, +0.906347169019147157946142717268914412664134293); DVK(KP049126849, +0.049126849769467254105343321271313617079695752); DVK(KP989176509, +0.989176509964780973451673738016243063983689533); DVK(KP803207531, +0.803207531480644909806676512963141923879569427); DVK(KP741650546, +0.741650546272035369581266691172079863842265220); DVK(KP956940335, +0.956940335732208864935797886980269969482849206); DVK(KP148335987, +0.148335987538347428753676511486911367000625355); DVK(KP881921264, +0.881921264348355029712756863660388349508442621); DVK(KP303346683, +0.303346683607342391675883946941299872384187453); DVK(KP534511135, +0.534511135950791641089685961295362908582039528); DVK(KP773010453, +0.773010453362736960810906609758469800971041293); DVK(KP995184726, +0.995184726672196886244836953109479921575474869); DVK(KP820678790, +0.820678790828660330972281985331011598767386482); DVK(KP098491403, +0.098491403357164253077197521291327432293052451); DVK(KP831469612, +0.831469612302545237078788377617905756738560812); DVK(KP980785280, +0.980785280403230449126182236134239036973933731); DVK(KP198912367, +0.198912367379658006911597622644676228597850501); DVK(KP668178637, +0.668178637919298919997757686523080761552472251); DVK(KP923879532, +0.923879532511286756128183189396788286822416626); DVK(KP414213562, +0.414213562373095048801688724209698078569671875); DVK(KP707106781, +0.707106781186547524400844362104849039284835938); { INT i; const R *xi; R *xo; xi = ii; xo = io; for (i = v; i > 0; i = i - VL, xi = xi + (VL * ivs), xo = xo + (VL * ovs), MAKE_VOLATILE_STRIDE(256, is), MAKE_VOLATILE_STRIDE(256, os)) { V T6a, T5J, T6b, T5K, T6B, T6C, T6J, T6A, T6o, T6j, T6r, T68, T6d, T5O, T5R; V T6e, T6D, T6K; { V Tad, TcZ, T6Z, T8T, T4U, Tr, Tfq, TgG, Ted, Tgf, Td0, Tcc, T9k, T84, Tb6; V Tbt, Td8, TdK, TeK, Tgq, TeV, Tgt, T7q, T94, T3p, T5X, T7B, T97, T2G, T5U; V TbD, Tc0, Tdf, TdN, Tf5, Tgx, Tfg, TgA, T7J, T9b, T4E, T64, T7U, T9e, T3V; V T61, Td2, Td3, T85, T72, T4V, TI, Tcd, Tas, TgH, Tek, Tgg, Tft, T86, T75; V T4W, TZ, TaI, Tcf, Tdo, TdG, Tgi, Tet, Tgj, Teq, T8X, T7a, T5M, T1B, T8W; V T7d, T5N, T1s, TaX, Tcg, Tdr, TdH, Tgl, TeC, Tgm, Tez, T90, T7h, T5P, T2c; V T8Z, T7k, T5Q, T23, T3Y, T49, TdL, Tdb, Tbu, Tbl, Tgu, TeR, Tgr, TeY, Tf6; V TbG, T5V, T3s, T5Y, T3f, T95, T7E, T98, T7x, T4g, T4f, T4q, TbH, T41, TbI; V T44, T4h, T4j, T4k, Tf9, TbN; { V Tu, TF, Ty, TL, TW, Tah, Tx, Tag, Tee, Tz, TM, TN, Teh, Tan, TP; V TQ; { V TeG, T2A, Tbq, TeT, Tbp, TeH, T3m, T2x, Td6, T7o, T2q, T3l, T7z, Tbr, T2D; V T82, T83; { V Ta7, T3, Ta8, T4O, Taa, Tab, Ta, T4P, Te, Tc6, Th, Tc7, Tl, Tc9, Tca; V To; { V T1, T2, T4M, T4N; T1 = LD(&(xi[0]), ivs, &(xi[0])); T2 = LD(&(xi[WS(is, 64)]), ivs, &(xi[0])); T4M = LD(&(xi[WS(is, 32)]), ivs, &(xi[0])); T4N = LD(&(xi[WS(is, 96)]), ivs, &(xi[0])); { V T4, T5, T7, T8; T4 = LD(&(xi[WS(is, 16)]), ivs, &(xi[0])); T5 = LD(&(xi[WS(is, 80)]), ivs, &(xi[0])); T7 = LD(&(xi[WS(is, 112)]), ivs, &(xi[0])); T8 = LD(&(xi[WS(is, 48)]), ivs, &(xi[0])); { V Tc, T6, T9, Td, Tf, Tg; Tc = LD(&(xi[WS(is, 8)]), ivs, &(xi[0])); Ta7 = VADD(T1, T2); T3 = VSUB(T1, T2); Ta8 = VADD(T4M, T4N); T4O = VSUB(T4M, T4N); Taa = VADD(T4, T5); T6 = VSUB(T4, T5); Tab = VADD(T7, T8); T9 = VSUB(T7, T8); Td = LD(&(xi[WS(is, 72)]), ivs, &(xi[0])); Tf = LD(&(xi[WS(is, 40)]), ivs, &(xi[0])); Tg = LD(&(xi[WS(is, 104)]), ivs, &(xi[0])); { V Tj, Tk, Tm, Tn; Tj = LD(&(xi[WS(is, 120)]), ivs, &(xi[0])); Tk = LD(&(xi[WS(is, 56)]), ivs, &(xi[0])); Tm = LD(&(xi[WS(is, 88)]), ivs, &(xi[0])); Tn = LD(&(xi[WS(is, 24)]), ivs, &(xi[0])); Ta = VADD(T6, T9); T4P = VSUB(T6, T9); Te = VSUB(Tc, Td); Tc6 = VADD(Tc, Td); Th = VSUB(Tf, Tg); Tc7 = VADD(Tf, Tg); Tl = VSUB(Tj, Tk); Tc9 = VADD(Tj, Tk); Tca = VADD(Tn, Tm); To = VSUB(Tm, Tn); } } } } { V T6X, Tb, Te9, Ta9, Tc8, Tea, T4R, Ti, Tfo, Tac, Tp, T4S, Tcb, Teb, T4Q; T6X = VFNMS(LDK(KP707106781), Ta, T3); Tb = VFMA(LDK(KP707106781), Ta, T3); Te9 = VSUB(Ta7, Ta8); Ta9 = VADD(Ta7, Ta8); Tc8 = VADD(Tc6, Tc7); Tea = VSUB(Tc6, Tc7); T4R = VFMA(LDK(KP414213562), Te, Th); Ti = VFNMS(LDK(KP414213562), Th, Te); Tfo = VSUB(Taa, Tab); Tac = VADD(Taa, Tab); Tp = VFNMS(LDK(KP414213562), To, Tl); T4S = VFMA(LDK(KP414213562), Tl, To); Tcb = VADD(Tc9, Tca); Teb = VSUB(Tc9, Tca); T4Q = VFMA(LDK(KP707106781), T4P, T4O); T82 = VFNMS(LDK(KP707106781), T4P, T4O); { V T4T, T6Y, Tq, Tfp, Tec; T4T = VSUB(T4R, T4S); T6Y = VADD(T4R, T4S); T83 = VSUB(Ti, Tp); Tq = VADD(Ti, Tp); Tfp = VSUB(Tea, Teb); Tec = VADD(Tea, Teb); Tad = VSUB(Ta9, Tac); TcZ = VADD(Ta9, Tac); T6Z = VFMA(LDK(KP923879532), T6Y, T6X); T8T = VFNMS(LDK(KP923879532), T6Y, T6X); T4U = VFMA(LDK(KP923879532), T4T, T4Q); T6a = VFNMS(LDK(KP923879532), T4T, T4Q); Tr = VFMA(LDK(KP923879532), Tq, Tb); T5J = VFNMS(LDK(KP923879532), Tq, Tb); Tfq = VFMA(LDK(KP707106781), Tfp, Tfo); TgG = VFNMS(LDK(KP707106781), Tfp, Tfo); Ted = VFMA(LDK(KP707106781), Tec, Te9); Tgf = VFNMS(LDK(KP707106781), Tec, Te9); Td0 = VADD(Tc8, Tcb); Tcc = VSUB(Tc8, Tcb); } } } { V T2i, T3j, Tb2, T2B, Tb5, T3k, T2p, T2C; { V T2m, Tb0, Tb1, Tb3, T2l, T2n; { V T2g, T2h, T3h, T3i, T2j, T2k; T2g = LD(&(xi[WS(is, 1)]), ivs, &(xi[WS(is, 1)])); T2h = LD(&(xi[WS(is, 65)]), ivs, &(xi[WS(is, 1)])); T3h = LD(&(xi[WS(is, 33)]), ivs, &(xi[WS(is, 1)])); T3i = LD(&(xi[WS(is, 97)]), ivs, &(xi[WS(is, 1)])); T2j = LD(&(xi[WS(is, 17)]), ivs, &(xi[WS(is, 1)])); T2k = LD(&(xi[WS(is, 81)]), ivs, &(xi[WS(is, 1)])); T2m = LD(&(xi[WS(is, 113)]), ivs, &(xi[WS(is, 1)])); T9k = VFMA(LDK(KP923879532), T83, T82); T84 = VFNMS(LDK(KP923879532), T83, T82); T2i = VSUB(T2g, T2h); Tb0 = VADD(T2g, T2h); T3j = VSUB(T3h, T3i); Tb1 = VADD(T3h, T3i); Tb3 = VADD(T2j, T2k); T2l = VSUB(T2j, T2k); T2n = LD(&(xi[WS(is, 49)]), ivs, &(xi[WS(is, 1)])); } { V T2r, T2s, T2u, T2v; T2r = LD(&(xi[WS(is, 9)]), ivs, &(xi[WS(is, 1)])); T2s = LD(&(xi[WS(is, 73)]), ivs, &(xi[WS(is, 1)])); T2u = LD(&(xi[WS(is, 41)]), ivs, &(xi[WS(is, 1)])); T2v = LD(&(xi[WS(is, 105)]), ivs, &(xi[WS(is, 1)])); TeG = VSUB(Tb0, Tb1); Tb2 = VADD(Tb0, Tb1); { V T2y, T2z, Tb4, T2o, Tbn, T2t, Tbo, T2w; T2y = LD(&(xi[WS(is, 121)]), ivs, &(xi[WS(is, 1)])); T2z = LD(&(xi[WS(is, 57)]), ivs, &(xi[WS(is, 1)])); Tb4 = VADD(T2m, T2n); T2o = VSUB(T2m, T2n); Tbn = VADD(T2r, T2s); T2t = VSUB(T2r, T2s); Tbo = VADD(T2u, T2v); T2w = VSUB(T2u, T2v); T2B = LD(&(xi[WS(is, 25)]), ivs, &(xi[WS(is, 1)])); T2A = VSUB(T2y, T2z); Tbq = VADD(T2y, T2z); TeT = VSUB(Tb3, Tb4); Tb5 = VADD(Tb3, Tb4); T3k = VSUB(T2l, T2o); T2p = VADD(T2l, T2o); Tbp = VADD(Tbn, Tbo); TeH = VSUB(Tbn, Tbo); T3m = VFMA(LDK(KP414213562), T2t, T2w); T2x = VFNMS(LDK(KP414213562), T2w, T2t); T2C = LD(&(xi[WS(is, 89)]), ivs, &(xi[WS(is, 1)])); } } } Td6 = VADD(Tb2, Tb5); Tb6 = VSUB(Tb2, Tb5); T7o = VFNMS(LDK(KP707106781), T2p, T2i); T2q = VFMA(LDK(KP707106781), T2p, T2i); T3l = VFMA(LDK(KP707106781), T3k, T3j); T7z = VFNMS(LDK(KP707106781), T3k, T3j); Tbr = VADD(T2B, T2C); T2D = VSUB(T2B, T2C); } { V Tf1, Tfe, Tf2, TbZ, T3M, T4B, Tdd, T3F, T7H, T4A, T7S, TbW, Tf3, T4C, T3T; { V T3x, T4y, Tbz, T3Q, TbC, T4z, T3E, T3R, T3P, TbU, TbV, T3S; { V T3y, T3z, T3B, T3C; { V T3v, T3w, T4w, T4x; T3v = LD(&(xi[WS(is, 127)]), ivs, &(xi[WS(is, 1)])); T3w = LD(&(xi[WS(is, 63)]), ivs, &(xi[WS(is, 1)])); T4w = LD(&(xi[WS(is, 95)]), ivs, &(xi[WS(is, 1)])); T4x = LD(&(xi[WS(is, 31)]), ivs, &(xi[WS(is, 1)])); T3y = LD(&(xi[WS(is, 15)]), ivs, &(xi[WS(is, 1)])); { V Tbs, TeI, T3n, T2E, Tbx; Tbs = VADD(Tbq, Tbr); TeI = VSUB(Tbq, Tbr); T3n = VFNMS(LDK(KP414213562), T2A, T2D); T2E = VFMA(LDK(KP414213562), T2D, T2A); T3x = VSUB(T3v, T3w); Tbx = VADD(T3v, T3w); { V Tby, Td7, TeJ, TeU; T4y = VSUB(T4w, T4x); Tby = VADD(T4x, T4w); Td7 = VADD(Tbp, Tbs); Tbt = VSUB(Tbp, Tbs); TeJ = VADD(TeH, TeI); TeU = VSUB(TeH, TeI); { V T7p, T3o, T7A, T2F; T7p = VSUB(T3m, T3n); T3o = VADD(T3m, T3n); T7A = VSUB(T2x, T2E); T2F = VADD(T2x, T2E); Tbz = VADD(Tbx, Tby); Tf1 = VSUB(Tbx, Tby); Td8 = VADD(Td6, Td7); TdK = VSUB(Td6, Td7); TeK = VFMA(LDK(KP707106781), TeJ, TeG); Tgq = VFNMS(LDK(KP707106781), TeJ, TeG); TeV = VFMA(LDK(KP707106781), TeU, TeT); Tgt = VFNMS(LDK(KP707106781), TeU, TeT); T7q = VFMA(LDK(KP923879532), T7p, T7o); T94 = VFNMS(LDK(KP923879532), T7p, T7o); T3p = VFMA(LDK(KP923879532), T3o, T3l); T5X = VFNMS(LDK(KP923879532), T3o, T3l); T7B = VFNMS(LDK(KP923879532), T7A, T7z); T97 = VFMA(LDK(KP923879532), T7A, T7z); T2G = VFMA(LDK(KP923879532), T2F, T2q); T5U = VFNMS(LDK(KP923879532), T2F, T2q); T3z = LD(&(xi[WS(is, 79)]), ivs, &(xi[WS(is, 1)])); } } } T3B = LD(&(xi[WS(is, 111)]), ivs, &(xi[WS(is, 1)])); T3C = LD(&(xi[WS(is, 47)]), ivs, &(xi[WS(is, 1)])); } { V T3G, T3H, T3J, T3K; T3G = LD(&(xi[WS(is, 7)]), ivs, &(xi[WS(is, 1)])); T3H = LD(&(xi[WS(is, 71)]), ivs, &(xi[WS(is, 1)])); T3J = LD(&(xi[WS(is, 103)]), ivs, &(xi[WS(is, 1)])); T3K = LD(&(xi[WS(is, 39)]), ivs, &(xi[WS(is, 1)])); { V T3N, T3A, TbA, T3D, TbB, T3I, TbX, T3L, TbY, T3O; T3N = LD(&(xi[WS(is, 119)]), ivs, &(xi[WS(is, 1)])); T3A = VSUB(T3y, T3z); TbA = VADD(T3y, T3z); T3D = VSUB(T3B, T3C); TbB = VADD(T3B, T3C); T3I = VSUB(T3G, T3H); TbX = VADD(T3G, T3H); T3L = VSUB(T3J, T3K); TbY = VADD(T3K, T3J); T3O = LD(&(xi[WS(is, 55)]), ivs, &(xi[WS(is, 1)])); T3Q = LD(&(xi[WS(is, 87)]), ivs, &(xi[WS(is, 1)])); Tfe = VSUB(TbB, TbA); TbC = VADD(TbA, TbB); T4z = VSUB(T3D, T3A); T3E = VADD(T3A, T3D); T3R = LD(&(xi[WS(is, 23)]), ivs, &(xi[WS(is, 1)])); Tf2 = VSUB(TbX, TbY); TbZ = VADD(TbX, TbY); T3M = VFMA(LDK(KP414213562), T3L, T3I); T4B = VFNMS(LDK(KP414213562), T3I, T3L); T3P = VSUB(T3N, T3O); TbU = VADD(T3N, T3O); } } } Tdd = VADD(Tbz, TbC); TbD = VSUB(Tbz, TbC); TbV = VADD(T3R, T3Q); T3S = VSUB(T3Q, T3R); T3F = VFMA(LDK(KP707106781), T3E, T3x); T7H = VFNMS(LDK(KP707106781), T3E, T3x); T4A = VFMA(LDK(KP707106781), T4z, T4y); T7S = VFNMS(LDK(KP707106781), T4z, T4y); TbW = VADD(TbU, TbV); Tf3 = VSUB(TbU, TbV); T4C = VFMA(LDK(KP414213562), T3P, T3S); T3T = VFNMS(LDK(KP414213562), T3S, T3P); } { V TD, Tae, TE, TJ, TK, TU, TV; { V Ts, Tt, Tde, Tf4, Tff; Ts = LD(&(xi[WS(is, 4)]), ivs, &(xi[0])); Tt = LD(&(xi[WS(is, 68)]), ivs, &(xi[0])); TD = LD(&(xi[WS(is, 36)]), ivs, &(xi[0])); Tde = VADD(TbZ, TbW); Tc0 = VSUB(TbW, TbZ); Tf4 = VADD(Tf2, Tf3); Tff = VSUB(Tf3, Tf2); { V T7I, T4D, T7T, T3U; T7I = VSUB(T4C, T4B); T4D = VADD(T4B, T4C); T7T = VSUB(T3T, T3M); T3U = VADD(T3M, T3T); Tae = VADD(Ts, Tt); Tu = VSUB(Ts, Tt); Tdf = VADD(Tdd, Tde); TdN = VSUB(Tdd, Tde); Tf5 = VFMA(LDK(KP707106781), Tf4, Tf1); Tgx = VFNMS(LDK(KP707106781), Tf4, Tf1); Tfg = VFMA(LDK(KP707106781), Tff, Tfe); TgA = VFNMS(LDK(KP707106781), Tff, Tfe); T7J = VFMA(LDK(KP923879532), T7I, T7H); T9b = VFNMS(LDK(KP923879532), T7I, T7H); T4E = VFMA(LDK(KP923879532), T4D, T4A); T64 = VFNMS(LDK(KP923879532), T4D, T4A); T7U = VFNMS(LDK(KP923879532), T7T, T7S); T9e = VFMA(LDK(KP923879532), T7T, T7S); T3V = VFMA(LDK(KP923879532), T3U, T3F); T61 = VFNMS(LDK(KP923879532), T3U, T3F); TE = LD(&(xi[WS(is, 100)]), ivs, &(xi[0])); } } TJ = LD(&(xi[WS(is, 124)]), ivs, &(xi[0])); TK = LD(&(xi[WS(is, 60)]), ivs, &(xi[0])); TU = LD(&(xi[WS(is, 92)]), ivs, &(xi[0])); TV = LD(&(xi[WS(is, 28)]), ivs, &(xi[0])); { V Tal, Tam, Tv, Tw, Taf; Tv = LD(&(xi[WS(is, 20)]), ivs, &(xi[0])); Tw = LD(&(xi[WS(is, 84)]), ivs, &(xi[0])); Taf = VADD(TD, TE); TF = VSUB(TD, TE); Ty = LD(&(xi[WS(is, 116)]), ivs, &(xi[0])); TL = VSUB(TJ, TK); Tal = VADD(TJ, TK); TW = VSUB(TU, TV); Tam = VADD(TV, TU); Tah = VADD(Tv, Tw); Tx = VSUB(Tv, Tw); Tag = VADD(Tae, Taf); Tee = VSUB(Tae, Taf); Tz = LD(&(xi[WS(is, 52)]), ivs, &(xi[0])); TM = LD(&(xi[WS(is, 12)]), ivs, &(xi[0])); TN = LD(&(xi[WS(is, 76)]), ivs, &(xi[0])); Teh = VSUB(Tal, Tam); Tan = VADD(Tal, Tam); TP = LD(&(xi[WS(is, 108)]), ivs, &(xi[0])); TQ = LD(&(xi[WS(is, 44)]), ivs, &(xi[0])); } } } } { V Tev, TeA, Tdp, TaP, Tew, TaV, T1U, T29, T7f, T1N, T28, T7i, Tex, TaS, T21; V T2a; { V Tem, Ter, Ten, TaD, T1j, T1y, TaA, Tdm, T1c, T78, T7b, T1x, TaG, Teo, T1z; V T1q; { V T14, T1v, Taw, Taz, T1b, T1w, T1n, T1o, T1m, TaE, TaF, T1p; { V Tau, Tav, T15, T16, T18, T19; { V T12, Tai, TA, Tao, TO, T13; T12 = LD(&(xi[WS(is, 2)]), ivs, &(xi[0])); Tai = VADD(Ty, Tz); TA = VSUB(Ty, Tz); Tao = VADD(TM, TN); TO = VSUB(TM, TN); T13 = LD(&(xi[WS(is, 66)]), ivs, &(xi[0])); { V T1t, Tap, TR, Taj, Tef, TG, TB, T1u; T1t = LD(&(xi[WS(is, 34)]), ivs, &(xi[0])); Tap = VADD(TP, TQ); TR = VSUB(TP, TQ); Taj = VADD(Tah, Tai); Tef = VSUB(Tah, Tai); TG = VSUB(Tx, TA); TB = VADD(Tx, TA); Tau = VADD(T12, T13); T14 = VSUB(T12, T13); T1u = LD(&(xi[WS(is, 98)]), ivs, &(xi[0])); { V Taq, Tei, TX, TS, Tak; Taq = VADD(Tao, Tap); Tei = VSUB(Tap, Tao); TX = VSUB(TR, TO); TS = VADD(TO, TR); Tak = VSUB(Tag, Taj); Td2 = VADD(Tag, Taj); { V Teg, Tfr, T71, TH; Teg = VFNMS(LDK(KP414213562), Tef, Tee); Tfr = VFMA(LDK(KP414213562), Tee, Tef); T71 = VFNMS(LDK(KP707106781), TG, TF); TH = VFMA(LDK(KP707106781), TG, TF); { V T70, TC, Tar, Tej, Tfs; T70 = VFNMS(LDK(KP707106781), TB, Tu); TC = VFMA(LDK(KP707106781), TB, Tu); Tar = VSUB(Tan, Taq); Td3 = VADD(Tan, Taq); Tej = VFNMS(LDK(KP414213562), Tei, Teh); Tfs = VFMA(LDK(KP414213562), Teh, Tei); { V T74, TY, T73, TT; T74 = VFNMS(LDK(KP707106781), TX, TW); TY = VFMA(LDK(KP707106781), TX, TW); T73 = VFNMS(LDK(KP707106781), TS, TL); TT = VFMA(LDK(KP707106781), TS, TL); T85 = VFNMS(LDK(KP668178637), T70, T71); T72 = VFMA(LDK(KP668178637), T71, T70); T4V = VFMA(LDK(KP198912367), TC, TH); TI = VFNMS(LDK(KP198912367), TH, TC); Tcd = VSUB(Tak, Tar); Tas = VADD(Tak, Tar); TgH = VSUB(Teg, Tej); Tek = VADD(Teg, Tej); Tgg = VADD(Tfr, Tfs); Tft = VSUB(Tfr, Tfs); T86 = VFNMS(LDK(KP668178637), T73, T74); T75 = VFMA(LDK(KP668178637), T74, T73); T4W = VFMA(LDK(KP198912367), TT, TY); TZ = VFNMS(LDK(KP198912367), TY, TT); Tav = VADD(T1t, T1u); T1v = VSUB(T1t, T1u); } } } } } } T15 = LD(&(xi[WS(is, 18)]), ivs, &(xi[0])); T16 = LD(&(xi[WS(is, 82)]), ivs, &(xi[0])); T18 = LD(&(xi[WS(is, 114)]), ivs, &(xi[0])); T19 = LD(&(xi[WS(is, 50)]), ivs, &(xi[0])); { V T1d, T1e, T1g, T1h, Tax, T17, Tay, T1a; T1d = LD(&(xi[WS(is, 10)]), ivs, &(xi[0])); Taw = VADD(Tau, Tav); Tem = VSUB(Tau, Tav); T1e = LD(&(xi[WS(is, 74)]), ivs, &(xi[0])); T1g = LD(&(xi[WS(is, 42)]), ivs, &(xi[0])); T1h = LD(&(xi[WS(is, 106)]), ivs, &(xi[0])); Tax = VADD(T15, T16); T17 = VSUB(T15, T16); Tay = VADD(T18, T19); T1a = VSUB(T18, T19); { V T1k, T1f, TaB, T1i, TaC, T1l; T1k = LD(&(xi[WS(is, 122)]), ivs, &(xi[0])); T1f = VSUB(T1d, T1e); TaB = VADD(T1d, T1e); T1i = VSUB(T1g, T1h); TaC = VADD(T1g, T1h); T1l = LD(&(xi[WS(is, 58)]), ivs, &(xi[0])); Taz = VADD(Tax, Tay); Ter = VSUB(Tax, Tay); T1b = VADD(T17, T1a); T1w = VSUB(T17, T1a); T1n = LD(&(xi[WS(is, 26)]), ivs, &(xi[0])); T1o = LD(&(xi[WS(is, 90)]), ivs, &(xi[0])); Ten = VSUB(TaB, TaC); TaD = VADD(TaB, TaC); T1j = VFNMS(LDK(KP414213562), T1i, T1f); T1y = VFMA(LDK(KP414213562), T1f, T1i); T1m = VSUB(T1k, T1l); TaE = VADD(T1k, T1l); } } } TaA = VSUB(Taw, Taz); Tdm = VADD(Taw, Taz); TaF = VADD(T1n, T1o); T1p = VSUB(T1n, T1o); T1c = VFMA(LDK(KP707106781), T1b, T14); T78 = VFNMS(LDK(KP707106781), T1b, T14); T7b = VFNMS(LDK(KP707106781), T1w, T1v); T1x = VFMA(LDK(KP707106781), T1w, T1v); TaG = VADD(TaE, TaF); Teo = VSUB(TaE, TaF); T1z = VFNMS(LDK(KP414213562), T1m, T1p); T1q = VFMA(LDK(KP414213562), T1p, T1m); } { V T1F, T26, T1Q, TaT, TaL, TaO, T27, T1M, T1Y, T1Z, TaU, T1T, TaQ, T1X, T20; V TaR; { V T24, TaJ, T25, T1G, T1H, T1J, T1K, T1D, T1E; T1D = LD(&(xi[WS(is, 126)]), ivs, &(xi[0])); T1E = LD(&(xi[WS(is, 62)]), ivs, &(xi[0])); T24 = LD(&(xi[WS(is, 94)]), ivs, &(xi[0])); { V TaH, Tdn, Tes, Tep; TaH = VSUB(TaD, TaG); Tdn = VADD(TaD, TaG); Tes = VSUB(Ten, Teo); Tep = VADD(Ten, Teo); { V T79, T1A, T7c, T1r; T79 = VSUB(T1y, T1z); T1A = VADD(T1y, T1z); T7c = VSUB(T1j, T1q); T1r = VADD(T1j, T1q); TaJ = VADD(T1D, T1E); T1F = VSUB(T1D, T1E); TaI = VFNMS(LDK(KP414213562), TaH, TaA); Tcf = VFMA(LDK(KP414213562), TaA, TaH); Tdo = VADD(Tdm, Tdn); TdG = VSUB(Tdm, Tdn); Tgi = VFNMS(LDK(KP707106781), Tes, Ter); Tet = VFMA(LDK(KP707106781), Tes, Ter); Tgj = VFNMS(LDK(KP707106781), Tep, Tem); Teq = VFMA(LDK(KP707106781), Tep, Tem); T8X = VFNMS(LDK(KP923879532), T79, T78); T7a = VFMA(LDK(KP923879532), T79, T78); T5M = VFNMS(LDK(KP923879532), T1A, T1x); T1B = VFMA(LDK(KP923879532), T1A, T1x); T8W = VFMA(LDK(KP923879532), T7c, T7b); T7d = VFNMS(LDK(KP923879532), T7c, T7b); T5N = VFNMS(LDK(KP923879532), T1r, T1c); T1s = VFMA(LDK(KP923879532), T1r, T1c); T25 = LD(&(xi[WS(is, 30)]), ivs, &(xi[0])); } } T1G = LD(&(xi[WS(is, 14)]), ivs, &(xi[0])); T1H = LD(&(xi[WS(is, 78)]), ivs, &(xi[0])); T1J = LD(&(xi[WS(is, 110)]), ivs, &(xi[0])); T1K = LD(&(xi[WS(is, 46)]), ivs, &(xi[0])); { V T1R, T1I, TaM, T1L, TaN, T1S, T1O, T1P, TaK, T1V, T1W; T1O = LD(&(xi[WS(is, 6)]), ivs, &(xi[0])); T1P = LD(&(xi[WS(is, 70)]), ivs, &(xi[0])); T26 = VSUB(T24, T25); TaK = VADD(T25, T24); T1R = LD(&(xi[WS(is, 102)]), ivs, &(xi[0])); T1I = VSUB(T1G, T1H); TaM = VADD(T1G, T1H); T1L = VSUB(T1J, T1K); TaN = VADD(T1J, T1K); T1Q = VSUB(T1O, T1P); TaT = VADD(T1O, T1P); Tev = VSUB(TaJ, TaK); TaL = VADD(TaJ, TaK); T1S = LD(&(xi[WS(is, 38)]), ivs, &(xi[0])); T1V = LD(&(xi[WS(is, 118)]), ivs, &(xi[0])); T1W = LD(&(xi[WS(is, 54)]), ivs, &(xi[0])); TeA = VSUB(TaN, TaM); TaO = VADD(TaM, TaN); T27 = VSUB(T1L, T1I); T1M = VADD(T1I, T1L); T1Y = LD(&(xi[WS(is, 86)]), ivs, &(xi[0])); T1Z = LD(&(xi[WS(is, 22)]), ivs, &(xi[0])); TaU = VADD(T1S, T1R); T1T = VSUB(T1R, T1S); TaQ = VADD(T1V, T1W); T1X = VSUB(T1V, T1W); } } Tdp = VADD(TaL, TaO); TaP = VSUB(TaL, TaO); T20 = VSUB(T1Y, T1Z); TaR = VADD(T1Z, T1Y); Tew = VSUB(TaT, TaU); TaV = VADD(TaT, TaU); T1U = VFMA(LDK(KP414213562), T1T, T1Q); T29 = VFNMS(LDK(KP414213562), T1Q, T1T); T7f = VFNMS(LDK(KP707106781), T1M, T1F); T1N = VFMA(LDK(KP707106781), T1M, T1F); T28 = VFMA(LDK(KP707106781), T27, T26); T7i = VFNMS(LDK(KP707106781), T27, T26); Tex = VSUB(TaQ, TaR); TaS = VADD(TaQ, TaR); T21 = VFNMS(LDK(KP414213562), T20, T1X); T2a = VFMA(LDK(KP414213562), T1X, T20); } } { V T2J, T2U, T30, T3b, TeL, Tb9, TeO, Tbg, T2M, Tba, T2P, Tbb, T34, Tbh, T33; V T35; { V T2H, T2I, T2S, T2T, T2Y, T2Z, T39, T3a; T2H = LD(&(xi[WS(is, 5)]), ivs, &(xi[WS(is, 1)])); { V Tdq, TaW, Tey, TeB; Tdq = VADD(TaV, TaS); TaW = VSUB(TaS, TaV); Tey = VADD(Tew, Tex); TeB = VSUB(Tex, Tew); { V T2b, T7g, T22, T7j; T2b = VADD(T29, T2a); T7g = VSUB(T2a, T29); T22 = VADD(T1U, T21); T7j = VSUB(T21, T1U); TaX = VFNMS(LDK(KP414213562), TaW, TaP); Tcg = VFMA(LDK(KP414213562), TaP, TaW); Tdr = VADD(Tdp, Tdq); TdH = VSUB(Tdp, Tdq); Tgl = VFNMS(LDK(KP707106781), TeB, TeA); TeC = VFMA(LDK(KP707106781), TeB, TeA); Tgm = VFNMS(LDK(KP707106781), Tey, Tev); Tez = VFMA(LDK(KP707106781), Tey, Tev); T90 = VFNMS(LDK(KP923879532), T7g, T7f); T7h = VFMA(LDK(KP923879532), T7g, T7f); T5P = VFNMS(LDK(KP923879532), T2b, T28); T2c = VFMA(LDK(KP923879532), T2b, T28); T8Z = VFMA(LDK(KP923879532), T7j, T7i); T7k = VFNMS(LDK(KP923879532), T7j, T7i); T5Q = VFNMS(LDK(KP923879532), T22, T1N); T23 = VFMA(LDK(KP923879532), T22, T1N); T2I = LD(&(xi[WS(is, 69)]), ivs, &(xi[WS(is, 1)])); } } T2S = LD(&(xi[WS(is, 37)]), ivs, &(xi[WS(is, 1)])); T2T = LD(&(xi[WS(is, 101)]), ivs, &(xi[WS(is, 1)])); T2Y = LD(&(xi[WS(is, 125)]), ivs, &(xi[WS(is, 1)])); T2Z = LD(&(xi[WS(is, 61)]), ivs, &(xi[WS(is, 1)])); T39 = LD(&(xi[WS(is, 29)]), ivs, &(xi[WS(is, 1)])); T3a = LD(&(xi[WS(is, 93)]), ivs, &(xi[WS(is, 1)])); { V T2K, Tbe, Tbf, T2L, T2N, T2O, Tb7, Tb8, T31, T32; T2K = LD(&(xi[WS(is, 21)]), ivs, &(xi[WS(is, 1)])); T2J = VSUB(T2H, T2I); Tb7 = VADD(T2H, T2I); T2U = VSUB(T2S, T2T); Tb8 = VADD(T2S, T2T); T30 = VSUB(T2Y, T2Z); Tbe = VADD(T2Y, T2Z); T3b = VSUB(T39, T3a); Tbf = VADD(T39, T3a); T2L = LD(&(xi[WS(is, 85)]), ivs, &(xi[WS(is, 1)])); T2N = LD(&(xi[WS(is, 117)]), ivs, &(xi[WS(is, 1)])); T2O = LD(&(xi[WS(is, 53)]), ivs, &(xi[WS(is, 1)])); TeL = VSUB(Tb7, Tb8); Tb9 = VADD(Tb7, Tb8); T31 = LD(&(xi[WS(is, 13)]), ivs, &(xi[WS(is, 1)])); T32 = LD(&(xi[WS(is, 77)]), ivs, &(xi[WS(is, 1)])); TeO = VSUB(Tbe, Tbf); Tbg = VADD(Tbe, Tbf); T2M = VSUB(T2K, T2L); Tba = VADD(T2K, T2L); T2P = VSUB(T2N, T2O); Tbb = VADD(T2N, T2O); T34 = LD(&(xi[WS(is, 109)]), ivs, &(xi[WS(is, 1)])); Tbh = VADD(T31, T32); T33 = VSUB(T31, T32); T35 = LD(&(xi[WS(is, 45)]), ivs, &(xi[WS(is, 1)])); } } { V T4d, T4e, T4o, T4p; { V T2X, T3q, T7t, T7C, T3r, T3e, T7D, T7w; { V T47, TbE, Tbd, Td9, TeW, TeN, T7s, T2W, T7r, T2R, TeP, Tbj, T37, T3c, T48; { V T3W, T3X, TeM, Tbc, T2Q, T2V, Tbi, T36; T3W = LD(&(xi[WS(is, 3)]), ivs, &(xi[WS(is, 1)])); T3X = LD(&(xi[WS(is, 67)]), ivs, &(xi[WS(is, 1)])); TeM = VSUB(Tba, Tbb); Tbc = VADD(Tba, Tbb); T2Q = VADD(T2M, T2P); T2V = VSUB(T2M, T2P); T47 = LD(&(xi[WS(is, 99)]), ivs, &(xi[WS(is, 1)])); Tbi = VADD(T34, T35); T36 = VSUB(T34, T35); TbE = VADD(T3W, T3X); T3Y = VSUB(T3W, T3X); Tbd = VSUB(Tb9, Tbc); Td9 = VADD(Tb9, Tbc); TeW = VFMA(LDK(KP414213562), TeL, TeM); TeN = VFNMS(LDK(KP414213562), TeM, TeL); T7s = VFNMS(LDK(KP707106781), T2V, T2U); T2W = VFMA(LDK(KP707106781), T2V, T2U); T7r = VFNMS(LDK(KP707106781), T2Q, T2J); T2R = VFMA(LDK(KP707106781), T2Q, T2J); TeP = VSUB(Tbh, Tbi); Tbj = VADD(Tbh, Tbi); T37 = VADD(T33, T36); T3c = VSUB(T33, T36); T48 = LD(&(xi[WS(is, 35)]), ivs, &(xi[WS(is, 1)])); } T2X = VFNMS(LDK(KP198912367), T2W, T2R); T3q = VFMA(LDK(KP198912367), T2R, T2W); T7t = VFMA(LDK(KP668178637), T7s, T7r); T7C = VFNMS(LDK(KP668178637), T7r, T7s); { V Tbk, Tda, TeX, TeQ; Tbk = VSUB(Tbg, Tbj); Tda = VADD(Tbg, Tbj); TeX = VFNMS(LDK(KP414213562), TeO, TeP); TeQ = VFMA(LDK(KP414213562), TeP, TeO); { V T7v, T3d, T7u, T38, TbF; T7v = VFNMS(LDK(KP707106781), T3c, T3b); T3d = VFMA(LDK(KP707106781), T3c, T3b); T7u = VFNMS(LDK(KP707106781), T37, T30); T38 = VFMA(LDK(KP707106781), T37, T30); T49 = VSUB(T47, T48); TbF = VADD(T48, T47); TdL = VSUB(Td9, Tda); Tdb = VADD(Td9, Tda); Tbu = VSUB(Tbd, Tbk); Tbl = VADD(Tbd, Tbk); Tgu = VSUB(TeN, TeQ); TeR = VADD(TeN, TeQ); Tgr = VSUB(TeW, TeX); TeY = VADD(TeW, TeX); T3r = VFNMS(LDK(KP198912367), T38, T3d); T3e = VFMA(LDK(KP198912367), T3d, T38); T7D = VFMA(LDK(KP668178637), T7u, T7v); T7w = VFNMS(LDK(KP668178637), T7v, T7u); Tf6 = VSUB(TbE, TbF); TbG = VADD(TbE, TbF); } } } T4d = LD(&(xi[WS(is, 123)]), ivs, &(xi[WS(is, 1)])); T5V = VSUB(T3q, T3r); T3s = VADD(T3q, T3r); T5Y = VSUB(T2X, T3e); T3f = VADD(T2X, T3e); T95 = VSUB(T7D, T7C); T7E = VADD(T7C, T7D); T98 = VSUB(T7t, T7w); T7x = VADD(T7t, T7w); T4e = LD(&(xi[WS(is, 59)]), ivs, &(xi[WS(is, 1)])); T4o = LD(&(xi[WS(is, 91)]), ivs, &(xi[WS(is, 1)])); T4p = LD(&(xi[WS(is, 27)]), ivs, &(xi[WS(is, 1)])); } { V T3Z, T40, T42, T43, TbL, TbM; T3Z = LD(&(xi[WS(is, 19)]), ivs, &(xi[WS(is, 1)])); T40 = LD(&(xi[WS(is, 83)]), ivs, &(xi[WS(is, 1)])); T42 = LD(&(xi[WS(is, 115)]), ivs, &(xi[WS(is, 1)])); T43 = LD(&(xi[WS(is, 51)]), ivs, &(xi[WS(is, 1)])); T4g = LD(&(xi[WS(is, 11)]), ivs, &(xi[WS(is, 1)])); T4f = VSUB(T4d, T4e); TbL = VADD(T4d, T4e); T4q = VSUB(T4o, T4p); TbM = VADD(T4p, T4o); TbH = VADD(T3Z, T40); T41 = VSUB(T3Z, T40); TbI = VADD(T42, T43); T44 = VSUB(T42, T43); T4h = LD(&(xi[WS(is, 75)]), ivs, &(xi[WS(is, 1)])); T4j = LD(&(xi[WS(is, 107)]), ivs, &(xi[WS(is, 1)])); T4k = LD(&(xi[WS(is, 43)]), ivs, &(xi[WS(is, 1)])); Tf9 = VSUB(TbL, TbM); TbN = VADD(TbL, TbM); } } } } } { V TgB, Tgy, T62, T4H, T65, T4u, T9c, T7X, T9f, T7Q, Tg0, Tga, TfF, TeF, TfT; V TfU, TfP, Tg7, TfI, Tfy, Tfz, Tf0, TfA, Tfl, Tg1, TfS; { V Tc1, TbS, Tfc, Tfj, TdX, Te5, TdZ, TdR, Te7, Te3, TdU, Te4; { V TdF, TdS, Tdx, Td5, TdO, TdE, TdC, Tdt, Tdk; { V Tdc, TdA, T4F, T4c, T7V, T7M, T4G, T4t, T7W, T7P, TdB, Tdj; { V Td1, Tdg, TbK, Tf8, Tfh, T4b, T7L, T46, T7K, TbQ, Tfa, T4r, T4m, Td4; TdF = VSUB(TcZ, Td0); Td1 = VADD(TcZ, Td0); { V TbJ, Tf7, T4a, T45; TbJ = VADD(TbH, TbI); Tf7 = VSUB(TbI, TbH); T4a = VSUB(T44, T41); T45 = VADD(T41, T44); { V TbO, T4i, TbP, T4l; TbO = VADD(T4g, T4h); T4i = VSUB(T4g, T4h); TbP = VADD(T4j, T4k); T4l = VSUB(T4j, T4k); Tdg = VADD(TbG, TbJ); TbK = VSUB(TbG, TbJ); Tf8 = VFMA(LDK(KP414213562), Tf7, Tf6); Tfh = VFNMS(LDK(KP414213562), Tf6, Tf7); T4b = VFMA(LDK(KP707106781), T4a, T49); T7L = VFNMS(LDK(KP707106781), T4a, T49); T46 = VFMA(LDK(KP707106781), T45, T3Y); T7K = VFNMS(LDK(KP707106781), T45, T3Y); TbQ = VADD(TbO, TbP); Tfa = VSUB(TbP, TbO); T4r = VSUB(T4l, T4i); T4m = VADD(T4i, T4l); Td4 = VADD(Td2, Td3); TdS = VSUB(Td2, Td3); } } Tdc = VSUB(Td8, Tdb); TdA = VADD(Td8, Tdb); T4F = VFNMS(LDK(KP198912367), T46, T4b); T4c = VFMA(LDK(KP198912367), T4b, T46); T7V = VFMA(LDK(KP668178637), T7K, T7L); T7M = VFNMS(LDK(KP668178637), T7L, T7K); { V Tdh, TbR, Tfb, Tfi; Tdh = VADD(TbN, TbQ); TbR = VSUB(TbN, TbQ); Tfb = VFNMS(LDK(KP414213562), Tfa, Tf9); Tfi = VFMA(LDK(KP414213562), Tf9, Tfa); { V T4s, T7O, T4n, T7N, Tdi; T4s = VFMA(LDK(KP707106781), T4r, T4q); T7O = VFNMS(LDK(KP707106781), T4r, T4q); T4n = VFMA(LDK(KP707106781), T4m, T4f); T7N = VFNMS(LDK(KP707106781), T4m, T4f); Tdx = VADD(Td1, Td4); Td5 = VSUB(Td1, Td4); TdO = VSUB(Tdh, Tdg); Tdi = VADD(Tdg, Tdh); Tc1 = VSUB(TbR, TbK); TbS = VADD(TbK, TbR); TgB = VSUB(Tfb, Tf8); Tfc = VADD(Tf8, Tfb); Tgy = VSUB(Tfi, Tfh); Tfj = VADD(Tfh, Tfi); T4G = VFMA(LDK(KP198912367), T4n, T4s); T4t = VFNMS(LDK(KP198912367), T4s, T4n); T7W = VFNMS(LDK(KP668178637), T7N, T7O); T7P = VFMA(LDK(KP668178637), T7O, T7N); TdB = VADD(Tdf, Tdi); Tdj = VSUB(Tdf, Tdi); } } } T62 = VSUB(T4G, T4F); T4H = VADD(T4F, T4G); T65 = VSUB(T4t, T4c); T4u = VADD(T4c, T4t); T9c = VSUB(T7V, T7W); T7X = VADD(T7V, T7W); T9f = VSUB(T7P, T7M); T7Q = VADD(T7M, T7P); TdE = VADD(TdA, TdB); TdC = VSUB(TdA, TdB); Tdt = VSUB(Tdc, Tdj); Tdk = VADD(Tdc, Tdj); } { V TdT, Tdl, Tdv, TdJ, Te1, Te2, TdQ, Tdz, TdD, Tdu, Tdw; { V TdI, TdM, TdV, TdW, TdP, Tds, Tdy; TdI = VADD(TdG, TdH); TdT = VSUB(TdG, TdH); TdM = VFNMS(LDK(KP414213562), TdL, TdK); TdV = VFMA(LDK(KP414213562), TdK, TdL); TdW = VFMA(LDK(KP414213562), TdN, TdO); TdP = VFNMS(LDK(KP414213562), TdO, TdN); Tdl = VFNMS(LDK(KP707106781), Tdk, Td5); Tdv = VFMA(LDK(KP707106781), Tdk, Td5); Tds = VSUB(Tdo, Tdr); Tdy = VADD(Tdo, Tdr); TdJ = VFMA(LDK(KP707106781), TdI, TdF); Te1 = VFNMS(LDK(KP707106781), TdI, TdF); TdX = VSUB(TdV, TdW); Te2 = VADD(TdV, TdW); Te5 = VSUB(TdM, TdP); TdQ = VADD(TdM, TdP); Tdz = VSUB(Tdx, Tdy); TdD = VADD(Tdx, Tdy); Tdu = VFNMS(LDK(KP707106781), Tdt, Tds); Tdw = VFMA(LDK(KP707106781), Tdt, Tds); } TdZ = VFMA(LDK(KP923879532), TdQ, TdJ); TdR = VFNMS(LDK(KP923879532), TdQ, TdJ); Te7 = VFMA(LDK(KP923879532), Te2, Te1); Te3 = VFNMS(LDK(KP923879532), Te2, Te1); ST(&(xo[0]), VADD(TdD, TdE), ovs, &(xo[0])); ST(&(xo[WS(os, 64)]), VSUB(TdD, TdE), ovs, &(xo[0])); ST(&(xo[WS(os, 32)]), VFMAI(TdC, Tdz), ovs, &(xo[0])); ST(&(xo[WS(os, 96)]), VFNMSI(TdC, Tdz), ovs, &(xo[0])); ST(&(xo[WS(os, 112)]), VFNMSI(Tdw, Tdv), ovs, &(xo[0])); ST(&(xo[WS(os, 16)]), VFMAI(Tdw, Tdv), ovs, &(xo[0])); ST(&(xo[WS(os, 80)]), VFMAI(Tdu, Tdl), ovs, &(xo[0])); ST(&(xo[WS(os, 48)]), VFNMSI(Tdu, Tdl), ovs, &(xo[0])); TdU = VFMA(LDK(KP707106781), TdT, TdS); Te4 = VFNMS(LDK(KP707106781), TdT, TdS); } } { V Tcx, TcJ, TcI, Tcy, TcA, Tbm, Tcp, TaZ, Tcs, Tci, Tbv, TcB, TcD, TbT, Tc2; V TcE, Tat, TaY; Tcx = VFNMS(LDK(KP707106781), Tas, Tad); Tat = VFMA(LDK(KP707106781), Tas, Tad); TaY = VADD(TaI, TaX); TcJ = VSUB(TaI, TaX); { V Tce, Tch, Te8, Te6, TdY, Te0; TcI = VFNMS(LDK(KP707106781), Tcd, Tcc); Tce = VFMA(LDK(KP707106781), Tcd, Tcc); Tch = VSUB(Tcf, Tcg); Tcy = VADD(Tcf, Tcg); Te8 = VFNMS(LDK(KP923879532), Te5, Te4); Te6 = VFMA(LDK(KP923879532), Te5, Te4); TdY = VFNMS(LDK(KP923879532), TdX, TdU); Te0 = VFMA(LDK(KP923879532), TdX, TdU); TcA = VFNMS(LDK(KP707106781), Tbl, Tb6); Tbm = VFMA(LDK(KP707106781), Tbl, Tb6); Tcp = VFNMS(LDK(KP923879532), TaY, Tat); TaZ = VFMA(LDK(KP923879532), TaY, Tat); Tcs = VFNMS(LDK(KP923879532), Tch, Tce); Tci = VFMA(LDK(KP923879532), Tch, Tce); ST(&(xo[WS(os, 88)]), VFNMSI(Te6, Te3), ovs, &(xo[0])); ST(&(xo[WS(os, 40)]), VFMAI(Te6, Te3), ovs, &(xo[0])); ST(&(xo[WS(os, 104)]), VFMAI(Te8, Te7), ovs, &(xo[0])); ST(&(xo[WS(os, 24)]), VFNMSI(Te8, Te7), ovs, &(xo[0])); ST(&(xo[WS(os, 8)]), VFMAI(Te0, TdZ), ovs, &(xo[0])); ST(&(xo[WS(os, 120)]), VFNMSI(Te0, TdZ), ovs, &(xo[0])); ST(&(xo[WS(os, 72)]), VFMAI(TdY, TdR), ovs, &(xo[0])); ST(&(xo[WS(os, 56)]), VFNMSI(TdY, TdR), ovs, &(xo[0])); Tbv = VFMA(LDK(KP707106781), Tbu, Tbt); TcB = VFNMS(LDK(KP707106781), Tbu, Tbt); TcD = VFNMS(LDK(KP707106781), TbS, TbD); TbT = VFMA(LDK(KP707106781), TbS, TbD); Tc2 = VFMA(LDK(KP707106781), Tc1, Tc0); TcE = VFNMS(LDK(KP707106781), Tc1, Tc0); } { V TcR, Tcz, TcU, TcK, Tcq, Tcl, Tct, Tc4; { V Tcj, Tbw, Tck, Tc3; Tcj = VFMA(LDK(KP198912367), Tbm, Tbv); Tbw = VFNMS(LDK(KP198912367), Tbv, Tbm); Tck = VFMA(LDK(KP198912367), TbT, Tc2); Tc3 = VFNMS(LDK(KP198912367), Tc2, TbT); TcR = VFNMS(LDK(KP923879532), Tcy, Tcx); Tcz = VFMA(LDK(KP923879532), Tcy, Tcx); TcU = VFMA(LDK(KP923879532), TcJ, TcI); TcK = VFNMS(LDK(KP923879532), TcJ, TcI); Tcq = VADD(Tcj, Tck); Tcl = VSUB(Tcj, Tck); Tct = VSUB(Tbw, Tc3); Tc4 = VADD(Tbw, Tc3); } { V TfN, Tel, TfY, Tfu, Tfv, Tfw, TcT, TcX, TcQ, TcO, TcW, TcY, TcP, TcH, TfZ; V TeE; { V Teu, TcS, TcN, TcV, TcG, TeD; TfN = VFNMS(LDK(KP923879532), Tek, Ted); Tel = VFMA(LDK(KP923879532), Tek, Ted); { V TcL, TcC, Tcr, Tcv; TcL = VFNMS(LDK(KP668178637), TcA, TcB); TcC = VFMA(LDK(KP668178637), TcB, TcA); Tcr = VFNMS(LDK(KP980785280), Tcq, Tcp); Tcv = VFMA(LDK(KP980785280), Tcq, Tcp); { V Tco, Tcm, Tcu, Tcw; Tco = VFMA(LDK(KP980785280), Tcl, Tci); Tcm = VFNMS(LDK(KP980785280), Tcl, Tci); Tcu = VFMA(LDK(KP980785280), Tct, Tcs); Tcw = VFNMS(LDK(KP980785280), Tct, Tcs); { V Tcn, Tc5, TcM, TcF; Tcn = VFMA(LDK(KP980785280), Tc4, TaZ); Tc5 = VFNMS(LDK(KP980785280), Tc4, TaZ); TcM = VFNMS(LDK(KP668178637), TcD, TcE); TcF = VFMA(LDK(KP668178637), TcE, TcD); TfY = VFNMS(LDK(KP923879532), Tft, Tfq); Tfu = VFMA(LDK(KP923879532), Tft, Tfq); Tfv = VFMA(LDK(KP198912367), Teq, Tet); Teu = VFNMS(LDK(KP198912367), Tet, Teq); ST(&(xo[WS(os, 92)]), VFNMSI(Tcu, Tcr), ovs, &(xo[0])); ST(&(xo[WS(os, 36)]), VFMAI(Tcu, Tcr), ovs, &(xo[0])); ST(&(xo[WS(os, 100)]), VFMAI(Tcw, Tcv), ovs, &(xo[0])); ST(&(xo[WS(os, 28)]), VFNMSI(Tcw, Tcv), ovs, &(xo[0])); ST(&(xo[WS(os, 4)]), VFMAI(Tco, Tcn), ovs, &(xo[0])); ST(&(xo[WS(os, 124)]), VFNMSI(Tco, Tcn), ovs, &(xo[0])); ST(&(xo[WS(os, 68)]), VFMAI(Tcm, Tc5), ovs, &(xo[0])); ST(&(xo[WS(os, 60)]), VFNMSI(Tcm, Tc5), ovs, &(xo[0])); TcS = VADD(TcL, TcM); TcN = VSUB(TcL, TcM); TcV = VSUB(TcC, TcF); TcG = VADD(TcC, TcF); TeD = VFNMS(LDK(KP198912367), TeC, Tez); Tfw = VFMA(LDK(KP198912367), Tez, TeC); } } } TcT = VFMA(LDK(KP831469612), TcS, TcR); TcX = VFNMS(LDK(KP831469612), TcS, TcR); TcQ = VFMA(LDK(KP831469612), TcN, TcK); TcO = VFNMS(LDK(KP831469612), TcN, TcK); TcW = VFNMS(LDK(KP831469612), TcV, TcU); TcY = VFMA(LDK(KP831469612), TcV, TcU); TcP = VFMA(LDK(KP831469612), TcG, Tcz); TcH = VFNMS(LDK(KP831469612), TcG, Tcz); TfZ = VSUB(Teu, TeD); TeE = VADD(Teu, TeD); } { V TfQ, TeS, TfO, Tfx, TeZ, TfR, Tfd, Tfk; TfQ = VFNMS(LDK(KP923879532), TeR, TeK); TeS = VFMA(LDK(KP923879532), TeR, TeK); ST(&(xo[WS(os, 84)]), VFMAI(TcW, TcT), ovs, &(xo[0])); ST(&(xo[WS(os, 44)]), VFNMSI(TcW, TcT), ovs, &(xo[0])); ST(&(xo[WS(os, 108)]), VFNMSI(TcY, TcX), ovs, &(xo[0])); ST(&(xo[WS(os, 20)]), VFMAI(TcY, TcX), ovs, &(xo[0])); ST(&(xo[WS(os, 116)]), VFMAI(TcQ, TcP), ovs, &(xo[0])); ST(&(xo[WS(os, 12)]), VFNMSI(TcQ, TcP), ovs, &(xo[0])); ST(&(xo[WS(os, 52)]), VFMAI(TcO, TcH), ovs, &(xo[0])); ST(&(xo[WS(os, 76)]), VFNMSI(TcO, TcH), ovs, &(xo[0])); Tg0 = VFNMS(LDK(KP980785280), TfZ, TfY); Tga = VFMA(LDK(KP980785280), TfZ, TfY); TfF = VFNMS(LDK(KP980785280), TeE, Tel); TeF = VFMA(LDK(KP980785280), TeE, Tel); TfO = VADD(Tfv, Tfw); Tfx = VSUB(Tfv, Tfw); TeZ = VFMA(LDK(KP923879532), TeY, TeV); TfR = VFNMS(LDK(KP923879532), TeY, TeV); TfT = VFNMS(LDK(KP923879532), Tfc, Tf5); Tfd = VFMA(LDK(KP923879532), Tfc, Tf5); Tfk = VFMA(LDK(KP923879532), Tfj, Tfg); TfU = VFNMS(LDK(KP923879532), Tfj, Tfg); TfP = VFMA(LDK(KP980785280), TfO, TfN); Tg7 = VFNMS(LDK(KP980785280), TfO, TfN); TfI = VFNMS(LDK(KP980785280), Tfx, Tfu); Tfy = VFMA(LDK(KP980785280), Tfx, Tfu); Tfz = VFMA(LDK(KP098491403), TeS, TeZ); Tf0 = VFNMS(LDK(KP098491403), TeZ, TeS); TfA = VFMA(LDK(KP098491403), Tfd, Tfk); Tfl = VFNMS(LDK(KP098491403), Tfk, Tfd); Tg1 = VFNMS(LDK(KP820678790), TfQ, TfR); TfS = VFMA(LDK(KP820678790), TfR, TfQ); } } } } } { V T8x, T8y, T8F, T8w, T8k, T8f, T8n, T80, T9l, T76, T87, T8U, T89, T7e, T7l; V T8a; { V The, Tho, TgT, Tgp, Th7, Th8, Thf, Th6, Th3, Thl, TgW, TgM, TgU, TgP, TgX; V TgE; { V Th1, TgI, TgJ, TgK; { V Tgh, Thc, Tgk, TfG, TfB, TfJ, Tfm, Tg2, TfV, Tgn, TfL, TfH; Th1 = VFMA(LDK(KP923879532), Tgg, Tgf); Tgh = VFNMS(LDK(KP923879532), Tgg, Tgf); Thc = VFNMS(LDK(KP923879532), TgH, TgG); TgI = VFMA(LDK(KP923879532), TgH, TgG); TgJ = VFMA(LDK(KP668178637), Tgi, Tgj); Tgk = VFNMS(LDK(KP668178637), Tgj, Tgi); TfG = VADD(Tfz, TfA); TfB = VSUB(Tfz, TfA); TfJ = VSUB(Tf0, Tfl); Tfm = VADD(Tf0, Tfl); Tg2 = VFNMS(LDK(KP820678790), TfT, TfU); TfV = VFMA(LDK(KP820678790), TfU, TfT); Tgn = VFNMS(LDK(KP668178637), Tgm, Tgl); TgK = VFMA(LDK(KP668178637), Tgl, Tgm); TfL = VFMA(LDK(KP995184726), TfG, TfF); TfH = VFNMS(LDK(KP995184726), TfG, TfF); { V TfE, TfC, TfM, TfK; TfE = VFMA(LDK(KP995184726), TfB, Tfy); TfC = VFNMS(LDK(KP995184726), TfB, Tfy); TfM = VFNMS(LDK(KP995184726), TfJ, TfI); TfK = VFMA(LDK(KP995184726), TfJ, TfI); { V TfD, Tfn, Tg8, Tg3; TfD = VFMA(LDK(KP995184726), Tfm, TeF); Tfn = VFNMS(LDK(KP995184726), Tfm, TeF); Tg8 = VADD(Tg1, Tg2); Tg3 = VSUB(Tg1, Tg2); { V Tgb, TfW, Thd, Tgo; Tgb = VSUB(TfS, TfV); TfW = VADD(TfS, TfV); Thd = VSUB(Tgk, Tgn); Tgo = VADD(Tgk, Tgn); ST(&(xo[WS(os, 98)]), VFMAI(TfM, TfL), ovs, &(xo[0])); ST(&(xo[WS(os, 30)]), VFNMSI(TfM, TfL), ovs, &(xo[0])); ST(&(xo[WS(os, 94)]), VFNMSI(TfK, TfH), ovs, &(xo[0])); ST(&(xo[WS(os, 34)]), VFMAI(TfK, TfH), ovs, &(xo[0])); ST(&(xo[WS(os, 2)]), VFMAI(TfE, TfD), ovs, &(xo[0])); ST(&(xo[WS(os, 126)]), VFNMSI(TfE, TfD), ovs, &(xo[0])); ST(&(xo[WS(os, 66)]), VFMAI(TfC, Tfn), ovs, &(xo[0])); ST(&(xo[WS(os, 62)]), VFNMSI(TfC, Tfn), ovs, &(xo[0])); { V Tgd, Tg9, Tg6, Tg4; Tgd = VFNMS(LDK(KP773010453), Tg8, Tg7); Tg9 = VFMA(LDK(KP773010453), Tg8, Tg7); Tg6 = VFMA(LDK(KP773010453), Tg3, Tg0); Tg4 = VFNMS(LDK(KP773010453), Tg3, Tg0); { V Tge, Tgc, Tg5, TfX; Tge = VFMA(LDK(KP773010453), Tgb, Tga); Tgc = VFNMS(LDK(KP773010453), Tgb, Tga); Tg5 = VFMA(LDK(KP773010453), TfW, TfP); TfX = VFNMS(LDK(KP773010453), TfW, TfP); The = VFMA(LDK(KP831469612), Thd, Thc); Tho = VFNMS(LDK(KP831469612), Thd, Thc); TgT = VFMA(LDK(KP831469612), Tgo, Tgh); Tgp = VFNMS(LDK(KP831469612), Tgo, Tgh); ST(&(xo[WS(os, 110)]), VFNMSI(Tge, Tgd), ovs, &(xo[0])); ST(&(xo[WS(os, 18)]), VFMAI(Tge, Tgd), ovs, &(xo[0])); ST(&(xo[WS(os, 82)]), VFMAI(Tgc, Tg9), ovs, &(xo[0])); ST(&(xo[WS(os, 46)]), VFNMSI(Tgc, Tg9), ovs, &(xo[0])); ST(&(xo[WS(os, 114)]), VFMAI(Tg6, Tg5), ovs, &(xo[0])); ST(&(xo[WS(os, 14)]), VFNMSI(Tg6, Tg5), ovs, &(xo[0])); ST(&(xo[WS(os, 50)]), VFMAI(Tg4, TfX), ovs, &(xo[0])); ST(&(xo[WS(os, 78)]), VFNMSI(Tg4, TfX), ovs, &(xo[0])); } } } } } } { V Th4, Tgs, Tgv, Th5, Tgz, TgC, Th2, TgL; Th4 = VFMA(LDK(KP923879532), Tgr, Tgq); Tgs = VFNMS(LDK(KP923879532), Tgr, Tgq); Tgv = VFMA(LDK(KP923879532), Tgu, Tgt); Th5 = VFNMS(LDK(KP923879532), Tgu, Tgt); Th7 = VFMA(LDK(KP923879532), Tgy, Tgx); Tgz = VFNMS(LDK(KP923879532), Tgy, Tgx); TgC = VFMA(LDK(KP923879532), TgB, TgA); Th8 = VFNMS(LDK(KP923879532), TgB, TgA); Th2 = VADD(TgJ, TgK); TgL = VSUB(TgJ, TgK); { V TgN, Tgw, TgO, TgD; TgN = VFMA(LDK(KP534511135), Tgs, Tgv); Tgw = VFNMS(LDK(KP534511135), Tgv, Tgs); TgO = VFMA(LDK(KP534511135), Tgz, TgC); TgD = VFNMS(LDK(KP534511135), TgC, Tgz); Thf = VFNMS(LDK(KP303346683), Th4, Th5); Th6 = VFMA(LDK(KP303346683), Th5, Th4); Th3 = VFMA(LDK(KP831469612), Th2, Th1); Thl = VFNMS(LDK(KP831469612), Th2, Th1); TgW = VFNMS(LDK(KP831469612), TgL, TgI); TgM = VFMA(LDK(KP831469612), TgL, TgI); TgU = VADD(TgN, TgO); TgP = VSUB(TgN, TgO); TgX = VSUB(Tgw, TgD); TgE = VADD(Tgw, TgD); } } } { V T8u, T8v, T7R, T8d, T7G, Thm, Thh, Thp, Tha, T7Y, Thr, Thn; { V T7y, T7F, TgZ, TgV; T8u = VFNMS(LDK(KP831469612), T7x, T7q); T7y = VFMA(LDK(KP831469612), T7x, T7q); T7F = VFMA(LDK(KP831469612), T7E, T7B); T8v = VFNMS(LDK(KP831469612), T7E, T7B); T8x = VFNMS(LDK(KP831469612), T7Q, T7J); T7R = VFMA(LDK(KP831469612), T7Q, T7J); TgZ = VFMA(LDK(KP881921264), TgU, TgT); TgV = VFNMS(LDK(KP881921264), TgU, TgT); { V TgS, TgQ, Th0, TgY; TgS = VFMA(LDK(KP881921264), TgP, TgM); TgQ = VFNMS(LDK(KP881921264), TgP, TgM); Th0 = VFNMS(LDK(KP881921264), TgX, TgW); TgY = VFMA(LDK(KP881921264), TgX, TgW); { V TgR, TgF, Thg, Th9; TgR = VFMA(LDK(KP881921264), TgE, Tgp); TgF = VFNMS(LDK(KP881921264), TgE, Tgp); Thg = VFNMS(LDK(KP303346683), Th7, Th8); Th9 = VFMA(LDK(KP303346683), Th8, Th7); T8d = VFNMS(LDK(KP148335987), T7y, T7F); T7G = VFMA(LDK(KP148335987), T7F, T7y); ST(&(xo[WS(os, 106)]), VFMAI(Th0, TgZ), ovs, &(xo[0])); ST(&(xo[WS(os, 22)]), VFNMSI(Th0, TgZ), ovs, &(xo[0])); ST(&(xo[WS(os, 86)]), VFNMSI(TgY, TgV), ovs, &(xo[0])); ST(&(xo[WS(os, 42)]), VFMAI(TgY, TgV), ovs, &(xo[0])); ST(&(xo[WS(os, 10)]), VFMAI(TgS, TgR), ovs, &(xo[0])); ST(&(xo[WS(os, 118)]), VFNMSI(TgS, TgR), ovs, &(xo[0])); ST(&(xo[WS(os, 74)]), VFMAI(TgQ, TgF), ovs, &(xo[0])); ST(&(xo[WS(os, 54)]), VFNMSI(TgQ, TgF), ovs, &(xo[0])); Thm = VADD(Thf, Thg); Thh = VSUB(Thf, Thg); Thp = VSUB(Th6, Th9); Tha = VADD(Th6, Th9); T7Y = VFMA(LDK(KP831469612), T7X, T7U); T8y = VFNMS(LDK(KP831469612), T7X, T7U); } } } Thr = VFNMS(LDK(KP956940335), Thm, Thl); Thn = VFMA(LDK(KP956940335), Thm, Thl); { V Thk, Thi, Ths, Thq; Thk = VFMA(LDK(KP956940335), Thh, The); Thi = VFNMS(LDK(KP956940335), Thh, The); Ths = VFMA(LDK(KP956940335), Thp, Tho); Thq = VFNMS(LDK(KP956940335), Thp, Tho); { V Thj, Thb, T8e, T7Z; Thj = VFMA(LDK(KP956940335), Tha, Th3); Thb = VFNMS(LDK(KP956940335), Tha, Th3); T8e = VFNMS(LDK(KP148335987), T7R, T7Y); T7Z = VFMA(LDK(KP148335987), T7Y, T7R); T8F = VFMA(LDK(KP741650546), T8u, T8v); T8w = VFNMS(LDK(KP741650546), T8v, T8u); ST(&(xo[WS(os, 102)]), VFNMSI(Ths, Thr), ovs, &(xo[0])); ST(&(xo[WS(os, 26)]), VFMAI(Ths, Thr), ovs, &(xo[0])); ST(&(xo[WS(os, 90)]), VFMAI(Thq, Thn), ovs, &(xo[0])); ST(&(xo[WS(os, 38)]), VFNMSI(Thq, Thn), ovs, &(xo[0])); ST(&(xo[WS(os, 122)]), VFMAI(Thk, Thj), ovs, &(xo[0])); ST(&(xo[WS(os, 6)]), VFNMSI(Thk, Thj), ovs, &(xo[0])); ST(&(xo[WS(os, 58)]), VFMAI(Thi, Thb), ovs, &(xo[0])); ST(&(xo[WS(os, 70)]), VFNMSI(Thi, Thb), ovs, &(xo[0])); T8k = VADD(T8d, T8e); T8f = VSUB(T8d, T8e); T8n = VSUB(T7G, T7Z); T80 = VADD(T7G, T7Z); } } T9l = VSUB(T72, T75); T76 = VADD(T72, T75); T87 = VSUB(T85, T86); T8U = VADD(T85, T86); T89 = VFNMS(LDK(KP303346683), T7a, T7d); T7e = VFMA(LDK(KP303346683), T7d, T7a); T7l = VFMA(LDK(KP303346683), T7k, T7h); T8a = VFNMS(LDK(KP303346683), T7h, T7k); } } { V T11, T5h, T5a, T55, T5d, T4K, T5C, T5x, T5F, T5q, T4X, T4Z, T1C, T2d, T50; { V T5k, T3g, T3t, T5l, T5n, T4v, T4I, T5o, T8G, T8z; T5k = VFNMS(LDK(KP980785280), T3f, T2G); T3g = VFMA(LDK(KP980785280), T3f, T2G); T8G = VFMA(LDK(KP741650546), T8x, T8y); T8z = VFNMS(LDK(KP741650546), T8y, T8x); { V T8r, T77, T8C, T88; T8r = VFNMS(LDK(KP831469612), T76, T6Z); T77 = VFMA(LDK(KP831469612), T76, T6Z); T8C = VFNMS(LDK(KP831469612), T87, T84); T88 = VFMA(LDK(KP831469612), T87, T84); { V T8D, T7m, T8s, T8b; T8D = VSUB(T7e, T7l); T7m = VADD(T7e, T7l); T8s = VADD(T89, T8a); T8b = VSUB(T89, T8a); { V T8M, T8H, T8P, T8A; T8M = VADD(T8F, T8G); T8H = VSUB(T8F, T8G); T8P = VSUB(T8w, T8z); T8A = VADD(T8w, T8z); { V T8E, T8O, T8j, T7n; T8E = VFMA(LDK(KP956940335), T8D, T8C); T8O = VFNMS(LDK(KP956940335), T8D, T8C); T8j = VFNMS(LDK(KP956940335), T7m, T77); T7n = VFMA(LDK(KP956940335), T7m, T77); { V T8t, T8L, T8m, T8c; T8t = VFNMS(LDK(KP956940335), T8s, T8r); T8L = VFMA(LDK(KP956940335), T8s, T8r); T8m = VFNMS(LDK(KP956940335), T8b, T88); T8c = VFMA(LDK(KP956940335), T8b, T88); { V T8K, T8I, T8S, T8Q; T8K = VFMA(LDK(KP803207531), T8H, T8E); T8I = VFNMS(LDK(KP803207531), T8H, T8E); T8S = VFNMS(LDK(KP803207531), T8P, T8O); T8Q = VFMA(LDK(KP803207531), T8P, T8O); { V T8p, T8l, T8h, T81; T8p = VFNMS(LDK(KP989176509), T8k, T8j); T8l = VFMA(LDK(KP989176509), T8k, T8j); T8h = VFMA(LDK(KP989176509), T80, T7n); T81 = VFNMS(LDK(KP989176509), T80, T7n); { V T8J, T8B, T8R, T8N; T8J = VFMA(LDK(KP803207531), T8A, T8t); T8B = VFNMS(LDK(KP803207531), T8A, T8t); T8R = VFMA(LDK(KP803207531), T8M, T8L); T8N = VFNMS(LDK(KP803207531), T8M, T8L); { V T8q, T8o, T8i, T8g; T8q = VFMA(LDK(KP989176509), T8n, T8m); T8o = VFNMS(LDK(KP989176509), T8n, T8m); T8i = VFMA(LDK(KP989176509), T8f, T8c); T8g = VFNMS(LDK(KP989176509), T8f, T8c); ST(&(xo[WS(os, 13)]), VFMAI(T8K, T8J), ovs, &(xo[WS(os, 1)])); ST(&(xo[WS(os, 115)]), VFNMSI(T8K, T8J), ovs, &(xo[WS(os, 1)])); ST(&(xo[WS(os, 77)]), VFMAI(T8I, T8B), ovs, &(xo[WS(os, 1)])); ST(&(xo[WS(os, 51)]), VFNMSI(T8I, T8B), ovs, &(xo[WS(os, 1)])); ST(&(xo[WS(os, 109)]), VFMAI(T8S, T8R), ovs, &(xo[WS(os, 1)])); ST(&(xo[WS(os, 19)]), VFNMSI(T8S, T8R), ovs, &(xo[WS(os, 1)])); ST(&(xo[WS(os, 83)]), VFNMSI(T8Q, T8N), ovs, &(xo[WS(os, 1)])); ST(&(xo[WS(os, 45)]), VFMAI(T8Q, T8N), ovs, &(xo[WS(os, 1)])); ST(&(xo[WS(os, 99)]), VFNMSI(T8q, T8p), ovs, &(xo[WS(os, 1)])); ST(&(xo[WS(os, 29)]), VFMAI(T8q, T8p), ovs, &(xo[WS(os, 1)])); ST(&(xo[WS(os, 93)]), VFMAI(T8o, T8l), ovs, &(xo[WS(os, 1)])); ST(&(xo[WS(os, 35)]), VFNMSI(T8o, T8l), ovs, &(xo[WS(os, 1)])); ST(&(xo[WS(os, 125)]), VFMAI(T8i, T8h), ovs, &(xo[WS(os, 1)])); ST(&(xo[WS(os, 3)]), VFNMSI(T8i, T8h), ovs, &(xo[WS(os, 1)])); ST(&(xo[WS(os, 61)]), VFMAI(T8g, T81), ovs, &(xo[WS(os, 1)])); ST(&(xo[WS(os, 67)]), VFNMSI(T8g, T81), ovs, &(xo[WS(os, 1)])); T3t = VFMA(LDK(KP980785280), T3s, T3p); T5l = VFNMS(LDK(KP980785280), T3s, T3p); } } } } } } } } } T5n = VFNMS(LDK(KP980785280), T4u, T3V); T4v = VFMA(LDK(KP980785280), T4u, T3V); T4I = VFMA(LDK(KP980785280), T4H, T4E); T5o = VFNMS(LDK(KP980785280), T4H, T4E); { V T53, T3u, T54, T4J, T5v, T5m, T5w, T5p, T10; T6b = VSUB(TI, TZ); T10 = VADD(TI, TZ); T53 = VFMA(LDK(KP049126849), T3g, T3t); T3u = VFNMS(LDK(KP049126849), T3t, T3g); T54 = VFMA(LDK(KP049126849), T4v, T4I); T4J = VFNMS(LDK(KP049126849), T4I, T4v); T5v = VFNMS(LDK(KP906347169), T5k, T5l); T5m = VFMA(LDK(KP906347169), T5l, T5k); T5w = VFNMS(LDK(KP906347169), T5n, T5o); T5p = VFMA(LDK(KP906347169), T5o, T5n); T11 = VFMA(LDK(KP980785280), T10, Tr); T5h = VFNMS(LDK(KP980785280), T10, Tr); T5a = VADD(T53, T54); T55 = VSUB(T53, T54); T5d = VSUB(T3u, T4J); T4K = VADD(T3u, T4J); T5C = VADD(T5v, T5w); T5x = VSUB(T5v, T5w); T5F = VSUB(T5m, T5p); T5q = VADD(T5m, T5p); T4X = VSUB(T4V, T4W); T5K = VADD(T4V, T4W); } T4Z = VFMA(LDK(KP098491403), T1s, T1B); T1C = VFNMS(LDK(KP098491403), T1B, T1s); T2d = VFNMS(LDK(KP098491403), T2c, T23); T50 = VFMA(LDK(KP098491403), T23, T2c); } { V T9y, T9t, T9B, T9i, T9n, T9o, T9F, T8V, T9Q, T9m, T9R, T92, Ta0, T9V, Ta3; V T9O; { V T9I, T9J, T9L, T9d, T5s, T4Y, T5t, T2e, T5i, T51, T9r, T9a, T9g, T9M, T96; V T99; T9I = VFMA(LDK(KP831469612), T95, T94); T96 = VFNMS(LDK(KP831469612), T95, T94); T99 = VFNMS(LDK(KP831469612), T98, T97); T9J = VFMA(LDK(KP831469612), T98, T97); T9L = VFMA(LDK(KP831469612), T9c, T9b); T9d = VFNMS(LDK(KP831469612), T9c, T9b); T5s = VFNMS(LDK(KP980785280), T4X, T4U); T4Y = VFMA(LDK(KP980785280), T4X, T4U); T5t = VSUB(T1C, T2d); T2e = VADD(T1C, T2d); T5i = VADD(T4Z, T50); T51 = VSUB(T4Z, T50); T9r = VFNMS(LDK(KP599376933), T96, T99); T9a = VFMA(LDK(KP599376933), T99, T96); T9g = VFNMS(LDK(KP831469612), T9f, T9e); T9M = VFMA(LDK(KP831469612), T9f, T9e); { V T5u, T5E, T8Y, T91; T5u = VFNMS(LDK(KP995184726), T5t, T5s); T5E = VFMA(LDK(KP995184726), T5t, T5s); { V T59, T2f, T5j, T5B; T59 = VFNMS(LDK(KP995184726), T2e, T11); T2f = VFMA(LDK(KP995184726), T2e, T11); T5j = VFMA(LDK(KP995184726), T5i, T5h); T5B = VFNMS(LDK(KP995184726), T5i, T5h); { V T5c, T52, T9s, T9h; T5c = VFNMS(LDK(KP995184726), T51, T4Y); T52 = VFMA(LDK(KP995184726), T51, T4Y); T9s = VFNMS(LDK(KP599376933), T9d, T9g); T9h = VFMA(LDK(KP599376933), T9g, T9d); { V T5A, T5y, T5I, T5G; T5A = VFMA(LDK(KP740951125), T5x, T5u); T5y = VFNMS(LDK(KP740951125), T5x, T5u); T5I = VFMA(LDK(KP740951125), T5F, T5E); T5G = VFNMS(LDK(KP740951125), T5F, T5E); { V T5f, T5b, T57, T4L; T5f = VFMA(LDK(KP998795456), T5a, T59); T5b = VFNMS(LDK(KP998795456), T5a, T59); T57 = VFMA(LDK(KP998795456), T4K, T2f); T4L = VFNMS(LDK(KP998795456), T4K, T2f); { V T5z, T5r, T5H, T5D; T5z = VFMA(LDK(KP740951125), T5q, T5j); T5r = VFNMS(LDK(KP740951125), T5q, T5j); T5H = VFNMS(LDK(KP740951125), T5C, T5B); T5D = VFMA(LDK(KP740951125), T5C, T5B); { V T5g, T5e, T58, T56; T5g = VFNMS(LDK(KP998795456), T5d, T5c); T5e = VFMA(LDK(KP998795456), T5d, T5c); T58 = VFMA(LDK(KP998795456), T55, T52); T56 = VFNMS(LDK(KP998795456), T55, T52); T9y = VADD(T9r, T9s); T9t = VSUB(T9r, T9s); T9B = VSUB(T9a, T9h); T9i = VADD(T9a, T9h); ST(&(xo[WS(os, 113)]), VFMAI(T5A, T5z), ovs, &(xo[WS(os, 1)])); ST(&(xo[WS(os, 15)]), VFNMSI(T5A, T5z), ovs, &(xo[WS(os, 1)])); ST(&(xo[WS(os, 49)]), VFMAI(T5y, T5r), ovs, &(xo[WS(os, 1)])); ST(&(xo[WS(os, 79)]), VFNMSI(T5y, T5r), ovs, &(xo[WS(os, 1)])); ST(&(xo[WS(os, 111)]), VFNMSI(T5I, T5H), ovs, &(xo[WS(os, 1)])); ST(&(xo[WS(os, 17)]), VFMAI(T5I, T5H), ovs, &(xo[WS(os, 1)])); ST(&(xo[WS(os, 81)]), VFMAI(T5G, T5D), ovs, &(xo[WS(os, 1)])); ST(&(xo[WS(os, 47)]), VFNMSI(T5G, T5D), ovs, &(xo[WS(os, 1)])); ST(&(xo[WS(os, 97)]), VFMAI(T5g, T5f), ovs, &(xo[WS(os, 1)])); ST(&(xo[WS(os, 31)]), VFNMSI(T5g, T5f), ovs, &(xo[WS(os, 1)])); ST(&(xo[WS(os, 95)]), VFNMSI(T5e, T5b), ovs, &(xo[WS(os, 1)])); ST(&(xo[WS(os, 33)]), VFMAI(T5e, T5b), ovs, &(xo[WS(os, 1)])); ST(&(xo[WS(os, 1)]), VFMAI(T58, T57), ovs, &(xo[WS(os, 1)])); ST(&(xo[WS(os, 127)]), VFNMSI(T58, T57), ovs, &(xo[WS(os, 1)])); ST(&(xo[WS(os, 65)]), VFMAI(T56, T4L), ovs, &(xo[WS(os, 1)])); ST(&(xo[WS(os, 63)]), VFNMSI(T56, T4L), ovs, &(xo[WS(os, 1)])); } } } } } } T9n = VFNMS(LDK(KP534511135), T8W, T8X); T8Y = VFMA(LDK(KP534511135), T8X, T8W); T91 = VFMA(LDK(KP534511135), T90, T8Z); T9o = VFNMS(LDK(KP534511135), T8Z, T90); { V T9T, T9K, T9U, T9N; T9T = VFMA(LDK(KP250486960), T9I, T9J); T9K = VFNMS(LDK(KP250486960), T9J, T9I); T9U = VFMA(LDK(KP250486960), T9L, T9M); T9N = VFNMS(LDK(KP250486960), T9M, T9L); T9F = VFNMS(LDK(KP831469612), T8U, T8T); T8V = VFMA(LDK(KP831469612), T8U, T8T); T9Q = VFMA(LDK(KP831469612), T9l, T9k); T9m = VFNMS(LDK(KP831469612), T9l, T9k); T9R = VSUB(T8Y, T91); T92 = VADD(T8Y, T91); Ta0 = VADD(T9T, T9U); T9V = VSUB(T9T, T9U); Ta3 = VSUB(T9K, T9N); T9O = VADD(T9K, T9N); } } } { V T6y, T6z, T63, T9Y, T9W, Ta6, Ta4, T9D, T9z, T9v, T9j, T6h, T60, T9H, T9Z; V T9A, T9q, T66, T9X, T9P; { V T5W, T9S, Ta2, T9x, T93, T5Z, T9G, T9p; T6y = VFMA(LDK(KP980785280), T5V, T5U); T5W = VFNMS(LDK(KP980785280), T5V, T5U); T9S = VFMA(LDK(KP881921264), T9R, T9Q); Ta2 = VFNMS(LDK(KP881921264), T9R, T9Q); T9x = VFNMS(LDK(KP881921264), T92, T8V); T93 = VFMA(LDK(KP881921264), T92, T8V); T5Z = VFMA(LDK(KP980785280), T5Y, T5X); T6z = VFNMS(LDK(KP980785280), T5Y, T5X); T6B = VFMA(LDK(KP980785280), T62, T61); T63 = VFNMS(LDK(KP980785280), T62, T61); T9G = VADD(T9n, T9o); T9p = VSUB(T9n, T9o); T9Y = VFMA(LDK(KP970031253), T9V, T9S); T9W = VFNMS(LDK(KP970031253), T9V, T9S); Ta6 = VFNMS(LDK(KP970031253), Ta3, Ta2); Ta4 = VFMA(LDK(KP970031253), Ta3, Ta2); T9D = VFNMS(LDK(KP857728610), T9y, T9x); T9z = VFMA(LDK(KP857728610), T9y, T9x); T9v = VFMA(LDK(KP857728610), T9i, T93); T9j = VFNMS(LDK(KP857728610), T9i, T93); T6h = VFMA(LDK(KP472964775), T5W, T5Z); T60 = VFNMS(LDK(KP472964775), T5Z, T5W); T9H = VFMA(LDK(KP881921264), T9G, T9F); T9Z = VFNMS(LDK(KP881921264), T9G, T9F); T9A = VFMA(LDK(KP881921264), T9p, T9m); T9q = VFNMS(LDK(KP881921264), T9p, T9m); T66 = VFMA(LDK(KP980785280), T65, T64); T6C = VFNMS(LDK(KP980785280), T65, T64); } T9X = VFMA(LDK(KP970031253), T9O, T9H); T9P = VFNMS(LDK(KP970031253), T9O, T9H); { V Ta5, Ta1, T9E, T9C; Ta5 = VFMA(LDK(KP970031253), Ta0, T9Z); Ta1 = VFNMS(LDK(KP970031253), Ta0, T9Z); T9E = VFMA(LDK(KP857728610), T9B, T9A); T9C = VFNMS(LDK(KP857728610), T9B, T9A); { V T9w, T9u, T6i, T67; T9w = VFMA(LDK(KP857728610), T9t, T9q); T9u = VFNMS(LDK(KP857728610), T9t, T9q); T6i = VFMA(LDK(KP472964775), T63, T66); T67 = VFNMS(LDK(KP472964775), T66, T63); T6J = VFNMS(LDK(KP357805721), T6y, T6z); T6A = VFMA(LDK(KP357805721), T6z, T6y); ST(&(xo[WS(os, 5)]), VFMAI(T9Y, T9X), ovs, &(xo[WS(os, 1)])); ST(&(xo[WS(os, 123)]), VFNMSI(T9Y, T9X), ovs, &(xo[WS(os, 1)])); ST(&(xo[WS(os, 69)]), VFMAI(T9W, T9P), ovs, &(xo[WS(os, 1)])); ST(&(xo[WS(os, 59)]), VFNMSI(T9W, T9P), ovs, &(xo[WS(os, 1)])); ST(&(xo[WS(os, 101)]), VFMAI(Ta6, Ta5), ovs, &(xo[WS(os, 1)])); ST(&(xo[WS(os, 27)]), VFNMSI(Ta6, Ta5), ovs, &(xo[WS(os, 1)])); ST(&(xo[WS(os, 91)]), VFNMSI(Ta4, Ta1), ovs, &(xo[WS(os, 1)])); ST(&(xo[WS(os, 37)]), VFMAI(Ta4, Ta1), ovs, &(xo[WS(os, 1)])); ST(&(xo[WS(os, 107)]), VFNMSI(T9E, T9D), ovs, &(xo[WS(os, 1)])); ST(&(xo[WS(os, 21)]), VFMAI(T9E, T9D), ovs, &(xo[WS(os, 1)])); ST(&(xo[WS(os, 85)]), VFMAI(T9C, T9z), ovs, &(xo[WS(os, 1)])); ST(&(xo[WS(os, 43)]), VFNMSI(T9C, T9z), ovs, &(xo[WS(os, 1)])); ST(&(xo[WS(os, 117)]), VFMAI(T9w, T9v), ovs, &(xo[WS(os, 1)])); ST(&(xo[WS(os, 11)]), VFNMSI(T9w, T9v), ovs, &(xo[WS(os, 1)])); ST(&(xo[WS(os, 53)]), VFMAI(T9u, T9j), ovs, &(xo[WS(os, 1)])); ST(&(xo[WS(os, 75)]), VFNMSI(T9u, T9j), ovs, &(xo[WS(os, 1)])); T6o = VADD(T6h, T6i); T6j = VSUB(T6h, T6i); T6r = VSUB(T60, T67); T68 = VADD(T60, T67); } } T6d = VFMA(LDK(KP820678790), T5M, T5N); T5O = VFNMS(LDK(KP820678790), T5N, T5M); T5R = VFNMS(LDK(KP820678790), T5Q, T5P); T6e = VFMA(LDK(KP820678790), T5P, T5Q); } } } } } } T6D = VFMA(LDK(KP357805721), T6C, T6B); T6K = VFNMS(LDK(KP357805721), T6B, T6C); { V T5L, T6v, T6c, T6G; T5L = VFNMS(LDK(KP980785280), T5K, T5J); T6v = VFMA(LDK(KP980785280), T5K, T5J); T6c = VFMA(LDK(KP980785280), T6b, T6a); T6G = VFNMS(LDK(KP980785280), T6b, T6a); { V T5S, T6H, T6f, T6w; T5S = VADD(T5O, T5R); T6H = VSUB(T5O, T5R); T6f = VSUB(T6d, T6e); T6w = VADD(T6d, T6e); { V T6L, T6Q, T6E, T6T; T6L = VSUB(T6J, T6K); T6Q = VADD(T6J, T6K); T6E = VADD(T6A, T6D); T6T = VSUB(T6A, T6D); { V T6S, T6I, T5T, T6n; T6S = VFNMS(LDK(KP773010453), T6H, T6G); T6I = VFMA(LDK(KP773010453), T6H, T6G); T5T = VFNMS(LDK(KP773010453), T5S, T5L); T6n = VFMA(LDK(KP773010453), T5S, T5L); { V T6P, T6x, T6g, T6q; T6P = VFNMS(LDK(KP773010453), T6w, T6v); T6x = VFMA(LDK(KP773010453), T6w, T6v); T6g = VFMA(LDK(KP773010453), T6f, T6c); T6q = VFNMS(LDK(KP773010453), T6f, T6c); { V T6M, T6O, T6U, T6W; T6M = VFNMS(LDK(KP941544065), T6L, T6I); T6O = VFMA(LDK(KP941544065), T6L, T6I); T6U = VFNMS(LDK(KP941544065), T6T, T6S); T6W = VFMA(LDK(KP941544065), T6T, T6S); { V T6p, T6t, T69, T6l; T6p = VFNMS(LDK(KP903989293), T6o, T6n); T6t = VFMA(LDK(KP903989293), T6o, T6n); T69 = VFNMS(LDK(KP903989293), T68, T5T); T6l = VFMA(LDK(KP903989293), T68, T5T); { V T6F, T6N, T6R, T6V; T6F = VFNMS(LDK(KP941544065), T6E, T6x); T6N = VFMA(LDK(KP941544065), T6E, T6x); T6R = VFMA(LDK(KP941544065), T6Q, T6P); T6V = VFNMS(LDK(KP941544065), T6Q, T6P); { V T6s, T6u, T6k, T6m; T6s = VFMA(LDK(KP903989293), T6r, T6q); T6u = VFNMS(LDK(KP903989293), T6r, T6q); T6k = VFNMS(LDK(KP903989293), T6j, T6g); T6m = VFMA(LDK(KP903989293), T6j, T6g); ST(&(xo[WS(os, 121)]), VFMAI(T6O, T6N), ovs, &(xo[WS(os, 1)])); ST(&(xo[WS(os, 7)]), VFNMSI(T6O, T6N), ovs, &(xo[WS(os, 1)])); ST(&(xo[WS(os, 57)]), VFMAI(T6M, T6F), ovs, &(xo[WS(os, 1)])); ST(&(xo[WS(os, 71)]), VFNMSI(T6M, T6F), ovs, &(xo[WS(os, 1)])); ST(&(xo[WS(os, 103)]), VFNMSI(T6W, T6V), ovs, &(xo[WS(os, 1)])); ST(&(xo[WS(os, 25)]), VFMAI(T6W, T6V), ovs, &(xo[WS(os, 1)])); ST(&(xo[WS(os, 89)]), VFMAI(T6U, T6R), ovs, &(xo[WS(os, 1)])); ST(&(xo[WS(os, 39)]), VFNMSI(T6U, T6R), ovs, &(xo[WS(os, 1)])); ST(&(xo[WS(os, 105)]), VFMAI(T6u, T6t), ovs, &(xo[WS(os, 1)])); ST(&(xo[WS(os, 23)]), VFNMSI(T6u, T6t), ovs, &(xo[WS(os, 1)])); ST(&(xo[WS(os, 87)]), VFNMSI(T6s, T6p), ovs, &(xo[WS(os, 1)])); ST(&(xo[WS(os, 41)]), VFMAI(T6s, T6p), ovs, &(xo[WS(os, 1)])); ST(&(xo[WS(os, 9)]), VFMAI(T6m, T6l), ovs, &(xo[WS(os, 1)])); ST(&(xo[WS(os, 119)]), VFNMSI(T6m, T6l), ovs, &(xo[WS(os, 1)])); ST(&(xo[WS(os, 73)]), VFMAI(T6k, T69), ovs, &(xo[WS(os, 1)])); ST(&(xo[WS(os, 55)]), VFNMSI(T6k, T69), ovs, &(xo[WS(os, 1)])); } } } } } } } } } } } VLEAVE(); } static const kdft_desc desc = { 128, XSIMD_STRING("n1bv_128"), {440, 0, 642, 0}, &GENUS, 0, 0, 0, 0 }; void XSIMD(codelet_n1bv_128) (planner *p) { X(kdft_register) (p, n1bv_128, &desc); } #else /* HAVE_FMA */ /* Generated by: ../../../genfft/gen_notw_c.native -simd -compact -variables 4 -pipeline-latency 8 -sign 1 -n 128 -name n1bv_128 -include n1b.h */ /* * This function contains 1082 FP additions, 330 FP multiplications, * (or, 938 additions, 186 multiplications, 144 fused multiply/add), * 194 stack variables, 31 constants, and 256 memory accesses */ #include "n1b.h" static void n1bv_128(const R *ri, const R *ii, R *ro, R *io, stride is, stride os, INT v, INT ivs, INT ovs) { DVK(KP146730474, +0.146730474455361751658850129646717819706215317); DVK(KP989176509, +0.989176509964780973451673738016243063983689533); DVK(KP595699304, +0.595699304492433343467036528829969889511926338); DVK(KP803207531, +0.803207531480644909806676512963141923879569427); DVK(KP049067674, +0.049067674327418014254954976942682658314745363); DVK(KP998795456, +0.998795456205172392714771604759100694443203615); DVK(KP671558954, +0.671558954847018400625376850427421803228750632); DVK(KP740951125, +0.740951125354959091175616897495162729728955309); DVK(KP514102744, +0.514102744193221726593693838968815772608049120); DVK(KP857728610, +0.857728610000272069902269984284770137042490799); DVK(KP242980179, +0.242980179903263889948274162077471118320990783); DVK(KP970031253, +0.970031253194543992603984207286100251456865962); DVK(KP427555093, +0.427555093430282094320966856888798534304578629); DVK(KP903989293, +0.903989293123443331586200297230537048710132025); DVK(KP336889853, +0.336889853392220050689253212619147570477766780); DVK(KP941544065, +0.941544065183020778412509402599502357185589796); DVK(KP634393284, +0.634393284163645498215171613225493370675687095); DVK(KP773010453, +0.773010453362736960810906609758469800971041293); DVK(KP098017140, +0.098017140329560601994195563888641845861136673); DVK(KP995184726, +0.995184726672196886244836953109479921575474869); DVK(KP290284677, +0.290284677254462367636192375817395274691476278); DVK(KP956940335, +0.956940335732208864935797886980269969482849206); DVK(KP471396736, +0.471396736825997648556387625905254377657460319); DVK(KP881921264, +0.881921264348355029712756863660388349508442621); DVK(KP980785280, +0.980785280403230449126182236134239036973933731); DVK(KP195090322, +0.195090322016128267848284868477022240927691618); DVK(KP555570233, +0.555570233019602224742830813948532874374937191); DVK(KP831469612, +0.831469612302545237078788377617905756738560812); DVK(KP382683432, +0.382683432365089771728459984030398866761344562); DVK(KP923879532, +0.923879532511286756128183189396788286822416626); DVK(KP707106781, +0.707106781186547524400844362104849039284835938); { INT i; const R *xi; R *xo; xi = ii; xo = io; for (i = v; i > 0; i = i - VL, xi = xi + (VL * ivs), xo = xo + (VL * ovs), MAKE_VOLATILE_STRIDE(256, is), MAKE_VOLATILE_STRIDE(256, os)) { V T49, T6e, Tev, TgK, TfA, TgL, T4U, T5J, T7R, T9o, Tah, TdG, Tcw, TdB, T84; V T8T, Tfk, Tfo, T1G, T64, Tgs, Th6, T2p, T62, T7t, T9c, Tce, Tdm, T7i, T9e; V Tc8, Tdp, TgF, TgG, T4q, T4V, TeC, Tfx, T4H, T4W, T7X, T86, Tcr, TdH, T7U; V T85, Taw, TdC, Tf3, Tf7, Tr, T5X, Tgl, Th3, T1a, T5V, T7a, T95, TbD, Tdf; V T6Z, T97, Tbx, Tdi, Tgy, Tgz, TgA, TaN, Tdv, TeK, Tfu, T2W, T5M, T35, T5N; V T7F, T8X, TaI, Tdu, T7C, T8W, TgB, TgC, TgD, Tb4, Tdy, TeR, Tfv, T3x, T5P; V T3G, T5Q, T7M, T90, TaZ, Tdx, T7J, T8Z, Tbm, Tdg, TbG, Tdj, Tgo, Th4, Tf0; V Tf8, T76, T98, T7d, T94, T10, T5Y, T1d, T5U, TbX, Tdn, Tch, Tdq, Tgv, Th7; V Tfh, Tfp, T7p, T9f, T7w, T9b, T2f, T65, T2s, T61; { V T47, Ta8, T4O, Ta7, T44, Tcu, T4P, Tct, Taa, Tab, T3P, Tac, T4R, Tad, Tae; V T3W, Taf, T4S; { V T45, T46, T4M, T4N; T45 = LD(&(xi[WS(is, 32)]), ivs, &(xi[0])); T46 = LD(&(xi[WS(is, 96)]), ivs, &(xi[0])); T47 = VSUB(T45, T46); Ta8 = VADD(T45, T46); T4M = LD(&(xi[0]), ivs, &(xi[0])); T4N = LD(&(xi[WS(is, 64)]), ivs, &(xi[0])); T4O = VSUB(T4M, T4N); Ta7 = VADD(T4M, T4N); } { V T3Y, T3Z, T40, T41, T42, T43; T3Y = LD(&(xi[WS(is, 16)]), ivs, &(xi[0])); T3Z = LD(&(xi[WS(is, 80)]), ivs, &(xi[0])); T40 = VSUB(T3Y, T3Z); T41 = LD(&(xi[WS(is, 112)]), ivs, &(xi[0])); T42 = LD(&(xi[WS(is, 48)]), ivs, &(xi[0])); T43 = VSUB(T41, T42); T44 = VMUL(LDK(KP707106781), VSUB(T40, T43)); Tcu = VADD(T41, T42); T4P = VMUL(LDK(KP707106781), VADD(T40, T43)); Tct = VADD(T3Y, T3Z); } { V T3L, T3O, T3S, T3V; { V T3J, T3K, T3M, T3N; T3J = LD(&(xi[WS(is, 8)]), ivs, &(xi[0])); T3K = LD(&(xi[WS(is, 72)]), ivs, &(xi[0])); T3L = VSUB(T3J, T3K); Taa = VADD(T3J, T3K); T3M = LD(&(xi[WS(is, 40)]), ivs, &(xi[0])); T3N = LD(&(xi[WS(is, 104)]), ivs, &(xi[0])); T3O = VSUB(T3M, T3N); Tab = VADD(T3M, T3N); } T3P = VFNMS(LDK(KP382683432), T3O, VMUL(LDK(KP923879532), T3L)); Tac = VSUB(Taa, Tab); T4R = VFMA(LDK(KP382683432), T3L, VMUL(LDK(KP923879532), T3O)); { V T3Q, T3R, T3T, T3U; T3Q = LD(&(xi[WS(is, 120)]), ivs, &(xi[0])); T3R = LD(&(xi[WS(is, 56)]), ivs, &(xi[0])); T3S = VSUB(T3Q, T3R); Tad = VADD(T3Q, T3R); T3T = LD(&(xi[WS(is, 24)]), ivs, &(xi[0])); T3U = LD(&(xi[WS(is, 88)]), ivs, &(xi[0])); T3V = VSUB(T3T, T3U); Tae = VADD(T3T, T3U); } T3W = VFMA(LDK(KP923879532), T3S, VMUL(LDK(KP382683432), T3V)); Taf = VSUB(Tad, Tae); T4S = VFNMS(LDK(KP382683432), T3S, VMUL(LDK(KP923879532), T3V)); } { V T3X, T48, Tet, Teu; T3X = VSUB(T3P, T3W); T48 = VSUB(T44, T47); T49 = VSUB(T3X, T48); T6e = VADD(T48, T3X); Tet = VADD(Ta7, Ta8); Teu = VADD(Tct, Tcu); Tev = VSUB(Tet, Teu); TgK = VADD(Tet, Teu); } { V Tfy, Tfz, T4Q, T4T; Tfy = VADD(Taa, Tab); Tfz = VADD(Tad, Tae); TfA = VSUB(Tfy, Tfz); TgL = VADD(Tfy, Tfz); T4Q = VSUB(T4O, T4P); T4T = VSUB(T4R, T4S); T4U = VSUB(T4Q, T4T); T5J = VADD(T4Q, T4T); } { V T7P, T7Q, Ta9, Tag; T7P = VADD(T4R, T4S); T7Q = VADD(T47, T44); T7R = VSUB(T7P, T7Q); T9o = VADD(T7Q, T7P); Ta9 = VSUB(Ta7, Ta8); Tag = VMUL(LDK(KP707106781), VADD(Tac, Taf)); Tah = VSUB(Ta9, Tag); TdG = VADD(Ta9, Tag); } { V Tcs, Tcv, T82, T83; Tcs = VMUL(LDK(KP707106781), VSUB(Tac, Taf)); Tcv = VSUB(Tct, Tcu); Tcw = VSUB(Tcs, Tcv); TdB = VADD(Tcv, Tcs); T82 = VADD(T4O, T4P); T83 = VADD(T3P, T3W); T84 = VSUB(T82, T83); T8T = VADD(T82, T83); } } { V Tca, Tcb, T1i, Tfm, T2n, Tc5, Tc6, T1p, Tfn, T2k, T1x, Tfi, T2h, Tc0, T1E; V Tfj, T2i, Tc3, T1l, T1o, Tcc, Tcd; { V T1g, T1h, T2l, T2m; T1g = LD(&(xi[WS(is, 127)]), ivs, &(xi[WS(is, 1)])); T1h = LD(&(xi[WS(is, 63)]), ivs, &(xi[WS(is, 1)])); Tca = VADD(T1g, T1h); T2l = LD(&(xi[WS(is, 31)]), ivs, &(xi[WS(is, 1)])); T2m = LD(&(xi[WS(is, 95)]), ivs, &(xi[WS(is, 1)])); Tcb = VADD(T2l, T2m); T1i = VSUB(T1g, T1h); Tfm = VADD(Tca, Tcb); T2n = VSUB(T2l, T2m); } { V T1j, T1k, T1m, T1n; T1j = LD(&(xi[WS(is, 15)]), ivs, &(xi[WS(is, 1)])); T1k = LD(&(xi[WS(is, 79)]), ivs, &(xi[WS(is, 1)])); T1l = VSUB(T1j, T1k); Tc5 = VADD(T1j, T1k); T1m = LD(&(xi[WS(is, 111)]), ivs, &(xi[WS(is, 1)])); T1n = LD(&(xi[WS(is, 47)]), ivs, &(xi[WS(is, 1)])); T1o = VSUB(T1m, T1n); Tc6 = VADD(T1m, T1n); } T1p = VMUL(LDK(KP707106781), VADD(T1l, T1o)); Tfn = VADD(Tc5, Tc6); T2k = VMUL(LDK(KP707106781), VSUB(T1l, T1o)); { V T1t, TbY, T1w, TbZ; { V T1r, T1s, T1u, T1v; T1r = LD(&(xi[WS(is, 7)]), ivs, &(xi[WS(is, 1)])); T1s = LD(&(xi[WS(is, 71)]), ivs, &(xi[WS(is, 1)])); T1t = VSUB(T1r, T1s); TbY = VADD(T1r, T1s); T1u = LD(&(xi[WS(is, 39)]), ivs, &(xi[WS(is, 1)])); T1v = LD(&(xi[WS(is, 103)]), ivs, &(xi[WS(is, 1)])); T1w = VSUB(T1u, T1v); TbZ = VADD(T1u, T1v); } T1x = VFMA(LDK(KP382683432), T1t, VMUL(LDK(KP923879532), T1w)); Tfi = VADD(TbY, TbZ); T2h = VFNMS(LDK(KP382683432), T1w, VMUL(LDK(KP923879532), T1t)); Tc0 = VSUB(TbY, TbZ); } { V T1A, Tc2, T1D, Tc1; { V T1y, T1z, T1B, T1C; T1y = LD(&(xi[WS(is, 23)]), ivs, &(xi[WS(is, 1)])); T1z = LD(&(xi[WS(is, 87)]), ivs, &(xi[WS(is, 1)])); T1A = VSUB(T1y, T1z); Tc2 = VADD(T1y, T1z); T1B = LD(&(xi[WS(is, 119)]), ivs, &(xi[WS(is, 1)])); T1C = LD(&(xi[WS(is, 55)]), ivs, &(xi[WS(is, 1)])); T1D = VSUB(T1B, T1C); Tc1 = VADD(T1B, T1C); } T1E = VFNMS(LDK(KP382683432), T1D, VMUL(LDK(KP923879532), T1A)); Tfj = VADD(Tc1, Tc2); T2i = VFMA(LDK(KP923879532), T1D, VMUL(LDK(KP382683432), T1A)); Tc3 = VSUB(Tc1, Tc2); } Tfk = VSUB(Tfi, Tfj); Tfo = VSUB(Tfm, Tfn); { V T1q, T1F, Tgq, Tgr; T1q = VSUB(T1i, T1p); T1F = VSUB(T1x, T1E); T1G = VSUB(T1q, T1F); T64 = VADD(T1q, T1F); Tgq = VADD(Tfm, Tfn); Tgr = VADD(Tfi, Tfj); Tgs = VSUB(Tgq, Tgr); Th6 = VADD(Tgq, Tgr); } { V T2j, T2o, T7r, T7s; T2j = VSUB(T2h, T2i); T2o = VSUB(T2k, T2n); T2p = VSUB(T2j, T2o); T62 = VADD(T2o, T2j); T7r = VADD(T1x, T1E); T7s = VADD(T2n, T2k); T7t = VSUB(T7r, T7s); T9c = VADD(T7s, T7r); } Tcc = VSUB(Tca, Tcb); Tcd = VMUL(LDK(KP707106781), VADD(Tc0, Tc3)); Tce = VSUB(Tcc, Tcd); Tdm = VADD(Tcc, Tcd); { V T7g, T7h, Tc4, Tc7; T7g = VADD(T1i, T1p); T7h = VADD(T2h, T2i); T7i = VSUB(T7g, T7h); T9e = VADD(T7g, T7h); Tc4 = VMUL(LDK(KP707106781), VSUB(Tc0, Tc3)); Tc7 = VSUB(Tc5, Tc6); Tc8 = VSUB(Tc4, Tc7); Tdp = VADD(Tc7, Tc4); } } { V T4c, Tew, T4o, Tak, T4A, Tez, T4E, Tau, T4j, Tex, T4l, Tan, T4x, TeA, T4F; V Tar, Tcp, Tcq; { V T4a, T4b, Tai, T4m, T4n, Taj; T4a = LD(&(xi[WS(is, 4)]), ivs, &(xi[0])); T4b = LD(&(xi[WS(is, 68)]), ivs, &(xi[0])); Tai = VADD(T4a, T4b); T4m = LD(&(xi[WS(is, 36)]), ivs, &(xi[0])); T4n = LD(&(xi[WS(is, 100)]), ivs, &(xi[0])); Taj = VADD(T4m, T4n); T4c = VSUB(T4a, T4b); Tew = VADD(Tai, Taj); T4o = VSUB(T4m, T4n); Tak = VSUB(Tai, Taj); } { V T4y, T4z, Tat, T4C, T4D, Tas; T4y = LD(&(xi[WS(is, 28)]), ivs, &(xi[0])); T4z = LD(&(xi[WS(is, 92)]), ivs, &(xi[0])); Tat = VADD(T4y, T4z); T4C = LD(&(xi[WS(is, 124)]), ivs, &(xi[0])); T4D = LD(&(xi[WS(is, 60)]), ivs, &(xi[0])); Tas = VADD(T4C, T4D); T4A = VSUB(T4y, T4z); Tez = VADD(Tas, Tat); T4E = VSUB(T4C, T4D); Tau = VSUB(Tas, Tat); } { V T4f, Tal, T4i, Tam; { V T4d, T4e, T4g, T4h; T4d = LD(&(xi[WS(is, 20)]), ivs, &(xi[0])); T4e = LD(&(xi[WS(is, 84)]), ivs, &(xi[0])); T4f = VSUB(T4d, T4e); Tal = VADD(T4d, T4e); T4g = LD(&(xi[WS(is, 116)]), ivs, &(xi[0])); T4h = LD(&(xi[WS(is, 52)]), ivs, &(xi[0])); T4i = VSUB(T4g, T4h); Tam = VADD(T4g, T4h); } T4j = VMUL(LDK(KP707106781), VADD(T4f, T4i)); Tex = VADD(Tal, Tam); T4l = VMUL(LDK(KP707106781), VSUB(T4f, T4i)); Tan = VSUB(Tal, Tam); } { V T4t, Tap, T4w, Taq; { V T4r, T4s, T4u, T4v; T4r = LD(&(xi[WS(is, 12)]), ivs, &(xi[0])); T4s = LD(&(xi[WS(is, 76)]), ivs, &(xi[0])); T4t = VSUB(T4r, T4s); Tap = VADD(T4r, T4s); T4u = LD(&(xi[WS(is, 108)]), ivs, &(xi[0])); T4v = LD(&(xi[WS(is, 44)]), ivs, &(xi[0])); T4w = VSUB(T4u, T4v); Taq = VADD(T4u, T4v); } T4x = VMUL(LDK(KP707106781), VSUB(T4t, T4w)); TeA = VADD(Tap, Taq); T4F = VMUL(LDK(KP707106781), VADD(T4t, T4w)); Tar = VSUB(Tap, Taq); } TgF = VADD(Tew, Tex); TgG = VADD(Tez, TeA); { V T4k, T4p, Tey, TeB; T4k = VSUB(T4c, T4j); T4p = VSUB(T4l, T4o); T4q = VFNMS(LDK(KP555570233), T4p, VMUL(LDK(KP831469612), T4k)); T4V = VFMA(LDK(KP831469612), T4p, VMUL(LDK(KP555570233), T4k)); Tey = VSUB(Tew, Tex); TeB = VSUB(Tez, TeA); TeC = VMUL(LDK(KP707106781), VADD(Tey, TeB)); Tfx = VMUL(LDK(KP707106781), VSUB(Tey, TeB)); } { V T4B, T4G, T7V, T7W; T4B = VSUB(T4x, T4A); T4G = VSUB(T4E, T4F); T4H = VFMA(LDK(KP555570233), T4B, VMUL(LDK(KP831469612), T4G)); T4W = VFNMS(LDK(KP555570233), T4G, VMUL(LDK(KP831469612), T4B)); T7V = VADD(T4A, T4x); T7W = VADD(T4E, T4F); T7X = VFMA(LDK(KP195090322), T7V, VMUL(LDK(KP980785280), T7W)); T86 = VFNMS(LDK(KP195090322), T7W, VMUL(LDK(KP980785280), T7V)); } Tcp = VFNMS(LDK(KP382683432), Tan, VMUL(LDK(KP923879532), Tak)); Tcq = VFMA(LDK(KP923879532), Tau, VMUL(LDK(KP382683432), Tar)); Tcr = VSUB(Tcp, Tcq); TdH = VADD(Tcp, Tcq); { V T7S, T7T, Tao, Tav; T7S = VADD(T4c, T4j); T7T = VADD(T4o, T4l); T7U = VFNMS(LDK(KP195090322), T7T, VMUL(LDK(KP980785280), T7S)); T85 = VFMA(LDK(KP980785280), T7T, VMUL(LDK(KP195090322), T7S)); Tao = VFMA(LDK(KP382683432), Tak, VMUL(LDK(KP923879532), Tan)); Tav = VFNMS(LDK(KP382683432), Tau, VMUL(LDK(KP923879532), Tar)); Taw = VSUB(Tao, Tav); TdC = VADD(Tao, Tav); } } { V Tbz, TbA, T3, Tf5, T18, Tbu, Tbv, Ta, Tf6, T15, Ti, Tf1, T12, Tbp, Tp; V Tf2, T13, Tbs, T6, T9, TbB, TbC; { V T1, T2, T16, T17; T1 = LD(&(xi[WS(is, 1)]), ivs, &(xi[WS(is, 1)])); T2 = LD(&(xi[WS(is, 65)]), ivs, &(xi[WS(is, 1)])); Tbz = VADD(T1, T2); T16 = LD(&(xi[WS(is, 33)]), ivs, &(xi[WS(is, 1)])); T17 = LD(&(xi[WS(is, 97)]), ivs, &(xi[WS(is, 1)])); TbA = VADD(T16, T17); T3 = VSUB(T1, T2); Tf5 = VADD(Tbz, TbA); T18 = VSUB(T16, T17); } { V T4, T5, T7, T8; T4 = LD(&(xi[WS(is, 17)]), ivs, &(xi[WS(is, 1)])); T5 = LD(&(xi[WS(is, 81)]), ivs, &(xi[WS(is, 1)])); T6 = VSUB(T4, T5); Tbu = VADD(T4, T5); T7 = LD(&(xi[WS(is, 113)]), ivs, &(xi[WS(is, 1)])); T8 = LD(&(xi[WS(is, 49)]), ivs, &(xi[WS(is, 1)])); T9 = VSUB(T7, T8); Tbv = VADD(T7, T8); } Ta = VMUL(LDK(KP707106781), VADD(T6, T9)); Tf6 = VADD(Tbu, Tbv); T15 = VMUL(LDK(KP707106781), VSUB(T6, T9)); { V Te, Tbn, Th, Tbo; { V Tc, Td, Tf, Tg; Tc = LD(&(xi[WS(is, 9)]), ivs, &(xi[WS(is, 1)])); Td = LD(&(xi[WS(is, 73)]), ivs, &(xi[WS(is, 1)])); Te = VSUB(Tc, Td); Tbn = VADD(Tc, Td); Tf = LD(&(xi[WS(is, 41)]), ivs, &(xi[WS(is, 1)])); Tg = LD(&(xi[WS(is, 105)]), ivs, &(xi[WS(is, 1)])); Th = VSUB(Tf, Tg); Tbo = VADD(Tf, Tg); } Ti = VFMA(LDK(KP382683432), Te, VMUL(LDK(KP923879532), Th)); Tf1 = VADD(Tbn, Tbo); T12 = VFNMS(LDK(KP382683432), Th, VMUL(LDK(KP923879532), Te)); Tbp = VSUB(Tbn, Tbo); } { V Tl, Tbr, To, Tbq; { V Tj, Tk, Tm, Tn; Tj = LD(&(xi[WS(is, 25)]), ivs, &(xi[WS(is, 1)])); Tk = LD(&(xi[WS(is, 89)]), ivs, &(xi[WS(is, 1)])); Tl = VSUB(Tj, Tk); Tbr = VADD(Tj, Tk); Tm = LD(&(xi[WS(is, 121)]), ivs, &(xi[WS(is, 1)])); Tn = LD(&(xi[WS(is, 57)]), ivs, &(xi[WS(is, 1)])); To = VSUB(Tm, Tn); Tbq = VADD(Tm, Tn); } Tp = VFNMS(LDK(KP382683432), To, VMUL(LDK(KP923879532), Tl)); Tf2 = VADD(Tbq, Tbr); T13 = VFMA(LDK(KP923879532), To, VMUL(LDK(KP382683432), Tl)); Tbs = VSUB(Tbq, Tbr); } Tf3 = VSUB(Tf1, Tf2); Tf7 = VSUB(Tf5, Tf6); { V Tb, Tq, Tgj, Tgk; Tb = VSUB(T3, Ta); Tq = VSUB(Ti, Tp); Tr = VSUB(Tb, Tq); T5X = VADD(Tb, Tq); Tgj = VADD(Tf5, Tf6); Tgk = VADD(Tf1, Tf2); Tgl = VSUB(Tgj, Tgk); Th3 = VADD(Tgj, Tgk); } { V T14, T19, T78, T79; T14 = VSUB(T12, T13); T19 = VSUB(T15, T18); T1a = VSUB(T14, T19); T5V = VADD(T19, T14); T78 = VADD(Ti, Tp); T79 = VADD(T18, T15); T7a = VSUB(T78, T79); T95 = VADD(T79, T78); } TbB = VSUB(Tbz, TbA); TbC = VMUL(LDK(KP707106781), VADD(Tbp, Tbs)); TbD = VSUB(TbB, TbC); Tdf = VADD(TbB, TbC); { V T6X, T6Y, Tbt, Tbw; T6X = VADD(T3, Ta); T6Y = VADD(T12, T13); T6Z = VSUB(T6X, T6Y); T97 = VADD(T6X, T6Y); Tbt = VMUL(LDK(KP707106781), VSUB(Tbp, Tbs)); Tbw = VSUB(Tbu, Tbv); Tbx = VSUB(Tbt, Tbw); Tdi = VADD(Tbw, Tbt); } } { V TaK, TaJ, T2U, TeE, T2Z, TaF, TaG, T2R, TeF, T30, T2C, TeH, T32, TaA, T2J; V TeI, T33, TaD, T2N, T2Q, TaL, TaM; { V T2S, T2T, T2X, T2Y; T2S = LD(&(xi[WS(is, 34)]), ivs, &(xi[0])); T2T = LD(&(xi[WS(is, 98)]), ivs, &(xi[0])); TaK = VADD(T2S, T2T); T2X = LD(&(xi[WS(is, 2)]), ivs, &(xi[0])); T2Y = LD(&(xi[WS(is, 66)]), ivs, &(xi[0])); TaJ = VADD(T2X, T2Y); T2U = VSUB(T2S, T2T); TeE = VADD(TaJ, TaK); T2Z = VSUB(T2X, T2Y); } { V T2L, T2M, T2O, T2P; T2L = LD(&(xi[WS(is, 18)]), ivs, &(xi[0])); T2M = LD(&(xi[WS(is, 82)]), ivs, &(xi[0])); T2N = VSUB(T2L, T2M); TaF = VADD(T2L, T2M); T2O = LD(&(xi[WS(is, 114)]), ivs, &(xi[0])); T2P = LD(&(xi[WS(is, 50)]), ivs, &(xi[0])); T2Q = VSUB(T2O, T2P); TaG = VADD(T2O, T2P); } T2R = VMUL(LDK(KP707106781), VSUB(T2N, T2Q)); TeF = VADD(TaF, TaG); T30 = VMUL(LDK(KP707106781), VADD(T2N, T2Q)); { V T2y, Tay, T2B, Taz; { V T2w, T2x, T2z, T2A; T2w = LD(&(xi[WS(is, 10)]), ivs, &(xi[0])); T2x = LD(&(xi[WS(is, 74)]), ivs, &(xi[0])); T2y = VSUB(T2w, T2x); Tay = VADD(T2w, T2x); T2z = LD(&(xi[WS(is, 42)]), ivs, &(xi[0])); T2A = LD(&(xi[WS(is, 106)]), ivs, &(xi[0])); T2B = VSUB(T2z, T2A); Taz = VADD(T2z, T2A); } T2C = VFNMS(LDK(KP382683432), T2B, VMUL(LDK(KP923879532), T2y)); TeH = VADD(Tay, Taz); T32 = VFMA(LDK(KP382683432), T2y, VMUL(LDK(KP923879532), T2B)); TaA = VSUB(Tay, Taz); } { V T2F, TaB, T2I, TaC; { V T2D, T2E, T2G, T2H; T2D = LD(&(xi[WS(is, 122)]), ivs, &(xi[0])); T2E = LD(&(xi[WS(is, 58)]), ivs, &(xi[0])); T2F = VSUB(T2D, T2E); TaB = VADD(T2D, T2E); T2G = LD(&(xi[WS(is, 26)]), ivs, &(xi[0])); T2H = LD(&(xi[WS(is, 90)]), ivs, &(xi[0])); T2I = VSUB(T2G, T2H); TaC = VADD(T2G, T2H); } T2J = VFMA(LDK(KP923879532), T2F, VMUL(LDK(KP382683432), T2I)); TeI = VADD(TaB, TaC); T33 = VFNMS(LDK(KP382683432), T2F, VMUL(LDK(KP923879532), T2I)); TaD = VSUB(TaB, TaC); } Tgy = VADD(TeE, TeF); Tgz = VADD(TeH, TeI); TgA = VSUB(Tgy, Tgz); TaL = VSUB(TaJ, TaK); TaM = VMUL(LDK(KP707106781), VADD(TaA, TaD)); TaN = VSUB(TaL, TaM); Tdv = VADD(TaL, TaM); { V TeG, TeJ, T2K, T2V; TeG = VSUB(TeE, TeF); TeJ = VSUB(TeH, TeI); TeK = VFMA(LDK(KP382683432), TeG, VMUL(LDK(KP923879532), TeJ)); Tfu = VFNMS(LDK(KP382683432), TeJ, VMUL(LDK(KP923879532), TeG)); T2K = VSUB(T2C, T2J); T2V = VSUB(T2R, T2U); T2W = VSUB(T2K, T2V); T5M = VADD(T2V, T2K); } { V T31, T34, T7D, T7E; T31 = VSUB(T2Z, T30); T34 = VSUB(T32, T33); T35 = VSUB(T31, T34); T5N = VADD(T31, T34); T7D = VADD(T32, T33); T7E = VADD(T2U, T2R); T7F = VSUB(T7D, T7E); T8X = VADD(T7E, T7D); } { V TaE, TaH, T7A, T7B; TaE = VMUL(LDK(KP707106781), VSUB(TaA, TaD)); TaH = VSUB(TaF, TaG); TaI = VSUB(TaE, TaH); Tdu = VADD(TaH, TaE); T7A = VADD(T2Z, T30); T7B = VADD(T2C, T2J); T7C = VSUB(T7A, T7B); T8W = VADD(T7A, T7B); } } { V Tb1, Tb0, T3v, TeO, T3A, TaW, TaX, T3s, TeP, T3B, T3d, TeL, T3D, TaR, T3k; V TeM, T3E, TaU, T3o, T3r, Tb2, Tb3; { V T3t, T3u, T3y, T3z; T3t = LD(&(xi[WS(is, 30)]), ivs, &(xi[0])); T3u = LD(&(xi[WS(is, 94)]), ivs, &(xi[0])); Tb1 = VADD(T3t, T3u); T3y = LD(&(xi[WS(is, 126)]), ivs, &(xi[0])); T3z = LD(&(xi[WS(is, 62)]), ivs, &(xi[0])); Tb0 = VADD(T3y, T3z); T3v = VSUB(T3t, T3u); TeO = VADD(Tb0, Tb1); T3A = VSUB(T3y, T3z); } { V T3m, T3n, T3p, T3q; T3m = LD(&(xi[WS(is, 14)]), ivs, &(xi[0])); T3n = LD(&(xi[WS(is, 78)]), ivs, &(xi[0])); T3o = VSUB(T3m, T3n); TaW = VADD(T3m, T3n); T3p = LD(&(xi[WS(is, 110)]), ivs, &(xi[0])); T3q = LD(&(xi[WS(is, 46)]), ivs, &(xi[0])); T3r = VSUB(T3p, T3q); TaX = VADD(T3p, T3q); } T3s = VMUL(LDK(KP707106781), VSUB(T3o, T3r)); TeP = VADD(TaW, TaX); T3B = VMUL(LDK(KP707106781), VADD(T3o, T3r)); { V T39, TaP, T3c, TaQ; { V T37, T38, T3a, T3b; T37 = LD(&(xi[WS(is, 6)]), ivs, &(xi[0])); T38 = LD(&(xi[WS(is, 70)]), ivs, &(xi[0])); T39 = VSUB(T37, T38); TaP = VADD(T37, T38); T3a = LD(&(xi[WS(is, 38)]), ivs, &(xi[0])); T3b = LD(&(xi[WS(is, 102)]), ivs, &(xi[0])); T3c = VSUB(T3a, T3b); TaQ = VADD(T3a, T3b); } T3d = VFNMS(LDK(KP382683432), T3c, VMUL(LDK(KP923879532), T39)); TeL = VADD(TaP, TaQ); T3D = VFMA(LDK(KP382683432), T39, VMUL(LDK(KP923879532), T3c)); TaR = VSUB(TaP, TaQ); } { V T3g, TaS, T3j, TaT; { V T3e, T3f, T3h, T3i; T3e = LD(&(xi[WS(is, 118)]), ivs, &(xi[0])); T3f = LD(&(xi[WS(is, 54)]), ivs, &(xi[0])); T3g = VSUB(T3e, T3f); TaS = VADD(T3e, T3f); T3h = LD(&(xi[WS(is, 22)]), ivs, &(xi[0])); T3i = LD(&(xi[WS(is, 86)]), ivs, &(xi[0])); T3j = VSUB(T3h, T3i); TaT = VADD(T3h, T3i); } T3k = VFMA(LDK(KP923879532), T3g, VMUL(LDK(KP382683432), T3j)); TeM = VADD(TaS, TaT); T3E = VFNMS(LDK(KP382683432), T3g, VMUL(LDK(KP923879532), T3j)); TaU = VSUB(TaS, TaT); } TgB = VADD(TeO, TeP); TgC = VADD(TeL, TeM); TgD = VSUB(TgB, TgC); Tb2 = VSUB(Tb0, Tb1); Tb3 = VMUL(LDK(KP707106781), VADD(TaR, TaU)); Tb4 = VSUB(Tb2, Tb3); Tdy = VADD(Tb2, Tb3); { V TeN, TeQ, T3l, T3w; TeN = VSUB(TeL, TeM); TeQ = VSUB(TeO, TeP); TeR = VFNMS(LDK(KP382683432), TeQ, VMUL(LDK(KP923879532), TeN)); Tfv = VFMA(LDK(KP923879532), TeQ, VMUL(LDK(KP382683432), TeN)); T3l = VSUB(T3d, T3k); T3w = VSUB(T3s, T3v); T3x = VSUB(T3l, T3w); T5P = VADD(T3w, T3l); } { V T3C, T3F, T7K, T7L; T3C = VSUB(T3A, T3B); T3F = VSUB(T3D, T3E); T3G = VSUB(T3C, T3F); T5Q = VADD(T3C, T3F); T7K = VADD(T3A, T3B); T7L = VADD(T3d, T3k); T7M = VSUB(T7K, T7L); T90 = VADD(T7K, T7L); } { V TaV, TaY, T7H, T7I; TaV = VMUL(LDK(KP707106781), VSUB(TaR, TaU)); TaY = VSUB(TaW, TaX); TaZ = VSUB(TaV, TaY); Tdx = VADD(TaY, TaV); T7H = VADD(T3D, T3E); T7I = VADD(T3v, T3s); T7J = VSUB(T7H, T7I); T8Z = VADD(T7I, T7H); } } { V TB, TeU, TF, Tba, TS, TeX, TW, Tbh, Ty, TeV, TG, Tbd, TP, TeY, TX; V Tbk; { V Tz, TA, Tb9, TD, TE, Tb8; Tz = LD(&(xi[WS(is, 37)]), ivs, &(xi[WS(is, 1)])); TA = LD(&(xi[WS(is, 101)]), ivs, &(xi[WS(is, 1)])); Tb9 = VADD(Tz, TA); TD = LD(&(xi[WS(is, 5)]), ivs, &(xi[WS(is, 1)])); TE = LD(&(xi[WS(is, 69)]), ivs, &(xi[WS(is, 1)])); Tb8 = VADD(TD, TE); TB = VSUB(Tz, TA); TeU = VADD(Tb8, Tb9); TF = VSUB(TD, TE); Tba = VSUB(Tb8, Tb9); } { V TQ, TR, Tbg, TU, TV, Tbf; TQ = LD(&(xi[WS(is, 29)]), ivs, &(xi[WS(is, 1)])); TR = LD(&(xi[WS(is, 93)]), ivs, &(xi[WS(is, 1)])); Tbg = VADD(TQ, TR); TU = LD(&(xi[WS(is, 125)]), ivs, &(xi[WS(is, 1)])); TV = LD(&(xi[WS(is, 61)]), ivs, &(xi[WS(is, 1)])); Tbf = VADD(TU, TV); TS = VSUB(TQ, TR); TeX = VADD(Tbf, Tbg); TW = VSUB(TU, TV); Tbh = VSUB(Tbf, Tbg); } { V Tu, Tbb, Tx, Tbc; { V Ts, Tt, Tv, Tw; Ts = LD(&(xi[WS(is, 21)]), ivs, &(xi[WS(is, 1)])); Tt = LD(&(xi[WS(is, 85)]), ivs, &(xi[WS(is, 1)])); Tu = VSUB(Ts, Tt); Tbb = VADD(Ts, Tt); Tv = LD(&(xi[WS(is, 117)]), ivs, &(xi[WS(is, 1)])); Tw = LD(&(xi[WS(is, 53)]), ivs, &(xi[WS(is, 1)])); Tx = VSUB(Tv, Tw); Tbc = VADD(Tv, Tw); } Ty = VMUL(LDK(KP707106781), VSUB(Tu, Tx)); TeV = VADD(Tbb, Tbc); TG = VMUL(LDK(KP707106781), VADD(Tu, Tx)); Tbd = VSUB(Tbb, Tbc); } { V TL, Tbi, TO, Tbj; { V TJ, TK, TM, TN; TJ = LD(&(xi[WS(is, 13)]), ivs, &(xi[WS(is, 1)])); TK = LD(&(xi[WS(is, 77)]), ivs, &(xi[WS(is, 1)])); TL = VSUB(TJ, TK); Tbi = VADD(TJ, TK); TM = LD(&(xi[WS(is, 109)]), ivs, &(xi[WS(is, 1)])); TN = LD(&(xi[WS(is, 45)]), ivs, &(xi[WS(is, 1)])); TO = VSUB(TM, TN); Tbj = VADD(TM, TN); } TP = VMUL(LDK(KP707106781), VSUB(TL, TO)); TeY = VADD(Tbi, Tbj); TX = VMUL(LDK(KP707106781), VADD(TL, TO)); Tbk = VSUB(Tbi, Tbj); } { V Tbe, Tbl, TeW, TeZ; Tbe = VFNMS(LDK(KP382683432), Tbd, VMUL(LDK(KP923879532), Tba)); Tbl = VFMA(LDK(KP923879532), Tbh, VMUL(LDK(KP382683432), Tbk)); Tbm = VSUB(Tbe, Tbl); Tdg = VADD(Tbe, Tbl); { V TbE, TbF, Tgm, Tgn; TbE = VFMA(LDK(KP382683432), Tba, VMUL(LDK(KP923879532), Tbd)); TbF = VFNMS(LDK(KP382683432), Tbh, VMUL(LDK(KP923879532), Tbk)); TbG = VSUB(TbE, TbF); Tdj = VADD(TbE, TbF); Tgm = VADD(TeU, TeV); Tgn = VADD(TeX, TeY); Tgo = VSUB(Tgm, Tgn); Th4 = VADD(Tgm, Tgn); } TeW = VSUB(TeU, TeV); TeZ = VSUB(TeX, TeY); Tf0 = VMUL(LDK(KP707106781), VSUB(TeW, TeZ)); Tf8 = VMUL(LDK(KP707106781), VADD(TeW, TeZ)); { V T72, T7b, T75, T7c; { V T70, T71, T73, T74; T70 = VADD(TB, Ty); T71 = VADD(TF, TG); T72 = VFMA(LDK(KP980785280), T70, VMUL(LDK(KP195090322), T71)); T7b = VFNMS(LDK(KP195090322), T70, VMUL(LDK(KP980785280), T71)); T73 = VADD(TS, TP); T74 = VADD(TW, TX); T75 = VFNMS(LDK(KP195090322), T74, VMUL(LDK(KP980785280), T73)); T7c = VFMA(LDK(KP195090322), T73, VMUL(LDK(KP980785280), T74)); } T76 = VSUB(T72, T75); T98 = VADD(T7b, T7c); T7d = VSUB(T7b, T7c); T94 = VADD(T72, T75); } { V TI, T1b, TZ, T1c; { V TC, TH, TT, TY; TC = VSUB(Ty, TB); TH = VSUB(TF, TG); TI = VFMA(LDK(KP831469612), TC, VMUL(LDK(KP555570233), TH)); T1b = VFNMS(LDK(KP555570233), TC, VMUL(LDK(KP831469612), TH)); TT = VSUB(TP, TS); TY = VSUB(TW, TX); TZ = VFNMS(LDK(KP555570233), TY, VMUL(LDK(KP831469612), TT)); T1c = VFMA(LDK(KP555570233), TT, VMUL(LDK(KP831469612), TY)); } T10 = VSUB(TI, TZ); T5Y = VADD(T1b, T1c); T1d = VSUB(T1b, T1c); T5U = VADD(TI, TZ); } } } { V T1Q, Tfb, T1U, TbL, T27, Tfe, T2b, TbS, T1N, Tfc, T1V, TbO, T24, Tff, T2c; V TbV; { V T1O, T1P, TbK, T1S, T1T, TbJ; T1O = LD(&(xi[WS(is, 35)]), ivs, &(xi[WS(is, 1)])); T1P = LD(&(xi[WS(is, 99)]), ivs, &(xi[WS(is, 1)])); TbK = VADD(T1O, T1P); T1S = LD(&(xi[WS(is, 3)]), ivs, &(xi[WS(is, 1)])); T1T = LD(&(xi[WS(is, 67)]), ivs, &(xi[WS(is, 1)])); TbJ = VADD(T1S, T1T); T1Q = VSUB(T1O, T1P); Tfb = VADD(TbJ, TbK); T1U = VSUB(T1S, T1T); TbL = VSUB(TbJ, TbK); } { V T25, T26, TbR, T29, T2a, TbQ; T25 = LD(&(xi[WS(is, 27)]), ivs, &(xi[WS(is, 1)])); T26 = LD(&(xi[WS(is, 91)]), ivs, &(xi[WS(is, 1)])); TbR = VADD(T25, T26); T29 = LD(&(xi[WS(is, 123)]), ivs, &(xi[WS(is, 1)])); T2a = LD(&(xi[WS(is, 59)]), ivs, &(xi[WS(is, 1)])); TbQ = VADD(T29, T2a); T27 = VSUB(T25, T26); Tfe = VADD(TbQ, TbR); T2b = VSUB(T29, T2a); TbS = VSUB(TbQ, TbR); } { V T1J, TbM, T1M, TbN; { V T1H, T1I, T1K, T1L; T1H = LD(&(xi[WS(is, 19)]), ivs, &(xi[WS(is, 1)])); T1I = LD(&(xi[WS(is, 83)]), ivs, &(xi[WS(is, 1)])); T1J = VSUB(T1H, T1I); TbM = VADD(T1H, T1I); T1K = LD(&(xi[WS(is, 115)]), ivs, &(xi[WS(is, 1)])); T1L = LD(&(xi[WS(is, 51)]), ivs, &(xi[WS(is, 1)])); T1M = VSUB(T1K, T1L); TbN = VADD(T1K, T1L); } T1N = VMUL(LDK(KP707106781), VSUB(T1J, T1M)); Tfc = VADD(TbM, TbN); T1V = VMUL(LDK(KP707106781), VADD(T1J, T1M)); TbO = VSUB(TbM, TbN); } { V T20, TbT, T23, TbU; { V T1Y, T1Z, T21, T22; T1Y = LD(&(xi[WS(is, 11)]), ivs, &(xi[WS(is, 1)])); T1Z = LD(&(xi[WS(is, 75)]), ivs, &(xi[WS(is, 1)])); T20 = VSUB(T1Y, T1Z); TbT = VADD(T1Y, T1Z); T21 = LD(&(xi[WS(is, 107)]), ivs, &(xi[WS(is, 1)])); T22 = LD(&(xi[WS(is, 43)]), ivs, &(xi[WS(is, 1)])); T23 = VSUB(T21, T22); TbU = VADD(T21, T22); } T24 = VMUL(LDK(KP707106781), VSUB(T20, T23)); Tff = VADD(TbT, TbU); T2c = VMUL(LDK(KP707106781), VADD(T20, T23)); TbV = VSUB(TbT, TbU); } { V TbP, TbW, Tfd, Tfg; TbP = VFNMS(LDK(KP382683432), TbO, VMUL(LDK(KP923879532), TbL)); TbW = VFMA(LDK(KP923879532), TbS, VMUL(LDK(KP382683432), TbV)); TbX = VSUB(TbP, TbW); Tdn = VADD(TbP, TbW); { V Tcf, Tcg, Tgt, Tgu; Tcf = VFMA(LDK(KP382683432), TbL, VMUL(LDK(KP923879532), TbO)); Tcg = VFNMS(LDK(KP382683432), TbS, VMUL(LDK(KP923879532), TbV)); Tch = VSUB(Tcf, Tcg); Tdq = VADD(Tcf, Tcg); Tgt = VADD(Tfb, Tfc); Tgu = VADD(Tfe, Tff); Tgv = VSUB(Tgt, Tgu); Th7 = VADD(Tgt, Tgu); } Tfd = VSUB(Tfb, Tfc); Tfg = VSUB(Tfe, Tff); Tfh = VMUL(LDK(KP707106781), VSUB(Tfd, Tfg)); Tfp = VMUL(LDK(KP707106781), VADD(Tfd, Tfg)); { V T7l, T7u, T7o, T7v; { V T7j, T7k, T7m, T7n; T7j = VADD(T1Q, T1N); T7k = VADD(T1U, T1V); T7l = VFMA(LDK(KP980785280), T7j, VMUL(LDK(KP195090322), T7k)); T7u = VFNMS(LDK(KP195090322), T7j, VMUL(LDK(KP980785280), T7k)); T7m = VADD(T27, T24); T7n = VADD(T2b, T2c); T7o = VFNMS(LDK(KP195090322), T7n, VMUL(LDK(KP980785280), T7m)); T7v = VFMA(LDK(KP195090322), T7m, VMUL(LDK(KP980785280), T7n)); } T7p = VSUB(T7l, T7o); T9f = VADD(T7u, T7v); T7w = VSUB(T7u, T7v); T9b = VADD(T7l, T7o); } { V T1X, T2q, T2e, T2r; { V T1R, T1W, T28, T2d; T1R = VSUB(T1N, T1Q); T1W = VSUB(T1U, T1V); T1X = VFMA(LDK(KP831469612), T1R, VMUL(LDK(KP555570233), T1W)); T2q = VFNMS(LDK(KP555570233), T1R, VMUL(LDK(KP831469612), T1W)); T28 = VSUB(T24, T27); T2d = VSUB(T2b, T2c); T2e = VFNMS(LDK(KP555570233), T2d, VMUL(LDK(KP831469612), T28)); T2r = VFMA(LDK(KP555570233), T28, VMUL(LDK(KP831469612), T2d)); } T2f = VSUB(T1X, T2e); T65 = VADD(T2q, T2r); T2s = VSUB(T2q, T2r); T61 = VADD(T1X, T2e); } } } { V Tgx, TgW, TgR, TgZ, TgI, TgY, TgO, TgV; { V Tgp, Tgw, TgP, TgQ; Tgp = VFNMS(LDK(KP382683432), Tgo, VMUL(LDK(KP923879532), Tgl)); Tgw = VFMA(LDK(KP923879532), Tgs, VMUL(LDK(KP382683432), Tgv)); Tgx = VSUB(Tgp, Tgw); TgW = VADD(Tgp, Tgw); TgP = VFMA(LDK(KP382683432), Tgl, VMUL(LDK(KP923879532), Tgo)); TgQ = VFNMS(LDK(KP382683432), Tgs, VMUL(LDK(KP923879532), Tgv)); TgR = VSUB(TgP, TgQ); TgZ = VADD(TgP, TgQ); } { V TgE, TgH, TgM, TgN; TgE = VMUL(LDK(KP707106781), VSUB(TgA, TgD)); TgH = VSUB(TgF, TgG); TgI = VSUB(TgE, TgH); TgY = VADD(TgH, TgE); TgM = VSUB(TgK, TgL); TgN = VMUL(LDK(KP707106781), VADD(TgA, TgD)); TgO = VSUB(TgM, TgN); TgV = VADD(TgM, TgN); } { V TgJ, TgS, Th1, Th2; TgJ = VBYI(VSUB(Tgx, TgI)); TgS = VSUB(TgO, TgR); ST(&(xo[WS(os, 40)]), VADD(TgJ, TgS), ovs, &(xo[0])); ST(&(xo[WS(os, 88)]), VSUB(TgS, TgJ), ovs, &(xo[0])); Th1 = VSUB(TgV, TgW); Th2 = VBYI(VSUB(TgZ, TgY)); ST(&(xo[WS(os, 72)]), VSUB(Th1, Th2), ovs, &(xo[0])); ST(&(xo[WS(os, 56)]), VADD(Th1, Th2), ovs, &(xo[0])); } { V TgT, TgU, TgX, Th0; TgT = VBYI(VADD(TgI, Tgx)); TgU = VADD(TgO, TgR); ST(&(xo[WS(os, 24)]), VADD(TgT, TgU), ovs, &(xo[0])); ST(&(xo[WS(os, 104)]), VSUB(TgU, TgT), ovs, &(xo[0])); TgX = VADD(TgV, TgW); Th0 = VBYI(VADD(TgY, TgZ)); ST(&(xo[WS(os, 120)]), VSUB(TgX, Th0), ovs, &(xo[0])); ST(&(xo[WS(os, 8)]), VADD(TgX, Th0), ovs, &(xo[0])); } } { V Th9, Thh, Thq, Ths, Thc, Thm, Thg, Thl, Thn, Thr; { V Th5, Th8, Tho, Thp; Th5 = VSUB(Th3, Th4); Th8 = VSUB(Th6, Th7); Th9 = VMUL(LDK(KP707106781), VSUB(Th5, Th8)); Thh = VMUL(LDK(KP707106781), VADD(Th5, Th8)); Tho = VADD(Th3, Th4); Thp = VADD(Th6, Th7); Thq = VBYI(VSUB(Tho, Thp)); Ths = VADD(Tho, Thp); } { V Tha, Thb, The, Thf; Tha = VADD(Tgy, Tgz); Thb = VADD(TgB, TgC); Thc = VSUB(Tha, Thb); Thm = VADD(Tha, Thb); The = VADD(TgK, TgL); Thf = VADD(TgF, TgG); Thg = VSUB(The, Thf); Thl = VADD(The, Thf); } Thn = VSUB(Thl, Thm); ST(&(xo[WS(os, 96)]), VSUB(Thn, Thq), ovs, &(xo[0])); ST(&(xo[WS(os, 32)]), VADD(Thn, Thq), ovs, &(xo[0])); Thr = VADD(Thl, Thm); ST(&(xo[WS(os, 64)]), VSUB(Thr, Ths), ovs, &(xo[0])); ST(&(xo[0]), VADD(Thr, Ths), ovs, &(xo[0])); { V Thd, Thi, Thj, Thk; Thd = VBYI(VSUB(Th9, Thc)); Thi = VSUB(Thg, Thh); ST(&(xo[WS(os, 48)]), VADD(Thd, Thi), ovs, &(xo[0])); ST(&(xo[WS(os, 80)]), VSUB(Thi, Thd), ovs, &(xo[0])); Thj = VBYI(VADD(Thc, Th9)); Thk = VADD(Thg, Thh); ST(&(xo[WS(os, 16)]), VADD(Thj, Thk), ovs, &(xo[0])); ST(&(xo[WS(os, 112)]), VSUB(Thk, Thj), ovs, &(xo[0])); } } { V TeT, TfM, TfC, TfK, Tfs, TfN, TfF, TfJ; { V TeD, TeS, Tfw, TfB; TeD = VSUB(Tev, TeC); TeS = VSUB(TeK, TeR); TeT = VSUB(TeD, TeS); TfM = VADD(TeD, TeS); Tfw = VSUB(Tfu, Tfv); TfB = VSUB(Tfx, TfA); TfC = VSUB(Tfw, TfB); TfK = VADD(TfB, Tfw); { V Tfa, TfD, Tfr, TfE; { V Tf4, Tf9, Tfl, Tfq; Tf4 = VSUB(Tf0, Tf3); Tf9 = VSUB(Tf7, Tf8); Tfa = VFMA(LDK(KP831469612), Tf4, VMUL(LDK(KP555570233), Tf9)); TfD = VFNMS(LDK(KP555570233), Tf4, VMUL(LDK(KP831469612), Tf9)); Tfl = VSUB(Tfh, Tfk); Tfq = VSUB(Tfo, Tfp); Tfr = VFNMS(LDK(KP555570233), Tfq, VMUL(LDK(KP831469612), Tfl)); TfE = VFMA(LDK(KP555570233), Tfl, VMUL(LDK(KP831469612), Tfq)); } Tfs = VSUB(Tfa, Tfr); TfN = VADD(TfD, TfE); TfF = VSUB(TfD, TfE); TfJ = VADD(Tfa, Tfr); } } { V Tft, TfG, TfP, TfQ; Tft = VADD(TeT, Tfs); TfG = VBYI(VADD(TfC, TfF)); ST(&(xo[WS(os, 108)]), VSUB(Tft, TfG), ovs, &(xo[0])); ST(&(xo[WS(os, 20)]), VADD(Tft, TfG), ovs, &(xo[0])); TfP = VBYI(VADD(TfK, TfJ)); TfQ = VADD(TfM, TfN); ST(&(xo[WS(os, 12)]), VADD(TfP, TfQ), ovs, &(xo[0])); ST(&(xo[WS(os, 116)]), VSUB(TfQ, TfP), ovs, &(xo[0])); } { V TfH, TfI, TfL, TfO; TfH = VSUB(TeT, Tfs); TfI = VBYI(VSUB(TfF, TfC)); ST(&(xo[WS(os, 84)]), VSUB(TfH, TfI), ovs, &(xo[0])); ST(&(xo[WS(os, 44)]), VADD(TfH, TfI), ovs, &(xo[0])); TfL = VBYI(VSUB(TfJ, TfK)); TfO = VSUB(TfM, TfN); ST(&(xo[WS(os, 52)]), VADD(TfL, TfO), ovs, &(xo[0])); ST(&(xo[WS(os, 76)]), VSUB(TfO, TfL), ovs, &(xo[0])); } } { V TfT, Tge, Tg4, Tgc, Tg0, Tgf, Tg7, Tgb; { V TfR, TfS, Tg2, Tg3; TfR = VADD(Tev, TeC); TfS = VADD(Tfu, Tfv); TfT = VSUB(TfR, TfS); Tge = VADD(TfR, TfS); Tg2 = VADD(TeK, TeR); Tg3 = VADD(TfA, Tfx); Tg4 = VSUB(Tg2, Tg3); Tgc = VADD(Tg3, Tg2); { V TfW, Tg5, TfZ, Tg6; { V TfU, TfV, TfX, TfY; TfU = VADD(Tf3, Tf0); TfV = VADD(Tf7, Tf8); TfW = VFMA(LDK(KP980785280), TfU, VMUL(LDK(KP195090322), TfV)); Tg5 = VFNMS(LDK(KP195090322), TfU, VMUL(LDK(KP980785280), TfV)); TfX = VADD(Tfk, Tfh); TfY = VADD(Tfo, Tfp); TfZ = VFNMS(LDK(KP195090322), TfY, VMUL(LDK(KP980785280), TfX)); Tg6 = VFMA(LDK(KP195090322), TfX, VMUL(LDK(KP980785280), TfY)); } Tg0 = VSUB(TfW, TfZ); Tgf = VADD(Tg5, Tg6); Tg7 = VSUB(Tg5, Tg6); Tgb = VADD(TfW, TfZ); } } { V Tg1, Tg8, Tgh, Tgi; Tg1 = VADD(TfT, Tg0); Tg8 = VBYI(VADD(Tg4, Tg7)); ST(&(xo[WS(os, 100)]), VSUB(Tg1, Tg8), ovs, &(xo[0])); ST(&(xo[WS(os, 28)]), VADD(Tg1, Tg8), ovs, &(xo[0])); Tgh = VBYI(VADD(Tgc, Tgb)); Tgi = VADD(Tge, Tgf); ST(&(xo[WS(os, 4)]), VADD(Tgh, Tgi), ovs, &(xo[0])); ST(&(xo[WS(os, 124)]), VSUB(Tgi, Tgh), ovs, &(xo[0])); } { V Tg9, Tga, Tgd, Tgg; Tg9 = VSUB(TfT, Tg0); Tga = VBYI(VSUB(Tg7, Tg4)); ST(&(xo[WS(os, 92)]), VSUB(Tg9, Tga), ovs, &(xo[0])); ST(&(xo[WS(os, 36)]), VADD(Tg9, Tga), ovs, &(xo[0])); Tgd = VBYI(VSUB(Tgb, Tgc)); Tgg = VSUB(Tge, Tgf); ST(&(xo[WS(os, 60)]), VADD(Tgd, Tgg), ovs, &(xo[0])); ST(&(xo[WS(os, 68)]), VSUB(Tgg, Tgd), ovs, &(xo[0])); } } { V Tb7, Td8, TcI, Td0, Tcy, Tda, TcG, TcP, Tck, TcJ, TcB, TcF, TcW, Tdb, Td3; V Td7; { V Tax, TcZ, Tb6, TcY, TaO, Tb5; Tax = VSUB(Tah, Taw); TcZ = VADD(Tcw, Tcr); TaO = VFMA(LDK(KP831469612), TaI, VMUL(LDK(KP555570233), TaN)); Tb5 = VFNMS(LDK(KP555570233), Tb4, VMUL(LDK(KP831469612), TaZ)); Tb6 = VSUB(TaO, Tb5); TcY = VADD(TaO, Tb5); Tb7 = VSUB(Tax, Tb6); Td8 = VADD(TcZ, TcY); TcI = VADD(Tax, Tb6); Td0 = VSUB(TcY, TcZ); } { V Tcx, TcN, Tco, TcO, Tcm, Tcn; Tcx = VSUB(Tcr, Tcw); TcN = VADD(Tah, Taw); Tcm = VFNMS(LDK(KP555570233), TaI, VMUL(LDK(KP831469612), TaN)); Tcn = VFMA(LDK(KP555570233), TaZ, VMUL(LDK(KP831469612), Tb4)); Tco = VSUB(Tcm, Tcn); TcO = VADD(Tcm, Tcn); Tcy = VSUB(Tco, Tcx); Tda = VADD(TcN, TcO); TcG = VADD(Tcx, Tco); TcP = VSUB(TcN, TcO); } { V TbI, Tcz, Tcj, TcA; { V Tby, TbH, Tc9, Tci; Tby = VSUB(Tbm, Tbx); TbH = VSUB(TbD, TbG); TbI = VFMA(LDK(KP881921264), Tby, VMUL(LDK(KP471396736), TbH)); Tcz = VFNMS(LDK(KP471396736), Tby, VMUL(LDK(KP881921264), TbH)); Tc9 = VSUB(TbX, Tc8); Tci = VSUB(Tce, Tch); Tcj = VFNMS(LDK(KP471396736), Tci, VMUL(LDK(KP881921264), Tc9)); TcA = VFMA(LDK(KP471396736), Tc9, VMUL(LDK(KP881921264), Tci)); } Tck = VSUB(TbI, Tcj); TcJ = VADD(Tcz, TcA); TcB = VSUB(Tcz, TcA); TcF = VADD(TbI, Tcj); } { V TcS, Td1, TcV, Td2; { V TcQ, TcR, TcT, TcU; TcQ = VADD(Tbx, Tbm); TcR = VADD(TbD, TbG); TcS = VFMA(LDK(KP956940335), TcQ, VMUL(LDK(KP290284677), TcR)); Td1 = VFNMS(LDK(KP290284677), TcQ, VMUL(LDK(KP956940335), TcR)); TcT = VADD(Tc8, TbX); TcU = VADD(Tce, Tch); TcV = VFNMS(LDK(KP290284677), TcU, VMUL(LDK(KP956940335), TcT)); Td2 = VFMA(LDK(KP290284677), TcT, VMUL(LDK(KP956940335), TcU)); } TcW = VSUB(TcS, TcV); Tdb = VADD(Td1, Td2); Td3 = VSUB(Td1, Td2); Td7 = VADD(TcS, TcV); } { V Tcl, TcC, Td9, Tdc; Tcl = VADD(Tb7, Tck); TcC = VBYI(VADD(Tcy, TcB)); ST(&(xo[WS(os, 106)]), VSUB(Tcl, TcC), ovs, &(xo[0])); ST(&(xo[WS(os, 22)]), VADD(Tcl, TcC), ovs, &(xo[0])); Td9 = VBYI(VSUB(Td7, Td8)); Tdc = VSUB(Tda, Tdb); ST(&(xo[WS(os, 58)]), VADD(Td9, Tdc), ovs, &(xo[0])); ST(&(xo[WS(os, 70)]), VSUB(Tdc, Td9), ovs, &(xo[0])); } { V Tdd, Tde, TcD, TcE; Tdd = VBYI(VADD(Td8, Td7)); Tde = VADD(Tda, Tdb); ST(&(xo[WS(os, 6)]), VADD(Tdd, Tde), ovs, &(xo[0])); ST(&(xo[WS(os, 122)]), VSUB(Tde, Tdd), ovs, &(xo[0])); TcD = VSUB(Tb7, Tck); TcE = VBYI(VSUB(TcB, Tcy)); ST(&(xo[WS(os, 86)]), VSUB(TcD, TcE), ovs, &(xo[0])); ST(&(xo[WS(os, 42)]), VADD(TcD, TcE), ovs, &(xo[0])); } { V TcH, TcK, TcX, Td4; TcH = VBYI(VSUB(TcF, TcG)); TcK = VSUB(TcI, TcJ); ST(&(xo[WS(os, 54)]), VADD(TcH, TcK), ovs, &(xo[0])); ST(&(xo[WS(os, 74)]), VSUB(TcK, TcH), ovs, &(xo[0])); TcX = VADD(TcP, TcW); Td4 = VBYI(VADD(Td0, Td3)); ST(&(xo[WS(os, 102)]), VSUB(TcX, Td4), ovs, &(xo[0])); ST(&(xo[WS(os, 26)]), VADD(TcX, Td4), ovs, &(xo[0])); } { V Td5, Td6, TcL, TcM; Td5 = VSUB(TcP, TcW); Td6 = VBYI(VSUB(Td3, Td0)); ST(&(xo[WS(os, 90)]), VSUB(Td5, Td6), ovs, &(xo[0])); ST(&(xo[WS(os, 38)]), VADD(Td5, Td6), ovs, &(xo[0])); TcL = VBYI(VADD(TcG, TcF)); TcM = VADD(TcI, TcJ); ST(&(xo[WS(os, 10)]), VADD(TcL, TcM), ovs, &(xo[0])); ST(&(xo[WS(os, 118)]), VSUB(TcM, TcL), ovs, &(xo[0])); } } { V TdE, Tel, TdW, Tee, TdM, Teo, TdT, Tea, Tdt, TdX, TdP, TdU, Te7, Tep, Teh; V Tem; { V TdD, Tec, TdA, Ted, Tdw, Tdz; TdD = VADD(TdB, TdC); Tec = VSUB(TdG, TdH); Tdw = VFMA(LDK(KP980785280), Tdu, VMUL(LDK(KP195090322), Tdv)); Tdz = VFNMS(LDK(KP195090322), Tdy, VMUL(LDK(KP980785280), Tdx)); TdA = VADD(Tdw, Tdz); Ted = VSUB(Tdw, Tdz); TdE = VSUB(TdA, TdD); Tel = VADD(Tec, Ted); TdW = VADD(TdD, TdA); Tee = VSUB(Tec, Ted); } { V TdI, Te9, TdL, Te8, TdJ, TdK; TdI = VADD(TdG, TdH); Te9 = VSUB(TdC, TdB); TdJ = VFNMS(LDK(KP195090322), Tdu, VMUL(LDK(KP980785280), Tdv)); TdK = VFMA(LDK(KP195090322), Tdx, VMUL(LDK(KP980785280), Tdy)); TdL = VADD(TdJ, TdK); Te8 = VSUB(TdJ, TdK); TdM = VSUB(TdI, TdL); Teo = VADD(Te9, Te8); TdT = VADD(TdI, TdL); Tea = VSUB(Te8, Te9); } { V Tdl, TdN, Tds, TdO; { V Tdh, Tdk, Tdo, Tdr; Tdh = VADD(Tdf, Tdg); Tdk = VADD(Tdi, Tdj); Tdl = VFNMS(LDK(KP098017140), Tdk, VMUL(LDK(KP995184726), Tdh)); TdN = VFMA(LDK(KP098017140), Tdh, VMUL(LDK(KP995184726), Tdk)); Tdo = VADD(Tdm, Tdn); Tdr = VADD(Tdp, Tdq); Tds = VFMA(LDK(KP995184726), Tdo, VMUL(LDK(KP098017140), Tdr)); TdO = VFNMS(LDK(KP098017140), Tdo, VMUL(LDK(KP995184726), Tdr)); } Tdt = VSUB(Tdl, Tds); TdX = VADD(TdN, TdO); TdP = VSUB(TdN, TdO); TdU = VADD(Tdl, Tds); } { V Te3, Tef, Te6, Teg; { V Te1, Te2, Te4, Te5; Te1 = VSUB(Tdf, Tdg); Te2 = VSUB(Tdj, Tdi); Te3 = VFNMS(LDK(KP634393284), Te2, VMUL(LDK(KP773010453), Te1)); Tef = VFMA(LDK(KP634393284), Te1, VMUL(LDK(KP773010453), Te2)); Te4 = VSUB(Tdm, Tdn); Te5 = VSUB(Tdq, Tdp); Te6 = VFMA(LDK(KP773010453), Te4, VMUL(LDK(KP634393284), Te5)); Teg = VFNMS(LDK(KP634393284), Te4, VMUL(LDK(KP773010453), Te5)); } Te7 = VSUB(Te3, Te6); Tep = VADD(Tef, Teg); Teh = VSUB(Tef, Teg); Tem = VADD(Te3, Te6); } { V TdF, TdQ, Ten, Teq; TdF = VBYI(VSUB(Tdt, TdE)); TdQ = VSUB(TdM, TdP); ST(&(xo[WS(os, 34)]), VADD(TdF, TdQ), ovs, &(xo[0])); ST(&(xo[WS(os, 94)]), VSUB(TdQ, TdF), ovs, &(xo[0])); Ten = VADD(Tel, Tem); Teq = VBYI(VADD(Teo, Tep)); ST(&(xo[WS(os, 114)]), VSUB(Ten, Teq), ovs, &(xo[0])); ST(&(xo[WS(os, 14)]), VADD(Ten, Teq), ovs, &(xo[0])); } { V Ter, Tes, TdR, TdS; Ter = VSUB(Tel, Tem); Tes = VBYI(VSUB(Tep, Teo)); ST(&(xo[WS(os, 78)]), VSUB(Ter, Tes), ovs, &(xo[0])); ST(&(xo[WS(os, 50)]), VADD(Ter, Tes), ovs, &(xo[0])); TdR = VBYI(VADD(TdE, Tdt)); TdS = VADD(TdM, TdP); ST(&(xo[WS(os, 30)]), VADD(TdR, TdS), ovs, &(xo[0])); ST(&(xo[WS(os, 98)]), VSUB(TdS, TdR), ovs, &(xo[0])); } { V TdV, TdY, Teb, Tei; TdV = VADD(TdT, TdU); TdY = VBYI(VADD(TdW, TdX)); ST(&(xo[WS(os, 126)]), VSUB(TdV, TdY), ovs, &(xo[0])); ST(&(xo[WS(os, 2)]), VADD(TdV, TdY), ovs, &(xo[0])); Teb = VBYI(VSUB(Te7, Tea)); Tei = VSUB(Tee, Teh); ST(&(xo[WS(os, 46)]), VADD(Teb, Tei), ovs, &(xo[0])); ST(&(xo[WS(os, 82)]), VSUB(Tei, Teb), ovs, &(xo[0])); } { V Tej, Tek, TdZ, Te0; Tej = VBYI(VADD(Tea, Te7)); Tek = VADD(Tee, Teh); ST(&(xo[WS(os, 18)]), VADD(Tej, Tek), ovs, &(xo[0])); ST(&(xo[WS(os, 110)]), VSUB(Tek, Tej), ovs, &(xo[0])); TdZ = VSUB(TdT, TdU); Te0 = VBYI(VSUB(TdX, TdW)); ST(&(xo[WS(os, 66)]), VSUB(TdZ, Te0), ovs, &(xo[0])); ST(&(xo[WS(os, 62)]), VADD(TdZ, Te0), ovs, &(xo[0])); } } { V T7z, T8n, T8f, T8k, T8x, T8P, T8H, T8M, T80, T8L, T8O, T8c, T8j, T8A, T8E; V T8m; { V T7f, T8d, T7y, T8e; { V T77, T7e, T7q, T7x; T77 = VADD(T6Z, T76); T7e = VADD(T7a, T7d); T7f = VFNMS(LDK(KP336889853), T7e, VMUL(LDK(KP941544065), T77)); T8d = VFMA(LDK(KP336889853), T77, VMUL(LDK(KP941544065), T7e)); T7q = VADD(T7i, T7p); T7x = VADD(T7t, T7w); T7y = VFMA(LDK(KP941544065), T7q, VMUL(LDK(KP336889853), T7x)); T8e = VFNMS(LDK(KP336889853), T7q, VMUL(LDK(KP941544065), T7x)); } T7z = VSUB(T7f, T7y); T8n = VADD(T8d, T8e); T8f = VSUB(T8d, T8e); T8k = VADD(T7f, T7y); } { V T8t, T8F, T8w, T8G; { V T8r, T8s, T8u, T8v; T8r = VSUB(T6Z, T76); T8s = VSUB(T7d, T7a); T8t = VFNMS(LDK(KP427555093), T8s, VMUL(LDK(KP903989293), T8r)); T8F = VFMA(LDK(KP427555093), T8r, VMUL(LDK(KP903989293), T8s)); T8u = VSUB(T7i, T7p); T8v = VSUB(T7w, T7t); T8w = VFMA(LDK(KP903989293), T8u, VMUL(LDK(KP427555093), T8v)); T8G = VFNMS(LDK(KP427555093), T8u, VMUL(LDK(KP903989293), T8v)); } T8x = VSUB(T8t, T8w); T8P = VADD(T8F, T8G); T8H = VSUB(T8F, T8G); T8M = VADD(T8t, T8w); } { V T7Z, T8z, T88, T8C, T7O, T8D, T8b, T8y, T7Y, T87; T7Y = VSUB(T7U, T7X); T7Z = VADD(T7R, T7Y); T8z = VSUB(T7Y, T7R); T87 = VSUB(T85, T86); T88 = VADD(T84, T87); T8C = VSUB(T84, T87); { V T7G, T7N, T89, T8a; T7G = VFMA(LDK(KP634393284), T7C, VMUL(LDK(KP773010453), T7F)); T7N = VFNMS(LDK(KP634393284), T7M, VMUL(LDK(KP773010453), T7J)); T7O = VADD(T7G, T7N); T8D = VSUB(T7G, T7N); T89 = VFNMS(LDK(KP634393284), T7F, VMUL(LDK(KP773010453), T7C)); T8a = VFMA(LDK(KP773010453), T7M, VMUL(LDK(KP634393284), T7J)); T8b = VADD(T89, T8a); T8y = VSUB(T89, T8a); } T80 = VSUB(T7O, T7Z); T8L = VADD(T8C, T8D); T8O = VADD(T8z, T8y); T8c = VSUB(T88, T8b); T8j = VADD(T88, T8b); T8A = VSUB(T8y, T8z); T8E = VSUB(T8C, T8D); T8m = VADD(T7Z, T7O); } { V T81, T8g, T8N, T8Q; T81 = VBYI(VSUB(T7z, T80)); T8g = VSUB(T8c, T8f); ST(&(xo[WS(os, 39)]), VADD(T81, T8g), ovs, &(xo[WS(os, 1)])); ST(&(xo[WS(os, 89)]), VSUB(T8g, T81), ovs, &(xo[WS(os, 1)])); T8N = VADD(T8L, T8M); T8Q = VBYI(VADD(T8O, T8P)); ST(&(xo[WS(os, 119)]), VSUB(T8N, T8Q), ovs, &(xo[WS(os, 1)])); ST(&(xo[WS(os, 9)]), VADD(T8N, T8Q), ovs, &(xo[WS(os, 1)])); } { V T8R, T8S, T8h, T8i; T8R = VSUB(T8L, T8M); T8S = VBYI(VSUB(T8P, T8O)); ST(&(xo[WS(os, 73)]), VSUB(T8R, T8S), ovs, &(xo[WS(os, 1)])); ST(&(xo[WS(os, 55)]), VADD(T8R, T8S), ovs, &(xo[WS(os, 1)])); T8h = VBYI(VADD(T80, T7z)); T8i = VADD(T8c, T8f); ST(&(xo[WS(os, 25)]), VADD(T8h, T8i), ovs, &(xo[WS(os, 1)])); ST(&(xo[WS(os, 103)]), VSUB(T8i, T8h), ovs, &(xo[WS(os, 1)])); } { V T8l, T8o, T8B, T8I; T8l = VADD(T8j, T8k); T8o = VBYI(VADD(T8m, T8n)); ST(&(xo[WS(os, 121)]), VSUB(T8l, T8o), ovs, &(xo[WS(os, 1)])); ST(&(xo[WS(os, 7)]), VADD(T8l, T8o), ovs, &(xo[WS(os, 1)])); T8B = VBYI(VSUB(T8x, T8A)); T8I = VSUB(T8E, T8H); ST(&(xo[WS(os, 41)]), VADD(T8B, T8I), ovs, &(xo[WS(os, 1)])); ST(&(xo[WS(os, 87)]), VSUB(T8I, T8B), ovs, &(xo[WS(os, 1)])); } { V T8J, T8K, T8p, T8q; T8J = VBYI(VADD(T8A, T8x)); T8K = VADD(T8E, T8H); ST(&(xo[WS(os, 23)]), VADD(T8J, T8K), ovs, &(xo[WS(os, 1)])); ST(&(xo[WS(os, 105)]), VSUB(T8K, T8J), ovs, &(xo[WS(os, 1)])); T8p = VSUB(T8j, T8k); T8q = VBYI(VSUB(T8n, T8m)); ST(&(xo[WS(os, 71)]), VSUB(T8p, T8q), ovs, &(xo[WS(os, 1)])); ST(&(xo[WS(os, 57)]), VADD(T8p, T8q), ovs, &(xo[WS(os, 1)])); } } { V T2v, T5d, T55, T5a, T5n, T5F, T5x, T5C, T4K, T5B, T5E, T52, T59, T5q, T5u; V T5c; { V T1f, T53, T2u, T54; { V T11, T1e, T2g, T2t; T11 = VADD(Tr, T10); T1e = VADD(T1a, T1d); T1f = VFNMS(LDK(KP242980179), T1e, VMUL(LDK(KP970031253), T11)); T53 = VFMA(LDK(KP242980179), T11, VMUL(LDK(KP970031253), T1e)); T2g = VADD(T1G, T2f); T2t = VADD(T2p, T2s); T2u = VFMA(LDK(KP970031253), T2g, VMUL(LDK(KP242980179), T2t)); T54 = VFNMS(LDK(KP242980179), T2g, VMUL(LDK(KP970031253), T2t)); } T2v = VSUB(T1f, T2u); T5d = VADD(T53, T54); T55 = VSUB(T53, T54); T5a = VADD(T1f, T2u); } { V T5j, T5v, T5m, T5w; { V T5h, T5i, T5k, T5l; T5h = VSUB(Tr, T10); T5i = VSUB(T1d, T1a); T5j = VFNMS(LDK(KP514102744), T5i, VMUL(LDK(KP857728610), T5h)); T5v = VFMA(LDK(KP514102744), T5h, VMUL(LDK(KP857728610), T5i)); T5k = VSUB(T1G, T2f); T5l = VSUB(T2s, T2p); T5m = VFMA(LDK(KP857728610), T5k, VMUL(LDK(KP514102744), T5l)); T5w = VFNMS(LDK(KP514102744), T5k, VMUL(LDK(KP857728610), T5l)); } T5n = VSUB(T5j, T5m); T5F = VADD(T5v, T5w); T5x = VSUB(T5v, T5w); T5C = VADD(T5j, T5m); } { V T4J, T5p, T4Y, T5s, T3I, T5t, T51, T5o, T4I, T4X; T4I = VSUB(T4q, T4H); T4J = VADD(T49, T4I); T5p = VSUB(T4I, T49); T4X = VSUB(T4V, T4W); T4Y = VADD(T4U, T4X); T5s = VSUB(T4U, T4X); { V T36, T3H, T4Z, T50; T36 = VFMA(LDK(KP881921264), T2W, VMUL(LDK(KP471396736), T35)); T3H = VFNMS(LDK(KP471396736), T3G, VMUL(LDK(KP881921264), T3x)); T3I = VADD(T36, T3H); T5t = VSUB(T36, T3H); T4Z = VFNMS(LDK(KP471396736), T2W, VMUL(LDK(KP881921264), T35)); T50 = VFMA(LDK(KP471396736), T3x, VMUL(LDK(KP881921264), T3G)); T51 = VADD(T4Z, T50); T5o = VSUB(T4Z, T50); } T4K = VSUB(T3I, T4J); T5B = VADD(T5s, T5t); T5E = VADD(T5p, T5o); T52 = VSUB(T4Y, T51); T59 = VADD(T4Y, T51); T5q = VSUB(T5o, T5p); T5u = VSUB(T5s, T5t); T5c = VADD(T4J, T3I); } { V T4L, T56, T5D, T5G; T4L = VBYI(VSUB(T2v, T4K)); T56 = VSUB(T52, T55); ST(&(xo[WS(os, 37)]), VADD(T4L, T56), ovs, &(xo[WS(os, 1)])); ST(&(xo[WS(os, 91)]), VSUB(T56, T4L), ovs, &(xo[WS(os, 1)])); T5D = VADD(T5B, T5C); T5G = VBYI(VADD(T5E, T5F)); ST(&(xo[WS(os, 117)]), VSUB(T5D, T5G), ovs, &(xo[WS(os, 1)])); ST(&(xo[WS(os, 11)]), VADD(T5D, T5G), ovs, &(xo[WS(os, 1)])); } { V T5H, T5I, T57, T58; T5H = VSUB(T5B, T5C); T5I = VBYI(VSUB(T5F, T5E)); ST(&(xo[WS(os, 75)]), VSUB(T5H, T5I), ovs, &(xo[WS(os, 1)])); ST(&(xo[WS(os, 53)]), VADD(T5H, T5I), ovs, &(xo[WS(os, 1)])); T57 = VBYI(VADD(T4K, T2v)); T58 = VADD(T52, T55); ST(&(xo[WS(os, 27)]), VADD(T57, T58), ovs, &(xo[WS(os, 1)])); ST(&(xo[WS(os, 101)]), VSUB(T58, T57), ovs, &(xo[WS(os, 1)])); } { V T5b, T5e, T5r, T5y; T5b = VADD(T59, T5a); T5e = VBYI(VADD(T5c, T5d)); ST(&(xo[WS(os, 123)]), VSUB(T5b, T5e), ovs, &(xo[WS(os, 1)])); ST(&(xo[WS(os, 5)]), VADD(T5b, T5e), ovs, &(xo[WS(os, 1)])); T5r = VBYI(VSUB(T5n, T5q)); T5y = VSUB(T5u, T5x); ST(&(xo[WS(os, 43)]), VADD(T5r, T5y), ovs, &(xo[WS(os, 1)])); ST(&(xo[WS(os, 85)]), VSUB(T5y, T5r), ovs, &(xo[WS(os, 1)])); } { V T5z, T5A, T5f, T5g; T5z = VBYI(VADD(T5q, T5n)); T5A = VADD(T5u, T5x); ST(&(xo[WS(os, 21)]), VADD(T5z, T5A), ovs, &(xo[WS(os, 1)])); ST(&(xo[WS(os, 107)]), VSUB(T5A, T5z), ovs, &(xo[WS(os, 1)])); T5f = VSUB(T59, T5a); T5g = VBYI(VSUB(T5d, T5c)); ST(&(xo[WS(os, 69)]), VSUB(T5f, T5g), ovs, &(xo[WS(os, 1)])); ST(&(xo[WS(os, 59)]), VADD(T5f, T5g), ovs, &(xo[WS(os, 1)])); } } { V T9i, T9B, T9t, T9x, T9O, Ta3, T9V, T9Z, T93, Ta0, Ta2, T9q, T9y, T9H, T9S; V T9A; { V T9a, T9r, T9h, T9s; { V T96, T99, T9d, T9g; T96 = VSUB(T94, T95); T99 = VSUB(T97, T98); T9a = VFMA(LDK(KP740951125), T96, VMUL(LDK(KP671558954), T99)); T9r = VFNMS(LDK(KP671558954), T96, VMUL(LDK(KP740951125), T99)); T9d = VSUB(T9b, T9c); T9g = VSUB(T9e, T9f); T9h = VFNMS(LDK(KP671558954), T9g, VMUL(LDK(KP740951125), T9d)); T9s = VFMA(LDK(KP671558954), T9d, VMUL(LDK(KP740951125), T9g)); } T9i = VSUB(T9a, T9h); T9B = VADD(T9r, T9s); T9t = VSUB(T9r, T9s); T9x = VADD(T9a, T9h); } { V T9K, T9T, T9N, T9U; { V T9I, T9J, T9L, T9M; T9I = VADD(T95, T94); T9J = VADD(T97, T98); T9K = VFMA(LDK(KP998795456), T9I, VMUL(LDK(KP049067674), T9J)); T9T = VFNMS(LDK(KP049067674), T9I, VMUL(LDK(KP998795456), T9J)); T9L = VADD(T9c, T9b); T9M = VADD(T9e, T9f); T9N = VFNMS(LDK(KP049067674), T9M, VMUL(LDK(KP998795456), T9L)); T9U = VFMA(LDK(KP049067674), T9L, VMUL(LDK(KP998795456), T9M)); } T9O = VSUB(T9K, T9N); Ta3 = VADD(T9T, T9U); T9V = VSUB(T9T, T9U); T9Z = VADD(T9K, T9N); } { V T8V, T9F, T9p, T9R, T92, T9Q, T9m, T9G, T8U, T9n; T8U = VADD(T7U, T7X); T8V = VSUB(T8T, T8U); T9F = VADD(T8T, T8U); T9n = VADD(T85, T86); T9p = VSUB(T9n, T9o); T9R = VADD(T9o, T9n); { V T8Y, T91, T9k, T9l; T8Y = VFMA(LDK(KP098017140), T8W, VMUL(LDK(KP995184726), T8X)); T91 = VFNMS(LDK(KP098017140), T90, VMUL(LDK(KP995184726), T8Z)); T92 = VSUB(T8Y, T91); T9Q = VADD(T8Y, T91); T9k = VFNMS(LDK(KP098017140), T8X, VMUL(LDK(KP995184726), T8W)); T9l = VFMA(LDK(KP995184726), T90, VMUL(LDK(KP098017140), T8Z)); T9m = VSUB(T9k, T9l); T9G = VADD(T9k, T9l); } T93 = VSUB(T8V, T92); Ta0 = VADD(T9R, T9Q); Ta2 = VADD(T9F, T9G); T9q = VSUB(T9m, T9p); T9y = VADD(T9p, T9m); T9H = VSUB(T9F, T9G); T9S = VSUB(T9Q, T9R); T9A = VADD(T8V, T92); } { V T9j, T9u, Ta1, Ta4; T9j = VADD(T93, T9i); T9u = VBYI(VADD(T9q, T9t)); ST(&(xo[WS(os, 111)]), VSUB(T9j, T9u), ovs, &(xo[WS(os, 1)])); ST(&(xo[WS(os, 17)]), VADD(T9j, T9u), ovs, &(xo[WS(os, 1)])); Ta1 = VBYI(VSUB(T9Z, Ta0)); Ta4 = VSUB(Ta2, Ta3); ST(&(xo[WS(os, 63)]), VADD(Ta1, Ta4), ovs, &(xo[WS(os, 1)])); ST(&(xo[WS(os, 65)]), VSUB(Ta4, Ta1), ovs, &(xo[WS(os, 1)])); } { V Ta5, Ta6, T9v, T9w; Ta5 = VBYI(VADD(Ta0, T9Z)); Ta6 = VADD(Ta2, Ta3); ST(&(xo[WS(os, 1)]), VADD(Ta5, Ta6), ovs, &(xo[WS(os, 1)])); ST(&(xo[WS(os, 127)]), VSUB(Ta6, Ta5), ovs, &(xo[WS(os, 1)])); T9v = VSUB(T93, T9i); T9w = VBYI(VSUB(T9t, T9q)); ST(&(xo[WS(os, 81)]), VSUB(T9v, T9w), ovs, &(xo[WS(os, 1)])); ST(&(xo[WS(os, 47)]), VADD(T9v, T9w), ovs, &(xo[WS(os, 1)])); } { V T9z, T9C, T9P, T9W; T9z = VBYI(VSUB(T9x, T9y)); T9C = VSUB(T9A, T9B); ST(&(xo[WS(os, 49)]), VADD(T9z, T9C), ovs, &(xo[WS(os, 1)])); ST(&(xo[WS(os, 79)]), VSUB(T9C, T9z), ovs, &(xo[WS(os, 1)])); T9P = VADD(T9H, T9O); T9W = VBYI(VADD(T9S, T9V)); ST(&(xo[WS(os, 97)]), VSUB(T9P, T9W), ovs, &(xo[WS(os, 1)])); ST(&(xo[WS(os, 31)]), VADD(T9P, T9W), ovs, &(xo[WS(os, 1)])); } { V T9X, T9Y, T9D, T9E; T9X = VSUB(T9H, T9O); T9Y = VBYI(VSUB(T9V, T9S)); ST(&(xo[WS(os, 95)]), VSUB(T9X, T9Y), ovs, &(xo[WS(os, 1)])); ST(&(xo[WS(os, 33)]), VADD(T9X, T9Y), ovs, &(xo[WS(os, 1)])); T9D = VBYI(VADD(T9y, T9x)); T9E = VADD(T9A, T9B); ST(&(xo[WS(os, 15)]), VADD(T9D, T9E), ovs, &(xo[WS(os, 1)])); ST(&(xo[WS(os, 113)]), VSUB(T9E, T9D), ovs, &(xo[WS(os, 1)])); } } { V T68, T6r, T6j, T6n, T6E, T6T, T6L, T6P, T5T, T6Q, T6S, T6g, T6o, T6x, T6I; V T6q; { V T60, T6h, T67, T6i; { V T5W, T5Z, T63, T66; T5W = VSUB(T5U, T5V); T5Z = VSUB(T5X, T5Y); T60 = VFMA(LDK(KP803207531), T5W, VMUL(LDK(KP595699304), T5Z)); T6h = VFNMS(LDK(KP595699304), T5W, VMUL(LDK(KP803207531), T5Z)); T63 = VSUB(T61, T62); T66 = VSUB(T64, T65); T67 = VFNMS(LDK(KP595699304), T66, VMUL(LDK(KP803207531), T63)); T6i = VFMA(LDK(KP595699304), T63, VMUL(LDK(KP803207531), T66)); } T68 = VSUB(T60, T67); T6r = VADD(T6h, T6i); T6j = VSUB(T6h, T6i); T6n = VADD(T60, T67); } { V T6A, T6J, T6D, T6K; { V T6y, T6z, T6B, T6C; T6y = VADD(T5V, T5U); T6z = VADD(T5X, T5Y); T6A = VFMA(LDK(KP989176509), T6y, VMUL(LDK(KP146730474), T6z)); T6J = VFNMS(LDK(KP146730474), T6y, VMUL(LDK(KP989176509), T6z)); T6B = VADD(T62, T61); T6C = VADD(T64, T65); T6D = VFNMS(LDK(KP146730474), T6C, VMUL(LDK(KP989176509), T6B)); T6K = VFMA(LDK(KP146730474), T6B, VMUL(LDK(KP989176509), T6C)); } T6E = VSUB(T6A, T6D); T6T = VADD(T6J, T6K); T6L = VSUB(T6J, T6K); T6P = VADD(T6A, T6D); } { V T5L, T6v, T6f, T6H, T5S, T6G, T6c, T6w, T5K, T6d; T5K = VADD(T4q, T4H); T5L = VSUB(T5J, T5K); T6v = VADD(T5J, T5K); T6d = VADD(T4V, T4W); T6f = VSUB(T6d, T6e); T6H = VADD(T6e, T6d); { V T5O, T5R, T6a, T6b; T5O = VFMA(LDK(KP956940335), T5M, VMUL(LDK(KP290284677), T5N)); T5R = VFNMS(LDK(KP290284677), T5Q, VMUL(LDK(KP956940335), T5P)); T5S = VSUB(T5O, T5R); T6G = VADD(T5O, T5R); T6a = VFNMS(LDK(KP290284677), T5M, VMUL(LDK(KP956940335), T5N)); T6b = VFMA(LDK(KP290284677), T5P, VMUL(LDK(KP956940335), T5Q)); T6c = VSUB(T6a, T6b); T6w = VADD(T6a, T6b); } T5T = VSUB(T5L, T5S); T6Q = VADD(T6H, T6G); T6S = VADD(T6v, T6w); T6g = VSUB(T6c, T6f); T6o = VADD(T6f, T6c); T6x = VSUB(T6v, T6w); T6I = VSUB(T6G, T6H); T6q = VADD(T5L, T5S); } { V T69, T6k, T6R, T6U; T69 = VADD(T5T, T68); T6k = VBYI(VADD(T6g, T6j)); ST(&(xo[WS(os, 109)]), VSUB(T69, T6k), ovs, &(xo[WS(os, 1)])); ST(&(xo[WS(os, 19)]), VADD(T69, T6k), ovs, &(xo[WS(os, 1)])); T6R = VBYI(VSUB(T6P, T6Q)); T6U = VSUB(T6S, T6T); ST(&(xo[WS(os, 61)]), VADD(T6R, T6U), ovs, &(xo[WS(os, 1)])); ST(&(xo[WS(os, 67)]), VSUB(T6U, T6R), ovs, &(xo[WS(os, 1)])); } { V T6V, T6W, T6l, T6m; T6V = VBYI(VADD(T6Q, T6P)); T6W = VADD(T6S, T6T); ST(&(xo[WS(os, 3)]), VADD(T6V, T6W), ovs, &(xo[WS(os, 1)])); ST(&(xo[WS(os, 125)]), VSUB(T6W, T6V), ovs, &(xo[WS(os, 1)])); T6l = VSUB(T5T, T68); T6m = VBYI(VSUB(T6j, T6g)); ST(&(xo[WS(os, 83)]), VSUB(T6l, T6m), ovs, &(xo[WS(os, 1)])); ST(&(xo[WS(os, 45)]), VADD(T6l, T6m), ovs, &(xo[WS(os, 1)])); } { V T6p, T6s, T6F, T6M; T6p = VBYI(VSUB(T6n, T6o)); T6s = VSUB(T6q, T6r); ST(&(xo[WS(os, 51)]), VADD(T6p, T6s), ovs, &(xo[WS(os, 1)])); ST(&(xo[WS(os, 77)]), VSUB(T6s, T6p), ovs, &(xo[WS(os, 1)])); T6F = VADD(T6x, T6E); T6M = VBYI(VADD(T6I, T6L)); ST(&(xo[WS(os, 99)]), VSUB(T6F, T6M), ovs, &(xo[WS(os, 1)])); ST(&(xo[WS(os, 29)]), VADD(T6F, T6M), ovs, &(xo[WS(os, 1)])); } { V T6N, T6O, T6t, T6u; T6N = VSUB(T6x, T6E); T6O = VBYI(VSUB(T6L, T6I)); ST(&(xo[WS(os, 93)]), VSUB(T6N, T6O), ovs, &(xo[WS(os, 1)])); ST(&(xo[WS(os, 35)]), VADD(T6N, T6O), ovs, &(xo[WS(os, 1)])); T6t = VBYI(VADD(T6o, T6n)); T6u = VADD(T6q, T6r); ST(&(xo[WS(os, 13)]), VADD(T6t, T6u), ovs, &(xo[WS(os, 1)])); ST(&(xo[WS(os, 115)]), VSUB(T6u, T6t), ovs, &(xo[WS(os, 1)])); } } } } VLEAVE(); } static const kdft_desc desc = { 128, XSIMD_STRING("n1bv_128"), {938, 186, 144, 0}, &GENUS, 0, 0, 0, 0 }; void XSIMD(codelet_n1bv_128) (planner *p) { X(kdft_register) (p, n1bv_128, &desc); } #endif /* HAVE_FMA */ fftw-3.3.3/dft/simd/common/n1bv_20.c0000644000175000017500000003542612054410372013733 00000000000000/* * Copyright (c) 2003, 2007-11 Matteo Frigo * Copyright (c) 2003, 2007-11 Massachusetts Institute of Technology * * This program is free software; you can redistribute it and/or modify * it under the terms of the GNU General Public License as published by * the Free Software Foundation; either version 2 of the License, or * (at your option) any later version. * * This program is distributed in the hope that it will be useful, * but WITHOUT ANY WARRANTY; without even the implied warranty of * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the * GNU General Public License for more details. * * You should have received a copy of the GNU General Public License * along with this program; if not, write to the Free Software * Foundation, Inc., 51 Franklin Street, Fifth Floor, Boston, MA 02110-1301 USA * */ /* This file was automatically generated --- DO NOT EDIT */ /* Generated on Sun Nov 25 07:37:14 EST 2012 */ #include "codelet-dft.h" #ifdef HAVE_FMA /* Generated by: ../../../genfft/gen_notw_c.native -fma -reorder-insns -schedule-for-pipeline -simd -compact -variables 4 -pipeline-latency 8 -sign 1 -n 20 -name n1bv_20 -include n1b.h */ /* * This function contains 104 FP additions, 50 FP multiplications, * (or, 58 additions, 4 multiplications, 46 fused multiply/add), * 71 stack variables, 4 constants, and 40 memory accesses */ #include "n1b.h" static void n1bv_20(const R *ri, const R *ii, R *ro, R *io, stride is, stride os, INT v, INT ivs, INT ovs) { DVK(KP559016994, +0.559016994374947424102293417182819058860154590); DVK(KP618033988, +0.618033988749894848204586834365638117720309180); DVK(KP951056516, +0.951056516295153572116439333379382143405698634); DVK(KP250000000, +0.250000000000000000000000000000000000000000000); { INT i; const R *xi; R *xo; xi = ii; xo = io; for (i = v; i > 0; i = i - VL, xi = xi + (VL * ivs), xo = xo + (VL * ovs), MAKE_VOLATILE_STRIDE(40, is), MAKE_VOLATILE_STRIDE(40, os)) { V TS, TA, TN, TV, TK, TU, TR, Tl; { V T3, TE, T1r, T13, Ta, TL, Tz, TG, Ts, TF, Th, TM, T1u, T1C, T1n; V T1a, T1m, T1h, T1x, T1D, Tk, Ti; { V T1, T2, TC, TD; T1 = LD(&(xi[0]), ivs, &(xi[0])); T2 = LD(&(xi[WS(is, 10)]), ivs, &(xi[0])); TC = LD(&(xi[WS(is, 5)]), ivs, &(xi[WS(is, 1)])); TD = LD(&(xi[WS(is, 15)]), ivs, &(xi[WS(is, 1)])); { V T14, T6, T1c, Tv, Tm, T1f, Ty, T17, T9, Tn, Tp, T1b, Td, Tq, Te; V Tf, T15, To; { V Tw, Tx, T7, T8, Tb, Tc; { V T4, T5, Tt, Tu, T11, T12; T4 = LD(&(xi[WS(is, 4)]), ivs, &(xi[0])); T5 = LD(&(xi[WS(is, 14)]), ivs, &(xi[0])); Tt = LD(&(xi[WS(is, 13)]), ivs, &(xi[WS(is, 1)])); Tu = LD(&(xi[WS(is, 3)]), ivs, &(xi[WS(is, 1)])); Tw = LD(&(xi[WS(is, 17)]), ivs, &(xi[WS(is, 1)])); T3 = VSUB(T1, T2); T11 = VADD(T1, T2); TE = VSUB(TC, TD); T12 = VADD(TC, TD); T14 = VADD(T4, T5); T6 = VSUB(T4, T5); T1c = VADD(Tt, Tu); Tv = VSUB(Tt, Tu); Tx = LD(&(xi[WS(is, 7)]), ivs, &(xi[WS(is, 1)])); T7 = LD(&(xi[WS(is, 16)]), ivs, &(xi[0])); T8 = LD(&(xi[WS(is, 6)]), ivs, &(xi[0])); T1r = VADD(T11, T12); T13 = VSUB(T11, T12); } Tb = LD(&(xi[WS(is, 8)]), ivs, &(xi[0])); Tc = LD(&(xi[WS(is, 18)]), ivs, &(xi[0])); Tm = LD(&(xi[WS(is, 9)]), ivs, &(xi[WS(is, 1)])); T1f = VADD(Tw, Tx); Ty = VSUB(Tw, Tx); T17 = VADD(T7, T8); T9 = VSUB(T7, T8); Tn = LD(&(xi[WS(is, 19)]), ivs, &(xi[WS(is, 1)])); Tp = LD(&(xi[WS(is, 1)]), ivs, &(xi[WS(is, 1)])); T1b = VADD(Tb, Tc); Td = VSUB(Tb, Tc); Tq = LD(&(xi[WS(is, 11)]), ivs, &(xi[WS(is, 1)])); Te = LD(&(xi[WS(is, 12)]), ivs, &(xi[0])); Tf = LD(&(xi[WS(is, 2)]), ivs, &(xi[0])); } Ta = VADD(T6, T9); TL = VSUB(T6, T9); T15 = VADD(Tm, Tn); To = VSUB(Tm, Tn); Tz = VSUB(Tv, Ty); TG = VADD(Tv, Ty); { V T1d, T1v, T18, Tr, T1e, Tg, T16, T1s; T1d = VSUB(T1b, T1c); T1v = VADD(T1b, T1c); T18 = VADD(Tp, Tq); Tr = VSUB(Tp, Tq); T1e = VADD(Te, Tf); Tg = VSUB(Te, Tf); T16 = VSUB(T14, T15); T1s = VADD(T14, T15); { V T1t, T19, T1w, T1g; T1t = VADD(T17, T18); T19 = VSUB(T17, T18); Ts = VSUB(To, Tr); TF = VADD(To, Tr); T1w = VADD(T1e, T1f); T1g = VSUB(T1e, T1f); Th = VADD(Td, Tg); TM = VSUB(Td, Tg); T1u = VADD(T1s, T1t); T1C = VSUB(T1s, T1t); T1n = VSUB(T16, T19); T1a = VADD(T16, T19); T1m = VSUB(T1d, T1g); T1h = VADD(T1d, T1g); T1x = VADD(T1v, T1w); T1D = VSUB(T1v, T1w); } } } } Tk = VSUB(Ta, Th); Ti = VADD(Ta, Th); { V TJ, T1k, T1A, TZ, Tj, T1E, T1G, TI, T10, T1j, T1z, T1i, T1y, TH; TJ = VSUB(TF, TG); TH = VADD(TF, TG); T1i = VADD(T1a, T1h); T1k = VSUB(T1a, T1h); T1y = VADD(T1u, T1x); T1A = VSUB(T1u, T1x); TZ = VADD(T3, Ti); Tj = VFNMS(LDK(KP250000000), Ti, T3); T1E = VMUL(LDK(KP951056516), VFMA(LDK(KP618033988), T1D, T1C)); T1G = VMUL(LDK(KP951056516), VFNMS(LDK(KP618033988), T1C, T1D)); TI = VFNMS(LDK(KP250000000), TH, TE); T10 = VADD(TE, TH); T1j = VFNMS(LDK(KP250000000), T1i, T13); ST(&(xo[0]), VADD(T1r, T1y), ovs, &(xo[0])); T1z = VFNMS(LDK(KP250000000), T1y, T1r); ST(&(xo[WS(os, 10)]), VADD(T13, T1i), ovs, &(xo[0])); { V T1p, T1l, T1o, T1q, T1F, T1B; TS = VFNMS(LDK(KP618033988), Ts, Tz); TA = VFMA(LDK(KP618033988), Tz, Ts); TN = VFMA(LDK(KP618033988), TM, TL); TV = VFNMS(LDK(KP618033988), TL, TM); ST(&(xo[WS(os, 5)]), VFMAI(T10, TZ), ovs, &(xo[WS(os, 1)])); ST(&(xo[WS(os, 15)]), VFNMSI(T10, TZ), ovs, &(xo[WS(os, 1)])); T1p = VFMA(LDK(KP559016994), T1k, T1j); T1l = VFNMS(LDK(KP559016994), T1k, T1j); T1o = VMUL(LDK(KP951056516), VFNMS(LDK(KP618033988), T1n, T1m)); T1q = VMUL(LDK(KP951056516), VFMA(LDK(KP618033988), T1m, T1n)); T1F = VFNMS(LDK(KP559016994), T1A, T1z); T1B = VFMA(LDK(KP559016994), T1A, T1z); ST(&(xo[WS(os, 14)]), VFNMSI(T1q, T1p), ovs, &(xo[0])); ST(&(xo[WS(os, 6)]), VFMAI(T1q, T1p), ovs, &(xo[0])); ST(&(xo[WS(os, 18)]), VFMAI(T1o, T1l), ovs, &(xo[0])); ST(&(xo[WS(os, 2)]), VFNMSI(T1o, T1l), ovs, &(xo[0])); ST(&(xo[WS(os, 16)]), VFMAI(T1E, T1B), ovs, &(xo[0])); ST(&(xo[WS(os, 4)]), VFNMSI(T1E, T1B), ovs, &(xo[0])); ST(&(xo[WS(os, 12)]), VFNMSI(T1G, T1F), ovs, &(xo[0])); ST(&(xo[WS(os, 8)]), VFMAI(T1G, T1F), ovs, &(xo[0])); TK = VFMA(LDK(KP559016994), TJ, TI); TU = VFNMS(LDK(KP559016994), TJ, TI); TR = VFNMS(LDK(KP559016994), Tk, Tj); Tl = VFMA(LDK(KP559016994), Tk, Tj); } } } { V TY, TW, TO, TQ, TB, TP, TX, TT; TY = VFMA(LDK(KP951056516), TV, TU); TW = VFNMS(LDK(KP951056516), TV, TU); TO = VFMA(LDK(KP951056516), TN, TK); TQ = VFNMS(LDK(KP951056516), TN, TK); TB = VFNMS(LDK(KP951056516), TA, Tl); TP = VFMA(LDK(KP951056516), TA, Tl); TX = VFNMS(LDK(KP951056516), TS, TR); TT = VFMA(LDK(KP951056516), TS, TR); ST(&(xo[WS(os, 9)]), VFMAI(TQ, TP), ovs, &(xo[WS(os, 1)])); ST(&(xo[WS(os, 11)]), VFNMSI(TQ, TP), ovs, &(xo[WS(os, 1)])); ST(&(xo[WS(os, 1)]), VFMAI(TO, TB), ovs, &(xo[WS(os, 1)])); ST(&(xo[WS(os, 19)]), VFNMSI(TO, TB), ovs, &(xo[WS(os, 1)])); ST(&(xo[WS(os, 17)]), VFMAI(TW, TT), ovs, &(xo[WS(os, 1)])); ST(&(xo[WS(os, 3)]), VFNMSI(TW, TT), ovs, &(xo[WS(os, 1)])); ST(&(xo[WS(os, 13)]), VFMAI(TY, TX), ovs, &(xo[WS(os, 1)])); ST(&(xo[WS(os, 7)]), VFNMSI(TY, TX), ovs, &(xo[WS(os, 1)])); } } } VLEAVE(); } static const kdft_desc desc = { 20, XSIMD_STRING("n1bv_20"), {58, 4, 46, 0}, &GENUS, 0, 0, 0, 0 }; void XSIMD(codelet_n1bv_20) (planner *p) { X(kdft_register) (p, n1bv_20, &desc); } #else /* HAVE_FMA */ /* Generated by: ../../../genfft/gen_notw_c.native -simd -compact -variables 4 -pipeline-latency 8 -sign 1 -n 20 -name n1bv_20 -include n1b.h */ /* * This function contains 104 FP additions, 24 FP multiplications, * (or, 92 additions, 12 multiplications, 12 fused multiply/add), * 53 stack variables, 4 constants, and 40 memory accesses */ #include "n1b.h" static void n1bv_20(const R *ri, const R *ii, R *ro, R *io, stride is, stride os, INT v, INT ivs, INT ovs) { DVK(KP587785252, +0.587785252292473129168705954639072768597652438); DVK(KP951056516, +0.951056516295153572116439333379382143405698634); DVK(KP250000000, +0.250000000000000000000000000000000000000000000); DVK(KP559016994, +0.559016994374947424102293417182819058860154590); { INT i; const R *xi; R *xo; xi = ii; xo = io; for (i = v; i > 0; i = i - VL, xi = xi + (VL * ivs), xo = xo + (VL * ovs), MAKE_VOLATILE_STRIDE(40, is), MAKE_VOLATILE_STRIDE(40, os)) { V T3, T1y, TH, T1i, Ts, TL, TM, Tz, T13, T16, T1j, T1u, T1v, T1w, T1r; V T1s, T1t, T1a, T1d, T1k, Ti, Tk, TE, TI, TZ, T10; { V T1, T2, T1g, TF, TG, T1h; T1 = LD(&(xi[0]), ivs, &(xi[0])); T2 = LD(&(xi[WS(is, 10)]), ivs, &(xi[0])); T1g = VADD(T1, T2); TF = LD(&(xi[WS(is, 5)]), ivs, &(xi[WS(is, 1)])); TG = LD(&(xi[WS(is, 15)]), ivs, &(xi[WS(is, 1)])); T1h = VADD(TF, TG); T3 = VSUB(T1, T2); T1y = VADD(T1g, T1h); TH = VSUB(TF, TG); T1i = VSUB(T1g, T1h); } { V T6, T11, Tv, T19, Ty, T1c, T9, T14, Td, T18, To, T12, Tr, T15, Tg; V T1b; { V T4, T5, Tt, Tu; T4 = LD(&(xi[WS(is, 4)]), ivs, &(xi[0])); T5 = LD(&(xi[WS(is, 14)]), ivs, &(xi[0])); T6 = VSUB(T4, T5); T11 = VADD(T4, T5); Tt = LD(&(xi[WS(is, 13)]), ivs, &(xi[WS(is, 1)])); Tu = LD(&(xi[WS(is, 3)]), ivs, &(xi[WS(is, 1)])); Tv = VSUB(Tt, Tu); T19 = VADD(Tt, Tu); } { V Tw, Tx, T7, T8; Tw = LD(&(xi[WS(is, 17)]), ivs, &(xi[WS(is, 1)])); Tx = LD(&(xi[WS(is, 7)]), ivs, &(xi[WS(is, 1)])); Ty = VSUB(Tw, Tx); T1c = VADD(Tw, Tx); T7 = LD(&(xi[WS(is, 16)]), ivs, &(xi[0])); T8 = LD(&(xi[WS(is, 6)]), ivs, &(xi[0])); T9 = VSUB(T7, T8); T14 = VADD(T7, T8); } { V Tb, Tc, Tm, Tn; Tb = LD(&(xi[WS(is, 8)]), ivs, &(xi[0])); Tc = LD(&(xi[WS(is, 18)]), ivs, &(xi[0])); Td = VSUB(Tb, Tc); T18 = VADD(Tb, Tc); Tm = LD(&(xi[WS(is, 9)]), ivs, &(xi[WS(is, 1)])); Tn = LD(&(xi[WS(is, 19)]), ivs, &(xi[WS(is, 1)])); To = VSUB(Tm, Tn); T12 = VADD(Tm, Tn); } { V Tp, Tq, Te, Tf; Tp = LD(&(xi[WS(is, 1)]), ivs, &(xi[WS(is, 1)])); Tq = LD(&(xi[WS(is, 11)]), ivs, &(xi[WS(is, 1)])); Tr = VSUB(Tp, Tq); T15 = VADD(Tp, Tq); Te = LD(&(xi[WS(is, 12)]), ivs, &(xi[0])); Tf = LD(&(xi[WS(is, 2)]), ivs, &(xi[0])); Tg = VSUB(Te, Tf); T1b = VADD(Te, Tf); } Ts = VSUB(To, Tr); TL = VSUB(T6, T9); TM = VSUB(Td, Tg); Tz = VSUB(Tv, Ty); T13 = VSUB(T11, T12); T16 = VSUB(T14, T15); T1j = VADD(T13, T16); T1u = VADD(T18, T19); T1v = VADD(T1b, T1c); T1w = VADD(T1u, T1v); T1r = VADD(T11, T12); T1s = VADD(T14, T15); T1t = VADD(T1r, T1s); T1a = VSUB(T18, T19); T1d = VSUB(T1b, T1c); T1k = VADD(T1a, T1d); { V Ta, Th, TC, TD; Ta = VADD(T6, T9); Th = VADD(Td, Tg); Ti = VADD(Ta, Th); Tk = VMUL(LDK(KP559016994), VSUB(Ta, Th)); TC = VADD(To, Tr); TD = VADD(Tv, Ty); TE = VMUL(LDK(KP559016994), VSUB(TC, TD)); TI = VADD(TC, TD); } } TZ = VADD(T3, Ti); T10 = VBYI(VADD(TH, TI)); ST(&(xo[WS(os, 15)]), VSUB(TZ, T10), ovs, &(xo[WS(os, 1)])); ST(&(xo[WS(os, 5)]), VADD(TZ, T10), ovs, &(xo[WS(os, 1)])); { V T1x, T1z, T1A, T1E, T1G, T1C, T1D, T1F, T1B; T1x = VMUL(LDK(KP559016994), VSUB(T1t, T1w)); T1z = VADD(T1t, T1w); T1A = VFNMS(LDK(KP250000000), T1z, T1y); T1C = VSUB(T1r, T1s); T1D = VSUB(T1u, T1v); T1E = VBYI(VFMA(LDK(KP951056516), T1C, VMUL(LDK(KP587785252), T1D))); T1G = VBYI(VFNMS(LDK(KP951056516), T1D, VMUL(LDK(KP587785252), T1C))); ST(&(xo[0]), VADD(T1y, T1z), ovs, &(xo[0])); T1F = VSUB(T1A, T1x); ST(&(xo[WS(os, 8)]), VSUB(T1F, T1G), ovs, &(xo[0])); ST(&(xo[WS(os, 12)]), VADD(T1G, T1F), ovs, &(xo[0])); T1B = VADD(T1x, T1A); ST(&(xo[WS(os, 4)]), VSUB(T1B, T1E), ovs, &(xo[0])); ST(&(xo[WS(os, 16)]), VADD(T1E, T1B), ovs, &(xo[0])); } { V T1n, T1l, T1m, T1f, T1p, T17, T1e, T1q, T1o; T1n = VMUL(LDK(KP559016994), VSUB(T1j, T1k)); T1l = VADD(T1j, T1k); T1m = VFNMS(LDK(KP250000000), T1l, T1i); T17 = VSUB(T13, T16); T1e = VSUB(T1a, T1d); T1f = VBYI(VFNMS(LDK(KP951056516), T1e, VMUL(LDK(KP587785252), T17))); T1p = VBYI(VFMA(LDK(KP951056516), T17, VMUL(LDK(KP587785252), T1e))); ST(&(xo[WS(os, 10)]), VADD(T1i, T1l), ovs, &(xo[0])); T1q = VADD(T1n, T1m); ST(&(xo[WS(os, 6)]), VADD(T1p, T1q), ovs, &(xo[0])); ST(&(xo[WS(os, 14)]), VSUB(T1q, T1p), ovs, &(xo[0])); T1o = VSUB(T1m, T1n); ST(&(xo[WS(os, 2)]), VADD(T1f, T1o), ovs, &(xo[0])); ST(&(xo[WS(os, 18)]), VSUB(T1o, T1f), ovs, &(xo[0])); } { V TA, TN, TU, TS, TK, TV, Tl, TR, TJ, Tj; TA = VFNMS(LDK(KP951056516), Tz, VMUL(LDK(KP587785252), Ts)); TN = VFNMS(LDK(KP951056516), TM, VMUL(LDK(KP587785252), TL)); TU = VFMA(LDK(KP951056516), TL, VMUL(LDK(KP587785252), TM)); TS = VFMA(LDK(KP951056516), Ts, VMUL(LDK(KP587785252), Tz)); TJ = VFNMS(LDK(KP250000000), TI, TH); TK = VSUB(TE, TJ); TV = VADD(TE, TJ); Tj = VFNMS(LDK(KP250000000), Ti, T3); Tl = VSUB(Tj, Tk); TR = VADD(Tk, Tj); { V TB, TO, TX, TY; TB = VSUB(Tl, TA); TO = VBYI(VSUB(TK, TN)); ST(&(xo[WS(os, 17)]), VSUB(TB, TO), ovs, &(xo[WS(os, 1)])); ST(&(xo[WS(os, 3)]), VADD(TB, TO), ovs, &(xo[WS(os, 1)])); TX = VADD(TR, TS); TY = VBYI(VSUB(TV, TU)); ST(&(xo[WS(os, 11)]), VSUB(TX, TY), ovs, &(xo[WS(os, 1)])); ST(&(xo[WS(os, 9)]), VADD(TX, TY), ovs, &(xo[WS(os, 1)])); } { V TP, TQ, TT, TW; TP = VADD(Tl, TA); TQ = VBYI(VADD(TN, TK)); ST(&(xo[WS(os, 13)]), VSUB(TP, TQ), ovs, &(xo[WS(os, 1)])); ST(&(xo[WS(os, 7)]), VADD(TP, TQ), ovs, &(xo[WS(os, 1)])); TT = VSUB(TR, TS); TW = VBYI(VADD(TU, TV)); ST(&(xo[WS(os, 19)]), VSUB(TT, TW), ovs, &(xo[WS(os, 1)])); ST(&(xo[WS(os, 1)]), VADD(TT, TW), ovs, &(xo[WS(os, 1)])); } } } } VLEAVE(); } static const kdft_desc desc = { 20, XSIMD_STRING("n1bv_20"), {92, 12, 12, 0}, &GENUS, 0, 0, 0, 0 }; void XSIMD(codelet_n1bv_20) (planner *p) { X(kdft_register) (p, n1bv_20, &desc); } #endif /* HAVE_FMA */ fftw-3.3.3/dft/simd/common/n1bv_25.c0000644000175000017500000010736312054410401013731 00000000000000/* * Copyright (c) 2003, 2007-11 Matteo Frigo * Copyright (c) 2003, 2007-11 Massachusetts Institute of Technology * * This program is free software; you can redistribute it and/or modify * it under the terms of the GNU General Public License as published by * the Free Software Foundation; either version 2 of the License, or * (at your option) any later version. * * This program is distributed in the hope that it will be useful, * but WITHOUT ANY WARRANTY; without even the implied warranty of * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the * GNU General Public License for more details. * * You should have received a copy of the GNU General Public License * along with this program; if not, write to the Free Software * Foundation, Inc., 51 Franklin Street, Fifth Floor, Boston, MA 02110-1301 USA * */ /* This file was automatically generated --- DO NOT EDIT */ /* Generated on Sun Nov 25 07:37:14 EST 2012 */ #include "codelet-dft.h" #ifdef HAVE_FMA /* Generated by: ../../../genfft/gen_notw_c.native -fma -reorder-insns -schedule-for-pipeline -simd -compact -variables 4 -pipeline-latency 8 -sign 1 -n 25 -name n1bv_25 -include n1b.h */ /* * This function contains 224 FP additions, 193 FP multiplications, * (or, 43 additions, 12 multiplications, 181 fused multiply/add), * 215 stack variables, 67 constants, and 50 memory accesses */ #include "n1b.h" static void n1bv_25(const R *ri, const R *ii, R *ro, R *io, stride is, stride os, INT v, INT ivs, INT ovs) { DVK(KP792626838, +0.792626838241819413632131824093538848057784557); DVK(KP876091699, +0.876091699473550838204498029706869638173524346); DVK(KP803003575, +0.803003575438660414833440593570376004635464850); DVK(KP617882369, +0.617882369114440893914546919006756321695042882); DVK(KP242145790, +0.242145790282157779872542093866183953459003101); DVK(KP968583161, +0.968583161128631119490168375464735813836012403); DVK(KP999544308, +0.999544308746292983948881682379742149196758193); DVK(KP683113946, +0.683113946453479238701949862233725244439656928); DVK(KP559154169, +0.559154169276087864842202529084232643714075927); DVK(KP904730450, +0.904730450839922351881287709692877908104763647); DVK(KP829049696, +0.829049696159252993975487806364305442437946767); DVK(KP831864738, +0.831864738706457140726048799369896829771167132); DVK(KP916574801, +0.916574801383451584742370439148878693530976769); DVK(KP894834959, +0.894834959464455102997960030820114611498661386); DVK(KP809385824, +0.809385824416008241660603814668679683846476688); DVK(KP447417479, +0.447417479732227551498980015410057305749330693); DVK(KP860541664, +0.860541664367944677098261680920518816412804187); DVK(KP897376177, +0.897376177523557693138608077137219684419427330); DVK(KP876306680, +0.876306680043863587308115903922062583399064238); DVK(KP681693190, +0.681693190061530575150324149145440022633095390); DVK(KP560319534, +0.560319534973832390111614715371676131169633784); DVK(KP855719849, +0.855719849902058969314654733608091555096772472); DVK(KP237294955, +0.237294955877110315393888866460840817927895961); DVK(KP949179823, +0.949179823508441261575555465843363271711583843); DVK(KP904508497, +0.904508497187473712051146708591409529430077295); DVK(KP997675361, +0.997675361079556513670859573984492383596555031); DVK(KP262346850, +0.262346850930607871785420028382979691334784273); DVK(KP763932022, +0.763932022500210303590826331268723764559381640); DVK(KP992114701, +0.992114701314477831049793042785778521453036709); DVK(KP690983005, +0.690983005625052575897706582817180941139845410); DVK(KP952936919, +0.952936919628306576880750665357914584765951388); DVK(KP998026728, +0.998026728428271561952336806863450553336905220); DVK(KP570584518, +0.570584518783621657366766175430996792655723863); DVK(KP669429328, +0.669429328479476605641803240971985825917022098); DVK(KP923225144, +0.923225144846402650453449441572664695995209956); DVK(KP906616052, +0.906616052148196230441134447086066874408359177); DVK(KP956723877, +0.956723877038460305821989399535483155872969262); DVK(KP522616830, +0.522616830205754336872861364785224694908468440); DVK(KP945422727, +0.945422727388575946270360266328811958657216298); DVK(KP912575812, +0.912575812670962425556968549836277086778922727); DVK(KP982009705, +0.982009705009746369461829878184175962711969869); DVK(KP921078979, +0.921078979742360627699756128143719920817673854); DVK(KP734762448, +0.734762448793050413546343770063151342619912334); DVK(KP951056516, +0.951056516295153572116439333379382143405698634); DVK(KP958953096, +0.958953096729998668045963838399037225970891871); DVK(KP867381224, +0.867381224396525206773171885031575671309956167); DVK(KP269969613, +0.269969613759572083574752974412347470060951301); DVK(KP244189809, +0.244189809627953270309879511234821255780225091); DVK(KP845997307, +0.845997307939530944175097360758058292389769300); DVK(KP772036680, +0.772036680810363904029489473607579825330539880); DVK(KP132830569, +0.132830569247582714407653942074819768844536507); DVK(KP120146378, +0.120146378570687701782758537356596213647956445); DVK(KP987388751, +0.987388751065621252324603216482382109400433949); DVK(KP893101515, +0.893101515366181661711202267938416198338079437); DVK(KP786782374, +0.786782374965295178365099601674911834788448471); DVK(KP869845200, +0.869845200362138853122720822420327157933056305); DVK(KP447533225, +0.447533225982656890041886979663652563063114397); DVK(KP494780565, +0.494780565770515410344588413655324772219443730); DVK(KP578046249, +0.578046249379945007321754579646815604023525655); DVK(KP522847744, +0.522847744331509716623755382187077770911012542); DVK(KP059835404, +0.059835404262124915169548397419498386427871950); DVK(KP066152395, +0.066152395967733048213034281011006031460903353); DVK(KP603558818, +0.603558818296015001454675132653458027918768137); DVK(KP667278218, +0.667278218140296670899089292254759909713898805); DVK(KP559016994, +0.559016994374947424102293417182819058860154590); DVK(KP250000000, +0.250000000000000000000000000000000000000000000); DVK(KP618033988, +0.618033988749894848204586834365638117720309180); { INT i; const R *xi; R *xo; xi = ii; xo = io; for (i = v; i > 0; i = i - VL, xi = xi + (VL * ivs), xo = xo + (VL * ovs), MAKE_VOLATILE_STRIDE(50, is), MAKE_VOLATILE_STRIDE(50, os)) { V T1g, T1k, T1I, T24, T2a, T1G, T1A, T1l, T1B, T1H, T1d; { V T2z, T1q, Ta, T9, T3n, Ty, Tl, T2O, T2W, T2l, T2s, TV, T1i, T1K, T1S; V T3z, T3t, Tk, T3o, Tp, T2g, T2N, T2V, T2o, T2t, T1a, T1j, T1J, T1R, Tz; V Tt, TA, Tw; { V T1, T5, T6, T2, T3; T1 = LD(&(xi[0]), ivs, &(xi[0])); T5 = LD(&(xi[WS(is, 10)]), ivs, &(xi[0])); T6 = LD(&(xi[WS(is, 15)]), ivs, &(xi[WS(is, 1)])); T2 = LD(&(xi[WS(is, 5)]), ivs, &(xi[WS(is, 1)])); T3 = LD(&(xi[WS(is, 20)]), ivs, &(xi[0])); { V TH, TW, TK, TS, T10, T8, TN, TT, T17, TZ, T11; TH = LD(&(xi[WS(is, 2)]), ivs, &(xi[0])); TW = LD(&(xi[WS(is, 3)]), ivs, &(xi[WS(is, 1)])); { V TI, TJ, TL, T7, T1p, T4, T1o, TM, TX, TY; TI = LD(&(xi[WS(is, 7)]), ivs, &(xi[WS(is, 1)])); TJ = LD(&(xi[WS(is, 22)]), ivs, &(xi[0])); TL = LD(&(xi[WS(is, 12)]), ivs, &(xi[0])); T7 = VADD(T5, T6); T1p = VSUB(T5, T6); T4 = VADD(T2, T3); T1o = VSUB(T2, T3); TM = LD(&(xi[WS(is, 17)]), ivs, &(xi[WS(is, 1)])); TX = LD(&(xi[WS(is, 8)]), ivs, &(xi[0])); TK = VADD(TI, TJ); TS = VSUB(TI, TJ); TY = LD(&(xi[WS(is, 23)]), ivs, &(xi[WS(is, 1)])); T10 = LD(&(xi[WS(is, 13)]), ivs, &(xi[WS(is, 1)])); T2z = VFNMS(LDK(KP618033988), T1o, T1p); T1q = VFMA(LDK(KP618033988), T1p, T1o); Ta = VSUB(T4, T7); T8 = VADD(T4, T7); TN = VADD(TL, TM); TT = VSUB(TM, TL); T17 = VSUB(TX, TY); TZ = VADD(TX, TY); T11 = LD(&(xi[WS(is, 18)]), ivs, &(xi[0])); } { V Tc, T2m, T19, Tn, To, Tr, Tj, T16, T2n, Ts, Tu, Tv; { V TU, T2j, TO, TQ, T12, T18; Tc = LD(&(xi[WS(is, 1)]), ivs, &(xi[WS(is, 1)])); T9 = VFNMS(LDK(KP250000000), T8, T1); T3n = VADD(T1, T8); TU = VFNMS(LDK(KP618033988), TT, TS); T2j = VFMA(LDK(KP618033988), TS, TT); TO = VADD(TK, TN); TQ = VSUB(TN, TK); T12 = VADD(T10, T11); T18 = VSUB(T10, T11); Ty = LD(&(xi[WS(is, 4)]), ivs, &(xi[0])); { V T3r, T15, T13, Tf, Ti, T2k, TR, TP, T3s, T14; { V Td, Te, Tg, Th; Td = LD(&(xi[WS(is, 6)]), ivs, &(xi[0])); Te = LD(&(xi[WS(is, 21)]), ivs, &(xi[WS(is, 1)])); Tg = LD(&(xi[WS(is, 11)]), ivs, &(xi[WS(is, 1)])); Th = LD(&(xi[WS(is, 16)]), ivs, &(xi[0])); TP = VFNMS(LDK(KP250000000), TO, TH); T3r = VADD(TH, TO); T2m = VFNMS(LDK(KP618033988), T17, T18); T19 = VFMA(LDK(KP618033988), T18, T17); T15 = VSUB(T12, TZ); T13 = VADD(TZ, T12); Tf = VADD(Td, Te); Tn = VSUB(Td, Te); To = VSUB(Th, Tg); Ti = VADD(Tg, Th); } T2k = VFMA(LDK(KP559016994), TQ, TP); TR = VFNMS(LDK(KP559016994), TQ, TP); Tr = LD(&(xi[WS(is, 24)]), ivs, &(xi[0])); T3s = VADD(TW, T13); T14 = VFNMS(LDK(KP250000000), T13, TW); Tj = VADD(Tf, Ti); Tl = VSUB(Tf, Ti); T2O = VFNMS(LDK(KP667278218), T2k, T2j); T2W = VFMA(LDK(KP603558818), T2j, T2k); T2l = VFMA(LDK(KP066152395), T2k, T2j); T2s = VFNMS(LDK(KP059835404), T2j, T2k); TV = VFNMS(LDK(KP522847744), TU, TR); T1i = VFMA(LDK(KP578046249), TR, TU); T1K = VFNMS(LDK(KP494780565), TR, TU); T1S = VFMA(LDK(KP447533225), TU, TR); T16 = VFNMS(LDK(KP559016994), T15, T14); T2n = VFMA(LDK(KP559016994), T15, T14); T3z = VSUB(T3r, T3s); T3t = VADD(T3r, T3s); Ts = LD(&(xi[WS(is, 9)]), ivs, &(xi[WS(is, 1)])); Tu = LD(&(xi[WS(is, 19)]), ivs, &(xi[WS(is, 1)])); Tv = LD(&(xi[WS(is, 14)]), ivs, &(xi[0])); } } Tk = VFNMS(LDK(KP250000000), Tj, Tc); T3o = VADD(Tc, Tj); Tp = VFNMS(LDK(KP618033988), To, Tn); T2g = VFMA(LDK(KP618033988), Tn, To); T2N = VFMA(LDK(KP066152395), T2n, T2m); T2V = VFNMS(LDK(KP059835404), T2m, T2n); T2o = VFMA(LDK(KP869845200), T2n, T2m); T2t = VFNMS(LDK(KP786782374), T2m, T2n); T1a = VFNMS(LDK(KP893101515), T19, T16); T1j = VFMA(LDK(KP987388751), T16, T19); T1J = VFNMS(LDK(KP120146378), T19, T16); T1R = VFMA(LDK(KP132830569), T16, T19); Tz = VADD(Ts, Tr); Tt = VSUB(Tr, Ts); TA = VADD(Tv, Tu); Tw = VSUB(Tu, Tv); } } } { V T2p, T2I, T2u, T2C, Tx, T2d, T2X, T34, T2P, T3b, T2b, Tb, T2Q, T2Z, T2h; V T2w, Tq, T1e, T1M, T1U, TE, T2c, T3q, T3y; T2p = VFNMS(LDK(KP772036680), T2o, T2l); T2I = VFMA(LDK(KP772036680), T2o, T2l); T2u = VFMA(LDK(KP772036680), T2t, T2s); T2C = VFNMS(LDK(KP772036680), T2t, T2s); { V TD, TB, Tm, T2f, T3p, TC; Tx = VFMA(LDK(KP618033988), Tw, Tt); T2d = VFNMS(LDK(KP618033988), Tt, Tw); TD = VSUB(Tz, TA); TB = VADD(Tz, TA); Tm = VFMA(LDK(KP559016994), Tl, Tk); T2f = VFNMS(LDK(KP559016994), Tl, Tk); T2X = VFMA(LDK(KP845997307), T2W, T2V); T34 = VFNMS(LDK(KP845997307), T2W, T2V); T2P = VFNMS(LDK(KP845997307), T2O, T2N); T3b = VFMA(LDK(KP845997307), T2O, T2N); T2b = VFNMS(LDK(KP559016994), Ta, T9); Tb = VFMA(LDK(KP559016994), Ta, T9); T3p = VADD(Ty, TB); TC = VFMS(LDK(KP250000000), TB, Ty); T2Q = VFNMS(LDK(KP522847744), T2g, T2f); T2Z = VFMA(LDK(KP578046249), T2f, T2g); T2h = VFMA(LDK(KP893101515), T2g, T2f); T2w = VFNMS(LDK(KP987388751), T2f, T2g); Tq = VFNMS(LDK(KP244189809), Tp, Tm); T1e = VFMA(LDK(KP269969613), Tm, Tp); T1M = VFMA(LDK(KP667278218), Tm, Tp); T1U = VFNMS(LDK(KP603558818), Tp, Tm); TE = VFNMS(LDK(KP559016994), TD, TC); T2c = VFMA(LDK(KP559016994), TD, TC); T3q = VADD(T3o, T3p); T3y = VSUB(T3o, T3p); } { V T1Z, T25, T1P, T22, T1X, TG, T1b, T28, T1t, T1y, T1x, T1E, T1Q, T1Y; { V T26, T1L, T1T, TF, T1f, T1W, T3m, T3g, T2M, T2G, T39, T3j, T21, T1O, T20; V T27; T26 = VFMA(LDK(KP867381224), T1K, T1J); T1L = VFNMS(LDK(KP867381224), T1K, T1J); T20 = VFNMS(LDK(KP958953096), T1S, T1R); T1T = VFMA(LDK(KP958953096), T1S, T1R); { V T2R, T2Y, T2e, T2v, T1N, T1V; T2R = VFNMS(LDK(KP494780565), T2c, T2d); T2Y = VFMA(LDK(KP447533225), T2d, T2c); T2e = VFMA(LDK(KP120146378), T2d, T2c); T2v = VFNMS(LDK(KP132830569), T2c, T2d); TF = VFNMS(LDK(KP667278218), TE, Tx); T1f = VFMA(LDK(KP603558818), Tx, TE); T1N = VFMA(LDK(KP869845200), TE, Tx); T1V = VFNMS(LDK(KP786782374), Tx, TE); { V T3A, T3C, T3w, T3u; T3A = VMUL(LDK(KP951056516), VFMA(LDK(KP618033988), T3z, T3y)); T3C = VMUL(LDK(KP951056516), VFNMS(LDK(KP618033988), T3y, T3z)); T3w = VSUB(T3q, T3t); T3u = VADD(T3q, T3t); { V T2B, T2x, T2H, T2i; T2B = VFMA(LDK(KP734762448), T2w, T2v); T2x = VFNMS(LDK(KP734762448), T2w, T2v); T2H = VFNMS(LDK(KP734762448), T2h, T2e); T2i = VFMA(LDK(KP734762448), T2h, T2e); { V T30, T35, T3c, T2S, T3v; T30 = VFNMS(LDK(KP921078979), T2Z, T2Y); T35 = VFMA(LDK(KP921078979), T2Z, T2Y); T3c = VFMA(LDK(KP982009705), T2R, T2Q); T2S = VFNMS(LDK(KP982009705), T2R, T2Q); T1W = VFMA(LDK(KP912575812), T1V, T1U); T1Z = VFNMS(LDK(KP912575812), T1V, T1U); T1O = VFMA(LDK(KP912575812), T1N, T1M); T25 = VFNMS(LDK(KP912575812), T1N, T1M); ST(&(xo[0]), VADD(T3u, T3n), ovs, &(xo[0])); T3v = VFNMS(LDK(KP250000000), T3u, T3n); { V T2y, T2J, T2q, T2D; T2y = VFMA(LDK(KP945422727), T2x, T2u); T2J = VFMA(LDK(KP522616830), T2x, T2I); T2q = VFMA(LDK(KP956723877), T2p, T2i); T2D = VFNMS(LDK(KP522616830), T2i, T2C); { V T3e, T31, T36, T2T; T3e = VFMA(LDK(KP906616052), T30, T2X); T31 = VFNMS(LDK(KP906616052), T30, T2X); T36 = VFNMS(LDK(KP923225144), T2S, T2P); T2T = VFMA(LDK(KP923225144), T2S, T2P); { V T3k, T3d, T3x, T3B; T3k = VFNMS(LDK(KP669429328), T3b, T3c); T3d = VFMA(LDK(KP570584518), T3c, T3b); T3x = VFMA(LDK(KP559016994), T3w, T3v); T3B = VFNMS(LDK(KP559016994), T3w, T3v); { V T2A, T2K, T2r, T2E; T2A = VMUL(LDK(KP998026728), VFMA(LDK(KP952936919), T2z, T2y)); T2K = VFNMS(LDK(KP690983005), T2J, T2u); T2r = VFMA(LDK(KP992114701), T2q, T2b); T2E = VFMA(LDK(KP763932022), T2D, T2p); { V T32, T3a, T37, T3h; T32 = VMUL(LDK(KP998026728), VFNMS(LDK(KP952936919), T2z, T31)); T3a = VFMA(LDK(KP262346850), T31, T2z); T37 = VFNMS(LDK(KP997675361), T36, T35); T3h = VFNMS(LDK(KP904508497), T36, T34); { V T2U, T33, T3l, T3f; T2U = VFMA(LDK(KP949179823), T2T, T2b); T33 = VFNMS(LDK(KP237294955), T2T, T2b); T3l = VFNMS(LDK(KP669429328), T3e, T3k); T3f = VFMA(LDK(KP618033988), T3e, T3d); ST(&(xo[WS(os, 20)]), VFNMSI(T3A, T3x), ovs, &(xo[0])); ST(&(xo[WS(os, 5)]), VFMAI(T3A, T3x), ovs, &(xo[WS(os, 1)])); ST(&(xo[WS(os, 15)]), VFMAI(T3C, T3B), ovs, &(xo[WS(os, 1)])); ST(&(xo[WS(os, 10)]), VFNMSI(T3C, T3B), ovs, &(xo[0])); { V T2L, T2F, T38, T3i; T2L = VFMA(LDK(KP855719849), T2K, T2H); ST(&(xo[WS(os, 3)]), VFMAI(T2A, T2r), ovs, &(xo[WS(os, 1)])); ST(&(xo[WS(os, 22)]), VFNMSI(T2A, T2r), ovs, &(xo[0])); T2F = VFNMS(LDK(KP855719849), T2E, T2B); T38 = VFMA(LDK(KP560319534), T37, T34); T3i = VFNMS(LDK(KP681693190), T3h, T35); ST(&(xo[WS(os, 2)]), VFMAI(T32, T2U), ovs, &(xo[0])); ST(&(xo[WS(os, 23)]), VFNMSI(T32, T2U), ovs, &(xo[WS(os, 1)])); T3m = VMUL(LDK(KP951056516), VFNMS(LDK(KP876306680), T3l, T3a)); T3g = VMUL(LDK(KP951056516), VFNMS(LDK(KP949179823), T3f, T3a)); T2M = VMUL(LDK(KP951056516), VFNMS(LDK(KP992114701), T2L, T2z)); T2G = VFMA(LDK(KP897376177), T2F, T2b); T39 = VFNMS(LDK(KP949179823), T38, T33); T3j = VFNMS(LDK(KP860541664), T3i, T33); T21 = VFMA(LDK(KP447417479), T1O, T20); } } } } } } } } } } } T1P = VFNMS(LDK(KP809385824), T1O, T1L); ST(&(xo[WS(os, 17)]), VFNMSI(T2M, T2G), ovs, &(xo[WS(os, 1)])); ST(&(xo[WS(os, 8)]), VFMAI(T2M, T2G), ovs, &(xo[0])); ST(&(xo[WS(os, 13)]), VFMAI(T3g, T39), ovs, &(xo[WS(os, 1)])); ST(&(xo[WS(os, 12)]), VFNMSI(T3g, T39), ovs, &(xo[0])); ST(&(xo[WS(os, 18)]), VFMAI(T3m, T3j), ovs, &(xo[0])); ST(&(xo[WS(os, 7)]), VFNMSI(T3m, T3j), ovs, &(xo[WS(os, 1)])); T22 = VFMA(LDK(KP690983005), T21, T1L); T27 = VFMA(LDK(KP447417479), T1W, T26); T1X = VFMA(LDK(KP894834959), T1W, T1T); { V T1r, T1s, T1v, T1w; T1r = VFNMS(LDK(KP916574801), T1f, T1e); T1g = VFMA(LDK(KP916574801), T1f, T1e); T1k = VFNMS(LDK(KP831864738), T1j, T1i); T1s = VFMA(LDK(KP831864738), T1j, T1i); T1v = VFNMS(LDK(KP829049696), TF, Tq); TG = VFMA(LDK(KP829049696), TF, Tq); T1b = VFMA(LDK(KP831864738), T1a, TV); T1w = VFNMS(LDK(KP831864738), T1a, TV); T28 = VFNMS(LDK(KP763932022), T27, T1T); T1t = VFMA(LDK(KP904730450), T1s, T1r); T1y = VFNMS(LDK(KP904730450), T1s, T1r); T1x = VFMA(LDK(KP559154169), T1w, T1v); T1E = VFNMS(LDK(KP683113946), T1v, T1w); } } T1Q = VFNMS(LDK(KP992114701), T1P, Tb); T1Y = VMUL(LDK(KP951056516), VFNMS(LDK(KP992114701), T1X, T1q)); { V T1u, T1F, T1z, T1h, T1c, T23, T29; T23 = VFNMS(LDK(KP999544308), T22, T1Z); T29 = VFNMS(LDK(KP999544308), T28, T25); T1I = VMUL(LDK(KP951056516), VFMA(LDK(KP968583161), T1t, T1q)); T1u = VFNMS(LDK(KP242145790), T1t, T1q); T1F = VFMA(LDK(KP617882369), T1y, T1E); T1z = VFMA(LDK(KP559016994), T1y, T1x); T1h = VFNMS(LDK(KP904730450), T1b, TG); T1c = VFMA(LDK(KP904730450), T1b, TG); ST(&(xo[WS(os, 21)]), VFMAI(T1Y, T1Q), ovs, &(xo[WS(os, 1)])); ST(&(xo[WS(os, 4)]), VFNMSI(T1Y, T1Q), ovs, &(xo[0])); T24 = VFNMS(LDK(KP803003575), T23, Tb); T2a = VMUL(LDK(KP951056516), VFNMS(LDK(KP803003575), T29, T1q)); T1G = VMUL(LDK(KP951056516), VFNMS(LDK(KP876306680), T1F, T1u)); T1A = VMUL(LDK(KP951056516), VFMA(LDK(KP968583161), T1z, T1u)); T1l = VFNMS(LDK(KP904730450), T1k, T1h); T1B = VADD(T1g, T1h); T1H = VFMA(LDK(KP968583161), T1c, Tb); T1d = VFNMS(LDK(KP242145790), T1c, Tb); } } } } ST(&(xo[WS(os, 16)]), VFMAI(T2a, T24), ovs, &(xo[0])); ST(&(xo[WS(os, 9)]), VFNMSI(T2a, T24), ovs, &(xo[WS(os, 1)])); { V T1m, T1C, T1n, T1D; T1m = VFNMS(LDK(KP618033988), T1l, T1g); T1C = VFNMS(LDK(KP683113946), T1B, T1k); ST(&(xo[WS(os, 24)]), VFNMSI(T1I, T1H), ovs, &(xo[0])); ST(&(xo[WS(os, 1)]), VFMAI(T1I, T1H), ovs, &(xo[WS(os, 1)])); T1n = VFNMS(LDK(KP876091699), T1m, T1d); T1D = VFMA(LDK(KP792626838), T1C, T1d); ST(&(xo[WS(os, 19)]), VFNMSI(T1A, T1n), ovs, &(xo[WS(os, 1)])); ST(&(xo[WS(os, 6)]), VFMAI(T1A, T1n), ovs, &(xo[0])); ST(&(xo[WS(os, 14)]), VFNMSI(T1G, T1D), ovs, &(xo[0])); ST(&(xo[WS(os, 11)]), VFMAI(T1G, T1D), ovs, &(xo[WS(os, 1)])); } } } VLEAVE(); } static const kdft_desc desc = { 25, XSIMD_STRING("n1bv_25"), {43, 12, 181, 0}, &GENUS, 0, 0, 0, 0 }; void XSIMD(codelet_n1bv_25) (planner *p) { X(kdft_register) (p, n1bv_25, &desc); } #else /* HAVE_FMA */ /* Generated by: ../../../genfft/gen_notw_c.native -simd -compact -variables 4 -pipeline-latency 8 -sign 1 -n 25 -name n1bv_25 -include n1b.h */ /* * This function contains 224 FP additions, 140 FP multiplications, * (or, 147 additions, 63 multiplications, 77 fused multiply/add), * 115 stack variables, 40 constants, and 50 memory accesses */ #include "n1b.h" static void n1bv_25(const R *ri, const R *ii, R *ro, R *io, stride is, stride os, INT v, INT ivs, INT ovs) { DVK(KP809016994, +0.809016994374947424102293417182819058860154590); DVK(KP309016994, +0.309016994374947424102293417182819058860154590); DVK(KP637423989, +0.637423989748689710176712811676016195434917298); DVK(KP1_541026485, +1.541026485551578461606019272792355694543335344); DVK(KP125333233, +0.125333233564304245373118759816508793942918247); DVK(KP1_984229402, +1.984229402628955662099586085571557042906073418); DVK(KP770513242, +0.770513242775789230803009636396177847271667672); DVK(KP1_274847979, +1.274847979497379420353425623352032390869834596); DVK(KP992114701, +0.992114701314477831049793042785778521453036709); DVK(KP250666467, +0.250666467128608490746237519633017587885836494); DVK(KP851558583, +0.851558583130145297725004891488503407959946084); DVK(KP904827052, +0.904827052466019527713668647932697593970413911); DVK(KP425779291, +0.425779291565072648862502445744251703979973042); DVK(KP1_809654104, +1.809654104932039055427337295865395187940827822); DVK(KP497379774, +0.497379774329709576484567492012895936835134813); DVK(KP968583161, +0.968583161128631119490168375464735813836012403); DVK(KP248689887, +0.248689887164854788242283746006447968417567406); DVK(KP1_937166322, +1.937166322257262238980336750929471627672024806); DVK(KP1_688655851, +1.688655851004030157097116127933363010763318483); DVK(KP535826794, +0.535826794978996618271308767867639978063575346); DVK(KP481753674, +0.481753674101715274987191502872129653528542010); DVK(KP1_752613360, +1.752613360087727174616231807844125166798128477); DVK(KP844327925, +0.844327925502015078548558063966681505381659241); DVK(KP1_071653589, +1.071653589957993236542617535735279956127150691); DVK(KP963507348, +0.963507348203430549974383005744259307057084020); DVK(KP876306680, +0.876306680043863587308115903922062583399064238); DVK(KP1_996053456, +1.996053456856543123904673613726901106673810439); DVK(KP062790519, +0.062790519529313376076178224565631133122484832); DVK(KP684547105, +0.684547105928688673732283357621209269889519233); DVK(KP1_457937254, +1.457937254842823046293460638110518222745143328); DVK(KP998026728, +0.998026728428271561952336806863450553336905220); DVK(KP125581039, +0.125581039058626752152356449131262266244969664); DVK(KP1_369094211, +1.369094211857377347464566715242418539779038465); DVK(KP728968627, +0.728968627421411523146730319055259111372571664); DVK(KP293892626, +0.293892626146236564584352977319536384298826219); DVK(KP475528258, +0.475528258147576786058219666689691071702849317); DVK(KP250000000, +0.250000000000000000000000000000000000000000000); DVK(KP951056516, +0.951056516295153572116439333379382143405698634); DVK(KP587785252, +0.587785252292473129168705954639072768597652438); DVK(KP559016994, +0.559016994374947424102293417182819058860154590); { INT i; const R *xi; R *xo; xi = ii; xo = io; for (i = v; i > 0; i = i - VL, xi = xi + (VL * ivs), xo = xo + (VL * ovs), MAKE_VOLATILE_STRIDE(50, is), MAKE_VOLATILE_STRIDE(50, os)) { V T1b, T2o, T1v, T1e, T2W, T2P, T2Q, T2U, T11, T27, TY, T26, T12, T2f, T1j; V T28, TM, T24, TJ, T23, TN, T2e, T1i, T25, T2M, T2N, T2T, Tm, T1W, Tt; V T1X, Tu, T20, Tw, T1Y, T7, T1U, Te, T1T, Tf, T21, Tx, T1V; { V T1c, T1a, T1t, T17, T1r; T1c = LD(&(xi[0]), ivs, &(xi[0])); { V T18, T19, T15, T16; T18 = LD(&(xi[WS(is, 10)]), ivs, &(xi[0])); T19 = LD(&(xi[WS(is, 15)]), ivs, &(xi[WS(is, 1)])); T1a = VADD(T18, T19); T1t = VSUB(T18, T19); T15 = LD(&(xi[WS(is, 5)]), ivs, &(xi[WS(is, 1)])); T16 = LD(&(xi[WS(is, 20)]), ivs, &(xi[0])); T17 = VADD(T15, T16); T1r = VSUB(T15, T16); } { V T2n, T1s, T1u, T1d; T1b = VMUL(LDK(KP559016994), VSUB(T17, T1a)); T2n = VMUL(LDK(KP587785252), T1r); T2o = VFNMS(LDK(KP951056516), T1t, T2n); T1s = VMUL(LDK(KP951056516), T1r); T1u = VMUL(LDK(KP587785252), T1t); T1v = VADD(T1s, T1u); T1d = VADD(T17, T1a); T1e = VFNMS(LDK(KP250000000), T1d, T1c); T2W = VADD(T1c, T1d); } } { V TG, TV, TF, TL, TH, TK, TU, T10, TW, TZ, TX, TI; TG = LD(&(xi[WS(is, 2)]), ivs, &(xi[0])); TV = LD(&(xi[WS(is, 3)]), ivs, &(xi[WS(is, 1)])); { V Tz, TA, TB, TC, TD, TE; Tz = LD(&(xi[WS(is, 7)]), ivs, &(xi[WS(is, 1)])); TA = LD(&(xi[WS(is, 22)]), ivs, &(xi[0])); TB = VADD(Tz, TA); TC = LD(&(xi[WS(is, 12)]), ivs, &(xi[0])); TD = LD(&(xi[WS(is, 17)]), ivs, &(xi[WS(is, 1)])); TE = VADD(TC, TD); TF = VMUL(LDK(KP559016994), VSUB(TB, TE)); TL = VSUB(TC, TD); TH = VADD(TB, TE); TK = VSUB(Tz, TA); } { V TO, TP, TQ, TR, TS, TT; TO = LD(&(xi[WS(is, 8)]), ivs, &(xi[0])); TP = LD(&(xi[WS(is, 23)]), ivs, &(xi[WS(is, 1)])); TQ = VADD(TO, TP); TR = LD(&(xi[WS(is, 13)]), ivs, &(xi[WS(is, 1)])); TS = LD(&(xi[WS(is, 18)]), ivs, &(xi[0])); TT = VADD(TR, TS); TU = VMUL(LDK(KP559016994), VSUB(TQ, TT)); T10 = VSUB(TR, TS); TW = VADD(TQ, TT); TZ = VSUB(TO, TP); } T2P = VADD(TG, TH); T2Q = VADD(TV, TW); T2U = VADD(T2P, T2Q); T11 = VFMA(LDK(KP475528258), TZ, VMUL(LDK(KP293892626), T10)); T27 = VFNMS(LDK(KP475528258), T10, VMUL(LDK(KP293892626), TZ)); TX = VFNMS(LDK(KP250000000), TW, TV); TY = VADD(TU, TX); T26 = VSUB(TX, TU); T12 = VFNMS(LDK(KP1_369094211), T11, VMUL(LDK(KP728968627), TY)); T2f = VFMA(LDK(KP125581039), T27, VMUL(LDK(KP998026728), T26)); T1j = VFMA(LDK(KP1_457937254), T11, VMUL(LDK(KP684547105), TY)); T28 = VFNMS(LDK(KP1_996053456), T27, VMUL(LDK(KP062790519), T26)); TM = VFMA(LDK(KP475528258), TK, VMUL(LDK(KP293892626), TL)); T24 = VFNMS(LDK(KP475528258), TL, VMUL(LDK(KP293892626), TK)); TI = VFNMS(LDK(KP250000000), TH, TG); TJ = VADD(TF, TI); T23 = VSUB(TI, TF); TN = VFNMS(LDK(KP963507348), TM, VMUL(LDK(KP876306680), TJ)); T2e = VFMA(LDK(KP1_071653589), T24, VMUL(LDK(KP844327925), T23)); T1i = VFMA(LDK(KP1_752613360), TM, VMUL(LDK(KP481753674), TJ)); T25 = VFNMS(LDK(KP1_688655851), T24, VMUL(LDK(KP535826794), T23)); } { V Tb, Tq, T3, Tc, T6, Ta, Ti, Tr, Tl, Tp, Ts, Td; Tb = LD(&(xi[WS(is, 4)]), ivs, &(xi[0])); Tq = LD(&(xi[WS(is, 1)]), ivs, &(xi[WS(is, 1)])); { V T1, T2, T8, T4, T5, T9; T1 = LD(&(xi[WS(is, 9)]), ivs, &(xi[WS(is, 1)])); T2 = LD(&(xi[WS(is, 24)]), ivs, &(xi[0])); T8 = VADD(T1, T2); T4 = LD(&(xi[WS(is, 14)]), ivs, &(xi[0])); T5 = LD(&(xi[WS(is, 19)]), ivs, &(xi[WS(is, 1)])); T9 = VADD(T4, T5); T3 = VSUB(T1, T2); Tc = VADD(T8, T9); T6 = VSUB(T4, T5); Ta = VMUL(LDK(KP559016994), VSUB(T8, T9)); } { V Tg, Th, Tn, Tj, Tk, To; Tg = LD(&(xi[WS(is, 6)]), ivs, &(xi[0])); Th = LD(&(xi[WS(is, 21)]), ivs, &(xi[WS(is, 1)])); Tn = VADD(Tg, Th); Tj = LD(&(xi[WS(is, 11)]), ivs, &(xi[WS(is, 1)])); Tk = LD(&(xi[WS(is, 16)]), ivs, &(xi[0])); To = VADD(Tj, Tk); Ti = VSUB(Tg, Th); Tr = VADD(Tn, To); Tl = VSUB(Tj, Tk); Tp = VMUL(LDK(KP559016994), VSUB(Tn, To)); } T2M = VADD(Tq, Tr); T2N = VADD(Tb, Tc); T2T = VADD(T2M, T2N); Tm = VFMA(LDK(KP475528258), Ti, VMUL(LDK(KP293892626), Tl)); T1W = VFNMS(LDK(KP475528258), Tl, VMUL(LDK(KP293892626), Ti)); Ts = VFNMS(LDK(KP250000000), Tr, Tq); Tt = VADD(Tp, Ts); T1X = VSUB(Ts, Tp); Tu = VFMA(LDK(KP1_937166322), Tm, VMUL(LDK(KP248689887), Tt)); T20 = VFNMS(LDK(KP963507348), T1W, VMUL(LDK(KP876306680), T1X)); Tw = VFNMS(LDK(KP497379774), Tm, VMUL(LDK(KP968583161), Tt)); T1Y = VFMA(LDK(KP1_752613360), T1W, VMUL(LDK(KP481753674), T1X)); T7 = VFMA(LDK(KP475528258), T3, VMUL(LDK(KP293892626), T6)); T1U = VFNMS(LDK(KP475528258), T6, VMUL(LDK(KP293892626), T3)); Td = VFNMS(LDK(KP250000000), Tc, Tb); Te = VADD(Ta, Td); T1T = VSUB(Td, Ta); Tf = VFMA(LDK(KP1_071653589), T7, VMUL(LDK(KP844327925), Te)); T21 = VFMA(LDK(KP1_809654104), T1U, VMUL(LDK(KP425779291), T1T)); Tx = VFNMS(LDK(KP1_688655851), T7, VMUL(LDK(KP535826794), Te)); T1V = VFNMS(LDK(KP851558583), T1U, VMUL(LDK(KP904827052), T1T)); } { V T2V, T2X, T2Y, T2S, T30, T2O, T2R, T31, T2Z; T2V = VMUL(LDK(KP559016994), VSUB(T2T, T2U)); T2X = VADD(T2T, T2U); T2Y = VFNMS(LDK(KP250000000), T2X, T2W); T2O = VSUB(T2M, T2N); T2R = VSUB(T2P, T2Q); T2S = VBYI(VFMA(LDK(KP951056516), T2O, VMUL(LDK(KP587785252), T2R))); T30 = VBYI(VFNMS(LDK(KP951056516), T2R, VMUL(LDK(KP587785252), T2O))); ST(&(xo[0]), VADD(T2W, T2X), ovs, &(xo[0])); T31 = VSUB(T2Y, T2V); ST(&(xo[WS(os, 10)]), VADD(T30, T31), ovs, &(xo[0])); ST(&(xo[WS(os, 15)]), VSUB(T31, T30), ovs, &(xo[WS(os, 1)])); T2Z = VADD(T2V, T2Y); ST(&(xo[WS(os, 5)]), VADD(T2S, T2Z), ovs, &(xo[WS(os, 1)])); ST(&(xo[WS(os, 20)]), VSUB(T2Z, T2S), ovs, &(xo[0])); } { V T1Z, T2i, T2j, T2g, T2w, T2x, T2y, T2G, T2H, T2I, T2D, T2E, T2F, T2z, T2A; V T2B, T2p, T2m, T2q, T2b, T2c, T2a, T2d, T2h, T2r; T1Z = VSUB(T1V, T1Y); T2i = VADD(T20, T21); T2j = VSUB(T25, T28); T2g = VSUB(T2e, T2f); T2w = VFMA(LDK(KP1_369094211), T1W, VMUL(LDK(KP728968627), T1X)); T2x = VFNMS(LDK(KP992114701), T1T, VMUL(LDK(KP250666467), T1U)); T2y = VADD(T2w, T2x); T2G = VFNMS(LDK(KP125581039), T24, VMUL(LDK(KP998026728), T23)); T2H = VFMA(LDK(KP1_274847979), T27, VMUL(LDK(KP770513242), T26)); T2I = VADD(T2G, T2H); T2D = VFNMS(LDK(KP1_457937254), T1W, VMUL(LDK(KP684547105), T1X)); T2E = VFMA(LDK(KP1_984229402), T1U, VMUL(LDK(KP125333233), T1T)); T2F = VADD(T2D, T2E); T2z = VFMA(LDK(KP1_996053456), T24, VMUL(LDK(KP062790519), T23)); T2A = VFNMS(LDK(KP637423989), T26, VMUL(LDK(KP1_541026485), T27)); T2B = VADD(T2z, T2A); { V T2k, T2l, T22, T29; T2k = VADD(T1Y, T1V); T2l = VADD(T2e, T2f); T2p = VADD(T2k, T2l); T2m = VMUL(LDK(KP559016994), VSUB(T2k, T2l)); T2q = VFNMS(LDK(KP250000000), T2p, T2o); T2b = VSUB(T1e, T1b); T22 = VSUB(T20, T21); T29 = VADD(T25, T28); T2c = VADD(T22, T29); T2a = VMUL(LDK(KP559016994), VSUB(T22, T29)); T2d = VFNMS(LDK(KP250000000), T2c, T2b); } { V T2u, T2v, T2C, T2J; T2u = VADD(T2b, T2c); T2v = VBYI(VADD(T2o, T2p)); ST(&(xo[WS(os, 23)]), VSUB(T2u, T2v), ovs, &(xo[WS(os, 1)])); ST(&(xo[WS(os, 2)]), VADD(T2u, T2v), ovs, &(xo[0])); T2C = VADD(T2b, VADD(T2y, T2B)); T2J = VBYI(VSUB(VADD(T2F, T2I), T2o)); ST(&(xo[WS(os, 22)]), VSUB(T2C, T2J), ovs, &(xo[0])); ST(&(xo[WS(os, 3)]), VADD(T2C, T2J), ovs, &(xo[WS(os, 1)])); } T2h = VFMA(LDK(KP951056516), T1Z, VADD(T2a, VFNMS(LDK(KP587785252), T2g, T2d))); T2r = VBYI(VADD(VFMA(LDK(KP951056516), T2i, VMUL(LDK(KP587785252), T2j)), VADD(T2m, T2q))); ST(&(xo[WS(os, 18)]), VSUB(T2h, T2r), ovs, &(xo[0])); ST(&(xo[WS(os, 7)]), VADD(T2h, T2r), ovs, &(xo[WS(os, 1)])); { V T2s, T2t, T2K, T2L; T2s = VFMA(LDK(KP587785252), T1Z, VFMA(LDK(KP951056516), T2g, VSUB(T2d, T2a))); T2t = VBYI(VADD(VFNMS(LDK(KP951056516), T2j, VMUL(LDK(KP587785252), T2i)), VSUB(T2q, T2m))); ST(&(xo[WS(os, 13)]), VSUB(T2s, T2t), ovs, &(xo[WS(os, 1)])); ST(&(xo[WS(os, 12)]), VADD(T2s, T2t), ovs, &(xo[0])); T2K = VBYI(VSUB(VFMA(LDK(KP951056516), VSUB(T2w, T2x), VFMA(LDK(KP309016994), T2F, VFNMS(LDK(KP809016994), T2I, VMUL(LDK(KP587785252), VSUB(T2z, T2A))))), T2o)); T2L = VFMA(LDK(KP309016994), T2y, VFMA(LDK(KP951056516), VSUB(T2E, T2D), VFMA(LDK(KP587785252), VSUB(T2H, T2G), VFNMS(LDK(KP809016994), T2B, T2b)))); ST(&(xo[WS(os, 8)]), VADD(T2K, T2L), ovs, &(xo[0])); ST(&(xo[WS(os, 17)]), VSUB(T2L, T2K), ovs, &(xo[WS(os, 1)])); } } { V Tv, T1m, T1n, T1k, T1D, T1E, T1F, T1N, T1O, T1P, T1K, T1L, T1M, T1G, T1H; V T1I, T1w, T1q, T1x, T1f, T1g, T14, T1h, T1l, T1y; Tv = VSUB(Tf, Tu); T1m = VSUB(Tw, Tx); T1n = VSUB(TN, T12); T1k = VSUB(T1i, T1j); T1D = VFMA(LDK(KP1_688655851), Tm, VMUL(LDK(KP535826794), Tt)); T1E = VFMA(LDK(KP1_541026485), T7, VMUL(LDK(KP637423989), Te)); T1F = VSUB(T1D, T1E); T1N = VFMA(LDK(KP851558583), TM, VMUL(LDK(KP904827052), TJ)); T1O = VFMA(LDK(KP1_984229402), T11, VMUL(LDK(KP125333233), TY)); T1P = VADD(T1N, T1O); T1K = VFNMS(LDK(KP1_071653589), Tm, VMUL(LDK(KP844327925), Tt)); T1L = VFNMS(LDK(KP770513242), Te, VMUL(LDK(KP1_274847979), T7)); T1M = VADD(T1K, T1L); T1G = VFNMS(LDK(KP425779291), TJ, VMUL(LDK(KP1_809654104), TM)); T1H = VFNMS(LDK(KP992114701), TY, VMUL(LDK(KP250666467), T11)); T1I = VADD(T1G, T1H); { V T1o, T1p, Ty, T13; T1o = VADD(Tu, Tf); T1p = VADD(T1i, T1j); T1w = VADD(T1o, T1p); T1q = VMUL(LDK(KP559016994), VSUB(T1o, T1p)); T1x = VFNMS(LDK(KP250000000), T1w, T1v); T1f = VADD(T1b, T1e); Ty = VADD(Tw, Tx); T13 = VADD(TN, T12); T1g = VADD(Ty, T13); T14 = VMUL(LDK(KP559016994), VSUB(Ty, T13)); T1h = VFNMS(LDK(KP250000000), T1g, T1f); } { V T1B, T1C, T1J, T1Q; T1B = VADD(T1f, T1g); T1C = VBYI(VADD(T1v, T1w)); ST(&(xo[WS(os, 24)]), VSUB(T1B, T1C), ovs, &(xo[0])); ST(&(xo[WS(os, 1)]), VADD(T1B, T1C), ovs, &(xo[WS(os, 1)])); T1J = VADD(T1f, VADD(T1F, T1I)); T1Q = VBYI(VSUB(VADD(T1M, T1P), T1v)); ST(&(xo[WS(os, 21)]), VSUB(T1J, T1Q), ovs, &(xo[WS(os, 1)])); ST(&(xo[WS(os, 4)]), VADD(T1J, T1Q), ovs, &(xo[0])); } T1l = VFMA(LDK(KP951056516), Tv, VADD(T14, VFNMS(LDK(KP587785252), T1k, T1h))); T1y = VBYI(VADD(VFMA(LDK(KP951056516), T1m, VMUL(LDK(KP587785252), T1n)), VADD(T1q, T1x))); ST(&(xo[WS(os, 19)]), VSUB(T1l, T1y), ovs, &(xo[WS(os, 1)])); ST(&(xo[WS(os, 6)]), VADD(T1l, T1y), ovs, &(xo[0])); { V T1z, T1A, T1R, T1S; T1z = VFMA(LDK(KP587785252), Tv, VFMA(LDK(KP951056516), T1k, VSUB(T1h, T14))); T1A = VBYI(VADD(VFNMS(LDK(KP951056516), T1n, VMUL(LDK(KP587785252), T1m)), VSUB(T1x, T1q))); ST(&(xo[WS(os, 14)]), VSUB(T1z, T1A), ovs, &(xo[0])); ST(&(xo[WS(os, 11)]), VADD(T1z, T1A), ovs, &(xo[WS(os, 1)])); T1R = VBYI(VSUB(VFMA(LDK(KP309016994), T1M, VFMA(LDK(KP951056516), VADD(T1D, T1E), VFNMS(LDK(KP809016994), T1P, VMUL(LDK(KP587785252), VSUB(T1G, T1H))))), T1v)); T1S = VFMA(LDK(KP951056516), VSUB(T1L, T1K), VFMA(LDK(KP309016994), T1F, VFMA(LDK(KP587785252), VSUB(T1O, T1N), VFNMS(LDK(KP809016994), T1I, T1f)))); ST(&(xo[WS(os, 9)]), VADD(T1R, T1S), ovs, &(xo[WS(os, 1)])); ST(&(xo[WS(os, 16)]), VSUB(T1S, T1R), ovs, &(xo[0])); } } } } VLEAVE(); } static const kdft_desc desc = { 25, XSIMD_STRING("n1bv_25"), {147, 63, 77, 0}, &GENUS, 0, 0, 0, 0 }; void XSIMD(codelet_n1bv_25) (planner *p) { X(kdft_register) (p, n1bv_25, &desc); } #endif /* HAVE_FMA */ fftw-3.3.3/dft/simd/common/n2fv_2.c0000644000175000017500000000666212054410401013651 00000000000000/* * Copyright (c) 2003, 2007-11 Matteo Frigo * Copyright (c) 2003, 2007-11 Massachusetts Institute of Technology * * This program is free software; you can redistribute it and/or modify * it under the terms of the GNU General Public License as published by * the Free Software Foundation; either version 2 of the License, or * (at your option) any later version. * * This program is distributed in the hope that it will be useful, * but WITHOUT ANY WARRANTY; without even the implied warranty of * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the * GNU General Public License for more details. * * You should have received a copy of the GNU General Public License * along with this program; if not, write to the Free Software * Foundation, Inc., 51 Franklin Street, Fifth Floor, Boston, MA 02110-1301 USA * */ /* This file was automatically generated --- DO NOT EDIT */ /* Generated on Sun Nov 25 07:37:21 EST 2012 */ #include "codelet-dft.h" #ifdef HAVE_FMA /* Generated by: ../../../genfft/gen_notw_c.native -fma -reorder-insns -schedule-for-pipeline -simd -compact -variables 4 -pipeline-latency 8 -n 2 -name n2fv_2 -with-ostride 2 -include n2f.h -store-multiple 2 */ /* * This function contains 2 FP additions, 0 FP multiplications, * (or, 2 additions, 0 multiplications, 0 fused multiply/add), * 7 stack variables, 0 constants, and 5 memory accesses */ #include "n2f.h" static void n2fv_2(const R *ri, const R *ii, R *ro, R *io, stride is, stride os, INT v, INT ivs, INT ovs) { { INT i; const R *xi; R *xo; xi = ri; xo = ro; for (i = v; i > 0; i = i - VL, xi = xi + (VL * ivs), xo = xo + (VL * ovs), MAKE_VOLATILE_STRIDE(4, is), MAKE_VOLATILE_STRIDE(4, os)) { V T1, T2, T3, T4; T1 = LD(&(xi[0]), ivs, &(xi[0])); T2 = LD(&(xi[WS(is, 1)]), ivs, &(xi[WS(is, 1)])); T3 = VADD(T1, T2); STM2(&(xo[0]), T3, ovs, &(xo[0])); T4 = VSUB(T1, T2); STM2(&(xo[2]), T4, ovs, &(xo[2])); STN2(&(xo[0]), T3, T4, ovs); } } VLEAVE(); } static const kdft_desc desc = { 2, XSIMD_STRING("n2fv_2"), {2, 0, 0, 0}, &GENUS, 0, 2, 0, 0 }; void XSIMD(codelet_n2fv_2) (planner *p) { X(kdft_register) (p, n2fv_2, &desc); } #else /* HAVE_FMA */ /* Generated by: ../../../genfft/gen_notw_c.native -simd -compact -variables 4 -pipeline-latency 8 -n 2 -name n2fv_2 -with-ostride 2 -include n2f.h -store-multiple 2 */ /* * This function contains 2 FP additions, 0 FP multiplications, * (or, 2 additions, 0 multiplications, 0 fused multiply/add), * 7 stack variables, 0 constants, and 5 memory accesses */ #include "n2f.h" static void n2fv_2(const R *ri, const R *ii, R *ro, R *io, stride is, stride os, INT v, INT ivs, INT ovs) { { INT i; const R *xi; R *xo; xi = ri; xo = ro; for (i = v; i > 0; i = i - VL, xi = xi + (VL * ivs), xo = xo + (VL * ovs), MAKE_VOLATILE_STRIDE(4, is), MAKE_VOLATILE_STRIDE(4, os)) { V T1, T2, T3, T4; T1 = LD(&(xi[0]), ivs, &(xi[0])); T2 = LD(&(xi[WS(is, 1)]), ivs, &(xi[WS(is, 1)])); T3 = VSUB(T1, T2); STM2(&(xo[2]), T3, ovs, &(xo[2])); T4 = VADD(T1, T2); STM2(&(xo[0]), T4, ovs, &(xo[0])); STN2(&(xo[0]), T4, T3, ovs); } } VLEAVE(); } static const kdft_desc desc = { 2, XSIMD_STRING("n2fv_2"), {2, 0, 0, 0}, &GENUS, 0, 2, 0, 0 }; void XSIMD(codelet_n2fv_2) (planner *p) { X(kdft_register) (p, n2fv_2, &desc); } #endif /* HAVE_FMA */ fftw-3.3.3/dft/simd/common/n2fv_4.c0000644000175000017500000001044712054410401013647 00000000000000/* * Copyright (c) 2003, 2007-11 Matteo Frigo * Copyright (c) 2003, 2007-11 Massachusetts Institute of Technology * * This program is free software; you can redistribute it and/or modify * it under the terms of the GNU General Public License as published by * the Free Software Foundation; either version 2 of the License, or * (at your option) any later version. * * This program is distributed in the hope that it will be useful, * but WITHOUT ANY WARRANTY; without even the implied warranty of * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the * GNU General Public License for more details. * * You should have received a copy of the GNU General Public License * along with this program; if not, write to the Free Software * Foundation, Inc., 51 Franklin Street, Fifth Floor, Boston, MA 02110-1301 USA * */ /* This file was automatically generated --- DO NOT EDIT */ /* Generated on Sun Nov 25 07:37:21 EST 2012 */ #include "codelet-dft.h" #ifdef HAVE_FMA /* Generated by: ../../../genfft/gen_notw_c.native -fma -reorder-insns -schedule-for-pipeline -simd -compact -variables 4 -pipeline-latency 8 -n 4 -name n2fv_4 -with-ostride 2 -include n2f.h -store-multiple 2 */ /* * This function contains 8 FP additions, 2 FP multiplications, * (or, 6 additions, 0 multiplications, 2 fused multiply/add), * 15 stack variables, 0 constants, and 10 memory accesses */ #include "n2f.h" static void n2fv_4(const R *ri, const R *ii, R *ro, R *io, stride is, stride os, INT v, INT ivs, INT ovs) { { INT i; const R *xi; R *xo; xi = ri; xo = ro; for (i = v; i > 0; i = i - VL, xi = xi + (VL * ivs), xo = xo + (VL * ovs), MAKE_VOLATILE_STRIDE(8, is), MAKE_VOLATILE_STRIDE(8, os)) { V T1, T2, T4, T5; T1 = LD(&(xi[0]), ivs, &(xi[0])); T2 = LD(&(xi[WS(is, 2)]), ivs, &(xi[0])); T4 = LD(&(xi[WS(is, 1)]), ivs, &(xi[WS(is, 1)])); T5 = LD(&(xi[WS(is, 3)]), ivs, &(xi[WS(is, 1)])); { V T3, T7, T6, T8; T3 = VSUB(T1, T2); T7 = VADD(T1, T2); T6 = VSUB(T4, T5); T8 = VADD(T4, T5); { V T9, Ta, Tb, Tc; T9 = VSUB(T7, T8); STM2(&(xo[4]), T9, ovs, &(xo[0])); Ta = VADD(T7, T8); STM2(&(xo[0]), Ta, ovs, &(xo[0])); Tb = VFMAI(T6, T3); STM2(&(xo[6]), Tb, ovs, &(xo[2])); STN2(&(xo[4]), T9, Tb, ovs); Tc = VFNMSI(T6, T3); STM2(&(xo[2]), Tc, ovs, &(xo[2])); STN2(&(xo[0]), Ta, Tc, ovs); } } } } VLEAVE(); } static const kdft_desc desc = { 4, XSIMD_STRING("n2fv_4"), {6, 0, 2, 0}, &GENUS, 0, 2, 0, 0 }; void XSIMD(codelet_n2fv_4) (planner *p) { X(kdft_register) (p, n2fv_4, &desc); } #else /* HAVE_FMA */ /* Generated by: ../../../genfft/gen_notw_c.native -simd -compact -variables 4 -pipeline-latency 8 -n 4 -name n2fv_4 -with-ostride 2 -include n2f.h -store-multiple 2 */ /* * This function contains 8 FP additions, 0 FP multiplications, * (or, 8 additions, 0 multiplications, 0 fused multiply/add), * 11 stack variables, 0 constants, and 10 memory accesses */ #include "n2f.h" static void n2fv_4(const R *ri, const R *ii, R *ro, R *io, stride is, stride os, INT v, INT ivs, INT ovs) { { INT i; const R *xi; R *xo; xi = ri; xo = ro; for (i = v; i > 0; i = i - VL, xi = xi + (VL * ivs), xo = xo + (VL * ovs), MAKE_VOLATILE_STRIDE(8, is), MAKE_VOLATILE_STRIDE(8, os)) { V T3, T7, T6, T8; { V T1, T2, T4, T5; T1 = LD(&(xi[0]), ivs, &(xi[0])); T2 = LD(&(xi[WS(is, 2)]), ivs, &(xi[0])); T3 = VSUB(T1, T2); T7 = VADD(T1, T2); T4 = LD(&(xi[WS(is, 1)]), ivs, &(xi[WS(is, 1)])); T5 = LD(&(xi[WS(is, 3)]), ivs, &(xi[WS(is, 1)])); T6 = VBYI(VSUB(T4, T5)); T8 = VADD(T4, T5); } { V T9, Ta, Tb, Tc; T9 = VSUB(T3, T6); STM2(&(xo[2]), T9, ovs, &(xo[2])); Ta = VADD(T7, T8); STM2(&(xo[0]), Ta, ovs, &(xo[0])); STN2(&(xo[0]), Ta, T9, ovs); Tb = VADD(T3, T6); STM2(&(xo[6]), Tb, ovs, &(xo[2])); Tc = VSUB(T7, T8); STM2(&(xo[4]), Tc, ovs, &(xo[0])); STN2(&(xo[4]), Tc, Tb, ovs); } } } VLEAVE(); } static const kdft_desc desc = { 4, XSIMD_STRING("n2fv_4"), {8, 0, 0, 0}, &GENUS, 0, 2, 0, 0 }; void XSIMD(codelet_n2fv_4) (planner *p) { X(kdft_register) (p, n2fv_4, &desc); } #endif /* HAVE_FMA */ fftw-3.3.3/dft/simd/common/n2fv_6.c0000644000175000017500000001362712054410401013654 00000000000000/* * Copyright (c) 2003, 2007-11 Matteo Frigo * Copyright (c) 2003, 2007-11 Massachusetts Institute of Technology * * This program is free software; you can redistribute it and/or modify * it under the terms of the GNU General Public License as published by * the Free Software Foundation; either version 2 of the License, or * (at your option) any later version. * * This program is distributed in the hope that it will be useful, * but WITHOUT ANY WARRANTY; without even the implied warranty of * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the * GNU General Public License for more details. * * You should have received a copy of the GNU General Public License * along with this program; if not, write to the Free Software * Foundation, Inc., 51 Franklin Street, Fifth Floor, Boston, MA 02110-1301 USA * */ /* This file was automatically generated --- DO NOT EDIT */ /* Generated on Sun Nov 25 07:37:21 EST 2012 */ #include "codelet-dft.h" #ifdef HAVE_FMA /* Generated by: ../../../genfft/gen_notw_c.native -fma -reorder-insns -schedule-for-pipeline -simd -compact -variables 4 -pipeline-latency 8 -n 6 -name n2fv_6 -with-ostride 2 -include n2f.h -store-multiple 2 */ /* * This function contains 18 FP additions, 8 FP multiplications, * (or, 12 additions, 2 multiplications, 6 fused multiply/add), * 29 stack variables, 2 constants, and 15 memory accesses */ #include "n2f.h" static void n2fv_6(const R *ri, const R *ii, R *ro, R *io, stride is, stride os, INT v, INT ivs, INT ovs) { DVK(KP500000000, +0.500000000000000000000000000000000000000000000); DVK(KP866025403, +0.866025403784438646763723170752936183471402627); { INT i; const R *xi; R *xo; xi = ri; xo = ro; for (i = v; i > 0; i = i - VL, xi = xi + (VL * ivs), xo = xo + (VL * ovs), MAKE_VOLATILE_STRIDE(12, is), MAKE_VOLATILE_STRIDE(12, os)) { V T1, T2, T4, T5, T7, T8; T1 = LD(&(xi[0]), ivs, &(xi[0])); T2 = LD(&(xi[WS(is, 3)]), ivs, &(xi[WS(is, 1)])); T4 = LD(&(xi[WS(is, 2)]), ivs, &(xi[0])); T5 = LD(&(xi[WS(is, 5)]), ivs, &(xi[WS(is, 1)])); T7 = LD(&(xi[WS(is, 4)]), ivs, &(xi[0])); T8 = LD(&(xi[WS(is, 1)]), ivs, &(xi[WS(is, 1)])); { V T3, Td, T6, Te, T9, Tf; T3 = VSUB(T1, T2); Td = VADD(T1, T2); T6 = VSUB(T4, T5); Te = VADD(T4, T5); T9 = VSUB(T7, T8); Tf = VADD(T7, T8); { V Tg, Ti, Ta, Tc; Tg = VADD(Te, Tf); Ti = VMUL(LDK(KP866025403), VSUB(Tf, Te)); Ta = VADD(T6, T9); Tc = VMUL(LDK(KP866025403), VSUB(T9, T6)); { V Th, Tj, Tb, Tk; Th = VFNMS(LDK(KP500000000), Tg, Td); Tj = VADD(Td, Tg); STM2(&(xo[0]), Tj, ovs, &(xo[0])); Tb = VFNMS(LDK(KP500000000), Ta, T3); Tk = VADD(T3, Ta); STM2(&(xo[6]), Tk, ovs, &(xo[2])); { V Tl, Tm, Tn, To; Tl = VFMAI(Ti, Th); STM2(&(xo[8]), Tl, ovs, &(xo[0])); Tm = VFNMSI(Ti, Th); STM2(&(xo[4]), Tm, ovs, &(xo[0])); STN2(&(xo[4]), Tm, Tk, ovs); Tn = VFMAI(Tc, Tb); STM2(&(xo[2]), Tn, ovs, &(xo[2])); STN2(&(xo[0]), Tj, Tn, ovs); To = VFNMSI(Tc, Tb); STM2(&(xo[10]), To, ovs, &(xo[2])); STN2(&(xo[8]), Tl, To, ovs); } } } } } } VLEAVE(); } static const kdft_desc desc = { 6, XSIMD_STRING("n2fv_6"), {12, 2, 6, 0}, &GENUS, 0, 2, 0, 0 }; void XSIMD(codelet_n2fv_6) (planner *p) { X(kdft_register) (p, n2fv_6, &desc); } #else /* HAVE_FMA */ /* Generated by: ../../../genfft/gen_notw_c.native -simd -compact -variables 4 -pipeline-latency 8 -n 6 -name n2fv_6 -with-ostride 2 -include n2f.h -store-multiple 2 */ /* * This function contains 18 FP additions, 4 FP multiplications, * (or, 16 additions, 2 multiplications, 2 fused multiply/add), * 25 stack variables, 2 constants, and 15 memory accesses */ #include "n2f.h" static void n2fv_6(const R *ri, const R *ii, R *ro, R *io, stride is, stride os, INT v, INT ivs, INT ovs) { DVK(KP866025403, +0.866025403784438646763723170752936183471402627); DVK(KP500000000, +0.500000000000000000000000000000000000000000000); { INT i; const R *xi; R *xo; xi = ri; xo = ro; for (i = v; i > 0; i = i - VL, xi = xi + (VL * ivs), xo = xo + (VL * ovs), MAKE_VOLATILE_STRIDE(12, is), MAKE_VOLATILE_STRIDE(12, os)) { V T3, Td, T6, Te, T9, Tf, Ta, Tg, T1, T2, Tj, Tk; T1 = LD(&(xi[0]), ivs, &(xi[0])); T2 = LD(&(xi[WS(is, 3)]), ivs, &(xi[WS(is, 1)])); T3 = VSUB(T1, T2); Td = VADD(T1, T2); { V T4, T5, T7, T8; T4 = LD(&(xi[WS(is, 2)]), ivs, &(xi[0])); T5 = LD(&(xi[WS(is, 5)]), ivs, &(xi[WS(is, 1)])); T6 = VSUB(T4, T5); Te = VADD(T4, T5); T7 = LD(&(xi[WS(is, 4)]), ivs, &(xi[0])); T8 = LD(&(xi[WS(is, 1)]), ivs, &(xi[WS(is, 1)])); T9 = VSUB(T7, T8); Tf = VADD(T7, T8); } Ta = VADD(T6, T9); Tg = VADD(Te, Tf); Tj = VADD(T3, Ta); STM2(&(xo[6]), Tj, ovs, &(xo[2])); Tk = VADD(Td, Tg); STM2(&(xo[0]), Tk, ovs, &(xo[0])); { V Tl, Tb, Tc, Tm; Tb = VFNMS(LDK(KP500000000), Ta, T3); Tc = VBYI(VMUL(LDK(KP866025403), VSUB(T9, T6))); Tl = VSUB(Tb, Tc); STM2(&(xo[10]), Tl, ovs, &(xo[2])); Tm = VADD(Tb, Tc); STM2(&(xo[2]), Tm, ovs, &(xo[2])); STN2(&(xo[0]), Tk, Tm, ovs); { V Th, Ti, Tn, To; Th = VFNMS(LDK(KP500000000), Tg, Td); Ti = VBYI(VMUL(LDK(KP866025403), VSUB(Tf, Te))); Tn = VSUB(Th, Ti); STM2(&(xo[4]), Tn, ovs, &(xo[0])); STN2(&(xo[4]), Tn, Tj, ovs); To = VADD(Th, Ti); STM2(&(xo[8]), To, ovs, &(xo[0])); STN2(&(xo[8]), To, Tl, ovs); } } } } VLEAVE(); } static const kdft_desc desc = { 6, XSIMD_STRING("n2fv_6"), {16, 2, 2, 0}, &GENUS, 0, 2, 0, 0 }; void XSIMD(codelet_n2fv_6) (planner *p) { X(kdft_register) (p, n2fv_6, &desc); } #endif /* HAVE_FMA */ fftw-3.3.3/dft/simd/common/n2fv_8.c0000644000175000017500000001524012054410402013650 00000000000000/* * Copyright (c) 2003, 2007-11 Matteo Frigo * Copyright (c) 2003, 2007-11 Massachusetts Institute of Technology * * This program is free software; you can redistribute it and/or modify * it under the terms of the GNU General Public License as published by * the Free Software Foundation; either version 2 of the License, or * (at your option) any later version. * * This program is distributed in the hope that it will be useful, * but WITHOUT ANY WARRANTY; without even the implied warranty of * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the * GNU General Public License for more details. * * You should have received a copy of the GNU General Public License * along with this program; if not, write to the Free Software * Foundation, Inc., 51 Franklin Street, Fifth Floor, Boston, MA 02110-1301 USA * */ /* This file was automatically generated --- DO NOT EDIT */ /* Generated on Sun Nov 25 07:37:22 EST 2012 */ #include "codelet-dft.h" #ifdef HAVE_FMA /* Generated by: ../../../genfft/gen_notw_c.native -fma -reorder-insns -schedule-for-pipeline -simd -compact -variables 4 -pipeline-latency 8 -n 8 -name n2fv_8 -with-ostride 2 -include n2f.h -store-multiple 2 */ /* * This function contains 26 FP additions, 10 FP multiplications, * (or, 16 additions, 0 multiplications, 10 fused multiply/add), * 38 stack variables, 1 constants, and 20 memory accesses */ #include "n2f.h" static void n2fv_8(const R *ri, const R *ii, R *ro, R *io, stride is, stride os, INT v, INT ivs, INT ovs) { DVK(KP707106781, +0.707106781186547524400844362104849039284835938); { INT i; const R *xi; R *xo; xi = ri; xo = ro; for (i = v; i > 0; i = i - VL, xi = xi + (VL * ivs), xo = xo + (VL * ovs), MAKE_VOLATILE_STRIDE(16, is), MAKE_VOLATILE_STRIDE(16, os)) { V T1, T2, Tc, Td, T4, T5, T7, T8; T1 = LD(&(xi[0]), ivs, &(xi[0])); T2 = LD(&(xi[WS(is, 4)]), ivs, &(xi[0])); Tc = LD(&(xi[WS(is, 2)]), ivs, &(xi[0])); Td = LD(&(xi[WS(is, 6)]), ivs, &(xi[0])); T4 = LD(&(xi[WS(is, 1)]), ivs, &(xi[WS(is, 1)])); T5 = LD(&(xi[WS(is, 5)]), ivs, &(xi[WS(is, 1)])); T7 = LD(&(xi[WS(is, 7)]), ivs, &(xi[WS(is, 1)])); T8 = LD(&(xi[WS(is, 3)]), ivs, &(xi[WS(is, 1)])); { V T3, Tj, Te, Tk, T6, Tm, T9, Tn, Tp, Tl; T3 = VSUB(T1, T2); Tj = VADD(T1, T2); Te = VSUB(Tc, Td); Tk = VADD(Tc, Td); T6 = VSUB(T4, T5); Tm = VADD(T4, T5); T9 = VSUB(T7, T8); Tn = VADD(T7, T8); Tp = VSUB(Tj, Tk); Tl = VADD(Tj, Tk); { V Tq, To, Ta, Tf; Tq = VSUB(Tn, Tm); To = VADD(Tm, Tn); Ta = VADD(T6, T9); Tf = VSUB(T9, T6); { V Tr, Ts, Tt, Tu, Tg, Ti, Tb, Th; Tr = VADD(Tl, To); STM2(&(xo[0]), Tr, ovs, &(xo[0])); Ts = VSUB(Tl, To); STM2(&(xo[8]), Ts, ovs, &(xo[0])); Tt = VFMAI(Tq, Tp); STM2(&(xo[4]), Tt, ovs, &(xo[0])); Tu = VFNMSI(Tq, Tp); STM2(&(xo[12]), Tu, ovs, &(xo[0])); Tg = VFNMS(LDK(KP707106781), Tf, Te); Ti = VFMA(LDK(KP707106781), Tf, Te); Tb = VFMA(LDK(KP707106781), Ta, T3); Th = VFNMS(LDK(KP707106781), Ta, T3); { V Tv, Tw, Tx, Ty; Tv = VFMAI(Ti, Th); STM2(&(xo[6]), Tv, ovs, &(xo[2])); STN2(&(xo[4]), Tt, Tv, ovs); Tw = VFNMSI(Ti, Th); STM2(&(xo[10]), Tw, ovs, &(xo[2])); STN2(&(xo[8]), Ts, Tw, ovs); Tx = VFMAI(Tg, Tb); STM2(&(xo[14]), Tx, ovs, &(xo[2])); STN2(&(xo[12]), Tu, Tx, ovs); Ty = VFNMSI(Tg, Tb); STM2(&(xo[2]), Ty, ovs, &(xo[2])); STN2(&(xo[0]), Tr, Ty, ovs); } } } } } } VLEAVE(); } static const kdft_desc desc = { 8, XSIMD_STRING("n2fv_8"), {16, 0, 10, 0}, &GENUS, 0, 2, 0, 0 }; void XSIMD(codelet_n2fv_8) (planner *p) { X(kdft_register) (p, n2fv_8, &desc); } #else /* HAVE_FMA */ /* Generated by: ../../../genfft/gen_notw_c.native -simd -compact -variables 4 -pipeline-latency 8 -n 8 -name n2fv_8 -with-ostride 2 -include n2f.h -store-multiple 2 */ /* * This function contains 26 FP additions, 2 FP multiplications, * (or, 26 additions, 2 multiplications, 0 fused multiply/add), * 24 stack variables, 1 constants, and 20 memory accesses */ #include "n2f.h" static void n2fv_8(const R *ri, const R *ii, R *ro, R *io, stride is, stride os, INT v, INT ivs, INT ovs) { DVK(KP707106781, +0.707106781186547524400844362104849039284835938); { INT i; const R *xi; R *xo; xi = ri; xo = ro; for (i = v; i > 0; i = i - VL, xi = xi + (VL * ivs), xo = xo + (VL * ovs), MAKE_VOLATILE_STRIDE(16, is), MAKE_VOLATILE_STRIDE(16, os)) { V T3, Tj, Tf, Tk, Ta, Tn, Tc, Tm, Ts, Tu; { V T1, T2, Td, Te; T1 = LD(&(xi[0]), ivs, &(xi[0])); T2 = LD(&(xi[WS(is, 4)]), ivs, &(xi[0])); T3 = VSUB(T1, T2); Tj = VADD(T1, T2); Td = LD(&(xi[WS(is, 2)]), ivs, &(xi[0])); Te = LD(&(xi[WS(is, 6)]), ivs, &(xi[0])); Tf = VSUB(Td, Te); Tk = VADD(Td, Te); { V T4, T5, T6, T7, T8, T9; T4 = LD(&(xi[WS(is, 1)]), ivs, &(xi[WS(is, 1)])); T5 = LD(&(xi[WS(is, 5)]), ivs, &(xi[WS(is, 1)])); T6 = VSUB(T4, T5); T7 = LD(&(xi[WS(is, 7)]), ivs, &(xi[WS(is, 1)])); T8 = LD(&(xi[WS(is, 3)]), ivs, &(xi[WS(is, 1)])); T9 = VSUB(T7, T8); Ta = VMUL(LDK(KP707106781), VADD(T6, T9)); Tn = VADD(T7, T8); Tc = VMUL(LDK(KP707106781), VSUB(T9, T6)); Tm = VADD(T4, T5); } } { V Tr, Tb, Tg, Tp, Tq, Tt; Tb = VADD(T3, Ta); Tg = VBYI(VSUB(Tc, Tf)); Tr = VSUB(Tb, Tg); STM2(&(xo[14]), Tr, ovs, &(xo[2])); Ts = VADD(Tb, Tg); STM2(&(xo[2]), Ts, ovs, &(xo[2])); Tp = VSUB(Tj, Tk); Tq = VBYI(VSUB(Tn, Tm)); Tt = VSUB(Tp, Tq); STM2(&(xo[12]), Tt, ovs, &(xo[0])); STN2(&(xo[12]), Tt, Tr, ovs); Tu = VADD(Tp, Tq); STM2(&(xo[4]), Tu, ovs, &(xo[0])); } { V Tv, Th, Ti, Tw; Th = VSUB(T3, Ta); Ti = VBYI(VADD(Tf, Tc)); Tv = VSUB(Th, Ti); STM2(&(xo[10]), Tv, ovs, &(xo[2])); Tw = VADD(Th, Ti); STM2(&(xo[6]), Tw, ovs, &(xo[2])); STN2(&(xo[4]), Tu, Tw, ovs); { V Tl, To, Tx, Ty; Tl = VADD(Tj, Tk); To = VADD(Tm, Tn); Tx = VSUB(Tl, To); STM2(&(xo[8]), Tx, ovs, &(xo[0])); STN2(&(xo[8]), Tx, Tv, ovs); Ty = VADD(Tl, To); STM2(&(xo[0]), Ty, ovs, &(xo[0])); STN2(&(xo[0]), Ty, Ts, ovs); } } } } VLEAVE(); } static const kdft_desc desc = { 8, XSIMD_STRING("n2fv_8"), {26, 2, 0, 0}, &GENUS, 0, 2, 0, 0 }; void XSIMD(codelet_n2fv_8) (planner *p) { X(kdft_register) (p, n2fv_8, &desc); } #endif /* HAVE_FMA */ fftw-3.3.3/dft/simd/common/n2fv_10.c0000644000175000017500000002206612054410402013725 00000000000000/* * Copyright (c) 2003, 2007-11 Matteo Frigo * Copyright (c) 2003, 2007-11 Massachusetts Institute of Technology * * This program is free software; you can redistribute it and/or modify * it under the terms of the GNU General Public License as published by * the Free Software Foundation; either version 2 of the License, or * (at your option) any later version. * * This program is distributed in the hope that it will be useful, * but WITHOUT ANY WARRANTY; without even the implied warranty of * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the * GNU General Public License for more details. * * You should have received a copy of the GNU General Public License * along with this program; if not, write to the Free Software * Foundation, Inc., 51 Franklin Street, Fifth Floor, Boston, MA 02110-1301 USA * */ /* This file was automatically generated --- DO NOT EDIT */ /* Generated on Sun Nov 25 07:37:22 EST 2012 */ #include "codelet-dft.h" #ifdef HAVE_FMA /* Generated by: ../../../genfft/gen_notw_c.native -fma -reorder-insns -schedule-for-pipeline -simd -compact -variables 4 -pipeline-latency 8 -n 10 -name n2fv_10 -with-ostride 2 -include n2f.h -store-multiple 2 */ /* * This function contains 42 FP additions, 22 FP multiplications, * (or, 24 additions, 4 multiplications, 18 fused multiply/add), * 53 stack variables, 4 constants, and 25 memory accesses */ #include "n2f.h" static void n2fv_10(const R *ri, const R *ii, R *ro, R *io, stride is, stride os, INT v, INT ivs, INT ovs) { DVK(KP559016994, +0.559016994374947424102293417182819058860154590); DVK(KP250000000, +0.250000000000000000000000000000000000000000000); DVK(KP618033988, +0.618033988749894848204586834365638117720309180); DVK(KP951056516, +0.951056516295153572116439333379382143405698634); { INT i; const R *xi; R *xo; xi = ri; xo = ro; for (i = v; i > 0; i = i - VL, xi = xi + (VL * ivs), xo = xo + (VL * ovs), MAKE_VOLATILE_STRIDE(20, is), MAKE_VOLATILE_STRIDE(20, os)) { V Tb, Tr, T3, Ts, T6, Tw, Tg, Tt, T9, Tc, T1, T2; T1 = LD(&(xi[0]), ivs, &(xi[0])); T2 = LD(&(xi[WS(is, 5)]), ivs, &(xi[WS(is, 1)])); { V T4, T5, Te, Tf, T7, T8; T4 = LD(&(xi[WS(is, 2)]), ivs, &(xi[0])); T5 = LD(&(xi[WS(is, 7)]), ivs, &(xi[WS(is, 1)])); Te = LD(&(xi[WS(is, 6)]), ivs, &(xi[0])); Tf = LD(&(xi[WS(is, 1)]), ivs, &(xi[WS(is, 1)])); T7 = LD(&(xi[WS(is, 8)]), ivs, &(xi[0])); T8 = LD(&(xi[WS(is, 3)]), ivs, &(xi[WS(is, 1)])); Tb = LD(&(xi[WS(is, 4)]), ivs, &(xi[0])); Tr = VADD(T1, T2); T3 = VSUB(T1, T2); Ts = VADD(T4, T5); T6 = VSUB(T4, T5); Tw = VADD(Te, Tf); Tg = VSUB(Te, Tf); Tt = VADD(T7, T8); T9 = VSUB(T7, T8); Tc = LD(&(xi[WS(is, 9)]), ivs, &(xi[WS(is, 1)])); } { V TD, Tu, Tm, Ta, Td, Tv; TD = VSUB(Ts, Tt); Tu = VADD(Ts, Tt); Tm = VSUB(T6, T9); Ta = VADD(T6, T9); Td = VSUB(Tb, Tc); Tv = VADD(Tb, Tc); { V TC, Tx, Tn, Th; TC = VSUB(Tv, Tw); Tx = VADD(Tv, Tw); Tn = VSUB(Td, Tg); Th = VADD(Td, Tg); { V Ty, TA, TE, TG, Ti, Tk, To, Tq; Ty = VADD(Tu, Tx); TA = VSUB(Tu, Tx); TE = VMUL(LDK(KP951056516), VFNMS(LDK(KP618033988), TD, TC)); TG = VMUL(LDK(KP951056516), VFMA(LDK(KP618033988), TC, TD)); Ti = VADD(Ta, Th); Tk = VSUB(Ta, Th); To = VMUL(LDK(KP951056516), VFMA(LDK(KP618033988), Tn, Tm)); Tq = VMUL(LDK(KP951056516), VFNMS(LDK(KP618033988), Tm, Tn)); { V Tz, TH, Tj, TI; Tz = VFNMS(LDK(KP250000000), Ty, Tr); TH = VADD(Tr, Ty); STM2(&(xo[0]), TH, ovs, &(xo[0])); Tj = VFNMS(LDK(KP250000000), Ti, T3); TI = VADD(T3, Ti); STM2(&(xo[10]), TI, ovs, &(xo[2])); { V TB, TF, Tl, Tp; TB = VFNMS(LDK(KP559016994), TA, Tz); TF = VFMA(LDK(KP559016994), TA, Tz); Tl = VFMA(LDK(KP559016994), Tk, Tj); Tp = VFNMS(LDK(KP559016994), Tk, Tj); { V TJ, TK, TL, TM; TJ = VFMAI(TG, TF); STM2(&(xo[8]), TJ, ovs, &(xo[0])); STN2(&(xo[8]), TJ, TI, ovs); TK = VFNMSI(TG, TF); STM2(&(xo[12]), TK, ovs, &(xo[0])); TL = VFNMSI(TE, TB); STM2(&(xo[16]), TL, ovs, &(xo[0])); TM = VFMAI(TE, TB); STM2(&(xo[4]), TM, ovs, &(xo[0])); { V TN, TO, TP, TQ; TN = VFNMSI(Tq, Tp); STM2(&(xo[6]), TN, ovs, &(xo[2])); STN2(&(xo[4]), TM, TN, ovs); TO = VFMAI(Tq, Tp); STM2(&(xo[14]), TO, ovs, &(xo[2])); STN2(&(xo[12]), TK, TO, ovs); TP = VFMAI(To, Tl); STM2(&(xo[18]), TP, ovs, &(xo[2])); STN2(&(xo[16]), TL, TP, ovs); TQ = VFNMSI(To, Tl); STM2(&(xo[2]), TQ, ovs, &(xo[2])); STN2(&(xo[0]), TH, TQ, ovs); } } } } } } } } } VLEAVE(); } static const kdft_desc desc = { 10, XSIMD_STRING("n2fv_10"), {24, 4, 18, 0}, &GENUS, 0, 2, 0, 0 }; void XSIMD(codelet_n2fv_10) (planner *p) { X(kdft_register) (p, n2fv_10, &desc); } #else /* HAVE_FMA */ /* Generated by: ../../../genfft/gen_notw_c.native -simd -compact -variables 4 -pipeline-latency 8 -n 10 -name n2fv_10 -with-ostride 2 -include n2f.h -store-multiple 2 */ /* * This function contains 42 FP additions, 12 FP multiplications, * (or, 36 additions, 6 multiplications, 6 fused multiply/add), * 36 stack variables, 4 constants, and 25 memory accesses */ #include "n2f.h" static void n2fv_10(const R *ri, const R *ii, R *ro, R *io, stride is, stride os, INT v, INT ivs, INT ovs) { DVK(KP250000000, +0.250000000000000000000000000000000000000000000); DVK(KP559016994, +0.559016994374947424102293417182819058860154590); DVK(KP587785252, +0.587785252292473129168705954639072768597652438); DVK(KP951056516, +0.951056516295153572116439333379382143405698634); { INT i; const R *xi; R *xo; xi = ri; xo = ro; for (i = v; i > 0; i = i - VL, xi = xi + (VL * ivs), xo = xo + (VL * ovs), MAKE_VOLATILE_STRIDE(20, is), MAKE_VOLATILE_STRIDE(20, os)) { V Ti, Ty, Tm, Tn, Tw, Tt, Tz, TA, TB, T7, Te, Tj, Tg, Th; Tg = LD(&(xi[0]), ivs, &(xi[0])); Th = LD(&(xi[WS(is, 5)]), ivs, &(xi[WS(is, 1)])); Ti = VSUB(Tg, Th); Ty = VADD(Tg, Th); { V T3, Tu, Td, Ts, T6, Tv, Ta, Tr; { V T1, T2, Tb, Tc; T1 = LD(&(xi[WS(is, 2)]), ivs, &(xi[0])); T2 = LD(&(xi[WS(is, 7)]), ivs, &(xi[WS(is, 1)])); T3 = VSUB(T1, T2); Tu = VADD(T1, T2); Tb = LD(&(xi[WS(is, 6)]), ivs, &(xi[0])); Tc = LD(&(xi[WS(is, 1)]), ivs, &(xi[WS(is, 1)])); Td = VSUB(Tb, Tc); Ts = VADD(Tb, Tc); } { V T4, T5, T8, T9; T4 = LD(&(xi[WS(is, 8)]), ivs, &(xi[0])); T5 = LD(&(xi[WS(is, 3)]), ivs, &(xi[WS(is, 1)])); T6 = VSUB(T4, T5); Tv = VADD(T4, T5); T8 = LD(&(xi[WS(is, 4)]), ivs, &(xi[0])); T9 = LD(&(xi[WS(is, 9)]), ivs, &(xi[WS(is, 1)])); Ta = VSUB(T8, T9); Tr = VADD(T8, T9); } Tm = VSUB(T3, T6); Tn = VSUB(Ta, Td); Tw = VSUB(Tu, Tv); Tt = VSUB(Tr, Ts); Tz = VADD(Tu, Tv); TA = VADD(Tr, Ts); TB = VADD(Tz, TA); T7 = VADD(T3, T6); Te = VADD(Ta, Td); Tj = VADD(T7, Te); } { V TH, TI, TK, TL, TM; TH = VADD(Ti, Tj); STM2(&(xo[10]), TH, ovs, &(xo[2])); TI = VADD(Ty, TB); STM2(&(xo[0]), TI, ovs, &(xo[0])); { V To, Tq, Tl, Tp, Tf, Tk, TJ; To = VBYI(VFMA(LDK(KP951056516), Tm, VMUL(LDK(KP587785252), Tn))); Tq = VBYI(VFNMS(LDK(KP587785252), Tm, VMUL(LDK(KP951056516), Tn))); Tf = VMUL(LDK(KP559016994), VSUB(T7, Te)); Tk = VFNMS(LDK(KP250000000), Tj, Ti); Tl = VADD(Tf, Tk); Tp = VSUB(Tk, Tf); TJ = VSUB(Tl, To); STM2(&(xo[2]), TJ, ovs, &(xo[2])); STN2(&(xo[0]), TI, TJ, ovs); TK = VADD(Tq, Tp); STM2(&(xo[14]), TK, ovs, &(xo[2])); TL = VADD(To, Tl); STM2(&(xo[18]), TL, ovs, &(xo[2])); TM = VSUB(Tp, Tq); STM2(&(xo[6]), TM, ovs, &(xo[2])); } { V Tx, TF, TE, TG, TC, TD; Tx = VBYI(VFNMS(LDK(KP587785252), Tw, VMUL(LDK(KP951056516), Tt))); TF = VBYI(VFMA(LDK(KP951056516), Tw, VMUL(LDK(KP587785252), Tt))); TC = VFNMS(LDK(KP250000000), TB, Ty); TD = VMUL(LDK(KP559016994), VSUB(Tz, TA)); TE = VSUB(TC, TD); TG = VADD(TD, TC); { V TN, TO, TP, TQ; TN = VADD(Tx, TE); STM2(&(xo[4]), TN, ovs, &(xo[0])); STN2(&(xo[4]), TN, TM, ovs); TO = VSUB(TG, TF); STM2(&(xo[12]), TO, ovs, &(xo[0])); STN2(&(xo[12]), TO, TK, ovs); TP = VSUB(TE, Tx); STM2(&(xo[16]), TP, ovs, &(xo[0])); STN2(&(xo[16]), TP, TL, ovs); TQ = VADD(TF, TG); STM2(&(xo[8]), TQ, ovs, &(xo[0])); STN2(&(xo[8]), TQ, TH, ovs); } } } } } VLEAVE(); } static const kdft_desc desc = { 10, XSIMD_STRING("n2fv_10"), {36, 6, 6, 0}, &GENUS, 0, 2, 0, 0 }; void XSIMD(codelet_n2fv_10) (planner *p) { X(kdft_register) (p, n2fv_10, &desc); } #endif /* HAVE_FMA */ fftw-3.3.3/dft/simd/common/n2fv_12.c0000644000175000017500000002301312054410402013720 00000000000000/* * Copyright (c) 2003, 2007-11 Matteo Frigo * Copyright (c) 2003, 2007-11 Massachusetts Institute of Technology * * This program is free software; you can redistribute it and/or modify * it under the terms of the GNU General Public License as published by * the Free Software Foundation; either version 2 of the License, or * (at your option) any later version. * * This program is distributed in the hope that it will be useful, * but WITHOUT ANY WARRANTY; without even the implied warranty of * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the * GNU General Public License for more details. * * You should have received a copy of the GNU General Public License * along with this program; if not, write to the Free Software * Foundation, Inc., 51 Franklin Street, Fifth Floor, Boston, MA 02110-1301 USA * */ /* This file was automatically generated --- DO NOT EDIT */ /* Generated on Sun Nov 25 07:37:22 EST 2012 */ #include "codelet-dft.h" #ifdef HAVE_FMA /* Generated by: ../../../genfft/gen_notw_c.native -fma -reorder-insns -schedule-for-pipeline -simd -compact -variables 4 -pipeline-latency 8 -n 12 -name n2fv_12 -with-ostride 2 -include n2f.h -store-multiple 2 */ /* * This function contains 48 FP additions, 20 FP multiplications, * (or, 30 additions, 2 multiplications, 18 fused multiply/add), * 61 stack variables, 2 constants, and 30 memory accesses */ #include "n2f.h" static void n2fv_12(const R *ri, const R *ii, R *ro, R *io, stride is, stride os, INT v, INT ivs, INT ovs) { DVK(KP866025403, +0.866025403784438646763723170752936183471402627); DVK(KP500000000, +0.500000000000000000000000000000000000000000000); { INT i; const R *xi; R *xo; xi = ri; xo = ro; for (i = v; i > 0; i = i - VL, xi = xi + (VL * ivs), xo = xo + (VL * ovs), MAKE_VOLATILE_STRIDE(24, is), MAKE_VOLATILE_STRIDE(24, os)) { V T1, T6, Tk, Tn, Tc, Td, Tf, Tr, T4, Ts, T9, Tg, Te, Tl; { V T2, T3, T7, T8; T1 = LD(&(xi[0]), ivs, &(xi[0])); T6 = LD(&(xi[WS(is, 6)]), ivs, &(xi[0])); T2 = LD(&(xi[WS(is, 4)]), ivs, &(xi[0])); T3 = LD(&(xi[WS(is, 8)]), ivs, &(xi[0])); T7 = LD(&(xi[WS(is, 10)]), ivs, &(xi[0])); T8 = LD(&(xi[WS(is, 2)]), ivs, &(xi[0])); Tk = LD(&(xi[WS(is, 3)]), ivs, &(xi[WS(is, 1)])); Tn = LD(&(xi[WS(is, 9)]), ivs, &(xi[WS(is, 1)])); Tc = LD(&(xi[WS(is, 11)]), ivs, &(xi[WS(is, 1)])); Td = LD(&(xi[WS(is, 7)]), ivs, &(xi[WS(is, 1)])); Tf = LD(&(xi[WS(is, 1)]), ivs, &(xi[WS(is, 1)])); Tr = VSUB(T3, T2); T4 = VADD(T2, T3); Ts = VSUB(T8, T7); T9 = VADD(T7, T8); Tg = LD(&(xi[WS(is, 5)]), ivs, &(xi[WS(is, 1)])); } Te = VSUB(Tc, Td); Tl = VADD(Td, Tc); { V T5, TF, TB, Tt, Ta, TG, Th, To, Tm, TI; T5 = VFNMS(LDK(KP500000000), T4, T1); TF = VADD(T1, T4); TB = VADD(Tr, Ts); Tt = VSUB(Tr, Ts); Ta = VFNMS(LDK(KP500000000), T9, T6); TG = VADD(T6, T9); Th = VSUB(Tf, Tg); To = VADD(Tf, Tg); Tm = VFNMS(LDK(KP500000000), Tl, Tk); TI = VADD(Tk, Tl); { V TH, TL, Tb, Tx, TJ, Tp, Ti, TA; TH = VSUB(TF, TG); TL = VADD(TF, TG); Tb = VSUB(T5, Ta); Tx = VADD(T5, Ta); TJ = VADD(Tn, To); Tp = VFNMS(LDK(KP500000000), To, Tn); Ti = VADD(Te, Th); TA = VSUB(Te, Th); { V Tq, Ty, TK, TM; Tq = VSUB(Tm, Tp); Ty = VADD(Tm, Tp); TK = VSUB(TI, TJ); TM = VADD(TI, TJ); { V TC, TE, Tj, Tv; TC = VMUL(LDK(KP866025403), VSUB(TA, TB)); TE = VMUL(LDK(KP866025403), VADD(TB, TA)); Tj = VFMA(LDK(KP866025403), Ti, Tb); Tv = VFNMS(LDK(KP866025403), Ti, Tb); { V Tz, TD, Tu, Tw; Tz = VSUB(Tx, Ty); TD = VADD(Tx, Ty); Tu = VFNMS(LDK(KP866025403), Tt, Tq); Tw = VFMA(LDK(KP866025403), Tt, Tq); { V TN, TO, TP, TQ; TN = VADD(TL, TM); STM2(&(xo[0]), TN, ovs, &(xo[0])); TO = VSUB(TL, TM); STM2(&(xo[12]), TO, ovs, &(xo[0])); TP = VFMAI(TK, TH); STM2(&(xo[6]), TP, ovs, &(xo[2])); TQ = VFNMSI(TK, TH); STM2(&(xo[18]), TQ, ovs, &(xo[2])); { V TR, TS, TT, TU; TR = VFMAI(TE, TD); STM2(&(xo[8]), TR, ovs, &(xo[0])); TS = VFNMSI(TE, TD); STM2(&(xo[16]), TS, ovs, &(xo[0])); STN2(&(xo[16]), TS, TQ, ovs); TT = VFNMSI(TC, Tz); STM2(&(xo[20]), TT, ovs, &(xo[0])); TU = VFMAI(TC, Tz); STM2(&(xo[4]), TU, ovs, &(xo[0])); STN2(&(xo[4]), TU, TP, ovs); { V TV, TW, TX, TY; TV = VFNMSI(Tw, Tv); STM2(&(xo[10]), TV, ovs, &(xo[2])); STN2(&(xo[8]), TR, TV, ovs); TW = VFMAI(Tw, Tv); STM2(&(xo[14]), TW, ovs, &(xo[2])); STN2(&(xo[12]), TO, TW, ovs); TX = VFMAI(Tu, Tj); STM2(&(xo[22]), TX, ovs, &(xo[2])); STN2(&(xo[20]), TT, TX, ovs); TY = VFNMSI(Tu, Tj); STM2(&(xo[2]), TY, ovs, &(xo[2])); STN2(&(xo[0]), TN, TY, ovs); } } } } } } } } } } VLEAVE(); } static const kdft_desc desc = { 12, XSIMD_STRING("n2fv_12"), {30, 2, 18, 0}, &GENUS, 0, 2, 0, 0 }; void XSIMD(codelet_n2fv_12) (planner *p) { X(kdft_register) (p, n2fv_12, &desc); } #else /* HAVE_FMA */ /* Generated by: ../../../genfft/gen_notw_c.native -simd -compact -variables 4 -pipeline-latency 8 -n 12 -name n2fv_12 -with-ostride 2 -include n2f.h -store-multiple 2 */ /* * This function contains 48 FP additions, 8 FP multiplications, * (or, 44 additions, 4 multiplications, 4 fused multiply/add), * 33 stack variables, 2 constants, and 30 memory accesses */ #include "n2f.h" static void n2fv_12(const R *ri, const R *ii, R *ro, R *io, stride is, stride os, INT v, INT ivs, INT ovs) { DVK(KP500000000, +0.500000000000000000000000000000000000000000000); DVK(KP866025403, +0.866025403784438646763723170752936183471402627); { INT i; const R *xi; R *xo; xi = ri; xo = ro; for (i = v; i > 0; i = i - VL, xi = xi + (VL * ivs), xo = xo + (VL * ovs), MAKE_VOLATILE_STRIDE(24, is), MAKE_VOLATILE_STRIDE(24, os)) { V T5, Ta, TJ, Ty, Tq, Tp, Tg, Tl, TI, TA, Tz, Tu; { V T1, T6, T4, Tw, T9, Tx; T1 = LD(&(xi[0]), ivs, &(xi[0])); T6 = LD(&(xi[WS(is, 6)]), ivs, &(xi[0])); { V T2, T3, T7, T8; T2 = LD(&(xi[WS(is, 4)]), ivs, &(xi[0])); T3 = LD(&(xi[WS(is, 8)]), ivs, &(xi[0])); T4 = VADD(T2, T3); Tw = VSUB(T3, T2); T7 = LD(&(xi[WS(is, 10)]), ivs, &(xi[0])); T8 = LD(&(xi[WS(is, 2)]), ivs, &(xi[0])); T9 = VADD(T7, T8); Tx = VSUB(T8, T7); } T5 = VADD(T1, T4); Ta = VADD(T6, T9); TJ = VADD(Tw, Tx); Ty = VMUL(LDK(KP866025403), VSUB(Tw, Tx)); Tq = VFNMS(LDK(KP500000000), T9, T6); Tp = VFNMS(LDK(KP500000000), T4, T1); } { V Tc, Th, Tf, Ts, Tk, Tt; Tc = LD(&(xi[WS(is, 3)]), ivs, &(xi[WS(is, 1)])); Th = LD(&(xi[WS(is, 9)]), ivs, &(xi[WS(is, 1)])); { V Td, Te, Ti, Tj; Td = LD(&(xi[WS(is, 7)]), ivs, &(xi[WS(is, 1)])); Te = LD(&(xi[WS(is, 11)]), ivs, &(xi[WS(is, 1)])); Tf = VADD(Td, Te); Ts = VSUB(Te, Td); Ti = LD(&(xi[WS(is, 1)]), ivs, &(xi[WS(is, 1)])); Tj = LD(&(xi[WS(is, 5)]), ivs, &(xi[WS(is, 1)])); Tk = VADD(Ti, Tj); Tt = VSUB(Tj, Ti); } Tg = VADD(Tc, Tf); Tl = VADD(Th, Tk); TI = VADD(Ts, Tt); TA = VFNMS(LDK(KP500000000), Tk, Th); Tz = VFNMS(LDK(KP500000000), Tf, Tc); Tu = VMUL(LDK(KP866025403), VSUB(Ts, Tt)); } { V TN, TO, TP, TQ, TR, TS; { V Tb, Tm, Tn, To; Tb = VSUB(T5, Ta); Tm = VBYI(VSUB(Tg, Tl)); TN = VSUB(Tb, Tm); STM2(&(xo[18]), TN, ovs, &(xo[2])); TO = VADD(Tb, Tm); STM2(&(xo[6]), TO, ovs, &(xo[2])); Tn = VADD(T5, Ta); To = VADD(Tg, Tl); TP = VSUB(Tn, To); STM2(&(xo[12]), TP, ovs, &(xo[0])); TQ = VADD(Tn, To); STM2(&(xo[0]), TQ, ovs, &(xo[0])); } { V Tv, TE, TC, TD, Tr, TB, TT, TU; Tr = VSUB(Tp, Tq); Tv = VSUB(Tr, Tu); TE = VADD(Tr, Tu); TB = VSUB(Tz, TA); TC = VBYI(VADD(Ty, TB)); TD = VBYI(VSUB(Ty, TB)); TR = VSUB(Tv, TC); STM2(&(xo[10]), TR, ovs, &(xo[2])); TS = VSUB(TE, TD); STM2(&(xo[22]), TS, ovs, &(xo[2])); TT = VADD(TC, Tv); STM2(&(xo[14]), TT, ovs, &(xo[2])); STN2(&(xo[12]), TP, TT, ovs); TU = VADD(TD, TE); STM2(&(xo[2]), TU, ovs, &(xo[2])); STN2(&(xo[0]), TQ, TU, ovs); } { V TK, TM, TH, TL, TF, TG; TK = VBYI(VMUL(LDK(KP866025403), VSUB(TI, TJ))); TM = VBYI(VMUL(LDK(KP866025403), VADD(TJ, TI))); TF = VADD(Tp, Tq); TG = VADD(Tz, TA); TH = VSUB(TF, TG); TL = VADD(TF, TG); { V TV, TW, TX, TY; TV = VSUB(TH, TK); STM2(&(xo[20]), TV, ovs, &(xo[0])); STN2(&(xo[20]), TV, TS, ovs); TW = VADD(TL, TM); STM2(&(xo[8]), TW, ovs, &(xo[0])); STN2(&(xo[8]), TW, TR, ovs); TX = VADD(TH, TK); STM2(&(xo[4]), TX, ovs, &(xo[0])); STN2(&(xo[4]), TX, TO, ovs); TY = VSUB(TL, TM); STM2(&(xo[16]), TY, ovs, &(xo[0])); STN2(&(xo[16]), TY, TN, ovs); } } } } } VLEAVE(); } static const kdft_desc desc = { 12, XSIMD_STRING("n2fv_12"), {44, 4, 4, 0}, &GENUS, 0, 2, 0, 0 }; void XSIMD(codelet_n2fv_12) (planner *p) { X(kdft_register) (p, n2fv_12, &desc); } #endif /* HAVE_FMA */ fftw-3.3.3/dft/simd/common/n2fv_14.c0000644000175000017500000003133412054410403013730 00000000000000/* * Copyright (c) 2003, 2007-11 Matteo Frigo * Copyright (c) 2003, 2007-11 Massachusetts Institute of Technology * * This program is free software; you can redistribute it and/or modify * it under the terms of the GNU General Public License as published by * the Free Software Foundation; either version 2 of the License, or * (at your option) any later version. * * This program is distributed in the hope that it will be useful, * but WITHOUT ANY WARRANTY; without even the implied warranty of * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the * GNU General Public License for more details. * * You should have received a copy of the GNU General Public License * along with this program; if not, write to the Free Software * Foundation, Inc., 51 Franklin Street, Fifth Floor, Boston, MA 02110-1301 USA * */ /* This file was automatically generated --- DO NOT EDIT */ /* Generated on Sun Nov 25 07:37:22 EST 2012 */ #include "codelet-dft.h" #ifdef HAVE_FMA /* Generated by: ../../../genfft/gen_notw_c.native -fma -reorder-insns -schedule-for-pipeline -simd -compact -variables 4 -pipeline-latency 8 -n 14 -name n2fv_14 -with-ostride 2 -include n2f.h -store-multiple 2 */ /* * This function contains 74 FP additions, 48 FP multiplications, * (or, 32 additions, 6 multiplications, 42 fused multiply/add), * 65 stack variables, 6 constants, and 35 memory accesses */ #include "n2f.h" static void n2fv_14(const R *ri, const R *ii, R *ro, R *io, stride is, stride os, INT v, INT ivs, INT ovs) { DVK(KP900968867, +0.900968867902419126236102319507445051165919162); DVK(KP801937735, +0.801937735804838252472204639014890102331838324); DVK(KP974927912, +0.974927912181823607018131682993931217232785801); DVK(KP692021471, +0.692021471630095869627814897002069140197260599); DVK(KP554958132, +0.554958132087371191422194871006410481067288862); DVK(KP356895867, +0.356895867892209443894399510021300583399127187); { INT i; const R *xi; R *xo; xi = ri; xo = ro; for (i = v; i > 0; i = i - VL, xi = xi + (VL * ivs), xo = xo + (VL * ovs), MAKE_VOLATILE_STRIDE(28, is), MAKE_VOLATILE_STRIDE(28, os)) { V TH, T3, TP, Tn, Ta, Ts, TW, TK, TO, Tk, TM, Tg, TL, Td, T1; V T2; T1 = LD(&(xi[0]), ivs, &(xi[0])); T2 = LD(&(xi[WS(is, 7)]), ivs, &(xi[WS(is, 1)])); { V Ti, TI, T6, TJ, T9, Tj, Te, Tf, Tb, Tc; { V T4, T5, T7, T8, Tl, Tm; T4 = LD(&(xi[WS(is, 2)]), ivs, &(xi[0])); T5 = LD(&(xi[WS(is, 9)]), ivs, &(xi[WS(is, 1)])); T7 = LD(&(xi[WS(is, 12)]), ivs, &(xi[0])); T8 = LD(&(xi[WS(is, 5)]), ivs, &(xi[WS(is, 1)])); Tl = LD(&(xi[WS(is, 8)]), ivs, &(xi[0])); Tm = LD(&(xi[WS(is, 1)]), ivs, &(xi[WS(is, 1)])); Ti = LD(&(xi[WS(is, 6)]), ivs, &(xi[0])); TH = VADD(T1, T2); T3 = VSUB(T1, T2); TI = VADD(T4, T5); T6 = VSUB(T4, T5); TJ = VADD(T7, T8); T9 = VSUB(T7, T8); TP = VADD(Tl, Tm); Tn = VSUB(Tl, Tm); Tj = LD(&(xi[WS(is, 13)]), ivs, &(xi[WS(is, 1)])); Te = LD(&(xi[WS(is, 10)]), ivs, &(xi[0])); Tf = LD(&(xi[WS(is, 3)]), ivs, &(xi[WS(is, 1)])); Tb = LD(&(xi[WS(is, 4)]), ivs, &(xi[0])); Tc = LD(&(xi[WS(is, 11)]), ivs, &(xi[WS(is, 1)])); } Ta = VADD(T6, T9); Ts = VSUB(T9, T6); TW = VSUB(TJ, TI); TK = VADD(TI, TJ); TO = VADD(Ti, Tj); Tk = VSUB(Ti, Tj); TM = VADD(Te, Tf); Tg = VSUB(Te, Tf); TL = VADD(Tb, Tc); Td = VSUB(Tb, Tc); } { V T19, T1a, T18, TB, T13, TY, TG, Tw, T11, Tr, T16, TT, Tz, TE, TU; V TQ; TU = VSUB(TO, TP); TQ = VADD(TO, TP); { V Tt, To, TV, TN; Tt = VSUB(Tn, Tk); To = VADD(Tk, Tn); TV = VSUB(TL, TM); TN = VADD(TL, TM); { V Tu, Th, TZ, T17; Tu = VSUB(Tg, Td); Th = VADD(Td, Tg); TZ = VFNMS(LDK(KP356895867), TK, TQ); T17 = VFNMS(LDK(KP554958132), TU, TW); { V Tp, TA, T14, TR; Tp = VFNMS(LDK(KP356895867), Ta, To); TA = VFMA(LDK(KP554958132), Tt, Ts); T19 = VADD(TH, VADD(TK, VADD(TN, TQ))); STM2(&(xo[0]), T19, ovs, &(xo[0])); T14 = VFNMS(LDK(KP356895867), TN, TK); TR = VFNMS(LDK(KP356895867), TQ, TN); { V T12, TX, Tx, TC; T12 = VFMA(LDK(KP554958132), TV, TU); TX = VFMA(LDK(KP554958132), TW, TV); T1a = VADD(T3, VADD(Ta, VADD(Th, To))); STM2(&(xo[14]), T1a, ovs, &(xo[2])); Tx = VFNMS(LDK(KP356895867), Th, Ta); TC = VFNMS(LDK(KP356895867), To, Th); { V TF, Tv, T10, Tq; TF = VFNMS(LDK(KP554958132), Ts, Tu); Tv = VFMA(LDK(KP554958132), Tu, Tt); T10 = VFNMS(LDK(KP692021471), TZ, TN); T18 = VMUL(LDK(KP974927912), VFNMS(LDK(KP801937735), T17, TV)); Tq = VFNMS(LDK(KP692021471), Tp, Th); TB = VMUL(LDK(KP974927912), VFMA(LDK(KP801937735), TA, Tu)); { V T15, TS, Ty, TD; T15 = VFNMS(LDK(KP692021471), T14, TQ); TS = VFNMS(LDK(KP692021471), TR, TK); T13 = VMUL(LDK(KP974927912), VFMA(LDK(KP801937735), T12, TW)); TY = VMUL(LDK(KP974927912), VFNMS(LDK(KP801937735), TX, TU)); Ty = VFNMS(LDK(KP692021471), Tx, To); TD = VFNMS(LDK(KP692021471), TC, Ta); TG = VMUL(LDK(KP974927912), VFNMS(LDK(KP801937735), TF, Tt)); Tw = VMUL(LDK(KP974927912), VFNMS(LDK(KP801937735), Tv, Ts)); T11 = VFNMS(LDK(KP900968867), T10, TH); Tr = VFNMS(LDK(KP900968867), Tq, T3); T16 = VFNMS(LDK(KP900968867), T15, TH); TT = VFNMS(LDK(KP900968867), TS, TH); Tz = VFNMS(LDK(KP900968867), Ty, T3); TE = VFNMS(LDK(KP900968867), TD, T3); } } } } } } { V T1b, T1c, T1d, T1e; T1b = VFNMSI(T13, T11); STM2(&(xo[24]), T1b, ovs, &(xo[0])); T1c = VFMAI(T13, T11); STM2(&(xo[4]), T1c, ovs, &(xo[0])); T1d = VFMAI(Tw, Tr); STM2(&(xo[18]), T1d, ovs, &(xo[2])); T1e = VFNMSI(Tw, Tr); STM2(&(xo[10]), T1e, ovs, &(xo[2])); { V T1f, T1g, T1h, T1i; T1f = VFNMSI(T18, T16); STM2(&(xo[16]), T1f, ovs, &(xo[0])); STN2(&(xo[16]), T1f, T1d, ovs); T1g = VFMAI(T18, T16); STM2(&(xo[12]), T1g, ovs, &(xo[0])); STN2(&(xo[12]), T1g, T1a, ovs); T1h = VFNMSI(TY, TT); STM2(&(xo[20]), T1h, ovs, &(xo[0])); T1i = VFMAI(TY, TT); STM2(&(xo[8]), T1i, ovs, &(xo[0])); STN2(&(xo[8]), T1i, T1e, ovs); { V T1j, T1k, T1l, T1m; T1j = VFMAI(TB, Tz); STM2(&(xo[2]), T1j, ovs, &(xo[2])); STN2(&(xo[0]), T19, T1j, ovs); T1k = VFNMSI(TB, Tz); STM2(&(xo[26]), T1k, ovs, &(xo[2])); STN2(&(xo[24]), T1b, T1k, ovs); T1l = VFMAI(TG, TE); STM2(&(xo[6]), T1l, ovs, &(xo[2])); STN2(&(xo[4]), T1c, T1l, ovs); T1m = VFNMSI(TG, TE); STM2(&(xo[22]), T1m, ovs, &(xo[2])); STN2(&(xo[20]), T1h, T1m, ovs); } } } } } } VLEAVE(); } static const kdft_desc desc = { 14, XSIMD_STRING("n2fv_14"), {32, 6, 42, 0}, &GENUS, 0, 2, 0, 0 }; void XSIMD(codelet_n2fv_14) (planner *p) { X(kdft_register) (p, n2fv_14, &desc); } #else /* HAVE_FMA */ /* Generated by: ../../../genfft/gen_notw_c.native -simd -compact -variables 4 -pipeline-latency 8 -n 14 -name n2fv_14 -with-ostride 2 -include n2f.h -store-multiple 2 */ /* * This function contains 74 FP additions, 36 FP multiplications, * (or, 50 additions, 12 multiplications, 24 fused multiply/add), * 39 stack variables, 6 constants, and 35 memory accesses */ #include "n2f.h" static void n2fv_14(const R *ri, const R *ii, R *ro, R *io, stride is, stride os, INT v, INT ivs, INT ovs) { DVK(KP222520933, +0.222520933956314404288902564496794759466355569); DVK(KP900968867, +0.900968867902419126236102319507445051165919162); DVK(KP623489801, +0.623489801858733530525004884004239810632274731); DVK(KP433883739, +0.433883739117558120475768332848358754609990728); DVK(KP781831482, +0.781831482468029808708444526674057750232334519); DVK(KP974927912, +0.974927912181823607018131682993931217232785801); { INT i; const R *xi; R *xo; xi = ri; xo = ro; for (i = v; i > 0; i = i - VL, xi = xi + (VL * ivs), xo = xo + (VL * ovs), MAKE_VOLATILE_STRIDE(28, is), MAKE_VOLATILE_STRIDE(28, os)) { V T3, Ty, To, TK, Tr, TE, Ta, TJ, Tq, TB, Th, TL, Ts, TH, T1; V T2; T1 = LD(&(xi[0]), ivs, &(xi[0])); T2 = LD(&(xi[WS(is, 7)]), ivs, &(xi[WS(is, 1)])); T3 = VSUB(T1, T2); Ty = VADD(T1, T2); { V Tk, TC, Tn, TD; { V Ti, Tj, Tl, Tm; Ti = LD(&(xi[WS(is, 6)]), ivs, &(xi[0])); Tj = LD(&(xi[WS(is, 13)]), ivs, &(xi[WS(is, 1)])); Tk = VSUB(Ti, Tj); TC = VADD(Ti, Tj); Tl = LD(&(xi[WS(is, 8)]), ivs, &(xi[0])); Tm = LD(&(xi[WS(is, 1)]), ivs, &(xi[WS(is, 1)])); Tn = VSUB(Tl, Tm); TD = VADD(Tl, Tm); } To = VADD(Tk, Tn); TK = VSUB(TC, TD); Tr = VSUB(Tn, Tk); TE = VADD(TC, TD); } { V T6, Tz, T9, TA; { V T4, T5, T7, T8; T4 = LD(&(xi[WS(is, 2)]), ivs, &(xi[0])); T5 = LD(&(xi[WS(is, 9)]), ivs, &(xi[WS(is, 1)])); T6 = VSUB(T4, T5); Tz = VADD(T4, T5); T7 = LD(&(xi[WS(is, 12)]), ivs, &(xi[0])); T8 = LD(&(xi[WS(is, 5)]), ivs, &(xi[WS(is, 1)])); T9 = VSUB(T7, T8); TA = VADD(T7, T8); } Ta = VADD(T6, T9); TJ = VSUB(TA, Tz); Tq = VSUB(T9, T6); TB = VADD(Tz, TA); } { V Td, TF, Tg, TG; { V Tb, Tc, Te, Tf; Tb = LD(&(xi[WS(is, 4)]), ivs, &(xi[0])); Tc = LD(&(xi[WS(is, 11)]), ivs, &(xi[WS(is, 1)])); Td = VSUB(Tb, Tc); TF = VADD(Tb, Tc); Te = LD(&(xi[WS(is, 10)]), ivs, &(xi[0])); Tf = LD(&(xi[WS(is, 3)]), ivs, &(xi[WS(is, 1)])); Tg = VSUB(Te, Tf); TG = VADD(Te, Tf); } Th = VADD(Td, Tg); TL = VSUB(TF, TG); Ts = VSUB(Tg, Td); TH = VADD(TF, TG); } { V TR, TS, TT, TU, TV, TW; TR = VADD(T3, VADD(Ta, VADD(Th, To))); STM2(&(xo[14]), TR, ovs, &(xo[2])); TS = VADD(Ty, VADD(TB, VADD(TH, TE))); STM2(&(xo[0]), TS, ovs, &(xo[0])); { V Tt, Tp, TP, TQ; Tt = VBYI(VFNMS(LDK(KP781831482), Tr, VFNMS(LDK(KP433883739), Ts, VMUL(LDK(KP974927912), Tq)))); Tp = VFMA(LDK(KP623489801), To, VFNMS(LDK(KP900968867), Th, VFNMS(LDK(KP222520933), Ta, T3))); TT = VSUB(Tp, Tt); STM2(&(xo[10]), TT, ovs, &(xo[2])); TU = VADD(Tp, Tt); STM2(&(xo[18]), TU, ovs, &(xo[2])); TP = VBYI(VFMA(LDK(KP974927912), TJ, VFMA(LDK(KP433883739), TL, VMUL(LDK(KP781831482), TK)))); TQ = VFMA(LDK(KP623489801), TE, VFNMS(LDK(KP900968867), TH, VFNMS(LDK(KP222520933), TB, Ty))); TV = VADD(TP, TQ); STM2(&(xo[4]), TV, ovs, &(xo[0])); TW = VSUB(TQ, TP); STM2(&(xo[24]), TW, ovs, &(xo[0])); } { V Tv, Tu, TX, TY; Tv = VBYI(VFMA(LDK(KP781831482), Tq, VFMA(LDK(KP974927912), Ts, VMUL(LDK(KP433883739), Tr)))); Tu = VFMA(LDK(KP623489801), Ta, VFNMS(LDK(KP900968867), To, VFNMS(LDK(KP222520933), Th, T3))); TX = VSUB(Tu, Tv); STM2(&(xo[26]), TX, ovs, &(xo[2])); STN2(&(xo[24]), TW, TX, ovs); TY = VADD(Tu, Tv); STM2(&(xo[2]), TY, ovs, &(xo[2])); STN2(&(xo[0]), TS, TY, ovs); } { V TM, TI, TZ, T10; TM = VBYI(VFNMS(LDK(KP433883739), TK, VFNMS(LDK(KP974927912), TL, VMUL(LDK(KP781831482), TJ)))); TI = VFMA(LDK(KP623489801), TB, VFNMS(LDK(KP900968867), TE, VFNMS(LDK(KP222520933), TH, Ty))); TZ = VSUB(TI, TM); STM2(&(xo[12]), TZ, ovs, &(xo[0])); STN2(&(xo[12]), TZ, TR, ovs); T10 = VADD(TM, TI); STM2(&(xo[16]), T10, ovs, &(xo[0])); STN2(&(xo[16]), T10, TU, ovs); } { V T12, TO, TN, T11; TO = VBYI(VFMA(LDK(KP433883739), TJ, VFNMS(LDK(KP974927912), TK, VMUL(LDK(KP781831482), TL)))); TN = VFMA(LDK(KP623489801), TH, VFNMS(LDK(KP222520933), TE, VFNMS(LDK(KP900968867), TB, Ty))); T11 = VSUB(TN, TO); STM2(&(xo[8]), T11, ovs, &(xo[0])); STN2(&(xo[8]), T11, TT, ovs); T12 = VADD(TO, TN); STM2(&(xo[20]), T12, ovs, &(xo[0])); { V Tx, Tw, T13, T14; Tx = VBYI(VFMA(LDK(KP433883739), Tq, VFNMS(LDK(KP781831482), Ts, VMUL(LDK(KP974927912), Tr)))); Tw = VFMA(LDK(KP623489801), Th, VFNMS(LDK(KP222520933), To, VFNMS(LDK(KP900968867), Ta, T3))); T13 = VSUB(Tw, Tx); STM2(&(xo[22]), T13, ovs, &(xo[2])); STN2(&(xo[20]), T12, T13, ovs); T14 = VADD(Tw, Tx); STM2(&(xo[6]), T14, ovs, &(xo[2])); STN2(&(xo[4]), TV, T14, ovs); } } } } } VLEAVE(); } static const kdft_desc desc = { 14, XSIMD_STRING("n2fv_14"), {50, 12, 24, 0}, &GENUS, 0, 2, 0, 0 }; void XSIMD(codelet_n2fv_14) (planner *p) { X(kdft_register) (p, n2fv_14, &desc); } #endif /* HAVE_FMA */ fftw-3.3.3/dft/simd/common/n2fv_16.c0000644000175000017500000003174512054410404013741 00000000000000/* * Copyright (c) 2003, 2007-11 Matteo Frigo * Copyright (c) 2003, 2007-11 Massachusetts Institute of Technology * * This program is free software; you can redistribute it and/or modify * it under the terms of the GNU General Public License as published by * the Free Software Foundation; either version 2 of the License, or * (at your option) any later version. * * This program is distributed in the hope that it will be useful, * but WITHOUT ANY WARRANTY; without even the implied warranty of * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the * GNU General Public License for more details. * * You should have received a copy of the GNU General Public License * along with this program; if not, write to the Free Software * Foundation, Inc., 51 Franklin Street, Fifth Floor, Boston, MA 02110-1301 USA * */ /* This file was automatically generated --- DO NOT EDIT */ /* Generated on Sun Nov 25 07:37:23 EST 2012 */ #include "codelet-dft.h" #ifdef HAVE_FMA /* Generated by: ../../../genfft/gen_notw_c.native -fma -reorder-insns -schedule-for-pipeline -simd -compact -variables 4 -pipeline-latency 8 -n 16 -name n2fv_16 -with-ostride 2 -include n2f.h -store-multiple 2 */ /* * This function contains 72 FP additions, 34 FP multiplications, * (or, 38 additions, 0 multiplications, 34 fused multiply/add), * 62 stack variables, 3 constants, and 40 memory accesses */ #include "n2f.h" static void n2fv_16(const R *ri, const R *ii, R *ro, R *io, stride is, stride os, INT v, INT ivs, INT ovs) { DVK(KP923879532, +0.923879532511286756128183189396788286822416626); DVK(KP414213562, +0.414213562373095048801688724209698078569671875); DVK(KP707106781, +0.707106781186547524400844362104849039284835938); { INT i; const R *xi; R *xo; xi = ri; xo = ro; for (i = v; i > 0; i = i - VL, xi = xi + (VL * ivs), xo = xo + (VL * ovs), MAKE_VOLATILE_STRIDE(32, is), MAKE_VOLATILE_STRIDE(32, os)) { V T7, Tu, TF, TB, T13, TL, TO, TX, TC, Te, TP, Th, TQ, Tk, TW; V T16; { V TH, TU, Tz, Tf, TK, TV, TA, TM, Ta, TN, Td, Tg, Ti, Tj; { V T1, T2, T4, T5, To, Tp, Tr, Ts; T1 = LD(&(xi[0]), ivs, &(xi[0])); T2 = LD(&(xi[WS(is, 8)]), ivs, &(xi[0])); T4 = LD(&(xi[WS(is, 4)]), ivs, &(xi[0])); T5 = LD(&(xi[WS(is, 12)]), ivs, &(xi[0])); To = LD(&(xi[WS(is, 14)]), ivs, &(xi[0])); Tp = LD(&(xi[WS(is, 6)]), ivs, &(xi[0])); Tr = LD(&(xi[WS(is, 2)]), ivs, &(xi[0])); Ts = LD(&(xi[WS(is, 10)]), ivs, &(xi[0])); { V T8, TJ, Tq, TI, Tt, T9, Tb, Tc, T3, T6; T8 = LD(&(xi[WS(is, 1)]), ivs, &(xi[WS(is, 1)])); TH = VSUB(T1, T2); T3 = VADD(T1, T2); TU = VSUB(T4, T5); T6 = VADD(T4, T5); TJ = VSUB(To, Tp); Tq = VADD(To, Tp); TI = VSUB(Tr, Ts); Tt = VADD(Tr, Ts); T9 = LD(&(xi[WS(is, 9)]), ivs, &(xi[WS(is, 1)])); Tb = LD(&(xi[WS(is, 5)]), ivs, &(xi[WS(is, 1)])); Tc = LD(&(xi[WS(is, 13)]), ivs, &(xi[WS(is, 1)])); T7 = VSUB(T3, T6); Tz = VADD(T3, T6); Tf = LD(&(xi[WS(is, 15)]), ivs, &(xi[WS(is, 1)])); TK = VADD(TI, TJ); TV = VSUB(TJ, TI); TA = VADD(Tt, Tq); Tu = VSUB(Tq, Tt); TM = VSUB(T8, T9); Ta = VADD(T8, T9); TN = VSUB(Tb, Tc); Td = VADD(Tb, Tc); Tg = LD(&(xi[WS(is, 7)]), ivs, &(xi[WS(is, 1)])); Ti = LD(&(xi[WS(is, 3)]), ivs, &(xi[WS(is, 1)])); Tj = LD(&(xi[WS(is, 11)]), ivs, &(xi[WS(is, 1)])); } } TF = VSUB(Tz, TA); TB = VADD(Tz, TA); T13 = VFNMS(LDK(KP707106781), TK, TH); TL = VFMA(LDK(KP707106781), TK, TH); TO = VFNMS(LDK(KP414213562), TN, TM); TX = VFMA(LDK(KP414213562), TM, TN); TC = VADD(Ta, Td); Te = VSUB(Ta, Td); TP = VSUB(Tf, Tg); Th = VADD(Tf, Tg); TQ = VSUB(Tj, Ti); Tk = VADD(Ti, Tj); TW = VFNMS(LDK(KP707106781), TV, TU); T16 = VFMA(LDK(KP707106781), TV, TU); } { V TY, TR, Tl, TD; TY = VFMA(LDK(KP414213562), TP, TQ); TR = VFNMS(LDK(KP414213562), TQ, TP); Tl = VSUB(Th, Tk); TD = VADD(Th, Tk); { V TS, T17, TZ, T14; TS = VADD(TO, TR); T17 = VSUB(TR, TO); TZ = VSUB(TX, TY); T14 = VADD(TX, TY); { V TE, TG, Tm, Tv; TE = VADD(TC, TD); TG = VSUB(TD, TC); Tm = VADD(Te, Tl); Tv = VSUB(Tl, Te); { V T18, T1a, TT, T11; T18 = VFNMS(LDK(KP923879532), T17, T16); T1a = VFMA(LDK(KP923879532), T17, T16); TT = VFNMS(LDK(KP923879532), TS, TL); T11 = VFMA(LDK(KP923879532), TS, TL); { V T15, T19, T10, T12; T15 = VFNMS(LDK(KP923879532), T14, T13); T19 = VFMA(LDK(KP923879532), T14, T13); T10 = VFNMS(LDK(KP923879532), TZ, TW); T12 = VFMA(LDK(KP923879532), TZ, TW); { V T1b, T1c, T1d, T1e; T1b = VFMAI(TG, TF); STM2(&(xo[8]), T1b, ovs, &(xo[0])); T1c = VFNMSI(TG, TF); STM2(&(xo[24]), T1c, ovs, &(xo[0])); T1d = VADD(TB, TE); STM2(&(xo[0]), T1d, ovs, &(xo[0])); T1e = VSUB(TB, TE); STM2(&(xo[16]), T1e, ovs, &(xo[0])); { V Tw, Ty, Tn, Tx; Tw = VFNMS(LDK(KP707106781), Tv, Tu); Ty = VFMA(LDK(KP707106781), Tv, Tu); Tn = VFNMS(LDK(KP707106781), Tm, T7); Tx = VFMA(LDK(KP707106781), Tm, T7); { V T1f, T1g, T1h, T1i; T1f = VFMAI(T1a, T19); STM2(&(xo[6]), T1f, ovs, &(xo[2])); T1g = VFNMSI(T1a, T19); STM2(&(xo[26]), T1g, ovs, &(xo[2])); STN2(&(xo[24]), T1c, T1g, ovs); T1h = VFMAI(T18, T15); STM2(&(xo[22]), T1h, ovs, &(xo[2])); T1i = VFNMSI(T18, T15); STM2(&(xo[10]), T1i, ovs, &(xo[2])); STN2(&(xo[8]), T1b, T1i, ovs); { V T1j, T1k, T1l, T1m; T1j = VFNMSI(T12, T11); STM2(&(xo[2]), T1j, ovs, &(xo[2])); STN2(&(xo[0]), T1d, T1j, ovs); T1k = VFMAI(T12, T11); STM2(&(xo[30]), T1k, ovs, &(xo[2])); T1l = VFMAI(T10, TT); STM2(&(xo[14]), T1l, ovs, &(xo[2])); T1m = VFNMSI(T10, TT); STM2(&(xo[18]), T1m, ovs, &(xo[2])); STN2(&(xo[16]), T1e, T1m, ovs); { V T1n, T1o, T1p, T1q; T1n = VFNMSI(Ty, Tx); STM2(&(xo[28]), T1n, ovs, &(xo[0])); STN2(&(xo[28]), T1n, T1k, ovs); T1o = VFMAI(Ty, Tx); STM2(&(xo[4]), T1o, ovs, &(xo[0])); STN2(&(xo[4]), T1o, T1f, ovs); T1p = VFMAI(Tw, Tn); STM2(&(xo[20]), T1p, ovs, &(xo[0])); STN2(&(xo[20]), T1p, T1h, ovs); T1q = VFNMSI(Tw, Tn); STM2(&(xo[12]), T1q, ovs, &(xo[0])); STN2(&(xo[12]), T1q, T1l, ovs); } } } } } } } } } } } } VLEAVE(); } static const kdft_desc desc = { 16, XSIMD_STRING("n2fv_16"), {38, 0, 34, 0}, &GENUS, 0, 2, 0, 0 }; void XSIMD(codelet_n2fv_16) (planner *p) { X(kdft_register) (p, n2fv_16, &desc); } #else /* HAVE_FMA */ /* Generated by: ../../../genfft/gen_notw_c.native -simd -compact -variables 4 -pipeline-latency 8 -n 16 -name n2fv_16 -with-ostride 2 -include n2f.h -store-multiple 2 */ /* * This function contains 72 FP additions, 12 FP multiplications, * (or, 68 additions, 8 multiplications, 4 fused multiply/add), * 38 stack variables, 3 constants, and 40 memory accesses */ #include "n2f.h" static void n2fv_16(const R *ri, const R *ii, R *ro, R *io, stride is, stride os, INT v, INT ivs, INT ovs) { DVK(KP923879532, +0.923879532511286756128183189396788286822416626); DVK(KP382683432, +0.382683432365089771728459984030398866761344562); DVK(KP707106781, +0.707106781186547524400844362104849039284835938); { INT i; const R *xi; R *xo; xi = ri; xo = ro; for (i = v; i > 0; i = i - VL, xi = xi + (VL * ivs), xo = xo + (VL * ovs), MAKE_VOLATILE_STRIDE(32, is), MAKE_VOLATILE_STRIDE(32, os)) { V Tp, T13, Tu, TN, Tm, T14, Tv, TY, T7, T17, Ty, TT, Te, T16, Tx; V TQ; { V Tn, To, TM, Ts, Tt, TL; Tn = LD(&(xi[WS(is, 4)]), ivs, &(xi[0])); To = LD(&(xi[WS(is, 12)]), ivs, &(xi[0])); TM = VADD(Tn, To); Ts = LD(&(xi[0]), ivs, &(xi[0])); Tt = LD(&(xi[WS(is, 8)]), ivs, &(xi[0])); TL = VADD(Ts, Tt); Tp = VSUB(Tn, To); T13 = VADD(TL, TM); Tu = VSUB(Ts, Tt); TN = VSUB(TL, TM); } { V Ti, TW, Tl, TX; { V Tg, Th, Tj, Tk; Tg = LD(&(xi[WS(is, 14)]), ivs, &(xi[0])); Th = LD(&(xi[WS(is, 6)]), ivs, &(xi[0])); Ti = VSUB(Tg, Th); TW = VADD(Tg, Th); Tj = LD(&(xi[WS(is, 2)]), ivs, &(xi[0])); Tk = LD(&(xi[WS(is, 10)]), ivs, &(xi[0])); Tl = VSUB(Tj, Tk); TX = VADD(Tj, Tk); } Tm = VMUL(LDK(KP707106781), VSUB(Ti, Tl)); T14 = VADD(TX, TW); Tv = VMUL(LDK(KP707106781), VADD(Tl, Ti)); TY = VSUB(TW, TX); } { V T3, TR, T6, TS; { V T1, T2, T4, T5; T1 = LD(&(xi[WS(is, 15)]), ivs, &(xi[WS(is, 1)])); T2 = LD(&(xi[WS(is, 7)]), ivs, &(xi[WS(is, 1)])); T3 = VSUB(T1, T2); TR = VADD(T1, T2); T4 = LD(&(xi[WS(is, 3)]), ivs, &(xi[WS(is, 1)])); T5 = LD(&(xi[WS(is, 11)]), ivs, &(xi[WS(is, 1)])); T6 = VSUB(T4, T5); TS = VADD(T4, T5); } T7 = VFNMS(LDK(KP923879532), T6, VMUL(LDK(KP382683432), T3)); T17 = VADD(TR, TS); Ty = VFMA(LDK(KP923879532), T3, VMUL(LDK(KP382683432), T6)); TT = VSUB(TR, TS); } { V Ta, TO, Td, TP; { V T8, T9, Tb, Tc; T8 = LD(&(xi[WS(is, 1)]), ivs, &(xi[WS(is, 1)])); T9 = LD(&(xi[WS(is, 9)]), ivs, &(xi[WS(is, 1)])); Ta = VSUB(T8, T9); TO = VADD(T8, T9); Tb = LD(&(xi[WS(is, 5)]), ivs, &(xi[WS(is, 1)])); Tc = LD(&(xi[WS(is, 13)]), ivs, &(xi[WS(is, 1)])); Td = VSUB(Tb, Tc); TP = VADD(Tb, Tc); } Te = VFMA(LDK(KP382683432), Ta, VMUL(LDK(KP923879532), Td)); T16 = VADD(TO, TP); Tx = VFNMS(LDK(KP382683432), Td, VMUL(LDK(KP923879532), Ta)); TQ = VSUB(TO, TP); } { V T1b, T1c, T1d, T1e; { V T15, T18, T19, T1a; T15 = VADD(T13, T14); T18 = VADD(T16, T17); T1b = VSUB(T15, T18); STM2(&(xo[16]), T1b, ovs, &(xo[0])); T1c = VADD(T15, T18); STM2(&(xo[0]), T1c, ovs, &(xo[0])); T19 = VSUB(T13, T14); T1a = VBYI(VSUB(T17, T16)); T1d = VSUB(T19, T1a); STM2(&(xo[24]), T1d, ovs, &(xo[0])); T1e = VADD(T19, T1a); STM2(&(xo[8]), T1e, ovs, &(xo[0])); } { V T1f, T1g, T1h, T1i; { V TV, T11, T10, T12, TU, TZ; TU = VMUL(LDK(KP707106781), VADD(TQ, TT)); TV = VADD(TN, TU); T11 = VSUB(TN, TU); TZ = VMUL(LDK(KP707106781), VSUB(TT, TQ)); T10 = VBYI(VADD(TY, TZ)); T12 = VBYI(VSUB(TZ, TY)); T1f = VSUB(TV, T10); STM2(&(xo[28]), T1f, ovs, &(xo[0])); T1g = VADD(T11, T12); STM2(&(xo[12]), T1g, ovs, &(xo[0])); T1h = VADD(TV, T10); STM2(&(xo[4]), T1h, ovs, &(xo[0])); T1i = VSUB(T11, T12); STM2(&(xo[20]), T1i, ovs, &(xo[0])); } { V Tr, TB, TA, TC; { V Tf, Tq, Tw, Tz; Tf = VSUB(T7, Te); Tq = VSUB(Tm, Tp); Tr = VBYI(VSUB(Tf, Tq)); TB = VBYI(VADD(Tq, Tf)); Tw = VADD(Tu, Tv); Tz = VADD(Tx, Ty); TA = VSUB(Tw, Tz); TC = VADD(Tw, Tz); } { V T1j, T1k, T1l, T1m; T1j = VADD(Tr, TA); STM2(&(xo[14]), T1j, ovs, &(xo[2])); STN2(&(xo[12]), T1g, T1j, ovs); T1k = VSUB(TC, TB); STM2(&(xo[30]), T1k, ovs, &(xo[2])); STN2(&(xo[28]), T1f, T1k, ovs); T1l = VSUB(TA, Tr); STM2(&(xo[18]), T1l, ovs, &(xo[2])); STN2(&(xo[16]), T1b, T1l, ovs); T1m = VADD(TB, TC); STM2(&(xo[2]), T1m, ovs, &(xo[2])); STN2(&(xo[0]), T1c, T1m, ovs); } } { V TF, TJ, TI, TK; { V TD, TE, TG, TH; TD = VSUB(Tu, Tv); TE = VADD(Te, T7); TF = VADD(TD, TE); TJ = VSUB(TD, TE); TG = VADD(Tp, Tm); TH = VSUB(Ty, Tx); TI = VBYI(VADD(TG, TH)); TK = VBYI(VSUB(TH, TG)); } { V T1n, T1o, T1p, T1q; T1n = VSUB(TF, TI); STM2(&(xo[26]), T1n, ovs, &(xo[2])); STN2(&(xo[24]), T1d, T1n, ovs); T1o = VADD(TJ, TK); STM2(&(xo[10]), T1o, ovs, &(xo[2])); STN2(&(xo[8]), T1e, T1o, ovs); T1p = VADD(TF, TI); STM2(&(xo[6]), T1p, ovs, &(xo[2])); STN2(&(xo[4]), T1h, T1p, ovs); T1q = VSUB(TJ, TK); STM2(&(xo[22]), T1q, ovs, &(xo[2])); STN2(&(xo[20]), T1i, T1q, ovs); } } } } } } VLEAVE(); } static const kdft_desc desc = { 16, XSIMD_STRING("n2fv_16"), {68, 8, 4, 0}, &GENUS, 0, 2, 0, 0 }; void XSIMD(codelet_n2fv_16) (planner *p) { X(kdft_register) (p, n2fv_16, &desc); } #endif /* HAVE_FMA */ fftw-3.3.3/dft/simd/common/n2fv_32.c0000644000175000017500000006554412054410410013740 00000000000000/* * Copyright (c) 2003, 2007-11 Matteo Frigo * Copyright (c) 2003, 2007-11 Massachusetts Institute of Technology * * This program is free software; you can redistribute it and/or modify * it under the terms of the GNU General Public License as published by * the Free Software Foundation; either version 2 of the License, or * (at your option) any later version. * * This program is distributed in the hope that it will be useful, * but WITHOUT ANY WARRANTY; without even the implied warranty of * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the * GNU General Public License for more details. * * You should have received a copy of the GNU General Public License * along with this program; if not, write to the Free Software * Foundation, Inc., 51 Franklin Street, Fifth Floor, Boston, MA 02110-1301 USA * */ /* This file was automatically generated --- DO NOT EDIT */ /* Generated on Sun Nov 25 07:37:24 EST 2012 */ #include "codelet-dft.h" #ifdef HAVE_FMA /* Generated by: ../../../genfft/gen_notw_c.native -fma -reorder-insns -schedule-for-pipeline -simd -compact -variables 4 -pipeline-latency 8 -n 32 -name n2fv_32 -with-ostride 2 -include n2f.h -store-multiple 2 */ /* * This function contains 186 FP additions, 98 FP multiplications, * (or, 88 additions, 0 multiplications, 98 fused multiply/add), * 120 stack variables, 7 constants, and 80 memory accesses */ #include "n2f.h" static void n2fv_32(const R *ri, const R *ii, R *ro, R *io, stride is, stride os, INT v, INT ivs, INT ovs) { DVK(KP980785280, +0.980785280403230449126182236134239036973933731); DVK(KP831469612, +0.831469612302545237078788377617905756738560812); DVK(KP198912367, +0.198912367379658006911597622644676228597850501); DVK(KP668178637, +0.668178637919298919997757686523080761552472251); DVK(KP923879532, +0.923879532511286756128183189396788286822416626); DVK(KP414213562, +0.414213562373095048801688724209698078569671875); DVK(KP707106781, +0.707106781186547524400844362104849039284835938); { INT i; const R *xi; R *xo; xi = ri; xo = ro; for (i = v; i > 0; i = i - VL, xi = xi + (VL * ivs), xo = xo + (VL * ovs), MAKE_VOLATILE_STRIDE(64, is), MAKE_VOLATILE_STRIDE(64, os)) { V T31, T32, T33, T34, T35, T36, T37, T38, T39, T3a, T3b, T3c, T1h, Tr, T3d; V T3e, T3f, T3g, T1a, T1k, TI, T1b, T1L, T1P, T1I, T1G, T1O, T1Q, T1H, T1z; V T1c, TZ; { V T2x, T1T, T2K, T1W, T1p, Tb, T1A, T16, Tu, TF, T2N, T2H, T2b, T2t, TY; V T1w, TT, T1v, T20, T2C, Tj, Te, T2h, To, T2f, T23, T2D, TB, TG, Th; V T2i, Tk; { V TL, TW, TP, TQ, T2F, T27, T28, TO; { V T1, T2, T12, T13, T4, T5, T7, T8; T1 = LD(&(xi[0]), ivs, &(xi[0])); T2 = LD(&(xi[WS(is, 16)]), ivs, &(xi[0])); T12 = LD(&(xi[WS(is, 8)]), ivs, &(xi[0])); T13 = LD(&(xi[WS(is, 24)]), ivs, &(xi[0])); T4 = LD(&(xi[WS(is, 4)]), ivs, &(xi[0])); T5 = LD(&(xi[WS(is, 20)]), ivs, &(xi[0])); T7 = LD(&(xi[WS(is, 28)]), ivs, &(xi[0])); T8 = LD(&(xi[WS(is, 12)]), ivs, &(xi[0])); { V TM, T25, T26, TN; { V TJ, T3, T14, T1U, T6, T1V, T9, TK, TU, TV, T1R, T1S, Ta, T15; TJ = LD(&(xi[WS(is, 31)]), ivs, &(xi[WS(is, 1)])); T1R = VADD(T1, T2); T3 = VSUB(T1, T2); T1S = VADD(T12, T13); T14 = VSUB(T12, T13); T1U = VADD(T4, T5); T6 = VSUB(T4, T5); T1V = VADD(T7, T8); T9 = VSUB(T7, T8); TK = LD(&(xi[WS(is, 15)]), ivs, &(xi[WS(is, 1)])); TU = LD(&(xi[WS(is, 23)]), ivs, &(xi[WS(is, 1)])); T2x = VSUB(T1R, T1S); T1T = VADD(T1R, T1S); TV = LD(&(xi[WS(is, 7)]), ivs, &(xi[WS(is, 1)])); TM = LD(&(xi[WS(is, 3)]), ivs, &(xi[WS(is, 1)])); T2K = VSUB(T1V, T1U); T1W = VADD(T1U, T1V); Ta = VADD(T6, T9); T15 = VSUB(T9, T6); T25 = VADD(TJ, TK); TL = VSUB(TJ, TK); T26 = VADD(TV, TU); TW = VSUB(TU, TV); TN = LD(&(xi[WS(is, 19)]), ivs, &(xi[WS(is, 1)])); TP = LD(&(xi[WS(is, 27)]), ivs, &(xi[WS(is, 1)])); T1p = VFNMS(LDK(KP707106781), Ta, T3); Tb = VFMA(LDK(KP707106781), Ta, T3); T1A = VFMA(LDK(KP707106781), T15, T14); T16 = VFNMS(LDK(KP707106781), T15, T14); TQ = LD(&(xi[WS(is, 11)]), ivs, &(xi[WS(is, 1)])); } T2F = VSUB(T25, T26); T27 = VADD(T25, T26); T28 = VADD(TM, TN); TO = VSUB(TM, TN); } } { V Ty, T21, Tx, Tz, T1Y, T1Z; { V Ts, Tt, TD, T29, TR, TE, Tv, Tw; Ts = LD(&(xi[WS(is, 1)]), ivs, &(xi[WS(is, 1)])); Tt = LD(&(xi[WS(is, 17)]), ivs, &(xi[WS(is, 1)])); TD = LD(&(xi[WS(is, 9)]), ivs, &(xi[WS(is, 1)])); T29 = VADD(TP, TQ); TR = VSUB(TP, TQ); TE = LD(&(xi[WS(is, 25)]), ivs, &(xi[WS(is, 1)])); Tv = LD(&(xi[WS(is, 5)]), ivs, &(xi[WS(is, 1)])); Tw = LD(&(xi[WS(is, 21)]), ivs, &(xi[WS(is, 1)])); Ty = LD(&(xi[WS(is, 29)]), ivs, &(xi[WS(is, 1)])); T1Y = VADD(Ts, Tt); Tu = VSUB(Ts, Tt); { V T2G, T2a, TX, TS; T2G = VSUB(T29, T28); T2a = VADD(T28, T29); TX = VSUB(TR, TO); TS = VADD(TO, TR); T1Z = VADD(TD, TE); TF = VSUB(TD, TE); T21 = VADD(Tv, Tw); Tx = VSUB(Tv, Tw); T2N = VFMA(LDK(KP414213562), T2F, T2G); T2H = VFNMS(LDK(KP414213562), T2G, T2F); T2b = VSUB(T27, T2a); T2t = VADD(T27, T2a); TY = VFMA(LDK(KP707106781), TX, TW); T1w = VFNMS(LDK(KP707106781), TX, TW); TT = VFMA(LDK(KP707106781), TS, TL); T1v = VFNMS(LDK(KP707106781), TS, TL); Tz = LD(&(xi[WS(is, 13)]), ivs, &(xi[WS(is, 1)])); } } T20 = VADD(T1Y, T1Z); T2C = VSUB(T1Y, T1Z); { V Tc, Td, Tm, Tn; Tc = LD(&(xi[WS(is, 2)]), ivs, &(xi[0])); Td = LD(&(xi[WS(is, 18)]), ivs, &(xi[0])); Tm = LD(&(xi[WS(is, 22)]), ivs, &(xi[0])); Tn = LD(&(xi[WS(is, 6)]), ivs, &(xi[0])); { V Tf, TA, T22, Tg; Tf = LD(&(xi[WS(is, 10)]), ivs, &(xi[0])); TA = VSUB(Ty, Tz); T22 = VADD(Ty, Tz); Tg = LD(&(xi[WS(is, 26)]), ivs, &(xi[0])); Tj = LD(&(xi[WS(is, 30)]), ivs, &(xi[0])); Te = VSUB(Tc, Td); T2h = VADD(Tc, Td); To = VSUB(Tm, Tn); T2f = VADD(Tn, Tm); T23 = VADD(T21, T22); T2D = VSUB(T21, T22); TB = VADD(Tx, TA); TG = VSUB(Tx, TA); Th = VSUB(Tf, Tg); T2i = VADD(Tf, Tg); Tk = LD(&(xi[WS(is, 14)]), ivs, &(xi[0])); } } } } { V T1t, TH, T1s, TC, T2P, T2U, T2n, T2d, T2w, T2u, T1q, T19, T1B, Tq, T2W; V T2M, T2B, T2T, T2v, T2r, T2o, T2m, T2X, T2I; { V T1X, T2p, T2E, T2O, T2s, T2y, T2j, T17, Ti, T2e, Tl, T2c, T2l, T24; T1X = VSUB(T1T, T1W); T2p = VADD(T1T, T1W); T2E = VFNMS(LDK(KP414213562), T2D, T2C); T2O = VFMA(LDK(KP414213562), T2C, T2D); T2s = VADD(T20, T23); T24 = VSUB(T20, T23); T1t = VFNMS(LDK(KP707106781), TG, TF); TH = VFMA(LDK(KP707106781), TG, TF); T1s = VFNMS(LDK(KP707106781), TB, Tu); TC = VFMA(LDK(KP707106781), TB, Tu); T2y = VSUB(T2h, T2i); T2j = VADD(T2h, T2i); T17 = VFMA(LDK(KP414213562), Te, Th); Ti = VFNMS(LDK(KP414213562), Th, Te); T2e = VADD(Tj, Tk); Tl = VSUB(Tj, Tk); T2c = VADD(T24, T2b); T2l = VSUB(T2b, T24); { V T2L, T2A, T2q, T2k; T2P = VSUB(T2N, T2O); T2U = VADD(T2O, T2N); { V T2z, T2g, T18, Tp; T2z = VSUB(T2e, T2f); T2g = VADD(T2e, T2f); T18 = VFMA(LDK(KP414213562), Tl, To); Tp = VFNMS(LDK(KP414213562), To, Tl); T2n = VFMA(LDK(KP707106781), T2c, T1X); T2d = VFNMS(LDK(KP707106781), T2c, T1X); T2w = VSUB(T2t, T2s); T2u = VADD(T2s, T2t); T2L = VSUB(T2z, T2y); T2A = VADD(T2y, T2z); T2q = VADD(T2j, T2g); T2k = VSUB(T2g, T2j); T1q = VADD(T17, T18); T19 = VSUB(T17, T18); T1B = VSUB(Tp, Ti); Tq = VADD(Ti, Tp); } T2W = VFNMS(LDK(KP707106781), T2L, T2K); T2M = VFMA(LDK(KP707106781), T2L, T2K); T2B = VFMA(LDK(KP707106781), T2A, T2x); T2T = VFNMS(LDK(KP707106781), T2A, T2x); T2v = VSUB(T2p, T2q); T2r = VADD(T2p, T2q); T2o = VFMA(LDK(KP707106781), T2l, T2k); T2m = VFNMS(LDK(KP707106781), T2l, T2k); T2X = VSUB(T2H, T2E); T2I = VADD(T2E, T2H); } } { V T2V, T2Z, T2Y, T30, T2R, T2J; T2V = VFNMS(LDK(KP923879532), T2U, T2T); T2Z = VFMA(LDK(KP923879532), T2U, T2T); T31 = VFNMSI(T2w, T2v); STM2(&(xo[48]), T31, ovs, &(xo[0])); T32 = VFMAI(T2w, T2v); STM2(&(xo[16]), T32, ovs, &(xo[0])); T33 = VADD(T2r, T2u); STM2(&(xo[0]), T33, ovs, &(xo[0])); T34 = VSUB(T2r, T2u); STM2(&(xo[32]), T34, ovs, &(xo[0])); T35 = VFNMSI(T2o, T2n); STM2(&(xo[56]), T35, ovs, &(xo[0])); T36 = VFMAI(T2o, T2n); STM2(&(xo[8]), T36, ovs, &(xo[0])); T37 = VFMAI(T2m, T2d); STM2(&(xo[40]), T37, ovs, &(xo[0])); T38 = VFNMSI(T2m, T2d); STM2(&(xo[24]), T38, ovs, &(xo[0])); T2Y = VFMA(LDK(KP923879532), T2X, T2W); T30 = VFNMS(LDK(KP923879532), T2X, T2W); T2R = VFMA(LDK(KP923879532), T2I, T2B); T2J = VFNMS(LDK(KP923879532), T2I, T2B); { V T1J, T1r, T1C, T1M, T2S, T2Q, T1u, T1D, T1E, T1x; T1J = VFNMS(LDK(KP923879532), T1q, T1p); T1r = VFMA(LDK(KP923879532), T1q, T1p); T1C = VFMA(LDK(KP923879532), T1B, T1A); T1M = VFNMS(LDK(KP923879532), T1B, T1A); T39 = VFNMSI(T30, T2Z); STM2(&(xo[12]), T39, ovs, &(xo[0])); T3a = VFMAI(T30, T2Z); STM2(&(xo[52]), T3a, ovs, &(xo[0])); T3b = VFNMSI(T2Y, T2V); STM2(&(xo[44]), T3b, ovs, &(xo[0])); T3c = VFMAI(T2Y, T2V); STM2(&(xo[20]), T3c, ovs, &(xo[0])); T2S = VFMA(LDK(KP923879532), T2P, T2M); T2Q = VFNMS(LDK(KP923879532), T2P, T2M); T1u = VFMA(LDK(KP668178637), T1t, T1s); T1D = VFNMS(LDK(KP668178637), T1s, T1t); T1E = VFNMS(LDK(KP668178637), T1v, T1w); T1x = VFMA(LDK(KP668178637), T1w, T1v); { V T1K, T1F, T1N, T1y; T1h = VFNMS(LDK(KP923879532), Tq, Tb); Tr = VFMA(LDK(KP923879532), Tq, Tb); T3d = VFNMSI(T2S, T2R); STM2(&(xo[60]), T3d, ovs, &(xo[0])); T3e = VFMAI(T2S, T2R); STM2(&(xo[4]), T3e, ovs, &(xo[0])); T3f = VFMAI(T2Q, T2J); STM2(&(xo[36]), T3f, ovs, &(xo[0])); T3g = VFNMSI(T2Q, T2J); STM2(&(xo[28]), T3g, ovs, &(xo[0])); T1K = VADD(T1D, T1E); T1F = VSUB(T1D, T1E); T1N = VSUB(T1x, T1u); T1y = VADD(T1u, T1x); T1a = VFMA(LDK(KP923879532), T19, T16); T1k = VFNMS(LDK(KP923879532), T19, T16); TI = VFNMS(LDK(KP198912367), TH, TC); T1b = VFMA(LDK(KP198912367), TC, TH); T1L = VFMA(LDK(KP831469612), T1K, T1J); T1P = VFNMS(LDK(KP831469612), T1K, T1J); T1I = VFMA(LDK(KP831469612), T1F, T1C); T1G = VFNMS(LDK(KP831469612), T1F, T1C); T1O = VFMA(LDK(KP831469612), T1N, T1M); T1Q = VFNMS(LDK(KP831469612), T1N, T1M); T1H = VFMA(LDK(KP831469612), T1y, T1r); T1z = VFNMS(LDK(KP831469612), T1y, T1r); T1c = VFMA(LDK(KP198912367), TT, TY); TZ = VFNMS(LDK(KP198912367), TY, TT); } } } } } { V T1d, T1i, T10, T1l; { V T3h, T3i, T3j, T3k; T3h = VFNMSI(T1O, T1L); STM2(&(xo[42]), T3h, ovs, &(xo[2])); STN2(&(xo[40]), T37, T3h, ovs); T3i = VFMAI(T1O, T1L); STM2(&(xo[22]), T3i, ovs, &(xo[2])); STN2(&(xo[20]), T3c, T3i, ovs); T3j = VFMAI(T1Q, T1P); STM2(&(xo[54]), T3j, ovs, &(xo[2])); STN2(&(xo[52]), T3a, T3j, ovs); T3k = VFNMSI(T1Q, T1P); STM2(&(xo[10]), T3k, ovs, &(xo[2])); STN2(&(xo[8]), T36, T3k, ovs); { V T3l, T3m, T3n, T3o; T3l = VFMAI(T1I, T1H); STM2(&(xo[6]), T3l, ovs, &(xo[2])); STN2(&(xo[4]), T3e, T3l, ovs); T3m = VFNMSI(T1I, T1H); STM2(&(xo[58]), T3m, ovs, &(xo[2])); STN2(&(xo[56]), T35, T3m, ovs); T3n = VFMAI(T1G, T1z); STM2(&(xo[38]), T3n, ovs, &(xo[2])); STN2(&(xo[36]), T3f, T3n, ovs); T3o = VFNMSI(T1G, T1z); STM2(&(xo[26]), T3o, ovs, &(xo[2])); STN2(&(xo[24]), T38, T3o, ovs); T1d = VSUB(T1b, T1c); T1i = VADD(T1b, T1c); T10 = VADD(TI, TZ); T1l = VSUB(TZ, TI); } } { V T1n, T1j, T1e, T1g, T1o, T1m, T11, T1f; T1n = VFMA(LDK(KP980785280), T1i, T1h); T1j = VFNMS(LDK(KP980785280), T1i, T1h); T1e = VFNMS(LDK(KP980785280), T1d, T1a); T1g = VFMA(LDK(KP980785280), T1d, T1a); T1o = VFMA(LDK(KP980785280), T1l, T1k); T1m = VFNMS(LDK(KP980785280), T1l, T1k); T11 = VFNMS(LDK(KP980785280), T10, Tr); T1f = VFMA(LDK(KP980785280), T10, Tr); { V T3p, T3q, T3r, T3s; T3p = VFMAI(T1m, T1j); STM2(&(xo[46]), T3p, ovs, &(xo[2])); STN2(&(xo[44]), T3b, T3p, ovs); T3q = VFNMSI(T1m, T1j); STM2(&(xo[18]), T3q, ovs, &(xo[2])); STN2(&(xo[16]), T32, T3q, ovs); T3r = VFNMSI(T1o, T1n); STM2(&(xo[50]), T3r, ovs, &(xo[2])); STN2(&(xo[48]), T31, T3r, ovs); T3s = VFMAI(T1o, T1n); STM2(&(xo[14]), T3s, ovs, &(xo[2])); STN2(&(xo[12]), T39, T3s, ovs); { V T3t, T3u, T3v, T3w; T3t = VFMAI(T1g, T1f); STM2(&(xo[62]), T3t, ovs, &(xo[2])); STN2(&(xo[60]), T3d, T3t, ovs); T3u = VFNMSI(T1g, T1f); STM2(&(xo[2]), T3u, ovs, &(xo[2])); STN2(&(xo[0]), T33, T3u, ovs); T3v = VFMAI(T1e, T11); STM2(&(xo[30]), T3v, ovs, &(xo[2])); STN2(&(xo[28]), T3g, T3v, ovs); T3w = VFNMSI(T1e, T11); STM2(&(xo[34]), T3w, ovs, &(xo[2])); STN2(&(xo[32]), T34, T3w, ovs); } } } } } } VLEAVE(); } static const kdft_desc desc = { 32, XSIMD_STRING("n2fv_32"), {88, 0, 98, 0}, &GENUS, 0, 2, 0, 0 }; void XSIMD(codelet_n2fv_32) (planner *p) { X(kdft_register) (p, n2fv_32, &desc); } #else /* HAVE_FMA */ /* Generated by: ../../../genfft/gen_notw_c.native -simd -compact -variables 4 -pipeline-latency 8 -n 32 -name n2fv_32 -with-ostride 2 -include n2f.h -store-multiple 2 */ /* * This function contains 186 FP additions, 42 FP multiplications, * (or, 170 additions, 26 multiplications, 16 fused multiply/add), * 72 stack variables, 7 constants, and 80 memory accesses */ #include "n2f.h" static void n2fv_32(const R *ri, const R *ii, R *ro, R *io, stride is, stride os, INT v, INT ivs, INT ovs) { DVK(KP195090322, +0.195090322016128267848284868477022240927691618); DVK(KP980785280, +0.980785280403230449126182236134239036973933731); DVK(KP555570233, +0.555570233019602224742830813948532874374937191); DVK(KP831469612, +0.831469612302545237078788377617905756738560812); DVK(KP382683432, +0.382683432365089771728459984030398866761344562); DVK(KP923879532, +0.923879532511286756128183189396788286822416626); DVK(KP707106781, +0.707106781186547524400844362104849039284835938); { INT i; const R *xi; R *xo; xi = ri; xo = ro; for (i = v; i > 0; i = i - VL, xi = xi + (VL * ivs), xo = xo + (VL * ovs), MAKE_VOLATILE_STRIDE(64, is), MAKE_VOLATILE_STRIDE(64, os)) { V T1T, T1W, T2K, T2x, T16, T1A, Tb, T1p, TT, T1v, TY, T1w, T27, T2a, T2b; V T2H, T2O, TC, T1s, TH, T1t, T20, T23, T24, T2E, T2N, T2g, T2j, Tq, T1B; V T19, T1q, T2A, T2L; { V T3, T1R, T15, T1S, T6, T1U, T9, T1V, T12, Ta; { V T1, T2, T13, T14; T1 = LD(&(xi[0]), ivs, &(xi[0])); T2 = LD(&(xi[WS(is, 16)]), ivs, &(xi[0])); T3 = VSUB(T1, T2); T1R = VADD(T1, T2); T13 = LD(&(xi[WS(is, 8)]), ivs, &(xi[0])); T14 = LD(&(xi[WS(is, 24)]), ivs, &(xi[0])); T15 = VSUB(T13, T14); T1S = VADD(T13, T14); } { V T4, T5, T7, T8; T4 = LD(&(xi[WS(is, 4)]), ivs, &(xi[0])); T5 = LD(&(xi[WS(is, 20)]), ivs, &(xi[0])); T6 = VSUB(T4, T5); T1U = VADD(T4, T5); T7 = LD(&(xi[WS(is, 28)]), ivs, &(xi[0])); T8 = LD(&(xi[WS(is, 12)]), ivs, &(xi[0])); T9 = VSUB(T7, T8); T1V = VADD(T7, T8); } T1T = VADD(T1R, T1S); T1W = VADD(T1U, T1V); T2K = VSUB(T1V, T1U); T2x = VSUB(T1R, T1S); T12 = VMUL(LDK(KP707106781), VSUB(T9, T6)); T16 = VSUB(T12, T15); T1A = VADD(T15, T12); Ta = VMUL(LDK(KP707106781), VADD(T6, T9)); Tb = VADD(T3, Ta); T1p = VSUB(T3, Ta); } { V TL, T25, TX, T26, TO, T28, TR, T29; { V TJ, TK, TV, TW; TJ = LD(&(xi[WS(is, 31)]), ivs, &(xi[WS(is, 1)])); TK = LD(&(xi[WS(is, 15)]), ivs, &(xi[WS(is, 1)])); TL = VSUB(TJ, TK); T25 = VADD(TJ, TK); TV = LD(&(xi[WS(is, 7)]), ivs, &(xi[WS(is, 1)])); TW = LD(&(xi[WS(is, 23)]), ivs, &(xi[WS(is, 1)])); TX = VSUB(TV, TW); T26 = VADD(TV, TW); } { V TM, TN, TP, TQ; TM = LD(&(xi[WS(is, 3)]), ivs, &(xi[WS(is, 1)])); TN = LD(&(xi[WS(is, 19)]), ivs, &(xi[WS(is, 1)])); TO = VSUB(TM, TN); T28 = VADD(TM, TN); TP = LD(&(xi[WS(is, 27)]), ivs, &(xi[WS(is, 1)])); TQ = LD(&(xi[WS(is, 11)]), ivs, &(xi[WS(is, 1)])); TR = VSUB(TP, TQ); T29 = VADD(TP, TQ); } { V TS, TU, T2F, T2G; TS = VMUL(LDK(KP707106781), VADD(TO, TR)); TT = VADD(TL, TS); T1v = VSUB(TL, TS); TU = VMUL(LDK(KP707106781), VSUB(TR, TO)); TY = VSUB(TU, TX); T1w = VADD(TX, TU); T27 = VADD(T25, T26); T2a = VADD(T28, T29); T2b = VSUB(T27, T2a); T2F = VSUB(T25, T26); T2G = VSUB(T29, T28); T2H = VFNMS(LDK(KP382683432), T2G, VMUL(LDK(KP923879532), T2F)); T2O = VFMA(LDK(KP382683432), T2F, VMUL(LDK(KP923879532), T2G)); } } { V Tu, T1Y, TG, T1Z, Tx, T21, TA, T22; { V Ts, Tt, TE, TF; Ts = LD(&(xi[WS(is, 1)]), ivs, &(xi[WS(is, 1)])); Tt = LD(&(xi[WS(is, 17)]), ivs, &(xi[WS(is, 1)])); Tu = VSUB(Ts, Tt); T1Y = VADD(Ts, Tt); TE = LD(&(xi[WS(is, 9)]), ivs, &(xi[WS(is, 1)])); TF = LD(&(xi[WS(is, 25)]), ivs, &(xi[WS(is, 1)])); TG = VSUB(TE, TF); T1Z = VADD(TE, TF); } { V Tv, Tw, Ty, Tz; Tv = LD(&(xi[WS(is, 5)]), ivs, &(xi[WS(is, 1)])); Tw = LD(&(xi[WS(is, 21)]), ivs, &(xi[WS(is, 1)])); Tx = VSUB(Tv, Tw); T21 = VADD(Tv, Tw); Ty = LD(&(xi[WS(is, 29)]), ivs, &(xi[WS(is, 1)])); Tz = LD(&(xi[WS(is, 13)]), ivs, &(xi[WS(is, 1)])); TA = VSUB(Ty, Tz); T22 = VADD(Ty, Tz); } { V TB, TD, T2C, T2D; TB = VMUL(LDK(KP707106781), VADD(Tx, TA)); TC = VADD(Tu, TB); T1s = VSUB(Tu, TB); TD = VMUL(LDK(KP707106781), VSUB(TA, Tx)); TH = VSUB(TD, TG); T1t = VADD(TG, TD); T20 = VADD(T1Y, T1Z); T23 = VADD(T21, T22); T24 = VSUB(T20, T23); T2C = VSUB(T1Y, T1Z); T2D = VSUB(T22, T21); T2E = VFMA(LDK(KP923879532), T2C, VMUL(LDK(KP382683432), T2D)); T2N = VFNMS(LDK(KP382683432), T2C, VMUL(LDK(KP923879532), T2D)); } } { V Te, T2h, To, T2f, Th, T2i, Tl, T2e, Ti, Tp; { V Tc, Td, Tm, Tn; Tc = LD(&(xi[WS(is, 2)]), ivs, &(xi[0])); Td = LD(&(xi[WS(is, 18)]), ivs, &(xi[0])); Te = VSUB(Tc, Td); T2h = VADD(Tc, Td); Tm = LD(&(xi[WS(is, 6)]), ivs, &(xi[0])); Tn = LD(&(xi[WS(is, 22)]), ivs, &(xi[0])); To = VSUB(Tm, Tn); T2f = VADD(Tm, Tn); } { V Tf, Tg, Tj, Tk; Tf = LD(&(xi[WS(is, 10)]), ivs, &(xi[0])); Tg = LD(&(xi[WS(is, 26)]), ivs, &(xi[0])); Th = VSUB(Tf, Tg); T2i = VADD(Tf, Tg); Tj = LD(&(xi[WS(is, 30)]), ivs, &(xi[0])); Tk = LD(&(xi[WS(is, 14)]), ivs, &(xi[0])); Tl = VSUB(Tj, Tk); T2e = VADD(Tj, Tk); } T2g = VADD(T2e, T2f); T2j = VADD(T2h, T2i); Ti = VFNMS(LDK(KP382683432), Th, VMUL(LDK(KP923879532), Te)); Tp = VFMA(LDK(KP923879532), Tl, VMUL(LDK(KP382683432), To)); Tq = VADD(Ti, Tp); T1B = VSUB(Tp, Ti); { V T17, T18, T2y, T2z; T17 = VFNMS(LDK(KP923879532), To, VMUL(LDK(KP382683432), Tl)); T18 = VFMA(LDK(KP382683432), Te, VMUL(LDK(KP923879532), Th)); T19 = VSUB(T17, T18); T1q = VADD(T18, T17); T2y = VSUB(T2h, T2i); T2z = VSUB(T2e, T2f); T2A = VMUL(LDK(KP707106781), VADD(T2y, T2z)); T2L = VMUL(LDK(KP707106781), VSUB(T2z, T2y)); } } { V T31, T32, T33, T34, T35, T36, T37, T38, T39, T3a, T3b, T3c; { V T2d, T2n, T2m, T2o; { V T1X, T2c, T2k, T2l; T1X = VSUB(T1T, T1W); T2c = VMUL(LDK(KP707106781), VADD(T24, T2b)); T2d = VADD(T1X, T2c); T2n = VSUB(T1X, T2c); T2k = VSUB(T2g, T2j); T2l = VMUL(LDK(KP707106781), VSUB(T2b, T24)); T2m = VBYI(VADD(T2k, T2l)); T2o = VBYI(VSUB(T2l, T2k)); } T31 = VSUB(T2d, T2m); STM2(&(xo[56]), T31, ovs, &(xo[0])); T32 = VADD(T2n, T2o); STM2(&(xo[24]), T32, ovs, &(xo[0])); T33 = VADD(T2d, T2m); STM2(&(xo[8]), T33, ovs, &(xo[0])); T34 = VSUB(T2n, T2o); STM2(&(xo[40]), T34, ovs, &(xo[0])); } { V T2r, T2v, T2u, T2w; { V T2p, T2q, T2s, T2t; T2p = VADD(T1T, T1W); T2q = VADD(T2j, T2g); T2r = VADD(T2p, T2q); T2v = VSUB(T2p, T2q); T2s = VADD(T20, T23); T2t = VADD(T27, T2a); T2u = VADD(T2s, T2t); T2w = VBYI(VSUB(T2t, T2s)); } T35 = VSUB(T2r, T2u); STM2(&(xo[32]), T35, ovs, &(xo[0])); T36 = VADD(T2v, T2w); STM2(&(xo[16]), T36, ovs, &(xo[0])); T37 = VADD(T2r, T2u); STM2(&(xo[0]), T37, ovs, &(xo[0])); T38 = VSUB(T2v, T2w); STM2(&(xo[48]), T38, ovs, &(xo[0])); } { V T2V, T2Z, T2Y, T30; { V T2T, T2U, T2W, T2X; T2T = VSUB(T2H, T2E); T2U = VSUB(T2L, T2K); T2V = VBYI(VSUB(T2T, T2U)); T2Z = VBYI(VADD(T2U, T2T)); T2W = VSUB(T2x, T2A); T2X = VSUB(T2O, T2N); T2Y = VSUB(T2W, T2X); T30 = VADD(T2W, T2X); } T39 = VADD(T2V, T2Y); STM2(&(xo[20]), T39, ovs, &(xo[0])); T3a = VSUB(T30, T2Z); STM2(&(xo[52]), T3a, ovs, &(xo[0])); T3b = VSUB(T2Y, T2V); STM2(&(xo[44]), T3b, ovs, &(xo[0])); T3c = VADD(T2Z, T30); STM2(&(xo[12]), T3c, ovs, &(xo[0])); } { V T3d, T3e, T3f, T3g; { V T2J, T2R, T2Q, T2S; { V T2B, T2I, T2M, T2P; T2B = VADD(T2x, T2A); T2I = VADD(T2E, T2H); T2J = VADD(T2B, T2I); T2R = VSUB(T2B, T2I); T2M = VADD(T2K, T2L); T2P = VADD(T2N, T2O); T2Q = VBYI(VADD(T2M, T2P)); T2S = VBYI(VSUB(T2P, T2M)); } T3d = VSUB(T2J, T2Q); STM2(&(xo[60]), T3d, ovs, &(xo[0])); T3e = VADD(T2R, T2S); STM2(&(xo[28]), T3e, ovs, &(xo[0])); T3f = VADD(T2J, T2Q); STM2(&(xo[4]), T3f, ovs, &(xo[0])); T3g = VSUB(T2R, T2S); STM2(&(xo[36]), T3g, ovs, &(xo[0])); } { V T1r, T1C, T1M, T1K, T1F, T1N, T1y, T1J; T1r = VADD(T1p, T1q); T1C = VADD(T1A, T1B); T1M = VSUB(T1p, T1q); T1K = VSUB(T1B, T1A); { V T1D, T1E, T1u, T1x; T1D = VFNMS(LDK(KP555570233), T1s, VMUL(LDK(KP831469612), T1t)); T1E = VFMA(LDK(KP555570233), T1v, VMUL(LDK(KP831469612), T1w)); T1F = VADD(T1D, T1E); T1N = VSUB(T1E, T1D); T1u = VFMA(LDK(KP831469612), T1s, VMUL(LDK(KP555570233), T1t)); T1x = VFNMS(LDK(KP555570233), T1w, VMUL(LDK(KP831469612), T1v)); T1y = VADD(T1u, T1x); T1J = VSUB(T1x, T1u); } { V T1z, T1G, T3h, T3i; T1z = VADD(T1r, T1y); T1G = VBYI(VADD(T1C, T1F)); T3h = VSUB(T1z, T1G); STM2(&(xo[58]), T3h, ovs, &(xo[2])); STN2(&(xo[56]), T31, T3h, ovs); T3i = VADD(T1z, T1G); STM2(&(xo[6]), T3i, ovs, &(xo[2])); STN2(&(xo[4]), T3f, T3i, ovs); } { V T1P, T1Q, T3j, T3k; T1P = VBYI(VADD(T1K, T1J)); T1Q = VADD(T1M, T1N); T3j = VADD(T1P, T1Q); STM2(&(xo[10]), T3j, ovs, &(xo[2])); STN2(&(xo[8]), T33, T3j, ovs); T3k = VSUB(T1Q, T1P); STM2(&(xo[54]), T3k, ovs, &(xo[2])); STN2(&(xo[52]), T3a, T3k, ovs); } { V T1H, T1I, T3l, T3m; T1H = VSUB(T1r, T1y); T1I = VBYI(VSUB(T1F, T1C)); T3l = VSUB(T1H, T1I); STM2(&(xo[38]), T3l, ovs, &(xo[2])); STN2(&(xo[36]), T3g, T3l, ovs); T3m = VADD(T1H, T1I); STM2(&(xo[26]), T3m, ovs, &(xo[2])); STN2(&(xo[24]), T32, T3m, ovs); } { V T1L, T1O, T3n, T3o; T1L = VBYI(VSUB(T1J, T1K)); T1O = VSUB(T1M, T1N); T3n = VADD(T1L, T1O); STM2(&(xo[22]), T3n, ovs, &(xo[2])); STN2(&(xo[20]), T39, T3n, ovs); T3o = VSUB(T1O, T1L); STM2(&(xo[42]), T3o, ovs, &(xo[2])); STN2(&(xo[40]), T34, T3o, ovs); } } { V Tr, T1a, T1k, T1i, T1d, T1l, T10, T1h; Tr = VADD(Tb, Tq); T1a = VADD(T16, T19); T1k = VSUB(Tb, Tq); T1i = VSUB(T19, T16); { V T1b, T1c, TI, TZ; T1b = VFNMS(LDK(KP195090322), TC, VMUL(LDK(KP980785280), TH)); T1c = VFMA(LDK(KP195090322), TT, VMUL(LDK(KP980785280), TY)); T1d = VADD(T1b, T1c); T1l = VSUB(T1c, T1b); TI = VFMA(LDK(KP980785280), TC, VMUL(LDK(KP195090322), TH)); TZ = VFNMS(LDK(KP195090322), TY, VMUL(LDK(KP980785280), TT)); T10 = VADD(TI, TZ); T1h = VSUB(TZ, TI); } { V T11, T1e, T3p, T3q; T11 = VADD(Tr, T10); T1e = VBYI(VADD(T1a, T1d)); T3p = VSUB(T11, T1e); STM2(&(xo[62]), T3p, ovs, &(xo[2])); STN2(&(xo[60]), T3d, T3p, ovs); T3q = VADD(T11, T1e); STM2(&(xo[2]), T3q, ovs, &(xo[2])); STN2(&(xo[0]), T37, T3q, ovs); } { V T1n, T1o, T3r, T3s; T1n = VBYI(VADD(T1i, T1h)); T1o = VADD(T1k, T1l); T3r = VADD(T1n, T1o); STM2(&(xo[14]), T3r, ovs, &(xo[2])); STN2(&(xo[12]), T3c, T3r, ovs); T3s = VSUB(T1o, T1n); STM2(&(xo[50]), T3s, ovs, &(xo[2])); STN2(&(xo[48]), T38, T3s, ovs); } { V T1f, T1g, T3t, T3u; T1f = VSUB(Tr, T10); T1g = VBYI(VSUB(T1d, T1a)); T3t = VSUB(T1f, T1g); STM2(&(xo[34]), T3t, ovs, &(xo[2])); STN2(&(xo[32]), T35, T3t, ovs); T3u = VADD(T1f, T1g); STM2(&(xo[30]), T3u, ovs, &(xo[2])); STN2(&(xo[28]), T3e, T3u, ovs); } { V T1j, T1m, T3v, T3w; T1j = VBYI(VSUB(T1h, T1i)); T1m = VSUB(T1k, T1l); T3v = VADD(T1j, T1m); STM2(&(xo[18]), T3v, ovs, &(xo[2])); STN2(&(xo[16]), T36, T3v, ovs); T3w = VSUB(T1m, T1j); STM2(&(xo[46]), T3w, ovs, &(xo[2])); STN2(&(xo[44]), T3b, T3w, ovs); } } } } } } VLEAVE(); } static const kdft_desc desc = { 32, XSIMD_STRING("n2fv_32"), {170, 26, 16, 0}, &GENUS, 0, 2, 0, 0 }; void XSIMD(codelet_n2fv_32) (planner *p) { X(kdft_register) (p, n2fv_32, &desc); } #endif /* HAVE_FMA */ fftw-3.3.3/dft/simd/common/n2fv_64.c0000644000175000017500000017301612054410432013743 00000000000000/* * Copyright (c) 2003, 2007-11 Matteo Frigo * Copyright (c) 2003, 2007-11 Massachusetts Institute of Technology * * This program is free software; you can redistribute it and/or modify * it under the terms of the GNU General Public License as published by * the Free Software Foundation; either version 2 of the License, or * (at your option) any later version. * * This program is distributed in the hope that it will be useful, * but WITHOUT ANY WARRANTY; without even the implied warranty of * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the * GNU General Public License for more details. * * You should have received a copy of the GNU General Public License * along with this program; if not, write to the Free Software * Foundation, Inc., 51 Franklin Street, Fifth Floor, Boston, MA 02110-1301 USA * */ /* This file was automatically generated --- DO NOT EDIT */ /* Generated on Sun Nov 25 07:37:25 EST 2012 */ #include "codelet-dft.h" #ifdef HAVE_FMA /* Generated by: ../../../genfft/gen_notw_c.native -fma -reorder-insns -schedule-for-pipeline -simd -compact -variables 4 -pipeline-latency 8 -n 64 -name n2fv_64 -with-ostride 2 -include n2f.h -store-multiple 2 */ /* * This function contains 456 FP additions, 258 FP multiplications, * (or, 198 additions, 0 multiplications, 258 fused multiply/add), * 178 stack variables, 15 constants, and 160 memory accesses */ #include "n2f.h" static void n2fv_64(const R *ri, const R *ii, R *ro, R *io, stride is, stride os, INT v, INT ivs, INT ovs) { DVK(KP773010453, +0.773010453362736960810906609758469800971041293); DVK(KP820678790, +0.820678790828660330972281985331011598767386482); DVK(KP881921264, +0.881921264348355029712756863660388349508442621); DVK(KP534511135, +0.534511135950791641089685961295362908582039528); DVK(KP995184726, +0.995184726672196886244836953109479921575474869); DVK(KP098491403, +0.098491403357164253077197521291327432293052451); DVK(KP980785280, +0.980785280403230449126182236134239036973933731); DVK(KP956940335, +0.956940335732208864935797886980269969482849206); DVK(KP303346683, +0.303346683607342391675883946941299872384187453); DVK(KP831469612, +0.831469612302545237078788377617905756738560812); DVK(KP923879532, +0.923879532511286756128183189396788286822416626); DVK(KP668178637, +0.668178637919298919997757686523080761552472251); DVK(KP198912367, +0.198912367379658006911597622644676228597850501); DVK(KP414213562, +0.414213562373095048801688724209698078569671875); DVK(KP707106781, +0.707106781186547524400844362104849039284835938); { INT i; const R *xi; R *xo; xi = ri; xo = ro; for (i = v; i > 0; i = i - VL, xi = xi + (VL * ivs), xo = xo + (VL * ovs), MAKE_VOLATILE_STRIDE(128, is), MAKE_VOLATILE_STRIDE(128, os)) { V T7r, T7s, T7t, T7u, T5T, T5S, T5X, T65, T8a, T8b, T8e, T8g, T5Z, T5R, T67; V T63, T5U, T64; { V T7, T26, T5k, T6A, T47, T69, T2V, T3z, T6B, T4e, T6a, T5n, T3M, T2Y, T27; V Tm, T3A, T3l, T2a, TC, T5p, T4o, T6E, T6e, T3i, T3B, TR, T29, T4x, T5q; V T6h, T6D, T39, T3H, T3I, T3c, T5N, T57, T72, T6w, T5O, T5e, T71, T6t, T2y; V T1W, T2x, T1N, T33, T34, T3E, T32, T1p, T2v, T1g, T2u, T4M, T5K, T6p, T6Z; V T6m, T6Y, T5L, T4T; { V T4g, T4l, T3j, Tu, Tx, T4h, TA, T4i; { V T1, T2, T23, T24, T4, T5, T20, T21; T1 = LD(&(xi[0]), ivs, &(xi[0])); T2 = LD(&(xi[WS(is, 32)]), ivs, &(xi[0])); T23 = LD(&(xi[WS(is, 8)]), ivs, &(xi[0])); T24 = LD(&(xi[WS(is, 40)]), ivs, &(xi[0])); T4 = LD(&(xi[WS(is, 16)]), ivs, &(xi[0])); T5 = LD(&(xi[WS(is, 48)]), ivs, &(xi[0])); T20 = LD(&(xi[WS(is, 56)]), ivs, &(xi[0])); T21 = LD(&(xi[WS(is, 24)]), ivs, &(xi[0])); { V Ta, T48, Tk, T4c, T49, Td, Tf, Tg; { V T8, T43, T3, T44, T25, T5i, T6, T45, T22, T9, Ti, Tj, Tb, Tc; T8 = LD(&(xi[WS(is, 4)]), ivs, &(xi[0])); T43 = VSUB(T1, T2); T3 = VADD(T1, T2); T44 = VSUB(T23, T24); T25 = VADD(T23, T24); T5i = VSUB(T4, T5); T6 = VADD(T4, T5); T45 = VSUB(T20, T21); T22 = VADD(T20, T21); T9 = LD(&(xi[WS(is, 36)]), ivs, &(xi[0])); Ti = LD(&(xi[WS(is, 12)]), ivs, &(xi[0])); Tj = LD(&(xi[WS(is, 44)]), ivs, &(xi[0])); Tb = LD(&(xi[WS(is, 20)]), ivs, &(xi[0])); Tc = LD(&(xi[WS(is, 52)]), ivs, &(xi[0])); { V T2T, T46, T5j, T2U; T7 = VSUB(T3, T6); T2T = VADD(T3, T6); T46 = VADD(T44, T45); T5j = VSUB(T45, T44); T26 = VSUB(T22, T25); T2U = VADD(T25, T22); Ta = VADD(T8, T9); T48 = VSUB(T8, T9); Tk = VADD(Ti, Tj); T4c = VSUB(Tj, Ti); T5k = VFNMS(LDK(KP707106781), T5j, T5i); T6A = VFMA(LDK(KP707106781), T5j, T5i); T47 = VFMA(LDK(KP707106781), T46, T43); T69 = VFNMS(LDK(KP707106781), T46, T43); T2V = VADD(T2T, T2U); T3z = VSUB(T2T, T2U); T49 = VSUB(Tb, Tc); Td = VADD(Tb, Tc); } Tf = LD(&(xi[WS(is, 60)]), ivs, &(xi[0])); Tg = LD(&(xi[WS(is, 28)]), ivs, &(xi[0])); } { V Te, T2W, T5l, T4a, Tq, Tt, Tv, Tw, T5m, T4d, Tl, T2X, Ty, Tz, To; V Tp; To = LD(&(xi[WS(is, 2)]), ivs, &(xi[0])); Tp = LD(&(xi[WS(is, 34)]), ivs, &(xi[0])); { V Th, T4b, Tr, Ts; Tr = LD(&(xi[WS(is, 18)]), ivs, &(xi[0])); Ts = LD(&(xi[WS(is, 50)]), ivs, &(xi[0])); Te = VSUB(Ta, Td); T2W = VADD(Ta, Td); T5l = VFMA(LDK(KP414213562), T48, T49); T4a = VFNMS(LDK(KP414213562), T49, T48); Th = VADD(Tf, Tg); T4b = VSUB(Tf, Tg); Tq = VADD(To, Tp); T4g = VSUB(To, Tp); T4l = VSUB(Tr, Ts); Tt = VADD(Tr, Ts); Tv = LD(&(xi[WS(is, 10)]), ivs, &(xi[0])); Tw = LD(&(xi[WS(is, 42)]), ivs, &(xi[0])); T5m = VFMA(LDK(KP414213562), T4b, T4c); T4d = VFNMS(LDK(KP414213562), T4c, T4b); Tl = VSUB(Th, Tk); T2X = VADD(Th, Tk); Ty = LD(&(xi[WS(is, 58)]), ivs, &(xi[0])); Tz = LD(&(xi[WS(is, 26)]), ivs, &(xi[0])); } T3j = VADD(Tq, Tt); Tu = VSUB(Tq, Tt); Tx = VADD(Tv, Tw); T4h = VSUB(Tv, Tw); T6B = VSUB(T4d, T4a); T4e = VADD(T4a, T4d); T6a = VADD(T5l, T5m); T5n = VSUB(T5l, T5m); T3M = VSUB(T2X, T2W); T2Y = VADD(T2W, T2X); T27 = VSUB(Tl, Te); Tm = VADD(Te, Tl); TA = VADD(Ty, Tz); T4i = VSUB(Ty, Tz); } } } { V TK, T4p, T4u, T4k, T6d, T4n, T6c, TL, TN, TO, T3g, TJ, TF, TI; { V TD, TE, TG, TH; TD = LD(&(xi[WS(is, 62)]), ivs, &(xi[0])); TE = LD(&(xi[WS(is, 30)]), ivs, &(xi[0])); TG = LD(&(xi[WS(is, 14)]), ivs, &(xi[0])); TH = LD(&(xi[WS(is, 46)]), ivs, &(xi[0])); TK = LD(&(xi[WS(is, 54)]), ivs, &(xi[0])); { V T3k, TB, T4j, T4m; T3k = VADD(Tx, TA); TB = VSUB(Tx, TA); T4j = VADD(T4h, T4i); T4m = VSUB(T4h, T4i); T4p = VSUB(TD, TE); TF = VADD(TD, TE); T4u = VSUB(TH, TG); TI = VADD(TG, TH); T3A = VSUB(T3j, T3k); T3l = VADD(T3j, T3k); T2a = VFMA(LDK(KP414213562), Tu, TB); TC = VFNMS(LDK(KP414213562), TB, Tu); T4k = VFMA(LDK(KP707106781), T4j, T4g); T6d = VFNMS(LDK(KP707106781), T4j, T4g); T4n = VFMA(LDK(KP707106781), T4m, T4l); T6c = VFNMS(LDK(KP707106781), T4m, T4l); TL = LD(&(xi[WS(is, 22)]), ivs, &(xi[0])); } TN = LD(&(xi[WS(is, 6)]), ivs, &(xi[0])); TO = LD(&(xi[WS(is, 38)]), ivs, &(xi[0])); } T3g = VADD(TF, TI); TJ = VSUB(TF, TI); { V T3a, T1E, T52, T5b, T1x, T4Z, T6r, T6u, T5a, T1U, T55, T5c, T1L, T3b; { V T4V, T1t, T58, T1w, T1Q, T1T, T1I, T4Y, T59, T1J, T53, T1H; { V T1r, TM, T4r, TP, T4q, T1s, T1u, T1v; T1r = LD(&(xi[WS(is, 63)]), ivs, &(xi[WS(is, 1)])); T5p = VFMA(LDK(KP198912367), T4k, T4n); T4o = VFNMS(LDK(KP198912367), T4n, T4k); T6E = VFMA(LDK(KP668178637), T6c, T6d); T6e = VFNMS(LDK(KP668178637), T6d, T6c); TM = VADD(TK, TL); T4r = VSUB(TK, TL); TP = VADD(TN, TO); T4q = VSUB(TN, TO); T1s = LD(&(xi[WS(is, 31)]), ivs, &(xi[WS(is, 1)])); T1u = LD(&(xi[WS(is, 15)]), ivs, &(xi[WS(is, 1)])); T1v = LD(&(xi[WS(is, 47)]), ivs, &(xi[WS(is, 1)])); { V T1R, T4X, T6g, T4t, T6f, T4w, T1S, T1O, T1P; T1O = LD(&(xi[WS(is, 55)]), ivs, &(xi[WS(is, 1)])); T1P = LD(&(xi[WS(is, 23)]), ivs, &(xi[WS(is, 1)])); T1R = LD(&(xi[WS(is, 7)]), ivs, &(xi[WS(is, 1)])); { V T3h, TQ, T4s, T4v; T3h = VADD(TP, TM); TQ = VSUB(TM, TP); T4s = VADD(T4q, T4r); T4v = VSUB(T4r, T4q); T4V = VSUB(T1r, T1s); T1t = VADD(T1r, T1s); T58 = VSUB(T1v, T1u); T1w = VADD(T1u, T1v); T4X = VSUB(T1O, T1P); T1Q = VADD(T1O, T1P); T3i = VADD(T3g, T3h); T3B = VSUB(T3g, T3h); TR = VFNMS(LDK(KP414213562), TQ, TJ); T29 = VFMA(LDK(KP414213562), TJ, TQ); T6g = VFNMS(LDK(KP707106781), T4s, T4p); T4t = VFMA(LDK(KP707106781), T4s, T4p); T6f = VFNMS(LDK(KP707106781), T4v, T4u); T4w = VFMA(LDK(KP707106781), T4v, T4u); T1S = LD(&(xi[WS(is, 39)]), ivs, &(xi[WS(is, 1)])); } { V T4W, T1A, T50, T51, T1D, T1F, T1G; { V T1y, T1z, T1B, T1C; T1y = LD(&(xi[WS(is, 3)]), ivs, &(xi[WS(is, 1)])); T1z = LD(&(xi[WS(is, 35)]), ivs, &(xi[WS(is, 1)])); T1B = LD(&(xi[WS(is, 19)]), ivs, &(xi[WS(is, 1)])); T1C = LD(&(xi[WS(is, 51)]), ivs, &(xi[WS(is, 1)])); T4x = VFNMS(LDK(KP198912367), T4w, T4t); T5q = VFMA(LDK(KP198912367), T4t, T4w); T6h = VFNMS(LDK(KP668178637), T6g, T6f); T6D = VFMA(LDK(KP668178637), T6f, T6g); T4W = VSUB(T1R, T1S); T1T = VADD(T1R, T1S); T1A = VADD(T1y, T1z); T50 = VSUB(T1y, T1z); T51 = VSUB(T1C, T1B); T1D = VADD(T1B, T1C); } T1F = LD(&(xi[WS(is, 59)]), ivs, &(xi[WS(is, 1)])); T1G = LD(&(xi[WS(is, 27)]), ivs, &(xi[WS(is, 1)])); T1I = LD(&(xi[WS(is, 11)]), ivs, &(xi[WS(is, 1)])); T4Y = VADD(T4W, T4X); T59 = VSUB(T4X, T4W); T1J = LD(&(xi[WS(is, 43)]), ivs, &(xi[WS(is, 1)])); T3a = VADD(T1A, T1D); T1E = VSUB(T1A, T1D); T52 = VFMA(LDK(KP414213562), T51, T50); T5b = VFNMS(LDK(KP414213562), T50, T51); T53 = VSUB(T1F, T1G); T1H = VADD(T1F, T1G); } } } { V T37, T54, T1K, T38; T1x = VSUB(T1t, T1w); T37 = VADD(T1t, T1w); T4Z = VFMA(LDK(KP707106781), T4Y, T4V); T6r = VFNMS(LDK(KP707106781), T4Y, T4V); T54 = VSUB(T1J, T1I); T1K = VADD(T1I, T1J); T6u = VFNMS(LDK(KP707106781), T59, T58); T5a = VFMA(LDK(KP707106781), T59, T58); T38 = VADD(T1T, T1Q); T1U = VSUB(T1Q, T1T); T55 = VFNMS(LDK(KP414213562), T54, T53); T5c = VFMA(LDK(KP414213562), T53, T54); T1L = VSUB(T1H, T1K); T3b = VADD(T1H, T1K); T39 = VADD(T37, T38); T3H = VSUB(T37, T38); } } { V T4A, TW, T4N, TZ, T1j, T1m, T4O, T4D, T13, T4F, T16, T4G, T1a, T4I, T4J; V T1d; { V TU, TV, TX, TY, T56, T6v; TU = LD(&(xi[WS(is, 1)]), ivs, &(xi[WS(is, 1)])); T56 = VADD(T52, T55); T6v = VSUB(T55, T52); { V T5d, T6s, T1V, T1M; T5d = VADD(T5b, T5c); T6s = VSUB(T5c, T5b); T1V = VSUB(T1L, T1E); T1M = VADD(T1E, T1L); T3I = VSUB(T3b, T3a); T3c = VADD(T3a, T3b); T5N = VFNMS(LDK(KP923879532), T56, T4Z); T57 = VFMA(LDK(KP923879532), T56, T4Z); T72 = VFNMS(LDK(KP923879532), T6v, T6u); T6w = VFMA(LDK(KP923879532), T6v, T6u); T5O = VFNMS(LDK(KP923879532), T5d, T5a); T5e = VFMA(LDK(KP923879532), T5d, T5a); T71 = VFMA(LDK(KP923879532), T6s, T6r); T6t = VFNMS(LDK(KP923879532), T6s, T6r); T2y = VFNMS(LDK(KP707106781), T1V, T1U); T1W = VFMA(LDK(KP707106781), T1V, T1U); T2x = VFNMS(LDK(KP707106781), T1M, T1x); T1N = VFMA(LDK(KP707106781), T1M, T1x); TV = LD(&(xi[WS(is, 33)]), ivs, &(xi[WS(is, 1)])); } TX = LD(&(xi[WS(is, 17)]), ivs, &(xi[WS(is, 1)])); TY = LD(&(xi[WS(is, 49)]), ivs, &(xi[WS(is, 1)])); { V T1h, T1i, T1k, T1l; T1h = LD(&(xi[WS(is, 9)]), ivs, &(xi[WS(is, 1)])); T1i = LD(&(xi[WS(is, 41)]), ivs, &(xi[WS(is, 1)])); T1k = LD(&(xi[WS(is, 57)]), ivs, &(xi[WS(is, 1)])); T1l = LD(&(xi[WS(is, 25)]), ivs, &(xi[WS(is, 1)])); { V T11, T4B, T4C, T12, T14, T15; T11 = LD(&(xi[WS(is, 5)]), ivs, &(xi[WS(is, 1)])); T4A = VSUB(TU, TV); TW = VADD(TU, TV); T4N = VSUB(TX, TY); TZ = VADD(TX, TY); T1j = VADD(T1h, T1i); T4B = VSUB(T1h, T1i); T1m = VADD(T1k, T1l); T4C = VSUB(T1k, T1l); T12 = LD(&(xi[WS(is, 37)]), ivs, &(xi[WS(is, 1)])); T14 = LD(&(xi[WS(is, 21)]), ivs, &(xi[WS(is, 1)])); T15 = LD(&(xi[WS(is, 53)]), ivs, &(xi[WS(is, 1)])); { V T18, T19, T1b, T1c; T18 = LD(&(xi[WS(is, 61)]), ivs, &(xi[WS(is, 1)])); T19 = LD(&(xi[WS(is, 29)]), ivs, &(xi[WS(is, 1)])); T1b = LD(&(xi[WS(is, 13)]), ivs, &(xi[WS(is, 1)])); T1c = LD(&(xi[WS(is, 45)]), ivs, &(xi[WS(is, 1)])); T4O = VSUB(T4B, T4C); T4D = VADD(T4B, T4C); T13 = VADD(T11, T12); T4F = VSUB(T11, T12); T16 = VADD(T14, T15); T4G = VSUB(T14, T15); T1a = VADD(T18, T19); T4I = VSUB(T18, T19); T4J = VSUB(T1b, T1c); T1d = VADD(T1b, T1c); } } } } { V T30, T10, T6k, T4E, T4Q, T4H, T17, T6n, T4P, T1e, T4K, T4R, T1n, T31; T30 = VADD(TW, TZ); T10 = VSUB(TW, TZ); T6k = VFNMS(LDK(KP707106781), T4D, T4A); T4E = VFMA(LDK(KP707106781), T4D, T4A); T4Q = VFMA(LDK(KP414213562), T4F, T4G); T4H = VFNMS(LDK(KP414213562), T4G, T4F); T33 = VADD(T13, T16); T17 = VSUB(T13, T16); T6n = VFNMS(LDK(KP707106781), T4O, T4N); T4P = VFMA(LDK(KP707106781), T4O, T4N); T34 = VADD(T1a, T1d); T1e = VSUB(T1a, T1d); T4K = VFMA(LDK(KP414213562), T4J, T4I); T4R = VFNMS(LDK(KP414213562), T4I, T4J); T1n = VSUB(T1j, T1m); T31 = VADD(T1j, T1m); { V T1f, T1o, T6o, T4L, T4S, T6l; T1f = VADD(T17, T1e); T1o = VSUB(T17, T1e); T6o = VSUB(T4H, T4K); T4L = VADD(T4H, T4K); T4S = VADD(T4Q, T4R); T6l = VSUB(T4Q, T4R); T3E = VSUB(T30, T31); T32 = VADD(T30, T31); T1p = VFMA(LDK(KP707106781), T1o, T1n); T2v = VFNMS(LDK(KP707106781), T1o, T1n); T1g = VFMA(LDK(KP707106781), T1f, T10); T2u = VFNMS(LDK(KP707106781), T1f, T10); T4M = VFMA(LDK(KP923879532), T4L, T4E); T5K = VFNMS(LDK(KP923879532), T4L, T4E); T6p = VFMA(LDK(KP923879532), T6o, T6n); T6Z = VFNMS(LDK(KP923879532), T6o, T6n); T6m = VFNMS(LDK(KP923879532), T6l, T6k); T6Y = VFMA(LDK(KP923879532), T6l, T6k); T5L = VFNMS(LDK(KP923879532), T4S, T4P); T4T = VFMA(LDK(KP923879532), T4S, T4P); } } } } } } { V T6b, T6F, T7n, T7o, T7p, T7q, T7v, T7w, T7x, T7y, T7z, T7A, T7B, T7C, T7f; V T6X, T70, T79, T7a, T73, T6C, T76, T77, T6i; { V T2Z, T3r, T3s, T3m, T3d, T3v; T2Z = VSUB(T2V, T2Y); T3r = VADD(T2V, T2Y); T3s = VADD(T3l, T3i); T3m = VSUB(T3i, T3l); T3d = VSUB(T39, T3c); T3v = VADD(T39, T3c); { V T3x, T3t, T3P, T3J, T3D, T3V, T3Q, T3G, T36, T3u, T3Y, T3O, T6V, T6W; { V T3N, T3C, T3F, T35; T3N = VSUB(T3B, T3A); T3C = VADD(T3A, T3B); T3F = VSUB(T33, T34); T35 = VADD(T33, T34); T3x = VSUB(T3r, T3s); T3t = VADD(T3r, T3s); T3P = VFMA(LDK(KP414213562), T3H, T3I); T3J = VFNMS(LDK(KP414213562), T3I, T3H); T3D = VFMA(LDK(KP707106781), T3C, T3z); T3V = VFNMS(LDK(KP707106781), T3C, T3z); T3Q = VFMA(LDK(KP414213562), T3E, T3F); T3G = VFNMS(LDK(KP414213562), T3F, T3E); T36 = VSUB(T32, T35); T3u = VADD(T32, T35); T3Y = VFNMS(LDK(KP707106781), T3N, T3M); T3O = VFMA(LDK(KP707106781), T3N, T3M); } T6b = VFNMS(LDK(KP923879532), T6a, T69); T6V = VFMA(LDK(KP923879532), T6a, T69); T6W = VADD(T6E, T6D); T6F = VSUB(T6D, T6E); { V T3K, T3Z, T3e, T3n; T3K = VADD(T3G, T3J); T3Z = VSUB(T3J, T3G); T3e = VADD(T36, T3d); T3n = VSUB(T3d, T36); { V T3w, T3y, T3R, T3W; T3w = VADD(T3u, T3v); T3y = VSUB(T3v, T3u); T3R = VSUB(T3P, T3Q); T3W = VADD(T3Q, T3P); { V T42, T40, T3L, T3T; T42 = VFNMS(LDK(KP923879532), T3Z, T3Y); T40 = VFMA(LDK(KP923879532), T3Z, T3Y); T3L = VFNMS(LDK(KP923879532), T3K, T3D); T3T = VFMA(LDK(KP923879532), T3K, T3D); { V T3o, T3q, T3f, T3p; T3o = VFNMS(LDK(KP707106781), T3n, T3m); T3q = VFMA(LDK(KP707106781), T3n, T3m); T3f = VFNMS(LDK(KP707106781), T3e, T2Z); T3p = VFMA(LDK(KP707106781), T3e, T2Z); T7n = VFNMSI(T3y, T3x); STM2(&(xo[96]), T7n, ovs, &(xo[0])); T7o = VFMAI(T3y, T3x); STM2(&(xo[32]), T7o, ovs, &(xo[0])); T7p = VADD(T3t, T3w); STM2(&(xo[0]), T7p, ovs, &(xo[0])); T7q = VSUB(T3t, T3w); STM2(&(xo[64]), T7q, ovs, &(xo[0])); { V T41, T3X, T3S, T3U; T41 = VFMA(LDK(KP923879532), T3W, T3V); T3X = VFNMS(LDK(KP923879532), T3W, T3V); T3S = VFNMS(LDK(KP923879532), T3R, T3O); T3U = VFMA(LDK(KP923879532), T3R, T3O); T7r = VFMAI(T3q, T3p); STM2(&(xo[16]), T7r, ovs, &(xo[0])); T7s = VFNMSI(T3q, T3p); STM2(&(xo[112]), T7s, ovs, &(xo[0])); T7t = VFMAI(T3o, T3f); STM2(&(xo[80]), T7t, ovs, &(xo[0])); T7u = VFNMSI(T3o, T3f); STM2(&(xo[48]), T7u, ovs, &(xo[0])); T7v = VFNMSI(T40, T3X); STM2(&(xo[88]), T7v, ovs, &(xo[0])); T7w = VFMAI(T40, T3X); STM2(&(xo[40]), T7w, ovs, &(xo[0])); T7x = VFMAI(T42, T41); STM2(&(xo[104]), T7x, ovs, &(xo[0])); T7y = VFNMSI(T42, T41); STM2(&(xo[24]), T7y, ovs, &(xo[0])); T7z = VFMAI(T3U, T3T); STM2(&(xo[8]), T7z, ovs, &(xo[0])); T7A = VFNMSI(T3U, T3T); STM2(&(xo[120]), T7A, ovs, &(xo[0])); T7B = VFMAI(T3S, T3L); STM2(&(xo[72]), T7B, ovs, &(xo[0])); T7C = VFNMSI(T3S, T3L); STM2(&(xo[56]), T7C, ovs, &(xo[0])); T7f = VFNMS(LDK(KP831469612), T6W, T6V); T6X = VFMA(LDK(KP831469612), T6W, T6V); } } } } } T70 = VFMA(LDK(KP303346683), T6Z, T6Y); T79 = VFNMS(LDK(KP303346683), T6Y, T6Z); T7a = VFNMS(LDK(KP303346683), T71, T72); T73 = VFMA(LDK(KP303346683), T72, T71); T6C = VFNMS(LDK(KP923879532), T6B, T6A); T76 = VFMA(LDK(KP923879532), T6B, T6A); T77 = VSUB(T6e, T6h); T6i = VADD(T6e, T6h); } } { V T2r, T2D, T2C, T2s, T5H, T5o, T5v, T5D, T7L, T7O, T7Q, T7S, T5r, T5I, T5x; V T5h, T5F, T5B; { V TT, T2f, T7E, T7F, T7H, T7J, T2n, T1Y, T28, T2b, T2l, T2p, T2j, T2k; { V T1X, T2d, T7h, T7l, T2e, T1q, T75, T7d, T7m, T7k, T7c, T7e, Tn, TS; T2r = VFNMS(LDK(KP707106781), Tm, T7); Tn = VFMA(LDK(KP707106781), Tm, T7); TS = VADD(TC, TR); T2D = VSUB(TR, TC); { V T7b, T7j, T74, T7i, T78, T7g; T1X = VFNMS(LDK(KP198912367), T1W, T1N); T2d = VFMA(LDK(KP198912367), T1N, T1W); T7g = VADD(T79, T7a); T7b = VSUB(T79, T7a); T7j = VSUB(T73, T70); T74 = VADD(T70, T73); T7i = VFNMS(LDK(KP831469612), T77, T76); T78 = VFMA(LDK(KP831469612), T77, T76); T2j = VFNMS(LDK(KP923879532), TS, Tn); TT = VFMA(LDK(KP923879532), TS, Tn); T7h = VFMA(LDK(KP956940335), T7g, T7f); T7l = VFNMS(LDK(KP956940335), T7g, T7f); T2e = VFMA(LDK(KP198912367), T1g, T1p); T1q = VFNMS(LDK(KP198912367), T1p, T1g); T75 = VFNMS(LDK(KP956940335), T74, T6X); T7d = VFMA(LDK(KP956940335), T74, T6X); T7m = VFNMS(LDK(KP956940335), T7j, T7i); T7k = VFMA(LDK(KP956940335), T7j, T7i); T7c = VFNMS(LDK(KP956940335), T7b, T78); T7e = VFMA(LDK(KP956940335), T7b, T78); } T2k = VADD(T2e, T2d); T2f = VSUB(T2d, T2e); { V T7D, T7G, T7I, T7K; T7D = VFNMSI(T7k, T7h); STM2(&(xo[90]), T7D, ovs, &(xo[2])); STN2(&(xo[88]), T7v, T7D, ovs); T7E = VFMAI(T7k, T7h); STM2(&(xo[38]), T7E, ovs, &(xo[2])); T7F = VFMAI(T7m, T7l); STM2(&(xo[102]), T7F, ovs, &(xo[2])); T7G = VFNMSI(T7m, T7l); STM2(&(xo[26]), T7G, ovs, &(xo[2])); STN2(&(xo[24]), T7y, T7G, ovs); T7H = VFMAI(T7e, T7d); STM2(&(xo[6]), T7H, ovs, &(xo[2])); T7I = VFNMSI(T7e, T7d); STM2(&(xo[122]), T7I, ovs, &(xo[2])); STN2(&(xo[120]), T7A, T7I, ovs); T7J = VFMAI(T7c, T75); STM2(&(xo[70]), T7J, ovs, &(xo[2])); T7K = VFNMSI(T7c, T75); STM2(&(xo[58]), T7K, ovs, &(xo[2])); STN2(&(xo[56]), T7C, T7K, ovs); T2n = VSUB(T1X, T1q); T1Y = VADD(T1q, T1X); } T2C = VFNMS(LDK(KP707106781), T27, T26); T28 = VFMA(LDK(KP707106781), T27, T26); T2b = VSUB(T29, T2a); T2s = VADD(T2a, T29); } T2l = VFNMS(LDK(KP980785280), T2k, T2j); T2p = VFMA(LDK(KP980785280), T2k, T2j); { V T5z, T4z, T5A, T5g; { V T4f, T4y, T1Z, T2h, T4U, T5t, T2m, T2c, T5u, T5f; T5H = VFNMS(LDK(KP923879532), T4e, T47); T4f = VFMA(LDK(KP923879532), T4e, T47); T4y = VADD(T4o, T4x); T5T = VSUB(T4x, T4o); T1Z = VFNMS(LDK(KP980785280), T1Y, TT); T2h = VFMA(LDK(KP980785280), T1Y, TT); T4U = VFNMS(LDK(KP098491403), T4T, T4M); T5t = VFMA(LDK(KP098491403), T4M, T4T); T2m = VFNMS(LDK(KP923879532), T2b, T28); T2c = VFMA(LDK(KP923879532), T2b, T28); T5u = VFMA(LDK(KP098491403), T57, T5e); T5f = VFNMS(LDK(KP098491403), T5e, T57); T5z = VFNMS(LDK(KP980785280), T4y, T4f); T4z = VFMA(LDK(KP980785280), T4y, T4f); T5S = VFNMS(LDK(KP923879532), T5n, T5k); T5o = VFMA(LDK(KP923879532), T5n, T5k); { V T2o, T2q, T2i, T2g; T2o = VFMA(LDK(KP980785280), T2n, T2m); T2q = VFNMS(LDK(KP980785280), T2n, T2m); T2i = VFMA(LDK(KP980785280), T2f, T2c); T2g = VFNMS(LDK(KP980785280), T2f, T2c); T5A = VADD(T5t, T5u); T5v = VSUB(T5t, T5u); T5D = VSUB(T5f, T4U); T5g = VADD(T4U, T5f); T7L = VFNMSI(T2o, T2l); STM2(&(xo[92]), T7L, ovs, &(xo[0])); { V T7M, T7N, T7P, T7R; T7M = VFMAI(T2o, T2l); STM2(&(xo[36]), T7M, ovs, &(xo[0])); STN2(&(xo[36]), T7M, T7E, ovs); T7N = VFMAI(T2q, T2p); STM2(&(xo[100]), T7N, ovs, &(xo[0])); STN2(&(xo[100]), T7N, T7F, ovs); T7O = VFNMSI(T2q, T2p); STM2(&(xo[28]), T7O, ovs, &(xo[0])); T7P = VFMAI(T2i, T2h); STM2(&(xo[4]), T7P, ovs, &(xo[0])); STN2(&(xo[4]), T7P, T7H, ovs); T7Q = VFNMSI(T2i, T2h); STM2(&(xo[124]), T7Q, ovs, &(xo[0])); T7R = VFMAI(T2g, T1Z); STM2(&(xo[68]), T7R, ovs, &(xo[0])); STN2(&(xo[68]), T7R, T7J, ovs); T7S = VFNMSI(T2g, T1Z); STM2(&(xo[60]), T7S, ovs, &(xo[0])); T5r = VSUB(T5p, T5q); T5I = VADD(T5p, T5q); } } } T5x = VFMA(LDK(KP995184726), T5g, T4z); T5h = VFNMS(LDK(KP995184726), T5g, T4z); T5F = VFMA(LDK(KP995184726), T5A, T5z); T5B = VFNMS(LDK(KP995184726), T5A, T5z); } } { V T6J, T6R, T6L, T6z, T6T, T6P; { V T6N, T6j, T6O, T6y; { V T6q, T6H, T5C, T5s, T6I, T6x; T6q = VFNMS(LDK(KP534511135), T6p, T6m); T6H = VFMA(LDK(KP534511135), T6m, T6p); T5C = VFNMS(LDK(KP980785280), T5r, T5o); T5s = VFMA(LDK(KP980785280), T5r, T5o); T6I = VFMA(LDK(KP534511135), T6t, T6w); T6x = VFNMS(LDK(KP534511135), T6w, T6t); T6N = VFMA(LDK(KP831469612), T6i, T6b); T6j = VFNMS(LDK(KP831469612), T6i, T6b); { V T5E, T5G, T5y, T5w; T5E = VFNMS(LDK(KP995184726), T5D, T5C); T5G = VFMA(LDK(KP995184726), T5D, T5C); T5y = VFMA(LDK(KP995184726), T5v, T5s); T5w = VFNMS(LDK(KP995184726), T5v, T5s); T6O = VADD(T6H, T6I); T6J = VSUB(T6H, T6I); T6R = VSUB(T6x, T6q); T6y = VADD(T6q, T6x); { V T7T, T7U, T7V, T7W; T7T = VFMAI(T5E, T5B); STM2(&(xo[94]), T7T, ovs, &(xo[2])); STN2(&(xo[92]), T7L, T7T, ovs); T7U = VFNMSI(T5E, T5B); STM2(&(xo[34]), T7U, ovs, &(xo[2])); STN2(&(xo[32]), T7o, T7U, ovs); T7V = VFNMSI(T5G, T5F); STM2(&(xo[98]), T7V, ovs, &(xo[2])); STN2(&(xo[96]), T7n, T7V, ovs); T7W = VFMAI(T5G, T5F); STM2(&(xo[30]), T7W, ovs, &(xo[2])); STN2(&(xo[28]), T7O, T7W, ovs); { V T7X, T7Y, T7Z, T80; T7X = VFMAI(T5y, T5x); STM2(&(xo[126]), T7X, ovs, &(xo[2])); STN2(&(xo[124]), T7Q, T7X, ovs); T7Y = VFNMSI(T5y, T5x); STM2(&(xo[2]), T7Y, ovs, &(xo[2])); STN2(&(xo[0]), T7p, T7Y, ovs); T7Z = VFMAI(T5w, T5h); STM2(&(xo[62]), T7Z, ovs, &(xo[2])); STN2(&(xo[60]), T7S, T7Z, ovs); T80 = VFNMSI(T5w, T5h); STM2(&(xo[66]), T80, ovs, &(xo[2])); STN2(&(xo[64]), T7q, T80, ovs); } } } } T6L = VFMA(LDK(KP881921264), T6y, T6j); T6z = VFNMS(LDK(KP881921264), T6y, T6j); T6T = VFMA(LDK(KP881921264), T6O, T6N); T6P = VFNMS(LDK(KP881921264), T6O, T6N); } { V T2H, T2P, T81, T84, T85, T87, T2J, T2B, T2R, T2N; { V T2L, T2t, T2M, T2A; { V T2z, T2F, T6Q, T6G, T2G, T2w; T2z = VFMA(LDK(KP668178637), T2y, T2x); T2F = VFNMS(LDK(KP668178637), T2x, T2y); T6Q = VFMA(LDK(KP831469612), T6F, T6C); T6G = VFNMS(LDK(KP831469612), T6F, T6C); T2G = VFNMS(LDK(KP668178637), T2u, T2v); T2w = VFMA(LDK(KP668178637), T2v, T2u); T2L = VFNMS(LDK(KP923879532), T2s, T2r); T2t = VFMA(LDK(KP923879532), T2s, T2r); { V T6S, T6U, T6M, T6K; T6S = VFNMS(LDK(KP881921264), T6R, T6Q); T6U = VFMA(LDK(KP881921264), T6R, T6Q); T6M = VFMA(LDK(KP881921264), T6J, T6G); T6K = VFNMS(LDK(KP881921264), T6J, T6G); T2M = VADD(T2G, T2F); T2H = VSUB(T2F, T2G); T2P = VSUB(T2z, T2w); T2A = VADD(T2w, T2z); T81 = VFMAI(T6S, T6P); STM2(&(xo[86]), T81, ovs, &(xo[2])); { V T82, T83, T86, T88; T82 = VFNMSI(T6S, T6P); STM2(&(xo[42]), T82, ovs, &(xo[2])); STN2(&(xo[40]), T7w, T82, ovs); T83 = VFNMSI(T6U, T6T); STM2(&(xo[106]), T83, ovs, &(xo[2])); STN2(&(xo[104]), T7x, T83, ovs); T84 = VFMAI(T6U, T6T); STM2(&(xo[22]), T84, ovs, &(xo[2])); T85 = VFMAI(T6M, T6L); STM2(&(xo[118]), T85, ovs, &(xo[2])); T86 = VFNMSI(T6M, T6L); STM2(&(xo[10]), T86, ovs, &(xo[2])); STN2(&(xo[8]), T7z, T86, ovs); T87 = VFMAI(T6K, T6z); STM2(&(xo[54]), T87, ovs, &(xo[2])); T88 = VFNMSI(T6K, T6z); STM2(&(xo[74]), T88, ovs, &(xo[2])); STN2(&(xo[72]), T7B, T88, ovs); } } } T2J = VFMA(LDK(KP831469612), T2A, T2t); T2B = VFNMS(LDK(KP831469612), T2A, T2t); T2R = VFNMS(LDK(KP831469612), T2M, T2L); T2N = VFMA(LDK(KP831469612), T2M, T2L); } { V T61, T5J, T62, T5Q; { V T5M, T5V, T2O, T2E, T5W, T5P; T5M = VFMA(LDK(KP820678790), T5L, T5K); T5V = VFNMS(LDK(KP820678790), T5K, T5L); T2O = VFMA(LDK(KP923879532), T2D, T2C); T2E = VFNMS(LDK(KP923879532), T2D, T2C); T5W = VFNMS(LDK(KP820678790), T5N, T5O); T5P = VFMA(LDK(KP820678790), T5O, T5N); T61 = VFNMS(LDK(KP980785280), T5I, T5H); T5J = VFMA(LDK(KP980785280), T5I, T5H); { V T2Q, T2S, T2K, T2I; T2Q = VFNMS(LDK(KP831469612), T2P, T2O); T2S = VFMA(LDK(KP831469612), T2P, T2O); T2K = VFMA(LDK(KP831469612), T2H, T2E); T2I = VFNMS(LDK(KP831469612), T2H, T2E); T62 = VADD(T5V, T5W); T5X = VSUB(T5V, T5W); T65 = VSUB(T5P, T5M); T5Q = VADD(T5M, T5P); { V T89, T8c, T8d, T8f; T89 = VFMAI(T2Q, T2N); STM2(&(xo[84]), T89, ovs, &(xo[0])); STN2(&(xo[84]), T89, T81, ovs); T8a = VFNMSI(T2Q, T2N); STM2(&(xo[44]), T8a, ovs, &(xo[0])); T8b = VFNMSI(T2S, T2R); STM2(&(xo[108]), T8b, ovs, &(xo[0])); T8c = VFMAI(T2S, T2R); STM2(&(xo[20]), T8c, ovs, &(xo[0])); STN2(&(xo[20]), T8c, T84, ovs); T8d = VFMAI(T2K, T2J); STM2(&(xo[116]), T8d, ovs, &(xo[0])); STN2(&(xo[116]), T8d, T85, ovs); T8e = VFNMSI(T2K, T2J); STM2(&(xo[12]), T8e, ovs, &(xo[0])); T8f = VFMAI(T2I, T2B); STM2(&(xo[52]), T8f, ovs, &(xo[0])); STN2(&(xo[52]), T8f, T87, ovs); T8g = VFNMSI(T2I, T2B); STM2(&(xo[76]), T8g, ovs, &(xo[0])); } } } T5Z = VFMA(LDK(KP773010453), T5Q, T5J); T5R = VFNMS(LDK(KP773010453), T5Q, T5J); T67 = VFNMS(LDK(KP773010453), T62, T61); T63 = VFMA(LDK(KP773010453), T62, T61); } } } } } } T5U = VFMA(LDK(KP980785280), T5T, T5S); T64 = VFNMS(LDK(KP980785280), T5T, T5S); { V T68, T66, T5Y, T60; T68 = VFNMS(LDK(KP773010453), T65, T64); T66 = VFMA(LDK(KP773010453), T65, T64); T5Y = VFNMS(LDK(KP773010453), T5X, T5U); T60 = VFMA(LDK(KP773010453), T5X, T5U); { V T8h, T8i, T8j, T8k; T8h = VFNMSI(T66, T63); STM2(&(xo[82]), T8h, ovs, &(xo[2])); STN2(&(xo[80]), T7t, T8h, ovs); T8i = VFMAI(T66, T63); STM2(&(xo[46]), T8i, ovs, &(xo[2])); STN2(&(xo[44]), T8a, T8i, ovs); T8j = VFMAI(T68, T67); STM2(&(xo[110]), T8j, ovs, &(xo[2])); STN2(&(xo[108]), T8b, T8j, ovs); T8k = VFNMSI(T68, T67); STM2(&(xo[18]), T8k, ovs, &(xo[2])); STN2(&(xo[16]), T7r, T8k, ovs); { V T8l, T8m, T8n, T8o; T8l = VFMAI(T60, T5Z); STM2(&(xo[14]), T8l, ovs, &(xo[2])); STN2(&(xo[12]), T8e, T8l, ovs); T8m = VFNMSI(T60, T5Z); STM2(&(xo[114]), T8m, ovs, &(xo[2])); STN2(&(xo[112]), T7s, T8m, ovs); T8n = VFMAI(T5Y, T5R); STM2(&(xo[78]), T8n, ovs, &(xo[2])); STN2(&(xo[76]), T8g, T8n, ovs); T8o = VFNMSI(T5Y, T5R); STM2(&(xo[50]), T8o, ovs, &(xo[2])); STN2(&(xo[48]), T7u, T8o, ovs); } } } } } VLEAVE(); } static const kdft_desc desc = { 64, XSIMD_STRING("n2fv_64"), {198, 0, 258, 0}, &GENUS, 0, 2, 0, 0 }; void XSIMD(codelet_n2fv_64) (planner *p) { X(kdft_register) (p, n2fv_64, &desc); } #else /* HAVE_FMA */ /* Generated by: ../../../genfft/gen_notw_c.native -simd -compact -variables 4 -pipeline-latency 8 -n 64 -name n2fv_64 -with-ostride 2 -include n2f.h -store-multiple 2 */ /* * This function contains 456 FP additions, 124 FP multiplications, * (or, 404 additions, 72 multiplications, 52 fused multiply/add), * 128 stack variables, 15 constants, and 160 memory accesses */ #include "n2f.h" static void n2fv_64(const R *ri, const R *ii, R *ro, R *io, stride is, stride os, INT v, INT ivs, INT ovs) { DVK(KP471396736, +0.471396736825997648556387625905254377657460319); DVK(KP881921264, +0.881921264348355029712756863660388349508442621); DVK(KP098017140, +0.098017140329560601994195563888641845861136673); DVK(KP995184726, +0.995184726672196886244836953109479921575474869); DVK(KP290284677, +0.290284677254462367636192375817395274691476278); DVK(KP956940335, +0.956940335732208864935797886980269969482849206); DVK(KP634393284, +0.634393284163645498215171613225493370675687095); DVK(KP773010453, +0.773010453362736960810906609758469800971041293); DVK(KP555570233, +0.555570233019602224742830813948532874374937191); DVK(KP831469612, +0.831469612302545237078788377617905756738560812); DVK(KP980785280, +0.980785280403230449126182236134239036973933731); DVK(KP195090322, +0.195090322016128267848284868477022240927691618); DVK(KP382683432, +0.382683432365089771728459984030398866761344562); DVK(KP923879532, +0.923879532511286756128183189396788286822416626); DVK(KP707106781, +0.707106781186547524400844362104849039284835938); { INT i; const R *xi; R *xo; xi = ri; xo = ro; for (i = v; i > 0; i = i - VL, xi = xi + (VL * ivs), xo = xo + (VL * ovs), MAKE_VOLATILE_STRIDE(128, is), MAKE_VOLATILE_STRIDE(128, os)) { V T4p, T5q, Tb, T39, T2n, T3A, T6f, T6T, Tq, T3B, T6i, T76, T2i, T3a, T4w; V T5r, TI, T2p, T6C, T6V, T3h, T3E, T4L, T5u, TZ, T2q, T6F, T6U, T3e, T3D; V T4E, T5t, T23, T2N, T6t, T71, T6w, T72, T2c, T2O, T3t, T41, T5f, T5R, T5k; V T5S, T3w, T42, T1s, T2K, T6m, T6Y, T6p, T6Z, T1B, T2L, T3m, T3Y, T4Y, T5O; V T53, T5P, T3p, T3Z; { V T3, T4n, T2m, T4o, T6, T5p, T9, T5o; { V T1, T2, T2k, T2l; T1 = LD(&(xi[0]), ivs, &(xi[0])); T2 = LD(&(xi[WS(is, 32)]), ivs, &(xi[0])); T3 = VSUB(T1, T2); T4n = VADD(T1, T2); T2k = LD(&(xi[WS(is, 16)]), ivs, &(xi[0])); T2l = LD(&(xi[WS(is, 48)]), ivs, &(xi[0])); T2m = VSUB(T2k, T2l); T4o = VADD(T2k, T2l); } { V T4, T5, T7, T8; T4 = LD(&(xi[WS(is, 8)]), ivs, &(xi[0])); T5 = LD(&(xi[WS(is, 40)]), ivs, &(xi[0])); T6 = VSUB(T4, T5); T5p = VADD(T4, T5); T7 = LD(&(xi[WS(is, 56)]), ivs, &(xi[0])); T8 = LD(&(xi[WS(is, 24)]), ivs, &(xi[0])); T9 = VSUB(T7, T8); T5o = VADD(T7, T8); } T4p = VSUB(T4n, T4o); T5q = VSUB(T5o, T5p); { V Ta, T2j, T6d, T6e; Ta = VMUL(LDK(KP707106781), VADD(T6, T9)); Tb = VADD(T3, Ta); T39 = VSUB(T3, Ta); T2j = VMUL(LDK(KP707106781), VSUB(T9, T6)); T2n = VSUB(T2j, T2m); T3A = VADD(T2m, T2j); T6d = VADD(T4n, T4o); T6e = VADD(T5p, T5o); T6f = VADD(T6d, T6e); T6T = VSUB(T6d, T6e); } } { V Te, T4q, To, T4u, Th, T4r, Tl, T4t; { V Tc, Td, Tm, Tn; Tc = LD(&(xi[WS(is, 4)]), ivs, &(xi[0])); Td = LD(&(xi[WS(is, 36)]), ivs, &(xi[0])); Te = VSUB(Tc, Td); T4q = VADD(Tc, Td); Tm = LD(&(xi[WS(is, 12)]), ivs, &(xi[0])); Tn = LD(&(xi[WS(is, 44)]), ivs, &(xi[0])); To = VSUB(Tm, Tn); T4u = VADD(Tm, Tn); } { V Tf, Tg, Tj, Tk; Tf = LD(&(xi[WS(is, 20)]), ivs, &(xi[0])); Tg = LD(&(xi[WS(is, 52)]), ivs, &(xi[0])); Th = VSUB(Tf, Tg); T4r = VADD(Tf, Tg); Tj = LD(&(xi[WS(is, 60)]), ivs, &(xi[0])); Tk = LD(&(xi[WS(is, 28)]), ivs, &(xi[0])); Tl = VSUB(Tj, Tk); T4t = VADD(Tj, Tk); } { V Ti, Tp, T6g, T6h; Ti = VFNMS(LDK(KP382683432), Th, VMUL(LDK(KP923879532), Te)); Tp = VFMA(LDK(KP923879532), Tl, VMUL(LDK(KP382683432), To)); Tq = VADD(Ti, Tp); T3B = VSUB(Tp, Ti); T6g = VADD(T4q, T4r); T6h = VADD(T4t, T4u); T6i = VADD(T6g, T6h); T76 = VSUB(T6h, T6g); } { V T2g, T2h, T4s, T4v; T2g = VFNMS(LDK(KP923879532), To, VMUL(LDK(KP382683432), Tl)); T2h = VFMA(LDK(KP382683432), Te, VMUL(LDK(KP923879532), Th)); T2i = VSUB(T2g, T2h); T3a = VADD(T2h, T2g); T4s = VSUB(T4q, T4r); T4v = VSUB(T4t, T4u); T4w = VMUL(LDK(KP707106781), VADD(T4s, T4v)); T5r = VMUL(LDK(KP707106781), VSUB(T4v, T4s)); } } { V Tu, T4F, TG, T4G, TB, T4J, TD, T4I; { V Ts, Tt, TE, TF; Ts = LD(&(xi[WS(is, 62)]), ivs, &(xi[0])); Tt = LD(&(xi[WS(is, 30)]), ivs, &(xi[0])); Tu = VSUB(Ts, Tt); T4F = VADD(Ts, Tt); TE = LD(&(xi[WS(is, 14)]), ivs, &(xi[0])); TF = LD(&(xi[WS(is, 46)]), ivs, &(xi[0])); TG = VSUB(TE, TF); T4G = VADD(TE, TF); { V Tv, Tw, Tx, Ty, Tz, TA; Tv = LD(&(xi[WS(is, 6)]), ivs, &(xi[0])); Tw = LD(&(xi[WS(is, 38)]), ivs, &(xi[0])); Tx = VSUB(Tv, Tw); Ty = LD(&(xi[WS(is, 54)]), ivs, &(xi[0])); Tz = LD(&(xi[WS(is, 22)]), ivs, &(xi[0])); TA = VSUB(Ty, Tz); TB = VMUL(LDK(KP707106781), VADD(Tx, TA)); T4J = VADD(Tv, Tw); TD = VMUL(LDK(KP707106781), VSUB(TA, Tx)); T4I = VADD(Ty, Tz); } } { V TC, TH, T6A, T6B; TC = VADD(Tu, TB); TH = VSUB(TD, TG); TI = VFMA(LDK(KP195090322), TC, VMUL(LDK(KP980785280), TH)); T2p = VFNMS(LDK(KP195090322), TH, VMUL(LDK(KP980785280), TC)); T6A = VADD(T4F, T4G); T6B = VADD(T4J, T4I); T6C = VADD(T6A, T6B); T6V = VSUB(T6A, T6B); } { V T3f, T3g, T4H, T4K; T3f = VSUB(Tu, TB); T3g = VADD(TG, TD); T3h = VFNMS(LDK(KP555570233), T3g, VMUL(LDK(KP831469612), T3f)); T3E = VFMA(LDK(KP555570233), T3f, VMUL(LDK(KP831469612), T3g)); T4H = VSUB(T4F, T4G); T4K = VSUB(T4I, T4J); T4L = VFNMS(LDK(KP382683432), T4K, VMUL(LDK(KP923879532), T4H)); T5u = VFMA(LDK(KP382683432), T4H, VMUL(LDK(KP923879532), T4K)); } } { V TS, T4z, TW, T4y, TP, T4C, TX, T4B; { V TQ, TR, TU, TV; TQ = LD(&(xi[WS(is, 18)]), ivs, &(xi[0])); TR = LD(&(xi[WS(is, 50)]), ivs, &(xi[0])); TS = VSUB(TQ, TR); T4z = VADD(TQ, TR); TU = LD(&(xi[WS(is, 2)]), ivs, &(xi[0])); TV = LD(&(xi[WS(is, 34)]), ivs, &(xi[0])); TW = VSUB(TU, TV); T4y = VADD(TU, TV); { V TJ, TK, TL, TM, TN, TO; TJ = LD(&(xi[WS(is, 58)]), ivs, &(xi[0])); TK = LD(&(xi[WS(is, 26)]), ivs, &(xi[0])); TL = VSUB(TJ, TK); TM = LD(&(xi[WS(is, 10)]), ivs, &(xi[0])); TN = LD(&(xi[WS(is, 42)]), ivs, &(xi[0])); TO = VSUB(TM, TN); TP = VMUL(LDK(KP707106781), VSUB(TL, TO)); T4C = VADD(TM, TN); TX = VMUL(LDK(KP707106781), VADD(TO, TL)); T4B = VADD(TJ, TK); } } { V TT, TY, T6D, T6E; TT = VSUB(TP, TS); TY = VADD(TW, TX); TZ = VFNMS(LDK(KP195090322), TY, VMUL(LDK(KP980785280), TT)); T2q = VFMA(LDK(KP980785280), TY, VMUL(LDK(KP195090322), TT)); T6D = VADD(T4y, T4z); T6E = VADD(T4C, T4B); T6F = VADD(T6D, T6E); T6U = VSUB(T6D, T6E); } { V T3c, T3d, T4A, T4D; T3c = VSUB(TW, TX); T3d = VADD(TS, TP); T3e = VFMA(LDK(KP831469612), T3c, VMUL(LDK(KP555570233), T3d)); T3D = VFNMS(LDK(KP555570233), T3c, VMUL(LDK(KP831469612), T3d)); T4A = VSUB(T4y, T4z); T4D = VSUB(T4B, T4C); T4E = VFMA(LDK(KP923879532), T4A, VMUL(LDK(KP382683432), T4D)); T5t = VFNMS(LDK(KP382683432), T4A, VMUL(LDK(KP923879532), T4D)); } } { V T1F, T55, T2a, T56, T1M, T5h, T27, T5g, T58, T59, T1U, T5a, T25, T5b, T5c; V T21, T5d, T24; { V T1D, T1E, T28, T29; T1D = LD(&(xi[WS(is, 63)]), ivs, &(xi[WS(is, 1)])); T1E = LD(&(xi[WS(is, 31)]), ivs, &(xi[WS(is, 1)])); T1F = VSUB(T1D, T1E); T55 = VADD(T1D, T1E); T28 = LD(&(xi[WS(is, 15)]), ivs, &(xi[WS(is, 1)])); T29 = LD(&(xi[WS(is, 47)]), ivs, &(xi[WS(is, 1)])); T2a = VSUB(T28, T29); T56 = VADD(T28, T29); } { V T1G, T1H, T1I, T1J, T1K, T1L; T1G = LD(&(xi[WS(is, 7)]), ivs, &(xi[WS(is, 1)])); T1H = LD(&(xi[WS(is, 39)]), ivs, &(xi[WS(is, 1)])); T1I = VSUB(T1G, T1H); T1J = LD(&(xi[WS(is, 55)]), ivs, &(xi[WS(is, 1)])); T1K = LD(&(xi[WS(is, 23)]), ivs, &(xi[WS(is, 1)])); T1L = VSUB(T1J, T1K); T1M = VMUL(LDK(KP707106781), VADD(T1I, T1L)); T5h = VADD(T1G, T1H); T27 = VMUL(LDK(KP707106781), VSUB(T1L, T1I)); T5g = VADD(T1J, T1K); } { V T1Q, T1T, T1X, T20; { V T1O, T1P, T1R, T1S; T1O = LD(&(xi[WS(is, 3)]), ivs, &(xi[WS(is, 1)])); T1P = LD(&(xi[WS(is, 35)]), ivs, &(xi[WS(is, 1)])); T1Q = VSUB(T1O, T1P); T58 = VADD(T1O, T1P); T1R = LD(&(xi[WS(is, 19)]), ivs, &(xi[WS(is, 1)])); T1S = LD(&(xi[WS(is, 51)]), ivs, &(xi[WS(is, 1)])); T1T = VSUB(T1R, T1S); T59 = VADD(T1R, T1S); } T1U = VFNMS(LDK(KP382683432), T1T, VMUL(LDK(KP923879532), T1Q)); T5a = VSUB(T58, T59); T25 = VFMA(LDK(KP382683432), T1Q, VMUL(LDK(KP923879532), T1T)); { V T1V, T1W, T1Y, T1Z; T1V = LD(&(xi[WS(is, 59)]), ivs, &(xi[WS(is, 1)])); T1W = LD(&(xi[WS(is, 27)]), ivs, &(xi[WS(is, 1)])); T1X = VSUB(T1V, T1W); T5b = VADD(T1V, T1W); T1Y = LD(&(xi[WS(is, 11)]), ivs, &(xi[WS(is, 1)])); T1Z = LD(&(xi[WS(is, 43)]), ivs, &(xi[WS(is, 1)])); T20 = VSUB(T1Y, T1Z); T5c = VADD(T1Y, T1Z); } T21 = VFMA(LDK(KP923879532), T1X, VMUL(LDK(KP382683432), T20)); T5d = VSUB(T5b, T5c); T24 = VFNMS(LDK(KP923879532), T20, VMUL(LDK(KP382683432), T1X)); } { V T1N, T22, T6r, T6s; T1N = VADD(T1F, T1M); T22 = VADD(T1U, T21); T23 = VSUB(T1N, T22); T2N = VADD(T1N, T22); T6r = VADD(T55, T56); T6s = VADD(T5h, T5g); T6t = VADD(T6r, T6s); T71 = VSUB(T6r, T6s); } { V T6u, T6v, T26, T2b; T6u = VADD(T58, T59); T6v = VADD(T5b, T5c); T6w = VADD(T6u, T6v); T72 = VSUB(T6v, T6u); T26 = VSUB(T24, T25); T2b = VSUB(T27, T2a); T2c = VSUB(T26, T2b); T2O = VADD(T2b, T26); } { V T3r, T3s, T57, T5e; T3r = VSUB(T1F, T1M); T3s = VADD(T25, T24); T3t = VADD(T3r, T3s); T41 = VSUB(T3r, T3s); T57 = VSUB(T55, T56); T5e = VMUL(LDK(KP707106781), VADD(T5a, T5d)); T5f = VADD(T57, T5e); T5R = VSUB(T57, T5e); } { V T5i, T5j, T3u, T3v; T5i = VSUB(T5g, T5h); T5j = VMUL(LDK(KP707106781), VSUB(T5d, T5a)); T5k = VADD(T5i, T5j); T5S = VSUB(T5j, T5i); T3u = VADD(T2a, T27); T3v = VSUB(T21, T1U); T3w = VADD(T3u, T3v); T42 = VSUB(T3v, T3u); } } { V T1q, T4P, T1v, T4O, T1n, T50, T1w, T4Z, T4U, T4V, T18, T4W, T1z, T4R, T4S; V T1f, T4T, T1y; { V T1o, T1p, T1t, T1u; T1o = LD(&(xi[WS(is, 17)]), ivs, &(xi[WS(is, 1)])); T1p = LD(&(xi[WS(is, 49)]), ivs, &(xi[WS(is, 1)])); T1q = VSUB(T1o, T1p); T4P = VADD(T1o, T1p); T1t = LD(&(xi[WS(is, 1)]), ivs, &(xi[WS(is, 1)])); T1u = LD(&(xi[WS(is, 33)]), ivs, &(xi[WS(is, 1)])); T1v = VSUB(T1t, T1u); T4O = VADD(T1t, T1u); } { V T1h, T1i, T1j, T1k, T1l, T1m; T1h = LD(&(xi[WS(is, 57)]), ivs, &(xi[WS(is, 1)])); T1i = LD(&(xi[WS(is, 25)]), ivs, &(xi[WS(is, 1)])); T1j = VSUB(T1h, T1i); T1k = LD(&(xi[WS(is, 9)]), ivs, &(xi[WS(is, 1)])); T1l = LD(&(xi[WS(is, 41)]), ivs, &(xi[WS(is, 1)])); T1m = VSUB(T1k, T1l); T1n = VMUL(LDK(KP707106781), VSUB(T1j, T1m)); T50 = VADD(T1k, T1l); T1w = VMUL(LDK(KP707106781), VADD(T1m, T1j)); T4Z = VADD(T1h, T1i); } { V T14, T17, T1b, T1e; { V T12, T13, T15, T16; T12 = LD(&(xi[WS(is, 61)]), ivs, &(xi[WS(is, 1)])); T13 = LD(&(xi[WS(is, 29)]), ivs, &(xi[WS(is, 1)])); T14 = VSUB(T12, T13); T4U = VADD(T12, T13); T15 = LD(&(xi[WS(is, 13)]), ivs, &(xi[WS(is, 1)])); T16 = LD(&(xi[WS(is, 45)]), ivs, &(xi[WS(is, 1)])); T17 = VSUB(T15, T16); T4V = VADD(T15, T16); } T18 = VFNMS(LDK(KP923879532), T17, VMUL(LDK(KP382683432), T14)); T4W = VSUB(T4U, T4V); T1z = VFMA(LDK(KP923879532), T14, VMUL(LDK(KP382683432), T17)); { V T19, T1a, T1c, T1d; T19 = LD(&(xi[WS(is, 5)]), ivs, &(xi[WS(is, 1)])); T1a = LD(&(xi[WS(is, 37)]), ivs, &(xi[WS(is, 1)])); T1b = VSUB(T19, T1a); T4R = VADD(T19, T1a); T1c = LD(&(xi[WS(is, 21)]), ivs, &(xi[WS(is, 1)])); T1d = LD(&(xi[WS(is, 53)]), ivs, &(xi[WS(is, 1)])); T1e = VSUB(T1c, T1d); T4S = VADD(T1c, T1d); } T1f = VFMA(LDK(KP382683432), T1b, VMUL(LDK(KP923879532), T1e)); T4T = VSUB(T4R, T4S); T1y = VFNMS(LDK(KP382683432), T1e, VMUL(LDK(KP923879532), T1b)); } { V T1g, T1r, T6k, T6l; T1g = VSUB(T18, T1f); T1r = VSUB(T1n, T1q); T1s = VSUB(T1g, T1r); T2K = VADD(T1r, T1g); T6k = VADD(T4O, T4P); T6l = VADD(T50, T4Z); T6m = VADD(T6k, T6l); T6Y = VSUB(T6k, T6l); } { V T6n, T6o, T1x, T1A; T6n = VADD(T4R, T4S); T6o = VADD(T4U, T4V); T6p = VADD(T6n, T6o); T6Z = VSUB(T6o, T6n); T1x = VADD(T1v, T1w); T1A = VADD(T1y, T1z); T1B = VSUB(T1x, T1A); T2L = VADD(T1x, T1A); } { V T3k, T3l, T4Q, T4X; T3k = VSUB(T1v, T1w); T3l = VADD(T1f, T18); T3m = VADD(T3k, T3l); T3Y = VSUB(T3k, T3l); T4Q = VSUB(T4O, T4P); T4X = VMUL(LDK(KP707106781), VADD(T4T, T4W)); T4Y = VADD(T4Q, T4X); T5O = VSUB(T4Q, T4X); } { V T51, T52, T3n, T3o; T51 = VSUB(T4Z, T50); T52 = VMUL(LDK(KP707106781), VSUB(T4W, T4T)); T53 = VADD(T51, T52); T5P = VSUB(T52, T51); T3n = VADD(T1q, T1n); T3o = VSUB(T1z, T1y); T3p = VADD(T3n, T3o); T3Z = VSUB(T3o, T3n); } } { V T7n, T7o, T7p, T7q, T7r, T7s, T7t, T7u, T7v, T7w, T7x, T7y, T7z, T7A, T7B; V T7C, T7D, T7E, T7F, T7G, T7H, T7I, T7J, T7K; { V T6N, T6R, T6Q, T6S; { V T6L, T6M, T6O, T6P; T6L = VADD(T6f, T6i); T6M = VADD(T6F, T6C); T6N = VADD(T6L, T6M); T6R = VSUB(T6L, T6M); T6O = VADD(T6m, T6p); T6P = VADD(T6t, T6w); T6Q = VADD(T6O, T6P); T6S = VBYI(VSUB(T6P, T6O)); } T7n = VSUB(T6N, T6Q); STM2(&(xo[64]), T7n, ovs, &(xo[0])); T7o = VADD(T6R, T6S); STM2(&(xo[32]), T7o, ovs, &(xo[0])); T7p = VADD(T6N, T6Q); STM2(&(xo[0]), T7p, ovs, &(xo[0])); T7q = VSUB(T6R, T6S); STM2(&(xo[96]), T7q, ovs, &(xo[0])); } { V T6j, T6G, T6y, T6H, T6q, T6x; T6j = VSUB(T6f, T6i); T6G = VSUB(T6C, T6F); T6q = VSUB(T6m, T6p); T6x = VSUB(T6t, T6w); T6y = VMUL(LDK(KP707106781), VADD(T6q, T6x)); T6H = VMUL(LDK(KP707106781), VSUB(T6x, T6q)); { V T6z, T6I, T6J, T6K; T6z = VADD(T6j, T6y); T6I = VBYI(VADD(T6G, T6H)); T7r = VSUB(T6z, T6I); STM2(&(xo[112]), T7r, ovs, &(xo[0])); T7s = VADD(T6z, T6I); STM2(&(xo[16]), T7s, ovs, &(xo[0])); T6J = VSUB(T6j, T6y); T6K = VBYI(VSUB(T6H, T6G)); T7t = VSUB(T6J, T6K); STM2(&(xo[80]), T7t, ovs, &(xo[0])); T7u = VADD(T6J, T6K); STM2(&(xo[48]), T7u, ovs, &(xo[0])); } } { V T6X, T7i, T78, T7g, T74, T7f, T7b, T7j, T6W, T77; T6W = VMUL(LDK(KP707106781), VADD(T6U, T6V)); T6X = VADD(T6T, T6W); T7i = VSUB(T6T, T6W); T77 = VMUL(LDK(KP707106781), VSUB(T6V, T6U)); T78 = VADD(T76, T77); T7g = VSUB(T77, T76); { V T70, T73, T79, T7a; T70 = VFMA(LDK(KP923879532), T6Y, VMUL(LDK(KP382683432), T6Z)); T73 = VFNMS(LDK(KP382683432), T72, VMUL(LDK(KP923879532), T71)); T74 = VADD(T70, T73); T7f = VSUB(T73, T70); T79 = VFNMS(LDK(KP382683432), T6Y, VMUL(LDK(KP923879532), T6Z)); T7a = VFMA(LDK(KP382683432), T71, VMUL(LDK(KP923879532), T72)); T7b = VADD(T79, T7a); T7j = VSUB(T7a, T79); } { V T75, T7c, T7l, T7m; T75 = VADD(T6X, T74); T7c = VBYI(VADD(T78, T7b)); T7v = VSUB(T75, T7c); STM2(&(xo[120]), T7v, ovs, &(xo[0])); T7w = VADD(T75, T7c); STM2(&(xo[8]), T7w, ovs, &(xo[0])); T7l = VBYI(VADD(T7g, T7f)); T7m = VADD(T7i, T7j); T7x = VADD(T7l, T7m); STM2(&(xo[24]), T7x, ovs, &(xo[0])); T7y = VSUB(T7m, T7l); STM2(&(xo[104]), T7y, ovs, &(xo[0])); } { V T7d, T7e, T7h, T7k; T7d = VSUB(T6X, T74); T7e = VBYI(VSUB(T7b, T78)); T7z = VSUB(T7d, T7e); STM2(&(xo[72]), T7z, ovs, &(xo[0])); T7A = VADD(T7d, T7e); STM2(&(xo[56]), T7A, ovs, &(xo[0])); T7h = VBYI(VSUB(T7f, T7g)); T7k = VSUB(T7i, T7j); T7B = VADD(T7h, T7k); STM2(&(xo[40]), T7B, ovs, &(xo[0])); T7C = VSUB(T7k, T7h); STM2(&(xo[88]), T7C, ovs, &(xo[0])); } } { V T5N, T68, T61, T69, T5U, T65, T5Y, T66; { V T5L, T5M, T5Z, T60; T5L = VSUB(T4p, T4w); T5M = VSUB(T5u, T5t); T5N = VADD(T5L, T5M); T68 = VSUB(T5L, T5M); T5Z = VFNMS(LDK(KP555570233), T5O, VMUL(LDK(KP831469612), T5P)); T60 = VFMA(LDK(KP555570233), T5R, VMUL(LDK(KP831469612), T5S)); T61 = VADD(T5Z, T60); T69 = VSUB(T60, T5Z); } { V T5Q, T5T, T5W, T5X; T5Q = VFMA(LDK(KP831469612), T5O, VMUL(LDK(KP555570233), T5P)); T5T = VFNMS(LDK(KP555570233), T5S, VMUL(LDK(KP831469612), T5R)); T5U = VADD(T5Q, T5T); T65 = VSUB(T5T, T5Q); T5W = VSUB(T5r, T5q); T5X = VSUB(T4L, T4E); T5Y = VADD(T5W, T5X); T66 = VSUB(T5X, T5W); } { V T5V, T62, T6b, T6c; T5V = VADD(T5N, T5U); T62 = VBYI(VADD(T5Y, T61)); T7D = VSUB(T5V, T62); STM2(&(xo[116]), T7D, ovs, &(xo[0])); T7E = VADD(T5V, T62); STM2(&(xo[12]), T7E, ovs, &(xo[0])); T6b = VBYI(VADD(T66, T65)); T6c = VADD(T68, T69); T7F = VADD(T6b, T6c); STM2(&(xo[20]), T7F, ovs, &(xo[0])); T7G = VSUB(T6c, T6b); STM2(&(xo[108]), T7G, ovs, &(xo[0])); } { V T63, T64, T67, T6a; T63 = VSUB(T5N, T5U); T64 = VBYI(VSUB(T61, T5Y)); T7H = VSUB(T63, T64); STM2(&(xo[76]), T7H, ovs, &(xo[0])); T7I = VADD(T63, T64); STM2(&(xo[52]), T7I, ovs, &(xo[0])); T67 = VBYI(VSUB(T65, T66)); T6a = VSUB(T68, T69); T7J = VADD(T67, T6a); STM2(&(xo[44]), T7J, ovs, &(xo[0])); T7K = VSUB(T6a, T67); STM2(&(xo[84]), T7K, ovs, &(xo[0])); } } { V T7U, T7W, T7X, T7Z; { V T11, T2C, T2v, T2D, T2e, T2z, T2s, T2A; { V Tr, T10, T2t, T2u; Tr = VSUB(Tb, Tq); T10 = VSUB(TI, TZ); T11 = VADD(Tr, T10); T2C = VSUB(Tr, T10); T2t = VFNMS(LDK(KP634393284), T1B, VMUL(LDK(KP773010453), T1s)); T2u = VFMA(LDK(KP773010453), T2c, VMUL(LDK(KP634393284), T23)); T2v = VADD(T2t, T2u); T2D = VSUB(T2u, T2t); } { V T1C, T2d, T2o, T2r; T1C = VFMA(LDK(KP634393284), T1s, VMUL(LDK(KP773010453), T1B)); T2d = VFNMS(LDK(KP634393284), T2c, VMUL(LDK(KP773010453), T23)); T2e = VADD(T1C, T2d); T2z = VSUB(T2d, T1C); T2o = VSUB(T2i, T2n); T2r = VSUB(T2p, T2q); T2s = VADD(T2o, T2r); T2A = VSUB(T2r, T2o); } { V T2f, T2w, T7L, T7M; T2f = VADD(T11, T2e); T2w = VBYI(VADD(T2s, T2v)); T7L = VSUB(T2f, T2w); STM2(&(xo[114]), T7L, ovs, &(xo[2])); STN2(&(xo[112]), T7r, T7L, ovs); T7M = VADD(T2f, T2w); STM2(&(xo[14]), T7M, ovs, &(xo[2])); STN2(&(xo[12]), T7E, T7M, ovs); } { V T2F, T2G, T7N, T7O; T2F = VBYI(VADD(T2A, T2z)); T2G = VADD(T2C, T2D); T7N = VADD(T2F, T2G); STM2(&(xo[18]), T7N, ovs, &(xo[2])); STN2(&(xo[16]), T7s, T7N, ovs); T7O = VSUB(T2G, T2F); STM2(&(xo[110]), T7O, ovs, &(xo[2])); STN2(&(xo[108]), T7G, T7O, ovs); } { V T2x, T2y, T7P, T7Q; T2x = VSUB(T11, T2e); T2y = VBYI(VSUB(T2v, T2s)); T7P = VSUB(T2x, T2y); STM2(&(xo[78]), T7P, ovs, &(xo[2])); STN2(&(xo[76]), T7H, T7P, ovs); T7Q = VADD(T2x, T2y); STM2(&(xo[50]), T7Q, ovs, &(xo[2])); STN2(&(xo[48]), T7u, T7Q, ovs); } { V T2B, T2E, T7R, T7S; T2B = VBYI(VSUB(T2z, T2A)); T2E = VSUB(T2C, T2D); T7R = VADD(T2B, T2E); STM2(&(xo[46]), T7R, ovs, &(xo[2])); STN2(&(xo[44]), T7J, T7R, ovs); T7S = VSUB(T2E, T2B); STM2(&(xo[82]), T7S, ovs, &(xo[2])); STN2(&(xo[80]), T7t, T7S, ovs); } } { V T3j, T3Q, T3J, T3R, T3y, T3N, T3G, T3O; { V T3b, T3i, T3H, T3I; T3b = VADD(T39, T3a); T3i = VADD(T3e, T3h); T3j = VADD(T3b, T3i); T3Q = VSUB(T3b, T3i); T3H = VFNMS(LDK(KP290284677), T3m, VMUL(LDK(KP956940335), T3p)); T3I = VFMA(LDK(KP290284677), T3t, VMUL(LDK(KP956940335), T3w)); T3J = VADD(T3H, T3I); T3R = VSUB(T3I, T3H); } { V T3q, T3x, T3C, T3F; T3q = VFMA(LDK(KP956940335), T3m, VMUL(LDK(KP290284677), T3p)); T3x = VFNMS(LDK(KP290284677), T3w, VMUL(LDK(KP956940335), T3t)); T3y = VADD(T3q, T3x); T3N = VSUB(T3x, T3q); T3C = VADD(T3A, T3B); T3F = VADD(T3D, T3E); T3G = VADD(T3C, T3F); T3O = VSUB(T3F, T3C); } { V T3z, T3K, T7T, T3T, T3U, T7V; T3z = VADD(T3j, T3y); T3K = VBYI(VADD(T3G, T3J)); T7T = VSUB(T3z, T3K); STM2(&(xo[122]), T7T, ovs, &(xo[2])); STN2(&(xo[120]), T7v, T7T, ovs); T7U = VADD(T3z, T3K); STM2(&(xo[6]), T7U, ovs, &(xo[2])); T3T = VBYI(VADD(T3O, T3N)); T3U = VADD(T3Q, T3R); T7V = VADD(T3T, T3U); STM2(&(xo[26]), T7V, ovs, &(xo[2])); STN2(&(xo[24]), T7x, T7V, ovs); T7W = VSUB(T3U, T3T); STM2(&(xo[102]), T7W, ovs, &(xo[2])); } { V T3L, T3M, T7Y, T3P, T3S, T80; T3L = VSUB(T3j, T3y); T3M = VBYI(VSUB(T3J, T3G)); T7X = VSUB(T3L, T3M); STM2(&(xo[70]), T7X, ovs, &(xo[2])); T7Y = VADD(T3L, T3M); STM2(&(xo[58]), T7Y, ovs, &(xo[2])); STN2(&(xo[56]), T7A, T7Y, ovs); T3P = VBYI(VSUB(T3N, T3O)); T3S = VSUB(T3Q, T3R); T7Z = VADD(T3P, T3S); STM2(&(xo[38]), T7Z, ovs, &(xo[2])); T80 = VSUB(T3S, T3P); STM2(&(xo[90]), T80, ovs, &(xo[2])); STN2(&(xo[88]), T7C, T80, ovs); } } { V T81, T83, T86, T88; { V T4N, T5G, T5z, T5H, T5m, T5D, T5w, T5E; { V T4x, T4M, T5x, T5y; T4x = VADD(T4p, T4w); T4M = VADD(T4E, T4L); T4N = VADD(T4x, T4M); T5G = VSUB(T4x, T4M); T5x = VFNMS(LDK(KP195090322), T4Y, VMUL(LDK(KP980785280), T53)); T5y = VFMA(LDK(KP195090322), T5f, VMUL(LDK(KP980785280), T5k)); T5z = VADD(T5x, T5y); T5H = VSUB(T5y, T5x); } { V T54, T5l, T5s, T5v; T54 = VFMA(LDK(KP980785280), T4Y, VMUL(LDK(KP195090322), T53)); T5l = VFNMS(LDK(KP195090322), T5k, VMUL(LDK(KP980785280), T5f)); T5m = VADD(T54, T5l); T5D = VSUB(T5l, T54); T5s = VADD(T5q, T5r); T5v = VADD(T5t, T5u); T5w = VADD(T5s, T5v); T5E = VSUB(T5v, T5s); } { V T5n, T5A, T82, T5J, T5K, T84; T5n = VADD(T4N, T5m); T5A = VBYI(VADD(T5w, T5z)); T81 = VSUB(T5n, T5A); STM2(&(xo[124]), T81, ovs, &(xo[0])); T82 = VADD(T5n, T5A); STM2(&(xo[4]), T82, ovs, &(xo[0])); STN2(&(xo[4]), T82, T7U, ovs); T5J = VBYI(VADD(T5E, T5D)); T5K = VADD(T5G, T5H); T83 = VADD(T5J, T5K); STM2(&(xo[28]), T83, ovs, &(xo[0])); T84 = VSUB(T5K, T5J); STM2(&(xo[100]), T84, ovs, &(xo[0])); STN2(&(xo[100]), T84, T7W, ovs); } { V T5B, T5C, T85, T5F, T5I, T87; T5B = VSUB(T4N, T5m); T5C = VBYI(VSUB(T5z, T5w)); T85 = VSUB(T5B, T5C); STM2(&(xo[68]), T85, ovs, &(xo[0])); STN2(&(xo[68]), T85, T7X, ovs); T86 = VADD(T5B, T5C); STM2(&(xo[60]), T86, ovs, &(xo[0])); T5F = VBYI(VSUB(T5D, T5E)); T5I = VSUB(T5G, T5H); T87 = VADD(T5F, T5I); STM2(&(xo[36]), T87, ovs, &(xo[0])); STN2(&(xo[36]), T87, T7Z, ovs); T88 = VSUB(T5I, T5F); STM2(&(xo[92]), T88, ovs, &(xo[0])); } } { V T2J, T34, T2X, T35, T2Q, T31, T2U, T32; { V T2H, T2I, T2V, T2W; T2H = VADD(Tb, Tq); T2I = VADD(T2q, T2p); T2J = VADD(T2H, T2I); T34 = VSUB(T2H, T2I); T2V = VFNMS(LDK(KP098017140), T2L, VMUL(LDK(KP995184726), T2K)); T2W = VFMA(LDK(KP995184726), T2O, VMUL(LDK(KP098017140), T2N)); T2X = VADD(T2V, T2W); T35 = VSUB(T2W, T2V); } { V T2M, T2P, T2S, T2T; T2M = VFMA(LDK(KP098017140), T2K, VMUL(LDK(KP995184726), T2L)); T2P = VFNMS(LDK(KP098017140), T2O, VMUL(LDK(KP995184726), T2N)); T2Q = VADD(T2M, T2P); T31 = VSUB(T2P, T2M); T2S = VADD(T2n, T2i); T2T = VADD(TZ, TI); T2U = VADD(T2S, T2T); T32 = VSUB(T2T, T2S); } { V T2R, T2Y, T89, T8a; T2R = VADD(T2J, T2Q); T2Y = VBYI(VADD(T2U, T2X)); T89 = VSUB(T2R, T2Y); STM2(&(xo[126]), T89, ovs, &(xo[2])); STN2(&(xo[124]), T81, T89, ovs); T8a = VADD(T2R, T2Y); STM2(&(xo[2]), T8a, ovs, &(xo[2])); STN2(&(xo[0]), T7p, T8a, ovs); } { V T37, T38, T8b, T8c; T37 = VBYI(VADD(T32, T31)); T38 = VADD(T34, T35); T8b = VADD(T37, T38); STM2(&(xo[30]), T8b, ovs, &(xo[2])); STN2(&(xo[28]), T83, T8b, ovs); T8c = VSUB(T38, T37); STM2(&(xo[98]), T8c, ovs, &(xo[2])); STN2(&(xo[96]), T7q, T8c, ovs); } { V T2Z, T30, T8d, T8e; T2Z = VSUB(T2J, T2Q); T30 = VBYI(VSUB(T2X, T2U)); T8d = VSUB(T2Z, T30); STM2(&(xo[66]), T8d, ovs, &(xo[2])); STN2(&(xo[64]), T7n, T8d, ovs); T8e = VADD(T2Z, T30); STM2(&(xo[62]), T8e, ovs, &(xo[2])); STN2(&(xo[60]), T86, T8e, ovs); } { V T33, T36, T8f, T8g; T33 = VBYI(VSUB(T31, T32)); T36 = VSUB(T34, T35); T8f = VADD(T33, T36); STM2(&(xo[34]), T8f, ovs, &(xo[2])); STN2(&(xo[32]), T7o, T8f, ovs); T8g = VSUB(T36, T33); STM2(&(xo[94]), T8g, ovs, &(xo[2])); STN2(&(xo[92]), T88, T8g, ovs); } } { V T3X, T4i, T4b, T4j, T44, T4f, T48, T4g; { V T3V, T3W, T49, T4a; T3V = VSUB(T39, T3a); T3W = VSUB(T3E, T3D); T3X = VADD(T3V, T3W); T4i = VSUB(T3V, T3W); T49 = VFNMS(LDK(KP471396736), T3Y, VMUL(LDK(KP881921264), T3Z)); T4a = VFMA(LDK(KP471396736), T41, VMUL(LDK(KP881921264), T42)); T4b = VADD(T49, T4a); T4j = VSUB(T4a, T49); } { V T40, T43, T46, T47; T40 = VFMA(LDK(KP881921264), T3Y, VMUL(LDK(KP471396736), T3Z)); T43 = VFNMS(LDK(KP471396736), T42, VMUL(LDK(KP881921264), T41)); T44 = VADD(T40, T43); T4f = VSUB(T43, T40); T46 = VSUB(T3B, T3A); T47 = VSUB(T3h, T3e); T48 = VADD(T46, T47); T4g = VSUB(T47, T46); } { V T45, T4c, T8h, T8i; T45 = VADD(T3X, T44); T4c = VBYI(VADD(T48, T4b)); T8h = VSUB(T45, T4c); STM2(&(xo[118]), T8h, ovs, &(xo[2])); STN2(&(xo[116]), T7D, T8h, ovs); T8i = VADD(T45, T4c); STM2(&(xo[10]), T8i, ovs, &(xo[2])); STN2(&(xo[8]), T7w, T8i, ovs); } { V T4l, T4m, T8j, T8k; T4l = VBYI(VADD(T4g, T4f)); T4m = VADD(T4i, T4j); T8j = VADD(T4l, T4m); STM2(&(xo[22]), T8j, ovs, &(xo[2])); STN2(&(xo[20]), T7F, T8j, ovs); T8k = VSUB(T4m, T4l); STM2(&(xo[106]), T8k, ovs, &(xo[2])); STN2(&(xo[104]), T7y, T8k, ovs); } { V T4d, T4e, T8l, T8m; T4d = VSUB(T3X, T44); T4e = VBYI(VSUB(T4b, T48)); T8l = VSUB(T4d, T4e); STM2(&(xo[74]), T8l, ovs, &(xo[2])); STN2(&(xo[72]), T7z, T8l, ovs); T8m = VADD(T4d, T4e); STM2(&(xo[54]), T8m, ovs, &(xo[2])); STN2(&(xo[52]), T7I, T8m, ovs); } { V T4h, T4k, T8n, T8o; T4h = VBYI(VSUB(T4f, T4g)); T4k = VSUB(T4i, T4j); T8n = VADD(T4h, T4k); STM2(&(xo[42]), T8n, ovs, &(xo[2])); STN2(&(xo[40]), T7B, T8n, ovs); T8o = VSUB(T4k, T4h); STM2(&(xo[86]), T8o, ovs, &(xo[2])); STN2(&(xo[84]), T7K, T8o, ovs); } } } } } } } VLEAVE(); } static const kdft_desc desc = { 64, XSIMD_STRING("n2fv_64"), {404, 72, 52, 0}, &GENUS, 0, 2, 0, 0 }; void XSIMD(codelet_n2fv_64) (planner *p) { X(kdft_register) (p, n2fv_64, &desc); } #endif /* HAVE_FMA */ fftw-3.3.3/dft/simd/common/n2fv_20.c0000644000175000017500000004056412054410411013731 00000000000000/* * Copyright (c) 2003, 2007-11 Matteo Frigo * Copyright (c) 2003, 2007-11 Massachusetts Institute of Technology * * This program is free software; you can redistribute it and/or modify * it under the terms of the GNU General Public License as published by * the Free Software Foundation; either version 2 of the License, or * (at your option) any later version. * * This program is distributed in the hope that it will be useful, * but WITHOUT ANY WARRANTY; without even the implied warranty of * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the * GNU General Public License for more details. * * You should have received a copy of the GNU General Public License * along with this program; if not, write to the Free Software * Foundation, Inc., 51 Franklin Street, Fifth Floor, Boston, MA 02110-1301 USA * */ /* This file was automatically generated --- DO NOT EDIT */ /* Generated on Sun Nov 25 07:37:28 EST 2012 */ #include "codelet-dft.h" #ifdef HAVE_FMA /* Generated by: ../../../genfft/gen_notw_c.native -fma -reorder-insns -schedule-for-pipeline -simd -compact -variables 4 -pipeline-latency 8 -n 20 -name n2fv_20 -with-ostride 2 -include n2f.h -store-multiple 2 */ /* * This function contains 104 FP additions, 50 FP multiplications, * (or, 58 additions, 4 multiplications, 46 fused multiply/add), * 79 stack variables, 4 constants, and 50 memory accesses */ #include "n2f.h" static void n2fv_20(const R *ri, const R *ii, R *ro, R *io, stride is, stride os, INT v, INT ivs, INT ovs) { DVK(KP559016994, +0.559016994374947424102293417182819058860154590); DVK(KP618033988, +0.618033988749894848204586834365638117720309180); DVK(KP951056516, +0.951056516295153572116439333379382143405698634); DVK(KP250000000, +0.250000000000000000000000000000000000000000000); { INT i; const R *xi; R *xo; xi = ri; xo = ro; for (i = v; i > 0; i = i - VL, xi = xi + (VL * ivs), xo = xo + (VL * ovs), MAKE_VOLATILE_STRIDE(40, is), MAKE_VOLATILE_STRIDE(40, os)) { V T1H, T1I, TU, TI, TP, TX, T1M, T1N, T1O, T1P, T1R, T1S, TM, TW, TT; V TF; { V T3, Tm, T1r, T13, Ta, TN, TH, TA, TG, Tt, Th, TO, T1u, T1C, T1n; V T1a, T1m, T1h, T1x, T1D, TE, Ti; { V T1, T2, Tk, Tl; T1 = LD(&(xi[0]), ivs, &(xi[0])); T2 = LD(&(xi[WS(is, 10)]), ivs, &(xi[0])); Tk = LD(&(xi[WS(is, 5)]), ivs, &(xi[WS(is, 1)])); Tl = LD(&(xi[WS(is, 15)]), ivs, &(xi[WS(is, 1)])); { V T14, T6, T1c, Tw, Tn, T1f, Tz, T17, T9, To, Tq, T1b, Td, Tr, Te; V Tf, T15, Tp; { V Tx, Ty, T7, T8, Tb, Tc; { V T4, T5, Tu, Tv, T11, T12; T4 = LD(&(xi[WS(is, 4)]), ivs, &(xi[0])); T5 = LD(&(xi[WS(is, 14)]), ivs, &(xi[0])); Tu = LD(&(xi[WS(is, 13)]), ivs, &(xi[WS(is, 1)])); Tv = LD(&(xi[WS(is, 3)]), ivs, &(xi[WS(is, 1)])); Tx = LD(&(xi[WS(is, 17)]), ivs, &(xi[WS(is, 1)])); T3 = VSUB(T1, T2); T11 = VADD(T1, T2); Tm = VSUB(Tk, Tl); T12 = VADD(Tk, Tl); T14 = VADD(T4, T5); T6 = VSUB(T4, T5); T1c = VADD(Tu, Tv); Tw = VSUB(Tu, Tv); Ty = LD(&(xi[WS(is, 7)]), ivs, &(xi[WS(is, 1)])); T7 = LD(&(xi[WS(is, 16)]), ivs, &(xi[0])); T8 = LD(&(xi[WS(is, 6)]), ivs, &(xi[0])); T1r = VADD(T11, T12); T13 = VSUB(T11, T12); } Tb = LD(&(xi[WS(is, 8)]), ivs, &(xi[0])); Tc = LD(&(xi[WS(is, 18)]), ivs, &(xi[0])); Tn = LD(&(xi[WS(is, 9)]), ivs, &(xi[WS(is, 1)])); T1f = VADD(Tx, Ty); Tz = VSUB(Tx, Ty); T17 = VADD(T7, T8); T9 = VSUB(T7, T8); To = LD(&(xi[WS(is, 19)]), ivs, &(xi[WS(is, 1)])); Tq = LD(&(xi[WS(is, 1)]), ivs, &(xi[WS(is, 1)])); T1b = VADD(Tb, Tc); Td = VSUB(Tb, Tc); Tr = LD(&(xi[WS(is, 11)]), ivs, &(xi[WS(is, 1)])); Te = LD(&(xi[WS(is, 12)]), ivs, &(xi[0])); Tf = LD(&(xi[WS(is, 2)]), ivs, &(xi[0])); } Ta = VADD(T6, T9); TN = VSUB(T6, T9); T15 = VADD(Tn, To); Tp = VSUB(Tn, To); TH = VSUB(Tz, Tw); TA = VADD(Tw, Tz); { V T1d, T1v, T18, Ts, T1e, Tg, T16, T1s; T1d = VSUB(T1b, T1c); T1v = VADD(T1b, T1c); T18 = VADD(Tq, Tr); Ts = VSUB(Tq, Tr); T1e = VADD(Te, Tf); Tg = VSUB(Te, Tf); T16 = VSUB(T14, T15); T1s = VADD(T14, T15); { V T1t, T19, T1w, T1g; T1t = VADD(T17, T18); T19 = VSUB(T17, T18); TG = VSUB(Ts, Tp); Tt = VADD(Tp, Ts); T1w = VADD(T1e, T1f); T1g = VSUB(T1e, T1f); Th = VADD(Td, Tg); TO = VSUB(Td, Tg); T1u = VADD(T1s, T1t); T1C = VSUB(T1s, T1t); T1n = VSUB(T16, T19); T1a = VADD(T16, T19); T1m = VSUB(T1d, T1g); T1h = VADD(T1d, T1g); T1x = VADD(T1v, T1w); T1D = VSUB(T1v, T1w); } } } } TE = VSUB(Ta, Th); Ti = VADD(Ta, Th); { V TL, T1k, T1A, Tj, TD, T1E, T1G, TK, TC, T1j, T1z, T1i, T1y, TB; TL = VSUB(TA, Tt); TB = VADD(Tt, TA); T1i = VADD(T1a, T1h); T1k = VSUB(T1a, T1h); T1y = VADD(T1u, T1x); T1A = VSUB(T1u, T1x); Tj = VADD(T3, Ti); TD = VFNMS(LDK(KP250000000), Ti, T3); T1E = VMUL(LDK(KP951056516), VFMA(LDK(KP618033988), T1D, T1C)); T1G = VMUL(LDK(KP951056516), VFNMS(LDK(KP618033988), T1C, T1D)); TK = VFNMS(LDK(KP250000000), TB, Tm); TC = VADD(Tm, TB); T1j = VFNMS(LDK(KP250000000), T1i, T13); T1H = VADD(T1r, T1y); STM2(&(xo[0]), T1H, ovs, &(xo[0])); T1z = VFNMS(LDK(KP250000000), T1y, T1r); T1I = VADD(T13, T1i); STM2(&(xo[20]), T1I, ovs, &(xo[0])); { V T1J, T1K, T1p, T1l, T1o, T1q, T1F, T1B, T1L, T1Q; TU = VFNMS(LDK(KP618033988), TG, TH); TI = VFMA(LDK(KP618033988), TH, TG); TP = VFMA(LDK(KP618033988), TO, TN); TX = VFNMS(LDK(KP618033988), TN, TO); T1J = VFMAI(TC, Tj); STM2(&(xo[30]), T1J, ovs, &(xo[2])); T1K = VFNMSI(TC, Tj); STM2(&(xo[10]), T1K, ovs, &(xo[2])); T1p = VFMA(LDK(KP559016994), T1k, T1j); T1l = VFNMS(LDK(KP559016994), T1k, T1j); T1o = VMUL(LDK(KP951056516), VFNMS(LDK(KP618033988), T1n, T1m)); T1q = VMUL(LDK(KP951056516), VFMA(LDK(KP618033988), T1m, T1n)); T1F = VFNMS(LDK(KP559016994), T1A, T1z); T1B = VFMA(LDK(KP559016994), T1A, T1z); T1L = VFMAI(T1q, T1p); STM2(&(xo[28]), T1L, ovs, &(xo[0])); STN2(&(xo[28]), T1L, T1J, ovs); T1M = VFNMSI(T1q, T1p); STM2(&(xo[12]), T1M, ovs, &(xo[0])); T1N = VFNMSI(T1o, T1l); STM2(&(xo[36]), T1N, ovs, &(xo[0])); T1O = VFMAI(T1o, T1l); STM2(&(xo[4]), T1O, ovs, &(xo[0])); T1P = VFNMSI(T1E, T1B); STM2(&(xo[32]), T1P, ovs, &(xo[0])); T1Q = VFMAI(T1E, T1B); STM2(&(xo[8]), T1Q, ovs, &(xo[0])); STN2(&(xo[8]), T1Q, T1K, ovs); T1R = VFMAI(T1G, T1F); STM2(&(xo[24]), T1R, ovs, &(xo[0])); T1S = VFNMSI(T1G, T1F); STM2(&(xo[16]), T1S, ovs, &(xo[0])); TM = VFNMS(LDK(KP559016994), TL, TK); TW = VFMA(LDK(KP559016994), TL, TK); TT = VFNMS(LDK(KP559016994), TE, TD); TF = VFMA(LDK(KP559016994), TE, TD); } } } { V T10, TY, TQ, TS, TJ, TR, TZ, TV; T10 = VFMA(LDK(KP951056516), TX, TW); TY = VFNMS(LDK(KP951056516), TX, TW); TQ = VFMA(LDK(KP951056516), TP, TM); TS = VFNMS(LDK(KP951056516), TP, TM); TJ = VFMA(LDK(KP951056516), TI, TF); TR = VFNMS(LDK(KP951056516), TI, TF); TZ = VFMA(LDK(KP951056516), TU, TT); TV = VFNMS(LDK(KP951056516), TU, TT); { V T1T, T1U, T1V, T1W; T1T = VFMAI(TS, TR); STM2(&(xo[22]), T1T, ovs, &(xo[2])); STN2(&(xo[20]), T1I, T1T, ovs); T1U = VFNMSI(TS, TR); STM2(&(xo[18]), T1U, ovs, &(xo[2])); STN2(&(xo[16]), T1S, T1U, ovs); T1V = VFMAI(TQ, TJ); STM2(&(xo[38]), T1V, ovs, &(xo[2])); STN2(&(xo[36]), T1N, T1V, ovs); T1W = VFNMSI(TQ, TJ); STM2(&(xo[2]), T1W, ovs, &(xo[2])); STN2(&(xo[0]), T1H, T1W, ovs); { V T1X, T1Y, T1Z, T20; T1X = VFMAI(TY, TV); STM2(&(xo[6]), T1X, ovs, &(xo[2])); STN2(&(xo[4]), T1O, T1X, ovs); T1Y = VFNMSI(TY, TV); STM2(&(xo[34]), T1Y, ovs, &(xo[2])); STN2(&(xo[32]), T1P, T1Y, ovs); T1Z = VFMAI(T10, TZ); STM2(&(xo[14]), T1Z, ovs, &(xo[2])); STN2(&(xo[12]), T1M, T1Z, ovs); T20 = VFNMSI(T10, TZ); STM2(&(xo[26]), T20, ovs, &(xo[2])); STN2(&(xo[24]), T1R, T20, ovs); } } } } } VLEAVE(); } static const kdft_desc desc = { 20, XSIMD_STRING("n2fv_20"), {58, 4, 46, 0}, &GENUS, 0, 2, 0, 0 }; void XSIMD(codelet_n2fv_20) (planner *p) { X(kdft_register) (p, n2fv_20, &desc); } #else /* HAVE_FMA */ /* Generated by: ../../../genfft/gen_notw_c.native -simd -compact -variables 4 -pipeline-latency 8 -n 20 -name n2fv_20 -with-ostride 2 -include n2f.h -store-multiple 2 */ /* * This function contains 104 FP additions, 24 FP multiplications, * (or, 92 additions, 12 multiplications, 12 fused multiply/add), * 57 stack variables, 4 constants, and 50 memory accesses */ #include "n2f.h" static void n2fv_20(const R *ri, const R *ii, R *ro, R *io, stride is, stride os, INT v, INT ivs, INT ovs) { DVK(KP587785252, +0.587785252292473129168705954639072768597652438); DVK(KP951056516, +0.951056516295153572116439333379382143405698634); DVK(KP250000000, +0.250000000000000000000000000000000000000000000); DVK(KP559016994, +0.559016994374947424102293417182819058860154590); { INT i; const R *xi; R *xo; xi = ri; xo = ro; for (i = v; i > 0; i = i - VL, xi = xi + (VL * ivs), xo = xo + (VL * ovs), MAKE_VOLATILE_STRIDE(40, is), MAKE_VOLATILE_STRIDE(40, os)) { V T3, T1B, Tm, T1i, TG, TN, TO, TH, T13, T16, T1k, T1u, T1v, T1z, T1r; V T1s, T1y, T1a, T1d, T1j, Ti, TD, TB, TL; { V T1, T2, T1g, Tk, Tl, T1h; T1 = LD(&(xi[0]), ivs, &(xi[0])); T2 = LD(&(xi[WS(is, 10)]), ivs, &(xi[0])); T1g = VADD(T1, T2); Tk = LD(&(xi[WS(is, 5)]), ivs, &(xi[WS(is, 1)])); Tl = LD(&(xi[WS(is, 15)]), ivs, &(xi[WS(is, 1)])); T1h = VADD(Tk, Tl); T3 = VSUB(T1, T2); T1B = VADD(T1g, T1h); Tm = VSUB(Tk, Tl); T1i = VSUB(T1g, T1h); } { V T6, T18, Tw, T12, Tz, T15, T9, T1b, Td, T11, Tp, T19, Ts, T1c, Tg; V T14; { V T4, T5, Tu, Tv; T4 = LD(&(xi[WS(is, 4)]), ivs, &(xi[0])); T5 = LD(&(xi[WS(is, 14)]), ivs, &(xi[0])); T6 = VSUB(T4, T5); T18 = VADD(T4, T5); Tu = LD(&(xi[WS(is, 13)]), ivs, &(xi[WS(is, 1)])); Tv = LD(&(xi[WS(is, 3)]), ivs, &(xi[WS(is, 1)])); Tw = VSUB(Tu, Tv); T12 = VADD(Tu, Tv); } { V Tx, Ty, T7, T8; Tx = LD(&(xi[WS(is, 17)]), ivs, &(xi[WS(is, 1)])); Ty = LD(&(xi[WS(is, 7)]), ivs, &(xi[WS(is, 1)])); Tz = VSUB(Tx, Ty); T15 = VADD(Tx, Ty); T7 = LD(&(xi[WS(is, 16)]), ivs, &(xi[0])); T8 = LD(&(xi[WS(is, 6)]), ivs, &(xi[0])); T9 = VSUB(T7, T8); T1b = VADD(T7, T8); } { V Tb, Tc, Tn, To; Tb = LD(&(xi[WS(is, 8)]), ivs, &(xi[0])); Tc = LD(&(xi[WS(is, 18)]), ivs, &(xi[0])); Td = VSUB(Tb, Tc); T11 = VADD(Tb, Tc); Tn = LD(&(xi[WS(is, 9)]), ivs, &(xi[WS(is, 1)])); To = LD(&(xi[WS(is, 19)]), ivs, &(xi[WS(is, 1)])); Tp = VSUB(Tn, To); T19 = VADD(Tn, To); } { V Tq, Tr, Te, Tf; Tq = LD(&(xi[WS(is, 1)]), ivs, &(xi[WS(is, 1)])); Tr = LD(&(xi[WS(is, 11)]), ivs, &(xi[WS(is, 1)])); Ts = VSUB(Tq, Tr); T1c = VADD(Tq, Tr); Te = LD(&(xi[WS(is, 12)]), ivs, &(xi[0])); Tf = LD(&(xi[WS(is, 2)]), ivs, &(xi[0])); Tg = VSUB(Te, Tf); T14 = VADD(Te, Tf); } TG = VSUB(Ts, Tp); TN = VSUB(T6, T9); TO = VSUB(Td, Tg); TH = VSUB(Tz, Tw); T13 = VSUB(T11, T12); T16 = VSUB(T14, T15); T1k = VADD(T13, T16); T1u = VADD(T11, T12); T1v = VADD(T14, T15); T1z = VADD(T1u, T1v); T1r = VADD(T18, T19); T1s = VADD(T1b, T1c); T1y = VADD(T1r, T1s); T1a = VSUB(T18, T19); T1d = VSUB(T1b, T1c); T1j = VADD(T1a, T1d); { V Ta, Th, Tt, TA; Ta = VADD(T6, T9); Th = VADD(Td, Tg); Ti = VADD(Ta, Th); TD = VMUL(LDK(KP559016994), VSUB(Ta, Th)); Tt = VADD(Tp, Ts); TA = VADD(Tw, Tz); TB = VADD(Tt, TA); TL = VMUL(LDK(KP559016994), VSUB(TA, Tt)); } } { V T1I, T1J, T1K, T1L, T1N, T1H, Tj, TC; Tj = VADD(T3, Ti); TC = VBYI(VADD(Tm, TB)); T1H = VSUB(Tj, TC); STM2(&(xo[10]), T1H, ovs, &(xo[2])); T1I = VADD(Tj, TC); STM2(&(xo[30]), T1I, ovs, &(xo[2])); { V T1A, T1C, T1D, T1x, T1G, T1t, T1w, T1F, T1E, T1M; T1A = VMUL(LDK(KP559016994), VSUB(T1y, T1z)); T1C = VADD(T1y, T1z); T1D = VFNMS(LDK(KP250000000), T1C, T1B); T1t = VSUB(T1r, T1s); T1w = VSUB(T1u, T1v); T1x = VBYI(VFMA(LDK(KP951056516), T1t, VMUL(LDK(KP587785252), T1w))); T1G = VBYI(VFNMS(LDK(KP587785252), T1t, VMUL(LDK(KP951056516), T1w))); T1J = VADD(T1B, T1C); STM2(&(xo[0]), T1J, ovs, &(xo[0])); T1F = VSUB(T1D, T1A); T1K = VSUB(T1F, T1G); STM2(&(xo[16]), T1K, ovs, &(xo[0])); T1L = VADD(T1G, T1F); STM2(&(xo[24]), T1L, ovs, &(xo[0])); T1E = VADD(T1A, T1D); T1M = VADD(T1x, T1E); STM2(&(xo[8]), T1M, ovs, &(xo[0])); STN2(&(xo[8]), T1M, T1H, ovs); T1N = VSUB(T1E, T1x); STM2(&(xo[32]), T1N, ovs, &(xo[0])); } { V T1O, T1P, T1R, T1S; { V T1n, T1l, T1m, T1f, T1q, T17, T1e, T1p, T1Q, T1o; T1n = VMUL(LDK(KP559016994), VSUB(T1j, T1k)); T1l = VADD(T1j, T1k); T1m = VFNMS(LDK(KP250000000), T1l, T1i); T17 = VSUB(T13, T16); T1e = VSUB(T1a, T1d); T1f = VBYI(VFNMS(LDK(KP587785252), T1e, VMUL(LDK(KP951056516), T17))); T1q = VBYI(VFMA(LDK(KP951056516), T1e, VMUL(LDK(KP587785252), T17))); T1O = VADD(T1i, T1l); STM2(&(xo[20]), T1O, ovs, &(xo[0])); T1p = VADD(T1n, T1m); T1P = VSUB(T1p, T1q); STM2(&(xo[12]), T1P, ovs, &(xo[0])); T1Q = VADD(T1q, T1p); STM2(&(xo[28]), T1Q, ovs, &(xo[0])); STN2(&(xo[28]), T1Q, T1I, ovs); T1o = VSUB(T1m, T1n); T1R = VADD(T1f, T1o); STM2(&(xo[4]), T1R, ovs, &(xo[0])); T1S = VSUB(T1o, T1f); STM2(&(xo[36]), T1S, ovs, &(xo[0])); } { V TI, TP, TX, TU, TM, TW, TF, TT, TK, TE; TI = VFMA(LDK(KP951056516), TG, VMUL(LDK(KP587785252), TH)); TP = VFMA(LDK(KP951056516), TN, VMUL(LDK(KP587785252), TO)); TX = VFNMS(LDK(KP587785252), TN, VMUL(LDK(KP951056516), TO)); TU = VFNMS(LDK(KP587785252), TG, VMUL(LDK(KP951056516), TH)); TK = VFMS(LDK(KP250000000), TB, Tm); TM = VADD(TK, TL); TW = VSUB(TL, TK); TE = VFNMS(LDK(KP250000000), Ti, T3); TF = VADD(TD, TE); TT = VSUB(TE, TD); { V TJ, TQ, T1T, T1U; TJ = VADD(TF, TI); TQ = VBYI(VSUB(TM, TP)); T1T = VSUB(TJ, TQ); STM2(&(xo[38]), T1T, ovs, &(xo[2])); STN2(&(xo[36]), T1S, T1T, ovs); T1U = VADD(TJ, TQ); STM2(&(xo[2]), T1U, ovs, &(xo[2])); STN2(&(xo[0]), T1J, T1U, ovs); } { V TZ, T10, T1V, T1W; TZ = VADD(TT, TU); T10 = VBYI(VADD(TX, TW)); T1V = VSUB(TZ, T10); STM2(&(xo[26]), T1V, ovs, &(xo[2])); STN2(&(xo[24]), T1L, T1V, ovs); T1W = VADD(TZ, T10); STM2(&(xo[14]), T1W, ovs, &(xo[2])); STN2(&(xo[12]), T1P, T1W, ovs); } { V TR, TS, T1X, T1Y; TR = VSUB(TF, TI); TS = VBYI(VADD(TP, TM)); T1X = VSUB(TR, TS); STM2(&(xo[22]), T1X, ovs, &(xo[2])); STN2(&(xo[20]), T1O, T1X, ovs); T1Y = VADD(TR, TS); STM2(&(xo[18]), T1Y, ovs, &(xo[2])); STN2(&(xo[16]), T1K, T1Y, ovs); } { V TV, TY, T1Z, T20; TV = VSUB(TT, TU); TY = VBYI(VSUB(TW, TX)); T1Z = VSUB(TV, TY); STM2(&(xo[34]), T1Z, ovs, &(xo[2])); STN2(&(xo[32]), T1N, T1Z, ovs); T20 = VADD(TV, TY); STM2(&(xo[6]), T20, ovs, &(xo[2])); STN2(&(xo[4]), T1R, T20, ovs); } } } } } } VLEAVE(); } static const kdft_desc desc = { 20, XSIMD_STRING("n2fv_20"), {92, 12, 12, 0}, &GENUS, 0, 2, 0, 0 }; void XSIMD(codelet_n2fv_20) (planner *p) { X(kdft_register) (p, n2fv_20, &desc); } #endif /* HAVE_FMA */ fftw-3.3.3/dft/simd/common/n2bv_2.c0000644000175000017500000000670212054410411013641 00000000000000/* * Copyright (c) 2003, 2007-11 Matteo Frigo * Copyright (c) 2003, 2007-11 Massachusetts Institute of Technology * * This program is free software; you can redistribute it and/or modify * it under the terms of the GNU General Public License as published by * the Free Software Foundation; either version 2 of the License, or * (at your option) any later version. * * This program is distributed in the hope that it will be useful, * but WITHOUT ANY WARRANTY; without even the implied warranty of * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the * GNU General Public License for more details. * * You should have received a copy of the GNU General Public License * along with this program; if not, write to the Free Software * Foundation, Inc., 51 Franklin Street, Fifth Floor, Boston, MA 02110-1301 USA * */ /* This file was automatically generated --- DO NOT EDIT */ /* Generated on Sun Nov 25 07:37:29 EST 2012 */ #include "codelet-dft.h" #ifdef HAVE_FMA /* Generated by: ../../../genfft/gen_notw_c.native -fma -reorder-insns -schedule-for-pipeline -simd -compact -variables 4 -pipeline-latency 8 -sign 1 -n 2 -name n2bv_2 -with-ostride 2 -include n2b.h -store-multiple 2 */ /* * This function contains 2 FP additions, 0 FP multiplications, * (or, 2 additions, 0 multiplications, 0 fused multiply/add), * 7 stack variables, 0 constants, and 5 memory accesses */ #include "n2b.h" static void n2bv_2(const R *ri, const R *ii, R *ro, R *io, stride is, stride os, INT v, INT ivs, INT ovs) { { INT i; const R *xi; R *xo; xi = ii; xo = io; for (i = v; i > 0; i = i - VL, xi = xi + (VL * ivs), xo = xo + (VL * ovs), MAKE_VOLATILE_STRIDE(4, is), MAKE_VOLATILE_STRIDE(4, os)) { V T1, T2, T3, T4; T1 = LD(&(xi[0]), ivs, &(xi[0])); T2 = LD(&(xi[WS(is, 1)]), ivs, &(xi[WS(is, 1)])); T3 = VADD(T1, T2); STM2(&(xo[0]), T3, ovs, &(xo[0])); T4 = VSUB(T1, T2); STM2(&(xo[2]), T4, ovs, &(xo[2])); STN2(&(xo[0]), T3, T4, ovs); } } VLEAVE(); } static const kdft_desc desc = { 2, XSIMD_STRING("n2bv_2"), {2, 0, 0, 0}, &GENUS, 0, 2, 0, 0 }; void XSIMD(codelet_n2bv_2) (planner *p) { X(kdft_register) (p, n2bv_2, &desc); } #else /* HAVE_FMA */ /* Generated by: ../../../genfft/gen_notw_c.native -simd -compact -variables 4 -pipeline-latency 8 -sign 1 -n 2 -name n2bv_2 -with-ostride 2 -include n2b.h -store-multiple 2 */ /* * This function contains 2 FP additions, 0 FP multiplications, * (or, 2 additions, 0 multiplications, 0 fused multiply/add), * 7 stack variables, 0 constants, and 5 memory accesses */ #include "n2b.h" static void n2bv_2(const R *ri, const R *ii, R *ro, R *io, stride is, stride os, INT v, INT ivs, INT ovs) { { INT i; const R *xi; R *xo; xi = ii; xo = io; for (i = v; i > 0; i = i - VL, xi = xi + (VL * ivs), xo = xo + (VL * ovs), MAKE_VOLATILE_STRIDE(4, is), MAKE_VOLATILE_STRIDE(4, os)) { V T1, T2, T3, T4; T1 = LD(&(xi[0]), ivs, &(xi[0])); T2 = LD(&(xi[WS(is, 1)]), ivs, &(xi[WS(is, 1)])); T3 = VSUB(T1, T2); STM2(&(xo[2]), T3, ovs, &(xo[2])); T4 = VADD(T1, T2); STM2(&(xo[0]), T4, ovs, &(xo[0])); STN2(&(xo[0]), T4, T3, ovs); } } VLEAVE(); } static const kdft_desc desc = { 2, XSIMD_STRING("n2bv_2"), {2, 0, 0, 0}, &GENUS, 0, 2, 0, 0 }; void XSIMD(codelet_n2bv_2) (planner *p) { X(kdft_register) (p, n2bv_2, &desc); } #endif /* HAVE_FMA */ fftw-3.3.3/dft/simd/common/n2bv_4.c0000644000175000017500000001046712054410411013646 00000000000000/* * Copyright (c) 2003, 2007-11 Matteo Frigo * Copyright (c) 2003, 2007-11 Massachusetts Institute of Technology * * This program is free software; you can redistribute it and/or modify * it under the terms of the GNU General Public License as published by * the Free Software Foundation; either version 2 of the License, or * (at your option) any later version. * * This program is distributed in the hope that it will be useful, * but WITHOUT ANY WARRANTY; without even the implied warranty of * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the * GNU General Public License for more details. * * You should have received a copy of the GNU General Public License * along with this program; if not, write to the Free Software * Foundation, Inc., 51 Franklin Street, Fifth Floor, Boston, MA 02110-1301 USA * */ /* This file was automatically generated --- DO NOT EDIT */ /* Generated on Sun Nov 25 07:37:29 EST 2012 */ #include "codelet-dft.h" #ifdef HAVE_FMA /* Generated by: ../../../genfft/gen_notw_c.native -fma -reorder-insns -schedule-for-pipeline -simd -compact -variables 4 -pipeline-latency 8 -sign 1 -n 4 -name n2bv_4 -with-ostride 2 -include n2b.h -store-multiple 2 */ /* * This function contains 8 FP additions, 2 FP multiplications, * (or, 6 additions, 0 multiplications, 2 fused multiply/add), * 15 stack variables, 0 constants, and 10 memory accesses */ #include "n2b.h" static void n2bv_4(const R *ri, const R *ii, R *ro, R *io, stride is, stride os, INT v, INT ivs, INT ovs) { { INT i; const R *xi; R *xo; xi = ii; xo = io; for (i = v; i > 0; i = i - VL, xi = xi + (VL * ivs), xo = xo + (VL * ovs), MAKE_VOLATILE_STRIDE(8, is), MAKE_VOLATILE_STRIDE(8, os)) { V T1, T2, T4, T5; T1 = LD(&(xi[0]), ivs, &(xi[0])); T2 = LD(&(xi[WS(is, 2)]), ivs, &(xi[0])); T4 = LD(&(xi[WS(is, 1)]), ivs, &(xi[WS(is, 1)])); T5 = LD(&(xi[WS(is, 3)]), ivs, &(xi[WS(is, 1)])); { V T3, T7, T6, T8; T3 = VSUB(T1, T2); T7 = VADD(T1, T2); T6 = VSUB(T4, T5); T8 = VADD(T4, T5); { V T9, Ta, Tb, Tc; T9 = VSUB(T7, T8); STM2(&(xo[4]), T9, ovs, &(xo[0])); Ta = VADD(T7, T8); STM2(&(xo[0]), Ta, ovs, &(xo[0])); Tb = VFMAI(T6, T3); STM2(&(xo[2]), Tb, ovs, &(xo[2])); STN2(&(xo[0]), Ta, Tb, ovs); Tc = VFNMSI(T6, T3); STM2(&(xo[6]), Tc, ovs, &(xo[2])); STN2(&(xo[4]), T9, Tc, ovs); } } } } VLEAVE(); } static const kdft_desc desc = { 4, XSIMD_STRING("n2bv_4"), {6, 0, 2, 0}, &GENUS, 0, 2, 0, 0 }; void XSIMD(codelet_n2bv_4) (planner *p) { X(kdft_register) (p, n2bv_4, &desc); } #else /* HAVE_FMA */ /* Generated by: ../../../genfft/gen_notw_c.native -simd -compact -variables 4 -pipeline-latency 8 -sign 1 -n 4 -name n2bv_4 -with-ostride 2 -include n2b.h -store-multiple 2 */ /* * This function contains 8 FP additions, 0 FP multiplications, * (or, 8 additions, 0 multiplications, 0 fused multiply/add), * 11 stack variables, 0 constants, and 10 memory accesses */ #include "n2b.h" static void n2bv_4(const R *ri, const R *ii, R *ro, R *io, stride is, stride os, INT v, INT ivs, INT ovs) { { INT i; const R *xi; R *xo; xi = ii; xo = io; for (i = v; i > 0; i = i - VL, xi = xi + (VL * ivs), xo = xo + (VL * ovs), MAKE_VOLATILE_STRIDE(8, is), MAKE_VOLATILE_STRIDE(8, os)) { V T3, T7, T6, T8; { V T1, T2, T4, T5; T1 = LD(&(xi[0]), ivs, &(xi[0])); T2 = LD(&(xi[WS(is, 2)]), ivs, &(xi[0])); T3 = VSUB(T1, T2); T7 = VADD(T1, T2); T4 = LD(&(xi[WS(is, 1)]), ivs, &(xi[WS(is, 1)])); T5 = LD(&(xi[WS(is, 3)]), ivs, &(xi[WS(is, 1)])); T6 = VBYI(VSUB(T4, T5)); T8 = VADD(T4, T5); } { V T9, Ta, Tb, Tc; T9 = VSUB(T3, T6); STM2(&(xo[6]), T9, ovs, &(xo[2])); Ta = VADD(T7, T8); STM2(&(xo[0]), Ta, ovs, &(xo[0])); Tb = VADD(T3, T6); STM2(&(xo[2]), Tb, ovs, &(xo[2])); STN2(&(xo[0]), Ta, Tb, ovs); Tc = VSUB(T7, T8); STM2(&(xo[4]), Tc, ovs, &(xo[0])); STN2(&(xo[4]), Tc, T9, ovs); } } } VLEAVE(); } static const kdft_desc desc = { 4, XSIMD_STRING("n2bv_4"), {8, 0, 0, 0}, &GENUS, 0, 2, 0, 0 }; void XSIMD(codelet_n2bv_4) (planner *p) { X(kdft_register) (p, n2bv_4, &desc); } #endif /* HAVE_FMA */ fftw-3.3.3/dft/simd/common/n2bv_6.c0000644000175000017500000001364712054410411013653 00000000000000/* * Copyright (c) 2003, 2007-11 Matteo Frigo * Copyright (c) 2003, 2007-11 Massachusetts Institute of Technology * * This program is free software; you can redistribute it and/or modify * it under the terms of the GNU General Public License as published by * the Free Software Foundation; either version 2 of the License, or * (at your option) any later version. * * This program is distributed in the hope that it will be useful, * but WITHOUT ANY WARRANTY; without even the implied warranty of * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the * GNU General Public License for more details. * * You should have received a copy of the GNU General Public License * along with this program; if not, write to the Free Software * Foundation, Inc., 51 Franklin Street, Fifth Floor, Boston, MA 02110-1301 USA * */ /* This file was automatically generated --- DO NOT EDIT */ /* Generated on Sun Nov 25 07:37:29 EST 2012 */ #include "codelet-dft.h" #ifdef HAVE_FMA /* Generated by: ../../../genfft/gen_notw_c.native -fma -reorder-insns -schedule-for-pipeline -simd -compact -variables 4 -pipeline-latency 8 -sign 1 -n 6 -name n2bv_6 -with-ostride 2 -include n2b.h -store-multiple 2 */ /* * This function contains 18 FP additions, 8 FP multiplications, * (or, 12 additions, 2 multiplications, 6 fused multiply/add), * 29 stack variables, 2 constants, and 15 memory accesses */ #include "n2b.h" static void n2bv_6(const R *ri, const R *ii, R *ro, R *io, stride is, stride os, INT v, INT ivs, INT ovs) { DVK(KP500000000, +0.500000000000000000000000000000000000000000000); DVK(KP866025403, +0.866025403784438646763723170752936183471402627); { INT i; const R *xi; R *xo; xi = ii; xo = io; for (i = v; i > 0; i = i - VL, xi = xi + (VL * ivs), xo = xo + (VL * ovs), MAKE_VOLATILE_STRIDE(12, is), MAKE_VOLATILE_STRIDE(12, os)) { V T1, T2, T4, T5, T7, T8; T1 = LD(&(xi[0]), ivs, &(xi[0])); T2 = LD(&(xi[WS(is, 3)]), ivs, &(xi[WS(is, 1)])); T4 = LD(&(xi[WS(is, 2)]), ivs, &(xi[0])); T5 = LD(&(xi[WS(is, 5)]), ivs, &(xi[WS(is, 1)])); T7 = LD(&(xi[WS(is, 4)]), ivs, &(xi[0])); T8 = LD(&(xi[WS(is, 1)]), ivs, &(xi[WS(is, 1)])); { V T3, Td, T6, Te, T9, Tf; T3 = VSUB(T1, T2); Td = VADD(T1, T2); T6 = VSUB(T4, T5); Te = VADD(T4, T5); T9 = VSUB(T7, T8); Tf = VADD(T7, T8); { V Tg, Ti, Ta, Tc; Tg = VADD(Te, Tf); Ti = VMUL(LDK(KP866025403), VSUB(Te, Tf)); Ta = VADD(T6, T9); Tc = VMUL(LDK(KP866025403), VSUB(T6, T9)); { V Th, Tj, Tb, Tk; Th = VFNMS(LDK(KP500000000), Tg, Td); Tj = VADD(Td, Tg); STM2(&(xo[0]), Tj, ovs, &(xo[0])); Tb = VFNMS(LDK(KP500000000), Ta, T3); Tk = VADD(T3, Ta); STM2(&(xo[6]), Tk, ovs, &(xo[2])); { V Tl, Tm, Tn, To; Tl = VFMAI(Ti, Th); STM2(&(xo[8]), Tl, ovs, &(xo[0])); Tm = VFNMSI(Ti, Th); STM2(&(xo[4]), Tm, ovs, &(xo[0])); STN2(&(xo[4]), Tm, Tk, ovs); Tn = VFNMSI(Tc, Tb); STM2(&(xo[10]), Tn, ovs, &(xo[2])); STN2(&(xo[8]), Tl, Tn, ovs); To = VFMAI(Tc, Tb); STM2(&(xo[2]), To, ovs, &(xo[2])); STN2(&(xo[0]), Tj, To, ovs); } } } } } } VLEAVE(); } static const kdft_desc desc = { 6, XSIMD_STRING("n2bv_6"), {12, 2, 6, 0}, &GENUS, 0, 2, 0, 0 }; void XSIMD(codelet_n2bv_6) (planner *p) { X(kdft_register) (p, n2bv_6, &desc); } #else /* HAVE_FMA */ /* Generated by: ../../../genfft/gen_notw_c.native -simd -compact -variables 4 -pipeline-latency 8 -sign 1 -n 6 -name n2bv_6 -with-ostride 2 -include n2b.h -store-multiple 2 */ /* * This function contains 18 FP additions, 4 FP multiplications, * (or, 16 additions, 2 multiplications, 2 fused multiply/add), * 25 stack variables, 2 constants, and 15 memory accesses */ #include "n2b.h" static void n2bv_6(const R *ri, const R *ii, R *ro, R *io, stride is, stride os, INT v, INT ivs, INT ovs) { DVK(KP500000000, +0.500000000000000000000000000000000000000000000); DVK(KP866025403, +0.866025403784438646763723170752936183471402627); { INT i; const R *xi; R *xo; xi = ii; xo = io; for (i = v; i > 0; i = i - VL, xi = xi + (VL * ivs), xo = xo + (VL * ovs), MAKE_VOLATILE_STRIDE(12, is), MAKE_VOLATILE_STRIDE(12, os)) { V Ta, Td, T3, Te, T6, Tf, Tb, Tg, T8, T9, Tj, Tk; T8 = LD(&(xi[0]), ivs, &(xi[0])); T9 = LD(&(xi[WS(is, 3)]), ivs, &(xi[WS(is, 1)])); Ta = VSUB(T8, T9); Td = VADD(T8, T9); { V T1, T2, T4, T5; T1 = LD(&(xi[WS(is, 2)]), ivs, &(xi[0])); T2 = LD(&(xi[WS(is, 5)]), ivs, &(xi[WS(is, 1)])); T3 = VSUB(T1, T2); Te = VADD(T1, T2); T4 = LD(&(xi[WS(is, 4)]), ivs, &(xi[0])); T5 = LD(&(xi[WS(is, 1)]), ivs, &(xi[WS(is, 1)])); T6 = VSUB(T4, T5); Tf = VADD(T4, T5); } Tb = VADD(T3, T6); Tg = VADD(Te, Tf); Tj = VADD(Ta, Tb); STM2(&(xo[6]), Tj, ovs, &(xo[2])); Tk = VADD(Td, Tg); STM2(&(xo[0]), Tk, ovs, &(xo[0])); { V Tm, T7, Tc, Tl; T7 = VBYI(VMUL(LDK(KP866025403), VSUB(T3, T6))); Tc = VFNMS(LDK(KP500000000), Tb, Ta); Tl = VADD(T7, Tc); STM2(&(xo[2]), Tl, ovs, &(xo[2])); STN2(&(xo[0]), Tk, Tl, ovs); Tm = VSUB(Tc, T7); STM2(&(xo[10]), Tm, ovs, &(xo[2])); { V Th, Ti, Tn, To; Th = VFNMS(LDK(KP500000000), Tg, Td); Ti = VBYI(VMUL(LDK(KP866025403), VSUB(Te, Tf))); Tn = VSUB(Th, Ti); STM2(&(xo[4]), Tn, ovs, &(xo[0])); STN2(&(xo[4]), Tn, Tj, ovs); To = VADD(Ti, Th); STM2(&(xo[8]), To, ovs, &(xo[0])); STN2(&(xo[8]), To, Tm, ovs); } } } } VLEAVE(); } static const kdft_desc desc = { 6, XSIMD_STRING("n2bv_6"), {16, 2, 2, 0}, &GENUS, 0, 2, 0, 0 }; void XSIMD(codelet_n2bv_6) (planner *p) { X(kdft_register) (p, n2bv_6, &desc); } #endif /* HAVE_FMA */ fftw-3.3.3/dft/simd/common/n2bv_8.c0000644000175000017500000001526012054410411013646 00000000000000/* * Copyright (c) 2003, 2007-11 Matteo Frigo * Copyright (c) 2003, 2007-11 Massachusetts Institute of Technology * * This program is free software; you can redistribute it and/or modify * it under the terms of the GNU General Public License as published by * the Free Software Foundation; either version 2 of the License, or * (at your option) any later version. * * This program is distributed in the hope that it will be useful, * but WITHOUT ANY WARRANTY; without even the implied warranty of * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the * GNU General Public License for more details. * * You should have received a copy of the GNU General Public License * along with this program; if not, write to the Free Software * Foundation, Inc., 51 Franklin Street, Fifth Floor, Boston, MA 02110-1301 USA * */ /* This file was automatically generated --- DO NOT EDIT */ /* Generated on Sun Nov 25 07:37:29 EST 2012 */ #include "codelet-dft.h" #ifdef HAVE_FMA /* Generated by: ../../../genfft/gen_notw_c.native -fma -reorder-insns -schedule-for-pipeline -simd -compact -variables 4 -pipeline-latency 8 -sign 1 -n 8 -name n2bv_8 -with-ostride 2 -include n2b.h -store-multiple 2 */ /* * This function contains 26 FP additions, 10 FP multiplications, * (or, 16 additions, 0 multiplications, 10 fused multiply/add), * 38 stack variables, 1 constants, and 20 memory accesses */ #include "n2b.h" static void n2bv_8(const R *ri, const R *ii, R *ro, R *io, stride is, stride os, INT v, INT ivs, INT ovs) { DVK(KP707106781, +0.707106781186547524400844362104849039284835938); { INT i; const R *xi; R *xo; xi = ii; xo = io; for (i = v; i > 0; i = i - VL, xi = xi + (VL * ivs), xo = xo + (VL * ovs), MAKE_VOLATILE_STRIDE(16, is), MAKE_VOLATILE_STRIDE(16, os)) { V T1, T2, Tc, Td, T4, T5, T7, T8; T1 = LD(&(xi[0]), ivs, &(xi[0])); T2 = LD(&(xi[WS(is, 4)]), ivs, &(xi[0])); Tc = LD(&(xi[WS(is, 2)]), ivs, &(xi[0])); Td = LD(&(xi[WS(is, 6)]), ivs, &(xi[0])); T4 = LD(&(xi[WS(is, 1)]), ivs, &(xi[WS(is, 1)])); T5 = LD(&(xi[WS(is, 5)]), ivs, &(xi[WS(is, 1)])); T7 = LD(&(xi[WS(is, 7)]), ivs, &(xi[WS(is, 1)])); T8 = LD(&(xi[WS(is, 3)]), ivs, &(xi[WS(is, 1)])); { V T3, Tj, Te, Tk, T6, Tm, T9, Tn, Tp, Tl; T3 = VSUB(T1, T2); Tj = VADD(T1, T2); Te = VSUB(Tc, Td); Tk = VADD(Tc, Td); T6 = VSUB(T4, T5); Tm = VADD(T4, T5); T9 = VSUB(T7, T8); Tn = VADD(T7, T8); Tp = VADD(Tj, Tk); Tl = VSUB(Tj, Tk); { V Tq, To, Ta, Tf; Tq = VADD(Tm, Tn); To = VSUB(Tm, Tn); Ta = VADD(T6, T9); Tf = VSUB(T6, T9); { V Tr, Ts, Tt, Tu, Tg, Ti, Tb, Th; Tr = VFMAI(To, Tl); STM2(&(xo[4]), Tr, ovs, &(xo[0])); Ts = VFNMSI(To, Tl); STM2(&(xo[12]), Ts, ovs, &(xo[0])); Tt = VADD(Tp, Tq); STM2(&(xo[0]), Tt, ovs, &(xo[0])); Tu = VSUB(Tp, Tq); STM2(&(xo[8]), Tu, ovs, &(xo[0])); Tg = VFNMS(LDK(KP707106781), Tf, Te); Ti = VFMA(LDK(KP707106781), Tf, Te); Tb = VFNMS(LDK(KP707106781), Ta, T3); Th = VFMA(LDK(KP707106781), Ta, T3); { V Tv, Tw, Tx, Ty; Tv = VFNMSI(Ti, Th); STM2(&(xo[14]), Tv, ovs, &(xo[2])); STN2(&(xo[12]), Ts, Tv, ovs); Tw = VFMAI(Ti, Th); STM2(&(xo[2]), Tw, ovs, &(xo[2])); STN2(&(xo[0]), Tt, Tw, ovs); Tx = VFMAI(Tg, Tb); STM2(&(xo[10]), Tx, ovs, &(xo[2])); STN2(&(xo[8]), Tu, Tx, ovs); Ty = VFNMSI(Tg, Tb); STM2(&(xo[6]), Ty, ovs, &(xo[2])); STN2(&(xo[4]), Tr, Ty, ovs); } } } } } } VLEAVE(); } static const kdft_desc desc = { 8, XSIMD_STRING("n2bv_8"), {16, 0, 10, 0}, &GENUS, 0, 2, 0, 0 }; void XSIMD(codelet_n2bv_8) (planner *p) { X(kdft_register) (p, n2bv_8, &desc); } #else /* HAVE_FMA */ /* Generated by: ../../../genfft/gen_notw_c.native -simd -compact -variables 4 -pipeline-latency 8 -sign 1 -n 8 -name n2bv_8 -with-ostride 2 -include n2b.h -store-multiple 2 */ /* * This function contains 26 FP additions, 2 FP multiplications, * (or, 26 additions, 2 multiplications, 0 fused multiply/add), * 24 stack variables, 1 constants, and 20 memory accesses */ #include "n2b.h" static void n2bv_8(const R *ri, const R *ii, R *ro, R *io, stride is, stride os, INT v, INT ivs, INT ovs) { DVK(KP707106781, +0.707106781186547524400844362104849039284835938); { INT i; const R *xi; R *xo; xi = ii; xo = io; for (i = v; i > 0; i = i - VL, xi = xi + (VL * ivs), xo = xo + (VL * ovs), MAKE_VOLATILE_STRIDE(16, is), MAKE_VOLATILE_STRIDE(16, os)) { V Ta, Tk, Te, Tj, T7, Tn, Tf, Tm, Tr, Tu; { V T8, T9, Tc, Td; T8 = LD(&(xi[WS(is, 2)]), ivs, &(xi[0])); T9 = LD(&(xi[WS(is, 6)]), ivs, &(xi[0])); Ta = VSUB(T8, T9); Tk = VADD(T8, T9); Tc = LD(&(xi[0]), ivs, &(xi[0])); Td = LD(&(xi[WS(is, 4)]), ivs, &(xi[0])); Te = VSUB(Tc, Td); Tj = VADD(Tc, Td); { V T1, T2, T3, T4, T5, T6; T1 = LD(&(xi[WS(is, 1)]), ivs, &(xi[WS(is, 1)])); T2 = LD(&(xi[WS(is, 5)]), ivs, &(xi[WS(is, 1)])); T3 = VSUB(T1, T2); T4 = LD(&(xi[WS(is, 7)]), ivs, &(xi[WS(is, 1)])); T5 = LD(&(xi[WS(is, 3)]), ivs, &(xi[WS(is, 1)])); T6 = VSUB(T4, T5); T7 = VMUL(LDK(KP707106781), VSUB(T3, T6)); Tn = VADD(T4, T5); Tf = VMUL(LDK(KP707106781), VADD(T3, T6)); Tm = VADD(T1, T2); } } { V Ts, Tb, Tg, Tp, Tq, Tt; Tb = VBYI(VSUB(T7, Ta)); Tg = VSUB(Te, Tf); Tr = VADD(Tb, Tg); STM2(&(xo[6]), Tr, ovs, &(xo[2])); Ts = VSUB(Tg, Tb); STM2(&(xo[10]), Ts, ovs, &(xo[2])); Tp = VADD(Tj, Tk); Tq = VADD(Tm, Tn); Tt = VSUB(Tp, Tq); STM2(&(xo[8]), Tt, ovs, &(xo[0])); STN2(&(xo[8]), Tt, Ts, ovs); Tu = VADD(Tp, Tq); STM2(&(xo[0]), Tu, ovs, &(xo[0])); } { V Tw, Th, Ti, Tv; Th = VBYI(VADD(Ta, T7)); Ti = VADD(Te, Tf); Tv = VADD(Th, Ti); STM2(&(xo[2]), Tv, ovs, &(xo[2])); STN2(&(xo[0]), Tu, Tv, ovs); Tw = VSUB(Ti, Th); STM2(&(xo[14]), Tw, ovs, &(xo[2])); { V Tl, To, Tx, Ty; Tl = VSUB(Tj, Tk); To = VBYI(VSUB(Tm, Tn)); Tx = VSUB(Tl, To); STM2(&(xo[12]), Tx, ovs, &(xo[0])); STN2(&(xo[12]), Tx, Tw, ovs); Ty = VADD(Tl, To); STM2(&(xo[4]), Ty, ovs, &(xo[0])); STN2(&(xo[4]), Ty, Tr, ovs); } } } } VLEAVE(); } static const kdft_desc desc = { 8, XSIMD_STRING("n2bv_8"), {26, 2, 0, 0}, &GENUS, 0, 2, 0, 0 }; void XSIMD(codelet_n2bv_8) (planner *p) { X(kdft_register) (p, n2bv_8, &desc); } #endif /* HAVE_FMA */ fftw-3.3.3/dft/simd/common/n2bv_10.c0000644000175000017500000002210612054410412013715 00000000000000/* * Copyright (c) 2003, 2007-11 Matteo Frigo * Copyright (c) 2003, 2007-11 Massachusetts Institute of Technology * * This program is free software; you can redistribute it and/or modify * it under the terms of the GNU General Public License as published by * the Free Software Foundation; either version 2 of the License, or * (at your option) any later version. * * This program is distributed in the hope that it will be useful, * but WITHOUT ANY WARRANTY; without even the implied warranty of * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the * GNU General Public License for more details. * * You should have received a copy of the GNU General Public License * along with this program; if not, write to the Free Software * Foundation, Inc., 51 Franklin Street, Fifth Floor, Boston, MA 02110-1301 USA * */ /* This file was automatically generated --- DO NOT EDIT */ /* Generated on Sun Nov 25 07:37:29 EST 2012 */ #include "codelet-dft.h" #ifdef HAVE_FMA /* Generated by: ../../../genfft/gen_notw_c.native -fma -reorder-insns -schedule-for-pipeline -simd -compact -variables 4 -pipeline-latency 8 -sign 1 -n 10 -name n2bv_10 -with-ostride 2 -include n2b.h -store-multiple 2 */ /* * This function contains 42 FP additions, 22 FP multiplications, * (or, 24 additions, 4 multiplications, 18 fused multiply/add), * 53 stack variables, 4 constants, and 25 memory accesses */ #include "n2b.h" static void n2bv_10(const R *ri, const R *ii, R *ro, R *io, stride is, stride os, INT v, INT ivs, INT ovs) { DVK(KP559016994, +0.559016994374947424102293417182819058860154590); DVK(KP250000000, +0.250000000000000000000000000000000000000000000); DVK(KP618033988, +0.618033988749894848204586834365638117720309180); DVK(KP951056516, +0.951056516295153572116439333379382143405698634); { INT i; const R *xi; R *xo; xi = ii; xo = io; for (i = v; i > 0; i = i - VL, xi = xi + (VL * ivs), xo = xo + (VL * ovs), MAKE_VOLATILE_STRIDE(20, is), MAKE_VOLATILE_STRIDE(20, os)) { V Tb, Tr, T3, Ts, T6, Tw, Tg, Tt, T9, Tc, T1, T2; T1 = LD(&(xi[0]), ivs, &(xi[0])); T2 = LD(&(xi[WS(is, 5)]), ivs, &(xi[WS(is, 1)])); { V T4, T5, Te, Tf, T7, T8; T4 = LD(&(xi[WS(is, 2)]), ivs, &(xi[0])); T5 = LD(&(xi[WS(is, 7)]), ivs, &(xi[WS(is, 1)])); Te = LD(&(xi[WS(is, 6)]), ivs, &(xi[0])); Tf = LD(&(xi[WS(is, 1)]), ivs, &(xi[WS(is, 1)])); T7 = LD(&(xi[WS(is, 8)]), ivs, &(xi[0])); T8 = LD(&(xi[WS(is, 3)]), ivs, &(xi[WS(is, 1)])); Tb = LD(&(xi[WS(is, 4)]), ivs, &(xi[0])); Tr = VADD(T1, T2); T3 = VSUB(T1, T2); Ts = VADD(T4, T5); T6 = VSUB(T4, T5); Tw = VADD(Te, Tf); Tg = VSUB(Te, Tf); Tt = VADD(T7, T8); T9 = VSUB(T7, T8); Tc = LD(&(xi[WS(is, 9)]), ivs, &(xi[WS(is, 1)])); } { V TD, Tu, Tm, Ta, Td, Tv; TD = VSUB(Ts, Tt); Tu = VADD(Ts, Tt); Tm = VSUB(T6, T9); Ta = VADD(T6, T9); Td = VSUB(Tb, Tc); Tv = VADD(Tb, Tc); { V TC, Tx, Tn, Th; TC = VSUB(Tv, Tw); Tx = VADD(Tv, Tw); Tn = VSUB(Td, Tg); Th = VADD(Td, Tg); { V Ty, TA, TE, TG, Ti, Tk, To, Tq; Ty = VADD(Tu, Tx); TA = VSUB(Tu, Tx); TE = VMUL(LDK(KP951056516), VFNMS(LDK(KP618033988), TD, TC)); TG = VMUL(LDK(KP951056516), VFMA(LDK(KP618033988), TC, TD)); Ti = VADD(Ta, Th); Tk = VSUB(Ta, Th); To = VMUL(LDK(KP951056516), VFMA(LDK(KP618033988), Tn, Tm)); Tq = VMUL(LDK(KP951056516), VFNMS(LDK(KP618033988), Tm, Tn)); { V Tz, TH, Tj, TI; Tz = VFNMS(LDK(KP250000000), Ty, Tr); TH = VADD(Tr, Ty); STM2(&(xo[0]), TH, ovs, &(xo[0])); Tj = VFNMS(LDK(KP250000000), Ti, T3); TI = VADD(T3, Ti); STM2(&(xo[10]), TI, ovs, &(xo[2])); { V TB, TF, Tl, Tp; TB = VFNMS(LDK(KP559016994), TA, Tz); TF = VFMA(LDK(KP559016994), TA, Tz); Tl = VFMA(LDK(KP559016994), Tk, Tj); Tp = VFNMS(LDK(KP559016994), Tk, Tj); { V TJ, TK, TL, TM; TJ = VFNMSI(TG, TF); STM2(&(xo[8]), TJ, ovs, &(xo[0])); STN2(&(xo[8]), TJ, TI, ovs); TK = VFMAI(TG, TF); STM2(&(xo[12]), TK, ovs, &(xo[0])); TL = VFMAI(TE, TB); STM2(&(xo[16]), TL, ovs, &(xo[0])); TM = VFNMSI(TE, TB); STM2(&(xo[4]), TM, ovs, &(xo[0])); { V TN, TO, TP, TQ; TN = VFMAI(Tq, Tp); STM2(&(xo[6]), TN, ovs, &(xo[2])); STN2(&(xo[4]), TM, TN, ovs); TO = VFNMSI(Tq, Tp); STM2(&(xo[14]), TO, ovs, &(xo[2])); STN2(&(xo[12]), TK, TO, ovs); TP = VFNMSI(To, Tl); STM2(&(xo[18]), TP, ovs, &(xo[2])); STN2(&(xo[16]), TL, TP, ovs); TQ = VFMAI(To, Tl); STM2(&(xo[2]), TQ, ovs, &(xo[2])); STN2(&(xo[0]), TH, TQ, ovs); } } } } } } } } } VLEAVE(); } static const kdft_desc desc = { 10, XSIMD_STRING("n2bv_10"), {24, 4, 18, 0}, &GENUS, 0, 2, 0, 0 }; void XSIMD(codelet_n2bv_10) (planner *p) { X(kdft_register) (p, n2bv_10, &desc); } #else /* HAVE_FMA */ /* Generated by: ../../../genfft/gen_notw_c.native -simd -compact -variables 4 -pipeline-latency 8 -sign 1 -n 10 -name n2bv_10 -with-ostride 2 -include n2b.h -store-multiple 2 */ /* * This function contains 42 FP additions, 12 FP multiplications, * (or, 36 additions, 6 multiplications, 6 fused multiply/add), * 36 stack variables, 4 constants, and 25 memory accesses */ #include "n2b.h" static void n2bv_10(const R *ri, const R *ii, R *ro, R *io, stride is, stride os, INT v, INT ivs, INT ovs) { DVK(KP250000000, +0.250000000000000000000000000000000000000000000); DVK(KP559016994, +0.559016994374947424102293417182819058860154590); DVK(KP587785252, +0.587785252292473129168705954639072768597652438); DVK(KP951056516, +0.951056516295153572116439333379382143405698634); { INT i; const R *xi; R *xo; xi = ii; xo = io; for (i = v; i > 0; i = i - VL, xi = xi + (VL * ivs), xo = xo + (VL * ovs), MAKE_VOLATILE_STRIDE(20, is), MAKE_VOLATILE_STRIDE(20, os)) { V Tl, Ty, T7, Te, Tw, Tt, Tz, TA, TB, Tg, Th, Tm, Tj, Tk; Tj = LD(&(xi[0]), ivs, &(xi[0])); Tk = LD(&(xi[WS(is, 5)]), ivs, &(xi[WS(is, 1)])); Tl = VSUB(Tj, Tk); Ty = VADD(Tj, Tk); { V T3, Tr, Td, Tv, T6, Ts, Ta, Tu; { V T1, T2, Tb, Tc; T1 = LD(&(xi[WS(is, 2)]), ivs, &(xi[0])); T2 = LD(&(xi[WS(is, 7)]), ivs, &(xi[WS(is, 1)])); T3 = VSUB(T1, T2); Tr = VADD(T1, T2); Tb = LD(&(xi[WS(is, 6)]), ivs, &(xi[0])); Tc = LD(&(xi[WS(is, 1)]), ivs, &(xi[WS(is, 1)])); Td = VSUB(Tb, Tc); Tv = VADD(Tb, Tc); } { V T4, T5, T8, T9; T4 = LD(&(xi[WS(is, 8)]), ivs, &(xi[0])); T5 = LD(&(xi[WS(is, 3)]), ivs, &(xi[WS(is, 1)])); T6 = VSUB(T4, T5); Ts = VADD(T4, T5); T8 = LD(&(xi[WS(is, 4)]), ivs, &(xi[0])); T9 = LD(&(xi[WS(is, 9)]), ivs, &(xi[WS(is, 1)])); Ta = VSUB(T8, T9); Tu = VADD(T8, T9); } T7 = VSUB(T3, T6); Te = VSUB(Ta, Td); Tw = VSUB(Tu, Tv); Tt = VSUB(Tr, Ts); Tz = VADD(Tr, Ts); TA = VADD(Tu, Tv); TB = VADD(Tz, TA); Tg = VADD(T3, T6); Th = VADD(Ta, Td); Tm = VADD(Tg, Th); } { V TH, TI, TK, TL, TM; TH = VADD(Tl, Tm); STM2(&(xo[10]), TH, ovs, &(xo[2])); TI = VADD(Ty, TB); STM2(&(xo[0]), TI, ovs, &(xo[0])); { V Tf, Tq, To, Tp, Ti, Tn, TJ; Tf = VBYI(VFMA(LDK(KP951056516), T7, VMUL(LDK(KP587785252), Te))); Tq = VBYI(VFNMS(LDK(KP951056516), Te, VMUL(LDK(KP587785252), T7))); Ti = VMUL(LDK(KP559016994), VSUB(Tg, Th)); Tn = VFNMS(LDK(KP250000000), Tm, Tl); To = VADD(Ti, Tn); Tp = VSUB(Tn, Ti); TJ = VADD(Tf, To); STM2(&(xo[2]), TJ, ovs, &(xo[2])); STN2(&(xo[0]), TI, TJ, ovs); TK = VADD(Tq, Tp); STM2(&(xo[14]), TK, ovs, &(xo[2])); TL = VSUB(To, Tf); STM2(&(xo[18]), TL, ovs, &(xo[2])); TM = VSUB(Tp, Tq); STM2(&(xo[6]), TM, ovs, &(xo[2])); } { V Tx, TG, TE, TF, TC, TD; Tx = VBYI(VFNMS(LDK(KP951056516), Tw, VMUL(LDK(KP587785252), Tt))); TG = VBYI(VFMA(LDK(KP951056516), Tt, VMUL(LDK(KP587785252), Tw))); TC = VFNMS(LDK(KP250000000), TB, Ty); TD = VMUL(LDK(KP559016994), VSUB(Tz, TA)); TE = VSUB(TC, TD); TF = VADD(TD, TC); { V TN, TO, TP, TQ; TN = VADD(Tx, TE); STM2(&(xo[4]), TN, ovs, &(xo[0])); STN2(&(xo[4]), TN, TM, ovs); TO = VADD(TG, TF); STM2(&(xo[12]), TO, ovs, &(xo[0])); STN2(&(xo[12]), TO, TK, ovs); TP = VSUB(TE, Tx); STM2(&(xo[16]), TP, ovs, &(xo[0])); STN2(&(xo[16]), TP, TL, ovs); TQ = VSUB(TF, TG); STM2(&(xo[8]), TQ, ovs, &(xo[0])); STN2(&(xo[8]), TQ, TH, ovs); } } } } } VLEAVE(); } static const kdft_desc desc = { 10, XSIMD_STRING("n2bv_10"), {36, 6, 6, 0}, &GENUS, 0, 2, 0, 0 }; void XSIMD(codelet_n2bv_10) (planner *p) { X(kdft_register) (p, n2bv_10, &desc); } #endif /* HAVE_FMA */ fftw-3.3.3/dft/simd/common/n2bv_12.c0000644000175000017500000002267012054410412013725 00000000000000/* * Copyright (c) 2003, 2007-11 Matteo Frigo * Copyright (c) 2003, 2007-11 Massachusetts Institute of Technology * * This program is free software; you can redistribute it and/or modify * it under the terms of the GNU General Public License as published by * the Free Software Foundation; either version 2 of the License, or * (at your option) any later version. * * This program is distributed in the hope that it will be useful, * but WITHOUT ANY WARRANTY; without even the implied warranty of * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the * GNU General Public License for more details. * * You should have received a copy of the GNU General Public License * along with this program; if not, write to the Free Software * Foundation, Inc., 51 Franklin Street, Fifth Floor, Boston, MA 02110-1301 USA * */ /* This file was automatically generated --- DO NOT EDIT */ /* Generated on Sun Nov 25 07:37:30 EST 2012 */ #include "codelet-dft.h" #ifdef HAVE_FMA /* Generated by: ../../../genfft/gen_notw_c.native -fma -reorder-insns -schedule-for-pipeline -simd -compact -variables 4 -pipeline-latency 8 -sign 1 -n 12 -name n2bv_12 -with-ostride 2 -include n2b.h -store-multiple 2 */ /* * This function contains 48 FP additions, 20 FP multiplications, * (or, 30 additions, 2 multiplications, 18 fused multiply/add), * 61 stack variables, 2 constants, and 30 memory accesses */ #include "n2b.h" static void n2bv_12(const R *ri, const R *ii, R *ro, R *io, stride is, stride os, INT v, INT ivs, INT ovs) { DVK(KP866025403, +0.866025403784438646763723170752936183471402627); DVK(KP500000000, +0.500000000000000000000000000000000000000000000); { INT i; const R *xi; R *xo; xi = ii; xo = io; for (i = v; i > 0; i = i - VL, xi = xi + (VL * ivs), xo = xo + (VL * ovs), MAKE_VOLATILE_STRIDE(24, is), MAKE_VOLATILE_STRIDE(24, os)) { V T1, T6, Tc, Th, Td, Te, Ti, Tz, T4, TA, T9, Tj, Tf, Tw; { V T2, T3, T7, T8; T1 = LD(&(xi[0]), ivs, &(xi[0])); T6 = LD(&(xi[WS(is, 6)]), ivs, &(xi[0])); T2 = LD(&(xi[WS(is, 4)]), ivs, &(xi[0])); T3 = LD(&(xi[WS(is, 8)]), ivs, &(xi[0])); T7 = LD(&(xi[WS(is, 10)]), ivs, &(xi[0])); T8 = LD(&(xi[WS(is, 2)]), ivs, &(xi[0])); Tc = LD(&(xi[WS(is, 3)]), ivs, &(xi[WS(is, 1)])); Th = LD(&(xi[WS(is, 9)]), ivs, &(xi[WS(is, 1)])); Td = LD(&(xi[WS(is, 7)]), ivs, &(xi[WS(is, 1)])); Te = LD(&(xi[WS(is, 11)]), ivs, &(xi[WS(is, 1)])); Ti = LD(&(xi[WS(is, 1)]), ivs, &(xi[WS(is, 1)])); Tz = VSUB(T2, T3); T4 = VADD(T2, T3); TA = VSUB(T7, T8); T9 = VADD(T7, T8); Tj = LD(&(xi[WS(is, 5)]), ivs, &(xi[WS(is, 1)])); } Tf = VADD(Td, Te); Tw = VSUB(Td, Te); { V T5, Tp, TJ, TB, Ta, Tq, Tk, Tx, Tg, Ts; T5 = VADD(T1, T4); Tp = VFNMS(LDK(KP500000000), T4, T1); TJ = VSUB(Tz, TA); TB = VADD(Tz, TA); Ta = VADD(T6, T9); Tq = VFNMS(LDK(KP500000000), T9, T6); Tk = VADD(Ti, Tj); Tx = VSUB(Tj, Ti); Tg = VADD(Tc, Tf); Ts = VFNMS(LDK(KP500000000), Tf, Tc); { V Tr, TF, Tb, Tn, TG, Ty, Tl, Tt; Tr = VADD(Tp, Tq); TF = VSUB(Tp, Tq); Tb = VSUB(T5, Ta); Tn = VADD(T5, Ta); TG = VADD(Tw, Tx); Ty = VSUB(Tw, Tx); Tl = VADD(Th, Tk); Tt = VFNMS(LDK(KP500000000), Tk, Th); { V TC, TE, TH, TL, Tu, TI, Tm, To; TC = VMUL(LDK(KP866025403), VSUB(Ty, TB)); TE = VMUL(LDK(KP866025403), VADD(TB, Ty)); TH = VFNMS(LDK(KP866025403), TG, TF); TL = VFMA(LDK(KP866025403), TG, TF); Tu = VADD(Ts, Tt); TI = VSUB(Ts, Tt); Tm = VSUB(Tg, Tl); To = VADD(Tg, Tl); { V TK, TM, Tv, TD; TK = VFMA(LDK(KP866025403), TJ, TI); TM = VFNMS(LDK(KP866025403), TJ, TI); Tv = VSUB(Tr, Tu); TD = VADD(Tr, Tu); { V TN, TO, TP, TQ; TN = VADD(Tn, To); STM2(&(xo[0]), TN, ovs, &(xo[0])); TO = VSUB(Tn, To); STM2(&(xo[12]), TO, ovs, &(xo[0])); TP = VFMAI(Tm, Tb); STM2(&(xo[18]), TP, ovs, &(xo[2])); TQ = VFNMSI(Tm, Tb); STM2(&(xo[6]), TQ, ovs, &(xo[2])); { V TR, TS, TT, TU; TR = VFMAI(TM, TL); STM2(&(xo[10]), TR, ovs, &(xo[2])); TS = VFNMSI(TM, TL); STM2(&(xo[14]), TS, ovs, &(xo[2])); STN2(&(xo[12]), TO, TS, ovs); TT = VFNMSI(TK, TH); STM2(&(xo[22]), TT, ovs, &(xo[2])); TU = VFMAI(TK, TH); STM2(&(xo[2]), TU, ovs, &(xo[2])); STN2(&(xo[0]), TN, TU, ovs); { V TV, TW, TX, TY; TV = VFNMSI(TE, TD); STM2(&(xo[16]), TV, ovs, &(xo[0])); STN2(&(xo[16]), TV, TP, ovs); TW = VFMAI(TE, TD); STM2(&(xo[8]), TW, ovs, &(xo[0])); STN2(&(xo[8]), TW, TR, ovs); TX = VFMAI(TC, Tv); STM2(&(xo[4]), TX, ovs, &(xo[0])); STN2(&(xo[4]), TX, TQ, ovs); TY = VFNMSI(TC, Tv); STM2(&(xo[20]), TY, ovs, &(xo[0])); STN2(&(xo[20]), TY, TT, ovs); } } } } } } } } } VLEAVE(); } static const kdft_desc desc = { 12, XSIMD_STRING("n2bv_12"), {30, 2, 18, 0}, &GENUS, 0, 2, 0, 0 }; void XSIMD(codelet_n2bv_12) (planner *p) { X(kdft_register) (p, n2bv_12, &desc); } #else /* HAVE_FMA */ /* Generated by: ../../../genfft/gen_notw_c.native -simd -compact -variables 4 -pipeline-latency 8 -sign 1 -n 12 -name n2bv_12 -with-ostride 2 -include n2b.h -store-multiple 2 */ /* * This function contains 48 FP additions, 8 FP multiplications, * (or, 44 additions, 4 multiplications, 4 fused multiply/add), * 33 stack variables, 2 constants, and 30 memory accesses */ #include "n2b.h" static void n2bv_12(const R *ri, const R *ii, R *ro, R *io, stride is, stride os, INT v, INT ivs, INT ovs) { DVK(KP866025403, +0.866025403784438646763723170752936183471402627); DVK(KP500000000, +0.500000000000000000000000000000000000000000000); { INT i; const R *xi; R *xo; xi = ii; xo = io; for (i = v; i > 0; i = i - VL, xi = xi + (VL * ivs), xo = xo + (VL * ovs), MAKE_VOLATILE_STRIDE(24, is), MAKE_VOLATILE_STRIDE(24, os)) { V T5, Ta, TG, TF, Ty, Tm, Ti, Tp, TJ, TI, Tx, Ts; { V T1, T6, T4, Tk, T9, Tl; T1 = LD(&(xi[0]), ivs, &(xi[0])); T6 = LD(&(xi[WS(is, 6)]), ivs, &(xi[0])); { V T2, T3, T7, T8; T2 = LD(&(xi[WS(is, 4)]), ivs, &(xi[0])); T3 = LD(&(xi[WS(is, 8)]), ivs, &(xi[0])); T4 = VADD(T2, T3); Tk = VSUB(T2, T3); T7 = LD(&(xi[WS(is, 10)]), ivs, &(xi[0])); T8 = LD(&(xi[WS(is, 2)]), ivs, &(xi[0])); T9 = VADD(T7, T8); Tl = VSUB(T7, T8); } T5 = VFNMS(LDK(KP500000000), T4, T1); Ta = VFNMS(LDK(KP500000000), T9, T6); TG = VADD(T6, T9); TF = VADD(T1, T4); Ty = VADD(Tk, Tl); Tm = VMUL(LDK(KP866025403), VSUB(Tk, Tl)); } { V Tn, Tq, Te, To, Th, Tr; Tn = LD(&(xi[WS(is, 3)]), ivs, &(xi[WS(is, 1)])); Tq = LD(&(xi[WS(is, 9)]), ivs, &(xi[WS(is, 1)])); { V Tc, Td, Tf, Tg; Tc = LD(&(xi[WS(is, 7)]), ivs, &(xi[WS(is, 1)])); Td = LD(&(xi[WS(is, 11)]), ivs, &(xi[WS(is, 1)])); Te = VSUB(Tc, Td); To = VADD(Tc, Td); Tf = LD(&(xi[WS(is, 1)]), ivs, &(xi[WS(is, 1)])); Tg = LD(&(xi[WS(is, 5)]), ivs, &(xi[WS(is, 1)])); Th = VSUB(Tf, Tg); Tr = VADD(Tf, Tg); } Ti = VMUL(LDK(KP866025403), VSUB(Te, Th)); Tp = VFNMS(LDK(KP500000000), To, Tn); TJ = VADD(Tq, Tr); TI = VADD(Tn, To); Tx = VADD(Te, Th); Ts = VFNMS(LDK(KP500000000), Tr, Tq); } { V TN, TO, TP, TQ, TR, TS; { V TH, TK, TL, TM; TH = VSUB(TF, TG); TK = VBYI(VSUB(TI, TJ)); TN = VSUB(TH, TK); STM2(&(xo[6]), TN, ovs, &(xo[2])); TO = VADD(TH, TK); STM2(&(xo[18]), TO, ovs, &(xo[2])); TL = VADD(TF, TG); TM = VADD(TI, TJ); TP = VSUB(TL, TM); STM2(&(xo[12]), TP, ovs, &(xo[0])); TQ = VADD(TL, TM); STM2(&(xo[0]), TQ, ovs, &(xo[0])); } { V Tj, Tv, Tu, Tw, Tb, Tt, TT, TU; Tb = VSUB(T5, Ta); Tj = VSUB(Tb, Ti); Tv = VADD(Tb, Ti); Tt = VSUB(Tp, Ts); Tu = VBYI(VADD(Tm, Tt)); Tw = VBYI(VSUB(Tt, Tm)); TR = VSUB(Tj, Tu); STM2(&(xo[22]), TR, ovs, &(xo[2])); TS = VADD(Tv, Tw); STM2(&(xo[10]), TS, ovs, &(xo[2])); TT = VADD(Tj, Tu); STM2(&(xo[2]), TT, ovs, &(xo[2])); STN2(&(xo[0]), TQ, TT, ovs); TU = VSUB(Tv, Tw); STM2(&(xo[14]), TU, ovs, &(xo[2])); STN2(&(xo[12]), TP, TU, ovs); } { V Tz, TD, TC, TE, TA, TB; Tz = VBYI(VMUL(LDK(KP866025403), VSUB(Tx, Ty))); TD = VBYI(VMUL(LDK(KP866025403), VADD(Ty, Tx))); TA = VADD(T5, Ta); TB = VADD(Tp, Ts); TC = VSUB(TA, TB); TE = VADD(TA, TB); { V TV, TW, TX, TY; TV = VADD(Tz, TC); STM2(&(xo[4]), TV, ovs, &(xo[0])); STN2(&(xo[4]), TV, TN, ovs); TW = VSUB(TE, TD); STM2(&(xo[16]), TW, ovs, &(xo[0])); STN2(&(xo[16]), TW, TO, ovs); TX = VSUB(TC, Tz); STM2(&(xo[20]), TX, ovs, &(xo[0])); STN2(&(xo[20]), TX, TR, ovs); TY = VADD(TD, TE); STM2(&(xo[8]), TY, ovs, &(xo[0])); STN2(&(xo[8]), TY, TS, ovs); } } } } } VLEAVE(); } static const kdft_desc desc = { 12, XSIMD_STRING("n2bv_12"), {44, 4, 4, 0}, &GENUS, 0, 2, 0, 0 }; void XSIMD(codelet_n2bv_12) (planner *p) { X(kdft_register) (p, n2bv_12, &desc); } #endif /* HAVE_FMA */ fftw-3.3.3/dft/simd/common/n2bv_14.c0000644000175000017500000003146112054410413013726 00000000000000/* * Copyright (c) 2003, 2007-11 Matteo Frigo * Copyright (c) 2003, 2007-11 Massachusetts Institute of Technology * * This program is free software; you can redistribute it and/or modify * it under the terms of the GNU General Public License as published by * the Free Software Foundation; either version 2 of the License, or * (at your option) any later version. * * This program is distributed in the hope that it will be useful, * but WITHOUT ANY WARRANTY; without even the implied warranty of * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the * GNU General Public License for more details. * * You should have received a copy of the GNU General Public License * along with this program; if not, write to the Free Software * Foundation, Inc., 51 Franklin Street, Fifth Floor, Boston, MA 02110-1301 USA * */ /* This file was automatically generated --- DO NOT EDIT */ /* Generated on Sun Nov 25 07:37:30 EST 2012 */ #include "codelet-dft.h" #ifdef HAVE_FMA /* Generated by: ../../../genfft/gen_notw_c.native -fma -reorder-insns -schedule-for-pipeline -simd -compact -variables 4 -pipeline-latency 8 -sign 1 -n 14 -name n2bv_14 -with-ostride 2 -include n2b.h -store-multiple 2 */ /* * This function contains 74 FP additions, 48 FP multiplications, * (or, 32 additions, 6 multiplications, 42 fused multiply/add), * 65 stack variables, 6 constants, and 35 memory accesses */ #include "n2b.h" static void n2bv_14(const R *ri, const R *ii, R *ro, R *io, stride is, stride os, INT v, INT ivs, INT ovs) { DVK(KP900968867, +0.900968867902419126236102319507445051165919162); DVK(KP801937735, +0.801937735804838252472204639014890102331838324); DVK(KP974927912, +0.974927912181823607018131682993931217232785801); DVK(KP692021471, +0.692021471630095869627814897002069140197260599); DVK(KP554958132, +0.554958132087371191422194871006410481067288862); DVK(KP356895867, +0.356895867892209443894399510021300583399127187); { INT i; const R *xi; R *xo; xi = ii; xo = io; for (i = v; i > 0; i = i - VL, xi = xi + (VL * ivs), xo = xo + (VL * ovs), MAKE_VOLATILE_STRIDE(28, is), MAKE_VOLATILE_STRIDE(28, os)) { V TH, T3, TP, Tn, Ta, Tu, TU, TK, TO, Tk, TM, Tg, TL, Td, T1; V T2; T1 = LD(&(xi[0]), ivs, &(xi[0])); T2 = LD(&(xi[WS(is, 7)]), ivs, &(xi[WS(is, 1)])); { V Ti, TI, T6, TJ, T9, Tj, Te, Tf, Tb, Tc; { V T4, T5, T7, T8, Tl, Tm; T4 = LD(&(xi[WS(is, 2)]), ivs, &(xi[0])); T5 = LD(&(xi[WS(is, 9)]), ivs, &(xi[WS(is, 1)])); T7 = LD(&(xi[WS(is, 12)]), ivs, &(xi[0])); T8 = LD(&(xi[WS(is, 5)]), ivs, &(xi[WS(is, 1)])); Tl = LD(&(xi[WS(is, 8)]), ivs, &(xi[0])); Tm = LD(&(xi[WS(is, 1)]), ivs, &(xi[WS(is, 1)])); Ti = LD(&(xi[WS(is, 6)]), ivs, &(xi[0])); TH = VADD(T1, T2); T3 = VSUB(T1, T2); TI = VADD(T4, T5); T6 = VSUB(T4, T5); TJ = VADD(T7, T8); T9 = VSUB(T7, T8); TP = VADD(Tl, Tm); Tn = VSUB(Tl, Tm); Tj = LD(&(xi[WS(is, 13)]), ivs, &(xi[WS(is, 1)])); Te = LD(&(xi[WS(is, 10)]), ivs, &(xi[0])); Tf = LD(&(xi[WS(is, 3)]), ivs, &(xi[WS(is, 1)])); Tb = LD(&(xi[WS(is, 4)]), ivs, &(xi[0])); Tc = LD(&(xi[WS(is, 11)]), ivs, &(xi[WS(is, 1)])); } Ta = VADD(T6, T9); Tu = VSUB(T6, T9); TU = VSUB(TI, TJ); TK = VADD(TI, TJ); TO = VADD(Ti, Tj); Tk = VSUB(Ti, Tj); TM = VADD(Te, Tf); Tg = VSUB(Te, Tf); TL = VADD(Tb, Tc); Td = VSUB(Tb, Tc); } { V T19, T1a, T13, TG, TY, T18, TB, Tw, TT, Tz, T11, T16, TE, Tr, TV; V TQ; TV = VSUB(TP, TO); TQ = VADD(TO, TP); { V Ts, To, TW, TN; Ts = VSUB(Tk, Tn); To = VADD(Tk, Tn); TW = VSUB(TM, TL); TN = VADD(TL, TM); { V Tt, Th, TR, T12; Tt = VSUB(Td, Tg); Th = VADD(Td, Tg); TR = VFNMS(LDK(KP356895867), TK, TQ); T12 = VFNMS(LDK(KP554958132), TV, TU); { V Tx, TF, TZ, T14; Tx = VFNMS(LDK(KP356895867), Ta, To); TF = VFMA(LDK(KP554958132), Ts, Tu); T19 = VADD(TH, VADD(TK, VADD(TN, TQ))); STM2(&(xo[0]), T19, ovs, &(xo[0])); TZ = VFNMS(LDK(KP356895867), TN, TK); T14 = VFNMS(LDK(KP356895867), TQ, TN); { V TX, T17, TC, Tp; TX = VFMA(LDK(KP554958132), TW, TV); T17 = VFMA(LDK(KP554958132), TU, TW); T1a = VADD(T3, VADD(Ta, VADD(Th, To))); STM2(&(xo[14]), T1a, ovs, &(xo[2])); TC = VFNMS(LDK(KP356895867), Th, Ta); Tp = VFNMS(LDK(KP356895867), To, Th); { V TA, Tv, TS, Ty; TA = VFMA(LDK(KP554958132), Tt, Ts); Tv = VFNMS(LDK(KP554958132), Tu, Tt); TS = VFNMS(LDK(KP692021471), TR, TN); T13 = VMUL(LDK(KP974927912), VFNMS(LDK(KP801937735), T12, TW)); Ty = VFNMS(LDK(KP692021471), Tx, Th); TG = VMUL(LDK(KP974927912), VFMA(LDK(KP801937735), TF, Tt)); { V T10, T15, TD, Tq; T10 = VFNMS(LDK(KP692021471), TZ, TQ); T15 = VFNMS(LDK(KP692021471), T14, TK); TY = VMUL(LDK(KP974927912), VFMA(LDK(KP801937735), TX, TU)); T18 = VMUL(LDK(KP974927912), VFNMS(LDK(KP801937735), T17, TV)); TD = VFNMS(LDK(KP692021471), TC, To); Tq = VFNMS(LDK(KP692021471), Tp, Ta); TB = VMUL(LDK(KP974927912), VFNMS(LDK(KP801937735), TA, Tu)); Tw = VMUL(LDK(KP974927912), VFNMS(LDK(KP801937735), Tv, Ts)); TT = VFNMS(LDK(KP900968867), TS, TH); Tz = VFNMS(LDK(KP900968867), Ty, T3); T11 = VFNMS(LDK(KP900968867), T10, TH); T16 = VFNMS(LDK(KP900968867), T15, TH); TE = VFNMS(LDK(KP900968867), TD, T3); Tr = VFNMS(LDK(KP900968867), Tq, T3); } } } } } } { V T1b, T1c, T1d, T1e; T1b = VFMAI(TY, TT); STM2(&(xo[4]), T1b, ovs, &(xo[0])); T1c = VFNMSI(TY, TT); STM2(&(xo[24]), T1c, ovs, &(xo[0])); T1d = VFMAI(TB, Tz); STM2(&(xo[18]), T1d, ovs, &(xo[2])); T1e = VFNMSI(TB, Tz); STM2(&(xo[10]), T1e, ovs, &(xo[2])); { V T1f, T1g, T1h, T1i; T1f = VFMAI(T13, T11); STM2(&(xo[12]), T1f, ovs, &(xo[0])); STN2(&(xo[12]), T1f, T1a, ovs); T1g = VFNMSI(T13, T11); STM2(&(xo[16]), T1g, ovs, &(xo[0])); STN2(&(xo[16]), T1g, T1d, ovs); T1h = VFMAI(T18, T16); STM2(&(xo[8]), T1h, ovs, &(xo[0])); STN2(&(xo[8]), T1h, T1e, ovs); T1i = VFNMSI(T18, T16); STM2(&(xo[20]), T1i, ovs, &(xo[0])); { V T1j, T1k, T1l, T1m; T1j = VFNMSI(TG, TE); STM2(&(xo[26]), T1j, ovs, &(xo[2])); STN2(&(xo[24]), T1c, T1j, ovs); T1k = VFMAI(TG, TE); STM2(&(xo[2]), T1k, ovs, &(xo[2])); STN2(&(xo[0]), T19, T1k, ovs); T1l = VFNMSI(Tw, Tr); STM2(&(xo[22]), T1l, ovs, &(xo[2])); STN2(&(xo[20]), T1i, T1l, ovs); T1m = VFMAI(Tw, Tr); STM2(&(xo[6]), T1m, ovs, &(xo[2])); STN2(&(xo[4]), T1b, T1m, ovs); } } } } } } VLEAVE(); } static const kdft_desc desc = { 14, XSIMD_STRING("n2bv_14"), {32, 6, 42, 0}, &GENUS, 0, 2, 0, 0 }; void XSIMD(codelet_n2bv_14) (planner *p) { X(kdft_register) (p, n2bv_14, &desc); } #else /* HAVE_FMA */ /* Generated by: ../../../genfft/gen_notw_c.native -simd -compact -variables 4 -pipeline-latency 8 -sign 1 -n 14 -name n2bv_14 -with-ostride 2 -include n2b.h -store-multiple 2 */ /* * This function contains 74 FP additions, 36 FP multiplications, * (or, 50 additions, 12 multiplications, 24 fused multiply/add), * 41 stack variables, 6 constants, and 35 memory accesses */ #include "n2b.h" static void n2bv_14(const R *ri, const R *ii, R *ro, R *io, stride is, stride os, INT v, INT ivs, INT ovs) { DVK(KP900968867, +0.900968867902419126236102319507445051165919162); DVK(KP222520933, +0.222520933956314404288902564496794759466355569); DVK(KP623489801, +0.623489801858733530525004884004239810632274731); DVK(KP781831482, +0.781831482468029808708444526674057750232334519); DVK(KP974927912, +0.974927912181823607018131682993931217232785801); DVK(KP433883739, +0.433883739117558120475768332848358754609990728); { INT i; const R *xi; R *xo; xi = ii; xo = io; for (i = v; i > 0; i = i - VL, xi = xi + (VL * ivs), xo = xo + (VL * ovs), MAKE_VOLATILE_STRIDE(28, is), MAKE_VOLATILE_STRIDE(28, os)) { V Tp, Ty, Tl, TL, Tq, TE, T7, TJ, Ts, TB, Te, TK, Tr, TH, Tn; V To; Tn = LD(&(xi[0]), ivs, &(xi[0])); To = LD(&(xi[WS(is, 7)]), ivs, &(xi[WS(is, 1)])); Tp = VSUB(Tn, To); Ty = VADD(Tn, To); { V Th, TC, Tk, TD; { V Tf, Tg, Ti, Tj; Tf = LD(&(xi[WS(is, 4)]), ivs, &(xi[0])); Tg = LD(&(xi[WS(is, 11)]), ivs, &(xi[WS(is, 1)])); Th = VSUB(Tf, Tg); TC = VADD(Tf, Tg); Ti = LD(&(xi[WS(is, 10)]), ivs, &(xi[0])); Tj = LD(&(xi[WS(is, 3)]), ivs, &(xi[WS(is, 1)])); Tk = VSUB(Ti, Tj); TD = VADD(Ti, Tj); } Tl = VSUB(Th, Tk); TL = VSUB(TD, TC); Tq = VADD(Th, Tk); TE = VADD(TC, TD); } { V T3, Tz, T6, TA; { V T1, T2, T4, T5; T1 = LD(&(xi[WS(is, 2)]), ivs, &(xi[0])); T2 = LD(&(xi[WS(is, 9)]), ivs, &(xi[WS(is, 1)])); T3 = VSUB(T1, T2); Tz = VADD(T1, T2); T4 = LD(&(xi[WS(is, 12)]), ivs, &(xi[0])); T5 = LD(&(xi[WS(is, 5)]), ivs, &(xi[WS(is, 1)])); T6 = VSUB(T4, T5); TA = VADD(T4, T5); } T7 = VSUB(T3, T6); TJ = VSUB(Tz, TA); Ts = VADD(T3, T6); TB = VADD(Tz, TA); } { V Ta, TF, Td, TG; { V T8, T9, Tb, Tc; T8 = LD(&(xi[WS(is, 6)]), ivs, &(xi[0])); T9 = LD(&(xi[WS(is, 13)]), ivs, &(xi[WS(is, 1)])); Ta = VSUB(T8, T9); TF = VADD(T8, T9); Tb = LD(&(xi[WS(is, 8)]), ivs, &(xi[0])); Tc = LD(&(xi[WS(is, 1)]), ivs, &(xi[WS(is, 1)])); Td = VSUB(Tb, Tc); TG = VADD(Tb, Tc); } Te = VSUB(Ta, Td); TK = VSUB(TG, TF); Tr = VADD(Ta, Td); TH = VADD(TF, TG); } { V TR, TS, TU, TV; TR = VADD(Tp, VADD(Ts, VADD(Tq, Tr))); STM2(&(xo[14]), TR, ovs, &(xo[2])); TS = VADD(Ty, VADD(TB, VADD(TE, TH))); STM2(&(xo[0]), TS, ovs, &(xo[0])); { V TT, Tm, Tt, TQ, TP, TW; Tm = VBYI(VFMA(LDK(KP433883739), T7, VFNMS(LDK(KP781831482), Tl, VMUL(LDK(KP974927912), Te)))); Tt = VFMA(LDK(KP623489801), Tq, VFNMS(LDK(KP222520933), Tr, VFNMS(LDK(KP900968867), Ts, Tp))); TT = VADD(Tm, Tt); STM2(&(xo[6]), TT, ovs, &(xo[2])); TU = VSUB(Tt, Tm); STM2(&(xo[22]), TU, ovs, &(xo[2])); TQ = VBYI(VFMA(LDK(KP974927912), TJ, VFMA(LDK(KP433883739), TL, VMUL(LDK(KP781831482), TK)))); TP = VFMA(LDK(KP623489801), TH, VFNMS(LDK(KP900968867), TE, VFNMS(LDK(KP222520933), TB, Ty))); TV = VSUB(TP, TQ); STM2(&(xo[24]), TV, ovs, &(xo[0])); TW = VADD(TP, TQ); STM2(&(xo[4]), TW, ovs, &(xo[0])); STN2(&(xo[4]), TW, TT, ovs); } { V T10, TM, TI, TZ; { V Tu, Tv, TX, TY; Tu = VBYI(VFMA(LDK(KP781831482), T7, VFMA(LDK(KP974927912), Tl, VMUL(LDK(KP433883739), Te)))); Tv = VFMA(LDK(KP623489801), Ts, VFNMS(LDK(KP900968867), Tr, VFNMS(LDK(KP222520933), Tq, Tp))); TX = VADD(Tu, Tv); STM2(&(xo[2]), TX, ovs, &(xo[2])); STN2(&(xo[0]), TS, TX, ovs); TY = VSUB(Tv, Tu); STM2(&(xo[26]), TY, ovs, &(xo[2])); STN2(&(xo[24]), TV, TY, ovs); } TM = VBYI(VFNMS(LDK(KP433883739), TK, VFNMS(LDK(KP974927912), TL, VMUL(LDK(KP781831482), TJ)))); TI = VFMA(LDK(KP623489801), TB, VFNMS(LDK(KP900968867), TH, VFNMS(LDK(KP222520933), TE, Ty))); TZ = VSUB(TI, TM); STM2(&(xo[12]), TZ, ovs, &(xo[0])); STN2(&(xo[12]), TZ, TR, ovs); T10 = VADD(TI, TM); STM2(&(xo[16]), T10, ovs, &(xo[0])); { V T11, TO, TN, T12; TO = VBYI(VFMA(LDK(KP433883739), TJ, VFNMS(LDK(KP974927912), TK, VMUL(LDK(KP781831482), TL)))); TN = VFMA(LDK(KP623489801), TE, VFNMS(LDK(KP222520933), TH, VFNMS(LDK(KP900968867), TB, Ty))); T11 = VSUB(TN, TO); STM2(&(xo[8]), T11, ovs, &(xo[0])); T12 = VADD(TN, TO); STM2(&(xo[20]), T12, ovs, &(xo[0])); STN2(&(xo[20]), T12, TU, ovs); { V Tx, Tw, T13, T14; Tx = VBYI(VFNMS(LDK(KP781831482), Te, VFNMS(LDK(KP433883739), Tl, VMUL(LDK(KP974927912), T7)))); Tw = VFMA(LDK(KP623489801), Tr, VFNMS(LDK(KP900968867), Tq, VFNMS(LDK(KP222520933), Ts, Tp))); T13 = VSUB(Tw, Tx); STM2(&(xo[10]), T13, ovs, &(xo[2])); STN2(&(xo[8]), T11, T13, ovs); T14 = VADD(Tx, Tw); STM2(&(xo[18]), T14, ovs, &(xo[2])); STN2(&(xo[16]), T10, T14, ovs); } } } } } } VLEAVE(); } static const kdft_desc desc = { 14, XSIMD_STRING("n2bv_14"), {50, 12, 24, 0}, &GENUS, 0, 2, 0, 0 }; void XSIMD(codelet_n2bv_14) (planner *p) { X(kdft_register) (p, n2bv_14, &desc); } #endif /* HAVE_FMA */ fftw-3.3.3/dft/simd/common/n2bv_16.c0000644000175000017500000003176512054410414013740 00000000000000/* * Copyright (c) 2003, 2007-11 Matteo Frigo * Copyright (c) 2003, 2007-11 Massachusetts Institute of Technology * * This program is free software; you can redistribute it and/or modify * it under the terms of the GNU General Public License as published by * the Free Software Foundation; either version 2 of the License, or * (at your option) any later version. * * This program is distributed in the hope that it will be useful, * but WITHOUT ANY WARRANTY; without even the implied warranty of * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the * GNU General Public License for more details. * * You should have received a copy of the GNU General Public License * along with this program; if not, write to the Free Software * Foundation, Inc., 51 Franklin Street, Fifth Floor, Boston, MA 02110-1301 USA * */ /* This file was automatically generated --- DO NOT EDIT */ /* Generated on Sun Nov 25 07:37:31 EST 2012 */ #include "codelet-dft.h" #ifdef HAVE_FMA /* Generated by: ../../../genfft/gen_notw_c.native -fma -reorder-insns -schedule-for-pipeline -simd -compact -variables 4 -pipeline-latency 8 -sign 1 -n 16 -name n2bv_16 -with-ostride 2 -include n2b.h -store-multiple 2 */ /* * This function contains 72 FP additions, 34 FP multiplications, * (or, 38 additions, 0 multiplications, 34 fused multiply/add), * 62 stack variables, 3 constants, and 40 memory accesses */ #include "n2b.h" static void n2bv_16(const R *ri, const R *ii, R *ro, R *io, stride is, stride os, INT v, INT ivs, INT ovs) { DVK(KP923879532, +0.923879532511286756128183189396788286822416626); DVK(KP414213562, +0.414213562373095048801688724209698078569671875); DVK(KP707106781, +0.707106781186547524400844362104849039284835938); { INT i; const R *xi; R *xo; xi = ii; xo = io; for (i = v; i > 0; i = i - VL, xi = xi + (VL * ivs), xo = xo + (VL * ovs), MAKE_VOLATILE_STRIDE(32, is), MAKE_VOLATILE_STRIDE(32, os)) { V T7, Tu, TF, TB, T13, TL, TO, TX, TC, Te, TP, Th, TQ, Tk, TW; V T16; { V TH, TU, Tz, Tf, TK, TV, TA, TM, Ta, TN, Td, Tg, Ti, Tj; { V T1, T2, T4, T5, To, Tp, Tr, Ts; T1 = LD(&(xi[0]), ivs, &(xi[0])); T2 = LD(&(xi[WS(is, 8)]), ivs, &(xi[0])); T4 = LD(&(xi[WS(is, 4)]), ivs, &(xi[0])); T5 = LD(&(xi[WS(is, 12)]), ivs, &(xi[0])); To = LD(&(xi[WS(is, 2)]), ivs, &(xi[0])); Tp = LD(&(xi[WS(is, 10)]), ivs, &(xi[0])); Tr = LD(&(xi[WS(is, 14)]), ivs, &(xi[0])); Ts = LD(&(xi[WS(is, 6)]), ivs, &(xi[0])); { V T8, TI, Tq, TJ, Tt, T9, Tb, Tc, T3, T6; T8 = LD(&(xi[WS(is, 1)]), ivs, &(xi[WS(is, 1)])); TH = VSUB(T1, T2); T3 = VADD(T1, T2); TU = VSUB(T4, T5); T6 = VADD(T4, T5); TI = VSUB(To, Tp); Tq = VADD(To, Tp); TJ = VSUB(Tr, Ts); Tt = VADD(Tr, Ts); T9 = LD(&(xi[WS(is, 9)]), ivs, &(xi[WS(is, 1)])); Tb = LD(&(xi[WS(is, 5)]), ivs, &(xi[WS(is, 1)])); Tc = LD(&(xi[WS(is, 13)]), ivs, &(xi[WS(is, 1)])); T7 = VSUB(T3, T6); Tz = VADD(T3, T6); Tf = LD(&(xi[WS(is, 15)]), ivs, &(xi[WS(is, 1)])); TK = VADD(TI, TJ); TV = VSUB(TI, TJ); TA = VADD(Tq, Tt); Tu = VSUB(Tq, Tt); TM = VSUB(T8, T9); Ta = VADD(T8, T9); TN = VSUB(Tb, Tc); Td = VADD(Tb, Tc); Tg = LD(&(xi[WS(is, 7)]), ivs, &(xi[WS(is, 1)])); Ti = LD(&(xi[WS(is, 3)]), ivs, &(xi[WS(is, 1)])); Tj = LD(&(xi[WS(is, 11)]), ivs, &(xi[WS(is, 1)])); } } TF = VADD(Tz, TA); TB = VSUB(Tz, TA); T13 = VFNMS(LDK(KP707106781), TK, TH); TL = VFMA(LDK(KP707106781), TK, TH); TO = VFNMS(LDK(KP414213562), TN, TM); TX = VFMA(LDK(KP414213562), TM, TN); TC = VADD(Ta, Td); Te = VSUB(Ta, Td); TP = VSUB(Tf, Tg); Th = VADD(Tf, Tg); TQ = VSUB(Tj, Ti); Tk = VADD(Ti, Tj); TW = VFMA(LDK(KP707106781), TV, TU); T16 = VFNMS(LDK(KP707106781), TV, TU); } { V TY, TR, Tl, TD; TY = VFMA(LDK(KP414213562), TP, TQ); TR = VFNMS(LDK(KP414213562), TQ, TP); Tl = VSUB(Th, Tk); TD = VADD(Th, Tk); { V TS, T17, TZ, T14; TS = VADD(TO, TR); T17 = VSUB(TO, TR); TZ = VSUB(TX, TY); T14 = VADD(TX, TY); { V TE, TG, Tm, Tv; TE = VSUB(TC, TD); TG = VADD(TC, TD); Tm = VADD(Te, Tl); Tv = VSUB(Te, Tl); { V T18, T1a, TT, T11; T18 = VFMA(LDK(KP923879532), T17, T16); T1a = VFNMS(LDK(KP923879532), T17, T16); TT = VFNMS(LDK(KP923879532), TS, TL); T11 = VFMA(LDK(KP923879532), TS, TL); { V T15, T19, T10, T12; T15 = VFNMS(LDK(KP923879532), T14, T13); T19 = VFMA(LDK(KP923879532), T14, T13); T10 = VFNMS(LDK(KP923879532), TZ, TW); T12 = VFMA(LDK(KP923879532), TZ, TW); { V T1b, T1c, T1d, T1e; T1b = VADD(TF, TG); STM2(&(xo[0]), T1b, ovs, &(xo[0])); T1c = VSUB(TF, TG); STM2(&(xo[16]), T1c, ovs, &(xo[0])); T1d = VFMAI(TE, TB); STM2(&(xo[8]), T1d, ovs, &(xo[0])); T1e = VFNMSI(TE, TB); STM2(&(xo[24]), T1e, ovs, &(xo[0])); { V Tw, Ty, Tn, Tx; Tw = VFNMS(LDK(KP707106781), Tv, Tu); Ty = VFMA(LDK(KP707106781), Tv, Tu); Tn = VFNMS(LDK(KP707106781), Tm, T7); Tx = VFMA(LDK(KP707106781), Tm, T7); { V T1f, T1g, T1h, T1i; T1f = VFNMSI(T1a, T19); STM2(&(xo[6]), T1f, ovs, &(xo[2])); T1g = VFMAI(T1a, T19); STM2(&(xo[26]), T1g, ovs, &(xo[2])); STN2(&(xo[24]), T1e, T1g, ovs); T1h = VFNMSI(T18, T15); STM2(&(xo[22]), T1h, ovs, &(xo[2])); T1i = VFMAI(T18, T15); STM2(&(xo[10]), T1i, ovs, &(xo[2])); STN2(&(xo[8]), T1d, T1i, ovs); { V T1j, T1k, T1l, T1m; T1j = VFNMSI(T12, T11); STM2(&(xo[30]), T1j, ovs, &(xo[2])); T1k = VFMAI(T12, T11); STM2(&(xo[2]), T1k, ovs, &(xo[2])); STN2(&(xo[0]), T1b, T1k, ovs); T1l = VFMAI(T10, TT); STM2(&(xo[18]), T1l, ovs, &(xo[2])); STN2(&(xo[16]), T1c, T1l, ovs); T1m = VFNMSI(T10, TT); STM2(&(xo[14]), T1m, ovs, &(xo[2])); { V T1n, T1o, T1p, T1q; T1n = VFMAI(Ty, Tx); STM2(&(xo[4]), T1n, ovs, &(xo[0])); STN2(&(xo[4]), T1n, T1f, ovs); T1o = VFNMSI(Ty, Tx); STM2(&(xo[28]), T1o, ovs, &(xo[0])); STN2(&(xo[28]), T1o, T1j, ovs); T1p = VFMAI(Tw, Tn); STM2(&(xo[20]), T1p, ovs, &(xo[0])); STN2(&(xo[20]), T1p, T1h, ovs); T1q = VFNMSI(Tw, Tn); STM2(&(xo[12]), T1q, ovs, &(xo[0])); STN2(&(xo[12]), T1q, T1m, ovs); } } } } } } } } } } } } VLEAVE(); } static const kdft_desc desc = { 16, XSIMD_STRING("n2bv_16"), {38, 0, 34, 0}, &GENUS, 0, 2, 0, 0 }; void XSIMD(codelet_n2bv_16) (planner *p) { X(kdft_register) (p, n2bv_16, &desc); } #else /* HAVE_FMA */ /* Generated by: ../../../genfft/gen_notw_c.native -simd -compact -variables 4 -pipeline-latency 8 -sign 1 -n 16 -name n2bv_16 -with-ostride 2 -include n2b.h -store-multiple 2 */ /* * This function contains 72 FP additions, 12 FP multiplications, * (or, 68 additions, 8 multiplications, 4 fused multiply/add), * 38 stack variables, 3 constants, and 40 memory accesses */ #include "n2b.h" static void n2bv_16(const R *ri, const R *ii, R *ro, R *io, stride is, stride os, INT v, INT ivs, INT ovs) { DVK(KP382683432, +0.382683432365089771728459984030398866761344562); DVK(KP923879532, +0.923879532511286756128183189396788286822416626); DVK(KP707106781, +0.707106781186547524400844362104849039284835938); { INT i; const R *xi; R *xo; xi = ii; xo = io; for (i = v; i > 0; i = i - VL, xi = xi + (VL * ivs), xo = xo + (VL * ovs), MAKE_VOLATILE_STRIDE(32, is), MAKE_VOLATILE_STRIDE(32, os)) { V Tp, T13, Tu, TY, Tm, T14, Tv, TU, T7, T16, Tx, TN, Te, T17, Ty; V TQ; { V Tn, To, TX, Ts, Tt, TW; Tn = LD(&(xi[WS(is, 4)]), ivs, &(xi[0])); To = LD(&(xi[WS(is, 12)]), ivs, &(xi[0])); TX = VADD(Tn, To); Ts = LD(&(xi[0]), ivs, &(xi[0])); Tt = LD(&(xi[WS(is, 8)]), ivs, &(xi[0])); TW = VADD(Ts, Tt); Tp = VSUB(Tn, To); T13 = VADD(TW, TX); Tu = VSUB(Ts, Tt); TY = VSUB(TW, TX); } { V Ti, TS, Tl, TT; { V Tg, Th, Tj, Tk; Tg = LD(&(xi[WS(is, 2)]), ivs, &(xi[0])); Th = LD(&(xi[WS(is, 10)]), ivs, &(xi[0])); Ti = VSUB(Tg, Th); TS = VADD(Tg, Th); Tj = LD(&(xi[WS(is, 14)]), ivs, &(xi[0])); Tk = LD(&(xi[WS(is, 6)]), ivs, &(xi[0])); Tl = VSUB(Tj, Tk); TT = VADD(Tj, Tk); } Tm = VMUL(LDK(KP707106781), VSUB(Ti, Tl)); T14 = VADD(TS, TT); Tv = VMUL(LDK(KP707106781), VADD(Ti, Tl)); TU = VSUB(TS, TT); } { V T3, TL, T6, TM; { V T1, T2, T4, T5; T1 = LD(&(xi[WS(is, 1)]), ivs, &(xi[WS(is, 1)])); T2 = LD(&(xi[WS(is, 9)]), ivs, &(xi[WS(is, 1)])); T3 = VSUB(T1, T2); TL = VADD(T1, T2); T4 = LD(&(xi[WS(is, 5)]), ivs, &(xi[WS(is, 1)])); T5 = LD(&(xi[WS(is, 13)]), ivs, &(xi[WS(is, 1)])); T6 = VSUB(T4, T5); TM = VADD(T4, T5); } T7 = VFNMS(LDK(KP382683432), T6, VMUL(LDK(KP923879532), T3)); T16 = VADD(TL, TM); Tx = VFMA(LDK(KP382683432), T3, VMUL(LDK(KP923879532), T6)); TN = VSUB(TL, TM); } { V Ta, TO, Td, TP; { V T8, T9, Tb, Tc; T8 = LD(&(xi[WS(is, 15)]), ivs, &(xi[WS(is, 1)])); T9 = LD(&(xi[WS(is, 7)]), ivs, &(xi[WS(is, 1)])); Ta = VSUB(T8, T9); TO = VADD(T8, T9); Tb = LD(&(xi[WS(is, 3)]), ivs, &(xi[WS(is, 1)])); Tc = LD(&(xi[WS(is, 11)]), ivs, &(xi[WS(is, 1)])); Td = VSUB(Tb, Tc); TP = VADD(Tb, Tc); } Te = VFMA(LDK(KP923879532), Ta, VMUL(LDK(KP382683432), Td)); T17 = VADD(TO, TP); Ty = VFNMS(LDK(KP382683432), Ta, VMUL(LDK(KP923879532), Td)); TQ = VSUB(TO, TP); } { V T1b, T1c, T1d, T1e; { V T15, T18, T19, T1a; T15 = VSUB(T13, T14); T18 = VBYI(VSUB(T16, T17)); T1b = VSUB(T15, T18); STM2(&(xo[24]), T1b, ovs, &(xo[0])); T1c = VADD(T15, T18); STM2(&(xo[8]), T1c, ovs, &(xo[0])); T19 = VADD(T13, T14); T1a = VADD(T16, T17); T1d = VSUB(T19, T1a); STM2(&(xo[16]), T1d, ovs, &(xo[0])); T1e = VADD(T19, T1a); STM2(&(xo[0]), T1e, ovs, &(xo[0])); } { V T1f, T1g, T1h, T1i; { V TV, T11, T10, T12, TR, TZ; TR = VMUL(LDK(KP707106781), VSUB(TN, TQ)); TV = VBYI(VSUB(TR, TU)); T11 = VBYI(VADD(TU, TR)); TZ = VMUL(LDK(KP707106781), VADD(TN, TQ)); T10 = VSUB(TY, TZ); T12 = VADD(TY, TZ); T1f = VADD(TV, T10); STM2(&(xo[12]), T1f, ovs, &(xo[0])); T1g = VSUB(T12, T11); STM2(&(xo[28]), T1g, ovs, &(xo[0])); T1h = VSUB(T10, TV); STM2(&(xo[20]), T1h, ovs, &(xo[0])); T1i = VADD(T11, T12); STM2(&(xo[4]), T1i, ovs, &(xo[0])); } { V Tr, TB, TA, TC; { V Tf, Tq, Tw, Tz; Tf = VSUB(T7, Te); Tq = VSUB(Tm, Tp); Tr = VBYI(VSUB(Tf, Tq)); TB = VBYI(VADD(Tq, Tf)); Tw = VSUB(Tu, Tv); Tz = VSUB(Tx, Ty); TA = VSUB(Tw, Tz); TC = VADD(Tw, Tz); } { V T1j, T1k, T1l, T1m; T1j = VADD(Tr, TA); STM2(&(xo[10]), T1j, ovs, &(xo[2])); STN2(&(xo[8]), T1c, T1j, ovs); T1k = VSUB(TC, TB); STM2(&(xo[26]), T1k, ovs, &(xo[2])); STN2(&(xo[24]), T1b, T1k, ovs); T1l = VSUB(TA, Tr); STM2(&(xo[22]), T1l, ovs, &(xo[2])); STN2(&(xo[20]), T1h, T1l, ovs); T1m = VADD(TB, TC); STM2(&(xo[6]), T1m, ovs, &(xo[2])); STN2(&(xo[4]), T1i, T1m, ovs); } } { V TF, TJ, TI, TK; { V TD, TE, TG, TH; TD = VADD(Tu, Tv); TE = VADD(T7, Te); TF = VADD(TD, TE); TJ = VSUB(TD, TE); TG = VADD(Tp, Tm); TH = VADD(Tx, Ty); TI = VBYI(VADD(TG, TH)); TK = VBYI(VSUB(TH, TG)); } { V T1n, T1o, T1p, T1q; T1n = VSUB(TF, TI); STM2(&(xo[30]), T1n, ovs, &(xo[2])); STN2(&(xo[28]), T1g, T1n, ovs); T1o = VADD(TJ, TK); STM2(&(xo[14]), T1o, ovs, &(xo[2])); STN2(&(xo[12]), T1f, T1o, ovs); T1p = VADD(TF, TI); STM2(&(xo[2]), T1p, ovs, &(xo[2])); STN2(&(xo[0]), T1e, T1p, ovs); T1q = VSUB(TJ, TK); STM2(&(xo[18]), T1q, ovs, &(xo[2])); STN2(&(xo[16]), T1d, T1q, ovs); } } } } } } VLEAVE(); } static const kdft_desc desc = { 16, XSIMD_STRING("n2bv_16"), {68, 8, 4, 0}, &GENUS, 0, 2, 0, 0 }; void XSIMD(codelet_n2bv_16) (planner *p) { X(kdft_register) (p, n2bv_16, &desc); } #endif /* HAVE_FMA */ fftw-3.3.3/dft/simd/common/n2bv_32.c0000644000175000017500000006556412054410420013737 00000000000000/* * Copyright (c) 2003, 2007-11 Matteo Frigo * Copyright (c) 2003, 2007-11 Massachusetts Institute of Technology * * This program is free software; you can redistribute it and/or modify * it under the terms of the GNU General Public License as published by * the Free Software Foundation; either version 2 of the License, or * (at your option) any later version. * * This program is distributed in the hope that it will be useful, * but WITHOUT ANY WARRANTY; without even the implied warranty of * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the * GNU General Public License for more details. * * You should have received a copy of the GNU General Public License * along with this program; if not, write to the Free Software * Foundation, Inc., 51 Franklin Street, Fifth Floor, Boston, MA 02110-1301 USA * */ /* This file was automatically generated --- DO NOT EDIT */ /* Generated on Sun Nov 25 07:37:32 EST 2012 */ #include "codelet-dft.h" #ifdef HAVE_FMA /* Generated by: ../../../genfft/gen_notw_c.native -fma -reorder-insns -schedule-for-pipeline -simd -compact -variables 4 -pipeline-latency 8 -sign 1 -n 32 -name n2bv_32 -with-ostride 2 -include n2b.h -store-multiple 2 */ /* * This function contains 186 FP additions, 98 FP multiplications, * (or, 88 additions, 0 multiplications, 98 fused multiply/add), * 120 stack variables, 7 constants, and 80 memory accesses */ #include "n2b.h" static void n2bv_32(const R *ri, const R *ii, R *ro, R *io, stride is, stride os, INT v, INT ivs, INT ovs) { DVK(KP980785280, +0.980785280403230449126182236134239036973933731); DVK(KP831469612, +0.831469612302545237078788377617905756738560812); DVK(KP198912367, +0.198912367379658006911597622644676228597850501); DVK(KP668178637, +0.668178637919298919997757686523080761552472251); DVK(KP923879532, +0.923879532511286756128183189396788286822416626); DVK(KP414213562, +0.414213562373095048801688724209698078569671875); DVK(KP707106781, +0.707106781186547524400844362104849039284835938); { INT i; const R *xi; R *xo; xi = ii; xo = io; for (i = v; i > 0; i = i - VL, xi = xi + (VL * ivs), xo = xo + (VL * ovs), MAKE_VOLATILE_STRIDE(64, is), MAKE_VOLATILE_STRIDE(64, os)) { V T31, T32, T33, T34, T35, T36, T37, T38, T39, T3a, T3b, T3c, T1h, Tr, T3d; V T3e, T3f, T3g, T1a, T1k, TI, T1b, T1L, T1P, T1I, T1G, T1O, T1Q, T1H, T1z; V T1c, TZ; { V T2x, T1T, T2K, T1W, T1p, Tb, T1A, T16, Tu, TF, T2O, T2H, T2b, T2t, TY; V T1w, TT, T1v, T20, T2C, Tj, Te, T2e, To, T2i, T23, T2D, TB, TG, Th; V T2f, Tk; { V TL, TW, TP, TQ, T2F, T27, T28, TO; { V T1, T2, T12, T13, T4, T5, T7, T8; T1 = LD(&(xi[0]), ivs, &(xi[0])); T2 = LD(&(xi[WS(is, 16)]), ivs, &(xi[0])); T12 = LD(&(xi[WS(is, 8)]), ivs, &(xi[0])); T13 = LD(&(xi[WS(is, 24)]), ivs, &(xi[0])); T4 = LD(&(xi[WS(is, 4)]), ivs, &(xi[0])); T5 = LD(&(xi[WS(is, 20)]), ivs, &(xi[0])); T7 = LD(&(xi[WS(is, 28)]), ivs, &(xi[0])); T8 = LD(&(xi[WS(is, 12)]), ivs, &(xi[0])); { V TM, T25, T26, TN; { V TJ, T3, T14, T1U, T6, T1V, T9, TK, TU, TV, T1R, T1S, Ta, T15; TJ = LD(&(xi[WS(is, 31)]), ivs, &(xi[WS(is, 1)])); T1R = VADD(T1, T2); T3 = VSUB(T1, T2); T1S = VADD(T12, T13); T14 = VSUB(T12, T13); T1U = VADD(T4, T5); T6 = VSUB(T4, T5); T1V = VADD(T7, T8); T9 = VSUB(T7, T8); TK = LD(&(xi[WS(is, 15)]), ivs, &(xi[WS(is, 1)])); TU = LD(&(xi[WS(is, 23)]), ivs, &(xi[WS(is, 1)])); T2x = VSUB(T1R, T1S); T1T = VADD(T1R, T1S); TV = LD(&(xi[WS(is, 7)]), ivs, &(xi[WS(is, 1)])); TM = LD(&(xi[WS(is, 3)]), ivs, &(xi[WS(is, 1)])); T2K = VSUB(T1U, T1V); T1W = VADD(T1U, T1V); Ta = VADD(T6, T9); T15 = VSUB(T6, T9); T25 = VADD(TJ, TK); TL = VSUB(TJ, TK); T26 = VADD(TV, TU); TW = VSUB(TU, TV); TN = LD(&(xi[WS(is, 19)]), ivs, &(xi[WS(is, 1)])); TP = LD(&(xi[WS(is, 27)]), ivs, &(xi[WS(is, 1)])); T1p = VFNMS(LDK(KP707106781), Ta, T3); Tb = VFMA(LDK(KP707106781), Ta, T3); T1A = VFNMS(LDK(KP707106781), T15, T14); T16 = VFMA(LDK(KP707106781), T15, T14); TQ = LD(&(xi[WS(is, 11)]), ivs, &(xi[WS(is, 1)])); } T2F = VSUB(T25, T26); T27 = VADD(T25, T26); T28 = VADD(TM, TN); TO = VSUB(TM, TN); } } { V Ty, T21, Tx, Tz, T1Y, T1Z; { V Ts, Tt, TD, T29, TR, TE, Tv, Tw; Ts = LD(&(xi[WS(is, 1)]), ivs, &(xi[WS(is, 1)])); Tt = LD(&(xi[WS(is, 17)]), ivs, &(xi[WS(is, 1)])); TD = LD(&(xi[WS(is, 9)]), ivs, &(xi[WS(is, 1)])); T29 = VADD(TP, TQ); TR = VSUB(TP, TQ); TE = LD(&(xi[WS(is, 25)]), ivs, &(xi[WS(is, 1)])); Tv = LD(&(xi[WS(is, 5)]), ivs, &(xi[WS(is, 1)])); Tw = LD(&(xi[WS(is, 21)]), ivs, &(xi[WS(is, 1)])); Ty = LD(&(xi[WS(is, 29)]), ivs, &(xi[WS(is, 1)])); T1Y = VADD(Ts, Tt); Tu = VSUB(Ts, Tt); { V T2G, T2a, TX, TS; T2G = VSUB(T29, T28); T2a = VADD(T28, T29); TX = VSUB(TR, TO); TS = VADD(TO, TR); T1Z = VADD(TD, TE); TF = VSUB(TD, TE); T21 = VADD(Tv, Tw); Tx = VSUB(Tv, Tw); T2O = VFMA(LDK(KP414213562), T2F, T2G); T2H = VFNMS(LDK(KP414213562), T2G, T2F); T2b = VSUB(T27, T2a); T2t = VADD(T27, T2a); TY = VFMA(LDK(KP707106781), TX, TW); T1w = VFNMS(LDK(KP707106781), TX, TW); TT = VFMA(LDK(KP707106781), TS, TL); T1v = VFNMS(LDK(KP707106781), TS, TL); Tz = LD(&(xi[WS(is, 13)]), ivs, &(xi[WS(is, 1)])); } } T20 = VADD(T1Y, T1Z); T2C = VSUB(T1Y, T1Z); { V Tc, Td, Tm, Tn; Tc = LD(&(xi[WS(is, 2)]), ivs, &(xi[0])); Td = LD(&(xi[WS(is, 18)]), ivs, &(xi[0])); Tm = LD(&(xi[WS(is, 22)]), ivs, &(xi[0])); Tn = LD(&(xi[WS(is, 6)]), ivs, &(xi[0])); { V Tf, TA, T22, Tg; Tf = LD(&(xi[WS(is, 10)]), ivs, &(xi[0])); TA = VSUB(Ty, Tz); T22 = VADD(Ty, Tz); Tg = LD(&(xi[WS(is, 26)]), ivs, &(xi[0])); Tj = LD(&(xi[WS(is, 30)]), ivs, &(xi[0])); Te = VSUB(Tc, Td); T2e = VADD(Tc, Td); To = VSUB(Tm, Tn); T2i = VADD(Tn, Tm); T23 = VADD(T21, T22); T2D = VSUB(T21, T22); TB = VADD(Tx, TA); TG = VSUB(Tx, TA); Th = VSUB(Tf, Tg); T2f = VADD(Tf, Tg); Tk = LD(&(xi[WS(is, 14)]), ivs, &(xi[0])); } } } } { V T1t, TH, T1s, TC, T2P, T2U, T2n, T2d, T2w, T2u, T1q, T19, T1B, Tq, T2W; V T2M, T2B, T2T, T2v, T2r, T2o, T2m, T2X, T2I; { V T1X, T2p, T2E, T2N, T2s, T2y, T2g, T17, Ti, T2h, Tl, T2c, T2l, T24; T1X = VSUB(T1T, T1W); T2p = VADD(T1T, T1W); T2E = VFNMS(LDK(KP414213562), T2D, T2C); T2N = VFMA(LDK(KP414213562), T2C, T2D); T2s = VADD(T20, T23); T24 = VSUB(T20, T23); T1t = VFNMS(LDK(KP707106781), TG, TF); TH = VFMA(LDK(KP707106781), TG, TF); T1s = VFNMS(LDK(KP707106781), TB, Tu); TC = VFMA(LDK(KP707106781), TB, Tu); T2y = VSUB(T2e, T2f); T2g = VADD(T2e, T2f); T17 = VFMA(LDK(KP414213562), Te, Th); Ti = VFNMS(LDK(KP414213562), Th, Te); T2h = VADD(Tj, Tk); Tl = VSUB(Tj, Tk); T2c = VADD(T24, T2b); T2l = VSUB(T24, T2b); { V T2L, T2A, T2q, T2k; T2P = VSUB(T2N, T2O); T2U = VADD(T2N, T2O); { V T2z, T2j, T18, Tp; T2z = VSUB(T2h, T2i); T2j = VADD(T2h, T2i); T18 = VFMA(LDK(KP414213562), Tl, To); Tp = VFNMS(LDK(KP414213562), To, Tl); T2n = VFMA(LDK(KP707106781), T2c, T1X); T2d = VFNMS(LDK(KP707106781), T2c, T1X); T2w = VADD(T2s, T2t); T2u = VSUB(T2s, T2t); T2L = VSUB(T2y, T2z); T2A = VADD(T2y, T2z); T2q = VADD(T2g, T2j); T2k = VSUB(T2g, T2j); T1q = VADD(T17, T18); T19 = VSUB(T17, T18); T1B = VSUB(Ti, Tp); Tq = VADD(Ti, Tp); } T2W = VFNMS(LDK(KP707106781), T2L, T2K); T2M = VFMA(LDK(KP707106781), T2L, T2K); T2B = VFMA(LDK(KP707106781), T2A, T2x); T2T = VFNMS(LDK(KP707106781), T2A, T2x); T2v = VADD(T2p, T2q); T2r = VSUB(T2p, T2q); T2o = VFMA(LDK(KP707106781), T2l, T2k); T2m = VFNMS(LDK(KP707106781), T2l, T2k); T2X = VSUB(T2E, T2H); T2I = VADD(T2E, T2H); } } { V T2V, T2Z, T2Y, T30, T2R, T2J; T2V = VFNMS(LDK(KP923879532), T2U, T2T); T2Z = VFMA(LDK(KP923879532), T2U, T2T); T31 = VSUB(T2v, T2w); STM2(&(xo[32]), T31, ovs, &(xo[0])); T32 = VADD(T2v, T2w); STM2(&(xo[0]), T32, ovs, &(xo[0])); T33 = VFMAI(T2u, T2r); STM2(&(xo[16]), T33, ovs, &(xo[0])); T34 = VFNMSI(T2u, T2r); STM2(&(xo[48]), T34, ovs, &(xo[0])); T35 = VFMAI(T2o, T2n); STM2(&(xo[8]), T35, ovs, &(xo[0])); T36 = VFNMSI(T2o, T2n); STM2(&(xo[56]), T36, ovs, &(xo[0])); T37 = VFMAI(T2m, T2d); STM2(&(xo[40]), T37, ovs, &(xo[0])); T38 = VFNMSI(T2m, T2d); STM2(&(xo[24]), T38, ovs, &(xo[0])); T2Y = VFMA(LDK(KP923879532), T2X, T2W); T30 = VFNMS(LDK(KP923879532), T2X, T2W); T2R = VFMA(LDK(KP923879532), T2I, T2B); T2J = VFNMS(LDK(KP923879532), T2I, T2B); { V T1J, T1r, T1C, T1M, T2S, T2Q, T1u, T1D, T1E, T1x; T1J = VFNMS(LDK(KP923879532), T1q, T1p); T1r = VFMA(LDK(KP923879532), T1q, T1p); T1C = VFNMS(LDK(KP923879532), T1B, T1A); T1M = VFMA(LDK(KP923879532), T1B, T1A); T39 = VFNMSI(T30, T2Z); STM2(&(xo[12]), T39, ovs, &(xo[0])); T3a = VFMAI(T30, T2Z); STM2(&(xo[52]), T3a, ovs, &(xo[0])); T3b = VFNMSI(T2Y, T2V); STM2(&(xo[44]), T3b, ovs, &(xo[0])); T3c = VFMAI(T2Y, T2V); STM2(&(xo[20]), T3c, ovs, &(xo[0])); T2S = VFMA(LDK(KP923879532), T2P, T2M); T2Q = VFNMS(LDK(KP923879532), T2P, T2M); T1u = VFMA(LDK(KP668178637), T1t, T1s); T1D = VFNMS(LDK(KP668178637), T1s, T1t); T1E = VFNMS(LDK(KP668178637), T1v, T1w); T1x = VFMA(LDK(KP668178637), T1w, T1v); { V T1K, T1F, T1N, T1y; T1h = VFNMS(LDK(KP923879532), Tq, Tb); Tr = VFMA(LDK(KP923879532), Tq, Tb); T3d = VFNMSI(T2S, T2R); STM2(&(xo[60]), T3d, ovs, &(xo[0])); T3e = VFMAI(T2S, T2R); STM2(&(xo[4]), T3e, ovs, &(xo[0])); T3f = VFMAI(T2Q, T2J); STM2(&(xo[36]), T3f, ovs, &(xo[0])); T3g = VFNMSI(T2Q, T2J); STM2(&(xo[28]), T3g, ovs, &(xo[0])); T1K = VADD(T1D, T1E); T1F = VSUB(T1D, T1E); T1N = VSUB(T1u, T1x); T1y = VADD(T1u, T1x); T1a = VFMA(LDK(KP923879532), T19, T16); T1k = VFNMS(LDK(KP923879532), T19, T16); TI = VFNMS(LDK(KP198912367), TH, TC); T1b = VFMA(LDK(KP198912367), TC, TH); T1L = VFMA(LDK(KP831469612), T1K, T1J); T1P = VFNMS(LDK(KP831469612), T1K, T1J); T1I = VFMA(LDK(KP831469612), T1F, T1C); T1G = VFNMS(LDK(KP831469612), T1F, T1C); T1O = VFNMS(LDK(KP831469612), T1N, T1M); T1Q = VFMA(LDK(KP831469612), T1N, T1M); T1H = VFMA(LDK(KP831469612), T1y, T1r); T1z = VFNMS(LDK(KP831469612), T1y, T1r); T1c = VFMA(LDK(KP198912367), TT, TY); TZ = VFNMS(LDK(KP198912367), TY, TT); } } } } } { V T1d, T1i, T10, T1l; { V T3h, T3i, T3j, T3k; T3h = VFMAI(T1O, T1L); STM2(&(xo[42]), T3h, ovs, &(xo[2])); STN2(&(xo[40]), T37, T3h, ovs); T3i = VFNMSI(T1O, T1L); STM2(&(xo[22]), T3i, ovs, &(xo[2])); STN2(&(xo[20]), T3c, T3i, ovs); T3j = VFNMSI(T1Q, T1P); STM2(&(xo[54]), T3j, ovs, &(xo[2])); STN2(&(xo[52]), T3a, T3j, ovs); T3k = VFMAI(T1Q, T1P); STM2(&(xo[10]), T3k, ovs, &(xo[2])); STN2(&(xo[8]), T35, T3k, ovs); { V T3l, T3m, T3n, T3o; T3l = VFMAI(T1I, T1H); STM2(&(xo[58]), T3l, ovs, &(xo[2])); STN2(&(xo[56]), T36, T3l, ovs); T3m = VFNMSI(T1I, T1H); STM2(&(xo[6]), T3m, ovs, &(xo[2])); STN2(&(xo[4]), T3e, T3m, ovs); T3n = VFMAI(T1G, T1z); STM2(&(xo[26]), T3n, ovs, &(xo[2])); STN2(&(xo[24]), T38, T3n, ovs); T3o = VFNMSI(T1G, T1z); STM2(&(xo[38]), T3o, ovs, &(xo[2])); STN2(&(xo[36]), T3f, T3o, ovs); T1d = VSUB(T1b, T1c); T1i = VADD(T1b, T1c); T10 = VADD(TI, TZ); T1l = VSUB(TI, TZ); } } { V T1n, T1j, T1e, T1g, T1o, T1m, T11, T1f; T1n = VFMA(LDK(KP980785280), T1i, T1h); T1j = VFNMS(LDK(KP980785280), T1i, T1h); T1e = VFNMS(LDK(KP980785280), T1d, T1a); T1g = VFMA(LDK(KP980785280), T1d, T1a); T1o = VFNMS(LDK(KP980785280), T1l, T1k); T1m = VFMA(LDK(KP980785280), T1l, T1k); T11 = VFNMS(LDK(KP980785280), T10, Tr); T1f = VFMA(LDK(KP980785280), T10, Tr); { V T3p, T3q, T3r, T3s; T3p = VFNMSI(T1m, T1j); STM2(&(xo[46]), T3p, ovs, &(xo[2])); STN2(&(xo[44]), T3b, T3p, ovs); T3q = VFMAI(T1m, T1j); STM2(&(xo[18]), T3q, ovs, &(xo[2])); STN2(&(xo[16]), T33, T3q, ovs); T3r = VFMAI(T1o, T1n); STM2(&(xo[50]), T3r, ovs, &(xo[2])); STN2(&(xo[48]), T34, T3r, ovs); T3s = VFNMSI(T1o, T1n); STM2(&(xo[14]), T3s, ovs, &(xo[2])); STN2(&(xo[12]), T39, T3s, ovs); { V T3t, T3u, T3v, T3w; T3t = VFMAI(T1g, T1f); STM2(&(xo[2]), T3t, ovs, &(xo[2])); STN2(&(xo[0]), T32, T3t, ovs); T3u = VFNMSI(T1g, T1f); STM2(&(xo[62]), T3u, ovs, &(xo[2])); STN2(&(xo[60]), T3d, T3u, ovs); T3v = VFMAI(T1e, T11); STM2(&(xo[34]), T3v, ovs, &(xo[2])); STN2(&(xo[32]), T31, T3v, ovs); T3w = VFNMSI(T1e, T11); STM2(&(xo[30]), T3w, ovs, &(xo[2])); STN2(&(xo[28]), T3g, T3w, ovs); } } } } } } VLEAVE(); } static const kdft_desc desc = { 32, XSIMD_STRING("n2bv_32"), {88, 0, 98, 0}, &GENUS, 0, 2, 0, 0 }; void XSIMD(codelet_n2bv_32) (planner *p) { X(kdft_register) (p, n2bv_32, &desc); } #else /* HAVE_FMA */ /* Generated by: ../../../genfft/gen_notw_c.native -simd -compact -variables 4 -pipeline-latency 8 -sign 1 -n 32 -name n2bv_32 -with-ostride 2 -include n2b.h -store-multiple 2 */ /* * This function contains 186 FP additions, 42 FP multiplications, * (or, 170 additions, 26 multiplications, 16 fused multiply/add), * 72 stack variables, 7 constants, and 80 memory accesses */ #include "n2b.h" static void n2bv_32(const R *ri, const R *ii, R *ro, R *io, stride is, stride os, INT v, INT ivs, INT ovs) { DVK(KP555570233, +0.555570233019602224742830813948532874374937191); DVK(KP831469612, +0.831469612302545237078788377617905756738560812); DVK(KP195090322, +0.195090322016128267848284868477022240927691618); DVK(KP980785280, +0.980785280403230449126182236134239036973933731); DVK(KP382683432, +0.382683432365089771728459984030398866761344562); DVK(KP923879532, +0.923879532511286756128183189396788286822416626); DVK(KP707106781, +0.707106781186547524400844362104849039284835938); { INT i; const R *xi; R *xo; xi = ii; xo = io; for (i = v; i > 0; i = i - VL, xi = xi + (VL * ivs), xo = xo + (VL * ovs), MAKE_VOLATILE_STRIDE(64, is), MAKE_VOLATILE_STRIDE(64, os)) { V T2f, T2k, T2N, T2M, T19, T1B, Tb, T1p, TT, T1v, TY, T1w, T2E, T2F, T2G; V T24, T2o, TC, T1s, TH, T1t, T2B, T2C, T2D, T1X, T2n, T2I, T2J, Tq, T1A; V T14, T1q, T2c, T2l; { V T3, T2i, T18, T2j, T6, T2d, T9, T2e, T15, Ta; { V T1, T2, T16, T17; T1 = LD(&(xi[0]), ivs, &(xi[0])); T2 = LD(&(xi[WS(is, 16)]), ivs, &(xi[0])); T3 = VSUB(T1, T2); T2i = VADD(T1, T2); T16 = LD(&(xi[WS(is, 8)]), ivs, &(xi[0])); T17 = LD(&(xi[WS(is, 24)]), ivs, &(xi[0])); T18 = VSUB(T16, T17); T2j = VADD(T16, T17); } { V T4, T5, T7, T8; T4 = LD(&(xi[WS(is, 4)]), ivs, &(xi[0])); T5 = LD(&(xi[WS(is, 20)]), ivs, &(xi[0])); T6 = VSUB(T4, T5); T2d = VADD(T4, T5); T7 = LD(&(xi[WS(is, 28)]), ivs, &(xi[0])); T8 = LD(&(xi[WS(is, 12)]), ivs, &(xi[0])); T9 = VSUB(T7, T8); T2e = VADD(T7, T8); } T2f = VSUB(T2d, T2e); T2k = VSUB(T2i, T2j); T2N = VADD(T2d, T2e); T2M = VADD(T2i, T2j); T15 = VMUL(LDK(KP707106781), VSUB(T6, T9)); T19 = VSUB(T15, T18); T1B = VADD(T18, T15); Ta = VMUL(LDK(KP707106781), VADD(T6, T9)); Tb = VSUB(T3, Ta); T1p = VADD(T3, Ta); } { V TL, T21, TW, T1Y, TO, T22, TS, T1Z; { V TJ, TK, TU, TV; TJ = LD(&(xi[WS(is, 3)]), ivs, &(xi[WS(is, 1)])); TK = LD(&(xi[WS(is, 19)]), ivs, &(xi[WS(is, 1)])); TL = VSUB(TJ, TK); T21 = VADD(TJ, TK); TU = LD(&(xi[WS(is, 31)]), ivs, &(xi[WS(is, 1)])); TV = LD(&(xi[WS(is, 15)]), ivs, &(xi[WS(is, 1)])); TW = VSUB(TU, TV); T1Y = VADD(TU, TV); } { V TM, TN, TQ, TR; TM = LD(&(xi[WS(is, 27)]), ivs, &(xi[WS(is, 1)])); TN = LD(&(xi[WS(is, 11)]), ivs, &(xi[WS(is, 1)])); TO = VSUB(TM, TN); T22 = VADD(TM, TN); TQ = LD(&(xi[WS(is, 7)]), ivs, &(xi[WS(is, 1)])); TR = LD(&(xi[WS(is, 23)]), ivs, &(xi[WS(is, 1)])); TS = VSUB(TQ, TR); T1Z = VADD(TQ, TR); } { V TP, TX, T20, T23; TP = VMUL(LDK(KP707106781), VSUB(TL, TO)); TT = VSUB(TP, TS); T1v = VADD(TS, TP); TX = VMUL(LDK(KP707106781), VADD(TL, TO)); TY = VSUB(TW, TX); T1w = VADD(TW, TX); T2E = VADD(T1Y, T1Z); T2F = VADD(T21, T22); T2G = VSUB(T2E, T2F); T20 = VSUB(T1Y, T1Z); T23 = VSUB(T21, T22); T24 = VFMA(LDK(KP923879532), T20, VMUL(LDK(KP382683432), T23)); T2o = VFNMS(LDK(KP382683432), T20, VMUL(LDK(KP923879532), T23)); } } { V Tu, T1U, TF, T1R, Tx, T1V, TB, T1S; { V Ts, Tt, TD, TE; Ts = LD(&(xi[WS(is, 5)]), ivs, &(xi[WS(is, 1)])); Tt = LD(&(xi[WS(is, 21)]), ivs, &(xi[WS(is, 1)])); Tu = VSUB(Ts, Tt); T1U = VADD(Ts, Tt); TD = LD(&(xi[WS(is, 1)]), ivs, &(xi[WS(is, 1)])); TE = LD(&(xi[WS(is, 17)]), ivs, &(xi[WS(is, 1)])); TF = VSUB(TD, TE); T1R = VADD(TD, TE); } { V Tv, Tw, Tz, TA; Tv = LD(&(xi[WS(is, 29)]), ivs, &(xi[WS(is, 1)])); Tw = LD(&(xi[WS(is, 13)]), ivs, &(xi[WS(is, 1)])); Tx = VSUB(Tv, Tw); T1V = VADD(Tv, Tw); Tz = LD(&(xi[WS(is, 9)]), ivs, &(xi[WS(is, 1)])); TA = LD(&(xi[WS(is, 25)]), ivs, &(xi[WS(is, 1)])); TB = VSUB(Tz, TA); T1S = VADD(Tz, TA); } { V Ty, TG, T1T, T1W; Ty = VMUL(LDK(KP707106781), VSUB(Tu, Tx)); TC = VSUB(Ty, TB); T1s = VADD(TB, Ty); TG = VMUL(LDK(KP707106781), VADD(Tu, Tx)); TH = VSUB(TF, TG); T1t = VADD(TF, TG); T2B = VADD(T1R, T1S); T2C = VADD(T1U, T1V); T2D = VSUB(T2B, T2C); T1T = VSUB(T1R, T1S); T1W = VSUB(T1U, T1V); T1X = VFNMS(LDK(KP382683432), T1W, VMUL(LDK(KP923879532), T1T)); T2n = VFMA(LDK(KP382683432), T1T, VMUL(LDK(KP923879532), T1W)); } } { V Te, T26, To, T29, Th, T27, Tl, T2a, Ti, Tp; { V Tc, Td, Tm, Tn; Tc = LD(&(xi[WS(is, 2)]), ivs, &(xi[0])); Td = LD(&(xi[WS(is, 18)]), ivs, &(xi[0])); Te = VSUB(Tc, Td); T26 = VADD(Tc, Td); Tm = LD(&(xi[WS(is, 30)]), ivs, &(xi[0])); Tn = LD(&(xi[WS(is, 14)]), ivs, &(xi[0])); To = VSUB(Tm, Tn); T29 = VADD(Tm, Tn); } { V Tf, Tg, Tj, Tk; Tf = LD(&(xi[WS(is, 10)]), ivs, &(xi[0])); Tg = LD(&(xi[WS(is, 26)]), ivs, &(xi[0])); Th = VSUB(Tf, Tg); T27 = VADD(Tf, Tg); Tj = LD(&(xi[WS(is, 6)]), ivs, &(xi[0])); Tk = LD(&(xi[WS(is, 22)]), ivs, &(xi[0])); Tl = VSUB(Tj, Tk); T2a = VADD(Tj, Tk); } T2I = VADD(T26, T27); T2J = VADD(T29, T2a); Ti = VFMA(LDK(KP382683432), Te, VMUL(LDK(KP923879532), Th)); Tp = VFNMS(LDK(KP382683432), To, VMUL(LDK(KP923879532), Tl)); Tq = VSUB(Ti, Tp); T1A = VADD(Ti, Tp); { V T12, T13, T28, T2b; T12 = VFNMS(LDK(KP382683432), Th, VMUL(LDK(KP923879532), Te)); T13 = VFMA(LDK(KP923879532), To, VMUL(LDK(KP382683432), Tl)); T14 = VSUB(T12, T13); T1q = VADD(T12, T13); T28 = VSUB(T26, T27); T2b = VSUB(T29, T2a); T2c = VMUL(LDK(KP707106781), VSUB(T28, T2b)); T2l = VMUL(LDK(KP707106781), VADD(T28, T2b)); } } { V T31, T32, T33, T34, T35, T36, T37, T38, T39, T3a, T3b, T3c; { V T2L, T2R, T2Q, T2S; { V T2H, T2K, T2O, T2P; T2H = VMUL(LDK(KP707106781), VSUB(T2D, T2G)); T2K = VSUB(T2I, T2J); T2L = VBYI(VSUB(T2H, T2K)); T2R = VBYI(VADD(T2K, T2H)); T2O = VSUB(T2M, T2N); T2P = VMUL(LDK(KP707106781), VADD(T2D, T2G)); T2Q = VSUB(T2O, T2P); T2S = VADD(T2O, T2P); } T31 = VADD(T2L, T2Q); STM2(&(xo[24]), T31, ovs, &(xo[0])); T32 = VSUB(T2S, T2R); STM2(&(xo[56]), T32, ovs, &(xo[0])); T33 = VSUB(T2Q, T2L); STM2(&(xo[40]), T33, ovs, &(xo[0])); T34 = VADD(T2R, T2S); STM2(&(xo[8]), T34, ovs, &(xo[0])); } { V T2h, T2r, T2q, T2s; { V T25, T2g, T2m, T2p; T25 = VSUB(T1X, T24); T2g = VSUB(T2c, T2f); T2h = VBYI(VSUB(T25, T2g)); T2r = VBYI(VADD(T2g, T25)); T2m = VSUB(T2k, T2l); T2p = VSUB(T2n, T2o); T2q = VSUB(T2m, T2p); T2s = VADD(T2m, T2p); } T35 = VADD(T2h, T2q); STM2(&(xo[20]), T35, ovs, &(xo[0])); T36 = VSUB(T2s, T2r); STM2(&(xo[52]), T36, ovs, &(xo[0])); T37 = VSUB(T2q, T2h); STM2(&(xo[44]), T37, ovs, &(xo[0])); T38 = VADD(T2r, T2s); STM2(&(xo[12]), T38, ovs, &(xo[0])); } { V T2V, T2Z, T2Y, T30; { V T2T, T2U, T2W, T2X; T2T = VADD(T2M, T2N); T2U = VADD(T2I, T2J); T2V = VSUB(T2T, T2U); T2Z = VADD(T2T, T2U); T2W = VADD(T2B, T2C); T2X = VADD(T2E, T2F); T2Y = VBYI(VSUB(T2W, T2X)); T30 = VADD(T2W, T2X); } T39 = VSUB(T2V, T2Y); STM2(&(xo[48]), T39, ovs, &(xo[0])); T3a = VADD(T2Z, T30); STM2(&(xo[0]), T3a, ovs, &(xo[0])); T3b = VADD(T2V, T2Y); STM2(&(xo[16]), T3b, ovs, &(xo[0])); T3c = VSUB(T2Z, T30); STM2(&(xo[32]), T3c, ovs, &(xo[0])); } { V T3d, T3e, T3f, T3g; { V T2v, T2z, T2y, T2A; { V T2t, T2u, T2w, T2x; T2t = VADD(T2k, T2l); T2u = VADD(T1X, T24); T2v = VADD(T2t, T2u); T2z = VSUB(T2t, T2u); T2w = VADD(T2f, T2c); T2x = VADD(T2n, T2o); T2y = VBYI(VADD(T2w, T2x)); T2A = VBYI(VSUB(T2x, T2w)); } T3d = VSUB(T2v, T2y); STM2(&(xo[60]), T3d, ovs, &(xo[0])); T3e = VADD(T2z, T2A); STM2(&(xo[28]), T3e, ovs, &(xo[0])); T3f = VADD(T2v, T2y); STM2(&(xo[4]), T3f, ovs, &(xo[0])); T3g = VSUB(T2z, T2A); STM2(&(xo[36]), T3g, ovs, &(xo[0])); } { V T1r, T1C, T1M, T1K, T1F, T1N, T1y, T1J; T1r = VSUB(T1p, T1q); T1C = VSUB(T1A, T1B); T1M = VADD(T1p, T1q); T1K = VADD(T1B, T1A); { V T1D, T1E, T1u, T1x; T1D = VFNMS(LDK(KP195090322), T1s, VMUL(LDK(KP980785280), T1t)); T1E = VFMA(LDK(KP195090322), T1v, VMUL(LDK(KP980785280), T1w)); T1F = VSUB(T1D, T1E); T1N = VADD(T1D, T1E); T1u = VFMA(LDK(KP980785280), T1s, VMUL(LDK(KP195090322), T1t)); T1x = VFNMS(LDK(KP195090322), T1w, VMUL(LDK(KP980785280), T1v)); T1y = VSUB(T1u, T1x); T1J = VADD(T1u, T1x); } { V T1z, T1G, T3h, T3i; T1z = VADD(T1r, T1y); T1G = VBYI(VADD(T1C, T1F)); T3h = VSUB(T1z, T1G); STM2(&(xo[50]), T3h, ovs, &(xo[2])); STN2(&(xo[48]), T39, T3h, ovs); T3i = VADD(T1z, T1G); STM2(&(xo[14]), T3i, ovs, &(xo[2])); STN2(&(xo[12]), T38, T3i, ovs); } { V T1P, T1Q, T3j, T3k; T1P = VBYI(VADD(T1K, T1J)); T1Q = VADD(T1M, T1N); T3j = VADD(T1P, T1Q); STM2(&(xo[2]), T3j, ovs, &(xo[2])); STN2(&(xo[0]), T3a, T3j, ovs); T3k = VSUB(T1Q, T1P); STM2(&(xo[62]), T3k, ovs, &(xo[2])); STN2(&(xo[60]), T3d, T3k, ovs); } { V T1H, T1I, T3l, T3m; T1H = VSUB(T1r, T1y); T1I = VBYI(VSUB(T1F, T1C)); T3l = VSUB(T1H, T1I); STM2(&(xo[46]), T3l, ovs, &(xo[2])); STN2(&(xo[44]), T37, T3l, ovs); T3m = VADD(T1H, T1I); STM2(&(xo[18]), T3m, ovs, &(xo[2])); STN2(&(xo[16]), T3b, T3m, ovs); } { V T1L, T1O, T3n, T3o; T1L = VBYI(VSUB(T1J, T1K)); T1O = VSUB(T1M, T1N); T3n = VADD(T1L, T1O); STM2(&(xo[30]), T3n, ovs, &(xo[2])); STN2(&(xo[28]), T3e, T3n, ovs); T3o = VSUB(T1O, T1L); STM2(&(xo[34]), T3o, ovs, &(xo[2])); STN2(&(xo[32]), T3c, T3o, ovs); } } { V Tr, T1a, T1k, T1i, T1d, T1l, T10, T1h; Tr = VSUB(Tb, Tq); T1a = VSUB(T14, T19); T1k = VADD(Tb, Tq); T1i = VADD(T19, T14); { V T1b, T1c, TI, TZ; T1b = VFNMS(LDK(KP555570233), TC, VMUL(LDK(KP831469612), TH)); T1c = VFMA(LDK(KP555570233), TT, VMUL(LDK(KP831469612), TY)); T1d = VSUB(T1b, T1c); T1l = VADD(T1b, T1c); TI = VFMA(LDK(KP831469612), TC, VMUL(LDK(KP555570233), TH)); TZ = VFNMS(LDK(KP555570233), TY, VMUL(LDK(KP831469612), TT)); T10 = VSUB(TI, TZ); T1h = VADD(TI, TZ); } { V T11, T1e, T3p, T3q; T11 = VADD(Tr, T10); T1e = VBYI(VADD(T1a, T1d)); T3p = VSUB(T11, T1e); STM2(&(xo[54]), T3p, ovs, &(xo[2])); STN2(&(xo[52]), T36, T3p, ovs); T3q = VADD(T11, T1e); STM2(&(xo[10]), T3q, ovs, &(xo[2])); STN2(&(xo[8]), T34, T3q, ovs); } { V T1n, T1o, T3r, T3s; T1n = VBYI(VADD(T1i, T1h)); T1o = VADD(T1k, T1l); T3r = VADD(T1n, T1o); STM2(&(xo[6]), T3r, ovs, &(xo[2])); STN2(&(xo[4]), T3f, T3r, ovs); T3s = VSUB(T1o, T1n); STM2(&(xo[58]), T3s, ovs, &(xo[2])); STN2(&(xo[56]), T32, T3s, ovs); } { V T1f, T1g, T3t, T3u; T1f = VSUB(Tr, T10); T1g = VBYI(VSUB(T1d, T1a)); T3t = VSUB(T1f, T1g); STM2(&(xo[42]), T3t, ovs, &(xo[2])); STN2(&(xo[40]), T33, T3t, ovs); T3u = VADD(T1f, T1g); STM2(&(xo[22]), T3u, ovs, &(xo[2])); STN2(&(xo[20]), T35, T3u, ovs); } { V T1j, T1m, T3v, T3w; T1j = VBYI(VSUB(T1h, T1i)); T1m = VSUB(T1k, T1l); T3v = VADD(T1j, T1m); STM2(&(xo[26]), T3v, ovs, &(xo[2])); STN2(&(xo[24]), T31, T3v, ovs); T3w = VSUB(T1m, T1j); STM2(&(xo[38]), T3w, ovs, &(xo[2])); STN2(&(xo[36]), T3g, T3w, ovs); } } } } } } VLEAVE(); } static const kdft_desc desc = { 32, XSIMD_STRING("n2bv_32"), {170, 26, 16, 0}, &GENUS, 0, 2, 0, 0 }; void XSIMD(codelet_n2bv_32) (planner *p) { X(kdft_register) (p, n2bv_32, &desc); } #endif /* HAVE_FMA */ fftw-3.3.3/dft/simd/common/n2bv_64.c0000644000175000017500000017301612054410447013745 00000000000000/* * Copyright (c) 2003, 2007-11 Matteo Frigo * Copyright (c) 2003, 2007-11 Massachusetts Institute of Technology * * This program is free software; you can redistribute it and/or modify * it under the terms of the GNU General Public License as published by * the Free Software Foundation; either version 2 of the License, or * (at your option) any later version. * * This program is distributed in the hope that it will be useful, * but WITHOUT ANY WARRANTY; without even the implied warranty of * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the * GNU General Public License for more details. * * You should have received a copy of the GNU General Public License * along with this program; if not, write to the Free Software * Foundation, Inc., 51 Franklin Street, Fifth Floor, Boston, MA 02110-1301 USA * */ /* This file was automatically generated --- DO NOT EDIT */ /* Generated on Sun Nov 25 07:37:36 EST 2012 */ #include "codelet-dft.h" #ifdef HAVE_FMA /* Generated by: ../../../genfft/gen_notw_c.native -fma -reorder-insns -schedule-for-pipeline -simd -compact -variables 4 -pipeline-latency 8 -sign 1 -n 64 -name n2bv_64 -with-ostride 2 -include n2b.h -store-multiple 2 */ /* * This function contains 456 FP additions, 258 FP multiplications, * (or, 198 additions, 0 multiplications, 258 fused multiply/add), * 178 stack variables, 15 constants, and 160 memory accesses */ #include "n2b.h" static void n2bv_64(const R *ri, const R *ii, R *ro, R *io, stride is, stride os, INT v, INT ivs, INT ovs) { DVK(KP773010453, +0.773010453362736960810906609758469800971041293); DVK(KP820678790, +0.820678790828660330972281985331011598767386482); DVK(KP881921264, +0.881921264348355029712756863660388349508442621); DVK(KP534511135, +0.534511135950791641089685961295362908582039528); DVK(KP995184726, +0.995184726672196886244836953109479921575474869); DVK(KP098491403, +0.098491403357164253077197521291327432293052451); DVK(KP980785280, +0.980785280403230449126182236134239036973933731); DVK(KP956940335, +0.956940335732208864935797886980269969482849206); DVK(KP303346683, +0.303346683607342391675883946941299872384187453); DVK(KP831469612, +0.831469612302545237078788377617905756738560812); DVK(KP923879532, +0.923879532511286756128183189396788286822416626); DVK(KP668178637, +0.668178637919298919997757686523080761552472251); DVK(KP198912367, +0.198912367379658006911597622644676228597850501); DVK(KP414213562, +0.414213562373095048801688724209698078569671875); DVK(KP707106781, +0.707106781186547524400844362104849039284835938); { INT i; const R *xi; R *xo; xi = ii; xo = io; for (i = v; i > 0; i = i - VL, xi = xi + (VL * ivs), xo = xo + (VL * ovs), MAKE_VOLATILE_STRIDE(128, is), MAKE_VOLATILE_STRIDE(128, os)) { V T7z, T7A, T7B, T7C, T5T, T5S, T5X, T65, T8a, T8b, T8e, T8g, T5Z, T5R, T67; V T63, T5U, T64; { V T7, T26, T5k, T6A, T47, T69, T2V, T3z, T6B, T4e, T6a, T5n, T3M, T2Y, T27; V Tm, T3A, T3i, T29, TC, T5p, T4o, T6D, T6e, T3l, T3B, TR, T2a, T4x, T5q; V T6h, T6E, T39, T3H, T3I, T3c, T5N, T57, T72, T6w, T5O, T5e, T71, T6t, T2y; V T1W, T2x, T1N, T33, T34, T3E, T32, T1p, T2v, T1g, T2u, T4M, T5K, T6p, T6Z; V T6m, T6Y, T5L, T4T; { V T4g, T4l, T3g, Tu, Tx, T4h, TA, T4i; { V T1, T2, T23, T24, T4, T5, T20, T21; T1 = LD(&(xi[0]), ivs, &(xi[0])); T2 = LD(&(xi[WS(is, 32)]), ivs, &(xi[0])); T23 = LD(&(xi[WS(is, 56)]), ivs, &(xi[0])); T24 = LD(&(xi[WS(is, 24)]), ivs, &(xi[0])); T4 = LD(&(xi[WS(is, 16)]), ivs, &(xi[0])); T5 = LD(&(xi[WS(is, 48)]), ivs, &(xi[0])); T20 = LD(&(xi[WS(is, 8)]), ivs, &(xi[0])); T21 = LD(&(xi[WS(is, 40)]), ivs, &(xi[0])); { V Ta, T48, Tk, T4c, T49, Td, Tf, Tg; { V T8, T43, T3, T45, T25, T5i, T6, T44, T22, T9, Ti, Tj, Tb, Tc; T8 = LD(&(xi[WS(is, 4)]), ivs, &(xi[0])); T43 = VSUB(T1, T2); T3 = VADD(T1, T2); T45 = VSUB(T23, T24); T25 = VADD(T23, T24); T5i = VSUB(T4, T5); T6 = VADD(T4, T5); T44 = VSUB(T20, T21); T22 = VADD(T20, T21); T9 = LD(&(xi[WS(is, 36)]), ivs, &(xi[0])); Ti = LD(&(xi[WS(is, 12)]), ivs, &(xi[0])); Tj = LD(&(xi[WS(is, 44)]), ivs, &(xi[0])); Tb = LD(&(xi[WS(is, 20)]), ivs, &(xi[0])); Tc = LD(&(xi[WS(is, 52)]), ivs, &(xi[0])); { V T2T, T46, T5j, T2U; T7 = VSUB(T3, T6); T2T = VADD(T3, T6); T46 = VADD(T44, T45); T5j = VSUB(T44, T45); T26 = VSUB(T22, T25); T2U = VADD(T22, T25); Ta = VADD(T8, T9); T48 = VSUB(T8, T9); Tk = VADD(Ti, Tj); T4c = VSUB(Tj, Ti); T5k = VFMA(LDK(KP707106781), T5j, T5i); T6A = VFNMS(LDK(KP707106781), T5j, T5i); T47 = VFMA(LDK(KP707106781), T46, T43); T69 = VFNMS(LDK(KP707106781), T46, T43); T2V = VADD(T2T, T2U); T3z = VSUB(T2T, T2U); T49 = VSUB(Tb, Tc); Td = VADD(Tb, Tc); } Tf = LD(&(xi[WS(is, 60)]), ivs, &(xi[0])); Tg = LD(&(xi[WS(is, 28)]), ivs, &(xi[0])); } { V Te, T2W, T5l, T4a, Tq, Tt, Tv, Tw, T5m, T4d, Tl, T2X, Ty, Tz, To; V Tp; To = LD(&(xi[WS(is, 2)]), ivs, &(xi[0])); Tp = LD(&(xi[WS(is, 34)]), ivs, &(xi[0])); { V Th, T4b, Tr, Ts; Tr = LD(&(xi[WS(is, 18)]), ivs, &(xi[0])); Ts = LD(&(xi[WS(is, 50)]), ivs, &(xi[0])); Te = VSUB(Ta, Td); T2W = VADD(Ta, Td); T5l = VFMA(LDK(KP414213562), T48, T49); T4a = VFNMS(LDK(KP414213562), T49, T48); Th = VADD(Tf, Tg); T4b = VSUB(Tf, Tg); Tq = VADD(To, Tp); T4g = VSUB(To, Tp); T4l = VSUB(Tr, Ts); Tt = VADD(Tr, Ts); Tv = LD(&(xi[WS(is, 10)]), ivs, &(xi[0])); Tw = LD(&(xi[WS(is, 42)]), ivs, &(xi[0])); T5m = VFMA(LDK(KP414213562), T4b, T4c); T4d = VFNMS(LDK(KP414213562), T4c, T4b); Tl = VSUB(Th, Tk); T2X = VADD(Th, Tk); Ty = LD(&(xi[WS(is, 58)]), ivs, &(xi[0])); Tz = LD(&(xi[WS(is, 26)]), ivs, &(xi[0])); } T3g = VADD(Tq, Tt); Tu = VSUB(Tq, Tt); Tx = VADD(Tv, Tw); T4h = VSUB(Tv, Tw); T6B = VSUB(T4a, T4d); T4e = VADD(T4a, T4d); T6a = VADD(T5l, T5m); T5n = VSUB(T5l, T5m); T3M = VSUB(T2W, T2X); T2Y = VADD(T2W, T2X); T27 = VSUB(Te, Tl); Tm = VADD(Te, Tl); TA = VADD(Ty, Tz); T4i = VSUB(Ty, Tz); } } } { V TK, T4p, T4u, T4k, T6d, T4n, T6c, TL, TN, TO, T3j, TJ, TF, TI; { V TD, TE, TG, TH; TD = LD(&(xi[WS(is, 62)]), ivs, &(xi[0])); TE = LD(&(xi[WS(is, 30)]), ivs, &(xi[0])); TG = LD(&(xi[WS(is, 14)]), ivs, &(xi[0])); TH = LD(&(xi[WS(is, 46)]), ivs, &(xi[0])); TK = LD(&(xi[WS(is, 54)]), ivs, &(xi[0])); { V T3h, TB, T4j, T4m; T3h = VADD(Tx, TA); TB = VSUB(Tx, TA); T4j = VADD(T4h, T4i); T4m = VSUB(T4h, T4i); T4p = VSUB(TD, TE); TF = VADD(TD, TE); T4u = VSUB(TH, TG); TI = VADD(TG, TH); T3A = VSUB(T3g, T3h); T3i = VADD(T3g, T3h); T29 = VFMA(LDK(KP414213562), Tu, TB); TC = VFNMS(LDK(KP414213562), TB, Tu); T4k = VFMA(LDK(KP707106781), T4j, T4g); T6d = VFNMS(LDK(KP707106781), T4j, T4g); T4n = VFMA(LDK(KP707106781), T4m, T4l); T6c = VFNMS(LDK(KP707106781), T4m, T4l); TL = LD(&(xi[WS(is, 22)]), ivs, &(xi[0])); } TN = LD(&(xi[WS(is, 6)]), ivs, &(xi[0])); TO = LD(&(xi[WS(is, 38)]), ivs, &(xi[0])); } T3j = VADD(TF, TI); TJ = VSUB(TF, TI); { V T3a, T1E, T52, T5b, T1x, T4Z, T6r, T6u, T5a, T1U, T55, T5c, T1L, T3b; { V T4V, T1t, T58, T1w, T1Q, T1T, T1I, T4Y, T59, T1J, T53, T1H; { V T1r, TM, T4r, TP, T4q, T1s, T1u, T1v; T1r = LD(&(xi[WS(is, 63)]), ivs, &(xi[WS(is, 1)])); T5p = VFMA(LDK(KP198912367), T4k, T4n); T4o = VFNMS(LDK(KP198912367), T4n, T4k); T6D = VFMA(LDK(KP668178637), T6c, T6d); T6e = VFNMS(LDK(KP668178637), T6d, T6c); TM = VADD(TK, TL); T4r = VSUB(TK, TL); TP = VADD(TN, TO); T4q = VSUB(TN, TO); T1s = LD(&(xi[WS(is, 31)]), ivs, &(xi[WS(is, 1)])); T1u = LD(&(xi[WS(is, 15)]), ivs, &(xi[WS(is, 1)])); T1v = LD(&(xi[WS(is, 47)]), ivs, &(xi[WS(is, 1)])); { V T1R, T4X, T6g, T4t, T6f, T4w, T1S, T1O, T1P; T1O = LD(&(xi[WS(is, 55)]), ivs, &(xi[WS(is, 1)])); T1P = LD(&(xi[WS(is, 23)]), ivs, &(xi[WS(is, 1)])); T1R = LD(&(xi[WS(is, 7)]), ivs, &(xi[WS(is, 1)])); { V T3k, TQ, T4s, T4v; T3k = VADD(TP, TM); TQ = VSUB(TM, TP); T4s = VADD(T4q, T4r); T4v = VSUB(T4r, T4q); T4V = VSUB(T1r, T1s); T1t = VADD(T1r, T1s); T58 = VSUB(T1v, T1u); T1w = VADD(T1u, T1v); T4X = VSUB(T1O, T1P); T1Q = VADD(T1O, T1P); T3l = VADD(T3j, T3k); T3B = VSUB(T3j, T3k); TR = VFNMS(LDK(KP414213562), TQ, TJ); T2a = VFMA(LDK(KP414213562), TJ, TQ); T6g = VFNMS(LDK(KP707106781), T4s, T4p); T4t = VFMA(LDK(KP707106781), T4s, T4p); T6f = VFNMS(LDK(KP707106781), T4v, T4u); T4w = VFMA(LDK(KP707106781), T4v, T4u); T1S = LD(&(xi[WS(is, 39)]), ivs, &(xi[WS(is, 1)])); } { V T4W, T1A, T50, T51, T1D, T1F, T1G; { V T1y, T1z, T1B, T1C; T1y = LD(&(xi[WS(is, 3)]), ivs, &(xi[WS(is, 1)])); T1z = LD(&(xi[WS(is, 35)]), ivs, &(xi[WS(is, 1)])); T1B = LD(&(xi[WS(is, 19)]), ivs, &(xi[WS(is, 1)])); T1C = LD(&(xi[WS(is, 51)]), ivs, &(xi[WS(is, 1)])); T4x = VFNMS(LDK(KP198912367), T4w, T4t); T5q = VFMA(LDK(KP198912367), T4t, T4w); T6h = VFNMS(LDK(KP668178637), T6g, T6f); T6E = VFMA(LDK(KP668178637), T6f, T6g); T4W = VSUB(T1R, T1S); T1T = VADD(T1R, T1S); T1A = VADD(T1y, T1z); T50 = VSUB(T1y, T1z); T51 = VSUB(T1C, T1B); T1D = VADD(T1B, T1C); } T1F = LD(&(xi[WS(is, 59)]), ivs, &(xi[WS(is, 1)])); T1G = LD(&(xi[WS(is, 27)]), ivs, &(xi[WS(is, 1)])); T1I = LD(&(xi[WS(is, 11)]), ivs, &(xi[WS(is, 1)])); T4Y = VADD(T4W, T4X); T59 = VSUB(T4X, T4W); T1J = LD(&(xi[WS(is, 43)]), ivs, &(xi[WS(is, 1)])); T3a = VADD(T1A, T1D); T1E = VSUB(T1A, T1D); T52 = VFMA(LDK(KP414213562), T51, T50); T5b = VFNMS(LDK(KP414213562), T50, T51); T53 = VSUB(T1F, T1G); T1H = VADD(T1F, T1G); } } } { V T37, T54, T1K, T38; T1x = VSUB(T1t, T1w); T37 = VADD(T1t, T1w); T4Z = VFMA(LDK(KP707106781), T4Y, T4V); T6r = VFNMS(LDK(KP707106781), T4Y, T4V); T54 = VSUB(T1J, T1I); T1K = VADD(T1I, T1J); T6u = VFNMS(LDK(KP707106781), T59, T58); T5a = VFMA(LDK(KP707106781), T59, T58); T38 = VADD(T1T, T1Q); T1U = VSUB(T1Q, T1T); T55 = VFNMS(LDK(KP414213562), T54, T53); T5c = VFMA(LDK(KP414213562), T53, T54); T1L = VSUB(T1H, T1K); T3b = VADD(T1H, T1K); T39 = VADD(T37, T38); T3H = VSUB(T37, T38); } } { V T4A, TW, T4N, TZ, T1j, T1m, T4O, T4D, T13, T4F, T16, T4G, T1a, T4I, T4J; V T1d; { V TU, TV, TX, TY, T56, T6v; TU = LD(&(xi[WS(is, 1)]), ivs, &(xi[WS(is, 1)])); T56 = VADD(T52, T55); T6v = VSUB(T55, T52); { V T5d, T6s, T1V, T1M; T5d = VADD(T5b, T5c); T6s = VSUB(T5c, T5b); T1V = VSUB(T1L, T1E); T1M = VADD(T1E, T1L); T3I = VSUB(T3b, T3a); T3c = VADD(T3a, T3b); T5N = VFNMS(LDK(KP923879532), T56, T4Z); T57 = VFMA(LDK(KP923879532), T56, T4Z); T72 = VFNMS(LDK(KP923879532), T6v, T6u); T6w = VFMA(LDK(KP923879532), T6v, T6u); T5O = VFNMS(LDK(KP923879532), T5d, T5a); T5e = VFMA(LDK(KP923879532), T5d, T5a); T71 = VFMA(LDK(KP923879532), T6s, T6r); T6t = VFNMS(LDK(KP923879532), T6s, T6r); T2y = VFNMS(LDK(KP707106781), T1V, T1U); T1W = VFMA(LDK(KP707106781), T1V, T1U); T2x = VFNMS(LDK(KP707106781), T1M, T1x); T1N = VFMA(LDK(KP707106781), T1M, T1x); TV = LD(&(xi[WS(is, 33)]), ivs, &(xi[WS(is, 1)])); } TX = LD(&(xi[WS(is, 17)]), ivs, &(xi[WS(is, 1)])); TY = LD(&(xi[WS(is, 49)]), ivs, &(xi[WS(is, 1)])); { V T1h, T1i, T1k, T1l; T1h = LD(&(xi[WS(is, 9)]), ivs, &(xi[WS(is, 1)])); T1i = LD(&(xi[WS(is, 41)]), ivs, &(xi[WS(is, 1)])); T1k = LD(&(xi[WS(is, 57)]), ivs, &(xi[WS(is, 1)])); T1l = LD(&(xi[WS(is, 25)]), ivs, &(xi[WS(is, 1)])); { V T11, T4B, T4C, T12, T14, T15; T11 = LD(&(xi[WS(is, 5)]), ivs, &(xi[WS(is, 1)])); T4A = VSUB(TU, TV); TW = VADD(TU, TV); T4N = VSUB(TX, TY); TZ = VADD(TX, TY); T1j = VADD(T1h, T1i); T4B = VSUB(T1h, T1i); T1m = VADD(T1k, T1l); T4C = VSUB(T1k, T1l); T12 = LD(&(xi[WS(is, 37)]), ivs, &(xi[WS(is, 1)])); T14 = LD(&(xi[WS(is, 21)]), ivs, &(xi[WS(is, 1)])); T15 = LD(&(xi[WS(is, 53)]), ivs, &(xi[WS(is, 1)])); { V T18, T19, T1b, T1c; T18 = LD(&(xi[WS(is, 61)]), ivs, &(xi[WS(is, 1)])); T19 = LD(&(xi[WS(is, 29)]), ivs, &(xi[WS(is, 1)])); T1b = LD(&(xi[WS(is, 13)]), ivs, &(xi[WS(is, 1)])); T1c = LD(&(xi[WS(is, 45)]), ivs, &(xi[WS(is, 1)])); T4O = VSUB(T4B, T4C); T4D = VADD(T4B, T4C); T13 = VADD(T11, T12); T4F = VSUB(T11, T12); T16 = VADD(T14, T15); T4G = VSUB(T14, T15); T1a = VADD(T18, T19); T4I = VSUB(T18, T19); T4J = VSUB(T1b, T1c); T1d = VADD(T1b, T1c); } } } } { V T30, T10, T6k, T4E, T4Q, T4H, T17, T6n, T4P, T1e, T4K, T4R, T1n, T31; T30 = VADD(TW, TZ); T10 = VSUB(TW, TZ); T6k = VFNMS(LDK(KP707106781), T4D, T4A); T4E = VFMA(LDK(KP707106781), T4D, T4A); T4Q = VFMA(LDK(KP414213562), T4F, T4G); T4H = VFNMS(LDK(KP414213562), T4G, T4F); T33 = VADD(T13, T16); T17 = VSUB(T13, T16); T6n = VFNMS(LDK(KP707106781), T4O, T4N); T4P = VFMA(LDK(KP707106781), T4O, T4N); T34 = VADD(T1a, T1d); T1e = VSUB(T1a, T1d); T4K = VFMA(LDK(KP414213562), T4J, T4I); T4R = VFNMS(LDK(KP414213562), T4I, T4J); T1n = VSUB(T1j, T1m); T31 = VADD(T1j, T1m); { V T1f, T1o, T6o, T4L, T4S, T6l; T1f = VADD(T17, T1e); T1o = VSUB(T17, T1e); T6o = VSUB(T4H, T4K); T4L = VADD(T4H, T4K); T4S = VADD(T4Q, T4R); T6l = VSUB(T4Q, T4R); T3E = VSUB(T30, T31); T32 = VADD(T30, T31); T1p = VFMA(LDK(KP707106781), T1o, T1n); T2v = VFNMS(LDK(KP707106781), T1o, T1n); T1g = VFMA(LDK(KP707106781), T1f, T10); T2u = VFNMS(LDK(KP707106781), T1f, T10); T4M = VFMA(LDK(KP923879532), T4L, T4E); T5K = VFNMS(LDK(KP923879532), T4L, T4E); T6p = VFMA(LDK(KP923879532), T6o, T6n); T6Z = VFNMS(LDK(KP923879532), T6o, T6n); T6m = VFNMS(LDK(KP923879532), T6l, T6k); T6Y = VFMA(LDK(KP923879532), T6l, T6k); T5L = VFNMS(LDK(KP923879532), T4S, T4P); T4T = VFMA(LDK(KP923879532), T4S, T4P); } } } } } } { V T6b, T6F, T7n, T7o, T7p, T7q, T7r, T7s, T7t, T7u, T7v, T7w, T7x, T7y, T7f; V T6X, T70, T79, T7a, T73, T6C, T76, T77, T6i; { V T2Z, T3r, T3s, T3m, T3d, T3v; T2Z = VSUB(T2V, T2Y); T3r = VADD(T2V, T2Y); T3s = VADD(T3i, T3l); T3m = VSUB(T3i, T3l); T3d = VSUB(T39, T3c); T3v = VADD(T39, T3c); { V T3x, T3t, T3Q, T3J, T3D, T3V, T3G, T3P, T3u, T36, T3O, T3Y, T6V, T6W; { V T3N, T3C, T3F, T35; T3N = VSUB(T3A, T3B); T3C = VADD(T3A, T3B); T3F = VSUB(T33, T34); T35 = VADD(T33, T34); T3x = VADD(T3r, T3s); T3t = VSUB(T3r, T3s); T3Q = VFMA(LDK(KP414213562), T3H, T3I); T3J = VFNMS(LDK(KP414213562), T3I, T3H); T3D = VFMA(LDK(KP707106781), T3C, T3z); T3V = VFNMS(LDK(KP707106781), T3C, T3z); T3G = VFNMS(LDK(KP414213562), T3F, T3E); T3P = VFMA(LDK(KP414213562), T3E, T3F); T3u = VADD(T32, T35); T36 = VSUB(T32, T35); T3O = VFMA(LDK(KP707106781), T3N, T3M); T3Y = VFNMS(LDK(KP707106781), T3N, T3M); } T6b = VFNMS(LDK(KP923879532), T6a, T69); T6V = VFMA(LDK(KP923879532), T6a, T69); T6W = VADD(T6D, T6E); T6F = VSUB(T6D, T6E); { V T3R, T3W, T3K, T3Z; T3R = VSUB(T3P, T3Q); T3W = VADD(T3P, T3Q); T3K = VADD(T3G, T3J); T3Z = VSUB(T3G, T3J); { V T3e, T3n, T3w, T3y; T3e = VADD(T36, T3d); T3n = VSUB(T36, T3d); T3w = VSUB(T3u, T3v); T3y = VADD(T3u, T3v); { V T41, T3X, T3S, T3U; T41 = VFMA(LDK(KP923879532), T3W, T3V); T3X = VFNMS(LDK(KP923879532), T3W, T3V); T3S = VFNMS(LDK(KP923879532), T3R, T3O); T3U = VFMA(LDK(KP923879532), T3R, T3O); { V T42, T40, T3L, T3T; T42 = VFNMS(LDK(KP923879532), T3Z, T3Y); T40 = VFMA(LDK(KP923879532), T3Z, T3Y); T3L = VFNMS(LDK(KP923879532), T3K, T3D); T3T = VFMA(LDK(KP923879532), T3K, T3D); { V T3o, T3q, T3f, T3p; T3o = VFNMS(LDK(KP707106781), T3n, T3m); T3q = VFMA(LDK(KP707106781), T3n, T3m); T3f = VFNMS(LDK(KP707106781), T3e, T2Z); T3p = VFMA(LDK(KP707106781), T3e, T2Z); T7n = VSUB(T3x, T3y); STM2(&(xo[64]), T7n, ovs, &(xo[0])); T7o = VADD(T3x, T3y); STM2(&(xo[0]), T7o, ovs, &(xo[0])); T7p = VFMAI(T3w, T3t); STM2(&(xo[32]), T7p, ovs, &(xo[0])); T7q = VFNMSI(T3w, T3t); STM2(&(xo[96]), T7q, ovs, &(xo[0])); T7r = VFNMSI(T40, T3X); STM2(&(xo[88]), T7r, ovs, &(xo[0])); T7s = VFMAI(T40, T3X); STM2(&(xo[40]), T7s, ovs, &(xo[0])); T7t = VFMAI(T42, T41); STM2(&(xo[104]), T7t, ovs, &(xo[0])); T7u = VFNMSI(T42, T41); STM2(&(xo[24]), T7u, ovs, &(xo[0])); T7v = VFMAI(T3U, T3T); STM2(&(xo[8]), T7v, ovs, &(xo[0])); T7w = VFNMSI(T3U, T3T); STM2(&(xo[120]), T7w, ovs, &(xo[0])); T7x = VFMAI(T3S, T3L); STM2(&(xo[72]), T7x, ovs, &(xo[0])); T7y = VFNMSI(T3S, T3L); STM2(&(xo[56]), T7y, ovs, &(xo[0])); T7z = VFNMSI(T3q, T3p); STM2(&(xo[112]), T7z, ovs, &(xo[0])); T7A = VFMAI(T3q, T3p); STM2(&(xo[16]), T7A, ovs, &(xo[0])); T7B = VFMAI(T3o, T3f); STM2(&(xo[80]), T7B, ovs, &(xo[0])); T7C = VFNMSI(T3o, T3f); STM2(&(xo[48]), T7C, ovs, &(xo[0])); T7f = VFNMS(LDK(KP831469612), T6W, T6V); T6X = VFMA(LDK(KP831469612), T6W, T6V); } } } } } T70 = VFMA(LDK(KP303346683), T6Z, T6Y); T79 = VFNMS(LDK(KP303346683), T6Y, T6Z); T7a = VFNMS(LDK(KP303346683), T71, T72); T73 = VFMA(LDK(KP303346683), T72, T71); T6C = VFMA(LDK(KP923879532), T6B, T6A); T76 = VFNMS(LDK(KP923879532), T6B, T6A); T77 = VSUB(T6e, T6h); T6i = VADD(T6e, T6h); } } { V T2r, T2D, T2C, T2s, T5H, T5o, T5v, T5D, T7L, T7O, T7Q, T7S, T5r, T5I, T5x; V T5h, T5F, T5B; { V TT, T2f, T7E, T7F, T7I, T7K, T2n, T1Y, T28, T2b, T2l, T2p, T2j, T2k; { V T1q, T2d, T7h, T7l, T2e, T1X, T75, T7d, T7m, T7k, T7c, T7e, Tn, TS; T2r = VFNMS(LDK(KP707106781), Tm, T7); Tn = VFMA(LDK(KP707106781), Tm, T7); TS = VADD(TC, TR); T2D = VSUB(TC, TR); { V T7b, T7j, T74, T7i, T78, T7g; T1q = VFNMS(LDK(KP198912367), T1p, T1g); T2d = VFMA(LDK(KP198912367), T1g, T1p); T7g = VADD(T79, T7a); T7b = VSUB(T79, T7a); T7j = VSUB(T70, T73); T74 = VADD(T70, T73); T7i = VFNMS(LDK(KP831469612), T77, T76); T78 = VFMA(LDK(KP831469612), T77, T76); T2j = VFNMS(LDK(KP923879532), TS, Tn); TT = VFMA(LDK(KP923879532), TS, Tn); T7h = VFMA(LDK(KP956940335), T7g, T7f); T7l = VFNMS(LDK(KP956940335), T7g, T7f); T2e = VFMA(LDK(KP198912367), T1N, T1W); T1X = VFNMS(LDK(KP198912367), T1W, T1N); T75 = VFNMS(LDK(KP956940335), T74, T6X); T7d = VFMA(LDK(KP956940335), T74, T6X); T7m = VFMA(LDK(KP956940335), T7j, T7i); T7k = VFNMS(LDK(KP956940335), T7j, T7i); T7c = VFNMS(LDK(KP956940335), T7b, T78); T7e = VFMA(LDK(KP956940335), T7b, T78); } T2k = VADD(T2d, T2e); T2f = VSUB(T2d, T2e); { V T7D, T7G, T7H, T7J; T7D = VFMAI(T7k, T7h); STM2(&(xo[90]), T7D, ovs, &(xo[2])); STN2(&(xo[88]), T7r, T7D, ovs); T7E = VFNMSI(T7k, T7h); STM2(&(xo[38]), T7E, ovs, &(xo[2])); T7F = VFNMSI(T7m, T7l); STM2(&(xo[102]), T7F, ovs, &(xo[2])); T7G = VFMAI(T7m, T7l); STM2(&(xo[26]), T7G, ovs, &(xo[2])); STN2(&(xo[24]), T7u, T7G, ovs); T7H = VFMAI(T7e, T7d); STM2(&(xo[122]), T7H, ovs, &(xo[2])); STN2(&(xo[120]), T7w, T7H, ovs); T7I = VFNMSI(T7e, T7d); STM2(&(xo[6]), T7I, ovs, &(xo[2])); T7J = VFMAI(T7c, T75); STM2(&(xo[58]), T7J, ovs, &(xo[2])); STN2(&(xo[56]), T7y, T7J, ovs); T7K = VFNMSI(T7c, T75); STM2(&(xo[70]), T7K, ovs, &(xo[2])); T2n = VSUB(T1q, T1X); T1Y = VADD(T1q, T1X); } T2C = VFNMS(LDK(KP707106781), T27, T26); T28 = VFMA(LDK(KP707106781), T27, T26); T2b = VSUB(T29, T2a); T2s = VADD(T29, T2a); } T2l = VFNMS(LDK(KP980785280), T2k, T2j); T2p = VFMA(LDK(KP980785280), T2k, T2j); { V T5z, T4z, T5A, T5g; { V T4f, T4y, T1Z, T2h, T4U, T5t, T2m, T2c, T5u, T5f; T5H = VFNMS(LDK(KP923879532), T4e, T47); T4f = VFMA(LDK(KP923879532), T4e, T47); T4y = VADD(T4o, T4x); T5T = VSUB(T4o, T4x); T1Z = VFNMS(LDK(KP980785280), T1Y, TT); T2h = VFMA(LDK(KP980785280), T1Y, TT); T4U = VFNMS(LDK(KP098491403), T4T, T4M); T5t = VFMA(LDK(KP098491403), T4M, T4T); T2m = VFNMS(LDK(KP923879532), T2b, T28); T2c = VFMA(LDK(KP923879532), T2b, T28); T5u = VFMA(LDK(KP098491403), T57, T5e); T5f = VFNMS(LDK(KP098491403), T5e, T57); T5z = VFNMS(LDK(KP980785280), T4y, T4f); T4z = VFMA(LDK(KP980785280), T4y, T4f); T5S = VFNMS(LDK(KP923879532), T5n, T5k); T5o = VFMA(LDK(KP923879532), T5n, T5k); { V T2o, T2q, T2i, T2g; T2o = VFMA(LDK(KP980785280), T2n, T2m); T2q = VFNMS(LDK(KP980785280), T2n, T2m); T2i = VFMA(LDK(KP980785280), T2f, T2c); T2g = VFNMS(LDK(KP980785280), T2f, T2c); T5A = VADD(T5t, T5u); T5v = VSUB(T5t, T5u); T5D = VSUB(T4U, T5f); T5g = VADD(T4U, T5f); T7L = VFNMSI(T2o, T2l); STM2(&(xo[92]), T7L, ovs, &(xo[0])); { V T7M, T7N, T7P, T7R; T7M = VFMAI(T2o, T2l); STM2(&(xo[36]), T7M, ovs, &(xo[0])); STN2(&(xo[36]), T7M, T7E, ovs); T7N = VFMAI(T2q, T2p); STM2(&(xo[100]), T7N, ovs, &(xo[0])); STN2(&(xo[100]), T7N, T7F, ovs); T7O = VFNMSI(T2q, T2p); STM2(&(xo[28]), T7O, ovs, &(xo[0])); T7P = VFMAI(T2i, T2h); STM2(&(xo[4]), T7P, ovs, &(xo[0])); STN2(&(xo[4]), T7P, T7I, ovs); T7Q = VFNMSI(T2i, T2h); STM2(&(xo[124]), T7Q, ovs, &(xo[0])); T7R = VFMAI(T2g, T1Z); STM2(&(xo[68]), T7R, ovs, &(xo[0])); STN2(&(xo[68]), T7R, T7K, ovs); T7S = VFNMSI(T2g, T1Z); STM2(&(xo[60]), T7S, ovs, &(xo[0])); T5r = VSUB(T5p, T5q); T5I = VADD(T5p, T5q); } } } T5x = VFMA(LDK(KP995184726), T5g, T4z); T5h = VFNMS(LDK(KP995184726), T5g, T4z); T5F = VFMA(LDK(KP995184726), T5A, T5z); T5B = VFNMS(LDK(KP995184726), T5A, T5z); } } { V T6J, T6R, T6L, T6z, T6T, T6P; { V T6N, T6j, T6O, T6y; { V T6q, T6H, T5C, T5s, T6I, T6x; T6q = VFNMS(LDK(KP534511135), T6p, T6m); T6H = VFMA(LDK(KP534511135), T6m, T6p); T5C = VFNMS(LDK(KP980785280), T5r, T5o); T5s = VFMA(LDK(KP980785280), T5r, T5o); T6I = VFMA(LDK(KP534511135), T6t, T6w); T6x = VFNMS(LDK(KP534511135), T6w, T6t); T6N = VFMA(LDK(KP831469612), T6i, T6b); T6j = VFNMS(LDK(KP831469612), T6i, T6b); { V T5E, T5G, T5y, T5w; T5E = VFMA(LDK(KP995184726), T5D, T5C); T5G = VFNMS(LDK(KP995184726), T5D, T5C); T5y = VFMA(LDK(KP995184726), T5v, T5s); T5w = VFNMS(LDK(KP995184726), T5v, T5s); T6O = VADD(T6H, T6I); T6J = VSUB(T6H, T6I); T6R = VSUB(T6q, T6x); T6y = VADD(T6q, T6x); { V T7T, T7U, T7V, T7W; T7T = VFNMSI(T5E, T5B); STM2(&(xo[94]), T7T, ovs, &(xo[2])); STN2(&(xo[92]), T7L, T7T, ovs); T7U = VFMAI(T5E, T5B); STM2(&(xo[34]), T7U, ovs, &(xo[2])); STN2(&(xo[32]), T7p, T7U, ovs); T7V = VFMAI(T5G, T5F); STM2(&(xo[98]), T7V, ovs, &(xo[2])); STN2(&(xo[96]), T7q, T7V, ovs); T7W = VFNMSI(T5G, T5F); STM2(&(xo[30]), T7W, ovs, &(xo[2])); STN2(&(xo[28]), T7O, T7W, ovs); { V T7X, T7Y, T7Z, T80; T7X = VFMAI(T5y, T5x); STM2(&(xo[2]), T7X, ovs, &(xo[2])); STN2(&(xo[0]), T7o, T7X, ovs); T7Y = VFNMSI(T5y, T5x); STM2(&(xo[126]), T7Y, ovs, &(xo[2])); STN2(&(xo[124]), T7Q, T7Y, ovs); T7Z = VFMAI(T5w, T5h); STM2(&(xo[66]), T7Z, ovs, &(xo[2])); STN2(&(xo[64]), T7n, T7Z, ovs); T80 = VFNMSI(T5w, T5h); STM2(&(xo[62]), T80, ovs, &(xo[2])); STN2(&(xo[60]), T7S, T80, ovs); } } } } T6L = VFMA(LDK(KP881921264), T6y, T6j); T6z = VFNMS(LDK(KP881921264), T6y, T6j); T6T = VFMA(LDK(KP881921264), T6O, T6N); T6P = VFNMS(LDK(KP881921264), T6O, T6N); } { V T2H, T2P, T81, T84, T86, T88, T2J, T2B, T2R, T2N; { V T2L, T2t, T2M, T2A; { V T2w, T2F, T6Q, T6G, T2G, T2z; T2w = VFMA(LDK(KP668178637), T2v, T2u); T2F = VFNMS(LDK(KP668178637), T2u, T2v); T6Q = VFNMS(LDK(KP831469612), T6F, T6C); T6G = VFMA(LDK(KP831469612), T6F, T6C); T2G = VFNMS(LDK(KP668178637), T2x, T2y); T2z = VFMA(LDK(KP668178637), T2y, T2x); T2L = VFNMS(LDK(KP923879532), T2s, T2r); T2t = VFMA(LDK(KP923879532), T2s, T2r); { V T6S, T6U, T6M, T6K; T6S = VFMA(LDK(KP881921264), T6R, T6Q); T6U = VFNMS(LDK(KP881921264), T6R, T6Q); T6M = VFMA(LDK(KP881921264), T6J, T6G); T6K = VFNMS(LDK(KP881921264), T6J, T6G); T2M = VADD(T2F, T2G); T2H = VSUB(T2F, T2G); T2P = VSUB(T2w, T2z); T2A = VADD(T2w, T2z); T81 = VFNMSI(T6S, T6P); STM2(&(xo[86]), T81, ovs, &(xo[2])); { V T82, T83, T85, T87; T82 = VFMAI(T6S, T6P); STM2(&(xo[42]), T82, ovs, &(xo[2])); STN2(&(xo[40]), T7s, T82, ovs); T83 = VFMAI(T6U, T6T); STM2(&(xo[106]), T83, ovs, &(xo[2])); STN2(&(xo[104]), T7t, T83, ovs); T84 = VFNMSI(T6U, T6T); STM2(&(xo[22]), T84, ovs, &(xo[2])); T85 = VFMAI(T6M, T6L); STM2(&(xo[10]), T85, ovs, &(xo[2])); STN2(&(xo[8]), T7v, T85, ovs); T86 = VFNMSI(T6M, T6L); STM2(&(xo[118]), T86, ovs, &(xo[2])); T87 = VFMAI(T6K, T6z); STM2(&(xo[74]), T87, ovs, &(xo[2])); STN2(&(xo[72]), T7x, T87, ovs); T88 = VFNMSI(T6K, T6z); STM2(&(xo[54]), T88, ovs, &(xo[2])); } } } T2J = VFMA(LDK(KP831469612), T2A, T2t); T2B = VFNMS(LDK(KP831469612), T2A, T2t); T2R = VFNMS(LDK(KP831469612), T2M, T2L); T2N = VFMA(LDK(KP831469612), T2M, T2L); } { V T61, T5J, T62, T5Q; { V T5M, T5V, T2O, T2E, T5W, T5P; T5M = VFMA(LDK(KP820678790), T5L, T5K); T5V = VFNMS(LDK(KP820678790), T5K, T5L); T2O = VFMA(LDK(KP923879532), T2D, T2C); T2E = VFNMS(LDK(KP923879532), T2D, T2C); T5W = VFNMS(LDK(KP820678790), T5N, T5O); T5P = VFMA(LDK(KP820678790), T5O, T5N); T61 = VFNMS(LDK(KP980785280), T5I, T5H); T5J = VFMA(LDK(KP980785280), T5I, T5H); { V T2Q, T2S, T2K, T2I; T2Q = VFNMS(LDK(KP831469612), T2P, T2O); T2S = VFMA(LDK(KP831469612), T2P, T2O); T2K = VFMA(LDK(KP831469612), T2H, T2E); T2I = VFNMS(LDK(KP831469612), T2H, T2E); T62 = VADD(T5V, T5W); T5X = VSUB(T5V, T5W); T65 = VSUB(T5M, T5P); T5Q = VADD(T5M, T5P); { V T89, T8c, T8d, T8f; T89 = VFMAI(T2Q, T2N); STM2(&(xo[84]), T89, ovs, &(xo[0])); STN2(&(xo[84]), T89, T81, ovs); T8a = VFNMSI(T2Q, T2N); STM2(&(xo[44]), T8a, ovs, &(xo[0])); T8b = VFNMSI(T2S, T2R); STM2(&(xo[108]), T8b, ovs, &(xo[0])); T8c = VFMAI(T2S, T2R); STM2(&(xo[20]), T8c, ovs, &(xo[0])); STN2(&(xo[20]), T8c, T84, ovs); T8d = VFMAI(T2K, T2J); STM2(&(xo[116]), T8d, ovs, &(xo[0])); STN2(&(xo[116]), T8d, T86, ovs); T8e = VFNMSI(T2K, T2J); STM2(&(xo[12]), T8e, ovs, &(xo[0])); T8f = VFMAI(T2I, T2B); STM2(&(xo[52]), T8f, ovs, &(xo[0])); STN2(&(xo[52]), T8f, T88, ovs); T8g = VFNMSI(T2I, T2B); STM2(&(xo[76]), T8g, ovs, &(xo[0])); } } } T5Z = VFMA(LDK(KP773010453), T5Q, T5J); T5R = VFNMS(LDK(KP773010453), T5Q, T5J); T67 = VFNMS(LDK(KP773010453), T62, T61); T63 = VFMA(LDK(KP773010453), T62, T61); } } } } } } T5U = VFNMS(LDK(KP980785280), T5T, T5S); T64 = VFMA(LDK(KP980785280), T5T, T5S); { V T68, T66, T5Y, T60; T68 = VFMA(LDK(KP773010453), T65, T64); T66 = VFNMS(LDK(KP773010453), T65, T64); T5Y = VFNMS(LDK(KP773010453), T5X, T5U); T60 = VFMA(LDK(KP773010453), T5X, T5U); { V T8h, T8i, T8j, T8k; T8h = VFMAI(T66, T63); STM2(&(xo[82]), T8h, ovs, &(xo[2])); STN2(&(xo[80]), T7B, T8h, ovs); T8i = VFNMSI(T66, T63); STM2(&(xo[46]), T8i, ovs, &(xo[2])); STN2(&(xo[44]), T8a, T8i, ovs); T8j = VFNMSI(T68, T67); STM2(&(xo[110]), T8j, ovs, &(xo[2])); STN2(&(xo[108]), T8b, T8j, ovs); T8k = VFMAI(T68, T67); STM2(&(xo[18]), T8k, ovs, &(xo[2])); STN2(&(xo[16]), T7A, T8k, ovs); { V T8l, T8m, T8n, T8o; T8l = VFMAI(T60, T5Z); STM2(&(xo[114]), T8l, ovs, &(xo[2])); STN2(&(xo[112]), T7z, T8l, ovs); T8m = VFNMSI(T60, T5Z); STM2(&(xo[14]), T8m, ovs, &(xo[2])); STN2(&(xo[12]), T8e, T8m, ovs); T8n = VFMAI(T5Y, T5R); STM2(&(xo[50]), T8n, ovs, &(xo[2])); STN2(&(xo[48]), T7C, T8n, ovs); T8o = VFNMSI(T5Y, T5R); STM2(&(xo[78]), T8o, ovs, &(xo[2])); STN2(&(xo[76]), T8g, T8o, ovs); } } } } } VLEAVE(); } static const kdft_desc desc = { 64, XSIMD_STRING("n2bv_64"), {198, 0, 258, 0}, &GENUS, 0, 2, 0, 0 }; void XSIMD(codelet_n2bv_64) (planner *p) { X(kdft_register) (p, n2bv_64, &desc); } #else /* HAVE_FMA */ /* Generated by: ../../../genfft/gen_notw_c.native -simd -compact -variables 4 -pipeline-latency 8 -sign 1 -n 64 -name n2bv_64 -with-ostride 2 -include n2b.h -store-multiple 2 */ /* * This function contains 456 FP additions, 124 FP multiplications, * (or, 404 additions, 72 multiplications, 52 fused multiply/add), * 128 stack variables, 15 constants, and 160 memory accesses */ #include "n2b.h" static void n2bv_64(const R *ri, const R *ii, R *ro, R *io, stride is, stride os, INT v, INT ivs, INT ovs) { DVK(KP634393284, +0.634393284163645498215171613225493370675687095); DVK(KP773010453, +0.773010453362736960810906609758469800971041293); DVK(KP290284677, +0.290284677254462367636192375817395274691476278); DVK(KP956940335, +0.956940335732208864935797886980269969482849206); DVK(KP098017140, +0.098017140329560601994195563888641845861136673); DVK(KP995184726, +0.995184726672196886244836953109479921575474869); DVK(KP471396736, +0.471396736825997648556387625905254377657460319); DVK(KP881921264, +0.881921264348355029712756863660388349508442621); DVK(KP195090322, +0.195090322016128267848284868477022240927691618); DVK(KP980785280, +0.980785280403230449126182236134239036973933731); DVK(KP555570233, +0.555570233019602224742830813948532874374937191); DVK(KP831469612, +0.831469612302545237078788377617905756738560812); DVK(KP923879532, +0.923879532511286756128183189396788286822416626); DVK(KP382683432, +0.382683432365089771728459984030398866761344562); DVK(KP707106781, +0.707106781186547524400844362104849039284835938); { INT i; const R *xi; R *xo; xi = ii; xo = io; for (i = v; i > 0; i = i - VL, xi = xi + (VL * ivs), xo = xo + (VL * ovs), MAKE_VOLATILE_STRIDE(128, is), MAKE_VOLATILE_STRIDE(128, os)) { V T4p, T5u, Tb, T3A, T2q, T3v, T6G, T78, Tq, T3w, T6B, T79, T2l, T3B, T4w; V T5r, TI, T2g, T6u, T74, T3q, T3D, T4E, T5o, TZ, T2h, T6x, T75, T3t, T3E; V T4L, T5p, T23, T2N, T6m, T70, T6p, T71, T2c, T2O, T3i, T3Y, T5f, T5R, T5k; V T5S, T3l, T3Z, T1s, T2K, T6f, T6X, T6i, T6Y, T1B, T2L, T3b, T3V, T4Y, T5O; V T53, T5P, T3e, T3W; { V T3, T4n, T2p, T4o, T6, T5s, T9, T5t; { V T1, T2, T2n, T2o; T1 = LD(&(xi[0]), ivs, &(xi[0])); T2 = LD(&(xi[WS(is, 32)]), ivs, &(xi[0])); T3 = VSUB(T1, T2); T4n = VADD(T1, T2); T2n = LD(&(xi[WS(is, 16)]), ivs, &(xi[0])); T2o = LD(&(xi[WS(is, 48)]), ivs, &(xi[0])); T2p = VSUB(T2n, T2o); T4o = VADD(T2n, T2o); } { V T4, T5, T7, T8; T4 = LD(&(xi[WS(is, 8)]), ivs, &(xi[0])); T5 = LD(&(xi[WS(is, 40)]), ivs, &(xi[0])); T6 = VSUB(T4, T5); T5s = VADD(T4, T5); T7 = LD(&(xi[WS(is, 56)]), ivs, &(xi[0])); T8 = LD(&(xi[WS(is, 24)]), ivs, &(xi[0])); T9 = VSUB(T7, T8); T5t = VADD(T7, T8); } T4p = VSUB(T4n, T4o); T5u = VSUB(T5s, T5t); { V Ta, T2m, T6E, T6F; Ta = VMUL(LDK(KP707106781), VADD(T6, T9)); Tb = VSUB(T3, Ta); T3A = VADD(T3, Ta); T2m = VMUL(LDK(KP707106781), VSUB(T6, T9)); T2q = VSUB(T2m, T2p); T3v = VADD(T2p, T2m); T6E = VADD(T4n, T4o); T6F = VADD(T5s, T5t); T6G = VSUB(T6E, T6F); T78 = VADD(T6E, T6F); } } { V Te, T4q, To, T4t, Th, T4r, Tl, T4u; { V Tc, Td, Tm, Tn; Tc = LD(&(xi[WS(is, 4)]), ivs, &(xi[0])); Td = LD(&(xi[WS(is, 36)]), ivs, &(xi[0])); Te = VSUB(Tc, Td); T4q = VADD(Tc, Td); Tm = LD(&(xi[WS(is, 60)]), ivs, &(xi[0])); Tn = LD(&(xi[WS(is, 28)]), ivs, &(xi[0])); To = VSUB(Tm, Tn); T4t = VADD(Tm, Tn); } { V Tf, Tg, Tj, Tk; Tf = LD(&(xi[WS(is, 20)]), ivs, &(xi[0])); Tg = LD(&(xi[WS(is, 52)]), ivs, &(xi[0])); Th = VSUB(Tf, Tg); T4r = VADD(Tf, Tg); Tj = LD(&(xi[WS(is, 12)]), ivs, &(xi[0])); Tk = LD(&(xi[WS(is, 44)]), ivs, &(xi[0])); Tl = VSUB(Tj, Tk); T4u = VADD(Tj, Tk); } { V Ti, Tp, T6z, T6A; Ti = VFMA(LDK(KP382683432), Te, VMUL(LDK(KP923879532), Th)); Tp = VFNMS(LDK(KP382683432), To, VMUL(LDK(KP923879532), Tl)); Tq = VSUB(Ti, Tp); T3w = VADD(Ti, Tp); T6z = VADD(T4q, T4r); T6A = VADD(T4t, T4u); T6B = VSUB(T6z, T6A); T79 = VADD(T6z, T6A); } { V T2j, T2k, T4s, T4v; T2j = VFNMS(LDK(KP382683432), Th, VMUL(LDK(KP923879532), Te)); T2k = VFMA(LDK(KP923879532), To, VMUL(LDK(KP382683432), Tl)); T2l = VSUB(T2j, T2k); T3B = VADD(T2j, T2k); T4s = VSUB(T4q, T4r); T4v = VSUB(T4t, T4u); T4w = VMUL(LDK(KP707106781), VADD(T4s, T4v)); T5r = VMUL(LDK(KP707106781), VSUB(T4s, T4v)); } } { V TB, T4z, TF, T4y, Ty, T4C, TG, T4B; { V Tz, TA, TD, TE; Tz = LD(&(xi[WS(is, 18)]), ivs, &(xi[0])); TA = LD(&(xi[WS(is, 50)]), ivs, &(xi[0])); TB = VSUB(Tz, TA); T4z = VADD(Tz, TA); TD = LD(&(xi[WS(is, 2)]), ivs, &(xi[0])); TE = LD(&(xi[WS(is, 34)]), ivs, &(xi[0])); TF = VSUB(TD, TE); T4y = VADD(TD, TE); { V Ts, Tt, Tu, Tv, Tw, Tx; Ts = LD(&(xi[WS(is, 10)]), ivs, &(xi[0])); Tt = LD(&(xi[WS(is, 42)]), ivs, &(xi[0])); Tu = VSUB(Ts, Tt); Tv = LD(&(xi[WS(is, 58)]), ivs, &(xi[0])); Tw = LD(&(xi[WS(is, 26)]), ivs, &(xi[0])); Tx = VSUB(Tv, Tw); Ty = VMUL(LDK(KP707106781), VSUB(Tu, Tx)); T4C = VADD(Tv, Tw); TG = VMUL(LDK(KP707106781), VADD(Tu, Tx)); T4B = VADD(Ts, Tt); } } { V TC, TH, T6s, T6t; TC = VSUB(Ty, TB); TH = VSUB(TF, TG); TI = VFMA(LDK(KP831469612), TC, VMUL(LDK(KP555570233), TH)); T2g = VFNMS(LDK(KP555570233), TC, VMUL(LDK(KP831469612), TH)); T6s = VADD(T4y, T4z); T6t = VADD(T4B, T4C); T6u = VSUB(T6s, T6t); T74 = VADD(T6s, T6t); } { V T3o, T3p, T4A, T4D; T3o = VADD(TB, Ty); T3p = VADD(TF, TG); T3q = VFMA(LDK(KP980785280), T3o, VMUL(LDK(KP195090322), T3p)); T3D = VFNMS(LDK(KP195090322), T3o, VMUL(LDK(KP980785280), T3p)); T4A = VSUB(T4y, T4z); T4D = VSUB(T4B, T4C); T4E = VFMA(LDK(KP382683432), T4A, VMUL(LDK(KP923879532), T4D)); T5o = VFNMS(LDK(KP382683432), T4D, VMUL(LDK(KP923879532), T4A)); } } { V TS, T4J, TW, T4I, TP, T4G, TX, T4F; { V TQ, TR, TU, TV; TQ = LD(&(xi[WS(is, 14)]), ivs, &(xi[0])); TR = LD(&(xi[WS(is, 46)]), ivs, &(xi[0])); TS = VSUB(TQ, TR); T4J = VADD(TQ, TR); TU = LD(&(xi[WS(is, 62)]), ivs, &(xi[0])); TV = LD(&(xi[WS(is, 30)]), ivs, &(xi[0])); TW = VSUB(TU, TV); T4I = VADD(TU, TV); { V TJ, TK, TL, TM, TN, TO; TJ = LD(&(xi[WS(is, 6)]), ivs, &(xi[0])); TK = LD(&(xi[WS(is, 38)]), ivs, &(xi[0])); TL = VSUB(TJ, TK); TM = LD(&(xi[WS(is, 54)]), ivs, &(xi[0])); TN = LD(&(xi[WS(is, 22)]), ivs, &(xi[0])); TO = VSUB(TM, TN); TP = VMUL(LDK(KP707106781), VSUB(TL, TO)); T4G = VADD(TM, TN); TX = VMUL(LDK(KP707106781), VADD(TL, TO)); T4F = VADD(TJ, TK); } } { V TT, TY, T6v, T6w; TT = VSUB(TP, TS); TY = VSUB(TW, TX); TZ = VFNMS(LDK(KP555570233), TY, VMUL(LDK(KP831469612), TT)); T2h = VFMA(LDK(KP555570233), TT, VMUL(LDK(KP831469612), TY)); T6v = VADD(T4I, T4J); T6w = VADD(T4F, T4G); T6x = VSUB(T6v, T6w); T75 = VADD(T6v, T6w); } { V T3r, T3s, T4H, T4K; T3r = VADD(TS, TP); T3s = VADD(TW, TX); T3t = VFNMS(LDK(KP195090322), T3s, VMUL(LDK(KP980785280), T3r)); T3E = VFMA(LDK(KP195090322), T3r, VMUL(LDK(KP980785280), T3s)); T4H = VSUB(T4F, T4G); T4K = VSUB(T4I, T4J); T4L = VFNMS(LDK(KP382683432), T4K, VMUL(LDK(KP923879532), T4H)); T5p = VFMA(LDK(KP923879532), T4K, VMUL(LDK(KP382683432), T4H)); } } { V T21, T5h, T26, T5g, T1Y, T5d, T27, T5c, T55, T56, T1J, T57, T29, T58, T59; V T1Q, T5a, T2a; { V T1Z, T20, T24, T25; T1Z = LD(&(xi[WS(is, 15)]), ivs, &(xi[WS(is, 1)])); T20 = LD(&(xi[WS(is, 47)]), ivs, &(xi[WS(is, 1)])); T21 = VSUB(T1Z, T20); T5h = VADD(T1Z, T20); T24 = LD(&(xi[WS(is, 63)]), ivs, &(xi[WS(is, 1)])); T25 = LD(&(xi[WS(is, 31)]), ivs, &(xi[WS(is, 1)])); T26 = VSUB(T24, T25); T5g = VADD(T24, T25); } { V T1S, T1T, T1U, T1V, T1W, T1X; T1S = LD(&(xi[WS(is, 7)]), ivs, &(xi[WS(is, 1)])); T1T = LD(&(xi[WS(is, 39)]), ivs, &(xi[WS(is, 1)])); T1U = VSUB(T1S, T1T); T1V = LD(&(xi[WS(is, 55)]), ivs, &(xi[WS(is, 1)])); T1W = LD(&(xi[WS(is, 23)]), ivs, &(xi[WS(is, 1)])); T1X = VSUB(T1V, T1W); T1Y = VMUL(LDK(KP707106781), VSUB(T1U, T1X)); T5d = VADD(T1V, T1W); T27 = VMUL(LDK(KP707106781), VADD(T1U, T1X)); T5c = VADD(T1S, T1T); } { V T1F, T1I, T1M, T1P; { V T1D, T1E, T1G, T1H; T1D = LD(&(xi[WS(is, 3)]), ivs, &(xi[WS(is, 1)])); T1E = LD(&(xi[WS(is, 35)]), ivs, &(xi[WS(is, 1)])); T1F = VSUB(T1D, T1E); T55 = VADD(T1D, T1E); T1G = LD(&(xi[WS(is, 19)]), ivs, &(xi[WS(is, 1)])); T1H = LD(&(xi[WS(is, 51)]), ivs, &(xi[WS(is, 1)])); T1I = VSUB(T1G, T1H); T56 = VADD(T1G, T1H); } T1J = VFNMS(LDK(KP382683432), T1I, VMUL(LDK(KP923879532), T1F)); T57 = VSUB(T55, T56); T29 = VFMA(LDK(KP382683432), T1F, VMUL(LDK(KP923879532), T1I)); { V T1K, T1L, T1N, T1O; T1K = LD(&(xi[WS(is, 59)]), ivs, &(xi[WS(is, 1)])); T1L = LD(&(xi[WS(is, 27)]), ivs, &(xi[WS(is, 1)])); T1M = VSUB(T1K, T1L); T58 = VADD(T1K, T1L); T1N = LD(&(xi[WS(is, 11)]), ivs, &(xi[WS(is, 1)])); T1O = LD(&(xi[WS(is, 43)]), ivs, &(xi[WS(is, 1)])); T1P = VSUB(T1N, T1O); T59 = VADD(T1N, T1O); } T1Q = VFMA(LDK(KP923879532), T1M, VMUL(LDK(KP382683432), T1P)); T5a = VSUB(T58, T59); T2a = VFNMS(LDK(KP382683432), T1M, VMUL(LDK(KP923879532), T1P)); } { V T1R, T22, T6k, T6l; T1R = VSUB(T1J, T1Q); T22 = VSUB(T1Y, T21); T23 = VSUB(T1R, T22); T2N = VADD(T22, T1R); T6k = VADD(T5g, T5h); T6l = VADD(T5c, T5d); T6m = VSUB(T6k, T6l); T70 = VADD(T6k, T6l); } { V T6n, T6o, T28, T2b; T6n = VADD(T55, T56); T6o = VADD(T58, T59); T6p = VSUB(T6n, T6o); T71 = VADD(T6n, T6o); T28 = VSUB(T26, T27); T2b = VSUB(T29, T2a); T2c = VSUB(T28, T2b); T2O = VADD(T28, T2b); } { V T3g, T3h, T5b, T5e; T3g = VADD(T26, T27); T3h = VADD(T1J, T1Q); T3i = VADD(T3g, T3h); T3Y = VSUB(T3g, T3h); T5b = VMUL(LDK(KP707106781), VSUB(T57, T5a)); T5e = VSUB(T5c, T5d); T5f = VSUB(T5b, T5e); T5R = VADD(T5e, T5b); } { V T5i, T5j, T3j, T3k; T5i = VSUB(T5g, T5h); T5j = VMUL(LDK(KP707106781), VADD(T57, T5a)); T5k = VSUB(T5i, T5j); T5S = VADD(T5i, T5j); T3j = VADD(T21, T1Y); T3k = VADD(T29, T2a); T3l = VADD(T3j, T3k); T3Z = VSUB(T3k, T3j); } } { V T1q, T50, T1v, T4Z, T1n, T4W, T1w, T4V, T4O, T4P, T18, T4Q, T1y, T4R, T4S; V T1f, T4T, T1z; { V T1o, T1p, T1t, T1u; T1o = LD(&(xi[WS(is, 17)]), ivs, &(xi[WS(is, 1)])); T1p = LD(&(xi[WS(is, 49)]), ivs, &(xi[WS(is, 1)])); T1q = VSUB(T1o, T1p); T50 = VADD(T1o, T1p); T1t = LD(&(xi[WS(is, 1)]), ivs, &(xi[WS(is, 1)])); T1u = LD(&(xi[WS(is, 33)]), ivs, &(xi[WS(is, 1)])); T1v = VSUB(T1t, T1u); T4Z = VADD(T1t, T1u); } { V T1h, T1i, T1j, T1k, T1l, T1m; T1h = LD(&(xi[WS(is, 9)]), ivs, &(xi[WS(is, 1)])); T1i = LD(&(xi[WS(is, 41)]), ivs, &(xi[WS(is, 1)])); T1j = VSUB(T1h, T1i); T1k = LD(&(xi[WS(is, 57)]), ivs, &(xi[WS(is, 1)])); T1l = LD(&(xi[WS(is, 25)]), ivs, &(xi[WS(is, 1)])); T1m = VSUB(T1k, T1l); T1n = VMUL(LDK(KP707106781), VSUB(T1j, T1m)); T4W = VADD(T1k, T1l); T1w = VMUL(LDK(KP707106781), VADD(T1j, T1m)); T4V = VADD(T1h, T1i); } { V T14, T17, T1b, T1e; { V T12, T13, T15, T16; T12 = LD(&(xi[WS(is, 5)]), ivs, &(xi[WS(is, 1)])); T13 = LD(&(xi[WS(is, 37)]), ivs, &(xi[WS(is, 1)])); T14 = VSUB(T12, T13); T4O = VADD(T12, T13); T15 = LD(&(xi[WS(is, 21)]), ivs, &(xi[WS(is, 1)])); T16 = LD(&(xi[WS(is, 53)]), ivs, &(xi[WS(is, 1)])); T17 = VSUB(T15, T16); T4P = VADD(T15, T16); } T18 = VFNMS(LDK(KP382683432), T17, VMUL(LDK(KP923879532), T14)); T4Q = VSUB(T4O, T4P); T1y = VFMA(LDK(KP382683432), T14, VMUL(LDK(KP923879532), T17)); { V T19, T1a, T1c, T1d; T19 = LD(&(xi[WS(is, 61)]), ivs, &(xi[WS(is, 1)])); T1a = LD(&(xi[WS(is, 29)]), ivs, &(xi[WS(is, 1)])); T1b = VSUB(T19, T1a); T4R = VADD(T19, T1a); T1c = LD(&(xi[WS(is, 13)]), ivs, &(xi[WS(is, 1)])); T1d = LD(&(xi[WS(is, 45)]), ivs, &(xi[WS(is, 1)])); T1e = VSUB(T1c, T1d); T4S = VADD(T1c, T1d); } T1f = VFMA(LDK(KP923879532), T1b, VMUL(LDK(KP382683432), T1e)); T4T = VSUB(T4R, T4S); T1z = VFNMS(LDK(KP382683432), T1b, VMUL(LDK(KP923879532), T1e)); } { V T1g, T1r, T6d, T6e; T1g = VSUB(T18, T1f); T1r = VSUB(T1n, T1q); T1s = VSUB(T1g, T1r); T2K = VADD(T1r, T1g); T6d = VADD(T4Z, T50); T6e = VADD(T4V, T4W); T6f = VSUB(T6d, T6e); T6X = VADD(T6d, T6e); } { V T6g, T6h, T1x, T1A; T6g = VADD(T4O, T4P); T6h = VADD(T4R, T4S); T6i = VSUB(T6g, T6h); T6Y = VADD(T6g, T6h); T1x = VSUB(T1v, T1w); T1A = VSUB(T1y, T1z); T1B = VSUB(T1x, T1A); T2L = VADD(T1x, T1A); } { V T39, T3a, T4U, T4X; T39 = VADD(T1v, T1w); T3a = VADD(T18, T1f); T3b = VADD(T39, T3a); T3V = VSUB(T39, T3a); T4U = VMUL(LDK(KP707106781), VSUB(T4Q, T4T)); T4X = VSUB(T4V, T4W); T4Y = VSUB(T4U, T4X); T5O = VADD(T4X, T4U); } { V T51, T52, T3c, T3d; T51 = VSUB(T4Z, T50); T52 = VMUL(LDK(KP707106781), VADD(T4Q, T4T)); T53 = VSUB(T51, T52); T5P = VADD(T51, T52); T3c = VADD(T1q, T1n); T3d = VADD(T1y, T1z); T3e = VADD(T3c, T3d); T3W = VSUB(T3d, T3c); } } { V T7n, T7o, T7p, T7q, T7r, T7s, T7t, T7u, T7v, T7w, T7x, T7y, T7z, T7A, T7B; V T7C, T7D, T7E, T7F, T7G, T7H, T7I, T7J, T7K; { V T7h, T7l, T7k, T7m; { V T7f, T7g, T7i, T7j; T7f = VADD(T78, T79); T7g = VADD(T74, T75); T7h = VSUB(T7f, T7g); T7l = VADD(T7f, T7g); T7i = VADD(T6X, T6Y); T7j = VADD(T70, T71); T7k = VBYI(VSUB(T7i, T7j)); T7m = VADD(T7i, T7j); } T7n = VSUB(T7h, T7k); STM2(&(xo[96]), T7n, ovs, &(xo[0])); T7o = VADD(T7l, T7m); STM2(&(xo[0]), T7o, ovs, &(xo[0])); T7p = VADD(T7h, T7k); STM2(&(xo[32]), T7p, ovs, &(xo[0])); T7q = VSUB(T7l, T7m); STM2(&(xo[64]), T7q, ovs, &(xo[0])); } { V T76, T7a, T73, T7b, T6Z, T72; T76 = VSUB(T74, T75); T7a = VSUB(T78, T79); T6Z = VSUB(T6X, T6Y); T72 = VSUB(T70, T71); T73 = VMUL(LDK(KP707106781), VSUB(T6Z, T72)); T7b = VMUL(LDK(KP707106781), VADD(T6Z, T72)); { V T77, T7c, T7d, T7e; T77 = VBYI(VSUB(T73, T76)); T7c = VSUB(T7a, T7b); T7r = VADD(T77, T7c); STM2(&(xo[48]), T7r, ovs, &(xo[0])); T7s = VSUB(T7c, T77); STM2(&(xo[80]), T7s, ovs, &(xo[0])); T7d = VBYI(VADD(T76, T73)); T7e = VADD(T7a, T7b); T7t = VADD(T7d, T7e); STM2(&(xo[16]), T7t, ovs, &(xo[0])); T7u = VSUB(T7e, T7d); STM2(&(xo[112]), T7u, ovs, &(xo[0])); } } { V T6C, T6S, T6I, T6P, T6r, T6Q, T6L, T6T, T6y, T6H; T6y = VMUL(LDK(KP707106781), VSUB(T6u, T6x)); T6C = VSUB(T6y, T6B); T6S = VADD(T6B, T6y); T6H = VMUL(LDK(KP707106781), VADD(T6u, T6x)); T6I = VSUB(T6G, T6H); T6P = VADD(T6G, T6H); { V T6j, T6q, T6J, T6K; T6j = VFNMS(LDK(KP382683432), T6i, VMUL(LDK(KP923879532), T6f)); T6q = VFMA(LDK(KP923879532), T6m, VMUL(LDK(KP382683432), T6p)); T6r = VSUB(T6j, T6q); T6Q = VADD(T6j, T6q); T6J = VFMA(LDK(KP382683432), T6f, VMUL(LDK(KP923879532), T6i)); T6K = VFNMS(LDK(KP382683432), T6m, VMUL(LDK(KP923879532), T6p)); T6L = VSUB(T6J, T6K); T6T = VADD(T6J, T6K); } { V T6D, T6M, T6V, T6W; T6D = VBYI(VSUB(T6r, T6C)); T6M = VSUB(T6I, T6L); T7v = VADD(T6D, T6M); STM2(&(xo[40]), T7v, ovs, &(xo[0])); T7w = VSUB(T6M, T6D); STM2(&(xo[88]), T7w, ovs, &(xo[0])); T6V = VSUB(T6P, T6Q); T6W = VBYI(VSUB(T6T, T6S)); T7x = VSUB(T6V, T6W); STM2(&(xo[72]), T7x, ovs, &(xo[0])); T7y = VADD(T6V, T6W); STM2(&(xo[56]), T7y, ovs, &(xo[0])); } { V T6N, T6O, T6R, T6U; T6N = VBYI(VADD(T6C, T6r)); T6O = VADD(T6I, T6L); T7z = VADD(T6N, T6O); STM2(&(xo[24]), T7z, ovs, &(xo[0])); T7A = VSUB(T6O, T6N); STM2(&(xo[104]), T7A, ovs, &(xo[0])); T6R = VADD(T6P, T6Q); T6U = VBYI(VADD(T6S, T6T)); T7B = VSUB(T6R, T6U); STM2(&(xo[120]), T7B, ovs, &(xo[0])); T7C = VADD(T6R, T6U); STM2(&(xo[8]), T7C, ovs, &(xo[0])); } } { V T5N, T68, T61, T69, T5U, T65, T5Y, T66; { V T5L, T5M, T5Z, T60; T5L = VADD(T4p, T4w); T5M = VADD(T5o, T5p); T5N = VSUB(T5L, T5M); T68 = VADD(T5L, T5M); T5Z = VFNMS(LDK(KP195090322), T5O, VMUL(LDK(KP980785280), T5P)); T60 = VFMA(LDK(KP195090322), T5R, VMUL(LDK(KP980785280), T5S)); T61 = VSUB(T5Z, T60); T69 = VADD(T5Z, T60); } { V T5Q, T5T, T5W, T5X; T5Q = VFMA(LDK(KP980785280), T5O, VMUL(LDK(KP195090322), T5P)); T5T = VFNMS(LDK(KP195090322), T5S, VMUL(LDK(KP980785280), T5R)); T5U = VSUB(T5Q, T5T); T65 = VADD(T5Q, T5T); T5W = VADD(T4E, T4L); T5X = VADD(T5u, T5r); T5Y = VSUB(T5W, T5X); T66 = VADD(T5X, T5W); } { V T5V, T62, T6b, T6c; T5V = VADD(T5N, T5U); T62 = VBYI(VADD(T5Y, T61)); T7D = VSUB(T5V, T62); STM2(&(xo[100]), T7D, ovs, &(xo[0])); T7E = VADD(T5V, T62); STM2(&(xo[28]), T7E, ovs, &(xo[0])); T6b = VBYI(VADD(T66, T65)); T6c = VADD(T68, T69); T7F = VADD(T6b, T6c); STM2(&(xo[4]), T7F, ovs, &(xo[0])); T7G = VSUB(T6c, T6b); STM2(&(xo[124]), T7G, ovs, &(xo[0])); } { V T63, T64, T67, T6a; T63 = VSUB(T5N, T5U); T64 = VBYI(VSUB(T61, T5Y)); T7H = VSUB(T63, T64); STM2(&(xo[92]), T7H, ovs, &(xo[0])); T7I = VADD(T63, T64); STM2(&(xo[36]), T7I, ovs, &(xo[0])); T67 = VBYI(VSUB(T65, T66)); T6a = VSUB(T68, T69); T7J = VADD(T67, T6a); STM2(&(xo[60]), T7J, ovs, &(xo[0])); T7K = VSUB(T6a, T67); STM2(&(xo[68]), T7K, ovs, &(xo[0])); } } { V T7M, T7O, T7P, T7R; { V T11, T2C, T2v, T2D, T2e, T2z, T2s, T2A; { V Tr, T10, T2t, T2u; Tr = VSUB(Tb, Tq); T10 = VSUB(TI, TZ); T11 = VSUB(Tr, T10); T2C = VADD(Tr, T10); T2t = VFNMS(LDK(KP471396736), T1s, VMUL(LDK(KP881921264), T1B)); T2u = VFMA(LDK(KP471396736), T23, VMUL(LDK(KP881921264), T2c)); T2v = VSUB(T2t, T2u); T2D = VADD(T2t, T2u); } { V T1C, T2d, T2i, T2r; T1C = VFMA(LDK(KP881921264), T1s, VMUL(LDK(KP471396736), T1B)); T2d = VFNMS(LDK(KP471396736), T2c, VMUL(LDK(KP881921264), T23)); T2e = VSUB(T1C, T2d); T2z = VADD(T1C, T2d); T2i = VSUB(T2g, T2h); T2r = VSUB(T2l, T2q); T2s = VSUB(T2i, T2r); T2A = VADD(T2r, T2i); } { V T2f, T2w, T7L, T2F, T2G, T7N; T2f = VADD(T11, T2e); T2w = VBYI(VADD(T2s, T2v)); T7L = VSUB(T2f, T2w); STM2(&(xo[106]), T7L, ovs, &(xo[2])); STN2(&(xo[104]), T7A, T7L, ovs); T7M = VADD(T2f, T2w); STM2(&(xo[22]), T7M, ovs, &(xo[2])); T2F = VBYI(VADD(T2A, T2z)); T2G = VADD(T2C, T2D); T7N = VADD(T2F, T2G); STM2(&(xo[10]), T7N, ovs, &(xo[2])); STN2(&(xo[8]), T7C, T7N, ovs); T7O = VSUB(T2G, T2F); STM2(&(xo[118]), T7O, ovs, &(xo[2])); } { V T2x, T2y, T7Q, T2B, T2E, T7S; T2x = VSUB(T11, T2e); T2y = VBYI(VSUB(T2v, T2s)); T7P = VSUB(T2x, T2y); STM2(&(xo[86]), T7P, ovs, &(xo[2])); T7Q = VADD(T2x, T2y); STM2(&(xo[42]), T7Q, ovs, &(xo[2])); STN2(&(xo[40]), T7v, T7Q, ovs); T2B = VBYI(VSUB(T2z, T2A)); T2E = VSUB(T2C, T2D); T7R = VADD(T2B, T2E); STM2(&(xo[54]), T7R, ovs, &(xo[2])); T7S = VSUB(T2E, T2B); STM2(&(xo[74]), T7S, ovs, &(xo[2])); STN2(&(xo[72]), T7x, T7S, ovs); } } { V T3n, T3O, T3J, T3R, T3y, T3Q, T3G, T3N; { V T3f, T3m, T3H, T3I; T3f = VFNMS(LDK(KP098017140), T3e, VMUL(LDK(KP995184726), T3b)); T3m = VFMA(LDK(KP995184726), T3i, VMUL(LDK(KP098017140), T3l)); T3n = VSUB(T3f, T3m); T3O = VADD(T3f, T3m); T3H = VFMA(LDK(KP098017140), T3b, VMUL(LDK(KP995184726), T3e)); T3I = VFNMS(LDK(KP098017140), T3i, VMUL(LDK(KP995184726), T3l)); T3J = VSUB(T3H, T3I); T3R = VADD(T3H, T3I); } { V T3u, T3x, T3C, T3F; T3u = VADD(T3q, T3t); T3x = VADD(T3v, T3w); T3y = VSUB(T3u, T3x); T3Q = VADD(T3x, T3u); T3C = VADD(T3A, T3B); T3F = VADD(T3D, T3E); T3G = VSUB(T3C, T3F); T3N = VADD(T3C, T3F); } { V T3z, T3K, T7T, T7U; T3z = VBYI(VSUB(T3n, T3y)); T3K = VSUB(T3G, T3J); T7T = VADD(T3z, T3K); STM2(&(xo[34]), T7T, ovs, &(xo[2])); STN2(&(xo[32]), T7p, T7T, ovs); T7U = VSUB(T3K, T3z); STM2(&(xo[94]), T7U, ovs, &(xo[2])); STN2(&(xo[92]), T7H, T7U, ovs); } { V T3T, T3U, T7V, T7W; T3T = VSUB(T3N, T3O); T3U = VBYI(VSUB(T3R, T3Q)); T7V = VSUB(T3T, T3U); STM2(&(xo[66]), T7V, ovs, &(xo[2])); STN2(&(xo[64]), T7q, T7V, ovs); T7W = VADD(T3T, T3U); STM2(&(xo[62]), T7W, ovs, &(xo[2])); STN2(&(xo[60]), T7J, T7W, ovs); } { V T3L, T3M, T7X, T7Y; T3L = VBYI(VADD(T3y, T3n)); T3M = VADD(T3G, T3J); T7X = VADD(T3L, T3M); STM2(&(xo[30]), T7X, ovs, &(xo[2])); STN2(&(xo[28]), T7E, T7X, ovs); T7Y = VSUB(T3M, T3L); STM2(&(xo[98]), T7Y, ovs, &(xo[2])); STN2(&(xo[96]), T7n, T7Y, ovs); } { V T3P, T3S, T7Z, T80; T3P = VADD(T3N, T3O); T3S = VBYI(VADD(T3Q, T3R)); T7Z = VSUB(T3P, T3S); STM2(&(xo[126]), T7Z, ovs, &(xo[2])); STN2(&(xo[124]), T7G, T7Z, ovs); T80 = VADD(T3P, T3S); STM2(&(xo[2]), T80, ovs, &(xo[2])); STN2(&(xo[0]), T7o, T80, ovs); } } { V T81, T83, T86, T88; { V T4N, T5G, T5z, T5H, T5m, T5D, T5w, T5E; { V T4x, T4M, T5x, T5y; T4x = VSUB(T4p, T4w); T4M = VSUB(T4E, T4L); T4N = VSUB(T4x, T4M); T5G = VADD(T4x, T4M); T5x = VFNMS(LDK(KP555570233), T4Y, VMUL(LDK(KP831469612), T53)); T5y = VFMA(LDK(KP555570233), T5f, VMUL(LDK(KP831469612), T5k)); T5z = VSUB(T5x, T5y); T5H = VADD(T5x, T5y); } { V T54, T5l, T5q, T5v; T54 = VFMA(LDK(KP831469612), T4Y, VMUL(LDK(KP555570233), T53)); T5l = VFNMS(LDK(KP555570233), T5k, VMUL(LDK(KP831469612), T5f)); T5m = VSUB(T54, T5l); T5D = VADD(T54, T5l); T5q = VSUB(T5o, T5p); T5v = VSUB(T5r, T5u); T5w = VSUB(T5q, T5v); T5E = VADD(T5v, T5q); } { V T5n, T5A, T82, T5J, T5K, T84; T5n = VADD(T4N, T5m); T5A = VBYI(VADD(T5w, T5z)); T81 = VSUB(T5n, T5A); STM2(&(xo[108]), T81, ovs, &(xo[0])); T82 = VADD(T5n, T5A); STM2(&(xo[20]), T82, ovs, &(xo[0])); STN2(&(xo[20]), T82, T7M, ovs); T5J = VBYI(VADD(T5E, T5D)); T5K = VADD(T5G, T5H); T83 = VADD(T5J, T5K); STM2(&(xo[12]), T83, ovs, &(xo[0])); T84 = VSUB(T5K, T5J); STM2(&(xo[116]), T84, ovs, &(xo[0])); STN2(&(xo[116]), T84, T7O, ovs); } { V T5B, T5C, T85, T5F, T5I, T87; T5B = VSUB(T4N, T5m); T5C = VBYI(VSUB(T5z, T5w)); T85 = VSUB(T5B, T5C); STM2(&(xo[84]), T85, ovs, &(xo[0])); STN2(&(xo[84]), T85, T7P, ovs); T86 = VADD(T5B, T5C); STM2(&(xo[44]), T86, ovs, &(xo[0])); T5F = VBYI(VSUB(T5D, T5E)); T5I = VSUB(T5G, T5H); T87 = VADD(T5F, T5I); STM2(&(xo[52]), T87, ovs, &(xo[0])); STN2(&(xo[52]), T87, T7R, ovs); T88 = VSUB(T5I, T5F); STM2(&(xo[76]), T88, ovs, &(xo[0])); } } { V T2J, T34, T2X, T35, T2Q, T31, T2U, T32; { V T2H, T2I, T2V, T2W; T2H = VADD(Tb, Tq); T2I = VADD(T2g, T2h); T2J = VSUB(T2H, T2I); T34 = VADD(T2H, T2I); T2V = VFNMS(LDK(KP290284677), T2K, VMUL(LDK(KP956940335), T2L)); T2W = VFMA(LDK(KP290284677), T2N, VMUL(LDK(KP956940335), T2O)); T2X = VSUB(T2V, T2W); T35 = VADD(T2V, T2W); } { V T2M, T2P, T2S, T2T; T2M = VFMA(LDK(KP956940335), T2K, VMUL(LDK(KP290284677), T2L)); T2P = VFNMS(LDK(KP290284677), T2O, VMUL(LDK(KP956940335), T2N)); T2Q = VSUB(T2M, T2P); T31 = VADD(T2M, T2P); T2S = VADD(TI, TZ); T2T = VADD(T2q, T2l); T2U = VSUB(T2S, T2T); T32 = VADD(T2T, T2S); } { V T2R, T2Y, T89, T8a; T2R = VADD(T2J, T2Q); T2Y = VBYI(VADD(T2U, T2X)); T89 = VSUB(T2R, T2Y); STM2(&(xo[102]), T89, ovs, &(xo[2])); STN2(&(xo[100]), T7D, T89, ovs); T8a = VADD(T2R, T2Y); STM2(&(xo[26]), T8a, ovs, &(xo[2])); STN2(&(xo[24]), T7z, T8a, ovs); } { V T37, T38, T8b, T8c; T37 = VBYI(VADD(T32, T31)); T38 = VADD(T34, T35); T8b = VADD(T37, T38); STM2(&(xo[6]), T8b, ovs, &(xo[2])); STN2(&(xo[4]), T7F, T8b, ovs); T8c = VSUB(T38, T37); STM2(&(xo[122]), T8c, ovs, &(xo[2])); STN2(&(xo[120]), T7B, T8c, ovs); } { V T2Z, T30, T8d, T8e; T2Z = VSUB(T2J, T2Q); T30 = VBYI(VSUB(T2X, T2U)); T8d = VSUB(T2Z, T30); STM2(&(xo[90]), T8d, ovs, &(xo[2])); STN2(&(xo[88]), T7w, T8d, ovs); T8e = VADD(T2Z, T30); STM2(&(xo[38]), T8e, ovs, &(xo[2])); STN2(&(xo[36]), T7I, T8e, ovs); } { V T33, T36, T8f, T8g; T33 = VBYI(VSUB(T31, T32)); T36 = VSUB(T34, T35); T8f = VADD(T33, T36); STM2(&(xo[58]), T8f, ovs, &(xo[2])); STN2(&(xo[56]), T7y, T8f, ovs); T8g = VSUB(T36, T33); STM2(&(xo[70]), T8g, ovs, &(xo[2])); STN2(&(xo[68]), T7K, T8g, ovs); } } { V T41, T4g, T4b, T4j, T44, T4i, T48, T4f; { V T3X, T40, T49, T4a; T3X = VFNMS(LDK(KP634393284), T3W, VMUL(LDK(KP773010453), T3V)); T40 = VFMA(LDK(KP773010453), T3Y, VMUL(LDK(KP634393284), T3Z)); T41 = VSUB(T3X, T40); T4g = VADD(T3X, T40); T49 = VFMA(LDK(KP634393284), T3V, VMUL(LDK(KP773010453), T3W)); T4a = VFNMS(LDK(KP634393284), T3Y, VMUL(LDK(KP773010453), T3Z)); T4b = VSUB(T49, T4a); T4j = VADD(T49, T4a); } { V T42, T43, T46, T47; T42 = VSUB(T3D, T3E); T43 = VSUB(T3w, T3v); T44 = VSUB(T42, T43); T4i = VADD(T43, T42); T46 = VSUB(T3A, T3B); T47 = VSUB(T3q, T3t); T48 = VSUB(T46, T47); T4f = VADD(T46, T47); } { V T45, T4c, T8h, T8i; T45 = VBYI(VSUB(T41, T44)); T4c = VSUB(T48, T4b); T8h = VADD(T45, T4c); STM2(&(xo[46]), T8h, ovs, &(xo[2])); STN2(&(xo[44]), T86, T8h, ovs); T8i = VSUB(T4c, T45); STM2(&(xo[82]), T8i, ovs, &(xo[2])); STN2(&(xo[80]), T7s, T8i, ovs); } { V T4l, T4m, T8j, T8k; T4l = VSUB(T4f, T4g); T4m = VBYI(VSUB(T4j, T4i)); T8j = VSUB(T4l, T4m); STM2(&(xo[78]), T8j, ovs, &(xo[2])); STN2(&(xo[76]), T88, T8j, ovs); T8k = VADD(T4l, T4m); STM2(&(xo[50]), T8k, ovs, &(xo[2])); STN2(&(xo[48]), T7r, T8k, ovs); } { V T4d, T4e, T8l, T8m; T4d = VBYI(VADD(T44, T41)); T4e = VADD(T48, T4b); T8l = VADD(T4d, T4e); STM2(&(xo[18]), T8l, ovs, &(xo[2])); STN2(&(xo[16]), T7t, T8l, ovs); T8m = VSUB(T4e, T4d); STM2(&(xo[110]), T8m, ovs, &(xo[2])); STN2(&(xo[108]), T81, T8m, ovs); } { V T4h, T4k, T8n, T8o; T4h = VADD(T4f, T4g); T4k = VBYI(VADD(T4i, T4j)); T8n = VSUB(T4h, T4k); STM2(&(xo[114]), T8n, ovs, &(xo[2])); STN2(&(xo[112]), T7u, T8n, ovs); T8o = VADD(T4h, T4k); STM2(&(xo[14]), T8o, ovs, &(xo[2])); STN2(&(xo[12]), T83, T8o, ovs); } } } } } } } VLEAVE(); } static const kdft_desc desc = { 64, XSIMD_STRING("n2bv_64"), {404, 72, 52, 0}, &GENUS, 0, 2, 0, 0 }; void XSIMD(codelet_n2bv_64) (planner *p) { X(kdft_register) (p, n2bv_64, &desc); } #endif /* HAVE_FMA */ fftw-3.3.3/dft/simd/common/n2bv_20.c0000644000175000017500000004060512054410433013725 00000000000000/* * Copyright (c) 2003, 2007-11 Matteo Frigo * Copyright (c) 2003, 2007-11 Massachusetts Institute of Technology * * This program is free software; you can redistribute it and/or modify * it under the terms of the GNU General Public License as published by * the Free Software Foundation; either version 2 of the License, or * (at your option) any later version. * * This program is distributed in the hope that it will be useful, * but WITHOUT ANY WARRANTY; without even the implied warranty of * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the * GNU General Public License for more details. * * You should have received a copy of the GNU General Public License * along with this program; if not, write to the Free Software * Foundation, Inc., 51 Franklin Street, Fifth Floor, Boston, MA 02110-1301 USA * */ /* This file was automatically generated --- DO NOT EDIT */ /* Generated on Sun Nov 25 07:37:46 EST 2012 */ #include "codelet-dft.h" #ifdef HAVE_FMA /* Generated by: ../../../genfft/gen_notw_c.native -fma -reorder-insns -schedule-for-pipeline -simd -compact -variables 4 -pipeline-latency 8 -sign 1 -n 20 -name n2bv_20 -with-ostride 2 -include n2b.h -store-multiple 2 */ /* * This function contains 104 FP additions, 50 FP multiplications, * (or, 58 additions, 4 multiplications, 46 fused multiply/add), * 79 stack variables, 4 constants, and 50 memory accesses */ #include "n2b.h" static void n2bv_20(const R *ri, const R *ii, R *ro, R *io, stride is, stride os, INT v, INT ivs, INT ovs) { DVK(KP559016994, +0.559016994374947424102293417182819058860154590); DVK(KP618033988, +0.618033988749894848204586834365638117720309180); DVK(KP951056516, +0.951056516295153572116439333379382143405698634); DVK(KP250000000, +0.250000000000000000000000000000000000000000000); { INT i; const R *xi; R *xo; xi = ii; xo = io; for (i = v; i > 0; i = i - VL, xi = xi + (VL * ivs), xo = xo + (VL * ovs), MAKE_VOLATILE_STRIDE(40, is), MAKE_VOLATILE_STRIDE(40, os)) { V T1H, T1I, TS, TA, TN, TV, T1M, T1N, T1O, T1P, T1R, T1S, TK, TU, TR; V Tl; { V T3, TE, T1r, T13, Ta, TL, Tz, TG, Ts, TF, Th, TM, T1u, T1C, T1n; V T1a, T1m, T1h, T1x, T1D, Tk, Ti; { V T1, T2, TC, TD; T1 = LD(&(xi[0]), ivs, &(xi[0])); T2 = LD(&(xi[WS(is, 10)]), ivs, &(xi[0])); TC = LD(&(xi[WS(is, 5)]), ivs, &(xi[WS(is, 1)])); TD = LD(&(xi[WS(is, 15)]), ivs, &(xi[WS(is, 1)])); { V T14, T6, T1c, Tv, Tm, T1f, Ty, T17, T9, Tn, Tp, T1b, Td, Tq, Te; V Tf, T15, To; { V Tw, Tx, T7, T8, Tb, Tc; { V T4, T5, Tt, Tu, T11, T12; T4 = LD(&(xi[WS(is, 4)]), ivs, &(xi[0])); T5 = LD(&(xi[WS(is, 14)]), ivs, &(xi[0])); Tt = LD(&(xi[WS(is, 13)]), ivs, &(xi[WS(is, 1)])); Tu = LD(&(xi[WS(is, 3)]), ivs, &(xi[WS(is, 1)])); Tw = LD(&(xi[WS(is, 17)]), ivs, &(xi[WS(is, 1)])); T3 = VSUB(T1, T2); T11 = VADD(T1, T2); TE = VSUB(TC, TD); T12 = VADD(TC, TD); T14 = VADD(T4, T5); T6 = VSUB(T4, T5); T1c = VADD(Tt, Tu); Tv = VSUB(Tt, Tu); Tx = LD(&(xi[WS(is, 7)]), ivs, &(xi[WS(is, 1)])); T7 = LD(&(xi[WS(is, 16)]), ivs, &(xi[0])); T8 = LD(&(xi[WS(is, 6)]), ivs, &(xi[0])); T1r = VADD(T11, T12); T13 = VSUB(T11, T12); } Tb = LD(&(xi[WS(is, 8)]), ivs, &(xi[0])); Tc = LD(&(xi[WS(is, 18)]), ivs, &(xi[0])); Tm = LD(&(xi[WS(is, 9)]), ivs, &(xi[WS(is, 1)])); T1f = VADD(Tw, Tx); Ty = VSUB(Tw, Tx); T17 = VADD(T7, T8); T9 = VSUB(T7, T8); Tn = LD(&(xi[WS(is, 19)]), ivs, &(xi[WS(is, 1)])); Tp = LD(&(xi[WS(is, 1)]), ivs, &(xi[WS(is, 1)])); T1b = VADD(Tb, Tc); Td = VSUB(Tb, Tc); Tq = LD(&(xi[WS(is, 11)]), ivs, &(xi[WS(is, 1)])); Te = LD(&(xi[WS(is, 12)]), ivs, &(xi[0])); Tf = LD(&(xi[WS(is, 2)]), ivs, &(xi[0])); } Ta = VADD(T6, T9); TL = VSUB(T6, T9); T15 = VADD(Tm, Tn); To = VSUB(Tm, Tn); Tz = VSUB(Tv, Ty); TG = VADD(Tv, Ty); { V T1d, T1v, T18, Tr, T1e, Tg, T16, T1s; T1d = VSUB(T1b, T1c); T1v = VADD(T1b, T1c); T18 = VADD(Tp, Tq); Tr = VSUB(Tp, Tq); T1e = VADD(Te, Tf); Tg = VSUB(Te, Tf); T16 = VSUB(T14, T15); T1s = VADD(T14, T15); { V T1t, T19, T1w, T1g; T1t = VADD(T17, T18); T19 = VSUB(T17, T18); Ts = VSUB(To, Tr); TF = VADD(To, Tr); T1w = VADD(T1e, T1f); T1g = VSUB(T1e, T1f); Th = VADD(Td, Tg); TM = VSUB(Td, Tg); T1u = VADD(T1s, T1t); T1C = VSUB(T1s, T1t); T1n = VSUB(T16, T19); T1a = VADD(T16, T19); T1m = VSUB(T1d, T1g); T1h = VADD(T1d, T1g); T1x = VADD(T1v, T1w); T1D = VSUB(T1v, T1w); } } } } Tk = VSUB(Ta, Th); Ti = VADD(Ta, Th); { V TJ, T1k, T1A, TZ, Tj, T1E, T1G, TI, T10, T1j, T1z, T1i, T1y, TH; TJ = VSUB(TF, TG); TH = VADD(TF, TG); T1i = VADD(T1a, T1h); T1k = VSUB(T1a, T1h); T1y = VADD(T1u, T1x); T1A = VSUB(T1u, T1x); TZ = VADD(T3, Ti); Tj = VFNMS(LDK(KP250000000), Ti, T3); T1E = VMUL(LDK(KP951056516), VFMA(LDK(KP618033988), T1D, T1C)); T1G = VMUL(LDK(KP951056516), VFNMS(LDK(KP618033988), T1C, T1D)); TI = VFNMS(LDK(KP250000000), TH, TE); T10 = VADD(TE, TH); T1j = VFNMS(LDK(KP250000000), T1i, T13); T1H = VADD(T1r, T1y); STM2(&(xo[0]), T1H, ovs, &(xo[0])); T1z = VFNMS(LDK(KP250000000), T1y, T1r); T1I = VADD(T13, T1i); STM2(&(xo[20]), T1I, ovs, &(xo[0])); { V T1J, T1K, T1p, T1l, T1o, T1q, T1F, T1B, T1L, T1Q; TS = VFNMS(LDK(KP618033988), Ts, Tz); TA = VFMA(LDK(KP618033988), Tz, Ts); TN = VFMA(LDK(KP618033988), TM, TL); TV = VFNMS(LDK(KP618033988), TL, TM); T1J = VFMAI(T10, TZ); STM2(&(xo[10]), T1J, ovs, &(xo[2])); T1K = VFNMSI(T10, TZ); STM2(&(xo[30]), T1K, ovs, &(xo[2])); T1p = VFMA(LDK(KP559016994), T1k, T1j); T1l = VFNMS(LDK(KP559016994), T1k, T1j); T1o = VMUL(LDK(KP951056516), VFNMS(LDK(KP618033988), T1n, T1m)); T1q = VMUL(LDK(KP951056516), VFMA(LDK(KP618033988), T1m, T1n)); T1F = VFNMS(LDK(KP559016994), T1A, T1z); T1B = VFMA(LDK(KP559016994), T1A, T1z); T1L = VFNMSI(T1q, T1p); STM2(&(xo[28]), T1L, ovs, &(xo[0])); STN2(&(xo[28]), T1L, T1K, ovs); T1M = VFMAI(T1q, T1p); STM2(&(xo[12]), T1M, ovs, &(xo[0])); T1N = VFMAI(T1o, T1l); STM2(&(xo[36]), T1N, ovs, &(xo[0])); T1O = VFNMSI(T1o, T1l); STM2(&(xo[4]), T1O, ovs, &(xo[0])); T1P = VFMAI(T1E, T1B); STM2(&(xo[32]), T1P, ovs, &(xo[0])); T1Q = VFNMSI(T1E, T1B); STM2(&(xo[8]), T1Q, ovs, &(xo[0])); STN2(&(xo[8]), T1Q, T1J, ovs); T1R = VFNMSI(T1G, T1F); STM2(&(xo[24]), T1R, ovs, &(xo[0])); T1S = VFMAI(T1G, T1F); STM2(&(xo[16]), T1S, ovs, &(xo[0])); TK = VFMA(LDK(KP559016994), TJ, TI); TU = VFNMS(LDK(KP559016994), TJ, TI); TR = VFNMS(LDK(KP559016994), Tk, Tj); Tl = VFMA(LDK(KP559016994), Tk, Tj); } } } { V TY, TW, TO, TQ, TB, TP, TX, TT; TY = VFMA(LDK(KP951056516), TV, TU); TW = VFNMS(LDK(KP951056516), TV, TU); TO = VFMA(LDK(KP951056516), TN, TK); TQ = VFNMS(LDK(KP951056516), TN, TK); TB = VFNMS(LDK(KP951056516), TA, Tl); TP = VFMA(LDK(KP951056516), TA, Tl); TX = VFNMS(LDK(KP951056516), TS, TR); TT = VFMA(LDK(KP951056516), TS, TR); { V T1T, T1U, T1V, T1W; T1T = VFMAI(TQ, TP); STM2(&(xo[18]), T1T, ovs, &(xo[2])); STN2(&(xo[16]), T1S, T1T, ovs); T1U = VFNMSI(TQ, TP); STM2(&(xo[22]), T1U, ovs, &(xo[2])); STN2(&(xo[20]), T1I, T1U, ovs); T1V = VFMAI(TO, TB); STM2(&(xo[2]), T1V, ovs, &(xo[2])); STN2(&(xo[0]), T1H, T1V, ovs); T1W = VFNMSI(TO, TB); STM2(&(xo[38]), T1W, ovs, &(xo[2])); STN2(&(xo[36]), T1N, T1W, ovs); { V T1X, T1Y, T1Z, T20; T1X = VFMAI(TW, TT); STM2(&(xo[34]), T1X, ovs, &(xo[2])); STN2(&(xo[32]), T1P, T1X, ovs); T1Y = VFNMSI(TW, TT); STM2(&(xo[6]), T1Y, ovs, &(xo[2])); STN2(&(xo[4]), T1O, T1Y, ovs); T1Z = VFMAI(TY, TX); STM2(&(xo[26]), T1Z, ovs, &(xo[2])); STN2(&(xo[24]), T1R, T1Z, ovs); T20 = VFNMSI(TY, TX); STM2(&(xo[14]), T20, ovs, &(xo[2])); STN2(&(xo[12]), T1M, T20, ovs); } } } } } VLEAVE(); } static const kdft_desc desc = { 20, XSIMD_STRING("n2bv_20"), {58, 4, 46, 0}, &GENUS, 0, 2, 0, 0 }; void XSIMD(codelet_n2bv_20) (planner *p) { X(kdft_register) (p, n2bv_20, &desc); } #else /* HAVE_FMA */ /* Generated by: ../../../genfft/gen_notw_c.native -simd -compact -variables 4 -pipeline-latency 8 -sign 1 -n 20 -name n2bv_20 -with-ostride 2 -include n2b.h -store-multiple 2 */ /* * This function contains 104 FP additions, 24 FP multiplications, * (or, 92 additions, 12 multiplications, 12 fused multiply/add), * 57 stack variables, 4 constants, and 50 memory accesses */ #include "n2b.h" static void n2bv_20(const R *ri, const R *ii, R *ro, R *io, stride is, stride os, INT v, INT ivs, INT ovs) { DVK(KP587785252, +0.587785252292473129168705954639072768597652438); DVK(KP951056516, +0.951056516295153572116439333379382143405698634); DVK(KP250000000, +0.250000000000000000000000000000000000000000000); DVK(KP559016994, +0.559016994374947424102293417182819058860154590); { INT i; const R *xi; R *xo; xi = ii; xo = io; for (i = v; i > 0; i = i - VL, xi = xi + (VL * ivs), xo = xo + (VL * ovs), MAKE_VOLATILE_STRIDE(40, is), MAKE_VOLATILE_STRIDE(40, os)) { V T3, T1y, TH, T1i, Ts, TL, TM, Tz, T13, T16, T1j, T1u, T1v, T1w, T1r; V T1s, T1t, T1a, T1d, T1k, Ti, Tk, TE, TI; { V T1, T2, T1g, TF, TG, T1h; T1 = LD(&(xi[0]), ivs, &(xi[0])); T2 = LD(&(xi[WS(is, 10)]), ivs, &(xi[0])); T1g = VADD(T1, T2); TF = LD(&(xi[WS(is, 5)]), ivs, &(xi[WS(is, 1)])); TG = LD(&(xi[WS(is, 15)]), ivs, &(xi[WS(is, 1)])); T1h = VADD(TF, TG); T3 = VSUB(T1, T2); T1y = VADD(T1g, T1h); TH = VSUB(TF, TG); T1i = VSUB(T1g, T1h); } { V T6, T11, Tv, T19, Ty, T1c, T9, T14, Td, T18, To, T12, Tr, T15, Tg; V T1b; { V T4, T5, Tt, Tu; T4 = LD(&(xi[WS(is, 4)]), ivs, &(xi[0])); T5 = LD(&(xi[WS(is, 14)]), ivs, &(xi[0])); T6 = VSUB(T4, T5); T11 = VADD(T4, T5); Tt = LD(&(xi[WS(is, 13)]), ivs, &(xi[WS(is, 1)])); Tu = LD(&(xi[WS(is, 3)]), ivs, &(xi[WS(is, 1)])); Tv = VSUB(Tt, Tu); T19 = VADD(Tt, Tu); } { V Tw, Tx, T7, T8; Tw = LD(&(xi[WS(is, 17)]), ivs, &(xi[WS(is, 1)])); Tx = LD(&(xi[WS(is, 7)]), ivs, &(xi[WS(is, 1)])); Ty = VSUB(Tw, Tx); T1c = VADD(Tw, Tx); T7 = LD(&(xi[WS(is, 16)]), ivs, &(xi[0])); T8 = LD(&(xi[WS(is, 6)]), ivs, &(xi[0])); T9 = VSUB(T7, T8); T14 = VADD(T7, T8); } { V Tb, Tc, Tm, Tn; Tb = LD(&(xi[WS(is, 8)]), ivs, &(xi[0])); Tc = LD(&(xi[WS(is, 18)]), ivs, &(xi[0])); Td = VSUB(Tb, Tc); T18 = VADD(Tb, Tc); Tm = LD(&(xi[WS(is, 9)]), ivs, &(xi[WS(is, 1)])); Tn = LD(&(xi[WS(is, 19)]), ivs, &(xi[WS(is, 1)])); To = VSUB(Tm, Tn); T12 = VADD(Tm, Tn); } { V Tp, Tq, Te, Tf; Tp = LD(&(xi[WS(is, 1)]), ivs, &(xi[WS(is, 1)])); Tq = LD(&(xi[WS(is, 11)]), ivs, &(xi[WS(is, 1)])); Tr = VSUB(Tp, Tq); T15 = VADD(Tp, Tq); Te = LD(&(xi[WS(is, 12)]), ivs, &(xi[0])); Tf = LD(&(xi[WS(is, 2)]), ivs, &(xi[0])); Tg = VSUB(Te, Tf); T1b = VADD(Te, Tf); } Ts = VSUB(To, Tr); TL = VSUB(T6, T9); TM = VSUB(Td, Tg); Tz = VSUB(Tv, Ty); T13 = VSUB(T11, T12); T16 = VSUB(T14, T15); T1j = VADD(T13, T16); T1u = VADD(T18, T19); T1v = VADD(T1b, T1c); T1w = VADD(T1u, T1v); T1r = VADD(T11, T12); T1s = VADD(T14, T15); T1t = VADD(T1r, T1s); T1a = VSUB(T18, T19); T1d = VSUB(T1b, T1c); T1k = VADD(T1a, T1d); { V Ta, Th, TC, TD; Ta = VADD(T6, T9); Th = VADD(Td, Tg); Ti = VADD(Ta, Th); Tk = VMUL(LDK(KP559016994), VSUB(Ta, Th)); TC = VADD(To, Tr); TD = VADD(Tv, Ty); TE = VMUL(LDK(KP559016994), VSUB(TC, TD)); TI = VADD(TC, TD); } } { V T1H, T1J, T1K, T1L, T1N, T1I, TZ, T10; TZ = VADD(T3, Ti); T10 = VBYI(VADD(TH, TI)); T1H = VSUB(TZ, T10); STM2(&(xo[30]), T1H, ovs, &(xo[2])); T1I = VADD(TZ, T10); STM2(&(xo[10]), T1I, ovs, &(xo[2])); { V T1x, T1z, T1A, T1E, T1G, T1C, T1D, T1F, T1B, T1M; T1x = VMUL(LDK(KP559016994), VSUB(T1t, T1w)); T1z = VADD(T1t, T1w); T1A = VFNMS(LDK(KP250000000), T1z, T1y); T1C = VSUB(T1r, T1s); T1D = VSUB(T1u, T1v); T1E = VBYI(VFMA(LDK(KP951056516), T1C, VMUL(LDK(KP587785252), T1D))); T1G = VBYI(VFNMS(LDK(KP951056516), T1D, VMUL(LDK(KP587785252), T1C))); T1J = VADD(T1y, T1z); STM2(&(xo[0]), T1J, ovs, &(xo[0])); T1F = VSUB(T1A, T1x); T1K = VSUB(T1F, T1G); STM2(&(xo[16]), T1K, ovs, &(xo[0])); T1L = VADD(T1G, T1F); STM2(&(xo[24]), T1L, ovs, &(xo[0])); T1B = VADD(T1x, T1A); T1M = VSUB(T1B, T1E); STM2(&(xo[8]), T1M, ovs, &(xo[0])); STN2(&(xo[8]), T1M, T1I, ovs); T1N = VADD(T1E, T1B); STM2(&(xo[32]), T1N, ovs, &(xo[0])); } { V T1O, T1P, T1R, T1S; { V T1n, T1l, T1m, T1f, T1p, T17, T1e, T1q, T1Q, T1o; T1n = VMUL(LDK(KP559016994), VSUB(T1j, T1k)); T1l = VADD(T1j, T1k); T1m = VFNMS(LDK(KP250000000), T1l, T1i); T17 = VSUB(T13, T16); T1e = VSUB(T1a, T1d); T1f = VBYI(VFNMS(LDK(KP951056516), T1e, VMUL(LDK(KP587785252), T17))); T1p = VBYI(VFMA(LDK(KP951056516), T17, VMUL(LDK(KP587785252), T1e))); T1O = VADD(T1i, T1l); STM2(&(xo[20]), T1O, ovs, &(xo[0])); T1q = VADD(T1n, T1m); T1P = VADD(T1p, T1q); STM2(&(xo[12]), T1P, ovs, &(xo[0])); T1Q = VSUB(T1q, T1p); STM2(&(xo[28]), T1Q, ovs, &(xo[0])); STN2(&(xo[28]), T1Q, T1H, ovs); T1o = VSUB(T1m, T1n); T1R = VADD(T1f, T1o); STM2(&(xo[4]), T1R, ovs, &(xo[0])); T1S = VSUB(T1o, T1f); STM2(&(xo[36]), T1S, ovs, &(xo[0])); } { V TA, TN, TU, TS, TK, TV, Tl, TR, TJ, Tj; TA = VFNMS(LDK(KP951056516), Tz, VMUL(LDK(KP587785252), Ts)); TN = VFNMS(LDK(KP951056516), TM, VMUL(LDK(KP587785252), TL)); TU = VFMA(LDK(KP951056516), TL, VMUL(LDK(KP587785252), TM)); TS = VFMA(LDK(KP951056516), Ts, VMUL(LDK(KP587785252), Tz)); TJ = VFNMS(LDK(KP250000000), TI, TH); TK = VSUB(TE, TJ); TV = VADD(TE, TJ); Tj = VFNMS(LDK(KP250000000), Ti, T3); Tl = VSUB(Tj, Tk); TR = VADD(Tk, Tj); { V TB, TO, T1T, T1U; TB = VSUB(Tl, TA); TO = VBYI(VSUB(TK, TN)); T1T = VSUB(TB, TO); STM2(&(xo[34]), T1T, ovs, &(xo[2])); STN2(&(xo[32]), T1N, T1T, ovs); T1U = VADD(TB, TO); STM2(&(xo[6]), T1U, ovs, &(xo[2])); STN2(&(xo[4]), T1R, T1U, ovs); } { V TX, TY, T1V, T1W; TX = VADD(TR, TS); TY = VBYI(VSUB(TV, TU)); T1V = VSUB(TX, TY); STM2(&(xo[22]), T1V, ovs, &(xo[2])); STN2(&(xo[20]), T1O, T1V, ovs); T1W = VADD(TX, TY); STM2(&(xo[18]), T1W, ovs, &(xo[2])); STN2(&(xo[16]), T1K, T1W, ovs); } { V TP, TQ, T1X, T1Y; TP = VADD(Tl, TA); TQ = VBYI(VADD(TN, TK)); T1X = VSUB(TP, TQ); STM2(&(xo[26]), T1X, ovs, &(xo[2])); STN2(&(xo[24]), T1L, T1X, ovs); T1Y = VADD(TP, TQ); STM2(&(xo[14]), T1Y, ovs, &(xo[2])); STN2(&(xo[12]), T1P, T1Y, ovs); } { V TT, TW, T1Z, T20; TT = VSUB(TR, TS); TW = VBYI(VADD(TU, TV)); T1Z = VSUB(TT, TW); STM2(&(xo[38]), T1Z, ovs, &(xo[2])); STN2(&(xo[36]), T1S, T1Z, ovs); T20 = VADD(TT, TW); STM2(&(xo[2]), T20, ovs, &(xo[2])); STN2(&(xo[0]), T1J, T20, ovs); } } } } } } VLEAVE(); } static const kdft_desc desc = { 20, XSIMD_STRING("n2bv_20"), {92, 12, 12, 0}, &GENUS, 0, 2, 0, 0 }; void XSIMD(codelet_n2bv_20) (planner *p) { X(kdft_register) (p, n2bv_20, &desc); } #endif /* HAVE_FMA */ fftw-3.3.3/dft/simd/common/n2sv_4.c0000644000175000017500000001323312054410433013665 00000000000000/* * Copyright (c) 2003, 2007-11 Matteo Frigo * Copyright (c) 2003, 2007-11 Massachusetts Institute of Technology * * This program is free software; you can redistribute it and/or modify * it under the terms of the GNU General Public License as published by * the Free Software Foundation; either version 2 of the License, or * (at your option) any later version. * * This program is distributed in the hope that it will be useful, * but WITHOUT ANY WARRANTY; without even the implied warranty of * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the * GNU General Public License for more details. * * You should have received a copy of the GNU General Public License * along with this program; if not, write to the Free Software * Foundation, Inc., 51 Franklin Street, Fifth Floor, Boston, MA 02110-1301 USA * */ /* This file was automatically generated --- DO NOT EDIT */ /* Generated on Sun Nov 25 07:37:47 EST 2012 */ #include "codelet-dft.h" #ifdef HAVE_FMA /* Generated by: ../../../genfft/gen_notw.native -fma -reorder-insns -schedule-for-pipeline -simd -compact -variables 4 -pipeline-latency 8 -n 4 -name n2sv_4 -with-ostride 1 -include n2s.h -store-multiple 4 */ /* * This function contains 16 FP additions, 0 FP multiplications, * (or, 16 additions, 0 multiplications, 0 fused multiply/add), * 25 stack variables, 0 constants, and 18 memory accesses */ #include "n2s.h" static void n2sv_4(const R *ri, const R *ii, R *ro, R *io, stride is, stride os, INT v, INT ivs, INT ovs) { { INT i; for (i = v; i > 0; i = i - (2 * VL), ri = ri + ((2 * VL) * ivs), ii = ii + ((2 * VL) * ivs), ro = ro + ((2 * VL) * ovs), io = io + ((2 * VL) * ovs), MAKE_VOLATILE_STRIDE(16, is), MAKE_VOLATILE_STRIDE(16, os)) { V T1, T2, T7, T8, T4, T5, Tc, Td; T1 = LD(&(ri[0]), ivs, &(ri[0])); T2 = LD(&(ri[WS(is, 2)]), ivs, &(ri[0])); T7 = LD(&(ii[0]), ivs, &(ii[0])); T8 = LD(&(ii[WS(is, 2)]), ivs, &(ii[0])); T4 = LD(&(ri[WS(is, 1)]), ivs, &(ri[WS(is, 1)])); T5 = LD(&(ri[WS(is, 3)]), ivs, &(ri[WS(is, 1)])); Tc = LD(&(ii[WS(is, 1)]), ivs, &(ii[WS(is, 1)])); Td = LD(&(ii[WS(is, 3)]), ivs, &(ii[WS(is, 1)])); { V T3, Tb, T9, Tf, T6, Ta, Te, Tg; T3 = VADD(T1, T2); Tb = VSUB(T1, T2); T9 = VSUB(T7, T8); Tf = VADD(T7, T8); T6 = VADD(T4, T5); Ta = VSUB(T4, T5); Te = VSUB(Tc, Td); Tg = VADD(Tc, Td); { V Th, Ti, Tj, Tk; Th = VADD(Ta, T9); STM4(&(io[3]), Th, ovs, &(io[1])); Ti = VSUB(T9, Ta); STM4(&(io[1]), Ti, ovs, &(io[1])); Tj = VADD(T3, T6); STM4(&(ro[0]), Tj, ovs, &(ro[0])); Tk = VSUB(T3, T6); STM4(&(ro[2]), Tk, ovs, &(ro[0])); { V Tl, Tm, Tn, To; Tl = VADD(Tf, Tg); STM4(&(io[0]), Tl, ovs, &(io[0])); Tm = VSUB(Tf, Tg); STM4(&(io[2]), Tm, ovs, &(io[0])); STN4(&(io[0]), Tl, Ti, Tm, Th, ovs); Tn = VSUB(Tb, Te); STM4(&(ro[3]), Tn, ovs, &(ro[1])); To = VADD(Tb, Te); STM4(&(ro[1]), To, ovs, &(ro[1])); STN4(&(ro[0]), Tj, To, Tk, Tn, ovs); } } } } } VLEAVE(); } static const kdft_desc desc = { 4, XSIMD_STRING("n2sv_4"), {16, 0, 0, 0}, &GENUS, 0, 1, 0, 0 }; void XSIMD(codelet_n2sv_4) (planner *p) { X(kdft_register) (p, n2sv_4, &desc); } #else /* HAVE_FMA */ /* Generated by: ../../../genfft/gen_notw.native -simd -compact -variables 4 -pipeline-latency 8 -n 4 -name n2sv_4 -with-ostride 1 -include n2s.h -store-multiple 4 */ /* * This function contains 16 FP additions, 0 FP multiplications, * (or, 16 additions, 0 multiplications, 0 fused multiply/add), * 17 stack variables, 0 constants, and 18 memory accesses */ #include "n2s.h" static void n2sv_4(const R *ri, const R *ii, R *ro, R *io, stride is, stride os, INT v, INT ivs, INT ovs) { { INT i; for (i = v; i > 0; i = i - (2 * VL), ri = ri + ((2 * VL) * ivs), ii = ii + ((2 * VL) * ivs), ro = ro + ((2 * VL) * ovs), io = io + ((2 * VL) * ovs), MAKE_VOLATILE_STRIDE(16, is), MAKE_VOLATILE_STRIDE(16, os)) { V T3, Tb, T9, Tf, T6, Ta, Te, Tg; { V T1, T2, T7, T8; T1 = LD(&(ri[0]), ivs, &(ri[0])); T2 = LD(&(ri[WS(is, 2)]), ivs, &(ri[0])); T3 = VADD(T1, T2); Tb = VSUB(T1, T2); T7 = LD(&(ii[0]), ivs, &(ii[0])); T8 = LD(&(ii[WS(is, 2)]), ivs, &(ii[0])); T9 = VSUB(T7, T8); Tf = VADD(T7, T8); } { V T4, T5, Tc, Td; T4 = LD(&(ri[WS(is, 1)]), ivs, &(ri[WS(is, 1)])); T5 = LD(&(ri[WS(is, 3)]), ivs, &(ri[WS(is, 1)])); T6 = VADD(T4, T5); Ta = VSUB(T4, T5); Tc = LD(&(ii[WS(is, 1)]), ivs, &(ii[WS(is, 1)])); Td = LD(&(ii[WS(is, 3)]), ivs, &(ii[WS(is, 1)])); Te = VSUB(Tc, Td); Tg = VADD(Tc, Td); } { V Th, Ti, Tj, Tk; Th = VSUB(T3, T6); STM4(&(ro[2]), Th, ovs, &(ro[0])); Ti = VSUB(Tf, Tg); STM4(&(io[2]), Ti, ovs, &(io[0])); Tj = VADD(T3, T6); STM4(&(ro[0]), Tj, ovs, &(ro[0])); Tk = VADD(Tf, Tg); STM4(&(io[0]), Tk, ovs, &(io[0])); { V Tl, Tm, Tn, To; Tl = VSUB(T9, Ta); STM4(&(io[1]), Tl, ovs, &(io[1])); Tm = VADD(Tb, Te); STM4(&(ro[1]), Tm, ovs, &(ro[1])); Tn = VADD(Ta, T9); STM4(&(io[3]), Tn, ovs, &(io[1])); STN4(&(io[0]), Tk, Tl, Ti, Tn, ovs); To = VSUB(Tb, Te); STM4(&(ro[3]), To, ovs, &(ro[1])); STN4(&(ro[0]), Tj, Tm, Th, To, ovs); } } } } VLEAVE(); } static const kdft_desc desc = { 4, XSIMD_STRING("n2sv_4"), {16, 0, 0, 0}, &GENUS, 0, 1, 0, 0 }; void XSIMD(codelet_n2sv_4) (planner *p) { X(kdft_register) (p, n2sv_4, &desc); } #endif /* HAVE_FMA */ fftw-3.3.3/dft/simd/common/n2sv_8.c0000644000175000017500000002374012054410434013676 00000000000000/* * Copyright (c) 2003, 2007-11 Matteo Frigo * Copyright (c) 2003, 2007-11 Massachusetts Institute of Technology * * This program is free software; you can redistribute it and/or modify * it under the terms of the GNU General Public License as published by * the Free Software Foundation; either version 2 of the License, or * (at your option) any later version. * * This program is distributed in the hope that it will be useful, * but WITHOUT ANY WARRANTY; without even the implied warranty of * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the * GNU General Public License for more details. * * You should have received a copy of the GNU General Public License * along with this program; if not, write to the Free Software * Foundation, Inc., 51 Franklin Street, Fifth Floor, Boston, MA 02110-1301 USA * */ /* This file was automatically generated --- DO NOT EDIT */ /* Generated on Sun Nov 25 07:37:47 EST 2012 */ #include "codelet-dft.h" #ifdef HAVE_FMA /* Generated by: ../../../genfft/gen_notw.native -fma -reorder-insns -schedule-for-pipeline -simd -compact -variables 4 -pipeline-latency 8 -n 8 -name n2sv_8 -with-ostride 1 -include n2s.h -store-multiple 4 */ /* * This function contains 52 FP additions, 8 FP multiplications, * (or, 44 additions, 0 multiplications, 8 fused multiply/add), * 58 stack variables, 1 constants, and 36 memory accesses */ #include "n2s.h" static void n2sv_8(const R *ri, const R *ii, R *ro, R *io, stride is, stride os, INT v, INT ivs, INT ovs) { DVK(KP707106781, +0.707106781186547524400844362104849039284835938); { INT i; for (i = v; i > 0; i = i - (2 * VL), ri = ri + ((2 * VL) * ivs), ii = ii + ((2 * VL) * ivs), ro = ro + ((2 * VL) * ovs), io = io + ((2 * VL) * ovs), MAKE_VOLATILE_STRIDE(32, is), MAKE_VOLATILE_STRIDE(32, os)) { V TF, TJ, TD, TR, TS, TT, TU, TV, TW, TE, TX, TY, TK, TI, TZ; V T10, T11, T12; { V Tb, Tn, T3, TC, Ti, TB, T6, To, Tl, Tc, Tw, Tx, T8, T9, Tr; V Ts; { V T1, T2, Tg, Th, T4, T5, Tj, Tk; T1 = LD(&(ri[0]), ivs, &(ri[0])); T2 = LD(&(ri[WS(is, 4)]), ivs, &(ri[0])); Tg = LD(&(ii[0]), ivs, &(ii[0])); Th = LD(&(ii[WS(is, 4)]), ivs, &(ii[0])); T4 = LD(&(ri[WS(is, 2)]), ivs, &(ri[0])); T5 = LD(&(ri[WS(is, 6)]), ivs, &(ri[0])); Tj = LD(&(ii[WS(is, 2)]), ivs, &(ii[0])); Tk = LD(&(ii[WS(is, 6)]), ivs, &(ii[0])); Tb = LD(&(ri[WS(is, 7)]), ivs, &(ri[WS(is, 1)])); Tn = VSUB(T1, T2); T3 = VADD(T1, T2); TC = VSUB(Tg, Th); Ti = VADD(Tg, Th); TB = VSUB(T4, T5); T6 = VADD(T4, T5); To = VSUB(Tj, Tk); Tl = VADD(Tj, Tk); Tc = LD(&(ri[WS(is, 3)]), ivs, &(ri[WS(is, 1)])); Tw = LD(&(ii[WS(is, 7)]), ivs, &(ii[WS(is, 1)])); Tx = LD(&(ii[WS(is, 3)]), ivs, &(ii[WS(is, 1)])); T8 = LD(&(ri[WS(is, 1)]), ivs, &(ri[WS(is, 1)])); T9 = LD(&(ri[WS(is, 5)]), ivs, &(ri[WS(is, 1)])); Tr = LD(&(ii[WS(is, 1)]), ivs, &(ii[WS(is, 1)])); Ts = LD(&(ii[WS(is, 5)]), ivs, &(ii[WS(is, 1)])); } { V TL, T7, TP, Tm, Tz, TH, Te, Tf, TO, TQ, TG, Tu, Tp, TA; { V Td, Tv, TN, Ty, Ta, Tq, TM, Tt; TL = VSUB(T3, T6); T7 = VADD(T3, T6); Td = VADD(Tb, Tc); Tv = VSUB(Tb, Tc); TN = VADD(Tw, Tx); Ty = VSUB(Tw, Tx); Ta = VADD(T8, T9); Tq = VSUB(T8, T9); TM = VADD(Tr, Ts); Tt = VSUB(Tr, Ts); TP = VADD(Ti, Tl); Tm = VSUB(Ti, Tl); Tz = VSUB(Tv, Ty); TH = VADD(Tv, Ty); Te = VADD(Ta, Td); Tf = VSUB(Td, Ta); TO = VSUB(TM, TN); TQ = VADD(TM, TN); TG = VSUB(Tt, Tq); Tu = VADD(Tq, Tt); } TF = VSUB(Tn, To); Tp = VADD(Tn, To); TJ = VSUB(TC, TB); TD = VADD(TB, TC); TR = VSUB(Tm, Tf); STM4(&(io[6]), TR, ovs, &(io[0])); TS = VADD(Tf, Tm); STM4(&(io[2]), TS, ovs, &(io[0])); TT = VADD(T7, Te); STM4(&(ro[0]), TT, ovs, &(ro[0])); TU = VSUB(T7, Te); STM4(&(ro[4]), TU, ovs, &(ro[0])); TV = VADD(TP, TQ); STM4(&(io[0]), TV, ovs, &(io[0])); TW = VSUB(TP, TQ); STM4(&(io[4]), TW, ovs, &(io[0])); TE = VSUB(Tz, Tu); TA = VADD(Tu, Tz); TX = VADD(TL, TO); STM4(&(ro[2]), TX, ovs, &(ro[0])); TY = VSUB(TL, TO); STM4(&(ro[6]), TY, ovs, &(ro[0])); TK = VADD(TG, TH); TI = VSUB(TG, TH); TZ = VFMA(LDK(KP707106781), TA, Tp); STM4(&(ro[1]), TZ, ovs, &(ro[1])); T10 = VFNMS(LDK(KP707106781), TA, Tp); STM4(&(ro[5]), T10, ovs, &(ro[1])); } } T11 = VFMA(LDK(KP707106781), TK, TJ); STM4(&(io[1]), T11, ovs, &(io[1])); T12 = VFNMS(LDK(KP707106781), TK, TJ); STM4(&(io[5]), T12, ovs, &(io[1])); { V T13, T14, T15, T16; T13 = VFMA(LDK(KP707106781), TE, TD); STM4(&(io[3]), T13, ovs, &(io[1])); STN4(&(io[0]), TV, T11, TS, T13, ovs); T14 = VFNMS(LDK(KP707106781), TE, TD); STM4(&(io[7]), T14, ovs, &(io[1])); STN4(&(io[4]), TW, T12, TR, T14, ovs); T15 = VFMA(LDK(KP707106781), TI, TF); STM4(&(ro[3]), T15, ovs, &(ro[1])); STN4(&(ro[0]), TT, TZ, TX, T15, ovs); T16 = VFNMS(LDK(KP707106781), TI, TF); STM4(&(ro[7]), T16, ovs, &(ro[1])); STN4(&(ro[4]), TU, T10, TY, T16, ovs); } } } VLEAVE(); } static const kdft_desc desc = { 8, XSIMD_STRING("n2sv_8"), {44, 0, 8, 0}, &GENUS, 0, 1, 0, 0 }; void XSIMD(codelet_n2sv_8) (planner *p) { X(kdft_register) (p, n2sv_8, &desc); } #else /* HAVE_FMA */ /* Generated by: ../../../genfft/gen_notw.native -simd -compact -variables 4 -pipeline-latency 8 -n 8 -name n2sv_8 -with-ostride 1 -include n2s.h -store-multiple 4 */ /* * This function contains 52 FP additions, 4 FP multiplications, * (or, 52 additions, 4 multiplications, 0 fused multiply/add), * 34 stack variables, 1 constants, and 36 memory accesses */ #include "n2s.h" static void n2sv_8(const R *ri, const R *ii, R *ro, R *io, stride is, stride os, INT v, INT ivs, INT ovs) { DVK(KP707106781, +0.707106781186547524400844362104849039284835938); { INT i; for (i = v; i > 0; i = i - (2 * VL), ri = ri + ((2 * VL) * ivs), ii = ii + ((2 * VL) * ivs), ro = ro + ((2 * VL) * ovs), io = io + ((2 * VL) * ovs), MAKE_VOLATILE_STRIDE(32, is), MAKE_VOLATILE_STRIDE(32, os)) { V T3, Tn, Ti, TC, T6, TB, Tl, To, Td, TN, Tz, TH, Ta, TM, Tu; V TG; { V T1, T2, Tj, Tk; T1 = LD(&(ri[0]), ivs, &(ri[0])); T2 = LD(&(ri[WS(is, 4)]), ivs, &(ri[0])); T3 = VADD(T1, T2); Tn = VSUB(T1, T2); { V Tg, Th, T4, T5; Tg = LD(&(ii[0]), ivs, &(ii[0])); Th = LD(&(ii[WS(is, 4)]), ivs, &(ii[0])); Ti = VADD(Tg, Th); TC = VSUB(Tg, Th); T4 = LD(&(ri[WS(is, 2)]), ivs, &(ri[0])); T5 = LD(&(ri[WS(is, 6)]), ivs, &(ri[0])); T6 = VADD(T4, T5); TB = VSUB(T4, T5); } Tj = LD(&(ii[WS(is, 2)]), ivs, &(ii[0])); Tk = LD(&(ii[WS(is, 6)]), ivs, &(ii[0])); Tl = VADD(Tj, Tk); To = VSUB(Tj, Tk); { V Tb, Tc, Tv, Tw, Tx, Ty; Tb = LD(&(ri[WS(is, 7)]), ivs, &(ri[WS(is, 1)])); Tc = LD(&(ri[WS(is, 3)]), ivs, &(ri[WS(is, 1)])); Tv = VSUB(Tb, Tc); Tw = LD(&(ii[WS(is, 7)]), ivs, &(ii[WS(is, 1)])); Tx = LD(&(ii[WS(is, 3)]), ivs, &(ii[WS(is, 1)])); Ty = VSUB(Tw, Tx); Td = VADD(Tb, Tc); TN = VADD(Tw, Tx); Tz = VSUB(Tv, Ty); TH = VADD(Tv, Ty); } { V T8, T9, Tq, Tr, Ts, Tt; T8 = LD(&(ri[WS(is, 1)]), ivs, &(ri[WS(is, 1)])); T9 = LD(&(ri[WS(is, 5)]), ivs, &(ri[WS(is, 1)])); Tq = VSUB(T8, T9); Tr = LD(&(ii[WS(is, 1)]), ivs, &(ii[WS(is, 1)])); Ts = LD(&(ii[WS(is, 5)]), ivs, &(ii[WS(is, 1)])); Tt = VSUB(Tr, Ts); Ta = VADD(T8, T9); TM = VADD(Tr, Ts); Tu = VADD(Tq, Tt); TG = VSUB(Tt, Tq); } } { V TR, TS, TT, TU, TV, TW, TX, TY; { V T7, Te, TP, TQ; T7 = VADD(T3, T6); Te = VADD(Ta, Td); TR = VSUB(T7, Te); STM4(&(ro[4]), TR, ovs, &(ro[0])); TS = VADD(T7, Te); STM4(&(ro[0]), TS, ovs, &(ro[0])); TP = VADD(Ti, Tl); TQ = VADD(TM, TN); TT = VSUB(TP, TQ); STM4(&(io[4]), TT, ovs, &(io[0])); TU = VADD(TP, TQ); STM4(&(io[0]), TU, ovs, &(io[0])); } { V Tf, Tm, TL, TO; Tf = VSUB(Td, Ta); Tm = VSUB(Ti, Tl); TV = VADD(Tf, Tm); STM4(&(io[2]), TV, ovs, &(io[0])); TW = VSUB(Tm, Tf); STM4(&(io[6]), TW, ovs, &(io[0])); TL = VSUB(T3, T6); TO = VSUB(TM, TN); TX = VSUB(TL, TO); STM4(&(ro[6]), TX, ovs, &(ro[0])); TY = VADD(TL, TO); STM4(&(ro[2]), TY, ovs, &(ro[0])); } { V TZ, T10, T11, T12; { V Tp, TA, TJ, TK; Tp = VADD(Tn, To); TA = VMUL(LDK(KP707106781), VADD(Tu, Tz)); TZ = VSUB(Tp, TA); STM4(&(ro[5]), TZ, ovs, &(ro[1])); T10 = VADD(Tp, TA); STM4(&(ro[1]), T10, ovs, &(ro[1])); TJ = VSUB(TC, TB); TK = VMUL(LDK(KP707106781), VADD(TG, TH)); T11 = VSUB(TJ, TK); STM4(&(io[5]), T11, ovs, &(io[1])); T12 = VADD(TJ, TK); STM4(&(io[1]), T12, ovs, &(io[1])); } { V TD, TE, T13, T14; TD = VADD(TB, TC); TE = VMUL(LDK(KP707106781), VSUB(Tz, Tu)); T13 = VSUB(TD, TE); STM4(&(io[7]), T13, ovs, &(io[1])); STN4(&(io[4]), TT, T11, TW, T13, ovs); T14 = VADD(TD, TE); STM4(&(io[3]), T14, ovs, &(io[1])); STN4(&(io[0]), TU, T12, TV, T14, ovs); } { V TF, TI, T15, T16; TF = VSUB(Tn, To); TI = VMUL(LDK(KP707106781), VSUB(TG, TH)); T15 = VSUB(TF, TI); STM4(&(ro[7]), T15, ovs, &(ro[1])); STN4(&(ro[4]), TR, TZ, TX, T15, ovs); T16 = VADD(TF, TI); STM4(&(ro[3]), T16, ovs, &(ro[1])); STN4(&(ro[0]), TS, T10, TY, T16, ovs); } } } } } VLEAVE(); } static const kdft_desc desc = { 8, XSIMD_STRING("n2sv_8"), {52, 4, 0, 0}, &GENUS, 0, 1, 0, 0 }; void XSIMD(codelet_n2sv_8) (planner *p) { X(kdft_register) (p, n2sv_8, &desc); } #endif /* HAVE_FMA */ fftw-3.3.3/dft/simd/common/n2sv_16.c0000644000175000017500000005320112054410435013751 00000000000000/* * Copyright (c) 2003, 2007-11 Matteo Frigo * Copyright (c) 2003, 2007-11 Massachusetts Institute of Technology * * This program is free software; you can redistribute it and/or modify * it under the terms of the GNU General Public License as published by * the Free Software Foundation; either version 2 of the License, or * (at your option) any later version. * * This program is distributed in the hope that it will be useful, * but WITHOUT ANY WARRANTY; without even the implied warranty of * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the * GNU General Public License for more details. * * You should have received a copy of the GNU General Public License * along with this program; if not, write to the Free Software * Foundation, Inc., 51 Franklin Street, Fifth Floor, Boston, MA 02110-1301 USA * */ /* This file was automatically generated --- DO NOT EDIT */ /* Generated on Sun Nov 25 07:37:48 EST 2012 */ #include "codelet-dft.h" #ifdef HAVE_FMA /* Generated by: ../../../genfft/gen_notw.native -fma -reorder-insns -schedule-for-pipeline -simd -compact -variables 4 -pipeline-latency 8 -n 16 -name n2sv_16 -with-ostride 1 -include n2s.h -store-multiple 4 */ /* * This function contains 144 FP additions, 40 FP multiplications, * (or, 104 additions, 0 multiplications, 40 fused multiply/add), * 110 stack variables, 3 constants, and 72 memory accesses */ #include "n2s.h" static void n2sv_16(const R *ri, const R *ii, R *ro, R *io, stride is, stride os, INT v, INT ivs, INT ovs) { DVK(KP923879532, +0.923879532511286756128183189396788286822416626); DVK(KP707106781, +0.707106781186547524400844362104849039284835938); DVK(KP414213562, +0.414213562373095048801688724209698078569671875); { INT i; for (i = v; i > 0; i = i - (2 * VL), ri = ri + ((2 * VL) * ivs), ii = ii + ((2 * VL) * ivs), ro = ro + ((2 * VL) * ovs), io = io + ((2 * VL) * ovs), MAKE_VOLATILE_STRIDE(64, is), MAKE_VOLATILE_STRIDE(64, os)) { V T2p, T2q, T2r, T2s, T2x, T2y, T2z, T2A, T1M, T1N, T1L, T1P, T2F, T2G, T2H; V T2I, T1O, T1Q; { V T1l, T1H, T1R, T7, T1x, TN, TC, T25, T1E, T1b, T1Z, Tt, T2h, T22, T1D; V T1g, T1n, TQ, T11, Ti, Te, T26, T1m, TT, T1S, TJ, TZ, T1V, TW, Tl; V T12, T13; { V Tq, T1c, Tp, T20, T1a, Tr, T1d, T1e; { V T1, T2, Tw, Tx, T4, T5, Tz, TA; T1 = LD(&(ri[0]), ivs, &(ri[0])); T2 = LD(&(ri[WS(is, 8)]), ivs, &(ri[0])); Tw = LD(&(ii[0]), ivs, &(ii[0])); Tx = LD(&(ii[WS(is, 8)]), ivs, &(ii[0])); T4 = LD(&(ri[WS(is, 4)]), ivs, &(ri[0])); T5 = LD(&(ri[WS(is, 12)]), ivs, &(ri[0])); Tz = LD(&(ii[WS(is, 4)]), ivs, &(ii[0])); TA = LD(&(ii[WS(is, 12)]), ivs, &(ii[0])); { V Tn, TL, T3, T1k, Ty, T1j, T6, TM, TB, To, T18, T19; Tn = LD(&(ri[WS(is, 15)]), ivs, &(ri[WS(is, 1)])); TL = VSUB(T1, T2); T3 = VADD(T1, T2); T1k = VSUB(Tw, Tx); Ty = VADD(Tw, Tx); T1j = VSUB(T4, T5); T6 = VADD(T4, T5); TM = VSUB(Tz, TA); TB = VADD(Tz, TA); To = LD(&(ri[WS(is, 7)]), ivs, &(ri[WS(is, 1)])); T18 = LD(&(ii[WS(is, 15)]), ivs, &(ii[WS(is, 1)])); T19 = LD(&(ii[WS(is, 7)]), ivs, &(ii[WS(is, 1)])); Tq = LD(&(ri[WS(is, 3)]), ivs, &(ri[WS(is, 1)])); T1l = VADD(T1j, T1k); T1H = VSUB(T1k, T1j); T1R = VSUB(T3, T6); T7 = VADD(T3, T6); T1x = VADD(TL, TM); TN = VSUB(TL, TM); TC = VADD(Ty, TB); T25 = VSUB(Ty, TB); T1c = VSUB(Tn, To); Tp = VADD(Tn, To); T20 = VADD(T18, T19); T1a = VSUB(T18, T19); Tr = LD(&(ri[WS(is, 11)]), ivs, &(ri[WS(is, 1)])); T1d = LD(&(ii[WS(is, 3)]), ivs, &(ii[WS(is, 1)])); T1e = LD(&(ii[WS(is, 11)]), ivs, &(ii[WS(is, 1)])); } } { V Tb, Ta, TF, Tc, TG, TH, TP, TO; { V T8, T9, TD, TE; T8 = LD(&(ri[WS(is, 2)]), ivs, &(ri[0])); T9 = LD(&(ri[WS(is, 10)]), ivs, &(ri[0])); TD = LD(&(ii[WS(is, 2)]), ivs, &(ii[0])); TE = LD(&(ii[WS(is, 10)]), ivs, &(ii[0])); Tb = LD(&(ri[WS(is, 14)]), ivs, &(ri[0])); { V T17, Ts, T21, T1f; T17 = VSUB(Tq, Tr); Ts = VADD(Tq, Tr); T21 = VADD(T1d, T1e); T1f = VSUB(T1d, T1e); TP = VSUB(T8, T9); Ta = VADD(T8, T9); TO = VSUB(TD, TE); TF = VADD(TD, TE); T1E = VSUB(T1a, T17); T1b = VADD(T17, T1a); T1Z = VSUB(Tp, Ts); Tt = VADD(Tp, Ts); T2h = VADD(T20, T21); T22 = VSUB(T20, T21); T1D = VADD(T1c, T1f); T1g = VSUB(T1c, T1f); Tc = LD(&(ri[WS(is, 6)]), ivs, &(ri[0])); } TG = LD(&(ii[WS(is, 14)]), ivs, &(ii[0])); TH = LD(&(ii[WS(is, 6)]), ivs, &(ii[0])); } T1n = VADD(TP, TO); TQ = VSUB(TO, TP); { V Tg, Th, TX, TR, Td, TS, TI, TY, Tj, Tk; Tg = LD(&(ri[WS(is, 1)]), ivs, &(ri[WS(is, 1)])); Th = LD(&(ri[WS(is, 9)]), ivs, &(ri[WS(is, 1)])); TX = LD(&(ii[WS(is, 1)]), ivs, &(ii[WS(is, 1)])); TR = VSUB(Tb, Tc); Td = VADD(Tb, Tc); TS = VSUB(TG, TH); TI = VADD(TG, TH); TY = LD(&(ii[WS(is, 9)]), ivs, &(ii[WS(is, 1)])); Tj = LD(&(ri[WS(is, 5)]), ivs, &(ri[WS(is, 1)])); T11 = VSUB(Tg, Th); Ti = VADD(Tg, Th); Tk = LD(&(ri[WS(is, 13)]), ivs, &(ri[WS(is, 1)])); Te = VADD(Ta, Td); T26 = VSUB(Td, Ta); T1m = VSUB(TR, TS); TT = VADD(TR, TS); T1S = VSUB(TF, TI); TJ = VADD(TF, TI); TZ = VSUB(TX, TY); T1V = VADD(TX, TY); TW = VSUB(Tj, Tk); Tl = VADD(Tj, Tk); T12 = LD(&(ii[WS(is, 5)]), ivs, &(ii[WS(is, 1)])); T13 = LD(&(ii[WS(is, 13)]), ivs, &(ii[WS(is, 1)])); } } } { V T2f, Tf, T2j, TK, Tm, T1U, T10, T1B, T14, T1W; T2f = VSUB(T7, Te); Tf = VADD(T7, Te); T2j = VADD(TC, TJ); TK = VSUB(TC, TJ); Tm = VADD(Ti, Tl); T1U = VSUB(Ti, Tl); T10 = VADD(TW, TZ); T1B = VSUB(TZ, TW); T14 = VSUB(T12, T13); T1W = VADD(T12, T13); { V T29, T1T, T27, T2d, T2b, T23, T15, T1A, T2l, T2m, T2n, T2o, T2i, T2k, T1Y; V T2a; { V Tv, Tu, T1X, T2g; T29 = VSUB(T1R, T1S); T1T = VADD(T1R, T1S); T27 = VSUB(T25, T26); T2d = VADD(T26, T25); T2b = VADD(T1Z, T22); T23 = VSUB(T1Z, T22); Tv = VSUB(Tt, Tm); Tu = VADD(Tm, Tt); T1X = VSUB(T1V, T1W); T2g = VADD(T1V, T1W); T15 = VSUB(T11, T14); T1A = VADD(T11, T14); T2l = VSUB(TK, Tv); STM4(&(io[12]), T2l, ovs, &(io[0])); T2m = VADD(Tv, TK); STM4(&(io[4]), T2m, ovs, &(io[0])); T2n = VADD(Tf, Tu); STM4(&(ro[0]), T2n, ovs, &(ro[0])); T2o = VSUB(Tf, Tu); STM4(&(ro[8]), T2o, ovs, &(ro[0])); T2i = VSUB(T2g, T2h); T2k = VADD(T2g, T2h); T1Y = VADD(T1U, T1X); T2a = VSUB(T1X, T1U); } { V T1I, T1y, T1t, T16, T1v, TV, T1r, T1p, T2t, T2u, T2v, T2w, T1h, T1s, TU; V T1o; T1I = VADD(TQ, TT); TU = VSUB(TQ, TT); T1o = VSUB(T1m, T1n); T1y = VADD(T1n, T1m); T1t = VFNMS(LDK(KP414213562), T10, T15); T16 = VFMA(LDK(KP414213562), T15, T10); T2p = VADD(T2f, T2i); STM4(&(ro[4]), T2p, ovs, &(ro[0])); T2q = VSUB(T2f, T2i); STM4(&(ro[12]), T2q, ovs, &(ro[0])); T2r = VADD(T2j, T2k); STM4(&(io[0]), T2r, ovs, &(io[0])); T2s = VSUB(T2j, T2k); STM4(&(io[8]), T2s, ovs, &(io[0])); { V T28, T24, T2e, T2c; T28 = VSUB(T23, T1Y); T24 = VADD(T1Y, T23); T2e = VADD(T2a, T2b); T2c = VSUB(T2a, T2b); T1v = VFNMS(LDK(KP707106781), TU, TN); TV = VFMA(LDK(KP707106781), TU, TN); T1r = VFMA(LDK(KP707106781), T1o, T1l); T1p = VFNMS(LDK(KP707106781), T1o, T1l); T2t = VFNMS(LDK(KP707106781), T28, T27); STM4(&(io[14]), T2t, ovs, &(io[0])); T2u = VFMA(LDK(KP707106781), T28, T27); STM4(&(io[6]), T2u, ovs, &(io[0])); T2v = VFMA(LDK(KP707106781), T24, T1T); STM4(&(ro[2]), T2v, ovs, &(ro[0])); T2w = VFNMS(LDK(KP707106781), T24, T1T); STM4(&(ro[10]), T2w, ovs, &(ro[0])); T2x = VFNMS(LDK(KP707106781), T2e, T2d); STM4(&(io[10]), T2x, ovs, &(io[0])); T2y = VFMA(LDK(KP707106781), T2e, T2d); STM4(&(io[2]), T2y, ovs, &(io[0])); T2z = VFMA(LDK(KP707106781), T2c, T29); STM4(&(ro[6]), T2z, ovs, &(ro[0])); T2A = VFNMS(LDK(KP707106781), T2c, T29); STM4(&(ro[14]), T2A, ovs, &(ro[0])); T1h = VFNMS(LDK(KP414213562), T1g, T1b); T1s = VFMA(LDK(KP414213562), T1b, T1g); } { V T1z, T1J, T1K, T1G, T2B, T2C, T2D, T2E, T1C, T1F; T1M = VFNMS(LDK(KP414213562), T1A, T1B); T1C = VFMA(LDK(KP414213562), T1B, T1A); T1F = VFNMS(LDK(KP414213562), T1E, T1D); T1N = VFMA(LDK(KP414213562), T1D, T1E); { V T1q, T1i, T1w, T1u; T1q = VADD(T16, T1h); T1i = VSUB(T16, T1h); T1w = VADD(T1t, T1s); T1u = VSUB(T1s, T1t); T1L = VFNMS(LDK(KP707106781), T1y, T1x); T1z = VFMA(LDK(KP707106781), T1y, T1x); T1P = VFMA(LDK(KP707106781), T1I, T1H); T1J = VFNMS(LDK(KP707106781), T1I, T1H); T1K = VSUB(T1F, T1C); T1G = VADD(T1C, T1F); T2B = VFMA(LDK(KP923879532), T1q, T1p); STM4(&(io[15]), T2B, ovs, &(io[1])); T2C = VFNMS(LDK(KP923879532), T1q, T1p); STM4(&(io[7]), T2C, ovs, &(io[1])); T2D = VFMA(LDK(KP923879532), T1i, TV); STM4(&(ro[3]), T2D, ovs, &(ro[1])); T2E = VFNMS(LDK(KP923879532), T1i, TV); STM4(&(ro[11]), T2E, ovs, &(ro[1])); T2F = VFMA(LDK(KP923879532), T1w, T1v); STM4(&(ro[15]), T2F, ovs, &(ro[1])); T2G = VFNMS(LDK(KP923879532), T1w, T1v); STM4(&(ro[7]), T2G, ovs, &(ro[1])); T2H = VFMA(LDK(KP923879532), T1u, T1r); STM4(&(io[3]), T2H, ovs, &(io[1])); T2I = VFNMS(LDK(KP923879532), T1u, T1r); STM4(&(io[11]), T2I, ovs, &(io[1])); } { V T2J, T2K, T2L, T2M; T2J = VFNMS(LDK(KP923879532), T1G, T1z); STM4(&(ro[9]), T2J, ovs, &(ro[1])); STN4(&(ro[8]), T2o, T2J, T2w, T2E, ovs); T2K = VFMA(LDK(KP923879532), T1G, T1z); STM4(&(ro[1]), T2K, ovs, &(ro[1])); STN4(&(ro[0]), T2n, T2K, T2v, T2D, ovs); T2L = VFNMS(LDK(KP923879532), T1K, T1J); STM4(&(io[13]), T2L, ovs, &(io[1])); STN4(&(io[12]), T2l, T2L, T2t, T2B, ovs); T2M = VFMA(LDK(KP923879532), T1K, T1J); STM4(&(io[5]), T2M, ovs, &(io[1])); STN4(&(io[4]), T2m, T2M, T2u, T2C, ovs); } } } } } } T1O = VSUB(T1M, T1N); T1Q = VADD(T1M, T1N); { V T2N, T2O, T2P, T2Q; T2N = VFMA(LDK(KP923879532), T1Q, T1P); STM4(&(io[1]), T2N, ovs, &(io[1])); STN4(&(io[0]), T2r, T2N, T2y, T2H, ovs); T2O = VFNMS(LDK(KP923879532), T1Q, T1P); STM4(&(io[9]), T2O, ovs, &(io[1])); STN4(&(io[8]), T2s, T2O, T2x, T2I, ovs); T2P = VFMA(LDK(KP923879532), T1O, T1L); STM4(&(ro[5]), T2P, ovs, &(ro[1])); STN4(&(ro[4]), T2p, T2P, T2z, T2G, ovs); T2Q = VFNMS(LDK(KP923879532), T1O, T1L); STM4(&(ro[13]), T2Q, ovs, &(ro[1])); STN4(&(ro[12]), T2q, T2Q, T2A, T2F, ovs); } } } VLEAVE(); } static const kdft_desc desc = { 16, XSIMD_STRING("n2sv_16"), {104, 0, 40, 0}, &GENUS, 0, 1, 0, 0 }; void XSIMD(codelet_n2sv_16) (planner *p) { X(kdft_register) (p, n2sv_16, &desc); } #else /* HAVE_FMA */ /* Generated by: ../../../genfft/gen_notw.native -simd -compact -variables 4 -pipeline-latency 8 -n 16 -name n2sv_16 -with-ostride 1 -include n2s.h -store-multiple 4 */ /* * This function contains 144 FP additions, 24 FP multiplications, * (or, 136 additions, 16 multiplications, 8 fused multiply/add), * 74 stack variables, 3 constants, and 72 memory accesses */ #include "n2s.h" static void n2sv_16(const R *ri, const R *ii, R *ro, R *io, stride is, stride os, INT v, INT ivs, INT ovs) { DVK(KP382683432, +0.382683432365089771728459984030398866761344562); DVK(KP923879532, +0.923879532511286756128183189396788286822416626); DVK(KP707106781, +0.707106781186547524400844362104849039284835938); { INT i; for (i = v; i > 0; i = i - (2 * VL), ri = ri + ((2 * VL) * ivs), ii = ii + ((2 * VL) * ivs), ro = ro + ((2 * VL) * ovs), io = io + ((2 * VL) * ovs), MAKE_VOLATILE_STRIDE(64, is), MAKE_VOLATILE_STRIDE(64, os)) { V T7, T1R, T25, TC, TN, T1x, T1H, T1l, Tt, T22, T2h, T1b, T1g, T1E, T1Z; V T1D, Te, T1S, T26, TJ, TQ, T1m, T1n, TT, Tm, T1X, T2g, T10, T15, T1B; V T1U, T1A; { V T3, TL, Ty, T1k, T6, T1j, TB, TM; { V T1, T2, Tw, Tx; T1 = LD(&(ri[0]), ivs, &(ri[0])); T2 = LD(&(ri[WS(is, 8)]), ivs, &(ri[0])); T3 = VADD(T1, T2); TL = VSUB(T1, T2); Tw = LD(&(ii[0]), ivs, &(ii[0])); Tx = LD(&(ii[WS(is, 8)]), ivs, &(ii[0])); Ty = VADD(Tw, Tx); T1k = VSUB(Tw, Tx); } { V T4, T5, Tz, TA; T4 = LD(&(ri[WS(is, 4)]), ivs, &(ri[0])); T5 = LD(&(ri[WS(is, 12)]), ivs, &(ri[0])); T6 = VADD(T4, T5); T1j = VSUB(T4, T5); Tz = LD(&(ii[WS(is, 4)]), ivs, &(ii[0])); TA = LD(&(ii[WS(is, 12)]), ivs, &(ii[0])); TB = VADD(Tz, TA); TM = VSUB(Tz, TA); } T7 = VADD(T3, T6); T1R = VSUB(T3, T6); T25 = VSUB(Ty, TB); TC = VADD(Ty, TB); TN = VSUB(TL, TM); T1x = VADD(TL, TM); T1H = VSUB(T1k, T1j); T1l = VADD(T1j, T1k); } { V Tp, T17, T1f, T20, Ts, T1c, T1a, T21; { V Tn, To, T1d, T1e; Tn = LD(&(ri[WS(is, 15)]), ivs, &(ri[WS(is, 1)])); To = LD(&(ri[WS(is, 7)]), ivs, &(ri[WS(is, 1)])); Tp = VADD(Tn, To); T17 = VSUB(Tn, To); T1d = LD(&(ii[WS(is, 15)]), ivs, &(ii[WS(is, 1)])); T1e = LD(&(ii[WS(is, 7)]), ivs, &(ii[WS(is, 1)])); T1f = VSUB(T1d, T1e); T20 = VADD(T1d, T1e); } { V Tq, Tr, T18, T19; Tq = LD(&(ri[WS(is, 3)]), ivs, &(ri[WS(is, 1)])); Tr = LD(&(ri[WS(is, 11)]), ivs, &(ri[WS(is, 1)])); Ts = VADD(Tq, Tr); T1c = VSUB(Tq, Tr); T18 = LD(&(ii[WS(is, 3)]), ivs, &(ii[WS(is, 1)])); T19 = LD(&(ii[WS(is, 11)]), ivs, &(ii[WS(is, 1)])); T1a = VSUB(T18, T19); T21 = VADD(T18, T19); } Tt = VADD(Tp, Ts); T22 = VSUB(T20, T21); T2h = VADD(T20, T21); T1b = VSUB(T17, T1a); T1g = VADD(T1c, T1f); T1E = VSUB(T1f, T1c); T1Z = VSUB(Tp, Ts); T1D = VADD(T17, T1a); } { V Ta, TP, TF, TO, Td, TR, TI, TS; { V T8, T9, TD, TE; T8 = LD(&(ri[WS(is, 2)]), ivs, &(ri[0])); T9 = LD(&(ri[WS(is, 10)]), ivs, &(ri[0])); Ta = VADD(T8, T9); TP = VSUB(T8, T9); TD = LD(&(ii[WS(is, 2)]), ivs, &(ii[0])); TE = LD(&(ii[WS(is, 10)]), ivs, &(ii[0])); TF = VADD(TD, TE); TO = VSUB(TD, TE); } { V Tb, Tc, TG, TH; Tb = LD(&(ri[WS(is, 14)]), ivs, &(ri[0])); Tc = LD(&(ri[WS(is, 6)]), ivs, &(ri[0])); Td = VADD(Tb, Tc); TR = VSUB(Tb, Tc); TG = LD(&(ii[WS(is, 14)]), ivs, &(ii[0])); TH = LD(&(ii[WS(is, 6)]), ivs, &(ii[0])); TI = VADD(TG, TH); TS = VSUB(TG, TH); } Te = VADD(Ta, Td); T1S = VSUB(TF, TI); T26 = VSUB(Td, Ta); TJ = VADD(TF, TI); TQ = VSUB(TO, TP); T1m = VSUB(TR, TS); T1n = VADD(TP, TO); TT = VADD(TR, TS); } { V Ti, T11, TZ, T1V, Tl, TW, T14, T1W; { V Tg, Th, TX, TY; Tg = LD(&(ri[WS(is, 1)]), ivs, &(ri[WS(is, 1)])); Th = LD(&(ri[WS(is, 9)]), ivs, &(ri[WS(is, 1)])); Ti = VADD(Tg, Th); T11 = VSUB(Tg, Th); TX = LD(&(ii[WS(is, 1)]), ivs, &(ii[WS(is, 1)])); TY = LD(&(ii[WS(is, 9)]), ivs, &(ii[WS(is, 1)])); TZ = VSUB(TX, TY); T1V = VADD(TX, TY); } { V Tj, Tk, T12, T13; Tj = LD(&(ri[WS(is, 5)]), ivs, &(ri[WS(is, 1)])); Tk = LD(&(ri[WS(is, 13)]), ivs, &(ri[WS(is, 1)])); Tl = VADD(Tj, Tk); TW = VSUB(Tj, Tk); T12 = LD(&(ii[WS(is, 5)]), ivs, &(ii[WS(is, 1)])); T13 = LD(&(ii[WS(is, 13)]), ivs, &(ii[WS(is, 1)])); T14 = VSUB(T12, T13); T1W = VADD(T12, T13); } Tm = VADD(Ti, Tl); T1X = VSUB(T1V, T1W); T2g = VADD(T1V, T1W); T10 = VADD(TW, TZ); T15 = VSUB(T11, T14); T1B = VADD(T11, T14); T1U = VSUB(Ti, Tl); T1A = VSUB(TZ, TW); } { V T2l, T2m, T2n, T2o, T2p, T2q, T2r, T2s; { V Tf, Tu, T2j, T2k; Tf = VADD(T7, Te); Tu = VADD(Tm, Tt); T2l = VSUB(Tf, Tu); STM4(&(ro[8]), T2l, ovs, &(ro[0])); T2m = VADD(Tf, Tu); STM4(&(ro[0]), T2m, ovs, &(ro[0])); T2j = VADD(TC, TJ); T2k = VADD(T2g, T2h); T2n = VSUB(T2j, T2k); STM4(&(io[8]), T2n, ovs, &(io[0])); T2o = VADD(T2j, T2k); STM4(&(io[0]), T2o, ovs, &(io[0])); } { V Tv, TK, T2f, T2i; Tv = VSUB(Tt, Tm); TK = VSUB(TC, TJ); T2p = VADD(Tv, TK); STM4(&(io[4]), T2p, ovs, &(io[0])); T2q = VSUB(TK, Tv); STM4(&(io[12]), T2q, ovs, &(io[0])); T2f = VSUB(T7, Te); T2i = VSUB(T2g, T2h); T2r = VSUB(T2f, T2i); STM4(&(ro[12]), T2r, ovs, &(ro[0])); T2s = VADD(T2f, T2i); STM4(&(ro[4]), T2s, ovs, &(ro[0])); } { V T2t, T2u, T2v, T2w, T2x, T2y, T2z, T2A; { V T1T, T27, T24, T28, T1Y, T23; T1T = VADD(T1R, T1S); T27 = VSUB(T25, T26); T1Y = VADD(T1U, T1X); T23 = VSUB(T1Z, T22); T24 = VMUL(LDK(KP707106781), VADD(T1Y, T23)); T28 = VMUL(LDK(KP707106781), VSUB(T23, T1Y)); T2t = VSUB(T1T, T24); STM4(&(ro[10]), T2t, ovs, &(ro[0])); T2u = VADD(T27, T28); STM4(&(io[6]), T2u, ovs, &(io[0])); T2v = VADD(T1T, T24); STM4(&(ro[2]), T2v, ovs, &(ro[0])); T2w = VSUB(T27, T28); STM4(&(io[14]), T2w, ovs, &(io[0])); } { V T29, T2d, T2c, T2e, T2a, T2b; T29 = VSUB(T1R, T1S); T2d = VADD(T26, T25); T2a = VSUB(T1X, T1U); T2b = VADD(T1Z, T22); T2c = VMUL(LDK(KP707106781), VSUB(T2a, T2b)); T2e = VMUL(LDK(KP707106781), VADD(T2a, T2b)); T2x = VSUB(T29, T2c); STM4(&(ro[14]), T2x, ovs, &(ro[0])); T2y = VADD(T2d, T2e); STM4(&(io[2]), T2y, ovs, &(io[0])); T2z = VADD(T29, T2c); STM4(&(ro[6]), T2z, ovs, &(ro[0])); T2A = VSUB(T2d, T2e); STM4(&(io[10]), T2A, ovs, &(io[0])); } { V T2B, T2C, T2D, T2E, T2F, T2G, T2H, T2I; { V TV, T1r, T1p, T1v, T1i, T1q, T1u, T1w, TU, T1o; TU = VMUL(LDK(KP707106781), VSUB(TQ, TT)); TV = VADD(TN, TU); T1r = VSUB(TN, TU); T1o = VMUL(LDK(KP707106781), VSUB(T1m, T1n)); T1p = VSUB(T1l, T1o); T1v = VADD(T1l, T1o); { V T16, T1h, T1s, T1t; T16 = VFMA(LDK(KP923879532), T10, VMUL(LDK(KP382683432), T15)); T1h = VFNMS(LDK(KP923879532), T1g, VMUL(LDK(KP382683432), T1b)); T1i = VADD(T16, T1h); T1q = VSUB(T1h, T16); T1s = VFNMS(LDK(KP923879532), T15, VMUL(LDK(KP382683432), T10)); T1t = VFMA(LDK(KP382683432), T1g, VMUL(LDK(KP923879532), T1b)); T1u = VSUB(T1s, T1t); T1w = VADD(T1s, T1t); } T2B = VSUB(TV, T1i); STM4(&(ro[11]), T2B, ovs, &(ro[1])); T2C = VSUB(T1v, T1w); STM4(&(io[11]), T2C, ovs, &(io[1])); T2D = VADD(TV, T1i); STM4(&(ro[3]), T2D, ovs, &(ro[1])); T2E = VADD(T1v, T1w); STM4(&(io[3]), T2E, ovs, &(io[1])); T2F = VSUB(T1p, T1q); STM4(&(io[15]), T2F, ovs, &(io[1])); T2G = VSUB(T1r, T1u); STM4(&(ro[15]), T2G, ovs, &(ro[1])); T2H = VADD(T1p, T1q); STM4(&(io[7]), T2H, ovs, &(io[1])); T2I = VADD(T1r, T1u); STM4(&(ro[7]), T2I, ovs, &(ro[1])); } { V T1z, T1L, T1J, T1P, T1G, T1K, T1O, T1Q, T1y, T1I; T1y = VMUL(LDK(KP707106781), VADD(T1n, T1m)); T1z = VADD(T1x, T1y); T1L = VSUB(T1x, T1y); T1I = VMUL(LDK(KP707106781), VADD(TQ, TT)); T1J = VSUB(T1H, T1I); T1P = VADD(T1H, T1I); { V T1C, T1F, T1M, T1N; T1C = VFMA(LDK(KP382683432), T1A, VMUL(LDK(KP923879532), T1B)); T1F = VFNMS(LDK(KP382683432), T1E, VMUL(LDK(KP923879532), T1D)); T1G = VADD(T1C, T1F); T1K = VSUB(T1F, T1C); T1M = VFNMS(LDK(KP382683432), T1B, VMUL(LDK(KP923879532), T1A)); T1N = VFMA(LDK(KP923879532), T1E, VMUL(LDK(KP382683432), T1D)); T1O = VSUB(T1M, T1N); T1Q = VADD(T1M, T1N); } { V T2J, T2K, T2L, T2M; T2J = VSUB(T1z, T1G); STM4(&(ro[9]), T2J, ovs, &(ro[1])); STN4(&(ro[8]), T2l, T2J, T2t, T2B, ovs); T2K = VSUB(T1P, T1Q); STM4(&(io[9]), T2K, ovs, &(io[1])); STN4(&(io[8]), T2n, T2K, T2A, T2C, ovs); T2L = VADD(T1z, T1G); STM4(&(ro[1]), T2L, ovs, &(ro[1])); STN4(&(ro[0]), T2m, T2L, T2v, T2D, ovs); T2M = VADD(T1P, T1Q); STM4(&(io[1]), T2M, ovs, &(io[1])); STN4(&(io[0]), T2o, T2M, T2y, T2E, ovs); } { V T2N, T2O, T2P, T2Q; T2N = VSUB(T1J, T1K); STM4(&(io[13]), T2N, ovs, &(io[1])); STN4(&(io[12]), T2q, T2N, T2w, T2F, ovs); T2O = VSUB(T1L, T1O); STM4(&(ro[13]), T2O, ovs, &(ro[1])); STN4(&(ro[12]), T2r, T2O, T2x, T2G, ovs); T2P = VADD(T1J, T1K); STM4(&(io[5]), T2P, ovs, &(io[1])); STN4(&(io[4]), T2p, T2P, T2u, T2H, ovs); T2Q = VADD(T1L, T1O); STM4(&(ro[5]), T2Q, ovs, &(ro[1])); STN4(&(ro[4]), T2s, T2Q, T2z, T2I, ovs); } } } } } } } VLEAVE(); } static const kdft_desc desc = { 16, XSIMD_STRING("n2sv_16"), {136, 16, 8, 0}, &GENUS, 0, 1, 0, 0 }; void XSIMD(codelet_n2sv_16) (planner *p) { X(kdft_register) (p, n2sv_16, &desc); } #endif /* HAVE_FMA */ fftw-3.3.3/dft/simd/common/n2sv_32.c0000644000175000017500000014206012054410445013752 00000000000000/* * Copyright (c) 2003, 2007-11 Matteo Frigo * Copyright (c) 2003, 2007-11 Massachusetts Institute of Technology * * This program is free software; you can redistribute it and/or modify * it under the terms of the GNU General Public License as published by * the Free Software Foundation; either version 2 of the License, or * (at your option) any later version. * * This program is distributed in the hope that it will be useful, * but WITHOUT ANY WARRANTY; without even the implied warranty of * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the * GNU General Public License for more details. * * You should have received a copy of the GNU General Public License * along with this program; if not, write to the Free Software * Foundation, Inc., 51 Franklin Street, Fifth Floor, Boston, MA 02110-1301 USA * */ /* This file was automatically generated --- DO NOT EDIT */ /* Generated on Sun Nov 25 07:37:49 EST 2012 */ #include "codelet-dft.h" #ifdef HAVE_FMA /* Generated by: ../../../genfft/gen_notw.native -fma -reorder-insns -schedule-for-pipeline -simd -compact -variables 4 -pipeline-latency 8 -n 32 -name n2sv_32 -with-ostride 1 -include n2s.h -store-multiple 4 */ /* * This function contains 372 FP additions, 136 FP multiplications, * (or, 236 additions, 0 multiplications, 136 fused multiply/add), * 194 stack variables, 7 constants, and 144 memory accesses */ #include "n2s.h" static void n2sv_32(const R *ri, const R *ii, R *ro, R *io, stride is, stride os, INT v, INT ivs, INT ovs) { DVK(KP980785280, +0.980785280403230449126182236134239036973933731); DVK(KP198912367, +0.198912367379658006911597622644676228597850501); DVK(KP831469612, +0.831469612302545237078788377617905756738560812); DVK(KP668178637, +0.668178637919298919997757686523080761552472251); DVK(KP923879532, +0.923879532511286756128183189396788286822416626); DVK(KP707106781, +0.707106781186547524400844362104849039284835938); DVK(KP414213562, +0.414213562373095048801688724209698078569671875); { INT i; for (i = v; i > 0; i = i - (2 * VL), ri = ri + ((2 * VL) * ivs), ii = ii + ((2 * VL) * ivs), ro = ro + ((2 * VL) * ovs), io = io + ((2 * VL) * ovs), MAKE_VOLATILE_STRIDE(128, is), MAKE_VOLATILE_STRIDE(128, os)) { V T61, T62, T63, T64, T65, T66, T67, T68, T69, T6a, T6b, T6c, T6d, T6e, T6f; V T6g, T6h, T6i, T6j, T6k, T6l, T6m, T6n, T6o, T6p, T6q, T6r, T6s, T6t, T6u; V T6v, T6w, T3g, T3f, T6x, T6y, T6z, T6A, T6B, T6C, T6D, T6E, T4p, T49, T4l; V T4j, T6F, T6G, T6H, T6I, T6J, T6K, T6L, T6M, T3n, T3b, T3r, T3l, T3o, T3e; V T4q, T4o, T4k, T4g, T3h, T3p; { V T2T, T3T, T4r, T7, T3t, T1z, T18, T4Z, Te, T50, T1f, T4s, T1G, T3U, T2W; V T3u, Tm, T1n, T3X, T3y, T2Z, T1O, T53, T4w, Tt, T1u, T3W, T3B, T2Y, T1V; V T52, T4z, T3O, T2t, T3L, T2K, TZ, T5F, T4R, T5k, T5j, T4W, T5I, T5X, T2E; V T3M, T2N, T3P, T3H, T22, T3E, T2j, T4G, T5h, TK, T5A, T5D, T5W, T2d, T3F; V T4L, T5g, T3I, T2m; { V T1L, T1j, T1k, T1l, T4v, T1K, T3w; { V T1, T2, T12, T13, T4, T5, T15, T16; T1 = LD(&(ri[0]), ivs, &(ri[0])); T2 = LD(&(ri[WS(is, 16)]), ivs, &(ri[0])); T12 = LD(&(ii[0]), ivs, &(ii[0])); T13 = LD(&(ii[WS(is, 16)]), ivs, &(ii[0])); T4 = LD(&(ri[WS(is, 8)]), ivs, &(ri[0])); T5 = LD(&(ri[WS(is, 24)]), ivs, &(ri[0])); T15 = LD(&(ii[WS(is, 8)]), ivs, &(ii[0])); T16 = LD(&(ii[WS(is, 24)]), ivs, &(ii[0])); { V Tb, T1A, Ta, T1B, T1b, Tc, T1c, T1d; { V T8, T1x, T3, T2R, T14, T2S, T6, T1y, T17, T9, T19, T1a; T8 = LD(&(ri[WS(is, 4)]), ivs, &(ri[0])); T1x = VSUB(T1, T2); T3 = VADD(T1, T2); T2R = VSUB(T12, T13); T14 = VADD(T12, T13); T2S = VSUB(T4, T5); T6 = VADD(T4, T5); T1y = VSUB(T15, T16); T17 = VADD(T15, T16); T9 = LD(&(ri[WS(is, 20)]), ivs, &(ri[0])); T19 = LD(&(ii[WS(is, 4)]), ivs, &(ii[0])); T1a = LD(&(ii[WS(is, 20)]), ivs, &(ii[0])); Tb = LD(&(ri[WS(is, 28)]), ivs, &(ri[0])); T2T = VSUB(T2R, T2S); T3T = VADD(T2S, T2R); T4r = VSUB(T3, T6); T7 = VADD(T3, T6); T3t = VSUB(T1x, T1y); T1z = VADD(T1x, T1y); T18 = VADD(T14, T17); T4Z = VSUB(T14, T17); T1A = VSUB(T8, T9); Ta = VADD(T8, T9); T1B = VSUB(T19, T1a); T1b = VADD(T19, T1a); Tc = LD(&(ri[WS(is, 12)]), ivs, &(ri[0])); T1c = LD(&(ii[WS(is, 28)]), ivs, &(ii[0])); T1d = LD(&(ii[WS(is, 12)]), ivs, &(ii[0])); } { V Ti, T1I, T1J, Tl; { V T1h, T1C, T2U, T1D, Td, T1E, T1e, T1i, Tg, Th; Tg = LD(&(ri[WS(is, 2)]), ivs, &(ri[0])); Th = LD(&(ri[WS(is, 18)]), ivs, &(ri[0])); T1h = LD(&(ii[WS(is, 2)]), ivs, &(ii[0])); T1C = VADD(T1A, T1B); T2U = VSUB(T1B, T1A); T1D = VSUB(Tb, Tc); Td = VADD(Tb, Tc); T1E = VSUB(T1c, T1d); T1e = VADD(T1c, T1d); T1L = VSUB(Tg, Th); Ti = VADD(Tg, Th); T1i = LD(&(ii[WS(is, 18)]), ivs, &(ii[0])); { V T2V, T1F, Tj, Tk; Tj = LD(&(ri[WS(is, 10)]), ivs, &(ri[0])); Tk = LD(&(ri[WS(is, 26)]), ivs, &(ri[0])); Te = VADD(Ta, Td); T50 = VSUB(Td, Ta); T2V = VADD(T1D, T1E); T1F = VSUB(T1D, T1E); T1f = VADD(T1b, T1e); T4s = VSUB(T1b, T1e); T1j = VADD(T1h, T1i); T1I = VSUB(T1h, T1i); T1J = VSUB(Tj, Tk); Tl = VADD(Tj, Tk); T1G = VADD(T1C, T1F); T3U = VSUB(T1F, T1C); T2W = VADD(T2U, T2V); T3u = VSUB(T2U, T2V); T1k = LD(&(ii[WS(is, 10)]), ivs, &(ii[0])); T1l = LD(&(ii[WS(is, 26)]), ivs, &(ii[0])); } } T4v = VSUB(Ti, Tl); Tm = VADD(Ti, Tl); T1K = VSUB(T1I, T1J); T3w = VADD(T1J, T1I); } } } { V T1r, T1S, T1q, T1s, T4x, T1R, T3z; { V Tp, T1P, T1Q, Ts; { V Tn, To, T1o, T1M, T1m, T1p; Tn = LD(&(ri[WS(is, 30)]), ivs, &(ri[0])); To = LD(&(ri[WS(is, 14)]), ivs, &(ri[0])); T1o = LD(&(ii[WS(is, 30)]), ivs, &(ii[0])); T1M = VSUB(T1k, T1l); T1m = VADD(T1k, T1l); T1p = LD(&(ii[WS(is, 14)]), ivs, &(ii[0])); { V Tq, Tr, T3x, T1N, T4u; Tq = LD(&(ri[WS(is, 6)]), ivs, &(ri[0])); Tr = LD(&(ri[WS(is, 22)]), ivs, &(ri[0])); T1r = LD(&(ii[WS(is, 6)]), ivs, &(ii[0])); T1S = VSUB(Tn, To); Tp = VADD(Tn, To); T3x = VSUB(T1L, T1M); T1N = VADD(T1L, T1M); T4u = VSUB(T1j, T1m); T1n = VADD(T1j, T1m); T1P = VSUB(T1o, T1p); T1q = VADD(T1o, T1p); T1Q = VSUB(Tq, Tr); Ts = VADD(Tq, Tr); T3X = VFNMS(LDK(KP414213562), T3w, T3x); T3y = VFMA(LDK(KP414213562), T3x, T3w); T2Z = VFMA(LDK(KP414213562), T1K, T1N); T1O = VFNMS(LDK(KP414213562), T1N, T1K); T53 = VADD(T4v, T4u); T4w = VSUB(T4u, T4v); T1s = LD(&(ii[WS(is, 22)]), ivs, &(ii[0])); } } T4x = VSUB(Tp, Ts); Tt = VADD(Tp, Ts); T1R = VSUB(T1P, T1Q); T3z = VADD(T1Q, T1P); } { V T4S, T5G, T2y, T2L, T4V, T5H, T2D, T2M; { V T2G, TN, T4N, T2r, T2s, TQ, T2A, T4O, T2J, T2x, TU, T4T, T2w, T2z, TX; V T2B, T2H, T2I, TR; { V TL, TM, T2p, T1T, T1t, T2q; TL = LD(&(ri[WS(is, 31)]), ivs, &(ri[WS(is, 1)])); TM = LD(&(ri[WS(is, 15)]), ivs, &(ri[WS(is, 1)])); T2p = LD(&(ii[WS(is, 31)]), ivs, &(ii[WS(is, 1)])); T1T = VSUB(T1r, T1s); T1t = VADD(T1r, T1s); T2q = LD(&(ii[WS(is, 15)]), ivs, &(ii[WS(is, 1)])); { V TO, TP, T3A, T1U, T4y; TO = LD(&(ri[WS(is, 7)]), ivs, &(ri[WS(is, 1)])); TP = LD(&(ri[WS(is, 23)]), ivs, &(ri[WS(is, 1)])); T2H = LD(&(ii[WS(is, 7)]), ivs, &(ii[WS(is, 1)])); T2G = VSUB(TL, TM); TN = VADD(TL, TM); T3A = VSUB(T1S, T1T); T1U = VADD(T1S, T1T); T4y = VSUB(T1q, T1t); T1u = VADD(T1q, T1t); T4N = VADD(T2p, T2q); T2r = VSUB(T2p, T2q); T2s = VSUB(TO, TP); TQ = VADD(TO, TP); T3W = VFMA(LDK(KP414213562), T3z, T3A); T3B = VFNMS(LDK(KP414213562), T3A, T3z); T2Y = VFNMS(LDK(KP414213562), T1R, T1U); T1V = VFMA(LDK(KP414213562), T1U, T1R); T52 = VSUB(T4x, T4y); T4z = VADD(T4x, T4y); T2I = LD(&(ii[WS(is, 23)]), ivs, &(ii[WS(is, 1)])); } } { V TS, TT, T2u, T2v, TV, TW; TS = LD(&(ri[WS(is, 3)]), ivs, &(ri[WS(is, 1)])); TT = LD(&(ri[WS(is, 19)]), ivs, &(ri[WS(is, 1)])); T2u = LD(&(ii[WS(is, 3)]), ivs, &(ii[WS(is, 1)])); T2v = LD(&(ii[WS(is, 19)]), ivs, &(ii[WS(is, 1)])); TV = LD(&(ri[WS(is, 27)]), ivs, &(ri[WS(is, 1)])); TW = LD(&(ri[WS(is, 11)]), ivs, &(ri[WS(is, 1)])); T2A = LD(&(ii[WS(is, 27)]), ivs, &(ii[WS(is, 1)])); T4O = VADD(T2H, T2I); T2J = VSUB(T2H, T2I); T2x = VSUB(TS, TT); TU = VADD(TS, TT); T4T = VADD(T2u, T2v); T2w = VSUB(T2u, T2v); T2z = VSUB(TV, TW); TX = VADD(TV, TW); T2B = LD(&(ii[WS(is, 11)]), ivs, &(ii[WS(is, 1)])); } T3O = VADD(T2s, T2r); T2t = VSUB(T2r, T2s); T3L = VSUB(T2G, T2J); T2K = VADD(T2G, T2J); T4S = VSUB(TN, TQ); TR = VADD(TN, TQ); { V T4P, T4Q, TY, T4U, T2C; T5G = VADD(T4N, T4O); T4P = VSUB(T4N, T4O); T4Q = VSUB(TX, TU); TY = VADD(TU, TX); T4U = VADD(T2A, T2B); T2C = VSUB(T2A, T2B); T2y = VSUB(T2w, T2x); T2L = VADD(T2x, T2w); TZ = VADD(TR, TY); T5F = VSUB(TR, TY); T4V = VSUB(T4T, T4U); T5H = VADD(T4T, T4U); T2D = VADD(T2z, T2C); T2M = VSUB(T2z, T2C); T4R = VSUB(T4P, T4Q); T5k = VADD(T4Q, T4P); } } { V T2f, Ty, T23, T4C, T20, T21, TB, T4D, T2i, T26, TF, T24, TG, TH, T29; V T2a; { V T1Y, T1Z, Tz, TA, T2g, T2h, Tw, Tx, TD, TE; Tw = LD(&(ri[WS(is, 1)]), ivs, &(ri[WS(is, 1)])); Tx = LD(&(ri[WS(is, 17)]), ivs, &(ri[WS(is, 1)])); T5j = VADD(T4S, T4V); T4W = VSUB(T4S, T4V); T5I = VSUB(T5G, T5H); T5X = VADD(T5G, T5H); T2E = VADD(T2y, T2D); T3M = VSUB(T2D, T2y); T2N = VADD(T2L, T2M); T3P = VSUB(T2L, T2M); T2f = VSUB(Tw, Tx); Ty = VADD(Tw, Tx); T1Y = LD(&(ii[WS(is, 1)]), ivs, &(ii[WS(is, 1)])); T1Z = LD(&(ii[WS(is, 17)]), ivs, &(ii[WS(is, 1)])); Tz = LD(&(ri[WS(is, 9)]), ivs, &(ri[WS(is, 1)])); TA = LD(&(ri[WS(is, 25)]), ivs, &(ri[WS(is, 1)])); T2g = LD(&(ii[WS(is, 9)]), ivs, &(ii[WS(is, 1)])); T2h = LD(&(ii[WS(is, 25)]), ivs, &(ii[WS(is, 1)])); TD = LD(&(ri[WS(is, 5)]), ivs, &(ri[WS(is, 1)])); TE = LD(&(ri[WS(is, 21)]), ivs, &(ri[WS(is, 1)])); T23 = LD(&(ii[WS(is, 5)]), ivs, &(ii[WS(is, 1)])); T4C = VADD(T1Y, T1Z); T20 = VSUB(T1Y, T1Z); T21 = VSUB(Tz, TA); TB = VADD(Tz, TA); T4D = VADD(T2g, T2h); T2i = VSUB(T2g, T2h); T26 = VSUB(TD, TE); TF = VADD(TD, TE); T24 = LD(&(ii[WS(is, 21)]), ivs, &(ii[WS(is, 1)])); TG = LD(&(ri[WS(is, 29)]), ivs, &(ri[WS(is, 1)])); TH = LD(&(ri[WS(is, 13)]), ivs, &(ri[WS(is, 1)])); T29 = LD(&(ii[WS(is, 29)]), ivs, &(ii[WS(is, 1)])); T2a = LD(&(ii[WS(is, 13)]), ivs, &(ii[WS(is, 1)])); } { V T4I, T25, T28, TI, T4J, T2b, T4H, TC, T5B, T4E; T3H = VADD(T21, T20); T22 = VSUB(T20, T21); T3E = VSUB(T2f, T2i); T2j = VADD(T2f, T2i); T4I = VADD(T23, T24); T25 = VSUB(T23, T24); T28 = VSUB(TG, TH); TI = VADD(TG, TH); T4J = VADD(T29, T2a); T2b = VSUB(T29, T2a); T4H = VSUB(Ty, TB); TC = VADD(Ty, TB); T5B = VADD(T4C, T4D); T4E = VSUB(T4C, T4D); { V T27, T2k, TJ, T4F, T4K, T5C, T2c, T2l; T27 = VSUB(T25, T26); T2k = VADD(T26, T25); TJ = VADD(TF, TI); T4F = VSUB(TI, TF); T4K = VSUB(T4I, T4J); T5C = VADD(T4I, T4J); T2c = VADD(T28, T2b); T2l = VSUB(T28, T2b); T4G = VSUB(T4E, T4F); T5h = VADD(T4F, T4E); TK = VADD(TC, TJ); T5A = VSUB(TC, TJ); T5D = VSUB(T5B, T5C); T5W = VADD(T5B, T5C); T2d = VADD(T27, T2c); T3F = VSUB(T2c, T27); T4L = VSUB(T4H, T4K); T5g = VADD(T4H, T4K); T3I = VSUB(T2k, T2l); T2m = VADD(T2k, T2l); } } } } } } { V T1v, T1g, T5V, Tv, T60, T5Y, T11, T10; { V T5o, T5n, T5i, T5r, T5f, T5l, T5w, T5u; { V T5d, T4t, T4A, T4X, T58, T51, T4M, T59, T54, T5e, T5b, T4B; T5d = VADD(T4r, T4s); T4t = VSUB(T4r, T4s); T4A = VSUB(T4w, T4z); T5o = VADD(T4w, T4z); T4X = VFNMS(LDK(KP414213562), T4W, T4R); T58 = VFMA(LDK(KP414213562), T4R, T4W); T5n = VADD(T50, T4Z); T51 = VSUB(T4Z, T50); T4M = VFMA(LDK(KP414213562), T4L, T4G); T59 = VFNMS(LDK(KP414213562), T4G, T4L); T54 = VSUB(T52, T53); T5e = VADD(T53, T52); T5b = VFNMS(LDK(KP707106781), T4A, T4t); T4B = VFMA(LDK(KP707106781), T4A, T4t); { V T5s, T56, T4Y, T5c, T5a, T57, T55, T5t; T5i = VFMA(LDK(KP414213562), T5h, T5g); T5s = VFNMS(LDK(KP414213562), T5g, T5h); T56 = VADD(T4M, T4X); T4Y = VSUB(T4M, T4X); T5c = VADD(T59, T58); T5a = VSUB(T58, T59); T57 = VFMA(LDK(KP707106781), T54, T51); T55 = VFNMS(LDK(KP707106781), T54, T51); T5r = VFNMS(LDK(KP707106781), T5e, T5d); T5f = VFMA(LDK(KP707106781), T5e, T5d); T5t = VFMA(LDK(KP414213562), T5j, T5k); T5l = VFNMS(LDK(KP414213562), T5k, T5j); T61 = VFMA(LDK(KP923879532), T4Y, T4B); STM4(&(ro[6]), T61, ovs, &(ro[0])); T62 = VFNMS(LDK(KP923879532), T4Y, T4B); STM4(&(ro[22]), T62, ovs, &(ro[0])); T63 = VFMA(LDK(KP923879532), T5c, T5b); STM4(&(ro[30]), T63, ovs, &(ro[0])); T64 = VFNMS(LDK(KP923879532), T5c, T5b); STM4(&(ro[14]), T64, ovs, &(ro[0])); T65 = VFMA(LDK(KP923879532), T5a, T57); STM4(&(io[6]), T65, ovs, &(io[0])); T66 = VFNMS(LDK(KP923879532), T5a, T57); STM4(&(io[22]), T66, ovs, &(io[0])); T67 = VFMA(LDK(KP923879532), T56, T55); STM4(&(io[30]), T67, ovs, &(io[0])); T68 = VFNMS(LDK(KP923879532), T56, T55); STM4(&(io[14]), T68, ovs, &(io[0])); T5w = VADD(T5s, T5t); T5u = VSUB(T5s, T5t); } } { V Tf, T5P, T5z, T5S, T5U, T5O, T5K, T5L, T5M, Tu, T5T, T5N; { V T5E, T5Q, T5q, T5m, T5v, T5p, T5R, T5J, T5x, T5y; Tf = VADD(T7, Te); T5x = VSUB(T7, Te); T5y = VSUB(T1n, T1u); T1v = VADD(T1n, T1u); T69 = VFMA(LDK(KP923879532), T5u, T5r); STM4(&(ro[10]), T69, ovs, &(ro[0])); T6a = VFNMS(LDK(KP923879532), T5u, T5r); STM4(&(ro[26]), T6a, ovs, &(ro[0])); T5E = VADD(T5A, T5D); T5Q = VSUB(T5D, T5A); T5q = VSUB(T5l, T5i); T5m = VADD(T5i, T5l); T5v = VFMA(LDK(KP707106781), T5o, T5n); T5p = VFNMS(LDK(KP707106781), T5o, T5n); T5P = VSUB(T5x, T5y); T5z = VADD(T5x, T5y); T5R = VADD(T5F, T5I); T5J = VSUB(T5F, T5I); T6b = VFMA(LDK(KP923879532), T5m, T5f); STM4(&(ro[2]), T6b, ovs, &(ro[0])); T6c = VFNMS(LDK(KP923879532), T5m, T5f); STM4(&(ro[18]), T6c, ovs, &(ro[0])); T6d = VFMA(LDK(KP923879532), T5w, T5v); STM4(&(io[2]), T6d, ovs, &(io[0])); T6e = VFNMS(LDK(KP923879532), T5w, T5v); STM4(&(io[18]), T6e, ovs, &(io[0])); T6f = VFMA(LDK(KP923879532), T5q, T5p); STM4(&(io[10]), T6f, ovs, &(io[0])); T6g = VFNMS(LDK(KP923879532), T5q, T5p); STM4(&(io[26]), T6g, ovs, &(io[0])); T5S = VSUB(T5Q, T5R); T5U = VADD(T5Q, T5R); T5O = VSUB(T5J, T5E); T5K = VADD(T5E, T5J); T1g = VADD(T18, T1f); T5L = VSUB(T18, T1f); T5M = VSUB(Tt, Tm); Tu = VADD(Tm, Tt); } T6h = VFMA(LDK(KP707106781), T5S, T5P); STM4(&(ro[12]), T6h, ovs, &(ro[0])); T6i = VFNMS(LDK(KP707106781), T5S, T5P); STM4(&(ro[28]), T6i, ovs, &(ro[0])); T6j = VFMA(LDK(KP707106781), T5K, T5z); STM4(&(ro[4]), T6j, ovs, &(ro[0])); T6k = VFNMS(LDK(KP707106781), T5K, T5z); STM4(&(ro[20]), T6k, ovs, &(ro[0])); T5T = VADD(T5M, T5L); T5N = VSUB(T5L, T5M); T5V = VSUB(Tf, Tu); Tv = VADD(Tf, Tu); T6l = VFMA(LDK(KP707106781), T5U, T5T); STM4(&(io[4]), T6l, ovs, &(io[0])); T6m = VFNMS(LDK(KP707106781), T5U, T5T); STM4(&(io[20]), T6m, ovs, &(io[0])); T6n = VFMA(LDK(KP707106781), T5O, T5N); STM4(&(io[12]), T6n, ovs, &(io[0])); T6o = VFNMS(LDK(KP707106781), T5O, T5N); STM4(&(io[28]), T6o, ovs, &(io[0])); T60 = VADD(T5W, T5X); T5Y = VSUB(T5W, T5X); T11 = VSUB(TZ, TK); T10 = VADD(TK, TZ); } } { V T39, T3k, T3j, T3a, T1X, T37, T33, T31, T3d, T3c, T47, T4i, T4h, T48, T4b; V T4a, T4e, T3N, T41, T3D, T45, T3Z, T38, T36, T32, T2Q, T42, T3K, T3Q, T4d; { V T2e, T2n, T2F, T2O, T1w, T5Z; { V T1H, T1W, T2X, T30; T39 = VFMA(LDK(KP707106781), T1G, T1z); T1H = VFNMS(LDK(KP707106781), T1G, T1z); T1W = VSUB(T1O, T1V); T3k = VADD(T1O, T1V); T3j = VFMA(LDK(KP707106781), T2W, T2T); T2X = VFNMS(LDK(KP707106781), T2W, T2T); T30 = VSUB(T2Y, T2Z); T3a = VADD(T2Z, T2Y); T6p = VSUB(T5V, T5Y); STM4(&(ro[24]), T6p, ovs, &(ro[0])); T6q = VADD(T5V, T5Y); STM4(&(ro[8]), T6q, ovs, &(ro[0])); T6r = VADD(Tv, T10); STM4(&(ro[0]), T6r, ovs, &(ro[0])); T6s = VSUB(Tv, T10); STM4(&(ro[16]), T6s, ovs, &(ro[0])); T1w = VSUB(T1g, T1v); T5Z = VADD(T1g, T1v); T1X = VFMA(LDK(KP923879532), T1W, T1H); T37 = VFNMS(LDK(KP923879532), T1W, T1H); T33 = VFMA(LDK(KP923879532), T30, T2X); T31 = VFNMS(LDK(KP923879532), T30, T2X); } T3d = VFMA(LDK(KP707106781), T2d, T22); T2e = VFNMS(LDK(KP707106781), T2d, T22); T2n = VFNMS(LDK(KP707106781), T2m, T2j); T3c = VFMA(LDK(KP707106781), T2m, T2j); T6t = VADD(T5Z, T60); STM4(&(io[0]), T6t, ovs, &(io[0])); T6u = VSUB(T5Z, T60); STM4(&(io[16]), T6u, ovs, &(io[0])); T6v = VSUB(T1w, T11); STM4(&(io[24]), T6v, ovs, &(io[0])); T6w = VADD(T11, T1w); STM4(&(io[8]), T6w, ovs, &(io[0])); T3g = VFMA(LDK(KP707106781), T2E, T2t); T2F = VFNMS(LDK(KP707106781), T2E, T2t); T2O = VFNMS(LDK(KP707106781), T2N, T2K); T3f = VFMA(LDK(KP707106781), T2N, T2K); { V T3v, T35, T2o, T3C, T3V, T3Y; T47 = VFNMS(LDK(KP707106781), T3u, T3t); T3v = VFMA(LDK(KP707106781), T3u, T3t); T35 = VFNMS(LDK(KP668178637), T2e, T2n); T2o = VFMA(LDK(KP668178637), T2n, T2e); T3C = VSUB(T3y, T3B); T4i = VADD(T3y, T3B); T4h = VFNMS(LDK(KP707106781), T3U, T3T); T3V = VFMA(LDK(KP707106781), T3U, T3T); T3Y = VSUB(T3W, T3X); T48 = VADD(T3X, T3W); { V T3G, T34, T2P, T3J; T4b = VFMA(LDK(KP707106781), T3F, T3E); T3G = VFNMS(LDK(KP707106781), T3F, T3E); T34 = VFMA(LDK(KP668178637), T2F, T2O); T2P = VFNMS(LDK(KP668178637), T2O, T2F); T3J = VFNMS(LDK(KP707106781), T3I, T3H); T4a = VFMA(LDK(KP707106781), T3I, T3H); T4e = VFMA(LDK(KP707106781), T3M, T3L); T3N = VFNMS(LDK(KP707106781), T3M, T3L); T41 = VFNMS(LDK(KP923879532), T3C, T3v); T3D = VFMA(LDK(KP923879532), T3C, T3v); T45 = VFMA(LDK(KP923879532), T3Y, T3V); T3Z = VFNMS(LDK(KP923879532), T3Y, T3V); T38 = VADD(T35, T34); T36 = VSUB(T34, T35); T32 = VADD(T2o, T2P); T2Q = VSUB(T2o, T2P); T42 = VFNMS(LDK(KP668178637), T3G, T3J); T3K = VFMA(LDK(KP668178637), T3J, T3G); T3Q = VFNMS(LDK(KP707106781), T3P, T3O); T4d = VFMA(LDK(KP707106781), T3P, T3O); } } } { V T4n, T4c, T43, T3R, T4m, T4f; T6x = VFMA(LDK(KP831469612), T38, T37); STM4(&(ro[29]), T6x, ovs, &(ro[1])); T6y = VFNMS(LDK(KP831469612), T38, T37); STM4(&(ro[13]), T6y, ovs, &(ro[1])); T6z = VFMA(LDK(KP831469612), T36, T33); STM4(&(io[5]), T6z, ovs, &(io[1])); T6A = VFNMS(LDK(KP831469612), T36, T33); STM4(&(io[21]), T6A, ovs, &(io[1])); T6B = VFMA(LDK(KP831469612), T32, T31); STM4(&(io[29]), T6B, ovs, &(io[1])); T6C = VFNMS(LDK(KP831469612), T32, T31); STM4(&(io[13]), T6C, ovs, &(io[1])); T6D = VFMA(LDK(KP831469612), T2Q, T1X); STM4(&(ro[5]), T6D, ovs, &(ro[1])); T6E = VFNMS(LDK(KP831469612), T2Q, T1X); STM4(&(ro[21]), T6E, ovs, &(ro[1])); T43 = VFMA(LDK(KP668178637), T3N, T3Q); T3R = VFNMS(LDK(KP668178637), T3Q, T3N); { V T44, T46, T40, T3S; T44 = VSUB(T42, T43); T46 = VADD(T42, T43); T40 = VSUB(T3R, T3K); T3S = VADD(T3K, T3R); T4p = VFMA(LDK(KP923879532), T48, T47); T49 = VFNMS(LDK(KP923879532), T48, T47); T4l = VFNMS(LDK(KP923879532), T4i, T4h); T4j = VFMA(LDK(KP923879532), T4i, T4h); T4n = VFNMS(LDK(KP198912367), T4a, T4b); T4c = VFMA(LDK(KP198912367), T4b, T4a); T6F = VFMA(LDK(KP831469612), T44, T41); STM4(&(ro[11]), T6F, ovs, &(ro[1])); T6G = VFNMS(LDK(KP831469612), T44, T41); STM4(&(ro[27]), T6G, ovs, &(ro[1])); T6H = VFMA(LDK(KP831469612), T46, T45); STM4(&(io[3]), T6H, ovs, &(io[1])); T6I = VFNMS(LDK(KP831469612), T46, T45); STM4(&(io[19]), T6I, ovs, &(io[1])); T6J = VFMA(LDK(KP831469612), T40, T3Z); STM4(&(io[11]), T6J, ovs, &(io[1])); T6K = VFNMS(LDK(KP831469612), T40, T3Z); STM4(&(io[27]), T6K, ovs, &(io[1])); T6L = VFMA(LDK(KP831469612), T3S, T3D); STM4(&(ro[3]), T6L, ovs, &(ro[1])); T6M = VFNMS(LDK(KP831469612), T3S, T3D); STM4(&(ro[19]), T6M, ovs, &(ro[1])); } T4m = VFMA(LDK(KP198912367), T4d, T4e); T4f = VFNMS(LDK(KP198912367), T4e, T4d); T3n = VFNMS(LDK(KP923879532), T3a, T39); T3b = VFMA(LDK(KP923879532), T3a, T39); T3r = VFMA(LDK(KP923879532), T3k, T3j); T3l = VFNMS(LDK(KP923879532), T3k, T3j); T3o = VFNMS(LDK(KP198912367), T3c, T3d); T3e = VFMA(LDK(KP198912367), T3d, T3c); T4q = VADD(T4n, T4m); T4o = VSUB(T4m, T4n); T4k = VADD(T4c, T4f); T4g = VSUB(T4c, T4f); } } } } { V T6N, T6O, T6P, T6Q; T6N = VFMA(LDK(KP980785280), T4q, T4p); STM4(&(ro[31]), T6N, ovs, &(ro[1])); STN4(&(ro[28]), T6i, T6x, T63, T6N, ovs); T6O = VFNMS(LDK(KP980785280), T4q, T4p); STM4(&(ro[15]), T6O, ovs, &(ro[1])); STN4(&(ro[12]), T6h, T6y, T64, T6O, ovs); T6P = VFMA(LDK(KP980785280), T4o, T4l); STM4(&(io[7]), T6P, ovs, &(io[1])); STN4(&(io[4]), T6l, T6z, T65, T6P, ovs); T6Q = VFNMS(LDK(KP980785280), T4o, T4l); STM4(&(io[23]), T6Q, ovs, &(io[1])); STN4(&(io[20]), T6m, T6A, T66, T6Q, ovs); { V T6R, T6S, T6T, T6U; T6R = VFMA(LDK(KP980785280), T4k, T4j); STM4(&(io[31]), T6R, ovs, &(io[1])); STN4(&(io[28]), T6o, T6B, T67, T6R, ovs); T6S = VFNMS(LDK(KP980785280), T4k, T4j); STM4(&(io[15]), T6S, ovs, &(io[1])); STN4(&(io[12]), T6n, T6C, T68, T6S, ovs); T6T = VFMA(LDK(KP980785280), T4g, T49); STM4(&(ro[7]), T6T, ovs, &(ro[1])); STN4(&(ro[4]), T6j, T6D, T61, T6T, ovs); T6U = VFNMS(LDK(KP980785280), T4g, T49); STM4(&(ro[23]), T6U, ovs, &(ro[1])); STN4(&(ro[20]), T6k, T6E, T62, T6U, ovs); T3h = VFNMS(LDK(KP198912367), T3g, T3f); T3p = VFMA(LDK(KP198912367), T3f, T3g); } } { V T3s, T3q, T3i, T3m; T3s = VADD(T3o, T3p); T3q = VSUB(T3o, T3p); T3i = VADD(T3e, T3h); T3m = VSUB(T3h, T3e); { V T6V, T6W, T6X, T6Y; T6V = VFMA(LDK(KP980785280), T3q, T3n); STM4(&(ro[9]), T6V, ovs, &(ro[1])); STN4(&(ro[8]), T6q, T6V, T69, T6F, ovs); T6W = VFNMS(LDK(KP980785280), T3q, T3n); STM4(&(ro[25]), T6W, ovs, &(ro[1])); STN4(&(ro[24]), T6p, T6W, T6a, T6G, ovs); T6X = VFMA(LDK(KP980785280), T3s, T3r); STM4(&(io[1]), T6X, ovs, &(io[1])); STN4(&(io[0]), T6t, T6X, T6d, T6H, ovs); T6Y = VFNMS(LDK(KP980785280), T3s, T3r); STM4(&(io[17]), T6Y, ovs, &(io[1])); STN4(&(io[16]), T6u, T6Y, T6e, T6I, ovs); { V T6Z, T70, T71, T72; T6Z = VFMA(LDK(KP980785280), T3m, T3l); STM4(&(io[9]), T6Z, ovs, &(io[1])); STN4(&(io[8]), T6w, T6Z, T6f, T6J, ovs); T70 = VFNMS(LDK(KP980785280), T3m, T3l); STM4(&(io[25]), T70, ovs, &(io[1])); STN4(&(io[24]), T6v, T70, T6g, T6K, ovs); T71 = VFMA(LDK(KP980785280), T3i, T3b); STM4(&(ro[1]), T71, ovs, &(ro[1])); STN4(&(ro[0]), T6r, T71, T6b, T6L, ovs); T72 = VFNMS(LDK(KP980785280), T3i, T3b); STM4(&(ro[17]), T72, ovs, &(ro[1])); STN4(&(ro[16]), T6s, T72, T6c, T6M, ovs); } } } } } VLEAVE(); } static const kdft_desc desc = { 32, XSIMD_STRING("n2sv_32"), {236, 0, 136, 0}, &GENUS, 0, 1, 0, 0 }; void XSIMD(codelet_n2sv_32) (planner *p) { X(kdft_register) (p, n2sv_32, &desc); } #else /* HAVE_FMA */ /* Generated by: ../../../genfft/gen_notw.native -simd -compact -variables 4 -pipeline-latency 8 -n 32 -name n2sv_32 -with-ostride 1 -include n2s.h -store-multiple 4 */ /* * This function contains 372 FP additions, 84 FP multiplications, * (or, 340 additions, 52 multiplications, 32 fused multiply/add), * 130 stack variables, 7 constants, and 144 memory accesses */ #include "n2s.h" static void n2sv_32(const R *ri, const R *ii, R *ro, R *io, stride is, stride os, INT v, INT ivs, INT ovs) { DVK(KP831469612, +0.831469612302545237078788377617905756738560812); DVK(KP555570233, +0.555570233019602224742830813948532874374937191); DVK(KP195090322, +0.195090322016128267848284868477022240927691618); DVK(KP980785280, +0.980785280403230449126182236134239036973933731); DVK(KP923879532, +0.923879532511286756128183189396788286822416626); DVK(KP382683432, +0.382683432365089771728459984030398866761344562); DVK(KP707106781, +0.707106781186547524400844362104849039284835938); { INT i; for (i = v; i > 0; i = i - (2 * VL), ri = ri + ((2 * VL) * ivs), ii = ii + ((2 * VL) * ivs), ro = ro + ((2 * VL) * ovs), io = io + ((2 * VL) * ovs), MAKE_VOLATILE_STRIDE(128, is), MAKE_VOLATILE_STRIDE(128, os)) { V T7, T4r, T4Z, T18, T1z, T3t, T3T, T2T, Te, T1f, T50, T4s, T2W, T3u, T1G; V T3U, Tm, T1n, T1O, T2Z, T3y, T3X, T4w, T53, Tt, T1u, T1V, T2Y, T3B, T3W; V T4z, T52, T2t, T3L, T3O, T2K, TR, TY, T5F, T5G, T5H, T5I, T4R, T5j, T2E; V T3P, T4W, T5k, T2N, T3M, T22, T3E, T3H, T2j, TC, TJ, T5A, T5B, T5C, T5D; V T4G, T5g, T2d, T3F, T4L, T5h, T2m, T3I; { V T3, T1x, T14, T2S, T6, T2R, T17, T1y; { V T1, T2, T12, T13; T1 = LD(&(ri[0]), ivs, &(ri[0])); T2 = LD(&(ri[WS(is, 16)]), ivs, &(ri[0])); T3 = VADD(T1, T2); T1x = VSUB(T1, T2); T12 = LD(&(ii[0]), ivs, &(ii[0])); T13 = LD(&(ii[WS(is, 16)]), ivs, &(ii[0])); T14 = VADD(T12, T13); T2S = VSUB(T12, T13); } { V T4, T5, T15, T16; T4 = LD(&(ri[WS(is, 8)]), ivs, &(ri[0])); T5 = LD(&(ri[WS(is, 24)]), ivs, &(ri[0])); T6 = VADD(T4, T5); T2R = VSUB(T4, T5); T15 = LD(&(ii[WS(is, 8)]), ivs, &(ii[0])); T16 = LD(&(ii[WS(is, 24)]), ivs, &(ii[0])); T17 = VADD(T15, T16); T1y = VSUB(T15, T16); } T7 = VADD(T3, T6); T4r = VSUB(T3, T6); T4Z = VSUB(T14, T17); T18 = VADD(T14, T17); T1z = VSUB(T1x, T1y); T3t = VADD(T1x, T1y); T3T = VSUB(T2S, T2R); T2T = VADD(T2R, T2S); } { V Ta, T1B, T1b, T1A, Td, T1D, T1e, T1E; { V T8, T9, T19, T1a; T8 = LD(&(ri[WS(is, 4)]), ivs, &(ri[0])); T9 = LD(&(ri[WS(is, 20)]), ivs, &(ri[0])); Ta = VADD(T8, T9); T1B = VSUB(T8, T9); T19 = LD(&(ii[WS(is, 4)]), ivs, &(ii[0])); T1a = LD(&(ii[WS(is, 20)]), ivs, &(ii[0])); T1b = VADD(T19, T1a); T1A = VSUB(T19, T1a); } { V Tb, Tc, T1c, T1d; Tb = LD(&(ri[WS(is, 28)]), ivs, &(ri[0])); Tc = LD(&(ri[WS(is, 12)]), ivs, &(ri[0])); Td = VADD(Tb, Tc); T1D = VSUB(Tb, Tc); T1c = LD(&(ii[WS(is, 28)]), ivs, &(ii[0])); T1d = LD(&(ii[WS(is, 12)]), ivs, &(ii[0])); T1e = VADD(T1c, T1d); T1E = VSUB(T1c, T1d); } Te = VADD(Ta, Td); T1f = VADD(T1b, T1e); T50 = VSUB(Td, Ta); T4s = VSUB(T1b, T1e); { V T2U, T2V, T1C, T1F; T2U = VSUB(T1D, T1E); T2V = VADD(T1B, T1A); T2W = VMUL(LDK(KP707106781), VSUB(T2U, T2V)); T3u = VMUL(LDK(KP707106781), VADD(T2V, T2U)); T1C = VSUB(T1A, T1B); T1F = VADD(T1D, T1E); T1G = VMUL(LDK(KP707106781), VSUB(T1C, T1F)); T3U = VMUL(LDK(KP707106781), VADD(T1C, T1F)); } } { V Ti, T1L, T1j, T1J, Tl, T1I, T1m, T1M, T1K, T1N; { V Tg, Th, T1h, T1i; Tg = LD(&(ri[WS(is, 2)]), ivs, &(ri[0])); Th = LD(&(ri[WS(is, 18)]), ivs, &(ri[0])); Ti = VADD(Tg, Th); T1L = VSUB(Tg, Th); T1h = LD(&(ii[WS(is, 2)]), ivs, &(ii[0])); T1i = LD(&(ii[WS(is, 18)]), ivs, &(ii[0])); T1j = VADD(T1h, T1i); T1J = VSUB(T1h, T1i); } { V Tj, Tk, T1k, T1l; Tj = LD(&(ri[WS(is, 10)]), ivs, &(ri[0])); Tk = LD(&(ri[WS(is, 26)]), ivs, &(ri[0])); Tl = VADD(Tj, Tk); T1I = VSUB(Tj, Tk); T1k = LD(&(ii[WS(is, 10)]), ivs, &(ii[0])); T1l = LD(&(ii[WS(is, 26)]), ivs, &(ii[0])); T1m = VADD(T1k, T1l); T1M = VSUB(T1k, T1l); } Tm = VADD(Ti, Tl); T1n = VADD(T1j, T1m); T1K = VADD(T1I, T1J); T1N = VSUB(T1L, T1M); T1O = VFNMS(LDK(KP923879532), T1N, VMUL(LDK(KP382683432), T1K)); T2Z = VFMA(LDK(KP923879532), T1K, VMUL(LDK(KP382683432), T1N)); { V T3w, T3x, T4u, T4v; T3w = VSUB(T1J, T1I); T3x = VADD(T1L, T1M); T3y = VFNMS(LDK(KP382683432), T3x, VMUL(LDK(KP923879532), T3w)); T3X = VFMA(LDK(KP382683432), T3w, VMUL(LDK(KP923879532), T3x)); T4u = VSUB(T1j, T1m); T4v = VSUB(Ti, Tl); T4w = VSUB(T4u, T4v); T53 = VADD(T4v, T4u); } } { V Tp, T1S, T1q, T1Q, Ts, T1P, T1t, T1T, T1R, T1U; { V Tn, To, T1o, T1p; Tn = LD(&(ri[WS(is, 30)]), ivs, &(ri[0])); To = LD(&(ri[WS(is, 14)]), ivs, &(ri[0])); Tp = VADD(Tn, To); T1S = VSUB(Tn, To); T1o = LD(&(ii[WS(is, 30)]), ivs, &(ii[0])); T1p = LD(&(ii[WS(is, 14)]), ivs, &(ii[0])); T1q = VADD(T1o, T1p); T1Q = VSUB(T1o, T1p); } { V Tq, Tr, T1r, T1s; Tq = LD(&(ri[WS(is, 6)]), ivs, &(ri[0])); Tr = LD(&(ri[WS(is, 22)]), ivs, &(ri[0])); Ts = VADD(Tq, Tr); T1P = VSUB(Tq, Tr); T1r = LD(&(ii[WS(is, 6)]), ivs, &(ii[0])); T1s = LD(&(ii[WS(is, 22)]), ivs, &(ii[0])); T1t = VADD(T1r, T1s); T1T = VSUB(T1r, T1s); } Tt = VADD(Tp, Ts); T1u = VADD(T1q, T1t); T1R = VADD(T1P, T1Q); T1U = VSUB(T1S, T1T); T1V = VFMA(LDK(KP382683432), T1R, VMUL(LDK(KP923879532), T1U)); T2Y = VFNMS(LDK(KP923879532), T1R, VMUL(LDK(KP382683432), T1U)); { V T3z, T3A, T4x, T4y; T3z = VSUB(T1Q, T1P); T3A = VADD(T1S, T1T); T3B = VFMA(LDK(KP923879532), T3z, VMUL(LDK(KP382683432), T3A)); T3W = VFNMS(LDK(KP382683432), T3z, VMUL(LDK(KP923879532), T3A)); T4x = VSUB(Tp, Ts); T4y = VSUB(T1q, T1t); T4z = VADD(T4x, T4y); T52 = VSUB(T4x, T4y); } } { V TN, T2p, T2J, T4S, TQ, T2G, T2s, T4T, TU, T2x, T2w, T4O, TX, T2z, T2C; V T4P; { V TL, TM, T2H, T2I; TL = LD(&(ri[WS(is, 31)]), ivs, &(ri[WS(is, 1)])); TM = LD(&(ri[WS(is, 15)]), ivs, &(ri[WS(is, 1)])); TN = VADD(TL, TM); T2p = VSUB(TL, TM); T2H = LD(&(ii[WS(is, 31)]), ivs, &(ii[WS(is, 1)])); T2I = LD(&(ii[WS(is, 15)]), ivs, &(ii[WS(is, 1)])); T2J = VSUB(T2H, T2I); T4S = VADD(T2H, T2I); } { V TO, TP, T2q, T2r; TO = LD(&(ri[WS(is, 7)]), ivs, &(ri[WS(is, 1)])); TP = LD(&(ri[WS(is, 23)]), ivs, &(ri[WS(is, 1)])); TQ = VADD(TO, TP); T2G = VSUB(TO, TP); T2q = LD(&(ii[WS(is, 7)]), ivs, &(ii[WS(is, 1)])); T2r = LD(&(ii[WS(is, 23)]), ivs, &(ii[WS(is, 1)])); T2s = VSUB(T2q, T2r); T4T = VADD(T2q, T2r); } { V TS, TT, T2u, T2v; TS = LD(&(ri[WS(is, 3)]), ivs, &(ri[WS(is, 1)])); TT = LD(&(ri[WS(is, 19)]), ivs, &(ri[WS(is, 1)])); TU = VADD(TS, TT); T2x = VSUB(TS, TT); T2u = LD(&(ii[WS(is, 3)]), ivs, &(ii[WS(is, 1)])); T2v = LD(&(ii[WS(is, 19)]), ivs, &(ii[WS(is, 1)])); T2w = VSUB(T2u, T2v); T4O = VADD(T2u, T2v); } { V TV, TW, T2A, T2B; TV = LD(&(ri[WS(is, 27)]), ivs, &(ri[WS(is, 1)])); TW = LD(&(ri[WS(is, 11)]), ivs, &(ri[WS(is, 1)])); TX = VADD(TV, TW); T2z = VSUB(TV, TW); T2A = LD(&(ii[WS(is, 27)]), ivs, &(ii[WS(is, 1)])); T2B = LD(&(ii[WS(is, 11)]), ivs, &(ii[WS(is, 1)])); T2C = VSUB(T2A, T2B); T4P = VADD(T2A, T2B); } T2t = VSUB(T2p, T2s); T3L = VADD(T2p, T2s); T3O = VSUB(T2J, T2G); T2K = VADD(T2G, T2J); TR = VADD(TN, TQ); TY = VADD(TU, TX); T5F = VSUB(TR, TY); { V T4N, T4Q, T2y, T2D; T5G = VADD(T4S, T4T); T5H = VADD(T4O, T4P); T5I = VSUB(T5G, T5H); T4N = VSUB(TN, TQ); T4Q = VSUB(T4O, T4P); T4R = VSUB(T4N, T4Q); T5j = VADD(T4N, T4Q); T2y = VSUB(T2w, T2x); T2D = VADD(T2z, T2C); T2E = VMUL(LDK(KP707106781), VSUB(T2y, T2D)); T3P = VMUL(LDK(KP707106781), VADD(T2y, T2D)); { V T4U, T4V, T2L, T2M; T4U = VSUB(T4S, T4T); T4V = VSUB(TX, TU); T4W = VSUB(T4U, T4V); T5k = VADD(T4V, T4U); T2L = VSUB(T2z, T2C); T2M = VADD(T2x, T2w); T2N = VMUL(LDK(KP707106781), VSUB(T2L, T2M)); T3M = VMUL(LDK(KP707106781), VADD(T2M, T2L)); } } } { V Ty, T2f, T21, T4C, TB, T1Y, T2i, T4D, TF, T28, T2b, T4I, TI, T23, T26; V T4J; { V Tw, Tx, T1Z, T20; Tw = LD(&(ri[WS(is, 1)]), ivs, &(ri[WS(is, 1)])); Tx = LD(&(ri[WS(is, 17)]), ivs, &(ri[WS(is, 1)])); Ty = VADD(Tw, Tx); T2f = VSUB(Tw, Tx); T1Z = LD(&(ii[WS(is, 1)]), ivs, &(ii[WS(is, 1)])); T20 = LD(&(ii[WS(is, 17)]), ivs, &(ii[WS(is, 1)])); T21 = VSUB(T1Z, T20); T4C = VADD(T1Z, T20); } { V Tz, TA, T2g, T2h; Tz = LD(&(ri[WS(is, 9)]), ivs, &(ri[WS(is, 1)])); TA = LD(&(ri[WS(is, 25)]), ivs, &(ri[WS(is, 1)])); TB = VADD(Tz, TA); T1Y = VSUB(Tz, TA); T2g = LD(&(ii[WS(is, 9)]), ivs, &(ii[WS(is, 1)])); T2h = LD(&(ii[WS(is, 25)]), ivs, &(ii[WS(is, 1)])); T2i = VSUB(T2g, T2h); T4D = VADD(T2g, T2h); } { V TD, TE, T29, T2a; TD = LD(&(ri[WS(is, 5)]), ivs, &(ri[WS(is, 1)])); TE = LD(&(ri[WS(is, 21)]), ivs, &(ri[WS(is, 1)])); TF = VADD(TD, TE); T28 = VSUB(TD, TE); T29 = LD(&(ii[WS(is, 5)]), ivs, &(ii[WS(is, 1)])); T2a = LD(&(ii[WS(is, 21)]), ivs, &(ii[WS(is, 1)])); T2b = VSUB(T29, T2a); T4I = VADD(T29, T2a); } { V TG, TH, T24, T25; TG = LD(&(ri[WS(is, 29)]), ivs, &(ri[WS(is, 1)])); TH = LD(&(ri[WS(is, 13)]), ivs, &(ri[WS(is, 1)])); TI = VADD(TG, TH); T23 = VSUB(TG, TH); T24 = LD(&(ii[WS(is, 29)]), ivs, &(ii[WS(is, 1)])); T25 = LD(&(ii[WS(is, 13)]), ivs, &(ii[WS(is, 1)])); T26 = VSUB(T24, T25); T4J = VADD(T24, T25); } T22 = VADD(T1Y, T21); T3E = VADD(T2f, T2i); T3H = VSUB(T21, T1Y); T2j = VSUB(T2f, T2i); TC = VADD(Ty, TB); TJ = VADD(TF, TI); T5A = VSUB(TC, TJ); { V T4E, T4F, T27, T2c; T5B = VADD(T4C, T4D); T5C = VADD(T4I, T4J); T5D = VSUB(T5B, T5C); T4E = VSUB(T4C, T4D); T4F = VSUB(TI, TF); T4G = VSUB(T4E, T4F); T5g = VADD(T4F, T4E); T27 = VSUB(T23, T26); T2c = VADD(T28, T2b); T2d = VMUL(LDK(KP707106781), VSUB(T27, T2c)); T3F = VMUL(LDK(KP707106781), VADD(T2c, T27)); { V T4H, T4K, T2k, T2l; T4H = VSUB(Ty, TB); T4K = VSUB(T4I, T4J); T4L = VSUB(T4H, T4K); T5h = VADD(T4H, T4K); T2k = VSUB(T2b, T28); T2l = VADD(T23, T26); T2m = VMUL(LDK(KP707106781), VSUB(T2k, T2l)); T3I = VMUL(LDK(KP707106781), VADD(T2k, T2l)); } } } { V T61, T62, T63, T64, T65, T66, T67, T68, T69, T6a, T6b, T6c, T6d, T6e, T6f; V T6g, T6h, T6i, T6j, T6k, T6l, T6m, T6n, T6o, T6p, T6q, T6r, T6s, T6t, T6u; V T6v, T6w; { V T4B, T57, T5a, T5c, T4Y, T56, T55, T5b; { V T4t, T4A, T58, T59; T4t = VSUB(T4r, T4s); T4A = VMUL(LDK(KP707106781), VSUB(T4w, T4z)); T4B = VADD(T4t, T4A); T57 = VSUB(T4t, T4A); T58 = VFNMS(LDK(KP923879532), T4L, VMUL(LDK(KP382683432), T4G)); T59 = VFMA(LDK(KP382683432), T4W, VMUL(LDK(KP923879532), T4R)); T5a = VSUB(T58, T59); T5c = VADD(T58, T59); } { V T4M, T4X, T51, T54; T4M = VFMA(LDK(KP923879532), T4G, VMUL(LDK(KP382683432), T4L)); T4X = VFNMS(LDK(KP923879532), T4W, VMUL(LDK(KP382683432), T4R)); T4Y = VADD(T4M, T4X); T56 = VSUB(T4X, T4M); T51 = VSUB(T4Z, T50); T54 = VMUL(LDK(KP707106781), VSUB(T52, T53)); T55 = VSUB(T51, T54); T5b = VADD(T51, T54); } T61 = VSUB(T4B, T4Y); STM4(&(ro[22]), T61, ovs, &(ro[0])); T62 = VSUB(T5b, T5c); STM4(&(io[22]), T62, ovs, &(io[0])); T63 = VADD(T4B, T4Y); STM4(&(ro[6]), T63, ovs, &(ro[0])); T64 = VADD(T5b, T5c); STM4(&(io[6]), T64, ovs, &(io[0])); T65 = VSUB(T55, T56); STM4(&(io[30]), T65, ovs, &(io[0])); T66 = VSUB(T57, T5a); STM4(&(ro[30]), T66, ovs, &(ro[0])); T67 = VADD(T55, T56); STM4(&(io[14]), T67, ovs, &(io[0])); T68 = VADD(T57, T5a); STM4(&(ro[14]), T68, ovs, &(ro[0])); } { V T5f, T5r, T5u, T5w, T5m, T5q, T5p, T5v; { V T5d, T5e, T5s, T5t; T5d = VADD(T4r, T4s); T5e = VMUL(LDK(KP707106781), VADD(T53, T52)); T5f = VADD(T5d, T5e); T5r = VSUB(T5d, T5e); T5s = VFNMS(LDK(KP382683432), T5h, VMUL(LDK(KP923879532), T5g)); T5t = VFMA(LDK(KP923879532), T5k, VMUL(LDK(KP382683432), T5j)); T5u = VSUB(T5s, T5t); T5w = VADD(T5s, T5t); } { V T5i, T5l, T5n, T5o; T5i = VFMA(LDK(KP382683432), T5g, VMUL(LDK(KP923879532), T5h)); T5l = VFNMS(LDK(KP382683432), T5k, VMUL(LDK(KP923879532), T5j)); T5m = VADD(T5i, T5l); T5q = VSUB(T5l, T5i); T5n = VADD(T50, T4Z); T5o = VMUL(LDK(KP707106781), VADD(T4w, T4z)); T5p = VSUB(T5n, T5o); T5v = VADD(T5n, T5o); } T69 = VSUB(T5f, T5m); STM4(&(ro[18]), T69, ovs, &(ro[0])); T6a = VSUB(T5v, T5w); STM4(&(io[18]), T6a, ovs, &(io[0])); T6b = VADD(T5f, T5m); STM4(&(ro[2]), T6b, ovs, &(ro[0])); T6c = VADD(T5v, T5w); STM4(&(io[2]), T6c, ovs, &(io[0])); T6d = VSUB(T5p, T5q); STM4(&(io[26]), T6d, ovs, &(io[0])); T6e = VSUB(T5r, T5u); STM4(&(ro[26]), T6e, ovs, &(ro[0])); T6f = VADD(T5p, T5q); STM4(&(io[10]), T6f, ovs, &(io[0])); T6g = VADD(T5r, T5u); STM4(&(ro[10]), T6g, ovs, &(ro[0])); } { V T5z, T5P, T5S, T5U, T5K, T5O, T5N, T5T; { V T5x, T5y, T5Q, T5R; T5x = VSUB(T7, Te); T5y = VSUB(T1n, T1u); T5z = VADD(T5x, T5y); T5P = VSUB(T5x, T5y); T5Q = VSUB(T5D, T5A); T5R = VADD(T5F, T5I); T5S = VMUL(LDK(KP707106781), VSUB(T5Q, T5R)); T5U = VMUL(LDK(KP707106781), VADD(T5Q, T5R)); } { V T5E, T5J, T5L, T5M; T5E = VADD(T5A, T5D); T5J = VSUB(T5F, T5I); T5K = VMUL(LDK(KP707106781), VADD(T5E, T5J)); T5O = VMUL(LDK(KP707106781), VSUB(T5J, T5E)); T5L = VSUB(T18, T1f); T5M = VSUB(Tt, Tm); T5N = VSUB(T5L, T5M); T5T = VADD(T5M, T5L); } T6h = VSUB(T5z, T5K); STM4(&(ro[20]), T6h, ovs, &(ro[0])); T6i = VSUB(T5T, T5U); STM4(&(io[20]), T6i, ovs, &(io[0])); T6j = VADD(T5z, T5K); STM4(&(ro[4]), T6j, ovs, &(ro[0])); T6k = VADD(T5T, T5U); STM4(&(io[4]), T6k, ovs, &(io[0])); T6l = VSUB(T5N, T5O); STM4(&(io[28]), T6l, ovs, &(io[0])); T6m = VSUB(T5P, T5S); STM4(&(ro[28]), T6m, ovs, &(ro[0])); T6n = VADD(T5N, T5O); STM4(&(io[12]), T6n, ovs, &(io[0])); T6o = VADD(T5P, T5S); STM4(&(ro[12]), T6o, ovs, &(ro[0])); } { V Tv, T5V, T5Y, T60, T10, T11, T1w, T5Z; { V Tf, Tu, T5W, T5X; Tf = VADD(T7, Te); Tu = VADD(Tm, Tt); Tv = VADD(Tf, Tu); T5V = VSUB(Tf, Tu); T5W = VADD(T5B, T5C); T5X = VADD(T5G, T5H); T5Y = VSUB(T5W, T5X); T60 = VADD(T5W, T5X); } { V TK, TZ, T1g, T1v; TK = VADD(TC, TJ); TZ = VADD(TR, TY); T10 = VADD(TK, TZ); T11 = VSUB(TZ, TK); T1g = VADD(T18, T1f); T1v = VADD(T1n, T1u); T1w = VSUB(T1g, T1v); T5Z = VADD(T1g, T1v); } T6p = VSUB(Tv, T10); STM4(&(ro[16]), T6p, ovs, &(ro[0])); T6q = VSUB(T5Z, T60); STM4(&(io[16]), T6q, ovs, &(io[0])); T6r = VADD(Tv, T10); STM4(&(ro[0]), T6r, ovs, &(ro[0])); T6s = VADD(T5Z, T60); STM4(&(io[0]), T6s, ovs, &(io[0])); T6t = VADD(T11, T1w); STM4(&(io[8]), T6t, ovs, &(io[0])); T6u = VADD(T5V, T5Y); STM4(&(ro[8]), T6u, ovs, &(ro[0])); T6v = VSUB(T1w, T11); STM4(&(io[24]), T6v, ovs, &(io[0])); T6w = VSUB(T5V, T5Y); STM4(&(ro[24]), T6w, ovs, &(ro[0])); } { V T6x, T6y, T6z, T6A, T6B, T6C, T6D, T6E; { V T1X, T33, T31, T37, T2o, T34, T2P, T35; { V T1H, T1W, T2X, T30; T1H = VSUB(T1z, T1G); T1W = VSUB(T1O, T1V); T1X = VADD(T1H, T1W); T33 = VSUB(T1H, T1W); T2X = VSUB(T2T, T2W); T30 = VSUB(T2Y, T2Z); T31 = VSUB(T2X, T30); T37 = VADD(T2X, T30); } { V T2e, T2n, T2F, T2O; T2e = VSUB(T22, T2d); T2n = VSUB(T2j, T2m); T2o = VFMA(LDK(KP980785280), T2e, VMUL(LDK(KP195090322), T2n)); T34 = VFNMS(LDK(KP980785280), T2n, VMUL(LDK(KP195090322), T2e)); T2F = VSUB(T2t, T2E); T2O = VSUB(T2K, T2N); T2P = VFNMS(LDK(KP980785280), T2O, VMUL(LDK(KP195090322), T2F)); T35 = VFMA(LDK(KP195090322), T2O, VMUL(LDK(KP980785280), T2F)); } { V T2Q, T38, T32, T36; T2Q = VADD(T2o, T2P); T6x = VSUB(T1X, T2Q); STM4(&(ro[23]), T6x, ovs, &(ro[1])); T6y = VADD(T1X, T2Q); STM4(&(ro[7]), T6y, ovs, &(ro[1])); T38 = VADD(T34, T35); T6z = VSUB(T37, T38); STM4(&(io[23]), T6z, ovs, &(io[1])); T6A = VADD(T37, T38); STM4(&(io[7]), T6A, ovs, &(io[1])); T32 = VSUB(T2P, T2o); T6B = VSUB(T31, T32); STM4(&(io[31]), T6B, ovs, &(io[1])); T6C = VADD(T31, T32); STM4(&(io[15]), T6C, ovs, &(io[1])); T36 = VSUB(T34, T35); T6D = VSUB(T33, T36); STM4(&(ro[31]), T6D, ovs, &(ro[1])); T6E = VADD(T33, T36); STM4(&(ro[15]), T6E, ovs, &(ro[1])); } } { V T3D, T41, T3Z, T45, T3K, T42, T3R, T43; { V T3v, T3C, T3V, T3Y; T3v = VSUB(T3t, T3u); T3C = VSUB(T3y, T3B); T3D = VADD(T3v, T3C); T41 = VSUB(T3v, T3C); T3V = VSUB(T3T, T3U); T3Y = VSUB(T3W, T3X); T3Z = VSUB(T3V, T3Y); T45 = VADD(T3V, T3Y); } { V T3G, T3J, T3N, T3Q; T3G = VSUB(T3E, T3F); T3J = VSUB(T3H, T3I); T3K = VFMA(LDK(KP555570233), T3G, VMUL(LDK(KP831469612), T3J)); T42 = VFNMS(LDK(KP831469612), T3G, VMUL(LDK(KP555570233), T3J)); T3N = VSUB(T3L, T3M); T3Q = VSUB(T3O, T3P); T3R = VFNMS(LDK(KP831469612), T3Q, VMUL(LDK(KP555570233), T3N)); T43 = VFMA(LDK(KP831469612), T3N, VMUL(LDK(KP555570233), T3Q)); } { V T3S, T6F, T6G, T46, T6H, T6I; T3S = VADD(T3K, T3R); T6F = VSUB(T3D, T3S); STM4(&(ro[21]), T6F, ovs, &(ro[1])); STN4(&(ro[20]), T6h, T6F, T61, T6x, ovs); T6G = VADD(T3D, T3S); STM4(&(ro[5]), T6G, ovs, &(ro[1])); STN4(&(ro[4]), T6j, T6G, T63, T6y, ovs); T46 = VADD(T42, T43); T6H = VSUB(T45, T46); STM4(&(io[21]), T6H, ovs, &(io[1])); STN4(&(io[20]), T6i, T6H, T62, T6z, ovs); T6I = VADD(T45, T46); STM4(&(io[5]), T6I, ovs, &(io[1])); STN4(&(io[4]), T6k, T6I, T64, T6A, ovs); } { V T40, T6J, T6K, T44, T6L, T6M; T40 = VSUB(T3R, T3K); T6J = VSUB(T3Z, T40); STM4(&(io[29]), T6J, ovs, &(io[1])); STN4(&(io[28]), T6l, T6J, T65, T6B, ovs); T6K = VADD(T3Z, T40); STM4(&(io[13]), T6K, ovs, &(io[1])); STN4(&(io[12]), T6n, T6K, T67, T6C, ovs); T44 = VSUB(T42, T43); T6L = VSUB(T41, T44); STM4(&(ro[29]), T6L, ovs, &(ro[1])); STN4(&(ro[28]), T6m, T6L, T66, T6D, ovs); T6M = VADD(T41, T44); STM4(&(ro[13]), T6M, ovs, &(ro[1])); STN4(&(ro[12]), T6o, T6M, T68, T6E, ovs); } } } { V T6N, T6O, T6P, T6Q, T6R, T6S, T6T, T6U; { V T49, T4l, T4j, T4p, T4c, T4m, T4f, T4n; { V T47, T48, T4h, T4i; T47 = VADD(T3t, T3u); T48 = VADD(T3X, T3W); T49 = VADD(T47, T48); T4l = VSUB(T47, T48); T4h = VADD(T3T, T3U); T4i = VADD(T3y, T3B); T4j = VSUB(T4h, T4i); T4p = VADD(T4h, T4i); } { V T4a, T4b, T4d, T4e; T4a = VADD(T3E, T3F); T4b = VADD(T3H, T3I); T4c = VFMA(LDK(KP980785280), T4a, VMUL(LDK(KP195090322), T4b)); T4m = VFNMS(LDK(KP195090322), T4a, VMUL(LDK(KP980785280), T4b)); T4d = VADD(T3L, T3M); T4e = VADD(T3O, T3P); T4f = VFNMS(LDK(KP195090322), T4e, VMUL(LDK(KP980785280), T4d)); T4n = VFMA(LDK(KP195090322), T4d, VMUL(LDK(KP980785280), T4e)); } { V T4g, T4q, T4k, T4o; T4g = VADD(T4c, T4f); T6N = VSUB(T49, T4g); STM4(&(ro[17]), T6N, ovs, &(ro[1])); T6O = VADD(T49, T4g); STM4(&(ro[1]), T6O, ovs, &(ro[1])); T4q = VADD(T4m, T4n); T6P = VSUB(T4p, T4q); STM4(&(io[17]), T6P, ovs, &(io[1])); T6Q = VADD(T4p, T4q); STM4(&(io[1]), T6Q, ovs, &(io[1])); T4k = VSUB(T4f, T4c); T6R = VSUB(T4j, T4k); STM4(&(io[25]), T6R, ovs, &(io[1])); T6S = VADD(T4j, T4k); STM4(&(io[9]), T6S, ovs, &(io[1])); T4o = VSUB(T4m, T4n); T6T = VSUB(T4l, T4o); STM4(&(ro[25]), T6T, ovs, &(ro[1])); T6U = VADD(T4l, T4o); STM4(&(ro[9]), T6U, ovs, &(ro[1])); } } { V T3b, T3n, T3l, T3r, T3e, T3o, T3h, T3p; { V T39, T3a, T3j, T3k; T39 = VADD(T1z, T1G); T3a = VADD(T2Z, T2Y); T3b = VADD(T39, T3a); T3n = VSUB(T39, T3a); T3j = VADD(T2T, T2W); T3k = VADD(T1O, T1V); T3l = VSUB(T3j, T3k); T3r = VADD(T3j, T3k); } { V T3c, T3d, T3f, T3g; T3c = VADD(T22, T2d); T3d = VADD(T2j, T2m); T3e = VFMA(LDK(KP555570233), T3c, VMUL(LDK(KP831469612), T3d)); T3o = VFNMS(LDK(KP555570233), T3d, VMUL(LDK(KP831469612), T3c)); T3f = VADD(T2t, T2E); T3g = VADD(T2K, T2N); T3h = VFNMS(LDK(KP555570233), T3g, VMUL(LDK(KP831469612), T3f)); T3p = VFMA(LDK(KP831469612), T3g, VMUL(LDK(KP555570233), T3f)); } { V T3i, T6V, T6W, T3s, T6X, T6Y; T3i = VADD(T3e, T3h); T6V = VSUB(T3b, T3i); STM4(&(ro[19]), T6V, ovs, &(ro[1])); STN4(&(ro[16]), T6p, T6N, T69, T6V, ovs); T6W = VADD(T3b, T3i); STM4(&(ro[3]), T6W, ovs, &(ro[1])); STN4(&(ro[0]), T6r, T6O, T6b, T6W, ovs); T3s = VADD(T3o, T3p); T6X = VSUB(T3r, T3s); STM4(&(io[19]), T6X, ovs, &(io[1])); STN4(&(io[16]), T6q, T6P, T6a, T6X, ovs); T6Y = VADD(T3r, T3s); STM4(&(io[3]), T6Y, ovs, &(io[1])); STN4(&(io[0]), T6s, T6Q, T6c, T6Y, ovs); } { V T3m, T6Z, T70, T3q, T71, T72; T3m = VSUB(T3h, T3e); T6Z = VSUB(T3l, T3m); STM4(&(io[27]), T6Z, ovs, &(io[1])); STN4(&(io[24]), T6v, T6R, T6d, T6Z, ovs); T70 = VADD(T3l, T3m); STM4(&(io[11]), T70, ovs, &(io[1])); STN4(&(io[8]), T6t, T6S, T6f, T70, ovs); T3q = VSUB(T3o, T3p); T71 = VSUB(T3n, T3q); STM4(&(ro[27]), T71, ovs, &(ro[1])); STN4(&(ro[24]), T6w, T6T, T6e, T71, ovs); T72 = VADD(T3n, T3q); STM4(&(ro[11]), T72, ovs, &(ro[1])); STN4(&(ro[8]), T6u, T6U, T6g, T72, ovs); } } } } } } VLEAVE(); } static const kdft_desc desc = { 32, XSIMD_STRING("n2sv_32"), {340, 52, 32, 0}, &GENUS, 0, 1, 0, 0 }; void XSIMD(codelet_n2sv_32) (planner *p) { X(kdft_register) (p, n2sv_32, &desc); } #endif /* HAVE_FMA */ fftw-3.3.3/dft/simd/common/n2sv_64.c0000644000175000017500000034464412054410513013767 00000000000000/* * Copyright (c) 2003, 2007-11 Matteo Frigo * Copyright (c) 2003, 2007-11 Massachusetts Institute of Technology * * This program is free software; you can redistribute it and/or modify * it under the terms of the GNU General Public License as published by * the Free Software Foundation; either version 2 of the License, or * (at your option) any later version. * * This program is distributed in the hope that it will be useful, * but WITHOUT ANY WARRANTY; without even the implied warranty of * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the * GNU General Public License for more details. * * You should have received a copy of the GNU General Public License * along with this program; if not, write to the Free Software * Foundation, Inc., 51 Franklin Street, Fifth Floor, Boston, MA 02110-1301 USA * */ /* This file was automatically generated --- DO NOT EDIT */ /* Generated on Sun Nov 25 07:37:57 EST 2012 */ #include "codelet-dft.h" #ifdef HAVE_FMA /* Generated by: ../../../genfft/gen_notw.native -fma -reorder-insns -schedule-for-pipeline -simd -compact -variables 4 -pipeline-latency 8 -n 64 -name n2sv_64 -with-ostride 1 -include n2s.h -store-multiple 4 */ /* * This function contains 912 FP additions, 392 FP multiplications, * (or, 520 additions, 0 multiplications, 392 fused multiply/add), * 310 stack variables, 15 constants, and 288 memory accesses */ #include "n2s.h" static void n2sv_64(const R *ri, const R *ii, R *ro, R *io, stride is, stride os, INT v, INT ivs, INT ovs) { DVK(KP881921264, +0.881921264348355029712756863660388349508442621); DVK(KP956940335, +0.956940335732208864935797886980269969482849206); DVK(KP534511135, +0.534511135950791641089685961295362908582039528); DVK(KP303346683, +0.303346683607342391675883946941299872384187453); DVK(KP773010453, +0.773010453362736960810906609758469800971041293); DVK(KP995184726, +0.995184726672196886244836953109479921575474869); DVK(KP820678790, +0.820678790828660330972281985331011598767386482); DVK(KP098491403, +0.098491403357164253077197521291327432293052451); DVK(KP831469612, +0.831469612302545237078788377617905756738560812); DVK(KP980785280, +0.980785280403230449126182236134239036973933731); DVK(KP668178637, +0.668178637919298919997757686523080761552472251); DVK(KP198912367, +0.198912367379658006911597622644676228597850501); DVK(KP923879532, +0.923879532511286756128183189396788286822416626); DVK(KP707106781, +0.707106781186547524400844362104849039284835938); DVK(KP414213562, +0.414213562373095048801688724209698078569671875); { INT i; for (i = v; i > 0; i = i - (2 * VL), ri = ri + ((2 * VL) * ivs), ii = ii + ((2 * VL) * ivs), ro = ro + ((2 * VL) * ovs), io = io + ((2 * VL) * ovs), MAKE_VOLATILE_STRIDE(256, is), MAKE_VOLATILE_STRIDE(256, os)) { V TeJ, TeK, TeP, TeQ, TfH, TfI, TfJ, TfK, Tgj, Tgk, Tgv, Tgw, T9a, T99, T9e; V T9b; { V T7B, T37, T5Z, T8F, TbB, TcB, Tf, Td9, T62, T7C, T2i, TdH, Tcb, Tah, T8G; V T3e, Tak, TbC, T65, T3m, TdI, Tu, Tda, T2x, TbD, Tan, T8I, T7G, T8J, T7J; V T64, T3t, Tas, Tce, TK, Tdd, Tav, Tcf, Tdc, T2N, T3G, T6G, T9k, T7O, T9l; V T7R, T6H, T3N, T1L, TdA, Tdx, Teo, Tbs, Tct, T5Q, T6V, T8y, T9z, T5j, T6Y; V Tbb, Tcw, T8n, T9C, Tch, Taz, Tdf, TZ, Tdg, T32, Tci, TaC, T6J, T3Z, T9n; V T7V, T9o, T7Y, T6K, T46, Tdp, T1g, Tej, Tdm, Tcm, Tb1, Tcp, TaK, T6O, T4X; V T9s, T8f, T6R, T4q, T9v, T84, Tdn, T1v, Tek, Tds, Tcn, TaV, Tcq, Tb4, T9t; V T8b, T9w, T8i, T6S, T50, T6P, T4N, T5k, T1V, T1S, TdB, Tbi, T5s, Tbt, Tbg; V T5F, T5R, T5p, T1Y, Tbj, T5n, T8z, T8q; { V Tba, T57, T8l, Tb7, T5M, T8w, T8m, T5P, T8x, T5i; { V T2p, T7F, T7E, Tal, T2w, Tam, T3s, T7H, T7I, T3p, T3d, T3a; { V T8, T35, T3, T5Y, T26, T5X, T6, T36, T29, T9, T2b, T2c, Tb, Tc, T2e; V T2f; { V T1, T2, T24, T25, T4, T5, T27, T28; T1 = LD(&(ri[0]), ivs, &(ri[0])); T2 = LD(&(ri[WS(is, 32)]), ivs, &(ri[0])); T24 = LD(&(ii[0]), ivs, &(ii[0])); T25 = LD(&(ii[WS(is, 32)]), ivs, &(ii[0])); T4 = LD(&(ri[WS(is, 16)]), ivs, &(ri[0])); T5 = LD(&(ri[WS(is, 48)]), ivs, &(ri[0])); T27 = LD(&(ii[WS(is, 16)]), ivs, &(ii[0])); T28 = LD(&(ii[WS(is, 48)]), ivs, &(ii[0])); T8 = LD(&(ri[WS(is, 8)]), ivs, &(ri[0])); T35 = VSUB(T1, T2); T3 = VADD(T1, T2); T5Y = VSUB(T24, T25); T26 = VADD(T24, T25); T5X = VSUB(T4, T5); T6 = VADD(T4, T5); T36 = VSUB(T27, T28); T29 = VADD(T27, T28); T9 = LD(&(ri[WS(is, 40)]), ivs, &(ri[0])); T2b = LD(&(ii[WS(is, 8)]), ivs, &(ii[0])); T2c = LD(&(ii[WS(is, 40)]), ivs, &(ii[0])); Tb = LD(&(ri[WS(is, 56)]), ivs, &(ri[0])); Tc = LD(&(ri[WS(is, 24)]), ivs, &(ri[0])); T2e = LD(&(ii[WS(is, 56)]), ivs, &(ii[0])); T2f = LD(&(ii[WS(is, 24)]), ivs, &(ii[0])); } { V T39, Ta, T38, T2d, T3b, Td, T3c, T2g, Taf, T7; T7B = VADD(T35, T36); T37 = VSUB(T35, T36); T39 = VSUB(T8, T9); Ta = VADD(T8, T9); T38 = VSUB(T2b, T2c); T2d = VADD(T2b, T2c); T3b = VSUB(Tb, Tc); Td = VADD(Tb, Tc); T3c = VSUB(T2e, T2f); T2g = VADD(T2e, T2f); T5Z = VADD(T5X, T5Y); T8F = VSUB(T5Y, T5X); Taf = VSUB(T3, T6); T7 = VADD(T3, T6); { V TbA, T2a, Te, Tbz, T60, T61, T2h, Tag; TbA = VSUB(T26, T29); T2a = VADD(T26, T29); Te = VADD(Ta, Td); Tbz = VSUB(Td, Ta); T3d = VADD(T3b, T3c); T60 = VSUB(T3b, T3c); T61 = VADD(T39, T38); T3a = VSUB(T38, T39); T2h = VADD(T2d, T2g); Tag = VSUB(T2d, T2g); TbB = VADD(Tbz, TbA); TcB = VSUB(TbA, Tbz); Tf = VADD(T7, Te); Td9 = VSUB(T7, Te); T62 = VSUB(T60, T61); T7C = VADD(T61, T60); T2i = VADD(T2a, T2h); TdH = VSUB(T2a, T2h); Tcb = VSUB(Taf, Tag); Tah = VADD(Taf, Tag); } } } { V T3j, Ti, T3h, T2l, T3g, Tl, T2t, T3k, T2o, T3q, Tp, T3o, T2s, T3n, Ts; V T2u, T2m, T2n; { V Tg, Th, T2j, T2k, Tj, Tk; Tg = LD(&(ri[WS(is, 4)]), ivs, &(ri[0])); Th = LD(&(ri[WS(is, 36)]), ivs, &(ri[0])); T2j = LD(&(ii[WS(is, 4)]), ivs, &(ii[0])); T2k = LD(&(ii[WS(is, 36)]), ivs, &(ii[0])); Tj = LD(&(ri[WS(is, 20)]), ivs, &(ri[0])); Tk = LD(&(ri[WS(is, 52)]), ivs, &(ri[0])); T2m = LD(&(ii[WS(is, 20)]), ivs, &(ii[0])); T8G = VADD(T3a, T3d); T3e = VSUB(T3a, T3d); T3j = VSUB(Tg, Th); Ti = VADD(Tg, Th); T3h = VSUB(T2j, T2k); T2l = VADD(T2j, T2k); T3g = VSUB(Tj, Tk); Tl = VADD(Tj, Tk); T2n = LD(&(ii[WS(is, 52)]), ivs, &(ii[0])); } { V Tn, To, T2q, T2r, Tq, Tr; Tn = LD(&(ri[WS(is, 60)]), ivs, &(ri[0])); To = LD(&(ri[WS(is, 28)]), ivs, &(ri[0])); T2q = LD(&(ii[WS(is, 60)]), ivs, &(ii[0])); T2r = LD(&(ii[WS(is, 28)]), ivs, &(ii[0])); Tq = LD(&(ri[WS(is, 12)]), ivs, &(ri[0])); Tr = LD(&(ri[WS(is, 44)]), ivs, &(ri[0])); T2t = LD(&(ii[WS(is, 12)]), ivs, &(ii[0])); T3k = VSUB(T2m, T2n); T2o = VADD(T2m, T2n); T3q = VSUB(Tn, To); Tp = VADD(Tn, To); T3o = VSUB(T2q, T2r); T2s = VADD(T2q, T2r); T3n = VSUB(Tq, Tr); Ts = VADD(Tq, Tr); T2u = LD(&(ii[WS(is, 44)]), ivs, &(ii[0])); } { V Tai, Tm, Taj, T3r; Tai = VSUB(Ti, Tl); Tm = VADD(Ti, Tl); T2p = VADD(T2l, T2o); Taj = VSUB(T2l, T2o); { V T3i, T3l, Tt, T2v; T7F = VSUB(T3h, T3g); T3i = VADD(T3g, T3h); T3l = VSUB(T3j, T3k); T7E = VADD(T3j, T3k); Tt = VADD(Tp, Ts); Tal = VSUB(Tp, Ts); T2v = VADD(T2t, T2u); T3r = VSUB(T2t, T2u); Tak = VADD(Tai, Taj); TbC = VSUB(Taj, Tai); T65 = VFNMS(LDK(KP414213562), T3i, T3l); T3m = VFMA(LDK(KP414213562), T3l, T3i); TdI = VSUB(Tt, Tm); Tu = VADD(Tm, Tt); T2w = VADD(T2s, T2v); Tam = VSUB(T2s, T2v); } T3s = VSUB(T3q, T3r); T7H = VADD(T3q, T3r); T7I = VSUB(T3o, T3n); T3p = VADD(T3n, T3o); } } { V T7M, T7Q, T7N, T3M, T3J, T7P; { V TG, T3H, Ty, T3x, T2B, T3w, TB, T3I, T2E, TH, T2J, T2K, TD, TE, T2G; V T2H; { V Tw, Tx, T2z, T2A, Tz, TA, T2C, T2D; Tw = LD(&(ri[WS(is, 2)]), ivs, &(ri[0])); Tda = VSUB(T2p, T2w); T2x = VADD(T2p, T2w); TbD = VADD(Tal, Tam); Tan = VSUB(Tal, Tam); T8I = VFNMS(LDK(KP414213562), T7E, T7F); T7G = VFMA(LDK(KP414213562), T7F, T7E); T8J = VFMA(LDK(KP414213562), T7H, T7I); T7J = VFNMS(LDK(KP414213562), T7I, T7H); T64 = VFMA(LDK(KP414213562), T3p, T3s); T3t = VFNMS(LDK(KP414213562), T3s, T3p); Tx = LD(&(ri[WS(is, 34)]), ivs, &(ri[0])); T2z = LD(&(ii[WS(is, 2)]), ivs, &(ii[0])); T2A = LD(&(ii[WS(is, 34)]), ivs, &(ii[0])); Tz = LD(&(ri[WS(is, 18)]), ivs, &(ri[0])); TA = LD(&(ri[WS(is, 50)]), ivs, &(ri[0])); T2C = LD(&(ii[WS(is, 18)]), ivs, &(ii[0])); T2D = LD(&(ii[WS(is, 50)]), ivs, &(ii[0])); TG = LD(&(ri[WS(is, 58)]), ivs, &(ri[0])); T3H = VSUB(Tw, Tx); Ty = VADD(Tw, Tx); T3x = VSUB(T2z, T2A); T2B = VADD(T2z, T2A); T3w = VSUB(Tz, TA); TB = VADD(Tz, TA); T3I = VSUB(T2C, T2D); T2E = VADD(T2C, T2D); TH = LD(&(ri[WS(is, 26)]), ivs, &(ri[0])); T2J = LD(&(ii[WS(is, 58)]), ivs, &(ii[0])); T2K = LD(&(ii[WS(is, 26)]), ivs, &(ii[0])); TD = LD(&(ri[WS(is, 10)]), ivs, &(ri[0])); TE = LD(&(ri[WS(is, 42)]), ivs, &(ri[0])); T2G = LD(&(ii[WS(is, 10)]), ivs, &(ii[0])); T2H = LD(&(ii[WS(is, 42)]), ivs, &(ii[0])); } { V Tat, TC, Tar, T2F, T3K, T3E, TJ, Taq, T2M, Tau, T3B, T3L, T3y, T3F; { V TI, T3C, T2L, T3D, TF, T3z, T2I, T3A; Tat = VSUB(Ty, TB); TC = VADD(Ty, TB); TI = VADD(TG, TH); T3C = VSUB(TG, TH); T2L = VADD(T2J, T2K); T3D = VSUB(T2J, T2K); TF = VADD(TD, TE); T3z = VSUB(TD, TE); T2I = VADD(T2G, T2H); T3A = VSUB(T2G, T2H); Tar = VSUB(T2B, T2E); T2F = VADD(T2B, T2E); T3K = VADD(T3C, T3D); T3E = VSUB(T3C, T3D); TJ = VADD(TF, TI); Taq = VSUB(TI, TF); T2M = VADD(T2I, T2L); Tau = VSUB(T2I, T2L); T3B = VADD(T3z, T3A); T3L = VSUB(T3A, T3z); } T7M = VSUB(T3x, T3w); T3y = VADD(T3w, T3x); Tas = VADD(Taq, Tar); Tce = VSUB(Tar, Taq); TK = VADD(TC, TJ); Tdd = VSUB(TC, TJ); Tav = VADD(Tat, Tau); Tcf = VSUB(Tat, Tau); T7Q = VADD(T3B, T3E); T3F = VSUB(T3B, T3E); Tdc = VSUB(T2F, T2M); T2N = VADD(T2F, T2M); T7N = VADD(T3L, T3K); T3M = VSUB(T3K, T3L); T3J = VSUB(T3H, T3I); T7P = VADD(T3H, T3I); T3G = VFNMS(LDK(KP707106781), T3F, T3y); T6G = VFMA(LDK(KP707106781), T3F, T3y); } } { V T1H, T5I, T1z, Tb8, T56, T53, T1C, Tb9, T5L, T1I, T5e, T5f, T1E, T1F, T59; V T5a; { V T1x, T1y, T54, T55, T1A, T1B, T5J, T5K; T1x = LD(&(ri[WS(is, 63)]), ivs, &(ri[WS(is, 1)])); T9k = VFNMS(LDK(KP707106781), T7N, T7M); T7O = VFMA(LDK(KP707106781), T7N, T7M); T9l = VFNMS(LDK(KP707106781), T7Q, T7P); T7R = VFMA(LDK(KP707106781), T7Q, T7P); T6H = VFMA(LDK(KP707106781), T3M, T3J); T3N = VFNMS(LDK(KP707106781), T3M, T3J); T1y = LD(&(ri[WS(is, 31)]), ivs, &(ri[WS(is, 1)])); T54 = LD(&(ii[WS(is, 63)]), ivs, &(ii[WS(is, 1)])); T55 = LD(&(ii[WS(is, 31)]), ivs, &(ii[WS(is, 1)])); T1A = LD(&(ri[WS(is, 15)]), ivs, &(ri[WS(is, 1)])); T1B = LD(&(ri[WS(is, 47)]), ivs, &(ri[WS(is, 1)])); T5J = LD(&(ii[WS(is, 15)]), ivs, &(ii[WS(is, 1)])); T5K = LD(&(ii[WS(is, 47)]), ivs, &(ii[WS(is, 1)])); T1H = LD(&(ri[WS(is, 55)]), ivs, &(ri[WS(is, 1)])); T5I = VSUB(T1x, T1y); T1z = VADD(T1x, T1y); Tb8 = VADD(T54, T55); T56 = VSUB(T54, T55); T53 = VSUB(T1A, T1B); T1C = VADD(T1A, T1B); Tb9 = VADD(T5J, T5K); T5L = VSUB(T5J, T5K); T1I = LD(&(ri[WS(is, 23)]), ivs, &(ri[WS(is, 1)])); T5e = LD(&(ii[WS(is, 55)]), ivs, &(ii[WS(is, 1)])); T5f = LD(&(ii[WS(is, 23)]), ivs, &(ii[WS(is, 1)])); T1E = LD(&(ri[WS(is, 7)]), ivs, &(ri[WS(is, 1)])); T1F = LD(&(ri[WS(is, 39)]), ivs, &(ri[WS(is, 1)])); T59 = LD(&(ii[WS(is, 7)]), ivs, &(ii[WS(is, 1)])); T5a = LD(&(ii[WS(is, 39)]), ivs, &(ii[WS(is, 1)])); } { V Tbo, T1D, Tdv, T5h, T5N, T1K, Tdw, Tbr, T5O, T5c; { V T1J, T5d, Tbq, T5g, T1G, T58, Tbp, T5b; Tbo = VSUB(T1z, T1C); T1D = VADD(T1z, T1C); T1J = VADD(T1H, T1I); T5d = VSUB(T1H, T1I); Tbq = VADD(T5e, T5f); T5g = VSUB(T5e, T5f); T1G = VADD(T1E, T1F); T58 = VSUB(T1E, T1F); Tbp = VADD(T59, T5a); T5b = VSUB(T59, T5a); Tba = VSUB(Tb8, Tb9); Tdv = VADD(Tb8, Tb9); T57 = VADD(T53, T56); T8l = VSUB(T56, T53); T5h = VSUB(T5d, T5g); T5N = VADD(T5d, T5g); Tb7 = VSUB(T1J, T1G); T1K = VADD(T1G, T1J); Tdw = VADD(Tbp, Tbq); Tbr = VSUB(Tbp, Tbq); T5O = VSUB(T5b, T58); T5c = VADD(T58, T5b); } T5M = VSUB(T5I, T5L); T8w = VADD(T5I, T5L); T1L = VADD(T1D, T1K); TdA = VSUB(T1D, T1K); Tdx = VSUB(Tdv, Tdw); Teo = VADD(Tdv, Tdw); Tbs = VADD(Tbo, Tbr); Tct = VSUB(Tbo, Tbr); T8m = VADD(T5O, T5N); T5P = VSUB(T5N, T5O); T8x = VADD(T5c, T5h); T5i = VSUB(T5c, T5h); } } } } { V T4e, T82, T8d, T4T, T4W, T83, T4p, T8e; { V T7T, T3R, T42, T7W, T3Y, T7X, T45, T7U; { V T40, TN, T2Y, T3Q, T2Q, T3P, TQ, T41, T2T, T3V, TX, T2Z, TS, TT, T2V; V T2W; { V T2O, T2P, TO, TP, TL, TM; TL = LD(&(ri[WS(is, 62)]), ivs, &(ri[0])); TM = LD(&(ri[WS(is, 30)]), ivs, &(ri[0])); T5Q = VFNMS(LDK(KP707106781), T5P, T5M); T6V = VFMA(LDK(KP707106781), T5P, T5M); T8y = VFMA(LDK(KP707106781), T8x, T8w); T9z = VFNMS(LDK(KP707106781), T8x, T8w); T5j = VFNMS(LDK(KP707106781), T5i, T57); T6Y = VFMA(LDK(KP707106781), T5i, T57); Tbb = VADD(Tb7, Tba); Tcw = VSUB(Tba, Tb7); T8n = VFMA(LDK(KP707106781), T8m, T8l); T9C = VFNMS(LDK(KP707106781), T8m, T8l); T40 = VSUB(TL, TM); TN = VADD(TL, TM); T2O = LD(&(ii[WS(is, 62)]), ivs, &(ii[0])); T2P = LD(&(ii[WS(is, 30)]), ivs, &(ii[0])); TO = LD(&(ri[WS(is, 14)]), ivs, &(ri[0])); TP = LD(&(ri[WS(is, 46)]), ivs, &(ri[0])); { V T2R, T2S, TV, TW; T2R = LD(&(ii[WS(is, 14)]), ivs, &(ii[0])); T2S = LD(&(ii[WS(is, 46)]), ivs, &(ii[0])); TV = LD(&(ri[WS(is, 54)]), ivs, &(ri[0])); TW = LD(&(ri[WS(is, 22)]), ivs, &(ri[0])); T2Y = LD(&(ii[WS(is, 54)]), ivs, &(ii[0])); T3Q = VSUB(T2O, T2P); T2Q = VADD(T2O, T2P); T3P = VSUB(TO, TP); TQ = VADD(TO, TP); T41 = VSUB(T2R, T2S); T2T = VADD(T2R, T2S); T3V = VSUB(TV, TW); TX = VADD(TV, TW); T2Z = LD(&(ii[WS(is, 22)]), ivs, &(ii[0])); TS = LD(&(ri[WS(is, 6)]), ivs, &(ri[0])); TT = LD(&(ri[WS(is, 38)]), ivs, &(ri[0])); T2V = LD(&(ii[WS(is, 6)]), ivs, &(ii[0])); T2W = LD(&(ii[WS(is, 38)]), ivs, &(ii[0])); } } { V TaA, TR, Tay, T2U, T3W, T30, TU, T3S, T2X, T3T; TaA = VSUB(TN, TQ); TR = VADD(TN, TQ); Tay = VSUB(T2Q, T2T); T2U = VADD(T2Q, T2T); T3W = VSUB(T2Y, T2Z); T30 = VADD(T2Y, T2Z); TU = VADD(TS, TT); T3S = VSUB(TS, TT); T2X = VADD(T2V, T2W); T3T = VSUB(T2V, T2W); { V T3X, T43, Tax, TY, T31, TaB, T3U, T44; T7T = VSUB(T3Q, T3P); T3R = VADD(T3P, T3Q); T3X = VSUB(T3V, T3W); T43 = VADD(T3V, T3W); Tax = VSUB(TX, TU); TY = VADD(TU, TX); T31 = VADD(T2X, T30); TaB = VSUB(T2X, T30); T3U = VADD(T3S, T3T); T44 = VSUB(T3T, T3S); T42 = VSUB(T40, T41); T7W = VADD(T40, T41); Tch = VSUB(Tay, Tax); Taz = VADD(Tax, Tay); Tdf = VSUB(TR, TY); TZ = VADD(TR, TY); Tdg = VSUB(T2U, T31); T32 = VADD(T2U, T31); Tci = VSUB(TaA, TaB); TaC = VADD(TaA, TaB); T3Y = VSUB(T3U, T3X); T7X = VADD(T3U, T3X); T45 = VSUB(T43, T44); T7U = VADD(T44, T43); } } } { V T4P, T14, T4l, TaH, T4d, T4a, T17, TaI, T4S, T4k, T1e, T4m, T19, T1a, T4g; V T4h; { V T4b, T4c, T15, T16, T12, T13; T12 = LD(&(ri[WS(is, 1)]), ivs, &(ri[WS(is, 1)])); T13 = LD(&(ri[WS(is, 33)]), ivs, &(ri[WS(is, 1)])); T4b = LD(&(ii[WS(is, 1)]), ivs, &(ii[WS(is, 1)])); T6J = VFMA(LDK(KP707106781), T3Y, T3R); T3Z = VFNMS(LDK(KP707106781), T3Y, T3R); T9n = VFNMS(LDK(KP707106781), T7U, T7T); T7V = VFMA(LDK(KP707106781), T7U, T7T); T9o = VFNMS(LDK(KP707106781), T7X, T7W); T7Y = VFMA(LDK(KP707106781), T7X, T7W); T6K = VFMA(LDK(KP707106781), T45, T42); T46 = VFNMS(LDK(KP707106781), T45, T42); T4P = VSUB(T12, T13); T14 = VADD(T12, T13); T4c = LD(&(ii[WS(is, 33)]), ivs, &(ii[WS(is, 1)])); T15 = LD(&(ri[WS(is, 17)]), ivs, &(ri[WS(is, 1)])); T16 = LD(&(ri[WS(is, 49)]), ivs, &(ri[WS(is, 1)])); { V T4Q, T4R, T1c, T1d; T4Q = LD(&(ii[WS(is, 17)]), ivs, &(ii[WS(is, 1)])); T4R = LD(&(ii[WS(is, 49)]), ivs, &(ii[WS(is, 1)])); T1c = LD(&(ri[WS(is, 57)]), ivs, &(ri[WS(is, 1)])); T1d = LD(&(ri[WS(is, 25)]), ivs, &(ri[WS(is, 1)])); T4l = LD(&(ii[WS(is, 57)]), ivs, &(ii[WS(is, 1)])); TaH = VADD(T4b, T4c); T4d = VSUB(T4b, T4c); T4a = VSUB(T15, T16); T17 = VADD(T15, T16); TaI = VADD(T4Q, T4R); T4S = VSUB(T4Q, T4R); T4k = VSUB(T1c, T1d); T1e = VADD(T1c, T1d); T4m = LD(&(ii[WS(is, 25)]), ivs, &(ii[WS(is, 1)])); T19 = LD(&(ri[WS(is, 9)]), ivs, &(ri[WS(is, 1)])); T1a = LD(&(ri[WS(is, 41)]), ivs, &(ri[WS(is, 1)])); T4g = LD(&(ii[WS(is, 9)]), ivs, &(ii[WS(is, 1)])); T4h = LD(&(ii[WS(is, 41)]), ivs, &(ii[WS(is, 1)])); } } { V TaX, T18, T4n, TaZ, TaJ, Tdk, T1b, T4f, TaY, T4i; TaX = VSUB(T14, T17); T18 = VADD(T14, T17); T4n = VSUB(T4l, T4m); TaZ = VADD(T4l, T4m); TaJ = VSUB(TaH, TaI); Tdk = VADD(TaH, TaI); T1b = VADD(T19, T1a); T4f = VSUB(T19, T1a); TaY = VADD(T4g, T4h); T4i = VSUB(T4g, T4h); T4e = VADD(T4a, T4d); T82 = VSUB(T4d, T4a); { V T4U, T4o, T1f, TaG, Tdl, Tb0, T4V, T4j; T8d = VADD(T4P, T4S); T4T = VSUB(T4P, T4S); T4U = VADD(T4k, T4n); T4o = VSUB(T4k, T4n); T1f = VADD(T1b, T1e); TaG = VSUB(T1e, T1b); Tdl = VADD(TaY, TaZ); Tb0 = VSUB(TaY, TaZ); T4V = VSUB(T4i, T4f); T4j = VADD(T4f, T4i); Tdp = VSUB(T18, T1f); T1g = VADD(T18, T1f); Tej = VADD(Tdk, Tdl); Tdm = VSUB(Tdk, Tdl); Tcm = VSUB(TaX, Tb0); Tb1 = VADD(TaX, Tb0); T4W = VSUB(T4U, T4V); T83 = VADD(T4V, T4U); T4p = VSUB(T4j, T4o); T8e = VADD(T4j, T4o); Tcp = VSUB(TaJ, TaG); TaK = VADD(TaG, TaJ); } } } } { V T1n, Tdq, T4r, T1q, TaR, T4z, Tb2, TaP, T4M, T4Y, T4w, T1t, TaS, T4u, T8g; V T87; { V T1r, T85, T4L, TaO, TaN, T86, T4G, T1s, T4s, T4t; { V T1h, T1i, T4I, T4J, T1k, T1l, T4D, T4E; T1h = LD(&(ri[WS(is, 5)]), ivs, &(ri[WS(is, 1)])); T6O = VFMA(LDK(KP707106781), T4W, T4T); T4X = VFNMS(LDK(KP707106781), T4W, T4T); T9s = VFNMS(LDK(KP707106781), T8e, T8d); T8f = VFMA(LDK(KP707106781), T8e, T8d); T6R = VFMA(LDK(KP707106781), T4p, T4e); T4q = VFNMS(LDK(KP707106781), T4p, T4e); T9v = VFNMS(LDK(KP707106781), T83, T82); T84 = VFMA(LDK(KP707106781), T83, T82); T1i = LD(&(ri[WS(is, 37)]), ivs, &(ri[WS(is, 1)])); T4I = LD(&(ii[WS(is, 5)]), ivs, &(ii[WS(is, 1)])); T4J = LD(&(ii[WS(is, 37)]), ivs, &(ii[WS(is, 1)])); T1k = LD(&(ri[WS(is, 21)]), ivs, &(ri[WS(is, 1)])); T1l = LD(&(ri[WS(is, 53)]), ivs, &(ri[WS(is, 1)])); T4D = LD(&(ii[WS(is, 21)]), ivs, &(ii[WS(is, 1)])); T4E = LD(&(ii[WS(is, 53)]), ivs, &(ii[WS(is, 1)])); { V T1o, T4C, T1j, TaL, T4K, T4H, T1m, TaM, T4F, T1p, T4x, T4y; T1o = LD(&(ri[WS(is, 61)]), ivs, &(ri[WS(is, 1)])); T4C = VSUB(T1h, T1i); T1j = VADD(T1h, T1i); TaL = VADD(T4I, T4J); T4K = VSUB(T4I, T4J); T4H = VSUB(T1k, T1l); T1m = VADD(T1k, T1l); TaM = VADD(T4D, T4E); T4F = VSUB(T4D, T4E); T1p = LD(&(ri[WS(is, 29)]), ivs, &(ri[WS(is, 1)])); T4x = LD(&(ii[WS(is, 61)]), ivs, &(ii[WS(is, 1)])); T4y = LD(&(ii[WS(is, 29)]), ivs, &(ii[WS(is, 1)])); T1r = LD(&(ri[WS(is, 13)]), ivs, &(ri[WS(is, 1)])); T85 = VSUB(T4K, T4H); T4L = VADD(T4H, T4K); TaO = VSUB(T1j, T1m); T1n = VADD(T1j, T1m); Tdq = VADD(TaL, TaM); TaN = VSUB(TaL, TaM); T86 = VADD(T4C, T4F); T4G = VSUB(T4C, T4F); T4r = VSUB(T1o, T1p); T1q = VADD(T1o, T1p); TaR = VADD(T4x, T4y); T4z = VSUB(T4x, T4y); T1s = LD(&(ri[WS(is, 45)]), ivs, &(ri[WS(is, 1)])); T4s = LD(&(ii[WS(is, 13)]), ivs, &(ii[WS(is, 1)])); T4t = LD(&(ii[WS(is, 45)]), ivs, &(ii[WS(is, 1)])); } } Tb2 = VADD(TaO, TaN); TaP = VSUB(TaN, TaO); T4M = VFNMS(LDK(KP414213562), T4L, T4G); T4Y = VFMA(LDK(KP414213562), T4G, T4L); T4w = VSUB(T1r, T1s); T1t = VADD(T1r, T1s); TaS = VADD(T4s, T4t); T4u = VSUB(T4s, T4t); T8g = VFMA(LDK(KP414213562), T85, T86); T87 = VFNMS(LDK(KP414213562), T86, T85); } { V T1W, T8o, T5E, Tbf, Tbe, T8p, T5z, T1X, T5l, T5m; { V T5B, T5v, T1O, T5C, T1P, T1Q, T5w, T5x; { V T1M, T88, T4A, T1u, TaQ, Tdr, TaT, T89, T4v, T1N, TaU, Tb3; T1M = LD(&(ri[WS(is, 3)]), ivs, &(ri[WS(is, 1)])); T88 = VSUB(T4z, T4w); T4A = VADD(T4w, T4z); T1u = VADD(T1q, T1t); TaQ = VSUB(T1q, T1t); Tdr = VADD(TaR, TaS); TaT = VSUB(TaR, TaS); T89 = VADD(T4r, T4u); T4v = VSUB(T4r, T4u); T1N = LD(&(ri[WS(is, 35)]), ivs, &(ri[WS(is, 1)])); T5B = LD(&(ii[WS(is, 3)]), ivs, &(ii[WS(is, 1)])); Tdn = VSUB(T1u, T1n); T1v = VADD(T1n, T1u); Tek = VADD(Tdq, Tdr); Tds = VSUB(Tdq, Tdr); TaU = VADD(TaQ, TaT); Tb3 = VSUB(TaQ, TaT); { V T8a, T8h, T4Z, T4B; T8a = VFMA(LDK(KP414213562), T89, T88); T8h = VFNMS(LDK(KP414213562), T88, T89); T4Z = VFNMS(LDK(KP414213562), T4v, T4A); T4B = VFMA(LDK(KP414213562), T4A, T4v); T5v = VSUB(T1M, T1N); T1O = VADD(T1M, T1N); Tcn = VSUB(TaU, TaP); TaV = VADD(TaP, TaU); Tcq = VSUB(Tb2, Tb3); Tb4 = VADD(Tb2, Tb3); T9t = VSUB(T8a, T87); T8b = VADD(T87, T8a); T9w = VSUB(T8g, T8h); T8i = VADD(T8g, T8h); T6S = VADD(T4Y, T4Z); T50 = VSUB(T4Y, T4Z); T6P = VADD(T4M, T4B); T4N = VSUB(T4B, T4M); T5C = LD(&(ii[WS(is, 35)]), ivs, &(ii[WS(is, 1)])); } } T1P = LD(&(ri[WS(is, 19)]), ivs, &(ri[WS(is, 1)])); T1Q = LD(&(ri[WS(is, 51)]), ivs, &(ri[WS(is, 1)])); T5w = LD(&(ii[WS(is, 19)]), ivs, &(ii[WS(is, 1)])); T5x = LD(&(ii[WS(is, 51)]), ivs, &(ii[WS(is, 1)])); { V T5q, Tbc, T5D, T5A, T1R, Tbd, T5y, T5r, T1T, T1U; T1T = LD(&(ri[WS(is, 59)]), ivs, &(ri[WS(is, 1)])); T1U = LD(&(ri[WS(is, 27)]), ivs, &(ri[WS(is, 1)])); T5q = LD(&(ii[WS(is, 59)]), ivs, &(ii[WS(is, 1)])); Tbc = VADD(T5B, T5C); T5D = VSUB(T5B, T5C); T5A = VSUB(T1P, T1Q); T1R = VADD(T1P, T1Q); Tbd = VADD(T5w, T5x); T5y = VSUB(T5w, T5x); T5k = VSUB(T1T, T1U); T1V = VADD(T1T, T1U); T5r = LD(&(ii[WS(is, 27)]), ivs, &(ii[WS(is, 1)])); T1W = LD(&(ri[WS(is, 11)]), ivs, &(ri[WS(is, 1)])); T8o = VSUB(T5D, T5A); T5E = VADD(T5A, T5D); Tbf = VSUB(T1O, T1R); T1S = VADD(T1O, T1R); TdB = VADD(Tbc, Tbd); Tbe = VSUB(Tbc, Tbd); T8p = VADD(T5v, T5y); T5z = VSUB(T5v, T5y); Tbi = VADD(T5q, T5r); T5s = VSUB(T5q, T5r); T1X = LD(&(ri[WS(is, 43)]), ivs, &(ri[WS(is, 1)])); T5l = LD(&(ii[WS(is, 11)]), ivs, &(ii[WS(is, 1)])); T5m = LD(&(ii[WS(is, 43)]), ivs, &(ii[WS(is, 1)])); } } Tbt = VADD(Tbf, Tbe); Tbg = VSUB(Tbe, Tbf); T5F = VFNMS(LDK(KP414213562), T5E, T5z); T5R = VFMA(LDK(KP414213562), T5z, T5E); T5p = VSUB(T1W, T1X); T1Y = VADD(T1W, T1X); Tbj = VADD(T5l, T5m); T5n = VSUB(T5l, T5m); T8z = VFMA(LDK(KP414213562), T8o, T8p); T8q = VFNMS(LDK(KP414213562), T8p, T8o); } } } } { V Tbm, Tbv, T9A, T8u, T9D, T8B, T6Z, T5T, T6W, T5G, TeL, TeM, TeN, TeO, TeR; V TeS, TeT, TeU, TeV, TeW, TeX, TeY, TeZ, Tf0, Tf1, Tf2, Tf3, Tf4, Tf5, Tf6; V Tf7, Tf8, Tf9, Tfa, Tfb, Tfc, TbE, Tao, Tfd, Tfe, Td7, Td8, Tff, Tfg, Tfh; V Tfi, Tfj, Tfk, Tfl, Tfm, Tfn, Tfo, Tfp, Tfq, Tfr, Tfs; { V Tel, Tdy, TdD, Tcu, Tcx, Teq, Tei, Ten, Tex, Teh, TeB, Tev, Te9, Tec; { V Tef, Teu, TeE, TeD, T11, TeF, T1w, T21, Tet, T2y, T33, Teg, T20; { V Tv, T8r, T5t, T1Z, Tbh, TdC, Tbk, T8s, T5o, T10, Tep, Tbl, Tbu; Tef = VSUB(Tf, Tu); Tv = VADD(Tf, Tu); T8r = VSUB(T5s, T5p); T5t = VADD(T5p, T5s); T1Z = VADD(T1V, T1Y); Tbh = VSUB(T1V, T1Y); TdC = VADD(Tbi, Tbj); Tbk = VSUB(Tbi, Tbj); T8s = VADD(T5k, T5n); T5o = VSUB(T5k, T5n); T10 = VADD(TK, TZ); Teu = VSUB(TZ, TK); Tel = VSUB(Tej, Tek); TeE = VADD(Tej, Tek); Tdy = VSUB(T1Z, T1S); T20 = VADD(T1S, T1Z); Tep = VADD(TdB, TdC); TdD = VSUB(TdB, TdC); Tbl = VADD(Tbh, Tbk); Tbu = VSUB(Tbh, Tbk); { V T8t, T8A, T5S, T5u; T8t = VFMA(LDK(KP414213562), T8s, T8r); T8A = VFNMS(LDK(KP414213562), T8r, T8s); T5S = VFNMS(LDK(KP414213562), T5o, T5t); T5u = VFMA(LDK(KP414213562), T5t, T5o); TeD = VSUB(Tv, T10); T11 = VADD(Tv, T10); Tcu = VSUB(Tbl, Tbg); Tbm = VADD(Tbg, Tbl); Tcx = VSUB(Tbt, Tbu); Tbv = VADD(Tbt, Tbu); T9A = VSUB(T8t, T8q); T8u = VADD(T8q, T8t); T9D = VSUB(T8z, T8A); T8B = VADD(T8z, T8A); T6Z = VADD(T5R, T5S); T5T = VSUB(T5R, T5S); T6W = VADD(T5F, T5u); T5G = VSUB(T5u, T5F); TeF = VADD(Teo, Tep); Teq = VSUB(Teo, Tep); } } Tei = VSUB(T1g, T1v); T1w = VADD(T1g, T1v); T21 = VADD(T1L, T20); Ten = VSUB(T1L, T20); Tet = VSUB(T2i, T2x); T2y = VADD(T2i, T2x); T33 = VADD(T2N, T32); Teg = VSUB(T2N, T32); { V TeI, TeG, T23, T22, TeH, T34; TeI = VADD(TeE, TeF); TeG = VSUB(TeE, TeF); T23 = VSUB(T21, T1w); T22 = VADD(T1w, T21); TeH = VADD(T2y, T33); T34 = VSUB(T2y, T33); Tex = VSUB(Tef, Teg); Teh = VADD(Tef, Teg); TeJ = VSUB(TeD, TeG); STM4(&(ro[48]), TeJ, ovs, &(ro[0])); TeK = VADD(TeD, TeG); STM4(&(ro[16]), TeK, ovs, &(ro[0])); TeL = VADD(T11, T22); STM4(&(ro[0]), TeL, ovs, &(ro[0])); TeM = VSUB(T11, T22); STM4(&(ro[32]), TeM, ovs, &(ro[0])); TeN = VADD(TeH, TeI); STM4(&(io[0]), TeN, ovs, &(io[0])); TeO = VSUB(TeH, TeI); STM4(&(io[32]), TeO, ovs, &(io[0])); TeP = VSUB(T34, T23); STM4(&(io[48]), TeP, ovs, &(io[0])); TeQ = VADD(T23, T34); STM4(&(io[16]), TeQ, ovs, &(io[0])); TeB = VADD(Teu, Tet); Tev = VSUB(Tet, Teu); } } { V TdV, Tdb, TdJ, Te5, TdE, Tdz, TdZ, Tdo, Te6, Tdi, Teb, Te3, TdW, TdM, Tdt; V TdY; { V TdL, Tde, Tey, Tem, Tez, Ter, Tdh, TdK, Te1, Te2; TdV = VADD(Td9, Tda); Tdb = VSUB(Td9, Tda); TdJ = VSUB(TdH, TdI); Te5 = VADD(TdI, TdH); TdL = VADD(Tdd, Tdc); Tde = VSUB(Tdc, Tdd); Tey = VSUB(Tel, Tei); Tem = VADD(Tei, Tel); Tez = VADD(Ten, Teq); Ter = VSUB(Ten, Teq); Tdh = VADD(Tdf, Tdg); TdK = VSUB(Tdf, Tdg); TdE = VSUB(TdA, TdD); Te1 = VADD(TdA, TdD); Te2 = VADD(Tdy, Tdx); Tdz = VSUB(Tdx, Tdy); TdZ = VADD(Tdn, Tdm); Tdo = VSUB(Tdm, Tdn); { V TeA, TeC, Tew, Tes; TeA = VSUB(Tey, Tez); TeC = VADD(Tey, Tez); Tew = VSUB(Ter, Tem); Tes = VADD(Tem, Ter); Te6 = VADD(Tde, Tdh); Tdi = VSUB(Tde, Tdh); Teb = VFMA(LDK(KP414213562), Te1, Te2); Te3 = VFNMS(LDK(KP414213562), Te2, Te1); TdW = VADD(TdL, TdK); TdM = VSUB(TdK, TdL); TeR = VFMA(LDK(KP707106781), TeA, Tex); STM4(&(ro[24]), TeR, ovs, &(ro[0])); TeS = VFNMS(LDK(KP707106781), TeA, Tex); STM4(&(ro[56]), TeS, ovs, &(ro[0])); TeT = VFMA(LDK(KP707106781), TeC, TeB); STM4(&(io[8]), TeT, ovs, &(io[0])); TeU = VFNMS(LDK(KP707106781), TeC, TeB); STM4(&(io[40]), TeU, ovs, &(io[0])); TeV = VFMA(LDK(KP707106781), Tew, Tev); STM4(&(io[24]), TeV, ovs, &(io[0])); TeW = VFNMS(LDK(KP707106781), Tew, Tev); STM4(&(io[56]), TeW, ovs, &(io[0])); TeX = VFMA(LDK(KP707106781), Tes, Teh); STM4(&(ro[8]), TeX, ovs, &(ro[0])); TeY = VFNMS(LDK(KP707106781), Tes, Teh); STM4(&(ro[40]), TeY, ovs, &(ro[0])); Tdt = VSUB(Tdp, Tds); TdY = VADD(Tdp, Tds); } } { V TdT, Tdj, TdP, TdN, TdR, Tdu, Tea, Te0, TdQ, TdF, TdX, Ted, Te7; TdT = VFNMS(LDK(KP707106781), Tdi, Tdb); Tdj = VFMA(LDK(KP707106781), Tdi, Tdb); TdP = VFMA(LDK(KP707106781), TdM, TdJ); TdN = VFNMS(LDK(KP707106781), TdM, TdJ); TdR = VFNMS(LDK(KP414213562), Tdo, Tdt); Tdu = VFMA(LDK(KP414213562), Tdt, Tdo); Tea = VFNMS(LDK(KP414213562), TdY, TdZ); Te0 = VFMA(LDK(KP414213562), TdZ, TdY); TdQ = VFMA(LDK(KP414213562), Tdz, TdE); TdF = VFNMS(LDK(KP414213562), TdE, Tdz); Te9 = VFNMS(LDK(KP707106781), TdW, TdV); TdX = VFMA(LDK(KP707106781), TdW, TdV); Ted = VFMA(LDK(KP707106781), Te6, Te5); Te7 = VFNMS(LDK(KP707106781), Te6, Te5); { V Tee, Te8, Te4, TdU, TdS, TdO, TdG; Tee = VADD(Tea, Teb); Tec = VSUB(Tea, Teb); Te8 = VSUB(Te3, Te0); Te4 = VADD(Te0, Te3); TdU = VADD(TdR, TdQ); TdS = VSUB(TdQ, TdR); TdO = VADD(Tdu, TdF); TdG = VSUB(Tdu, TdF); TeZ = VFMA(LDK(KP923879532), Tee, Ted); STM4(&(io[4]), TeZ, ovs, &(io[0])); Tf0 = VFNMS(LDK(KP923879532), Tee, Ted); STM4(&(io[36]), Tf0, ovs, &(io[0])); Tf1 = VFMA(LDK(KP923879532), Te4, TdX); STM4(&(ro[4]), Tf1, ovs, &(ro[0])); Tf2 = VFNMS(LDK(KP923879532), Te4, TdX); STM4(&(ro[36]), Tf2, ovs, &(ro[0])); Tf3 = VFMA(LDK(KP923879532), TdU, TdT); STM4(&(ro[60]), Tf3, ovs, &(ro[0])); Tf4 = VFNMS(LDK(KP923879532), TdU, TdT); STM4(&(ro[28]), Tf4, ovs, &(ro[0])); Tf5 = VFMA(LDK(KP923879532), TdS, TdP); STM4(&(io[12]), Tf5, ovs, &(io[0])); Tf6 = VFNMS(LDK(KP923879532), TdS, TdP); STM4(&(io[44]), Tf6, ovs, &(io[0])); Tf7 = VFMA(LDK(KP923879532), TdO, TdN); STM4(&(io[60]), Tf7, ovs, &(io[0])); Tf8 = VFNMS(LDK(KP923879532), TdO, TdN); STM4(&(io[28]), Tf8, ovs, &(io[0])); Tf9 = VFMA(LDK(KP923879532), TdG, Tdj); STM4(&(ro[12]), Tf9, ovs, &(ro[0])); Tfa = VFNMS(LDK(KP923879532), TdG, Tdj); STM4(&(ro[44]), Tfa, ovs, &(ro[0])); Tfb = VFMA(LDK(KP923879532), Te8, Te7); STM4(&(io[20]), Tfb, ovs, &(io[0])); Tfc = VFNMS(LDK(KP923879532), Te8, Te7); STM4(&(io[52]), Tfc, ovs, &(io[0])); } } } { V TcF, TcE, Tcy, Tcv, TcT, Tco, TcP, Tcd, TcZ, TcD, Td0, Tck, Td4, TcX, Tcr; V TcS; { V Tcc, TcC, Tcg, Tcj, TcV, TcW; TbE = VADD(TbC, TbD); Tcc = VSUB(TbC, TbD); TcC = VSUB(Tan, Tak); Tao = VADD(Tak, Tan); TcF = VFNMS(LDK(KP414213562), Tce, Tcf); Tcg = VFMA(LDK(KP414213562), Tcf, Tce); Tcj = VFNMS(LDK(KP414213562), Tci, Tch); TcE = VFMA(LDK(KP414213562), Tch, Tci); Tcy = VFNMS(LDK(KP707106781), Tcx, Tcw); TcV = VFMA(LDK(KP707106781), Tcx, Tcw); TcW = VFMA(LDK(KP707106781), Tcu, Tct); Tcv = VFNMS(LDK(KP707106781), Tcu, Tct); TcT = VFMA(LDK(KP707106781), Tcn, Tcm); Tco = VFNMS(LDK(KP707106781), Tcn, Tcm); Tfd = VFMA(LDK(KP923879532), Tec, Te9); STM4(&(ro[20]), Tfd, ovs, &(ro[0])); Tfe = VFNMS(LDK(KP923879532), Tec, Te9); STM4(&(ro[52]), Tfe, ovs, &(ro[0])); TcP = VFNMS(LDK(KP707106781), Tcc, Tcb); Tcd = VFMA(LDK(KP707106781), Tcc, Tcb); TcZ = VFNMS(LDK(KP707106781), TcC, TcB); TcD = VFMA(LDK(KP707106781), TcC, TcB); Td0 = VADD(Tcg, Tcj); Tck = VSUB(Tcg, Tcj); Td4 = VFMA(LDK(KP198912367), TcV, TcW); TcX = VFNMS(LDK(KP198912367), TcW, TcV); Tcr = VFNMS(LDK(KP707106781), Tcq, Tcp); TcS = VFMA(LDK(KP707106781), Tcq, Tcp); } { V TcJ, Tcl, TcK, Tcs, TcQ, TcG, Td5, TcU, TcL, Tcz; TcJ = VFNMS(LDK(KP923879532), Tck, Tcd); Tcl = VFMA(LDK(KP923879532), Tck, Tcd); TcK = VFNMS(LDK(KP668178637), Tco, Tcr); Tcs = VFMA(LDK(KP668178637), Tcr, Tco); TcQ = VADD(TcF, TcE); TcG = VSUB(TcE, TcF); Td5 = VFNMS(LDK(KP198912367), TcS, TcT); TcU = VFMA(LDK(KP198912367), TcT, TcS); TcL = VFMA(LDK(KP668178637), Tcv, Tcy); Tcz = VFNMS(LDK(KP668178637), Tcy, Tcv); { V Td1, Td3, TcR, TcN, TcH, Td2, TcY, TcM, TcO, TcI, TcA, Td6; Td1 = VFMA(LDK(KP923879532), Td0, TcZ); Td3 = VFNMS(LDK(KP923879532), Td0, TcZ); TcR = VFNMS(LDK(KP923879532), TcQ, TcP); Td7 = VFMA(LDK(KP923879532), TcQ, TcP); TcN = VFMA(LDK(KP923879532), TcG, TcD); TcH = VFNMS(LDK(KP923879532), TcG, TcD); Td2 = VADD(TcU, TcX); TcY = VSUB(TcU, TcX); TcM = VSUB(TcK, TcL); TcO = VADD(TcK, TcL); TcI = VSUB(Tcz, Tcs); TcA = VADD(Tcs, Tcz); Td6 = VSUB(Td4, Td5); Td8 = VADD(Td5, Td4); Tff = VFMA(LDK(KP980785280), TcY, TcR); STM4(&(ro[14]), Tff, ovs, &(ro[0])); Tfg = VFNMS(LDK(KP980785280), TcY, TcR); STM4(&(ro[46]), Tfg, ovs, &(ro[0])); Tfh = VFMA(LDK(KP831469612), TcM, TcJ); STM4(&(ro[22]), Tfh, ovs, &(ro[0])); Tfi = VFNMS(LDK(KP831469612), TcM, TcJ); STM4(&(ro[54]), Tfi, ovs, &(ro[0])); Tfj = VFMA(LDK(KP831469612), TcO, TcN); STM4(&(io[6]), Tfj, ovs, &(io[0])); Tfk = VFNMS(LDK(KP831469612), TcO, TcN); STM4(&(io[38]), Tfk, ovs, &(io[0])); Tfl = VFMA(LDK(KP831469612), TcI, TcH); STM4(&(io[22]), Tfl, ovs, &(io[0])); Tfm = VFNMS(LDK(KP831469612), TcI, TcH); STM4(&(io[54]), Tfm, ovs, &(io[0])); Tfn = VFMA(LDK(KP831469612), TcA, Tcl); STM4(&(ro[6]), Tfn, ovs, &(ro[0])); Tfo = VFNMS(LDK(KP831469612), TcA, Tcl); STM4(&(ro[38]), Tfo, ovs, &(ro[0])); Tfp = VFMA(LDK(KP980785280), Td6, Td3); STM4(&(io[14]), Tfp, ovs, &(io[0])); Tfq = VFNMS(LDK(KP980785280), Td6, Td3); STM4(&(io[46]), Tfq, ovs, &(io[0])); Tfr = VFNMS(LDK(KP980785280), Td2, Td1); STM4(&(io[30]), Tfr, ovs, &(io[0])); Tfs = VFMA(LDK(KP980785280), Td2, Td1); STM4(&(io[62]), Tfs, ovs, &(io[0])); } } } } { V Tft, Tfu, Tfv, Tfw, Tfx, Tfy, Tfz, TfA, TfB, TfC, TfD, TfE, TfF, TfG, T3f; V T66, T63, T3u, TfL, TfM, TfN, TfO, TfP, TfQ, TfR, TfS, TfT, TfU, TfV, TfW; V TfX, TfY, TfZ, Tg0, Tc5, Tc8; { V TbH, TbG, Tbw, Tbn, TbV, TaW, TbR, Tap, Tc1, TbF, Tc2, TaE, Tc7, TbZ, Tb5; V TbU; { V Taw, TaD, TbX, TbY; TbH = VFMA(LDK(KP414213562), Tas, Tav); Taw = VFNMS(LDK(KP414213562), Tav, Tas); TaD = VFMA(LDK(KP414213562), TaC, Taz); TbG = VFNMS(LDK(KP414213562), Taz, TaC); Tbw = VFNMS(LDK(KP707106781), Tbv, Tbs); TbX = VFMA(LDK(KP707106781), Tbv, Tbs); TbY = VFMA(LDK(KP707106781), Tbm, Tbb); Tbn = VFNMS(LDK(KP707106781), Tbm, Tbb); TbV = VFMA(LDK(KP707106781), TaV, TaK); TaW = VFNMS(LDK(KP707106781), TaV, TaK); Tft = VFMA(LDK(KP980785280), Td8, Td7); STM4(&(ro[62]), Tft, ovs, &(ro[0])); Tfu = VFNMS(LDK(KP980785280), Td8, Td7); STM4(&(ro[30]), Tfu, ovs, &(ro[0])); TbR = VFMA(LDK(KP707106781), Tao, Tah); Tap = VFNMS(LDK(KP707106781), Tao, Tah); Tc1 = VFMA(LDK(KP707106781), TbE, TbB); TbF = VFNMS(LDK(KP707106781), TbE, TbB); Tc2 = VADD(Taw, TaD); TaE = VSUB(Taw, TaD); Tc7 = VFMA(LDK(KP198912367), TbX, TbY); TbZ = VFNMS(LDK(KP198912367), TbY, TbX); Tb5 = VFNMS(LDK(KP707106781), Tb4, Tb1); TbU = VFMA(LDK(KP707106781), Tb4, Tb1); } { V TbP, TaF, TbN, Tb6, TbS, TbI, Tc6, TbW, TbM, Tbx; TbP = VFNMS(LDK(KP923879532), TaE, Tap); TaF = VFMA(LDK(KP923879532), TaE, Tap); TbN = VFNMS(LDK(KP668178637), TaW, Tb5); Tb6 = VFMA(LDK(KP668178637), Tb5, TaW); TbS = VADD(TbH, TbG); TbI = VSUB(TbG, TbH); Tc6 = VFNMS(LDK(KP198912367), TbU, TbV); TbW = VFMA(LDK(KP198912367), TbV, TbU); TbM = VFMA(LDK(KP668178637), Tbn, Tbw); Tbx = VFNMS(LDK(KP668178637), Tbw, Tbn); { V Tc3, Tc9, TbT, TbL, TbJ, Tc4, Tc0, TbQ, TbO, TbK, Tby, Tca; Tc3 = VFNMS(LDK(KP923879532), Tc2, Tc1); Tc9 = VFMA(LDK(KP923879532), Tc2, Tc1); TbT = VFMA(LDK(KP923879532), TbS, TbR); Tc5 = VFNMS(LDK(KP923879532), TbS, TbR); TbL = VFMA(LDK(KP923879532), TbI, TbF); TbJ = VFNMS(LDK(KP923879532), TbI, TbF); Tc4 = VSUB(TbZ, TbW); Tc0 = VADD(TbW, TbZ); TbQ = VADD(TbN, TbM); TbO = VSUB(TbM, TbN); TbK = VADD(Tb6, Tbx); Tby = VSUB(Tb6, Tbx); Tca = VADD(Tc6, Tc7); Tc8 = VSUB(Tc6, Tc7); Tfv = VFMA(LDK(KP980785280), Tc0, TbT); STM4(&(ro[2]), Tfv, ovs, &(ro[0])); Tfw = VFNMS(LDK(KP980785280), Tc0, TbT); STM4(&(ro[34]), Tfw, ovs, &(ro[0])); Tfx = VFMA(LDK(KP831469612), TbQ, TbP); STM4(&(ro[58]), Tfx, ovs, &(ro[0])); Tfy = VFNMS(LDK(KP831469612), TbQ, TbP); STM4(&(ro[26]), Tfy, ovs, &(ro[0])); Tfz = VFMA(LDK(KP831469612), TbO, TbL); STM4(&(io[10]), Tfz, ovs, &(io[0])); TfA = VFNMS(LDK(KP831469612), TbO, TbL); STM4(&(io[42]), TfA, ovs, &(io[0])); TfB = VFMA(LDK(KP831469612), TbK, TbJ); STM4(&(io[58]), TfB, ovs, &(io[0])); TfC = VFNMS(LDK(KP831469612), TbK, TbJ); STM4(&(io[26]), TfC, ovs, &(io[0])); TfD = VFMA(LDK(KP831469612), Tby, TaF); STM4(&(ro[10]), TfD, ovs, &(ro[0])); TfE = VFNMS(LDK(KP831469612), Tby, TaF); STM4(&(ro[42]), TfE, ovs, &(ro[0])); TfF = VFMA(LDK(KP980785280), Tca, Tc9); STM4(&(io[2]), TfF, ovs, &(io[0])); TfG = VFNMS(LDK(KP980785280), Tca, Tc9); STM4(&(io[34]), TfG, ovs, &(io[0])); TfH = VFNMS(LDK(KP980785280), Tc4, Tc3); STM4(&(io[50]), TfH, ovs, &(io[0])); TfI = VFMA(LDK(KP980785280), Tc4, Tc3); STM4(&(io[18]), TfI, ovs, &(io[0])); } } } { V T70, T6X, T7h, T6F, T7x, T7m, T7w, T7p, T7s, T6M, T7c, T6U, T7r, T75, T7i; V T78, T7b, T6N; { V T6T, T6Q, T77, T6I, T6L, T76, T73, T74; { V T6D, T6E, T7k, T7l, T7n, T7o; T3f = VFMA(LDK(KP707106781), T3e, T37); T6D = VFNMS(LDK(KP707106781), T3e, T37); T6E = VADD(T65, T64); T66 = VSUB(T64, T65); T6T = VFNMS(LDK(KP923879532), T6S, T6R); T7k = VFMA(LDK(KP923879532), T6S, T6R); T7l = VFMA(LDK(KP923879532), T6P, T6O); T6Q = VFNMS(LDK(KP923879532), T6P, T6O); T70 = VFNMS(LDK(KP923879532), T6Z, T6Y); T7n = VFMA(LDK(KP923879532), T6Z, T6Y); T7o = VFMA(LDK(KP923879532), T6W, T6V); T6X = VFNMS(LDK(KP923879532), T6W, T6V); T77 = VFNMS(LDK(KP198912367), T6G, T6H); T6I = VFMA(LDK(KP198912367), T6H, T6G); TfJ = VFMA(LDK(KP980785280), Tc8, Tc5); STM4(&(ro[18]), TfJ, ovs, &(ro[0])); TfK = VFNMS(LDK(KP980785280), Tc8, Tc5); STM4(&(ro[50]), TfK, ovs, &(ro[0])); T7h = VFMA(LDK(KP923879532), T6E, T6D); T6F = VFNMS(LDK(KP923879532), T6E, T6D); T7x = VFNMS(LDK(KP098491403), T7k, T7l); T7m = VFMA(LDK(KP098491403), T7l, T7k); T7w = VFMA(LDK(KP098491403), T7n, T7o); T7p = VFNMS(LDK(KP098491403), T7o, T7n); T6L = VFNMS(LDK(KP198912367), T6K, T6J); T76 = VFMA(LDK(KP198912367), T6J, T6K); } T63 = VFMA(LDK(KP707106781), T62, T5Z); T73 = VFNMS(LDK(KP707106781), T62, T5Z); T74 = VADD(T3m, T3t); T3u = VSUB(T3m, T3t); T7s = VADD(T6I, T6L); T6M = VSUB(T6I, T6L); T7c = VFNMS(LDK(KP820678790), T6Q, T6T); T6U = VFMA(LDK(KP820678790), T6T, T6Q); T7r = VFMA(LDK(KP923879532), T74, T73); T75 = VFNMS(LDK(KP923879532), T74, T73); T7i = VADD(T77, T76); T78 = VSUB(T76, T77); } T7b = VFNMS(LDK(KP980785280), T6M, T6F); T6N = VFMA(LDK(KP980785280), T6M, T6F); { V T7u, T7q, T7v, T7t, T7A, T7y, T7j, T7z, T7f, T79, T71, T7d; T7u = VADD(T7m, T7p); T7q = VSUB(T7m, T7p); T7v = VFNMS(LDK(KP980785280), T7s, T7r); T7t = VFMA(LDK(KP980785280), T7s, T7r); T7A = VADD(T7x, T7w); T7y = VSUB(T7w, T7x); T7j = VFNMS(LDK(KP980785280), T7i, T7h); T7z = VFMA(LDK(KP980785280), T7i, T7h); T7f = VFMA(LDK(KP980785280), T78, T75); T79 = VFNMS(LDK(KP980785280), T78, T75); T71 = VFNMS(LDK(KP820678790), T70, T6X); T7d = VFMA(LDK(KP820678790), T6X, T70); { V T7g, T7e, T72, T7a; TfL = VFMA(LDK(KP995184726), T7y, T7v); STM4(&(io[15]), TfL, ovs, &(io[1])); TfM = VFNMS(LDK(KP995184726), T7y, T7v); STM4(&(io[47]), TfM, ovs, &(io[1])); TfN = VFMA(LDK(KP995184726), T7q, T7j); STM4(&(ro[15]), TfN, ovs, &(ro[1])); TfO = VFNMS(LDK(KP995184726), T7q, T7j); STM4(&(ro[47]), TfO, ovs, &(ro[1])); T7g = VADD(T7c, T7d); T7e = VSUB(T7c, T7d); T72 = VADD(T6U, T71); T7a = VSUB(T71, T6U); TfP = VFNMS(LDK(KP995184726), T7u, T7t); STM4(&(io[31]), TfP, ovs, &(io[1])); TfQ = VFMA(LDK(KP995184726), T7u, T7t); STM4(&(io[63]), TfQ, ovs, &(io[1])); TfR = VFMA(LDK(KP773010453), T7e, T7b); STM4(&(ro[23]), TfR, ovs, &(ro[1])); TfS = VFNMS(LDK(KP773010453), T7e, T7b); STM4(&(ro[55]), TfS, ovs, &(ro[1])); TfT = VFMA(LDK(KP773010453), T7g, T7f); STM4(&(io[7]), TfT, ovs, &(io[1])); TfU = VFNMS(LDK(KP773010453), T7g, T7f); STM4(&(io[39]), TfU, ovs, &(io[1])); TfV = VFMA(LDK(KP773010453), T7a, T79); STM4(&(io[23]), TfV, ovs, &(io[1])); TfW = VFNMS(LDK(KP773010453), T7a, T79); STM4(&(io[55]), TfW, ovs, &(io[1])); TfX = VFMA(LDK(KP773010453), T72, T6N); STM4(&(ro[7]), TfX, ovs, &(ro[1])); TfY = VFNMS(LDK(KP773010453), T72, T6N); STM4(&(ro[39]), TfY, ovs, &(ro[1])); TfZ = VFNMS(LDK(KP995184726), T7A, T7z); STM4(&(ro[31]), TfZ, ovs, &(ro[1])); Tg0 = VFMA(LDK(KP995184726), T7A, T7z); STM4(&(ro[63]), Tg0, ovs, &(ro[1])); } } } { V T7D, T8K, T8H, T7K, Ta8, Ta7, Tae, Tad; { V T9x, T9u, T9E, T9B, T9L, T9K, T9V, T9j, Tab, Ta0, Taa, Ta3, Ta6, T9q, T9H; V T9I; { V T9h, T9i, T9Y, T9Z, Ta1, Ta2, T9m, T9p; T7D = VFMA(LDK(KP707106781), T7C, T7B); T9h = VFNMS(LDK(KP707106781), T7C, T7B); T9i = VSUB(T8I, T8J); T8K = VADD(T8I, T8J); T9x = VFNMS(LDK(KP923879532), T9w, T9v); T9Y = VFMA(LDK(KP923879532), T9w, T9v); T9Z = VFMA(LDK(KP923879532), T9t, T9s); T9u = VFNMS(LDK(KP923879532), T9t, T9s); T9E = VFNMS(LDK(KP923879532), T9D, T9C); Ta1 = VFMA(LDK(KP923879532), T9D, T9C); Ta2 = VFMA(LDK(KP923879532), T9A, T9z); T9B = VFNMS(LDK(KP923879532), T9A, T9z); T9L = VFNMS(LDK(KP668178637), T9k, T9l); T9m = VFMA(LDK(KP668178637), T9l, T9k); T9p = VFNMS(LDK(KP668178637), T9o, T9n); T9K = VFMA(LDK(KP668178637), T9n, T9o); T9V = VFNMS(LDK(KP923879532), T9i, T9h); T9j = VFMA(LDK(KP923879532), T9i, T9h); Tab = VFNMS(LDK(KP303346683), T9Y, T9Z); Ta0 = VFMA(LDK(KP303346683), T9Z, T9Y); Taa = VFMA(LDK(KP303346683), Ta1, Ta2); Ta3 = VFNMS(LDK(KP303346683), Ta2, Ta1); Ta6 = VADD(T9m, T9p); T9q = VSUB(T9m, T9p); T8H = VFMA(LDK(KP707106781), T8G, T8F); T9H = VFNMS(LDK(KP707106781), T8G, T8F); T9I = VSUB(T7J, T7G); T7K = VADD(T7G, T7J); } { V T9P, T9r, T9Q, T9y, Ta5, T9J, T9W, T9M, T9R, T9F; T9P = VFNMS(LDK(KP831469612), T9q, T9j); T9r = VFMA(LDK(KP831469612), T9q, T9j); T9Q = VFNMS(LDK(KP534511135), T9u, T9x); T9y = VFMA(LDK(KP534511135), T9x, T9u); Ta5 = VFNMS(LDK(KP923879532), T9I, T9H); T9J = VFMA(LDK(KP923879532), T9I, T9H); T9W = VADD(T9L, T9K); T9M = VSUB(T9K, T9L); T9R = VFMA(LDK(KP534511135), T9B, T9E); T9F = VFNMS(LDK(KP534511135), T9E, T9B); { V T9T, T9N, T9U, T9S, T9G, T9O; { V Ta4, Ta9, Tac, T9X; Ta8 = VADD(Ta0, Ta3); Ta4 = VSUB(Ta0, Ta3); Ta9 = VFNMS(LDK(KP831469612), Ta6, Ta5); Ta7 = VFMA(LDK(KP831469612), Ta6, Ta5); Tae = VADD(Tab, Taa); Tac = VSUB(Taa, Tab); T9X = VFNMS(LDK(KP831469612), T9W, T9V); Tad = VFMA(LDK(KP831469612), T9W, T9V); T9T = VFMA(LDK(KP831469612), T9M, T9J); T9N = VFNMS(LDK(KP831469612), T9M, T9J); T9U = VADD(T9Q, T9R); T9S = VSUB(T9Q, T9R); T9G = VADD(T9y, T9F); T9O = VSUB(T9F, T9y); { V Tg1, Tg2, Tg3, Tg4; Tg1 = VFNMS(LDK(KP956940335), Tac, Ta9); STM4(&(io[45]), Tg1, ovs, &(io[1])); STN4(&(io[44]), Tf6, Tg1, Tfq, TfM, ovs); Tg2 = VFMA(LDK(KP956940335), Ta4, T9X); STM4(&(ro[13]), Tg2, ovs, &(ro[1])); STN4(&(ro[12]), Tf9, Tg2, Tff, TfN, ovs); Tg3 = VFNMS(LDK(KP956940335), Ta4, T9X); STM4(&(ro[45]), Tg3, ovs, &(ro[1])); STN4(&(ro[44]), Tfa, Tg3, Tfg, TfO, ovs); Tg4 = VFMA(LDK(KP956940335), Tac, Ta9); STM4(&(io[13]), Tg4, ovs, &(io[1])); STN4(&(io[12]), Tf5, Tg4, Tfp, TfL, ovs); } } { V Tg5, Tg6, Tg7, Tg8; Tg5 = VFMA(LDK(KP881921264), T9S, T9P); STM4(&(ro[21]), Tg5, ovs, &(ro[1])); STN4(&(ro[20]), Tfd, Tg5, Tfh, TfR, ovs); Tg6 = VFNMS(LDK(KP881921264), T9S, T9P); STM4(&(ro[53]), Tg6, ovs, &(ro[1])); STN4(&(ro[52]), Tfe, Tg6, Tfi, TfS, ovs); Tg7 = VFMA(LDK(KP881921264), T9U, T9T); STM4(&(io[5]), Tg7, ovs, &(io[1])); STN4(&(io[4]), TeZ, Tg7, Tfj, TfT, ovs); Tg8 = VFNMS(LDK(KP881921264), T9U, T9T); STM4(&(io[37]), Tg8, ovs, &(io[1])); STN4(&(io[36]), Tf0, Tg8, Tfk, TfU, ovs); { V Tg9, Tga, Tgb, Tgc; Tg9 = VFMA(LDK(KP881921264), T9O, T9N); STM4(&(io[21]), Tg9, ovs, &(io[1])); STN4(&(io[20]), Tfb, Tg9, Tfl, TfV, ovs); Tga = VFNMS(LDK(KP881921264), T9O, T9N); STM4(&(io[53]), Tga, ovs, &(io[1])); STN4(&(io[52]), Tfc, Tga, Tfm, TfW, ovs); Tgb = VFMA(LDK(KP881921264), T9G, T9r); STM4(&(ro[5]), Tgb, ovs, &(ro[1])); STN4(&(ro[4]), Tf1, Tgb, Tfn, TfX, ovs); Tgc = VFNMS(LDK(KP881921264), T9G, T9r); STM4(&(ro[37]), Tgc, ovs, &(ro[1])); STN4(&(ro[36]), Tf2, Tgc, Tfo, TfY, ovs); } } } } } { V Tgh, Tgi, Tgl, Tgm, Tgn, Tgo, Tgp, Tgq, Tgr, Tgs, Tgt, Tgu; { V T5U, T6j, T3v, T6y, T6o, T5H, T69, T68, T6z, T6r, T6u, T48, T6f, T52, T6t; V T67, T6h, T49; { V T51, T4O, T6p, T6q, T3O, T47, T6m, T6n; T51 = VFNMS(LDK(KP923879532), T50, T4X); T6m = VFMA(LDK(KP923879532), T50, T4X); T6n = VFMA(LDK(KP923879532), T4N, T4q); T4O = VFNMS(LDK(KP923879532), T4N, T4q); T5U = VFNMS(LDK(KP923879532), T5T, T5Q); T6p = VFMA(LDK(KP923879532), T5T, T5Q); { V Tgd, Tge, Tgf, Tgg; Tgd = VFMA(LDK(KP956940335), Ta8, Ta7); STM4(&(io[61]), Tgd, ovs, &(io[1])); STN4(&(io[60]), Tf7, Tgd, Tfs, TfQ, ovs); Tge = VFNMS(LDK(KP956940335), Ta8, Ta7); STM4(&(io[29]), Tge, ovs, &(io[1])); STN4(&(io[28]), Tf8, Tge, Tfr, TfP, ovs); Tgf = VFMA(LDK(KP956940335), Tae, Tad); STM4(&(ro[61]), Tgf, ovs, &(ro[1])); STN4(&(ro[60]), Tf3, Tgf, Tft, Tg0, ovs); Tgg = VFNMS(LDK(KP956940335), Tae, Tad); STM4(&(ro[29]), Tgg, ovs, &(ro[1])); STN4(&(ro[28]), Tf4, Tgg, Tfu, TfZ, ovs); T6j = VFMA(LDK(KP923879532), T3u, T3f); T3v = VFNMS(LDK(KP923879532), T3u, T3f); T6y = VFNMS(LDK(KP303346683), T6m, T6n); T6o = VFMA(LDK(KP303346683), T6n, T6m); T6q = VFMA(LDK(KP923879532), T5G, T5j); T5H = VFNMS(LDK(KP923879532), T5G, T5j); } T69 = VFMA(LDK(KP668178637), T3G, T3N); T3O = VFNMS(LDK(KP668178637), T3N, T3G); T47 = VFMA(LDK(KP668178637), T46, T3Z); T68 = VFNMS(LDK(KP668178637), T3Z, T46); T6z = VFMA(LDK(KP303346683), T6p, T6q); T6r = VFNMS(LDK(KP303346683), T6q, T6p); T6u = VADD(T3O, T47); T48 = VSUB(T3O, T47); T6f = VFNMS(LDK(KP534511135), T4O, T51); T52 = VFMA(LDK(KP534511135), T51, T4O); T6t = VFMA(LDK(KP923879532), T66, T63); T67 = VFNMS(LDK(KP923879532), T66, T63); } T6h = VFNMS(LDK(KP831469612), T48, T3v); T49 = VFMA(LDK(KP831469612), T48, T3v); { V T6w, T6s, T6B, T6v, T6A, T6C, T6k, T6a, T6e, T5V; T6w = VSUB(T6r, T6o); T6s = VADD(T6o, T6r); T6B = VFMA(LDK(KP831469612), T6u, T6t); T6v = VFNMS(LDK(KP831469612), T6u, T6t); T6A = VSUB(T6y, T6z); T6C = VADD(T6y, T6z); T6k = VADD(T69, T68); T6a = VSUB(T68, T69); T6e = VFMA(LDK(KP534511135), T5H, T5U); T5V = VFNMS(LDK(KP534511135), T5U, T5H); Tgh = VFMA(LDK(KP956940335), T6C, T6B); STM4(&(io[3]), Tgh, ovs, &(io[1])); Tgi = VFNMS(LDK(KP956940335), T6C, T6B); STM4(&(io[35]), Tgi, ovs, &(io[1])); { V T6l, T6x, T6d, T6b; T6l = VFMA(LDK(KP831469612), T6k, T6j); T6x = VFNMS(LDK(KP831469612), T6k, T6j); T6d = VFMA(LDK(KP831469612), T6a, T67); T6b = VFNMS(LDK(KP831469612), T6a, T67); { V T6g, T6i, T5W, T6c; T6g = VSUB(T6e, T6f); T6i = VADD(T6f, T6e); T5W = VSUB(T52, T5V); T6c = VADD(T52, T5V); Tgj = VFMA(LDK(KP956940335), T6w, T6v); STM4(&(io[19]), Tgj, ovs, &(io[1])); Tgk = VFNMS(LDK(KP956940335), T6w, T6v); STM4(&(io[51]), Tgk, ovs, &(io[1])); Tgl = VFMA(LDK(KP956940335), T6s, T6l); STM4(&(ro[3]), Tgl, ovs, &(ro[1])); Tgm = VFNMS(LDK(KP956940335), T6s, T6l); STM4(&(ro[35]), Tgm, ovs, &(ro[1])); Tgn = VFMA(LDK(KP881921264), T6i, T6h); STM4(&(ro[59]), Tgn, ovs, &(ro[1])); Tgo = VFNMS(LDK(KP881921264), T6i, T6h); STM4(&(ro[27]), Tgo, ovs, &(ro[1])); Tgp = VFMA(LDK(KP881921264), T6g, T6d); STM4(&(io[11]), Tgp, ovs, &(io[1])); Tgq = VFNMS(LDK(KP881921264), T6g, T6d); STM4(&(io[43]), Tgq, ovs, &(io[1])); Tgr = VFMA(LDK(KP881921264), T6c, T6b); STM4(&(io[59]), Tgr, ovs, &(io[1])); Tgs = VFNMS(LDK(KP881921264), T6c, T6b); STM4(&(io[27]), Tgs, ovs, &(io[1])); Tgt = VFMA(LDK(KP881921264), T5W, T49); STM4(&(ro[11]), Tgt, ovs, &(ro[1])); Tgu = VFNMS(LDK(KP881921264), T5W, T49); STM4(&(ro[43]), Tgu, ovs, &(ro[1])); Tgv = VFNMS(LDK(KP956940335), T6A, T6x); STM4(&(ro[51]), Tgv, ovs, &(ro[1])); Tgw = VFMA(LDK(KP956940335), T6A, T6x); STM4(&(ro[19]), Tgw, ovs, &(ro[1])); } } } } { V T8j, T8c, T8C, T8v, T8N, T8M, T8X, T7L, T9c, T92, T9d, T95, T98, T80; { V T90, T91, T93, T94, T7S, T7Z; T8j = VFNMS(LDK(KP923879532), T8i, T8f); T90 = VFMA(LDK(KP923879532), T8i, T8f); T91 = VFMA(LDK(KP923879532), T8b, T84); T8c = VFNMS(LDK(KP923879532), T8b, T84); T8C = VFNMS(LDK(KP923879532), T8B, T8y); T93 = VFMA(LDK(KP923879532), T8B, T8y); T94 = VFMA(LDK(KP923879532), T8u, T8n); T8v = VFNMS(LDK(KP923879532), T8u, T8n); T8N = VFMA(LDK(KP198912367), T7O, T7R); T7S = VFNMS(LDK(KP198912367), T7R, T7O); T7Z = VFMA(LDK(KP198912367), T7Y, T7V); T8M = VFNMS(LDK(KP198912367), T7V, T7Y); T8X = VFMA(LDK(KP923879532), T7K, T7D); T7L = VFNMS(LDK(KP923879532), T7K, T7D); T9c = VFNMS(LDK(KP098491403), T90, T91); T92 = VFMA(LDK(KP098491403), T91, T90); T9d = VFMA(LDK(KP098491403), T93, T94); T95 = VFNMS(LDK(KP098491403), T94, T93); T98 = VADD(T7S, T7Z); T80 = VSUB(T7S, T7Z); } { V T8V, T81, T8T, T8k, T97, T8L, T8Y, T8O, T8S, T8D; T8V = VFNMS(LDK(KP980785280), T80, T7L); T81 = VFMA(LDK(KP980785280), T80, T7L); T8T = VFNMS(LDK(KP820678790), T8c, T8j); T8k = VFMA(LDK(KP820678790), T8j, T8c); T97 = VFMA(LDK(KP923879532), T8K, T8H); T8L = VFNMS(LDK(KP923879532), T8K, T8H); T8Y = VADD(T8N, T8M); T8O = VSUB(T8M, T8N); T8S = VFMA(LDK(KP820678790), T8v, T8C); T8D = VFNMS(LDK(KP820678790), T8C, T8v); { V T8R, T8P, T8U, T8W, T8E, T8Q; { V T96, T9f, T9g, T8Z; T9a = VSUB(T95, T92); T96 = VADD(T92, T95); T9f = VFMA(LDK(KP980785280), T98, T97); T99 = VFNMS(LDK(KP980785280), T98, T97); T9e = VSUB(T9c, T9d); T9g = VADD(T9c, T9d); T8Z = VFMA(LDK(KP980785280), T8Y, T8X); T9b = VFNMS(LDK(KP980785280), T8Y, T8X); T8R = VFMA(LDK(KP980785280), T8O, T8L); T8P = VFNMS(LDK(KP980785280), T8O, T8L); T8U = VSUB(T8S, T8T); T8W = VADD(T8T, T8S); T8E = VSUB(T8k, T8D); T8Q = VADD(T8k, T8D); { V Tgx, Tgy, Tgz, TgA; Tgx = VFNMS(LDK(KP995184726), T9g, T9f); STM4(&(io[33]), Tgx, ovs, &(io[1])); STN4(&(io[32]), TeO, Tgx, TfG, Tgi, ovs); Tgy = VFMA(LDK(KP995184726), T96, T8Z); STM4(&(ro[1]), Tgy, ovs, &(ro[1])); STN4(&(ro[0]), TeL, Tgy, Tfv, Tgl, ovs); Tgz = VFNMS(LDK(KP995184726), T96, T8Z); STM4(&(ro[33]), Tgz, ovs, &(ro[1])); STN4(&(ro[32]), TeM, Tgz, Tfw, Tgm, ovs); TgA = VFMA(LDK(KP995184726), T9g, T9f); STM4(&(io[1]), TgA, ovs, &(io[1])); STN4(&(io[0]), TeN, TgA, TfF, Tgh, ovs); } } { V TgB, TgC, TgD, TgE; TgB = VFMA(LDK(KP773010453), T8W, T8V); STM4(&(ro[57]), TgB, ovs, &(ro[1])); STN4(&(ro[56]), TeS, TgB, Tfx, Tgn, ovs); TgC = VFNMS(LDK(KP773010453), T8W, T8V); STM4(&(ro[25]), TgC, ovs, &(ro[1])); STN4(&(ro[24]), TeR, TgC, Tfy, Tgo, ovs); TgD = VFMA(LDK(KP773010453), T8U, T8R); STM4(&(io[9]), TgD, ovs, &(io[1])); STN4(&(io[8]), TeT, TgD, Tfz, Tgp, ovs); TgE = VFNMS(LDK(KP773010453), T8U, T8R); STM4(&(io[41]), TgE, ovs, &(io[1])); STN4(&(io[40]), TeU, TgE, TfA, Tgq, ovs); { V TgF, TgG, TgH, TgI; TgF = VFMA(LDK(KP773010453), T8Q, T8P); STM4(&(io[57]), TgF, ovs, &(io[1])); STN4(&(io[56]), TeW, TgF, TfB, Tgr, ovs); TgG = VFNMS(LDK(KP773010453), T8Q, T8P); STM4(&(io[25]), TgG, ovs, &(io[1])); STN4(&(io[24]), TeV, TgG, TfC, Tgs, ovs); TgH = VFMA(LDK(KP773010453), T8E, T81); STM4(&(ro[9]), TgH, ovs, &(ro[1])); STN4(&(ro[8]), TeX, TgH, TfD, Tgt, ovs); TgI = VFNMS(LDK(KP773010453), T8E, T81); STM4(&(ro[41]), TgI, ovs, &(ro[1])); STN4(&(ro[40]), TeY, TgI, TfE, Tgu, ovs); } } } } } } } } } } { V TgJ, TgK, TgL, TgM; TgJ = VFMA(LDK(KP995184726), T9a, T99); STM4(&(io[17]), TgJ, ovs, &(io[1])); STN4(&(io[16]), TeQ, TgJ, TfI, Tgj, ovs); TgK = VFNMS(LDK(KP995184726), T9a, T99); STM4(&(io[49]), TgK, ovs, &(io[1])); STN4(&(io[48]), TeP, TgK, TfH, Tgk, ovs); TgL = VFMA(LDK(KP995184726), T9e, T9b); STM4(&(ro[17]), TgL, ovs, &(ro[1])); STN4(&(ro[16]), TeK, TgL, TfJ, Tgw, ovs); TgM = VFNMS(LDK(KP995184726), T9e, T9b); STM4(&(ro[49]), TgM, ovs, &(ro[1])); STN4(&(ro[48]), TeJ, TgM, TfK, Tgv, ovs); } } } VLEAVE(); } static const kdft_desc desc = { 64, XSIMD_STRING("n2sv_64"), {520, 0, 392, 0}, &GENUS, 0, 1, 0, 0 }; void XSIMD(codelet_n2sv_64) (planner *p) { X(kdft_register) (p, n2sv_64, &desc); } #else /* HAVE_FMA */ /* Generated by: ../../../genfft/gen_notw.native -simd -compact -variables 4 -pipeline-latency 8 -n 64 -name n2sv_64 -with-ostride 1 -include n2s.h -store-multiple 4 */ /* * This function contains 912 FP additions, 248 FP multiplications, * (or, 808 additions, 144 multiplications, 104 fused multiply/add), * 260 stack variables, 15 constants, and 288 memory accesses */ #include "n2s.h" static void n2sv_64(const R *ri, const R *ii, R *ro, R *io, stride is, stride os, INT v, INT ivs, INT ovs) { DVK(KP773010453, +0.773010453362736960810906609758469800971041293); DVK(KP634393284, +0.634393284163645498215171613225493370675687095); DVK(KP098017140, +0.098017140329560601994195563888641845861136673); DVK(KP995184726, +0.995184726672196886244836953109479921575474869); DVK(KP881921264, +0.881921264348355029712756863660388349508442621); DVK(KP471396736, +0.471396736825997648556387625905254377657460319); DVK(KP290284677, +0.290284677254462367636192375817395274691476278); DVK(KP956940335, +0.956940335732208864935797886980269969482849206); DVK(KP831469612, +0.831469612302545237078788377617905756738560812); DVK(KP555570233, +0.555570233019602224742830813948532874374937191); DVK(KP195090322, +0.195090322016128267848284868477022240927691618); DVK(KP980785280, +0.980785280403230449126182236134239036973933731); DVK(KP923879532, +0.923879532511286756128183189396788286822416626); DVK(KP382683432, +0.382683432365089771728459984030398866761344562); DVK(KP707106781, +0.707106781186547524400844362104849039284835938); { INT i; for (i = v; i > 0; i = i - (2 * VL), ri = ri + ((2 * VL) * ivs), ii = ii + ((2 * VL) * ivs), ro = ro + ((2 * VL) * ovs), io = io + ((2 * VL) * ovs), MAKE_VOLATILE_STRIDE(256, is), MAKE_VOLATILE_STRIDE(256, os)) { V T37, T7B, T8F, T5Z, Tf, Td9, TbB, TcB, T62, T7C, T2i, TdH, Tah, Tcb, T3e; V T8G, Tu, TdI, Tak, TbD, Tan, TbC, T2x, Tda, T3m, T65, T7G, T8J, T7J, T8I; V T3t, T64, TK, Tdd, Tas, Tce, Tav, Tcf, T2N, Tdc, T3G, T6G, T7O, T9k, T7R; V T9l, T3N, T6H, T1L, Tdv, Tbs, Tcw, TdC, Teo, T5j, T6V, T5Q, T6Y, T8y, T9C; V Tbb, Tct, T8n, T9z, TZ, Tdf, Taz, Tch, TaC, Tci, T32, Tdg, T3Z, T6J, T7V; V T9n, T7Y, T9o, T46, T6K, T1g, Tdp, Tb1, Tcm, Tdm, Tej, T4q, T6R, T4X, T6O; V T8f, T9s, TaK, Tcp, T84, T9v, T1v, Tdn, Tb4, Tcq, Tds, Tek, T4N, T6P, T50; V T6S, T8i, T9w, TaV, Tcn, T8b, T9t, T20, TdD, Tbv, Tcu, Tdy, Tep, T5G, T6Z; V T5T, T6W, T8B, T9A, Tbm, Tcx, T8u, T9D; { V T3, T35, T26, T5Y, T6, T5X, T29, T36, Ta, T39, T2d, T38, Td, T3b, T2g; V T3c; { V T1, T2, T24, T25; T1 = LD(&(ri[0]), ivs, &(ri[0])); T2 = LD(&(ri[WS(is, 32)]), ivs, &(ri[0])); T3 = VADD(T1, T2); T35 = VSUB(T1, T2); T24 = LD(&(ii[0]), ivs, &(ii[0])); T25 = LD(&(ii[WS(is, 32)]), ivs, &(ii[0])); T26 = VADD(T24, T25); T5Y = VSUB(T24, T25); } { V T4, T5, T27, T28; T4 = LD(&(ri[WS(is, 16)]), ivs, &(ri[0])); T5 = LD(&(ri[WS(is, 48)]), ivs, &(ri[0])); T6 = VADD(T4, T5); T5X = VSUB(T4, T5); T27 = LD(&(ii[WS(is, 16)]), ivs, &(ii[0])); T28 = LD(&(ii[WS(is, 48)]), ivs, &(ii[0])); T29 = VADD(T27, T28); T36 = VSUB(T27, T28); } { V T8, T9, T2b, T2c; T8 = LD(&(ri[WS(is, 8)]), ivs, &(ri[0])); T9 = LD(&(ri[WS(is, 40)]), ivs, &(ri[0])); Ta = VADD(T8, T9); T39 = VSUB(T8, T9); T2b = LD(&(ii[WS(is, 8)]), ivs, &(ii[0])); T2c = LD(&(ii[WS(is, 40)]), ivs, &(ii[0])); T2d = VADD(T2b, T2c); T38 = VSUB(T2b, T2c); } { V Tb, Tc, T2e, T2f; Tb = LD(&(ri[WS(is, 56)]), ivs, &(ri[0])); Tc = LD(&(ri[WS(is, 24)]), ivs, &(ri[0])); Td = VADD(Tb, Tc); T3b = VSUB(Tb, Tc); T2e = LD(&(ii[WS(is, 56)]), ivs, &(ii[0])); T2f = LD(&(ii[WS(is, 24)]), ivs, &(ii[0])); T2g = VADD(T2e, T2f); T3c = VSUB(T2e, T2f); } { V T7, Te, T2a, T2h; T37 = VSUB(T35, T36); T7B = VADD(T35, T36); T8F = VSUB(T5Y, T5X); T5Z = VADD(T5X, T5Y); T7 = VADD(T3, T6); Te = VADD(Ta, Td); Tf = VADD(T7, Te); Td9 = VSUB(T7, Te); { V Tbz, TbA, T60, T61; Tbz = VSUB(T26, T29); TbA = VSUB(Td, Ta); TbB = VSUB(Tbz, TbA); TcB = VADD(TbA, Tbz); T60 = VSUB(T3b, T3c); T61 = VADD(T39, T38); T62 = VMUL(LDK(KP707106781), VSUB(T60, T61)); T7C = VMUL(LDK(KP707106781), VADD(T61, T60)); } T2a = VADD(T26, T29); T2h = VADD(T2d, T2g); T2i = VADD(T2a, T2h); TdH = VSUB(T2a, T2h); { V Taf, Tag, T3a, T3d; Taf = VSUB(T3, T6); Tag = VSUB(T2d, T2g); Tah = VSUB(Taf, Tag); Tcb = VADD(Taf, Tag); T3a = VSUB(T38, T39); T3d = VADD(T3b, T3c); T3e = VMUL(LDK(KP707106781), VSUB(T3a, T3d)); T8G = VMUL(LDK(KP707106781), VADD(T3a, T3d)); } } } { V Ti, T3j, T2l, T3h, Tl, T3g, T2o, T3k, Tp, T3q, T2s, T3o, Ts, T3n, T2v; V T3r; { V Tg, Th, T2j, T2k; Tg = LD(&(ri[WS(is, 4)]), ivs, &(ri[0])); Th = LD(&(ri[WS(is, 36)]), ivs, &(ri[0])); Ti = VADD(Tg, Th); T3j = VSUB(Tg, Th); T2j = LD(&(ii[WS(is, 4)]), ivs, &(ii[0])); T2k = LD(&(ii[WS(is, 36)]), ivs, &(ii[0])); T2l = VADD(T2j, T2k); T3h = VSUB(T2j, T2k); } { V Tj, Tk, T2m, T2n; Tj = LD(&(ri[WS(is, 20)]), ivs, &(ri[0])); Tk = LD(&(ri[WS(is, 52)]), ivs, &(ri[0])); Tl = VADD(Tj, Tk); T3g = VSUB(Tj, Tk); T2m = LD(&(ii[WS(is, 20)]), ivs, &(ii[0])); T2n = LD(&(ii[WS(is, 52)]), ivs, &(ii[0])); T2o = VADD(T2m, T2n); T3k = VSUB(T2m, T2n); } { V Tn, To, T2q, T2r; Tn = LD(&(ri[WS(is, 60)]), ivs, &(ri[0])); To = LD(&(ri[WS(is, 28)]), ivs, &(ri[0])); Tp = VADD(Tn, To); T3q = VSUB(Tn, To); T2q = LD(&(ii[WS(is, 60)]), ivs, &(ii[0])); T2r = LD(&(ii[WS(is, 28)]), ivs, &(ii[0])); T2s = VADD(T2q, T2r); T3o = VSUB(T2q, T2r); } { V Tq, Tr, T2t, T2u; Tq = LD(&(ri[WS(is, 12)]), ivs, &(ri[0])); Tr = LD(&(ri[WS(is, 44)]), ivs, &(ri[0])); Ts = VADD(Tq, Tr); T3n = VSUB(Tq, Tr); T2t = LD(&(ii[WS(is, 12)]), ivs, &(ii[0])); T2u = LD(&(ii[WS(is, 44)]), ivs, &(ii[0])); T2v = VADD(T2t, T2u); T3r = VSUB(T2t, T2u); } { V Tm, Tt, Tai, Taj; Tm = VADD(Ti, Tl); Tt = VADD(Tp, Ts); Tu = VADD(Tm, Tt); TdI = VSUB(Tt, Tm); Tai = VSUB(T2l, T2o); Taj = VSUB(Ti, Tl); Tak = VSUB(Tai, Taj); TbD = VADD(Taj, Tai); } { V Tal, Tam, T2p, T2w; Tal = VSUB(Tp, Ts); Tam = VSUB(T2s, T2v); Tan = VADD(Tal, Tam); TbC = VSUB(Tal, Tam); T2p = VADD(T2l, T2o); T2w = VADD(T2s, T2v); T2x = VADD(T2p, T2w); Tda = VSUB(T2p, T2w); } { V T3i, T3l, T7E, T7F; T3i = VADD(T3g, T3h); T3l = VSUB(T3j, T3k); T3m = VFNMS(LDK(KP923879532), T3l, VMUL(LDK(KP382683432), T3i)); T65 = VFMA(LDK(KP923879532), T3i, VMUL(LDK(KP382683432), T3l)); T7E = VSUB(T3h, T3g); T7F = VADD(T3j, T3k); T7G = VFNMS(LDK(KP382683432), T7F, VMUL(LDK(KP923879532), T7E)); T8J = VFMA(LDK(KP382683432), T7E, VMUL(LDK(KP923879532), T7F)); } { V T7H, T7I, T3p, T3s; T7H = VSUB(T3o, T3n); T7I = VADD(T3q, T3r); T7J = VFMA(LDK(KP923879532), T7H, VMUL(LDK(KP382683432), T7I)); T8I = VFNMS(LDK(KP382683432), T7H, VMUL(LDK(KP923879532), T7I)); T3p = VADD(T3n, T3o); T3s = VSUB(T3q, T3r); T3t = VFMA(LDK(KP382683432), T3p, VMUL(LDK(KP923879532), T3s)); T64 = VFNMS(LDK(KP923879532), T3p, VMUL(LDK(KP382683432), T3s)); } } { V Ty, T3H, T2B, T3x, TB, T3w, T2E, T3I, TI, T3L, T2L, T3B, TF, T3K, T2I; V T3E; { V Tw, Tx, T2C, T2D; Tw = LD(&(ri[WS(is, 2)]), ivs, &(ri[0])); Tx = LD(&(ri[WS(is, 34)]), ivs, &(ri[0])); Ty = VADD(Tw, Tx); T3H = VSUB(Tw, Tx); { V T2z, T2A, Tz, TA; T2z = LD(&(ii[WS(is, 2)]), ivs, &(ii[0])); T2A = LD(&(ii[WS(is, 34)]), ivs, &(ii[0])); T2B = VADD(T2z, T2A); T3x = VSUB(T2z, T2A); Tz = LD(&(ri[WS(is, 18)]), ivs, &(ri[0])); TA = LD(&(ri[WS(is, 50)]), ivs, &(ri[0])); TB = VADD(Tz, TA); T3w = VSUB(Tz, TA); } T2C = LD(&(ii[WS(is, 18)]), ivs, &(ii[0])); T2D = LD(&(ii[WS(is, 50)]), ivs, &(ii[0])); T2E = VADD(T2C, T2D); T3I = VSUB(T2C, T2D); { V TG, TH, T3z, T2J, T2K, T3A; TG = LD(&(ri[WS(is, 58)]), ivs, &(ri[0])); TH = LD(&(ri[WS(is, 26)]), ivs, &(ri[0])); T3z = VSUB(TG, TH); T2J = LD(&(ii[WS(is, 58)]), ivs, &(ii[0])); T2K = LD(&(ii[WS(is, 26)]), ivs, &(ii[0])); T3A = VSUB(T2J, T2K); TI = VADD(TG, TH); T3L = VADD(T3z, T3A); T2L = VADD(T2J, T2K); T3B = VSUB(T3z, T3A); } { V TD, TE, T3C, T2G, T2H, T3D; TD = LD(&(ri[WS(is, 10)]), ivs, &(ri[0])); TE = LD(&(ri[WS(is, 42)]), ivs, &(ri[0])); T3C = VSUB(TD, TE); T2G = LD(&(ii[WS(is, 10)]), ivs, &(ii[0])); T2H = LD(&(ii[WS(is, 42)]), ivs, &(ii[0])); T3D = VSUB(T2G, T2H); TF = VADD(TD, TE); T3K = VSUB(T3D, T3C); T2I = VADD(T2G, T2H); T3E = VADD(T3C, T3D); } } { V TC, TJ, Taq, Tar; TC = VADD(Ty, TB); TJ = VADD(TF, TI); TK = VADD(TC, TJ); Tdd = VSUB(TC, TJ); Taq = VSUB(T2B, T2E); Tar = VSUB(TI, TF); Tas = VSUB(Taq, Tar); Tce = VADD(Tar, Taq); } { V Tat, Tau, T2F, T2M; Tat = VSUB(Ty, TB); Tau = VSUB(T2I, T2L); Tav = VSUB(Tat, Tau); Tcf = VADD(Tat, Tau); T2F = VADD(T2B, T2E); T2M = VADD(T2I, T2L); T2N = VADD(T2F, T2M); Tdc = VSUB(T2F, T2M); } { V T3y, T3F, T7M, T7N; T3y = VADD(T3w, T3x); T3F = VMUL(LDK(KP707106781), VSUB(T3B, T3E)); T3G = VSUB(T3y, T3F); T6G = VADD(T3y, T3F); T7M = VSUB(T3x, T3w); T7N = VMUL(LDK(KP707106781), VADD(T3K, T3L)); T7O = VSUB(T7M, T7N); T9k = VADD(T7M, T7N); } { V T7P, T7Q, T3J, T3M; T7P = VADD(T3H, T3I); T7Q = VMUL(LDK(KP707106781), VADD(T3E, T3B)); T7R = VSUB(T7P, T7Q); T9l = VADD(T7P, T7Q); T3J = VSUB(T3H, T3I); T3M = VMUL(LDK(KP707106781), VSUB(T3K, T3L)); T3N = VSUB(T3J, T3M); T6H = VADD(T3J, T3M); } } { V T1z, T53, T5L, Tbo, T1C, T5I, T56, Tbp, T1J, Tb9, T5h, T5N, T1G, Tb8, T5c; V T5O; { V T1x, T1y, T54, T55; T1x = LD(&(ri[WS(is, 63)]), ivs, &(ri[WS(is, 1)])); T1y = LD(&(ri[WS(is, 31)]), ivs, &(ri[WS(is, 1)])); T1z = VADD(T1x, T1y); T53 = VSUB(T1x, T1y); { V T5J, T5K, T1A, T1B; T5J = LD(&(ii[WS(is, 63)]), ivs, &(ii[WS(is, 1)])); T5K = LD(&(ii[WS(is, 31)]), ivs, &(ii[WS(is, 1)])); T5L = VSUB(T5J, T5K); Tbo = VADD(T5J, T5K); T1A = LD(&(ri[WS(is, 15)]), ivs, &(ri[WS(is, 1)])); T1B = LD(&(ri[WS(is, 47)]), ivs, &(ri[WS(is, 1)])); T1C = VADD(T1A, T1B); T5I = VSUB(T1A, T1B); } T54 = LD(&(ii[WS(is, 15)]), ivs, &(ii[WS(is, 1)])); T55 = LD(&(ii[WS(is, 47)]), ivs, &(ii[WS(is, 1)])); T56 = VSUB(T54, T55); Tbp = VADD(T54, T55); { V T1H, T1I, T5d, T5e, T5f, T5g; T1H = LD(&(ri[WS(is, 55)]), ivs, &(ri[WS(is, 1)])); T1I = LD(&(ri[WS(is, 23)]), ivs, &(ri[WS(is, 1)])); T5d = VSUB(T1H, T1I); T5e = LD(&(ii[WS(is, 55)]), ivs, &(ii[WS(is, 1)])); T5f = LD(&(ii[WS(is, 23)]), ivs, &(ii[WS(is, 1)])); T5g = VSUB(T5e, T5f); T1J = VADD(T1H, T1I); Tb9 = VADD(T5e, T5f); T5h = VADD(T5d, T5g); T5N = VSUB(T5d, T5g); } { V T1E, T1F, T5b, T58, T59, T5a; T1E = LD(&(ri[WS(is, 7)]), ivs, &(ri[WS(is, 1)])); T1F = LD(&(ri[WS(is, 39)]), ivs, &(ri[WS(is, 1)])); T5b = VSUB(T1E, T1F); T58 = LD(&(ii[WS(is, 7)]), ivs, &(ii[WS(is, 1)])); T59 = LD(&(ii[WS(is, 39)]), ivs, &(ii[WS(is, 1)])); T5a = VSUB(T58, T59); T1G = VADD(T1E, T1F); Tb8 = VADD(T58, T59); T5c = VSUB(T5a, T5b); T5O = VADD(T5b, T5a); } } { V T1D, T1K, Tbq, Tbr; T1D = VADD(T1z, T1C); T1K = VADD(T1G, T1J); T1L = VADD(T1D, T1K); Tdv = VSUB(T1D, T1K); Tbq = VSUB(Tbo, Tbp); Tbr = VSUB(T1J, T1G); Tbs = VSUB(Tbq, Tbr); Tcw = VADD(Tbr, Tbq); } { V TdA, TdB, T57, T5i; TdA = VADD(Tbo, Tbp); TdB = VADD(Tb8, Tb9); TdC = VSUB(TdA, TdB); Teo = VADD(TdA, TdB); T57 = VSUB(T53, T56); T5i = VMUL(LDK(KP707106781), VSUB(T5c, T5h)); T5j = VSUB(T57, T5i); T6V = VADD(T57, T5i); } { V T5M, T5P, T8w, T8x; T5M = VADD(T5I, T5L); T5P = VMUL(LDK(KP707106781), VSUB(T5N, T5O)); T5Q = VSUB(T5M, T5P); T6Y = VADD(T5M, T5P); T8w = VSUB(T5L, T5I); T8x = VMUL(LDK(KP707106781), VADD(T5c, T5h)); T8y = VSUB(T8w, T8x); T9C = VADD(T8w, T8x); } { V Tb7, Tba, T8l, T8m; Tb7 = VSUB(T1z, T1C); Tba = VSUB(Tb8, Tb9); Tbb = VSUB(Tb7, Tba); Tct = VADD(Tb7, Tba); T8l = VADD(T53, T56); T8m = VMUL(LDK(KP707106781), VADD(T5O, T5N)); T8n = VSUB(T8l, T8m); T9z = VADD(T8l, T8m); } } { V TN, T40, T2Q, T3Q, TQ, T3P, T2T, T41, TX, T44, T30, T3U, TU, T43, T2X; V T3X; { V TL, TM, T2R, T2S; TL = LD(&(ri[WS(is, 62)]), ivs, &(ri[0])); TM = LD(&(ri[WS(is, 30)]), ivs, &(ri[0])); TN = VADD(TL, TM); T40 = VSUB(TL, TM); { V T2O, T2P, TO, TP; T2O = LD(&(ii[WS(is, 62)]), ivs, &(ii[0])); T2P = LD(&(ii[WS(is, 30)]), ivs, &(ii[0])); T2Q = VADD(T2O, T2P); T3Q = VSUB(T2O, T2P); TO = LD(&(ri[WS(is, 14)]), ivs, &(ri[0])); TP = LD(&(ri[WS(is, 46)]), ivs, &(ri[0])); TQ = VADD(TO, TP); T3P = VSUB(TO, TP); } T2R = LD(&(ii[WS(is, 14)]), ivs, &(ii[0])); T2S = LD(&(ii[WS(is, 46)]), ivs, &(ii[0])); T2T = VADD(T2R, T2S); T41 = VSUB(T2R, T2S); { V TV, TW, T3S, T2Y, T2Z, T3T; TV = LD(&(ri[WS(is, 54)]), ivs, &(ri[0])); TW = LD(&(ri[WS(is, 22)]), ivs, &(ri[0])); T3S = VSUB(TV, TW); T2Y = LD(&(ii[WS(is, 54)]), ivs, &(ii[0])); T2Z = LD(&(ii[WS(is, 22)]), ivs, &(ii[0])); T3T = VSUB(T2Y, T2Z); TX = VADD(TV, TW); T44 = VADD(T3S, T3T); T30 = VADD(T2Y, T2Z); T3U = VSUB(T3S, T3T); } { V TS, TT, T3V, T2V, T2W, T3W; TS = LD(&(ri[WS(is, 6)]), ivs, &(ri[0])); TT = LD(&(ri[WS(is, 38)]), ivs, &(ri[0])); T3V = VSUB(TS, TT); T2V = LD(&(ii[WS(is, 6)]), ivs, &(ii[0])); T2W = LD(&(ii[WS(is, 38)]), ivs, &(ii[0])); T3W = VSUB(T2V, T2W); TU = VADD(TS, TT); T43 = VSUB(T3W, T3V); T2X = VADD(T2V, T2W); T3X = VADD(T3V, T3W); } } { V TR, TY, Tax, Tay; TR = VADD(TN, TQ); TY = VADD(TU, TX); TZ = VADD(TR, TY); Tdf = VSUB(TR, TY); Tax = VSUB(T2Q, T2T); Tay = VSUB(TX, TU); Taz = VSUB(Tax, Tay); Tch = VADD(Tay, Tax); } { V TaA, TaB, T2U, T31; TaA = VSUB(TN, TQ); TaB = VSUB(T2X, T30); TaC = VSUB(TaA, TaB); Tci = VADD(TaA, TaB); T2U = VADD(T2Q, T2T); T31 = VADD(T2X, T30); T32 = VADD(T2U, T31); Tdg = VSUB(T2U, T31); } { V T3R, T3Y, T7T, T7U; T3R = VADD(T3P, T3Q); T3Y = VMUL(LDK(KP707106781), VSUB(T3U, T3X)); T3Z = VSUB(T3R, T3Y); T6J = VADD(T3R, T3Y); T7T = VADD(T40, T41); T7U = VMUL(LDK(KP707106781), VADD(T3X, T3U)); T7V = VSUB(T7T, T7U); T9n = VADD(T7T, T7U); } { V T7W, T7X, T42, T45; T7W = VSUB(T3Q, T3P); T7X = VMUL(LDK(KP707106781), VADD(T43, T44)); T7Y = VSUB(T7W, T7X); T9o = VADD(T7W, T7X); T42 = VSUB(T40, T41); T45 = VMUL(LDK(KP707106781), VSUB(T43, T44)); T46 = VSUB(T42, T45); T6K = VADD(T42, T45); } } { V T14, T4P, T4d, TaG, T17, T4a, T4S, TaH, T1e, TaZ, T4j, T4V, T1b, TaY, T4o; V T4U; { V T12, T13, T4Q, T4R; T12 = LD(&(ri[WS(is, 1)]), ivs, &(ri[WS(is, 1)])); T13 = LD(&(ri[WS(is, 33)]), ivs, &(ri[WS(is, 1)])); T14 = VADD(T12, T13); T4P = VSUB(T12, T13); { V T4b, T4c, T15, T16; T4b = LD(&(ii[WS(is, 1)]), ivs, &(ii[WS(is, 1)])); T4c = LD(&(ii[WS(is, 33)]), ivs, &(ii[WS(is, 1)])); T4d = VSUB(T4b, T4c); TaG = VADD(T4b, T4c); T15 = LD(&(ri[WS(is, 17)]), ivs, &(ri[WS(is, 1)])); T16 = LD(&(ri[WS(is, 49)]), ivs, &(ri[WS(is, 1)])); T17 = VADD(T15, T16); T4a = VSUB(T15, T16); } T4Q = LD(&(ii[WS(is, 17)]), ivs, &(ii[WS(is, 1)])); T4R = LD(&(ii[WS(is, 49)]), ivs, &(ii[WS(is, 1)])); T4S = VSUB(T4Q, T4R); TaH = VADD(T4Q, T4R); { V T1c, T1d, T4f, T4g, T4h, T4i; T1c = LD(&(ri[WS(is, 57)]), ivs, &(ri[WS(is, 1)])); T1d = LD(&(ri[WS(is, 25)]), ivs, &(ri[WS(is, 1)])); T4f = VSUB(T1c, T1d); T4g = LD(&(ii[WS(is, 57)]), ivs, &(ii[WS(is, 1)])); T4h = LD(&(ii[WS(is, 25)]), ivs, &(ii[WS(is, 1)])); T4i = VSUB(T4g, T4h); T1e = VADD(T1c, T1d); TaZ = VADD(T4g, T4h); T4j = VSUB(T4f, T4i); T4V = VADD(T4f, T4i); } { V T19, T1a, T4k, T4l, T4m, T4n; T19 = LD(&(ri[WS(is, 9)]), ivs, &(ri[WS(is, 1)])); T1a = LD(&(ri[WS(is, 41)]), ivs, &(ri[WS(is, 1)])); T4k = VSUB(T19, T1a); T4l = LD(&(ii[WS(is, 9)]), ivs, &(ii[WS(is, 1)])); T4m = LD(&(ii[WS(is, 41)]), ivs, &(ii[WS(is, 1)])); T4n = VSUB(T4l, T4m); T1b = VADD(T19, T1a); TaY = VADD(T4l, T4m); T4o = VADD(T4k, T4n); T4U = VSUB(T4n, T4k); } } { V T18, T1f, TaX, Tb0; T18 = VADD(T14, T17); T1f = VADD(T1b, T1e); T1g = VADD(T18, T1f); Tdp = VSUB(T18, T1f); TaX = VSUB(T14, T17); Tb0 = VSUB(TaY, TaZ); Tb1 = VSUB(TaX, Tb0); Tcm = VADD(TaX, Tb0); } { V Tdk, Tdl, T4e, T4p; Tdk = VADD(TaG, TaH); Tdl = VADD(TaY, TaZ); Tdm = VSUB(Tdk, Tdl); Tej = VADD(Tdk, Tdl); T4e = VADD(T4a, T4d); T4p = VMUL(LDK(KP707106781), VSUB(T4j, T4o)); T4q = VSUB(T4e, T4p); T6R = VADD(T4e, T4p); } { V T4T, T4W, T8d, T8e; T4T = VSUB(T4P, T4S); T4W = VMUL(LDK(KP707106781), VSUB(T4U, T4V)); T4X = VSUB(T4T, T4W); T6O = VADD(T4T, T4W); T8d = VADD(T4P, T4S); T8e = VMUL(LDK(KP707106781), VADD(T4o, T4j)); T8f = VSUB(T8d, T8e); T9s = VADD(T8d, T8e); } { V TaI, TaJ, T82, T83; TaI = VSUB(TaG, TaH); TaJ = VSUB(T1e, T1b); TaK = VSUB(TaI, TaJ); Tcp = VADD(TaJ, TaI); T82 = VSUB(T4d, T4a); T83 = VMUL(LDK(KP707106781), VADD(T4U, T4V)); T84 = VSUB(T82, T83); T9v = VADD(T82, T83); } } { V T1j, TaR, T1m, TaS, T4G, T4L, TaT, TaQ, T89, T88, T1q, TaM, T1t, TaN, T4v; V T4A, TaO, TaL, T86, T85; { V T4H, T4F, T4C, T4K; { V T1h, T1i, T4D, T4E; T1h = LD(&(ri[WS(is, 5)]), ivs, &(ri[WS(is, 1)])); T1i = LD(&(ri[WS(is, 37)]), ivs, &(ri[WS(is, 1)])); T1j = VADD(T1h, T1i); T4H = VSUB(T1h, T1i); T4D = LD(&(ii[WS(is, 5)]), ivs, &(ii[WS(is, 1)])); T4E = LD(&(ii[WS(is, 37)]), ivs, &(ii[WS(is, 1)])); T4F = VSUB(T4D, T4E); TaR = VADD(T4D, T4E); } { V T1k, T1l, T4I, T4J; T1k = LD(&(ri[WS(is, 21)]), ivs, &(ri[WS(is, 1)])); T1l = LD(&(ri[WS(is, 53)]), ivs, &(ri[WS(is, 1)])); T1m = VADD(T1k, T1l); T4C = VSUB(T1k, T1l); T4I = LD(&(ii[WS(is, 21)]), ivs, &(ii[WS(is, 1)])); T4J = LD(&(ii[WS(is, 53)]), ivs, &(ii[WS(is, 1)])); T4K = VSUB(T4I, T4J); TaS = VADD(T4I, T4J); } T4G = VADD(T4C, T4F); T4L = VSUB(T4H, T4K); TaT = VSUB(TaR, TaS); TaQ = VSUB(T1j, T1m); T89 = VADD(T4H, T4K); T88 = VSUB(T4F, T4C); } { V T4r, T4z, T4w, T4u; { V T1o, T1p, T4x, T4y; T1o = LD(&(ri[WS(is, 61)]), ivs, &(ri[WS(is, 1)])); T1p = LD(&(ri[WS(is, 29)]), ivs, &(ri[WS(is, 1)])); T1q = VADD(T1o, T1p); T4r = VSUB(T1o, T1p); T4x = LD(&(ii[WS(is, 61)]), ivs, &(ii[WS(is, 1)])); T4y = LD(&(ii[WS(is, 29)]), ivs, &(ii[WS(is, 1)])); T4z = VSUB(T4x, T4y); TaM = VADD(T4x, T4y); } { V T1r, T1s, T4s, T4t; T1r = LD(&(ri[WS(is, 13)]), ivs, &(ri[WS(is, 1)])); T1s = LD(&(ri[WS(is, 45)]), ivs, &(ri[WS(is, 1)])); T1t = VADD(T1r, T1s); T4w = VSUB(T1r, T1s); T4s = LD(&(ii[WS(is, 13)]), ivs, &(ii[WS(is, 1)])); T4t = LD(&(ii[WS(is, 45)]), ivs, &(ii[WS(is, 1)])); T4u = VSUB(T4s, T4t); TaN = VADD(T4s, T4t); } T4v = VSUB(T4r, T4u); T4A = VADD(T4w, T4z); TaO = VSUB(TaM, TaN); TaL = VSUB(T1q, T1t); T86 = VSUB(T4z, T4w); T85 = VADD(T4r, T4u); } { V T1n, T1u, Tb2, Tb3; T1n = VADD(T1j, T1m); T1u = VADD(T1q, T1t); T1v = VADD(T1n, T1u); Tdn = VSUB(T1u, T1n); Tb2 = VSUB(TaT, TaQ); Tb3 = VADD(TaL, TaO); Tb4 = VMUL(LDK(KP707106781), VSUB(Tb2, Tb3)); Tcq = VMUL(LDK(KP707106781), VADD(Tb2, Tb3)); } { V Tdq, Tdr, T4B, T4M; Tdq = VADD(TaR, TaS); Tdr = VADD(TaM, TaN); Tds = VSUB(Tdq, Tdr); Tek = VADD(Tdq, Tdr); T4B = VFNMS(LDK(KP923879532), T4A, VMUL(LDK(KP382683432), T4v)); T4M = VFMA(LDK(KP923879532), T4G, VMUL(LDK(KP382683432), T4L)); T4N = VSUB(T4B, T4M); T6P = VADD(T4M, T4B); } { V T4Y, T4Z, T8g, T8h; T4Y = VFNMS(LDK(KP923879532), T4L, VMUL(LDK(KP382683432), T4G)); T4Z = VFMA(LDK(KP382683432), T4A, VMUL(LDK(KP923879532), T4v)); T50 = VSUB(T4Y, T4Z); T6S = VADD(T4Y, T4Z); T8g = VFNMS(LDK(KP382683432), T89, VMUL(LDK(KP923879532), T88)); T8h = VFMA(LDK(KP923879532), T86, VMUL(LDK(KP382683432), T85)); T8i = VSUB(T8g, T8h); T9w = VADD(T8g, T8h); } { V TaP, TaU, T87, T8a; TaP = VSUB(TaL, TaO); TaU = VADD(TaQ, TaT); TaV = VMUL(LDK(KP707106781), VSUB(TaP, TaU)); Tcn = VMUL(LDK(KP707106781), VADD(TaU, TaP)); T87 = VFNMS(LDK(KP382683432), T86, VMUL(LDK(KP923879532), T85)); T8a = VFMA(LDK(KP382683432), T88, VMUL(LDK(KP923879532), T89)); T8b = VSUB(T87, T8a); T9t = VADD(T8a, T87); } } { V T1O, Tbc, T1R, Tbd, T5o, T5t, Tbf, Tbe, T8p, T8o, T1V, Tbi, T1Y, Tbj, T5z; V T5E, Tbk, Tbh, T8s, T8r; { V T5p, T5n, T5k, T5s; { V T1M, T1N, T5l, T5m; T1M = LD(&(ri[WS(is, 3)]), ivs, &(ri[WS(is, 1)])); T1N = LD(&(ri[WS(is, 35)]), ivs, &(ri[WS(is, 1)])); T1O = VADD(T1M, T1N); T5p = VSUB(T1M, T1N); T5l = LD(&(ii[WS(is, 3)]), ivs, &(ii[WS(is, 1)])); T5m = LD(&(ii[WS(is, 35)]), ivs, &(ii[WS(is, 1)])); T5n = VSUB(T5l, T5m); Tbc = VADD(T5l, T5m); } { V T1P, T1Q, T5q, T5r; T1P = LD(&(ri[WS(is, 19)]), ivs, &(ri[WS(is, 1)])); T1Q = LD(&(ri[WS(is, 51)]), ivs, &(ri[WS(is, 1)])); T1R = VADD(T1P, T1Q); T5k = VSUB(T1P, T1Q); T5q = LD(&(ii[WS(is, 19)]), ivs, &(ii[WS(is, 1)])); T5r = LD(&(ii[WS(is, 51)]), ivs, &(ii[WS(is, 1)])); T5s = VSUB(T5q, T5r); Tbd = VADD(T5q, T5r); } T5o = VADD(T5k, T5n); T5t = VSUB(T5p, T5s); Tbf = VSUB(T1O, T1R); Tbe = VSUB(Tbc, Tbd); T8p = VADD(T5p, T5s); T8o = VSUB(T5n, T5k); } { V T5A, T5y, T5v, T5D; { V T1T, T1U, T5w, T5x; T1T = LD(&(ri[WS(is, 59)]), ivs, &(ri[WS(is, 1)])); T1U = LD(&(ri[WS(is, 27)]), ivs, &(ri[WS(is, 1)])); T1V = VADD(T1T, T1U); T5A = VSUB(T1T, T1U); T5w = LD(&(ii[WS(is, 59)]), ivs, &(ii[WS(is, 1)])); T5x = LD(&(ii[WS(is, 27)]), ivs, &(ii[WS(is, 1)])); T5y = VSUB(T5w, T5x); Tbi = VADD(T5w, T5x); } { V T1W, T1X, T5B, T5C; T1W = LD(&(ri[WS(is, 11)]), ivs, &(ri[WS(is, 1)])); T1X = LD(&(ri[WS(is, 43)]), ivs, &(ri[WS(is, 1)])); T1Y = VADD(T1W, T1X); T5v = VSUB(T1W, T1X); T5B = LD(&(ii[WS(is, 11)]), ivs, &(ii[WS(is, 1)])); T5C = LD(&(ii[WS(is, 43)]), ivs, &(ii[WS(is, 1)])); T5D = VSUB(T5B, T5C); Tbj = VADD(T5B, T5C); } T5z = VADD(T5v, T5y); T5E = VSUB(T5A, T5D); Tbk = VSUB(Tbi, Tbj); Tbh = VSUB(T1V, T1Y); T8s = VADD(T5A, T5D); T8r = VSUB(T5y, T5v); } { V T1S, T1Z, Tbt, Tbu; T1S = VADD(T1O, T1R); T1Z = VADD(T1V, T1Y); T20 = VADD(T1S, T1Z); TdD = VSUB(T1Z, T1S); Tbt = VSUB(Tbh, Tbk); Tbu = VADD(Tbf, Tbe); Tbv = VMUL(LDK(KP707106781), VSUB(Tbt, Tbu)); Tcu = VMUL(LDK(KP707106781), VADD(Tbu, Tbt)); } { V Tdw, Tdx, T5u, T5F; Tdw = VADD(Tbc, Tbd); Tdx = VADD(Tbi, Tbj); Tdy = VSUB(Tdw, Tdx); Tep = VADD(Tdw, Tdx); T5u = VFNMS(LDK(KP923879532), T5t, VMUL(LDK(KP382683432), T5o)); T5F = VFMA(LDK(KP382683432), T5z, VMUL(LDK(KP923879532), T5E)); T5G = VSUB(T5u, T5F); T6Z = VADD(T5u, T5F); } { V T5R, T5S, T8z, T8A; T5R = VFNMS(LDK(KP923879532), T5z, VMUL(LDK(KP382683432), T5E)); T5S = VFMA(LDK(KP923879532), T5o, VMUL(LDK(KP382683432), T5t)); T5T = VSUB(T5R, T5S); T6W = VADD(T5S, T5R); T8z = VFNMS(LDK(KP382683432), T8r, VMUL(LDK(KP923879532), T8s)); T8A = VFMA(LDK(KP382683432), T8o, VMUL(LDK(KP923879532), T8p)); T8B = VSUB(T8z, T8A); T9A = VADD(T8A, T8z); } { V Tbg, Tbl, T8q, T8t; Tbg = VSUB(Tbe, Tbf); Tbl = VADD(Tbh, Tbk); Tbm = VMUL(LDK(KP707106781), VSUB(Tbg, Tbl)); Tcx = VMUL(LDK(KP707106781), VADD(Tbg, Tbl)); T8q = VFNMS(LDK(KP382683432), T8p, VMUL(LDK(KP923879532), T8o)); T8t = VFMA(LDK(KP923879532), T8r, VMUL(LDK(KP382683432), T8s)); T8u = VSUB(T8q, T8t); T9D = VADD(T8q, T8t); } } { V TeJ, TeK, TeL, TeM, TeN, TeO, TeP, TeQ, TeR, TeS, TeT, TeU, TeV, TeW, TeX; V TeY, TeZ, Tf0, Tf1, Tf2, Tf3, Tf4, Tf5, Tf6, Tf7, Tf8, Tf9, Tfa, Tfb, Tfc; V Tfd, Tfe, Tff, Tfg, Tfh, Tfi, Tfj, Tfk, Tfl, Tfm, Tfn, Tfo, Tfp, Tfq, Tfr; V Tfs, Tft, Tfu; { V T11, TeD, TeG, TeI, T22, T23, T34, TeH; { V Tv, T10, TeE, TeF; Tv = VADD(Tf, Tu); T10 = VADD(TK, TZ); T11 = VADD(Tv, T10); TeD = VSUB(Tv, T10); TeE = VADD(Tej, Tek); TeF = VADD(Teo, Tep); TeG = VSUB(TeE, TeF); TeI = VADD(TeE, TeF); } { V T1w, T21, T2y, T33; T1w = VADD(T1g, T1v); T21 = VADD(T1L, T20); T22 = VADD(T1w, T21); T23 = VSUB(T21, T1w); T2y = VADD(T2i, T2x); T33 = VADD(T2N, T32); T34 = VSUB(T2y, T33); TeH = VADD(T2y, T33); } TeJ = VSUB(T11, T22); STM4(&(ro[32]), TeJ, ovs, &(ro[0])); TeK = VSUB(TeH, TeI); STM4(&(io[32]), TeK, ovs, &(io[0])); TeL = VADD(T11, T22); STM4(&(ro[0]), TeL, ovs, &(ro[0])); TeM = VADD(TeH, TeI); STM4(&(io[0]), TeM, ovs, &(io[0])); TeN = VADD(T23, T34); STM4(&(io[16]), TeN, ovs, &(io[0])); TeO = VADD(TeD, TeG); STM4(&(ro[16]), TeO, ovs, &(ro[0])); TeP = VSUB(T34, T23); STM4(&(io[48]), TeP, ovs, &(io[0])); TeQ = VSUB(TeD, TeG); STM4(&(ro[48]), TeQ, ovs, &(ro[0])); } { V Teh, Tex, Tev, TeB, Tem, Tey, Ter, Tez; { V Tef, Teg, Tet, Teu; Tef = VSUB(Tf, Tu); Teg = VSUB(T2N, T32); Teh = VADD(Tef, Teg); Tex = VSUB(Tef, Teg); Tet = VSUB(T2i, T2x); Teu = VSUB(TZ, TK); Tev = VSUB(Tet, Teu); TeB = VADD(Teu, Tet); } { V Tei, Tel, Ten, Teq; Tei = VSUB(T1g, T1v); Tel = VSUB(Tej, Tek); Tem = VADD(Tei, Tel); Tey = VSUB(Tel, Tei); Ten = VSUB(T1L, T20); Teq = VSUB(Teo, Tep); Ter = VSUB(Ten, Teq); Tez = VADD(Ten, Teq); } { V Tes, TeC, Tew, TeA; Tes = VMUL(LDK(KP707106781), VADD(Tem, Ter)); TeR = VSUB(Teh, Tes); STM4(&(ro[40]), TeR, ovs, &(ro[0])); TeS = VADD(Teh, Tes); STM4(&(ro[8]), TeS, ovs, &(ro[0])); TeC = VMUL(LDK(KP707106781), VADD(Tey, Tez)); TeT = VSUB(TeB, TeC); STM4(&(io[40]), TeT, ovs, &(io[0])); TeU = VADD(TeB, TeC); STM4(&(io[8]), TeU, ovs, &(io[0])); Tew = VMUL(LDK(KP707106781), VSUB(Ter, Tem)); TeV = VSUB(Tev, Tew); STM4(&(io[56]), TeV, ovs, &(io[0])); TeW = VADD(Tev, Tew); STM4(&(io[24]), TeW, ovs, &(io[0])); TeA = VMUL(LDK(KP707106781), VSUB(Tey, Tez)); TeX = VSUB(Tex, TeA); STM4(&(ro[56]), TeX, ovs, &(ro[0])); TeY = VADD(Tex, TeA); STM4(&(ro[24]), TeY, ovs, &(ro[0])); } } { V Tdb, TdV, Te5, TdJ, Tdi, Te6, Te3, Teb, TdM, TdW, Tdu, TdQ, Te0, Tea, TdF; V TdR; { V Tde, Tdh, Tdo, Tdt; Tdb = VSUB(Td9, Tda); TdV = VADD(Td9, Tda); Te5 = VADD(TdI, TdH); TdJ = VSUB(TdH, TdI); Tde = VSUB(Tdc, Tdd); Tdh = VADD(Tdf, Tdg); Tdi = VMUL(LDK(KP707106781), VSUB(Tde, Tdh)); Te6 = VMUL(LDK(KP707106781), VADD(Tde, Tdh)); { V Te1, Te2, TdK, TdL; Te1 = VADD(Tdv, Tdy); Te2 = VADD(TdD, TdC); Te3 = VFNMS(LDK(KP382683432), Te2, VMUL(LDK(KP923879532), Te1)); Teb = VFMA(LDK(KP923879532), Te2, VMUL(LDK(KP382683432), Te1)); TdK = VSUB(Tdf, Tdg); TdL = VADD(Tdd, Tdc); TdM = VMUL(LDK(KP707106781), VSUB(TdK, TdL)); TdW = VMUL(LDK(KP707106781), VADD(TdL, TdK)); } Tdo = VSUB(Tdm, Tdn); Tdt = VSUB(Tdp, Tds); Tdu = VFMA(LDK(KP923879532), Tdo, VMUL(LDK(KP382683432), Tdt)); TdQ = VFNMS(LDK(KP923879532), Tdt, VMUL(LDK(KP382683432), Tdo)); { V TdY, TdZ, Tdz, TdE; TdY = VADD(Tdn, Tdm); TdZ = VADD(Tdp, Tds); Te0 = VFMA(LDK(KP382683432), TdY, VMUL(LDK(KP923879532), TdZ)); Tea = VFNMS(LDK(KP382683432), TdZ, VMUL(LDK(KP923879532), TdY)); Tdz = VSUB(Tdv, Tdy); TdE = VSUB(TdC, TdD); TdF = VFNMS(LDK(KP923879532), TdE, VMUL(LDK(KP382683432), Tdz)); TdR = VFMA(LDK(KP382683432), TdE, VMUL(LDK(KP923879532), Tdz)); } } { V Tdj, TdG, TdT, TdU; Tdj = VADD(Tdb, Tdi); TdG = VADD(Tdu, TdF); TeZ = VSUB(Tdj, TdG); STM4(&(ro[44]), TeZ, ovs, &(ro[0])); Tf0 = VADD(Tdj, TdG); STM4(&(ro[12]), Tf0, ovs, &(ro[0])); TdT = VADD(TdJ, TdM); TdU = VADD(TdQ, TdR); Tf1 = VSUB(TdT, TdU); STM4(&(io[44]), Tf1, ovs, &(io[0])); Tf2 = VADD(TdT, TdU); STM4(&(io[12]), Tf2, ovs, &(io[0])); } { V TdN, TdO, TdP, TdS; TdN = VSUB(TdJ, TdM); TdO = VSUB(TdF, Tdu); Tf3 = VSUB(TdN, TdO); STM4(&(io[60]), Tf3, ovs, &(io[0])); Tf4 = VADD(TdN, TdO); STM4(&(io[28]), Tf4, ovs, &(io[0])); TdP = VSUB(Tdb, Tdi); TdS = VSUB(TdQ, TdR); Tf5 = VSUB(TdP, TdS); STM4(&(ro[60]), Tf5, ovs, &(ro[0])); Tf6 = VADD(TdP, TdS); STM4(&(ro[28]), Tf6, ovs, &(ro[0])); } { V TdX, Te4, Ted, Tee; TdX = VADD(TdV, TdW); Te4 = VADD(Te0, Te3); Tf7 = VSUB(TdX, Te4); STM4(&(ro[36]), Tf7, ovs, &(ro[0])); Tf8 = VADD(TdX, Te4); STM4(&(ro[4]), Tf8, ovs, &(ro[0])); Ted = VADD(Te5, Te6); Tee = VADD(Tea, Teb); Tf9 = VSUB(Ted, Tee); STM4(&(io[36]), Tf9, ovs, &(io[0])); Tfa = VADD(Ted, Tee); STM4(&(io[4]), Tfa, ovs, &(io[0])); } { V Te7, Te8, Te9, Tec; Te7 = VSUB(Te5, Te6); Te8 = VSUB(Te3, Te0); Tfb = VSUB(Te7, Te8); STM4(&(io[52]), Tfb, ovs, &(io[0])); Tfc = VADD(Te7, Te8); STM4(&(io[20]), Tfc, ovs, &(io[0])); Te9 = VSUB(TdV, TdW); Tec = VSUB(Tea, Teb); Tfd = VSUB(Te9, Tec); STM4(&(ro[52]), Tfd, ovs, &(ro[0])); Tfe = VADD(Te9, Tec); STM4(&(ro[20]), Tfe, ovs, &(ro[0])); } } { V Tcd, TcP, TcD, TcZ, Tck, Td0, TcX, Td5, Tcs, TcK, TcG, TcQ, TcU, Td4, Tcz; V TcL, Tcc, TcC; Tcc = VMUL(LDK(KP707106781), VADD(TbD, TbC)); Tcd = VSUB(Tcb, Tcc); TcP = VADD(Tcb, Tcc); TcC = VMUL(LDK(KP707106781), VADD(Tak, Tan)); TcD = VSUB(TcB, TcC); TcZ = VADD(TcB, TcC); { V Tcg, Tcj, TcV, TcW; Tcg = VFNMS(LDK(KP382683432), Tcf, VMUL(LDK(KP923879532), Tce)); Tcj = VFMA(LDK(KP923879532), Tch, VMUL(LDK(KP382683432), Tci)); Tck = VSUB(Tcg, Tcj); Td0 = VADD(Tcg, Tcj); TcV = VADD(Tct, Tcu); TcW = VADD(Tcw, Tcx); TcX = VFNMS(LDK(KP195090322), TcW, VMUL(LDK(KP980785280), TcV)); Td5 = VFMA(LDK(KP195090322), TcV, VMUL(LDK(KP980785280), TcW)); } { V Tco, Tcr, TcE, TcF; Tco = VSUB(Tcm, Tcn); Tcr = VSUB(Tcp, Tcq); Tcs = VFMA(LDK(KP555570233), Tco, VMUL(LDK(KP831469612), Tcr)); TcK = VFNMS(LDK(KP831469612), Tco, VMUL(LDK(KP555570233), Tcr)); TcE = VFNMS(LDK(KP382683432), Tch, VMUL(LDK(KP923879532), Tci)); TcF = VFMA(LDK(KP382683432), Tce, VMUL(LDK(KP923879532), Tcf)); TcG = VSUB(TcE, TcF); TcQ = VADD(TcF, TcE); } { V TcS, TcT, Tcv, Tcy; TcS = VADD(Tcm, Tcn); TcT = VADD(Tcp, Tcq); TcU = VFMA(LDK(KP980785280), TcS, VMUL(LDK(KP195090322), TcT)); Td4 = VFNMS(LDK(KP195090322), TcS, VMUL(LDK(KP980785280), TcT)); Tcv = VSUB(Tct, Tcu); Tcy = VSUB(Tcw, Tcx); Tcz = VFNMS(LDK(KP831469612), Tcy, VMUL(LDK(KP555570233), Tcv)); TcL = VFMA(LDK(KP831469612), Tcv, VMUL(LDK(KP555570233), Tcy)); } { V Tcl, TcA, TcN, TcO; Tcl = VADD(Tcd, Tck); TcA = VADD(Tcs, Tcz); Tff = VSUB(Tcl, TcA); STM4(&(ro[42]), Tff, ovs, &(ro[0])); Tfg = VADD(Tcl, TcA); STM4(&(ro[10]), Tfg, ovs, &(ro[0])); TcN = VADD(TcD, TcG); TcO = VADD(TcK, TcL); Tfh = VSUB(TcN, TcO); STM4(&(io[42]), Tfh, ovs, &(io[0])); Tfi = VADD(TcN, TcO); STM4(&(io[10]), Tfi, ovs, &(io[0])); } { V TcH, TcI, TcJ, TcM; TcH = VSUB(TcD, TcG); TcI = VSUB(Tcz, Tcs); Tfj = VSUB(TcH, TcI); STM4(&(io[58]), Tfj, ovs, &(io[0])); Tfk = VADD(TcH, TcI); STM4(&(io[26]), Tfk, ovs, &(io[0])); TcJ = VSUB(Tcd, Tck); TcM = VSUB(TcK, TcL); Tfl = VSUB(TcJ, TcM); STM4(&(ro[58]), Tfl, ovs, &(ro[0])); Tfm = VADD(TcJ, TcM); STM4(&(ro[26]), Tfm, ovs, &(ro[0])); } { V TcR, TcY, Td7, Td8; TcR = VADD(TcP, TcQ); TcY = VADD(TcU, TcX); Tfn = VSUB(TcR, TcY); STM4(&(ro[34]), Tfn, ovs, &(ro[0])); Tfo = VADD(TcR, TcY); STM4(&(ro[2]), Tfo, ovs, &(ro[0])); Td7 = VADD(TcZ, Td0); Td8 = VADD(Td4, Td5); Tfp = VSUB(Td7, Td8); STM4(&(io[34]), Tfp, ovs, &(io[0])); Tfq = VADD(Td7, Td8); STM4(&(io[2]), Tfq, ovs, &(io[0])); } { V Td1, Td2, Td3, Td6; Td1 = VSUB(TcZ, Td0); Td2 = VSUB(TcX, TcU); Tfr = VSUB(Td1, Td2); STM4(&(io[50]), Tfr, ovs, &(io[0])); Tfs = VADD(Td1, Td2); STM4(&(io[18]), Tfs, ovs, &(io[0])); Td3 = VSUB(TcP, TcQ); Td6 = VSUB(Td4, Td5); Tft = VSUB(Td3, Td6); STM4(&(ro[50]), Tft, ovs, &(ro[0])); Tfu = VADD(Td3, Td6); STM4(&(ro[18]), Tfu, ovs, &(ro[0])); } } { V Tfv, Tfw, Tfx, Tfy, Tfz, TfA, TfB, TfC, TfD, TfE, TfF, TfG, TfH, TfI, TfJ; V TfK, TfL, TfM, TfN, TfO, TfP, TfQ, TfR, TfS, TfT, TfU, TfV, TfW, TfX, TfY; V TfZ, Tg0; { V Tap, TbR, TbF, Tc1, TaE, Tc2, TbZ, Tc7, Tb6, TbM, TbI, TbS, TbW, Tc6, Tbx; V TbN, Tao, TbE; Tao = VMUL(LDK(KP707106781), VSUB(Tak, Tan)); Tap = VSUB(Tah, Tao); TbR = VADD(Tah, Tao); TbE = VMUL(LDK(KP707106781), VSUB(TbC, TbD)); TbF = VSUB(TbB, TbE); Tc1 = VADD(TbB, TbE); { V Taw, TaD, TbX, TbY; Taw = VFNMS(LDK(KP923879532), Tav, VMUL(LDK(KP382683432), Tas)); TaD = VFMA(LDK(KP382683432), Taz, VMUL(LDK(KP923879532), TaC)); TaE = VSUB(Taw, TaD); Tc2 = VADD(Taw, TaD); TbX = VADD(Tbb, Tbm); TbY = VADD(Tbs, Tbv); TbZ = VFNMS(LDK(KP555570233), TbY, VMUL(LDK(KP831469612), TbX)); Tc7 = VFMA(LDK(KP831469612), TbY, VMUL(LDK(KP555570233), TbX)); } { V TaW, Tb5, TbG, TbH; TaW = VSUB(TaK, TaV); Tb5 = VSUB(Tb1, Tb4); Tb6 = VFMA(LDK(KP980785280), TaW, VMUL(LDK(KP195090322), Tb5)); TbM = VFNMS(LDK(KP980785280), Tb5, VMUL(LDK(KP195090322), TaW)); TbG = VFNMS(LDK(KP923879532), Taz, VMUL(LDK(KP382683432), TaC)); TbH = VFMA(LDK(KP923879532), Tas, VMUL(LDK(KP382683432), Tav)); TbI = VSUB(TbG, TbH); TbS = VADD(TbH, TbG); } { V TbU, TbV, Tbn, Tbw; TbU = VADD(TaK, TaV); TbV = VADD(Tb1, Tb4); TbW = VFMA(LDK(KP555570233), TbU, VMUL(LDK(KP831469612), TbV)); Tc6 = VFNMS(LDK(KP555570233), TbV, VMUL(LDK(KP831469612), TbU)); Tbn = VSUB(Tbb, Tbm); Tbw = VSUB(Tbs, Tbv); Tbx = VFNMS(LDK(KP980785280), Tbw, VMUL(LDK(KP195090322), Tbn)); TbN = VFMA(LDK(KP195090322), Tbw, VMUL(LDK(KP980785280), Tbn)); } { V TaF, Tby, TbP, TbQ; TaF = VADD(Tap, TaE); Tby = VADD(Tb6, Tbx); Tfv = VSUB(TaF, Tby); STM4(&(ro[46]), Tfv, ovs, &(ro[0])); Tfw = VADD(TaF, Tby); STM4(&(ro[14]), Tfw, ovs, &(ro[0])); TbP = VADD(TbF, TbI); TbQ = VADD(TbM, TbN); Tfx = VSUB(TbP, TbQ); STM4(&(io[46]), Tfx, ovs, &(io[0])); Tfy = VADD(TbP, TbQ); STM4(&(io[14]), Tfy, ovs, &(io[0])); } { V TbJ, TbK, TbL, TbO; TbJ = VSUB(TbF, TbI); TbK = VSUB(Tbx, Tb6); Tfz = VSUB(TbJ, TbK); STM4(&(io[62]), Tfz, ovs, &(io[0])); TfA = VADD(TbJ, TbK); STM4(&(io[30]), TfA, ovs, &(io[0])); TbL = VSUB(Tap, TaE); TbO = VSUB(TbM, TbN); TfB = VSUB(TbL, TbO); STM4(&(ro[62]), TfB, ovs, &(ro[0])); TfC = VADD(TbL, TbO); STM4(&(ro[30]), TfC, ovs, &(ro[0])); } { V TbT, Tc0, Tc9, Tca; TbT = VADD(TbR, TbS); Tc0 = VADD(TbW, TbZ); TfD = VSUB(TbT, Tc0); STM4(&(ro[38]), TfD, ovs, &(ro[0])); TfE = VADD(TbT, Tc0); STM4(&(ro[6]), TfE, ovs, &(ro[0])); Tc9 = VADD(Tc1, Tc2); Tca = VADD(Tc6, Tc7); TfF = VSUB(Tc9, Tca); STM4(&(io[38]), TfF, ovs, &(io[0])); TfG = VADD(Tc9, Tca); STM4(&(io[6]), TfG, ovs, &(io[0])); } { V Tc3, Tc4, Tc5, Tc8; Tc3 = VSUB(Tc1, Tc2); Tc4 = VSUB(TbZ, TbW); TfH = VSUB(Tc3, Tc4); STM4(&(io[54]), TfH, ovs, &(io[0])); TfI = VADD(Tc3, Tc4); STM4(&(io[22]), TfI, ovs, &(io[0])); Tc5 = VSUB(TbR, TbS); Tc8 = VSUB(Tc6, Tc7); TfJ = VSUB(Tc5, Tc8); STM4(&(ro[54]), TfJ, ovs, &(ro[0])); TfK = VADD(Tc5, Tc8); STM4(&(ro[22]), TfK, ovs, &(ro[0])); } } { V T6F, T7h, T7m, T7w, T7p, T7x, T6M, T7s, T6U, T7c, T75, T7r, T78, T7i, T71; V T7d; { V T6D, T6E, T7k, T7l; T6D = VADD(T37, T3e); T6E = VADD(T65, T64); T6F = VSUB(T6D, T6E); T7h = VADD(T6D, T6E); T7k = VADD(T6O, T6P); T7l = VADD(T6R, T6S); T7m = VFMA(LDK(KP956940335), T7k, VMUL(LDK(KP290284677), T7l)); T7w = VFNMS(LDK(KP290284677), T7k, VMUL(LDK(KP956940335), T7l)); } { V T7n, T7o, T6I, T6L; T7n = VADD(T6V, T6W); T7o = VADD(T6Y, T6Z); T7p = VFNMS(LDK(KP290284677), T7o, VMUL(LDK(KP956940335), T7n)); T7x = VFMA(LDK(KP290284677), T7n, VMUL(LDK(KP956940335), T7o)); T6I = VFNMS(LDK(KP555570233), T6H, VMUL(LDK(KP831469612), T6G)); T6L = VFMA(LDK(KP831469612), T6J, VMUL(LDK(KP555570233), T6K)); T6M = VSUB(T6I, T6L); T7s = VADD(T6I, T6L); } { V T6Q, T6T, T73, T74; T6Q = VSUB(T6O, T6P); T6T = VSUB(T6R, T6S); T6U = VFMA(LDK(KP471396736), T6Q, VMUL(LDK(KP881921264), T6T)); T7c = VFNMS(LDK(KP881921264), T6Q, VMUL(LDK(KP471396736), T6T)); T73 = VADD(T5Z, T62); T74 = VADD(T3m, T3t); T75 = VSUB(T73, T74); T7r = VADD(T73, T74); } { V T76, T77, T6X, T70; T76 = VFNMS(LDK(KP555570233), T6J, VMUL(LDK(KP831469612), T6K)); T77 = VFMA(LDK(KP555570233), T6G, VMUL(LDK(KP831469612), T6H)); T78 = VSUB(T76, T77); T7i = VADD(T77, T76); T6X = VSUB(T6V, T6W); T70 = VSUB(T6Y, T6Z); T71 = VFNMS(LDK(KP881921264), T70, VMUL(LDK(KP471396736), T6X)); T7d = VFMA(LDK(KP881921264), T6X, VMUL(LDK(KP471396736), T70)); } { V T6N, T72, T7f, T7g; T6N = VADD(T6F, T6M); T72 = VADD(T6U, T71); TfL = VSUB(T6N, T72); STM4(&(ro[43]), TfL, ovs, &(ro[1])); TfM = VADD(T6N, T72); STM4(&(ro[11]), TfM, ovs, &(ro[1])); T7f = VADD(T75, T78); T7g = VADD(T7c, T7d); TfN = VSUB(T7f, T7g); STM4(&(io[43]), TfN, ovs, &(io[1])); TfO = VADD(T7f, T7g); STM4(&(io[11]), TfO, ovs, &(io[1])); } { V T79, T7a, T7b, T7e; T79 = VSUB(T75, T78); T7a = VSUB(T71, T6U); TfP = VSUB(T79, T7a); STM4(&(io[59]), TfP, ovs, &(io[1])); TfQ = VADD(T79, T7a); STM4(&(io[27]), TfQ, ovs, &(io[1])); T7b = VSUB(T6F, T6M); T7e = VSUB(T7c, T7d); TfR = VSUB(T7b, T7e); STM4(&(ro[59]), TfR, ovs, &(ro[1])); TfS = VADD(T7b, T7e); STM4(&(ro[27]), TfS, ovs, &(ro[1])); } { V T7j, T7q, T7z, T7A; T7j = VADD(T7h, T7i); T7q = VADD(T7m, T7p); TfT = VSUB(T7j, T7q); STM4(&(ro[35]), TfT, ovs, &(ro[1])); TfU = VADD(T7j, T7q); STM4(&(ro[3]), TfU, ovs, &(ro[1])); T7z = VADD(T7r, T7s); T7A = VADD(T7w, T7x); TfV = VSUB(T7z, T7A); STM4(&(io[35]), TfV, ovs, &(io[1])); TfW = VADD(T7z, T7A); STM4(&(io[3]), TfW, ovs, &(io[1])); } { V T7t, T7u, T7v, T7y; T7t = VSUB(T7r, T7s); T7u = VSUB(T7p, T7m); TfX = VSUB(T7t, T7u); STM4(&(io[51]), TfX, ovs, &(io[1])); TfY = VADD(T7t, T7u); STM4(&(io[19]), TfY, ovs, &(io[1])); T7v = VSUB(T7h, T7i); T7y = VSUB(T7w, T7x); TfZ = VSUB(T7v, T7y); STM4(&(ro[51]), TfZ, ovs, &(ro[1])); Tg0 = VADD(T7v, T7y); STM4(&(ro[19]), Tg0, ovs, &(ro[1])); } } { V T9j, T9V, Ta0, Taa, Ta3, Tab, T9q, Ta6, T9y, T9Q, T9J, Ta5, T9M, T9W, T9F; V T9R; { V T9h, T9i, T9Y, T9Z; T9h = VADD(T7B, T7C); T9i = VADD(T8J, T8I); T9j = VSUB(T9h, T9i); T9V = VADD(T9h, T9i); T9Y = VADD(T9s, T9t); T9Z = VADD(T9v, T9w); Ta0 = VFMA(LDK(KP995184726), T9Y, VMUL(LDK(KP098017140), T9Z)); Taa = VFNMS(LDK(KP098017140), T9Y, VMUL(LDK(KP995184726), T9Z)); } { V Ta1, Ta2, T9m, T9p; Ta1 = VADD(T9z, T9A); Ta2 = VADD(T9C, T9D); Ta3 = VFNMS(LDK(KP098017140), Ta2, VMUL(LDK(KP995184726), Ta1)); Tab = VFMA(LDK(KP098017140), Ta1, VMUL(LDK(KP995184726), Ta2)); T9m = VFNMS(LDK(KP195090322), T9l, VMUL(LDK(KP980785280), T9k)); T9p = VFMA(LDK(KP195090322), T9n, VMUL(LDK(KP980785280), T9o)); T9q = VSUB(T9m, T9p); Ta6 = VADD(T9m, T9p); } { V T9u, T9x, T9H, T9I; T9u = VSUB(T9s, T9t); T9x = VSUB(T9v, T9w); T9y = VFMA(LDK(KP634393284), T9u, VMUL(LDK(KP773010453), T9x)); T9Q = VFNMS(LDK(KP773010453), T9u, VMUL(LDK(KP634393284), T9x)); T9H = VADD(T8F, T8G); T9I = VADD(T7G, T7J); T9J = VSUB(T9H, T9I); Ta5 = VADD(T9H, T9I); } { V T9K, T9L, T9B, T9E; T9K = VFNMS(LDK(KP195090322), T9o, VMUL(LDK(KP980785280), T9n)); T9L = VFMA(LDK(KP980785280), T9l, VMUL(LDK(KP195090322), T9k)); T9M = VSUB(T9K, T9L); T9W = VADD(T9L, T9K); T9B = VSUB(T9z, T9A); T9E = VSUB(T9C, T9D); T9F = VFNMS(LDK(KP773010453), T9E, VMUL(LDK(KP634393284), T9B)); T9R = VFMA(LDK(KP773010453), T9B, VMUL(LDK(KP634393284), T9E)); } { V T9r, T9G, Tg1, Tg2; T9r = VADD(T9j, T9q); T9G = VADD(T9y, T9F); Tg1 = VSUB(T9r, T9G); STM4(&(ro[41]), Tg1, ovs, &(ro[1])); STN4(&(ro[40]), TeR, Tg1, Tff, TfL, ovs); Tg2 = VADD(T9r, T9G); STM4(&(ro[9]), Tg2, ovs, &(ro[1])); STN4(&(ro[8]), TeS, Tg2, Tfg, TfM, ovs); } { V T9T, T9U, Tg3, Tg4; T9T = VADD(T9J, T9M); T9U = VADD(T9Q, T9R); Tg3 = VSUB(T9T, T9U); STM4(&(io[41]), Tg3, ovs, &(io[1])); STN4(&(io[40]), TeT, Tg3, Tfh, TfN, ovs); Tg4 = VADD(T9T, T9U); STM4(&(io[9]), Tg4, ovs, &(io[1])); STN4(&(io[8]), TeU, Tg4, Tfi, TfO, ovs); } { V T9N, T9O, Tg5, Tg6; T9N = VSUB(T9J, T9M); T9O = VSUB(T9F, T9y); Tg5 = VSUB(T9N, T9O); STM4(&(io[57]), Tg5, ovs, &(io[1])); STN4(&(io[56]), TeV, Tg5, Tfj, TfP, ovs); Tg6 = VADD(T9N, T9O); STM4(&(io[25]), Tg6, ovs, &(io[1])); STN4(&(io[24]), TeW, Tg6, Tfk, TfQ, ovs); } { V T9P, T9S, Tg7, Tg8; T9P = VSUB(T9j, T9q); T9S = VSUB(T9Q, T9R); Tg7 = VSUB(T9P, T9S); STM4(&(ro[57]), Tg7, ovs, &(ro[1])); STN4(&(ro[56]), TeX, Tg7, Tfl, TfR, ovs); Tg8 = VADD(T9P, T9S); STM4(&(ro[25]), Tg8, ovs, &(ro[1])); STN4(&(ro[24]), TeY, Tg8, Tfm, TfS, ovs); } { V T9X, Ta4, Tg9, Tga; T9X = VADD(T9V, T9W); Ta4 = VADD(Ta0, Ta3); Tg9 = VSUB(T9X, Ta4); STM4(&(ro[33]), Tg9, ovs, &(ro[1])); STN4(&(ro[32]), TeJ, Tg9, Tfn, TfT, ovs); Tga = VADD(T9X, Ta4); STM4(&(ro[1]), Tga, ovs, &(ro[1])); STN4(&(ro[0]), TeL, Tga, Tfo, TfU, ovs); } { V Tad, Tae, Tgb, Tgc; Tad = VADD(Ta5, Ta6); Tae = VADD(Taa, Tab); Tgb = VSUB(Tad, Tae); STM4(&(io[33]), Tgb, ovs, &(io[1])); STN4(&(io[32]), TeK, Tgb, Tfp, TfV, ovs); Tgc = VADD(Tad, Tae); STM4(&(io[1]), Tgc, ovs, &(io[1])); STN4(&(io[0]), TeM, Tgc, Tfq, TfW, ovs); } { V Ta7, Ta8, Tgd, Tge; Ta7 = VSUB(Ta5, Ta6); Ta8 = VSUB(Ta3, Ta0); Tgd = VSUB(Ta7, Ta8); STM4(&(io[49]), Tgd, ovs, &(io[1])); STN4(&(io[48]), TeP, Tgd, Tfr, TfX, ovs); Tge = VADD(Ta7, Ta8); STM4(&(io[17]), Tge, ovs, &(io[1])); STN4(&(io[16]), TeN, Tge, Tfs, TfY, ovs); } { V Ta9, Tac, Tgf, Tgg; Ta9 = VSUB(T9V, T9W); Tac = VSUB(Taa, Tab); Tgf = VSUB(Ta9, Tac); STM4(&(ro[49]), Tgf, ovs, &(ro[1])); STN4(&(ro[48]), TeQ, Tgf, Tft, TfZ, ovs); Tgg = VADD(Ta9, Tac); STM4(&(ro[17]), Tgg, ovs, &(ro[1])); STN4(&(ro[16]), TeO, Tgg, Tfu, Tg0, ovs); } } { V Tgh, Tgi, Tgj, Tgk, Tgl, Tgm, Tgn, Tgo, Tgp, Tgq, Tgr, Tgs, Tgt, Tgu, Tgv; V Tgw; { V T3v, T6j, T6o, T6y, T6r, T6z, T48, T6u, T52, T6e, T67, T6t, T6a, T6k, T5V; V T6f; { V T3f, T3u, T6m, T6n; T3f = VSUB(T37, T3e); T3u = VSUB(T3m, T3t); T3v = VSUB(T3f, T3u); T6j = VADD(T3f, T3u); T6m = VADD(T4q, T4N); T6n = VADD(T4X, T50); T6o = VFMA(LDK(KP634393284), T6m, VMUL(LDK(KP773010453), T6n)); T6y = VFNMS(LDK(KP634393284), T6n, VMUL(LDK(KP773010453), T6m)); } { V T6p, T6q, T3O, T47; T6p = VADD(T5j, T5G); T6q = VADD(T5Q, T5T); T6r = VFNMS(LDK(KP634393284), T6q, VMUL(LDK(KP773010453), T6p)); T6z = VFMA(LDK(KP773010453), T6q, VMUL(LDK(KP634393284), T6p)); T3O = VFNMS(LDK(KP980785280), T3N, VMUL(LDK(KP195090322), T3G)); T47 = VFMA(LDK(KP195090322), T3Z, VMUL(LDK(KP980785280), T46)); T48 = VSUB(T3O, T47); T6u = VADD(T3O, T47); } { V T4O, T51, T63, T66; T4O = VSUB(T4q, T4N); T51 = VSUB(T4X, T50); T52 = VFMA(LDK(KP995184726), T4O, VMUL(LDK(KP098017140), T51)); T6e = VFNMS(LDK(KP995184726), T51, VMUL(LDK(KP098017140), T4O)); T63 = VSUB(T5Z, T62); T66 = VSUB(T64, T65); T67 = VSUB(T63, T66); T6t = VADD(T63, T66); } { V T68, T69, T5H, T5U; T68 = VFNMS(LDK(KP980785280), T3Z, VMUL(LDK(KP195090322), T46)); T69 = VFMA(LDK(KP980785280), T3G, VMUL(LDK(KP195090322), T3N)); T6a = VSUB(T68, T69); T6k = VADD(T69, T68); T5H = VSUB(T5j, T5G); T5U = VSUB(T5Q, T5T); T5V = VFNMS(LDK(KP995184726), T5U, VMUL(LDK(KP098017140), T5H)); T6f = VFMA(LDK(KP098017140), T5U, VMUL(LDK(KP995184726), T5H)); } { V T49, T5W, T6h, T6i; T49 = VADD(T3v, T48); T5W = VADD(T52, T5V); Tgh = VSUB(T49, T5W); STM4(&(ro[47]), Tgh, ovs, &(ro[1])); Tgi = VADD(T49, T5W); STM4(&(ro[15]), Tgi, ovs, &(ro[1])); T6h = VADD(T67, T6a); T6i = VADD(T6e, T6f); Tgj = VSUB(T6h, T6i); STM4(&(io[47]), Tgj, ovs, &(io[1])); Tgk = VADD(T6h, T6i); STM4(&(io[15]), Tgk, ovs, &(io[1])); } { V T6b, T6c, T6d, T6g; T6b = VSUB(T67, T6a); T6c = VSUB(T5V, T52); Tgl = VSUB(T6b, T6c); STM4(&(io[63]), Tgl, ovs, &(io[1])); Tgm = VADD(T6b, T6c); STM4(&(io[31]), Tgm, ovs, &(io[1])); T6d = VSUB(T3v, T48); T6g = VSUB(T6e, T6f); Tgn = VSUB(T6d, T6g); STM4(&(ro[63]), Tgn, ovs, &(ro[1])); Tgo = VADD(T6d, T6g); STM4(&(ro[31]), Tgo, ovs, &(ro[1])); } { V T6l, T6s, T6B, T6C; T6l = VADD(T6j, T6k); T6s = VADD(T6o, T6r); Tgp = VSUB(T6l, T6s); STM4(&(ro[39]), Tgp, ovs, &(ro[1])); Tgq = VADD(T6l, T6s); STM4(&(ro[7]), Tgq, ovs, &(ro[1])); T6B = VADD(T6t, T6u); T6C = VADD(T6y, T6z); Tgr = VSUB(T6B, T6C); STM4(&(io[39]), Tgr, ovs, &(io[1])); Tgs = VADD(T6B, T6C); STM4(&(io[7]), Tgs, ovs, &(io[1])); } { V T6v, T6w, T6x, T6A; T6v = VSUB(T6t, T6u); T6w = VSUB(T6r, T6o); Tgt = VSUB(T6v, T6w); STM4(&(io[55]), Tgt, ovs, &(io[1])); Tgu = VADD(T6v, T6w); STM4(&(io[23]), Tgu, ovs, &(io[1])); T6x = VSUB(T6j, T6k); T6A = VSUB(T6y, T6z); Tgv = VSUB(T6x, T6A); STM4(&(ro[55]), Tgv, ovs, &(ro[1])); Tgw = VADD(T6x, T6A); STM4(&(ro[23]), Tgw, ovs, &(ro[1])); } } { V T7L, T8X, T92, T9c, T95, T9d, T80, T98, T8k, T8S, T8L, T97, T8O, T8Y, T8D; V T8T; { V T7D, T7K, T90, T91; T7D = VSUB(T7B, T7C); T7K = VSUB(T7G, T7J); T7L = VSUB(T7D, T7K); T8X = VADD(T7D, T7K); T90 = VADD(T84, T8b); T91 = VADD(T8f, T8i); T92 = VFMA(LDK(KP471396736), T90, VMUL(LDK(KP881921264), T91)); T9c = VFNMS(LDK(KP471396736), T91, VMUL(LDK(KP881921264), T90)); } { V T93, T94, T7S, T7Z; T93 = VADD(T8n, T8u); T94 = VADD(T8y, T8B); T95 = VFNMS(LDK(KP471396736), T94, VMUL(LDK(KP881921264), T93)); T9d = VFMA(LDK(KP881921264), T94, VMUL(LDK(KP471396736), T93)); T7S = VFNMS(LDK(KP831469612), T7R, VMUL(LDK(KP555570233), T7O)); T7Z = VFMA(LDK(KP831469612), T7V, VMUL(LDK(KP555570233), T7Y)); T80 = VSUB(T7S, T7Z); T98 = VADD(T7S, T7Z); } { V T8c, T8j, T8H, T8K; T8c = VSUB(T84, T8b); T8j = VSUB(T8f, T8i); T8k = VFMA(LDK(KP956940335), T8c, VMUL(LDK(KP290284677), T8j)); T8S = VFNMS(LDK(KP956940335), T8j, VMUL(LDK(KP290284677), T8c)); T8H = VSUB(T8F, T8G); T8K = VSUB(T8I, T8J); T8L = VSUB(T8H, T8K); T97 = VADD(T8H, T8K); } { V T8M, T8N, T8v, T8C; T8M = VFNMS(LDK(KP831469612), T7Y, VMUL(LDK(KP555570233), T7V)); T8N = VFMA(LDK(KP555570233), T7R, VMUL(LDK(KP831469612), T7O)); T8O = VSUB(T8M, T8N); T8Y = VADD(T8N, T8M); T8v = VSUB(T8n, T8u); T8C = VSUB(T8y, T8B); T8D = VFNMS(LDK(KP956940335), T8C, VMUL(LDK(KP290284677), T8v)); T8T = VFMA(LDK(KP290284677), T8C, VMUL(LDK(KP956940335), T8v)); } { V T81, T8E, Tgx, Tgy; T81 = VADD(T7L, T80); T8E = VADD(T8k, T8D); Tgx = VSUB(T81, T8E); STM4(&(ro[45]), Tgx, ovs, &(ro[1])); STN4(&(ro[44]), TeZ, Tgx, Tfv, Tgh, ovs); Tgy = VADD(T81, T8E); STM4(&(ro[13]), Tgy, ovs, &(ro[1])); STN4(&(ro[12]), Tf0, Tgy, Tfw, Tgi, ovs); } { V T8V, T8W, Tgz, TgA; T8V = VADD(T8L, T8O); T8W = VADD(T8S, T8T); Tgz = VSUB(T8V, T8W); STM4(&(io[45]), Tgz, ovs, &(io[1])); STN4(&(io[44]), Tf1, Tgz, Tfx, Tgj, ovs); TgA = VADD(T8V, T8W); STM4(&(io[13]), TgA, ovs, &(io[1])); STN4(&(io[12]), Tf2, TgA, Tfy, Tgk, ovs); } { V T8P, T8Q, TgB, TgC; T8P = VSUB(T8L, T8O); T8Q = VSUB(T8D, T8k); TgB = VSUB(T8P, T8Q); STM4(&(io[61]), TgB, ovs, &(io[1])); STN4(&(io[60]), Tf3, TgB, Tfz, Tgl, ovs); TgC = VADD(T8P, T8Q); STM4(&(io[29]), TgC, ovs, &(io[1])); STN4(&(io[28]), Tf4, TgC, TfA, Tgm, ovs); } { V T8R, T8U, TgD, TgE; T8R = VSUB(T7L, T80); T8U = VSUB(T8S, T8T); TgD = VSUB(T8R, T8U); STM4(&(ro[61]), TgD, ovs, &(ro[1])); STN4(&(ro[60]), Tf5, TgD, TfB, Tgn, ovs); TgE = VADD(T8R, T8U); STM4(&(ro[29]), TgE, ovs, &(ro[1])); STN4(&(ro[28]), Tf6, TgE, TfC, Tgo, ovs); } { V T8Z, T96, TgF, TgG; T8Z = VADD(T8X, T8Y); T96 = VADD(T92, T95); TgF = VSUB(T8Z, T96); STM4(&(ro[37]), TgF, ovs, &(ro[1])); STN4(&(ro[36]), Tf7, TgF, TfD, Tgp, ovs); TgG = VADD(T8Z, T96); STM4(&(ro[5]), TgG, ovs, &(ro[1])); STN4(&(ro[4]), Tf8, TgG, TfE, Tgq, ovs); } { V T9f, T9g, TgH, TgI; T9f = VADD(T97, T98); T9g = VADD(T9c, T9d); TgH = VSUB(T9f, T9g); STM4(&(io[37]), TgH, ovs, &(io[1])); STN4(&(io[36]), Tf9, TgH, TfF, Tgr, ovs); TgI = VADD(T9f, T9g); STM4(&(io[5]), TgI, ovs, &(io[1])); STN4(&(io[4]), Tfa, TgI, TfG, Tgs, ovs); } { V T99, T9a, TgJ, TgK; T99 = VSUB(T97, T98); T9a = VSUB(T95, T92); TgJ = VSUB(T99, T9a); STM4(&(io[53]), TgJ, ovs, &(io[1])); STN4(&(io[52]), Tfb, TgJ, TfH, Tgt, ovs); TgK = VADD(T99, T9a); STM4(&(io[21]), TgK, ovs, &(io[1])); STN4(&(io[20]), Tfc, TgK, TfI, Tgu, ovs); } { V T9b, T9e, TgL, TgM; T9b = VSUB(T8X, T8Y); T9e = VSUB(T9c, T9d); TgL = VSUB(T9b, T9e); STM4(&(ro[53]), TgL, ovs, &(ro[1])); STN4(&(ro[52]), Tfd, TgL, TfJ, Tgv, ovs); TgM = VADD(T9b, T9e); STM4(&(ro[21]), TgM, ovs, &(ro[1])); STN4(&(ro[20]), Tfe, TgM, TfK, Tgw, ovs); } } } } } } } VLEAVE(); } static const kdft_desc desc = { 64, XSIMD_STRING("n2sv_64"), {808, 144, 104, 0}, &GENUS, 0, 1, 0, 0 }; void XSIMD(codelet_n2sv_64) (planner *p) { X(kdft_register) (p, n2sv_64, &desc); } #endif /* HAVE_FMA */ fftw-3.3.3/dft/simd/common/t1fuv_2.c0000644000175000017500000000655412054410447014055 00000000000000/* * Copyright (c) 2003, 2007-11 Matteo Frigo * Copyright (c) 2003, 2007-11 Massachusetts Institute of Technology * * This program is free software; you can redistribute it and/or modify * it under the terms of the GNU General Public License as published by * the Free Software Foundation; either version 2 of the License, or * (at your option) any later version. * * This program is distributed in the hope that it will be useful, * but WITHOUT ANY WARRANTY; without even the implied warranty of * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the * GNU General Public License for more details. * * You should have received a copy of the GNU General Public License * along with this program; if not, write to the Free Software * Foundation, Inc., 51 Franklin Street, Fifth Floor, Boston, MA 02110-1301 USA * */ /* This file was automatically generated --- DO NOT EDIT */ /* Generated on Sun Nov 25 07:37:59 EST 2012 */ #include "codelet-dft.h" #ifdef HAVE_FMA /* Generated by: ../../../genfft/gen_twiddle_c.native -fma -reorder-insns -schedule-for-pipeline -simd -compact -variables 4 -pipeline-latency 8 -n 2 -name t1fuv_2 -include t1fu.h */ /* * This function contains 3 FP additions, 2 FP multiplications, * (or, 3 additions, 2 multiplications, 0 fused multiply/add), * 5 stack variables, 0 constants, and 4 memory accesses */ #include "t1fu.h" static void t1fuv_2(R *ri, R *ii, const R *W, stride rs, INT mb, INT me, INT ms) { { INT m; R *x; x = ri; for (m = mb, W = W + (mb * ((TWVL / VL) * 2)); m < me; m = m + VL, x = x + (VL * ms), W = W + (TWVL * 2), MAKE_VOLATILE_STRIDE(2, rs)) { V T1, T2, T3; T1 = LD(&(x[0]), ms, &(x[0])); T2 = LD(&(x[WS(rs, 1)]), ms, &(x[WS(rs, 1)])); T3 = BYTWJ(&(W[0]), T2); ST(&(x[0]), VADD(T1, T3), ms, &(x[0])); ST(&(x[WS(rs, 1)]), VSUB(T1, T3), ms, &(x[WS(rs, 1)])); } } VLEAVE(); } static const tw_instr twinstr[] = { VTW(0, 1), {TW_NEXT, VL, 0} }; static const ct_desc desc = { 2, XSIMD_STRING("t1fuv_2"), twinstr, &GENUS, {3, 2, 0, 0}, 0, 0, 0 }; void XSIMD(codelet_t1fuv_2) (planner *p) { X(kdft_dit_register) (p, t1fuv_2, &desc); } #else /* HAVE_FMA */ /* Generated by: ../../../genfft/gen_twiddle_c.native -simd -compact -variables 4 -pipeline-latency 8 -n 2 -name t1fuv_2 -include t1fu.h */ /* * This function contains 3 FP additions, 2 FP multiplications, * (or, 3 additions, 2 multiplications, 0 fused multiply/add), * 5 stack variables, 0 constants, and 4 memory accesses */ #include "t1fu.h" static void t1fuv_2(R *ri, R *ii, const R *W, stride rs, INT mb, INT me, INT ms) { { INT m; R *x; x = ri; for (m = mb, W = W + (mb * ((TWVL / VL) * 2)); m < me; m = m + VL, x = x + (VL * ms), W = W + (TWVL * 2), MAKE_VOLATILE_STRIDE(2, rs)) { V T1, T3, T2; T1 = LD(&(x[0]), ms, &(x[0])); T2 = LD(&(x[WS(rs, 1)]), ms, &(x[WS(rs, 1)])); T3 = BYTWJ(&(W[0]), T2); ST(&(x[WS(rs, 1)]), VSUB(T1, T3), ms, &(x[WS(rs, 1)])); ST(&(x[0]), VADD(T1, T3), ms, &(x[0])); } } VLEAVE(); } static const tw_instr twinstr[] = { VTW(0, 1), {TW_NEXT, VL, 0} }; static const ct_desc desc = { 2, XSIMD_STRING("t1fuv_2"), twinstr, &GENUS, {3, 2, 0, 0}, 0, 0, 0 }; void XSIMD(codelet_t1fuv_2) (planner *p) { X(kdft_dit_register) (p, t1fuv_2, &desc); } #endif /* HAVE_FMA */ fftw-3.3.3/dft/simd/common/t1fuv_3.c0000644000175000017500000001040512054410447014044 00000000000000/* * Copyright (c) 2003, 2007-11 Matteo Frigo * Copyright (c) 2003, 2007-11 Massachusetts Institute of Technology * * This program is free software; you can redistribute it and/or modify * it under the terms of the GNU General Public License as published by * the Free Software Foundation; either version 2 of the License, or * (at your option) any later version. * * This program is distributed in the hope that it will be useful, * but WITHOUT ANY WARRANTY; without even the implied warranty of * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the * GNU General Public License for more details. * * You should have received a copy of the GNU General Public License * along with this program; if not, write to the Free Software * Foundation, Inc., 51 Franklin Street, Fifth Floor, Boston, MA 02110-1301 USA * */ /* This file was automatically generated --- DO NOT EDIT */ /* Generated on Sun Nov 25 07:37:59 EST 2012 */ #include "codelet-dft.h" #ifdef HAVE_FMA /* Generated by: ../../../genfft/gen_twiddle_c.native -fma -reorder-insns -schedule-for-pipeline -simd -compact -variables 4 -pipeline-latency 8 -n 3 -name t1fuv_3 -include t1fu.h */ /* * This function contains 8 FP additions, 8 FP multiplications, * (or, 5 additions, 5 multiplications, 3 fused multiply/add), * 12 stack variables, 2 constants, and 6 memory accesses */ #include "t1fu.h" static void t1fuv_3(R *ri, R *ii, const R *W, stride rs, INT mb, INT me, INT ms) { DVK(KP500000000, +0.500000000000000000000000000000000000000000000); DVK(KP866025403, +0.866025403784438646763723170752936183471402627); { INT m; R *x; x = ri; for (m = mb, W = W + (mb * ((TWVL / VL) * 4)); m < me; m = m + VL, x = x + (VL * ms), W = W + (TWVL * 4), MAKE_VOLATILE_STRIDE(3, rs)) { V T1, T2, T4; T1 = LD(&(x[0]), ms, &(x[0])); T2 = LD(&(x[WS(rs, 1)]), ms, &(x[WS(rs, 1)])); T4 = LD(&(x[WS(rs, 2)]), ms, &(x[0])); { V T3, T5, T8, T6, T7; T3 = BYTWJ(&(W[0]), T2); T5 = BYTWJ(&(W[TWVL * 2]), T4); T8 = VMUL(LDK(KP866025403), VSUB(T5, T3)); T6 = VADD(T3, T5); T7 = VFNMS(LDK(KP500000000), T6, T1); ST(&(x[0]), VADD(T1, T6), ms, &(x[0])); ST(&(x[WS(rs, 1)]), VFMAI(T8, T7), ms, &(x[WS(rs, 1)])); ST(&(x[WS(rs, 2)]), VFNMSI(T8, T7), ms, &(x[0])); } } } VLEAVE(); } static const tw_instr twinstr[] = { VTW(0, 1), VTW(0, 2), {TW_NEXT, VL, 0} }; static const ct_desc desc = { 3, XSIMD_STRING("t1fuv_3"), twinstr, &GENUS, {5, 5, 3, 0}, 0, 0, 0 }; void XSIMD(codelet_t1fuv_3) (planner *p) { X(kdft_dit_register) (p, t1fuv_3, &desc); } #else /* HAVE_FMA */ /* Generated by: ../../../genfft/gen_twiddle_c.native -simd -compact -variables 4 -pipeline-latency 8 -n 3 -name t1fuv_3 -include t1fu.h */ /* * This function contains 8 FP additions, 6 FP multiplications, * (or, 7 additions, 5 multiplications, 1 fused multiply/add), * 12 stack variables, 2 constants, and 6 memory accesses */ #include "t1fu.h" static void t1fuv_3(R *ri, R *ii, const R *W, stride rs, INT mb, INT me, INT ms) { DVK(KP866025403, +0.866025403784438646763723170752936183471402627); DVK(KP500000000, +0.500000000000000000000000000000000000000000000); { INT m; R *x; x = ri; for (m = mb, W = W + (mb * ((TWVL / VL) * 4)); m < me; m = m + VL, x = x + (VL * ms), W = W + (TWVL * 4), MAKE_VOLATILE_STRIDE(3, rs)) { V T1, T3, T5, T6, T2, T4, T7, T8; T1 = LD(&(x[0]), ms, &(x[0])); T2 = LD(&(x[WS(rs, 1)]), ms, &(x[WS(rs, 1)])); T3 = BYTWJ(&(W[0]), T2); T4 = LD(&(x[WS(rs, 2)]), ms, &(x[0])); T5 = BYTWJ(&(W[TWVL * 2]), T4); T6 = VADD(T3, T5); ST(&(x[0]), VADD(T1, T6), ms, &(x[0])); T7 = VFNMS(LDK(KP500000000), T6, T1); T8 = VBYI(VMUL(LDK(KP866025403), VSUB(T5, T3))); ST(&(x[WS(rs, 2)]), VSUB(T7, T8), ms, &(x[0])); ST(&(x[WS(rs, 1)]), VADD(T7, T8), ms, &(x[WS(rs, 1)])); } } VLEAVE(); } static const tw_instr twinstr[] = { VTW(0, 1), VTW(0, 2), {TW_NEXT, VL, 0} }; static const ct_desc desc = { 3, XSIMD_STRING("t1fuv_3"), twinstr, &GENUS, {7, 5, 1, 0}, 0, 0, 0 }; void XSIMD(codelet_t1fuv_3) (planner *p) { X(kdft_dit_register) (p, t1fuv_3, &desc); } #endif /* HAVE_FMA */ fftw-3.3.3/dft/simd/common/t1fuv_4.c0000644000175000017500000001050712054410447014050 00000000000000/* * Copyright (c) 2003, 2007-11 Matteo Frigo * Copyright (c) 2003, 2007-11 Massachusetts Institute of Technology * * This program is free software; you can redistribute it and/or modify * it under the terms of the GNU General Public License as published by * the Free Software Foundation; either version 2 of the License, or * (at your option) any later version. * * This program is distributed in the hope that it will be useful, * but WITHOUT ANY WARRANTY; without even the implied warranty of * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the * GNU General Public License for more details. * * You should have received a copy of the GNU General Public License * along with this program; if not, write to the Free Software * Foundation, Inc., 51 Franklin Street, Fifth Floor, Boston, MA 02110-1301 USA * */ /* This file was automatically generated --- DO NOT EDIT */ /* Generated on Sun Nov 25 07:37:59 EST 2012 */ #include "codelet-dft.h" #ifdef HAVE_FMA /* Generated by: ../../../genfft/gen_twiddle_c.native -fma -reorder-insns -schedule-for-pipeline -simd -compact -variables 4 -pipeline-latency 8 -n 4 -name t1fuv_4 -include t1fu.h */ /* * This function contains 11 FP additions, 8 FP multiplications, * (or, 9 additions, 6 multiplications, 2 fused multiply/add), * 13 stack variables, 0 constants, and 8 memory accesses */ #include "t1fu.h" static void t1fuv_4(R *ri, R *ii, const R *W, stride rs, INT mb, INT me, INT ms) { { INT m; R *x; x = ri; for (m = mb, W = W + (mb * ((TWVL / VL) * 6)); m < me; m = m + VL, x = x + (VL * ms), W = W + (TWVL * 6), MAKE_VOLATILE_STRIDE(4, rs)) { V T1, T7, T2, T5, T8, T3, T6; T1 = LD(&(x[0]), ms, &(x[0])); T7 = LD(&(x[WS(rs, 3)]), ms, &(x[WS(rs, 1)])); T2 = LD(&(x[WS(rs, 2)]), ms, &(x[0])); T5 = LD(&(x[WS(rs, 1)]), ms, &(x[WS(rs, 1)])); T8 = BYTWJ(&(W[TWVL * 4]), T7); T3 = BYTWJ(&(W[TWVL * 2]), T2); T6 = BYTWJ(&(W[0]), T5); { V Ta, T4, Tb, T9; Ta = VADD(T1, T3); T4 = VSUB(T1, T3); Tb = VADD(T6, T8); T9 = VSUB(T6, T8); ST(&(x[0]), VADD(Ta, Tb), ms, &(x[0])); ST(&(x[WS(rs, 2)]), VSUB(Ta, Tb), ms, &(x[0])); ST(&(x[WS(rs, 3)]), VFMAI(T9, T4), ms, &(x[WS(rs, 1)])); ST(&(x[WS(rs, 1)]), VFNMSI(T9, T4), ms, &(x[WS(rs, 1)])); } } } VLEAVE(); } static const tw_instr twinstr[] = { VTW(0, 1), VTW(0, 2), VTW(0, 3), {TW_NEXT, VL, 0} }; static const ct_desc desc = { 4, XSIMD_STRING("t1fuv_4"), twinstr, &GENUS, {9, 6, 2, 0}, 0, 0, 0 }; void XSIMD(codelet_t1fuv_4) (planner *p) { X(kdft_dit_register) (p, t1fuv_4, &desc); } #else /* HAVE_FMA */ /* Generated by: ../../../genfft/gen_twiddle_c.native -simd -compact -variables 4 -pipeline-latency 8 -n 4 -name t1fuv_4 -include t1fu.h */ /* * This function contains 11 FP additions, 6 FP multiplications, * (or, 11 additions, 6 multiplications, 0 fused multiply/add), * 13 stack variables, 0 constants, and 8 memory accesses */ #include "t1fu.h" static void t1fuv_4(R *ri, R *ii, const R *W, stride rs, INT mb, INT me, INT ms) { { INT m; R *x; x = ri; for (m = mb, W = W + (mb * ((TWVL / VL) * 6)); m < me; m = m + VL, x = x + (VL * ms), W = W + (TWVL * 6), MAKE_VOLATILE_STRIDE(4, rs)) { V T1, T8, T3, T6, T7, T2, T5; T1 = LD(&(x[0]), ms, &(x[0])); T7 = LD(&(x[WS(rs, 3)]), ms, &(x[WS(rs, 1)])); T8 = BYTWJ(&(W[TWVL * 4]), T7); T2 = LD(&(x[WS(rs, 2)]), ms, &(x[0])); T3 = BYTWJ(&(W[TWVL * 2]), T2); T5 = LD(&(x[WS(rs, 1)]), ms, &(x[WS(rs, 1)])); T6 = BYTWJ(&(W[0]), T5); { V T4, T9, Ta, Tb; T4 = VSUB(T1, T3); T9 = VBYI(VSUB(T6, T8)); ST(&(x[WS(rs, 1)]), VSUB(T4, T9), ms, &(x[WS(rs, 1)])); ST(&(x[WS(rs, 3)]), VADD(T4, T9), ms, &(x[WS(rs, 1)])); Ta = VADD(T1, T3); Tb = VADD(T6, T8); ST(&(x[WS(rs, 2)]), VSUB(Ta, Tb), ms, &(x[0])); ST(&(x[0]), VADD(Ta, Tb), ms, &(x[0])); } } } VLEAVE(); } static const tw_instr twinstr[] = { VTW(0, 1), VTW(0, 2), VTW(0, 3), {TW_NEXT, VL, 0} }; static const ct_desc desc = { 4, XSIMD_STRING("t1fuv_4"), twinstr, &GENUS, {11, 6, 0, 0}, 0, 0, 0 }; void XSIMD(codelet_t1fuv_4) (planner *p) { X(kdft_dit_register) (p, t1fuv_4, &desc); } #endif /* HAVE_FMA */ fftw-3.3.3/dft/simd/common/t1fuv_5.c0000644000175000017500000001374412054410447014057 00000000000000/* * Copyright (c) 2003, 2007-11 Matteo Frigo * Copyright (c) 2003, 2007-11 Massachusetts Institute of Technology * * This program is free software; you can redistribute it and/or modify * it under the terms of the GNU General Public License as published by * the Free Software Foundation; either version 2 of the License, or * (at your option) any later version. * * This program is distributed in the hope that it will be useful, * but WITHOUT ANY WARRANTY; without even the implied warranty of * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the * GNU General Public License for more details. * * You should have received a copy of the GNU General Public License * along with this program; if not, write to the Free Software * Foundation, Inc., 51 Franklin Street, Fifth Floor, Boston, MA 02110-1301 USA * */ /* This file was automatically generated --- DO NOT EDIT */ /* Generated on Sun Nov 25 07:37:59 EST 2012 */ #include "codelet-dft.h" #ifdef HAVE_FMA /* Generated by: ../../../genfft/gen_twiddle_c.native -fma -reorder-insns -schedule-for-pipeline -simd -compact -variables 4 -pipeline-latency 8 -n 5 -name t1fuv_5 -include t1fu.h */ /* * This function contains 20 FP additions, 19 FP multiplications, * (or, 11 additions, 10 multiplications, 9 fused multiply/add), * 26 stack variables, 4 constants, and 10 memory accesses */ #include "t1fu.h" static void t1fuv_5(R *ri, R *ii, const R *W, stride rs, INT mb, INT me, INT ms) { DVK(KP559016994, +0.559016994374947424102293417182819058860154590); DVK(KP250000000, +0.250000000000000000000000000000000000000000000); DVK(KP618033988, +0.618033988749894848204586834365638117720309180); DVK(KP951056516, +0.951056516295153572116439333379382143405698634); { INT m; R *x; x = ri; for (m = mb, W = W + (mb * ((TWVL / VL) * 8)); m < me; m = m + VL, x = x + (VL * ms), W = W + (TWVL * 8), MAKE_VOLATILE_STRIDE(5, rs)) { V T1, T2, T9, T4, T7; T1 = LD(&(x[0]), ms, &(x[0])); T2 = LD(&(x[WS(rs, 1)]), ms, &(x[WS(rs, 1)])); T9 = LD(&(x[WS(rs, 3)]), ms, &(x[WS(rs, 1)])); T4 = LD(&(x[WS(rs, 4)]), ms, &(x[0])); T7 = LD(&(x[WS(rs, 2)]), ms, &(x[0])); { V T3, Ta, T5, T8; T3 = BYTWJ(&(W[0]), T2); Ta = BYTWJ(&(W[TWVL * 4]), T9); T5 = BYTWJ(&(W[TWVL * 6]), T4); T8 = BYTWJ(&(W[TWVL * 2]), T7); { V T6, Tg, Tb, Th; T6 = VADD(T3, T5); Tg = VSUB(T3, T5); Tb = VADD(T8, Ta); Th = VSUB(T8, Ta); { V Te, Tc, Tk, Ti, Td, Tj, Tf; Te = VSUB(T6, Tb); Tc = VADD(T6, Tb); Tk = VMUL(LDK(KP951056516), VFNMS(LDK(KP618033988), Tg, Th)); Ti = VMUL(LDK(KP951056516), VFMA(LDK(KP618033988), Th, Tg)); Td = VFNMS(LDK(KP250000000), Tc, T1); ST(&(x[0]), VADD(T1, Tc), ms, &(x[0])); Tj = VFNMS(LDK(KP559016994), Te, Td); Tf = VFMA(LDK(KP559016994), Te, Td); ST(&(x[WS(rs, 2)]), VFMAI(Tk, Tj), ms, &(x[0])); ST(&(x[WS(rs, 3)]), VFNMSI(Tk, Tj), ms, &(x[WS(rs, 1)])); ST(&(x[WS(rs, 4)]), VFMAI(Ti, Tf), ms, &(x[0])); ST(&(x[WS(rs, 1)]), VFNMSI(Ti, Tf), ms, &(x[WS(rs, 1)])); } } } } } VLEAVE(); } static const tw_instr twinstr[] = { VTW(0, 1), VTW(0, 2), VTW(0, 3), VTW(0, 4), {TW_NEXT, VL, 0} }; static const ct_desc desc = { 5, XSIMD_STRING("t1fuv_5"), twinstr, &GENUS, {11, 10, 9, 0}, 0, 0, 0 }; void XSIMD(codelet_t1fuv_5) (planner *p) { X(kdft_dit_register) (p, t1fuv_5, &desc); } #else /* HAVE_FMA */ /* Generated by: ../../../genfft/gen_twiddle_c.native -simd -compact -variables 4 -pipeline-latency 8 -n 5 -name t1fuv_5 -include t1fu.h */ /* * This function contains 20 FP additions, 14 FP multiplications, * (or, 17 additions, 11 multiplications, 3 fused multiply/add), * 20 stack variables, 4 constants, and 10 memory accesses */ #include "t1fu.h" static void t1fuv_5(R *ri, R *ii, const R *W, stride rs, INT mb, INT me, INT ms) { DVK(KP250000000, +0.250000000000000000000000000000000000000000000); DVK(KP559016994, +0.559016994374947424102293417182819058860154590); DVK(KP587785252, +0.587785252292473129168705954639072768597652438); DVK(KP951056516, +0.951056516295153572116439333379382143405698634); { INT m; R *x; x = ri; for (m = mb, W = W + (mb * ((TWVL / VL) * 8)); m < me; m = m + VL, x = x + (VL * ms), W = W + (TWVL * 8), MAKE_VOLATILE_STRIDE(5, rs)) { V Tc, Tg, Th, T5, Ta, Td; Tc = LD(&(x[0]), ms, &(x[0])); { V T2, T9, T4, T7; { V T1, T8, T3, T6; T1 = LD(&(x[WS(rs, 1)]), ms, &(x[WS(rs, 1)])); T2 = BYTWJ(&(W[0]), T1); T8 = LD(&(x[WS(rs, 3)]), ms, &(x[WS(rs, 1)])); T9 = BYTWJ(&(W[TWVL * 4]), T8); T3 = LD(&(x[WS(rs, 4)]), ms, &(x[0])); T4 = BYTWJ(&(W[TWVL * 6]), T3); T6 = LD(&(x[WS(rs, 2)]), ms, &(x[0])); T7 = BYTWJ(&(W[TWVL * 2]), T6); } Tg = VSUB(T2, T4); Th = VSUB(T7, T9); T5 = VADD(T2, T4); Ta = VADD(T7, T9); Td = VADD(T5, Ta); } ST(&(x[0]), VADD(Tc, Td), ms, &(x[0])); { V Ti, Tj, Tf, Tk, Tb, Te; Ti = VBYI(VFMA(LDK(KP951056516), Tg, VMUL(LDK(KP587785252), Th))); Tj = VBYI(VFNMS(LDK(KP587785252), Tg, VMUL(LDK(KP951056516), Th))); Tb = VMUL(LDK(KP559016994), VSUB(T5, Ta)); Te = VFNMS(LDK(KP250000000), Td, Tc); Tf = VADD(Tb, Te); Tk = VSUB(Te, Tb); ST(&(x[WS(rs, 1)]), VSUB(Tf, Ti), ms, &(x[WS(rs, 1)])); ST(&(x[WS(rs, 3)]), VSUB(Tk, Tj), ms, &(x[WS(rs, 1)])); ST(&(x[WS(rs, 4)]), VADD(Ti, Tf), ms, &(x[0])); ST(&(x[WS(rs, 2)]), VADD(Tj, Tk), ms, &(x[0])); } } } VLEAVE(); } static const tw_instr twinstr[] = { VTW(0, 1), VTW(0, 2), VTW(0, 3), VTW(0, 4), {TW_NEXT, VL, 0} }; static const ct_desc desc = { 5, XSIMD_STRING("t1fuv_5"), twinstr, &GENUS, {17, 11, 3, 0}, 0, 0, 0 }; void XSIMD(codelet_t1fuv_5) (planner *p) { X(kdft_dit_register) (p, t1fuv_5, &desc); } #endif /* HAVE_FMA */ fftw-3.3.3/dft/simd/common/t1fuv_6.c0000644000175000017500000001403112054410447014046 00000000000000/* * Copyright (c) 2003, 2007-11 Matteo Frigo * Copyright (c) 2003, 2007-11 Massachusetts Institute of Technology * * This program is free software; you can redistribute it and/or modify * it under the terms of the GNU General Public License as published by * the Free Software Foundation; either version 2 of the License, or * (at your option) any later version. * * This program is distributed in the hope that it will be useful, * but WITHOUT ANY WARRANTY; without even the implied warranty of * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the * GNU General Public License for more details. * * You should have received a copy of the GNU General Public License * along with this program; if not, write to the Free Software * Foundation, Inc., 51 Franklin Street, Fifth Floor, Boston, MA 02110-1301 USA * */ /* This file was automatically generated --- DO NOT EDIT */ /* Generated on Sun Nov 25 07:37:59 EST 2012 */ #include "codelet-dft.h" #ifdef HAVE_FMA /* Generated by: ../../../genfft/gen_twiddle_c.native -fma -reorder-insns -schedule-for-pipeline -simd -compact -variables 4 -pipeline-latency 8 -n 6 -name t1fuv_6 -include t1fu.h */ /* * This function contains 23 FP additions, 18 FP multiplications, * (or, 17 additions, 12 multiplications, 6 fused multiply/add), * 27 stack variables, 2 constants, and 12 memory accesses */ #include "t1fu.h" static void t1fuv_6(R *ri, R *ii, const R *W, stride rs, INT mb, INT me, INT ms) { DVK(KP500000000, +0.500000000000000000000000000000000000000000000); DVK(KP866025403, +0.866025403784438646763723170752936183471402627); { INT m; R *x; x = ri; for (m = mb, W = W + (mb * ((TWVL / VL) * 10)); m < me; m = m + VL, x = x + (VL * ms), W = W + (TWVL * 10), MAKE_VOLATILE_STRIDE(6, rs)) { V T1, T2, Ta, Tc, T5, T7; T1 = LD(&(x[0]), ms, &(x[0])); T2 = LD(&(x[WS(rs, 3)]), ms, &(x[WS(rs, 1)])); Ta = LD(&(x[WS(rs, 4)]), ms, &(x[0])); Tc = LD(&(x[WS(rs, 1)]), ms, &(x[WS(rs, 1)])); T5 = LD(&(x[WS(rs, 2)]), ms, &(x[0])); T7 = LD(&(x[WS(rs, 5)]), ms, &(x[WS(rs, 1)])); { V T3, Tb, Td, T6, T8; T3 = BYTWJ(&(W[TWVL * 4]), T2); Tb = BYTWJ(&(W[TWVL * 6]), Ta); Td = BYTWJ(&(W[0]), Tc); T6 = BYTWJ(&(W[TWVL * 2]), T5); T8 = BYTWJ(&(W[TWVL * 8]), T7); { V Ti, T4, Tk, Te, Tj, T9; Ti = VADD(T1, T3); T4 = VSUB(T1, T3); Tk = VADD(Tb, Td); Te = VSUB(Tb, Td); Tj = VADD(T6, T8); T9 = VSUB(T6, T8); { V Tl, Tn, Tf, Th, Tm, Tg; Tl = VADD(Tj, Tk); Tn = VMUL(LDK(KP866025403), VSUB(Tk, Tj)); Tf = VADD(T9, Te); Th = VMUL(LDK(KP866025403), VSUB(Te, T9)); ST(&(x[0]), VADD(Ti, Tl), ms, &(x[0])); Tm = VFNMS(LDK(KP500000000), Tl, Ti); ST(&(x[WS(rs, 3)]), VADD(T4, Tf), ms, &(x[WS(rs, 1)])); Tg = VFNMS(LDK(KP500000000), Tf, T4); ST(&(x[WS(rs, 2)]), VFNMSI(Tn, Tm), ms, &(x[0])); ST(&(x[WS(rs, 4)]), VFMAI(Tn, Tm), ms, &(x[0])); ST(&(x[WS(rs, 5)]), VFNMSI(Th, Tg), ms, &(x[WS(rs, 1)])); ST(&(x[WS(rs, 1)]), VFMAI(Th, Tg), ms, &(x[WS(rs, 1)])); } } } } } VLEAVE(); } static const tw_instr twinstr[] = { VTW(0, 1), VTW(0, 2), VTW(0, 3), VTW(0, 4), VTW(0, 5), {TW_NEXT, VL, 0} }; static const ct_desc desc = { 6, XSIMD_STRING("t1fuv_6"), twinstr, &GENUS, {17, 12, 6, 0}, 0, 0, 0 }; void XSIMD(codelet_t1fuv_6) (planner *p) { X(kdft_dit_register) (p, t1fuv_6, &desc); } #else /* HAVE_FMA */ /* Generated by: ../../../genfft/gen_twiddle_c.native -simd -compact -variables 4 -pipeline-latency 8 -n 6 -name t1fuv_6 -include t1fu.h */ /* * This function contains 23 FP additions, 14 FP multiplications, * (or, 21 additions, 12 multiplications, 2 fused multiply/add), * 19 stack variables, 2 constants, and 12 memory accesses */ #include "t1fu.h" static void t1fuv_6(R *ri, R *ii, const R *W, stride rs, INT mb, INT me, INT ms) { DVK(KP500000000, +0.500000000000000000000000000000000000000000000); DVK(KP866025403, +0.866025403784438646763723170752936183471402627); { INT m; R *x; x = ri; for (m = mb, W = W + (mb * ((TWVL / VL) * 10)); m < me; m = m + VL, x = x + (VL * ms), W = W + (TWVL * 10), MAKE_VOLATILE_STRIDE(6, rs)) { V T4, Ti, Te, Tk, T9, Tj, T1, T3, T2; T1 = LD(&(x[0]), ms, &(x[0])); T2 = LD(&(x[WS(rs, 3)]), ms, &(x[WS(rs, 1)])); T3 = BYTWJ(&(W[TWVL * 4]), T2); T4 = VSUB(T1, T3); Ti = VADD(T1, T3); { V Tb, Td, Ta, Tc; Ta = LD(&(x[WS(rs, 4)]), ms, &(x[0])); Tb = BYTWJ(&(W[TWVL * 6]), Ta); Tc = LD(&(x[WS(rs, 1)]), ms, &(x[WS(rs, 1)])); Td = BYTWJ(&(W[0]), Tc); Te = VSUB(Tb, Td); Tk = VADD(Tb, Td); } { V T6, T8, T5, T7; T5 = LD(&(x[WS(rs, 2)]), ms, &(x[0])); T6 = BYTWJ(&(W[TWVL * 2]), T5); T7 = LD(&(x[WS(rs, 5)]), ms, &(x[WS(rs, 1)])); T8 = BYTWJ(&(W[TWVL * 8]), T7); T9 = VSUB(T6, T8); Tj = VADD(T6, T8); } { V Th, Tf, Tg, Tn, Tl, Tm; Th = VBYI(VMUL(LDK(KP866025403), VSUB(Te, T9))); Tf = VADD(T9, Te); Tg = VFNMS(LDK(KP500000000), Tf, T4); ST(&(x[WS(rs, 3)]), VADD(T4, Tf), ms, &(x[WS(rs, 1)])); ST(&(x[WS(rs, 1)]), VADD(Tg, Th), ms, &(x[WS(rs, 1)])); ST(&(x[WS(rs, 5)]), VSUB(Tg, Th), ms, &(x[WS(rs, 1)])); Tn = VBYI(VMUL(LDK(KP866025403), VSUB(Tk, Tj))); Tl = VADD(Tj, Tk); Tm = VFNMS(LDK(KP500000000), Tl, Ti); ST(&(x[0]), VADD(Ti, Tl), ms, &(x[0])); ST(&(x[WS(rs, 4)]), VADD(Tm, Tn), ms, &(x[0])); ST(&(x[WS(rs, 2)]), VSUB(Tm, Tn), ms, &(x[0])); } } } VLEAVE(); } static const tw_instr twinstr[] = { VTW(0, 1), VTW(0, 2), VTW(0, 3), VTW(0, 4), VTW(0, 5), {TW_NEXT, VL, 0} }; static const ct_desc desc = { 6, XSIMD_STRING("t1fuv_6"), twinstr, &GENUS, {21, 12, 2, 0}, 0, 0, 0 }; void XSIMD(codelet_t1fuv_6) (planner *p) { X(kdft_dit_register) (p, t1fuv_6, &desc); } #endif /* HAVE_FMA */ fftw-3.3.3/dft/simd/common/t1fuv_7.c0000644000175000017500000001762112054410450014051 00000000000000/* * Copyright (c) 2003, 2007-11 Matteo Frigo * Copyright (c) 2003, 2007-11 Massachusetts Institute of Technology * * This program is free software; you can redistribute it and/or modify * it under the terms of the GNU General Public License as published by * the Free Software Foundation; either version 2 of the License, or * (at your option) any later version. * * This program is distributed in the hope that it will be useful, * but WITHOUT ANY WARRANTY; without even the implied warranty of * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the * GNU General Public License for more details. * * You should have received a copy of the GNU General Public License * along with this program; if not, write to the Free Software * Foundation, Inc., 51 Franklin Street, Fifth Floor, Boston, MA 02110-1301 USA * */ /* This file was automatically generated --- DO NOT EDIT */ /* Generated on Sun Nov 25 07:37:59 EST 2012 */ #include "codelet-dft.h" #ifdef HAVE_FMA /* Generated by: ../../../genfft/gen_twiddle_c.native -fma -reorder-insns -schedule-for-pipeline -simd -compact -variables 4 -pipeline-latency 8 -n 7 -name t1fuv_7 -include t1fu.h */ /* * This function contains 36 FP additions, 36 FP multiplications, * (or, 15 additions, 15 multiplications, 21 fused multiply/add), * 42 stack variables, 6 constants, and 14 memory accesses */ #include "t1fu.h" static void t1fuv_7(R *ri, R *ii, const R *W, stride rs, INT mb, INT me, INT ms) { DVK(KP900968867, +0.900968867902419126236102319507445051165919162); DVK(KP801937735, +0.801937735804838252472204639014890102331838324); DVK(KP974927912, +0.974927912181823607018131682993931217232785801); DVK(KP692021471, +0.692021471630095869627814897002069140197260599); DVK(KP554958132, +0.554958132087371191422194871006410481067288862); DVK(KP356895867, +0.356895867892209443894399510021300583399127187); { INT m; R *x; x = ri; for (m = mb, W = W + (mb * ((TWVL / VL) * 12)); m < me; m = m + VL, x = x + (VL * ms), W = W + (TWVL * 12), MAKE_VOLATILE_STRIDE(7, rs)) { V T1, T2, T4, Te, Tc, T9, T7; T1 = LD(&(x[0]), ms, &(x[0])); T2 = LD(&(x[WS(rs, 1)]), ms, &(x[WS(rs, 1)])); T4 = LD(&(x[WS(rs, 6)]), ms, &(x[0])); Te = LD(&(x[WS(rs, 4)]), ms, &(x[0])); Tc = LD(&(x[WS(rs, 3)]), ms, &(x[WS(rs, 1)])); T9 = LD(&(x[WS(rs, 5)]), ms, &(x[WS(rs, 1)])); T7 = LD(&(x[WS(rs, 2)]), ms, &(x[0])); { V T3, T5, Tf, Td, Ta, T8; T3 = BYTWJ(&(W[0]), T2); T5 = BYTWJ(&(W[TWVL * 10]), T4); Tf = BYTWJ(&(W[TWVL * 6]), Te); Td = BYTWJ(&(W[TWVL * 4]), Tc); Ta = BYTWJ(&(W[TWVL * 8]), T9); T8 = BYTWJ(&(W[TWVL * 2]), T7); { V T6, Tk, Tg, Tl, Tb, Tm; T6 = VADD(T3, T5); Tk = VSUB(T5, T3); Tg = VADD(Td, Tf); Tl = VSUB(Tf, Td); Tb = VADD(T8, Ta); Tm = VSUB(Ta, T8); { V Th, Ts, Tp, Tu, Tn, Tx, Ti, Tt; Th = VFNMS(LDK(KP356895867), T6, Tg); Ts = VFMA(LDK(KP554958132), Tl, Tk); ST(&(x[0]), VADD(T1, VADD(T6, VADD(Tb, Tg))), ms, &(x[0])); Tp = VFNMS(LDK(KP356895867), Tb, T6); Tu = VFNMS(LDK(KP356895867), Tg, Tb); Tn = VFMA(LDK(KP554958132), Tm, Tl); Tx = VFNMS(LDK(KP554958132), Tk, Tm); Ti = VFNMS(LDK(KP692021471), Th, Tb); Tt = VMUL(LDK(KP974927912), VFMA(LDK(KP801937735), Ts, Tm)); { V Tq, Tv, To, Ty, Tj, Tr, Tw; Tq = VFNMS(LDK(KP692021471), Tp, Tg); Tv = VFNMS(LDK(KP692021471), Tu, T6); To = VMUL(LDK(KP974927912), VFNMS(LDK(KP801937735), Tn, Tk)); Ty = VMUL(LDK(KP974927912), VFNMS(LDK(KP801937735), Tx, Tl)); Tj = VFNMS(LDK(KP900968867), Ti, T1); Tr = VFNMS(LDK(KP900968867), Tq, T1); Tw = VFNMS(LDK(KP900968867), Tv, T1); ST(&(x[WS(rs, 2)]), VFMAI(To, Tj), ms, &(x[0])); ST(&(x[WS(rs, 5)]), VFNMSI(To, Tj), ms, &(x[WS(rs, 1)])); ST(&(x[WS(rs, 1)]), VFMAI(Tt, Tr), ms, &(x[WS(rs, 1)])); ST(&(x[WS(rs, 6)]), VFNMSI(Tt, Tr), ms, &(x[0])); ST(&(x[WS(rs, 3)]), VFMAI(Ty, Tw), ms, &(x[WS(rs, 1)])); ST(&(x[WS(rs, 4)]), VFNMSI(Ty, Tw), ms, &(x[0])); } } } } } } VLEAVE(); } static const tw_instr twinstr[] = { VTW(0, 1), VTW(0, 2), VTW(0, 3), VTW(0, 4), VTW(0, 5), VTW(0, 6), {TW_NEXT, VL, 0} }; static const ct_desc desc = { 7, XSIMD_STRING("t1fuv_7"), twinstr, &GENUS, {15, 15, 21, 0}, 0, 0, 0 }; void XSIMD(codelet_t1fuv_7) (planner *p) { X(kdft_dit_register) (p, t1fuv_7, &desc); } #else /* HAVE_FMA */ /* Generated by: ../../../genfft/gen_twiddle_c.native -simd -compact -variables 4 -pipeline-latency 8 -n 7 -name t1fuv_7 -include t1fu.h */ /* * This function contains 36 FP additions, 30 FP multiplications, * (or, 24 additions, 18 multiplications, 12 fused multiply/add), * 21 stack variables, 6 constants, and 14 memory accesses */ #include "t1fu.h" static void t1fuv_7(R *ri, R *ii, const R *W, stride rs, INT mb, INT me, INT ms) { DVK(KP900968867, +0.900968867902419126236102319507445051165919162); DVK(KP222520933, +0.222520933956314404288902564496794759466355569); DVK(KP623489801, +0.623489801858733530525004884004239810632274731); DVK(KP781831482, +0.781831482468029808708444526674057750232334519); DVK(KP974927912, +0.974927912181823607018131682993931217232785801); DVK(KP433883739, +0.433883739117558120475768332848358754609990728); { INT m; R *x; x = ri; for (m = mb, W = W + (mb * ((TWVL / VL) * 12)); m < me; m = m + VL, x = x + (VL * ms), W = W + (TWVL * 12), MAKE_VOLATILE_STRIDE(7, rs)) { V T1, Tg, Tj, T6, Ti, Tb, Tk, Tp, To; T1 = LD(&(x[0]), ms, &(x[0])); { V Td, Tf, Tc, Te; Tc = LD(&(x[WS(rs, 3)]), ms, &(x[WS(rs, 1)])); Td = BYTWJ(&(W[TWVL * 4]), Tc); Te = LD(&(x[WS(rs, 4)]), ms, &(x[0])); Tf = BYTWJ(&(W[TWVL * 6]), Te); Tg = VADD(Td, Tf); Tj = VSUB(Tf, Td); } { V T3, T5, T2, T4; T2 = LD(&(x[WS(rs, 1)]), ms, &(x[WS(rs, 1)])); T3 = BYTWJ(&(W[0]), T2); T4 = LD(&(x[WS(rs, 6)]), ms, &(x[0])); T5 = BYTWJ(&(W[TWVL * 10]), T4); T6 = VADD(T3, T5); Ti = VSUB(T5, T3); } { V T8, Ta, T7, T9; T7 = LD(&(x[WS(rs, 2)]), ms, &(x[0])); T8 = BYTWJ(&(W[TWVL * 2]), T7); T9 = LD(&(x[WS(rs, 5)]), ms, &(x[WS(rs, 1)])); Ta = BYTWJ(&(W[TWVL * 8]), T9); Tb = VADD(T8, Ta); Tk = VSUB(Ta, T8); } ST(&(x[0]), VADD(T1, VADD(T6, VADD(Tb, Tg))), ms, &(x[0])); Tp = VBYI(VFMA(LDK(KP433883739), Ti, VFNMS(LDK(KP781831482), Tk, VMUL(LDK(KP974927912), Tj)))); To = VFMA(LDK(KP623489801), Tb, VFNMS(LDK(KP222520933), Tg, VFNMS(LDK(KP900968867), T6, T1))); ST(&(x[WS(rs, 4)]), VSUB(To, Tp), ms, &(x[0])); ST(&(x[WS(rs, 3)]), VADD(To, Tp), ms, &(x[WS(rs, 1)])); { V Tl, Th, Tn, Tm; Tl = VBYI(VFNMS(LDK(KP781831482), Tj, VFNMS(LDK(KP433883739), Tk, VMUL(LDK(KP974927912), Ti)))); Th = VFMA(LDK(KP623489801), Tg, VFNMS(LDK(KP900968867), Tb, VFNMS(LDK(KP222520933), T6, T1))); ST(&(x[WS(rs, 5)]), VSUB(Th, Tl), ms, &(x[WS(rs, 1)])); ST(&(x[WS(rs, 2)]), VADD(Th, Tl), ms, &(x[0])); Tn = VBYI(VFMA(LDK(KP781831482), Ti, VFMA(LDK(KP974927912), Tk, VMUL(LDK(KP433883739), Tj)))); Tm = VFMA(LDK(KP623489801), T6, VFNMS(LDK(KP900968867), Tg, VFNMS(LDK(KP222520933), Tb, T1))); ST(&(x[WS(rs, 6)]), VSUB(Tm, Tn), ms, &(x[0])); ST(&(x[WS(rs, 1)]), VADD(Tm, Tn), ms, &(x[WS(rs, 1)])); } } } VLEAVE(); } static const tw_instr twinstr[] = { VTW(0, 1), VTW(0, 2), VTW(0, 3), VTW(0, 4), VTW(0, 5), VTW(0, 6), {TW_NEXT, VL, 0} }; static const ct_desc desc = { 7, XSIMD_STRING("t1fuv_7"), twinstr, &GENUS, {24, 18, 12, 0}, 0, 0, 0 }; void XSIMD(codelet_t1fuv_7) (planner *p) { X(kdft_dit_register) (p, t1fuv_7, &desc); } #endif /* HAVE_FMA */ fftw-3.3.3/dft/simd/common/t1fuv_8.c0000644000175000017500000001562112054410450014050 00000000000000/* * Copyright (c) 2003, 2007-11 Matteo Frigo * Copyright (c) 2003, 2007-11 Massachusetts Institute of Technology * * This program is free software; you can redistribute it and/or modify * it under the terms of the GNU General Public License as published by * the Free Software Foundation; either version 2 of the License, or * (at your option) any later version. * * This program is distributed in the hope that it will be useful, * but WITHOUT ANY WARRANTY; without even the implied warranty of * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the * GNU General Public License for more details. * * You should have received a copy of the GNU General Public License * along with this program; if not, write to the Free Software * Foundation, Inc., 51 Franklin Street, Fifth Floor, Boston, MA 02110-1301 USA * */ /* This file was automatically generated --- DO NOT EDIT */ /* Generated on Sun Nov 25 07:38:00 EST 2012 */ #include "codelet-dft.h" #ifdef HAVE_FMA /* Generated by: ../../../genfft/gen_twiddle_c.native -fma -reorder-insns -schedule-for-pipeline -simd -compact -variables 4 -pipeline-latency 8 -n 8 -name t1fuv_8 -include t1fu.h */ /* * This function contains 33 FP additions, 24 FP multiplications, * (or, 23 additions, 14 multiplications, 10 fused multiply/add), * 36 stack variables, 1 constants, and 16 memory accesses */ #include "t1fu.h" static void t1fuv_8(R *ri, R *ii, const R *W, stride rs, INT mb, INT me, INT ms) { DVK(KP707106781, +0.707106781186547524400844362104849039284835938); { INT m; R *x; x = ri; for (m = mb, W = W + (mb * ((TWVL / VL) * 14)); m < me; m = m + VL, x = x + (VL * ms), W = W + (TWVL * 14), MAKE_VOLATILE_STRIDE(8, rs)) { V T1, T2, Th, Tj, T5, T7, Ta, Tc; T1 = LD(&(x[0]), ms, &(x[0])); T2 = LD(&(x[WS(rs, 4)]), ms, &(x[0])); Th = LD(&(x[WS(rs, 2)]), ms, &(x[0])); Tj = LD(&(x[WS(rs, 6)]), ms, &(x[0])); T5 = LD(&(x[WS(rs, 1)]), ms, &(x[WS(rs, 1)])); T7 = LD(&(x[WS(rs, 5)]), ms, &(x[WS(rs, 1)])); Ta = LD(&(x[WS(rs, 7)]), ms, &(x[WS(rs, 1)])); Tc = LD(&(x[WS(rs, 3)]), ms, &(x[WS(rs, 1)])); { V T3, Ti, Tk, T6, T8, Tb, Td; T3 = BYTWJ(&(W[TWVL * 6]), T2); Ti = BYTWJ(&(W[TWVL * 2]), Th); Tk = BYTWJ(&(W[TWVL * 10]), Tj); T6 = BYTWJ(&(W[0]), T5); T8 = BYTWJ(&(W[TWVL * 8]), T7); Tb = BYTWJ(&(W[TWVL * 12]), Ta); Td = BYTWJ(&(W[TWVL * 4]), Tc); { V Tq, T4, Tr, Tl, Tt, T9, Tu, Te, Tw, Ts; Tq = VADD(T1, T3); T4 = VSUB(T1, T3); Tr = VADD(Ti, Tk); Tl = VSUB(Ti, Tk); Tt = VADD(T6, T8); T9 = VSUB(T6, T8); Tu = VADD(Tb, Td); Te = VSUB(Tb, Td); Tw = VSUB(Tq, Tr); Ts = VADD(Tq, Tr); { V Tx, Tv, Tm, Tf; Tx = VSUB(Tu, Tt); Tv = VADD(Tt, Tu); Tm = VSUB(Te, T9); Tf = VADD(T9, Te); { V Tp, Tn, To, Tg; ST(&(x[WS(rs, 2)]), VFMAI(Tx, Tw), ms, &(x[0])); ST(&(x[WS(rs, 6)]), VFNMSI(Tx, Tw), ms, &(x[0])); ST(&(x[0]), VADD(Ts, Tv), ms, &(x[0])); ST(&(x[WS(rs, 4)]), VSUB(Ts, Tv), ms, &(x[0])); Tp = VFMA(LDK(KP707106781), Tm, Tl); Tn = VFNMS(LDK(KP707106781), Tm, Tl); To = VFNMS(LDK(KP707106781), Tf, T4); Tg = VFMA(LDK(KP707106781), Tf, T4); ST(&(x[WS(rs, 5)]), VFNMSI(Tp, To), ms, &(x[WS(rs, 1)])); ST(&(x[WS(rs, 3)]), VFMAI(Tp, To), ms, &(x[WS(rs, 1)])); ST(&(x[WS(rs, 7)]), VFMAI(Tn, Tg), ms, &(x[WS(rs, 1)])); ST(&(x[WS(rs, 1)]), VFNMSI(Tn, Tg), ms, &(x[WS(rs, 1)])); } } } } } } VLEAVE(); } static const tw_instr twinstr[] = { VTW(0, 1), VTW(0, 2), VTW(0, 3), VTW(0, 4), VTW(0, 5), VTW(0, 6), VTW(0, 7), {TW_NEXT, VL, 0} }; static const ct_desc desc = { 8, XSIMD_STRING("t1fuv_8"), twinstr, &GENUS, {23, 14, 10, 0}, 0, 0, 0 }; void XSIMD(codelet_t1fuv_8) (planner *p) { X(kdft_dit_register) (p, t1fuv_8, &desc); } #else /* HAVE_FMA */ /* Generated by: ../../../genfft/gen_twiddle_c.native -simd -compact -variables 4 -pipeline-latency 8 -n 8 -name t1fuv_8 -include t1fu.h */ /* * This function contains 33 FP additions, 16 FP multiplications, * (or, 33 additions, 16 multiplications, 0 fused multiply/add), * 24 stack variables, 1 constants, and 16 memory accesses */ #include "t1fu.h" static void t1fuv_8(R *ri, R *ii, const R *W, stride rs, INT mb, INT me, INT ms) { DVK(KP707106781, +0.707106781186547524400844362104849039284835938); { INT m; R *x; x = ri; for (m = mb, W = W + (mb * ((TWVL / VL) * 14)); m < me; m = m + VL, x = x + (VL * ms), W = W + (TWVL * 14), MAKE_VOLATILE_STRIDE(8, rs)) { V T4, Tq, Tm, Tr, T9, Tt, Te, Tu, T1, T3, T2; T1 = LD(&(x[0]), ms, &(x[0])); T2 = LD(&(x[WS(rs, 4)]), ms, &(x[0])); T3 = BYTWJ(&(W[TWVL * 6]), T2); T4 = VSUB(T1, T3); Tq = VADD(T1, T3); { V Tj, Tl, Ti, Tk; Ti = LD(&(x[WS(rs, 2)]), ms, &(x[0])); Tj = BYTWJ(&(W[TWVL * 2]), Ti); Tk = LD(&(x[WS(rs, 6)]), ms, &(x[0])); Tl = BYTWJ(&(W[TWVL * 10]), Tk); Tm = VSUB(Tj, Tl); Tr = VADD(Tj, Tl); } { V T6, T8, T5, T7; T5 = LD(&(x[WS(rs, 1)]), ms, &(x[WS(rs, 1)])); T6 = BYTWJ(&(W[0]), T5); T7 = LD(&(x[WS(rs, 5)]), ms, &(x[WS(rs, 1)])); T8 = BYTWJ(&(W[TWVL * 8]), T7); T9 = VSUB(T6, T8); Tt = VADD(T6, T8); } { V Tb, Td, Ta, Tc; Ta = LD(&(x[WS(rs, 7)]), ms, &(x[WS(rs, 1)])); Tb = BYTWJ(&(W[TWVL * 12]), Ta); Tc = LD(&(x[WS(rs, 3)]), ms, &(x[WS(rs, 1)])); Td = BYTWJ(&(W[TWVL * 4]), Tc); Te = VSUB(Tb, Td); Tu = VADD(Tb, Td); } { V Ts, Tv, Tw, Tx; Ts = VADD(Tq, Tr); Tv = VADD(Tt, Tu); ST(&(x[WS(rs, 4)]), VSUB(Ts, Tv), ms, &(x[0])); ST(&(x[0]), VADD(Ts, Tv), ms, &(x[0])); Tw = VSUB(Tq, Tr); Tx = VBYI(VSUB(Tu, Tt)); ST(&(x[WS(rs, 6)]), VSUB(Tw, Tx), ms, &(x[0])); ST(&(x[WS(rs, 2)]), VADD(Tw, Tx), ms, &(x[0])); { V Tg, To, Tn, Tp, Tf, Th; Tf = VMUL(LDK(KP707106781), VADD(T9, Te)); Tg = VADD(T4, Tf); To = VSUB(T4, Tf); Th = VMUL(LDK(KP707106781), VSUB(Te, T9)); Tn = VBYI(VSUB(Th, Tm)); Tp = VBYI(VADD(Tm, Th)); ST(&(x[WS(rs, 7)]), VSUB(Tg, Tn), ms, &(x[WS(rs, 1)])); ST(&(x[WS(rs, 3)]), VADD(To, Tp), ms, &(x[WS(rs, 1)])); ST(&(x[WS(rs, 1)]), VADD(Tg, Tn), ms, &(x[WS(rs, 1)])); ST(&(x[WS(rs, 5)]), VSUB(To, Tp), ms, &(x[WS(rs, 1)])); } } } } VLEAVE(); } static const tw_instr twinstr[] = { VTW(0, 1), VTW(0, 2), VTW(0, 3), VTW(0, 4), VTW(0, 5), VTW(0, 6), VTW(0, 7), {TW_NEXT, VL, 0} }; static const ct_desc desc = { 8, XSIMD_STRING("t1fuv_8"), twinstr, &GENUS, {33, 16, 0, 0}, 0, 0, 0 }; void XSIMD(codelet_t1fuv_8) (planner *p) { X(kdft_dit_register) (p, t1fuv_8, &desc); } #endif /* HAVE_FMA */ fftw-3.3.3/dft/simd/common/t1fuv_9.c0000644000175000017500000002655512054410451014062 00000000000000/* * Copyright (c) 2003, 2007-11 Matteo Frigo * Copyright (c) 2003, 2007-11 Massachusetts Institute of Technology * * This program is free software; you can redistribute it and/or modify * it under the terms of the GNU General Public License as published by * the Free Software Foundation; either version 2 of the License, or * (at your option) any later version. * * This program is distributed in the hope that it will be useful, * but WITHOUT ANY WARRANTY; without even the implied warranty of * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the * GNU General Public License for more details. * * You should have received a copy of the GNU General Public License * along with this program; if not, write to the Free Software * Foundation, Inc., 51 Franklin Street, Fifth Floor, Boston, MA 02110-1301 USA * */ /* This file was automatically generated --- DO NOT EDIT */ /* Generated on Sun Nov 25 07:38:00 EST 2012 */ #include "codelet-dft.h" #ifdef HAVE_FMA /* Generated by: ../../../genfft/gen_twiddle_c.native -fma -reorder-insns -schedule-for-pipeline -simd -compact -variables 4 -pipeline-latency 8 -n 9 -name t1fuv_9 -include t1fu.h */ /* * This function contains 54 FP additions, 54 FP multiplications, * (or, 20 additions, 20 multiplications, 34 fused multiply/add), * 67 stack variables, 19 constants, and 18 memory accesses */ #include "t1fu.h" static void t1fuv_9(R *ri, R *ii, const R *W, stride rs, INT mb, INT me, INT ms) { DVK(KP939692620, +0.939692620785908384054109277324731469936208134); DVK(KP826351822, +0.826351822333069651148283373230685203999624323); DVK(KP879385241, +0.879385241571816768108218554649462939872416269); DVK(KP984807753, +0.984807753012208059366743024589523013670643252); DVK(KP666666666, +0.666666666666666666666666666666666666666666667); DVK(KP852868531, +0.852868531952443209628250963940074071936020296); DVK(KP907603734, +0.907603734547952313649323976213898122064543220); DVK(KP420276625, +0.420276625461206169731530603237061658838781920); DVK(KP673648177, +0.673648177666930348851716626769314796000375677); DVK(KP898197570, +0.898197570222573798468955502359086394667167570); DVK(KP347296355, +0.347296355333860697703433253538629592000751354); DVK(KP866025403, +0.866025403784438646763723170752936183471402627); DVK(KP439692620, +0.439692620785908384054109277324731469936208134); DVK(KP203604859, +0.203604859554852403062088995281827210665664861); DVK(KP152703644, +0.152703644666139302296566746461370407999248646); DVK(KP586256827, +0.586256827714544512072145703099641959914944179); DVK(KP968908795, +0.968908795874236621082202410917456709164223497); DVK(KP726681596, +0.726681596905677465811651808188092531873167623); DVK(KP500000000, +0.500000000000000000000000000000000000000000000); { INT m; R *x; x = ri; for (m = mb, W = W + (mb * ((TWVL / VL) * 16)); m < me; m = m + VL, x = x + (VL * ms), W = W + (TWVL * 16), MAKE_VOLATILE_STRIDE(9, rs)) { V T1, T3, T5, T9, Th, Tb, Td, Tj, Tl, TD, T6; T1 = LD(&(x[0]), ms, &(x[0])); { V T2, T4, T8, Tg; T2 = LD(&(x[WS(rs, 3)]), ms, &(x[WS(rs, 1)])); T4 = LD(&(x[WS(rs, 6)]), ms, &(x[0])); T8 = LD(&(x[WS(rs, 1)]), ms, &(x[WS(rs, 1)])); Tg = LD(&(x[WS(rs, 2)]), ms, &(x[0])); { V Ta, Tc, Ti, Tk; Ta = LD(&(x[WS(rs, 4)]), ms, &(x[0])); Tc = LD(&(x[WS(rs, 7)]), ms, &(x[WS(rs, 1)])); Ti = LD(&(x[WS(rs, 5)]), ms, &(x[WS(rs, 1)])); Tk = LD(&(x[WS(rs, 8)]), ms, &(x[0])); T3 = BYTWJ(&(W[TWVL * 4]), T2); T5 = BYTWJ(&(W[TWVL * 10]), T4); T9 = BYTWJ(&(W[0]), T8); Th = BYTWJ(&(W[TWVL * 2]), Tg); Tb = BYTWJ(&(W[TWVL * 6]), Ta); Td = BYTWJ(&(W[TWVL * 12]), Tc); Tj = BYTWJ(&(W[TWVL * 8]), Ti); Tl = BYTWJ(&(W[TWVL * 14]), Tk); } } TD = VSUB(T5, T3); T6 = VADD(T3, T5); { V Tt, Te, Tu, Tm, Tr, T7; Tt = VSUB(Tb, Td); Te = VADD(Tb, Td); Tu = VSUB(Tl, Tj); Tm = VADD(Tj, Tl); Tr = VFNMS(LDK(KP500000000), T6, T1); T7 = VADD(T1, T6); { V Tv, Tf, Ts, Tn; Tv = VFNMS(LDK(KP500000000), Te, T9); Tf = VADD(T9, Te); Ts = VFNMS(LDK(KP500000000), Tm, Th); Tn = VADD(Th, Tm); { V TG, TK, Tw, TJ, TF, TA, To, Tq; TG = VFNMS(LDK(KP726681596), Tt, Tv); TK = VFMA(LDK(KP968908795), Tv, Tt); Tw = VFNMS(LDK(KP586256827), Tv, Tu); TJ = VFNMS(LDK(KP152703644), Tu, Ts); TF = VFMA(LDK(KP203604859), Ts, Tu); TA = VFNMS(LDK(KP439692620), Tt, Ts); To = VADD(Tf, Tn); Tq = VMUL(LDK(KP866025403), VSUB(Tn, Tf)); { V TQ, TH, TL, TN, TB, Tp, Ty, TI, Tx; Tx = VFNMS(LDK(KP347296355), Tw, Tt); TQ = VFNMS(LDK(KP898197570), TG, TF); TH = VFMA(LDK(KP898197570), TG, TF); TL = VFMA(LDK(KP673648177), TK, TJ); TN = VFNMS(LDK(KP673648177), TK, TJ); TB = VFNMS(LDK(KP420276625), TA, Tu); ST(&(x[0]), VADD(T7, To), ms, &(x[0])); Tp = VFNMS(LDK(KP500000000), To, T7); Ty = VFNMS(LDK(KP907603734), Tx, Ts); TI = VFMA(LDK(KP852868531), TH, Tr); { V TO, TR, TM, TC, Tz, TP, TS, TE; TO = VFNMS(LDK(KP500000000), TH, TN); TR = VFMA(LDK(KP666666666), TL, TQ); TM = VMUL(LDK(KP984807753), VFNMS(LDK(KP879385241), TD, TL)); TC = VFNMS(LDK(KP826351822), TB, Tv); ST(&(x[WS(rs, 6)]), VFNMSI(Tq, Tp), ms, &(x[0])); ST(&(x[WS(rs, 3)]), VFMAI(Tq, Tp), ms, &(x[WS(rs, 1)])); Tz = VFNMS(LDK(KP939692620), Ty, Tr); TP = VFMA(LDK(KP852868531), TO, Tr); TS = VMUL(LDK(KP866025403), VFMA(LDK(KP852868531), TR, TD)); ST(&(x[WS(rs, 8)]), VFMAI(TM, TI), ms, &(x[0])); ST(&(x[WS(rs, 1)]), VFNMSI(TM, TI), ms, &(x[WS(rs, 1)])); TE = VMUL(LDK(KP984807753), VFMA(LDK(KP879385241), TD, TC)); ST(&(x[WS(rs, 4)]), VFMAI(TS, TP), ms, &(x[0])); ST(&(x[WS(rs, 5)]), VFNMSI(TS, TP), ms, &(x[WS(rs, 1)])); ST(&(x[WS(rs, 7)]), VFMAI(TE, Tz), ms, &(x[WS(rs, 1)])); ST(&(x[WS(rs, 2)]), VFNMSI(TE, Tz), ms, &(x[0])); } } } } } } } VLEAVE(); } static const tw_instr twinstr[] = { VTW(0, 1), VTW(0, 2), VTW(0, 3), VTW(0, 4), VTW(0, 5), VTW(0, 6), VTW(0, 7), VTW(0, 8), {TW_NEXT, VL, 0} }; static const ct_desc desc = { 9, XSIMD_STRING("t1fuv_9"), twinstr, &GENUS, {20, 20, 34, 0}, 0, 0, 0 }; void XSIMD(codelet_t1fuv_9) (planner *p) { X(kdft_dit_register) (p, t1fuv_9, &desc); } #else /* HAVE_FMA */ /* Generated by: ../../../genfft/gen_twiddle_c.native -simd -compact -variables 4 -pipeline-latency 8 -n 9 -name t1fuv_9 -include t1fu.h */ /* * This function contains 54 FP additions, 42 FP multiplications, * (or, 38 additions, 26 multiplications, 16 fused multiply/add), * 38 stack variables, 14 constants, and 18 memory accesses */ #include "t1fu.h" static void t1fuv_9(R *ri, R *ii, const R *W, stride rs, INT mb, INT me, INT ms) { DVK(KP939692620, +0.939692620785908384054109277324731469936208134); DVK(KP296198132, +0.296198132726023843175338011893050938967728390); DVK(KP852868531, +0.852868531952443209628250963940074071936020296); DVK(KP173648177, +0.173648177666930348851716626769314796000375677); DVK(KP556670399, +0.556670399226419366452912952047023132968291906); DVK(KP766044443, +0.766044443118978035202392650555416673935832457); DVK(KP642787609, +0.642787609686539326322643409907263432907559884); DVK(KP663413948, +0.663413948168938396205421319635891297216863310); DVK(KP984807753, +0.984807753012208059366743024589523013670643252); DVK(KP150383733, +0.150383733180435296639271897612501926072238258); DVK(KP342020143, +0.342020143325668733044099614682259580763083368); DVK(KP813797681, +0.813797681349373692844693217248393223289101568); DVK(KP500000000, +0.500000000000000000000000000000000000000000000); DVK(KP866025403, +0.866025403784438646763723170752936183471402627); { INT m; R *x; x = ri; for (m = mb, W = W + (mb * ((TWVL / VL) * 16)); m < me; m = m + VL, x = x + (VL * ms), W = W + (TWVL * 16), MAKE_VOLATILE_STRIDE(9, rs)) { V T1, T6, TA, Tt, Tf, Ts, Tw, Tn, Tv; T1 = LD(&(x[0]), ms, &(x[0])); { V T3, T5, T2, T4; T2 = LD(&(x[WS(rs, 3)]), ms, &(x[WS(rs, 1)])); T3 = BYTWJ(&(W[TWVL * 4]), T2); T4 = LD(&(x[WS(rs, 6)]), ms, &(x[0])); T5 = BYTWJ(&(W[TWVL * 10]), T4); T6 = VADD(T3, T5); TA = VMUL(LDK(KP866025403), VSUB(T5, T3)); } { V T9, Td, Tb, T8, Tc, Ta, Te; T8 = LD(&(x[WS(rs, 1)]), ms, &(x[WS(rs, 1)])); T9 = BYTWJ(&(W[0]), T8); Tc = LD(&(x[WS(rs, 7)]), ms, &(x[WS(rs, 1)])); Td = BYTWJ(&(W[TWVL * 12]), Tc); Ta = LD(&(x[WS(rs, 4)]), ms, &(x[0])); Tb = BYTWJ(&(W[TWVL * 6]), Ta); Tt = VSUB(Td, Tb); Te = VADD(Tb, Td); Tf = VADD(T9, Te); Ts = VFNMS(LDK(KP500000000), Te, T9); } { V Th, Tl, Tj, Tg, Tk, Ti, Tm; Tg = LD(&(x[WS(rs, 2)]), ms, &(x[0])); Th = BYTWJ(&(W[TWVL * 2]), Tg); Tk = LD(&(x[WS(rs, 8)]), ms, &(x[0])); Tl = BYTWJ(&(W[TWVL * 14]), Tk); Ti = LD(&(x[WS(rs, 5)]), ms, &(x[WS(rs, 1)])); Tj = BYTWJ(&(W[TWVL * 8]), Ti); Tw = VSUB(Tl, Tj); Tm = VADD(Tj, Tl); Tn = VADD(Th, Tm); Tv = VFNMS(LDK(KP500000000), Tm, Th); } { V Tq, T7, To, Tp; Tq = VBYI(VMUL(LDK(KP866025403), VSUB(Tn, Tf))); T7 = VADD(T1, T6); To = VADD(Tf, Tn); Tp = VFNMS(LDK(KP500000000), To, T7); ST(&(x[0]), VADD(T7, To), ms, &(x[0])); ST(&(x[WS(rs, 3)]), VADD(Tp, Tq), ms, &(x[WS(rs, 1)])); ST(&(x[WS(rs, 6)]), VSUB(Tp, Tq), ms, &(x[0])); } { V TI, TB, TC, TD, Tu, Tx, Ty, Tr, TH; TI = VBYI(VSUB(VFNMS(LDK(KP342020143), Tv, VFNMS(LDK(KP150383733), Tt, VFNMS(LDK(KP984807753), Ts, VMUL(LDK(KP813797681), Tw)))), TA)); TB = VFNMS(LDK(KP642787609), Ts, VMUL(LDK(KP663413948), Tt)); TC = VFNMS(LDK(KP984807753), Tv, VMUL(LDK(KP150383733), Tw)); TD = VADD(TB, TC); Tu = VFMA(LDK(KP766044443), Ts, VMUL(LDK(KP556670399), Tt)); Tx = VFMA(LDK(KP173648177), Tv, VMUL(LDK(KP852868531), Tw)); Ty = VADD(Tu, Tx); Tr = VFNMS(LDK(KP500000000), T6, T1); TH = VFMA(LDK(KP173648177), Ts, VFNMS(LDK(KP296198132), Tw, VFNMS(LDK(KP939692620), Tv, VFNMS(LDK(KP852868531), Tt, Tr)))); ST(&(x[WS(rs, 7)]), VSUB(TH, TI), ms, &(x[WS(rs, 1)])); ST(&(x[WS(rs, 2)]), VADD(TH, TI), ms, &(x[0])); { V Tz, TE, TF, TG; Tz = VADD(Tr, Ty); TE = VBYI(VADD(TA, TD)); ST(&(x[WS(rs, 8)]), VSUB(Tz, TE), ms, &(x[0])); ST(&(x[WS(rs, 1)]), VADD(TE, Tz), ms, &(x[WS(rs, 1)])); TF = VFMA(LDK(KP866025403), VSUB(TB, TC), VFNMS(LDK(KP500000000), Ty, Tr)); TG = VBYI(VADD(TA, VFNMS(LDK(KP500000000), TD, VMUL(LDK(KP866025403), VSUB(Tx, Tu))))); ST(&(x[WS(rs, 5)]), VSUB(TF, TG), ms, &(x[WS(rs, 1)])); ST(&(x[WS(rs, 4)]), VADD(TF, TG), ms, &(x[0])); } } } } VLEAVE(); } static const tw_instr twinstr[] = { VTW(0, 1), VTW(0, 2), VTW(0, 3), VTW(0, 4), VTW(0, 5), VTW(0, 6), VTW(0, 7), VTW(0, 8), {TW_NEXT, VL, 0} }; static const ct_desc desc = { 9, XSIMD_STRING("t1fuv_9"), twinstr, &GENUS, {38, 26, 16, 0}, 0, 0, 0 }; void XSIMD(codelet_t1fuv_9) (planner *p) { X(kdft_dit_register) (p, t1fuv_9, &desc); } #endif /* HAVE_FMA */ fftw-3.3.3/dft/simd/common/t1fuv_10.c0000644000175000017500000002256312054410451014125 00000000000000/* * Copyright (c) 2003, 2007-11 Matteo Frigo * Copyright (c) 2003, 2007-11 Massachusetts Institute of Technology * * This program is free software; you can redistribute it and/or modify * it under the terms of the GNU General Public License as published by * the Free Software Foundation; either version 2 of the License, or * (at your option) any later version. * * This program is distributed in the hope that it will be useful, * but WITHOUT ANY WARRANTY; without even the implied warranty of * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the * GNU General Public License for more details. * * You should have received a copy of the GNU General Public License * along with this program; if not, write to the Free Software * Foundation, Inc., 51 Franklin Street, Fifth Floor, Boston, MA 02110-1301 USA * */ /* This file was automatically generated --- DO NOT EDIT */ /* Generated on Sun Nov 25 07:38:01 EST 2012 */ #include "codelet-dft.h" #ifdef HAVE_FMA /* Generated by: ../../../genfft/gen_twiddle_c.native -fma -reorder-insns -schedule-for-pipeline -simd -compact -variables 4 -pipeline-latency 8 -n 10 -name t1fuv_10 -include t1fu.h */ /* * This function contains 51 FP additions, 40 FP multiplications, * (or, 33 additions, 22 multiplications, 18 fused multiply/add), * 43 stack variables, 4 constants, and 20 memory accesses */ #include "t1fu.h" static void t1fuv_10(R *ri, R *ii, const R *W, stride rs, INT mb, INT me, INT ms) { DVK(KP559016994, +0.559016994374947424102293417182819058860154590); DVK(KP250000000, +0.250000000000000000000000000000000000000000000); DVK(KP618033988, +0.618033988749894848204586834365638117720309180); DVK(KP951056516, +0.951056516295153572116439333379382143405698634); { INT m; R *x; x = ri; for (m = mb, W = W + (mb * ((TWVL / VL) * 18)); m < me; m = m + VL, x = x + (VL * ms), W = W + (TWVL * 18), MAKE_VOLATILE_STRIDE(10, rs)) { V Td, TA, T4, Ta, Tk, TE, Tp, TF, TB, T9, T1, T2, Tb; T1 = LD(&(x[0]), ms, &(x[0])); T2 = LD(&(x[WS(rs, 5)]), ms, &(x[WS(rs, 1)])); { V Tg, Tn, Ti, Tl; Tg = LD(&(x[WS(rs, 4)]), ms, &(x[0])); Tn = LD(&(x[WS(rs, 1)]), ms, &(x[WS(rs, 1)])); Ti = LD(&(x[WS(rs, 9)]), ms, &(x[WS(rs, 1)])); Tl = LD(&(x[WS(rs, 6)]), ms, &(x[0])); { V T6, T8, T5, Tc; T5 = LD(&(x[WS(rs, 2)]), ms, &(x[0])); Tc = LD(&(x[WS(rs, 3)]), ms, &(x[WS(rs, 1)])); { V T3, Th, To, Tj, Tm, T7; T7 = LD(&(x[WS(rs, 7)]), ms, &(x[WS(rs, 1)])); T3 = BYTWJ(&(W[TWVL * 8]), T2); Th = BYTWJ(&(W[TWVL * 6]), Tg); To = BYTWJ(&(W[0]), Tn); Tj = BYTWJ(&(W[TWVL * 16]), Ti); Tm = BYTWJ(&(W[TWVL * 10]), Tl); T6 = BYTWJ(&(W[TWVL * 2]), T5); Td = BYTWJ(&(W[TWVL * 4]), Tc); T8 = BYTWJ(&(W[TWVL * 12]), T7); TA = VADD(T1, T3); T4 = VSUB(T1, T3); Ta = LD(&(x[WS(rs, 8)]), ms, &(x[0])); Tk = VSUB(Th, Tj); TE = VADD(Th, Tj); Tp = VSUB(Tm, To); TF = VADD(Tm, To); } TB = VADD(T6, T8); T9 = VSUB(T6, T8); } } Tb = BYTWJ(&(W[TWVL * 14]), Ta); { V TL, TG, Tw, Tq, TC, Te; TL = VSUB(TE, TF); TG = VADD(TE, TF); Tw = VSUB(Tk, Tp); Tq = VADD(Tk, Tp); TC = VADD(Tb, Td); Te = VSUB(Tb, Td); { V TM, TD, Tv, Tf; TM = VSUB(TB, TC); TD = VADD(TB, TC); Tv = VSUB(T9, Te); Tf = VADD(T9, Te); { V TP, TN, TH, TJ, Tz, Tx, Tr, Tt, TI, Ts; TP = VMUL(LDK(KP951056516), VFMA(LDK(KP618033988), TL, TM)); TN = VMUL(LDK(KP951056516), VFNMS(LDK(KP618033988), TM, TL)); TH = VADD(TD, TG); TJ = VSUB(TD, TG); Tz = VMUL(LDK(KP951056516), VFNMS(LDK(KP618033988), Tv, Tw)); Tx = VMUL(LDK(KP951056516), VFMA(LDK(KP618033988), Tw, Tv)); Tr = VADD(Tf, Tq); Tt = VSUB(Tf, Tq); ST(&(x[0]), VADD(TA, TH), ms, &(x[0])); TI = VFNMS(LDK(KP250000000), TH, TA); ST(&(x[WS(rs, 5)]), VADD(T4, Tr), ms, &(x[WS(rs, 1)])); Ts = VFNMS(LDK(KP250000000), Tr, T4); { V TK, TO, Tu, Ty; TK = VFNMS(LDK(KP559016994), TJ, TI); TO = VFMA(LDK(KP559016994), TJ, TI); Tu = VFMA(LDK(KP559016994), Tt, Ts); Ty = VFNMS(LDK(KP559016994), Tt, Ts); ST(&(x[WS(rs, 8)]), VFNMSI(TN, TK), ms, &(x[0])); ST(&(x[WS(rs, 2)]), VFMAI(TN, TK), ms, &(x[0])); ST(&(x[WS(rs, 6)]), VFNMSI(TP, TO), ms, &(x[0])); ST(&(x[WS(rs, 4)]), VFMAI(TP, TO), ms, &(x[0])); ST(&(x[WS(rs, 9)]), VFMAI(Tx, Tu), ms, &(x[WS(rs, 1)])); ST(&(x[WS(rs, 1)]), VFNMSI(Tx, Tu), ms, &(x[WS(rs, 1)])); ST(&(x[WS(rs, 7)]), VFMAI(Tz, Ty), ms, &(x[WS(rs, 1)])); ST(&(x[WS(rs, 3)]), VFNMSI(Tz, Ty), ms, &(x[WS(rs, 1)])); } } } } } } VLEAVE(); } static const tw_instr twinstr[] = { VTW(0, 1), VTW(0, 2), VTW(0, 3), VTW(0, 4), VTW(0, 5), VTW(0, 6), VTW(0, 7), VTW(0, 8), VTW(0, 9), {TW_NEXT, VL, 0} }; static const ct_desc desc = { 10, XSIMD_STRING("t1fuv_10"), twinstr, &GENUS, {33, 22, 18, 0}, 0, 0, 0 }; void XSIMD(codelet_t1fuv_10) (planner *p) { X(kdft_dit_register) (p, t1fuv_10, &desc); } #else /* HAVE_FMA */ /* Generated by: ../../../genfft/gen_twiddle_c.native -simd -compact -variables 4 -pipeline-latency 8 -n 10 -name t1fuv_10 -include t1fu.h */ /* * This function contains 51 FP additions, 30 FP multiplications, * (or, 45 additions, 24 multiplications, 6 fused multiply/add), * 32 stack variables, 4 constants, and 20 memory accesses */ #include "t1fu.h" static void t1fuv_10(R *ri, R *ii, const R *W, stride rs, INT mb, INT me, INT ms) { DVK(KP587785252, +0.587785252292473129168705954639072768597652438); DVK(KP951056516, +0.951056516295153572116439333379382143405698634); DVK(KP250000000, +0.250000000000000000000000000000000000000000000); DVK(KP559016994, +0.559016994374947424102293417182819058860154590); { INT m; R *x; x = ri; for (m = mb, W = W + (mb * ((TWVL / VL) * 18)); m < me; m = m + VL, x = x + (VL * ms), W = W + (TWVL * 18), MAKE_VOLATILE_STRIDE(10, rs)) { V Tr, TH, Tg, Tl, Tm, TA, TB, TJ, T5, Ta, Tb, TD, TE, TI, To; V Tq, Tp; To = LD(&(x[0]), ms, &(x[0])); Tp = LD(&(x[WS(rs, 5)]), ms, &(x[WS(rs, 1)])); Tq = BYTWJ(&(W[TWVL * 8]), Tp); Tr = VSUB(To, Tq); TH = VADD(To, Tq); { V Td, Tk, Tf, Ti; { V Tc, Tj, Te, Th; Tc = LD(&(x[WS(rs, 4)]), ms, &(x[0])); Td = BYTWJ(&(W[TWVL * 6]), Tc); Tj = LD(&(x[WS(rs, 1)]), ms, &(x[WS(rs, 1)])); Tk = BYTWJ(&(W[0]), Tj); Te = LD(&(x[WS(rs, 9)]), ms, &(x[WS(rs, 1)])); Tf = BYTWJ(&(W[TWVL * 16]), Te); Th = LD(&(x[WS(rs, 6)]), ms, &(x[0])); Ti = BYTWJ(&(W[TWVL * 10]), Th); } Tg = VSUB(Td, Tf); Tl = VSUB(Ti, Tk); Tm = VADD(Tg, Tl); TA = VADD(Td, Tf); TB = VADD(Ti, Tk); TJ = VADD(TA, TB); } { V T2, T9, T4, T7; { V T1, T8, T3, T6; T1 = LD(&(x[WS(rs, 2)]), ms, &(x[0])); T2 = BYTWJ(&(W[TWVL * 2]), T1); T8 = LD(&(x[WS(rs, 3)]), ms, &(x[WS(rs, 1)])); T9 = BYTWJ(&(W[TWVL * 4]), T8); T3 = LD(&(x[WS(rs, 7)]), ms, &(x[WS(rs, 1)])); T4 = BYTWJ(&(W[TWVL * 12]), T3); T6 = LD(&(x[WS(rs, 8)]), ms, &(x[0])); T7 = BYTWJ(&(W[TWVL * 14]), T6); } T5 = VSUB(T2, T4); Ta = VSUB(T7, T9); Tb = VADD(T5, Ta); TD = VADD(T2, T4); TE = VADD(T7, T9); TI = VADD(TD, TE); } { V Tn, Ts, Tt, Tx, Tz, Tv, Tw, Ty, Tu; Tn = VMUL(LDK(KP559016994), VSUB(Tb, Tm)); Ts = VADD(Tb, Tm); Tt = VFNMS(LDK(KP250000000), Ts, Tr); Tv = VSUB(T5, Ta); Tw = VSUB(Tg, Tl); Tx = VBYI(VFMA(LDK(KP951056516), Tv, VMUL(LDK(KP587785252), Tw))); Tz = VBYI(VFNMS(LDK(KP587785252), Tv, VMUL(LDK(KP951056516), Tw))); ST(&(x[WS(rs, 5)]), VADD(Tr, Ts), ms, &(x[WS(rs, 1)])); Ty = VSUB(Tt, Tn); ST(&(x[WS(rs, 3)]), VSUB(Ty, Tz), ms, &(x[WS(rs, 1)])); ST(&(x[WS(rs, 7)]), VADD(Tz, Ty), ms, &(x[WS(rs, 1)])); Tu = VADD(Tn, Tt); ST(&(x[WS(rs, 1)]), VSUB(Tu, Tx), ms, &(x[WS(rs, 1)])); ST(&(x[WS(rs, 9)]), VADD(Tx, Tu), ms, &(x[WS(rs, 1)])); } { V TM, TK, TL, TG, TO, TC, TF, TP, TN; TM = VMUL(LDK(KP559016994), VSUB(TI, TJ)); TK = VADD(TI, TJ); TL = VFNMS(LDK(KP250000000), TK, TH); TC = VSUB(TA, TB); TF = VSUB(TD, TE); TG = VBYI(VFNMS(LDK(KP587785252), TF, VMUL(LDK(KP951056516), TC))); TO = VBYI(VFMA(LDK(KP951056516), TF, VMUL(LDK(KP587785252), TC))); ST(&(x[0]), VADD(TH, TK), ms, &(x[0])); TP = VADD(TM, TL); ST(&(x[WS(rs, 4)]), VADD(TO, TP), ms, &(x[0])); ST(&(x[WS(rs, 6)]), VSUB(TP, TO), ms, &(x[0])); TN = VSUB(TL, TM); ST(&(x[WS(rs, 2)]), VADD(TG, TN), ms, &(x[0])); ST(&(x[WS(rs, 8)]), VSUB(TN, TG), ms, &(x[0])); } } } VLEAVE(); } static const tw_instr twinstr[] = { VTW(0, 1), VTW(0, 2), VTW(0, 3), VTW(0, 4), VTW(0, 5), VTW(0, 6), VTW(0, 7), VTW(0, 8), VTW(0, 9), {TW_NEXT, VL, 0} }; static const ct_desc desc = { 10, XSIMD_STRING("t1fuv_10"), twinstr, &GENUS, {45, 24, 6, 0}, 0, 0, 0 }; void XSIMD(codelet_t1fuv_10) (planner *p) { X(kdft_dit_register) (p, t1fuv_10, &desc); } #endif /* HAVE_FMA */ fftw-3.3.3/dft/simd/common/t1fv_2.c0000644000175000017500000000653612054410451013663 00000000000000/* * Copyright (c) 2003, 2007-11 Matteo Frigo * Copyright (c) 2003, 2007-11 Massachusetts Institute of Technology * * This program is free software; you can redistribute it and/or modify * it under the terms of the GNU General Public License as published by * the Free Software Foundation; either version 2 of the License, or * (at your option) any later version. * * This program is distributed in the hope that it will be useful, * but WITHOUT ANY WARRANTY; without even the implied warranty of * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the * GNU General Public License for more details. * * You should have received a copy of the GNU General Public License * along with this program; if not, write to the Free Software * Foundation, Inc., 51 Franklin Street, Fifth Floor, Boston, MA 02110-1301 USA * */ /* This file was automatically generated --- DO NOT EDIT */ /* Generated on Sun Nov 25 07:38:01 EST 2012 */ #include "codelet-dft.h" #ifdef HAVE_FMA /* Generated by: ../../../genfft/gen_twiddle_c.native -fma -reorder-insns -schedule-for-pipeline -simd -compact -variables 4 -pipeline-latency 8 -n 2 -name t1fv_2 -include t1f.h */ /* * This function contains 3 FP additions, 2 FP multiplications, * (or, 3 additions, 2 multiplications, 0 fused multiply/add), * 5 stack variables, 0 constants, and 4 memory accesses */ #include "t1f.h" static void t1fv_2(R *ri, R *ii, const R *W, stride rs, INT mb, INT me, INT ms) { { INT m; R *x; x = ri; for (m = mb, W = W + (mb * ((TWVL / VL) * 2)); m < me; m = m + VL, x = x + (VL * ms), W = W + (TWVL * 2), MAKE_VOLATILE_STRIDE(2, rs)) { V T1, T2, T3; T1 = LD(&(x[0]), ms, &(x[0])); T2 = LD(&(x[WS(rs, 1)]), ms, &(x[WS(rs, 1)])); T3 = BYTWJ(&(W[0]), T2); ST(&(x[0]), VADD(T1, T3), ms, &(x[0])); ST(&(x[WS(rs, 1)]), VSUB(T1, T3), ms, &(x[WS(rs, 1)])); } } VLEAVE(); } static const tw_instr twinstr[] = { VTW(0, 1), {TW_NEXT, VL, 0} }; static const ct_desc desc = { 2, XSIMD_STRING("t1fv_2"), twinstr, &GENUS, {3, 2, 0, 0}, 0, 0, 0 }; void XSIMD(codelet_t1fv_2) (planner *p) { X(kdft_dit_register) (p, t1fv_2, &desc); } #else /* HAVE_FMA */ /* Generated by: ../../../genfft/gen_twiddle_c.native -simd -compact -variables 4 -pipeline-latency 8 -n 2 -name t1fv_2 -include t1f.h */ /* * This function contains 3 FP additions, 2 FP multiplications, * (or, 3 additions, 2 multiplications, 0 fused multiply/add), * 5 stack variables, 0 constants, and 4 memory accesses */ #include "t1f.h" static void t1fv_2(R *ri, R *ii, const R *W, stride rs, INT mb, INT me, INT ms) { { INT m; R *x; x = ri; for (m = mb, W = W + (mb * ((TWVL / VL) * 2)); m < me; m = m + VL, x = x + (VL * ms), W = W + (TWVL * 2), MAKE_VOLATILE_STRIDE(2, rs)) { V T1, T3, T2; T1 = LD(&(x[0]), ms, &(x[0])); T2 = LD(&(x[WS(rs, 1)]), ms, &(x[WS(rs, 1)])); T3 = BYTWJ(&(W[0]), T2); ST(&(x[WS(rs, 1)]), VSUB(T1, T3), ms, &(x[WS(rs, 1)])); ST(&(x[0]), VADD(T1, T3), ms, &(x[0])); } } VLEAVE(); } static const tw_instr twinstr[] = { VTW(0, 1), {TW_NEXT, VL, 0} }; static const ct_desc desc = { 2, XSIMD_STRING("t1fv_2"), twinstr, &GENUS, {3, 2, 0, 0}, 0, 0, 0 }; void XSIMD(codelet_t1fv_2) (planner *p) { X(kdft_dit_register) (p, t1fv_2, &desc); } #endif /* HAVE_FMA */ fftw-3.3.3/dft/simd/common/t1fv_3.c0000644000175000017500000001036712054410451013661 00000000000000/* * Copyright (c) 2003, 2007-11 Matteo Frigo * Copyright (c) 2003, 2007-11 Massachusetts Institute of Technology * * This program is free software; you can redistribute it and/or modify * it under the terms of the GNU General Public License as published by * the Free Software Foundation; either version 2 of the License, or * (at your option) any later version. * * This program is distributed in the hope that it will be useful, * but WITHOUT ANY WARRANTY; without even the implied warranty of * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the * GNU General Public License for more details. * * You should have received a copy of the GNU General Public License * along with this program; if not, write to the Free Software * Foundation, Inc., 51 Franklin Street, Fifth Floor, Boston, MA 02110-1301 USA * */ /* This file was automatically generated --- DO NOT EDIT */ /* Generated on Sun Nov 25 07:38:01 EST 2012 */ #include "codelet-dft.h" #ifdef HAVE_FMA /* Generated by: ../../../genfft/gen_twiddle_c.native -fma -reorder-insns -schedule-for-pipeline -simd -compact -variables 4 -pipeline-latency 8 -n 3 -name t1fv_3 -include t1f.h */ /* * This function contains 8 FP additions, 8 FP multiplications, * (or, 5 additions, 5 multiplications, 3 fused multiply/add), * 12 stack variables, 2 constants, and 6 memory accesses */ #include "t1f.h" static void t1fv_3(R *ri, R *ii, const R *W, stride rs, INT mb, INT me, INT ms) { DVK(KP500000000, +0.500000000000000000000000000000000000000000000); DVK(KP866025403, +0.866025403784438646763723170752936183471402627); { INT m; R *x; x = ri; for (m = mb, W = W + (mb * ((TWVL / VL) * 4)); m < me; m = m + VL, x = x + (VL * ms), W = W + (TWVL * 4), MAKE_VOLATILE_STRIDE(3, rs)) { V T1, T2, T4; T1 = LD(&(x[0]), ms, &(x[0])); T2 = LD(&(x[WS(rs, 1)]), ms, &(x[WS(rs, 1)])); T4 = LD(&(x[WS(rs, 2)]), ms, &(x[0])); { V T3, T5, T8, T6, T7; T3 = BYTWJ(&(W[0]), T2); T5 = BYTWJ(&(W[TWVL * 2]), T4); T8 = VMUL(LDK(KP866025403), VSUB(T5, T3)); T6 = VADD(T3, T5); T7 = VFNMS(LDK(KP500000000), T6, T1); ST(&(x[0]), VADD(T1, T6), ms, &(x[0])); ST(&(x[WS(rs, 1)]), VFMAI(T8, T7), ms, &(x[WS(rs, 1)])); ST(&(x[WS(rs, 2)]), VFNMSI(T8, T7), ms, &(x[0])); } } } VLEAVE(); } static const tw_instr twinstr[] = { VTW(0, 1), VTW(0, 2), {TW_NEXT, VL, 0} }; static const ct_desc desc = { 3, XSIMD_STRING("t1fv_3"), twinstr, &GENUS, {5, 5, 3, 0}, 0, 0, 0 }; void XSIMD(codelet_t1fv_3) (planner *p) { X(kdft_dit_register) (p, t1fv_3, &desc); } #else /* HAVE_FMA */ /* Generated by: ../../../genfft/gen_twiddle_c.native -simd -compact -variables 4 -pipeline-latency 8 -n 3 -name t1fv_3 -include t1f.h */ /* * This function contains 8 FP additions, 6 FP multiplications, * (or, 7 additions, 5 multiplications, 1 fused multiply/add), * 12 stack variables, 2 constants, and 6 memory accesses */ #include "t1f.h" static void t1fv_3(R *ri, R *ii, const R *W, stride rs, INT mb, INT me, INT ms) { DVK(KP866025403, +0.866025403784438646763723170752936183471402627); DVK(KP500000000, +0.500000000000000000000000000000000000000000000); { INT m; R *x; x = ri; for (m = mb, W = W + (mb * ((TWVL / VL) * 4)); m < me; m = m + VL, x = x + (VL * ms), W = W + (TWVL * 4), MAKE_VOLATILE_STRIDE(3, rs)) { V T1, T3, T5, T6, T2, T4, T7, T8; T1 = LD(&(x[0]), ms, &(x[0])); T2 = LD(&(x[WS(rs, 1)]), ms, &(x[WS(rs, 1)])); T3 = BYTWJ(&(W[0]), T2); T4 = LD(&(x[WS(rs, 2)]), ms, &(x[0])); T5 = BYTWJ(&(W[TWVL * 2]), T4); T6 = VADD(T3, T5); ST(&(x[0]), VADD(T1, T6), ms, &(x[0])); T7 = VFNMS(LDK(KP500000000), T6, T1); T8 = VBYI(VMUL(LDK(KP866025403), VSUB(T5, T3))); ST(&(x[WS(rs, 2)]), VSUB(T7, T8), ms, &(x[0])); ST(&(x[WS(rs, 1)]), VADD(T7, T8), ms, &(x[WS(rs, 1)])); } } VLEAVE(); } static const tw_instr twinstr[] = { VTW(0, 1), VTW(0, 2), {TW_NEXT, VL, 0} }; static const ct_desc desc = { 3, XSIMD_STRING("t1fv_3"), twinstr, &GENUS, {7, 5, 1, 0}, 0, 0, 0 }; void XSIMD(codelet_t1fv_3) (planner *p) { X(kdft_dit_register) (p, t1fv_3, &desc); } #endif /* HAVE_FMA */ fftw-3.3.3/dft/simd/common/t1fv_4.c0000644000175000017500000001047112054410451013656 00000000000000/* * Copyright (c) 2003, 2007-11 Matteo Frigo * Copyright (c) 2003, 2007-11 Massachusetts Institute of Technology * * This program is free software; you can redistribute it and/or modify * it under the terms of the GNU General Public License as published by * the Free Software Foundation; either version 2 of the License, or * (at your option) any later version. * * This program is distributed in the hope that it will be useful, * but WITHOUT ANY WARRANTY; without even the implied warranty of * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the * GNU General Public License for more details. * * You should have received a copy of the GNU General Public License * along with this program; if not, write to the Free Software * Foundation, Inc., 51 Franklin Street, Fifth Floor, Boston, MA 02110-1301 USA * */ /* This file was automatically generated --- DO NOT EDIT */ /* Generated on Sun Nov 25 07:38:01 EST 2012 */ #include "codelet-dft.h" #ifdef HAVE_FMA /* Generated by: ../../../genfft/gen_twiddle_c.native -fma -reorder-insns -schedule-for-pipeline -simd -compact -variables 4 -pipeline-latency 8 -n 4 -name t1fv_4 -include t1f.h */ /* * This function contains 11 FP additions, 8 FP multiplications, * (or, 9 additions, 6 multiplications, 2 fused multiply/add), * 13 stack variables, 0 constants, and 8 memory accesses */ #include "t1f.h" static void t1fv_4(R *ri, R *ii, const R *W, stride rs, INT mb, INT me, INT ms) { { INT m; R *x; x = ri; for (m = mb, W = W + (mb * ((TWVL / VL) * 6)); m < me; m = m + VL, x = x + (VL * ms), W = W + (TWVL * 6), MAKE_VOLATILE_STRIDE(4, rs)) { V T1, T7, T2, T5, T8, T3, T6; T1 = LD(&(x[0]), ms, &(x[0])); T7 = LD(&(x[WS(rs, 3)]), ms, &(x[WS(rs, 1)])); T2 = LD(&(x[WS(rs, 2)]), ms, &(x[0])); T5 = LD(&(x[WS(rs, 1)]), ms, &(x[WS(rs, 1)])); T8 = BYTWJ(&(W[TWVL * 4]), T7); T3 = BYTWJ(&(W[TWVL * 2]), T2); T6 = BYTWJ(&(W[0]), T5); { V Ta, T4, Tb, T9; Ta = VADD(T1, T3); T4 = VSUB(T1, T3); Tb = VADD(T6, T8); T9 = VSUB(T6, T8); ST(&(x[0]), VADD(Ta, Tb), ms, &(x[0])); ST(&(x[WS(rs, 2)]), VSUB(Ta, Tb), ms, &(x[0])); ST(&(x[WS(rs, 3)]), VFMAI(T9, T4), ms, &(x[WS(rs, 1)])); ST(&(x[WS(rs, 1)]), VFNMSI(T9, T4), ms, &(x[WS(rs, 1)])); } } } VLEAVE(); } static const tw_instr twinstr[] = { VTW(0, 1), VTW(0, 2), VTW(0, 3), {TW_NEXT, VL, 0} }; static const ct_desc desc = { 4, XSIMD_STRING("t1fv_4"), twinstr, &GENUS, {9, 6, 2, 0}, 0, 0, 0 }; void XSIMD(codelet_t1fv_4) (planner *p) { X(kdft_dit_register) (p, t1fv_4, &desc); } #else /* HAVE_FMA */ /* Generated by: ../../../genfft/gen_twiddle_c.native -simd -compact -variables 4 -pipeline-latency 8 -n 4 -name t1fv_4 -include t1f.h */ /* * This function contains 11 FP additions, 6 FP multiplications, * (or, 11 additions, 6 multiplications, 0 fused multiply/add), * 13 stack variables, 0 constants, and 8 memory accesses */ #include "t1f.h" static void t1fv_4(R *ri, R *ii, const R *W, stride rs, INT mb, INT me, INT ms) { { INT m; R *x; x = ri; for (m = mb, W = W + (mb * ((TWVL / VL) * 6)); m < me; m = m + VL, x = x + (VL * ms), W = W + (TWVL * 6), MAKE_VOLATILE_STRIDE(4, rs)) { V T1, T8, T3, T6, T7, T2, T5; T1 = LD(&(x[0]), ms, &(x[0])); T7 = LD(&(x[WS(rs, 3)]), ms, &(x[WS(rs, 1)])); T8 = BYTWJ(&(W[TWVL * 4]), T7); T2 = LD(&(x[WS(rs, 2)]), ms, &(x[0])); T3 = BYTWJ(&(W[TWVL * 2]), T2); T5 = LD(&(x[WS(rs, 1)]), ms, &(x[WS(rs, 1)])); T6 = BYTWJ(&(W[0]), T5); { V T4, T9, Ta, Tb; T4 = VSUB(T1, T3); T9 = VBYI(VSUB(T6, T8)); ST(&(x[WS(rs, 1)]), VSUB(T4, T9), ms, &(x[WS(rs, 1)])); ST(&(x[WS(rs, 3)]), VADD(T4, T9), ms, &(x[WS(rs, 1)])); Ta = VADD(T1, T3); Tb = VADD(T6, T8); ST(&(x[WS(rs, 2)]), VSUB(Ta, Tb), ms, &(x[0])); ST(&(x[0]), VADD(Ta, Tb), ms, &(x[0])); } } } VLEAVE(); } static const tw_instr twinstr[] = { VTW(0, 1), VTW(0, 2), VTW(0, 3), {TW_NEXT, VL, 0} }; static const ct_desc desc = { 4, XSIMD_STRING("t1fv_4"), twinstr, &GENUS, {11, 6, 0, 0}, 0, 0, 0 }; void XSIMD(codelet_t1fv_4) (planner *p) { X(kdft_dit_register) (p, t1fv_4, &desc); } #endif /* HAVE_FMA */ fftw-3.3.3/dft/simd/common/t1fv_5.c0000644000175000017500000001372612054410451013665 00000000000000/* * Copyright (c) 2003, 2007-11 Matteo Frigo * Copyright (c) 2003, 2007-11 Massachusetts Institute of Technology * * This program is free software; you can redistribute it and/or modify * it under the terms of the GNU General Public License as published by * the Free Software Foundation; either version 2 of the License, or * (at your option) any later version. * * This program is distributed in the hope that it will be useful, * but WITHOUT ANY WARRANTY; without even the implied warranty of * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the * GNU General Public License for more details. * * You should have received a copy of the GNU General Public License * along with this program; if not, write to the Free Software * Foundation, Inc., 51 Franklin Street, Fifth Floor, Boston, MA 02110-1301 USA * */ /* This file was automatically generated --- DO NOT EDIT */ /* Generated on Sun Nov 25 07:38:01 EST 2012 */ #include "codelet-dft.h" #ifdef HAVE_FMA /* Generated by: ../../../genfft/gen_twiddle_c.native -fma -reorder-insns -schedule-for-pipeline -simd -compact -variables 4 -pipeline-latency 8 -n 5 -name t1fv_5 -include t1f.h */ /* * This function contains 20 FP additions, 19 FP multiplications, * (or, 11 additions, 10 multiplications, 9 fused multiply/add), * 26 stack variables, 4 constants, and 10 memory accesses */ #include "t1f.h" static void t1fv_5(R *ri, R *ii, const R *W, stride rs, INT mb, INT me, INT ms) { DVK(KP559016994, +0.559016994374947424102293417182819058860154590); DVK(KP250000000, +0.250000000000000000000000000000000000000000000); DVK(KP618033988, +0.618033988749894848204586834365638117720309180); DVK(KP951056516, +0.951056516295153572116439333379382143405698634); { INT m; R *x; x = ri; for (m = mb, W = W + (mb * ((TWVL / VL) * 8)); m < me; m = m + VL, x = x + (VL * ms), W = W + (TWVL * 8), MAKE_VOLATILE_STRIDE(5, rs)) { V T1, T2, T9, T4, T7; T1 = LD(&(x[0]), ms, &(x[0])); T2 = LD(&(x[WS(rs, 1)]), ms, &(x[WS(rs, 1)])); T9 = LD(&(x[WS(rs, 3)]), ms, &(x[WS(rs, 1)])); T4 = LD(&(x[WS(rs, 4)]), ms, &(x[0])); T7 = LD(&(x[WS(rs, 2)]), ms, &(x[0])); { V T3, Ta, T5, T8; T3 = BYTWJ(&(W[0]), T2); Ta = BYTWJ(&(W[TWVL * 4]), T9); T5 = BYTWJ(&(W[TWVL * 6]), T4); T8 = BYTWJ(&(W[TWVL * 2]), T7); { V T6, Tg, Tb, Th; T6 = VADD(T3, T5); Tg = VSUB(T3, T5); Tb = VADD(T8, Ta); Th = VSUB(T8, Ta); { V Te, Tc, Tk, Ti, Td, Tj, Tf; Te = VSUB(T6, Tb); Tc = VADD(T6, Tb); Tk = VMUL(LDK(KP951056516), VFNMS(LDK(KP618033988), Tg, Th)); Ti = VMUL(LDK(KP951056516), VFMA(LDK(KP618033988), Th, Tg)); Td = VFNMS(LDK(KP250000000), Tc, T1); ST(&(x[0]), VADD(T1, Tc), ms, &(x[0])); Tj = VFNMS(LDK(KP559016994), Te, Td); Tf = VFMA(LDK(KP559016994), Te, Td); ST(&(x[WS(rs, 2)]), VFMAI(Tk, Tj), ms, &(x[0])); ST(&(x[WS(rs, 3)]), VFNMSI(Tk, Tj), ms, &(x[WS(rs, 1)])); ST(&(x[WS(rs, 4)]), VFMAI(Ti, Tf), ms, &(x[0])); ST(&(x[WS(rs, 1)]), VFNMSI(Ti, Tf), ms, &(x[WS(rs, 1)])); } } } } } VLEAVE(); } static const tw_instr twinstr[] = { VTW(0, 1), VTW(0, 2), VTW(0, 3), VTW(0, 4), {TW_NEXT, VL, 0} }; static const ct_desc desc = { 5, XSIMD_STRING("t1fv_5"), twinstr, &GENUS, {11, 10, 9, 0}, 0, 0, 0 }; void XSIMD(codelet_t1fv_5) (planner *p) { X(kdft_dit_register) (p, t1fv_5, &desc); } #else /* HAVE_FMA */ /* Generated by: ../../../genfft/gen_twiddle_c.native -simd -compact -variables 4 -pipeline-latency 8 -n 5 -name t1fv_5 -include t1f.h */ /* * This function contains 20 FP additions, 14 FP multiplications, * (or, 17 additions, 11 multiplications, 3 fused multiply/add), * 20 stack variables, 4 constants, and 10 memory accesses */ #include "t1f.h" static void t1fv_5(R *ri, R *ii, const R *W, stride rs, INT mb, INT me, INT ms) { DVK(KP250000000, +0.250000000000000000000000000000000000000000000); DVK(KP559016994, +0.559016994374947424102293417182819058860154590); DVK(KP587785252, +0.587785252292473129168705954639072768597652438); DVK(KP951056516, +0.951056516295153572116439333379382143405698634); { INT m; R *x; x = ri; for (m = mb, W = W + (mb * ((TWVL / VL) * 8)); m < me; m = m + VL, x = x + (VL * ms), W = W + (TWVL * 8), MAKE_VOLATILE_STRIDE(5, rs)) { V Tc, Tg, Th, T5, Ta, Td; Tc = LD(&(x[0]), ms, &(x[0])); { V T2, T9, T4, T7; { V T1, T8, T3, T6; T1 = LD(&(x[WS(rs, 1)]), ms, &(x[WS(rs, 1)])); T2 = BYTWJ(&(W[0]), T1); T8 = LD(&(x[WS(rs, 3)]), ms, &(x[WS(rs, 1)])); T9 = BYTWJ(&(W[TWVL * 4]), T8); T3 = LD(&(x[WS(rs, 4)]), ms, &(x[0])); T4 = BYTWJ(&(W[TWVL * 6]), T3); T6 = LD(&(x[WS(rs, 2)]), ms, &(x[0])); T7 = BYTWJ(&(W[TWVL * 2]), T6); } Tg = VSUB(T2, T4); Th = VSUB(T7, T9); T5 = VADD(T2, T4); Ta = VADD(T7, T9); Td = VADD(T5, Ta); } ST(&(x[0]), VADD(Tc, Td), ms, &(x[0])); { V Ti, Tj, Tf, Tk, Tb, Te; Ti = VBYI(VFMA(LDK(KP951056516), Tg, VMUL(LDK(KP587785252), Th))); Tj = VBYI(VFNMS(LDK(KP587785252), Tg, VMUL(LDK(KP951056516), Th))); Tb = VMUL(LDK(KP559016994), VSUB(T5, Ta)); Te = VFNMS(LDK(KP250000000), Td, Tc); Tf = VADD(Tb, Te); Tk = VSUB(Te, Tb); ST(&(x[WS(rs, 1)]), VSUB(Tf, Ti), ms, &(x[WS(rs, 1)])); ST(&(x[WS(rs, 3)]), VSUB(Tk, Tj), ms, &(x[WS(rs, 1)])); ST(&(x[WS(rs, 4)]), VADD(Ti, Tf), ms, &(x[0])); ST(&(x[WS(rs, 2)]), VADD(Tj, Tk), ms, &(x[0])); } } } VLEAVE(); } static const tw_instr twinstr[] = { VTW(0, 1), VTW(0, 2), VTW(0, 3), VTW(0, 4), {TW_NEXT, VL, 0} }; static const ct_desc desc = { 5, XSIMD_STRING("t1fv_5"), twinstr, &GENUS, {17, 11, 3, 0}, 0, 0, 0 }; void XSIMD(codelet_t1fv_5) (planner *p) { X(kdft_dit_register) (p, t1fv_5, &desc); } #endif /* HAVE_FMA */ fftw-3.3.3/dft/simd/common/t1fv_6.c0000644000175000017500000001401312054410451013654 00000000000000/* * Copyright (c) 2003, 2007-11 Matteo Frigo * Copyright (c) 2003, 2007-11 Massachusetts Institute of Technology * * This program is free software; you can redistribute it and/or modify * it under the terms of the GNU General Public License as published by * the Free Software Foundation; either version 2 of the License, or * (at your option) any later version. * * This program is distributed in the hope that it will be useful, * but WITHOUT ANY WARRANTY; without even the implied warranty of * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the * GNU General Public License for more details. * * You should have received a copy of the GNU General Public License * along with this program; if not, write to the Free Software * Foundation, Inc., 51 Franklin Street, Fifth Floor, Boston, MA 02110-1301 USA * */ /* This file was automatically generated --- DO NOT EDIT */ /* Generated on Sun Nov 25 07:38:01 EST 2012 */ #include "codelet-dft.h" #ifdef HAVE_FMA /* Generated by: ../../../genfft/gen_twiddle_c.native -fma -reorder-insns -schedule-for-pipeline -simd -compact -variables 4 -pipeline-latency 8 -n 6 -name t1fv_6 -include t1f.h */ /* * This function contains 23 FP additions, 18 FP multiplications, * (or, 17 additions, 12 multiplications, 6 fused multiply/add), * 27 stack variables, 2 constants, and 12 memory accesses */ #include "t1f.h" static void t1fv_6(R *ri, R *ii, const R *W, stride rs, INT mb, INT me, INT ms) { DVK(KP500000000, +0.500000000000000000000000000000000000000000000); DVK(KP866025403, +0.866025403784438646763723170752936183471402627); { INT m; R *x; x = ri; for (m = mb, W = W + (mb * ((TWVL / VL) * 10)); m < me; m = m + VL, x = x + (VL * ms), W = W + (TWVL * 10), MAKE_VOLATILE_STRIDE(6, rs)) { V T1, T2, Ta, Tc, T5, T7; T1 = LD(&(x[0]), ms, &(x[0])); T2 = LD(&(x[WS(rs, 3)]), ms, &(x[WS(rs, 1)])); Ta = LD(&(x[WS(rs, 4)]), ms, &(x[0])); Tc = LD(&(x[WS(rs, 1)]), ms, &(x[WS(rs, 1)])); T5 = LD(&(x[WS(rs, 2)]), ms, &(x[0])); T7 = LD(&(x[WS(rs, 5)]), ms, &(x[WS(rs, 1)])); { V T3, Tb, Td, T6, T8; T3 = BYTWJ(&(W[TWVL * 4]), T2); Tb = BYTWJ(&(W[TWVL * 6]), Ta); Td = BYTWJ(&(W[0]), Tc); T6 = BYTWJ(&(W[TWVL * 2]), T5); T8 = BYTWJ(&(W[TWVL * 8]), T7); { V Ti, T4, Tk, Te, Tj, T9; Ti = VADD(T1, T3); T4 = VSUB(T1, T3); Tk = VADD(Tb, Td); Te = VSUB(Tb, Td); Tj = VADD(T6, T8); T9 = VSUB(T6, T8); { V Tl, Tn, Tf, Th, Tm, Tg; Tl = VADD(Tj, Tk); Tn = VMUL(LDK(KP866025403), VSUB(Tk, Tj)); Tf = VADD(T9, Te); Th = VMUL(LDK(KP866025403), VSUB(Te, T9)); ST(&(x[0]), VADD(Ti, Tl), ms, &(x[0])); Tm = VFNMS(LDK(KP500000000), Tl, Ti); ST(&(x[WS(rs, 3)]), VADD(T4, Tf), ms, &(x[WS(rs, 1)])); Tg = VFNMS(LDK(KP500000000), Tf, T4); ST(&(x[WS(rs, 2)]), VFNMSI(Tn, Tm), ms, &(x[0])); ST(&(x[WS(rs, 4)]), VFMAI(Tn, Tm), ms, &(x[0])); ST(&(x[WS(rs, 5)]), VFNMSI(Th, Tg), ms, &(x[WS(rs, 1)])); ST(&(x[WS(rs, 1)]), VFMAI(Th, Tg), ms, &(x[WS(rs, 1)])); } } } } } VLEAVE(); } static const tw_instr twinstr[] = { VTW(0, 1), VTW(0, 2), VTW(0, 3), VTW(0, 4), VTW(0, 5), {TW_NEXT, VL, 0} }; static const ct_desc desc = { 6, XSIMD_STRING("t1fv_6"), twinstr, &GENUS, {17, 12, 6, 0}, 0, 0, 0 }; void XSIMD(codelet_t1fv_6) (planner *p) { X(kdft_dit_register) (p, t1fv_6, &desc); } #else /* HAVE_FMA */ /* Generated by: ../../../genfft/gen_twiddle_c.native -simd -compact -variables 4 -pipeline-latency 8 -n 6 -name t1fv_6 -include t1f.h */ /* * This function contains 23 FP additions, 14 FP multiplications, * (or, 21 additions, 12 multiplications, 2 fused multiply/add), * 19 stack variables, 2 constants, and 12 memory accesses */ #include "t1f.h" static void t1fv_6(R *ri, R *ii, const R *W, stride rs, INT mb, INT me, INT ms) { DVK(KP500000000, +0.500000000000000000000000000000000000000000000); DVK(KP866025403, +0.866025403784438646763723170752936183471402627); { INT m; R *x; x = ri; for (m = mb, W = W + (mb * ((TWVL / VL) * 10)); m < me; m = m + VL, x = x + (VL * ms), W = W + (TWVL * 10), MAKE_VOLATILE_STRIDE(6, rs)) { V T4, Ti, Te, Tk, T9, Tj, T1, T3, T2; T1 = LD(&(x[0]), ms, &(x[0])); T2 = LD(&(x[WS(rs, 3)]), ms, &(x[WS(rs, 1)])); T3 = BYTWJ(&(W[TWVL * 4]), T2); T4 = VSUB(T1, T3); Ti = VADD(T1, T3); { V Tb, Td, Ta, Tc; Ta = LD(&(x[WS(rs, 4)]), ms, &(x[0])); Tb = BYTWJ(&(W[TWVL * 6]), Ta); Tc = LD(&(x[WS(rs, 1)]), ms, &(x[WS(rs, 1)])); Td = BYTWJ(&(W[0]), Tc); Te = VSUB(Tb, Td); Tk = VADD(Tb, Td); } { V T6, T8, T5, T7; T5 = LD(&(x[WS(rs, 2)]), ms, &(x[0])); T6 = BYTWJ(&(W[TWVL * 2]), T5); T7 = LD(&(x[WS(rs, 5)]), ms, &(x[WS(rs, 1)])); T8 = BYTWJ(&(W[TWVL * 8]), T7); T9 = VSUB(T6, T8); Tj = VADD(T6, T8); } { V Th, Tf, Tg, Tn, Tl, Tm; Th = VBYI(VMUL(LDK(KP866025403), VSUB(Te, T9))); Tf = VADD(T9, Te); Tg = VFNMS(LDK(KP500000000), Tf, T4); ST(&(x[WS(rs, 3)]), VADD(T4, Tf), ms, &(x[WS(rs, 1)])); ST(&(x[WS(rs, 1)]), VADD(Tg, Th), ms, &(x[WS(rs, 1)])); ST(&(x[WS(rs, 5)]), VSUB(Tg, Th), ms, &(x[WS(rs, 1)])); Tn = VBYI(VMUL(LDK(KP866025403), VSUB(Tk, Tj))); Tl = VADD(Tj, Tk); Tm = VFNMS(LDK(KP500000000), Tl, Ti); ST(&(x[0]), VADD(Ti, Tl), ms, &(x[0])); ST(&(x[WS(rs, 4)]), VADD(Tm, Tn), ms, &(x[0])); ST(&(x[WS(rs, 2)]), VSUB(Tm, Tn), ms, &(x[0])); } } } VLEAVE(); } static const tw_instr twinstr[] = { VTW(0, 1), VTW(0, 2), VTW(0, 3), VTW(0, 4), VTW(0, 5), {TW_NEXT, VL, 0} }; static const ct_desc desc = { 6, XSIMD_STRING("t1fv_6"), twinstr, &GENUS, {21, 12, 2, 0}, 0, 0, 0 }; void XSIMD(codelet_t1fv_6) (planner *p) { X(kdft_dit_register) (p, t1fv_6, &desc); } #endif /* HAVE_FMA */ fftw-3.3.3/dft/simd/common/t1fv_7.c0000644000175000017500000001760312054410452013666 00000000000000/* * Copyright (c) 2003, 2007-11 Matteo Frigo * Copyright (c) 2003, 2007-11 Massachusetts Institute of Technology * * This program is free software; you can redistribute it and/or modify * it under the terms of the GNU General Public License as published by * the Free Software Foundation; either version 2 of the License, or * (at your option) any later version. * * This program is distributed in the hope that it will be useful, * but WITHOUT ANY WARRANTY; without even the implied warranty of * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the * GNU General Public License for more details. * * You should have received a copy of the GNU General Public License * along with this program; if not, write to the Free Software * Foundation, Inc., 51 Franklin Street, Fifth Floor, Boston, MA 02110-1301 USA * */ /* This file was automatically generated --- DO NOT EDIT */ /* Generated on Sun Nov 25 07:38:01 EST 2012 */ #include "codelet-dft.h" #ifdef HAVE_FMA /* Generated by: ../../../genfft/gen_twiddle_c.native -fma -reorder-insns -schedule-for-pipeline -simd -compact -variables 4 -pipeline-latency 8 -n 7 -name t1fv_7 -include t1f.h */ /* * This function contains 36 FP additions, 36 FP multiplications, * (or, 15 additions, 15 multiplications, 21 fused multiply/add), * 42 stack variables, 6 constants, and 14 memory accesses */ #include "t1f.h" static void t1fv_7(R *ri, R *ii, const R *W, stride rs, INT mb, INT me, INT ms) { DVK(KP900968867, +0.900968867902419126236102319507445051165919162); DVK(KP801937735, +0.801937735804838252472204639014890102331838324); DVK(KP974927912, +0.974927912181823607018131682993931217232785801); DVK(KP692021471, +0.692021471630095869627814897002069140197260599); DVK(KP554958132, +0.554958132087371191422194871006410481067288862); DVK(KP356895867, +0.356895867892209443894399510021300583399127187); { INT m; R *x; x = ri; for (m = mb, W = W + (mb * ((TWVL / VL) * 12)); m < me; m = m + VL, x = x + (VL * ms), W = W + (TWVL * 12), MAKE_VOLATILE_STRIDE(7, rs)) { V T1, T2, T4, Te, Tc, T9, T7; T1 = LD(&(x[0]), ms, &(x[0])); T2 = LD(&(x[WS(rs, 1)]), ms, &(x[WS(rs, 1)])); T4 = LD(&(x[WS(rs, 6)]), ms, &(x[0])); Te = LD(&(x[WS(rs, 4)]), ms, &(x[0])); Tc = LD(&(x[WS(rs, 3)]), ms, &(x[WS(rs, 1)])); T9 = LD(&(x[WS(rs, 5)]), ms, &(x[WS(rs, 1)])); T7 = LD(&(x[WS(rs, 2)]), ms, &(x[0])); { V T3, T5, Tf, Td, Ta, T8; T3 = BYTWJ(&(W[0]), T2); T5 = BYTWJ(&(W[TWVL * 10]), T4); Tf = BYTWJ(&(W[TWVL * 6]), Te); Td = BYTWJ(&(W[TWVL * 4]), Tc); Ta = BYTWJ(&(W[TWVL * 8]), T9); T8 = BYTWJ(&(W[TWVL * 2]), T7); { V T6, Tk, Tg, Tl, Tb, Tm; T6 = VADD(T3, T5); Tk = VSUB(T5, T3); Tg = VADD(Td, Tf); Tl = VSUB(Tf, Td); Tb = VADD(T8, Ta); Tm = VSUB(Ta, T8); { V Th, Ts, Tp, Tu, Tn, Tx, Ti, Tt; Th = VFNMS(LDK(KP356895867), T6, Tg); Ts = VFMA(LDK(KP554958132), Tl, Tk); ST(&(x[0]), VADD(T1, VADD(T6, VADD(Tb, Tg))), ms, &(x[0])); Tp = VFNMS(LDK(KP356895867), Tb, T6); Tu = VFNMS(LDK(KP356895867), Tg, Tb); Tn = VFMA(LDK(KP554958132), Tm, Tl); Tx = VFNMS(LDK(KP554958132), Tk, Tm); Ti = VFNMS(LDK(KP692021471), Th, Tb); Tt = VMUL(LDK(KP974927912), VFMA(LDK(KP801937735), Ts, Tm)); { V Tq, Tv, To, Ty, Tj, Tr, Tw; Tq = VFNMS(LDK(KP692021471), Tp, Tg); Tv = VFNMS(LDK(KP692021471), Tu, T6); To = VMUL(LDK(KP974927912), VFNMS(LDK(KP801937735), Tn, Tk)); Ty = VMUL(LDK(KP974927912), VFNMS(LDK(KP801937735), Tx, Tl)); Tj = VFNMS(LDK(KP900968867), Ti, T1); Tr = VFNMS(LDK(KP900968867), Tq, T1); Tw = VFNMS(LDK(KP900968867), Tv, T1); ST(&(x[WS(rs, 2)]), VFMAI(To, Tj), ms, &(x[0])); ST(&(x[WS(rs, 5)]), VFNMSI(To, Tj), ms, &(x[WS(rs, 1)])); ST(&(x[WS(rs, 1)]), VFMAI(Tt, Tr), ms, &(x[WS(rs, 1)])); ST(&(x[WS(rs, 6)]), VFNMSI(Tt, Tr), ms, &(x[0])); ST(&(x[WS(rs, 3)]), VFMAI(Ty, Tw), ms, &(x[WS(rs, 1)])); ST(&(x[WS(rs, 4)]), VFNMSI(Ty, Tw), ms, &(x[0])); } } } } } } VLEAVE(); } static const tw_instr twinstr[] = { VTW(0, 1), VTW(0, 2), VTW(0, 3), VTW(0, 4), VTW(0, 5), VTW(0, 6), {TW_NEXT, VL, 0} }; static const ct_desc desc = { 7, XSIMD_STRING("t1fv_7"), twinstr, &GENUS, {15, 15, 21, 0}, 0, 0, 0 }; void XSIMD(codelet_t1fv_7) (planner *p) { X(kdft_dit_register) (p, t1fv_7, &desc); } #else /* HAVE_FMA */ /* Generated by: ../../../genfft/gen_twiddle_c.native -simd -compact -variables 4 -pipeline-latency 8 -n 7 -name t1fv_7 -include t1f.h */ /* * This function contains 36 FP additions, 30 FP multiplications, * (or, 24 additions, 18 multiplications, 12 fused multiply/add), * 21 stack variables, 6 constants, and 14 memory accesses */ #include "t1f.h" static void t1fv_7(R *ri, R *ii, const R *W, stride rs, INT mb, INT me, INT ms) { DVK(KP900968867, +0.900968867902419126236102319507445051165919162); DVK(KP222520933, +0.222520933956314404288902564496794759466355569); DVK(KP623489801, +0.623489801858733530525004884004239810632274731); DVK(KP781831482, +0.781831482468029808708444526674057750232334519); DVK(KP974927912, +0.974927912181823607018131682993931217232785801); DVK(KP433883739, +0.433883739117558120475768332848358754609990728); { INT m; R *x; x = ri; for (m = mb, W = W + (mb * ((TWVL / VL) * 12)); m < me; m = m + VL, x = x + (VL * ms), W = W + (TWVL * 12), MAKE_VOLATILE_STRIDE(7, rs)) { V T1, Tg, Tj, T6, Ti, Tb, Tk, Tp, To; T1 = LD(&(x[0]), ms, &(x[0])); { V Td, Tf, Tc, Te; Tc = LD(&(x[WS(rs, 3)]), ms, &(x[WS(rs, 1)])); Td = BYTWJ(&(W[TWVL * 4]), Tc); Te = LD(&(x[WS(rs, 4)]), ms, &(x[0])); Tf = BYTWJ(&(W[TWVL * 6]), Te); Tg = VADD(Td, Tf); Tj = VSUB(Tf, Td); } { V T3, T5, T2, T4; T2 = LD(&(x[WS(rs, 1)]), ms, &(x[WS(rs, 1)])); T3 = BYTWJ(&(W[0]), T2); T4 = LD(&(x[WS(rs, 6)]), ms, &(x[0])); T5 = BYTWJ(&(W[TWVL * 10]), T4); T6 = VADD(T3, T5); Ti = VSUB(T5, T3); } { V T8, Ta, T7, T9; T7 = LD(&(x[WS(rs, 2)]), ms, &(x[0])); T8 = BYTWJ(&(W[TWVL * 2]), T7); T9 = LD(&(x[WS(rs, 5)]), ms, &(x[WS(rs, 1)])); Ta = BYTWJ(&(W[TWVL * 8]), T9); Tb = VADD(T8, Ta); Tk = VSUB(Ta, T8); } ST(&(x[0]), VADD(T1, VADD(T6, VADD(Tb, Tg))), ms, &(x[0])); Tp = VBYI(VFMA(LDK(KP433883739), Ti, VFNMS(LDK(KP781831482), Tk, VMUL(LDK(KP974927912), Tj)))); To = VFMA(LDK(KP623489801), Tb, VFNMS(LDK(KP222520933), Tg, VFNMS(LDK(KP900968867), T6, T1))); ST(&(x[WS(rs, 4)]), VSUB(To, Tp), ms, &(x[0])); ST(&(x[WS(rs, 3)]), VADD(To, Tp), ms, &(x[WS(rs, 1)])); { V Tl, Th, Tn, Tm; Tl = VBYI(VFNMS(LDK(KP781831482), Tj, VFNMS(LDK(KP433883739), Tk, VMUL(LDK(KP974927912), Ti)))); Th = VFMA(LDK(KP623489801), Tg, VFNMS(LDK(KP900968867), Tb, VFNMS(LDK(KP222520933), T6, T1))); ST(&(x[WS(rs, 5)]), VSUB(Th, Tl), ms, &(x[WS(rs, 1)])); ST(&(x[WS(rs, 2)]), VADD(Th, Tl), ms, &(x[0])); Tn = VBYI(VFMA(LDK(KP781831482), Ti, VFMA(LDK(KP974927912), Tk, VMUL(LDK(KP433883739), Tj)))); Tm = VFMA(LDK(KP623489801), T6, VFNMS(LDK(KP900968867), Tg, VFNMS(LDK(KP222520933), Tb, T1))); ST(&(x[WS(rs, 6)]), VSUB(Tm, Tn), ms, &(x[0])); ST(&(x[WS(rs, 1)]), VADD(Tm, Tn), ms, &(x[WS(rs, 1)])); } } } VLEAVE(); } static const tw_instr twinstr[] = { VTW(0, 1), VTW(0, 2), VTW(0, 3), VTW(0, 4), VTW(0, 5), VTW(0, 6), {TW_NEXT, VL, 0} }; static const ct_desc desc = { 7, XSIMD_STRING("t1fv_7"), twinstr, &GENUS, {24, 18, 12, 0}, 0, 0, 0 }; void XSIMD(codelet_t1fv_7) (planner *p) { X(kdft_dit_register) (p, t1fv_7, &desc); } #endif /* HAVE_FMA */ fftw-3.3.3/dft/simd/common/t1fv_8.c0000644000175000017500000001560312054410452013665 00000000000000/* * Copyright (c) 2003, 2007-11 Matteo Frigo * Copyright (c) 2003, 2007-11 Massachusetts Institute of Technology * * This program is free software; you can redistribute it and/or modify * it under the terms of the GNU General Public License as published by * the Free Software Foundation; either version 2 of the License, or * (at your option) any later version. * * This program is distributed in the hope that it will be useful, * but WITHOUT ANY WARRANTY; without even the implied warranty of * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the * GNU General Public License for more details. * * You should have received a copy of the GNU General Public License * along with this program; if not, write to the Free Software * Foundation, Inc., 51 Franklin Street, Fifth Floor, Boston, MA 02110-1301 USA * */ /* This file was automatically generated --- DO NOT EDIT */ /* Generated on Sun Nov 25 07:38:02 EST 2012 */ #include "codelet-dft.h" #ifdef HAVE_FMA /* Generated by: ../../../genfft/gen_twiddle_c.native -fma -reorder-insns -schedule-for-pipeline -simd -compact -variables 4 -pipeline-latency 8 -n 8 -name t1fv_8 -include t1f.h */ /* * This function contains 33 FP additions, 24 FP multiplications, * (or, 23 additions, 14 multiplications, 10 fused multiply/add), * 36 stack variables, 1 constants, and 16 memory accesses */ #include "t1f.h" static void t1fv_8(R *ri, R *ii, const R *W, stride rs, INT mb, INT me, INT ms) { DVK(KP707106781, +0.707106781186547524400844362104849039284835938); { INT m; R *x; x = ri; for (m = mb, W = W + (mb * ((TWVL / VL) * 14)); m < me; m = m + VL, x = x + (VL * ms), W = W + (TWVL * 14), MAKE_VOLATILE_STRIDE(8, rs)) { V T1, T2, Th, Tj, T5, T7, Ta, Tc; T1 = LD(&(x[0]), ms, &(x[0])); T2 = LD(&(x[WS(rs, 4)]), ms, &(x[0])); Th = LD(&(x[WS(rs, 2)]), ms, &(x[0])); Tj = LD(&(x[WS(rs, 6)]), ms, &(x[0])); T5 = LD(&(x[WS(rs, 1)]), ms, &(x[WS(rs, 1)])); T7 = LD(&(x[WS(rs, 5)]), ms, &(x[WS(rs, 1)])); Ta = LD(&(x[WS(rs, 7)]), ms, &(x[WS(rs, 1)])); Tc = LD(&(x[WS(rs, 3)]), ms, &(x[WS(rs, 1)])); { V T3, Ti, Tk, T6, T8, Tb, Td; T3 = BYTWJ(&(W[TWVL * 6]), T2); Ti = BYTWJ(&(W[TWVL * 2]), Th); Tk = BYTWJ(&(W[TWVL * 10]), Tj); T6 = BYTWJ(&(W[0]), T5); T8 = BYTWJ(&(W[TWVL * 8]), T7); Tb = BYTWJ(&(W[TWVL * 12]), Ta); Td = BYTWJ(&(W[TWVL * 4]), Tc); { V Tq, T4, Tr, Tl, Tt, T9, Tu, Te, Tw, Ts; Tq = VADD(T1, T3); T4 = VSUB(T1, T3); Tr = VADD(Ti, Tk); Tl = VSUB(Ti, Tk); Tt = VADD(T6, T8); T9 = VSUB(T6, T8); Tu = VADD(Tb, Td); Te = VSUB(Tb, Td); Tw = VSUB(Tq, Tr); Ts = VADD(Tq, Tr); { V Tx, Tv, Tm, Tf; Tx = VSUB(Tu, Tt); Tv = VADD(Tt, Tu); Tm = VSUB(Te, T9); Tf = VADD(T9, Te); { V Tp, Tn, To, Tg; ST(&(x[WS(rs, 2)]), VFMAI(Tx, Tw), ms, &(x[0])); ST(&(x[WS(rs, 6)]), VFNMSI(Tx, Tw), ms, &(x[0])); ST(&(x[0]), VADD(Ts, Tv), ms, &(x[0])); ST(&(x[WS(rs, 4)]), VSUB(Ts, Tv), ms, &(x[0])); Tp = VFMA(LDK(KP707106781), Tm, Tl); Tn = VFNMS(LDK(KP707106781), Tm, Tl); To = VFNMS(LDK(KP707106781), Tf, T4); Tg = VFMA(LDK(KP707106781), Tf, T4); ST(&(x[WS(rs, 5)]), VFNMSI(Tp, To), ms, &(x[WS(rs, 1)])); ST(&(x[WS(rs, 3)]), VFMAI(Tp, To), ms, &(x[WS(rs, 1)])); ST(&(x[WS(rs, 7)]), VFMAI(Tn, Tg), ms, &(x[WS(rs, 1)])); ST(&(x[WS(rs, 1)]), VFNMSI(Tn, Tg), ms, &(x[WS(rs, 1)])); } } } } } } VLEAVE(); } static const tw_instr twinstr[] = { VTW(0, 1), VTW(0, 2), VTW(0, 3), VTW(0, 4), VTW(0, 5), VTW(0, 6), VTW(0, 7), {TW_NEXT, VL, 0} }; static const ct_desc desc = { 8, XSIMD_STRING("t1fv_8"), twinstr, &GENUS, {23, 14, 10, 0}, 0, 0, 0 }; void XSIMD(codelet_t1fv_8) (planner *p) { X(kdft_dit_register) (p, t1fv_8, &desc); } #else /* HAVE_FMA */ /* Generated by: ../../../genfft/gen_twiddle_c.native -simd -compact -variables 4 -pipeline-latency 8 -n 8 -name t1fv_8 -include t1f.h */ /* * This function contains 33 FP additions, 16 FP multiplications, * (or, 33 additions, 16 multiplications, 0 fused multiply/add), * 24 stack variables, 1 constants, and 16 memory accesses */ #include "t1f.h" static void t1fv_8(R *ri, R *ii, const R *W, stride rs, INT mb, INT me, INT ms) { DVK(KP707106781, +0.707106781186547524400844362104849039284835938); { INT m; R *x; x = ri; for (m = mb, W = W + (mb * ((TWVL / VL) * 14)); m < me; m = m + VL, x = x + (VL * ms), W = W + (TWVL * 14), MAKE_VOLATILE_STRIDE(8, rs)) { V T4, Tq, Tm, Tr, T9, Tt, Te, Tu, T1, T3, T2; T1 = LD(&(x[0]), ms, &(x[0])); T2 = LD(&(x[WS(rs, 4)]), ms, &(x[0])); T3 = BYTWJ(&(W[TWVL * 6]), T2); T4 = VSUB(T1, T3); Tq = VADD(T1, T3); { V Tj, Tl, Ti, Tk; Ti = LD(&(x[WS(rs, 2)]), ms, &(x[0])); Tj = BYTWJ(&(W[TWVL * 2]), Ti); Tk = LD(&(x[WS(rs, 6)]), ms, &(x[0])); Tl = BYTWJ(&(W[TWVL * 10]), Tk); Tm = VSUB(Tj, Tl); Tr = VADD(Tj, Tl); } { V T6, T8, T5, T7; T5 = LD(&(x[WS(rs, 1)]), ms, &(x[WS(rs, 1)])); T6 = BYTWJ(&(W[0]), T5); T7 = LD(&(x[WS(rs, 5)]), ms, &(x[WS(rs, 1)])); T8 = BYTWJ(&(W[TWVL * 8]), T7); T9 = VSUB(T6, T8); Tt = VADD(T6, T8); } { V Tb, Td, Ta, Tc; Ta = LD(&(x[WS(rs, 7)]), ms, &(x[WS(rs, 1)])); Tb = BYTWJ(&(W[TWVL * 12]), Ta); Tc = LD(&(x[WS(rs, 3)]), ms, &(x[WS(rs, 1)])); Td = BYTWJ(&(W[TWVL * 4]), Tc); Te = VSUB(Tb, Td); Tu = VADD(Tb, Td); } { V Ts, Tv, Tw, Tx; Ts = VADD(Tq, Tr); Tv = VADD(Tt, Tu); ST(&(x[WS(rs, 4)]), VSUB(Ts, Tv), ms, &(x[0])); ST(&(x[0]), VADD(Ts, Tv), ms, &(x[0])); Tw = VSUB(Tq, Tr); Tx = VBYI(VSUB(Tu, Tt)); ST(&(x[WS(rs, 6)]), VSUB(Tw, Tx), ms, &(x[0])); ST(&(x[WS(rs, 2)]), VADD(Tw, Tx), ms, &(x[0])); { V Tg, To, Tn, Tp, Tf, Th; Tf = VMUL(LDK(KP707106781), VADD(T9, Te)); Tg = VADD(T4, Tf); To = VSUB(T4, Tf); Th = VMUL(LDK(KP707106781), VSUB(Te, T9)); Tn = VBYI(VSUB(Th, Tm)); Tp = VBYI(VADD(Tm, Th)); ST(&(x[WS(rs, 7)]), VSUB(Tg, Tn), ms, &(x[WS(rs, 1)])); ST(&(x[WS(rs, 3)]), VADD(To, Tp), ms, &(x[WS(rs, 1)])); ST(&(x[WS(rs, 1)]), VADD(Tg, Tn), ms, &(x[WS(rs, 1)])); ST(&(x[WS(rs, 5)]), VSUB(To, Tp), ms, &(x[WS(rs, 1)])); } } } } VLEAVE(); } static const tw_instr twinstr[] = { VTW(0, 1), VTW(0, 2), VTW(0, 3), VTW(0, 4), VTW(0, 5), VTW(0, 6), VTW(0, 7), {TW_NEXT, VL, 0} }; static const ct_desc desc = { 8, XSIMD_STRING("t1fv_8"), twinstr, &GENUS, {33, 16, 0, 0}, 0, 0, 0 }; void XSIMD(codelet_t1fv_8) (planner *p) { X(kdft_dit_register) (p, t1fv_8, &desc); } #endif /* HAVE_FMA */ fftw-3.3.3/dft/simd/common/t1fv_9.c0000644000175000017500000002653712054410453013677 00000000000000/* * Copyright (c) 2003, 2007-11 Matteo Frigo * Copyright (c) 2003, 2007-11 Massachusetts Institute of Technology * * This program is free software; you can redistribute it and/or modify * it under the terms of the GNU General Public License as published by * the Free Software Foundation; either version 2 of the License, or * (at your option) any later version. * * This program is distributed in the hope that it will be useful, * but WITHOUT ANY WARRANTY; without even the implied warranty of * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the * GNU General Public License for more details. * * You should have received a copy of the GNU General Public License * along with this program; if not, write to the Free Software * Foundation, Inc., 51 Franklin Street, Fifth Floor, Boston, MA 02110-1301 USA * */ /* This file was automatically generated --- DO NOT EDIT */ /* Generated on Sun Nov 25 07:38:02 EST 2012 */ #include "codelet-dft.h" #ifdef HAVE_FMA /* Generated by: ../../../genfft/gen_twiddle_c.native -fma -reorder-insns -schedule-for-pipeline -simd -compact -variables 4 -pipeline-latency 8 -n 9 -name t1fv_9 -include t1f.h */ /* * This function contains 54 FP additions, 54 FP multiplications, * (or, 20 additions, 20 multiplications, 34 fused multiply/add), * 67 stack variables, 19 constants, and 18 memory accesses */ #include "t1f.h" static void t1fv_9(R *ri, R *ii, const R *W, stride rs, INT mb, INT me, INT ms) { DVK(KP939692620, +0.939692620785908384054109277324731469936208134); DVK(KP826351822, +0.826351822333069651148283373230685203999624323); DVK(KP879385241, +0.879385241571816768108218554649462939872416269); DVK(KP984807753, +0.984807753012208059366743024589523013670643252); DVK(KP666666666, +0.666666666666666666666666666666666666666666667); DVK(KP852868531, +0.852868531952443209628250963940074071936020296); DVK(KP907603734, +0.907603734547952313649323976213898122064543220); DVK(KP420276625, +0.420276625461206169731530603237061658838781920); DVK(KP673648177, +0.673648177666930348851716626769314796000375677); DVK(KP898197570, +0.898197570222573798468955502359086394667167570); DVK(KP347296355, +0.347296355333860697703433253538629592000751354); DVK(KP866025403, +0.866025403784438646763723170752936183471402627); DVK(KP439692620, +0.439692620785908384054109277324731469936208134); DVK(KP203604859, +0.203604859554852403062088995281827210665664861); DVK(KP152703644, +0.152703644666139302296566746461370407999248646); DVK(KP586256827, +0.586256827714544512072145703099641959914944179); DVK(KP968908795, +0.968908795874236621082202410917456709164223497); DVK(KP726681596, +0.726681596905677465811651808188092531873167623); DVK(KP500000000, +0.500000000000000000000000000000000000000000000); { INT m; R *x; x = ri; for (m = mb, W = W + (mb * ((TWVL / VL) * 16)); m < me; m = m + VL, x = x + (VL * ms), W = W + (TWVL * 16), MAKE_VOLATILE_STRIDE(9, rs)) { V T1, T3, T5, T9, Th, Tb, Td, Tj, Tl, TD, T6; T1 = LD(&(x[0]), ms, &(x[0])); { V T2, T4, T8, Tg; T2 = LD(&(x[WS(rs, 3)]), ms, &(x[WS(rs, 1)])); T4 = LD(&(x[WS(rs, 6)]), ms, &(x[0])); T8 = LD(&(x[WS(rs, 1)]), ms, &(x[WS(rs, 1)])); Tg = LD(&(x[WS(rs, 2)]), ms, &(x[0])); { V Ta, Tc, Ti, Tk; Ta = LD(&(x[WS(rs, 4)]), ms, &(x[0])); Tc = LD(&(x[WS(rs, 7)]), ms, &(x[WS(rs, 1)])); Ti = LD(&(x[WS(rs, 5)]), ms, &(x[WS(rs, 1)])); Tk = LD(&(x[WS(rs, 8)]), ms, &(x[0])); T3 = BYTWJ(&(W[TWVL * 4]), T2); T5 = BYTWJ(&(W[TWVL * 10]), T4); T9 = BYTWJ(&(W[0]), T8); Th = BYTWJ(&(W[TWVL * 2]), Tg); Tb = BYTWJ(&(W[TWVL * 6]), Ta); Td = BYTWJ(&(W[TWVL * 12]), Tc); Tj = BYTWJ(&(W[TWVL * 8]), Ti); Tl = BYTWJ(&(W[TWVL * 14]), Tk); } } TD = VSUB(T5, T3); T6 = VADD(T3, T5); { V Tt, Te, Tu, Tm, Tr, T7; Tt = VSUB(Tb, Td); Te = VADD(Tb, Td); Tu = VSUB(Tl, Tj); Tm = VADD(Tj, Tl); Tr = VFNMS(LDK(KP500000000), T6, T1); T7 = VADD(T1, T6); { V Tv, Tf, Ts, Tn; Tv = VFNMS(LDK(KP500000000), Te, T9); Tf = VADD(T9, Te); Ts = VFNMS(LDK(KP500000000), Tm, Th); Tn = VADD(Th, Tm); { V TG, TK, Tw, TJ, TF, TA, To, Tq; TG = VFNMS(LDK(KP726681596), Tt, Tv); TK = VFMA(LDK(KP968908795), Tv, Tt); Tw = VFNMS(LDK(KP586256827), Tv, Tu); TJ = VFNMS(LDK(KP152703644), Tu, Ts); TF = VFMA(LDK(KP203604859), Ts, Tu); TA = VFNMS(LDK(KP439692620), Tt, Ts); To = VADD(Tf, Tn); Tq = VMUL(LDK(KP866025403), VSUB(Tn, Tf)); { V TQ, TH, TL, TN, TB, Tp, Ty, TI, Tx; Tx = VFNMS(LDK(KP347296355), Tw, Tt); TQ = VFNMS(LDK(KP898197570), TG, TF); TH = VFMA(LDK(KP898197570), TG, TF); TL = VFMA(LDK(KP673648177), TK, TJ); TN = VFNMS(LDK(KP673648177), TK, TJ); TB = VFNMS(LDK(KP420276625), TA, Tu); ST(&(x[0]), VADD(T7, To), ms, &(x[0])); Tp = VFNMS(LDK(KP500000000), To, T7); Ty = VFNMS(LDK(KP907603734), Tx, Ts); TI = VFMA(LDK(KP852868531), TH, Tr); { V TO, TR, TM, TC, Tz, TP, TS, TE; TO = VFNMS(LDK(KP500000000), TH, TN); TR = VFMA(LDK(KP666666666), TL, TQ); TM = VMUL(LDK(KP984807753), VFNMS(LDK(KP879385241), TD, TL)); TC = VFNMS(LDK(KP826351822), TB, Tv); ST(&(x[WS(rs, 6)]), VFNMSI(Tq, Tp), ms, &(x[0])); ST(&(x[WS(rs, 3)]), VFMAI(Tq, Tp), ms, &(x[WS(rs, 1)])); Tz = VFNMS(LDK(KP939692620), Ty, Tr); TP = VFMA(LDK(KP852868531), TO, Tr); TS = VMUL(LDK(KP866025403), VFMA(LDK(KP852868531), TR, TD)); ST(&(x[WS(rs, 8)]), VFMAI(TM, TI), ms, &(x[0])); ST(&(x[WS(rs, 1)]), VFNMSI(TM, TI), ms, &(x[WS(rs, 1)])); TE = VMUL(LDK(KP984807753), VFMA(LDK(KP879385241), TD, TC)); ST(&(x[WS(rs, 4)]), VFMAI(TS, TP), ms, &(x[0])); ST(&(x[WS(rs, 5)]), VFNMSI(TS, TP), ms, &(x[WS(rs, 1)])); ST(&(x[WS(rs, 7)]), VFMAI(TE, Tz), ms, &(x[WS(rs, 1)])); ST(&(x[WS(rs, 2)]), VFNMSI(TE, Tz), ms, &(x[0])); } } } } } } } VLEAVE(); } static const tw_instr twinstr[] = { VTW(0, 1), VTW(0, 2), VTW(0, 3), VTW(0, 4), VTW(0, 5), VTW(0, 6), VTW(0, 7), VTW(0, 8), {TW_NEXT, VL, 0} }; static const ct_desc desc = { 9, XSIMD_STRING("t1fv_9"), twinstr, &GENUS, {20, 20, 34, 0}, 0, 0, 0 }; void XSIMD(codelet_t1fv_9) (planner *p) { X(kdft_dit_register) (p, t1fv_9, &desc); } #else /* HAVE_FMA */ /* Generated by: ../../../genfft/gen_twiddle_c.native -simd -compact -variables 4 -pipeline-latency 8 -n 9 -name t1fv_9 -include t1f.h */ /* * This function contains 54 FP additions, 42 FP multiplications, * (or, 38 additions, 26 multiplications, 16 fused multiply/add), * 38 stack variables, 14 constants, and 18 memory accesses */ #include "t1f.h" static void t1fv_9(R *ri, R *ii, const R *W, stride rs, INT mb, INT me, INT ms) { DVK(KP939692620, +0.939692620785908384054109277324731469936208134); DVK(KP296198132, +0.296198132726023843175338011893050938967728390); DVK(KP852868531, +0.852868531952443209628250963940074071936020296); DVK(KP173648177, +0.173648177666930348851716626769314796000375677); DVK(KP556670399, +0.556670399226419366452912952047023132968291906); DVK(KP766044443, +0.766044443118978035202392650555416673935832457); DVK(KP642787609, +0.642787609686539326322643409907263432907559884); DVK(KP663413948, +0.663413948168938396205421319635891297216863310); DVK(KP984807753, +0.984807753012208059366743024589523013670643252); DVK(KP150383733, +0.150383733180435296639271897612501926072238258); DVK(KP342020143, +0.342020143325668733044099614682259580763083368); DVK(KP813797681, +0.813797681349373692844693217248393223289101568); DVK(KP500000000, +0.500000000000000000000000000000000000000000000); DVK(KP866025403, +0.866025403784438646763723170752936183471402627); { INT m; R *x; x = ri; for (m = mb, W = W + (mb * ((TWVL / VL) * 16)); m < me; m = m + VL, x = x + (VL * ms), W = W + (TWVL * 16), MAKE_VOLATILE_STRIDE(9, rs)) { V T1, T6, TA, Tt, Tf, Ts, Tw, Tn, Tv; T1 = LD(&(x[0]), ms, &(x[0])); { V T3, T5, T2, T4; T2 = LD(&(x[WS(rs, 3)]), ms, &(x[WS(rs, 1)])); T3 = BYTWJ(&(W[TWVL * 4]), T2); T4 = LD(&(x[WS(rs, 6)]), ms, &(x[0])); T5 = BYTWJ(&(W[TWVL * 10]), T4); T6 = VADD(T3, T5); TA = VMUL(LDK(KP866025403), VSUB(T5, T3)); } { V T9, Td, Tb, T8, Tc, Ta, Te; T8 = LD(&(x[WS(rs, 1)]), ms, &(x[WS(rs, 1)])); T9 = BYTWJ(&(W[0]), T8); Tc = LD(&(x[WS(rs, 7)]), ms, &(x[WS(rs, 1)])); Td = BYTWJ(&(W[TWVL * 12]), Tc); Ta = LD(&(x[WS(rs, 4)]), ms, &(x[0])); Tb = BYTWJ(&(W[TWVL * 6]), Ta); Tt = VSUB(Td, Tb); Te = VADD(Tb, Td); Tf = VADD(T9, Te); Ts = VFNMS(LDK(KP500000000), Te, T9); } { V Th, Tl, Tj, Tg, Tk, Ti, Tm; Tg = LD(&(x[WS(rs, 2)]), ms, &(x[0])); Th = BYTWJ(&(W[TWVL * 2]), Tg); Tk = LD(&(x[WS(rs, 8)]), ms, &(x[0])); Tl = BYTWJ(&(W[TWVL * 14]), Tk); Ti = LD(&(x[WS(rs, 5)]), ms, &(x[WS(rs, 1)])); Tj = BYTWJ(&(W[TWVL * 8]), Ti); Tw = VSUB(Tl, Tj); Tm = VADD(Tj, Tl); Tn = VADD(Th, Tm); Tv = VFNMS(LDK(KP500000000), Tm, Th); } { V Tq, T7, To, Tp; Tq = VBYI(VMUL(LDK(KP866025403), VSUB(Tn, Tf))); T7 = VADD(T1, T6); To = VADD(Tf, Tn); Tp = VFNMS(LDK(KP500000000), To, T7); ST(&(x[0]), VADD(T7, To), ms, &(x[0])); ST(&(x[WS(rs, 3)]), VADD(Tp, Tq), ms, &(x[WS(rs, 1)])); ST(&(x[WS(rs, 6)]), VSUB(Tp, Tq), ms, &(x[0])); } { V TI, TB, TC, TD, Tu, Tx, Ty, Tr, TH; TI = VBYI(VSUB(VFNMS(LDK(KP342020143), Tv, VFNMS(LDK(KP150383733), Tt, VFNMS(LDK(KP984807753), Ts, VMUL(LDK(KP813797681), Tw)))), TA)); TB = VFNMS(LDK(KP642787609), Ts, VMUL(LDK(KP663413948), Tt)); TC = VFNMS(LDK(KP984807753), Tv, VMUL(LDK(KP150383733), Tw)); TD = VADD(TB, TC); Tu = VFMA(LDK(KP766044443), Ts, VMUL(LDK(KP556670399), Tt)); Tx = VFMA(LDK(KP173648177), Tv, VMUL(LDK(KP852868531), Tw)); Ty = VADD(Tu, Tx); Tr = VFNMS(LDK(KP500000000), T6, T1); TH = VFMA(LDK(KP173648177), Ts, VFNMS(LDK(KP296198132), Tw, VFNMS(LDK(KP939692620), Tv, VFNMS(LDK(KP852868531), Tt, Tr)))); ST(&(x[WS(rs, 7)]), VSUB(TH, TI), ms, &(x[WS(rs, 1)])); ST(&(x[WS(rs, 2)]), VADD(TH, TI), ms, &(x[0])); { V Tz, TE, TF, TG; Tz = VADD(Tr, Ty); TE = VBYI(VADD(TA, TD)); ST(&(x[WS(rs, 8)]), VSUB(Tz, TE), ms, &(x[0])); ST(&(x[WS(rs, 1)]), VADD(TE, Tz), ms, &(x[WS(rs, 1)])); TF = VFMA(LDK(KP866025403), VSUB(TB, TC), VFNMS(LDK(KP500000000), Ty, Tr)); TG = VBYI(VADD(TA, VFNMS(LDK(KP500000000), TD, VMUL(LDK(KP866025403), VSUB(Tx, Tu))))); ST(&(x[WS(rs, 5)]), VSUB(TF, TG), ms, &(x[WS(rs, 1)])); ST(&(x[WS(rs, 4)]), VADD(TF, TG), ms, &(x[0])); } } } } VLEAVE(); } static const tw_instr twinstr[] = { VTW(0, 1), VTW(0, 2), VTW(0, 3), VTW(0, 4), VTW(0, 5), VTW(0, 6), VTW(0, 7), VTW(0, 8), {TW_NEXT, VL, 0} }; static const ct_desc desc = { 9, XSIMD_STRING("t1fv_9"), twinstr, &GENUS, {38, 26, 16, 0}, 0, 0, 0 }; void XSIMD(codelet_t1fv_9) (planner *p) { X(kdft_dit_register) (p, t1fv_9, &desc); } #endif /* HAVE_FMA */ fftw-3.3.3/dft/simd/common/t1fv_10.c0000644000175000017500000002254512054410453013742 00000000000000/* * Copyright (c) 2003, 2007-11 Matteo Frigo * Copyright (c) 2003, 2007-11 Massachusetts Institute of Technology * * This program is free software; you can redistribute it and/or modify * it under the terms of the GNU General Public License as published by * the Free Software Foundation; either version 2 of the License, or * (at your option) any later version. * * This program is distributed in the hope that it will be useful, * but WITHOUT ANY WARRANTY; without even the implied warranty of * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the * GNU General Public License for more details. * * You should have received a copy of the GNU General Public License * along with this program; if not, write to the Free Software * Foundation, Inc., 51 Franklin Street, Fifth Floor, Boston, MA 02110-1301 USA * */ /* This file was automatically generated --- DO NOT EDIT */ /* Generated on Sun Nov 25 07:38:03 EST 2012 */ #include "codelet-dft.h" #ifdef HAVE_FMA /* Generated by: ../../../genfft/gen_twiddle_c.native -fma -reorder-insns -schedule-for-pipeline -simd -compact -variables 4 -pipeline-latency 8 -n 10 -name t1fv_10 -include t1f.h */ /* * This function contains 51 FP additions, 40 FP multiplications, * (or, 33 additions, 22 multiplications, 18 fused multiply/add), * 43 stack variables, 4 constants, and 20 memory accesses */ #include "t1f.h" static void t1fv_10(R *ri, R *ii, const R *W, stride rs, INT mb, INT me, INT ms) { DVK(KP559016994, +0.559016994374947424102293417182819058860154590); DVK(KP250000000, +0.250000000000000000000000000000000000000000000); DVK(KP618033988, +0.618033988749894848204586834365638117720309180); DVK(KP951056516, +0.951056516295153572116439333379382143405698634); { INT m; R *x; x = ri; for (m = mb, W = W + (mb * ((TWVL / VL) * 18)); m < me; m = m + VL, x = x + (VL * ms), W = W + (TWVL * 18), MAKE_VOLATILE_STRIDE(10, rs)) { V Td, TA, T4, Ta, Tk, TE, Tp, TF, TB, T9, T1, T2, Tb; T1 = LD(&(x[0]), ms, &(x[0])); T2 = LD(&(x[WS(rs, 5)]), ms, &(x[WS(rs, 1)])); { V Tg, Tn, Ti, Tl; Tg = LD(&(x[WS(rs, 4)]), ms, &(x[0])); Tn = LD(&(x[WS(rs, 1)]), ms, &(x[WS(rs, 1)])); Ti = LD(&(x[WS(rs, 9)]), ms, &(x[WS(rs, 1)])); Tl = LD(&(x[WS(rs, 6)]), ms, &(x[0])); { V T6, T8, T5, Tc; T5 = LD(&(x[WS(rs, 2)]), ms, &(x[0])); Tc = LD(&(x[WS(rs, 3)]), ms, &(x[WS(rs, 1)])); { V T3, Th, To, Tj, Tm, T7; T7 = LD(&(x[WS(rs, 7)]), ms, &(x[WS(rs, 1)])); T3 = BYTWJ(&(W[TWVL * 8]), T2); Th = BYTWJ(&(W[TWVL * 6]), Tg); To = BYTWJ(&(W[0]), Tn); Tj = BYTWJ(&(W[TWVL * 16]), Ti); Tm = BYTWJ(&(W[TWVL * 10]), Tl); T6 = BYTWJ(&(W[TWVL * 2]), T5); Td = BYTWJ(&(W[TWVL * 4]), Tc); T8 = BYTWJ(&(W[TWVL * 12]), T7); TA = VADD(T1, T3); T4 = VSUB(T1, T3); Ta = LD(&(x[WS(rs, 8)]), ms, &(x[0])); Tk = VSUB(Th, Tj); TE = VADD(Th, Tj); Tp = VSUB(Tm, To); TF = VADD(Tm, To); } TB = VADD(T6, T8); T9 = VSUB(T6, T8); } } Tb = BYTWJ(&(W[TWVL * 14]), Ta); { V TL, TG, Tw, Tq, TC, Te; TL = VSUB(TE, TF); TG = VADD(TE, TF); Tw = VSUB(Tk, Tp); Tq = VADD(Tk, Tp); TC = VADD(Tb, Td); Te = VSUB(Tb, Td); { V TM, TD, Tv, Tf; TM = VSUB(TB, TC); TD = VADD(TB, TC); Tv = VSUB(T9, Te); Tf = VADD(T9, Te); { V TP, TN, TH, TJ, Tz, Tx, Tr, Tt, TI, Ts; TP = VMUL(LDK(KP951056516), VFMA(LDK(KP618033988), TL, TM)); TN = VMUL(LDK(KP951056516), VFNMS(LDK(KP618033988), TM, TL)); TH = VADD(TD, TG); TJ = VSUB(TD, TG); Tz = VMUL(LDK(KP951056516), VFNMS(LDK(KP618033988), Tv, Tw)); Tx = VMUL(LDK(KP951056516), VFMA(LDK(KP618033988), Tw, Tv)); Tr = VADD(Tf, Tq); Tt = VSUB(Tf, Tq); ST(&(x[0]), VADD(TA, TH), ms, &(x[0])); TI = VFNMS(LDK(KP250000000), TH, TA); ST(&(x[WS(rs, 5)]), VADD(T4, Tr), ms, &(x[WS(rs, 1)])); Ts = VFNMS(LDK(KP250000000), Tr, T4); { V TK, TO, Tu, Ty; TK = VFNMS(LDK(KP559016994), TJ, TI); TO = VFMA(LDK(KP559016994), TJ, TI); Tu = VFMA(LDK(KP559016994), Tt, Ts); Ty = VFNMS(LDK(KP559016994), Tt, Ts); ST(&(x[WS(rs, 8)]), VFNMSI(TN, TK), ms, &(x[0])); ST(&(x[WS(rs, 2)]), VFMAI(TN, TK), ms, &(x[0])); ST(&(x[WS(rs, 6)]), VFNMSI(TP, TO), ms, &(x[0])); ST(&(x[WS(rs, 4)]), VFMAI(TP, TO), ms, &(x[0])); ST(&(x[WS(rs, 9)]), VFMAI(Tx, Tu), ms, &(x[WS(rs, 1)])); ST(&(x[WS(rs, 1)]), VFNMSI(Tx, Tu), ms, &(x[WS(rs, 1)])); ST(&(x[WS(rs, 7)]), VFMAI(Tz, Ty), ms, &(x[WS(rs, 1)])); ST(&(x[WS(rs, 3)]), VFNMSI(Tz, Ty), ms, &(x[WS(rs, 1)])); } } } } } } VLEAVE(); } static const tw_instr twinstr[] = { VTW(0, 1), VTW(0, 2), VTW(0, 3), VTW(0, 4), VTW(0, 5), VTW(0, 6), VTW(0, 7), VTW(0, 8), VTW(0, 9), {TW_NEXT, VL, 0} }; static const ct_desc desc = { 10, XSIMD_STRING("t1fv_10"), twinstr, &GENUS, {33, 22, 18, 0}, 0, 0, 0 }; void XSIMD(codelet_t1fv_10) (planner *p) { X(kdft_dit_register) (p, t1fv_10, &desc); } #else /* HAVE_FMA */ /* Generated by: ../../../genfft/gen_twiddle_c.native -simd -compact -variables 4 -pipeline-latency 8 -n 10 -name t1fv_10 -include t1f.h */ /* * This function contains 51 FP additions, 30 FP multiplications, * (or, 45 additions, 24 multiplications, 6 fused multiply/add), * 32 stack variables, 4 constants, and 20 memory accesses */ #include "t1f.h" static void t1fv_10(R *ri, R *ii, const R *W, stride rs, INT mb, INT me, INT ms) { DVK(KP587785252, +0.587785252292473129168705954639072768597652438); DVK(KP951056516, +0.951056516295153572116439333379382143405698634); DVK(KP250000000, +0.250000000000000000000000000000000000000000000); DVK(KP559016994, +0.559016994374947424102293417182819058860154590); { INT m; R *x; x = ri; for (m = mb, W = W + (mb * ((TWVL / VL) * 18)); m < me; m = m + VL, x = x + (VL * ms), W = W + (TWVL * 18), MAKE_VOLATILE_STRIDE(10, rs)) { V Tr, TH, Tg, Tl, Tm, TA, TB, TJ, T5, Ta, Tb, TD, TE, TI, To; V Tq, Tp; To = LD(&(x[0]), ms, &(x[0])); Tp = LD(&(x[WS(rs, 5)]), ms, &(x[WS(rs, 1)])); Tq = BYTWJ(&(W[TWVL * 8]), Tp); Tr = VSUB(To, Tq); TH = VADD(To, Tq); { V Td, Tk, Tf, Ti; { V Tc, Tj, Te, Th; Tc = LD(&(x[WS(rs, 4)]), ms, &(x[0])); Td = BYTWJ(&(W[TWVL * 6]), Tc); Tj = LD(&(x[WS(rs, 1)]), ms, &(x[WS(rs, 1)])); Tk = BYTWJ(&(W[0]), Tj); Te = LD(&(x[WS(rs, 9)]), ms, &(x[WS(rs, 1)])); Tf = BYTWJ(&(W[TWVL * 16]), Te); Th = LD(&(x[WS(rs, 6)]), ms, &(x[0])); Ti = BYTWJ(&(W[TWVL * 10]), Th); } Tg = VSUB(Td, Tf); Tl = VSUB(Ti, Tk); Tm = VADD(Tg, Tl); TA = VADD(Td, Tf); TB = VADD(Ti, Tk); TJ = VADD(TA, TB); } { V T2, T9, T4, T7; { V T1, T8, T3, T6; T1 = LD(&(x[WS(rs, 2)]), ms, &(x[0])); T2 = BYTWJ(&(W[TWVL * 2]), T1); T8 = LD(&(x[WS(rs, 3)]), ms, &(x[WS(rs, 1)])); T9 = BYTWJ(&(W[TWVL * 4]), T8); T3 = LD(&(x[WS(rs, 7)]), ms, &(x[WS(rs, 1)])); T4 = BYTWJ(&(W[TWVL * 12]), T3); T6 = LD(&(x[WS(rs, 8)]), ms, &(x[0])); T7 = BYTWJ(&(W[TWVL * 14]), T6); } T5 = VSUB(T2, T4); Ta = VSUB(T7, T9); Tb = VADD(T5, Ta); TD = VADD(T2, T4); TE = VADD(T7, T9); TI = VADD(TD, TE); } { V Tn, Ts, Tt, Tx, Tz, Tv, Tw, Ty, Tu; Tn = VMUL(LDK(KP559016994), VSUB(Tb, Tm)); Ts = VADD(Tb, Tm); Tt = VFNMS(LDK(KP250000000), Ts, Tr); Tv = VSUB(T5, Ta); Tw = VSUB(Tg, Tl); Tx = VBYI(VFMA(LDK(KP951056516), Tv, VMUL(LDK(KP587785252), Tw))); Tz = VBYI(VFNMS(LDK(KP587785252), Tv, VMUL(LDK(KP951056516), Tw))); ST(&(x[WS(rs, 5)]), VADD(Tr, Ts), ms, &(x[WS(rs, 1)])); Ty = VSUB(Tt, Tn); ST(&(x[WS(rs, 3)]), VSUB(Ty, Tz), ms, &(x[WS(rs, 1)])); ST(&(x[WS(rs, 7)]), VADD(Tz, Ty), ms, &(x[WS(rs, 1)])); Tu = VADD(Tn, Tt); ST(&(x[WS(rs, 1)]), VSUB(Tu, Tx), ms, &(x[WS(rs, 1)])); ST(&(x[WS(rs, 9)]), VADD(Tx, Tu), ms, &(x[WS(rs, 1)])); } { V TM, TK, TL, TG, TO, TC, TF, TP, TN; TM = VMUL(LDK(KP559016994), VSUB(TI, TJ)); TK = VADD(TI, TJ); TL = VFNMS(LDK(KP250000000), TK, TH); TC = VSUB(TA, TB); TF = VSUB(TD, TE); TG = VBYI(VFNMS(LDK(KP587785252), TF, VMUL(LDK(KP951056516), TC))); TO = VBYI(VFMA(LDK(KP951056516), TF, VMUL(LDK(KP587785252), TC))); ST(&(x[0]), VADD(TH, TK), ms, &(x[0])); TP = VADD(TM, TL); ST(&(x[WS(rs, 4)]), VADD(TO, TP), ms, &(x[0])); ST(&(x[WS(rs, 6)]), VSUB(TP, TO), ms, &(x[0])); TN = VSUB(TL, TM); ST(&(x[WS(rs, 2)]), VADD(TG, TN), ms, &(x[0])); ST(&(x[WS(rs, 8)]), VSUB(TN, TG), ms, &(x[0])); } } } VLEAVE(); } static const tw_instr twinstr[] = { VTW(0, 1), VTW(0, 2), VTW(0, 3), VTW(0, 4), VTW(0, 5), VTW(0, 6), VTW(0, 7), VTW(0, 8), VTW(0, 9), {TW_NEXT, VL, 0} }; static const ct_desc desc = { 10, XSIMD_STRING("t1fv_10"), twinstr, &GENUS, {45, 24, 6, 0}, 0, 0, 0 }; void XSIMD(codelet_t1fv_10) (planner *p) { X(kdft_dit_register) (p, t1fv_10, &desc); } #endif /* HAVE_FMA */ fftw-3.3.3/dft/simd/common/t1fv_12.c0000644000175000017500000002363312054410454013744 00000000000000/* * Copyright (c) 2003, 2007-11 Matteo Frigo * Copyright (c) 2003, 2007-11 Massachusetts Institute of Technology * * This program is free software; you can redistribute it and/or modify * it under the terms of the GNU General Public License as published by * the Free Software Foundation; either version 2 of the License, or * (at your option) any later version. * * This program is distributed in the hope that it will be useful, * but WITHOUT ANY WARRANTY; without even the implied warranty of * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the * GNU General Public License for more details. * * You should have received a copy of the GNU General Public License * along with this program; if not, write to the Free Software * Foundation, Inc., 51 Franklin Street, Fifth Floor, Boston, MA 02110-1301 USA * */ /* This file was automatically generated --- DO NOT EDIT */ /* Generated on Sun Nov 25 07:38:03 EST 2012 */ #include "codelet-dft.h" #ifdef HAVE_FMA /* Generated by: ../../../genfft/gen_twiddle_c.native -fma -reorder-insns -schedule-for-pipeline -simd -compact -variables 4 -pipeline-latency 8 -n 12 -name t1fv_12 -include t1f.h */ /* * This function contains 59 FP additions, 42 FP multiplications, * (or, 41 additions, 24 multiplications, 18 fused multiply/add), * 41 stack variables, 2 constants, and 24 memory accesses */ #include "t1f.h" static void t1fv_12(R *ri, R *ii, const R *W, stride rs, INT mb, INT me, INT ms) { DVK(KP866025403, +0.866025403784438646763723170752936183471402627); DVK(KP500000000, +0.500000000000000000000000000000000000000000000); { INT m; R *x; x = ri; for (m = mb, W = W + (mb * ((TWVL / VL) * 22)); m < me; m = m + VL, x = x + (VL * ms), W = W + (TWVL * 22), MAKE_VOLATILE_STRIDE(12, rs)) { V Tq, Ti, T7, TQ, Tu, TA, TU, Tk, TR, Tf, TE, TM; { V T9, TC, Tj, TD, Te; { V T1, T4, T2, Tm, Tx, To; T1 = LD(&(x[0]), ms, &(x[0])); T4 = LD(&(x[WS(rs, 8)]), ms, &(x[0])); T2 = LD(&(x[WS(rs, 4)]), ms, &(x[0])); Tm = LD(&(x[WS(rs, 1)]), ms, &(x[WS(rs, 1)])); Tx = LD(&(x[WS(rs, 9)]), ms, &(x[WS(rs, 1)])); To = LD(&(x[WS(rs, 5)]), ms, &(x[WS(rs, 1)])); { V T5, T3, Tn, Ty, Tp, Td, Tb, T8, Tc, Ta; T8 = LD(&(x[WS(rs, 6)]), ms, &(x[0])); Tc = LD(&(x[WS(rs, 2)]), ms, &(x[0])); Ta = LD(&(x[WS(rs, 10)]), ms, &(x[0])); T5 = BYTWJ(&(W[TWVL * 14]), T4); T3 = BYTWJ(&(W[TWVL * 6]), T2); Tn = BYTWJ(&(W[0]), Tm); Ty = BYTWJ(&(W[TWVL * 16]), Tx); Tp = BYTWJ(&(W[TWVL * 8]), To); T9 = BYTWJ(&(W[TWVL * 10]), T8); Td = BYTWJ(&(W[TWVL * 2]), Tc); Tb = BYTWJ(&(W[TWVL * 18]), Ta); { V Th, T6, Tt, Tz; Th = LD(&(x[WS(rs, 11)]), ms, &(x[WS(rs, 1)])); TC = VSUB(T5, T3); T6 = VADD(T3, T5); Tt = LD(&(x[WS(rs, 3)]), ms, &(x[WS(rs, 1)])); Tz = VADD(Tn, Tp); Tq = VSUB(Tn, Tp); Tj = LD(&(x[WS(rs, 7)]), ms, &(x[WS(rs, 1)])); TD = VSUB(Td, Tb); Te = VADD(Tb, Td); Ti = BYTWJ(&(W[TWVL * 20]), Th); T7 = VFNMS(LDK(KP500000000), T6, T1); TQ = VADD(T1, T6); Tu = BYTWJ(&(W[TWVL * 4]), Tt); TA = VFNMS(LDK(KP500000000), Tz, Ty); TU = VADD(Ty, Tz); } } } Tk = BYTWJ(&(W[TWVL * 12]), Tj); TR = VADD(T9, Te); Tf = VFNMS(LDK(KP500000000), Te, T9); TE = VSUB(TC, TD); TM = VADD(TC, TD); } { V Tv, Tl, TI, Tg, TW, TS; Tv = VADD(Tk, Ti); Tl = VSUB(Ti, Tk); TI = VADD(T7, Tf); Tg = VSUB(T7, Tf); TW = VADD(TQ, TR); TS = VSUB(TQ, TR); { V TT, Tw, TL, Tr; TT = VADD(Tu, Tv); Tw = VFNMS(LDK(KP500000000), Tv, Tu); TL = VSUB(Tl, Tq); Tr = VADD(Tl, Tq); { V TP, TN, TG, Ts, TO, TK, TH, TF; { V TX, TV, TJ, TB; TX = VADD(TT, TU); TV = VSUB(TT, TU); TJ = VADD(Tw, TA); TB = VSUB(Tw, TA); TP = VMUL(LDK(KP866025403), VADD(TM, TL)); TN = VMUL(LDK(KP866025403), VSUB(TL, TM)); TG = VFNMS(LDK(KP866025403), Tr, Tg); Ts = VFMA(LDK(KP866025403), Tr, Tg); ST(&(x[WS(rs, 6)]), VSUB(TW, TX), ms, &(x[0])); ST(&(x[0]), VADD(TW, TX), ms, &(x[0])); ST(&(x[WS(rs, 3)]), VFMAI(TV, TS), ms, &(x[WS(rs, 1)])); ST(&(x[WS(rs, 9)]), VFNMSI(TV, TS), ms, &(x[WS(rs, 1)])); TO = VADD(TI, TJ); TK = VSUB(TI, TJ); TH = VFMA(LDK(KP866025403), TE, TB); TF = VFNMS(LDK(KP866025403), TE, TB); } ST(&(x[WS(rs, 4)]), VFMAI(TP, TO), ms, &(x[0])); ST(&(x[WS(rs, 8)]), VFNMSI(TP, TO), ms, &(x[0])); ST(&(x[WS(rs, 10)]), VFNMSI(TN, TK), ms, &(x[0])); ST(&(x[WS(rs, 2)]), VFMAI(TN, TK), ms, &(x[0])); ST(&(x[WS(rs, 5)]), VFNMSI(TH, TG), ms, &(x[WS(rs, 1)])); ST(&(x[WS(rs, 7)]), VFMAI(TH, TG), ms, &(x[WS(rs, 1)])); ST(&(x[WS(rs, 11)]), VFMAI(TF, Ts), ms, &(x[WS(rs, 1)])); ST(&(x[WS(rs, 1)]), VFNMSI(TF, Ts), ms, &(x[WS(rs, 1)])); } } } } } VLEAVE(); } static const tw_instr twinstr[] = { VTW(0, 1), VTW(0, 2), VTW(0, 3), VTW(0, 4), VTW(0, 5), VTW(0, 6), VTW(0, 7), VTW(0, 8), VTW(0, 9), VTW(0, 10), VTW(0, 11), {TW_NEXT, VL, 0} }; static const ct_desc desc = { 12, XSIMD_STRING("t1fv_12"), twinstr, &GENUS, {41, 24, 18, 0}, 0, 0, 0 }; void XSIMD(codelet_t1fv_12) (planner *p) { X(kdft_dit_register) (p, t1fv_12, &desc); } #else /* HAVE_FMA */ /* Generated by: ../../../genfft/gen_twiddle_c.native -simd -compact -variables 4 -pipeline-latency 8 -n 12 -name t1fv_12 -include t1f.h */ /* * This function contains 59 FP additions, 30 FP multiplications, * (or, 55 additions, 26 multiplications, 4 fused multiply/add), * 28 stack variables, 2 constants, and 24 memory accesses */ #include "t1f.h" static void t1fv_12(R *ri, R *ii, const R *W, stride rs, INT mb, INT me, INT ms) { DVK(KP866025403, +0.866025403784438646763723170752936183471402627); DVK(KP500000000, +0.500000000000000000000000000000000000000000000); { INT m; R *x; x = ri; for (m = mb, W = W + (mb * ((TWVL / VL) * 22)); m < me; m = m + VL, x = x + (VL * ms), W = W + (TWVL * 22), MAKE_VOLATILE_STRIDE(12, rs)) { V T1, TH, T6, TA, Tq, TE, Tv, TL, T9, TI, Te, TB, Ti, TD, Tn; V TK; { V T5, T3, T4, T2; T1 = LD(&(x[0]), ms, &(x[0])); T4 = LD(&(x[WS(rs, 8)]), ms, &(x[0])); T5 = BYTWJ(&(W[TWVL * 14]), T4); T2 = LD(&(x[WS(rs, 4)]), ms, &(x[0])); T3 = BYTWJ(&(W[TWVL * 6]), T2); TH = VSUB(T5, T3); T6 = VADD(T3, T5); TA = VFNMS(LDK(KP500000000), T6, T1); } { V Tu, Ts, Tp, Tt, Tr; Tp = LD(&(x[WS(rs, 9)]), ms, &(x[WS(rs, 1)])); Tq = BYTWJ(&(W[TWVL * 16]), Tp); Tt = LD(&(x[WS(rs, 5)]), ms, &(x[WS(rs, 1)])); Tu = BYTWJ(&(W[TWVL * 8]), Tt); Tr = LD(&(x[WS(rs, 1)]), ms, &(x[WS(rs, 1)])); Ts = BYTWJ(&(W[0]), Tr); TE = VSUB(Tu, Ts); Tv = VADD(Ts, Tu); TL = VFNMS(LDK(KP500000000), Tv, Tq); } { V Td, Tb, T8, Tc, Ta; T8 = LD(&(x[WS(rs, 6)]), ms, &(x[0])); T9 = BYTWJ(&(W[TWVL * 10]), T8); Tc = LD(&(x[WS(rs, 2)]), ms, &(x[0])); Td = BYTWJ(&(W[TWVL * 2]), Tc); Ta = LD(&(x[WS(rs, 10)]), ms, &(x[0])); Tb = BYTWJ(&(W[TWVL * 18]), Ta); TI = VSUB(Td, Tb); Te = VADD(Tb, Td); TB = VFNMS(LDK(KP500000000), Te, T9); } { V Tm, Tk, Th, Tl, Tj; Th = LD(&(x[WS(rs, 3)]), ms, &(x[WS(rs, 1)])); Ti = BYTWJ(&(W[TWVL * 4]), Th); Tl = LD(&(x[WS(rs, 11)]), ms, &(x[WS(rs, 1)])); Tm = BYTWJ(&(W[TWVL * 20]), Tl); Tj = LD(&(x[WS(rs, 7)]), ms, &(x[WS(rs, 1)])); Tk = BYTWJ(&(W[TWVL * 12]), Tj); TD = VSUB(Tm, Tk); Tn = VADD(Tk, Tm); TK = VFNMS(LDK(KP500000000), Tn, Ti); } { V Tg, Ty, Tx, Tz; { V T7, Tf, To, Tw; T7 = VADD(T1, T6); Tf = VADD(T9, Te); Tg = VSUB(T7, Tf); Ty = VADD(T7, Tf); To = VADD(Ti, Tn); Tw = VADD(Tq, Tv); Tx = VBYI(VSUB(To, Tw)); Tz = VADD(To, Tw); } ST(&(x[WS(rs, 9)]), VSUB(Tg, Tx), ms, &(x[WS(rs, 1)])); ST(&(x[0]), VADD(Ty, Tz), ms, &(x[0])); ST(&(x[WS(rs, 3)]), VADD(Tg, Tx), ms, &(x[WS(rs, 1)])); ST(&(x[WS(rs, 6)]), VSUB(Ty, Tz), ms, &(x[0])); } { V TS, TW, TV, TX; { V TQ, TR, TT, TU; TQ = VADD(TA, TB); TR = VADD(TK, TL); TS = VSUB(TQ, TR); TW = VADD(TQ, TR); TT = VADD(TD, TE); TU = VADD(TH, TI); TV = VBYI(VMUL(LDK(KP866025403), VSUB(TT, TU))); TX = VBYI(VMUL(LDK(KP866025403), VADD(TU, TT))); } ST(&(x[WS(rs, 10)]), VSUB(TS, TV), ms, &(x[0])); ST(&(x[WS(rs, 4)]), VADD(TW, TX), ms, &(x[0])); ST(&(x[WS(rs, 2)]), VADD(TS, TV), ms, &(x[0])); ST(&(x[WS(rs, 8)]), VSUB(TW, TX), ms, &(x[0])); } { V TG, TP, TN, TO; { V TC, TF, TJ, TM; TC = VSUB(TA, TB); TF = VMUL(LDK(KP866025403), VSUB(TD, TE)); TG = VSUB(TC, TF); TP = VADD(TC, TF); TJ = VMUL(LDK(KP866025403), VSUB(TH, TI)); TM = VSUB(TK, TL); TN = VBYI(VADD(TJ, TM)); TO = VBYI(VSUB(TJ, TM)); } ST(&(x[WS(rs, 5)]), VSUB(TG, TN), ms, &(x[WS(rs, 1)])); ST(&(x[WS(rs, 11)]), VSUB(TP, TO), ms, &(x[WS(rs, 1)])); ST(&(x[WS(rs, 7)]), VADD(TN, TG), ms, &(x[WS(rs, 1)])); ST(&(x[WS(rs, 1)]), VADD(TO, TP), ms, &(x[WS(rs, 1)])); } } } VLEAVE(); } static const tw_instr twinstr[] = { VTW(0, 1), VTW(0, 2), VTW(0, 3), VTW(0, 4), VTW(0, 5), VTW(0, 6), VTW(0, 7), VTW(0, 8), VTW(0, 9), VTW(0, 10), VTW(0, 11), {TW_NEXT, VL, 0} }; static const ct_desc desc = { 12, XSIMD_STRING("t1fv_12"), twinstr, &GENUS, {55, 26, 4, 0}, 0, 0, 0 }; void XSIMD(codelet_t1fv_12) (planner *p) { X(kdft_dit_register) (p, t1fv_12, &desc); } #endif /* HAVE_FMA */ fftw-3.3.3/dft/simd/common/t1fv_15.c0000644000175000017500000003524012054410455013745 00000000000000/* * Copyright (c) 2003, 2007-11 Matteo Frigo * Copyright (c) 2003, 2007-11 Massachusetts Institute of Technology * * This program is free software; you can redistribute it and/or modify * it under the terms of the GNU General Public License as published by * the Free Software Foundation; either version 2 of the License, or * (at your option) any later version. * * This program is distributed in the hope that it will be useful, * but WITHOUT ANY WARRANTY; without even the implied warranty of * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the * GNU General Public License for more details. * * You should have received a copy of the GNU General Public License * along with this program; if not, write to the Free Software * Foundation, Inc., 51 Franklin Street, Fifth Floor, Boston, MA 02110-1301 USA * */ /* This file was automatically generated --- DO NOT EDIT */ /* Generated on Sun Nov 25 07:38:04 EST 2012 */ #include "codelet-dft.h" #ifdef HAVE_FMA /* Generated by: ../../../genfft/gen_twiddle_c.native -fma -reorder-insns -schedule-for-pipeline -simd -compact -variables 4 -pipeline-latency 8 -n 15 -name t1fv_15 -include t1f.h */ /* * This function contains 92 FP additions, 77 FP multiplications, * (or, 50 additions, 35 multiplications, 42 fused multiply/add), * 81 stack variables, 8 constants, and 30 memory accesses */ #include "t1f.h" static void t1fv_15(R *ri, R *ii, const R *W, stride rs, INT mb, INT me, INT ms) { DVK(KP823639103, +0.823639103546331925877420039278190003029660514); DVK(KP910592997, +0.910592997310029334643087372129977886038870291); DVK(KP559016994, +0.559016994374947424102293417182819058860154590); DVK(KP951056516, +0.951056516295153572116439333379382143405698634); DVK(KP866025403, +0.866025403784438646763723170752936183471402627); DVK(KP250000000, +0.250000000000000000000000000000000000000000000); DVK(KP618033988, +0.618033988749894848204586834365638117720309180); DVK(KP500000000, +0.500000000000000000000000000000000000000000000); { INT m; R *x; x = ri; for (m = mb, W = W + (mb * ((TWVL / VL) * 28)); m < me; m = m + VL, x = x + (VL * ms), W = W + (TWVL * 28), MAKE_VOLATILE_STRIDE(15, rs)) { V Tq, Ty, Th, T1b, T10, Ts, TP, T7, Tu, TA, TC, Tj, Tk, TQ, Tf; { V T1, T4, T2, T9, Te; T1 = LD(&(x[0]), ms, &(x[0])); T4 = LD(&(x[WS(rs, 10)]), ms, &(x[0])); T2 = LD(&(x[WS(rs, 5)]), ms, &(x[WS(rs, 1)])); { V T8, Tp, Tx, Tg; T8 = LD(&(x[WS(rs, 3)]), ms, &(x[WS(rs, 1)])); Tp = LD(&(x[WS(rs, 6)]), ms, &(x[0])); Tx = LD(&(x[WS(rs, 9)]), ms, &(x[WS(rs, 1)])); Tg = LD(&(x[WS(rs, 12)]), ms, &(x[0])); { V Tb, Td, Tr, T6, Tt, Tz, TB, Ti; { V T5, T3, Ta, Tc; Ta = LD(&(x[WS(rs, 8)]), ms, &(x[0])); Tc = LD(&(x[WS(rs, 13)]), ms, &(x[WS(rs, 1)])); T5 = BYTWJ(&(W[TWVL * 18]), T4); T3 = BYTWJ(&(W[TWVL * 8]), T2); T9 = BYTWJ(&(W[TWVL * 4]), T8); Tq = BYTWJ(&(W[TWVL * 10]), Tp); Ty = BYTWJ(&(W[TWVL * 16]), Tx); Th = BYTWJ(&(W[TWVL * 22]), Tg); Tb = BYTWJ(&(W[TWVL * 14]), Ta); Td = BYTWJ(&(W[TWVL * 24]), Tc); Tr = LD(&(x[WS(rs, 11)]), ms, &(x[WS(rs, 1)])); T1b = VSUB(T5, T3); T6 = VADD(T3, T5); Tt = LD(&(x[WS(rs, 1)]), ms, &(x[WS(rs, 1)])); } Tz = LD(&(x[WS(rs, 14)]), ms, &(x[0])); TB = LD(&(x[WS(rs, 4)]), ms, &(x[0])); Ti = LD(&(x[WS(rs, 2)]), ms, &(x[0])); Te = VADD(Tb, Td); T10 = VSUB(Td, Tb); Ts = BYTWJ(&(W[TWVL * 20]), Tr); TP = VFNMS(LDK(KP500000000), T6, T1); T7 = VADD(T1, T6); Tu = BYTWJ(&(W[0]), Tt); TA = BYTWJ(&(W[TWVL * 26]), Tz); TC = BYTWJ(&(W[TWVL * 6]), TB); Tj = BYTWJ(&(W[TWVL * 2]), Ti); Tk = LD(&(x[WS(rs, 7)]), ms, &(x[WS(rs, 1)])); } } TQ = VFNMS(LDK(KP500000000), Te, T9); Tf = VADD(T9, Te); } { V Tv, T13, TD, T14, Tl; Tv = VADD(Ts, Tu); T13 = VSUB(Tu, Ts); TD = VADD(TA, TC); T14 = VSUB(TC, TA); Tl = BYTWJ(&(W[TWVL * 12]), Tk); { V TT, Tw, T1d, T15, TU, TE, T11, Tm; TT = VFNMS(LDK(KP500000000), Tv, Tq); Tw = VADD(Tq, Tv); T1d = VADD(T13, T14); T15 = VSUB(T13, T14); TU = VFNMS(LDK(KP500000000), TD, Ty); TE = VADD(Ty, TD); T11 = VSUB(Tl, Tj); Tm = VADD(Tj, Tl); { V T19, TV, TK, TF, T1c, T12, TR, Tn; T19 = VSUB(TT, TU); TV = VADD(TT, TU); TK = VSUB(Tw, TE); TF = VADD(Tw, TE); T1c = VADD(T10, T11); T12 = VSUB(T10, T11); TR = VFNMS(LDK(KP500000000), Tm, Th); Tn = VADD(Th, Tm); { V T1g, T1e, T1m, T16, T18, TS, TL, To, T1f, T1u; T1g = VSUB(T1c, T1d); T1e = VADD(T1c, T1d); T1m = VFNMS(LDK(KP618033988), T12, T15); T16 = VFMA(LDK(KP618033988), T15, T12); T18 = VSUB(TQ, TR); TS = VADD(TQ, TR); TL = VSUB(Tf, Tn); To = VADD(Tf, Tn); T1f = VFNMS(LDK(KP250000000), T1e, T1b); T1u = VMUL(LDK(KP866025403), VADD(T1b, T1e)); { V T1o, T1a, TY, TO, TM, TG, TI, T1p, T1h, T1t, TX, TW; T1o = VFNMS(LDK(KP618033988), T18, T19); T1a = VFMA(LDK(KP618033988), T19, T18); TW = VADD(TS, TV); TY = VSUB(TS, TV); TO = VMUL(LDK(KP951056516), VFMA(LDK(KP618033988), TK, TL)); TM = VMUL(LDK(KP951056516), VFNMS(LDK(KP618033988), TL, TK)); TG = VADD(To, TF); TI = VSUB(To, TF); T1p = VFNMS(LDK(KP559016994), T1g, T1f); T1h = VFMA(LDK(KP559016994), T1g, T1f); T1t = VADD(TP, TW); TX = VFNMS(LDK(KP250000000), TW, TP); { V T1q, T1s, T1k, T1i, T1l, TZ, TJ, TN, TH; ST(&(x[0]), VADD(T7, TG), ms, &(x[0])); TH = VFNMS(LDK(KP250000000), TG, T7); T1q = VMUL(LDK(KP951056516), VFNMS(LDK(KP910592997), T1p, T1o)); T1s = VMUL(LDK(KP951056516), VFMA(LDK(KP910592997), T1p, T1o)); T1k = VMUL(LDK(KP951056516), VFMA(LDK(KP910592997), T1h, T1a)); T1i = VMUL(LDK(KP951056516), VFNMS(LDK(KP910592997), T1h, T1a)); ST(&(x[WS(rs, 10)]), VFMAI(T1u, T1t), ms, &(x[0])); ST(&(x[WS(rs, 5)]), VFNMSI(T1u, T1t), ms, &(x[WS(rs, 1)])); T1l = VFNMS(LDK(KP559016994), TY, TX); TZ = VFMA(LDK(KP559016994), TY, TX); TJ = VFNMS(LDK(KP559016994), TI, TH); TN = VFMA(LDK(KP559016994), TI, TH); { V T1n, T1r, T1j, T17; T1n = VFMA(LDK(KP823639103), T1m, T1l); T1r = VFNMS(LDK(KP823639103), T1m, T1l); T1j = VFNMS(LDK(KP823639103), T16, TZ); T17 = VFMA(LDK(KP823639103), T16, TZ); ST(&(x[WS(rs, 12)]), VFMAI(TM, TJ), ms, &(x[0])); ST(&(x[WS(rs, 3)]), VFNMSI(TM, TJ), ms, &(x[WS(rs, 1)])); ST(&(x[WS(rs, 9)]), VFMAI(TO, TN), ms, &(x[WS(rs, 1)])); ST(&(x[WS(rs, 6)]), VFNMSI(TO, TN), ms, &(x[0])); ST(&(x[WS(rs, 2)]), VFMAI(T1q, T1n), ms, &(x[0])); ST(&(x[WS(rs, 13)]), VFNMSI(T1q, T1n), ms, &(x[WS(rs, 1)])); ST(&(x[WS(rs, 7)]), VFMAI(T1s, T1r), ms, &(x[WS(rs, 1)])); ST(&(x[WS(rs, 8)]), VFNMSI(T1s, T1r), ms, &(x[0])); ST(&(x[WS(rs, 4)]), VFMAI(T1k, T1j), ms, &(x[0])); ST(&(x[WS(rs, 11)]), VFNMSI(T1k, T1j), ms, &(x[WS(rs, 1)])); ST(&(x[WS(rs, 14)]), VFMAI(T1i, T17), ms, &(x[0])); ST(&(x[WS(rs, 1)]), VFNMSI(T1i, T17), ms, &(x[WS(rs, 1)])); } } } } } } } } } VLEAVE(); } static const tw_instr twinstr[] = { VTW(0, 1), VTW(0, 2), VTW(0, 3), VTW(0, 4), VTW(0, 5), VTW(0, 6), VTW(0, 7), VTW(0, 8), VTW(0, 9), VTW(0, 10), VTW(0, 11), VTW(0, 12), VTW(0, 13), VTW(0, 14), {TW_NEXT, VL, 0} }; static const ct_desc desc = { 15, XSIMD_STRING("t1fv_15"), twinstr, &GENUS, {50, 35, 42, 0}, 0, 0, 0 }; void XSIMD(codelet_t1fv_15) (planner *p) { X(kdft_dit_register) (p, t1fv_15, &desc); } #else /* HAVE_FMA */ /* Generated by: ../../../genfft/gen_twiddle_c.native -simd -compact -variables 4 -pipeline-latency 8 -n 15 -name t1fv_15 -include t1f.h */ /* * This function contains 92 FP additions, 53 FP multiplications, * (or, 78 additions, 39 multiplications, 14 fused multiply/add), * 52 stack variables, 10 constants, and 30 memory accesses */ #include "t1f.h" static void t1fv_15(R *ri, R *ii, const R *W, stride rs, INT mb, INT me, INT ms) { DVK(KP216506350, +0.216506350946109661690930792688234045867850657); DVK(KP484122918, +0.484122918275927110647408174972799951354115213); DVK(KP866025403, +0.866025403784438646763723170752936183471402627); DVK(KP509036960, +0.509036960455127183450980863393907648510733164); DVK(KP823639103, +0.823639103546331925877420039278190003029660514); DVK(KP587785252, +0.587785252292473129168705954639072768597652438); DVK(KP951056516, +0.951056516295153572116439333379382143405698634); DVK(KP250000000, +0.250000000000000000000000000000000000000000000); DVK(KP559016994, +0.559016994374947424102293417182819058860154590); DVK(KP500000000, +0.500000000000000000000000000000000000000000000); { INT m; R *x; x = ri; for (m = mb, W = W + (mb * ((TWVL / VL) * 28)); m < me; m = m + VL, x = x + (VL * ms), W = W + (TWVL * 28), MAKE_VOLATILE_STRIDE(15, rs)) { V T1e, T7, TP, T12, T15, Tf, Tn, To, T1b, T1c, T1f, TQ, TR, TS, Tw; V TE, TF, TT, TU, TV; { V T1, T5, T3, T4, T2, T6; T1 = LD(&(x[0]), ms, &(x[0])); T4 = LD(&(x[WS(rs, 10)]), ms, &(x[0])); T5 = BYTWJ(&(W[TWVL * 18]), T4); T2 = LD(&(x[WS(rs, 5)]), ms, &(x[WS(rs, 1)])); T3 = BYTWJ(&(W[TWVL * 8]), T2); T1e = VSUB(T5, T3); T6 = VADD(T3, T5); T7 = VADD(T1, T6); TP = VFNMS(LDK(KP500000000), T6, T1); } { V T9, Tq, Ty, Th, Te, T13, Tv, T10, TD, T11, Tm, T14; { V T8, Tp, Tx, Tg; T8 = LD(&(x[WS(rs, 3)]), ms, &(x[WS(rs, 1)])); T9 = BYTWJ(&(W[TWVL * 4]), T8); Tp = LD(&(x[WS(rs, 6)]), ms, &(x[0])); Tq = BYTWJ(&(W[TWVL * 10]), Tp); Tx = LD(&(x[WS(rs, 9)]), ms, &(x[WS(rs, 1)])); Ty = BYTWJ(&(W[TWVL * 16]), Tx); Tg = LD(&(x[WS(rs, 12)]), ms, &(x[0])); Th = BYTWJ(&(W[TWVL * 22]), Tg); } { V Tb, Td, Ta, Tc; Ta = LD(&(x[WS(rs, 8)]), ms, &(x[0])); Tb = BYTWJ(&(W[TWVL * 14]), Ta); Tc = LD(&(x[WS(rs, 13)]), ms, &(x[WS(rs, 1)])); Td = BYTWJ(&(W[TWVL * 24]), Tc); Te = VADD(Tb, Td); T13 = VSUB(Td, Tb); } { V Ts, Tu, Tr, Tt; Tr = LD(&(x[WS(rs, 11)]), ms, &(x[WS(rs, 1)])); Ts = BYTWJ(&(W[TWVL * 20]), Tr); Tt = LD(&(x[WS(rs, 1)]), ms, &(x[WS(rs, 1)])); Tu = BYTWJ(&(W[0]), Tt); Tv = VADD(Ts, Tu); T10 = VSUB(Tu, Ts); } { V TA, TC, Tz, TB; Tz = LD(&(x[WS(rs, 14)]), ms, &(x[0])); TA = BYTWJ(&(W[TWVL * 26]), Tz); TB = LD(&(x[WS(rs, 4)]), ms, &(x[0])); TC = BYTWJ(&(W[TWVL * 6]), TB); TD = VADD(TA, TC); T11 = VSUB(TC, TA); } { V Tj, Tl, Ti, Tk; Ti = LD(&(x[WS(rs, 2)]), ms, &(x[0])); Tj = BYTWJ(&(W[TWVL * 2]), Ti); Tk = LD(&(x[WS(rs, 7)]), ms, &(x[WS(rs, 1)])); Tl = BYTWJ(&(W[TWVL * 12]), Tk); Tm = VADD(Tj, Tl); T14 = VSUB(Tl, Tj); } T12 = VSUB(T10, T11); T15 = VSUB(T13, T14); Tf = VADD(T9, Te); Tn = VADD(Th, Tm); To = VADD(Tf, Tn); T1b = VADD(T13, T14); T1c = VADD(T10, T11); T1f = VADD(T1b, T1c); TQ = VFNMS(LDK(KP500000000), Te, T9); TR = VFNMS(LDK(KP500000000), Tm, Th); TS = VADD(TQ, TR); Tw = VADD(Tq, Tv); TE = VADD(Ty, TD); TF = VADD(Tw, TE); TT = VFNMS(LDK(KP500000000), Tv, Tq); TU = VFNMS(LDK(KP500000000), TD, Ty); TV = VADD(TT, TU); } { V TI, TG, TH, TM, TO, TK, TL, TN, TJ; TI = VMUL(LDK(KP559016994), VSUB(To, TF)); TG = VADD(To, TF); TH = VFNMS(LDK(KP250000000), TG, T7); TK = VSUB(Tw, TE); TL = VSUB(Tf, Tn); TM = VBYI(VFNMS(LDK(KP587785252), TL, VMUL(LDK(KP951056516), TK))); TO = VBYI(VFMA(LDK(KP951056516), TL, VMUL(LDK(KP587785252), TK))); ST(&(x[0]), VADD(T7, TG), ms, &(x[0])); TN = VADD(TI, TH); ST(&(x[WS(rs, 6)]), VSUB(TN, TO), ms, &(x[0])); ST(&(x[WS(rs, 9)]), VADD(TO, TN), ms, &(x[WS(rs, 1)])); TJ = VSUB(TH, TI); ST(&(x[WS(rs, 3)]), VSUB(TJ, TM), ms, &(x[WS(rs, 1)])); ST(&(x[WS(rs, 12)]), VADD(TM, TJ), ms, &(x[0])); } { V T16, T1m, T1u, T1h, T1o, T1a, T1p, TZ, T1t, T1l, T1d, T1g; T16 = VFNMS(LDK(KP509036960), T15, VMUL(LDK(KP823639103), T12)); T1m = VFMA(LDK(KP823639103), T15, VMUL(LDK(KP509036960), T12)); T1u = VBYI(VMUL(LDK(KP866025403), VADD(T1e, T1f))); T1d = VMUL(LDK(KP484122918), VSUB(T1b, T1c)); T1g = VFNMS(LDK(KP216506350), T1f, VMUL(LDK(KP866025403), T1e)); T1h = VSUB(T1d, T1g); T1o = VADD(T1d, T1g); { V T18, T19, TY, TW, TX; T18 = VSUB(TT, TU); T19 = VSUB(TQ, TR); T1a = VFNMS(LDK(KP587785252), T19, VMUL(LDK(KP951056516), T18)); T1p = VFMA(LDK(KP951056516), T19, VMUL(LDK(KP587785252), T18)); TY = VMUL(LDK(KP559016994), VSUB(TS, TV)); TW = VADD(TS, TV); TX = VFNMS(LDK(KP250000000), TW, TP); TZ = VSUB(TX, TY); T1t = VADD(TP, TW); T1l = VADD(TY, TX); } { V T17, T1i, T1r, T1s; ST(&(x[WS(rs, 5)]), VSUB(T1t, T1u), ms, &(x[WS(rs, 1)])); ST(&(x[WS(rs, 10)]), VADD(T1t, T1u), ms, &(x[0])); T17 = VSUB(TZ, T16); T1i = VBYI(VSUB(T1a, T1h)); ST(&(x[WS(rs, 8)]), VSUB(T17, T1i), ms, &(x[0])); ST(&(x[WS(rs, 7)]), VADD(T17, T1i), ms, &(x[WS(rs, 1)])); T1r = VSUB(T1l, T1m); T1s = VBYI(VADD(T1p, T1o)); ST(&(x[WS(rs, 11)]), VSUB(T1r, T1s), ms, &(x[WS(rs, 1)])); ST(&(x[WS(rs, 4)]), VADD(T1r, T1s), ms, &(x[0])); { V T1n, T1q, T1j, T1k; T1n = VADD(T1l, T1m); T1q = VBYI(VSUB(T1o, T1p)); ST(&(x[WS(rs, 14)]), VSUB(T1n, T1q), ms, &(x[0])); ST(&(x[WS(rs, 1)]), VADD(T1n, T1q), ms, &(x[WS(rs, 1)])); T1j = VADD(TZ, T16); T1k = VBYI(VADD(T1a, T1h)); ST(&(x[WS(rs, 13)]), VSUB(T1j, T1k), ms, &(x[WS(rs, 1)])); ST(&(x[WS(rs, 2)]), VADD(T1j, T1k), ms, &(x[0])); } } } } } VLEAVE(); } static const tw_instr twinstr[] = { VTW(0, 1), VTW(0, 2), VTW(0, 3), VTW(0, 4), VTW(0, 5), VTW(0, 6), VTW(0, 7), VTW(0, 8), VTW(0, 9), VTW(0, 10), VTW(0, 11), VTW(0, 12), VTW(0, 13), VTW(0, 14), {TW_NEXT, VL, 0} }; static const ct_desc desc = { 15, XSIMD_STRING("t1fv_15"), twinstr, &GENUS, {78, 39, 14, 0}, 0, 0, 0 }; void XSIMD(codelet_t1fv_15) (planner *p) { X(kdft_dit_register) (p, t1fv_15, &desc); } #endif /* HAVE_FMA */ fftw-3.3.3/dft/simd/common/t1fv_16.c0000644000175000017500000003221512054410455013745 00000000000000/* * Copyright (c) 2003, 2007-11 Matteo Frigo * Copyright (c) 2003, 2007-11 Massachusetts Institute of Technology * * This program is free software; you can redistribute it and/or modify * it under the terms of the GNU General Public License as published by * the Free Software Foundation; either version 2 of the License, or * (at your option) any later version. * * This program is distributed in the hope that it will be useful, * but WITHOUT ANY WARRANTY; without even the implied warranty of * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the * GNU General Public License for more details. * * You should have received a copy of the GNU General Public License * along with this program; if not, write to the Free Software * Foundation, Inc., 51 Franklin Street, Fifth Floor, Boston, MA 02110-1301 USA * */ /* This file was automatically generated --- DO NOT EDIT */ /* Generated on Sun Nov 25 07:38:05 EST 2012 */ #include "codelet-dft.h" #ifdef HAVE_FMA /* Generated by: ../../../genfft/gen_twiddle_c.native -fma -reorder-insns -schedule-for-pipeline -simd -compact -variables 4 -pipeline-latency 8 -n 16 -name t1fv_16 -include t1f.h */ /* * This function contains 87 FP additions, 64 FP multiplications, * (or, 53 additions, 30 multiplications, 34 fused multiply/add), * 61 stack variables, 3 constants, and 32 memory accesses */ #include "t1f.h" static void t1fv_16(R *ri, R *ii, const R *W, stride rs, INT mb, INT me, INT ms) { DVK(KP923879532, +0.923879532511286756128183189396788286822416626); DVK(KP414213562, +0.414213562373095048801688724209698078569671875); DVK(KP707106781, +0.707106781186547524400844362104849039284835938); { INT m; R *x; x = ri; for (m = mb, W = W + (mb * ((TWVL / VL) * 30)); m < me; m = m + VL, x = x + (VL * ms), W = W + (TWVL * 30), MAKE_VOLATILE_STRIDE(16, rs)) { V TO, Ta, TJ, TP, T14, Tq, T1i, T10, T1b, T1l, T13, T1c, TR, Tl, T15; V Tv; { V Tc, TW, T4, T19, T9, TD, TI, Tj, TZ, T1a, Te, Th, Tn, Tr, Tu; V Tp; { V T1, T2, T5, T7; T1 = LD(&(x[0]), ms, &(x[0])); T2 = LD(&(x[WS(rs, 8)]), ms, &(x[0])); T5 = LD(&(x[WS(rs, 4)]), ms, &(x[0])); T7 = LD(&(x[WS(rs, 12)]), ms, &(x[0])); { V Tz, TG, TB, TE; Tz = LD(&(x[WS(rs, 14)]), ms, &(x[0])); TG = LD(&(x[WS(rs, 10)]), ms, &(x[0])); TB = LD(&(x[WS(rs, 6)]), ms, &(x[0])); TE = LD(&(x[WS(rs, 2)]), ms, &(x[0])); { V Ti, TY, TX, Td, Tg, Tm, Tt, To; { V T3, T6, T8, TA, TH, TC, TF, Tb; Tb = LD(&(x[WS(rs, 1)]), ms, &(x[WS(rs, 1)])); T3 = BYTWJ(&(W[TWVL * 14]), T2); T6 = BYTWJ(&(W[TWVL * 6]), T5); T8 = BYTWJ(&(W[TWVL * 22]), T7); TA = BYTWJ(&(W[TWVL * 26]), Tz); TH = BYTWJ(&(W[TWVL * 18]), TG); TC = BYTWJ(&(W[TWVL * 10]), TB); TF = BYTWJ(&(W[TWVL * 2]), TE); Tc = BYTWJ(&(W[0]), Tb); TW = VSUB(T1, T3); T4 = VADD(T1, T3); T19 = VSUB(T6, T8); T9 = VADD(T6, T8); Ti = LD(&(x[WS(rs, 13)]), ms, &(x[WS(rs, 1)])); TD = VADD(TA, TC); TY = VSUB(TA, TC); TI = VADD(TF, TH); TX = VSUB(TF, TH); } Td = LD(&(x[WS(rs, 9)]), ms, &(x[WS(rs, 1)])); Tg = LD(&(x[WS(rs, 5)]), ms, &(x[WS(rs, 1)])); Tm = LD(&(x[WS(rs, 15)]), ms, &(x[WS(rs, 1)])); Tj = BYTWJ(&(W[TWVL * 24]), Ti); Tt = LD(&(x[WS(rs, 11)]), ms, &(x[WS(rs, 1)])); To = LD(&(x[WS(rs, 7)]), ms, &(x[WS(rs, 1)])); TZ = VADD(TX, TY); T1a = VSUB(TY, TX); Te = BYTWJ(&(W[TWVL * 16]), Td); Th = BYTWJ(&(W[TWVL * 8]), Tg); Tn = BYTWJ(&(W[TWVL * 28]), Tm); Tr = LD(&(x[WS(rs, 3)]), ms, &(x[WS(rs, 1)])); Tu = BYTWJ(&(W[TWVL * 20]), Tt); Tp = BYTWJ(&(W[TWVL * 12]), To); } } } { V Tf, T11, Tk, T12, Ts; TO = VADD(T4, T9); Ta = VSUB(T4, T9); TJ = VSUB(TD, TI); TP = VADD(TI, TD); Tf = VADD(Tc, Te); T11 = VSUB(Tc, Te); Tk = VADD(Th, Tj); T12 = VSUB(Th, Tj); Ts = BYTWJ(&(W[TWVL * 4]), Tr); T14 = VSUB(Tn, Tp); Tq = VADD(Tn, Tp); T1i = VFNMS(LDK(KP707106781), TZ, TW); T10 = VFMA(LDK(KP707106781), TZ, TW); T1b = VFNMS(LDK(KP707106781), T1a, T19); T1l = VFMA(LDK(KP707106781), T1a, T19); T13 = VFNMS(LDK(KP414213562), T12, T11); T1c = VFMA(LDK(KP414213562), T11, T12); TR = VADD(Tf, Tk); Tl = VSUB(Tf, Tk); T15 = VSUB(Tu, Ts); Tv = VADD(Ts, Tu); } } { V T1d, T16, TS, Tw, TU, TQ; T1d = VFMA(LDK(KP414213562), T14, T15); T16 = VFNMS(LDK(KP414213562), T15, T14); TS = VADD(Tq, Tv); Tw = VSUB(Tq, Tv); TU = VSUB(TO, TP); TQ = VADD(TO, TP); { V T1e, T1j, T17, T1m; T1e = VSUB(T1c, T1d); T1j = VADD(T1c, T1d); T17 = VADD(T13, T16); T1m = VSUB(T16, T13); { V TV, TT, TK, Tx; TV = VSUB(TS, TR); TT = VADD(TR, TS); TK = VSUB(Tw, Tl); Tx = VADD(Tl, Tw); { V T1h, T1f, T1o, T1k; T1h = VFMA(LDK(KP923879532), T1e, T1b); T1f = VFNMS(LDK(KP923879532), T1e, T1b); T1o = VFMA(LDK(KP923879532), T1j, T1i); T1k = VFNMS(LDK(KP923879532), T1j, T1i); { V T1g, T18, T1p, T1n; T1g = VFMA(LDK(KP923879532), T17, T10); T18 = VFNMS(LDK(KP923879532), T17, T10); T1p = VFMA(LDK(KP923879532), T1m, T1l); T1n = VFNMS(LDK(KP923879532), T1m, T1l); ST(&(x[WS(rs, 12)]), VFNMSI(TV, TU), ms, &(x[0])); ST(&(x[WS(rs, 4)]), VFMAI(TV, TU), ms, &(x[0])); ST(&(x[0]), VADD(TQ, TT), ms, &(x[0])); ST(&(x[WS(rs, 8)]), VSUB(TQ, TT), ms, &(x[0])); { V TN, TL, TM, Ty; TN = VFMA(LDK(KP707106781), TK, TJ); TL = VFNMS(LDK(KP707106781), TK, TJ); TM = VFMA(LDK(KP707106781), Tx, Ta); Ty = VFNMS(LDK(KP707106781), Tx, Ta); ST(&(x[WS(rs, 1)]), VFNMSI(T1h, T1g), ms, &(x[WS(rs, 1)])); ST(&(x[WS(rs, 15)]), VFMAI(T1h, T1g), ms, &(x[WS(rs, 1)])); ST(&(x[WS(rs, 7)]), VFMAI(T1f, T18), ms, &(x[WS(rs, 1)])); ST(&(x[WS(rs, 9)]), VFNMSI(T1f, T18), ms, &(x[WS(rs, 1)])); ST(&(x[WS(rs, 3)]), VFMAI(T1p, T1o), ms, &(x[WS(rs, 1)])); ST(&(x[WS(rs, 13)]), VFNMSI(T1p, T1o), ms, &(x[WS(rs, 1)])); ST(&(x[WS(rs, 11)]), VFMAI(T1n, T1k), ms, &(x[WS(rs, 1)])); ST(&(x[WS(rs, 5)]), VFNMSI(T1n, T1k), ms, &(x[WS(rs, 1)])); ST(&(x[WS(rs, 14)]), VFNMSI(TN, TM), ms, &(x[0])); ST(&(x[WS(rs, 2)]), VFMAI(TN, TM), ms, &(x[0])); ST(&(x[WS(rs, 10)]), VFMAI(TL, Ty), ms, &(x[0])); ST(&(x[WS(rs, 6)]), VFNMSI(TL, Ty), ms, &(x[0])); } } } } } } } } VLEAVE(); } static const tw_instr twinstr[] = { VTW(0, 1), VTW(0, 2), VTW(0, 3), VTW(0, 4), VTW(0, 5), VTW(0, 6), VTW(0, 7), VTW(0, 8), VTW(0, 9), VTW(0, 10), VTW(0, 11), VTW(0, 12), VTW(0, 13), VTW(0, 14), VTW(0, 15), {TW_NEXT, VL, 0} }; static const ct_desc desc = { 16, XSIMD_STRING("t1fv_16"), twinstr, &GENUS, {53, 30, 34, 0}, 0, 0, 0 }; void XSIMD(codelet_t1fv_16) (planner *p) { X(kdft_dit_register) (p, t1fv_16, &desc); } #else /* HAVE_FMA */ /* Generated by: ../../../genfft/gen_twiddle_c.native -simd -compact -variables 4 -pipeline-latency 8 -n 16 -name t1fv_16 -include t1f.h */ /* * This function contains 87 FP additions, 42 FP multiplications, * (or, 83 additions, 38 multiplications, 4 fused multiply/add), * 36 stack variables, 3 constants, and 32 memory accesses */ #include "t1f.h" static void t1fv_16(R *ri, R *ii, const R *W, stride rs, INT mb, INT me, INT ms) { DVK(KP923879532, +0.923879532511286756128183189396788286822416626); DVK(KP382683432, +0.382683432365089771728459984030398866761344562); DVK(KP707106781, +0.707106781186547524400844362104849039284835938); { INT m; R *x; x = ri; for (m = mb, W = W + (mb * ((TWVL / VL) * 30)); m < me; m = m + VL, x = x + (VL * ms), W = W + (TWVL * 30), MAKE_VOLATILE_STRIDE(16, rs)) { V TJ, T10, TD, T11, T1b, T1c, Ty, TK, T16, T17, T18, Tb, TN, T13, T14; V T15, Tm, TM, TG, TI, TH; TG = LD(&(x[0]), ms, &(x[0])); TH = LD(&(x[WS(rs, 8)]), ms, &(x[0])); TI = BYTWJ(&(W[TWVL * 14]), TH); TJ = VSUB(TG, TI); T10 = VADD(TG, TI); { V TA, TC, Tz, TB; Tz = LD(&(x[WS(rs, 4)]), ms, &(x[0])); TA = BYTWJ(&(W[TWVL * 6]), Tz); TB = LD(&(x[WS(rs, 12)]), ms, &(x[0])); TC = BYTWJ(&(W[TWVL * 22]), TB); TD = VSUB(TA, TC); T11 = VADD(TA, TC); } { V Tp, Tw, Tr, Tu, Ts, Tx; { V To, Tv, Tq, Tt; To = LD(&(x[WS(rs, 14)]), ms, &(x[0])); Tp = BYTWJ(&(W[TWVL * 26]), To); Tv = LD(&(x[WS(rs, 10)]), ms, &(x[0])); Tw = BYTWJ(&(W[TWVL * 18]), Tv); Tq = LD(&(x[WS(rs, 6)]), ms, &(x[0])); Tr = BYTWJ(&(W[TWVL * 10]), Tq); Tt = LD(&(x[WS(rs, 2)]), ms, &(x[0])); Tu = BYTWJ(&(W[TWVL * 2]), Tt); } T1b = VADD(Tp, Tr); T1c = VADD(Tu, Tw); Ts = VSUB(Tp, Tr); Tx = VSUB(Tu, Tw); Ty = VMUL(LDK(KP707106781), VSUB(Ts, Tx)); TK = VMUL(LDK(KP707106781), VADD(Tx, Ts)); } { V T2, T9, T4, T7, T5, Ta; { V T1, T8, T3, T6; T1 = LD(&(x[WS(rs, 15)]), ms, &(x[WS(rs, 1)])); T2 = BYTWJ(&(W[TWVL * 28]), T1); T8 = LD(&(x[WS(rs, 11)]), ms, &(x[WS(rs, 1)])); T9 = BYTWJ(&(W[TWVL * 20]), T8); T3 = LD(&(x[WS(rs, 7)]), ms, &(x[WS(rs, 1)])); T4 = BYTWJ(&(W[TWVL * 12]), T3); T6 = LD(&(x[WS(rs, 3)]), ms, &(x[WS(rs, 1)])); T7 = BYTWJ(&(W[TWVL * 4]), T6); } T16 = VADD(T2, T4); T17 = VADD(T7, T9); T18 = VSUB(T16, T17); T5 = VSUB(T2, T4); Ta = VSUB(T7, T9); Tb = VFNMS(LDK(KP923879532), Ta, VMUL(LDK(KP382683432), T5)); TN = VFMA(LDK(KP923879532), T5, VMUL(LDK(KP382683432), Ta)); } { V Td, Tk, Tf, Ti, Tg, Tl; { V Tc, Tj, Te, Th; Tc = LD(&(x[WS(rs, 1)]), ms, &(x[WS(rs, 1)])); Td = BYTWJ(&(W[0]), Tc); Tj = LD(&(x[WS(rs, 13)]), ms, &(x[WS(rs, 1)])); Tk = BYTWJ(&(W[TWVL * 24]), Tj); Te = LD(&(x[WS(rs, 9)]), ms, &(x[WS(rs, 1)])); Tf = BYTWJ(&(W[TWVL * 16]), Te); Th = LD(&(x[WS(rs, 5)]), ms, &(x[WS(rs, 1)])); Ti = BYTWJ(&(W[TWVL * 8]), Th); } T13 = VADD(Td, Tf); T14 = VADD(Ti, Tk); T15 = VSUB(T13, T14); Tg = VSUB(Td, Tf); Tl = VSUB(Ti, Tk); Tm = VFMA(LDK(KP382683432), Tg, VMUL(LDK(KP923879532), Tl)); TM = VFNMS(LDK(KP382683432), Tl, VMUL(LDK(KP923879532), Tg)); } { V T1a, T1g, T1f, T1h; { V T12, T19, T1d, T1e; T12 = VSUB(T10, T11); T19 = VMUL(LDK(KP707106781), VADD(T15, T18)); T1a = VADD(T12, T19); T1g = VSUB(T12, T19); T1d = VSUB(T1b, T1c); T1e = VMUL(LDK(KP707106781), VSUB(T18, T15)); T1f = VBYI(VADD(T1d, T1e)); T1h = VBYI(VSUB(T1e, T1d)); } ST(&(x[WS(rs, 14)]), VSUB(T1a, T1f), ms, &(x[0])); ST(&(x[WS(rs, 6)]), VADD(T1g, T1h), ms, &(x[0])); ST(&(x[WS(rs, 2)]), VADD(T1a, T1f), ms, &(x[0])); ST(&(x[WS(rs, 10)]), VSUB(T1g, T1h), ms, &(x[0])); } { V T1k, T1o, T1n, T1p; { V T1i, T1j, T1l, T1m; T1i = VADD(T10, T11); T1j = VADD(T1c, T1b); T1k = VADD(T1i, T1j); T1o = VSUB(T1i, T1j); T1l = VADD(T13, T14); T1m = VADD(T16, T17); T1n = VADD(T1l, T1m); T1p = VBYI(VSUB(T1m, T1l)); } ST(&(x[WS(rs, 8)]), VSUB(T1k, T1n), ms, &(x[0])); ST(&(x[WS(rs, 4)]), VADD(T1o, T1p), ms, &(x[0])); ST(&(x[0]), VADD(T1k, T1n), ms, &(x[0])); ST(&(x[WS(rs, 12)]), VSUB(T1o, T1p), ms, &(x[0])); } { V TF, TQ, TP, TR; { V Tn, TE, TL, TO; Tn = VSUB(Tb, Tm); TE = VSUB(Ty, TD); TF = VBYI(VSUB(Tn, TE)); TQ = VBYI(VADD(TE, Tn)); TL = VADD(TJ, TK); TO = VADD(TM, TN); TP = VSUB(TL, TO); TR = VADD(TL, TO); } ST(&(x[WS(rs, 7)]), VADD(TF, TP), ms, &(x[WS(rs, 1)])); ST(&(x[WS(rs, 15)]), VSUB(TR, TQ), ms, &(x[WS(rs, 1)])); ST(&(x[WS(rs, 9)]), VSUB(TP, TF), ms, &(x[WS(rs, 1)])); ST(&(x[WS(rs, 1)]), VADD(TQ, TR), ms, &(x[WS(rs, 1)])); } { V TU, TY, TX, TZ; { V TS, TT, TV, TW; TS = VSUB(TJ, TK); TT = VADD(Tm, Tb); TU = VADD(TS, TT); TY = VSUB(TS, TT); TV = VADD(TD, Ty); TW = VSUB(TN, TM); TX = VBYI(VADD(TV, TW)); TZ = VBYI(VSUB(TW, TV)); } ST(&(x[WS(rs, 13)]), VSUB(TU, TX), ms, &(x[WS(rs, 1)])); ST(&(x[WS(rs, 5)]), VADD(TY, TZ), ms, &(x[WS(rs, 1)])); ST(&(x[WS(rs, 3)]), VADD(TU, TX), ms, &(x[WS(rs, 1)])); ST(&(x[WS(rs, 11)]), VSUB(TY, TZ), ms, &(x[WS(rs, 1)])); } } } VLEAVE(); } static const tw_instr twinstr[] = { VTW(0, 1), VTW(0, 2), VTW(0, 3), VTW(0, 4), VTW(0, 5), VTW(0, 6), VTW(0, 7), VTW(0, 8), VTW(0, 9), VTW(0, 10), VTW(0, 11), VTW(0, 12), VTW(0, 13), VTW(0, 14), VTW(0, 15), {TW_NEXT, VL, 0} }; static const ct_desc desc = { 16, XSIMD_STRING("t1fv_16"), twinstr, &GENUS, {83, 38, 4, 0}, 0, 0, 0 }; void XSIMD(codelet_t1fv_16) (planner *p) { X(kdft_dit_register) (p, t1fv_16, &desc); } #endif /* HAVE_FMA */ fftw-3.3.3/dft/simd/common/t1fv_32.c0000644000175000017500000007041312054410462013743 00000000000000/* * Copyright (c) 2003, 2007-11 Matteo Frigo * Copyright (c) 2003, 2007-11 Massachusetts Institute of Technology * * This program is free software; you can redistribute it and/or modify * it under the terms of the GNU General Public License as published by * the Free Software Foundation; either version 2 of the License, or * (at your option) any later version. * * This program is distributed in the hope that it will be useful, * but WITHOUT ANY WARRANTY; without even the implied warranty of * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the * GNU General Public License for more details. * * You should have received a copy of the GNU General Public License * along with this program; if not, write to the Free Software * Foundation, Inc., 51 Franklin Street, Fifth Floor, Boston, MA 02110-1301 USA * */ /* This file was automatically generated --- DO NOT EDIT */ /* Generated on Sun Nov 25 07:38:06 EST 2012 */ #include "codelet-dft.h" #ifdef HAVE_FMA /* Generated by: ../../../genfft/gen_twiddle_c.native -fma -reorder-insns -schedule-for-pipeline -simd -compact -variables 4 -pipeline-latency 8 -n 32 -name t1fv_32 -include t1f.h */ /* * This function contains 217 FP additions, 160 FP multiplications, * (or, 119 additions, 62 multiplications, 98 fused multiply/add), * 112 stack variables, 7 constants, and 64 memory accesses */ #include "t1f.h" static void t1fv_32(R *ri, R *ii, const R *W, stride rs, INT mb, INT me, INT ms) { DVK(KP831469612, +0.831469612302545237078788377617905756738560812); DVK(KP980785280, +0.980785280403230449126182236134239036973933731); DVK(KP668178637, +0.668178637919298919997757686523080761552472251); DVK(KP198912367, +0.198912367379658006911597622644676228597850501); DVK(KP923879532, +0.923879532511286756128183189396788286822416626); DVK(KP707106781, +0.707106781186547524400844362104849039284835938); DVK(KP414213562, +0.414213562373095048801688724209698078569671875); { INT m; R *x; x = ri; for (m = mb, W = W + (mb * ((TWVL / VL) * 62)); m < me; m = m + VL, x = x + (VL * ms), W = W + (TWVL * 62), MAKE_VOLATILE_STRIDE(32, rs)) { V T26, T25, T1Z, T22, T1W, T2a, T2k, T2g; { V T4, T1z, T2o, T32, T2r, T3f, Tf, T1A, T34, T2L, T1D, TC, T33, T2O, T1C; V Tr, T2C, T3a, T2F, T3b, T1r, T21, T1k, T20, TQ, TM, TS, TL, T2t, TJ; V T10, T2u; { V Tt, T9, T2p, Te, T2q, TA, Tu, Tx; { V T1, T1x, T2, T1v; T1 = LD(&(x[0]), ms, &(x[0])); T1x = LD(&(x[WS(rs, 24)]), ms, &(x[0])); T2 = LD(&(x[WS(rs, 16)]), ms, &(x[0])); T1v = LD(&(x[WS(rs, 8)]), ms, &(x[0])); { V T5, Tc, T7, Ta, T2m, T2n; T5 = LD(&(x[WS(rs, 4)]), ms, &(x[0])); Tc = LD(&(x[WS(rs, 12)]), ms, &(x[0])); T7 = LD(&(x[WS(rs, 20)]), ms, &(x[0])); Ta = LD(&(x[WS(rs, 28)]), ms, &(x[0])); { V T1y, T3, T1w, T6, Td, T8, Tb, Ts, Tz; Ts = LD(&(x[WS(rs, 30)]), ms, &(x[0])); T1y = BYTWJ(&(W[TWVL * 46]), T1x); T3 = BYTWJ(&(W[TWVL * 30]), T2); T1w = BYTWJ(&(W[TWVL * 14]), T1v); T6 = BYTWJ(&(W[TWVL * 6]), T5); Td = BYTWJ(&(W[TWVL * 22]), Tc); T8 = BYTWJ(&(W[TWVL * 38]), T7); Tb = BYTWJ(&(W[TWVL * 54]), Ta); Tt = BYTWJ(&(W[TWVL * 58]), Ts); Tz = LD(&(x[WS(rs, 6)]), ms, &(x[0])); T4 = VSUB(T1, T3); T2m = VADD(T1, T3); T1z = VSUB(T1w, T1y); T2n = VADD(T1w, T1y); T9 = VSUB(T6, T8); T2p = VADD(T6, T8); Te = VSUB(Tb, Td); T2q = VADD(Tb, Td); TA = BYTWJ(&(W[TWVL * 10]), Tz); } Tu = LD(&(x[WS(rs, 14)]), ms, &(x[0])); T2o = VADD(T2m, T2n); T32 = VSUB(T2m, T2n); Tx = LD(&(x[WS(rs, 22)]), ms, &(x[0])); } } { V Tv, To, Ty, Ti, Tj, Tm, Th; Th = LD(&(x[WS(rs, 2)]), ms, &(x[0])); T2r = VADD(T2p, T2q); T3f = VSUB(T2q, T2p); Tf = VADD(T9, Te); T1A = VSUB(Te, T9); Tv = BYTWJ(&(W[TWVL * 26]), Tu); To = LD(&(x[WS(rs, 26)]), ms, &(x[0])); Ty = BYTWJ(&(W[TWVL * 42]), Tx); Ti = BYTWJ(&(W[TWVL * 2]), Th); Tj = LD(&(x[WS(rs, 18)]), ms, &(x[0])); Tm = LD(&(x[WS(rs, 10)]), ms, &(x[0])); { V T1f, T1h, T1a, T1c, T18, T2A, T2B, T1p; { V T15, T17, T1o, T1m; { V Tw, T2J, Tp, T2K, TB, Tk, Tn, T1n, T14, T16; T14 = LD(&(x[WS(rs, 31)]), ms, &(x[WS(rs, 1)])); T16 = LD(&(x[WS(rs, 15)]), ms, &(x[WS(rs, 1)])); Tw = VSUB(Tt, Tv); T2J = VADD(Tt, Tv); Tp = BYTWJ(&(W[TWVL * 50]), To); T2K = VADD(TA, Ty); TB = VSUB(Ty, TA); Tk = BYTWJ(&(W[TWVL * 34]), Tj); Tn = BYTWJ(&(W[TWVL * 18]), Tm); T15 = BYTWJ(&(W[TWVL * 60]), T14); T17 = BYTWJ(&(W[TWVL * 28]), T16); T1n = LD(&(x[WS(rs, 7)]), ms, &(x[WS(rs, 1)])); { V T2M, Tl, T2N, Tq, T1l; T1l = LD(&(x[WS(rs, 23)]), ms, &(x[WS(rs, 1)])); T34 = VSUB(T2J, T2K); T2L = VADD(T2J, T2K); T1D = VFMA(LDK(KP414213562), Tw, TB); TC = VFNMS(LDK(KP414213562), TB, Tw); T2M = VADD(Ti, Tk); Tl = VSUB(Ti, Tk); T2N = VADD(Tn, Tp); Tq = VSUB(Tn, Tp); T1o = BYTWJ(&(W[TWVL * 12]), T1n); T1m = BYTWJ(&(W[TWVL * 44]), T1l); { V T1e, T1g, T19, T1b; T1e = LD(&(x[WS(rs, 27)]), ms, &(x[WS(rs, 1)])); T1g = LD(&(x[WS(rs, 11)]), ms, &(x[WS(rs, 1)])); T19 = LD(&(x[WS(rs, 3)]), ms, &(x[WS(rs, 1)])); T1b = LD(&(x[WS(rs, 19)]), ms, &(x[WS(rs, 1)])); T33 = VSUB(T2M, T2N); T2O = VADD(T2M, T2N); T1C = VFMA(LDK(KP414213562), Tl, Tq); Tr = VFNMS(LDK(KP414213562), Tq, Tl); T1f = BYTWJ(&(W[TWVL * 52]), T1e); T1h = BYTWJ(&(W[TWVL * 20]), T1g); T1a = BYTWJ(&(W[TWVL * 4]), T19); T1c = BYTWJ(&(W[TWVL * 36]), T1b); } } } T18 = VSUB(T15, T17); T2A = VADD(T15, T17); T2B = VADD(T1o, T1m); T1p = VSUB(T1m, T1o); } { V TG, TI, TZ, TX; { V T1i, T2E, T1d, T2D, TH, TY, TF; TF = LD(&(x[WS(rs, 1)]), ms, &(x[WS(rs, 1)])); T1i = VSUB(T1f, T1h); T2E = VADD(T1f, T1h); T1d = VSUB(T1a, T1c); T2D = VADD(T1a, T1c); TH = LD(&(x[WS(rs, 17)]), ms, &(x[WS(rs, 1)])); TY = LD(&(x[WS(rs, 25)]), ms, &(x[WS(rs, 1)])); T2C = VADD(T2A, T2B); T3a = VSUB(T2A, T2B); TG = BYTWJ(&(W[0]), TF); { V TW, T1j, T1q, TP, TR, TK; TW = LD(&(x[WS(rs, 9)]), ms, &(x[WS(rs, 1)])); T2F = VADD(T2D, T2E); T3b = VSUB(T2E, T2D); T1j = VADD(T1d, T1i); T1q = VSUB(T1i, T1d); TI = BYTWJ(&(W[TWVL * 32]), TH); TZ = BYTWJ(&(W[TWVL * 48]), TY); TP = LD(&(x[WS(rs, 29)]), ms, &(x[WS(rs, 1)])); TX = BYTWJ(&(W[TWVL * 16]), TW); TR = LD(&(x[WS(rs, 13)]), ms, &(x[WS(rs, 1)])); TK = LD(&(x[WS(rs, 5)]), ms, &(x[WS(rs, 1)])); T1r = VFMA(LDK(KP707106781), T1q, T1p); T21 = VFNMS(LDK(KP707106781), T1q, T1p); T1k = VFMA(LDK(KP707106781), T1j, T18); T20 = VFNMS(LDK(KP707106781), T1j, T18); TQ = BYTWJ(&(W[TWVL * 56]), TP); TM = LD(&(x[WS(rs, 21)]), ms, &(x[WS(rs, 1)])); TS = BYTWJ(&(W[TWVL * 24]), TR); TL = BYTWJ(&(W[TWVL * 8]), TK); } } T2t = VADD(TG, TI); TJ = VSUB(TG, TI); T10 = VSUB(TX, TZ); T2u = VADD(TX, TZ); } } } } { V T2s, TT, T2x, T2P, T2Y, T2G, T37, T2v, T2w, TO, T2W, T30, T2U, TN, T2V; T2s = VSUB(T2o, T2r); T2U = VADD(T2o, T2r); TN = BYTWJ(&(W[TWVL * 40]), TM); TT = VSUB(TQ, TS); T2x = VADD(TQ, TS); T2P = VSUB(T2L, T2O); T2V = VADD(T2O, T2L); T2Y = VADD(T2C, T2F); T2G = VSUB(T2C, T2F); T37 = VSUB(T2t, T2u); T2v = VADD(T2t, T2u); T2w = VADD(TL, TN); TO = VSUB(TL, TN); T2W = VADD(T2U, T2V); T30 = VSUB(T2U, T2V); { V T3i, T3o, T36, T3r, T3h, T3j, T12, T1Y, TV, T1X, T3s, T3d, T2Q, T2H, T31; V T2Z; { V T35, T3g, T38, T2y, T11, TU; T35 = VADD(T33, T34); T3g = VSUB(T34, T33); T38 = VSUB(T2w, T2x); T2y = VADD(T2w, T2x); T11 = VSUB(TO, TT); TU = VADD(TO, TT); { V T3c, T39, T2X, T2z; T3c = VFNMS(LDK(KP414213562), T3b, T3a); T3i = VFMA(LDK(KP414213562), T3a, T3b); T3o = VFNMS(LDK(KP707106781), T35, T32); T36 = VFMA(LDK(KP707106781), T35, T32); T3r = VFNMS(LDK(KP707106781), T3g, T3f); T3h = VFMA(LDK(KP707106781), T3g, T3f); T39 = VFNMS(LDK(KP414213562), T38, T37); T3j = VFMA(LDK(KP414213562), T37, T38); T2X = VADD(T2v, T2y); T2z = VSUB(T2v, T2y); T12 = VFMA(LDK(KP707106781), T11, T10); T1Y = VFNMS(LDK(KP707106781), T11, T10); TV = VFMA(LDK(KP707106781), TU, TJ); T1X = VFNMS(LDK(KP707106781), TU, TJ); T3s = VSUB(T3c, T39); T3d = VADD(T39, T3c); T2Q = VSUB(T2G, T2z); T2H = VADD(T2z, T2G); T31 = VSUB(T2Y, T2X); T2Z = VADD(T2X, T2Y); } } { V Tg, T1U, TD, T1G, T13, T1s, T1H, T1B, T1V, T1E, T3k, T3p, T2e, T2f; Tg = VFMA(LDK(KP707106781), Tf, T4); T1U = VFNMS(LDK(KP707106781), Tf, T4); T3k = VSUB(T3i, T3j); T3p = VADD(T3j, T3i); { V T3v, T3t, T3e, T3m; T3v = VFNMS(LDK(KP923879532), T3s, T3r); T3t = VFMA(LDK(KP923879532), T3s, T3r); T3e = VFNMS(LDK(KP923879532), T3d, T36); T3m = VFMA(LDK(KP923879532), T3d, T36); { V T2R, T2T, T2I, T2S; T2R = VFNMS(LDK(KP707106781), T2Q, T2P); T2T = VFMA(LDK(KP707106781), T2Q, T2P); T2I = VFNMS(LDK(KP707106781), T2H, T2s); T2S = VFMA(LDK(KP707106781), T2H, T2s); ST(&(x[WS(rs, 24)]), VFNMSI(T31, T30), ms, &(x[0])); ST(&(x[WS(rs, 8)]), VFMAI(T31, T30), ms, &(x[0])); ST(&(x[0]), VADD(T2W, T2Z), ms, &(x[0])); ST(&(x[WS(rs, 16)]), VSUB(T2W, T2Z), ms, &(x[0])); { V T3u, T3q, T3l, T3n; T3u = VFMA(LDK(KP923879532), T3p, T3o); T3q = VFNMS(LDK(KP923879532), T3p, T3o); T3l = VFNMS(LDK(KP923879532), T3k, T3h); T3n = VFMA(LDK(KP923879532), T3k, T3h); ST(&(x[WS(rs, 4)]), VFMAI(T2T, T2S), ms, &(x[0])); ST(&(x[WS(rs, 28)]), VFNMSI(T2T, T2S), ms, &(x[0])); ST(&(x[WS(rs, 20)]), VFMAI(T2R, T2I), ms, &(x[0])); ST(&(x[WS(rs, 12)]), VFNMSI(T2R, T2I), ms, &(x[0])); ST(&(x[WS(rs, 22)]), VFNMSI(T3t, T3q), ms, &(x[0])); ST(&(x[WS(rs, 10)]), VFMAI(T3t, T3q), ms, &(x[0])); ST(&(x[WS(rs, 26)]), VFMAI(T3v, T3u), ms, &(x[0])); ST(&(x[WS(rs, 6)]), VFNMSI(T3v, T3u), ms, &(x[0])); ST(&(x[WS(rs, 2)]), VFMAI(T3n, T3m), ms, &(x[0])); ST(&(x[WS(rs, 30)]), VFNMSI(T3n, T3m), ms, &(x[0])); ST(&(x[WS(rs, 18)]), VFMAI(T3l, T3e), ms, &(x[0])); ST(&(x[WS(rs, 14)]), VFNMSI(T3l, T3e), ms, &(x[0])); T26 = VSUB(TC, Tr); TD = VADD(Tr, TC); } } } T1G = VFMA(LDK(KP198912367), TV, T12); T13 = VFNMS(LDK(KP198912367), T12, TV); T1s = VFNMS(LDK(KP198912367), T1r, T1k); T1H = VFMA(LDK(KP198912367), T1k, T1r); T1B = VFNMS(LDK(KP707106781), T1A, T1z); T25 = VFMA(LDK(KP707106781), T1A, T1z); T1V = VADD(T1C, T1D); T1E = VSUB(T1C, T1D); { V T1S, T1O, T1K, T1u, T1R, T1T, T1L, T1J; { V TE, T1M, T1I, T1N, T1t, T1Q, T1F, T1P, T28, T29; TE = VFMA(LDK(KP923879532), TD, Tg); T1M = VFNMS(LDK(KP923879532), TD, Tg); T1I = VSUB(T1G, T1H); T1N = VADD(T1G, T1H); T1t = VADD(T13, T1s); T1Q = VSUB(T1s, T13); T1F = VFMA(LDK(KP923879532), T1E, T1B); T1P = VFNMS(LDK(KP923879532), T1E, T1B); T28 = VFNMS(LDK(KP668178637), T1X, T1Y); T1Z = VFMA(LDK(KP668178637), T1Y, T1X); T1S = VFMA(LDK(KP980785280), T1N, T1M); T1O = VFNMS(LDK(KP980785280), T1N, T1M); T22 = VFMA(LDK(KP668178637), T21, T20); T29 = VFNMS(LDK(KP668178637), T20, T21); T1K = VFMA(LDK(KP980785280), T1t, TE); T1u = VFNMS(LDK(KP980785280), T1t, TE); T1R = VFNMS(LDK(KP980785280), T1Q, T1P); T1T = VFMA(LDK(KP980785280), T1Q, T1P); T1L = VFMA(LDK(KP980785280), T1I, T1F); T1J = VFNMS(LDK(KP980785280), T1I, T1F); T2e = VFNMS(LDK(KP923879532), T1V, T1U); T1W = VFMA(LDK(KP923879532), T1V, T1U); T2a = VSUB(T28, T29); T2f = VADD(T28, T29); } ST(&(x[WS(rs, 23)]), VFMAI(T1R, T1O), ms, &(x[WS(rs, 1)])); ST(&(x[WS(rs, 9)]), VFNMSI(T1R, T1O), ms, &(x[WS(rs, 1)])); ST(&(x[WS(rs, 25)]), VFNMSI(T1T, T1S), ms, &(x[WS(rs, 1)])); ST(&(x[WS(rs, 7)]), VFMAI(T1T, T1S), ms, &(x[WS(rs, 1)])); ST(&(x[WS(rs, 31)]), VFMAI(T1L, T1K), ms, &(x[WS(rs, 1)])); ST(&(x[WS(rs, 1)]), VFNMSI(T1L, T1K), ms, &(x[WS(rs, 1)])); ST(&(x[WS(rs, 15)]), VFMAI(T1J, T1u), ms, &(x[WS(rs, 1)])); ST(&(x[WS(rs, 17)]), VFNMSI(T1J, T1u), ms, &(x[WS(rs, 1)])); } T2k = VFNMS(LDK(KP831469612), T2f, T2e); T2g = VFMA(LDK(KP831469612), T2f, T2e); } } } } { V T2i, T23, T2h, T27; T2i = VSUB(T22, T1Z); T23 = VADD(T1Z, T22); T2h = VFNMS(LDK(KP923879532), T26, T25); T27 = VFMA(LDK(KP923879532), T26, T25); { V T2c, T24, T2j, T2l, T2d, T2b; T2c = VFMA(LDK(KP831469612), T23, T1W); T24 = VFNMS(LDK(KP831469612), T23, T1W); T2j = VFMA(LDK(KP831469612), T2i, T2h); T2l = VFNMS(LDK(KP831469612), T2i, T2h); T2d = VFMA(LDK(KP831469612), T2a, T27); T2b = VFNMS(LDK(KP831469612), T2a, T27); ST(&(x[WS(rs, 21)]), VFNMSI(T2j, T2g), ms, &(x[WS(rs, 1)])); ST(&(x[WS(rs, 11)]), VFMAI(T2j, T2g), ms, &(x[WS(rs, 1)])); ST(&(x[WS(rs, 27)]), VFMAI(T2l, T2k), ms, &(x[WS(rs, 1)])); ST(&(x[WS(rs, 5)]), VFNMSI(T2l, T2k), ms, &(x[WS(rs, 1)])); ST(&(x[WS(rs, 3)]), VFMAI(T2d, T2c), ms, &(x[WS(rs, 1)])); ST(&(x[WS(rs, 29)]), VFNMSI(T2d, T2c), ms, &(x[WS(rs, 1)])); ST(&(x[WS(rs, 19)]), VFMAI(T2b, T24), ms, &(x[WS(rs, 1)])); ST(&(x[WS(rs, 13)]), VFNMSI(T2b, T24), ms, &(x[WS(rs, 1)])); } } } } VLEAVE(); } static const tw_instr twinstr[] = { VTW(0, 1), VTW(0, 2), VTW(0, 3), VTW(0, 4), VTW(0, 5), VTW(0, 6), VTW(0, 7), VTW(0, 8), VTW(0, 9), VTW(0, 10), VTW(0, 11), VTW(0, 12), VTW(0, 13), VTW(0, 14), VTW(0, 15), VTW(0, 16), VTW(0, 17), VTW(0, 18), VTW(0, 19), VTW(0, 20), VTW(0, 21), VTW(0, 22), VTW(0, 23), VTW(0, 24), VTW(0, 25), VTW(0, 26), VTW(0, 27), VTW(0, 28), VTW(0, 29), VTW(0, 30), VTW(0, 31), {TW_NEXT, VL, 0} }; static const ct_desc desc = { 32, XSIMD_STRING("t1fv_32"), twinstr, &GENUS, {119, 62, 98, 0}, 0, 0, 0 }; void XSIMD(codelet_t1fv_32) (planner *p) { X(kdft_dit_register) (p, t1fv_32, &desc); } #else /* HAVE_FMA */ /* Generated by: ../../../genfft/gen_twiddle_c.native -simd -compact -variables 4 -pipeline-latency 8 -n 32 -name t1fv_32 -include t1f.h */ /* * This function contains 217 FP additions, 104 FP multiplications, * (or, 201 additions, 88 multiplications, 16 fused multiply/add), * 59 stack variables, 7 constants, and 64 memory accesses */ #include "t1f.h" static void t1fv_32(R *ri, R *ii, const R *W, stride rs, INT mb, INT me, INT ms) { DVK(KP555570233, +0.555570233019602224742830813948532874374937191); DVK(KP831469612, +0.831469612302545237078788377617905756738560812); DVK(KP195090322, +0.195090322016128267848284868477022240927691618); DVK(KP980785280, +0.980785280403230449126182236134239036973933731); DVK(KP382683432, +0.382683432365089771728459984030398866761344562); DVK(KP923879532, +0.923879532511286756128183189396788286822416626); DVK(KP707106781, +0.707106781186547524400844362104849039284835938); { INT m; R *x; x = ri; for (m = mb, W = W + (mb * ((TWVL / VL) * 62)); m < me; m = m + VL, x = x + (VL * ms), W = W + (TWVL * 62), MAKE_VOLATILE_STRIDE(32, rs)) { V T4, T1A, T2o, T32, Tf, T1v, T2r, T3f, TC, T1C, T2L, T34, Tr, T1D, T2O; V T33, T1k, T20, T2F, T3b, T1r, T21, T2C, T3a, TV, T1X, T2y, T38, T12, T1Y; V T2v, T37; { V T1, T1z, T3, T1x, T1y, T2, T1w, T2m, T2n; T1 = LD(&(x[0]), ms, &(x[0])); T1y = LD(&(x[WS(rs, 24)]), ms, &(x[0])); T1z = BYTWJ(&(W[TWVL * 46]), T1y); T2 = LD(&(x[WS(rs, 16)]), ms, &(x[0])); T3 = BYTWJ(&(W[TWVL * 30]), T2); T1w = LD(&(x[WS(rs, 8)]), ms, &(x[0])); T1x = BYTWJ(&(W[TWVL * 14]), T1w); T4 = VSUB(T1, T3); T1A = VSUB(T1x, T1z); T2m = VADD(T1, T3); T2n = VADD(T1x, T1z); T2o = VADD(T2m, T2n); T32 = VSUB(T2m, T2n); } { V T6, Td, T8, Tb; { V T5, Tc, T7, Ta; T5 = LD(&(x[WS(rs, 4)]), ms, &(x[0])); T6 = BYTWJ(&(W[TWVL * 6]), T5); Tc = LD(&(x[WS(rs, 12)]), ms, &(x[0])); Td = BYTWJ(&(W[TWVL * 22]), Tc); T7 = LD(&(x[WS(rs, 20)]), ms, &(x[0])); T8 = BYTWJ(&(W[TWVL * 38]), T7); Ta = LD(&(x[WS(rs, 28)]), ms, &(x[0])); Tb = BYTWJ(&(W[TWVL * 54]), Ta); } { V T9, Te, T2p, T2q; T9 = VSUB(T6, T8); Te = VSUB(Tb, Td); Tf = VMUL(LDK(KP707106781), VADD(T9, Te)); T1v = VMUL(LDK(KP707106781), VSUB(Te, T9)); T2p = VADD(T6, T8); T2q = VADD(Tb, Td); T2r = VADD(T2p, T2q); T3f = VSUB(T2q, T2p); } } { V Tt, TA, Tv, Ty; { V Ts, Tz, Tu, Tx; Ts = LD(&(x[WS(rs, 30)]), ms, &(x[0])); Tt = BYTWJ(&(W[TWVL * 58]), Ts); Tz = LD(&(x[WS(rs, 22)]), ms, &(x[0])); TA = BYTWJ(&(W[TWVL * 42]), Tz); Tu = LD(&(x[WS(rs, 14)]), ms, &(x[0])); Tv = BYTWJ(&(W[TWVL * 26]), Tu); Tx = LD(&(x[WS(rs, 6)]), ms, &(x[0])); Ty = BYTWJ(&(W[TWVL * 10]), Tx); } { V Tw, TB, T2J, T2K; Tw = VSUB(Tt, Tv); TB = VSUB(Ty, TA); TC = VFMA(LDK(KP923879532), Tw, VMUL(LDK(KP382683432), TB)); T1C = VFNMS(LDK(KP923879532), TB, VMUL(LDK(KP382683432), Tw)); T2J = VADD(Tt, Tv); T2K = VADD(Ty, TA); T2L = VADD(T2J, T2K); T34 = VSUB(T2J, T2K); } } { V Ti, Tp, Tk, Tn; { V Th, To, Tj, Tm; Th = LD(&(x[WS(rs, 2)]), ms, &(x[0])); Ti = BYTWJ(&(W[TWVL * 2]), Th); To = LD(&(x[WS(rs, 26)]), ms, &(x[0])); Tp = BYTWJ(&(W[TWVL * 50]), To); Tj = LD(&(x[WS(rs, 18)]), ms, &(x[0])); Tk = BYTWJ(&(W[TWVL * 34]), Tj); Tm = LD(&(x[WS(rs, 10)]), ms, &(x[0])); Tn = BYTWJ(&(W[TWVL * 18]), Tm); } { V Tl, Tq, T2M, T2N; Tl = VSUB(Ti, Tk); Tq = VSUB(Tn, Tp); Tr = VFNMS(LDK(KP382683432), Tq, VMUL(LDK(KP923879532), Tl)); T1D = VFMA(LDK(KP382683432), Tl, VMUL(LDK(KP923879532), Tq)); T2M = VADD(Ti, Tk); T2N = VADD(Tn, Tp); T2O = VADD(T2M, T2N); T33 = VSUB(T2M, T2N); } } { V T15, T17, T1p, T1n, T1f, T1h, T1i, T1a, T1c, T1d; { V T14, T16, T1o, T1m; T14 = LD(&(x[WS(rs, 31)]), ms, &(x[WS(rs, 1)])); T15 = BYTWJ(&(W[TWVL * 60]), T14); T16 = LD(&(x[WS(rs, 15)]), ms, &(x[WS(rs, 1)])); T17 = BYTWJ(&(W[TWVL * 28]), T16); T1o = LD(&(x[WS(rs, 23)]), ms, &(x[WS(rs, 1)])); T1p = BYTWJ(&(W[TWVL * 44]), T1o); T1m = LD(&(x[WS(rs, 7)]), ms, &(x[WS(rs, 1)])); T1n = BYTWJ(&(W[TWVL * 12]), T1m); { V T1e, T1g, T19, T1b; T1e = LD(&(x[WS(rs, 27)]), ms, &(x[WS(rs, 1)])); T1f = BYTWJ(&(W[TWVL * 52]), T1e); T1g = LD(&(x[WS(rs, 11)]), ms, &(x[WS(rs, 1)])); T1h = BYTWJ(&(W[TWVL * 20]), T1g); T1i = VSUB(T1f, T1h); T19 = LD(&(x[WS(rs, 3)]), ms, &(x[WS(rs, 1)])); T1a = BYTWJ(&(W[TWVL * 4]), T19); T1b = LD(&(x[WS(rs, 19)]), ms, &(x[WS(rs, 1)])); T1c = BYTWJ(&(W[TWVL * 36]), T1b); T1d = VSUB(T1a, T1c); } } { V T18, T1j, T2D, T2E; T18 = VSUB(T15, T17); T1j = VMUL(LDK(KP707106781), VADD(T1d, T1i)); T1k = VADD(T18, T1j); T20 = VSUB(T18, T1j); T2D = VADD(T1a, T1c); T2E = VADD(T1f, T1h); T2F = VADD(T2D, T2E); T3b = VSUB(T2E, T2D); } { V T1l, T1q, T2A, T2B; T1l = VMUL(LDK(KP707106781), VSUB(T1i, T1d)); T1q = VSUB(T1n, T1p); T1r = VSUB(T1l, T1q); T21 = VADD(T1q, T1l); T2A = VADD(T15, T17); T2B = VADD(T1n, T1p); T2C = VADD(T2A, T2B); T3a = VSUB(T2A, T2B); } } { V TG, TI, T10, TY, TQ, TS, TT, TL, TN, TO; { V TF, TH, TZ, TX; TF = LD(&(x[WS(rs, 1)]), ms, &(x[WS(rs, 1)])); TG = BYTWJ(&(W[0]), TF); TH = LD(&(x[WS(rs, 17)]), ms, &(x[WS(rs, 1)])); TI = BYTWJ(&(W[TWVL * 32]), TH); TZ = LD(&(x[WS(rs, 25)]), ms, &(x[WS(rs, 1)])); T10 = BYTWJ(&(W[TWVL * 48]), TZ); TX = LD(&(x[WS(rs, 9)]), ms, &(x[WS(rs, 1)])); TY = BYTWJ(&(W[TWVL * 16]), TX); { V TP, TR, TK, TM; TP = LD(&(x[WS(rs, 29)]), ms, &(x[WS(rs, 1)])); TQ = BYTWJ(&(W[TWVL * 56]), TP); TR = LD(&(x[WS(rs, 13)]), ms, &(x[WS(rs, 1)])); TS = BYTWJ(&(W[TWVL * 24]), TR); TT = VSUB(TQ, TS); TK = LD(&(x[WS(rs, 5)]), ms, &(x[WS(rs, 1)])); TL = BYTWJ(&(W[TWVL * 8]), TK); TM = LD(&(x[WS(rs, 21)]), ms, &(x[WS(rs, 1)])); TN = BYTWJ(&(W[TWVL * 40]), TM); TO = VSUB(TL, TN); } } { V TJ, TU, T2w, T2x; TJ = VSUB(TG, TI); TU = VMUL(LDK(KP707106781), VADD(TO, TT)); TV = VADD(TJ, TU); T1X = VSUB(TJ, TU); T2w = VADD(TL, TN); T2x = VADD(TQ, TS); T2y = VADD(T2w, T2x); T38 = VSUB(T2x, T2w); } { V TW, T11, T2t, T2u; TW = VMUL(LDK(KP707106781), VSUB(TT, TO)); T11 = VSUB(TY, T10); T12 = VSUB(TW, T11); T1Y = VADD(T11, TW); T2t = VADD(TG, TI); T2u = VADD(TY, T10); T2v = VADD(T2t, T2u); T37 = VSUB(T2t, T2u); } } { V T2W, T30, T2Z, T31; { V T2U, T2V, T2X, T2Y; T2U = VADD(T2o, T2r); T2V = VADD(T2O, T2L); T2W = VADD(T2U, T2V); T30 = VSUB(T2U, T2V); T2X = VADD(T2v, T2y); T2Y = VADD(T2C, T2F); T2Z = VADD(T2X, T2Y); T31 = VBYI(VSUB(T2Y, T2X)); } ST(&(x[WS(rs, 16)]), VSUB(T2W, T2Z), ms, &(x[0])); ST(&(x[WS(rs, 8)]), VADD(T30, T31), ms, &(x[0])); ST(&(x[0]), VADD(T2W, T2Z), ms, &(x[0])); ST(&(x[WS(rs, 24)]), VSUB(T30, T31), ms, &(x[0])); } { V T2s, T2P, T2H, T2Q, T2z, T2G; T2s = VSUB(T2o, T2r); T2P = VSUB(T2L, T2O); T2z = VSUB(T2v, T2y); T2G = VSUB(T2C, T2F); T2H = VMUL(LDK(KP707106781), VADD(T2z, T2G)); T2Q = VMUL(LDK(KP707106781), VSUB(T2G, T2z)); { V T2I, T2R, T2S, T2T; T2I = VADD(T2s, T2H); T2R = VBYI(VADD(T2P, T2Q)); ST(&(x[WS(rs, 28)]), VSUB(T2I, T2R), ms, &(x[0])); ST(&(x[WS(rs, 4)]), VADD(T2I, T2R), ms, &(x[0])); T2S = VSUB(T2s, T2H); T2T = VBYI(VSUB(T2Q, T2P)); ST(&(x[WS(rs, 20)]), VSUB(T2S, T2T), ms, &(x[0])); ST(&(x[WS(rs, 12)]), VADD(T2S, T2T), ms, &(x[0])); } } { V T36, T3r, T3h, T3p, T3d, T3o, T3k, T3s, T35, T3g; T35 = VMUL(LDK(KP707106781), VADD(T33, T34)); T36 = VADD(T32, T35); T3r = VSUB(T32, T35); T3g = VMUL(LDK(KP707106781), VSUB(T34, T33)); T3h = VADD(T3f, T3g); T3p = VSUB(T3g, T3f); { V T39, T3c, T3i, T3j; T39 = VFMA(LDK(KP923879532), T37, VMUL(LDK(KP382683432), T38)); T3c = VFNMS(LDK(KP382683432), T3b, VMUL(LDK(KP923879532), T3a)); T3d = VADD(T39, T3c); T3o = VSUB(T3c, T39); T3i = VFNMS(LDK(KP382683432), T37, VMUL(LDK(KP923879532), T38)); T3j = VFMA(LDK(KP382683432), T3a, VMUL(LDK(KP923879532), T3b)); T3k = VADD(T3i, T3j); T3s = VSUB(T3j, T3i); } { V T3e, T3l, T3u, T3v; T3e = VADD(T36, T3d); T3l = VBYI(VADD(T3h, T3k)); ST(&(x[WS(rs, 30)]), VSUB(T3e, T3l), ms, &(x[0])); ST(&(x[WS(rs, 2)]), VADD(T3e, T3l), ms, &(x[0])); T3u = VBYI(VADD(T3p, T3o)); T3v = VADD(T3r, T3s); ST(&(x[WS(rs, 6)]), VADD(T3u, T3v), ms, &(x[0])); ST(&(x[WS(rs, 26)]), VSUB(T3v, T3u), ms, &(x[0])); } { V T3m, T3n, T3q, T3t; T3m = VSUB(T36, T3d); T3n = VBYI(VSUB(T3k, T3h)); ST(&(x[WS(rs, 18)]), VSUB(T3m, T3n), ms, &(x[0])); ST(&(x[WS(rs, 14)]), VADD(T3m, T3n), ms, &(x[0])); T3q = VBYI(VSUB(T3o, T3p)); T3t = VSUB(T3r, T3s); ST(&(x[WS(rs, 10)]), VADD(T3q, T3t), ms, &(x[0])); ST(&(x[WS(rs, 22)]), VSUB(T3t, T3q), ms, &(x[0])); } } { V TE, T1P, T1I, T1Q, T1t, T1M, T1F, T1N; { V Tg, TD, T1G, T1H; Tg = VADD(T4, Tf); TD = VADD(Tr, TC); TE = VADD(Tg, TD); T1P = VSUB(Tg, TD); T1G = VFNMS(LDK(KP195090322), TV, VMUL(LDK(KP980785280), T12)); T1H = VFMA(LDK(KP195090322), T1k, VMUL(LDK(KP980785280), T1r)); T1I = VADD(T1G, T1H); T1Q = VSUB(T1H, T1G); } { V T13, T1s, T1B, T1E; T13 = VFMA(LDK(KP980785280), TV, VMUL(LDK(KP195090322), T12)); T1s = VFNMS(LDK(KP195090322), T1r, VMUL(LDK(KP980785280), T1k)); T1t = VADD(T13, T1s); T1M = VSUB(T1s, T13); T1B = VSUB(T1v, T1A); T1E = VSUB(T1C, T1D); T1F = VADD(T1B, T1E); T1N = VSUB(T1E, T1B); } { V T1u, T1J, T1S, T1T; T1u = VADD(TE, T1t); T1J = VBYI(VADD(T1F, T1I)); ST(&(x[WS(rs, 31)]), VSUB(T1u, T1J), ms, &(x[WS(rs, 1)])); ST(&(x[WS(rs, 1)]), VADD(T1u, T1J), ms, &(x[WS(rs, 1)])); T1S = VBYI(VADD(T1N, T1M)); T1T = VADD(T1P, T1Q); ST(&(x[WS(rs, 7)]), VADD(T1S, T1T), ms, &(x[WS(rs, 1)])); ST(&(x[WS(rs, 25)]), VSUB(T1T, T1S), ms, &(x[WS(rs, 1)])); } { V T1K, T1L, T1O, T1R; T1K = VSUB(TE, T1t); T1L = VBYI(VSUB(T1I, T1F)); ST(&(x[WS(rs, 17)]), VSUB(T1K, T1L), ms, &(x[WS(rs, 1)])); ST(&(x[WS(rs, 15)]), VADD(T1K, T1L), ms, &(x[WS(rs, 1)])); T1O = VBYI(VSUB(T1M, T1N)); T1R = VSUB(T1P, T1Q); ST(&(x[WS(rs, 9)]), VADD(T1O, T1R), ms, &(x[WS(rs, 1)])); ST(&(x[WS(rs, 23)]), VSUB(T1R, T1O), ms, &(x[WS(rs, 1)])); } } { V T1W, T2h, T2a, T2i, T23, T2e, T27, T2f; { V T1U, T1V, T28, T29; T1U = VSUB(T4, Tf); T1V = VADD(T1D, T1C); T1W = VADD(T1U, T1V); T2h = VSUB(T1U, T1V); T28 = VFNMS(LDK(KP555570233), T1X, VMUL(LDK(KP831469612), T1Y)); T29 = VFMA(LDK(KP555570233), T20, VMUL(LDK(KP831469612), T21)); T2a = VADD(T28, T29); T2i = VSUB(T29, T28); } { V T1Z, T22, T25, T26; T1Z = VFMA(LDK(KP831469612), T1X, VMUL(LDK(KP555570233), T1Y)); T22 = VFNMS(LDK(KP555570233), T21, VMUL(LDK(KP831469612), T20)); T23 = VADD(T1Z, T22); T2e = VSUB(T22, T1Z); T25 = VADD(T1A, T1v); T26 = VSUB(TC, Tr); T27 = VADD(T25, T26); T2f = VSUB(T26, T25); } { V T24, T2b, T2k, T2l; T24 = VADD(T1W, T23); T2b = VBYI(VADD(T27, T2a)); ST(&(x[WS(rs, 29)]), VSUB(T24, T2b), ms, &(x[WS(rs, 1)])); ST(&(x[WS(rs, 3)]), VADD(T24, T2b), ms, &(x[WS(rs, 1)])); T2k = VBYI(VADD(T2f, T2e)); T2l = VADD(T2h, T2i); ST(&(x[WS(rs, 5)]), VADD(T2k, T2l), ms, &(x[WS(rs, 1)])); ST(&(x[WS(rs, 27)]), VSUB(T2l, T2k), ms, &(x[WS(rs, 1)])); } { V T2c, T2d, T2g, T2j; T2c = VSUB(T1W, T23); T2d = VBYI(VSUB(T2a, T27)); ST(&(x[WS(rs, 19)]), VSUB(T2c, T2d), ms, &(x[WS(rs, 1)])); ST(&(x[WS(rs, 13)]), VADD(T2c, T2d), ms, &(x[WS(rs, 1)])); T2g = VBYI(VSUB(T2e, T2f)); T2j = VSUB(T2h, T2i); ST(&(x[WS(rs, 11)]), VADD(T2g, T2j), ms, &(x[WS(rs, 1)])); ST(&(x[WS(rs, 21)]), VSUB(T2j, T2g), ms, &(x[WS(rs, 1)])); } } } } VLEAVE(); } static const tw_instr twinstr[] = { VTW(0, 1), VTW(0, 2), VTW(0, 3), VTW(0, 4), VTW(0, 5), VTW(0, 6), VTW(0, 7), VTW(0, 8), VTW(0, 9), VTW(0, 10), VTW(0, 11), VTW(0, 12), VTW(0, 13), VTW(0, 14), VTW(0, 15), VTW(0, 16), VTW(0, 17), VTW(0, 18), VTW(0, 19), VTW(0, 20), VTW(0, 21), VTW(0, 22), VTW(0, 23), VTW(0, 24), VTW(0, 25), VTW(0, 26), VTW(0, 27), VTW(0, 28), VTW(0, 29), VTW(0, 30), VTW(0, 31), {TW_NEXT, VL, 0} }; static const ct_desc desc = { 32, XSIMD_STRING("t1fv_32"), twinstr, &GENUS, {201, 88, 16, 0}, 0, 0, 0 }; void XSIMD(codelet_t1fv_32) (planner *p) { X(kdft_dit_register) (p, t1fv_32, &desc); } #endif /* HAVE_FMA */ fftw-3.3.3/dft/simd/common/t1fv_64.c0000644000175000017500000017600512054410510013746 00000000000000/* * Copyright (c) 2003, 2007-11 Matteo Frigo * Copyright (c) 2003, 2007-11 Massachusetts Institute of Technology * * This program is free software; you can redistribute it and/or modify * it under the terms of the GNU General Public License as published by * the Free Software Foundation; either version 2 of the License, or * (at your option) any later version. * * This program is distributed in the hope that it will be useful, * but WITHOUT ANY WARRANTY; without even the implied warranty of * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the * GNU General Public License for more details. * * You should have received a copy of the GNU General Public License * along with this program; if not, write to the Free Software * Foundation, Inc., 51 Franklin Street, Fifth Floor, Boston, MA 02110-1301 USA * */ /* This file was automatically generated --- DO NOT EDIT */ /* Generated on Sun Nov 25 07:38:10 EST 2012 */ #include "codelet-dft.h" #ifdef HAVE_FMA /* Generated by: ../../../genfft/gen_twiddle_c.native -fma -reorder-insns -schedule-for-pipeline -simd -compact -variables 4 -pipeline-latency 8 -n 64 -name t1fv_64 -include t1f.h */ /* * This function contains 519 FP additions, 384 FP multiplications, * (or, 261 additions, 126 multiplications, 258 fused multiply/add), * 187 stack variables, 15 constants, and 128 memory accesses */ #include "t1f.h" static void t1fv_64(R *ri, R *ii, const R *W, stride rs, INT mb, INT me, INT ms) { DVK(KP773010453, +0.773010453362736960810906609758469800971041293); DVK(KP995184726, +0.995184726672196886244836953109479921575474869); DVK(KP820678790, +0.820678790828660330972281985331011598767386482); DVK(KP098491403, +0.098491403357164253077197521291327432293052451); DVK(KP956940335, +0.956940335732208864935797886980269969482849206); DVK(KP881921264, +0.881921264348355029712756863660388349508442621); DVK(KP303346683, +0.303346683607342391675883946941299872384187453); DVK(KP534511135, +0.534511135950791641089685961295362908582039528); DVK(KP831469612, +0.831469612302545237078788377617905756738560812); DVK(KP980785280, +0.980785280403230449126182236134239036973933731); DVK(KP668178637, +0.668178637919298919997757686523080761552472251); DVK(KP198912367, +0.198912367379658006911597622644676228597850501); DVK(KP923879532, +0.923879532511286756128183189396788286822416626); DVK(KP414213562, +0.414213562373095048801688724209698078569671875); DVK(KP707106781, +0.707106781186547524400844362104849039284835938); { INT m; R *x; x = ri; for (m = mb, W = W + (mb * ((TWVL / VL) * 126)); m < me; m = m + VL, x = x + (VL * ms), W = W + (TWVL * 126), MAKE_VOLATILE_STRIDE(64, rs)) { V T6L, T6M, T6O, T6P, T75, T6V, T5A, T6A, T72, T6K, T6t, T6D, T6w, T6B, T6h; V T6E; { V Ta, T3U, T3V, T37, T7a, T58, T7B, T6l, T1v, T24, T5Q, T7o, T5F, T7l, T43; V T4F, T2i, T2R, T6b, T7v, T60, T7s, T4a, T4I, T5u, T7h, T5x, T7g, T1i, T3a; V T4j, T4C, T7e, T5l, T7d, T5o, T3b, TV, T4B, T4m, T3X, T3Y, T6o, T7b, T5f; V T7C, Tx, T38, T2p, T61, T2n, T65, T2D, T7p, T5M, T7m, T5T, T4G, T46, T25; V T1S, T2q, T2u, T2w; { V T5q, T10, T5v, T15, T1b, T5s, T1c, T1e; { V T1V, T1p, T5B, T5O, T1u, T1X, T20, T21; { V T1, T2, T7, T5, T32, T34, T2X, T2Z; T1 = LD(&(x[0]), ms, &(x[0])); T2 = LD(&(x[WS(rs, 32)]), ms, &(x[0])); T7 = LD(&(x[WS(rs, 48)]), ms, &(x[0])); T5 = LD(&(x[WS(rs, 16)]), ms, &(x[0])); T32 = LD(&(x[WS(rs, 8)]), ms, &(x[0])); T34 = LD(&(x[WS(rs, 40)]), ms, &(x[0])); T2X = LD(&(x[WS(rs, 56)]), ms, &(x[0])); T2Z = LD(&(x[WS(rs, 24)]), ms, &(x[0])); { V T1m, T54, T6j, T36, T55, T31, T56, T1n, T1q, T1s, T4, T9; { V T3, T8, T6, T33, T35, T2Y, T30, T1l; T1l = LD(&(x[WS(rs, 1)]), ms, &(x[WS(rs, 1)])); T3 = BYTWJ(&(W[TWVL * 62]), T2); T8 = BYTWJ(&(W[TWVL * 94]), T7); T6 = BYTWJ(&(W[TWVL * 30]), T5); T33 = BYTWJ(&(W[TWVL * 14]), T32); T35 = BYTWJ(&(W[TWVL * 78]), T34); T2Y = BYTWJ(&(W[TWVL * 110]), T2X); T30 = BYTWJ(&(W[TWVL * 46]), T2Z); T1m = BYTWJ(&(W[0]), T1l); T54 = VSUB(T1, T3); T4 = VADD(T1, T3); T6j = VSUB(T6, T8); T9 = VADD(T6, T8); T36 = VADD(T33, T35); T55 = VSUB(T33, T35); T31 = VADD(T2Y, T30); T56 = VSUB(T2Y, T30); T1n = LD(&(x[WS(rs, 33)]), ms, &(x[WS(rs, 1)])); } T1q = LD(&(x[WS(rs, 17)]), ms, &(x[WS(rs, 1)])); T1s = LD(&(x[WS(rs, 49)]), ms, &(x[WS(rs, 1)])); Ta = VSUB(T4, T9); T3U = VADD(T4, T9); { V T57, T6k, T1o, T1r, T1t, T1W, T1U, T1Z; T1U = LD(&(x[WS(rs, 9)]), ms, &(x[WS(rs, 1)])); T3V = VADD(T36, T31); T37 = VSUB(T31, T36); T57 = VADD(T55, T56); T6k = VSUB(T56, T55); T1o = BYTWJ(&(W[TWVL * 64]), T1n); T1r = BYTWJ(&(W[TWVL * 32]), T1q); T1t = BYTWJ(&(W[TWVL * 96]), T1s); T1V = BYTWJ(&(W[TWVL * 16]), T1U); T1W = LD(&(x[WS(rs, 41)]), ms, &(x[WS(rs, 1)])); T1Z = LD(&(x[WS(rs, 57)]), ms, &(x[WS(rs, 1)])); T7a = VFNMS(LDK(KP707106781), T57, T54); T58 = VFMA(LDK(KP707106781), T57, T54); T7B = VFMA(LDK(KP707106781), T6k, T6j); T6l = VFNMS(LDK(KP707106781), T6k, T6j); T1p = VADD(T1m, T1o); T5B = VSUB(T1m, T1o); T5O = VSUB(T1r, T1t); T1u = VADD(T1r, T1t); T1X = BYTWJ(&(W[TWVL * 80]), T1W); T20 = BYTWJ(&(W[TWVL * 112]), T1Z); T21 = LD(&(x[WS(rs, 25)]), ms, &(x[WS(rs, 1)])); } } } { V T5W, T2N, T69, T2L, T5Y, T2P, T48, T2c, T2h; { V T41, T1Y, T5C, T22, T2d, T29, T2b, T2f, T28, T2a, T2H, T2J; T28 = LD(&(x[WS(rs, 63)]), ms, &(x[WS(rs, 1)])); T2a = LD(&(x[WS(rs, 31)]), ms, &(x[WS(rs, 1)])); T1v = VSUB(T1p, T1u); T41 = VADD(T1p, T1u); T1Y = VADD(T1V, T1X); T5C = VSUB(T1V, T1X); T22 = BYTWJ(&(W[TWVL * 48]), T21); T2d = LD(&(x[WS(rs, 15)]), ms, &(x[WS(rs, 1)])); T29 = BYTWJ(&(W[TWVL * 124]), T28); T2b = BYTWJ(&(W[TWVL * 60]), T2a); T2f = LD(&(x[WS(rs, 47)]), ms, &(x[WS(rs, 1)])); T2H = LD(&(x[WS(rs, 55)]), ms, &(x[WS(rs, 1)])); T2J = LD(&(x[WS(rs, 23)]), ms, &(x[WS(rs, 1)])); { V T23, T5D, T2e, T2g, T2I, T2K, T2M; T2M = LD(&(x[WS(rs, 7)]), ms, &(x[WS(rs, 1)])); T23 = VADD(T20, T22); T5D = VSUB(T20, T22); T2e = BYTWJ(&(W[TWVL * 28]), T2d); T2c = VADD(T29, T2b); T5W = VSUB(T29, T2b); T2g = BYTWJ(&(W[TWVL * 92]), T2f); T2I = BYTWJ(&(W[TWVL * 108]), T2H); T2K = BYTWJ(&(W[TWVL * 44]), T2J); T2N = BYTWJ(&(W[TWVL * 12]), T2M); { V T5E, T5P, T42, T2O; T5E = VADD(T5C, T5D); T5P = VSUB(T5C, T5D); T24 = VSUB(T1Y, T23); T42 = VADD(T1Y, T23); T69 = VSUB(T2g, T2e); T2h = VADD(T2e, T2g); T2O = LD(&(x[WS(rs, 39)]), ms, &(x[WS(rs, 1)])); T2L = VADD(T2I, T2K); T5Y = VSUB(T2I, T2K); T5Q = VFMA(LDK(KP707106781), T5P, T5O); T7o = VFNMS(LDK(KP707106781), T5P, T5O); T5F = VFMA(LDK(KP707106781), T5E, T5B); T7l = VFNMS(LDK(KP707106781), T5E, T5B); T43 = VADD(T41, T42); T4F = VSUB(T41, T42); T2P = BYTWJ(&(W[TWVL * 76]), T2O); } } } T2i = VSUB(T2c, T2h); T48 = VADD(T2c, T2h); { V TW, TY, T11, T2Q, T5X, T13; TW = LD(&(x[WS(rs, 62)]), ms, &(x[0])); TY = LD(&(x[WS(rs, 30)]), ms, &(x[0])); T11 = LD(&(x[WS(rs, 14)]), ms, &(x[0])); T2Q = VADD(T2N, T2P); T5X = VSUB(T2N, T2P); T13 = LD(&(x[WS(rs, 46)]), ms, &(x[0])); { V T12, T5Z, T6a, T49, T14, T18, T1a; { V T17, T19, TX, TZ; T17 = LD(&(x[WS(rs, 54)]), ms, &(x[0])); T19 = LD(&(x[WS(rs, 22)]), ms, &(x[0])); TX = BYTWJ(&(W[TWVL * 122]), TW); TZ = BYTWJ(&(W[TWVL * 58]), TY); T12 = BYTWJ(&(W[TWVL * 26]), T11); T5Z = VADD(T5X, T5Y); T6a = VSUB(T5Y, T5X); T2R = VSUB(T2L, T2Q); T49 = VADD(T2Q, T2L); T14 = BYTWJ(&(W[TWVL * 90]), T13); T18 = BYTWJ(&(W[TWVL * 106]), T17); T5q = VSUB(TX, TZ); T10 = VADD(TX, TZ); T1a = BYTWJ(&(W[TWVL * 42]), T19); } T6b = VFMA(LDK(KP707106781), T6a, T69); T7v = VFNMS(LDK(KP707106781), T6a, T69); T60 = VFMA(LDK(KP707106781), T5Z, T5W); T7s = VFNMS(LDK(KP707106781), T5Z, T5W); T4a = VADD(T48, T49); T4I = VSUB(T48, T49); T5v = VSUB(T14, T12); T15 = VADD(T12, T14); T1b = VADD(T18, T1a); T5s = VSUB(T18, T1a); } T1c = LD(&(x[WS(rs, 6)]), ms, &(x[0])); T1e = LD(&(x[WS(rs, 38)]), ms, &(x[0])); } } } { V Th, T59, Tf, Tv, T5d, Tj, Tm, To; { V T5h, TQ, T5m, T5i, TO, TS, TJ, T4k, TD, TI; { V T4h, T16, TB, T1d, T1f, TE, TG, TA, Tz, TK, TM, TC; Tz = LD(&(x[WS(rs, 2)]), ms, &(x[0])); T4h = VADD(T10, T15); T16 = VSUB(T10, T15); TB = LD(&(x[WS(rs, 34)]), ms, &(x[0])); T1d = BYTWJ(&(W[TWVL * 10]), T1c); T1f = BYTWJ(&(W[TWVL * 74]), T1e); TE = LD(&(x[WS(rs, 18)]), ms, &(x[0])); TG = LD(&(x[WS(rs, 50)]), ms, &(x[0])); TA = BYTWJ(&(W[TWVL * 2]), Tz); TK = LD(&(x[WS(rs, 10)]), ms, &(x[0])); TM = LD(&(x[WS(rs, 42)]), ms, &(x[0])); TC = BYTWJ(&(W[TWVL * 66]), TB); { V T1g, T5r, TF, TH, TL, TN, TP; TP = LD(&(x[WS(rs, 58)]), ms, &(x[0])); T1g = VADD(T1d, T1f); T5r = VSUB(T1d, T1f); TF = BYTWJ(&(W[TWVL * 34]), TE); TH = BYTWJ(&(W[TWVL * 98]), TG); TL = BYTWJ(&(W[TWVL * 18]), TK); TN = BYTWJ(&(W[TWVL * 82]), TM); T5h = VSUB(TA, TC); TD = VADD(TA, TC); TQ = BYTWJ(&(W[TWVL * 114]), TP); { V T5w, T5t, T4i, T1h, TR; T5w = VSUB(T5s, T5r); T5t = VADD(T5r, T5s); T4i = VADD(T1g, T1b); T1h = VSUB(T1b, T1g); T5m = VSUB(TF, TH); TI = VADD(TF, TH); T5i = VSUB(TL, TN); TO = VADD(TL, TN); TR = LD(&(x[WS(rs, 26)]), ms, &(x[0])); T5u = VFMA(LDK(KP707106781), T5t, T5q); T7h = VFNMS(LDK(KP707106781), T5t, T5q); T5x = VFMA(LDK(KP707106781), T5w, T5v); T7g = VFNMS(LDK(KP707106781), T5w, T5v); T1i = VFNMS(LDK(KP414213562), T1h, T16); T3a = VFMA(LDK(KP414213562), T16, T1h); T4j = VADD(T4h, T4i); T4C = VSUB(T4h, T4i); TS = BYTWJ(&(W[TWVL * 50]), TR); } } } TJ = VSUB(TD, TI); T4k = VADD(TD, TI); { V Tb, Td, Tr, T5j, TT, Tt, Tg; Tb = LD(&(x[WS(rs, 4)]), ms, &(x[0])); Td = LD(&(x[WS(rs, 36)]), ms, &(x[0])); Tr = LD(&(x[WS(rs, 12)]), ms, &(x[0])); T5j = VSUB(TQ, TS); TT = VADD(TQ, TS); Tt = LD(&(x[WS(rs, 44)]), ms, &(x[0])); Tg = LD(&(x[WS(rs, 20)]), ms, &(x[0])); { V Ti, Tc, Te, Ts; Ti = LD(&(x[WS(rs, 52)]), ms, &(x[0])); Tc = BYTWJ(&(W[TWVL * 6]), Tb); Te = BYTWJ(&(W[TWVL * 70]), Td); Ts = BYTWJ(&(W[TWVL * 22]), Tr); { V T5k, T5n, TU, T4l, Tu; T5k = VADD(T5i, T5j); T5n = VSUB(T5i, T5j); TU = VSUB(TO, TT); T4l = VADD(TO, TT); Tu = BYTWJ(&(W[TWVL * 86]), Tt); Th = BYTWJ(&(W[TWVL * 38]), Tg); T59 = VSUB(Tc, Te); Tf = VADD(Tc, Te); T7e = VFNMS(LDK(KP707106781), T5k, T5h); T5l = VFMA(LDK(KP707106781), T5k, T5h); T7d = VFNMS(LDK(KP707106781), T5n, T5m); T5o = VFMA(LDK(KP707106781), T5n, T5m); T3b = VFMA(LDK(KP414213562), TJ, TU); TV = VFNMS(LDK(KP414213562), TU, TJ); T4B = VSUB(T4k, T4l); T4m = VADD(T4k, T4l); Tv = VADD(Ts, Tu); T5d = VSUB(Tu, Ts); Tj = BYTWJ(&(W[TWVL * 102]), Ti); } } Tm = LD(&(x[WS(rs, 60)]), ms, &(x[0])); To = LD(&(x[WS(rs, 28)]), ms, &(x[0])); } } { V T5b, T6m, Tl, T1A, T5G, T1Q, T5K, T1C, T1D, T5e, T6n, Tw, T1H, T1J; { V T1w, T1y, T1M, T1O, Tq, T5c, T1B; T1w = LD(&(x[WS(rs, 5)]), ms, &(x[WS(rs, 1)])); T1y = LD(&(x[WS(rs, 37)]), ms, &(x[WS(rs, 1)])); T1M = LD(&(x[WS(rs, 13)]), ms, &(x[WS(rs, 1)])); T1O = LD(&(x[WS(rs, 45)]), ms, &(x[WS(rs, 1)])); T1B = LD(&(x[WS(rs, 21)]), ms, &(x[WS(rs, 1)])); { V Tk, T5a, Tn, Tp; Tk = VADD(Th, Tj); T5a = VSUB(Th, Tj); Tn = BYTWJ(&(W[TWVL * 118]), Tm); Tp = BYTWJ(&(W[TWVL * 54]), To); { V T1x, T1z, T1N, T1P; T1x = BYTWJ(&(W[TWVL * 8]), T1w); T1z = BYTWJ(&(W[TWVL * 72]), T1y); T1N = BYTWJ(&(W[TWVL * 24]), T1M); T1P = BYTWJ(&(W[TWVL * 88]), T1O); T5b = VFNMS(LDK(KP414213562), T5a, T59); T6m = VFMA(LDK(KP414213562), T59, T5a); T3X = VADD(Tf, Tk); Tl = VSUB(Tf, Tk); Tq = VADD(Tn, Tp); T5c = VSUB(Tn, Tp); T1A = VADD(T1x, T1z); T5G = VSUB(T1x, T1z); T1Q = VADD(T1N, T1P); T5K = VSUB(T1N, T1P); T1C = BYTWJ(&(W[TWVL * 40]), T1B); } } T1D = LD(&(x[WS(rs, 53)]), ms, &(x[WS(rs, 1)])); T5e = VFNMS(LDK(KP414213562), T5d, T5c); T6n = VFMA(LDK(KP414213562), T5c, T5d); T3Y = VADD(Tq, Tv); Tw = VSUB(Tq, Tv); T1H = LD(&(x[WS(rs, 61)]), ms, &(x[WS(rs, 1)])); T1J = LD(&(x[WS(rs, 29)]), ms, &(x[WS(rs, 1)])); } { V T1I, T1K, T1F, T5H, T2k, T2l, T2z, T2B, T2j, T1E; T2j = LD(&(x[WS(rs, 3)]), ms, &(x[WS(rs, 1)])); T1E = BYTWJ(&(W[TWVL * 104]), T1D); T6o = VSUB(T6m, T6n); T7b = VADD(T6m, T6n); T5f = VADD(T5b, T5e); T7C = VSUB(T5e, T5b); Tx = VADD(Tl, Tw); T38 = VSUB(Tw, Tl); T1I = BYTWJ(&(W[TWVL * 120]), T1H); T1K = BYTWJ(&(W[TWVL * 56]), T1J); T1F = VADD(T1C, T1E); T5H = VSUB(T1C, T1E); T2k = BYTWJ(&(W[TWVL * 4]), T2j); T2l = LD(&(x[WS(rs, 35)]), ms, &(x[WS(rs, 1)])); T2z = LD(&(x[WS(rs, 11)]), ms, &(x[WS(rs, 1)])); T2B = LD(&(x[WS(rs, 43)]), ms, &(x[WS(rs, 1)])); { V T5I, T5R, T44, T1G, T2m, T2A, T2C, T5S, T5L, T1R, T45, T2o, T5J, T1L; T2o = LD(&(x[WS(rs, 19)]), ms, &(x[WS(rs, 1)])); T5J = VSUB(T1I, T1K); T1L = VADD(T1I, T1K); T5I = VFNMS(LDK(KP414213562), T5H, T5G); T5R = VFMA(LDK(KP414213562), T5G, T5H); T44 = VADD(T1A, T1F); T1G = VSUB(T1A, T1F); T2m = BYTWJ(&(W[TWVL * 68]), T2l); T2A = BYTWJ(&(W[TWVL * 20]), T2z); T2C = BYTWJ(&(W[TWVL * 84]), T2B); T5S = VFNMS(LDK(KP414213562), T5J, T5K); T5L = VFMA(LDK(KP414213562), T5K, T5J); T1R = VSUB(T1L, T1Q); T45 = VADD(T1L, T1Q); T2p = BYTWJ(&(W[TWVL * 36]), T2o); T61 = VSUB(T2k, T2m); T2n = VADD(T2k, T2m); T65 = VSUB(T2C, T2A); T2D = VADD(T2A, T2C); T7p = VSUB(T5I, T5L); T5M = VADD(T5I, T5L); T7m = VSUB(T5R, T5S); T5T = VADD(T5R, T5S); T4G = VSUB(T44, T45); T46 = VADD(T44, T45); T25 = VSUB(T1G, T1R); T1S = VADD(T1G, T1R); T2q = LD(&(x[WS(rs, 51)]), ms, &(x[WS(rs, 1)])); } T2u = LD(&(x[WS(rs, 59)]), ms, &(x[WS(rs, 1)])); T2w = LD(&(x[WS(rs, 27)]), ms, &(x[WS(rs, 1)])); } } } } { V T67, T7w, T6e, T7t, T3s, T3E, T39, T3D, T1k, T3k, T3t, T3c, T1T, T3v, T3w; V T26, T2G, T3y, T3z, T2T; { V T4A, T4N, T47, T4v, T2r, T2v, T2x, T4s, T40, T3W, T3Z; T4A = VSUB(T3U, T3V); T3W = VADD(T3U, T3V); T3Z = VADD(T3X, T3Y); T4N = VSUB(T3Y, T3X); T47 = VSUB(T43, T46); T4v = VADD(T43, T46); T2r = BYTWJ(&(W[TWVL * 100]), T2q); T2v = BYTWJ(&(W[TWVL * 116]), T2u); T2x = BYTWJ(&(W[TWVL * 52]), T2w); T4s = VADD(T3W, T3Z); T40 = VSUB(T3W, T3Z); { V T4O, T4n, T4R, T4H, T4E, T4W, T4u, T4y, T4d, T4J, T2F, T2S; { V T6c, T63, T2t, T4b, T6d, T66, T2E, T4c; { V T4D, T62, T2s, T64, T2y, T4t; T4O = VSUB(T4C, T4B); T4D = VADD(T4B, T4C); T62 = VSUB(T2r, T2p); T2s = VADD(T2p, T2r); T64 = VSUB(T2v, T2x); T2y = VADD(T2v, T2x); T4t = VADD(T4m, T4j); T4n = VSUB(T4j, T4m); T4R = VFMA(LDK(KP414213562), T4F, T4G); T4H = VFNMS(LDK(KP414213562), T4G, T4F); T4E = VFMA(LDK(KP707106781), T4D, T4A); T4W = VFNMS(LDK(KP707106781), T4D, T4A); T6c = VFNMS(LDK(KP414213562), T61, T62); T63 = VFMA(LDK(KP414213562), T62, T61); T2t = VSUB(T2n, T2s); T4b = VADD(T2n, T2s); T6d = VFMA(LDK(KP414213562), T64, T65); T66 = VFNMS(LDK(KP414213562), T65, T64); T2E = VSUB(T2y, T2D); T4c = VADD(T2y, T2D); T4u = VADD(T4s, T4t); T4y = VSUB(T4s, T4t); } T67 = VADD(T63, T66); T7w = VSUB(T66, T63); T6e = VADD(T6c, T6d); T7t = VSUB(T6d, T6c); T4d = VADD(T4b, T4c); T4J = VSUB(T4c, T4b); T2F = VADD(T2t, T2E); T2S = VSUB(T2E, T2t); } { V Ty, T1j, T4Q, T4K; Ty = VFMA(LDK(KP707106781), Tx, Ta); T3s = VFNMS(LDK(KP707106781), Tx, Ta); T3E = VSUB(T1i, TV); T1j = VADD(TV, T1i); T39 = VFMA(LDK(KP707106781), T38, T37); T3D = VFNMS(LDK(KP707106781), T38, T37); T4Q = VFMA(LDK(KP414213562), T4I, T4J); T4K = VFNMS(LDK(KP414213562), T4J, T4I); { V T4w, T4e, T4P, T4Z; T4w = VADD(T4a, T4d); T4e = VSUB(T4a, T4d); T4P = VFMA(LDK(KP707106781), T4O, T4N); T4Z = VFNMS(LDK(KP707106781), T4O, T4N); T1k = VFMA(LDK(KP923879532), T1j, Ty); T3k = VFNMS(LDK(KP923879532), T1j, Ty); { V T4L, T50, T4S, T4X; T4L = VADD(T4H, T4K); T50 = VSUB(T4K, T4H); T4S = VSUB(T4Q, T4R); T4X = VADD(T4R, T4Q); { V T4f, T4o, T4x, T4z; T4f = VADD(T47, T4e); T4o = VSUB(T4e, T47); T4x = VADD(T4v, T4w); T4z = VSUB(T4w, T4v); { V T53, T51, T4M, T4U; T53 = VFNMS(LDK(KP923879532), T50, T4Z); T51 = VFMA(LDK(KP923879532), T50, T4Z); T4M = VFNMS(LDK(KP923879532), T4L, T4E); T4U = VFMA(LDK(KP923879532), T4L, T4E); { V T52, T4Y, T4T, T4V; T52 = VFMA(LDK(KP923879532), T4X, T4W); T4Y = VFNMS(LDK(KP923879532), T4X, T4W); T4T = VFNMS(LDK(KP923879532), T4S, T4P); T4V = VFMA(LDK(KP923879532), T4S, T4P); { V T4p, T4r, T4g, T4q; T4p = VFNMS(LDK(KP707106781), T4o, T4n); T4r = VFMA(LDK(KP707106781), T4o, T4n); T4g = VFNMS(LDK(KP707106781), T4f, T40); T4q = VFMA(LDK(KP707106781), T4f, T40); ST(&(x[WS(rs, 16)]), VFMAI(T4z, T4y), ms, &(x[0])); ST(&(x[WS(rs, 48)]), VFNMSI(T4z, T4y), ms, &(x[0])); ST(&(x[0]), VADD(T4u, T4x), ms, &(x[0])); ST(&(x[WS(rs, 32)]), VSUB(T4u, T4x), ms, &(x[0])); ST(&(x[WS(rs, 44)]), VFNMSI(T51, T4Y), ms, &(x[0])); ST(&(x[WS(rs, 20)]), VFMAI(T51, T4Y), ms, &(x[0])); ST(&(x[WS(rs, 52)]), VFMAI(T53, T52), ms, &(x[0])); ST(&(x[WS(rs, 12)]), VFNMSI(T53, T52), ms, &(x[0])); ST(&(x[WS(rs, 4)]), VFMAI(T4V, T4U), ms, &(x[0])); ST(&(x[WS(rs, 60)]), VFNMSI(T4V, T4U), ms, &(x[0])); ST(&(x[WS(rs, 36)]), VFMAI(T4T, T4M), ms, &(x[0])); ST(&(x[WS(rs, 28)]), VFNMSI(T4T, T4M), ms, &(x[0])); ST(&(x[WS(rs, 8)]), VFMAI(T4r, T4q), ms, &(x[0])); ST(&(x[WS(rs, 56)]), VFNMSI(T4r, T4q), ms, &(x[0])); ST(&(x[WS(rs, 40)]), VFMAI(T4p, T4g), ms, &(x[0])); ST(&(x[WS(rs, 24)]), VFNMSI(T4p, T4g), ms, &(x[0])); T3t = VADD(T3b, T3a); T3c = VSUB(T3a, T3b); } } } } } } T1T = VFMA(LDK(KP707106781), T1S, T1v); T3v = VFNMS(LDK(KP707106781), T1S, T1v); T3w = VFNMS(LDK(KP707106781), T25, T24); T26 = VFMA(LDK(KP707106781), T25, T24); T2G = VFMA(LDK(KP707106781), T2F, T2i); T3y = VFNMS(LDK(KP707106781), T2F, T2i); T3z = VFNMS(LDK(KP707106781), T2S, T2R); T2T = VFMA(LDK(KP707106781), T2S, T2R); } } } { V T3u, T3M, T3F, T3P, T3x, T3H, T3q, T3m, T3h, T3j, T3r, T3p, T2W, T3i; { V T3d, T3n, T27, T3f, T2U, T3e; T3d = VFMA(LDK(KP923879532), T3c, T39); T3n = VFNMS(LDK(KP923879532), T3c, T39); T27 = VFNMS(LDK(KP198912367), T26, T1T); T3f = VFMA(LDK(KP198912367), T1T, T26); T2U = VFNMS(LDK(KP198912367), T2T, T2G); T3e = VFMA(LDK(KP198912367), T2G, T2T); T3u = VFMA(LDK(KP923879532), T3t, T3s); T3M = VFNMS(LDK(KP923879532), T3t, T3s); { V T3g, T3l, T2V, T3o; T3g = VSUB(T3e, T3f); T3l = VADD(T3f, T3e); T2V = VADD(T27, T2U); T3o = VSUB(T2U, T27); T3F = VFNMS(LDK(KP923879532), T3E, T3D); T3P = VFMA(LDK(KP923879532), T3E, T3D); T3x = VFMA(LDK(KP668178637), T3w, T3v); T3H = VFNMS(LDK(KP668178637), T3v, T3w); T3q = VFMA(LDK(KP980785280), T3l, T3k); T3m = VFNMS(LDK(KP980785280), T3l, T3k); T3h = VFNMS(LDK(KP980785280), T3g, T3d); T3j = VFMA(LDK(KP980785280), T3g, T3d); T3r = VFNMS(LDK(KP980785280), T3o, T3n); T3p = VFMA(LDK(KP980785280), T3o, T3n); T2W = VFNMS(LDK(KP980785280), T2V, T1k); T3i = VFMA(LDK(KP980785280), T2V, T1k); } } { V T7n, T7Z, T8j, T89, T7k, T7O, T8g, T7Y, T7H, T7R, T80, T7q, T7u, T82, T83; V T7x; { V T7c, T7W, T7D, T87, T7f, T7F, T3A, T3G, T7E, T7i; T7c = VFNMS(LDK(KP923879532), T7b, T7a); T7W = VFMA(LDK(KP923879532), T7b, T7a); T7D = VFNMS(LDK(KP923879532), T7C, T7B); T87 = VFMA(LDK(KP923879532), T7C, T7B); T7f = VFNMS(LDK(KP668178637), T7e, T7d); T7F = VFMA(LDK(KP668178637), T7d, T7e); ST(&(x[WS(rs, 46)]), VFNMSI(T3p, T3m), ms, &(x[0])); ST(&(x[WS(rs, 18)]), VFMAI(T3p, T3m), ms, &(x[0])); ST(&(x[WS(rs, 50)]), VFMAI(T3r, T3q), ms, &(x[0])); ST(&(x[WS(rs, 14)]), VFNMSI(T3r, T3q), ms, &(x[0])); ST(&(x[WS(rs, 2)]), VFMAI(T3j, T3i), ms, &(x[0])); ST(&(x[WS(rs, 62)]), VFNMSI(T3j, T3i), ms, &(x[0])); ST(&(x[WS(rs, 34)]), VFMAI(T3h, T2W), ms, &(x[0])); ST(&(x[WS(rs, 30)]), VFNMSI(T3h, T2W), ms, &(x[0])); T3A = VFMA(LDK(KP668178637), T3z, T3y); T3G = VFNMS(LDK(KP668178637), T3y, T3z); T7E = VFMA(LDK(KP668178637), T7g, T7h); T7i = VFNMS(LDK(KP668178637), T7h, T7g); T7n = VFNMS(LDK(KP923879532), T7m, T7l); T7Z = VFMA(LDK(KP923879532), T7m, T7l); { V T3I, T3N, T3B, T3Q; T3I = VSUB(T3G, T3H); T3N = VADD(T3H, T3G); T3B = VADD(T3x, T3A); T3Q = VSUB(T3A, T3x); { V T7j, T88, T7G, T7X; T7j = VADD(T7f, T7i); T88 = VSUB(T7f, T7i); T7G = VSUB(T7E, T7F); T7X = VADD(T7F, T7E); { V T3S, T3O, T3J, T3L; T3S = VFNMS(LDK(KP831469612), T3N, T3M); T3O = VFMA(LDK(KP831469612), T3N, T3M); T3J = VFNMS(LDK(KP831469612), T3I, T3F); T3L = VFMA(LDK(KP831469612), T3I, T3F); { V T3T, T3R, T3C, T3K; T3T = VFMA(LDK(KP831469612), T3Q, T3P); T3R = VFNMS(LDK(KP831469612), T3Q, T3P); T3C = VFNMS(LDK(KP831469612), T3B, T3u); T3K = VFMA(LDK(KP831469612), T3B, T3u); T8j = VFNMS(LDK(KP831469612), T88, T87); T89 = VFMA(LDK(KP831469612), T88, T87); T7k = VFNMS(LDK(KP831469612), T7j, T7c); T7O = VFMA(LDK(KP831469612), T7j, T7c); T8g = VFNMS(LDK(KP831469612), T7X, T7W); T7Y = VFMA(LDK(KP831469612), T7X, T7W); T7H = VFNMS(LDK(KP831469612), T7G, T7D); T7R = VFMA(LDK(KP831469612), T7G, T7D); ST(&(x[WS(rs, 42)]), VFMAI(T3R, T3O), ms, &(x[0])); ST(&(x[WS(rs, 22)]), VFNMSI(T3R, T3O), ms, &(x[0])); ST(&(x[WS(rs, 54)]), VFNMSI(T3T, T3S), ms, &(x[0])); ST(&(x[WS(rs, 10)]), VFMAI(T3T, T3S), ms, &(x[0])); ST(&(x[WS(rs, 58)]), VFMAI(T3L, T3K), ms, &(x[0])); ST(&(x[WS(rs, 6)]), VFNMSI(T3L, T3K), ms, &(x[0])); ST(&(x[WS(rs, 26)]), VFMAI(T3J, T3C), ms, &(x[0])); ST(&(x[WS(rs, 38)]), VFNMSI(T3J, T3C), ms, &(x[0])); T80 = VFNMS(LDK(KP923879532), T7p, T7o); T7q = VFMA(LDK(KP923879532), T7p, T7o); } } } } T7u = VFNMS(LDK(KP923879532), T7t, T7s); T82 = VFMA(LDK(KP923879532), T7t, T7s); T83 = VFNMS(LDK(KP923879532), T7w, T7v); T7x = VFMA(LDK(KP923879532), T7w, T7v); } { V T5g, T6I, T6p, T6T, T5p, T6q, T6r, T5y; T5g = VFMA(LDK(KP923879532), T5f, T58); T6I = VFNMS(LDK(KP923879532), T5f, T58); { V T7r, T7I, T7y, T7J; T7r = VFNMS(LDK(KP534511135), T7q, T7n); T7I = VFMA(LDK(KP534511135), T7n, T7q); T7y = VFNMS(LDK(KP534511135), T7x, T7u); T7J = VFMA(LDK(KP534511135), T7u, T7x); { V T81, T8a, T84, T8b; T81 = VFMA(LDK(KP303346683), T80, T7Z); T8a = VFNMS(LDK(KP303346683), T7Z, T80); T84 = VFMA(LDK(KP303346683), T83, T82); T8b = VFNMS(LDK(KP303346683), T82, T83); T6p = VFMA(LDK(KP923879532), T6o, T6l); T6T = VFNMS(LDK(KP923879532), T6o, T6l); T5p = VFNMS(LDK(KP198912367), T5o, T5l); T6q = VFMA(LDK(KP198912367), T5l, T5o); { V T7K, T7P, T7z, T7S; T7K = VSUB(T7I, T7J); T7P = VADD(T7I, T7J); T7z = VADD(T7r, T7y); T7S = VSUB(T7y, T7r); { V T8c, T8h, T85, T8k; T8c = VSUB(T8a, T8b); T8h = VADD(T8a, T8b); T85 = VADD(T81, T84); T8k = VSUB(T84, T81); { V T7Q, T7U, T7L, T7N; T7Q = VFNMS(LDK(KP881921264), T7P, T7O); T7U = VFMA(LDK(KP881921264), T7P, T7O); T7L = VFNMS(LDK(KP881921264), T7K, T7H); T7N = VFMA(LDK(KP881921264), T7K, T7H); { V T7T, T7V, T7A, T7M; T7T = VFNMS(LDK(KP881921264), T7S, T7R); T7V = VFMA(LDK(KP881921264), T7S, T7R); T7A = VFNMS(LDK(KP881921264), T7z, T7k); T7M = VFMA(LDK(KP881921264), T7z, T7k); { V T8i, T8m, T8d, T8f; T8i = VFMA(LDK(KP956940335), T8h, T8g); T8m = VFNMS(LDK(KP956940335), T8h, T8g); T8d = VFNMS(LDK(KP956940335), T8c, T89); T8f = VFMA(LDK(KP956940335), T8c, T89); { V T8l, T8n, T86, T8e; T8l = VFMA(LDK(KP956940335), T8k, T8j); T8n = VFNMS(LDK(KP956940335), T8k, T8j); T86 = VFNMS(LDK(KP956940335), T85, T7Y); T8e = VFMA(LDK(KP956940335), T85, T7Y); ST(&(x[WS(rs, 53)]), VFNMSI(T7V, T7U), ms, &(x[WS(rs, 1)])); ST(&(x[WS(rs, 11)]), VFMAI(T7V, T7U), ms, &(x[WS(rs, 1)])); ST(&(x[WS(rs, 43)]), VFMAI(T7T, T7Q), ms, &(x[WS(rs, 1)])); ST(&(x[WS(rs, 21)]), VFNMSI(T7T, T7Q), ms, &(x[WS(rs, 1)])); ST(&(x[WS(rs, 59)]), VFMAI(T7N, T7M), ms, &(x[WS(rs, 1)])); ST(&(x[WS(rs, 5)]), VFNMSI(T7N, T7M), ms, &(x[WS(rs, 1)])); ST(&(x[WS(rs, 27)]), VFMAI(T7L, T7A), ms, &(x[WS(rs, 1)])); ST(&(x[WS(rs, 37)]), VFNMSI(T7L, T7A), ms, &(x[WS(rs, 1)])); ST(&(x[WS(rs, 51)]), VFMAI(T8n, T8m), ms, &(x[WS(rs, 1)])); ST(&(x[WS(rs, 13)]), VFNMSI(T8n, T8m), ms, &(x[WS(rs, 1)])); ST(&(x[WS(rs, 45)]), VFNMSI(T8l, T8i), ms, &(x[WS(rs, 1)])); ST(&(x[WS(rs, 19)]), VFMAI(T8l, T8i), ms, &(x[WS(rs, 1)])); ST(&(x[WS(rs, 3)]), VFMAI(T8f, T8e), ms, &(x[WS(rs, 1)])); ST(&(x[WS(rs, 61)]), VFNMSI(T8f, T8e), ms, &(x[WS(rs, 1)])); ST(&(x[WS(rs, 35)]), VFMAI(T8d, T86), ms, &(x[WS(rs, 1)])); ST(&(x[WS(rs, 29)]), VFNMSI(T8d, T86), ms, &(x[WS(rs, 1)])); T6r = VFMA(LDK(KP198912367), T5u, T5x); T5y = VFNMS(LDK(KP198912367), T5x, T5u); } } } } } } } } { V T5N, T5U, T68, T5z, T6U, T6f; T5N = VFMA(LDK(KP923879532), T5M, T5F); T6L = VFNMS(LDK(KP923879532), T5M, T5F); T6M = VFNMS(LDK(KP923879532), T5T, T5Q); T5U = VFMA(LDK(KP923879532), T5T, T5Q); T68 = VFMA(LDK(KP923879532), T67, T60); T6O = VFNMS(LDK(KP923879532), T67, T60); T5z = VADD(T5p, T5y); T6U = VSUB(T5y, T5p); T6P = VFNMS(LDK(KP923879532), T6e, T6b); T6f = VFMA(LDK(KP923879532), T6e, T6b); { V T5V, T6u, T6g, T6v, T6s, T6J; T6s = VSUB(T6q, T6r); T6J = VADD(T6q, T6r); T5V = VFNMS(LDK(KP098491403), T5U, T5N); T6u = VFMA(LDK(KP098491403), T5N, T5U); T75 = VFNMS(LDK(KP980785280), T6U, T6T); T6V = VFMA(LDK(KP980785280), T6U, T6T); T5A = VFMA(LDK(KP980785280), T5z, T5g); T6A = VFNMS(LDK(KP980785280), T5z, T5g); T6g = VFNMS(LDK(KP098491403), T6f, T68); T6v = VFMA(LDK(KP098491403), T68, T6f); T72 = VFNMS(LDK(KP980785280), T6J, T6I); T6K = VFMA(LDK(KP980785280), T6J, T6I); T6t = VFMA(LDK(KP980785280), T6s, T6p); T6D = VFNMS(LDK(KP980785280), T6s, T6p); T6w = VSUB(T6u, T6v); T6B = VADD(T6u, T6v); T6h = VADD(T5V, T6g); T6E = VSUB(T6g, T5V); } } } } } } } { V T6W, T6N, T6G, T6C, T6z, T6x, T6H, T6F, T6y, T6i, T6X, T6Q; T6W = VFNMS(LDK(KP820678790), T6L, T6M); T6N = VFMA(LDK(KP820678790), T6M, T6L); T6G = VFMA(LDK(KP995184726), T6B, T6A); T6C = VFNMS(LDK(KP995184726), T6B, T6A); T6z = VFMA(LDK(KP995184726), T6w, T6t); T6x = VFNMS(LDK(KP995184726), T6w, T6t); T6H = VFMA(LDK(KP995184726), T6E, T6D); T6F = VFNMS(LDK(KP995184726), T6E, T6D); T6y = VFMA(LDK(KP995184726), T6h, T5A); T6i = VFNMS(LDK(KP995184726), T6h, T5A); T6X = VFNMS(LDK(KP820678790), T6O, T6P); T6Q = VFMA(LDK(KP820678790), T6P, T6O); { V T73, T6Y, T76, T6R; ST(&(x[WS(rs, 49)]), VFNMSI(T6H, T6G), ms, &(x[WS(rs, 1)])); ST(&(x[WS(rs, 15)]), VFMAI(T6H, T6G), ms, &(x[WS(rs, 1)])); ST(&(x[WS(rs, 47)]), VFMAI(T6F, T6C), ms, &(x[WS(rs, 1)])); ST(&(x[WS(rs, 17)]), VFNMSI(T6F, T6C), ms, &(x[WS(rs, 1)])); ST(&(x[WS(rs, 63)]), VFMAI(T6z, T6y), ms, &(x[WS(rs, 1)])); ST(&(x[WS(rs, 1)]), VFNMSI(T6z, T6y), ms, &(x[WS(rs, 1)])); ST(&(x[WS(rs, 31)]), VFMAI(T6x, T6i), ms, &(x[WS(rs, 1)])); ST(&(x[WS(rs, 33)]), VFNMSI(T6x, T6i), ms, &(x[WS(rs, 1)])); T73 = VADD(T6W, T6X); T6Y = VSUB(T6W, T6X); T76 = VSUB(T6Q, T6N); T6R = VADD(T6N, T6Q); { V T78, T74, T71, T6Z, T79, T77, T70, T6S; T78 = VFNMS(LDK(KP773010453), T73, T72); T74 = VFMA(LDK(KP773010453), T73, T72); T71 = VFMA(LDK(KP773010453), T6Y, T6V); T6Z = VFNMS(LDK(KP773010453), T6Y, T6V); T79 = VFNMS(LDK(KP773010453), T76, T75); T77 = VFMA(LDK(KP773010453), T76, T75); T70 = VFMA(LDK(KP773010453), T6R, T6K); T6S = VFNMS(LDK(KP773010453), T6R, T6K); ST(&(x[WS(rs, 55)]), VFMAI(T79, T78), ms, &(x[WS(rs, 1)])); ST(&(x[WS(rs, 9)]), VFNMSI(T79, T78), ms, &(x[WS(rs, 1)])); ST(&(x[WS(rs, 41)]), VFNMSI(T77, T74), ms, &(x[WS(rs, 1)])); ST(&(x[WS(rs, 23)]), VFMAI(T77, T74), ms, &(x[WS(rs, 1)])); ST(&(x[WS(rs, 7)]), VFMAI(T71, T70), ms, &(x[WS(rs, 1)])); ST(&(x[WS(rs, 57)]), VFNMSI(T71, T70), ms, &(x[WS(rs, 1)])); ST(&(x[WS(rs, 39)]), VFMAI(T6Z, T6S), ms, &(x[WS(rs, 1)])); ST(&(x[WS(rs, 25)]), VFNMSI(T6Z, T6S), ms, &(x[WS(rs, 1)])); } } } } } VLEAVE(); } static const tw_instr twinstr[] = { VTW(0, 1), VTW(0, 2), VTW(0, 3), VTW(0, 4), VTW(0, 5), VTW(0, 6), VTW(0, 7), VTW(0, 8), VTW(0, 9), VTW(0, 10), VTW(0, 11), VTW(0, 12), VTW(0, 13), VTW(0, 14), VTW(0, 15), VTW(0, 16), VTW(0, 17), VTW(0, 18), VTW(0, 19), VTW(0, 20), VTW(0, 21), VTW(0, 22), VTW(0, 23), VTW(0, 24), VTW(0, 25), VTW(0, 26), VTW(0, 27), VTW(0, 28), VTW(0, 29), VTW(0, 30), VTW(0, 31), VTW(0, 32), VTW(0, 33), VTW(0, 34), VTW(0, 35), VTW(0, 36), VTW(0, 37), VTW(0, 38), VTW(0, 39), VTW(0, 40), VTW(0, 41), VTW(0, 42), VTW(0, 43), VTW(0, 44), VTW(0, 45), VTW(0, 46), VTW(0, 47), VTW(0, 48), VTW(0, 49), VTW(0, 50), VTW(0, 51), VTW(0, 52), VTW(0, 53), VTW(0, 54), VTW(0, 55), VTW(0, 56), VTW(0, 57), VTW(0, 58), VTW(0, 59), VTW(0, 60), VTW(0, 61), VTW(0, 62), VTW(0, 63), {TW_NEXT, VL, 0} }; static const ct_desc desc = { 64, XSIMD_STRING("t1fv_64"), twinstr, &GENUS, {261, 126, 258, 0}, 0, 0, 0 }; void XSIMD(codelet_t1fv_64) (planner *p) { X(kdft_dit_register) (p, t1fv_64, &desc); } #else /* HAVE_FMA */ /* Generated by: ../../../genfft/gen_twiddle_c.native -simd -compact -variables 4 -pipeline-latency 8 -n 64 -name t1fv_64 -include t1f.h */ /* * This function contains 519 FP additions, 250 FP multiplications, * (or, 467 additions, 198 multiplications, 52 fused multiply/add), * 107 stack variables, 15 constants, and 128 memory accesses */ #include "t1f.h" static void t1fv_64(R *ri, R *ii, const R *W, stride rs, INT mb, INT me, INT ms) { DVK(KP995184726, +0.995184726672196886244836953109479921575474869); DVK(KP098017140, +0.098017140329560601994195563888641845861136673); DVK(KP773010453, +0.773010453362736960810906609758469800971041293); DVK(KP634393284, +0.634393284163645498215171613225493370675687095); DVK(KP471396736, +0.471396736825997648556387625905254377657460319); DVK(KP881921264, +0.881921264348355029712756863660388349508442621); DVK(KP290284677, +0.290284677254462367636192375817395274691476278); DVK(KP956940335, +0.956940335732208864935797886980269969482849206); DVK(KP555570233, +0.555570233019602224742830813948532874374937191); DVK(KP831469612, +0.831469612302545237078788377617905756738560812); DVK(KP195090322, +0.195090322016128267848284868477022240927691618); DVK(KP980785280, +0.980785280403230449126182236134239036973933731); DVK(KP382683432, +0.382683432365089771728459984030398866761344562); DVK(KP923879532, +0.923879532511286756128183189396788286822416626); DVK(KP707106781, +0.707106781186547524400844362104849039284835938); { INT m; R *x; x = ri; for (m = mb, W = W + (mb * ((TWVL / VL) * 126)); m < me; m = m + VL, x = x + (VL * ms), W = W + (TWVL * 126), MAKE_VOLATILE_STRIDE(64, rs)) { V Tg, T4a, T6r, T7f, T3o, T4B, T5q, T7e, T5R, T62, T28, T4o, T2g, T4l, T7n; V T7Z, T68, T6j, T2C, T4s, T3a, T4v, T7u, T82, T7E, T7F, T7V, T5F, T6u, T1k; V T4e, T1r, T4d, T7B, T7C, T7W, T5M, T6v, TV, T4g, T12, T4h, T7h, T7i, TD; V T4C, T3h, T4b, T5x, T6s, T1R, T4m, T7q, T80, T2j, T4p, T5Y, T63, T2Z, T4w; V T7x, T83, T33, T4t, T6f, T6k; { V T1, T3, T3m, T3k, Tb, Td, Te, T6, T8, T9, T2, T3l, T3j; T1 = LD(&(x[0]), ms, &(x[0])); T2 = LD(&(x[WS(rs, 32)]), ms, &(x[0])); T3 = BYTWJ(&(W[TWVL * 62]), T2); T3l = LD(&(x[WS(rs, 48)]), ms, &(x[0])); T3m = BYTWJ(&(W[TWVL * 94]), T3l); T3j = LD(&(x[WS(rs, 16)]), ms, &(x[0])); T3k = BYTWJ(&(W[TWVL * 30]), T3j); { V Ta, Tc, T5, T7; Ta = LD(&(x[WS(rs, 56)]), ms, &(x[0])); Tb = BYTWJ(&(W[TWVL * 110]), Ta); Tc = LD(&(x[WS(rs, 24)]), ms, &(x[0])); Td = BYTWJ(&(W[TWVL * 46]), Tc); Te = VSUB(Tb, Td); T5 = LD(&(x[WS(rs, 8)]), ms, &(x[0])); T6 = BYTWJ(&(W[TWVL * 14]), T5); T7 = LD(&(x[WS(rs, 40)]), ms, &(x[0])); T8 = BYTWJ(&(W[TWVL * 78]), T7); T9 = VSUB(T6, T8); } { V T4, Tf, T6p, T6q; T4 = VSUB(T1, T3); Tf = VMUL(LDK(KP707106781), VADD(T9, Te)); Tg = VADD(T4, Tf); T4a = VSUB(T4, Tf); T6p = VADD(Tb, Td); T6q = VADD(T6, T8); T6r = VSUB(T6p, T6q); T7f = VADD(T6q, T6p); } { V T3i, T3n, T5o, T5p; T3i = VMUL(LDK(KP707106781), VSUB(Te, T9)); T3n = VSUB(T3k, T3m); T3o = VSUB(T3i, T3n); T4B = VADD(T3n, T3i); T5o = VADD(T1, T3); T5p = VADD(T3k, T3m); T5q = VSUB(T5o, T5p); T7e = VADD(T5o, T5p); } } { V T24, T26, T5Q, T2b, T2d, T5P, T1W, T60, T21, T61, T22, T27; { V T23, T25, T2a, T2c; T23 = LD(&(x[WS(rs, 17)]), ms, &(x[WS(rs, 1)])); T24 = BYTWJ(&(W[TWVL * 32]), T23); T25 = LD(&(x[WS(rs, 49)]), ms, &(x[WS(rs, 1)])); T26 = BYTWJ(&(W[TWVL * 96]), T25); T5Q = VADD(T24, T26); T2a = LD(&(x[WS(rs, 1)]), ms, &(x[WS(rs, 1)])); T2b = BYTWJ(&(W[0]), T2a); T2c = LD(&(x[WS(rs, 33)]), ms, &(x[WS(rs, 1)])); T2d = BYTWJ(&(W[TWVL * 64]), T2c); T5P = VADD(T2b, T2d); } { V T1T, T1V, T1S, T1U; T1S = LD(&(x[WS(rs, 57)]), ms, &(x[WS(rs, 1)])); T1T = BYTWJ(&(W[TWVL * 112]), T1S); T1U = LD(&(x[WS(rs, 25)]), ms, &(x[WS(rs, 1)])); T1V = BYTWJ(&(W[TWVL * 48]), T1U); T1W = VSUB(T1T, T1V); T60 = VADD(T1T, T1V); } { V T1Y, T20, T1X, T1Z; T1X = LD(&(x[WS(rs, 9)]), ms, &(x[WS(rs, 1)])); T1Y = BYTWJ(&(W[TWVL * 16]), T1X); T1Z = LD(&(x[WS(rs, 41)]), ms, &(x[WS(rs, 1)])); T20 = BYTWJ(&(W[TWVL * 80]), T1Z); T21 = VSUB(T1Y, T20); T61 = VADD(T1Y, T20); } T5R = VSUB(T5P, T5Q); T62 = VSUB(T60, T61); T22 = VMUL(LDK(KP707106781), VSUB(T1W, T21)); T27 = VSUB(T24, T26); T28 = VSUB(T22, T27); T4o = VADD(T27, T22); { V T2e, T2f, T7l, T7m; T2e = VSUB(T2b, T2d); T2f = VMUL(LDK(KP707106781), VADD(T21, T1W)); T2g = VADD(T2e, T2f); T4l = VSUB(T2e, T2f); T7l = VADD(T5P, T5Q); T7m = VADD(T61, T60); T7n = VADD(T7l, T7m); T7Z = VSUB(T7l, T7m); } } { V T2n, T2p, T66, T36, T38, T67, T2v, T6i, T2A, T6h, T2q, T2B; { V T2m, T2o, T35, T37; T2m = LD(&(x[WS(rs, 63)]), ms, &(x[WS(rs, 1)])); T2n = BYTWJ(&(W[TWVL * 124]), T2m); T2o = LD(&(x[WS(rs, 31)]), ms, &(x[WS(rs, 1)])); T2p = BYTWJ(&(W[TWVL * 60]), T2o); T66 = VADD(T2n, T2p); T35 = LD(&(x[WS(rs, 15)]), ms, &(x[WS(rs, 1)])); T36 = BYTWJ(&(W[TWVL * 28]), T35); T37 = LD(&(x[WS(rs, 47)]), ms, &(x[WS(rs, 1)])); T38 = BYTWJ(&(W[TWVL * 92]), T37); T67 = VADD(T36, T38); } { V T2s, T2u, T2r, T2t; T2r = LD(&(x[WS(rs, 7)]), ms, &(x[WS(rs, 1)])); T2s = BYTWJ(&(W[TWVL * 12]), T2r); T2t = LD(&(x[WS(rs, 39)]), ms, &(x[WS(rs, 1)])); T2u = BYTWJ(&(W[TWVL * 76]), T2t); T2v = VSUB(T2s, T2u); T6i = VADD(T2s, T2u); } { V T2x, T2z, T2w, T2y; T2w = LD(&(x[WS(rs, 55)]), ms, &(x[WS(rs, 1)])); T2x = BYTWJ(&(W[TWVL * 108]), T2w); T2y = LD(&(x[WS(rs, 23)]), ms, &(x[WS(rs, 1)])); T2z = BYTWJ(&(W[TWVL * 44]), T2y); T2A = VSUB(T2x, T2z); T6h = VADD(T2x, T2z); } T68 = VSUB(T66, T67); T6j = VSUB(T6h, T6i); T2q = VSUB(T2n, T2p); T2B = VMUL(LDK(KP707106781), VADD(T2v, T2A)); T2C = VADD(T2q, T2B); T4s = VSUB(T2q, T2B); { V T34, T39, T7s, T7t; T34 = VMUL(LDK(KP707106781), VSUB(T2A, T2v)); T39 = VSUB(T36, T38); T3a = VSUB(T34, T39); T4v = VADD(T39, T34); T7s = VADD(T66, T67); T7t = VADD(T6i, T6h); T7u = VADD(T7s, T7t); T82 = VSUB(T7s, T7t); } } { V T1g, T1i, T5A, T1m, T1o, T5z, T18, T5C, T1d, T5D, T5B, T5E; { V T1f, T1h, T1l, T1n; T1f = LD(&(x[WS(rs, 18)]), ms, &(x[0])); T1g = BYTWJ(&(W[TWVL * 34]), T1f); T1h = LD(&(x[WS(rs, 50)]), ms, &(x[0])); T1i = BYTWJ(&(W[TWVL * 98]), T1h); T5A = VADD(T1g, T1i); T1l = LD(&(x[WS(rs, 2)]), ms, &(x[0])); T1m = BYTWJ(&(W[TWVL * 2]), T1l); T1n = LD(&(x[WS(rs, 34)]), ms, &(x[0])); T1o = BYTWJ(&(W[TWVL * 66]), T1n); T5z = VADD(T1m, T1o); } { V T15, T17, T14, T16; T14 = LD(&(x[WS(rs, 58)]), ms, &(x[0])); T15 = BYTWJ(&(W[TWVL * 114]), T14); T16 = LD(&(x[WS(rs, 26)]), ms, &(x[0])); T17 = BYTWJ(&(W[TWVL * 50]), T16); T18 = VSUB(T15, T17); T5C = VADD(T15, T17); } { V T1a, T1c, T19, T1b; T19 = LD(&(x[WS(rs, 10)]), ms, &(x[0])); T1a = BYTWJ(&(W[TWVL * 18]), T19); T1b = LD(&(x[WS(rs, 42)]), ms, &(x[0])); T1c = BYTWJ(&(W[TWVL * 82]), T1b); T1d = VSUB(T1a, T1c); T5D = VADD(T1a, T1c); } T7E = VADD(T5z, T5A); T7F = VADD(T5D, T5C); T7V = VSUB(T7E, T7F); T5B = VSUB(T5z, T5A); T5E = VSUB(T5C, T5D); T5F = VFMA(LDK(KP923879532), T5B, VMUL(LDK(KP382683432), T5E)); T6u = VFNMS(LDK(KP382683432), T5B, VMUL(LDK(KP923879532), T5E)); { V T1e, T1j, T1p, T1q; T1e = VMUL(LDK(KP707106781), VSUB(T18, T1d)); T1j = VSUB(T1g, T1i); T1k = VSUB(T1e, T1j); T4e = VADD(T1j, T1e); T1p = VSUB(T1m, T1o); T1q = VMUL(LDK(KP707106781), VADD(T1d, T18)); T1r = VADD(T1p, T1q); T4d = VSUB(T1p, T1q); } } { V TG, TI, T5G, TY, T10, T5H, TO, T5K, TT, T5J, T5I, T5L; { V TF, TH, TX, TZ; TF = LD(&(x[WS(rs, 62)]), ms, &(x[0])); TG = BYTWJ(&(W[TWVL * 122]), TF); TH = LD(&(x[WS(rs, 30)]), ms, &(x[0])); TI = BYTWJ(&(W[TWVL * 58]), TH); T5G = VADD(TG, TI); TX = LD(&(x[WS(rs, 14)]), ms, &(x[0])); TY = BYTWJ(&(W[TWVL * 26]), TX); TZ = LD(&(x[WS(rs, 46)]), ms, &(x[0])); T10 = BYTWJ(&(W[TWVL * 90]), TZ); T5H = VADD(TY, T10); } { V TL, TN, TK, TM; TK = LD(&(x[WS(rs, 6)]), ms, &(x[0])); TL = BYTWJ(&(W[TWVL * 10]), TK); TM = LD(&(x[WS(rs, 38)]), ms, &(x[0])); TN = BYTWJ(&(W[TWVL * 74]), TM); TO = VSUB(TL, TN); T5K = VADD(TL, TN); } { V TQ, TS, TP, TR; TP = LD(&(x[WS(rs, 54)]), ms, &(x[0])); TQ = BYTWJ(&(W[TWVL * 106]), TP); TR = LD(&(x[WS(rs, 22)]), ms, &(x[0])); TS = BYTWJ(&(W[TWVL * 42]), TR); TT = VSUB(TQ, TS); T5J = VADD(TQ, TS); } T7B = VADD(T5G, T5H); T7C = VADD(T5K, T5J); T7W = VSUB(T7B, T7C); T5I = VSUB(T5G, T5H); T5L = VSUB(T5J, T5K); T5M = VFNMS(LDK(KP382683432), T5L, VMUL(LDK(KP923879532), T5I)); T6v = VFMA(LDK(KP382683432), T5I, VMUL(LDK(KP923879532), T5L)); { V TJ, TU, TW, T11; TJ = VSUB(TG, TI); TU = VMUL(LDK(KP707106781), VADD(TO, TT)); TV = VADD(TJ, TU); T4g = VSUB(TJ, TU); TW = VMUL(LDK(KP707106781), VSUB(TT, TO)); T11 = VSUB(TY, T10); T12 = VSUB(TW, T11); T4h = VADD(T11, TW); } } { V Tl, T5r, TB, T5v, Tq, T5s, Tw, T5u, Tr, TC; { V Ti, Tk, Th, Tj; Th = LD(&(x[WS(rs, 4)]), ms, &(x[0])); Ti = BYTWJ(&(W[TWVL * 6]), Th); Tj = LD(&(x[WS(rs, 36)]), ms, &(x[0])); Tk = BYTWJ(&(W[TWVL * 70]), Tj); Tl = VSUB(Ti, Tk); T5r = VADD(Ti, Tk); } { V Ty, TA, Tx, Tz; Tx = LD(&(x[WS(rs, 12)]), ms, &(x[0])); Ty = BYTWJ(&(W[TWVL * 22]), Tx); Tz = LD(&(x[WS(rs, 44)]), ms, &(x[0])); TA = BYTWJ(&(W[TWVL * 86]), Tz); TB = VSUB(Ty, TA); T5v = VADD(Ty, TA); } { V Tn, Tp, Tm, To; Tm = LD(&(x[WS(rs, 20)]), ms, &(x[0])); Tn = BYTWJ(&(W[TWVL * 38]), Tm); To = LD(&(x[WS(rs, 52)]), ms, &(x[0])); Tp = BYTWJ(&(W[TWVL * 102]), To); Tq = VSUB(Tn, Tp); T5s = VADD(Tn, Tp); } { V Tt, Tv, Ts, Tu; Ts = LD(&(x[WS(rs, 60)]), ms, &(x[0])); Tt = BYTWJ(&(W[TWVL * 118]), Ts); Tu = LD(&(x[WS(rs, 28)]), ms, &(x[0])); Tv = BYTWJ(&(W[TWVL * 54]), Tu); Tw = VSUB(Tt, Tv); T5u = VADD(Tt, Tv); } T7h = VADD(T5r, T5s); T7i = VADD(T5u, T5v); Tr = VFNMS(LDK(KP382683432), Tq, VMUL(LDK(KP923879532), Tl)); TC = VFMA(LDK(KP923879532), Tw, VMUL(LDK(KP382683432), TB)); TD = VADD(Tr, TC); T4C = VSUB(TC, Tr); { V T3f, T3g, T5t, T5w; T3f = VFNMS(LDK(KP923879532), TB, VMUL(LDK(KP382683432), Tw)); T3g = VFMA(LDK(KP382683432), Tl, VMUL(LDK(KP923879532), Tq)); T3h = VSUB(T3f, T3g); T4b = VADD(T3g, T3f); T5t = VSUB(T5r, T5s); T5w = VSUB(T5u, T5v); T5x = VMUL(LDK(KP707106781), VADD(T5t, T5w)); T6s = VMUL(LDK(KP707106781), VSUB(T5w, T5t)); } } { V T1z, T5V, T1P, T5T, T1E, T5W, T1K, T5S; { V T1w, T1y, T1v, T1x; T1v = LD(&(x[WS(rs, 61)]), ms, &(x[WS(rs, 1)])); T1w = BYTWJ(&(W[TWVL * 120]), T1v); T1x = LD(&(x[WS(rs, 29)]), ms, &(x[WS(rs, 1)])); T1y = BYTWJ(&(W[TWVL * 56]), T1x); T1z = VSUB(T1w, T1y); T5V = VADD(T1w, T1y); } { V T1M, T1O, T1L, T1N; T1L = LD(&(x[WS(rs, 21)]), ms, &(x[WS(rs, 1)])); T1M = BYTWJ(&(W[TWVL * 40]), T1L); T1N = LD(&(x[WS(rs, 53)]), ms, &(x[WS(rs, 1)])); T1O = BYTWJ(&(W[TWVL * 104]), T1N); T1P = VSUB(T1M, T1O); T5T = VADD(T1M, T1O); } { V T1B, T1D, T1A, T1C; T1A = LD(&(x[WS(rs, 13)]), ms, &(x[WS(rs, 1)])); T1B = BYTWJ(&(W[TWVL * 24]), T1A); T1C = LD(&(x[WS(rs, 45)]), ms, &(x[WS(rs, 1)])); T1D = BYTWJ(&(W[TWVL * 88]), T1C); T1E = VSUB(T1B, T1D); T5W = VADD(T1B, T1D); } { V T1H, T1J, T1G, T1I; T1G = LD(&(x[WS(rs, 5)]), ms, &(x[WS(rs, 1)])); T1H = BYTWJ(&(W[TWVL * 8]), T1G); T1I = LD(&(x[WS(rs, 37)]), ms, &(x[WS(rs, 1)])); T1J = BYTWJ(&(W[TWVL * 72]), T1I); T1K = VSUB(T1H, T1J); T5S = VADD(T1H, T1J); } { V T1F, T1Q, T7o, T7p; T1F = VFNMS(LDK(KP923879532), T1E, VMUL(LDK(KP382683432), T1z)); T1Q = VFMA(LDK(KP382683432), T1K, VMUL(LDK(KP923879532), T1P)); T1R = VSUB(T1F, T1Q); T4m = VADD(T1Q, T1F); T7o = VADD(T5S, T5T); T7p = VADD(T5V, T5W); T7q = VADD(T7o, T7p); T80 = VSUB(T7p, T7o); } { V T2h, T2i, T5U, T5X; T2h = VFNMS(LDK(KP382683432), T1P, VMUL(LDK(KP923879532), T1K)); T2i = VFMA(LDK(KP923879532), T1z, VMUL(LDK(KP382683432), T1E)); T2j = VADD(T2h, T2i); T4p = VSUB(T2i, T2h); T5U = VSUB(T5S, T5T); T5X = VSUB(T5V, T5W); T5Y = VMUL(LDK(KP707106781), VADD(T5U, T5X)); T63 = VMUL(LDK(KP707106781), VSUB(T5X, T5U)); } } { V T2H, T69, T2X, T6d, T2M, T6a, T2S, T6c; { V T2E, T2G, T2D, T2F; T2D = LD(&(x[WS(rs, 3)]), ms, &(x[WS(rs, 1)])); T2E = BYTWJ(&(W[TWVL * 4]), T2D); T2F = LD(&(x[WS(rs, 35)]), ms, &(x[WS(rs, 1)])); T2G = BYTWJ(&(W[TWVL * 68]), T2F); T2H = VSUB(T2E, T2G); T69 = VADD(T2E, T2G); } { V T2U, T2W, T2T, T2V; T2T = LD(&(x[WS(rs, 11)]), ms, &(x[WS(rs, 1)])); T2U = BYTWJ(&(W[TWVL * 20]), T2T); T2V = LD(&(x[WS(rs, 43)]), ms, &(x[WS(rs, 1)])); T2W = BYTWJ(&(W[TWVL * 84]), T2V); T2X = VSUB(T2U, T2W); T6d = VADD(T2U, T2W); } { V T2J, T2L, T2I, T2K; T2I = LD(&(x[WS(rs, 19)]), ms, &(x[WS(rs, 1)])); T2J = BYTWJ(&(W[TWVL * 36]), T2I); T2K = LD(&(x[WS(rs, 51)]), ms, &(x[WS(rs, 1)])); T2L = BYTWJ(&(W[TWVL * 100]), T2K); T2M = VSUB(T2J, T2L); T6a = VADD(T2J, T2L); } { V T2P, T2R, T2O, T2Q; T2O = LD(&(x[WS(rs, 59)]), ms, &(x[WS(rs, 1)])); T2P = BYTWJ(&(W[TWVL * 116]), T2O); T2Q = LD(&(x[WS(rs, 27)]), ms, &(x[WS(rs, 1)])); T2R = BYTWJ(&(W[TWVL * 52]), T2Q); T2S = VSUB(T2P, T2R); T6c = VADD(T2P, T2R); } { V T2N, T2Y, T7v, T7w; T2N = VFNMS(LDK(KP382683432), T2M, VMUL(LDK(KP923879532), T2H)); T2Y = VFMA(LDK(KP923879532), T2S, VMUL(LDK(KP382683432), T2X)); T2Z = VADD(T2N, T2Y); T4w = VSUB(T2Y, T2N); T7v = VADD(T69, T6a); T7w = VADD(T6c, T6d); T7x = VADD(T7v, T7w); T83 = VSUB(T7w, T7v); } { V T31, T32, T6b, T6e; T31 = VFNMS(LDK(KP923879532), T2X, VMUL(LDK(KP382683432), T2S)); T32 = VFMA(LDK(KP382683432), T2H, VMUL(LDK(KP923879532), T2M)); T33 = VSUB(T31, T32); T4t = VADD(T32, T31); T6b = VSUB(T69, T6a); T6e = VSUB(T6c, T6d); T6f = VMUL(LDK(KP707106781), VADD(T6b, T6e)); T6k = VMUL(LDK(KP707106781), VSUB(T6e, T6b)); } } { V T7k, T7M, T7R, T7T, T7z, T7I, T7H, T7N, T7O, T7S; { V T7g, T7j, T7P, T7Q; T7g = VADD(T7e, T7f); T7j = VADD(T7h, T7i); T7k = VSUB(T7g, T7j); T7M = VADD(T7g, T7j); T7P = VADD(T7n, T7q); T7Q = VADD(T7u, T7x); T7R = VADD(T7P, T7Q); T7T = VBYI(VSUB(T7Q, T7P)); } { V T7r, T7y, T7D, T7G; T7r = VSUB(T7n, T7q); T7y = VSUB(T7u, T7x); T7z = VMUL(LDK(KP707106781), VADD(T7r, T7y)); T7I = VMUL(LDK(KP707106781), VSUB(T7y, T7r)); T7D = VADD(T7B, T7C); T7G = VADD(T7E, T7F); T7H = VSUB(T7D, T7G); T7N = VADD(T7G, T7D); } T7O = VADD(T7M, T7N); ST(&(x[WS(rs, 32)]), VSUB(T7O, T7R), ms, &(x[0])); ST(&(x[0]), VADD(T7O, T7R), ms, &(x[0])); T7S = VSUB(T7M, T7N); ST(&(x[WS(rs, 48)]), VSUB(T7S, T7T), ms, &(x[0])); ST(&(x[WS(rs, 16)]), VADD(T7S, T7T), ms, &(x[0])); { V T7A, T7J, T7K, T7L; T7A = VADD(T7k, T7z); T7J = VBYI(VADD(T7H, T7I)); ST(&(x[WS(rs, 56)]), VSUB(T7A, T7J), ms, &(x[0])); ST(&(x[WS(rs, 8)]), VADD(T7A, T7J), ms, &(x[0])); T7K = VSUB(T7k, T7z); T7L = VBYI(VSUB(T7I, T7H)); ST(&(x[WS(rs, 40)]), VSUB(T7K, T7L), ms, &(x[0])); ST(&(x[WS(rs, 24)]), VADD(T7K, T7L), ms, &(x[0])); } } { V T7Y, T8j, T8c, T8k, T85, T8g, T89, T8h; { V T7U, T7X, T8a, T8b; T7U = VSUB(T7e, T7f); T7X = VMUL(LDK(KP707106781), VADD(T7V, T7W)); T7Y = VADD(T7U, T7X); T8j = VSUB(T7U, T7X); T8a = VFNMS(LDK(KP382683432), T7Z, VMUL(LDK(KP923879532), T80)); T8b = VFMA(LDK(KP382683432), T82, VMUL(LDK(KP923879532), T83)); T8c = VADD(T8a, T8b); T8k = VSUB(T8b, T8a); } { V T81, T84, T87, T88; T81 = VFMA(LDK(KP923879532), T7Z, VMUL(LDK(KP382683432), T80)); T84 = VFNMS(LDK(KP382683432), T83, VMUL(LDK(KP923879532), T82)); T85 = VADD(T81, T84); T8g = VSUB(T84, T81); T87 = VSUB(T7i, T7h); T88 = VMUL(LDK(KP707106781), VSUB(T7W, T7V)); T89 = VADD(T87, T88); T8h = VSUB(T88, T87); } { V T86, T8d, T8m, T8n; T86 = VADD(T7Y, T85); T8d = VBYI(VADD(T89, T8c)); ST(&(x[WS(rs, 60)]), VSUB(T86, T8d), ms, &(x[0])); ST(&(x[WS(rs, 4)]), VADD(T86, T8d), ms, &(x[0])); T8m = VBYI(VADD(T8h, T8g)); T8n = VADD(T8j, T8k); ST(&(x[WS(rs, 12)]), VADD(T8m, T8n), ms, &(x[0])); ST(&(x[WS(rs, 52)]), VSUB(T8n, T8m), ms, &(x[0])); } { V T8e, T8f, T8i, T8l; T8e = VSUB(T7Y, T85); T8f = VBYI(VSUB(T8c, T89)); ST(&(x[WS(rs, 36)]), VSUB(T8e, T8f), ms, &(x[0])); ST(&(x[WS(rs, 28)]), VADD(T8e, T8f), ms, &(x[0])); T8i = VBYI(VSUB(T8g, T8h)); T8l = VSUB(T8j, T8k); ST(&(x[WS(rs, 20)]), VADD(T8i, T8l), ms, &(x[0])); ST(&(x[WS(rs, 44)]), VSUB(T8l, T8i), ms, &(x[0])); } } { V T5O, T6H, T6x, T6F, T6n, T6I, T6A, T6E; { V T5y, T5N, T6t, T6w; T5y = VADD(T5q, T5x); T5N = VADD(T5F, T5M); T5O = VADD(T5y, T5N); T6H = VSUB(T5y, T5N); T6t = VADD(T6r, T6s); T6w = VADD(T6u, T6v); T6x = VADD(T6t, T6w); T6F = VSUB(T6w, T6t); { V T65, T6y, T6m, T6z; { V T5Z, T64, T6g, T6l; T5Z = VADD(T5R, T5Y); T64 = VADD(T62, T63); T65 = VFMA(LDK(KP980785280), T5Z, VMUL(LDK(KP195090322), T64)); T6y = VFNMS(LDK(KP195090322), T5Z, VMUL(LDK(KP980785280), T64)); T6g = VADD(T68, T6f); T6l = VADD(T6j, T6k); T6m = VFNMS(LDK(KP195090322), T6l, VMUL(LDK(KP980785280), T6g)); T6z = VFMA(LDK(KP195090322), T6g, VMUL(LDK(KP980785280), T6l)); } T6n = VADD(T65, T6m); T6I = VSUB(T6z, T6y); T6A = VADD(T6y, T6z); T6E = VSUB(T6m, T65); } } { V T6o, T6B, T6K, T6L; T6o = VADD(T5O, T6n); T6B = VBYI(VADD(T6x, T6A)); ST(&(x[WS(rs, 62)]), VSUB(T6o, T6B), ms, &(x[0])); ST(&(x[WS(rs, 2)]), VADD(T6o, T6B), ms, &(x[0])); T6K = VBYI(VADD(T6F, T6E)); T6L = VADD(T6H, T6I); ST(&(x[WS(rs, 14)]), VADD(T6K, T6L), ms, &(x[0])); ST(&(x[WS(rs, 50)]), VSUB(T6L, T6K), ms, &(x[0])); } { V T6C, T6D, T6G, T6J; T6C = VSUB(T5O, T6n); T6D = VBYI(VSUB(T6A, T6x)); ST(&(x[WS(rs, 34)]), VSUB(T6C, T6D), ms, &(x[0])); ST(&(x[WS(rs, 30)]), VADD(T6C, T6D), ms, &(x[0])); T6G = VBYI(VSUB(T6E, T6F)); T6J = VSUB(T6H, T6I); ST(&(x[WS(rs, 18)]), VADD(T6G, T6J), ms, &(x[0])); ST(&(x[WS(rs, 46)]), VSUB(T6J, T6G), ms, &(x[0])); } } { V T6O, T79, T6Z, T77, T6V, T7a, T72, T76; { V T6M, T6N, T6X, T6Y; T6M = VSUB(T5q, T5x); T6N = VSUB(T6v, T6u); T6O = VADD(T6M, T6N); T79 = VSUB(T6M, T6N); T6X = VSUB(T6s, T6r); T6Y = VSUB(T5M, T5F); T6Z = VADD(T6X, T6Y); T77 = VSUB(T6Y, T6X); { V T6R, T70, T6U, T71; { V T6P, T6Q, T6S, T6T; T6P = VSUB(T5R, T5Y); T6Q = VSUB(T63, T62); T6R = VFMA(LDK(KP831469612), T6P, VMUL(LDK(KP555570233), T6Q)); T70 = VFNMS(LDK(KP555570233), T6P, VMUL(LDK(KP831469612), T6Q)); T6S = VSUB(T68, T6f); T6T = VSUB(T6k, T6j); T6U = VFNMS(LDK(KP555570233), T6T, VMUL(LDK(KP831469612), T6S)); T71 = VFMA(LDK(KP555570233), T6S, VMUL(LDK(KP831469612), T6T)); } T6V = VADD(T6R, T6U); T7a = VSUB(T71, T70); T72 = VADD(T70, T71); T76 = VSUB(T6U, T6R); } } { V T6W, T73, T7c, T7d; T6W = VADD(T6O, T6V); T73 = VBYI(VADD(T6Z, T72)); ST(&(x[WS(rs, 58)]), VSUB(T6W, T73), ms, &(x[0])); ST(&(x[WS(rs, 6)]), VADD(T6W, T73), ms, &(x[0])); T7c = VBYI(VADD(T77, T76)); T7d = VADD(T79, T7a); ST(&(x[WS(rs, 10)]), VADD(T7c, T7d), ms, &(x[0])); ST(&(x[WS(rs, 54)]), VSUB(T7d, T7c), ms, &(x[0])); } { V T74, T75, T78, T7b; T74 = VSUB(T6O, T6V); T75 = VBYI(VSUB(T72, T6Z)); ST(&(x[WS(rs, 38)]), VSUB(T74, T75), ms, &(x[0])); ST(&(x[WS(rs, 26)]), VADD(T74, T75), ms, &(x[0])); T78 = VBYI(VSUB(T76, T77)); T7b = VSUB(T79, T7a); ST(&(x[WS(rs, 22)]), VADD(T78, T7b), ms, &(x[0])); ST(&(x[WS(rs, 42)]), VSUB(T7b, T78), ms, &(x[0])); } } { V T4k, T5h, T4R, T59, T4H, T5j, T4P, T4Y, T4z, T4S, T4K, T4O, T55, T5k, T5c; V T5g; { V T4c, T57, T4j, T58, T4f, T4i; T4c = VADD(T4a, T4b); T57 = VSUB(T4C, T4B); T4f = VFMA(LDK(KP831469612), T4d, VMUL(LDK(KP555570233), T4e)); T4i = VFNMS(LDK(KP555570233), T4h, VMUL(LDK(KP831469612), T4g)); T4j = VADD(T4f, T4i); T58 = VSUB(T4i, T4f); T4k = VADD(T4c, T4j); T5h = VSUB(T58, T57); T4R = VSUB(T4c, T4j); T59 = VADD(T57, T58); } { V T4D, T4W, T4G, T4X, T4E, T4F; T4D = VADD(T4B, T4C); T4W = VSUB(T4a, T4b); T4E = VFNMS(LDK(KP555570233), T4d, VMUL(LDK(KP831469612), T4e)); T4F = VFMA(LDK(KP555570233), T4g, VMUL(LDK(KP831469612), T4h)); T4G = VADD(T4E, T4F); T4X = VSUB(T4F, T4E); T4H = VADD(T4D, T4G); T5j = VSUB(T4W, T4X); T4P = VSUB(T4G, T4D); T4Y = VADD(T4W, T4X); } { V T4r, T4I, T4y, T4J; { V T4n, T4q, T4u, T4x; T4n = VADD(T4l, T4m); T4q = VADD(T4o, T4p); T4r = VFMA(LDK(KP956940335), T4n, VMUL(LDK(KP290284677), T4q)); T4I = VFNMS(LDK(KP290284677), T4n, VMUL(LDK(KP956940335), T4q)); T4u = VADD(T4s, T4t); T4x = VADD(T4v, T4w); T4y = VFNMS(LDK(KP290284677), T4x, VMUL(LDK(KP956940335), T4u)); T4J = VFMA(LDK(KP290284677), T4u, VMUL(LDK(KP956940335), T4x)); } T4z = VADD(T4r, T4y); T4S = VSUB(T4J, T4I); T4K = VADD(T4I, T4J); T4O = VSUB(T4y, T4r); } { V T51, T5a, T54, T5b; { V T4Z, T50, T52, T53; T4Z = VSUB(T4l, T4m); T50 = VSUB(T4p, T4o); T51 = VFMA(LDK(KP881921264), T4Z, VMUL(LDK(KP471396736), T50)); T5a = VFNMS(LDK(KP471396736), T4Z, VMUL(LDK(KP881921264), T50)); T52 = VSUB(T4s, T4t); T53 = VSUB(T4w, T4v); T54 = VFNMS(LDK(KP471396736), T53, VMUL(LDK(KP881921264), T52)); T5b = VFMA(LDK(KP471396736), T52, VMUL(LDK(KP881921264), T53)); } T55 = VADD(T51, T54); T5k = VSUB(T5b, T5a); T5c = VADD(T5a, T5b); T5g = VSUB(T54, T51); } { V T4A, T4L, T5i, T5l; T4A = VADD(T4k, T4z); T4L = VBYI(VADD(T4H, T4K)); ST(&(x[WS(rs, 61)]), VSUB(T4A, T4L), ms, &(x[WS(rs, 1)])); ST(&(x[WS(rs, 3)]), VADD(T4A, T4L), ms, &(x[WS(rs, 1)])); T5i = VBYI(VSUB(T5g, T5h)); T5l = VSUB(T5j, T5k); ST(&(x[WS(rs, 21)]), VADD(T5i, T5l), ms, &(x[WS(rs, 1)])); ST(&(x[WS(rs, 43)]), VSUB(T5l, T5i), ms, &(x[WS(rs, 1)])); } { V T5m, T5n, T4M, T4N; T5m = VBYI(VADD(T5h, T5g)); T5n = VADD(T5j, T5k); ST(&(x[WS(rs, 11)]), VADD(T5m, T5n), ms, &(x[WS(rs, 1)])); ST(&(x[WS(rs, 53)]), VSUB(T5n, T5m), ms, &(x[WS(rs, 1)])); T4M = VSUB(T4k, T4z); T4N = VBYI(VSUB(T4K, T4H)); ST(&(x[WS(rs, 35)]), VSUB(T4M, T4N), ms, &(x[WS(rs, 1)])); ST(&(x[WS(rs, 29)]), VADD(T4M, T4N), ms, &(x[WS(rs, 1)])); } { V T4Q, T4T, T56, T5d; T4Q = VBYI(VSUB(T4O, T4P)); T4T = VSUB(T4R, T4S); ST(&(x[WS(rs, 19)]), VADD(T4Q, T4T), ms, &(x[WS(rs, 1)])); ST(&(x[WS(rs, 45)]), VSUB(T4T, T4Q), ms, &(x[WS(rs, 1)])); T56 = VADD(T4Y, T55); T5d = VBYI(VADD(T59, T5c)); ST(&(x[WS(rs, 59)]), VSUB(T56, T5d), ms, &(x[WS(rs, 1)])); ST(&(x[WS(rs, 5)]), VADD(T56, T5d), ms, &(x[WS(rs, 1)])); } { V T5e, T5f, T4U, T4V; T5e = VSUB(T4Y, T55); T5f = VBYI(VSUB(T5c, T59)); ST(&(x[WS(rs, 37)]), VSUB(T5e, T5f), ms, &(x[WS(rs, 1)])); ST(&(x[WS(rs, 27)]), VADD(T5e, T5f), ms, &(x[WS(rs, 1)])); T4U = VBYI(VADD(T4P, T4O)); T4V = VADD(T4R, T4S); ST(&(x[WS(rs, 13)]), VADD(T4U, T4V), ms, &(x[WS(rs, 1)])); ST(&(x[WS(rs, 51)]), VSUB(T4V, T4U), ms, &(x[WS(rs, 1)])); } } { V T1u, T43, T3D, T3V, T3t, T45, T3B, T3K, T3d, T3E, T3w, T3A, T3R, T46, T3Y; V T42; { V TE, T3T, T1t, T3U, T13, T1s; TE = VSUB(Tg, TD); T3T = VADD(T3o, T3h); T13 = VFMA(LDK(KP195090322), TV, VMUL(LDK(KP980785280), T12)); T1s = VFNMS(LDK(KP195090322), T1r, VMUL(LDK(KP980785280), T1k)); T1t = VSUB(T13, T1s); T3U = VADD(T1s, T13); T1u = VADD(TE, T1t); T43 = VSUB(T3U, T3T); T3D = VSUB(TE, T1t); T3V = VADD(T3T, T3U); } { V T3p, T3I, T3s, T3J, T3q, T3r; T3p = VSUB(T3h, T3o); T3I = VADD(Tg, TD); T3q = VFNMS(LDK(KP195090322), T12, VMUL(LDK(KP980785280), TV)); T3r = VFMA(LDK(KP980785280), T1r, VMUL(LDK(KP195090322), T1k)); T3s = VSUB(T3q, T3r); T3J = VADD(T3r, T3q); T3t = VADD(T3p, T3s); T45 = VSUB(T3I, T3J); T3B = VSUB(T3s, T3p); T3K = VADD(T3I, T3J); } { V T2l, T3u, T3c, T3v; { V T29, T2k, T30, T3b; T29 = VSUB(T1R, T28); T2k = VSUB(T2g, T2j); T2l = VFMA(LDK(KP634393284), T29, VMUL(LDK(KP773010453), T2k)); T3u = VFNMS(LDK(KP634393284), T2k, VMUL(LDK(KP773010453), T29)); T30 = VSUB(T2C, T2Z); T3b = VSUB(T33, T3a); T3c = VFNMS(LDK(KP634393284), T3b, VMUL(LDK(KP773010453), T30)); T3v = VFMA(LDK(KP773010453), T3b, VMUL(LDK(KP634393284), T30)); } T3d = VADD(T2l, T3c); T3E = VSUB(T3v, T3u); T3w = VADD(T3u, T3v); T3A = VSUB(T3c, T2l); } { V T3N, T3W, T3Q, T3X; { V T3L, T3M, T3O, T3P; T3L = VADD(T28, T1R); T3M = VADD(T2g, T2j); T3N = VFMA(LDK(KP098017140), T3L, VMUL(LDK(KP995184726), T3M)); T3W = VFNMS(LDK(KP098017140), T3M, VMUL(LDK(KP995184726), T3L)); T3O = VADD(T2C, T2Z); T3P = VADD(T3a, T33); T3Q = VFNMS(LDK(KP098017140), T3P, VMUL(LDK(KP995184726), T3O)); T3X = VFMA(LDK(KP995184726), T3P, VMUL(LDK(KP098017140), T3O)); } T3R = VADD(T3N, T3Q); T46 = VSUB(T3X, T3W); T3Y = VADD(T3W, T3X); T42 = VSUB(T3Q, T3N); } { V T3e, T3x, T44, T47; T3e = VADD(T1u, T3d); T3x = VBYI(VADD(T3t, T3w)); ST(&(x[WS(rs, 57)]), VSUB(T3e, T3x), ms, &(x[WS(rs, 1)])); ST(&(x[WS(rs, 7)]), VADD(T3e, T3x), ms, &(x[WS(rs, 1)])); T44 = VBYI(VSUB(T42, T43)); T47 = VSUB(T45, T46); ST(&(x[WS(rs, 17)]), VADD(T44, T47), ms, &(x[WS(rs, 1)])); ST(&(x[WS(rs, 47)]), VSUB(T47, T44), ms, &(x[WS(rs, 1)])); } { V T48, T49, T3y, T3z; T48 = VBYI(VADD(T43, T42)); T49 = VADD(T45, T46); ST(&(x[WS(rs, 15)]), VADD(T48, T49), ms, &(x[WS(rs, 1)])); ST(&(x[WS(rs, 49)]), VSUB(T49, T48), ms, &(x[WS(rs, 1)])); T3y = VSUB(T1u, T3d); T3z = VBYI(VSUB(T3w, T3t)); ST(&(x[WS(rs, 39)]), VSUB(T3y, T3z), ms, &(x[WS(rs, 1)])); ST(&(x[WS(rs, 25)]), VADD(T3y, T3z), ms, &(x[WS(rs, 1)])); } { V T3C, T3F, T3S, T3Z; T3C = VBYI(VSUB(T3A, T3B)); T3F = VSUB(T3D, T3E); ST(&(x[WS(rs, 23)]), VADD(T3C, T3F), ms, &(x[WS(rs, 1)])); ST(&(x[WS(rs, 41)]), VSUB(T3F, T3C), ms, &(x[WS(rs, 1)])); T3S = VADD(T3K, T3R); T3Z = VBYI(VADD(T3V, T3Y)); ST(&(x[WS(rs, 63)]), VSUB(T3S, T3Z), ms, &(x[WS(rs, 1)])); ST(&(x[WS(rs, 1)]), VADD(T3S, T3Z), ms, &(x[WS(rs, 1)])); } { V T40, T41, T3G, T3H; T40 = VSUB(T3K, T3R); T41 = VBYI(VSUB(T3Y, T3V)); ST(&(x[WS(rs, 33)]), VSUB(T40, T41), ms, &(x[WS(rs, 1)])); ST(&(x[WS(rs, 31)]), VADD(T40, T41), ms, &(x[WS(rs, 1)])); T3G = VBYI(VADD(T3B, T3A)); T3H = VADD(T3D, T3E); ST(&(x[WS(rs, 9)]), VADD(T3G, T3H), ms, &(x[WS(rs, 1)])); ST(&(x[WS(rs, 55)]), VSUB(T3H, T3G), ms, &(x[WS(rs, 1)])); } } } } VLEAVE(); } static const tw_instr twinstr[] = { VTW(0, 1), VTW(0, 2), VTW(0, 3), VTW(0, 4), VTW(0, 5), VTW(0, 6), VTW(0, 7), VTW(0, 8), VTW(0, 9), VTW(0, 10), VTW(0, 11), VTW(0, 12), VTW(0, 13), VTW(0, 14), VTW(0, 15), VTW(0, 16), VTW(0, 17), VTW(0, 18), VTW(0, 19), VTW(0, 20), VTW(0, 21), VTW(0, 22), VTW(0, 23), VTW(0, 24), VTW(0, 25), VTW(0, 26), VTW(0, 27), VTW(0, 28), VTW(0, 29), VTW(0, 30), VTW(0, 31), VTW(0, 32), VTW(0, 33), VTW(0, 34), VTW(0, 35), VTW(0, 36), VTW(0, 37), VTW(0, 38), VTW(0, 39), VTW(0, 40), VTW(0, 41), VTW(0, 42), VTW(0, 43), VTW(0, 44), VTW(0, 45), VTW(0, 46), VTW(0, 47), VTW(0, 48), VTW(0, 49), VTW(0, 50), VTW(0, 51), VTW(0, 52), VTW(0, 53), VTW(0, 54), VTW(0, 55), VTW(0, 56), VTW(0, 57), VTW(0, 58), VTW(0, 59), VTW(0, 60), VTW(0, 61), VTW(0, 62), VTW(0, 63), {TW_NEXT, VL, 0} }; static const ct_desc desc = { 64, XSIMD_STRING("t1fv_64"), twinstr, &GENUS, {467, 198, 52, 0}, 0, 0, 0 }; void XSIMD(codelet_t1fv_64) (planner *p) { X(kdft_dit_register) (p, t1fv_64, &desc); } #endif /* HAVE_FMA */ fftw-3.3.3/dft/simd/common/t1fv_20.c0000644000175000017500000004200212054410510013723 00000000000000/* * Copyright (c) 2003, 2007-11 Matteo Frigo * Copyright (c) 2003, 2007-11 Massachusetts Institute of Technology * * This program is free software; you can redistribute it and/or modify * it under the terms of the GNU General Public License as published by * the Free Software Foundation; either version 2 of the License, or * (at your option) any later version. * * This program is distributed in the hope that it will be useful, * but WITHOUT ANY WARRANTY; without even the implied warranty of * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the * GNU General Public License for more details. * * You should have received a copy of the GNU General Public License * along with this program; if not, write to the Free Software * Foundation, Inc., 51 Franklin Street, Fifth Floor, Boston, MA 02110-1301 USA * */ /* This file was automatically generated --- DO NOT EDIT */ /* Generated on Sun Nov 25 07:38:32 EST 2012 */ #include "codelet-dft.h" #ifdef HAVE_FMA /* Generated by: ../../../genfft/gen_twiddle_c.native -fma -reorder-insns -schedule-for-pipeline -simd -compact -variables 4 -pipeline-latency 8 -n 20 -name t1fv_20 -include t1f.h */ /* * This function contains 123 FP additions, 88 FP multiplications, * (or, 77 additions, 42 multiplications, 46 fused multiply/add), * 68 stack variables, 4 constants, and 40 memory accesses */ #include "t1f.h" static void t1fv_20(R *ri, R *ii, const R *W, stride rs, INT mb, INT me, INT ms) { DVK(KP951056516, +0.951056516295153572116439333379382143405698634); DVK(KP559016994, +0.559016994374947424102293417182819058860154590); DVK(KP250000000, +0.250000000000000000000000000000000000000000000); DVK(KP618033988, +0.618033988749894848204586834365638117720309180); { INT m; R *x; x = ri; for (m = mb, W = W + (mb * ((TWVL / VL) * 38)); m < me; m = m + VL, x = x + (VL * ms), W = W + (TWVL * 38), MAKE_VOLATILE_STRIDE(20, rs)) { V T4, Tx, T1m, T1K, T1y, Tk, Tf, T16, T10, TT, T1O, T1w, T1L, T1p, T1M; V T1s, TZ, TI, T1x, Tp; { V T1, Tv, T2, Tt; T1 = LD(&(x[0]), ms, &(x[0])); Tv = LD(&(x[WS(rs, 15)]), ms, &(x[WS(rs, 1)])); T2 = LD(&(x[WS(rs, 10)]), ms, &(x[0])); Tt = LD(&(x[WS(rs, 5)]), ms, &(x[WS(rs, 1)])); { V T9, T1n, TN, T1v, TS, Te, T1q, T1u, TE, TG, Tm, T1o, TC, Tn, T1r; V TH, To; { V TP, TR, Ta, Tc; { V T5, T7, TJ, TL, T1k, T1l; T5 = LD(&(x[WS(rs, 4)]), ms, &(x[0])); T7 = LD(&(x[WS(rs, 14)]), ms, &(x[0])); TJ = LD(&(x[WS(rs, 13)]), ms, &(x[WS(rs, 1)])); TL = LD(&(x[WS(rs, 3)]), ms, &(x[WS(rs, 1)])); { V Tw, T3, Tu, T6, T8, TK, TM, TO, TQ; TO = LD(&(x[WS(rs, 17)]), ms, &(x[WS(rs, 1)])); Tw = BYTWJ(&(W[TWVL * 28]), Tv); T3 = BYTWJ(&(W[TWVL * 18]), T2); Tu = BYTWJ(&(W[TWVL * 8]), Tt); T6 = BYTWJ(&(W[TWVL * 6]), T5); T8 = BYTWJ(&(W[TWVL * 26]), T7); TK = BYTWJ(&(W[TWVL * 24]), TJ); TM = BYTWJ(&(W[TWVL * 4]), TL); TP = BYTWJ(&(W[TWVL * 32]), TO); TQ = LD(&(x[WS(rs, 7)]), ms, &(x[WS(rs, 1)])); T4 = VSUB(T1, T3); T1k = VADD(T1, T3); Tx = VSUB(Tu, Tw); T1l = VADD(Tu, Tw); T9 = VSUB(T6, T8); T1n = VADD(T6, T8); TN = VSUB(TK, TM); T1v = VADD(TK, TM); TR = BYTWJ(&(W[TWVL * 12]), TQ); } Ta = LD(&(x[WS(rs, 16)]), ms, &(x[0])); T1m = VSUB(T1k, T1l); T1K = VADD(T1k, T1l); Tc = LD(&(x[WS(rs, 6)]), ms, &(x[0])); } { V Tb, TA, Td, Th, Tj, Tz, Tg, Ti, Ty; Tg = LD(&(x[WS(rs, 8)]), ms, &(x[0])); Ti = LD(&(x[WS(rs, 18)]), ms, &(x[0])); Ty = LD(&(x[WS(rs, 9)]), ms, &(x[WS(rs, 1)])); TS = VSUB(TP, TR); T1y = VADD(TP, TR); Tb = BYTWJ(&(W[TWVL * 30]), Ta); TA = LD(&(x[WS(rs, 19)]), ms, &(x[WS(rs, 1)])); Td = BYTWJ(&(W[TWVL * 10]), Tc); Th = BYTWJ(&(W[TWVL * 14]), Tg); Tj = BYTWJ(&(W[TWVL * 34]), Ti); Tz = BYTWJ(&(W[TWVL * 16]), Ty); { V TD, TF, TB, Tl; TD = LD(&(x[WS(rs, 1)]), ms, &(x[WS(rs, 1)])); TF = LD(&(x[WS(rs, 11)]), ms, &(x[WS(rs, 1)])); Tl = LD(&(x[WS(rs, 12)]), ms, &(x[0])); TB = BYTWJ(&(W[TWVL * 36]), TA); Te = VSUB(Tb, Td); T1q = VADD(Tb, Td); Tk = VSUB(Th, Tj); T1u = VADD(Th, Tj); TE = BYTWJ(&(W[0]), TD); TG = BYTWJ(&(W[TWVL * 20]), TF); Tm = BYTWJ(&(W[TWVL * 22]), Tl); T1o = VADD(Tz, TB); TC = VSUB(Tz, TB); Tn = LD(&(x[WS(rs, 2)]), ms, &(x[0])); } } } Tf = VADD(T9, Te); T16 = VSUB(T9, Te); T10 = VSUB(TS, TN); TT = VADD(TN, TS); T1r = VADD(TE, TG); TH = VSUB(TE, TG); T1O = VADD(T1u, T1v); T1w = VSUB(T1u, T1v); To = BYTWJ(&(W[TWVL * 2]), Tn); T1L = VADD(T1n, T1o); T1p = VSUB(T1n, T1o); T1M = VADD(T1q, T1r); T1s = VSUB(T1q, T1r); TZ = VSUB(TH, TC); TI = VADD(TC, TH); T1x = VADD(Tm, To); Tp = VSUB(Tm, To); } } { V T1V, T1N, T14, T1d, T11, T1G, T1t, T1z, T1P, Tq, T17, T13, TV, TU; T1V = VSUB(T1L, T1M); T1N = VADD(T1L, T1M); T14 = VSUB(TT, TI); TU = VADD(TI, TT); T1d = VFNMS(LDK(KP618033988), TZ, T10); T11 = VFMA(LDK(KP618033988), T10, TZ); T1G = VSUB(T1p, T1s); T1t = VADD(T1p, T1s); T1z = VSUB(T1x, T1y); T1P = VADD(T1x, T1y); Tq = VADD(Tk, Tp); T17 = VSUB(Tk, Tp); T13 = VFNMS(LDK(KP250000000), TU, Tx); TV = VADD(Tx, TU); { V T1J, T1H, T1D, T1Z, T1X, T1T, T1h, T1j, T1b, T19, T1C, T1S, T1c, TY, T1F; V T1A; T1F = VSUB(T1w, T1z); T1A = VADD(T1w, T1z); { V T1W, T1Q, TX, Tr; T1W = VSUB(T1O, T1P); T1Q = VADD(T1O, T1P); TX = VSUB(Tf, Tq); Tr = VADD(Tf, Tq); { V T1g, T18, T1f, T15; T1g = VFNMS(LDK(KP618033988), T16, T17); T18 = VFMA(LDK(KP618033988), T17, T16); T1f = VFMA(LDK(KP559016994), T14, T13); T15 = VFNMS(LDK(KP559016994), T14, T13); T1J = VMUL(LDK(KP951056516), VFMA(LDK(KP618033988), T1F, T1G)); T1H = VMUL(LDK(KP951056516), VFNMS(LDK(KP618033988), T1G, T1F)); { V T1B, T1R, TW, Ts; T1B = VADD(T1t, T1A); T1D = VSUB(T1t, T1A); T1Z = VMUL(LDK(KP951056516), VFNMS(LDK(KP618033988), T1V, T1W)); T1X = VMUL(LDK(KP951056516), VFMA(LDK(KP618033988), T1W, T1V)); T1R = VADD(T1N, T1Q); T1T = VSUB(T1N, T1Q); TW = VFNMS(LDK(KP250000000), Tr, T4); Ts = VADD(T4, Tr); T1h = VFNMS(LDK(KP951056516), T1g, T1f); T1j = VFMA(LDK(KP951056516), T1g, T1f); T1b = VFNMS(LDK(KP951056516), T18, T15); T19 = VFMA(LDK(KP951056516), T18, T15); ST(&(x[WS(rs, 10)]), VADD(T1m, T1B), ms, &(x[0])); T1C = VFNMS(LDK(KP250000000), T1B, T1m); ST(&(x[0]), VADD(T1K, T1R), ms, &(x[0])); T1S = VFNMS(LDK(KP250000000), T1R, T1K); T1c = VFNMS(LDK(KP559016994), TX, TW); TY = VFMA(LDK(KP559016994), TX, TW); ST(&(x[WS(rs, 15)]), VFMAI(TV, Ts), ms, &(x[WS(rs, 1)])); ST(&(x[WS(rs, 5)]), VFNMSI(TV, Ts), ms, &(x[WS(rs, 1)])); } } } { V T1E, T1I, T1U, T1Y; T1E = VFNMS(LDK(KP559016994), T1D, T1C); T1I = VFMA(LDK(KP559016994), T1D, T1C); T1U = VFMA(LDK(KP559016994), T1T, T1S); T1Y = VFNMS(LDK(KP559016994), T1T, T1S); { V T1e, T1i, T1a, T12; T1e = VFNMS(LDK(KP951056516), T1d, T1c); T1i = VFMA(LDK(KP951056516), T1d, T1c); T1a = VFNMS(LDK(KP951056516), T11, TY); T12 = VFMA(LDK(KP951056516), T11, TY); ST(&(x[WS(rs, 18)]), VFNMSI(T1H, T1E), ms, &(x[0])); ST(&(x[WS(rs, 2)]), VFMAI(T1H, T1E), ms, &(x[0])); ST(&(x[WS(rs, 14)]), VFMAI(T1J, T1I), ms, &(x[0])); ST(&(x[WS(rs, 6)]), VFNMSI(T1J, T1I), ms, &(x[0])); ST(&(x[WS(rs, 16)]), VFNMSI(T1X, T1U), ms, &(x[0])); ST(&(x[WS(rs, 4)]), VFMAI(T1X, T1U), ms, &(x[0])); ST(&(x[WS(rs, 12)]), VFMAI(T1Z, T1Y), ms, &(x[0])); ST(&(x[WS(rs, 8)]), VFNMSI(T1Z, T1Y), ms, &(x[0])); ST(&(x[WS(rs, 3)]), VFMAI(T1h, T1e), ms, &(x[WS(rs, 1)])); ST(&(x[WS(rs, 17)]), VFNMSI(T1h, T1e), ms, &(x[WS(rs, 1)])); ST(&(x[WS(rs, 7)]), VFMAI(T1j, T1i), ms, &(x[WS(rs, 1)])); ST(&(x[WS(rs, 13)]), VFNMSI(T1j, T1i), ms, &(x[WS(rs, 1)])); ST(&(x[WS(rs, 11)]), VFMAI(T1b, T1a), ms, &(x[WS(rs, 1)])); ST(&(x[WS(rs, 9)]), VFNMSI(T1b, T1a), ms, &(x[WS(rs, 1)])); ST(&(x[WS(rs, 19)]), VFMAI(T19, T12), ms, &(x[WS(rs, 1)])); ST(&(x[WS(rs, 1)]), VFNMSI(T19, T12), ms, &(x[WS(rs, 1)])); } } } } } } VLEAVE(); } static const tw_instr twinstr[] = { VTW(0, 1), VTW(0, 2), VTW(0, 3), VTW(0, 4), VTW(0, 5), VTW(0, 6), VTW(0, 7), VTW(0, 8), VTW(0, 9), VTW(0, 10), VTW(0, 11), VTW(0, 12), VTW(0, 13), VTW(0, 14), VTW(0, 15), VTW(0, 16), VTW(0, 17), VTW(0, 18), VTW(0, 19), {TW_NEXT, VL, 0} }; static const ct_desc desc = { 20, XSIMD_STRING("t1fv_20"), twinstr, &GENUS, {77, 42, 46, 0}, 0, 0, 0 }; void XSIMD(codelet_t1fv_20) (planner *p) { X(kdft_dit_register) (p, t1fv_20, &desc); } #else /* HAVE_FMA */ /* Generated by: ../../../genfft/gen_twiddle_c.native -simd -compact -variables 4 -pipeline-latency 8 -n 20 -name t1fv_20 -include t1f.h */ /* * This function contains 123 FP additions, 62 FP multiplications, * (or, 111 additions, 50 multiplications, 12 fused multiply/add), * 54 stack variables, 4 constants, and 40 memory accesses */ #include "t1f.h" static void t1fv_20(R *ri, R *ii, const R *W, stride rs, INT mb, INT me, INT ms) { DVK(KP587785252, +0.587785252292473129168705954639072768597652438); DVK(KP951056516, +0.951056516295153572116439333379382143405698634); DVK(KP250000000, +0.250000000000000000000000000000000000000000000); DVK(KP559016994, +0.559016994374947424102293417182819058860154590); { INT m; R *x; x = ri; for (m = mb, W = W + (mb * ((TWVL / VL) * 38)); m < me; m = m + VL, x = x + (VL * ms), W = W + (TWVL * 38), MAKE_VOLATILE_STRIDE(20, rs)) { V T4, Tx, T1B, T1U, TZ, T16, T17, T10, Tf, Tq, Tr, T1N, T1O, T1S, T1t; V T1w, T1C, TI, TT, TU, T1K, T1L, T1R, T1m, T1p, T1D, Ts, TV; { V T1, Tw, T3, Tu, Tv, T2, Tt, T1z, T1A; T1 = LD(&(x[0]), ms, &(x[0])); Tv = LD(&(x[WS(rs, 15)]), ms, &(x[WS(rs, 1)])); Tw = BYTWJ(&(W[TWVL * 28]), Tv); T2 = LD(&(x[WS(rs, 10)]), ms, &(x[0])); T3 = BYTWJ(&(W[TWVL * 18]), T2); Tt = LD(&(x[WS(rs, 5)]), ms, &(x[WS(rs, 1)])); Tu = BYTWJ(&(W[TWVL * 8]), Tt); T4 = VSUB(T1, T3); Tx = VSUB(Tu, Tw); T1z = VADD(T1, T3); T1A = VADD(Tu, Tw); T1B = VSUB(T1z, T1A); T1U = VADD(T1z, T1A); } { V T9, T1r, TN, T1l, TS, T1o, Te, T1u, Tk, T1k, TC, T1s, TH, T1v, Tp; V T1n; { V T6, T8, T5, T7; T5 = LD(&(x[WS(rs, 4)]), ms, &(x[0])); T6 = BYTWJ(&(W[TWVL * 6]), T5); T7 = LD(&(x[WS(rs, 14)]), ms, &(x[0])); T8 = BYTWJ(&(W[TWVL * 26]), T7); T9 = VSUB(T6, T8); T1r = VADD(T6, T8); } { V TK, TM, TJ, TL; TJ = LD(&(x[WS(rs, 13)]), ms, &(x[WS(rs, 1)])); TK = BYTWJ(&(W[TWVL * 24]), TJ); TL = LD(&(x[WS(rs, 3)]), ms, &(x[WS(rs, 1)])); TM = BYTWJ(&(W[TWVL * 4]), TL); TN = VSUB(TK, TM); T1l = VADD(TK, TM); } { V TP, TR, TO, TQ; TO = LD(&(x[WS(rs, 17)]), ms, &(x[WS(rs, 1)])); TP = BYTWJ(&(W[TWVL * 32]), TO); TQ = LD(&(x[WS(rs, 7)]), ms, &(x[WS(rs, 1)])); TR = BYTWJ(&(W[TWVL * 12]), TQ); TS = VSUB(TP, TR); T1o = VADD(TP, TR); } { V Tb, Td, Ta, Tc; Ta = LD(&(x[WS(rs, 16)]), ms, &(x[0])); Tb = BYTWJ(&(W[TWVL * 30]), Ta); Tc = LD(&(x[WS(rs, 6)]), ms, &(x[0])); Td = BYTWJ(&(W[TWVL * 10]), Tc); Te = VSUB(Tb, Td); T1u = VADD(Tb, Td); } { V Th, Tj, Tg, Ti; Tg = LD(&(x[WS(rs, 8)]), ms, &(x[0])); Th = BYTWJ(&(W[TWVL * 14]), Tg); Ti = LD(&(x[WS(rs, 18)]), ms, &(x[0])); Tj = BYTWJ(&(W[TWVL * 34]), Ti); Tk = VSUB(Th, Tj); T1k = VADD(Th, Tj); } { V Tz, TB, Ty, TA; Ty = LD(&(x[WS(rs, 9)]), ms, &(x[WS(rs, 1)])); Tz = BYTWJ(&(W[TWVL * 16]), Ty); TA = LD(&(x[WS(rs, 19)]), ms, &(x[WS(rs, 1)])); TB = BYTWJ(&(W[TWVL * 36]), TA); TC = VSUB(Tz, TB); T1s = VADD(Tz, TB); } { V TE, TG, TD, TF; TD = LD(&(x[WS(rs, 1)]), ms, &(x[WS(rs, 1)])); TE = BYTWJ(&(W[0]), TD); TF = LD(&(x[WS(rs, 11)]), ms, &(x[WS(rs, 1)])); TG = BYTWJ(&(W[TWVL * 20]), TF); TH = VSUB(TE, TG); T1v = VADD(TE, TG); } { V Tm, To, Tl, Tn; Tl = LD(&(x[WS(rs, 12)]), ms, &(x[0])); Tm = BYTWJ(&(W[TWVL * 22]), Tl); Tn = LD(&(x[WS(rs, 2)]), ms, &(x[0])); To = BYTWJ(&(W[TWVL * 2]), Tn); Tp = VSUB(Tm, To); T1n = VADD(Tm, To); } TZ = VSUB(TH, TC); T16 = VSUB(T9, Te); T17 = VSUB(Tk, Tp); T10 = VSUB(TS, TN); Tf = VADD(T9, Te); Tq = VADD(Tk, Tp); Tr = VADD(Tf, Tq); T1N = VADD(T1k, T1l); T1O = VADD(T1n, T1o); T1S = VADD(T1N, T1O); T1t = VSUB(T1r, T1s); T1w = VSUB(T1u, T1v); T1C = VADD(T1t, T1w); TI = VADD(TC, TH); TT = VADD(TN, TS); TU = VADD(TI, TT); T1K = VADD(T1r, T1s); T1L = VADD(T1u, T1v); T1R = VADD(T1K, T1L); T1m = VSUB(T1k, T1l); T1p = VSUB(T1n, T1o); T1D = VADD(T1m, T1p); } Ts = VADD(T4, Tr); TV = VBYI(VADD(Tx, TU)); ST(&(x[WS(rs, 5)]), VSUB(Ts, TV), ms, &(x[WS(rs, 1)])); ST(&(x[WS(rs, 15)]), VADD(Ts, TV), ms, &(x[WS(rs, 1)])); { V T1T, T1V, T1W, T1Q, T1Z, T1M, T1P, T1Y, T1X; T1T = VMUL(LDK(KP559016994), VSUB(T1R, T1S)); T1V = VADD(T1R, T1S); T1W = VFNMS(LDK(KP250000000), T1V, T1U); T1M = VSUB(T1K, T1L); T1P = VSUB(T1N, T1O); T1Q = VBYI(VFMA(LDK(KP951056516), T1M, VMUL(LDK(KP587785252), T1P))); T1Z = VBYI(VFNMS(LDK(KP587785252), T1M, VMUL(LDK(KP951056516), T1P))); ST(&(x[0]), VADD(T1U, T1V), ms, &(x[0])); T1Y = VSUB(T1W, T1T); ST(&(x[WS(rs, 8)]), VSUB(T1Y, T1Z), ms, &(x[0])); ST(&(x[WS(rs, 12)]), VADD(T1Z, T1Y), ms, &(x[0])); T1X = VADD(T1T, T1W); ST(&(x[WS(rs, 4)]), VADD(T1Q, T1X), ms, &(x[0])); ST(&(x[WS(rs, 16)]), VSUB(T1X, T1Q), ms, &(x[0])); } { V T1G, T1E, T1F, T1y, T1J, T1q, T1x, T1I, T1H; T1G = VMUL(LDK(KP559016994), VSUB(T1C, T1D)); T1E = VADD(T1C, T1D); T1F = VFNMS(LDK(KP250000000), T1E, T1B); T1q = VSUB(T1m, T1p); T1x = VSUB(T1t, T1w); T1y = VBYI(VFNMS(LDK(KP587785252), T1x, VMUL(LDK(KP951056516), T1q))); T1J = VBYI(VFMA(LDK(KP951056516), T1x, VMUL(LDK(KP587785252), T1q))); ST(&(x[WS(rs, 10)]), VADD(T1B, T1E), ms, &(x[0])); T1I = VADD(T1G, T1F); ST(&(x[WS(rs, 6)]), VSUB(T1I, T1J), ms, &(x[0])); ST(&(x[WS(rs, 14)]), VADD(T1J, T1I), ms, &(x[0])); T1H = VSUB(T1F, T1G); ST(&(x[WS(rs, 2)]), VADD(T1y, T1H), ms, &(x[0])); ST(&(x[WS(rs, 18)]), VSUB(T1H, T1y), ms, &(x[0])); } { V T11, T18, T1g, T1d, T15, T1f, TY, T1c; T11 = VFMA(LDK(KP951056516), TZ, VMUL(LDK(KP587785252), T10)); T18 = VFMA(LDK(KP951056516), T16, VMUL(LDK(KP587785252), T17)); T1g = VFNMS(LDK(KP587785252), T16, VMUL(LDK(KP951056516), T17)); T1d = VFNMS(LDK(KP587785252), TZ, VMUL(LDK(KP951056516), T10)); { V T13, T14, TW, TX; T13 = VFMS(LDK(KP250000000), TU, Tx); T14 = VMUL(LDK(KP559016994), VSUB(TT, TI)); T15 = VADD(T13, T14); T1f = VSUB(T14, T13); TW = VMUL(LDK(KP559016994), VSUB(Tf, Tq)); TX = VFNMS(LDK(KP250000000), Tr, T4); TY = VADD(TW, TX); T1c = VSUB(TX, TW); } { V T12, T19, T1i, T1j; T12 = VADD(TY, T11); T19 = VBYI(VSUB(T15, T18)); ST(&(x[WS(rs, 19)]), VSUB(T12, T19), ms, &(x[WS(rs, 1)])); ST(&(x[WS(rs, 1)]), VADD(T12, T19), ms, &(x[WS(rs, 1)])); T1i = VADD(T1c, T1d); T1j = VBYI(VADD(T1g, T1f)); ST(&(x[WS(rs, 13)]), VSUB(T1i, T1j), ms, &(x[WS(rs, 1)])); ST(&(x[WS(rs, 7)]), VADD(T1i, T1j), ms, &(x[WS(rs, 1)])); } { V T1a, T1b, T1e, T1h; T1a = VSUB(TY, T11); T1b = VBYI(VADD(T18, T15)); ST(&(x[WS(rs, 11)]), VSUB(T1a, T1b), ms, &(x[WS(rs, 1)])); ST(&(x[WS(rs, 9)]), VADD(T1a, T1b), ms, &(x[WS(rs, 1)])); T1e = VSUB(T1c, T1d); T1h = VBYI(VSUB(T1f, T1g)); ST(&(x[WS(rs, 17)]), VSUB(T1e, T1h), ms, &(x[WS(rs, 1)])); ST(&(x[WS(rs, 3)]), VADD(T1e, T1h), ms, &(x[WS(rs, 1)])); } } } } VLEAVE(); } static const tw_instr twinstr[] = { VTW(0, 1), VTW(0, 2), VTW(0, 3), VTW(0, 4), VTW(0, 5), VTW(0, 6), VTW(0, 7), VTW(0, 8), VTW(0, 9), VTW(0, 10), VTW(0, 11), VTW(0, 12), VTW(0, 13), VTW(0, 14), VTW(0, 15), VTW(0, 16), VTW(0, 17), VTW(0, 18), VTW(0, 19), {TW_NEXT, VL, 0} }; static const ct_desc desc = { 20, XSIMD_STRING("t1fv_20"), twinstr, &GENUS, {111, 50, 12, 0}, 0, 0, 0 }; void XSIMD(codelet_t1fv_20) (planner *p) { X(kdft_dit_register) (p, t1fv_20, &desc); } #endif /* HAVE_FMA */ fftw-3.3.3/dft/simd/common/t1fv_25.c0000644000175000017500000011471012054410520013737 00000000000000/* * Copyright (c) 2003, 2007-11 Matteo Frigo * Copyright (c) 2003, 2007-11 Massachusetts Institute of Technology * * This program is free software; you can redistribute it and/or modify * it under the terms of the GNU General Public License as published by * the Free Software Foundation; either version 2 of the License, or * (at your option) any later version. * * This program is distributed in the hope that it will be useful, * but WITHOUT ANY WARRANTY; without even the implied warranty of * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the * GNU General Public License for more details. * * You should have received a copy of the GNU General Public License * along with this program; if not, write to the Free Software * Foundation, Inc., 51 Franklin Street, Fifth Floor, Boston, MA 02110-1301 USA * */ /* This file was automatically generated --- DO NOT EDIT */ /* Generated on Sun Nov 25 07:38:33 EST 2012 */ #include "codelet-dft.h" #ifdef HAVE_FMA /* Generated by: ../../../genfft/gen_twiddle_c.native -fma -reorder-insns -schedule-for-pipeline -simd -compact -variables 4 -pipeline-latency 8 -n 25 -name t1fv_25 -include t1f.h */ /* * This function contains 248 FP additions, 241 FP multiplications, * (or, 67 additions, 60 multiplications, 181 fused multiply/add), * 208 stack variables, 67 constants, and 50 memory accesses */ #include "t1f.h" static void t1fv_25(R *ri, R *ii, const R *W, stride rs, INT mb, INT me, INT ms) { DVK(KP792626838, +0.792626838241819413632131824093538848057784557); DVK(KP876091699, +0.876091699473550838204498029706869638173524346); DVK(KP617882369, +0.617882369114440893914546919006756321695042882); DVK(KP803003575, +0.803003575438660414833440593570376004635464850); DVK(KP242145790, +0.242145790282157779872542093866183953459003101); DVK(KP968583161, +0.968583161128631119490168375464735813836012403); DVK(KP999544308, +0.999544308746292983948881682379742149196758193); DVK(KP916574801, +0.916574801383451584742370439148878693530976769); DVK(KP904730450, +0.904730450839922351881287709692877908104763647); DVK(KP809385824, +0.809385824416008241660603814668679683846476688); DVK(KP447417479, +0.447417479732227551498980015410057305749330693); DVK(KP894834959, +0.894834959464455102997960030820114611498661386); DVK(KP867381224, +0.867381224396525206773171885031575671309956167); DVK(KP683113946, +0.683113946453479238701949862233725244439656928); DVK(KP559154169, +0.559154169276087864842202529084232643714075927); DVK(KP958953096, +0.958953096729998668045963838399037225970891871); DVK(KP831864738, +0.831864738706457140726048799369896829771167132); DVK(KP829049696, +0.829049696159252993975487806364305442437946767); DVK(KP860541664, +0.860541664367944677098261680920518816412804187); DVK(KP897376177, +0.897376177523557693138608077137219684419427330); DVK(KP876306680, +0.876306680043863587308115903922062583399064238); DVK(KP681693190, +0.681693190061530575150324149145440022633095390); DVK(KP560319534, +0.560319534973832390111614715371676131169633784); DVK(KP855719849, +0.855719849902058969314654733608091555096772472); DVK(KP237294955, +0.237294955877110315393888866460840817927895961); DVK(KP949179823, +0.949179823508441261575555465843363271711583843); DVK(KP904508497, +0.904508497187473712051146708591409529430077295); DVK(KP997675361, +0.997675361079556513670859573984492383596555031); DVK(KP763932022, +0.763932022500210303590826331268723764559381640); DVK(KP690983005, +0.690983005625052575897706582817180941139845410); DVK(KP992114701, +0.992114701314477831049793042785778521453036709); DVK(KP952936919, +0.952936919628306576880750665357914584765951388); DVK(KP998026728, +0.998026728428271561952336806863450553336905220); DVK(KP262346850, +0.262346850930607871785420028382979691334784273); DVK(KP570584518, +0.570584518783621657366766175430996792655723863); DVK(KP669429328, +0.669429328479476605641803240971985825917022098); DVK(KP923225144, +0.923225144846402650453449441572664695995209956); DVK(KP945422727, +0.945422727388575946270360266328811958657216298); DVK(KP522616830, +0.522616830205754336872861364785224694908468440); DVK(KP956723877, +0.956723877038460305821989399535483155872969262); DVK(KP906616052, +0.906616052148196230441134447086066874408359177); DVK(KP772036680, +0.772036680810363904029489473607579825330539880); DVK(KP845997307, +0.845997307939530944175097360758058292389769300); DVK(KP951056516, +0.951056516295153572116439333379382143405698634); DVK(KP921078979, +0.921078979742360627699756128143719920817673854); DVK(KP912575812, +0.912575812670962425556968549836277086778922727); DVK(KP982009705, +0.982009705009746369461829878184175962711969869); DVK(KP734762448, +0.734762448793050413546343770063151342619912334); DVK(KP494780565, +0.494780565770515410344588413655324772219443730); DVK(KP447533225, +0.447533225982656890041886979663652563063114397); DVK(KP269969613, +0.269969613759572083574752974412347470060951301); DVK(KP244189809, +0.244189809627953270309879511234821255780225091); DVK(KP667278218, +0.667278218140296670899089292254759909713898805); DVK(KP603558818, +0.603558818296015001454675132653458027918768137); DVK(KP522847744, +0.522847744331509716623755382187077770911012542); DVK(KP578046249, +0.578046249379945007321754579646815604023525655); DVK(KP987388751, +0.987388751065621252324603216482382109400433949); DVK(KP893101515, +0.893101515366181661711202267938416198338079437); DVK(KP120146378, +0.120146378570687701782758537356596213647956445); DVK(KP132830569, +0.132830569247582714407653942074819768844536507); DVK(KP869845200, +0.869845200362138853122720822420327157933056305); DVK(KP786782374, +0.786782374965295178365099601674911834788448471); DVK(KP066152395, +0.066152395967733048213034281011006031460903353); DVK(KP059835404, +0.059835404262124915169548397419498386427871950); DVK(KP559016994, +0.559016994374947424102293417182819058860154590); DVK(KP250000000, +0.250000000000000000000000000000000000000000000); DVK(KP618033988, +0.618033988749894848204586834365638117720309180); { INT m; R *x; x = ri; for (m = mb, W = W + (mb * ((TWVL / VL) * 48)); m < me; m = m + VL, x = x + (VL * ms), W = W + (TWVL * 48), MAKE_VOLATILE_STRIDE(25, rs)) { V T25, T1B, T2y, T1K, T2s, T23, T1S, T26, T20, T1X; { V T1O, T2X, Te, T3L, Td, T3Q, T3j, T3b, T2R, T2M, T2f, T27, T1y, T1H, T3M; V TW, TR, TK, T2B, T3n, T3e, T2U, T2F, T2i, T2a, Tz, T1C, T3N, TQ, T11; V T1b, T1c, T16; { V T1, T1g, T1i, T1p, T1k, T1m, Tb, T1N, T6, T1M; { V T7, T9, T2, T4, T1f, T1h, T1o; T1 = LD(&(x[0]), ms, &(x[0])); T7 = LD(&(x[WS(rs, 10)]), ms, &(x[0])); T9 = LD(&(x[WS(rs, 15)]), ms, &(x[WS(rs, 1)])); T2 = LD(&(x[WS(rs, 5)]), ms, &(x[WS(rs, 1)])); T4 = LD(&(x[WS(rs, 20)]), ms, &(x[0])); T1f = LD(&(x[WS(rs, 3)]), ms, &(x[WS(rs, 1)])); T1h = LD(&(x[WS(rs, 8)]), ms, &(x[0])); T1o = LD(&(x[WS(rs, 18)]), ms, &(x[0])); { V T8, Ta, T3, T5, T1j; T1j = LD(&(x[WS(rs, 23)]), ms, &(x[WS(rs, 1)])); T8 = BYTWJ(&(W[TWVL * 18]), T7); Ta = BYTWJ(&(W[TWVL * 28]), T9); T3 = BYTWJ(&(W[TWVL * 8]), T2); T5 = BYTWJ(&(W[TWVL * 38]), T4); T1g = BYTWJ(&(W[TWVL * 4]), T1f); T1i = BYTWJ(&(W[TWVL * 14]), T1h); T1p = BYTWJ(&(W[TWVL * 34]), T1o); T1k = BYTWJ(&(W[TWVL * 44]), T1j); T1m = LD(&(x[WS(rs, 13)]), ms, &(x[WS(rs, 1)])); Tb = VADD(T8, Ta); T1N = VSUB(T8, Ta); T6 = VADD(T3, T5); T1M = VSUB(T3, T5); } } { V T1v, T1l, Th, Tj, T1w, T1q, Tq, Tk, Tn, Tg; Tg = LD(&(x[WS(rs, 1)]), ms, &(x[WS(rs, 1)])); { V Tc, Ti, T1n, Tp; Ti = LD(&(x[WS(rs, 6)]), ms, &(x[0])); T1v = VSUB(T1i, T1k); T1l = VADD(T1i, T1k); T1n = BYTWJ(&(W[TWVL * 24]), T1m); Tp = LD(&(x[WS(rs, 16)]), ms, &(x[0])); T1O = VFMA(LDK(KP618033988), T1N, T1M); T2X = VFNMS(LDK(KP618033988), T1M, T1N); Te = VSUB(T6, Tb); Tc = VADD(T6, Tb); Th = BYTWJ(&(W[0]), Tg); Tj = BYTWJ(&(W[TWVL * 10]), Ti); T1w = VSUB(T1n, T1p); T1q = VADD(T1n, T1p); Tq = BYTWJ(&(W[TWVL * 30]), Tp); Tk = LD(&(x[WS(rs, 21)]), ms, &(x[WS(rs, 1)])); T3L = VADD(T1, Tc); Td = VFNMS(LDK(KP250000000), Tc, T1); Tn = LD(&(x[WS(rs, 11)]), ms, &(x[WS(rs, 1)])); } { V T1x, T2K, TM, TB, Tw, Tm, Tx, Tr, TI, T2L, T1u, TD, TF, TL; TL = LD(&(x[WS(rs, 4)]), ms, &(x[0])); { V T1t, Tl, To, TH, T1s, T1r, TA, TC; TA = LD(&(x[WS(rs, 24)]), ms, &(x[0])); T1r = VADD(T1l, T1q); T1t = VSUB(T1q, T1l); T1x = VFMA(LDK(KP618033988), T1w, T1v); T2K = VFNMS(LDK(KP618033988), T1v, T1w); Tl = BYTWJ(&(W[TWVL * 40]), Tk); To = BYTWJ(&(W[TWVL * 20]), Tn); TM = BYTWJ(&(W[TWVL * 6]), TL); TB = BYTWJ(&(W[TWVL * 46]), TA); TH = LD(&(x[WS(rs, 14)]), ms, &(x[0])); T1s = VFNMS(LDK(KP250000000), T1r, T1g); T3Q = VADD(T1g, T1r); TC = LD(&(x[WS(rs, 9)]), ms, &(x[WS(rs, 1)])); Tw = VSUB(Tj, Tl); Tm = VADD(Tj, Tl); Tx = VSUB(Tq, To); Tr = VADD(To, Tq); TI = BYTWJ(&(W[TWVL * 26]), TH); T2L = VFMA(LDK(KP559016994), T1t, T1s); T1u = VFNMS(LDK(KP559016994), T1t, T1s); TD = BYTWJ(&(W[TWVL * 16]), TC); TF = LD(&(x[WS(rs, 19)]), ms, &(x[WS(rs, 1)])); } { V Tu, Ty, T2E, TE, TN, TG, Tt, TV, Ts; TV = LD(&(x[WS(rs, 2)]), ms, &(x[0])); Ts = VADD(Tm, Tr); Tu = VSUB(Tm, Tr); Ty = VFNMS(LDK(KP618033988), Tx, Tw); T2E = VFMA(LDK(KP618033988), Tw, Tx); T3j = VFNMS(LDK(KP059835404), T2K, T2L); T3b = VFMA(LDK(KP066152395), T2L, T2K); T2R = VFNMS(LDK(KP786782374), T2K, T2L); T2M = VFMA(LDK(KP869845200), T2L, T2K); T2f = VFMA(LDK(KP132830569), T1u, T1x); T27 = VFNMS(LDK(KP120146378), T1x, T1u); T1y = VFNMS(LDK(KP893101515), T1x, T1u); T1H = VFMA(LDK(KP987388751), T1u, T1x); TE = VSUB(TB, TD); TN = VADD(TD, TB); TG = BYTWJ(&(W[TWVL * 36]), TF); Tt = VFNMS(LDK(KP250000000), Ts, Th); T3M = VADD(Th, Ts); TW = BYTWJ(&(W[TWVL * 2]), TV); { V TJ, TO, Tv, T2D, TY, T15, T10, T13, TP; { V TX, T14, TZ, T12; TX = LD(&(x[WS(rs, 7)]), ms, &(x[WS(rs, 1)])); T14 = LD(&(x[WS(rs, 17)]), ms, &(x[WS(rs, 1)])); TZ = LD(&(x[WS(rs, 22)]), ms, &(x[0])); T12 = LD(&(x[WS(rs, 12)]), ms, &(x[0])); TJ = VSUB(TG, TI); TO = VADD(TI, TG); Tv = VFMA(LDK(KP559016994), Tu, Tt); T2D = VFNMS(LDK(KP559016994), Tu, Tt); TY = BYTWJ(&(W[TWVL * 12]), TX); T15 = BYTWJ(&(W[TWVL * 32]), T14); T10 = BYTWJ(&(W[TWVL * 42]), TZ); T13 = BYTWJ(&(W[TWVL * 22]), T12); } TP = VADD(TN, TO); TR = VSUB(TN, TO); TK = VFMA(LDK(KP618033988), TJ, TE); T2B = VFNMS(LDK(KP618033988), TE, TJ); T3n = VFMA(LDK(KP578046249), T2D, T2E); T3e = VFNMS(LDK(KP522847744), T2E, T2D); T2U = VFNMS(LDK(KP987388751), T2D, T2E); T2F = VFMA(LDK(KP893101515), T2E, T2D); T2i = VFNMS(LDK(KP603558818), Ty, Tv); T2a = VFMA(LDK(KP667278218), Tv, Ty); Tz = VFNMS(LDK(KP244189809), Ty, Tv); T1C = VFMA(LDK(KP269969613), Tv, Ty); T3N = VADD(TM, TP); TQ = VFMS(LDK(KP250000000), TP, TM); T11 = VADD(TY, T10); T1b = VSUB(TY, T10); T1c = VSUB(T15, T13); T16 = VADD(T13, T15); } } } } } { V T2z, Tf, T3W, T3O, T1d, T2H, T3m, T2j, T2b, TT, T1D, T2G, T35, T2V, T2Z; V T3A, T3g, T2I, T1a, T3R, T3X; T2z = VFNMS(LDK(KP559016994), Te, Td); Tf = VFMA(LDK(KP559016994), Te, Td); { V TS, T2A, T17, T19; TS = VFNMS(LDK(KP559016994), TR, TQ); T2A = VFMA(LDK(KP559016994), TR, TQ); T3W = VSUB(T3M, T3N); T3O = VADD(T3M, T3N); T1d = VFNMS(LDK(KP618033988), T1c, T1b); T2H = VFMA(LDK(KP618033988), T1b, T1c); T17 = VADD(T11, T16); T19 = VSUB(T16, T11); { V T3f, T2T, T2C, T18, T3P; T3m = VFMA(LDK(KP447533225), T2B, T2A); T3f = VFNMS(LDK(KP494780565), T2A, T2B); T2T = VFNMS(LDK(KP132830569), T2A, T2B); T2C = VFMA(LDK(KP120146378), T2B, T2A); T2j = VFNMS(LDK(KP786782374), TK, TS); T2b = VFMA(LDK(KP869845200), TS, TK); TT = VFNMS(LDK(KP667278218), TS, TK); T1D = VFMA(LDK(KP603558818), TK, TS); T18 = VFNMS(LDK(KP250000000), T17, TW); T3P = VADD(TW, T17); T2G = VFMA(LDK(KP734762448), T2F, T2C); T35 = VFNMS(LDK(KP734762448), T2F, T2C); T2V = VFNMS(LDK(KP734762448), T2U, T2T); T2Z = VFMA(LDK(KP734762448), T2U, T2T); T3A = VFMA(LDK(KP982009705), T3f, T3e); T3g = VFNMS(LDK(KP982009705), T3f, T3e); T2I = VFMA(LDK(KP559016994), T19, T18); T1a = VFNMS(LDK(KP559016994), T19, T18); T3R = VADD(T3P, T3Q); T3X = VSUB(T3P, T3Q); } } { V T2n, T2t, T1V, T22, T2l, T2d, T1Q, T1I, T2w, T1A, T1F, T2q; { V T2k, T1G, T28, T2g, T3K, T3E, T3a, T34, T3x, T3H, T2c, TU, T1T, T1U, T1z; V T3o, T3t; T2n = VFNMS(LDK(KP912575812), T2j, T2i); T2k = VFMA(LDK(KP912575812), T2j, T2i); T3o = VFNMS(LDK(KP921078979), T3n, T3m); T3t = VFMA(LDK(KP921078979), T3n, T3m); { V T3c, T2Q, T2J, T3k, T1e; T3c = VFNMS(LDK(KP667278218), T2I, T2H); T2Q = VFNMS(LDK(KP059835404), T2H, T2I); T2J = VFMA(LDK(KP066152395), T2I, T2H); T3k = VFMA(LDK(KP603558818), T2H, T2I); T1G = VFMA(LDK(KP578046249), T1a, T1d); T1e = VFNMS(LDK(KP522847744), T1d, T1a); T28 = VFNMS(LDK(KP494780565), T1a, T1d); T2g = VFMA(LDK(KP447533225), T1d, T1a); { V T3U, T3S, T40, T3Y; T3U = VSUB(T3O, T3R); T3S = VADD(T3O, T3R); T40 = VMUL(LDK(KP951056516), VFNMS(LDK(KP618033988), T3W, T3X)); T3Y = VMUL(LDK(KP951056516), VFMA(LDK(KP618033988), T3X, T3W)); { V T3s, T3l, T2N, T36; T3s = VFNMS(LDK(KP845997307), T3k, T3j); T3l = VFMA(LDK(KP845997307), T3k, T3j); T2N = VFNMS(LDK(KP772036680), T2M, T2J); T36 = VFMA(LDK(KP772036680), T2M, T2J); { V T30, T2S, T3d, T3z, T3T; T30 = VFNMS(LDK(KP772036680), T2R, T2Q); T2S = VFMA(LDK(KP772036680), T2R, T2Q); T3d = VFNMS(LDK(KP845997307), T3c, T3b); T3z = VFMA(LDK(KP845997307), T3c, T3b); ST(&(x[0]), VADD(T3S, T3L), ms, &(x[0])); T3T = VFNMS(LDK(KP250000000), T3S, T3L); { V T3C, T3p, T2O, T37; T3C = VFMA(LDK(KP906616052), T3o, T3l); T3p = VFNMS(LDK(KP906616052), T3o, T3l); T2O = VFMA(LDK(KP956723877), T2N, T2G); T37 = VFMA(LDK(KP522616830), T2V, T36); { V T31, T2W, T3u, T3h; T31 = VFNMS(LDK(KP522616830), T2G, T30); T2W = VFMA(LDK(KP945422727), T2V, T2S); T3u = VFNMS(LDK(KP923225144), T3g, T3d); T3h = VFMA(LDK(KP923225144), T3g, T3d); { V T3I, T3B, T3V, T3Z; T3I = VFNMS(LDK(KP669429328), T3z, T3A); T3B = VFMA(LDK(KP570584518), T3A, T3z); T3V = VFMA(LDK(KP559016994), T3U, T3T); T3Z = VFNMS(LDK(KP559016994), T3U, T3T); { V T3y, T3q, T2P, T38; T3y = VFMA(LDK(KP262346850), T3p, T2X); T3q = VMUL(LDK(KP998026728), VFNMS(LDK(KP952936919), T2X, T3p)); T2P = VFMA(LDK(KP992114701), T2O, T2z); T38 = VFNMS(LDK(KP690983005), T37, T2S); { V T32, T2Y, T3v, T3F; T32 = VFMA(LDK(KP763932022), T31, T2N); T2Y = VMUL(LDK(KP998026728), VFMA(LDK(KP952936919), T2X, T2W)); T3v = VFNMS(LDK(KP997675361), T3u, T3t); T3F = VFNMS(LDK(KP904508497), T3u, T3s); { V T3i, T3r, T3J, T3D; T3i = VFMA(LDK(KP949179823), T3h, T2z); T3r = VFNMS(LDK(KP237294955), T3h, T2z); T3J = VFNMS(LDK(KP669429328), T3C, T3I); T3D = VFMA(LDK(KP618033988), T3C, T3B); ST(&(x[WS(rs, 20)]), VFMAI(T3Y, T3V), ms, &(x[0])); ST(&(x[WS(rs, 5)]), VFNMSI(T3Y, T3V), ms, &(x[WS(rs, 1)])); ST(&(x[WS(rs, 15)]), VFNMSI(T40, T3Z), ms, &(x[WS(rs, 1)])); ST(&(x[WS(rs, 10)]), VFMAI(T40, T3Z), ms, &(x[0])); { V T39, T33, T3w, T3G; T39 = VFMA(LDK(KP855719849), T38, T35); T33 = VFNMS(LDK(KP855719849), T32, T2Z); ST(&(x[WS(rs, 22)]), VFMAI(T2Y, T2P), ms, &(x[0])); ST(&(x[WS(rs, 3)]), VFNMSI(T2Y, T2P), ms, &(x[WS(rs, 1)])); T3w = VFMA(LDK(KP560319534), T3v, T3s); T3G = VFNMS(LDK(KP681693190), T3F, T3t); ST(&(x[WS(rs, 23)]), VFMAI(T3q, T3i), ms, &(x[WS(rs, 1)])); ST(&(x[WS(rs, 2)]), VFNMSI(T3q, T3i), ms, &(x[0])); T3K = VMUL(LDK(KP951056516), VFNMS(LDK(KP876306680), T3J, T3y)); T3E = VMUL(LDK(KP951056516), VFNMS(LDK(KP949179823), T3D, T3y)); T3a = VMUL(LDK(KP951056516), VFNMS(LDK(KP992114701), T39, T2X)); T34 = VFMA(LDK(KP897376177), T33, T2z); T3x = VFNMS(LDK(KP949179823), T3w, T3r); T3H = VFNMS(LDK(KP860541664), T3G, T3r); T2t = VFNMS(LDK(KP912575812), T2b, T2a); T2c = VFMA(LDK(KP912575812), T2b, T2a); TU = VFMA(LDK(KP829049696), TT, Tz); T1T = VFNMS(LDK(KP829049696), TT, Tz); T1U = VFNMS(LDK(KP831864738), T1y, T1e); T1z = VFMA(LDK(KP831864738), T1y, T1e); } } } } } } } } } } } { V T2o, T2h, T29, T2u, T2v, T2p; T2o = VFNMS(LDK(KP958953096), T2g, T2f); T2h = VFMA(LDK(KP958953096), T2g, T2f); ST(&(x[WS(rs, 17)]), VFMAI(T3a, T34), ms, &(x[WS(rs, 1)])); ST(&(x[WS(rs, 8)]), VFNMSI(T3a, T34), ms, &(x[0])); ST(&(x[WS(rs, 12)]), VFMAI(T3E, T3x), ms, &(x[0])); ST(&(x[WS(rs, 13)]), VFNMSI(T3E, T3x), ms, &(x[WS(rs, 1)])); ST(&(x[WS(rs, 18)]), VFNMSI(T3K, T3H), ms, &(x[0])); ST(&(x[WS(rs, 7)]), VFMAI(T3K, T3H), ms, &(x[WS(rs, 1)])); T1V = VFMA(LDK(KP559154169), T1U, T1T); T22 = VFNMS(LDK(KP683113946), T1T, T1U); T29 = VFNMS(LDK(KP867381224), T28, T27); T2u = VFMA(LDK(KP867381224), T28, T27); T2l = VFMA(LDK(KP894834959), T2k, T2h); T2v = VFMA(LDK(KP447417479), T2k, T2u); T2d = VFNMS(LDK(KP809385824), T2c, T29); T2p = VFMA(LDK(KP447417479), T2c, T2o); T1Q = VFMA(LDK(KP831864738), T1H, T1G); T1I = VFNMS(LDK(KP831864738), T1H, T1G); T2w = VFNMS(LDK(KP763932022), T2v, T2h); T1A = VFMA(LDK(KP904730450), T1z, TU); T1F = VFNMS(LDK(KP904730450), T1z, TU); T2q = VFMA(LDK(KP690983005), T2p, T29); } } { V T2e, T1E, T1P, T2m; T2e = VFNMS(LDK(KP992114701), T2d, Tf); T1E = VFMA(LDK(KP916574801), T1D, T1C); T1P = VFNMS(LDK(KP916574801), T1D, T1C); T2m = VMUL(LDK(KP951056516), VFNMS(LDK(KP992114701), T2l, T1O)); { V T1J, T2r, T1R, T1W, T1Z, T2x; T2x = VFNMS(LDK(KP999544308), T2w, T2t); T1J = VFNMS(LDK(KP904730450), T1I, T1F); T25 = VFMA(LDK(KP968583161), T1A, Tf); T1B = VFNMS(LDK(KP242145790), T1A, Tf); T2r = VFNMS(LDK(KP999544308), T2q, T2n); T1R = VFMA(LDK(KP904730450), T1Q, T1P); T1W = VFNMS(LDK(KP904730450), T1Q, T1P); T1Z = VADD(T1E, T1F); ST(&(x[WS(rs, 21)]), VFNMSI(T2m, T2e), ms, &(x[WS(rs, 1)])); ST(&(x[WS(rs, 4)]), VFMAI(T2m, T2e), ms, &(x[0])); T2y = VMUL(LDK(KP951056516), VFNMS(LDK(KP803003575), T2x, T1O)); T1K = VFNMS(LDK(KP618033988), T1J, T1E); T2s = VFNMS(LDK(KP803003575), T2r, Tf); T23 = VFMA(LDK(KP617882369), T1W, T22); T1S = VFNMS(LDK(KP242145790), T1R, T1O); T26 = VMUL(LDK(KP951056516), VFMA(LDK(KP968583161), T1R, T1O)); T20 = VFNMS(LDK(KP683113946), T1Z, T1I); T1X = VFMA(LDK(KP559016994), T1W, T1V); } } } } } { V T1L, T24, T21, T1Y; T1L = VFNMS(LDK(KP876091699), T1K, T1B); ST(&(x[WS(rs, 9)]), VFMAI(T2y, T2s), ms, &(x[WS(rs, 1)])); ST(&(x[WS(rs, 16)]), VFNMSI(T2y, T2s), ms, &(x[0])); T24 = VMUL(LDK(KP951056516), VFNMS(LDK(KP876306680), T23, T1S)); ST(&(x[WS(rs, 24)]), VFMAI(T26, T25), ms, &(x[0])); ST(&(x[WS(rs, 1)]), VFNMSI(T26, T25), ms, &(x[WS(rs, 1)])); T21 = VFMA(LDK(KP792626838), T20, T1B); T1Y = VMUL(LDK(KP951056516), VFMA(LDK(KP968583161), T1X, T1S)); ST(&(x[WS(rs, 11)]), VFNMSI(T24, T21), ms, &(x[WS(rs, 1)])); ST(&(x[WS(rs, 14)]), VFMAI(T24, T21), ms, &(x[0])); ST(&(x[WS(rs, 19)]), VFMAI(T1Y, T1L), ms, &(x[WS(rs, 1)])); ST(&(x[WS(rs, 6)]), VFNMSI(T1Y, T1L), ms, &(x[0])); } } } VLEAVE(); } static const tw_instr twinstr[] = { VTW(0, 1), VTW(0, 2), VTW(0, 3), VTW(0, 4), VTW(0, 5), VTW(0, 6), VTW(0, 7), VTW(0, 8), VTW(0, 9), VTW(0, 10), VTW(0, 11), VTW(0, 12), VTW(0, 13), VTW(0, 14), VTW(0, 15), VTW(0, 16), VTW(0, 17), VTW(0, 18), VTW(0, 19), VTW(0, 20), VTW(0, 21), VTW(0, 22), VTW(0, 23), VTW(0, 24), {TW_NEXT, VL, 0} }; static const ct_desc desc = { 25, XSIMD_STRING("t1fv_25"), twinstr, &GENUS, {67, 60, 181, 0}, 0, 0, 0 }; void XSIMD(codelet_t1fv_25) (planner *p) { X(kdft_dit_register) (p, t1fv_25, &desc); } #else /* HAVE_FMA */ /* Generated by: ../../../genfft/gen_twiddle_c.native -simd -compact -variables 4 -pipeline-latency 8 -n 25 -name t1fv_25 -include t1f.h */ /* * This function contains 248 FP additions, 188 FP multiplications, * (or, 170 additions, 110 multiplications, 78 fused multiply/add), * 99 stack variables, 40 constants, and 50 memory accesses */ #include "t1f.h" static void t1fv_25(R *ri, R *ii, const R *W, stride rs, INT mb, INT me, INT ms) { DVK(KP998026728, +0.998026728428271561952336806863450553336905220); DVK(KP125581039, +0.125581039058626752152356449131262266244969664); DVK(KP1_996053456, +1.996053456856543123904673613726901106673810439); DVK(KP062790519, +0.062790519529313376076178224565631133122484832); DVK(KP809016994, +0.809016994374947424102293417182819058860154590); DVK(KP309016994, +0.309016994374947424102293417182819058860154590); DVK(KP1_369094211, +1.369094211857377347464566715242418539779038465); DVK(KP728968627, +0.728968627421411523146730319055259111372571664); DVK(KP963507348, +0.963507348203430549974383005744259307057084020); DVK(KP876306680, +0.876306680043863587308115903922062583399064238); DVK(KP497379774, +0.497379774329709576484567492012895936835134813); DVK(KP968583161, +0.968583161128631119490168375464735813836012403); DVK(KP684547105, +0.684547105928688673732283357621209269889519233); DVK(KP1_457937254, +1.457937254842823046293460638110518222745143328); DVK(KP481753674, +0.481753674101715274987191502872129653528542010); DVK(KP1_752613360, +1.752613360087727174616231807844125166798128477); DVK(KP248689887, +0.248689887164854788242283746006447968417567406); DVK(KP1_937166322, +1.937166322257262238980336750929471627672024806); DVK(KP992114701, +0.992114701314477831049793042785778521453036709); DVK(KP250666467, +0.250666467128608490746237519633017587885836494); DVK(KP425779291, +0.425779291565072648862502445744251703979973042); DVK(KP1_809654104, +1.809654104932039055427337295865395187940827822); DVK(KP1_274847979, +1.274847979497379420353425623352032390869834596); DVK(KP770513242, +0.770513242775789230803009636396177847271667672); DVK(KP844327925, +0.844327925502015078548558063966681505381659241); DVK(KP1_071653589, +1.071653589957993236542617535735279956127150691); DVK(KP125333233, +0.125333233564304245373118759816508793942918247); DVK(KP1_984229402, +1.984229402628955662099586085571557042906073418); DVK(KP904827052, +0.904827052466019527713668647932697593970413911); DVK(KP851558583, +0.851558583130145297725004891488503407959946084); DVK(KP637423989, +0.637423989748689710176712811676016195434917298); DVK(KP1_541026485, +1.541026485551578461606019272792355694543335344); DVK(KP535826794, +0.535826794978996618271308767867639978063575346); DVK(KP1_688655851, +1.688655851004030157097116127933363010763318483); DVK(KP293892626, +0.293892626146236564584352977319536384298826219); DVK(KP475528258, +0.475528258147576786058219666689691071702849317); DVK(KP587785252, +0.587785252292473129168705954639072768597652438); DVK(KP951056516, +0.951056516295153572116439333379382143405698634); DVK(KP250000000, +0.250000000000000000000000000000000000000000000); DVK(KP559016994, +0.559016994374947424102293417182819058860154590); { INT m; R *x; x = ri; for (m = mb, W = W + (mb * ((TWVL / VL) * 48)); m < me; m = m + VL, x = x + (VL * ms), W = W + (TWVL * 48), MAKE_VOLATILE_STRIDE(25, rs)) { V Tc, Tb, Td, Te, T1C, T2t, T1E, T1x, T2m, T1u, T3c, T2n, Ty, T2i, Tv; V T38, T2j, TS, T2f, TP, T39, T2g, T1d, T2p, T1a, T3b, T2q; { V T7, T9, Ta, T2, T4, T5, T1D; Tc = LD(&(x[0]), ms, &(x[0])); { V T6, T8, T1, T3; T6 = LD(&(x[WS(rs, 10)]), ms, &(x[0])); T7 = BYTWJ(&(W[TWVL * 18]), T6); T8 = LD(&(x[WS(rs, 15)]), ms, &(x[WS(rs, 1)])); T9 = BYTWJ(&(W[TWVL * 28]), T8); Ta = VADD(T7, T9); T1 = LD(&(x[WS(rs, 5)]), ms, &(x[WS(rs, 1)])); T2 = BYTWJ(&(W[TWVL * 8]), T1); T3 = LD(&(x[WS(rs, 20)]), ms, &(x[0])); T4 = BYTWJ(&(W[TWVL * 38]), T3); T5 = VADD(T2, T4); } Tb = VMUL(LDK(KP559016994), VSUB(T5, Ta)); Td = VADD(T5, Ta); Te = VFNMS(LDK(KP250000000), Td, Tc); T1C = VSUB(T2, T4); T1D = VSUB(T7, T9); T2t = VMUL(LDK(KP951056516), T1D); T1E = VFMA(LDK(KP951056516), T1C, VMUL(LDK(KP587785252), T1D)); } { V T1r, T1l, T1n, T1o, T1g, T1i, T1j, T1q; T1q = LD(&(x[WS(rs, 3)]), ms, &(x[WS(rs, 1)])); T1r = BYTWJ(&(W[TWVL * 4]), T1q); { V T1k, T1m, T1f, T1h; T1k = LD(&(x[WS(rs, 13)]), ms, &(x[WS(rs, 1)])); T1l = BYTWJ(&(W[TWVL * 24]), T1k); T1m = LD(&(x[WS(rs, 18)]), ms, &(x[0])); T1n = BYTWJ(&(W[TWVL * 34]), T1m); T1o = VADD(T1l, T1n); T1f = LD(&(x[WS(rs, 8)]), ms, &(x[0])); T1g = BYTWJ(&(W[TWVL * 14]), T1f); T1h = LD(&(x[WS(rs, 23)]), ms, &(x[WS(rs, 1)])); T1i = BYTWJ(&(W[TWVL * 44]), T1h); T1j = VADD(T1g, T1i); } { V T1v, T1w, T1p, T1s, T1t; T1v = VSUB(T1g, T1i); T1w = VSUB(T1l, T1n); T1x = VFMA(LDK(KP475528258), T1v, VMUL(LDK(KP293892626), T1w)); T2m = VFNMS(LDK(KP293892626), T1v, VMUL(LDK(KP475528258), T1w)); T1p = VMUL(LDK(KP559016994), VSUB(T1j, T1o)); T1s = VADD(T1j, T1o); T1t = VFNMS(LDK(KP250000000), T1s, T1r); T1u = VADD(T1p, T1t); T3c = VADD(T1r, T1s); T2n = VSUB(T1t, T1p); } } { V Ts, Tm, To, Tp, Th, Tj, Tk, Tr; Tr = LD(&(x[WS(rs, 1)]), ms, &(x[WS(rs, 1)])); Ts = BYTWJ(&(W[0]), Tr); { V Tl, Tn, Tg, Ti; Tl = LD(&(x[WS(rs, 11)]), ms, &(x[WS(rs, 1)])); Tm = BYTWJ(&(W[TWVL * 20]), Tl); Tn = LD(&(x[WS(rs, 16)]), ms, &(x[0])); To = BYTWJ(&(W[TWVL * 30]), Tn); Tp = VADD(Tm, To); Tg = LD(&(x[WS(rs, 6)]), ms, &(x[0])); Th = BYTWJ(&(W[TWVL * 10]), Tg); Ti = LD(&(x[WS(rs, 21)]), ms, &(x[WS(rs, 1)])); Tj = BYTWJ(&(W[TWVL * 40]), Ti); Tk = VADD(Th, Tj); } { V Tw, Tx, Tq, Tt, Tu; Tw = VSUB(Th, Tj); Tx = VSUB(Tm, To); Ty = VFMA(LDK(KP475528258), Tw, VMUL(LDK(KP293892626), Tx)); T2i = VFNMS(LDK(KP293892626), Tw, VMUL(LDK(KP475528258), Tx)); Tq = VMUL(LDK(KP559016994), VSUB(Tk, Tp)); Tt = VADD(Tk, Tp); Tu = VFNMS(LDK(KP250000000), Tt, Ts); Tv = VADD(Tq, Tu); T38 = VADD(Ts, Tt); T2j = VSUB(Tu, Tq); } } { V TM, TG, TI, TJ, TB, TD, TE, TL; TL = LD(&(x[WS(rs, 4)]), ms, &(x[0])); TM = BYTWJ(&(W[TWVL * 6]), TL); { V TF, TH, TA, TC; TF = LD(&(x[WS(rs, 14)]), ms, &(x[0])); TG = BYTWJ(&(W[TWVL * 26]), TF); TH = LD(&(x[WS(rs, 19)]), ms, &(x[WS(rs, 1)])); TI = BYTWJ(&(W[TWVL * 36]), TH); TJ = VADD(TG, TI); TA = LD(&(x[WS(rs, 9)]), ms, &(x[WS(rs, 1)])); TB = BYTWJ(&(W[TWVL * 16]), TA); TC = LD(&(x[WS(rs, 24)]), ms, &(x[0])); TD = BYTWJ(&(W[TWVL * 46]), TC); TE = VADD(TB, TD); } { V TQ, TR, TK, TN, TO; TQ = VSUB(TB, TD); TR = VSUB(TG, TI); TS = VFMA(LDK(KP475528258), TQ, VMUL(LDK(KP293892626), TR)); T2f = VFNMS(LDK(KP293892626), TQ, VMUL(LDK(KP475528258), TR)); TK = VMUL(LDK(KP559016994), VSUB(TE, TJ)); TN = VADD(TE, TJ); TO = VFNMS(LDK(KP250000000), TN, TM); TP = VADD(TK, TO); T39 = VADD(TM, TN); T2g = VSUB(TO, TK); } } { V T17, T11, T13, T14, TW, TY, TZ, T16; T16 = LD(&(x[WS(rs, 2)]), ms, &(x[0])); T17 = BYTWJ(&(W[TWVL * 2]), T16); { V T10, T12, TV, TX; T10 = LD(&(x[WS(rs, 12)]), ms, &(x[0])); T11 = BYTWJ(&(W[TWVL * 22]), T10); T12 = LD(&(x[WS(rs, 17)]), ms, &(x[WS(rs, 1)])); T13 = BYTWJ(&(W[TWVL * 32]), T12); T14 = VADD(T11, T13); TV = LD(&(x[WS(rs, 7)]), ms, &(x[WS(rs, 1)])); TW = BYTWJ(&(W[TWVL * 12]), TV); TX = LD(&(x[WS(rs, 22)]), ms, &(x[0])); TY = BYTWJ(&(W[TWVL * 42]), TX); TZ = VADD(TW, TY); } { V T1b, T1c, T15, T18, T19; T1b = VSUB(TW, TY); T1c = VSUB(T11, T13); T1d = VFMA(LDK(KP475528258), T1b, VMUL(LDK(KP293892626), T1c)); T2p = VFNMS(LDK(KP293892626), T1b, VMUL(LDK(KP475528258), T1c)); T15 = VMUL(LDK(KP559016994), VSUB(TZ, T14)); T18 = VADD(TZ, T14); T19 = VFNMS(LDK(KP250000000), T18, T17); T1a = VADD(T15, T19); T3b = VADD(T17, T18); T2q = VSUB(T19, T15); } } { V T3l, T3m, T3f, T3g, T3e, T3h, T3n, T3i; { V T3j, T3k, T3a, T3d; T3j = VSUB(T38, T39); T3k = VSUB(T3b, T3c); T3l = VBYI(VFMA(LDK(KP951056516), T3j, VMUL(LDK(KP587785252), T3k))); T3m = VBYI(VFNMS(LDK(KP587785252), T3j, VMUL(LDK(KP951056516), T3k))); T3f = VADD(Tc, Td); T3a = VADD(T38, T39); T3d = VADD(T3b, T3c); T3g = VADD(T3a, T3d); T3e = VMUL(LDK(KP559016994), VSUB(T3a, T3d)); T3h = VFNMS(LDK(KP250000000), T3g, T3f); } ST(&(x[0]), VADD(T3f, T3g), ms, &(x[0])); T3n = VSUB(T3h, T3e); ST(&(x[WS(rs, 10)]), VADD(T3m, T3n), ms, &(x[0])); ST(&(x[WS(rs, 15)]), VSUB(T3n, T3m), ms, &(x[WS(rs, 1)])); T3i = VADD(T3e, T3h); ST(&(x[WS(rs, 5)]), VSUB(T3i, T3l), ms, &(x[WS(rs, 1)])); ST(&(x[WS(rs, 20)]), VADD(T3l, T3i), ms, &(x[0])); } { V Tf, T1Z, T20, T21, T29, T2a, T2b, T26, T27, T28, T22, T23, T24, T1L, T1U; V T1Q, T1S, T1A, T1V, T1N, T1O, T2d, T2e; Tf = VADD(Tb, Te); T1Z = VFMA(LDK(KP1_688655851), Ty, VMUL(LDK(KP535826794), Tv)); T20 = VFMA(LDK(KP1_541026485), TS, VMUL(LDK(KP637423989), TP)); T21 = VSUB(T1Z, T20); T29 = VFMA(LDK(KP851558583), T1d, VMUL(LDK(KP904827052), T1a)); T2a = VFMA(LDK(KP1_984229402), T1x, VMUL(LDK(KP125333233), T1u)); T2b = VADD(T29, T2a); T26 = VFNMS(LDK(KP844327925), Tv, VMUL(LDK(KP1_071653589), Ty)); T27 = VFNMS(LDK(KP1_274847979), TS, VMUL(LDK(KP770513242), TP)); T28 = VADD(T26, T27); T22 = VFNMS(LDK(KP425779291), T1a, VMUL(LDK(KP1_809654104), T1d)); T23 = VFNMS(LDK(KP992114701), T1u, VMUL(LDK(KP250666467), T1x)); T24 = VADD(T22, T23); { V T1F, T1G, T1H, T1I, T1J, T1K; T1F = VFMA(LDK(KP1_937166322), Ty, VMUL(LDK(KP248689887), Tv)); T1G = VFMA(LDK(KP1_071653589), TS, VMUL(LDK(KP844327925), TP)); T1H = VADD(T1F, T1G); T1I = VFMA(LDK(KP1_752613360), T1d, VMUL(LDK(KP481753674), T1a)); T1J = VFMA(LDK(KP1_457937254), T1x, VMUL(LDK(KP684547105), T1u)); T1K = VADD(T1I, T1J); T1L = VADD(T1H, T1K); T1U = VSUB(T1J, T1I); T1Q = VMUL(LDK(KP559016994), VSUB(T1K, T1H)); T1S = VSUB(T1G, T1F); } { V Tz, TT, TU, T1e, T1y, T1z; Tz = VFNMS(LDK(KP497379774), Ty, VMUL(LDK(KP968583161), Tv)); TT = VFNMS(LDK(KP1_688655851), TS, VMUL(LDK(KP535826794), TP)); TU = VADD(Tz, TT); T1e = VFNMS(LDK(KP963507348), T1d, VMUL(LDK(KP876306680), T1a)); T1y = VFNMS(LDK(KP1_369094211), T1x, VMUL(LDK(KP728968627), T1u)); T1z = VADD(T1e, T1y); T1A = VADD(TU, T1z); T1V = VMUL(LDK(KP559016994), VSUB(TU, T1z)); T1N = VSUB(TT, Tz); T1O = VSUB(T1e, T1y); } { V T1B, T1M, T25, T2c; T1B = VADD(Tf, T1A); T1M = VBYI(VADD(T1E, T1L)); ST(&(x[WS(rs, 1)]), VSUB(T1B, T1M), ms, &(x[WS(rs, 1)])); ST(&(x[WS(rs, 24)]), VADD(T1B, T1M), ms, &(x[0])); T25 = VADD(Tf, VADD(T21, T24)); T2c = VBYI(VADD(T1E, VSUB(T28, T2b))); ST(&(x[WS(rs, 21)]), VSUB(T25, T2c), ms, &(x[WS(rs, 1)])); ST(&(x[WS(rs, 4)]), VADD(T25, T2c), ms, &(x[0])); } T2d = VBYI(VADD(T1E, VFMA(LDK(KP309016994), T28, VFMA(LDK(KP587785252), VSUB(T23, T22), VFNMS(LDK(KP951056516), VADD(T1Z, T20), VMUL(LDK(KP809016994), T2b)))))); T2e = VFMA(LDK(KP309016994), T21, VFMA(LDK(KP951056516), VSUB(T26, T27), VFMA(LDK(KP587785252), VSUB(T2a, T29), VFNMS(LDK(KP809016994), T24, Tf)))); ST(&(x[WS(rs, 9)]), VADD(T2d, T2e), ms, &(x[WS(rs, 1)])); ST(&(x[WS(rs, 16)]), VSUB(T2e, T2d), ms, &(x[0])); { V T1R, T1X, T1W, T1Y, T1P, T1T; T1P = VFMS(LDK(KP250000000), T1L, T1E); T1R = VBYI(VADD(VFMA(LDK(KP587785252), T1N, VMUL(LDK(KP951056516), T1O)), VSUB(T1P, T1Q))); T1X = VBYI(VADD(VFNMS(LDK(KP587785252), T1O, VMUL(LDK(KP951056516), T1N)), VADD(T1P, T1Q))); T1T = VFNMS(LDK(KP250000000), T1A, Tf); T1W = VFMA(LDK(KP587785252), T1S, VFNMS(LDK(KP951056516), T1U, VSUB(T1T, T1V))); T1Y = VFMA(LDK(KP951056516), T1S, VADD(T1V, VFMA(LDK(KP587785252), T1U, T1T))); ST(&(x[WS(rs, 11)]), VADD(T1R, T1W), ms, &(x[WS(rs, 1)])); ST(&(x[WS(rs, 19)]), VSUB(T1Y, T1X), ms, &(x[WS(rs, 1)])); ST(&(x[WS(rs, 14)]), VSUB(T1W, T1R), ms, &(x[0])); ST(&(x[WS(rs, 6)]), VADD(T1X, T1Y), ms, &(x[0])); } } { V T2u, T2w, T2h, T2k, T2l, T2A, T2B, T2C, T2o, T2r, T2s, T2x, T2y, T2z, T2M; V T2X, T2N, T2W, T2R, T31, T2U, T30, T2E, T2F; T2u = VFNMS(LDK(KP587785252), T1C, T2t); T2w = VSUB(Te, Tb); T2h = VFNMS(LDK(KP125333233), T2g, VMUL(LDK(KP1_984229402), T2f)); T2k = VFMA(LDK(KP1_457937254), T2i, VMUL(LDK(KP684547105), T2j)); T2l = VSUB(T2h, T2k); T2A = VFNMS(LDK(KP1_996053456), T2p, VMUL(LDK(KP062790519), T2q)); T2B = VFMA(LDK(KP1_541026485), T2m, VMUL(LDK(KP637423989), T2n)); T2C = VSUB(T2A, T2B); T2o = VFNMS(LDK(KP770513242), T2n, VMUL(LDK(KP1_274847979), T2m)); T2r = VFMA(LDK(KP125581039), T2p, VMUL(LDK(KP998026728), T2q)); T2s = VSUB(T2o, T2r); T2x = VFNMS(LDK(KP1_369094211), T2i, VMUL(LDK(KP728968627), T2j)); T2y = VFMA(LDK(KP250666467), T2f, VMUL(LDK(KP992114701), T2g)); T2z = VSUB(T2x, T2y); { V T2G, T2H, T2I, T2J, T2K, T2L; T2G = VFNMS(LDK(KP481753674), T2j, VMUL(LDK(KP1_752613360), T2i)); T2H = VFMA(LDK(KP851558583), T2f, VMUL(LDK(KP904827052), T2g)); T2I = VSUB(T2G, T2H); T2J = VFNMS(LDK(KP844327925), T2q, VMUL(LDK(KP1_071653589), T2p)); T2K = VFNMS(LDK(KP998026728), T2n, VMUL(LDK(KP125581039), T2m)); T2L = VADD(T2J, T2K); T2M = VMUL(LDK(KP559016994), VSUB(T2I, T2L)); T2X = VSUB(T2J, T2K); T2N = VADD(T2I, T2L); T2W = VADD(T2G, T2H); } { V T2P, T2Q, T2Y, T2S, T2T, T2Z; T2P = VFNMS(LDK(KP425779291), T2g, VMUL(LDK(KP1_809654104), T2f)); T2Q = VFMA(LDK(KP963507348), T2i, VMUL(LDK(KP876306680), T2j)); T2Y = VADD(T2Q, T2P); T2S = VFMA(LDK(KP1_688655851), T2p, VMUL(LDK(KP535826794), T2q)); T2T = VFMA(LDK(KP1_996053456), T2m, VMUL(LDK(KP062790519), T2n)); T2Z = VADD(T2S, T2T); T2R = VSUB(T2P, T2Q); T31 = VADD(T2Y, T2Z); T2U = VSUB(T2S, T2T); T30 = VMUL(LDK(KP559016994), VSUB(T2Y, T2Z)); } { V T36, T37, T2v, T2D; T36 = VBYI(VADD(T2u, T2N)); T37 = VADD(T2w, T31); ST(&(x[WS(rs, 2)]), VADD(T36, T37), ms, &(x[0])); ST(&(x[WS(rs, 23)]), VSUB(T37, T36), ms, &(x[WS(rs, 1)])); T2v = VBYI(VSUB(VADD(T2l, T2s), T2u)); T2D = VADD(T2w, VADD(T2z, T2C)); ST(&(x[WS(rs, 3)]), VADD(T2v, T2D), ms, &(x[WS(rs, 1)])); ST(&(x[WS(rs, 22)]), VSUB(T2D, T2v), ms, &(x[0])); } T2E = VFMA(LDK(KP309016994), T2z, VFNMS(LDK(KP809016994), T2C, VFNMS(LDK(KP587785252), VADD(T2r, T2o), VFNMS(LDK(KP951056516), VADD(T2k, T2h), T2w)))); T2F = VBYI(VSUB(VFNMS(LDK(KP587785252), VADD(T2A, T2B), VFNMS(LDK(KP809016994), T2s, VFNMS(LDK(KP951056516), VADD(T2x, T2y), VMUL(LDK(KP309016994), T2l)))), T2u)); ST(&(x[WS(rs, 17)]), VSUB(T2E, T2F), ms, &(x[WS(rs, 1)])); ST(&(x[WS(rs, 8)]), VADD(T2E, T2F), ms, &(x[0])); { V T2V, T34, T33, T35, T2O, T32; T2O = VFNMS(LDK(KP250000000), T2N, T2u); T2V = VBYI(VADD(T2M, VADD(T2O, VFNMS(LDK(KP587785252), T2U, VMUL(LDK(KP951056516), T2R))))); T34 = VBYI(VADD(T2O, VSUB(VFMA(LDK(KP587785252), T2R, VMUL(LDK(KP951056516), T2U)), T2M))); T32 = VFNMS(LDK(KP250000000), T31, T2w); T33 = VFMA(LDK(KP951056516), T2W, VFMA(LDK(KP587785252), T2X, VADD(T30, T32))); T35 = VFMA(LDK(KP587785252), T2W, VSUB(VFNMS(LDK(KP951056516), T2X, T32), T30)); ST(&(x[WS(rs, 7)]), VADD(T2V, T33), ms, &(x[WS(rs, 1)])); ST(&(x[WS(rs, 13)]), VSUB(T35, T34), ms, &(x[WS(rs, 1)])); ST(&(x[WS(rs, 18)]), VSUB(T33, T2V), ms, &(x[0])); ST(&(x[WS(rs, 12)]), VADD(T34, T35), ms, &(x[0])); } } } } VLEAVE(); } static const tw_instr twinstr[] = { VTW(0, 1), VTW(0, 2), VTW(0, 3), VTW(0, 4), VTW(0, 5), VTW(0, 6), VTW(0, 7), VTW(0, 8), VTW(0, 9), VTW(0, 10), VTW(0, 11), VTW(0, 12), VTW(0, 13), VTW(0, 14), VTW(0, 15), VTW(0, 16), VTW(0, 17), VTW(0, 18), VTW(0, 19), VTW(0, 20), VTW(0, 21), VTW(0, 22), VTW(0, 23), VTW(0, 24), {TW_NEXT, VL, 0} }; static const ct_desc desc = { 25, XSIMD_STRING("t1fv_25"), twinstr, &GENUS, {170, 110, 78, 0}, 0, 0, 0 }; void XSIMD(codelet_t1fv_25) (planner *p) { X(kdft_dit_register) (p, t1fv_25, &desc); } #endif /* HAVE_FMA */ fftw-3.3.3/dft/simd/common/t2fv_2.c0000644000175000017500000000653612054410513013663 00000000000000/* * Copyright (c) 2003, 2007-11 Matteo Frigo * Copyright (c) 2003, 2007-11 Massachusetts Institute of Technology * * This program is free software; you can redistribute it and/or modify * it under the terms of the GNU General Public License as published by * the Free Software Foundation; either version 2 of the License, or * (at your option) any later version. * * This program is distributed in the hope that it will be useful, * but WITHOUT ANY WARRANTY; without even the implied warranty of * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the * GNU General Public License for more details. * * You should have received a copy of the GNU General Public License * along with this program; if not, write to the Free Software * Foundation, Inc., 51 Franklin Street, Fifth Floor, Boston, MA 02110-1301 USA * */ /* This file was automatically generated --- DO NOT EDIT */ /* Generated on Sun Nov 25 07:38:35 EST 2012 */ #include "codelet-dft.h" #ifdef HAVE_FMA /* Generated by: ../../../genfft/gen_twiddle_c.native -fma -reorder-insns -schedule-for-pipeline -simd -compact -variables 4 -pipeline-latency 8 -n 2 -name t2fv_2 -include t2f.h */ /* * This function contains 3 FP additions, 2 FP multiplications, * (or, 3 additions, 2 multiplications, 0 fused multiply/add), * 5 stack variables, 0 constants, and 4 memory accesses */ #include "t2f.h" static void t2fv_2(R *ri, R *ii, const R *W, stride rs, INT mb, INT me, INT ms) { { INT m; R *x; x = ri; for (m = mb, W = W + (mb * ((TWVL / VL) * 2)); m < me; m = m + VL, x = x + (VL * ms), W = W + (TWVL * 2), MAKE_VOLATILE_STRIDE(2, rs)) { V T1, T2, T3; T1 = LD(&(x[0]), ms, &(x[0])); T2 = LD(&(x[WS(rs, 1)]), ms, &(x[WS(rs, 1)])); T3 = BYTWJ(&(W[0]), T2); ST(&(x[0]), VADD(T1, T3), ms, &(x[0])); ST(&(x[WS(rs, 1)]), VSUB(T1, T3), ms, &(x[WS(rs, 1)])); } } VLEAVE(); } static const tw_instr twinstr[] = { VTW(0, 1), {TW_NEXT, VL, 0} }; static const ct_desc desc = { 2, XSIMD_STRING("t2fv_2"), twinstr, &GENUS, {3, 2, 0, 0}, 0, 0, 0 }; void XSIMD(codelet_t2fv_2) (planner *p) { X(kdft_dit_register) (p, t2fv_2, &desc); } #else /* HAVE_FMA */ /* Generated by: ../../../genfft/gen_twiddle_c.native -simd -compact -variables 4 -pipeline-latency 8 -n 2 -name t2fv_2 -include t2f.h */ /* * This function contains 3 FP additions, 2 FP multiplications, * (or, 3 additions, 2 multiplications, 0 fused multiply/add), * 5 stack variables, 0 constants, and 4 memory accesses */ #include "t2f.h" static void t2fv_2(R *ri, R *ii, const R *W, stride rs, INT mb, INT me, INT ms) { { INT m; R *x; x = ri; for (m = mb, W = W + (mb * ((TWVL / VL) * 2)); m < me; m = m + VL, x = x + (VL * ms), W = W + (TWVL * 2), MAKE_VOLATILE_STRIDE(2, rs)) { V T1, T3, T2; T1 = LD(&(x[0]), ms, &(x[0])); T2 = LD(&(x[WS(rs, 1)]), ms, &(x[WS(rs, 1)])); T3 = BYTWJ(&(W[0]), T2); ST(&(x[WS(rs, 1)]), VSUB(T1, T3), ms, &(x[WS(rs, 1)])); ST(&(x[0]), VADD(T1, T3), ms, &(x[0])); } } VLEAVE(); } static const tw_instr twinstr[] = { VTW(0, 1), {TW_NEXT, VL, 0} }; static const ct_desc desc = { 2, XSIMD_STRING("t2fv_2"), twinstr, &GENUS, {3, 2, 0, 0}, 0, 0, 0 }; void XSIMD(codelet_t2fv_2) (planner *p) { X(kdft_dit_register) (p, t2fv_2, &desc); } #endif /* HAVE_FMA */ fftw-3.3.3/dft/simd/common/t2fv_4.c0000644000175000017500000001047112054410513013656 00000000000000/* * Copyright (c) 2003, 2007-11 Matteo Frigo * Copyright (c) 2003, 2007-11 Massachusetts Institute of Technology * * This program is free software; you can redistribute it and/or modify * it under the terms of the GNU General Public License as published by * the Free Software Foundation; either version 2 of the License, or * (at your option) any later version. * * This program is distributed in the hope that it will be useful, * but WITHOUT ANY WARRANTY; without even the implied warranty of * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the * GNU General Public License for more details. * * You should have received a copy of the GNU General Public License * along with this program; if not, write to the Free Software * Foundation, Inc., 51 Franklin Street, Fifth Floor, Boston, MA 02110-1301 USA * */ /* This file was automatically generated --- DO NOT EDIT */ /* Generated on Sun Nov 25 07:38:35 EST 2012 */ #include "codelet-dft.h" #ifdef HAVE_FMA /* Generated by: ../../../genfft/gen_twiddle_c.native -fma -reorder-insns -schedule-for-pipeline -simd -compact -variables 4 -pipeline-latency 8 -n 4 -name t2fv_4 -include t2f.h */ /* * This function contains 11 FP additions, 8 FP multiplications, * (or, 9 additions, 6 multiplications, 2 fused multiply/add), * 13 stack variables, 0 constants, and 8 memory accesses */ #include "t2f.h" static void t2fv_4(R *ri, R *ii, const R *W, stride rs, INT mb, INT me, INT ms) { { INT m; R *x; x = ri; for (m = mb, W = W + (mb * ((TWVL / VL) * 6)); m < me; m = m + VL, x = x + (VL * ms), W = W + (TWVL * 6), MAKE_VOLATILE_STRIDE(4, rs)) { V T1, T7, T2, T5, T8, T3, T6; T1 = LD(&(x[0]), ms, &(x[0])); T7 = LD(&(x[WS(rs, 3)]), ms, &(x[WS(rs, 1)])); T2 = LD(&(x[WS(rs, 2)]), ms, &(x[0])); T5 = LD(&(x[WS(rs, 1)]), ms, &(x[WS(rs, 1)])); T8 = BYTWJ(&(W[TWVL * 4]), T7); T3 = BYTWJ(&(W[TWVL * 2]), T2); T6 = BYTWJ(&(W[0]), T5); { V Ta, T4, Tb, T9; Ta = VADD(T1, T3); T4 = VSUB(T1, T3); Tb = VADD(T6, T8); T9 = VSUB(T6, T8); ST(&(x[0]), VADD(Ta, Tb), ms, &(x[0])); ST(&(x[WS(rs, 2)]), VSUB(Ta, Tb), ms, &(x[0])); ST(&(x[WS(rs, 3)]), VFMAI(T9, T4), ms, &(x[WS(rs, 1)])); ST(&(x[WS(rs, 1)]), VFNMSI(T9, T4), ms, &(x[WS(rs, 1)])); } } } VLEAVE(); } static const tw_instr twinstr[] = { VTW(0, 1), VTW(0, 2), VTW(0, 3), {TW_NEXT, VL, 0} }; static const ct_desc desc = { 4, XSIMD_STRING("t2fv_4"), twinstr, &GENUS, {9, 6, 2, 0}, 0, 0, 0 }; void XSIMD(codelet_t2fv_4) (planner *p) { X(kdft_dit_register) (p, t2fv_4, &desc); } #else /* HAVE_FMA */ /* Generated by: ../../../genfft/gen_twiddle_c.native -simd -compact -variables 4 -pipeline-latency 8 -n 4 -name t2fv_4 -include t2f.h */ /* * This function contains 11 FP additions, 6 FP multiplications, * (or, 11 additions, 6 multiplications, 0 fused multiply/add), * 13 stack variables, 0 constants, and 8 memory accesses */ #include "t2f.h" static void t2fv_4(R *ri, R *ii, const R *W, stride rs, INT mb, INT me, INT ms) { { INT m; R *x; x = ri; for (m = mb, W = W + (mb * ((TWVL / VL) * 6)); m < me; m = m + VL, x = x + (VL * ms), W = W + (TWVL * 6), MAKE_VOLATILE_STRIDE(4, rs)) { V T1, T8, T3, T6, T7, T2, T5; T1 = LD(&(x[0]), ms, &(x[0])); T7 = LD(&(x[WS(rs, 3)]), ms, &(x[WS(rs, 1)])); T8 = BYTWJ(&(W[TWVL * 4]), T7); T2 = LD(&(x[WS(rs, 2)]), ms, &(x[0])); T3 = BYTWJ(&(W[TWVL * 2]), T2); T5 = LD(&(x[WS(rs, 1)]), ms, &(x[WS(rs, 1)])); T6 = BYTWJ(&(W[0]), T5); { V T4, T9, Ta, Tb; T4 = VSUB(T1, T3); T9 = VBYI(VSUB(T6, T8)); ST(&(x[WS(rs, 1)]), VSUB(T4, T9), ms, &(x[WS(rs, 1)])); ST(&(x[WS(rs, 3)]), VADD(T4, T9), ms, &(x[WS(rs, 1)])); Ta = VADD(T1, T3); Tb = VADD(T6, T8); ST(&(x[WS(rs, 2)]), VSUB(Ta, Tb), ms, &(x[0])); ST(&(x[0]), VADD(Ta, Tb), ms, &(x[0])); } } } VLEAVE(); } static const tw_instr twinstr[] = { VTW(0, 1), VTW(0, 2), VTW(0, 3), {TW_NEXT, VL, 0} }; static const ct_desc desc = { 4, XSIMD_STRING("t2fv_4"), twinstr, &GENUS, {11, 6, 0, 0}, 0, 0, 0 }; void XSIMD(codelet_t2fv_4) (planner *p) { X(kdft_dit_register) (p, t2fv_4, &desc); } #endif /* HAVE_FMA */ fftw-3.3.3/dft/simd/common/t2fv_8.c0000644000175000017500000001560312054410513013664 00000000000000/* * Copyright (c) 2003, 2007-11 Matteo Frigo * Copyright (c) 2003, 2007-11 Massachusetts Institute of Technology * * This program is free software; you can redistribute it and/or modify * it under the terms of the GNU General Public License as published by * the Free Software Foundation; either version 2 of the License, or * (at your option) any later version. * * This program is distributed in the hope that it will be useful, * but WITHOUT ANY WARRANTY; without even the implied warranty of * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the * GNU General Public License for more details. * * You should have received a copy of the GNU General Public License * along with this program; if not, write to the Free Software * Foundation, Inc., 51 Franklin Street, Fifth Floor, Boston, MA 02110-1301 USA * */ /* This file was automatically generated --- DO NOT EDIT */ /* Generated on Sun Nov 25 07:38:35 EST 2012 */ #include "codelet-dft.h" #ifdef HAVE_FMA /* Generated by: ../../../genfft/gen_twiddle_c.native -fma -reorder-insns -schedule-for-pipeline -simd -compact -variables 4 -pipeline-latency 8 -n 8 -name t2fv_8 -include t2f.h */ /* * This function contains 33 FP additions, 24 FP multiplications, * (or, 23 additions, 14 multiplications, 10 fused multiply/add), * 36 stack variables, 1 constants, and 16 memory accesses */ #include "t2f.h" static void t2fv_8(R *ri, R *ii, const R *W, stride rs, INT mb, INT me, INT ms) { DVK(KP707106781, +0.707106781186547524400844362104849039284835938); { INT m; R *x; x = ri; for (m = mb, W = W + (mb * ((TWVL / VL) * 14)); m < me; m = m + VL, x = x + (VL * ms), W = W + (TWVL * 14), MAKE_VOLATILE_STRIDE(8, rs)) { V T1, T2, Th, Tj, T5, T7, Ta, Tc; T1 = LD(&(x[0]), ms, &(x[0])); T2 = LD(&(x[WS(rs, 4)]), ms, &(x[0])); Th = LD(&(x[WS(rs, 2)]), ms, &(x[0])); Tj = LD(&(x[WS(rs, 6)]), ms, &(x[0])); T5 = LD(&(x[WS(rs, 1)]), ms, &(x[WS(rs, 1)])); T7 = LD(&(x[WS(rs, 5)]), ms, &(x[WS(rs, 1)])); Ta = LD(&(x[WS(rs, 7)]), ms, &(x[WS(rs, 1)])); Tc = LD(&(x[WS(rs, 3)]), ms, &(x[WS(rs, 1)])); { V T3, Ti, Tk, T6, T8, Tb, Td; T3 = BYTWJ(&(W[TWVL * 6]), T2); Ti = BYTWJ(&(W[TWVL * 2]), Th); Tk = BYTWJ(&(W[TWVL * 10]), Tj); T6 = BYTWJ(&(W[0]), T5); T8 = BYTWJ(&(W[TWVL * 8]), T7); Tb = BYTWJ(&(W[TWVL * 12]), Ta); Td = BYTWJ(&(W[TWVL * 4]), Tc); { V Tq, T4, Tr, Tl, Tt, T9, Tu, Te, Tw, Ts; Tq = VADD(T1, T3); T4 = VSUB(T1, T3); Tr = VADD(Ti, Tk); Tl = VSUB(Ti, Tk); Tt = VADD(T6, T8); T9 = VSUB(T6, T8); Tu = VADD(Tb, Td); Te = VSUB(Tb, Td); Tw = VSUB(Tq, Tr); Ts = VADD(Tq, Tr); { V Tx, Tv, Tm, Tf; Tx = VSUB(Tu, Tt); Tv = VADD(Tt, Tu); Tm = VSUB(Te, T9); Tf = VADD(T9, Te); { V Tp, Tn, To, Tg; ST(&(x[WS(rs, 2)]), VFMAI(Tx, Tw), ms, &(x[0])); ST(&(x[WS(rs, 6)]), VFNMSI(Tx, Tw), ms, &(x[0])); ST(&(x[0]), VADD(Ts, Tv), ms, &(x[0])); ST(&(x[WS(rs, 4)]), VSUB(Ts, Tv), ms, &(x[0])); Tp = VFMA(LDK(KP707106781), Tm, Tl); Tn = VFNMS(LDK(KP707106781), Tm, Tl); To = VFNMS(LDK(KP707106781), Tf, T4); Tg = VFMA(LDK(KP707106781), Tf, T4); ST(&(x[WS(rs, 5)]), VFNMSI(Tp, To), ms, &(x[WS(rs, 1)])); ST(&(x[WS(rs, 3)]), VFMAI(Tp, To), ms, &(x[WS(rs, 1)])); ST(&(x[WS(rs, 7)]), VFMAI(Tn, Tg), ms, &(x[WS(rs, 1)])); ST(&(x[WS(rs, 1)]), VFNMSI(Tn, Tg), ms, &(x[WS(rs, 1)])); } } } } } } VLEAVE(); } static const tw_instr twinstr[] = { VTW(0, 1), VTW(0, 2), VTW(0, 3), VTW(0, 4), VTW(0, 5), VTW(0, 6), VTW(0, 7), {TW_NEXT, VL, 0} }; static const ct_desc desc = { 8, XSIMD_STRING("t2fv_8"), twinstr, &GENUS, {23, 14, 10, 0}, 0, 0, 0 }; void XSIMD(codelet_t2fv_8) (planner *p) { X(kdft_dit_register) (p, t2fv_8, &desc); } #else /* HAVE_FMA */ /* Generated by: ../../../genfft/gen_twiddle_c.native -simd -compact -variables 4 -pipeline-latency 8 -n 8 -name t2fv_8 -include t2f.h */ /* * This function contains 33 FP additions, 16 FP multiplications, * (or, 33 additions, 16 multiplications, 0 fused multiply/add), * 24 stack variables, 1 constants, and 16 memory accesses */ #include "t2f.h" static void t2fv_8(R *ri, R *ii, const R *W, stride rs, INT mb, INT me, INT ms) { DVK(KP707106781, +0.707106781186547524400844362104849039284835938); { INT m; R *x; x = ri; for (m = mb, W = W + (mb * ((TWVL / VL) * 14)); m < me; m = m + VL, x = x + (VL * ms), W = W + (TWVL * 14), MAKE_VOLATILE_STRIDE(8, rs)) { V T4, Tq, Tm, Tr, T9, Tt, Te, Tu, T1, T3, T2; T1 = LD(&(x[0]), ms, &(x[0])); T2 = LD(&(x[WS(rs, 4)]), ms, &(x[0])); T3 = BYTWJ(&(W[TWVL * 6]), T2); T4 = VSUB(T1, T3); Tq = VADD(T1, T3); { V Tj, Tl, Ti, Tk; Ti = LD(&(x[WS(rs, 2)]), ms, &(x[0])); Tj = BYTWJ(&(W[TWVL * 2]), Ti); Tk = LD(&(x[WS(rs, 6)]), ms, &(x[0])); Tl = BYTWJ(&(W[TWVL * 10]), Tk); Tm = VSUB(Tj, Tl); Tr = VADD(Tj, Tl); } { V T6, T8, T5, T7; T5 = LD(&(x[WS(rs, 1)]), ms, &(x[WS(rs, 1)])); T6 = BYTWJ(&(W[0]), T5); T7 = LD(&(x[WS(rs, 5)]), ms, &(x[WS(rs, 1)])); T8 = BYTWJ(&(W[TWVL * 8]), T7); T9 = VSUB(T6, T8); Tt = VADD(T6, T8); } { V Tb, Td, Ta, Tc; Ta = LD(&(x[WS(rs, 7)]), ms, &(x[WS(rs, 1)])); Tb = BYTWJ(&(W[TWVL * 12]), Ta); Tc = LD(&(x[WS(rs, 3)]), ms, &(x[WS(rs, 1)])); Td = BYTWJ(&(W[TWVL * 4]), Tc); Te = VSUB(Tb, Td); Tu = VADD(Tb, Td); } { V Ts, Tv, Tw, Tx; Ts = VADD(Tq, Tr); Tv = VADD(Tt, Tu); ST(&(x[WS(rs, 4)]), VSUB(Ts, Tv), ms, &(x[0])); ST(&(x[0]), VADD(Ts, Tv), ms, &(x[0])); Tw = VSUB(Tq, Tr); Tx = VBYI(VSUB(Tu, Tt)); ST(&(x[WS(rs, 6)]), VSUB(Tw, Tx), ms, &(x[0])); ST(&(x[WS(rs, 2)]), VADD(Tw, Tx), ms, &(x[0])); { V Tg, To, Tn, Tp, Tf, Th; Tf = VMUL(LDK(KP707106781), VADD(T9, Te)); Tg = VADD(T4, Tf); To = VSUB(T4, Tf); Th = VMUL(LDK(KP707106781), VSUB(Te, T9)); Tn = VBYI(VSUB(Th, Tm)); Tp = VBYI(VADD(Tm, Th)); ST(&(x[WS(rs, 7)]), VSUB(Tg, Tn), ms, &(x[WS(rs, 1)])); ST(&(x[WS(rs, 3)]), VADD(To, Tp), ms, &(x[WS(rs, 1)])); ST(&(x[WS(rs, 1)]), VADD(Tg, Tn), ms, &(x[WS(rs, 1)])); ST(&(x[WS(rs, 5)]), VSUB(To, Tp), ms, &(x[WS(rs, 1)])); } } } } VLEAVE(); } static const tw_instr twinstr[] = { VTW(0, 1), VTW(0, 2), VTW(0, 3), VTW(0, 4), VTW(0, 5), VTW(0, 6), VTW(0, 7), {TW_NEXT, VL, 0} }; static const ct_desc desc = { 8, XSIMD_STRING("t2fv_8"), twinstr, &GENUS, {33, 16, 0, 0}, 0, 0, 0 }; void XSIMD(codelet_t2fv_8) (planner *p) { X(kdft_dit_register) (p, t2fv_8, &desc); } #endif /* HAVE_FMA */ fftw-3.3.3/dft/simd/common/t2fv_16.c0000644000175000017500000003221512054410514013742 00000000000000/* * Copyright (c) 2003, 2007-11 Matteo Frigo * Copyright (c) 2003, 2007-11 Massachusetts Institute of Technology * * This program is free software; you can redistribute it and/or modify * it under the terms of the GNU General Public License as published by * the Free Software Foundation; either version 2 of the License, or * (at your option) any later version. * * This program is distributed in the hope that it will be useful, * but WITHOUT ANY WARRANTY; without even the implied warranty of * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the * GNU General Public License for more details. * * You should have received a copy of the GNU General Public License * along with this program; if not, write to the Free Software * Foundation, Inc., 51 Franklin Street, Fifth Floor, Boston, MA 02110-1301 USA * */ /* This file was automatically generated --- DO NOT EDIT */ /* Generated on Sun Nov 25 07:38:35 EST 2012 */ #include "codelet-dft.h" #ifdef HAVE_FMA /* Generated by: ../../../genfft/gen_twiddle_c.native -fma -reorder-insns -schedule-for-pipeline -simd -compact -variables 4 -pipeline-latency 8 -n 16 -name t2fv_16 -include t2f.h */ /* * This function contains 87 FP additions, 64 FP multiplications, * (or, 53 additions, 30 multiplications, 34 fused multiply/add), * 61 stack variables, 3 constants, and 32 memory accesses */ #include "t2f.h" static void t2fv_16(R *ri, R *ii, const R *W, stride rs, INT mb, INT me, INT ms) { DVK(KP923879532, +0.923879532511286756128183189396788286822416626); DVK(KP414213562, +0.414213562373095048801688724209698078569671875); DVK(KP707106781, +0.707106781186547524400844362104849039284835938); { INT m; R *x; x = ri; for (m = mb, W = W + (mb * ((TWVL / VL) * 30)); m < me; m = m + VL, x = x + (VL * ms), W = W + (TWVL * 30), MAKE_VOLATILE_STRIDE(16, rs)) { V TO, Ta, TJ, TP, T14, Tq, T1i, T10, T1b, T1l, T13, T1c, TR, Tl, T15; V Tv; { V Tc, TW, T4, T19, T9, TD, TI, Tj, TZ, T1a, Te, Th, Tn, Tr, Tu; V Tp; { V T1, T2, T5, T7; T1 = LD(&(x[0]), ms, &(x[0])); T2 = LD(&(x[WS(rs, 8)]), ms, &(x[0])); T5 = LD(&(x[WS(rs, 4)]), ms, &(x[0])); T7 = LD(&(x[WS(rs, 12)]), ms, &(x[0])); { V Tz, TG, TB, TE; Tz = LD(&(x[WS(rs, 14)]), ms, &(x[0])); TG = LD(&(x[WS(rs, 10)]), ms, &(x[0])); TB = LD(&(x[WS(rs, 6)]), ms, &(x[0])); TE = LD(&(x[WS(rs, 2)]), ms, &(x[0])); { V Ti, TY, TX, Td, Tg, Tm, Tt, To; { V T3, T6, T8, TA, TH, TC, TF, Tb; Tb = LD(&(x[WS(rs, 1)]), ms, &(x[WS(rs, 1)])); T3 = BYTWJ(&(W[TWVL * 14]), T2); T6 = BYTWJ(&(W[TWVL * 6]), T5); T8 = BYTWJ(&(W[TWVL * 22]), T7); TA = BYTWJ(&(W[TWVL * 26]), Tz); TH = BYTWJ(&(W[TWVL * 18]), TG); TC = BYTWJ(&(W[TWVL * 10]), TB); TF = BYTWJ(&(W[TWVL * 2]), TE); Tc = BYTWJ(&(W[0]), Tb); TW = VSUB(T1, T3); T4 = VADD(T1, T3); T19 = VSUB(T6, T8); T9 = VADD(T6, T8); Ti = LD(&(x[WS(rs, 13)]), ms, &(x[WS(rs, 1)])); TD = VADD(TA, TC); TY = VSUB(TA, TC); TI = VADD(TF, TH); TX = VSUB(TF, TH); } Td = LD(&(x[WS(rs, 9)]), ms, &(x[WS(rs, 1)])); Tg = LD(&(x[WS(rs, 5)]), ms, &(x[WS(rs, 1)])); Tm = LD(&(x[WS(rs, 15)]), ms, &(x[WS(rs, 1)])); Tj = BYTWJ(&(W[TWVL * 24]), Ti); Tt = LD(&(x[WS(rs, 11)]), ms, &(x[WS(rs, 1)])); To = LD(&(x[WS(rs, 7)]), ms, &(x[WS(rs, 1)])); TZ = VADD(TX, TY); T1a = VSUB(TY, TX); Te = BYTWJ(&(W[TWVL * 16]), Td); Th = BYTWJ(&(W[TWVL * 8]), Tg); Tn = BYTWJ(&(W[TWVL * 28]), Tm); Tr = LD(&(x[WS(rs, 3)]), ms, &(x[WS(rs, 1)])); Tu = BYTWJ(&(W[TWVL * 20]), Tt); Tp = BYTWJ(&(W[TWVL * 12]), To); } } } { V Tf, T11, Tk, T12, Ts; TO = VADD(T4, T9); Ta = VSUB(T4, T9); TJ = VSUB(TD, TI); TP = VADD(TI, TD); Tf = VADD(Tc, Te); T11 = VSUB(Tc, Te); Tk = VADD(Th, Tj); T12 = VSUB(Th, Tj); Ts = BYTWJ(&(W[TWVL * 4]), Tr); T14 = VSUB(Tn, Tp); Tq = VADD(Tn, Tp); T1i = VFNMS(LDK(KP707106781), TZ, TW); T10 = VFMA(LDK(KP707106781), TZ, TW); T1b = VFNMS(LDK(KP707106781), T1a, T19); T1l = VFMA(LDK(KP707106781), T1a, T19); T13 = VFNMS(LDK(KP414213562), T12, T11); T1c = VFMA(LDK(KP414213562), T11, T12); TR = VADD(Tf, Tk); Tl = VSUB(Tf, Tk); T15 = VSUB(Tu, Ts); Tv = VADD(Ts, Tu); } } { V T1d, T16, TS, Tw, TU, TQ; T1d = VFMA(LDK(KP414213562), T14, T15); T16 = VFNMS(LDK(KP414213562), T15, T14); TS = VADD(Tq, Tv); Tw = VSUB(Tq, Tv); TU = VSUB(TO, TP); TQ = VADD(TO, TP); { V T1e, T1j, T17, T1m; T1e = VSUB(T1c, T1d); T1j = VADD(T1c, T1d); T17 = VADD(T13, T16); T1m = VSUB(T16, T13); { V TV, TT, TK, Tx; TV = VSUB(TS, TR); TT = VADD(TR, TS); TK = VSUB(Tw, Tl); Tx = VADD(Tl, Tw); { V T1h, T1f, T1o, T1k; T1h = VFMA(LDK(KP923879532), T1e, T1b); T1f = VFNMS(LDK(KP923879532), T1e, T1b); T1o = VFMA(LDK(KP923879532), T1j, T1i); T1k = VFNMS(LDK(KP923879532), T1j, T1i); { V T1g, T18, T1p, T1n; T1g = VFMA(LDK(KP923879532), T17, T10); T18 = VFNMS(LDK(KP923879532), T17, T10); T1p = VFMA(LDK(KP923879532), T1m, T1l); T1n = VFNMS(LDK(KP923879532), T1m, T1l); ST(&(x[WS(rs, 12)]), VFNMSI(TV, TU), ms, &(x[0])); ST(&(x[WS(rs, 4)]), VFMAI(TV, TU), ms, &(x[0])); ST(&(x[0]), VADD(TQ, TT), ms, &(x[0])); ST(&(x[WS(rs, 8)]), VSUB(TQ, TT), ms, &(x[0])); { V TN, TL, TM, Ty; TN = VFMA(LDK(KP707106781), TK, TJ); TL = VFNMS(LDK(KP707106781), TK, TJ); TM = VFMA(LDK(KP707106781), Tx, Ta); Ty = VFNMS(LDK(KP707106781), Tx, Ta); ST(&(x[WS(rs, 1)]), VFNMSI(T1h, T1g), ms, &(x[WS(rs, 1)])); ST(&(x[WS(rs, 15)]), VFMAI(T1h, T1g), ms, &(x[WS(rs, 1)])); ST(&(x[WS(rs, 7)]), VFMAI(T1f, T18), ms, &(x[WS(rs, 1)])); ST(&(x[WS(rs, 9)]), VFNMSI(T1f, T18), ms, &(x[WS(rs, 1)])); ST(&(x[WS(rs, 3)]), VFMAI(T1p, T1o), ms, &(x[WS(rs, 1)])); ST(&(x[WS(rs, 13)]), VFNMSI(T1p, T1o), ms, &(x[WS(rs, 1)])); ST(&(x[WS(rs, 11)]), VFMAI(T1n, T1k), ms, &(x[WS(rs, 1)])); ST(&(x[WS(rs, 5)]), VFNMSI(T1n, T1k), ms, &(x[WS(rs, 1)])); ST(&(x[WS(rs, 14)]), VFNMSI(TN, TM), ms, &(x[0])); ST(&(x[WS(rs, 2)]), VFMAI(TN, TM), ms, &(x[0])); ST(&(x[WS(rs, 10)]), VFMAI(TL, Ty), ms, &(x[0])); ST(&(x[WS(rs, 6)]), VFNMSI(TL, Ty), ms, &(x[0])); } } } } } } } } VLEAVE(); } static const tw_instr twinstr[] = { VTW(0, 1), VTW(0, 2), VTW(0, 3), VTW(0, 4), VTW(0, 5), VTW(0, 6), VTW(0, 7), VTW(0, 8), VTW(0, 9), VTW(0, 10), VTW(0, 11), VTW(0, 12), VTW(0, 13), VTW(0, 14), VTW(0, 15), {TW_NEXT, VL, 0} }; static const ct_desc desc = { 16, XSIMD_STRING("t2fv_16"), twinstr, &GENUS, {53, 30, 34, 0}, 0, 0, 0 }; void XSIMD(codelet_t2fv_16) (planner *p) { X(kdft_dit_register) (p, t2fv_16, &desc); } #else /* HAVE_FMA */ /* Generated by: ../../../genfft/gen_twiddle_c.native -simd -compact -variables 4 -pipeline-latency 8 -n 16 -name t2fv_16 -include t2f.h */ /* * This function contains 87 FP additions, 42 FP multiplications, * (or, 83 additions, 38 multiplications, 4 fused multiply/add), * 36 stack variables, 3 constants, and 32 memory accesses */ #include "t2f.h" static void t2fv_16(R *ri, R *ii, const R *W, stride rs, INT mb, INT me, INT ms) { DVK(KP923879532, +0.923879532511286756128183189396788286822416626); DVK(KP382683432, +0.382683432365089771728459984030398866761344562); DVK(KP707106781, +0.707106781186547524400844362104849039284835938); { INT m; R *x; x = ri; for (m = mb, W = W + (mb * ((TWVL / VL) * 30)); m < me; m = m + VL, x = x + (VL * ms), W = W + (TWVL * 30), MAKE_VOLATILE_STRIDE(16, rs)) { V TJ, T10, TD, T11, T1b, T1c, Ty, TK, T16, T17, T18, Tb, TN, T13, T14; V T15, Tm, TM, TG, TI, TH; TG = LD(&(x[0]), ms, &(x[0])); TH = LD(&(x[WS(rs, 8)]), ms, &(x[0])); TI = BYTWJ(&(W[TWVL * 14]), TH); TJ = VSUB(TG, TI); T10 = VADD(TG, TI); { V TA, TC, Tz, TB; Tz = LD(&(x[WS(rs, 4)]), ms, &(x[0])); TA = BYTWJ(&(W[TWVL * 6]), Tz); TB = LD(&(x[WS(rs, 12)]), ms, &(x[0])); TC = BYTWJ(&(W[TWVL * 22]), TB); TD = VSUB(TA, TC); T11 = VADD(TA, TC); } { V Tp, Tw, Tr, Tu, Ts, Tx; { V To, Tv, Tq, Tt; To = LD(&(x[WS(rs, 14)]), ms, &(x[0])); Tp = BYTWJ(&(W[TWVL * 26]), To); Tv = LD(&(x[WS(rs, 10)]), ms, &(x[0])); Tw = BYTWJ(&(W[TWVL * 18]), Tv); Tq = LD(&(x[WS(rs, 6)]), ms, &(x[0])); Tr = BYTWJ(&(W[TWVL * 10]), Tq); Tt = LD(&(x[WS(rs, 2)]), ms, &(x[0])); Tu = BYTWJ(&(W[TWVL * 2]), Tt); } T1b = VADD(Tp, Tr); T1c = VADD(Tu, Tw); Ts = VSUB(Tp, Tr); Tx = VSUB(Tu, Tw); Ty = VMUL(LDK(KP707106781), VSUB(Ts, Tx)); TK = VMUL(LDK(KP707106781), VADD(Tx, Ts)); } { V T2, T9, T4, T7, T5, Ta; { V T1, T8, T3, T6; T1 = LD(&(x[WS(rs, 15)]), ms, &(x[WS(rs, 1)])); T2 = BYTWJ(&(W[TWVL * 28]), T1); T8 = LD(&(x[WS(rs, 11)]), ms, &(x[WS(rs, 1)])); T9 = BYTWJ(&(W[TWVL * 20]), T8); T3 = LD(&(x[WS(rs, 7)]), ms, &(x[WS(rs, 1)])); T4 = BYTWJ(&(W[TWVL * 12]), T3); T6 = LD(&(x[WS(rs, 3)]), ms, &(x[WS(rs, 1)])); T7 = BYTWJ(&(W[TWVL * 4]), T6); } T16 = VADD(T2, T4); T17 = VADD(T7, T9); T18 = VSUB(T16, T17); T5 = VSUB(T2, T4); Ta = VSUB(T7, T9); Tb = VFNMS(LDK(KP923879532), Ta, VMUL(LDK(KP382683432), T5)); TN = VFMA(LDK(KP923879532), T5, VMUL(LDK(KP382683432), Ta)); } { V Td, Tk, Tf, Ti, Tg, Tl; { V Tc, Tj, Te, Th; Tc = LD(&(x[WS(rs, 1)]), ms, &(x[WS(rs, 1)])); Td = BYTWJ(&(W[0]), Tc); Tj = LD(&(x[WS(rs, 13)]), ms, &(x[WS(rs, 1)])); Tk = BYTWJ(&(W[TWVL * 24]), Tj); Te = LD(&(x[WS(rs, 9)]), ms, &(x[WS(rs, 1)])); Tf = BYTWJ(&(W[TWVL * 16]), Te); Th = LD(&(x[WS(rs, 5)]), ms, &(x[WS(rs, 1)])); Ti = BYTWJ(&(W[TWVL * 8]), Th); } T13 = VADD(Td, Tf); T14 = VADD(Ti, Tk); T15 = VSUB(T13, T14); Tg = VSUB(Td, Tf); Tl = VSUB(Ti, Tk); Tm = VFMA(LDK(KP382683432), Tg, VMUL(LDK(KP923879532), Tl)); TM = VFNMS(LDK(KP382683432), Tl, VMUL(LDK(KP923879532), Tg)); } { V T1a, T1g, T1f, T1h; { V T12, T19, T1d, T1e; T12 = VSUB(T10, T11); T19 = VMUL(LDK(KP707106781), VADD(T15, T18)); T1a = VADD(T12, T19); T1g = VSUB(T12, T19); T1d = VSUB(T1b, T1c); T1e = VMUL(LDK(KP707106781), VSUB(T18, T15)); T1f = VBYI(VADD(T1d, T1e)); T1h = VBYI(VSUB(T1e, T1d)); } ST(&(x[WS(rs, 14)]), VSUB(T1a, T1f), ms, &(x[0])); ST(&(x[WS(rs, 6)]), VADD(T1g, T1h), ms, &(x[0])); ST(&(x[WS(rs, 2)]), VADD(T1a, T1f), ms, &(x[0])); ST(&(x[WS(rs, 10)]), VSUB(T1g, T1h), ms, &(x[0])); } { V T1k, T1o, T1n, T1p; { V T1i, T1j, T1l, T1m; T1i = VADD(T10, T11); T1j = VADD(T1c, T1b); T1k = VADD(T1i, T1j); T1o = VSUB(T1i, T1j); T1l = VADD(T13, T14); T1m = VADD(T16, T17); T1n = VADD(T1l, T1m); T1p = VBYI(VSUB(T1m, T1l)); } ST(&(x[WS(rs, 8)]), VSUB(T1k, T1n), ms, &(x[0])); ST(&(x[WS(rs, 4)]), VADD(T1o, T1p), ms, &(x[0])); ST(&(x[0]), VADD(T1k, T1n), ms, &(x[0])); ST(&(x[WS(rs, 12)]), VSUB(T1o, T1p), ms, &(x[0])); } { V TF, TQ, TP, TR; { V Tn, TE, TL, TO; Tn = VSUB(Tb, Tm); TE = VSUB(Ty, TD); TF = VBYI(VSUB(Tn, TE)); TQ = VBYI(VADD(TE, Tn)); TL = VADD(TJ, TK); TO = VADD(TM, TN); TP = VSUB(TL, TO); TR = VADD(TL, TO); } ST(&(x[WS(rs, 7)]), VADD(TF, TP), ms, &(x[WS(rs, 1)])); ST(&(x[WS(rs, 15)]), VSUB(TR, TQ), ms, &(x[WS(rs, 1)])); ST(&(x[WS(rs, 9)]), VSUB(TP, TF), ms, &(x[WS(rs, 1)])); ST(&(x[WS(rs, 1)]), VADD(TQ, TR), ms, &(x[WS(rs, 1)])); } { V TU, TY, TX, TZ; { V TS, TT, TV, TW; TS = VSUB(TJ, TK); TT = VADD(Tm, Tb); TU = VADD(TS, TT); TY = VSUB(TS, TT); TV = VADD(TD, Ty); TW = VSUB(TN, TM); TX = VBYI(VADD(TV, TW)); TZ = VBYI(VSUB(TW, TV)); } ST(&(x[WS(rs, 13)]), VSUB(TU, TX), ms, &(x[WS(rs, 1)])); ST(&(x[WS(rs, 5)]), VADD(TY, TZ), ms, &(x[WS(rs, 1)])); ST(&(x[WS(rs, 3)]), VADD(TU, TX), ms, &(x[WS(rs, 1)])); ST(&(x[WS(rs, 11)]), VSUB(TY, TZ), ms, &(x[WS(rs, 1)])); } } } VLEAVE(); } static const tw_instr twinstr[] = { VTW(0, 1), VTW(0, 2), VTW(0, 3), VTW(0, 4), VTW(0, 5), VTW(0, 6), VTW(0, 7), VTW(0, 8), VTW(0, 9), VTW(0, 10), VTW(0, 11), VTW(0, 12), VTW(0, 13), VTW(0, 14), VTW(0, 15), {TW_NEXT, VL, 0} }; static const ct_desc desc = { 16, XSIMD_STRING("t2fv_16"), twinstr, &GENUS, {83, 38, 4, 0}, 0, 0, 0 }; void XSIMD(codelet_t2fv_16) (planner *p) { X(kdft_dit_register) (p, t2fv_16, &desc); } #endif /* HAVE_FMA */ fftw-3.3.3/dft/simd/common/t2fv_32.c0000644000175000017500000007041312054410521013740 00000000000000/* * Copyright (c) 2003, 2007-11 Matteo Frigo * Copyright (c) 2003, 2007-11 Massachusetts Institute of Technology * * This program is free software; you can redistribute it and/or modify * it under the terms of the GNU General Public License as published by * the Free Software Foundation; either version 2 of the License, or * (at your option) any later version. * * This program is distributed in the hope that it will be useful, * but WITHOUT ANY WARRANTY; without even the implied warranty of * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the * GNU General Public License for more details. * * You should have received a copy of the GNU General Public License * along with this program; if not, write to the Free Software * Foundation, Inc., 51 Franklin Street, Fifth Floor, Boston, MA 02110-1301 USA * */ /* This file was automatically generated --- DO NOT EDIT */ /* Generated on Sun Nov 25 07:38:36 EST 2012 */ #include "codelet-dft.h" #ifdef HAVE_FMA /* Generated by: ../../../genfft/gen_twiddle_c.native -fma -reorder-insns -schedule-for-pipeline -simd -compact -variables 4 -pipeline-latency 8 -n 32 -name t2fv_32 -include t2f.h */ /* * This function contains 217 FP additions, 160 FP multiplications, * (or, 119 additions, 62 multiplications, 98 fused multiply/add), * 112 stack variables, 7 constants, and 64 memory accesses */ #include "t2f.h" static void t2fv_32(R *ri, R *ii, const R *W, stride rs, INT mb, INT me, INT ms) { DVK(KP831469612, +0.831469612302545237078788377617905756738560812); DVK(KP980785280, +0.980785280403230449126182236134239036973933731); DVK(KP668178637, +0.668178637919298919997757686523080761552472251); DVK(KP198912367, +0.198912367379658006911597622644676228597850501); DVK(KP923879532, +0.923879532511286756128183189396788286822416626); DVK(KP707106781, +0.707106781186547524400844362104849039284835938); DVK(KP414213562, +0.414213562373095048801688724209698078569671875); { INT m; R *x; x = ri; for (m = mb, W = W + (mb * ((TWVL / VL) * 62)); m < me; m = m + VL, x = x + (VL * ms), W = W + (TWVL * 62), MAKE_VOLATILE_STRIDE(32, rs)) { V T26, T25, T1Z, T22, T1W, T2a, T2k, T2g; { V T4, T1z, T2o, T32, T2r, T3f, Tf, T1A, T34, T2L, T1D, TC, T33, T2O, T1C; V Tr, T2C, T3a, T2F, T3b, T1r, T21, T1k, T20, TQ, TM, TS, TL, T2t, TJ; V T10, T2u; { V Tt, T9, T2p, Te, T2q, TA, Tu, Tx; { V T1, T1x, T2, T1v; T1 = LD(&(x[0]), ms, &(x[0])); T1x = LD(&(x[WS(rs, 24)]), ms, &(x[0])); T2 = LD(&(x[WS(rs, 16)]), ms, &(x[0])); T1v = LD(&(x[WS(rs, 8)]), ms, &(x[0])); { V T5, Tc, T7, Ta, T2m, T2n; T5 = LD(&(x[WS(rs, 4)]), ms, &(x[0])); Tc = LD(&(x[WS(rs, 12)]), ms, &(x[0])); T7 = LD(&(x[WS(rs, 20)]), ms, &(x[0])); Ta = LD(&(x[WS(rs, 28)]), ms, &(x[0])); { V T1y, T3, T1w, T6, Td, T8, Tb, Ts, Tz; Ts = LD(&(x[WS(rs, 30)]), ms, &(x[0])); T1y = BYTWJ(&(W[TWVL * 46]), T1x); T3 = BYTWJ(&(W[TWVL * 30]), T2); T1w = BYTWJ(&(W[TWVL * 14]), T1v); T6 = BYTWJ(&(W[TWVL * 6]), T5); Td = BYTWJ(&(W[TWVL * 22]), Tc); T8 = BYTWJ(&(W[TWVL * 38]), T7); Tb = BYTWJ(&(W[TWVL * 54]), Ta); Tt = BYTWJ(&(W[TWVL * 58]), Ts); Tz = LD(&(x[WS(rs, 6)]), ms, &(x[0])); T4 = VSUB(T1, T3); T2m = VADD(T1, T3); T1z = VSUB(T1w, T1y); T2n = VADD(T1w, T1y); T9 = VSUB(T6, T8); T2p = VADD(T6, T8); Te = VSUB(Tb, Td); T2q = VADD(Tb, Td); TA = BYTWJ(&(W[TWVL * 10]), Tz); } Tu = LD(&(x[WS(rs, 14)]), ms, &(x[0])); T2o = VADD(T2m, T2n); T32 = VSUB(T2m, T2n); Tx = LD(&(x[WS(rs, 22)]), ms, &(x[0])); } } { V Tv, To, Ty, Ti, Tj, Tm, Th; Th = LD(&(x[WS(rs, 2)]), ms, &(x[0])); T2r = VADD(T2p, T2q); T3f = VSUB(T2q, T2p); Tf = VADD(T9, Te); T1A = VSUB(Te, T9); Tv = BYTWJ(&(W[TWVL * 26]), Tu); To = LD(&(x[WS(rs, 26)]), ms, &(x[0])); Ty = BYTWJ(&(W[TWVL * 42]), Tx); Ti = BYTWJ(&(W[TWVL * 2]), Th); Tj = LD(&(x[WS(rs, 18)]), ms, &(x[0])); Tm = LD(&(x[WS(rs, 10)]), ms, &(x[0])); { V T1f, T1h, T1a, T1c, T18, T2A, T2B, T1p; { V T15, T17, T1o, T1m; { V Tw, T2J, Tp, T2K, TB, Tk, Tn, T1n, T14, T16; T14 = LD(&(x[WS(rs, 31)]), ms, &(x[WS(rs, 1)])); T16 = LD(&(x[WS(rs, 15)]), ms, &(x[WS(rs, 1)])); Tw = VSUB(Tt, Tv); T2J = VADD(Tt, Tv); Tp = BYTWJ(&(W[TWVL * 50]), To); T2K = VADD(TA, Ty); TB = VSUB(Ty, TA); Tk = BYTWJ(&(W[TWVL * 34]), Tj); Tn = BYTWJ(&(W[TWVL * 18]), Tm); T15 = BYTWJ(&(W[TWVL * 60]), T14); T17 = BYTWJ(&(W[TWVL * 28]), T16); T1n = LD(&(x[WS(rs, 7)]), ms, &(x[WS(rs, 1)])); { V T2M, Tl, T2N, Tq, T1l; T1l = LD(&(x[WS(rs, 23)]), ms, &(x[WS(rs, 1)])); T34 = VSUB(T2J, T2K); T2L = VADD(T2J, T2K); T1D = VFMA(LDK(KP414213562), Tw, TB); TC = VFNMS(LDK(KP414213562), TB, Tw); T2M = VADD(Ti, Tk); Tl = VSUB(Ti, Tk); T2N = VADD(Tn, Tp); Tq = VSUB(Tn, Tp); T1o = BYTWJ(&(W[TWVL * 12]), T1n); T1m = BYTWJ(&(W[TWVL * 44]), T1l); { V T1e, T1g, T19, T1b; T1e = LD(&(x[WS(rs, 27)]), ms, &(x[WS(rs, 1)])); T1g = LD(&(x[WS(rs, 11)]), ms, &(x[WS(rs, 1)])); T19 = LD(&(x[WS(rs, 3)]), ms, &(x[WS(rs, 1)])); T1b = LD(&(x[WS(rs, 19)]), ms, &(x[WS(rs, 1)])); T33 = VSUB(T2M, T2N); T2O = VADD(T2M, T2N); T1C = VFMA(LDK(KP414213562), Tl, Tq); Tr = VFNMS(LDK(KP414213562), Tq, Tl); T1f = BYTWJ(&(W[TWVL * 52]), T1e); T1h = BYTWJ(&(W[TWVL * 20]), T1g); T1a = BYTWJ(&(W[TWVL * 4]), T19); T1c = BYTWJ(&(W[TWVL * 36]), T1b); } } } T18 = VSUB(T15, T17); T2A = VADD(T15, T17); T2B = VADD(T1o, T1m); T1p = VSUB(T1m, T1o); } { V TG, TI, TZ, TX; { V T1i, T2E, T1d, T2D, TH, TY, TF; TF = LD(&(x[WS(rs, 1)]), ms, &(x[WS(rs, 1)])); T1i = VSUB(T1f, T1h); T2E = VADD(T1f, T1h); T1d = VSUB(T1a, T1c); T2D = VADD(T1a, T1c); TH = LD(&(x[WS(rs, 17)]), ms, &(x[WS(rs, 1)])); TY = LD(&(x[WS(rs, 25)]), ms, &(x[WS(rs, 1)])); T2C = VADD(T2A, T2B); T3a = VSUB(T2A, T2B); TG = BYTWJ(&(W[0]), TF); { V TW, T1j, T1q, TP, TR, TK; TW = LD(&(x[WS(rs, 9)]), ms, &(x[WS(rs, 1)])); T2F = VADD(T2D, T2E); T3b = VSUB(T2E, T2D); T1j = VADD(T1d, T1i); T1q = VSUB(T1i, T1d); TI = BYTWJ(&(W[TWVL * 32]), TH); TZ = BYTWJ(&(W[TWVL * 48]), TY); TP = LD(&(x[WS(rs, 29)]), ms, &(x[WS(rs, 1)])); TX = BYTWJ(&(W[TWVL * 16]), TW); TR = LD(&(x[WS(rs, 13)]), ms, &(x[WS(rs, 1)])); TK = LD(&(x[WS(rs, 5)]), ms, &(x[WS(rs, 1)])); T1r = VFMA(LDK(KP707106781), T1q, T1p); T21 = VFNMS(LDK(KP707106781), T1q, T1p); T1k = VFMA(LDK(KP707106781), T1j, T18); T20 = VFNMS(LDK(KP707106781), T1j, T18); TQ = BYTWJ(&(W[TWVL * 56]), TP); TM = LD(&(x[WS(rs, 21)]), ms, &(x[WS(rs, 1)])); TS = BYTWJ(&(W[TWVL * 24]), TR); TL = BYTWJ(&(W[TWVL * 8]), TK); } } T2t = VADD(TG, TI); TJ = VSUB(TG, TI); T10 = VSUB(TX, TZ); T2u = VADD(TX, TZ); } } } } { V T2s, TT, T2x, T2P, T2Y, T2G, T37, T2v, T2w, TO, T2W, T30, T2U, TN, T2V; T2s = VSUB(T2o, T2r); T2U = VADD(T2o, T2r); TN = BYTWJ(&(W[TWVL * 40]), TM); TT = VSUB(TQ, TS); T2x = VADD(TQ, TS); T2P = VSUB(T2L, T2O); T2V = VADD(T2O, T2L); T2Y = VADD(T2C, T2F); T2G = VSUB(T2C, T2F); T37 = VSUB(T2t, T2u); T2v = VADD(T2t, T2u); T2w = VADD(TL, TN); TO = VSUB(TL, TN); T2W = VADD(T2U, T2V); T30 = VSUB(T2U, T2V); { V T3i, T3o, T36, T3r, T3h, T3j, T12, T1Y, TV, T1X, T3s, T3d, T2Q, T2H, T31; V T2Z; { V T35, T3g, T38, T2y, T11, TU; T35 = VADD(T33, T34); T3g = VSUB(T34, T33); T38 = VSUB(T2w, T2x); T2y = VADD(T2w, T2x); T11 = VSUB(TO, TT); TU = VADD(TO, TT); { V T3c, T39, T2X, T2z; T3c = VFNMS(LDK(KP414213562), T3b, T3a); T3i = VFMA(LDK(KP414213562), T3a, T3b); T3o = VFNMS(LDK(KP707106781), T35, T32); T36 = VFMA(LDK(KP707106781), T35, T32); T3r = VFNMS(LDK(KP707106781), T3g, T3f); T3h = VFMA(LDK(KP707106781), T3g, T3f); T39 = VFNMS(LDK(KP414213562), T38, T37); T3j = VFMA(LDK(KP414213562), T37, T38); T2X = VADD(T2v, T2y); T2z = VSUB(T2v, T2y); T12 = VFMA(LDK(KP707106781), T11, T10); T1Y = VFNMS(LDK(KP707106781), T11, T10); TV = VFMA(LDK(KP707106781), TU, TJ); T1X = VFNMS(LDK(KP707106781), TU, TJ); T3s = VSUB(T3c, T39); T3d = VADD(T39, T3c); T2Q = VSUB(T2G, T2z); T2H = VADD(T2z, T2G); T31 = VSUB(T2Y, T2X); T2Z = VADD(T2X, T2Y); } } { V Tg, T1U, TD, T1G, T13, T1s, T1H, T1B, T1V, T1E, T3k, T3p, T2e, T2f; Tg = VFMA(LDK(KP707106781), Tf, T4); T1U = VFNMS(LDK(KP707106781), Tf, T4); T3k = VSUB(T3i, T3j); T3p = VADD(T3j, T3i); { V T3v, T3t, T3e, T3m; T3v = VFNMS(LDK(KP923879532), T3s, T3r); T3t = VFMA(LDK(KP923879532), T3s, T3r); T3e = VFNMS(LDK(KP923879532), T3d, T36); T3m = VFMA(LDK(KP923879532), T3d, T36); { V T2R, T2T, T2I, T2S; T2R = VFNMS(LDK(KP707106781), T2Q, T2P); T2T = VFMA(LDK(KP707106781), T2Q, T2P); T2I = VFNMS(LDK(KP707106781), T2H, T2s); T2S = VFMA(LDK(KP707106781), T2H, T2s); ST(&(x[WS(rs, 24)]), VFNMSI(T31, T30), ms, &(x[0])); ST(&(x[WS(rs, 8)]), VFMAI(T31, T30), ms, &(x[0])); ST(&(x[0]), VADD(T2W, T2Z), ms, &(x[0])); ST(&(x[WS(rs, 16)]), VSUB(T2W, T2Z), ms, &(x[0])); { V T3u, T3q, T3l, T3n; T3u = VFMA(LDK(KP923879532), T3p, T3o); T3q = VFNMS(LDK(KP923879532), T3p, T3o); T3l = VFNMS(LDK(KP923879532), T3k, T3h); T3n = VFMA(LDK(KP923879532), T3k, T3h); ST(&(x[WS(rs, 4)]), VFMAI(T2T, T2S), ms, &(x[0])); ST(&(x[WS(rs, 28)]), VFNMSI(T2T, T2S), ms, &(x[0])); ST(&(x[WS(rs, 20)]), VFMAI(T2R, T2I), ms, &(x[0])); ST(&(x[WS(rs, 12)]), VFNMSI(T2R, T2I), ms, &(x[0])); ST(&(x[WS(rs, 22)]), VFNMSI(T3t, T3q), ms, &(x[0])); ST(&(x[WS(rs, 10)]), VFMAI(T3t, T3q), ms, &(x[0])); ST(&(x[WS(rs, 26)]), VFMAI(T3v, T3u), ms, &(x[0])); ST(&(x[WS(rs, 6)]), VFNMSI(T3v, T3u), ms, &(x[0])); ST(&(x[WS(rs, 2)]), VFMAI(T3n, T3m), ms, &(x[0])); ST(&(x[WS(rs, 30)]), VFNMSI(T3n, T3m), ms, &(x[0])); ST(&(x[WS(rs, 18)]), VFMAI(T3l, T3e), ms, &(x[0])); ST(&(x[WS(rs, 14)]), VFNMSI(T3l, T3e), ms, &(x[0])); T26 = VSUB(TC, Tr); TD = VADD(Tr, TC); } } } T1G = VFMA(LDK(KP198912367), TV, T12); T13 = VFNMS(LDK(KP198912367), T12, TV); T1s = VFNMS(LDK(KP198912367), T1r, T1k); T1H = VFMA(LDK(KP198912367), T1k, T1r); T1B = VFNMS(LDK(KP707106781), T1A, T1z); T25 = VFMA(LDK(KP707106781), T1A, T1z); T1V = VADD(T1C, T1D); T1E = VSUB(T1C, T1D); { V T1S, T1O, T1K, T1u, T1R, T1T, T1L, T1J; { V TE, T1M, T1I, T1N, T1t, T1Q, T1F, T1P, T28, T29; TE = VFMA(LDK(KP923879532), TD, Tg); T1M = VFNMS(LDK(KP923879532), TD, Tg); T1I = VSUB(T1G, T1H); T1N = VADD(T1G, T1H); T1t = VADD(T13, T1s); T1Q = VSUB(T1s, T13); T1F = VFMA(LDK(KP923879532), T1E, T1B); T1P = VFNMS(LDK(KP923879532), T1E, T1B); T28 = VFNMS(LDK(KP668178637), T1X, T1Y); T1Z = VFMA(LDK(KP668178637), T1Y, T1X); T1S = VFMA(LDK(KP980785280), T1N, T1M); T1O = VFNMS(LDK(KP980785280), T1N, T1M); T22 = VFMA(LDK(KP668178637), T21, T20); T29 = VFNMS(LDK(KP668178637), T20, T21); T1K = VFMA(LDK(KP980785280), T1t, TE); T1u = VFNMS(LDK(KP980785280), T1t, TE); T1R = VFNMS(LDK(KP980785280), T1Q, T1P); T1T = VFMA(LDK(KP980785280), T1Q, T1P); T1L = VFMA(LDK(KP980785280), T1I, T1F); T1J = VFNMS(LDK(KP980785280), T1I, T1F); T2e = VFNMS(LDK(KP923879532), T1V, T1U); T1W = VFMA(LDK(KP923879532), T1V, T1U); T2a = VSUB(T28, T29); T2f = VADD(T28, T29); } ST(&(x[WS(rs, 23)]), VFMAI(T1R, T1O), ms, &(x[WS(rs, 1)])); ST(&(x[WS(rs, 9)]), VFNMSI(T1R, T1O), ms, &(x[WS(rs, 1)])); ST(&(x[WS(rs, 25)]), VFNMSI(T1T, T1S), ms, &(x[WS(rs, 1)])); ST(&(x[WS(rs, 7)]), VFMAI(T1T, T1S), ms, &(x[WS(rs, 1)])); ST(&(x[WS(rs, 31)]), VFMAI(T1L, T1K), ms, &(x[WS(rs, 1)])); ST(&(x[WS(rs, 1)]), VFNMSI(T1L, T1K), ms, &(x[WS(rs, 1)])); ST(&(x[WS(rs, 15)]), VFMAI(T1J, T1u), ms, &(x[WS(rs, 1)])); ST(&(x[WS(rs, 17)]), VFNMSI(T1J, T1u), ms, &(x[WS(rs, 1)])); } T2k = VFNMS(LDK(KP831469612), T2f, T2e); T2g = VFMA(LDK(KP831469612), T2f, T2e); } } } } { V T2i, T23, T2h, T27; T2i = VSUB(T22, T1Z); T23 = VADD(T1Z, T22); T2h = VFNMS(LDK(KP923879532), T26, T25); T27 = VFMA(LDK(KP923879532), T26, T25); { V T2c, T24, T2j, T2l, T2d, T2b; T2c = VFMA(LDK(KP831469612), T23, T1W); T24 = VFNMS(LDK(KP831469612), T23, T1W); T2j = VFMA(LDK(KP831469612), T2i, T2h); T2l = VFNMS(LDK(KP831469612), T2i, T2h); T2d = VFMA(LDK(KP831469612), T2a, T27); T2b = VFNMS(LDK(KP831469612), T2a, T27); ST(&(x[WS(rs, 21)]), VFNMSI(T2j, T2g), ms, &(x[WS(rs, 1)])); ST(&(x[WS(rs, 11)]), VFMAI(T2j, T2g), ms, &(x[WS(rs, 1)])); ST(&(x[WS(rs, 27)]), VFMAI(T2l, T2k), ms, &(x[WS(rs, 1)])); ST(&(x[WS(rs, 5)]), VFNMSI(T2l, T2k), ms, &(x[WS(rs, 1)])); ST(&(x[WS(rs, 3)]), VFMAI(T2d, T2c), ms, &(x[WS(rs, 1)])); ST(&(x[WS(rs, 29)]), VFNMSI(T2d, T2c), ms, &(x[WS(rs, 1)])); ST(&(x[WS(rs, 19)]), VFMAI(T2b, T24), ms, &(x[WS(rs, 1)])); ST(&(x[WS(rs, 13)]), VFNMSI(T2b, T24), ms, &(x[WS(rs, 1)])); } } } } VLEAVE(); } static const tw_instr twinstr[] = { VTW(0, 1), VTW(0, 2), VTW(0, 3), VTW(0, 4), VTW(0, 5), VTW(0, 6), VTW(0, 7), VTW(0, 8), VTW(0, 9), VTW(0, 10), VTW(0, 11), VTW(0, 12), VTW(0, 13), VTW(0, 14), VTW(0, 15), VTW(0, 16), VTW(0, 17), VTW(0, 18), VTW(0, 19), VTW(0, 20), VTW(0, 21), VTW(0, 22), VTW(0, 23), VTW(0, 24), VTW(0, 25), VTW(0, 26), VTW(0, 27), VTW(0, 28), VTW(0, 29), VTW(0, 30), VTW(0, 31), {TW_NEXT, VL, 0} }; static const ct_desc desc = { 32, XSIMD_STRING("t2fv_32"), twinstr, &GENUS, {119, 62, 98, 0}, 0, 0, 0 }; void XSIMD(codelet_t2fv_32) (planner *p) { X(kdft_dit_register) (p, t2fv_32, &desc); } #else /* HAVE_FMA */ /* Generated by: ../../../genfft/gen_twiddle_c.native -simd -compact -variables 4 -pipeline-latency 8 -n 32 -name t2fv_32 -include t2f.h */ /* * This function contains 217 FP additions, 104 FP multiplications, * (or, 201 additions, 88 multiplications, 16 fused multiply/add), * 59 stack variables, 7 constants, and 64 memory accesses */ #include "t2f.h" static void t2fv_32(R *ri, R *ii, const R *W, stride rs, INT mb, INT me, INT ms) { DVK(KP555570233, +0.555570233019602224742830813948532874374937191); DVK(KP831469612, +0.831469612302545237078788377617905756738560812); DVK(KP195090322, +0.195090322016128267848284868477022240927691618); DVK(KP980785280, +0.980785280403230449126182236134239036973933731); DVK(KP382683432, +0.382683432365089771728459984030398866761344562); DVK(KP923879532, +0.923879532511286756128183189396788286822416626); DVK(KP707106781, +0.707106781186547524400844362104849039284835938); { INT m; R *x; x = ri; for (m = mb, W = W + (mb * ((TWVL / VL) * 62)); m < me; m = m + VL, x = x + (VL * ms), W = W + (TWVL * 62), MAKE_VOLATILE_STRIDE(32, rs)) { V T4, T1A, T2o, T32, Tf, T1v, T2r, T3f, TC, T1C, T2L, T34, Tr, T1D, T2O; V T33, T1k, T20, T2F, T3b, T1r, T21, T2C, T3a, TV, T1X, T2y, T38, T12, T1Y; V T2v, T37; { V T1, T1z, T3, T1x, T1y, T2, T1w, T2m, T2n; T1 = LD(&(x[0]), ms, &(x[0])); T1y = LD(&(x[WS(rs, 24)]), ms, &(x[0])); T1z = BYTWJ(&(W[TWVL * 46]), T1y); T2 = LD(&(x[WS(rs, 16)]), ms, &(x[0])); T3 = BYTWJ(&(W[TWVL * 30]), T2); T1w = LD(&(x[WS(rs, 8)]), ms, &(x[0])); T1x = BYTWJ(&(W[TWVL * 14]), T1w); T4 = VSUB(T1, T3); T1A = VSUB(T1x, T1z); T2m = VADD(T1, T3); T2n = VADD(T1x, T1z); T2o = VADD(T2m, T2n); T32 = VSUB(T2m, T2n); } { V T6, Td, T8, Tb; { V T5, Tc, T7, Ta; T5 = LD(&(x[WS(rs, 4)]), ms, &(x[0])); T6 = BYTWJ(&(W[TWVL * 6]), T5); Tc = LD(&(x[WS(rs, 12)]), ms, &(x[0])); Td = BYTWJ(&(W[TWVL * 22]), Tc); T7 = LD(&(x[WS(rs, 20)]), ms, &(x[0])); T8 = BYTWJ(&(W[TWVL * 38]), T7); Ta = LD(&(x[WS(rs, 28)]), ms, &(x[0])); Tb = BYTWJ(&(W[TWVL * 54]), Ta); } { V T9, Te, T2p, T2q; T9 = VSUB(T6, T8); Te = VSUB(Tb, Td); Tf = VMUL(LDK(KP707106781), VADD(T9, Te)); T1v = VMUL(LDK(KP707106781), VSUB(Te, T9)); T2p = VADD(T6, T8); T2q = VADD(Tb, Td); T2r = VADD(T2p, T2q); T3f = VSUB(T2q, T2p); } } { V Tt, TA, Tv, Ty; { V Ts, Tz, Tu, Tx; Ts = LD(&(x[WS(rs, 30)]), ms, &(x[0])); Tt = BYTWJ(&(W[TWVL * 58]), Ts); Tz = LD(&(x[WS(rs, 22)]), ms, &(x[0])); TA = BYTWJ(&(W[TWVL * 42]), Tz); Tu = LD(&(x[WS(rs, 14)]), ms, &(x[0])); Tv = BYTWJ(&(W[TWVL * 26]), Tu); Tx = LD(&(x[WS(rs, 6)]), ms, &(x[0])); Ty = BYTWJ(&(W[TWVL * 10]), Tx); } { V Tw, TB, T2J, T2K; Tw = VSUB(Tt, Tv); TB = VSUB(Ty, TA); TC = VFMA(LDK(KP923879532), Tw, VMUL(LDK(KP382683432), TB)); T1C = VFNMS(LDK(KP923879532), TB, VMUL(LDK(KP382683432), Tw)); T2J = VADD(Tt, Tv); T2K = VADD(Ty, TA); T2L = VADD(T2J, T2K); T34 = VSUB(T2J, T2K); } } { V Ti, Tp, Tk, Tn; { V Th, To, Tj, Tm; Th = LD(&(x[WS(rs, 2)]), ms, &(x[0])); Ti = BYTWJ(&(W[TWVL * 2]), Th); To = LD(&(x[WS(rs, 26)]), ms, &(x[0])); Tp = BYTWJ(&(W[TWVL * 50]), To); Tj = LD(&(x[WS(rs, 18)]), ms, &(x[0])); Tk = BYTWJ(&(W[TWVL * 34]), Tj); Tm = LD(&(x[WS(rs, 10)]), ms, &(x[0])); Tn = BYTWJ(&(W[TWVL * 18]), Tm); } { V Tl, Tq, T2M, T2N; Tl = VSUB(Ti, Tk); Tq = VSUB(Tn, Tp); Tr = VFNMS(LDK(KP382683432), Tq, VMUL(LDK(KP923879532), Tl)); T1D = VFMA(LDK(KP382683432), Tl, VMUL(LDK(KP923879532), Tq)); T2M = VADD(Ti, Tk); T2N = VADD(Tn, Tp); T2O = VADD(T2M, T2N); T33 = VSUB(T2M, T2N); } } { V T15, T17, T1p, T1n, T1f, T1h, T1i, T1a, T1c, T1d; { V T14, T16, T1o, T1m; T14 = LD(&(x[WS(rs, 31)]), ms, &(x[WS(rs, 1)])); T15 = BYTWJ(&(W[TWVL * 60]), T14); T16 = LD(&(x[WS(rs, 15)]), ms, &(x[WS(rs, 1)])); T17 = BYTWJ(&(W[TWVL * 28]), T16); T1o = LD(&(x[WS(rs, 23)]), ms, &(x[WS(rs, 1)])); T1p = BYTWJ(&(W[TWVL * 44]), T1o); T1m = LD(&(x[WS(rs, 7)]), ms, &(x[WS(rs, 1)])); T1n = BYTWJ(&(W[TWVL * 12]), T1m); { V T1e, T1g, T19, T1b; T1e = LD(&(x[WS(rs, 27)]), ms, &(x[WS(rs, 1)])); T1f = BYTWJ(&(W[TWVL * 52]), T1e); T1g = LD(&(x[WS(rs, 11)]), ms, &(x[WS(rs, 1)])); T1h = BYTWJ(&(W[TWVL * 20]), T1g); T1i = VSUB(T1f, T1h); T19 = LD(&(x[WS(rs, 3)]), ms, &(x[WS(rs, 1)])); T1a = BYTWJ(&(W[TWVL * 4]), T19); T1b = LD(&(x[WS(rs, 19)]), ms, &(x[WS(rs, 1)])); T1c = BYTWJ(&(W[TWVL * 36]), T1b); T1d = VSUB(T1a, T1c); } } { V T18, T1j, T2D, T2E; T18 = VSUB(T15, T17); T1j = VMUL(LDK(KP707106781), VADD(T1d, T1i)); T1k = VADD(T18, T1j); T20 = VSUB(T18, T1j); T2D = VADD(T1a, T1c); T2E = VADD(T1f, T1h); T2F = VADD(T2D, T2E); T3b = VSUB(T2E, T2D); } { V T1l, T1q, T2A, T2B; T1l = VMUL(LDK(KP707106781), VSUB(T1i, T1d)); T1q = VSUB(T1n, T1p); T1r = VSUB(T1l, T1q); T21 = VADD(T1q, T1l); T2A = VADD(T15, T17); T2B = VADD(T1n, T1p); T2C = VADD(T2A, T2B); T3a = VSUB(T2A, T2B); } } { V TG, TI, T10, TY, TQ, TS, TT, TL, TN, TO; { V TF, TH, TZ, TX; TF = LD(&(x[WS(rs, 1)]), ms, &(x[WS(rs, 1)])); TG = BYTWJ(&(W[0]), TF); TH = LD(&(x[WS(rs, 17)]), ms, &(x[WS(rs, 1)])); TI = BYTWJ(&(W[TWVL * 32]), TH); TZ = LD(&(x[WS(rs, 25)]), ms, &(x[WS(rs, 1)])); T10 = BYTWJ(&(W[TWVL * 48]), TZ); TX = LD(&(x[WS(rs, 9)]), ms, &(x[WS(rs, 1)])); TY = BYTWJ(&(W[TWVL * 16]), TX); { V TP, TR, TK, TM; TP = LD(&(x[WS(rs, 29)]), ms, &(x[WS(rs, 1)])); TQ = BYTWJ(&(W[TWVL * 56]), TP); TR = LD(&(x[WS(rs, 13)]), ms, &(x[WS(rs, 1)])); TS = BYTWJ(&(W[TWVL * 24]), TR); TT = VSUB(TQ, TS); TK = LD(&(x[WS(rs, 5)]), ms, &(x[WS(rs, 1)])); TL = BYTWJ(&(W[TWVL * 8]), TK); TM = LD(&(x[WS(rs, 21)]), ms, &(x[WS(rs, 1)])); TN = BYTWJ(&(W[TWVL * 40]), TM); TO = VSUB(TL, TN); } } { V TJ, TU, T2w, T2x; TJ = VSUB(TG, TI); TU = VMUL(LDK(KP707106781), VADD(TO, TT)); TV = VADD(TJ, TU); T1X = VSUB(TJ, TU); T2w = VADD(TL, TN); T2x = VADD(TQ, TS); T2y = VADD(T2w, T2x); T38 = VSUB(T2x, T2w); } { V TW, T11, T2t, T2u; TW = VMUL(LDK(KP707106781), VSUB(TT, TO)); T11 = VSUB(TY, T10); T12 = VSUB(TW, T11); T1Y = VADD(T11, TW); T2t = VADD(TG, TI); T2u = VADD(TY, T10); T2v = VADD(T2t, T2u); T37 = VSUB(T2t, T2u); } } { V T2W, T30, T2Z, T31; { V T2U, T2V, T2X, T2Y; T2U = VADD(T2o, T2r); T2V = VADD(T2O, T2L); T2W = VADD(T2U, T2V); T30 = VSUB(T2U, T2V); T2X = VADD(T2v, T2y); T2Y = VADD(T2C, T2F); T2Z = VADD(T2X, T2Y); T31 = VBYI(VSUB(T2Y, T2X)); } ST(&(x[WS(rs, 16)]), VSUB(T2W, T2Z), ms, &(x[0])); ST(&(x[WS(rs, 8)]), VADD(T30, T31), ms, &(x[0])); ST(&(x[0]), VADD(T2W, T2Z), ms, &(x[0])); ST(&(x[WS(rs, 24)]), VSUB(T30, T31), ms, &(x[0])); } { V T2s, T2P, T2H, T2Q, T2z, T2G; T2s = VSUB(T2o, T2r); T2P = VSUB(T2L, T2O); T2z = VSUB(T2v, T2y); T2G = VSUB(T2C, T2F); T2H = VMUL(LDK(KP707106781), VADD(T2z, T2G)); T2Q = VMUL(LDK(KP707106781), VSUB(T2G, T2z)); { V T2I, T2R, T2S, T2T; T2I = VADD(T2s, T2H); T2R = VBYI(VADD(T2P, T2Q)); ST(&(x[WS(rs, 28)]), VSUB(T2I, T2R), ms, &(x[0])); ST(&(x[WS(rs, 4)]), VADD(T2I, T2R), ms, &(x[0])); T2S = VSUB(T2s, T2H); T2T = VBYI(VSUB(T2Q, T2P)); ST(&(x[WS(rs, 20)]), VSUB(T2S, T2T), ms, &(x[0])); ST(&(x[WS(rs, 12)]), VADD(T2S, T2T), ms, &(x[0])); } } { V T36, T3r, T3h, T3p, T3d, T3o, T3k, T3s, T35, T3g; T35 = VMUL(LDK(KP707106781), VADD(T33, T34)); T36 = VADD(T32, T35); T3r = VSUB(T32, T35); T3g = VMUL(LDK(KP707106781), VSUB(T34, T33)); T3h = VADD(T3f, T3g); T3p = VSUB(T3g, T3f); { V T39, T3c, T3i, T3j; T39 = VFMA(LDK(KP923879532), T37, VMUL(LDK(KP382683432), T38)); T3c = VFNMS(LDK(KP382683432), T3b, VMUL(LDK(KP923879532), T3a)); T3d = VADD(T39, T3c); T3o = VSUB(T3c, T39); T3i = VFNMS(LDK(KP382683432), T37, VMUL(LDK(KP923879532), T38)); T3j = VFMA(LDK(KP382683432), T3a, VMUL(LDK(KP923879532), T3b)); T3k = VADD(T3i, T3j); T3s = VSUB(T3j, T3i); } { V T3e, T3l, T3u, T3v; T3e = VADD(T36, T3d); T3l = VBYI(VADD(T3h, T3k)); ST(&(x[WS(rs, 30)]), VSUB(T3e, T3l), ms, &(x[0])); ST(&(x[WS(rs, 2)]), VADD(T3e, T3l), ms, &(x[0])); T3u = VBYI(VADD(T3p, T3o)); T3v = VADD(T3r, T3s); ST(&(x[WS(rs, 6)]), VADD(T3u, T3v), ms, &(x[0])); ST(&(x[WS(rs, 26)]), VSUB(T3v, T3u), ms, &(x[0])); } { V T3m, T3n, T3q, T3t; T3m = VSUB(T36, T3d); T3n = VBYI(VSUB(T3k, T3h)); ST(&(x[WS(rs, 18)]), VSUB(T3m, T3n), ms, &(x[0])); ST(&(x[WS(rs, 14)]), VADD(T3m, T3n), ms, &(x[0])); T3q = VBYI(VSUB(T3o, T3p)); T3t = VSUB(T3r, T3s); ST(&(x[WS(rs, 10)]), VADD(T3q, T3t), ms, &(x[0])); ST(&(x[WS(rs, 22)]), VSUB(T3t, T3q), ms, &(x[0])); } } { V TE, T1P, T1I, T1Q, T1t, T1M, T1F, T1N; { V Tg, TD, T1G, T1H; Tg = VADD(T4, Tf); TD = VADD(Tr, TC); TE = VADD(Tg, TD); T1P = VSUB(Tg, TD); T1G = VFNMS(LDK(KP195090322), TV, VMUL(LDK(KP980785280), T12)); T1H = VFMA(LDK(KP195090322), T1k, VMUL(LDK(KP980785280), T1r)); T1I = VADD(T1G, T1H); T1Q = VSUB(T1H, T1G); } { V T13, T1s, T1B, T1E; T13 = VFMA(LDK(KP980785280), TV, VMUL(LDK(KP195090322), T12)); T1s = VFNMS(LDK(KP195090322), T1r, VMUL(LDK(KP980785280), T1k)); T1t = VADD(T13, T1s); T1M = VSUB(T1s, T13); T1B = VSUB(T1v, T1A); T1E = VSUB(T1C, T1D); T1F = VADD(T1B, T1E); T1N = VSUB(T1E, T1B); } { V T1u, T1J, T1S, T1T; T1u = VADD(TE, T1t); T1J = VBYI(VADD(T1F, T1I)); ST(&(x[WS(rs, 31)]), VSUB(T1u, T1J), ms, &(x[WS(rs, 1)])); ST(&(x[WS(rs, 1)]), VADD(T1u, T1J), ms, &(x[WS(rs, 1)])); T1S = VBYI(VADD(T1N, T1M)); T1T = VADD(T1P, T1Q); ST(&(x[WS(rs, 7)]), VADD(T1S, T1T), ms, &(x[WS(rs, 1)])); ST(&(x[WS(rs, 25)]), VSUB(T1T, T1S), ms, &(x[WS(rs, 1)])); } { V T1K, T1L, T1O, T1R; T1K = VSUB(TE, T1t); T1L = VBYI(VSUB(T1I, T1F)); ST(&(x[WS(rs, 17)]), VSUB(T1K, T1L), ms, &(x[WS(rs, 1)])); ST(&(x[WS(rs, 15)]), VADD(T1K, T1L), ms, &(x[WS(rs, 1)])); T1O = VBYI(VSUB(T1M, T1N)); T1R = VSUB(T1P, T1Q); ST(&(x[WS(rs, 9)]), VADD(T1O, T1R), ms, &(x[WS(rs, 1)])); ST(&(x[WS(rs, 23)]), VSUB(T1R, T1O), ms, &(x[WS(rs, 1)])); } } { V T1W, T2h, T2a, T2i, T23, T2e, T27, T2f; { V T1U, T1V, T28, T29; T1U = VSUB(T4, Tf); T1V = VADD(T1D, T1C); T1W = VADD(T1U, T1V); T2h = VSUB(T1U, T1V); T28 = VFNMS(LDK(KP555570233), T1X, VMUL(LDK(KP831469612), T1Y)); T29 = VFMA(LDK(KP555570233), T20, VMUL(LDK(KP831469612), T21)); T2a = VADD(T28, T29); T2i = VSUB(T29, T28); } { V T1Z, T22, T25, T26; T1Z = VFMA(LDK(KP831469612), T1X, VMUL(LDK(KP555570233), T1Y)); T22 = VFNMS(LDK(KP555570233), T21, VMUL(LDK(KP831469612), T20)); T23 = VADD(T1Z, T22); T2e = VSUB(T22, T1Z); T25 = VADD(T1A, T1v); T26 = VSUB(TC, Tr); T27 = VADD(T25, T26); T2f = VSUB(T26, T25); } { V T24, T2b, T2k, T2l; T24 = VADD(T1W, T23); T2b = VBYI(VADD(T27, T2a)); ST(&(x[WS(rs, 29)]), VSUB(T24, T2b), ms, &(x[WS(rs, 1)])); ST(&(x[WS(rs, 3)]), VADD(T24, T2b), ms, &(x[WS(rs, 1)])); T2k = VBYI(VADD(T2f, T2e)); T2l = VADD(T2h, T2i); ST(&(x[WS(rs, 5)]), VADD(T2k, T2l), ms, &(x[WS(rs, 1)])); ST(&(x[WS(rs, 27)]), VSUB(T2l, T2k), ms, &(x[WS(rs, 1)])); } { V T2c, T2d, T2g, T2j; T2c = VSUB(T1W, T23); T2d = VBYI(VSUB(T2a, T27)); ST(&(x[WS(rs, 19)]), VSUB(T2c, T2d), ms, &(x[WS(rs, 1)])); ST(&(x[WS(rs, 13)]), VADD(T2c, T2d), ms, &(x[WS(rs, 1)])); T2g = VBYI(VSUB(T2e, T2f)); T2j = VSUB(T2h, T2i); ST(&(x[WS(rs, 11)]), VADD(T2g, T2j), ms, &(x[WS(rs, 1)])); ST(&(x[WS(rs, 21)]), VSUB(T2j, T2g), ms, &(x[WS(rs, 1)])); } } } } VLEAVE(); } static const tw_instr twinstr[] = { VTW(0, 1), VTW(0, 2), VTW(0, 3), VTW(0, 4), VTW(0, 5), VTW(0, 6), VTW(0, 7), VTW(0, 8), VTW(0, 9), VTW(0, 10), VTW(0, 11), VTW(0, 12), VTW(0, 13), VTW(0, 14), VTW(0, 15), VTW(0, 16), VTW(0, 17), VTW(0, 18), VTW(0, 19), VTW(0, 20), VTW(0, 21), VTW(0, 22), VTW(0, 23), VTW(0, 24), VTW(0, 25), VTW(0, 26), VTW(0, 27), VTW(0, 28), VTW(0, 29), VTW(0, 30), VTW(0, 31), {TW_NEXT, VL, 0} }; static const ct_desc desc = { 32, XSIMD_STRING("t2fv_32"), twinstr, &GENUS, {201, 88, 16, 0}, 0, 0, 0 }; void XSIMD(codelet_t2fv_32) (planner *p) { X(kdft_dit_register) (p, t2fv_32, &desc); } #endif /* HAVE_FMA */ fftw-3.3.3/dft/simd/common/t2fv_64.c0000644000175000017500000017600512054410546013760 00000000000000/* * Copyright (c) 2003, 2007-11 Matteo Frigo * Copyright (c) 2003, 2007-11 Massachusetts Institute of Technology * * This program is free software; you can redistribute it and/or modify * it under the terms of the GNU General Public License as published by * the Free Software Foundation; either version 2 of the License, or * (at your option) any later version. * * This program is distributed in the hope that it will be useful, * but WITHOUT ANY WARRANTY; without even the implied warranty of * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the * GNU General Public License for more details. * * You should have received a copy of the GNU General Public License * along with this program; if not, write to the Free Software * Foundation, Inc., 51 Franklin Street, Fifth Floor, Boston, MA 02110-1301 USA * */ /* This file was automatically generated --- DO NOT EDIT */ /* Generated on Sun Nov 25 07:38:40 EST 2012 */ #include "codelet-dft.h" #ifdef HAVE_FMA /* Generated by: ../../../genfft/gen_twiddle_c.native -fma -reorder-insns -schedule-for-pipeline -simd -compact -variables 4 -pipeline-latency 8 -n 64 -name t2fv_64 -include t2f.h */ /* * This function contains 519 FP additions, 384 FP multiplications, * (or, 261 additions, 126 multiplications, 258 fused multiply/add), * 187 stack variables, 15 constants, and 128 memory accesses */ #include "t2f.h" static void t2fv_64(R *ri, R *ii, const R *W, stride rs, INT mb, INT me, INT ms) { DVK(KP773010453, +0.773010453362736960810906609758469800971041293); DVK(KP995184726, +0.995184726672196886244836953109479921575474869); DVK(KP820678790, +0.820678790828660330972281985331011598767386482); DVK(KP098491403, +0.098491403357164253077197521291327432293052451); DVK(KP956940335, +0.956940335732208864935797886980269969482849206); DVK(KP881921264, +0.881921264348355029712756863660388349508442621); DVK(KP303346683, +0.303346683607342391675883946941299872384187453); DVK(KP534511135, +0.534511135950791641089685961295362908582039528); DVK(KP831469612, +0.831469612302545237078788377617905756738560812); DVK(KP980785280, +0.980785280403230449126182236134239036973933731); DVK(KP668178637, +0.668178637919298919997757686523080761552472251); DVK(KP198912367, +0.198912367379658006911597622644676228597850501); DVK(KP923879532, +0.923879532511286756128183189396788286822416626); DVK(KP414213562, +0.414213562373095048801688724209698078569671875); DVK(KP707106781, +0.707106781186547524400844362104849039284835938); { INT m; R *x; x = ri; for (m = mb, W = W + (mb * ((TWVL / VL) * 126)); m < me; m = m + VL, x = x + (VL * ms), W = W + (TWVL * 126), MAKE_VOLATILE_STRIDE(64, rs)) { V T6L, T6M, T6O, T6P, T75, T6V, T5A, T6A, T72, T6K, T6t, T6D, T6w, T6B, T6h; V T6E; { V Ta, T3U, T3V, T37, T7a, T58, T7B, T6l, T1v, T24, T5Q, T7o, T5F, T7l, T43; V T4F, T2i, T2R, T6b, T7v, T60, T7s, T4a, T4I, T5u, T7h, T5x, T7g, T1i, T3a; V T4j, T4C, T7e, T5l, T7d, T5o, T3b, TV, T4B, T4m, T3X, T3Y, T6o, T7b, T5f; V T7C, Tx, T38, T2p, T61, T2n, T65, T2D, T7p, T5M, T7m, T5T, T4G, T46, T25; V T1S, T2q, T2u, T2w; { V T5q, T10, T5v, T15, T1b, T5s, T1c, T1e; { V T1V, T1p, T5B, T5O, T1u, T1X, T20, T21; { V T1, T2, T7, T5, T32, T34, T2X, T2Z; T1 = LD(&(x[0]), ms, &(x[0])); T2 = LD(&(x[WS(rs, 32)]), ms, &(x[0])); T7 = LD(&(x[WS(rs, 48)]), ms, &(x[0])); T5 = LD(&(x[WS(rs, 16)]), ms, &(x[0])); T32 = LD(&(x[WS(rs, 8)]), ms, &(x[0])); T34 = LD(&(x[WS(rs, 40)]), ms, &(x[0])); T2X = LD(&(x[WS(rs, 56)]), ms, &(x[0])); T2Z = LD(&(x[WS(rs, 24)]), ms, &(x[0])); { V T1m, T54, T6j, T36, T55, T31, T56, T1n, T1q, T1s, T4, T9; { V T3, T8, T6, T33, T35, T2Y, T30, T1l; T1l = LD(&(x[WS(rs, 1)]), ms, &(x[WS(rs, 1)])); T3 = BYTWJ(&(W[TWVL * 62]), T2); T8 = BYTWJ(&(W[TWVL * 94]), T7); T6 = BYTWJ(&(W[TWVL * 30]), T5); T33 = BYTWJ(&(W[TWVL * 14]), T32); T35 = BYTWJ(&(W[TWVL * 78]), T34); T2Y = BYTWJ(&(W[TWVL * 110]), T2X); T30 = BYTWJ(&(W[TWVL * 46]), T2Z); T1m = BYTWJ(&(W[0]), T1l); T54 = VSUB(T1, T3); T4 = VADD(T1, T3); T6j = VSUB(T6, T8); T9 = VADD(T6, T8); T36 = VADD(T33, T35); T55 = VSUB(T33, T35); T31 = VADD(T2Y, T30); T56 = VSUB(T2Y, T30); T1n = LD(&(x[WS(rs, 33)]), ms, &(x[WS(rs, 1)])); } T1q = LD(&(x[WS(rs, 17)]), ms, &(x[WS(rs, 1)])); T1s = LD(&(x[WS(rs, 49)]), ms, &(x[WS(rs, 1)])); Ta = VSUB(T4, T9); T3U = VADD(T4, T9); { V T57, T6k, T1o, T1r, T1t, T1W, T1U, T1Z; T1U = LD(&(x[WS(rs, 9)]), ms, &(x[WS(rs, 1)])); T3V = VADD(T36, T31); T37 = VSUB(T31, T36); T57 = VADD(T55, T56); T6k = VSUB(T56, T55); T1o = BYTWJ(&(W[TWVL * 64]), T1n); T1r = BYTWJ(&(W[TWVL * 32]), T1q); T1t = BYTWJ(&(W[TWVL * 96]), T1s); T1V = BYTWJ(&(W[TWVL * 16]), T1U); T1W = LD(&(x[WS(rs, 41)]), ms, &(x[WS(rs, 1)])); T1Z = LD(&(x[WS(rs, 57)]), ms, &(x[WS(rs, 1)])); T7a = VFNMS(LDK(KP707106781), T57, T54); T58 = VFMA(LDK(KP707106781), T57, T54); T7B = VFMA(LDK(KP707106781), T6k, T6j); T6l = VFNMS(LDK(KP707106781), T6k, T6j); T1p = VADD(T1m, T1o); T5B = VSUB(T1m, T1o); T5O = VSUB(T1r, T1t); T1u = VADD(T1r, T1t); T1X = BYTWJ(&(W[TWVL * 80]), T1W); T20 = BYTWJ(&(W[TWVL * 112]), T1Z); T21 = LD(&(x[WS(rs, 25)]), ms, &(x[WS(rs, 1)])); } } } { V T5W, T2N, T69, T2L, T5Y, T2P, T48, T2c, T2h; { V T41, T1Y, T5C, T22, T2d, T29, T2b, T2f, T28, T2a, T2H, T2J; T28 = LD(&(x[WS(rs, 63)]), ms, &(x[WS(rs, 1)])); T2a = LD(&(x[WS(rs, 31)]), ms, &(x[WS(rs, 1)])); T1v = VSUB(T1p, T1u); T41 = VADD(T1p, T1u); T1Y = VADD(T1V, T1X); T5C = VSUB(T1V, T1X); T22 = BYTWJ(&(W[TWVL * 48]), T21); T2d = LD(&(x[WS(rs, 15)]), ms, &(x[WS(rs, 1)])); T29 = BYTWJ(&(W[TWVL * 124]), T28); T2b = BYTWJ(&(W[TWVL * 60]), T2a); T2f = LD(&(x[WS(rs, 47)]), ms, &(x[WS(rs, 1)])); T2H = LD(&(x[WS(rs, 55)]), ms, &(x[WS(rs, 1)])); T2J = LD(&(x[WS(rs, 23)]), ms, &(x[WS(rs, 1)])); { V T23, T5D, T2e, T2g, T2I, T2K, T2M; T2M = LD(&(x[WS(rs, 7)]), ms, &(x[WS(rs, 1)])); T23 = VADD(T20, T22); T5D = VSUB(T20, T22); T2e = BYTWJ(&(W[TWVL * 28]), T2d); T2c = VADD(T29, T2b); T5W = VSUB(T29, T2b); T2g = BYTWJ(&(W[TWVL * 92]), T2f); T2I = BYTWJ(&(W[TWVL * 108]), T2H); T2K = BYTWJ(&(W[TWVL * 44]), T2J); T2N = BYTWJ(&(W[TWVL * 12]), T2M); { V T5E, T5P, T42, T2O; T5E = VADD(T5C, T5D); T5P = VSUB(T5C, T5D); T24 = VSUB(T1Y, T23); T42 = VADD(T1Y, T23); T69 = VSUB(T2g, T2e); T2h = VADD(T2e, T2g); T2O = LD(&(x[WS(rs, 39)]), ms, &(x[WS(rs, 1)])); T2L = VADD(T2I, T2K); T5Y = VSUB(T2I, T2K); T5Q = VFMA(LDK(KP707106781), T5P, T5O); T7o = VFNMS(LDK(KP707106781), T5P, T5O); T5F = VFMA(LDK(KP707106781), T5E, T5B); T7l = VFNMS(LDK(KP707106781), T5E, T5B); T43 = VADD(T41, T42); T4F = VSUB(T41, T42); T2P = BYTWJ(&(W[TWVL * 76]), T2O); } } } T2i = VSUB(T2c, T2h); T48 = VADD(T2c, T2h); { V TW, TY, T11, T2Q, T5X, T13; TW = LD(&(x[WS(rs, 62)]), ms, &(x[0])); TY = LD(&(x[WS(rs, 30)]), ms, &(x[0])); T11 = LD(&(x[WS(rs, 14)]), ms, &(x[0])); T2Q = VADD(T2N, T2P); T5X = VSUB(T2N, T2P); T13 = LD(&(x[WS(rs, 46)]), ms, &(x[0])); { V T12, T5Z, T6a, T49, T14, T18, T1a; { V T17, T19, TX, TZ; T17 = LD(&(x[WS(rs, 54)]), ms, &(x[0])); T19 = LD(&(x[WS(rs, 22)]), ms, &(x[0])); TX = BYTWJ(&(W[TWVL * 122]), TW); TZ = BYTWJ(&(W[TWVL * 58]), TY); T12 = BYTWJ(&(W[TWVL * 26]), T11); T5Z = VADD(T5X, T5Y); T6a = VSUB(T5Y, T5X); T2R = VSUB(T2L, T2Q); T49 = VADD(T2Q, T2L); T14 = BYTWJ(&(W[TWVL * 90]), T13); T18 = BYTWJ(&(W[TWVL * 106]), T17); T5q = VSUB(TX, TZ); T10 = VADD(TX, TZ); T1a = BYTWJ(&(W[TWVL * 42]), T19); } T6b = VFMA(LDK(KP707106781), T6a, T69); T7v = VFNMS(LDK(KP707106781), T6a, T69); T60 = VFMA(LDK(KP707106781), T5Z, T5W); T7s = VFNMS(LDK(KP707106781), T5Z, T5W); T4a = VADD(T48, T49); T4I = VSUB(T48, T49); T5v = VSUB(T14, T12); T15 = VADD(T12, T14); T1b = VADD(T18, T1a); T5s = VSUB(T18, T1a); } T1c = LD(&(x[WS(rs, 6)]), ms, &(x[0])); T1e = LD(&(x[WS(rs, 38)]), ms, &(x[0])); } } } { V Th, T59, Tf, Tv, T5d, Tj, Tm, To; { V T5h, TQ, T5m, T5i, TO, TS, TJ, T4k, TD, TI; { V T4h, T16, TB, T1d, T1f, TE, TG, TA, Tz, TK, TM, TC; Tz = LD(&(x[WS(rs, 2)]), ms, &(x[0])); T4h = VADD(T10, T15); T16 = VSUB(T10, T15); TB = LD(&(x[WS(rs, 34)]), ms, &(x[0])); T1d = BYTWJ(&(W[TWVL * 10]), T1c); T1f = BYTWJ(&(W[TWVL * 74]), T1e); TE = LD(&(x[WS(rs, 18)]), ms, &(x[0])); TG = LD(&(x[WS(rs, 50)]), ms, &(x[0])); TA = BYTWJ(&(W[TWVL * 2]), Tz); TK = LD(&(x[WS(rs, 10)]), ms, &(x[0])); TM = LD(&(x[WS(rs, 42)]), ms, &(x[0])); TC = BYTWJ(&(W[TWVL * 66]), TB); { V T1g, T5r, TF, TH, TL, TN, TP; TP = LD(&(x[WS(rs, 58)]), ms, &(x[0])); T1g = VADD(T1d, T1f); T5r = VSUB(T1d, T1f); TF = BYTWJ(&(W[TWVL * 34]), TE); TH = BYTWJ(&(W[TWVL * 98]), TG); TL = BYTWJ(&(W[TWVL * 18]), TK); TN = BYTWJ(&(W[TWVL * 82]), TM); T5h = VSUB(TA, TC); TD = VADD(TA, TC); TQ = BYTWJ(&(W[TWVL * 114]), TP); { V T5w, T5t, T4i, T1h, TR; T5w = VSUB(T5s, T5r); T5t = VADD(T5r, T5s); T4i = VADD(T1g, T1b); T1h = VSUB(T1b, T1g); T5m = VSUB(TF, TH); TI = VADD(TF, TH); T5i = VSUB(TL, TN); TO = VADD(TL, TN); TR = LD(&(x[WS(rs, 26)]), ms, &(x[0])); T5u = VFMA(LDK(KP707106781), T5t, T5q); T7h = VFNMS(LDK(KP707106781), T5t, T5q); T5x = VFMA(LDK(KP707106781), T5w, T5v); T7g = VFNMS(LDK(KP707106781), T5w, T5v); T1i = VFNMS(LDK(KP414213562), T1h, T16); T3a = VFMA(LDK(KP414213562), T16, T1h); T4j = VADD(T4h, T4i); T4C = VSUB(T4h, T4i); TS = BYTWJ(&(W[TWVL * 50]), TR); } } } TJ = VSUB(TD, TI); T4k = VADD(TD, TI); { V Tb, Td, Tr, T5j, TT, Tt, Tg; Tb = LD(&(x[WS(rs, 4)]), ms, &(x[0])); Td = LD(&(x[WS(rs, 36)]), ms, &(x[0])); Tr = LD(&(x[WS(rs, 12)]), ms, &(x[0])); T5j = VSUB(TQ, TS); TT = VADD(TQ, TS); Tt = LD(&(x[WS(rs, 44)]), ms, &(x[0])); Tg = LD(&(x[WS(rs, 20)]), ms, &(x[0])); { V Ti, Tc, Te, Ts; Ti = LD(&(x[WS(rs, 52)]), ms, &(x[0])); Tc = BYTWJ(&(W[TWVL * 6]), Tb); Te = BYTWJ(&(W[TWVL * 70]), Td); Ts = BYTWJ(&(W[TWVL * 22]), Tr); { V T5k, T5n, TU, T4l, Tu; T5k = VADD(T5i, T5j); T5n = VSUB(T5i, T5j); TU = VSUB(TO, TT); T4l = VADD(TO, TT); Tu = BYTWJ(&(W[TWVL * 86]), Tt); Th = BYTWJ(&(W[TWVL * 38]), Tg); T59 = VSUB(Tc, Te); Tf = VADD(Tc, Te); T7e = VFNMS(LDK(KP707106781), T5k, T5h); T5l = VFMA(LDK(KP707106781), T5k, T5h); T7d = VFNMS(LDK(KP707106781), T5n, T5m); T5o = VFMA(LDK(KP707106781), T5n, T5m); T3b = VFMA(LDK(KP414213562), TJ, TU); TV = VFNMS(LDK(KP414213562), TU, TJ); T4B = VSUB(T4k, T4l); T4m = VADD(T4k, T4l); Tv = VADD(Ts, Tu); T5d = VSUB(Tu, Ts); Tj = BYTWJ(&(W[TWVL * 102]), Ti); } } Tm = LD(&(x[WS(rs, 60)]), ms, &(x[0])); To = LD(&(x[WS(rs, 28)]), ms, &(x[0])); } } { V T5b, T6m, Tl, T1A, T5G, T1Q, T5K, T1C, T1D, T5e, T6n, Tw, T1H, T1J; { V T1w, T1y, T1M, T1O, Tq, T5c, T1B; T1w = LD(&(x[WS(rs, 5)]), ms, &(x[WS(rs, 1)])); T1y = LD(&(x[WS(rs, 37)]), ms, &(x[WS(rs, 1)])); T1M = LD(&(x[WS(rs, 13)]), ms, &(x[WS(rs, 1)])); T1O = LD(&(x[WS(rs, 45)]), ms, &(x[WS(rs, 1)])); T1B = LD(&(x[WS(rs, 21)]), ms, &(x[WS(rs, 1)])); { V Tk, T5a, Tn, Tp; Tk = VADD(Th, Tj); T5a = VSUB(Th, Tj); Tn = BYTWJ(&(W[TWVL * 118]), Tm); Tp = BYTWJ(&(W[TWVL * 54]), To); { V T1x, T1z, T1N, T1P; T1x = BYTWJ(&(W[TWVL * 8]), T1w); T1z = BYTWJ(&(W[TWVL * 72]), T1y); T1N = BYTWJ(&(W[TWVL * 24]), T1M); T1P = BYTWJ(&(W[TWVL * 88]), T1O); T5b = VFNMS(LDK(KP414213562), T5a, T59); T6m = VFMA(LDK(KP414213562), T59, T5a); T3X = VADD(Tf, Tk); Tl = VSUB(Tf, Tk); Tq = VADD(Tn, Tp); T5c = VSUB(Tn, Tp); T1A = VADD(T1x, T1z); T5G = VSUB(T1x, T1z); T1Q = VADD(T1N, T1P); T5K = VSUB(T1N, T1P); T1C = BYTWJ(&(W[TWVL * 40]), T1B); } } T1D = LD(&(x[WS(rs, 53)]), ms, &(x[WS(rs, 1)])); T5e = VFNMS(LDK(KP414213562), T5d, T5c); T6n = VFMA(LDK(KP414213562), T5c, T5d); T3Y = VADD(Tq, Tv); Tw = VSUB(Tq, Tv); T1H = LD(&(x[WS(rs, 61)]), ms, &(x[WS(rs, 1)])); T1J = LD(&(x[WS(rs, 29)]), ms, &(x[WS(rs, 1)])); } { V T1I, T1K, T1F, T5H, T2k, T2l, T2z, T2B, T2j, T1E; T2j = LD(&(x[WS(rs, 3)]), ms, &(x[WS(rs, 1)])); T1E = BYTWJ(&(W[TWVL * 104]), T1D); T6o = VSUB(T6m, T6n); T7b = VADD(T6m, T6n); T5f = VADD(T5b, T5e); T7C = VSUB(T5e, T5b); Tx = VADD(Tl, Tw); T38 = VSUB(Tw, Tl); T1I = BYTWJ(&(W[TWVL * 120]), T1H); T1K = BYTWJ(&(W[TWVL * 56]), T1J); T1F = VADD(T1C, T1E); T5H = VSUB(T1C, T1E); T2k = BYTWJ(&(W[TWVL * 4]), T2j); T2l = LD(&(x[WS(rs, 35)]), ms, &(x[WS(rs, 1)])); T2z = LD(&(x[WS(rs, 11)]), ms, &(x[WS(rs, 1)])); T2B = LD(&(x[WS(rs, 43)]), ms, &(x[WS(rs, 1)])); { V T5I, T5R, T44, T1G, T2m, T2A, T2C, T5S, T5L, T1R, T45, T2o, T5J, T1L; T2o = LD(&(x[WS(rs, 19)]), ms, &(x[WS(rs, 1)])); T5J = VSUB(T1I, T1K); T1L = VADD(T1I, T1K); T5I = VFNMS(LDK(KP414213562), T5H, T5G); T5R = VFMA(LDK(KP414213562), T5G, T5H); T44 = VADD(T1A, T1F); T1G = VSUB(T1A, T1F); T2m = BYTWJ(&(W[TWVL * 68]), T2l); T2A = BYTWJ(&(W[TWVL * 20]), T2z); T2C = BYTWJ(&(W[TWVL * 84]), T2B); T5S = VFNMS(LDK(KP414213562), T5J, T5K); T5L = VFMA(LDK(KP414213562), T5K, T5J); T1R = VSUB(T1L, T1Q); T45 = VADD(T1L, T1Q); T2p = BYTWJ(&(W[TWVL * 36]), T2o); T61 = VSUB(T2k, T2m); T2n = VADD(T2k, T2m); T65 = VSUB(T2C, T2A); T2D = VADD(T2A, T2C); T7p = VSUB(T5I, T5L); T5M = VADD(T5I, T5L); T7m = VSUB(T5R, T5S); T5T = VADD(T5R, T5S); T4G = VSUB(T44, T45); T46 = VADD(T44, T45); T25 = VSUB(T1G, T1R); T1S = VADD(T1G, T1R); T2q = LD(&(x[WS(rs, 51)]), ms, &(x[WS(rs, 1)])); } T2u = LD(&(x[WS(rs, 59)]), ms, &(x[WS(rs, 1)])); T2w = LD(&(x[WS(rs, 27)]), ms, &(x[WS(rs, 1)])); } } } } { V T67, T7w, T6e, T7t, T3s, T3E, T39, T3D, T1k, T3k, T3t, T3c, T1T, T3v, T3w; V T26, T2G, T3y, T3z, T2T; { V T4A, T4N, T47, T4v, T2r, T2v, T2x, T4s, T40, T3W, T3Z; T4A = VSUB(T3U, T3V); T3W = VADD(T3U, T3V); T3Z = VADD(T3X, T3Y); T4N = VSUB(T3Y, T3X); T47 = VSUB(T43, T46); T4v = VADD(T43, T46); T2r = BYTWJ(&(W[TWVL * 100]), T2q); T2v = BYTWJ(&(W[TWVL * 116]), T2u); T2x = BYTWJ(&(W[TWVL * 52]), T2w); T4s = VADD(T3W, T3Z); T40 = VSUB(T3W, T3Z); { V T4O, T4n, T4R, T4H, T4E, T4W, T4u, T4y, T4d, T4J, T2F, T2S; { V T6c, T63, T2t, T4b, T6d, T66, T2E, T4c; { V T4D, T62, T2s, T64, T2y, T4t; T4O = VSUB(T4C, T4B); T4D = VADD(T4B, T4C); T62 = VSUB(T2r, T2p); T2s = VADD(T2p, T2r); T64 = VSUB(T2v, T2x); T2y = VADD(T2v, T2x); T4t = VADD(T4m, T4j); T4n = VSUB(T4j, T4m); T4R = VFMA(LDK(KP414213562), T4F, T4G); T4H = VFNMS(LDK(KP414213562), T4G, T4F); T4E = VFMA(LDK(KP707106781), T4D, T4A); T4W = VFNMS(LDK(KP707106781), T4D, T4A); T6c = VFNMS(LDK(KP414213562), T61, T62); T63 = VFMA(LDK(KP414213562), T62, T61); T2t = VSUB(T2n, T2s); T4b = VADD(T2n, T2s); T6d = VFMA(LDK(KP414213562), T64, T65); T66 = VFNMS(LDK(KP414213562), T65, T64); T2E = VSUB(T2y, T2D); T4c = VADD(T2y, T2D); T4u = VADD(T4s, T4t); T4y = VSUB(T4s, T4t); } T67 = VADD(T63, T66); T7w = VSUB(T66, T63); T6e = VADD(T6c, T6d); T7t = VSUB(T6d, T6c); T4d = VADD(T4b, T4c); T4J = VSUB(T4c, T4b); T2F = VADD(T2t, T2E); T2S = VSUB(T2E, T2t); } { V Ty, T1j, T4Q, T4K; Ty = VFMA(LDK(KP707106781), Tx, Ta); T3s = VFNMS(LDK(KP707106781), Tx, Ta); T3E = VSUB(T1i, TV); T1j = VADD(TV, T1i); T39 = VFMA(LDK(KP707106781), T38, T37); T3D = VFNMS(LDK(KP707106781), T38, T37); T4Q = VFMA(LDK(KP414213562), T4I, T4J); T4K = VFNMS(LDK(KP414213562), T4J, T4I); { V T4w, T4e, T4P, T4Z; T4w = VADD(T4a, T4d); T4e = VSUB(T4a, T4d); T4P = VFMA(LDK(KP707106781), T4O, T4N); T4Z = VFNMS(LDK(KP707106781), T4O, T4N); T1k = VFMA(LDK(KP923879532), T1j, Ty); T3k = VFNMS(LDK(KP923879532), T1j, Ty); { V T4L, T50, T4S, T4X; T4L = VADD(T4H, T4K); T50 = VSUB(T4K, T4H); T4S = VSUB(T4Q, T4R); T4X = VADD(T4R, T4Q); { V T4f, T4o, T4x, T4z; T4f = VADD(T47, T4e); T4o = VSUB(T4e, T47); T4x = VADD(T4v, T4w); T4z = VSUB(T4w, T4v); { V T53, T51, T4M, T4U; T53 = VFNMS(LDK(KP923879532), T50, T4Z); T51 = VFMA(LDK(KP923879532), T50, T4Z); T4M = VFNMS(LDK(KP923879532), T4L, T4E); T4U = VFMA(LDK(KP923879532), T4L, T4E); { V T52, T4Y, T4T, T4V; T52 = VFMA(LDK(KP923879532), T4X, T4W); T4Y = VFNMS(LDK(KP923879532), T4X, T4W); T4T = VFNMS(LDK(KP923879532), T4S, T4P); T4V = VFMA(LDK(KP923879532), T4S, T4P); { V T4p, T4r, T4g, T4q; T4p = VFNMS(LDK(KP707106781), T4o, T4n); T4r = VFMA(LDK(KP707106781), T4o, T4n); T4g = VFNMS(LDK(KP707106781), T4f, T40); T4q = VFMA(LDK(KP707106781), T4f, T40); ST(&(x[WS(rs, 16)]), VFMAI(T4z, T4y), ms, &(x[0])); ST(&(x[WS(rs, 48)]), VFNMSI(T4z, T4y), ms, &(x[0])); ST(&(x[0]), VADD(T4u, T4x), ms, &(x[0])); ST(&(x[WS(rs, 32)]), VSUB(T4u, T4x), ms, &(x[0])); ST(&(x[WS(rs, 44)]), VFNMSI(T51, T4Y), ms, &(x[0])); ST(&(x[WS(rs, 20)]), VFMAI(T51, T4Y), ms, &(x[0])); ST(&(x[WS(rs, 52)]), VFMAI(T53, T52), ms, &(x[0])); ST(&(x[WS(rs, 12)]), VFNMSI(T53, T52), ms, &(x[0])); ST(&(x[WS(rs, 4)]), VFMAI(T4V, T4U), ms, &(x[0])); ST(&(x[WS(rs, 60)]), VFNMSI(T4V, T4U), ms, &(x[0])); ST(&(x[WS(rs, 36)]), VFMAI(T4T, T4M), ms, &(x[0])); ST(&(x[WS(rs, 28)]), VFNMSI(T4T, T4M), ms, &(x[0])); ST(&(x[WS(rs, 8)]), VFMAI(T4r, T4q), ms, &(x[0])); ST(&(x[WS(rs, 56)]), VFNMSI(T4r, T4q), ms, &(x[0])); ST(&(x[WS(rs, 40)]), VFMAI(T4p, T4g), ms, &(x[0])); ST(&(x[WS(rs, 24)]), VFNMSI(T4p, T4g), ms, &(x[0])); T3t = VADD(T3b, T3a); T3c = VSUB(T3a, T3b); } } } } } } T1T = VFMA(LDK(KP707106781), T1S, T1v); T3v = VFNMS(LDK(KP707106781), T1S, T1v); T3w = VFNMS(LDK(KP707106781), T25, T24); T26 = VFMA(LDK(KP707106781), T25, T24); T2G = VFMA(LDK(KP707106781), T2F, T2i); T3y = VFNMS(LDK(KP707106781), T2F, T2i); T3z = VFNMS(LDK(KP707106781), T2S, T2R); T2T = VFMA(LDK(KP707106781), T2S, T2R); } } } { V T3u, T3M, T3F, T3P, T3x, T3H, T3q, T3m, T3h, T3j, T3r, T3p, T2W, T3i; { V T3d, T3n, T27, T3f, T2U, T3e; T3d = VFMA(LDK(KP923879532), T3c, T39); T3n = VFNMS(LDK(KP923879532), T3c, T39); T27 = VFNMS(LDK(KP198912367), T26, T1T); T3f = VFMA(LDK(KP198912367), T1T, T26); T2U = VFNMS(LDK(KP198912367), T2T, T2G); T3e = VFMA(LDK(KP198912367), T2G, T2T); T3u = VFMA(LDK(KP923879532), T3t, T3s); T3M = VFNMS(LDK(KP923879532), T3t, T3s); { V T3g, T3l, T2V, T3o; T3g = VSUB(T3e, T3f); T3l = VADD(T3f, T3e); T2V = VADD(T27, T2U); T3o = VSUB(T2U, T27); T3F = VFNMS(LDK(KP923879532), T3E, T3D); T3P = VFMA(LDK(KP923879532), T3E, T3D); T3x = VFMA(LDK(KP668178637), T3w, T3v); T3H = VFNMS(LDK(KP668178637), T3v, T3w); T3q = VFMA(LDK(KP980785280), T3l, T3k); T3m = VFNMS(LDK(KP980785280), T3l, T3k); T3h = VFNMS(LDK(KP980785280), T3g, T3d); T3j = VFMA(LDK(KP980785280), T3g, T3d); T3r = VFNMS(LDK(KP980785280), T3o, T3n); T3p = VFMA(LDK(KP980785280), T3o, T3n); T2W = VFNMS(LDK(KP980785280), T2V, T1k); T3i = VFMA(LDK(KP980785280), T2V, T1k); } } { V T7n, T7Z, T8j, T89, T7k, T7O, T8g, T7Y, T7H, T7R, T80, T7q, T7u, T82, T83; V T7x; { V T7c, T7W, T7D, T87, T7f, T7F, T3A, T3G, T7E, T7i; T7c = VFNMS(LDK(KP923879532), T7b, T7a); T7W = VFMA(LDK(KP923879532), T7b, T7a); T7D = VFNMS(LDK(KP923879532), T7C, T7B); T87 = VFMA(LDK(KP923879532), T7C, T7B); T7f = VFNMS(LDK(KP668178637), T7e, T7d); T7F = VFMA(LDK(KP668178637), T7d, T7e); ST(&(x[WS(rs, 46)]), VFNMSI(T3p, T3m), ms, &(x[0])); ST(&(x[WS(rs, 18)]), VFMAI(T3p, T3m), ms, &(x[0])); ST(&(x[WS(rs, 50)]), VFMAI(T3r, T3q), ms, &(x[0])); ST(&(x[WS(rs, 14)]), VFNMSI(T3r, T3q), ms, &(x[0])); ST(&(x[WS(rs, 2)]), VFMAI(T3j, T3i), ms, &(x[0])); ST(&(x[WS(rs, 62)]), VFNMSI(T3j, T3i), ms, &(x[0])); ST(&(x[WS(rs, 34)]), VFMAI(T3h, T2W), ms, &(x[0])); ST(&(x[WS(rs, 30)]), VFNMSI(T3h, T2W), ms, &(x[0])); T3A = VFMA(LDK(KP668178637), T3z, T3y); T3G = VFNMS(LDK(KP668178637), T3y, T3z); T7E = VFMA(LDK(KP668178637), T7g, T7h); T7i = VFNMS(LDK(KP668178637), T7h, T7g); T7n = VFNMS(LDK(KP923879532), T7m, T7l); T7Z = VFMA(LDK(KP923879532), T7m, T7l); { V T3I, T3N, T3B, T3Q; T3I = VSUB(T3G, T3H); T3N = VADD(T3H, T3G); T3B = VADD(T3x, T3A); T3Q = VSUB(T3A, T3x); { V T7j, T88, T7G, T7X; T7j = VADD(T7f, T7i); T88 = VSUB(T7f, T7i); T7G = VSUB(T7E, T7F); T7X = VADD(T7F, T7E); { V T3S, T3O, T3J, T3L; T3S = VFNMS(LDK(KP831469612), T3N, T3M); T3O = VFMA(LDK(KP831469612), T3N, T3M); T3J = VFNMS(LDK(KP831469612), T3I, T3F); T3L = VFMA(LDK(KP831469612), T3I, T3F); { V T3T, T3R, T3C, T3K; T3T = VFMA(LDK(KP831469612), T3Q, T3P); T3R = VFNMS(LDK(KP831469612), T3Q, T3P); T3C = VFNMS(LDK(KP831469612), T3B, T3u); T3K = VFMA(LDK(KP831469612), T3B, T3u); T8j = VFNMS(LDK(KP831469612), T88, T87); T89 = VFMA(LDK(KP831469612), T88, T87); T7k = VFNMS(LDK(KP831469612), T7j, T7c); T7O = VFMA(LDK(KP831469612), T7j, T7c); T8g = VFNMS(LDK(KP831469612), T7X, T7W); T7Y = VFMA(LDK(KP831469612), T7X, T7W); T7H = VFNMS(LDK(KP831469612), T7G, T7D); T7R = VFMA(LDK(KP831469612), T7G, T7D); ST(&(x[WS(rs, 42)]), VFMAI(T3R, T3O), ms, &(x[0])); ST(&(x[WS(rs, 22)]), VFNMSI(T3R, T3O), ms, &(x[0])); ST(&(x[WS(rs, 54)]), VFNMSI(T3T, T3S), ms, &(x[0])); ST(&(x[WS(rs, 10)]), VFMAI(T3T, T3S), ms, &(x[0])); ST(&(x[WS(rs, 58)]), VFMAI(T3L, T3K), ms, &(x[0])); ST(&(x[WS(rs, 6)]), VFNMSI(T3L, T3K), ms, &(x[0])); ST(&(x[WS(rs, 26)]), VFMAI(T3J, T3C), ms, &(x[0])); ST(&(x[WS(rs, 38)]), VFNMSI(T3J, T3C), ms, &(x[0])); T80 = VFNMS(LDK(KP923879532), T7p, T7o); T7q = VFMA(LDK(KP923879532), T7p, T7o); } } } } T7u = VFNMS(LDK(KP923879532), T7t, T7s); T82 = VFMA(LDK(KP923879532), T7t, T7s); T83 = VFNMS(LDK(KP923879532), T7w, T7v); T7x = VFMA(LDK(KP923879532), T7w, T7v); } { V T5g, T6I, T6p, T6T, T5p, T6q, T6r, T5y; T5g = VFMA(LDK(KP923879532), T5f, T58); T6I = VFNMS(LDK(KP923879532), T5f, T58); { V T7r, T7I, T7y, T7J; T7r = VFNMS(LDK(KP534511135), T7q, T7n); T7I = VFMA(LDK(KP534511135), T7n, T7q); T7y = VFNMS(LDK(KP534511135), T7x, T7u); T7J = VFMA(LDK(KP534511135), T7u, T7x); { V T81, T8a, T84, T8b; T81 = VFMA(LDK(KP303346683), T80, T7Z); T8a = VFNMS(LDK(KP303346683), T7Z, T80); T84 = VFMA(LDK(KP303346683), T83, T82); T8b = VFNMS(LDK(KP303346683), T82, T83); T6p = VFMA(LDK(KP923879532), T6o, T6l); T6T = VFNMS(LDK(KP923879532), T6o, T6l); T5p = VFNMS(LDK(KP198912367), T5o, T5l); T6q = VFMA(LDK(KP198912367), T5l, T5o); { V T7K, T7P, T7z, T7S; T7K = VSUB(T7I, T7J); T7P = VADD(T7I, T7J); T7z = VADD(T7r, T7y); T7S = VSUB(T7y, T7r); { V T8c, T8h, T85, T8k; T8c = VSUB(T8a, T8b); T8h = VADD(T8a, T8b); T85 = VADD(T81, T84); T8k = VSUB(T84, T81); { V T7Q, T7U, T7L, T7N; T7Q = VFNMS(LDK(KP881921264), T7P, T7O); T7U = VFMA(LDK(KP881921264), T7P, T7O); T7L = VFNMS(LDK(KP881921264), T7K, T7H); T7N = VFMA(LDK(KP881921264), T7K, T7H); { V T7T, T7V, T7A, T7M; T7T = VFNMS(LDK(KP881921264), T7S, T7R); T7V = VFMA(LDK(KP881921264), T7S, T7R); T7A = VFNMS(LDK(KP881921264), T7z, T7k); T7M = VFMA(LDK(KP881921264), T7z, T7k); { V T8i, T8m, T8d, T8f; T8i = VFMA(LDK(KP956940335), T8h, T8g); T8m = VFNMS(LDK(KP956940335), T8h, T8g); T8d = VFNMS(LDK(KP956940335), T8c, T89); T8f = VFMA(LDK(KP956940335), T8c, T89); { V T8l, T8n, T86, T8e; T8l = VFMA(LDK(KP956940335), T8k, T8j); T8n = VFNMS(LDK(KP956940335), T8k, T8j); T86 = VFNMS(LDK(KP956940335), T85, T7Y); T8e = VFMA(LDK(KP956940335), T85, T7Y); ST(&(x[WS(rs, 53)]), VFNMSI(T7V, T7U), ms, &(x[WS(rs, 1)])); ST(&(x[WS(rs, 11)]), VFMAI(T7V, T7U), ms, &(x[WS(rs, 1)])); ST(&(x[WS(rs, 43)]), VFMAI(T7T, T7Q), ms, &(x[WS(rs, 1)])); ST(&(x[WS(rs, 21)]), VFNMSI(T7T, T7Q), ms, &(x[WS(rs, 1)])); ST(&(x[WS(rs, 59)]), VFMAI(T7N, T7M), ms, &(x[WS(rs, 1)])); ST(&(x[WS(rs, 5)]), VFNMSI(T7N, T7M), ms, &(x[WS(rs, 1)])); ST(&(x[WS(rs, 27)]), VFMAI(T7L, T7A), ms, &(x[WS(rs, 1)])); ST(&(x[WS(rs, 37)]), VFNMSI(T7L, T7A), ms, &(x[WS(rs, 1)])); ST(&(x[WS(rs, 51)]), VFMAI(T8n, T8m), ms, &(x[WS(rs, 1)])); ST(&(x[WS(rs, 13)]), VFNMSI(T8n, T8m), ms, &(x[WS(rs, 1)])); ST(&(x[WS(rs, 45)]), VFNMSI(T8l, T8i), ms, &(x[WS(rs, 1)])); ST(&(x[WS(rs, 19)]), VFMAI(T8l, T8i), ms, &(x[WS(rs, 1)])); ST(&(x[WS(rs, 3)]), VFMAI(T8f, T8e), ms, &(x[WS(rs, 1)])); ST(&(x[WS(rs, 61)]), VFNMSI(T8f, T8e), ms, &(x[WS(rs, 1)])); ST(&(x[WS(rs, 35)]), VFMAI(T8d, T86), ms, &(x[WS(rs, 1)])); ST(&(x[WS(rs, 29)]), VFNMSI(T8d, T86), ms, &(x[WS(rs, 1)])); T6r = VFMA(LDK(KP198912367), T5u, T5x); T5y = VFNMS(LDK(KP198912367), T5x, T5u); } } } } } } } } { V T5N, T5U, T68, T5z, T6U, T6f; T5N = VFMA(LDK(KP923879532), T5M, T5F); T6L = VFNMS(LDK(KP923879532), T5M, T5F); T6M = VFNMS(LDK(KP923879532), T5T, T5Q); T5U = VFMA(LDK(KP923879532), T5T, T5Q); T68 = VFMA(LDK(KP923879532), T67, T60); T6O = VFNMS(LDK(KP923879532), T67, T60); T5z = VADD(T5p, T5y); T6U = VSUB(T5y, T5p); T6P = VFNMS(LDK(KP923879532), T6e, T6b); T6f = VFMA(LDK(KP923879532), T6e, T6b); { V T5V, T6u, T6g, T6v, T6s, T6J; T6s = VSUB(T6q, T6r); T6J = VADD(T6q, T6r); T5V = VFNMS(LDK(KP098491403), T5U, T5N); T6u = VFMA(LDK(KP098491403), T5N, T5U); T75 = VFNMS(LDK(KP980785280), T6U, T6T); T6V = VFMA(LDK(KP980785280), T6U, T6T); T5A = VFMA(LDK(KP980785280), T5z, T5g); T6A = VFNMS(LDK(KP980785280), T5z, T5g); T6g = VFNMS(LDK(KP098491403), T6f, T68); T6v = VFMA(LDK(KP098491403), T68, T6f); T72 = VFNMS(LDK(KP980785280), T6J, T6I); T6K = VFMA(LDK(KP980785280), T6J, T6I); T6t = VFMA(LDK(KP980785280), T6s, T6p); T6D = VFNMS(LDK(KP980785280), T6s, T6p); T6w = VSUB(T6u, T6v); T6B = VADD(T6u, T6v); T6h = VADD(T5V, T6g); T6E = VSUB(T6g, T5V); } } } } } } } { V T6W, T6N, T6G, T6C, T6z, T6x, T6H, T6F, T6y, T6i, T6X, T6Q; T6W = VFNMS(LDK(KP820678790), T6L, T6M); T6N = VFMA(LDK(KP820678790), T6M, T6L); T6G = VFMA(LDK(KP995184726), T6B, T6A); T6C = VFNMS(LDK(KP995184726), T6B, T6A); T6z = VFMA(LDK(KP995184726), T6w, T6t); T6x = VFNMS(LDK(KP995184726), T6w, T6t); T6H = VFMA(LDK(KP995184726), T6E, T6D); T6F = VFNMS(LDK(KP995184726), T6E, T6D); T6y = VFMA(LDK(KP995184726), T6h, T5A); T6i = VFNMS(LDK(KP995184726), T6h, T5A); T6X = VFNMS(LDK(KP820678790), T6O, T6P); T6Q = VFMA(LDK(KP820678790), T6P, T6O); { V T73, T6Y, T76, T6R; ST(&(x[WS(rs, 49)]), VFNMSI(T6H, T6G), ms, &(x[WS(rs, 1)])); ST(&(x[WS(rs, 15)]), VFMAI(T6H, T6G), ms, &(x[WS(rs, 1)])); ST(&(x[WS(rs, 47)]), VFMAI(T6F, T6C), ms, &(x[WS(rs, 1)])); ST(&(x[WS(rs, 17)]), VFNMSI(T6F, T6C), ms, &(x[WS(rs, 1)])); ST(&(x[WS(rs, 63)]), VFMAI(T6z, T6y), ms, &(x[WS(rs, 1)])); ST(&(x[WS(rs, 1)]), VFNMSI(T6z, T6y), ms, &(x[WS(rs, 1)])); ST(&(x[WS(rs, 31)]), VFMAI(T6x, T6i), ms, &(x[WS(rs, 1)])); ST(&(x[WS(rs, 33)]), VFNMSI(T6x, T6i), ms, &(x[WS(rs, 1)])); T73 = VADD(T6W, T6X); T6Y = VSUB(T6W, T6X); T76 = VSUB(T6Q, T6N); T6R = VADD(T6N, T6Q); { V T78, T74, T71, T6Z, T79, T77, T70, T6S; T78 = VFNMS(LDK(KP773010453), T73, T72); T74 = VFMA(LDK(KP773010453), T73, T72); T71 = VFMA(LDK(KP773010453), T6Y, T6V); T6Z = VFNMS(LDK(KP773010453), T6Y, T6V); T79 = VFNMS(LDK(KP773010453), T76, T75); T77 = VFMA(LDK(KP773010453), T76, T75); T70 = VFMA(LDK(KP773010453), T6R, T6K); T6S = VFNMS(LDK(KP773010453), T6R, T6K); ST(&(x[WS(rs, 55)]), VFMAI(T79, T78), ms, &(x[WS(rs, 1)])); ST(&(x[WS(rs, 9)]), VFNMSI(T79, T78), ms, &(x[WS(rs, 1)])); ST(&(x[WS(rs, 41)]), VFNMSI(T77, T74), ms, &(x[WS(rs, 1)])); ST(&(x[WS(rs, 23)]), VFMAI(T77, T74), ms, &(x[WS(rs, 1)])); ST(&(x[WS(rs, 7)]), VFMAI(T71, T70), ms, &(x[WS(rs, 1)])); ST(&(x[WS(rs, 57)]), VFNMSI(T71, T70), ms, &(x[WS(rs, 1)])); ST(&(x[WS(rs, 39)]), VFMAI(T6Z, T6S), ms, &(x[WS(rs, 1)])); ST(&(x[WS(rs, 25)]), VFNMSI(T6Z, T6S), ms, &(x[WS(rs, 1)])); } } } } } VLEAVE(); } static const tw_instr twinstr[] = { VTW(0, 1), VTW(0, 2), VTW(0, 3), VTW(0, 4), VTW(0, 5), VTW(0, 6), VTW(0, 7), VTW(0, 8), VTW(0, 9), VTW(0, 10), VTW(0, 11), VTW(0, 12), VTW(0, 13), VTW(0, 14), VTW(0, 15), VTW(0, 16), VTW(0, 17), VTW(0, 18), VTW(0, 19), VTW(0, 20), VTW(0, 21), VTW(0, 22), VTW(0, 23), VTW(0, 24), VTW(0, 25), VTW(0, 26), VTW(0, 27), VTW(0, 28), VTW(0, 29), VTW(0, 30), VTW(0, 31), VTW(0, 32), VTW(0, 33), VTW(0, 34), VTW(0, 35), VTW(0, 36), VTW(0, 37), VTW(0, 38), VTW(0, 39), VTW(0, 40), VTW(0, 41), VTW(0, 42), VTW(0, 43), VTW(0, 44), VTW(0, 45), VTW(0, 46), VTW(0, 47), VTW(0, 48), VTW(0, 49), VTW(0, 50), VTW(0, 51), VTW(0, 52), VTW(0, 53), VTW(0, 54), VTW(0, 55), VTW(0, 56), VTW(0, 57), VTW(0, 58), VTW(0, 59), VTW(0, 60), VTW(0, 61), VTW(0, 62), VTW(0, 63), {TW_NEXT, VL, 0} }; static const ct_desc desc = { 64, XSIMD_STRING("t2fv_64"), twinstr, &GENUS, {261, 126, 258, 0}, 0, 0, 0 }; void XSIMD(codelet_t2fv_64) (planner *p) { X(kdft_dit_register) (p, t2fv_64, &desc); } #else /* HAVE_FMA */ /* Generated by: ../../../genfft/gen_twiddle_c.native -simd -compact -variables 4 -pipeline-latency 8 -n 64 -name t2fv_64 -include t2f.h */ /* * This function contains 519 FP additions, 250 FP multiplications, * (or, 467 additions, 198 multiplications, 52 fused multiply/add), * 107 stack variables, 15 constants, and 128 memory accesses */ #include "t2f.h" static void t2fv_64(R *ri, R *ii, const R *W, stride rs, INT mb, INT me, INT ms) { DVK(KP995184726, +0.995184726672196886244836953109479921575474869); DVK(KP098017140, +0.098017140329560601994195563888641845861136673); DVK(KP773010453, +0.773010453362736960810906609758469800971041293); DVK(KP634393284, +0.634393284163645498215171613225493370675687095); DVK(KP471396736, +0.471396736825997648556387625905254377657460319); DVK(KP881921264, +0.881921264348355029712756863660388349508442621); DVK(KP290284677, +0.290284677254462367636192375817395274691476278); DVK(KP956940335, +0.956940335732208864935797886980269969482849206); DVK(KP555570233, +0.555570233019602224742830813948532874374937191); DVK(KP831469612, +0.831469612302545237078788377617905756738560812); DVK(KP195090322, +0.195090322016128267848284868477022240927691618); DVK(KP980785280, +0.980785280403230449126182236134239036973933731); DVK(KP382683432, +0.382683432365089771728459984030398866761344562); DVK(KP923879532, +0.923879532511286756128183189396788286822416626); DVK(KP707106781, +0.707106781186547524400844362104849039284835938); { INT m; R *x; x = ri; for (m = mb, W = W + (mb * ((TWVL / VL) * 126)); m < me; m = m + VL, x = x + (VL * ms), W = W + (TWVL * 126), MAKE_VOLATILE_STRIDE(64, rs)) { V Tg, T4a, T6r, T7f, T3o, T4B, T5q, T7e, T5R, T62, T28, T4o, T2g, T4l, T7n; V T7Z, T68, T6j, T2C, T4s, T3a, T4v, T7u, T82, T7E, T7F, T7V, T5F, T6u, T1k; V T4e, T1r, T4d, T7B, T7C, T7W, T5M, T6v, TV, T4g, T12, T4h, T7h, T7i, TD; V T4C, T3h, T4b, T5x, T6s, T1R, T4m, T7q, T80, T2j, T4p, T5Y, T63, T2Z, T4w; V T7x, T83, T33, T4t, T6f, T6k; { V T1, T3, T3m, T3k, Tb, Td, Te, T6, T8, T9, T2, T3l, T3j; T1 = LD(&(x[0]), ms, &(x[0])); T2 = LD(&(x[WS(rs, 32)]), ms, &(x[0])); T3 = BYTWJ(&(W[TWVL * 62]), T2); T3l = LD(&(x[WS(rs, 48)]), ms, &(x[0])); T3m = BYTWJ(&(W[TWVL * 94]), T3l); T3j = LD(&(x[WS(rs, 16)]), ms, &(x[0])); T3k = BYTWJ(&(W[TWVL * 30]), T3j); { V Ta, Tc, T5, T7; Ta = LD(&(x[WS(rs, 56)]), ms, &(x[0])); Tb = BYTWJ(&(W[TWVL * 110]), Ta); Tc = LD(&(x[WS(rs, 24)]), ms, &(x[0])); Td = BYTWJ(&(W[TWVL * 46]), Tc); Te = VSUB(Tb, Td); T5 = LD(&(x[WS(rs, 8)]), ms, &(x[0])); T6 = BYTWJ(&(W[TWVL * 14]), T5); T7 = LD(&(x[WS(rs, 40)]), ms, &(x[0])); T8 = BYTWJ(&(W[TWVL * 78]), T7); T9 = VSUB(T6, T8); } { V T4, Tf, T6p, T6q; T4 = VSUB(T1, T3); Tf = VMUL(LDK(KP707106781), VADD(T9, Te)); Tg = VADD(T4, Tf); T4a = VSUB(T4, Tf); T6p = VADD(Tb, Td); T6q = VADD(T6, T8); T6r = VSUB(T6p, T6q); T7f = VADD(T6q, T6p); } { V T3i, T3n, T5o, T5p; T3i = VMUL(LDK(KP707106781), VSUB(Te, T9)); T3n = VSUB(T3k, T3m); T3o = VSUB(T3i, T3n); T4B = VADD(T3n, T3i); T5o = VADD(T1, T3); T5p = VADD(T3k, T3m); T5q = VSUB(T5o, T5p); T7e = VADD(T5o, T5p); } } { V T24, T26, T5Q, T2b, T2d, T5P, T1W, T60, T21, T61, T22, T27; { V T23, T25, T2a, T2c; T23 = LD(&(x[WS(rs, 17)]), ms, &(x[WS(rs, 1)])); T24 = BYTWJ(&(W[TWVL * 32]), T23); T25 = LD(&(x[WS(rs, 49)]), ms, &(x[WS(rs, 1)])); T26 = BYTWJ(&(W[TWVL * 96]), T25); T5Q = VADD(T24, T26); T2a = LD(&(x[WS(rs, 1)]), ms, &(x[WS(rs, 1)])); T2b = BYTWJ(&(W[0]), T2a); T2c = LD(&(x[WS(rs, 33)]), ms, &(x[WS(rs, 1)])); T2d = BYTWJ(&(W[TWVL * 64]), T2c); T5P = VADD(T2b, T2d); } { V T1T, T1V, T1S, T1U; T1S = LD(&(x[WS(rs, 57)]), ms, &(x[WS(rs, 1)])); T1T = BYTWJ(&(W[TWVL * 112]), T1S); T1U = LD(&(x[WS(rs, 25)]), ms, &(x[WS(rs, 1)])); T1V = BYTWJ(&(W[TWVL * 48]), T1U); T1W = VSUB(T1T, T1V); T60 = VADD(T1T, T1V); } { V T1Y, T20, T1X, T1Z; T1X = LD(&(x[WS(rs, 9)]), ms, &(x[WS(rs, 1)])); T1Y = BYTWJ(&(W[TWVL * 16]), T1X); T1Z = LD(&(x[WS(rs, 41)]), ms, &(x[WS(rs, 1)])); T20 = BYTWJ(&(W[TWVL * 80]), T1Z); T21 = VSUB(T1Y, T20); T61 = VADD(T1Y, T20); } T5R = VSUB(T5P, T5Q); T62 = VSUB(T60, T61); T22 = VMUL(LDK(KP707106781), VSUB(T1W, T21)); T27 = VSUB(T24, T26); T28 = VSUB(T22, T27); T4o = VADD(T27, T22); { V T2e, T2f, T7l, T7m; T2e = VSUB(T2b, T2d); T2f = VMUL(LDK(KP707106781), VADD(T21, T1W)); T2g = VADD(T2e, T2f); T4l = VSUB(T2e, T2f); T7l = VADD(T5P, T5Q); T7m = VADD(T61, T60); T7n = VADD(T7l, T7m); T7Z = VSUB(T7l, T7m); } } { V T2n, T2p, T66, T36, T38, T67, T2v, T6i, T2A, T6h, T2q, T2B; { V T2m, T2o, T35, T37; T2m = LD(&(x[WS(rs, 63)]), ms, &(x[WS(rs, 1)])); T2n = BYTWJ(&(W[TWVL * 124]), T2m); T2o = LD(&(x[WS(rs, 31)]), ms, &(x[WS(rs, 1)])); T2p = BYTWJ(&(W[TWVL * 60]), T2o); T66 = VADD(T2n, T2p); T35 = LD(&(x[WS(rs, 15)]), ms, &(x[WS(rs, 1)])); T36 = BYTWJ(&(W[TWVL * 28]), T35); T37 = LD(&(x[WS(rs, 47)]), ms, &(x[WS(rs, 1)])); T38 = BYTWJ(&(W[TWVL * 92]), T37); T67 = VADD(T36, T38); } { V T2s, T2u, T2r, T2t; T2r = LD(&(x[WS(rs, 7)]), ms, &(x[WS(rs, 1)])); T2s = BYTWJ(&(W[TWVL * 12]), T2r); T2t = LD(&(x[WS(rs, 39)]), ms, &(x[WS(rs, 1)])); T2u = BYTWJ(&(W[TWVL * 76]), T2t); T2v = VSUB(T2s, T2u); T6i = VADD(T2s, T2u); } { V T2x, T2z, T2w, T2y; T2w = LD(&(x[WS(rs, 55)]), ms, &(x[WS(rs, 1)])); T2x = BYTWJ(&(W[TWVL * 108]), T2w); T2y = LD(&(x[WS(rs, 23)]), ms, &(x[WS(rs, 1)])); T2z = BYTWJ(&(W[TWVL * 44]), T2y); T2A = VSUB(T2x, T2z); T6h = VADD(T2x, T2z); } T68 = VSUB(T66, T67); T6j = VSUB(T6h, T6i); T2q = VSUB(T2n, T2p); T2B = VMUL(LDK(KP707106781), VADD(T2v, T2A)); T2C = VADD(T2q, T2B); T4s = VSUB(T2q, T2B); { V T34, T39, T7s, T7t; T34 = VMUL(LDK(KP707106781), VSUB(T2A, T2v)); T39 = VSUB(T36, T38); T3a = VSUB(T34, T39); T4v = VADD(T39, T34); T7s = VADD(T66, T67); T7t = VADD(T6i, T6h); T7u = VADD(T7s, T7t); T82 = VSUB(T7s, T7t); } } { V T1g, T1i, T5A, T1m, T1o, T5z, T18, T5C, T1d, T5D, T5B, T5E; { V T1f, T1h, T1l, T1n; T1f = LD(&(x[WS(rs, 18)]), ms, &(x[0])); T1g = BYTWJ(&(W[TWVL * 34]), T1f); T1h = LD(&(x[WS(rs, 50)]), ms, &(x[0])); T1i = BYTWJ(&(W[TWVL * 98]), T1h); T5A = VADD(T1g, T1i); T1l = LD(&(x[WS(rs, 2)]), ms, &(x[0])); T1m = BYTWJ(&(W[TWVL * 2]), T1l); T1n = LD(&(x[WS(rs, 34)]), ms, &(x[0])); T1o = BYTWJ(&(W[TWVL * 66]), T1n); T5z = VADD(T1m, T1o); } { V T15, T17, T14, T16; T14 = LD(&(x[WS(rs, 58)]), ms, &(x[0])); T15 = BYTWJ(&(W[TWVL * 114]), T14); T16 = LD(&(x[WS(rs, 26)]), ms, &(x[0])); T17 = BYTWJ(&(W[TWVL * 50]), T16); T18 = VSUB(T15, T17); T5C = VADD(T15, T17); } { V T1a, T1c, T19, T1b; T19 = LD(&(x[WS(rs, 10)]), ms, &(x[0])); T1a = BYTWJ(&(W[TWVL * 18]), T19); T1b = LD(&(x[WS(rs, 42)]), ms, &(x[0])); T1c = BYTWJ(&(W[TWVL * 82]), T1b); T1d = VSUB(T1a, T1c); T5D = VADD(T1a, T1c); } T7E = VADD(T5z, T5A); T7F = VADD(T5D, T5C); T7V = VSUB(T7E, T7F); T5B = VSUB(T5z, T5A); T5E = VSUB(T5C, T5D); T5F = VFMA(LDK(KP923879532), T5B, VMUL(LDK(KP382683432), T5E)); T6u = VFNMS(LDK(KP382683432), T5B, VMUL(LDK(KP923879532), T5E)); { V T1e, T1j, T1p, T1q; T1e = VMUL(LDK(KP707106781), VSUB(T18, T1d)); T1j = VSUB(T1g, T1i); T1k = VSUB(T1e, T1j); T4e = VADD(T1j, T1e); T1p = VSUB(T1m, T1o); T1q = VMUL(LDK(KP707106781), VADD(T1d, T18)); T1r = VADD(T1p, T1q); T4d = VSUB(T1p, T1q); } } { V TG, TI, T5G, TY, T10, T5H, TO, T5K, TT, T5J, T5I, T5L; { V TF, TH, TX, TZ; TF = LD(&(x[WS(rs, 62)]), ms, &(x[0])); TG = BYTWJ(&(W[TWVL * 122]), TF); TH = LD(&(x[WS(rs, 30)]), ms, &(x[0])); TI = BYTWJ(&(W[TWVL * 58]), TH); T5G = VADD(TG, TI); TX = LD(&(x[WS(rs, 14)]), ms, &(x[0])); TY = BYTWJ(&(W[TWVL * 26]), TX); TZ = LD(&(x[WS(rs, 46)]), ms, &(x[0])); T10 = BYTWJ(&(W[TWVL * 90]), TZ); T5H = VADD(TY, T10); } { V TL, TN, TK, TM; TK = LD(&(x[WS(rs, 6)]), ms, &(x[0])); TL = BYTWJ(&(W[TWVL * 10]), TK); TM = LD(&(x[WS(rs, 38)]), ms, &(x[0])); TN = BYTWJ(&(W[TWVL * 74]), TM); TO = VSUB(TL, TN); T5K = VADD(TL, TN); } { V TQ, TS, TP, TR; TP = LD(&(x[WS(rs, 54)]), ms, &(x[0])); TQ = BYTWJ(&(W[TWVL * 106]), TP); TR = LD(&(x[WS(rs, 22)]), ms, &(x[0])); TS = BYTWJ(&(W[TWVL * 42]), TR); TT = VSUB(TQ, TS); T5J = VADD(TQ, TS); } T7B = VADD(T5G, T5H); T7C = VADD(T5K, T5J); T7W = VSUB(T7B, T7C); T5I = VSUB(T5G, T5H); T5L = VSUB(T5J, T5K); T5M = VFNMS(LDK(KP382683432), T5L, VMUL(LDK(KP923879532), T5I)); T6v = VFMA(LDK(KP382683432), T5I, VMUL(LDK(KP923879532), T5L)); { V TJ, TU, TW, T11; TJ = VSUB(TG, TI); TU = VMUL(LDK(KP707106781), VADD(TO, TT)); TV = VADD(TJ, TU); T4g = VSUB(TJ, TU); TW = VMUL(LDK(KP707106781), VSUB(TT, TO)); T11 = VSUB(TY, T10); T12 = VSUB(TW, T11); T4h = VADD(T11, TW); } } { V Tl, T5r, TB, T5v, Tq, T5s, Tw, T5u, Tr, TC; { V Ti, Tk, Th, Tj; Th = LD(&(x[WS(rs, 4)]), ms, &(x[0])); Ti = BYTWJ(&(W[TWVL * 6]), Th); Tj = LD(&(x[WS(rs, 36)]), ms, &(x[0])); Tk = BYTWJ(&(W[TWVL * 70]), Tj); Tl = VSUB(Ti, Tk); T5r = VADD(Ti, Tk); } { V Ty, TA, Tx, Tz; Tx = LD(&(x[WS(rs, 12)]), ms, &(x[0])); Ty = BYTWJ(&(W[TWVL * 22]), Tx); Tz = LD(&(x[WS(rs, 44)]), ms, &(x[0])); TA = BYTWJ(&(W[TWVL * 86]), Tz); TB = VSUB(Ty, TA); T5v = VADD(Ty, TA); } { V Tn, Tp, Tm, To; Tm = LD(&(x[WS(rs, 20)]), ms, &(x[0])); Tn = BYTWJ(&(W[TWVL * 38]), Tm); To = LD(&(x[WS(rs, 52)]), ms, &(x[0])); Tp = BYTWJ(&(W[TWVL * 102]), To); Tq = VSUB(Tn, Tp); T5s = VADD(Tn, Tp); } { V Tt, Tv, Ts, Tu; Ts = LD(&(x[WS(rs, 60)]), ms, &(x[0])); Tt = BYTWJ(&(W[TWVL * 118]), Ts); Tu = LD(&(x[WS(rs, 28)]), ms, &(x[0])); Tv = BYTWJ(&(W[TWVL * 54]), Tu); Tw = VSUB(Tt, Tv); T5u = VADD(Tt, Tv); } T7h = VADD(T5r, T5s); T7i = VADD(T5u, T5v); Tr = VFNMS(LDK(KP382683432), Tq, VMUL(LDK(KP923879532), Tl)); TC = VFMA(LDK(KP923879532), Tw, VMUL(LDK(KP382683432), TB)); TD = VADD(Tr, TC); T4C = VSUB(TC, Tr); { V T3f, T3g, T5t, T5w; T3f = VFNMS(LDK(KP923879532), TB, VMUL(LDK(KP382683432), Tw)); T3g = VFMA(LDK(KP382683432), Tl, VMUL(LDK(KP923879532), Tq)); T3h = VSUB(T3f, T3g); T4b = VADD(T3g, T3f); T5t = VSUB(T5r, T5s); T5w = VSUB(T5u, T5v); T5x = VMUL(LDK(KP707106781), VADD(T5t, T5w)); T6s = VMUL(LDK(KP707106781), VSUB(T5w, T5t)); } } { V T1z, T5V, T1P, T5T, T1E, T5W, T1K, T5S; { V T1w, T1y, T1v, T1x; T1v = LD(&(x[WS(rs, 61)]), ms, &(x[WS(rs, 1)])); T1w = BYTWJ(&(W[TWVL * 120]), T1v); T1x = LD(&(x[WS(rs, 29)]), ms, &(x[WS(rs, 1)])); T1y = BYTWJ(&(W[TWVL * 56]), T1x); T1z = VSUB(T1w, T1y); T5V = VADD(T1w, T1y); } { V T1M, T1O, T1L, T1N; T1L = LD(&(x[WS(rs, 21)]), ms, &(x[WS(rs, 1)])); T1M = BYTWJ(&(W[TWVL * 40]), T1L); T1N = LD(&(x[WS(rs, 53)]), ms, &(x[WS(rs, 1)])); T1O = BYTWJ(&(W[TWVL * 104]), T1N); T1P = VSUB(T1M, T1O); T5T = VADD(T1M, T1O); } { V T1B, T1D, T1A, T1C; T1A = LD(&(x[WS(rs, 13)]), ms, &(x[WS(rs, 1)])); T1B = BYTWJ(&(W[TWVL * 24]), T1A); T1C = LD(&(x[WS(rs, 45)]), ms, &(x[WS(rs, 1)])); T1D = BYTWJ(&(W[TWVL * 88]), T1C); T1E = VSUB(T1B, T1D); T5W = VADD(T1B, T1D); } { V T1H, T1J, T1G, T1I; T1G = LD(&(x[WS(rs, 5)]), ms, &(x[WS(rs, 1)])); T1H = BYTWJ(&(W[TWVL * 8]), T1G); T1I = LD(&(x[WS(rs, 37)]), ms, &(x[WS(rs, 1)])); T1J = BYTWJ(&(W[TWVL * 72]), T1I); T1K = VSUB(T1H, T1J); T5S = VADD(T1H, T1J); } { V T1F, T1Q, T7o, T7p; T1F = VFNMS(LDK(KP923879532), T1E, VMUL(LDK(KP382683432), T1z)); T1Q = VFMA(LDK(KP382683432), T1K, VMUL(LDK(KP923879532), T1P)); T1R = VSUB(T1F, T1Q); T4m = VADD(T1Q, T1F); T7o = VADD(T5S, T5T); T7p = VADD(T5V, T5W); T7q = VADD(T7o, T7p); T80 = VSUB(T7p, T7o); } { V T2h, T2i, T5U, T5X; T2h = VFNMS(LDK(KP382683432), T1P, VMUL(LDK(KP923879532), T1K)); T2i = VFMA(LDK(KP923879532), T1z, VMUL(LDK(KP382683432), T1E)); T2j = VADD(T2h, T2i); T4p = VSUB(T2i, T2h); T5U = VSUB(T5S, T5T); T5X = VSUB(T5V, T5W); T5Y = VMUL(LDK(KP707106781), VADD(T5U, T5X)); T63 = VMUL(LDK(KP707106781), VSUB(T5X, T5U)); } } { V T2H, T69, T2X, T6d, T2M, T6a, T2S, T6c; { V T2E, T2G, T2D, T2F; T2D = LD(&(x[WS(rs, 3)]), ms, &(x[WS(rs, 1)])); T2E = BYTWJ(&(W[TWVL * 4]), T2D); T2F = LD(&(x[WS(rs, 35)]), ms, &(x[WS(rs, 1)])); T2G = BYTWJ(&(W[TWVL * 68]), T2F); T2H = VSUB(T2E, T2G); T69 = VADD(T2E, T2G); } { V T2U, T2W, T2T, T2V; T2T = LD(&(x[WS(rs, 11)]), ms, &(x[WS(rs, 1)])); T2U = BYTWJ(&(W[TWVL * 20]), T2T); T2V = LD(&(x[WS(rs, 43)]), ms, &(x[WS(rs, 1)])); T2W = BYTWJ(&(W[TWVL * 84]), T2V); T2X = VSUB(T2U, T2W); T6d = VADD(T2U, T2W); } { V T2J, T2L, T2I, T2K; T2I = LD(&(x[WS(rs, 19)]), ms, &(x[WS(rs, 1)])); T2J = BYTWJ(&(W[TWVL * 36]), T2I); T2K = LD(&(x[WS(rs, 51)]), ms, &(x[WS(rs, 1)])); T2L = BYTWJ(&(W[TWVL * 100]), T2K); T2M = VSUB(T2J, T2L); T6a = VADD(T2J, T2L); } { V T2P, T2R, T2O, T2Q; T2O = LD(&(x[WS(rs, 59)]), ms, &(x[WS(rs, 1)])); T2P = BYTWJ(&(W[TWVL * 116]), T2O); T2Q = LD(&(x[WS(rs, 27)]), ms, &(x[WS(rs, 1)])); T2R = BYTWJ(&(W[TWVL * 52]), T2Q); T2S = VSUB(T2P, T2R); T6c = VADD(T2P, T2R); } { V T2N, T2Y, T7v, T7w; T2N = VFNMS(LDK(KP382683432), T2M, VMUL(LDK(KP923879532), T2H)); T2Y = VFMA(LDK(KP923879532), T2S, VMUL(LDK(KP382683432), T2X)); T2Z = VADD(T2N, T2Y); T4w = VSUB(T2Y, T2N); T7v = VADD(T69, T6a); T7w = VADD(T6c, T6d); T7x = VADD(T7v, T7w); T83 = VSUB(T7w, T7v); } { V T31, T32, T6b, T6e; T31 = VFNMS(LDK(KP923879532), T2X, VMUL(LDK(KP382683432), T2S)); T32 = VFMA(LDK(KP382683432), T2H, VMUL(LDK(KP923879532), T2M)); T33 = VSUB(T31, T32); T4t = VADD(T32, T31); T6b = VSUB(T69, T6a); T6e = VSUB(T6c, T6d); T6f = VMUL(LDK(KP707106781), VADD(T6b, T6e)); T6k = VMUL(LDK(KP707106781), VSUB(T6e, T6b)); } } { V T7k, T7M, T7R, T7T, T7z, T7I, T7H, T7N, T7O, T7S; { V T7g, T7j, T7P, T7Q; T7g = VADD(T7e, T7f); T7j = VADD(T7h, T7i); T7k = VSUB(T7g, T7j); T7M = VADD(T7g, T7j); T7P = VADD(T7n, T7q); T7Q = VADD(T7u, T7x); T7R = VADD(T7P, T7Q); T7T = VBYI(VSUB(T7Q, T7P)); } { V T7r, T7y, T7D, T7G; T7r = VSUB(T7n, T7q); T7y = VSUB(T7u, T7x); T7z = VMUL(LDK(KP707106781), VADD(T7r, T7y)); T7I = VMUL(LDK(KP707106781), VSUB(T7y, T7r)); T7D = VADD(T7B, T7C); T7G = VADD(T7E, T7F); T7H = VSUB(T7D, T7G); T7N = VADD(T7G, T7D); } T7O = VADD(T7M, T7N); ST(&(x[WS(rs, 32)]), VSUB(T7O, T7R), ms, &(x[0])); ST(&(x[0]), VADD(T7O, T7R), ms, &(x[0])); T7S = VSUB(T7M, T7N); ST(&(x[WS(rs, 48)]), VSUB(T7S, T7T), ms, &(x[0])); ST(&(x[WS(rs, 16)]), VADD(T7S, T7T), ms, &(x[0])); { V T7A, T7J, T7K, T7L; T7A = VADD(T7k, T7z); T7J = VBYI(VADD(T7H, T7I)); ST(&(x[WS(rs, 56)]), VSUB(T7A, T7J), ms, &(x[0])); ST(&(x[WS(rs, 8)]), VADD(T7A, T7J), ms, &(x[0])); T7K = VSUB(T7k, T7z); T7L = VBYI(VSUB(T7I, T7H)); ST(&(x[WS(rs, 40)]), VSUB(T7K, T7L), ms, &(x[0])); ST(&(x[WS(rs, 24)]), VADD(T7K, T7L), ms, &(x[0])); } } { V T7Y, T8j, T8c, T8k, T85, T8g, T89, T8h; { V T7U, T7X, T8a, T8b; T7U = VSUB(T7e, T7f); T7X = VMUL(LDK(KP707106781), VADD(T7V, T7W)); T7Y = VADD(T7U, T7X); T8j = VSUB(T7U, T7X); T8a = VFNMS(LDK(KP382683432), T7Z, VMUL(LDK(KP923879532), T80)); T8b = VFMA(LDK(KP382683432), T82, VMUL(LDK(KP923879532), T83)); T8c = VADD(T8a, T8b); T8k = VSUB(T8b, T8a); } { V T81, T84, T87, T88; T81 = VFMA(LDK(KP923879532), T7Z, VMUL(LDK(KP382683432), T80)); T84 = VFNMS(LDK(KP382683432), T83, VMUL(LDK(KP923879532), T82)); T85 = VADD(T81, T84); T8g = VSUB(T84, T81); T87 = VSUB(T7i, T7h); T88 = VMUL(LDK(KP707106781), VSUB(T7W, T7V)); T89 = VADD(T87, T88); T8h = VSUB(T88, T87); } { V T86, T8d, T8m, T8n; T86 = VADD(T7Y, T85); T8d = VBYI(VADD(T89, T8c)); ST(&(x[WS(rs, 60)]), VSUB(T86, T8d), ms, &(x[0])); ST(&(x[WS(rs, 4)]), VADD(T86, T8d), ms, &(x[0])); T8m = VBYI(VADD(T8h, T8g)); T8n = VADD(T8j, T8k); ST(&(x[WS(rs, 12)]), VADD(T8m, T8n), ms, &(x[0])); ST(&(x[WS(rs, 52)]), VSUB(T8n, T8m), ms, &(x[0])); } { V T8e, T8f, T8i, T8l; T8e = VSUB(T7Y, T85); T8f = VBYI(VSUB(T8c, T89)); ST(&(x[WS(rs, 36)]), VSUB(T8e, T8f), ms, &(x[0])); ST(&(x[WS(rs, 28)]), VADD(T8e, T8f), ms, &(x[0])); T8i = VBYI(VSUB(T8g, T8h)); T8l = VSUB(T8j, T8k); ST(&(x[WS(rs, 20)]), VADD(T8i, T8l), ms, &(x[0])); ST(&(x[WS(rs, 44)]), VSUB(T8l, T8i), ms, &(x[0])); } } { V T5O, T6H, T6x, T6F, T6n, T6I, T6A, T6E; { V T5y, T5N, T6t, T6w; T5y = VADD(T5q, T5x); T5N = VADD(T5F, T5M); T5O = VADD(T5y, T5N); T6H = VSUB(T5y, T5N); T6t = VADD(T6r, T6s); T6w = VADD(T6u, T6v); T6x = VADD(T6t, T6w); T6F = VSUB(T6w, T6t); { V T65, T6y, T6m, T6z; { V T5Z, T64, T6g, T6l; T5Z = VADD(T5R, T5Y); T64 = VADD(T62, T63); T65 = VFMA(LDK(KP980785280), T5Z, VMUL(LDK(KP195090322), T64)); T6y = VFNMS(LDK(KP195090322), T5Z, VMUL(LDK(KP980785280), T64)); T6g = VADD(T68, T6f); T6l = VADD(T6j, T6k); T6m = VFNMS(LDK(KP195090322), T6l, VMUL(LDK(KP980785280), T6g)); T6z = VFMA(LDK(KP195090322), T6g, VMUL(LDK(KP980785280), T6l)); } T6n = VADD(T65, T6m); T6I = VSUB(T6z, T6y); T6A = VADD(T6y, T6z); T6E = VSUB(T6m, T65); } } { V T6o, T6B, T6K, T6L; T6o = VADD(T5O, T6n); T6B = VBYI(VADD(T6x, T6A)); ST(&(x[WS(rs, 62)]), VSUB(T6o, T6B), ms, &(x[0])); ST(&(x[WS(rs, 2)]), VADD(T6o, T6B), ms, &(x[0])); T6K = VBYI(VADD(T6F, T6E)); T6L = VADD(T6H, T6I); ST(&(x[WS(rs, 14)]), VADD(T6K, T6L), ms, &(x[0])); ST(&(x[WS(rs, 50)]), VSUB(T6L, T6K), ms, &(x[0])); } { V T6C, T6D, T6G, T6J; T6C = VSUB(T5O, T6n); T6D = VBYI(VSUB(T6A, T6x)); ST(&(x[WS(rs, 34)]), VSUB(T6C, T6D), ms, &(x[0])); ST(&(x[WS(rs, 30)]), VADD(T6C, T6D), ms, &(x[0])); T6G = VBYI(VSUB(T6E, T6F)); T6J = VSUB(T6H, T6I); ST(&(x[WS(rs, 18)]), VADD(T6G, T6J), ms, &(x[0])); ST(&(x[WS(rs, 46)]), VSUB(T6J, T6G), ms, &(x[0])); } } { V T6O, T79, T6Z, T77, T6V, T7a, T72, T76; { V T6M, T6N, T6X, T6Y; T6M = VSUB(T5q, T5x); T6N = VSUB(T6v, T6u); T6O = VADD(T6M, T6N); T79 = VSUB(T6M, T6N); T6X = VSUB(T6s, T6r); T6Y = VSUB(T5M, T5F); T6Z = VADD(T6X, T6Y); T77 = VSUB(T6Y, T6X); { V T6R, T70, T6U, T71; { V T6P, T6Q, T6S, T6T; T6P = VSUB(T5R, T5Y); T6Q = VSUB(T63, T62); T6R = VFMA(LDK(KP831469612), T6P, VMUL(LDK(KP555570233), T6Q)); T70 = VFNMS(LDK(KP555570233), T6P, VMUL(LDK(KP831469612), T6Q)); T6S = VSUB(T68, T6f); T6T = VSUB(T6k, T6j); T6U = VFNMS(LDK(KP555570233), T6T, VMUL(LDK(KP831469612), T6S)); T71 = VFMA(LDK(KP555570233), T6S, VMUL(LDK(KP831469612), T6T)); } T6V = VADD(T6R, T6U); T7a = VSUB(T71, T70); T72 = VADD(T70, T71); T76 = VSUB(T6U, T6R); } } { V T6W, T73, T7c, T7d; T6W = VADD(T6O, T6V); T73 = VBYI(VADD(T6Z, T72)); ST(&(x[WS(rs, 58)]), VSUB(T6W, T73), ms, &(x[0])); ST(&(x[WS(rs, 6)]), VADD(T6W, T73), ms, &(x[0])); T7c = VBYI(VADD(T77, T76)); T7d = VADD(T79, T7a); ST(&(x[WS(rs, 10)]), VADD(T7c, T7d), ms, &(x[0])); ST(&(x[WS(rs, 54)]), VSUB(T7d, T7c), ms, &(x[0])); } { V T74, T75, T78, T7b; T74 = VSUB(T6O, T6V); T75 = VBYI(VSUB(T72, T6Z)); ST(&(x[WS(rs, 38)]), VSUB(T74, T75), ms, &(x[0])); ST(&(x[WS(rs, 26)]), VADD(T74, T75), ms, &(x[0])); T78 = VBYI(VSUB(T76, T77)); T7b = VSUB(T79, T7a); ST(&(x[WS(rs, 22)]), VADD(T78, T7b), ms, &(x[0])); ST(&(x[WS(rs, 42)]), VSUB(T7b, T78), ms, &(x[0])); } } { V T4k, T5h, T4R, T59, T4H, T5j, T4P, T4Y, T4z, T4S, T4K, T4O, T55, T5k, T5c; V T5g; { V T4c, T57, T4j, T58, T4f, T4i; T4c = VADD(T4a, T4b); T57 = VSUB(T4C, T4B); T4f = VFMA(LDK(KP831469612), T4d, VMUL(LDK(KP555570233), T4e)); T4i = VFNMS(LDK(KP555570233), T4h, VMUL(LDK(KP831469612), T4g)); T4j = VADD(T4f, T4i); T58 = VSUB(T4i, T4f); T4k = VADD(T4c, T4j); T5h = VSUB(T58, T57); T4R = VSUB(T4c, T4j); T59 = VADD(T57, T58); } { V T4D, T4W, T4G, T4X, T4E, T4F; T4D = VADD(T4B, T4C); T4W = VSUB(T4a, T4b); T4E = VFNMS(LDK(KP555570233), T4d, VMUL(LDK(KP831469612), T4e)); T4F = VFMA(LDK(KP555570233), T4g, VMUL(LDK(KP831469612), T4h)); T4G = VADD(T4E, T4F); T4X = VSUB(T4F, T4E); T4H = VADD(T4D, T4G); T5j = VSUB(T4W, T4X); T4P = VSUB(T4G, T4D); T4Y = VADD(T4W, T4X); } { V T4r, T4I, T4y, T4J; { V T4n, T4q, T4u, T4x; T4n = VADD(T4l, T4m); T4q = VADD(T4o, T4p); T4r = VFMA(LDK(KP956940335), T4n, VMUL(LDK(KP290284677), T4q)); T4I = VFNMS(LDK(KP290284677), T4n, VMUL(LDK(KP956940335), T4q)); T4u = VADD(T4s, T4t); T4x = VADD(T4v, T4w); T4y = VFNMS(LDK(KP290284677), T4x, VMUL(LDK(KP956940335), T4u)); T4J = VFMA(LDK(KP290284677), T4u, VMUL(LDK(KP956940335), T4x)); } T4z = VADD(T4r, T4y); T4S = VSUB(T4J, T4I); T4K = VADD(T4I, T4J); T4O = VSUB(T4y, T4r); } { V T51, T5a, T54, T5b; { V T4Z, T50, T52, T53; T4Z = VSUB(T4l, T4m); T50 = VSUB(T4p, T4o); T51 = VFMA(LDK(KP881921264), T4Z, VMUL(LDK(KP471396736), T50)); T5a = VFNMS(LDK(KP471396736), T4Z, VMUL(LDK(KP881921264), T50)); T52 = VSUB(T4s, T4t); T53 = VSUB(T4w, T4v); T54 = VFNMS(LDK(KP471396736), T53, VMUL(LDK(KP881921264), T52)); T5b = VFMA(LDK(KP471396736), T52, VMUL(LDK(KP881921264), T53)); } T55 = VADD(T51, T54); T5k = VSUB(T5b, T5a); T5c = VADD(T5a, T5b); T5g = VSUB(T54, T51); } { V T4A, T4L, T5i, T5l; T4A = VADD(T4k, T4z); T4L = VBYI(VADD(T4H, T4K)); ST(&(x[WS(rs, 61)]), VSUB(T4A, T4L), ms, &(x[WS(rs, 1)])); ST(&(x[WS(rs, 3)]), VADD(T4A, T4L), ms, &(x[WS(rs, 1)])); T5i = VBYI(VSUB(T5g, T5h)); T5l = VSUB(T5j, T5k); ST(&(x[WS(rs, 21)]), VADD(T5i, T5l), ms, &(x[WS(rs, 1)])); ST(&(x[WS(rs, 43)]), VSUB(T5l, T5i), ms, &(x[WS(rs, 1)])); } { V T5m, T5n, T4M, T4N; T5m = VBYI(VADD(T5h, T5g)); T5n = VADD(T5j, T5k); ST(&(x[WS(rs, 11)]), VADD(T5m, T5n), ms, &(x[WS(rs, 1)])); ST(&(x[WS(rs, 53)]), VSUB(T5n, T5m), ms, &(x[WS(rs, 1)])); T4M = VSUB(T4k, T4z); T4N = VBYI(VSUB(T4K, T4H)); ST(&(x[WS(rs, 35)]), VSUB(T4M, T4N), ms, &(x[WS(rs, 1)])); ST(&(x[WS(rs, 29)]), VADD(T4M, T4N), ms, &(x[WS(rs, 1)])); } { V T4Q, T4T, T56, T5d; T4Q = VBYI(VSUB(T4O, T4P)); T4T = VSUB(T4R, T4S); ST(&(x[WS(rs, 19)]), VADD(T4Q, T4T), ms, &(x[WS(rs, 1)])); ST(&(x[WS(rs, 45)]), VSUB(T4T, T4Q), ms, &(x[WS(rs, 1)])); T56 = VADD(T4Y, T55); T5d = VBYI(VADD(T59, T5c)); ST(&(x[WS(rs, 59)]), VSUB(T56, T5d), ms, &(x[WS(rs, 1)])); ST(&(x[WS(rs, 5)]), VADD(T56, T5d), ms, &(x[WS(rs, 1)])); } { V T5e, T5f, T4U, T4V; T5e = VSUB(T4Y, T55); T5f = VBYI(VSUB(T5c, T59)); ST(&(x[WS(rs, 37)]), VSUB(T5e, T5f), ms, &(x[WS(rs, 1)])); ST(&(x[WS(rs, 27)]), VADD(T5e, T5f), ms, &(x[WS(rs, 1)])); T4U = VBYI(VADD(T4P, T4O)); T4V = VADD(T4R, T4S); ST(&(x[WS(rs, 13)]), VADD(T4U, T4V), ms, &(x[WS(rs, 1)])); ST(&(x[WS(rs, 51)]), VSUB(T4V, T4U), ms, &(x[WS(rs, 1)])); } } { V T1u, T43, T3D, T3V, T3t, T45, T3B, T3K, T3d, T3E, T3w, T3A, T3R, T46, T3Y; V T42; { V TE, T3T, T1t, T3U, T13, T1s; TE = VSUB(Tg, TD); T3T = VADD(T3o, T3h); T13 = VFMA(LDK(KP195090322), TV, VMUL(LDK(KP980785280), T12)); T1s = VFNMS(LDK(KP195090322), T1r, VMUL(LDK(KP980785280), T1k)); T1t = VSUB(T13, T1s); T3U = VADD(T1s, T13); T1u = VADD(TE, T1t); T43 = VSUB(T3U, T3T); T3D = VSUB(TE, T1t); T3V = VADD(T3T, T3U); } { V T3p, T3I, T3s, T3J, T3q, T3r; T3p = VSUB(T3h, T3o); T3I = VADD(Tg, TD); T3q = VFNMS(LDK(KP195090322), T12, VMUL(LDK(KP980785280), TV)); T3r = VFMA(LDK(KP980785280), T1r, VMUL(LDK(KP195090322), T1k)); T3s = VSUB(T3q, T3r); T3J = VADD(T3r, T3q); T3t = VADD(T3p, T3s); T45 = VSUB(T3I, T3J); T3B = VSUB(T3s, T3p); T3K = VADD(T3I, T3J); } { V T2l, T3u, T3c, T3v; { V T29, T2k, T30, T3b; T29 = VSUB(T1R, T28); T2k = VSUB(T2g, T2j); T2l = VFMA(LDK(KP634393284), T29, VMUL(LDK(KP773010453), T2k)); T3u = VFNMS(LDK(KP634393284), T2k, VMUL(LDK(KP773010453), T29)); T30 = VSUB(T2C, T2Z); T3b = VSUB(T33, T3a); T3c = VFNMS(LDK(KP634393284), T3b, VMUL(LDK(KP773010453), T30)); T3v = VFMA(LDK(KP773010453), T3b, VMUL(LDK(KP634393284), T30)); } T3d = VADD(T2l, T3c); T3E = VSUB(T3v, T3u); T3w = VADD(T3u, T3v); T3A = VSUB(T3c, T2l); } { V T3N, T3W, T3Q, T3X; { V T3L, T3M, T3O, T3P; T3L = VADD(T28, T1R); T3M = VADD(T2g, T2j); T3N = VFMA(LDK(KP098017140), T3L, VMUL(LDK(KP995184726), T3M)); T3W = VFNMS(LDK(KP098017140), T3M, VMUL(LDK(KP995184726), T3L)); T3O = VADD(T2C, T2Z); T3P = VADD(T3a, T33); T3Q = VFNMS(LDK(KP098017140), T3P, VMUL(LDK(KP995184726), T3O)); T3X = VFMA(LDK(KP995184726), T3P, VMUL(LDK(KP098017140), T3O)); } T3R = VADD(T3N, T3Q); T46 = VSUB(T3X, T3W); T3Y = VADD(T3W, T3X); T42 = VSUB(T3Q, T3N); } { V T3e, T3x, T44, T47; T3e = VADD(T1u, T3d); T3x = VBYI(VADD(T3t, T3w)); ST(&(x[WS(rs, 57)]), VSUB(T3e, T3x), ms, &(x[WS(rs, 1)])); ST(&(x[WS(rs, 7)]), VADD(T3e, T3x), ms, &(x[WS(rs, 1)])); T44 = VBYI(VSUB(T42, T43)); T47 = VSUB(T45, T46); ST(&(x[WS(rs, 17)]), VADD(T44, T47), ms, &(x[WS(rs, 1)])); ST(&(x[WS(rs, 47)]), VSUB(T47, T44), ms, &(x[WS(rs, 1)])); } { V T48, T49, T3y, T3z; T48 = VBYI(VADD(T43, T42)); T49 = VADD(T45, T46); ST(&(x[WS(rs, 15)]), VADD(T48, T49), ms, &(x[WS(rs, 1)])); ST(&(x[WS(rs, 49)]), VSUB(T49, T48), ms, &(x[WS(rs, 1)])); T3y = VSUB(T1u, T3d); T3z = VBYI(VSUB(T3w, T3t)); ST(&(x[WS(rs, 39)]), VSUB(T3y, T3z), ms, &(x[WS(rs, 1)])); ST(&(x[WS(rs, 25)]), VADD(T3y, T3z), ms, &(x[WS(rs, 1)])); } { V T3C, T3F, T3S, T3Z; T3C = VBYI(VSUB(T3A, T3B)); T3F = VSUB(T3D, T3E); ST(&(x[WS(rs, 23)]), VADD(T3C, T3F), ms, &(x[WS(rs, 1)])); ST(&(x[WS(rs, 41)]), VSUB(T3F, T3C), ms, &(x[WS(rs, 1)])); T3S = VADD(T3K, T3R); T3Z = VBYI(VADD(T3V, T3Y)); ST(&(x[WS(rs, 63)]), VSUB(T3S, T3Z), ms, &(x[WS(rs, 1)])); ST(&(x[WS(rs, 1)]), VADD(T3S, T3Z), ms, &(x[WS(rs, 1)])); } { V T40, T41, T3G, T3H; T40 = VSUB(T3K, T3R); T41 = VBYI(VSUB(T3Y, T3V)); ST(&(x[WS(rs, 33)]), VSUB(T40, T41), ms, &(x[WS(rs, 1)])); ST(&(x[WS(rs, 31)]), VADD(T40, T41), ms, &(x[WS(rs, 1)])); T3G = VBYI(VADD(T3B, T3A)); T3H = VADD(T3D, T3E); ST(&(x[WS(rs, 9)]), VADD(T3G, T3H), ms, &(x[WS(rs, 1)])); ST(&(x[WS(rs, 55)]), VSUB(T3H, T3G), ms, &(x[WS(rs, 1)])); } } } } VLEAVE(); } static const tw_instr twinstr[] = { VTW(0, 1), VTW(0, 2), VTW(0, 3), VTW(0, 4), VTW(0, 5), VTW(0, 6), VTW(0, 7), VTW(0, 8), VTW(0, 9), VTW(0, 10), VTW(0, 11), VTW(0, 12), VTW(0, 13), VTW(0, 14), VTW(0, 15), VTW(0, 16), VTW(0, 17), VTW(0, 18), VTW(0, 19), VTW(0, 20), VTW(0, 21), VTW(0, 22), VTW(0, 23), VTW(0, 24), VTW(0, 25), VTW(0, 26), VTW(0, 27), VTW(0, 28), VTW(0, 29), VTW(0, 30), VTW(0, 31), VTW(0, 32), VTW(0, 33), VTW(0, 34), VTW(0, 35), VTW(0, 36), VTW(0, 37), VTW(0, 38), VTW(0, 39), VTW(0, 40), VTW(0, 41), VTW(0, 42), VTW(0, 43), VTW(0, 44), VTW(0, 45), VTW(0, 46), VTW(0, 47), VTW(0, 48), VTW(0, 49), VTW(0, 50), VTW(0, 51), VTW(0, 52), VTW(0, 53), VTW(0, 54), VTW(0, 55), VTW(0, 56), VTW(0, 57), VTW(0, 58), VTW(0, 59), VTW(0, 60), VTW(0, 61), VTW(0, 62), VTW(0, 63), {TW_NEXT, VL, 0} }; static const ct_desc desc = { 64, XSIMD_STRING("t2fv_64"), twinstr, &GENUS, {467, 198, 52, 0}, 0, 0, 0 }; void XSIMD(codelet_t2fv_64) (planner *p) { X(kdft_dit_register) (p, t2fv_64, &desc); } #endif /* HAVE_FMA */ fftw-3.3.3/dft/simd/common/t2fv_5.c0000644000175000017500000001372612054410521013664 00000000000000/* * Copyright (c) 2003, 2007-11 Matteo Frigo * Copyright (c) 2003, 2007-11 Massachusetts Institute of Technology * * This program is free software; you can redistribute it and/or modify * it under the terms of the GNU General Public License as published by * the Free Software Foundation; either version 2 of the License, or * (at your option) any later version. * * This program is distributed in the hope that it will be useful, * but WITHOUT ANY WARRANTY; without even the implied warranty of * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the * GNU General Public License for more details. * * You should have received a copy of the GNU General Public License * along with this program; if not, write to the Free Software * Foundation, Inc., 51 Franklin Street, Fifth Floor, Boston, MA 02110-1301 USA * */ /* This file was automatically generated --- DO NOT EDIT */ /* Generated on Sun Nov 25 07:38:41 EST 2012 */ #include "codelet-dft.h" #ifdef HAVE_FMA /* Generated by: ../../../genfft/gen_twiddle_c.native -fma -reorder-insns -schedule-for-pipeline -simd -compact -variables 4 -pipeline-latency 8 -n 5 -name t2fv_5 -include t2f.h */ /* * This function contains 20 FP additions, 19 FP multiplications, * (or, 11 additions, 10 multiplications, 9 fused multiply/add), * 26 stack variables, 4 constants, and 10 memory accesses */ #include "t2f.h" static void t2fv_5(R *ri, R *ii, const R *W, stride rs, INT mb, INT me, INT ms) { DVK(KP559016994, +0.559016994374947424102293417182819058860154590); DVK(KP250000000, +0.250000000000000000000000000000000000000000000); DVK(KP618033988, +0.618033988749894848204586834365638117720309180); DVK(KP951056516, +0.951056516295153572116439333379382143405698634); { INT m; R *x; x = ri; for (m = mb, W = W + (mb * ((TWVL / VL) * 8)); m < me; m = m + VL, x = x + (VL * ms), W = W + (TWVL * 8), MAKE_VOLATILE_STRIDE(5, rs)) { V T1, T2, T9, T4, T7; T1 = LD(&(x[0]), ms, &(x[0])); T2 = LD(&(x[WS(rs, 1)]), ms, &(x[WS(rs, 1)])); T9 = LD(&(x[WS(rs, 3)]), ms, &(x[WS(rs, 1)])); T4 = LD(&(x[WS(rs, 4)]), ms, &(x[0])); T7 = LD(&(x[WS(rs, 2)]), ms, &(x[0])); { V T3, Ta, T5, T8; T3 = BYTWJ(&(W[0]), T2); Ta = BYTWJ(&(W[TWVL * 4]), T9); T5 = BYTWJ(&(W[TWVL * 6]), T4); T8 = BYTWJ(&(W[TWVL * 2]), T7); { V T6, Tg, Tb, Th; T6 = VADD(T3, T5); Tg = VSUB(T3, T5); Tb = VADD(T8, Ta); Th = VSUB(T8, Ta); { V Te, Tc, Tk, Ti, Td, Tj, Tf; Te = VSUB(T6, Tb); Tc = VADD(T6, Tb); Tk = VMUL(LDK(KP951056516), VFNMS(LDK(KP618033988), Tg, Th)); Ti = VMUL(LDK(KP951056516), VFMA(LDK(KP618033988), Th, Tg)); Td = VFNMS(LDK(KP250000000), Tc, T1); ST(&(x[0]), VADD(T1, Tc), ms, &(x[0])); Tj = VFNMS(LDK(KP559016994), Te, Td); Tf = VFMA(LDK(KP559016994), Te, Td); ST(&(x[WS(rs, 2)]), VFMAI(Tk, Tj), ms, &(x[0])); ST(&(x[WS(rs, 3)]), VFNMSI(Tk, Tj), ms, &(x[WS(rs, 1)])); ST(&(x[WS(rs, 4)]), VFMAI(Ti, Tf), ms, &(x[0])); ST(&(x[WS(rs, 1)]), VFNMSI(Ti, Tf), ms, &(x[WS(rs, 1)])); } } } } } VLEAVE(); } static const tw_instr twinstr[] = { VTW(0, 1), VTW(0, 2), VTW(0, 3), VTW(0, 4), {TW_NEXT, VL, 0} }; static const ct_desc desc = { 5, XSIMD_STRING("t2fv_5"), twinstr, &GENUS, {11, 10, 9, 0}, 0, 0, 0 }; void XSIMD(codelet_t2fv_5) (planner *p) { X(kdft_dit_register) (p, t2fv_5, &desc); } #else /* HAVE_FMA */ /* Generated by: ../../../genfft/gen_twiddle_c.native -simd -compact -variables 4 -pipeline-latency 8 -n 5 -name t2fv_5 -include t2f.h */ /* * This function contains 20 FP additions, 14 FP multiplications, * (or, 17 additions, 11 multiplications, 3 fused multiply/add), * 20 stack variables, 4 constants, and 10 memory accesses */ #include "t2f.h" static void t2fv_5(R *ri, R *ii, const R *W, stride rs, INT mb, INT me, INT ms) { DVK(KP250000000, +0.250000000000000000000000000000000000000000000); DVK(KP559016994, +0.559016994374947424102293417182819058860154590); DVK(KP587785252, +0.587785252292473129168705954639072768597652438); DVK(KP951056516, +0.951056516295153572116439333379382143405698634); { INT m; R *x; x = ri; for (m = mb, W = W + (mb * ((TWVL / VL) * 8)); m < me; m = m + VL, x = x + (VL * ms), W = W + (TWVL * 8), MAKE_VOLATILE_STRIDE(5, rs)) { V Tc, Tg, Th, T5, Ta, Td; Tc = LD(&(x[0]), ms, &(x[0])); { V T2, T9, T4, T7; { V T1, T8, T3, T6; T1 = LD(&(x[WS(rs, 1)]), ms, &(x[WS(rs, 1)])); T2 = BYTWJ(&(W[0]), T1); T8 = LD(&(x[WS(rs, 3)]), ms, &(x[WS(rs, 1)])); T9 = BYTWJ(&(W[TWVL * 4]), T8); T3 = LD(&(x[WS(rs, 4)]), ms, &(x[0])); T4 = BYTWJ(&(W[TWVL * 6]), T3); T6 = LD(&(x[WS(rs, 2)]), ms, &(x[0])); T7 = BYTWJ(&(W[TWVL * 2]), T6); } Tg = VSUB(T2, T4); Th = VSUB(T7, T9); T5 = VADD(T2, T4); Ta = VADD(T7, T9); Td = VADD(T5, Ta); } ST(&(x[0]), VADD(Tc, Td), ms, &(x[0])); { V Ti, Tj, Tf, Tk, Tb, Te; Ti = VBYI(VFMA(LDK(KP951056516), Tg, VMUL(LDK(KP587785252), Th))); Tj = VBYI(VFNMS(LDK(KP587785252), Tg, VMUL(LDK(KP951056516), Th))); Tb = VMUL(LDK(KP559016994), VSUB(T5, Ta)); Te = VFNMS(LDK(KP250000000), Td, Tc); Tf = VADD(Tb, Te); Tk = VSUB(Te, Tb); ST(&(x[WS(rs, 1)]), VSUB(Tf, Ti), ms, &(x[WS(rs, 1)])); ST(&(x[WS(rs, 3)]), VSUB(Tk, Tj), ms, &(x[WS(rs, 1)])); ST(&(x[WS(rs, 4)]), VADD(Ti, Tf), ms, &(x[0])); ST(&(x[WS(rs, 2)]), VADD(Tj, Tk), ms, &(x[0])); } } } VLEAVE(); } static const tw_instr twinstr[] = { VTW(0, 1), VTW(0, 2), VTW(0, 3), VTW(0, 4), {TW_NEXT, VL, 0} }; static const ct_desc desc = { 5, XSIMD_STRING("t2fv_5"), twinstr, &GENUS, {17, 11, 3, 0}, 0, 0, 0 }; void XSIMD(codelet_t2fv_5) (planner *p) { X(kdft_dit_register) (p, t2fv_5, &desc); } #endif /* HAVE_FMA */ fftw-3.3.3/dft/simd/common/t2fv_10.c0000644000175000017500000002254512054410521013737 00000000000000/* * Copyright (c) 2003, 2007-11 Matteo Frigo * Copyright (c) 2003, 2007-11 Massachusetts Institute of Technology * * This program is free software; you can redistribute it and/or modify * it under the terms of the GNU General Public License as published by * the Free Software Foundation; either version 2 of the License, or * (at your option) any later version. * * This program is distributed in the hope that it will be useful, * but WITHOUT ANY WARRANTY; without even the implied warranty of * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the * GNU General Public License for more details. * * You should have received a copy of the GNU General Public License * along with this program; if not, write to the Free Software * Foundation, Inc., 51 Franklin Street, Fifth Floor, Boston, MA 02110-1301 USA * */ /* This file was automatically generated --- DO NOT EDIT */ /* Generated on Sun Nov 25 07:38:41 EST 2012 */ #include "codelet-dft.h" #ifdef HAVE_FMA /* Generated by: ../../../genfft/gen_twiddle_c.native -fma -reorder-insns -schedule-for-pipeline -simd -compact -variables 4 -pipeline-latency 8 -n 10 -name t2fv_10 -include t2f.h */ /* * This function contains 51 FP additions, 40 FP multiplications, * (or, 33 additions, 22 multiplications, 18 fused multiply/add), * 43 stack variables, 4 constants, and 20 memory accesses */ #include "t2f.h" static void t2fv_10(R *ri, R *ii, const R *W, stride rs, INT mb, INT me, INT ms) { DVK(KP559016994, +0.559016994374947424102293417182819058860154590); DVK(KP250000000, +0.250000000000000000000000000000000000000000000); DVK(KP618033988, +0.618033988749894848204586834365638117720309180); DVK(KP951056516, +0.951056516295153572116439333379382143405698634); { INT m; R *x; x = ri; for (m = mb, W = W + (mb * ((TWVL / VL) * 18)); m < me; m = m + VL, x = x + (VL * ms), W = W + (TWVL * 18), MAKE_VOLATILE_STRIDE(10, rs)) { V Td, TA, T4, Ta, Tk, TE, Tp, TF, TB, T9, T1, T2, Tb; T1 = LD(&(x[0]), ms, &(x[0])); T2 = LD(&(x[WS(rs, 5)]), ms, &(x[WS(rs, 1)])); { V Tg, Tn, Ti, Tl; Tg = LD(&(x[WS(rs, 4)]), ms, &(x[0])); Tn = LD(&(x[WS(rs, 1)]), ms, &(x[WS(rs, 1)])); Ti = LD(&(x[WS(rs, 9)]), ms, &(x[WS(rs, 1)])); Tl = LD(&(x[WS(rs, 6)]), ms, &(x[0])); { V T6, T8, T5, Tc; T5 = LD(&(x[WS(rs, 2)]), ms, &(x[0])); Tc = LD(&(x[WS(rs, 3)]), ms, &(x[WS(rs, 1)])); { V T3, Th, To, Tj, Tm, T7; T7 = LD(&(x[WS(rs, 7)]), ms, &(x[WS(rs, 1)])); T3 = BYTWJ(&(W[TWVL * 8]), T2); Th = BYTWJ(&(W[TWVL * 6]), Tg); To = BYTWJ(&(W[0]), Tn); Tj = BYTWJ(&(W[TWVL * 16]), Ti); Tm = BYTWJ(&(W[TWVL * 10]), Tl); T6 = BYTWJ(&(W[TWVL * 2]), T5); Td = BYTWJ(&(W[TWVL * 4]), Tc); T8 = BYTWJ(&(W[TWVL * 12]), T7); TA = VADD(T1, T3); T4 = VSUB(T1, T3); Ta = LD(&(x[WS(rs, 8)]), ms, &(x[0])); Tk = VSUB(Th, Tj); TE = VADD(Th, Tj); Tp = VSUB(Tm, To); TF = VADD(Tm, To); } TB = VADD(T6, T8); T9 = VSUB(T6, T8); } } Tb = BYTWJ(&(W[TWVL * 14]), Ta); { V TL, TG, Tw, Tq, TC, Te; TL = VSUB(TE, TF); TG = VADD(TE, TF); Tw = VSUB(Tk, Tp); Tq = VADD(Tk, Tp); TC = VADD(Tb, Td); Te = VSUB(Tb, Td); { V TM, TD, Tv, Tf; TM = VSUB(TB, TC); TD = VADD(TB, TC); Tv = VSUB(T9, Te); Tf = VADD(T9, Te); { V TP, TN, TH, TJ, Tz, Tx, Tr, Tt, TI, Ts; TP = VMUL(LDK(KP951056516), VFMA(LDK(KP618033988), TL, TM)); TN = VMUL(LDK(KP951056516), VFNMS(LDK(KP618033988), TM, TL)); TH = VADD(TD, TG); TJ = VSUB(TD, TG); Tz = VMUL(LDK(KP951056516), VFNMS(LDK(KP618033988), Tv, Tw)); Tx = VMUL(LDK(KP951056516), VFMA(LDK(KP618033988), Tw, Tv)); Tr = VADD(Tf, Tq); Tt = VSUB(Tf, Tq); ST(&(x[0]), VADD(TA, TH), ms, &(x[0])); TI = VFNMS(LDK(KP250000000), TH, TA); ST(&(x[WS(rs, 5)]), VADD(T4, Tr), ms, &(x[WS(rs, 1)])); Ts = VFNMS(LDK(KP250000000), Tr, T4); { V TK, TO, Tu, Ty; TK = VFNMS(LDK(KP559016994), TJ, TI); TO = VFMA(LDK(KP559016994), TJ, TI); Tu = VFMA(LDK(KP559016994), Tt, Ts); Ty = VFNMS(LDK(KP559016994), Tt, Ts); ST(&(x[WS(rs, 8)]), VFNMSI(TN, TK), ms, &(x[0])); ST(&(x[WS(rs, 2)]), VFMAI(TN, TK), ms, &(x[0])); ST(&(x[WS(rs, 6)]), VFNMSI(TP, TO), ms, &(x[0])); ST(&(x[WS(rs, 4)]), VFMAI(TP, TO), ms, &(x[0])); ST(&(x[WS(rs, 9)]), VFMAI(Tx, Tu), ms, &(x[WS(rs, 1)])); ST(&(x[WS(rs, 1)]), VFNMSI(Tx, Tu), ms, &(x[WS(rs, 1)])); ST(&(x[WS(rs, 7)]), VFMAI(Tz, Ty), ms, &(x[WS(rs, 1)])); ST(&(x[WS(rs, 3)]), VFNMSI(Tz, Ty), ms, &(x[WS(rs, 1)])); } } } } } } VLEAVE(); } static const tw_instr twinstr[] = { VTW(0, 1), VTW(0, 2), VTW(0, 3), VTW(0, 4), VTW(0, 5), VTW(0, 6), VTW(0, 7), VTW(0, 8), VTW(0, 9), {TW_NEXT, VL, 0} }; static const ct_desc desc = { 10, XSIMD_STRING("t2fv_10"), twinstr, &GENUS, {33, 22, 18, 0}, 0, 0, 0 }; void XSIMD(codelet_t2fv_10) (planner *p) { X(kdft_dit_register) (p, t2fv_10, &desc); } #else /* HAVE_FMA */ /* Generated by: ../../../genfft/gen_twiddle_c.native -simd -compact -variables 4 -pipeline-latency 8 -n 10 -name t2fv_10 -include t2f.h */ /* * This function contains 51 FP additions, 30 FP multiplications, * (or, 45 additions, 24 multiplications, 6 fused multiply/add), * 32 stack variables, 4 constants, and 20 memory accesses */ #include "t2f.h" static void t2fv_10(R *ri, R *ii, const R *W, stride rs, INT mb, INT me, INT ms) { DVK(KP587785252, +0.587785252292473129168705954639072768597652438); DVK(KP951056516, +0.951056516295153572116439333379382143405698634); DVK(KP250000000, +0.250000000000000000000000000000000000000000000); DVK(KP559016994, +0.559016994374947424102293417182819058860154590); { INT m; R *x; x = ri; for (m = mb, W = W + (mb * ((TWVL / VL) * 18)); m < me; m = m + VL, x = x + (VL * ms), W = W + (TWVL * 18), MAKE_VOLATILE_STRIDE(10, rs)) { V Tr, TH, Tg, Tl, Tm, TA, TB, TJ, T5, Ta, Tb, TD, TE, TI, To; V Tq, Tp; To = LD(&(x[0]), ms, &(x[0])); Tp = LD(&(x[WS(rs, 5)]), ms, &(x[WS(rs, 1)])); Tq = BYTWJ(&(W[TWVL * 8]), Tp); Tr = VSUB(To, Tq); TH = VADD(To, Tq); { V Td, Tk, Tf, Ti; { V Tc, Tj, Te, Th; Tc = LD(&(x[WS(rs, 4)]), ms, &(x[0])); Td = BYTWJ(&(W[TWVL * 6]), Tc); Tj = LD(&(x[WS(rs, 1)]), ms, &(x[WS(rs, 1)])); Tk = BYTWJ(&(W[0]), Tj); Te = LD(&(x[WS(rs, 9)]), ms, &(x[WS(rs, 1)])); Tf = BYTWJ(&(W[TWVL * 16]), Te); Th = LD(&(x[WS(rs, 6)]), ms, &(x[0])); Ti = BYTWJ(&(W[TWVL * 10]), Th); } Tg = VSUB(Td, Tf); Tl = VSUB(Ti, Tk); Tm = VADD(Tg, Tl); TA = VADD(Td, Tf); TB = VADD(Ti, Tk); TJ = VADD(TA, TB); } { V T2, T9, T4, T7; { V T1, T8, T3, T6; T1 = LD(&(x[WS(rs, 2)]), ms, &(x[0])); T2 = BYTWJ(&(W[TWVL * 2]), T1); T8 = LD(&(x[WS(rs, 3)]), ms, &(x[WS(rs, 1)])); T9 = BYTWJ(&(W[TWVL * 4]), T8); T3 = LD(&(x[WS(rs, 7)]), ms, &(x[WS(rs, 1)])); T4 = BYTWJ(&(W[TWVL * 12]), T3); T6 = LD(&(x[WS(rs, 8)]), ms, &(x[0])); T7 = BYTWJ(&(W[TWVL * 14]), T6); } T5 = VSUB(T2, T4); Ta = VSUB(T7, T9); Tb = VADD(T5, Ta); TD = VADD(T2, T4); TE = VADD(T7, T9); TI = VADD(TD, TE); } { V Tn, Ts, Tt, Tx, Tz, Tv, Tw, Ty, Tu; Tn = VMUL(LDK(KP559016994), VSUB(Tb, Tm)); Ts = VADD(Tb, Tm); Tt = VFNMS(LDK(KP250000000), Ts, Tr); Tv = VSUB(T5, Ta); Tw = VSUB(Tg, Tl); Tx = VBYI(VFMA(LDK(KP951056516), Tv, VMUL(LDK(KP587785252), Tw))); Tz = VBYI(VFNMS(LDK(KP587785252), Tv, VMUL(LDK(KP951056516), Tw))); ST(&(x[WS(rs, 5)]), VADD(Tr, Ts), ms, &(x[WS(rs, 1)])); Ty = VSUB(Tt, Tn); ST(&(x[WS(rs, 3)]), VSUB(Ty, Tz), ms, &(x[WS(rs, 1)])); ST(&(x[WS(rs, 7)]), VADD(Tz, Ty), ms, &(x[WS(rs, 1)])); Tu = VADD(Tn, Tt); ST(&(x[WS(rs, 1)]), VSUB(Tu, Tx), ms, &(x[WS(rs, 1)])); ST(&(x[WS(rs, 9)]), VADD(Tx, Tu), ms, &(x[WS(rs, 1)])); } { V TM, TK, TL, TG, TO, TC, TF, TP, TN; TM = VMUL(LDK(KP559016994), VSUB(TI, TJ)); TK = VADD(TI, TJ); TL = VFNMS(LDK(KP250000000), TK, TH); TC = VSUB(TA, TB); TF = VSUB(TD, TE); TG = VBYI(VFNMS(LDK(KP587785252), TF, VMUL(LDK(KP951056516), TC))); TO = VBYI(VFMA(LDK(KP951056516), TF, VMUL(LDK(KP587785252), TC))); ST(&(x[0]), VADD(TH, TK), ms, &(x[0])); TP = VADD(TM, TL); ST(&(x[WS(rs, 4)]), VADD(TO, TP), ms, &(x[0])); ST(&(x[WS(rs, 6)]), VSUB(TP, TO), ms, &(x[0])); TN = VSUB(TL, TM); ST(&(x[WS(rs, 2)]), VADD(TG, TN), ms, &(x[0])); ST(&(x[WS(rs, 8)]), VSUB(TN, TG), ms, &(x[0])); } } } VLEAVE(); } static const tw_instr twinstr[] = { VTW(0, 1), VTW(0, 2), VTW(0, 3), VTW(0, 4), VTW(0, 5), VTW(0, 6), VTW(0, 7), VTW(0, 8), VTW(0, 9), {TW_NEXT, VL, 0} }; static const ct_desc desc = { 10, XSIMD_STRING("t2fv_10"), twinstr, &GENUS, {45, 24, 6, 0}, 0, 0, 0 }; void XSIMD(codelet_t2fv_10) (planner *p) { X(kdft_dit_register) (p, t2fv_10, &desc); } #endif /* HAVE_FMA */ fftw-3.3.3/dft/simd/common/t2fv_20.c0000644000175000017500000004200212054410522013727 00000000000000/* * Copyright (c) 2003, 2007-11 Matteo Frigo * Copyright (c) 2003, 2007-11 Massachusetts Institute of Technology * * This program is free software; you can redistribute it and/or modify * it under the terms of the GNU General Public License as published by * the Free Software Foundation; either version 2 of the License, or * (at your option) any later version. * * This program is distributed in the hope that it will be useful, * but WITHOUT ANY WARRANTY; without even the implied warranty of * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the * GNU General Public License for more details. * * You should have received a copy of the GNU General Public License * along with this program; if not, write to the Free Software * Foundation, Inc., 51 Franklin Street, Fifth Floor, Boston, MA 02110-1301 USA * */ /* This file was automatically generated --- DO NOT EDIT */ /* Generated on Sun Nov 25 07:38:41 EST 2012 */ #include "codelet-dft.h" #ifdef HAVE_FMA /* Generated by: ../../../genfft/gen_twiddle_c.native -fma -reorder-insns -schedule-for-pipeline -simd -compact -variables 4 -pipeline-latency 8 -n 20 -name t2fv_20 -include t2f.h */ /* * This function contains 123 FP additions, 88 FP multiplications, * (or, 77 additions, 42 multiplications, 46 fused multiply/add), * 68 stack variables, 4 constants, and 40 memory accesses */ #include "t2f.h" static void t2fv_20(R *ri, R *ii, const R *W, stride rs, INT mb, INT me, INT ms) { DVK(KP951056516, +0.951056516295153572116439333379382143405698634); DVK(KP559016994, +0.559016994374947424102293417182819058860154590); DVK(KP250000000, +0.250000000000000000000000000000000000000000000); DVK(KP618033988, +0.618033988749894848204586834365638117720309180); { INT m; R *x; x = ri; for (m = mb, W = W + (mb * ((TWVL / VL) * 38)); m < me; m = m + VL, x = x + (VL * ms), W = W + (TWVL * 38), MAKE_VOLATILE_STRIDE(20, rs)) { V T4, Tx, T1m, T1K, T1y, Tk, Tf, T16, T10, TT, T1O, T1w, T1L, T1p, T1M; V T1s, TZ, TI, T1x, Tp; { V T1, Tv, T2, Tt; T1 = LD(&(x[0]), ms, &(x[0])); Tv = LD(&(x[WS(rs, 15)]), ms, &(x[WS(rs, 1)])); T2 = LD(&(x[WS(rs, 10)]), ms, &(x[0])); Tt = LD(&(x[WS(rs, 5)]), ms, &(x[WS(rs, 1)])); { V T9, T1n, TN, T1v, TS, Te, T1q, T1u, TE, TG, Tm, T1o, TC, Tn, T1r; V TH, To; { V TP, TR, Ta, Tc; { V T5, T7, TJ, TL, T1k, T1l; T5 = LD(&(x[WS(rs, 4)]), ms, &(x[0])); T7 = LD(&(x[WS(rs, 14)]), ms, &(x[0])); TJ = LD(&(x[WS(rs, 13)]), ms, &(x[WS(rs, 1)])); TL = LD(&(x[WS(rs, 3)]), ms, &(x[WS(rs, 1)])); { V Tw, T3, Tu, T6, T8, TK, TM, TO, TQ; TO = LD(&(x[WS(rs, 17)]), ms, &(x[WS(rs, 1)])); Tw = BYTWJ(&(W[TWVL * 28]), Tv); T3 = BYTWJ(&(W[TWVL * 18]), T2); Tu = BYTWJ(&(W[TWVL * 8]), Tt); T6 = BYTWJ(&(W[TWVL * 6]), T5); T8 = BYTWJ(&(W[TWVL * 26]), T7); TK = BYTWJ(&(W[TWVL * 24]), TJ); TM = BYTWJ(&(W[TWVL * 4]), TL); TP = BYTWJ(&(W[TWVL * 32]), TO); TQ = LD(&(x[WS(rs, 7)]), ms, &(x[WS(rs, 1)])); T4 = VSUB(T1, T3); T1k = VADD(T1, T3); Tx = VSUB(Tu, Tw); T1l = VADD(Tu, Tw); T9 = VSUB(T6, T8); T1n = VADD(T6, T8); TN = VSUB(TK, TM); T1v = VADD(TK, TM); TR = BYTWJ(&(W[TWVL * 12]), TQ); } Ta = LD(&(x[WS(rs, 16)]), ms, &(x[0])); T1m = VSUB(T1k, T1l); T1K = VADD(T1k, T1l); Tc = LD(&(x[WS(rs, 6)]), ms, &(x[0])); } { V Tb, TA, Td, Th, Tj, Tz, Tg, Ti, Ty; Tg = LD(&(x[WS(rs, 8)]), ms, &(x[0])); Ti = LD(&(x[WS(rs, 18)]), ms, &(x[0])); Ty = LD(&(x[WS(rs, 9)]), ms, &(x[WS(rs, 1)])); TS = VSUB(TP, TR); T1y = VADD(TP, TR); Tb = BYTWJ(&(W[TWVL * 30]), Ta); TA = LD(&(x[WS(rs, 19)]), ms, &(x[WS(rs, 1)])); Td = BYTWJ(&(W[TWVL * 10]), Tc); Th = BYTWJ(&(W[TWVL * 14]), Tg); Tj = BYTWJ(&(W[TWVL * 34]), Ti); Tz = BYTWJ(&(W[TWVL * 16]), Ty); { V TD, TF, TB, Tl; TD = LD(&(x[WS(rs, 1)]), ms, &(x[WS(rs, 1)])); TF = LD(&(x[WS(rs, 11)]), ms, &(x[WS(rs, 1)])); Tl = LD(&(x[WS(rs, 12)]), ms, &(x[0])); TB = BYTWJ(&(W[TWVL * 36]), TA); Te = VSUB(Tb, Td); T1q = VADD(Tb, Td); Tk = VSUB(Th, Tj); T1u = VADD(Th, Tj); TE = BYTWJ(&(W[0]), TD); TG = BYTWJ(&(W[TWVL * 20]), TF); Tm = BYTWJ(&(W[TWVL * 22]), Tl); T1o = VADD(Tz, TB); TC = VSUB(Tz, TB); Tn = LD(&(x[WS(rs, 2)]), ms, &(x[0])); } } } Tf = VADD(T9, Te); T16 = VSUB(T9, Te); T10 = VSUB(TS, TN); TT = VADD(TN, TS); T1r = VADD(TE, TG); TH = VSUB(TE, TG); T1O = VADD(T1u, T1v); T1w = VSUB(T1u, T1v); To = BYTWJ(&(W[TWVL * 2]), Tn); T1L = VADD(T1n, T1o); T1p = VSUB(T1n, T1o); T1M = VADD(T1q, T1r); T1s = VSUB(T1q, T1r); TZ = VSUB(TH, TC); TI = VADD(TC, TH); T1x = VADD(Tm, To); Tp = VSUB(Tm, To); } } { V T1V, T1N, T14, T1d, T11, T1G, T1t, T1z, T1P, Tq, T17, T13, TV, TU; T1V = VSUB(T1L, T1M); T1N = VADD(T1L, T1M); T14 = VSUB(TT, TI); TU = VADD(TI, TT); T1d = VFNMS(LDK(KP618033988), TZ, T10); T11 = VFMA(LDK(KP618033988), T10, TZ); T1G = VSUB(T1p, T1s); T1t = VADD(T1p, T1s); T1z = VSUB(T1x, T1y); T1P = VADD(T1x, T1y); Tq = VADD(Tk, Tp); T17 = VSUB(Tk, Tp); T13 = VFNMS(LDK(KP250000000), TU, Tx); TV = VADD(Tx, TU); { V T1J, T1H, T1D, T1Z, T1X, T1T, T1h, T1j, T1b, T19, T1C, T1S, T1c, TY, T1F; V T1A; T1F = VSUB(T1w, T1z); T1A = VADD(T1w, T1z); { V T1W, T1Q, TX, Tr; T1W = VSUB(T1O, T1P); T1Q = VADD(T1O, T1P); TX = VSUB(Tf, Tq); Tr = VADD(Tf, Tq); { V T1g, T18, T1f, T15; T1g = VFNMS(LDK(KP618033988), T16, T17); T18 = VFMA(LDK(KP618033988), T17, T16); T1f = VFMA(LDK(KP559016994), T14, T13); T15 = VFNMS(LDK(KP559016994), T14, T13); T1J = VMUL(LDK(KP951056516), VFMA(LDK(KP618033988), T1F, T1G)); T1H = VMUL(LDK(KP951056516), VFNMS(LDK(KP618033988), T1G, T1F)); { V T1B, T1R, TW, Ts; T1B = VADD(T1t, T1A); T1D = VSUB(T1t, T1A); T1Z = VMUL(LDK(KP951056516), VFNMS(LDK(KP618033988), T1V, T1W)); T1X = VMUL(LDK(KP951056516), VFMA(LDK(KP618033988), T1W, T1V)); T1R = VADD(T1N, T1Q); T1T = VSUB(T1N, T1Q); TW = VFNMS(LDK(KP250000000), Tr, T4); Ts = VADD(T4, Tr); T1h = VFNMS(LDK(KP951056516), T1g, T1f); T1j = VFMA(LDK(KP951056516), T1g, T1f); T1b = VFNMS(LDK(KP951056516), T18, T15); T19 = VFMA(LDK(KP951056516), T18, T15); ST(&(x[WS(rs, 10)]), VADD(T1m, T1B), ms, &(x[0])); T1C = VFNMS(LDK(KP250000000), T1B, T1m); ST(&(x[0]), VADD(T1K, T1R), ms, &(x[0])); T1S = VFNMS(LDK(KP250000000), T1R, T1K); T1c = VFNMS(LDK(KP559016994), TX, TW); TY = VFMA(LDK(KP559016994), TX, TW); ST(&(x[WS(rs, 15)]), VFMAI(TV, Ts), ms, &(x[WS(rs, 1)])); ST(&(x[WS(rs, 5)]), VFNMSI(TV, Ts), ms, &(x[WS(rs, 1)])); } } } { V T1E, T1I, T1U, T1Y; T1E = VFNMS(LDK(KP559016994), T1D, T1C); T1I = VFMA(LDK(KP559016994), T1D, T1C); T1U = VFMA(LDK(KP559016994), T1T, T1S); T1Y = VFNMS(LDK(KP559016994), T1T, T1S); { V T1e, T1i, T1a, T12; T1e = VFNMS(LDK(KP951056516), T1d, T1c); T1i = VFMA(LDK(KP951056516), T1d, T1c); T1a = VFNMS(LDK(KP951056516), T11, TY); T12 = VFMA(LDK(KP951056516), T11, TY); ST(&(x[WS(rs, 18)]), VFNMSI(T1H, T1E), ms, &(x[0])); ST(&(x[WS(rs, 2)]), VFMAI(T1H, T1E), ms, &(x[0])); ST(&(x[WS(rs, 14)]), VFMAI(T1J, T1I), ms, &(x[0])); ST(&(x[WS(rs, 6)]), VFNMSI(T1J, T1I), ms, &(x[0])); ST(&(x[WS(rs, 16)]), VFNMSI(T1X, T1U), ms, &(x[0])); ST(&(x[WS(rs, 4)]), VFMAI(T1X, T1U), ms, &(x[0])); ST(&(x[WS(rs, 12)]), VFMAI(T1Z, T1Y), ms, &(x[0])); ST(&(x[WS(rs, 8)]), VFNMSI(T1Z, T1Y), ms, &(x[0])); ST(&(x[WS(rs, 3)]), VFMAI(T1h, T1e), ms, &(x[WS(rs, 1)])); ST(&(x[WS(rs, 17)]), VFNMSI(T1h, T1e), ms, &(x[WS(rs, 1)])); ST(&(x[WS(rs, 7)]), VFMAI(T1j, T1i), ms, &(x[WS(rs, 1)])); ST(&(x[WS(rs, 13)]), VFNMSI(T1j, T1i), ms, &(x[WS(rs, 1)])); ST(&(x[WS(rs, 11)]), VFMAI(T1b, T1a), ms, &(x[WS(rs, 1)])); ST(&(x[WS(rs, 9)]), VFNMSI(T1b, T1a), ms, &(x[WS(rs, 1)])); ST(&(x[WS(rs, 19)]), VFMAI(T19, T12), ms, &(x[WS(rs, 1)])); ST(&(x[WS(rs, 1)]), VFNMSI(T19, T12), ms, &(x[WS(rs, 1)])); } } } } } } VLEAVE(); } static const tw_instr twinstr[] = { VTW(0, 1), VTW(0, 2), VTW(0, 3), VTW(0, 4), VTW(0, 5), VTW(0, 6), VTW(0, 7), VTW(0, 8), VTW(0, 9), VTW(0, 10), VTW(0, 11), VTW(0, 12), VTW(0, 13), VTW(0, 14), VTW(0, 15), VTW(0, 16), VTW(0, 17), VTW(0, 18), VTW(0, 19), {TW_NEXT, VL, 0} }; static const ct_desc desc = { 20, XSIMD_STRING("t2fv_20"), twinstr, &GENUS, {77, 42, 46, 0}, 0, 0, 0 }; void XSIMD(codelet_t2fv_20) (planner *p) { X(kdft_dit_register) (p, t2fv_20, &desc); } #else /* HAVE_FMA */ /* Generated by: ../../../genfft/gen_twiddle_c.native -simd -compact -variables 4 -pipeline-latency 8 -n 20 -name t2fv_20 -include t2f.h */ /* * This function contains 123 FP additions, 62 FP multiplications, * (or, 111 additions, 50 multiplications, 12 fused multiply/add), * 54 stack variables, 4 constants, and 40 memory accesses */ #include "t2f.h" static void t2fv_20(R *ri, R *ii, const R *W, stride rs, INT mb, INT me, INT ms) { DVK(KP587785252, +0.587785252292473129168705954639072768597652438); DVK(KP951056516, +0.951056516295153572116439333379382143405698634); DVK(KP250000000, +0.250000000000000000000000000000000000000000000); DVK(KP559016994, +0.559016994374947424102293417182819058860154590); { INT m; R *x; x = ri; for (m = mb, W = W + (mb * ((TWVL / VL) * 38)); m < me; m = m + VL, x = x + (VL * ms), W = W + (TWVL * 38), MAKE_VOLATILE_STRIDE(20, rs)) { V T4, Tx, T1B, T1U, TZ, T16, T17, T10, Tf, Tq, Tr, T1N, T1O, T1S, T1t; V T1w, T1C, TI, TT, TU, T1K, T1L, T1R, T1m, T1p, T1D, Ts, TV; { V T1, Tw, T3, Tu, Tv, T2, Tt, T1z, T1A; T1 = LD(&(x[0]), ms, &(x[0])); Tv = LD(&(x[WS(rs, 15)]), ms, &(x[WS(rs, 1)])); Tw = BYTWJ(&(W[TWVL * 28]), Tv); T2 = LD(&(x[WS(rs, 10)]), ms, &(x[0])); T3 = BYTWJ(&(W[TWVL * 18]), T2); Tt = LD(&(x[WS(rs, 5)]), ms, &(x[WS(rs, 1)])); Tu = BYTWJ(&(W[TWVL * 8]), Tt); T4 = VSUB(T1, T3); Tx = VSUB(Tu, Tw); T1z = VADD(T1, T3); T1A = VADD(Tu, Tw); T1B = VSUB(T1z, T1A); T1U = VADD(T1z, T1A); } { V T9, T1r, TN, T1l, TS, T1o, Te, T1u, Tk, T1k, TC, T1s, TH, T1v, Tp; V T1n; { V T6, T8, T5, T7; T5 = LD(&(x[WS(rs, 4)]), ms, &(x[0])); T6 = BYTWJ(&(W[TWVL * 6]), T5); T7 = LD(&(x[WS(rs, 14)]), ms, &(x[0])); T8 = BYTWJ(&(W[TWVL * 26]), T7); T9 = VSUB(T6, T8); T1r = VADD(T6, T8); } { V TK, TM, TJ, TL; TJ = LD(&(x[WS(rs, 13)]), ms, &(x[WS(rs, 1)])); TK = BYTWJ(&(W[TWVL * 24]), TJ); TL = LD(&(x[WS(rs, 3)]), ms, &(x[WS(rs, 1)])); TM = BYTWJ(&(W[TWVL * 4]), TL); TN = VSUB(TK, TM); T1l = VADD(TK, TM); } { V TP, TR, TO, TQ; TO = LD(&(x[WS(rs, 17)]), ms, &(x[WS(rs, 1)])); TP = BYTWJ(&(W[TWVL * 32]), TO); TQ = LD(&(x[WS(rs, 7)]), ms, &(x[WS(rs, 1)])); TR = BYTWJ(&(W[TWVL * 12]), TQ); TS = VSUB(TP, TR); T1o = VADD(TP, TR); } { V Tb, Td, Ta, Tc; Ta = LD(&(x[WS(rs, 16)]), ms, &(x[0])); Tb = BYTWJ(&(W[TWVL * 30]), Ta); Tc = LD(&(x[WS(rs, 6)]), ms, &(x[0])); Td = BYTWJ(&(W[TWVL * 10]), Tc); Te = VSUB(Tb, Td); T1u = VADD(Tb, Td); } { V Th, Tj, Tg, Ti; Tg = LD(&(x[WS(rs, 8)]), ms, &(x[0])); Th = BYTWJ(&(W[TWVL * 14]), Tg); Ti = LD(&(x[WS(rs, 18)]), ms, &(x[0])); Tj = BYTWJ(&(W[TWVL * 34]), Ti); Tk = VSUB(Th, Tj); T1k = VADD(Th, Tj); } { V Tz, TB, Ty, TA; Ty = LD(&(x[WS(rs, 9)]), ms, &(x[WS(rs, 1)])); Tz = BYTWJ(&(W[TWVL * 16]), Ty); TA = LD(&(x[WS(rs, 19)]), ms, &(x[WS(rs, 1)])); TB = BYTWJ(&(W[TWVL * 36]), TA); TC = VSUB(Tz, TB); T1s = VADD(Tz, TB); } { V TE, TG, TD, TF; TD = LD(&(x[WS(rs, 1)]), ms, &(x[WS(rs, 1)])); TE = BYTWJ(&(W[0]), TD); TF = LD(&(x[WS(rs, 11)]), ms, &(x[WS(rs, 1)])); TG = BYTWJ(&(W[TWVL * 20]), TF); TH = VSUB(TE, TG); T1v = VADD(TE, TG); } { V Tm, To, Tl, Tn; Tl = LD(&(x[WS(rs, 12)]), ms, &(x[0])); Tm = BYTWJ(&(W[TWVL * 22]), Tl); Tn = LD(&(x[WS(rs, 2)]), ms, &(x[0])); To = BYTWJ(&(W[TWVL * 2]), Tn); Tp = VSUB(Tm, To); T1n = VADD(Tm, To); } TZ = VSUB(TH, TC); T16 = VSUB(T9, Te); T17 = VSUB(Tk, Tp); T10 = VSUB(TS, TN); Tf = VADD(T9, Te); Tq = VADD(Tk, Tp); Tr = VADD(Tf, Tq); T1N = VADD(T1k, T1l); T1O = VADD(T1n, T1o); T1S = VADD(T1N, T1O); T1t = VSUB(T1r, T1s); T1w = VSUB(T1u, T1v); T1C = VADD(T1t, T1w); TI = VADD(TC, TH); TT = VADD(TN, TS); TU = VADD(TI, TT); T1K = VADD(T1r, T1s); T1L = VADD(T1u, T1v); T1R = VADD(T1K, T1L); T1m = VSUB(T1k, T1l); T1p = VSUB(T1n, T1o); T1D = VADD(T1m, T1p); } Ts = VADD(T4, Tr); TV = VBYI(VADD(Tx, TU)); ST(&(x[WS(rs, 5)]), VSUB(Ts, TV), ms, &(x[WS(rs, 1)])); ST(&(x[WS(rs, 15)]), VADD(Ts, TV), ms, &(x[WS(rs, 1)])); { V T1T, T1V, T1W, T1Q, T1Z, T1M, T1P, T1Y, T1X; T1T = VMUL(LDK(KP559016994), VSUB(T1R, T1S)); T1V = VADD(T1R, T1S); T1W = VFNMS(LDK(KP250000000), T1V, T1U); T1M = VSUB(T1K, T1L); T1P = VSUB(T1N, T1O); T1Q = VBYI(VFMA(LDK(KP951056516), T1M, VMUL(LDK(KP587785252), T1P))); T1Z = VBYI(VFNMS(LDK(KP587785252), T1M, VMUL(LDK(KP951056516), T1P))); ST(&(x[0]), VADD(T1U, T1V), ms, &(x[0])); T1Y = VSUB(T1W, T1T); ST(&(x[WS(rs, 8)]), VSUB(T1Y, T1Z), ms, &(x[0])); ST(&(x[WS(rs, 12)]), VADD(T1Z, T1Y), ms, &(x[0])); T1X = VADD(T1T, T1W); ST(&(x[WS(rs, 4)]), VADD(T1Q, T1X), ms, &(x[0])); ST(&(x[WS(rs, 16)]), VSUB(T1X, T1Q), ms, &(x[0])); } { V T1G, T1E, T1F, T1y, T1J, T1q, T1x, T1I, T1H; T1G = VMUL(LDK(KP559016994), VSUB(T1C, T1D)); T1E = VADD(T1C, T1D); T1F = VFNMS(LDK(KP250000000), T1E, T1B); T1q = VSUB(T1m, T1p); T1x = VSUB(T1t, T1w); T1y = VBYI(VFNMS(LDK(KP587785252), T1x, VMUL(LDK(KP951056516), T1q))); T1J = VBYI(VFMA(LDK(KP951056516), T1x, VMUL(LDK(KP587785252), T1q))); ST(&(x[WS(rs, 10)]), VADD(T1B, T1E), ms, &(x[0])); T1I = VADD(T1G, T1F); ST(&(x[WS(rs, 6)]), VSUB(T1I, T1J), ms, &(x[0])); ST(&(x[WS(rs, 14)]), VADD(T1J, T1I), ms, &(x[0])); T1H = VSUB(T1F, T1G); ST(&(x[WS(rs, 2)]), VADD(T1y, T1H), ms, &(x[0])); ST(&(x[WS(rs, 18)]), VSUB(T1H, T1y), ms, &(x[0])); } { V T11, T18, T1g, T1d, T15, T1f, TY, T1c; T11 = VFMA(LDK(KP951056516), TZ, VMUL(LDK(KP587785252), T10)); T18 = VFMA(LDK(KP951056516), T16, VMUL(LDK(KP587785252), T17)); T1g = VFNMS(LDK(KP587785252), T16, VMUL(LDK(KP951056516), T17)); T1d = VFNMS(LDK(KP587785252), TZ, VMUL(LDK(KP951056516), T10)); { V T13, T14, TW, TX; T13 = VFMS(LDK(KP250000000), TU, Tx); T14 = VMUL(LDK(KP559016994), VSUB(TT, TI)); T15 = VADD(T13, T14); T1f = VSUB(T14, T13); TW = VMUL(LDK(KP559016994), VSUB(Tf, Tq)); TX = VFNMS(LDK(KP250000000), Tr, T4); TY = VADD(TW, TX); T1c = VSUB(TX, TW); } { V T12, T19, T1i, T1j; T12 = VADD(TY, T11); T19 = VBYI(VSUB(T15, T18)); ST(&(x[WS(rs, 19)]), VSUB(T12, T19), ms, &(x[WS(rs, 1)])); ST(&(x[WS(rs, 1)]), VADD(T12, T19), ms, &(x[WS(rs, 1)])); T1i = VADD(T1c, T1d); T1j = VBYI(VADD(T1g, T1f)); ST(&(x[WS(rs, 13)]), VSUB(T1i, T1j), ms, &(x[WS(rs, 1)])); ST(&(x[WS(rs, 7)]), VADD(T1i, T1j), ms, &(x[WS(rs, 1)])); } { V T1a, T1b, T1e, T1h; T1a = VSUB(TY, T11); T1b = VBYI(VADD(T18, T15)); ST(&(x[WS(rs, 11)]), VSUB(T1a, T1b), ms, &(x[WS(rs, 1)])); ST(&(x[WS(rs, 9)]), VADD(T1a, T1b), ms, &(x[WS(rs, 1)])); T1e = VSUB(T1c, T1d); T1h = VBYI(VSUB(T1f, T1g)); ST(&(x[WS(rs, 17)]), VSUB(T1e, T1h), ms, &(x[WS(rs, 1)])); ST(&(x[WS(rs, 3)]), VADD(T1e, T1h), ms, &(x[WS(rs, 1)])); } } } } VLEAVE(); } static const tw_instr twinstr[] = { VTW(0, 1), VTW(0, 2), VTW(0, 3), VTW(0, 4), VTW(0, 5), VTW(0, 6), VTW(0, 7), VTW(0, 8), VTW(0, 9), VTW(0, 10), VTW(0, 11), VTW(0, 12), VTW(0, 13), VTW(0, 14), VTW(0, 15), VTW(0, 16), VTW(0, 17), VTW(0, 18), VTW(0, 19), {TW_NEXT, VL, 0} }; static const ct_desc desc = { 20, XSIMD_STRING("t2fv_20"), twinstr, &GENUS, {111, 50, 12, 0}, 0, 0, 0 }; void XSIMD(codelet_t2fv_20) (planner *p) { X(kdft_dit_register) (p, t2fv_20, &desc); } #endif /* HAVE_FMA */ fftw-3.3.3/dft/simd/common/t2fv_25.c0000644000175000017500000011471012054410531013742 00000000000000/* * Copyright (c) 2003, 2007-11 Matteo Frigo * Copyright (c) 2003, 2007-11 Massachusetts Institute of Technology * * This program is free software; you can redistribute it and/or modify * it under the terms of the GNU General Public License as published by * the Free Software Foundation; either version 2 of the License, or * (at your option) any later version. * * This program is distributed in the hope that it will be useful, * but WITHOUT ANY WARRANTY; without even the implied warranty of * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the * GNU General Public License for more details. * * You should have received a copy of the GNU General Public License * along with this program; if not, write to the Free Software * Foundation, Inc., 51 Franklin Street, Fifth Floor, Boston, MA 02110-1301 USA * */ /* This file was automatically generated --- DO NOT EDIT */ /* Generated on Sun Nov 25 07:38:42 EST 2012 */ #include "codelet-dft.h" #ifdef HAVE_FMA /* Generated by: ../../../genfft/gen_twiddle_c.native -fma -reorder-insns -schedule-for-pipeline -simd -compact -variables 4 -pipeline-latency 8 -n 25 -name t2fv_25 -include t2f.h */ /* * This function contains 248 FP additions, 241 FP multiplications, * (or, 67 additions, 60 multiplications, 181 fused multiply/add), * 208 stack variables, 67 constants, and 50 memory accesses */ #include "t2f.h" static void t2fv_25(R *ri, R *ii, const R *W, stride rs, INT mb, INT me, INT ms) { DVK(KP792626838, +0.792626838241819413632131824093538848057784557); DVK(KP876091699, +0.876091699473550838204498029706869638173524346); DVK(KP617882369, +0.617882369114440893914546919006756321695042882); DVK(KP803003575, +0.803003575438660414833440593570376004635464850); DVK(KP242145790, +0.242145790282157779872542093866183953459003101); DVK(KP968583161, +0.968583161128631119490168375464735813836012403); DVK(KP999544308, +0.999544308746292983948881682379742149196758193); DVK(KP916574801, +0.916574801383451584742370439148878693530976769); DVK(KP904730450, +0.904730450839922351881287709692877908104763647); DVK(KP809385824, +0.809385824416008241660603814668679683846476688); DVK(KP447417479, +0.447417479732227551498980015410057305749330693); DVK(KP894834959, +0.894834959464455102997960030820114611498661386); DVK(KP867381224, +0.867381224396525206773171885031575671309956167); DVK(KP683113946, +0.683113946453479238701949862233725244439656928); DVK(KP559154169, +0.559154169276087864842202529084232643714075927); DVK(KP958953096, +0.958953096729998668045963838399037225970891871); DVK(KP831864738, +0.831864738706457140726048799369896829771167132); DVK(KP829049696, +0.829049696159252993975487806364305442437946767); DVK(KP860541664, +0.860541664367944677098261680920518816412804187); DVK(KP897376177, +0.897376177523557693138608077137219684419427330); DVK(KP876306680, +0.876306680043863587308115903922062583399064238); DVK(KP681693190, +0.681693190061530575150324149145440022633095390); DVK(KP560319534, +0.560319534973832390111614715371676131169633784); DVK(KP855719849, +0.855719849902058969314654733608091555096772472); DVK(KP237294955, +0.237294955877110315393888866460840817927895961); DVK(KP949179823, +0.949179823508441261575555465843363271711583843); DVK(KP904508497, +0.904508497187473712051146708591409529430077295); DVK(KP997675361, +0.997675361079556513670859573984492383596555031); DVK(KP763932022, +0.763932022500210303590826331268723764559381640); DVK(KP690983005, +0.690983005625052575897706582817180941139845410); DVK(KP992114701, +0.992114701314477831049793042785778521453036709); DVK(KP952936919, +0.952936919628306576880750665357914584765951388); DVK(KP998026728, +0.998026728428271561952336806863450553336905220); DVK(KP262346850, +0.262346850930607871785420028382979691334784273); DVK(KP570584518, +0.570584518783621657366766175430996792655723863); DVK(KP669429328, +0.669429328479476605641803240971985825917022098); DVK(KP923225144, +0.923225144846402650453449441572664695995209956); DVK(KP945422727, +0.945422727388575946270360266328811958657216298); DVK(KP522616830, +0.522616830205754336872861364785224694908468440); DVK(KP956723877, +0.956723877038460305821989399535483155872969262); DVK(KP906616052, +0.906616052148196230441134447086066874408359177); DVK(KP772036680, +0.772036680810363904029489473607579825330539880); DVK(KP845997307, +0.845997307939530944175097360758058292389769300); DVK(KP951056516, +0.951056516295153572116439333379382143405698634); DVK(KP921078979, +0.921078979742360627699756128143719920817673854); DVK(KP912575812, +0.912575812670962425556968549836277086778922727); DVK(KP982009705, +0.982009705009746369461829878184175962711969869); DVK(KP734762448, +0.734762448793050413546343770063151342619912334); DVK(KP494780565, +0.494780565770515410344588413655324772219443730); DVK(KP447533225, +0.447533225982656890041886979663652563063114397); DVK(KP269969613, +0.269969613759572083574752974412347470060951301); DVK(KP244189809, +0.244189809627953270309879511234821255780225091); DVK(KP667278218, +0.667278218140296670899089292254759909713898805); DVK(KP603558818, +0.603558818296015001454675132653458027918768137); DVK(KP522847744, +0.522847744331509716623755382187077770911012542); DVK(KP578046249, +0.578046249379945007321754579646815604023525655); DVK(KP987388751, +0.987388751065621252324603216482382109400433949); DVK(KP893101515, +0.893101515366181661711202267938416198338079437); DVK(KP120146378, +0.120146378570687701782758537356596213647956445); DVK(KP132830569, +0.132830569247582714407653942074819768844536507); DVK(KP869845200, +0.869845200362138853122720822420327157933056305); DVK(KP786782374, +0.786782374965295178365099601674911834788448471); DVK(KP066152395, +0.066152395967733048213034281011006031460903353); DVK(KP059835404, +0.059835404262124915169548397419498386427871950); DVK(KP559016994, +0.559016994374947424102293417182819058860154590); DVK(KP250000000, +0.250000000000000000000000000000000000000000000); DVK(KP618033988, +0.618033988749894848204586834365638117720309180); { INT m; R *x; x = ri; for (m = mb, W = W + (mb * ((TWVL / VL) * 48)); m < me; m = m + VL, x = x + (VL * ms), W = W + (TWVL * 48), MAKE_VOLATILE_STRIDE(25, rs)) { V T25, T1B, T2y, T1K, T2s, T23, T1S, T26, T20, T1X; { V T1O, T2X, Te, T3L, Td, T3Q, T3j, T3b, T2R, T2M, T2f, T27, T1y, T1H, T3M; V TW, TR, TK, T2B, T3n, T3e, T2U, T2F, T2i, T2a, Tz, T1C, T3N, TQ, T11; V T1b, T1c, T16; { V T1, T1g, T1i, T1p, T1k, T1m, Tb, T1N, T6, T1M; { V T7, T9, T2, T4, T1f, T1h, T1o; T1 = LD(&(x[0]), ms, &(x[0])); T7 = LD(&(x[WS(rs, 10)]), ms, &(x[0])); T9 = LD(&(x[WS(rs, 15)]), ms, &(x[WS(rs, 1)])); T2 = LD(&(x[WS(rs, 5)]), ms, &(x[WS(rs, 1)])); T4 = LD(&(x[WS(rs, 20)]), ms, &(x[0])); T1f = LD(&(x[WS(rs, 3)]), ms, &(x[WS(rs, 1)])); T1h = LD(&(x[WS(rs, 8)]), ms, &(x[0])); T1o = LD(&(x[WS(rs, 18)]), ms, &(x[0])); { V T8, Ta, T3, T5, T1j; T1j = LD(&(x[WS(rs, 23)]), ms, &(x[WS(rs, 1)])); T8 = BYTWJ(&(W[TWVL * 18]), T7); Ta = BYTWJ(&(W[TWVL * 28]), T9); T3 = BYTWJ(&(W[TWVL * 8]), T2); T5 = BYTWJ(&(W[TWVL * 38]), T4); T1g = BYTWJ(&(W[TWVL * 4]), T1f); T1i = BYTWJ(&(W[TWVL * 14]), T1h); T1p = BYTWJ(&(W[TWVL * 34]), T1o); T1k = BYTWJ(&(W[TWVL * 44]), T1j); T1m = LD(&(x[WS(rs, 13)]), ms, &(x[WS(rs, 1)])); Tb = VADD(T8, Ta); T1N = VSUB(T8, Ta); T6 = VADD(T3, T5); T1M = VSUB(T3, T5); } } { V T1v, T1l, Th, Tj, T1w, T1q, Tq, Tk, Tn, Tg; Tg = LD(&(x[WS(rs, 1)]), ms, &(x[WS(rs, 1)])); { V Tc, Ti, T1n, Tp; Ti = LD(&(x[WS(rs, 6)]), ms, &(x[0])); T1v = VSUB(T1i, T1k); T1l = VADD(T1i, T1k); T1n = BYTWJ(&(W[TWVL * 24]), T1m); Tp = LD(&(x[WS(rs, 16)]), ms, &(x[0])); T1O = VFMA(LDK(KP618033988), T1N, T1M); T2X = VFNMS(LDK(KP618033988), T1M, T1N); Te = VSUB(T6, Tb); Tc = VADD(T6, Tb); Th = BYTWJ(&(W[0]), Tg); Tj = BYTWJ(&(W[TWVL * 10]), Ti); T1w = VSUB(T1n, T1p); T1q = VADD(T1n, T1p); Tq = BYTWJ(&(W[TWVL * 30]), Tp); Tk = LD(&(x[WS(rs, 21)]), ms, &(x[WS(rs, 1)])); T3L = VADD(T1, Tc); Td = VFNMS(LDK(KP250000000), Tc, T1); Tn = LD(&(x[WS(rs, 11)]), ms, &(x[WS(rs, 1)])); } { V T1x, T2K, TM, TB, Tw, Tm, Tx, Tr, TI, T2L, T1u, TD, TF, TL; TL = LD(&(x[WS(rs, 4)]), ms, &(x[0])); { V T1t, Tl, To, TH, T1s, T1r, TA, TC; TA = LD(&(x[WS(rs, 24)]), ms, &(x[0])); T1r = VADD(T1l, T1q); T1t = VSUB(T1q, T1l); T1x = VFMA(LDK(KP618033988), T1w, T1v); T2K = VFNMS(LDK(KP618033988), T1v, T1w); Tl = BYTWJ(&(W[TWVL * 40]), Tk); To = BYTWJ(&(W[TWVL * 20]), Tn); TM = BYTWJ(&(W[TWVL * 6]), TL); TB = BYTWJ(&(W[TWVL * 46]), TA); TH = LD(&(x[WS(rs, 14)]), ms, &(x[0])); T1s = VFNMS(LDK(KP250000000), T1r, T1g); T3Q = VADD(T1g, T1r); TC = LD(&(x[WS(rs, 9)]), ms, &(x[WS(rs, 1)])); Tw = VSUB(Tj, Tl); Tm = VADD(Tj, Tl); Tx = VSUB(Tq, To); Tr = VADD(To, Tq); TI = BYTWJ(&(W[TWVL * 26]), TH); T2L = VFMA(LDK(KP559016994), T1t, T1s); T1u = VFNMS(LDK(KP559016994), T1t, T1s); TD = BYTWJ(&(W[TWVL * 16]), TC); TF = LD(&(x[WS(rs, 19)]), ms, &(x[WS(rs, 1)])); } { V Tu, Ty, T2E, TE, TN, TG, Tt, TV, Ts; TV = LD(&(x[WS(rs, 2)]), ms, &(x[0])); Ts = VADD(Tm, Tr); Tu = VSUB(Tm, Tr); Ty = VFNMS(LDK(KP618033988), Tx, Tw); T2E = VFMA(LDK(KP618033988), Tw, Tx); T3j = VFNMS(LDK(KP059835404), T2K, T2L); T3b = VFMA(LDK(KP066152395), T2L, T2K); T2R = VFNMS(LDK(KP786782374), T2K, T2L); T2M = VFMA(LDK(KP869845200), T2L, T2K); T2f = VFMA(LDK(KP132830569), T1u, T1x); T27 = VFNMS(LDK(KP120146378), T1x, T1u); T1y = VFNMS(LDK(KP893101515), T1x, T1u); T1H = VFMA(LDK(KP987388751), T1u, T1x); TE = VSUB(TB, TD); TN = VADD(TD, TB); TG = BYTWJ(&(W[TWVL * 36]), TF); Tt = VFNMS(LDK(KP250000000), Ts, Th); T3M = VADD(Th, Ts); TW = BYTWJ(&(W[TWVL * 2]), TV); { V TJ, TO, Tv, T2D, TY, T15, T10, T13, TP; { V TX, T14, TZ, T12; TX = LD(&(x[WS(rs, 7)]), ms, &(x[WS(rs, 1)])); T14 = LD(&(x[WS(rs, 17)]), ms, &(x[WS(rs, 1)])); TZ = LD(&(x[WS(rs, 22)]), ms, &(x[0])); T12 = LD(&(x[WS(rs, 12)]), ms, &(x[0])); TJ = VSUB(TG, TI); TO = VADD(TI, TG); Tv = VFMA(LDK(KP559016994), Tu, Tt); T2D = VFNMS(LDK(KP559016994), Tu, Tt); TY = BYTWJ(&(W[TWVL * 12]), TX); T15 = BYTWJ(&(W[TWVL * 32]), T14); T10 = BYTWJ(&(W[TWVL * 42]), TZ); T13 = BYTWJ(&(W[TWVL * 22]), T12); } TP = VADD(TN, TO); TR = VSUB(TN, TO); TK = VFMA(LDK(KP618033988), TJ, TE); T2B = VFNMS(LDK(KP618033988), TE, TJ); T3n = VFMA(LDK(KP578046249), T2D, T2E); T3e = VFNMS(LDK(KP522847744), T2E, T2D); T2U = VFNMS(LDK(KP987388751), T2D, T2E); T2F = VFMA(LDK(KP893101515), T2E, T2D); T2i = VFNMS(LDK(KP603558818), Ty, Tv); T2a = VFMA(LDK(KP667278218), Tv, Ty); Tz = VFNMS(LDK(KP244189809), Ty, Tv); T1C = VFMA(LDK(KP269969613), Tv, Ty); T3N = VADD(TM, TP); TQ = VFMS(LDK(KP250000000), TP, TM); T11 = VADD(TY, T10); T1b = VSUB(TY, T10); T1c = VSUB(T15, T13); T16 = VADD(T13, T15); } } } } } { V T2z, Tf, T3W, T3O, T1d, T2H, T3m, T2j, T2b, TT, T1D, T2G, T35, T2V, T2Z; V T3A, T3g, T2I, T1a, T3R, T3X; T2z = VFNMS(LDK(KP559016994), Te, Td); Tf = VFMA(LDK(KP559016994), Te, Td); { V TS, T2A, T17, T19; TS = VFNMS(LDK(KP559016994), TR, TQ); T2A = VFMA(LDK(KP559016994), TR, TQ); T3W = VSUB(T3M, T3N); T3O = VADD(T3M, T3N); T1d = VFNMS(LDK(KP618033988), T1c, T1b); T2H = VFMA(LDK(KP618033988), T1b, T1c); T17 = VADD(T11, T16); T19 = VSUB(T16, T11); { V T3f, T2T, T2C, T18, T3P; T3m = VFMA(LDK(KP447533225), T2B, T2A); T3f = VFNMS(LDK(KP494780565), T2A, T2B); T2T = VFNMS(LDK(KP132830569), T2A, T2B); T2C = VFMA(LDK(KP120146378), T2B, T2A); T2j = VFNMS(LDK(KP786782374), TK, TS); T2b = VFMA(LDK(KP869845200), TS, TK); TT = VFNMS(LDK(KP667278218), TS, TK); T1D = VFMA(LDK(KP603558818), TK, TS); T18 = VFNMS(LDK(KP250000000), T17, TW); T3P = VADD(TW, T17); T2G = VFMA(LDK(KP734762448), T2F, T2C); T35 = VFNMS(LDK(KP734762448), T2F, T2C); T2V = VFNMS(LDK(KP734762448), T2U, T2T); T2Z = VFMA(LDK(KP734762448), T2U, T2T); T3A = VFMA(LDK(KP982009705), T3f, T3e); T3g = VFNMS(LDK(KP982009705), T3f, T3e); T2I = VFMA(LDK(KP559016994), T19, T18); T1a = VFNMS(LDK(KP559016994), T19, T18); T3R = VADD(T3P, T3Q); T3X = VSUB(T3P, T3Q); } } { V T2n, T2t, T1V, T22, T2l, T2d, T1Q, T1I, T2w, T1A, T1F, T2q; { V T2k, T1G, T28, T2g, T3K, T3E, T3a, T34, T3x, T3H, T2c, TU, T1T, T1U, T1z; V T3o, T3t; T2n = VFNMS(LDK(KP912575812), T2j, T2i); T2k = VFMA(LDK(KP912575812), T2j, T2i); T3o = VFNMS(LDK(KP921078979), T3n, T3m); T3t = VFMA(LDK(KP921078979), T3n, T3m); { V T3c, T2Q, T2J, T3k, T1e; T3c = VFNMS(LDK(KP667278218), T2I, T2H); T2Q = VFNMS(LDK(KP059835404), T2H, T2I); T2J = VFMA(LDK(KP066152395), T2I, T2H); T3k = VFMA(LDK(KP603558818), T2H, T2I); T1G = VFMA(LDK(KP578046249), T1a, T1d); T1e = VFNMS(LDK(KP522847744), T1d, T1a); T28 = VFNMS(LDK(KP494780565), T1a, T1d); T2g = VFMA(LDK(KP447533225), T1d, T1a); { V T3U, T3S, T40, T3Y; T3U = VSUB(T3O, T3R); T3S = VADD(T3O, T3R); T40 = VMUL(LDK(KP951056516), VFNMS(LDK(KP618033988), T3W, T3X)); T3Y = VMUL(LDK(KP951056516), VFMA(LDK(KP618033988), T3X, T3W)); { V T3s, T3l, T2N, T36; T3s = VFNMS(LDK(KP845997307), T3k, T3j); T3l = VFMA(LDK(KP845997307), T3k, T3j); T2N = VFNMS(LDK(KP772036680), T2M, T2J); T36 = VFMA(LDK(KP772036680), T2M, T2J); { V T30, T2S, T3d, T3z, T3T; T30 = VFNMS(LDK(KP772036680), T2R, T2Q); T2S = VFMA(LDK(KP772036680), T2R, T2Q); T3d = VFNMS(LDK(KP845997307), T3c, T3b); T3z = VFMA(LDK(KP845997307), T3c, T3b); ST(&(x[0]), VADD(T3S, T3L), ms, &(x[0])); T3T = VFNMS(LDK(KP250000000), T3S, T3L); { V T3C, T3p, T2O, T37; T3C = VFMA(LDK(KP906616052), T3o, T3l); T3p = VFNMS(LDK(KP906616052), T3o, T3l); T2O = VFMA(LDK(KP956723877), T2N, T2G); T37 = VFMA(LDK(KP522616830), T2V, T36); { V T31, T2W, T3u, T3h; T31 = VFNMS(LDK(KP522616830), T2G, T30); T2W = VFMA(LDK(KP945422727), T2V, T2S); T3u = VFNMS(LDK(KP923225144), T3g, T3d); T3h = VFMA(LDK(KP923225144), T3g, T3d); { V T3I, T3B, T3V, T3Z; T3I = VFNMS(LDK(KP669429328), T3z, T3A); T3B = VFMA(LDK(KP570584518), T3A, T3z); T3V = VFMA(LDK(KP559016994), T3U, T3T); T3Z = VFNMS(LDK(KP559016994), T3U, T3T); { V T3y, T3q, T2P, T38; T3y = VFMA(LDK(KP262346850), T3p, T2X); T3q = VMUL(LDK(KP998026728), VFNMS(LDK(KP952936919), T2X, T3p)); T2P = VFMA(LDK(KP992114701), T2O, T2z); T38 = VFNMS(LDK(KP690983005), T37, T2S); { V T32, T2Y, T3v, T3F; T32 = VFMA(LDK(KP763932022), T31, T2N); T2Y = VMUL(LDK(KP998026728), VFMA(LDK(KP952936919), T2X, T2W)); T3v = VFNMS(LDK(KP997675361), T3u, T3t); T3F = VFNMS(LDK(KP904508497), T3u, T3s); { V T3i, T3r, T3J, T3D; T3i = VFMA(LDK(KP949179823), T3h, T2z); T3r = VFNMS(LDK(KP237294955), T3h, T2z); T3J = VFNMS(LDK(KP669429328), T3C, T3I); T3D = VFMA(LDK(KP618033988), T3C, T3B); ST(&(x[WS(rs, 20)]), VFMAI(T3Y, T3V), ms, &(x[0])); ST(&(x[WS(rs, 5)]), VFNMSI(T3Y, T3V), ms, &(x[WS(rs, 1)])); ST(&(x[WS(rs, 15)]), VFNMSI(T40, T3Z), ms, &(x[WS(rs, 1)])); ST(&(x[WS(rs, 10)]), VFMAI(T40, T3Z), ms, &(x[0])); { V T39, T33, T3w, T3G; T39 = VFMA(LDK(KP855719849), T38, T35); T33 = VFNMS(LDK(KP855719849), T32, T2Z); ST(&(x[WS(rs, 22)]), VFMAI(T2Y, T2P), ms, &(x[0])); ST(&(x[WS(rs, 3)]), VFNMSI(T2Y, T2P), ms, &(x[WS(rs, 1)])); T3w = VFMA(LDK(KP560319534), T3v, T3s); T3G = VFNMS(LDK(KP681693190), T3F, T3t); ST(&(x[WS(rs, 23)]), VFMAI(T3q, T3i), ms, &(x[WS(rs, 1)])); ST(&(x[WS(rs, 2)]), VFNMSI(T3q, T3i), ms, &(x[0])); T3K = VMUL(LDK(KP951056516), VFNMS(LDK(KP876306680), T3J, T3y)); T3E = VMUL(LDK(KP951056516), VFNMS(LDK(KP949179823), T3D, T3y)); T3a = VMUL(LDK(KP951056516), VFNMS(LDK(KP992114701), T39, T2X)); T34 = VFMA(LDK(KP897376177), T33, T2z); T3x = VFNMS(LDK(KP949179823), T3w, T3r); T3H = VFNMS(LDK(KP860541664), T3G, T3r); T2t = VFNMS(LDK(KP912575812), T2b, T2a); T2c = VFMA(LDK(KP912575812), T2b, T2a); TU = VFMA(LDK(KP829049696), TT, Tz); T1T = VFNMS(LDK(KP829049696), TT, Tz); T1U = VFNMS(LDK(KP831864738), T1y, T1e); T1z = VFMA(LDK(KP831864738), T1y, T1e); } } } } } } } } } } } { V T2o, T2h, T29, T2u, T2v, T2p; T2o = VFNMS(LDK(KP958953096), T2g, T2f); T2h = VFMA(LDK(KP958953096), T2g, T2f); ST(&(x[WS(rs, 17)]), VFMAI(T3a, T34), ms, &(x[WS(rs, 1)])); ST(&(x[WS(rs, 8)]), VFNMSI(T3a, T34), ms, &(x[0])); ST(&(x[WS(rs, 12)]), VFMAI(T3E, T3x), ms, &(x[0])); ST(&(x[WS(rs, 13)]), VFNMSI(T3E, T3x), ms, &(x[WS(rs, 1)])); ST(&(x[WS(rs, 18)]), VFNMSI(T3K, T3H), ms, &(x[0])); ST(&(x[WS(rs, 7)]), VFMAI(T3K, T3H), ms, &(x[WS(rs, 1)])); T1V = VFMA(LDK(KP559154169), T1U, T1T); T22 = VFNMS(LDK(KP683113946), T1T, T1U); T29 = VFNMS(LDK(KP867381224), T28, T27); T2u = VFMA(LDK(KP867381224), T28, T27); T2l = VFMA(LDK(KP894834959), T2k, T2h); T2v = VFMA(LDK(KP447417479), T2k, T2u); T2d = VFNMS(LDK(KP809385824), T2c, T29); T2p = VFMA(LDK(KP447417479), T2c, T2o); T1Q = VFMA(LDK(KP831864738), T1H, T1G); T1I = VFNMS(LDK(KP831864738), T1H, T1G); T2w = VFNMS(LDK(KP763932022), T2v, T2h); T1A = VFMA(LDK(KP904730450), T1z, TU); T1F = VFNMS(LDK(KP904730450), T1z, TU); T2q = VFMA(LDK(KP690983005), T2p, T29); } } { V T2e, T1E, T1P, T2m; T2e = VFNMS(LDK(KP992114701), T2d, Tf); T1E = VFMA(LDK(KP916574801), T1D, T1C); T1P = VFNMS(LDK(KP916574801), T1D, T1C); T2m = VMUL(LDK(KP951056516), VFNMS(LDK(KP992114701), T2l, T1O)); { V T1J, T2r, T1R, T1W, T1Z, T2x; T2x = VFNMS(LDK(KP999544308), T2w, T2t); T1J = VFNMS(LDK(KP904730450), T1I, T1F); T25 = VFMA(LDK(KP968583161), T1A, Tf); T1B = VFNMS(LDK(KP242145790), T1A, Tf); T2r = VFNMS(LDK(KP999544308), T2q, T2n); T1R = VFMA(LDK(KP904730450), T1Q, T1P); T1W = VFNMS(LDK(KP904730450), T1Q, T1P); T1Z = VADD(T1E, T1F); ST(&(x[WS(rs, 21)]), VFNMSI(T2m, T2e), ms, &(x[WS(rs, 1)])); ST(&(x[WS(rs, 4)]), VFMAI(T2m, T2e), ms, &(x[0])); T2y = VMUL(LDK(KP951056516), VFNMS(LDK(KP803003575), T2x, T1O)); T1K = VFNMS(LDK(KP618033988), T1J, T1E); T2s = VFNMS(LDK(KP803003575), T2r, Tf); T23 = VFMA(LDK(KP617882369), T1W, T22); T1S = VFNMS(LDK(KP242145790), T1R, T1O); T26 = VMUL(LDK(KP951056516), VFMA(LDK(KP968583161), T1R, T1O)); T20 = VFNMS(LDK(KP683113946), T1Z, T1I); T1X = VFMA(LDK(KP559016994), T1W, T1V); } } } } } { V T1L, T24, T21, T1Y; T1L = VFNMS(LDK(KP876091699), T1K, T1B); ST(&(x[WS(rs, 9)]), VFMAI(T2y, T2s), ms, &(x[WS(rs, 1)])); ST(&(x[WS(rs, 16)]), VFNMSI(T2y, T2s), ms, &(x[0])); T24 = VMUL(LDK(KP951056516), VFNMS(LDK(KP876306680), T23, T1S)); ST(&(x[WS(rs, 24)]), VFMAI(T26, T25), ms, &(x[0])); ST(&(x[WS(rs, 1)]), VFNMSI(T26, T25), ms, &(x[WS(rs, 1)])); T21 = VFMA(LDK(KP792626838), T20, T1B); T1Y = VMUL(LDK(KP951056516), VFMA(LDK(KP968583161), T1X, T1S)); ST(&(x[WS(rs, 11)]), VFNMSI(T24, T21), ms, &(x[WS(rs, 1)])); ST(&(x[WS(rs, 14)]), VFMAI(T24, T21), ms, &(x[0])); ST(&(x[WS(rs, 19)]), VFMAI(T1Y, T1L), ms, &(x[WS(rs, 1)])); ST(&(x[WS(rs, 6)]), VFNMSI(T1Y, T1L), ms, &(x[0])); } } } VLEAVE(); } static const tw_instr twinstr[] = { VTW(0, 1), VTW(0, 2), VTW(0, 3), VTW(0, 4), VTW(0, 5), VTW(0, 6), VTW(0, 7), VTW(0, 8), VTW(0, 9), VTW(0, 10), VTW(0, 11), VTW(0, 12), VTW(0, 13), VTW(0, 14), VTW(0, 15), VTW(0, 16), VTW(0, 17), VTW(0, 18), VTW(0, 19), VTW(0, 20), VTW(0, 21), VTW(0, 22), VTW(0, 23), VTW(0, 24), {TW_NEXT, VL, 0} }; static const ct_desc desc = { 25, XSIMD_STRING("t2fv_25"), twinstr, &GENUS, {67, 60, 181, 0}, 0, 0, 0 }; void XSIMD(codelet_t2fv_25) (planner *p) { X(kdft_dit_register) (p, t2fv_25, &desc); } #else /* HAVE_FMA */ /* Generated by: ../../../genfft/gen_twiddle_c.native -simd -compact -variables 4 -pipeline-latency 8 -n 25 -name t2fv_25 -include t2f.h */ /* * This function contains 248 FP additions, 188 FP multiplications, * (or, 170 additions, 110 multiplications, 78 fused multiply/add), * 99 stack variables, 40 constants, and 50 memory accesses */ #include "t2f.h" static void t2fv_25(R *ri, R *ii, const R *W, stride rs, INT mb, INT me, INT ms) { DVK(KP998026728, +0.998026728428271561952336806863450553336905220); DVK(KP125581039, +0.125581039058626752152356449131262266244969664); DVK(KP1_996053456, +1.996053456856543123904673613726901106673810439); DVK(KP062790519, +0.062790519529313376076178224565631133122484832); DVK(KP809016994, +0.809016994374947424102293417182819058860154590); DVK(KP309016994, +0.309016994374947424102293417182819058860154590); DVK(KP1_369094211, +1.369094211857377347464566715242418539779038465); DVK(KP728968627, +0.728968627421411523146730319055259111372571664); DVK(KP963507348, +0.963507348203430549974383005744259307057084020); DVK(KP876306680, +0.876306680043863587308115903922062583399064238); DVK(KP497379774, +0.497379774329709576484567492012895936835134813); DVK(KP968583161, +0.968583161128631119490168375464735813836012403); DVK(KP684547105, +0.684547105928688673732283357621209269889519233); DVK(KP1_457937254, +1.457937254842823046293460638110518222745143328); DVK(KP481753674, +0.481753674101715274987191502872129653528542010); DVK(KP1_752613360, +1.752613360087727174616231807844125166798128477); DVK(KP248689887, +0.248689887164854788242283746006447968417567406); DVK(KP1_937166322, +1.937166322257262238980336750929471627672024806); DVK(KP992114701, +0.992114701314477831049793042785778521453036709); DVK(KP250666467, +0.250666467128608490746237519633017587885836494); DVK(KP425779291, +0.425779291565072648862502445744251703979973042); DVK(KP1_809654104, +1.809654104932039055427337295865395187940827822); DVK(KP1_274847979, +1.274847979497379420353425623352032390869834596); DVK(KP770513242, +0.770513242775789230803009636396177847271667672); DVK(KP844327925, +0.844327925502015078548558063966681505381659241); DVK(KP1_071653589, +1.071653589957993236542617535735279956127150691); DVK(KP125333233, +0.125333233564304245373118759816508793942918247); DVK(KP1_984229402, +1.984229402628955662099586085571557042906073418); DVK(KP904827052, +0.904827052466019527713668647932697593970413911); DVK(KP851558583, +0.851558583130145297725004891488503407959946084); DVK(KP637423989, +0.637423989748689710176712811676016195434917298); DVK(KP1_541026485, +1.541026485551578461606019272792355694543335344); DVK(KP535826794, +0.535826794978996618271308767867639978063575346); DVK(KP1_688655851, +1.688655851004030157097116127933363010763318483); DVK(KP293892626, +0.293892626146236564584352977319536384298826219); DVK(KP475528258, +0.475528258147576786058219666689691071702849317); DVK(KP587785252, +0.587785252292473129168705954639072768597652438); DVK(KP951056516, +0.951056516295153572116439333379382143405698634); DVK(KP250000000, +0.250000000000000000000000000000000000000000000); DVK(KP559016994, +0.559016994374947424102293417182819058860154590); { INT m; R *x; x = ri; for (m = mb, W = W + (mb * ((TWVL / VL) * 48)); m < me; m = m + VL, x = x + (VL * ms), W = W + (TWVL * 48), MAKE_VOLATILE_STRIDE(25, rs)) { V Tc, Tb, Td, Te, T1C, T2t, T1E, T1x, T2m, T1u, T3c, T2n, Ty, T2i, Tv; V T38, T2j, TS, T2f, TP, T39, T2g, T1d, T2p, T1a, T3b, T2q; { V T7, T9, Ta, T2, T4, T5, T1D; Tc = LD(&(x[0]), ms, &(x[0])); { V T6, T8, T1, T3; T6 = LD(&(x[WS(rs, 10)]), ms, &(x[0])); T7 = BYTWJ(&(W[TWVL * 18]), T6); T8 = LD(&(x[WS(rs, 15)]), ms, &(x[WS(rs, 1)])); T9 = BYTWJ(&(W[TWVL * 28]), T8); Ta = VADD(T7, T9); T1 = LD(&(x[WS(rs, 5)]), ms, &(x[WS(rs, 1)])); T2 = BYTWJ(&(W[TWVL * 8]), T1); T3 = LD(&(x[WS(rs, 20)]), ms, &(x[0])); T4 = BYTWJ(&(W[TWVL * 38]), T3); T5 = VADD(T2, T4); } Tb = VMUL(LDK(KP559016994), VSUB(T5, Ta)); Td = VADD(T5, Ta); Te = VFNMS(LDK(KP250000000), Td, Tc); T1C = VSUB(T2, T4); T1D = VSUB(T7, T9); T2t = VMUL(LDK(KP951056516), T1D); T1E = VFMA(LDK(KP951056516), T1C, VMUL(LDK(KP587785252), T1D)); } { V T1r, T1l, T1n, T1o, T1g, T1i, T1j, T1q; T1q = LD(&(x[WS(rs, 3)]), ms, &(x[WS(rs, 1)])); T1r = BYTWJ(&(W[TWVL * 4]), T1q); { V T1k, T1m, T1f, T1h; T1k = LD(&(x[WS(rs, 13)]), ms, &(x[WS(rs, 1)])); T1l = BYTWJ(&(W[TWVL * 24]), T1k); T1m = LD(&(x[WS(rs, 18)]), ms, &(x[0])); T1n = BYTWJ(&(W[TWVL * 34]), T1m); T1o = VADD(T1l, T1n); T1f = LD(&(x[WS(rs, 8)]), ms, &(x[0])); T1g = BYTWJ(&(W[TWVL * 14]), T1f); T1h = LD(&(x[WS(rs, 23)]), ms, &(x[WS(rs, 1)])); T1i = BYTWJ(&(W[TWVL * 44]), T1h); T1j = VADD(T1g, T1i); } { V T1v, T1w, T1p, T1s, T1t; T1v = VSUB(T1g, T1i); T1w = VSUB(T1l, T1n); T1x = VFMA(LDK(KP475528258), T1v, VMUL(LDK(KP293892626), T1w)); T2m = VFNMS(LDK(KP293892626), T1v, VMUL(LDK(KP475528258), T1w)); T1p = VMUL(LDK(KP559016994), VSUB(T1j, T1o)); T1s = VADD(T1j, T1o); T1t = VFNMS(LDK(KP250000000), T1s, T1r); T1u = VADD(T1p, T1t); T3c = VADD(T1r, T1s); T2n = VSUB(T1t, T1p); } } { V Ts, Tm, To, Tp, Th, Tj, Tk, Tr; Tr = LD(&(x[WS(rs, 1)]), ms, &(x[WS(rs, 1)])); Ts = BYTWJ(&(W[0]), Tr); { V Tl, Tn, Tg, Ti; Tl = LD(&(x[WS(rs, 11)]), ms, &(x[WS(rs, 1)])); Tm = BYTWJ(&(W[TWVL * 20]), Tl); Tn = LD(&(x[WS(rs, 16)]), ms, &(x[0])); To = BYTWJ(&(W[TWVL * 30]), Tn); Tp = VADD(Tm, To); Tg = LD(&(x[WS(rs, 6)]), ms, &(x[0])); Th = BYTWJ(&(W[TWVL * 10]), Tg); Ti = LD(&(x[WS(rs, 21)]), ms, &(x[WS(rs, 1)])); Tj = BYTWJ(&(W[TWVL * 40]), Ti); Tk = VADD(Th, Tj); } { V Tw, Tx, Tq, Tt, Tu; Tw = VSUB(Th, Tj); Tx = VSUB(Tm, To); Ty = VFMA(LDK(KP475528258), Tw, VMUL(LDK(KP293892626), Tx)); T2i = VFNMS(LDK(KP293892626), Tw, VMUL(LDK(KP475528258), Tx)); Tq = VMUL(LDK(KP559016994), VSUB(Tk, Tp)); Tt = VADD(Tk, Tp); Tu = VFNMS(LDK(KP250000000), Tt, Ts); Tv = VADD(Tq, Tu); T38 = VADD(Ts, Tt); T2j = VSUB(Tu, Tq); } } { V TM, TG, TI, TJ, TB, TD, TE, TL; TL = LD(&(x[WS(rs, 4)]), ms, &(x[0])); TM = BYTWJ(&(W[TWVL * 6]), TL); { V TF, TH, TA, TC; TF = LD(&(x[WS(rs, 14)]), ms, &(x[0])); TG = BYTWJ(&(W[TWVL * 26]), TF); TH = LD(&(x[WS(rs, 19)]), ms, &(x[WS(rs, 1)])); TI = BYTWJ(&(W[TWVL * 36]), TH); TJ = VADD(TG, TI); TA = LD(&(x[WS(rs, 9)]), ms, &(x[WS(rs, 1)])); TB = BYTWJ(&(W[TWVL * 16]), TA); TC = LD(&(x[WS(rs, 24)]), ms, &(x[0])); TD = BYTWJ(&(W[TWVL * 46]), TC); TE = VADD(TB, TD); } { V TQ, TR, TK, TN, TO; TQ = VSUB(TB, TD); TR = VSUB(TG, TI); TS = VFMA(LDK(KP475528258), TQ, VMUL(LDK(KP293892626), TR)); T2f = VFNMS(LDK(KP293892626), TQ, VMUL(LDK(KP475528258), TR)); TK = VMUL(LDK(KP559016994), VSUB(TE, TJ)); TN = VADD(TE, TJ); TO = VFNMS(LDK(KP250000000), TN, TM); TP = VADD(TK, TO); T39 = VADD(TM, TN); T2g = VSUB(TO, TK); } } { V T17, T11, T13, T14, TW, TY, TZ, T16; T16 = LD(&(x[WS(rs, 2)]), ms, &(x[0])); T17 = BYTWJ(&(W[TWVL * 2]), T16); { V T10, T12, TV, TX; T10 = LD(&(x[WS(rs, 12)]), ms, &(x[0])); T11 = BYTWJ(&(W[TWVL * 22]), T10); T12 = LD(&(x[WS(rs, 17)]), ms, &(x[WS(rs, 1)])); T13 = BYTWJ(&(W[TWVL * 32]), T12); T14 = VADD(T11, T13); TV = LD(&(x[WS(rs, 7)]), ms, &(x[WS(rs, 1)])); TW = BYTWJ(&(W[TWVL * 12]), TV); TX = LD(&(x[WS(rs, 22)]), ms, &(x[0])); TY = BYTWJ(&(W[TWVL * 42]), TX); TZ = VADD(TW, TY); } { V T1b, T1c, T15, T18, T19; T1b = VSUB(TW, TY); T1c = VSUB(T11, T13); T1d = VFMA(LDK(KP475528258), T1b, VMUL(LDK(KP293892626), T1c)); T2p = VFNMS(LDK(KP293892626), T1b, VMUL(LDK(KP475528258), T1c)); T15 = VMUL(LDK(KP559016994), VSUB(TZ, T14)); T18 = VADD(TZ, T14); T19 = VFNMS(LDK(KP250000000), T18, T17); T1a = VADD(T15, T19); T3b = VADD(T17, T18); T2q = VSUB(T19, T15); } } { V T3l, T3m, T3f, T3g, T3e, T3h, T3n, T3i; { V T3j, T3k, T3a, T3d; T3j = VSUB(T38, T39); T3k = VSUB(T3b, T3c); T3l = VBYI(VFMA(LDK(KP951056516), T3j, VMUL(LDK(KP587785252), T3k))); T3m = VBYI(VFNMS(LDK(KP587785252), T3j, VMUL(LDK(KP951056516), T3k))); T3f = VADD(Tc, Td); T3a = VADD(T38, T39); T3d = VADD(T3b, T3c); T3g = VADD(T3a, T3d); T3e = VMUL(LDK(KP559016994), VSUB(T3a, T3d)); T3h = VFNMS(LDK(KP250000000), T3g, T3f); } ST(&(x[0]), VADD(T3f, T3g), ms, &(x[0])); T3n = VSUB(T3h, T3e); ST(&(x[WS(rs, 10)]), VADD(T3m, T3n), ms, &(x[0])); ST(&(x[WS(rs, 15)]), VSUB(T3n, T3m), ms, &(x[WS(rs, 1)])); T3i = VADD(T3e, T3h); ST(&(x[WS(rs, 5)]), VSUB(T3i, T3l), ms, &(x[WS(rs, 1)])); ST(&(x[WS(rs, 20)]), VADD(T3l, T3i), ms, &(x[0])); } { V Tf, T1Z, T20, T21, T29, T2a, T2b, T26, T27, T28, T22, T23, T24, T1L, T1U; V T1Q, T1S, T1A, T1V, T1N, T1O, T2d, T2e; Tf = VADD(Tb, Te); T1Z = VFMA(LDK(KP1_688655851), Ty, VMUL(LDK(KP535826794), Tv)); T20 = VFMA(LDK(KP1_541026485), TS, VMUL(LDK(KP637423989), TP)); T21 = VSUB(T1Z, T20); T29 = VFMA(LDK(KP851558583), T1d, VMUL(LDK(KP904827052), T1a)); T2a = VFMA(LDK(KP1_984229402), T1x, VMUL(LDK(KP125333233), T1u)); T2b = VADD(T29, T2a); T26 = VFNMS(LDK(KP844327925), Tv, VMUL(LDK(KP1_071653589), Ty)); T27 = VFNMS(LDK(KP1_274847979), TS, VMUL(LDK(KP770513242), TP)); T28 = VADD(T26, T27); T22 = VFNMS(LDK(KP425779291), T1a, VMUL(LDK(KP1_809654104), T1d)); T23 = VFNMS(LDK(KP992114701), T1u, VMUL(LDK(KP250666467), T1x)); T24 = VADD(T22, T23); { V T1F, T1G, T1H, T1I, T1J, T1K; T1F = VFMA(LDK(KP1_937166322), Ty, VMUL(LDK(KP248689887), Tv)); T1G = VFMA(LDK(KP1_071653589), TS, VMUL(LDK(KP844327925), TP)); T1H = VADD(T1F, T1G); T1I = VFMA(LDK(KP1_752613360), T1d, VMUL(LDK(KP481753674), T1a)); T1J = VFMA(LDK(KP1_457937254), T1x, VMUL(LDK(KP684547105), T1u)); T1K = VADD(T1I, T1J); T1L = VADD(T1H, T1K); T1U = VSUB(T1J, T1I); T1Q = VMUL(LDK(KP559016994), VSUB(T1K, T1H)); T1S = VSUB(T1G, T1F); } { V Tz, TT, TU, T1e, T1y, T1z; Tz = VFNMS(LDK(KP497379774), Ty, VMUL(LDK(KP968583161), Tv)); TT = VFNMS(LDK(KP1_688655851), TS, VMUL(LDK(KP535826794), TP)); TU = VADD(Tz, TT); T1e = VFNMS(LDK(KP963507348), T1d, VMUL(LDK(KP876306680), T1a)); T1y = VFNMS(LDK(KP1_369094211), T1x, VMUL(LDK(KP728968627), T1u)); T1z = VADD(T1e, T1y); T1A = VADD(TU, T1z); T1V = VMUL(LDK(KP559016994), VSUB(TU, T1z)); T1N = VSUB(TT, Tz); T1O = VSUB(T1e, T1y); } { V T1B, T1M, T25, T2c; T1B = VADD(Tf, T1A); T1M = VBYI(VADD(T1E, T1L)); ST(&(x[WS(rs, 1)]), VSUB(T1B, T1M), ms, &(x[WS(rs, 1)])); ST(&(x[WS(rs, 24)]), VADD(T1B, T1M), ms, &(x[0])); T25 = VADD(Tf, VADD(T21, T24)); T2c = VBYI(VADD(T1E, VSUB(T28, T2b))); ST(&(x[WS(rs, 21)]), VSUB(T25, T2c), ms, &(x[WS(rs, 1)])); ST(&(x[WS(rs, 4)]), VADD(T25, T2c), ms, &(x[0])); } T2d = VBYI(VADD(T1E, VFMA(LDK(KP309016994), T28, VFMA(LDK(KP587785252), VSUB(T23, T22), VFNMS(LDK(KP951056516), VADD(T1Z, T20), VMUL(LDK(KP809016994), T2b)))))); T2e = VFMA(LDK(KP309016994), T21, VFMA(LDK(KP951056516), VSUB(T26, T27), VFMA(LDK(KP587785252), VSUB(T2a, T29), VFNMS(LDK(KP809016994), T24, Tf)))); ST(&(x[WS(rs, 9)]), VADD(T2d, T2e), ms, &(x[WS(rs, 1)])); ST(&(x[WS(rs, 16)]), VSUB(T2e, T2d), ms, &(x[0])); { V T1R, T1X, T1W, T1Y, T1P, T1T; T1P = VFMS(LDK(KP250000000), T1L, T1E); T1R = VBYI(VADD(VFMA(LDK(KP587785252), T1N, VMUL(LDK(KP951056516), T1O)), VSUB(T1P, T1Q))); T1X = VBYI(VADD(VFNMS(LDK(KP587785252), T1O, VMUL(LDK(KP951056516), T1N)), VADD(T1P, T1Q))); T1T = VFNMS(LDK(KP250000000), T1A, Tf); T1W = VFMA(LDK(KP587785252), T1S, VFNMS(LDK(KP951056516), T1U, VSUB(T1T, T1V))); T1Y = VFMA(LDK(KP951056516), T1S, VADD(T1V, VFMA(LDK(KP587785252), T1U, T1T))); ST(&(x[WS(rs, 11)]), VADD(T1R, T1W), ms, &(x[WS(rs, 1)])); ST(&(x[WS(rs, 19)]), VSUB(T1Y, T1X), ms, &(x[WS(rs, 1)])); ST(&(x[WS(rs, 14)]), VSUB(T1W, T1R), ms, &(x[0])); ST(&(x[WS(rs, 6)]), VADD(T1X, T1Y), ms, &(x[0])); } } { V T2u, T2w, T2h, T2k, T2l, T2A, T2B, T2C, T2o, T2r, T2s, T2x, T2y, T2z, T2M; V T2X, T2N, T2W, T2R, T31, T2U, T30, T2E, T2F; T2u = VFNMS(LDK(KP587785252), T1C, T2t); T2w = VSUB(Te, Tb); T2h = VFNMS(LDK(KP125333233), T2g, VMUL(LDK(KP1_984229402), T2f)); T2k = VFMA(LDK(KP1_457937254), T2i, VMUL(LDK(KP684547105), T2j)); T2l = VSUB(T2h, T2k); T2A = VFNMS(LDK(KP1_996053456), T2p, VMUL(LDK(KP062790519), T2q)); T2B = VFMA(LDK(KP1_541026485), T2m, VMUL(LDK(KP637423989), T2n)); T2C = VSUB(T2A, T2B); T2o = VFNMS(LDK(KP770513242), T2n, VMUL(LDK(KP1_274847979), T2m)); T2r = VFMA(LDK(KP125581039), T2p, VMUL(LDK(KP998026728), T2q)); T2s = VSUB(T2o, T2r); T2x = VFNMS(LDK(KP1_369094211), T2i, VMUL(LDK(KP728968627), T2j)); T2y = VFMA(LDK(KP250666467), T2f, VMUL(LDK(KP992114701), T2g)); T2z = VSUB(T2x, T2y); { V T2G, T2H, T2I, T2J, T2K, T2L; T2G = VFNMS(LDK(KP481753674), T2j, VMUL(LDK(KP1_752613360), T2i)); T2H = VFMA(LDK(KP851558583), T2f, VMUL(LDK(KP904827052), T2g)); T2I = VSUB(T2G, T2H); T2J = VFNMS(LDK(KP844327925), T2q, VMUL(LDK(KP1_071653589), T2p)); T2K = VFNMS(LDK(KP998026728), T2n, VMUL(LDK(KP125581039), T2m)); T2L = VADD(T2J, T2K); T2M = VMUL(LDK(KP559016994), VSUB(T2I, T2L)); T2X = VSUB(T2J, T2K); T2N = VADD(T2I, T2L); T2W = VADD(T2G, T2H); } { V T2P, T2Q, T2Y, T2S, T2T, T2Z; T2P = VFNMS(LDK(KP425779291), T2g, VMUL(LDK(KP1_809654104), T2f)); T2Q = VFMA(LDK(KP963507348), T2i, VMUL(LDK(KP876306680), T2j)); T2Y = VADD(T2Q, T2P); T2S = VFMA(LDK(KP1_688655851), T2p, VMUL(LDK(KP535826794), T2q)); T2T = VFMA(LDK(KP1_996053456), T2m, VMUL(LDK(KP062790519), T2n)); T2Z = VADD(T2S, T2T); T2R = VSUB(T2P, T2Q); T31 = VADD(T2Y, T2Z); T2U = VSUB(T2S, T2T); T30 = VMUL(LDK(KP559016994), VSUB(T2Y, T2Z)); } { V T36, T37, T2v, T2D; T36 = VBYI(VADD(T2u, T2N)); T37 = VADD(T2w, T31); ST(&(x[WS(rs, 2)]), VADD(T36, T37), ms, &(x[0])); ST(&(x[WS(rs, 23)]), VSUB(T37, T36), ms, &(x[WS(rs, 1)])); T2v = VBYI(VSUB(VADD(T2l, T2s), T2u)); T2D = VADD(T2w, VADD(T2z, T2C)); ST(&(x[WS(rs, 3)]), VADD(T2v, T2D), ms, &(x[WS(rs, 1)])); ST(&(x[WS(rs, 22)]), VSUB(T2D, T2v), ms, &(x[0])); } T2E = VFMA(LDK(KP309016994), T2z, VFNMS(LDK(KP809016994), T2C, VFNMS(LDK(KP587785252), VADD(T2r, T2o), VFNMS(LDK(KP951056516), VADD(T2k, T2h), T2w)))); T2F = VBYI(VSUB(VFNMS(LDK(KP587785252), VADD(T2A, T2B), VFNMS(LDK(KP809016994), T2s, VFNMS(LDK(KP951056516), VADD(T2x, T2y), VMUL(LDK(KP309016994), T2l)))), T2u)); ST(&(x[WS(rs, 17)]), VSUB(T2E, T2F), ms, &(x[WS(rs, 1)])); ST(&(x[WS(rs, 8)]), VADD(T2E, T2F), ms, &(x[0])); { V T2V, T34, T33, T35, T2O, T32; T2O = VFNMS(LDK(KP250000000), T2N, T2u); T2V = VBYI(VADD(T2M, VADD(T2O, VFNMS(LDK(KP587785252), T2U, VMUL(LDK(KP951056516), T2R))))); T34 = VBYI(VADD(T2O, VSUB(VFMA(LDK(KP587785252), T2R, VMUL(LDK(KP951056516), T2U)), T2M))); T32 = VFNMS(LDK(KP250000000), T31, T2w); T33 = VFMA(LDK(KP951056516), T2W, VFMA(LDK(KP587785252), T2X, VADD(T30, T32))); T35 = VFMA(LDK(KP587785252), T2W, VSUB(VFNMS(LDK(KP951056516), T2X, T32), T30)); ST(&(x[WS(rs, 7)]), VADD(T2V, T33), ms, &(x[WS(rs, 1)])); ST(&(x[WS(rs, 13)]), VSUB(T35, T34), ms, &(x[WS(rs, 1)])); ST(&(x[WS(rs, 18)]), VSUB(T33, T2V), ms, &(x[0])); ST(&(x[WS(rs, 12)]), VADD(T34, T35), ms, &(x[0])); } } } } VLEAVE(); } static const tw_instr twinstr[] = { VTW(0, 1), VTW(0, 2), VTW(0, 3), VTW(0, 4), VTW(0, 5), VTW(0, 6), VTW(0, 7), VTW(0, 8), VTW(0, 9), VTW(0, 10), VTW(0, 11), VTW(0, 12), VTW(0, 13), VTW(0, 14), VTW(0, 15), VTW(0, 16), VTW(0, 17), VTW(0, 18), VTW(0, 19), VTW(0, 20), VTW(0, 21), VTW(0, 22), VTW(0, 23), VTW(0, 24), {TW_NEXT, VL, 0} }; static const ct_desc desc = { 25, XSIMD_STRING("t2fv_25"), twinstr, &GENUS, {170, 110, 78, 0}, 0, 0, 0 }; void XSIMD(codelet_t2fv_25) (planner *p) { X(kdft_dit_register) (p, t2fv_25, &desc); } #endif /* HAVE_FMA */ fftw-3.3.3/dft/simd/common/t3fv_4.c0000644000175000017500000001102412054410531013652 00000000000000/* * Copyright (c) 2003, 2007-11 Matteo Frigo * Copyright (c) 2003, 2007-11 Massachusetts Institute of Technology * * This program is free software; you can redistribute it and/or modify * it under the terms of the GNU General Public License as published by * the Free Software Foundation; either version 2 of the License, or * (at your option) any later version. * * This program is distributed in the hope that it will be useful, * but WITHOUT ANY WARRANTY; without even the implied warranty of * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the * GNU General Public License for more details. * * You should have received a copy of the GNU General Public License * along with this program; if not, write to the Free Software * Foundation, Inc., 51 Franklin Street, Fifth Floor, Boston, MA 02110-1301 USA * */ /* This file was automatically generated --- DO NOT EDIT */ /* Generated on Sun Nov 25 07:38:49 EST 2012 */ #include "codelet-dft.h" #ifdef HAVE_FMA /* Generated by: ../../../genfft/gen_twiddle_c.native -fma -reorder-insns -schedule-for-pipeline -simd -compact -variables 4 -pipeline-latency 8 -twiddle-log3 -precompute-twiddles -no-generate-bytw -n 4 -name t3fv_4 -include t3f.h */ /* * This function contains 12 FP additions, 10 FP multiplications, * (or, 10 additions, 8 multiplications, 2 fused multiply/add), * 16 stack variables, 0 constants, and 8 memory accesses */ #include "t3f.h" static void t3fv_4(R *ri, R *ii, const R *W, stride rs, INT mb, INT me, INT ms) { { INT m; R *x; x = ri; for (m = mb, W = W + (mb * ((TWVL / VL) * 4)); m < me; m = m + VL, x = x + (VL * ms), W = W + (TWVL * 4), MAKE_VOLATILE_STRIDE(4, rs)) { V T2, T3, T1, Ta, T5, T8; T2 = LDW(&(W[0])); T3 = LDW(&(W[TWVL * 2])); T1 = LD(&(x[0]), ms, &(x[0])); Ta = LD(&(x[WS(rs, 3)]), ms, &(x[WS(rs, 1)])); T5 = LD(&(x[WS(rs, 2)]), ms, &(x[0])); T8 = LD(&(x[WS(rs, 1)]), ms, &(x[WS(rs, 1)])); { V T4, Tb, T9, T6; T4 = VZMULJ(T2, T3); Tb = VZMULJ(T3, Ta); T9 = VZMULJ(T2, T8); T6 = VZMULJ(T4, T5); { V Tc, Te, T7, Td; Tc = VSUB(T9, Tb); Te = VADD(T9, Tb); T7 = VSUB(T1, T6); Td = VADD(T1, T6); ST(&(x[0]), VADD(Td, Te), ms, &(x[0])); ST(&(x[WS(rs, 2)]), VSUB(Td, Te), ms, &(x[0])); ST(&(x[WS(rs, 3)]), VFMAI(Tc, T7), ms, &(x[WS(rs, 1)])); ST(&(x[WS(rs, 1)]), VFNMSI(Tc, T7), ms, &(x[WS(rs, 1)])); } } } } VLEAVE(); } static const tw_instr twinstr[] = { VTW(0, 1), VTW(0, 3), {TW_NEXT, VL, 0} }; static const ct_desc desc = { 4, XSIMD_STRING("t3fv_4"), twinstr, &GENUS, {10, 8, 2, 0}, 0, 0, 0 }; void XSIMD(codelet_t3fv_4) (planner *p) { X(kdft_dit_register) (p, t3fv_4, &desc); } #else /* HAVE_FMA */ /* Generated by: ../../../genfft/gen_twiddle_c.native -simd -compact -variables 4 -pipeline-latency 8 -twiddle-log3 -precompute-twiddles -no-generate-bytw -n 4 -name t3fv_4 -include t3f.h */ /* * This function contains 12 FP additions, 8 FP multiplications, * (or, 12 additions, 8 multiplications, 0 fused multiply/add), * 16 stack variables, 0 constants, and 8 memory accesses */ #include "t3f.h" static void t3fv_4(R *ri, R *ii, const R *W, stride rs, INT mb, INT me, INT ms) { { INT m; R *x; x = ri; for (m = mb, W = W + (mb * ((TWVL / VL) * 4)); m < me; m = m + VL, x = x + (VL * ms), W = W + (TWVL * 4), MAKE_VOLATILE_STRIDE(4, rs)) { V T2, T3, T4; T2 = LDW(&(W[0])); T3 = LDW(&(W[TWVL * 2])); T4 = VZMULJ(T2, T3); { V T1, Tb, T6, T9, Ta, T5, T8; T1 = LD(&(x[0]), ms, &(x[0])); Ta = LD(&(x[WS(rs, 3)]), ms, &(x[WS(rs, 1)])); Tb = VZMULJ(T3, Ta); T5 = LD(&(x[WS(rs, 2)]), ms, &(x[0])); T6 = VZMULJ(T4, T5); T8 = LD(&(x[WS(rs, 1)]), ms, &(x[WS(rs, 1)])); T9 = VZMULJ(T2, T8); { V T7, Tc, Td, Te; T7 = VSUB(T1, T6); Tc = VBYI(VSUB(T9, Tb)); ST(&(x[WS(rs, 1)]), VSUB(T7, Tc), ms, &(x[WS(rs, 1)])); ST(&(x[WS(rs, 3)]), VADD(T7, Tc), ms, &(x[WS(rs, 1)])); Td = VADD(T1, T6); Te = VADD(T9, Tb); ST(&(x[WS(rs, 2)]), VSUB(Td, Te), ms, &(x[0])); ST(&(x[0]), VADD(Td, Te), ms, &(x[0])); } } } } VLEAVE(); } static const tw_instr twinstr[] = { VTW(0, 1), VTW(0, 3), {TW_NEXT, VL, 0} }; static const ct_desc desc = { 4, XSIMD_STRING("t3fv_4"), twinstr, &GENUS, {12, 8, 0, 0}, 0, 0, 0 }; void XSIMD(codelet_t3fv_4) (planner *p) { X(kdft_dit_register) (p, t3fv_4, &desc); } #endif /* HAVE_FMA */ fftw-3.3.3/dft/simd/common/t3fv_8.c0000644000175000017500000001627112054410531013667 00000000000000/* * Copyright (c) 2003, 2007-11 Matteo Frigo * Copyright (c) 2003, 2007-11 Massachusetts Institute of Technology * * This program is free software; you can redistribute it and/or modify * it under the terms of the GNU General Public License as published by * the Free Software Foundation; either version 2 of the License, or * (at your option) any later version. * * This program is distributed in the hope that it will be useful, * but WITHOUT ANY WARRANTY; without even the implied warranty of * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the * GNU General Public License for more details. * * You should have received a copy of the GNU General Public License * along with this program; if not, write to the Free Software * Foundation, Inc., 51 Franklin Street, Fifth Floor, Boston, MA 02110-1301 USA * */ /* This file was automatically generated --- DO NOT EDIT */ /* Generated on Sun Nov 25 07:38:49 EST 2012 */ #include "codelet-dft.h" #ifdef HAVE_FMA /* Generated by: ../../../genfft/gen_twiddle_c.native -fma -reorder-insns -schedule-for-pipeline -simd -compact -variables 4 -pipeline-latency 8 -twiddle-log3 -precompute-twiddles -no-generate-bytw -n 8 -name t3fv_8 -include t3f.h */ /* * This function contains 37 FP additions, 32 FP multiplications, * (or, 27 additions, 22 multiplications, 10 fused multiply/add), * 43 stack variables, 1 constants, and 16 memory accesses */ #include "t3f.h" static void t3fv_8(R *ri, R *ii, const R *W, stride rs, INT mb, INT me, INT ms) { DVK(KP707106781, +0.707106781186547524400844362104849039284835938); { INT m; R *x; x = ri; for (m = mb, W = W + (mb * ((TWVL / VL) * 6)); m < me; m = m + VL, x = x + (VL * ms), W = W + (TWVL * 6), MAKE_VOLATILE_STRIDE(8, rs)) { V T2, T3, Tb, T1, T5, Tn, Tq, T8, Td, T4, Ta, Tp, Tg, Ti, T9; T2 = LDW(&(W[0])); T3 = LDW(&(W[TWVL * 2])); Tb = LDW(&(W[TWVL * 4])); T1 = LD(&(x[0]), ms, &(x[0])); T5 = LD(&(x[WS(rs, 4)]), ms, &(x[0])); Tn = LD(&(x[WS(rs, 2)]), ms, &(x[0])); Tq = LD(&(x[WS(rs, 6)]), ms, &(x[0])); T8 = LD(&(x[WS(rs, 1)]), ms, &(x[WS(rs, 1)])); Td = LD(&(x[WS(rs, 5)]), ms, &(x[WS(rs, 1)])); T4 = VZMUL(T2, T3); Ta = VZMULJ(T2, T3); Tp = VZMULJ(T2, Tb); Tg = LD(&(x[WS(rs, 7)]), ms, &(x[WS(rs, 1)])); Ti = LD(&(x[WS(rs, 3)]), ms, &(x[WS(rs, 1)])); T9 = VZMULJ(T2, T8); { V T6, To, Tc, Tr, Th, Tj; T6 = VZMULJ(T4, T5); To = VZMULJ(Ta, Tn); Tc = VZMULJ(Ta, Tb); Tr = VZMULJ(Tp, Tq); Th = VZMULJ(Tb, Tg); Tj = VZMULJ(T3, Ti); { V Tx, T7, Te, Ts, Ty, Tk, TB; Tx = VADD(T1, T6); T7 = VSUB(T1, T6); Te = VZMULJ(Tc, Td); Ts = VSUB(To, Tr); Ty = VADD(To, Tr); Tk = VSUB(Th, Tj); TB = VADD(Th, Tj); { V Tf, TA, Tz, TD; Tf = VSUB(T9, Te); TA = VADD(T9, Te); Tz = VADD(Tx, Ty); TD = VSUB(Tx, Ty); { V TC, TE, Tl, Tt; TC = VADD(TA, TB); TE = VSUB(TB, TA); Tl = VADD(Tf, Tk); Tt = VSUB(Tk, Tf); { V Tu, Tw, Tm, Tv; ST(&(x[WS(rs, 2)]), VFMAI(TE, TD), ms, &(x[0])); ST(&(x[WS(rs, 6)]), VFNMSI(TE, TD), ms, &(x[0])); ST(&(x[0]), VADD(Tz, TC), ms, &(x[0])); ST(&(x[WS(rs, 4)]), VSUB(Tz, TC), ms, &(x[0])); Tu = VFNMS(LDK(KP707106781), Tt, Ts); Tw = VFMA(LDK(KP707106781), Tt, Ts); Tm = VFMA(LDK(KP707106781), Tl, T7); Tv = VFNMS(LDK(KP707106781), Tl, T7); ST(&(x[WS(rs, 5)]), VFNMSI(Tw, Tv), ms, &(x[WS(rs, 1)])); ST(&(x[WS(rs, 3)]), VFMAI(Tw, Tv), ms, &(x[WS(rs, 1)])); ST(&(x[WS(rs, 7)]), VFMAI(Tu, Tm), ms, &(x[WS(rs, 1)])); ST(&(x[WS(rs, 1)]), VFNMSI(Tu, Tm), ms, &(x[WS(rs, 1)])); } } } } } } } VLEAVE(); } static const tw_instr twinstr[] = { VTW(0, 1), VTW(0, 3), VTW(0, 7), {TW_NEXT, VL, 0} }; static const ct_desc desc = { 8, XSIMD_STRING("t3fv_8"), twinstr, &GENUS, {27, 22, 10, 0}, 0, 0, 0 }; void XSIMD(codelet_t3fv_8) (planner *p) { X(kdft_dit_register) (p, t3fv_8, &desc); } #else /* HAVE_FMA */ /* Generated by: ../../../genfft/gen_twiddle_c.native -simd -compact -variables 4 -pipeline-latency 8 -twiddle-log3 -precompute-twiddles -no-generate-bytw -n 8 -name t3fv_8 -include t3f.h */ /* * This function contains 37 FP additions, 24 FP multiplications, * (or, 37 additions, 24 multiplications, 0 fused multiply/add), * 31 stack variables, 1 constants, and 16 memory accesses */ #include "t3f.h" static void t3fv_8(R *ri, R *ii, const R *W, stride rs, INT mb, INT me, INT ms) { DVK(KP707106781, +0.707106781186547524400844362104849039284835938); { INT m; R *x; x = ri; for (m = mb, W = W + (mb * ((TWVL / VL) * 6)); m < me; m = m + VL, x = x + (VL * ms), W = W + (TWVL * 6), MAKE_VOLATILE_STRIDE(8, rs)) { V T2, T3, Ta, T4, Tb, Tc, Tq; T2 = LDW(&(W[0])); T3 = LDW(&(W[TWVL * 2])); Ta = VZMULJ(T2, T3); T4 = VZMUL(T2, T3); Tb = LDW(&(W[TWVL * 4])); Tc = VZMULJ(Ta, Tb); Tq = VZMULJ(T2, Tb); { V T7, Tx, Tt, Ty, Tf, TA, Tk, TB, T1, T6, T5; T1 = LD(&(x[0]), ms, &(x[0])); T5 = LD(&(x[WS(rs, 4)]), ms, &(x[0])); T6 = VZMULJ(T4, T5); T7 = VSUB(T1, T6); Tx = VADD(T1, T6); { V Tp, Ts, To, Tr; To = LD(&(x[WS(rs, 2)]), ms, &(x[0])); Tp = VZMULJ(Ta, To); Tr = LD(&(x[WS(rs, 6)]), ms, &(x[0])); Ts = VZMULJ(Tq, Tr); Tt = VSUB(Tp, Ts); Ty = VADD(Tp, Ts); } { V T9, Te, T8, Td; T8 = LD(&(x[WS(rs, 1)]), ms, &(x[WS(rs, 1)])); T9 = VZMULJ(T2, T8); Td = LD(&(x[WS(rs, 5)]), ms, &(x[WS(rs, 1)])); Te = VZMULJ(Tc, Td); Tf = VSUB(T9, Te); TA = VADD(T9, Te); } { V Th, Tj, Tg, Ti; Tg = LD(&(x[WS(rs, 7)]), ms, &(x[WS(rs, 1)])); Th = VZMULJ(Tb, Tg); Ti = LD(&(x[WS(rs, 3)]), ms, &(x[WS(rs, 1)])); Tj = VZMULJ(T3, Ti); Tk = VSUB(Th, Tj); TB = VADD(Th, Tj); } { V Tz, TC, TD, TE; Tz = VADD(Tx, Ty); TC = VADD(TA, TB); ST(&(x[WS(rs, 4)]), VSUB(Tz, TC), ms, &(x[0])); ST(&(x[0]), VADD(Tz, TC), ms, &(x[0])); TD = VSUB(Tx, Ty); TE = VBYI(VSUB(TB, TA)); ST(&(x[WS(rs, 6)]), VSUB(TD, TE), ms, &(x[0])); ST(&(x[WS(rs, 2)]), VADD(TD, TE), ms, &(x[0])); { V Tm, Tv, Tu, Tw, Tl, Tn; Tl = VMUL(LDK(KP707106781), VADD(Tf, Tk)); Tm = VADD(T7, Tl); Tv = VSUB(T7, Tl); Tn = VMUL(LDK(KP707106781), VSUB(Tk, Tf)); Tu = VBYI(VSUB(Tn, Tt)); Tw = VBYI(VADD(Tt, Tn)); ST(&(x[WS(rs, 7)]), VSUB(Tm, Tu), ms, &(x[WS(rs, 1)])); ST(&(x[WS(rs, 3)]), VADD(Tv, Tw), ms, &(x[WS(rs, 1)])); ST(&(x[WS(rs, 1)]), VADD(Tm, Tu), ms, &(x[WS(rs, 1)])); ST(&(x[WS(rs, 5)]), VSUB(Tv, Tw), ms, &(x[WS(rs, 1)])); } } } } } VLEAVE(); } static const tw_instr twinstr[] = { VTW(0, 1), VTW(0, 3), VTW(0, 7), {TW_NEXT, VL, 0} }; static const ct_desc desc = { 8, XSIMD_STRING("t3fv_8"), twinstr, &GENUS, {37, 24, 0, 0}, 0, 0, 0 }; void XSIMD(codelet_t3fv_8) (planner *p) { X(kdft_dit_register) (p, t3fv_8, &desc); } #endif /* HAVE_FMA */ fftw-3.3.3/dft/simd/common/t3fv_16.c0000644000175000017500000003375712054410532013757 00000000000000/* * Copyright (c) 2003, 2007-11 Matteo Frigo * Copyright (c) 2003, 2007-11 Massachusetts Institute of Technology * * This program is free software; you can redistribute it and/or modify * it under the terms of the GNU General Public License as published by * the Free Software Foundation; either version 2 of the License, or * (at your option) any later version. * * This program is distributed in the hope that it will be useful, * but WITHOUT ANY WARRANTY; without even the implied warranty of * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the * GNU General Public License for more details. * * You should have received a copy of the GNU General Public License * along with this program; if not, write to the Free Software * Foundation, Inc., 51 Franklin Street, Fifth Floor, Boston, MA 02110-1301 USA * */ /* This file was automatically generated --- DO NOT EDIT */ /* Generated on Sun Nov 25 07:38:49 EST 2012 */ #include "codelet-dft.h" #ifdef HAVE_FMA /* Generated by: ../../../genfft/gen_twiddle_c.native -fma -reorder-insns -schedule-for-pipeline -simd -compact -variables 4 -pipeline-latency 8 -twiddle-log3 -precompute-twiddles -no-generate-bytw -n 16 -name t3fv_16 -include t3f.h */ /* * This function contains 98 FP additions, 86 FP multiplications, * (or, 64 additions, 52 multiplications, 34 fused multiply/add), * 70 stack variables, 3 constants, and 32 memory accesses */ #include "t3f.h" static void t3fv_16(R *ri, R *ii, const R *W, stride rs, INT mb, INT me, INT ms) { DVK(KP923879532, +0.923879532511286756128183189396788286822416626); DVK(KP414213562, +0.414213562373095048801688724209698078569671875); DVK(KP707106781, +0.707106781186547524400844362104849039284835938); { INT m; R *x; x = ri; for (m = mb, W = W + (mb * ((TWVL / VL) * 8)); m < me; m = m + VL, x = x + (VL * ms), W = W + (TWVL * 8), MAKE_VOLATILE_STRIDE(16, rs)) { V T13, Tg, TY, T14, T1A, T1q, T1f, T1x, T1r, T1i, Tt, T16, TB, T1j, T1k; V TH; { V T2, T8, Tu, T3; T2 = LDW(&(W[0])); T8 = LDW(&(W[TWVL * 2])); Tu = LDW(&(W[TWVL * 6])); T3 = LDW(&(W[TWVL * 4])); { V Ty, T1o, Tf, T1b, T7, Tr, TR, TX, T1g, Tl, To, Tw, TG, Tz, T1p; V T1e, TC; { V T1, T5, Ta, Td; T1 = LD(&(x[0]), ms, &(x[0])); T5 = LD(&(x[WS(rs, 8)]), ms, &(x[0])); Ta = LD(&(x[WS(rs, 4)]), ms, &(x[0])); Td = LD(&(x[WS(rs, 12)]), ms, &(x[0])); { V Tx, TO, TE, Tb, Tm, Tp, TN, Te, T6, TW, TP, TS; { V TM, T9, TL, Tc, TU, T4, TV; TM = LD(&(x[WS(rs, 14)]), ms, &(x[0])); Tx = VZMULJ(T2, T8); T9 = VZMUL(T2, T8); TL = VZMULJ(T2, Tu); TO = VZMULJ(T8, T3); Tc = VZMUL(T8, T3); TU = VZMUL(T2, T3); T4 = VZMULJ(T2, T3); TV = LD(&(x[WS(rs, 10)]), ms, &(x[0])); TE = VZMUL(Tx, T3); Ty = VZMULJ(Tx, T3); Tb = VZMULJ(T9, Ta); Tm = VZMULJ(T9, T3); Tp = VZMUL(T9, T3); TN = VZMULJ(TL, TM); Te = VZMULJ(Tc, Td); T6 = VZMULJ(T4, T5); TW = VZMULJ(TU, TV); } TP = LD(&(x[WS(rs, 6)]), ms, &(x[0])); TS = LD(&(x[WS(rs, 2)]), ms, &(x[0])); { V TQ, TT, Ti, Tk, Tn, Th, Tq, Tj; Th = LD(&(x[WS(rs, 1)]), ms, &(x[WS(rs, 1)])); Tq = LD(&(x[WS(rs, 13)]), ms, &(x[WS(rs, 1)])); Tj = LD(&(x[WS(rs, 9)]), ms, &(x[WS(rs, 1)])); T1o = VSUB(Tb, Te); Tf = VADD(Tb, Te); T1b = VSUB(T1, T6); T7 = VADD(T1, T6); TQ = VZMULJ(TO, TP); TT = VZMULJ(Tx, TS); Ti = VZMULJ(T2, Th); Tr = VZMULJ(Tp, Tq); Tk = VZMULJ(T3, Tj); Tn = LD(&(x[WS(rs, 5)]), ms, &(x[WS(rs, 1)])); { V T1d, T1c, Tv, TF; Tv = LD(&(x[WS(rs, 15)]), ms, &(x[WS(rs, 1)])); TF = LD(&(x[WS(rs, 11)]), ms, &(x[WS(rs, 1)])); T1d = VSUB(TN, TQ); TR = VADD(TN, TQ); T1c = VSUB(TT, TW); TX = VADD(TT, TW); T1g = VSUB(Ti, Tk); Tl = VADD(Ti, Tk); To = VZMULJ(Tm, Tn); Tw = VZMULJ(Tu, Tv); TG = VZMULJ(TE, TF); Tz = LD(&(x[WS(rs, 7)]), ms, &(x[WS(rs, 1)])); T1p = VSUB(T1d, T1c); T1e = VADD(T1c, T1d); TC = LD(&(x[WS(rs, 3)]), ms, &(x[WS(rs, 1)])); } } } } { V T1h, Ts, TA, TD; T13 = VADD(T7, Tf); Tg = VSUB(T7, Tf); T1h = VSUB(To, Tr); Ts = VADD(To, Tr); TY = VSUB(TR, TX); T14 = VADD(TX, TR); TA = VZMULJ(Ty, Tz); T1A = VFMA(LDK(KP707106781), T1p, T1o); T1q = VFNMS(LDK(KP707106781), T1p, T1o); T1f = VFMA(LDK(KP707106781), T1e, T1b); T1x = VFNMS(LDK(KP707106781), T1e, T1b); TD = VZMULJ(T8, TC); T1r = VFMA(LDK(KP414213562), T1g, T1h); T1i = VFNMS(LDK(KP414213562), T1h, T1g); Tt = VSUB(Tl, Ts); T16 = VADD(Tl, Ts); TB = VADD(Tw, TA); T1j = VSUB(Tw, TA); T1k = VSUB(TG, TD); TH = VADD(TD, TG); } } } { V T15, T19, T1l, T1s, TI, T17; T15 = VADD(T13, T14); T19 = VSUB(T13, T14); T1l = VFNMS(LDK(KP414213562), T1k, T1j); T1s = VFMA(LDK(KP414213562), T1j, T1k); TI = VSUB(TB, TH); T17 = VADD(TB, TH); { V T1y, T1t, T1B, T1m; T1y = VADD(T1r, T1s); T1t = VSUB(T1r, T1s); T1B = VSUB(T1l, T1i); T1m = VADD(T1i, T1l); { V T18, T1a, TJ, TZ; T18 = VADD(T16, T17); T1a = VSUB(T17, T16); TJ = VADD(Tt, TI); TZ = VSUB(TI, Tt); { V T1u, T1w, T1z, T1D; T1u = VFNMS(LDK(KP923879532), T1t, T1q); T1w = VFMA(LDK(KP923879532), T1t, T1q); T1z = VFNMS(LDK(KP923879532), T1y, T1x); T1D = VFMA(LDK(KP923879532), T1y, T1x); { V T1n, T1v, T1C, T1E; T1n = VFNMS(LDK(KP923879532), T1m, T1f); T1v = VFMA(LDK(KP923879532), T1m, T1f); T1C = VFNMS(LDK(KP923879532), T1B, T1A); T1E = VFMA(LDK(KP923879532), T1B, T1A); ST(&(x[WS(rs, 12)]), VFNMSI(T1a, T19), ms, &(x[0])); ST(&(x[WS(rs, 4)]), VFMAI(T1a, T19), ms, &(x[0])); ST(&(x[0]), VADD(T15, T18), ms, &(x[0])); ST(&(x[WS(rs, 8)]), VSUB(T15, T18), ms, &(x[0])); { V T10, T12, TK, T11; T10 = VFNMS(LDK(KP707106781), TZ, TY); T12 = VFMA(LDK(KP707106781), TZ, TY); TK = VFNMS(LDK(KP707106781), TJ, Tg); T11 = VFMA(LDK(KP707106781), TJ, Tg); ST(&(x[WS(rs, 1)]), VFNMSI(T1w, T1v), ms, &(x[WS(rs, 1)])); ST(&(x[WS(rs, 15)]), VFMAI(T1w, T1v), ms, &(x[WS(rs, 1)])); ST(&(x[WS(rs, 7)]), VFMAI(T1u, T1n), ms, &(x[WS(rs, 1)])); ST(&(x[WS(rs, 9)]), VFNMSI(T1u, T1n), ms, &(x[WS(rs, 1)])); ST(&(x[WS(rs, 3)]), VFMAI(T1E, T1D), ms, &(x[WS(rs, 1)])); ST(&(x[WS(rs, 13)]), VFNMSI(T1E, T1D), ms, &(x[WS(rs, 1)])); ST(&(x[WS(rs, 11)]), VFMAI(T1C, T1z), ms, &(x[WS(rs, 1)])); ST(&(x[WS(rs, 5)]), VFNMSI(T1C, T1z), ms, &(x[WS(rs, 1)])); ST(&(x[WS(rs, 14)]), VFNMSI(T12, T11), ms, &(x[0])); ST(&(x[WS(rs, 2)]), VFMAI(T12, T11), ms, &(x[0])); ST(&(x[WS(rs, 10)]), VFMAI(T10, TK), ms, &(x[0])); ST(&(x[WS(rs, 6)]), VFNMSI(T10, TK), ms, &(x[0])); } } } } } } } } VLEAVE(); } static const tw_instr twinstr[] = { VTW(0, 1), VTW(0, 3), VTW(0, 9), VTW(0, 15), {TW_NEXT, VL, 0} }; static const ct_desc desc = { 16, XSIMD_STRING("t3fv_16"), twinstr, &GENUS, {64, 52, 34, 0}, 0, 0, 0 }; void XSIMD(codelet_t3fv_16) (planner *p) { X(kdft_dit_register) (p, t3fv_16, &desc); } #else /* HAVE_FMA */ /* Generated by: ../../../genfft/gen_twiddle_c.native -simd -compact -variables 4 -pipeline-latency 8 -twiddle-log3 -precompute-twiddles -no-generate-bytw -n 16 -name t3fv_16 -include t3f.h */ /* * This function contains 98 FP additions, 64 FP multiplications, * (or, 94 additions, 60 multiplications, 4 fused multiply/add), * 51 stack variables, 3 constants, and 32 memory accesses */ #include "t3f.h" static void t3fv_16(R *ri, R *ii, const R *W, stride rs, INT mb, INT me, INT ms) { DVK(KP923879532, +0.923879532511286756128183189396788286822416626); DVK(KP382683432, +0.382683432365089771728459984030398866761344562); DVK(KP707106781, +0.707106781186547524400844362104849039284835938); { INT m; R *x; x = ri; for (m = mb, W = W + (mb * ((TWVL / VL) * 8)); m < me; m = m + VL, x = x + (VL * ms), W = W + (TWVL * 8), MAKE_VOLATILE_STRIDE(16, rs)) { V T4, T5, T6, To, T1, Ty, T7, T8, TO, TV, Te, Tp, TB, TH, Ts; T4 = LDW(&(W[0])); T5 = LDW(&(W[TWVL * 2])); T6 = VZMULJ(T4, T5); To = VZMUL(T4, T5); T1 = LDW(&(W[TWVL * 6])); Ty = VZMULJ(T4, T1); T7 = LDW(&(W[TWVL * 4])); T8 = VZMULJ(T6, T7); TO = VZMUL(T5, T7); TV = VZMULJ(T4, T7); Te = VZMUL(T6, T7); Tp = VZMULJ(To, T7); TB = VZMULJ(T5, T7); TH = VZMUL(T4, T7); Ts = VZMUL(To, T7); { V TY, T1f, TR, T1g, T1q, T1r, TL, TZ, T1l, T1m, T1n, Ti, T12, T1i, T1j; V T1k, Tw, T11, TU, TX, TW; TU = LD(&(x[0]), ms, &(x[0])); TW = LD(&(x[WS(rs, 8)]), ms, &(x[0])); TX = VZMULJ(TV, TW); TY = VSUB(TU, TX); T1f = VADD(TU, TX); { V TN, TQ, TM, TP; TM = LD(&(x[WS(rs, 4)]), ms, &(x[0])); TN = VZMULJ(To, TM); TP = LD(&(x[WS(rs, 12)]), ms, &(x[0])); TQ = VZMULJ(TO, TP); TR = VSUB(TN, TQ); T1g = VADD(TN, TQ); } { V TA, TJ, TD, TG, TE, TK; { V Tz, TI, TC, TF; Tz = LD(&(x[WS(rs, 14)]), ms, &(x[0])); TA = VZMULJ(Ty, Tz); TI = LD(&(x[WS(rs, 10)]), ms, &(x[0])); TJ = VZMULJ(TH, TI); TC = LD(&(x[WS(rs, 6)]), ms, &(x[0])); TD = VZMULJ(TB, TC); TF = LD(&(x[WS(rs, 2)]), ms, &(x[0])); TG = VZMULJ(T6, TF); } T1q = VADD(TA, TD); T1r = VADD(TG, TJ); TE = VSUB(TA, TD); TK = VSUB(TG, TJ); TL = VMUL(LDK(KP707106781), VSUB(TE, TK)); TZ = VMUL(LDK(KP707106781), VADD(TK, TE)); } { V T3, Tg, Ta, Td, Tb, Th; { V T2, Tf, T9, Tc; T2 = LD(&(x[WS(rs, 15)]), ms, &(x[WS(rs, 1)])); T3 = VZMULJ(T1, T2); Tf = LD(&(x[WS(rs, 11)]), ms, &(x[WS(rs, 1)])); Tg = VZMULJ(Te, Tf); T9 = LD(&(x[WS(rs, 7)]), ms, &(x[WS(rs, 1)])); Ta = VZMULJ(T8, T9); Tc = LD(&(x[WS(rs, 3)]), ms, &(x[WS(rs, 1)])); Td = VZMULJ(T5, Tc); } T1l = VADD(T3, Ta); T1m = VADD(Td, Tg); T1n = VSUB(T1l, T1m); Tb = VSUB(T3, Ta); Th = VSUB(Td, Tg); Ti = VFNMS(LDK(KP923879532), Th, VMUL(LDK(KP382683432), Tb)); T12 = VFMA(LDK(KP923879532), Tb, VMUL(LDK(KP382683432), Th)); } { V Tk, Tu, Tm, Tr, Tn, Tv; { V Tj, Tt, Tl, Tq; Tj = LD(&(x[WS(rs, 1)]), ms, &(x[WS(rs, 1)])); Tk = VZMULJ(T4, Tj); Tt = LD(&(x[WS(rs, 13)]), ms, &(x[WS(rs, 1)])); Tu = VZMULJ(Ts, Tt); Tl = LD(&(x[WS(rs, 9)]), ms, &(x[WS(rs, 1)])); Tm = VZMULJ(T7, Tl); Tq = LD(&(x[WS(rs, 5)]), ms, &(x[WS(rs, 1)])); Tr = VZMULJ(Tp, Tq); } T1i = VADD(Tk, Tm); T1j = VADD(Tr, Tu); T1k = VSUB(T1i, T1j); Tn = VSUB(Tk, Tm); Tv = VSUB(Tr, Tu); Tw = VFMA(LDK(KP382683432), Tn, VMUL(LDK(KP923879532), Tv)); T11 = VFNMS(LDK(KP382683432), Tv, VMUL(LDK(KP923879532), Tn)); } { V T1p, T1v, T1u, T1w; { V T1h, T1o, T1s, T1t; T1h = VSUB(T1f, T1g); T1o = VMUL(LDK(KP707106781), VADD(T1k, T1n)); T1p = VADD(T1h, T1o); T1v = VSUB(T1h, T1o); T1s = VSUB(T1q, T1r); T1t = VMUL(LDK(KP707106781), VSUB(T1n, T1k)); T1u = VBYI(VADD(T1s, T1t)); T1w = VBYI(VSUB(T1t, T1s)); } ST(&(x[WS(rs, 14)]), VSUB(T1p, T1u), ms, &(x[0])); ST(&(x[WS(rs, 6)]), VADD(T1v, T1w), ms, &(x[0])); ST(&(x[WS(rs, 2)]), VADD(T1p, T1u), ms, &(x[0])); ST(&(x[WS(rs, 10)]), VSUB(T1v, T1w), ms, &(x[0])); } { V T1z, T1D, T1C, T1E; { V T1x, T1y, T1A, T1B; T1x = VADD(T1f, T1g); T1y = VADD(T1r, T1q); T1z = VADD(T1x, T1y); T1D = VSUB(T1x, T1y); T1A = VADD(T1i, T1j); T1B = VADD(T1l, T1m); T1C = VADD(T1A, T1B); T1E = VBYI(VSUB(T1B, T1A)); } ST(&(x[WS(rs, 8)]), VSUB(T1z, T1C), ms, &(x[0])); ST(&(x[WS(rs, 4)]), VADD(T1D, T1E), ms, &(x[0])); ST(&(x[0]), VADD(T1z, T1C), ms, &(x[0])); ST(&(x[WS(rs, 12)]), VSUB(T1D, T1E), ms, &(x[0])); } { V TT, T15, T14, T16; { V Tx, TS, T10, T13; Tx = VSUB(Ti, Tw); TS = VSUB(TL, TR); TT = VBYI(VSUB(Tx, TS)); T15 = VBYI(VADD(TS, Tx)); T10 = VADD(TY, TZ); T13 = VADD(T11, T12); T14 = VSUB(T10, T13); T16 = VADD(T10, T13); } ST(&(x[WS(rs, 7)]), VADD(TT, T14), ms, &(x[WS(rs, 1)])); ST(&(x[WS(rs, 15)]), VSUB(T16, T15), ms, &(x[WS(rs, 1)])); ST(&(x[WS(rs, 9)]), VSUB(T14, TT), ms, &(x[WS(rs, 1)])); ST(&(x[WS(rs, 1)]), VADD(T15, T16), ms, &(x[WS(rs, 1)])); } { V T19, T1d, T1c, T1e; { V T17, T18, T1a, T1b; T17 = VSUB(TY, TZ); T18 = VADD(Tw, Ti); T19 = VADD(T17, T18); T1d = VSUB(T17, T18); T1a = VADD(TR, TL); T1b = VSUB(T12, T11); T1c = VBYI(VADD(T1a, T1b)); T1e = VBYI(VSUB(T1b, T1a)); } ST(&(x[WS(rs, 13)]), VSUB(T19, T1c), ms, &(x[WS(rs, 1)])); ST(&(x[WS(rs, 5)]), VADD(T1d, T1e), ms, &(x[WS(rs, 1)])); ST(&(x[WS(rs, 3)]), VADD(T19, T1c), ms, &(x[WS(rs, 1)])); ST(&(x[WS(rs, 11)]), VSUB(T1d, T1e), ms, &(x[WS(rs, 1)])); } } } } VLEAVE(); } static const tw_instr twinstr[] = { VTW(0, 1), VTW(0, 3), VTW(0, 9), VTW(0, 15), {TW_NEXT, VL, 0} }; static const ct_desc desc = { 16, XSIMD_STRING("t3fv_16"), twinstr, &GENUS, {94, 60, 4, 0}, 0, 0, 0 }; void XSIMD(codelet_t3fv_16) (planner *p) { X(kdft_dit_register) (p, t3fv_16, &desc); } #endif /* HAVE_FMA */ fftw-3.3.3/dft/simd/common/t3fv_32.c0000644000175000017500000007443012054410536013752 00000000000000/* * Copyright (c) 2003, 2007-11 Matteo Frigo * Copyright (c) 2003, 2007-11 Massachusetts Institute of Technology * * This program is free software; you can redistribute it and/or modify * it under the terms of the GNU General Public License as published by * the Free Software Foundation; either version 2 of the License, or * (at your option) any later version. * * This program is distributed in the hope that it will be useful, * but WITHOUT ANY WARRANTY; without even the implied warranty of * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the * GNU General Public License for more details. * * You should have received a copy of the GNU General Public License * along with this program; if not, write to the Free Software * Foundation, Inc., 51 Franklin Street, Fifth Floor, Boston, MA 02110-1301 USA * */ /* This file was automatically generated --- DO NOT EDIT */ /* Generated on Sun Nov 25 07:38:50 EST 2012 */ #include "codelet-dft.h" #ifdef HAVE_FMA /* Generated by: ../../../genfft/gen_twiddle_c.native -fma -reorder-insns -schedule-for-pipeline -simd -compact -variables 4 -pipeline-latency 8 -twiddle-log3 -precompute-twiddles -no-generate-bytw -n 32 -name t3fv_32 -include t3f.h */ /* * This function contains 244 FP additions, 214 FP multiplications, * (or, 146 additions, 116 multiplications, 98 fused multiply/add), * 118 stack variables, 7 constants, and 64 memory accesses */ #include "t3f.h" static void t3fv_32(R *ri, R *ii, const R *W, stride rs, INT mb, INT me, INT ms) { DVK(KP831469612, +0.831469612302545237078788377617905756738560812); DVK(KP980785280, +0.980785280403230449126182236134239036973933731); DVK(KP668178637, +0.668178637919298919997757686523080761552472251); DVK(KP198912367, +0.198912367379658006911597622644676228597850501); DVK(KP923879532, +0.923879532511286756128183189396788286822416626); DVK(KP707106781, +0.707106781186547524400844362104849039284835938); DVK(KP414213562, +0.414213562373095048801688724209698078569671875); { INT m; R *x; x = ri; for (m = mb, W = W + (mb * ((TWVL / VL) * 8)); m < me; m = m + VL, x = x + (VL * ms), W = W + (TWVL * 8), MAKE_VOLATILE_STRIDE(32, rs)) { V T2B, T2A, T2u, T2x, T2r, T2F, T2L, T2P; { V T2, T5, T3, T7; T2 = LDW(&(W[0])); T5 = LDW(&(W[TWVL * 4])); T3 = LDW(&(W[TWVL * 2])); T7 = LDW(&(W[TWVL * 6])); { V T24, Tb, T3x, T2T, T3K, T2W, T25, Tr, T3z, T3g, T28, TX, T3y, T3j, T27; V TG, T37, T3F, T3G, T3a, T2Y, T15, T1p, T2Z, T2w, T1V, T2v, T1N, T32, T1h; V T17, T1a; { V T1, Tz, TT, T4, TC, Tv, T12, T1D, T1w, T18, T1t, T1O, TK, TP, T1c; V T1m, Tf, T6, Te, TL, TQ, T2S, Tp, TU, Ti, Ta, TM, TR, Tm, TJ; V T22, T9, T1Z; T1 = LD(&(x[0]), ms, &(x[0])); T22 = LD(&(x[WS(rs, 24)]), ms, &(x[0])); T9 = LD(&(x[WS(rs, 16)]), ms, &(x[0])); T1Z = LD(&(x[WS(rs, 8)]), ms, &(x[0])); { V Tn, TH, Tk, To, Th, Tg, T8, Tl, T20, T23, TI; { V Td, T1C, Tc, T21; Td = LD(&(x[WS(rs, 4)]), ms, &(x[0])); Tz = VZMUL(T2, T5); T1C = VZMULJ(T2, T5); Tn = VZMUL(T3, T5); TT = VZMULJ(T3, T5); Tc = VZMUL(T2, T3); T4 = VZMULJ(T2, T3); TH = VZMUL(T3, T7); T21 = VZMULJ(T3, T7); Tk = VZMUL(T2, T7); TC = VZMULJ(T2, T7); Tv = VZMULJ(T5, T7); T12 = VZMULJ(Tz, T7); T20 = VZMULJ(T1C, T1Z); T1D = VZMULJ(T1C, T7); T1w = VZMULJ(Tn, T7); T18 = VZMULJ(TT, T7); T1t = VZMUL(Tc, T7); T1O = VZMULJ(Tc, T7); TK = VZMUL(Tc, T5); TP = VZMULJ(Tc, T5); T1c = VZMUL(T4, T7); T1m = VZMULJ(T4, T7); Tf = VZMULJ(T4, T5); T6 = VZMUL(T4, T5); T23 = VZMULJ(T21, T22); Te = VZMULJ(Tc, Td); } TL = VZMULJ(TK, T7); TQ = VZMULJ(TP, T7); To = LD(&(x[WS(rs, 12)]), ms, &(x[0])); Th = LD(&(x[WS(rs, 20)]), ms, &(x[0])); Tg = VZMULJ(Tf, T7); T8 = VZMULJ(T6, T7); T2S = VADD(T20, T23); T24 = VSUB(T20, T23); Tl = LD(&(x[WS(rs, 28)]), ms, &(x[0])); TI = LD(&(x[WS(rs, 30)]), ms, &(x[0])); Tp = VZMULJ(Tn, To); TU = LD(&(x[WS(rs, 6)]), ms, &(x[0])); Ti = VZMULJ(Tg, Th); Ta = VZMULJ(T8, T9); TM = LD(&(x[WS(rs, 14)]), ms, &(x[0])); TR = LD(&(x[WS(rs, 22)]), ms, &(x[0])); Tm = VZMULJ(Tk, Tl); TJ = VZMULJ(TH, TI); } { V Tu, TE, Tw, TA; { V T3e, TO, T3f, TW; { V TV, T2U, Tj, T2R, TN, TS, T2V, Tq, Tt, TD; Tt = LD(&(x[WS(rs, 2)]), ms, &(x[0])); TV = VZMULJ(TT, TU); T2U = VADD(Te, Ti); Tj = VSUB(Te, Ti); T2R = VADD(T1, Ta); Tb = VSUB(T1, Ta); TN = VZMULJ(TL, TM); TS = VZMULJ(TQ, TR); T2V = VADD(Tm, Tp); Tq = VSUB(Tm, Tp); Tu = VZMULJ(T4, Tt); TD = LD(&(x[WS(rs, 26)]), ms, &(x[0])); T3x = VSUB(T2R, T2S); T2T = VADD(T2R, T2S); T3e = VADD(TJ, TN); TO = VSUB(TJ, TN); T3f = VADD(TV, TS); TW = VSUB(TS, TV); T3K = VSUB(T2V, T2U); T2W = VADD(T2U, T2V); T25 = VSUB(Tq, Tj); Tr = VADD(Tj, Tq); TE = VZMULJ(TC, TD); } Tw = LD(&(x[WS(rs, 18)]), ms, &(x[0])); T3z = VSUB(T3e, T3f); T3g = VADD(T3e, T3f); T28 = VFMA(LDK(KP414213562), TO, TW); TX = VFNMS(LDK(KP414213562), TW, TO); TA = LD(&(x[WS(rs, 10)]), ms, &(x[0])); } { V T35, T1z, T1T, T36, T39, T1L, T1B, T1F; { V T1v, T1y, Ty, T3h, T1S, T1Q, T1I, T3i, TF, T1K, T1A, T1E; { V T1u, T1x, Tx, T1R; T1u = LD(&(x[WS(rs, 31)]), ms, &(x[WS(rs, 1)])); T1x = LD(&(x[WS(rs, 15)]), ms, &(x[WS(rs, 1)])); Tx = VZMULJ(Tv, Tw); T1R = LD(&(x[WS(rs, 7)]), ms, &(x[WS(rs, 1)])); { V T1P, T1H, T1J, TB; T1P = LD(&(x[WS(rs, 23)]), ms, &(x[WS(rs, 1)])); T1H = LD(&(x[WS(rs, 27)]), ms, &(x[WS(rs, 1)])); T1J = LD(&(x[WS(rs, 11)]), ms, &(x[WS(rs, 1)])); TB = VZMULJ(Tz, TA); T1v = VZMULJ(T1t, T1u); T1y = VZMULJ(T1w, T1x); Ty = VSUB(Tu, Tx); T3h = VADD(Tu, Tx); T1S = VZMULJ(Tf, T1R); T1Q = VZMULJ(T1O, T1P); T1I = VZMULJ(T7, T1H); T3i = VADD(TB, TE); TF = VSUB(TB, TE); T1K = VZMULJ(T6, T1J); T1A = LD(&(x[WS(rs, 3)]), ms, &(x[WS(rs, 1)])); T1E = LD(&(x[WS(rs, 19)]), ms, &(x[WS(rs, 1)])); } } T35 = VADD(T1v, T1y); T1z = VSUB(T1v, T1y); T1T = VSUB(T1Q, T1S); T36 = VADD(T1S, T1Q); T3y = VSUB(T3h, T3i); T3j = VADD(T3h, T3i); T27 = VFMA(LDK(KP414213562), Ty, TF); TG = VFNMS(LDK(KP414213562), TF, Ty); T39 = VADD(T1I, T1K); T1L = VSUB(T1I, T1K); T1B = VZMULJ(T3, T1A); T1F = VZMULJ(T1D, T1E); } { V T11, T14, T1o, T1l, T1e, T1U, T1M, T1g, T16, T19; { V T10, T13, T1n, T1k; T10 = LD(&(x[WS(rs, 1)]), ms, &(x[WS(rs, 1)])); T13 = LD(&(x[WS(rs, 17)]), ms, &(x[WS(rs, 1)])); T1n = LD(&(x[WS(rs, 25)]), ms, &(x[WS(rs, 1)])); T1k = LD(&(x[WS(rs, 9)]), ms, &(x[WS(rs, 1)])); { V T1d, T1f, T1G, T38; T1d = LD(&(x[WS(rs, 29)]), ms, &(x[WS(rs, 1)])); T1f = LD(&(x[WS(rs, 13)]), ms, &(x[WS(rs, 1)])); T1G = VSUB(T1B, T1F); T38 = VADD(T1B, T1F); T37 = VADD(T35, T36); T3F = VSUB(T35, T36); T11 = VZMULJ(T2, T10); T14 = VZMULJ(T12, T13); T1o = VZMULJ(T1m, T1n); T1l = VZMULJ(T5, T1k); T1e = VZMULJ(T1c, T1d); T3G = VSUB(T39, T38); T3a = VADD(T38, T39); T1U = VSUB(T1L, T1G); T1M = VADD(T1G, T1L); T1g = VZMULJ(TK, T1f); } T16 = LD(&(x[WS(rs, 5)]), ms, &(x[WS(rs, 1)])); T19 = LD(&(x[WS(rs, 21)]), ms, &(x[WS(rs, 1)])); } T2Y = VADD(T11, T14); T15 = VSUB(T11, T14); T1p = VSUB(T1l, T1o); T2Z = VADD(T1l, T1o); T2w = VFNMS(LDK(KP707106781), T1U, T1T); T1V = VFMA(LDK(KP707106781), T1U, T1T); T2v = VFNMS(LDK(KP707106781), T1M, T1z); T1N = VFMA(LDK(KP707106781), T1M, T1z); T32 = VADD(T1e, T1g); T1h = VSUB(T1e, T1g); T17 = VZMULJ(TP, T16); T1a = VZMULJ(T18, T19); } } } } { V T2X, T3k, T3b, T3t, T1b, T31, T30, T3C, T3r, T3v, T3p, T3q; T2X = VSUB(T2T, T2W); T3p = VADD(T2T, T2W); T3q = VADD(T3j, T3g); T3k = VSUB(T3g, T3j); T3b = VSUB(T37, T3a); T3t = VADD(T37, T3a); T1b = VSUB(T17, T1a); T31 = VADD(T17, T1a); T30 = VADD(T2Y, T2Z); T3C = VSUB(T2Y, T2Z); T3r = VADD(T3p, T3q); T3v = VSUB(T3p, T3q); { V T3N, T3B, T3T, T3M, T3W, T3O, T2t, T1r, T2s, T1j, T3I, T3X, T3c, T3l, T3u; V T3w; { V T3L, T3A, T33, T3D, T1i, T1q; T3L = VSUB(T3z, T3y); T3A = VADD(T3y, T3z); T33 = VADD(T31, T32); T3D = VSUB(T31, T32); T1i = VADD(T1b, T1h); T1q = VSUB(T1b, T1h); { V T3H, T3E, T34, T3s; T3N = VFMA(LDK(KP414213562), T3F, T3G); T3H = VFNMS(LDK(KP414213562), T3G, T3F); T3B = VFMA(LDK(KP707106781), T3A, T3x); T3T = VFNMS(LDK(KP707106781), T3A, T3x); T3M = VFMA(LDK(KP707106781), T3L, T3K); T3W = VFNMS(LDK(KP707106781), T3L, T3K); T3O = VFMA(LDK(KP414213562), T3C, T3D); T3E = VFNMS(LDK(KP414213562), T3D, T3C); T34 = VSUB(T30, T33); T3s = VADD(T30, T33); T2t = VFNMS(LDK(KP707106781), T1q, T1p); T1r = VFMA(LDK(KP707106781), T1q, T1p); T2s = VFNMS(LDK(KP707106781), T1i, T15); T1j = VFMA(LDK(KP707106781), T1i, T15); T3I = VADD(T3E, T3H); T3X = VSUB(T3H, T3E); T3c = VADD(T34, T3b); T3l = VSUB(T3b, T34); T3u = VADD(T3s, T3t); T3w = VSUB(T3t, T3s); } } { V T2p, Ts, TY, T1s, T2b, T2c, T1W, T26, T29, T2q, T3U, T3P, T2J, T2K; T2p = VFNMS(LDK(KP707106781), Tr, Tb); Ts = VFMA(LDK(KP707106781), Tr, Tb); T3U = VADD(T3O, T3N); T3P = VSUB(T3N, T3O); { V T3Y, T40, T3R, T3J; T3Y = VFMA(LDK(KP923879532), T3X, T3W); T40 = VFNMS(LDK(KP923879532), T3X, T3W); T3R = VFMA(LDK(KP923879532), T3I, T3B); T3J = VFNMS(LDK(KP923879532), T3I, T3B); { V T3o, T3m, T3n, T3d; T3o = VFMA(LDK(KP707106781), T3l, T3k); T3m = VFNMS(LDK(KP707106781), T3l, T3k); T3n = VFMA(LDK(KP707106781), T3c, T2X); T3d = VFNMS(LDK(KP707106781), T3c, T2X); ST(&(x[WS(rs, 24)]), VFNMSI(T3w, T3v), ms, &(x[0])); ST(&(x[WS(rs, 8)]), VFMAI(T3w, T3v), ms, &(x[0])); ST(&(x[0]), VADD(T3r, T3u), ms, &(x[0])); ST(&(x[WS(rs, 16)]), VSUB(T3r, T3u), ms, &(x[0])); { V T3V, T3Z, T3S, T3Q; T3V = VFNMS(LDK(KP923879532), T3U, T3T); T3Z = VFMA(LDK(KP923879532), T3U, T3T); T3S = VFMA(LDK(KP923879532), T3P, T3M); T3Q = VFNMS(LDK(KP923879532), T3P, T3M); ST(&(x[WS(rs, 4)]), VFMAI(T3o, T3n), ms, &(x[0])); ST(&(x[WS(rs, 28)]), VFNMSI(T3o, T3n), ms, &(x[0])); ST(&(x[WS(rs, 20)]), VFMAI(T3m, T3d), ms, &(x[0])); ST(&(x[WS(rs, 12)]), VFNMSI(T3m, T3d), ms, &(x[0])); ST(&(x[WS(rs, 22)]), VFNMSI(T3Y, T3V), ms, &(x[0])); ST(&(x[WS(rs, 10)]), VFMAI(T3Y, T3V), ms, &(x[0])); ST(&(x[WS(rs, 26)]), VFMAI(T40, T3Z), ms, &(x[0])); ST(&(x[WS(rs, 6)]), VFNMSI(T40, T3Z), ms, &(x[0])); ST(&(x[WS(rs, 2)]), VFMAI(T3S, T3R), ms, &(x[0])); ST(&(x[WS(rs, 30)]), VFNMSI(T3S, T3R), ms, &(x[0])); ST(&(x[WS(rs, 18)]), VFMAI(T3Q, T3J), ms, &(x[0])); ST(&(x[WS(rs, 14)]), VFNMSI(T3Q, T3J), ms, &(x[0])); TY = VADD(TG, TX); T2B = VSUB(TX, TG); } } } T1s = VFNMS(LDK(KP198912367), T1r, T1j); T2b = VFMA(LDK(KP198912367), T1j, T1r); T2c = VFMA(LDK(KP198912367), T1N, T1V); T1W = VFNMS(LDK(KP198912367), T1V, T1N); T2A = VFMA(LDK(KP707106781), T25, T24); T26 = VFNMS(LDK(KP707106781), T25, T24); T29 = VSUB(T27, T28); T2q = VADD(T27, T28); { V T2j, T2n, T1Y, T2f, T2o, T2m, T2e, T2g; { V T2h, TZ, T2i, T2d, T2l, T1X, T2k, T2a, T2D, T2E; T2h = VFNMS(LDK(KP923879532), TY, Ts); TZ = VFMA(LDK(KP923879532), TY, Ts); T2i = VADD(T2b, T2c); T2d = VSUB(T2b, T2c); T2l = VSUB(T1W, T1s); T1X = VADD(T1s, T1W); T2k = VFNMS(LDK(KP923879532), T29, T26); T2a = VFMA(LDK(KP923879532), T29, T26); T2u = VFMA(LDK(KP668178637), T2t, T2s); T2D = VFNMS(LDK(KP668178637), T2s, T2t); T2j = VFNMS(LDK(KP980785280), T2i, T2h); T2n = VFMA(LDK(KP980785280), T2i, T2h); T2E = VFNMS(LDK(KP668178637), T2v, T2w); T2x = VFMA(LDK(KP668178637), T2w, T2v); T1Y = VFNMS(LDK(KP980785280), T1X, TZ); T2f = VFMA(LDK(KP980785280), T1X, TZ); T2o = VFMA(LDK(KP980785280), T2l, T2k); T2m = VFNMS(LDK(KP980785280), T2l, T2k); T2e = VFNMS(LDK(KP980785280), T2d, T2a); T2g = VFMA(LDK(KP980785280), T2d, T2a); T2r = VFMA(LDK(KP923879532), T2q, T2p); T2J = VFNMS(LDK(KP923879532), T2q, T2p); T2K = VADD(T2D, T2E); T2F = VSUB(T2D, T2E); } ST(&(x[WS(rs, 23)]), VFMAI(T2m, T2j), ms, &(x[WS(rs, 1)])); ST(&(x[WS(rs, 9)]), VFNMSI(T2m, T2j), ms, &(x[WS(rs, 1)])); ST(&(x[WS(rs, 25)]), VFNMSI(T2o, T2n), ms, &(x[WS(rs, 1)])); ST(&(x[WS(rs, 7)]), VFMAI(T2o, T2n), ms, &(x[WS(rs, 1)])); ST(&(x[WS(rs, 31)]), VFMAI(T2g, T2f), ms, &(x[WS(rs, 1)])); ST(&(x[WS(rs, 1)]), VFNMSI(T2g, T2f), ms, &(x[WS(rs, 1)])); ST(&(x[WS(rs, 15)]), VFMAI(T2e, T1Y), ms, &(x[WS(rs, 1)])); ST(&(x[WS(rs, 17)]), VFNMSI(T2e, T1Y), ms, &(x[WS(rs, 1)])); } T2L = VFMA(LDK(KP831469612), T2K, T2J); T2P = VFNMS(LDK(KP831469612), T2K, T2J); } } } } } { V T2y, T2N, T2C, T2M; T2y = VADD(T2u, T2x); T2N = VSUB(T2x, T2u); T2C = VFMA(LDK(KP923879532), T2B, T2A); T2M = VFNMS(LDK(KP923879532), T2B, T2A); { V T2z, T2H, T2Q, T2O, T2G, T2I; T2z = VFNMS(LDK(KP831469612), T2y, T2r); T2H = VFMA(LDK(KP831469612), T2y, T2r); T2Q = VFNMS(LDK(KP831469612), T2N, T2M); T2O = VFMA(LDK(KP831469612), T2N, T2M); T2G = VFNMS(LDK(KP831469612), T2F, T2C); T2I = VFMA(LDK(KP831469612), T2F, T2C); ST(&(x[WS(rs, 21)]), VFNMSI(T2O, T2L), ms, &(x[WS(rs, 1)])); ST(&(x[WS(rs, 11)]), VFMAI(T2O, T2L), ms, &(x[WS(rs, 1)])); ST(&(x[WS(rs, 27)]), VFMAI(T2Q, T2P), ms, &(x[WS(rs, 1)])); ST(&(x[WS(rs, 5)]), VFNMSI(T2Q, T2P), ms, &(x[WS(rs, 1)])); ST(&(x[WS(rs, 3)]), VFMAI(T2I, T2H), ms, &(x[WS(rs, 1)])); ST(&(x[WS(rs, 29)]), VFNMSI(T2I, T2H), ms, &(x[WS(rs, 1)])); ST(&(x[WS(rs, 19)]), VFMAI(T2G, T2z), ms, &(x[WS(rs, 1)])); ST(&(x[WS(rs, 13)]), VFNMSI(T2G, T2z), ms, &(x[WS(rs, 1)])); } } } } VLEAVE(); } static const tw_instr twinstr[] = { VTW(0, 1), VTW(0, 3), VTW(0, 9), VTW(0, 27), {TW_NEXT, VL, 0} }; static const ct_desc desc = { 32, XSIMD_STRING("t3fv_32"), twinstr, &GENUS, {146, 116, 98, 0}, 0, 0, 0 }; void XSIMD(codelet_t3fv_32) (planner *p) { X(kdft_dit_register) (p, t3fv_32, &desc); } #else /* HAVE_FMA */ /* Generated by: ../../../genfft/gen_twiddle_c.native -simd -compact -variables 4 -pipeline-latency 8 -twiddle-log3 -precompute-twiddles -no-generate-bytw -n 32 -name t3fv_32 -include t3f.h */ /* * This function contains 244 FP additions, 158 FP multiplications, * (or, 228 additions, 142 multiplications, 16 fused multiply/add), * 90 stack variables, 7 constants, and 64 memory accesses */ #include "t3f.h" static void t3fv_32(R *ri, R *ii, const R *W, stride rs, INT mb, INT me, INT ms) { DVK(KP555570233, +0.555570233019602224742830813948532874374937191); DVK(KP831469612, +0.831469612302545237078788377617905756738560812); DVK(KP195090322, +0.195090322016128267848284868477022240927691618); DVK(KP980785280, +0.980785280403230449126182236134239036973933731); DVK(KP382683432, +0.382683432365089771728459984030398866761344562); DVK(KP923879532, +0.923879532511286756128183189396788286822416626); DVK(KP707106781, +0.707106781186547524400844362104849039284835938); { INT m; R *x; x = ri; for (m = mb, W = W + (mb * ((TWVL / VL) * 8)); m < me; m = m + VL, x = x + (VL * ms), W = W + (TWVL * 8), MAKE_VOLATILE_STRIDE(32, rs)) { V T2, T5, T3, T4, Tc, T1C, TP, Tz, Tn, T6, TS, Tf, TK, T7, T8; V Tv, T1w, T22, Tg, Tk, T1D, T1R, TC, T18, T12, T1t, TH, TL, TT, T1n; V T1c; T2 = LDW(&(W[0])); T5 = LDW(&(W[TWVL * 4])); T3 = LDW(&(W[TWVL * 2])); T4 = VZMULJ(T2, T3); Tc = VZMUL(T2, T3); T1C = VZMULJ(T2, T5); TP = VZMULJ(T3, T5); Tz = VZMUL(T2, T5); Tn = VZMUL(T3, T5); T6 = VZMUL(T4, T5); TS = VZMULJ(Tc, T5); Tf = VZMULJ(T4, T5); TK = VZMUL(Tc, T5); T7 = LDW(&(W[TWVL * 6])); T8 = VZMULJ(T6, T7); Tv = VZMULJ(T5, T7); T1w = VZMULJ(Tn, T7); T22 = VZMULJ(T3, T7); Tg = VZMULJ(Tf, T7); Tk = VZMUL(T2, T7); T1D = VZMULJ(T1C, T7); T1R = VZMULJ(Tc, T7); TC = VZMULJ(T2, T7); T18 = VZMULJ(TP, T7); T12 = VZMULJ(Tz, T7); T1t = VZMUL(Tc, T7); TH = VZMUL(T3, T7); TL = VZMULJ(TK, T7); TT = VZMULJ(TS, T7); T1n = VZMULJ(T4, T7); T1c = VZMUL(T4, T7); { V Tb, T25, T2T, T3x, Tr, T1Z, T2W, T3K, TX, T27, T3g, T3z, TG, T28, T3j; V T3y, T1N, T2v, T3a, T3G, T1V, T2w, T37, T3F, T1j, T2s, T33, T3D, T1r, T2t; V T30, T3C; { V T1, T24, Ta, T21, T23, T9, T20, T2R, T2S; T1 = LD(&(x[0]), ms, &(x[0])); T23 = LD(&(x[WS(rs, 24)]), ms, &(x[0])); T24 = VZMULJ(T22, T23); T9 = LD(&(x[WS(rs, 16)]), ms, &(x[0])); Ta = VZMULJ(T8, T9); T20 = LD(&(x[WS(rs, 8)]), ms, &(x[0])); T21 = VZMULJ(T1C, T20); Tb = VSUB(T1, Ta); T25 = VSUB(T21, T24); T2R = VADD(T1, Ta); T2S = VADD(T21, T24); T2T = VADD(T2R, T2S); T3x = VSUB(T2R, T2S); } { V Te, Tp, Ti, Tm; { V Td, To, Th, Tl; Td = LD(&(x[WS(rs, 4)]), ms, &(x[0])); Te = VZMULJ(Tc, Td); To = LD(&(x[WS(rs, 12)]), ms, &(x[0])); Tp = VZMULJ(Tn, To); Th = LD(&(x[WS(rs, 20)]), ms, &(x[0])); Ti = VZMULJ(Tg, Th); Tl = LD(&(x[WS(rs, 28)]), ms, &(x[0])); Tm = VZMULJ(Tk, Tl); } { V Tj, Tq, T2U, T2V; Tj = VSUB(Te, Ti); Tq = VSUB(Tm, Tp); Tr = VMUL(LDK(KP707106781), VADD(Tj, Tq)); T1Z = VMUL(LDK(KP707106781), VSUB(Tq, Tj)); T2U = VADD(Te, Ti); T2V = VADD(Tm, Tp); T2W = VADD(T2U, T2V); T3K = VSUB(T2V, T2U); } } { V TJ, TV, TN, TR; { V TI, TU, TM, TQ; TI = LD(&(x[WS(rs, 30)]), ms, &(x[0])); TJ = VZMULJ(TH, TI); TU = LD(&(x[WS(rs, 22)]), ms, &(x[0])); TV = VZMULJ(TT, TU); TM = LD(&(x[WS(rs, 14)]), ms, &(x[0])); TN = VZMULJ(TL, TM); TQ = LD(&(x[WS(rs, 6)]), ms, &(x[0])); TR = VZMULJ(TP, TQ); } { V TO, TW, T3e, T3f; TO = VSUB(TJ, TN); TW = VSUB(TR, TV); TX = VFMA(LDK(KP923879532), TO, VMUL(LDK(KP382683432), TW)); T27 = VFNMS(LDK(KP923879532), TW, VMUL(LDK(KP382683432), TO)); T3e = VADD(TJ, TN); T3f = VADD(TR, TV); T3g = VADD(T3e, T3f); T3z = VSUB(T3e, T3f); } } { V Tu, TE, Tx, TB; { V Tt, TD, Tw, TA; Tt = LD(&(x[WS(rs, 2)]), ms, &(x[0])); Tu = VZMULJ(T4, Tt); TD = LD(&(x[WS(rs, 26)]), ms, &(x[0])); TE = VZMULJ(TC, TD); Tw = LD(&(x[WS(rs, 18)]), ms, &(x[0])); Tx = VZMULJ(Tv, Tw); TA = LD(&(x[WS(rs, 10)]), ms, &(x[0])); TB = VZMULJ(Tz, TA); } { V Ty, TF, T3h, T3i; Ty = VSUB(Tu, Tx); TF = VSUB(TB, TE); TG = VFNMS(LDK(KP382683432), TF, VMUL(LDK(KP923879532), Ty)); T28 = VFMA(LDK(KP382683432), Ty, VMUL(LDK(KP923879532), TF)); T3h = VADD(Tu, Tx); T3i = VADD(TB, TE); T3j = VADD(T3h, T3i); T3y = VSUB(T3h, T3i); } } { V T1v, T1y, T1T, T1Q, T1I, T1K, T1L, T1B, T1F, T1G; { V T1u, T1x, T1S, T1P; T1u = LD(&(x[WS(rs, 31)]), ms, &(x[WS(rs, 1)])); T1v = VZMULJ(T1t, T1u); T1x = LD(&(x[WS(rs, 15)]), ms, &(x[WS(rs, 1)])); T1y = VZMULJ(T1w, T1x); T1S = LD(&(x[WS(rs, 23)]), ms, &(x[WS(rs, 1)])); T1T = VZMULJ(T1R, T1S); T1P = LD(&(x[WS(rs, 7)]), ms, &(x[WS(rs, 1)])); T1Q = VZMULJ(Tf, T1P); { V T1H, T1J, T1A, T1E; T1H = LD(&(x[WS(rs, 27)]), ms, &(x[WS(rs, 1)])); T1I = VZMULJ(T7, T1H); T1J = LD(&(x[WS(rs, 11)]), ms, &(x[WS(rs, 1)])); T1K = VZMULJ(T6, T1J); T1L = VSUB(T1I, T1K); T1A = LD(&(x[WS(rs, 3)]), ms, &(x[WS(rs, 1)])); T1B = VZMULJ(T3, T1A); T1E = LD(&(x[WS(rs, 19)]), ms, &(x[WS(rs, 1)])); T1F = VZMULJ(T1D, T1E); T1G = VSUB(T1B, T1F); } } { V T1z, T1M, T38, T39; T1z = VSUB(T1v, T1y); T1M = VMUL(LDK(KP707106781), VADD(T1G, T1L)); T1N = VADD(T1z, T1M); T2v = VSUB(T1z, T1M); T38 = VADD(T1B, T1F); T39 = VADD(T1I, T1K); T3a = VADD(T38, T39); T3G = VSUB(T39, T38); } { V T1O, T1U, T35, T36; T1O = VMUL(LDK(KP707106781), VSUB(T1L, T1G)); T1U = VSUB(T1Q, T1T); T1V = VSUB(T1O, T1U); T2w = VADD(T1U, T1O); T35 = VADD(T1v, T1y); T36 = VADD(T1Q, T1T); T37 = VADD(T35, T36); T3F = VSUB(T35, T36); } } { V T11, T14, T1p, T1m, T1e, T1g, T1h, T17, T1a, T1b; { V T10, T13, T1o, T1l; T10 = LD(&(x[WS(rs, 1)]), ms, &(x[WS(rs, 1)])); T11 = VZMULJ(T2, T10); T13 = LD(&(x[WS(rs, 17)]), ms, &(x[WS(rs, 1)])); T14 = VZMULJ(T12, T13); T1o = LD(&(x[WS(rs, 25)]), ms, &(x[WS(rs, 1)])); T1p = VZMULJ(T1n, T1o); T1l = LD(&(x[WS(rs, 9)]), ms, &(x[WS(rs, 1)])); T1m = VZMULJ(T5, T1l); { V T1d, T1f, T16, T19; T1d = LD(&(x[WS(rs, 29)]), ms, &(x[WS(rs, 1)])); T1e = VZMULJ(T1c, T1d); T1f = LD(&(x[WS(rs, 13)]), ms, &(x[WS(rs, 1)])); T1g = VZMULJ(TK, T1f); T1h = VSUB(T1e, T1g); T16 = LD(&(x[WS(rs, 5)]), ms, &(x[WS(rs, 1)])); T17 = VZMULJ(TS, T16); T19 = LD(&(x[WS(rs, 21)]), ms, &(x[WS(rs, 1)])); T1a = VZMULJ(T18, T19); T1b = VSUB(T17, T1a); } } { V T15, T1i, T31, T32; T15 = VSUB(T11, T14); T1i = VMUL(LDK(KP707106781), VADD(T1b, T1h)); T1j = VADD(T15, T1i); T2s = VSUB(T15, T1i); T31 = VADD(T17, T1a); T32 = VADD(T1e, T1g); T33 = VADD(T31, T32); T3D = VSUB(T32, T31); } { V T1k, T1q, T2Y, T2Z; T1k = VMUL(LDK(KP707106781), VSUB(T1h, T1b)); T1q = VSUB(T1m, T1p); T1r = VSUB(T1k, T1q); T2t = VADD(T1q, T1k); T2Y = VADD(T11, T14); T2Z = VADD(T1m, T1p); T30 = VADD(T2Y, T2Z); T3C = VSUB(T2Y, T2Z); } } { V T3r, T3v, T3u, T3w; { V T3p, T3q, T3s, T3t; T3p = VADD(T2T, T2W); T3q = VADD(T3j, T3g); T3r = VADD(T3p, T3q); T3v = VSUB(T3p, T3q); T3s = VADD(T30, T33); T3t = VADD(T37, T3a); T3u = VADD(T3s, T3t); T3w = VBYI(VSUB(T3t, T3s)); } ST(&(x[WS(rs, 16)]), VSUB(T3r, T3u), ms, &(x[0])); ST(&(x[WS(rs, 8)]), VADD(T3v, T3w), ms, &(x[0])); ST(&(x[0]), VADD(T3r, T3u), ms, &(x[0])); ST(&(x[WS(rs, 24)]), VSUB(T3v, T3w), ms, &(x[0])); } { V T2X, T3k, T3c, T3l, T34, T3b; T2X = VSUB(T2T, T2W); T3k = VSUB(T3g, T3j); T34 = VSUB(T30, T33); T3b = VSUB(T37, T3a); T3c = VMUL(LDK(KP707106781), VADD(T34, T3b)); T3l = VMUL(LDK(KP707106781), VSUB(T3b, T34)); { V T3d, T3m, T3n, T3o; T3d = VADD(T2X, T3c); T3m = VBYI(VADD(T3k, T3l)); ST(&(x[WS(rs, 28)]), VSUB(T3d, T3m), ms, &(x[0])); ST(&(x[WS(rs, 4)]), VADD(T3d, T3m), ms, &(x[0])); T3n = VSUB(T2X, T3c); T3o = VBYI(VSUB(T3l, T3k)); ST(&(x[WS(rs, 20)]), VSUB(T3n, T3o), ms, &(x[0])); ST(&(x[WS(rs, 12)]), VADD(T3n, T3o), ms, &(x[0])); } } { V T3B, T3W, T3M, T3U, T3I, T3T, T3P, T3X, T3A, T3L; T3A = VMUL(LDK(KP707106781), VADD(T3y, T3z)); T3B = VADD(T3x, T3A); T3W = VSUB(T3x, T3A); T3L = VMUL(LDK(KP707106781), VSUB(T3z, T3y)); T3M = VADD(T3K, T3L); T3U = VSUB(T3L, T3K); { V T3E, T3H, T3N, T3O; T3E = VFMA(LDK(KP923879532), T3C, VMUL(LDK(KP382683432), T3D)); T3H = VFNMS(LDK(KP382683432), T3G, VMUL(LDK(KP923879532), T3F)); T3I = VADD(T3E, T3H); T3T = VSUB(T3H, T3E); T3N = VFNMS(LDK(KP382683432), T3C, VMUL(LDK(KP923879532), T3D)); T3O = VFMA(LDK(KP382683432), T3F, VMUL(LDK(KP923879532), T3G)); T3P = VADD(T3N, T3O); T3X = VSUB(T3O, T3N); } { V T3J, T3Q, T3Z, T40; T3J = VADD(T3B, T3I); T3Q = VBYI(VADD(T3M, T3P)); ST(&(x[WS(rs, 30)]), VSUB(T3J, T3Q), ms, &(x[0])); ST(&(x[WS(rs, 2)]), VADD(T3J, T3Q), ms, &(x[0])); T3Z = VBYI(VADD(T3U, T3T)); T40 = VADD(T3W, T3X); ST(&(x[WS(rs, 6)]), VADD(T3Z, T40), ms, &(x[0])); ST(&(x[WS(rs, 26)]), VSUB(T40, T3Z), ms, &(x[0])); } { V T3R, T3S, T3V, T3Y; T3R = VSUB(T3B, T3I); T3S = VBYI(VSUB(T3P, T3M)); ST(&(x[WS(rs, 18)]), VSUB(T3R, T3S), ms, &(x[0])); ST(&(x[WS(rs, 14)]), VADD(T3R, T3S), ms, &(x[0])); T3V = VBYI(VSUB(T3T, T3U)); T3Y = VSUB(T3W, T3X); ST(&(x[WS(rs, 10)]), VADD(T3V, T3Y), ms, &(x[0])); ST(&(x[WS(rs, 22)]), VSUB(T3Y, T3V), ms, &(x[0])); } } { V TZ, T2k, T2d, T2l, T1X, T2h, T2a, T2i; { V Ts, TY, T2b, T2c; Ts = VADD(Tb, Tr); TY = VADD(TG, TX); TZ = VADD(Ts, TY); T2k = VSUB(Ts, TY); T2b = VFNMS(LDK(KP195090322), T1j, VMUL(LDK(KP980785280), T1r)); T2c = VFMA(LDK(KP195090322), T1N, VMUL(LDK(KP980785280), T1V)); T2d = VADD(T2b, T2c); T2l = VSUB(T2c, T2b); } { V T1s, T1W, T26, T29; T1s = VFMA(LDK(KP980785280), T1j, VMUL(LDK(KP195090322), T1r)); T1W = VFNMS(LDK(KP195090322), T1V, VMUL(LDK(KP980785280), T1N)); T1X = VADD(T1s, T1W); T2h = VSUB(T1W, T1s); T26 = VSUB(T1Z, T25); T29 = VSUB(T27, T28); T2a = VADD(T26, T29); T2i = VSUB(T29, T26); } { V T1Y, T2e, T2n, T2o; T1Y = VADD(TZ, T1X); T2e = VBYI(VADD(T2a, T2d)); ST(&(x[WS(rs, 31)]), VSUB(T1Y, T2e), ms, &(x[WS(rs, 1)])); ST(&(x[WS(rs, 1)]), VADD(T1Y, T2e), ms, &(x[WS(rs, 1)])); T2n = VBYI(VADD(T2i, T2h)); T2o = VADD(T2k, T2l); ST(&(x[WS(rs, 7)]), VADD(T2n, T2o), ms, &(x[WS(rs, 1)])); ST(&(x[WS(rs, 25)]), VSUB(T2o, T2n), ms, &(x[WS(rs, 1)])); } { V T2f, T2g, T2j, T2m; T2f = VSUB(TZ, T1X); T2g = VBYI(VSUB(T2d, T2a)); ST(&(x[WS(rs, 17)]), VSUB(T2f, T2g), ms, &(x[WS(rs, 1)])); ST(&(x[WS(rs, 15)]), VADD(T2f, T2g), ms, &(x[WS(rs, 1)])); T2j = VBYI(VSUB(T2h, T2i)); T2m = VSUB(T2k, T2l); ST(&(x[WS(rs, 9)]), VADD(T2j, T2m), ms, &(x[WS(rs, 1)])); ST(&(x[WS(rs, 23)]), VSUB(T2m, T2j), ms, &(x[WS(rs, 1)])); } } { V T2r, T2M, T2F, T2N, T2y, T2J, T2C, T2K; { V T2p, T2q, T2D, T2E; T2p = VSUB(Tb, Tr); T2q = VADD(T28, T27); T2r = VADD(T2p, T2q); T2M = VSUB(T2p, T2q); T2D = VFNMS(LDK(KP555570233), T2s, VMUL(LDK(KP831469612), T2t)); T2E = VFMA(LDK(KP555570233), T2v, VMUL(LDK(KP831469612), T2w)); T2F = VADD(T2D, T2E); T2N = VSUB(T2E, T2D); } { V T2u, T2x, T2A, T2B; T2u = VFMA(LDK(KP831469612), T2s, VMUL(LDK(KP555570233), T2t)); T2x = VFNMS(LDK(KP555570233), T2w, VMUL(LDK(KP831469612), T2v)); T2y = VADD(T2u, T2x); T2J = VSUB(T2x, T2u); T2A = VADD(T25, T1Z); T2B = VSUB(TX, TG); T2C = VADD(T2A, T2B); T2K = VSUB(T2B, T2A); } { V T2z, T2G, T2P, T2Q; T2z = VADD(T2r, T2y); T2G = VBYI(VADD(T2C, T2F)); ST(&(x[WS(rs, 29)]), VSUB(T2z, T2G), ms, &(x[WS(rs, 1)])); ST(&(x[WS(rs, 3)]), VADD(T2z, T2G), ms, &(x[WS(rs, 1)])); T2P = VBYI(VADD(T2K, T2J)); T2Q = VADD(T2M, T2N); ST(&(x[WS(rs, 5)]), VADD(T2P, T2Q), ms, &(x[WS(rs, 1)])); ST(&(x[WS(rs, 27)]), VSUB(T2Q, T2P), ms, &(x[WS(rs, 1)])); } { V T2H, T2I, T2L, T2O; T2H = VSUB(T2r, T2y); T2I = VBYI(VSUB(T2F, T2C)); ST(&(x[WS(rs, 19)]), VSUB(T2H, T2I), ms, &(x[WS(rs, 1)])); ST(&(x[WS(rs, 13)]), VADD(T2H, T2I), ms, &(x[WS(rs, 1)])); T2L = VBYI(VSUB(T2J, T2K)); T2O = VSUB(T2M, T2N); ST(&(x[WS(rs, 11)]), VADD(T2L, T2O), ms, &(x[WS(rs, 1)])); ST(&(x[WS(rs, 21)]), VSUB(T2O, T2L), ms, &(x[WS(rs, 1)])); } } } } } VLEAVE(); } static const tw_instr twinstr[] = { VTW(0, 1), VTW(0, 3), VTW(0, 9), VTW(0, 27), {TW_NEXT, VL, 0} }; static const ct_desc desc = { 32, XSIMD_STRING("t3fv_32"), twinstr, &GENUS, {228, 142, 16, 0}, 0, 0, 0 }; void XSIMD(codelet_t3fv_32) (planner *p) { X(kdft_dit_register) (p, t3fv_32, &desc); } #endif /* HAVE_FMA */ fftw-3.3.3/dft/simd/common/t3fv_5.c0000644000175000017500000001432512054410537013670 00000000000000/* * Copyright (c) 2003, 2007-11 Matteo Frigo * Copyright (c) 2003, 2007-11 Massachusetts Institute of Technology * * This program is free software; you can redistribute it and/or modify * it under the terms of the GNU General Public License as published by * the Free Software Foundation; either version 2 of the License, or * (at your option) any later version. * * This program is distributed in the hope that it will be useful, * but WITHOUT ANY WARRANTY; without even the implied warranty of * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the * GNU General Public License for more details. * * You should have received a copy of the GNU General Public License * along with this program; if not, write to the Free Software * Foundation, Inc., 51 Franklin Street, Fifth Floor, Boston, MA 02110-1301 USA * */ /* This file was automatically generated --- DO NOT EDIT */ /* Generated on Sun Nov 25 07:38:54 EST 2012 */ #include "codelet-dft.h" #ifdef HAVE_FMA /* Generated by: ../../../genfft/gen_twiddle_c.native -fma -reorder-insns -schedule-for-pipeline -simd -compact -variables 4 -pipeline-latency 8 -twiddle-log3 -precompute-twiddles -no-generate-bytw -n 5 -name t3fv_5 -include t3f.h */ /* * This function contains 22 FP additions, 23 FP multiplications, * (or, 13 additions, 14 multiplications, 9 fused multiply/add), * 30 stack variables, 4 constants, and 10 memory accesses */ #include "t3f.h" static void t3fv_5(R *ri, R *ii, const R *W, stride rs, INT mb, INT me, INT ms) { DVK(KP559016994, +0.559016994374947424102293417182819058860154590); DVK(KP250000000, +0.250000000000000000000000000000000000000000000); DVK(KP618033988, +0.618033988749894848204586834365638117720309180); DVK(KP951056516, +0.951056516295153572116439333379382143405698634); { INT m; R *x; x = ri; for (m = mb, W = W + (mb * ((TWVL / VL) * 4)); m < me; m = m + VL, x = x + (VL * ms), W = W + (TWVL * 4), MAKE_VOLATILE_STRIDE(5, rs)) { V T2, T5, T1, T3, Td, T7, Tb; T2 = LDW(&(W[0])); T5 = LDW(&(W[TWVL * 2])); T1 = LD(&(x[0]), ms, &(x[0])); T3 = LD(&(x[WS(rs, 1)]), ms, &(x[WS(rs, 1)])); Td = LD(&(x[WS(rs, 3)]), ms, &(x[WS(rs, 1)])); T7 = LD(&(x[WS(rs, 4)]), ms, &(x[0])); Tb = LD(&(x[WS(rs, 2)]), ms, &(x[0])); { V Ta, T6, T4, Te, Tc, T8; Ta = VZMULJ(T2, T5); T6 = VZMUL(T2, T5); T4 = VZMULJ(T2, T3); Te = VZMULJ(T5, Td); Tc = VZMULJ(Ta, Tb); T8 = VZMULJ(T6, T7); { V Tf, Tl, T9, Tk; Tf = VADD(Tc, Te); Tl = VSUB(Tc, Te); T9 = VADD(T4, T8); Tk = VSUB(T4, T8); { V Ti, Tg, To, Tm, Th, Tn, Tj; Ti = VSUB(T9, Tf); Tg = VADD(T9, Tf); To = VMUL(LDK(KP951056516), VFNMS(LDK(KP618033988), Tk, Tl)); Tm = VMUL(LDK(KP951056516), VFMA(LDK(KP618033988), Tl, Tk)); Th = VFNMS(LDK(KP250000000), Tg, T1); ST(&(x[0]), VADD(T1, Tg), ms, &(x[0])); Tn = VFNMS(LDK(KP559016994), Ti, Th); Tj = VFMA(LDK(KP559016994), Ti, Th); ST(&(x[WS(rs, 2)]), VFMAI(To, Tn), ms, &(x[0])); ST(&(x[WS(rs, 3)]), VFNMSI(To, Tn), ms, &(x[WS(rs, 1)])); ST(&(x[WS(rs, 4)]), VFMAI(Tm, Tj), ms, &(x[0])); ST(&(x[WS(rs, 1)]), VFNMSI(Tm, Tj), ms, &(x[WS(rs, 1)])); } } } } } VLEAVE(); } static const tw_instr twinstr[] = { VTW(0, 1), VTW(0, 3), {TW_NEXT, VL, 0} }; static const ct_desc desc = { 5, XSIMD_STRING("t3fv_5"), twinstr, &GENUS, {13, 14, 9, 0}, 0, 0, 0 }; void XSIMD(codelet_t3fv_5) (planner *p) { X(kdft_dit_register) (p, t3fv_5, &desc); } #else /* HAVE_FMA */ /* Generated by: ../../../genfft/gen_twiddle_c.native -simd -compact -variables 4 -pipeline-latency 8 -twiddle-log3 -precompute-twiddles -no-generate-bytw -n 5 -name t3fv_5 -include t3f.h */ /* * This function contains 22 FP additions, 18 FP multiplications, * (or, 19 additions, 15 multiplications, 3 fused multiply/add), * 24 stack variables, 4 constants, and 10 memory accesses */ #include "t3f.h" static void t3fv_5(R *ri, R *ii, const R *W, stride rs, INT mb, INT me, INT ms) { DVK(KP250000000, +0.250000000000000000000000000000000000000000000); DVK(KP559016994, +0.559016994374947424102293417182819058860154590); DVK(KP587785252, +0.587785252292473129168705954639072768597652438); DVK(KP951056516, +0.951056516295153572116439333379382143405698634); { INT m; R *x; x = ri; for (m = mb, W = W + (mb * ((TWVL / VL) * 4)); m < me; m = m + VL, x = x + (VL * ms), W = W + (TWVL * 4), MAKE_VOLATILE_STRIDE(5, rs)) { V T1, T4, T5, T9; T1 = LDW(&(W[0])); T4 = LDW(&(W[TWVL * 2])); T5 = VZMUL(T1, T4); T9 = VZMULJ(T1, T4); { V Tg, Tk, Tl, T8, Te, Th; Tg = LD(&(x[0]), ms, &(x[0])); { V T3, Td, T7, Tb; { V T2, Tc, T6, Ta; T2 = LD(&(x[WS(rs, 1)]), ms, &(x[WS(rs, 1)])); T3 = VZMULJ(T1, T2); Tc = LD(&(x[WS(rs, 3)]), ms, &(x[WS(rs, 1)])); Td = VZMULJ(T4, Tc); T6 = LD(&(x[WS(rs, 4)]), ms, &(x[0])); T7 = VZMULJ(T5, T6); Ta = LD(&(x[WS(rs, 2)]), ms, &(x[0])); Tb = VZMULJ(T9, Ta); } Tk = VSUB(T3, T7); Tl = VSUB(Tb, Td); T8 = VADD(T3, T7); Te = VADD(Tb, Td); Th = VADD(T8, Te); } ST(&(x[0]), VADD(Tg, Th), ms, &(x[0])); { V Tm, Tn, Tj, To, Tf, Ti; Tm = VBYI(VFMA(LDK(KP951056516), Tk, VMUL(LDK(KP587785252), Tl))); Tn = VBYI(VFNMS(LDK(KP587785252), Tk, VMUL(LDK(KP951056516), Tl))); Tf = VMUL(LDK(KP559016994), VSUB(T8, Te)); Ti = VFNMS(LDK(KP250000000), Th, Tg); Tj = VADD(Tf, Ti); To = VSUB(Ti, Tf); ST(&(x[WS(rs, 1)]), VSUB(Tj, Tm), ms, &(x[WS(rs, 1)])); ST(&(x[WS(rs, 3)]), VSUB(To, Tn), ms, &(x[WS(rs, 1)])); ST(&(x[WS(rs, 4)]), VADD(Tm, Tj), ms, &(x[0])); ST(&(x[WS(rs, 2)]), VADD(Tn, To), ms, &(x[0])); } } } } VLEAVE(); } static const tw_instr twinstr[] = { VTW(0, 1), VTW(0, 3), {TW_NEXT, VL, 0} }; static const ct_desc desc = { 5, XSIMD_STRING("t3fv_5"), twinstr, &GENUS, {19, 15, 3, 0}, 0, 0, 0 }; void XSIMD(codelet_t3fv_5) (planner *p) { X(kdft_dit_register) (p, t3fv_5, &desc); } #endif /* HAVE_FMA */ fftw-3.3.3/dft/simd/common/t3fv_10.c0000644000175000017500000002306512054410537013745 00000000000000/* * Copyright (c) 2003, 2007-11 Matteo Frigo * Copyright (c) 2003, 2007-11 Massachusetts Institute of Technology * * This program is free software; you can redistribute it and/or modify * it under the terms of the GNU General Public License as published by * the Free Software Foundation; either version 2 of the License, or * (at your option) any later version. * * This program is distributed in the hope that it will be useful, * but WITHOUT ANY WARRANTY; without even the implied warranty of * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the * GNU General Public License for more details. * * You should have received a copy of the GNU General Public License * along with this program; if not, write to the Free Software * Foundation, Inc., 51 Franklin Street, Fifth Floor, Boston, MA 02110-1301 USA * */ /* This file was automatically generated --- DO NOT EDIT */ /* Generated on Sun Nov 25 07:38:55 EST 2012 */ #include "codelet-dft.h" #ifdef HAVE_FMA /* Generated by: ../../../genfft/gen_twiddle_c.native -fma -reorder-insns -schedule-for-pipeline -simd -compact -variables 4 -pipeline-latency 8 -twiddle-log3 -precompute-twiddles -no-generate-bytw -n 10 -name t3fv_10 -include t3f.h */ /* * This function contains 57 FP additions, 52 FP multiplications, * (or, 39 additions, 34 multiplications, 18 fused multiply/add), * 57 stack variables, 4 constants, and 20 memory accesses */ #include "t3f.h" static void t3fv_10(R *ri, R *ii, const R *W, stride rs, INT mb, INT me, INT ms) { DVK(KP559016994, +0.559016994374947424102293417182819058860154590); DVK(KP250000000, +0.250000000000000000000000000000000000000000000); DVK(KP618033988, +0.618033988749894848204586834365638117720309180); DVK(KP951056516, +0.951056516295153572116439333379382143405698634); { INT m; R *x; x = ri; for (m = mb, W = W + (mb * ((TWVL / VL) * 6)); m < me; m = m + VL, x = x + (VL * ms), W = W + (TWVL * 6), MAKE_VOLATILE_STRIDE(10, rs)) { V T1, T7, Th, Tx, Tr, Td, Tp, T6, Tv, Tc, Te, Ti, Tl, T2, T3; V T5; T2 = LDW(&(W[0])); T3 = LDW(&(W[TWVL * 2])); T5 = LDW(&(W[TWVL * 4])); T1 = LD(&(x[0]), ms, &(x[0])); T7 = LD(&(x[WS(rs, 5)]), ms, &(x[WS(rs, 1)])); { V To, Tw, Tq, Tu, Ta, T4, Tt, Tk, Tb; To = LD(&(x[WS(rs, 4)]), ms, &(x[0])); Tw = LD(&(x[WS(rs, 1)]), ms, &(x[WS(rs, 1)])); Tq = LD(&(x[WS(rs, 9)]), ms, &(x[WS(rs, 1)])); Tu = LD(&(x[WS(rs, 6)]), ms, &(x[0])); Ta = VZMULJ(T2, T3); T4 = VZMUL(T2, T3); Th = VZMULJ(T2, T5); Tt = VZMULJ(T3, T5); Tb = LD(&(x[WS(rs, 2)]), ms, &(x[0])); Tx = VZMULJ(T2, Tw); Tr = VZMULJ(T5, Tq); Tk = LD(&(x[WS(rs, 3)]), ms, &(x[WS(rs, 1)])); Td = VZMULJ(Ta, T5); Tp = VZMULJ(T4, To); T6 = VZMULJ(T4, T5); Tv = VZMULJ(Tt, Tu); Tc = VZMULJ(Ta, Tb); Te = LD(&(x[WS(rs, 7)]), ms, &(x[WS(rs, 1)])); Ti = LD(&(x[WS(rs, 8)]), ms, &(x[0])); Tl = VZMULJ(T3, Tk); } { V TN, Ts, T8, Ty, TO, Tf, Tj; TN = VADD(Tp, Tr); Ts = VSUB(Tp, Tr); T8 = VZMULJ(T6, T7); Ty = VSUB(Tv, Tx); TO = VADD(Tv, Tx); Tf = VZMULJ(Td, Te); Tj = VZMULJ(Th, Ti); { V T9, TJ, TP, TU, Tz, TF, Tg, TK, Tm, TL; T9 = VSUB(T1, T8); TJ = VADD(T1, T8); TP = VADD(TN, TO); TU = VSUB(TN, TO); Tz = VADD(Ts, Ty); TF = VSUB(Ts, Ty); Tg = VSUB(Tc, Tf); TK = VADD(Tc, Tf); Tm = VSUB(Tj, Tl); TL = VADD(Tj, Tl); { V TM, TV, Tn, TE; TM = VADD(TK, TL); TV = VSUB(TK, TL); Tn = VADD(Tg, Tm); TE = VSUB(Tg, Tm); { V TW, TY, TS, TQ, TG, TI, TC, TA, TR, TB; TW = VMUL(LDK(KP951056516), VFNMS(LDK(KP618033988), TV, TU)); TY = VMUL(LDK(KP951056516), VFMA(LDK(KP618033988), TU, TV)); TS = VSUB(TM, TP); TQ = VADD(TM, TP); TG = VMUL(LDK(KP951056516), VFMA(LDK(KP618033988), TF, TE)); TI = VMUL(LDK(KP951056516), VFNMS(LDK(KP618033988), TE, TF)); TC = VSUB(Tn, Tz); TA = VADD(Tn, Tz); ST(&(x[0]), VADD(TJ, TQ), ms, &(x[0])); TR = VFNMS(LDK(KP250000000), TQ, TJ); ST(&(x[WS(rs, 5)]), VADD(T9, TA), ms, &(x[WS(rs, 1)])); TB = VFNMS(LDK(KP250000000), TA, T9); { V TX, TT, TH, TD; TX = VFMA(LDK(KP559016994), TS, TR); TT = VFNMS(LDK(KP559016994), TS, TR); TH = VFNMS(LDK(KP559016994), TC, TB); TD = VFMA(LDK(KP559016994), TC, TB); ST(&(x[WS(rs, 8)]), VFNMSI(TW, TT), ms, &(x[0])); ST(&(x[WS(rs, 2)]), VFMAI(TW, TT), ms, &(x[0])); ST(&(x[WS(rs, 6)]), VFNMSI(TY, TX), ms, &(x[0])); ST(&(x[WS(rs, 4)]), VFMAI(TY, TX), ms, &(x[0])); ST(&(x[WS(rs, 9)]), VFMAI(TG, TD), ms, &(x[WS(rs, 1)])); ST(&(x[WS(rs, 1)]), VFNMSI(TG, TD), ms, &(x[WS(rs, 1)])); ST(&(x[WS(rs, 7)]), VFMAI(TI, TH), ms, &(x[WS(rs, 1)])); ST(&(x[WS(rs, 3)]), VFNMSI(TI, TH), ms, &(x[WS(rs, 1)])); } } } } } } } VLEAVE(); } static const tw_instr twinstr[] = { VTW(0, 1), VTW(0, 3), VTW(0, 9), {TW_NEXT, VL, 0} }; static const ct_desc desc = { 10, XSIMD_STRING("t3fv_10"), twinstr, &GENUS, {39, 34, 18, 0}, 0, 0, 0 }; void XSIMD(codelet_t3fv_10) (planner *p) { X(kdft_dit_register) (p, t3fv_10, &desc); } #else /* HAVE_FMA */ /* Generated by: ../../../genfft/gen_twiddle_c.native -simd -compact -variables 4 -pipeline-latency 8 -twiddle-log3 -precompute-twiddles -no-generate-bytw -n 10 -name t3fv_10 -include t3f.h */ /* * This function contains 57 FP additions, 42 FP multiplications, * (or, 51 additions, 36 multiplications, 6 fused multiply/add), * 41 stack variables, 4 constants, and 20 memory accesses */ #include "t3f.h" static void t3fv_10(R *ri, R *ii, const R *W, stride rs, INT mb, INT me, INT ms) { DVK(KP587785252, +0.587785252292473129168705954639072768597652438); DVK(KP951056516, +0.951056516295153572116439333379382143405698634); DVK(KP250000000, +0.250000000000000000000000000000000000000000000); DVK(KP559016994, +0.559016994374947424102293417182819058860154590); { INT m; R *x; x = ri; for (m = mb, W = W + (mb * ((TWVL / VL) * 6)); m < me; m = m + VL, x = x + (VL * ms), W = W + (TWVL * 6), MAKE_VOLATILE_STRIDE(10, rs)) { V T1, T2, T3, Ti, T6, T7, Tx, Tb, To; T1 = LDW(&(W[0])); T2 = LDW(&(W[TWVL * 2])); T3 = VZMULJ(T1, T2); Ti = VZMUL(T1, T2); T6 = LDW(&(W[TWVL * 4])); T7 = VZMULJ(T3, T6); Tx = VZMULJ(Ti, T6); Tb = VZMULJ(T1, T6); To = VZMULJ(T2, T6); { V TA, TQ, Tn, Tt, Tu, TJ, TK, TS, Ta, Tg, Th, TM, TN, TR, Tw; V Tz, Ty; Tw = LD(&(x[0]), ms, &(x[0])); Ty = LD(&(x[WS(rs, 5)]), ms, &(x[WS(rs, 1)])); Tz = VZMULJ(Tx, Ty); TA = VSUB(Tw, Tz); TQ = VADD(Tw, Tz); { V Tk, Ts, Tm, Tq; { V Tj, Tr, Tl, Tp; Tj = LD(&(x[WS(rs, 4)]), ms, &(x[0])); Tk = VZMULJ(Ti, Tj); Tr = LD(&(x[WS(rs, 1)]), ms, &(x[WS(rs, 1)])); Ts = VZMULJ(T1, Tr); Tl = LD(&(x[WS(rs, 9)]), ms, &(x[WS(rs, 1)])); Tm = VZMULJ(T6, Tl); Tp = LD(&(x[WS(rs, 6)]), ms, &(x[0])); Tq = VZMULJ(To, Tp); } Tn = VSUB(Tk, Tm); Tt = VSUB(Tq, Ts); Tu = VADD(Tn, Tt); TJ = VADD(Tk, Tm); TK = VADD(Tq, Ts); TS = VADD(TJ, TK); } { V T5, Tf, T9, Td; { V T4, Te, T8, Tc; T4 = LD(&(x[WS(rs, 2)]), ms, &(x[0])); T5 = VZMULJ(T3, T4); Te = LD(&(x[WS(rs, 3)]), ms, &(x[WS(rs, 1)])); Tf = VZMULJ(T2, Te); T8 = LD(&(x[WS(rs, 7)]), ms, &(x[WS(rs, 1)])); T9 = VZMULJ(T7, T8); Tc = LD(&(x[WS(rs, 8)]), ms, &(x[0])); Td = VZMULJ(Tb, Tc); } Ta = VSUB(T5, T9); Tg = VSUB(Td, Tf); Th = VADD(Ta, Tg); TM = VADD(T5, T9); TN = VADD(Td, Tf); TR = VADD(TM, TN); } { V Tv, TB, TC, TG, TI, TE, TF, TH, TD; Tv = VMUL(LDK(KP559016994), VSUB(Th, Tu)); TB = VADD(Th, Tu); TC = VFNMS(LDK(KP250000000), TB, TA); TE = VSUB(Ta, Tg); TF = VSUB(Tn, Tt); TG = VBYI(VFMA(LDK(KP951056516), TE, VMUL(LDK(KP587785252), TF))); TI = VBYI(VFNMS(LDK(KP587785252), TE, VMUL(LDK(KP951056516), TF))); ST(&(x[WS(rs, 5)]), VADD(TA, TB), ms, &(x[WS(rs, 1)])); TH = VSUB(TC, Tv); ST(&(x[WS(rs, 3)]), VSUB(TH, TI), ms, &(x[WS(rs, 1)])); ST(&(x[WS(rs, 7)]), VADD(TI, TH), ms, &(x[WS(rs, 1)])); TD = VADD(Tv, TC); ST(&(x[WS(rs, 1)]), VSUB(TD, TG), ms, &(x[WS(rs, 1)])); ST(&(x[WS(rs, 9)]), VADD(TG, TD), ms, &(x[WS(rs, 1)])); } { V TV, TT, TU, TP, TX, TL, TO, TY, TW; TV = VMUL(LDK(KP559016994), VSUB(TR, TS)); TT = VADD(TR, TS); TU = VFNMS(LDK(KP250000000), TT, TQ); TL = VSUB(TJ, TK); TO = VSUB(TM, TN); TP = VBYI(VFNMS(LDK(KP587785252), TO, VMUL(LDK(KP951056516), TL))); TX = VBYI(VFMA(LDK(KP951056516), TO, VMUL(LDK(KP587785252), TL))); ST(&(x[0]), VADD(TQ, TT), ms, &(x[0])); TY = VADD(TV, TU); ST(&(x[WS(rs, 4)]), VADD(TX, TY), ms, &(x[0])); ST(&(x[WS(rs, 6)]), VSUB(TY, TX), ms, &(x[0])); TW = VSUB(TU, TV); ST(&(x[WS(rs, 2)]), VADD(TP, TW), ms, &(x[0])); ST(&(x[WS(rs, 8)]), VSUB(TW, TP), ms, &(x[0])); } } } } VLEAVE(); } static const tw_instr twinstr[] = { VTW(0, 1), VTW(0, 3), VTW(0, 9), {TW_NEXT, VL, 0} }; static const ct_desc desc = { 10, XSIMD_STRING("t3fv_10"), twinstr, &GENUS, {51, 36, 6, 0}, 0, 0, 0 }; void XSIMD(codelet_t3fv_10) (planner *p) { X(kdft_dit_register) (p, t3fv_10, &desc); } #endif /* HAVE_FMA */ fftw-3.3.3/dft/simd/common/t3fv_20.c0000644000175000017500000004407512054410540013744 00000000000000/* * Copyright (c) 2003, 2007-11 Matteo Frigo * Copyright (c) 2003, 2007-11 Massachusetts Institute of Technology * * This program is free software; you can redistribute it and/or modify * it under the terms of the GNU General Public License as published by * the Free Software Foundation; either version 2 of the License, or * (at your option) any later version. * * This program is distributed in the hope that it will be useful, * but WITHOUT ANY WARRANTY; without even the implied warranty of * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the * GNU General Public License for more details. * * You should have received a copy of the GNU General Public License * along with this program; if not, write to the Free Software * Foundation, Inc., 51 Franklin Street, Fifth Floor, Boston, MA 02110-1301 USA * */ /* This file was automatically generated --- DO NOT EDIT */ /* Generated on Sun Nov 25 07:38:55 EST 2012 */ #include "codelet-dft.h" #ifdef HAVE_FMA /* Generated by: ../../../genfft/gen_twiddle_c.native -fma -reorder-insns -schedule-for-pipeline -simd -compact -variables 4 -pipeline-latency 8 -twiddle-log3 -precompute-twiddles -no-generate-bytw -n 20 -name t3fv_20 -include t3f.h */ /* * This function contains 138 FP additions, 118 FP multiplications, * (or, 92 additions, 72 multiplications, 46 fused multiply/add), * 90 stack variables, 4 constants, and 40 memory accesses */ #include "t3f.h" static void t3fv_20(R *ri, R *ii, const R *W, stride rs, INT mb, INT me, INT ms) { DVK(KP559016994, +0.559016994374947424102293417182819058860154590); DVK(KP951056516, +0.951056516295153572116439333379382143405698634); DVK(KP618033988, +0.618033988749894848204586834365638117720309180); DVK(KP250000000, +0.250000000000000000000000000000000000000000000); { INT m; R *x; x = ri; for (m = mb, W = W + (mb * ((TWVL / VL) * 8)); m < me; m = m + VL, x = x + (VL * ms), W = W + (TWVL * 8), MAKE_VOLATILE_STRIDE(20, rs)) { V T1k, T1w, T1r, T1z, T1o, T1y, T1v, T1h; { V T2, T8, T3, Td; T2 = LDW(&(W[0])); T8 = LDW(&(W[TWVL * 2])); T3 = LDW(&(W[TWVL * 4])); Td = LDW(&(W[TWVL * 6])); { V T7, TM, T1F, T23, T1p, Tp, T1j, T27, T1P, T1I, T1i, T1L, T28, T1S, T1q; V TE, T1n, T1d, T26, T2e; { V T1, TK, T5, TH; T1 = LD(&(x[0]), ms, &(x[0])); TK = LD(&(x[WS(rs, 15)]), ms, &(x[WS(rs, 1)])); T5 = LD(&(x[WS(rs, 10)]), ms, &(x[0])); TH = LD(&(x[WS(rs, 5)]), ms, &(x[WS(rs, 1)])); { V TA, Tx, TU, T1O, T14, Th, T1G, T1R, T1b, T1J, To, Ts, TV, Tv, TO; V TQ, TT, Ty, TB; { V Tq, Tt, T17, T1a, Tk, Tn; { V Tl, Ti, T15, T18, TZ, Tc, T6, Tb, Tf, T10, T12, TL; { V TJ, Ta, T9, T4; Ta = LD(&(x[WS(rs, 4)]), ms, &(x[0])); TA = VZMULJ(T2, T8); T9 = VZMUL(T2, T8); Tx = VZMUL(T8, T3); Tl = VZMULJ(T8, T3); T4 = VZMUL(T2, T3); Tq = VZMULJ(T2, T3); Tt = VZMULJ(T2, Td); Ti = VZMULJ(T8, Td); T15 = VZMULJ(TA, Td); T18 = VZMULJ(TA, T3); TU = VZMUL(TA, T3); TJ = VZMULJ(T9, Td); TZ = VZMUL(T9, T3); Tc = VZMULJ(T9, T3); T6 = VZMULJ(T4, T5); Tb = VZMULJ(T9, Ta); Tf = LD(&(x[WS(rs, 14)]), ms, &(x[0])); T10 = LD(&(x[WS(rs, 13)]), ms, &(x[WS(rs, 1)])); T12 = LD(&(x[WS(rs, 3)]), ms, &(x[WS(rs, 1)])); TL = VZMULJ(TJ, TK); } { V T1D, T11, T13, T19, T1E, Tg, T16, TI, Te, Tj, Tm; T16 = LD(&(x[WS(rs, 17)]), ms, &(x[WS(rs, 1)])); TI = VZMULJ(Tc, TH); Te = VZMULJ(Tc, Td); T7 = VSUB(T1, T6); T1D = VADD(T1, T6); T11 = VZMULJ(TZ, T10); T13 = VZMULJ(T8, T12); T19 = LD(&(x[WS(rs, 7)]), ms, &(x[WS(rs, 1)])); T17 = VZMULJ(T15, T16); TM = VSUB(TI, TL); T1E = VADD(TI, TL); Tg = VZMULJ(Te, Tf); Tj = LD(&(x[WS(rs, 16)]), ms, &(x[0])); Tm = LD(&(x[WS(rs, 6)]), ms, &(x[0])); T1O = VADD(T11, T13); T14 = VSUB(T11, T13); T1a = VZMULJ(T18, T19); T1F = VSUB(T1D, T1E); T23 = VADD(T1D, T1E); Th = VSUB(Tb, Tg); T1G = VADD(Tb, Tg); Tk = VZMULJ(Ti, Tj); Tn = VZMULJ(Tl, Tm); } } { V Tr, Tu, TN, TP, TS; Tr = LD(&(x[WS(rs, 8)]), ms, &(x[0])); T1R = VADD(T17, T1a); T1b = VSUB(T17, T1a); Tu = LD(&(x[WS(rs, 18)]), ms, &(x[0])); TN = LD(&(x[WS(rs, 9)]), ms, &(x[WS(rs, 1)])); TP = LD(&(x[WS(rs, 19)]), ms, &(x[WS(rs, 1)])); TS = LD(&(x[WS(rs, 1)]), ms, &(x[WS(rs, 1)])); T1J = VADD(Tk, Tn); To = VSUB(Tk, Tn); Ts = VZMULJ(Tq, Tr); TV = LD(&(x[WS(rs, 11)]), ms, &(x[WS(rs, 1)])); Tv = VZMULJ(Tt, Tu); TO = VZMULJ(T3, TN); TQ = VZMULJ(Td, TP); TT = VZMULJ(T2, TS); Ty = LD(&(x[WS(rs, 12)]), ms, &(x[0])); TB = LD(&(x[WS(rs, 2)]), ms, &(x[0])); } } { V T1N, Tw, T1H, TR, Tz, TC, T1c, TX, T1K, TW; T1p = VSUB(Th, To); Tp = VADD(Th, To); TW = VZMULJ(TU, TV); T1N = VADD(Ts, Tv); Tw = VSUB(Ts, Tv); T1H = VADD(TO, TQ); TR = VSUB(TO, TQ); Tz = VZMULJ(Tx, Ty); TC = VZMULJ(TA, TB); T1j = VSUB(T1b, T14); T1c = VADD(T14, T1b); TX = VSUB(TT, TW); T1K = VADD(TT, TW); T27 = VADD(T1N, T1O); T1P = VSUB(T1N, T1O); { V TD, T1Q, T24, TY, T25; TD = VSUB(Tz, TC); T1Q = VADD(Tz, TC); T1I = VSUB(T1G, T1H); T24 = VADD(T1G, T1H); TY = VADD(TR, TX); T1i = VSUB(TX, TR); T25 = VADD(T1J, T1K); T1L = VSUB(T1J, T1K); T28 = VADD(T1Q, T1R); T1S = VSUB(T1Q, T1R); T1q = VSUB(Tw, TD); TE = VADD(Tw, TD); T1n = VSUB(T1c, TY); T1d = VADD(TY, T1c); T26 = VADD(T24, T25); T2e = VSUB(T24, T25); } } } } { V T1M, T1Z, T1Y, T1T, T29, T2f, T1g, TF, T1m, T1e; T1M = VADD(T1I, T1L); T1Z = VSUB(T1I, T1L); T1Y = VSUB(T1P, T1S); T1T = VADD(T1P, T1S); T29 = VADD(T27, T28); T2f = VSUB(T27, T28); T1g = VSUB(Tp, TE); TF = VADD(Tp, TE); T1m = VFNMS(LDK(KP250000000), T1d, TM); T1e = VADD(TM, T1d); { V T1W, T2c, T1f, T2i, T2g, T22, T20, T1V, T2b, T1U, T2a, TG; T1k = VFMA(LDK(KP618033988), T1j, T1i); T1w = VFNMS(LDK(KP618033988), T1i, T1j); T1W = VSUB(T1M, T1T); T1U = VADD(T1M, T1T); T2c = VSUB(T26, T29); T2a = VADD(T26, T29); T1f = VFNMS(LDK(KP250000000), TF, T7); TG = VADD(T7, TF); T2i = VMUL(LDK(KP951056516), VFNMS(LDK(KP618033988), T2e, T2f)); T2g = VMUL(LDK(KP951056516), VFMA(LDK(KP618033988), T2f, T2e)); T22 = VMUL(LDK(KP951056516), VFMA(LDK(KP618033988), T1Y, T1Z)); T20 = VMUL(LDK(KP951056516), VFNMS(LDK(KP618033988), T1Z, T1Y)); ST(&(x[WS(rs, 10)]), VADD(T1F, T1U), ms, &(x[0])); T1V = VFNMS(LDK(KP250000000), T1U, T1F); ST(&(x[0]), VADD(T23, T2a), ms, &(x[0])); T2b = VFNMS(LDK(KP250000000), T2a, T23); ST(&(x[WS(rs, 15)]), VFMAI(T1e, TG), ms, &(x[WS(rs, 1)])); ST(&(x[WS(rs, 5)]), VFNMSI(T1e, TG), ms, &(x[WS(rs, 1)])); T1r = VFMA(LDK(KP618033988), T1q, T1p); T1z = VFNMS(LDK(KP618033988), T1p, T1q); { V T21, T1X, T2h, T2d; T21 = VFMA(LDK(KP559016994), T1W, T1V); T1X = VFNMS(LDK(KP559016994), T1W, T1V); T2h = VFNMS(LDK(KP559016994), T2c, T2b); T2d = VFMA(LDK(KP559016994), T2c, T2b); ST(&(x[WS(rs, 18)]), VFNMSI(T20, T1X), ms, &(x[0])); ST(&(x[WS(rs, 2)]), VFMAI(T20, T1X), ms, &(x[0])); ST(&(x[WS(rs, 14)]), VFMAI(T22, T21), ms, &(x[0])); ST(&(x[WS(rs, 6)]), VFNMSI(T22, T21), ms, &(x[0])); ST(&(x[WS(rs, 16)]), VFNMSI(T2g, T2d), ms, &(x[0])); ST(&(x[WS(rs, 4)]), VFMAI(T2g, T2d), ms, &(x[0])); ST(&(x[WS(rs, 12)]), VFMAI(T2i, T2h), ms, &(x[0])); ST(&(x[WS(rs, 8)]), VFNMSI(T2i, T2h), ms, &(x[0])); T1o = VFNMS(LDK(KP559016994), T1n, T1m); T1y = VFMA(LDK(KP559016994), T1n, T1m); T1v = VFNMS(LDK(KP559016994), T1g, T1f); T1h = VFMA(LDK(KP559016994), T1g, T1f); } } } } } { V T1C, T1A, T1s, T1u, T1l, T1t, T1B, T1x; T1C = VFMA(LDK(KP951056516), T1z, T1y); T1A = VFNMS(LDK(KP951056516), T1z, T1y); T1s = VFMA(LDK(KP951056516), T1r, T1o); T1u = VFNMS(LDK(KP951056516), T1r, T1o); T1l = VFMA(LDK(KP951056516), T1k, T1h); T1t = VFNMS(LDK(KP951056516), T1k, T1h); T1B = VFMA(LDK(KP951056516), T1w, T1v); T1x = VFNMS(LDK(KP951056516), T1w, T1v); ST(&(x[WS(rs, 11)]), VFMAI(T1u, T1t), ms, &(x[WS(rs, 1)])); ST(&(x[WS(rs, 9)]), VFNMSI(T1u, T1t), ms, &(x[WS(rs, 1)])); ST(&(x[WS(rs, 19)]), VFMAI(T1s, T1l), ms, &(x[WS(rs, 1)])); ST(&(x[WS(rs, 1)]), VFNMSI(T1s, T1l), ms, &(x[WS(rs, 1)])); ST(&(x[WS(rs, 3)]), VFMAI(T1A, T1x), ms, &(x[WS(rs, 1)])); ST(&(x[WS(rs, 17)]), VFNMSI(T1A, T1x), ms, &(x[WS(rs, 1)])); ST(&(x[WS(rs, 7)]), VFMAI(T1C, T1B), ms, &(x[WS(rs, 1)])); ST(&(x[WS(rs, 13)]), VFNMSI(T1C, T1B), ms, &(x[WS(rs, 1)])); } } } VLEAVE(); } static const tw_instr twinstr[] = { VTW(0, 1), VTW(0, 3), VTW(0, 9), VTW(0, 19), {TW_NEXT, VL, 0} }; static const ct_desc desc = { 20, XSIMD_STRING("t3fv_20"), twinstr, &GENUS, {92, 72, 46, 0}, 0, 0, 0 }; void XSIMD(codelet_t3fv_20) (planner *p) { X(kdft_dit_register) (p, t3fv_20, &desc); } #else /* HAVE_FMA */ /* Generated by: ../../../genfft/gen_twiddle_c.native -simd -compact -variables 4 -pipeline-latency 8 -twiddle-log3 -precompute-twiddles -no-generate-bytw -n 20 -name t3fv_20 -include t3f.h */ /* * This function contains 138 FP additions, 92 FP multiplications, * (or, 126 additions, 80 multiplications, 12 fused multiply/add), * 73 stack variables, 4 constants, and 40 memory accesses */ #include "t3f.h" static void t3fv_20(R *ri, R *ii, const R *W, stride rs, INT mb, INT me, INT ms) { DVK(KP587785252, +0.587785252292473129168705954639072768597652438); DVK(KP951056516, +0.951056516295153572116439333379382143405698634); DVK(KP250000000, +0.250000000000000000000000000000000000000000000); DVK(KP559016994, +0.559016994374947424102293417182819058860154590); { INT m; R *x; x = ri; for (m = mb, W = W + (mb * ((TWVL / VL) * 8)); m < me; m = m + VL, x = x + (VL * ms), W = W + (TWVL * 8), MAKE_VOLATILE_STRIDE(20, rs)) { V T2, T8, T9, TA, T3, Tc, T4, TZ, T18, Tl, Tq, Tx, TU, Td, Te; V T15, Ti, Tt, TJ; T2 = LDW(&(W[0])); T8 = LDW(&(W[TWVL * 2])); T9 = VZMUL(T2, T8); TA = VZMULJ(T2, T8); T3 = LDW(&(W[TWVL * 4])); Tc = VZMULJ(T9, T3); T4 = VZMUL(T2, T3); TZ = VZMUL(T9, T3); T18 = VZMULJ(TA, T3); Tl = VZMULJ(T8, T3); Tq = VZMULJ(T2, T3); Tx = VZMUL(T8, T3); TU = VZMUL(TA, T3); Td = LDW(&(W[TWVL * 6])); Te = VZMULJ(Tc, Td); T15 = VZMULJ(TA, Td); Ti = VZMULJ(T8, Td); Tt = VZMULJ(T2, Td); TJ = VZMULJ(T9, Td); { V T7, TM, T1U, T2d, T1i, T1p, T1q, T1j, Tp, TE, TF, T26, T27, T2b, T1M; V T1P, T1V, TY, T1c, T1d, T23, T24, T2a, T1F, T1I, T1W, TG, T1e; { V T1, TL, T6, TI, TK, T5, TH, T1S, T1T; T1 = LD(&(x[0]), ms, &(x[0])); TK = LD(&(x[WS(rs, 15)]), ms, &(x[WS(rs, 1)])); TL = VZMULJ(TJ, TK); T5 = LD(&(x[WS(rs, 10)]), ms, &(x[0])); T6 = VZMULJ(T4, T5); TH = LD(&(x[WS(rs, 5)]), ms, &(x[WS(rs, 1)])); TI = VZMULJ(Tc, TH); T7 = VSUB(T1, T6); TM = VSUB(TI, TL); T1S = VADD(T1, T6); T1T = VADD(TI, TL); T1U = VSUB(T1S, T1T); T2d = VADD(T1S, T1T); } { V Th, T1K, T14, T1E, T1b, T1H, To, T1N, Tw, T1D, TR, T1L, TX, T1O, TD; V T1G; { V Tb, Tg, Ta, Tf; Ta = LD(&(x[WS(rs, 4)]), ms, &(x[0])); Tb = VZMULJ(T9, Ta); Tf = LD(&(x[WS(rs, 14)]), ms, &(x[0])); Tg = VZMULJ(Te, Tf); Th = VSUB(Tb, Tg); T1K = VADD(Tb, Tg); } { V T11, T13, T10, T12; T10 = LD(&(x[WS(rs, 13)]), ms, &(x[WS(rs, 1)])); T11 = VZMULJ(TZ, T10); T12 = LD(&(x[WS(rs, 3)]), ms, &(x[WS(rs, 1)])); T13 = VZMULJ(T8, T12); T14 = VSUB(T11, T13); T1E = VADD(T11, T13); } { V T17, T1a, T16, T19; T16 = LD(&(x[WS(rs, 17)]), ms, &(x[WS(rs, 1)])); T17 = VZMULJ(T15, T16); T19 = LD(&(x[WS(rs, 7)]), ms, &(x[WS(rs, 1)])); T1a = VZMULJ(T18, T19); T1b = VSUB(T17, T1a); T1H = VADD(T17, T1a); } { V Tk, Tn, Tj, Tm; Tj = LD(&(x[WS(rs, 16)]), ms, &(x[0])); Tk = VZMULJ(Ti, Tj); Tm = LD(&(x[WS(rs, 6)]), ms, &(x[0])); Tn = VZMULJ(Tl, Tm); To = VSUB(Tk, Tn); T1N = VADD(Tk, Tn); } { V Ts, Tv, Tr, Tu; Tr = LD(&(x[WS(rs, 8)]), ms, &(x[0])); Ts = VZMULJ(Tq, Tr); Tu = LD(&(x[WS(rs, 18)]), ms, &(x[0])); Tv = VZMULJ(Tt, Tu); Tw = VSUB(Ts, Tv); T1D = VADD(Ts, Tv); } { V TO, TQ, TN, TP; TN = LD(&(x[WS(rs, 9)]), ms, &(x[WS(rs, 1)])); TO = VZMULJ(T3, TN); TP = LD(&(x[WS(rs, 19)]), ms, &(x[WS(rs, 1)])); TQ = VZMULJ(Td, TP); TR = VSUB(TO, TQ); T1L = VADD(TO, TQ); } { V TT, TW, TS, TV; TS = LD(&(x[WS(rs, 1)]), ms, &(x[WS(rs, 1)])); TT = VZMULJ(T2, TS); TV = LD(&(x[WS(rs, 11)]), ms, &(x[WS(rs, 1)])); TW = VZMULJ(TU, TV); TX = VSUB(TT, TW); T1O = VADD(TT, TW); } { V Tz, TC, Ty, TB; Ty = LD(&(x[WS(rs, 12)]), ms, &(x[0])); Tz = VZMULJ(Tx, Ty); TB = LD(&(x[WS(rs, 2)]), ms, &(x[0])); TC = VZMULJ(TA, TB); TD = VSUB(Tz, TC); T1G = VADD(Tz, TC); } T1i = VSUB(TX, TR); T1p = VSUB(Th, To); T1q = VSUB(Tw, TD); T1j = VSUB(T1b, T14); Tp = VADD(Th, To); TE = VADD(Tw, TD); TF = VADD(Tp, TE); T26 = VADD(T1D, T1E); T27 = VADD(T1G, T1H); T2b = VADD(T26, T27); T1M = VSUB(T1K, T1L); T1P = VSUB(T1N, T1O); T1V = VADD(T1M, T1P); TY = VADD(TR, TX); T1c = VADD(T14, T1b); T1d = VADD(TY, T1c); T23 = VADD(T1K, T1L); T24 = VADD(T1N, T1O); T2a = VADD(T23, T24); T1F = VSUB(T1D, T1E); T1I = VSUB(T1G, T1H); T1W = VADD(T1F, T1I); } TG = VADD(T7, TF); T1e = VBYI(VADD(TM, T1d)); ST(&(x[WS(rs, 5)]), VSUB(TG, T1e), ms, &(x[WS(rs, 1)])); ST(&(x[WS(rs, 15)]), VADD(TG, T1e), ms, &(x[WS(rs, 1)])); { V T2c, T2e, T2f, T29, T2i, T25, T28, T2h, T2g; T2c = VMUL(LDK(KP559016994), VSUB(T2a, T2b)); T2e = VADD(T2a, T2b); T2f = VFNMS(LDK(KP250000000), T2e, T2d); T25 = VSUB(T23, T24); T28 = VSUB(T26, T27); T29 = VBYI(VFMA(LDK(KP951056516), T25, VMUL(LDK(KP587785252), T28))); T2i = VBYI(VFNMS(LDK(KP587785252), T25, VMUL(LDK(KP951056516), T28))); ST(&(x[0]), VADD(T2d, T2e), ms, &(x[0])); T2h = VSUB(T2f, T2c); ST(&(x[WS(rs, 8)]), VSUB(T2h, T2i), ms, &(x[0])); ST(&(x[WS(rs, 12)]), VADD(T2i, T2h), ms, &(x[0])); T2g = VADD(T2c, T2f); ST(&(x[WS(rs, 4)]), VADD(T29, T2g), ms, &(x[0])); ST(&(x[WS(rs, 16)]), VSUB(T2g, T29), ms, &(x[0])); } { V T1Z, T1X, T1Y, T1R, T22, T1J, T1Q, T21, T20; T1Z = VMUL(LDK(KP559016994), VSUB(T1V, T1W)); T1X = VADD(T1V, T1W); T1Y = VFNMS(LDK(KP250000000), T1X, T1U); T1J = VSUB(T1F, T1I); T1Q = VSUB(T1M, T1P); T1R = VBYI(VFNMS(LDK(KP587785252), T1Q, VMUL(LDK(KP951056516), T1J))); T22 = VBYI(VFMA(LDK(KP951056516), T1Q, VMUL(LDK(KP587785252), T1J))); ST(&(x[WS(rs, 10)]), VADD(T1U, T1X), ms, &(x[0])); T21 = VADD(T1Z, T1Y); ST(&(x[WS(rs, 6)]), VSUB(T21, T22), ms, &(x[0])); ST(&(x[WS(rs, 14)]), VADD(T22, T21), ms, &(x[0])); T20 = VSUB(T1Y, T1Z); ST(&(x[WS(rs, 2)]), VADD(T1R, T20), ms, &(x[0])); ST(&(x[WS(rs, 18)]), VSUB(T20, T1R), ms, &(x[0])); } { V T1k, T1r, T1z, T1w, T1o, T1y, T1h, T1v; T1k = VFMA(LDK(KP951056516), T1i, VMUL(LDK(KP587785252), T1j)); T1r = VFMA(LDK(KP951056516), T1p, VMUL(LDK(KP587785252), T1q)); T1z = VFNMS(LDK(KP587785252), T1p, VMUL(LDK(KP951056516), T1q)); T1w = VFNMS(LDK(KP587785252), T1i, VMUL(LDK(KP951056516), T1j)); { V T1m, T1n, T1f, T1g; T1m = VFMS(LDK(KP250000000), T1d, TM); T1n = VMUL(LDK(KP559016994), VSUB(T1c, TY)); T1o = VADD(T1m, T1n); T1y = VSUB(T1n, T1m); T1f = VMUL(LDK(KP559016994), VSUB(Tp, TE)); T1g = VFNMS(LDK(KP250000000), TF, T7); T1h = VADD(T1f, T1g); T1v = VSUB(T1g, T1f); } { V T1l, T1s, T1B, T1C; T1l = VADD(T1h, T1k); T1s = VBYI(VSUB(T1o, T1r)); ST(&(x[WS(rs, 19)]), VSUB(T1l, T1s), ms, &(x[WS(rs, 1)])); ST(&(x[WS(rs, 1)]), VADD(T1l, T1s), ms, &(x[WS(rs, 1)])); T1B = VADD(T1v, T1w); T1C = VBYI(VADD(T1z, T1y)); ST(&(x[WS(rs, 13)]), VSUB(T1B, T1C), ms, &(x[WS(rs, 1)])); ST(&(x[WS(rs, 7)]), VADD(T1B, T1C), ms, &(x[WS(rs, 1)])); } { V T1t, T1u, T1x, T1A; T1t = VSUB(T1h, T1k); T1u = VBYI(VADD(T1r, T1o)); ST(&(x[WS(rs, 11)]), VSUB(T1t, T1u), ms, &(x[WS(rs, 1)])); ST(&(x[WS(rs, 9)]), VADD(T1t, T1u), ms, &(x[WS(rs, 1)])); T1x = VSUB(T1v, T1w); T1A = VBYI(VSUB(T1y, T1z)); ST(&(x[WS(rs, 17)]), VSUB(T1x, T1A), ms, &(x[WS(rs, 1)])); ST(&(x[WS(rs, 3)]), VADD(T1x, T1A), ms, &(x[WS(rs, 1)])); } } } } } VLEAVE(); } static const tw_instr twinstr[] = { VTW(0, 1), VTW(0, 3), VTW(0, 9), VTW(0, 19), {TW_NEXT, VL, 0} }; static const ct_desc desc = { 20, XSIMD_STRING("t3fv_20"), twinstr, &GENUS, {126, 80, 12, 0}, 0, 0, 0 }; void XSIMD(codelet_t3fv_20) (planner *p) { X(kdft_dit_register) (p, t3fv_20, &desc); } #endif /* HAVE_FMA */ fftw-3.3.3/dft/simd/common/t3fv_25.c0000644000175000017500000012025612054410547013754 00000000000000/* * Copyright (c) 2003, 2007-11 Matteo Frigo * Copyright (c) 2003, 2007-11 Massachusetts Institute of Technology * * This program is free software; you can redistribute it and/or modify * it under the terms of the GNU General Public License as published by * the Free Software Foundation; either version 2 of the License, or * (at your option) any later version. * * This program is distributed in the hope that it will be useful, * but WITHOUT ANY WARRANTY; without even the implied warranty of * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the * GNU General Public License for more details. * * You should have received a copy of the GNU General Public License * along with this program; if not, write to the Free Software * Foundation, Inc., 51 Franklin Street, Fifth Floor, Boston, MA 02110-1301 USA * */ /* This file was automatically generated --- DO NOT EDIT */ /* Generated on Sun Nov 25 07:38:56 EST 2012 */ #include "codelet-dft.h" #ifdef HAVE_FMA /* Generated by: ../../../genfft/gen_twiddle_c.native -fma -reorder-insns -schedule-for-pipeline -simd -compact -variables 4 -pipeline-latency 8 -twiddle-log3 -precompute-twiddles -no-generate-bytw -n 25 -name t3fv_25 -include t3f.h */ /* * This function contains 268 FP additions, 281 FP multiplications, * (or, 87 additions, 100 multiplications, 181 fused multiply/add), * 223 stack variables, 67 constants, and 50 memory accesses */ #include "t3f.h" static void t3fv_25(R *ri, R *ii, const R *W, stride rs, INT mb, INT me, INT ms) { DVK(KP792626838, +0.792626838241819413632131824093538848057784557); DVK(KP876091699, +0.876091699473550838204498029706869638173524346); DVK(KP617882369, +0.617882369114440893914546919006756321695042882); DVK(KP803003575, +0.803003575438660414833440593570376004635464850); DVK(KP242145790, +0.242145790282157779872542093866183953459003101); DVK(KP968583161, +0.968583161128631119490168375464735813836012403); DVK(KP999544308, +0.999544308746292983948881682379742149196758193); DVK(KP916574801, +0.916574801383451584742370439148878693530976769); DVK(KP904730450, +0.904730450839922351881287709692877908104763647); DVK(KP809385824, +0.809385824416008241660603814668679683846476688); DVK(KP447417479, +0.447417479732227551498980015410057305749330693); DVK(KP894834959, +0.894834959464455102997960030820114611498661386); DVK(KP867381224, +0.867381224396525206773171885031575671309956167); DVK(KP683113946, +0.683113946453479238701949862233725244439656928); DVK(KP559154169, +0.559154169276087864842202529084232643714075927); DVK(KP958953096, +0.958953096729998668045963838399037225970891871); DVK(KP831864738, +0.831864738706457140726048799369896829771167132); DVK(KP829049696, +0.829049696159252993975487806364305442437946767); DVK(KP860541664, +0.860541664367944677098261680920518816412804187); DVK(KP897376177, +0.897376177523557693138608077137219684419427330); DVK(KP876306680, +0.876306680043863587308115903922062583399064238); DVK(KP681693190, +0.681693190061530575150324149145440022633095390); DVK(KP560319534, +0.560319534973832390111614715371676131169633784); DVK(KP855719849, +0.855719849902058969314654733608091555096772472); DVK(KP237294955, +0.237294955877110315393888866460840817927895961); DVK(KP949179823, +0.949179823508441261575555465843363271711583843); DVK(KP904508497, +0.904508497187473712051146708591409529430077295); DVK(KP997675361, +0.997675361079556513670859573984492383596555031); DVK(KP763932022, +0.763932022500210303590826331268723764559381640); DVK(KP690983005, +0.690983005625052575897706582817180941139845410); DVK(KP992114701, +0.992114701314477831049793042785778521453036709); DVK(KP952936919, +0.952936919628306576880750665357914584765951388); DVK(KP998026728, +0.998026728428271561952336806863450553336905220); DVK(KP262346850, +0.262346850930607871785420028382979691334784273); DVK(KP570584518, +0.570584518783621657366766175430996792655723863); DVK(KP669429328, +0.669429328479476605641803240971985825917022098); DVK(KP923225144, +0.923225144846402650453449441572664695995209956); DVK(KP945422727, +0.945422727388575946270360266328811958657216298); DVK(KP522616830, +0.522616830205754336872861364785224694908468440); DVK(KP956723877, +0.956723877038460305821989399535483155872969262); DVK(KP906616052, +0.906616052148196230441134447086066874408359177); DVK(KP772036680, +0.772036680810363904029489473607579825330539880); DVK(KP845997307, +0.845997307939530944175097360758058292389769300); DVK(KP951056516, +0.951056516295153572116439333379382143405698634); DVK(KP912575812, +0.912575812670962425556968549836277086778922727); DVK(KP921078979, +0.921078979742360627699756128143719920817673854); DVK(KP982009705, +0.982009705009746369461829878184175962711969869); DVK(KP734762448, +0.734762448793050413546343770063151342619912334); DVK(KP494780565, +0.494780565770515410344588413655324772219443730); DVK(KP447533225, +0.447533225982656890041886979663652563063114397); DVK(KP603558818, +0.603558818296015001454675132653458027918768137); DVK(KP667278218, +0.667278218140296670899089292254759909713898805); DVK(KP244189809, +0.244189809627953270309879511234821255780225091); DVK(KP269969613, +0.269969613759572083574752974412347470060951301); DVK(KP578046249, +0.578046249379945007321754579646815604023525655); DVK(KP522847744, +0.522847744331509716623755382187077770911012542); DVK(KP132830569, +0.132830569247582714407653942074819768844536507); DVK(KP120146378, +0.120146378570687701782758537356596213647956445); DVK(KP893101515, +0.893101515366181661711202267938416198338079437); DVK(KP987388751, +0.987388751065621252324603216482382109400433949); DVK(KP059835404, +0.059835404262124915169548397419498386427871950); DVK(KP066152395, +0.066152395967733048213034281011006031460903353); DVK(KP786782374, +0.786782374965295178365099601674911834788448471); DVK(KP869845200, +0.869845200362138853122720822420327157933056305); DVK(KP559016994, +0.559016994374947424102293417182819058860154590); DVK(KP250000000, +0.250000000000000000000000000000000000000000000); DVK(KP618033988, +0.618033988749894848204586834365638117720309180); { INT m; R *x; x = ri; for (m = mb, W = W + (mb * ((TWVL / VL) * 8)); m < me; m = m + VL, x = x + (VL * ms), W = W + (TWVL * 8), MAKE_VOLATILE_STRIDE(25, rs)) { V T2t, T1Z, T2W, T28, T2Q, T2r, T2g, T2u, T2o, T2l; { V T2, T5, T3, T9; T2 = LDW(&(W[0])); T5 = LDW(&(W[TWVL * 4])); T3 = LDW(&(W[TWVL * 2])); T9 = LDW(&(W[TWVL * 6])); { V T2c, T3l, Tn, T49, Tm, T4e, TN, T32, T1d, T3a, T3f, T3z, T3H, T25, T1W; V T2v, T2D, T4a, T1g, T18, T2Z, T11, T31, TK, T1q, T1j, T1n, T4b, T17; { V T1, T1l, Tr, T4, Ty, T1E, Tu, TX, TD, T1h, Tz, T1e, T1I, T1o, TU; V Tk, T2b, T1B, T1D, T1N, T1F, Td, T2a, T1J; { V T7, Tb, TC, Tg, T1L, Ta, T6, Tj, T1A; T1 = LD(&(x[0]), ms, &(x[0])); { V Tf, Ti, Te, Th; Tf = LD(&(x[WS(rs, 10)]), ms, &(x[0])); Ti = LD(&(x[WS(rs, 15)]), ms, &(x[WS(rs, 1)])); T7 = LD(&(x[WS(rs, 5)]), ms, &(x[WS(rs, 1)])); Tb = LD(&(x[WS(rs, 20)]), ms, &(x[0])); Te = VZMUL(T2, T5); TC = VZMULJ(T2, T5); T1l = VZMUL(T3, T5); Tr = VZMULJ(T3, T5); T4 = VZMUL(T2, T3); Ty = VZMULJ(T2, T3); T1E = VZMULJ(T2, T9); Th = VZMULJ(T5, T9); Tu = VZMULJ(T3, T9); Tg = VZMULJ(Te, Tf); TX = VZMULJ(Te, T9); TD = VZMULJ(TC, T9); T1h = VZMULJ(Ty, T9); Tz = VZMUL(Ty, T5); T1e = VZMULJ(Ty, T5); T1L = VZMULJ(Tr, T9); Ta = VZMULJ(T4, T9); T1I = VZMUL(T4, T5); T6 = VZMULJ(T4, T5); Tj = VZMULJ(Th, Ti); } T1A = LD(&(x[WS(rs, 3)]), ms, &(x[WS(rs, 1)])); T1o = VZMULJ(T1e, T9); { V Tc, T8, T1C, T1M; T1C = LD(&(x[WS(rs, 8)]), ms, &(x[0])); T1M = LD(&(x[WS(rs, 18)]), ms, &(x[0])); Tc = VZMULJ(Ta, Tb); T8 = VZMULJ(T6, T7); TU = VZMULJ(T6, T9); Tk = VADD(Tg, Tj); T2b = VSUB(Tg, Tj); T1B = VZMULJ(T3, T1A); T1D = VZMULJ(TC, T1C); T1N = VZMULJ(T1L, T1M); T1F = LD(&(x[WS(rs, 23)]), ms, &(x[WS(rs, 1)])); Td = VADD(T8, Tc); T2a = VSUB(T8, Tc); T1J = LD(&(x[WS(rs, 13)]), ms, &(x[WS(rs, 1)])); } } { V Tq, Tt, TF, T1T, T1H, Tw, T1U, T1O, TA, Tp, Ts, TE; Tp = LD(&(x[WS(rs, 1)]), ms, &(x[WS(rs, 1)])); Ts = LD(&(x[WS(rs, 6)]), ms, &(x[0])); TE = LD(&(x[WS(rs, 16)]), ms, &(x[0])); { V T1K, Tv, T1G, Tl; Tv = LD(&(x[WS(rs, 21)]), ms, &(x[WS(rs, 1)])); T1G = VZMULJ(T1E, T1F); T2c = VFMA(LDK(KP618033988), T2b, T2a); T3l = VFNMS(LDK(KP618033988), T2a, T2b); Tn = VSUB(Td, Tk); Tl = VADD(Td, Tk); T1K = VZMULJ(T1I, T1J); Tq = VZMULJ(T2, Tp); Tt = VZMULJ(Tr, Ts); TF = VZMULJ(TD, TE); T1T = VSUB(T1D, T1G); T1H = VADD(T1D, T1G); T49 = VADD(T1, Tl); Tm = VFNMS(LDK(KP250000000), Tl, T1); Tw = VZMULJ(Tu, Tv); T1U = VSUB(T1K, T1N); T1O = VADD(T1K, T1N); TA = LD(&(x[WS(rs, 11)]), ms, &(x[WS(rs, 1)])); } { V Tx, TL, T1R, T38, T1V, T13, TQ, TZ, TS, T1Q, TV, TG, TM, T12, T1c; V T16; T12 = LD(&(x[WS(rs, 4)]), ms, &(x[0])); { V TP, TY, T1P, TB, TR; TP = LD(&(x[WS(rs, 24)]), ms, &(x[0])); TY = LD(&(x[WS(rs, 14)]), ms, &(x[0])); TR = LD(&(x[WS(rs, 9)]), ms, &(x[WS(rs, 1)])); Tx = VADD(Tt, Tw); TL = VSUB(Tt, Tw); T1R = VSUB(T1O, T1H); T1P = VADD(T1H, T1O); T38 = VFNMS(LDK(KP618033988), T1T, T1U); T1V = VFMA(LDK(KP618033988), T1U, T1T); TB = VZMULJ(Tz, TA); T13 = VZMULJ(T4, T12); TQ = VZMULJ(T9, TP); TZ = VZMULJ(TX, TY); TS = VZMULJ(T5, TR); T4e = VADD(T1B, T1P); T1Q = VFNMS(LDK(KP250000000), T1P, T1B); TV = LD(&(x[WS(rs, 19)]), ms, &(x[WS(rs, 1)])); TG = VADD(TB, TF); TM = VSUB(TF, TB); } T1c = LD(&(x[WS(rs, 2)]), ms, &(x[0])); { V T14, TT, TJ, T15, T10, TI, T1p, T1f, T1i, T1m; T1f = LD(&(x[WS(rs, 7)]), ms, &(x[WS(rs, 1)])); T14 = VADD(TS, TQ); TT = VSUB(TQ, TS); { V T39, T1S, TW, TH; T39 = VFMA(LDK(KP559016994), T1R, T1Q); T1S = VFNMS(LDK(KP559016994), T1R, T1Q); TW = VZMULJ(TU, TV); TH = VADD(Tx, TG); TJ = VSUB(Tx, TG); TN = VFNMS(LDK(KP618033988), TM, TL); T32 = VFMA(LDK(KP618033988), TL, TM); T1d = VZMULJ(Ty, T1c); T3a = VFMA(LDK(KP869845200), T39, T38); T3f = VFNMS(LDK(KP786782374), T38, T39); T3z = VFMA(LDK(KP066152395), T39, T38); T3H = VFNMS(LDK(KP059835404), T38, T39); T25 = VFMA(LDK(KP987388751), T1S, T1V); T1W = VFNMS(LDK(KP893101515), T1V, T1S); T2v = VFNMS(LDK(KP120146378), T1V, T1S); T2D = VFMA(LDK(KP132830569), T1S, T1V); T15 = VADD(TZ, TW); T10 = VSUB(TW, TZ); TI = VFNMS(LDK(KP250000000), TH, Tq); T4a = VADD(Tq, TH); T1g = VZMULJ(T1e, T1f); } T1p = LD(&(x[WS(rs, 17)]), ms, &(x[WS(rs, 1)])); T1i = LD(&(x[WS(rs, 22)]), ms, &(x[0])); T1m = LD(&(x[WS(rs, 12)]), ms, &(x[0])); T18 = VSUB(T14, T15); T16 = VADD(T14, T15); T2Z = VFNMS(LDK(KP618033988), TT, T10); T11 = VFMA(LDK(KP618033988), T10, TT); T31 = VFNMS(LDK(KP559016994), TJ, TI); TK = VFMA(LDK(KP559016994), TJ, TI); T1q = VZMULJ(T1o, T1p); T1j = VZMULJ(T1h, T1i); T1n = VZMULJ(T1l, T1m); } T4b = VADD(T13, T16); T17 = VFMS(LDK(KP250000000), T16, T13); } } } { V T33, T3i, T3C, T3L, T20, TO, T2y, T2G, T1k, T1w, T1r, T1x, T2Y, T19, T4k; V T4c; T33 = VFMA(LDK(KP893101515), T32, T31); T3i = VFNMS(LDK(KP987388751), T31, T32); T3C = VFNMS(LDK(KP522847744), T32, T31); T3L = VFMA(LDK(KP578046249), T31, T32); T20 = VFMA(LDK(KP269969613), TK, TN); TO = VFNMS(LDK(KP244189809), TN, TK); T2y = VFMA(LDK(KP667278218), TK, TN); T2G = VFNMS(LDK(KP603558818), TN, TK); T1k = VADD(T1g, T1j); T1w = VSUB(T1g, T1j); T1r = VADD(T1n, T1q); T1x = VSUB(T1q, T1n); T2Y = VFMA(LDK(KP559016994), T18, T17); T19 = VFNMS(LDK(KP559016994), T18, T17); T4k = VSUB(T4a, T4b); T4c = VADD(T4a, T4b); { V T2X, To, T35, T1y, T2H, T2z, T1a, T21, T3t, T34, T3n, T3j, T3E, T3Y, T3M; V T3R, T1v, T36, T4l, T4f, T1u, T1s; T2X = VFNMS(LDK(KP559016994), Tn, Tm); To = VFMA(LDK(KP559016994), Tn, Tm); T1u = VSUB(T1r, T1k); T1s = VADD(T1k, T1r); T35 = VFMA(LDK(KP618033988), T1w, T1x); T1y = VFNMS(LDK(KP618033988), T1x, T1w); { V T3K, T30, T3h, T3D, T4d, T1t; T3K = VFMA(LDK(KP447533225), T2Z, T2Y); T30 = VFMA(LDK(KP120146378), T2Z, T2Y); T3h = VFNMS(LDK(KP132830569), T2Y, T2Z); T3D = VFNMS(LDK(KP494780565), T2Y, T2Z); T2H = VFNMS(LDK(KP786782374), T11, T19); T2z = VFMA(LDK(KP869845200), T19, T11); T1a = VFNMS(LDK(KP667278218), T19, T11); T21 = VFMA(LDK(KP603558818), T11, T19); T4d = VADD(T1d, T1s); T1t = VFNMS(LDK(KP250000000), T1s, T1d); T3t = VFNMS(LDK(KP734762448), T33, T30); T34 = VFMA(LDK(KP734762448), T33, T30); T3n = VFMA(LDK(KP734762448), T3i, T3h); T3j = VFNMS(LDK(KP734762448), T3i, T3h); T3E = VFNMS(LDK(KP982009705), T3D, T3C); T3Y = VFMA(LDK(KP982009705), T3D, T3C); T3M = VFNMS(LDK(KP921078979), T3L, T3K); T3R = VFMA(LDK(KP921078979), T3L, T3K); T1v = VFNMS(LDK(KP559016994), T1u, T1t); T36 = VFMA(LDK(KP559016994), T1u, T1t); T4l = VSUB(T4d, T4e); T4f = VADD(T4d, T4e); } { V T2L, T2R, T2j, T2q, T2J, T2B, T2e, T26, T2U, T1Y, T23, T2O; { V T2I, T24, T2w, T2E, T48, T42, T3y, T3s, T3V, T45, T2A, T1b, T2h, T2i, T1X; T2L = VFNMS(LDK(KP912575812), T2H, T2G); T2I = VFMA(LDK(KP912575812), T2H, T2G); { V T3A, T3e, T37, T3I, T1z; T3A = VFNMS(LDK(KP667278218), T36, T35); T3e = VFNMS(LDK(KP059835404), T35, T36); T37 = VFMA(LDK(KP066152395), T36, T35); T3I = VFMA(LDK(KP603558818), T35, T36); T24 = VFMA(LDK(KP578046249), T1v, T1y); T1z = VFNMS(LDK(KP522847744), T1y, T1v); T2w = VFNMS(LDK(KP494780565), T1v, T1y); T2E = VFMA(LDK(KP447533225), T1y, T1v); { V T4i, T4g, T4o, T4m; T4i = VSUB(T4c, T4f); T4g = VADD(T4c, T4f); T4o = VMUL(LDK(KP951056516), VFNMS(LDK(KP618033988), T4k, T4l)); T4m = VMUL(LDK(KP951056516), VFMA(LDK(KP618033988), T4l, T4k)); { V T3Q, T3J, T3b, T3u; T3Q = VFNMS(LDK(KP845997307), T3I, T3H); T3J = VFMA(LDK(KP845997307), T3I, T3H); T3b = VFNMS(LDK(KP772036680), T3a, T37); T3u = VFMA(LDK(KP772036680), T3a, T37); { V T3o, T3g, T3B, T3X, T4h; T3o = VFNMS(LDK(KP772036680), T3f, T3e); T3g = VFMA(LDK(KP772036680), T3f, T3e); T3B = VFNMS(LDK(KP845997307), T3A, T3z); T3X = VFMA(LDK(KP845997307), T3A, T3z); ST(&(x[0]), VADD(T4g, T49), ms, &(x[0])); T4h = VFNMS(LDK(KP250000000), T4g, T49); { V T40, T3N, T3c, T3v; T40 = VFMA(LDK(KP906616052), T3M, T3J); T3N = VFNMS(LDK(KP906616052), T3M, T3J); T3c = VFMA(LDK(KP956723877), T3b, T34); T3v = VFMA(LDK(KP522616830), T3j, T3u); { V T3p, T3k, T3S, T3F; T3p = VFNMS(LDK(KP522616830), T34, T3o); T3k = VFMA(LDK(KP945422727), T3j, T3g); T3S = VFNMS(LDK(KP923225144), T3E, T3B); T3F = VFMA(LDK(KP923225144), T3E, T3B); { V T46, T3Z, T4j, T4n; T46 = VFNMS(LDK(KP669429328), T3X, T3Y); T3Z = VFMA(LDK(KP570584518), T3Y, T3X); T4j = VFMA(LDK(KP559016994), T4i, T4h); T4n = VFNMS(LDK(KP559016994), T4i, T4h); { V T3W, T3O, T3d, T3w; T3W = VFMA(LDK(KP262346850), T3N, T3l); T3O = VMUL(LDK(KP998026728), VFNMS(LDK(KP952936919), T3l, T3N)); T3d = VFMA(LDK(KP992114701), T3c, T2X); T3w = VFNMS(LDK(KP690983005), T3v, T3g); { V T3q, T3m, T3T, T43; T3q = VFMA(LDK(KP763932022), T3p, T3b); T3m = VMUL(LDK(KP998026728), VFMA(LDK(KP952936919), T3l, T3k)); T3T = VFNMS(LDK(KP997675361), T3S, T3R); T43 = VFNMS(LDK(KP904508497), T3S, T3Q); { V T3G, T3P, T47, T41; T3G = VFMA(LDK(KP949179823), T3F, T2X); T3P = VFNMS(LDK(KP237294955), T3F, T2X); T47 = VFNMS(LDK(KP669429328), T40, T46); T41 = VFMA(LDK(KP618033988), T40, T3Z); ST(&(x[WS(rs, 20)]), VFMAI(T4m, T4j), ms, &(x[0])); ST(&(x[WS(rs, 5)]), VFNMSI(T4m, T4j), ms, &(x[WS(rs, 1)])); ST(&(x[WS(rs, 15)]), VFNMSI(T4o, T4n), ms, &(x[WS(rs, 1)])); ST(&(x[WS(rs, 10)]), VFMAI(T4o, T4n), ms, &(x[0])); { V T3x, T3r, T3U, T44; T3x = VFMA(LDK(KP855719849), T3w, T3t); T3r = VFNMS(LDK(KP855719849), T3q, T3n); ST(&(x[WS(rs, 22)]), VFMAI(T3m, T3d), ms, &(x[0])); ST(&(x[WS(rs, 3)]), VFNMSI(T3m, T3d), ms, &(x[WS(rs, 1)])); T3U = VFMA(LDK(KP560319534), T3T, T3Q); T44 = VFNMS(LDK(KP681693190), T43, T3R); ST(&(x[WS(rs, 23)]), VFMAI(T3O, T3G), ms, &(x[WS(rs, 1)])); ST(&(x[WS(rs, 2)]), VFNMSI(T3O, T3G), ms, &(x[0])); T48 = VMUL(LDK(KP951056516), VFNMS(LDK(KP876306680), T47, T3W)); T42 = VMUL(LDK(KP951056516), VFNMS(LDK(KP949179823), T41, T3W)); T3y = VMUL(LDK(KP951056516), VFNMS(LDK(KP992114701), T3x, T3l)); T3s = VFMA(LDK(KP897376177), T3r, T2X); T3V = VFNMS(LDK(KP949179823), T3U, T3P); T45 = VFNMS(LDK(KP860541664), T44, T3P); T2R = VFNMS(LDK(KP912575812), T2z, T2y); T2A = VFMA(LDK(KP912575812), T2z, T2y); T1b = VFMA(LDK(KP829049696), T1a, TO); T2h = VFNMS(LDK(KP829049696), T1a, TO); T2i = VFNMS(LDK(KP831864738), T1W, T1z); T1X = VFMA(LDK(KP831864738), T1W, T1z); } } } } } } } } } } } { V T2M, T2F, T2x, T2S, T2T, T2N; T2M = VFNMS(LDK(KP958953096), T2E, T2D); T2F = VFMA(LDK(KP958953096), T2E, T2D); ST(&(x[WS(rs, 17)]), VFMAI(T3y, T3s), ms, &(x[WS(rs, 1)])); ST(&(x[WS(rs, 8)]), VFNMSI(T3y, T3s), ms, &(x[0])); ST(&(x[WS(rs, 12)]), VFMAI(T42, T3V), ms, &(x[0])); ST(&(x[WS(rs, 13)]), VFNMSI(T42, T3V), ms, &(x[WS(rs, 1)])); ST(&(x[WS(rs, 18)]), VFNMSI(T48, T45), ms, &(x[0])); ST(&(x[WS(rs, 7)]), VFMAI(T48, T45), ms, &(x[WS(rs, 1)])); T2j = VFMA(LDK(KP559154169), T2i, T2h); T2q = VFNMS(LDK(KP683113946), T2h, T2i); T2x = VFNMS(LDK(KP867381224), T2w, T2v); T2S = VFMA(LDK(KP867381224), T2w, T2v); T2J = VFMA(LDK(KP894834959), T2I, T2F); T2T = VFMA(LDK(KP447417479), T2I, T2S); T2B = VFNMS(LDK(KP809385824), T2A, T2x); T2N = VFMA(LDK(KP447417479), T2A, T2M); T2e = VFMA(LDK(KP831864738), T25, T24); T26 = VFNMS(LDK(KP831864738), T25, T24); T2U = VFNMS(LDK(KP763932022), T2T, T2F); T1Y = VFMA(LDK(KP904730450), T1X, T1b); T23 = VFNMS(LDK(KP904730450), T1X, T1b); T2O = VFMA(LDK(KP690983005), T2N, T2x); } } { V T2C, T22, T2d, T2K; T2C = VFNMS(LDK(KP992114701), T2B, To); T22 = VFMA(LDK(KP916574801), T21, T20); T2d = VFNMS(LDK(KP916574801), T21, T20); T2K = VMUL(LDK(KP951056516), VFNMS(LDK(KP992114701), T2J, T2c)); { V T27, T2P, T2f, T2k, T2n, T2V; T2V = VFNMS(LDK(KP999544308), T2U, T2R); T27 = VFNMS(LDK(KP904730450), T26, T23); T2t = VFMA(LDK(KP968583161), T1Y, To); T1Z = VFNMS(LDK(KP242145790), T1Y, To); T2P = VFNMS(LDK(KP999544308), T2O, T2L); T2f = VFMA(LDK(KP904730450), T2e, T2d); T2k = VFNMS(LDK(KP904730450), T2e, T2d); T2n = VADD(T22, T23); ST(&(x[WS(rs, 21)]), VFNMSI(T2K, T2C), ms, &(x[WS(rs, 1)])); ST(&(x[WS(rs, 4)]), VFMAI(T2K, T2C), ms, &(x[0])); T2W = VMUL(LDK(KP951056516), VFNMS(LDK(KP803003575), T2V, T2c)); T28 = VFNMS(LDK(KP618033988), T27, T22); T2Q = VFNMS(LDK(KP803003575), T2P, To); T2r = VFMA(LDK(KP617882369), T2k, T2q); T2g = VFNMS(LDK(KP242145790), T2f, T2c); T2u = VMUL(LDK(KP951056516), VFMA(LDK(KP968583161), T2f, T2c)); T2o = VFNMS(LDK(KP683113946), T2n, T26); T2l = VFMA(LDK(KP559016994), T2k, T2j); } } } } } } } { V T29, T2s, T2p, T2m; T29 = VFNMS(LDK(KP876091699), T28, T1Z); ST(&(x[WS(rs, 9)]), VFMAI(T2W, T2Q), ms, &(x[WS(rs, 1)])); ST(&(x[WS(rs, 16)]), VFNMSI(T2W, T2Q), ms, &(x[0])); T2s = VMUL(LDK(KP951056516), VFNMS(LDK(KP876306680), T2r, T2g)); ST(&(x[WS(rs, 24)]), VFMAI(T2u, T2t), ms, &(x[0])); ST(&(x[WS(rs, 1)]), VFNMSI(T2u, T2t), ms, &(x[WS(rs, 1)])); T2p = VFMA(LDK(KP792626838), T2o, T1Z); T2m = VMUL(LDK(KP951056516), VFMA(LDK(KP968583161), T2l, T2g)); ST(&(x[WS(rs, 11)]), VFNMSI(T2s, T2p), ms, &(x[WS(rs, 1)])); ST(&(x[WS(rs, 14)]), VFMAI(T2s, T2p), ms, &(x[0])); ST(&(x[WS(rs, 19)]), VFMAI(T2m, T29), ms, &(x[WS(rs, 1)])); ST(&(x[WS(rs, 6)]), VFNMSI(T2m, T29), ms, &(x[0])); } } } VLEAVE(); } static const tw_instr twinstr[] = { VTW(0, 1), VTW(0, 3), VTW(0, 9), VTW(0, 24), {TW_NEXT, VL, 0} }; static const ct_desc desc = { 25, XSIMD_STRING("t3fv_25"), twinstr, &GENUS, {87, 100, 181, 0}, 0, 0, 0 }; void XSIMD(codelet_t3fv_25) (planner *p) { X(kdft_dit_register) (p, t3fv_25, &desc); } #else /* HAVE_FMA */ /* Generated by: ../../../genfft/gen_twiddle_c.native -simd -compact -variables 4 -pipeline-latency 8 -twiddle-log3 -precompute-twiddles -no-generate-bytw -n 25 -name t3fv_25 -include t3f.h */ /* * This function contains 268 FP additions, 228 FP multiplications, * (or, 190 additions, 150 multiplications, 78 fused multiply/add), * 123 stack variables, 40 constants, and 50 memory accesses */ #include "t3f.h" static void t3fv_25(R *ri, R *ii, const R *W, stride rs, INT mb, INT me, INT ms) { DVK(KP998026728, +0.998026728428271561952336806863450553336905220); DVK(KP125581039, +0.125581039058626752152356449131262266244969664); DVK(KP1_996053456, +1.996053456856543123904673613726901106673810439); DVK(KP062790519, +0.062790519529313376076178224565631133122484832); DVK(KP809016994, +0.809016994374947424102293417182819058860154590); DVK(KP309016994, +0.309016994374947424102293417182819058860154590); DVK(KP1_369094211, +1.369094211857377347464566715242418539779038465); DVK(KP728968627, +0.728968627421411523146730319055259111372571664); DVK(KP963507348, +0.963507348203430549974383005744259307057084020); DVK(KP876306680, +0.876306680043863587308115903922062583399064238); DVK(KP497379774, +0.497379774329709576484567492012895936835134813); DVK(KP968583161, +0.968583161128631119490168375464735813836012403); DVK(KP684547105, +0.684547105928688673732283357621209269889519233); DVK(KP1_457937254, +1.457937254842823046293460638110518222745143328); DVK(KP481753674, +0.481753674101715274987191502872129653528542010); DVK(KP1_752613360, +1.752613360087727174616231807844125166798128477); DVK(KP248689887, +0.248689887164854788242283746006447968417567406); DVK(KP1_937166322, +1.937166322257262238980336750929471627672024806); DVK(KP992114701, +0.992114701314477831049793042785778521453036709); DVK(KP250666467, +0.250666467128608490746237519633017587885836494); DVK(KP425779291, +0.425779291565072648862502445744251703979973042); DVK(KP1_809654104, +1.809654104932039055427337295865395187940827822); DVK(KP1_274847979, +1.274847979497379420353425623352032390869834596); DVK(KP770513242, +0.770513242775789230803009636396177847271667672); DVK(KP844327925, +0.844327925502015078548558063966681505381659241); DVK(KP1_071653589, +1.071653589957993236542617535735279956127150691); DVK(KP125333233, +0.125333233564304245373118759816508793942918247); DVK(KP1_984229402, +1.984229402628955662099586085571557042906073418); DVK(KP904827052, +0.904827052466019527713668647932697593970413911); DVK(KP851558583, +0.851558583130145297725004891488503407959946084); DVK(KP637423989, +0.637423989748689710176712811676016195434917298); DVK(KP1_541026485, +1.541026485551578461606019272792355694543335344); DVK(KP535826794, +0.535826794978996618271308767867639978063575346); DVK(KP1_688655851, +1.688655851004030157097116127933363010763318483); DVK(KP293892626, +0.293892626146236564584352977319536384298826219); DVK(KP475528258, +0.475528258147576786058219666689691071702849317); DVK(KP587785252, +0.587785252292473129168705954639072768597652438); DVK(KP951056516, +0.951056516295153572116439333379382143405698634); DVK(KP250000000, +0.250000000000000000000000000000000000000000000); DVK(KP559016994, +0.559016994374947424102293417182819058860154590); { INT m; R *x; x = ri; for (m = mb, W = W + (mb * ((TWVL / VL) * 8)); m < me; m = m + VL, x = x + (VL * ms), W = W + (TWVL * 8), MAKE_VOLATILE_STRIDE(25, rs)) { V T1, T4, T2, T3, TA, Td, Tp, Tw, Tx, T1G, T1j, T5, T1c, T8, T9; V Ts, T1J, Tg, T1C, T1m, TX, TB, T1f, TU; T1 = LDW(&(W[0])); T4 = LDW(&(W[TWVL * 4])); T2 = LDW(&(W[TWVL * 2])); T3 = VZMUL(T1, T2); TA = VZMULJ(T1, T4); Td = VZMUL(T1, T4); Tp = VZMULJ(T2, T4); Tw = VZMULJ(T1, T2); Tx = VZMUL(Tw, T4); T1G = VZMUL(T3, T4); T1j = VZMUL(T2, T4); T5 = VZMULJ(T3, T4); T1c = VZMULJ(Tw, T4); T8 = LDW(&(W[TWVL * 6])); T9 = VZMULJ(T3, T8); Ts = VZMULJ(T2, T8); T1J = VZMULJ(Tp, T8); Tg = VZMULJ(T4, T8); T1C = VZMULJ(T1, T8); T1m = VZMULJ(T1c, T8); TX = VZMULJ(T5, T8); TB = VZMULJ(TA, T8); T1f = VZMULJ(Tw, T8); TU = VZMULJ(Td, T8); { V Tl, Tk, Tm, Tn, T20, T2R, T22, T1V, T2K, T1S, T3A, T2L, TN, T2G, TK; V T3w, T2H, T19, T2D, T16, T3x, T2E, T1y, T2N, T1v, T3z, T2O; { V Tf, Ti, Tj, T7, Tb, Tc, T21; Tl = LD(&(x[0]), ms, &(x[0])); { V Te, Th, T6, Ta; Te = LD(&(x[WS(rs, 10)]), ms, &(x[0])); Tf = VZMULJ(Td, Te); Th = LD(&(x[WS(rs, 15)]), ms, &(x[WS(rs, 1)])); Ti = VZMULJ(Tg, Th); Tj = VADD(Tf, Ti); T6 = LD(&(x[WS(rs, 5)]), ms, &(x[WS(rs, 1)])); T7 = VZMULJ(T5, T6); Ta = LD(&(x[WS(rs, 20)]), ms, &(x[0])); Tb = VZMULJ(T9, Ta); Tc = VADD(T7, Tb); } Tk = VMUL(LDK(KP559016994), VSUB(Tc, Tj)); Tm = VADD(Tc, Tj); Tn = VFNMS(LDK(KP250000000), Tm, Tl); T20 = VSUB(T7, Tb); T21 = VSUB(Tf, Ti); T2R = VMUL(LDK(KP951056516), T21); T22 = VFMA(LDK(KP951056516), T20, VMUL(LDK(KP587785252), T21)); } { V T1P, T1I, T1L, T1M, T1B, T1E, T1F, T1O; T1O = LD(&(x[WS(rs, 3)]), ms, &(x[WS(rs, 1)])); T1P = VZMULJ(T2, T1O); { V T1H, T1K, T1A, T1D; T1H = LD(&(x[WS(rs, 13)]), ms, &(x[WS(rs, 1)])); T1I = VZMULJ(T1G, T1H); T1K = LD(&(x[WS(rs, 18)]), ms, &(x[0])); T1L = VZMULJ(T1J, T1K); T1M = VADD(T1I, T1L); T1A = LD(&(x[WS(rs, 8)]), ms, &(x[0])); T1B = VZMULJ(TA, T1A); T1D = LD(&(x[WS(rs, 23)]), ms, &(x[WS(rs, 1)])); T1E = VZMULJ(T1C, T1D); T1F = VADD(T1B, T1E); } { V T1T, T1U, T1N, T1Q, T1R; T1T = VSUB(T1B, T1E); T1U = VSUB(T1I, T1L); T1V = VFMA(LDK(KP475528258), T1T, VMUL(LDK(KP293892626), T1U)); T2K = VFNMS(LDK(KP293892626), T1T, VMUL(LDK(KP475528258), T1U)); T1N = VMUL(LDK(KP559016994), VSUB(T1F, T1M)); T1Q = VADD(T1F, T1M); T1R = VFNMS(LDK(KP250000000), T1Q, T1P); T1S = VADD(T1N, T1R); T3A = VADD(T1P, T1Q); T2L = VSUB(T1R, T1N); } } { V TH, Tz, TD, TE, Tr, Tu, Tv, TG; TG = LD(&(x[WS(rs, 1)]), ms, &(x[WS(rs, 1)])); TH = VZMULJ(T1, TG); { V Ty, TC, Tq, Tt; Ty = LD(&(x[WS(rs, 11)]), ms, &(x[WS(rs, 1)])); Tz = VZMULJ(Tx, Ty); TC = LD(&(x[WS(rs, 16)]), ms, &(x[0])); TD = VZMULJ(TB, TC); TE = VADD(Tz, TD); Tq = LD(&(x[WS(rs, 6)]), ms, &(x[0])); Tr = VZMULJ(Tp, Tq); Tt = LD(&(x[WS(rs, 21)]), ms, &(x[WS(rs, 1)])); Tu = VZMULJ(Ts, Tt); Tv = VADD(Tr, Tu); } { V TL, TM, TF, TI, TJ; TL = VSUB(Tr, Tu); TM = VSUB(Tz, TD); TN = VFMA(LDK(KP475528258), TL, VMUL(LDK(KP293892626), TM)); T2G = VFNMS(LDK(KP293892626), TL, VMUL(LDK(KP475528258), TM)); TF = VMUL(LDK(KP559016994), VSUB(Tv, TE)); TI = VADD(Tv, TE); TJ = VFNMS(LDK(KP250000000), TI, TH); TK = VADD(TF, TJ); T3w = VADD(TH, TI); T2H = VSUB(TJ, TF); } } { V T13, TW, TZ, T10, TQ, TS, TT, T12; T12 = LD(&(x[WS(rs, 4)]), ms, &(x[0])); T13 = VZMULJ(T3, T12); { V TV, TY, TP, TR; TV = LD(&(x[WS(rs, 14)]), ms, &(x[0])); TW = VZMULJ(TU, TV); TY = LD(&(x[WS(rs, 19)]), ms, &(x[WS(rs, 1)])); TZ = VZMULJ(TX, TY); T10 = VADD(TW, TZ); TP = LD(&(x[WS(rs, 9)]), ms, &(x[WS(rs, 1)])); TQ = VZMULJ(T4, TP); TR = LD(&(x[WS(rs, 24)]), ms, &(x[0])); TS = VZMULJ(T8, TR); TT = VADD(TQ, TS); } { V T17, T18, T11, T14, T15; T17 = VSUB(TQ, TS); T18 = VSUB(TW, TZ); T19 = VFMA(LDK(KP475528258), T17, VMUL(LDK(KP293892626), T18)); T2D = VFNMS(LDK(KP293892626), T17, VMUL(LDK(KP475528258), T18)); T11 = VMUL(LDK(KP559016994), VSUB(TT, T10)); T14 = VADD(TT, T10); T15 = VFNMS(LDK(KP250000000), T14, T13); T16 = VADD(T11, T15); T3x = VADD(T13, T14); T2E = VSUB(T15, T11); } } { V T1s, T1l, T1o, T1p, T1e, T1h, T1i, T1r; T1r = LD(&(x[WS(rs, 2)]), ms, &(x[0])); T1s = VZMULJ(Tw, T1r); { V T1k, T1n, T1d, T1g; T1k = LD(&(x[WS(rs, 12)]), ms, &(x[0])); T1l = VZMULJ(T1j, T1k); T1n = LD(&(x[WS(rs, 17)]), ms, &(x[WS(rs, 1)])); T1o = VZMULJ(T1m, T1n); T1p = VADD(T1l, T1o); T1d = LD(&(x[WS(rs, 7)]), ms, &(x[WS(rs, 1)])); T1e = VZMULJ(T1c, T1d); T1g = LD(&(x[WS(rs, 22)]), ms, &(x[0])); T1h = VZMULJ(T1f, T1g); T1i = VADD(T1e, T1h); } { V T1w, T1x, T1q, T1t, T1u; T1w = VSUB(T1e, T1h); T1x = VSUB(T1l, T1o); T1y = VFMA(LDK(KP475528258), T1w, VMUL(LDK(KP293892626), T1x)); T2N = VFNMS(LDK(KP293892626), T1w, VMUL(LDK(KP475528258), T1x)); T1q = VMUL(LDK(KP559016994), VSUB(T1i, T1p)); T1t = VADD(T1i, T1p); T1u = VFNMS(LDK(KP250000000), T1t, T1s); T1v = VADD(T1q, T1u); T3z = VADD(T1s, T1t); T2O = VSUB(T1u, T1q); } } { V T3J, T3K, T3D, T3E, T3C, T3F, T3L, T3G; { V T3H, T3I, T3y, T3B; T3H = VSUB(T3w, T3x); T3I = VSUB(T3z, T3A); T3J = VBYI(VFMA(LDK(KP951056516), T3H, VMUL(LDK(KP587785252), T3I))); T3K = VBYI(VFNMS(LDK(KP587785252), T3H, VMUL(LDK(KP951056516), T3I))); T3D = VADD(Tl, Tm); T3y = VADD(T3w, T3x); T3B = VADD(T3z, T3A); T3E = VADD(T3y, T3B); T3C = VMUL(LDK(KP559016994), VSUB(T3y, T3B)); T3F = VFNMS(LDK(KP250000000), T3E, T3D); } ST(&(x[0]), VADD(T3D, T3E), ms, &(x[0])); T3L = VSUB(T3F, T3C); ST(&(x[WS(rs, 10)]), VADD(T3K, T3L), ms, &(x[0])); ST(&(x[WS(rs, 15)]), VSUB(T3L, T3K), ms, &(x[WS(rs, 1)])); T3G = VADD(T3C, T3F); ST(&(x[WS(rs, 5)]), VSUB(T3G, T3J), ms, &(x[WS(rs, 1)])); ST(&(x[WS(rs, 20)]), VADD(T3J, T3G), ms, &(x[0])); } { V To, T2n, T2o, T2p, T2x, T2y, T2z, T2u, T2v, T2w, T2q, T2r, T2s, T29, T2i; V T2e, T2g, T1Y, T2j, T2b, T2c, T2B, T2C; To = VADD(Tk, Tn); T2n = VFMA(LDK(KP1_688655851), TN, VMUL(LDK(KP535826794), TK)); T2o = VFMA(LDK(KP1_541026485), T19, VMUL(LDK(KP637423989), T16)); T2p = VSUB(T2n, T2o); T2x = VFMA(LDK(KP851558583), T1y, VMUL(LDK(KP904827052), T1v)); T2y = VFMA(LDK(KP1_984229402), T1V, VMUL(LDK(KP125333233), T1S)); T2z = VADD(T2x, T2y); T2u = VFNMS(LDK(KP844327925), TK, VMUL(LDK(KP1_071653589), TN)); T2v = VFNMS(LDK(KP1_274847979), T19, VMUL(LDK(KP770513242), T16)); T2w = VADD(T2u, T2v); T2q = VFNMS(LDK(KP425779291), T1v, VMUL(LDK(KP1_809654104), T1y)); T2r = VFNMS(LDK(KP992114701), T1S, VMUL(LDK(KP250666467), T1V)); T2s = VADD(T2q, T2r); { V T23, T24, T25, T26, T27, T28; T23 = VFMA(LDK(KP1_937166322), TN, VMUL(LDK(KP248689887), TK)); T24 = VFMA(LDK(KP1_071653589), T19, VMUL(LDK(KP844327925), T16)); T25 = VADD(T23, T24); T26 = VFMA(LDK(KP1_752613360), T1y, VMUL(LDK(KP481753674), T1v)); T27 = VFMA(LDK(KP1_457937254), T1V, VMUL(LDK(KP684547105), T1S)); T28 = VADD(T26, T27); T29 = VADD(T25, T28); T2i = VSUB(T27, T26); T2e = VMUL(LDK(KP559016994), VSUB(T28, T25)); T2g = VSUB(T24, T23); } { V TO, T1a, T1b, T1z, T1W, T1X; TO = VFNMS(LDK(KP497379774), TN, VMUL(LDK(KP968583161), TK)); T1a = VFNMS(LDK(KP1_688655851), T19, VMUL(LDK(KP535826794), T16)); T1b = VADD(TO, T1a); T1z = VFNMS(LDK(KP963507348), T1y, VMUL(LDK(KP876306680), T1v)); T1W = VFNMS(LDK(KP1_369094211), T1V, VMUL(LDK(KP728968627), T1S)); T1X = VADD(T1z, T1W); T1Y = VADD(T1b, T1X); T2j = VMUL(LDK(KP559016994), VSUB(T1b, T1X)); T2b = VSUB(T1a, TO); T2c = VSUB(T1z, T1W); } { V T1Z, T2a, T2t, T2A; T1Z = VADD(To, T1Y); T2a = VBYI(VADD(T22, T29)); ST(&(x[WS(rs, 1)]), VSUB(T1Z, T2a), ms, &(x[WS(rs, 1)])); ST(&(x[WS(rs, 24)]), VADD(T1Z, T2a), ms, &(x[0])); T2t = VADD(To, VADD(T2p, T2s)); T2A = VBYI(VADD(T22, VSUB(T2w, T2z))); ST(&(x[WS(rs, 21)]), VSUB(T2t, T2A), ms, &(x[WS(rs, 1)])); ST(&(x[WS(rs, 4)]), VADD(T2t, T2A), ms, &(x[0])); } T2B = VBYI(VADD(T22, VFMA(LDK(KP309016994), T2w, VFMA(LDK(KP587785252), VSUB(T2r, T2q), VFNMS(LDK(KP951056516), VADD(T2n, T2o), VMUL(LDK(KP809016994), T2z)))))); T2C = VFMA(LDK(KP309016994), T2p, VFMA(LDK(KP951056516), VSUB(T2u, T2v), VFMA(LDK(KP587785252), VSUB(T2y, T2x), VFNMS(LDK(KP809016994), T2s, To)))); ST(&(x[WS(rs, 9)]), VADD(T2B, T2C), ms, &(x[WS(rs, 1)])); ST(&(x[WS(rs, 16)]), VSUB(T2C, T2B), ms, &(x[0])); { V T2f, T2l, T2k, T2m, T2d, T2h; T2d = VFMS(LDK(KP250000000), T29, T22); T2f = VBYI(VADD(VFMA(LDK(KP587785252), T2b, VMUL(LDK(KP951056516), T2c)), VSUB(T2d, T2e))); T2l = VBYI(VADD(VFNMS(LDK(KP587785252), T2c, VMUL(LDK(KP951056516), T2b)), VADD(T2d, T2e))); T2h = VFNMS(LDK(KP250000000), T1Y, To); T2k = VFMA(LDK(KP587785252), T2g, VFNMS(LDK(KP951056516), T2i, VSUB(T2h, T2j))); T2m = VFMA(LDK(KP951056516), T2g, VADD(T2j, VFMA(LDK(KP587785252), T2i, T2h))); ST(&(x[WS(rs, 11)]), VADD(T2f, T2k), ms, &(x[WS(rs, 1)])); ST(&(x[WS(rs, 19)]), VSUB(T2m, T2l), ms, &(x[WS(rs, 1)])); ST(&(x[WS(rs, 14)]), VSUB(T2k, T2f), ms, &(x[0])); ST(&(x[WS(rs, 6)]), VADD(T2l, T2m), ms, &(x[0])); } } { V T2S, T2U, T2F, T2I, T2J, T2Y, T2Z, T30, T2M, T2P, T2Q, T2V, T2W, T2X, T3a; V T3l, T3b, T3k, T3f, T3p, T3i, T3o, T32, T33; T2S = VFNMS(LDK(KP587785252), T20, T2R); T2U = VSUB(Tn, Tk); T2F = VFNMS(LDK(KP125333233), T2E, VMUL(LDK(KP1_984229402), T2D)); T2I = VFMA(LDK(KP1_457937254), T2G, VMUL(LDK(KP684547105), T2H)); T2J = VSUB(T2F, T2I); T2Y = VFNMS(LDK(KP1_996053456), T2N, VMUL(LDK(KP062790519), T2O)); T2Z = VFMA(LDK(KP1_541026485), T2K, VMUL(LDK(KP637423989), T2L)); T30 = VSUB(T2Y, T2Z); T2M = VFNMS(LDK(KP770513242), T2L, VMUL(LDK(KP1_274847979), T2K)); T2P = VFMA(LDK(KP125581039), T2N, VMUL(LDK(KP998026728), T2O)); T2Q = VSUB(T2M, T2P); T2V = VFNMS(LDK(KP1_369094211), T2G, VMUL(LDK(KP728968627), T2H)); T2W = VFMA(LDK(KP250666467), T2D, VMUL(LDK(KP992114701), T2E)); T2X = VSUB(T2V, T2W); { V T34, T35, T36, T37, T38, T39; T34 = VFNMS(LDK(KP481753674), T2H, VMUL(LDK(KP1_752613360), T2G)); T35 = VFMA(LDK(KP851558583), T2D, VMUL(LDK(KP904827052), T2E)); T36 = VSUB(T34, T35); T37 = VFNMS(LDK(KP844327925), T2O, VMUL(LDK(KP1_071653589), T2N)); T38 = VFNMS(LDK(KP998026728), T2L, VMUL(LDK(KP125581039), T2K)); T39 = VADD(T37, T38); T3a = VMUL(LDK(KP559016994), VSUB(T36, T39)); T3l = VSUB(T37, T38); T3b = VADD(T36, T39); T3k = VADD(T34, T35); } { V T3d, T3e, T3m, T3g, T3h, T3n; T3d = VFNMS(LDK(KP425779291), T2E, VMUL(LDK(KP1_809654104), T2D)); T3e = VFMA(LDK(KP963507348), T2G, VMUL(LDK(KP876306680), T2H)); T3m = VADD(T3e, T3d); T3g = VFMA(LDK(KP1_688655851), T2N, VMUL(LDK(KP535826794), T2O)); T3h = VFMA(LDK(KP1_996053456), T2K, VMUL(LDK(KP062790519), T2L)); T3n = VADD(T3g, T3h); T3f = VSUB(T3d, T3e); T3p = VADD(T3m, T3n); T3i = VSUB(T3g, T3h); T3o = VMUL(LDK(KP559016994), VSUB(T3m, T3n)); } { V T3u, T3v, T2T, T31; T3u = VBYI(VADD(T2S, T3b)); T3v = VADD(T2U, T3p); ST(&(x[WS(rs, 2)]), VADD(T3u, T3v), ms, &(x[0])); ST(&(x[WS(rs, 23)]), VSUB(T3v, T3u), ms, &(x[WS(rs, 1)])); T2T = VBYI(VSUB(VADD(T2J, T2Q), T2S)); T31 = VADD(T2U, VADD(T2X, T30)); ST(&(x[WS(rs, 3)]), VADD(T2T, T31), ms, &(x[WS(rs, 1)])); ST(&(x[WS(rs, 22)]), VSUB(T31, T2T), ms, &(x[0])); } T32 = VFMA(LDK(KP309016994), T2X, VFNMS(LDK(KP809016994), T30, VFNMS(LDK(KP587785252), VADD(T2P, T2M), VFNMS(LDK(KP951056516), VADD(T2I, T2F), T2U)))); T33 = VBYI(VSUB(VFNMS(LDK(KP587785252), VADD(T2Y, T2Z), VFNMS(LDK(KP809016994), T2Q, VFNMS(LDK(KP951056516), VADD(T2V, T2W), VMUL(LDK(KP309016994), T2J)))), T2S)); ST(&(x[WS(rs, 17)]), VSUB(T32, T33), ms, &(x[WS(rs, 1)])); ST(&(x[WS(rs, 8)]), VADD(T32, T33), ms, &(x[0])); { V T3j, T3s, T3r, T3t, T3c, T3q; T3c = VFNMS(LDK(KP250000000), T3b, T2S); T3j = VBYI(VADD(T3a, VADD(T3c, VFNMS(LDK(KP587785252), T3i, VMUL(LDK(KP951056516), T3f))))); T3s = VBYI(VADD(T3c, VSUB(VFMA(LDK(KP587785252), T3f, VMUL(LDK(KP951056516), T3i)), T3a))); T3q = VFNMS(LDK(KP250000000), T3p, T2U); T3r = VFMA(LDK(KP951056516), T3k, VFMA(LDK(KP587785252), T3l, VADD(T3o, T3q))); T3t = VFMA(LDK(KP587785252), T3k, VSUB(VFNMS(LDK(KP951056516), T3l, T3q), T3o)); ST(&(x[WS(rs, 7)]), VADD(T3j, T3r), ms, &(x[WS(rs, 1)])); ST(&(x[WS(rs, 13)]), VSUB(T3t, T3s), ms, &(x[WS(rs, 1)])); ST(&(x[WS(rs, 18)]), VSUB(T3r, T3j), ms, &(x[0])); ST(&(x[WS(rs, 12)]), VADD(T3s, T3t), ms, &(x[0])); } } } } } VLEAVE(); } static const tw_instr twinstr[] = { VTW(0, 1), VTW(0, 3), VTW(0, 9), VTW(0, 24), {TW_NEXT, VL, 0} }; static const ct_desc desc = { 25, XSIMD_STRING("t3fv_25"), twinstr, &GENUS, {190, 150, 78, 0}, 0, 0, 0 }; void XSIMD(codelet_t3fv_25) (planner *p) { X(kdft_dit_register) (p, t3fv_25, &desc); } #endif /* HAVE_FMA */ fftw-3.3.3/dft/simd/common/t1buv_2.c0000644000175000017500000000657212054410546014051 00000000000000/* * Copyright (c) 2003, 2007-11 Matteo Frigo * Copyright (c) 2003, 2007-11 Massachusetts Institute of Technology * * This program is free software; you can redistribute it and/or modify * it under the terms of the GNU General Public License as published by * the Free Software Foundation; either version 2 of the License, or * (at your option) any later version. * * This program is distributed in the hope that it will be useful, * but WITHOUT ANY WARRANTY; without even the implied warranty of * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the * GNU General Public License for more details. * * You should have received a copy of the GNU General Public License * along with this program; if not, write to the Free Software * Foundation, Inc., 51 Franklin Street, Fifth Floor, Boston, MA 02110-1301 USA * */ /* This file was automatically generated --- DO NOT EDIT */ /* Generated on Sun Nov 25 07:39:02 EST 2012 */ #include "codelet-dft.h" #ifdef HAVE_FMA /* Generated by: ../../../genfft/gen_twiddle_c.native -fma -reorder-insns -schedule-for-pipeline -simd -compact -variables 4 -pipeline-latency 8 -n 2 -name t1buv_2 -include t1bu.h -sign 1 */ /* * This function contains 3 FP additions, 2 FP multiplications, * (or, 3 additions, 2 multiplications, 0 fused multiply/add), * 5 stack variables, 0 constants, and 4 memory accesses */ #include "t1bu.h" static void t1buv_2(R *ri, R *ii, const R *W, stride rs, INT mb, INT me, INT ms) { { INT m; R *x; x = ii; for (m = mb, W = W + (mb * ((TWVL / VL) * 2)); m < me; m = m + VL, x = x + (VL * ms), W = W + (TWVL * 2), MAKE_VOLATILE_STRIDE(2, rs)) { V T1, T2, T3; T1 = LD(&(x[0]), ms, &(x[0])); T2 = LD(&(x[WS(rs, 1)]), ms, &(x[WS(rs, 1)])); T3 = BYTW(&(W[0]), T2); ST(&(x[0]), VADD(T1, T3), ms, &(x[0])); ST(&(x[WS(rs, 1)]), VSUB(T1, T3), ms, &(x[WS(rs, 1)])); } } VLEAVE(); } static const tw_instr twinstr[] = { VTW(0, 1), {TW_NEXT, VL, 0} }; static const ct_desc desc = { 2, XSIMD_STRING("t1buv_2"), twinstr, &GENUS, {3, 2, 0, 0}, 0, 0, 0 }; void XSIMD(codelet_t1buv_2) (planner *p) { X(kdft_dit_register) (p, t1buv_2, &desc); } #else /* HAVE_FMA */ /* Generated by: ../../../genfft/gen_twiddle_c.native -simd -compact -variables 4 -pipeline-latency 8 -n 2 -name t1buv_2 -include t1bu.h -sign 1 */ /* * This function contains 3 FP additions, 2 FP multiplications, * (or, 3 additions, 2 multiplications, 0 fused multiply/add), * 5 stack variables, 0 constants, and 4 memory accesses */ #include "t1bu.h" static void t1buv_2(R *ri, R *ii, const R *W, stride rs, INT mb, INT me, INT ms) { { INT m; R *x; x = ii; for (m = mb, W = W + (mb * ((TWVL / VL) * 2)); m < me; m = m + VL, x = x + (VL * ms), W = W + (TWVL * 2), MAKE_VOLATILE_STRIDE(2, rs)) { V T1, T3, T2; T1 = LD(&(x[0]), ms, &(x[0])); T2 = LD(&(x[WS(rs, 1)]), ms, &(x[WS(rs, 1)])); T3 = BYTW(&(W[0]), T2); ST(&(x[WS(rs, 1)]), VSUB(T1, T3), ms, &(x[WS(rs, 1)])); ST(&(x[0]), VADD(T1, T3), ms, &(x[0])); } } VLEAVE(); } static const tw_instr twinstr[] = { VTW(0, 1), {TW_NEXT, VL, 0} }; static const ct_desc desc = { 2, XSIMD_STRING("t1buv_2"), twinstr, &GENUS, {3, 2, 0, 0}, 0, 0, 0 }; void XSIMD(codelet_t1buv_2) (planner *p) { X(kdft_dit_register) (p, t1buv_2, &desc); } #endif /* HAVE_FMA */ fftw-3.3.3/dft/simd/common/t1buv_3.c0000644000175000017500000001042112054410546014036 00000000000000/* * Copyright (c) 2003, 2007-11 Matteo Frigo * Copyright (c) 2003, 2007-11 Massachusetts Institute of Technology * * This program is free software; you can redistribute it and/or modify * it under the terms of the GNU General Public License as published by * the Free Software Foundation; either version 2 of the License, or * (at your option) any later version. * * This program is distributed in the hope that it will be useful, * but WITHOUT ANY WARRANTY; without even the implied warranty of * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the * GNU General Public License for more details. * * You should have received a copy of the GNU General Public License * along with this program; if not, write to the Free Software * Foundation, Inc., 51 Franklin Street, Fifth Floor, Boston, MA 02110-1301 USA * */ /* This file was automatically generated --- DO NOT EDIT */ /* Generated on Sun Nov 25 07:39:02 EST 2012 */ #include "codelet-dft.h" #ifdef HAVE_FMA /* Generated by: ../../../genfft/gen_twiddle_c.native -fma -reorder-insns -schedule-for-pipeline -simd -compact -variables 4 -pipeline-latency 8 -n 3 -name t1buv_3 -include t1bu.h -sign 1 */ /* * This function contains 8 FP additions, 8 FP multiplications, * (or, 5 additions, 5 multiplications, 3 fused multiply/add), * 12 stack variables, 2 constants, and 6 memory accesses */ #include "t1bu.h" static void t1buv_3(R *ri, R *ii, const R *W, stride rs, INT mb, INT me, INT ms) { DVK(KP500000000, +0.500000000000000000000000000000000000000000000); DVK(KP866025403, +0.866025403784438646763723170752936183471402627); { INT m; R *x; x = ii; for (m = mb, W = W + (mb * ((TWVL / VL) * 4)); m < me; m = m + VL, x = x + (VL * ms), W = W + (TWVL * 4), MAKE_VOLATILE_STRIDE(3, rs)) { V T1, T2, T4; T1 = LD(&(x[0]), ms, &(x[0])); T2 = LD(&(x[WS(rs, 1)]), ms, &(x[WS(rs, 1)])); T4 = LD(&(x[WS(rs, 2)]), ms, &(x[0])); { V T3, T5, T8, T6, T7; T3 = BYTW(&(W[0]), T2); T5 = BYTW(&(W[TWVL * 2]), T4); T8 = VMUL(LDK(KP866025403), VSUB(T3, T5)); T6 = VADD(T3, T5); T7 = VFNMS(LDK(KP500000000), T6, T1); ST(&(x[0]), VADD(T1, T6), ms, &(x[0])); ST(&(x[WS(rs, 2)]), VFNMSI(T8, T7), ms, &(x[0])); ST(&(x[WS(rs, 1)]), VFMAI(T8, T7), ms, &(x[WS(rs, 1)])); } } } VLEAVE(); } static const tw_instr twinstr[] = { VTW(0, 1), VTW(0, 2), {TW_NEXT, VL, 0} }; static const ct_desc desc = { 3, XSIMD_STRING("t1buv_3"), twinstr, &GENUS, {5, 5, 3, 0}, 0, 0, 0 }; void XSIMD(codelet_t1buv_3) (planner *p) { X(kdft_dit_register) (p, t1buv_3, &desc); } #else /* HAVE_FMA */ /* Generated by: ../../../genfft/gen_twiddle_c.native -simd -compact -variables 4 -pipeline-latency 8 -n 3 -name t1buv_3 -include t1bu.h -sign 1 */ /* * This function contains 8 FP additions, 6 FP multiplications, * (or, 7 additions, 5 multiplications, 1 fused multiply/add), * 12 stack variables, 2 constants, and 6 memory accesses */ #include "t1bu.h" static void t1buv_3(R *ri, R *ii, const R *W, stride rs, INT mb, INT me, INT ms) { DVK(KP500000000, +0.500000000000000000000000000000000000000000000); DVK(KP866025403, +0.866025403784438646763723170752936183471402627); { INT m; R *x; x = ii; for (m = mb, W = W + (mb * ((TWVL / VL) * 4)); m < me; m = m + VL, x = x + (VL * ms), W = W + (TWVL * 4), MAKE_VOLATILE_STRIDE(3, rs)) { V T6, T2, T4, T7, T1, T3, T5, T8; T6 = LD(&(x[0]), ms, &(x[0])); T1 = LD(&(x[WS(rs, 1)]), ms, &(x[WS(rs, 1)])); T2 = BYTW(&(W[0]), T1); T3 = LD(&(x[WS(rs, 2)]), ms, &(x[0])); T4 = BYTW(&(W[TWVL * 2]), T3); T7 = VADD(T2, T4); ST(&(x[0]), VADD(T6, T7), ms, &(x[0])); T5 = VBYI(VMUL(LDK(KP866025403), VSUB(T2, T4))); T8 = VFNMS(LDK(KP500000000), T7, T6); ST(&(x[WS(rs, 1)]), VADD(T5, T8), ms, &(x[WS(rs, 1)])); ST(&(x[WS(rs, 2)]), VSUB(T8, T5), ms, &(x[0])); } } VLEAVE(); } static const tw_instr twinstr[] = { VTW(0, 1), VTW(0, 2), {TW_NEXT, VL, 0} }; static const ct_desc desc = { 3, XSIMD_STRING("t1buv_3"), twinstr, &GENUS, {7, 5, 1, 0}, 0, 0, 0 }; void XSIMD(codelet_t1buv_3) (planner *p) { X(kdft_dit_register) (p, t1buv_3, &desc); } #endif /* HAVE_FMA */ fftw-3.3.3/dft/simd/common/t1buv_4.c0000644000175000017500000001052112054410546014040 00000000000000/* * Copyright (c) 2003, 2007-11 Matteo Frigo * Copyright (c) 2003, 2007-11 Massachusetts Institute of Technology * * This program is free software; you can redistribute it and/or modify * it under the terms of the GNU General Public License as published by * the Free Software Foundation; either version 2 of the License, or * (at your option) any later version. * * This program is distributed in the hope that it will be useful, * but WITHOUT ANY WARRANTY; without even the implied warranty of * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the * GNU General Public License for more details. * * You should have received a copy of the GNU General Public License * along with this program; if not, write to the Free Software * Foundation, Inc., 51 Franklin Street, Fifth Floor, Boston, MA 02110-1301 USA * */ /* This file was automatically generated --- DO NOT EDIT */ /* Generated on Sun Nov 25 07:39:02 EST 2012 */ #include "codelet-dft.h" #ifdef HAVE_FMA /* Generated by: ../../../genfft/gen_twiddle_c.native -fma -reorder-insns -schedule-for-pipeline -simd -compact -variables 4 -pipeline-latency 8 -n 4 -name t1buv_4 -include t1bu.h -sign 1 */ /* * This function contains 11 FP additions, 8 FP multiplications, * (or, 9 additions, 6 multiplications, 2 fused multiply/add), * 13 stack variables, 0 constants, and 8 memory accesses */ #include "t1bu.h" static void t1buv_4(R *ri, R *ii, const R *W, stride rs, INT mb, INT me, INT ms) { { INT m; R *x; x = ii; for (m = mb, W = W + (mb * ((TWVL / VL) * 6)); m < me; m = m + VL, x = x + (VL * ms), W = W + (TWVL * 6), MAKE_VOLATILE_STRIDE(4, rs)) { V T1, T7, T2, T5, T8, T3, T6; T1 = LD(&(x[0]), ms, &(x[0])); T7 = LD(&(x[WS(rs, 3)]), ms, &(x[WS(rs, 1)])); T2 = LD(&(x[WS(rs, 2)]), ms, &(x[0])); T5 = LD(&(x[WS(rs, 1)]), ms, &(x[WS(rs, 1)])); T8 = BYTW(&(W[TWVL * 4]), T7); T3 = BYTW(&(W[TWVL * 2]), T2); T6 = BYTW(&(W[0]), T5); { V Ta, T4, Tb, T9; Ta = VADD(T1, T3); T4 = VSUB(T1, T3); Tb = VADD(T6, T8); T9 = VSUB(T6, T8); ST(&(x[0]), VADD(Ta, Tb), ms, &(x[0])); ST(&(x[WS(rs, 2)]), VSUB(Ta, Tb), ms, &(x[0])); ST(&(x[WS(rs, 1)]), VFMAI(T9, T4), ms, &(x[WS(rs, 1)])); ST(&(x[WS(rs, 3)]), VFNMSI(T9, T4), ms, &(x[WS(rs, 1)])); } } } VLEAVE(); } static const tw_instr twinstr[] = { VTW(0, 1), VTW(0, 2), VTW(0, 3), {TW_NEXT, VL, 0} }; static const ct_desc desc = { 4, XSIMD_STRING("t1buv_4"), twinstr, &GENUS, {9, 6, 2, 0}, 0, 0, 0 }; void XSIMD(codelet_t1buv_4) (planner *p) { X(kdft_dit_register) (p, t1buv_4, &desc); } #else /* HAVE_FMA */ /* Generated by: ../../../genfft/gen_twiddle_c.native -simd -compact -variables 4 -pipeline-latency 8 -n 4 -name t1buv_4 -include t1bu.h -sign 1 */ /* * This function contains 11 FP additions, 6 FP multiplications, * (or, 11 additions, 6 multiplications, 0 fused multiply/add), * 13 stack variables, 0 constants, and 8 memory accesses */ #include "t1bu.h" static void t1buv_4(R *ri, R *ii, const R *W, stride rs, INT mb, INT me, INT ms) { { INT m; R *x; x = ii; for (m = mb, W = W + (mb * ((TWVL / VL) * 6)); m < me; m = m + VL, x = x + (VL * ms), W = W + (TWVL * 6), MAKE_VOLATILE_STRIDE(4, rs)) { V T1, T8, T3, T6, T7, T2, T5; T1 = LD(&(x[0]), ms, &(x[0])); T7 = LD(&(x[WS(rs, 3)]), ms, &(x[WS(rs, 1)])); T8 = BYTW(&(W[TWVL * 4]), T7); T2 = LD(&(x[WS(rs, 2)]), ms, &(x[0])); T3 = BYTW(&(W[TWVL * 2]), T2); T5 = LD(&(x[WS(rs, 1)]), ms, &(x[WS(rs, 1)])); T6 = BYTW(&(W[0]), T5); { V T4, T9, Ta, Tb; T4 = VSUB(T1, T3); T9 = VBYI(VSUB(T6, T8)); ST(&(x[WS(rs, 3)]), VSUB(T4, T9), ms, &(x[WS(rs, 1)])); ST(&(x[WS(rs, 1)]), VADD(T4, T9), ms, &(x[WS(rs, 1)])); Ta = VADD(T1, T3); Tb = VADD(T6, T8); ST(&(x[WS(rs, 2)]), VSUB(Ta, Tb), ms, &(x[0])); ST(&(x[0]), VADD(Ta, Tb), ms, &(x[0])); } } } VLEAVE(); } static const tw_instr twinstr[] = { VTW(0, 1), VTW(0, 2), VTW(0, 3), {TW_NEXT, VL, 0} }; static const ct_desc desc = { 4, XSIMD_STRING("t1buv_4"), twinstr, &GENUS, {11, 6, 0, 0}, 0, 0, 0 }; void XSIMD(codelet_t1buv_4) (planner *p) { X(kdft_dit_register) (p, t1buv_4, &desc); } #endif /* HAVE_FMA */ fftw-3.3.3/dft/simd/common/t1buv_5.c0000644000175000017500000001375412054410546014054 00000000000000/* * Copyright (c) 2003, 2007-11 Matteo Frigo * Copyright (c) 2003, 2007-11 Massachusetts Institute of Technology * * This program is free software; you can redistribute it and/or modify * it under the terms of the GNU General Public License as published by * the Free Software Foundation; either version 2 of the License, or * (at your option) any later version. * * This program is distributed in the hope that it will be useful, * but WITHOUT ANY WARRANTY; without even the implied warranty of * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the * GNU General Public License for more details. * * You should have received a copy of the GNU General Public License * along with this program; if not, write to the Free Software * Foundation, Inc., 51 Franklin Street, Fifth Floor, Boston, MA 02110-1301 USA * */ /* This file was automatically generated --- DO NOT EDIT */ /* Generated on Sun Nov 25 07:39:02 EST 2012 */ #include "codelet-dft.h" #ifdef HAVE_FMA /* Generated by: ../../../genfft/gen_twiddle_c.native -fma -reorder-insns -schedule-for-pipeline -simd -compact -variables 4 -pipeline-latency 8 -n 5 -name t1buv_5 -include t1bu.h -sign 1 */ /* * This function contains 20 FP additions, 19 FP multiplications, * (or, 11 additions, 10 multiplications, 9 fused multiply/add), * 26 stack variables, 4 constants, and 10 memory accesses */ #include "t1bu.h" static void t1buv_5(R *ri, R *ii, const R *W, stride rs, INT mb, INT me, INT ms) { DVK(KP559016994, +0.559016994374947424102293417182819058860154590); DVK(KP250000000, +0.250000000000000000000000000000000000000000000); DVK(KP618033988, +0.618033988749894848204586834365638117720309180); DVK(KP951056516, +0.951056516295153572116439333379382143405698634); { INT m; R *x; x = ii; for (m = mb, W = W + (mb * ((TWVL / VL) * 8)); m < me; m = m + VL, x = x + (VL * ms), W = W + (TWVL * 8), MAKE_VOLATILE_STRIDE(5, rs)) { V T1, T2, T9, T4, T7; T1 = LD(&(x[0]), ms, &(x[0])); T2 = LD(&(x[WS(rs, 1)]), ms, &(x[WS(rs, 1)])); T9 = LD(&(x[WS(rs, 3)]), ms, &(x[WS(rs, 1)])); T4 = LD(&(x[WS(rs, 4)]), ms, &(x[0])); T7 = LD(&(x[WS(rs, 2)]), ms, &(x[0])); { V T3, Ta, T5, T8; T3 = BYTW(&(W[0]), T2); Ta = BYTW(&(W[TWVL * 4]), T9); T5 = BYTW(&(W[TWVL * 6]), T4); T8 = BYTW(&(W[TWVL * 2]), T7); { V T6, Tg, Tb, Th; T6 = VADD(T3, T5); Tg = VSUB(T3, T5); Tb = VADD(T8, Ta); Th = VSUB(T8, Ta); { V Te, Tc, Tk, Ti, Td, Tj, Tf; Te = VSUB(T6, Tb); Tc = VADD(T6, Tb); Tk = VMUL(LDK(KP951056516), VFNMS(LDK(KP618033988), Tg, Th)); Ti = VMUL(LDK(KP951056516), VFMA(LDK(KP618033988), Th, Tg)); Td = VFNMS(LDK(KP250000000), Tc, T1); ST(&(x[0]), VADD(T1, Tc), ms, &(x[0])); Tj = VFNMS(LDK(KP559016994), Te, Td); Tf = VFMA(LDK(KP559016994), Te, Td); ST(&(x[WS(rs, 2)]), VFNMSI(Tk, Tj), ms, &(x[0])); ST(&(x[WS(rs, 3)]), VFMAI(Tk, Tj), ms, &(x[WS(rs, 1)])); ST(&(x[WS(rs, 4)]), VFNMSI(Ti, Tf), ms, &(x[0])); ST(&(x[WS(rs, 1)]), VFMAI(Ti, Tf), ms, &(x[WS(rs, 1)])); } } } } } VLEAVE(); } static const tw_instr twinstr[] = { VTW(0, 1), VTW(0, 2), VTW(0, 3), VTW(0, 4), {TW_NEXT, VL, 0} }; static const ct_desc desc = { 5, XSIMD_STRING("t1buv_5"), twinstr, &GENUS, {11, 10, 9, 0}, 0, 0, 0 }; void XSIMD(codelet_t1buv_5) (planner *p) { X(kdft_dit_register) (p, t1buv_5, &desc); } #else /* HAVE_FMA */ /* Generated by: ../../../genfft/gen_twiddle_c.native -simd -compact -variables 4 -pipeline-latency 8 -n 5 -name t1buv_5 -include t1bu.h -sign 1 */ /* * This function contains 20 FP additions, 14 FP multiplications, * (or, 17 additions, 11 multiplications, 3 fused multiply/add), * 20 stack variables, 4 constants, and 10 memory accesses */ #include "t1bu.h" static void t1buv_5(R *ri, R *ii, const R *W, stride rs, INT mb, INT me, INT ms) { DVK(KP250000000, +0.250000000000000000000000000000000000000000000); DVK(KP559016994, +0.559016994374947424102293417182819058860154590); DVK(KP587785252, +0.587785252292473129168705954639072768597652438); DVK(KP951056516, +0.951056516295153572116439333379382143405698634); { INT m; R *x; x = ii; for (m = mb, W = W + (mb * ((TWVL / VL) * 8)); m < me; m = m + VL, x = x + (VL * ms), W = W + (TWVL * 8), MAKE_VOLATILE_STRIDE(5, rs)) { V Tf, T5, Ta, Tc, Td, Tg; Tf = LD(&(x[0]), ms, &(x[0])); { V T2, T9, T4, T7; { V T1, T8, T3, T6; T1 = LD(&(x[WS(rs, 1)]), ms, &(x[WS(rs, 1)])); T2 = BYTW(&(W[0]), T1); T8 = LD(&(x[WS(rs, 3)]), ms, &(x[WS(rs, 1)])); T9 = BYTW(&(W[TWVL * 4]), T8); T3 = LD(&(x[WS(rs, 4)]), ms, &(x[0])); T4 = BYTW(&(W[TWVL * 6]), T3); T6 = LD(&(x[WS(rs, 2)]), ms, &(x[0])); T7 = BYTW(&(W[TWVL * 2]), T6); } T5 = VSUB(T2, T4); Ta = VSUB(T7, T9); Tc = VADD(T2, T4); Td = VADD(T7, T9); Tg = VADD(Tc, Td); } ST(&(x[0]), VADD(Tf, Tg), ms, &(x[0])); { V Tb, Tj, Ti, Tk, Te, Th; Tb = VBYI(VFMA(LDK(KP951056516), T5, VMUL(LDK(KP587785252), Ta))); Tj = VBYI(VFNMS(LDK(KP951056516), Ta, VMUL(LDK(KP587785252), T5))); Te = VMUL(LDK(KP559016994), VSUB(Tc, Td)); Th = VFNMS(LDK(KP250000000), Tg, Tf); Ti = VADD(Te, Th); Tk = VSUB(Th, Te); ST(&(x[WS(rs, 1)]), VADD(Tb, Ti), ms, &(x[WS(rs, 1)])); ST(&(x[WS(rs, 3)]), VSUB(Tk, Tj), ms, &(x[WS(rs, 1)])); ST(&(x[WS(rs, 4)]), VSUB(Ti, Tb), ms, &(x[0])); ST(&(x[WS(rs, 2)]), VADD(Tj, Tk), ms, &(x[0])); } } } VLEAVE(); } static const tw_instr twinstr[] = { VTW(0, 1), VTW(0, 2), VTW(0, 3), VTW(0, 4), {TW_NEXT, VL, 0} }; static const ct_desc desc = { 5, XSIMD_STRING("t1buv_5"), twinstr, &GENUS, {17, 11, 3, 0}, 0, 0, 0 }; void XSIMD(codelet_t1buv_5) (planner *p) { X(kdft_dit_register) (p, t1buv_5, &desc); } #endif /* HAVE_FMA */ fftw-3.3.3/dft/simd/common/t1buv_6.c0000644000175000017500000001403712054410546014050 00000000000000/* * Copyright (c) 2003, 2007-11 Matteo Frigo * Copyright (c) 2003, 2007-11 Massachusetts Institute of Technology * * This program is free software; you can redistribute it and/or modify * it under the terms of the GNU General Public License as published by * the Free Software Foundation; either version 2 of the License, or * (at your option) any later version. * * This program is distributed in the hope that it will be useful, * but WITHOUT ANY WARRANTY; without even the implied warranty of * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the * GNU General Public License for more details. * * You should have received a copy of the GNU General Public License * along with this program; if not, write to the Free Software * Foundation, Inc., 51 Franklin Street, Fifth Floor, Boston, MA 02110-1301 USA * */ /* This file was automatically generated --- DO NOT EDIT */ /* Generated on Sun Nov 25 07:39:02 EST 2012 */ #include "codelet-dft.h" #ifdef HAVE_FMA /* Generated by: ../../../genfft/gen_twiddle_c.native -fma -reorder-insns -schedule-for-pipeline -simd -compact -variables 4 -pipeline-latency 8 -n 6 -name t1buv_6 -include t1bu.h -sign 1 */ /* * This function contains 23 FP additions, 18 FP multiplications, * (or, 17 additions, 12 multiplications, 6 fused multiply/add), * 27 stack variables, 2 constants, and 12 memory accesses */ #include "t1bu.h" static void t1buv_6(R *ri, R *ii, const R *W, stride rs, INT mb, INT me, INT ms) { DVK(KP500000000, +0.500000000000000000000000000000000000000000000); DVK(KP866025403, +0.866025403784438646763723170752936183471402627); { INT m; R *x; x = ii; for (m = mb, W = W + (mb * ((TWVL / VL) * 10)); m < me; m = m + VL, x = x + (VL * ms), W = W + (TWVL * 10), MAKE_VOLATILE_STRIDE(6, rs)) { V T1, T2, Ta, Tc, T5, T7; T1 = LD(&(x[0]), ms, &(x[0])); T2 = LD(&(x[WS(rs, 3)]), ms, &(x[WS(rs, 1)])); Ta = LD(&(x[WS(rs, 4)]), ms, &(x[0])); Tc = LD(&(x[WS(rs, 1)]), ms, &(x[WS(rs, 1)])); T5 = LD(&(x[WS(rs, 2)]), ms, &(x[0])); T7 = LD(&(x[WS(rs, 5)]), ms, &(x[WS(rs, 1)])); { V T3, Tb, Td, T6, T8; T3 = BYTW(&(W[TWVL * 4]), T2); Tb = BYTW(&(W[TWVL * 6]), Ta); Td = BYTW(&(W[0]), Tc); T6 = BYTW(&(W[TWVL * 2]), T5); T8 = BYTW(&(W[TWVL * 8]), T7); { V Ti, T4, Tk, Te, Tj, T9; Ti = VADD(T1, T3); T4 = VSUB(T1, T3); Tk = VADD(Tb, Td); Te = VSUB(Tb, Td); Tj = VADD(T6, T8); T9 = VSUB(T6, T8); { V Tl, Tn, Tf, Th, Tm, Tg; Tl = VADD(Tj, Tk); Tn = VMUL(LDK(KP866025403), VSUB(Tj, Tk)); Tf = VADD(T9, Te); Th = VMUL(LDK(KP866025403), VSUB(T9, Te)); ST(&(x[0]), VADD(Ti, Tl), ms, &(x[0])); Tm = VFNMS(LDK(KP500000000), Tl, Ti); ST(&(x[WS(rs, 3)]), VADD(T4, Tf), ms, &(x[WS(rs, 1)])); Tg = VFNMS(LDK(KP500000000), Tf, T4); ST(&(x[WS(rs, 4)]), VFMAI(Tn, Tm), ms, &(x[0])); ST(&(x[WS(rs, 2)]), VFNMSI(Tn, Tm), ms, &(x[0])); ST(&(x[WS(rs, 5)]), VFNMSI(Th, Tg), ms, &(x[WS(rs, 1)])); ST(&(x[WS(rs, 1)]), VFMAI(Th, Tg), ms, &(x[WS(rs, 1)])); } } } } } VLEAVE(); } static const tw_instr twinstr[] = { VTW(0, 1), VTW(0, 2), VTW(0, 3), VTW(0, 4), VTW(0, 5), {TW_NEXT, VL, 0} }; static const ct_desc desc = { 6, XSIMD_STRING("t1buv_6"), twinstr, &GENUS, {17, 12, 6, 0}, 0, 0, 0 }; void XSIMD(codelet_t1buv_6) (planner *p) { X(kdft_dit_register) (p, t1buv_6, &desc); } #else /* HAVE_FMA */ /* Generated by: ../../../genfft/gen_twiddle_c.native -simd -compact -variables 4 -pipeline-latency 8 -n 6 -name t1buv_6 -include t1bu.h -sign 1 */ /* * This function contains 23 FP additions, 14 FP multiplications, * (or, 21 additions, 12 multiplications, 2 fused multiply/add), * 19 stack variables, 2 constants, and 12 memory accesses */ #include "t1bu.h" static void t1buv_6(R *ri, R *ii, const R *W, stride rs, INT mb, INT me, INT ms) { DVK(KP500000000, +0.500000000000000000000000000000000000000000000); DVK(KP866025403, +0.866025403784438646763723170752936183471402627); { INT m; R *x; x = ii; for (m = mb, W = W + (mb * ((TWVL / VL) * 10)); m < me; m = m + VL, x = x + (VL * ms), W = W + (TWVL * 10), MAKE_VOLATILE_STRIDE(6, rs)) { V Tf, Ti, Ta, Tk, T5, Tj, Tc, Te, Td; Tc = LD(&(x[0]), ms, &(x[0])); Td = LD(&(x[WS(rs, 3)]), ms, &(x[WS(rs, 1)])); Te = BYTW(&(W[TWVL * 4]), Td); Tf = VSUB(Tc, Te); Ti = VADD(Tc, Te); { V T7, T9, T6, T8; T6 = LD(&(x[WS(rs, 4)]), ms, &(x[0])); T7 = BYTW(&(W[TWVL * 6]), T6); T8 = LD(&(x[WS(rs, 1)]), ms, &(x[WS(rs, 1)])); T9 = BYTW(&(W[0]), T8); Ta = VSUB(T7, T9); Tk = VADD(T7, T9); } { V T2, T4, T1, T3; T1 = LD(&(x[WS(rs, 2)]), ms, &(x[0])); T2 = BYTW(&(W[TWVL * 2]), T1); T3 = LD(&(x[WS(rs, 5)]), ms, &(x[WS(rs, 1)])); T4 = BYTW(&(W[TWVL * 8]), T3); T5 = VSUB(T2, T4); Tj = VADD(T2, T4); } { V Tb, Tg, Th, Tn, Tl, Tm; Tb = VBYI(VMUL(LDK(KP866025403), VSUB(T5, Ta))); Tg = VADD(T5, Ta); Th = VFNMS(LDK(KP500000000), Tg, Tf); ST(&(x[WS(rs, 1)]), VADD(Tb, Th), ms, &(x[WS(rs, 1)])); ST(&(x[WS(rs, 3)]), VADD(Tf, Tg), ms, &(x[WS(rs, 1)])); ST(&(x[WS(rs, 5)]), VSUB(Th, Tb), ms, &(x[WS(rs, 1)])); Tn = VBYI(VMUL(LDK(KP866025403), VSUB(Tj, Tk))); Tl = VADD(Tj, Tk); Tm = VFNMS(LDK(KP500000000), Tl, Ti); ST(&(x[WS(rs, 2)]), VSUB(Tm, Tn), ms, &(x[0])); ST(&(x[0]), VADD(Ti, Tl), ms, &(x[0])); ST(&(x[WS(rs, 4)]), VADD(Tn, Tm), ms, &(x[0])); } } } VLEAVE(); } static const tw_instr twinstr[] = { VTW(0, 1), VTW(0, 2), VTW(0, 3), VTW(0, 4), VTW(0, 5), {TW_NEXT, VL, 0} }; static const ct_desc desc = { 6, XSIMD_STRING("t1buv_6"), twinstr, &GENUS, {21, 12, 2, 0}, 0, 0, 0 }; void XSIMD(codelet_t1buv_6) (planner *p) { X(kdft_dit_register) (p, t1buv_6, &desc); } #endif /* HAVE_FMA */ fftw-3.3.3/dft/simd/common/t1buv_7.c0000644000175000017500000001762512054410547014060 00000000000000/* * Copyright (c) 2003, 2007-11 Matteo Frigo * Copyright (c) 2003, 2007-11 Massachusetts Institute of Technology * * This program is free software; you can redistribute it and/or modify * it under the terms of the GNU General Public License as published by * the Free Software Foundation; either version 2 of the License, or * (at your option) any later version. * * This program is distributed in the hope that it will be useful, * but WITHOUT ANY WARRANTY; without even the implied warranty of * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the * GNU General Public License for more details. * * You should have received a copy of the GNU General Public License * along with this program; if not, write to the Free Software * Foundation, Inc., 51 Franklin Street, Fifth Floor, Boston, MA 02110-1301 USA * */ /* This file was automatically generated --- DO NOT EDIT */ /* Generated on Sun Nov 25 07:39:02 EST 2012 */ #include "codelet-dft.h" #ifdef HAVE_FMA /* Generated by: ../../../genfft/gen_twiddle_c.native -fma -reorder-insns -schedule-for-pipeline -simd -compact -variables 4 -pipeline-latency 8 -n 7 -name t1buv_7 -include t1bu.h -sign 1 */ /* * This function contains 36 FP additions, 36 FP multiplications, * (or, 15 additions, 15 multiplications, 21 fused multiply/add), * 42 stack variables, 6 constants, and 14 memory accesses */ #include "t1bu.h" static void t1buv_7(R *ri, R *ii, const R *W, stride rs, INT mb, INT me, INT ms) { DVK(KP900968867, +0.900968867902419126236102319507445051165919162); DVK(KP801937735, +0.801937735804838252472204639014890102331838324); DVK(KP974927912, +0.974927912181823607018131682993931217232785801); DVK(KP692021471, +0.692021471630095869627814897002069140197260599); DVK(KP554958132, +0.554958132087371191422194871006410481067288862); DVK(KP356895867, +0.356895867892209443894399510021300583399127187); { INT m; R *x; x = ii; for (m = mb, W = W + (mb * ((TWVL / VL) * 12)); m < me; m = m + VL, x = x + (VL * ms), W = W + (TWVL * 12), MAKE_VOLATILE_STRIDE(7, rs)) { V T1, T2, T4, Te, Tc, T9, T7; T1 = LD(&(x[0]), ms, &(x[0])); T2 = LD(&(x[WS(rs, 1)]), ms, &(x[WS(rs, 1)])); T4 = LD(&(x[WS(rs, 6)]), ms, &(x[0])); Te = LD(&(x[WS(rs, 4)]), ms, &(x[0])); Tc = LD(&(x[WS(rs, 3)]), ms, &(x[WS(rs, 1)])); T9 = LD(&(x[WS(rs, 5)]), ms, &(x[WS(rs, 1)])); T7 = LD(&(x[WS(rs, 2)]), ms, &(x[0])); { V T3, T5, Tf, Td, Ta, T8; T3 = BYTW(&(W[0]), T2); T5 = BYTW(&(W[TWVL * 10]), T4); Tf = BYTW(&(W[TWVL * 6]), Te); Td = BYTW(&(W[TWVL * 4]), Tc); Ta = BYTW(&(W[TWVL * 8]), T9); T8 = BYTW(&(W[TWVL * 2]), T7); { V T6, Tm, Tg, Tk, Tb, Tl; T6 = VADD(T3, T5); Tm = VSUB(T3, T5); Tg = VADD(Td, Tf); Tk = VSUB(Td, Tf); Tb = VADD(T8, Ta); Tl = VSUB(T8, Ta); { V Tp, Tx, Tu, Th, Ts, Tn, Tq, Ty; Tp = VFNMS(LDK(KP356895867), T6, Tg); Tx = VFMA(LDK(KP554958132), Tk, Tm); ST(&(x[0]), VADD(T1, VADD(T6, VADD(Tb, Tg))), ms, &(x[0])); Tu = VFNMS(LDK(KP356895867), Tb, T6); Th = VFNMS(LDK(KP356895867), Tg, Tb); Ts = VFMA(LDK(KP554958132), Tl, Tk); Tn = VFNMS(LDK(KP554958132), Tm, Tl); Tq = VFNMS(LDK(KP692021471), Tp, Tb); Ty = VMUL(LDK(KP974927912), VFMA(LDK(KP801937735), Tx, Tl)); { V Tv, Ti, Tt, To, Tr, Tw, Tj; Tv = VFNMS(LDK(KP692021471), Tu, Tg); Ti = VFNMS(LDK(KP692021471), Th, T6); Tt = VMUL(LDK(KP974927912), VFNMS(LDK(KP801937735), Ts, Tm)); To = VMUL(LDK(KP974927912), VFNMS(LDK(KP801937735), Tn, Tk)); Tr = VFNMS(LDK(KP900968867), Tq, T1); Tw = VFNMS(LDK(KP900968867), Tv, T1); Tj = VFNMS(LDK(KP900968867), Ti, T1); ST(&(x[WS(rs, 5)]), VFNMSI(Tt, Tr), ms, &(x[WS(rs, 1)])); ST(&(x[WS(rs, 2)]), VFMAI(Tt, Tr), ms, &(x[0])); ST(&(x[WS(rs, 6)]), VFNMSI(Ty, Tw), ms, &(x[0])); ST(&(x[WS(rs, 1)]), VFMAI(Ty, Tw), ms, &(x[WS(rs, 1)])); ST(&(x[WS(rs, 4)]), VFNMSI(To, Tj), ms, &(x[0])); ST(&(x[WS(rs, 3)]), VFMAI(To, Tj), ms, &(x[WS(rs, 1)])); } } } } } } VLEAVE(); } static const tw_instr twinstr[] = { VTW(0, 1), VTW(0, 2), VTW(0, 3), VTW(0, 4), VTW(0, 5), VTW(0, 6), {TW_NEXT, VL, 0} }; static const ct_desc desc = { 7, XSIMD_STRING("t1buv_7"), twinstr, &GENUS, {15, 15, 21, 0}, 0, 0, 0 }; void XSIMD(codelet_t1buv_7) (planner *p) { X(kdft_dit_register) (p, t1buv_7, &desc); } #else /* HAVE_FMA */ /* Generated by: ../../../genfft/gen_twiddle_c.native -simd -compact -variables 4 -pipeline-latency 8 -n 7 -name t1buv_7 -include t1bu.h -sign 1 */ /* * This function contains 36 FP additions, 30 FP multiplications, * (or, 24 additions, 18 multiplications, 12 fused multiply/add), * 21 stack variables, 6 constants, and 14 memory accesses */ #include "t1bu.h" static void t1buv_7(R *ri, R *ii, const R *W, stride rs, INT mb, INT me, INT ms) { DVK(KP222520933, +0.222520933956314404288902564496794759466355569); DVK(KP900968867, +0.900968867902419126236102319507445051165919162); DVK(KP623489801, +0.623489801858733530525004884004239810632274731); DVK(KP433883739, +0.433883739117558120475768332848358754609990728); DVK(KP781831482, +0.781831482468029808708444526674057750232334519); DVK(KP974927912, +0.974927912181823607018131682993931217232785801); { INT m; R *x; x = ii; for (m = mb, W = W + (mb * ((TWVL / VL) * 12)); m < me; m = m + VL, x = x + (VL * ms), W = W + (TWVL * 12), MAKE_VOLATILE_STRIDE(7, rs)) { V Th, Tf, Ti, T5, Tk, Ta, Tj, To, Tp; Th = LD(&(x[0]), ms, &(x[0])); { V Tc, Te, Tb, Td; Tb = LD(&(x[WS(rs, 2)]), ms, &(x[0])); Tc = BYTW(&(W[TWVL * 2]), Tb); Td = LD(&(x[WS(rs, 5)]), ms, &(x[WS(rs, 1)])); Te = BYTW(&(W[TWVL * 8]), Td); Tf = VSUB(Tc, Te); Ti = VADD(Tc, Te); } { V T2, T4, T1, T3; T1 = LD(&(x[WS(rs, 1)]), ms, &(x[WS(rs, 1)])); T2 = BYTW(&(W[0]), T1); T3 = LD(&(x[WS(rs, 6)]), ms, &(x[0])); T4 = BYTW(&(W[TWVL * 10]), T3); T5 = VSUB(T2, T4); Tk = VADD(T2, T4); } { V T7, T9, T6, T8; T6 = LD(&(x[WS(rs, 3)]), ms, &(x[WS(rs, 1)])); T7 = BYTW(&(W[TWVL * 4]), T6); T8 = LD(&(x[WS(rs, 4)]), ms, &(x[0])); T9 = BYTW(&(W[TWVL * 6]), T8); Ta = VSUB(T7, T9); Tj = VADD(T7, T9); } ST(&(x[0]), VADD(Th, VADD(Tk, VADD(Ti, Tj))), ms, &(x[0])); To = VBYI(VFNMS(LDK(KP781831482), Ta, VFNMS(LDK(KP433883739), Tf, VMUL(LDK(KP974927912), T5)))); Tp = VFMA(LDK(KP623489801), Tj, VFNMS(LDK(KP900968867), Ti, VFNMS(LDK(KP222520933), Tk, Th))); ST(&(x[WS(rs, 2)]), VADD(To, Tp), ms, &(x[0])); ST(&(x[WS(rs, 5)]), VSUB(Tp, To), ms, &(x[WS(rs, 1)])); { V Tg, Tl, Tm, Tn; Tg = VBYI(VFMA(LDK(KP433883739), T5, VFNMS(LDK(KP781831482), Tf, VMUL(LDK(KP974927912), Ta)))); Tl = VFMA(LDK(KP623489801), Ti, VFNMS(LDK(KP222520933), Tj, VFNMS(LDK(KP900968867), Tk, Th))); ST(&(x[WS(rs, 3)]), VADD(Tg, Tl), ms, &(x[WS(rs, 1)])); ST(&(x[WS(rs, 4)]), VSUB(Tl, Tg), ms, &(x[0])); Tm = VBYI(VFMA(LDK(KP781831482), T5, VFMA(LDK(KP974927912), Tf, VMUL(LDK(KP433883739), Ta)))); Tn = VFMA(LDK(KP623489801), Tk, VFNMS(LDK(KP900968867), Tj, VFNMS(LDK(KP222520933), Ti, Th))); ST(&(x[WS(rs, 1)]), VADD(Tm, Tn), ms, &(x[WS(rs, 1)])); ST(&(x[WS(rs, 6)]), VSUB(Tn, Tm), ms, &(x[0])); } } } VLEAVE(); } static const tw_instr twinstr[] = { VTW(0, 1), VTW(0, 2), VTW(0, 3), VTW(0, 4), VTW(0, 5), VTW(0, 6), {TW_NEXT, VL, 0} }; static const ct_desc desc = { 7, XSIMD_STRING("t1buv_7"), twinstr, &GENUS, {24, 18, 12, 0}, 0, 0, 0 }; void XSIMD(codelet_t1buv_7) (planner *p) { X(kdft_dit_register) (p, t1buv_7, &desc); } #endif /* HAVE_FMA */ fftw-3.3.3/dft/simd/common/t1buv_8.c0000644000175000017500000001562312054410547014055 00000000000000/* * Copyright (c) 2003, 2007-11 Matteo Frigo * Copyright (c) 2003, 2007-11 Massachusetts Institute of Technology * * This program is free software; you can redistribute it and/or modify * it under the terms of the GNU General Public License as published by * the Free Software Foundation; either version 2 of the License, or * (at your option) any later version. * * This program is distributed in the hope that it will be useful, * but WITHOUT ANY WARRANTY; without even the implied warranty of * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the * GNU General Public License for more details. * * You should have received a copy of the GNU General Public License * along with this program; if not, write to the Free Software * Foundation, Inc., 51 Franklin Street, Fifth Floor, Boston, MA 02110-1301 USA * */ /* This file was automatically generated --- DO NOT EDIT */ /* Generated on Sun Nov 25 07:39:03 EST 2012 */ #include "codelet-dft.h" #ifdef HAVE_FMA /* Generated by: ../../../genfft/gen_twiddle_c.native -fma -reorder-insns -schedule-for-pipeline -simd -compact -variables 4 -pipeline-latency 8 -n 8 -name t1buv_8 -include t1bu.h -sign 1 */ /* * This function contains 33 FP additions, 24 FP multiplications, * (or, 23 additions, 14 multiplications, 10 fused multiply/add), * 36 stack variables, 1 constants, and 16 memory accesses */ #include "t1bu.h" static void t1buv_8(R *ri, R *ii, const R *W, stride rs, INT mb, INT me, INT ms) { DVK(KP707106781, +0.707106781186547524400844362104849039284835938); { INT m; R *x; x = ii; for (m = mb, W = W + (mb * ((TWVL / VL) * 14)); m < me; m = m + VL, x = x + (VL * ms), W = W + (TWVL * 14), MAKE_VOLATILE_STRIDE(8, rs)) { V T1, T2, Th, Tj, T5, T7, Ta, Tc; T1 = LD(&(x[0]), ms, &(x[0])); T2 = LD(&(x[WS(rs, 4)]), ms, &(x[0])); Th = LD(&(x[WS(rs, 2)]), ms, &(x[0])); Tj = LD(&(x[WS(rs, 6)]), ms, &(x[0])); T5 = LD(&(x[WS(rs, 1)]), ms, &(x[WS(rs, 1)])); T7 = LD(&(x[WS(rs, 5)]), ms, &(x[WS(rs, 1)])); Ta = LD(&(x[WS(rs, 7)]), ms, &(x[WS(rs, 1)])); Tc = LD(&(x[WS(rs, 3)]), ms, &(x[WS(rs, 1)])); { V T3, Ti, Tk, T6, T8, Tb, Td; T3 = BYTW(&(W[TWVL * 6]), T2); Ti = BYTW(&(W[TWVL * 2]), Th); Tk = BYTW(&(W[TWVL * 10]), Tj); T6 = BYTW(&(W[0]), T5); T8 = BYTW(&(W[TWVL * 8]), T7); Tb = BYTW(&(W[TWVL * 12]), Ta); Td = BYTW(&(W[TWVL * 4]), Tc); { V Tq, T4, Tr, Tl, Tt, T9, Tu, Te, Tw, Ts; Tq = VADD(T1, T3); T4 = VSUB(T1, T3); Tr = VADD(Ti, Tk); Tl = VSUB(Ti, Tk); Tt = VADD(T6, T8); T9 = VSUB(T6, T8); Tu = VADD(Tb, Td); Te = VSUB(Tb, Td); Tw = VADD(Tq, Tr); Ts = VSUB(Tq, Tr); { V Tx, Tv, Tm, Tf; Tx = VADD(Tt, Tu); Tv = VSUB(Tt, Tu); Tm = VSUB(T9, Te); Tf = VADD(T9, Te); { V Tp, Tn, To, Tg; ST(&(x[0]), VADD(Tw, Tx), ms, &(x[0])); ST(&(x[WS(rs, 4)]), VSUB(Tw, Tx), ms, &(x[0])); ST(&(x[WS(rs, 2)]), VFMAI(Tv, Ts), ms, &(x[0])); ST(&(x[WS(rs, 6)]), VFNMSI(Tv, Ts), ms, &(x[0])); Tp = VFMA(LDK(KP707106781), Tm, Tl); Tn = VFNMS(LDK(KP707106781), Tm, Tl); To = VFMA(LDK(KP707106781), Tf, T4); Tg = VFNMS(LDK(KP707106781), Tf, T4); ST(&(x[WS(rs, 1)]), VFMAI(Tp, To), ms, &(x[WS(rs, 1)])); ST(&(x[WS(rs, 7)]), VFNMSI(Tp, To), ms, &(x[WS(rs, 1)])); ST(&(x[WS(rs, 5)]), VFMAI(Tn, Tg), ms, &(x[WS(rs, 1)])); ST(&(x[WS(rs, 3)]), VFNMSI(Tn, Tg), ms, &(x[WS(rs, 1)])); } } } } } } VLEAVE(); } static const tw_instr twinstr[] = { VTW(0, 1), VTW(0, 2), VTW(0, 3), VTW(0, 4), VTW(0, 5), VTW(0, 6), VTW(0, 7), {TW_NEXT, VL, 0} }; static const ct_desc desc = { 8, XSIMD_STRING("t1buv_8"), twinstr, &GENUS, {23, 14, 10, 0}, 0, 0, 0 }; void XSIMD(codelet_t1buv_8) (planner *p) { X(kdft_dit_register) (p, t1buv_8, &desc); } #else /* HAVE_FMA */ /* Generated by: ../../../genfft/gen_twiddle_c.native -simd -compact -variables 4 -pipeline-latency 8 -n 8 -name t1buv_8 -include t1bu.h -sign 1 */ /* * This function contains 33 FP additions, 16 FP multiplications, * (or, 33 additions, 16 multiplications, 0 fused multiply/add), * 24 stack variables, 1 constants, and 16 memory accesses */ #include "t1bu.h" static void t1buv_8(R *ri, R *ii, const R *W, stride rs, INT mb, INT me, INT ms) { DVK(KP707106781, +0.707106781186547524400844362104849039284835938); { INT m; R *x; x = ii; for (m = mb, W = W + (mb * ((TWVL / VL) * 14)); m < me; m = m + VL, x = x + (VL * ms), W = W + (TWVL * 14), MAKE_VOLATILE_STRIDE(8, rs)) { V Tl, Tq, Tg, Tr, T5, Tt, Ta, Tu, Ti, Tk, Tj; Ti = LD(&(x[0]), ms, &(x[0])); Tj = LD(&(x[WS(rs, 4)]), ms, &(x[0])); Tk = BYTW(&(W[TWVL * 6]), Tj); Tl = VSUB(Ti, Tk); Tq = VADD(Ti, Tk); { V Td, Tf, Tc, Te; Tc = LD(&(x[WS(rs, 2)]), ms, &(x[0])); Td = BYTW(&(W[TWVL * 2]), Tc); Te = LD(&(x[WS(rs, 6)]), ms, &(x[0])); Tf = BYTW(&(W[TWVL * 10]), Te); Tg = VSUB(Td, Tf); Tr = VADD(Td, Tf); } { V T2, T4, T1, T3; T1 = LD(&(x[WS(rs, 1)]), ms, &(x[WS(rs, 1)])); T2 = BYTW(&(W[0]), T1); T3 = LD(&(x[WS(rs, 5)]), ms, &(x[WS(rs, 1)])); T4 = BYTW(&(W[TWVL * 8]), T3); T5 = VSUB(T2, T4); Tt = VADD(T2, T4); } { V T7, T9, T6, T8; T6 = LD(&(x[WS(rs, 7)]), ms, &(x[WS(rs, 1)])); T7 = BYTW(&(W[TWVL * 12]), T6); T8 = LD(&(x[WS(rs, 3)]), ms, &(x[WS(rs, 1)])); T9 = BYTW(&(W[TWVL * 4]), T8); Ta = VSUB(T7, T9); Tu = VADD(T7, T9); } { V Ts, Tv, Tw, Tx; Ts = VSUB(Tq, Tr); Tv = VBYI(VSUB(Tt, Tu)); ST(&(x[WS(rs, 6)]), VSUB(Ts, Tv), ms, &(x[0])); ST(&(x[WS(rs, 2)]), VADD(Ts, Tv), ms, &(x[0])); Tw = VADD(Tq, Tr); Tx = VADD(Tt, Tu); ST(&(x[WS(rs, 4)]), VSUB(Tw, Tx), ms, &(x[0])); ST(&(x[0]), VADD(Tw, Tx), ms, &(x[0])); { V Th, To, Tn, Tp, Tb, Tm; Tb = VMUL(LDK(KP707106781), VSUB(T5, Ta)); Th = VBYI(VSUB(Tb, Tg)); To = VBYI(VADD(Tg, Tb)); Tm = VMUL(LDK(KP707106781), VADD(T5, Ta)); Tn = VSUB(Tl, Tm); Tp = VADD(Tl, Tm); ST(&(x[WS(rs, 3)]), VADD(Th, Tn), ms, &(x[WS(rs, 1)])); ST(&(x[WS(rs, 7)]), VSUB(Tp, To), ms, &(x[WS(rs, 1)])); ST(&(x[WS(rs, 5)]), VSUB(Tn, Th), ms, &(x[WS(rs, 1)])); ST(&(x[WS(rs, 1)]), VADD(To, Tp), ms, &(x[WS(rs, 1)])); } } } } VLEAVE(); } static const tw_instr twinstr[] = { VTW(0, 1), VTW(0, 2), VTW(0, 3), VTW(0, 4), VTW(0, 5), VTW(0, 6), VTW(0, 7), {TW_NEXT, VL, 0} }; static const ct_desc desc = { 8, XSIMD_STRING("t1buv_8"), twinstr, &GENUS, {33, 16, 0, 0}, 0, 0, 0 }; void XSIMD(codelet_t1buv_8) (planner *p) { X(kdft_dit_register) (p, t1buv_8, &desc); } #endif /* HAVE_FMA */ fftw-3.3.3/dft/simd/common/t1buv_9.c0000644000175000017500000002655012054410547014057 00000000000000/* * Copyright (c) 2003, 2007-11 Matteo Frigo * Copyright (c) 2003, 2007-11 Massachusetts Institute of Technology * * This program is free software; you can redistribute it and/or modify * it under the terms of the GNU General Public License as published by * the Free Software Foundation; either version 2 of the License, or * (at your option) any later version. * * This program is distributed in the hope that it will be useful, * but WITHOUT ANY WARRANTY; without even the implied warranty of * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the * GNU General Public License for more details. * * You should have received a copy of the GNU General Public License * along with this program; if not, write to the Free Software * Foundation, Inc., 51 Franklin Street, Fifth Floor, Boston, MA 02110-1301 USA * */ /* This file was automatically generated --- DO NOT EDIT */ /* Generated on Sun Nov 25 07:39:03 EST 2012 */ #include "codelet-dft.h" #ifdef HAVE_FMA /* Generated by: ../../../genfft/gen_twiddle_c.native -fma -reorder-insns -schedule-for-pipeline -simd -compact -variables 4 -pipeline-latency 8 -n 9 -name t1buv_9 -include t1bu.h -sign 1 */ /* * This function contains 54 FP additions, 54 FP multiplications, * (or, 20 additions, 20 multiplications, 34 fused multiply/add), * 67 stack variables, 19 constants, and 18 memory accesses */ #include "t1bu.h" static void t1buv_9(R *ri, R *ii, const R *W, stride rs, INT mb, INT me, INT ms) { DVK(KP939692620, +0.939692620785908384054109277324731469936208134); DVK(KP907603734, +0.907603734547952313649323976213898122064543220); DVK(KP666666666, +0.666666666666666666666666666666666666666666667); DVK(KP852868531, +0.852868531952443209628250963940074071936020296); DVK(KP879385241, +0.879385241571816768108218554649462939872416269); DVK(KP984807753, +0.984807753012208059366743024589523013670643252); DVK(KP826351822, +0.826351822333069651148283373230685203999624323); DVK(KP347296355, +0.347296355333860697703433253538629592000751354); DVK(KP898197570, +0.898197570222573798468955502359086394667167570); DVK(KP673648177, +0.673648177666930348851716626769314796000375677); DVK(KP420276625, +0.420276625461206169731530603237061658838781920); DVK(KP866025403, +0.866025403784438646763723170752936183471402627); DVK(KP586256827, +0.586256827714544512072145703099641959914944179); DVK(KP968908795, +0.968908795874236621082202410917456709164223497); DVK(KP726681596, +0.726681596905677465811651808188092531873167623); DVK(KP439692620, +0.439692620785908384054109277324731469936208134); DVK(KP203604859, +0.203604859554852403062088995281827210665664861); DVK(KP152703644, +0.152703644666139302296566746461370407999248646); DVK(KP500000000, +0.500000000000000000000000000000000000000000000); { INT m; R *x; x = ii; for (m = mb, W = W + (mb * ((TWVL / VL) * 16)); m < me; m = m + VL, x = x + (VL * ms), W = W + (TWVL * 16), MAKE_VOLATILE_STRIDE(9, rs)) { V T1, T3, T5, T9, Tn, Tb, Td, Th, Tj, Tx, T6; T1 = LD(&(x[0]), ms, &(x[0])); { V T2, T4, T8, Tm; T2 = LD(&(x[WS(rs, 3)]), ms, &(x[WS(rs, 1)])); T4 = LD(&(x[WS(rs, 6)]), ms, &(x[0])); T8 = LD(&(x[WS(rs, 2)]), ms, &(x[0])); Tm = LD(&(x[WS(rs, 1)]), ms, &(x[WS(rs, 1)])); { V Ta, Tc, Tg, Ti; Ta = LD(&(x[WS(rs, 5)]), ms, &(x[WS(rs, 1)])); Tc = LD(&(x[WS(rs, 8)]), ms, &(x[0])); Tg = LD(&(x[WS(rs, 4)]), ms, &(x[0])); Ti = LD(&(x[WS(rs, 7)]), ms, &(x[WS(rs, 1)])); T3 = BYTW(&(W[TWVL * 4]), T2); T5 = BYTW(&(W[TWVL * 10]), T4); T9 = BYTW(&(W[TWVL * 2]), T8); Tn = BYTW(&(W[0]), Tm); Tb = BYTW(&(W[TWVL * 8]), Ta); Td = BYTW(&(W[TWVL * 14]), Tc); Th = BYTW(&(W[TWVL * 6]), Tg); Tj = BYTW(&(W[TWVL * 12]), Ti); } } Tx = VSUB(T3, T5); T6 = VADD(T3, T5); { V Tl, Te, Tk, To, T7, TN; Tl = VSUB(Td, Tb); Te = VADD(Tb, Td); Tk = VSUB(Th, Tj); To = VADD(Th, Tj); T7 = VFNMS(LDK(KP500000000), T6, T1); TN = VADD(T1, T6); { V Tf, TP, Tp, TO; Tf = VFNMS(LDK(KP500000000), Te, T9); TP = VADD(T9, Te); Tp = VFNMS(LDK(KP500000000), To, Tn); TO = VADD(Tn, To); { V Tz, TC, Tu, TD, TA, Tq, TQ, TS; Tz = VFNMS(LDK(KP152703644), Tl, Tf); TC = VFMA(LDK(KP203604859), Tf, Tl); Tu = VFNMS(LDK(KP439692620), Tk, Tf); TD = VFNMS(LDK(KP726681596), Tk, Tp); TA = VFMA(LDK(KP968908795), Tp, Tk); Tq = VFNMS(LDK(KP586256827), Tp, Tl); TQ = VADD(TO, TP); TS = VMUL(LDK(KP866025403), VSUB(TO, TP)); { V TI, TB, TH, TE, Tr, TR, Tw, Tv; Tv = VFNMS(LDK(KP420276625), Tu, Tl); TI = VFMA(LDK(KP673648177), TA, Tz); TB = VFNMS(LDK(KP673648177), TA, Tz); TH = VFNMS(LDK(KP898197570), TD, TC); TE = VFMA(LDK(KP898197570), TD, TC); Tr = VFNMS(LDK(KP347296355), Tq, Tk); ST(&(x[0]), VADD(TQ, TN), ms, &(x[0])); TR = VFNMS(LDK(KP500000000), TQ, TN); Tw = VFNMS(LDK(KP826351822), Tv, Tp); { V TM, TL, TF, TJ, Ts, Ty, TG, TK, Tt; TM = VMUL(LDK(KP984807753), VFMA(LDK(KP879385241), Tx, TI)); TL = VFMA(LDK(KP852868531), TE, T7); TF = VFNMS(LDK(KP500000000), TE, TB); TJ = VFMA(LDK(KP666666666), TI, TH); Ts = VFNMS(LDK(KP907603734), Tr, Tf); ST(&(x[WS(rs, 6)]), VFNMSI(TS, TR), ms, &(x[0])); ST(&(x[WS(rs, 3)]), VFMAI(TS, TR), ms, &(x[WS(rs, 1)])); Ty = VMUL(LDK(KP984807753), VFNMS(LDK(KP879385241), Tx, Tw)); ST(&(x[WS(rs, 8)]), VFNMSI(TM, TL), ms, &(x[0])); ST(&(x[WS(rs, 1)]), VFMAI(TM, TL), ms, &(x[WS(rs, 1)])); TG = VFMA(LDK(KP852868531), TF, T7); TK = VMUL(LDK(KP866025403), VFNMS(LDK(KP852868531), TJ, Tx)); Tt = VFNMS(LDK(KP939692620), Ts, T7); ST(&(x[WS(rs, 5)]), VFNMSI(TK, TG), ms, &(x[WS(rs, 1)])); ST(&(x[WS(rs, 4)]), VFMAI(TK, TG), ms, &(x[0])); ST(&(x[WS(rs, 2)]), VFMAI(Ty, Tt), ms, &(x[0])); ST(&(x[WS(rs, 7)]), VFNMSI(Ty, Tt), ms, &(x[WS(rs, 1)])); } } } } } } } VLEAVE(); } static const tw_instr twinstr[] = { VTW(0, 1), VTW(0, 2), VTW(0, 3), VTW(0, 4), VTW(0, 5), VTW(0, 6), VTW(0, 7), VTW(0, 8), {TW_NEXT, VL, 0} }; static const ct_desc desc = { 9, XSIMD_STRING("t1buv_9"), twinstr, &GENUS, {20, 20, 34, 0}, 0, 0, 0 }; void XSIMD(codelet_t1buv_9) (planner *p) { X(kdft_dit_register) (p, t1buv_9, &desc); } #else /* HAVE_FMA */ /* Generated by: ../../../genfft/gen_twiddle_c.native -simd -compact -variables 4 -pipeline-latency 8 -n 9 -name t1buv_9 -include t1bu.h -sign 1 */ /* * This function contains 54 FP additions, 42 FP multiplications, * (or, 38 additions, 26 multiplications, 16 fused multiply/add), * 38 stack variables, 14 constants, and 18 memory accesses */ #include "t1bu.h" static void t1buv_9(R *ri, R *ii, const R *W, stride rs, INT mb, INT me, INT ms) { DVK(KP939692620, +0.939692620785908384054109277324731469936208134); DVK(KP296198132, +0.296198132726023843175338011893050938967728390); DVK(KP852868531, +0.852868531952443209628250963940074071936020296); DVK(KP173648177, +0.173648177666930348851716626769314796000375677); DVK(KP556670399, +0.556670399226419366452912952047023132968291906); DVK(KP766044443, +0.766044443118978035202392650555416673935832457); DVK(KP642787609, +0.642787609686539326322643409907263432907559884); DVK(KP663413948, +0.663413948168938396205421319635891297216863310); DVK(KP150383733, +0.150383733180435296639271897612501926072238258); DVK(KP342020143, +0.342020143325668733044099614682259580763083368); DVK(KP813797681, +0.813797681349373692844693217248393223289101568); DVK(KP984807753, +0.984807753012208059366743024589523013670643252); DVK(KP500000000, +0.500000000000000000000000000000000000000000000); DVK(KP866025403, +0.866025403784438646763723170752936183471402627); { INT m; R *x; x = ii; for (m = mb, W = W + (mb * ((TWVL / VL) * 16)); m < me; m = m + VL, x = x + (VL * ms), W = W + (TWVL * 16), MAKE_VOLATILE_STRIDE(9, rs)) { V T1, T6, Tu, Tg, Tf, TD, Tq, Tp, TE; T1 = LD(&(x[0]), ms, &(x[0])); { V T3, T5, T2, T4; T2 = LD(&(x[WS(rs, 3)]), ms, &(x[WS(rs, 1)])); T3 = BYTW(&(W[TWVL * 4]), T2); T4 = LD(&(x[WS(rs, 6)]), ms, &(x[0])); T5 = BYTW(&(W[TWVL * 10]), T4); T6 = VADD(T3, T5); Tu = VMUL(LDK(KP866025403), VSUB(T3, T5)); } { V T9, Td, Tb, T8, Tc, Ta, Te; T8 = LD(&(x[WS(rs, 1)]), ms, &(x[WS(rs, 1)])); T9 = BYTW(&(W[0]), T8); Tc = LD(&(x[WS(rs, 7)]), ms, &(x[WS(rs, 1)])); Td = BYTW(&(W[TWVL * 12]), Tc); Ta = LD(&(x[WS(rs, 4)]), ms, &(x[0])); Tb = BYTW(&(W[TWVL * 6]), Ta); Tg = VSUB(Tb, Td); Te = VADD(Tb, Td); Tf = VFNMS(LDK(KP500000000), Te, T9); TD = VADD(T9, Te); } { V Tj, Tn, Tl, Ti, Tm, Tk, To; Ti = LD(&(x[WS(rs, 2)]), ms, &(x[0])); Tj = BYTW(&(W[TWVL * 2]), Ti); Tm = LD(&(x[WS(rs, 8)]), ms, &(x[0])); Tn = BYTW(&(W[TWVL * 14]), Tm); Tk = LD(&(x[WS(rs, 5)]), ms, &(x[WS(rs, 1)])); Tl = BYTW(&(W[TWVL * 8]), Tk); Tq = VSUB(Tl, Tn); To = VADD(Tl, Tn); Tp = VFNMS(LDK(KP500000000), To, Tj); TE = VADD(Tj, To); } { V TF, TG, TH, TI; TF = VBYI(VMUL(LDK(KP866025403), VSUB(TD, TE))); TG = VADD(T1, T6); TH = VADD(TD, TE); TI = VFNMS(LDK(KP500000000), TH, TG); ST(&(x[WS(rs, 3)]), VADD(TF, TI), ms, &(x[WS(rs, 1)])); ST(&(x[0]), VADD(TG, TH), ms, &(x[0])); ST(&(x[WS(rs, 6)]), VSUB(TI, TF), ms, &(x[0])); } { V TC, Tv, Tw, Tx, Th, Tr, Ts, T7, TB; TC = VBYI(VSUB(VFMA(LDK(KP984807753), Tf, VFMA(LDK(KP813797681), Tq, VFNMS(LDK(KP150383733), Tg, VMUL(LDK(KP342020143), Tp)))), Tu)); Tv = VFMA(LDK(KP663413948), Tg, VMUL(LDK(KP642787609), Tf)); Tw = VFMA(LDK(KP150383733), Tq, VMUL(LDK(KP984807753), Tp)); Tx = VADD(Tv, Tw); Th = VFNMS(LDK(KP556670399), Tg, VMUL(LDK(KP766044443), Tf)); Tr = VFNMS(LDK(KP852868531), Tq, VMUL(LDK(KP173648177), Tp)); Ts = VADD(Th, Tr); T7 = VFNMS(LDK(KP500000000), T6, T1); TB = VFMA(LDK(KP852868531), Tg, VFMA(LDK(KP173648177), Tf, VFMA(LDK(KP296198132), Tq, VFNMS(LDK(KP939692620), Tp, T7)))); ST(&(x[WS(rs, 7)]), VSUB(TB, TC), ms, &(x[WS(rs, 1)])); ST(&(x[WS(rs, 2)]), VADD(TB, TC), ms, &(x[0])); { V Tt, Ty, Tz, TA; Tt = VADD(T7, Ts); Ty = VBYI(VADD(Tu, Tx)); ST(&(x[WS(rs, 8)]), VSUB(Tt, Ty), ms, &(x[0])); ST(&(x[WS(rs, 1)]), VADD(Tt, Ty), ms, &(x[WS(rs, 1)])); Tz = VBYI(VADD(Tu, VFNMS(LDK(KP500000000), Tx, VMUL(LDK(KP866025403), VSUB(Th, Tr))))); TA = VFMA(LDK(KP866025403), VSUB(Tw, Tv), VFNMS(LDK(KP500000000), Ts, T7)); ST(&(x[WS(rs, 4)]), VADD(Tz, TA), ms, &(x[0])); ST(&(x[WS(rs, 5)]), VSUB(TA, Tz), ms, &(x[WS(rs, 1)])); } } } } VLEAVE(); } static const tw_instr twinstr[] = { VTW(0, 1), VTW(0, 2), VTW(0, 3), VTW(0, 4), VTW(0, 5), VTW(0, 6), VTW(0, 7), VTW(0, 8), {TW_NEXT, VL, 0} }; static const ct_desc desc = { 9, XSIMD_STRING("t1buv_9"), twinstr, &GENUS, {38, 26, 16, 0}, 0, 0, 0 }; void XSIMD(codelet_t1buv_9) (planner *p) { X(kdft_dit_register) (p, t1buv_9, &desc); } #endif /* HAVE_FMA */ fftw-3.3.3/dft/simd/common/t1buv_10.c0000644000175000017500000002256112054410547014125 00000000000000/* * Copyright (c) 2003, 2007-11 Matteo Frigo * Copyright (c) 2003, 2007-11 Massachusetts Institute of Technology * * This program is free software; you can redistribute it and/or modify * it under the terms of the GNU General Public License as published by * the Free Software Foundation; either version 2 of the License, or * (at your option) any later version. * * This program is distributed in the hope that it will be useful, * but WITHOUT ANY WARRANTY; without even the implied warranty of * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the * GNU General Public License for more details. * * You should have received a copy of the GNU General Public License * along with this program; if not, write to the Free Software * Foundation, Inc., 51 Franklin Street, Fifth Floor, Boston, MA 02110-1301 USA * */ /* This file was automatically generated --- DO NOT EDIT */ /* Generated on Sun Nov 25 07:39:03 EST 2012 */ #include "codelet-dft.h" #ifdef HAVE_FMA /* Generated by: ../../../genfft/gen_twiddle_c.native -fma -reorder-insns -schedule-for-pipeline -simd -compact -variables 4 -pipeline-latency 8 -n 10 -name t1buv_10 -include t1bu.h -sign 1 */ /* * This function contains 51 FP additions, 40 FP multiplications, * (or, 33 additions, 22 multiplications, 18 fused multiply/add), * 43 stack variables, 4 constants, and 20 memory accesses */ #include "t1bu.h" static void t1buv_10(R *ri, R *ii, const R *W, stride rs, INT mb, INT me, INT ms) { DVK(KP559016994, +0.559016994374947424102293417182819058860154590); DVK(KP250000000, +0.250000000000000000000000000000000000000000000); DVK(KP618033988, +0.618033988749894848204586834365638117720309180); DVK(KP951056516, +0.951056516295153572116439333379382143405698634); { INT m; R *x; x = ii; for (m = mb, W = W + (mb * ((TWVL / VL) * 18)); m < me; m = m + VL, x = x + (VL * ms), W = W + (TWVL * 18), MAKE_VOLATILE_STRIDE(10, rs)) { V Td, TA, T4, Ta, Tk, TE, Tp, TF, TB, T9, T1, T2, Tb; T1 = LD(&(x[0]), ms, &(x[0])); T2 = LD(&(x[WS(rs, 5)]), ms, &(x[WS(rs, 1)])); { V Tg, Tn, Ti, Tl; Tg = LD(&(x[WS(rs, 4)]), ms, &(x[0])); Tn = LD(&(x[WS(rs, 1)]), ms, &(x[WS(rs, 1)])); Ti = LD(&(x[WS(rs, 9)]), ms, &(x[WS(rs, 1)])); Tl = LD(&(x[WS(rs, 6)]), ms, &(x[0])); { V T6, T8, T5, Tc; T5 = LD(&(x[WS(rs, 2)]), ms, &(x[0])); Tc = LD(&(x[WS(rs, 3)]), ms, &(x[WS(rs, 1)])); { V T3, Th, To, Tj, Tm, T7; T7 = LD(&(x[WS(rs, 7)]), ms, &(x[WS(rs, 1)])); T3 = BYTW(&(W[TWVL * 8]), T2); Th = BYTW(&(W[TWVL * 6]), Tg); To = BYTW(&(W[0]), Tn); Tj = BYTW(&(W[TWVL * 16]), Ti); Tm = BYTW(&(W[TWVL * 10]), Tl); T6 = BYTW(&(W[TWVL * 2]), T5); Td = BYTW(&(W[TWVL * 4]), Tc); T8 = BYTW(&(W[TWVL * 12]), T7); TA = VADD(T1, T3); T4 = VSUB(T1, T3); Ta = LD(&(x[WS(rs, 8)]), ms, &(x[0])); Tk = VSUB(Th, Tj); TE = VADD(Th, Tj); Tp = VSUB(Tm, To); TF = VADD(Tm, To); } TB = VADD(T6, T8); T9 = VSUB(T6, T8); } } Tb = BYTW(&(W[TWVL * 14]), Ta); { V TL, TG, Tw, Tq, TC, Te; TL = VSUB(TE, TF); TG = VADD(TE, TF); Tw = VSUB(Tk, Tp); Tq = VADD(Tk, Tp); TC = VADD(Tb, Td); Te = VSUB(Tb, Td); { V TM, TD, Tv, Tf; TM = VSUB(TB, TC); TD = VADD(TB, TC); Tv = VSUB(T9, Te); Tf = VADD(T9, Te); { V TP, TN, TH, TJ, Tz, Tx, Tr, Tt, TI, Ts; TP = VMUL(LDK(KP951056516), VFMA(LDK(KP618033988), TL, TM)); TN = VMUL(LDK(KP951056516), VFNMS(LDK(KP618033988), TM, TL)); TH = VADD(TD, TG); TJ = VSUB(TD, TG); Tz = VMUL(LDK(KP951056516), VFNMS(LDK(KP618033988), Tv, Tw)); Tx = VMUL(LDK(KP951056516), VFMA(LDK(KP618033988), Tw, Tv)); Tr = VADD(Tf, Tq); Tt = VSUB(Tf, Tq); ST(&(x[0]), VADD(TA, TH), ms, &(x[0])); TI = VFNMS(LDK(KP250000000), TH, TA); ST(&(x[WS(rs, 5)]), VADD(T4, Tr), ms, &(x[WS(rs, 1)])); Ts = VFNMS(LDK(KP250000000), Tr, T4); { V TK, TO, Tu, Ty; TK = VFNMS(LDK(KP559016994), TJ, TI); TO = VFMA(LDK(KP559016994), TJ, TI); Tu = VFMA(LDK(KP559016994), Tt, Ts); Ty = VFNMS(LDK(KP559016994), Tt, Ts); ST(&(x[WS(rs, 8)]), VFMAI(TN, TK), ms, &(x[0])); ST(&(x[WS(rs, 2)]), VFNMSI(TN, TK), ms, &(x[0])); ST(&(x[WS(rs, 6)]), VFMAI(TP, TO), ms, &(x[0])); ST(&(x[WS(rs, 4)]), VFNMSI(TP, TO), ms, &(x[0])); ST(&(x[WS(rs, 9)]), VFNMSI(Tx, Tu), ms, &(x[WS(rs, 1)])); ST(&(x[WS(rs, 1)]), VFMAI(Tx, Tu), ms, &(x[WS(rs, 1)])); ST(&(x[WS(rs, 7)]), VFNMSI(Tz, Ty), ms, &(x[WS(rs, 1)])); ST(&(x[WS(rs, 3)]), VFMAI(Tz, Ty), ms, &(x[WS(rs, 1)])); } } } } } } VLEAVE(); } static const tw_instr twinstr[] = { VTW(0, 1), VTW(0, 2), VTW(0, 3), VTW(0, 4), VTW(0, 5), VTW(0, 6), VTW(0, 7), VTW(0, 8), VTW(0, 9), {TW_NEXT, VL, 0} }; static const ct_desc desc = { 10, XSIMD_STRING("t1buv_10"), twinstr, &GENUS, {33, 22, 18, 0}, 0, 0, 0 }; void XSIMD(codelet_t1buv_10) (planner *p) { X(kdft_dit_register) (p, t1buv_10, &desc); } #else /* HAVE_FMA */ /* Generated by: ../../../genfft/gen_twiddle_c.native -simd -compact -variables 4 -pipeline-latency 8 -n 10 -name t1buv_10 -include t1bu.h -sign 1 */ /* * This function contains 51 FP additions, 30 FP multiplications, * (or, 45 additions, 24 multiplications, 6 fused multiply/add), * 32 stack variables, 4 constants, and 20 memory accesses */ #include "t1bu.h" static void t1buv_10(R *ri, R *ii, const R *W, stride rs, INT mb, INT me, INT ms) { DVK(KP587785252, +0.587785252292473129168705954639072768597652438); DVK(KP951056516, +0.951056516295153572116439333379382143405698634); DVK(KP250000000, +0.250000000000000000000000000000000000000000000); DVK(KP559016994, +0.559016994374947424102293417182819058860154590); { INT m; R *x; x = ii; for (m = mb, W = W + (mb * ((TWVL / VL) * 18)); m < me; m = m + VL, x = x + (VL * ms), W = W + (TWVL * 18), MAKE_VOLATILE_STRIDE(10, rs)) { V Tu, TH, Tg, Tl, Tp, TD, TE, TJ, T5, Ta, To, TA, TB, TI, Tr; V Tt, Ts; Tr = LD(&(x[0]), ms, &(x[0])); Ts = LD(&(x[WS(rs, 5)]), ms, &(x[WS(rs, 1)])); Tt = BYTW(&(W[TWVL * 8]), Ts); Tu = VSUB(Tr, Tt); TH = VADD(Tr, Tt); { V Td, Tk, Tf, Ti; { V Tc, Tj, Te, Th; Tc = LD(&(x[WS(rs, 4)]), ms, &(x[0])); Td = BYTW(&(W[TWVL * 6]), Tc); Tj = LD(&(x[WS(rs, 1)]), ms, &(x[WS(rs, 1)])); Tk = BYTW(&(W[0]), Tj); Te = LD(&(x[WS(rs, 9)]), ms, &(x[WS(rs, 1)])); Tf = BYTW(&(W[TWVL * 16]), Te); Th = LD(&(x[WS(rs, 6)]), ms, &(x[0])); Ti = BYTW(&(W[TWVL * 10]), Th); } Tg = VSUB(Td, Tf); Tl = VSUB(Ti, Tk); Tp = VADD(Tg, Tl); TD = VADD(Td, Tf); TE = VADD(Ti, Tk); TJ = VADD(TD, TE); } { V T2, T9, T4, T7; { V T1, T8, T3, T6; T1 = LD(&(x[WS(rs, 2)]), ms, &(x[0])); T2 = BYTW(&(W[TWVL * 2]), T1); T8 = LD(&(x[WS(rs, 3)]), ms, &(x[WS(rs, 1)])); T9 = BYTW(&(W[TWVL * 4]), T8); T3 = LD(&(x[WS(rs, 7)]), ms, &(x[WS(rs, 1)])); T4 = BYTW(&(W[TWVL * 12]), T3); T6 = LD(&(x[WS(rs, 8)]), ms, &(x[0])); T7 = BYTW(&(W[TWVL * 14]), T6); } T5 = VSUB(T2, T4); Ta = VSUB(T7, T9); To = VADD(T5, Ta); TA = VADD(T2, T4); TB = VADD(T7, T9); TI = VADD(TA, TB); } { V Tq, Tv, Tw, Tn, Tz, Tb, Tm, Ty, Tx; Tq = VMUL(LDK(KP559016994), VSUB(To, Tp)); Tv = VADD(To, Tp); Tw = VFNMS(LDK(KP250000000), Tv, Tu); Tb = VSUB(T5, Ta); Tm = VSUB(Tg, Tl); Tn = VBYI(VFMA(LDK(KP951056516), Tb, VMUL(LDK(KP587785252), Tm))); Tz = VBYI(VFNMS(LDK(KP951056516), Tm, VMUL(LDK(KP587785252), Tb))); ST(&(x[WS(rs, 5)]), VADD(Tu, Tv), ms, &(x[WS(rs, 1)])); Ty = VSUB(Tw, Tq); ST(&(x[WS(rs, 3)]), VSUB(Ty, Tz), ms, &(x[WS(rs, 1)])); ST(&(x[WS(rs, 7)]), VADD(Tz, Ty), ms, &(x[WS(rs, 1)])); Tx = VADD(Tq, Tw); ST(&(x[WS(rs, 1)]), VADD(Tn, Tx), ms, &(x[WS(rs, 1)])); ST(&(x[WS(rs, 9)]), VSUB(Tx, Tn), ms, &(x[WS(rs, 1)])); } { V TM, TK, TL, TG, TP, TC, TF, TO, TN; TM = VMUL(LDK(KP559016994), VSUB(TI, TJ)); TK = VADD(TI, TJ); TL = VFNMS(LDK(KP250000000), TK, TH); TC = VSUB(TA, TB); TF = VSUB(TD, TE); TG = VBYI(VFNMS(LDK(KP951056516), TF, VMUL(LDK(KP587785252), TC))); TP = VBYI(VFMA(LDK(KP951056516), TC, VMUL(LDK(KP587785252), TF))); ST(&(x[0]), VADD(TH, TK), ms, &(x[0])); TO = VADD(TM, TL); ST(&(x[WS(rs, 4)]), VSUB(TO, TP), ms, &(x[0])); ST(&(x[WS(rs, 6)]), VADD(TP, TO), ms, &(x[0])); TN = VSUB(TL, TM); ST(&(x[WS(rs, 2)]), VADD(TG, TN), ms, &(x[0])); ST(&(x[WS(rs, 8)]), VSUB(TN, TG), ms, &(x[0])); } } } VLEAVE(); } static const tw_instr twinstr[] = { VTW(0, 1), VTW(0, 2), VTW(0, 3), VTW(0, 4), VTW(0, 5), VTW(0, 6), VTW(0, 7), VTW(0, 8), VTW(0, 9), {TW_NEXT, VL, 0} }; static const ct_desc desc = { 10, XSIMD_STRING("t1buv_10"), twinstr, &GENUS, {45, 24, 6, 0}, 0, 0, 0 }; void XSIMD(codelet_t1buv_10) (planner *p) { X(kdft_dit_register) (p, t1buv_10, &desc); } #endif /* HAVE_FMA */ fftw-3.3.3/dft/simd/common/t1bv_2.c0000644000175000017500000000655412054410547013665 00000000000000/* * Copyright (c) 2003, 2007-11 Matteo Frigo * Copyright (c) 2003, 2007-11 Massachusetts Institute of Technology * * This program is free software; you can redistribute it and/or modify * it under the terms of the GNU General Public License as published by * the Free Software Foundation; either version 2 of the License, or * (at your option) any later version. * * This program is distributed in the hope that it will be useful, * but WITHOUT ANY WARRANTY; without even the implied warranty of * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the * GNU General Public License for more details. * * You should have received a copy of the GNU General Public License * along with this program; if not, write to the Free Software * Foundation, Inc., 51 Franklin Street, Fifth Floor, Boston, MA 02110-1301 USA * */ /* This file was automatically generated --- DO NOT EDIT */ /* Generated on Sun Nov 25 07:39:03 EST 2012 */ #include "codelet-dft.h" #ifdef HAVE_FMA /* Generated by: ../../../genfft/gen_twiddle_c.native -fma -reorder-insns -schedule-for-pipeline -simd -compact -variables 4 -pipeline-latency 8 -n 2 -name t1bv_2 -include t1b.h -sign 1 */ /* * This function contains 3 FP additions, 2 FP multiplications, * (or, 3 additions, 2 multiplications, 0 fused multiply/add), * 5 stack variables, 0 constants, and 4 memory accesses */ #include "t1b.h" static void t1bv_2(R *ri, R *ii, const R *W, stride rs, INT mb, INT me, INT ms) { { INT m; R *x; x = ii; for (m = mb, W = W + (mb * ((TWVL / VL) * 2)); m < me; m = m + VL, x = x + (VL * ms), W = W + (TWVL * 2), MAKE_VOLATILE_STRIDE(2, rs)) { V T1, T2, T3; T1 = LD(&(x[0]), ms, &(x[0])); T2 = LD(&(x[WS(rs, 1)]), ms, &(x[WS(rs, 1)])); T3 = BYTW(&(W[0]), T2); ST(&(x[0]), VADD(T1, T3), ms, &(x[0])); ST(&(x[WS(rs, 1)]), VSUB(T1, T3), ms, &(x[WS(rs, 1)])); } } VLEAVE(); } static const tw_instr twinstr[] = { VTW(0, 1), {TW_NEXT, VL, 0} }; static const ct_desc desc = { 2, XSIMD_STRING("t1bv_2"), twinstr, &GENUS, {3, 2, 0, 0}, 0, 0, 0 }; void XSIMD(codelet_t1bv_2) (planner *p) { X(kdft_dit_register) (p, t1bv_2, &desc); } #else /* HAVE_FMA */ /* Generated by: ../../../genfft/gen_twiddle_c.native -simd -compact -variables 4 -pipeline-latency 8 -n 2 -name t1bv_2 -include t1b.h -sign 1 */ /* * This function contains 3 FP additions, 2 FP multiplications, * (or, 3 additions, 2 multiplications, 0 fused multiply/add), * 5 stack variables, 0 constants, and 4 memory accesses */ #include "t1b.h" static void t1bv_2(R *ri, R *ii, const R *W, stride rs, INT mb, INT me, INT ms) { { INT m; R *x; x = ii; for (m = mb, W = W + (mb * ((TWVL / VL) * 2)); m < me; m = m + VL, x = x + (VL * ms), W = W + (TWVL * 2), MAKE_VOLATILE_STRIDE(2, rs)) { V T1, T3, T2; T1 = LD(&(x[0]), ms, &(x[0])); T2 = LD(&(x[WS(rs, 1)]), ms, &(x[WS(rs, 1)])); T3 = BYTW(&(W[0]), T2); ST(&(x[WS(rs, 1)]), VSUB(T1, T3), ms, &(x[WS(rs, 1)])); ST(&(x[0]), VADD(T1, T3), ms, &(x[0])); } } VLEAVE(); } static const tw_instr twinstr[] = { VTW(0, 1), {TW_NEXT, VL, 0} }; static const ct_desc desc = { 2, XSIMD_STRING("t1bv_2"), twinstr, &GENUS, {3, 2, 0, 0}, 0, 0, 0 }; void XSIMD(codelet_t1bv_2) (planner *p) { X(kdft_dit_register) (p, t1bv_2, &desc); } #endif /* HAVE_FMA */ fftw-3.3.3/dft/simd/common/t1bv_3.c0000644000175000017500000001040312054410547013652 00000000000000/* * Copyright (c) 2003, 2007-11 Matteo Frigo * Copyright (c) 2003, 2007-11 Massachusetts Institute of Technology * * This program is free software; you can redistribute it and/or modify * it under the terms of the GNU General Public License as published by * the Free Software Foundation; either version 2 of the License, or * (at your option) any later version. * * This program is distributed in the hope that it will be useful, * but WITHOUT ANY WARRANTY; without even the implied warranty of * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the * GNU General Public License for more details. * * You should have received a copy of the GNU General Public License * along with this program; if not, write to the Free Software * Foundation, Inc., 51 Franklin Street, Fifth Floor, Boston, MA 02110-1301 USA * */ /* This file was automatically generated --- DO NOT EDIT */ /* Generated on Sun Nov 25 07:39:03 EST 2012 */ #include "codelet-dft.h" #ifdef HAVE_FMA /* Generated by: ../../../genfft/gen_twiddle_c.native -fma -reorder-insns -schedule-for-pipeline -simd -compact -variables 4 -pipeline-latency 8 -n 3 -name t1bv_3 -include t1b.h -sign 1 */ /* * This function contains 8 FP additions, 8 FP multiplications, * (or, 5 additions, 5 multiplications, 3 fused multiply/add), * 12 stack variables, 2 constants, and 6 memory accesses */ #include "t1b.h" static void t1bv_3(R *ri, R *ii, const R *W, stride rs, INT mb, INT me, INT ms) { DVK(KP500000000, +0.500000000000000000000000000000000000000000000); DVK(KP866025403, +0.866025403784438646763723170752936183471402627); { INT m; R *x; x = ii; for (m = mb, W = W + (mb * ((TWVL / VL) * 4)); m < me; m = m + VL, x = x + (VL * ms), W = W + (TWVL * 4), MAKE_VOLATILE_STRIDE(3, rs)) { V T1, T2, T4; T1 = LD(&(x[0]), ms, &(x[0])); T2 = LD(&(x[WS(rs, 1)]), ms, &(x[WS(rs, 1)])); T4 = LD(&(x[WS(rs, 2)]), ms, &(x[0])); { V T3, T5, T8, T6, T7; T3 = BYTW(&(W[0]), T2); T5 = BYTW(&(W[TWVL * 2]), T4); T8 = VMUL(LDK(KP866025403), VSUB(T3, T5)); T6 = VADD(T3, T5); T7 = VFNMS(LDK(KP500000000), T6, T1); ST(&(x[0]), VADD(T1, T6), ms, &(x[0])); ST(&(x[WS(rs, 2)]), VFNMSI(T8, T7), ms, &(x[0])); ST(&(x[WS(rs, 1)]), VFMAI(T8, T7), ms, &(x[WS(rs, 1)])); } } } VLEAVE(); } static const tw_instr twinstr[] = { VTW(0, 1), VTW(0, 2), {TW_NEXT, VL, 0} }; static const ct_desc desc = { 3, XSIMD_STRING("t1bv_3"), twinstr, &GENUS, {5, 5, 3, 0}, 0, 0, 0 }; void XSIMD(codelet_t1bv_3) (planner *p) { X(kdft_dit_register) (p, t1bv_3, &desc); } #else /* HAVE_FMA */ /* Generated by: ../../../genfft/gen_twiddle_c.native -simd -compact -variables 4 -pipeline-latency 8 -n 3 -name t1bv_3 -include t1b.h -sign 1 */ /* * This function contains 8 FP additions, 6 FP multiplications, * (or, 7 additions, 5 multiplications, 1 fused multiply/add), * 12 stack variables, 2 constants, and 6 memory accesses */ #include "t1b.h" static void t1bv_3(R *ri, R *ii, const R *W, stride rs, INT mb, INT me, INT ms) { DVK(KP500000000, +0.500000000000000000000000000000000000000000000); DVK(KP866025403, +0.866025403784438646763723170752936183471402627); { INT m; R *x; x = ii; for (m = mb, W = W + (mb * ((TWVL / VL) * 4)); m < me; m = m + VL, x = x + (VL * ms), W = W + (TWVL * 4), MAKE_VOLATILE_STRIDE(3, rs)) { V T6, T2, T4, T7, T1, T3, T5, T8; T6 = LD(&(x[0]), ms, &(x[0])); T1 = LD(&(x[WS(rs, 1)]), ms, &(x[WS(rs, 1)])); T2 = BYTW(&(W[0]), T1); T3 = LD(&(x[WS(rs, 2)]), ms, &(x[0])); T4 = BYTW(&(W[TWVL * 2]), T3); T7 = VADD(T2, T4); ST(&(x[0]), VADD(T6, T7), ms, &(x[0])); T5 = VBYI(VMUL(LDK(KP866025403), VSUB(T2, T4))); T8 = VFNMS(LDK(KP500000000), T7, T6); ST(&(x[WS(rs, 1)]), VADD(T5, T8), ms, &(x[WS(rs, 1)])); ST(&(x[WS(rs, 2)]), VSUB(T8, T5), ms, &(x[0])); } } VLEAVE(); } static const tw_instr twinstr[] = { VTW(0, 1), VTW(0, 2), {TW_NEXT, VL, 0} }; static const ct_desc desc = { 3, XSIMD_STRING("t1bv_3"), twinstr, &GENUS, {7, 5, 1, 0}, 0, 0, 0 }; void XSIMD(codelet_t1bv_3) (planner *p) { X(kdft_dit_register) (p, t1bv_3, &desc); } #endif /* HAVE_FMA */ fftw-3.3.3/dft/simd/common/t1bv_4.c0000644000175000017500000001050312054410547013654 00000000000000/* * Copyright (c) 2003, 2007-11 Matteo Frigo * Copyright (c) 2003, 2007-11 Massachusetts Institute of Technology * * This program is free software; you can redistribute it and/or modify * it under the terms of the GNU General Public License as published by * the Free Software Foundation; either version 2 of the License, or * (at your option) any later version. * * This program is distributed in the hope that it will be useful, * but WITHOUT ANY WARRANTY; without even the implied warranty of * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the * GNU General Public License for more details. * * You should have received a copy of the GNU General Public License * along with this program; if not, write to the Free Software * Foundation, Inc., 51 Franklin Street, Fifth Floor, Boston, MA 02110-1301 USA * */ /* This file was automatically generated --- DO NOT EDIT */ /* Generated on Sun Nov 25 07:39:03 EST 2012 */ #include "codelet-dft.h" #ifdef HAVE_FMA /* Generated by: ../../../genfft/gen_twiddle_c.native -fma -reorder-insns -schedule-for-pipeline -simd -compact -variables 4 -pipeline-latency 8 -n 4 -name t1bv_4 -include t1b.h -sign 1 */ /* * This function contains 11 FP additions, 8 FP multiplications, * (or, 9 additions, 6 multiplications, 2 fused multiply/add), * 13 stack variables, 0 constants, and 8 memory accesses */ #include "t1b.h" static void t1bv_4(R *ri, R *ii, const R *W, stride rs, INT mb, INT me, INT ms) { { INT m; R *x; x = ii; for (m = mb, W = W + (mb * ((TWVL / VL) * 6)); m < me; m = m + VL, x = x + (VL * ms), W = W + (TWVL * 6), MAKE_VOLATILE_STRIDE(4, rs)) { V T1, T7, T2, T5, T8, T3, T6; T1 = LD(&(x[0]), ms, &(x[0])); T7 = LD(&(x[WS(rs, 3)]), ms, &(x[WS(rs, 1)])); T2 = LD(&(x[WS(rs, 2)]), ms, &(x[0])); T5 = LD(&(x[WS(rs, 1)]), ms, &(x[WS(rs, 1)])); T8 = BYTW(&(W[TWVL * 4]), T7); T3 = BYTW(&(W[TWVL * 2]), T2); T6 = BYTW(&(W[0]), T5); { V Ta, T4, Tb, T9; Ta = VADD(T1, T3); T4 = VSUB(T1, T3); Tb = VADD(T6, T8); T9 = VSUB(T6, T8); ST(&(x[0]), VADD(Ta, Tb), ms, &(x[0])); ST(&(x[WS(rs, 2)]), VSUB(Ta, Tb), ms, &(x[0])); ST(&(x[WS(rs, 1)]), VFMAI(T9, T4), ms, &(x[WS(rs, 1)])); ST(&(x[WS(rs, 3)]), VFNMSI(T9, T4), ms, &(x[WS(rs, 1)])); } } } VLEAVE(); } static const tw_instr twinstr[] = { VTW(0, 1), VTW(0, 2), VTW(0, 3), {TW_NEXT, VL, 0} }; static const ct_desc desc = { 4, XSIMD_STRING("t1bv_4"), twinstr, &GENUS, {9, 6, 2, 0}, 0, 0, 0 }; void XSIMD(codelet_t1bv_4) (planner *p) { X(kdft_dit_register) (p, t1bv_4, &desc); } #else /* HAVE_FMA */ /* Generated by: ../../../genfft/gen_twiddle_c.native -simd -compact -variables 4 -pipeline-latency 8 -n 4 -name t1bv_4 -include t1b.h -sign 1 */ /* * This function contains 11 FP additions, 6 FP multiplications, * (or, 11 additions, 6 multiplications, 0 fused multiply/add), * 13 stack variables, 0 constants, and 8 memory accesses */ #include "t1b.h" static void t1bv_4(R *ri, R *ii, const R *W, stride rs, INT mb, INT me, INT ms) { { INT m; R *x; x = ii; for (m = mb, W = W + (mb * ((TWVL / VL) * 6)); m < me; m = m + VL, x = x + (VL * ms), W = W + (TWVL * 6), MAKE_VOLATILE_STRIDE(4, rs)) { V T1, T8, T3, T6, T7, T2, T5; T1 = LD(&(x[0]), ms, &(x[0])); T7 = LD(&(x[WS(rs, 3)]), ms, &(x[WS(rs, 1)])); T8 = BYTW(&(W[TWVL * 4]), T7); T2 = LD(&(x[WS(rs, 2)]), ms, &(x[0])); T3 = BYTW(&(W[TWVL * 2]), T2); T5 = LD(&(x[WS(rs, 1)]), ms, &(x[WS(rs, 1)])); T6 = BYTW(&(W[0]), T5); { V T4, T9, Ta, Tb; T4 = VSUB(T1, T3); T9 = VBYI(VSUB(T6, T8)); ST(&(x[WS(rs, 3)]), VSUB(T4, T9), ms, &(x[WS(rs, 1)])); ST(&(x[WS(rs, 1)]), VADD(T4, T9), ms, &(x[WS(rs, 1)])); Ta = VADD(T1, T3); Tb = VADD(T6, T8); ST(&(x[WS(rs, 2)]), VSUB(Ta, Tb), ms, &(x[0])); ST(&(x[0]), VADD(Ta, Tb), ms, &(x[0])); } } } VLEAVE(); } static const tw_instr twinstr[] = { VTW(0, 1), VTW(0, 2), VTW(0, 3), {TW_NEXT, VL, 0} }; static const ct_desc desc = { 4, XSIMD_STRING("t1bv_4"), twinstr, &GENUS, {11, 6, 0, 0}, 0, 0, 0 }; void XSIMD(codelet_t1bv_4) (planner *p) { X(kdft_dit_register) (p, t1bv_4, &desc); } #endif /* HAVE_FMA */ fftw-3.3.3/dft/simd/common/t1bv_5.c0000644000175000017500000001373612054410550013662 00000000000000/* * Copyright (c) 2003, 2007-11 Matteo Frigo * Copyright (c) 2003, 2007-11 Massachusetts Institute of Technology * * This program is free software; you can redistribute it and/or modify * it under the terms of the GNU General Public License as published by * the Free Software Foundation; either version 2 of the License, or * (at your option) any later version. * * This program is distributed in the hope that it will be useful, * but WITHOUT ANY WARRANTY; without even the implied warranty of * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the * GNU General Public License for more details. * * You should have received a copy of the GNU General Public License * along with this program; if not, write to the Free Software * Foundation, Inc., 51 Franklin Street, Fifth Floor, Boston, MA 02110-1301 USA * */ /* This file was automatically generated --- DO NOT EDIT */ /* Generated on Sun Nov 25 07:39:03 EST 2012 */ #include "codelet-dft.h" #ifdef HAVE_FMA /* Generated by: ../../../genfft/gen_twiddle_c.native -fma -reorder-insns -schedule-for-pipeline -simd -compact -variables 4 -pipeline-latency 8 -n 5 -name t1bv_5 -include t1b.h -sign 1 */ /* * This function contains 20 FP additions, 19 FP multiplications, * (or, 11 additions, 10 multiplications, 9 fused multiply/add), * 26 stack variables, 4 constants, and 10 memory accesses */ #include "t1b.h" static void t1bv_5(R *ri, R *ii, const R *W, stride rs, INT mb, INT me, INT ms) { DVK(KP559016994, +0.559016994374947424102293417182819058860154590); DVK(KP250000000, +0.250000000000000000000000000000000000000000000); DVK(KP618033988, +0.618033988749894848204586834365638117720309180); DVK(KP951056516, +0.951056516295153572116439333379382143405698634); { INT m; R *x; x = ii; for (m = mb, W = W + (mb * ((TWVL / VL) * 8)); m < me; m = m + VL, x = x + (VL * ms), W = W + (TWVL * 8), MAKE_VOLATILE_STRIDE(5, rs)) { V T1, T2, T9, T4, T7; T1 = LD(&(x[0]), ms, &(x[0])); T2 = LD(&(x[WS(rs, 1)]), ms, &(x[WS(rs, 1)])); T9 = LD(&(x[WS(rs, 3)]), ms, &(x[WS(rs, 1)])); T4 = LD(&(x[WS(rs, 4)]), ms, &(x[0])); T7 = LD(&(x[WS(rs, 2)]), ms, &(x[0])); { V T3, Ta, T5, T8; T3 = BYTW(&(W[0]), T2); Ta = BYTW(&(W[TWVL * 4]), T9); T5 = BYTW(&(W[TWVL * 6]), T4); T8 = BYTW(&(W[TWVL * 2]), T7); { V T6, Tg, Tb, Th; T6 = VADD(T3, T5); Tg = VSUB(T3, T5); Tb = VADD(T8, Ta); Th = VSUB(T8, Ta); { V Te, Tc, Tk, Ti, Td, Tj, Tf; Te = VSUB(T6, Tb); Tc = VADD(T6, Tb); Tk = VMUL(LDK(KP951056516), VFNMS(LDK(KP618033988), Tg, Th)); Ti = VMUL(LDK(KP951056516), VFMA(LDK(KP618033988), Th, Tg)); Td = VFNMS(LDK(KP250000000), Tc, T1); ST(&(x[0]), VADD(T1, Tc), ms, &(x[0])); Tj = VFNMS(LDK(KP559016994), Te, Td); Tf = VFMA(LDK(KP559016994), Te, Td); ST(&(x[WS(rs, 2)]), VFNMSI(Tk, Tj), ms, &(x[0])); ST(&(x[WS(rs, 3)]), VFMAI(Tk, Tj), ms, &(x[WS(rs, 1)])); ST(&(x[WS(rs, 4)]), VFNMSI(Ti, Tf), ms, &(x[0])); ST(&(x[WS(rs, 1)]), VFMAI(Ti, Tf), ms, &(x[WS(rs, 1)])); } } } } } VLEAVE(); } static const tw_instr twinstr[] = { VTW(0, 1), VTW(0, 2), VTW(0, 3), VTW(0, 4), {TW_NEXT, VL, 0} }; static const ct_desc desc = { 5, XSIMD_STRING("t1bv_5"), twinstr, &GENUS, {11, 10, 9, 0}, 0, 0, 0 }; void XSIMD(codelet_t1bv_5) (planner *p) { X(kdft_dit_register) (p, t1bv_5, &desc); } #else /* HAVE_FMA */ /* Generated by: ../../../genfft/gen_twiddle_c.native -simd -compact -variables 4 -pipeline-latency 8 -n 5 -name t1bv_5 -include t1b.h -sign 1 */ /* * This function contains 20 FP additions, 14 FP multiplications, * (or, 17 additions, 11 multiplications, 3 fused multiply/add), * 20 stack variables, 4 constants, and 10 memory accesses */ #include "t1b.h" static void t1bv_5(R *ri, R *ii, const R *W, stride rs, INT mb, INT me, INT ms) { DVK(KP250000000, +0.250000000000000000000000000000000000000000000); DVK(KP559016994, +0.559016994374947424102293417182819058860154590); DVK(KP587785252, +0.587785252292473129168705954639072768597652438); DVK(KP951056516, +0.951056516295153572116439333379382143405698634); { INT m; R *x; x = ii; for (m = mb, W = W + (mb * ((TWVL / VL) * 8)); m < me; m = m + VL, x = x + (VL * ms), W = W + (TWVL * 8), MAKE_VOLATILE_STRIDE(5, rs)) { V Tf, T5, Ta, Tc, Td, Tg; Tf = LD(&(x[0]), ms, &(x[0])); { V T2, T9, T4, T7; { V T1, T8, T3, T6; T1 = LD(&(x[WS(rs, 1)]), ms, &(x[WS(rs, 1)])); T2 = BYTW(&(W[0]), T1); T8 = LD(&(x[WS(rs, 3)]), ms, &(x[WS(rs, 1)])); T9 = BYTW(&(W[TWVL * 4]), T8); T3 = LD(&(x[WS(rs, 4)]), ms, &(x[0])); T4 = BYTW(&(W[TWVL * 6]), T3); T6 = LD(&(x[WS(rs, 2)]), ms, &(x[0])); T7 = BYTW(&(W[TWVL * 2]), T6); } T5 = VSUB(T2, T4); Ta = VSUB(T7, T9); Tc = VADD(T2, T4); Td = VADD(T7, T9); Tg = VADD(Tc, Td); } ST(&(x[0]), VADD(Tf, Tg), ms, &(x[0])); { V Tb, Tj, Ti, Tk, Te, Th; Tb = VBYI(VFMA(LDK(KP951056516), T5, VMUL(LDK(KP587785252), Ta))); Tj = VBYI(VFNMS(LDK(KP951056516), Ta, VMUL(LDK(KP587785252), T5))); Te = VMUL(LDK(KP559016994), VSUB(Tc, Td)); Th = VFNMS(LDK(KP250000000), Tg, Tf); Ti = VADD(Te, Th); Tk = VSUB(Th, Te); ST(&(x[WS(rs, 1)]), VADD(Tb, Ti), ms, &(x[WS(rs, 1)])); ST(&(x[WS(rs, 3)]), VSUB(Tk, Tj), ms, &(x[WS(rs, 1)])); ST(&(x[WS(rs, 4)]), VSUB(Ti, Tb), ms, &(x[0])); ST(&(x[WS(rs, 2)]), VADD(Tj, Tk), ms, &(x[0])); } } } VLEAVE(); } static const tw_instr twinstr[] = { VTW(0, 1), VTW(0, 2), VTW(0, 3), VTW(0, 4), {TW_NEXT, VL, 0} }; static const ct_desc desc = { 5, XSIMD_STRING("t1bv_5"), twinstr, &GENUS, {17, 11, 3, 0}, 0, 0, 0 }; void XSIMD(codelet_t1bv_5) (planner *p) { X(kdft_dit_register) (p, t1bv_5, &desc); } #endif /* HAVE_FMA */ fftw-3.3.3/dft/simd/common/t1bv_6.c0000644000175000017500000001402112054410550013647 00000000000000/* * Copyright (c) 2003, 2007-11 Matteo Frigo * Copyright (c) 2003, 2007-11 Massachusetts Institute of Technology * * This program is free software; you can redistribute it and/or modify * it under the terms of the GNU General Public License as published by * the Free Software Foundation; either version 2 of the License, or * (at your option) any later version. * * This program is distributed in the hope that it will be useful, * but WITHOUT ANY WARRANTY; without even the implied warranty of * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the * GNU General Public License for more details. * * You should have received a copy of the GNU General Public License * along with this program; if not, write to the Free Software * Foundation, Inc., 51 Franklin Street, Fifth Floor, Boston, MA 02110-1301 USA * */ /* This file was automatically generated --- DO NOT EDIT */ /* Generated on Sun Nov 25 07:39:03 EST 2012 */ #include "codelet-dft.h" #ifdef HAVE_FMA /* Generated by: ../../../genfft/gen_twiddle_c.native -fma -reorder-insns -schedule-for-pipeline -simd -compact -variables 4 -pipeline-latency 8 -n 6 -name t1bv_6 -include t1b.h -sign 1 */ /* * This function contains 23 FP additions, 18 FP multiplications, * (or, 17 additions, 12 multiplications, 6 fused multiply/add), * 27 stack variables, 2 constants, and 12 memory accesses */ #include "t1b.h" static void t1bv_6(R *ri, R *ii, const R *W, stride rs, INT mb, INT me, INT ms) { DVK(KP500000000, +0.500000000000000000000000000000000000000000000); DVK(KP866025403, +0.866025403784438646763723170752936183471402627); { INT m; R *x; x = ii; for (m = mb, W = W + (mb * ((TWVL / VL) * 10)); m < me; m = m + VL, x = x + (VL * ms), W = W + (TWVL * 10), MAKE_VOLATILE_STRIDE(6, rs)) { V T1, T2, Ta, Tc, T5, T7; T1 = LD(&(x[0]), ms, &(x[0])); T2 = LD(&(x[WS(rs, 3)]), ms, &(x[WS(rs, 1)])); Ta = LD(&(x[WS(rs, 4)]), ms, &(x[0])); Tc = LD(&(x[WS(rs, 1)]), ms, &(x[WS(rs, 1)])); T5 = LD(&(x[WS(rs, 2)]), ms, &(x[0])); T7 = LD(&(x[WS(rs, 5)]), ms, &(x[WS(rs, 1)])); { V T3, Tb, Td, T6, T8; T3 = BYTW(&(W[TWVL * 4]), T2); Tb = BYTW(&(W[TWVL * 6]), Ta); Td = BYTW(&(W[0]), Tc); T6 = BYTW(&(W[TWVL * 2]), T5); T8 = BYTW(&(W[TWVL * 8]), T7); { V Ti, T4, Tk, Te, Tj, T9; Ti = VADD(T1, T3); T4 = VSUB(T1, T3); Tk = VADD(Tb, Td); Te = VSUB(Tb, Td); Tj = VADD(T6, T8); T9 = VSUB(T6, T8); { V Tl, Tn, Tf, Th, Tm, Tg; Tl = VADD(Tj, Tk); Tn = VMUL(LDK(KP866025403), VSUB(Tj, Tk)); Tf = VADD(T9, Te); Th = VMUL(LDK(KP866025403), VSUB(T9, Te)); ST(&(x[0]), VADD(Ti, Tl), ms, &(x[0])); Tm = VFNMS(LDK(KP500000000), Tl, Ti); ST(&(x[WS(rs, 3)]), VADD(T4, Tf), ms, &(x[WS(rs, 1)])); Tg = VFNMS(LDK(KP500000000), Tf, T4); ST(&(x[WS(rs, 4)]), VFMAI(Tn, Tm), ms, &(x[0])); ST(&(x[WS(rs, 2)]), VFNMSI(Tn, Tm), ms, &(x[0])); ST(&(x[WS(rs, 5)]), VFNMSI(Th, Tg), ms, &(x[WS(rs, 1)])); ST(&(x[WS(rs, 1)]), VFMAI(Th, Tg), ms, &(x[WS(rs, 1)])); } } } } } VLEAVE(); } static const tw_instr twinstr[] = { VTW(0, 1), VTW(0, 2), VTW(0, 3), VTW(0, 4), VTW(0, 5), {TW_NEXT, VL, 0} }; static const ct_desc desc = { 6, XSIMD_STRING("t1bv_6"), twinstr, &GENUS, {17, 12, 6, 0}, 0, 0, 0 }; void XSIMD(codelet_t1bv_6) (planner *p) { X(kdft_dit_register) (p, t1bv_6, &desc); } #else /* HAVE_FMA */ /* Generated by: ../../../genfft/gen_twiddle_c.native -simd -compact -variables 4 -pipeline-latency 8 -n 6 -name t1bv_6 -include t1b.h -sign 1 */ /* * This function contains 23 FP additions, 14 FP multiplications, * (or, 21 additions, 12 multiplications, 2 fused multiply/add), * 19 stack variables, 2 constants, and 12 memory accesses */ #include "t1b.h" static void t1bv_6(R *ri, R *ii, const R *W, stride rs, INT mb, INT me, INT ms) { DVK(KP500000000, +0.500000000000000000000000000000000000000000000); DVK(KP866025403, +0.866025403784438646763723170752936183471402627); { INT m; R *x; x = ii; for (m = mb, W = W + (mb * ((TWVL / VL) * 10)); m < me; m = m + VL, x = x + (VL * ms), W = W + (TWVL * 10), MAKE_VOLATILE_STRIDE(6, rs)) { V Tf, Ti, Ta, Tk, T5, Tj, Tc, Te, Td; Tc = LD(&(x[0]), ms, &(x[0])); Td = LD(&(x[WS(rs, 3)]), ms, &(x[WS(rs, 1)])); Te = BYTW(&(W[TWVL * 4]), Td); Tf = VSUB(Tc, Te); Ti = VADD(Tc, Te); { V T7, T9, T6, T8; T6 = LD(&(x[WS(rs, 4)]), ms, &(x[0])); T7 = BYTW(&(W[TWVL * 6]), T6); T8 = LD(&(x[WS(rs, 1)]), ms, &(x[WS(rs, 1)])); T9 = BYTW(&(W[0]), T8); Ta = VSUB(T7, T9); Tk = VADD(T7, T9); } { V T2, T4, T1, T3; T1 = LD(&(x[WS(rs, 2)]), ms, &(x[0])); T2 = BYTW(&(W[TWVL * 2]), T1); T3 = LD(&(x[WS(rs, 5)]), ms, &(x[WS(rs, 1)])); T4 = BYTW(&(W[TWVL * 8]), T3); T5 = VSUB(T2, T4); Tj = VADD(T2, T4); } { V Tb, Tg, Th, Tn, Tl, Tm; Tb = VBYI(VMUL(LDK(KP866025403), VSUB(T5, Ta))); Tg = VADD(T5, Ta); Th = VFNMS(LDK(KP500000000), Tg, Tf); ST(&(x[WS(rs, 1)]), VADD(Tb, Th), ms, &(x[WS(rs, 1)])); ST(&(x[WS(rs, 3)]), VADD(Tf, Tg), ms, &(x[WS(rs, 1)])); ST(&(x[WS(rs, 5)]), VSUB(Th, Tb), ms, &(x[WS(rs, 1)])); Tn = VBYI(VMUL(LDK(KP866025403), VSUB(Tj, Tk))); Tl = VADD(Tj, Tk); Tm = VFNMS(LDK(KP500000000), Tl, Ti); ST(&(x[WS(rs, 2)]), VSUB(Tm, Tn), ms, &(x[0])); ST(&(x[0]), VADD(Ti, Tl), ms, &(x[0])); ST(&(x[WS(rs, 4)]), VADD(Tn, Tm), ms, &(x[0])); } } } VLEAVE(); } static const tw_instr twinstr[] = { VTW(0, 1), VTW(0, 2), VTW(0, 3), VTW(0, 4), VTW(0, 5), {TW_NEXT, VL, 0} }; static const ct_desc desc = { 6, XSIMD_STRING("t1bv_6"), twinstr, &GENUS, {21, 12, 2, 0}, 0, 0, 0 }; void XSIMD(codelet_t1bv_6) (planner *p) { X(kdft_dit_register) (p, t1bv_6, &desc); } #endif /* HAVE_FMA */ fftw-3.3.3/dft/simd/common/t1bv_7.c0000644000175000017500000001760712054410550013665 00000000000000/* * Copyright (c) 2003, 2007-11 Matteo Frigo * Copyright (c) 2003, 2007-11 Massachusetts Institute of Technology * * This program is free software; you can redistribute it and/or modify * it under the terms of the GNU General Public License as published by * the Free Software Foundation; either version 2 of the License, or * (at your option) any later version. * * This program is distributed in the hope that it will be useful, * but WITHOUT ANY WARRANTY; without even the implied warranty of * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the * GNU General Public License for more details. * * You should have received a copy of the GNU General Public License * along with this program; if not, write to the Free Software * Foundation, Inc., 51 Franklin Street, Fifth Floor, Boston, MA 02110-1301 USA * */ /* This file was automatically generated --- DO NOT EDIT */ /* Generated on Sun Nov 25 07:39:03 EST 2012 */ #include "codelet-dft.h" #ifdef HAVE_FMA /* Generated by: ../../../genfft/gen_twiddle_c.native -fma -reorder-insns -schedule-for-pipeline -simd -compact -variables 4 -pipeline-latency 8 -n 7 -name t1bv_7 -include t1b.h -sign 1 */ /* * This function contains 36 FP additions, 36 FP multiplications, * (or, 15 additions, 15 multiplications, 21 fused multiply/add), * 42 stack variables, 6 constants, and 14 memory accesses */ #include "t1b.h" static void t1bv_7(R *ri, R *ii, const R *W, stride rs, INT mb, INT me, INT ms) { DVK(KP900968867, +0.900968867902419126236102319507445051165919162); DVK(KP801937735, +0.801937735804838252472204639014890102331838324); DVK(KP974927912, +0.974927912181823607018131682993931217232785801); DVK(KP692021471, +0.692021471630095869627814897002069140197260599); DVK(KP554958132, +0.554958132087371191422194871006410481067288862); DVK(KP356895867, +0.356895867892209443894399510021300583399127187); { INT m; R *x; x = ii; for (m = mb, W = W + (mb * ((TWVL / VL) * 12)); m < me; m = m + VL, x = x + (VL * ms), W = W + (TWVL * 12), MAKE_VOLATILE_STRIDE(7, rs)) { V T1, T2, T4, Te, Tc, T9, T7; T1 = LD(&(x[0]), ms, &(x[0])); T2 = LD(&(x[WS(rs, 1)]), ms, &(x[WS(rs, 1)])); T4 = LD(&(x[WS(rs, 6)]), ms, &(x[0])); Te = LD(&(x[WS(rs, 4)]), ms, &(x[0])); Tc = LD(&(x[WS(rs, 3)]), ms, &(x[WS(rs, 1)])); T9 = LD(&(x[WS(rs, 5)]), ms, &(x[WS(rs, 1)])); T7 = LD(&(x[WS(rs, 2)]), ms, &(x[0])); { V T3, T5, Tf, Td, Ta, T8; T3 = BYTW(&(W[0]), T2); T5 = BYTW(&(W[TWVL * 10]), T4); Tf = BYTW(&(W[TWVL * 6]), Te); Td = BYTW(&(W[TWVL * 4]), Tc); Ta = BYTW(&(W[TWVL * 8]), T9); T8 = BYTW(&(W[TWVL * 2]), T7); { V T6, Tm, Tg, Tk, Tb, Tl; T6 = VADD(T3, T5); Tm = VSUB(T3, T5); Tg = VADD(Td, Tf); Tk = VSUB(Td, Tf); Tb = VADD(T8, Ta); Tl = VSUB(T8, Ta); { V Tp, Tx, Tu, Th, Ts, Tn, Tq, Ty; Tp = VFNMS(LDK(KP356895867), T6, Tg); Tx = VFMA(LDK(KP554958132), Tk, Tm); ST(&(x[0]), VADD(T1, VADD(T6, VADD(Tb, Tg))), ms, &(x[0])); Tu = VFNMS(LDK(KP356895867), Tb, T6); Th = VFNMS(LDK(KP356895867), Tg, Tb); Ts = VFMA(LDK(KP554958132), Tl, Tk); Tn = VFNMS(LDK(KP554958132), Tm, Tl); Tq = VFNMS(LDK(KP692021471), Tp, Tb); Ty = VMUL(LDK(KP974927912), VFMA(LDK(KP801937735), Tx, Tl)); { V Tv, Ti, Tt, To, Tr, Tw, Tj; Tv = VFNMS(LDK(KP692021471), Tu, Tg); Ti = VFNMS(LDK(KP692021471), Th, T6); Tt = VMUL(LDK(KP974927912), VFNMS(LDK(KP801937735), Ts, Tm)); To = VMUL(LDK(KP974927912), VFNMS(LDK(KP801937735), Tn, Tk)); Tr = VFNMS(LDK(KP900968867), Tq, T1); Tw = VFNMS(LDK(KP900968867), Tv, T1); Tj = VFNMS(LDK(KP900968867), Ti, T1); ST(&(x[WS(rs, 5)]), VFNMSI(Tt, Tr), ms, &(x[WS(rs, 1)])); ST(&(x[WS(rs, 2)]), VFMAI(Tt, Tr), ms, &(x[0])); ST(&(x[WS(rs, 6)]), VFNMSI(Ty, Tw), ms, &(x[0])); ST(&(x[WS(rs, 1)]), VFMAI(Ty, Tw), ms, &(x[WS(rs, 1)])); ST(&(x[WS(rs, 4)]), VFNMSI(To, Tj), ms, &(x[0])); ST(&(x[WS(rs, 3)]), VFMAI(To, Tj), ms, &(x[WS(rs, 1)])); } } } } } } VLEAVE(); } static const tw_instr twinstr[] = { VTW(0, 1), VTW(0, 2), VTW(0, 3), VTW(0, 4), VTW(0, 5), VTW(0, 6), {TW_NEXT, VL, 0} }; static const ct_desc desc = { 7, XSIMD_STRING("t1bv_7"), twinstr, &GENUS, {15, 15, 21, 0}, 0, 0, 0 }; void XSIMD(codelet_t1bv_7) (planner *p) { X(kdft_dit_register) (p, t1bv_7, &desc); } #else /* HAVE_FMA */ /* Generated by: ../../../genfft/gen_twiddle_c.native -simd -compact -variables 4 -pipeline-latency 8 -n 7 -name t1bv_7 -include t1b.h -sign 1 */ /* * This function contains 36 FP additions, 30 FP multiplications, * (or, 24 additions, 18 multiplications, 12 fused multiply/add), * 21 stack variables, 6 constants, and 14 memory accesses */ #include "t1b.h" static void t1bv_7(R *ri, R *ii, const R *W, stride rs, INT mb, INT me, INT ms) { DVK(KP222520933, +0.222520933956314404288902564496794759466355569); DVK(KP900968867, +0.900968867902419126236102319507445051165919162); DVK(KP623489801, +0.623489801858733530525004884004239810632274731); DVK(KP433883739, +0.433883739117558120475768332848358754609990728); DVK(KP781831482, +0.781831482468029808708444526674057750232334519); DVK(KP974927912, +0.974927912181823607018131682993931217232785801); { INT m; R *x; x = ii; for (m = mb, W = W + (mb * ((TWVL / VL) * 12)); m < me; m = m + VL, x = x + (VL * ms), W = W + (TWVL * 12), MAKE_VOLATILE_STRIDE(7, rs)) { V Th, Tf, Ti, T5, Tk, Ta, Tj, To, Tp; Th = LD(&(x[0]), ms, &(x[0])); { V Tc, Te, Tb, Td; Tb = LD(&(x[WS(rs, 2)]), ms, &(x[0])); Tc = BYTW(&(W[TWVL * 2]), Tb); Td = LD(&(x[WS(rs, 5)]), ms, &(x[WS(rs, 1)])); Te = BYTW(&(W[TWVL * 8]), Td); Tf = VSUB(Tc, Te); Ti = VADD(Tc, Te); } { V T2, T4, T1, T3; T1 = LD(&(x[WS(rs, 1)]), ms, &(x[WS(rs, 1)])); T2 = BYTW(&(W[0]), T1); T3 = LD(&(x[WS(rs, 6)]), ms, &(x[0])); T4 = BYTW(&(W[TWVL * 10]), T3); T5 = VSUB(T2, T4); Tk = VADD(T2, T4); } { V T7, T9, T6, T8; T6 = LD(&(x[WS(rs, 3)]), ms, &(x[WS(rs, 1)])); T7 = BYTW(&(W[TWVL * 4]), T6); T8 = LD(&(x[WS(rs, 4)]), ms, &(x[0])); T9 = BYTW(&(W[TWVL * 6]), T8); Ta = VSUB(T7, T9); Tj = VADD(T7, T9); } ST(&(x[0]), VADD(Th, VADD(Tk, VADD(Ti, Tj))), ms, &(x[0])); To = VBYI(VFNMS(LDK(KP781831482), Ta, VFNMS(LDK(KP433883739), Tf, VMUL(LDK(KP974927912), T5)))); Tp = VFMA(LDK(KP623489801), Tj, VFNMS(LDK(KP900968867), Ti, VFNMS(LDK(KP222520933), Tk, Th))); ST(&(x[WS(rs, 2)]), VADD(To, Tp), ms, &(x[0])); ST(&(x[WS(rs, 5)]), VSUB(Tp, To), ms, &(x[WS(rs, 1)])); { V Tg, Tl, Tm, Tn; Tg = VBYI(VFMA(LDK(KP433883739), T5, VFNMS(LDK(KP781831482), Tf, VMUL(LDK(KP974927912), Ta)))); Tl = VFMA(LDK(KP623489801), Ti, VFNMS(LDK(KP222520933), Tj, VFNMS(LDK(KP900968867), Tk, Th))); ST(&(x[WS(rs, 3)]), VADD(Tg, Tl), ms, &(x[WS(rs, 1)])); ST(&(x[WS(rs, 4)]), VSUB(Tl, Tg), ms, &(x[0])); Tm = VBYI(VFMA(LDK(KP781831482), T5, VFMA(LDK(KP974927912), Tf, VMUL(LDK(KP433883739), Ta)))); Tn = VFMA(LDK(KP623489801), Tk, VFNMS(LDK(KP900968867), Tj, VFNMS(LDK(KP222520933), Ti, Th))); ST(&(x[WS(rs, 1)]), VADD(Tm, Tn), ms, &(x[WS(rs, 1)])); ST(&(x[WS(rs, 6)]), VSUB(Tn, Tm), ms, &(x[0])); } } } VLEAVE(); } static const tw_instr twinstr[] = { VTW(0, 1), VTW(0, 2), VTW(0, 3), VTW(0, 4), VTW(0, 5), VTW(0, 6), {TW_NEXT, VL, 0} }; static const ct_desc desc = { 7, XSIMD_STRING("t1bv_7"), twinstr, &GENUS, {24, 18, 12, 0}, 0, 0, 0 }; void XSIMD(codelet_t1bv_7) (planner *p) { X(kdft_dit_register) (p, t1bv_7, &desc); } #endif /* HAVE_FMA */ fftw-3.3.3/dft/simd/common/t1bv_8.c0000644000175000017500000001560512054410550013662 00000000000000/* * Copyright (c) 2003, 2007-11 Matteo Frigo * Copyright (c) 2003, 2007-11 Massachusetts Institute of Technology * * This program is free software; you can redistribute it and/or modify * it under the terms of the GNU General Public License as published by * the Free Software Foundation; either version 2 of the License, or * (at your option) any later version. * * This program is distributed in the hope that it will be useful, * but WITHOUT ANY WARRANTY; without even the implied warranty of * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the * GNU General Public License for more details. * * You should have received a copy of the GNU General Public License * along with this program; if not, write to the Free Software * Foundation, Inc., 51 Franklin Street, Fifth Floor, Boston, MA 02110-1301 USA * */ /* This file was automatically generated --- DO NOT EDIT */ /* Generated on Sun Nov 25 07:39:04 EST 2012 */ #include "codelet-dft.h" #ifdef HAVE_FMA /* Generated by: ../../../genfft/gen_twiddle_c.native -fma -reorder-insns -schedule-for-pipeline -simd -compact -variables 4 -pipeline-latency 8 -n 8 -name t1bv_8 -include t1b.h -sign 1 */ /* * This function contains 33 FP additions, 24 FP multiplications, * (or, 23 additions, 14 multiplications, 10 fused multiply/add), * 36 stack variables, 1 constants, and 16 memory accesses */ #include "t1b.h" static void t1bv_8(R *ri, R *ii, const R *W, stride rs, INT mb, INT me, INT ms) { DVK(KP707106781, +0.707106781186547524400844362104849039284835938); { INT m; R *x; x = ii; for (m = mb, W = W + (mb * ((TWVL / VL) * 14)); m < me; m = m + VL, x = x + (VL * ms), W = W + (TWVL * 14), MAKE_VOLATILE_STRIDE(8, rs)) { V T1, T2, Th, Tj, T5, T7, Ta, Tc; T1 = LD(&(x[0]), ms, &(x[0])); T2 = LD(&(x[WS(rs, 4)]), ms, &(x[0])); Th = LD(&(x[WS(rs, 2)]), ms, &(x[0])); Tj = LD(&(x[WS(rs, 6)]), ms, &(x[0])); T5 = LD(&(x[WS(rs, 1)]), ms, &(x[WS(rs, 1)])); T7 = LD(&(x[WS(rs, 5)]), ms, &(x[WS(rs, 1)])); Ta = LD(&(x[WS(rs, 7)]), ms, &(x[WS(rs, 1)])); Tc = LD(&(x[WS(rs, 3)]), ms, &(x[WS(rs, 1)])); { V T3, Ti, Tk, T6, T8, Tb, Td; T3 = BYTW(&(W[TWVL * 6]), T2); Ti = BYTW(&(W[TWVL * 2]), Th); Tk = BYTW(&(W[TWVL * 10]), Tj); T6 = BYTW(&(W[0]), T5); T8 = BYTW(&(W[TWVL * 8]), T7); Tb = BYTW(&(W[TWVL * 12]), Ta); Td = BYTW(&(W[TWVL * 4]), Tc); { V Tq, T4, Tr, Tl, Tt, T9, Tu, Te, Tw, Ts; Tq = VADD(T1, T3); T4 = VSUB(T1, T3); Tr = VADD(Ti, Tk); Tl = VSUB(Ti, Tk); Tt = VADD(T6, T8); T9 = VSUB(T6, T8); Tu = VADD(Tb, Td); Te = VSUB(Tb, Td); Tw = VADD(Tq, Tr); Ts = VSUB(Tq, Tr); { V Tx, Tv, Tm, Tf; Tx = VADD(Tt, Tu); Tv = VSUB(Tt, Tu); Tm = VSUB(T9, Te); Tf = VADD(T9, Te); { V Tp, Tn, To, Tg; ST(&(x[0]), VADD(Tw, Tx), ms, &(x[0])); ST(&(x[WS(rs, 4)]), VSUB(Tw, Tx), ms, &(x[0])); ST(&(x[WS(rs, 2)]), VFMAI(Tv, Ts), ms, &(x[0])); ST(&(x[WS(rs, 6)]), VFNMSI(Tv, Ts), ms, &(x[0])); Tp = VFMA(LDK(KP707106781), Tm, Tl); Tn = VFNMS(LDK(KP707106781), Tm, Tl); To = VFMA(LDK(KP707106781), Tf, T4); Tg = VFNMS(LDK(KP707106781), Tf, T4); ST(&(x[WS(rs, 1)]), VFMAI(Tp, To), ms, &(x[WS(rs, 1)])); ST(&(x[WS(rs, 7)]), VFNMSI(Tp, To), ms, &(x[WS(rs, 1)])); ST(&(x[WS(rs, 5)]), VFMAI(Tn, Tg), ms, &(x[WS(rs, 1)])); ST(&(x[WS(rs, 3)]), VFNMSI(Tn, Tg), ms, &(x[WS(rs, 1)])); } } } } } } VLEAVE(); } static const tw_instr twinstr[] = { VTW(0, 1), VTW(0, 2), VTW(0, 3), VTW(0, 4), VTW(0, 5), VTW(0, 6), VTW(0, 7), {TW_NEXT, VL, 0} }; static const ct_desc desc = { 8, XSIMD_STRING("t1bv_8"), twinstr, &GENUS, {23, 14, 10, 0}, 0, 0, 0 }; void XSIMD(codelet_t1bv_8) (planner *p) { X(kdft_dit_register) (p, t1bv_8, &desc); } #else /* HAVE_FMA */ /* Generated by: ../../../genfft/gen_twiddle_c.native -simd -compact -variables 4 -pipeline-latency 8 -n 8 -name t1bv_8 -include t1b.h -sign 1 */ /* * This function contains 33 FP additions, 16 FP multiplications, * (or, 33 additions, 16 multiplications, 0 fused multiply/add), * 24 stack variables, 1 constants, and 16 memory accesses */ #include "t1b.h" static void t1bv_8(R *ri, R *ii, const R *W, stride rs, INT mb, INT me, INT ms) { DVK(KP707106781, +0.707106781186547524400844362104849039284835938); { INT m; R *x; x = ii; for (m = mb, W = W + (mb * ((TWVL / VL) * 14)); m < me; m = m + VL, x = x + (VL * ms), W = W + (TWVL * 14), MAKE_VOLATILE_STRIDE(8, rs)) { V Tl, Tq, Tg, Tr, T5, Tt, Ta, Tu, Ti, Tk, Tj; Ti = LD(&(x[0]), ms, &(x[0])); Tj = LD(&(x[WS(rs, 4)]), ms, &(x[0])); Tk = BYTW(&(W[TWVL * 6]), Tj); Tl = VSUB(Ti, Tk); Tq = VADD(Ti, Tk); { V Td, Tf, Tc, Te; Tc = LD(&(x[WS(rs, 2)]), ms, &(x[0])); Td = BYTW(&(W[TWVL * 2]), Tc); Te = LD(&(x[WS(rs, 6)]), ms, &(x[0])); Tf = BYTW(&(W[TWVL * 10]), Te); Tg = VSUB(Td, Tf); Tr = VADD(Td, Tf); } { V T2, T4, T1, T3; T1 = LD(&(x[WS(rs, 1)]), ms, &(x[WS(rs, 1)])); T2 = BYTW(&(W[0]), T1); T3 = LD(&(x[WS(rs, 5)]), ms, &(x[WS(rs, 1)])); T4 = BYTW(&(W[TWVL * 8]), T3); T5 = VSUB(T2, T4); Tt = VADD(T2, T4); } { V T7, T9, T6, T8; T6 = LD(&(x[WS(rs, 7)]), ms, &(x[WS(rs, 1)])); T7 = BYTW(&(W[TWVL * 12]), T6); T8 = LD(&(x[WS(rs, 3)]), ms, &(x[WS(rs, 1)])); T9 = BYTW(&(W[TWVL * 4]), T8); Ta = VSUB(T7, T9); Tu = VADD(T7, T9); } { V Ts, Tv, Tw, Tx; Ts = VSUB(Tq, Tr); Tv = VBYI(VSUB(Tt, Tu)); ST(&(x[WS(rs, 6)]), VSUB(Ts, Tv), ms, &(x[0])); ST(&(x[WS(rs, 2)]), VADD(Ts, Tv), ms, &(x[0])); Tw = VADD(Tq, Tr); Tx = VADD(Tt, Tu); ST(&(x[WS(rs, 4)]), VSUB(Tw, Tx), ms, &(x[0])); ST(&(x[0]), VADD(Tw, Tx), ms, &(x[0])); { V Th, To, Tn, Tp, Tb, Tm; Tb = VMUL(LDK(KP707106781), VSUB(T5, Ta)); Th = VBYI(VSUB(Tb, Tg)); To = VBYI(VADD(Tg, Tb)); Tm = VMUL(LDK(KP707106781), VADD(T5, Ta)); Tn = VSUB(Tl, Tm); Tp = VADD(Tl, Tm); ST(&(x[WS(rs, 3)]), VADD(Th, Tn), ms, &(x[WS(rs, 1)])); ST(&(x[WS(rs, 7)]), VSUB(Tp, To), ms, &(x[WS(rs, 1)])); ST(&(x[WS(rs, 5)]), VSUB(Tn, Th), ms, &(x[WS(rs, 1)])); ST(&(x[WS(rs, 1)]), VADD(To, Tp), ms, &(x[WS(rs, 1)])); } } } } VLEAVE(); } static const tw_instr twinstr[] = { VTW(0, 1), VTW(0, 2), VTW(0, 3), VTW(0, 4), VTW(0, 5), VTW(0, 6), VTW(0, 7), {TW_NEXT, VL, 0} }; static const ct_desc desc = { 8, XSIMD_STRING("t1bv_8"), twinstr, &GENUS, {33, 16, 0, 0}, 0, 0, 0 }; void XSIMD(codelet_t1bv_8) (planner *p) { X(kdft_dit_register) (p, t1bv_8, &desc); } #endif /* HAVE_FMA */ fftw-3.3.3/dft/simd/common/t1bv_9.c0000644000175000017500000002653212054410550013664 00000000000000/* * Copyright (c) 2003, 2007-11 Matteo Frigo * Copyright (c) 2003, 2007-11 Massachusetts Institute of Technology * * This program is free software; you can redistribute it and/or modify * it under the terms of the GNU General Public License as published by * the Free Software Foundation; either version 2 of the License, or * (at your option) any later version. * * This program is distributed in the hope that it will be useful, * but WITHOUT ANY WARRANTY; without even the implied warranty of * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the * GNU General Public License for more details. * * You should have received a copy of the GNU General Public License * along with this program; if not, write to the Free Software * Foundation, Inc., 51 Franklin Street, Fifth Floor, Boston, MA 02110-1301 USA * */ /* This file was automatically generated --- DO NOT EDIT */ /* Generated on Sun Nov 25 07:39:04 EST 2012 */ #include "codelet-dft.h" #ifdef HAVE_FMA /* Generated by: ../../../genfft/gen_twiddle_c.native -fma -reorder-insns -schedule-for-pipeline -simd -compact -variables 4 -pipeline-latency 8 -n 9 -name t1bv_9 -include t1b.h -sign 1 */ /* * This function contains 54 FP additions, 54 FP multiplications, * (or, 20 additions, 20 multiplications, 34 fused multiply/add), * 67 stack variables, 19 constants, and 18 memory accesses */ #include "t1b.h" static void t1bv_9(R *ri, R *ii, const R *W, stride rs, INT mb, INT me, INT ms) { DVK(KP939692620, +0.939692620785908384054109277324731469936208134); DVK(KP907603734, +0.907603734547952313649323976213898122064543220); DVK(KP666666666, +0.666666666666666666666666666666666666666666667); DVK(KP852868531, +0.852868531952443209628250963940074071936020296); DVK(KP879385241, +0.879385241571816768108218554649462939872416269); DVK(KP984807753, +0.984807753012208059366743024589523013670643252); DVK(KP826351822, +0.826351822333069651148283373230685203999624323); DVK(KP347296355, +0.347296355333860697703433253538629592000751354); DVK(KP898197570, +0.898197570222573798468955502359086394667167570); DVK(KP673648177, +0.673648177666930348851716626769314796000375677); DVK(KP420276625, +0.420276625461206169731530603237061658838781920); DVK(KP866025403, +0.866025403784438646763723170752936183471402627); DVK(KP586256827, +0.586256827714544512072145703099641959914944179); DVK(KP968908795, +0.968908795874236621082202410917456709164223497); DVK(KP726681596, +0.726681596905677465811651808188092531873167623); DVK(KP439692620, +0.439692620785908384054109277324731469936208134); DVK(KP203604859, +0.203604859554852403062088995281827210665664861); DVK(KP152703644, +0.152703644666139302296566746461370407999248646); DVK(KP500000000, +0.500000000000000000000000000000000000000000000); { INT m; R *x; x = ii; for (m = mb, W = W + (mb * ((TWVL / VL) * 16)); m < me; m = m + VL, x = x + (VL * ms), W = W + (TWVL * 16), MAKE_VOLATILE_STRIDE(9, rs)) { V T1, T3, T5, T9, Tn, Tb, Td, Th, Tj, Tx, T6; T1 = LD(&(x[0]), ms, &(x[0])); { V T2, T4, T8, Tm; T2 = LD(&(x[WS(rs, 3)]), ms, &(x[WS(rs, 1)])); T4 = LD(&(x[WS(rs, 6)]), ms, &(x[0])); T8 = LD(&(x[WS(rs, 2)]), ms, &(x[0])); Tm = LD(&(x[WS(rs, 1)]), ms, &(x[WS(rs, 1)])); { V Ta, Tc, Tg, Ti; Ta = LD(&(x[WS(rs, 5)]), ms, &(x[WS(rs, 1)])); Tc = LD(&(x[WS(rs, 8)]), ms, &(x[0])); Tg = LD(&(x[WS(rs, 4)]), ms, &(x[0])); Ti = LD(&(x[WS(rs, 7)]), ms, &(x[WS(rs, 1)])); T3 = BYTW(&(W[TWVL * 4]), T2); T5 = BYTW(&(W[TWVL * 10]), T4); T9 = BYTW(&(W[TWVL * 2]), T8); Tn = BYTW(&(W[0]), Tm); Tb = BYTW(&(W[TWVL * 8]), Ta); Td = BYTW(&(W[TWVL * 14]), Tc); Th = BYTW(&(W[TWVL * 6]), Tg); Tj = BYTW(&(W[TWVL * 12]), Ti); } } Tx = VSUB(T3, T5); T6 = VADD(T3, T5); { V Tl, Te, Tk, To, T7, TN; Tl = VSUB(Td, Tb); Te = VADD(Tb, Td); Tk = VSUB(Th, Tj); To = VADD(Th, Tj); T7 = VFNMS(LDK(KP500000000), T6, T1); TN = VADD(T1, T6); { V Tf, TP, Tp, TO; Tf = VFNMS(LDK(KP500000000), Te, T9); TP = VADD(T9, Te); Tp = VFNMS(LDK(KP500000000), To, Tn); TO = VADD(Tn, To); { V Tz, TC, Tu, TD, TA, Tq, TQ, TS; Tz = VFNMS(LDK(KP152703644), Tl, Tf); TC = VFMA(LDK(KP203604859), Tf, Tl); Tu = VFNMS(LDK(KP439692620), Tk, Tf); TD = VFNMS(LDK(KP726681596), Tk, Tp); TA = VFMA(LDK(KP968908795), Tp, Tk); Tq = VFNMS(LDK(KP586256827), Tp, Tl); TQ = VADD(TO, TP); TS = VMUL(LDK(KP866025403), VSUB(TO, TP)); { V TI, TB, TH, TE, Tr, TR, Tw, Tv; Tv = VFNMS(LDK(KP420276625), Tu, Tl); TI = VFMA(LDK(KP673648177), TA, Tz); TB = VFNMS(LDK(KP673648177), TA, Tz); TH = VFNMS(LDK(KP898197570), TD, TC); TE = VFMA(LDK(KP898197570), TD, TC); Tr = VFNMS(LDK(KP347296355), Tq, Tk); ST(&(x[0]), VADD(TQ, TN), ms, &(x[0])); TR = VFNMS(LDK(KP500000000), TQ, TN); Tw = VFNMS(LDK(KP826351822), Tv, Tp); { V TM, TL, TF, TJ, Ts, Ty, TG, TK, Tt; TM = VMUL(LDK(KP984807753), VFMA(LDK(KP879385241), Tx, TI)); TL = VFMA(LDK(KP852868531), TE, T7); TF = VFNMS(LDK(KP500000000), TE, TB); TJ = VFMA(LDK(KP666666666), TI, TH); Ts = VFNMS(LDK(KP907603734), Tr, Tf); ST(&(x[WS(rs, 6)]), VFNMSI(TS, TR), ms, &(x[0])); ST(&(x[WS(rs, 3)]), VFMAI(TS, TR), ms, &(x[WS(rs, 1)])); Ty = VMUL(LDK(KP984807753), VFNMS(LDK(KP879385241), Tx, Tw)); ST(&(x[WS(rs, 8)]), VFNMSI(TM, TL), ms, &(x[0])); ST(&(x[WS(rs, 1)]), VFMAI(TM, TL), ms, &(x[WS(rs, 1)])); TG = VFMA(LDK(KP852868531), TF, T7); TK = VMUL(LDK(KP866025403), VFNMS(LDK(KP852868531), TJ, Tx)); Tt = VFNMS(LDK(KP939692620), Ts, T7); ST(&(x[WS(rs, 5)]), VFNMSI(TK, TG), ms, &(x[WS(rs, 1)])); ST(&(x[WS(rs, 4)]), VFMAI(TK, TG), ms, &(x[0])); ST(&(x[WS(rs, 2)]), VFMAI(Ty, Tt), ms, &(x[0])); ST(&(x[WS(rs, 7)]), VFNMSI(Ty, Tt), ms, &(x[WS(rs, 1)])); } } } } } } } VLEAVE(); } static const tw_instr twinstr[] = { VTW(0, 1), VTW(0, 2), VTW(0, 3), VTW(0, 4), VTW(0, 5), VTW(0, 6), VTW(0, 7), VTW(0, 8), {TW_NEXT, VL, 0} }; static const ct_desc desc = { 9, XSIMD_STRING("t1bv_9"), twinstr, &GENUS, {20, 20, 34, 0}, 0, 0, 0 }; void XSIMD(codelet_t1bv_9) (planner *p) { X(kdft_dit_register) (p, t1bv_9, &desc); } #else /* HAVE_FMA */ /* Generated by: ../../../genfft/gen_twiddle_c.native -simd -compact -variables 4 -pipeline-latency 8 -n 9 -name t1bv_9 -include t1b.h -sign 1 */ /* * This function contains 54 FP additions, 42 FP multiplications, * (or, 38 additions, 26 multiplications, 16 fused multiply/add), * 38 stack variables, 14 constants, and 18 memory accesses */ #include "t1b.h" static void t1bv_9(R *ri, R *ii, const R *W, stride rs, INT mb, INT me, INT ms) { DVK(KP939692620, +0.939692620785908384054109277324731469936208134); DVK(KP296198132, +0.296198132726023843175338011893050938967728390); DVK(KP852868531, +0.852868531952443209628250963940074071936020296); DVK(KP173648177, +0.173648177666930348851716626769314796000375677); DVK(KP556670399, +0.556670399226419366452912952047023132968291906); DVK(KP766044443, +0.766044443118978035202392650555416673935832457); DVK(KP642787609, +0.642787609686539326322643409907263432907559884); DVK(KP663413948, +0.663413948168938396205421319635891297216863310); DVK(KP150383733, +0.150383733180435296639271897612501926072238258); DVK(KP342020143, +0.342020143325668733044099614682259580763083368); DVK(KP813797681, +0.813797681349373692844693217248393223289101568); DVK(KP984807753, +0.984807753012208059366743024589523013670643252); DVK(KP500000000, +0.500000000000000000000000000000000000000000000); DVK(KP866025403, +0.866025403784438646763723170752936183471402627); { INT m; R *x; x = ii; for (m = mb, W = W + (mb * ((TWVL / VL) * 16)); m < me; m = m + VL, x = x + (VL * ms), W = W + (TWVL * 16), MAKE_VOLATILE_STRIDE(9, rs)) { V T1, T6, Tu, Tg, Tf, TD, Tq, Tp, TE; T1 = LD(&(x[0]), ms, &(x[0])); { V T3, T5, T2, T4; T2 = LD(&(x[WS(rs, 3)]), ms, &(x[WS(rs, 1)])); T3 = BYTW(&(W[TWVL * 4]), T2); T4 = LD(&(x[WS(rs, 6)]), ms, &(x[0])); T5 = BYTW(&(W[TWVL * 10]), T4); T6 = VADD(T3, T5); Tu = VMUL(LDK(KP866025403), VSUB(T3, T5)); } { V T9, Td, Tb, T8, Tc, Ta, Te; T8 = LD(&(x[WS(rs, 1)]), ms, &(x[WS(rs, 1)])); T9 = BYTW(&(W[0]), T8); Tc = LD(&(x[WS(rs, 7)]), ms, &(x[WS(rs, 1)])); Td = BYTW(&(W[TWVL * 12]), Tc); Ta = LD(&(x[WS(rs, 4)]), ms, &(x[0])); Tb = BYTW(&(W[TWVL * 6]), Ta); Tg = VSUB(Tb, Td); Te = VADD(Tb, Td); Tf = VFNMS(LDK(KP500000000), Te, T9); TD = VADD(T9, Te); } { V Tj, Tn, Tl, Ti, Tm, Tk, To; Ti = LD(&(x[WS(rs, 2)]), ms, &(x[0])); Tj = BYTW(&(W[TWVL * 2]), Ti); Tm = LD(&(x[WS(rs, 8)]), ms, &(x[0])); Tn = BYTW(&(W[TWVL * 14]), Tm); Tk = LD(&(x[WS(rs, 5)]), ms, &(x[WS(rs, 1)])); Tl = BYTW(&(W[TWVL * 8]), Tk); Tq = VSUB(Tl, Tn); To = VADD(Tl, Tn); Tp = VFNMS(LDK(KP500000000), To, Tj); TE = VADD(Tj, To); } { V TF, TG, TH, TI; TF = VBYI(VMUL(LDK(KP866025403), VSUB(TD, TE))); TG = VADD(T1, T6); TH = VADD(TD, TE); TI = VFNMS(LDK(KP500000000), TH, TG); ST(&(x[WS(rs, 3)]), VADD(TF, TI), ms, &(x[WS(rs, 1)])); ST(&(x[0]), VADD(TG, TH), ms, &(x[0])); ST(&(x[WS(rs, 6)]), VSUB(TI, TF), ms, &(x[0])); } { V TC, Tv, Tw, Tx, Th, Tr, Ts, T7, TB; TC = VBYI(VSUB(VFMA(LDK(KP984807753), Tf, VFMA(LDK(KP813797681), Tq, VFNMS(LDK(KP150383733), Tg, VMUL(LDK(KP342020143), Tp)))), Tu)); Tv = VFMA(LDK(KP663413948), Tg, VMUL(LDK(KP642787609), Tf)); Tw = VFMA(LDK(KP150383733), Tq, VMUL(LDK(KP984807753), Tp)); Tx = VADD(Tv, Tw); Th = VFNMS(LDK(KP556670399), Tg, VMUL(LDK(KP766044443), Tf)); Tr = VFNMS(LDK(KP852868531), Tq, VMUL(LDK(KP173648177), Tp)); Ts = VADD(Th, Tr); T7 = VFNMS(LDK(KP500000000), T6, T1); TB = VFMA(LDK(KP852868531), Tg, VFMA(LDK(KP173648177), Tf, VFMA(LDK(KP296198132), Tq, VFNMS(LDK(KP939692620), Tp, T7)))); ST(&(x[WS(rs, 7)]), VSUB(TB, TC), ms, &(x[WS(rs, 1)])); ST(&(x[WS(rs, 2)]), VADD(TB, TC), ms, &(x[0])); { V Tt, Ty, Tz, TA; Tt = VADD(T7, Ts); Ty = VBYI(VADD(Tu, Tx)); ST(&(x[WS(rs, 8)]), VSUB(Tt, Ty), ms, &(x[0])); ST(&(x[WS(rs, 1)]), VADD(Tt, Ty), ms, &(x[WS(rs, 1)])); Tz = VBYI(VADD(Tu, VFNMS(LDK(KP500000000), Tx, VMUL(LDK(KP866025403), VSUB(Th, Tr))))); TA = VFMA(LDK(KP866025403), VSUB(Tw, Tv), VFNMS(LDK(KP500000000), Ts, T7)); ST(&(x[WS(rs, 4)]), VADD(Tz, TA), ms, &(x[0])); ST(&(x[WS(rs, 5)]), VSUB(TA, Tz), ms, &(x[WS(rs, 1)])); } } } } VLEAVE(); } static const tw_instr twinstr[] = { VTW(0, 1), VTW(0, 2), VTW(0, 3), VTW(0, 4), VTW(0, 5), VTW(0, 6), VTW(0, 7), VTW(0, 8), {TW_NEXT, VL, 0} }; static const ct_desc desc = { 9, XSIMD_STRING("t1bv_9"), twinstr, &GENUS, {38, 26, 16, 0}, 0, 0, 0 }; void XSIMD(codelet_t1bv_9) (planner *p) { X(kdft_dit_register) (p, t1bv_9, &desc); } #endif /* HAVE_FMA */ fftw-3.3.3/dft/simd/common/t1bv_10.c0000644000175000017500000002254312054410550013732 00000000000000/* * Copyright (c) 2003, 2007-11 Matteo Frigo * Copyright (c) 2003, 2007-11 Massachusetts Institute of Technology * * This program is free software; you can redistribute it and/or modify * it under the terms of the GNU General Public License as published by * the Free Software Foundation; either version 2 of the License, or * (at your option) any later version. * * This program is distributed in the hope that it will be useful, * but WITHOUT ANY WARRANTY; without even the implied warranty of * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the * GNU General Public License for more details. * * You should have received a copy of the GNU General Public License * along with this program; if not, write to the Free Software * Foundation, Inc., 51 Franklin Street, Fifth Floor, Boston, MA 02110-1301 USA * */ /* This file was automatically generated --- DO NOT EDIT */ /* Generated on Sun Nov 25 07:39:04 EST 2012 */ #include "codelet-dft.h" #ifdef HAVE_FMA /* Generated by: ../../../genfft/gen_twiddle_c.native -fma -reorder-insns -schedule-for-pipeline -simd -compact -variables 4 -pipeline-latency 8 -n 10 -name t1bv_10 -include t1b.h -sign 1 */ /* * This function contains 51 FP additions, 40 FP multiplications, * (or, 33 additions, 22 multiplications, 18 fused multiply/add), * 43 stack variables, 4 constants, and 20 memory accesses */ #include "t1b.h" static void t1bv_10(R *ri, R *ii, const R *W, stride rs, INT mb, INT me, INT ms) { DVK(KP559016994, +0.559016994374947424102293417182819058860154590); DVK(KP250000000, +0.250000000000000000000000000000000000000000000); DVK(KP618033988, +0.618033988749894848204586834365638117720309180); DVK(KP951056516, +0.951056516295153572116439333379382143405698634); { INT m; R *x; x = ii; for (m = mb, W = W + (mb * ((TWVL / VL) * 18)); m < me; m = m + VL, x = x + (VL * ms), W = W + (TWVL * 18), MAKE_VOLATILE_STRIDE(10, rs)) { V Td, TA, T4, Ta, Tk, TE, Tp, TF, TB, T9, T1, T2, Tb; T1 = LD(&(x[0]), ms, &(x[0])); T2 = LD(&(x[WS(rs, 5)]), ms, &(x[WS(rs, 1)])); { V Tg, Tn, Ti, Tl; Tg = LD(&(x[WS(rs, 4)]), ms, &(x[0])); Tn = LD(&(x[WS(rs, 1)]), ms, &(x[WS(rs, 1)])); Ti = LD(&(x[WS(rs, 9)]), ms, &(x[WS(rs, 1)])); Tl = LD(&(x[WS(rs, 6)]), ms, &(x[0])); { V T6, T8, T5, Tc; T5 = LD(&(x[WS(rs, 2)]), ms, &(x[0])); Tc = LD(&(x[WS(rs, 3)]), ms, &(x[WS(rs, 1)])); { V T3, Th, To, Tj, Tm, T7; T7 = LD(&(x[WS(rs, 7)]), ms, &(x[WS(rs, 1)])); T3 = BYTW(&(W[TWVL * 8]), T2); Th = BYTW(&(W[TWVL * 6]), Tg); To = BYTW(&(W[0]), Tn); Tj = BYTW(&(W[TWVL * 16]), Ti); Tm = BYTW(&(W[TWVL * 10]), Tl); T6 = BYTW(&(W[TWVL * 2]), T5); Td = BYTW(&(W[TWVL * 4]), Tc); T8 = BYTW(&(W[TWVL * 12]), T7); TA = VADD(T1, T3); T4 = VSUB(T1, T3); Ta = LD(&(x[WS(rs, 8)]), ms, &(x[0])); Tk = VSUB(Th, Tj); TE = VADD(Th, Tj); Tp = VSUB(Tm, To); TF = VADD(Tm, To); } TB = VADD(T6, T8); T9 = VSUB(T6, T8); } } Tb = BYTW(&(W[TWVL * 14]), Ta); { V TL, TG, Tw, Tq, TC, Te; TL = VSUB(TE, TF); TG = VADD(TE, TF); Tw = VSUB(Tk, Tp); Tq = VADD(Tk, Tp); TC = VADD(Tb, Td); Te = VSUB(Tb, Td); { V TM, TD, Tv, Tf; TM = VSUB(TB, TC); TD = VADD(TB, TC); Tv = VSUB(T9, Te); Tf = VADD(T9, Te); { V TP, TN, TH, TJ, Tz, Tx, Tr, Tt, TI, Ts; TP = VMUL(LDK(KP951056516), VFMA(LDK(KP618033988), TL, TM)); TN = VMUL(LDK(KP951056516), VFNMS(LDK(KP618033988), TM, TL)); TH = VADD(TD, TG); TJ = VSUB(TD, TG); Tz = VMUL(LDK(KP951056516), VFNMS(LDK(KP618033988), Tv, Tw)); Tx = VMUL(LDK(KP951056516), VFMA(LDK(KP618033988), Tw, Tv)); Tr = VADD(Tf, Tq); Tt = VSUB(Tf, Tq); ST(&(x[0]), VADD(TA, TH), ms, &(x[0])); TI = VFNMS(LDK(KP250000000), TH, TA); ST(&(x[WS(rs, 5)]), VADD(T4, Tr), ms, &(x[WS(rs, 1)])); Ts = VFNMS(LDK(KP250000000), Tr, T4); { V TK, TO, Tu, Ty; TK = VFNMS(LDK(KP559016994), TJ, TI); TO = VFMA(LDK(KP559016994), TJ, TI); Tu = VFMA(LDK(KP559016994), Tt, Ts); Ty = VFNMS(LDK(KP559016994), Tt, Ts); ST(&(x[WS(rs, 8)]), VFMAI(TN, TK), ms, &(x[0])); ST(&(x[WS(rs, 2)]), VFNMSI(TN, TK), ms, &(x[0])); ST(&(x[WS(rs, 6)]), VFMAI(TP, TO), ms, &(x[0])); ST(&(x[WS(rs, 4)]), VFNMSI(TP, TO), ms, &(x[0])); ST(&(x[WS(rs, 9)]), VFNMSI(Tx, Tu), ms, &(x[WS(rs, 1)])); ST(&(x[WS(rs, 1)]), VFMAI(Tx, Tu), ms, &(x[WS(rs, 1)])); ST(&(x[WS(rs, 7)]), VFNMSI(Tz, Ty), ms, &(x[WS(rs, 1)])); ST(&(x[WS(rs, 3)]), VFMAI(Tz, Ty), ms, &(x[WS(rs, 1)])); } } } } } } VLEAVE(); } static const tw_instr twinstr[] = { VTW(0, 1), VTW(0, 2), VTW(0, 3), VTW(0, 4), VTW(0, 5), VTW(0, 6), VTW(0, 7), VTW(0, 8), VTW(0, 9), {TW_NEXT, VL, 0} }; static const ct_desc desc = { 10, XSIMD_STRING("t1bv_10"), twinstr, &GENUS, {33, 22, 18, 0}, 0, 0, 0 }; void XSIMD(codelet_t1bv_10) (planner *p) { X(kdft_dit_register) (p, t1bv_10, &desc); } #else /* HAVE_FMA */ /* Generated by: ../../../genfft/gen_twiddle_c.native -simd -compact -variables 4 -pipeline-latency 8 -n 10 -name t1bv_10 -include t1b.h -sign 1 */ /* * This function contains 51 FP additions, 30 FP multiplications, * (or, 45 additions, 24 multiplications, 6 fused multiply/add), * 32 stack variables, 4 constants, and 20 memory accesses */ #include "t1b.h" static void t1bv_10(R *ri, R *ii, const R *W, stride rs, INT mb, INT me, INT ms) { DVK(KP587785252, +0.587785252292473129168705954639072768597652438); DVK(KP951056516, +0.951056516295153572116439333379382143405698634); DVK(KP250000000, +0.250000000000000000000000000000000000000000000); DVK(KP559016994, +0.559016994374947424102293417182819058860154590); { INT m; R *x; x = ii; for (m = mb, W = W + (mb * ((TWVL / VL) * 18)); m < me; m = m + VL, x = x + (VL * ms), W = W + (TWVL * 18), MAKE_VOLATILE_STRIDE(10, rs)) { V Tu, TH, Tg, Tl, Tp, TD, TE, TJ, T5, Ta, To, TA, TB, TI, Tr; V Tt, Ts; Tr = LD(&(x[0]), ms, &(x[0])); Ts = LD(&(x[WS(rs, 5)]), ms, &(x[WS(rs, 1)])); Tt = BYTW(&(W[TWVL * 8]), Ts); Tu = VSUB(Tr, Tt); TH = VADD(Tr, Tt); { V Td, Tk, Tf, Ti; { V Tc, Tj, Te, Th; Tc = LD(&(x[WS(rs, 4)]), ms, &(x[0])); Td = BYTW(&(W[TWVL * 6]), Tc); Tj = LD(&(x[WS(rs, 1)]), ms, &(x[WS(rs, 1)])); Tk = BYTW(&(W[0]), Tj); Te = LD(&(x[WS(rs, 9)]), ms, &(x[WS(rs, 1)])); Tf = BYTW(&(W[TWVL * 16]), Te); Th = LD(&(x[WS(rs, 6)]), ms, &(x[0])); Ti = BYTW(&(W[TWVL * 10]), Th); } Tg = VSUB(Td, Tf); Tl = VSUB(Ti, Tk); Tp = VADD(Tg, Tl); TD = VADD(Td, Tf); TE = VADD(Ti, Tk); TJ = VADD(TD, TE); } { V T2, T9, T4, T7; { V T1, T8, T3, T6; T1 = LD(&(x[WS(rs, 2)]), ms, &(x[0])); T2 = BYTW(&(W[TWVL * 2]), T1); T8 = LD(&(x[WS(rs, 3)]), ms, &(x[WS(rs, 1)])); T9 = BYTW(&(W[TWVL * 4]), T8); T3 = LD(&(x[WS(rs, 7)]), ms, &(x[WS(rs, 1)])); T4 = BYTW(&(W[TWVL * 12]), T3); T6 = LD(&(x[WS(rs, 8)]), ms, &(x[0])); T7 = BYTW(&(W[TWVL * 14]), T6); } T5 = VSUB(T2, T4); Ta = VSUB(T7, T9); To = VADD(T5, Ta); TA = VADD(T2, T4); TB = VADD(T7, T9); TI = VADD(TA, TB); } { V Tq, Tv, Tw, Tn, Tz, Tb, Tm, Ty, Tx; Tq = VMUL(LDK(KP559016994), VSUB(To, Tp)); Tv = VADD(To, Tp); Tw = VFNMS(LDK(KP250000000), Tv, Tu); Tb = VSUB(T5, Ta); Tm = VSUB(Tg, Tl); Tn = VBYI(VFMA(LDK(KP951056516), Tb, VMUL(LDK(KP587785252), Tm))); Tz = VBYI(VFNMS(LDK(KP951056516), Tm, VMUL(LDK(KP587785252), Tb))); ST(&(x[WS(rs, 5)]), VADD(Tu, Tv), ms, &(x[WS(rs, 1)])); Ty = VSUB(Tw, Tq); ST(&(x[WS(rs, 3)]), VSUB(Ty, Tz), ms, &(x[WS(rs, 1)])); ST(&(x[WS(rs, 7)]), VADD(Tz, Ty), ms, &(x[WS(rs, 1)])); Tx = VADD(Tq, Tw); ST(&(x[WS(rs, 1)]), VADD(Tn, Tx), ms, &(x[WS(rs, 1)])); ST(&(x[WS(rs, 9)]), VSUB(Tx, Tn), ms, &(x[WS(rs, 1)])); } { V TM, TK, TL, TG, TP, TC, TF, TO, TN; TM = VMUL(LDK(KP559016994), VSUB(TI, TJ)); TK = VADD(TI, TJ); TL = VFNMS(LDK(KP250000000), TK, TH); TC = VSUB(TA, TB); TF = VSUB(TD, TE); TG = VBYI(VFNMS(LDK(KP951056516), TF, VMUL(LDK(KP587785252), TC))); TP = VBYI(VFMA(LDK(KP951056516), TC, VMUL(LDK(KP587785252), TF))); ST(&(x[0]), VADD(TH, TK), ms, &(x[0])); TO = VADD(TM, TL); ST(&(x[WS(rs, 4)]), VSUB(TO, TP), ms, &(x[0])); ST(&(x[WS(rs, 6)]), VADD(TP, TO), ms, &(x[0])); TN = VSUB(TL, TM); ST(&(x[WS(rs, 2)]), VADD(TG, TN), ms, &(x[0])); ST(&(x[WS(rs, 8)]), VSUB(TN, TG), ms, &(x[0])); } } } VLEAVE(); } static const tw_instr twinstr[] = { VTW(0, 1), VTW(0, 2), VTW(0, 3), VTW(0, 4), VTW(0, 5), VTW(0, 6), VTW(0, 7), VTW(0, 8), VTW(0, 9), {TW_NEXT, VL, 0} }; static const ct_desc desc = { 10, XSIMD_STRING("t1bv_10"), twinstr, &GENUS, {45, 24, 6, 0}, 0, 0, 0 }; void XSIMD(codelet_t1bv_10) (planner *p) { X(kdft_dit_register) (p, t1bv_10, &desc); } #endif /* HAVE_FMA */ fftw-3.3.3/dft/simd/common/t1bv_12.c0000644000175000017500000002362512054410550013736 00000000000000/* * Copyright (c) 2003, 2007-11 Matteo Frigo * Copyright (c) 2003, 2007-11 Massachusetts Institute of Technology * * This program is free software; you can redistribute it and/or modify * it under the terms of the GNU General Public License as published by * the Free Software Foundation; either version 2 of the License, or * (at your option) any later version. * * This program is distributed in the hope that it will be useful, * but WITHOUT ANY WARRANTY; without even the implied warranty of * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the * GNU General Public License for more details. * * You should have received a copy of the GNU General Public License * along with this program; if not, write to the Free Software * Foundation, Inc., 51 Franklin Street, Fifth Floor, Boston, MA 02110-1301 USA * */ /* This file was automatically generated --- DO NOT EDIT */ /* Generated on Sun Nov 25 07:39:04 EST 2012 */ #include "codelet-dft.h" #ifdef HAVE_FMA /* Generated by: ../../../genfft/gen_twiddle_c.native -fma -reorder-insns -schedule-for-pipeline -simd -compact -variables 4 -pipeline-latency 8 -n 12 -name t1bv_12 -include t1b.h -sign 1 */ /* * This function contains 59 FP additions, 42 FP multiplications, * (or, 41 additions, 24 multiplications, 18 fused multiply/add), * 41 stack variables, 2 constants, and 24 memory accesses */ #include "t1b.h" static void t1bv_12(R *ri, R *ii, const R *W, stride rs, INT mb, INT me, INT ms) { DVK(KP866025403, +0.866025403784438646763723170752936183471402627); DVK(KP500000000, +0.500000000000000000000000000000000000000000000); { INT m; R *x; x = ii; for (m = mb, W = W + (mb * ((TWVL / VL) * 22)); m < me; m = m + VL, x = x + (VL * ms), W = W + (TWVL * 22), MAKE_VOLATILE_STRIDE(12, rs)) { V TI, Ti, TA, T7, Tm, TE, Tw, Tk, Tf, TB, TU, TM; { V T9, TK, Tj, TL, Te; { V T1, T4, T2, Tp, Tt, Tr; T1 = LD(&(x[0]), ms, &(x[0])); T4 = LD(&(x[WS(rs, 8)]), ms, &(x[0])); T2 = LD(&(x[WS(rs, 4)]), ms, &(x[0])); Tp = LD(&(x[WS(rs, 9)]), ms, &(x[WS(rs, 1)])); Tt = LD(&(x[WS(rs, 5)]), ms, &(x[WS(rs, 1)])); Tr = LD(&(x[WS(rs, 1)]), ms, &(x[WS(rs, 1)])); { V T5, T3, Tq, Tu, Ts, Td, Tb, T8, Tc, Ta; T8 = LD(&(x[WS(rs, 6)]), ms, &(x[0])); Tc = LD(&(x[WS(rs, 2)]), ms, &(x[0])); Ta = LD(&(x[WS(rs, 10)]), ms, &(x[0])); T5 = BYTW(&(W[TWVL * 14]), T4); T3 = BYTW(&(W[TWVL * 6]), T2); Tq = BYTW(&(W[TWVL * 16]), Tp); Tu = BYTW(&(W[TWVL * 8]), Tt); Ts = BYTW(&(W[0]), Tr); T9 = BYTW(&(W[TWVL * 10]), T8); Td = BYTW(&(W[TWVL * 2]), Tc); Tb = BYTW(&(W[TWVL * 18]), Ta); { V Th, T6, Tl, Tv; Th = LD(&(x[WS(rs, 3)]), ms, &(x[WS(rs, 1)])); TK = VSUB(T3, T5); T6 = VADD(T3, T5); Tl = LD(&(x[WS(rs, 11)]), ms, &(x[WS(rs, 1)])); Tv = VADD(Ts, Tu); TI = VSUB(Tu, Ts); Tj = LD(&(x[WS(rs, 7)]), ms, &(x[WS(rs, 1)])); TL = VSUB(Tb, Td); Te = VADD(Tb, Td); Ti = BYTW(&(W[TWVL * 4]), Th); TA = VFNMS(LDK(KP500000000), T6, T1); T7 = VADD(T1, T6); Tm = BYTW(&(W[TWVL * 20]), Tl); TE = VFNMS(LDK(KP500000000), Tv, Tq); Tw = VADD(Tq, Tv); } } } Tk = BYTW(&(W[TWVL * 12]), Tj); Tf = VADD(T9, Te); TB = VFNMS(LDK(KP500000000), Te, T9); TU = VSUB(TK, TL); TM = VADD(TK, TL); } { V Tn, TH, TC, TQ, Ty, Tg; Tn = VADD(Tk, Tm); TH = VSUB(Tk, Tm); TC = VADD(TA, TB); TQ = VSUB(TA, TB); Ty = VADD(T7, Tf); Tg = VSUB(T7, Tf); { V To, TD, TJ, TR; To = VADD(Ti, Tn); TD = VFNMS(LDK(KP500000000), Tn, Ti); TJ = VSUB(TH, TI); TR = VADD(TH, TI); { V TP, TN, TW, TS, TO, TG, TX, TV; { V Tz, Tx, TF, TT; Tz = VADD(To, Tw); Tx = VSUB(To, Tw); TF = VADD(TD, TE); TT = VSUB(TD, TE); TP = VMUL(LDK(KP866025403), VADD(TM, TJ)); TN = VMUL(LDK(KP866025403), VSUB(TJ, TM)); TW = VFMA(LDK(KP866025403), TR, TQ); TS = VFNMS(LDK(KP866025403), TR, TQ); ST(&(x[WS(rs, 6)]), VSUB(Ty, Tz), ms, &(x[0])); ST(&(x[0]), VADD(Ty, Tz), ms, &(x[0])); ST(&(x[WS(rs, 9)]), VFMAI(Tx, Tg), ms, &(x[WS(rs, 1)])); ST(&(x[WS(rs, 3)]), VFNMSI(Tx, Tg), ms, &(x[WS(rs, 1)])); TO = VADD(TC, TF); TG = VSUB(TC, TF); TX = VFNMS(LDK(KP866025403), TU, TT); TV = VFMA(LDK(KP866025403), TU, TT); } ST(&(x[WS(rs, 8)]), VFNMSI(TP, TO), ms, &(x[0])); ST(&(x[WS(rs, 4)]), VFMAI(TP, TO), ms, &(x[0])); ST(&(x[WS(rs, 2)]), VFMAI(TN, TG), ms, &(x[0])); ST(&(x[WS(rs, 10)]), VFNMSI(TN, TG), ms, &(x[0])); ST(&(x[WS(rs, 5)]), VFMAI(TX, TW), ms, &(x[WS(rs, 1)])); ST(&(x[WS(rs, 7)]), VFNMSI(TX, TW), ms, &(x[WS(rs, 1)])); ST(&(x[WS(rs, 11)]), VFNMSI(TV, TS), ms, &(x[WS(rs, 1)])); ST(&(x[WS(rs, 1)]), VFMAI(TV, TS), ms, &(x[WS(rs, 1)])); } } } } } VLEAVE(); } static const tw_instr twinstr[] = { VTW(0, 1), VTW(0, 2), VTW(0, 3), VTW(0, 4), VTW(0, 5), VTW(0, 6), VTW(0, 7), VTW(0, 8), VTW(0, 9), VTW(0, 10), VTW(0, 11), {TW_NEXT, VL, 0} }; static const ct_desc desc = { 12, XSIMD_STRING("t1bv_12"), twinstr, &GENUS, {41, 24, 18, 0}, 0, 0, 0 }; void XSIMD(codelet_t1bv_12) (planner *p) { X(kdft_dit_register) (p, t1bv_12, &desc); } #else /* HAVE_FMA */ /* Generated by: ../../../genfft/gen_twiddle_c.native -simd -compact -variables 4 -pipeline-latency 8 -n 12 -name t1bv_12 -include t1b.h -sign 1 */ /* * This function contains 59 FP additions, 30 FP multiplications, * (or, 55 additions, 26 multiplications, 4 fused multiply/add), * 28 stack variables, 2 constants, and 24 memory accesses */ #include "t1b.h" static void t1bv_12(R *ri, R *ii, const R *W, stride rs, INT mb, INT me, INT ms) { DVK(KP866025403, +0.866025403784438646763723170752936183471402627); DVK(KP500000000, +0.500000000000000000000000000000000000000000000); { INT m; R *x; x = ii; for (m = mb, W = W + (mb * ((TWVL / VL) * 22)); m < me; m = m + VL, x = x + (VL * ms), W = W + (TWVL * 22), MAKE_VOLATILE_STRIDE(12, rs)) { V T1, Tt, T6, T7, TB, Tq, TC, TD, T9, Tu, Te, Tf, Tx, Tl, Ty; V Tz; { V T5, T3, T4, T2; T1 = LD(&(x[0]), ms, &(x[0])); T4 = LD(&(x[WS(rs, 8)]), ms, &(x[0])); T5 = BYTW(&(W[TWVL * 14]), T4); T2 = LD(&(x[WS(rs, 4)]), ms, &(x[0])); T3 = BYTW(&(W[TWVL * 6]), T2); Tt = VSUB(T3, T5); T6 = VADD(T3, T5); T7 = VFNMS(LDK(KP500000000), T6, T1); } { V Tn, Tp, Tm, TA, To; Tm = LD(&(x[WS(rs, 1)]), ms, &(x[WS(rs, 1)])); Tn = BYTW(&(W[0]), Tm); TA = LD(&(x[WS(rs, 9)]), ms, &(x[WS(rs, 1)])); TB = BYTW(&(W[TWVL * 16]), TA); To = LD(&(x[WS(rs, 5)]), ms, &(x[WS(rs, 1)])); Tp = BYTW(&(W[TWVL * 8]), To); Tq = VSUB(Tn, Tp); TC = VADD(Tn, Tp); TD = VFNMS(LDK(KP500000000), TC, TB); } { V Td, Tb, T8, Tc, Ta; T8 = LD(&(x[WS(rs, 6)]), ms, &(x[0])); T9 = BYTW(&(W[TWVL * 10]), T8); Tc = LD(&(x[WS(rs, 2)]), ms, &(x[0])); Td = BYTW(&(W[TWVL * 2]), Tc); Ta = LD(&(x[WS(rs, 10)]), ms, &(x[0])); Tb = BYTW(&(W[TWVL * 18]), Ta); Tu = VSUB(Tb, Td); Te = VADD(Tb, Td); Tf = VFNMS(LDK(KP500000000), Te, T9); } { V Ti, Tk, Th, Tw, Tj; Th = LD(&(x[WS(rs, 7)]), ms, &(x[WS(rs, 1)])); Ti = BYTW(&(W[TWVL * 12]), Th); Tw = LD(&(x[WS(rs, 3)]), ms, &(x[WS(rs, 1)])); Tx = BYTW(&(W[TWVL * 4]), Tw); Tj = LD(&(x[WS(rs, 11)]), ms, &(x[WS(rs, 1)])); Tk = BYTW(&(W[TWVL * 20]), Tj); Tl = VSUB(Ti, Tk); Ty = VADD(Ti, Tk); Tz = VFNMS(LDK(KP500000000), Ty, Tx); } { V Ts, TG, TF, TH; { V Tg, Tr, Tv, TE; Tg = VSUB(T7, Tf); Tr = VMUL(LDK(KP866025403), VSUB(Tl, Tq)); Ts = VSUB(Tg, Tr); TG = VADD(Tg, Tr); Tv = VMUL(LDK(KP866025403), VSUB(Tt, Tu)); TE = VSUB(Tz, TD); TF = VBYI(VADD(Tv, TE)); TH = VBYI(VSUB(TE, Tv)); } ST(&(x[WS(rs, 11)]), VSUB(Ts, TF), ms, &(x[WS(rs, 1)])); ST(&(x[WS(rs, 5)]), VADD(TG, TH), ms, &(x[WS(rs, 1)])); ST(&(x[WS(rs, 1)]), VADD(Ts, TF), ms, &(x[WS(rs, 1)])); ST(&(x[WS(rs, 7)]), VSUB(TG, TH), ms, &(x[WS(rs, 1)])); } { V TS, TW, TV, TX; { V TQ, TR, TT, TU; TQ = VADD(T1, T6); TR = VADD(T9, Te); TS = VSUB(TQ, TR); TW = VADD(TQ, TR); TT = VADD(Tx, Ty); TU = VADD(TB, TC); TV = VBYI(VSUB(TT, TU)); TX = VADD(TT, TU); } ST(&(x[WS(rs, 3)]), VSUB(TS, TV), ms, &(x[WS(rs, 1)])); ST(&(x[0]), VADD(TW, TX), ms, &(x[0])); ST(&(x[WS(rs, 9)]), VADD(TS, TV), ms, &(x[WS(rs, 1)])); ST(&(x[WS(rs, 6)]), VSUB(TW, TX), ms, &(x[0])); } { V TK, TO, TN, TP; { V TI, TJ, TL, TM; TI = VADD(Tl, Tq); TJ = VADD(Tt, Tu); TK = VBYI(VMUL(LDK(KP866025403), VSUB(TI, TJ))); TO = VBYI(VMUL(LDK(KP866025403), VADD(TJ, TI))); TL = VADD(T7, Tf); TM = VADD(Tz, TD); TN = VSUB(TL, TM); TP = VADD(TL, TM); } ST(&(x[WS(rs, 2)]), VADD(TK, TN), ms, &(x[0])); ST(&(x[WS(rs, 8)]), VSUB(TP, TO), ms, &(x[0])); ST(&(x[WS(rs, 10)]), VSUB(TN, TK), ms, &(x[0])); ST(&(x[WS(rs, 4)]), VADD(TO, TP), ms, &(x[0])); } } } VLEAVE(); } static const tw_instr twinstr[] = { VTW(0, 1), VTW(0, 2), VTW(0, 3), VTW(0, 4), VTW(0, 5), VTW(0, 6), VTW(0, 7), VTW(0, 8), VTW(0, 9), VTW(0, 10), VTW(0, 11), {TW_NEXT, VL, 0} }; static const ct_desc desc = { 12, XSIMD_STRING("t1bv_12"), twinstr, &GENUS, {55, 26, 4, 0}, 0, 0, 0 }; void XSIMD(codelet_t1bv_12) (planner *p) { X(kdft_dit_register) (p, t1bv_12, &desc); } #endif /* HAVE_FMA */ fftw-3.3.3/dft/simd/common/t1bv_15.c0000644000175000017500000003522412054410551013740 00000000000000/* * Copyright (c) 2003, 2007-11 Matteo Frigo * Copyright (c) 2003, 2007-11 Massachusetts Institute of Technology * * This program is free software; you can redistribute it and/or modify * it under the terms of the GNU General Public License as published by * the Free Software Foundation; either version 2 of the License, or * (at your option) any later version. * * This program is distributed in the hope that it will be useful, * but WITHOUT ANY WARRANTY; without even the implied warranty of * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the * GNU General Public License for more details. * * You should have received a copy of the GNU General Public License * along with this program; if not, write to the Free Software * Foundation, Inc., 51 Franklin Street, Fifth Floor, Boston, MA 02110-1301 USA * */ /* This file was automatically generated --- DO NOT EDIT */ /* Generated on Sun Nov 25 07:39:04 EST 2012 */ #include "codelet-dft.h" #ifdef HAVE_FMA /* Generated by: ../../../genfft/gen_twiddle_c.native -fma -reorder-insns -schedule-for-pipeline -simd -compact -variables 4 -pipeline-latency 8 -n 15 -name t1bv_15 -include t1b.h -sign 1 */ /* * This function contains 92 FP additions, 77 FP multiplications, * (or, 50 additions, 35 multiplications, 42 fused multiply/add), * 81 stack variables, 8 constants, and 30 memory accesses */ #include "t1b.h" static void t1bv_15(R *ri, R *ii, const R *W, stride rs, INT mb, INT me, INT ms) { DVK(KP823639103, +0.823639103546331925877420039278190003029660514); DVK(KP910592997, +0.910592997310029334643087372129977886038870291); DVK(KP559016994, +0.559016994374947424102293417182819058860154590); DVK(KP951056516, +0.951056516295153572116439333379382143405698634); DVK(KP866025403, +0.866025403784438646763723170752936183471402627); DVK(KP250000000, +0.250000000000000000000000000000000000000000000); DVK(KP618033988, +0.618033988749894848204586834365638117720309180); DVK(KP500000000, +0.500000000000000000000000000000000000000000000); { INT m; R *x; x = ii; for (m = mb, W = W + (mb * ((TWVL / VL) * 28)); m < me; m = m + VL, x = x + (VL * ms), W = W + (TWVL * 28), MAKE_VOLATILE_STRIDE(15, rs)) { V Tq, Ty, Th, TV, TK, Ts, T1f, T7, Tu, TA, TC, Tj, Tk, T1g, Tf; { V T1, T4, T2, T9, Te; T1 = LD(&(x[0]), ms, &(x[0])); T4 = LD(&(x[WS(rs, 10)]), ms, &(x[0])); T2 = LD(&(x[WS(rs, 5)]), ms, &(x[WS(rs, 1)])); { V T8, Tp, Tx, Tg; T8 = LD(&(x[WS(rs, 3)]), ms, &(x[WS(rs, 1)])); Tp = LD(&(x[WS(rs, 6)]), ms, &(x[0])); Tx = LD(&(x[WS(rs, 9)]), ms, &(x[WS(rs, 1)])); Tg = LD(&(x[WS(rs, 12)]), ms, &(x[0])); { V Tb, Td, Tr, T6, Tt, Tz, TB, Ti; { V T5, T3, Ta, Tc; Ta = LD(&(x[WS(rs, 8)]), ms, &(x[0])); Tc = LD(&(x[WS(rs, 13)]), ms, &(x[WS(rs, 1)])); T5 = BYTW(&(W[TWVL * 18]), T4); T3 = BYTW(&(W[TWVL * 8]), T2); T9 = BYTW(&(W[TWVL * 4]), T8); Tq = BYTW(&(W[TWVL * 10]), Tp); Ty = BYTW(&(W[TWVL * 16]), Tx); Th = BYTW(&(W[TWVL * 22]), Tg); Tb = BYTW(&(W[TWVL * 14]), Ta); Td = BYTW(&(W[TWVL * 24]), Tc); Tr = LD(&(x[WS(rs, 11)]), ms, &(x[WS(rs, 1)])); TV = VSUB(T3, T5); T6 = VADD(T3, T5); Tt = LD(&(x[WS(rs, 1)]), ms, &(x[WS(rs, 1)])); } Tz = LD(&(x[WS(rs, 14)]), ms, &(x[0])); TB = LD(&(x[WS(rs, 4)]), ms, &(x[0])); Ti = LD(&(x[WS(rs, 2)]), ms, &(x[0])); Te = VADD(Tb, Td); TK = VSUB(Tb, Td); Ts = BYTW(&(W[TWVL * 20]), Tr); T1f = VADD(T1, T6); T7 = VFNMS(LDK(KP500000000), T6, T1); Tu = BYTW(&(W[0]), Tt); TA = BYTW(&(W[TWVL * 26]), Tz); TC = BYTW(&(W[TWVL * 6]), TB); Tj = BYTW(&(W[TWVL * 2]), Ti); Tk = LD(&(x[WS(rs, 7)]), ms, &(x[WS(rs, 1)])); } } T1g = VADD(T9, Te); Tf = VFNMS(LDK(KP500000000), Te, T9); } { V Tv, TN, TD, TO, Tl; Tv = VADD(Ts, Tu); TN = VSUB(Ts, Tu); TD = VADD(TA, TC); TO = VSUB(TA, TC); Tl = BYTW(&(W[TWVL * 12]), Tk); { V Tw, T1j, TX, TP, TE, T1k, TL, Tm; Tw = VFNMS(LDK(KP500000000), Tv, Tq); T1j = VADD(Tq, Tv); TX = VADD(TN, TO); TP = VSUB(TN, TO); TE = VFNMS(LDK(KP500000000), TD, Ty); T1k = VADD(Ty, TD); TL = VSUB(Tj, Tl); Tm = VADD(Tj, Tl); { V TT, TF, T1q, T1l, TW, TM, T1h, Tn; TT = VSUB(Tw, TE); TF = VADD(Tw, TE); T1q = VSUB(T1j, T1k); T1l = VADD(T1j, T1k); TW = VADD(TK, TL); TM = VSUB(TK, TL); T1h = VADD(Th, Tm); Tn = VFNMS(LDK(KP500000000), Tm, Th); { V T10, TY, T16, TQ, T1r, T1i, TS, To, TZ, T1e; T10 = VSUB(TW, TX); TY = VADD(TW, TX); T16 = VFNMS(LDK(KP618033988), TM, TP); TQ = VFMA(LDK(KP618033988), TP, TM); T1r = VSUB(T1g, T1h); T1i = VADD(T1g, T1h); TS = VSUB(Tf, Tn); To = VADD(Tf, Tn); TZ = VFNMS(LDK(KP250000000), TY, TV); T1e = VMUL(LDK(KP866025403), VADD(TV, TY)); { V T1u, T1s, T1o, T18, TU, TG, TI, T19, T11, T1n, T1m; T1u = VMUL(LDK(KP951056516), VFMA(LDK(KP618033988), T1q, T1r)); T1s = VMUL(LDK(KP951056516), VFNMS(LDK(KP618033988), T1r, T1q)); T1m = VADD(T1i, T1l); T1o = VSUB(T1i, T1l); T18 = VFNMS(LDK(KP618033988), TS, TT); TU = VFMA(LDK(KP618033988), TT, TS); TG = VADD(To, TF); TI = VSUB(To, TF); T19 = VFNMS(LDK(KP559016994), T10, TZ); T11 = VFMA(LDK(KP559016994), T10, TZ); ST(&(x[0]), VADD(T1f, T1m), ms, &(x[0])); T1n = VFNMS(LDK(KP250000000), T1m, T1f); { V T1a, T1c, T14, T12, T1p, T1t, T15, TJ, T1d, TH; T1d = VADD(T7, TG); TH = VFNMS(LDK(KP250000000), TG, T7); T1a = VMUL(LDK(KP951056516), VFMA(LDK(KP910592997), T19, T18)); T1c = VMUL(LDK(KP951056516), VFNMS(LDK(KP910592997), T19, T18)); T14 = VMUL(LDK(KP951056516), VFNMS(LDK(KP910592997), T11, TU)); T12 = VMUL(LDK(KP951056516), VFMA(LDK(KP910592997), T11, TU)); T1p = VFNMS(LDK(KP559016994), T1o, T1n); T1t = VFMA(LDK(KP559016994), T1o, T1n); ST(&(x[WS(rs, 10)]), VFMAI(T1e, T1d), ms, &(x[0])); ST(&(x[WS(rs, 5)]), VFNMSI(T1e, T1d), ms, &(x[WS(rs, 1)])); T15 = VFNMS(LDK(KP559016994), TI, TH); TJ = VFMA(LDK(KP559016994), TI, TH); { V T17, T1b, T13, TR; ST(&(x[WS(rs, 12)]), VFNMSI(T1s, T1p), ms, &(x[0])); ST(&(x[WS(rs, 3)]), VFMAI(T1s, T1p), ms, &(x[WS(rs, 1)])); ST(&(x[WS(rs, 9)]), VFNMSI(T1u, T1t), ms, &(x[WS(rs, 1)])); ST(&(x[WS(rs, 6)]), VFMAI(T1u, T1t), ms, &(x[0])); T17 = VFNMS(LDK(KP823639103), T16, T15); T1b = VFMA(LDK(KP823639103), T16, T15); T13 = VFMA(LDK(KP823639103), TQ, TJ); TR = VFNMS(LDK(KP823639103), TQ, TJ); ST(&(x[WS(rs, 13)]), VFMAI(T1a, T17), ms, &(x[WS(rs, 1)])); ST(&(x[WS(rs, 2)]), VFNMSI(T1a, T17), ms, &(x[0])); ST(&(x[WS(rs, 8)]), VFMAI(T1c, T1b), ms, &(x[0])); ST(&(x[WS(rs, 7)]), VFNMSI(T1c, T1b), ms, &(x[WS(rs, 1)])); ST(&(x[WS(rs, 11)]), VFMAI(T14, T13), ms, &(x[WS(rs, 1)])); ST(&(x[WS(rs, 4)]), VFNMSI(T14, T13), ms, &(x[0])); ST(&(x[WS(rs, 14)]), VFNMSI(T12, TR), ms, &(x[0])); ST(&(x[WS(rs, 1)]), VFMAI(T12, TR), ms, &(x[WS(rs, 1)])); } } } } } } } } } VLEAVE(); } static const tw_instr twinstr[] = { VTW(0, 1), VTW(0, 2), VTW(0, 3), VTW(0, 4), VTW(0, 5), VTW(0, 6), VTW(0, 7), VTW(0, 8), VTW(0, 9), VTW(0, 10), VTW(0, 11), VTW(0, 12), VTW(0, 13), VTW(0, 14), {TW_NEXT, VL, 0} }; static const ct_desc desc = { 15, XSIMD_STRING("t1bv_15"), twinstr, &GENUS, {50, 35, 42, 0}, 0, 0, 0 }; void XSIMD(codelet_t1bv_15) (planner *p) { X(kdft_dit_register) (p, t1bv_15, &desc); } #else /* HAVE_FMA */ /* Generated by: ../../../genfft/gen_twiddle_c.native -simd -compact -variables 4 -pipeline-latency 8 -n 15 -name t1bv_15 -include t1b.h -sign 1 */ /* * This function contains 92 FP additions, 53 FP multiplications, * (or, 78 additions, 39 multiplications, 14 fused multiply/add), * 52 stack variables, 10 constants, and 30 memory accesses */ #include "t1b.h" static void t1bv_15(R *ri, R *ii, const R *W, stride rs, INT mb, INT me, INT ms) { DVK(KP216506350, +0.216506350946109661690930792688234045867850657); DVK(KP484122918, +0.484122918275927110647408174972799951354115213); DVK(KP866025403, +0.866025403784438646763723170752936183471402627); DVK(KP509036960, +0.509036960455127183450980863393907648510733164); DVK(KP823639103, +0.823639103546331925877420039278190003029660514); DVK(KP951056516, +0.951056516295153572116439333379382143405698634); DVK(KP587785252, +0.587785252292473129168705954639072768597652438); DVK(KP250000000, +0.250000000000000000000000000000000000000000000); DVK(KP559016994, +0.559016994374947424102293417182819058860154590); DVK(KP500000000, +0.500000000000000000000000000000000000000000000); { INT m; R *x; x = ii; for (m = mb, W = W + (mb * ((TWVL / VL) * 28)); m < me; m = m + VL, x = x + (VL * ms), W = W + (TWVL * 28), MAKE_VOLATILE_STRIDE(15, rs)) { V Ts, TV, T1f, TZ, T10, Tb, Tm, Tt, T1j, T1k, T1l, TI, TM, TR, Tz; V TD, TQ, T1g, T1h, T1i; { V TT, Tr, Tp, Tq, To, TU; TT = LD(&(x[0]), ms, &(x[0])); Tq = LD(&(x[WS(rs, 10)]), ms, &(x[0])); Tr = BYTW(&(W[TWVL * 18]), Tq); To = LD(&(x[WS(rs, 5)]), ms, &(x[WS(rs, 1)])); Tp = BYTW(&(W[TWVL * 8]), To); Ts = VSUB(Tp, Tr); TU = VADD(Tp, Tr); TV = VFNMS(LDK(KP500000000), TU, TT); T1f = VADD(TT, TU); } { V Tx, TG, TK, TB, T5, Ty, Tg, TH, Tl, TL, Ta, TC; { V Tw, TF, TJ, TA; Tw = LD(&(x[WS(rs, 3)]), ms, &(x[WS(rs, 1)])); Tx = BYTW(&(W[TWVL * 4]), Tw); TF = LD(&(x[WS(rs, 6)]), ms, &(x[0])); TG = BYTW(&(W[TWVL * 10]), TF); TJ = LD(&(x[WS(rs, 9)]), ms, &(x[WS(rs, 1)])); TK = BYTW(&(W[TWVL * 16]), TJ); TA = LD(&(x[WS(rs, 12)]), ms, &(x[0])); TB = BYTW(&(W[TWVL * 22]), TA); } { V T2, T4, T1, T3; T1 = LD(&(x[WS(rs, 8)]), ms, &(x[0])); T2 = BYTW(&(W[TWVL * 14]), T1); T3 = LD(&(x[WS(rs, 13)]), ms, &(x[WS(rs, 1)])); T4 = BYTW(&(W[TWVL * 24]), T3); T5 = VSUB(T2, T4); Ty = VADD(T2, T4); } { V Td, Tf, Tc, Te; Tc = LD(&(x[WS(rs, 11)]), ms, &(x[WS(rs, 1)])); Td = BYTW(&(W[TWVL * 20]), Tc); Te = LD(&(x[WS(rs, 1)]), ms, &(x[WS(rs, 1)])); Tf = BYTW(&(W[0]), Te); Tg = VSUB(Td, Tf); TH = VADD(Td, Tf); } { V Ti, Tk, Th, Tj; Th = LD(&(x[WS(rs, 14)]), ms, &(x[0])); Ti = BYTW(&(W[TWVL * 26]), Th); Tj = LD(&(x[WS(rs, 4)]), ms, &(x[0])); Tk = BYTW(&(W[TWVL * 6]), Tj); Tl = VSUB(Ti, Tk); TL = VADD(Ti, Tk); } { V T7, T9, T6, T8; T6 = LD(&(x[WS(rs, 2)]), ms, &(x[0])); T7 = BYTW(&(W[TWVL * 2]), T6); T8 = LD(&(x[WS(rs, 7)]), ms, &(x[WS(rs, 1)])); T9 = BYTW(&(W[TWVL * 12]), T8); Ta = VSUB(T7, T9); TC = VADD(T7, T9); } TZ = VSUB(T5, Ta); T10 = VSUB(Tg, Tl); Tb = VADD(T5, Ta); Tm = VADD(Tg, Tl); Tt = VADD(Tb, Tm); T1j = VADD(TG, TH); T1k = VADD(TK, TL); T1l = VADD(T1j, T1k); TI = VFNMS(LDK(KP500000000), TH, TG); TM = VFNMS(LDK(KP500000000), TL, TK); TR = VADD(TI, TM); Tz = VFNMS(LDK(KP500000000), Ty, Tx); TD = VFNMS(LDK(KP500000000), TC, TB); TQ = VADD(Tz, TD); T1g = VADD(Tx, Ty); T1h = VADD(TB, TC); T1i = VADD(T1g, T1h); } { V T1o, T1m, T1n, T1s, T1t, T1q, T1r, T1u, T1p; T1o = VMUL(LDK(KP559016994), VSUB(T1i, T1l)); T1m = VADD(T1i, T1l); T1n = VFNMS(LDK(KP250000000), T1m, T1f); T1q = VSUB(T1g, T1h); T1r = VSUB(T1j, T1k); T1s = VBYI(VFNMS(LDK(KP951056516), T1r, VMUL(LDK(KP587785252), T1q))); T1t = VBYI(VFMA(LDK(KP951056516), T1q, VMUL(LDK(KP587785252), T1r))); ST(&(x[0]), VADD(T1f, T1m), ms, &(x[0])); T1u = VADD(T1o, T1n); ST(&(x[WS(rs, 6)]), VADD(T1t, T1u), ms, &(x[0])); ST(&(x[WS(rs, 9)]), VSUB(T1u, T1t), ms, &(x[WS(rs, 1)])); T1p = VSUB(T1n, T1o); ST(&(x[WS(rs, 3)]), VSUB(T1p, T1s), ms, &(x[WS(rs, 1)])); ST(&(x[WS(rs, 12)]), VADD(T1s, T1p), ms, &(x[0])); } { V T11, T18, T1e, TO, T16, Tv, T15, TY, T1d, T19, TE, TN; T11 = VFMA(LDK(KP823639103), TZ, VMUL(LDK(KP509036960), T10)); T18 = VFNMS(LDK(KP823639103), T10, VMUL(LDK(KP509036960), TZ)); T1e = VBYI(VMUL(LDK(KP866025403), VADD(Ts, Tt))); TE = VSUB(Tz, TD); TN = VSUB(TI, TM); TO = VFMA(LDK(KP951056516), TE, VMUL(LDK(KP587785252), TN)); T16 = VFNMS(LDK(KP951056516), TN, VMUL(LDK(KP587785252), TE)); { V Tn, Tu, TS, TW, TX; Tn = VMUL(LDK(KP484122918), VSUB(Tb, Tm)); Tu = VFNMS(LDK(KP216506350), Tt, VMUL(LDK(KP866025403), Ts)); Tv = VADD(Tn, Tu); T15 = VSUB(Tn, Tu); TS = VMUL(LDK(KP559016994), VSUB(TQ, TR)); TW = VADD(TQ, TR); TX = VFNMS(LDK(KP250000000), TW, TV); TY = VADD(TS, TX); T1d = VADD(TV, TW); T19 = VSUB(TX, TS); } { V TP, T12, T1b, T1c; ST(&(x[WS(rs, 5)]), VSUB(T1d, T1e), ms, &(x[WS(rs, 1)])); ST(&(x[WS(rs, 10)]), VADD(T1e, T1d), ms, &(x[0])); TP = VBYI(VADD(Tv, TO)); T12 = VSUB(TY, T11); ST(&(x[WS(rs, 1)]), VADD(TP, T12), ms, &(x[WS(rs, 1)])); ST(&(x[WS(rs, 14)]), VSUB(T12, TP), ms, &(x[0])); T1b = VBYI(VSUB(T16, T15)); T1c = VSUB(T19, T18); ST(&(x[WS(rs, 7)]), VADD(T1b, T1c), ms, &(x[WS(rs, 1)])); ST(&(x[WS(rs, 8)]), VSUB(T1c, T1b), ms, &(x[0])); { V T17, T1a, T13, T14; T17 = VBYI(VADD(T15, T16)); T1a = VADD(T18, T19); ST(&(x[WS(rs, 2)]), VADD(T17, T1a), ms, &(x[0])); ST(&(x[WS(rs, 13)]), VSUB(T1a, T17), ms, &(x[WS(rs, 1)])); T13 = VBYI(VSUB(Tv, TO)); T14 = VADD(T11, TY); ST(&(x[WS(rs, 4)]), VADD(T13, T14), ms, &(x[0])); ST(&(x[WS(rs, 11)]), VSUB(T14, T13), ms, &(x[WS(rs, 1)])); } } } } } VLEAVE(); } static const tw_instr twinstr[] = { VTW(0, 1), VTW(0, 2), VTW(0, 3), VTW(0, 4), VTW(0, 5), VTW(0, 6), VTW(0, 7), VTW(0, 8), VTW(0, 9), VTW(0, 10), VTW(0, 11), VTW(0, 12), VTW(0, 13), VTW(0, 14), {TW_NEXT, VL, 0} }; static const ct_desc desc = { 15, XSIMD_STRING("t1bv_15"), twinstr, &GENUS, {78, 39, 14, 0}, 0, 0, 0 }; void XSIMD(codelet_t1bv_15) (planner *p) { X(kdft_dit_register) (p, t1bv_15, &desc); } #endif /* HAVE_FMA */ fftw-3.3.3/dft/simd/common/t1bv_16.c0000644000175000017500000003217712054410551013745 00000000000000/* * Copyright (c) 2003, 2007-11 Matteo Frigo * Copyright (c) 2003, 2007-11 Massachusetts Institute of Technology * * This program is free software; you can redistribute it and/or modify * it under the terms of the GNU General Public License as published by * the Free Software Foundation; either version 2 of the License, or * (at your option) any later version. * * This program is distributed in the hope that it will be useful, * but WITHOUT ANY WARRANTY; without even the implied warranty of * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the * GNU General Public License for more details. * * You should have received a copy of the GNU General Public License * along with this program; if not, write to the Free Software * Foundation, Inc., 51 Franklin Street, Fifth Floor, Boston, MA 02110-1301 USA * */ /* This file was automatically generated --- DO NOT EDIT */ /* Generated on Sun Nov 25 07:39:04 EST 2012 */ #include "codelet-dft.h" #ifdef HAVE_FMA /* Generated by: ../../../genfft/gen_twiddle_c.native -fma -reorder-insns -schedule-for-pipeline -simd -compact -variables 4 -pipeline-latency 8 -n 16 -name t1bv_16 -include t1b.h -sign 1 */ /* * This function contains 87 FP additions, 64 FP multiplications, * (or, 53 additions, 30 multiplications, 34 fused multiply/add), * 61 stack variables, 3 constants, and 32 memory accesses */ #include "t1b.h" static void t1bv_16(R *ri, R *ii, const R *W, stride rs, INT mb, INT me, INT ms) { DVK(KP923879532, +0.923879532511286756128183189396788286822416626); DVK(KP414213562, +0.414213562373095048801688724209698078569671875); DVK(KP707106781, +0.707106781186547524400844362104849039284835938); { INT m; R *x; x = ii; for (m = mb, W = W + (mb * ((TWVL / VL) * 30)); m < me; m = m + VL, x = x + (VL * ms), W = W + (TWVL * 30), MAKE_VOLATILE_STRIDE(16, rs)) { V TO, Ta, TJ, TP, T14, Tq, T1i, T10, T1b, T1l, T13, T1c, TR, Tl, T15; V Tv; { V Tc, TW, T4, T19, T9, TD, TI, Tj, TZ, T1a, Te, Th, Tn, Tr, Tu; V Tp; { V T1, T2, T5, T7; T1 = LD(&(x[0]), ms, &(x[0])); T2 = LD(&(x[WS(rs, 8)]), ms, &(x[0])); T5 = LD(&(x[WS(rs, 4)]), ms, &(x[0])); T7 = LD(&(x[WS(rs, 12)]), ms, &(x[0])); { V Tz, TG, TB, TE; Tz = LD(&(x[WS(rs, 2)]), ms, &(x[0])); TG = LD(&(x[WS(rs, 6)]), ms, &(x[0])); TB = LD(&(x[WS(rs, 10)]), ms, &(x[0])); TE = LD(&(x[WS(rs, 14)]), ms, &(x[0])); { V Ti, TX, TY, Td, Tg, Tm, Tt, To; { V T3, T6, T8, TA, TH, TC, TF, Tb; Tb = LD(&(x[WS(rs, 1)]), ms, &(x[WS(rs, 1)])); T3 = BYTW(&(W[TWVL * 14]), T2); T6 = BYTW(&(W[TWVL * 6]), T5); T8 = BYTW(&(W[TWVL * 22]), T7); TA = BYTW(&(W[TWVL * 2]), Tz); TH = BYTW(&(W[TWVL * 10]), TG); TC = BYTW(&(W[TWVL * 18]), TB); TF = BYTW(&(W[TWVL * 26]), TE); Tc = BYTW(&(W[0]), Tb); TW = VSUB(T1, T3); T4 = VADD(T1, T3); T19 = VSUB(T6, T8); T9 = VADD(T6, T8); Ti = LD(&(x[WS(rs, 13)]), ms, &(x[WS(rs, 1)])); TD = VADD(TA, TC); TX = VSUB(TA, TC); TI = VADD(TF, TH); TY = VSUB(TF, TH); } Td = LD(&(x[WS(rs, 9)]), ms, &(x[WS(rs, 1)])); Tg = LD(&(x[WS(rs, 5)]), ms, &(x[WS(rs, 1)])); Tm = LD(&(x[WS(rs, 15)]), ms, &(x[WS(rs, 1)])); Tj = BYTW(&(W[TWVL * 24]), Ti); Tt = LD(&(x[WS(rs, 11)]), ms, &(x[WS(rs, 1)])); To = LD(&(x[WS(rs, 7)]), ms, &(x[WS(rs, 1)])); TZ = VADD(TX, TY); T1a = VSUB(TX, TY); Te = BYTW(&(W[TWVL * 16]), Td); Th = BYTW(&(W[TWVL * 8]), Tg); Tn = BYTW(&(W[TWVL * 28]), Tm); Tr = LD(&(x[WS(rs, 3)]), ms, &(x[WS(rs, 1)])); Tu = BYTW(&(W[TWVL * 20]), Tt); Tp = BYTW(&(W[TWVL * 12]), To); } } } { V Tf, T11, Tk, T12, Ts; TO = VADD(T4, T9); Ta = VSUB(T4, T9); TJ = VSUB(TD, TI); TP = VADD(TD, TI); Tf = VADD(Tc, Te); T11 = VSUB(Tc, Te); Tk = VADD(Th, Tj); T12 = VSUB(Th, Tj); Ts = BYTW(&(W[TWVL * 4]), Tr); T14 = VSUB(Tn, Tp); Tq = VADD(Tn, Tp); T1i = VFNMS(LDK(KP707106781), TZ, TW); T10 = VFMA(LDK(KP707106781), TZ, TW); T1b = VFMA(LDK(KP707106781), T1a, T19); T1l = VFNMS(LDK(KP707106781), T1a, T19); T13 = VFNMS(LDK(KP414213562), T12, T11); T1c = VFMA(LDK(KP414213562), T11, T12); TR = VADD(Tf, Tk); Tl = VSUB(Tf, Tk); T15 = VSUB(Tu, Ts); Tv = VADD(Ts, Tu); } } { V T1d, T16, TS, Tw, TU, TQ; T1d = VFMA(LDK(KP414213562), T14, T15); T16 = VFNMS(LDK(KP414213562), T15, T14); TS = VADD(Tq, Tv); Tw = VSUB(Tq, Tv); TU = VADD(TO, TP); TQ = VSUB(TO, TP); { V T1e, T1j, T17, T1m; T1e = VSUB(T1c, T1d); T1j = VADD(T1c, T1d); T17 = VADD(T13, T16); T1m = VSUB(T13, T16); { V TV, TT, TK, Tx; TV = VADD(TR, TS); TT = VSUB(TR, TS); TK = VSUB(Tl, Tw); Tx = VADD(Tl, Tw); { V T1h, T1f, T1o, T1k; T1h = VFMA(LDK(KP923879532), T1e, T1b); T1f = VFNMS(LDK(KP923879532), T1e, T1b); T1o = VFMA(LDK(KP923879532), T1j, T1i); T1k = VFNMS(LDK(KP923879532), T1j, T1i); { V T1g, T18, T1p, T1n; T1g = VFMA(LDK(KP923879532), T17, T10); T18 = VFNMS(LDK(KP923879532), T17, T10); T1p = VFNMS(LDK(KP923879532), T1m, T1l); T1n = VFMA(LDK(KP923879532), T1m, T1l); ST(&(x[WS(rs, 8)]), VSUB(TU, TV), ms, &(x[0])); ST(&(x[0]), VADD(TU, TV), ms, &(x[0])); ST(&(x[WS(rs, 4)]), VFMAI(TT, TQ), ms, &(x[0])); ST(&(x[WS(rs, 12)]), VFNMSI(TT, TQ), ms, &(x[0])); { V TN, TL, TM, Ty; TN = VFMA(LDK(KP707106781), TK, TJ); TL = VFNMS(LDK(KP707106781), TK, TJ); TM = VFMA(LDK(KP707106781), Tx, Ta); Ty = VFNMS(LDK(KP707106781), Tx, Ta); ST(&(x[WS(rs, 15)]), VFNMSI(T1h, T1g), ms, &(x[WS(rs, 1)])); ST(&(x[WS(rs, 1)]), VFMAI(T1h, T1g), ms, &(x[WS(rs, 1)])); ST(&(x[WS(rs, 9)]), VFMAI(T1f, T18), ms, &(x[WS(rs, 1)])); ST(&(x[WS(rs, 7)]), VFNMSI(T1f, T18), ms, &(x[WS(rs, 1)])); ST(&(x[WS(rs, 3)]), VFNMSI(T1p, T1o), ms, &(x[WS(rs, 1)])); ST(&(x[WS(rs, 13)]), VFMAI(T1p, T1o), ms, &(x[WS(rs, 1)])); ST(&(x[WS(rs, 11)]), VFNMSI(T1n, T1k), ms, &(x[WS(rs, 1)])); ST(&(x[WS(rs, 5)]), VFMAI(T1n, T1k), ms, &(x[WS(rs, 1)])); ST(&(x[WS(rs, 2)]), VFMAI(TN, TM), ms, &(x[0])); ST(&(x[WS(rs, 14)]), VFNMSI(TN, TM), ms, &(x[0])); ST(&(x[WS(rs, 10)]), VFMAI(TL, Ty), ms, &(x[0])); ST(&(x[WS(rs, 6)]), VFNMSI(TL, Ty), ms, &(x[0])); } } } } } } } } VLEAVE(); } static const tw_instr twinstr[] = { VTW(0, 1), VTW(0, 2), VTW(0, 3), VTW(0, 4), VTW(0, 5), VTW(0, 6), VTW(0, 7), VTW(0, 8), VTW(0, 9), VTW(0, 10), VTW(0, 11), VTW(0, 12), VTW(0, 13), VTW(0, 14), VTW(0, 15), {TW_NEXT, VL, 0} }; static const ct_desc desc = { 16, XSIMD_STRING("t1bv_16"), twinstr, &GENUS, {53, 30, 34, 0}, 0, 0, 0 }; void XSIMD(codelet_t1bv_16) (planner *p) { X(kdft_dit_register) (p, t1bv_16, &desc); } #else /* HAVE_FMA */ /* Generated by: ../../../genfft/gen_twiddle_c.native -simd -compact -variables 4 -pipeline-latency 8 -n 16 -name t1bv_16 -include t1b.h -sign 1 */ /* * This function contains 87 FP additions, 42 FP multiplications, * (or, 83 additions, 38 multiplications, 4 fused multiply/add), * 36 stack variables, 3 constants, and 32 memory accesses */ #include "t1b.h" static void t1bv_16(R *ri, R *ii, const R *W, stride rs, INT mb, INT me, INT ms) { DVK(KP382683432, +0.382683432365089771728459984030398866761344562); DVK(KP923879532, +0.923879532511286756128183189396788286822416626); DVK(KP707106781, +0.707106781186547524400844362104849039284835938); { INT m; R *x; x = ii; for (m = mb, W = W + (mb * ((TWVL / VL) * 30)); m < me; m = m + VL, x = x + (VL * ms), W = W + (TWVL * 30), MAKE_VOLATILE_STRIDE(16, rs)) { V TJ, T1b, TD, T1c, T17, T18, Ty, TK, T10, T11, T12, Tb, TM, T13, T14; V T15, Tm, TN, TG, TI, TH; TG = LD(&(x[0]), ms, &(x[0])); TH = LD(&(x[WS(rs, 8)]), ms, &(x[0])); TI = BYTW(&(W[TWVL * 14]), TH); TJ = VSUB(TG, TI); T1b = VADD(TG, TI); { V TA, TC, Tz, TB; Tz = LD(&(x[WS(rs, 4)]), ms, &(x[0])); TA = BYTW(&(W[TWVL * 6]), Tz); TB = LD(&(x[WS(rs, 12)]), ms, &(x[0])); TC = BYTW(&(W[TWVL * 22]), TB); TD = VSUB(TA, TC); T1c = VADD(TA, TC); } { V Tp, Tw, Tr, Tu, Ts, Tx; { V To, Tv, Tq, Tt; To = LD(&(x[WS(rs, 2)]), ms, &(x[0])); Tp = BYTW(&(W[TWVL * 2]), To); Tv = LD(&(x[WS(rs, 6)]), ms, &(x[0])); Tw = BYTW(&(W[TWVL * 10]), Tv); Tq = LD(&(x[WS(rs, 10)]), ms, &(x[0])); Tr = BYTW(&(W[TWVL * 18]), Tq); Tt = LD(&(x[WS(rs, 14)]), ms, &(x[0])); Tu = BYTW(&(W[TWVL * 26]), Tt); } T17 = VADD(Tp, Tr); T18 = VADD(Tu, Tw); Ts = VSUB(Tp, Tr); Tx = VSUB(Tu, Tw); Ty = VMUL(LDK(KP707106781), VSUB(Ts, Tx)); TK = VMUL(LDK(KP707106781), VADD(Ts, Tx)); } { V T2, T9, T4, T7, T5, Ta; { V T1, T8, T3, T6; T1 = LD(&(x[WS(rs, 1)]), ms, &(x[WS(rs, 1)])); T2 = BYTW(&(W[0]), T1); T8 = LD(&(x[WS(rs, 13)]), ms, &(x[WS(rs, 1)])); T9 = BYTW(&(W[TWVL * 24]), T8); T3 = LD(&(x[WS(rs, 9)]), ms, &(x[WS(rs, 1)])); T4 = BYTW(&(W[TWVL * 16]), T3); T6 = LD(&(x[WS(rs, 5)]), ms, &(x[WS(rs, 1)])); T7 = BYTW(&(W[TWVL * 8]), T6); } T10 = VADD(T2, T4); T11 = VADD(T7, T9); T12 = VSUB(T10, T11); T5 = VSUB(T2, T4); Ta = VSUB(T7, T9); Tb = VFNMS(LDK(KP382683432), Ta, VMUL(LDK(KP923879532), T5)); TM = VFMA(LDK(KP382683432), T5, VMUL(LDK(KP923879532), Ta)); } { V Td, Tk, Tf, Ti, Tg, Tl; { V Tc, Tj, Te, Th; Tc = LD(&(x[WS(rs, 15)]), ms, &(x[WS(rs, 1)])); Td = BYTW(&(W[TWVL * 28]), Tc); Tj = LD(&(x[WS(rs, 11)]), ms, &(x[WS(rs, 1)])); Tk = BYTW(&(W[TWVL * 20]), Tj); Te = LD(&(x[WS(rs, 7)]), ms, &(x[WS(rs, 1)])); Tf = BYTW(&(W[TWVL * 12]), Te); Th = LD(&(x[WS(rs, 3)]), ms, &(x[WS(rs, 1)])); Ti = BYTW(&(W[TWVL * 4]), Th); } T13 = VADD(Td, Tf); T14 = VADD(Ti, Tk); T15 = VSUB(T13, T14); Tg = VSUB(Td, Tf); Tl = VSUB(Ti, Tk); Tm = VFMA(LDK(KP923879532), Tg, VMUL(LDK(KP382683432), Tl)); TN = VFNMS(LDK(KP382683432), Tg, VMUL(LDK(KP923879532), Tl)); } { V T1a, T1g, T1f, T1h; { V T16, T19, T1d, T1e; T16 = VMUL(LDK(KP707106781), VSUB(T12, T15)); T19 = VSUB(T17, T18); T1a = VBYI(VSUB(T16, T19)); T1g = VBYI(VADD(T19, T16)); T1d = VSUB(T1b, T1c); T1e = VMUL(LDK(KP707106781), VADD(T12, T15)); T1f = VSUB(T1d, T1e); T1h = VADD(T1d, T1e); } ST(&(x[WS(rs, 6)]), VADD(T1a, T1f), ms, &(x[0])); ST(&(x[WS(rs, 14)]), VSUB(T1h, T1g), ms, &(x[0])); ST(&(x[WS(rs, 10)]), VSUB(T1f, T1a), ms, &(x[0])); ST(&(x[WS(rs, 2)]), VADD(T1g, T1h), ms, &(x[0])); } { V T1k, T1o, T1n, T1p; { V T1i, T1j, T1l, T1m; T1i = VADD(T1b, T1c); T1j = VADD(T17, T18); T1k = VSUB(T1i, T1j); T1o = VADD(T1i, T1j); T1l = VADD(T10, T11); T1m = VADD(T13, T14); T1n = VBYI(VSUB(T1l, T1m)); T1p = VADD(T1l, T1m); } ST(&(x[WS(rs, 12)]), VSUB(T1k, T1n), ms, &(x[0])); ST(&(x[0]), VADD(T1o, T1p), ms, &(x[0])); ST(&(x[WS(rs, 4)]), VADD(T1k, T1n), ms, &(x[0])); ST(&(x[WS(rs, 8)]), VSUB(T1o, T1p), ms, &(x[0])); } { V TF, TQ, TP, TR; { V Tn, TE, TL, TO; Tn = VSUB(Tb, Tm); TE = VSUB(Ty, TD); TF = VBYI(VSUB(Tn, TE)); TQ = VBYI(VADD(TE, Tn)); TL = VSUB(TJ, TK); TO = VSUB(TM, TN); TP = VSUB(TL, TO); TR = VADD(TL, TO); } ST(&(x[WS(rs, 5)]), VADD(TF, TP), ms, &(x[WS(rs, 1)])); ST(&(x[WS(rs, 13)]), VSUB(TR, TQ), ms, &(x[WS(rs, 1)])); ST(&(x[WS(rs, 11)]), VSUB(TP, TF), ms, &(x[WS(rs, 1)])); ST(&(x[WS(rs, 3)]), VADD(TQ, TR), ms, &(x[WS(rs, 1)])); } { V TU, TY, TX, TZ; { V TS, TT, TV, TW; TS = VADD(TJ, TK); TT = VADD(Tb, Tm); TU = VADD(TS, TT); TY = VSUB(TS, TT); TV = VADD(TD, Ty); TW = VADD(TM, TN); TX = VBYI(VADD(TV, TW)); TZ = VBYI(VSUB(TW, TV)); } ST(&(x[WS(rs, 15)]), VSUB(TU, TX), ms, &(x[WS(rs, 1)])); ST(&(x[WS(rs, 7)]), VADD(TY, TZ), ms, &(x[WS(rs, 1)])); ST(&(x[WS(rs, 1)]), VADD(TU, TX), ms, &(x[WS(rs, 1)])); ST(&(x[WS(rs, 9)]), VSUB(TY, TZ), ms, &(x[WS(rs, 1)])); } } } VLEAVE(); } static const tw_instr twinstr[] = { VTW(0, 1), VTW(0, 2), VTW(0, 3), VTW(0, 4), VTW(0, 5), VTW(0, 6), VTW(0, 7), VTW(0, 8), VTW(0, 9), VTW(0, 10), VTW(0, 11), VTW(0, 12), VTW(0, 13), VTW(0, 14), VTW(0, 15), {TW_NEXT, VL, 0} }; static const ct_desc desc = { 16, XSIMD_STRING("t1bv_16"), twinstr, &GENUS, {83, 38, 4, 0}, 0, 0, 0 }; void XSIMD(codelet_t1bv_16) (planner *p) { X(kdft_dit_register) (p, t1bv_16, &desc); } #endif /* HAVE_FMA */ fftw-3.3.3/dft/simd/common/t1bv_32.c0000644000175000017500000007043712054410555013750 00000000000000/* * Copyright (c) 2003, 2007-11 Matteo Frigo * Copyright (c) 2003, 2007-11 Massachusetts Institute of Technology * * This program is free software; you can redistribute it and/or modify * it under the terms of the GNU General Public License as published by * the Free Software Foundation; either version 2 of the License, or * (at your option) any later version. * * This program is distributed in the hope that it will be useful, * but WITHOUT ANY WARRANTY; without even the implied warranty of * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the * GNU General Public License for more details. * * You should have received a copy of the GNU General Public License * along with this program; if not, write to the Free Software * Foundation, Inc., 51 Franklin Street, Fifth Floor, Boston, MA 02110-1301 USA * */ /* This file was automatically generated --- DO NOT EDIT */ /* Generated on Sun Nov 25 07:39:04 EST 2012 */ #include "codelet-dft.h" #ifdef HAVE_FMA /* Generated by: ../../../genfft/gen_twiddle_c.native -fma -reorder-insns -schedule-for-pipeline -simd -compact -variables 4 -pipeline-latency 8 -n 32 -name t1bv_32 -include t1b.h -sign 1 */ /* * This function contains 217 FP additions, 160 FP multiplications, * (or, 119 additions, 62 multiplications, 98 fused multiply/add), * 104 stack variables, 7 constants, and 64 memory accesses */ #include "t1b.h" static void t1bv_32(R *ri, R *ii, const R *W, stride rs, INT mb, INT me, INT ms) { DVK(KP831469612, +0.831469612302545237078788377617905756738560812); DVK(KP980785280, +0.980785280403230449126182236134239036973933731); DVK(KP668178637, +0.668178637919298919997757686523080761552472251); DVK(KP198912367, +0.198912367379658006911597622644676228597850501); DVK(KP923879532, +0.923879532511286756128183189396788286822416626); DVK(KP707106781, +0.707106781186547524400844362104849039284835938); DVK(KP414213562, +0.414213562373095048801688724209698078569671875); { INT m; R *x; x = ii; for (m = mb, W = W + (mb * ((TWVL / VL) * 62)); m < me; m = m + VL, x = x + (VL * ms), W = W + (TWVL * 62), MAKE_VOLATILE_STRIDE(32, rs)) { V T26, T25, T2a, T2i, T24, T2c, T2g, T2k, T2h, T27; { V T4, T1z, T2o, T32, T2r, T3f, Tf, T1A, T34, T2O, T1D, TC, T33, T2L, T1C; V Tr, T2C, T3a, T2F, T3b, T1r, T21, T1k, T20, TQ, TM, TS, TL, T2t, TJ; V T10, T2u; { V Tt, T9, T2p, Te, T2q, TA, Tu, Tx; { V T1, T1x, T2, T1v; T1 = LD(&(x[0]), ms, &(x[0])); T1x = LD(&(x[WS(rs, 24)]), ms, &(x[0])); T2 = LD(&(x[WS(rs, 16)]), ms, &(x[0])); T1v = LD(&(x[WS(rs, 8)]), ms, &(x[0])); { V T5, Tc, T7, Ta, T2m, T2n; T5 = LD(&(x[WS(rs, 4)]), ms, &(x[0])); Tc = LD(&(x[WS(rs, 12)]), ms, &(x[0])); T7 = LD(&(x[WS(rs, 20)]), ms, &(x[0])); Ta = LD(&(x[WS(rs, 28)]), ms, &(x[0])); { V T1y, T3, T1w, T6, Td, T8, Tb, Ts, Tz; Ts = LD(&(x[WS(rs, 30)]), ms, &(x[0])); T1y = BYTW(&(W[TWVL * 46]), T1x); T3 = BYTW(&(W[TWVL * 30]), T2); T1w = BYTW(&(W[TWVL * 14]), T1v); T6 = BYTW(&(W[TWVL * 6]), T5); Td = BYTW(&(W[TWVL * 22]), Tc); T8 = BYTW(&(W[TWVL * 38]), T7); Tb = BYTW(&(W[TWVL * 54]), Ta); Tt = BYTW(&(W[TWVL * 58]), Ts); Tz = LD(&(x[WS(rs, 6)]), ms, &(x[0])); T4 = VSUB(T1, T3); T2m = VADD(T1, T3); T1z = VSUB(T1w, T1y); T2n = VADD(T1w, T1y); T9 = VSUB(T6, T8); T2p = VADD(T6, T8); Te = VSUB(Tb, Td); T2q = VADD(Tb, Td); TA = BYTW(&(W[TWVL * 10]), Tz); } Tu = LD(&(x[WS(rs, 14)]), ms, &(x[0])); T2o = VADD(T2m, T2n); T32 = VSUB(T2m, T2n); Tx = LD(&(x[WS(rs, 22)]), ms, &(x[0])); } } { V Tv, To, Ty, Ti, Tj, Tm, Th; Th = LD(&(x[WS(rs, 2)]), ms, &(x[0])); T2r = VADD(T2p, T2q); T3f = VSUB(T2p, T2q); Tf = VADD(T9, Te); T1A = VSUB(T9, Te); Tv = BYTW(&(W[TWVL * 26]), Tu); To = LD(&(x[WS(rs, 26)]), ms, &(x[0])); Ty = BYTW(&(W[TWVL * 42]), Tx); Ti = BYTW(&(W[TWVL * 2]), Th); Tj = LD(&(x[WS(rs, 18)]), ms, &(x[0])); Tm = LD(&(x[WS(rs, 10)]), ms, &(x[0])); { V T1f, T1h, T1a, T1c, T18, T2A, T2B, T1p; { V T15, T17, T1o, T1m; { V Tw, T2M, Tp, T2N, TB, Tk, Tn, T1n, T14, T16; T14 = LD(&(x[WS(rs, 31)]), ms, &(x[WS(rs, 1)])); T16 = LD(&(x[WS(rs, 15)]), ms, &(x[WS(rs, 1)])); Tw = VSUB(Tt, Tv); T2M = VADD(Tt, Tv); Tp = BYTW(&(W[TWVL * 50]), To); T2N = VADD(TA, Ty); TB = VSUB(Ty, TA); Tk = BYTW(&(W[TWVL * 34]), Tj); Tn = BYTW(&(W[TWVL * 18]), Tm); T15 = BYTW(&(W[TWVL * 60]), T14); T17 = BYTW(&(W[TWVL * 28]), T16); T1n = LD(&(x[WS(rs, 7)]), ms, &(x[WS(rs, 1)])); { V T2J, Tl, T2K, Tq, T1l; T1l = LD(&(x[WS(rs, 23)]), ms, &(x[WS(rs, 1)])); T34 = VSUB(T2M, T2N); T2O = VADD(T2M, T2N); T1D = VFMA(LDK(KP414213562), Tw, TB); TC = VFNMS(LDK(KP414213562), TB, Tw); T2J = VADD(Ti, Tk); Tl = VSUB(Ti, Tk); T2K = VADD(Tn, Tp); Tq = VSUB(Tn, Tp); T1o = BYTW(&(W[TWVL * 12]), T1n); T1m = BYTW(&(W[TWVL * 44]), T1l); { V T1e, T1g, T19, T1b; T1e = LD(&(x[WS(rs, 27)]), ms, &(x[WS(rs, 1)])); T1g = LD(&(x[WS(rs, 11)]), ms, &(x[WS(rs, 1)])); T19 = LD(&(x[WS(rs, 3)]), ms, &(x[WS(rs, 1)])); T1b = LD(&(x[WS(rs, 19)]), ms, &(x[WS(rs, 1)])); T33 = VSUB(T2J, T2K); T2L = VADD(T2J, T2K); T1C = VFMA(LDK(KP414213562), Tl, Tq); Tr = VFNMS(LDK(KP414213562), Tq, Tl); T1f = BYTW(&(W[TWVL * 52]), T1e); T1h = BYTW(&(W[TWVL * 20]), T1g); T1a = BYTW(&(W[TWVL * 4]), T19); T1c = BYTW(&(W[TWVL * 36]), T1b); } } } T18 = VSUB(T15, T17); T2A = VADD(T15, T17); T2B = VADD(T1o, T1m); T1p = VSUB(T1m, T1o); } { V TG, TI, TZ, TX; { V T1i, T2E, T1d, T2D, TH, TY, TF; TF = LD(&(x[WS(rs, 1)]), ms, &(x[WS(rs, 1)])); T1i = VSUB(T1f, T1h); T2E = VADD(T1f, T1h); T1d = VSUB(T1a, T1c); T2D = VADD(T1a, T1c); TH = LD(&(x[WS(rs, 17)]), ms, &(x[WS(rs, 1)])); TY = LD(&(x[WS(rs, 25)]), ms, &(x[WS(rs, 1)])); T2C = VADD(T2A, T2B); T3a = VSUB(T2A, T2B); TG = BYTW(&(W[0]), TF); { V TW, T1j, T1q, TP, TR, TK; TW = LD(&(x[WS(rs, 9)]), ms, &(x[WS(rs, 1)])); T2F = VADD(T2D, T2E); T3b = VSUB(T2E, T2D); T1j = VADD(T1d, T1i); T1q = VSUB(T1i, T1d); TI = BYTW(&(W[TWVL * 32]), TH); TZ = BYTW(&(W[TWVL * 48]), TY); TP = LD(&(x[WS(rs, 29)]), ms, &(x[WS(rs, 1)])); TX = BYTW(&(W[TWVL * 16]), TW); TR = LD(&(x[WS(rs, 13)]), ms, &(x[WS(rs, 1)])); TK = LD(&(x[WS(rs, 5)]), ms, &(x[WS(rs, 1)])); T1r = VFMA(LDK(KP707106781), T1q, T1p); T21 = VFNMS(LDK(KP707106781), T1q, T1p); T1k = VFMA(LDK(KP707106781), T1j, T18); T20 = VFNMS(LDK(KP707106781), T1j, T18); TQ = BYTW(&(W[TWVL * 56]), TP); TM = LD(&(x[WS(rs, 21)]), ms, &(x[WS(rs, 1)])); TS = BYTW(&(W[TWVL * 24]), TR); TL = BYTW(&(W[TWVL * 8]), TK); } } T2t = VADD(TG, TI); TJ = VSUB(TG, TI); T10 = VSUB(TX, TZ); T2u = VADD(TX, TZ); } } } } { V T2s, TT, T2x, T2P, T2Y, T2G, T37, T2v, T2w, TO, T2W, T30, T2U, TN, T2V; T2s = VSUB(T2o, T2r); T2U = VADD(T2o, T2r); TN = BYTW(&(W[TWVL * 40]), TM); TT = VSUB(TQ, TS); T2x = VADD(TQ, TS); T2P = VSUB(T2L, T2O); T2V = VADD(T2L, T2O); T2Y = VADD(T2C, T2F); T2G = VSUB(T2C, T2F); T37 = VSUB(T2t, T2u); T2v = VADD(T2t, T2u); T2w = VADD(TL, TN); TO = VSUB(TL, TN); T2W = VSUB(T2U, T2V); T30 = VADD(T2U, T2V); { V T1Y, T12, T1X, TV, T3n, T3t, T3m, T3q; { V T3o, T36, T3r, T3h, T3k, T3p, T3d, T3s, T2H, T2Q, T2Z, T31; { V T35, T3g, T38, T2y, T11, TU, T3c, T3j; T35 = VADD(T33, T34); T3g = VSUB(T33, T34); T38 = VSUB(T2w, T2x); T2y = VADD(T2w, T2x); T11 = VSUB(TO, TT); TU = VADD(TO, TT); T3c = VFNMS(LDK(KP414213562), T3b, T3a); T3j = VFMA(LDK(KP414213562), T3a, T3b); T3o = VFNMS(LDK(KP707106781), T35, T32); T36 = VFMA(LDK(KP707106781), T35, T32); T3r = VFNMS(LDK(KP707106781), T3g, T3f); T3h = VFMA(LDK(KP707106781), T3g, T3f); { V T3i, T39, T2z, T2X; T3i = VFMA(LDK(KP414213562), T37, T38); T39 = VFNMS(LDK(KP414213562), T38, T37); T2z = VSUB(T2v, T2y); T2X = VADD(T2v, T2y); T1Y = VFNMS(LDK(KP707106781), T11, T10); T12 = VFMA(LDK(KP707106781), T11, T10); T1X = VFNMS(LDK(KP707106781), TU, TJ); TV = VFMA(LDK(KP707106781), TU, TJ); T3k = VSUB(T3i, T3j); T3p = VADD(T3i, T3j); T3d = VADD(T39, T3c); T3s = VSUB(T39, T3c); T2H = VADD(T2z, T2G); T2Q = VSUB(T2z, T2G); T2Z = VSUB(T2X, T2Y); T31 = VADD(T2X, T2Y); } } { V T3v, T3u, T3l, T3e; T3l = VFNMS(LDK(KP923879532), T3k, T3h); T3n = VFMA(LDK(KP923879532), T3k, T3h); T3t = VFMA(LDK(KP923879532), T3s, T3r); T3v = VFNMS(LDK(KP923879532), T3s, T3r); T3e = VFNMS(LDK(KP923879532), T3d, T36); T3m = VFMA(LDK(KP923879532), T3d, T36); { V T2R, T2T, T2I, T2S; T2R = VFNMS(LDK(KP707106781), T2Q, T2P); T2T = VFMA(LDK(KP707106781), T2Q, T2P); T2I = VFNMS(LDK(KP707106781), T2H, T2s); T2S = VFMA(LDK(KP707106781), T2H, T2s); ST(&(x[WS(rs, 16)]), VSUB(T30, T31), ms, &(x[0])); ST(&(x[0]), VADD(T30, T31), ms, &(x[0])); ST(&(x[WS(rs, 8)]), VFMAI(T2Z, T2W), ms, &(x[0])); ST(&(x[WS(rs, 24)]), VFNMSI(T2Z, T2W), ms, &(x[0])); T3q = VFNMS(LDK(KP923879532), T3p, T3o); T3u = VFMA(LDK(KP923879532), T3p, T3o); ST(&(x[WS(rs, 18)]), VFMAI(T3l, T3e), ms, &(x[0])); ST(&(x[WS(rs, 14)]), VFNMSI(T3l, T3e), ms, &(x[0])); ST(&(x[WS(rs, 28)]), VFNMSI(T2T, T2S), ms, &(x[0])); ST(&(x[WS(rs, 4)]), VFMAI(T2T, T2S), ms, &(x[0])); ST(&(x[WS(rs, 20)]), VFMAI(T2R, T2I), ms, &(x[0])); ST(&(x[WS(rs, 12)]), VFNMSI(T2R, T2I), ms, &(x[0])); } ST(&(x[WS(rs, 26)]), VFMAI(T3v, T3u), ms, &(x[0])); ST(&(x[WS(rs, 6)]), VFNMSI(T3v, T3u), ms, &(x[0])); } } { V T1U, T13, T1s, TE, T1M, T1I, T1N, T1B, T1V, T1E; { V Tg, TD, T1G, T1H; Tg = VFMA(LDK(KP707106781), Tf, T4); T1U = VFNMS(LDK(KP707106781), Tf, T4); T26 = VSUB(Tr, TC); TD = VADD(Tr, TC); T1G = VFMA(LDK(KP198912367), TV, T12); T13 = VFNMS(LDK(KP198912367), T12, TV); T1s = VFNMS(LDK(KP198912367), T1r, T1k); T1H = VFMA(LDK(KP198912367), T1k, T1r); ST(&(x[WS(rs, 2)]), VFMAI(T3n, T3m), ms, &(x[0])); ST(&(x[WS(rs, 30)]), VFNMSI(T3n, T3m), ms, &(x[0])); ST(&(x[WS(rs, 22)]), VFNMSI(T3t, T3q), ms, &(x[0])); ST(&(x[WS(rs, 10)]), VFMAI(T3t, T3q), ms, &(x[0])); TE = VFMA(LDK(KP923879532), TD, Tg); T1M = VFNMS(LDK(KP923879532), TD, Tg); T1I = VSUB(T1G, T1H); T1N = VADD(T1G, T1H); T1B = VFMA(LDK(KP707106781), T1A, T1z); T25 = VFNMS(LDK(KP707106781), T1A, T1z); T1V = VADD(T1C, T1D); T1E = VSUB(T1C, T1D); } { V T1W, T2e, T2f, T23; { V T28, T1Z, T1S, T1O, T1t, T1Q, T1F, T1P, T22, T29; T28 = VFNMS(LDK(KP668178637), T1X, T1Y); T1Z = VFMA(LDK(KP668178637), T1Y, T1X); T1S = VFMA(LDK(KP980785280), T1N, T1M); T1O = VFNMS(LDK(KP980785280), T1N, T1M); T1t = VADD(T13, T1s); T1Q = VSUB(T13, T1s); T1F = VFMA(LDK(KP923879532), T1E, T1B); T1P = VFNMS(LDK(KP923879532), T1E, T1B); T1W = VFMA(LDK(KP923879532), T1V, T1U); T2e = VFNMS(LDK(KP923879532), T1V, T1U); T22 = VFMA(LDK(KP668178637), T21, T20); T29 = VFNMS(LDK(KP668178637), T20, T21); { V T1K, T1u, T1R, T1T, T1L, T1J; T1K = VFMA(LDK(KP980785280), T1t, TE); T1u = VFNMS(LDK(KP980785280), T1t, TE); T1R = VFMA(LDK(KP980785280), T1Q, T1P); T1T = VFNMS(LDK(KP980785280), T1Q, T1P); T1L = VFMA(LDK(KP980785280), T1I, T1F); T1J = VFNMS(LDK(KP980785280), T1I, T1F); T2f = VADD(T28, T29); T2a = VSUB(T28, T29); T23 = VADD(T1Z, T22); T2i = VSUB(T1Z, T22); ST(&(x[WS(rs, 23)]), VFNMSI(T1R, T1O), ms, &(x[WS(rs, 1)])); ST(&(x[WS(rs, 9)]), VFMAI(T1R, T1O), ms, &(x[WS(rs, 1)])); ST(&(x[WS(rs, 25)]), VFMAI(T1T, T1S), ms, &(x[WS(rs, 1)])); ST(&(x[WS(rs, 7)]), VFNMSI(T1T, T1S), ms, &(x[WS(rs, 1)])); ST(&(x[WS(rs, 1)]), VFMAI(T1L, T1K), ms, &(x[WS(rs, 1)])); ST(&(x[WS(rs, 31)]), VFNMSI(T1L, T1K), ms, &(x[WS(rs, 1)])); ST(&(x[WS(rs, 17)]), VFMAI(T1J, T1u), ms, &(x[WS(rs, 1)])); ST(&(x[WS(rs, 15)]), VFNMSI(T1J, T1u), ms, &(x[WS(rs, 1)])); } } T24 = VFNMS(LDK(KP831469612), T23, T1W); T2c = VFMA(LDK(KP831469612), T23, T1W); T2g = VFMA(LDK(KP831469612), T2f, T2e); T2k = VFNMS(LDK(KP831469612), T2f, T2e); } } } } } T2h = VFMA(LDK(KP923879532), T26, T25); T27 = VFNMS(LDK(KP923879532), T26, T25); { V T2j, T2l, T2d, T2b; T2j = VFNMS(LDK(KP831469612), T2i, T2h); T2l = VFMA(LDK(KP831469612), T2i, T2h); T2d = VFMA(LDK(KP831469612), T2a, T27); T2b = VFNMS(LDK(KP831469612), T2a, T27); ST(&(x[WS(rs, 21)]), VFMAI(T2j, T2g), ms, &(x[WS(rs, 1)])); ST(&(x[WS(rs, 11)]), VFNMSI(T2j, T2g), ms, &(x[WS(rs, 1)])); ST(&(x[WS(rs, 27)]), VFNMSI(T2l, T2k), ms, &(x[WS(rs, 1)])); ST(&(x[WS(rs, 5)]), VFMAI(T2l, T2k), ms, &(x[WS(rs, 1)])); ST(&(x[WS(rs, 29)]), VFMAI(T2d, T2c), ms, &(x[WS(rs, 1)])); ST(&(x[WS(rs, 3)]), VFNMSI(T2d, T2c), ms, &(x[WS(rs, 1)])); ST(&(x[WS(rs, 13)]), VFMAI(T2b, T24), ms, &(x[WS(rs, 1)])); ST(&(x[WS(rs, 19)]), VFNMSI(T2b, T24), ms, &(x[WS(rs, 1)])); } } } VLEAVE(); } static const tw_instr twinstr[] = { VTW(0, 1), VTW(0, 2), VTW(0, 3), VTW(0, 4), VTW(0, 5), VTW(0, 6), VTW(0, 7), VTW(0, 8), VTW(0, 9), VTW(0, 10), VTW(0, 11), VTW(0, 12), VTW(0, 13), VTW(0, 14), VTW(0, 15), VTW(0, 16), VTW(0, 17), VTW(0, 18), VTW(0, 19), VTW(0, 20), VTW(0, 21), VTW(0, 22), VTW(0, 23), VTW(0, 24), VTW(0, 25), VTW(0, 26), VTW(0, 27), VTW(0, 28), VTW(0, 29), VTW(0, 30), VTW(0, 31), {TW_NEXT, VL, 0} }; static const ct_desc desc = { 32, XSIMD_STRING("t1bv_32"), twinstr, &GENUS, {119, 62, 98, 0}, 0, 0, 0 }; void XSIMD(codelet_t1bv_32) (planner *p) { X(kdft_dit_register) (p, t1bv_32, &desc); } #else /* HAVE_FMA */ /* Generated by: ../../../genfft/gen_twiddle_c.native -simd -compact -variables 4 -pipeline-latency 8 -n 32 -name t1bv_32 -include t1b.h -sign 1 */ /* * This function contains 217 FP additions, 104 FP multiplications, * (or, 201 additions, 88 multiplications, 16 fused multiply/add), * 59 stack variables, 7 constants, and 64 memory accesses */ #include "t1b.h" static void t1bv_32(R *ri, R *ii, const R *W, stride rs, INT mb, INT me, INT ms) { DVK(KP195090322, +0.195090322016128267848284868477022240927691618); DVK(KP980785280, +0.980785280403230449126182236134239036973933731); DVK(KP555570233, +0.555570233019602224742830813948532874374937191); DVK(KP831469612, +0.831469612302545237078788377617905756738560812); DVK(KP382683432, +0.382683432365089771728459984030398866761344562); DVK(KP923879532, +0.923879532511286756128183189396788286822416626); DVK(KP707106781, +0.707106781186547524400844362104849039284835938); { INT m; R *x; x = ii; for (m = mb, W = W + (mb * ((TWVL / VL) * 62)); m < me; m = m + VL, x = x + (VL * ms), W = W + (TWVL * 62), MAKE_VOLATILE_STRIDE(32, rs)) { V T4, T1D, T2P, T3h, Tf, T1y, T2K, T3i, TC, T1w, T2G, T3e, Tr, T1v, T2D; V T3d, T1k, T20, T2y, T3a, T1r, T21, T2v, T39, TV, T1X, T2r, T37, T12, T1Y; V T2o, T36; { V T1, T1C, T3, T1A, T1B, T2, T1z, T2N, T2O; T1 = LD(&(x[0]), ms, &(x[0])); T1B = LD(&(x[WS(rs, 24)]), ms, &(x[0])); T1C = BYTW(&(W[TWVL * 46]), T1B); T2 = LD(&(x[WS(rs, 16)]), ms, &(x[0])); T3 = BYTW(&(W[TWVL * 30]), T2); T1z = LD(&(x[WS(rs, 8)]), ms, &(x[0])); T1A = BYTW(&(W[TWVL * 14]), T1z); T4 = VSUB(T1, T3); T1D = VSUB(T1A, T1C); T2N = VADD(T1, T3); T2O = VADD(T1A, T1C); T2P = VSUB(T2N, T2O); T3h = VADD(T2N, T2O); } { V T6, Td, T8, Tb; { V T5, Tc, T7, Ta; T5 = LD(&(x[WS(rs, 4)]), ms, &(x[0])); T6 = BYTW(&(W[TWVL * 6]), T5); Tc = LD(&(x[WS(rs, 12)]), ms, &(x[0])); Td = BYTW(&(W[TWVL * 22]), Tc); T7 = LD(&(x[WS(rs, 20)]), ms, &(x[0])); T8 = BYTW(&(W[TWVL * 38]), T7); Ta = LD(&(x[WS(rs, 28)]), ms, &(x[0])); Tb = BYTW(&(W[TWVL * 54]), Ta); } { V T9, Te, T2I, T2J; T9 = VSUB(T6, T8); Te = VSUB(Tb, Td); Tf = VMUL(LDK(KP707106781), VADD(T9, Te)); T1y = VMUL(LDK(KP707106781), VSUB(T9, Te)); T2I = VADD(T6, T8); T2J = VADD(Tb, Td); T2K = VSUB(T2I, T2J); T3i = VADD(T2I, T2J); } } { V Tt, TA, Tv, Ty; { V Ts, Tz, Tu, Tx; Ts = LD(&(x[WS(rs, 6)]), ms, &(x[0])); Tt = BYTW(&(W[TWVL * 10]), Ts); Tz = LD(&(x[WS(rs, 14)]), ms, &(x[0])); TA = BYTW(&(W[TWVL * 26]), Tz); Tu = LD(&(x[WS(rs, 22)]), ms, &(x[0])); Tv = BYTW(&(W[TWVL * 42]), Tu); Tx = LD(&(x[WS(rs, 30)]), ms, &(x[0])); Ty = BYTW(&(W[TWVL * 58]), Tx); } { V Tw, TB, T2E, T2F; Tw = VSUB(Tt, Tv); TB = VSUB(Ty, TA); TC = VFNMS(LDK(KP382683432), TB, VMUL(LDK(KP923879532), Tw)); T1w = VFMA(LDK(KP923879532), TB, VMUL(LDK(KP382683432), Tw)); T2E = VADD(Ty, TA); T2F = VADD(Tt, Tv); T2G = VSUB(T2E, T2F); T3e = VADD(T2E, T2F); } } { V Ti, Tp, Tk, Tn; { V Th, To, Tj, Tm; Th = LD(&(x[WS(rs, 2)]), ms, &(x[0])); Ti = BYTW(&(W[TWVL * 2]), Th); To = LD(&(x[WS(rs, 26)]), ms, &(x[0])); Tp = BYTW(&(W[TWVL * 50]), To); Tj = LD(&(x[WS(rs, 18)]), ms, &(x[0])); Tk = BYTW(&(W[TWVL * 34]), Tj); Tm = LD(&(x[WS(rs, 10)]), ms, &(x[0])); Tn = BYTW(&(W[TWVL * 18]), Tm); } { V Tl, Tq, T2B, T2C; Tl = VSUB(Ti, Tk); Tq = VSUB(Tn, Tp); Tr = VFMA(LDK(KP382683432), Tl, VMUL(LDK(KP923879532), Tq)); T1v = VFNMS(LDK(KP382683432), Tq, VMUL(LDK(KP923879532), Tl)); T2B = VADD(Ti, Tk); T2C = VADD(Tn, Tp); T2D = VSUB(T2B, T2C); T3d = VADD(T2B, T2C); } } { V T1g, T1i, T1o, T1m, T1a, T1c, T1d, T15, T17, T18; { V T1f, T1h, T1n, T1l; T1f = LD(&(x[WS(rs, 7)]), ms, &(x[WS(rs, 1)])); T1g = BYTW(&(W[TWVL * 12]), T1f); T1h = LD(&(x[WS(rs, 23)]), ms, &(x[WS(rs, 1)])); T1i = BYTW(&(W[TWVL * 44]), T1h); T1n = LD(&(x[WS(rs, 15)]), ms, &(x[WS(rs, 1)])); T1o = BYTW(&(W[TWVL * 28]), T1n); T1l = LD(&(x[WS(rs, 31)]), ms, &(x[WS(rs, 1)])); T1m = BYTW(&(W[TWVL * 60]), T1l); { V T19, T1b, T14, T16; T19 = LD(&(x[WS(rs, 27)]), ms, &(x[WS(rs, 1)])); T1a = BYTW(&(W[TWVL * 52]), T19); T1b = LD(&(x[WS(rs, 11)]), ms, &(x[WS(rs, 1)])); T1c = BYTW(&(W[TWVL * 20]), T1b); T1d = VSUB(T1a, T1c); T14 = LD(&(x[WS(rs, 3)]), ms, &(x[WS(rs, 1)])); T15 = BYTW(&(W[TWVL * 4]), T14); T16 = LD(&(x[WS(rs, 19)]), ms, &(x[WS(rs, 1)])); T17 = BYTW(&(W[TWVL * 36]), T16); T18 = VSUB(T15, T17); } } { V T1e, T1j, T2w, T2x; T1e = VMUL(LDK(KP707106781), VSUB(T18, T1d)); T1j = VSUB(T1g, T1i); T1k = VSUB(T1e, T1j); T20 = VADD(T1j, T1e); T2w = VADD(T15, T17); T2x = VADD(T1a, T1c); T2y = VSUB(T2w, T2x); T3a = VADD(T2w, T2x); } { V T1p, T1q, T2t, T2u; T1p = VSUB(T1m, T1o); T1q = VMUL(LDK(KP707106781), VADD(T18, T1d)); T1r = VSUB(T1p, T1q); T21 = VADD(T1p, T1q); T2t = VADD(T1m, T1o); T2u = VADD(T1g, T1i); T2v = VSUB(T2t, T2u); T39 = VADD(T2t, T2u); } } { V TR, TT, TZ, TX, TL, TN, TO, TG, TI, TJ; { V TQ, TS, TY, TW; TQ = LD(&(x[WS(rs, 9)]), ms, &(x[WS(rs, 1)])); TR = BYTW(&(W[TWVL * 16]), TQ); TS = LD(&(x[WS(rs, 25)]), ms, &(x[WS(rs, 1)])); TT = BYTW(&(W[TWVL * 48]), TS); TY = LD(&(x[WS(rs, 17)]), ms, &(x[WS(rs, 1)])); TZ = BYTW(&(W[TWVL * 32]), TY); TW = LD(&(x[WS(rs, 1)]), ms, &(x[WS(rs, 1)])); TX = BYTW(&(W[0]), TW); { V TK, TM, TF, TH; TK = LD(&(x[WS(rs, 29)]), ms, &(x[WS(rs, 1)])); TL = BYTW(&(W[TWVL * 56]), TK); TM = LD(&(x[WS(rs, 13)]), ms, &(x[WS(rs, 1)])); TN = BYTW(&(W[TWVL * 24]), TM); TO = VSUB(TL, TN); TF = LD(&(x[WS(rs, 5)]), ms, &(x[WS(rs, 1)])); TG = BYTW(&(W[TWVL * 8]), TF); TH = LD(&(x[WS(rs, 21)]), ms, &(x[WS(rs, 1)])); TI = BYTW(&(W[TWVL * 40]), TH); TJ = VSUB(TG, TI); } } { V TP, TU, T2p, T2q; TP = VMUL(LDK(KP707106781), VSUB(TJ, TO)); TU = VSUB(TR, TT); TV = VSUB(TP, TU); T1X = VADD(TU, TP); T2p = VADD(TG, TI); T2q = VADD(TL, TN); T2r = VSUB(T2p, T2q); T37 = VADD(T2p, T2q); } { V T10, T11, T2m, T2n; T10 = VSUB(TX, TZ); T11 = VMUL(LDK(KP707106781), VADD(TJ, TO)); T12 = VSUB(T10, T11); T1Y = VADD(T10, T11); T2m = VADD(TX, TZ); T2n = VADD(TR, TT); T2o = VSUB(T2m, T2n); T36 = VADD(T2m, T2n); } } { V T3q, T3u, T3t, T3v; { V T3o, T3p, T3r, T3s; T3o = VADD(T3h, T3i); T3p = VADD(T3d, T3e); T3q = VSUB(T3o, T3p); T3u = VADD(T3o, T3p); T3r = VADD(T36, T37); T3s = VADD(T39, T3a); T3t = VBYI(VSUB(T3r, T3s)); T3v = VADD(T3r, T3s); } ST(&(x[WS(rs, 24)]), VSUB(T3q, T3t), ms, &(x[0])); ST(&(x[0]), VADD(T3u, T3v), ms, &(x[0])); ST(&(x[WS(rs, 8)]), VADD(T3q, T3t), ms, &(x[0])); ST(&(x[WS(rs, 16)]), VSUB(T3u, T3v), ms, &(x[0])); } { V T3f, T3j, T3c, T3k, T38, T3b; T3f = VSUB(T3d, T3e); T3j = VSUB(T3h, T3i); T38 = VSUB(T36, T37); T3b = VSUB(T39, T3a); T3c = VMUL(LDK(KP707106781), VSUB(T38, T3b)); T3k = VMUL(LDK(KP707106781), VADD(T38, T3b)); { V T3g, T3l, T3m, T3n; T3g = VBYI(VSUB(T3c, T3f)); T3l = VSUB(T3j, T3k); ST(&(x[WS(rs, 12)]), VADD(T3g, T3l), ms, &(x[0])); ST(&(x[WS(rs, 20)]), VSUB(T3l, T3g), ms, &(x[0])); T3m = VBYI(VADD(T3f, T3c)); T3n = VADD(T3j, T3k); ST(&(x[WS(rs, 4)]), VADD(T3m, T3n), ms, &(x[0])); ST(&(x[WS(rs, 28)]), VSUB(T3n, T3m), ms, &(x[0])); } } { V T2L, T31, T2R, T2Y, T2A, T2Z, T2U, T32, T2H, T2Q; T2H = VMUL(LDK(KP707106781), VSUB(T2D, T2G)); T2L = VSUB(T2H, T2K); T31 = VADD(T2K, T2H); T2Q = VMUL(LDK(KP707106781), VADD(T2D, T2G)); T2R = VSUB(T2P, T2Q); T2Y = VADD(T2P, T2Q); { V T2s, T2z, T2S, T2T; T2s = VFNMS(LDK(KP382683432), T2r, VMUL(LDK(KP923879532), T2o)); T2z = VFMA(LDK(KP923879532), T2v, VMUL(LDK(KP382683432), T2y)); T2A = VSUB(T2s, T2z); T2Z = VADD(T2s, T2z); T2S = VFMA(LDK(KP382683432), T2o, VMUL(LDK(KP923879532), T2r)); T2T = VFNMS(LDK(KP382683432), T2v, VMUL(LDK(KP923879532), T2y)); T2U = VSUB(T2S, T2T); T32 = VADD(T2S, T2T); } { V T2M, T2V, T34, T35; T2M = VBYI(VSUB(T2A, T2L)); T2V = VSUB(T2R, T2U); ST(&(x[WS(rs, 10)]), VADD(T2M, T2V), ms, &(x[0])); ST(&(x[WS(rs, 22)]), VSUB(T2V, T2M), ms, &(x[0])); T34 = VSUB(T2Y, T2Z); T35 = VBYI(VSUB(T32, T31)); ST(&(x[WS(rs, 18)]), VSUB(T34, T35), ms, &(x[0])); ST(&(x[WS(rs, 14)]), VADD(T34, T35), ms, &(x[0])); } { V T2W, T2X, T30, T33; T2W = VBYI(VADD(T2L, T2A)); T2X = VADD(T2R, T2U); ST(&(x[WS(rs, 6)]), VADD(T2W, T2X), ms, &(x[0])); ST(&(x[WS(rs, 26)]), VSUB(T2X, T2W), ms, &(x[0])); T30 = VADD(T2Y, T2Z); T33 = VBYI(VADD(T31, T32)); ST(&(x[WS(rs, 30)]), VSUB(T30, T33), ms, &(x[0])); ST(&(x[WS(rs, 2)]), VADD(T30, T33), ms, &(x[0])); } } { V TE, T1P, T1I, T1Q, T1t, T1M, T1F, T1N; { V Tg, TD, T1G, T1H; Tg = VSUB(T4, Tf); TD = VSUB(Tr, TC); TE = VSUB(Tg, TD); T1P = VADD(Tg, TD); T1G = VFNMS(LDK(KP555570233), TV, VMUL(LDK(KP831469612), T12)); T1H = VFMA(LDK(KP555570233), T1k, VMUL(LDK(KP831469612), T1r)); T1I = VSUB(T1G, T1H); T1Q = VADD(T1G, T1H); } { V T13, T1s, T1x, T1E; T13 = VFMA(LDK(KP831469612), TV, VMUL(LDK(KP555570233), T12)); T1s = VFNMS(LDK(KP555570233), T1r, VMUL(LDK(KP831469612), T1k)); T1t = VSUB(T13, T1s); T1M = VADD(T13, T1s); T1x = VSUB(T1v, T1w); T1E = VSUB(T1y, T1D); T1F = VSUB(T1x, T1E); T1N = VADD(T1E, T1x); } { V T1u, T1J, T1S, T1T; T1u = VADD(TE, T1t); T1J = VBYI(VADD(T1F, T1I)); ST(&(x[WS(rs, 27)]), VSUB(T1u, T1J), ms, &(x[WS(rs, 1)])); ST(&(x[WS(rs, 5)]), VADD(T1u, T1J), ms, &(x[WS(rs, 1)])); T1S = VBYI(VADD(T1N, T1M)); T1T = VADD(T1P, T1Q); ST(&(x[WS(rs, 3)]), VADD(T1S, T1T), ms, &(x[WS(rs, 1)])); ST(&(x[WS(rs, 29)]), VSUB(T1T, T1S), ms, &(x[WS(rs, 1)])); } { V T1K, T1L, T1O, T1R; T1K = VSUB(TE, T1t); T1L = VBYI(VSUB(T1I, T1F)); ST(&(x[WS(rs, 21)]), VSUB(T1K, T1L), ms, &(x[WS(rs, 1)])); ST(&(x[WS(rs, 11)]), VADD(T1K, T1L), ms, &(x[WS(rs, 1)])); T1O = VBYI(VSUB(T1M, T1N)); T1R = VSUB(T1P, T1Q); ST(&(x[WS(rs, 13)]), VADD(T1O, T1R), ms, &(x[WS(rs, 1)])); ST(&(x[WS(rs, 19)]), VSUB(T1R, T1O), ms, &(x[WS(rs, 1)])); } } { V T1W, T2h, T2a, T2i, T23, T2e, T27, T2f; { V T1U, T1V, T28, T29; T1U = VADD(T4, Tf); T1V = VADD(T1v, T1w); T1W = VSUB(T1U, T1V); T2h = VADD(T1U, T1V); T28 = VFNMS(LDK(KP195090322), T1X, VMUL(LDK(KP980785280), T1Y)); T29 = VFMA(LDK(KP195090322), T20, VMUL(LDK(KP980785280), T21)); T2a = VSUB(T28, T29); T2i = VADD(T28, T29); } { V T1Z, T22, T25, T26; T1Z = VFMA(LDK(KP980785280), T1X, VMUL(LDK(KP195090322), T1Y)); T22 = VFNMS(LDK(KP195090322), T21, VMUL(LDK(KP980785280), T20)); T23 = VSUB(T1Z, T22); T2e = VADD(T1Z, T22); T25 = VADD(Tr, TC); T26 = VADD(T1D, T1y); T27 = VSUB(T25, T26); T2f = VADD(T26, T25); } { V T24, T2b, T2k, T2l; T24 = VADD(T1W, T23); T2b = VBYI(VADD(T27, T2a)); ST(&(x[WS(rs, 25)]), VSUB(T24, T2b), ms, &(x[WS(rs, 1)])); ST(&(x[WS(rs, 7)]), VADD(T24, T2b), ms, &(x[WS(rs, 1)])); T2k = VBYI(VADD(T2f, T2e)); T2l = VADD(T2h, T2i); ST(&(x[WS(rs, 1)]), VADD(T2k, T2l), ms, &(x[WS(rs, 1)])); ST(&(x[WS(rs, 31)]), VSUB(T2l, T2k), ms, &(x[WS(rs, 1)])); } { V T2c, T2d, T2g, T2j; T2c = VSUB(T1W, T23); T2d = VBYI(VSUB(T2a, T27)); ST(&(x[WS(rs, 23)]), VSUB(T2c, T2d), ms, &(x[WS(rs, 1)])); ST(&(x[WS(rs, 9)]), VADD(T2c, T2d), ms, &(x[WS(rs, 1)])); T2g = VBYI(VSUB(T2e, T2f)); T2j = VSUB(T2h, T2i); ST(&(x[WS(rs, 15)]), VADD(T2g, T2j), ms, &(x[WS(rs, 1)])); ST(&(x[WS(rs, 17)]), VSUB(T2j, T2g), ms, &(x[WS(rs, 1)])); } } } } VLEAVE(); } static const tw_instr twinstr[] = { VTW(0, 1), VTW(0, 2), VTW(0, 3), VTW(0, 4), VTW(0, 5), VTW(0, 6), VTW(0, 7), VTW(0, 8), VTW(0, 9), VTW(0, 10), VTW(0, 11), VTW(0, 12), VTW(0, 13), VTW(0, 14), VTW(0, 15), VTW(0, 16), VTW(0, 17), VTW(0, 18), VTW(0, 19), VTW(0, 20), VTW(0, 21), VTW(0, 22), VTW(0, 23), VTW(0, 24), VTW(0, 25), VTW(0, 26), VTW(0, 27), VTW(0, 28), VTW(0, 29), VTW(0, 30), VTW(0, 31), {TW_NEXT, VL, 0} }; static const ct_desc desc = { 32, XSIMD_STRING("t1bv_32"), twinstr, &GENUS, {201, 88, 16, 0}, 0, 0, 0 }; void XSIMD(codelet_t1bv_32) (planner *p) { X(kdft_dit_register) (p, t1bv_32, &desc); } #endif /* HAVE_FMA */ fftw-3.3.3/dft/simd/common/t1bv_64.c0000644000175000017500000017562712054410576013767 00000000000000/* * Copyright (c) 2003, 2007-11 Matteo Frigo * Copyright (c) 2003, 2007-11 Massachusetts Institute of Technology * * This program is free software; you can redistribute it and/or modify * it under the terms of the GNU General Public License as published by * the Free Software Foundation; either version 2 of the License, or * (at your option) any later version. * * This program is distributed in the hope that it will be useful, * but WITHOUT ANY WARRANTY; without even the implied warranty of * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the * GNU General Public License for more details. * * You should have received a copy of the GNU General Public License * along with this program; if not, write to the Free Software * Foundation, Inc., 51 Franklin Street, Fifth Floor, Boston, MA 02110-1301 USA * */ /* This file was automatically generated --- DO NOT EDIT */ /* Generated on Sun Nov 25 07:39:05 EST 2012 */ #include "codelet-dft.h" #ifdef HAVE_FMA /* Generated by: ../../../genfft/gen_twiddle_c.native -fma -reorder-insns -schedule-for-pipeline -simd -compact -variables 4 -pipeline-latency 8 -n 64 -name t1bv_64 -include t1b.h -sign 1 */ /* * This function contains 519 FP additions, 384 FP multiplications, * (or, 261 additions, 126 multiplications, 258 fused multiply/add), * 187 stack variables, 15 constants, and 128 memory accesses */ #include "t1b.h" static void t1bv_64(R *ri, R *ii, const R *W, stride rs, INT mb, INT me, INT ms) { DVK(KP773010453, +0.773010453362736960810906609758469800971041293); DVK(KP995184726, +0.995184726672196886244836953109479921575474869); DVK(KP820678790, +0.820678790828660330972281985331011598767386482); DVK(KP098491403, +0.098491403357164253077197521291327432293052451); DVK(KP956940335, +0.956940335732208864935797886980269969482849206); DVK(KP881921264, +0.881921264348355029712756863660388349508442621); DVK(KP303346683, +0.303346683607342391675883946941299872384187453); DVK(KP534511135, +0.534511135950791641089685961295362908582039528); DVK(KP831469612, +0.831469612302545237078788377617905756738560812); DVK(KP980785280, +0.980785280403230449126182236134239036973933731); DVK(KP668178637, +0.668178637919298919997757686523080761552472251); DVK(KP198912367, +0.198912367379658006911597622644676228597850501); DVK(KP923879532, +0.923879532511286756128183189396788286822416626); DVK(KP414213562, +0.414213562373095048801688724209698078569671875); DVK(KP707106781, +0.707106781186547524400844362104849039284835938); { INT m; R *x; x = ii; for (m = mb, W = W + (mb * ((TWVL / VL) * 126)); m < me; m = m + VL, x = x + (VL * ms), W = W + (TWVL * 126), MAKE_VOLATILE_STRIDE(64, rs)) { V T6L, T6M, T6O, T6P, T75, T6V, T5A, T6A, T72, T6K, T6t, T6D, T6w, T6B, T6h; V T6E; { V Ta, T3U, T3V, T37, T7a, T58, T7B, T6l, T1v, T24, T5Q, T7o, T5F, T7l, T43; V T4F, T2i, T2R, T6b, T7v, T60, T7s, T4a, T4I, T5u, T7h, T5x, T7g, T1i, T3b; V T4m, T4C, T7e, T5l, T7d, T5o, T3a, TV, T4B, T4j, T3X, T3Y, T6o, T7b, T5f; V T7C, Tx, T38, T2p, T61, T2n, T65, T2D, T7p, T5M, T7m, T5T, T4G, T46, T25; V T1S, T2q, T2u, T2w; { V T5q, T10, T5v, T15, T1b, T5s, T1c, T1e; { V T1V, T1p, T5B, T5O, T1u, T1X, T20, T21; { V T1, T2, T7, T5, T32, T34, T2X, T2Z; T1 = LD(&(x[0]), ms, &(x[0])); T2 = LD(&(x[WS(rs, 32)]), ms, &(x[0])); T7 = LD(&(x[WS(rs, 48)]), ms, &(x[0])); T5 = LD(&(x[WS(rs, 16)]), ms, &(x[0])); T32 = LD(&(x[WS(rs, 56)]), ms, &(x[0])); T34 = LD(&(x[WS(rs, 24)]), ms, &(x[0])); T2X = LD(&(x[WS(rs, 8)]), ms, &(x[0])); T2Z = LD(&(x[WS(rs, 40)]), ms, &(x[0])); { V T1m, T54, T6j, T36, T56, T31, T55, T1n, T1q, T1s, T4, T9; { V T3, T8, T6, T33, T35, T2Y, T30, T1l; T1l = LD(&(x[WS(rs, 1)]), ms, &(x[WS(rs, 1)])); T3 = BYTW(&(W[TWVL * 62]), T2); T8 = BYTW(&(W[TWVL * 94]), T7); T6 = BYTW(&(W[TWVL * 30]), T5); T33 = BYTW(&(W[TWVL * 110]), T32); T35 = BYTW(&(W[TWVL * 46]), T34); T2Y = BYTW(&(W[TWVL * 14]), T2X); T30 = BYTW(&(W[TWVL * 78]), T2Z); T1m = BYTW(&(W[0]), T1l); T54 = VSUB(T1, T3); T4 = VADD(T1, T3); T6j = VSUB(T6, T8); T9 = VADD(T6, T8); T36 = VADD(T33, T35); T56 = VSUB(T33, T35); T31 = VADD(T2Y, T30); T55 = VSUB(T2Y, T30); T1n = LD(&(x[WS(rs, 33)]), ms, &(x[WS(rs, 1)])); } T1q = LD(&(x[WS(rs, 17)]), ms, &(x[WS(rs, 1)])); T1s = LD(&(x[WS(rs, 49)]), ms, &(x[WS(rs, 1)])); Ta = VSUB(T4, T9); T3U = VADD(T4, T9); { V T57, T6k, T1o, T1r, T1t, T1W, T1U, T1Z; T1U = LD(&(x[WS(rs, 9)]), ms, &(x[WS(rs, 1)])); T3V = VADD(T31, T36); T37 = VSUB(T31, T36); T57 = VADD(T55, T56); T6k = VSUB(T55, T56); T1o = BYTW(&(W[TWVL * 64]), T1n); T1r = BYTW(&(W[TWVL * 32]), T1q); T1t = BYTW(&(W[TWVL * 96]), T1s); T1V = BYTW(&(W[TWVL * 16]), T1U); T1W = LD(&(x[WS(rs, 41)]), ms, &(x[WS(rs, 1)])); T1Z = LD(&(x[WS(rs, 57)]), ms, &(x[WS(rs, 1)])); T7a = VFNMS(LDK(KP707106781), T57, T54); T58 = VFMA(LDK(KP707106781), T57, T54); T7B = VFNMS(LDK(KP707106781), T6k, T6j); T6l = VFMA(LDK(KP707106781), T6k, T6j); T1p = VADD(T1m, T1o); T5B = VSUB(T1m, T1o); T5O = VSUB(T1r, T1t); T1u = VADD(T1r, T1t); T1X = BYTW(&(W[TWVL * 80]), T1W); T20 = BYTW(&(W[TWVL * 112]), T1Z); T21 = LD(&(x[WS(rs, 25)]), ms, &(x[WS(rs, 1)])); } } } { V T5W, T2N, T69, T2L, T5Y, T2P, T48, T2c, T2h; { V T41, T1Y, T5C, T22, T2d, T29, T2b, T2f, T28, T2a, T2H, T2J; T28 = LD(&(x[WS(rs, 63)]), ms, &(x[WS(rs, 1)])); T2a = LD(&(x[WS(rs, 31)]), ms, &(x[WS(rs, 1)])); T1v = VSUB(T1p, T1u); T41 = VADD(T1p, T1u); T1Y = VADD(T1V, T1X); T5C = VSUB(T1V, T1X); T22 = BYTW(&(W[TWVL * 48]), T21); T2d = LD(&(x[WS(rs, 15)]), ms, &(x[WS(rs, 1)])); T29 = BYTW(&(W[TWVL * 124]), T28); T2b = BYTW(&(W[TWVL * 60]), T2a); T2f = LD(&(x[WS(rs, 47)]), ms, &(x[WS(rs, 1)])); T2H = LD(&(x[WS(rs, 55)]), ms, &(x[WS(rs, 1)])); T2J = LD(&(x[WS(rs, 23)]), ms, &(x[WS(rs, 1)])); { V T23, T5D, T2e, T2g, T2I, T2K, T2M; T2M = LD(&(x[WS(rs, 7)]), ms, &(x[WS(rs, 1)])); T23 = VADD(T20, T22); T5D = VSUB(T20, T22); T2e = BYTW(&(W[TWVL * 28]), T2d); T2c = VADD(T29, T2b); T5W = VSUB(T29, T2b); T2g = BYTW(&(W[TWVL * 92]), T2f); T2I = BYTW(&(W[TWVL * 108]), T2H); T2K = BYTW(&(W[TWVL * 44]), T2J); T2N = BYTW(&(W[TWVL * 12]), T2M); { V T5E, T5P, T42, T2O; T5E = VADD(T5C, T5D); T5P = VSUB(T5C, T5D); T24 = VSUB(T1Y, T23); T42 = VADD(T1Y, T23); T69 = VSUB(T2g, T2e); T2h = VADD(T2e, T2g); T2O = LD(&(x[WS(rs, 39)]), ms, &(x[WS(rs, 1)])); T2L = VADD(T2I, T2K); T5Y = VSUB(T2I, T2K); T5Q = VFMA(LDK(KP707106781), T5P, T5O); T7o = VFNMS(LDK(KP707106781), T5P, T5O); T5F = VFMA(LDK(KP707106781), T5E, T5B); T7l = VFNMS(LDK(KP707106781), T5E, T5B); T43 = VADD(T41, T42); T4F = VSUB(T41, T42); T2P = BYTW(&(W[TWVL * 76]), T2O); } } } T2i = VSUB(T2c, T2h); T48 = VADD(T2c, T2h); { V TW, TY, T11, T2Q, T5X, T13; TW = LD(&(x[WS(rs, 62)]), ms, &(x[0])); TY = LD(&(x[WS(rs, 30)]), ms, &(x[0])); T11 = LD(&(x[WS(rs, 14)]), ms, &(x[0])); T2Q = VADD(T2N, T2P); T5X = VSUB(T2N, T2P); T13 = LD(&(x[WS(rs, 46)]), ms, &(x[0])); { V T12, T5Z, T6a, T49, T14, T18, T1a; { V T17, T19, TX, TZ; T17 = LD(&(x[WS(rs, 54)]), ms, &(x[0])); T19 = LD(&(x[WS(rs, 22)]), ms, &(x[0])); TX = BYTW(&(W[TWVL * 122]), TW); TZ = BYTW(&(W[TWVL * 58]), TY); T12 = BYTW(&(W[TWVL * 26]), T11); T5Z = VADD(T5X, T5Y); T6a = VSUB(T5Y, T5X); T2R = VSUB(T2L, T2Q); T49 = VADD(T2Q, T2L); T14 = BYTW(&(W[TWVL * 90]), T13); T18 = BYTW(&(W[TWVL * 106]), T17); T5q = VSUB(TX, TZ); T10 = VADD(TX, TZ); T1a = BYTW(&(W[TWVL * 42]), T19); } T6b = VFMA(LDK(KP707106781), T6a, T69); T7v = VFNMS(LDK(KP707106781), T6a, T69); T60 = VFMA(LDK(KP707106781), T5Z, T5W); T7s = VFNMS(LDK(KP707106781), T5Z, T5W); T4a = VADD(T48, T49); T4I = VSUB(T48, T49); T5v = VSUB(T14, T12); T15 = VADD(T12, T14); T1b = VADD(T18, T1a); T5s = VSUB(T18, T1a); } T1c = LD(&(x[WS(rs, 6)]), ms, &(x[0])); T1e = LD(&(x[WS(rs, 38)]), ms, &(x[0])); } } } { V Th, T59, Tf, Tv, T5d, Tj, Tm, To; { V T5h, TQ, T5m, T5i, TO, TS, TJ, T4h, TD, TI; { V T4k, T16, TB, T1d, T1f, TE, TG, TA, Tz, TK, TM, TC; Tz = LD(&(x[WS(rs, 2)]), ms, &(x[0])); T4k = VADD(T10, T15); T16 = VSUB(T10, T15); TB = LD(&(x[WS(rs, 34)]), ms, &(x[0])); T1d = BYTW(&(W[TWVL * 10]), T1c); T1f = BYTW(&(W[TWVL * 74]), T1e); TE = LD(&(x[WS(rs, 18)]), ms, &(x[0])); TG = LD(&(x[WS(rs, 50)]), ms, &(x[0])); TA = BYTW(&(W[TWVL * 2]), Tz); TK = LD(&(x[WS(rs, 10)]), ms, &(x[0])); TM = LD(&(x[WS(rs, 42)]), ms, &(x[0])); TC = BYTW(&(W[TWVL * 66]), TB); { V T1g, T5r, TF, TH, TL, TN, TP; TP = LD(&(x[WS(rs, 58)]), ms, &(x[0])); T1g = VADD(T1d, T1f); T5r = VSUB(T1d, T1f); TF = BYTW(&(W[TWVL * 34]), TE); TH = BYTW(&(W[TWVL * 98]), TG); TL = BYTW(&(W[TWVL * 18]), TK); TN = BYTW(&(W[TWVL * 82]), TM); T5h = VSUB(TA, TC); TD = VADD(TA, TC); TQ = BYTW(&(W[TWVL * 114]), TP); { V T5w, T5t, T4l, T1h, TR; T5w = VSUB(T5s, T5r); T5t = VADD(T5r, T5s); T4l = VADD(T1g, T1b); T1h = VSUB(T1b, T1g); T5m = VSUB(TF, TH); TI = VADD(TF, TH); T5i = VSUB(TL, TN); TO = VADD(TL, TN); TR = LD(&(x[WS(rs, 26)]), ms, &(x[0])); T5u = VFMA(LDK(KP707106781), T5t, T5q); T7h = VFNMS(LDK(KP707106781), T5t, T5q); T5x = VFMA(LDK(KP707106781), T5w, T5v); T7g = VFNMS(LDK(KP707106781), T5w, T5v); T1i = VFNMS(LDK(KP414213562), T1h, T16); T3b = VFMA(LDK(KP414213562), T16, T1h); T4m = VADD(T4k, T4l); T4C = VSUB(T4k, T4l); TS = BYTW(&(W[TWVL * 50]), TR); } } } TJ = VSUB(TD, TI); T4h = VADD(TD, TI); { V Tb, Td, Tr, T5j, TT, Tt, Tg; Tb = LD(&(x[WS(rs, 4)]), ms, &(x[0])); Td = LD(&(x[WS(rs, 36)]), ms, &(x[0])); Tr = LD(&(x[WS(rs, 12)]), ms, &(x[0])); T5j = VSUB(TQ, TS); TT = VADD(TQ, TS); Tt = LD(&(x[WS(rs, 44)]), ms, &(x[0])); Tg = LD(&(x[WS(rs, 20)]), ms, &(x[0])); { V Ti, Tc, Te, Ts; Ti = LD(&(x[WS(rs, 52)]), ms, &(x[0])); Tc = BYTW(&(W[TWVL * 6]), Tb); Te = BYTW(&(W[TWVL * 70]), Td); Ts = BYTW(&(W[TWVL * 22]), Tr); { V T5k, T5n, TU, T4i, Tu; T5k = VADD(T5i, T5j); T5n = VSUB(T5i, T5j); TU = VSUB(TO, TT); T4i = VADD(TO, TT); Tu = BYTW(&(W[TWVL * 86]), Tt); Th = BYTW(&(W[TWVL * 38]), Tg); T59 = VSUB(Tc, Te); Tf = VADD(Tc, Te); T7e = VFNMS(LDK(KP707106781), T5k, T5h); T5l = VFMA(LDK(KP707106781), T5k, T5h); T7d = VFNMS(LDK(KP707106781), T5n, T5m); T5o = VFMA(LDK(KP707106781), T5n, T5m); T3a = VFMA(LDK(KP414213562), TJ, TU); TV = VFNMS(LDK(KP414213562), TU, TJ); T4B = VSUB(T4h, T4i); T4j = VADD(T4h, T4i); Tv = VADD(Ts, Tu); T5d = VSUB(Tu, Ts); Tj = BYTW(&(W[TWVL * 102]), Ti); } } Tm = LD(&(x[WS(rs, 60)]), ms, &(x[0])); To = LD(&(x[WS(rs, 28)]), ms, &(x[0])); } } { V T5b, T6m, Tl, T1A, T5G, T1Q, T5K, T1C, T1D, T5e, T6n, Tw, T1H, T1J; { V T1w, T1y, T1M, T1O, Tq, T5c, T1B; T1w = LD(&(x[WS(rs, 5)]), ms, &(x[WS(rs, 1)])); T1y = LD(&(x[WS(rs, 37)]), ms, &(x[WS(rs, 1)])); T1M = LD(&(x[WS(rs, 13)]), ms, &(x[WS(rs, 1)])); T1O = LD(&(x[WS(rs, 45)]), ms, &(x[WS(rs, 1)])); T1B = LD(&(x[WS(rs, 21)]), ms, &(x[WS(rs, 1)])); { V Tk, T5a, Tn, Tp; Tk = VADD(Th, Tj); T5a = VSUB(Th, Tj); Tn = BYTW(&(W[TWVL * 118]), Tm); Tp = BYTW(&(W[TWVL * 54]), To); { V T1x, T1z, T1N, T1P; T1x = BYTW(&(W[TWVL * 8]), T1w); T1z = BYTW(&(W[TWVL * 72]), T1y); T1N = BYTW(&(W[TWVL * 24]), T1M); T1P = BYTW(&(W[TWVL * 88]), T1O); T5b = VFNMS(LDK(KP414213562), T5a, T59); T6m = VFMA(LDK(KP414213562), T59, T5a); T3X = VADD(Tf, Tk); Tl = VSUB(Tf, Tk); Tq = VADD(Tn, Tp); T5c = VSUB(Tn, Tp); T1A = VADD(T1x, T1z); T5G = VSUB(T1x, T1z); T1Q = VADD(T1N, T1P); T5K = VSUB(T1N, T1P); T1C = BYTW(&(W[TWVL * 40]), T1B); } } T1D = LD(&(x[WS(rs, 53)]), ms, &(x[WS(rs, 1)])); T5e = VFNMS(LDK(KP414213562), T5d, T5c); T6n = VFMA(LDK(KP414213562), T5c, T5d); T3Y = VADD(Tq, Tv); Tw = VSUB(Tq, Tv); T1H = LD(&(x[WS(rs, 61)]), ms, &(x[WS(rs, 1)])); T1J = LD(&(x[WS(rs, 29)]), ms, &(x[WS(rs, 1)])); } { V T1I, T1K, T1F, T5H, T2k, T2l, T2z, T2B, T2j, T1E; T2j = LD(&(x[WS(rs, 3)]), ms, &(x[WS(rs, 1)])); T1E = BYTW(&(W[TWVL * 104]), T1D); T6o = VSUB(T6m, T6n); T7b = VADD(T6m, T6n); T5f = VADD(T5b, T5e); T7C = VSUB(T5b, T5e); Tx = VADD(Tl, Tw); T38 = VSUB(Tl, Tw); T1I = BYTW(&(W[TWVL * 120]), T1H); T1K = BYTW(&(W[TWVL * 56]), T1J); T1F = VADD(T1C, T1E); T5H = VSUB(T1C, T1E); T2k = BYTW(&(W[TWVL * 4]), T2j); T2l = LD(&(x[WS(rs, 35)]), ms, &(x[WS(rs, 1)])); T2z = LD(&(x[WS(rs, 11)]), ms, &(x[WS(rs, 1)])); T2B = LD(&(x[WS(rs, 43)]), ms, &(x[WS(rs, 1)])); { V T5I, T5R, T44, T1G, T2m, T2A, T2C, T5S, T5L, T1R, T45, T2o, T5J, T1L; T2o = LD(&(x[WS(rs, 19)]), ms, &(x[WS(rs, 1)])); T5J = VSUB(T1I, T1K); T1L = VADD(T1I, T1K); T5I = VFNMS(LDK(KP414213562), T5H, T5G); T5R = VFMA(LDK(KP414213562), T5G, T5H); T44 = VADD(T1A, T1F); T1G = VSUB(T1A, T1F); T2m = BYTW(&(W[TWVL * 68]), T2l); T2A = BYTW(&(W[TWVL * 20]), T2z); T2C = BYTW(&(W[TWVL * 84]), T2B); T5S = VFNMS(LDK(KP414213562), T5J, T5K); T5L = VFMA(LDK(KP414213562), T5K, T5J); T1R = VSUB(T1L, T1Q); T45 = VADD(T1L, T1Q); T2p = BYTW(&(W[TWVL * 36]), T2o); T61 = VSUB(T2k, T2m); T2n = VADD(T2k, T2m); T65 = VSUB(T2C, T2A); T2D = VADD(T2A, T2C); T7p = VSUB(T5I, T5L); T5M = VADD(T5I, T5L); T7m = VSUB(T5R, T5S); T5T = VADD(T5R, T5S); T4G = VSUB(T44, T45); T46 = VADD(T44, T45); T25 = VSUB(T1G, T1R); T1S = VADD(T1G, T1R); T2q = LD(&(x[WS(rs, 51)]), ms, &(x[WS(rs, 1)])); } T2u = LD(&(x[WS(rs, 59)]), ms, &(x[WS(rs, 1)])); T2w = LD(&(x[WS(rs, 27)]), ms, &(x[WS(rs, 1)])); } } } } { V T67, T7w, T6e, T7t, T3s, T3E, T39, T3D, T1k, T3k, T3t, T3c, T1T, T3v, T3w; V T26, T2G, T3y, T3z, T2T; { V T4A, T4N, T47, T4v, T2r, T2v, T2x, T4s, T40, T3W, T3Z; T4A = VSUB(T3U, T3V); T3W = VADD(T3U, T3V); T3Z = VADD(T3X, T3Y); T4N = VSUB(T3X, T3Y); T47 = VSUB(T43, T46); T4v = VADD(T43, T46); T2r = BYTW(&(W[TWVL * 100]), T2q); T2v = BYTW(&(W[TWVL * 116]), T2u); T2x = BYTW(&(W[TWVL * 52]), T2w); T4s = VADD(T3W, T3Z); T40 = VSUB(T3W, T3Z); { V T4O, T4n, T4Q, T4H, T4E, T4W, T4u, T4y, T4d, T4J, T2F, T2S; { V T6c, T63, T2t, T4b, T6d, T66, T2E, T4c; { V T4D, T62, T2s, T64, T2y, T4t; T4O = VSUB(T4B, T4C); T4D = VADD(T4B, T4C); T62 = VSUB(T2r, T2p); T2s = VADD(T2p, T2r); T64 = VSUB(T2v, T2x); T2y = VADD(T2v, T2x); T4t = VADD(T4j, T4m); T4n = VSUB(T4j, T4m); T4Q = VFMA(LDK(KP414213562), T4F, T4G); T4H = VFNMS(LDK(KP414213562), T4G, T4F); T4E = VFMA(LDK(KP707106781), T4D, T4A); T4W = VFNMS(LDK(KP707106781), T4D, T4A); T6c = VFNMS(LDK(KP414213562), T61, T62); T63 = VFMA(LDK(KP414213562), T62, T61); T2t = VSUB(T2n, T2s); T4b = VADD(T2n, T2s); T6d = VFMA(LDK(KP414213562), T64, T65); T66 = VFNMS(LDK(KP414213562), T65, T64); T2E = VSUB(T2y, T2D); T4c = VADD(T2y, T2D); T4u = VSUB(T4s, T4t); T4y = VADD(T4s, T4t); } T67 = VADD(T63, T66); T7w = VSUB(T66, T63); T6e = VADD(T6c, T6d); T7t = VSUB(T6d, T6c); T4d = VADD(T4b, T4c); T4J = VSUB(T4c, T4b); T2F = VADD(T2t, T2E); T2S = VSUB(T2E, T2t); } { V Ty, T1j, T4R, T4K; Ty = VFMA(LDK(KP707106781), Tx, Ta); T3s = VFNMS(LDK(KP707106781), Tx, Ta); T3E = VSUB(TV, T1i); T1j = VADD(TV, T1i); T39 = VFMA(LDK(KP707106781), T38, T37); T3D = VFNMS(LDK(KP707106781), T38, T37); T4R = VFMA(LDK(KP414213562), T4I, T4J); T4K = VFNMS(LDK(KP414213562), T4J, T4I); { V T4w, T4e, T4P, T4Z; T4w = VADD(T4a, T4d); T4e = VSUB(T4a, T4d); T4P = VFMA(LDK(KP707106781), T4O, T4N); T4Z = VFNMS(LDK(KP707106781), T4O, T4N); T1k = VFMA(LDK(KP923879532), T1j, Ty); T3k = VFNMS(LDK(KP923879532), T1j, Ty); { V T4L, T50, T4S, T4X; T4L = VADD(T4H, T4K); T50 = VSUB(T4H, T4K); T4S = VSUB(T4Q, T4R); T4X = VADD(T4Q, T4R); { V T4f, T4o, T4x, T4z; T4f = VADD(T47, T4e); T4o = VSUB(T47, T4e); T4x = VSUB(T4v, T4w); T4z = VADD(T4v, T4w); { V T53, T51, T4M, T4U; T53 = VFNMS(LDK(KP923879532), T50, T4Z); T51 = VFMA(LDK(KP923879532), T50, T4Z); T4M = VFNMS(LDK(KP923879532), T4L, T4E); T4U = VFMA(LDK(KP923879532), T4L, T4E); { V T52, T4Y, T4T, T4V; T52 = VFMA(LDK(KP923879532), T4X, T4W); T4Y = VFNMS(LDK(KP923879532), T4X, T4W); T4T = VFNMS(LDK(KP923879532), T4S, T4P); T4V = VFMA(LDK(KP923879532), T4S, T4P); { V T4p, T4r, T4g, T4q; T4p = VFNMS(LDK(KP707106781), T4o, T4n); T4r = VFMA(LDK(KP707106781), T4o, T4n); T4g = VFNMS(LDK(KP707106781), T4f, T40); T4q = VFMA(LDK(KP707106781), T4f, T40); ST(&(x[0]), VADD(T4y, T4z), ms, &(x[0])); ST(&(x[WS(rs, 32)]), VSUB(T4y, T4z), ms, &(x[0])); ST(&(x[WS(rs, 16)]), VFMAI(T4x, T4u), ms, &(x[0])); ST(&(x[WS(rs, 48)]), VFNMSI(T4x, T4u), ms, &(x[0])); ST(&(x[WS(rs, 44)]), VFNMSI(T51, T4Y), ms, &(x[0])); ST(&(x[WS(rs, 20)]), VFMAI(T51, T4Y), ms, &(x[0])); ST(&(x[WS(rs, 52)]), VFMAI(T53, T52), ms, &(x[0])); ST(&(x[WS(rs, 12)]), VFNMSI(T53, T52), ms, &(x[0])); ST(&(x[WS(rs, 4)]), VFMAI(T4V, T4U), ms, &(x[0])); ST(&(x[WS(rs, 60)]), VFNMSI(T4V, T4U), ms, &(x[0])); ST(&(x[WS(rs, 36)]), VFMAI(T4T, T4M), ms, &(x[0])); ST(&(x[WS(rs, 28)]), VFNMSI(T4T, T4M), ms, &(x[0])); ST(&(x[WS(rs, 56)]), VFNMSI(T4r, T4q), ms, &(x[0])); ST(&(x[WS(rs, 8)]), VFMAI(T4r, T4q), ms, &(x[0])); ST(&(x[WS(rs, 40)]), VFMAI(T4p, T4g), ms, &(x[0])); ST(&(x[WS(rs, 24)]), VFNMSI(T4p, T4g), ms, &(x[0])); T3t = VADD(T3a, T3b); T3c = VSUB(T3a, T3b); } } } } } } T1T = VFMA(LDK(KP707106781), T1S, T1v); T3v = VFNMS(LDK(KP707106781), T1S, T1v); T3w = VFNMS(LDK(KP707106781), T25, T24); T26 = VFMA(LDK(KP707106781), T25, T24); T2G = VFMA(LDK(KP707106781), T2F, T2i); T3y = VFNMS(LDK(KP707106781), T2F, T2i); T3z = VFNMS(LDK(KP707106781), T2S, T2R); T2T = VFMA(LDK(KP707106781), T2S, T2R); } } } { V T3u, T3M, T3F, T3P, T3x, T3G, T3q, T3m, T3h, T3j, T3r, T3p, T2W, T3i; { V T3d, T3n, T27, T3e, T2U, T3f; T3d = VFMA(LDK(KP923879532), T3c, T39); T3n = VFNMS(LDK(KP923879532), T3c, T39); T27 = VFNMS(LDK(KP198912367), T26, T1T); T3e = VFMA(LDK(KP198912367), T1T, T26); T2U = VFNMS(LDK(KP198912367), T2T, T2G); T3f = VFMA(LDK(KP198912367), T2G, T2T); T3u = VFMA(LDK(KP923879532), T3t, T3s); T3M = VFNMS(LDK(KP923879532), T3t, T3s); { V T3g, T3l, T2V, T3o; T3g = VSUB(T3e, T3f); T3l = VADD(T3e, T3f); T2V = VADD(T27, T2U); T3o = VSUB(T27, T2U); T3F = VFNMS(LDK(KP923879532), T3E, T3D); T3P = VFMA(LDK(KP923879532), T3E, T3D); T3x = VFMA(LDK(KP668178637), T3w, T3v); T3G = VFNMS(LDK(KP668178637), T3v, T3w); T3q = VFMA(LDK(KP980785280), T3l, T3k); T3m = VFNMS(LDK(KP980785280), T3l, T3k); T3h = VFNMS(LDK(KP980785280), T3g, T3d); T3j = VFMA(LDK(KP980785280), T3g, T3d); T3r = VFNMS(LDK(KP980785280), T3o, T3n); T3p = VFMA(LDK(KP980785280), T3o, T3n); T2W = VFNMS(LDK(KP980785280), T2V, T1k); T3i = VFMA(LDK(KP980785280), T2V, T1k); } } { V T7n, T7Z, T8j, T89, T7k, T7O, T8g, T7Y, T7H, T7R, T80, T7q, T7u, T82, T83; V T7x; { V T7c, T7W, T7D, T87, T7f, T7E, T3A, T3H, T7F, T7i; T7c = VFNMS(LDK(KP923879532), T7b, T7a); T7W = VFMA(LDK(KP923879532), T7b, T7a); T7D = VFMA(LDK(KP923879532), T7C, T7B); T87 = VFNMS(LDK(KP923879532), T7C, T7B); T7f = VFNMS(LDK(KP668178637), T7e, T7d); T7E = VFMA(LDK(KP668178637), T7d, T7e); ST(&(x[WS(rs, 46)]), VFNMSI(T3p, T3m), ms, &(x[0])); ST(&(x[WS(rs, 18)]), VFMAI(T3p, T3m), ms, &(x[0])); ST(&(x[WS(rs, 50)]), VFMAI(T3r, T3q), ms, &(x[0])); ST(&(x[WS(rs, 14)]), VFNMSI(T3r, T3q), ms, &(x[0])); ST(&(x[WS(rs, 2)]), VFMAI(T3j, T3i), ms, &(x[0])); ST(&(x[WS(rs, 62)]), VFNMSI(T3j, T3i), ms, &(x[0])); ST(&(x[WS(rs, 34)]), VFMAI(T3h, T2W), ms, &(x[0])); ST(&(x[WS(rs, 30)]), VFNMSI(T3h, T2W), ms, &(x[0])); T3A = VFMA(LDK(KP668178637), T3z, T3y); T3H = VFNMS(LDK(KP668178637), T3y, T3z); T7F = VFMA(LDK(KP668178637), T7g, T7h); T7i = VFNMS(LDK(KP668178637), T7h, T7g); T7n = VFNMS(LDK(KP923879532), T7m, T7l); T7Z = VFMA(LDK(KP923879532), T7m, T7l); { V T3I, T3N, T3B, T3Q; T3I = VSUB(T3G, T3H); T3N = VADD(T3G, T3H); T3B = VADD(T3x, T3A); T3Q = VSUB(T3x, T3A); { V T7j, T88, T7G, T7X; T7j = VADD(T7f, T7i); T88 = VSUB(T7f, T7i); T7G = VSUB(T7E, T7F); T7X = VADD(T7E, T7F); { V T3S, T3O, T3J, T3L; T3S = VFNMS(LDK(KP831469612), T3N, T3M); T3O = VFMA(LDK(KP831469612), T3N, T3M); T3J = VFNMS(LDK(KP831469612), T3I, T3F); T3L = VFMA(LDK(KP831469612), T3I, T3F); { V T3T, T3R, T3C, T3K; T3T = VFMA(LDK(KP831469612), T3Q, T3P); T3R = VFNMS(LDK(KP831469612), T3Q, T3P); T3C = VFNMS(LDK(KP831469612), T3B, T3u); T3K = VFMA(LDK(KP831469612), T3B, T3u); T8j = VFNMS(LDK(KP831469612), T88, T87); T89 = VFMA(LDK(KP831469612), T88, T87); T7k = VFNMS(LDK(KP831469612), T7j, T7c); T7O = VFMA(LDK(KP831469612), T7j, T7c); T8g = VFNMS(LDK(KP831469612), T7X, T7W); T7Y = VFMA(LDK(KP831469612), T7X, T7W); T7H = VFMA(LDK(KP831469612), T7G, T7D); T7R = VFNMS(LDK(KP831469612), T7G, T7D); ST(&(x[WS(rs, 42)]), VFMAI(T3R, T3O), ms, &(x[0])); ST(&(x[WS(rs, 22)]), VFNMSI(T3R, T3O), ms, &(x[0])); ST(&(x[WS(rs, 54)]), VFNMSI(T3T, T3S), ms, &(x[0])); ST(&(x[WS(rs, 10)]), VFMAI(T3T, T3S), ms, &(x[0])); ST(&(x[WS(rs, 58)]), VFMAI(T3L, T3K), ms, &(x[0])); ST(&(x[WS(rs, 6)]), VFNMSI(T3L, T3K), ms, &(x[0])); ST(&(x[WS(rs, 26)]), VFMAI(T3J, T3C), ms, &(x[0])); ST(&(x[WS(rs, 38)]), VFNMSI(T3J, T3C), ms, &(x[0])); T80 = VFNMS(LDK(KP923879532), T7p, T7o); T7q = VFMA(LDK(KP923879532), T7p, T7o); } } } } T7u = VFNMS(LDK(KP923879532), T7t, T7s); T82 = VFMA(LDK(KP923879532), T7t, T7s); T83 = VFNMS(LDK(KP923879532), T7w, T7v); T7x = VFMA(LDK(KP923879532), T7w, T7v); } { V T5g, T6I, T6p, T6T, T5p, T6q, T6r, T5y; T5g = VFMA(LDK(KP923879532), T5f, T58); T6I = VFNMS(LDK(KP923879532), T5f, T58); { V T7r, T7I, T7y, T7J; T7r = VFNMS(LDK(KP534511135), T7q, T7n); T7I = VFMA(LDK(KP534511135), T7n, T7q); T7y = VFNMS(LDK(KP534511135), T7x, T7u); T7J = VFMA(LDK(KP534511135), T7u, T7x); { V T81, T8a, T84, T8b; T81 = VFMA(LDK(KP303346683), T80, T7Z); T8a = VFNMS(LDK(KP303346683), T7Z, T80); T84 = VFMA(LDK(KP303346683), T83, T82); T8b = VFNMS(LDK(KP303346683), T82, T83); T6p = VFMA(LDK(KP923879532), T6o, T6l); T6T = VFNMS(LDK(KP923879532), T6o, T6l); T5p = VFNMS(LDK(KP198912367), T5o, T5l); T6q = VFMA(LDK(KP198912367), T5l, T5o); { V T7K, T7P, T7z, T7S; T7K = VSUB(T7I, T7J); T7P = VADD(T7I, T7J); T7z = VADD(T7r, T7y); T7S = VSUB(T7r, T7y); { V T8c, T8h, T85, T8k; T8c = VSUB(T8a, T8b); T8h = VADD(T8a, T8b); T85 = VADD(T81, T84); T8k = VSUB(T81, T84); { V T7Q, T7U, T7L, T7N; T7Q = VFNMS(LDK(KP881921264), T7P, T7O); T7U = VFMA(LDK(KP881921264), T7P, T7O); T7L = VFNMS(LDK(KP881921264), T7K, T7H); T7N = VFMA(LDK(KP881921264), T7K, T7H); { V T7T, T7V, T7A, T7M; T7T = VFMA(LDK(KP881921264), T7S, T7R); T7V = VFNMS(LDK(KP881921264), T7S, T7R); T7A = VFNMS(LDK(KP881921264), T7z, T7k); T7M = VFMA(LDK(KP881921264), T7z, T7k); { V T8i, T8m, T8d, T8f; T8i = VFMA(LDK(KP956940335), T8h, T8g); T8m = VFNMS(LDK(KP956940335), T8h, T8g); T8d = VFNMS(LDK(KP956940335), T8c, T89); T8f = VFMA(LDK(KP956940335), T8c, T89); { V T8l, T8n, T86, T8e; T8l = VFNMS(LDK(KP956940335), T8k, T8j); T8n = VFMA(LDK(KP956940335), T8k, T8j); T86 = VFNMS(LDK(KP956940335), T85, T7Y); T8e = VFMA(LDK(KP956940335), T85, T7Y); ST(&(x[WS(rs, 53)]), VFMAI(T7V, T7U), ms, &(x[WS(rs, 1)])); ST(&(x[WS(rs, 11)]), VFNMSI(T7V, T7U), ms, &(x[WS(rs, 1)])); ST(&(x[WS(rs, 43)]), VFNMSI(T7T, T7Q), ms, &(x[WS(rs, 1)])); ST(&(x[WS(rs, 21)]), VFMAI(T7T, T7Q), ms, &(x[WS(rs, 1)])); ST(&(x[WS(rs, 5)]), VFMAI(T7N, T7M), ms, &(x[WS(rs, 1)])); ST(&(x[WS(rs, 59)]), VFNMSI(T7N, T7M), ms, &(x[WS(rs, 1)])); ST(&(x[WS(rs, 37)]), VFMAI(T7L, T7A), ms, &(x[WS(rs, 1)])); ST(&(x[WS(rs, 27)]), VFNMSI(T7L, T7A), ms, &(x[WS(rs, 1)])); ST(&(x[WS(rs, 51)]), VFNMSI(T8n, T8m), ms, &(x[WS(rs, 1)])); ST(&(x[WS(rs, 13)]), VFMAI(T8n, T8m), ms, &(x[WS(rs, 1)])); ST(&(x[WS(rs, 45)]), VFMAI(T8l, T8i), ms, &(x[WS(rs, 1)])); ST(&(x[WS(rs, 19)]), VFNMSI(T8l, T8i), ms, &(x[WS(rs, 1)])); ST(&(x[WS(rs, 61)]), VFMAI(T8f, T8e), ms, &(x[WS(rs, 1)])); ST(&(x[WS(rs, 3)]), VFNMSI(T8f, T8e), ms, &(x[WS(rs, 1)])); ST(&(x[WS(rs, 29)]), VFMAI(T8d, T86), ms, &(x[WS(rs, 1)])); ST(&(x[WS(rs, 35)]), VFNMSI(T8d, T86), ms, &(x[WS(rs, 1)])); T6r = VFMA(LDK(KP198912367), T5u, T5x); T5y = VFNMS(LDK(KP198912367), T5x, T5u); } } } } } } } } { V T5N, T5U, T68, T5z, T6U, T6f; T5N = VFMA(LDK(KP923879532), T5M, T5F); T6L = VFNMS(LDK(KP923879532), T5M, T5F); T6M = VFNMS(LDK(KP923879532), T5T, T5Q); T5U = VFMA(LDK(KP923879532), T5T, T5Q); T68 = VFMA(LDK(KP923879532), T67, T60); T6O = VFNMS(LDK(KP923879532), T67, T60); T5z = VADD(T5p, T5y); T6U = VSUB(T5p, T5y); T6P = VFNMS(LDK(KP923879532), T6e, T6b); T6f = VFMA(LDK(KP923879532), T6e, T6b); { V T5V, T6u, T6g, T6v, T6s, T6J; T6s = VSUB(T6q, T6r); T6J = VADD(T6q, T6r); T5V = VFNMS(LDK(KP098491403), T5U, T5N); T6u = VFMA(LDK(KP098491403), T5N, T5U); T75 = VFMA(LDK(KP980785280), T6U, T6T); T6V = VFNMS(LDK(KP980785280), T6U, T6T); T5A = VFMA(LDK(KP980785280), T5z, T5g); T6A = VFNMS(LDK(KP980785280), T5z, T5g); T6g = VFNMS(LDK(KP098491403), T6f, T68); T6v = VFMA(LDK(KP098491403), T68, T6f); T72 = VFNMS(LDK(KP980785280), T6J, T6I); T6K = VFMA(LDK(KP980785280), T6J, T6I); T6t = VFMA(LDK(KP980785280), T6s, T6p); T6D = VFNMS(LDK(KP980785280), T6s, T6p); T6w = VSUB(T6u, T6v); T6B = VADD(T6u, T6v); T6h = VADD(T5V, T6g); T6E = VSUB(T5V, T6g); } } } } } } } { V T6W, T6N, T6G, T6C, T6z, T6x, T6H, T6F, T6y, T6i, T6X, T6Q; T6W = VFNMS(LDK(KP820678790), T6L, T6M); T6N = VFMA(LDK(KP820678790), T6M, T6L); T6G = VFMA(LDK(KP995184726), T6B, T6A); T6C = VFNMS(LDK(KP995184726), T6B, T6A); T6z = VFMA(LDK(KP995184726), T6w, T6t); T6x = VFNMS(LDK(KP995184726), T6w, T6t); T6H = VFNMS(LDK(KP995184726), T6E, T6D); T6F = VFMA(LDK(KP995184726), T6E, T6D); T6y = VFMA(LDK(KP995184726), T6h, T5A); T6i = VFNMS(LDK(KP995184726), T6h, T5A); T6X = VFNMS(LDK(KP820678790), T6O, T6P); T6Q = VFMA(LDK(KP820678790), T6P, T6O); { V T73, T6Y, T76, T6R; ST(&(x[WS(rs, 49)]), VFMAI(T6H, T6G), ms, &(x[WS(rs, 1)])); ST(&(x[WS(rs, 15)]), VFNMSI(T6H, T6G), ms, &(x[WS(rs, 1)])); ST(&(x[WS(rs, 47)]), VFNMSI(T6F, T6C), ms, &(x[WS(rs, 1)])); ST(&(x[WS(rs, 17)]), VFMAI(T6F, T6C), ms, &(x[WS(rs, 1)])); ST(&(x[WS(rs, 1)]), VFMAI(T6z, T6y), ms, &(x[WS(rs, 1)])); ST(&(x[WS(rs, 63)]), VFNMSI(T6z, T6y), ms, &(x[WS(rs, 1)])); ST(&(x[WS(rs, 33)]), VFMAI(T6x, T6i), ms, &(x[WS(rs, 1)])); ST(&(x[WS(rs, 31)]), VFNMSI(T6x, T6i), ms, &(x[WS(rs, 1)])); T73 = VADD(T6W, T6X); T6Y = VSUB(T6W, T6X); T76 = VSUB(T6N, T6Q); T6R = VADD(T6N, T6Q); { V T78, T74, T71, T6Z, T79, T77, T70, T6S; T78 = VFNMS(LDK(KP773010453), T73, T72); T74 = VFMA(LDK(KP773010453), T73, T72); T71 = VFMA(LDK(KP773010453), T6Y, T6V); T6Z = VFNMS(LDK(KP773010453), T6Y, T6V); T79 = VFMA(LDK(KP773010453), T76, T75); T77 = VFNMS(LDK(KP773010453), T76, T75); T70 = VFMA(LDK(KP773010453), T6R, T6K); T6S = VFNMS(LDK(KP773010453), T6R, T6K); ST(&(x[WS(rs, 55)]), VFNMSI(T79, T78), ms, &(x[WS(rs, 1)])); ST(&(x[WS(rs, 9)]), VFMAI(T79, T78), ms, &(x[WS(rs, 1)])); ST(&(x[WS(rs, 41)]), VFMAI(T77, T74), ms, &(x[WS(rs, 1)])); ST(&(x[WS(rs, 23)]), VFNMSI(T77, T74), ms, &(x[WS(rs, 1)])); ST(&(x[WS(rs, 57)]), VFMAI(T71, T70), ms, &(x[WS(rs, 1)])); ST(&(x[WS(rs, 7)]), VFNMSI(T71, T70), ms, &(x[WS(rs, 1)])); ST(&(x[WS(rs, 25)]), VFMAI(T6Z, T6S), ms, &(x[WS(rs, 1)])); ST(&(x[WS(rs, 39)]), VFNMSI(T6Z, T6S), ms, &(x[WS(rs, 1)])); } } } } } VLEAVE(); } static const tw_instr twinstr[] = { VTW(0, 1), VTW(0, 2), VTW(0, 3), VTW(0, 4), VTW(0, 5), VTW(0, 6), VTW(0, 7), VTW(0, 8), VTW(0, 9), VTW(0, 10), VTW(0, 11), VTW(0, 12), VTW(0, 13), VTW(0, 14), VTW(0, 15), VTW(0, 16), VTW(0, 17), VTW(0, 18), VTW(0, 19), VTW(0, 20), VTW(0, 21), VTW(0, 22), VTW(0, 23), VTW(0, 24), VTW(0, 25), VTW(0, 26), VTW(0, 27), VTW(0, 28), VTW(0, 29), VTW(0, 30), VTW(0, 31), VTW(0, 32), VTW(0, 33), VTW(0, 34), VTW(0, 35), VTW(0, 36), VTW(0, 37), VTW(0, 38), VTW(0, 39), VTW(0, 40), VTW(0, 41), VTW(0, 42), VTW(0, 43), VTW(0, 44), VTW(0, 45), VTW(0, 46), VTW(0, 47), VTW(0, 48), VTW(0, 49), VTW(0, 50), VTW(0, 51), VTW(0, 52), VTW(0, 53), VTW(0, 54), VTW(0, 55), VTW(0, 56), VTW(0, 57), VTW(0, 58), VTW(0, 59), VTW(0, 60), VTW(0, 61), VTW(0, 62), VTW(0, 63), {TW_NEXT, VL, 0} }; static const ct_desc desc = { 64, XSIMD_STRING("t1bv_64"), twinstr, &GENUS, {261, 126, 258, 0}, 0, 0, 0 }; void XSIMD(codelet_t1bv_64) (planner *p) { X(kdft_dit_register) (p, t1bv_64, &desc); } #else /* HAVE_FMA */ /* Generated by: ../../../genfft/gen_twiddle_c.native -simd -compact -variables 4 -pipeline-latency 8 -n 64 -name t1bv_64 -include t1b.h -sign 1 */ /* * This function contains 519 FP additions, 250 FP multiplications, * (or, 467 additions, 198 multiplications, 52 fused multiply/add), * 107 stack variables, 15 constants, and 128 memory accesses */ #include "t1b.h" static void t1bv_64(R *ri, R *ii, const R *W, stride rs, INT mb, INT me, INT ms) { DVK(KP290284677, +0.290284677254462367636192375817395274691476278); DVK(KP956940335, +0.956940335732208864935797886980269969482849206); DVK(KP471396736, +0.471396736825997648556387625905254377657460319); DVK(KP881921264, +0.881921264348355029712756863660388349508442621); DVK(KP634393284, +0.634393284163645498215171613225493370675687095); DVK(KP773010453, +0.773010453362736960810906609758469800971041293); DVK(KP098017140, +0.098017140329560601994195563888641845861136673); DVK(KP995184726, +0.995184726672196886244836953109479921575474869); DVK(KP195090322, +0.195090322016128267848284868477022240927691618); DVK(KP980785280, +0.980785280403230449126182236134239036973933731); DVK(KP555570233, +0.555570233019602224742830813948532874374937191); DVK(KP831469612, +0.831469612302545237078788377617905756738560812); DVK(KP382683432, +0.382683432365089771728459984030398866761344562); DVK(KP923879532, +0.923879532511286756128183189396788286822416626); DVK(KP707106781, +0.707106781186547524400844362104849039284835938); { INT m; R *x; x = ii; for (m = mb, W = W + (mb * ((TWVL / VL) * 126)); m < me; m = m + VL, x = x + (VL * ms), W = W + (TWVL * 126), MAKE_VOLATILE_STRIDE(64, rs)) { V Tg, T4B, T6v, T7G, T3r, T4w, T5q, T7F, T5Y, T62, T28, T4d, T2g, T4a, T7g; V T7Y, T6f, T6j, T2Z, T4k, T37, T4h, T7n, T81, T7w, T7x, T7y, T5M, T6q, T1k; V T4s, T1r, T4t, T7t, T7u, T7v, T5F, T6p, TV, T4p, T12, T4q, T7A, T7B, TD; V T4x, T3k, T4C, T5x, T6s, T1R, T4b, T7j, T7Z, T2j, T4e, T5V, T63, T2I, T4i; V T7q, T82, T3a, T4l, T6c, T6k; { V T1, T3, T3p, T3n, Tb, Td, Te, T6, T8, T9, T2, T3o, T3m; T1 = LD(&(x[0]), ms, &(x[0])); T2 = LD(&(x[WS(rs, 32)]), ms, &(x[0])); T3 = BYTW(&(W[TWVL * 62]), T2); T3o = LD(&(x[WS(rs, 48)]), ms, &(x[0])); T3p = BYTW(&(W[TWVL * 94]), T3o); T3m = LD(&(x[WS(rs, 16)]), ms, &(x[0])); T3n = BYTW(&(W[TWVL * 30]), T3m); { V Ta, Tc, T5, T7; Ta = LD(&(x[WS(rs, 56)]), ms, &(x[0])); Tb = BYTW(&(W[TWVL * 110]), Ta); Tc = LD(&(x[WS(rs, 24)]), ms, &(x[0])); Td = BYTW(&(W[TWVL * 46]), Tc); Te = VSUB(Tb, Td); T5 = LD(&(x[WS(rs, 8)]), ms, &(x[0])); T6 = BYTW(&(W[TWVL * 14]), T5); T7 = LD(&(x[WS(rs, 40)]), ms, &(x[0])); T8 = BYTW(&(W[TWVL * 78]), T7); T9 = VSUB(T6, T8); } { V T4, Tf, T6t, T6u; T4 = VSUB(T1, T3); Tf = VMUL(LDK(KP707106781), VADD(T9, Te)); Tg = VSUB(T4, Tf); T4B = VADD(T4, Tf); T6t = VADD(T6, T8); T6u = VADD(Tb, Td); T6v = VSUB(T6t, T6u); T7G = VADD(T6t, T6u); } { V T3l, T3q, T5o, T5p; T3l = VMUL(LDK(KP707106781), VSUB(T9, Te)); T3q = VSUB(T3n, T3p); T3r = VSUB(T3l, T3q); T4w = VADD(T3q, T3l); T5o = VADD(T1, T3); T5p = VADD(T3n, T3p); T5q = VSUB(T5o, T5p); T7F = VADD(T5o, T5p); } } { V T24, T26, T61, T2b, T2d, T60, T1W, T5W, T21, T5X, T22, T27; { V T23, T25, T2a, T2c; T23 = LD(&(x[WS(rs, 17)]), ms, &(x[WS(rs, 1)])); T24 = BYTW(&(W[TWVL * 32]), T23); T25 = LD(&(x[WS(rs, 49)]), ms, &(x[WS(rs, 1)])); T26 = BYTW(&(W[TWVL * 96]), T25); T61 = VADD(T24, T26); T2a = LD(&(x[WS(rs, 1)]), ms, &(x[WS(rs, 1)])); T2b = BYTW(&(W[0]), T2a); T2c = LD(&(x[WS(rs, 33)]), ms, &(x[WS(rs, 1)])); T2d = BYTW(&(W[TWVL * 64]), T2c); T60 = VADD(T2b, T2d); } { V T1T, T1V, T1S, T1U; T1S = LD(&(x[WS(rs, 9)]), ms, &(x[WS(rs, 1)])); T1T = BYTW(&(W[TWVL * 16]), T1S); T1U = LD(&(x[WS(rs, 41)]), ms, &(x[WS(rs, 1)])); T1V = BYTW(&(W[TWVL * 80]), T1U); T1W = VSUB(T1T, T1V); T5W = VADD(T1T, T1V); } { V T1Y, T20, T1X, T1Z; T1X = LD(&(x[WS(rs, 57)]), ms, &(x[WS(rs, 1)])); T1Y = BYTW(&(W[TWVL * 112]), T1X); T1Z = LD(&(x[WS(rs, 25)]), ms, &(x[WS(rs, 1)])); T20 = BYTW(&(W[TWVL * 48]), T1Z); T21 = VSUB(T1Y, T20); T5X = VADD(T1Y, T20); } T5Y = VSUB(T5W, T5X); T62 = VSUB(T60, T61); T22 = VMUL(LDK(KP707106781), VSUB(T1W, T21)); T27 = VSUB(T24, T26); T28 = VSUB(T22, T27); T4d = VADD(T27, T22); { V T2e, T2f, T7e, T7f; T2e = VSUB(T2b, T2d); T2f = VMUL(LDK(KP707106781), VADD(T1W, T21)); T2g = VSUB(T2e, T2f); T4a = VADD(T2e, T2f); T7e = VADD(T60, T61); T7f = VADD(T5W, T5X); T7g = VSUB(T7e, T7f); T7Y = VADD(T7e, T7f); } } { V T2V, T2X, T6i, T32, T34, T6h, T2N, T6d, T2S, T6e, T2T, T2Y; { V T2U, T2W, T31, T33; T2U = LD(&(x[WS(rs, 15)]), ms, &(x[WS(rs, 1)])); T2V = BYTW(&(W[TWVL * 28]), T2U); T2W = LD(&(x[WS(rs, 47)]), ms, &(x[WS(rs, 1)])); T2X = BYTW(&(W[TWVL * 92]), T2W); T6i = VADD(T2V, T2X); T31 = LD(&(x[WS(rs, 63)]), ms, &(x[WS(rs, 1)])); T32 = BYTW(&(W[TWVL * 124]), T31); T33 = LD(&(x[WS(rs, 31)]), ms, &(x[WS(rs, 1)])); T34 = BYTW(&(W[TWVL * 60]), T33); T6h = VADD(T32, T34); } { V T2K, T2M, T2J, T2L; T2J = LD(&(x[WS(rs, 7)]), ms, &(x[WS(rs, 1)])); T2K = BYTW(&(W[TWVL * 12]), T2J); T2L = LD(&(x[WS(rs, 39)]), ms, &(x[WS(rs, 1)])); T2M = BYTW(&(W[TWVL * 76]), T2L); T2N = VSUB(T2K, T2M); T6d = VADD(T2K, T2M); } { V T2P, T2R, T2O, T2Q; T2O = LD(&(x[WS(rs, 55)]), ms, &(x[WS(rs, 1)])); T2P = BYTW(&(W[TWVL * 108]), T2O); T2Q = LD(&(x[WS(rs, 23)]), ms, &(x[WS(rs, 1)])); T2R = BYTW(&(W[TWVL * 44]), T2Q); T2S = VSUB(T2P, T2R); T6e = VADD(T2P, T2R); } T6f = VSUB(T6d, T6e); T6j = VSUB(T6h, T6i); T2T = VMUL(LDK(KP707106781), VSUB(T2N, T2S)); T2Y = VSUB(T2V, T2X); T2Z = VSUB(T2T, T2Y); T4k = VADD(T2Y, T2T); { V T35, T36, T7l, T7m; T35 = VSUB(T32, T34); T36 = VMUL(LDK(KP707106781), VADD(T2N, T2S)); T37 = VSUB(T35, T36); T4h = VADD(T35, T36); T7l = VADD(T6h, T6i); T7m = VADD(T6d, T6e); T7n = VSUB(T7l, T7m); T81 = VADD(T7l, T7m); } } { V T1g, T1i, T5K, T1m, T1o, T5J, T18, T5G, T1d, T5H, T5I, T5L; { V T1f, T1h, T1l, T1n; T1f = LD(&(x[WS(rs, 14)]), ms, &(x[0])); T1g = BYTW(&(W[TWVL * 26]), T1f); T1h = LD(&(x[WS(rs, 46)]), ms, &(x[0])); T1i = BYTW(&(W[TWVL * 90]), T1h); T5K = VADD(T1g, T1i); T1l = LD(&(x[WS(rs, 62)]), ms, &(x[0])); T1m = BYTW(&(W[TWVL * 122]), T1l); T1n = LD(&(x[WS(rs, 30)]), ms, &(x[0])); T1o = BYTW(&(W[TWVL * 58]), T1n); T5J = VADD(T1m, T1o); } { V T15, T17, T14, T16; T14 = LD(&(x[WS(rs, 6)]), ms, &(x[0])); T15 = BYTW(&(W[TWVL * 10]), T14); T16 = LD(&(x[WS(rs, 38)]), ms, &(x[0])); T17 = BYTW(&(W[TWVL * 74]), T16); T18 = VSUB(T15, T17); T5G = VADD(T15, T17); } { V T1a, T1c, T19, T1b; T19 = LD(&(x[WS(rs, 54)]), ms, &(x[0])); T1a = BYTW(&(W[TWVL * 106]), T19); T1b = LD(&(x[WS(rs, 22)]), ms, &(x[0])); T1c = BYTW(&(W[TWVL * 42]), T1b); T1d = VSUB(T1a, T1c); T5H = VADD(T1a, T1c); } T7w = VADD(T5J, T5K); T7x = VADD(T5G, T5H); T7y = VSUB(T7w, T7x); T5I = VSUB(T5G, T5H); T5L = VSUB(T5J, T5K); T5M = VFNMS(LDK(KP382683432), T5L, VMUL(LDK(KP923879532), T5I)); T6q = VFMA(LDK(KP923879532), T5L, VMUL(LDK(KP382683432), T5I)); { V T1e, T1j, T1p, T1q; T1e = VMUL(LDK(KP707106781), VSUB(T18, T1d)); T1j = VSUB(T1g, T1i); T1k = VSUB(T1e, T1j); T4s = VADD(T1j, T1e); T1p = VSUB(T1m, T1o); T1q = VMUL(LDK(KP707106781), VADD(T18, T1d)); T1r = VSUB(T1p, T1q); T4t = VADD(T1p, T1q); } } { V TR, TT, T5A, TX, TZ, T5z, TJ, T5C, TO, T5D, T5B, T5E; { V TQ, TS, TW, TY; TQ = LD(&(x[WS(rs, 18)]), ms, &(x[0])); TR = BYTW(&(W[TWVL * 34]), TQ); TS = LD(&(x[WS(rs, 50)]), ms, &(x[0])); TT = BYTW(&(W[TWVL * 98]), TS); T5A = VADD(TR, TT); TW = LD(&(x[WS(rs, 2)]), ms, &(x[0])); TX = BYTW(&(W[TWVL * 2]), TW); TY = LD(&(x[WS(rs, 34)]), ms, &(x[0])); TZ = BYTW(&(W[TWVL * 66]), TY); T5z = VADD(TX, TZ); } { V TG, TI, TF, TH; TF = LD(&(x[WS(rs, 10)]), ms, &(x[0])); TG = BYTW(&(W[TWVL * 18]), TF); TH = LD(&(x[WS(rs, 42)]), ms, &(x[0])); TI = BYTW(&(W[TWVL * 82]), TH); TJ = VSUB(TG, TI); T5C = VADD(TG, TI); } { V TL, TN, TK, TM; TK = LD(&(x[WS(rs, 58)]), ms, &(x[0])); TL = BYTW(&(W[TWVL * 114]), TK); TM = LD(&(x[WS(rs, 26)]), ms, &(x[0])); TN = BYTW(&(W[TWVL * 50]), TM); TO = VSUB(TL, TN); T5D = VADD(TL, TN); } T7t = VADD(T5z, T5A); T7u = VADD(T5C, T5D); T7v = VSUB(T7t, T7u); T5B = VSUB(T5z, T5A); T5E = VSUB(T5C, T5D); T5F = VFMA(LDK(KP382683432), T5B, VMUL(LDK(KP923879532), T5E)); T6p = VFNMS(LDK(KP382683432), T5E, VMUL(LDK(KP923879532), T5B)); { V TP, TU, T10, T11; TP = VMUL(LDK(KP707106781), VSUB(TJ, TO)); TU = VSUB(TR, TT); TV = VSUB(TP, TU); T4p = VADD(TU, TP); T10 = VSUB(TX, TZ); T11 = VMUL(LDK(KP707106781), VADD(TJ, TO)); T12 = VSUB(T10, T11); T4q = VADD(T10, T11); } } { V Tl, T5r, TB, T5u, Tq, T5s, Tw, T5v, Tr, TC; { V Ti, Tk, Th, Tj; Th = LD(&(x[WS(rs, 4)]), ms, &(x[0])); Ti = BYTW(&(W[TWVL * 6]), Th); Tj = LD(&(x[WS(rs, 36)]), ms, &(x[0])); Tk = BYTW(&(W[TWVL * 70]), Tj); Tl = VSUB(Ti, Tk); T5r = VADD(Ti, Tk); } { V Ty, TA, Tx, Tz; Tx = LD(&(x[WS(rs, 60)]), ms, &(x[0])); Ty = BYTW(&(W[TWVL * 118]), Tx); Tz = LD(&(x[WS(rs, 28)]), ms, &(x[0])); TA = BYTW(&(W[TWVL * 54]), Tz); TB = VSUB(Ty, TA); T5u = VADD(Ty, TA); } { V Tn, Tp, Tm, To; Tm = LD(&(x[WS(rs, 20)]), ms, &(x[0])); Tn = BYTW(&(W[TWVL * 38]), Tm); To = LD(&(x[WS(rs, 52)]), ms, &(x[0])); Tp = BYTW(&(W[TWVL * 102]), To); Tq = VSUB(Tn, Tp); T5s = VADD(Tn, Tp); } { V Tt, Tv, Ts, Tu; Ts = LD(&(x[WS(rs, 12)]), ms, &(x[0])); Tt = BYTW(&(W[TWVL * 22]), Ts); Tu = LD(&(x[WS(rs, 44)]), ms, &(x[0])); Tv = BYTW(&(W[TWVL * 86]), Tu); Tw = VSUB(Tt, Tv); T5v = VADD(Tt, Tv); } T7A = VADD(T5r, T5s); T7B = VADD(T5u, T5v); Tr = VFMA(LDK(KP382683432), Tl, VMUL(LDK(KP923879532), Tq)); TC = VFNMS(LDK(KP382683432), TB, VMUL(LDK(KP923879532), Tw)); TD = VSUB(Tr, TC); T4x = VADD(Tr, TC); { V T3i, T3j, T5t, T5w; T3i = VFNMS(LDK(KP382683432), Tq, VMUL(LDK(KP923879532), Tl)); T3j = VFMA(LDK(KP923879532), TB, VMUL(LDK(KP382683432), Tw)); T3k = VSUB(T3i, T3j); T4C = VADD(T3i, T3j); T5t = VSUB(T5r, T5s); T5w = VSUB(T5u, T5v); T5x = VMUL(LDK(KP707106781), VADD(T5t, T5w)); T6s = VMUL(LDK(KP707106781), VSUB(T5t, T5w)); } } { V T1z, T5P, T1P, T5T, T1E, T5Q, T1K, T5S; { V T1w, T1y, T1v, T1x; T1v = LD(&(x[WS(rs, 5)]), ms, &(x[WS(rs, 1)])); T1w = BYTW(&(W[TWVL * 8]), T1v); T1x = LD(&(x[WS(rs, 37)]), ms, &(x[WS(rs, 1)])); T1y = BYTW(&(W[TWVL * 72]), T1x); T1z = VSUB(T1w, T1y); T5P = VADD(T1w, T1y); } { V T1M, T1O, T1L, T1N; T1L = LD(&(x[WS(rs, 13)]), ms, &(x[WS(rs, 1)])); T1M = BYTW(&(W[TWVL * 24]), T1L); T1N = LD(&(x[WS(rs, 45)]), ms, &(x[WS(rs, 1)])); T1O = BYTW(&(W[TWVL * 88]), T1N); T1P = VSUB(T1M, T1O); T5T = VADD(T1M, T1O); } { V T1B, T1D, T1A, T1C; T1A = LD(&(x[WS(rs, 21)]), ms, &(x[WS(rs, 1)])); T1B = BYTW(&(W[TWVL * 40]), T1A); T1C = LD(&(x[WS(rs, 53)]), ms, &(x[WS(rs, 1)])); T1D = BYTW(&(W[TWVL * 104]), T1C); T1E = VSUB(T1B, T1D); T5Q = VADD(T1B, T1D); } { V T1H, T1J, T1G, T1I; T1G = LD(&(x[WS(rs, 61)]), ms, &(x[WS(rs, 1)])); T1H = BYTW(&(W[TWVL * 120]), T1G); T1I = LD(&(x[WS(rs, 29)]), ms, &(x[WS(rs, 1)])); T1J = BYTW(&(W[TWVL * 56]), T1I); T1K = VSUB(T1H, T1J); T5S = VADD(T1H, T1J); } { V T1F, T1Q, T7h, T7i; T1F = VFNMS(LDK(KP382683432), T1E, VMUL(LDK(KP923879532), T1z)); T1Q = VFMA(LDK(KP923879532), T1K, VMUL(LDK(KP382683432), T1P)); T1R = VSUB(T1F, T1Q); T4b = VADD(T1F, T1Q); T7h = VADD(T5P, T5Q); T7i = VADD(T5S, T5T); T7j = VSUB(T7h, T7i); T7Z = VADD(T7h, T7i); } { V T2h, T2i, T5R, T5U; T2h = VFMA(LDK(KP382683432), T1z, VMUL(LDK(KP923879532), T1E)); T2i = VFNMS(LDK(KP382683432), T1K, VMUL(LDK(KP923879532), T1P)); T2j = VSUB(T2h, T2i); T4e = VADD(T2h, T2i); T5R = VSUB(T5P, T5Q); T5U = VSUB(T5S, T5T); T5V = VMUL(LDK(KP707106781), VSUB(T5R, T5U)); T63 = VMUL(LDK(KP707106781), VADD(T5R, T5U)); } } { V T2q, T66, T2G, T6a, T2v, T67, T2B, T69; { V T2n, T2p, T2m, T2o; T2m = LD(&(x[WS(rs, 3)]), ms, &(x[WS(rs, 1)])); T2n = BYTW(&(W[TWVL * 4]), T2m); T2o = LD(&(x[WS(rs, 35)]), ms, &(x[WS(rs, 1)])); T2p = BYTW(&(W[TWVL * 68]), T2o); T2q = VSUB(T2n, T2p); T66 = VADD(T2n, T2p); } { V T2D, T2F, T2C, T2E; T2C = LD(&(x[WS(rs, 11)]), ms, &(x[WS(rs, 1)])); T2D = BYTW(&(W[TWVL * 20]), T2C); T2E = LD(&(x[WS(rs, 43)]), ms, &(x[WS(rs, 1)])); T2F = BYTW(&(W[TWVL * 84]), T2E); T2G = VSUB(T2D, T2F); T6a = VADD(T2D, T2F); } { V T2s, T2u, T2r, T2t; T2r = LD(&(x[WS(rs, 19)]), ms, &(x[WS(rs, 1)])); T2s = BYTW(&(W[TWVL * 36]), T2r); T2t = LD(&(x[WS(rs, 51)]), ms, &(x[WS(rs, 1)])); T2u = BYTW(&(W[TWVL * 100]), T2t); T2v = VSUB(T2s, T2u); T67 = VADD(T2s, T2u); } { V T2y, T2A, T2x, T2z; T2x = LD(&(x[WS(rs, 59)]), ms, &(x[WS(rs, 1)])); T2y = BYTW(&(W[TWVL * 116]), T2x); T2z = LD(&(x[WS(rs, 27)]), ms, &(x[WS(rs, 1)])); T2A = BYTW(&(W[TWVL * 52]), T2z); T2B = VSUB(T2y, T2A); T69 = VADD(T2y, T2A); } { V T2w, T2H, T7o, T7p; T2w = VFNMS(LDK(KP382683432), T2v, VMUL(LDK(KP923879532), T2q)); T2H = VFMA(LDK(KP923879532), T2B, VMUL(LDK(KP382683432), T2G)); T2I = VSUB(T2w, T2H); T4i = VADD(T2w, T2H); T7o = VADD(T66, T67); T7p = VADD(T69, T6a); T7q = VSUB(T7o, T7p); T82 = VADD(T7o, T7p); } { V T38, T39, T68, T6b; T38 = VFMA(LDK(KP382683432), T2q, VMUL(LDK(KP923879532), T2v)); T39 = VFNMS(LDK(KP382683432), T2B, VMUL(LDK(KP923879532), T2G)); T3a = VSUB(T38, T39); T4l = VADD(T38, T39); T68 = VSUB(T66, T67); T6b = VSUB(T69, T6a); T6c = VMUL(LDK(KP707106781), VSUB(T68, T6b)); T6k = VMUL(LDK(KP707106781), VADD(T68, T6b)); } } { V T7s, T7R, T7M, T7U, T7D, T7T, T7J, T7Q; { V T7k, T7r, T7K, T7L; T7k = VFNMS(LDK(KP382683432), T7j, VMUL(LDK(KP923879532), T7g)); T7r = VFMA(LDK(KP923879532), T7n, VMUL(LDK(KP382683432), T7q)); T7s = VSUB(T7k, T7r); T7R = VADD(T7k, T7r); T7K = VFMA(LDK(KP382683432), T7g, VMUL(LDK(KP923879532), T7j)); T7L = VFNMS(LDK(KP382683432), T7n, VMUL(LDK(KP923879532), T7q)); T7M = VSUB(T7K, T7L); T7U = VADD(T7K, T7L); } { V T7z, T7C, T7H, T7I; T7z = VMUL(LDK(KP707106781), VSUB(T7v, T7y)); T7C = VSUB(T7A, T7B); T7D = VSUB(T7z, T7C); T7T = VADD(T7C, T7z); T7H = VSUB(T7F, T7G); T7I = VMUL(LDK(KP707106781), VADD(T7v, T7y)); T7J = VSUB(T7H, T7I); T7Q = VADD(T7H, T7I); } { V T7E, T7N, T7W, T7X; T7E = VBYI(VSUB(T7s, T7D)); T7N = VSUB(T7J, T7M); ST(&(x[WS(rs, 20)]), VADD(T7E, T7N), ms, &(x[0])); ST(&(x[WS(rs, 44)]), VSUB(T7N, T7E), ms, &(x[0])); T7W = VSUB(T7Q, T7R); T7X = VBYI(VSUB(T7U, T7T)); ST(&(x[WS(rs, 36)]), VSUB(T7W, T7X), ms, &(x[0])); ST(&(x[WS(rs, 28)]), VADD(T7W, T7X), ms, &(x[0])); } { V T7O, T7P, T7S, T7V; T7O = VBYI(VADD(T7D, T7s)); T7P = VADD(T7J, T7M); ST(&(x[WS(rs, 12)]), VADD(T7O, T7P), ms, &(x[0])); ST(&(x[WS(rs, 52)]), VSUB(T7P, T7O), ms, &(x[0])); T7S = VADD(T7Q, T7R); T7V = VBYI(VADD(T7T, T7U)); ST(&(x[WS(rs, 60)]), VSUB(T7S, T7V), ms, &(x[0])); ST(&(x[WS(rs, 4)]), VADD(T7S, T7V), ms, &(x[0])); } } { V T84, T8c, T8l, T8n, T87, T8h, T8b, T8g, T8i, T8m; { V T80, T83, T8j, T8k; T80 = VSUB(T7Y, T7Z); T83 = VSUB(T81, T82); T84 = VMUL(LDK(KP707106781), VSUB(T80, T83)); T8c = VMUL(LDK(KP707106781), VADD(T80, T83)); T8j = VADD(T7Y, T7Z); T8k = VADD(T81, T82); T8l = VBYI(VSUB(T8j, T8k)); T8n = VADD(T8j, T8k); } { V T85, T86, T89, T8a; T85 = VADD(T7t, T7u); T86 = VADD(T7w, T7x); T87 = VSUB(T85, T86); T8h = VADD(T85, T86); T89 = VADD(T7F, T7G); T8a = VADD(T7A, T7B); T8b = VSUB(T89, T8a); T8g = VADD(T89, T8a); } T8i = VSUB(T8g, T8h); ST(&(x[WS(rs, 48)]), VSUB(T8i, T8l), ms, &(x[0])); ST(&(x[WS(rs, 16)]), VADD(T8i, T8l), ms, &(x[0])); T8m = VADD(T8g, T8h); ST(&(x[WS(rs, 32)]), VSUB(T8m, T8n), ms, &(x[0])); ST(&(x[0]), VADD(T8m, T8n), ms, &(x[0])); { V T88, T8d, T8e, T8f; T88 = VBYI(VSUB(T84, T87)); T8d = VSUB(T8b, T8c); ST(&(x[WS(rs, 24)]), VADD(T88, T8d), ms, &(x[0])); ST(&(x[WS(rs, 40)]), VSUB(T8d, T88), ms, &(x[0])); T8e = VBYI(VADD(T87, T84)); T8f = VADD(T8b, T8c); ST(&(x[WS(rs, 8)]), VADD(T8e, T8f), ms, &(x[0])); ST(&(x[WS(rs, 56)]), VSUB(T8f, T8e), ms, &(x[0])); } } { V T5O, T6H, T6x, T6F, T6n, T6I, T6A, T6E; { V T5y, T5N, T6r, T6w; T5y = VSUB(T5q, T5x); T5N = VSUB(T5F, T5M); T5O = VSUB(T5y, T5N); T6H = VADD(T5y, T5N); T6r = VSUB(T6p, T6q); T6w = VSUB(T6s, T6v); T6x = VSUB(T6r, T6w); T6F = VADD(T6w, T6r); { V T65, T6y, T6m, T6z; { V T5Z, T64, T6g, T6l; T5Z = VSUB(T5V, T5Y); T64 = VSUB(T62, T63); T65 = VFMA(LDK(KP831469612), T5Z, VMUL(LDK(KP555570233), T64)); T6y = VFNMS(LDK(KP555570233), T5Z, VMUL(LDK(KP831469612), T64)); T6g = VSUB(T6c, T6f); T6l = VSUB(T6j, T6k); T6m = VFNMS(LDK(KP555570233), T6l, VMUL(LDK(KP831469612), T6g)); T6z = VFMA(LDK(KP555570233), T6g, VMUL(LDK(KP831469612), T6l)); } T6n = VSUB(T65, T6m); T6I = VADD(T6y, T6z); T6A = VSUB(T6y, T6z); T6E = VADD(T65, T6m); } } { V T6o, T6B, T6K, T6L; T6o = VADD(T5O, T6n); T6B = VBYI(VADD(T6x, T6A)); ST(&(x[WS(rs, 54)]), VSUB(T6o, T6B), ms, &(x[0])); ST(&(x[WS(rs, 10)]), VADD(T6o, T6B), ms, &(x[0])); T6K = VBYI(VADD(T6F, T6E)); T6L = VADD(T6H, T6I); ST(&(x[WS(rs, 6)]), VADD(T6K, T6L), ms, &(x[0])); ST(&(x[WS(rs, 58)]), VSUB(T6L, T6K), ms, &(x[0])); } { V T6C, T6D, T6G, T6J; T6C = VSUB(T5O, T6n); T6D = VBYI(VSUB(T6A, T6x)); ST(&(x[WS(rs, 42)]), VSUB(T6C, T6D), ms, &(x[0])); ST(&(x[WS(rs, 22)]), VADD(T6C, T6D), ms, &(x[0])); T6G = VBYI(VSUB(T6E, T6F)); T6J = VSUB(T6H, T6I); ST(&(x[WS(rs, 26)]), VADD(T6G, T6J), ms, &(x[0])); ST(&(x[WS(rs, 38)]), VSUB(T6J, T6G), ms, &(x[0])); } } { V T6O, T79, T6Z, T77, T6V, T7a, T72, T76; { V T6M, T6N, T6X, T6Y; T6M = VADD(T5q, T5x); T6N = VADD(T6p, T6q); T6O = VSUB(T6M, T6N); T79 = VADD(T6M, T6N); T6X = VADD(T5F, T5M); T6Y = VADD(T6v, T6s); T6Z = VSUB(T6X, T6Y); T77 = VADD(T6Y, T6X); { V T6R, T70, T6U, T71; { V T6P, T6Q, T6S, T6T; T6P = VADD(T5Y, T5V); T6Q = VADD(T62, T63); T6R = VFMA(LDK(KP980785280), T6P, VMUL(LDK(KP195090322), T6Q)); T70 = VFNMS(LDK(KP195090322), T6P, VMUL(LDK(KP980785280), T6Q)); T6S = VADD(T6f, T6c); T6T = VADD(T6j, T6k); T6U = VFNMS(LDK(KP195090322), T6T, VMUL(LDK(KP980785280), T6S)); T71 = VFMA(LDK(KP195090322), T6S, VMUL(LDK(KP980785280), T6T)); } T6V = VSUB(T6R, T6U); T7a = VADD(T70, T71); T72 = VSUB(T70, T71); T76 = VADD(T6R, T6U); } } { V T6W, T73, T7c, T7d; T6W = VADD(T6O, T6V); T73 = VBYI(VADD(T6Z, T72)); ST(&(x[WS(rs, 50)]), VSUB(T6W, T73), ms, &(x[0])); ST(&(x[WS(rs, 14)]), VADD(T6W, T73), ms, &(x[0])); T7c = VBYI(VADD(T77, T76)); T7d = VADD(T79, T7a); ST(&(x[WS(rs, 2)]), VADD(T7c, T7d), ms, &(x[0])); ST(&(x[WS(rs, 62)]), VSUB(T7d, T7c), ms, &(x[0])); } { V T74, T75, T78, T7b; T74 = VSUB(T6O, T6V); T75 = VBYI(VSUB(T72, T6Z)); ST(&(x[WS(rs, 46)]), VSUB(T74, T75), ms, &(x[0])); ST(&(x[WS(rs, 18)]), VADD(T74, T75), ms, &(x[0])); T78 = VBYI(VSUB(T76, T77)); T7b = VSUB(T79, T7a); ST(&(x[WS(rs, 30)]), VADD(T78, T7b), ms, &(x[0])); ST(&(x[WS(rs, 34)]), VSUB(T7b, T78), ms, &(x[0])); } } { V T4z, T5g, T4R, T59, T4H, T5j, T4O, T55, T4o, T4S, T4K, T4P, T52, T5k, T5c; V T5h; { V T4y, T57, T4v, T58, T4r, T4u; T4y = VADD(T4w, T4x); T57 = VSUB(T4B, T4C); T4r = VFMA(LDK(KP980785280), T4p, VMUL(LDK(KP195090322), T4q)); T4u = VFNMS(LDK(KP195090322), T4t, VMUL(LDK(KP980785280), T4s)); T4v = VADD(T4r, T4u); T58 = VSUB(T4r, T4u); T4z = VSUB(T4v, T4y); T5g = VADD(T57, T58); T4R = VADD(T4y, T4v); T59 = VSUB(T57, T58); } { V T4D, T54, T4G, T53, T4E, T4F; T4D = VADD(T4B, T4C); T54 = VSUB(T4x, T4w); T4E = VFNMS(LDK(KP195090322), T4p, VMUL(LDK(KP980785280), T4q)); T4F = VFMA(LDK(KP195090322), T4s, VMUL(LDK(KP980785280), T4t)); T4G = VADD(T4E, T4F); T53 = VSUB(T4E, T4F); T4H = VSUB(T4D, T4G); T5j = VADD(T54, T53); T4O = VADD(T4D, T4G); T55 = VSUB(T53, T54); } { V T4g, T4I, T4n, T4J; { V T4c, T4f, T4j, T4m; T4c = VADD(T4a, T4b); T4f = VADD(T4d, T4e); T4g = VFNMS(LDK(KP098017140), T4f, VMUL(LDK(KP995184726), T4c)); T4I = VFMA(LDK(KP098017140), T4c, VMUL(LDK(KP995184726), T4f)); T4j = VADD(T4h, T4i); T4m = VADD(T4k, T4l); T4n = VFMA(LDK(KP995184726), T4j, VMUL(LDK(KP098017140), T4m)); T4J = VFNMS(LDK(KP098017140), T4j, VMUL(LDK(KP995184726), T4m)); } T4o = VSUB(T4g, T4n); T4S = VADD(T4I, T4J); T4K = VSUB(T4I, T4J); T4P = VADD(T4g, T4n); } { V T4Y, T5a, T51, T5b; { V T4W, T4X, T4Z, T50; T4W = VSUB(T4a, T4b); T4X = VSUB(T4e, T4d); T4Y = VFNMS(LDK(KP634393284), T4X, VMUL(LDK(KP773010453), T4W)); T5a = VFMA(LDK(KP634393284), T4W, VMUL(LDK(KP773010453), T4X)); T4Z = VSUB(T4h, T4i); T50 = VSUB(T4l, T4k); T51 = VFMA(LDK(KP773010453), T4Z, VMUL(LDK(KP634393284), T50)); T5b = VFNMS(LDK(KP634393284), T4Z, VMUL(LDK(KP773010453), T50)); } T52 = VSUB(T4Y, T51); T5k = VADD(T5a, T5b); T5c = VSUB(T5a, T5b); T5h = VADD(T4Y, T51); } { V T4A, T4L, T5i, T5l; T4A = VBYI(VSUB(T4o, T4z)); T4L = VSUB(T4H, T4K); ST(&(x[WS(rs, 17)]), VADD(T4A, T4L), ms, &(x[WS(rs, 1)])); ST(&(x[WS(rs, 47)]), VSUB(T4L, T4A), ms, &(x[WS(rs, 1)])); T5i = VADD(T5g, T5h); T5l = VBYI(VADD(T5j, T5k)); ST(&(x[WS(rs, 57)]), VSUB(T5i, T5l), ms, &(x[WS(rs, 1)])); ST(&(x[WS(rs, 7)]), VADD(T5i, T5l), ms, &(x[WS(rs, 1)])); } { V T5m, T5n, T4M, T4N; T5m = VSUB(T5g, T5h); T5n = VBYI(VSUB(T5k, T5j)); ST(&(x[WS(rs, 39)]), VSUB(T5m, T5n), ms, &(x[WS(rs, 1)])); ST(&(x[WS(rs, 25)]), VADD(T5m, T5n), ms, &(x[WS(rs, 1)])); T4M = VBYI(VADD(T4z, T4o)); T4N = VADD(T4H, T4K); ST(&(x[WS(rs, 15)]), VADD(T4M, T4N), ms, &(x[WS(rs, 1)])); ST(&(x[WS(rs, 49)]), VSUB(T4N, T4M), ms, &(x[WS(rs, 1)])); } { V T4Q, T4T, T56, T5d; T4Q = VADD(T4O, T4P); T4T = VBYI(VADD(T4R, T4S)); ST(&(x[WS(rs, 63)]), VSUB(T4Q, T4T), ms, &(x[WS(rs, 1)])); ST(&(x[WS(rs, 1)]), VADD(T4Q, T4T), ms, &(x[WS(rs, 1)])); T56 = VBYI(VSUB(T52, T55)); T5d = VSUB(T59, T5c); ST(&(x[WS(rs, 23)]), VADD(T56, T5d), ms, &(x[WS(rs, 1)])); ST(&(x[WS(rs, 41)]), VSUB(T5d, T56), ms, &(x[WS(rs, 1)])); } { V T5e, T5f, T4U, T4V; T5e = VBYI(VADD(T55, T52)); T5f = VADD(T59, T5c); ST(&(x[WS(rs, 9)]), VADD(T5e, T5f), ms, &(x[WS(rs, 1)])); ST(&(x[WS(rs, 55)]), VSUB(T5f, T5e), ms, &(x[WS(rs, 1)])); T4U = VSUB(T4O, T4P); T4V = VBYI(VSUB(T4S, T4R)); ST(&(x[WS(rs, 33)]), VSUB(T4U, T4V), ms, &(x[WS(rs, 1)])); ST(&(x[WS(rs, 31)]), VADD(T4U, T4V), ms, &(x[WS(rs, 1)])); } } { V T1u, T43, T3D, T3V, T3t, T45, T3B, T3K, T3d, T3E, T3w, T3A, T3R, T46, T3Y; V T42; { V TE, T3U, T1t, T3T, T13, T1s; TE = VSUB(Tg, TD); T3U = VADD(T3r, T3k); T13 = VFMA(LDK(KP831469612), TV, VMUL(LDK(KP555570233), T12)); T1s = VFNMS(LDK(KP555570233), T1r, VMUL(LDK(KP831469612), T1k)); T1t = VSUB(T13, T1s); T3T = VADD(T13, T1s); T1u = VSUB(TE, T1t); T43 = VADD(T3U, T3T); T3D = VADD(TE, T1t); T3V = VSUB(T3T, T3U); } { V T3s, T3I, T3h, T3J, T3f, T3g; T3s = VSUB(T3k, T3r); T3I = VADD(Tg, TD); T3f = VFNMS(LDK(KP555570233), TV, VMUL(LDK(KP831469612), T12)); T3g = VFMA(LDK(KP555570233), T1k, VMUL(LDK(KP831469612), T1r)); T3h = VSUB(T3f, T3g); T3J = VADD(T3f, T3g); T3t = VSUB(T3h, T3s); T45 = VADD(T3I, T3J); T3B = VADD(T3s, T3h); T3K = VSUB(T3I, T3J); } { V T2l, T3u, T3c, T3v; { V T29, T2k, T30, T3b; T29 = VSUB(T1R, T28); T2k = VSUB(T2g, T2j); T2l = VFMA(LDK(KP881921264), T29, VMUL(LDK(KP471396736), T2k)); T3u = VFNMS(LDK(KP471396736), T29, VMUL(LDK(KP881921264), T2k)); T30 = VSUB(T2I, T2Z); T3b = VSUB(T37, T3a); T3c = VFNMS(LDK(KP471396736), T3b, VMUL(LDK(KP881921264), T30)); T3v = VFMA(LDK(KP471396736), T30, VMUL(LDK(KP881921264), T3b)); } T3d = VSUB(T2l, T3c); T3E = VADD(T3u, T3v); T3w = VSUB(T3u, T3v); T3A = VADD(T2l, T3c); } { V T3N, T3W, T3Q, T3X; { V T3L, T3M, T3O, T3P; T3L = VADD(T28, T1R); T3M = VADD(T2g, T2j); T3N = VFMA(LDK(KP956940335), T3L, VMUL(LDK(KP290284677), T3M)); T3W = VFNMS(LDK(KP290284677), T3L, VMUL(LDK(KP956940335), T3M)); T3O = VADD(T2Z, T2I); T3P = VADD(T37, T3a); T3Q = VFNMS(LDK(KP290284677), T3P, VMUL(LDK(KP956940335), T3O)); T3X = VFMA(LDK(KP290284677), T3O, VMUL(LDK(KP956940335), T3P)); } T3R = VSUB(T3N, T3Q); T46 = VADD(T3W, T3X); T3Y = VSUB(T3W, T3X); T42 = VADD(T3N, T3Q); } { V T3e, T3x, T44, T47; T3e = VADD(T1u, T3d); T3x = VBYI(VADD(T3t, T3w)); ST(&(x[WS(rs, 53)]), VSUB(T3e, T3x), ms, &(x[WS(rs, 1)])); ST(&(x[WS(rs, 11)]), VADD(T3e, T3x), ms, &(x[WS(rs, 1)])); T44 = VBYI(VSUB(T42, T43)); T47 = VSUB(T45, T46); ST(&(x[WS(rs, 29)]), VADD(T44, T47), ms, &(x[WS(rs, 1)])); ST(&(x[WS(rs, 35)]), VSUB(T47, T44), ms, &(x[WS(rs, 1)])); } { V T48, T49, T3y, T3z; T48 = VBYI(VADD(T43, T42)); T49 = VADD(T45, T46); ST(&(x[WS(rs, 3)]), VADD(T48, T49), ms, &(x[WS(rs, 1)])); ST(&(x[WS(rs, 61)]), VSUB(T49, T48), ms, &(x[WS(rs, 1)])); T3y = VSUB(T1u, T3d); T3z = VBYI(VSUB(T3w, T3t)); ST(&(x[WS(rs, 43)]), VSUB(T3y, T3z), ms, &(x[WS(rs, 1)])); ST(&(x[WS(rs, 21)]), VADD(T3y, T3z), ms, &(x[WS(rs, 1)])); } { V T3C, T3F, T3S, T3Z; T3C = VBYI(VSUB(T3A, T3B)); T3F = VSUB(T3D, T3E); ST(&(x[WS(rs, 27)]), VADD(T3C, T3F), ms, &(x[WS(rs, 1)])); ST(&(x[WS(rs, 37)]), VSUB(T3F, T3C), ms, &(x[WS(rs, 1)])); T3S = VADD(T3K, T3R); T3Z = VBYI(VADD(T3V, T3Y)); ST(&(x[WS(rs, 51)]), VSUB(T3S, T3Z), ms, &(x[WS(rs, 1)])); ST(&(x[WS(rs, 13)]), VADD(T3S, T3Z), ms, &(x[WS(rs, 1)])); } { V T40, T41, T3G, T3H; T40 = VSUB(T3K, T3R); T41 = VBYI(VSUB(T3Y, T3V)); ST(&(x[WS(rs, 45)]), VSUB(T40, T41), ms, &(x[WS(rs, 1)])); ST(&(x[WS(rs, 19)]), VADD(T40, T41), ms, &(x[WS(rs, 1)])); T3G = VBYI(VADD(T3B, T3A)); T3H = VADD(T3D, T3E); ST(&(x[WS(rs, 5)]), VADD(T3G, T3H), ms, &(x[WS(rs, 1)])); ST(&(x[WS(rs, 59)]), VSUB(T3H, T3G), ms, &(x[WS(rs, 1)])); } } } } VLEAVE(); } static const tw_instr twinstr[] = { VTW(0, 1), VTW(0, 2), VTW(0, 3), VTW(0, 4), VTW(0, 5), VTW(0, 6), VTW(0, 7), VTW(0, 8), VTW(0, 9), VTW(0, 10), VTW(0, 11), VTW(0, 12), VTW(0, 13), VTW(0, 14), VTW(0, 15), VTW(0, 16), VTW(0, 17), VTW(0, 18), VTW(0, 19), VTW(0, 20), VTW(0, 21), VTW(0, 22), VTW(0, 23), VTW(0, 24), VTW(0, 25), VTW(0, 26), VTW(0, 27), VTW(0, 28), VTW(0, 29), VTW(0, 30), VTW(0, 31), VTW(0, 32), VTW(0, 33), VTW(0, 34), VTW(0, 35), VTW(0, 36), VTW(0, 37), VTW(0, 38), VTW(0, 39), VTW(0, 40), VTW(0, 41), VTW(0, 42), VTW(0, 43), VTW(0, 44), VTW(0, 45), VTW(0, 46), VTW(0, 47), VTW(0, 48), VTW(0, 49), VTW(0, 50), VTW(0, 51), VTW(0, 52), VTW(0, 53), VTW(0, 54), VTW(0, 55), VTW(0, 56), VTW(0, 57), VTW(0, 58), VTW(0, 59), VTW(0, 60), VTW(0, 61), VTW(0, 62), VTW(0, 63), {TW_NEXT, VL, 0} }; static const ct_desc desc = { 64, XSIMD_STRING("t1bv_64"), twinstr, &GENUS, {467, 198, 52, 0}, 0, 0, 0 }; void XSIMD(codelet_t1bv_64) (planner *p) { X(kdft_dit_register) (p, t1bv_64, &desc); } #endif /* HAVE_FMA */ fftw-3.3.3/dft/simd/common/t1bv_20.c0000644000175000017500000004175512054410552013743 00000000000000/* * Copyright (c) 2003, 2007-11 Matteo Frigo * Copyright (c) 2003, 2007-11 Massachusetts Institute of Technology * * This program is free software; you can redistribute it and/or modify * it under the terms of the GNU General Public License as published by * the Free Software Foundation; either version 2 of the License, or * (at your option) any later version. * * This program is distributed in the hope that it will be useful, * but WITHOUT ANY WARRANTY; without even the implied warranty of * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the * GNU General Public License for more details. * * You should have received a copy of the GNU General Public License * along with this program; if not, write to the Free Software * Foundation, Inc., 51 Franklin Street, Fifth Floor, Boston, MA 02110-1301 USA * */ /* This file was automatically generated --- DO NOT EDIT */ /* Generated on Sun Nov 25 07:39:05 EST 2012 */ #include "codelet-dft.h" #ifdef HAVE_FMA /* Generated by: ../../../genfft/gen_twiddle_c.native -fma -reorder-insns -schedule-for-pipeline -simd -compact -variables 4 -pipeline-latency 8 -n 20 -name t1bv_20 -include t1b.h -sign 1 */ /* * This function contains 123 FP additions, 88 FP multiplications, * (or, 77 additions, 42 multiplications, 46 fused multiply/add), * 68 stack variables, 4 constants, and 40 memory accesses */ #include "t1b.h" static void t1bv_20(R *ri, R *ii, const R *W, stride rs, INT mb, INT me, INT ms) { DVK(KP951056516, +0.951056516295153572116439333379382143405698634); DVK(KP559016994, +0.559016994374947424102293417182819058860154590); DVK(KP250000000, +0.250000000000000000000000000000000000000000000); DVK(KP618033988, +0.618033988749894848204586834365638117720309180); { INT m; R *x; x = ii; for (m = mb, W = W + (mb * ((TWVL / VL) * 38)); m < me; m = m + VL, x = x + (VL * ms), W = W + (TWVL * 38), MAKE_VOLATILE_STRIDE(20, rs)) { V T4, TX, T1m, T1K, T1y, Tk, Tf, T14, TQ, TZ, T1O, T1w, T1L, T1p, T1M; V T1s, TF, TY, T1x, Tp; { V T1, TV, T2, TT; T1 = LD(&(x[0]), ms, &(x[0])); TV = LD(&(x[WS(rs, 15)]), ms, &(x[WS(rs, 1)])); T2 = LD(&(x[WS(rs, 10)]), ms, &(x[0])); TT = LD(&(x[WS(rs, 5)]), ms, &(x[WS(rs, 1)])); { V T9, T1n, TK, T1v, TP, Te, T1q, T1u, TB, TD, Tm, T1o, Tz, Tn, T1r; V TE, To; { V TM, TO, Ta, Tc; { V T5, T7, TG, TI, T1k, T1l; T5 = LD(&(x[WS(rs, 4)]), ms, &(x[0])); T7 = LD(&(x[WS(rs, 14)]), ms, &(x[0])); TG = LD(&(x[WS(rs, 13)]), ms, &(x[WS(rs, 1)])); TI = LD(&(x[WS(rs, 3)]), ms, &(x[WS(rs, 1)])); { V TW, T3, TU, T6, T8, TH, TJ, TL, TN; TL = LD(&(x[WS(rs, 17)]), ms, &(x[WS(rs, 1)])); TW = BYTW(&(W[TWVL * 28]), TV); T3 = BYTW(&(W[TWVL * 18]), T2); TU = BYTW(&(W[TWVL * 8]), TT); T6 = BYTW(&(W[TWVL * 6]), T5); T8 = BYTW(&(W[TWVL * 26]), T7); TH = BYTW(&(W[TWVL * 24]), TG); TJ = BYTW(&(W[TWVL * 4]), TI); TM = BYTW(&(W[TWVL * 32]), TL); TN = LD(&(x[WS(rs, 7)]), ms, &(x[WS(rs, 1)])); T4 = VSUB(T1, T3); T1k = VADD(T1, T3); TX = VSUB(TU, TW); T1l = VADD(TU, TW); T9 = VSUB(T6, T8); T1n = VADD(T6, T8); TK = VSUB(TH, TJ); T1v = VADD(TH, TJ); TO = BYTW(&(W[TWVL * 12]), TN); } Ta = LD(&(x[WS(rs, 16)]), ms, &(x[0])); T1m = VSUB(T1k, T1l); T1K = VADD(T1k, T1l); Tc = LD(&(x[WS(rs, 6)]), ms, &(x[0])); } { V Tb, Tx, Td, Th, Tj, Tw, Tg, Ti, Tv; Tg = LD(&(x[WS(rs, 8)]), ms, &(x[0])); Ti = LD(&(x[WS(rs, 18)]), ms, &(x[0])); Tv = LD(&(x[WS(rs, 9)]), ms, &(x[WS(rs, 1)])); TP = VSUB(TM, TO); T1y = VADD(TM, TO); Tb = BYTW(&(W[TWVL * 30]), Ta); Tx = LD(&(x[WS(rs, 19)]), ms, &(x[WS(rs, 1)])); Td = BYTW(&(W[TWVL * 10]), Tc); Th = BYTW(&(W[TWVL * 14]), Tg); Tj = BYTW(&(W[TWVL * 34]), Ti); Tw = BYTW(&(W[TWVL * 16]), Tv); { V TA, TC, Ty, Tl; TA = LD(&(x[WS(rs, 1)]), ms, &(x[WS(rs, 1)])); TC = LD(&(x[WS(rs, 11)]), ms, &(x[WS(rs, 1)])); Tl = LD(&(x[WS(rs, 12)]), ms, &(x[0])); Ty = BYTW(&(W[TWVL * 36]), Tx); Te = VSUB(Tb, Td); T1q = VADD(Tb, Td); Tk = VSUB(Th, Tj); T1u = VADD(Th, Tj); TB = BYTW(&(W[0]), TA); TD = BYTW(&(W[TWVL * 20]), TC); Tm = BYTW(&(W[TWVL * 22]), Tl); T1o = VADD(Tw, Ty); Tz = VSUB(Tw, Ty); Tn = LD(&(x[WS(rs, 2)]), ms, &(x[0])); } } } Tf = VADD(T9, Te); T14 = VSUB(T9, Te); TQ = VSUB(TK, TP); TZ = VADD(TK, TP); T1r = VADD(TB, TD); TE = VSUB(TB, TD); T1O = VADD(T1u, T1v); T1w = VSUB(T1u, T1v); To = BYTW(&(W[TWVL * 2]), Tn); T1L = VADD(T1n, T1o); T1p = VSUB(T1n, T1o); T1M = VADD(T1q, T1r); T1s = VSUB(T1q, T1r); TF = VSUB(Tz, TE); TY = VADD(Tz, TE); T1x = VADD(Tm, To); Tp = VSUB(Tm, To); } } { V T1V, T1N, T12, T1b, TR, T1G, T1t, T1z, T1P, Tq, T15, T11, T1j, T10; T1V = VSUB(T1L, T1M); T1N = VADD(T1L, T1M); T12 = VSUB(TY, TZ); T10 = VADD(TY, TZ); T1b = VFNMS(LDK(KP618033988), TF, TQ); TR = VFMA(LDK(KP618033988), TQ, TF); T1G = VSUB(T1p, T1s); T1t = VADD(T1p, T1s); T1z = VSUB(T1x, T1y); T1P = VADD(T1x, T1y); Tq = VADD(Tk, Tp); T15 = VSUB(Tk, Tp); T11 = VFNMS(LDK(KP250000000), T10, TX); T1j = VADD(TX, T10); { V T1J, T1H, T1D, T1Z, T1X, T1T, T1f, T1h, T19, T17, T1C, T1S, T1a, Tu, T1F; V T1A; T1F = VSUB(T1w, T1z); T1A = VADD(T1w, T1z); { V T1W, T1Q, Tt, Tr; T1W = VSUB(T1O, T1P); T1Q = VADD(T1O, T1P); Tt = VSUB(Tf, Tq); Tr = VADD(Tf, Tq); { V T1e, T16, T1d, T13; T1e = VFNMS(LDK(KP618033988), T14, T15); T16 = VFMA(LDK(KP618033988), T15, T14); T1d = VFNMS(LDK(KP559016994), T12, T11); T13 = VFMA(LDK(KP559016994), T12, T11); T1J = VMUL(LDK(KP951056516), VFMA(LDK(KP618033988), T1F, T1G)); T1H = VMUL(LDK(KP951056516), VFNMS(LDK(KP618033988), T1G, T1F)); { V T1B, T1R, Ts, T1i; T1B = VADD(T1t, T1A); T1D = VSUB(T1t, T1A); T1Z = VMUL(LDK(KP951056516), VFNMS(LDK(KP618033988), T1V, T1W)); T1X = VMUL(LDK(KP951056516), VFMA(LDK(KP618033988), T1W, T1V)); T1R = VADD(T1N, T1Q); T1T = VSUB(T1N, T1Q); Ts = VFNMS(LDK(KP250000000), Tr, T4); T1i = VADD(T4, Tr); T1f = VFNMS(LDK(KP951056516), T1e, T1d); T1h = VFMA(LDK(KP951056516), T1e, T1d); T19 = VFNMS(LDK(KP951056516), T16, T13); T17 = VFMA(LDK(KP951056516), T16, T13); ST(&(x[WS(rs, 10)]), VADD(T1m, T1B), ms, &(x[0])); T1C = VFNMS(LDK(KP250000000), T1B, T1m); ST(&(x[0]), VADD(T1K, T1R), ms, &(x[0])); T1S = VFNMS(LDK(KP250000000), T1R, T1K); T1a = VFNMS(LDK(KP559016994), Tt, Ts); Tu = VFMA(LDK(KP559016994), Tt, Ts); ST(&(x[WS(rs, 5)]), VFMAI(T1j, T1i), ms, &(x[WS(rs, 1)])); ST(&(x[WS(rs, 15)]), VFNMSI(T1j, T1i), ms, &(x[WS(rs, 1)])); } } } { V T1E, T1I, T1U, T1Y; T1E = VFNMS(LDK(KP559016994), T1D, T1C); T1I = VFMA(LDK(KP559016994), T1D, T1C); T1U = VFMA(LDK(KP559016994), T1T, T1S); T1Y = VFNMS(LDK(KP559016994), T1T, T1S); { V T1c, T1g, T18, TS; T1c = VFMA(LDK(KP951056516), T1b, T1a); T1g = VFNMS(LDK(KP951056516), T1b, T1a); T18 = VFMA(LDK(KP951056516), TR, Tu); TS = VFNMS(LDK(KP951056516), TR, Tu); ST(&(x[WS(rs, 18)]), VFMAI(T1H, T1E), ms, &(x[0])); ST(&(x[WS(rs, 2)]), VFNMSI(T1H, T1E), ms, &(x[0])); ST(&(x[WS(rs, 14)]), VFNMSI(T1J, T1I), ms, &(x[0])); ST(&(x[WS(rs, 6)]), VFMAI(T1J, T1I), ms, &(x[0])); ST(&(x[WS(rs, 16)]), VFMAI(T1X, T1U), ms, &(x[0])); ST(&(x[WS(rs, 4)]), VFNMSI(T1X, T1U), ms, &(x[0])); ST(&(x[WS(rs, 12)]), VFNMSI(T1Z, T1Y), ms, &(x[0])); ST(&(x[WS(rs, 8)]), VFMAI(T1Z, T1Y), ms, &(x[0])); ST(&(x[WS(rs, 17)]), VFMAI(T1f, T1c), ms, &(x[WS(rs, 1)])); ST(&(x[WS(rs, 3)]), VFNMSI(T1f, T1c), ms, &(x[WS(rs, 1)])); ST(&(x[WS(rs, 13)]), VFMAI(T1h, T1g), ms, &(x[WS(rs, 1)])); ST(&(x[WS(rs, 7)]), VFNMSI(T1h, T1g), ms, &(x[WS(rs, 1)])); ST(&(x[WS(rs, 9)]), VFMAI(T19, T18), ms, &(x[WS(rs, 1)])); ST(&(x[WS(rs, 11)]), VFNMSI(T19, T18), ms, &(x[WS(rs, 1)])); ST(&(x[WS(rs, 1)]), VFMAI(T17, TS), ms, &(x[WS(rs, 1)])); ST(&(x[WS(rs, 19)]), VFNMSI(T17, TS), ms, &(x[WS(rs, 1)])); } } } } } } VLEAVE(); } static const tw_instr twinstr[] = { VTW(0, 1), VTW(0, 2), VTW(0, 3), VTW(0, 4), VTW(0, 5), VTW(0, 6), VTW(0, 7), VTW(0, 8), VTW(0, 9), VTW(0, 10), VTW(0, 11), VTW(0, 12), VTW(0, 13), VTW(0, 14), VTW(0, 15), VTW(0, 16), VTW(0, 17), VTW(0, 18), VTW(0, 19), {TW_NEXT, VL, 0} }; static const ct_desc desc = { 20, XSIMD_STRING("t1bv_20"), twinstr, &GENUS, {77, 42, 46, 0}, 0, 0, 0 }; void XSIMD(codelet_t1bv_20) (planner *p) { X(kdft_dit_register) (p, t1bv_20, &desc); } #else /* HAVE_FMA */ /* Generated by: ../../../genfft/gen_twiddle_c.native -simd -compact -variables 4 -pipeline-latency 8 -n 20 -name t1bv_20 -include t1b.h -sign 1 */ /* * This function contains 123 FP additions, 62 FP multiplications, * (or, 111 additions, 50 multiplications, 12 fused multiply/add), * 54 stack variables, 4 constants, and 40 memory accesses */ #include "t1b.h" static void t1bv_20(R *ri, R *ii, const R *W, stride rs, INT mb, INT me, INT ms) { DVK(KP587785252, +0.587785252292473129168705954639072768597652438); DVK(KP951056516, +0.951056516295153572116439333379382143405698634); DVK(KP250000000, +0.250000000000000000000000000000000000000000000); DVK(KP559016994, +0.559016994374947424102293417182819058860154590); { INT m; R *x; x = ii; for (m = mb, W = W + (mb * ((TWVL / VL) * 38)); m < me; m = m + VL, x = x + (VL * ms), W = W + (TWVL * 38), MAKE_VOLATILE_STRIDE(20, rs)) { V T4, T10, T1B, T1R, TF, T14, T15, TQ, Tf, Tq, Tr, T1N, T1O, T1P, T1t; V T1w, T1D, TT, TU, T11, T1K, T1L, T1M, T1m, T1p, T1C, T1i, T1j; { V T1, TZ, T3, TX, TY, T2, TW, T1z, T1A; T1 = LD(&(x[0]), ms, &(x[0])); TY = LD(&(x[WS(rs, 15)]), ms, &(x[WS(rs, 1)])); TZ = BYTW(&(W[TWVL * 28]), TY); T2 = LD(&(x[WS(rs, 10)]), ms, &(x[0])); T3 = BYTW(&(W[TWVL * 18]), T2); TW = LD(&(x[WS(rs, 5)]), ms, &(x[WS(rs, 1)])); TX = BYTW(&(W[TWVL * 8]), TW); T4 = VSUB(T1, T3); T10 = VSUB(TX, TZ); T1z = VADD(T1, T3); T1A = VADD(TX, TZ); T1B = VSUB(T1z, T1A); T1R = VADD(T1z, T1A); } { V T9, T1k, TK, T1s, TP, T1v, Te, T1n, Tk, T1r, Tz, T1l, TE, T1o, Tp; V T1u; { V T6, T8, T5, T7; T5 = LD(&(x[WS(rs, 4)]), ms, &(x[0])); T6 = BYTW(&(W[TWVL * 6]), T5); T7 = LD(&(x[WS(rs, 14)]), ms, &(x[0])); T8 = BYTW(&(W[TWVL * 26]), T7); T9 = VSUB(T6, T8); T1k = VADD(T6, T8); } { V TH, TJ, TG, TI; TG = LD(&(x[WS(rs, 13)]), ms, &(x[WS(rs, 1)])); TH = BYTW(&(W[TWVL * 24]), TG); TI = LD(&(x[WS(rs, 3)]), ms, &(x[WS(rs, 1)])); TJ = BYTW(&(W[TWVL * 4]), TI); TK = VSUB(TH, TJ); T1s = VADD(TH, TJ); } { V TM, TO, TL, TN; TL = LD(&(x[WS(rs, 17)]), ms, &(x[WS(rs, 1)])); TM = BYTW(&(W[TWVL * 32]), TL); TN = LD(&(x[WS(rs, 7)]), ms, &(x[WS(rs, 1)])); TO = BYTW(&(W[TWVL * 12]), TN); TP = VSUB(TM, TO); T1v = VADD(TM, TO); } { V Tb, Td, Ta, Tc; Ta = LD(&(x[WS(rs, 16)]), ms, &(x[0])); Tb = BYTW(&(W[TWVL * 30]), Ta); Tc = LD(&(x[WS(rs, 6)]), ms, &(x[0])); Td = BYTW(&(W[TWVL * 10]), Tc); Te = VSUB(Tb, Td); T1n = VADD(Tb, Td); } { V Th, Tj, Tg, Ti; Tg = LD(&(x[WS(rs, 8)]), ms, &(x[0])); Th = BYTW(&(W[TWVL * 14]), Tg); Ti = LD(&(x[WS(rs, 18)]), ms, &(x[0])); Tj = BYTW(&(W[TWVL * 34]), Ti); Tk = VSUB(Th, Tj); T1r = VADD(Th, Tj); } { V Tw, Ty, Tv, Tx; Tv = LD(&(x[WS(rs, 9)]), ms, &(x[WS(rs, 1)])); Tw = BYTW(&(W[TWVL * 16]), Tv); Tx = LD(&(x[WS(rs, 19)]), ms, &(x[WS(rs, 1)])); Ty = BYTW(&(W[TWVL * 36]), Tx); Tz = VSUB(Tw, Ty); T1l = VADD(Tw, Ty); } { V TB, TD, TA, TC; TA = LD(&(x[WS(rs, 1)]), ms, &(x[WS(rs, 1)])); TB = BYTW(&(W[0]), TA); TC = LD(&(x[WS(rs, 11)]), ms, &(x[WS(rs, 1)])); TD = BYTW(&(W[TWVL * 20]), TC); TE = VSUB(TB, TD); T1o = VADD(TB, TD); } { V Tm, To, Tl, Tn; Tl = LD(&(x[WS(rs, 12)]), ms, &(x[0])); Tm = BYTW(&(W[TWVL * 22]), Tl); Tn = LD(&(x[WS(rs, 2)]), ms, &(x[0])); To = BYTW(&(W[TWVL * 2]), Tn); Tp = VSUB(Tm, To); T1u = VADD(Tm, To); } TF = VSUB(Tz, TE); T14 = VSUB(T9, Te); T15 = VSUB(Tk, Tp); TQ = VSUB(TK, TP); Tf = VADD(T9, Te); Tq = VADD(Tk, Tp); Tr = VADD(Tf, Tq); T1N = VADD(T1r, T1s); T1O = VADD(T1u, T1v); T1P = VADD(T1N, T1O); T1t = VSUB(T1r, T1s); T1w = VSUB(T1u, T1v); T1D = VADD(T1t, T1w); TT = VADD(Tz, TE); TU = VADD(TK, TP); T11 = VADD(TT, TU); T1K = VADD(T1k, T1l); T1L = VADD(T1n, T1o); T1M = VADD(T1K, T1L); T1m = VSUB(T1k, T1l); T1p = VSUB(T1n, T1o); T1C = VADD(T1m, T1p); } T1i = VADD(T4, Tr); T1j = VBYI(VADD(T10, T11)); ST(&(x[WS(rs, 15)]), VSUB(T1i, T1j), ms, &(x[WS(rs, 1)])); ST(&(x[WS(rs, 5)]), VADD(T1i, T1j), ms, &(x[WS(rs, 1)])); { V T1Q, T1S, T1T, T1X, T1Z, T1V, T1W, T1Y, T1U; T1Q = VMUL(LDK(KP559016994), VSUB(T1M, T1P)); T1S = VADD(T1M, T1P); T1T = VFNMS(LDK(KP250000000), T1S, T1R); T1V = VSUB(T1K, T1L); T1W = VSUB(T1N, T1O); T1X = VBYI(VFMA(LDK(KP951056516), T1V, VMUL(LDK(KP587785252), T1W))); T1Z = VBYI(VFNMS(LDK(KP951056516), T1W, VMUL(LDK(KP587785252), T1V))); ST(&(x[0]), VADD(T1R, T1S), ms, &(x[0])); T1Y = VSUB(T1T, T1Q); ST(&(x[WS(rs, 8)]), VSUB(T1Y, T1Z), ms, &(x[0])); ST(&(x[WS(rs, 12)]), VADD(T1Z, T1Y), ms, &(x[0])); T1U = VADD(T1Q, T1T); ST(&(x[WS(rs, 4)]), VSUB(T1U, T1X), ms, &(x[0])); ST(&(x[WS(rs, 16)]), VADD(T1X, T1U), ms, &(x[0])); } { V T1G, T1E, T1F, T1y, T1I, T1q, T1x, T1J, T1H; T1G = VMUL(LDK(KP559016994), VSUB(T1C, T1D)); T1E = VADD(T1C, T1D); T1F = VFNMS(LDK(KP250000000), T1E, T1B); T1q = VSUB(T1m, T1p); T1x = VSUB(T1t, T1w); T1y = VBYI(VFNMS(LDK(KP951056516), T1x, VMUL(LDK(KP587785252), T1q))); T1I = VBYI(VFMA(LDK(KP951056516), T1q, VMUL(LDK(KP587785252), T1x))); ST(&(x[WS(rs, 10)]), VADD(T1B, T1E), ms, &(x[0])); T1J = VADD(T1G, T1F); ST(&(x[WS(rs, 6)]), VADD(T1I, T1J), ms, &(x[0])); ST(&(x[WS(rs, 14)]), VSUB(T1J, T1I), ms, &(x[0])); T1H = VSUB(T1F, T1G); ST(&(x[WS(rs, 2)]), VADD(T1y, T1H), ms, &(x[0])); ST(&(x[WS(rs, 18)]), VSUB(T1H, T1y), ms, &(x[0])); } { V TR, T16, T1d, T1b, T13, T1e, Tu, T1a; TR = VFNMS(LDK(KP951056516), TQ, VMUL(LDK(KP587785252), TF)); T16 = VFNMS(LDK(KP951056516), T15, VMUL(LDK(KP587785252), T14)); T1d = VFMA(LDK(KP951056516), T14, VMUL(LDK(KP587785252), T15)); T1b = VFMA(LDK(KP951056516), TF, VMUL(LDK(KP587785252), TQ)); { V TV, T12, Ts, Tt; TV = VMUL(LDK(KP559016994), VSUB(TT, TU)); T12 = VFNMS(LDK(KP250000000), T11, T10); T13 = VSUB(TV, T12); T1e = VADD(TV, T12); Ts = VFNMS(LDK(KP250000000), Tr, T4); Tt = VMUL(LDK(KP559016994), VSUB(Tf, Tq)); Tu = VSUB(Ts, Tt); T1a = VADD(Tt, Ts); } { V TS, T17, T1g, T1h; TS = VSUB(Tu, TR); T17 = VBYI(VSUB(T13, T16)); ST(&(x[WS(rs, 17)]), VSUB(TS, T17), ms, &(x[WS(rs, 1)])); ST(&(x[WS(rs, 3)]), VADD(TS, T17), ms, &(x[WS(rs, 1)])); T1g = VADD(T1a, T1b); T1h = VBYI(VSUB(T1e, T1d)); ST(&(x[WS(rs, 11)]), VSUB(T1g, T1h), ms, &(x[WS(rs, 1)])); ST(&(x[WS(rs, 9)]), VADD(T1g, T1h), ms, &(x[WS(rs, 1)])); } { V T18, T19, T1c, T1f; T18 = VADD(Tu, TR); T19 = VBYI(VADD(T16, T13)); ST(&(x[WS(rs, 13)]), VSUB(T18, T19), ms, &(x[WS(rs, 1)])); ST(&(x[WS(rs, 7)]), VADD(T18, T19), ms, &(x[WS(rs, 1)])); T1c = VSUB(T1a, T1b); T1f = VBYI(VADD(T1d, T1e)); ST(&(x[WS(rs, 19)]), VSUB(T1c, T1f), ms, &(x[WS(rs, 1)])); ST(&(x[WS(rs, 1)]), VADD(T1c, T1f), ms, &(x[WS(rs, 1)])); } } } } VLEAVE(); } static const tw_instr twinstr[] = { VTW(0, 1), VTW(0, 2), VTW(0, 3), VTW(0, 4), VTW(0, 5), VTW(0, 6), VTW(0, 7), VTW(0, 8), VTW(0, 9), VTW(0, 10), VTW(0, 11), VTW(0, 12), VTW(0, 13), VTW(0, 14), VTW(0, 15), VTW(0, 16), VTW(0, 17), VTW(0, 18), VTW(0, 19), {TW_NEXT, VL, 0} }; static const ct_desc desc = { 20, XSIMD_STRING("t1bv_20"), twinstr, &GENUS, {111, 50, 12, 0}, 0, 0, 0 }; void XSIMD(codelet_t1bv_20) (planner *p) { X(kdft_dit_register) (p, t1bv_20, &desc); } #endif /* HAVE_FMA */ fftw-3.3.3/dft/simd/common/t1bv_25.c0000644000175000017500000011472712054410561013750 00000000000000/* * Copyright (c) 2003, 2007-11 Matteo Frigo * Copyright (c) 2003, 2007-11 Massachusetts Institute of Technology * * This program is free software; you can redistribute it and/or modify * it under the terms of the GNU General Public License as published by * the Free Software Foundation; either version 2 of the License, or * (at your option) any later version. * * This program is distributed in the hope that it will be useful, * but WITHOUT ANY WARRANTY; without even the implied warranty of * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the * GNU General Public License for more details. * * You should have received a copy of the GNU General Public License * along with this program; if not, write to the Free Software * Foundation, Inc., 51 Franklin Street, Fifth Floor, Boston, MA 02110-1301 USA * */ /* This file was automatically generated --- DO NOT EDIT */ /* Generated on Sun Nov 25 07:39:06 EST 2012 */ #include "codelet-dft.h" #ifdef HAVE_FMA /* Generated by: ../../../genfft/gen_twiddle_c.native -fma -reorder-insns -schedule-for-pipeline -simd -compact -variables 4 -pipeline-latency 8 -n 25 -name t1bv_25 -include t1b.h -sign 1 */ /* * This function contains 248 FP additions, 241 FP multiplications, * (or, 67 additions, 60 multiplications, 181 fused multiply/add), * 208 stack variables, 67 constants, and 50 memory accesses */ #include "t1b.h" static void t1bv_25(R *ri, R *ii, const R *W, stride rs, INT mb, INT me, INT ms) { DVK(KP792626838, +0.792626838241819413632131824093538848057784557); DVK(KP876091699, +0.876091699473550838204498029706869638173524346); DVK(KP617882369, +0.617882369114440893914546919006756321695042882); DVK(KP803003575, +0.803003575438660414833440593570376004635464850); DVK(KP242145790, +0.242145790282157779872542093866183953459003101); DVK(KP968583161, +0.968583161128631119490168375464735813836012403); DVK(KP999544308, +0.999544308746292983948881682379742149196758193); DVK(KP916574801, +0.916574801383451584742370439148878693530976769); DVK(KP904730450, +0.904730450839922351881287709692877908104763647); DVK(KP809385824, +0.809385824416008241660603814668679683846476688); DVK(KP447417479, +0.447417479732227551498980015410057305749330693); DVK(KP894834959, +0.894834959464455102997960030820114611498661386); DVK(KP867381224, +0.867381224396525206773171885031575671309956167); DVK(KP683113946, +0.683113946453479238701949862233725244439656928); DVK(KP559154169, +0.559154169276087864842202529084232643714075927); DVK(KP958953096, +0.958953096729998668045963838399037225970891871); DVK(KP831864738, +0.831864738706457140726048799369896829771167132); DVK(KP829049696, +0.829049696159252993975487806364305442437946767); DVK(KP860541664, +0.860541664367944677098261680920518816412804187); DVK(KP897376177, +0.897376177523557693138608077137219684419427330); DVK(KP876306680, +0.876306680043863587308115903922062583399064238); DVK(KP681693190, +0.681693190061530575150324149145440022633095390); DVK(KP560319534, +0.560319534973832390111614715371676131169633784); DVK(KP855719849, +0.855719849902058969314654733608091555096772472); DVK(KP237294955, +0.237294955877110315393888866460840817927895961); DVK(KP949179823, +0.949179823508441261575555465843363271711583843); DVK(KP904508497, +0.904508497187473712051146708591409529430077295); DVK(KP997675361, +0.997675361079556513670859573984492383596555031); DVK(KP763932022, +0.763932022500210303590826331268723764559381640); DVK(KP690983005, +0.690983005625052575897706582817180941139845410); DVK(KP992114701, +0.992114701314477831049793042785778521453036709); DVK(KP952936919, +0.952936919628306576880750665357914584765951388); DVK(KP998026728, +0.998026728428271561952336806863450553336905220); DVK(KP262346850, +0.262346850930607871785420028382979691334784273); DVK(KP570584518, +0.570584518783621657366766175430996792655723863); DVK(KP669429328, +0.669429328479476605641803240971985825917022098); DVK(KP923225144, +0.923225144846402650453449441572664695995209956); DVK(KP945422727, +0.945422727388575946270360266328811958657216298); DVK(KP522616830, +0.522616830205754336872861364785224694908468440); DVK(KP956723877, +0.956723877038460305821989399535483155872969262); DVK(KP906616052, +0.906616052148196230441134447086066874408359177); DVK(KP772036680, +0.772036680810363904029489473607579825330539880); DVK(KP845997307, +0.845997307939530944175097360758058292389769300); DVK(KP951056516, +0.951056516295153572116439333379382143405698634); DVK(KP921078979, +0.921078979742360627699756128143719920817673854); DVK(KP912575812, +0.912575812670962425556968549836277086778922727); DVK(KP982009705, +0.982009705009746369461829878184175962711969869); DVK(KP734762448, +0.734762448793050413546343770063151342619912334); DVK(KP494780565, +0.494780565770515410344588413655324772219443730); DVK(KP447533225, +0.447533225982656890041886979663652563063114397); DVK(KP269969613, +0.269969613759572083574752974412347470060951301); DVK(KP244189809, +0.244189809627953270309879511234821255780225091); DVK(KP667278218, +0.667278218140296670899089292254759909713898805); DVK(KP603558818, +0.603558818296015001454675132653458027918768137); DVK(KP522847744, +0.522847744331509716623755382187077770911012542); DVK(KP578046249, +0.578046249379945007321754579646815604023525655); DVK(KP987388751, +0.987388751065621252324603216482382109400433949); DVK(KP893101515, +0.893101515366181661711202267938416198338079437); DVK(KP120146378, +0.120146378570687701782758537356596213647956445); DVK(KP132830569, +0.132830569247582714407653942074819768844536507); DVK(KP869845200, +0.869845200362138853122720822420327157933056305); DVK(KP786782374, +0.786782374965295178365099601674911834788448471); DVK(KP066152395, +0.066152395967733048213034281011006031460903353); DVK(KP059835404, +0.059835404262124915169548397419498386427871950); DVK(KP559016994, +0.559016994374947424102293417182819058860154590); DVK(KP250000000, +0.250000000000000000000000000000000000000000000); DVK(KP618033988, +0.618033988749894848204586834365638117720309180); { INT m; R *x; x = ii; for (m = mb, W = W + (mb * ((TWVL / VL) * 48)); m < me; m = m + VL, x = x + (VL * ms), W = W + (TWVL * 48), MAKE_VOLATILE_STRIDE(25, rs)) { V T25, T1B, T2y, T1K, T2s, T23, T1S, T26, T20, T1X; { V T1O, T2X, Te, T3L, Td, T3Q, T3j, T3b, T2R, T2M, T2f, T27, T1y, T1H, T3M; V TW, TR, TK, T2B, T3n, T3e, T2U, T2F, T2i, T2a, Tz, T1C, T3N, TQ, T11; V T1b, T1c, T16; { V T1, T1g, T1i, T1p, T1k, T1m, Tb, T1N, T6, T1M; { V T7, T9, T2, T4, T1f, T1h, T1o; T1 = LD(&(x[0]), ms, &(x[0])); T7 = LD(&(x[WS(rs, 10)]), ms, &(x[0])); T9 = LD(&(x[WS(rs, 15)]), ms, &(x[WS(rs, 1)])); T2 = LD(&(x[WS(rs, 5)]), ms, &(x[WS(rs, 1)])); T4 = LD(&(x[WS(rs, 20)]), ms, &(x[0])); T1f = LD(&(x[WS(rs, 3)]), ms, &(x[WS(rs, 1)])); T1h = LD(&(x[WS(rs, 8)]), ms, &(x[0])); T1o = LD(&(x[WS(rs, 18)]), ms, &(x[0])); { V T8, Ta, T3, T5, T1j; T1j = LD(&(x[WS(rs, 23)]), ms, &(x[WS(rs, 1)])); T8 = BYTW(&(W[TWVL * 18]), T7); Ta = BYTW(&(W[TWVL * 28]), T9); T3 = BYTW(&(W[TWVL * 8]), T2); T5 = BYTW(&(W[TWVL * 38]), T4); T1g = BYTW(&(W[TWVL * 4]), T1f); T1i = BYTW(&(W[TWVL * 14]), T1h); T1p = BYTW(&(W[TWVL * 34]), T1o); T1k = BYTW(&(W[TWVL * 44]), T1j); T1m = LD(&(x[WS(rs, 13)]), ms, &(x[WS(rs, 1)])); Tb = VADD(T8, Ta); T1N = VSUB(T8, Ta); T6 = VADD(T3, T5); T1M = VSUB(T3, T5); } } { V T1v, T1l, Th, Tj, T1w, T1q, Tq, Tk, Tn, Tg; Tg = LD(&(x[WS(rs, 1)]), ms, &(x[WS(rs, 1)])); { V Tc, Ti, T1n, Tp; Ti = LD(&(x[WS(rs, 6)]), ms, &(x[0])); T1v = VSUB(T1i, T1k); T1l = VADD(T1i, T1k); T1n = BYTW(&(W[TWVL * 24]), T1m); Tp = LD(&(x[WS(rs, 16)]), ms, &(x[0])); T1O = VFMA(LDK(KP618033988), T1N, T1M); T2X = VFNMS(LDK(KP618033988), T1M, T1N); Te = VSUB(T6, Tb); Tc = VADD(T6, Tb); Th = BYTW(&(W[0]), Tg); Tj = BYTW(&(W[TWVL * 10]), Ti); T1w = VSUB(T1n, T1p); T1q = VADD(T1n, T1p); Tq = BYTW(&(W[TWVL * 30]), Tp); Tk = LD(&(x[WS(rs, 21)]), ms, &(x[WS(rs, 1)])); T3L = VADD(T1, Tc); Td = VFNMS(LDK(KP250000000), Tc, T1); Tn = LD(&(x[WS(rs, 11)]), ms, &(x[WS(rs, 1)])); } { V T1x, T2K, TM, TB, Tw, Tm, Tx, Tr, TI, T2L, T1u, TD, TF, TL; TL = LD(&(x[WS(rs, 4)]), ms, &(x[0])); { V T1t, Tl, To, TH, T1s, T1r, TA, TC; TA = LD(&(x[WS(rs, 24)]), ms, &(x[0])); T1r = VADD(T1l, T1q); T1t = VSUB(T1q, T1l); T1x = VFMA(LDK(KP618033988), T1w, T1v); T2K = VFNMS(LDK(KP618033988), T1v, T1w); Tl = BYTW(&(W[TWVL * 40]), Tk); To = BYTW(&(W[TWVL * 20]), Tn); TM = BYTW(&(W[TWVL * 6]), TL); TB = BYTW(&(W[TWVL * 46]), TA); TH = LD(&(x[WS(rs, 14)]), ms, &(x[0])); T1s = VFNMS(LDK(KP250000000), T1r, T1g); T3Q = VADD(T1g, T1r); TC = LD(&(x[WS(rs, 9)]), ms, &(x[WS(rs, 1)])); Tw = VSUB(Tj, Tl); Tm = VADD(Tj, Tl); Tx = VSUB(Tq, To); Tr = VADD(To, Tq); TI = BYTW(&(W[TWVL * 26]), TH); T2L = VFMA(LDK(KP559016994), T1t, T1s); T1u = VFNMS(LDK(KP559016994), T1t, T1s); TD = BYTW(&(W[TWVL * 16]), TC); TF = LD(&(x[WS(rs, 19)]), ms, &(x[WS(rs, 1)])); } { V Tu, Ty, T2E, TE, TN, TG, Tt, TV, Ts; TV = LD(&(x[WS(rs, 2)]), ms, &(x[0])); Ts = VADD(Tm, Tr); Tu = VSUB(Tm, Tr); Ty = VFNMS(LDK(KP618033988), Tx, Tw); T2E = VFMA(LDK(KP618033988), Tw, Tx); T3j = VFNMS(LDK(KP059835404), T2K, T2L); T3b = VFMA(LDK(KP066152395), T2L, T2K); T2R = VFNMS(LDK(KP786782374), T2K, T2L); T2M = VFMA(LDK(KP869845200), T2L, T2K); T2f = VFMA(LDK(KP132830569), T1u, T1x); T27 = VFNMS(LDK(KP120146378), T1x, T1u); T1y = VFNMS(LDK(KP893101515), T1x, T1u); T1H = VFMA(LDK(KP987388751), T1u, T1x); TE = VSUB(TB, TD); TN = VADD(TD, TB); TG = BYTW(&(W[TWVL * 36]), TF); Tt = VFNMS(LDK(KP250000000), Ts, Th); T3M = VADD(Th, Ts); TW = BYTW(&(W[TWVL * 2]), TV); { V TJ, TO, Tv, T2D, TY, T15, T10, T13, TP; { V TX, T14, TZ, T12; TX = LD(&(x[WS(rs, 7)]), ms, &(x[WS(rs, 1)])); T14 = LD(&(x[WS(rs, 17)]), ms, &(x[WS(rs, 1)])); TZ = LD(&(x[WS(rs, 22)]), ms, &(x[0])); T12 = LD(&(x[WS(rs, 12)]), ms, &(x[0])); TJ = VSUB(TG, TI); TO = VADD(TI, TG); Tv = VFMA(LDK(KP559016994), Tu, Tt); T2D = VFNMS(LDK(KP559016994), Tu, Tt); TY = BYTW(&(W[TWVL * 12]), TX); T15 = BYTW(&(W[TWVL * 32]), T14); T10 = BYTW(&(W[TWVL * 42]), TZ); T13 = BYTW(&(W[TWVL * 22]), T12); } TP = VADD(TN, TO); TR = VSUB(TN, TO); TK = VFMA(LDK(KP618033988), TJ, TE); T2B = VFNMS(LDK(KP618033988), TE, TJ); T3n = VFMA(LDK(KP578046249), T2D, T2E); T3e = VFNMS(LDK(KP522847744), T2E, T2D); T2U = VFNMS(LDK(KP987388751), T2D, T2E); T2F = VFMA(LDK(KP893101515), T2E, T2D); T2i = VFNMS(LDK(KP603558818), Ty, Tv); T2a = VFMA(LDK(KP667278218), Tv, Ty); Tz = VFNMS(LDK(KP244189809), Ty, Tv); T1C = VFMA(LDK(KP269969613), Tv, Ty); T3N = VADD(TM, TP); TQ = VFMS(LDK(KP250000000), TP, TM); T11 = VADD(TY, T10); T1b = VSUB(TY, T10); T1c = VSUB(T15, T13); T16 = VADD(T13, T15); } } } } } { V T2z, Tf, T3W, T3O, T1d, T2H, T3m, T2j, T2b, TT, T1D, T2G, T35, T2V, T2Z; V T3A, T3g, T2I, T1a, T3R, T3X; T2z = VFNMS(LDK(KP559016994), Te, Td); Tf = VFMA(LDK(KP559016994), Te, Td); { V TS, T2A, T17, T19; TS = VFNMS(LDK(KP559016994), TR, TQ); T2A = VFMA(LDK(KP559016994), TR, TQ); T3W = VSUB(T3M, T3N); T3O = VADD(T3M, T3N); T1d = VFNMS(LDK(KP618033988), T1c, T1b); T2H = VFMA(LDK(KP618033988), T1b, T1c); T17 = VADD(T11, T16); T19 = VSUB(T16, T11); { V T3f, T2T, T2C, T18, T3P; T3m = VFMA(LDK(KP447533225), T2B, T2A); T3f = VFNMS(LDK(KP494780565), T2A, T2B); T2T = VFNMS(LDK(KP132830569), T2A, T2B); T2C = VFMA(LDK(KP120146378), T2B, T2A); T2j = VFNMS(LDK(KP786782374), TK, TS); T2b = VFMA(LDK(KP869845200), TS, TK); TT = VFNMS(LDK(KP667278218), TS, TK); T1D = VFMA(LDK(KP603558818), TK, TS); T18 = VFNMS(LDK(KP250000000), T17, TW); T3P = VADD(TW, T17); T2G = VFMA(LDK(KP734762448), T2F, T2C); T35 = VFNMS(LDK(KP734762448), T2F, T2C); T2V = VFNMS(LDK(KP734762448), T2U, T2T); T2Z = VFMA(LDK(KP734762448), T2U, T2T); T3A = VFMA(LDK(KP982009705), T3f, T3e); T3g = VFNMS(LDK(KP982009705), T3f, T3e); T2I = VFMA(LDK(KP559016994), T19, T18); T1a = VFNMS(LDK(KP559016994), T19, T18); T3R = VADD(T3P, T3Q); T3X = VSUB(T3P, T3Q); } } { V T2n, T2t, T1V, T22, T2l, T2d, T1Q, T1I, T2w, T1A, T1F, T2q; { V T2k, T1G, T28, T2g, T3K, T3E, T3a, T34, T3x, T3H, T2c, TU, T1T, T1U, T1z; V T3o, T3t; T2n = VFNMS(LDK(KP912575812), T2j, T2i); T2k = VFMA(LDK(KP912575812), T2j, T2i); T3o = VFNMS(LDK(KP921078979), T3n, T3m); T3t = VFMA(LDK(KP921078979), T3n, T3m); { V T3c, T2Q, T2J, T3k, T1e; T3c = VFNMS(LDK(KP667278218), T2I, T2H); T2Q = VFNMS(LDK(KP059835404), T2H, T2I); T2J = VFMA(LDK(KP066152395), T2I, T2H); T3k = VFMA(LDK(KP603558818), T2H, T2I); T1G = VFMA(LDK(KP578046249), T1a, T1d); T1e = VFNMS(LDK(KP522847744), T1d, T1a); T28 = VFNMS(LDK(KP494780565), T1a, T1d); T2g = VFMA(LDK(KP447533225), T1d, T1a); { V T3U, T3S, T40, T3Y; T3U = VSUB(T3O, T3R); T3S = VADD(T3O, T3R); T40 = VMUL(LDK(KP951056516), VFNMS(LDK(KP618033988), T3W, T3X)); T3Y = VMUL(LDK(KP951056516), VFMA(LDK(KP618033988), T3X, T3W)); { V T3s, T3l, T2N, T36; T3s = VFNMS(LDK(KP845997307), T3k, T3j); T3l = VFMA(LDK(KP845997307), T3k, T3j); T2N = VFNMS(LDK(KP772036680), T2M, T2J); T36 = VFMA(LDK(KP772036680), T2M, T2J); { V T30, T2S, T3d, T3z, T3T; T30 = VFNMS(LDK(KP772036680), T2R, T2Q); T2S = VFMA(LDK(KP772036680), T2R, T2Q); T3d = VFNMS(LDK(KP845997307), T3c, T3b); T3z = VFMA(LDK(KP845997307), T3c, T3b); ST(&(x[0]), VADD(T3S, T3L), ms, &(x[0])); T3T = VFNMS(LDK(KP250000000), T3S, T3L); { V T3C, T3p, T2O, T37; T3C = VFMA(LDK(KP906616052), T3o, T3l); T3p = VFNMS(LDK(KP906616052), T3o, T3l); T2O = VFMA(LDK(KP956723877), T2N, T2G); T37 = VFMA(LDK(KP522616830), T2V, T36); { V T31, T2W, T3u, T3h; T31 = VFNMS(LDK(KP522616830), T2G, T30); T2W = VFMA(LDK(KP945422727), T2V, T2S); T3u = VFNMS(LDK(KP923225144), T3g, T3d); T3h = VFMA(LDK(KP923225144), T3g, T3d); { V T3I, T3B, T3V, T3Z; T3I = VFNMS(LDK(KP669429328), T3z, T3A); T3B = VFMA(LDK(KP570584518), T3A, T3z); T3V = VFMA(LDK(KP559016994), T3U, T3T); T3Z = VFNMS(LDK(KP559016994), T3U, T3T); { V T3y, T3q, T2P, T38; T3y = VFMA(LDK(KP262346850), T3p, T2X); T3q = VMUL(LDK(KP998026728), VFNMS(LDK(KP952936919), T2X, T3p)); T2P = VFMA(LDK(KP992114701), T2O, T2z); T38 = VFNMS(LDK(KP690983005), T37, T2S); { V T32, T2Y, T3v, T3F; T32 = VFMA(LDK(KP763932022), T31, T2N); T2Y = VMUL(LDK(KP998026728), VFMA(LDK(KP952936919), T2X, T2W)); T3v = VFNMS(LDK(KP997675361), T3u, T3t); T3F = VFNMS(LDK(KP904508497), T3u, T3s); { V T3i, T3r, T3J, T3D; T3i = VFMA(LDK(KP949179823), T3h, T2z); T3r = VFNMS(LDK(KP237294955), T3h, T2z); T3J = VFNMS(LDK(KP669429328), T3C, T3I); T3D = VFMA(LDK(KP618033988), T3C, T3B); ST(&(x[WS(rs, 20)]), VFNMSI(T3Y, T3V), ms, &(x[0])); ST(&(x[WS(rs, 5)]), VFMAI(T3Y, T3V), ms, &(x[WS(rs, 1)])); ST(&(x[WS(rs, 15)]), VFMAI(T40, T3Z), ms, &(x[WS(rs, 1)])); ST(&(x[WS(rs, 10)]), VFNMSI(T40, T3Z), ms, &(x[0])); { V T39, T33, T3w, T3G; T39 = VFMA(LDK(KP855719849), T38, T35); T33 = VFNMS(LDK(KP855719849), T32, T2Z); ST(&(x[WS(rs, 3)]), VFMAI(T2Y, T2P), ms, &(x[WS(rs, 1)])); ST(&(x[WS(rs, 22)]), VFNMSI(T2Y, T2P), ms, &(x[0])); T3w = VFMA(LDK(KP560319534), T3v, T3s); T3G = VFNMS(LDK(KP681693190), T3F, T3t); ST(&(x[WS(rs, 2)]), VFMAI(T3q, T3i), ms, &(x[0])); ST(&(x[WS(rs, 23)]), VFNMSI(T3q, T3i), ms, &(x[WS(rs, 1)])); T3K = VMUL(LDK(KP951056516), VFNMS(LDK(KP876306680), T3J, T3y)); T3E = VMUL(LDK(KP951056516), VFNMS(LDK(KP949179823), T3D, T3y)); T3a = VMUL(LDK(KP951056516), VFNMS(LDK(KP992114701), T39, T2X)); T34 = VFMA(LDK(KP897376177), T33, T2z); T3x = VFNMS(LDK(KP949179823), T3w, T3r); T3H = VFNMS(LDK(KP860541664), T3G, T3r); T2t = VFNMS(LDK(KP912575812), T2b, T2a); T2c = VFMA(LDK(KP912575812), T2b, T2a); TU = VFMA(LDK(KP829049696), TT, Tz); T1T = VFNMS(LDK(KP829049696), TT, Tz); T1U = VFNMS(LDK(KP831864738), T1y, T1e); T1z = VFMA(LDK(KP831864738), T1y, T1e); } } } } } } } } } } } { V T2o, T2h, T29, T2u, T2v, T2p; T2o = VFNMS(LDK(KP958953096), T2g, T2f); T2h = VFMA(LDK(KP958953096), T2g, T2f); ST(&(x[WS(rs, 17)]), VFNMSI(T3a, T34), ms, &(x[WS(rs, 1)])); ST(&(x[WS(rs, 8)]), VFMAI(T3a, T34), ms, &(x[0])); ST(&(x[WS(rs, 13)]), VFMAI(T3E, T3x), ms, &(x[WS(rs, 1)])); ST(&(x[WS(rs, 12)]), VFNMSI(T3E, T3x), ms, &(x[0])); ST(&(x[WS(rs, 7)]), VFNMSI(T3K, T3H), ms, &(x[WS(rs, 1)])); ST(&(x[WS(rs, 18)]), VFMAI(T3K, T3H), ms, &(x[0])); T1V = VFMA(LDK(KP559154169), T1U, T1T); T22 = VFNMS(LDK(KP683113946), T1T, T1U); T29 = VFNMS(LDK(KP867381224), T28, T27); T2u = VFMA(LDK(KP867381224), T28, T27); T2l = VFMA(LDK(KP894834959), T2k, T2h); T2v = VFMA(LDK(KP447417479), T2k, T2u); T2d = VFNMS(LDK(KP809385824), T2c, T29); T2p = VFMA(LDK(KP447417479), T2c, T2o); T1Q = VFMA(LDK(KP831864738), T1H, T1G); T1I = VFNMS(LDK(KP831864738), T1H, T1G); T2w = VFNMS(LDK(KP763932022), T2v, T2h); T1A = VFMA(LDK(KP904730450), T1z, TU); T1F = VFNMS(LDK(KP904730450), T1z, TU); T2q = VFMA(LDK(KP690983005), T2p, T29); } } { V T2e, T1E, T1P, T2m; T2e = VFNMS(LDK(KP992114701), T2d, Tf); T1E = VFMA(LDK(KP916574801), T1D, T1C); T1P = VFNMS(LDK(KP916574801), T1D, T1C); T2m = VMUL(LDK(KP951056516), VFNMS(LDK(KP992114701), T2l, T1O)); { V T1J, T2r, T1R, T1W, T1Z, T2x; T2x = VFNMS(LDK(KP999544308), T2w, T2t); T1J = VFNMS(LDK(KP904730450), T1I, T1F); T25 = VFMA(LDK(KP968583161), T1A, Tf); T1B = VFNMS(LDK(KP242145790), T1A, Tf); T2r = VFNMS(LDK(KP999544308), T2q, T2n); T1R = VFMA(LDK(KP904730450), T1Q, T1P); T1W = VFNMS(LDK(KP904730450), T1Q, T1P); T1Z = VADD(T1E, T1F); ST(&(x[WS(rs, 21)]), VFMAI(T2m, T2e), ms, &(x[WS(rs, 1)])); ST(&(x[WS(rs, 4)]), VFNMSI(T2m, T2e), ms, &(x[0])); T2y = VMUL(LDK(KP951056516), VFNMS(LDK(KP803003575), T2x, T1O)); T1K = VFNMS(LDK(KP618033988), T1J, T1E); T2s = VFNMS(LDK(KP803003575), T2r, Tf); T23 = VFMA(LDK(KP617882369), T1W, T22); T1S = VFNMS(LDK(KP242145790), T1R, T1O); T26 = VMUL(LDK(KP951056516), VFMA(LDK(KP968583161), T1R, T1O)); T20 = VFNMS(LDK(KP683113946), T1Z, T1I); T1X = VFMA(LDK(KP559016994), T1W, T1V); } } } } } { V T1L, T24, T21, T1Y; T1L = VFNMS(LDK(KP876091699), T1K, T1B); ST(&(x[WS(rs, 16)]), VFMAI(T2y, T2s), ms, &(x[0])); ST(&(x[WS(rs, 9)]), VFNMSI(T2y, T2s), ms, &(x[WS(rs, 1)])); T24 = VMUL(LDK(KP951056516), VFNMS(LDK(KP876306680), T23, T1S)); ST(&(x[WS(rs, 24)]), VFNMSI(T26, T25), ms, &(x[0])); ST(&(x[WS(rs, 1)]), VFMAI(T26, T25), ms, &(x[WS(rs, 1)])); T21 = VFMA(LDK(KP792626838), T20, T1B); T1Y = VMUL(LDK(KP951056516), VFMA(LDK(KP968583161), T1X, T1S)); ST(&(x[WS(rs, 11)]), VFMAI(T24, T21), ms, &(x[WS(rs, 1)])); ST(&(x[WS(rs, 14)]), VFNMSI(T24, T21), ms, &(x[0])); ST(&(x[WS(rs, 19)]), VFNMSI(T1Y, T1L), ms, &(x[WS(rs, 1)])); ST(&(x[WS(rs, 6)]), VFMAI(T1Y, T1L), ms, &(x[0])); } } } VLEAVE(); } static const tw_instr twinstr[] = { VTW(0, 1), VTW(0, 2), VTW(0, 3), VTW(0, 4), VTW(0, 5), VTW(0, 6), VTW(0, 7), VTW(0, 8), VTW(0, 9), VTW(0, 10), VTW(0, 11), VTW(0, 12), VTW(0, 13), VTW(0, 14), VTW(0, 15), VTW(0, 16), VTW(0, 17), VTW(0, 18), VTW(0, 19), VTW(0, 20), VTW(0, 21), VTW(0, 22), VTW(0, 23), VTW(0, 24), {TW_NEXT, VL, 0} }; static const ct_desc desc = { 25, XSIMD_STRING("t1bv_25"), twinstr, &GENUS, {67, 60, 181, 0}, 0, 0, 0 }; void XSIMD(codelet_t1bv_25) (planner *p) { X(kdft_dit_register) (p, t1bv_25, &desc); } #else /* HAVE_FMA */ /* Generated by: ../../../genfft/gen_twiddle_c.native -simd -compact -variables 4 -pipeline-latency 8 -n 25 -name t1bv_25 -include t1b.h -sign 1 */ /* * This function contains 248 FP additions, 188 FP multiplications, * (or, 171 additions, 111 multiplications, 77 fused multiply/add), * 100 stack variables, 40 constants, and 50 memory accesses */ #include "t1b.h" static void t1bv_25(R *ri, R *ii, const R *W, stride rs, INT mb, INT me, INT ms) { DVK(KP497379774, +0.497379774329709576484567492012895936835134813); DVK(KP968583161, +0.968583161128631119490168375464735813836012403); DVK(KP248689887, +0.248689887164854788242283746006447968417567406); DVK(KP1_937166322, +1.937166322257262238980336750929471627672024806); DVK(KP809016994, +0.809016994374947424102293417182819058860154590); DVK(KP309016994, +0.309016994374947424102293417182819058860154590); DVK(KP1_688655851, +1.688655851004030157097116127933363010763318483); DVK(KP535826794, +0.535826794978996618271308767867639978063575346); DVK(KP425779291, +0.425779291565072648862502445744251703979973042); DVK(KP1_809654104, +1.809654104932039055427337295865395187940827822); DVK(KP963507348, +0.963507348203430549974383005744259307057084020); DVK(KP876306680, +0.876306680043863587308115903922062583399064238); DVK(KP844327925, +0.844327925502015078548558063966681505381659241); DVK(KP1_071653589, +1.071653589957993236542617535735279956127150691); DVK(KP481753674, +0.481753674101715274987191502872129653528542010); DVK(KP1_752613360, +1.752613360087727174616231807844125166798128477); DVK(KP851558583, +0.851558583130145297725004891488503407959946084); DVK(KP904827052, +0.904827052466019527713668647932697593970413911); DVK(KP125333233, +0.125333233564304245373118759816508793942918247); DVK(KP1_984229402, +1.984229402628955662099586085571557042906073418); DVK(KP1_457937254, +1.457937254842823046293460638110518222745143328); DVK(KP684547105, +0.684547105928688673732283357621209269889519233); DVK(KP637423989, +0.637423989748689710176712811676016195434917298); DVK(KP1_541026485, +1.541026485551578461606019272792355694543335344); DVK(KP062790519, +0.062790519529313376076178224565631133122484832); DVK(KP1_996053456, +1.996053456856543123904673613726901106673810439); DVK(KP770513242, +0.770513242775789230803009636396177847271667672); DVK(KP1_274847979, +1.274847979497379420353425623352032390869834596); DVK(KP125581039, +0.125581039058626752152356449131262266244969664); DVK(KP998026728, +0.998026728428271561952336806863450553336905220); DVK(KP992114701, +0.992114701314477831049793042785778521453036709); DVK(KP250666467, +0.250666467128608490746237519633017587885836494); DVK(KP728968627, +0.728968627421411523146730319055259111372571664); DVK(KP1_369094211, +1.369094211857377347464566715242418539779038465); DVK(KP293892626, +0.293892626146236564584352977319536384298826219); DVK(KP475528258, +0.475528258147576786058219666689691071702849317); DVK(KP951056516, +0.951056516295153572116439333379382143405698634); DVK(KP250000000, +0.250000000000000000000000000000000000000000000); DVK(KP587785252, +0.587785252292473129168705954639072768597652438); DVK(KP559016994, +0.559016994374947424102293417182819058860154590); { INT m; R *x; x = ii; for (m = mb, W = W + (mb * ((TWVL / VL) * 48)); m < me; m = m + VL, x = x + (VL * ms), W = W + (TWVL * 48), MAKE_VOLATILE_STRIDE(25, rs)) { V T1A, T1z, T1R, T1S, T1B, T1C, T1Q, T2L, T1l, T2v, T1i, T3e, T2u, Tb, T2i; V Tj, T3b, T2h, Tv, T2k, TD, T3a, T2l, T11, T2s, TY, T3d, T2r; { V T1v, T1x, T1y, T1q, T1s, T1t, T1P; T1A = LD(&(x[0]), ms, &(x[0])); { V T1u, T1w, T1p, T1r; T1u = LD(&(x[WS(rs, 10)]), ms, &(x[0])); T1v = BYTW(&(W[TWVL * 18]), T1u); T1w = LD(&(x[WS(rs, 15)]), ms, &(x[WS(rs, 1)])); T1x = BYTW(&(W[TWVL * 28]), T1w); T1y = VADD(T1v, T1x); T1p = LD(&(x[WS(rs, 5)]), ms, &(x[WS(rs, 1)])); T1q = BYTW(&(W[TWVL * 8]), T1p); T1r = LD(&(x[WS(rs, 20)]), ms, &(x[0])); T1s = BYTW(&(W[TWVL * 38]), T1r); T1t = VADD(T1q, T1s); } T1z = VMUL(LDK(KP559016994), VSUB(T1t, T1y)); T1R = VSUB(T1v, T1x); T1S = VMUL(LDK(KP587785252), T1R); T1B = VADD(T1t, T1y); T1C = VFNMS(LDK(KP250000000), T1B, T1A); T1P = VSUB(T1q, T1s); T1Q = VMUL(LDK(KP951056516), T1P); T2L = VMUL(LDK(KP587785252), T1P); } { V T1f, T19, T1b, T1c, T14, T16, T17, T1e; T1e = LD(&(x[WS(rs, 3)]), ms, &(x[WS(rs, 1)])); T1f = BYTW(&(W[TWVL * 4]), T1e); { V T18, T1a, T13, T15; T18 = LD(&(x[WS(rs, 13)]), ms, &(x[WS(rs, 1)])); T19 = BYTW(&(W[TWVL * 24]), T18); T1a = LD(&(x[WS(rs, 18)]), ms, &(x[0])); T1b = BYTW(&(W[TWVL * 34]), T1a); T1c = VADD(T19, T1b); T13 = LD(&(x[WS(rs, 8)]), ms, &(x[0])); T14 = BYTW(&(W[TWVL * 14]), T13); T15 = LD(&(x[WS(rs, 23)]), ms, &(x[WS(rs, 1)])); T16 = BYTW(&(W[TWVL * 44]), T15); T17 = VADD(T14, T16); } { V T1j, T1k, T1d, T1g, T1h; T1j = VSUB(T14, T16); T1k = VSUB(T19, T1b); T1l = VFMA(LDK(KP475528258), T1j, VMUL(LDK(KP293892626), T1k)); T2v = VFNMS(LDK(KP475528258), T1k, VMUL(LDK(KP293892626), T1j)); T1d = VMUL(LDK(KP559016994), VSUB(T17, T1c)); T1g = VADD(T17, T1c); T1h = VFNMS(LDK(KP250000000), T1g, T1f); T1i = VADD(T1d, T1h); T3e = VADD(T1f, T1g); T2u = VSUB(T1h, T1d); } } { V Tg, T7, T9, Td, T2, T4, Tc, Tf; Tf = LD(&(x[WS(rs, 4)]), ms, &(x[0])); Tg = BYTW(&(W[TWVL * 6]), Tf); { V T6, T8, T1, T3; T6 = LD(&(x[WS(rs, 14)]), ms, &(x[0])); T7 = BYTW(&(W[TWVL * 26]), T6); T8 = LD(&(x[WS(rs, 19)]), ms, &(x[WS(rs, 1)])); T9 = BYTW(&(W[TWVL * 36]), T8); Td = VADD(T7, T9); T1 = LD(&(x[WS(rs, 9)]), ms, &(x[WS(rs, 1)])); T2 = BYTW(&(W[TWVL * 16]), T1); T3 = LD(&(x[WS(rs, 24)]), ms, &(x[0])); T4 = BYTW(&(W[TWVL * 46]), T3); Tc = VADD(T2, T4); } { V T5, Ta, Te, Th, Ti; T5 = VSUB(T2, T4); Ta = VSUB(T7, T9); Tb = VFMA(LDK(KP475528258), T5, VMUL(LDK(KP293892626), Ta)); T2i = VFNMS(LDK(KP475528258), Ta, VMUL(LDK(KP293892626), T5)); Te = VMUL(LDK(KP559016994), VSUB(Tc, Td)); Th = VADD(Tc, Td); Ti = VFNMS(LDK(KP250000000), Th, Tg); Tj = VADD(Te, Ti); T3b = VADD(Tg, Th); T2h = VSUB(Ti, Te); } } { V TA, Tr, Tt, Tx, Tm, To, Tw, Tz; Tz = LD(&(x[WS(rs, 1)]), ms, &(x[WS(rs, 1)])); TA = BYTW(&(W[0]), Tz); { V Tq, Ts, Tl, Tn; Tq = LD(&(x[WS(rs, 11)]), ms, &(x[WS(rs, 1)])); Tr = BYTW(&(W[TWVL * 20]), Tq); Ts = LD(&(x[WS(rs, 16)]), ms, &(x[0])); Tt = BYTW(&(W[TWVL * 30]), Ts); Tx = VADD(Tr, Tt); Tl = LD(&(x[WS(rs, 6)]), ms, &(x[0])); Tm = BYTW(&(W[TWVL * 10]), Tl); Tn = LD(&(x[WS(rs, 21)]), ms, &(x[WS(rs, 1)])); To = BYTW(&(W[TWVL * 40]), Tn); Tw = VADD(Tm, To); } { V Tp, Tu, Ty, TB, TC; Tp = VSUB(Tm, To); Tu = VSUB(Tr, Tt); Tv = VFMA(LDK(KP475528258), Tp, VMUL(LDK(KP293892626), Tu)); T2k = VFNMS(LDK(KP475528258), Tu, VMUL(LDK(KP293892626), Tp)); Ty = VMUL(LDK(KP559016994), VSUB(Tw, Tx)); TB = VADD(Tw, Tx); TC = VFNMS(LDK(KP250000000), TB, TA); TD = VADD(Ty, TC); T3a = VADD(TA, TB); T2l = VSUB(TC, Ty); } } { V TV, TP, TR, TS, TK, TM, TN, TU; TU = LD(&(x[WS(rs, 2)]), ms, &(x[0])); TV = BYTW(&(W[TWVL * 2]), TU); { V TO, TQ, TJ, TL; TO = LD(&(x[WS(rs, 12)]), ms, &(x[0])); TP = BYTW(&(W[TWVL * 22]), TO); TQ = LD(&(x[WS(rs, 17)]), ms, &(x[WS(rs, 1)])); TR = BYTW(&(W[TWVL * 32]), TQ); TS = VADD(TP, TR); TJ = LD(&(x[WS(rs, 7)]), ms, &(x[WS(rs, 1)])); TK = BYTW(&(W[TWVL * 12]), TJ); TL = LD(&(x[WS(rs, 22)]), ms, &(x[0])); TM = BYTW(&(W[TWVL * 42]), TL); TN = VADD(TK, TM); } { V TZ, T10, TT, TW, TX; TZ = VSUB(TK, TM); T10 = VSUB(TP, TR); T11 = VFMA(LDK(KP475528258), TZ, VMUL(LDK(KP293892626), T10)); T2s = VFNMS(LDK(KP475528258), T10, VMUL(LDK(KP293892626), TZ)); TT = VMUL(LDK(KP559016994), VSUB(TN, TS)); TW = VADD(TN, TS); TX = VFNMS(LDK(KP250000000), TW, TV); TY = VADD(TT, TX); T3d = VADD(TV, TW); T2r = VSUB(TX, TT); } } { V T3g, T3o, T3k, T3l, T3j, T3m, T3p, T3n; { V T3c, T3f, T3h, T3i; T3c = VSUB(T3a, T3b); T3f = VSUB(T3d, T3e); T3g = VBYI(VFMA(LDK(KP951056516), T3c, VMUL(LDK(KP587785252), T3f))); T3o = VBYI(VFNMS(LDK(KP951056516), T3f, VMUL(LDK(KP587785252), T3c))); T3k = VADD(T1A, T1B); T3h = VADD(T3a, T3b); T3i = VADD(T3d, T3e); T3l = VADD(T3h, T3i); T3j = VMUL(LDK(KP559016994), VSUB(T3h, T3i)); T3m = VFNMS(LDK(KP250000000), T3l, T3k); } ST(&(x[0]), VADD(T3k, T3l), ms, &(x[0])); T3p = VSUB(T3m, T3j); ST(&(x[WS(rs, 10)]), VADD(T3o, T3p), ms, &(x[0])); ST(&(x[WS(rs, 15)]), VSUB(T3p, T3o), ms, &(x[WS(rs, 1)])); T3n = VADD(T3j, T3m); ST(&(x[WS(rs, 5)]), VADD(T3g, T3n), ms, &(x[WS(rs, 1)])); ST(&(x[WS(rs, 20)]), VSUB(T3n, T3g), ms, &(x[0])); } { V T2z, T2M, T2U, T2V, T2W, T34, T35, T36, T2X, T2Y, T2Z, T31, T32, T33, T2n; V T2N, T2E, T2K, T2y, T2H, T2A, T2G, T38, T39; T2z = VSUB(T1C, T1z); T2M = VFNMS(LDK(KP951056516), T1R, T2L); T2U = VFMA(LDK(KP1_369094211), T2k, VMUL(LDK(KP728968627), T2l)); T2V = VFNMS(LDK(KP992114701), T2h, VMUL(LDK(KP250666467), T2i)); T2W = VADD(T2U, T2V); T34 = VFNMS(LDK(KP125581039), T2s, VMUL(LDK(KP998026728), T2r)); T35 = VFMA(LDK(KP1_274847979), T2v, VMUL(LDK(KP770513242), T2u)); T36 = VADD(T34, T35); T2X = VFMA(LDK(KP1_996053456), T2s, VMUL(LDK(KP062790519), T2r)); T2Y = VFNMS(LDK(KP637423989), T2u, VMUL(LDK(KP1_541026485), T2v)); T2Z = VADD(T2X, T2Y); T31 = VFNMS(LDK(KP1_457937254), T2k, VMUL(LDK(KP684547105), T2l)); T32 = VFMA(LDK(KP1_984229402), T2i, VMUL(LDK(KP125333233), T2h)); T33 = VADD(T31, T32); { V T2j, T2m, T2I, T2C, T2D, T2J; T2j = VFNMS(LDK(KP851558583), T2i, VMUL(LDK(KP904827052), T2h)); T2m = VFMA(LDK(KP1_752613360), T2k, VMUL(LDK(KP481753674), T2l)); T2I = VADD(T2m, T2j); T2C = VFMA(LDK(KP1_071653589), T2s, VMUL(LDK(KP844327925), T2r)); T2D = VFMA(LDK(KP125581039), T2v, VMUL(LDK(KP998026728), T2u)); T2J = VADD(T2C, T2D); T2n = VSUB(T2j, T2m); T2N = VADD(T2I, T2J); T2E = VSUB(T2C, T2D); T2K = VMUL(LDK(KP559016994), VSUB(T2I, T2J)); } { V T2o, T2p, T2q, T2t, T2w, T2x; T2o = VFNMS(LDK(KP963507348), T2k, VMUL(LDK(KP876306680), T2l)); T2p = VFMA(LDK(KP1_809654104), T2i, VMUL(LDK(KP425779291), T2h)); T2q = VSUB(T2o, T2p); T2t = VFNMS(LDK(KP1_688655851), T2s, VMUL(LDK(KP535826794), T2r)); T2w = VFNMS(LDK(KP1_996053456), T2v, VMUL(LDK(KP062790519), T2u)); T2x = VADD(T2t, T2w); T2y = VMUL(LDK(KP559016994), VSUB(T2q, T2x)); T2H = VSUB(T2t, T2w); T2A = VADD(T2q, T2x); T2G = VADD(T2o, T2p); } { V T2S, T2T, T30, T37; T2S = VADD(T2z, T2A); T2T = VBYI(VADD(T2M, T2N)); ST(&(x[WS(rs, 23)]), VSUB(T2S, T2T), ms, &(x[WS(rs, 1)])); ST(&(x[WS(rs, 2)]), VADD(T2S, T2T), ms, &(x[0])); T30 = VADD(T2z, VADD(T2W, T2Z)); T37 = VBYI(VSUB(VADD(T33, T36), T2M)); ST(&(x[WS(rs, 22)]), VSUB(T30, T37), ms, &(x[0])); ST(&(x[WS(rs, 3)]), VADD(T30, T37), ms, &(x[WS(rs, 1)])); } T38 = VBYI(VSUB(VFMA(LDK(KP951056516), VSUB(T2U, T2V), VFMA(LDK(KP309016994), T33, VFNMS(LDK(KP809016994), T36, VMUL(LDK(KP587785252), VSUB(T2X, T2Y))))), T2M)); T39 = VFMA(LDK(KP309016994), T2W, VFMA(LDK(KP951056516), VSUB(T32, T31), VFMA(LDK(KP587785252), VSUB(T35, T34), VFNMS(LDK(KP809016994), T2Z, T2z)))); ST(&(x[WS(rs, 8)]), VADD(T38, T39), ms, &(x[0])); ST(&(x[WS(rs, 17)]), VSUB(T39, T38), ms, &(x[WS(rs, 1)])); { V T2F, T2Q, T2P, T2R, T2B, T2O; T2B = VFNMS(LDK(KP250000000), T2A, T2z); T2F = VFMA(LDK(KP951056516), T2n, VADD(T2y, VFNMS(LDK(KP587785252), T2E, T2B))); T2Q = VFMA(LDK(KP587785252), T2n, VFMA(LDK(KP951056516), T2E, VSUB(T2B, T2y))); T2O = VFNMS(LDK(KP250000000), T2N, T2M); T2P = VBYI(VADD(VFMA(LDK(KP951056516), T2G, VMUL(LDK(KP587785252), T2H)), VADD(T2K, T2O))); T2R = VBYI(VADD(VFNMS(LDK(KP951056516), T2H, VMUL(LDK(KP587785252), T2G)), VSUB(T2O, T2K))); ST(&(x[WS(rs, 18)]), VSUB(T2F, T2P), ms, &(x[0])); ST(&(x[WS(rs, 12)]), VADD(T2Q, T2R), ms, &(x[0])); ST(&(x[WS(rs, 7)]), VADD(T2F, T2P), ms, &(x[WS(rs, 1)])); ST(&(x[WS(rs, 13)]), VSUB(T2Q, T2R), ms, &(x[WS(rs, 1)])); } } { V T1D, T1T, T21, T22, T23, T2b, T2c, T2d, T24, T25, T26, T28, T29, T2a, TF; V T1U, T1I, T1O, T1o, T1L, T1E, T1K, T2f, T2g; T1D = VADD(T1z, T1C); T1T = VADD(T1Q, T1S); T21 = VFMA(LDK(KP1_688655851), Tv, VMUL(LDK(KP535826794), TD)); T22 = VFMA(LDK(KP1_541026485), Tb, VMUL(LDK(KP637423989), Tj)); T23 = VSUB(T21, T22); T2b = VFMA(LDK(KP851558583), T11, VMUL(LDK(KP904827052), TY)); T2c = VFMA(LDK(KP1_984229402), T1l, VMUL(LDK(KP125333233), T1i)); T2d = VADD(T2b, T2c); T24 = VFNMS(LDK(KP425779291), TY, VMUL(LDK(KP1_809654104), T11)); T25 = VFNMS(LDK(KP992114701), T1i, VMUL(LDK(KP250666467), T1l)); T26 = VADD(T24, T25); T28 = VFNMS(LDK(KP1_071653589), Tv, VMUL(LDK(KP844327925), TD)); T29 = VFNMS(LDK(KP770513242), Tj, VMUL(LDK(KP1_274847979), Tb)); T2a = VADD(T28, T29); { V Tk, TE, T1M, T1G, T1H, T1N; Tk = VFMA(LDK(KP1_071653589), Tb, VMUL(LDK(KP844327925), Tj)); TE = VFMA(LDK(KP1_937166322), Tv, VMUL(LDK(KP248689887), TD)); T1M = VADD(TE, Tk); T1G = VFMA(LDK(KP1_752613360), T11, VMUL(LDK(KP481753674), TY)); T1H = VFMA(LDK(KP1_457937254), T1l, VMUL(LDK(KP684547105), T1i)); T1N = VADD(T1G, T1H); TF = VSUB(Tk, TE); T1U = VADD(T1M, T1N); T1I = VSUB(T1G, T1H); T1O = VMUL(LDK(KP559016994), VSUB(T1M, T1N)); } { V TG, TH, TI, T12, T1m, T1n; TG = VFNMS(LDK(KP497379774), Tv, VMUL(LDK(KP968583161), TD)); TH = VFNMS(LDK(KP1_688655851), Tb, VMUL(LDK(KP535826794), Tj)); TI = VADD(TG, TH); T12 = VFNMS(LDK(KP963507348), T11, VMUL(LDK(KP876306680), TY)); T1m = VFNMS(LDK(KP1_369094211), T1l, VMUL(LDK(KP728968627), T1i)); T1n = VADD(T12, T1m); T1o = VMUL(LDK(KP559016994), VSUB(TI, T1n)); T1L = VSUB(T12, T1m); T1E = VADD(TI, T1n); T1K = VSUB(TG, TH); } { V T1Z, T20, T27, T2e; T1Z = VADD(T1D, T1E); T20 = VBYI(VADD(T1T, T1U)); ST(&(x[WS(rs, 24)]), VSUB(T1Z, T20), ms, &(x[0])); ST(&(x[WS(rs, 1)]), VADD(T1Z, T20), ms, &(x[WS(rs, 1)])); T27 = VADD(T1D, VADD(T23, T26)); T2e = VBYI(VSUB(VADD(T2a, T2d), T1T)); ST(&(x[WS(rs, 21)]), VSUB(T27, T2e), ms, &(x[WS(rs, 1)])); ST(&(x[WS(rs, 4)]), VADD(T27, T2e), ms, &(x[0])); } T2f = VBYI(VSUB(VFMA(LDK(KP309016994), T2a, VFMA(LDK(KP951056516), VADD(T21, T22), VFNMS(LDK(KP809016994), T2d, VMUL(LDK(KP587785252), VSUB(T24, T25))))), T1T)); T2g = VFMA(LDK(KP951056516), VSUB(T29, T28), VFMA(LDK(KP309016994), T23, VFMA(LDK(KP587785252), VSUB(T2c, T2b), VFNMS(LDK(KP809016994), T26, T1D)))); ST(&(x[WS(rs, 9)]), VADD(T2f, T2g), ms, &(x[WS(rs, 1)])); ST(&(x[WS(rs, 16)]), VSUB(T2g, T2f), ms, &(x[0])); { V T1J, T1X, T1W, T1Y, T1F, T1V; T1F = VFNMS(LDK(KP250000000), T1E, T1D); T1J = VFMA(LDK(KP951056516), TF, VADD(T1o, VFNMS(LDK(KP587785252), T1I, T1F))); T1X = VFMA(LDK(KP587785252), TF, VFMA(LDK(KP951056516), T1I, VSUB(T1F, T1o))); T1V = VFNMS(LDK(KP250000000), T1U, T1T); T1W = VBYI(VADD(VFMA(LDK(KP951056516), T1K, VMUL(LDK(KP587785252), T1L)), VADD(T1O, T1V))); T1Y = VBYI(VADD(VFNMS(LDK(KP951056516), T1L, VMUL(LDK(KP587785252), T1K)), VSUB(T1V, T1O))); ST(&(x[WS(rs, 19)]), VSUB(T1J, T1W), ms, &(x[WS(rs, 1)])); ST(&(x[WS(rs, 11)]), VADD(T1X, T1Y), ms, &(x[WS(rs, 1)])); ST(&(x[WS(rs, 6)]), VADD(T1J, T1W), ms, &(x[0])); ST(&(x[WS(rs, 14)]), VSUB(T1X, T1Y), ms, &(x[0])); } } } } VLEAVE(); } static const tw_instr twinstr[] = { VTW(0, 1), VTW(0, 2), VTW(0, 3), VTW(0, 4), VTW(0, 5), VTW(0, 6), VTW(0, 7), VTW(0, 8), VTW(0, 9), VTW(0, 10), VTW(0, 11), VTW(0, 12), VTW(0, 13), VTW(0, 14), VTW(0, 15), VTW(0, 16), VTW(0, 17), VTW(0, 18), VTW(0, 19), VTW(0, 20), VTW(0, 21), VTW(0, 22), VTW(0, 23), VTW(0, 24), {TW_NEXT, VL, 0} }; static const ct_desc desc = { 25, XSIMD_STRING("t1bv_25"), twinstr, &GENUS, {171, 111, 77, 0}, 0, 0, 0 }; void XSIMD(codelet_t1bv_25) (planner *p) { X(kdft_dit_register) (p, t1bv_25, &desc); } #endif /* HAVE_FMA */ fftw-3.3.3/dft/simd/common/t2bv_2.c0000644000175000017500000000655412054410555013665 00000000000000/* * Copyright (c) 2003, 2007-11 Matteo Frigo * Copyright (c) 2003, 2007-11 Massachusetts Institute of Technology * * This program is free software; you can redistribute it and/or modify * it under the terms of the GNU General Public License as published by * the Free Software Foundation; either version 2 of the License, or * (at your option) any later version. * * This program is distributed in the hope that it will be useful, * but WITHOUT ANY WARRANTY; without even the implied warranty of * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the * GNU General Public License for more details. * * You should have received a copy of the GNU General Public License * along with this program; if not, write to the Free Software * Foundation, Inc., 51 Franklin Street, Fifth Floor, Boston, MA 02110-1301 USA * */ /* This file was automatically generated --- DO NOT EDIT */ /* Generated on Sun Nov 25 07:39:09 EST 2012 */ #include "codelet-dft.h" #ifdef HAVE_FMA /* Generated by: ../../../genfft/gen_twiddle_c.native -fma -reorder-insns -schedule-for-pipeline -simd -compact -variables 4 -pipeline-latency 8 -n 2 -name t2bv_2 -include t2b.h -sign 1 */ /* * This function contains 3 FP additions, 2 FP multiplications, * (or, 3 additions, 2 multiplications, 0 fused multiply/add), * 5 stack variables, 0 constants, and 4 memory accesses */ #include "t2b.h" static void t2bv_2(R *ri, R *ii, const R *W, stride rs, INT mb, INT me, INT ms) { { INT m; R *x; x = ii; for (m = mb, W = W + (mb * ((TWVL / VL) * 2)); m < me; m = m + VL, x = x + (VL * ms), W = W + (TWVL * 2), MAKE_VOLATILE_STRIDE(2, rs)) { V T1, T2, T3; T1 = LD(&(x[0]), ms, &(x[0])); T2 = LD(&(x[WS(rs, 1)]), ms, &(x[WS(rs, 1)])); T3 = BYTW(&(W[0]), T2); ST(&(x[0]), VADD(T1, T3), ms, &(x[0])); ST(&(x[WS(rs, 1)]), VSUB(T1, T3), ms, &(x[WS(rs, 1)])); } } VLEAVE(); } static const tw_instr twinstr[] = { VTW(0, 1), {TW_NEXT, VL, 0} }; static const ct_desc desc = { 2, XSIMD_STRING("t2bv_2"), twinstr, &GENUS, {3, 2, 0, 0}, 0, 0, 0 }; void XSIMD(codelet_t2bv_2) (planner *p) { X(kdft_dit_register) (p, t2bv_2, &desc); } #else /* HAVE_FMA */ /* Generated by: ../../../genfft/gen_twiddle_c.native -simd -compact -variables 4 -pipeline-latency 8 -n 2 -name t2bv_2 -include t2b.h -sign 1 */ /* * This function contains 3 FP additions, 2 FP multiplications, * (or, 3 additions, 2 multiplications, 0 fused multiply/add), * 5 stack variables, 0 constants, and 4 memory accesses */ #include "t2b.h" static void t2bv_2(R *ri, R *ii, const R *W, stride rs, INT mb, INT me, INT ms) { { INT m; R *x; x = ii; for (m = mb, W = W + (mb * ((TWVL / VL) * 2)); m < me; m = m + VL, x = x + (VL * ms), W = W + (TWVL * 2), MAKE_VOLATILE_STRIDE(2, rs)) { V T1, T3, T2; T1 = LD(&(x[0]), ms, &(x[0])); T2 = LD(&(x[WS(rs, 1)]), ms, &(x[WS(rs, 1)])); T3 = BYTW(&(W[0]), T2); ST(&(x[WS(rs, 1)]), VSUB(T1, T3), ms, &(x[WS(rs, 1)])); ST(&(x[0]), VADD(T1, T3), ms, &(x[0])); } } VLEAVE(); } static const tw_instr twinstr[] = { VTW(0, 1), {TW_NEXT, VL, 0} }; static const ct_desc desc = { 2, XSIMD_STRING("t2bv_2"), twinstr, &GENUS, {3, 2, 0, 0}, 0, 0, 0 }; void XSIMD(codelet_t2bv_2) (planner *p) { X(kdft_dit_register) (p, t2bv_2, &desc); } #endif /* HAVE_FMA */ fftw-3.3.3/dft/simd/common/t2bv_4.c0000644000175000017500000001050312054410555013654 00000000000000/* * Copyright (c) 2003, 2007-11 Matteo Frigo * Copyright (c) 2003, 2007-11 Massachusetts Institute of Technology * * This program is free software; you can redistribute it and/or modify * it under the terms of the GNU General Public License as published by * the Free Software Foundation; either version 2 of the License, or * (at your option) any later version. * * This program is distributed in the hope that it will be useful, * but WITHOUT ANY WARRANTY; without even the implied warranty of * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the * GNU General Public License for more details. * * You should have received a copy of the GNU General Public License * along with this program; if not, write to the Free Software * Foundation, Inc., 51 Franklin Street, Fifth Floor, Boston, MA 02110-1301 USA * */ /* This file was automatically generated --- DO NOT EDIT */ /* Generated on Sun Nov 25 07:39:09 EST 2012 */ #include "codelet-dft.h" #ifdef HAVE_FMA /* Generated by: ../../../genfft/gen_twiddle_c.native -fma -reorder-insns -schedule-for-pipeline -simd -compact -variables 4 -pipeline-latency 8 -n 4 -name t2bv_4 -include t2b.h -sign 1 */ /* * This function contains 11 FP additions, 8 FP multiplications, * (or, 9 additions, 6 multiplications, 2 fused multiply/add), * 13 stack variables, 0 constants, and 8 memory accesses */ #include "t2b.h" static void t2bv_4(R *ri, R *ii, const R *W, stride rs, INT mb, INT me, INT ms) { { INT m; R *x; x = ii; for (m = mb, W = W + (mb * ((TWVL / VL) * 6)); m < me; m = m + VL, x = x + (VL * ms), W = W + (TWVL * 6), MAKE_VOLATILE_STRIDE(4, rs)) { V T1, T7, T2, T5, T8, T3, T6; T1 = LD(&(x[0]), ms, &(x[0])); T7 = LD(&(x[WS(rs, 3)]), ms, &(x[WS(rs, 1)])); T2 = LD(&(x[WS(rs, 2)]), ms, &(x[0])); T5 = LD(&(x[WS(rs, 1)]), ms, &(x[WS(rs, 1)])); T8 = BYTW(&(W[TWVL * 4]), T7); T3 = BYTW(&(W[TWVL * 2]), T2); T6 = BYTW(&(W[0]), T5); { V Ta, T4, Tb, T9; Ta = VADD(T1, T3); T4 = VSUB(T1, T3); Tb = VADD(T6, T8); T9 = VSUB(T6, T8); ST(&(x[0]), VADD(Ta, Tb), ms, &(x[0])); ST(&(x[WS(rs, 2)]), VSUB(Ta, Tb), ms, &(x[0])); ST(&(x[WS(rs, 1)]), VFMAI(T9, T4), ms, &(x[WS(rs, 1)])); ST(&(x[WS(rs, 3)]), VFNMSI(T9, T4), ms, &(x[WS(rs, 1)])); } } } VLEAVE(); } static const tw_instr twinstr[] = { VTW(0, 1), VTW(0, 2), VTW(0, 3), {TW_NEXT, VL, 0} }; static const ct_desc desc = { 4, XSIMD_STRING("t2bv_4"), twinstr, &GENUS, {9, 6, 2, 0}, 0, 0, 0 }; void XSIMD(codelet_t2bv_4) (planner *p) { X(kdft_dit_register) (p, t2bv_4, &desc); } #else /* HAVE_FMA */ /* Generated by: ../../../genfft/gen_twiddle_c.native -simd -compact -variables 4 -pipeline-latency 8 -n 4 -name t2bv_4 -include t2b.h -sign 1 */ /* * This function contains 11 FP additions, 6 FP multiplications, * (or, 11 additions, 6 multiplications, 0 fused multiply/add), * 13 stack variables, 0 constants, and 8 memory accesses */ #include "t2b.h" static void t2bv_4(R *ri, R *ii, const R *W, stride rs, INT mb, INT me, INT ms) { { INT m; R *x; x = ii; for (m = mb, W = W + (mb * ((TWVL / VL) * 6)); m < me; m = m + VL, x = x + (VL * ms), W = W + (TWVL * 6), MAKE_VOLATILE_STRIDE(4, rs)) { V T1, T8, T3, T6, T7, T2, T5; T1 = LD(&(x[0]), ms, &(x[0])); T7 = LD(&(x[WS(rs, 3)]), ms, &(x[WS(rs, 1)])); T8 = BYTW(&(W[TWVL * 4]), T7); T2 = LD(&(x[WS(rs, 2)]), ms, &(x[0])); T3 = BYTW(&(W[TWVL * 2]), T2); T5 = LD(&(x[WS(rs, 1)]), ms, &(x[WS(rs, 1)])); T6 = BYTW(&(W[0]), T5); { V T4, T9, Ta, Tb; T4 = VSUB(T1, T3); T9 = VBYI(VSUB(T6, T8)); ST(&(x[WS(rs, 3)]), VSUB(T4, T9), ms, &(x[WS(rs, 1)])); ST(&(x[WS(rs, 1)]), VADD(T4, T9), ms, &(x[WS(rs, 1)])); Ta = VADD(T1, T3); Tb = VADD(T6, T8); ST(&(x[WS(rs, 2)]), VSUB(Ta, Tb), ms, &(x[0])); ST(&(x[0]), VADD(Ta, Tb), ms, &(x[0])); } } } VLEAVE(); } static const tw_instr twinstr[] = { VTW(0, 1), VTW(0, 2), VTW(0, 3), {TW_NEXT, VL, 0} }; static const ct_desc desc = { 4, XSIMD_STRING("t2bv_4"), twinstr, &GENUS, {11, 6, 0, 0}, 0, 0, 0 }; void XSIMD(codelet_t2bv_4) (planner *p) { X(kdft_dit_register) (p, t2bv_4, &desc); } #endif /* HAVE_FMA */ fftw-3.3.3/dft/simd/common/t2bv_8.c0000644000175000017500000001560512054410555013670 00000000000000/* * Copyright (c) 2003, 2007-11 Matteo Frigo * Copyright (c) 2003, 2007-11 Massachusetts Institute of Technology * * This program is free software; you can redistribute it and/or modify * it under the terms of the GNU General Public License as published by * the Free Software Foundation; either version 2 of the License, or * (at your option) any later version. * * This program is distributed in the hope that it will be useful, * but WITHOUT ANY WARRANTY; without even the implied warranty of * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the * GNU General Public License for more details. * * You should have received a copy of the GNU General Public License * along with this program; if not, write to the Free Software * Foundation, Inc., 51 Franklin Street, Fifth Floor, Boston, MA 02110-1301 USA * */ /* This file was automatically generated --- DO NOT EDIT */ /* Generated on Sun Nov 25 07:39:09 EST 2012 */ #include "codelet-dft.h" #ifdef HAVE_FMA /* Generated by: ../../../genfft/gen_twiddle_c.native -fma -reorder-insns -schedule-for-pipeline -simd -compact -variables 4 -pipeline-latency 8 -n 8 -name t2bv_8 -include t2b.h -sign 1 */ /* * This function contains 33 FP additions, 24 FP multiplications, * (or, 23 additions, 14 multiplications, 10 fused multiply/add), * 36 stack variables, 1 constants, and 16 memory accesses */ #include "t2b.h" static void t2bv_8(R *ri, R *ii, const R *W, stride rs, INT mb, INT me, INT ms) { DVK(KP707106781, +0.707106781186547524400844362104849039284835938); { INT m; R *x; x = ii; for (m = mb, W = W + (mb * ((TWVL / VL) * 14)); m < me; m = m + VL, x = x + (VL * ms), W = W + (TWVL * 14), MAKE_VOLATILE_STRIDE(8, rs)) { V T1, T2, Th, Tj, T5, T7, Ta, Tc; T1 = LD(&(x[0]), ms, &(x[0])); T2 = LD(&(x[WS(rs, 4)]), ms, &(x[0])); Th = LD(&(x[WS(rs, 2)]), ms, &(x[0])); Tj = LD(&(x[WS(rs, 6)]), ms, &(x[0])); T5 = LD(&(x[WS(rs, 1)]), ms, &(x[WS(rs, 1)])); T7 = LD(&(x[WS(rs, 5)]), ms, &(x[WS(rs, 1)])); Ta = LD(&(x[WS(rs, 7)]), ms, &(x[WS(rs, 1)])); Tc = LD(&(x[WS(rs, 3)]), ms, &(x[WS(rs, 1)])); { V T3, Ti, Tk, T6, T8, Tb, Td; T3 = BYTW(&(W[TWVL * 6]), T2); Ti = BYTW(&(W[TWVL * 2]), Th); Tk = BYTW(&(W[TWVL * 10]), Tj); T6 = BYTW(&(W[0]), T5); T8 = BYTW(&(W[TWVL * 8]), T7); Tb = BYTW(&(W[TWVL * 12]), Ta); Td = BYTW(&(W[TWVL * 4]), Tc); { V Tq, T4, Tr, Tl, Tt, T9, Tu, Te, Tw, Ts; Tq = VADD(T1, T3); T4 = VSUB(T1, T3); Tr = VADD(Ti, Tk); Tl = VSUB(Ti, Tk); Tt = VADD(T6, T8); T9 = VSUB(T6, T8); Tu = VADD(Tb, Td); Te = VSUB(Tb, Td); Tw = VADD(Tq, Tr); Ts = VSUB(Tq, Tr); { V Tx, Tv, Tm, Tf; Tx = VADD(Tt, Tu); Tv = VSUB(Tt, Tu); Tm = VSUB(T9, Te); Tf = VADD(T9, Te); { V Tp, Tn, To, Tg; ST(&(x[0]), VADD(Tw, Tx), ms, &(x[0])); ST(&(x[WS(rs, 4)]), VSUB(Tw, Tx), ms, &(x[0])); ST(&(x[WS(rs, 2)]), VFMAI(Tv, Ts), ms, &(x[0])); ST(&(x[WS(rs, 6)]), VFNMSI(Tv, Ts), ms, &(x[0])); Tp = VFMA(LDK(KP707106781), Tm, Tl); Tn = VFNMS(LDK(KP707106781), Tm, Tl); To = VFMA(LDK(KP707106781), Tf, T4); Tg = VFNMS(LDK(KP707106781), Tf, T4); ST(&(x[WS(rs, 1)]), VFMAI(Tp, To), ms, &(x[WS(rs, 1)])); ST(&(x[WS(rs, 7)]), VFNMSI(Tp, To), ms, &(x[WS(rs, 1)])); ST(&(x[WS(rs, 5)]), VFMAI(Tn, Tg), ms, &(x[WS(rs, 1)])); ST(&(x[WS(rs, 3)]), VFNMSI(Tn, Tg), ms, &(x[WS(rs, 1)])); } } } } } } VLEAVE(); } static const tw_instr twinstr[] = { VTW(0, 1), VTW(0, 2), VTW(0, 3), VTW(0, 4), VTW(0, 5), VTW(0, 6), VTW(0, 7), {TW_NEXT, VL, 0} }; static const ct_desc desc = { 8, XSIMD_STRING("t2bv_8"), twinstr, &GENUS, {23, 14, 10, 0}, 0, 0, 0 }; void XSIMD(codelet_t2bv_8) (planner *p) { X(kdft_dit_register) (p, t2bv_8, &desc); } #else /* HAVE_FMA */ /* Generated by: ../../../genfft/gen_twiddle_c.native -simd -compact -variables 4 -pipeline-latency 8 -n 8 -name t2bv_8 -include t2b.h -sign 1 */ /* * This function contains 33 FP additions, 16 FP multiplications, * (or, 33 additions, 16 multiplications, 0 fused multiply/add), * 24 stack variables, 1 constants, and 16 memory accesses */ #include "t2b.h" static void t2bv_8(R *ri, R *ii, const R *W, stride rs, INT mb, INT me, INT ms) { DVK(KP707106781, +0.707106781186547524400844362104849039284835938); { INT m; R *x; x = ii; for (m = mb, W = W + (mb * ((TWVL / VL) * 14)); m < me; m = m + VL, x = x + (VL * ms), W = W + (TWVL * 14), MAKE_VOLATILE_STRIDE(8, rs)) { V Tl, Tq, Tg, Tr, T5, Tt, Ta, Tu, Ti, Tk, Tj; Ti = LD(&(x[0]), ms, &(x[0])); Tj = LD(&(x[WS(rs, 4)]), ms, &(x[0])); Tk = BYTW(&(W[TWVL * 6]), Tj); Tl = VSUB(Ti, Tk); Tq = VADD(Ti, Tk); { V Td, Tf, Tc, Te; Tc = LD(&(x[WS(rs, 2)]), ms, &(x[0])); Td = BYTW(&(W[TWVL * 2]), Tc); Te = LD(&(x[WS(rs, 6)]), ms, &(x[0])); Tf = BYTW(&(W[TWVL * 10]), Te); Tg = VSUB(Td, Tf); Tr = VADD(Td, Tf); } { V T2, T4, T1, T3; T1 = LD(&(x[WS(rs, 1)]), ms, &(x[WS(rs, 1)])); T2 = BYTW(&(W[0]), T1); T3 = LD(&(x[WS(rs, 5)]), ms, &(x[WS(rs, 1)])); T4 = BYTW(&(W[TWVL * 8]), T3); T5 = VSUB(T2, T4); Tt = VADD(T2, T4); } { V T7, T9, T6, T8; T6 = LD(&(x[WS(rs, 7)]), ms, &(x[WS(rs, 1)])); T7 = BYTW(&(W[TWVL * 12]), T6); T8 = LD(&(x[WS(rs, 3)]), ms, &(x[WS(rs, 1)])); T9 = BYTW(&(W[TWVL * 4]), T8); Ta = VSUB(T7, T9); Tu = VADD(T7, T9); } { V Ts, Tv, Tw, Tx; Ts = VSUB(Tq, Tr); Tv = VBYI(VSUB(Tt, Tu)); ST(&(x[WS(rs, 6)]), VSUB(Ts, Tv), ms, &(x[0])); ST(&(x[WS(rs, 2)]), VADD(Ts, Tv), ms, &(x[0])); Tw = VADD(Tq, Tr); Tx = VADD(Tt, Tu); ST(&(x[WS(rs, 4)]), VSUB(Tw, Tx), ms, &(x[0])); ST(&(x[0]), VADD(Tw, Tx), ms, &(x[0])); { V Th, To, Tn, Tp, Tb, Tm; Tb = VMUL(LDK(KP707106781), VSUB(T5, Ta)); Th = VBYI(VSUB(Tb, Tg)); To = VBYI(VADD(Tg, Tb)); Tm = VMUL(LDK(KP707106781), VADD(T5, Ta)); Tn = VSUB(Tl, Tm); Tp = VADD(Tl, Tm); ST(&(x[WS(rs, 3)]), VADD(Th, Tn), ms, &(x[WS(rs, 1)])); ST(&(x[WS(rs, 7)]), VSUB(Tp, To), ms, &(x[WS(rs, 1)])); ST(&(x[WS(rs, 5)]), VSUB(Tn, Th), ms, &(x[WS(rs, 1)])); ST(&(x[WS(rs, 1)]), VADD(To, Tp), ms, &(x[WS(rs, 1)])); } } } } VLEAVE(); } static const tw_instr twinstr[] = { VTW(0, 1), VTW(0, 2), VTW(0, 3), VTW(0, 4), VTW(0, 5), VTW(0, 6), VTW(0, 7), {TW_NEXT, VL, 0} }; static const ct_desc desc = { 8, XSIMD_STRING("t2bv_8"), twinstr, &GENUS, {33, 16, 0, 0}, 0, 0, 0 }; void XSIMD(codelet_t2bv_8) (planner *p) { X(kdft_dit_register) (p, t2bv_8, &desc); } #endif /* HAVE_FMA */ fftw-3.3.3/dft/simd/common/t2bv_16.c0000644000175000017500000003217712054410556013753 00000000000000/* * Copyright (c) 2003, 2007-11 Matteo Frigo * Copyright (c) 2003, 2007-11 Massachusetts Institute of Technology * * This program is free software; you can redistribute it and/or modify * it under the terms of the GNU General Public License as published by * the Free Software Foundation; either version 2 of the License, or * (at your option) any later version. * * This program is distributed in the hope that it will be useful, * but WITHOUT ANY WARRANTY; without even the implied warranty of * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the * GNU General Public License for more details. * * You should have received a copy of the GNU General Public License * along with this program; if not, write to the Free Software * Foundation, Inc., 51 Franklin Street, Fifth Floor, Boston, MA 02110-1301 USA * */ /* This file was automatically generated --- DO NOT EDIT */ /* Generated on Sun Nov 25 07:39:09 EST 2012 */ #include "codelet-dft.h" #ifdef HAVE_FMA /* Generated by: ../../../genfft/gen_twiddle_c.native -fma -reorder-insns -schedule-for-pipeline -simd -compact -variables 4 -pipeline-latency 8 -n 16 -name t2bv_16 -include t2b.h -sign 1 */ /* * This function contains 87 FP additions, 64 FP multiplications, * (or, 53 additions, 30 multiplications, 34 fused multiply/add), * 61 stack variables, 3 constants, and 32 memory accesses */ #include "t2b.h" static void t2bv_16(R *ri, R *ii, const R *W, stride rs, INT mb, INT me, INT ms) { DVK(KP923879532, +0.923879532511286756128183189396788286822416626); DVK(KP414213562, +0.414213562373095048801688724209698078569671875); DVK(KP707106781, +0.707106781186547524400844362104849039284835938); { INT m; R *x; x = ii; for (m = mb, W = W + (mb * ((TWVL / VL) * 30)); m < me; m = m + VL, x = x + (VL * ms), W = W + (TWVL * 30), MAKE_VOLATILE_STRIDE(16, rs)) { V TO, Ta, TJ, TP, T14, Tq, T1i, T10, T1b, T1l, T13, T1c, TR, Tl, T15; V Tv; { V Tc, TW, T4, T19, T9, TD, TI, Tj, TZ, T1a, Te, Th, Tn, Tr, Tu; V Tp; { V T1, T2, T5, T7; T1 = LD(&(x[0]), ms, &(x[0])); T2 = LD(&(x[WS(rs, 8)]), ms, &(x[0])); T5 = LD(&(x[WS(rs, 4)]), ms, &(x[0])); T7 = LD(&(x[WS(rs, 12)]), ms, &(x[0])); { V Tz, TG, TB, TE; Tz = LD(&(x[WS(rs, 2)]), ms, &(x[0])); TG = LD(&(x[WS(rs, 6)]), ms, &(x[0])); TB = LD(&(x[WS(rs, 10)]), ms, &(x[0])); TE = LD(&(x[WS(rs, 14)]), ms, &(x[0])); { V Ti, TX, TY, Td, Tg, Tm, Tt, To; { V T3, T6, T8, TA, TH, TC, TF, Tb; Tb = LD(&(x[WS(rs, 1)]), ms, &(x[WS(rs, 1)])); T3 = BYTW(&(W[TWVL * 14]), T2); T6 = BYTW(&(W[TWVL * 6]), T5); T8 = BYTW(&(W[TWVL * 22]), T7); TA = BYTW(&(W[TWVL * 2]), Tz); TH = BYTW(&(W[TWVL * 10]), TG); TC = BYTW(&(W[TWVL * 18]), TB); TF = BYTW(&(W[TWVL * 26]), TE); Tc = BYTW(&(W[0]), Tb); TW = VSUB(T1, T3); T4 = VADD(T1, T3); T19 = VSUB(T6, T8); T9 = VADD(T6, T8); Ti = LD(&(x[WS(rs, 13)]), ms, &(x[WS(rs, 1)])); TD = VADD(TA, TC); TX = VSUB(TA, TC); TI = VADD(TF, TH); TY = VSUB(TF, TH); } Td = LD(&(x[WS(rs, 9)]), ms, &(x[WS(rs, 1)])); Tg = LD(&(x[WS(rs, 5)]), ms, &(x[WS(rs, 1)])); Tm = LD(&(x[WS(rs, 15)]), ms, &(x[WS(rs, 1)])); Tj = BYTW(&(W[TWVL * 24]), Ti); Tt = LD(&(x[WS(rs, 11)]), ms, &(x[WS(rs, 1)])); To = LD(&(x[WS(rs, 7)]), ms, &(x[WS(rs, 1)])); TZ = VADD(TX, TY); T1a = VSUB(TX, TY); Te = BYTW(&(W[TWVL * 16]), Td); Th = BYTW(&(W[TWVL * 8]), Tg); Tn = BYTW(&(W[TWVL * 28]), Tm); Tr = LD(&(x[WS(rs, 3)]), ms, &(x[WS(rs, 1)])); Tu = BYTW(&(W[TWVL * 20]), Tt); Tp = BYTW(&(W[TWVL * 12]), To); } } } { V Tf, T11, Tk, T12, Ts; TO = VADD(T4, T9); Ta = VSUB(T4, T9); TJ = VSUB(TD, TI); TP = VADD(TD, TI); Tf = VADD(Tc, Te); T11 = VSUB(Tc, Te); Tk = VADD(Th, Tj); T12 = VSUB(Th, Tj); Ts = BYTW(&(W[TWVL * 4]), Tr); T14 = VSUB(Tn, Tp); Tq = VADD(Tn, Tp); T1i = VFNMS(LDK(KP707106781), TZ, TW); T10 = VFMA(LDK(KP707106781), TZ, TW); T1b = VFMA(LDK(KP707106781), T1a, T19); T1l = VFNMS(LDK(KP707106781), T1a, T19); T13 = VFNMS(LDK(KP414213562), T12, T11); T1c = VFMA(LDK(KP414213562), T11, T12); TR = VADD(Tf, Tk); Tl = VSUB(Tf, Tk); T15 = VSUB(Tu, Ts); Tv = VADD(Ts, Tu); } } { V T1d, T16, TS, Tw, TU, TQ; T1d = VFMA(LDK(KP414213562), T14, T15); T16 = VFNMS(LDK(KP414213562), T15, T14); TS = VADD(Tq, Tv); Tw = VSUB(Tq, Tv); TU = VADD(TO, TP); TQ = VSUB(TO, TP); { V T1e, T1j, T17, T1m; T1e = VSUB(T1c, T1d); T1j = VADD(T1c, T1d); T17 = VADD(T13, T16); T1m = VSUB(T13, T16); { V TV, TT, TK, Tx; TV = VADD(TR, TS); TT = VSUB(TR, TS); TK = VSUB(Tl, Tw); Tx = VADD(Tl, Tw); { V T1h, T1f, T1o, T1k; T1h = VFMA(LDK(KP923879532), T1e, T1b); T1f = VFNMS(LDK(KP923879532), T1e, T1b); T1o = VFMA(LDK(KP923879532), T1j, T1i); T1k = VFNMS(LDK(KP923879532), T1j, T1i); { V T1g, T18, T1p, T1n; T1g = VFMA(LDK(KP923879532), T17, T10); T18 = VFNMS(LDK(KP923879532), T17, T10); T1p = VFNMS(LDK(KP923879532), T1m, T1l); T1n = VFMA(LDK(KP923879532), T1m, T1l); ST(&(x[WS(rs, 8)]), VSUB(TU, TV), ms, &(x[0])); ST(&(x[0]), VADD(TU, TV), ms, &(x[0])); ST(&(x[WS(rs, 4)]), VFMAI(TT, TQ), ms, &(x[0])); ST(&(x[WS(rs, 12)]), VFNMSI(TT, TQ), ms, &(x[0])); { V TN, TL, TM, Ty; TN = VFMA(LDK(KP707106781), TK, TJ); TL = VFNMS(LDK(KP707106781), TK, TJ); TM = VFMA(LDK(KP707106781), Tx, Ta); Ty = VFNMS(LDK(KP707106781), Tx, Ta); ST(&(x[WS(rs, 15)]), VFNMSI(T1h, T1g), ms, &(x[WS(rs, 1)])); ST(&(x[WS(rs, 1)]), VFMAI(T1h, T1g), ms, &(x[WS(rs, 1)])); ST(&(x[WS(rs, 9)]), VFMAI(T1f, T18), ms, &(x[WS(rs, 1)])); ST(&(x[WS(rs, 7)]), VFNMSI(T1f, T18), ms, &(x[WS(rs, 1)])); ST(&(x[WS(rs, 3)]), VFNMSI(T1p, T1o), ms, &(x[WS(rs, 1)])); ST(&(x[WS(rs, 13)]), VFMAI(T1p, T1o), ms, &(x[WS(rs, 1)])); ST(&(x[WS(rs, 11)]), VFNMSI(T1n, T1k), ms, &(x[WS(rs, 1)])); ST(&(x[WS(rs, 5)]), VFMAI(T1n, T1k), ms, &(x[WS(rs, 1)])); ST(&(x[WS(rs, 2)]), VFMAI(TN, TM), ms, &(x[0])); ST(&(x[WS(rs, 14)]), VFNMSI(TN, TM), ms, &(x[0])); ST(&(x[WS(rs, 10)]), VFMAI(TL, Ty), ms, &(x[0])); ST(&(x[WS(rs, 6)]), VFNMSI(TL, Ty), ms, &(x[0])); } } } } } } } } VLEAVE(); } static const tw_instr twinstr[] = { VTW(0, 1), VTW(0, 2), VTW(0, 3), VTW(0, 4), VTW(0, 5), VTW(0, 6), VTW(0, 7), VTW(0, 8), VTW(0, 9), VTW(0, 10), VTW(0, 11), VTW(0, 12), VTW(0, 13), VTW(0, 14), VTW(0, 15), {TW_NEXT, VL, 0} }; static const ct_desc desc = { 16, XSIMD_STRING("t2bv_16"), twinstr, &GENUS, {53, 30, 34, 0}, 0, 0, 0 }; void XSIMD(codelet_t2bv_16) (planner *p) { X(kdft_dit_register) (p, t2bv_16, &desc); } #else /* HAVE_FMA */ /* Generated by: ../../../genfft/gen_twiddle_c.native -simd -compact -variables 4 -pipeline-latency 8 -n 16 -name t2bv_16 -include t2b.h -sign 1 */ /* * This function contains 87 FP additions, 42 FP multiplications, * (or, 83 additions, 38 multiplications, 4 fused multiply/add), * 36 stack variables, 3 constants, and 32 memory accesses */ #include "t2b.h" static void t2bv_16(R *ri, R *ii, const R *W, stride rs, INT mb, INT me, INT ms) { DVK(KP382683432, +0.382683432365089771728459984030398866761344562); DVK(KP923879532, +0.923879532511286756128183189396788286822416626); DVK(KP707106781, +0.707106781186547524400844362104849039284835938); { INT m; R *x; x = ii; for (m = mb, W = W + (mb * ((TWVL / VL) * 30)); m < me; m = m + VL, x = x + (VL * ms), W = W + (TWVL * 30), MAKE_VOLATILE_STRIDE(16, rs)) { V TJ, T1b, TD, T1c, T17, T18, Ty, TK, T10, T11, T12, Tb, TM, T13, T14; V T15, Tm, TN, TG, TI, TH; TG = LD(&(x[0]), ms, &(x[0])); TH = LD(&(x[WS(rs, 8)]), ms, &(x[0])); TI = BYTW(&(W[TWVL * 14]), TH); TJ = VSUB(TG, TI); T1b = VADD(TG, TI); { V TA, TC, Tz, TB; Tz = LD(&(x[WS(rs, 4)]), ms, &(x[0])); TA = BYTW(&(W[TWVL * 6]), Tz); TB = LD(&(x[WS(rs, 12)]), ms, &(x[0])); TC = BYTW(&(W[TWVL * 22]), TB); TD = VSUB(TA, TC); T1c = VADD(TA, TC); } { V Tp, Tw, Tr, Tu, Ts, Tx; { V To, Tv, Tq, Tt; To = LD(&(x[WS(rs, 2)]), ms, &(x[0])); Tp = BYTW(&(W[TWVL * 2]), To); Tv = LD(&(x[WS(rs, 6)]), ms, &(x[0])); Tw = BYTW(&(W[TWVL * 10]), Tv); Tq = LD(&(x[WS(rs, 10)]), ms, &(x[0])); Tr = BYTW(&(W[TWVL * 18]), Tq); Tt = LD(&(x[WS(rs, 14)]), ms, &(x[0])); Tu = BYTW(&(W[TWVL * 26]), Tt); } T17 = VADD(Tp, Tr); T18 = VADD(Tu, Tw); Ts = VSUB(Tp, Tr); Tx = VSUB(Tu, Tw); Ty = VMUL(LDK(KP707106781), VSUB(Ts, Tx)); TK = VMUL(LDK(KP707106781), VADD(Ts, Tx)); } { V T2, T9, T4, T7, T5, Ta; { V T1, T8, T3, T6; T1 = LD(&(x[WS(rs, 1)]), ms, &(x[WS(rs, 1)])); T2 = BYTW(&(W[0]), T1); T8 = LD(&(x[WS(rs, 13)]), ms, &(x[WS(rs, 1)])); T9 = BYTW(&(W[TWVL * 24]), T8); T3 = LD(&(x[WS(rs, 9)]), ms, &(x[WS(rs, 1)])); T4 = BYTW(&(W[TWVL * 16]), T3); T6 = LD(&(x[WS(rs, 5)]), ms, &(x[WS(rs, 1)])); T7 = BYTW(&(W[TWVL * 8]), T6); } T10 = VADD(T2, T4); T11 = VADD(T7, T9); T12 = VSUB(T10, T11); T5 = VSUB(T2, T4); Ta = VSUB(T7, T9); Tb = VFNMS(LDK(KP382683432), Ta, VMUL(LDK(KP923879532), T5)); TM = VFMA(LDK(KP382683432), T5, VMUL(LDK(KP923879532), Ta)); } { V Td, Tk, Tf, Ti, Tg, Tl; { V Tc, Tj, Te, Th; Tc = LD(&(x[WS(rs, 15)]), ms, &(x[WS(rs, 1)])); Td = BYTW(&(W[TWVL * 28]), Tc); Tj = LD(&(x[WS(rs, 11)]), ms, &(x[WS(rs, 1)])); Tk = BYTW(&(W[TWVL * 20]), Tj); Te = LD(&(x[WS(rs, 7)]), ms, &(x[WS(rs, 1)])); Tf = BYTW(&(W[TWVL * 12]), Te); Th = LD(&(x[WS(rs, 3)]), ms, &(x[WS(rs, 1)])); Ti = BYTW(&(W[TWVL * 4]), Th); } T13 = VADD(Td, Tf); T14 = VADD(Ti, Tk); T15 = VSUB(T13, T14); Tg = VSUB(Td, Tf); Tl = VSUB(Ti, Tk); Tm = VFMA(LDK(KP923879532), Tg, VMUL(LDK(KP382683432), Tl)); TN = VFNMS(LDK(KP382683432), Tg, VMUL(LDK(KP923879532), Tl)); } { V T1a, T1g, T1f, T1h; { V T16, T19, T1d, T1e; T16 = VMUL(LDK(KP707106781), VSUB(T12, T15)); T19 = VSUB(T17, T18); T1a = VBYI(VSUB(T16, T19)); T1g = VBYI(VADD(T19, T16)); T1d = VSUB(T1b, T1c); T1e = VMUL(LDK(KP707106781), VADD(T12, T15)); T1f = VSUB(T1d, T1e); T1h = VADD(T1d, T1e); } ST(&(x[WS(rs, 6)]), VADD(T1a, T1f), ms, &(x[0])); ST(&(x[WS(rs, 14)]), VSUB(T1h, T1g), ms, &(x[0])); ST(&(x[WS(rs, 10)]), VSUB(T1f, T1a), ms, &(x[0])); ST(&(x[WS(rs, 2)]), VADD(T1g, T1h), ms, &(x[0])); } { V T1k, T1o, T1n, T1p; { V T1i, T1j, T1l, T1m; T1i = VADD(T1b, T1c); T1j = VADD(T17, T18); T1k = VSUB(T1i, T1j); T1o = VADD(T1i, T1j); T1l = VADD(T10, T11); T1m = VADD(T13, T14); T1n = VBYI(VSUB(T1l, T1m)); T1p = VADD(T1l, T1m); } ST(&(x[WS(rs, 12)]), VSUB(T1k, T1n), ms, &(x[0])); ST(&(x[0]), VADD(T1o, T1p), ms, &(x[0])); ST(&(x[WS(rs, 4)]), VADD(T1k, T1n), ms, &(x[0])); ST(&(x[WS(rs, 8)]), VSUB(T1o, T1p), ms, &(x[0])); } { V TF, TQ, TP, TR; { V Tn, TE, TL, TO; Tn = VSUB(Tb, Tm); TE = VSUB(Ty, TD); TF = VBYI(VSUB(Tn, TE)); TQ = VBYI(VADD(TE, Tn)); TL = VSUB(TJ, TK); TO = VSUB(TM, TN); TP = VSUB(TL, TO); TR = VADD(TL, TO); } ST(&(x[WS(rs, 5)]), VADD(TF, TP), ms, &(x[WS(rs, 1)])); ST(&(x[WS(rs, 13)]), VSUB(TR, TQ), ms, &(x[WS(rs, 1)])); ST(&(x[WS(rs, 11)]), VSUB(TP, TF), ms, &(x[WS(rs, 1)])); ST(&(x[WS(rs, 3)]), VADD(TQ, TR), ms, &(x[WS(rs, 1)])); } { V TU, TY, TX, TZ; { V TS, TT, TV, TW; TS = VADD(TJ, TK); TT = VADD(Tb, Tm); TU = VADD(TS, TT); TY = VSUB(TS, TT); TV = VADD(TD, Ty); TW = VADD(TM, TN); TX = VBYI(VADD(TV, TW)); TZ = VBYI(VSUB(TW, TV)); } ST(&(x[WS(rs, 15)]), VSUB(TU, TX), ms, &(x[WS(rs, 1)])); ST(&(x[WS(rs, 7)]), VADD(TY, TZ), ms, &(x[WS(rs, 1)])); ST(&(x[WS(rs, 1)]), VADD(TU, TX), ms, &(x[WS(rs, 1)])); ST(&(x[WS(rs, 9)]), VSUB(TY, TZ), ms, &(x[WS(rs, 1)])); } } } VLEAVE(); } static const tw_instr twinstr[] = { VTW(0, 1), VTW(0, 2), VTW(0, 3), VTW(0, 4), VTW(0, 5), VTW(0, 6), VTW(0, 7), VTW(0, 8), VTW(0, 9), VTW(0, 10), VTW(0, 11), VTW(0, 12), VTW(0, 13), VTW(0, 14), VTW(0, 15), {TW_NEXT, VL, 0} }; static const ct_desc desc = { 16, XSIMD_STRING("t2bv_16"), twinstr, &GENUS, {83, 38, 4, 0}, 0, 0, 0 }; void XSIMD(codelet_t2bv_16) (planner *p) { X(kdft_dit_register) (p, t2bv_16, &desc); } #endif /* HAVE_FMA */ fftw-3.3.3/dft/simd/common/t2bv_32.c0000644000175000017500000007043712054410562013747 00000000000000/* * Copyright (c) 2003, 2007-11 Matteo Frigo * Copyright (c) 2003, 2007-11 Massachusetts Institute of Technology * * This program is free software; you can redistribute it and/or modify * it under the terms of the GNU General Public License as published by * the Free Software Foundation; either version 2 of the License, or * (at your option) any later version. * * This program is distributed in the hope that it will be useful, * but WITHOUT ANY WARRANTY; without even the implied warranty of * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the * GNU General Public License for more details. * * You should have received a copy of the GNU General Public License * along with this program; if not, write to the Free Software * Foundation, Inc., 51 Franklin Street, Fifth Floor, Boston, MA 02110-1301 USA * */ /* This file was automatically generated --- DO NOT EDIT */ /* Generated on Sun Nov 25 07:39:10 EST 2012 */ #include "codelet-dft.h" #ifdef HAVE_FMA /* Generated by: ../../../genfft/gen_twiddle_c.native -fma -reorder-insns -schedule-for-pipeline -simd -compact -variables 4 -pipeline-latency 8 -n 32 -name t2bv_32 -include t2b.h -sign 1 */ /* * This function contains 217 FP additions, 160 FP multiplications, * (or, 119 additions, 62 multiplications, 98 fused multiply/add), * 104 stack variables, 7 constants, and 64 memory accesses */ #include "t2b.h" static void t2bv_32(R *ri, R *ii, const R *W, stride rs, INT mb, INT me, INT ms) { DVK(KP831469612, +0.831469612302545237078788377617905756738560812); DVK(KP980785280, +0.980785280403230449126182236134239036973933731); DVK(KP668178637, +0.668178637919298919997757686523080761552472251); DVK(KP198912367, +0.198912367379658006911597622644676228597850501); DVK(KP923879532, +0.923879532511286756128183189396788286822416626); DVK(KP707106781, +0.707106781186547524400844362104849039284835938); DVK(KP414213562, +0.414213562373095048801688724209698078569671875); { INT m; R *x; x = ii; for (m = mb, W = W + (mb * ((TWVL / VL) * 62)); m < me; m = m + VL, x = x + (VL * ms), W = W + (TWVL * 62), MAKE_VOLATILE_STRIDE(32, rs)) { V T26, T25, T2a, T2i, T24, T2c, T2g, T2k, T2h, T27; { V T4, T1z, T2o, T32, T2r, T3f, Tf, T1A, T34, T2O, T1D, TC, T33, T2L, T1C; V Tr, T2C, T3a, T2F, T3b, T1r, T21, T1k, T20, TQ, TM, TS, TL, T2t, TJ; V T10, T2u; { V Tt, T9, T2p, Te, T2q, TA, Tu, Tx; { V T1, T1x, T2, T1v; T1 = LD(&(x[0]), ms, &(x[0])); T1x = LD(&(x[WS(rs, 24)]), ms, &(x[0])); T2 = LD(&(x[WS(rs, 16)]), ms, &(x[0])); T1v = LD(&(x[WS(rs, 8)]), ms, &(x[0])); { V T5, Tc, T7, Ta, T2m, T2n; T5 = LD(&(x[WS(rs, 4)]), ms, &(x[0])); Tc = LD(&(x[WS(rs, 12)]), ms, &(x[0])); T7 = LD(&(x[WS(rs, 20)]), ms, &(x[0])); Ta = LD(&(x[WS(rs, 28)]), ms, &(x[0])); { V T1y, T3, T1w, T6, Td, T8, Tb, Ts, Tz; Ts = LD(&(x[WS(rs, 30)]), ms, &(x[0])); T1y = BYTW(&(W[TWVL * 46]), T1x); T3 = BYTW(&(W[TWVL * 30]), T2); T1w = BYTW(&(W[TWVL * 14]), T1v); T6 = BYTW(&(W[TWVL * 6]), T5); Td = BYTW(&(W[TWVL * 22]), Tc); T8 = BYTW(&(W[TWVL * 38]), T7); Tb = BYTW(&(W[TWVL * 54]), Ta); Tt = BYTW(&(W[TWVL * 58]), Ts); Tz = LD(&(x[WS(rs, 6)]), ms, &(x[0])); T4 = VSUB(T1, T3); T2m = VADD(T1, T3); T1z = VSUB(T1w, T1y); T2n = VADD(T1w, T1y); T9 = VSUB(T6, T8); T2p = VADD(T6, T8); Te = VSUB(Tb, Td); T2q = VADD(Tb, Td); TA = BYTW(&(W[TWVL * 10]), Tz); } Tu = LD(&(x[WS(rs, 14)]), ms, &(x[0])); T2o = VADD(T2m, T2n); T32 = VSUB(T2m, T2n); Tx = LD(&(x[WS(rs, 22)]), ms, &(x[0])); } } { V Tv, To, Ty, Ti, Tj, Tm, Th; Th = LD(&(x[WS(rs, 2)]), ms, &(x[0])); T2r = VADD(T2p, T2q); T3f = VSUB(T2p, T2q); Tf = VADD(T9, Te); T1A = VSUB(T9, Te); Tv = BYTW(&(W[TWVL * 26]), Tu); To = LD(&(x[WS(rs, 26)]), ms, &(x[0])); Ty = BYTW(&(W[TWVL * 42]), Tx); Ti = BYTW(&(W[TWVL * 2]), Th); Tj = LD(&(x[WS(rs, 18)]), ms, &(x[0])); Tm = LD(&(x[WS(rs, 10)]), ms, &(x[0])); { V T1f, T1h, T1a, T1c, T18, T2A, T2B, T1p; { V T15, T17, T1o, T1m; { V Tw, T2M, Tp, T2N, TB, Tk, Tn, T1n, T14, T16; T14 = LD(&(x[WS(rs, 31)]), ms, &(x[WS(rs, 1)])); T16 = LD(&(x[WS(rs, 15)]), ms, &(x[WS(rs, 1)])); Tw = VSUB(Tt, Tv); T2M = VADD(Tt, Tv); Tp = BYTW(&(W[TWVL * 50]), To); T2N = VADD(TA, Ty); TB = VSUB(Ty, TA); Tk = BYTW(&(W[TWVL * 34]), Tj); Tn = BYTW(&(W[TWVL * 18]), Tm); T15 = BYTW(&(W[TWVL * 60]), T14); T17 = BYTW(&(W[TWVL * 28]), T16); T1n = LD(&(x[WS(rs, 7)]), ms, &(x[WS(rs, 1)])); { V T2J, Tl, T2K, Tq, T1l; T1l = LD(&(x[WS(rs, 23)]), ms, &(x[WS(rs, 1)])); T34 = VSUB(T2M, T2N); T2O = VADD(T2M, T2N); T1D = VFMA(LDK(KP414213562), Tw, TB); TC = VFNMS(LDK(KP414213562), TB, Tw); T2J = VADD(Ti, Tk); Tl = VSUB(Ti, Tk); T2K = VADD(Tn, Tp); Tq = VSUB(Tn, Tp); T1o = BYTW(&(W[TWVL * 12]), T1n); T1m = BYTW(&(W[TWVL * 44]), T1l); { V T1e, T1g, T19, T1b; T1e = LD(&(x[WS(rs, 27)]), ms, &(x[WS(rs, 1)])); T1g = LD(&(x[WS(rs, 11)]), ms, &(x[WS(rs, 1)])); T19 = LD(&(x[WS(rs, 3)]), ms, &(x[WS(rs, 1)])); T1b = LD(&(x[WS(rs, 19)]), ms, &(x[WS(rs, 1)])); T33 = VSUB(T2J, T2K); T2L = VADD(T2J, T2K); T1C = VFMA(LDK(KP414213562), Tl, Tq); Tr = VFNMS(LDK(KP414213562), Tq, Tl); T1f = BYTW(&(W[TWVL * 52]), T1e); T1h = BYTW(&(W[TWVL * 20]), T1g); T1a = BYTW(&(W[TWVL * 4]), T19); T1c = BYTW(&(W[TWVL * 36]), T1b); } } } T18 = VSUB(T15, T17); T2A = VADD(T15, T17); T2B = VADD(T1o, T1m); T1p = VSUB(T1m, T1o); } { V TG, TI, TZ, TX; { V T1i, T2E, T1d, T2D, TH, TY, TF; TF = LD(&(x[WS(rs, 1)]), ms, &(x[WS(rs, 1)])); T1i = VSUB(T1f, T1h); T2E = VADD(T1f, T1h); T1d = VSUB(T1a, T1c); T2D = VADD(T1a, T1c); TH = LD(&(x[WS(rs, 17)]), ms, &(x[WS(rs, 1)])); TY = LD(&(x[WS(rs, 25)]), ms, &(x[WS(rs, 1)])); T2C = VADD(T2A, T2B); T3a = VSUB(T2A, T2B); TG = BYTW(&(W[0]), TF); { V TW, T1j, T1q, TP, TR, TK; TW = LD(&(x[WS(rs, 9)]), ms, &(x[WS(rs, 1)])); T2F = VADD(T2D, T2E); T3b = VSUB(T2E, T2D); T1j = VADD(T1d, T1i); T1q = VSUB(T1i, T1d); TI = BYTW(&(W[TWVL * 32]), TH); TZ = BYTW(&(W[TWVL * 48]), TY); TP = LD(&(x[WS(rs, 29)]), ms, &(x[WS(rs, 1)])); TX = BYTW(&(W[TWVL * 16]), TW); TR = LD(&(x[WS(rs, 13)]), ms, &(x[WS(rs, 1)])); TK = LD(&(x[WS(rs, 5)]), ms, &(x[WS(rs, 1)])); T1r = VFMA(LDK(KP707106781), T1q, T1p); T21 = VFNMS(LDK(KP707106781), T1q, T1p); T1k = VFMA(LDK(KP707106781), T1j, T18); T20 = VFNMS(LDK(KP707106781), T1j, T18); TQ = BYTW(&(W[TWVL * 56]), TP); TM = LD(&(x[WS(rs, 21)]), ms, &(x[WS(rs, 1)])); TS = BYTW(&(W[TWVL * 24]), TR); TL = BYTW(&(W[TWVL * 8]), TK); } } T2t = VADD(TG, TI); TJ = VSUB(TG, TI); T10 = VSUB(TX, TZ); T2u = VADD(TX, TZ); } } } } { V T2s, TT, T2x, T2P, T2Y, T2G, T37, T2v, T2w, TO, T2W, T30, T2U, TN, T2V; T2s = VSUB(T2o, T2r); T2U = VADD(T2o, T2r); TN = BYTW(&(W[TWVL * 40]), TM); TT = VSUB(TQ, TS); T2x = VADD(TQ, TS); T2P = VSUB(T2L, T2O); T2V = VADD(T2L, T2O); T2Y = VADD(T2C, T2F); T2G = VSUB(T2C, T2F); T37 = VSUB(T2t, T2u); T2v = VADD(T2t, T2u); T2w = VADD(TL, TN); TO = VSUB(TL, TN); T2W = VSUB(T2U, T2V); T30 = VADD(T2U, T2V); { V T1Y, T12, T1X, TV, T3n, T3t, T3m, T3q; { V T3o, T36, T3r, T3h, T3k, T3p, T3d, T3s, T2H, T2Q, T2Z, T31; { V T35, T3g, T38, T2y, T11, TU, T3c, T3j; T35 = VADD(T33, T34); T3g = VSUB(T33, T34); T38 = VSUB(T2w, T2x); T2y = VADD(T2w, T2x); T11 = VSUB(TO, TT); TU = VADD(TO, TT); T3c = VFNMS(LDK(KP414213562), T3b, T3a); T3j = VFMA(LDK(KP414213562), T3a, T3b); T3o = VFNMS(LDK(KP707106781), T35, T32); T36 = VFMA(LDK(KP707106781), T35, T32); T3r = VFNMS(LDK(KP707106781), T3g, T3f); T3h = VFMA(LDK(KP707106781), T3g, T3f); { V T3i, T39, T2z, T2X; T3i = VFMA(LDK(KP414213562), T37, T38); T39 = VFNMS(LDK(KP414213562), T38, T37); T2z = VSUB(T2v, T2y); T2X = VADD(T2v, T2y); T1Y = VFNMS(LDK(KP707106781), T11, T10); T12 = VFMA(LDK(KP707106781), T11, T10); T1X = VFNMS(LDK(KP707106781), TU, TJ); TV = VFMA(LDK(KP707106781), TU, TJ); T3k = VSUB(T3i, T3j); T3p = VADD(T3i, T3j); T3d = VADD(T39, T3c); T3s = VSUB(T39, T3c); T2H = VADD(T2z, T2G); T2Q = VSUB(T2z, T2G); T2Z = VSUB(T2X, T2Y); T31 = VADD(T2X, T2Y); } } { V T3v, T3u, T3l, T3e; T3l = VFNMS(LDK(KP923879532), T3k, T3h); T3n = VFMA(LDK(KP923879532), T3k, T3h); T3t = VFMA(LDK(KP923879532), T3s, T3r); T3v = VFNMS(LDK(KP923879532), T3s, T3r); T3e = VFNMS(LDK(KP923879532), T3d, T36); T3m = VFMA(LDK(KP923879532), T3d, T36); { V T2R, T2T, T2I, T2S; T2R = VFNMS(LDK(KP707106781), T2Q, T2P); T2T = VFMA(LDK(KP707106781), T2Q, T2P); T2I = VFNMS(LDK(KP707106781), T2H, T2s); T2S = VFMA(LDK(KP707106781), T2H, T2s); ST(&(x[WS(rs, 16)]), VSUB(T30, T31), ms, &(x[0])); ST(&(x[0]), VADD(T30, T31), ms, &(x[0])); ST(&(x[WS(rs, 8)]), VFMAI(T2Z, T2W), ms, &(x[0])); ST(&(x[WS(rs, 24)]), VFNMSI(T2Z, T2W), ms, &(x[0])); T3q = VFNMS(LDK(KP923879532), T3p, T3o); T3u = VFMA(LDK(KP923879532), T3p, T3o); ST(&(x[WS(rs, 18)]), VFMAI(T3l, T3e), ms, &(x[0])); ST(&(x[WS(rs, 14)]), VFNMSI(T3l, T3e), ms, &(x[0])); ST(&(x[WS(rs, 28)]), VFNMSI(T2T, T2S), ms, &(x[0])); ST(&(x[WS(rs, 4)]), VFMAI(T2T, T2S), ms, &(x[0])); ST(&(x[WS(rs, 20)]), VFMAI(T2R, T2I), ms, &(x[0])); ST(&(x[WS(rs, 12)]), VFNMSI(T2R, T2I), ms, &(x[0])); } ST(&(x[WS(rs, 26)]), VFMAI(T3v, T3u), ms, &(x[0])); ST(&(x[WS(rs, 6)]), VFNMSI(T3v, T3u), ms, &(x[0])); } } { V T1U, T13, T1s, TE, T1M, T1I, T1N, T1B, T1V, T1E; { V Tg, TD, T1G, T1H; Tg = VFMA(LDK(KP707106781), Tf, T4); T1U = VFNMS(LDK(KP707106781), Tf, T4); T26 = VSUB(Tr, TC); TD = VADD(Tr, TC); T1G = VFMA(LDK(KP198912367), TV, T12); T13 = VFNMS(LDK(KP198912367), T12, TV); T1s = VFNMS(LDK(KP198912367), T1r, T1k); T1H = VFMA(LDK(KP198912367), T1k, T1r); ST(&(x[WS(rs, 2)]), VFMAI(T3n, T3m), ms, &(x[0])); ST(&(x[WS(rs, 30)]), VFNMSI(T3n, T3m), ms, &(x[0])); ST(&(x[WS(rs, 22)]), VFNMSI(T3t, T3q), ms, &(x[0])); ST(&(x[WS(rs, 10)]), VFMAI(T3t, T3q), ms, &(x[0])); TE = VFMA(LDK(KP923879532), TD, Tg); T1M = VFNMS(LDK(KP923879532), TD, Tg); T1I = VSUB(T1G, T1H); T1N = VADD(T1G, T1H); T1B = VFMA(LDK(KP707106781), T1A, T1z); T25 = VFNMS(LDK(KP707106781), T1A, T1z); T1V = VADD(T1C, T1D); T1E = VSUB(T1C, T1D); } { V T1W, T2e, T2f, T23; { V T28, T1Z, T1S, T1O, T1t, T1Q, T1F, T1P, T22, T29; T28 = VFNMS(LDK(KP668178637), T1X, T1Y); T1Z = VFMA(LDK(KP668178637), T1Y, T1X); T1S = VFMA(LDK(KP980785280), T1N, T1M); T1O = VFNMS(LDK(KP980785280), T1N, T1M); T1t = VADD(T13, T1s); T1Q = VSUB(T13, T1s); T1F = VFMA(LDK(KP923879532), T1E, T1B); T1P = VFNMS(LDK(KP923879532), T1E, T1B); T1W = VFMA(LDK(KP923879532), T1V, T1U); T2e = VFNMS(LDK(KP923879532), T1V, T1U); T22 = VFMA(LDK(KP668178637), T21, T20); T29 = VFNMS(LDK(KP668178637), T20, T21); { V T1K, T1u, T1R, T1T, T1L, T1J; T1K = VFMA(LDK(KP980785280), T1t, TE); T1u = VFNMS(LDK(KP980785280), T1t, TE); T1R = VFMA(LDK(KP980785280), T1Q, T1P); T1T = VFNMS(LDK(KP980785280), T1Q, T1P); T1L = VFMA(LDK(KP980785280), T1I, T1F); T1J = VFNMS(LDK(KP980785280), T1I, T1F); T2f = VADD(T28, T29); T2a = VSUB(T28, T29); T23 = VADD(T1Z, T22); T2i = VSUB(T1Z, T22); ST(&(x[WS(rs, 23)]), VFNMSI(T1R, T1O), ms, &(x[WS(rs, 1)])); ST(&(x[WS(rs, 9)]), VFMAI(T1R, T1O), ms, &(x[WS(rs, 1)])); ST(&(x[WS(rs, 25)]), VFMAI(T1T, T1S), ms, &(x[WS(rs, 1)])); ST(&(x[WS(rs, 7)]), VFNMSI(T1T, T1S), ms, &(x[WS(rs, 1)])); ST(&(x[WS(rs, 1)]), VFMAI(T1L, T1K), ms, &(x[WS(rs, 1)])); ST(&(x[WS(rs, 31)]), VFNMSI(T1L, T1K), ms, &(x[WS(rs, 1)])); ST(&(x[WS(rs, 17)]), VFMAI(T1J, T1u), ms, &(x[WS(rs, 1)])); ST(&(x[WS(rs, 15)]), VFNMSI(T1J, T1u), ms, &(x[WS(rs, 1)])); } } T24 = VFNMS(LDK(KP831469612), T23, T1W); T2c = VFMA(LDK(KP831469612), T23, T1W); T2g = VFMA(LDK(KP831469612), T2f, T2e); T2k = VFNMS(LDK(KP831469612), T2f, T2e); } } } } } T2h = VFMA(LDK(KP923879532), T26, T25); T27 = VFNMS(LDK(KP923879532), T26, T25); { V T2j, T2l, T2d, T2b; T2j = VFNMS(LDK(KP831469612), T2i, T2h); T2l = VFMA(LDK(KP831469612), T2i, T2h); T2d = VFMA(LDK(KP831469612), T2a, T27); T2b = VFNMS(LDK(KP831469612), T2a, T27); ST(&(x[WS(rs, 21)]), VFMAI(T2j, T2g), ms, &(x[WS(rs, 1)])); ST(&(x[WS(rs, 11)]), VFNMSI(T2j, T2g), ms, &(x[WS(rs, 1)])); ST(&(x[WS(rs, 27)]), VFNMSI(T2l, T2k), ms, &(x[WS(rs, 1)])); ST(&(x[WS(rs, 5)]), VFMAI(T2l, T2k), ms, &(x[WS(rs, 1)])); ST(&(x[WS(rs, 29)]), VFMAI(T2d, T2c), ms, &(x[WS(rs, 1)])); ST(&(x[WS(rs, 3)]), VFNMSI(T2d, T2c), ms, &(x[WS(rs, 1)])); ST(&(x[WS(rs, 13)]), VFMAI(T2b, T24), ms, &(x[WS(rs, 1)])); ST(&(x[WS(rs, 19)]), VFNMSI(T2b, T24), ms, &(x[WS(rs, 1)])); } } } VLEAVE(); } static const tw_instr twinstr[] = { VTW(0, 1), VTW(0, 2), VTW(0, 3), VTW(0, 4), VTW(0, 5), VTW(0, 6), VTW(0, 7), VTW(0, 8), VTW(0, 9), VTW(0, 10), VTW(0, 11), VTW(0, 12), VTW(0, 13), VTW(0, 14), VTW(0, 15), VTW(0, 16), VTW(0, 17), VTW(0, 18), VTW(0, 19), VTW(0, 20), VTW(0, 21), VTW(0, 22), VTW(0, 23), VTW(0, 24), VTW(0, 25), VTW(0, 26), VTW(0, 27), VTW(0, 28), VTW(0, 29), VTW(0, 30), VTW(0, 31), {TW_NEXT, VL, 0} }; static const ct_desc desc = { 32, XSIMD_STRING("t2bv_32"), twinstr, &GENUS, {119, 62, 98, 0}, 0, 0, 0 }; void XSIMD(codelet_t2bv_32) (planner *p) { X(kdft_dit_register) (p, t2bv_32, &desc); } #else /* HAVE_FMA */ /* Generated by: ../../../genfft/gen_twiddle_c.native -simd -compact -variables 4 -pipeline-latency 8 -n 32 -name t2bv_32 -include t2b.h -sign 1 */ /* * This function contains 217 FP additions, 104 FP multiplications, * (or, 201 additions, 88 multiplications, 16 fused multiply/add), * 59 stack variables, 7 constants, and 64 memory accesses */ #include "t2b.h" static void t2bv_32(R *ri, R *ii, const R *W, stride rs, INT mb, INT me, INT ms) { DVK(KP195090322, +0.195090322016128267848284868477022240927691618); DVK(KP980785280, +0.980785280403230449126182236134239036973933731); DVK(KP555570233, +0.555570233019602224742830813948532874374937191); DVK(KP831469612, +0.831469612302545237078788377617905756738560812); DVK(KP382683432, +0.382683432365089771728459984030398866761344562); DVK(KP923879532, +0.923879532511286756128183189396788286822416626); DVK(KP707106781, +0.707106781186547524400844362104849039284835938); { INT m; R *x; x = ii; for (m = mb, W = W + (mb * ((TWVL / VL) * 62)); m < me; m = m + VL, x = x + (VL * ms), W = W + (TWVL * 62), MAKE_VOLATILE_STRIDE(32, rs)) { V T4, T1D, T2P, T3h, Tf, T1y, T2K, T3i, TC, T1w, T2G, T3e, Tr, T1v, T2D; V T3d, T1k, T20, T2y, T3a, T1r, T21, T2v, T39, TV, T1X, T2r, T37, T12, T1Y; V T2o, T36; { V T1, T1C, T3, T1A, T1B, T2, T1z, T2N, T2O; T1 = LD(&(x[0]), ms, &(x[0])); T1B = LD(&(x[WS(rs, 24)]), ms, &(x[0])); T1C = BYTW(&(W[TWVL * 46]), T1B); T2 = LD(&(x[WS(rs, 16)]), ms, &(x[0])); T3 = BYTW(&(W[TWVL * 30]), T2); T1z = LD(&(x[WS(rs, 8)]), ms, &(x[0])); T1A = BYTW(&(W[TWVL * 14]), T1z); T4 = VSUB(T1, T3); T1D = VSUB(T1A, T1C); T2N = VADD(T1, T3); T2O = VADD(T1A, T1C); T2P = VSUB(T2N, T2O); T3h = VADD(T2N, T2O); } { V T6, Td, T8, Tb; { V T5, Tc, T7, Ta; T5 = LD(&(x[WS(rs, 4)]), ms, &(x[0])); T6 = BYTW(&(W[TWVL * 6]), T5); Tc = LD(&(x[WS(rs, 12)]), ms, &(x[0])); Td = BYTW(&(W[TWVL * 22]), Tc); T7 = LD(&(x[WS(rs, 20)]), ms, &(x[0])); T8 = BYTW(&(W[TWVL * 38]), T7); Ta = LD(&(x[WS(rs, 28)]), ms, &(x[0])); Tb = BYTW(&(W[TWVL * 54]), Ta); } { V T9, Te, T2I, T2J; T9 = VSUB(T6, T8); Te = VSUB(Tb, Td); Tf = VMUL(LDK(KP707106781), VADD(T9, Te)); T1y = VMUL(LDK(KP707106781), VSUB(T9, Te)); T2I = VADD(T6, T8); T2J = VADD(Tb, Td); T2K = VSUB(T2I, T2J); T3i = VADD(T2I, T2J); } } { V Tt, TA, Tv, Ty; { V Ts, Tz, Tu, Tx; Ts = LD(&(x[WS(rs, 6)]), ms, &(x[0])); Tt = BYTW(&(W[TWVL * 10]), Ts); Tz = LD(&(x[WS(rs, 14)]), ms, &(x[0])); TA = BYTW(&(W[TWVL * 26]), Tz); Tu = LD(&(x[WS(rs, 22)]), ms, &(x[0])); Tv = BYTW(&(W[TWVL * 42]), Tu); Tx = LD(&(x[WS(rs, 30)]), ms, &(x[0])); Ty = BYTW(&(W[TWVL * 58]), Tx); } { V Tw, TB, T2E, T2F; Tw = VSUB(Tt, Tv); TB = VSUB(Ty, TA); TC = VFNMS(LDK(KP382683432), TB, VMUL(LDK(KP923879532), Tw)); T1w = VFMA(LDK(KP923879532), TB, VMUL(LDK(KP382683432), Tw)); T2E = VADD(Ty, TA); T2F = VADD(Tt, Tv); T2G = VSUB(T2E, T2F); T3e = VADD(T2E, T2F); } } { V Ti, Tp, Tk, Tn; { V Th, To, Tj, Tm; Th = LD(&(x[WS(rs, 2)]), ms, &(x[0])); Ti = BYTW(&(W[TWVL * 2]), Th); To = LD(&(x[WS(rs, 26)]), ms, &(x[0])); Tp = BYTW(&(W[TWVL * 50]), To); Tj = LD(&(x[WS(rs, 18)]), ms, &(x[0])); Tk = BYTW(&(W[TWVL * 34]), Tj); Tm = LD(&(x[WS(rs, 10)]), ms, &(x[0])); Tn = BYTW(&(W[TWVL * 18]), Tm); } { V Tl, Tq, T2B, T2C; Tl = VSUB(Ti, Tk); Tq = VSUB(Tn, Tp); Tr = VFMA(LDK(KP382683432), Tl, VMUL(LDK(KP923879532), Tq)); T1v = VFNMS(LDK(KP382683432), Tq, VMUL(LDK(KP923879532), Tl)); T2B = VADD(Ti, Tk); T2C = VADD(Tn, Tp); T2D = VSUB(T2B, T2C); T3d = VADD(T2B, T2C); } } { V T1g, T1i, T1o, T1m, T1a, T1c, T1d, T15, T17, T18; { V T1f, T1h, T1n, T1l; T1f = LD(&(x[WS(rs, 7)]), ms, &(x[WS(rs, 1)])); T1g = BYTW(&(W[TWVL * 12]), T1f); T1h = LD(&(x[WS(rs, 23)]), ms, &(x[WS(rs, 1)])); T1i = BYTW(&(W[TWVL * 44]), T1h); T1n = LD(&(x[WS(rs, 15)]), ms, &(x[WS(rs, 1)])); T1o = BYTW(&(W[TWVL * 28]), T1n); T1l = LD(&(x[WS(rs, 31)]), ms, &(x[WS(rs, 1)])); T1m = BYTW(&(W[TWVL * 60]), T1l); { V T19, T1b, T14, T16; T19 = LD(&(x[WS(rs, 27)]), ms, &(x[WS(rs, 1)])); T1a = BYTW(&(W[TWVL * 52]), T19); T1b = LD(&(x[WS(rs, 11)]), ms, &(x[WS(rs, 1)])); T1c = BYTW(&(W[TWVL * 20]), T1b); T1d = VSUB(T1a, T1c); T14 = LD(&(x[WS(rs, 3)]), ms, &(x[WS(rs, 1)])); T15 = BYTW(&(W[TWVL * 4]), T14); T16 = LD(&(x[WS(rs, 19)]), ms, &(x[WS(rs, 1)])); T17 = BYTW(&(W[TWVL * 36]), T16); T18 = VSUB(T15, T17); } } { V T1e, T1j, T2w, T2x; T1e = VMUL(LDK(KP707106781), VSUB(T18, T1d)); T1j = VSUB(T1g, T1i); T1k = VSUB(T1e, T1j); T20 = VADD(T1j, T1e); T2w = VADD(T15, T17); T2x = VADD(T1a, T1c); T2y = VSUB(T2w, T2x); T3a = VADD(T2w, T2x); } { V T1p, T1q, T2t, T2u; T1p = VSUB(T1m, T1o); T1q = VMUL(LDK(KP707106781), VADD(T18, T1d)); T1r = VSUB(T1p, T1q); T21 = VADD(T1p, T1q); T2t = VADD(T1m, T1o); T2u = VADD(T1g, T1i); T2v = VSUB(T2t, T2u); T39 = VADD(T2t, T2u); } } { V TR, TT, TZ, TX, TL, TN, TO, TG, TI, TJ; { V TQ, TS, TY, TW; TQ = LD(&(x[WS(rs, 9)]), ms, &(x[WS(rs, 1)])); TR = BYTW(&(W[TWVL * 16]), TQ); TS = LD(&(x[WS(rs, 25)]), ms, &(x[WS(rs, 1)])); TT = BYTW(&(W[TWVL * 48]), TS); TY = LD(&(x[WS(rs, 17)]), ms, &(x[WS(rs, 1)])); TZ = BYTW(&(W[TWVL * 32]), TY); TW = LD(&(x[WS(rs, 1)]), ms, &(x[WS(rs, 1)])); TX = BYTW(&(W[0]), TW); { V TK, TM, TF, TH; TK = LD(&(x[WS(rs, 29)]), ms, &(x[WS(rs, 1)])); TL = BYTW(&(W[TWVL * 56]), TK); TM = LD(&(x[WS(rs, 13)]), ms, &(x[WS(rs, 1)])); TN = BYTW(&(W[TWVL * 24]), TM); TO = VSUB(TL, TN); TF = LD(&(x[WS(rs, 5)]), ms, &(x[WS(rs, 1)])); TG = BYTW(&(W[TWVL * 8]), TF); TH = LD(&(x[WS(rs, 21)]), ms, &(x[WS(rs, 1)])); TI = BYTW(&(W[TWVL * 40]), TH); TJ = VSUB(TG, TI); } } { V TP, TU, T2p, T2q; TP = VMUL(LDK(KP707106781), VSUB(TJ, TO)); TU = VSUB(TR, TT); TV = VSUB(TP, TU); T1X = VADD(TU, TP); T2p = VADD(TG, TI); T2q = VADD(TL, TN); T2r = VSUB(T2p, T2q); T37 = VADD(T2p, T2q); } { V T10, T11, T2m, T2n; T10 = VSUB(TX, TZ); T11 = VMUL(LDK(KP707106781), VADD(TJ, TO)); T12 = VSUB(T10, T11); T1Y = VADD(T10, T11); T2m = VADD(TX, TZ); T2n = VADD(TR, TT); T2o = VSUB(T2m, T2n); T36 = VADD(T2m, T2n); } } { V T3q, T3u, T3t, T3v; { V T3o, T3p, T3r, T3s; T3o = VADD(T3h, T3i); T3p = VADD(T3d, T3e); T3q = VSUB(T3o, T3p); T3u = VADD(T3o, T3p); T3r = VADD(T36, T37); T3s = VADD(T39, T3a); T3t = VBYI(VSUB(T3r, T3s)); T3v = VADD(T3r, T3s); } ST(&(x[WS(rs, 24)]), VSUB(T3q, T3t), ms, &(x[0])); ST(&(x[0]), VADD(T3u, T3v), ms, &(x[0])); ST(&(x[WS(rs, 8)]), VADD(T3q, T3t), ms, &(x[0])); ST(&(x[WS(rs, 16)]), VSUB(T3u, T3v), ms, &(x[0])); } { V T3f, T3j, T3c, T3k, T38, T3b; T3f = VSUB(T3d, T3e); T3j = VSUB(T3h, T3i); T38 = VSUB(T36, T37); T3b = VSUB(T39, T3a); T3c = VMUL(LDK(KP707106781), VSUB(T38, T3b)); T3k = VMUL(LDK(KP707106781), VADD(T38, T3b)); { V T3g, T3l, T3m, T3n; T3g = VBYI(VSUB(T3c, T3f)); T3l = VSUB(T3j, T3k); ST(&(x[WS(rs, 12)]), VADD(T3g, T3l), ms, &(x[0])); ST(&(x[WS(rs, 20)]), VSUB(T3l, T3g), ms, &(x[0])); T3m = VBYI(VADD(T3f, T3c)); T3n = VADD(T3j, T3k); ST(&(x[WS(rs, 4)]), VADD(T3m, T3n), ms, &(x[0])); ST(&(x[WS(rs, 28)]), VSUB(T3n, T3m), ms, &(x[0])); } } { V T2L, T31, T2R, T2Y, T2A, T2Z, T2U, T32, T2H, T2Q; T2H = VMUL(LDK(KP707106781), VSUB(T2D, T2G)); T2L = VSUB(T2H, T2K); T31 = VADD(T2K, T2H); T2Q = VMUL(LDK(KP707106781), VADD(T2D, T2G)); T2R = VSUB(T2P, T2Q); T2Y = VADD(T2P, T2Q); { V T2s, T2z, T2S, T2T; T2s = VFNMS(LDK(KP382683432), T2r, VMUL(LDK(KP923879532), T2o)); T2z = VFMA(LDK(KP923879532), T2v, VMUL(LDK(KP382683432), T2y)); T2A = VSUB(T2s, T2z); T2Z = VADD(T2s, T2z); T2S = VFMA(LDK(KP382683432), T2o, VMUL(LDK(KP923879532), T2r)); T2T = VFNMS(LDK(KP382683432), T2v, VMUL(LDK(KP923879532), T2y)); T2U = VSUB(T2S, T2T); T32 = VADD(T2S, T2T); } { V T2M, T2V, T34, T35; T2M = VBYI(VSUB(T2A, T2L)); T2V = VSUB(T2R, T2U); ST(&(x[WS(rs, 10)]), VADD(T2M, T2V), ms, &(x[0])); ST(&(x[WS(rs, 22)]), VSUB(T2V, T2M), ms, &(x[0])); T34 = VSUB(T2Y, T2Z); T35 = VBYI(VSUB(T32, T31)); ST(&(x[WS(rs, 18)]), VSUB(T34, T35), ms, &(x[0])); ST(&(x[WS(rs, 14)]), VADD(T34, T35), ms, &(x[0])); } { V T2W, T2X, T30, T33; T2W = VBYI(VADD(T2L, T2A)); T2X = VADD(T2R, T2U); ST(&(x[WS(rs, 6)]), VADD(T2W, T2X), ms, &(x[0])); ST(&(x[WS(rs, 26)]), VSUB(T2X, T2W), ms, &(x[0])); T30 = VADD(T2Y, T2Z); T33 = VBYI(VADD(T31, T32)); ST(&(x[WS(rs, 30)]), VSUB(T30, T33), ms, &(x[0])); ST(&(x[WS(rs, 2)]), VADD(T30, T33), ms, &(x[0])); } } { V TE, T1P, T1I, T1Q, T1t, T1M, T1F, T1N; { V Tg, TD, T1G, T1H; Tg = VSUB(T4, Tf); TD = VSUB(Tr, TC); TE = VSUB(Tg, TD); T1P = VADD(Tg, TD); T1G = VFNMS(LDK(KP555570233), TV, VMUL(LDK(KP831469612), T12)); T1H = VFMA(LDK(KP555570233), T1k, VMUL(LDK(KP831469612), T1r)); T1I = VSUB(T1G, T1H); T1Q = VADD(T1G, T1H); } { V T13, T1s, T1x, T1E; T13 = VFMA(LDK(KP831469612), TV, VMUL(LDK(KP555570233), T12)); T1s = VFNMS(LDK(KP555570233), T1r, VMUL(LDK(KP831469612), T1k)); T1t = VSUB(T13, T1s); T1M = VADD(T13, T1s); T1x = VSUB(T1v, T1w); T1E = VSUB(T1y, T1D); T1F = VSUB(T1x, T1E); T1N = VADD(T1E, T1x); } { V T1u, T1J, T1S, T1T; T1u = VADD(TE, T1t); T1J = VBYI(VADD(T1F, T1I)); ST(&(x[WS(rs, 27)]), VSUB(T1u, T1J), ms, &(x[WS(rs, 1)])); ST(&(x[WS(rs, 5)]), VADD(T1u, T1J), ms, &(x[WS(rs, 1)])); T1S = VBYI(VADD(T1N, T1M)); T1T = VADD(T1P, T1Q); ST(&(x[WS(rs, 3)]), VADD(T1S, T1T), ms, &(x[WS(rs, 1)])); ST(&(x[WS(rs, 29)]), VSUB(T1T, T1S), ms, &(x[WS(rs, 1)])); } { V T1K, T1L, T1O, T1R; T1K = VSUB(TE, T1t); T1L = VBYI(VSUB(T1I, T1F)); ST(&(x[WS(rs, 21)]), VSUB(T1K, T1L), ms, &(x[WS(rs, 1)])); ST(&(x[WS(rs, 11)]), VADD(T1K, T1L), ms, &(x[WS(rs, 1)])); T1O = VBYI(VSUB(T1M, T1N)); T1R = VSUB(T1P, T1Q); ST(&(x[WS(rs, 13)]), VADD(T1O, T1R), ms, &(x[WS(rs, 1)])); ST(&(x[WS(rs, 19)]), VSUB(T1R, T1O), ms, &(x[WS(rs, 1)])); } } { V T1W, T2h, T2a, T2i, T23, T2e, T27, T2f; { V T1U, T1V, T28, T29; T1U = VADD(T4, Tf); T1V = VADD(T1v, T1w); T1W = VSUB(T1U, T1V); T2h = VADD(T1U, T1V); T28 = VFNMS(LDK(KP195090322), T1X, VMUL(LDK(KP980785280), T1Y)); T29 = VFMA(LDK(KP195090322), T20, VMUL(LDK(KP980785280), T21)); T2a = VSUB(T28, T29); T2i = VADD(T28, T29); } { V T1Z, T22, T25, T26; T1Z = VFMA(LDK(KP980785280), T1X, VMUL(LDK(KP195090322), T1Y)); T22 = VFNMS(LDK(KP195090322), T21, VMUL(LDK(KP980785280), T20)); T23 = VSUB(T1Z, T22); T2e = VADD(T1Z, T22); T25 = VADD(Tr, TC); T26 = VADD(T1D, T1y); T27 = VSUB(T25, T26); T2f = VADD(T26, T25); } { V T24, T2b, T2k, T2l; T24 = VADD(T1W, T23); T2b = VBYI(VADD(T27, T2a)); ST(&(x[WS(rs, 25)]), VSUB(T24, T2b), ms, &(x[WS(rs, 1)])); ST(&(x[WS(rs, 7)]), VADD(T24, T2b), ms, &(x[WS(rs, 1)])); T2k = VBYI(VADD(T2f, T2e)); T2l = VADD(T2h, T2i); ST(&(x[WS(rs, 1)]), VADD(T2k, T2l), ms, &(x[WS(rs, 1)])); ST(&(x[WS(rs, 31)]), VSUB(T2l, T2k), ms, &(x[WS(rs, 1)])); } { V T2c, T2d, T2g, T2j; T2c = VSUB(T1W, T23); T2d = VBYI(VSUB(T2a, T27)); ST(&(x[WS(rs, 23)]), VSUB(T2c, T2d), ms, &(x[WS(rs, 1)])); ST(&(x[WS(rs, 9)]), VADD(T2c, T2d), ms, &(x[WS(rs, 1)])); T2g = VBYI(VSUB(T2e, T2f)); T2j = VSUB(T2h, T2i); ST(&(x[WS(rs, 15)]), VADD(T2g, T2j), ms, &(x[WS(rs, 1)])); ST(&(x[WS(rs, 17)]), VSUB(T2j, T2g), ms, &(x[WS(rs, 1)])); } } } } VLEAVE(); } static const tw_instr twinstr[] = { VTW(0, 1), VTW(0, 2), VTW(0, 3), VTW(0, 4), VTW(0, 5), VTW(0, 6), VTW(0, 7), VTW(0, 8), VTW(0, 9), VTW(0, 10), VTW(0, 11), VTW(0, 12), VTW(0, 13), VTW(0, 14), VTW(0, 15), VTW(0, 16), VTW(0, 17), VTW(0, 18), VTW(0, 19), VTW(0, 20), VTW(0, 21), VTW(0, 22), VTW(0, 23), VTW(0, 24), VTW(0, 25), VTW(0, 26), VTW(0, 27), VTW(0, 28), VTW(0, 29), VTW(0, 30), VTW(0, 31), {TW_NEXT, VL, 0} }; static const ct_desc desc = { 32, XSIMD_STRING("t2bv_32"), twinstr, &GENUS, {201, 88, 16, 0}, 0, 0, 0 }; void XSIMD(codelet_t2bv_32) (planner *p) { X(kdft_dit_register) (p, t2bv_32, &desc); } #endif /* HAVE_FMA */ fftw-3.3.3/dft/simd/common/t2bv_64.c0000644000175000017500000017562712054410606013762 00000000000000/* * Copyright (c) 2003, 2007-11 Matteo Frigo * Copyright (c) 2003, 2007-11 Massachusetts Institute of Technology * * This program is free software; you can redistribute it and/or modify * it under the terms of the GNU General Public License as published by * the Free Software Foundation; either version 2 of the License, or * (at your option) any later version. * * This program is distributed in the hope that it will be useful, * but WITHOUT ANY WARRANTY; without even the implied warranty of * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the * GNU General Public License for more details. * * You should have received a copy of the GNU General Public License * along with this program; if not, write to the Free Software * Foundation, Inc., 51 Franklin Street, Fifth Floor, Boston, MA 02110-1301 USA * */ /* This file was automatically generated --- DO NOT EDIT */ /* Generated on Sun Nov 25 07:39:13 EST 2012 */ #include "codelet-dft.h" #ifdef HAVE_FMA /* Generated by: ../../../genfft/gen_twiddle_c.native -fma -reorder-insns -schedule-for-pipeline -simd -compact -variables 4 -pipeline-latency 8 -n 64 -name t2bv_64 -include t2b.h -sign 1 */ /* * This function contains 519 FP additions, 384 FP multiplications, * (or, 261 additions, 126 multiplications, 258 fused multiply/add), * 187 stack variables, 15 constants, and 128 memory accesses */ #include "t2b.h" static void t2bv_64(R *ri, R *ii, const R *W, stride rs, INT mb, INT me, INT ms) { DVK(KP773010453, +0.773010453362736960810906609758469800971041293); DVK(KP995184726, +0.995184726672196886244836953109479921575474869); DVK(KP820678790, +0.820678790828660330972281985331011598767386482); DVK(KP098491403, +0.098491403357164253077197521291327432293052451); DVK(KP956940335, +0.956940335732208864935797886980269969482849206); DVK(KP881921264, +0.881921264348355029712756863660388349508442621); DVK(KP303346683, +0.303346683607342391675883946941299872384187453); DVK(KP534511135, +0.534511135950791641089685961295362908582039528); DVK(KP831469612, +0.831469612302545237078788377617905756738560812); DVK(KP980785280, +0.980785280403230449126182236134239036973933731); DVK(KP668178637, +0.668178637919298919997757686523080761552472251); DVK(KP198912367, +0.198912367379658006911597622644676228597850501); DVK(KP923879532, +0.923879532511286756128183189396788286822416626); DVK(KP414213562, +0.414213562373095048801688724209698078569671875); DVK(KP707106781, +0.707106781186547524400844362104849039284835938); { INT m; R *x; x = ii; for (m = mb, W = W + (mb * ((TWVL / VL) * 126)); m < me; m = m + VL, x = x + (VL * ms), W = W + (TWVL * 126), MAKE_VOLATILE_STRIDE(64, rs)) { V T6L, T6M, T6O, T6P, T75, T6V, T5A, T6A, T72, T6K, T6t, T6D, T6w, T6B, T6h; V T6E; { V Ta, T3U, T3V, T37, T7a, T58, T7B, T6l, T1v, T24, T5Q, T7o, T5F, T7l, T43; V T4F, T2i, T2R, T6b, T7v, T60, T7s, T4a, T4I, T5u, T7h, T5x, T7g, T1i, T3b; V T4m, T4C, T7e, T5l, T7d, T5o, T3a, TV, T4B, T4j, T3X, T3Y, T6o, T7b, T5f; V T7C, Tx, T38, T2p, T61, T2n, T65, T2D, T7p, T5M, T7m, T5T, T4G, T46, T25; V T1S, T2q, T2u, T2w; { V T5q, T10, T5v, T15, T1b, T5s, T1c, T1e; { V T1V, T1p, T5B, T5O, T1u, T1X, T20, T21; { V T1, T2, T7, T5, T32, T34, T2X, T2Z; T1 = LD(&(x[0]), ms, &(x[0])); T2 = LD(&(x[WS(rs, 32)]), ms, &(x[0])); T7 = LD(&(x[WS(rs, 48)]), ms, &(x[0])); T5 = LD(&(x[WS(rs, 16)]), ms, &(x[0])); T32 = LD(&(x[WS(rs, 56)]), ms, &(x[0])); T34 = LD(&(x[WS(rs, 24)]), ms, &(x[0])); T2X = LD(&(x[WS(rs, 8)]), ms, &(x[0])); T2Z = LD(&(x[WS(rs, 40)]), ms, &(x[0])); { V T1m, T54, T6j, T36, T56, T31, T55, T1n, T1q, T1s, T4, T9; { V T3, T8, T6, T33, T35, T2Y, T30, T1l; T1l = LD(&(x[WS(rs, 1)]), ms, &(x[WS(rs, 1)])); T3 = BYTW(&(W[TWVL * 62]), T2); T8 = BYTW(&(W[TWVL * 94]), T7); T6 = BYTW(&(W[TWVL * 30]), T5); T33 = BYTW(&(W[TWVL * 110]), T32); T35 = BYTW(&(W[TWVL * 46]), T34); T2Y = BYTW(&(W[TWVL * 14]), T2X); T30 = BYTW(&(W[TWVL * 78]), T2Z); T1m = BYTW(&(W[0]), T1l); T54 = VSUB(T1, T3); T4 = VADD(T1, T3); T6j = VSUB(T6, T8); T9 = VADD(T6, T8); T36 = VADD(T33, T35); T56 = VSUB(T33, T35); T31 = VADD(T2Y, T30); T55 = VSUB(T2Y, T30); T1n = LD(&(x[WS(rs, 33)]), ms, &(x[WS(rs, 1)])); } T1q = LD(&(x[WS(rs, 17)]), ms, &(x[WS(rs, 1)])); T1s = LD(&(x[WS(rs, 49)]), ms, &(x[WS(rs, 1)])); Ta = VSUB(T4, T9); T3U = VADD(T4, T9); { V T57, T6k, T1o, T1r, T1t, T1W, T1U, T1Z; T1U = LD(&(x[WS(rs, 9)]), ms, &(x[WS(rs, 1)])); T3V = VADD(T31, T36); T37 = VSUB(T31, T36); T57 = VADD(T55, T56); T6k = VSUB(T55, T56); T1o = BYTW(&(W[TWVL * 64]), T1n); T1r = BYTW(&(W[TWVL * 32]), T1q); T1t = BYTW(&(W[TWVL * 96]), T1s); T1V = BYTW(&(W[TWVL * 16]), T1U); T1W = LD(&(x[WS(rs, 41)]), ms, &(x[WS(rs, 1)])); T1Z = LD(&(x[WS(rs, 57)]), ms, &(x[WS(rs, 1)])); T7a = VFNMS(LDK(KP707106781), T57, T54); T58 = VFMA(LDK(KP707106781), T57, T54); T7B = VFNMS(LDK(KP707106781), T6k, T6j); T6l = VFMA(LDK(KP707106781), T6k, T6j); T1p = VADD(T1m, T1o); T5B = VSUB(T1m, T1o); T5O = VSUB(T1r, T1t); T1u = VADD(T1r, T1t); T1X = BYTW(&(W[TWVL * 80]), T1W); T20 = BYTW(&(W[TWVL * 112]), T1Z); T21 = LD(&(x[WS(rs, 25)]), ms, &(x[WS(rs, 1)])); } } } { V T5W, T2N, T69, T2L, T5Y, T2P, T48, T2c, T2h; { V T41, T1Y, T5C, T22, T2d, T29, T2b, T2f, T28, T2a, T2H, T2J; T28 = LD(&(x[WS(rs, 63)]), ms, &(x[WS(rs, 1)])); T2a = LD(&(x[WS(rs, 31)]), ms, &(x[WS(rs, 1)])); T1v = VSUB(T1p, T1u); T41 = VADD(T1p, T1u); T1Y = VADD(T1V, T1X); T5C = VSUB(T1V, T1X); T22 = BYTW(&(W[TWVL * 48]), T21); T2d = LD(&(x[WS(rs, 15)]), ms, &(x[WS(rs, 1)])); T29 = BYTW(&(W[TWVL * 124]), T28); T2b = BYTW(&(W[TWVL * 60]), T2a); T2f = LD(&(x[WS(rs, 47)]), ms, &(x[WS(rs, 1)])); T2H = LD(&(x[WS(rs, 55)]), ms, &(x[WS(rs, 1)])); T2J = LD(&(x[WS(rs, 23)]), ms, &(x[WS(rs, 1)])); { V T23, T5D, T2e, T2g, T2I, T2K, T2M; T2M = LD(&(x[WS(rs, 7)]), ms, &(x[WS(rs, 1)])); T23 = VADD(T20, T22); T5D = VSUB(T20, T22); T2e = BYTW(&(W[TWVL * 28]), T2d); T2c = VADD(T29, T2b); T5W = VSUB(T29, T2b); T2g = BYTW(&(W[TWVL * 92]), T2f); T2I = BYTW(&(W[TWVL * 108]), T2H); T2K = BYTW(&(W[TWVL * 44]), T2J); T2N = BYTW(&(W[TWVL * 12]), T2M); { V T5E, T5P, T42, T2O; T5E = VADD(T5C, T5D); T5P = VSUB(T5C, T5D); T24 = VSUB(T1Y, T23); T42 = VADD(T1Y, T23); T69 = VSUB(T2g, T2e); T2h = VADD(T2e, T2g); T2O = LD(&(x[WS(rs, 39)]), ms, &(x[WS(rs, 1)])); T2L = VADD(T2I, T2K); T5Y = VSUB(T2I, T2K); T5Q = VFMA(LDK(KP707106781), T5P, T5O); T7o = VFNMS(LDK(KP707106781), T5P, T5O); T5F = VFMA(LDK(KP707106781), T5E, T5B); T7l = VFNMS(LDK(KP707106781), T5E, T5B); T43 = VADD(T41, T42); T4F = VSUB(T41, T42); T2P = BYTW(&(W[TWVL * 76]), T2O); } } } T2i = VSUB(T2c, T2h); T48 = VADD(T2c, T2h); { V TW, TY, T11, T2Q, T5X, T13; TW = LD(&(x[WS(rs, 62)]), ms, &(x[0])); TY = LD(&(x[WS(rs, 30)]), ms, &(x[0])); T11 = LD(&(x[WS(rs, 14)]), ms, &(x[0])); T2Q = VADD(T2N, T2P); T5X = VSUB(T2N, T2P); T13 = LD(&(x[WS(rs, 46)]), ms, &(x[0])); { V T12, T5Z, T6a, T49, T14, T18, T1a; { V T17, T19, TX, TZ; T17 = LD(&(x[WS(rs, 54)]), ms, &(x[0])); T19 = LD(&(x[WS(rs, 22)]), ms, &(x[0])); TX = BYTW(&(W[TWVL * 122]), TW); TZ = BYTW(&(W[TWVL * 58]), TY); T12 = BYTW(&(W[TWVL * 26]), T11); T5Z = VADD(T5X, T5Y); T6a = VSUB(T5Y, T5X); T2R = VSUB(T2L, T2Q); T49 = VADD(T2Q, T2L); T14 = BYTW(&(W[TWVL * 90]), T13); T18 = BYTW(&(W[TWVL * 106]), T17); T5q = VSUB(TX, TZ); T10 = VADD(TX, TZ); T1a = BYTW(&(W[TWVL * 42]), T19); } T6b = VFMA(LDK(KP707106781), T6a, T69); T7v = VFNMS(LDK(KP707106781), T6a, T69); T60 = VFMA(LDK(KP707106781), T5Z, T5W); T7s = VFNMS(LDK(KP707106781), T5Z, T5W); T4a = VADD(T48, T49); T4I = VSUB(T48, T49); T5v = VSUB(T14, T12); T15 = VADD(T12, T14); T1b = VADD(T18, T1a); T5s = VSUB(T18, T1a); } T1c = LD(&(x[WS(rs, 6)]), ms, &(x[0])); T1e = LD(&(x[WS(rs, 38)]), ms, &(x[0])); } } } { V Th, T59, Tf, Tv, T5d, Tj, Tm, To; { V T5h, TQ, T5m, T5i, TO, TS, TJ, T4h, TD, TI; { V T4k, T16, TB, T1d, T1f, TE, TG, TA, Tz, TK, TM, TC; Tz = LD(&(x[WS(rs, 2)]), ms, &(x[0])); T4k = VADD(T10, T15); T16 = VSUB(T10, T15); TB = LD(&(x[WS(rs, 34)]), ms, &(x[0])); T1d = BYTW(&(W[TWVL * 10]), T1c); T1f = BYTW(&(W[TWVL * 74]), T1e); TE = LD(&(x[WS(rs, 18)]), ms, &(x[0])); TG = LD(&(x[WS(rs, 50)]), ms, &(x[0])); TA = BYTW(&(W[TWVL * 2]), Tz); TK = LD(&(x[WS(rs, 10)]), ms, &(x[0])); TM = LD(&(x[WS(rs, 42)]), ms, &(x[0])); TC = BYTW(&(W[TWVL * 66]), TB); { V T1g, T5r, TF, TH, TL, TN, TP; TP = LD(&(x[WS(rs, 58)]), ms, &(x[0])); T1g = VADD(T1d, T1f); T5r = VSUB(T1d, T1f); TF = BYTW(&(W[TWVL * 34]), TE); TH = BYTW(&(W[TWVL * 98]), TG); TL = BYTW(&(W[TWVL * 18]), TK); TN = BYTW(&(W[TWVL * 82]), TM); T5h = VSUB(TA, TC); TD = VADD(TA, TC); TQ = BYTW(&(W[TWVL * 114]), TP); { V T5w, T5t, T4l, T1h, TR; T5w = VSUB(T5s, T5r); T5t = VADD(T5r, T5s); T4l = VADD(T1g, T1b); T1h = VSUB(T1b, T1g); T5m = VSUB(TF, TH); TI = VADD(TF, TH); T5i = VSUB(TL, TN); TO = VADD(TL, TN); TR = LD(&(x[WS(rs, 26)]), ms, &(x[0])); T5u = VFMA(LDK(KP707106781), T5t, T5q); T7h = VFNMS(LDK(KP707106781), T5t, T5q); T5x = VFMA(LDK(KP707106781), T5w, T5v); T7g = VFNMS(LDK(KP707106781), T5w, T5v); T1i = VFNMS(LDK(KP414213562), T1h, T16); T3b = VFMA(LDK(KP414213562), T16, T1h); T4m = VADD(T4k, T4l); T4C = VSUB(T4k, T4l); TS = BYTW(&(W[TWVL * 50]), TR); } } } TJ = VSUB(TD, TI); T4h = VADD(TD, TI); { V Tb, Td, Tr, T5j, TT, Tt, Tg; Tb = LD(&(x[WS(rs, 4)]), ms, &(x[0])); Td = LD(&(x[WS(rs, 36)]), ms, &(x[0])); Tr = LD(&(x[WS(rs, 12)]), ms, &(x[0])); T5j = VSUB(TQ, TS); TT = VADD(TQ, TS); Tt = LD(&(x[WS(rs, 44)]), ms, &(x[0])); Tg = LD(&(x[WS(rs, 20)]), ms, &(x[0])); { V Ti, Tc, Te, Ts; Ti = LD(&(x[WS(rs, 52)]), ms, &(x[0])); Tc = BYTW(&(W[TWVL * 6]), Tb); Te = BYTW(&(W[TWVL * 70]), Td); Ts = BYTW(&(W[TWVL * 22]), Tr); { V T5k, T5n, TU, T4i, Tu; T5k = VADD(T5i, T5j); T5n = VSUB(T5i, T5j); TU = VSUB(TO, TT); T4i = VADD(TO, TT); Tu = BYTW(&(W[TWVL * 86]), Tt); Th = BYTW(&(W[TWVL * 38]), Tg); T59 = VSUB(Tc, Te); Tf = VADD(Tc, Te); T7e = VFNMS(LDK(KP707106781), T5k, T5h); T5l = VFMA(LDK(KP707106781), T5k, T5h); T7d = VFNMS(LDK(KP707106781), T5n, T5m); T5o = VFMA(LDK(KP707106781), T5n, T5m); T3a = VFMA(LDK(KP414213562), TJ, TU); TV = VFNMS(LDK(KP414213562), TU, TJ); T4B = VSUB(T4h, T4i); T4j = VADD(T4h, T4i); Tv = VADD(Ts, Tu); T5d = VSUB(Tu, Ts); Tj = BYTW(&(W[TWVL * 102]), Ti); } } Tm = LD(&(x[WS(rs, 60)]), ms, &(x[0])); To = LD(&(x[WS(rs, 28)]), ms, &(x[0])); } } { V T5b, T6m, Tl, T1A, T5G, T1Q, T5K, T1C, T1D, T5e, T6n, Tw, T1H, T1J; { V T1w, T1y, T1M, T1O, Tq, T5c, T1B; T1w = LD(&(x[WS(rs, 5)]), ms, &(x[WS(rs, 1)])); T1y = LD(&(x[WS(rs, 37)]), ms, &(x[WS(rs, 1)])); T1M = LD(&(x[WS(rs, 13)]), ms, &(x[WS(rs, 1)])); T1O = LD(&(x[WS(rs, 45)]), ms, &(x[WS(rs, 1)])); T1B = LD(&(x[WS(rs, 21)]), ms, &(x[WS(rs, 1)])); { V Tk, T5a, Tn, Tp; Tk = VADD(Th, Tj); T5a = VSUB(Th, Tj); Tn = BYTW(&(W[TWVL * 118]), Tm); Tp = BYTW(&(W[TWVL * 54]), To); { V T1x, T1z, T1N, T1P; T1x = BYTW(&(W[TWVL * 8]), T1w); T1z = BYTW(&(W[TWVL * 72]), T1y); T1N = BYTW(&(W[TWVL * 24]), T1M); T1P = BYTW(&(W[TWVL * 88]), T1O); T5b = VFNMS(LDK(KP414213562), T5a, T59); T6m = VFMA(LDK(KP414213562), T59, T5a); T3X = VADD(Tf, Tk); Tl = VSUB(Tf, Tk); Tq = VADD(Tn, Tp); T5c = VSUB(Tn, Tp); T1A = VADD(T1x, T1z); T5G = VSUB(T1x, T1z); T1Q = VADD(T1N, T1P); T5K = VSUB(T1N, T1P); T1C = BYTW(&(W[TWVL * 40]), T1B); } } T1D = LD(&(x[WS(rs, 53)]), ms, &(x[WS(rs, 1)])); T5e = VFNMS(LDK(KP414213562), T5d, T5c); T6n = VFMA(LDK(KP414213562), T5c, T5d); T3Y = VADD(Tq, Tv); Tw = VSUB(Tq, Tv); T1H = LD(&(x[WS(rs, 61)]), ms, &(x[WS(rs, 1)])); T1J = LD(&(x[WS(rs, 29)]), ms, &(x[WS(rs, 1)])); } { V T1I, T1K, T1F, T5H, T2k, T2l, T2z, T2B, T2j, T1E; T2j = LD(&(x[WS(rs, 3)]), ms, &(x[WS(rs, 1)])); T1E = BYTW(&(W[TWVL * 104]), T1D); T6o = VSUB(T6m, T6n); T7b = VADD(T6m, T6n); T5f = VADD(T5b, T5e); T7C = VSUB(T5b, T5e); Tx = VADD(Tl, Tw); T38 = VSUB(Tl, Tw); T1I = BYTW(&(W[TWVL * 120]), T1H); T1K = BYTW(&(W[TWVL * 56]), T1J); T1F = VADD(T1C, T1E); T5H = VSUB(T1C, T1E); T2k = BYTW(&(W[TWVL * 4]), T2j); T2l = LD(&(x[WS(rs, 35)]), ms, &(x[WS(rs, 1)])); T2z = LD(&(x[WS(rs, 11)]), ms, &(x[WS(rs, 1)])); T2B = LD(&(x[WS(rs, 43)]), ms, &(x[WS(rs, 1)])); { V T5I, T5R, T44, T1G, T2m, T2A, T2C, T5S, T5L, T1R, T45, T2o, T5J, T1L; T2o = LD(&(x[WS(rs, 19)]), ms, &(x[WS(rs, 1)])); T5J = VSUB(T1I, T1K); T1L = VADD(T1I, T1K); T5I = VFNMS(LDK(KP414213562), T5H, T5G); T5R = VFMA(LDK(KP414213562), T5G, T5H); T44 = VADD(T1A, T1F); T1G = VSUB(T1A, T1F); T2m = BYTW(&(W[TWVL * 68]), T2l); T2A = BYTW(&(W[TWVL * 20]), T2z); T2C = BYTW(&(W[TWVL * 84]), T2B); T5S = VFNMS(LDK(KP414213562), T5J, T5K); T5L = VFMA(LDK(KP414213562), T5K, T5J); T1R = VSUB(T1L, T1Q); T45 = VADD(T1L, T1Q); T2p = BYTW(&(W[TWVL * 36]), T2o); T61 = VSUB(T2k, T2m); T2n = VADD(T2k, T2m); T65 = VSUB(T2C, T2A); T2D = VADD(T2A, T2C); T7p = VSUB(T5I, T5L); T5M = VADD(T5I, T5L); T7m = VSUB(T5R, T5S); T5T = VADD(T5R, T5S); T4G = VSUB(T44, T45); T46 = VADD(T44, T45); T25 = VSUB(T1G, T1R); T1S = VADD(T1G, T1R); T2q = LD(&(x[WS(rs, 51)]), ms, &(x[WS(rs, 1)])); } T2u = LD(&(x[WS(rs, 59)]), ms, &(x[WS(rs, 1)])); T2w = LD(&(x[WS(rs, 27)]), ms, &(x[WS(rs, 1)])); } } } } { V T67, T7w, T6e, T7t, T3s, T3E, T39, T3D, T1k, T3k, T3t, T3c, T1T, T3v, T3w; V T26, T2G, T3y, T3z, T2T; { V T4A, T4N, T47, T4v, T2r, T2v, T2x, T4s, T40, T3W, T3Z; T4A = VSUB(T3U, T3V); T3W = VADD(T3U, T3V); T3Z = VADD(T3X, T3Y); T4N = VSUB(T3X, T3Y); T47 = VSUB(T43, T46); T4v = VADD(T43, T46); T2r = BYTW(&(W[TWVL * 100]), T2q); T2v = BYTW(&(W[TWVL * 116]), T2u); T2x = BYTW(&(W[TWVL * 52]), T2w); T4s = VADD(T3W, T3Z); T40 = VSUB(T3W, T3Z); { V T4O, T4n, T4Q, T4H, T4E, T4W, T4u, T4y, T4d, T4J, T2F, T2S; { V T6c, T63, T2t, T4b, T6d, T66, T2E, T4c; { V T4D, T62, T2s, T64, T2y, T4t; T4O = VSUB(T4B, T4C); T4D = VADD(T4B, T4C); T62 = VSUB(T2r, T2p); T2s = VADD(T2p, T2r); T64 = VSUB(T2v, T2x); T2y = VADD(T2v, T2x); T4t = VADD(T4j, T4m); T4n = VSUB(T4j, T4m); T4Q = VFMA(LDK(KP414213562), T4F, T4G); T4H = VFNMS(LDK(KP414213562), T4G, T4F); T4E = VFMA(LDK(KP707106781), T4D, T4A); T4W = VFNMS(LDK(KP707106781), T4D, T4A); T6c = VFNMS(LDK(KP414213562), T61, T62); T63 = VFMA(LDK(KP414213562), T62, T61); T2t = VSUB(T2n, T2s); T4b = VADD(T2n, T2s); T6d = VFMA(LDK(KP414213562), T64, T65); T66 = VFNMS(LDK(KP414213562), T65, T64); T2E = VSUB(T2y, T2D); T4c = VADD(T2y, T2D); T4u = VSUB(T4s, T4t); T4y = VADD(T4s, T4t); } T67 = VADD(T63, T66); T7w = VSUB(T66, T63); T6e = VADD(T6c, T6d); T7t = VSUB(T6d, T6c); T4d = VADD(T4b, T4c); T4J = VSUB(T4c, T4b); T2F = VADD(T2t, T2E); T2S = VSUB(T2E, T2t); } { V Ty, T1j, T4R, T4K; Ty = VFMA(LDK(KP707106781), Tx, Ta); T3s = VFNMS(LDK(KP707106781), Tx, Ta); T3E = VSUB(TV, T1i); T1j = VADD(TV, T1i); T39 = VFMA(LDK(KP707106781), T38, T37); T3D = VFNMS(LDK(KP707106781), T38, T37); T4R = VFMA(LDK(KP414213562), T4I, T4J); T4K = VFNMS(LDK(KP414213562), T4J, T4I); { V T4w, T4e, T4P, T4Z; T4w = VADD(T4a, T4d); T4e = VSUB(T4a, T4d); T4P = VFMA(LDK(KP707106781), T4O, T4N); T4Z = VFNMS(LDK(KP707106781), T4O, T4N); T1k = VFMA(LDK(KP923879532), T1j, Ty); T3k = VFNMS(LDK(KP923879532), T1j, Ty); { V T4L, T50, T4S, T4X; T4L = VADD(T4H, T4K); T50 = VSUB(T4H, T4K); T4S = VSUB(T4Q, T4R); T4X = VADD(T4Q, T4R); { V T4f, T4o, T4x, T4z; T4f = VADD(T47, T4e); T4o = VSUB(T47, T4e); T4x = VSUB(T4v, T4w); T4z = VADD(T4v, T4w); { V T53, T51, T4M, T4U; T53 = VFNMS(LDK(KP923879532), T50, T4Z); T51 = VFMA(LDK(KP923879532), T50, T4Z); T4M = VFNMS(LDK(KP923879532), T4L, T4E); T4U = VFMA(LDK(KP923879532), T4L, T4E); { V T52, T4Y, T4T, T4V; T52 = VFMA(LDK(KP923879532), T4X, T4W); T4Y = VFNMS(LDK(KP923879532), T4X, T4W); T4T = VFNMS(LDK(KP923879532), T4S, T4P); T4V = VFMA(LDK(KP923879532), T4S, T4P); { V T4p, T4r, T4g, T4q; T4p = VFNMS(LDK(KP707106781), T4o, T4n); T4r = VFMA(LDK(KP707106781), T4o, T4n); T4g = VFNMS(LDK(KP707106781), T4f, T40); T4q = VFMA(LDK(KP707106781), T4f, T40); ST(&(x[0]), VADD(T4y, T4z), ms, &(x[0])); ST(&(x[WS(rs, 32)]), VSUB(T4y, T4z), ms, &(x[0])); ST(&(x[WS(rs, 16)]), VFMAI(T4x, T4u), ms, &(x[0])); ST(&(x[WS(rs, 48)]), VFNMSI(T4x, T4u), ms, &(x[0])); ST(&(x[WS(rs, 44)]), VFNMSI(T51, T4Y), ms, &(x[0])); ST(&(x[WS(rs, 20)]), VFMAI(T51, T4Y), ms, &(x[0])); ST(&(x[WS(rs, 52)]), VFMAI(T53, T52), ms, &(x[0])); ST(&(x[WS(rs, 12)]), VFNMSI(T53, T52), ms, &(x[0])); ST(&(x[WS(rs, 4)]), VFMAI(T4V, T4U), ms, &(x[0])); ST(&(x[WS(rs, 60)]), VFNMSI(T4V, T4U), ms, &(x[0])); ST(&(x[WS(rs, 36)]), VFMAI(T4T, T4M), ms, &(x[0])); ST(&(x[WS(rs, 28)]), VFNMSI(T4T, T4M), ms, &(x[0])); ST(&(x[WS(rs, 56)]), VFNMSI(T4r, T4q), ms, &(x[0])); ST(&(x[WS(rs, 8)]), VFMAI(T4r, T4q), ms, &(x[0])); ST(&(x[WS(rs, 40)]), VFMAI(T4p, T4g), ms, &(x[0])); ST(&(x[WS(rs, 24)]), VFNMSI(T4p, T4g), ms, &(x[0])); T3t = VADD(T3a, T3b); T3c = VSUB(T3a, T3b); } } } } } } T1T = VFMA(LDK(KP707106781), T1S, T1v); T3v = VFNMS(LDK(KP707106781), T1S, T1v); T3w = VFNMS(LDK(KP707106781), T25, T24); T26 = VFMA(LDK(KP707106781), T25, T24); T2G = VFMA(LDK(KP707106781), T2F, T2i); T3y = VFNMS(LDK(KP707106781), T2F, T2i); T3z = VFNMS(LDK(KP707106781), T2S, T2R); T2T = VFMA(LDK(KP707106781), T2S, T2R); } } } { V T3u, T3M, T3F, T3P, T3x, T3G, T3q, T3m, T3h, T3j, T3r, T3p, T2W, T3i; { V T3d, T3n, T27, T3e, T2U, T3f; T3d = VFMA(LDK(KP923879532), T3c, T39); T3n = VFNMS(LDK(KP923879532), T3c, T39); T27 = VFNMS(LDK(KP198912367), T26, T1T); T3e = VFMA(LDK(KP198912367), T1T, T26); T2U = VFNMS(LDK(KP198912367), T2T, T2G); T3f = VFMA(LDK(KP198912367), T2G, T2T); T3u = VFMA(LDK(KP923879532), T3t, T3s); T3M = VFNMS(LDK(KP923879532), T3t, T3s); { V T3g, T3l, T2V, T3o; T3g = VSUB(T3e, T3f); T3l = VADD(T3e, T3f); T2V = VADD(T27, T2U); T3o = VSUB(T27, T2U); T3F = VFNMS(LDK(KP923879532), T3E, T3D); T3P = VFMA(LDK(KP923879532), T3E, T3D); T3x = VFMA(LDK(KP668178637), T3w, T3v); T3G = VFNMS(LDK(KP668178637), T3v, T3w); T3q = VFMA(LDK(KP980785280), T3l, T3k); T3m = VFNMS(LDK(KP980785280), T3l, T3k); T3h = VFNMS(LDK(KP980785280), T3g, T3d); T3j = VFMA(LDK(KP980785280), T3g, T3d); T3r = VFNMS(LDK(KP980785280), T3o, T3n); T3p = VFMA(LDK(KP980785280), T3o, T3n); T2W = VFNMS(LDK(KP980785280), T2V, T1k); T3i = VFMA(LDK(KP980785280), T2V, T1k); } } { V T7n, T7Z, T8j, T89, T7k, T7O, T8g, T7Y, T7H, T7R, T80, T7q, T7u, T82, T83; V T7x; { V T7c, T7W, T7D, T87, T7f, T7E, T3A, T3H, T7F, T7i; T7c = VFNMS(LDK(KP923879532), T7b, T7a); T7W = VFMA(LDK(KP923879532), T7b, T7a); T7D = VFMA(LDK(KP923879532), T7C, T7B); T87 = VFNMS(LDK(KP923879532), T7C, T7B); T7f = VFNMS(LDK(KP668178637), T7e, T7d); T7E = VFMA(LDK(KP668178637), T7d, T7e); ST(&(x[WS(rs, 46)]), VFNMSI(T3p, T3m), ms, &(x[0])); ST(&(x[WS(rs, 18)]), VFMAI(T3p, T3m), ms, &(x[0])); ST(&(x[WS(rs, 50)]), VFMAI(T3r, T3q), ms, &(x[0])); ST(&(x[WS(rs, 14)]), VFNMSI(T3r, T3q), ms, &(x[0])); ST(&(x[WS(rs, 2)]), VFMAI(T3j, T3i), ms, &(x[0])); ST(&(x[WS(rs, 62)]), VFNMSI(T3j, T3i), ms, &(x[0])); ST(&(x[WS(rs, 34)]), VFMAI(T3h, T2W), ms, &(x[0])); ST(&(x[WS(rs, 30)]), VFNMSI(T3h, T2W), ms, &(x[0])); T3A = VFMA(LDK(KP668178637), T3z, T3y); T3H = VFNMS(LDK(KP668178637), T3y, T3z); T7F = VFMA(LDK(KP668178637), T7g, T7h); T7i = VFNMS(LDK(KP668178637), T7h, T7g); T7n = VFNMS(LDK(KP923879532), T7m, T7l); T7Z = VFMA(LDK(KP923879532), T7m, T7l); { V T3I, T3N, T3B, T3Q; T3I = VSUB(T3G, T3H); T3N = VADD(T3G, T3H); T3B = VADD(T3x, T3A); T3Q = VSUB(T3x, T3A); { V T7j, T88, T7G, T7X; T7j = VADD(T7f, T7i); T88 = VSUB(T7f, T7i); T7G = VSUB(T7E, T7F); T7X = VADD(T7E, T7F); { V T3S, T3O, T3J, T3L; T3S = VFNMS(LDK(KP831469612), T3N, T3M); T3O = VFMA(LDK(KP831469612), T3N, T3M); T3J = VFNMS(LDK(KP831469612), T3I, T3F); T3L = VFMA(LDK(KP831469612), T3I, T3F); { V T3T, T3R, T3C, T3K; T3T = VFMA(LDK(KP831469612), T3Q, T3P); T3R = VFNMS(LDK(KP831469612), T3Q, T3P); T3C = VFNMS(LDK(KP831469612), T3B, T3u); T3K = VFMA(LDK(KP831469612), T3B, T3u); T8j = VFNMS(LDK(KP831469612), T88, T87); T89 = VFMA(LDK(KP831469612), T88, T87); T7k = VFNMS(LDK(KP831469612), T7j, T7c); T7O = VFMA(LDK(KP831469612), T7j, T7c); T8g = VFNMS(LDK(KP831469612), T7X, T7W); T7Y = VFMA(LDK(KP831469612), T7X, T7W); T7H = VFMA(LDK(KP831469612), T7G, T7D); T7R = VFNMS(LDK(KP831469612), T7G, T7D); ST(&(x[WS(rs, 42)]), VFMAI(T3R, T3O), ms, &(x[0])); ST(&(x[WS(rs, 22)]), VFNMSI(T3R, T3O), ms, &(x[0])); ST(&(x[WS(rs, 54)]), VFNMSI(T3T, T3S), ms, &(x[0])); ST(&(x[WS(rs, 10)]), VFMAI(T3T, T3S), ms, &(x[0])); ST(&(x[WS(rs, 58)]), VFMAI(T3L, T3K), ms, &(x[0])); ST(&(x[WS(rs, 6)]), VFNMSI(T3L, T3K), ms, &(x[0])); ST(&(x[WS(rs, 26)]), VFMAI(T3J, T3C), ms, &(x[0])); ST(&(x[WS(rs, 38)]), VFNMSI(T3J, T3C), ms, &(x[0])); T80 = VFNMS(LDK(KP923879532), T7p, T7o); T7q = VFMA(LDK(KP923879532), T7p, T7o); } } } } T7u = VFNMS(LDK(KP923879532), T7t, T7s); T82 = VFMA(LDK(KP923879532), T7t, T7s); T83 = VFNMS(LDK(KP923879532), T7w, T7v); T7x = VFMA(LDK(KP923879532), T7w, T7v); } { V T5g, T6I, T6p, T6T, T5p, T6q, T6r, T5y; T5g = VFMA(LDK(KP923879532), T5f, T58); T6I = VFNMS(LDK(KP923879532), T5f, T58); { V T7r, T7I, T7y, T7J; T7r = VFNMS(LDK(KP534511135), T7q, T7n); T7I = VFMA(LDK(KP534511135), T7n, T7q); T7y = VFNMS(LDK(KP534511135), T7x, T7u); T7J = VFMA(LDK(KP534511135), T7u, T7x); { V T81, T8a, T84, T8b; T81 = VFMA(LDK(KP303346683), T80, T7Z); T8a = VFNMS(LDK(KP303346683), T7Z, T80); T84 = VFMA(LDK(KP303346683), T83, T82); T8b = VFNMS(LDK(KP303346683), T82, T83); T6p = VFMA(LDK(KP923879532), T6o, T6l); T6T = VFNMS(LDK(KP923879532), T6o, T6l); T5p = VFNMS(LDK(KP198912367), T5o, T5l); T6q = VFMA(LDK(KP198912367), T5l, T5o); { V T7K, T7P, T7z, T7S; T7K = VSUB(T7I, T7J); T7P = VADD(T7I, T7J); T7z = VADD(T7r, T7y); T7S = VSUB(T7r, T7y); { V T8c, T8h, T85, T8k; T8c = VSUB(T8a, T8b); T8h = VADD(T8a, T8b); T85 = VADD(T81, T84); T8k = VSUB(T81, T84); { V T7Q, T7U, T7L, T7N; T7Q = VFNMS(LDK(KP881921264), T7P, T7O); T7U = VFMA(LDK(KP881921264), T7P, T7O); T7L = VFNMS(LDK(KP881921264), T7K, T7H); T7N = VFMA(LDK(KP881921264), T7K, T7H); { V T7T, T7V, T7A, T7M; T7T = VFMA(LDK(KP881921264), T7S, T7R); T7V = VFNMS(LDK(KP881921264), T7S, T7R); T7A = VFNMS(LDK(KP881921264), T7z, T7k); T7M = VFMA(LDK(KP881921264), T7z, T7k); { V T8i, T8m, T8d, T8f; T8i = VFMA(LDK(KP956940335), T8h, T8g); T8m = VFNMS(LDK(KP956940335), T8h, T8g); T8d = VFNMS(LDK(KP956940335), T8c, T89); T8f = VFMA(LDK(KP956940335), T8c, T89); { V T8l, T8n, T86, T8e; T8l = VFNMS(LDK(KP956940335), T8k, T8j); T8n = VFMA(LDK(KP956940335), T8k, T8j); T86 = VFNMS(LDK(KP956940335), T85, T7Y); T8e = VFMA(LDK(KP956940335), T85, T7Y); ST(&(x[WS(rs, 53)]), VFMAI(T7V, T7U), ms, &(x[WS(rs, 1)])); ST(&(x[WS(rs, 11)]), VFNMSI(T7V, T7U), ms, &(x[WS(rs, 1)])); ST(&(x[WS(rs, 43)]), VFNMSI(T7T, T7Q), ms, &(x[WS(rs, 1)])); ST(&(x[WS(rs, 21)]), VFMAI(T7T, T7Q), ms, &(x[WS(rs, 1)])); ST(&(x[WS(rs, 5)]), VFMAI(T7N, T7M), ms, &(x[WS(rs, 1)])); ST(&(x[WS(rs, 59)]), VFNMSI(T7N, T7M), ms, &(x[WS(rs, 1)])); ST(&(x[WS(rs, 37)]), VFMAI(T7L, T7A), ms, &(x[WS(rs, 1)])); ST(&(x[WS(rs, 27)]), VFNMSI(T7L, T7A), ms, &(x[WS(rs, 1)])); ST(&(x[WS(rs, 51)]), VFNMSI(T8n, T8m), ms, &(x[WS(rs, 1)])); ST(&(x[WS(rs, 13)]), VFMAI(T8n, T8m), ms, &(x[WS(rs, 1)])); ST(&(x[WS(rs, 45)]), VFMAI(T8l, T8i), ms, &(x[WS(rs, 1)])); ST(&(x[WS(rs, 19)]), VFNMSI(T8l, T8i), ms, &(x[WS(rs, 1)])); ST(&(x[WS(rs, 61)]), VFMAI(T8f, T8e), ms, &(x[WS(rs, 1)])); ST(&(x[WS(rs, 3)]), VFNMSI(T8f, T8e), ms, &(x[WS(rs, 1)])); ST(&(x[WS(rs, 29)]), VFMAI(T8d, T86), ms, &(x[WS(rs, 1)])); ST(&(x[WS(rs, 35)]), VFNMSI(T8d, T86), ms, &(x[WS(rs, 1)])); T6r = VFMA(LDK(KP198912367), T5u, T5x); T5y = VFNMS(LDK(KP198912367), T5x, T5u); } } } } } } } } { V T5N, T5U, T68, T5z, T6U, T6f; T5N = VFMA(LDK(KP923879532), T5M, T5F); T6L = VFNMS(LDK(KP923879532), T5M, T5F); T6M = VFNMS(LDK(KP923879532), T5T, T5Q); T5U = VFMA(LDK(KP923879532), T5T, T5Q); T68 = VFMA(LDK(KP923879532), T67, T60); T6O = VFNMS(LDK(KP923879532), T67, T60); T5z = VADD(T5p, T5y); T6U = VSUB(T5p, T5y); T6P = VFNMS(LDK(KP923879532), T6e, T6b); T6f = VFMA(LDK(KP923879532), T6e, T6b); { V T5V, T6u, T6g, T6v, T6s, T6J; T6s = VSUB(T6q, T6r); T6J = VADD(T6q, T6r); T5V = VFNMS(LDK(KP098491403), T5U, T5N); T6u = VFMA(LDK(KP098491403), T5N, T5U); T75 = VFMA(LDK(KP980785280), T6U, T6T); T6V = VFNMS(LDK(KP980785280), T6U, T6T); T5A = VFMA(LDK(KP980785280), T5z, T5g); T6A = VFNMS(LDK(KP980785280), T5z, T5g); T6g = VFNMS(LDK(KP098491403), T6f, T68); T6v = VFMA(LDK(KP098491403), T68, T6f); T72 = VFNMS(LDK(KP980785280), T6J, T6I); T6K = VFMA(LDK(KP980785280), T6J, T6I); T6t = VFMA(LDK(KP980785280), T6s, T6p); T6D = VFNMS(LDK(KP980785280), T6s, T6p); T6w = VSUB(T6u, T6v); T6B = VADD(T6u, T6v); T6h = VADD(T5V, T6g); T6E = VSUB(T5V, T6g); } } } } } } } { V T6W, T6N, T6G, T6C, T6z, T6x, T6H, T6F, T6y, T6i, T6X, T6Q; T6W = VFNMS(LDK(KP820678790), T6L, T6M); T6N = VFMA(LDK(KP820678790), T6M, T6L); T6G = VFMA(LDK(KP995184726), T6B, T6A); T6C = VFNMS(LDK(KP995184726), T6B, T6A); T6z = VFMA(LDK(KP995184726), T6w, T6t); T6x = VFNMS(LDK(KP995184726), T6w, T6t); T6H = VFNMS(LDK(KP995184726), T6E, T6D); T6F = VFMA(LDK(KP995184726), T6E, T6D); T6y = VFMA(LDK(KP995184726), T6h, T5A); T6i = VFNMS(LDK(KP995184726), T6h, T5A); T6X = VFNMS(LDK(KP820678790), T6O, T6P); T6Q = VFMA(LDK(KP820678790), T6P, T6O); { V T73, T6Y, T76, T6R; ST(&(x[WS(rs, 49)]), VFMAI(T6H, T6G), ms, &(x[WS(rs, 1)])); ST(&(x[WS(rs, 15)]), VFNMSI(T6H, T6G), ms, &(x[WS(rs, 1)])); ST(&(x[WS(rs, 47)]), VFNMSI(T6F, T6C), ms, &(x[WS(rs, 1)])); ST(&(x[WS(rs, 17)]), VFMAI(T6F, T6C), ms, &(x[WS(rs, 1)])); ST(&(x[WS(rs, 1)]), VFMAI(T6z, T6y), ms, &(x[WS(rs, 1)])); ST(&(x[WS(rs, 63)]), VFNMSI(T6z, T6y), ms, &(x[WS(rs, 1)])); ST(&(x[WS(rs, 33)]), VFMAI(T6x, T6i), ms, &(x[WS(rs, 1)])); ST(&(x[WS(rs, 31)]), VFNMSI(T6x, T6i), ms, &(x[WS(rs, 1)])); T73 = VADD(T6W, T6X); T6Y = VSUB(T6W, T6X); T76 = VSUB(T6N, T6Q); T6R = VADD(T6N, T6Q); { V T78, T74, T71, T6Z, T79, T77, T70, T6S; T78 = VFNMS(LDK(KP773010453), T73, T72); T74 = VFMA(LDK(KP773010453), T73, T72); T71 = VFMA(LDK(KP773010453), T6Y, T6V); T6Z = VFNMS(LDK(KP773010453), T6Y, T6V); T79 = VFMA(LDK(KP773010453), T76, T75); T77 = VFNMS(LDK(KP773010453), T76, T75); T70 = VFMA(LDK(KP773010453), T6R, T6K); T6S = VFNMS(LDK(KP773010453), T6R, T6K); ST(&(x[WS(rs, 55)]), VFNMSI(T79, T78), ms, &(x[WS(rs, 1)])); ST(&(x[WS(rs, 9)]), VFMAI(T79, T78), ms, &(x[WS(rs, 1)])); ST(&(x[WS(rs, 41)]), VFMAI(T77, T74), ms, &(x[WS(rs, 1)])); ST(&(x[WS(rs, 23)]), VFNMSI(T77, T74), ms, &(x[WS(rs, 1)])); ST(&(x[WS(rs, 57)]), VFMAI(T71, T70), ms, &(x[WS(rs, 1)])); ST(&(x[WS(rs, 7)]), VFNMSI(T71, T70), ms, &(x[WS(rs, 1)])); ST(&(x[WS(rs, 25)]), VFMAI(T6Z, T6S), ms, &(x[WS(rs, 1)])); ST(&(x[WS(rs, 39)]), VFNMSI(T6Z, T6S), ms, &(x[WS(rs, 1)])); } } } } } VLEAVE(); } static const tw_instr twinstr[] = { VTW(0, 1), VTW(0, 2), VTW(0, 3), VTW(0, 4), VTW(0, 5), VTW(0, 6), VTW(0, 7), VTW(0, 8), VTW(0, 9), VTW(0, 10), VTW(0, 11), VTW(0, 12), VTW(0, 13), VTW(0, 14), VTW(0, 15), VTW(0, 16), VTW(0, 17), VTW(0, 18), VTW(0, 19), VTW(0, 20), VTW(0, 21), VTW(0, 22), VTW(0, 23), VTW(0, 24), VTW(0, 25), VTW(0, 26), VTW(0, 27), VTW(0, 28), VTW(0, 29), VTW(0, 30), VTW(0, 31), VTW(0, 32), VTW(0, 33), VTW(0, 34), VTW(0, 35), VTW(0, 36), VTW(0, 37), VTW(0, 38), VTW(0, 39), VTW(0, 40), VTW(0, 41), VTW(0, 42), VTW(0, 43), VTW(0, 44), VTW(0, 45), VTW(0, 46), VTW(0, 47), VTW(0, 48), VTW(0, 49), VTW(0, 50), VTW(0, 51), VTW(0, 52), VTW(0, 53), VTW(0, 54), VTW(0, 55), VTW(0, 56), VTW(0, 57), VTW(0, 58), VTW(0, 59), VTW(0, 60), VTW(0, 61), VTW(0, 62), VTW(0, 63), {TW_NEXT, VL, 0} }; static const ct_desc desc = { 64, XSIMD_STRING("t2bv_64"), twinstr, &GENUS, {261, 126, 258, 0}, 0, 0, 0 }; void XSIMD(codelet_t2bv_64) (planner *p) { X(kdft_dit_register) (p, t2bv_64, &desc); } #else /* HAVE_FMA */ /* Generated by: ../../../genfft/gen_twiddle_c.native -simd -compact -variables 4 -pipeline-latency 8 -n 64 -name t2bv_64 -include t2b.h -sign 1 */ /* * This function contains 519 FP additions, 250 FP multiplications, * (or, 467 additions, 198 multiplications, 52 fused multiply/add), * 107 stack variables, 15 constants, and 128 memory accesses */ #include "t2b.h" static void t2bv_64(R *ri, R *ii, const R *W, stride rs, INT mb, INT me, INT ms) { DVK(KP290284677, +0.290284677254462367636192375817395274691476278); DVK(KP956940335, +0.956940335732208864935797886980269969482849206); DVK(KP471396736, +0.471396736825997648556387625905254377657460319); DVK(KP881921264, +0.881921264348355029712756863660388349508442621); DVK(KP634393284, +0.634393284163645498215171613225493370675687095); DVK(KP773010453, +0.773010453362736960810906609758469800971041293); DVK(KP098017140, +0.098017140329560601994195563888641845861136673); DVK(KP995184726, +0.995184726672196886244836953109479921575474869); DVK(KP195090322, +0.195090322016128267848284868477022240927691618); DVK(KP980785280, +0.980785280403230449126182236134239036973933731); DVK(KP555570233, +0.555570233019602224742830813948532874374937191); DVK(KP831469612, +0.831469612302545237078788377617905756738560812); DVK(KP382683432, +0.382683432365089771728459984030398866761344562); DVK(KP923879532, +0.923879532511286756128183189396788286822416626); DVK(KP707106781, +0.707106781186547524400844362104849039284835938); { INT m; R *x; x = ii; for (m = mb, W = W + (mb * ((TWVL / VL) * 126)); m < me; m = m + VL, x = x + (VL * ms), W = W + (TWVL * 126), MAKE_VOLATILE_STRIDE(64, rs)) { V Tg, T4B, T6v, T7G, T3r, T4w, T5q, T7F, T5Y, T62, T28, T4d, T2g, T4a, T7g; V T7Y, T6f, T6j, T2Z, T4k, T37, T4h, T7n, T81, T7w, T7x, T7y, T5M, T6q, T1k; V T4s, T1r, T4t, T7t, T7u, T7v, T5F, T6p, TV, T4p, T12, T4q, T7A, T7B, TD; V T4x, T3k, T4C, T5x, T6s, T1R, T4b, T7j, T7Z, T2j, T4e, T5V, T63, T2I, T4i; V T7q, T82, T3a, T4l, T6c, T6k; { V T1, T3, T3p, T3n, Tb, Td, Te, T6, T8, T9, T2, T3o, T3m; T1 = LD(&(x[0]), ms, &(x[0])); T2 = LD(&(x[WS(rs, 32)]), ms, &(x[0])); T3 = BYTW(&(W[TWVL * 62]), T2); T3o = LD(&(x[WS(rs, 48)]), ms, &(x[0])); T3p = BYTW(&(W[TWVL * 94]), T3o); T3m = LD(&(x[WS(rs, 16)]), ms, &(x[0])); T3n = BYTW(&(W[TWVL * 30]), T3m); { V Ta, Tc, T5, T7; Ta = LD(&(x[WS(rs, 56)]), ms, &(x[0])); Tb = BYTW(&(W[TWVL * 110]), Ta); Tc = LD(&(x[WS(rs, 24)]), ms, &(x[0])); Td = BYTW(&(W[TWVL * 46]), Tc); Te = VSUB(Tb, Td); T5 = LD(&(x[WS(rs, 8)]), ms, &(x[0])); T6 = BYTW(&(W[TWVL * 14]), T5); T7 = LD(&(x[WS(rs, 40)]), ms, &(x[0])); T8 = BYTW(&(W[TWVL * 78]), T7); T9 = VSUB(T6, T8); } { V T4, Tf, T6t, T6u; T4 = VSUB(T1, T3); Tf = VMUL(LDK(KP707106781), VADD(T9, Te)); Tg = VSUB(T4, Tf); T4B = VADD(T4, Tf); T6t = VADD(T6, T8); T6u = VADD(Tb, Td); T6v = VSUB(T6t, T6u); T7G = VADD(T6t, T6u); } { V T3l, T3q, T5o, T5p; T3l = VMUL(LDK(KP707106781), VSUB(T9, Te)); T3q = VSUB(T3n, T3p); T3r = VSUB(T3l, T3q); T4w = VADD(T3q, T3l); T5o = VADD(T1, T3); T5p = VADD(T3n, T3p); T5q = VSUB(T5o, T5p); T7F = VADD(T5o, T5p); } } { V T24, T26, T61, T2b, T2d, T60, T1W, T5W, T21, T5X, T22, T27; { V T23, T25, T2a, T2c; T23 = LD(&(x[WS(rs, 17)]), ms, &(x[WS(rs, 1)])); T24 = BYTW(&(W[TWVL * 32]), T23); T25 = LD(&(x[WS(rs, 49)]), ms, &(x[WS(rs, 1)])); T26 = BYTW(&(W[TWVL * 96]), T25); T61 = VADD(T24, T26); T2a = LD(&(x[WS(rs, 1)]), ms, &(x[WS(rs, 1)])); T2b = BYTW(&(W[0]), T2a); T2c = LD(&(x[WS(rs, 33)]), ms, &(x[WS(rs, 1)])); T2d = BYTW(&(W[TWVL * 64]), T2c); T60 = VADD(T2b, T2d); } { V T1T, T1V, T1S, T1U; T1S = LD(&(x[WS(rs, 9)]), ms, &(x[WS(rs, 1)])); T1T = BYTW(&(W[TWVL * 16]), T1S); T1U = LD(&(x[WS(rs, 41)]), ms, &(x[WS(rs, 1)])); T1V = BYTW(&(W[TWVL * 80]), T1U); T1W = VSUB(T1T, T1V); T5W = VADD(T1T, T1V); } { V T1Y, T20, T1X, T1Z; T1X = LD(&(x[WS(rs, 57)]), ms, &(x[WS(rs, 1)])); T1Y = BYTW(&(W[TWVL * 112]), T1X); T1Z = LD(&(x[WS(rs, 25)]), ms, &(x[WS(rs, 1)])); T20 = BYTW(&(W[TWVL * 48]), T1Z); T21 = VSUB(T1Y, T20); T5X = VADD(T1Y, T20); } T5Y = VSUB(T5W, T5X); T62 = VSUB(T60, T61); T22 = VMUL(LDK(KP707106781), VSUB(T1W, T21)); T27 = VSUB(T24, T26); T28 = VSUB(T22, T27); T4d = VADD(T27, T22); { V T2e, T2f, T7e, T7f; T2e = VSUB(T2b, T2d); T2f = VMUL(LDK(KP707106781), VADD(T1W, T21)); T2g = VSUB(T2e, T2f); T4a = VADD(T2e, T2f); T7e = VADD(T60, T61); T7f = VADD(T5W, T5X); T7g = VSUB(T7e, T7f); T7Y = VADD(T7e, T7f); } } { V T2V, T2X, T6i, T32, T34, T6h, T2N, T6d, T2S, T6e, T2T, T2Y; { V T2U, T2W, T31, T33; T2U = LD(&(x[WS(rs, 15)]), ms, &(x[WS(rs, 1)])); T2V = BYTW(&(W[TWVL * 28]), T2U); T2W = LD(&(x[WS(rs, 47)]), ms, &(x[WS(rs, 1)])); T2X = BYTW(&(W[TWVL * 92]), T2W); T6i = VADD(T2V, T2X); T31 = LD(&(x[WS(rs, 63)]), ms, &(x[WS(rs, 1)])); T32 = BYTW(&(W[TWVL * 124]), T31); T33 = LD(&(x[WS(rs, 31)]), ms, &(x[WS(rs, 1)])); T34 = BYTW(&(W[TWVL * 60]), T33); T6h = VADD(T32, T34); } { V T2K, T2M, T2J, T2L; T2J = LD(&(x[WS(rs, 7)]), ms, &(x[WS(rs, 1)])); T2K = BYTW(&(W[TWVL * 12]), T2J); T2L = LD(&(x[WS(rs, 39)]), ms, &(x[WS(rs, 1)])); T2M = BYTW(&(W[TWVL * 76]), T2L); T2N = VSUB(T2K, T2M); T6d = VADD(T2K, T2M); } { V T2P, T2R, T2O, T2Q; T2O = LD(&(x[WS(rs, 55)]), ms, &(x[WS(rs, 1)])); T2P = BYTW(&(W[TWVL * 108]), T2O); T2Q = LD(&(x[WS(rs, 23)]), ms, &(x[WS(rs, 1)])); T2R = BYTW(&(W[TWVL * 44]), T2Q); T2S = VSUB(T2P, T2R); T6e = VADD(T2P, T2R); } T6f = VSUB(T6d, T6e); T6j = VSUB(T6h, T6i); T2T = VMUL(LDK(KP707106781), VSUB(T2N, T2S)); T2Y = VSUB(T2V, T2X); T2Z = VSUB(T2T, T2Y); T4k = VADD(T2Y, T2T); { V T35, T36, T7l, T7m; T35 = VSUB(T32, T34); T36 = VMUL(LDK(KP707106781), VADD(T2N, T2S)); T37 = VSUB(T35, T36); T4h = VADD(T35, T36); T7l = VADD(T6h, T6i); T7m = VADD(T6d, T6e); T7n = VSUB(T7l, T7m); T81 = VADD(T7l, T7m); } } { V T1g, T1i, T5K, T1m, T1o, T5J, T18, T5G, T1d, T5H, T5I, T5L; { V T1f, T1h, T1l, T1n; T1f = LD(&(x[WS(rs, 14)]), ms, &(x[0])); T1g = BYTW(&(W[TWVL * 26]), T1f); T1h = LD(&(x[WS(rs, 46)]), ms, &(x[0])); T1i = BYTW(&(W[TWVL * 90]), T1h); T5K = VADD(T1g, T1i); T1l = LD(&(x[WS(rs, 62)]), ms, &(x[0])); T1m = BYTW(&(W[TWVL * 122]), T1l); T1n = LD(&(x[WS(rs, 30)]), ms, &(x[0])); T1o = BYTW(&(W[TWVL * 58]), T1n); T5J = VADD(T1m, T1o); } { V T15, T17, T14, T16; T14 = LD(&(x[WS(rs, 6)]), ms, &(x[0])); T15 = BYTW(&(W[TWVL * 10]), T14); T16 = LD(&(x[WS(rs, 38)]), ms, &(x[0])); T17 = BYTW(&(W[TWVL * 74]), T16); T18 = VSUB(T15, T17); T5G = VADD(T15, T17); } { V T1a, T1c, T19, T1b; T19 = LD(&(x[WS(rs, 54)]), ms, &(x[0])); T1a = BYTW(&(W[TWVL * 106]), T19); T1b = LD(&(x[WS(rs, 22)]), ms, &(x[0])); T1c = BYTW(&(W[TWVL * 42]), T1b); T1d = VSUB(T1a, T1c); T5H = VADD(T1a, T1c); } T7w = VADD(T5J, T5K); T7x = VADD(T5G, T5H); T7y = VSUB(T7w, T7x); T5I = VSUB(T5G, T5H); T5L = VSUB(T5J, T5K); T5M = VFNMS(LDK(KP382683432), T5L, VMUL(LDK(KP923879532), T5I)); T6q = VFMA(LDK(KP923879532), T5L, VMUL(LDK(KP382683432), T5I)); { V T1e, T1j, T1p, T1q; T1e = VMUL(LDK(KP707106781), VSUB(T18, T1d)); T1j = VSUB(T1g, T1i); T1k = VSUB(T1e, T1j); T4s = VADD(T1j, T1e); T1p = VSUB(T1m, T1o); T1q = VMUL(LDK(KP707106781), VADD(T18, T1d)); T1r = VSUB(T1p, T1q); T4t = VADD(T1p, T1q); } } { V TR, TT, T5A, TX, TZ, T5z, TJ, T5C, TO, T5D, T5B, T5E; { V TQ, TS, TW, TY; TQ = LD(&(x[WS(rs, 18)]), ms, &(x[0])); TR = BYTW(&(W[TWVL * 34]), TQ); TS = LD(&(x[WS(rs, 50)]), ms, &(x[0])); TT = BYTW(&(W[TWVL * 98]), TS); T5A = VADD(TR, TT); TW = LD(&(x[WS(rs, 2)]), ms, &(x[0])); TX = BYTW(&(W[TWVL * 2]), TW); TY = LD(&(x[WS(rs, 34)]), ms, &(x[0])); TZ = BYTW(&(W[TWVL * 66]), TY); T5z = VADD(TX, TZ); } { V TG, TI, TF, TH; TF = LD(&(x[WS(rs, 10)]), ms, &(x[0])); TG = BYTW(&(W[TWVL * 18]), TF); TH = LD(&(x[WS(rs, 42)]), ms, &(x[0])); TI = BYTW(&(W[TWVL * 82]), TH); TJ = VSUB(TG, TI); T5C = VADD(TG, TI); } { V TL, TN, TK, TM; TK = LD(&(x[WS(rs, 58)]), ms, &(x[0])); TL = BYTW(&(W[TWVL * 114]), TK); TM = LD(&(x[WS(rs, 26)]), ms, &(x[0])); TN = BYTW(&(W[TWVL * 50]), TM); TO = VSUB(TL, TN); T5D = VADD(TL, TN); } T7t = VADD(T5z, T5A); T7u = VADD(T5C, T5D); T7v = VSUB(T7t, T7u); T5B = VSUB(T5z, T5A); T5E = VSUB(T5C, T5D); T5F = VFMA(LDK(KP382683432), T5B, VMUL(LDK(KP923879532), T5E)); T6p = VFNMS(LDK(KP382683432), T5E, VMUL(LDK(KP923879532), T5B)); { V TP, TU, T10, T11; TP = VMUL(LDK(KP707106781), VSUB(TJ, TO)); TU = VSUB(TR, TT); TV = VSUB(TP, TU); T4p = VADD(TU, TP); T10 = VSUB(TX, TZ); T11 = VMUL(LDK(KP707106781), VADD(TJ, TO)); T12 = VSUB(T10, T11); T4q = VADD(T10, T11); } } { V Tl, T5r, TB, T5u, Tq, T5s, Tw, T5v, Tr, TC; { V Ti, Tk, Th, Tj; Th = LD(&(x[WS(rs, 4)]), ms, &(x[0])); Ti = BYTW(&(W[TWVL * 6]), Th); Tj = LD(&(x[WS(rs, 36)]), ms, &(x[0])); Tk = BYTW(&(W[TWVL * 70]), Tj); Tl = VSUB(Ti, Tk); T5r = VADD(Ti, Tk); } { V Ty, TA, Tx, Tz; Tx = LD(&(x[WS(rs, 60)]), ms, &(x[0])); Ty = BYTW(&(W[TWVL * 118]), Tx); Tz = LD(&(x[WS(rs, 28)]), ms, &(x[0])); TA = BYTW(&(W[TWVL * 54]), Tz); TB = VSUB(Ty, TA); T5u = VADD(Ty, TA); } { V Tn, Tp, Tm, To; Tm = LD(&(x[WS(rs, 20)]), ms, &(x[0])); Tn = BYTW(&(W[TWVL * 38]), Tm); To = LD(&(x[WS(rs, 52)]), ms, &(x[0])); Tp = BYTW(&(W[TWVL * 102]), To); Tq = VSUB(Tn, Tp); T5s = VADD(Tn, Tp); } { V Tt, Tv, Ts, Tu; Ts = LD(&(x[WS(rs, 12)]), ms, &(x[0])); Tt = BYTW(&(W[TWVL * 22]), Ts); Tu = LD(&(x[WS(rs, 44)]), ms, &(x[0])); Tv = BYTW(&(W[TWVL * 86]), Tu); Tw = VSUB(Tt, Tv); T5v = VADD(Tt, Tv); } T7A = VADD(T5r, T5s); T7B = VADD(T5u, T5v); Tr = VFMA(LDK(KP382683432), Tl, VMUL(LDK(KP923879532), Tq)); TC = VFNMS(LDK(KP382683432), TB, VMUL(LDK(KP923879532), Tw)); TD = VSUB(Tr, TC); T4x = VADD(Tr, TC); { V T3i, T3j, T5t, T5w; T3i = VFNMS(LDK(KP382683432), Tq, VMUL(LDK(KP923879532), Tl)); T3j = VFMA(LDK(KP923879532), TB, VMUL(LDK(KP382683432), Tw)); T3k = VSUB(T3i, T3j); T4C = VADD(T3i, T3j); T5t = VSUB(T5r, T5s); T5w = VSUB(T5u, T5v); T5x = VMUL(LDK(KP707106781), VADD(T5t, T5w)); T6s = VMUL(LDK(KP707106781), VSUB(T5t, T5w)); } } { V T1z, T5P, T1P, T5T, T1E, T5Q, T1K, T5S; { V T1w, T1y, T1v, T1x; T1v = LD(&(x[WS(rs, 5)]), ms, &(x[WS(rs, 1)])); T1w = BYTW(&(W[TWVL * 8]), T1v); T1x = LD(&(x[WS(rs, 37)]), ms, &(x[WS(rs, 1)])); T1y = BYTW(&(W[TWVL * 72]), T1x); T1z = VSUB(T1w, T1y); T5P = VADD(T1w, T1y); } { V T1M, T1O, T1L, T1N; T1L = LD(&(x[WS(rs, 13)]), ms, &(x[WS(rs, 1)])); T1M = BYTW(&(W[TWVL * 24]), T1L); T1N = LD(&(x[WS(rs, 45)]), ms, &(x[WS(rs, 1)])); T1O = BYTW(&(W[TWVL * 88]), T1N); T1P = VSUB(T1M, T1O); T5T = VADD(T1M, T1O); } { V T1B, T1D, T1A, T1C; T1A = LD(&(x[WS(rs, 21)]), ms, &(x[WS(rs, 1)])); T1B = BYTW(&(W[TWVL * 40]), T1A); T1C = LD(&(x[WS(rs, 53)]), ms, &(x[WS(rs, 1)])); T1D = BYTW(&(W[TWVL * 104]), T1C); T1E = VSUB(T1B, T1D); T5Q = VADD(T1B, T1D); } { V T1H, T1J, T1G, T1I; T1G = LD(&(x[WS(rs, 61)]), ms, &(x[WS(rs, 1)])); T1H = BYTW(&(W[TWVL * 120]), T1G); T1I = LD(&(x[WS(rs, 29)]), ms, &(x[WS(rs, 1)])); T1J = BYTW(&(W[TWVL * 56]), T1I); T1K = VSUB(T1H, T1J); T5S = VADD(T1H, T1J); } { V T1F, T1Q, T7h, T7i; T1F = VFNMS(LDK(KP382683432), T1E, VMUL(LDK(KP923879532), T1z)); T1Q = VFMA(LDK(KP923879532), T1K, VMUL(LDK(KP382683432), T1P)); T1R = VSUB(T1F, T1Q); T4b = VADD(T1F, T1Q); T7h = VADD(T5P, T5Q); T7i = VADD(T5S, T5T); T7j = VSUB(T7h, T7i); T7Z = VADD(T7h, T7i); } { V T2h, T2i, T5R, T5U; T2h = VFMA(LDK(KP382683432), T1z, VMUL(LDK(KP923879532), T1E)); T2i = VFNMS(LDK(KP382683432), T1K, VMUL(LDK(KP923879532), T1P)); T2j = VSUB(T2h, T2i); T4e = VADD(T2h, T2i); T5R = VSUB(T5P, T5Q); T5U = VSUB(T5S, T5T); T5V = VMUL(LDK(KP707106781), VSUB(T5R, T5U)); T63 = VMUL(LDK(KP707106781), VADD(T5R, T5U)); } } { V T2q, T66, T2G, T6a, T2v, T67, T2B, T69; { V T2n, T2p, T2m, T2o; T2m = LD(&(x[WS(rs, 3)]), ms, &(x[WS(rs, 1)])); T2n = BYTW(&(W[TWVL * 4]), T2m); T2o = LD(&(x[WS(rs, 35)]), ms, &(x[WS(rs, 1)])); T2p = BYTW(&(W[TWVL * 68]), T2o); T2q = VSUB(T2n, T2p); T66 = VADD(T2n, T2p); } { V T2D, T2F, T2C, T2E; T2C = LD(&(x[WS(rs, 11)]), ms, &(x[WS(rs, 1)])); T2D = BYTW(&(W[TWVL * 20]), T2C); T2E = LD(&(x[WS(rs, 43)]), ms, &(x[WS(rs, 1)])); T2F = BYTW(&(W[TWVL * 84]), T2E); T2G = VSUB(T2D, T2F); T6a = VADD(T2D, T2F); } { V T2s, T2u, T2r, T2t; T2r = LD(&(x[WS(rs, 19)]), ms, &(x[WS(rs, 1)])); T2s = BYTW(&(W[TWVL * 36]), T2r); T2t = LD(&(x[WS(rs, 51)]), ms, &(x[WS(rs, 1)])); T2u = BYTW(&(W[TWVL * 100]), T2t); T2v = VSUB(T2s, T2u); T67 = VADD(T2s, T2u); } { V T2y, T2A, T2x, T2z; T2x = LD(&(x[WS(rs, 59)]), ms, &(x[WS(rs, 1)])); T2y = BYTW(&(W[TWVL * 116]), T2x); T2z = LD(&(x[WS(rs, 27)]), ms, &(x[WS(rs, 1)])); T2A = BYTW(&(W[TWVL * 52]), T2z); T2B = VSUB(T2y, T2A); T69 = VADD(T2y, T2A); } { V T2w, T2H, T7o, T7p; T2w = VFNMS(LDK(KP382683432), T2v, VMUL(LDK(KP923879532), T2q)); T2H = VFMA(LDK(KP923879532), T2B, VMUL(LDK(KP382683432), T2G)); T2I = VSUB(T2w, T2H); T4i = VADD(T2w, T2H); T7o = VADD(T66, T67); T7p = VADD(T69, T6a); T7q = VSUB(T7o, T7p); T82 = VADD(T7o, T7p); } { V T38, T39, T68, T6b; T38 = VFMA(LDK(KP382683432), T2q, VMUL(LDK(KP923879532), T2v)); T39 = VFNMS(LDK(KP382683432), T2B, VMUL(LDK(KP923879532), T2G)); T3a = VSUB(T38, T39); T4l = VADD(T38, T39); T68 = VSUB(T66, T67); T6b = VSUB(T69, T6a); T6c = VMUL(LDK(KP707106781), VSUB(T68, T6b)); T6k = VMUL(LDK(KP707106781), VADD(T68, T6b)); } } { V T7s, T7R, T7M, T7U, T7D, T7T, T7J, T7Q; { V T7k, T7r, T7K, T7L; T7k = VFNMS(LDK(KP382683432), T7j, VMUL(LDK(KP923879532), T7g)); T7r = VFMA(LDK(KP923879532), T7n, VMUL(LDK(KP382683432), T7q)); T7s = VSUB(T7k, T7r); T7R = VADD(T7k, T7r); T7K = VFMA(LDK(KP382683432), T7g, VMUL(LDK(KP923879532), T7j)); T7L = VFNMS(LDK(KP382683432), T7n, VMUL(LDK(KP923879532), T7q)); T7M = VSUB(T7K, T7L); T7U = VADD(T7K, T7L); } { V T7z, T7C, T7H, T7I; T7z = VMUL(LDK(KP707106781), VSUB(T7v, T7y)); T7C = VSUB(T7A, T7B); T7D = VSUB(T7z, T7C); T7T = VADD(T7C, T7z); T7H = VSUB(T7F, T7G); T7I = VMUL(LDK(KP707106781), VADD(T7v, T7y)); T7J = VSUB(T7H, T7I); T7Q = VADD(T7H, T7I); } { V T7E, T7N, T7W, T7X; T7E = VBYI(VSUB(T7s, T7D)); T7N = VSUB(T7J, T7M); ST(&(x[WS(rs, 20)]), VADD(T7E, T7N), ms, &(x[0])); ST(&(x[WS(rs, 44)]), VSUB(T7N, T7E), ms, &(x[0])); T7W = VSUB(T7Q, T7R); T7X = VBYI(VSUB(T7U, T7T)); ST(&(x[WS(rs, 36)]), VSUB(T7W, T7X), ms, &(x[0])); ST(&(x[WS(rs, 28)]), VADD(T7W, T7X), ms, &(x[0])); } { V T7O, T7P, T7S, T7V; T7O = VBYI(VADD(T7D, T7s)); T7P = VADD(T7J, T7M); ST(&(x[WS(rs, 12)]), VADD(T7O, T7P), ms, &(x[0])); ST(&(x[WS(rs, 52)]), VSUB(T7P, T7O), ms, &(x[0])); T7S = VADD(T7Q, T7R); T7V = VBYI(VADD(T7T, T7U)); ST(&(x[WS(rs, 60)]), VSUB(T7S, T7V), ms, &(x[0])); ST(&(x[WS(rs, 4)]), VADD(T7S, T7V), ms, &(x[0])); } } { V T84, T8c, T8l, T8n, T87, T8h, T8b, T8g, T8i, T8m; { V T80, T83, T8j, T8k; T80 = VSUB(T7Y, T7Z); T83 = VSUB(T81, T82); T84 = VMUL(LDK(KP707106781), VSUB(T80, T83)); T8c = VMUL(LDK(KP707106781), VADD(T80, T83)); T8j = VADD(T7Y, T7Z); T8k = VADD(T81, T82); T8l = VBYI(VSUB(T8j, T8k)); T8n = VADD(T8j, T8k); } { V T85, T86, T89, T8a; T85 = VADD(T7t, T7u); T86 = VADD(T7w, T7x); T87 = VSUB(T85, T86); T8h = VADD(T85, T86); T89 = VADD(T7F, T7G); T8a = VADD(T7A, T7B); T8b = VSUB(T89, T8a); T8g = VADD(T89, T8a); } T8i = VSUB(T8g, T8h); ST(&(x[WS(rs, 48)]), VSUB(T8i, T8l), ms, &(x[0])); ST(&(x[WS(rs, 16)]), VADD(T8i, T8l), ms, &(x[0])); T8m = VADD(T8g, T8h); ST(&(x[WS(rs, 32)]), VSUB(T8m, T8n), ms, &(x[0])); ST(&(x[0]), VADD(T8m, T8n), ms, &(x[0])); { V T88, T8d, T8e, T8f; T88 = VBYI(VSUB(T84, T87)); T8d = VSUB(T8b, T8c); ST(&(x[WS(rs, 24)]), VADD(T88, T8d), ms, &(x[0])); ST(&(x[WS(rs, 40)]), VSUB(T8d, T88), ms, &(x[0])); T8e = VBYI(VADD(T87, T84)); T8f = VADD(T8b, T8c); ST(&(x[WS(rs, 8)]), VADD(T8e, T8f), ms, &(x[0])); ST(&(x[WS(rs, 56)]), VSUB(T8f, T8e), ms, &(x[0])); } } { V T5O, T6H, T6x, T6F, T6n, T6I, T6A, T6E; { V T5y, T5N, T6r, T6w; T5y = VSUB(T5q, T5x); T5N = VSUB(T5F, T5M); T5O = VSUB(T5y, T5N); T6H = VADD(T5y, T5N); T6r = VSUB(T6p, T6q); T6w = VSUB(T6s, T6v); T6x = VSUB(T6r, T6w); T6F = VADD(T6w, T6r); { V T65, T6y, T6m, T6z; { V T5Z, T64, T6g, T6l; T5Z = VSUB(T5V, T5Y); T64 = VSUB(T62, T63); T65 = VFMA(LDK(KP831469612), T5Z, VMUL(LDK(KP555570233), T64)); T6y = VFNMS(LDK(KP555570233), T5Z, VMUL(LDK(KP831469612), T64)); T6g = VSUB(T6c, T6f); T6l = VSUB(T6j, T6k); T6m = VFNMS(LDK(KP555570233), T6l, VMUL(LDK(KP831469612), T6g)); T6z = VFMA(LDK(KP555570233), T6g, VMUL(LDK(KP831469612), T6l)); } T6n = VSUB(T65, T6m); T6I = VADD(T6y, T6z); T6A = VSUB(T6y, T6z); T6E = VADD(T65, T6m); } } { V T6o, T6B, T6K, T6L; T6o = VADD(T5O, T6n); T6B = VBYI(VADD(T6x, T6A)); ST(&(x[WS(rs, 54)]), VSUB(T6o, T6B), ms, &(x[0])); ST(&(x[WS(rs, 10)]), VADD(T6o, T6B), ms, &(x[0])); T6K = VBYI(VADD(T6F, T6E)); T6L = VADD(T6H, T6I); ST(&(x[WS(rs, 6)]), VADD(T6K, T6L), ms, &(x[0])); ST(&(x[WS(rs, 58)]), VSUB(T6L, T6K), ms, &(x[0])); } { V T6C, T6D, T6G, T6J; T6C = VSUB(T5O, T6n); T6D = VBYI(VSUB(T6A, T6x)); ST(&(x[WS(rs, 42)]), VSUB(T6C, T6D), ms, &(x[0])); ST(&(x[WS(rs, 22)]), VADD(T6C, T6D), ms, &(x[0])); T6G = VBYI(VSUB(T6E, T6F)); T6J = VSUB(T6H, T6I); ST(&(x[WS(rs, 26)]), VADD(T6G, T6J), ms, &(x[0])); ST(&(x[WS(rs, 38)]), VSUB(T6J, T6G), ms, &(x[0])); } } { V T6O, T79, T6Z, T77, T6V, T7a, T72, T76; { V T6M, T6N, T6X, T6Y; T6M = VADD(T5q, T5x); T6N = VADD(T6p, T6q); T6O = VSUB(T6M, T6N); T79 = VADD(T6M, T6N); T6X = VADD(T5F, T5M); T6Y = VADD(T6v, T6s); T6Z = VSUB(T6X, T6Y); T77 = VADD(T6Y, T6X); { V T6R, T70, T6U, T71; { V T6P, T6Q, T6S, T6T; T6P = VADD(T5Y, T5V); T6Q = VADD(T62, T63); T6R = VFMA(LDK(KP980785280), T6P, VMUL(LDK(KP195090322), T6Q)); T70 = VFNMS(LDK(KP195090322), T6P, VMUL(LDK(KP980785280), T6Q)); T6S = VADD(T6f, T6c); T6T = VADD(T6j, T6k); T6U = VFNMS(LDK(KP195090322), T6T, VMUL(LDK(KP980785280), T6S)); T71 = VFMA(LDK(KP195090322), T6S, VMUL(LDK(KP980785280), T6T)); } T6V = VSUB(T6R, T6U); T7a = VADD(T70, T71); T72 = VSUB(T70, T71); T76 = VADD(T6R, T6U); } } { V T6W, T73, T7c, T7d; T6W = VADD(T6O, T6V); T73 = VBYI(VADD(T6Z, T72)); ST(&(x[WS(rs, 50)]), VSUB(T6W, T73), ms, &(x[0])); ST(&(x[WS(rs, 14)]), VADD(T6W, T73), ms, &(x[0])); T7c = VBYI(VADD(T77, T76)); T7d = VADD(T79, T7a); ST(&(x[WS(rs, 2)]), VADD(T7c, T7d), ms, &(x[0])); ST(&(x[WS(rs, 62)]), VSUB(T7d, T7c), ms, &(x[0])); } { V T74, T75, T78, T7b; T74 = VSUB(T6O, T6V); T75 = VBYI(VSUB(T72, T6Z)); ST(&(x[WS(rs, 46)]), VSUB(T74, T75), ms, &(x[0])); ST(&(x[WS(rs, 18)]), VADD(T74, T75), ms, &(x[0])); T78 = VBYI(VSUB(T76, T77)); T7b = VSUB(T79, T7a); ST(&(x[WS(rs, 30)]), VADD(T78, T7b), ms, &(x[0])); ST(&(x[WS(rs, 34)]), VSUB(T7b, T78), ms, &(x[0])); } } { V T4z, T5g, T4R, T59, T4H, T5j, T4O, T55, T4o, T4S, T4K, T4P, T52, T5k, T5c; V T5h; { V T4y, T57, T4v, T58, T4r, T4u; T4y = VADD(T4w, T4x); T57 = VSUB(T4B, T4C); T4r = VFMA(LDK(KP980785280), T4p, VMUL(LDK(KP195090322), T4q)); T4u = VFNMS(LDK(KP195090322), T4t, VMUL(LDK(KP980785280), T4s)); T4v = VADD(T4r, T4u); T58 = VSUB(T4r, T4u); T4z = VSUB(T4v, T4y); T5g = VADD(T57, T58); T4R = VADD(T4y, T4v); T59 = VSUB(T57, T58); } { V T4D, T54, T4G, T53, T4E, T4F; T4D = VADD(T4B, T4C); T54 = VSUB(T4x, T4w); T4E = VFNMS(LDK(KP195090322), T4p, VMUL(LDK(KP980785280), T4q)); T4F = VFMA(LDK(KP195090322), T4s, VMUL(LDK(KP980785280), T4t)); T4G = VADD(T4E, T4F); T53 = VSUB(T4E, T4F); T4H = VSUB(T4D, T4G); T5j = VADD(T54, T53); T4O = VADD(T4D, T4G); T55 = VSUB(T53, T54); } { V T4g, T4I, T4n, T4J; { V T4c, T4f, T4j, T4m; T4c = VADD(T4a, T4b); T4f = VADD(T4d, T4e); T4g = VFNMS(LDK(KP098017140), T4f, VMUL(LDK(KP995184726), T4c)); T4I = VFMA(LDK(KP098017140), T4c, VMUL(LDK(KP995184726), T4f)); T4j = VADD(T4h, T4i); T4m = VADD(T4k, T4l); T4n = VFMA(LDK(KP995184726), T4j, VMUL(LDK(KP098017140), T4m)); T4J = VFNMS(LDK(KP098017140), T4j, VMUL(LDK(KP995184726), T4m)); } T4o = VSUB(T4g, T4n); T4S = VADD(T4I, T4J); T4K = VSUB(T4I, T4J); T4P = VADD(T4g, T4n); } { V T4Y, T5a, T51, T5b; { V T4W, T4X, T4Z, T50; T4W = VSUB(T4a, T4b); T4X = VSUB(T4e, T4d); T4Y = VFNMS(LDK(KP634393284), T4X, VMUL(LDK(KP773010453), T4W)); T5a = VFMA(LDK(KP634393284), T4W, VMUL(LDK(KP773010453), T4X)); T4Z = VSUB(T4h, T4i); T50 = VSUB(T4l, T4k); T51 = VFMA(LDK(KP773010453), T4Z, VMUL(LDK(KP634393284), T50)); T5b = VFNMS(LDK(KP634393284), T4Z, VMUL(LDK(KP773010453), T50)); } T52 = VSUB(T4Y, T51); T5k = VADD(T5a, T5b); T5c = VSUB(T5a, T5b); T5h = VADD(T4Y, T51); } { V T4A, T4L, T5i, T5l; T4A = VBYI(VSUB(T4o, T4z)); T4L = VSUB(T4H, T4K); ST(&(x[WS(rs, 17)]), VADD(T4A, T4L), ms, &(x[WS(rs, 1)])); ST(&(x[WS(rs, 47)]), VSUB(T4L, T4A), ms, &(x[WS(rs, 1)])); T5i = VADD(T5g, T5h); T5l = VBYI(VADD(T5j, T5k)); ST(&(x[WS(rs, 57)]), VSUB(T5i, T5l), ms, &(x[WS(rs, 1)])); ST(&(x[WS(rs, 7)]), VADD(T5i, T5l), ms, &(x[WS(rs, 1)])); } { V T5m, T5n, T4M, T4N; T5m = VSUB(T5g, T5h); T5n = VBYI(VSUB(T5k, T5j)); ST(&(x[WS(rs, 39)]), VSUB(T5m, T5n), ms, &(x[WS(rs, 1)])); ST(&(x[WS(rs, 25)]), VADD(T5m, T5n), ms, &(x[WS(rs, 1)])); T4M = VBYI(VADD(T4z, T4o)); T4N = VADD(T4H, T4K); ST(&(x[WS(rs, 15)]), VADD(T4M, T4N), ms, &(x[WS(rs, 1)])); ST(&(x[WS(rs, 49)]), VSUB(T4N, T4M), ms, &(x[WS(rs, 1)])); } { V T4Q, T4T, T56, T5d; T4Q = VADD(T4O, T4P); T4T = VBYI(VADD(T4R, T4S)); ST(&(x[WS(rs, 63)]), VSUB(T4Q, T4T), ms, &(x[WS(rs, 1)])); ST(&(x[WS(rs, 1)]), VADD(T4Q, T4T), ms, &(x[WS(rs, 1)])); T56 = VBYI(VSUB(T52, T55)); T5d = VSUB(T59, T5c); ST(&(x[WS(rs, 23)]), VADD(T56, T5d), ms, &(x[WS(rs, 1)])); ST(&(x[WS(rs, 41)]), VSUB(T5d, T56), ms, &(x[WS(rs, 1)])); } { V T5e, T5f, T4U, T4V; T5e = VBYI(VADD(T55, T52)); T5f = VADD(T59, T5c); ST(&(x[WS(rs, 9)]), VADD(T5e, T5f), ms, &(x[WS(rs, 1)])); ST(&(x[WS(rs, 55)]), VSUB(T5f, T5e), ms, &(x[WS(rs, 1)])); T4U = VSUB(T4O, T4P); T4V = VBYI(VSUB(T4S, T4R)); ST(&(x[WS(rs, 33)]), VSUB(T4U, T4V), ms, &(x[WS(rs, 1)])); ST(&(x[WS(rs, 31)]), VADD(T4U, T4V), ms, &(x[WS(rs, 1)])); } } { V T1u, T43, T3D, T3V, T3t, T45, T3B, T3K, T3d, T3E, T3w, T3A, T3R, T46, T3Y; V T42; { V TE, T3U, T1t, T3T, T13, T1s; TE = VSUB(Tg, TD); T3U = VADD(T3r, T3k); T13 = VFMA(LDK(KP831469612), TV, VMUL(LDK(KP555570233), T12)); T1s = VFNMS(LDK(KP555570233), T1r, VMUL(LDK(KP831469612), T1k)); T1t = VSUB(T13, T1s); T3T = VADD(T13, T1s); T1u = VSUB(TE, T1t); T43 = VADD(T3U, T3T); T3D = VADD(TE, T1t); T3V = VSUB(T3T, T3U); } { V T3s, T3I, T3h, T3J, T3f, T3g; T3s = VSUB(T3k, T3r); T3I = VADD(Tg, TD); T3f = VFNMS(LDK(KP555570233), TV, VMUL(LDK(KP831469612), T12)); T3g = VFMA(LDK(KP555570233), T1k, VMUL(LDK(KP831469612), T1r)); T3h = VSUB(T3f, T3g); T3J = VADD(T3f, T3g); T3t = VSUB(T3h, T3s); T45 = VADD(T3I, T3J); T3B = VADD(T3s, T3h); T3K = VSUB(T3I, T3J); } { V T2l, T3u, T3c, T3v; { V T29, T2k, T30, T3b; T29 = VSUB(T1R, T28); T2k = VSUB(T2g, T2j); T2l = VFMA(LDK(KP881921264), T29, VMUL(LDK(KP471396736), T2k)); T3u = VFNMS(LDK(KP471396736), T29, VMUL(LDK(KP881921264), T2k)); T30 = VSUB(T2I, T2Z); T3b = VSUB(T37, T3a); T3c = VFNMS(LDK(KP471396736), T3b, VMUL(LDK(KP881921264), T30)); T3v = VFMA(LDK(KP471396736), T30, VMUL(LDK(KP881921264), T3b)); } T3d = VSUB(T2l, T3c); T3E = VADD(T3u, T3v); T3w = VSUB(T3u, T3v); T3A = VADD(T2l, T3c); } { V T3N, T3W, T3Q, T3X; { V T3L, T3M, T3O, T3P; T3L = VADD(T28, T1R); T3M = VADD(T2g, T2j); T3N = VFMA(LDK(KP956940335), T3L, VMUL(LDK(KP290284677), T3M)); T3W = VFNMS(LDK(KP290284677), T3L, VMUL(LDK(KP956940335), T3M)); T3O = VADD(T2Z, T2I); T3P = VADD(T37, T3a); T3Q = VFNMS(LDK(KP290284677), T3P, VMUL(LDK(KP956940335), T3O)); T3X = VFMA(LDK(KP290284677), T3O, VMUL(LDK(KP956940335), T3P)); } T3R = VSUB(T3N, T3Q); T46 = VADD(T3W, T3X); T3Y = VSUB(T3W, T3X); T42 = VADD(T3N, T3Q); } { V T3e, T3x, T44, T47; T3e = VADD(T1u, T3d); T3x = VBYI(VADD(T3t, T3w)); ST(&(x[WS(rs, 53)]), VSUB(T3e, T3x), ms, &(x[WS(rs, 1)])); ST(&(x[WS(rs, 11)]), VADD(T3e, T3x), ms, &(x[WS(rs, 1)])); T44 = VBYI(VSUB(T42, T43)); T47 = VSUB(T45, T46); ST(&(x[WS(rs, 29)]), VADD(T44, T47), ms, &(x[WS(rs, 1)])); ST(&(x[WS(rs, 35)]), VSUB(T47, T44), ms, &(x[WS(rs, 1)])); } { V T48, T49, T3y, T3z; T48 = VBYI(VADD(T43, T42)); T49 = VADD(T45, T46); ST(&(x[WS(rs, 3)]), VADD(T48, T49), ms, &(x[WS(rs, 1)])); ST(&(x[WS(rs, 61)]), VSUB(T49, T48), ms, &(x[WS(rs, 1)])); T3y = VSUB(T1u, T3d); T3z = VBYI(VSUB(T3w, T3t)); ST(&(x[WS(rs, 43)]), VSUB(T3y, T3z), ms, &(x[WS(rs, 1)])); ST(&(x[WS(rs, 21)]), VADD(T3y, T3z), ms, &(x[WS(rs, 1)])); } { V T3C, T3F, T3S, T3Z; T3C = VBYI(VSUB(T3A, T3B)); T3F = VSUB(T3D, T3E); ST(&(x[WS(rs, 27)]), VADD(T3C, T3F), ms, &(x[WS(rs, 1)])); ST(&(x[WS(rs, 37)]), VSUB(T3F, T3C), ms, &(x[WS(rs, 1)])); T3S = VADD(T3K, T3R); T3Z = VBYI(VADD(T3V, T3Y)); ST(&(x[WS(rs, 51)]), VSUB(T3S, T3Z), ms, &(x[WS(rs, 1)])); ST(&(x[WS(rs, 13)]), VADD(T3S, T3Z), ms, &(x[WS(rs, 1)])); } { V T40, T41, T3G, T3H; T40 = VSUB(T3K, T3R); T41 = VBYI(VSUB(T3Y, T3V)); ST(&(x[WS(rs, 45)]), VSUB(T40, T41), ms, &(x[WS(rs, 1)])); ST(&(x[WS(rs, 19)]), VADD(T40, T41), ms, &(x[WS(rs, 1)])); T3G = VBYI(VADD(T3B, T3A)); T3H = VADD(T3D, T3E); ST(&(x[WS(rs, 5)]), VADD(T3G, T3H), ms, &(x[WS(rs, 1)])); ST(&(x[WS(rs, 59)]), VSUB(T3H, T3G), ms, &(x[WS(rs, 1)])); } } } } VLEAVE(); } static const tw_instr twinstr[] = { VTW(0, 1), VTW(0, 2), VTW(0, 3), VTW(0, 4), VTW(0, 5), VTW(0, 6), VTW(0, 7), VTW(0, 8), VTW(0, 9), VTW(0, 10), VTW(0, 11), VTW(0, 12), VTW(0, 13), VTW(0, 14), VTW(0, 15), VTW(0, 16), VTW(0, 17), VTW(0, 18), VTW(0, 19), VTW(0, 20), VTW(0, 21), VTW(0, 22), VTW(0, 23), VTW(0, 24), VTW(0, 25), VTW(0, 26), VTW(0, 27), VTW(0, 28), VTW(0, 29), VTW(0, 30), VTW(0, 31), VTW(0, 32), VTW(0, 33), VTW(0, 34), VTW(0, 35), VTW(0, 36), VTW(0, 37), VTW(0, 38), VTW(0, 39), VTW(0, 40), VTW(0, 41), VTW(0, 42), VTW(0, 43), VTW(0, 44), VTW(0, 45), VTW(0, 46), VTW(0, 47), VTW(0, 48), VTW(0, 49), VTW(0, 50), VTW(0, 51), VTW(0, 52), VTW(0, 53), VTW(0, 54), VTW(0, 55), VTW(0, 56), VTW(0, 57), VTW(0, 58), VTW(0, 59), VTW(0, 60), VTW(0, 61), VTW(0, 62), VTW(0, 63), {TW_NEXT, VL, 0} }; static const ct_desc desc = { 64, XSIMD_STRING("t2bv_64"), twinstr, &GENUS, {467, 198, 52, 0}, 0, 0, 0 }; void XSIMD(codelet_t2bv_64) (planner *p) { X(kdft_dit_register) (p, t2bv_64, &desc); } #endif /* HAVE_FMA */ fftw-3.3.3/dft/simd/common/t2bv_5.c0000644000175000017500000001373612054410562013666 00000000000000/* * Copyright (c) 2003, 2007-11 Matteo Frigo * Copyright (c) 2003, 2007-11 Massachusetts Institute of Technology * * This program is free software; you can redistribute it and/or modify * it under the terms of the GNU General Public License as published by * the Free Software Foundation; either version 2 of the License, or * (at your option) any later version. * * This program is distributed in the hope that it will be useful, * but WITHOUT ANY WARRANTY; without even the implied warranty of * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the * GNU General Public License for more details. * * You should have received a copy of the GNU General Public License * along with this program; if not, write to the Free Software * Foundation, Inc., 51 Franklin Street, Fifth Floor, Boston, MA 02110-1301 USA * */ /* This file was automatically generated --- DO NOT EDIT */ /* Generated on Sun Nov 25 07:39:14 EST 2012 */ #include "codelet-dft.h" #ifdef HAVE_FMA /* Generated by: ../../../genfft/gen_twiddle_c.native -fma -reorder-insns -schedule-for-pipeline -simd -compact -variables 4 -pipeline-latency 8 -n 5 -name t2bv_5 -include t2b.h -sign 1 */ /* * This function contains 20 FP additions, 19 FP multiplications, * (or, 11 additions, 10 multiplications, 9 fused multiply/add), * 26 stack variables, 4 constants, and 10 memory accesses */ #include "t2b.h" static void t2bv_5(R *ri, R *ii, const R *W, stride rs, INT mb, INT me, INT ms) { DVK(KP559016994, +0.559016994374947424102293417182819058860154590); DVK(KP250000000, +0.250000000000000000000000000000000000000000000); DVK(KP618033988, +0.618033988749894848204586834365638117720309180); DVK(KP951056516, +0.951056516295153572116439333379382143405698634); { INT m; R *x; x = ii; for (m = mb, W = W + (mb * ((TWVL / VL) * 8)); m < me; m = m + VL, x = x + (VL * ms), W = W + (TWVL * 8), MAKE_VOLATILE_STRIDE(5, rs)) { V T1, T2, T9, T4, T7; T1 = LD(&(x[0]), ms, &(x[0])); T2 = LD(&(x[WS(rs, 1)]), ms, &(x[WS(rs, 1)])); T9 = LD(&(x[WS(rs, 3)]), ms, &(x[WS(rs, 1)])); T4 = LD(&(x[WS(rs, 4)]), ms, &(x[0])); T7 = LD(&(x[WS(rs, 2)]), ms, &(x[0])); { V T3, Ta, T5, T8; T3 = BYTW(&(W[0]), T2); Ta = BYTW(&(W[TWVL * 4]), T9); T5 = BYTW(&(W[TWVL * 6]), T4); T8 = BYTW(&(W[TWVL * 2]), T7); { V T6, Tg, Tb, Th; T6 = VADD(T3, T5); Tg = VSUB(T3, T5); Tb = VADD(T8, Ta); Th = VSUB(T8, Ta); { V Te, Tc, Tk, Ti, Td, Tj, Tf; Te = VSUB(T6, Tb); Tc = VADD(T6, Tb); Tk = VMUL(LDK(KP951056516), VFNMS(LDK(KP618033988), Tg, Th)); Ti = VMUL(LDK(KP951056516), VFMA(LDK(KP618033988), Th, Tg)); Td = VFNMS(LDK(KP250000000), Tc, T1); ST(&(x[0]), VADD(T1, Tc), ms, &(x[0])); Tj = VFNMS(LDK(KP559016994), Te, Td); Tf = VFMA(LDK(KP559016994), Te, Td); ST(&(x[WS(rs, 2)]), VFNMSI(Tk, Tj), ms, &(x[0])); ST(&(x[WS(rs, 3)]), VFMAI(Tk, Tj), ms, &(x[WS(rs, 1)])); ST(&(x[WS(rs, 4)]), VFNMSI(Ti, Tf), ms, &(x[0])); ST(&(x[WS(rs, 1)]), VFMAI(Ti, Tf), ms, &(x[WS(rs, 1)])); } } } } } VLEAVE(); } static const tw_instr twinstr[] = { VTW(0, 1), VTW(0, 2), VTW(0, 3), VTW(0, 4), {TW_NEXT, VL, 0} }; static const ct_desc desc = { 5, XSIMD_STRING("t2bv_5"), twinstr, &GENUS, {11, 10, 9, 0}, 0, 0, 0 }; void XSIMD(codelet_t2bv_5) (planner *p) { X(kdft_dit_register) (p, t2bv_5, &desc); } #else /* HAVE_FMA */ /* Generated by: ../../../genfft/gen_twiddle_c.native -simd -compact -variables 4 -pipeline-latency 8 -n 5 -name t2bv_5 -include t2b.h -sign 1 */ /* * This function contains 20 FP additions, 14 FP multiplications, * (or, 17 additions, 11 multiplications, 3 fused multiply/add), * 20 stack variables, 4 constants, and 10 memory accesses */ #include "t2b.h" static void t2bv_5(R *ri, R *ii, const R *W, stride rs, INT mb, INT me, INT ms) { DVK(KP250000000, +0.250000000000000000000000000000000000000000000); DVK(KP559016994, +0.559016994374947424102293417182819058860154590); DVK(KP587785252, +0.587785252292473129168705954639072768597652438); DVK(KP951056516, +0.951056516295153572116439333379382143405698634); { INT m; R *x; x = ii; for (m = mb, W = W + (mb * ((TWVL / VL) * 8)); m < me; m = m + VL, x = x + (VL * ms), W = W + (TWVL * 8), MAKE_VOLATILE_STRIDE(5, rs)) { V Tf, T5, Ta, Tc, Td, Tg; Tf = LD(&(x[0]), ms, &(x[0])); { V T2, T9, T4, T7; { V T1, T8, T3, T6; T1 = LD(&(x[WS(rs, 1)]), ms, &(x[WS(rs, 1)])); T2 = BYTW(&(W[0]), T1); T8 = LD(&(x[WS(rs, 3)]), ms, &(x[WS(rs, 1)])); T9 = BYTW(&(W[TWVL * 4]), T8); T3 = LD(&(x[WS(rs, 4)]), ms, &(x[0])); T4 = BYTW(&(W[TWVL * 6]), T3); T6 = LD(&(x[WS(rs, 2)]), ms, &(x[0])); T7 = BYTW(&(W[TWVL * 2]), T6); } T5 = VSUB(T2, T4); Ta = VSUB(T7, T9); Tc = VADD(T2, T4); Td = VADD(T7, T9); Tg = VADD(Tc, Td); } ST(&(x[0]), VADD(Tf, Tg), ms, &(x[0])); { V Tb, Tj, Ti, Tk, Te, Th; Tb = VBYI(VFMA(LDK(KP951056516), T5, VMUL(LDK(KP587785252), Ta))); Tj = VBYI(VFNMS(LDK(KP951056516), Ta, VMUL(LDK(KP587785252), T5))); Te = VMUL(LDK(KP559016994), VSUB(Tc, Td)); Th = VFNMS(LDK(KP250000000), Tg, Tf); Ti = VADD(Te, Th); Tk = VSUB(Th, Te); ST(&(x[WS(rs, 1)]), VADD(Tb, Ti), ms, &(x[WS(rs, 1)])); ST(&(x[WS(rs, 3)]), VSUB(Tk, Tj), ms, &(x[WS(rs, 1)])); ST(&(x[WS(rs, 4)]), VSUB(Ti, Tb), ms, &(x[0])); ST(&(x[WS(rs, 2)]), VADD(Tj, Tk), ms, &(x[0])); } } } VLEAVE(); } static const tw_instr twinstr[] = { VTW(0, 1), VTW(0, 2), VTW(0, 3), VTW(0, 4), {TW_NEXT, VL, 0} }; static const ct_desc desc = { 5, XSIMD_STRING("t2bv_5"), twinstr, &GENUS, {17, 11, 3, 0}, 0, 0, 0 }; void XSIMD(codelet_t2bv_5) (planner *p) { X(kdft_dit_register) (p, t2bv_5, &desc); } #endif /* HAVE_FMA */ fftw-3.3.3/dft/simd/common/t2bv_10.c0000644000175000017500000002254312054410562013736 00000000000000/* * Copyright (c) 2003, 2007-11 Matteo Frigo * Copyright (c) 2003, 2007-11 Massachusetts Institute of Technology * * This program is free software; you can redistribute it and/or modify * it under the terms of the GNU General Public License as published by * the Free Software Foundation; either version 2 of the License, or * (at your option) any later version. * * This program is distributed in the hope that it will be useful, * but WITHOUT ANY WARRANTY; without even the implied warranty of * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the * GNU General Public License for more details. * * You should have received a copy of the GNU General Public License * along with this program; if not, write to the Free Software * Foundation, Inc., 51 Franklin Street, Fifth Floor, Boston, MA 02110-1301 USA * */ /* This file was automatically generated --- DO NOT EDIT */ /* Generated on Sun Nov 25 07:39:14 EST 2012 */ #include "codelet-dft.h" #ifdef HAVE_FMA /* Generated by: ../../../genfft/gen_twiddle_c.native -fma -reorder-insns -schedule-for-pipeline -simd -compact -variables 4 -pipeline-latency 8 -n 10 -name t2bv_10 -include t2b.h -sign 1 */ /* * This function contains 51 FP additions, 40 FP multiplications, * (or, 33 additions, 22 multiplications, 18 fused multiply/add), * 43 stack variables, 4 constants, and 20 memory accesses */ #include "t2b.h" static void t2bv_10(R *ri, R *ii, const R *W, stride rs, INT mb, INT me, INT ms) { DVK(KP559016994, +0.559016994374947424102293417182819058860154590); DVK(KP250000000, +0.250000000000000000000000000000000000000000000); DVK(KP618033988, +0.618033988749894848204586834365638117720309180); DVK(KP951056516, +0.951056516295153572116439333379382143405698634); { INT m; R *x; x = ii; for (m = mb, W = W + (mb * ((TWVL / VL) * 18)); m < me; m = m + VL, x = x + (VL * ms), W = W + (TWVL * 18), MAKE_VOLATILE_STRIDE(10, rs)) { V Td, TA, T4, Ta, Tk, TE, Tp, TF, TB, T9, T1, T2, Tb; T1 = LD(&(x[0]), ms, &(x[0])); T2 = LD(&(x[WS(rs, 5)]), ms, &(x[WS(rs, 1)])); { V Tg, Tn, Ti, Tl; Tg = LD(&(x[WS(rs, 4)]), ms, &(x[0])); Tn = LD(&(x[WS(rs, 1)]), ms, &(x[WS(rs, 1)])); Ti = LD(&(x[WS(rs, 9)]), ms, &(x[WS(rs, 1)])); Tl = LD(&(x[WS(rs, 6)]), ms, &(x[0])); { V T6, T8, T5, Tc; T5 = LD(&(x[WS(rs, 2)]), ms, &(x[0])); Tc = LD(&(x[WS(rs, 3)]), ms, &(x[WS(rs, 1)])); { V T3, Th, To, Tj, Tm, T7; T7 = LD(&(x[WS(rs, 7)]), ms, &(x[WS(rs, 1)])); T3 = BYTW(&(W[TWVL * 8]), T2); Th = BYTW(&(W[TWVL * 6]), Tg); To = BYTW(&(W[0]), Tn); Tj = BYTW(&(W[TWVL * 16]), Ti); Tm = BYTW(&(W[TWVL * 10]), Tl); T6 = BYTW(&(W[TWVL * 2]), T5); Td = BYTW(&(W[TWVL * 4]), Tc); T8 = BYTW(&(W[TWVL * 12]), T7); TA = VADD(T1, T3); T4 = VSUB(T1, T3); Ta = LD(&(x[WS(rs, 8)]), ms, &(x[0])); Tk = VSUB(Th, Tj); TE = VADD(Th, Tj); Tp = VSUB(Tm, To); TF = VADD(Tm, To); } TB = VADD(T6, T8); T9 = VSUB(T6, T8); } } Tb = BYTW(&(W[TWVL * 14]), Ta); { V TL, TG, Tw, Tq, TC, Te; TL = VSUB(TE, TF); TG = VADD(TE, TF); Tw = VSUB(Tk, Tp); Tq = VADD(Tk, Tp); TC = VADD(Tb, Td); Te = VSUB(Tb, Td); { V TM, TD, Tv, Tf; TM = VSUB(TB, TC); TD = VADD(TB, TC); Tv = VSUB(T9, Te); Tf = VADD(T9, Te); { V TP, TN, TH, TJ, Tz, Tx, Tr, Tt, TI, Ts; TP = VMUL(LDK(KP951056516), VFMA(LDK(KP618033988), TL, TM)); TN = VMUL(LDK(KP951056516), VFNMS(LDK(KP618033988), TM, TL)); TH = VADD(TD, TG); TJ = VSUB(TD, TG); Tz = VMUL(LDK(KP951056516), VFNMS(LDK(KP618033988), Tv, Tw)); Tx = VMUL(LDK(KP951056516), VFMA(LDK(KP618033988), Tw, Tv)); Tr = VADD(Tf, Tq); Tt = VSUB(Tf, Tq); ST(&(x[0]), VADD(TA, TH), ms, &(x[0])); TI = VFNMS(LDK(KP250000000), TH, TA); ST(&(x[WS(rs, 5)]), VADD(T4, Tr), ms, &(x[WS(rs, 1)])); Ts = VFNMS(LDK(KP250000000), Tr, T4); { V TK, TO, Tu, Ty; TK = VFNMS(LDK(KP559016994), TJ, TI); TO = VFMA(LDK(KP559016994), TJ, TI); Tu = VFMA(LDK(KP559016994), Tt, Ts); Ty = VFNMS(LDK(KP559016994), Tt, Ts); ST(&(x[WS(rs, 8)]), VFMAI(TN, TK), ms, &(x[0])); ST(&(x[WS(rs, 2)]), VFNMSI(TN, TK), ms, &(x[0])); ST(&(x[WS(rs, 6)]), VFMAI(TP, TO), ms, &(x[0])); ST(&(x[WS(rs, 4)]), VFNMSI(TP, TO), ms, &(x[0])); ST(&(x[WS(rs, 9)]), VFNMSI(Tx, Tu), ms, &(x[WS(rs, 1)])); ST(&(x[WS(rs, 1)]), VFMAI(Tx, Tu), ms, &(x[WS(rs, 1)])); ST(&(x[WS(rs, 7)]), VFNMSI(Tz, Ty), ms, &(x[WS(rs, 1)])); ST(&(x[WS(rs, 3)]), VFMAI(Tz, Ty), ms, &(x[WS(rs, 1)])); } } } } } } VLEAVE(); } static const tw_instr twinstr[] = { VTW(0, 1), VTW(0, 2), VTW(0, 3), VTW(0, 4), VTW(0, 5), VTW(0, 6), VTW(0, 7), VTW(0, 8), VTW(0, 9), {TW_NEXT, VL, 0} }; static const ct_desc desc = { 10, XSIMD_STRING("t2bv_10"), twinstr, &GENUS, {33, 22, 18, 0}, 0, 0, 0 }; void XSIMD(codelet_t2bv_10) (planner *p) { X(kdft_dit_register) (p, t2bv_10, &desc); } #else /* HAVE_FMA */ /* Generated by: ../../../genfft/gen_twiddle_c.native -simd -compact -variables 4 -pipeline-latency 8 -n 10 -name t2bv_10 -include t2b.h -sign 1 */ /* * This function contains 51 FP additions, 30 FP multiplications, * (or, 45 additions, 24 multiplications, 6 fused multiply/add), * 32 stack variables, 4 constants, and 20 memory accesses */ #include "t2b.h" static void t2bv_10(R *ri, R *ii, const R *W, stride rs, INT mb, INT me, INT ms) { DVK(KP587785252, +0.587785252292473129168705954639072768597652438); DVK(KP951056516, +0.951056516295153572116439333379382143405698634); DVK(KP250000000, +0.250000000000000000000000000000000000000000000); DVK(KP559016994, +0.559016994374947424102293417182819058860154590); { INT m; R *x; x = ii; for (m = mb, W = W + (mb * ((TWVL / VL) * 18)); m < me; m = m + VL, x = x + (VL * ms), W = W + (TWVL * 18), MAKE_VOLATILE_STRIDE(10, rs)) { V Tu, TH, Tg, Tl, Tp, TD, TE, TJ, T5, Ta, To, TA, TB, TI, Tr; V Tt, Ts; Tr = LD(&(x[0]), ms, &(x[0])); Ts = LD(&(x[WS(rs, 5)]), ms, &(x[WS(rs, 1)])); Tt = BYTW(&(W[TWVL * 8]), Ts); Tu = VSUB(Tr, Tt); TH = VADD(Tr, Tt); { V Td, Tk, Tf, Ti; { V Tc, Tj, Te, Th; Tc = LD(&(x[WS(rs, 4)]), ms, &(x[0])); Td = BYTW(&(W[TWVL * 6]), Tc); Tj = LD(&(x[WS(rs, 1)]), ms, &(x[WS(rs, 1)])); Tk = BYTW(&(W[0]), Tj); Te = LD(&(x[WS(rs, 9)]), ms, &(x[WS(rs, 1)])); Tf = BYTW(&(W[TWVL * 16]), Te); Th = LD(&(x[WS(rs, 6)]), ms, &(x[0])); Ti = BYTW(&(W[TWVL * 10]), Th); } Tg = VSUB(Td, Tf); Tl = VSUB(Ti, Tk); Tp = VADD(Tg, Tl); TD = VADD(Td, Tf); TE = VADD(Ti, Tk); TJ = VADD(TD, TE); } { V T2, T9, T4, T7; { V T1, T8, T3, T6; T1 = LD(&(x[WS(rs, 2)]), ms, &(x[0])); T2 = BYTW(&(W[TWVL * 2]), T1); T8 = LD(&(x[WS(rs, 3)]), ms, &(x[WS(rs, 1)])); T9 = BYTW(&(W[TWVL * 4]), T8); T3 = LD(&(x[WS(rs, 7)]), ms, &(x[WS(rs, 1)])); T4 = BYTW(&(W[TWVL * 12]), T3); T6 = LD(&(x[WS(rs, 8)]), ms, &(x[0])); T7 = BYTW(&(W[TWVL * 14]), T6); } T5 = VSUB(T2, T4); Ta = VSUB(T7, T9); To = VADD(T5, Ta); TA = VADD(T2, T4); TB = VADD(T7, T9); TI = VADD(TA, TB); } { V Tq, Tv, Tw, Tn, Tz, Tb, Tm, Ty, Tx; Tq = VMUL(LDK(KP559016994), VSUB(To, Tp)); Tv = VADD(To, Tp); Tw = VFNMS(LDK(KP250000000), Tv, Tu); Tb = VSUB(T5, Ta); Tm = VSUB(Tg, Tl); Tn = VBYI(VFMA(LDK(KP951056516), Tb, VMUL(LDK(KP587785252), Tm))); Tz = VBYI(VFNMS(LDK(KP951056516), Tm, VMUL(LDK(KP587785252), Tb))); ST(&(x[WS(rs, 5)]), VADD(Tu, Tv), ms, &(x[WS(rs, 1)])); Ty = VSUB(Tw, Tq); ST(&(x[WS(rs, 3)]), VSUB(Ty, Tz), ms, &(x[WS(rs, 1)])); ST(&(x[WS(rs, 7)]), VADD(Tz, Ty), ms, &(x[WS(rs, 1)])); Tx = VADD(Tq, Tw); ST(&(x[WS(rs, 1)]), VADD(Tn, Tx), ms, &(x[WS(rs, 1)])); ST(&(x[WS(rs, 9)]), VSUB(Tx, Tn), ms, &(x[WS(rs, 1)])); } { V TM, TK, TL, TG, TP, TC, TF, TO, TN; TM = VMUL(LDK(KP559016994), VSUB(TI, TJ)); TK = VADD(TI, TJ); TL = VFNMS(LDK(KP250000000), TK, TH); TC = VSUB(TA, TB); TF = VSUB(TD, TE); TG = VBYI(VFNMS(LDK(KP951056516), TF, VMUL(LDK(KP587785252), TC))); TP = VBYI(VFMA(LDK(KP951056516), TC, VMUL(LDK(KP587785252), TF))); ST(&(x[0]), VADD(TH, TK), ms, &(x[0])); TO = VADD(TM, TL); ST(&(x[WS(rs, 4)]), VSUB(TO, TP), ms, &(x[0])); ST(&(x[WS(rs, 6)]), VADD(TP, TO), ms, &(x[0])); TN = VSUB(TL, TM); ST(&(x[WS(rs, 2)]), VADD(TG, TN), ms, &(x[0])); ST(&(x[WS(rs, 8)]), VSUB(TN, TG), ms, &(x[0])); } } } VLEAVE(); } static const tw_instr twinstr[] = { VTW(0, 1), VTW(0, 2), VTW(0, 3), VTW(0, 4), VTW(0, 5), VTW(0, 6), VTW(0, 7), VTW(0, 8), VTW(0, 9), {TW_NEXT, VL, 0} }; static const ct_desc desc = { 10, XSIMD_STRING("t2bv_10"), twinstr, &GENUS, {45, 24, 6, 0}, 0, 0, 0 }; void XSIMD(codelet_t2bv_10) (planner *p) { X(kdft_dit_register) (p, t2bv_10, &desc); } #endif /* HAVE_FMA */ fftw-3.3.3/dft/simd/common/t2bv_20.c0000644000175000017500000004175512054410563013746 00000000000000/* * Copyright (c) 2003, 2007-11 Matteo Frigo * Copyright (c) 2003, 2007-11 Massachusetts Institute of Technology * * This program is free software; you can redistribute it and/or modify * it under the terms of the GNU General Public License as published by * the Free Software Foundation; either version 2 of the License, or * (at your option) any later version. * * This program is distributed in the hope that it will be useful, * but WITHOUT ANY WARRANTY; without even the implied warranty of * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the * GNU General Public License for more details. * * You should have received a copy of the GNU General Public License * along with this program; if not, write to the Free Software * Foundation, Inc., 51 Franklin Street, Fifth Floor, Boston, MA 02110-1301 USA * */ /* This file was automatically generated --- DO NOT EDIT */ /* Generated on Sun Nov 25 07:39:14 EST 2012 */ #include "codelet-dft.h" #ifdef HAVE_FMA /* Generated by: ../../../genfft/gen_twiddle_c.native -fma -reorder-insns -schedule-for-pipeline -simd -compact -variables 4 -pipeline-latency 8 -n 20 -name t2bv_20 -include t2b.h -sign 1 */ /* * This function contains 123 FP additions, 88 FP multiplications, * (or, 77 additions, 42 multiplications, 46 fused multiply/add), * 68 stack variables, 4 constants, and 40 memory accesses */ #include "t2b.h" static void t2bv_20(R *ri, R *ii, const R *W, stride rs, INT mb, INT me, INT ms) { DVK(KP951056516, +0.951056516295153572116439333379382143405698634); DVK(KP559016994, +0.559016994374947424102293417182819058860154590); DVK(KP250000000, +0.250000000000000000000000000000000000000000000); DVK(KP618033988, +0.618033988749894848204586834365638117720309180); { INT m; R *x; x = ii; for (m = mb, W = W + (mb * ((TWVL / VL) * 38)); m < me; m = m + VL, x = x + (VL * ms), W = W + (TWVL * 38), MAKE_VOLATILE_STRIDE(20, rs)) { V T4, TX, T1m, T1K, T1y, Tk, Tf, T14, TQ, TZ, T1O, T1w, T1L, T1p, T1M; V T1s, TF, TY, T1x, Tp; { V T1, TV, T2, TT; T1 = LD(&(x[0]), ms, &(x[0])); TV = LD(&(x[WS(rs, 15)]), ms, &(x[WS(rs, 1)])); T2 = LD(&(x[WS(rs, 10)]), ms, &(x[0])); TT = LD(&(x[WS(rs, 5)]), ms, &(x[WS(rs, 1)])); { V T9, T1n, TK, T1v, TP, Te, T1q, T1u, TB, TD, Tm, T1o, Tz, Tn, T1r; V TE, To; { V TM, TO, Ta, Tc; { V T5, T7, TG, TI, T1k, T1l; T5 = LD(&(x[WS(rs, 4)]), ms, &(x[0])); T7 = LD(&(x[WS(rs, 14)]), ms, &(x[0])); TG = LD(&(x[WS(rs, 13)]), ms, &(x[WS(rs, 1)])); TI = LD(&(x[WS(rs, 3)]), ms, &(x[WS(rs, 1)])); { V TW, T3, TU, T6, T8, TH, TJ, TL, TN; TL = LD(&(x[WS(rs, 17)]), ms, &(x[WS(rs, 1)])); TW = BYTW(&(W[TWVL * 28]), TV); T3 = BYTW(&(W[TWVL * 18]), T2); TU = BYTW(&(W[TWVL * 8]), TT); T6 = BYTW(&(W[TWVL * 6]), T5); T8 = BYTW(&(W[TWVL * 26]), T7); TH = BYTW(&(W[TWVL * 24]), TG); TJ = BYTW(&(W[TWVL * 4]), TI); TM = BYTW(&(W[TWVL * 32]), TL); TN = LD(&(x[WS(rs, 7)]), ms, &(x[WS(rs, 1)])); T4 = VSUB(T1, T3); T1k = VADD(T1, T3); TX = VSUB(TU, TW); T1l = VADD(TU, TW); T9 = VSUB(T6, T8); T1n = VADD(T6, T8); TK = VSUB(TH, TJ); T1v = VADD(TH, TJ); TO = BYTW(&(W[TWVL * 12]), TN); } Ta = LD(&(x[WS(rs, 16)]), ms, &(x[0])); T1m = VSUB(T1k, T1l); T1K = VADD(T1k, T1l); Tc = LD(&(x[WS(rs, 6)]), ms, &(x[0])); } { V Tb, Tx, Td, Th, Tj, Tw, Tg, Ti, Tv; Tg = LD(&(x[WS(rs, 8)]), ms, &(x[0])); Ti = LD(&(x[WS(rs, 18)]), ms, &(x[0])); Tv = LD(&(x[WS(rs, 9)]), ms, &(x[WS(rs, 1)])); TP = VSUB(TM, TO); T1y = VADD(TM, TO); Tb = BYTW(&(W[TWVL * 30]), Ta); Tx = LD(&(x[WS(rs, 19)]), ms, &(x[WS(rs, 1)])); Td = BYTW(&(W[TWVL * 10]), Tc); Th = BYTW(&(W[TWVL * 14]), Tg); Tj = BYTW(&(W[TWVL * 34]), Ti); Tw = BYTW(&(W[TWVL * 16]), Tv); { V TA, TC, Ty, Tl; TA = LD(&(x[WS(rs, 1)]), ms, &(x[WS(rs, 1)])); TC = LD(&(x[WS(rs, 11)]), ms, &(x[WS(rs, 1)])); Tl = LD(&(x[WS(rs, 12)]), ms, &(x[0])); Ty = BYTW(&(W[TWVL * 36]), Tx); Te = VSUB(Tb, Td); T1q = VADD(Tb, Td); Tk = VSUB(Th, Tj); T1u = VADD(Th, Tj); TB = BYTW(&(W[0]), TA); TD = BYTW(&(W[TWVL * 20]), TC); Tm = BYTW(&(W[TWVL * 22]), Tl); T1o = VADD(Tw, Ty); Tz = VSUB(Tw, Ty); Tn = LD(&(x[WS(rs, 2)]), ms, &(x[0])); } } } Tf = VADD(T9, Te); T14 = VSUB(T9, Te); TQ = VSUB(TK, TP); TZ = VADD(TK, TP); T1r = VADD(TB, TD); TE = VSUB(TB, TD); T1O = VADD(T1u, T1v); T1w = VSUB(T1u, T1v); To = BYTW(&(W[TWVL * 2]), Tn); T1L = VADD(T1n, T1o); T1p = VSUB(T1n, T1o); T1M = VADD(T1q, T1r); T1s = VSUB(T1q, T1r); TF = VSUB(Tz, TE); TY = VADD(Tz, TE); T1x = VADD(Tm, To); Tp = VSUB(Tm, To); } } { V T1V, T1N, T12, T1b, TR, T1G, T1t, T1z, T1P, Tq, T15, T11, T1j, T10; T1V = VSUB(T1L, T1M); T1N = VADD(T1L, T1M); T12 = VSUB(TY, TZ); T10 = VADD(TY, TZ); T1b = VFNMS(LDK(KP618033988), TF, TQ); TR = VFMA(LDK(KP618033988), TQ, TF); T1G = VSUB(T1p, T1s); T1t = VADD(T1p, T1s); T1z = VSUB(T1x, T1y); T1P = VADD(T1x, T1y); Tq = VADD(Tk, Tp); T15 = VSUB(Tk, Tp); T11 = VFNMS(LDK(KP250000000), T10, TX); T1j = VADD(TX, T10); { V T1J, T1H, T1D, T1Z, T1X, T1T, T1f, T1h, T19, T17, T1C, T1S, T1a, Tu, T1F; V T1A; T1F = VSUB(T1w, T1z); T1A = VADD(T1w, T1z); { V T1W, T1Q, Tt, Tr; T1W = VSUB(T1O, T1P); T1Q = VADD(T1O, T1P); Tt = VSUB(Tf, Tq); Tr = VADD(Tf, Tq); { V T1e, T16, T1d, T13; T1e = VFNMS(LDK(KP618033988), T14, T15); T16 = VFMA(LDK(KP618033988), T15, T14); T1d = VFNMS(LDK(KP559016994), T12, T11); T13 = VFMA(LDK(KP559016994), T12, T11); T1J = VMUL(LDK(KP951056516), VFMA(LDK(KP618033988), T1F, T1G)); T1H = VMUL(LDK(KP951056516), VFNMS(LDK(KP618033988), T1G, T1F)); { V T1B, T1R, Ts, T1i; T1B = VADD(T1t, T1A); T1D = VSUB(T1t, T1A); T1Z = VMUL(LDK(KP951056516), VFNMS(LDK(KP618033988), T1V, T1W)); T1X = VMUL(LDK(KP951056516), VFMA(LDK(KP618033988), T1W, T1V)); T1R = VADD(T1N, T1Q); T1T = VSUB(T1N, T1Q); Ts = VFNMS(LDK(KP250000000), Tr, T4); T1i = VADD(T4, Tr); T1f = VFNMS(LDK(KP951056516), T1e, T1d); T1h = VFMA(LDK(KP951056516), T1e, T1d); T19 = VFNMS(LDK(KP951056516), T16, T13); T17 = VFMA(LDK(KP951056516), T16, T13); ST(&(x[WS(rs, 10)]), VADD(T1m, T1B), ms, &(x[0])); T1C = VFNMS(LDK(KP250000000), T1B, T1m); ST(&(x[0]), VADD(T1K, T1R), ms, &(x[0])); T1S = VFNMS(LDK(KP250000000), T1R, T1K); T1a = VFNMS(LDK(KP559016994), Tt, Ts); Tu = VFMA(LDK(KP559016994), Tt, Ts); ST(&(x[WS(rs, 5)]), VFMAI(T1j, T1i), ms, &(x[WS(rs, 1)])); ST(&(x[WS(rs, 15)]), VFNMSI(T1j, T1i), ms, &(x[WS(rs, 1)])); } } } { V T1E, T1I, T1U, T1Y; T1E = VFNMS(LDK(KP559016994), T1D, T1C); T1I = VFMA(LDK(KP559016994), T1D, T1C); T1U = VFMA(LDK(KP559016994), T1T, T1S); T1Y = VFNMS(LDK(KP559016994), T1T, T1S); { V T1c, T1g, T18, TS; T1c = VFMA(LDK(KP951056516), T1b, T1a); T1g = VFNMS(LDK(KP951056516), T1b, T1a); T18 = VFMA(LDK(KP951056516), TR, Tu); TS = VFNMS(LDK(KP951056516), TR, Tu); ST(&(x[WS(rs, 18)]), VFMAI(T1H, T1E), ms, &(x[0])); ST(&(x[WS(rs, 2)]), VFNMSI(T1H, T1E), ms, &(x[0])); ST(&(x[WS(rs, 14)]), VFNMSI(T1J, T1I), ms, &(x[0])); ST(&(x[WS(rs, 6)]), VFMAI(T1J, T1I), ms, &(x[0])); ST(&(x[WS(rs, 16)]), VFMAI(T1X, T1U), ms, &(x[0])); ST(&(x[WS(rs, 4)]), VFNMSI(T1X, T1U), ms, &(x[0])); ST(&(x[WS(rs, 12)]), VFNMSI(T1Z, T1Y), ms, &(x[0])); ST(&(x[WS(rs, 8)]), VFMAI(T1Z, T1Y), ms, &(x[0])); ST(&(x[WS(rs, 17)]), VFMAI(T1f, T1c), ms, &(x[WS(rs, 1)])); ST(&(x[WS(rs, 3)]), VFNMSI(T1f, T1c), ms, &(x[WS(rs, 1)])); ST(&(x[WS(rs, 13)]), VFMAI(T1h, T1g), ms, &(x[WS(rs, 1)])); ST(&(x[WS(rs, 7)]), VFNMSI(T1h, T1g), ms, &(x[WS(rs, 1)])); ST(&(x[WS(rs, 9)]), VFMAI(T19, T18), ms, &(x[WS(rs, 1)])); ST(&(x[WS(rs, 11)]), VFNMSI(T19, T18), ms, &(x[WS(rs, 1)])); ST(&(x[WS(rs, 1)]), VFMAI(T17, TS), ms, &(x[WS(rs, 1)])); ST(&(x[WS(rs, 19)]), VFNMSI(T17, TS), ms, &(x[WS(rs, 1)])); } } } } } } VLEAVE(); } static const tw_instr twinstr[] = { VTW(0, 1), VTW(0, 2), VTW(0, 3), VTW(0, 4), VTW(0, 5), VTW(0, 6), VTW(0, 7), VTW(0, 8), VTW(0, 9), VTW(0, 10), VTW(0, 11), VTW(0, 12), VTW(0, 13), VTW(0, 14), VTW(0, 15), VTW(0, 16), VTW(0, 17), VTW(0, 18), VTW(0, 19), {TW_NEXT, VL, 0} }; static const ct_desc desc = { 20, XSIMD_STRING("t2bv_20"), twinstr, &GENUS, {77, 42, 46, 0}, 0, 0, 0 }; void XSIMD(codelet_t2bv_20) (planner *p) { X(kdft_dit_register) (p, t2bv_20, &desc); } #else /* HAVE_FMA */ /* Generated by: ../../../genfft/gen_twiddle_c.native -simd -compact -variables 4 -pipeline-latency 8 -n 20 -name t2bv_20 -include t2b.h -sign 1 */ /* * This function contains 123 FP additions, 62 FP multiplications, * (or, 111 additions, 50 multiplications, 12 fused multiply/add), * 54 stack variables, 4 constants, and 40 memory accesses */ #include "t2b.h" static void t2bv_20(R *ri, R *ii, const R *W, stride rs, INT mb, INT me, INT ms) { DVK(KP587785252, +0.587785252292473129168705954639072768597652438); DVK(KP951056516, +0.951056516295153572116439333379382143405698634); DVK(KP250000000, +0.250000000000000000000000000000000000000000000); DVK(KP559016994, +0.559016994374947424102293417182819058860154590); { INT m; R *x; x = ii; for (m = mb, W = W + (mb * ((TWVL / VL) * 38)); m < me; m = m + VL, x = x + (VL * ms), W = W + (TWVL * 38), MAKE_VOLATILE_STRIDE(20, rs)) { V T4, T10, T1B, T1R, TF, T14, T15, TQ, Tf, Tq, Tr, T1N, T1O, T1P, T1t; V T1w, T1D, TT, TU, T11, T1K, T1L, T1M, T1m, T1p, T1C, T1i, T1j; { V T1, TZ, T3, TX, TY, T2, TW, T1z, T1A; T1 = LD(&(x[0]), ms, &(x[0])); TY = LD(&(x[WS(rs, 15)]), ms, &(x[WS(rs, 1)])); TZ = BYTW(&(W[TWVL * 28]), TY); T2 = LD(&(x[WS(rs, 10)]), ms, &(x[0])); T3 = BYTW(&(W[TWVL * 18]), T2); TW = LD(&(x[WS(rs, 5)]), ms, &(x[WS(rs, 1)])); TX = BYTW(&(W[TWVL * 8]), TW); T4 = VSUB(T1, T3); T10 = VSUB(TX, TZ); T1z = VADD(T1, T3); T1A = VADD(TX, TZ); T1B = VSUB(T1z, T1A); T1R = VADD(T1z, T1A); } { V T9, T1k, TK, T1s, TP, T1v, Te, T1n, Tk, T1r, Tz, T1l, TE, T1o, Tp; V T1u; { V T6, T8, T5, T7; T5 = LD(&(x[WS(rs, 4)]), ms, &(x[0])); T6 = BYTW(&(W[TWVL * 6]), T5); T7 = LD(&(x[WS(rs, 14)]), ms, &(x[0])); T8 = BYTW(&(W[TWVL * 26]), T7); T9 = VSUB(T6, T8); T1k = VADD(T6, T8); } { V TH, TJ, TG, TI; TG = LD(&(x[WS(rs, 13)]), ms, &(x[WS(rs, 1)])); TH = BYTW(&(W[TWVL * 24]), TG); TI = LD(&(x[WS(rs, 3)]), ms, &(x[WS(rs, 1)])); TJ = BYTW(&(W[TWVL * 4]), TI); TK = VSUB(TH, TJ); T1s = VADD(TH, TJ); } { V TM, TO, TL, TN; TL = LD(&(x[WS(rs, 17)]), ms, &(x[WS(rs, 1)])); TM = BYTW(&(W[TWVL * 32]), TL); TN = LD(&(x[WS(rs, 7)]), ms, &(x[WS(rs, 1)])); TO = BYTW(&(W[TWVL * 12]), TN); TP = VSUB(TM, TO); T1v = VADD(TM, TO); } { V Tb, Td, Ta, Tc; Ta = LD(&(x[WS(rs, 16)]), ms, &(x[0])); Tb = BYTW(&(W[TWVL * 30]), Ta); Tc = LD(&(x[WS(rs, 6)]), ms, &(x[0])); Td = BYTW(&(W[TWVL * 10]), Tc); Te = VSUB(Tb, Td); T1n = VADD(Tb, Td); } { V Th, Tj, Tg, Ti; Tg = LD(&(x[WS(rs, 8)]), ms, &(x[0])); Th = BYTW(&(W[TWVL * 14]), Tg); Ti = LD(&(x[WS(rs, 18)]), ms, &(x[0])); Tj = BYTW(&(W[TWVL * 34]), Ti); Tk = VSUB(Th, Tj); T1r = VADD(Th, Tj); } { V Tw, Ty, Tv, Tx; Tv = LD(&(x[WS(rs, 9)]), ms, &(x[WS(rs, 1)])); Tw = BYTW(&(W[TWVL * 16]), Tv); Tx = LD(&(x[WS(rs, 19)]), ms, &(x[WS(rs, 1)])); Ty = BYTW(&(W[TWVL * 36]), Tx); Tz = VSUB(Tw, Ty); T1l = VADD(Tw, Ty); } { V TB, TD, TA, TC; TA = LD(&(x[WS(rs, 1)]), ms, &(x[WS(rs, 1)])); TB = BYTW(&(W[0]), TA); TC = LD(&(x[WS(rs, 11)]), ms, &(x[WS(rs, 1)])); TD = BYTW(&(W[TWVL * 20]), TC); TE = VSUB(TB, TD); T1o = VADD(TB, TD); } { V Tm, To, Tl, Tn; Tl = LD(&(x[WS(rs, 12)]), ms, &(x[0])); Tm = BYTW(&(W[TWVL * 22]), Tl); Tn = LD(&(x[WS(rs, 2)]), ms, &(x[0])); To = BYTW(&(W[TWVL * 2]), Tn); Tp = VSUB(Tm, To); T1u = VADD(Tm, To); } TF = VSUB(Tz, TE); T14 = VSUB(T9, Te); T15 = VSUB(Tk, Tp); TQ = VSUB(TK, TP); Tf = VADD(T9, Te); Tq = VADD(Tk, Tp); Tr = VADD(Tf, Tq); T1N = VADD(T1r, T1s); T1O = VADD(T1u, T1v); T1P = VADD(T1N, T1O); T1t = VSUB(T1r, T1s); T1w = VSUB(T1u, T1v); T1D = VADD(T1t, T1w); TT = VADD(Tz, TE); TU = VADD(TK, TP); T11 = VADD(TT, TU); T1K = VADD(T1k, T1l); T1L = VADD(T1n, T1o); T1M = VADD(T1K, T1L); T1m = VSUB(T1k, T1l); T1p = VSUB(T1n, T1o); T1C = VADD(T1m, T1p); } T1i = VADD(T4, Tr); T1j = VBYI(VADD(T10, T11)); ST(&(x[WS(rs, 15)]), VSUB(T1i, T1j), ms, &(x[WS(rs, 1)])); ST(&(x[WS(rs, 5)]), VADD(T1i, T1j), ms, &(x[WS(rs, 1)])); { V T1Q, T1S, T1T, T1X, T1Z, T1V, T1W, T1Y, T1U; T1Q = VMUL(LDK(KP559016994), VSUB(T1M, T1P)); T1S = VADD(T1M, T1P); T1T = VFNMS(LDK(KP250000000), T1S, T1R); T1V = VSUB(T1K, T1L); T1W = VSUB(T1N, T1O); T1X = VBYI(VFMA(LDK(KP951056516), T1V, VMUL(LDK(KP587785252), T1W))); T1Z = VBYI(VFNMS(LDK(KP951056516), T1W, VMUL(LDK(KP587785252), T1V))); ST(&(x[0]), VADD(T1R, T1S), ms, &(x[0])); T1Y = VSUB(T1T, T1Q); ST(&(x[WS(rs, 8)]), VSUB(T1Y, T1Z), ms, &(x[0])); ST(&(x[WS(rs, 12)]), VADD(T1Z, T1Y), ms, &(x[0])); T1U = VADD(T1Q, T1T); ST(&(x[WS(rs, 4)]), VSUB(T1U, T1X), ms, &(x[0])); ST(&(x[WS(rs, 16)]), VADD(T1X, T1U), ms, &(x[0])); } { V T1G, T1E, T1F, T1y, T1I, T1q, T1x, T1J, T1H; T1G = VMUL(LDK(KP559016994), VSUB(T1C, T1D)); T1E = VADD(T1C, T1D); T1F = VFNMS(LDK(KP250000000), T1E, T1B); T1q = VSUB(T1m, T1p); T1x = VSUB(T1t, T1w); T1y = VBYI(VFNMS(LDK(KP951056516), T1x, VMUL(LDK(KP587785252), T1q))); T1I = VBYI(VFMA(LDK(KP951056516), T1q, VMUL(LDK(KP587785252), T1x))); ST(&(x[WS(rs, 10)]), VADD(T1B, T1E), ms, &(x[0])); T1J = VADD(T1G, T1F); ST(&(x[WS(rs, 6)]), VADD(T1I, T1J), ms, &(x[0])); ST(&(x[WS(rs, 14)]), VSUB(T1J, T1I), ms, &(x[0])); T1H = VSUB(T1F, T1G); ST(&(x[WS(rs, 2)]), VADD(T1y, T1H), ms, &(x[0])); ST(&(x[WS(rs, 18)]), VSUB(T1H, T1y), ms, &(x[0])); } { V TR, T16, T1d, T1b, T13, T1e, Tu, T1a; TR = VFNMS(LDK(KP951056516), TQ, VMUL(LDK(KP587785252), TF)); T16 = VFNMS(LDK(KP951056516), T15, VMUL(LDK(KP587785252), T14)); T1d = VFMA(LDK(KP951056516), T14, VMUL(LDK(KP587785252), T15)); T1b = VFMA(LDK(KP951056516), TF, VMUL(LDK(KP587785252), TQ)); { V TV, T12, Ts, Tt; TV = VMUL(LDK(KP559016994), VSUB(TT, TU)); T12 = VFNMS(LDK(KP250000000), T11, T10); T13 = VSUB(TV, T12); T1e = VADD(TV, T12); Ts = VFNMS(LDK(KP250000000), Tr, T4); Tt = VMUL(LDK(KP559016994), VSUB(Tf, Tq)); Tu = VSUB(Ts, Tt); T1a = VADD(Tt, Ts); } { V TS, T17, T1g, T1h; TS = VSUB(Tu, TR); T17 = VBYI(VSUB(T13, T16)); ST(&(x[WS(rs, 17)]), VSUB(TS, T17), ms, &(x[WS(rs, 1)])); ST(&(x[WS(rs, 3)]), VADD(TS, T17), ms, &(x[WS(rs, 1)])); T1g = VADD(T1a, T1b); T1h = VBYI(VSUB(T1e, T1d)); ST(&(x[WS(rs, 11)]), VSUB(T1g, T1h), ms, &(x[WS(rs, 1)])); ST(&(x[WS(rs, 9)]), VADD(T1g, T1h), ms, &(x[WS(rs, 1)])); } { V T18, T19, T1c, T1f; T18 = VADD(Tu, TR); T19 = VBYI(VADD(T16, T13)); ST(&(x[WS(rs, 13)]), VSUB(T18, T19), ms, &(x[WS(rs, 1)])); ST(&(x[WS(rs, 7)]), VADD(T18, T19), ms, &(x[WS(rs, 1)])); T1c = VSUB(T1a, T1b); T1f = VBYI(VADD(T1d, T1e)); ST(&(x[WS(rs, 19)]), VSUB(T1c, T1f), ms, &(x[WS(rs, 1)])); ST(&(x[WS(rs, 1)]), VADD(T1c, T1f), ms, &(x[WS(rs, 1)])); } } } } VLEAVE(); } static const tw_instr twinstr[] = { VTW(0, 1), VTW(0, 2), VTW(0, 3), VTW(0, 4), VTW(0, 5), VTW(0, 6), VTW(0, 7), VTW(0, 8), VTW(0, 9), VTW(0, 10), VTW(0, 11), VTW(0, 12), VTW(0, 13), VTW(0, 14), VTW(0, 15), VTW(0, 16), VTW(0, 17), VTW(0, 18), VTW(0, 19), {TW_NEXT, VL, 0} }; static const ct_desc desc = { 20, XSIMD_STRING("t2bv_20"), twinstr, &GENUS, {111, 50, 12, 0}, 0, 0, 0 }; void XSIMD(codelet_t2bv_20) (planner *p) { X(kdft_dit_register) (p, t2bv_20, &desc); } #endif /* HAVE_FMA */ fftw-3.3.3/dft/simd/common/t2bv_25.c0000644000175000017500000011472712054410572013753 00000000000000/* * Copyright (c) 2003, 2007-11 Matteo Frigo * Copyright (c) 2003, 2007-11 Massachusetts Institute of Technology * * This program is free software; you can redistribute it and/or modify * it under the terms of the GNU General Public License as published by * the Free Software Foundation; either version 2 of the License, or * (at your option) any later version. * * This program is distributed in the hope that it will be useful, * but WITHOUT ANY WARRANTY; without even the implied warranty of * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the * GNU General Public License for more details. * * You should have received a copy of the GNU General Public License * along with this program; if not, write to the Free Software * Foundation, Inc., 51 Franklin Street, Fifth Floor, Boston, MA 02110-1301 USA * */ /* This file was automatically generated --- DO NOT EDIT */ /* Generated on Sun Nov 25 07:39:15 EST 2012 */ #include "codelet-dft.h" #ifdef HAVE_FMA /* Generated by: ../../../genfft/gen_twiddle_c.native -fma -reorder-insns -schedule-for-pipeline -simd -compact -variables 4 -pipeline-latency 8 -n 25 -name t2bv_25 -include t2b.h -sign 1 */ /* * This function contains 248 FP additions, 241 FP multiplications, * (or, 67 additions, 60 multiplications, 181 fused multiply/add), * 208 stack variables, 67 constants, and 50 memory accesses */ #include "t2b.h" static void t2bv_25(R *ri, R *ii, const R *W, stride rs, INT mb, INT me, INT ms) { DVK(KP792626838, +0.792626838241819413632131824093538848057784557); DVK(KP876091699, +0.876091699473550838204498029706869638173524346); DVK(KP617882369, +0.617882369114440893914546919006756321695042882); DVK(KP803003575, +0.803003575438660414833440593570376004635464850); DVK(KP242145790, +0.242145790282157779872542093866183953459003101); DVK(KP968583161, +0.968583161128631119490168375464735813836012403); DVK(KP999544308, +0.999544308746292983948881682379742149196758193); DVK(KP916574801, +0.916574801383451584742370439148878693530976769); DVK(KP904730450, +0.904730450839922351881287709692877908104763647); DVK(KP809385824, +0.809385824416008241660603814668679683846476688); DVK(KP447417479, +0.447417479732227551498980015410057305749330693); DVK(KP894834959, +0.894834959464455102997960030820114611498661386); DVK(KP867381224, +0.867381224396525206773171885031575671309956167); DVK(KP683113946, +0.683113946453479238701949862233725244439656928); DVK(KP559154169, +0.559154169276087864842202529084232643714075927); DVK(KP958953096, +0.958953096729998668045963838399037225970891871); DVK(KP831864738, +0.831864738706457140726048799369896829771167132); DVK(KP829049696, +0.829049696159252993975487806364305442437946767); DVK(KP860541664, +0.860541664367944677098261680920518816412804187); DVK(KP897376177, +0.897376177523557693138608077137219684419427330); DVK(KP876306680, +0.876306680043863587308115903922062583399064238); DVK(KP681693190, +0.681693190061530575150324149145440022633095390); DVK(KP560319534, +0.560319534973832390111614715371676131169633784); DVK(KP855719849, +0.855719849902058969314654733608091555096772472); DVK(KP237294955, +0.237294955877110315393888866460840817927895961); DVK(KP949179823, +0.949179823508441261575555465843363271711583843); DVK(KP904508497, +0.904508497187473712051146708591409529430077295); DVK(KP997675361, +0.997675361079556513670859573984492383596555031); DVK(KP763932022, +0.763932022500210303590826331268723764559381640); DVK(KP690983005, +0.690983005625052575897706582817180941139845410); DVK(KP992114701, +0.992114701314477831049793042785778521453036709); DVK(KP952936919, +0.952936919628306576880750665357914584765951388); DVK(KP998026728, +0.998026728428271561952336806863450553336905220); DVK(KP262346850, +0.262346850930607871785420028382979691334784273); DVK(KP570584518, +0.570584518783621657366766175430996792655723863); DVK(KP669429328, +0.669429328479476605641803240971985825917022098); DVK(KP923225144, +0.923225144846402650453449441572664695995209956); DVK(KP945422727, +0.945422727388575946270360266328811958657216298); DVK(KP522616830, +0.522616830205754336872861364785224694908468440); DVK(KP956723877, +0.956723877038460305821989399535483155872969262); DVK(KP906616052, +0.906616052148196230441134447086066874408359177); DVK(KP772036680, +0.772036680810363904029489473607579825330539880); DVK(KP845997307, +0.845997307939530944175097360758058292389769300); DVK(KP951056516, +0.951056516295153572116439333379382143405698634); DVK(KP921078979, +0.921078979742360627699756128143719920817673854); DVK(KP912575812, +0.912575812670962425556968549836277086778922727); DVK(KP982009705, +0.982009705009746369461829878184175962711969869); DVK(KP734762448, +0.734762448793050413546343770063151342619912334); DVK(KP494780565, +0.494780565770515410344588413655324772219443730); DVK(KP447533225, +0.447533225982656890041886979663652563063114397); DVK(KP269969613, +0.269969613759572083574752974412347470060951301); DVK(KP244189809, +0.244189809627953270309879511234821255780225091); DVK(KP667278218, +0.667278218140296670899089292254759909713898805); DVK(KP603558818, +0.603558818296015001454675132653458027918768137); DVK(KP522847744, +0.522847744331509716623755382187077770911012542); DVK(KP578046249, +0.578046249379945007321754579646815604023525655); DVK(KP987388751, +0.987388751065621252324603216482382109400433949); DVK(KP893101515, +0.893101515366181661711202267938416198338079437); DVK(KP120146378, +0.120146378570687701782758537356596213647956445); DVK(KP132830569, +0.132830569247582714407653942074819768844536507); DVK(KP869845200, +0.869845200362138853122720822420327157933056305); DVK(KP786782374, +0.786782374965295178365099601674911834788448471); DVK(KP066152395, +0.066152395967733048213034281011006031460903353); DVK(KP059835404, +0.059835404262124915169548397419498386427871950); DVK(KP559016994, +0.559016994374947424102293417182819058860154590); DVK(KP250000000, +0.250000000000000000000000000000000000000000000); DVK(KP618033988, +0.618033988749894848204586834365638117720309180); { INT m; R *x; x = ii; for (m = mb, W = W + (mb * ((TWVL / VL) * 48)); m < me; m = m + VL, x = x + (VL * ms), W = W + (TWVL * 48), MAKE_VOLATILE_STRIDE(25, rs)) { V T25, T1B, T2y, T1K, T2s, T23, T1S, T26, T20, T1X; { V T1O, T2X, Te, T3L, Td, T3Q, T3j, T3b, T2R, T2M, T2f, T27, T1y, T1H, T3M; V TW, TR, TK, T2B, T3n, T3e, T2U, T2F, T2i, T2a, Tz, T1C, T3N, TQ, T11; V T1b, T1c, T16; { V T1, T1g, T1i, T1p, T1k, T1m, Tb, T1N, T6, T1M; { V T7, T9, T2, T4, T1f, T1h, T1o; T1 = LD(&(x[0]), ms, &(x[0])); T7 = LD(&(x[WS(rs, 10)]), ms, &(x[0])); T9 = LD(&(x[WS(rs, 15)]), ms, &(x[WS(rs, 1)])); T2 = LD(&(x[WS(rs, 5)]), ms, &(x[WS(rs, 1)])); T4 = LD(&(x[WS(rs, 20)]), ms, &(x[0])); T1f = LD(&(x[WS(rs, 3)]), ms, &(x[WS(rs, 1)])); T1h = LD(&(x[WS(rs, 8)]), ms, &(x[0])); T1o = LD(&(x[WS(rs, 18)]), ms, &(x[0])); { V T8, Ta, T3, T5, T1j; T1j = LD(&(x[WS(rs, 23)]), ms, &(x[WS(rs, 1)])); T8 = BYTW(&(W[TWVL * 18]), T7); Ta = BYTW(&(W[TWVL * 28]), T9); T3 = BYTW(&(W[TWVL * 8]), T2); T5 = BYTW(&(W[TWVL * 38]), T4); T1g = BYTW(&(W[TWVL * 4]), T1f); T1i = BYTW(&(W[TWVL * 14]), T1h); T1p = BYTW(&(W[TWVL * 34]), T1o); T1k = BYTW(&(W[TWVL * 44]), T1j); T1m = LD(&(x[WS(rs, 13)]), ms, &(x[WS(rs, 1)])); Tb = VADD(T8, Ta); T1N = VSUB(T8, Ta); T6 = VADD(T3, T5); T1M = VSUB(T3, T5); } } { V T1v, T1l, Th, Tj, T1w, T1q, Tq, Tk, Tn, Tg; Tg = LD(&(x[WS(rs, 1)]), ms, &(x[WS(rs, 1)])); { V Tc, Ti, T1n, Tp; Ti = LD(&(x[WS(rs, 6)]), ms, &(x[0])); T1v = VSUB(T1i, T1k); T1l = VADD(T1i, T1k); T1n = BYTW(&(W[TWVL * 24]), T1m); Tp = LD(&(x[WS(rs, 16)]), ms, &(x[0])); T1O = VFMA(LDK(KP618033988), T1N, T1M); T2X = VFNMS(LDK(KP618033988), T1M, T1N); Te = VSUB(T6, Tb); Tc = VADD(T6, Tb); Th = BYTW(&(W[0]), Tg); Tj = BYTW(&(W[TWVL * 10]), Ti); T1w = VSUB(T1n, T1p); T1q = VADD(T1n, T1p); Tq = BYTW(&(W[TWVL * 30]), Tp); Tk = LD(&(x[WS(rs, 21)]), ms, &(x[WS(rs, 1)])); T3L = VADD(T1, Tc); Td = VFNMS(LDK(KP250000000), Tc, T1); Tn = LD(&(x[WS(rs, 11)]), ms, &(x[WS(rs, 1)])); } { V T1x, T2K, TM, TB, Tw, Tm, Tx, Tr, TI, T2L, T1u, TD, TF, TL; TL = LD(&(x[WS(rs, 4)]), ms, &(x[0])); { V T1t, Tl, To, TH, T1s, T1r, TA, TC; TA = LD(&(x[WS(rs, 24)]), ms, &(x[0])); T1r = VADD(T1l, T1q); T1t = VSUB(T1q, T1l); T1x = VFMA(LDK(KP618033988), T1w, T1v); T2K = VFNMS(LDK(KP618033988), T1v, T1w); Tl = BYTW(&(W[TWVL * 40]), Tk); To = BYTW(&(W[TWVL * 20]), Tn); TM = BYTW(&(W[TWVL * 6]), TL); TB = BYTW(&(W[TWVL * 46]), TA); TH = LD(&(x[WS(rs, 14)]), ms, &(x[0])); T1s = VFNMS(LDK(KP250000000), T1r, T1g); T3Q = VADD(T1g, T1r); TC = LD(&(x[WS(rs, 9)]), ms, &(x[WS(rs, 1)])); Tw = VSUB(Tj, Tl); Tm = VADD(Tj, Tl); Tx = VSUB(Tq, To); Tr = VADD(To, Tq); TI = BYTW(&(W[TWVL * 26]), TH); T2L = VFMA(LDK(KP559016994), T1t, T1s); T1u = VFNMS(LDK(KP559016994), T1t, T1s); TD = BYTW(&(W[TWVL * 16]), TC); TF = LD(&(x[WS(rs, 19)]), ms, &(x[WS(rs, 1)])); } { V Tu, Ty, T2E, TE, TN, TG, Tt, TV, Ts; TV = LD(&(x[WS(rs, 2)]), ms, &(x[0])); Ts = VADD(Tm, Tr); Tu = VSUB(Tm, Tr); Ty = VFNMS(LDK(KP618033988), Tx, Tw); T2E = VFMA(LDK(KP618033988), Tw, Tx); T3j = VFNMS(LDK(KP059835404), T2K, T2L); T3b = VFMA(LDK(KP066152395), T2L, T2K); T2R = VFNMS(LDK(KP786782374), T2K, T2L); T2M = VFMA(LDK(KP869845200), T2L, T2K); T2f = VFMA(LDK(KP132830569), T1u, T1x); T27 = VFNMS(LDK(KP120146378), T1x, T1u); T1y = VFNMS(LDK(KP893101515), T1x, T1u); T1H = VFMA(LDK(KP987388751), T1u, T1x); TE = VSUB(TB, TD); TN = VADD(TD, TB); TG = BYTW(&(W[TWVL * 36]), TF); Tt = VFNMS(LDK(KP250000000), Ts, Th); T3M = VADD(Th, Ts); TW = BYTW(&(W[TWVL * 2]), TV); { V TJ, TO, Tv, T2D, TY, T15, T10, T13, TP; { V TX, T14, TZ, T12; TX = LD(&(x[WS(rs, 7)]), ms, &(x[WS(rs, 1)])); T14 = LD(&(x[WS(rs, 17)]), ms, &(x[WS(rs, 1)])); TZ = LD(&(x[WS(rs, 22)]), ms, &(x[0])); T12 = LD(&(x[WS(rs, 12)]), ms, &(x[0])); TJ = VSUB(TG, TI); TO = VADD(TI, TG); Tv = VFMA(LDK(KP559016994), Tu, Tt); T2D = VFNMS(LDK(KP559016994), Tu, Tt); TY = BYTW(&(W[TWVL * 12]), TX); T15 = BYTW(&(W[TWVL * 32]), T14); T10 = BYTW(&(W[TWVL * 42]), TZ); T13 = BYTW(&(W[TWVL * 22]), T12); } TP = VADD(TN, TO); TR = VSUB(TN, TO); TK = VFMA(LDK(KP618033988), TJ, TE); T2B = VFNMS(LDK(KP618033988), TE, TJ); T3n = VFMA(LDK(KP578046249), T2D, T2E); T3e = VFNMS(LDK(KP522847744), T2E, T2D); T2U = VFNMS(LDK(KP987388751), T2D, T2E); T2F = VFMA(LDK(KP893101515), T2E, T2D); T2i = VFNMS(LDK(KP603558818), Ty, Tv); T2a = VFMA(LDK(KP667278218), Tv, Ty); Tz = VFNMS(LDK(KP244189809), Ty, Tv); T1C = VFMA(LDK(KP269969613), Tv, Ty); T3N = VADD(TM, TP); TQ = VFMS(LDK(KP250000000), TP, TM); T11 = VADD(TY, T10); T1b = VSUB(TY, T10); T1c = VSUB(T15, T13); T16 = VADD(T13, T15); } } } } } { V T2z, Tf, T3W, T3O, T1d, T2H, T3m, T2j, T2b, TT, T1D, T2G, T35, T2V, T2Z; V T3A, T3g, T2I, T1a, T3R, T3X; T2z = VFNMS(LDK(KP559016994), Te, Td); Tf = VFMA(LDK(KP559016994), Te, Td); { V TS, T2A, T17, T19; TS = VFNMS(LDK(KP559016994), TR, TQ); T2A = VFMA(LDK(KP559016994), TR, TQ); T3W = VSUB(T3M, T3N); T3O = VADD(T3M, T3N); T1d = VFNMS(LDK(KP618033988), T1c, T1b); T2H = VFMA(LDK(KP618033988), T1b, T1c); T17 = VADD(T11, T16); T19 = VSUB(T16, T11); { V T3f, T2T, T2C, T18, T3P; T3m = VFMA(LDK(KP447533225), T2B, T2A); T3f = VFNMS(LDK(KP494780565), T2A, T2B); T2T = VFNMS(LDK(KP132830569), T2A, T2B); T2C = VFMA(LDK(KP120146378), T2B, T2A); T2j = VFNMS(LDK(KP786782374), TK, TS); T2b = VFMA(LDK(KP869845200), TS, TK); TT = VFNMS(LDK(KP667278218), TS, TK); T1D = VFMA(LDK(KP603558818), TK, TS); T18 = VFNMS(LDK(KP250000000), T17, TW); T3P = VADD(TW, T17); T2G = VFMA(LDK(KP734762448), T2F, T2C); T35 = VFNMS(LDK(KP734762448), T2F, T2C); T2V = VFNMS(LDK(KP734762448), T2U, T2T); T2Z = VFMA(LDK(KP734762448), T2U, T2T); T3A = VFMA(LDK(KP982009705), T3f, T3e); T3g = VFNMS(LDK(KP982009705), T3f, T3e); T2I = VFMA(LDK(KP559016994), T19, T18); T1a = VFNMS(LDK(KP559016994), T19, T18); T3R = VADD(T3P, T3Q); T3X = VSUB(T3P, T3Q); } } { V T2n, T2t, T1V, T22, T2l, T2d, T1Q, T1I, T2w, T1A, T1F, T2q; { V T2k, T1G, T28, T2g, T3K, T3E, T3a, T34, T3x, T3H, T2c, TU, T1T, T1U, T1z; V T3o, T3t; T2n = VFNMS(LDK(KP912575812), T2j, T2i); T2k = VFMA(LDK(KP912575812), T2j, T2i); T3o = VFNMS(LDK(KP921078979), T3n, T3m); T3t = VFMA(LDK(KP921078979), T3n, T3m); { V T3c, T2Q, T2J, T3k, T1e; T3c = VFNMS(LDK(KP667278218), T2I, T2H); T2Q = VFNMS(LDK(KP059835404), T2H, T2I); T2J = VFMA(LDK(KP066152395), T2I, T2H); T3k = VFMA(LDK(KP603558818), T2H, T2I); T1G = VFMA(LDK(KP578046249), T1a, T1d); T1e = VFNMS(LDK(KP522847744), T1d, T1a); T28 = VFNMS(LDK(KP494780565), T1a, T1d); T2g = VFMA(LDK(KP447533225), T1d, T1a); { V T3U, T3S, T40, T3Y; T3U = VSUB(T3O, T3R); T3S = VADD(T3O, T3R); T40 = VMUL(LDK(KP951056516), VFNMS(LDK(KP618033988), T3W, T3X)); T3Y = VMUL(LDK(KP951056516), VFMA(LDK(KP618033988), T3X, T3W)); { V T3s, T3l, T2N, T36; T3s = VFNMS(LDK(KP845997307), T3k, T3j); T3l = VFMA(LDK(KP845997307), T3k, T3j); T2N = VFNMS(LDK(KP772036680), T2M, T2J); T36 = VFMA(LDK(KP772036680), T2M, T2J); { V T30, T2S, T3d, T3z, T3T; T30 = VFNMS(LDK(KP772036680), T2R, T2Q); T2S = VFMA(LDK(KP772036680), T2R, T2Q); T3d = VFNMS(LDK(KP845997307), T3c, T3b); T3z = VFMA(LDK(KP845997307), T3c, T3b); ST(&(x[0]), VADD(T3S, T3L), ms, &(x[0])); T3T = VFNMS(LDK(KP250000000), T3S, T3L); { V T3C, T3p, T2O, T37; T3C = VFMA(LDK(KP906616052), T3o, T3l); T3p = VFNMS(LDK(KP906616052), T3o, T3l); T2O = VFMA(LDK(KP956723877), T2N, T2G); T37 = VFMA(LDK(KP522616830), T2V, T36); { V T31, T2W, T3u, T3h; T31 = VFNMS(LDK(KP522616830), T2G, T30); T2W = VFMA(LDK(KP945422727), T2V, T2S); T3u = VFNMS(LDK(KP923225144), T3g, T3d); T3h = VFMA(LDK(KP923225144), T3g, T3d); { V T3I, T3B, T3V, T3Z; T3I = VFNMS(LDK(KP669429328), T3z, T3A); T3B = VFMA(LDK(KP570584518), T3A, T3z); T3V = VFMA(LDK(KP559016994), T3U, T3T); T3Z = VFNMS(LDK(KP559016994), T3U, T3T); { V T3y, T3q, T2P, T38; T3y = VFMA(LDK(KP262346850), T3p, T2X); T3q = VMUL(LDK(KP998026728), VFNMS(LDK(KP952936919), T2X, T3p)); T2P = VFMA(LDK(KP992114701), T2O, T2z); T38 = VFNMS(LDK(KP690983005), T37, T2S); { V T32, T2Y, T3v, T3F; T32 = VFMA(LDK(KP763932022), T31, T2N); T2Y = VMUL(LDK(KP998026728), VFMA(LDK(KP952936919), T2X, T2W)); T3v = VFNMS(LDK(KP997675361), T3u, T3t); T3F = VFNMS(LDK(KP904508497), T3u, T3s); { V T3i, T3r, T3J, T3D; T3i = VFMA(LDK(KP949179823), T3h, T2z); T3r = VFNMS(LDK(KP237294955), T3h, T2z); T3J = VFNMS(LDK(KP669429328), T3C, T3I); T3D = VFMA(LDK(KP618033988), T3C, T3B); ST(&(x[WS(rs, 20)]), VFNMSI(T3Y, T3V), ms, &(x[0])); ST(&(x[WS(rs, 5)]), VFMAI(T3Y, T3V), ms, &(x[WS(rs, 1)])); ST(&(x[WS(rs, 15)]), VFMAI(T40, T3Z), ms, &(x[WS(rs, 1)])); ST(&(x[WS(rs, 10)]), VFNMSI(T40, T3Z), ms, &(x[0])); { V T39, T33, T3w, T3G; T39 = VFMA(LDK(KP855719849), T38, T35); T33 = VFNMS(LDK(KP855719849), T32, T2Z); ST(&(x[WS(rs, 3)]), VFMAI(T2Y, T2P), ms, &(x[WS(rs, 1)])); ST(&(x[WS(rs, 22)]), VFNMSI(T2Y, T2P), ms, &(x[0])); T3w = VFMA(LDK(KP560319534), T3v, T3s); T3G = VFNMS(LDK(KP681693190), T3F, T3t); ST(&(x[WS(rs, 2)]), VFMAI(T3q, T3i), ms, &(x[0])); ST(&(x[WS(rs, 23)]), VFNMSI(T3q, T3i), ms, &(x[WS(rs, 1)])); T3K = VMUL(LDK(KP951056516), VFNMS(LDK(KP876306680), T3J, T3y)); T3E = VMUL(LDK(KP951056516), VFNMS(LDK(KP949179823), T3D, T3y)); T3a = VMUL(LDK(KP951056516), VFNMS(LDK(KP992114701), T39, T2X)); T34 = VFMA(LDK(KP897376177), T33, T2z); T3x = VFNMS(LDK(KP949179823), T3w, T3r); T3H = VFNMS(LDK(KP860541664), T3G, T3r); T2t = VFNMS(LDK(KP912575812), T2b, T2a); T2c = VFMA(LDK(KP912575812), T2b, T2a); TU = VFMA(LDK(KP829049696), TT, Tz); T1T = VFNMS(LDK(KP829049696), TT, Tz); T1U = VFNMS(LDK(KP831864738), T1y, T1e); T1z = VFMA(LDK(KP831864738), T1y, T1e); } } } } } } } } } } } { V T2o, T2h, T29, T2u, T2v, T2p; T2o = VFNMS(LDK(KP958953096), T2g, T2f); T2h = VFMA(LDK(KP958953096), T2g, T2f); ST(&(x[WS(rs, 17)]), VFNMSI(T3a, T34), ms, &(x[WS(rs, 1)])); ST(&(x[WS(rs, 8)]), VFMAI(T3a, T34), ms, &(x[0])); ST(&(x[WS(rs, 13)]), VFMAI(T3E, T3x), ms, &(x[WS(rs, 1)])); ST(&(x[WS(rs, 12)]), VFNMSI(T3E, T3x), ms, &(x[0])); ST(&(x[WS(rs, 7)]), VFNMSI(T3K, T3H), ms, &(x[WS(rs, 1)])); ST(&(x[WS(rs, 18)]), VFMAI(T3K, T3H), ms, &(x[0])); T1V = VFMA(LDK(KP559154169), T1U, T1T); T22 = VFNMS(LDK(KP683113946), T1T, T1U); T29 = VFNMS(LDK(KP867381224), T28, T27); T2u = VFMA(LDK(KP867381224), T28, T27); T2l = VFMA(LDK(KP894834959), T2k, T2h); T2v = VFMA(LDK(KP447417479), T2k, T2u); T2d = VFNMS(LDK(KP809385824), T2c, T29); T2p = VFMA(LDK(KP447417479), T2c, T2o); T1Q = VFMA(LDK(KP831864738), T1H, T1G); T1I = VFNMS(LDK(KP831864738), T1H, T1G); T2w = VFNMS(LDK(KP763932022), T2v, T2h); T1A = VFMA(LDK(KP904730450), T1z, TU); T1F = VFNMS(LDK(KP904730450), T1z, TU); T2q = VFMA(LDK(KP690983005), T2p, T29); } } { V T2e, T1E, T1P, T2m; T2e = VFNMS(LDK(KP992114701), T2d, Tf); T1E = VFMA(LDK(KP916574801), T1D, T1C); T1P = VFNMS(LDK(KP916574801), T1D, T1C); T2m = VMUL(LDK(KP951056516), VFNMS(LDK(KP992114701), T2l, T1O)); { V T1J, T2r, T1R, T1W, T1Z, T2x; T2x = VFNMS(LDK(KP999544308), T2w, T2t); T1J = VFNMS(LDK(KP904730450), T1I, T1F); T25 = VFMA(LDK(KP968583161), T1A, Tf); T1B = VFNMS(LDK(KP242145790), T1A, Tf); T2r = VFNMS(LDK(KP999544308), T2q, T2n); T1R = VFMA(LDK(KP904730450), T1Q, T1P); T1W = VFNMS(LDK(KP904730450), T1Q, T1P); T1Z = VADD(T1E, T1F); ST(&(x[WS(rs, 21)]), VFMAI(T2m, T2e), ms, &(x[WS(rs, 1)])); ST(&(x[WS(rs, 4)]), VFNMSI(T2m, T2e), ms, &(x[0])); T2y = VMUL(LDK(KP951056516), VFNMS(LDK(KP803003575), T2x, T1O)); T1K = VFNMS(LDK(KP618033988), T1J, T1E); T2s = VFNMS(LDK(KP803003575), T2r, Tf); T23 = VFMA(LDK(KP617882369), T1W, T22); T1S = VFNMS(LDK(KP242145790), T1R, T1O); T26 = VMUL(LDK(KP951056516), VFMA(LDK(KP968583161), T1R, T1O)); T20 = VFNMS(LDK(KP683113946), T1Z, T1I); T1X = VFMA(LDK(KP559016994), T1W, T1V); } } } } } { V T1L, T24, T21, T1Y; T1L = VFNMS(LDK(KP876091699), T1K, T1B); ST(&(x[WS(rs, 16)]), VFMAI(T2y, T2s), ms, &(x[0])); ST(&(x[WS(rs, 9)]), VFNMSI(T2y, T2s), ms, &(x[WS(rs, 1)])); T24 = VMUL(LDK(KP951056516), VFNMS(LDK(KP876306680), T23, T1S)); ST(&(x[WS(rs, 24)]), VFNMSI(T26, T25), ms, &(x[0])); ST(&(x[WS(rs, 1)]), VFMAI(T26, T25), ms, &(x[WS(rs, 1)])); T21 = VFMA(LDK(KP792626838), T20, T1B); T1Y = VMUL(LDK(KP951056516), VFMA(LDK(KP968583161), T1X, T1S)); ST(&(x[WS(rs, 11)]), VFMAI(T24, T21), ms, &(x[WS(rs, 1)])); ST(&(x[WS(rs, 14)]), VFNMSI(T24, T21), ms, &(x[0])); ST(&(x[WS(rs, 19)]), VFNMSI(T1Y, T1L), ms, &(x[WS(rs, 1)])); ST(&(x[WS(rs, 6)]), VFMAI(T1Y, T1L), ms, &(x[0])); } } } VLEAVE(); } static const tw_instr twinstr[] = { VTW(0, 1), VTW(0, 2), VTW(0, 3), VTW(0, 4), VTW(0, 5), VTW(0, 6), VTW(0, 7), VTW(0, 8), VTW(0, 9), VTW(0, 10), VTW(0, 11), VTW(0, 12), VTW(0, 13), VTW(0, 14), VTW(0, 15), VTW(0, 16), VTW(0, 17), VTW(0, 18), VTW(0, 19), VTW(0, 20), VTW(0, 21), VTW(0, 22), VTW(0, 23), VTW(0, 24), {TW_NEXT, VL, 0} }; static const ct_desc desc = { 25, XSIMD_STRING("t2bv_25"), twinstr, &GENUS, {67, 60, 181, 0}, 0, 0, 0 }; void XSIMD(codelet_t2bv_25) (planner *p) { X(kdft_dit_register) (p, t2bv_25, &desc); } #else /* HAVE_FMA */ /* Generated by: ../../../genfft/gen_twiddle_c.native -simd -compact -variables 4 -pipeline-latency 8 -n 25 -name t2bv_25 -include t2b.h -sign 1 */ /* * This function contains 248 FP additions, 188 FP multiplications, * (or, 171 additions, 111 multiplications, 77 fused multiply/add), * 100 stack variables, 40 constants, and 50 memory accesses */ #include "t2b.h" static void t2bv_25(R *ri, R *ii, const R *W, stride rs, INT mb, INT me, INT ms) { DVK(KP497379774, +0.497379774329709576484567492012895936835134813); DVK(KP968583161, +0.968583161128631119490168375464735813836012403); DVK(KP248689887, +0.248689887164854788242283746006447968417567406); DVK(KP1_937166322, +1.937166322257262238980336750929471627672024806); DVK(KP809016994, +0.809016994374947424102293417182819058860154590); DVK(KP309016994, +0.309016994374947424102293417182819058860154590); DVK(KP1_688655851, +1.688655851004030157097116127933363010763318483); DVK(KP535826794, +0.535826794978996618271308767867639978063575346); DVK(KP425779291, +0.425779291565072648862502445744251703979973042); DVK(KP1_809654104, +1.809654104932039055427337295865395187940827822); DVK(KP963507348, +0.963507348203430549974383005744259307057084020); DVK(KP876306680, +0.876306680043863587308115903922062583399064238); DVK(KP844327925, +0.844327925502015078548558063966681505381659241); DVK(KP1_071653589, +1.071653589957993236542617535735279956127150691); DVK(KP481753674, +0.481753674101715274987191502872129653528542010); DVK(KP1_752613360, +1.752613360087727174616231807844125166798128477); DVK(KP851558583, +0.851558583130145297725004891488503407959946084); DVK(KP904827052, +0.904827052466019527713668647932697593970413911); DVK(KP125333233, +0.125333233564304245373118759816508793942918247); DVK(KP1_984229402, +1.984229402628955662099586085571557042906073418); DVK(KP1_457937254, +1.457937254842823046293460638110518222745143328); DVK(KP684547105, +0.684547105928688673732283357621209269889519233); DVK(KP637423989, +0.637423989748689710176712811676016195434917298); DVK(KP1_541026485, +1.541026485551578461606019272792355694543335344); DVK(KP062790519, +0.062790519529313376076178224565631133122484832); DVK(KP1_996053456, +1.996053456856543123904673613726901106673810439); DVK(KP770513242, +0.770513242775789230803009636396177847271667672); DVK(KP1_274847979, +1.274847979497379420353425623352032390869834596); DVK(KP125581039, +0.125581039058626752152356449131262266244969664); DVK(KP998026728, +0.998026728428271561952336806863450553336905220); DVK(KP992114701, +0.992114701314477831049793042785778521453036709); DVK(KP250666467, +0.250666467128608490746237519633017587885836494); DVK(KP728968627, +0.728968627421411523146730319055259111372571664); DVK(KP1_369094211, +1.369094211857377347464566715242418539779038465); DVK(KP293892626, +0.293892626146236564584352977319536384298826219); DVK(KP475528258, +0.475528258147576786058219666689691071702849317); DVK(KP951056516, +0.951056516295153572116439333379382143405698634); DVK(KP250000000, +0.250000000000000000000000000000000000000000000); DVK(KP587785252, +0.587785252292473129168705954639072768597652438); DVK(KP559016994, +0.559016994374947424102293417182819058860154590); { INT m; R *x; x = ii; for (m = mb, W = W + (mb * ((TWVL / VL) * 48)); m < me; m = m + VL, x = x + (VL * ms), W = W + (TWVL * 48), MAKE_VOLATILE_STRIDE(25, rs)) { V T1A, T1z, T1R, T1S, T1B, T1C, T1Q, T2L, T1l, T2v, T1i, T3e, T2u, Tb, T2i; V Tj, T3b, T2h, Tv, T2k, TD, T3a, T2l, T11, T2s, TY, T3d, T2r; { V T1v, T1x, T1y, T1q, T1s, T1t, T1P; T1A = LD(&(x[0]), ms, &(x[0])); { V T1u, T1w, T1p, T1r; T1u = LD(&(x[WS(rs, 10)]), ms, &(x[0])); T1v = BYTW(&(W[TWVL * 18]), T1u); T1w = LD(&(x[WS(rs, 15)]), ms, &(x[WS(rs, 1)])); T1x = BYTW(&(W[TWVL * 28]), T1w); T1y = VADD(T1v, T1x); T1p = LD(&(x[WS(rs, 5)]), ms, &(x[WS(rs, 1)])); T1q = BYTW(&(W[TWVL * 8]), T1p); T1r = LD(&(x[WS(rs, 20)]), ms, &(x[0])); T1s = BYTW(&(W[TWVL * 38]), T1r); T1t = VADD(T1q, T1s); } T1z = VMUL(LDK(KP559016994), VSUB(T1t, T1y)); T1R = VSUB(T1v, T1x); T1S = VMUL(LDK(KP587785252), T1R); T1B = VADD(T1t, T1y); T1C = VFNMS(LDK(KP250000000), T1B, T1A); T1P = VSUB(T1q, T1s); T1Q = VMUL(LDK(KP951056516), T1P); T2L = VMUL(LDK(KP587785252), T1P); } { V T1f, T19, T1b, T1c, T14, T16, T17, T1e; T1e = LD(&(x[WS(rs, 3)]), ms, &(x[WS(rs, 1)])); T1f = BYTW(&(W[TWVL * 4]), T1e); { V T18, T1a, T13, T15; T18 = LD(&(x[WS(rs, 13)]), ms, &(x[WS(rs, 1)])); T19 = BYTW(&(W[TWVL * 24]), T18); T1a = LD(&(x[WS(rs, 18)]), ms, &(x[0])); T1b = BYTW(&(W[TWVL * 34]), T1a); T1c = VADD(T19, T1b); T13 = LD(&(x[WS(rs, 8)]), ms, &(x[0])); T14 = BYTW(&(W[TWVL * 14]), T13); T15 = LD(&(x[WS(rs, 23)]), ms, &(x[WS(rs, 1)])); T16 = BYTW(&(W[TWVL * 44]), T15); T17 = VADD(T14, T16); } { V T1j, T1k, T1d, T1g, T1h; T1j = VSUB(T14, T16); T1k = VSUB(T19, T1b); T1l = VFMA(LDK(KP475528258), T1j, VMUL(LDK(KP293892626), T1k)); T2v = VFNMS(LDK(KP475528258), T1k, VMUL(LDK(KP293892626), T1j)); T1d = VMUL(LDK(KP559016994), VSUB(T17, T1c)); T1g = VADD(T17, T1c); T1h = VFNMS(LDK(KP250000000), T1g, T1f); T1i = VADD(T1d, T1h); T3e = VADD(T1f, T1g); T2u = VSUB(T1h, T1d); } } { V Tg, T7, T9, Td, T2, T4, Tc, Tf; Tf = LD(&(x[WS(rs, 4)]), ms, &(x[0])); Tg = BYTW(&(W[TWVL * 6]), Tf); { V T6, T8, T1, T3; T6 = LD(&(x[WS(rs, 14)]), ms, &(x[0])); T7 = BYTW(&(W[TWVL * 26]), T6); T8 = LD(&(x[WS(rs, 19)]), ms, &(x[WS(rs, 1)])); T9 = BYTW(&(W[TWVL * 36]), T8); Td = VADD(T7, T9); T1 = LD(&(x[WS(rs, 9)]), ms, &(x[WS(rs, 1)])); T2 = BYTW(&(W[TWVL * 16]), T1); T3 = LD(&(x[WS(rs, 24)]), ms, &(x[0])); T4 = BYTW(&(W[TWVL * 46]), T3); Tc = VADD(T2, T4); } { V T5, Ta, Te, Th, Ti; T5 = VSUB(T2, T4); Ta = VSUB(T7, T9); Tb = VFMA(LDK(KP475528258), T5, VMUL(LDK(KP293892626), Ta)); T2i = VFNMS(LDK(KP475528258), Ta, VMUL(LDK(KP293892626), T5)); Te = VMUL(LDK(KP559016994), VSUB(Tc, Td)); Th = VADD(Tc, Td); Ti = VFNMS(LDK(KP250000000), Th, Tg); Tj = VADD(Te, Ti); T3b = VADD(Tg, Th); T2h = VSUB(Ti, Te); } } { V TA, Tr, Tt, Tx, Tm, To, Tw, Tz; Tz = LD(&(x[WS(rs, 1)]), ms, &(x[WS(rs, 1)])); TA = BYTW(&(W[0]), Tz); { V Tq, Ts, Tl, Tn; Tq = LD(&(x[WS(rs, 11)]), ms, &(x[WS(rs, 1)])); Tr = BYTW(&(W[TWVL * 20]), Tq); Ts = LD(&(x[WS(rs, 16)]), ms, &(x[0])); Tt = BYTW(&(W[TWVL * 30]), Ts); Tx = VADD(Tr, Tt); Tl = LD(&(x[WS(rs, 6)]), ms, &(x[0])); Tm = BYTW(&(W[TWVL * 10]), Tl); Tn = LD(&(x[WS(rs, 21)]), ms, &(x[WS(rs, 1)])); To = BYTW(&(W[TWVL * 40]), Tn); Tw = VADD(Tm, To); } { V Tp, Tu, Ty, TB, TC; Tp = VSUB(Tm, To); Tu = VSUB(Tr, Tt); Tv = VFMA(LDK(KP475528258), Tp, VMUL(LDK(KP293892626), Tu)); T2k = VFNMS(LDK(KP475528258), Tu, VMUL(LDK(KP293892626), Tp)); Ty = VMUL(LDK(KP559016994), VSUB(Tw, Tx)); TB = VADD(Tw, Tx); TC = VFNMS(LDK(KP250000000), TB, TA); TD = VADD(Ty, TC); T3a = VADD(TA, TB); T2l = VSUB(TC, Ty); } } { V TV, TP, TR, TS, TK, TM, TN, TU; TU = LD(&(x[WS(rs, 2)]), ms, &(x[0])); TV = BYTW(&(W[TWVL * 2]), TU); { V TO, TQ, TJ, TL; TO = LD(&(x[WS(rs, 12)]), ms, &(x[0])); TP = BYTW(&(W[TWVL * 22]), TO); TQ = LD(&(x[WS(rs, 17)]), ms, &(x[WS(rs, 1)])); TR = BYTW(&(W[TWVL * 32]), TQ); TS = VADD(TP, TR); TJ = LD(&(x[WS(rs, 7)]), ms, &(x[WS(rs, 1)])); TK = BYTW(&(W[TWVL * 12]), TJ); TL = LD(&(x[WS(rs, 22)]), ms, &(x[0])); TM = BYTW(&(W[TWVL * 42]), TL); TN = VADD(TK, TM); } { V TZ, T10, TT, TW, TX; TZ = VSUB(TK, TM); T10 = VSUB(TP, TR); T11 = VFMA(LDK(KP475528258), TZ, VMUL(LDK(KP293892626), T10)); T2s = VFNMS(LDK(KP475528258), T10, VMUL(LDK(KP293892626), TZ)); TT = VMUL(LDK(KP559016994), VSUB(TN, TS)); TW = VADD(TN, TS); TX = VFNMS(LDK(KP250000000), TW, TV); TY = VADD(TT, TX); T3d = VADD(TV, TW); T2r = VSUB(TX, TT); } } { V T3g, T3o, T3k, T3l, T3j, T3m, T3p, T3n; { V T3c, T3f, T3h, T3i; T3c = VSUB(T3a, T3b); T3f = VSUB(T3d, T3e); T3g = VBYI(VFMA(LDK(KP951056516), T3c, VMUL(LDK(KP587785252), T3f))); T3o = VBYI(VFNMS(LDK(KP951056516), T3f, VMUL(LDK(KP587785252), T3c))); T3k = VADD(T1A, T1B); T3h = VADD(T3a, T3b); T3i = VADD(T3d, T3e); T3l = VADD(T3h, T3i); T3j = VMUL(LDK(KP559016994), VSUB(T3h, T3i)); T3m = VFNMS(LDK(KP250000000), T3l, T3k); } ST(&(x[0]), VADD(T3k, T3l), ms, &(x[0])); T3p = VSUB(T3m, T3j); ST(&(x[WS(rs, 10)]), VADD(T3o, T3p), ms, &(x[0])); ST(&(x[WS(rs, 15)]), VSUB(T3p, T3o), ms, &(x[WS(rs, 1)])); T3n = VADD(T3j, T3m); ST(&(x[WS(rs, 5)]), VADD(T3g, T3n), ms, &(x[WS(rs, 1)])); ST(&(x[WS(rs, 20)]), VSUB(T3n, T3g), ms, &(x[0])); } { V T2z, T2M, T2U, T2V, T2W, T34, T35, T36, T2X, T2Y, T2Z, T31, T32, T33, T2n; V T2N, T2E, T2K, T2y, T2H, T2A, T2G, T38, T39; T2z = VSUB(T1C, T1z); T2M = VFNMS(LDK(KP951056516), T1R, T2L); T2U = VFMA(LDK(KP1_369094211), T2k, VMUL(LDK(KP728968627), T2l)); T2V = VFNMS(LDK(KP992114701), T2h, VMUL(LDK(KP250666467), T2i)); T2W = VADD(T2U, T2V); T34 = VFNMS(LDK(KP125581039), T2s, VMUL(LDK(KP998026728), T2r)); T35 = VFMA(LDK(KP1_274847979), T2v, VMUL(LDK(KP770513242), T2u)); T36 = VADD(T34, T35); T2X = VFMA(LDK(KP1_996053456), T2s, VMUL(LDK(KP062790519), T2r)); T2Y = VFNMS(LDK(KP637423989), T2u, VMUL(LDK(KP1_541026485), T2v)); T2Z = VADD(T2X, T2Y); T31 = VFNMS(LDK(KP1_457937254), T2k, VMUL(LDK(KP684547105), T2l)); T32 = VFMA(LDK(KP1_984229402), T2i, VMUL(LDK(KP125333233), T2h)); T33 = VADD(T31, T32); { V T2j, T2m, T2I, T2C, T2D, T2J; T2j = VFNMS(LDK(KP851558583), T2i, VMUL(LDK(KP904827052), T2h)); T2m = VFMA(LDK(KP1_752613360), T2k, VMUL(LDK(KP481753674), T2l)); T2I = VADD(T2m, T2j); T2C = VFMA(LDK(KP1_071653589), T2s, VMUL(LDK(KP844327925), T2r)); T2D = VFMA(LDK(KP125581039), T2v, VMUL(LDK(KP998026728), T2u)); T2J = VADD(T2C, T2D); T2n = VSUB(T2j, T2m); T2N = VADD(T2I, T2J); T2E = VSUB(T2C, T2D); T2K = VMUL(LDK(KP559016994), VSUB(T2I, T2J)); } { V T2o, T2p, T2q, T2t, T2w, T2x; T2o = VFNMS(LDK(KP963507348), T2k, VMUL(LDK(KP876306680), T2l)); T2p = VFMA(LDK(KP1_809654104), T2i, VMUL(LDK(KP425779291), T2h)); T2q = VSUB(T2o, T2p); T2t = VFNMS(LDK(KP1_688655851), T2s, VMUL(LDK(KP535826794), T2r)); T2w = VFNMS(LDK(KP1_996053456), T2v, VMUL(LDK(KP062790519), T2u)); T2x = VADD(T2t, T2w); T2y = VMUL(LDK(KP559016994), VSUB(T2q, T2x)); T2H = VSUB(T2t, T2w); T2A = VADD(T2q, T2x); T2G = VADD(T2o, T2p); } { V T2S, T2T, T30, T37; T2S = VADD(T2z, T2A); T2T = VBYI(VADD(T2M, T2N)); ST(&(x[WS(rs, 23)]), VSUB(T2S, T2T), ms, &(x[WS(rs, 1)])); ST(&(x[WS(rs, 2)]), VADD(T2S, T2T), ms, &(x[0])); T30 = VADD(T2z, VADD(T2W, T2Z)); T37 = VBYI(VSUB(VADD(T33, T36), T2M)); ST(&(x[WS(rs, 22)]), VSUB(T30, T37), ms, &(x[0])); ST(&(x[WS(rs, 3)]), VADD(T30, T37), ms, &(x[WS(rs, 1)])); } T38 = VBYI(VSUB(VFMA(LDK(KP951056516), VSUB(T2U, T2V), VFMA(LDK(KP309016994), T33, VFNMS(LDK(KP809016994), T36, VMUL(LDK(KP587785252), VSUB(T2X, T2Y))))), T2M)); T39 = VFMA(LDK(KP309016994), T2W, VFMA(LDK(KP951056516), VSUB(T32, T31), VFMA(LDK(KP587785252), VSUB(T35, T34), VFNMS(LDK(KP809016994), T2Z, T2z)))); ST(&(x[WS(rs, 8)]), VADD(T38, T39), ms, &(x[0])); ST(&(x[WS(rs, 17)]), VSUB(T39, T38), ms, &(x[WS(rs, 1)])); { V T2F, T2Q, T2P, T2R, T2B, T2O; T2B = VFNMS(LDK(KP250000000), T2A, T2z); T2F = VFMA(LDK(KP951056516), T2n, VADD(T2y, VFNMS(LDK(KP587785252), T2E, T2B))); T2Q = VFMA(LDK(KP587785252), T2n, VFMA(LDK(KP951056516), T2E, VSUB(T2B, T2y))); T2O = VFNMS(LDK(KP250000000), T2N, T2M); T2P = VBYI(VADD(VFMA(LDK(KP951056516), T2G, VMUL(LDK(KP587785252), T2H)), VADD(T2K, T2O))); T2R = VBYI(VADD(VFNMS(LDK(KP951056516), T2H, VMUL(LDK(KP587785252), T2G)), VSUB(T2O, T2K))); ST(&(x[WS(rs, 18)]), VSUB(T2F, T2P), ms, &(x[0])); ST(&(x[WS(rs, 12)]), VADD(T2Q, T2R), ms, &(x[0])); ST(&(x[WS(rs, 7)]), VADD(T2F, T2P), ms, &(x[WS(rs, 1)])); ST(&(x[WS(rs, 13)]), VSUB(T2Q, T2R), ms, &(x[WS(rs, 1)])); } } { V T1D, T1T, T21, T22, T23, T2b, T2c, T2d, T24, T25, T26, T28, T29, T2a, TF; V T1U, T1I, T1O, T1o, T1L, T1E, T1K, T2f, T2g; T1D = VADD(T1z, T1C); T1T = VADD(T1Q, T1S); T21 = VFMA(LDK(KP1_688655851), Tv, VMUL(LDK(KP535826794), TD)); T22 = VFMA(LDK(KP1_541026485), Tb, VMUL(LDK(KP637423989), Tj)); T23 = VSUB(T21, T22); T2b = VFMA(LDK(KP851558583), T11, VMUL(LDK(KP904827052), TY)); T2c = VFMA(LDK(KP1_984229402), T1l, VMUL(LDK(KP125333233), T1i)); T2d = VADD(T2b, T2c); T24 = VFNMS(LDK(KP425779291), TY, VMUL(LDK(KP1_809654104), T11)); T25 = VFNMS(LDK(KP992114701), T1i, VMUL(LDK(KP250666467), T1l)); T26 = VADD(T24, T25); T28 = VFNMS(LDK(KP1_071653589), Tv, VMUL(LDK(KP844327925), TD)); T29 = VFNMS(LDK(KP770513242), Tj, VMUL(LDK(KP1_274847979), Tb)); T2a = VADD(T28, T29); { V Tk, TE, T1M, T1G, T1H, T1N; Tk = VFMA(LDK(KP1_071653589), Tb, VMUL(LDK(KP844327925), Tj)); TE = VFMA(LDK(KP1_937166322), Tv, VMUL(LDK(KP248689887), TD)); T1M = VADD(TE, Tk); T1G = VFMA(LDK(KP1_752613360), T11, VMUL(LDK(KP481753674), TY)); T1H = VFMA(LDK(KP1_457937254), T1l, VMUL(LDK(KP684547105), T1i)); T1N = VADD(T1G, T1H); TF = VSUB(Tk, TE); T1U = VADD(T1M, T1N); T1I = VSUB(T1G, T1H); T1O = VMUL(LDK(KP559016994), VSUB(T1M, T1N)); } { V TG, TH, TI, T12, T1m, T1n; TG = VFNMS(LDK(KP497379774), Tv, VMUL(LDK(KP968583161), TD)); TH = VFNMS(LDK(KP1_688655851), Tb, VMUL(LDK(KP535826794), Tj)); TI = VADD(TG, TH); T12 = VFNMS(LDK(KP963507348), T11, VMUL(LDK(KP876306680), TY)); T1m = VFNMS(LDK(KP1_369094211), T1l, VMUL(LDK(KP728968627), T1i)); T1n = VADD(T12, T1m); T1o = VMUL(LDK(KP559016994), VSUB(TI, T1n)); T1L = VSUB(T12, T1m); T1E = VADD(TI, T1n); T1K = VSUB(TG, TH); } { V T1Z, T20, T27, T2e; T1Z = VADD(T1D, T1E); T20 = VBYI(VADD(T1T, T1U)); ST(&(x[WS(rs, 24)]), VSUB(T1Z, T20), ms, &(x[0])); ST(&(x[WS(rs, 1)]), VADD(T1Z, T20), ms, &(x[WS(rs, 1)])); T27 = VADD(T1D, VADD(T23, T26)); T2e = VBYI(VSUB(VADD(T2a, T2d), T1T)); ST(&(x[WS(rs, 21)]), VSUB(T27, T2e), ms, &(x[WS(rs, 1)])); ST(&(x[WS(rs, 4)]), VADD(T27, T2e), ms, &(x[0])); } T2f = VBYI(VSUB(VFMA(LDK(KP309016994), T2a, VFMA(LDK(KP951056516), VADD(T21, T22), VFNMS(LDK(KP809016994), T2d, VMUL(LDK(KP587785252), VSUB(T24, T25))))), T1T)); T2g = VFMA(LDK(KP951056516), VSUB(T29, T28), VFMA(LDK(KP309016994), T23, VFMA(LDK(KP587785252), VSUB(T2c, T2b), VFNMS(LDK(KP809016994), T26, T1D)))); ST(&(x[WS(rs, 9)]), VADD(T2f, T2g), ms, &(x[WS(rs, 1)])); ST(&(x[WS(rs, 16)]), VSUB(T2g, T2f), ms, &(x[0])); { V T1J, T1X, T1W, T1Y, T1F, T1V; T1F = VFNMS(LDK(KP250000000), T1E, T1D); T1J = VFMA(LDK(KP951056516), TF, VADD(T1o, VFNMS(LDK(KP587785252), T1I, T1F))); T1X = VFMA(LDK(KP587785252), TF, VFMA(LDK(KP951056516), T1I, VSUB(T1F, T1o))); T1V = VFNMS(LDK(KP250000000), T1U, T1T); T1W = VBYI(VADD(VFMA(LDK(KP951056516), T1K, VMUL(LDK(KP587785252), T1L)), VADD(T1O, T1V))); T1Y = VBYI(VADD(VFNMS(LDK(KP951056516), T1L, VMUL(LDK(KP587785252), T1K)), VSUB(T1V, T1O))); ST(&(x[WS(rs, 19)]), VSUB(T1J, T1W), ms, &(x[WS(rs, 1)])); ST(&(x[WS(rs, 11)]), VADD(T1X, T1Y), ms, &(x[WS(rs, 1)])); ST(&(x[WS(rs, 6)]), VADD(T1J, T1W), ms, &(x[0])); ST(&(x[WS(rs, 14)]), VSUB(T1X, T1Y), ms, &(x[0])); } } } } VLEAVE(); } static const tw_instr twinstr[] = { VTW(0, 1), VTW(0, 2), VTW(0, 3), VTW(0, 4), VTW(0, 5), VTW(0, 6), VTW(0, 7), VTW(0, 8), VTW(0, 9), VTW(0, 10), VTW(0, 11), VTW(0, 12), VTW(0, 13), VTW(0, 14), VTW(0, 15), VTW(0, 16), VTW(0, 17), VTW(0, 18), VTW(0, 19), VTW(0, 20), VTW(0, 21), VTW(0, 22), VTW(0, 23), VTW(0, 24), {TW_NEXT, VL, 0} }; static const ct_desc desc = { 25, XSIMD_STRING("t2bv_25"), twinstr, &GENUS, {171, 111, 77, 0}, 0, 0, 0 }; void XSIMD(codelet_t2bv_25) (planner *p) { X(kdft_dit_register) (p, t2bv_25, &desc); } #endif /* HAVE_FMA */ fftw-3.3.3/dft/simd/common/t3bv_4.c0000644000175000017500000001103612054410566013661 00000000000000/* * Copyright (c) 2003, 2007-11 Matteo Frigo * Copyright (c) 2003, 2007-11 Massachusetts Institute of Technology * * This program is free software; you can redistribute it and/or modify * it under the terms of the GNU General Public License as published by * the Free Software Foundation; either version 2 of the License, or * (at your option) any later version. * * This program is distributed in the hope that it will be useful, * but WITHOUT ANY WARRANTY; without even the implied warranty of * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the * GNU General Public License for more details. * * You should have received a copy of the GNU General Public License * along with this program; if not, write to the Free Software * Foundation, Inc., 51 Franklin Street, Fifth Floor, Boston, MA 02110-1301 USA * */ /* This file was automatically generated --- DO NOT EDIT */ /* Generated on Sun Nov 25 07:39:18 EST 2012 */ #include "codelet-dft.h" #ifdef HAVE_FMA /* Generated by: ../../../genfft/gen_twiddle_c.native -fma -reorder-insns -schedule-for-pipeline -simd -compact -variables 4 -pipeline-latency 8 -twiddle-log3 -precompute-twiddles -no-generate-bytw -n 4 -name t3bv_4 -include t3b.h -sign 1 */ /* * This function contains 12 FP additions, 10 FP multiplications, * (or, 10 additions, 8 multiplications, 2 fused multiply/add), * 16 stack variables, 0 constants, and 8 memory accesses */ #include "t3b.h" static void t3bv_4(R *ri, R *ii, const R *W, stride rs, INT mb, INT me, INT ms) { { INT m; R *x; x = ii; for (m = mb, W = W + (mb * ((TWVL / VL) * 4)); m < me; m = m + VL, x = x + (VL * ms), W = W + (TWVL * 4), MAKE_VOLATILE_STRIDE(4, rs)) { V T2, T3, T1, Ta, T5, T8; T2 = LDW(&(W[0])); T3 = LDW(&(W[TWVL * 2])); T1 = LD(&(x[0]), ms, &(x[0])); Ta = LD(&(x[WS(rs, 3)]), ms, &(x[WS(rs, 1)])); T5 = LD(&(x[WS(rs, 2)]), ms, &(x[0])); T8 = LD(&(x[WS(rs, 1)]), ms, &(x[WS(rs, 1)])); { V T4, Tb, T9, T6; T4 = VZMULJ(T2, T3); Tb = VZMUL(T3, Ta); T9 = VZMUL(T2, T8); T6 = VZMUL(T4, T5); { V Tc, Te, T7, Td; Tc = VSUB(T9, Tb); Te = VADD(T9, Tb); T7 = VSUB(T1, T6); Td = VADD(T1, T6); ST(&(x[0]), VADD(Td, Te), ms, &(x[0])); ST(&(x[WS(rs, 2)]), VSUB(Td, Te), ms, &(x[0])); ST(&(x[WS(rs, 1)]), VFMAI(Tc, T7), ms, &(x[WS(rs, 1)])); ST(&(x[WS(rs, 3)]), VFNMSI(Tc, T7), ms, &(x[WS(rs, 1)])); } } } } VLEAVE(); } static const tw_instr twinstr[] = { VTW(0, 1), VTW(0, 3), {TW_NEXT, VL, 0} }; static const ct_desc desc = { 4, XSIMD_STRING("t3bv_4"), twinstr, &GENUS, {10, 8, 2, 0}, 0, 0, 0 }; void XSIMD(codelet_t3bv_4) (planner *p) { X(kdft_dit_register) (p, t3bv_4, &desc); } #else /* HAVE_FMA */ /* Generated by: ../../../genfft/gen_twiddle_c.native -simd -compact -variables 4 -pipeline-latency 8 -twiddle-log3 -precompute-twiddles -no-generate-bytw -n 4 -name t3bv_4 -include t3b.h -sign 1 */ /* * This function contains 12 FP additions, 8 FP multiplications, * (or, 12 additions, 8 multiplications, 0 fused multiply/add), * 16 stack variables, 0 constants, and 8 memory accesses */ #include "t3b.h" static void t3bv_4(R *ri, R *ii, const R *W, stride rs, INT mb, INT me, INT ms) { { INT m; R *x; x = ii; for (m = mb, W = W + (mb * ((TWVL / VL) * 4)); m < me; m = m + VL, x = x + (VL * ms), W = W + (TWVL * 4), MAKE_VOLATILE_STRIDE(4, rs)) { V T2, T3, T4; T2 = LDW(&(W[0])); T3 = LDW(&(W[TWVL * 2])); T4 = VZMULJ(T2, T3); { V T1, Tb, T6, T9, Ta, T5, T8; T1 = LD(&(x[0]), ms, &(x[0])); Ta = LD(&(x[WS(rs, 3)]), ms, &(x[WS(rs, 1)])); Tb = VZMUL(T3, Ta); T5 = LD(&(x[WS(rs, 2)]), ms, &(x[0])); T6 = VZMUL(T4, T5); T8 = LD(&(x[WS(rs, 1)]), ms, &(x[WS(rs, 1)])); T9 = VZMUL(T2, T8); { V T7, Tc, Td, Te; T7 = VSUB(T1, T6); Tc = VBYI(VSUB(T9, Tb)); ST(&(x[WS(rs, 3)]), VSUB(T7, Tc), ms, &(x[WS(rs, 1)])); ST(&(x[WS(rs, 1)]), VADD(T7, Tc), ms, &(x[WS(rs, 1)])); Td = VADD(T1, T6); Te = VADD(T9, Tb); ST(&(x[WS(rs, 2)]), VSUB(Td, Te), ms, &(x[0])); ST(&(x[0]), VADD(Td, Te), ms, &(x[0])); } } } } VLEAVE(); } static const tw_instr twinstr[] = { VTW(0, 1), VTW(0, 3), {TW_NEXT, VL, 0} }; static const ct_desc desc = { 4, XSIMD_STRING("t3bv_4"), twinstr, &GENUS, {12, 8, 0, 0}, 0, 0, 0 }; void XSIMD(codelet_t3bv_4) (planner *p) { X(kdft_dit_register) (p, t3bv_4, &desc); } #endif /* HAVE_FMA */ fftw-3.3.3/dft/simd/common/t3bv_8.c0000644000175000017500000001627312054410566013675 00000000000000/* * Copyright (c) 2003, 2007-11 Matteo Frigo * Copyright (c) 2003, 2007-11 Massachusetts Institute of Technology * * This program is free software; you can redistribute it and/or modify * it under the terms of the GNU General Public License as published by * the Free Software Foundation; either version 2 of the License, or * (at your option) any later version. * * This program is distributed in the hope that it will be useful, * but WITHOUT ANY WARRANTY; without even the implied warranty of * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the * GNU General Public License for more details. * * You should have received a copy of the GNU General Public License * along with this program; if not, write to the Free Software * Foundation, Inc., 51 Franklin Street, Fifth Floor, Boston, MA 02110-1301 USA * */ /* This file was automatically generated --- DO NOT EDIT */ /* Generated on Sun Nov 25 07:39:18 EST 2012 */ #include "codelet-dft.h" #ifdef HAVE_FMA /* Generated by: ../../../genfft/gen_twiddle_c.native -fma -reorder-insns -schedule-for-pipeline -simd -compact -variables 4 -pipeline-latency 8 -twiddle-log3 -precompute-twiddles -no-generate-bytw -n 8 -name t3bv_8 -include t3b.h -sign 1 */ /* * This function contains 37 FP additions, 32 FP multiplications, * (or, 27 additions, 22 multiplications, 10 fused multiply/add), * 43 stack variables, 1 constants, and 16 memory accesses */ #include "t3b.h" static void t3bv_8(R *ri, R *ii, const R *W, stride rs, INT mb, INT me, INT ms) { DVK(KP707106781, +0.707106781186547524400844362104849039284835938); { INT m; R *x; x = ii; for (m = mb, W = W + (mb * ((TWVL / VL) * 6)); m < me; m = m + VL, x = x + (VL * ms), W = W + (TWVL * 6), MAKE_VOLATILE_STRIDE(8, rs)) { V T2, T3, Tb, T1, T5, Tn, Tq, T8, Td, T4, Ta, Tp, Tg, Ti, T9; T2 = LDW(&(W[0])); T3 = LDW(&(W[TWVL * 2])); Tb = LDW(&(W[TWVL * 4])); T1 = LD(&(x[0]), ms, &(x[0])); T5 = LD(&(x[WS(rs, 4)]), ms, &(x[0])); Tn = LD(&(x[WS(rs, 2)]), ms, &(x[0])); Tq = LD(&(x[WS(rs, 6)]), ms, &(x[0])); T8 = LD(&(x[WS(rs, 1)]), ms, &(x[WS(rs, 1)])); Td = LD(&(x[WS(rs, 5)]), ms, &(x[WS(rs, 1)])); T4 = VZMUL(T2, T3); Ta = VZMULJ(T2, T3); Tp = VZMULJ(T2, Tb); Tg = LD(&(x[WS(rs, 7)]), ms, &(x[WS(rs, 1)])); Ti = LD(&(x[WS(rs, 3)]), ms, &(x[WS(rs, 1)])); T9 = VZMUL(T2, T8); { V T6, To, Tc, Tr, Th, Tj; T6 = VZMUL(T4, T5); To = VZMUL(Ta, Tn); Tc = VZMULJ(Ta, Tb); Tr = VZMUL(Tp, Tq); Th = VZMUL(Tb, Tg); Tj = VZMUL(T3, Ti); { V Tx, T7, Te, Ts, Ty, Tk, TB; Tx = VADD(T1, T6); T7 = VSUB(T1, T6); Te = VZMUL(Tc, Td); Ts = VSUB(To, Tr); Ty = VADD(To, Tr); Tk = VSUB(Th, Tj); TB = VADD(Th, Tj); { V Tf, TA, Tz, TD; Tf = VSUB(T9, Te); TA = VADD(T9, Te); Tz = VSUB(Tx, Ty); TD = VADD(Tx, Ty); { V TC, TE, Tl, Tt; TC = VSUB(TA, TB); TE = VADD(TA, TB); Tl = VADD(Tf, Tk); Tt = VSUB(Tf, Tk); { V Tu, Tw, Tm, Tv; ST(&(x[0]), VADD(TD, TE), ms, &(x[0])); ST(&(x[WS(rs, 4)]), VSUB(TD, TE), ms, &(x[0])); ST(&(x[WS(rs, 2)]), VFMAI(TC, Tz), ms, &(x[0])); ST(&(x[WS(rs, 6)]), VFNMSI(TC, Tz), ms, &(x[0])); Tu = VFNMS(LDK(KP707106781), Tt, Ts); Tw = VFMA(LDK(KP707106781), Tt, Ts); Tm = VFNMS(LDK(KP707106781), Tl, T7); Tv = VFMA(LDK(KP707106781), Tl, T7); ST(&(x[WS(rs, 1)]), VFMAI(Tw, Tv), ms, &(x[WS(rs, 1)])); ST(&(x[WS(rs, 7)]), VFNMSI(Tw, Tv), ms, &(x[WS(rs, 1)])); ST(&(x[WS(rs, 5)]), VFMAI(Tu, Tm), ms, &(x[WS(rs, 1)])); ST(&(x[WS(rs, 3)]), VFNMSI(Tu, Tm), ms, &(x[WS(rs, 1)])); } } } } } } } VLEAVE(); } static const tw_instr twinstr[] = { VTW(0, 1), VTW(0, 3), VTW(0, 7), {TW_NEXT, VL, 0} }; static const ct_desc desc = { 8, XSIMD_STRING("t3bv_8"), twinstr, &GENUS, {27, 22, 10, 0}, 0, 0, 0 }; void XSIMD(codelet_t3bv_8) (planner *p) { X(kdft_dit_register) (p, t3bv_8, &desc); } #else /* HAVE_FMA */ /* Generated by: ../../../genfft/gen_twiddle_c.native -simd -compact -variables 4 -pipeline-latency 8 -twiddle-log3 -precompute-twiddles -no-generate-bytw -n 8 -name t3bv_8 -include t3b.h -sign 1 */ /* * This function contains 37 FP additions, 24 FP multiplications, * (or, 37 additions, 24 multiplications, 0 fused multiply/add), * 31 stack variables, 1 constants, and 16 memory accesses */ #include "t3b.h" static void t3bv_8(R *ri, R *ii, const R *W, stride rs, INT mb, INT me, INT ms) { DVK(KP707106781, +0.707106781186547524400844362104849039284835938); { INT m; R *x; x = ii; for (m = mb, W = W + (mb * ((TWVL / VL) * 6)); m < me; m = m + VL, x = x + (VL * ms), W = W + (TWVL * 6), MAKE_VOLATILE_STRIDE(8, rs)) { V T1, T4, T5, Tp, T6, T7, Tj; T1 = LDW(&(W[0])); T4 = LDW(&(W[TWVL * 2])); T5 = VZMULJ(T1, T4); Tp = VZMUL(T1, T4); T6 = LDW(&(W[TWVL * 4])); T7 = VZMULJ(T5, T6); Tj = VZMULJ(T1, T6); { V Ts, Tx, Tm, Ty, Ta, TA, Tf, TB, To, Tr, Tq; To = LD(&(x[0]), ms, &(x[0])); Tq = LD(&(x[WS(rs, 4)]), ms, &(x[0])); Tr = VZMUL(Tp, Tq); Ts = VSUB(To, Tr); Tx = VADD(To, Tr); { V Ti, Tl, Th, Tk; Th = LD(&(x[WS(rs, 2)]), ms, &(x[0])); Ti = VZMUL(T5, Th); Tk = LD(&(x[WS(rs, 6)]), ms, &(x[0])); Tl = VZMUL(Tj, Tk); Tm = VSUB(Ti, Tl); Ty = VADD(Ti, Tl); } { V T3, T9, T2, T8; T2 = LD(&(x[WS(rs, 1)]), ms, &(x[WS(rs, 1)])); T3 = VZMUL(T1, T2); T8 = LD(&(x[WS(rs, 5)]), ms, &(x[WS(rs, 1)])); T9 = VZMUL(T7, T8); Ta = VSUB(T3, T9); TA = VADD(T3, T9); } { V Tc, Te, Tb, Td; Tb = LD(&(x[WS(rs, 7)]), ms, &(x[WS(rs, 1)])); Tc = VZMUL(T6, Tb); Td = LD(&(x[WS(rs, 3)]), ms, &(x[WS(rs, 1)])); Te = VZMUL(T4, Td); Tf = VSUB(Tc, Te); TB = VADD(Tc, Te); } { V Tz, TC, TD, TE; Tz = VSUB(Tx, Ty); TC = VBYI(VSUB(TA, TB)); ST(&(x[WS(rs, 6)]), VSUB(Tz, TC), ms, &(x[0])); ST(&(x[WS(rs, 2)]), VADD(Tz, TC), ms, &(x[0])); TD = VADD(Tx, Ty); TE = VADD(TA, TB); ST(&(x[WS(rs, 4)]), VSUB(TD, TE), ms, &(x[0])); ST(&(x[0]), VADD(TD, TE), ms, &(x[0])); { V Tn, Tv, Tu, Tw, Tg, Tt; Tg = VMUL(LDK(KP707106781), VSUB(Ta, Tf)); Tn = VBYI(VSUB(Tg, Tm)); Tv = VBYI(VADD(Tm, Tg)); Tt = VMUL(LDK(KP707106781), VADD(Ta, Tf)); Tu = VSUB(Ts, Tt); Tw = VADD(Ts, Tt); ST(&(x[WS(rs, 3)]), VADD(Tn, Tu), ms, &(x[WS(rs, 1)])); ST(&(x[WS(rs, 7)]), VSUB(Tw, Tv), ms, &(x[WS(rs, 1)])); ST(&(x[WS(rs, 5)]), VSUB(Tu, Tn), ms, &(x[WS(rs, 1)])); ST(&(x[WS(rs, 1)]), VADD(Tv, Tw), ms, &(x[WS(rs, 1)])); } } } } } VLEAVE(); } static const tw_instr twinstr[] = { VTW(0, 1), VTW(0, 3), VTW(0, 7), {TW_NEXT, VL, 0} }; static const ct_desc desc = { 8, XSIMD_STRING("t3bv_8"), twinstr, &GENUS, {37, 24, 0, 0}, 0, 0, 0 }; void XSIMD(codelet_t3bv_8) (planner *p) { X(kdft_dit_register) (p, t3bv_8, &desc); } #endif /* HAVE_FMA */ fftw-3.3.3/dft/simd/common/t3bv_16.c0000644000175000017500000003374112054410567013754 00000000000000/* * Copyright (c) 2003, 2007-11 Matteo Frigo * Copyright (c) 2003, 2007-11 Massachusetts Institute of Technology * * This program is free software; you can redistribute it and/or modify * it under the terms of the GNU General Public License as published by * the Free Software Foundation; either version 2 of the License, or * (at your option) any later version. * * This program is distributed in the hope that it will be useful, * but WITHOUT ANY WARRANTY; without even the implied warranty of * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the * GNU General Public License for more details. * * You should have received a copy of the GNU General Public License * along with this program; if not, write to the Free Software * Foundation, Inc., 51 Franklin Street, Fifth Floor, Boston, MA 02110-1301 USA * */ /* This file was automatically generated --- DO NOT EDIT */ /* Generated on Sun Nov 25 07:39:18 EST 2012 */ #include "codelet-dft.h" #ifdef HAVE_FMA /* Generated by: ../../../genfft/gen_twiddle_c.native -fma -reorder-insns -schedule-for-pipeline -simd -compact -variables 4 -pipeline-latency 8 -twiddle-log3 -precompute-twiddles -no-generate-bytw -n 16 -name t3bv_16 -include t3b.h -sign 1 */ /* * This function contains 98 FP additions, 86 FP multiplications, * (or, 64 additions, 52 multiplications, 34 fused multiply/add), * 70 stack variables, 3 constants, and 32 memory accesses */ #include "t3b.h" static void t3bv_16(R *ri, R *ii, const R *W, stride rs, INT mb, INT me, INT ms) { DVK(KP923879532, +0.923879532511286756128183189396788286822416626); DVK(KP414213562, +0.414213562373095048801688724209698078569671875); DVK(KP707106781, +0.707106781186547524400844362104849039284835938); { INT m; R *x; x = ii; for (m = mb, W = W + (mb * ((TWVL / VL) * 8)); m < me; m = m + VL, x = x + (VL * ms), W = W + (TWVL * 8), MAKE_VOLATILE_STRIDE(16, rs)) { V T13, Tg, TY, T14, T1A, T1q, T1f, T1x, T1r, T1i, Tt, T16, TB, T1j, T1k; V TH; { V T2, T8, Tu, T3; T2 = LDW(&(W[0])); T8 = LDW(&(W[TWVL * 2])); Tu = LDW(&(W[TWVL * 6])); T3 = LDW(&(W[TWVL * 4])); { V Ty, T1o, Tf, T1b, T7, Tr, TQ, TX, T1g, Tl, To, Tw, TG, Tz, T1p; V T1e, TC; { V T1, T5, Ta, Td; T1 = LD(&(x[0]), ms, &(x[0])); T5 = LD(&(x[WS(rs, 8)]), ms, &(x[0])); Ta = LD(&(x[WS(rs, 4)]), ms, &(x[0])); Td = LD(&(x[WS(rs, 12)]), ms, &(x[0])); { V TR, TN, TM, TE, Tb, Tp, Tm, Te, T6, TW, TO, TS; { V TL, Tx, T9, TU, Tc, T4, TV; TL = LD(&(x[WS(rs, 2)]), ms, &(x[0])); Tx = VZMULJ(T2, T8); T9 = VZMUL(T2, T8); TR = VZMULJ(T2, Tu); TU = VZMULJ(T8, T3); Tc = VZMUL(T8, T3); T4 = VZMULJ(T2, T3); TN = VZMUL(T2, T3); TV = LD(&(x[WS(rs, 6)]), ms, &(x[0])); TM = VZMUL(Tx, TL); Ty = VZMULJ(Tx, T3); TE = VZMUL(Tx, T3); Tb = VZMUL(T9, Ta); Tp = VZMUL(T9, T3); Tm = VZMULJ(T9, T3); Te = VZMUL(Tc, Td); T6 = VZMUL(T4, T5); TW = VZMUL(TU, TV); } TO = LD(&(x[WS(rs, 10)]), ms, &(x[0])); TS = LD(&(x[WS(rs, 14)]), ms, &(x[0])); { V TP, TT, Ti, Tk, Tn, Th, Tq, Tj; Th = LD(&(x[WS(rs, 1)]), ms, &(x[WS(rs, 1)])); Tq = LD(&(x[WS(rs, 13)]), ms, &(x[WS(rs, 1)])); Tj = LD(&(x[WS(rs, 9)]), ms, &(x[WS(rs, 1)])); T1o = VSUB(Tb, Te); Tf = VADD(Tb, Te); T1b = VSUB(T1, T6); T7 = VADD(T1, T6); TP = VZMUL(TN, TO); TT = VZMUL(TR, TS); Ti = VZMUL(T2, Th); Tr = VZMUL(Tp, Tq); Tk = VZMUL(T3, Tj); Tn = LD(&(x[WS(rs, 5)]), ms, &(x[WS(rs, 1)])); { V T1c, T1d, Tv, TF; Tv = LD(&(x[WS(rs, 15)]), ms, &(x[WS(rs, 1)])); TF = LD(&(x[WS(rs, 11)]), ms, &(x[WS(rs, 1)])); T1c = VSUB(TM, TP); TQ = VADD(TM, TP); T1d = VSUB(TT, TW); TX = VADD(TT, TW); T1g = VSUB(Ti, Tk); Tl = VADD(Ti, Tk); To = VZMUL(Tm, Tn); Tw = VZMUL(Tu, Tv); TG = VZMUL(TE, TF); Tz = LD(&(x[WS(rs, 7)]), ms, &(x[WS(rs, 1)])); T1p = VSUB(T1c, T1d); T1e = VADD(T1c, T1d); TC = LD(&(x[WS(rs, 3)]), ms, &(x[WS(rs, 1)])); } } } } { V T1h, Ts, TA, TD; T13 = VADD(T7, Tf); Tg = VSUB(T7, Tf); T1h = VSUB(To, Tr); Ts = VADD(To, Tr); TY = VSUB(TQ, TX); T14 = VADD(TQ, TX); TA = VZMUL(Ty, Tz); T1A = VFNMS(LDK(KP707106781), T1p, T1o); T1q = VFMA(LDK(KP707106781), T1p, T1o); T1f = VFMA(LDK(KP707106781), T1e, T1b); T1x = VFNMS(LDK(KP707106781), T1e, T1b); TD = VZMUL(T8, TC); T1r = VFMA(LDK(KP414213562), T1g, T1h); T1i = VFNMS(LDK(KP414213562), T1h, T1g); Tt = VSUB(Tl, Ts); T16 = VADD(Tl, Ts); TB = VADD(Tw, TA); T1j = VSUB(Tw, TA); T1k = VSUB(TG, TD); TH = VADD(TD, TG); } } } { V T15, T19, T1l, T1s, TI, T17; T15 = VSUB(T13, T14); T19 = VADD(T13, T14); T1l = VFNMS(LDK(KP414213562), T1k, T1j); T1s = VFMA(LDK(KP414213562), T1j, T1k); TI = VSUB(TB, TH); T17 = VADD(TB, TH); { V T1y, T1t, T1B, T1m; T1y = VADD(T1r, T1s); T1t = VSUB(T1r, T1s); T1B = VSUB(T1i, T1l); T1m = VADD(T1i, T1l); { V T18, T1a, TJ, TZ; T18 = VSUB(T16, T17); T1a = VADD(T16, T17); TJ = VADD(Tt, TI); TZ = VSUB(Tt, TI); { V T1u, T1w, T1z, T1D; T1u = VFNMS(LDK(KP923879532), T1t, T1q); T1w = VFMA(LDK(KP923879532), T1t, T1q); T1z = VFNMS(LDK(KP923879532), T1y, T1x); T1D = VFMA(LDK(KP923879532), T1y, T1x); { V T1n, T1v, T1C, T1E; T1n = VFNMS(LDK(KP923879532), T1m, T1f); T1v = VFMA(LDK(KP923879532), T1m, T1f); T1C = VFMA(LDK(KP923879532), T1B, T1A); T1E = VFNMS(LDK(KP923879532), T1B, T1A); ST(&(x[WS(rs, 8)]), VSUB(T19, T1a), ms, &(x[0])); ST(&(x[0]), VADD(T19, T1a), ms, &(x[0])); ST(&(x[WS(rs, 4)]), VFMAI(T18, T15), ms, &(x[0])); ST(&(x[WS(rs, 12)]), VFNMSI(T18, T15), ms, &(x[0])); { V T10, T12, TK, T11; T10 = VFNMS(LDK(KP707106781), TZ, TY); T12 = VFMA(LDK(KP707106781), TZ, TY); TK = VFNMS(LDK(KP707106781), TJ, Tg); T11 = VFMA(LDK(KP707106781), TJ, Tg); ST(&(x[WS(rs, 15)]), VFNMSI(T1w, T1v), ms, &(x[WS(rs, 1)])); ST(&(x[WS(rs, 1)]), VFMAI(T1w, T1v), ms, &(x[WS(rs, 1)])); ST(&(x[WS(rs, 9)]), VFMAI(T1u, T1n), ms, &(x[WS(rs, 1)])); ST(&(x[WS(rs, 7)]), VFNMSI(T1u, T1n), ms, &(x[WS(rs, 1)])); ST(&(x[WS(rs, 3)]), VFNMSI(T1E, T1D), ms, &(x[WS(rs, 1)])); ST(&(x[WS(rs, 13)]), VFMAI(T1E, T1D), ms, &(x[WS(rs, 1)])); ST(&(x[WS(rs, 11)]), VFNMSI(T1C, T1z), ms, &(x[WS(rs, 1)])); ST(&(x[WS(rs, 5)]), VFMAI(T1C, T1z), ms, &(x[WS(rs, 1)])); ST(&(x[WS(rs, 2)]), VFMAI(T12, T11), ms, &(x[0])); ST(&(x[WS(rs, 14)]), VFNMSI(T12, T11), ms, &(x[0])); ST(&(x[WS(rs, 10)]), VFMAI(T10, TK), ms, &(x[0])); ST(&(x[WS(rs, 6)]), VFNMSI(T10, TK), ms, &(x[0])); } } } } } } } } VLEAVE(); } static const tw_instr twinstr[] = { VTW(0, 1), VTW(0, 3), VTW(0, 9), VTW(0, 15), {TW_NEXT, VL, 0} }; static const ct_desc desc = { 16, XSIMD_STRING("t3bv_16"), twinstr, &GENUS, {64, 52, 34, 0}, 0, 0, 0 }; void XSIMD(codelet_t3bv_16) (planner *p) { X(kdft_dit_register) (p, t3bv_16, &desc); } #else /* HAVE_FMA */ /* Generated by: ../../../genfft/gen_twiddle_c.native -simd -compact -variables 4 -pipeline-latency 8 -twiddle-log3 -precompute-twiddles -no-generate-bytw -n 16 -name t3bv_16 -include t3b.h -sign 1 */ /* * This function contains 98 FP additions, 64 FP multiplications, * (or, 94 additions, 60 multiplications, 4 fused multiply/add), * 51 stack variables, 3 constants, and 32 memory accesses */ #include "t3b.h" static void t3bv_16(R *ri, R *ii, const R *W, stride rs, INT mb, INT me, INT ms) { DVK(KP382683432, +0.382683432365089771728459984030398866761344562); DVK(KP923879532, +0.923879532511286756128183189396788286822416626); DVK(KP707106781, +0.707106781186547524400844362104849039284835938); { INT m; R *x; x = ii; for (m = mb, W = W + (mb * ((TWVL / VL) * 8)); m < me; m = m + VL, x = x + (VL * ms), W = W + (TWVL * 8), MAKE_VOLATILE_STRIDE(16, rs)) { V T1, T8, T9, Tl, Ti, TE, T4, Ta, TO, TV, Td, Tm, TA, TH, Ts; T1 = LDW(&(W[0])); T8 = LDW(&(W[TWVL * 2])); T9 = VZMUL(T1, T8); Tl = VZMULJ(T1, T8); Ti = LDW(&(W[TWVL * 6])); TE = VZMULJ(T1, Ti); T4 = LDW(&(W[TWVL * 4])); Ta = VZMULJ(T9, T4); TO = VZMUL(T8, T4); TV = VZMULJ(T1, T4); Td = VZMUL(T9, T4); Tm = VZMULJ(Tl, T4); TA = VZMUL(T1, T4); TH = VZMULJ(T8, T4); Ts = VZMUL(Tl, T4); { V TY, T1q, TR, T1r, T1m, T1n, TL, TZ, T1f, T1g, T1h, Th, T11, T1i, T1j; V T1k, Tw, T12, TU, TX, TW; TU = LD(&(x[0]), ms, &(x[0])); TW = LD(&(x[WS(rs, 8)]), ms, &(x[0])); TX = VZMUL(TV, TW); TY = VSUB(TU, TX); T1q = VADD(TU, TX); { V TN, TQ, TM, TP; TM = LD(&(x[WS(rs, 4)]), ms, &(x[0])); TN = VZMUL(T9, TM); TP = LD(&(x[WS(rs, 12)]), ms, &(x[0])); TQ = VZMUL(TO, TP); TR = VSUB(TN, TQ); T1r = VADD(TN, TQ); } { V Tz, TJ, TC, TG, TD, TK; { V Ty, TI, TB, TF; Ty = LD(&(x[WS(rs, 2)]), ms, &(x[0])); Tz = VZMUL(Tl, Ty); TI = LD(&(x[WS(rs, 6)]), ms, &(x[0])); TJ = VZMUL(TH, TI); TB = LD(&(x[WS(rs, 10)]), ms, &(x[0])); TC = VZMUL(TA, TB); TF = LD(&(x[WS(rs, 14)]), ms, &(x[0])); TG = VZMUL(TE, TF); } T1m = VADD(Tz, TC); T1n = VADD(TG, TJ); TD = VSUB(Tz, TC); TK = VSUB(TG, TJ); TL = VMUL(LDK(KP707106781), VSUB(TD, TK)); TZ = VMUL(LDK(KP707106781), VADD(TD, TK)); } { V T3, Tf, T6, Tc, T7, Tg; { V T2, Te, T5, Tb; T2 = LD(&(x[WS(rs, 1)]), ms, &(x[WS(rs, 1)])); T3 = VZMUL(T1, T2); Te = LD(&(x[WS(rs, 13)]), ms, &(x[WS(rs, 1)])); Tf = VZMUL(Td, Te); T5 = LD(&(x[WS(rs, 9)]), ms, &(x[WS(rs, 1)])); T6 = VZMUL(T4, T5); Tb = LD(&(x[WS(rs, 5)]), ms, &(x[WS(rs, 1)])); Tc = VZMUL(Ta, Tb); } T1f = VADD(T3, T6); T1g = VADD(Tc, Tf); T1h = VSUB(T1f, T1g); T7 = VSUB(T3, T6); Tg = VSUB(Tc, Tf); Th = VFNMS(LDK(KP382683432), Tg, VMUL(LDK(KP923879532), T7)); T11 = VFMA(LDK(KP382683432), T7, VMUL(LDK(KP923879532), Tg)); } { V Tk, Tu, To, Tr, Tp, Tv; { V Tj, Tt, Tn, Tq; Tj = LD(&(x[WS(rs, 15)]), ms, &(x[WS(rs, 1)])); Tk = VZMUL(Ti, Tj); Tt = LD(&(x[WS(rs, 11)]), ms, &(x[WS(rs, 1)])); Tu = VZMUL(Ts, Tt); Tn = LD(&(x[WS(rs, 7)]), ms, &(x[WS(rs, 1)])); To = VZMUL(Tm, Tn); Tq = LD(&(x[WS(rs, 3)]), ms, &(x[WS(rs, 1)])); Tr = VZMUL(T8, Tq); } T1i = VADD(Tk, To); T1j = VADD(Tr, Tu); T1k = VSUB(T1i, T1j); Tp = VSUB(Tk, To); Tv = VSUB(Tr, Tu); Tw = VFMA(LDK(KP923879532), Tp, VMUL(LDK(KP382683432), Tv)); T12 = VFNMS(LDK(KP382683432), Tp, VMUL(LDK(KP923879532), Tv)); } { V T1p, T1v, T1u, T1w; { V T1l, T1o, T1s, T1t; T1l = VMUL(LDK(KP707106781), VSUB(T1h, T1k)); T1o = VSUB(T1m, T1n); T1p = VBYI(VSUB(T1l, T1o)); T1v = VBYI(VADD(T1o, T1l)); T1s = VSUB(T1q, T1r); T1t = VMUL(LDK(KP707106781), VADD(T1h, T1k)); T1u = VSUB(T1s, T1t); T1w = VADD(T1s, T1t); } ST(&(x[WS(rs, 6)]), VADD(T1p, T1u), ms, &(x[0])); ST(&(x[WS(rs, 14)]), VSUB(T1w, T1v), ms, &(x[0])); ST(&(x[WS(rs, 10)]), VSUB(T1u, T1p), ms, &(x[0])); ST(&(x[WS(rs, 2)]), VADD(T1v, T1w), ms, &(x[0])); } { V T1z, T1D, T1C, T1E; { V T1x, T1y, T1A, T1B; T1x = VADD(T1q, T1r); T1y = VADD(T1m, T1n); T1z = VSUB(T1x, T1y); T1D = VADD(T1x, T1y); T1A = VADD(T1f, T1g); T1B = VADD(T1i, T1j); T1C = VBYI(VSUB(T1A, T1B)); T1E = VADD(T1A, T1B); } ST(&(x[WS(rs, 12)]), VSUB(T1z, T1C), ms, &(x[0])); ST(&(x[0]), VADD(T1D, T1E), ms, &(x[0])); ST(&(x[WS(rs, 4)]), VADD(T1z, T1C), ms, &(x[0])); ST(&(x[WS(rs, 8)]), VSUB(T1D, T1E), ms, &(x[0])); } { V TT, T15, T14, T16; { V Tx, TS, T10, T13; Tx = VSUB(Th, Tw); TS = VSUB(TL, TR); TT = VBYI(VSUB(Tx, TS)); T15 = VBYI(VADD(TS, Tx)); T10 = VSUB(TY, TZ); T13 = VSUB(T11, T12); T14 = VSUB(T10, T13); T16 = VADD(T10, T13); } ST(&(x[WS(rs, 5)]), VADD(TT, T14), ms, &(x[WS(rs, 1)])); ST(&(x[WS(rs, 13)]), VSUB(T16, T15), ms, &(x[WS(rs, 1)])); ST(&(x[WS(rs, 11)]), VSUB(T14, TT), ms, &(x[WS(rs, 1)])); ST(&(x[WS(rs, 3)]), VADD(T15, T16), ms, &(x[WS(rs, 1)])); } { V T19, T1d, T1c, T1e; { V T17, T18, T1a, T1b; T17 = VADD(TY, TZ); T18 = VADD(Th, Tw); T19 = VADD(T17, T18); T1d = VSUB(T17, T18); T1a = VADD(TR, TL); T1b = VADD(T11, T12); T1c = VBYI(VADD(T1a, T1b)); T1e = VBYI(VSUB(T1b, T1a)); } ST(&(x[WS(rs, 15)]), VSUB(T19, T1c), ms, &(x[WS(rs, 1)])); ST(&(x[WS(rs, 7)]), VADD(T1d, T1e), ms, &(x[WS(rs, 1)])); ST(&(x[WS(rs, 1)]), VADD(T19, T1c), ms, &(x[WS(rs, 1)])); ST(&(x[WS(rs, 9)]), VSUB(T1d, T1e), ms, &(x[WS(rs, 1)])); } } } } VLEAVE(); } static const tw_instr twinstr[] = { VTW(0, 1), VTW(0, 3), VTW(0, 9), VTW(0, 15), {TW_NEXT, VL, 0} }; static const ct_desc desc = { 16, XSIMD_STRING("t3bv_16"), twinstr, &GENUS, {94, 60, 4, 0}, 0, 0, 0 }; void XSIMD(codelet_t3bv_16) (planner *p) { X(kdft_dit_register) (p, t3bv_16, &desc); } #endif /* HAVE_FMA */ fftw-3.3.3/dft/simd/common/t3bv_32.c0000644000175000017500000007460312054410573013751 00000000000000/* * Copyright (c) 2003, 2007-11 Matteo Frigo * Copyright (c) 2003, 2007-11 Massachusetts Institute of Technology * * This program is free software; you can redistribute it and/or modify * it under the terms of the GNU General Public License as published by * the Free Software Foundation; either version 2 of the License, or * (at your option) any later version. * * This program is distributed in the hope that it will be useful, * but WITHOUT ANY WARRANTY; without even the implied warranty of * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the * GNU General Public License for more details. * * You should have received a copy of the GNU General Public License * along with this program; if not, write to the Free Software * Foundation, Inc., 51 Franklin Street, Fifth Floor, Boston, MA 02110-1301 USA * */ /* This file was automatically generated --- DO NOT EDIT */ /* Generated on Sun Nov 25 07:39:19 EST 2012 */ #include "codelet-dft.h" #ifdef HAVE_FMA /* Generated by: ../../../genfft/gen_twiddle_c.native -fma -reorder-insns -schedule-for-pipeline -simd -compact -variables 4 -pipeline-latency 8 -twiddle-log3 -precompute-twiddles -no-generate-bytw -n 32 -name t3bv_32 -include t3b.h -sign 1 */ /* * This function contains 244 FP additions, 214 FP multiplications, * (or, 146 additions, 116 multiplications, 98 fused multiply/add), * 120 stack variables, 7 constants, and 64 memory accesses */ #include "t3b.h" static void t3bv_32(R *ri, R *ii, const R *W, stride rs, INT mb, INT me, INT ms) { DVK(KP831469612, +0.831469612302545237078788377617905756738560812); DVK(KP980785280, +0.980785280403230449126182236134239036973933731); DVK(KP668178637, +0.668178637919298919997757686523080761552472251); DVK(KP198912367, +0.198912367379658006911597622644676228597850501); DVK(KP923879532, +0.923879532511286756128183189396788286822416626); DVK(KP707106781, +0.707106781186547524400844362104849039284835938); DVK(KP414213562, +0.414213562373095048801688724209698078569671875); { INT m; R *x; x = ii; for (m = mb, W = W + (mb * ((TWVL / VL) * 8)); m < me; m = m + VL, x = x + (VL * ms), W = W + (TWVL * 8), MAKE_VOLATILE_STRIDE(32, rs)) { V T2B, T2A, T2F, T2N, T2H, T2z, T2P, T2L, T2C, T2M; { V T2, T5, T3, T7; T2 = LDW(&(W[0])); T5 = LDW(&(W[TWVL * 4])); T3 = LDW(&(W[TWVL * 2])); T7 = LDW(&(W[TWVL * 6])); { V T24, Tb, T3x, T2T, T3K, T2W, T25, Tr, T3z, T3j, T28, TX, T3y, T3g, T27; V TG, T37, T3F, T3G, T3a, T2Y, T15, T1p, T2Z, T2w, T1V, T2v, T1N, T32, T1h; V T17, T1a; { V T1, Tz, TT, T4, TC, Tv, T12, T1D, T1w, T18, T1t, T1O, TK, TP, T1c; V T1m, Tf, T6, Te, TL, TQ, T2S, Tp, TU, Ti, Ta, TM, TR, Tm, TJ; V T22, T9, T1Z; T1 = LD(&(x[0]), ms, &(x[0])); T22 = LD(&(x[WS(rs, 24)]), ms, &(x[0])); T9 = LD(&(x[WS(rs, 16)]), ms, &(x[0])); T1Z = LD(&(x[WS(rs, 8)]), ms, &(x[0])); { V Tn, TH, Tk, To, Th, Tg, T8, Tl, T20, T23, TI; { V Td, T1C, Tc, T21; Td = LD(&(x[WS(rs, 4)]), ms, &(x[0])); Tz = VZMUL(T2, T5); T1C = VZMULJ(T2, T5); Tn = VZMUL(T3, T5); TT = VZMULJ(T3, T5); Tc = VZMUL(T2, T3); T4 = VZMULJ(T2, T3); TH = VZMUL(T3, T7); T21 = VZMULJ(T3, T7); Tk = VZMUL(T2, T7); TC = VZMULJ(T2, T7); Tv = VZMULJ(T5, T7); T12 = VZMULJ(Tz, T7); T20 = VZMUL(T1C, T1Z); T1D = VZMULJ(T1C, T7); T1w = VZMULJ(Tn, T7); T18 = VZMULJ(TT, T7); T1t = VZMUL(Tc, T7); T1O = VZMULJ(Tc, T7); TK = VZMUL(Tc, T5); TP = VZMULJ(Tc, T5); T1c = VZMUL(T4, T7); T1m = VZMULJ(T4, T7); Tf = VZMULJ(T4, T5); T6 = VZMUL(T4, T5); T23 = VZMUL(T21, T22); Te = VZMUL(Tc, Td); } TL = VZMULJ(TK, T7); TQ = VZMULJ(TP, T7); To = LD(&(x[WS(rs, 12)]), ms, &(x[0])); Th = LD(&(x[WS(rs, 20)]), ms, &(x[0])); Tg = VZMULJ(Tf, T7); T8 = VZMULJ(T6, T7); T2S = VADD(T20, T23); T24 = VSUB(T20, T23); Tl = LD(&(x[WS(rs, 28)]), ms, &(x[0])); TI = LD(&(x[WS(rs, 30)]), ms, &(x[0])); Tp = VZMUL(Tn, To); TU = LD(&(x[WS(rs, 6)]), ms, &(x[0])); Ti = VZMUL(Tg, Th); Ta = VZMUL(T8, T9); TM = LD(&(x[WS(rs, 14)]), ms, &(x[0])); TR = LD(&(x[WS(rs, 22)]), ms, &(x[0])); Tm = VZMUL(Tk, Tl); TJ = VZMUL(TH, TI); } { V Tu, TE, Tw, TA; { V T3h, TO, T3i, TW; { V TV, T2U, Tj, T2R, TN, TS, T2V, Tq, Tt, TD; Tt = LD(&(x[WS(rs, 2)]), ms, &(x[0])); TV = VZMUL(TT, TU); T2U = VADD(Te, Ti); Tj = VSUB(Te, Ti); T2R = VADD(T1, Ta); Tb = VSUB(T1, Ta); TN = VZMUL(TL, TM); TS = VZMUL(TQ, TR); T2V = VADD(Tm, Tp); Tq = VSUB(Tm, Tp); Tu = VZMUL(T4, Tt); TD = LD(&(x[WS(rs, 26)]), ms, &(x[0])); T3x = VSUB(T2R, T2S); T2T = VADD(T2R, T2S); T3h = VADD(TJ, TN); TO = VSUB(TJ, TN); T3i = VADD(TV, TS); TW = VSUB(TS, TV); T3K = VSUB(T2U, T2V); T2W = VADD(T2U, T2V); T25 = VSUB(Tj, Tq); Tr = VADD(Tj, Tq); TE = VZMUL(TC, TD); } Tw = LD(&(x[WS(rs, 18)]), ms, &(x[0])); T3z = VSUB(T3h, T3i); T3j = VADD(T3h, T3i); T28 = VFMA(LDK(KP414213562), TO, TW); TX = VFNMS(LDK(KP414213562), TW, TO); TA = LD(&(x[WS(rs, 10)]), ms, &(x[0])); } { V T35, T1z, T1T, T36, T39, T1L, T1B, T1F; { V T1v, T1y, Ty, T3e, T1S, T1Q, T1I, T3f, TF, T1K, T1A, T1E; { V T1u, T1x, Tx, T1R; T1u = LD(&(x[WS(rs, 31)]), ms, &(x[WS(rs, 1)])); T1x = LD(&(x[WS(rs, 15)]), ms, &(x[WS(rs, 1)])); Tx = VZMUL(Tv, Tw); T1R = LD(&(x[WS(rs, 7)]), ms, &(x[WS(rs, 1)])); { V T1P, T1H, T1J, TB; T1P = LD(&(x[WS(rs, 23)]), ms, &(x[WS(rs, 1)])); T1H = LD(&(x[WS(rs, 27)]), ms, &(x[WS(rs, 1)])); T1J = LD(&(x[WS(rs, 11)]), ms, &(x[WS(rs, 1)])); TB = VZMUL(Tz, TA); T1v = VZMUL(T1t, T1u); T1y = VZMUL(T1w, T1x); Ty = VSUB(Tu, Tx); T3e = VADD(Tu, Tx); T1S = VZMUL(Tf, T1R); T1Q = VZMUL(T1O, T1P); T1I = VZMUL(T7, T1H); T3f = VADD(TB, TE); TF = VSUB(TB, TE); T1K = VZMUL(T6, T1J); T1A = LD(&(x[WS(rs, 3)]), ms, &(x[WS(rs, 1)])); T1E = LD(&(x[WS(rs, 19)]), ms, &(x[WS(rs, 1)])); } } T35 = VADD(T1v, T1y); T1z = VSUB(T1v, T1y); T1T = VSUB(T1Q, T1S); T36 = VADD(T1S, T1Q); T3y = VSUB(T3e, T3f); T3g = VADD(T3e, T3f); T27 = VFMA(LDK(KP414213562), Ty, TF); TG = VFNMS(LDK(KP414213562), TF, Ty); T39 = VADD(T1I, T1K); T1L = VSUB(T1I, T1K); T1B = VZMUL(T3, T1A); T1F = VZMUL(T1D, T1E); } { V T11, T14, T1o, T1l, T1e, T1U, T1M, T1g, T16, T19; { V T10, T13, T1n, T1k; T10 = LD(&(x[WS(rs, 1)]), ms, &(x[WS(rs, 1)])); T13 = LD(&(x[WS(rs, 17)]), ms, &(x[WS(rs, 1)])); T1n = LD(&(x[WS(rs, 25)]), ms, &(x[WS(rs, 1)])); T1k = LD(&(x[WS(rs, 9)]), ms, &(x[WS(rs, 1)])); { V T1d, T1f, T1G, T38; T1d = LD(&(x[WS(rs, 29)]), ms, &(x[WS(rs, 1)])); T1f = LD(&(x[WS(rs, 13)]), ms, &(x[WS(rs, 1)])); T1G = VSUB(T1B, T1F); T38 = VADD(T1B, T1F); T37 = VADD(T35, T36); T3F = VSUB(T35, T36); T11 = VZMUL(T2, T10); T14 = VZMUL(T12, T13); T1o = VZMUL(T1m, T1n); T1l = VZMUL(T5, T1k); T1e = VZMUL(T1c, T1d); T3G = VSUB(T39, T38); T3a = VADD(T38, T39); T1U = VSUB(T1L, T1G); T1M = VADD(T1G, T1L); T1g = VZMUL(TK, T1f); } T16 = LD(&(x[WS(rs, 5)]), ms, &(x[WS(rs, 1)])); T19 = LD(&(x[WS(rs, 21)]), ms, &(x[WS(rs, 1)])); } T2Y = VADD(T11, T14); T15 = VSUB(T11, T14); T1p = VSUB(T1l, T1o); T2Z = VADD(T1l, T1o); T2w = VFNMS(LDK(KP707106781), T1U, T1T); T1V = VFMA(LDK(KP707106781), T1U, T1T); T2v = VFNMS(LDK(KP707106781), T1M, T1z); T1N = VFMA(LDK(KP707106781), T1M, T1z); T32 = VADD(T1e, T1g); T1h = VSUB(T1e, T1g); T17 = VZMUL(TP, T16); T1a = VZMUL(T18, T19); } } } } { V T2X, T3k, T3b, T3t, T1b, T31, T30, T3C, T3r, T3v, T3p, T3q; T2X = VSUB(T2T, T2W); T3p = VADD(T2T, T2W); T3q = VADD(T3g, T3j); T3k = VSUB(T3g, T3j); T3b = VSUB(T37, T3a); T3t = VADD(T37, T3a); T1b = VSUB(T17, T1a); T31 = VADD(T17, T1a); T30 = VADD(T2Y, T2Z); T3C = VSUB(T2Y, T2Z); T3r = VSUB(T3p, T3q); T3v = VADD(T3p, T3q); { V T1r, T2t, T1j, T2s, T3S, T3Y, T3R, T3V; { V T3B, T3T, T3M, T3W, T3U, T3P, T3X, T3I, T3l, T3c, T3w, T3u; { V T3L, T3A, T33, T3D, T1i, T1q, T3O, T3H; T3L = VSUB(T3y, T3z); T3A = VADD(T3y, T3z); T33 = VADD(T31, T32); T3D = VSUB(T31, T32); T1i = VADD(T1b, T1h); T1q = VSUB(T1b, T1h); T3O = VFMA(LDK(KP414213562), T3F, T3G); T3H = VFNMS(LDK(KP414213562), T3G, T3F); T3B = VFMA(LDK(KP707106781), T3A, T3x); T3T = VFNMS(LDK(KP707106781), T3A, T3x); T3M = VFMA(LDK(KP707106781), T3L, T3K); T3W = VFNMS(LDK(KP707106781), T3L, T3K); { V T3E, T3N, T3s, T34; T3E = VFNMS(LDK(KP414213562), T3D, T3C); T3N = VFMA(LDK(KP414213562), T3C, T3D); T3s = VADD(T30, T33); T34 = VSUB(T30, T33); T1r = VFMA(LDK(KP707106781), T1q, T1p); T2t = VFNMS(LDK(KP707106781), T1q, T1p); T1j = VFMA(LDK(KP707106781), T1i, T15); T2s = VFNMS(LDK(KP707106781), T1i, T15); T3U = VADD(T3N, T3O); T3P = VSUB(T3N, T3O); T3X = VSUB(T3E, T3H); T3I = VADD(T3E, T3H); T3l = VSUB(T34, T3b); T3c = VADD(T34, T3b); T3w = VADD(T3s, T3t); T3u = VSUB(T3s, T3t); } } { V T40, T3Z, T3Q, T3J; T3S = VFMA(LDK(KP923879532), T3P, T3M); T3Q = VFNMS(LDK(KP923879532), T3P, T3M); T40 = VFNMS(LDK(KP923879532), T3X, T3W); T3Y = VFMA(LDK(KP923879532), T3X, T3W); T3R = VFMA(LDK(KP923879532), T3I, T3B); T3J = VFNMS(LDK(KP923879532), T3I, T3B); { V T3o, T3m, T3n, T3d; T3o = VFMA(LDK(KP707106781), T3l, T3k); T3m = VFNMS(LDK(KP707106781), T3l, T3k); T3n = VFMA(LDK(KP707106781), T3c, T2X); T3d = VFNMS(LDK(KP707106781), T3c, T2X); ST(&(x[WS(rs, 16)]), VSUB(T3v, T3w), ms, &(x[0])); ST(&(x[0]), VADD(T3v, T3w), ms, &(x[0])); ST(&(x[WS(rs, 8)]), VFMAI(T3u, T3r), ms, &(x[0])); ST(&(x[WS(rs, 24)]), VFNMSI(T3u, T3r), ms, &(x[0])); T3Z = VFMA(LDK(KP923879532), T3U, T3T); T3V = VFNMS(LDK(KP923879532), T3U, T3T); ST(&(x[WS(rs, 18)]), VFMAI(T3Q, T3J), ms, &(x[0])); ST(&(x[WS(rs, 14)]), VFNMSI(T3Q, T3J), ms, &(x[0])); ST(&(x[WS(rs, 28)]), VFNMSI(T3o, T3n), ms, &(x[0])); ST(&(x[WS(rs, 4)]), VFMAI(T3o, T3n), ms, &(x[0])); ST(&(x[WS(rs, 20)]), VFMAI(T3m, T3d), ms, &(x[0])); ST(&(x[WS(rs, 12)]), VFNMSI(T3m, T3d), ms, &(x[0])); } ST(&(x[WS(rs, 26)]), VFMAI(T40, T3Z), ms, &(x[0])); ST(&(x[WS(rs, 6)]), VFNMSI(T40, T3Z), ms, &(x[0])); } } { V T2p, T1s, T1W, T2h, TZ, T2i, T2d, T26, T29, T2q; { V Ts, TY, T2b, T2c; T2p = VFNMS(LDK(KP707106781), Tr, Tb); Ts = VFMA(LDK(KP707106781), Tr, Tb); TY = VADD(TG, TX); T2B = VSUB(TG, TX); T1s = VFNMS(LDK(KP198912367), T1r, T1j); T2b = VFMA(LDK(KP198912367), T1j, T1r); T2c = VFMA(LDK(KP198912367), T1N, T1V); T1W = VFNMS(LDK(KP198912367), T1V, T1N); ST(&(x[WS(rs, 2)]), VFMAI(T3S, T3R), ms, &(x[0])); ST(&(x[WS(rs, 30)]), VFNMSI(T3S, T3R), ms, &(x[0])); ST(&(x[WS(rs, 22)]), VFNMSI(T3Y, T3V), ms, &(x[0])); ST(&(x[WS(rs, 10)]), VFMAI(T3Y, T3V), ms, &(x[0])); T2h = VFNMS(LDK(KP923879532), TY, Ts); TZ = VFMA(LDK(KP923879532), TY, Ts); T2i = VADD(T2b, T2c); T2d = VSUB(T2b, T2c); T2A = VFNMS(LDK(KP707106781), T25, T24); T26 = VFMA(LDK(KP707106781), T25, T24); T29 = VSUB(T27, T28); T2q = VADD(T27, T28); } { V T2J, T2r, T2K, T2y; { V T2u, T2D, T2j, T2n, T2l, T1X, T2k, T2a, T2E, T2x; T2u = VFMA(LDK(KP668178637), T2t, T2s); T2D = VFNMS(LDK(KP668178637), T2s, T2t); T2j = VFNMS(LDK(KP980785280), T2i, T2h); T2n = VFMA(LDK(KP980785280), T2i, T2h); T2l = VSUB(T1s, T1W); T1X = VADD(T1s, T1W); T2k = VFNMS(LDK(KP923879532), T29, T26); T2a = VFMA(LDK(KP923879532), T29, T26); T2J = VFNMS(LDK(KP923879532), T2q, T2p); T2r = VFMA(LDK(KP923879532), T2q, T2p); T2E = VFNMS(LDK(KP668178637), T2v, T2w); T2x = VFMA(LDK(KP668178637), T2w, T2v); { V T1Y, T2f, T2o, T2m, T2e, T2g; T1Y = VFNMS(LDK(KP980785280), T1X, TZ); T2f = VFMA(LDK(KP980785280), T1X, TZ); T2o = VFNMS(LDK(KP980785280), T2l, T2k); T2m = VFMA(LDK(KP980785280), T2l, T2k); T2e = VFNMS(LDK(KP980785280), T2d, T2a); T2g = VFMA(LDK(KP980785280), T2d, T2a); T2F = VSUB(T2D, T2E); T2K = VADD(T2D, T2E); T2N = VSUB(T2u, T2x); T2y = VADD(T2u, T2x); ST(&(x[WS(rs, 23)]), VFNMSI(T2m, T2j), ms, &(x[WS(rs, 1)])); ST(&(x[WS(rs, 9)]), VFMAI(T2m, T2j), ms, &(x[WS(rs, 1)])); ST(&(x[WS(rs, 25)]), VFMAI(T2o, T2n), ms, &(x[WS(rs, 1)])); ST(&(x[WS(rs, 7)]), VFNMSI(T2o, T2n), ms, &(x[WS(rs, 1)])); ST(&(x[WS(rs, 1)]), VFMAI(T2g, T2f), ms, &(x[WS(rs, 1)])); ST(&(x[WS(rs, 31)]), VFNMSI(T2g, T2f), ms, &(x[WS(rs, 1)])); ST(&(x[WS(rs, 17)]), VFMAI(T2e, T1Y), ms, &(x[WS(rs, 1)])); ST(&(x[WS(rs, 15)]), VFNMSI(T2e, T1Y), ms, &(x[WS(rs, 1)])); } } T2H = VFMA(LDK(KP831469612), T2y, T2r); T2z = VFNMS(LDK(KP831469612), T2y, T2r); T2P = VFNMS(LDK(KP831469612), T2K, T2J); T2L = VFMA(LDK(KP831469612), T2K, T2J); } } } } } } T2C = VFNMS(LDK(KP923879532), T2B, T2A); T2M = VFMA(LDK(KP923879532), T2B, T2A); { V T2Q, T2O, T2G, T2I; T2Q = VFMA(LDK(KP831469612), T2N, T2M); T2O = VFNMS(LDK(KP831469612), T2N, T2M); T2G = VFNMS(LDK(KP831469612), T2F, T2C); T2I = VFMA(LDK(KP831469612), T2F, T2C); ST(&(x[WS(rs, 21)]), VFMAI(T2O, T2L), ms, &(x[WS(rs, 1)])); ST(&(x[WS(rs, 11)]), VFNMSI(T2O, T2L), ms, &(x[WS(rs, 1)])); ST(&(x[WS(rs, 27)]), VFNMSI(T2Q, T2P), ms, &(x[WS(rs, 1)])); ST(&(x[WS(rs, 5)]), VFMAI(T2Q, T2P), ms, &(x[WS(rs, 1)])); ST(&(x[WS(rs, 29)]), VFMAI(T2I, T2H), ms, &(x[WS(rs, 1)])); ST(&(x[WS(rs, 3)]), VFNMSI(T2I, T2H), ms, &(x[WS(rs, 1)])); ST(&(x[WS(rs, 13)]), VFMAI(T2G, T2z), ms, &(x[WS(rs, 1)])); ST(&(x[WS(rs, 19)]), VFNMSI(T2G, T2z), ms, &(x[WS(rs, 1)])); } } } VLEAVE(); } static const tw_instr twinstr[] = { VTW(0, 1), VTW(0, 3), VTW(0, 9), VTW(0, 27), {TW_NEXT, VL, 0} }; static const ct_desc desc = { 32, XSIMD_STRING("t3bv_32"), twinstr, &GENUS, {146, 116, 98, 0}, 0, 0, 0 }; void XSIMD(codelet_t3bv_32) (planner *p) { X(kdft_dit_register) (p, t3bv_32, &desc); } #else /* HAVE_FMA */ /* Generated by: ../../../genfft/gen_twiddle_c.native -simd -compact -variables 4 -pipeline-latency 8 -twiddle-log3 -precompute-twiddles -no-generate-bytw -n 32 -name t3bv_32 -include t3b.h -sign 1 */ /* * This function contains 244 FP additions, 158 FP multiplications, * (or, 228 additions, 142 multiplications, 16 fused multiply/add), * 90 stack variables, 7 constants, and 64 memory accesses */ #include "t3b.h" static void t3bv_32(R *ri, R *ii, const R *W, stride rs, INT mb, INT me, INT ms) { DVK(KP195090322, +0.195090322016128267848284868477022240927691618); DVK(KP980785280, +0.980785280403230449126182236134239036973933731); DVK(KP555570233, +0.555570233019602224742830813948532874374937191); DVK(KP831469612, +0.831469612302545237078788377617905756738560812); DVK(KP382683432, +0.382683432365089771728459984030398866761344562); DVK(KP923879532, +0.923879532511286756128183189396788286822416626); DVK(KP707106781, +0.707106781186547524400844362104849039284835938); { INT m; R *x; x = ii; for (m = mb, W = W + (mb * ((TWVL / VL) * 8)); m < me; m = m + VL, x = x + (VL * ms), W = W + (TWVL * 8), MAKE_VOLATILE_STRIDE(32, rs)) { V T2, T5, T3, T4, Tc, T1v, TH, Tz, Tn, T6, TS, Tf, TK, T7, T8; V Tv, T1I, T25, Tg, Tk, T1N, T1Q, TC, T16, T12, T1w, TL, TP, TT, T1m; V T1f; T2 = LDW(&(W[0])); T5 = LDW(&(W[TWVL * 4])); T3 = LDW(&(W[TWVL * 2])); T4 = VZMULJ(T2, T3); Tc = VZMUL(T2, T3); T1v = VZMULJ(T2, T5); TH = VZMULJ(T3, T5); Tz = VZMUL(T2, T5); Tn = VZMUL(T3, T5); T6 = VZMUL(T4, T5); TS = VZMUL(Tc, T5); Tf = VZMULJ(T4, T5); TK = VZMULJ(Tc, T5); T7 = LDW(&(W[TWVL * 6])); T8 = VZMULJ(T6, T7); Tv = VZMULJ(T5, T7); T1I = VZMULJ(Tc, T7); T25 = VZMULJ(T3, T7); Tg = VZMULJ(Tf, T7); Tk = VZMUL(T2, T7); T1N = VZMUL(Tc, T7); T1Q = VZMULJ(Tn, T7); TC = VZMULJ(T2, T7); T16 = VZMUL(T4, T7); T12 = VZMULJ(TH, T7); T1w = VZMULJ(T1v, T7); TL = VZMULJ(TK, T7); TP = VZMUL(T3, T7); TT = VZMULJ(TS, T7); T1m = VZMULJ(Tz, T7); T1f = VZMULJ(T4, T7); { V Tb, T28, T3k, T3M, Tr, T22, T3f, T3N, TX, T20, T3b, T3J, TG, T1Z, T38; V T3I, T1M, T2v, T33, T3F, T1V, T2w, T30, T3E, T1j, T2s, T2W, T3C, T1r, T2t; V T2T, T3B; { V T1, T27, Ta, T24, T26, T9, T23, T3i, T3j; T1 = LD(&(x[0]), ms, &(x[0])); T26 = LD(&(x[WS(rs, 24)]), ms, &(x[0])); T27 = VZMUL(T25, T26); T9 = LD(&(x[WS(rs, 16)]), ms, &(x[0])); Ta = VZMUL(T8, T9); T23 = LD(&(x[WS(rs, 8)]), ms, &(x[0])); T24 = VZMUL(T1v, T23); Tb = VSUB(T1, Ta); T28 = VSUB(T24, T27); T3i = VADD(T1, Ta); T3j = VADD(T24, T27); T3k = VSUB(T3i, T3j); T3M = VADD(T3i, T3j); } { V Te, Tp, Ti, Tm; { V Td, To, Th, Tl; Td = LD(&(x[WS(rs, 4)]), ms, &(x[0])); Te = VZMUL(Tc, Td); To = LD(&(x[WS(rs, 12)]), ms, &(x[0])); Tp = VZMUL(Tn, To); Th = LD(&(x[WS(rs, 20)]), ms, &(x[0])); Ti = VZMUL(Tg, Th); Tl = LD(&(x[WS(rs, 28)]), ms, &(x[0])); Tm = VZMUL(Tk, Tl); } { V Tj, Tq, T3d, T3e; Tj = VSUB(Te, Ti); Tq = VSUB(Tm, Tp); Tr = VMUL(LDK(KP707106781), VADD(Tj, Tq)); T22 = VMUL(LDK(KP707106781), VSUB(Tj, Tq)); T3d = VADD(Te, Ti); T3e = VADD(Tm, Tp); T3f = VSUB(T3d, T3e); T3N = VADD(T3d, T3e); } } { V TJ, TV, TN, TR; { V TI, TU, TM, TQ; TI = LD(&(x[WS(rs, 6)]), ms, &(x[0])); TJ = VZMUL(TH, TI); TU = LD(&(x[WS(rs, 14)]), ms, &(x[0])); TV = VZMUL(TT, TU); TM = LD(&(x[WS(rs, 22)]), ms, &(x[0])); TN = VZMUL(TL, TM); TQ = LD(&(x[WS(rs, 30)]), ms, &(x[0])); TR = VZMUL(TP, TQ); } { V TO, TW, T39, T3a; TO = VSUB(TJ, TN); TW = VSUB(TR, TV); TX = VFNMS(LDK(KP382683432), TW, VMUL(LDK(KP923879532), TO)); T20 = VFMA(LDK(KP923879532), TW, VMUL(LDK(KP382683432), TO)); T39 = VADD(TR, TV); T3a = VADD(TJ, TN); T3b = VSUB(T39, T3a); T3J = VADD(T39, T3a); } } { V Tu, TE, Tx, TB; { V Tt, TD, Tw, TA; Tt = LD(&(x[WS(rs, 2)]), ms, &(x[0])); Tu = VZMUL(T4, Tt); TD = LD(&(x[WS(rs, 26)]), ms, &(x[0])); TE = VZMUL(TC, TD); Tw = LD(&(x[WS(rs, 18)]), ms, &(x[0])); Tx = VZMUL(Tv, Tw); TA = LD(&(x[WS(rs, 10)]), ms, &(x[0])); TB = VZMUL(Tz, TA); } { V Ty, TF, T36, T37; Ty = VSUB(Tu, Tx); TF = VSUB(TB, TE); TG = VFMA(LDK(KP382683432), Ty, VMUL(LDK(KP923879532), TF)); T1Z = VFNMS(LDK(KP382683432), TF, VMUL(LDK(KP923879532), Ty)); T36 = VADD(Tu, Tx); T37 = VADD(TB, TE); T38 = VSUB(T36, T37); T3I = VADD(T36, T37); } } { V T1H, T1K, T1S, T1P, T1B, T1D, T1E, T1u, T1y, T1z; { V T1G, T1J, T1R, T1O; T1G = LD(&(x[WS(rs, 7)]), ms, &(x[WS(rs, 1)])); T1H = VZMUL(Tf, T1G); T1J = LD(&(x[WS(rs, 23)]), ms, &(x[WS(rs, 1)])); T1K = VZMUL(T1I, T1J); T1R = LD(&(x[WS(rs, 15)]), ms, &(x[WS(rs, 1)])); T1S = VZMUL(T1Q, T1R); T1O = LD(&(x[WS(rs, 31)]), ms, &(x[WS(rs, 1)])); T1P = VZMUL(T1N, T1O); { V T1A, T1C, T1t, T1x; T1A = LD(&(x[WS(rs, 27)]), ms, &(x[WS(rs, 1)])); T1B = VZMUL(T7, T1A); T1C = LD(&(x[WS(rs, 11)]), ms, &(x[WS(rs, 1)])); T1D = VZMUL(T6, T1C); T1E = VSUB(T1B, T1D); T1t = LD(&(x[WS(rs, 3)]), ms, &(x[WS(rs, 1)])); T1u = VZMUL(T3, T1t); T1x = LD(&(x[WS(rs, 19)]), ms, &(x[WS(rs, 1)])); T1y = VZMUL(T1w, T1x); T1z = VSUB(T1u, T1y); } } { V T1F, T1L, T31, T32; T1F = VMUL(LDK(KP707106781), VSUB(T1z, T1E)); T1L = VSUB(T1H, T1K); T1M = VSUB(T1F, T1L); T2v = VADD(T1L, T1F); T31 = VADD(T1u, T1y); T32 = VADD(T1B, T1D); T33 = VSUB(T31, T32); T3F = VADD(T31, T32); } { V T1T, T1U, T2Y, T2Z; T1T = VSUB(T1P, T1S); T1U = VMUL(LDK(KP707106781), VADD(T1z, T1E)); T1V = VSUB(T1T, T1U); T2w = VADD(T1T, T1U); T2Y = VADD(T1P, T1S); T2Z = VADD(T1H, T1K); T30 = VSUB(T2Y, T2Z); T3E = VADD(T2Y, T2Z); } } { V T1e, T1h, T1o, T1l, T18, T1a, T1b, T11, T14, T15; { V T1d, T1g, T1n, T1k; T1d = LD(&(x[WS(rs, 9)]), ms, &(x[WS(rs, 1)])); T1e = VZMUL(T5, T1d); T1g = LD(&(x[WS(rs, 25)]), ms, &(x[WS(rs, 1)])); T1h = VZMUL(T1f, T1g); T1n = LD(&(x[WS(rs, 17)]), ms, &(x[WS(rs, 1)])); T1o = VZMUL(T1m, T1n); T1k = LD(&(x[WS(rs, 1)]), ms, &(x[WS(rs, 1)])); T1l = VZMUL(T2, T1k); { V T17, T19, T10, T13; T17 = LD(&(x[WS(rs, 29)]), ms, &(x[WS(rs, 1)])); T18 = VZMUL(T16, T17); T19 = LD(&(x[WS(rs, 13)]), ms, &(x[WS(rs, 1)])); T1a = VZMUL(TS, T19); T1b = VSUB(T18, T1a); T10 = LD(&(x[WS(rs, 5)]), ms, &(x[WS(rs, 1)])); T11 = VZMUL(TK, T10); T13 = LD(&(x[WS(rs, 21)]), ms, &(x[WS(rs, 1)])); T14 = VZMUL(T12, T13); T15 = VSUB(T11, T14); } } { V T1c, T1i, T2U, T2V; T1c = VMUL(LDK(KP707106781), VSUB(T15, T1b)); T1i = VSUB(T1e, T1h); T1j = VSUB(T1c, T1i); T2s = VADD(T1i, T1c); T2U = VADD(T11, T14); T2V = VADD(T18, T1a); T2W = VSUB(T2U, T2V); T3C = VADD(T2U, T2V); } { V T1p, T1q, T2R, T2S; T1p = VSUB(T1l, T1o); T1q = VMUL(LDK(KP707106781), VADD(T15, T1b)); T1r = VSUB(T1p, T1q); T2t = VADD(T1p, T1q); T2R = VADD(T1l, T1o); T2S = VADD(T1e, T1h); T2T = VSUB(T2R, T2S); T3B = VADD(T2R, T2S); } } { V T3V, T3Z, T3Y, T40; { V T3T, T3U, T3W, T3X; T3T = VADD(T3M, T3N); T3U = VADD(T3I, T3J); T3V = VSUB(T3T, T3U); T3Z = VADD(T3T, T3U); T3W = VADD(T3B, T3C); T3X = VADD(T3E, T3F); T3Y = VBYI(VSUB(T3W, T3X)); T40 = VADD(T3W, T3X); } ST(&(x[WS(rs, 24)]), VSUB(T3V, T3Y), ms, &(x[0])); ST(&(x[0]), VADD(T3Z, T40), ms, &(x[0])); ST(&(x[WS(rs, 8)]), VADD(T3V, T3Y), ms, &(x[0])); ST(&(x[WS(rs, 16)]), VSUB(T3Z, T40), ms, &(x[0])); } { V T3K, T3O, T3H, T3P, T3D, T3G; T3K = VSUB(T3I, T3J); T3O = VSUB(T3M, T3N); T3D = VSUB(T3B, T3C); T3G = VSUB(T3E, T3F); T3H = VMUL(LDK(KP707106781), VSUB(T3D, T3G)); T3P = VMUL(LDK(KP707106781), VADD(T3D, T3G)); { V T3L, T3Q, T3R, T3S; T3L = VBYI(VSUB(T3H, T3K)); T3Q = VSUB(T3O, T3P); ST(&(x[WS(rs, 12)]), VADD(T3L, T3Q), ms, &(x[0])); ST(&(x[WS(rs, 20)]), VSUB(T3Q, T3L), ms, &(x[0])); T3R = VBYI(VADD(T3K, T3H)); T3S = VADD(T3O, T3P); ST(&(x[WS(rs, 4)]), VADD(T3R, T3S), ms, &(x[0])); ST(&(x[WS(rs, 28)]), VSUB(T3S, T3R), ms, &(x[0])); } } { V T3g, T3w, T3m, T3t, T35, T3u, T3p, T3x, T3c, T3l; T3c = VMUL(LDK(KP707106781), VSUB(T38, T3b)); T3g = VSUB(T3c, T3f); T3w = VADD(T3f, T3c); T3l = VMUL(LDK(KP707106781), VADD(T38, T3b)); T3m = VSUB(T3k, T3l); T3t = VADD(T3k, T3l); { V T2X, T34, T3n, T3o; T2X = VFNMS(LDK(KP382683432), T2W, VMUL(LDK(KP923879532), T2T)); T34 = VFMA(LDK(KP923879532), T30, VMUL(LDK(KP382683432), T33)); T35 = VSUB(T2X, T34); T3u = VADD(T2X, T34); T3n = VFMA(LDK(KP382683432), T2T, VMUL(LDK(KP923879532), T2W)); T3o = VFNMS(LDK(KP382683432), T30, VMUL(LDK(KP923879532), T33)); T3p = VSUB(T3n, T3o); T3x = VADD(T3n, T3o); } { V T3h, T3q, T3z, T3A; T3h = VBYI(VSUB(T35, T3g)); T3q = VSUB(T3m, T3p); ST(&(x[WS(rs, 10)]), VADD(T3h, T3q), ms, &(x[0])); ST(&(x[WS(rs, 22)]), VSUB(T3q, T3h), ms, &(x[0])); T3z = VSUB(T3t, T3u); T3A = VBYI(VSUB(T3x, T3w)); ST(&(x[WS(rs, 18)]), VSUB(T3z, T3A), ms, &(x[0])); ST(&(x[WS(rs, 14)]), VADD(T3z, T3A), ms, &(x[0])); } { V T3r, T3s, T3v, T3y; T3r = VBYI(VADD(T3g, T35)); T3s = VADD(T3m, T3p); ST(&(x[WS(rs, 6)]), VADD(T3r, T3s), ms, &(x[0])); ST(&(x[WS(rs, 26)]), VSUB(T3s, T3r), ms, &(x[0])); T3v = VADD(T3t, T3u); T3y = VBYI(VADD(T3w, T3x)); ST(&(x[WS(rs, 30)]), VSUB(T3v, T3y), ms, &(x[0])); ST(&(x[WS(rs, 2)]), VADD(T3v, T3y), ms, &(x[0])); } } { V TZ, T2k, T2d, T2l, T1X, T2h, T2a, T2i; { V Ts, TY, T2b, T2c; Ts = VSUB(Tb, Tr); TY = VSUB(TG, TX); TZ = VSUB(Ts, TY); T2k = VADD(Ts, TY); T2b = VFNMS(LDK(KP555570233), T1j, VMUL(LDK(KP831469612), T1r)); T2c = VFMA(LDK(KP555570233), T1M, VMUL(LDK(KP831469612), T1V)); T2d = VSUB(T2b, T2c); T2l = VADD(T2b, T2c); } { V T1s, T1W, T21, T29; T1s = VFMA(LDK(KP831469612), T1j, VMUL(LDK(KP555570233), T1r)); T1W = VFNMS(LDK(KP555570233), T1V, VMUL(LDK(KP831469612), T1M)); T1X = VSUB(T1s, T1W); T2h = VADD(T1s, T1W); T21 = VSUB(T1Z, T20); T29 = VSUB(T22, T28); T2a = VSUB(T21, T29); T2i = VADD(T29, T21); } { V T1Y, T2e, T2n, T2o; T1Y = VADD(TZ, T1X); T2e = VBYI(VADD(T2a, T2d)); ST(&(x[WS(rs, 27)]), VSUB(T1Y, T2e), ms, &(x[WS(rs, 1)])); ST(&(x[WS(rs, 5)]), VADD(T1Y, T2e), ms, &(x[WS(rs, 1)])); T2n = VBYI(VADD(T2i, T2h)); T2o = VADD(T2k, T2l); ST(&(x[WS(rs, 3)]), VADD(T2n, T2o), ms, &(x[WS(rs, 1)])); ST(&(x[WS(rs, 29)]), VSUB(T2o, T2n), ms, &(x[WS(rs, 1)])); } { V T2f, T2g, T2j, T2m; T2f = VSUB(TZ, T1X); T2g = VBYI(VSUB(T2d, T2a)); ST(&(x[WS(rs, 21)]), VSUB(T2f, T2g), ms, &(x[WS(rs, 1)])); ST(&(x[WS(rs, 11)]), VADD(T2f, T2g), ms, &(x[WS(rs, 1)])); T2j = VBYI(VSUB(T2h, T2i)); T2m = VSUB(T2k, T2l); ST(&(x[WS(rs, 13)]), VADD(T2j, T2m), ms, &(x[WS(rs, 1)])); ST(&(x[WS(rs, 19)]), VSUB(T2m, T2j), ms, &(x[WS(rs, 1)])); } } { V T2r, T2M, T2F, T2N, T2y, T2J, T2C, T2K; { V T2p, T2q, T2D, T2E; T2p = VADD(Tb, Tr); T2q = VADD(T1Z, T20); T2r = VSUB(T2p, T2q); T2M = VADD(T2p, T2q); T2D = VFNMS(LDK(KP195090322), T2s, VMUL(LDK(KP980785280), T2t)); T2E = VFMA(LDK(KP195090322), T2v, VMUL(LDK(KP980785280), T2w)); T2F = VSUB(T2D, T2E); T2N = VADD(T2D, T2E); } { V T2u, T2x, T2A, T2B; T2u = VFMA(LDK(KP980785280), T2s, VMUL(LDK(KP195090322), T2t)); T2x = VFNMS(LDK(KP195090322), T2w, VMUL(LDK(KP980785280), T2v)); T2y = VSUB(T2u, T2x); T2J = VADD(T2u, T2x); T2A = VADD(TG, TX); T2B = VADD(T28, T22); T2C = VSUB(T2A, T2B); T2K = VADD(T2B, T2A); } { V T2z, T2G, T2P, T2Q; T2z = VADD(T2r, T2y); T2G = VBYI(VADD(T2C, T2F)); ST(&(x[WS(rs, 25)]), VSUB(T2z, T2G), ms, &(x[WS(rs, 1)])); ST(&(x[WS(rs, 7)]), VADD(T2z, T2G), ms, &(x[WS(rs, 1)])); T2P = VBYI(VADD(T2K, T2J)); T2Q = VADD(T2M, T2N); ST(&(x[WS(rs, 1)]), VADD(T2P, T2Q), ms, &(x[WS(rs, 1)])); ST(&(x[WS(rs, 31)]), VSUB(T2Q, T2P), ms, &(x[WS(rs, 1)])); } { V T2H, T2I, T2L, T2O; T2H = VSUB(T2r, T2y); T2I = VBYI(VSUB(T2F, T2C)); ST(&(x[WS(rs, 23)]), VSUB(T2H, T2I), ms, &(x[WS(rs, 1)])); ST(&(x[WS(rs, 9)]), VADD(T2H, T2I), ms, &(x[WS(rs, 1)])); T2L = VBYI(VSUB(T2J, T2K)); T2O = VSUB(T2M, T2N); ST(&(x[WS(rs, 15)]), VADD(T2L, T2O), ms, &(x[WS(rs, 1)])); ST(&(x[WS(rs, 17)]), VSUB(T2O, T2L), ms, &(x[WS(rs, 1)])); } } } } } VLEAVE(); } static const tw_instr twinstr[] = { VTW(0, 1), VTW(0, 3), VTW(0, 9), VTW(0, 27), {TW_NEXT, VL, 0} }; static const ct_desc desc = { 32, XSIMD_STRING("t3bv_32"), twinstr, &GENUS, {228, 142, 16, 0}, 0, 0, 0 }; void XSIMD(codelet_t3bv_32) (planner *p) { X(kdft_dit_register) (p, t3bv_32, &desc); } #endif /* HAVE_FMA */ fftw-3.3.3/dft/simd/common/t3bv_5.c0000644000175000017500000001433512054410572013664 00000000000000/* * Copyright (c) 2003, 2007-11 Matteo Frigo * Copyright (c) 2003, 2007-11 Massachusetts Institute of Technology * * This program is free software; you can redistribute it and/or modify * it under the terms of the GNU General Public License as published by * the Free Software Foundation; either version 2 of the License, or * (at your option) any later version. * * This program is distributed in the hope that it will be useful, * but WITHOUT ANY WARRANTY; without even the implied warranty of * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the * GNU General Public License for more details. * * You should have received a copy of the GNU General Public License * along with this program; if not, write to the Free Software * Foundation, Inc., 51 Franklin Street, Fifth Floor, Boston, MA 02110-1301 USA * */ /* This file was automatically generated --- DO NOT EDIT */ /* Generated on Sun Nov 25 07:39:22 EST 2012 */ #include "codelet-dft.h" #ifdef HAVE_FMA /* Generated by: ../../../genfft/gen_twiddle_c.native -fma -reorder-insns -schedule-for-pipeline -simd -compact -variables 4 -pipeline-latency 8 -twiddle-log3 -precompute-twiddles -no-generate-bytw -n 5 -name t3bv_5 -include t3b.h -sign 1 */ /* * This function contains 22 FP additions, 23 FP multiplications, * (or, 13 additions, 14 multiplications, 9 fused multiply/add), * 30 stack variables, 4 constants, and 10 memory accesses */ #include "t3b.h" static void t3bv_5(R *ri, R *ii, const R *W, stride rs, INT mb, INT me, INT ms) { DVK(KP559016994, +0.559016994374947424102293417182819058860154590); DVK(KP250000000, +0.250000000000000000000000000000000000000000000); DVK(KP618033988, +0.618033988749894848204586834365638117720309180); DVK(KP951056516, +0.951056516295153572116439333379382143405698634); { INT m; R *x; x = ii; for (m = mb, W = W + (mb * ((TWVL / VL) * 4)); m < me; m = m + VL, x = x + (VL * ms), W = W + (TWVL * 4), MAKE_VOLATILE_STRIDE(5, rs)) { V T2, T5, T1, T3, Td, T7, Tb; T2 = LDW(&(W[0])); T5 = LDW(&(W[TWVL * 2])); T1 = LD(&(x[0]), ms, &(x[0])); T3 = LD(&(x[WS(rs, 1)]), ms, &(x[WS(rs, 1)])); Td = LD(&(x[WS(rs, 3)]), ms, &(x[WS(rs, 1)])); T7 = LD(&(x[WS(rs, 4)]), ms, &(x[0])); Tb = LD(&(x[WS(rs, 2)]), ms, &(x[0])); { V Ta, T6, T4, Te, Tc, T8; Ta = VZMULJ(T2, T5); T6 = VZMUL(T2, T5); T4 = VZMUL(T2, T3); Te = VZMUL(T5, Td); Tc = VZMUL(Ta, Tb); T8 = VZMUL(T6, T7); { V Tf, Tl, T9, Tk; Tf = VADD(Tc, Te); Tl = VSUB(Tc, Te); T9 = VADD(T4, T8); Tk = VSUB(T4, T8); { V Ti, Tg, To, Tm, Th, Tn, Tj; Ti = VSUB(T9, Tf); Tg = VADD(T9, Tf); To = VMUL(LDK(KP951056516), VFNMS(LDK(KP618033988), Tk, Tl)); Tm = VMUL(LDK(KP951056516), VFMA(LDK(KP618033988), Tl, Tk)); Th = VFNMS(LDK(KP250000000), Tg, T1); ST(&(x[0]), VADD(T1, Tg), ms, &(x[0])); Tn = VFNMS(LDK(KP559016994), Ti, Th); Tj = VFMA(LDK(KP559016994), Ti, Th); ST(&(x[WS(rs, 2)]), VFNMSI(To, Tn), ms, &(x[0])); ST(&(x[WS(rs, 3)]), VFMAI(To, Tn), ms, &(x[WS(rs, 1)])); ST(&(x[WS(rs, 4)]), VFNMSI(Tm, Tj), ms, &(x[0])); ST(&(x[WS(rs, 1)]), VFMAI(Tm, Tj), ms, &(x[WS(rs, 1)])); } } } } } VLEAVE(); } static const tw_instr twinstr[] = { VTW(0, 1), VTW(0, 3), {TW_NEXT, VL, 0} }; static const ct_desc desc = { 5, XSIMD_STRING("t3bv_5"), twinstr, &GENUS, {13, 14, 9, 0}, 0, 0, 0 }; void XSIMD(codelet_t3bv_5) (planner *p) { X(kdft_dit_register) (p, t3bv_5, &desc); } #else /* HAVE_FMA */ /* Generated by: ../../../genfft/gen_twiddle_c.native -simd -compact -variables 4 -pipeline-latency 8 -twiddle-log3 -precompute-twiddles -no-generate-bytw -n 5 -name t3bv_5 -include t3b.h -sign 1 */ /* * This function contains 22 FP additions, 18 FP multiplications, * (or, 19 additions, 15 multiplications, 3 fused multiply/add), * 24 stack variables, 4 constants, and 10 memory accesses */ #include "t3b.h" static void t3bv_5(R *ri, R *ii, const R *W, stride rs, INT mb, INT me, INT ms) { DVK(KP250000000, +0.250000000000000000000000000000000000000000000); DVK(KP559016994, +0.559016994374947424102293417182819058860154590); DVK(KP587785252, +0.587785252292473129168705954639072768597652438); DVK(KP951056516, +0.951056516295153572116439333379382143405698634); { INT m; R *x; x = ii; for (m = mb, W = W + (mb * ((TWVL / VL) * 4)); m < me; m = m + VL, x = x + (VL * ms), W = W + (TWVL * 4), MAKE_VOLATILE_STRIDE(5, rs)) { V T1, T4, T5, T9; T1 = LDW(&(W[0])); T4 = LDW(&(W[TWVL * 2])); T5 = VZMUL(T1, T4); T9 = VZMULJ(T1, T4); { V Tj, T8, Te, Tg, Th, Tk; Tj = LD(&(x[0]), ms, &(x[0])); { V T3, Td, T7, Tb; { V T2, Tc, T6, Ta; T2 = LD(&(x[WS(rs, 1)]), ms, &(x[WS(rs, 1)])); T3 = VZMUL(T1, T2); Tc = LD(&(x[WS(rs, 3)]), ms, &(x[WS(rs, 1)])); Td = VZMUL(T4, Tc); T6 = LD(&(x[WS(rs, 4)]), ms, &(x[0])); T7 = VZMUL(T5, T6); Ta = LD(&(x[WS(rs, 2)]), ms, &(x[0])); Tb = VZMUL(T9, Ta); } T8 = VSUB(T3, T7); Te = VSUB(Tb, Td); Tg = VADD(T3, T7); Th = VADD(Tb, Td); Tk = VADD(Tg, Th); } ST(&(x[0]), VADD(Tj, Tk), ms, &(x[0])); { V Tf, Tn, Tm, To, Ti, Tl; Tf = VBYI(VFMA(LDK(KP951056516), T8, VMUL(LDK(KP587785252), Te))); Tn = VBYI(VFNMS(LDK(KP951056516), Te, VMUL(LDK(KP587785252), T8))); Ti = VMUL(LDK(KP559016994), VSUB(Tg, Th)); Tl = VFNMS(LDK(KP250000000), Tk, Tj); Tm = VADD(Ti, Tl); To = VSUB(Tl, Ti); ST(&(x[WS(rs, 1)]), VADD(Tf, Tm), ms, &(x[WS(rs, 1)])); ST(&(x[WS(rs, 3)]), VSUB(To, Tn), ms, &(x[WS(rs, 1)])); ST(&(x[WS(rs, 4)]), VSUB(Tm, Tf), ms, &(x[0])); ST(&(x[WS(rs, 2)]), VADD(Tn, To), ms, &(x[0])); } } } } VLEAVE(); } static const tw_instr twinstr[] = { VTW(0, 1), VTW(0, 3), {TW_NEXT, VL, 0} }; static const ct_desc desc = { 5, XSIMD_STRING("t3bv_5"), twinstr, &GENUS, {19, 15, 3, 0}, 0, 0, 0 }; void XSIMD(codelet_t3bv_5) (planner *p) { X(kdft_dit_register) (p, t3bv_5, &desc); } #endif /* HAVE_FMA */ fftw-3.3.3/dft/simd/common/t3bv_10.c0000644000175000017500000002306312054410573013737 00000000000000/* * Copyright (c) 2003, 2007-11 Matteo Frigo * Copyright (c) 2003, 2007-11 Massachusetts Institute of Technology * * This program is free software; you can redistribute it and/or modify * it under the terms of the GNU General Public License as published by * the Free Software Foundation; either version 2 of the License, or * (at your option) any later version. * * This program is distributed in the hope that it will be useful, * but WITHOUT ANY WARRANTY; without even the implied warranty of * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the * GNU General Public License for more details. * * You should have received a copy of the GNU General Public License * along with this program; if not, write to the Free Software * Foundation, Inc., 51 Franklin Street, Fifth Floor, Boston, MA 02110-1301 USA * */ /* This file was automatically generated --- DO NOT EDIT */ /* Generated on Sun Nov 25 07:39:22 EST 2012 */ #include "codelet-dft.h" #ifdef HAVE_FMA /* Generated by: ../../../genfft/gen_twiddle_c.native -fma -reorder-insns -schedule-for-pipeline -simd -compact -variables 4 -pipeline-latency 8 -twiddle-log3 -precompute-twiddles -no-generate-bytw -n 10 -name t3bv_10 -include t3b.h -sign 1 */ /* * This function contains 57 FP additions, 52 FP multiplications, * (or, 39 additions, 34 multiplications, 18 fused multiply/add), * 57 stack variables, 4 constants, and 20 memory accesses */ #include "t3b.h" static void t3bv_10(R *ri, R *ii, const R *W, stride rs, INT mb, INT me, INT ms) { DVK(KP559016994, +0.559016994374947424102293417182819058860154590); DVK(KP250000000, +0.250000000000000000000000000000000000000000000); DVK(KP618033988, +0.618033988749894848204586834365638117720309180); DVK(KP951056516, +0.951056516295153572116439333379382143405698634); { INT m; R *x; x = ii; for (m = mb, W = W + (mb * ((TWVL / VL) * 6)); m < me; m = m + VL, x = x + (VL * ms), W = W + (TWVL * 6), MAKE_VOLATILE_STRIDE(10, rs)) { V T1, T7, Th, Tx, Tr, Td, Tp, T6, Tv, Tc, Te, Ti, Tl, T2, T3; V T5; T2 = LDW(&(W[0])); T3 = LDW(&(W[TWVL * 2])); T5 = LDW(&(W[TWVL * 4])); T1 = LD(&(x[0]), ms, &(x[0])); T7 = LD(&(x[WS(rs, 5)]), ms, &(x[WS(rs, 1)])); { V To, Tw, Tq, Tu, Ta, T4, Tt, Tk, Tb; To = LD(&(x[WS(rs, 4)]), ms, &(x[0])); Tw = LD(&(x[WS(rs, 1)]), ms, &(x[WS(rs, 1)])); Tq = LD(&(x[WS(rs, 9)]), ms, &(x[WS(rs, 1)])); Tu = LD(&(x[WS(rs, 6)]), ms, &(x[0])); Ta = VZMULJ(T2, T3); T4 = VZMUL(T2, T3); Th = VZMULJ(T2, T5); Tt = VZMULJ(T3, T5); Tb = LD(&(x[WS(rs, 2)]), ms, &(x[0])); Tx = VZMUL(T2, Tw); Tr = VZMUL(T5, Tq); Tk = LD(&(x[WS(rs, 3)]), ms, &(x[WS(rs, 1)])); Td = VZMULJ(Ta, T5); Tp = VZMUL(T4, To); T6 = VZMULJ(T4, T5); Tv = VZMUL(Tt, Tu); Tc = VZMUL(Ta, Tb); Te = LD(&(x[WS(rs, 7)]), ms, &(x[WS(rs, 1)])); Ti = LD(&(x[WS(rs, 8)]), ms, &(x[0])); Tl = VZMUL(T3, Tk); } { V TN, Ts, T8, Ty, TO, Tf, Tj; TN = VADD(Tp, Tr); Ts = VSUB(Tp, Tr); T8 = VZMUL(T6, T7); Ty = VSUB(Tv, Tx); TO = VADD(Tv, Tx); Tf = VZMUL(Td, Te); Tj = VZMUL(Th, Ti); { V T9, TJ, TP, TU, Tz, TF, Tg, TK, Tm, TL; T9 = VSUB(T1, T8); TJ = VADD(T1, T8); TP = VADD(TN, TO); TU = VSUB(TN, TO); Tz = VADD(Ts, Ty); TF = VSUB(Ts, Ty); Tg = VSUB(Tc, Tf); TK = VADD(Tc, Tf); Tm = VSUB(Tj, Tl); TL = VADD(Tj, Tl); { V TM, TV, Tn, TE; TM = VADD(TK, TL); TV = VSUB(TK, TL); Tn = VADD(Tg, Tm); TE = VSUB(Tg, Tm); { V TW, TY, TS, TQ, TG, TI, TC, TA, TR, TB; TW = VMUL(LDK(KP951056516), VFNMS(LDK(KP618033988), TV, TU)); TY = VMUL(LDK(KP951056516), VFMA(LDK(KP618033988), TU, TV)); TS = VSUB(TM, TP); TQ = VADD(TM, TP); TG = VMUL(LDK(KP951056516), VFMA(LDK(KP618033988), TF, TE)); TI = VMUL(LDK(KP951056516), VFNMS(LDK(KP618033988), TE, TF)); TC = VSUB(Tn, Tz); TA = VADD(Tn, Tz); ST(&(x[0]), VADD(TJ, TQ), ms, &(x[0])); TR = VFNMS(LDK(KP250000000), TQ, TJ); ST(&(x[WS(rs, 5)]), VADD(T9, TA), ms, &(x[WS(rs, 1)])); TB = VFNMS(LDK(KP250000000), TA, T9); { V TX, TT, TH, TD; TX = VFMA(LDK(KP559016994), TS, TR); TT = VFNMS(LDK(KP559016994), TS, TR); TH = VFNMS(LDK(KP559016994), TC, TB); TD = VFMA(LDK(KP559016994), TC, TB); ST(&(x[WS(rs, 8)]), VFMAI(TW, TT), ms, &(x[0])); ST(&(x[WS(rs, 2)]), VFNMSI(TW, TT), ms, &(x[0])); ST(&(x[WS(rs, 6)]), VFMAI(TY, TX), ms, &(x[0])); ST(&(x[WS(rs, 4)]), VFNMSI(TY, TX), ms, &(x[0])); ST(&(x[WS(rs, 9)]), VFNMSI(TG, TD), ms, &(x[WS(rs, 1)])); ST(&(x[WS(rs, 1)]), VFMAI(TG, TD), ms, &(x[WS(rs, 1)])); ST(&(x[WS(rs, 7)]), VFNMSI(TI, TH), ms, &(x[WS(rs, 1)])); ST(&(x[WS(rs, 3)]), VFMAI(TI, TH), ms, &(x[WS(rs, 1)])); } } } } } } } VLEAVE(); } static const tw_instr twinstr[] = { VTW(0, 1), VTW(0, 3), VTW(0, 9), {TW_NEXT, VL, 0} }; static const ct_desc desc = { 10, XSIMD_STRING("t3bv_10"), twinstr, &GENUS, {39, 34, 18, 0}, 0, 0, 0 }; void XSIMD(codelet_t3bv_10) (planner *p) { X(kdft_dit_register) (p, t3bv_10, &desc); } #else /* HAVE_FMA */ /* Generated by: ../../../genfft/gen_twiddle_c.native -simd -compact -variables 4 -pipeline-latency 8 -twiddle-log3 -precompute-twiddles -no-generate-bytw -n 10 -name t3bv_10 -include t3b.h -sign 1 */ /* * This function contains 57 FP additions, 42 FP multiplications, * (or, 51 additions, 36 multiplications, 6 fused multiply/add), * 41 stack variables, 4 constants, and 20 memory accesses */ #include "t3b.h" static void t3bv_10(R *ri, R *ii, const R *W, stride rs, INT mb, INT me, INT ms) { DVK(KP587785252, +0.587785252292473129168705954639072768597652438); DVK(KP951056516, +0.951056516295153572116439333379382143405698634); DVK(KP250000000, +0.250000000000000000000000000000000000000000000); DVK(KP559016994, +0.559016994374947424102293417182819058860154590); { INT m; R *x; x = ii; for (m = mb, W = W + (mb * ((TWVL / VL) * 6)); m < me; m = m + VL, x = x + (VL * ms), W = W + (TWVL * 6), MAKE_VOLATILE_STRIDE(10, rs)) { V T1, T2, T3, Ti, T6, T7, TA, Tb, To; T1 = LDW(&(W[0])); T2 = LDW(&(W[TWVL * 2])); T3 = VZMULJ(T1, T2); Ti = VZMUL(T1, T2); T6 = LDW(&(W[TWVL * 4])); T7 = VZMULJ(T3, T6); TA = VZMULJ(Ti, T6); Tb = VZMULJ(T1, T6); To = VZMULJ(T2, T6); { V TD, TQ, Tn, Tt, Tx, TM, TN, TS, Ta, Tg, Tw, TJ, TK, TR, Tz; V TC, TB; Tz = LD(&(x[0]), ms, &(x[0])); TB = LD(&(x[WS(rs, 5)]), ms, &(x[WS(rs, 1)])); TC = VZMUL(TA, TB); TD = VSUB(Tz, TC); TQ = VADD(Tz, TC); { V Tk, Ts, Tm, Tq; { V Tj, Tr, Tl, Tp; Tj = LD(&(x[WS(rs, 4)]), ms, &(x[0])); Tk = VZMUL(Ti, Tj); Tr = LD(&(x[WS(rs, 1)]), ms, &(x[WS(rs, 1)])); Ts = VZMUL(T1, Tr); Tl = LD(&(x[WS(rs, 9)]), ms, &(x[WS(rs, 1)])); Tm = VZMUL(T6, Tl); Tp = LD(&(x[WS(rs, 6)]), ms, &(x[0])); Tq = VZMUL(To, Tp); } Tn = VSUB(Tk, Tm); Tt = VSUB(Tq, Ts); Tx = VADD(Tn, Tt); TM = VADD(Tk, Tm); TN = VADD(Tq, Ts); TS = VADD(TM, TN); } { V T5, Tf, T9, Td; { V T4, Te, T8, Tc; T4 = LD(&(x[WS(rs, 2)]), ms, &(x[0])); T5 = VZMUL(T3, T4); Te = LD(&(x[WS(rs, 3)]), ms, &(x[WS(rs, 1)])); Tf = VZMUL(T2, Te); T8 = LD(&(x[WS(rs, 7)]), ms, &(x[WS(rs, 1)])); T9 = VZMUL(T7, T8); Tc = LD(&(x[WS(rs, 8)]), ms, &(x[0])); Td = VZMUL(Tb, Tc); } Ta = VSUB(T5, T9); Tg = VSUB(Td, Tf); Tw = VADD(Ta, Tg); TJ = VADD(T5, T9); TK = VADD(Td, Tf); TR = VADD(TJ, TK); } { V Ty, TE, TF, Tv, TI, Th, Tu, TH, TG; Ty = VMUL(LDK(KP559016994), VSUB(Tw, Tx)); TE = VADD(Tw, Tx); TF = VFNMS(LDK(KP250000000), TE, TD); Th = VSUB(Ta, Tg); Tu = VSUB(Tn, Tt); Tv = VBYI(VFMA(LDK(KP951056516), Th, VMUL(LDK(KP587785252), Tu))); TI = VBYI(VFNMS(LDK(KP951056516), Tu, VMUL(LDK(KP587785252), Th))); ST(&(x[WS(rs, 5)]), VADD(TD, TE), ms, &(x[WS(rs, 1)])); TH = VSUB(TF, Ty); ST(&(x[WS(rs, 3)]), VSUB(TH, TI), ms, &(x[WS(rs, 1)])); ST(&(x[WS(rs, 7)]), VADD(TI, TH), ms, &(x[WS(rs, 1)])); TG = VADD(Ty, TF); ST(&(x[WS(rs, 1)]), VADD(Tv, TG), ms, &(x[WS(rs, 1)])); ST(&(x[WS(rs, 9)]), VSUB(TG, Tv), ms, &(x[WS(rs, 1)])); } { V TV, TT, TU, TP, TY, TL, TO, TX, TW; TV = VMUL(LDK(KP559016994), VSUB(TR, TS)); TT = VADD(TR, TS); TU = VFNMS(LDK(KP250000000), TT, TQ); TL = VSUB(TJ, TK); TO = VSUB(TM, TN); TP = VBYI(VFNMS(LDK(KP951056516), TO, VMUL(LDK(KP587785252), TL))); TY = VBYI(VFMA(LDK(KP951056516), TL, VMUL(LDK(KP587785252), TO))); ST(&(x[0]), VADD(TQ, TT), ms, &(x[0])); TX = VADD(TV, TU); ST(&(x[WS(rs, 4)]), VSUB(TX, TY), ms, &(x[0])); ST(&(x[WS(rs, 6)]), VADD(TY, TX), ms, &(x[0])); TW = VSUB(TU, TV); ST(&(x[WS(rs, 2)]), VADD(TP, TW), ms, &(x[0])); ST(&(x[WS(rs, 8)]), VSUB(TW, TP), ms, &(x[0])); } } } } VLEAVE(); } static const tw_instr twinstr[] = { VTW(0, 1), VTW(0, 3), VTW(0, 9), {TW_NEXT, VL, 0} }; static const ct_desc desc = { 10, XSIMD_STRING("t3bv_10"), twinstr, &GENUS, {51, 36, 6, 0}, 0, 0, 0 }; void XSIMD(codelet_t3bv_10) (planner *p) { X(kdft_dit_register) (p, t3bv_10, &desc); } #endif /* HAVE_FMA */ fftw-3.3.3/dft/simd/common/t3bv_20.c0000644000175000017500000004404612054410573013744 00000000000000/* * Copyright (c) 2003, 2007-11 Matteo Frigo * Copyright (c) 2003, 2007-11 Massachusetts Institute of Technology * * This program is free software; you can redistribute it and/or modify * it under the terms of the GNU General Public License as published by * the Free Software Foundation; either version 2 of the License, or * (at your option) any later version. * * This program is distributed in the hope that it will be useful, * but WITHOUT ANY WARRANTY; without even the implied warranty of * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the * GNU General Public License for more details. * * You should have received a copy of the GNU General Public License * along with this program; if not, write to the Free Software * Foundation, Inc., 51 Franklin Street, Fifth Floor, Boston, MA 02110-1301 USA * */ /* This file was automatically generated --- DO NOT EDIT */ /* Generated on Sun Nov 25 07:39:23 EST 2012 */ #include "codelet-dft.h" #ifdef HAVE_FMA /* Generated by: ../../../genfft/gen_twiddle_c.native -fma -reorder-insns -schedule-for-pipeline -simd -compact -variables 4 -pipeline-latency 8 -twiddle-log3 -precompute-twiddles -no-generate-bytw -n 20 -name t3bv_20 -include t3b.h -sign 1 */ /* * This function contains 138 FP additions, 118 FP multiplications, * (or, 92 additions, 72 multiplications, 46 fused multiply/add), * 90 stack variables, 4 constants, and 40 memory accesses */ #include "t3b.h" static void t3bv_20(R *ri, R *ii, const R *W, stride rs, INT mb, INT me, INT ms) { DVK(KP559016994, +0.559016994374947424102293417182819058860154590); DVK(KP951056516, +0.951056516295153572116439333379382143405698634); DVK(KP618033988, +0.618033988749894848204586834365638117720309180); DVK(KP250000000, +0.250000000000000000000000000000000000000000000); { INT m; R *x; x = ii; for (m = mb, W = W + (mb * ((TWVL / VL) * 8)); m < me; m = m + VL, x = x + (VL * ms), W = W + (TWVL * 8), MAKE_VOLATILE_STRIDE(20, rs)) { V T19, T1u, T1p, T1x, T1m, T1w, T1t, TI; { V T2, T8, T3, Td; T2 = LDW(&(W[0])); T8 = LDW(&(W[TWVL * 2])); T3 = LDW(&(W[TWVL * 4])); Td = LDW(&(W[TWVL * 6])); { V T7, T1g, T1F, T23, T1n, Tp, T18, T27, T1P, T1I, TU, T1L, T28, T1S, T1o; V TE, T1l, T1j, T26, T2e; { V T1, T1e, T5, T1b; T1 = LD(&(x[0]), ms, &(x[0])); T1e = LD(&(x[WS(rs, 15)]), ms, &(x[WS(rs, 1)])); T5 = LD(&(x[WS(rs, 10)]), ms, &(x[0])); T1b = LD(&(x[WS(rs, 5)]), ms, &(x[WS(rs, 1)])); { V TA, Tx, TQ, T1O, T10, Th, T1G, T1R, T17, T1J, To, Ts, TR, Tv, TK; V TM, TP, Ty, TB; { V Tq, Tt, T13, T16, Tk, Tn; { V Tl, Ti, T11, T14, TV, Tc, T6, Tb, Tf, TW, TY, T1f; { V T1d, Ta, T9, T4; Ta = LD(&(x[WS(rs, 4)]), ms, &(x[0])); TA = VZMULJ(T2, T8); T9 = VZMUL(T2, T8); Tx = VZMUL(T8, T3); Tl = VZMULJ(T8, T3); T4 = VZMUL(T2, T3); Tq = VZMULJ(T2, T3); Tt = VZMULJ(T2, Td); Ti = VZMULJ(T8, Td); T11 = VZMULJ(TA, Td); T14 = VZMULJ(TA, T3); TQ = VZMUL(TA, T3); T1d = VZMULJ(T9, Td); TV = VZMUL(T9, T3); Tc = VZMULJ(T9, T3); T6 = VZMUL(T4, T5); Tb = VZMUL(T9, Ta); Tf = LD(&(x[WS(rs, 14)]), ms, &(x[0])); TW = LD(&(x[WS(rs, 13)]), ms, &(x[WS(rs, 1)])); TY = LD(&(x[WS(rs, 3)]), ms, &(x[WS(rs, 1)])); T1f = VZMUL(T1d, T1e); } { V T1D, TX, TZ, T15, T1E, Tg, T12, T1c, Te, Tj, Tm; T12 = LD(&(x[WS(rs, 17)]), ms, &(x[WS(rs, 1)])); T1c = VZMUL(Tc, T1b); Te = VZMULJ(Tc, Td); T7 = VSUB(T1, T6); T1D = VADD(T1, T6); TX = VZMUL(TV, TW); TZ = VZMUL(T8, TY); T15 = LD(&(x[WS(rs, 7)]), ms, &(x[WS(rs, 1)])); T13 = VZMUL(T11, T12); T1g = VSUB(T1c, T1f); T1E = VADD(T1c, T1f); Tg = VZMUL(Te, Tf); Tj = LD(&(x[WS(rs, 16)]), ms, &(x[0])); Tm = LD(&(x[WS(rs, 6)]), ms, &(x[0])); T1O = VADD(TX, TZ); T10 = VSUB(TX, TZ); T16 = VZMUL(T14, T15); T1F = VSUB(T1D, T1E); T23 = VADD(T1D, T1E); Th = VSUB(Tb, Tg); T1G = VADD(Tb, Tg); Tk = VZMUL(Ti, Tj); Tn = VZMUL(Tl, Tm); } } { V Tr, Tu, TJ, TL, TO; Tr = LD(&(x[WS(rs, 8)]), ms, &(x[0])); T1R = VADD(T13, T16); T17 = VSUB(T13, T16); Tu = LD(&(x[WS(rs, 18)]), ms, &(x[0])); TJ = LD(&(x[WS(rs, 9)]), ms, &(x[WS(rs, 1)])); TL = LD(&(x[WS(rs, 19)]), ms, &(x[WS(rs, 1)])); TO = LD(&(x[WS(rs, 1)]), ms, &(x[WS(rs, 1)])); T1J = VADD(Tk, Tn); To = VSUB(Tk, Tn); Ts = VZMUL(Tq, Tr); TR = LD(&(x[WS(rs, 11)]), ms, &(x[WS(rs, 1)])); Tv = VZMUL(Tt, Tu); TK = VZMUL(T3, TJ); TM = VZMUL(Td, TL); TP = VZMUL(T2, TO); Ty = LD(&(x[WS(rs, 12)]), ms, &(x[0])); TB = LD(&(x[WS(rs, 2)]), ms, &(x[0])); } } { V T1N, Tw, T1H, TN, Tz, TC, T1i, TT, T1K, TS; T1n = VSUB(Th, To); Tp = VADD(Th, To); TS = VZMUL(TQ, TR); T1N = VADD(Ts, Tv); Tw = VSUB(Ts, Tv); T1H = VADD(TK, TM); TN = VSUB(TK, TM); Tz = VZMUL(Tx, Ty); TC = VZMUL(TA, TB); T18 = VSUB(T10, T17); T1i = VADD(T10, T17); TT = VSUB(TP, TS); T1K = VADD(TP, TS); T27 = VADD(T1N, T1O); T1P = VSUB(T1N, T1O); { V TD, T1Q, T24, T1h, T25; TD = VSUB(Tz, TC); T1Q = VADD(Tz, TC); T1I = VSUB(T1G, T1H); T24 = VADD(T1G, T1H); T1h = VADD(TN, TT); TU = VSUB(TN, TT); T25 = VADD(T1J, T1K); T1L = VSUB(T1J, T1K); T28 = VADD(T1Q, T1R); T1S = VSUB(T1Q, T1R); T1o = VSUB(Tw, TD); TE = VADD(Tw, TD); T1l = VSUB(T1h, T1i); T1j = VADD(T1h, T1i); T26 = VADD(T24, T25); T2e = VSUB(T24, T25); } } } } { V T1M, T1Z, T1Y, T1T, T29, T2f, TH, TF, T1k, T1C; T1M = VADD(T1I, T1L); T1Z = VSUB(T1I, T1L); T1Y = VSUB(T1P, T1S); T1T = VADD(T1P, T1S); T29 = VADD(T27, T28); T2f = VSUB(T27, T28); TH = VSUB(Tp, TE); TF = VADD(Tp, TE); T1k = VFNMS(LDK(KP250000000), T1j, T1g); T1C = VADD(T1g, T1j); { V T1W, T2c, TG, T2i, T2g, T22, T20, T1V, T2b, T1U, T2a, T1B; T19 = VFMA(LDK(KP618033988), T18, TU); T1u = VFNMS(LDK(KP618033988), TU, T18); T1W = VSUB(T1M, T1T); T1U = VADD(T1M, T1T); T2c = VSUB(T26, T29); T2a = VADD(T26, T29); TG = VFNMS(LDK(KP250000000), TF, T7); T1B = VADD(T7, TF); T2i = VMUL(LDK(KP951056516), VFNMS(LDK(KP618033988), T2e, T2f)); T2g = VMUL(LDK(KP951056516), VFMA(LDK(KP618033988), T2f, T2e)); T22 = VMUL(LDK(KP951056516), VFMA(LDK(KP618033988), T1Y, T1Z)); T20 = VMUL(LDK(KP951056516), VFNMS(LDK(KP618033988), T1Z, T1Y)); ST(&(x[WS(rs, 10)]), VADD(T1F, T1U), ms, &(x[0])); T1V = VFNMS(LDK(KP250000000), T1U, T1F); ST(&(x[0]), VADD(T23, T2a), ms, &(x[0])); T2b = VFNMS(LDK(KP250000000), T2a, T23); ST(&(x[WS(rs, 5)]), VFMAI(T1C, T1B), ms, &(x[WS(rs, 1)])); ST(&(x[WS(rs, 15)]), VFNMSI(T1C, T1B), ms, &(x[WS(rs, 1)])); T1p = VFMA(LDK(KP618033988), T1o, T1n); T1x = VFNMS(LDK(KP618033988), T1n, T1o); { V T21, T1X, T2h, T2d; T21 = VFMA(LDK(KP559016994), T1W, T1V); T1X = VFNMS(LDK(KP559016994), T1W, T1V); T2h = VFNMS(LDK(KP559016994), T2c, T2b); T2d = VFMA(LDK(KP559016994), T2c, T2b); ST(&(x[WS(rs, 18)]), VFMAI(T20, T1X), ms, &(x[0])); ST(&(x[WS(rs, 2)]), VFNMSI(T20, T1X), ms, &(x[0])); ST(&(x[WS(rs, 14)]), VFNMSI(T22, T21), ms, &(x[0])); ST(&(x[WS(rs, 6)]), VFMAI(T22, T21), ms, &(x[0])); ST(&(x[WS(rs, 16)]), VFMAI(T2g, T2d), ms, &(x[0])); ST(&(x[WS(rs, 4)]), VFNMSI(T2g, T2d), ms, &(x[0])); ST(&(x[WS(rs, 12)]), VFNMSI(T2i, T2h), ms, &(x[0])); ST(&(x[WS(rs, 8)]), VFMAI(T2i, T2h), ms, &(x[0])); T1m = VFMA(LDK(KP559016994), T1l, T1k); T1w = VFNMS(LDK(KP559016994), T1l, T1k); T1t = VFNMS(LDK(KP559016994), TH, TG); TI = VFMA(LDK(KP559016994), TH, TG); } } } } } { V T1A, T1y, T1q, T1s, T1a, T1r, T1z, T1v; T1A = VFMA(LDK(KP951056516), T1x, T1w); T1y = VFNMS(LDK(KP951056516), T1x, T1w); T1q = VFMA(LDK(KP951056516), T1p, T1m); T1s = VFNMS(LDK(KP951056516), T1p, T1m); T1a = VFNMS(LDK(KP951056516), T19, TI); T1r = VFMA(LDK(KP951056516), T19, TI); T1z = VFNMS(LDK(KP951056516), T1u, T1t); T1v = VFMA(LDK(KP951056516), T1u, T1t); ST(&(x[WS(rs, 9)]), VFMAI(T1s, T1r), ms, &(x[WS(rs, 1)])); ST(&(x[WS(rs, 11)]), VFNMSI(T1s, T1r), ms, &(x[WS(rs, 1)])); ST(&(x[WS(rs, 1)]), VFMAI(T1q, T1a), ms, &(x[WS(rs, 1)])); ST(&(x[WS(rs, 19)]), VFNMSI(T1q, T1a), ms, &(x[WS(rs, 1)])); ST(&(x[WS(rs, 17)]), VFMAI(T1y, T1v), ms, &(x[WS(rs, 1)])); ST(&(x[WS(rs, 3)]), VFNMSI(T1y, T1v), ms, &(x[WS(rs, 1)])); ST(&(x[WS(rs, 13)]), VFMAI(T1A, T1z), ms, &(x[WS(rs, 1)])); ST(&(x[WS(rs, 7)]), VFNMSI(T1A, T1z), ms, &(x[WS(rs, 1)])); } } } VLEAVE(); } static const tw_instr twinstr[] = { VTW(0, 1), VTW(0, 3), VTW(0, 9), VTW(0, 19), {TW_NEXT, VL, 0} }; static const ct_desc desc = { 20, XSIMD_STRING("t3bv_20"), twinstr, &GENUS, {92, 72, 46, 0}, 0, 0, 0 }; void XSIMD(codelet_t3bv_20) (planner *p) { X(kdft_dit_register) (p, t3bv_20, &desc); } #else /* HAVE_FMA */ /* Generated by: ../../../genfft/gen_twiddle_c.native -simd -compact -variables 4 -pipeline-latency 8 -twiddle-log3 -precompute-twiddles -no-generate-bytw -n 20 -name t3bv_20 -include t3b.h -sign 1 */ /* * This function contains 138 FP additions, 92 FP multiplications, * (or, 126 additions, 80 multiplications, 12 fused multiply/add), * 73 stack variables, 4 constants, and 40 memory accesses */ #include "t3b.h" static void t3bv_20(R *ri, R *ii, const R *W, stride rs, INT mb, INT me, INT ms) { DVK(KP587785252, +0.587785252292473129168705954639072768597652438); DVK(KP951056516, +0.951056516295153572116439333379382143405698634); DVK(KP250000000, +0.250000000000000000000000000000000000000000000); DVK(KP559016994, +0.559016994374947424102293417182819058860154590); { INT m; R *x; x = ii; for (m = mb, W = W + (mb * ((TWVL / VL) * 8)); m < me; m = m + VL, x = x + (VL * ms), W = W + (TWVL * 8), MAKE_VOLATILE_STRIDE(20, rs)) { V T2, T8, T9, TA, T3, Tc, T4, TV, T14, Tl, Tq, Tx, TQ, Td, Te; V T1g, Ti, Tt, T11; T2 = LDW(&(W[0])); T8 = LDW(&(W[TWVL * 2])); T9 = VZMUL(T2, T8); TA = VZMULJ(T2, T8); T3 = LDW(&(W[TWVL * 4])); Tc = VZMULJ(T9, T3); T4 = VZMUL(T2, T3); TV = VZMUL(T9, T3); T14 = VZMULJ(TA, T3); Tl = VZMULJ(T8, T3); Tq = VZMULJ(T2, T3); Tx = VZMUL(T8, T3); TQ = VZMUL(TA, T3); Td = LDW(&(W[TWVL * 6])); Te = VZMULJ(Tc, Td); T1g = VZMULJ(T9, Td); Ti = VZMULJ(T8, Td); Tt = VZMULJ(T2, Td); T11 = VZMULJ(TA, Td); { V T7, T1j, T1U, T2a, TU, T1n, T1o, T18, Tp, TE, TF, T26, T27, T28, T1M; V T1P, T1W, T1b, T1c, T1k, T23, T24, T25, T1F, T1I, T1V, T1B, T1C; { V T1, T1i, T6, T1f, T1h, T5, T1e, T1S, T1T; T1 = LD(&(x[0]), ms, &(x[0])); T1h = LD(&(x[WS(rs, 15)]), ms, &(x[WS(rs, 1)])); T1i = VZMUL(T1g, T1h); T5 = LD(&(x[WS(rs, 10)]), ms, &(x[0])); T6 = VZMUL(T4, T5); T1e = LD(&(x[WS(rs, 5)]), ms, &(x[WS(rs, 1)])); T1f = VZMUL(Tc, T1e); T7 = VSUB(T1, T6); T1j = VSUB(T1f, T1i); T1S = VADD(T1, T6); T1T = VADD(T1f, T1i); T1U = VSUB(T1S, T1T); T2a = VADD(T1S, T1T); } { V Th, T1D, T10, T1L, T17, T1O, To, T1G, Tw, T1K, TN, T1E, TT, T1H, TD; V T1N; { V Tb, Tg, Ta, Tf; Ta = LD(&(x[WS(rs, 4)]), ms, &(x[0])); Tb = VZMUL(T9, Ta); Tf = LD(&(x[WS(rs, 14)]), ms, &(x[0])); Tg = VZMUL(Te, Tf); Th = VSUB(Tb, Tg); T1D = VADD(Tb, Tg); } { V TX, TZ, TW, TY; TW = LD(&(x[WS(rs, 13)]), ms, &(x[WS(rs, 1)])); TX = VZMUL(TV, TW); TY = LD(&(x[WS(rs, 3)]), ms, &(x[WS(rs, 1)])); TZ = VZMUL(T8, TY); T10 = VSUB(TX, TZ); T1L = VADD(TX, TZ); } { V T13, T16, T12, T15; T12 = LD(&(x[WS(rs, 17)]), ms, &(x[WS(rs, 1)])); T13 = VZMUL(T11, T12); T15 = LD(&(x[WS(rs, 7)]), ms, &(x[WS(rs, 1)])); T16 = VZMUL(T14, T15); T17 = VSUB(T13, T16); T1O = VADD(T13, T16); } { V Tk, Tn, Tj, Tm; Tj = LD(&(x[WS(rs, 16)]), ms, &(x[0])); Tk = VZMUL(Ti, Tj); Tm = LD(&(x[WS(rs, 6)]), ms, &(x[0])); Tn = VZMUL(Tl, Tm); To = VSUB(Tk, Tn); T1G = VADD(Tk, Tn); } { V Ts, Tv, Tr, Tu; Tr = LD(&(x[WS(rs, 8)]), ms, &(x[0])); Ts = VZMUL(Tq, Tr); Tu = LD(&(x[WS(rs, 18)]), ms, &(x[0])); Tv = VZMUL(Tt, Tu); Tw = VSUB(Ts, Tv); T1K = VADD(Ts, Tv); } { V TK, TM, TJ, TL; TJ = LD(&(x[WS(rs, 9)]), ms, &(x[WS(rs, 1)])); TK = VZMUL(T3, TJ); TL = LD(&(x[WS(rs, 19)]), ms, &(x[WS(rs, 1)])); TM = VZMUL(Td, TL); TN = VSUB(TK, TM); T1E = VADD(TK, TM); } { V TP, TS, TO, TR; TO = LD(&(x[WS(rs, 1)]), ms, &(x[WS(rs, 1)])); TP = VZMUL(T2, TO); TR = LD(&(x[WS(rs, 11)]), ms, &(x[WS(rs, 1)])); TS = VZMUL(TQ, TR); TT = VSUB(TP, TS); T1H = VADD(TP, TS); } { V Tz, TC, Ty, TB; Ty = LD(&(x[WS(rs, 12)]), ms, &(x[0])); Tz = VZMUL(Tx, Ty); TB = LD(&(x[WS(rs, 2)]), ms, &(x[0])); TC = VZMUL(TA, TB); TD = VSUB(Tz, TC); T1N = VADD(Tz, TC); } TU = VSUB(TN, TT); T1n = VSUB(Th, To); T1o = VSUB(Tw, TD); T18 = VSUB(T10, T17); Tp = VADD(Th, To); TE = VADD(Tw, TD); TF = VADD(Tp, TE); T26 = VADD(T1K, T1L); T27 = VADD(T1N, T1O); T28 = VADD(T26, T27); T1M = VSUB(T1K, T1L); T1P = VSUB(T1N, T1O); T1W = VADD(T1M, T1P); T1b = VADD(TN, TT); T1c = VADD(T10, T17); T1k = VADD(T1b, T1c); T23 = VADD(T1D, T1E); T24 = VADD(T1G, T1H); T25 = VADD(T23, T24); T1F = VSUB(T1D, T1E); T1I = VSUB(T1G, T1H); T1V = VADD(T1F, T1I); } T1B = VADD(T7, TF); T1C = VBYI(VADD(T1j, T1k)); ST(&(x[WS(rs, 15)]), VSUB(T1B, T1C), ms, &(x[WS(rs, 1)])); ST(&(x[WS(rs, 5)]), VADD(T1B, T1C), ms, &(x[WS(rs, 1)])); { V T29, T2b, T2c, T2g, T2i, T2e, T2f, T2h, T2d; T29 = VMUL(LDK(KP559016994), VSUB(T25, T28)); T2b = VADD(T25, T28); T2c = VFNMS(LDK(KP250000000), T2b, T2a); T2e = VSUB(T23, T24); T2f = VSUB(T26, T27); T2g = VBYI(VFMA(LDK(KP951056516), T2e, VMUL(LDK(KP587785252), T2f))); T2i = VBYI(VFNMS(LDK(KP951056516), T2f, VMUL(LDK(KP587785252), T2e))); ST(&(x[0]), VADD(T2a, T2b), ms, &(x[0])); T2h = VSUB(T2c, T29); ST(&(x[WS(rs, 8)]), VSUB(T2h, T2i), ms, &(x[0])); ST(&(x[WS(rs, 12)]), VADD(T2i, T2h), ms, &(x[0])); T2d = VADD(T29, T2c); ST(&(x[WS(rs, 4)]), VSUB(T2d, T2g), ms, &(x[0])); ST(&(x[WS(rs, 16)]), VADD(T2g, T2d), ms, &(x[0])); } { V T1Z, T1X, T1Y, T1R, T21, T1J, T1Q, T22, T20; T1Z = VMUL(LDK(KP559016994), VSUB(T1V, T1W)); T1X = VADD(T1V, T1W); T1Y = VFNMS(LDK(KP250000000), T1X, T1U); T1J = VSUB(T1F, T1I); T1Q = VSUB(T1M, T1P); T1R = VBYI(VFNMS(LDK(KP951056516), T1Q, VMUL(LDK(KP587785252), T1J))); T21 = VBYI(VFMA(LDK(KP951056516), T1J, VMUL(LDK(KP587785252), T1Q))); ST(&(x[WS(rs, 10)]), VADD(T1U, T1X), ms, &(x[0])); T22 = VADD(T1Z, T1Y); ST(&(x[WS(rs, 6)]), VADD(T21, T22), ms, &(x[0])); ST(&(x[WS(rs, 14)]), VSUB(T22, T21), ms, &(x[0])); T20 = VSUB(T1Y, T1Z); ST(&(x[WS(rs, 2)]), VADD(T1R, T20), ms, &(x[0])); ST(&(x[WS(rs, 18)]), VSUB(T20, T1R), ms, &(x[0])); } { V T19, T1p, T1w, T1u, T1m, T1x, TI, T1t; T19 = VFNMS(LDK(KP951056516), T18, VMUL(LDK(KP587785252), TU)); T1p = VFNMS(LDK(KP951056516), T1o, VMUL(LDK(KP587785252), T1n)); T1w = VFMA(LDK(KP951056516), T1n, VMUL(LDK(KP587785252), T1o)); T1u = VFMA(LDK(KP951056516), TU, VMUL(LDK(KP587785252), T18)); { V T1d, T1l, TG, TH; T1d = VMUL(LDK(KP559016994), VSUB(T1b, T1c)); T1l = VFNMS(LDK(KP250000000), T1k, T1j); T1m = VSUB(T1d, T1l); T1x = VADD(T1d, T1l); TG = VFNMS(LDK(KP250000000), TF, T7); TH = VMUL(LDK(KP559016994), VSUB(Tp, TE)); TI = VSUB(TG, TH); T1t = VADD(TH, TG); } { V T1a, T1q, T1z, T1A; T1a = VSUB(TI, T19); T1q = VBYI(VSUB(T1m, T1p)); ST(&(x[WS(rs, 17)]), VSUB(T1a, T1q), ms, &(x[WS(rs, 1)])); ST(&(x[WS(rs, 3)]), VADD(T1a, T1q), ms, &(x[WS(rs, 1)])); T1z = VADD(T1t, T1u); T1A = VBYI(VSUB(T1x, T1w)); ST(&(x[WS(rs, 11)]), VSUB(T1z, T1A), ms, &(x[WS(rs, 1)])); ST(&(x[WS(rs, 9)]), VADD(T1z, T1A), ms, &(x[WS(rs, 1)])); } { V T1r, T1s, T1v, T1y; T1r = VADD(TI, T19); T1s = VBYI(VADD(T1p, T1m)); ST(&(x[WS(rs, 13)]), VSUB(T1r, T1s), ms, &(x[WS(rs, 1)])); ST(&(x[WS(rs, 7)]), VADD(T1r, T1s), ms, &(x[WS(rs, 1)])); T1v = VSUB(T1t, T1u); T1y = VBYI(VADD(T1w, T1x)); ST(&(x[WS(rs, 19)]), VSUB(T1v, T1y), ms, &(x[WS(rs, 1)])); ST(&(x[WS(rs, 1)]), VADD(T1v, T1y), ms, &(x[WS(rs, 1)])); } } } } } VLEAVE(); } static const tw_instr twinstr[] = { VTW(0, 1), VTW(0, 3), VTW(0, 9), VTW(0, 19), {TW_NEXT, VL, 0} }; static const ct_desc desc = { 20, XSIMD_STRING("t3bv_20"), twinstr, &GENUS, {126, 80, 12, 0}, 0, 0, 0 }; void XSIMD(codelet_t3bv_20) (planner *p) { X(kdft_dit_register) (p, t3bv_20, &desc); } #endif /* HAVE_FMA */ fftw-3.3.3/dft/simd/common/t3bv_25.c0000644000175000017500000012026312054410602013736 00000000000000/* * Copyright (c) 2003, 2007-11 Matteo Frigo * Copyright (c) 2003, 2007-11 Massachusetts Institute of Technology * * This program is free software; you can redistribute it and/or modify * it under the terms of the GNU General Public License as published by * the Free Software Foundation; either version 2 of the License, or * (at your option) any later version. * * This program is distributed in the hope that it will be useful, * but WITHOUT ANY WARRANTY; without even the implied warranty of * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the * GNU General Public License for more details. * * You should have received a copy of the GNU General Public License * along with this program; if not, write to the Free Software * Foundation, Inc., 51 Franklin Street, Fifth Floor, Boston, MA 02110-1301 USA * */ /* This file was automatically generated --- DO NOT EDIT */ /* Generated on Sun Nov 25 07:39:23 EST 2012 */ #include "codelet-dft.h" #ifdef HAVE_FMA /* Generated by: ../../../genfft/gen_twiddle_c.native -fma -reorder-insns -schedule-for-pipeline -simd -compact -variables 4 -pipeline-latency 8 -twiddle-log3 -precompute-twiddles -no-generate-bytw -n 25 -name t3bv_25 -include t3b.h -sign 1 */ /* * This function contains 268 FP additions, 281 FP multiplications, * (or, 87 additions, 100 multiplications, 181 fused multiply/add), * 223 stack variables, 67 constants, and 50 memory accesses */ #include "t3b.h" static void t3bv_25(R *ri, R *ii, const R *W, stride rs, INT mb, INT me, INT ms) { DVK(KP792626838, +0.792626838241819413632131824093538848057784557); DVK(KP876091699, +0.876091699473550838204498029706869638173524346); DVK(KP617882369, +0.617882369114440893914546919006756321695042882); DVK(KP803003575, +0.803003575438660414833440593570376004635464850); DVK(KP242145790, +0.242145790282157779872542093866183953459003101); DVK(KP968583161, +0.968583161128631119490168375464735813836012403); DVK(KP999544308, +0.999544308746292983948881682379742149196758193); DVK(KP916574801, +0.916574801383451584742370439148878693530976769); DVK(KP904730450, +0.904730450839922351881287709692877908104763647); DVK(KP809385824, +0.809385824416008241660603814668679683846476688); DVK(KP447417479, +0.447417479732227551498980015410057305749330693); DVK(KP894834959, +0.894834959464455102997960030820114611498661386); DVK(KP867381224, +0.867381224396525206773171885031575671309956167); DVK(KP683113946, +0.683113946453479238701949862233725244439656928); DVK(KP559154169, +0.559154169276087864842202529084232643714075927); DVK(KP958953096, +0.958953096729998668045963838399037225970891871); DVK(KP831864738, +0.831864738706457140726048799369896829771167132); DVK(KP829049696, +0.829049696159252993975487806364305442437946767); DVK(KP860541664, +0.860541664367944677098261680920518816412804187); DVK(KP897376177, +0.897376177523557693138608077137219684419427330); DVK(KP876306680, +0.876306680043863587308115903922062583399064238); DVK(KP681693190, +0.681693190061530575150324149145440022633095390); DVK(KP560319534, +0.560319534973832390111614715371676131169633784); DVK(KP855719849, +0.855719849902058969314654733608091555096772472); DVK(KP237294955, +0.237294955877110315393888866460840817927895961); DVK(KP949179823, +0.949179823508441261575555465843363271711583843); DVK(KP904508497, +0.904508497187473712051146708591409529430077295); DVK(KP997675361, +0.997675361079556513670859573984492383596555031); DVK(KP763932022, +0.763932022500210303590826331268723764559381640); DVK(KP690983005, +0.690983005625052575897706582817180941139845410); DVK(KP992114701, +0.992114701314477831049793042785778521453036709); DVK(KP952936919, +0.952936919628306576880750665357914584765951388); DVK(KP998026728, +0.998026728428271561952336806863450553336905220); DVK(KP262346850, +0.262346850930607871785420028382979691334784273); DVK(KP570584518, +0.570584518783621657366766175430996792655723863); DVK(KP669429328, +0.669429328479476605641803240971985825917022098); DVK(KP923225144, +0.923225144846402650453449441572664695995209956); DVK(KP945422727, +0.945422727388575946270360266328811958657216298); DVK(KP522616830, +0.522616830205754336872861364785224694908468440); DVK(KP956723877, +0.956723877038460305821989399535483155872969262); DVK(KP906616052, +0.906616052148196230441134447086066874408359177); DVK(KP772036680, +0.772036680810363904029489473607579825330539880); DVK(KP845997307, +0.845997307939530944175097360758058292389769300); DVK(KP951056516, +0.951056516295153572116439333379382143405698634); DVK(KP912575812, +0.912575812670962425556968549836277086778922727); DVK(KP921078979, +0.921078979742360627699756128143719920817673854); DVK(KP982009705, +0.982009705009746369461829878184175962711969869); DVK(KP734762448, +0.734762448793050413546343770063151342619912334); DVK(KP494780565, +0.494780565770515410344588413655324772219443730); DVK(KP447533225, +0.447533225982656890041886979663652563063114397); DVK(KP603558818, +0.603558818296015001454675132653458027918768137); DVK(KP667278218, +0.667278218140296670899089292254759909713898805); DVK(KP244189809, +0.244189809627953270309879511234821255780225091); DVK(KP269969613, +0.269969613759572083574752974412347470060951301); DVK(KP578046249, +0.578046249379945007321754579646815604023525655); DVK(KP522847744, +0.522847744331509716623755382187077770911012542); DVK(KP132830569, +0.132830569247582714407653942074819768844536507); DVK(KP120146378, +0.120146378570687701782758537356596213647956445); DVK(KP893101515, +0.893101515366181661711202267938416198338079437); DVK(KP987388751, +0.987388751065621252324603216482382109400433949); DVK(KP059835404, +0.059835404262124915169548397419498386427871950); DVK(KP066152395, +0.066152395967733048213034281011006031460903353); DVK(KP786782374, +0.786782374965295178365099601674911834788448471); DVK(KP869845200, +0.869845200362138853122720822420327157933056305); DVK(KP559016994, +0.559016994374947424102293417182819058860154590); DVK(KP250000000, +0.250000000000000000000000000000000000000000000); DVK(KP618033988, +0.618033988749894848204586834365638117720309180); { INT m; R *x; x = ii; for (m = mb, W = W + (mb * ((TWVL / VL) * 8)); m < me; m = m + VL, x = x + (VL * ms), W = W + (TWVL * 8), MAKE_VOLATILE_STRIDE(25, rs)) { V T2t, T1Z, T2W, T28, T2Q, T2r, T2g, T2u, T2o, T2l; { V T2, T5, T3, T9; T2 = LDW(&(W[0])); T5 = LDW(&(W[TWVL * 4])); T3 = LDW(&(W[TWVL * 2])); T9 = LDW(&(W[TWVL * 6])); { V T2c, T3l, Tn, T49, Tm, T4e, TN, T32, T1d, T3a, T3f, T3z, T3H, T25, T1W; V T2v, T2D, T4a, T1g, T18, T2Z, T11, T31, TK, T1q, T1j, T1n, T4b, T17; { V T1, T1l, Tr, T4, Ty, T1E, Tu, TX, TD, T1h, Tz, T1e, T1I, T1o, TU; V Tk, T2b, T1B, T1D, T1N, T1F, Td, T2a, T1J; { V T7, Tb, TC, Tg, T1L, Ta, T6, Tj, T1A; T1 = LD(&(x[0]), ms, &(x[0])); { V Tf, Ti, Te, Th; Tf = LD(&(x[WS(rs, 10)]), ms, &(x[0])); Ti = LD(&(x[WS(rs, 15)]), ms, &(x[WS(rs, 1)])); T7 = LD(&(x[WS(rs, 5)]), ms, &(x[WS(rs, 1)])); Tb = LD(&(x[WS(rs, 20)]), ms, &(x[0])); Te = VZMUL(T2, T5); TC = VZMULJ(T2, T5); T1l = VZMUL(T3, T5); Tr = VZMULJ(T3, T5); T4 = VZMUL(T2, T3); Ty = VZMULJ(T2, T3); T1E = VZMULJ(T2, T9); Th = VZMULJ(T5, T9); Tu = VZMULJ(T3, T9); Tg = VZMUL(Te, Tf); TX = VZMULJ(Te, T9); TD = VZMULJ(TC, T9); T1h = VZMULJ(Ty, T9); Tz = VZMUL(Ty, T5); T1e = VZMULJ(Ty, T5); T1L = VZMULJ(Tr, T9); Ta = VZMULJ(T4, T9); T1I = VZMUL(T4, T5); T6 = VZMULJ(T4, T5); Tj = VZMUL(Th, Ti); } T1A = LD(&(x[WS(rs, 3)]), ms, &(x[WS(rs, 1)])); T1o = VZMULJ(T1e, T9); { V Tc, T8, T1C, T1M; T1C = LD(&(x[WS(rs, 8)]), ms, &(x[0])); T1M = LD(&(x[WS(rs, 18)]), ms, &(x[0])); Tc = VZMUL(Ta, Tb); T8 = VZMUL(T6, T7); TU = VZMULJ(T6, T9); Tk = VADD(Tg, Tj); T2b = VSUB(Tg, Tj); T1B = VZMUL(T3, T1A); T1D = VZMUL(TC, T1C); T1N = VZMUL(T1L, T1M); T1F = LD(&(x[WS(rs, 23)]), ms, &(x[WS(rs, 1)])); Td = VADD(T8, Tc); T2a = VSUB(T8, Tc); T1J = LD(&(x[WS(rs, 13)]), ms, &(x[WS(rs, 1)])); } } { V Tq, Tt, TF, T1T, T1H, Tw, T1U, T1O, TA, Tp, Ts, TE; Tp = LD(&(x[WS(rs, 1)]), ms, &(x[WS(rs, 1)])); Ts = LD(&(x[WS(rs, 6)]), ms, &(x[0])); TE = LD(&(x[WS(rs, 16)]), ms, &(x[0])); { V T1K, Tv, T1G, Tl; Tv = LD(&(x[WS(rs, 21)]), ms, &(x[WS(rs, 1)])); T1G = VZMUL(T1E, T1F); T2c = VFMA(LDK(KP618033988), T2b, T2a); T3l = VFNMS(LDK(KP618033988), T2a, T2b); Tn = VSUB(Td, Tk); Tl = VADD(Td, Tk); T1K = VZMUL(T1I, T1J); Tq = VZMUL(T2, Tp); Tt = VZMUL(Tr, Ts); TF = VZMUL(TD, TE); T1T = VSUB(T1D, T1G); T1H = VADD(T1D, T1G); T49 = VADD(T1, Tl); Tm = VFNMS(LDK(KP250000000), Tl, T1); Tw = VZMUL(Tu, Tv); T1U = VSUB(T1K, T1N); T1O = VADD(T1K, T1N); TA = LD(&(x[WS(rs, 11)]), ms, &(x[WS(rs, 1)])); } { V Tx, TL, T1R, T38, T1V, T13, TQ, TZ, TS, T1Q, TV, TG, TM, T12, T1c; V T16; T12 = LD(&(x[WS(rs, 4)]), ms, &(x[0])); { V TP, TY, T1P, TB, TR; TP = LD(&(x[WS(rs, 24)]), ms, &(x[0])); TY = LD(&(x[WS(rs, 14)]), ms, &(x[0])); TR = LD(&(x[WS(rs, 9)]), ms, &(x[WS(rs, 1)])); Tx = VADD(Tt, Tw); TL = VSUB(Tt, Tw); T1R = VSUB(T1O, T1H); T1P = VADD(T1H, T1O); T38 = VFNMS(LDK(KP618033988), T1T, T1U); T1V = VFMA(LDK(KP618033988), T1U, T1T); TB = VZMUL(Tz, TA); T13 = VZMUL(T4, T12); TQ = VZMUL(T9, TP); TZ = VZMUL(TX, TY); TS = VZMUL(T5, TR); T4e = VADD(T1B, T1P); T1Q = VFNMS(LDK(KP250000000), T1P, T1B); TV = LD(&(x[WS(rs, 19)]), ms, &(x[WS(rs, 1)])); TG = VADD(TB, TF); TM = VSUB(TF, TB); } T1c = LD(&(x[WS(rs, 2)]), ms, &(x[0])); { V T14, TT, TJ, T15, T10, TI, T1p, T1f, T1i, T1m; T1f = LD(&(x[WS(rs, 7)]), ms, &(x[WS(rs, 1)])); T14 = VADD(TS, TQ); TT = VSUB(TQ, TS); { V T39, T1S, TW, TH; T39 = VFMA(LDK(KP559016994), T1R, T1Q); T1S = VFNMS(LDK(KP559016994), T1R, T1Q); TW = VZMUL(TU, TV); TH = VADD(Tx, TG); TJ = VSUB(Tx, TG); TN = VFNMS(LDK(KP618033988), TM, TL); T32 = VFMA(LDK(KP618033988), TL, TM); T1d = VZMUL(Ty, T1c); T3a = VFMA(LDK(KP869845200), T39, T38); T3f = VFNMS(LDK(KP786782374), T38, T39); T3z = VFMA(LDK(KP066152395), T39, T38); T3H = VFNMS(LDK(KP059835404), T38, T39); T25 = VFMA(LDK(KP987388751), T1S, T1V); T1W = VFNMS(LDK(KP893101515), T1V, T1S); T2v = VFNMS(LDK(KP120146378), T1V, T1S); T2D = VFMA(LDK(KP132830569), T1S, T1V); T15 = VADD(TZ, TW); T10 = VSUB(TW, TZ); TI = VFNMS(LDK(KP250000000), TH, Tq); T4a = VADD(Tq, TH); T1g = VZMUL(T1e, T1f); } T1p = LD(&(x[WS(rs, 17)]), ms, &(x[WS(rs, 1)])); T1i = LD(&(x[WS(rs, 22)]), ms, &(x[0])); T1m = LD(&(x[WS(rs, 12)]), ms, &(x[0])); T18 = VSUB(T14, T15); T16 = VADD(T14, T15); T2Z = VFNMS(LDK(KP618033988), TT, T10); T11 = VFMA(LDK(KP618033988), T10, TT); T31 = VFNMS(LDK(KP559016994), TJ, TI); TK = VFMA(LDK(KP559016994), TJ, TI); T1q = VZMUL(T1o, T1p); T1j = VZMUL(T1h, T1i); T1n = VZMUL(T1l, T1m); } T4b = VADD(T13, T16); T17 = VFMS(LDK(KP250000000), T16, T13); } } } { V T33, T3i, T3C, T3L, T20, TO, T2y, T2G, T1k, T1w, T1r, T1x, T2Y, T19, T4k; V T4c; T33 = VFMA(LDK(KP893101515), T32, T31); T3i = VFNMS(LDK(KP987388751), T31, T32); T3C = VFNMS(LDK(KP522847744), T32, T31); T3L = VFMA(LDK(KP578046249), T31, T32); T20 = VFMA(LDK(KP269969613), TK, TN); TO = VFNMS(LDK(KP244189809), TN, TK); T2y = VFMA(LDK(KP667278218), TK, TN); T2G = VFNMS(LDK(KP603558818), TN, TK); T1k = VADD(T1g, T1j); T1w = VSUB(T1g, T1j); T1r = VADD(T1n, T1q); T1x = VSUB(T1q, T1n); T2Y = VFMA(LDK(KP559016994), T18, T17); T19 = VFNMS(LDK(KP559016994), T18, T17); T4k = VSUB(T4a, T4b); T4c = VADD(T4a, T4b); { V T2X, To, T35, T1y, T2H, T2z, T1a, T21, T3t, T34, T3n, T3j, T3E, T3Y, T3M; V T3R, T1v, T36, T4l, T4f, T1u, T1s; T2X = VFNMS(LDK(KP559016994), Tn, Tm); To = VFMA(LDK(KP559016994), Tn, Tm); T1u = VSUB(T1r, T1k); T1s = VADD(T1k, T1r); T35 = VFMA(LDK(KP618033988), T1w, T1x); T1y = VFNMS(LDK(KP618033988), T1x, T1w); { V T3K, T30, T3h, T3D, T4d, T1t; T3K = VFMA(LDK(KP447533225), T2Z, T2Y); T30 = VFMA(LDK(KP120146378), T2Z, T2Y); T3h = VFNMS(LDK(KP132830569), T2Y, T2Z); T3D = VFNMS(LDK(KP494780565), T2Y, T2Z); T2H = VFNMS(LDK(KP786782374), T11, T19); T2z = VFMA(LDK(KP869845200), T19, T11); T1a = VFNMS(LDK(KP667278218), T19, T11); T21 = VFMA(LDK(KP603558818), T11, T19); T4d = VADD(T1d, T1s); T1t = VFNMS(LDK(KP250000000), T1s, T1d); T3t = VFNMS(LDK(KP734762448), T33, T30); T34 = VFMA(LDK(KP734762448), T33, T30); T3n = VFMA(LDK(KP734762448), T3i, T3h); T3j = VFNMS(LDK(KP734762448), T3i, T3h); T3E = VFNMS(LDK(KP982009705), T3D, T3C); T3Y = VFMA(LDK(KP982009705), T3D, T3C); T3M = VFNMS(LDK(KP921078979), T3L, T3K); T3R = VFMA(LDK(KP921078979), T3L, T3K); T1v = VFNMS(LDK(KP559016994), T1u, T1t); T36 = VFMA(LDK(KP559016994), T1u, T1t); T4l = VSUB(T4d, T4e); T4f = VADD(T4d, T4e); } { V T2L, T2R, T2j, T2q, T2J, T2B, T2e, T26, T2U, T1Y, T23, T2O; { V T2I, T24, T2w, T2E, T48, T42, T3y, T3s, T3V, T45, T2A, T1b, T2h, T2i, T1X; T2L = VFNMS(LDK(KP912575812), T2H, T2G); T2I = VFMA(LDK(KP912575812), T2H, T2G); { V T3A, T3e, T37, T3I, T1z; T3A = VFNMS(LDK(KP667278218), T36, T35); T3e = VFNMS(LDK(KP059835404), T35, T36); T37 = VFMA(LDK(KP066152395), T36, T35); T3I = VFMA(LDK(KP603558818), T35, T36); T24 = VFMA(LDK(KP578046249), T1v, T1y); T1z = VFNMS(LDK(KP522847744), T1y, T1v); T2w = VFNMS(LDK(KP494780565), T1v, T1y); T2E = VFMA(LDK(KP447533225), T1y, T1v); { V T4i, T4g, T4o, T4m; T4i = VSUB(T4c, T4f); T4g = VADD(T4c, T4f); T4o = VMUL(LDK(KP951056516), VFNMS(LDK(KP618033988), T4k, T4l)); T4m = VMUL(LDK(KP951056516), VFMA(LDK(KP618033988), T4l, T4k)); { V T3Q, T3J, T3b, T3u; T3Q = VFNMS(LDK(KP845997307), T3I, T3H); T3J = VFMA(LDK(KP845997307), T3I, T3H); T3b = VFNMS(LDK(KP772036680), T3a, T37); T3u = VFMA(LDK(KP772036680), T3a, T37); { V T3o, T3g, T3B, T3X, T4h; T3o = VFNMS(LDK(KP772036680), T3f, T3e); T3g = VFMA(LDK(KP772036680), T3f, T3e); T3B = VFNMS(LDK(KP845997307), T3A, T3z); T3X = VFMA(LDK(KP845997307), T3A, T3z); ST(&(x[0]), VADD(T4g, T49), ms, &(x[0])); T4h = VFNMS(LDK(KP250000000), T4g, T49); { V T40, T3N, T3c, T3v; T40 = VFMA(LDK(KP906616052), T3M, T3J); T3N = VFNMS(LDK(KP906616052), T3M, T3J); T3c = VFMA(LDK(KP956723877), T3b, T34); T3v = VFMA(LDK(KP522616830), T3j, T3u); { V T3p, T3k, T3S, T3F; T3p = VFNMS(LDK(KP522616830), T34, T3o); T3k = VFMA(LDK(KP945422727), T3j, T3g); T3S = VFNMS(LDK(KP923225144), T3E, T3B); T3F = VFMA(LDK(KP923225144), T3E, T3B); { V T46, T3Z, T4j, T4n; T46 = VFNMS(LDK(KP669429328), T3X, T3Y); T3Z = VFMA(LDK(KP570584518), T3Y, T3X); T4j = VFMA(LDK(KP559016994), T4i, T4h); T4n = VFNMS(LDK(KP559016994), T4i, T4h); { V T3W, T3O, T3d, T3w; T3W = VFMA(LDK(KP262346850), T3N, T3l); T3O = VMUL(LDK(KP998026728), VFNMS(LDK(KP952936919), T3l, T3N)); T3d = VFMA(LDK(KP992114701), T3c, T2X); T3w = VFNMS(LDK(KP690983005), T3v, T3g); { V T3q, T3m, T3T, T43; T3q = VFMA(LDK(KP763932022), T3p, T3b); T3m = VMUL(LDK(KP998026728), VFMA(LDK(KP952936919), T3l, T3k)); T3T = VFNMS(LDK(KP997675361), T3S, T3R); T43 = VFNMS(LDK(KP904508497), T3S, T3Q); { V T3G, T3P, T47, T41; T3G = VFMA(LDK(KP949179823), T3F, T2X); T3P = VFNMS(LDK(KP237294955), T3F, T2X); T47 = VFNMS(LDK(KP669429328), T40, T46); T41 = VFMA(LDK(KP618033988), T40, T3Z); ST(&(x[WS(rs, 20)]), VFNMSI(T4m, T4j), ms, &(x[0])); ST(&(x[WS(rs, 5)]), VFMAI(T4m, T4j), ms, &(x[WS(rs, 1)])); ST(&(x[WS(rs, 15)]), VFMAI(T4o, T4n), ms, &(x[WS(rs, 1)])); ST(&(x[WS(rs, 10)]), VFNMSI(T4o, T4n), ms, &(x[0])); { V T3x, T3r, T3U, T44; T3x = VFMA(LDK(KP855719849), T3w, T3t); T3r = VFNMS(LDK(KP855719849), T3q, T3n); ST(&(x[WS(rs, 3)]), VFMAI(T3m, T3d), ms, &(x[WS(rs, 1)])); ST(&(x[WS(rs, 22)]), VFNMSI(T3m, T3d), ms, &(x[0])); T3U = VFMA(LDK(KP560319534), T3T, T3Q); T44 = VFNMS(LDK(KP681693190), T43, T3R); ST(&(x[WS(rs, 2)]), VFMAI(T3O, T3G), ms, &(x[0])); ST(&(x[WS(rs, 23)]), VFNMSI(T3O, T3G), ms, &(x[WS(rs, 1)])); T48 = VMUL(LDK(KP951056516), VFNMS(LDK(KP876306680), T47, T3W)); T42 = VMUL(LDK(KP951056516), VFNMS(LDK(KP949179823), T41, T3W)); T3y = VMUL(LDK(KP951056516), VFNMS(LDK(KP992114701), T3x, T3l)); T3s = VFMA(LDK(KP897376177), T3r, T2X); T3V = VFNMS(LDK(KP949179823), T3U, T3P); T45 = VFNMS(LDK(KP860541664), T44, T3P); T2R = VFNMS(LDK(KP912575812), T2z, T2y); T2A = VFMA(LDK(KP912575812), T2z, T2y); T1b = VFMA(LDK(KP829049696), T1a, TO); T2h = VFNMS(LDK(KP829049696), T1a, TO); T2i = VFNMS(LDK(KP831864738), T1W, T1z); T1X = VFMA(LDK(KP831864738), T1W, T1z); } } } } } } } } } } } { V T2M, T2F, T2x, T2S, T2T, T2N; T2M = VFNMS(LDK(KP958953096), T2E, T2D); T2F = VFMA(LDK(KP958953096), T2E, T2D); ST(&(x[WS(rs, 17)]), VFNMSI(T3y, T3s), ms, &(x[WS(rs, 1)])); ST(&(x[WS(rs, 8)]), VFMAI(T3y, T3s), ms, &(x[0])); ST(&(x[WS(rs, 13)]), VFMAI(T42, T3V), ms, &(x[WS(rs, 1)])); ST(&(x[WS(rs, 12)]), VFNMSI(T42, T3V), ms, &(x[0])); ST(&(x[WS(rs, 7)]), VFNMSI(T48, T45), ms, &(x[WS(rs, 1)])); ST(&(x[WS(rs, 18)]), VFMAI(T48, T45), ms, &(x[0])); T2j = VFMA(LDK(KP559154169), T2i, T2h); T2q = VFNMS(LDK(KP683113946), T2h, T2i); T2x = VFNMS(LDK(KP867381224), T2w, T2v); T2S = VFMA(LDK(KP867381224), T2w, T2v); T2J = VFMA(LDK(KP894834959), T2I, T2F); T2T = VFMA(LDK(KP447417479), T2I, T2S); T2B = VFNMS(LDK(KP809385824), T2A, T2x); T2N = VFMA(LDK(KP447417479), T2A, T2M); T2e = VFMA(LDK(KP831864738), T25, T24); T26 = VFNMS(LDK(KP831864738), T25, T24); T2U = VFNMS(LDK(KP763932022), T2T, T2F); T1Y = VFMA(LDK(KP904730450), T1X, T1b); T23 = VFNMS(LDK(KP904730450), T1X, T1b); T2O = VFMA(LDK(KP690983005), T2N, T2x); } } { V T2C, T22, T2d, T2K; T2C = VFNMS(LDK(KP992114701), T2B, To); T22 = VFMA(LDK(KP916574801), T21, T20); T2d = VFNMS(LDK(KP916574801), T21, T20); T2K = VMUL(LDK(KP951056516), VFNMS(LDK(KP992114701), T2J, T2c)); { V T27, T2P, T2f, T2k, T2n, T2V; T2V = VFNMS(LDK(KP999544308), T2U, T2R); T27 = VFNMS(LDK(KP904730450), T26, T23); T2t = VFMA(LDK(KP968583161), T1Y, To); T1Z = VFNMS(LDK(KP242145790), T1Y, To); T2P = VFNMS(LDK(KP999544308), T2O, T2L); T2f = VFMA(LDK(KP904730450), T2e, T2d); T2k = VFNMS(LDK(KP904730450), T2e, T2d); T2n = VADD(T22, T23); ST(&(x[WS(rs, 21)]), VFMAI(T2K, T2C), ms, &(x[WS(rs, 1)])); ST(&(x[WS(rs, 4)]), VFNMSI(T2K, T2C), ms, &(x[0])); T2W = VMUL(LDK(KP951056516), VFNMS(LDK(KP803003575), T2V, T2c)); T28 = VFNMS(LDK(KP618033988), T27, T22); T2Q = VFNMS(LDK(KP803003575), T2P, To); T2r = VFMA(LDK(KP617882369), T2k, T2q); T2g = VFNMS(LDK(KP242145790), T2f, T2c); T2u = VMUL(LDK(KP951056516), VFMA(LDK(KP968583161), T2f, T2c)); T2o = VFNMS(LDK(KP683113946), T2n, T26); T2l = VFMA(LDK(KP559016994), T2k, T2j); } } } } } } } { V T29, T2s, T2p, T2m; T29 = VFNMS(LDK(KP876091699), T28, T1Z); ST(&(x[WS(rs, 16)]), VFMAI(T2W, T2Q), ms, &(x[0])); ST(&(x[WS(rs, 9)]), VFNMSI(T2W, T2Q), ms, &(x[WS(rs, 1)])); T2s = VMUL(LDK(KP951056516), VFNMS(LDK(KP876306680), T2r, T2g)); ST(&(x[WS(rs, 24)]), VFNMSI(T2u, T2t), ms, &(x[0])); ST(&(x[WS(rs, 1)]), VFMAI(T2u, T2t), ms, &(x[WS(rs, 1)])); T2p = VFMA(LDK(KP792626838), T2o, T1Z); T2m = VMUL(LDK(KP951056516), VFMA(LDK(KP968583161), T2l, T2g)); ST(&(x[WS(rs, 11)]), VFMAI(T2s, T2p), ms, &(x[WS(rs, 1)])); ST(&(x[WS(rs, 14)]), VFNMSI(T2s, T2p), ms, &(x[0])); ST(&(x[WS(rs, 19)]), VFNMSI(T2m, T29), ms, &(x[WS(rs, 1)])); ST(&(x[WS(rs, 6)]), VFMAI(T2m, T29), ms, &(x[0])); } } } VLEAVE(); } static const tw_instr twinstr[] = { VTW(0, 1), VTW(0, 3), VTW(0, 9), VTW(0, 24), {TW_NEXT, VL, 0} }; static const ct_desc desc = { 25, XSIMD_STRING("t3bv_25"), twinstr, &GENUS, {87, 100, 181, 0}, 0, 0, 0 }; void XSIMD(codelet_t3bv_25) (planner *p) { X(kdft_dit_register) (p, t3bv_25, &desc); } #else /* HAVE_FMA */ /* Generated by: ../../../genfft/gen_twiddle_c.native -simd -compact -variables 4 -pipeline-latency 8 -twiddle-log3 -precompute-twiddles -no-generate-bytw -n 25 -name t3bv_25 -include t3b.h -sign 1 */ /* * This function contains 268 FP additions, 228 FP multiplications, * (or, 191 additions, 151 multiplications, 77 fused multiply/add), * 124 stack variables, 40 constants, and 50 memory accesses */ #include "t3b.h" static void t3bv_25(R *ri, R *ii, const R *W, stride rs, INT mb, INT me, INT ms) { DVK(KP497379774, +0.497379774329709576484567492012895936835134813); DVK(KP968583161, +0.968583161128631119490168375464735813836012403); DVK(KP248689887, +0.248689887164854788242283746006447968417567406); DVK(KP1_937166322, +1.937166322257262238980336750929471627672024806); DVK(KP809016994, +0.809016994374947424102293417182819058860154590); DVK(KP309016994, +0.309016994374947424102293417182819058860154590); DVK(KP1_688655851, +1.688655851004030157097116127933363010763318483); DVK(KP535826794, +0.535826794978996618271308767867639978063575346); DVK(KP425779291, +0.425779291565072648862502445744251703979973042); DVK(KP1_809654104, +1.809654104932039055427337295865395187940827822); DVK(KP963507348, +0.963507348203430549974383005744259307057084020); DVK(KP876306680, +0.876306680043863587308115903922062583399064238); DVK(KP844327925, +0.844327925502015078548558063966681505381659241); DVK(KP1_071653589, +1.071653589957993236542617535735279956127150691); DVK(KP481753674, +0.481753674101715274987191502872129653528542010); DVK(KP1_752613360, +1.752613360087727174616231807844125166798128477); DVK(KP851558583, +0.851558583130145297725004891488503407959946084); DVK(KP904827052, +0.904827052466019527713668647932697593970413911); DVK(KP125333233, +0.125333233564304245373118759816508793942918247); DVK(KP1_984229402, +1.984229402628955662099586085571557042906073418); DVK(KP1_457937254, +1.457937254842823046293460638110518222745143328); DVK(KP684547105, +0.684547105928688673732283357621209269889519233); DVK(KP637423989, +0.637423989748689710176712811676016195434917298); DVK(KP1_541026485, +1.541026485551578461606019272792355694543335344); DVK(KP062790519, +0.062790519529313376076178224565631133122484832); DVK(KP1_996053456, +1.996053456856543123904673613726901106673810439); DVK(KP770513242, +0.770513242775789230803009636396177847271667672); DVK(KP1_274847979, +1.274847979497379420353425623352032390869834596); DVK(KP125581039, +0.125581039058626752152356449131262266244969664); DVK(KP998026728, +0.998026728428271561952336806863450553336905220); DVK(KP992114701, +0.992114701314477831049793042785778521453036709); DVK(KP250666467, +0.250666467128608490746237519633017587885836494); DVK(KP728968627, +0.728968627421411523146730319055259111372571664); DVK(KP1_369094211, +1.369094211857377347464566715242418539779038465); DVK(KP293892626, +0.293892626146236564584352977319536384298826219); DVK(KP475528258, +0.475528258147576786058219666689691071702849317); DVK(KP951056516, +0.951056516295153572116439333379382143405698634); DVK(KP250000000, +0.250000000000000000000000000000000000000000000); DVK(KP587785252, +0.587785252292473129168705954639072768597652438); DVK(KP559016994, +0.559016994374947424102293417182819058860154590); { INT m; R *x; x = ii; for (m = mb, W = W + (mb * ((TWVL / VL) * 8)); m < me; m = m + VL, x = x + (VL * ms), W = W + (TWVL * 8), MAKE_VOLATILE_STRIDE(25, rs)) { V T1, Td, T8, T9, TF, Te, Tu, TB, TC, T1s, T15, Tf, TY, T4, Ta; V Tx, T1T, Tg, T1N, T1v, T18, TG, T1o, T11; T1 = LDW(&(W[TWVL * 4])); Td = LDW(&(W[TWVL * 2])); T8 = LDW(&(W[0])); T9 = VZMUL(T8, T1); TF = VZMULJ(T8, T1); Te = VZMUL(T8, Td); Tu = VZMULJ(Td, T1); TB = VZMULJ(T8, Td); TC = VZMUL(TB, T1); T1s = VZMUL(Te, T1); T15 = VZMUL(Td, T1); Tf = VZMULJ(Te, T1); TY = VZMULJ(TB, T1); T4 = LDW(&(W[TWVL * 6])); Ta = VZMULJ(T9, T4); Tx = VZMULJ(Td, T4); T1T = VZMULJ(T1, T4); Tg = VZMULJ(Tf, T4); T1N = VZMULJ(Te, T4); T1v = VZMULJ(Tu, T4); T18 = VZMULJ(TY, T4); TG = VZMULJ(TF, T4); T1o = VZMULJ(T8, T4); T11 = VZMULJ(TB, T4); { V T1Y, T1X, T2f, T2g, T1Z, T20, T2e, T39, T1H, T2T, T1E, T3C, T2S, Tk, T2G; V Ts, T3z, T2F, TK, T2I, TS, T3y, T2J, T1k, T2Q, T1h, T3B, T2P; { V T1S, T1V, T1W, T1M, T1P, T1Q, T2d; T1Y = LD(&(x[0]), ms, &(x[0])); { V T1R, T1U, T1L, T1O; T1R = LD(&(x[WS(rs, 10)]), ms, &(x[0])); T1S = VZMUL(T9, T1R); T1U = LD(&(x[WS(rs, 15)]), ms, &(x[WS(rs, 1)])); T1V = VZMUL(T1T, T1U); T1W = VADD(T1S, T1V); T1L = LD(&(x[WS(rs, 5)]), ms, &(x[WS(rs, 1)])); T1M = VZMUL(Tf, T1L); T1O = LD(&(x[WS(rs, 20)]), ms, &(x[0])); T1P = VZMUL(T1N, T1O); T1Q = VADD(T1M, T1P); } T1X = VMUL(LDK(KP559016994), VSUB(T1Q, T1W)); T2f = VSUB(T1S, T1V); T2g = VMUL(LDK(KP587785252), T2f); T1Z = VADD(T1Q, T1W); T20 = VFNMS(LDK(KP250000000), T1Z, T1Y); T2d = VSUB(T1M, T1P); T2e = VMUL(LDK(KP951056516), T2d); T39 = VMUL(LDK(KP587785252), T2d); } { V T1B, T1u, T1x, T1y, T1n, T1q, T1r, T1A; T1A = LD(&(x[WS(rs, 3)]), ms, &(x[WS(rs, 1)])); T1B = VZMUL(Td, T1A); { V T1t, T1w, T1m, T1p; T1t = LD(&(x[WS(rs, 13)]), ms, &(x[WS(rs, 1)])); T1u = VZMUL(T1s, T1t); T1w = LD(&(x[WS(rs, 18)]), ms, &(x[0])); T1x = VZMUL(T1v, T1w); T1y = VADD(T1u, T1x); T1m = LD(&(x[WS(rs, 8)]), ms, &(x[0])); T1n = VZMUL(TF, T1m); T1p = LD(&(x[WS(rs, 23)]), ms, &(x[WS(rs, 1)])); T1q = VZMUL(T1o, T1p); T1r = VADD(T1n, T1q); } { V T1F, T1G, T1z, T1C, T1D; T1F = VSUB(T1n, T1q); T1G = VSUB(T1u, T1x); T1H = VFMA(LDK(KP475528258), T1F, VMUL(LDK(KP293892626), T1G)); T2T = VFNMS(LDK(KP475528258), T1G, VMUL(LDK(KP293892626), T1F)); T1z = VMUL(LDK(KP559016994), VSUB(T1r, T1y)); T1C = VADD(T1r, T1y); T1D = VFNMS(LDK(KP250000000), T1C, T1B); T1E = VADD(T1z, T1D); T3C = VADD(T1B, T1C); T2S = VSUB(T1D, T1z); } } { V Tp, Tc, Ti, Tm, T3, T6, Tl, To; To = LD(&(x[WS(rs, 4)]), ms, &(x[0])); Tp = VZMUL(Te, To); { V Tb, Th, T2, T5; Tb = LD(&(x[WS(rs, 14)]), ms, &(x[0])); Tc = VZMUL(Ta, Tb); Th = LD(&(x[WS(rs, 19)]), ms, &(x[WS(rs, 1)])); Ti = VZMUL(Tg, Th); Tm = VADD(Tc, Ti); T2 = LD(&(x[WS(rs, 9)]), ms, &(x[WS(rs, 1)])); T3 = VZMUL(T1, T2); T5 = LD(&(x[WS(rs, 24)]), ms, &(x[0])); T6 = VZMUL(T4, T5); Tl = VADD(T3, T6); } { V T7, Tj, Tn, Tq, Tr; T7 = VSUB(T3, T6); Tj = VSUB(Tc, Ti); Tk = VFMA(LDK(KP475528258), T7, VMUL(LDK(KP293892626), Tj)); T2G = VFNMS(LDK(KP475528258), Tj, VMUL(LDK(KP293892626), T7)); Tn = VMUL(LDK(KP559016994), VSUB(Tl, Tm)); Tq = VADD(Tl, Tm); Tr = VFNMS(LDK(KP250000000), Tq, Tp); Ts = VADD(Tn, Tr); T3z = VADD(Tp, Tq); T2F = VSUB(Tr, Tn); } } { V TP, TE, TI, TM, Tw, Tz, TL, TO; TO = LD(&(x[WS(rs, 1)]), ms, &(x[WS(rs, 1)])); TP = VZMUL(T8, TO); { V TD, TH, Tv, Ty; TD = LD(&(x[WS(rs, 11)]), ms, &(x[WS(rs, 1)])); TE = VZMUL(TC, TD); TH = LD(&(x[WS(rs, 16)]), ms, &(x[0])); TI = VZMUL(TG, TH); TM = VADD(TE, TI); Tv = LD(&(x[WS(rs, 6)]), ms, &(x[0])); Tw = VZMUL(Tu, Tv); Ty = LD(&(x[WS(rs, 21)]), ms, &(x[WS(rs, 1)])); Tz = VZMUL(Tx, Ty); TL = VADD(Tw, Tz); } { V TA, TJ, TN, TQ, TR; TA = VSUB(Tw, Tz); TJ = VSUB(TE, TI); TK = VFMA(LDK(KP475528258), TA, VMUL(LDK(KP293892626), TJ)); T2I = VFNMS(LDK(KP475528258), TJ, VMUL(LDK(KP293892626), TA)); TN = VMUL(LDK(KP559016994), VSUB(TL, TM)); TQ = VADD(TL, TM); TR = VFNMS(LDK(KP250000000), TQ, TP); TS = VADD(TN, TR); T3y = VADD(TP, TQ); T2J = VSUB(TR, TN); } } { V T1e, T17, T1a, T1b, T10, T13, T14, T1d; T1d = LD(&(x[WS(rs, 2)]), ms, &(x[0])); T1e = VZMUL(TB, T1d); { V T16, T19, TZ, T12; T16 = LD(&(x[WS(rs, 12)]), ms, &(x[0])); T17 = VZMUL(T15, T16); T19 = LD(&(x[WS(rs, 17)]), ms, &(x[WS(rs, 1)])); T1a = VZMUL(T18, T19); T1b = VADD(T17, T1a); TZ = LD(&(x[WS(rs, 7)]), ms, &(x[WS(rs, 1)])); T10 = VZMUL(TY, TZ); T12 = LD(&(x[WS(rs, 22)]), ms, &(x[0])); T13 = VZMUL(T11, T12); T14 = VADD(T10, T13); } { V T1i, T1j, T1c, T1f, T1g; T1i = VSUB(T10, T13); T1j = VSUB(T17, T1a); T1k = VFMA(LDK(KP475528258), T1i, VMUL(LDK(KP293892626), T1j)); T2Q = VFNMS(LDK(KP475528258), T1j, VMUL(LDK(KP293892626), T1i)); T1c = VMUL(LDK(KP559016994), VSUB(T14, T1b)); T1f = VADD(T14, T1b); T1g = VFNMS(LDK(KP250000000), T1f, T1e); T1h = VADD(T1c, T1g); T3B = VADD(T1e, T1f); T2P = VSUB(T1g, T1c); } } { V T3E, T3M, T3I, T3J, T3H, T3K, T3N, T3L; { V T3A, T3D, T3F, T3G; T3A = VSUB(T3y, T3z); T3D = VSUB(T3B, T3C); T3E = VBYI(VFMA(LDK(KP951056516), T3A, VMUL(LDK(KP587785252), T3D))); T3M = VBYI(VFNMS(LDK(KP951056516), T3D, VMUL(LDK(KP587785252), T3A))); T3I = VADD(T1Y, T1Z); T3F = VADD(T3y, T3z); T3G = VADD(T3B, T3C); T3J = VADD(T3F, T3G); T3H = VMUL(LDK(KP559016994), VSUB(T3F, T3G)); T3K = VFNMS(LDK(KP250000000), T3J, T3I); } ST(&(x[0]), VADD(T3I, T3J), ms, &(x[0])); T3N = VSUB(T3K, T3H); ST(&(x[WS(rs, 10)]), VADD(T3M, T3N), ms, &(x[0])); ST(&(x[WS(rs, 15)]), VSUB(T3N, T3M), ms, &(x[WS(rs, 1)])); T3L = VADD(T3H, T3K); ST(&(x[WS(rs, 5)]), VADD(T3E, T3L), ms, &(x[WS(rs, 1)])); ST(&(x[WS(rs, 20)]), VSUB(T3L, T3E), ms, &(x[0])); } { V T2X, T3a, T3i, T3j, T3k, T3s, T3t, T3u, T3l, T3m, T3n, T3p, T3q, T3r, T2L; V T3b, T32, T38, T2W, T35, T2Y, T34, T3w, T3x; T2X = VSUB(T20, T1X); T3a = VFNMS(LDK(KP951056516), T2f, T39); T3i = VFMA(LDK(KP1_369094211), T2I, VMUL(LDK(KP728968627), T2J)); T3j = VFNMS(LDK(KP992114701), T2F, VMUL(LDK(KP250666467), T2G)); T3k = VADD(T3i, T3j); T3s = VFNMS(LDK(KP125581039), T2Q, VMUL(LDK(KP998026728), T2P)); T3t = VFMA(LDK(KP1_274847979), T2T, VMUL(LDK(KP770513242), T2S)); T3u = VADD(T3s, T3t); T3l = VFMA(LDK(KP1_996053456), T2Q, VMUL(LDK(KP062790519), T2P)); T3m = VFNMS(LDK(KP637423989), T2S, VMUL(LDK(KP1_541026485), T2T)); T3n = VADD(T3l, T3m); T3p = VFNMS(LDK(KP1_457937254), T2I, VMUL(LDK(KP684547105), T2J)); T3q = VFMA(LDK(KP1_984229402), T2G, VMUL(LDK(KP125333233), T2F)); T3r = VADD(T3p, T3q); { V T2H, T2K, T36, T30, T31, T37; T2H = VFNMS(LDK(KP851558583), T2G, VMUL(LDK(KP904827052), T2F)); T2K = VFMA(LDK(KP1_752613360), T2I, VMUL(LDK(KP481753674), T2J)); T36 = VADD(T2K, T2H); T30 = VFMA(LDK(KP1_071653589), T2Q, VMUL(LDK(KP844327925), T2P)); T31 = VFMA(LDK(KP125581039), T2T, VMUL(LDK(KP998026728), T2S)); T37 = VADD(T30, T31); T2L = VSUB(T2H, T2K); T3b = VADD(T36, T37); T32 = VSUB(T30, T31); T38 = VMUL(LDK(KP559016994), VSUB(T36, T37)); } { V T2M, T2N, T2O, T2R, T2U, T2V; T2M = VFNMS(LDK(KP963507348), T2I, VMUL(LDK(KP876306680), T2J)); T2N = VFMA(LDK(KP1_809654104), T2G, VMUL(LDK(KP425779291), T2F)); T2O = VSUB(T2M, T2N); T2R = VFNMS(LDK(KP1_688655851), T2Q, VMUL(LDK(KP535826794), T2P)); T2U = VFNMS(LDK(KP1_996053456), T2T, VMUL(LDK(KP062790519), T2S)); T2V = VADD(T2R, T2U); T2W = VMUL(LDK(KP559016994), VSUB(T2O, T2V)); T35 = VSUB(T2R, T2U); T2Y = VADD(T2O, T2V); T34 = VADD(T2M, T2N); } { V T3g, T3h, T3o, T3v; T3g = VADD(T2X, T2Y); T3h = VBYI(VADD(T3a, T3b)); ST(&(x[WS(rs, 23)]), VSUB(T3g, T3h), ms, &(x[WS(rs, 1)])); ST(&(x[WS(rs, 2)]), VADD(T3g, T3h), ms, &(x[0])); T3o = VADD(T2X, VADD(T3k, T3n)); T3v = VBYI(VSUB(VADD(T3r, T3u), T3a)); ST(&(x[WS(rs, 22)]), VSUB(T3o, T3v), ms, &(x[0])); ST(&(x[WS(rs, 3)]), VADD(T3o, T3v), ms, &(x[WS(rs, 1)])); } T3w = VBYI(VSUB(VFMA(LDK(KP951056516), VSUB(T3i, T3j), VFMA(LDK(KP309016994), T3r, VFNMS(LDK(KP809016994), T3u, VMUL(LDK(KP587785252), VSUB(T3l, T3m))))), T3a)); T3x = VFMA(LDK(KP309016994), T3k, VFMA(LDK(KP951056516), VSUB(T3q, T3p), VFMA(LDK(KP587785252), VSUB(T3t, T3s), VFNMS(LDK(KP809016994), T3n, T2X)))); ST(&(x[WS(rs, 8)]), VADD(T3w, T3x), ms, &(x[0])); ST(&(x[WS(rs, 17)]), VSUB(T3x, T3w), ms, &(x[WS(rs, 1)])); { V T33, T3e, T3d, T3f, T2Z, T3c; T2Z = VFNMS(LDK(KP250000000), T2Y, T2X); T33 = VFMA(LDK(KP951056516), T2L, VADD(T2W, VFNMS(LDK(KP587785252), T32, T2Z))); T3e = VFMA(LDK(KP587785252), T2L, VFMA(LDK(KP951056516), T32, VSUB(T2Z, T2W))); T3c = VFNMS(LDK(KP250000000), T3b, T3a); T3d = VBYI(VADD(VFMA(LDK(KP951056516), T34, VMUL(LDK(KP587785252), T35)), VADD(T38, T3c))); T3f = VBYI(VADD(VFNMS(LDK(KP951056516), T35, VMUL(LDK(KP587785252), T34)), VSUB(T3c, T38))); ST(&(x[WS(rs, 18)]), VSUB(T33, T3d), ms, &(x[0])); ST(&(x[WS(rs, 12)]), VADD(T3e, T3f), ms, &(x[0])); ST(&(x[WS(rs, 7)]), VADD(T33, T3d), ms, &(x[WS(rs, 1)])); ST(&(x[WS(rs, 13)]), VSUB(T3e, T3f), ms, &(x[WS(rs, 1)])); } } { V T21, T2h, T2p, T2q, T2r, T2z, T2A, T2B, T2s, T2t, T2u, T2w, T2x, T2y, TU; V T2i, T26, T2c, T1K, T29, T22, T28, T2D, T2E; T21 = VADD(T1X, T20); T2h = VADD(T2e, T2g); T2p = VFMA(LDK(KP1_688655851), TK, VMUL(LDK(KP535826794), TS)); T2q = VFMA(LDK(KP1_541026485), Tk, VMUL(LDK(KP637423989), Ts)); T2r = VSUB(T2p, T2q); T2z = VFMA(LDK(KP851558583), T1k, VMUL(LDK(KP904827052), T1h)); T2A = VFMA(LDK(KP1_984229402), T1H, VMUL(LDK(KP125333233), T1E)); T2B = VADD(T2z, T2A); T2s = VFNMS(LDK(KP425779291), T1h, VMUL(LDK(KP1_809654104), T1k)); T2t = VFNMS(LDK(KP992114701), T1E, VMUL(LDK(KP250666467), T1H)); T2u = VADD(T2s, T2t); T2w = VFNMS(LDK(KP1_071653589), TK, VMUL(LDK(KP844327925), TS)); T2x = VFNMS(LDK(KP770513242), Ts, VMUL(LDK(KP1_274847979), Tk)); T2y = VADD(T2w, T2x); { V Tt, TT, T2a, T24, T25, T2b; Tt = VFMA(LDK(KP1_071653589), Tk, VMUL(LDK(KP844327925), Ts)); TT = VFMA(LDK(KP1_937166322), TK, VMUL(LDK(KP248689887), TS)); T2a = VADD(TT, Tt); T24 = VFMA(LDK(KP1_752613360), T1k, VMUL(LDK(KP481753674), T1h)); T25 = VFMA(LDK(KP1_457937254), T1H, VMUL(LDK(KP684547105), T1E)); T2b = VADD(T24, T25); TU = VSUB(Tt, TT); T2i = VADD(T2a, T2b); T26 = VSUB(T24, T25); T2c = VMUL(LDK(KP559016994), VSUB(T2a, T2b)); } { V TV, TW, TX, T1l, T1I, T1J; TV = VFNMS(LDK(KP497379774), TK, VMUL(LDK(KP968583161), TS)); TW = VFNMS(LDK(KP1_688655851), Tk, VMUL(LDK(KP535826794), Ts)); TX = VADD(TV, TW); T1l = VFNMS(LDK(KP963507348), T1k, VMUL(LDK(KP876306680), T1h)); T1I = VFNMS(LDK(KP1_369094211), T1H, VMUL(LDK(KP728968627), T1E)); T1J = VADD(T1l, T1I); T1K = VMUL(LDK(KP559016994), VSUB(TX, T1J)); T29 = VSUB(T1l, T1I); T22 = VADD(TX, T1J); T28 = VSUB(TV, TW); } { V T2n, T2o, T2v, T2C; T2n = VADD(T21, T22); T2o = VBYI(VADD(T2h, T2i)); ST(&(x[WS(rs, 24)]), VSUB(T2n, T2o), ms, &(x[0])); ST(&(x[WS(rs, 1)]), VADD(T2n, T2o), ms, &(x[WS(rs, 1)])); T2v = VADD(T21, VADD(T2r, T2u)); T2C = VBYI(VSUB(VADD(T2y, T2B), T2h)); ST(&(x[WS(rs, 21)]), VSUB(T2v, T2C), ms, &(x[WS(rs, 1)])); ST(&(x[WS(rs, 4)]), VADD(T2v, T2C), ms, &(x[0])); } T2D = VBYI(VSUB(VFMA(LDK(KP309016994), T2y, VFMA(LDK(KP951056516), VADD(T2p, T2q), VFNMS(LDK(KP809016994), T2B, VMUL(LDK(KP587785252), VSUB(T2s, T2t))))), T2h)); T2E = VFMA(LDK(KP951056516), VSUB(T2x, T2w), VFMA(LDK(KP309016994), T2r, VFMA(LDK(KP587785252), VSUB(T2A, T2z), VFNMS(LDK(KP809016994), T2u, T21)))); ST(&(x[WS(rs, 9)]), VADD(T2D, T2E), ms, &(x[WS(rs, 1)])); ST(&(x[WS(rs, 16)]), VSUB(T2E, T2D), ms, &(x[0])); { V T27, T2l, T2k, T2m, T23, T2j; T23 = VFNMS(LDK(KP250000000), T22, T21); T27 = VFMA(LDK(KP951056516), TU, VADD(T1K, VFNMS(LDK(KP587785252), T26, T23))); T2l = VFMA(LDK(KP587785252), TU, VFMA(LDK(KP951056516), T26, VSUB(T23, T1K))); T2j = VFNMS(LDK(KP250000000), T2i, T2h); T2k = VBYI(VADD(VFMA(LDK(KP951056516), T28, VMUL(LDK(KP587785252), T29)), VADD(T2c, T2j))); T2m = VBYI(VADD(VFNMS(LDK(KP951056516), T29, VMUL(LDK(KP587785252), T28)), VSUB(T2j, T2c))); ST(&(x[WS(rs, 19)]), VSUB(T27, T2k), ms, &(x[WS(rs, 1)])); ST(&(x[WS(rs, 11)]), VADD(T2l, T2m), ms, &(x[WS(rs, 1)])); ST(&(x[WS(rs, 6)]), VADD(T27, T2k), ms, &(x[0])); ST(&(x[WS(rs, 14)]), VSUB(T2l, T2m), ms, &(x[0])); } } } } } VLEAVE(); } static const tw_instr twinstr[] = { VTW(0, 1), VTW(0, 3), VTW(0, 9), VTW(0, 24), {TW_NEXT, VL, 0} }; static const ct_desc desc = { 25, XSIMD_STRING("t3bv_25"), twinstr, &GENUS, {191, 151, 77, 0}, 0, 0, 0 }; void XSIMD(codelet_t3bv_25) (planner *p) { X(kdft_dit_register) (p, t3bv_25, &desc); } #endif /* HAVE_FMA */ fftw-3.3.3/dft/simd/common/t1sv_2.c0000644000175000017500000001017512054410573013677 00000000000000/* * Copyright (c) 2003, 2007-11 Matteo Frigo * Copyright (c) 2003, 2007-11 Massachusetts Institute of Technology * * This program is free software; you can redistribute it and/or modify * it under the terms of the GNU General Public License as published by * the Free Software Foundation; either version 2 of the License, or * (at your option) any later version. * * This program is distributed in the hope that it will be useful, * but WITHOUT ANY WARRANTY; without even the implied warranty of * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the * GNU General Public License for more details. * * You should have received a copy of the GNU General Public License * along with this program; if not, write to the Free Software * Foundation, Inc., 51 Franklin Street, Fifth Floor, Boston, MA 02110-1301 USA * */ /* This file was automatically generated --- DO NOT EDIT */ /* Generated on Sun Nov 25 07:39:23 EST 2012 */ #include "codelet-dft.h" #ifdef HAVE_FMA /* Generated by: ../../../genfft/gen_twiddle.native -fma -reorder-insns -schedule-for-pipeline -simd -compact -variables 4 -pipeline-latency 8 -n 2 -name t1sv_2 -include ts.h */ /* * This function contains 6 FP additions, 4 FP multiplications, * (or, 4 additions, 2 multiplications, 2 fused multiply/add), * 11 stack variables, 0 constants, and 8 memory accesses */ #include "ts.h" static void t1sv_2(R *ri, R *ii, const R *W, stride rs, INT mb, INT me, INT ms) { { INT m; for (m = mb, W = W + (mb * 2); m < me; m = m + (2 * VL), ri = ri + ((2 * VL) * ms), ii = ii + ((2 * VL) * ms), W = W + ((2 * VL) * 2), MAKE_VOLATILE_STRIDE(4, rs)) { V T1, Ta, T3, T6, T2, T5; T1 = LD(&(ri[0]), ms, &(ri[0])); Ta = LD(&(ii[0]), ms, &(ii[0])); T3 = LD(&(ri[WS(rs, 1)]), ms, &(ri[WS(rs, 1)])); T6 = LD(&(ii[WS(rs, 1)]), ms, &(ii[WS(rs, 1)])); T2 = LDW(&(W[0])); T5 = LDW(&(W[TWVL * 1])); { V T8, T4, T9, T7; T8 = VMUL(T2, T6); T4 = VMUL(T2, T3); T9 = VFNMS(T5, T3, T8); T7 = VFMA(T5, T6, T4); ST(&(ii[0]), VADD(T9, Ta), ms, &(ii[0])); ST(&(ii[WS(rs, 1)]), VSUB(Ta, T9), ms, &(ii[WS(rs, 1)])); ST(&(ri[0]), VADD(T1, T7), ms, &(ri[0])); ST(&(ri[WS(rs, 1)]), VSUB(T1, T7), ms, &(ri[WS(rs, 1)])); } } } VLEAVE(); } static const tw_instr twinstr[] = { VTW(0, 1), {TW_NEXT, (2 * VL), 0} }; static const ct_desc desc = { 2, XSIMD_STRING("t1sv_2"), twinstr, &GENUS, {4, 2, 2, 0}, 0, 0, 0 }; void XSIMD(codelet_t1sv_2) (planner *p) { X(kdft_dit_register) (p, t1sv_2, &desc); } #else /* HAVE_FMA */ /* Generated by: ../../../genfft/gen_twiddle.native -simd -compact -variables 4 -pipeline-latency 8 -n 2 -name t1sv_2 -include ts.h */ /* * This function contains 6 FP additions, 4 FP multiplications, * (or, 4 additions, 2 multiplications, 2 fused multiply/add), * 9 stack variables, 0 constants, and 8 memory accesses */ #include "ts.h" static void t1sv_2(R *ri, R *ii, const R *W, stride rs, INT mb, INT me, INT ms) { { INT m; for (m = mb, W = W + (mb * 2); m < me; m = m + (2 * VL), ri = ri + ((2 * VL) * ms), ii = ii + ((2 * VL) * ms), W = W + ((2 * VL) * 2), MAKE_VOLATILE_STRIDE(4, rs)) { V T1, T8, T6, T7; T1 = LD(&(ri[0]), ms, &(ri[0])); T8 = LD(&(ii[0]), ms, &(ii[0])); { V T3, T5, T2, T4; T3 = LD(&(ri[WS(rs, 1)]), ms, &(ri[WS(rs, 1)])); T5 = LD(&(ii[WS(rs, 1)]), ms, &(ii[WS(rs, 1)])); T2 = LDW(&(W[0])); T4 = LDW(&(W[TWVL * 1])); T6 = VFMA(T2, T3, VMUL(T4, T5)); T7 = VFNMS(T4, T3, VMUL(T2, T5)); } ST(&(ri[WS(rs, 1)]), VSUB(T1, T6), ms, &(ri[WS(rs, 1)])); ST(&(ii[WS(rs, 1)]), VSUB(T8, T7), ms, &(ii[WS(rs, 1)])); ST(&(ri[0]), VADD(T1, T6), ms, &(ri[0])); ST(&(ii[0]), VADD(T7, T8), ms, &(ii[0])); } } VLEAVE(); } static const tw_instr twinstr[] = { VTW(0, 1), {TW_NEXT, (2 * VL), 0} }; static const ct_desc desc = { 2, XSIMD_STRING("t1sv_2"), twinstr, &GENUS, {4, 2, 2, 0}, 0, 0, 0 }; void XSIMD(codelet_t1sv_2) (planner *p) { X(kdft_dit_register) (p, t1sv_2, &desc); } #endif /* HAVE_FMA */ fftw-3.3.3/dft/simd/common/t1sv_4.c0000644000175000017500000001470712054410574013707 00000000000000/* * Copyright (c) 2003, 2007-11 Matteo Frigo * Copyright (c) 2003, 2007-11 Massachusetts Institute of Technology * * This program is free software; you can redistribute it and/or modify * it under the terms of the GNU General Public License as published by * the Free Software Foundation; either version 2 of the License, or * (at your option) any later version. * * This program is distributed in the hope that it will be useful, * but WITHOUT ANY WARRANTY; without even the implied warranty of * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the * GNU General Public License for more details. * * You should have received a copy of the GNU General Public License * along with this program; if not, write to the Free Software * Foundation, Inc., 51 Franklin Street, Fifth Floor, Boston, MA 02110-1301 USA * */ /* This file was automatically generated --- DO NOT EDIT */ /* Generated on Sun Nov 25 07:39:24 EST 2012 */ #include "codelet-dft.h" #ifdef HAVE_FMA /* Generated by: ../../../genfft/gen_twiddle.native -fma -reorder-insns -schedule-for-pipeline -simd -compact -variables 4 -pipeline-latency 8 -n 4 -name t1sv_4 -include ts.h */ /* * This function contains 22 FP additions, 12 FP multiplications, * (or, 16 additions, 6 multiplications, 6 fused multiply/add), * 35 stack variables, 0 constants, and 16 memory accesses */ #include "ts.h" static void t1sv_4(R *ri, R *ii, const R *W, stride rs, INT mb, INT me, INT ms) { { INT m; for (m = mb, W = W + (mb * 6); m < me; m = m + (2 * VL), ri = ri + ((2 * VL) * ms), ii = ii + ((2 * VL) * ms), W = W + ((2 * VL) * 6), MAKE_VOLATILE_STRIDE(8, rs)) { V T1, Tv, T3, T6, T5, Ta, Td, Tc, Tg, Tj, Tt, T4, Tf, Ti, Tn; V Tb, T2, T9; T1 = LD(&(ri[0]), ms, &(ri[0])); Tv = LD(&(ii[0]), ms, &(ii[0])); T3 = LD(&(ri[WS(rs, 2)]), ms, &(ri[0])); T6 = LD(&(ii[WS(rs, 2)]), ms, &(ii[0])); T2 = LDW(&(W[TWVL * 2])); T5 = LDW(&(W[TWVL * 3])); Ta = LD(&(ri[WS(rs, 1)]), ms, &(ri[WS(rs, 1)])); Td = LD(&(ii[WS(rs, 1)]), ms, &(ii[WS(rs, 1)])); T9 = LDW(&(W[0])); Tc = LDW(&(W[TWVL * 1])); Tg = LD(&(ri[WS(rs, 3)]), ms, &(ri[WS(rs, 1)])); Tj = LD(&(ii[WS(rs, 3)]), ms, &(ii[WS(rs, 1)])); Tt = VMUL(T2, T6); T4 = VMUL(T2, T3); Tf = LDW(&(W[TWVL * 4])); Ti = LDW(&(W[TWVL * 5])); Tn = VMUL(T9, Td); Tb = VMUL(T9, Ta); { V Tu, T7, Tp, Th, To, Te; Tu = VFNMS(T5, T3, Tt); T7 = VFMA(T5, T6, T4); Tp = VMUL(Tf, Tj); Th = VMUL(Tf, Tg); To = VFNMS(Tc, Ta, Tn); Te = VFMA(Tc, Td, Tb); { V Tw, Tx, T8, Tm, Tq, Tk; Tw = VADD(Tu, Tv); Tx = VSUB(Tv, Tu); T8 = VADD(T1, T7); Tm = VSUB(T1, T7); Tq = VFNMS(Ti, Tg, Tp); Tk = VFMA(Ti, Tj, Th); { V Ts, Tr, Tl, Ty; Ts = VADD(To, Tq); Tr = VSUB(To, Tq); Tl = VADD(Te, Tk); Ty = VSUB(Te, Tk); ST(&(ri[WS(rs, 1)]), VADD(Tm, Tr), ms, &(ri[WS(rs, 1)])); ST(&(ri[WS(rs, 3)]), VSUB(Tm, Tr), ms, &(ri[WS(rs, 1)])); ST(&(ii[WS(rs, 2)]), VSUB(Tw, Ts), ms, &(ii[0])); ST(&(ii[0]), VADD(Ts, Tw), ms, &(ii[0])); ST(&(ii[WS(rs, 3)]), VADD(Ty, Tx), ms, &(ii[WS(rs, 1)])); ST(&(ii[WS(rs, 1)]), VSUB(Tx, Ty), ms, &(ii[WS(rs, 1)])); ST(&(ri[0]), VADD(T8, Tl), ms, &(ri[0])); ST(&(ri[WS(rs, 2)]), VSUB(T8, Tl), ms, &(ri[0])); } } } } } VLEAVE(); } static const tw_instr twinstr[] = { VTW(0, 1), VTW(0, 2), VTW(0, 3), {TW_NEXT, (2 * VL), 0} }; static const ct_desc desc = { 4, XSIMD_STRING("t1sv_4"), twinstr, &GENUS, {16, 6, 6, 0}, 0, 0, 0 }; void XSIMD(codelet_t1sv_4) (planner *p) { X(kdft_dit_register) (p, t1sv_4, &desc); } #else /* HAVE_FMA */ /* Generated by: ../../../genfft/gen_twiddle.native -simd -compact -variables 4 -pipeline-latency 8 -n 4 -name t1sv_4 -include ts.h */ /* * This function contains 22 FP additions, 12 FP multiplications, * (or, 16 additions, 6 multiplications, 6 fused multiply/add), * 13 stack variables, 0 constants, and 16 memory accesses */ #include "ts.h" static void t1sv_4(R *ri, R *ii, const R *W, stride rs, INT mb, INT me, INT ms) { { INT m; for (m = mb, W = W + (mb * 6); m < me; m = m + (2 * VL), ri = ri + ((2 * VL) * ms), ii = ii + ((2 * VL) * ms), W = W + ((2 * VL) * 6), MAKE_VOLATILE_STRIDE(8, rs)) { V T1, Tp, T6, To, Tc, Tk, Th, Tl; T1 = LD(&(ri[0]), ms, &(ri[0])); Tp = LD(&(ii[0]), ms, &(ii[0])); { V T3, T5, T2, T4; T3 = LD(&(ri[WS(rs, 2)]), ms, &(ri[0])); T5 = LD(&(ii[WS(rs, 2)]), ms, &(ii[0])); T2 = LDW(&(W[TWVL * 2])); T4 = LDW(&(W[TWVL * 3])); T6 = VFMA(T2, T3, VMUL(T4, T5)); To = VFNMS(T4, T3, VMUL(T2, T5)); } { V T9, Tb, T8, Ta; T9 = LD(&(ri[WS(rs, 1)]), ms, &(ri[WS(rs, 1)])); Tb = LD(&(ii[WS(rs, 1)]), ms, &(ii[WS(rs, 1)])); T8 = LDW(&(W[0])); Ta = LDW(&(W[TWVL * 1])); Tc = VFMA(T8, T9, VMUL(Ta, Tb)); Tk = VFNMS(Ta, T9, VMUL(T8, Tb)); } { V Te, Tg, Td, Tf; Te = LD(&(ri[WS(rs, 3)]), ms, &(ri[WS(rs, 1)])); Tg = LD(&(ii[WS(rs, 3)]), ms, &(ii[WS(rs, 1)])); Td = LDW(&(W[TWVL * 4])); Tf = LDW(&(W[TWVL * 5])); Th = VFMA(Td, Te, VMUL(Tf, Tg)); Tl = VFNMS(Tf, Te, VMUL(Td, Tg)); } { V T7, Ti, Tn, Tq; T7 = VADD(T1, T6); Ti = VADD(Tc, Th); ST(&(ri[WS(rs, 2)]), VSUB(T7, Ti), ms, &(ri[0])); ST(&(ri[0]), VADD(T7, Ti), ms, &(ri[0])); Tn = VADD(Tk, Tl); Tq = VADD(To, Tp); ST(&(ii[0]), VADD(Tn, Tq), ms, &(ii[0])); ST(&(ii[WS(rs, 2)]), VSUB(Tq, Tn), ms, &(ii[0])); } { V Tj, Tm, Tr, Ts; Tj = VSUB(T1, T6); Tm = VSUB(Tk, Tl); ST(&(ri[WS(rs, 3)]), VSUB(Tj, Tm), ms, &(ri[WS(rs, 1)])); ST(&(ri[WS(rs, 1)]), VADD(Tj, Tm), ms, &(ri[WS(rs, 1)])); Tr = VSUB(Tp, To); Ts = VSUB(Tc, Th); ST(&(ii[WS(rs, 1)]), VSUB(Tr, Ts), ms, &(ii[WS(rs, 1)])); ST(&(ii[WS(rs, 3)]), VADD(Ts, Tr), ms, &(ii[WS(rs, 1)])); } } } VLEAVE(); } static const tw_instr twinstr[] = { VTW(0, 1), VTW(0, 2), VTW(0, 3), {TW_NEXT, (2 * VL), 0} }; static const ct_desc desc = { 4, XSIMD_STRING("t1sv_4"), twinstr, &GENUS, {16, 6, 6, 0}, 0, 0, 0 }; void XSIMD(codelet_t1sv_4) (planner *p) { X(kdft_dit_register) (p, t1sv_4, &desc); } #endif /* HAVE_FMA */ fftw-3.3.3/dft/simd/common/t1sv_8.c0000644000175000017500000003020712054410574013704 00000000000000/* * Copyright (c) 2003, 2007-11 Matteo Frigo * Copyright (c) 2003, 2007-11 Massachusetts Institute of Technology * * This program is free software; you can redistribute it and/or modify * it under the terms of the GNU General Public License as published by * the Free Software Foundation; either version 2 of the License, or * (at your option) any later version. * * This program is distributed in the hope that it will be useful, * but WITHOUT ANY WARRANTY; without even the implied warranty of * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the * GNU General Public License for more details. * * You should have received a copy of the GNU General Public License * along with this program; if not, write to the Free Software * Foundation, Inc., 51 Franklin Street, Fifth Floor, Boston, MA 02110-1301 USA * */ /* This file was automatically generated --- DO NOT EDIT */ /* Generated on Sun Nov 25 07:39:24 EST 2012 */ #include "codelet-dft.h" #ifdef HAVE_FMA /* Generated by: ../../../genfft/gen_twiddle.native -fma -reorder-insns -schedule-for-pipeline -simd -compact -variables 4 -pipeline-latency 8 -n 8 -name t1sv_8 -include ts.h */ /* * This function contains 66 FP additions, 36 FP multiplications, * (or, 44 additions, 14 multiplications, 22 fused multiply/add), * 59 stack variables, 1 constants, and 32 memory accesses */ #include "ts.h" static void t1sv_8(R *ri, R *ii, const R *W, stride rs, INT mb, INT me, INT ms) { DVK(KP707106781, +0.707106781186547524400844362104849039284835938); { INT m; for (m = mb, W = W + (mb * 14); m < me; m = m + (2 * VL), ri = ri + ((2 * VL) * ms), ii = ii + ((2 * VL) * ms), W = W + ((2 * VL) * 14), MAKE_VOLATILE_STRIDE(16, rs)) { V T1, T1m, T1l, T7, TS, Tk, TQ, Te, To, Tr, Tu, T14, TF, Tx, T16; V TL, Tt, TW, Tp, Tq, Tw; { V T3, T6, T2, T5; T1 = LD(&(ri[0]), ms, &(ri[0])); T1m = LD(&(ii[0]), ms, &(ii[0])); T3 = LD(&(ri[WS(rs, 4)]), ms, &(ri[0])); T6 = LD(&(ii[WS(rs, 4)]), ms, &(ii[0])); T2 = LDW(&(W[TWVL * 6])); T5 = LDW(&(W[TWVL * 7])); { V Tg, Tj, Ti, Ta, Td, T1k, T4, T9, Tc, TR, Th, Tf; Tg = LD(&(ri[WS(rs, 6)]), ms, &(ri[0])); Tj = LD(&(ii[WS(rs, 6)]), ms, &(ii[0])); Tf = LDW(&(W[TWVL * 10])); Ti = LDW(&(W[TWVL * 11])); Ta = LD(&(ri[WS(rs, 2)]), ms, &(ri[0])); Td = LD(&(ii[WS(rs, 2)]), ms, &(ii[0])); T1k = VMUL(T2, T6); T4 = VMUL(T2, T3); T9 = LDW(&(W[TWVL * 2])); Tc = LDW(&(W[TWVL * 3])); TR = VMUL(Tf, Tj); Th = VMUL(Tf, Tg); { V TB, TE, TH, TK, TG, TD, TJ, T13, TC, TA, TP, Tb, T15, TI, Tn; TB = LD(&(ri[WS(rs, 7)]), ms, &(ri[WS(rs, 1)])); TE = LD(&(ii[WS(rs, 7)]), ms, &(ii[WS(rs, 1)])); T1l = VFNMS(T5, T3, T1k); T7 = VFMA(T5, T6, T4); TP = VMUL(T9, Td); Tb = VMUL(T9, Ta); TS = VFNMS(Ti, Tg, TR); Tk = VFMA(Ti, Tj, Th); TA = LDW(&(W[TWVL * 12])); TH = LD(&(ri[WS(rs, 3)]), ms, &(ri[WS(rs, 1)])); TK = LD(&(ii[WS(rs, 3)]), ms, &(ii[WS(rs, 1)])); TG = LDW(&(W[TWVL * 4])); TQ = VFNMS(Tc, Ta, TP); Te = VFMA(Tc, Td, Tb); TD = LDW(&(W[TWVL * 13])); TJ = LDW(&(W[TWVL * 5])); T13 = VMUL(TA, TE); TC = VMUL(TA, TB); To = LD(&(ri[WS(rs, 1)]), ms, &(ri[WS(rs, 1)])); T15 = VMUL(TG, TK); TI = VMUL(TG, TH); Tr = LD(&(ii[WS(rs, 1)]), ms, &(ii[WS(rs, 1)])); Tn = LDW(&(W[0])); Tu = LD(&(ri[WS(rs, 5)]), ms, &(ri[WS(rs, 1)])); T14 = VFNMS(TD, TB, T13); TF = VFMA(TD, TE, TC); Tx = LD(&(ii[WS(rs, 5)]), ms, &(ii[WS(rs, 1)])); T16 = VFNMS(TJ, TH, T15); TL = VFMA(TJ, TK, TI); Tt = LDW(&(W[TWVL * 8])); TW = VMUL(Tn, Tr); Tp = VMUL(Tn, To); Tq = LDW(&(W[TWVL * 1])); Tw = LDW(&(W[TWVL * 9])); } } } { V T8, T1g, TM, T1j, TX, Ts, T1n, T1r, T1s, Tl, T1c, T18, TZ, Ty, T1a; V TU; { V TO, T17, T12, TY, Tv, TT; T8 = VADD(T1, T7); TO = VSUB(T1, T7); T17 = VSUB(T14, T16); T1g = VADD(T14, T16); TM = VADD(TF, TL); T12 = VSUB(TF, TL); TY = VMUL(Tt, Tx); Tv = VMUL(Tt, Tu); TT = VSUB(TQ, TS); T1j = VADD(TQ, TS); TX = VFNMS(Tq, To, TW); Ts = VFMA(Tq, Tr, Tp); T1n = VADD(T1l, T1m); T1r = VSUB(T1m, T1l); T1s = VSUB(Te, Tk); Tl = VADD(Te, Tk); T1c = VADD(T12, T17); T18 = VSUB(T12, T17); TZ = VFNMS(Tw, Tu, TY); Ty = VFMA(Tw, Tx, Tv); T1a = VSUB(TO, TT); TU = VADD(TO, TT); } { V T1v, T1t, Tm, T1e, T1o, T1q, TN, T1p, T1d, T1u, T19, T1w, T1i, T1h; { V T10, T1f, Tz, TV, T11, T1b; T1v = VADD(T1s, T1r); T1t = VSUB(T1r, T1s); T10 = VSUB(TX, TZ); T1f = VADD(TX, TZ); Tz = VADD(Ts, Ty); TV = VSUB(Ts, Ty); T11 = VADD(TV, T10); T1b = VSUB(T10, TV); Tm = VADD(T8, Tl); T1e = VSUB(T8, Tl); T1o = VADD(T1j, T1n); T1q = VSUB(T1n, T1j); TN = VADD(Tz, TM); T1p = VSUB(TM, Tz); T1d = VSUB(T1b, T1c); T1u = VADD(T1b, T1c); T19 = VADD(T11, T18); T1w = VSUB(T18, T11); T1i = VADD(T1f, T1g); T1h = VSUB(T1f, T1g); } ST(&(ii[WS(rs, 6)]), VSUB(T1q, T1p), ms, &(ii[0])); ST(&(ri[0]), VADD(Tm, TN), ms, &(ri[0])); ST(&(ri[WS(rs, 4)]), VSUB(Tm, TN), ms, &(ri[0])); ST(&(ii[WS(rs, 1)]), VFMA(LDK(KP707106781), T1u, T1t), ms, &(ii[WS(rs, 1)])); ST(&(ii[WS(rs, 5)]), VFNMS(LDK(KP707106781), T1u, T1t), ms, &(ii[WS(rs, 1)])); ST(&(ri[WS(rs, 3)]), VFMA(LDK(KP707106781), T1d, T1a), ms, &(ri[WS(rs, 1)])); ST(&(ri[WS(rs, 7)]), VFNMS(LDK(KP707106781), T1d, T1a), ms, &(ri[WS(rs, 1)])); ST(&(ii[WS(rs, 3)]), VFMA(LDK(KP707106781), T1w, T1v), ms, &(ii[WS(rs, 1)])); ST(&(ii[WS(rs, 7)]), VFNMS(LDK(KP707106781), T1w, T1v), ms, &(ii[WS(rs, 1)])); ST(&(ri[WS(rs, 1)]), VFMA(LDK(KP707106781), T19, TU), ms, &(ri[WS(rs, 1)])); ST(&(ri[WS(rs, 5)]), VFNMS(LDK(KP707106781), T19, TU), ms, &(ri[WS(rs, 1)])); ST(&(ri[WS(rs, 6)]), VSUB(T1e, T1h), ms, &(ri[0])); ST(&(ii[0]), VADD(T1i, T1o), ms, &(ii[0])); ST(&(ii[WS(rs, 4)]), VSUB(T1o, T1i), ms, &(ii[0])); ST(&(ri[WS(rs, 2)]), VADD(T1e, T1h), ms, &(ri[0])); ST(&(ii[WS(rs, 2)]), VADD(T1p, T1q), ms, &(ii[0])); } } } } VLEAVE(); } static const tw_instr twinstr[] = { VTW(0, 1), VTW(0, 2), VTW(0, 3), VTW(0, 4), VTW(0, 5), VTW(0, 6), VTW(0, 7), {TW_NEXT, (2 * VL), 0} }; static const ct_desc desc = { 8, XSIMD_STRING("t1sv_8"), twinstr, &GENUS, {44, 14, 22, 0}, 0, 0, 0 }; void XSIMD(codelet_t1sv_8) (planner *p) { X(kdft_dit_register) (p, t1sv_8, &desc); } #else /* HAVE_FMA */ /* Generated by: ../../../genfft/gen_twiddle.native -simd -compact -variables 4 -pipeline-latency 8 -n 8 -name t1sv_8 -include ts.h */ /* * This function contains 66 FP additions, 32 FP multiplications, * (or, 52 additions, 18 multiplications, 14 fused multiply/add), * 28 stack variables, 1 constants, and 32 memory accesses */ #include "ts.h" static void t1sv_8(R *ri, R *ii, const R *W, stride rs, INT mb, INT me, INT ms) { DVK(KP707106781, +0.707106781186547524400844362104849039284835938); { INT m; for (m = mb, W = W + (mb * 14); m < me; m = m + (2 * VL), ri = ri + ((2 * VL) * ms), ii = ii + ((2 * VL) * ms), W = W + ((2 * VL) * 14), MAKE_VOLATILE_STRIDE(16, rs)) { V T7, T1e, TH, T19, TF, T13, TR, TU, Ti, T1f, TK, T16, Tu, T12, TM; V TP; { V T1, T18, T6, T17; T1 = LD(&(ri[0]), ms, &(ri[0])); T18 = LD(&(ii[0]), ms, &(ii[0])); { V T3, T5, T2, T4; T3 = LD(&(ri[WS(rs, 4)]), ms, &(ri[0])); T5 = LD(&(ii[WS(rs, 4)]), ms, &(ii[0])); T2 = LDW(&(W[TWVL * 6])); T4 = LDW(&(W[TWVL * 7])); T6 = VFMA(T2, T3, VMUL(T4, T5)); T17 = VFNMS(T4, T3, VMUL(T2, T5)); } T7 = VADD(T1, T6); T1e = VSUB(T18, T17); TH = VSUB(T1, T6); T19 = VADD(T17, T18); } { V Tz, TS, TE, TT; { V Tw, Ty, Tv, Tx; Tw = LD(&(ri[WS(rs, 7)]), ms, &(ri[WS(rs, 1)])); Ty = LD(&(ii[WS(rs, 7)]), ms, &(ii[WS(rs, 1)])); Tv = LDW(&(W[TWVL * 12])); Tx = LDW(&(W[TWVL * 13])); Tz = VFMA(Tv, Tw, VMUL(Tx, Ty)); TS = VFNMS(Tx, Tw, VMUL(Tv, Ty)); } { V TB, TD, TA, TC; TB = LD(&(ri[WS(rs, 3)]), ms, &(ri[WS(rs, 1)])); TD = LD(&(ii[WS(rs, 3)]), ms, &(ii[WS(rs, 1)])); TA = LDW(&(W[TWVL * 4])); TC = LDW(&(W[TWVL * 5])); TE = VFMA(TA, TB, VMUL(TC, TD)); TT = VFNMS(TC, TB, VMUL(TA, TD)); } TF = VADD(Tz, TE); T13 = VADD(TS, TT); TR = VSUB(Tz, TE); TU = VSUB(TS, TT); } { V Tc, TI, Th, TJ; { V T9, Tb, T8, Ta; T9 = LD(&(ri[WS(rs, 2)]), ms, &(ri[0])); Tb = LD(&(ii[WS(rs, 2)]), ms, &(ii[0])); T8 = LDW(&(W[TWVL * 2])); Ta = LDW(&(W[TWVL * 3])); Tc = VFMA(T8, T9, VMUL(Ta, Tb)); TI = VFNMS(Ta, T9, VMUL(T8, Tb)); } { V Te, Tg, Td, Tf; Te = LD(&(ri[WS(rs, 6)]), ms, &(ri[0])); Tg = LD(&(ii[WS(rs, 6)]), ms, &(ii[0])); Td = LDW(&(W[TWVL * 10])); Tf = LDW(&(W[TWVL * 11])); Th = VFMA(Td, Te, VMUL(Tf, Tg)); TJ = VFNMS(Tf, Te, VMUL(Td, Tg)); } Ti = VADD(Tc, Th); T1f = VSUB(Tc, Th); TK = VSUB(TI, TJ); T16 = VADD(TI, TJ); } { V To, TN, Tt, TO; { V Tl, Tn, Tk, Tm; Tl = LD(&(ri[WS(rs, 1)]), ms, &(ri[WS(rs, 1)])); Tn = LD(&(ii[WS(rs, 1)]), ms, &(ii[WS(rs, 1)])); Tk = LDW(&(W[0])); Tm = LDW(&(W[TWVL * 1])); To = VFMA(Tk, Tl, VMUL(Tm, Tn)); TN = VFNMS(Tm, Tl, VMUL(Tk, Tn)); } { V Tq, Ts, Tp, Tr; Tq = LD(&(ri[WS(rs, 5)]), ms, &(ri[WS(rs, 1)])); Ts = LD(&(ii[WS(rs, 5)]), ms, &(ii[WS(rs, 1)])); Tp = LDW(&(W[TWVL * 8])); Tr = LDW(&(W[TWVL * 9])); Tt = VFMA(Tp, Tq, VMUL(Tr, Ts)); TO = VFNMS(Tr, Tq, VMUL(Tp, Ts)); } Tu = VADD(To, Tt); T12 = VADD(TN, TO); TM = VSUB(To, Tt); TP = VSUB(TN, TO); } { V Tj, TG, T1b, T1c; Tj = VADD(T7, Ti); TG = VADD(Tu, TF); ST(&(ri[WS(rs, 4)]), VSUB(Tj, TG), ms, &(ri[0])); ST(&(ri[0]), VADD(Tj, TG), ms, &(ri[0])); { V T15, T1a, T11, T14; T15 = VADD(T12, T13); T1a = VADD(T16, T19); ST(&(ii[0]), VADD(T15, T1a), ms, &(ii[0])); ST(&(ii[WS(rs, 4)]), VSUB(T1a, T15), ms, &(ii[0])); T11 = VSUB(T7, Ti); T14 = VSUB(T12, T13); ST(&(ri[WS(rs, 6)]), VSUB(T11, T14), ms, &(ri[0])); ST(&(ri[WS(rs, 2)]), VADD(T11, T14), ms, &(ri[0])); } T1b = VSUB(TF, Tu); T1c = VSUB(T19, T16); ST(&(ii[WS(rs, 2)]), VADD(T1b, T1c), ms, &(ii[0])); ST(&(ii[WS(rs, 6)]), VSUB(T1c, T1b), ms, &(ii[0])); { V TX, T1g, T10, T1d, TY, TZ; TX = VSUB(TH, TK); T1g = VSUB(T1e, T1f); TY = VSUB(TP, TM); TZ = VADD(TR, TU); T10 = VMUL(LDK(KP707106781), VSUB(TY, TZ)); T1d = VMUL(LDK(KP707106781), VADD(TY, TZ)); ST(&(ri[WS(rs, 7)]), VSUB(TX, T10), ms, &(ri[WS(rs, 1)])); ST(&(ii[WS(rs, 5)]), VSUB(T1g, T1d), ms, &(ii[WS(rs, 1)])); ST(&(ri[WS(rs, 3)]), VADD(TX, T10), ms, &(ri[WS(rs, 1)])); ST(&(ii[WS(rs, 1)]), VADD(T1d, T1g), ms, &(ii[WS(rs, 1)])); } { V TL, T1i, TW, T1h, TQ, TV; TL = VADD(TH, TK); T1i = VADD(T1f, T1e); TQ = VADD(TM, TP); TV = VSUB(TR, TU); TW = VMUL(LDK(KP707106781), VADD(TQ, TV)); T1h = VMUL(LDK(KP707106781), VSUB(TV, TQ)); ST(&(ri[WS(rs, 5)]), VSUB(TL, TW), ms, &(ri[WS(rs, 1)])); ST(&(ii[WS(rs, 7)]), VSUB(T1i, T1h), ms, &(ii[WS(rs, 1)])); ST(&(ri[WS(rs, 1)]), VADD(TL, TW), ms, &(ri[WS(rs, 1)])); ST(&(ii[WS(rs, 3)]), VADD(T1h, T1i), ms, &(ii[WS(rs, 1)])); } } } } VLEAVE(); } static const tw_instr twinstr[] = { VTW(0, 1), VTW(0, 2), VTW(0, 3), VTW(0, 4), VTW(0, 5), VTW(0, 6), VTW(0, 7), {TW_NEXT, (2 * VL), 0} }; static const ct_desc desc = { 8, XSIMD_STRING("t1sv_8"), twinstr, &GENUS, {52, 18, 14, 0}, 0, 0, 0 }; void XSIMD(codelet_t1sv_8) (planner *p) { X(kdft_dit_register) (p, t1sv_8, &desc); } #endif /* HAVE_FMA */ fftw-3.3.3/dft/simd/common/t1sv_16.c0000644000175000017500000006464612054410575014002 00000000000000/* * Copyright (c) 2003, 2007-11 Matteo Frigo * Copyright (c) 2003, 2007-11 Massachusetts Institute of Technology * * This program is free software; you can redistribute it and/or modify * it under the terms of the GNU General Public License as published by * the Free Software Foundation; either version 2 of the License, or * (at your option) any later version. * * This program is distributed in the hope that it will be useful, * but WITHOUT ANY WARRANTY; without even the implied warranty of * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the * GNU General Public License for more details. * * You should have received a copy of the GNU General Public License * along with this program; if not, write to the Free Software * Foundation, Inc., 51 Franklin Street, Fifth Floor, Boston, MA 02110-1301 USA * */ /* This file was automatically generated --- DO NOT EDIT */ /* Generated on Sun Nov 25 07:39:24 EST 2012 */ #include "codelet-dft.h" #ifdef HAVE_FMA /* Generated by: ../../../genfft/gen_twiddle.native -fma -reorder-insns -schedule-for-pipeline -simd -compact -variables 4 -pipeline-latency 8 -n 16 -name t1sv_16 -include ts.h */ /* * This function contains 174 FP additions, 100 FP multiplications, * (or, 104 additions, 30 multiplications, 70 fused multiply/add), * 113 stack variables, 3 constants, and 64 memory accesses */ #include "ts.h" static void t1sv_16(R *ri, R *ii, const R *W, stride rs, INT mb, INT me, INT ms) { DVK(KP923879532, +0.923879532511286756128183189396788286822416626); DVK(KP414213562, +0.414213562373095048801688724209698078569671875); DVK(KP707106781, +0.707106781186547524400844362104849039284835938); { INT m; for (m = mb, W = W + (mb * 30); m < me; m = m + (2 * VL), ri = ri + ((2 * VL) * ms), ii = ii + ((2 * VL) * ms), W = W + ((2 * VL) * 30), MAKE_VOLATILE_STRIDE(32, rs)) { V T2S, T2O, T2B, T2j, T2A, T24, T3J, T3L, T2Q, T2I, T2R, T2L, T2C, T2y, T3D; V T3F; { V T3o, T3z, T1I, T8, T35, T2o, T1s, T2r, T36, T2w, T1F, T2p, T1N, T3k, Tl; V T3A, T2V, T1T, Tz, T1U, T30, T29, T11, T2c, TH, TK, TJ, T31, T2h, T1e; V T2a, T1Z, TI, T1Y, TF; { V Ta, Td, Tg, Tj, T2t, T1y, Tf, T1J, Tb, Tc, T2v, T1E, Ti; { V T1, T3n, T3, T6, T5, T1h, T1k, T1n, T1q, T1m, T3l, T4, T1j, T1p, T2k; V T1i, T2, T1g; T1 = LD(&(ri[0]), ms, &(ri[0])); T3n = LD(&(ii[0]), ms, &(ii[0])); T3 = LD(&(ri[WS(rs, 8)]), ms, &(ri[0])); T6 = LD(&(ii[WS(rs, 8)]), ms, &(ii[0])); T2 = LDW(&(W[TWVL * 14])); T5 = LDW(&(W[TWVL * 15])); T1h = LD(&(ri[WS(rs, 15)]), ms, &(ri[WS(rs, 1)])); T1k = LD(&(ii[WS(rs, 15)]), ms, &(ii[WS(rs, 1)])); T1g = LDW(&(W[TWVL * 28])); T1n = LD(&(ri[WS(rs, 7)]), ms, &(ri[WS(rs, 1)])); T1q = LD(&(ii[WS(rs, 7)]), ms, &(ii[WS(rs, 1)])); T1m = LDW(&(W[TWVL * 12])); T3l = VMUL(T2, T6); T4 = VMUL(T2, T3); T1j = LDW(&(W[TWVL * 29])); T1p = LDW(&(W[TWVL * 13])); T2k = VMUL(T1g, T1k); T1i = VMUL(T1g, T1h); { V T1u, T1x, T1A, T2s, T1v, T1D, T1z, T1w, T1C, T2u, T1B, T9; { V T2l, T1l, T1t, T2n, T1r; { V T2m, T1o, T3m, T7; T1u = LD(&(ri[WS(rs, 3)]), ms, &(ri[WS(rs, 1)])); T2m = VMUL(T1m, T1q); T1o = VMUL(T1m, T1n); T3m = VFNMS(T5, T3, T3l); T7 = VFMA(T5, T6, T4); T1x = LD(&(ii[WS(rs, 3)]), ms, &(ii[WS(rs, 1)])); T2l = VFNMS(T1j, T1h, T2k); T1l = VFMA(T1j, T1k, T1i); T1t = LDW(&(W[TWVL * 4])); T2n = VFNMS(T1p, T1n, T2m); T1r = VFMA(T1p, T1q, T1o); T3o = VADD(T3m, T3n); T3z = VSUB(T3n, T3m); T1I = VSUB(T1, T7); T8 = VADD(T1, T7); } T1A = LD(&(ri[WS(rs, 11)]), ms, &(ri[WS(rs, 1)])); T2s = VMUL(T1t, T1x); T1v = VMUL(T1t, T1u); T35 = VADD(T2l, T2n); T2o = VSUB(T2l, T2n); T1s = VADD(T1l, T1r); T2r = VSUB(T1l, T1r); T1D = LD(&(ii[WS(rs, 11)]), ms, &(ii[WS(rs, 1)])); T1z = LDW(&(W[TWVL * 20])); } T1w = LDW(&(W[TWVL * 5])); T1C = LDW(&(W[TWVL * 21])); Ta = LD(&(ri[WS(rs, 4)]), ms, &(ri[0])); Td = LD(&(ii[WS(rs, 4)]), ms, &(ii[0])); T9 = LDW(&(W[TWVL * 6])); Tg = LD(&(ri[WS(rs, 12)]), ms, &(ri[0])); Tj = LD(&(ii[WS(rs, 12)]), ms, &(ii[0])); T2u = VMUL(T1z, T1D); T1B = VMUL(T1z, T1A); T2t = VFNMS(T1w, T1u, T2s); T1y = VFMA(T1w, T1x, T1v); Tf = LDW(&(W[TWVL * 22])); T1J = VMUL(T9, Td); Tb = VMUL(T9, Ta); Tc = LDW(&(W[TWVL * 7])); T2v = VFNMS(T1C, T1A, T2u); T1E = VFMA(T1C, T1D, T1B); Ti = LDW(&(W[TWVL * 23])); } } { V TW, TZ, TY, T27, TX, T26, TU; { V To, Tr, Tu, Tx, Tq, Tw, T1P, Tp, T1R, Tv; { V T1K, Te, T1M, Tk, Tn, Tt, T1L, Th; To = LD(&(ri[WS(rs, 2)]), ms, &(ri[0])); T1L = VMUL(Tf, Tj); Th = VMUL(Tf, Tg); Tr = LD(&(ii[WS(rs, 2)]), ms, &(ii[0])); T1K = VFNMS(Tc, Ta, T1J); Te = VFMA(Tc, Td, Tb); T36 = VADD(T2t, T2v); T2w = VSUB(T2t, T2v); T1F = VADD(T1y, T1E); T2p = VSUB(T1y, T1E); T1M = VFNMS(Ti, Tg, T1L); Tk = VFMA(Ti, Tj, Th); Tn = LDW(&(W[TWVL * 2])); Tu = LD(&(ri[WS(rs, 10)]), ms, &(ri[0])); Tx = LD(&(ii[WS(rs, 10)]), ms, &(ii[0])); Tt = LDW(&(W[TWVL * 18])); Tq = LDW(&(W[TWVL * 3])); Tw = LDW(&(W[TWVL * 19])); T1N = VSUB(T1K, T1M); T3k = VADD(T1K, T1M); Tl = VADD(Te, Tk); T3A = VSUB(Te, Tk); T1P = VMUL(Tn, Tr); Tp = VMUL(Tn, To); T1R = VMUL(Tt, Tx); Tv = VMUL(Tt, Tu); } { V TQ, TT, T1Q, Ts, T1S, Ty, TV, T25, TR, TP, TS; TQ = LD(&(ri[WS(rs, 1)]), ms, &(ri[WS(rs, 1)])); TT = LD(&(ii[WS(rs, 1)]), ms, &(ii[WS(rs, 1)])); TP = LDW(&(W[0])); TW = LD(&(ri[WS(rs, 9)]), ms, &(ri[WS(rs, 1)])); T1Q = VFNMS(Tq, To, T1P); Ts = VFMA(Tq, Tr, Tp); T1S = VFNMS(Tw, Tu, T1R); Ty = VFMA(Tw, Tx, Tv); TZ = LD(&(ii[WS(rs, 9)]), ms, &(ii[WS(rs, 1)])); TV = LDW(&(W[TWVL * 16])); T25 = VMUL(TP, TT); TR = VMUL(TP, TQ); TS = LDW(&(W[TWVL * 1])); TY = LDW(&(W[TWVL * 17])); T2V = VADD(T1Q, T1S); T1T = VSUB(T1Q, T1S); Tz = VADD(Ts, Ty); T1U = VSUB(Ts, Ty); T27 = VMUL(TV, TZ); TX = VMUL(TV, TW); T26 = VFNMS(TS, TQ, T25); TU = VFMA(TS, TT, TR); } } { V T19, T1c, T1b, T2f, T1a, T2e, T17; { V T13, T16, T12, T28, T10, T18, T15, T2d, T14; T13 = LD(&(ri[WS(rs, 5)]), ms, &(ri[WS(rs, 1)])); T16 = LD(&(ii[WS(rs, 5)]), ms, &(ii[WS(rs, 1)])); T12 = LDW(&(W[TWVL * 8])); T19 = LD(&(ri[WS(rs, 13)]), ms, &(ri[WS(rs, 1)])); T28 = VFNMS(TY, TW, T27); T10 = VFMA(TY, TZ, TX); T1c = LD(&(ii[WS(rs, 13)]), ms, &(ii[WS(rs, 1)])); T18 = LDW(&(W[TWVL * 24])); T15 = LDW(&(W[TWVL * 9])); T1b = LDW(&(W[TWVL * 25])); T2d = VMUL(T12, T16); T14 = VMUL(T12, T13); T30 = VADD(T26, T28); T29 = VSUB(T26, T28); T11 = VADD(TU, T10); T2c = VSUB(TU, T10); T2f = VMUL(T18, T1c); T1a = VMUL(T18, T19); T2e = VFNMS(T15, T13, T2d); T17 = VFMA(T15, T16, T14); } { V TB, TE, TA, T2g, T1d, TG, TD, T1X, TC; TB = LD(&(ri[WS(rs, 14)]), ms, &(ri[0])); TE = LD(&(ii[WS(rs, 14)]), ms, &(ii[0])); TA = LDW(&(W[TWVL * 26])); TH = LD(&(ri[WS(rs, 6)]), ms, &(ri[0])); T2g = VFNMS(T1b, T19, T2f); T1d = VFMA(T1b, T1c, T1a); TK = LD(&(ii[WS(rs, 6)]), ms, &(ii[0])); TG = LDW(&(W[TWVL * 10])); TD = LDW(&(W[TWVL * 27])); TJ = LDW(&(W[TWVL * 11])); T1X = VMUL(TA, TE); TC = VMUL(TA, TB); T31 = VADD(T2e, T2g); T2h = VSUB(T2e, T2g); T1e = VADD(T17, T1d); T2a = VSUB(T17, T1d); T1Z = VMUL(TG, TK); TI = VMUL(TG, TH); T1Y = VFNMS(TD, TB, T1X); TF = VFMA(TD, TE, TC); } } } } { V T2U, Tm, T3p, T3u, T34, T1G, T1f, T2Z, T20, TL, T32, T3f, T3g, T37; T2U = VSUB(T8, Tl); Tm = VADD(T8, Tl); T3p = VADD(T3k, T3o); T3u = VSUB(T3o, T3k); T34 = VSUB(T1s, T1F); T1G = VADD(T1s, T1F); T1f = VADD(T11, T1e); T2Z = VSUB(T11, T1e); T20 = VFNMS(TJ, TH, T1Z); TL = VFMA(TJ, TK, TI); T32 = VSUB(T30, T31); T3f = VADD(T30, T31); T3g = VADD(T35, T36); T37 = VSUB(T35, T36); { V T3r, T1H, T21, T1W, T3i, T3h, T3j, T2X, TN, T3t, T2W, TM; T3r = VSUB(T1G, T1f); T1H = VADD(T1f, T1G); T21 = VSUB(T1Y, T20); T2W = VADD(T1Y, T20); T1W = VSUB(TF, TL); TM = VADD(TF, TL); T3i = VADD(T3f, T3g); T3h = VSUB(T3f, T3g); T3j = VADD(T2V, T2W); T2X = VSUB(T2V, T2W); TN = VADD(Tz, TM); T3t = VSUB(TM, Tz); { V T2E, T1O, T3B, T3H, T2x, T2q, T2K, T2J, T3C, T23, T3I, T2H; { V T2F, T1V, T22, T2G; T2E = VADD(T1I, T1N); T1O = VSUB(T1I, T1N); { V T3b, T33, T3c, T38; T3b = VSUB(T32, T2Z); T33 = VADD(T2Z, T32); T3c = VADD(T34, T37); T38 = VSUB(T34, T37); { V T3a, T2Y, T3s, T3q; T3a = VSUB(T2U, T2X); T2Y = VADD(T2U, T2X); T3s = VSUB(T3p, T3j); T3q = VADD(T3j, T3p); { V T3x, T3v, T3e, TO; T3x = VSUB(T3u, T3t); T3v = VADD(T3t, T3u); T3e = VSUB(Tm, TN); TO = VADD(Tm, TN); { V T3d, T3w, T3y, T39; T3d = VSUB(T3b, T3c); T3w = VADD(T3b, T3c); T3y = VSUB(T38, T33); T39 = VADD(T33, T38); ST(&(ii[WS(rs, 4)]), VADD(T3r, T3s), ms, &(ii[0])); ST(&(ii[WS(rs, 12)]), VSUB(T3s, T3r), ms, &(ii[0])); ST(&(ii[0]), VADD(T3i, T3q), ms, &(ii[0])); ST(&(ii[WS(rs, 8)]), VSUB(T3q, T3i), ms, &(ii[0])); ST(&(ri[WS(rs, 4)]), VADD(T3e, T3h), ms, &(ri[0])); ST(&(ri[WS(rs, 12)]), VSUB(T3e, T3h), ms, &(ri[0])); ST(&(ri[0]), VADD(TO, T1H), ms, &(ri[0])); ST(&(ri[WS(rs, 8)]), VSUB(TO, T1H), ms, &(ri[0])); ST(&(ri[WS(rs, 6)]), VFMA(LDK(KP707106781), T3d, T3a), ms, &(ri[0])); ST(&(ri[WS(rs, 14)]), VFNMS(LDK(KP707106781), T3d, T3a), ms, &(ri[0])); ST(&(ii[WS(rs, 10)]), VFNMS(LDK(KP707106781), T3w, T3v), ms, &(ii[0])); ST(&(ii[WS(rs, 2)]), VFMA(LDK(KP707106781), T3w, T3v), ms, &(ii[0])); ST(&(ii[WS(rs, 14)]), VFNMS(LDK(KP707106781), T3y, T3x), ms, &(ii[0])); ST(&(ii[WS(rs, 6)]), VFMA(LDK(KP707106781), T3y, T3x), ms, &(ii[0])); ST(&(ri[WS(rs, 2)]), VFMA(LDK(KP707106781), T39, T2Y), ms, &(ri[0])); ST(&(ri[WS(rs, 10)]), VFNMS(LDK(KP707106781), T39, T2Y), ms, &(ri[0])); T3B = VSUB(T3z, T3A); T3H = VADD(T3A, T3z); } } } } T2F = VADD(T1U, T1T); T1V = VSUB(T1T, T1U); T22 = VADD(T1W, T21); T2G = VSUB(T1W, T21); { V T2M, T2N, T2b, T2i; T2x = VSUB(T2r, T2w); T2M = VADD(T2r, T2w); T2N = VSUB(T2o, T2p); T2q = VADD(T2o, T2p); T2K = VSUB(T29, T2a); T2b = VADD(T29, T2a); T2i = VSUB(T2c, T2h); T2J = VADD(T2c, T2h); T3C = VADD(T1V, T22); T23 = VSUB(T1V, T22); T2S = VFMA(LDK(KP414213562), T2M, T2N); T2O = VFNMS(LDK(KP414213562), T2N, T2M); T3I = VSUB(T2G, T2F); T2H = VADD(T2F, T2G); T2B = VFNMS(LDK(KP414213562), T2b, T2i); T2j = VFMA(LDK(KP414213562), T2i, T2b); } } T2A = VFNMS(LDK(KP707106781), T23, T1O); T24 = VFMA(LDK(KP707106781), T23, T1O); T3J = VFMA(LDK(KP707106781), T3I, T3H); T3L = VFNMS(LDK(KP707106781), T3I, T3H); T2Q = VFNMS(LDK(KP707106781), T2H, T2E); T2I = VFMA(LDK(KP707106781), T2H, T2E); T2R = VFNMS(LDK(KP414213562), T2J, T2K); T2L = VFMA(LDK(KP414213562), T2K, T2J); T2C = VFMA(LDK(KP414213562), T2q, T2x); T2y = VFNMS(LDK(KP414213562), T2x, T2q); T3D = VFMA(LDK(KP707106781), T3C, T3B); T3F = VFNMS(LDK(KP707106781), T3C, T3B); } } } } { V T3E, T2T, T2P, T3G; T3E = VADD(T2R, T2S); T2T = VSUB(T2R, T2S); T2P = VADD(T2L, T2O); T3G = VSUB(T2O, T2L); { V T3K, T2D, T2z, T3M; T3K = VSUB(T2C, T2B); T2D = VADD(T2B, T2C); T2z = VSUB(T2j, T2y); T3M = VADD(T2j, T2y); ST(&(ri[WS(rs, 5)]), VFMA(LDK(KP923879532), T2T, T2Q), ms, &(ri[WS(rs, 1)])); ST(&(ri[WS(rs, 13)]), VFNMS(LDK(KP923879532), T2T, T2Q), ms, &(ri[WS(rs, 1)])); ST(&(ii[WS(rs, 9)]), VFNMS(LDK(KP923879532), T3E, T3D), ms, &(ii[WS(rs, 1)])); ST(&(ii[WS(rs, 1)]), VFMA(LDK(KP923879532), T3E, T3D), ms, &(ii[WS(rs, 1)])); ST(&(ii[WS(rs, 13)]), VFNMS(LDK(KP923879532), T3G, T3F), ms, &(ii[WS(rs, 1)])); ST(&(ii[WS(rs, 5)]), VFMA(LDK(KP923879532), T3G, T3F), ms, &(ii[WS(rs, 1)])); ST(&(ri[WS(rs, 1)]), VFMA(LDK(KP923879532), T2P, T2I), ms, &(ri[WS(rs, 1)])); ST(&(ri[WS(rs, 9)]), VFNMS(LDK(KP923879532), T2P, T2I), ms, &(ri[WS(rs, 1)])); ST(&(ri[WS(rs, 15)]), VFMA(LDK(KP923879532), T2D, T2A), ms, &(ri[WS(rs, 1)])); ST(&(ri[WS(rs, 7)]), VFNMS(LDK(KP923879532), T2D, T2A), ms, &(ri[WS(rs, 1)])); ST(&(ii[WS(rs, 11)]), VFNMS(LDK(KP923879532), T3K, T3J), ms, &(ii[WS(rs, 1)])); ST(&(ii[WS(rs, 3)]), VFMA(LDK(KP923879532), T3K, T3J), ms, &(ii[WS(rs, 1)])); ST(&(ii[WS(rs, 15)]), VFMA(LDK(KP923879532), T3M, T3L), ms, &(ii[WS(rs, 1)])); ST(&(ii[WS(rs, 7)]), VFNMS(LDK(KP923879532), T3M, T3L), ms, &(ii[WS(rs, 1)])); ST(&(ri[WS(rs, 3)]), VFMA(LDK(KP923879532), T2z, T24), ms, &(ri[WS(rs, 1)])); ST(&(ri[WS(rs, 11)]), VFNMS(LDK(KP923879532), T2z, T24), ms, &(ri[WS(rs, 1)])); } } } } VLEAVE(); } static const tw_instr twinstr[] = { VTW(0, 1), VTW(0, 2), VTW(0, 3), VTW(0, 4), VTW(0, 5), VTW(0, 6), VTW(0, 7), VTW(0, 8), VTW(0, 9), VTW(0, 10), VTW(0, 11), VTW(0, 12), VTW(0, 13), VTW(0, 14), VTW(0, 15), {TW_NEXT, (2 * VL), 0} }; static const ct_desc desc = { 16, XSIMD_STRING("t1sv_16"), twinstr, &GENUS, {104, 30, 70, 0}, 0, 0, 0 }; void XSIMD(codelet_t1sv_16) (planner *p) { X(kdft_dit_register) (p, t1sv_16, &desc); } #else /* HAVE_FMA */ /* Generated by: ../../../genfft/gen_twiddle.native -simd -compact -variables 4 -pipeline-latency 8 -n 16 -name t1sv_16 -include ts.h */ /* * This function contains 174 FP additions, 84 FP multiplications, * (or, 136 additions, 46 multiplications, 38 fused multiply/add), * 52 stack variables, 3 constants, and 64 memory accesses */ #include "ts.h" static void t1sv_16(R *ri, R *ii, const R *W, stride rs, INT mb, INT me, INT ms) { DVK(KP382683432, +0.382683432365089771728459984030398866761344562); DVK(KP923879532, +0.923879532511286756128183189396788286822416626); DVK(KP707106781, +0.707106781186547524400844362104849039284835938); { INT m; for (m = mb, W = W + (mb * 30); m < me; m = m + (2 * VL), ri = ri + ((2 * VL) * ms), ii = ii + ((2 * VL) * ms), W = W + ((2 * VL) * 30), MAKE_VOLATILE_STRIDE(32, rs)) { V T7, T37, T1t, T2U, Ti, T38, T1w, T2R, Tu, T2s, T1C, T2c, TF, T2t, T1H; V T2d, T1f, T1q, T2B, T2C, T2D, T2E, T1Z, T2j, T24, T2k, TS, T13, T2w, T2x; V T2y, T2z, T1O, T2g, T1T, T2h; { V T1, T2T, T6, T2S; T1 = LD(&(ri[0]), ms, &(ri[0])); T2T = LD(&(ii[0]), ms, &(ii[0])); { V T3, T5, T2, T4; T3 = LD(&(ri[WS(rs, 8)]), ms, &(ri[0])); T5 = LD(&(ii[WS(rs, 8)]), ms, &(ii[0])); T2 = LDW(&(W[TWVL * 14])); T4 = LDW(&(W[TWVL * 15])); T6 = VFMA(T2, T3, VMUL(T4, T5)); T2S = VFNMS(T4, T3, VMUL(T2, T5)); } T7 = VADD(T1, T6); T37 = VSUB(T2T, T2S); T1t = VSUB(T1, T6); T2U = VADD(T2S, T2T); } { V Tc, T1u, Th, T1v; { V T9, Tb, T8, Ta; T9 = LD(&(ri[WS(rs, 4)]), ms, &(ri[0])); Tb = LD(&(ii[WS(rs, 4)]), ms, &(ii[0])); T8 = LDW(&(W[TWVL * 6])); Ta = LDW(&(W[TWVL * 7])); Tc = VFMA(T8, T9, VMUL(Ta, Tb)); T1u = VFNMS(Ta, T9, VMUL(T8, Tb)); } { V Te, Tg, Td, Tf; Te = LD(&(ri[WS(rs, 12)]), ms, &(ri[0])); Tg = LD(&(ii[WS(rs, 12)]), ms, &(ii[0])); Td = LDW(&(W[TWVL * 22])); Tf = LDW(&(W[TWVL * 23])); Th = VFMA(Td, Te, VMUL(Tf, Tg)); T1v = VFNMS(Tf, Te, VMUL(Td, Tg)); } Ti = VADD(Tc, Th); T38 = VSUB(Tc, Th); T1w = VSUB(T1u, T1v); T2R = VADD(T1u, T1v); } { V To, T1y, Tt, T1z, T1A, T1B; { V Tl, Tn, Tk, Tm; Tl = LD(&(ri[WS(rs, 2)]), ms, &(ri[0])); Tn = LD(&(ii[WS(rs, 2)]), ms, &(ii[0])); Tk = LDW(&(W[TWVL * 2])); Tm = LDW(&(W[TWVL * 3])); To = VFMA(Tk, Tl, VMUL(Tm, Tn)); T1y = VFNMS(Tm, Tl, VMUL(Tk, Tn)); } { V Tq, Ts, Tp, Tr; Tq = LD(&(ri[WS(rs, 10)]), ms, &(ri[0])); Ts = LD(&(ii[WS(rs, 10)]), ms, &(ii[0])); Tp = LDW(&(W[TWVL * 18])); Tr = LDW(&(W[TWVL * 19])); Tt = VFMA(Tp, Tq, VMUL(Tr, Ts)); T1z = VFNMS(Tr, Tq, VMUL(Tp, Ts)); } Tu = VADD(To, Tt); T2s = VADD(T1y, T1z); T1A = VSUB(T1y, T1z); T1B = VSUB(To, Tt); T1C = VSUB(T1A, T1B); T2c = VADD(T1B, T1A); } { V Tz, T1E, TE, T1F, T1D, T1G; { V Tw, Ty, Tv, Tx; Tw = LD(&(ri[WS(rs, 14)]), ms, &(ri[0])); Ty = LD(&(ii[WS(rs, 14)]), ms, &(ii[0])); Tv = LDW(&(W[TWVL * 26])); Tx = LDW(&(W[TWVL * 27])); Tz = VFMA(Tv, Tw, VMUL(Tx, Ty)); T1E = VFNMS(Tx, Tw, VMUL(Tv, Ty)); } { V TB, TD, TA, TC; TB = LD(&(ri[WS(rs, 6)]), ms, &(ri[0])); TD = LD(&(ii[WS(rs, 6)]), ms, &(ii[0])); TA = LDW(&(W[TWVL * 10])); TC = LDW(&(W[TWVL * 11])); TE = VFMA(TA, TB, VMUL(TC, TD)); T1F = VFNMS(TC, TB, VMUL(TA, TD)); } TF = VADD(Tz, TE); T2t = VADD(T1E, T1F); T1D = VSUB(Tz, TE); T1G = VSUB(T1E, T1F); T1H = VADD(T1D, T1G); T2d = VSUB(T1D, T1G); } { V T19, T20, T1p, T1X, T1e, T21, T1k, T1W; { V T16, T18, T15, T17; T16 = LD(&(ri[WS(rs, 15)]), ms, &(ri[WS(rs, 1)])); T18 = LD(&(ii[WS(rs, 15)]), ms, &(ii[WS(rs, 1)])); T15 = LDW(&(W[TWVL * 28])); T17 = LDW(&(W[TWVL * 29])); T19 = VFMA(T15, T16, VMUL(T17, T18)); T20 = VFNMS(T17, T16, VMUL(T15, T18)); } { V T1m, T1o, T1l, T1n; T1m = LD(&(ri[WS(rs, 11)]), ms, &(ri[WS(rs, 1)])); T1o = LD(&(ii[WS(rs, 11)]), ms, &(ii[WS(rs, 1)])); T1l = LDW(&(W[TWVL * 20])); T1n = LDW(&(W[TWVL * 21])); T1p = VFMA(T1l, T1m, VMUL(T1n, T1o)); T1X = VFNMS(T1n, T1m, VMUL(T1l, T1o)); } { V T1b, T1d, T1a, T1c; T1b = LD(&(ri[WS(rs, 7)]), ms, &(ri[WS(rs, 1)])); T1d = LD(&(ii[WS(rs, 7)]), ms, &(ii[WS(rs, 1)])); T1a = LDW(&(W[TWVL * 12])); T1c = LDW(&(W[TWVL * 13])); T1e = VFMA(T1a, T1b, VMUL(T1c, T1d)); T21 = VFNMS(T1c, T1b, VMUL(T1a, T1d)); } { V T1h, T1j, T1g, T1i; T1h = LD(&(ri[WS(rs, 3)]), ms, &(ri[WS(rs, 1)])); T1j = LD(&(ii[WS(rs, 3)]), ms, &(ii[WS(rs, 1)])); T1g = LDW(&(W[TWVL * 4])); T1i = LDW(&(W[TWVL * 5])); T1k = VFMA(T1g, T1h, VMUL(T1i, T1j)); T1W = VFNMS(T1i, T1h, VMUL(T1g, T1j)); } T1f = VADD(T19, T1e); T1q = VADD(T1k, T1p); T2B = VSUB(T1f, T1q); T2C = VADD(T20, T21); T2D = VADD(T1W, T1X); T2E = VSUB(T2C, T2D); { V T1V, T1Y, T22, T23; T1V = VSUB(T19, T1e); T1Y = VSUB(T1W, T1X); T1Z = VSUB(T1V, T1Y); T2j = VADD(T1V, T1Y); T22 = VSUB(T20, T21); T23 = VSUB(T1k, T1p); T24 = VADD(T22, T23); T2k = VSUB(T22, T23); } } { V TM, T1K, T12, T1R, TR, T1L, TX, T1Q; { V TJ, TL, TI, TK; TJ = LD(&(ri[WS(rs, 1)]), ms, &(ri[WS(rs, 1)])); TL = LD(&(ii[WS(rs, 1)]), ms, &(ii[WS(rs, 1)])); TI = LDW(&(W[0])); TK = LDW(&(W[TWVL * 1])); TM = VFMA(TI, TJ, VMUL(TK, TL)); T1K = VFNMS(TK, TJ, VMUL(TI, TL)); } { V TZ, T11, TY, T10; TZ = LD(&(ri[WS(rs, 13)]), ms, &(ri[WS(rs, 1)])); T11 = LD(&(ii[WS(rs, 13)]), ms, &(ii[WS(rs, 1)])); TY = LDW(&(W[TWVL * 24])); T10 = LDW(&(W[TWVL * 25])); T12 = VFMA(TY, TZ, VMUL(T10, T11)); T1R = VFNMS(T10, TZ, VMUL(TY, T11)); } { V TO, TQ, TN, TP; TO = LD(&(ri[WS(rs, 9)]), ms, &(ri[WS(rs, 1)])); TQ = LD(&(ii[WS(rs, 9)]), ms, &(ii[WS(rs, 1)])); TN = LDW(&(W[TWVL * 16])); TP = LDW(&(W[TWVL * 17])); TR = VFMA(TN, TO, VMUL(TP, TQ)); T1L = VFNMS(TP, TO, VMUL(TN, TQ)); } { V TU, TW, TT, TV; TU = LD(&(ri[WS(rs, 5)]), ms, &(ri[WS(rs, 1)])); TW = LD(&(ii[WS(rs, 5)]), ms, &(ii[WS(rs, 1)])); TT = LDW(&(W[TWVL * 8])); TV = LDW(&(W[TWVL * 9])); TX = VFMA(TT, TU, VMUL(TV, TW)); T1Q = VFNMS(TV, TU, VMUL(TT, TW)); } TS = VADD(TM, TR); T13 = VADD(TX, T12); T2w = VSUB(TS, T13); T2x = VADD(T1K, T1L); T2y = VADD(T1Q, T1R); T2z = VSUB(T2x, T2y); { V T1M, T1N, T1P, T1S; T1M = VSUB(T1K, T1L); T1N = VSUB(TX, T12); T1O = VADD(T1M, T1N); T2g = VSUB(T1M, T1N); T1P = VSUB(TM, TR); T1S = VSUB(T1Q, T1R); T1T = VSUB(T1P, T1S); T2h = VADD(T1P, T1S); } } { V T1J, T27, T3g, T3i, T26, T3h, T2a, T3d; { V T1x, T1I, T3e, T3f; T1x = VSUB(T1t, T1w); T1I = VMUL(LDK(KP707106781), VSUB(T1C, T1H)); T1J = VADD(T1x, T1I); T27 = VSUB(T1x, T1I); T3e = VMUL(LDK(KP707106781), VSUB(T2d, T2c)); T3f = VADD(T38, T37); T3g = VADD(T3e, T3f); T3i = VSUB(T3f, T3e); } { V T1U, T25, T28, T29; T1U = VFMA(LDK(KP923879532), T1O, VMUL(LDK(KP382683432), T1T)); T25 = VFNMS(LDK(KP923879532), T24, VMUL(LDK(KP382683432), T1Z)); T26 = VADD(T1U, T25); T3h = VSUB(T25, T1U); T28 = VFNMS(LDK(KP923879532), T1T, VMUL(LDK(KP382683432), T1O)); T29 = VFMA(LDK(KP382683432), T24, VMUL(LDK(KP923879532), T1Z)); T2a = VSUB(T28, T29); T3d = VADD(T28, T29); } ST(&(ri[WS(rs, 11)]), VSUB(T1J, T26), ms, &(ri[WS(rs, 1)])); ST(&(ii[WS(rs, 11)]), VSUB(T3g, T3d), ms, &(ii[WS(rs, 1)])); ST(&(ri[WS(rs, 3)]), VADD(T1J, T26), ms, &(ri[WS(rs, 1)])); ST(&(ii[WS(rs, 3)]), VADD(T3d, T3g), ms, &(ii[WS(rs, 1)])); ST(&(ri[WS(rs, 15)]), VSUB(T27, T2a), ms, &(ri[WS(rs, 1)])); ST(&(ii[WS(rs, 15)]), VSUB(T3i, T3h), ms, &(ii[WS(rs, 1)])); ST(&(ri[WS(rs, 7)]), VADD(T27, T2a), ms, &(ri[WS(rs, 1)])); ST(&(ii[WS(rs, 7)]), VADD(T3h, T3i), ms, &(ii[WS(rs, 1)])); } { V T2v, T2H, T32, T34, T2G, T33, T2K, T2Z; { V T2r, T2u, T30, T31; T2r = VSUB(T7, Ti); T2u = VSUB(T2s, T2t); T2v = VADD(T2r, T2u); T2H = VSUB(T2r, T2u); T30 = VSUB(TF, Tu); T31 = VSUB(T2U, T2R); T32 = VADD(T30, T31); T34 = VSUB(T31, T30); } { V T2A, T2F, T2I, T2J; T2A = VADD(T2w, T2z); T2F = VSUB(T2B, T2E); T2G = VMUL(LDK(KP707106781), VADD(T2A, T2F)); T33 = VMUL(LDK(KP707106781), VSUB(T2F, T2A)); T2I = VSUB(T2z, T2w); T2J = VADD(T2B, T2E); T2K = VMUL(LDK(KP707106781), VSUB(T2I, T2J)); T2Z = VMUL(LDK(KP707106781), VADD(T2I, T2J)); } ST(&(ri[WS(rs, 10)]), VSUB(T2v, T2G), ms, &(ri[0])); ST(&(ii[WS(rs, 10)]), VSUB(T32, T2Z), ms, &(ii[0])); ST(&(ri[WS(rs, 2)]), VADD(T2v, T2G), ms, &(ri[0])); ST(&(ii[WS(rs, 2)]), VADD(T2Z, T32), ms, &(ii[0])); ST(&(ri[WS(rs, 14)]), VSUB(T2H, T2K), ms, &(ri[0])); ST(&(ii[WS(rs, 14)]), VSUB(T34, T33), ms, &(ii[0])); ST(&(ri[WS(rs, 6)]), VADD(T2H, T2K), ms, &(ri[0])); ST(&(ii[WS(rs, 6)]), VADD(T33, T34), ms, &(ii[0])); } { V T2f, T2n, T3a, T3c, T2m, T3b, T2q, T35; { V T2b, T2e, T36, T39; T2b = VADD(T1t, T1w); T2e = VMUL(LDK(KP707106781), VADD(T2c, T2d)); T2f = VADD(T2b, T2e); T2n = VSUB(T2b, T2e); T36 = VMUL(LDK(KP707106781), VADD(T1C, T1H)); T39 = VSUB(T37, T38); T3a = VADD(T36, T39); T3c = VSUB(T39, T36); } { V T2i, T2l, T2o, T2p; T2i = VFMA(LDK(KP382683432), T2g, VMUL(LDK(KP923879532), T2h)); T2l = VFNMS(LDK(KP382683432), T2k, VMUL(LDK(KP923879532), T2j)); T2m = VADD(T2i, T2l); T3b = VSUB(T2l, T2i); T2o = VFNMS(LDK(KP382683432), T2h, VMUL(LDK(KP923879532), T2g)); T2p = VFMA(LDK(KP923879532), T2k, VMUL(LDK(KP382683432), T2j)); T2q = VSUB(T2o, T2p); T35 = VADD(T2o, T2p); } ST(&(ri[WS(rs, 9)]), VSUB(T2f, T2m), ms, &(ri[WS(rs, 1)])); ST(&(ii[WS(rs, 9)]), VSUB(T3a, T35), ms, &(ii[WS(rs, 1)])); ST(&(ri[WS(rs, 1)]), VADD(T2f, T2m), ms, &(ri[WS(rs, 1)])); ST(&(ii[WS(rs, 1)]), VADD(T35, T3a), ms, &(ii[WS(rs, 1)])); ST(&(ri[WS(rs, 13)]), VSUB(T2n, T2q), ms, &(ri[WS(rs, 1)])); ST(&(ii[WS(rs, 13)]), VSUB(T3c, T3b), ms, &(ii[WS(rs, 1)])); ST(&(ri[WS(rs, 5)]), VADD(T2n, T2q), ms, &(ri[WS(rs, 1)])); ST(&(ii[WS(rs, 5)]), VADD(T3b, T3c), ms, &(ii[WS(rs, 1)])); } { V TH, T2L, T2W, T2Y, T1s, T2X, T2O, T2P; { V Tj, TG, T2Q, T2V; Tj = VADD(T7, Ti); TG = VADD(Tu, TF); TH = VADD(Tj, TG); T2L = VSUB(Tj, TG); T2Q = VADD(T2s, T2t); T2V = VADD(T2R, T2U); T2W = VADD(T2Q, T2V); T2Y = VSUB(T2V, T2Q); } { V T14, T1r, T2M, T2N; T14 = VADD(TS, T13); T1r = VADD(T1f, T1q); T1s = VADD(T14, T1r); T2X = VSUB(T1r, T14); T2M = VADD(T2x, T2y); T2N = VADD(T2C, T2D); T2O = VSUB(T2M, T2N); T2P = VADD(T2M, T2N); } ST(&(ri[WS(rs, 8)]), VSUB(TH, T1s), ms, &(ri[0])); ST(&(ii[WS(rs, 8)]), VSUB(T2W, T2P), ms, &(ii[0])); ST(&(ri[0]), VADD(TH, T1s), ms, &(ri[0])); ST(&(ii[0]), VADD(T2P, T2W), ms, &(ii[0])); ST(&(ri[WS(rs, 12)]), VSUB(T2L, T2O), ms, &(ri[0])); ST(&(ii[WS(rs, 12)]), VSUB(T2Y, T2X), ms, &(ii[0])); ST(&(ri[WS(rs, 4)]), VADD(T2L, T2O), ms, &(ri[0])); ST(&(ii[WS(rs, 4)]), VADD(T2X, T2Y), ms, &(ii[0])); } } } VLEAVE(); } static const tw_instr twinstr[] = { VTW(0, 1), VTW(0, 2), VTW(0, 3), VTW(0, 4), VTW(0, 5), VTW(0, 6), VTW(0, 7), VTW(0, 8), VTW(0, 9), VTW(0, 10), VTW(0, 11), VTW(0, 12), VTW(0, 13), VTW(0, 14), VTW(0, 15), {TW_NEXT, (2 * VL), 0} }; static const ct_desc desc = { 16, XSIMD_STRING("t1sv_16"), twinstr, &GENUS, {136, 46, 38, 0}, 0, 0, 0 }; void XSIMD(codelet_t1sv_16) (planner *p) { X(kdft_dit_register) (p, t1sv_16, &desc); } #endif /* HAVE_FMA */ fftw-3.3.3/dft/simd/common/t1sv_32.c0000644000175000017500000016676512054410605013777 00000000000000/* * Copyright (c) 2003, 2007-11 Matteo Frigo * Copyright (c) 2003, 2007-11 Massachusetts Institute of Technology * * This program is free software; you can redistribute it and/or modify * it under the terms of the GNU General Public License as published by * the Free Software Foundation; either version 2 of the License, or * (at your option) any later version. * * This program is distributed in the hope that it will be useful, * but WITHOUT ANY WARRANTY; without even the implied warranty of * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the * GNU General Public License for more details. * * You should have received a copy of the GNU General Public License * along with this program; if not, write to the Free Software * Foundation, Inc., 51 Franklin Street, Fifth Floor, Boston, MA 02110-1301 USA * */ /* This file was automatically generated --- DO NOT EDIT */ /* Generated on Sun Nov 25 07:39:25 EST 2012 */ #include "codelet-dft.h" #ifdef HAVE_FMA /* Generated by: ../../../genfft/gen_twiddle.native -fma -reorder-insns -schedule-for-pipeline -simd -compact -variables 4 -pipeline-latency 8 -n 32 -name t1sv_32 -include ts.h */ /* * This function contains 434 FP additions, 260 FP multiplications, * (or, 236 additions, 62 multiplications, 198 fused multiply/add), * 158 stack variables, 7 constants, and 128 memory accesses */ #include "ts.h" static void t1sv_32(R *ri, R *ii, const R *W, stride rs, INT mb, INT me, INT ms) { DVK(KP831469612, +0.831469612302545237078788377617905756738560812); DVK(KP980785280, +0.980785280403230449126182236134239036973933731); DVK(KP668178637, +0.668178637919298919997757686523080761552472251); DVK(KP198912367, +0.198912367379658006911597622644676228597850501); DVK(KP923879532, +0.923879532511286756128183189396788286822416626); DVK(KP414213562, +0.414213562373095048801688724209698078569671875); DVK(KP707106781, +0.707106781186547524400844362104849039284835938); { INT m; for (m = mb, W = W + (mb * 62); m < me; m = m + (2 * VL), ri = ri + ((2 * VL) * ms), ii = ii + ((2 * VL) * ms), W = W + ((2 * VL) * 62), MAKE_VOLATILE_STRIDE(64, rs)) { V T8Z, T90; { V T87, T8x, T3w, T8, T3B, T83, Tl, T8y, T6F, Tz, T3J, T5T, T6G, TM, T3Q; V T5U, T46, T5Y, T7D, T6L, T5X, T3Z, T6M, T1f, T4l, T61, T7E, T6R, T60, T4e; V T6O, T1G, T5r, T6c, T78, T7N, T54, T6f, T32, T7b, T4S, T65, T6X, T7I, T4v; V T68, T29, T70, T4x, T2f, T5b, T5s, T7O, T7e, T5t, T5i, T79, T3t, T2h, T2k; V T2j, T2o, T2r, T4H, T2y, T2n, T2q, T4y, T2i; { V T3U, TU, TW, TZ, TY, T13, T16, T12, T15, T3V, TX, T44, T1d; { V T1, T86, T3, T6, T5, Ta, Td, Tg, Tj, Tf, T84, T4, Tc, Ti, T3x; V Tb, T2, T9; T1 = LD(&(ri[0]), ms, &(ri[0])); T86 = LD(&(ii[0]), ms, &(ii[0])); T3 = LD(&(ri[WS(rs, 16)]), ms, &(ri[0])); T6 = LD(&(ii[WS(rs, 16)]), ms, &(ii[0])); T2 = LDW(&(W[TWVL * 30])); T5 = LDW(&(W[TWVL * 31])); Ta = LD(&(ri[WS(rs, 8)]), ms, &(ri[0])); Td = LD(&(ii[WS(rs, 8)]), ms, &(ii[0])); T9 = LDW(&(W[TWVL * 14])); Tg = LD(&(ri[WS(rs, 24)]), ms, &(ri[0])); Tj = LD(&(ii[WS(rs, 24)]), ms, &(ii[0])); Tf = LDW(&(W[TWVL * 46])); T84 = VMUL(T2, T6); T4 = VMUL(T2, T3); Tc = LDW(&(W[TWVL * 15])); Ti = LDW(&(W[TWVL * 47])); T3x = VMUL(T9, Td); Tb = VMUL(T9, Ta); { V Tu, Tx, T3F, Ts, Tt, Tw; { V To, Tr, Tq, T3E, Tp; { V T3y, Te, Tn, T3A, Tk; { V T3z, Th, T85, T7; To = LD(&(ri[WS(rs, 4)]), ms, &(ri[0])); T3z = VMUL(Tf, Tj); Th = VMUL(Tf, Tg); T85 = VFNMS(T5, T3, T84); T7 = VFMA(T5, T6, T4); Tr = LD(&(ii[WS(rs, 4)]), ms, &(ii[0])); T3y = VFNMS(Tc, Ta, T3x); Te = VFMA(Tc, Td, Tb); Tn = LDW(&(W[TWVL * 6])); T3A = VFNMS(Ti, Tg, T3z); Tk = VFMA(Ti, Tj, Th); T87 = VADD(T85, T86); T8x = VSUB(T86, T85); T3w = VSUB(T1, T7); T8 = VADD(T1, T7); } Tq = LDW(&(W[TWVL * 7])); T3E = VMUL(Tn, Tr); Tp = VMUL(Tn, To); T3B = VSUB(T3y, T3A); T83 = VADD(T3y, T3A); Tl = VADD(Te, Tk); T8y = VSUB(Te, Tk); } Tu = LD(&(ri[WS(rs, 20)]), ms, &(ri[0])); Tx = LD(&(ii[WS(rs, 20)]), ms, &(ii[0])); T3F = VFNMS(Tq, To, T3E); Ts = VFMA(Tq, Tr, Tp); Tt = LDW(&(W[TWVL * 38])); Tw = LDW(&(W[TWVL * 39])); } { V TB, TE, TD, TH, TK, T3G, Tv, TG, TJ, T3L, TC, TA; TB = LD(&(ri[WS(rs, 28)]), ms, &(ri[0])); TE = LD(&(ii[WS(rs, 28)]), ms, &(ii[0])); TA = LDW(&(W[TWVL * 54])); TD = LDW(&(W[TWVL * 55])); TH = LD(&(ri[WS(rs, 12)]), ms, &(ri[0])); TK = LD(&(ii[WS(rs, 12)]), ms, &(ii[0])); T3G = VMUL(Tt, Tx); Tv = VMUL(Tt, Tu); TG = LDW(&(W[TWVL * 22])); TJ = LDW(&(W[TWVL * 23])); T3L = VMUL(TA, TE); TC = VMUL(TA, TB); { V T19, T1c, T3P, T3K, T18, T1b, TV, T43, T1a; { V TQ, TT, T3M, TF, TS, T3I, T3D, T3O, TL, T3T, TR; { V T3H, Ty, T3N, TI, TP; TQ = LD(&(ri[WS(rs, 2)]), ms, &(ri[0])); TT = LD(&(ii[WS(rs, 2)]), ms, &(ii[0])); T3H = VFNMS(Tw, Tu, T3G); Ty = VFMA(Tw, Tx, Tv); T3N = VMUL(TG, TK); TI = VMUL(TG, TH); T3M = VFNMS(TD, TB, T3L); TF = VFMA(TD, TE, TC); TP = LDW(&(W[TWVL * 2])); TS = LDW(&(W[TWVL * 3])); T6F = VADD(T3F, T3H); T3I = VSUB(T3F, T3H); Tz = VADD(Ts, Ty); T3D = VSUB(Ts, Ty); T3O = VFNMS(TJ, TH, T3N); TL = VFMA(TJ, TK, TI); T3T = VMUL(TP, TT); TR = VMUL(TP, TQ); } T19 = LD(&(ri[WS(rs, 26)]), ms, &(ri[0])); T1c = LD(&(ii[WS(rs, 26)]), ms, &(ii[0])); T3J = VADD(T3D, T3I); T5T = VSUB(T3I, T3D); T6G = VADD(T3M, T3O); T3P = VSUB(T3M, T3O); TM = VADD(TF, TL); T3K = VSUB(TF, TL); T3U = VFNMS(TS, TQ, T3T); TU = VFMA(TS, TT, TR); T18 = LDW(&(W[TWVL * 50])); T1b = LDW(&(W[TWVL * 51])); } TW = LD(&(ri[WS(rs, 18)]), ms, &(ri[0])); TZ = LD(&(ii[WS(rs, 18)]), ms, &(ii[0])); T3Q = VSUB(T3K, T3P); T5U = VADD(T3K, T3P); TV = LDW(&(W[TWVL * 34])); TY = LDW(&(W[TWVL * 35])); T43 = VMUL(T18, T1c); T1a = VMUL(T18, T19); T13 = LD(&(ri[WS(rs, 10)]), ms, &(ri[0])); T16 = LD(&(ii[WS(rs, 10)]), ms, &(ii[0])); T12 = LDW(&(W[TWVL * 18])); T15 = LDW(&(W[TWVL * 19])); T3V = VMUL(TV, TZ); TX = VMUL(TV, TW); T44 = VFNMS(T1b, T19, T43); T1d = VFMA(T1b, T1c, T1a); } } } } { V T4Z, T2H, T2J, T2M, T2L, T2Q, T2T, T2P, T2S, T5p, T30, T50, T2K; { V T49, T1l, T1n, T1q, T1p, T1u, T1x, T4j, T1E, T1t, T1w, T4a, T1o; { V T1A, T1D, T1C, T4i, T1B, T1m; { V T1h, T1k, T41, T14, T3W, T10, T1g, T1j; T1h = LD(&(ri[WS(rs, 30)]), ms, &(ri[0])); T1k = LD(&(ii[WS(rs, 30)]), ms, &(ii[0])); T41 = VMUL(T12, T16); T14 = VMUL(T12, T13); T3W = VFNMS(TY, TW, T3V); T10 = VFMA(TY, TZ, TX); T1g = LDW(&(W[TWVL * 58])); T1j = LDW(&(W[TWVL * 59])); { V T6J, T3X, T11, T40, T48, T1i, T6K, T45, T1e, T3Y, T1z, T42, T17; T1A = LD(&(ri[WS(rs, 22)]), ms, &(ri[0])); T1D = LD(&(ii[WS(rs, 22)]), ms, &(ii[0])); T42 = VFNMS(T15, T13, T41); T17 = VFMA(T15, T16, T14); T6J = VADD(T3U, T3W); T3X = VSUB(T3U, T3W); T11 = VADD(TU, T10); T40 = VSUB(TU, T10); T48 = VMUL(T1g, T1k); T1i = VMUL(T1g, T1h); T6K = VADD(T42, T44); T45 = VSUB(T42, T44); T1e = VADD(T17, T1d); T3Y = VSUB(T17, T1d); T1z = LDW(&(W[TWVL * 42])); T1C = LDW(&(W[TWVL * 43])); T49 = VFNMS(T1j, T1h, T48); T1l = VFMA(T1j, T1k, T1i); T46 = VADD(T40, T45); T5Y = VSUB(T40, T45); T7D = VADD(T6J, T6K); T6L = VSUB(T6J, T6K); T5X = VADD(T3X, T3Y); T3Z = VSUB(T3X, T3Y); T6M = VSUB(T11, T1e); T1f = VADD(T11, T1e); T4i = VMUL(T1z, T1D); T1B = VMUL(T1z, T1A); } } T1n = LD(&(ri[WS(rs, 14)]), ms, &(ri[0])); T1q = LD(&(ii[WS(rs, 14)]), ms, &(ii[0])); T1m = LDW(&(W[TWVL * 26])); T1p = LDW(&(W[TWVL * 27])); T1u = LD(&(ri[WS(rs, 6)]), ms, &(ri[0])); T1x = LD(&(ii[WS(rs, 6)]), ms, &(ii[0])); T4j = VFNMS(T1C, T1A, T4i); T1E = VFMA(T1C, T1D, T1B); T1t = LDW(&(W[TWVL * 10])); T1w = LDW(&(W[TWVL * 11])); T4a = VMUL(T1m, T1q); T1o = VMUL(T1m, T1n); } { V T2W, T2Z, T6P, T4c, T1s, T4f, T6Q, T4k, T1F, T4d, T2V, T2Y, T5o, T2X, T2I; { V T2D, T2G, T2C, T2F, T4g, T1v, T4b, T1r; T2D = LD(&(ri[WS(rs, 31)]), ms, &(ri[WS(rs, 1)])); T2G = LD(&(ii[WS(rs, 31)]), ms, &(ii[WS(rs, 1)])); T2C = LDW(&(W[TWVL * 60])); T2F = LDW(&(W[TWVL * 61])); T4g = VMUL(T1t, T1x); T1v = VMUL(T1t, T1u); T4b = VFNMS(T1p, T1n, T4a); T1r = VFMA(T1p, T1q, T1o); T2W = LD(&(ri[WS(rs, 23)]), ms, &(ri[WS(rs, 1)])); T2Z = LD(&(ii[WS(rs, 23)]), ms, &(ii[WS(rs, 1)])); { V T4Y, T2E, T4h, T1y; T4Y = VMUL(T2C, T2G); T2E = VMUL(T2C, T2D); T4h = VFNMS(T1w, T1u, T4g); T1y = VFMA(T1w, T1x, T1v); T6P = VADD(T49, T4b); T4c = VSUB(T49, T4b); T1s = VADD(T1l, T1r); T4f = VSUB(T1l, T1r); T4Z = VFNMS(T2F, T2D, T4Y); T2H = VFMA(T2F, T2G, T2E); T6Q = VADD(T4h, T4j); T4k = VSUB(T4h, T4j); T1F = VADD(T1y, T1E); T4d = VSUB(T1y, T1E); T2V = LDW(&(W[TWVL * 44])); } T2Y = LDW(&(W[TWVL * 45])); } T2J = LD(&(ri[WS(rs, 15)]), ms, &(ri[WS(rs, 1)])); T2M = LD(&(ii[WS(rs, 15)]), ms, &(ii[WS(rs, 1)])); T4l = VADD(T4f, T4k); T61 = VSUB(T4f, T4k); T7E = VADD(T6P, T6Q); T6R = VSUB(T6P, T6Q); T60 = VADD(T4c, T4d); T4e = VSUB(T4c, T4d); T6O = VSUB(T1s, T1F); T1G = VADD(T1s, T1F); T5o = VMUL(T2V, T2Z); T2X = VMUL(T2V, T2W); T2I = LDW(&(W[TWVL * 28])); T2L = LDW(&(W[TWVL * 29])); T2Q = LD(&(ri[WS(rs, 7)]), ms, &(ri[WS(rs, 1)])); T2T = LD(&(ii[WS(rs, 7)]), ms, &(ii[WS(rs, 1)])); T2P = LDW(&(W[TWVL * 12])); T2S = LDW(&(W[TWVL * 13])); T5p = VFNMS(T2Y, T2W, T5o); T30 = VFMA(T2Y, T2Z, T2X); T50 = VMUL(T2I, T2M); T2K = VMUL(T2I, T2J); } } { V T4q, T1O, T1Q, T1T, T1S, T1X, T20, T4Q, T27, T1W, T1Z, T4r, T1R; { V T23, T26, T25, T4P, T24, T1P; { V T1K, T1N, T5m, T2R, T1J, T1M, T51, T2N; T1K = LD(&(ri[WS(rs, 1)]), ms, &(ri[WS(rs, 1)])); T1N = LD(&(ii[WS(rs, 1)]), ms, &(ii[WS(rs, 1)])); T5m = VMUL(T2P, T2T); T2R = VMUL(T2P, T2Q); T1J = LDW(&(W[0])); T1M = LDW(&(W[TWVL * 1])); T51 = VFNMS(T2L, T2J, T50); T2N = VFMA(T2L, T2M, T2K); { V T76, T52, T2O, T5l, T77, T5q, T31, T53, T22; T23 = LD(&(ri[WS(rs, 25)]), ms, &(ri[WS(rs, 1)])); T26 = LD(&(ii[WS(rs, 25)]), ms, &(ii[WS(rs, 1)])); { V T5n, T2U, T4p, T1L; T5n = VFNMS(T2S, T2Q, T5m); T2U = VFMA(T2S, T2T, T2R); T4p = VMUL(T1J, T1N); T1L = VMUL(T1J, T1K); T76 = VADD(T4Z, T51); T52 = VSUB(T4Z, T51); T2O = VADD(T2H, T2N); T5l = VSUB(T2H, T2N); T77 = VADD(T5n, T5p); T5q = VSUB(T5n, T5p); T31 = VADD(T2U, T30); T53 = VSUB(T2U, T30); T4q = VFNMS(T1M, T1K, T4p); T1O = VFMA(T1M, T1N, T1L); T22 = LDW(&(W[TWVL * 48])); } T25 = LDW(&(W[TWVL * 49])); T5r = VADD(T5l, T5q); T6c = VSUB(T5l, T5q); T78 = VSUB(T76, T77); T7N = VADD(T76, T77); T54 = VSUB(T52, T53); T6f = VADD(T52, T53); T32 = VADD(T2O, T31); T7b = VSUB(T2O, T31); T4P = VMUL(T22, T26); T24 = VMUL(T22, T23); } } T1Q = LD(&(ri[WS(rs, 17)]), ms, &(ri[WS(rs, 1)])); T1T = LD(&(ii[WS(rs, 17)]), ms, &(ii[WS(rs, 1)])); T1P = LDW(&(W[TWVL * 32])); T1S = LDW(&(W[TWVL * 33])); T1X = LD(&(ri[WS(rs, 9)]), ms, &(ri[WS(rs, 1)])); T20 = LD(&(ii[WS(rs, 9)]), ms, &(ii[WS(rs, 1)])); T4Q = VFNMS(T25, T23, T4P); T27 = VFMA(T25, T26, T24); T1W = LDW(&(W[TWVL * 16])); T1Z = LDW(&(W[TWVL * 17])); T4r = VMUL(T1P, T1T); T1R = VMUL(T1P, T1Q); } { V T56, T38, T3a, T3d, T3c, T3h, T3k, T3g, T3j, T5g, T3r, T57, T3b; { V T3n, T3q, T6V, T4t, T1V, T4M, T6W, T4R, T28, T4u, T3m, T3p, T5f, T3o, T39; { V T34, T37, T33, T36, T4N, T1Y, T4s, T1U; T34 = LD(&(ri[WS(rs, 3)]), ms, &(ri[WS(rs, 1)])); T37 = LD(&(ii[WS(rs, 3)]), ms, &(ii[WS(rs, 1)])); T33 = LDW(&(W[TWVL * 4])); T36 = LDW(&(W[TWVL * 5])); T4N = VMUL(T1W, T20); T1Y = VMUL(T1W, T1X); T4s = VFNMS(T1S, T1Q, T4r); T1U = VFMA(T1S, T1T, T1R); T3n = LD(&(ri[WS(rs, 11)]), ms, &(ri[WS(rs, 1)])); T3q = LD(&(ii[WS(rs, 11)]), ms, &(ii[WS(rs, 1)])); { V T55, T35, T4O, T21; T55 = VMUL(T33, T37); T35 = VMUL(T33, T34); T4O = VFNMS(T1Z, T1X, T4N); T21 = VFMA(T1Z, T20, T1Y); T6V = VADD(T4q, T4s); T4t = VSUB(T4q, T4s); T1V = VADD(T1O, T1U); T4M = VSUB(T1O, T1U); T56 = VFNMS(T36, T34, T55); T38 = VFMA(T36, T37, T35); T6W = VADD(T4O, T4Q); T4R = VSUB(T4O, T4Q); T28 = VADD(T21, T27); T4u = VSUB(T21, T27); T3m = LDW(&(W[TWVL * 20])); } T3p = LDW(&(W[TWVL * 21])); } T3a = LD(&(ri[WS(rs, 19)]), ms, &(ri[WS(rs, 1)])); T3d = LD(&(ii[WS(rs, 19)]), ms, &(ii[WS(rs, 1)])); T4S = VADD(T4M, T4R); T65 = VSUB(T4M, T4R); T6X = VSUB(T6V, T6W); T7I = VADD(T6V, T6W); T4v = VSUB(T4t, T4u); T68 = VADD(T4t, T4u); T29 = VADD(T1V, T28); T70 = VSUB(T1V, T28); T5f = VMUL(T3m, T3q); T3o = VMUL(T3m, T3n); T39 = LDW(&(W[TWVL * 36])); T3c = LDW(&(W[TWVL * 37])); T3h = LD(&(ri[WS(rs, 27)]), ms, &(ri[WS(rs, 1)])); T3k = LD(&(ii[WS(rs, 27)]), ms, &(ii[WS(rs, 1)])); T3g = LDW(&(W[TWVL * 52])); T3j = LDW(&(W[TWVL * 53])); T5g = VFNMS(T3p, T3n, T5f); T3r = VFMA(T3p, T3q, T3o); T57 = VMUL(T39, T3d); T3b = VMUL(T39, T3a); } { V T2u, T2x, T2w, T4G, T2v, T2g; { V T2b, T2e, T5d, T3i, T2a, T2d, T58, T3e, T2t; T2b = LD(&(ri[WS(rs, 5)]), ms, &(ri[WS(rs, 1)])); T2e = LD(&(ii[WS(rs, 5)]), ms, &(ii[WS(rs, 1)])); T5d = VMUL(T3g, T3k); T3i = VMUL(T3g, T3h); T2a = LDW(&(W[TWVL * 8])); T2d = LDW(&(W[TWVL * 9])); T58 = VFNMS(T3c, T3a, T57); T3e = VFMA(T3c, T3d, T3b); T2u = LD(&(ri[WS(rs, 13)]), ms, &(ri[WS(rs, 1)])); T2x = LD(&(ii[WS(rs, 13)]), ms, &(ii[WS(rs, 1)])); { V T5e, T3l, T4w, T2c; T5e = VFNMS(T3j, T3h, T5d); T3l = VFMA(T3j, T3k, T3i); T4w = VMUL(T2a, T2e); T2c = VMUL(T2a, T2b); { V T7c, T59, T3f, T5a; T7c = VADD(T56, T58); T59 = VSUB(T56, T58); T3f = VADD(T38, T3e); T5a = VSUB(T38, T3e); { V T7d, T5h, T3s, T5c; T7d = VADD(T5e, T5g); T5h = VSUB(T5e, T5g); T3s = VADD(T3l, T3r); T5c = VSUB(T3l, T3r); T4x = VFNMS(T2d, T2b, T4w); T2f = VFMA(T2d, T2e, T2c); T5b = VSUB(T59, T5a); T5s = VADD(T5a, T59); T2t = LDW(&(W[TWVL * 24])); T7O = VADD(T7c, T7d); T7e = VSUB(T7c, T7d); T5t = VSUB(T5c, T5h); T5i = VADD(T5c, T5h); T79 = VSUB(T3s, T3f); T3t = VADD(T3f, T3s); } } } T2w = LDW(&(W[TWVL * 25])); T4G = VMUL(T2t, T2x); T2v = VMUL(T2t, T2u); } T2h = LD(&(ri[WS(rs, 21)]), ms, &(ri[WS(rs, 1)])); T2k = LD(&(ii[WS(rs, 21)]), ms, &(ii[WS(rs, 1)])); T2g = LDW(&(W[TWVL * 40])); T2j = LDW(&(W[TWVL * 41])); T2o = LD(&(ri[WS(rs, 29)]), ms, &(ri[WS(rs, 1)])); T2r = LD(&(ii[WS(rs, 29)]), ms, &(ii[WS(rs, 1)])); T4H = VFNMS(T2w, T2u, T4G); T2y = VFMA(T2w, T2x, T2v); T2n = LDW(&(W[TWVL * 56])); T2q = LDW(&(W[TWVL * 57])); T4y = VMUL(T2g, T2k); T2i = VMUL(T2g, T2h); } } } } } { V T4C, T4T, T4U, T4J, T7A, T7w, T7j, T75, T7i, T6U, T8p, T8n, T8v, T8t, T7q; V T7y, T7t, T7z, T7g, T7k; { V T6E, T8j, T6H, T8k, T73, T6Y, T7S, T8i, T8h, T7V; { V T7P, T7Y, T7C, TO, T89, T8e, T3u, T7M, T8d, T1H, T7K, T7X, T2B, T7H; { V T71, T2m, T72, T4I, T2z, T4D, Tm, TN, T2A, T7J; T6E = VSUB(T8, Tl); Tm = VADD(T8, Tl); TN = VADD(Tz, TM); T8j = VSUB(TM, Tz); T7P = VSUB(T7N, T7O); T7Y = VADD(T7N, T7O); { V T82, T4E, T2p, T4z, T2l, T88; T82 = VADD(T6F, T6G); T6H = VSUB(T6F, T6G); T4E = VMUL(T2n, T2r); T2p = VMUL(T2n, T2o); T4z = VFNMS(T2j, T2h, T4y); T2l = VFMA(T2j, T2k, T2i); T8k = VSUB(T87, T83); T88 = VADD(T83, T87); T7C = VSUB(Tm, TN); TO = VADD(Tm, TN); { V T4F, T2s, T4A, T4B; T4F = VFNMS(T2q, T2o, T4E); T2s = VFMA(T2q, T2r, T2p); T71 = VADD(T4x, T4z); T4A = VSUB(T4x, T4z); T2m = VADD(T2f, T2l); T4B = VSUB(T2f, T2l); T89 = VADD(T82, T88); T8e = VSUB(T88, T82); T72 = VADD(T4F, T4H); T4I = VSUB(T4F, T4H); T2z = VADD(T2s, T2y); T4D = VSUB(T2s, T2y); T4C = VSUB(T4A, T4B); T4T = VADD(T4B, T4A); } } T3u = VADD(T32, T3t); T7M = VSUB(T32, T3t); T7J = VADD(T71, T72); T73 = VSUB(T71, T72); T4U = VSUB(T4D, T4I); T4J = VADD(T4D, T4I); T6Y = VSUB(T2z, T2m); T2A = VADD(T2m, T2z); T8d = VSUB(T1G, T1f); T1H = VADD(T1f, T1G); T7K = VSUB(T7I, T7J); T7X = VADD(T7I, T7J); T2B = VADD(T29, T2A); T7H = VSUB(T29, T2A); } { V T1I, T80, T7Q, T7U, T7F, T7L, T7T, T3v, T8b, T8c, T8a, T7W, T81, T7Z; T7W = VSUB(TO, T1H); T1I = VADD(TO, T1H); T7Z = VSUB(T7X, T7Y); T80 = VADD(T7X, T7Y); T7Q = VSUB(T7M, T7P); T7U = VADD(T7M, T7P); T7F = VSUB(T7D, T7E); T81 = VADD(T7D, T7E); T7L = VADD(T7H, T7K); T7T = VSUB(T7K, T7H); T3v = VADD(T2B, T3u); T8b = VSUB(T3u, T2B); ST(&(ri[WS(rs, 24)]), VSUB(T7W, T7Z), ms, &(ri[0])); ST(&(ri[WS(rs, 8)]), VADD(T7W, T7Z), ms, &(ri[0])); T8c = VSUB(T89, T81); T8a = VADD(T81, T89); { V T8f, T8g, T7G, T7R; T7S = VSUB(T7C, T7F); T7G = VADD(T7C, T7F); T7R = VADD(T7L, T7Q); T8i = VSUB(T7Q, T7L); T8h = VSUB(T8e, T8d); T8f = VADD(T8d, T8e); ST(&(ri[0]), VADD(T1I, T3v), ms, &(ri[0])); ST(&(ri[WS(rs, 16)]), VSUB(T1I, T3v), ms, &(ri[0])); T8g = VADD(T7T, T7U); T7V = VSUB(T7T, T7U); ST(&(ii[WS(rs, 16)]), VSUB(T8a, T80), ms, &(ii[0])); ST(&(ii[0]), VADD(T80, T8a), ms, &(ii[0])); ST(&(ii[WS(rs, 24)]), VSUB(T8c, T8b), ms, &(ii[0])); ST(&(ii[WS(rs, 8)]), VADD(T8b, T8c), ms, &(ii[0])); ST(&(ri[WS(rs, 4)]), VFMA(LDK(KP707106781), T7R, T7G), ms, &(ri[0])); ST(&(ri[WS(rs, 20)]), VFNMS(LDK(KP707106781), T7R, T7G), ms, &(ri[0])); ST(&(ii[WS(rs, 20)]), VFNMS(LDK(KP707106781), T8g, T8f), ms, &(ii[0])); ST(&(ii[WS(rs, 4)]), VFMA(LDK(KP707106781), T8g, T8f), ms, &(ii[0])); } } } { V T7f, T7a, T7m, T6I, T7s, T7r, T8r, T8l, T8m, T6T, T8s, T7p; { V T7n, T6N, T6S, T7o, T7u, T7v, T6Z, T74; T7f = VSUB(T7b, T7e); T7u = VADD(T7b, T7e); T7v = VADD(T78, T79); T7a = VSUB(T78, T79); ST(&(ri[WS(rs, 12)]), VFMA(LDK(KP707106781), T7V, T7S), ms, &(ri[0])); ST(&(ri[WS(rs, 28)]), VFNMS(LDK(KP707106781), T7V, T7S), ms, &(ri[0])); ST(&(ii[WS(rs, 28)]), VFNMS(LDK(KP707106781), T8i, T8h), ms, &(ii[0])); ST(&(ii[WS(rs, 12)]), VFMA(LDK(KP707106781), T8i, T8h), ms, &(ii[0])); T7m = VADD(T6E, T6H); T6I = VSUB(T6E, T6H); T7A = VFMA(LDK(KP414213562), T7u, T7v); T7w = VFNMS(LDK(KP414213562), T7v, T7u); T7n = VADD(T6M, T6L); T6N = VSUB(T6L, T6M); T6S = VADD(T6O, T6R); T7o = VSUB(T6O, T6R); T7s = VADD(T6X, T6Y); T6Z = VSUB(T6X, T6Y); T74 = VSUB(T70, T73); T7r = VADD(T70, T73); T8r = VSUB(T8k, T8j); T8l = VADD(T8j, T8k); T8m = VADD(T6N, T6S); T6T = VSUB(T6N, T6S); T7j = VFNMS(LDK(KP414213562), T6Z, T74); T75 = VFMA(LDK(KP414213562), T74, T6Z); T8s = VSUB(T7o, T7n); T7p = VADD(T7n, T7o); } T7i = VFNMS(LDK(KP707106781), T6T, T6I); T6U = VFMA(LDK(KP707106781), T6T, T6I); T8p = VFNMS(LDK(KP707106781), T8m, T8l); T8n = VFMA(LDK(KP707106781), T8m, T8l); T8v = VFNMS(LDK(KP707106781), T8s, T8r); T8t = VFMA(LDK(KP707106781), T8s, T8r); T7q = VFMA(LDK(KP707106781), T7p, T7m); T7y = VFNMS(LDK(KP707106781), T7p, T7m); T7t = VFMA(LDK(KP414213562), T7s, T7r); T7z = VFNMS(LDK(KP414213562), T7r, T7s); T7g = VFNMS(LDK(KP414213562), T7f, T7a); T7k = VFMA(LDK(KP414213562), T7a, T7f); } } { V T5S, T8O, T8N, T5V, T6d, T6g, T66, T4L, T5I, T69, T5y, T4o, T8J, T8L, T5M; V T5Q, T5A, T5w, T5H, T4W, T5O, T5G, T8D, T8F; { V T5C, T3S, T8C, T4n, T8H, T8B, T8I, T5F, T5L, T5k, T5K, T5v, T4V; { V T5D, T47, T4m, T5E, T8z, T8A, T3C, T3R, T5j, T5u, T4K; T5S = VSUB(T3w, T3B); T3C = VADD(T3w, T3B); T3R = VADD(T3J, T3Q); T8O = VSUB(T3Q, T3J); { V T8o, T7B, T7x, T8q; T8o = VADD(T7z, T7A); T7B = VSUB(T7z, T7A); T7x = VADD(T7t, T7w); T8q = VSUB(T7w, T7t); { V T8u, T7l, T7h, T8w; T8u = VSUB(T7k, T7j); T7l = VADD(T7j, T7k); T7h = VSUB(T75, T7g); T8w = VADD(T75, T7g); ST(&(ri[WS(rs, 10)]), VFMA(LDK(KP923879532), T7B, T7y), ms, &(ri[0])); ST(&(ri[WS(rs, 26)]), VFNMS(LDK(KP923879532), T7B, T7y), ms, &(ri[0])); ST(&(ii[WS(rs, 18)]), VFNMS(LDK(KP923879532), T8o, T8n), ms, &(ii[0])); ST(&(ii[WS(rs, 2)]), VFMA(LDK(KP923879532), T8o, T8n), ms, &(ii[0])); ST(&(ii[WS(rs, 26)]), VFNMS(LDK(KP923879532), T8q, T8p), ms, &(ii[0])); ST(&(ii[WS(rs, 10)]), VFMA(LDK(KP923879532), T8q, T8p), ms, &(ii[0])); ST(&(ri[WS(rs, 2)]), VFMA(LDK(KP923879532), T7x, T7q), ms, &(ri[0])); ST(&(ri[WS(rs, 18)]), VFNMS(LDK(KP923879532), T7x, T7q), ms, &(ri[0])); ST(&(ri[WS(rs, 30)]), VFMA(LDK(KP923879532), T7l, T7i), ms, &(ri[0])); ST(&(ri[WS(rs, 14)]), VFNMS(LDK(KP923879532), T7l, T7i), ms, &(ri[0])); ST(&(ii[WS(rs, 22)]), VFNMS(LDK(KP923879532), T8u, T8t), ms, &(ii[0])); ST(&(ii[WS(rs, 6)]), VFMA(LDK(KP923879532), T8u, T8t), ms, &(ii[0])); ST(&(ii[WS(rs, 30)]), VFMA(LDK(KP923879532), T8w, T8v), ms, &(ii[0])); ST(&(ii[WS(rs, 14)]), VFNMS(LDK(KP923879532), T8w, T8v), ms, &(ii[0])); ST(&(ri[WS(rs, 6)]), VFMA(LDK(KP923879532), T7h, T6U), ms, &(ri[0])); ST(&(ri[WS(rs, 22)]), VFNMS(LDK(KP923879532), T7h, T6U), ms, &(ri[0])); T5C = VFMA(LDK(KP707106781), T3R, T3C); T3S = VFNMS(LDK(KP707106781), T3R, T3C); } } T5D = VFMA(LDK(KP414213562), T3Z, T46); T47 = VFNMS(LDK(KP414213562), T46, T3Z); T4m = VFMA(LDK(KP414213562), T4l, T4e); T5E = VFNMS(LDK(KP414213562), T4e, T4l); T8N = VADD(T8y, T8x); T8z = VSUB(T8x, T8y); T8A = VADD(T5T, T5U); T5V = VSUB(T5T, T5U); T6d = VSUB(T5i, T5b); T5j = VADD(T5b, T5i); T5u = VADD(T5s, T5t); T6g = VSUB(T5s, T5t); T66 = VSUB(T4J, T4C); T4K = VADD(T4C, T4J); T8C = VADD(T47, T4m); T4n = VSUB(T47, T4m); T8H = VFNMS(LDK(KP707106781), T8A, T8z); T8B = VFMA(LDK(KP707106781), T8A, T8z); T8I = VSUB(T5E, T5D); T5F = VADD(T5D, T5E); T5L = VFMA(LDK(KP707106781), T5j, T54); T5k = VFNMS(LDK(KP707106781), T5j, T54); T5K = VFMA(LDK(KP707106781), T5u, T5r); T5v = VFNMS(LDK(KP707106781), T5u, T5r); T4L = VFNMS(LDK(KP707106781), T4K, T4v); T5I = VFMA(LDK(KP707106781), T4K, T4v); T4V = VADD(T4T, T4U); T69 = VSUB(T4T, T4U); } T5y = VFNMS(LDK(KP923879532), T4n, T3S); T4o = VFMA(LDK(KP923879532), T4n, T3S); T8J = VFMA(LDK(KP923879532), T8I, T8H); T8L = VFNMS(LDK(KP923879532), T8I, T8H); T5M = VFNMS(LDK(KP198912367), T5L, T5K); T5Q = VFMA(LDK(KP198912367), T5K, T5L); T5A = VFMA(LDK(KP668178637), T5k, T5v); T5w = VFNMS(LDK(KP668178637), T5v, T5k); T5H = VFMA(LDK(KP707106781), T4V, T4S); T4W = VFNMS(LDK(KP707106781), T4V, T4S); T5O = VFNMS(LDK(KP923879532), T5F, T5C); T5G = VFMA(LDK(KP923879532), T5F, T5C); T8D = VFMA(LDK(KP923879532), T8C, T8B); T8F = VFNMS(LDK(KP923879532), T8C, T8B); } { V T6p, T6q, T6o, T5W, T8W, T63; { V T5J, T5P, T5z, T4X, T5Z, T62; T5J = VFMA(LDK(KP198912367), T5I, T5H); T5P = VFNMS(LDK(KP198912367), T5H, T5I); T5z = VFNMS(LDK(KP668178637), T4L, T4W); T4X = VFMA(LDK(KP668178637), T4W, T4L); T6p = VFNMS(LDK(KP414213562), T5X, T5Y); T5Z = VFMA(LDK(KP414213562), T5Y, T5X); T62 = VFNMS(LDK(KP414213562), T61, T60); T6q = VFMA(LDK(KP414213562), T60, T61); { V T8G, T5N, T5R, T8E; T8G = VSUB(T5M, T5J); T5N = VADD(T5J, T5M); T5R = VSUB(T5P, T5Q); T8E = VADD(T5P, T5Q); { V T5B, T8K, T8M, T5x; T5B = VADD(T5z, T5A); T8K = VSUB(T5A, T5z); T8M = VADD(T4X, T5w); T5x = VSUB(T4X, T5w); T6o = VFNMS(LDK(KP707106781), T5V, T5S); T5W = VFMA(LDK(KP707106781), T5V, T5S); T8W = VADD(T5Z, T62); T63 = VSUB(T5Z, T62); ST(&(ii[WS(rs, 25)]), VFNMS(LDK(KP980785280), T8G, T8F), ms, &(ii[WS(rs, 1)])); ST(&(ii[WS(rs, 9)]), VFMA(LDK(KP980785280), T8G, T8F), ms, &(ii[WS(rs, 1)])); ST(&(ri[WS(rs, 1)]), VFMA(LDK(KP980785280), T5N, T5G), ms, &(ri[WS(rs, 1)])); ST(&(ri[WS(rs, 17)]), VFNMS(LDK(KP980785280), T5N, T5G), ms, &(ri[WS(rs, 1)])); ST(&(ri[WS(rs, 9)]), VFMA(LDK(KP980785280), T5R, T5O), ms, &(ri[WS(rs, 1)])); ST(&(ri[WS(rs, 25)]), VFNMS(LDK(KP980785280), T5R, T5O), ms, &(ri[WS(rs, 1)])); ST(&(ii[WS(rs, 17)]), VFNMS(LDK(KP980785280), T8E, T8D), ms, &(ii[WS(rs, 1)])); ST(&(ii[WS(rs, 1)]), VFMA(LDK(KP980785280), T8E, T8D), ms, &(ii[WS(rs, 1)])); ST(&(ri[WS(rs, 29)]), VFMA(LDK(KP831469612), T5B, T5y), ms, &(ri[WS(rs, 1)])); ST(&(ri[WS(rs, 13)]), VFNMS(LDK(KP831469612), T5B, T5y), ms, &(ri[WS(rs, 1)])); ST(&(ii[WS(rs, 21)]), VFNMS(LDK(KP831469612), T8K, T8J), ms, &(ii[WS(rs, 1)])); ST(&(ii[WS(rs, 5)]), VFMA(LDK(KP831469612), T8K, T8J), ms, &(ii[WS(rs, 1)])); ST(&(ii[WS(rs, 29)]), VFMA(LDK(KP831469612), T8M, T8L), ms, &(ii[WS(rs, 1)])); ST(&(ii[WS(rs, 13)]), VFNMS(LDK(KP831469612), T8M, T8L), ms, &(ii[WS(rs, 1)])); ST(&(ri[WS(rs, 5)]), VFMA(LDK(KP831469612), T5x, T4o), ms, &(ri[WS(rs, 1)])); ST(&(ri[WS(rs, 21)]), VFNMS(LDK(KP831469612), T5x, T4o), ms, &(ri[WS(rs, 1)])); } } } { V T6k, T64, T8V, T6r, T8R, T8T, T6y, T6C, T6m, T6i, T6v, T6B, T6l, T6b, T6A; V T6s, T8X; { V T6x, T6e, T6w, T6h, T6u, T67, T6t, T6a, T8P, T8Q; T6k = VFNMS(LDK(KP923879532), T63, T5W); T64 = VFMA(LDK(KP923879532), T63, T5W); T8V = VFNMS(LDK(KP707106781), T8O, T8N); T8P = VFMA(LDK(KP707106781), T8O, T8N); T8Q = VSUB(T6q, T6p); T6r = VADD(T6p, T6q); T6x = VFMA(LDK(KP707106781), T6d, T6c); T6e = VFNMS(LDK(KP707106781), T6d, T6c); T6w = VFMA(LDK(KP707106781), T6g, T6f); T6h = VFNMS(LDK(KP707106781), T6g, T6f); T6u = VFMA(LDK(KP707106781), T66, T65); T67 = VFNMS(LDK(KP707106781), T66, T65); T6t = VFMA(LDK(KP707106781), T69, T68); T6a = VFNMS(LDK(KP707106781), T69, T68); T8R = VFMA(LDK(KP923879532), T8Q, T8P); T8T = VFNMS(LDK(KP923879532), T8Q, T8P); T6y = VFNMS(LDK(KP198912367), T6x, T6w); T6C = VFMA(LDK(KP198912367), T6w, T6x); T6m = VFMA(LDK(KP668178637), T6e, T6h); T6i = VFNMS(LDK(KP668178637), T6h, T6e); T6v = VFMA(LDK(KP198912367), T6u, T6t); T6B = VFNMS(LDK(KP198912367), T6t, T6u); T6l = VFNMS(LDK(KP668178637), T67, T6a); T6b = VFMA(LDK(KP668178637), T6a, T67); } T6A = VFMA(LDK(KP923879532), T6r, T6o); T6s = VFNMS(LDK(KP923879532), T6r, T6o); T8X = VFNMS(LDK(KP923879532), T8W, T8V); T8Z = VFMA(LDK(KP923879532), T8W, T8V); { V T6z, T6D, T8Y, T6n, T8S, T8U, T6j; T6z = VSUB(T6v, T6y); T90 = VADD(T6v, T6y); T6D = VADD(T6B, T6C); T8Y = VSUB(T6C, T6B); T6n = VSUB(T6l, T6m); T8S = VADD(T6l, T6m); T8U = VSUB(T6i, T6b); T6j = VADD(T6b, T6i); ST(&(ri[WS(rs, 7)]), VFMA(LDK(KP980785280), T6z, T6s), ms, &(ri[WS(rs, 1)])); ST(&(ri[WS(rs, 23)]), VFNMS(LDK(KP980785280), T6z, T6s), ms, &(ri[WS(rs, 1)])); ST(&(ii[WS(rs, 23)]), VFNMS(LDK(KP980785280), T8Y, T8X), ms, &(ii[WS(rs, 1)])); ST(&(ii[WS(rs, 7)]), VFMA(LDK(KP980785280), T8Y, T8X), ms, &(ii[WS(rs, 1)])); ST(&(ri[WS(rs, 11)]), VFMA(LDK(KP831469612), T6n, T6k), ms, &(ri[WS(rs, 1)])); ST(&(ri[WS(rs, 27)]), VFNMS(LDK(KP831469612), T6n, T6k), ms, &(ri[WS(rs, 1)])); ST(&(ii[WS(rs, 19)]), VFNMS(LDK(KP831469612), T8S, T8R), ms, &(ii[WS(rs, 1)])); ST(&(ii[WS(rs, 3)]), VFMA(LDK(KP831469612), T8S, T8R), ms, &(ii[WS(rs, 1)])); ST(&(ii[WS(rs, 27)]), VFNMS(LDK(KP831469612), T8U, T8T), ms, &(ii[WS(rs, 1)])); ST(&(ii[WS(rs, 11)]), VFMA(LDK(KP831469612), T8U, T8T), ms, &(ii[WS(rs, 1)])); ST(&(ri[WS(rs, 3)]), VFMA(LDK(KP831469612), T6j, T64), ms, &(ri[WS(rs, 1)])); ST(&(ri[WS(rs, 19)]), VFNMS(LDK(KP831469612), T6j, T64), ms, &(ri[WS(rs, 1)])); ST(&(ri[WS(rs, 31)]), VFMA(LDK(KP980785280), T6D, T6A), ms, &(ri[WS(rs, 1)])); ST(&(ri[WS(rs, 15)]), VFNMS(LDK(KP980785280), T6D, T6A), ms, &(ri[WS(rs, 1)])); } } } } } } ST(&(ii[WS(rs, 31)]), VFMA(LDK(KP980785280), T90, T8Z), ms, &(ii[WS(rs, 1)])); ST(&(ii[WS(rs, 15)]), VFNMS(LDK(KP980785280), T90, T8Z), ms, &(ii[WS(rs, 1)])); } } VLEAVE(); } static const tw_instr twinstr[] = { VTW(0, 1), VTW(0, 2), VTW(0, 3), VTW(0, 4), VTW(0, 5), VTW(0, 6), VTW(0, 7), VTW(0, 8), VTW(0, 9), VTW(0, 10), VTW(0, 11), VTW(0, 12), VTW(0, 13), VTW(0, 14), VTW(0, 15), VTW(0, 16), VTW(0, 17), VTW(0, 18), VTW(0, 19), VTW(0, 20), VTW(0, 21), VTW(0, 22), VTW(0, 23), VTW(0, 24), VTW(0, 25), VTW(0, 26), VTW(0, 27), VTW(0, 28), VTW(0, 29), VTW(0, 30), VTW(0, 31), {TW_NEXT, (2 * VL), 0} }; static const ct_desc desc = { 32, XSIMD_STRING("t1sv_32"), twinstr, &GENUS, {236, 62, 198, 0}, 0, 0, 0 }; void XSIMD(codelet_t1sv_32) (planner *p) { X(kdft_dit_register) (p, t1sv_32, &desc); } #else /* HAVE_FMA */ /* Generated by: ../../../genfft/gen_twiddle.native -simd -compact -variables 4 -pipeline-latency 8 -n 32 -name t1sv_32 -include ts.h */ /* * This function contains 434 FP additions, 208 FP multiplications, * (or, 340 additions, 114 multiplications, 94 fused multiply/add), * 96 stack variables, 7 constants, and 128 memory accesses */ #include "ts.h" static void t1sv_32(R *ri, R *ii, const R *W, stride rs, INT mb, INT me, INT ms) { DVK(KP195090322, +0.195090322016128267848284868477022240927691618); DVK(KP980785280, +0.980785280403230449126182236134239036973933731); DVK(KP555570233, +0.555570233019602224742830813948532874374937191); DVK(KP831469612, +0.831469612302545237078788377617905756738560812); DVK(KP382683432, +0.382683432365089771728459984030398866761344562); DVK(KP923879532, +0.923879532511286756128183189396788286822416626); DVK(KP707106781, +0.707106781186547524400844362104849039284835938); { INT m; for (m = mb, W = W + (mb * 62); m < me; m = m + (2 * VL), ri = ri + ((2 * VL) * ms), ii = ii + ((2 * VL) * ms), W = W + ((2 * VL) * 62), MAKE_VOLATILE_STRIDE(64, rs)) { V Tj, T5F, T7C, T7Q, T35, T4T, T78, T7m, T1Q, T61, T5Y, T6J, T3K, T59, T41; V T56, T2B, T67, T6e, T6O, T4b, T5d, T4s, T5g, TG, T7l, T5I, T73, T3a, T4U; V T3f, T4V, T14, T5N, T5M, T6E, T3m, T4Y, T3r, T4Z, T1r, T5P, T5S, T6F, T3x; V T51, T3C, T52, T2d, T5Z, T64, T6K, T3V, T57, T44, T5a, T2Y, T6f, T6a, T6P; V T4m, T5h, T4v, T5e; { V T1, T76, T6, T75, Tc, T32, Th, T33; T1 = LD(&(ri[0]), ms, &(ri[0])); T76 = LD(&(ii[0]), ms, &(ii[0])); { V T3, T5, T2, T4; T3 = LD(&(ri[WS(rs, 16)]), ms, &(ri[0])); T5 = LD(&(ii[WS(rs, 16)]), ms, &(ii[0])); T2 = LDW(&(W[TWVL * 30])); T4 = LDW(&(W[TWVL * 31])); T6 = VFMA(T2, T3, VMUL(T4, T5)); T75 = VFNMS(T4, T3, VMUL(T2, T5)); } { V T9, Tb, T8, Ta; T9 = LD(&(ri[WS(rs, 8)]), ms, &(ri[0])); Tb = LD(&(ii[WS(rs, 8)]), ms, &(ii[0])); T8 = LDW(&(W[TWVL * 14])); Ta = LDW(&(W[TWVL * 15])); Tc = VFMA(T8, T9, VMUL(Ta, Tb)); T32 = VFNMS(Ta, T9, VMUL(T8, Tb)); } { V Te, Tg, Td, Tf; Te = LD(&(ri[WS(rs, 24)]), ms, &(ri[0])); Tg = LD(&(ii[WS(rs, 24)]), ms, &(ii[0])); Td = LDW(&(W[TWVL * 46])); Tf = LDW(&(W[TWVL * 47])); Th = VFMA(Td, Te, VMUL(Tf, Tg)); T33 = VFNMS(Tf, Te, VMUL(Td, Tg)); } { V T7, Ti, T7A, T7B; T7 = VADD(T1, T6); Ti = VADD(Tc, Th); Tj = VADD(T7, Ti); T5F = VSUB(T7, Ti); T7A = VSUB(T76, T75); T7B = VSUB(Tc, Th); T7C = VSUB(T7A, T7B); T7Q = VADD(T7B, T7A); } { V T31, T34, T74, T77; T31 = VSUB(T1, T6); T34 = VSUB(T32, T33); T35 = VSUB(T31, T34); T4T = VADD(T31, T34); T74 = VADD(T32, T33); T77 = VADD(T75, T76); T78 = VADD(T74, T77); T7m = VSUB(T77, T74); } } { V T1y, T3G, T1O, T3Z, T1D, T3H, T1J, T3Y; { V T1v, T1x, T1u, T1w; T1v = LD(&(ri[WS(rs, 1)]), ms, &(ri[WS(rs, 1)])); T1x = LD(&(ii[WS(rs, 1)]), ms, &(ii[WS(rs, 1)])); T1u = LDW(&(W[0])); T1w = LDW(&(W[TWVL * 1])); T1y = VFMA(T1u, T1v, VMUL(T1w, T1x)); T3G = VFNMS(T1w, T1v, VMUL(T1u, T1x)); } { V T1L, T1N, T1K, T1M; T1L = LD(&(ri[WS(rs, 25)]), ms, &(ri[WS(rs, 1)])); T1N = LD(&(ii[WS(rs, 25)]), ms, &(ii[WS(rs, 1)])); T1K = LDW(&(W[TWVL * 48])); T1M = LDW(&(W[TWVL * 49])); T1O = VFMA(T1K, T1L, VMUL(T1M, T1N)); T3Z = VFNMS(T1M, T1L, VMUL(T1K, T1N)); } { V T1A, T1C, T1z, T1B; T1A = LD(&(ri[WS(rs, 17)]), ms, &(ri[WS(rs, 1)])); T1C = LD(&(ii[WS(rs, 17)]), ms, &(ii[WS(rs, 1)])); T1z = LDW(&(W[TWVL * 32])); T1B = LDW(&(W[TWVL * 33])); T1D = VFMA(T1z, T1A, VMUL(T1B, T1C)); T3H = VFNMS(T1B, T1A, VMUL(T1z, T1C)); } { V T1G, T1I, T1F, T1H; T1G = LD(&(ri[WS(rs, 9)]), ms, &(ri[WS(rs, 1)])); T1I = LD(&(ii[WS(rs, 9)]), ms, &(ii[WS(rs, 1)])); T1F = LDW(&(W[TWVL * 16])); T1H = LDW(&(W[TWVL * 17])); T1J = VFMA(T1F, T1G, VMUL(T1H, T1I)); T3Y = VFNMS(T1H, T1G, VMUL(T1F, T1I)); } { V T1E, T1P, T5W, T5X; T1E = VADD(T1y, T1D); T1P = VADD(T1J, T1O); T1Q = VADD(T1E, T1P); T61 = VSUB(T1E, T1P); T5W = VADD(T3G, T3H); T5X = VADD(T3Y, T3Z); T5Y = VSUB(T5W, T5X); T6J = VADD(T5W, T5X); } { V T3I, T3J, T3X, T40; T3I = VSUB(T3G, T3H); T3J = VSUB(T1J, T1O); T3K = VADD(T3I, T3J); T59 = VSUB(T3I, T3J); T3X = VSUB(T1y, T1D); T40 = VSUB(T3Y, T3Z); T41 = VSUB(T3X, T40); T56 = VADD(T3X, T40); } } { V T2j, T4o, T2z, T49, T2o, T4p, T2u, T48; { V T2g, T2i, T2f, T2h; T2g = LD(&(ri[WS(rs, 31)]), ms, &(ri[WS(rs, 1)])); T2i = LD(&(ii[WS(rs, 31)]), ms, &(ii[WS(rs, 1)])); T2f = LDW(&(W[TWVL * 60])); T2h = LDW(&(W[TWVL * 61])); T2j = VFMA(T2f, T2g, VMUL(T2h, T2i)); T4o = VFNMS(T2h, T2g, VMUL(T2f, T2i)); } { V T2w, T2y, T2v, T2x; T2w = LD(&(ri[WS(rs, 23)]), ms, &(ri[WS(rs, 1)])); T2y = LD(&(ii[WS(rs, 23)]), ms, &(ii[WS(rs, 1)])); T2v = LDW(&(W[TWVL * 44])); T2x = LDW(&(W[TWVL * 45])); T2z = VFMA(T2v, T2w, VMUL(T2x, T2y)); T49 = VFNMS(T2x, T2w, VMUL(T2v, T2y)); } { V T2l, T2n, T2k, T2m; T2l = LD(&(ri[WS(rs, 15)]), ms, &(ri[WS(rs, 1)])); T2n = LD(&(ii[WS(rs, 15)]), ms, &(ii[WS(rs, 1)])); T2k = LDW(&(W[TWVL * 28])); T2m = LDW(&(W[TWVL * 29])); T2o = VFMA(T2k, T2l, VMUL(T2m, T2n)); T4p = VFNMS(T2m, T2l, VMUL(T2k, T2n)); } { V T2r, T2t, T2q, T2s; T2r = LD(&(ri[WS(rs, 7)]), ms, &(ri[WS(rs, 1)])); T2t = LD(&(ii[WS(rs, 7)]), ms, &(ii[WS(rs, 1)])); T2q = LDW(&(W[TWVL * 12])); T2s = LDW(&(W[TWVL * 13])); T2u = VFMA(T2q, T2r, VMUL(T2s, T2t)); T48 = VFNMS(T2s, T2r, VMUL(T2q, T2t)); } { V T2p, T2A, T6c, T6d; T2p = VADD(T2j, T2o); T2A = VADD(T2u, T2z); T2B = VADD(T2p, T2A); T67 = VSUB(T2p, T2A); T6c = VADD(T4o, T4p); T6d = VADD(T48, T49); T6e = VSUB(T6c, T6d); T6O = VADD(T6c, T6d); } { V T47, T4a, T4q, T4r; T47 = VSUB(T2j, T2o); T4a = VSUB(T48, T49); T4b = VSUB(T47, T4a); T5d = VADD(T47, T4a); T4q = VSUB(T4o, T4p); T4r = VSUB(T2u, T2z); T4s = VADD(T4q, T4r); T5g = VSUB(T4q, T4r); } } { V To, T36, TE, T3d, Tt, T37, Tz, T3c; { V Tl, Tn, Tk, Tm; Tl = LD(&(ri[WS(rs, 4)]), ms, &(ri[0])); Tn = LD(&(ii[WS(rs, 4)]), ms, &(ii[0])); Tk = LDW(&(W[TWVL * 6])); Tm = LDW(&(W[TWVL * 7])); To = VFMA(Tk, Tl, VMUL(Tm, Tn)); T36 = VFNMS(Tm, Tl, VMUL(Tk, Tn)); } { V TB, TD, TA, TC; TB = LD(&(ri[WS(rs, 12)]), ms, &(ri[0])); TD = LD(&(ii[WS(rs, 12)]), ms, &(ii[0])); TA = LDW(&(W[TWVL * 22])); TC = LDW(&(W[TWVL * 23])); TE = VFMA(TA, TB, VMUL(TC, TD)); T3d = VFNMS(TC, TB, VMUL(TA, TD)); } { V Tq, Ts, Tp, Tr; Tq = LD(&(ri[WS(rs, 20)]), ms, &(ri[0])); Ts = LD(&(ii[WS(rs, 20)]), ms, &(ii[0])); Tp = LDW(&(W[TWVL * 38])); Tr = LDW(&(W[TWVL * 39])); Tt = VFMA(Tp, Tq, VMUL(Tr, Ts)); T37 = VFNMS(Tr, Tq, VMUL(Tp, Ts)); } { V Tw, Ty, Tv, Tx; Tw = LD(&(ri[WS(rs, 28)]), ms, &(ri[0])); Ty = LD(&(ii[WS(rs, 28)]), ms, &(ii[0])); Tv = LDW(&(W[TWVL * 54])); Tx = LDW(&(W[TWVL * 55])); Tz = VFMA(Tv, Tw, VMUL(Tx, Ty)); T3c = VFNMS(Tx, Tw, VMUL(Tv, Ty)); } { V Tu, TF, T5G, T5H; Tu = VADD(To, Tt); TF = VADD(Tz, TE); TG = VADD(Tu, TF); T7l = VSUB(TF, Tu); T5G = VADD(T36, T37); T5H = VADD(T3c, T3d); T5I = VSUB(T5G, T5H); T73 = VADD(T5G, T5H); } { V T38, T39, T3b, T3e; T38 = VSUB(T36, T37); T39 = VSUB(To, Tt); T3a = VSUB(T38, T39); T4U = VADD(T39, T38); T3b = VSUB(Tz, TE); T3e = VSUB(T3c, T3d); T3f = VADD(T3b, T3e); T4V = VSUB(T3b, T3e); } } { V TM, T3i, T12, T3p, TR, T3j, TX, T3o; { V TJ, TL, TI, TK; TJ = LD(&(ri[WS(rs, 2)]), ms, &(ri[0])); TL = LD(&(ii[WS(rs, 2)]), ms, &(ii[0])); TI = LDW(&(W[TWVL * 2])); TK = LDW(&(W[TWVL * 3])); TM = VFMA(TI, TJ, VMUL(TK, TL)); T3i = VFNMS(TK, TJ, VMUL(TI, TL)); } { V TZ, T11, TY, T10; TZ = LD(&(ri[WS(rs, 26)]), ms, &(ri[0])); T11 = LD(&(ii[WS(rs, 26)]), ms, &(ii[0])); TY = LDW(&(W[TWVL * 50])); T10 = LDW(&(W[TWVL * 51])); T12 = VFMA(TY, TZ, VMUL(T10, T11)); T3p = VFNMS(T10, TZ, VMUL(TY, T11)); } { V TO, TQ, TN, TP; TO = LD(&(ri[WS(rs, 18)]), ms, &(ri[0])); TQ = LD(&(ii[WS(rs, 18)]), ms, &(ii[0])); TN = LDW(&(W[TWVL * 34])); TP = LDW(&(W[TWVL * 35])); TR = VFMA(TN, TO, VMUL(TP, TQ)); T3j = VFNMS(TP, TO, VMUL(TN, TQ)); } { V TU, TW, TT, TV; TU = LD(&(ri[WS(rs, 10)]), ms, &(ri[0])); TW = LD(&(ii[WS(rs, 10)]), ms, &(ii[0])); TT = LDW(&(W[TWVL * 18])); TV = LDW(&(W[TWVL * 19])); TX = VFMA(TT, TU, VMUL(TV, TW)); T3o = VFNMS(TV, TU, VMUL(TT, TW)); } { V TS, T13, T5K, T5L; TS = VADD(TM, TR); T13 = VADD(TX, T12); T14 = VADD(TS, T13); T5N = VSUB(TS, T13); T5K = VADD(T3i, T3j); T5L = VADD(T3o, T3p); T5M = VSUB(T5K, T5L); T6E = VADD(T5K, T5L); } { V T3k, T3l, T3n, T3q; T3k = VSUB(T3i, T3j); T3l = VSUB(TX, T12); T3m = VADD(T3k, T3l); T4Y = VSUB(T3k, T3l); T3n = VSUB(TM, TR); T3q = VSUB(T3o, T3p); T3r = VSUB(T3n, T3q); T4Z = VADD(T3n, T3q); } } { V T19, T3t, T1p, T3A, T1e, T3u, T1k, T3z; { V T16, T18, T15, T17; T16 = LD(&(ri[WS(rs, 30)]), ms, &(ri[0])); T18 = LD(&(ii[WS(rs, 30)]), ms, &(ii[0])); T15 = LDW(&(W[TWVL * 58])); T17 = LDW(&(W[TWVL * 59])); T19 = VFMA(T15, T16, VMUL(T17, T18)); T3t = VFNMS(T17, T16, VMUL(T15, T18)); } { V T1m, T1o, T1l, T1n; T1m = LD(&(ri[WS(rs, 22)]), ms, &(ri[0])); T1o = LD(&(ii[WS(rs, 22)]), ms, &(ii[0])); T1l = LDW(&(W[TWVL * 42])); T1n = LDW(&(W[TWVL * 43])); T1p = VFMA(T1l, T1m, VMUL(T1n, T1o)); T3A = VFNMS(T1n, T1m, VMUL(T1l, T1o)); } { V T1b, T1d, T1a, T1c; T1b = LD(&(ri[WS(rs, 14)]), ms, &(ri[0])); T1d = LD(&(ii[WS(rs, 14)]), ms, &(ii[0])); T1a = LDW(&(W[TWVL * 26])); T1c = LDW(&(W[TWVL * 27])); T1e = VFMA(T1a, T1b, VMUL(T1c, T1d)); T3u = VFNMS(T1c, T1b, VMUL(T1a, T1d)); } { V T1h, T1j, T1g, T1i; T1h = LD(&(ri[WS(rs, 6)]), ms, &(ri[0])); T1j = LD(&(ii[WS(rs, 6)]), ms, &(ii[0])); T1g = LDW(&(W[TWVL * 10])); T1i = LDW(&(W[TWVL * 11])); T1k = VFMA(T1g, T1h, VMUL(T1i, T1j)); T3z = VFNMS(T1i, T1h, VMUL(T1g, T1j)); } { V T1f, T1q, T5Q, T5R; T1f = VADD(T19, T1e); T1q = VADD(T1k, T1p); T1r = VADD(T1f, T1q); T5P = VSUB(T1f, T1q); T5Q = VADD(T3t, T3u); T5R = VADD(T3z, T3A); T5S = VSUB(T5Q, T5R); T6F = VADD(T5Q, T5R); } { V T3v, T3w, T3y, T3B; T3v = VSUB(T3t, T3u); T3w = VSUB(T1k, T1p); T3x = VADD(T3v, T3w); T51 = VSUB(T3v, T3w); T3y = VSUB(T19, T1e); T3B = VSUB(T3z, T3A); T3C = VSUB(T3y, T3B); T52 = VADD(T3y, T3B); } } { V T1V, T3R, T20, T3S, T3Q, T3T, T26, T3M, T2b, T3N, T3L, T3O; { V T1S, T1U, T1R, T1T; T1S = LD(&(ri[WS(rs, 5)]), ms, &(ri[WS(rs, 1)])); T1U = LD(&(ii[WS(rs, 5)]), ms, &(ii[WS(rs, 1)])); T1R = LDW(&(W[TWVL * 8])); T1T = LDW(&(W[TWVL * 9])); T1V = VFMA(T1R, T1S, VMUL(T1T, T1U)); T3R = VFNMS(T1T, T1S, VMUL(T1R, T1U)); } { V T1X, T1Z, T1W, T1Y; T1X = LD(&(ri[WS(rs, 21)]), ms, &(ri[WS(rs, 1)])); T1Z = LD(&(ii[WS(rs, 21)]), ms, &(ii[WS(rs, 1)])); T1W = LDW(&(W[TWVL * 40])); T1Y = LDW(&(W[TWVL * 41])); T20 = VFMA(T1W, T1X, VMUL(T1Y, T1Z)); T3S = VFNMS(T1Y, T1X, VMUL(T1W, T1Z)); } T3Q = VSUB(T1V, T20); T3T = VSUB(T3R, T3S); { V T23, T25, T22, T24; T23 = LD(&(ri[WS(rs, 29)]), ms, &(ri[WS(rs, 1)])); T25 = LD(&(ii[WS(rs, 29)]), ms, &(ii[WS(rs, 1)])); T22 = LDW(&(W[TWVL * 56])); T24 = LDW(&(W[TWVL * 57])); T26 = VFMA(T22, T23, VMUL(T24, T25)); T3M = VFNMS(T24, T23, VMUL(T22, T25)); } { V T28, T2a, T27, T29; T28 = LD(&(ri[WS(rs, 13)]), ms, &(ri[WS(rs, 1)])); T2a = LD(&(ii[WS(rs, 13)]), ms, &(ii[WS(rs, 1)])); T27 = LDW(&(W[TWVL * 24])); T29 = LDW(&(W[TWVL * 25])); T2b = VFMA(T27, T28, VMUL(T29, T2a)); T3N = VFNMS(T29, T28, VMUL(T27, T2a)); } T3L = VSUB(T26, T2b); T3O = VSUB(T3M, T3N); { V T21, T2c, T62, T63; T21 = VADD(T1V, T20); T2c = VADD(T26, T2b); T2d = VADD(T21, T2c); T5Z = VSUB(T2c, T21); T62 = VADD(T3R, T3S); T63 = VADD(T3M, T3N); T64 = VSUB(T62, T63); T6K = VADD(T62, T63); } { V T3P, T3U, T42, T43; T3P = VSUB(T3L, T3O); T3U = VADD(T3Q, T3T); T3V = VMUL(LDK(KP707106781), VSUB(T3P, T3U)); T57 = VMUL(LDK(KP707106781), VADD(T3U, T3P)); T42 = VSUB(T3T, T3Q); T43 = VADD(T3L, T3O); T44 = VMUL(LDK(KP707106781), VSUB(T42, T43)); T5a = VMUL(LDK(KP707106781), VADD(T42, T43)); } } { V T2G, T4c, T2L, T4d, T4e, T4f, T2R, T4i, T2W, T4j, T4h, T4k; { V T2D, T2F, T2C, T2E; T2D = LD(&(ri[WS(rs, 3)]), ms, &(ri[WS(rs, 1)])); T2F = LD(&(ii[WS(rs, 3)]), ms, &(ii[WS(rs, 1)])); T2C = LDW(&(W[TWVL * 4])); T2E = LDW(&(W[TWVL * 5])); T2G = VFMA(T2C, T2D, VMUL(T2E, T2F)); T4c = VFNMS(T2E, T2D, VMUL(T2C, T2F)); } { V T2I, T2K, T2H, T2J; T2I = LD(&(ri[WS(rs, 19)]), ms, &(ri[WS(rs, 1)])); T2K = LD(&(ii[WS(rs, 19)]), ms, &(ii[WS(rs, 1)])); T2H = LDW(&(W[TWVL * 36])); T2J = LDW(&(W[TWVL * 37])); T2L = VFMA(T2H, T2I, VMUL(T2J, T2K)); T4d = VFNMS(T2J, T2I, VMUL(T2H, T2K)); } T4e = VSUB(T4c, T4d); T4f = VSUB(T2G, T2L); { V T2O, T2Q, T2N, T2P; T2O = LD(&(ri[WS(rs, 27)]), ms, &(ri[WS(rs, 1)])); T2Q = LD(&(ii[WS(rs, 27)]), ms, &(ii[WS(rs, 1)])); T2N = LDW(&(W[TWVL * 52])); T2P = LDW(&(W[TWVL * 53])); T2R = VFMA(T2N, T2O, VMUL(T2P, T2Q)); T4i = VFNMS(T2P, T2O, VMUL(T2N, T2Q)); } { V T2T, T2V, T2S, T2U; T2T = LD(&(ri[WS(rs, 11)]), ms, &(ri[WS(rs, 1)])); T2V = LD(&(ii[WS(rs, 11)]), ms, &(ii[WS(rs, 1)])); T2S = LDW(&(W[TWVL * 20])); T2U = LDW(&(W[TWVL * 21])); T2W = VFMA(T2S, T2T, VMUL(T2U, T2V)); T4j = VFNMS(T2U, T2T, VMUL(T2S, T2V)); } T4h = VSUB(T2R, T2W); T4k = VSUB(T4i, T4j); { V T2M, T2X, T68, T69; T2M = VADD(T2G, T2L); T2X = VADD(T2R, T2W); T2Y = VADD(T2M, T2X); T6f = VSUB(T2X, T2M); T68 = VADD(T4c, T4d); T69 = VADD(T4i, T4j); T6a = VSUB(T68, T69); T6P = VADD(T68, T69); } { V T4g, T4l, T4t, T4u; T4g = VSUB(T4e, T4f); T4l = VADD(T4h, T4k); T4m = VMUL(LDK(KP707106781), VSUB(T4g, T4l)); T5h = VMUL(LDK(KP707106781), VADD(T4g, T4l)); T4t = VSUB(T4h, T4k); T4u = VADD(T4f, T4e); T4v = VMUL(LDK(KP707106781), VSUB(T4t, T4u)); T5e = VMUL(LDK(KP707106781), VADD(T4u, T4t)); } } { V T1t, T6X, T7a, T7c, T30, T7b, T70, T71; { V TH, T1s, T72, T79; TH = VADD(Tj, TG); T1s = VADD(T14, T1r); T1t = VADD(TH, T1s); T6X = VSUB(TH, T1s); T72 = VADD(T6E, T6F); T79 = VADD(T73, T78); T7a = VADD(T72, T79); T7c = VSUB(T79, T72); } { V T2e, T2Z, T6Y, T6Z; T2e = VADD(T1Q, T2d); T2Z = VADD(T2B, T2Y); T30 = VADD(T2e, T2Z); T7b = VSUB(T2Z, T2e); T6Y = VADD(T6J, T6K); T6Z = VADD(T6O, T6P); T70 = VSUB(T6Y, T6Z); T71 = VADD(T6Y, T6Z); } ST(&(ri[WS(rs, 16)]), VSUB(T1t, T30), ms, &(ri[0])); ST(&(ii[WS(rs, 16)]), VSUB(T7a, T71), ms, &(ii[0])); ST(&(ri[0]), VADD(T1t, T30), ms, &(ri[0])); ST(&(ii[0]), VADD(T71, T7a), ms, &(ii[0])); ST(&(ri[WS(rs, 24)]), VSUB(T6X, T70), ms, &(ri[0])); ST(&(ii[WS(rs, 24)]), VSUB(T7c, T7b), ms, &(ii[0])); ST(&(ri[WS(rs, 8)]), VADD(T6X, T70), ms, &(ri[0])); ST(&(ii[WS(rs, 8)]), VADD(T7b, T7c), ms, &(ii[0])); } { V T6H, T6T, T7g, T7i, T6M, T6U, T6R, T6V; { V T6D, T6G, T7e, T7f; T6D = VSUB(Tj, TG); T6G = VSUB(T6E, T6F); T6H = VADD(T6D, T6G); T6T = VSUB(T6D, T6G); T7e = VSUB(T1r, T14); T7f = VSUB(T78, T73); T7g = VADD(T7e, T7f); T7i = VSUB(T7f, T7e); } { V T6I, T6L, T6N, T6Q; T6I = VSUB(T1Q, T2d); T6L = VSUB(T6J, T6K); T6M = VADD(T6I, T6L); T6U = VSUB(T6L, T6I); T6N = VSUB(T2B, T2Y); T6Q = VSUB(T6O, T6P); T6R = VSUB(T6N, T6Q); T6V = VADD(T6N, T6Q); } { V T6S, T7d, T6W, T7h; T6S = VMUL(LDK(KP707106781), VADD(T6M, T6R)); ST(&(ri[WS(rs, 20)]), VSUB(T6H, T6S), ms, &(ri[0])); ST(&(ri[WS(rs, 4)]), VADD(T6H, T6S), ms, &(ri[0])); T7d = VMUL(LDK(KP707106781), VADD(T6U, T6V)); ST(&(ii[WS(rs, 4)]), VADD(T7d, T7g), ms, &(ii[0])); ST(&(ii[WS(rs, 20)]), VSUB(T7g, T7d), ms, &(ii[0])); T6W = VMUL(LDK(KP707106781), VSUB(T6U, T6V)); ST(&(ri[WS(rs, 28)]), VSUB(T6T, T6W), ms, &(ri[0])); ST(&(ri[WS(rs, 12)]), VADD(T6T, T6W), ms, &(ri[0])); T7h = VMUL(LDK(KP707106781), VSUB(T6R, T6M)); ST(&(ii[WS(rs, 12)]), VADD(T7h, T7i), ms, &(ii[0])); ST(&(ii[WS(rs, 28)]), VSUB(T7i, T7h), ms, &(ii[0])); } } { V T5J, T7n, T7t, T6n, T5U, T7k, T6x, T6B, T6q, T7s, T66, T6k, T6u, T6A, T6h; V T6l; { V T5O, T5T, T60, T65; T5J = VSUB(T5F, T5I); T7n = VADD(T7l, T7m); T7t = VSUB(T7m, T7l); T6n = VADD(T5F, T5I); T5O = VSUB(T5M, T5N); T5T = VADD(T5P, T5S); T5U = VMUL(LDK(KP707106781), VSUB(T5O, T5T)); T7k = VMUL(LDK(KP707106781), VADD(T5O, T5T)); { V T6v, T6w, T6o, T6p; T6v = VADD(T67, T6a); T6w = VADD(T6e, T6f); T6x = VFNMS(LDK(KP382683432), T6w, VMUL(LDK(KP923879532), T6v)); T6B = VFMA(LDK(KP923879532), T6w, VMUL(LDK(KP382683432), T6v)); T6o = VADD(T5N, T5M); T6p = VSUB(T5P, T5S); T6q = VMUL(LDK(KP707106781), VADD(T6o, T6p)); T7s = VMUL(LDK(KP707106781), VSUB(T6p, T6o)); } T60 = VSUB(T5Y, T5Z); T65 = VSUB(T61, T64); T66 = VFMA(LDK(KP923879532), T60, VMUL(LDK(KP382683432), T65)); T6k = VFNMS(LDK(KP923879532), T65, VMUL(LDK(KP382683432), T60)); { V T6s, T6t, T6b, T6g; T6s = VADD(T5Y, T5Z); T6t = VADD(T61, T64); T6u = VFMA(LDK(KP382683432), T6s, VMUL(LDK(KP923879532), T6t)); T6A = VFNMS(LDK(KP382683432), T6t, VMUL(LDK(KP923879532), T6s)); T6b = VSUB(T67, T6a); T6g = VSUB(T6e, T6f); T6h = VFNMS(LDK(KP923879532), T6g, VMUL(LDK(KP382683432), T6b)); T6l = VFMA(LDK(KP382683432), T6g, VMUL(LDK(KP923879532), T6b)); } } { V T5V, T6i, T7r, T7u; T5V = VADD(T5J, T5U); T6i = VADD(T66, T6h); ST(&(ri[WS(rs, 22)]), VSUB(T5V, T6i), ms, &(ri[0])); ST(&(ri[WS(rs, 6)]), VADD(T5V, T6i), ms, &(ri[0])); T7r = VADD(T6k, T6l); T7u = VADD(T7s, T7t); ST(&(ii[WS(rs, 6)]), VADD(T7r, T7u), ms, &(ii[0])); ST(&(ii[WS(rs, 22)]), VSUB(T7u, T7r), ms, &(ii[0])); } { V T6j, T6m, T7v, T7w; T6j = VSUB(T5J, T5U); T6m = VSUB(T6k, T6l); ST(&(ri[WS(rs, 30)]), VSUB(T6j, T6m), ms, &(ri[0])); ST(&(ri[WS(rs, 14)]), VADD(T6j, T6m), ms, &(ri[0])); T7v = VSUB(T6h, T66); T7w = VSUB(T7t, T7s); ST(&(ii[WS(rs, 14)]), VADD(T7v, T7w), ms, &(ii[0])); ST(&(ii[WS(rs, 30)]), VSUB(T7w, T7v), ms, &(ii[0])); } { V T6r, T6y, T7j, T7o; T6r = VADD(T6n, T6q); T6y = VADD(T6u, T6x); ST(&(ri[WS(rs, 18)]), VSUB(T6r, T6y), ms, &(ri[0])); ST(&(ri[WS(rs, 2)]), VADD(T6r, T6y), ms, &(ri[0])); T7j = VADD(T6A, T6B); T7o = VADD(T7k, T7n); ST(&(ii[WS(rs, 2)]), VADD(T7j, T7o), ms, &(ii[0])); ST(&(ii[WS(rs, 18)]), VSUB(T7o, T7j), ms, &(ii[0])); } { V T6z, T6C, T7p, T7q; T6z = VSUB(T6n, T6q); T6C = VSUB(T6A, T6B); ST(&(ri[WS(rs, 26)]), VSUB(T6z, T6C), ms, &(ri[0])); ST(&(ri[WS(rs, 10)]), VADD(T6z, T6C), ms, &(ri[0])); T7p = VSUB(T6x, T6u); T7q = VSUB(T7n, T7k); ST(&(ii[WS(rs, 10)]), VADD(T7p, T7q), ms, &(ii[0])); ST(&(ii[WS(rs, 26)]), VSUB(T7q, T7p), ms, &(ii[0])); } } { V T3h, T4D, T7R, T7X, T3E, T7O, T4N, T4R, T46, T4A, T4G, T7W, T4K, T4Q, T4x; V T4B, T3g, T7P; T3g = VMUL(LDK(KP707106781), VSUB(T3a, T3f)); T3h = VSUB(T35, T3g); T4D = VADD(T35, T3g); T7P = VMUL(LDK(KP707106781), VSUB(T4V, T4U)); T7R = VADD(T7P, T7Q); T7X = VSUB(T7Q, T7P); { V T3s, T3D, T4L, T4M; T3s = VFNMS(LDK(KP923879532), T3r, VMUL(LDK(KP382683432), T3m)); T3D = VFMA(LDK(KP382683432), T3x, VMUL(LDK(KP923879532), T3C)); T3E = VSUB(T3s, T3D); T7O = VADD(T3s, T3D); T4L = VADD(T4b, T4m); T4M = VADD(T4s, T4v); T4N = VFNMS(LDK(KP555570233), T4M, VMUL(LDK(KP831469612), T4L)); T4R = VFMA(LDK(KP831469612), T4M, VMUL(LDK(KP555570233), T4L)); } { V T3W, T45, T4E, T4F; T3W = VSUB(T3K, T3V); T45 = VSUB(T41, T44); T46 = VFMA(LDK(KP980785280), T3W, VMUL(LDK(KP195090322), T45)); T4A = VFNMS(LDK(KP980785280), T45, VMUL(LDK(KP195090322), T3W)); T4E = VFMA(LDK(KP923879532), T3m, VMUL(LDK(KP382683432), T3r)); T4F = VFNMS(LDK(KP923879532), T3x, VMUL(LDK(KP382683432), T3C)); T4G = VADD(T4E, T4F); T7W = VSUB(T4F, T4E); } { V T4I, T4J, T4n, T4w; T4I = VADD(T3K, T3V); T4J = VADD(T41, T44); T4K = VFMA(LDK(KP555570233), T4I, VMUL(LDK(KP831469612), T4J)); T4Q = VFNMS(LDK(KP555570233), T4J, VMUL(LDK(KP831469612), T4I)); T4n = VSUB(T4b, T4m); T4w = VSUB(T4s, T4v); T4x = VFNMS(LDK(KP980785280), T4w, VMUL(LDK(KP195090322), T4n)); T4B = VFMA(LDK(KP195090322), T4w, VMUL(LDK(KP980785280), T4n)); } { V T3F, T4y, T7V, T7Y; T3F = VADD(T3h, T3E); T4y = VADD(T46, T4x); ST(&(ri[WS(rs, 23)]), VSUB(T3F, T4y), ms, &(ri[WS(rs, 1)])); ST(&(ri[WS(rs, 7)]), VADD(T3F, T4y), ms, &(ri[WS(rs, 1)])); T7V = VADD(T4A, T4B); T7Y = VADD(T7W, T7X); ST(&(ii[WS(rs, 7)]), VADD(T7V, T7Y), ms, &(ii[WS(rs, 1)])); ST(&(ii[WS(rs, 23)]), VSUB(T7Y, T7V), ms, &(ii[WS(rs, 1)])); } { V T4z, T4C, T7Z, T80; T4z = VSUB(T3h, T3E); T4C = VSUB(T4A, T4B); ST(&(ri[WS(rs, 31)]), VSUB(T4z, T4C), ms, &(ri[WS(rs, 1)])); ST(&(ri[WS(rs, 15)]), VADD(T4z, T4C), ms, &(ri[WS(rs, 1)])); T7Z = VSUB(T4x, T46); T80 = VSUB(T7X, T7W); ST(&(ii[WS(rs, 15)]), VADD(T7Z, T80), ms, &(ii[WS(rs, 1)])); ST(&(ii[WS(rs, 31)]), VSUB(T80, T7Z), ms, &(ii[WS(rs, 1)])); } { V T4H, T4O, T7N, T7S; T4H = VADD(T4D, T4G); T4O = VADD(T4K, T4N); ST(&(ri[WS(rs, 19)]), VSUB(T4H, T4O), ms, &(ri[WS(rs, 1)])); ST(&(ri[WS(rs, 3)]), VADD(T4H, T4O), ms, &(ri[WS(rs, 1)])); T7N = VADD(T4Q, T4R); T7S = VADD(T7O, T7R); ST(&(ii[WS(rs, 3)]), VADD(T7N, T7S), ms, &(ii[WS(rs, 1)])); ST(&(ii[WS(rs, 19)]), VSUB(T7S, T7N), ms, &(ii[WS(rs, 1)])); } { V T4P, T4S, T7T, T7U; T4P = VSUB(T4D, T4G); T4S = VSUB(T4Q, T4R); ST(&(ri[WS(rs, 27)]), VSUB(T4P, T4S), ms, &(ri[WS(rs, 1)])); ST(&(ri[WS(rs, 11)]), VADD(T4P, T4S), ms, &(ri[WS(rs, 1)])); T7T = VSUB(T4N, T4K); T7U = VSUB(T7R, T7O); ST(&(ii[WS(rs, 11)]), VADD(T7T, T7U), ms, &(ii[WS(rs, 1)])); ST(&(ii[WS(rs, 27)]), VSUB(T7U, T7T), ms, &(ii[WS(rs, 1)])); } } { V T4X, T5p, T7D, T7J, T54, T7y, T5z, T5D, T5c, T5m, T5s, T7I, T5w, T5C, T5j; V T5n, T4W, T7z; T4W = VMUL(LDK(KP707106781), VADD(T4U, T4V)); T4X = VSUB(T4T, T4W); T5p = VADD(T4T, T4W); T7z = VMUL(LDK(KP707106781), VADD(T3a, T3f)); T7D = VADD(T7z, T7C); T7J = VSUB(T7C, T7z); { V T50, T53, T5x, T5y; T50 = VFNMS(LDK(KP382683432), T4Z, VMUL(LDK(KP923879532), T4Y)); T53 = VFMA(LDK(KP923879532), T51, VMUL(LDK(KP382683432), T52)); T54 = VSUB(T50, T53); T7y = VADD(T50, T53); T5x = VADD(T5d, T5e); T5y = VADD(T5g, T5h); T5z = VFNMS(LDK(KP195090322), T5y, VMUL(LDK(KP980785280), T5x)); T5D = VFMA(LDK(KP195090322), T5x, VMUL(LDK(KP980785280), T5y)); } { V T58, T5b, T5q, T5r; T58 = VSUB(T56, T57); T5b = VSUB(T59, T5a); T5c = VFMA(LDK(KP555570233), T58, VMUL(LDK(KP831469612), T5b)); T5m = VFNMS(LDK(KP831469612), T58, VMUL(LDK(KP555570233), T5b)); T5q = VFMA(LDK(KP382683432), T4Y, VMUL(LDK(KP923879532), T4Z)); T5r = VFNMS(LDK(KP382683432), T51, VMUL(LDK(KP923879532), T52)); T5s = VADD(T5q, T5r); T7I = VSUB(T5r, T5q); } { V T5u, T5v, T5f, T5i; T5u = VADD(T56, T57); T5v = VADD(T59, T5a); T5w = VFMA(LDK(KP980785280), T5u, VMUL(LDK(KP195090322), T5v)); T5C = VFNMS(LDK(KP195090322), T5u, VMUL(LDK(KP980785280), T5v)); T5f = VSUB(T5d, T5e); T5i = VSUB(T5g, T5h); T5j = VFNMS(LDK(KP831469612), T5i, VMUL(LDK(KP555570233), T5f)); T5n = VFMA(LDK(KP831469612), T5f, VMUL(LDK(KP555570233), T5i)); } { V T55, T5k, T7H, T7K; T55 = VADD(T4X, T54); T5k = VADD(T5c, T5j); ST(&(ri[WS(rs, 21)]), VSUB(T55, T5k), ms, &(ri[WS(rs, 1)])); ST(&(ri[WS(rs, 5)]), VADD(T55, T5k), ms, &(ri[WS(rs, 1)])); T7H = VADD(T5m, T5n); T7K = VADD(T7I, T7J); ST(&(ii[WS(rs, 5)]), VADD(T7H, T7K), ms, &(ii[WS(rs, 1)])); ST(&(ii[WS(rs, 21)]), VSUB(T7K, T7H), ms, &(ii[WS(rs, 1)])); } { V T5l, T5o, T7L, T7M; T5l = VSUB(T4X, T54); T5o = VSUB(T5m, T5n); ST(&(ri[WS(rs, 29)]), VSUB(T5l, T5o), ms, &(ri[WS(rs, 1)])); ST(&(ri[WS(rs, 13)]), VADD(T5l, T5o), ms, &(ri[WS(rs, 1)])); T7L = VSUB(T5j, T5c); T7M = VSUB(T7J, T7I); ST(&(ii[WS(rs, 13)]), VADD(T7L, T7M), ms, &(ii[WS(rs, 1)])); ST(&(ii[WS(rs, 29)]), VSUB(T7M, T7L), ms, &(ii[WS(rs, 1)])); } { V T5t, T5A, T7x, T7E; T5t = VADD(T5p, T5s); T5A = VADD(T5w, T5z); ST(&(ri[WS(rs, 17)]), VSUB(T5t, T5A), ms, &(ri[WS(rs, 1)])); ST(&(ri[WS(rs, 1)]), VADD(T5t, T5A), ms, &(ri[WS(rs, 1)])); T7x = VADD(T5C, T5D); T7E = VADD(T7y, T7D); ST(&(ii[WS(rs, 1)]), VADD(T7x, T7E), ms, &(ii[WS(rs, 1)])); ST(&(ii[WS(rs, 17)]), VSUB(T7E, T7x), ms, &(ii[WS(rs, 1)])); } { V T5B, T5E, T7F, T7G; T5B = VSUB(T5p, T5s); T5E = VSUB(T5C, T5D); ST(&(ri[WS(rs, 25)]), VSUB(T5B, T5E), ms, &(ri[WS(rs, 1)])); ST(&(ri[WS(rs, 9)]), VADD(T5B, T5E), ms, &(ri[WS(rs, 1)])); T7F = VSUB(T5z, T5w); T7G = VSUB(T7D, T7y); ST(&(ii[WS(rs, 9)]), VADD(T7F, T7G), ms, &(ii[WS(rs, 1)])); ST(&(ii[WS(rs, 25)]), VSUB(T7G, T7F), ms, &(ii[WS(rs, 1)])); } } } } VLEAVE(); } static const tw_instr twinstr[] = { VTW(0, 1), VTW(0, 2), VTW(0, 3), VTW(0, 4), VTW(0, 5), VTW(0, 6), VTW(0, 7), VTW(0, 8), VTW(0, 9), VTW(0, 10), VTW(0, 11), VTW(0, 12), VTW(0, 13), VTW(0, 14), VTW(0, 15), VTW(0, 16), VTW(0, 17), VTW(0, 18), VTW(0, 19), VTW(0, 20), VTW(0, 21), VTW(0, 22), VTW(0, 23), VTW(0, 24), VTW(0, 25), VTW(0, 26), VTW(0, 27), VTW(0, 28), VTW(0, 29), VTW(0, 30), VTW(0, 31), {TW_NEXT, (2 * VL), 0} }; static const ct_desc desc = { 32, XSIMD_STRING("t1sv_32"), twinstr, &GENUS, {340, 114, 94, 0}, 0, 0, 0 }; void XSIMD(codelet_t1sv_32) (planner *p) { X(kdft_dit_register) (p, t1sv_32, &desc); } #endif /* HAVE_FMA */ fftw-3.3.3/dft/simd/common/t2sv_4.c0000644000175000017500000001473212054410576013710 00000000000000/* * Copyright (c) 2003, 2007-11 Matteo Frigo * Copyright (c) 2003, 2007-11 Massachusetts Institute of Technology * * This program is free software; you can redistribute it and/or modify * it under the terms of the GNU General Public License as published by * the Free Software Foundation; either version 2 of the License, or * (at your option) any later version. * * This program is distributed in the hope that it will be useful, * but WITHOUT ANY WARRANTY; without even the implied warranty of * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the * GNU General Public License for more details. * * You should have received a copy of the GNU General Public License * along with this program; if not, write to the Free Software * Foundation, Inc., 51 Franklin Street, Fifth Floor, Boston, MA 02110-1301 USA * */ /* This file was automatically generated --- DO NOT EDIT */ /* Generated on Sun Nov 25 07:39:26 EST 2012 */ #include "codelet-dft.h" #ifdef HAVE_FMA /* Generated by: ../../../genfft/gen_twiddle.native -fma -reorder-insns -schedule-for-pipeline -simd -compact -variables 4 -pipeline-latency 8 -twiddle-log3 -precompute-twiddles -n 4 -name t2sv_4 -include ts.h */ /* * This function contains 24 FP additions, 16 FP multiplications, * (or, 16 additions, 8 multiplications, 8 fused multiply/add), * 37 stack variables, 0 constants, and 16 memory accesses */ #include "ts.h" static void t2sv_4(R *ri, R *ii, const R *W, stride rs, INT mb, INT me, INT ms) { { INT m; for (m = mb, W = W + (mb * 4); m < me; m = m + (2 * VL), ri = ri + ((2 * VL) * ms), ii = ii + ((2 * VL) * ms), W = W + ((2 * VL) * 4), MAKE_VOLATILE_STRIDE(8, rs)) { V T2, T6, T3, T5, T1, Tx, T8, Tc, Tf, Ta, T4, Th, Tj, Tl; T2 = LDW(&(W[0])); T6 = LDW(&(W[TWVL * 3])); T3 = LDW(&(W[TWVL * 2])); T5 = LDW(&(W[TWVL * 1])); T1 = LD(&(ri[0]), ms, &(ri[0])); Tx = LD(&(ii[0]), ms, &(ii[0])); T8 = LD(&(ri[WS(rs, 2)]), ms, &(ri[0])); Tc = LD(&(ii[WS(rs, 2)]), ms, &(ii[0])); Tf = LD(&(ri[WS(rs, 1)]), ms, &(ri[WS(rs, 1)])); Ta = VMUL(T2, T6); T4 = VMUL(T2, T3); Th = LD(&(ii[WS(rs, 1)]), ms, &(ii[WS(rs, 1)])); Tj = LD(&(ri[WS(rs, 3)]), ms, &(ri[WS(rs, 1)])); Tl = LD(&(ii[WS(rs, 3)]), ms, &(ii[WS(rs, 1)])); { V Tg, Tb, T7, Tp, Tk, Tr, Ti; Tg = VMUL(T2, Tf); Tb = VFNMS(T5, T3, Ta); T7 = VFMA(T5, T6, T4); Tp = VMUL(T2, Th); Tk = VMUL(T3, Tj); Tr = VMUL(T3, Tl); Ti = VFMA(T5, Th, Tg); { V Tv, T9, Tq, Tm, Ts, Tw, Td; Tv = VMUL(T7, Tc); T9 = VMUL(T7, T8); Tq = VFNMS(T5, Tf, Tp); Tm = VFMA(T6, Tl, Tk); Ts = VFNMS(T6, Tj, Tr); Tw = VFNMS(Tb, T8, Tv); Td = VFMA(Tb, Tc, T9); { V Tn, TA, Tu, Tt; Tn = VADD(Ti, Tm); TA = VSUB(Ti, Tm); Tu = VADD(Tq, Ts); Tt = VSUB(Tq, Ts); { V Ty, Tz, Te, To; Ty = VADD(Tw, Tx); Tz = VSUB(Tx, Tw); Te = VADD(T1, Td); To = VSUB(T1, Td); ST(&(ii[WS(rs, 3)]), VADD(TA, Tz), ms, &(ii[WS(rs, 1)])); ST(&(ii[WS(rs, 1)]), VSUB(Tz, TA), ms, &(ii[WS(rs, 1)])); ST(&(ii[WS(rs, 2)]), VSUB(Ty, Tu), ms, &(ii[0])); ST(&(ii[0]), VADD(Tu, Ty), ms, &(ii[0])); ST(&(ri[WS(rs, 1)]), VADD(To, Tt), ms, &(ri[WS(rs, 1)])); ST(&(ri[WS(rs, 3)]), VSUB(To, Tt), ms, &(ri[WS(rs, 1)])); ST(&(ri[0]), VADD(Te, Tn), ms, &(ri[0])); ST(&(ri[WS(rs, 2)]), VSUB(Te, Tn), ms, &(ri[0])); } } } } } } VLEAVE(); } static const tw_instr twinstr[] = { VTW(0, 1), VTW(0, 3), {TW_NEXT, (2 * VL), 0} }; static const ct_desc desc = { 4, XSIMD_STRING("t2sv_4"), twinstr, &GENUS, {16, 8, 8, 0}, 0, 0, 0 }; void XSIMD(codelet_t2sv_4) (planner *p) { X(kdft_dit_register) (p, t2sv_4, &desc); } #else /* HAVE_FMA */ /* Generated by: ../../../genfft/gen_twiddle.native -simd -compact -variables 4 -pipeline-latency 8 -twiddle-log3 -precompute-twiddles -n 4 -name t2sv_4 -include ts.h */ /* * This function contains 24 FP additions, 16 FP multiplications, * (or, 16 additions, 8 multiplications, 8 fused multiply/add), * 21 stack variables, 0 constants, and 16 memory accesses */ #include "ts.h" static void t2sv_4(R *ri, R *ii, const R *W, stride rs, INT mb, INT me, INT ms) { { INT m; for (m = mb, W = W + (mb * 4); m < me; m = m + (2 * VL), ri = ri + ((2 * VL) * ms), ii = ii + ((2 * VL) * ms), W = W + ((2 * VL) * 4), MAKE_VOLATILE_STRIDE(8, rs)) { V T2, T4, T3, T5, T6, T8; T2 = LDW(&(W[0])); T4 = LDW(&(W[TWVL * 1])); T3 = LDW(&(W[TWVL * 2])); T5 = LDW(&(W[TWVL * 3])); T6 = VFMA(T2, T3, VMUL(T4, T5)); T8 = VFNMS(T4, T3, VMUL(T2, T5)); { V T1, Tp, Ta, To, Te, Tk, Th, Tl, T7, T9; T1 = LD(&(ri[0]), ms, &(ri[0])); Tp = LD(&(ii[0]), ms, &(ii[0])); T7 = LD(&(ri[WS(rs, 2)]), ms, &(ri[0])); T9 = LD(&(ii[WS(rs, 2)]), ms, &(ii[0])); Ta = VFMA(T6, T7, VMUL(T8, T9)); To = VFNMS(T8, T7, VMUL(T6, T9)); { V Tc, Td, Tf, Tg; Tc = LD(&(ri[WS(rs, 1)]), ms, &(ri[WS(rs, 1)])); Td = LD(&(ii[WS(rs, 1)]), ms, &(ii[WS(rs, 1)])); Te = VFMA(T2, Tc, VMUL(T4, Td)); Tk = VFNMS(T4, Tc, VMUL(T2, Td)); Tf = LD(&(ri[WS(rs, 3)]), ms, &(ri[WS(rs, 1)])); Tg = LD(&(ii[WS(rs, 3)]), ms, &(ii[WS(rs, 1)])); Th = VFMA(T3, Tf, VMUL(T5, Tg)); Tl = VFNMS(T5, Tf, VMUL(T3, Tg)); } { V Tb, Ti, Tn, Tq; Tb = VADD(T1, Ta); Ti = VADD(Te, Th); ST(&(ri[WS(rs, 2)]), VSUB(Tb, Ti), ms, &(ri[0])); ST(&(ri[0]), VADD(Tb, Ti), ms, &(ri[0])); Tn = VADD(Tk, Tl); Tq = VADD(To, Tp); ST(&(ii[0]), VADD(Tn, Tq), ms, &(ii[0])); ST(&(ii[WS(rs, 2)]), VSUB(Tq, Tn), ms, &(ii[0])); } { V Tj, Tm, Tr, Ts; Tj = VSUB(T1, Ta); Tm = VSUB(Tk, Tl); ST(&(ri[WS(rs, 3)]), VSUB(Tj, Tm), ms, &(ri[WS(rs, 1)])); ST(&(ri[WS(rs, 1)]), VADD(Tj, Tm), ms, &(ri[WS(rs, 1)])); Tr = VSUB(Tp, To); Ts = VSUB(Te, Th); ST(&(ii[WS(rs, 1)]), VSUB(Tr, Ts), ms, &(ii[WS(rs, 1)])); ST(&(ii[WS(rs, 3)]), VADD(Ts, Tr), ms, &(ii[WS(rs, 1)])); } } } } VLEAVE(); } static const tw_instr twinstr[] = { VTW(0, 1), VTW(0, 3), {TW_NEXT, (2 * VL), 0} }; static const ct_desc desc = { 4, XSIMD_STRING("t2sv_4"), twinstr, &GENUS, {16, 8, 8, 0}, 0, 0, 0 }; void XSIMD(codelet_t2sv_4) (planner *p) { X(kdft_dit_register) (p, t2sv_4, &desc); } #endif /* HAVE_FMA */ fftw-3.3.3/dft/simd/common/t2sv_8.c0000644000175000017500000003141412054410576013710 00000000000000/* * Copyright (c) 2003, 2007-11 Matteo Frigo * Copyright (c) 2003, 2007-11 Massachusetts Institute of Technology * * This program is free software; you can redistribute it and/or modify * it under the terms of the GNU General Public License as published by * the Free Software Foundation; either version 2 of the License, or * (at your option) any later version. * * This program is distributed in the hope that it will be useful, * but WITHOUT ANY WARRANTY; without even the implied warranty of * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the * GNU General Public License for more details. * * You should have received a copy of the GNU General Public License * along with this program; if not, write to the Free Software * Foundation, Inc., 51 Franklin Street, Fifth Floor, Boston, MA 02110-1301 USA * */ /* This file was automatically generated --- DO NOT EDIT */ /* Generated on Sun Nov 25 07:39:26 EST 2012 */ #include "codelet-dft.h" #ifdef HAVE_FMA /* Generated by: ../../../genfft/gen_twiddle.native -fma -reorder-insns -schedule-for-pipeline -simd -compact -variables 4 -pipeline-latency 8 -twiddle-log3 -precompute-twiddles -n 8 -name t2sv_8 -include ts.h */ /* * This function contains 74 FP additions, 50 FP multiplications, * (or, 44 additions, 20 multiplications, 30 fused multiply/add), * 64 stack variables, 1 constants, and 32 memory accesses */ #include "ts.h" static void t2sv_8(R *ri, R *ii, const R *W, stride rs, INT mb, INT me, INT ms) { DVK(KP707106781, +0.707106781186547524400844362104849039284835938); { INT m; for (m = mb, W = W + (mb * 6); m < me; m = m + (2 * VL), ri = ri + ((2 * VL) * ms), ii = ii + ((2 * VL) * ms), W = W + ((2 * VL) * 6), MAKE_VOLATILE_STRIDE(16, rs)) { V T1m, T1l, T1k, T1u, T1n, T1o; { V T2, T3, Tl, Tn, T5, T6; T2 = LDW(&(W[0])); T3 = LDW(&(W[TWVL * 2])); Tl = LDW(&(W[TWVL * 4])); Tn = LDW(&(W[TWVL * 5])); T5 = LDW(&(W[TWVL * 1])); T6 = LDW(&(W[TWVL * 3])); { V T1, T1s, TK, T1r, Td, Tk, TG, TC, TY, Tu, TW, TL, TM, TO, TQ; V Tx, Tz, TD, TH; { V T8, T4, Tm, Tr, Tc, Ta; T1 = LD(&(ri[0]), ms, &(ri[0])); T1s = LD(&(ii[0]), ms, &(ii[0])); T8 = LD(&(ri[WS(rs, 4)]), ms, &(ri[0])); T4 = VMUL(T2, T3); Tm = VMUL(T2, Tl); Tr = VMUL(T2, Tn); Tc = LD(&(ii[WS(rs, 4)]), ms, &(ii[0])); Ta = VMUL(T2, T6); { V Tp, Tt, Tg, T7, Tf, To, Ts, Ti, Tb, Tj; Tp = LD(&(ri[WS(rs, 6)]), ms, &(ri[0])); Tt = LD(&(ii[WS(rs, 6)]), ms, &(ii[0])); Tg = LD(&(ri[WS(rs, 2)]), ms, &(ri[0])); T7 = VFNMS(T5, T6, T4); Tf = VFMA(T5, T6, T4); To = VFMA(T5, Tn, Tm); Ts = VFNMS(T5, Tl, Tr); Ti = VFNMS(T5, T3, Ta); Tb = VFMA(T5, T3, Ta); Tj = LD(&(ii[WS(rs, 2)]), ms, &(ii[0])); TK = LD(&(ri[WS(rs, 7)]), ms, &(ri[WS(rs, 1)])); { V T1q, T9, Th, TF; T1q = VMUL(T7, Tc); T9 = VMUL(T7, T8); Th = VMUL(Tf, Tg); TF = VMUL(Tf, Tn); { V TB, TX, Tq, TV; TB = VMUL(Tf, Tl); TX = VMUL(To, Tt); Tq = VMUL(To, Tp); TV = VMUL(Tf, Tj); T1r = VFNMS(Tb, T8, T1q); Td = VFMA(Tb, Tc, T9); Tk = VFMA(Ti, Tj, Th); TG = VFNMS(Ti, Tl, TF); TC = VFMA(Ti, Tn, TB); TY = VFNMS(Ts, Tp, TX); Tu = VFMA(Ts, Tt, Tq); TW = VFNMS(Ti, Tg, TV); TL = VMUL(Tl, TK); } } TM = LD(&(ii[WS(rs, 7)]), ms, &(ii[WS(rs, 1)])); TO = LD(&(ri[WS(rs, 3)]), ms, &(ri[WS(rs, 1)])); TQ = LD(&(ii[WS(rs, 3)]), ms, &(ii[WS(rs, 1)])); Tx = LD(&(ri[WS(rs, 1)]), ms, &(ri[WS(rs, 1)])); Tz = LD(&(ii[WS(rs, 1)]), ms, &(ii[WS(rs, 1)])); TD = LD(&(ri[WS(rs, 5)]), ms, &(ri[WS(rs, 1)])); TH = LD(&(ii[WS(rs, 5)]), ms, &(ii[WS(rs, 1)])); } } { V Te, T1p, T1g, T10, TS, T18, T1d, T1t, T1x, T1y, Tv, TJ, T11, T16; { V TN, T1a, TR, T1c, TA, T13, TI, T15; { V TU, T19, TP, T1b, Ty, T12, TE, T14, TZ; TU = VSUB(T1, Td); Te = VADD(T1, Td); TN = VFMA(Tn, TM, TL); T19 = VMUL(Tl, TM); TP = VMUL(T3, TO); T1b = VMUL(T3, TQ); Ty = VMUL(T2, Tx); T12 = VMUL(T2, Tz); TE = VMUL(TC, TD); T14 = VMUL(TC, TH); T1p = VADD(TW, TY); TZ = VSUB(TW, TY); T1a = VFNMS(Tn, TK, T19); TR = VFMA(T6, TQ, TP); T1c = VFNMS(T6, TO, T1b); TA = VFMA(T5, Tz, Ty); T13 = VFNMS(T5, Tx, T12); TI = VFMA(TG, TH, TE); T15 = VFNMS(TG, TD, T14); T1g = VSUB(TU, TZ); T10 = VADD(TU, TZ); } TS = VADD(TN, TR); T18 = VSUB(TN, TR); T1d = VSUB(T1a, T1c); T1m = VADD(T1a, T1c); T1t = VADD(T1r, T1s); T1x = VSUB(T1s, T1r); T1y = VSUB(Tk, Tu); Tv = VADD(Tk, Tu); TJ = VADD(TA, TI); T11 = VSUB(TA, TI); T16 = VSUB(T13, T15); T1l = VADD(T13, T15); } { V Tw, T1w, T1v, TT; { V T1i, T1e, T1B, T1z, T1h, T17; T1i = VADD(T18, T1d); T1e = VSUB(T18, T1d); T1B = VADD(T1y, T1x); T1z = VSUB(T1x, T1y); T1h = VSUB(T16, T11); T17 = VADD(T11, T16); T1k = VSUB(Te, Tv); Tw = VADD(Te, Tv); { V T1A, T1j, T1C, T1f; T1A = VADD(T1h, T1i); T1j = VSUB(T1h, T1i); T1C = VSUB(T1e, T17); T1f = VADD(T17, T1e); T1w = VSUB(T1t, T1p); T1u = VADD(T1p, T1t); T1v = VSUB(TS, TJ); TT = VADD(TJ, TS); ST(&(ii[WS(rs, 1)]), VFMA(LDK(KP707106781), T1A, T1z), ms, &(ii[WS(rs, 1)])); ST(&(ii[WS(rs, 5)]), VFNMS(LDK(KP707106781), T1A, T1z), ms, &(ii[WS(rs, 1)])); ST(&(ri[WS(rs, 3)]), VFMA(LDK(KP707106781), T1j, T1g), ms, &(ri[WS(rs, 1)])); ST(&(ri[WS(rs, 7)]), VFNMS(LDK(KP707106781), T1j, T1g), ms, &(ri[WS(rs, 1)])); ST(&(ii[WS(rs, 3)]), VFMA(LDK(KP707106781), T1C, T1B), ms, &(ii[WS(rs, 1)])); ST(&(ii[WS(rs, 7)]), VFNMS(LDK(KP707106781), T1C, T1B), ms, &(ii[WS(rs, 1)])); ST(&(ri[WS(rs, 1)]), VFMA(LDK(KP707106781), T1f, T10), ms, &(ri[WS(rs, 1)])); ST(&(ri[WS(rs, 5)]), VFNMS(LDK(KP707106781), T1f, T10), ms, &(ri[WS(rs, 1)])); } } ST(&(ri[WS(rs, 4)]), VSUB(Tw, TT), ms, &(ri[0])); ST(&(ri[0]), VADD(Tw, TT), ms, &(ri[0])); ST(&(ii[WS(rs, 6)]), VSUB(T1w, T1v), ms, &(ii[0])); ST(&(ii[WS(rs, 2)]), VADD(T1v, T1w), ms, &(ii[0])); } } } } T1n = VSUB(T1l, T1m); T1o = VADD(T1l, T1m); ST(&(ii[0]), VADD(T1o, T1u), ms, &(ii[0])); ST(&(ii[WS(rs, 4)]), VSUB(T1u, T1o), ms, &(ii[0])); ST(&(ri[WS(rs, 2)]), VADD(T1k, T1n), ms, &(ri[0])); ST(&(ri[WS(rs, 6)]), VSUB(T1k, T1n), ms, &(ri[0])); } } VLEAVE(); } static const tw_instr twinstr[] = { VTW(0, 1), VTW(0, 3), VTW(0, 7), {TW_NEXT, (2 * VL), 0} }; static const ct_desc desc = { 8, XSIMD_STRING("t2sv_8"), twinstr, &GENUS, {44, 20, 30, 0}, 0, 0, 0 }; void XSIMD(codelet_t2sv_8) (planner *p) { X(kdft_dit_register) (p, t2sv_8, &desc); } #else /* HAVE_FMA */ /* Generated by: ../../../genfft/gen_twiddle.native -simd -compact -variables 4 -pipeline-latency 8 -twiddle-log3 -precompute-twiddles -n 8 -name t2sv_8 -include ts.h */ /* * This function contains 74 FP additions, 44 FP multiplications, * (or, 56 additions, 26 multiplications, 18 fused multiply/add), * 42 stack variables, 1 constants, and 32 memory accesses */ #include "ts.h" static void t2sv_8(R *ri, R *ii, const R *W, stride rs, INT mb, INT me, INT ms) { DVK(KP707106781, +0.707106781186547524400844362104849039284835938); { INT m; for (m = mb, W = W + (mb * 6); m < me; m = m + (2 * VL), ri = ri + ((2 * VL) * ms), ii = ii + ((2 * VL) * ms), W = W + ((2 * VL) * 6), MAKE_VOLATILE_STRIDE(16, rs)) { V T2, T5, T3, T6, T8, Tc, Tg, Ti, Tl, Tm, Tn, Tz, Tp, Tx; { V T4, Tb, T7, Ta; T2 = LDW(&(W[0])); T5 = LDW(&(W[TWVL * 1])); T3 = LDW(&(W[TWVL * 2])); T6 = LDW(&(W[TWVL * 3])); T4 = VMUL(T2, T3); Tb = VMUL(T5, T3); T7 = VMUL(T5, T6); Ta = VMUL(T2, T6); T8 = VSUB(T4, T7); Tc = VADD(Ta, Tb); Tg = VADD(T4, T7); Ti = VSUB(Ta, Tb); Tl = LDW(&(W[TWVL * 4])); Tm = LDW(&(W[TWVL * 5])); Tn = VFMA(T2, Tl, VMUL(T5, Tm)); Tz = VFNMS(Ti, Tl, VMUL(Tg, Tm)); Tp = VFNMS(T5, Tl, VMUL(T2, Tm)); Tx = VFMA(Tg, Tl, VMUL(Ti, Tm)); } { V Tf, T1i, TL, T1d, TJ, T17, TV, TY, Ts, T1j, TO, T1a, TC, T16, TQ; V TT; { V T1, T1c, Te, T1b, T9, Td; T1 = LD(&(ri[0]), ms, &(ri[0])); T1c = LD(&(ii[0]), ms, &(ii[0])); T9 = LD(&(ri[WS(rs, 4)]), ms, &(ri[0])); Td = LD(&(ii[WS(rs, 4)]), ms, &(ii[0])); Te = VFMA(T8, T9, VMUL(Tc, Td)); T1b = VFNMS(Tc, T9, VMUL(T8, Td)); Tf = VADD(T1, Te); T1i = VSUB(T1c, T1b); TL = VSUB(T1, Te); T1d = VADD(T1b, T1c); } { V TF, TW, TI, TX; { V TD, TE, TG, TH; TD = LD(&(ri[WS(rs, 7)]), ms, &(ri[WS(rs, 1)])); TE = LD(&(ii[WS(rs, 7)]), ms, &(ii[WS(rs, 1)])); TF = VFMA(Tl, TD, VMUL(Tm, TE)); TW = VFNMS(Tm, TD, VMUL(Tl, TE)); TG = LD(&(ri[WS(rs, 3)]), ms, &(ri[WS(rs, 1)])); TH = LD(&(ii[WS(rs, 3)]), ms, &(ii[WS(rs, 1)])); TI = VFMA(T3, TG, VMUL(T6, TH)); TX = VFNMS(T6, TG, VMUL(T3, TH)); } TJ = VADD(TF, TI); T17 = VADD(TW, TX); TV = VSUB(TF, TI); TY = VSUB(TW, TX); } { V Tk, TM, Tr, TN; { V Th, Tj, To, Tq; Th = LD(&(ri[WS(rs, 2)]), ms, &(ri[0])); Tj = LD(&(ii[WS(rs, 2)]), ms, &(ii[0])); Tk = VFMA(Tg, Th, VMUL(Ti, Tj)); TM = VFNMS(Ti, Th, VMUL(Tg, Tj)); To = LD(&(ri[WS(rs, 6)]), ms, &(ri[0])); Tq = LD(&(ii[WS(rs, 6)]), ms, &(ii[0])); Tr = VFMA(Tn, To, VMUL(Tp, Tq)); TN = VFNMS(Tp, To, VMUL(Tn, Tq)); } Ts = VADD(Tk, Tr); T1j = VSUB(Tk, Tr); TO = VSUB(TM, TN); T1a = VADD(TM, TN); } { V Tw, TR, TB, TS; { V Tu, Tv, Ty, TA; Tu = LD(&(ri[WS(rs, 1)]), ms, &(ri[WS(rs, 1)])); Tv = LD(&(ii[WS(rs, 1)]), ms, &(ii[WS(rs, 1)])); Tw = VFMA(T2, Tu, VMUL(T5, Tv)); TR = VFNMS(T5, Tu, VMUL(T2, Tv)); Ty = LD(&(ri[WS(rs, 5)]), ms, &(ri[WS(rs, 1)])); TA = LD(&(ii[WS(rs, 5)]), ms, &(ii[WS(rs, 1)])); TB = VFMA(Tx, Ty, VMUL(Tz, TA)); TS = VFNMS(Tz, Ty, VMUL(Tx, TA)); } TC = VADD(Tw, TB); T16 = VADD(TR, TS); TQ = VSUB(Tw, TB); TT = VSUB(TR, TS); } { V Tt, TK, T1f, T1g; Tt = VADD(Tf, Ts); TK = VADD(TC, TJ); ST(&(ri[WS(rs, 4)]), VSUB(Tt, TK), ms, &(ri[0])); ST(&(ri[0]), VADD(Tt, TK), ms, &(ri[0])); { V T19, T1e, T15, T18; T19 = VADD(T16, T17); T1e = VADD(T1a, T1d); ST(&(ii[0]), VADD(T19, T1e), ms, &(ii[0])); ST(&(ii[WS(rs, 4)]), VSUB(T1e, T19), ms, &(ii[0])); T15 = VSUB(Tf, Ts); T18 = VSUB(T16, T17); ST(&(ri[WS(rs, 6)]), VSUB(T15, T18), ms, &(ri[0])); ST(&(ri[WS(rs, 2)]), VADD(T15, T18), ms, &(ri[0])); } T1f = VSUB(TJ, TC); T1g = VSUB(T1d, T1a); ST(&(ii[WS(rs, 2)]), VADD(T1f, T1g), ms, &(ii[0])); ST(&(ii[WS(rs, 6)]), VSUB(T1g, T1f), ms, &(ii[0])); { V T11, T1k, T14, T1h, T12, T13; T11 = VSUB(TL, TO); T1k = VSUB(T1i, T1j); T12 = VSUB(TT, TQ); T13 = VADD(TV, TY); T14 = VMUL(LDK(KP707106781), VSUB(T12, T13)); T1h = VMUL(LDK(KP707106781), VADD(T12, T13)); ST(&(ri[WS(rs, 7)]), VSUB(T11, T14), ms, &(ri[WS(rs, 1)])); ST(&(ii[WS(rs, 5)]), VSUB(T1k, T1h), ms, &(ii[WS(rs, 1)])); ST(&(ri[WS(rs, 3)]), VADD(T11, T14), ms, &(ri[WS(rs, 1)])); ST(&(ii[WS(rs, 1)]), VADD(T1h, T1k), ms, &(ii[WS(rs, 1)])); } { V TP, T1m, T10, T1l, TU, TZ; TP = VADD(TL, TO); T1m = VADD(T1j, T1i); TU = VADD(TQ, TT); TZ = VSUB(TV, TY); T10 = VMUL(LDK(KP707106781), VADD(TU, TZ)); T1l = VMUL(LDK(KP707106781), VSUB(TZ, TU)); ST(&(ri[WS(rs, 5)]), VSUB(TP, T10), ms, &(ri[WS(rs, 1)])); ST(&(ii[WS(rs, 7)]), VSUB(T1m, T1l), ms, &(ii[WS(rs, 1)])); ST(&(ri[WS(rs, 1)]), VADD(TP, T10), ms, &(ri[WS(rs, 1)])); ST(&(ii[WS(rs, 3)]), VADD(T1l, T1m), ms, &(ii[WS(rs, 1)])); } } } } } VLEAVE(); } static const tw_instr twinstr[] = { VTW(0, 1), VTW(0, 3), VTW(0, 7), {TW_NEXT, (2 * VL), 0} }; static const ct_desc desc = { 8, XSIMD_STRING("t2sv_8"), twinstr, &GENUS, {56, 26, 18, 0}, 0, 0, 0 }; void XSIMD(codelet_t2sv_8) (planner *p) { X(kdft_dit_register) (p, t2sv_8, &desc); } #endif /* HAVE_FMA */ fftw-3.3.3/dft/simd/common/t2sv_16.c0000644000175000017500000006717012054410600013763 00000000000000/* * Copyright (c) 2003, 2007-11 Matteo Frigo * Copyright (c) 2003, 2007-11 Massachusetts Institute of Technology * * This program is free software; you can redistribute it and/or modify * it under the terms of the GNU General Public License as published by * the Free Software Foundation; either version 2 of the License, or * (at your option) any later version. * * This program is distributed in the hope that it will be useful, * but WITHOUT ANY WARRANTY; without even the implied warranty of * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the * GNU General Public License for more details. * * You should have received a copy of the GNU General Public License * along with this program; if not, write to the Free Software * Foundation, Inc., 51 Franklin Street, Fifth Floor, Boston, MA 02110-1301 USA * */ /* This file was automatically generated --- DO NOT EDIT */ /* Generated on Sun Nov 25 07:39:26 EST 2012 */ #include "codelet-dft.h" #ifdef HAVE_FMA /* Generated by: ../../../genfft/gen_twiddle.native -fma -reorder-insns -schedule-for-pipeline -simd -compact -variables 4 -pipeline-latency 8 -twiddle-log3 -precompute-twiddles -n 16 -name t2sv_16 -include ts.h */ /* * This function contains 196 FP additions, 134 FP multiplications, * (or, 104 additions, 42 multiplications, 92 fused multiply/add), * 120 stack variables, 3 constants, and 64 memory accesses */ #include "ts.h" static void t2sv_16(R *ri, R *ii, const R *W, stride rs, INT mb, INT me, INT ms) { DVK(KP923879532, +0.923879532511286756128183189396788286822416626); DVK(KP414213562, +0.414213562373095048801688724209698078569671875); DVK(KP707106781, +0.707106781186547524400844362104849039284835938); { INT m; for (m = mb, W = W + (mb * 8); m < me; m = m + (2 * VL), ri = ri + ((2 * VL) * ms), ii = ii + ((2 * VL) * ms), W = W + ((2 * VL) * 8), MAKE_VOLATILE_STRIDE(32, rs)) { V T34, T30, T2N, T2v, T2M, T2g, T3V, T3X, T32, T2U, T33, T2X, T2O, T2K, T3P; V T3R; { V T2, Tf, TM, TO, T3, T6, T5, Th; T2 = LDW(&(W[0])); Tf = LDW(&(W[TWVL * 2])); TM = LDW(&(W[TWVL * 6])); TO = LDW(&(W[TWVL * 7])); T3 = LDW(&(W[TWVL * 4])); T6 = LDW(&(W[TWVL * 5])); T5 = LDW(&(W[TWVL * 1])); Th = LDW(&(W[TWVL * 3])); { V TW, TZ, Te, T1U, T3A, T3L, T2D, T1G, T3h, T2A, T2B, T1R, T3i, T2I, Tx; V T3M, T1Z, T3w, TL, T26, T25, T37, T1l, T2q, T1d, T2o, T2l, T3c, T1r, T2s; V TX, T10, TV, T2a; { V Tz, TP, TT, Tq, TF, Tu, TI, Tm, TC, T1j, T1p, T1m, T1f, T1O, T1M; V T1K, T2F, Tj, Tn, T1Q, T2G, Tk, T1V, Tr, Tv; { V T1, Ti, Tb, T3z, T8, Tc, T1u, T1D, T1L, T1z, T9, T3x, T1v, T1w, T1A; V T1E; { V T7, T1i, T1e, T1C, T1y; T1 = LD(&(ri[0]), ms, &(ri[0])); { V Tg, TN, TS, Tp; Tg = VMUL(T2, Tf); TN = VMUL(T2, TM); TS = VMUL(T2, TO); Tp = VMUL(Tf, T3); { V T4, Tt, Ta, Tl; T4 = VMUL(T2, T3); Tt = VMUL(Tf, T6); Ta = VMUL(T2, T6); Tl = VMUL(T2, Th); Ti = VFNMS(T5, Th, Tg); Tz = VFMA(T5, Th, Tg); TP = VFMA(T5, TO, TN); TT = VFNMS(T5, TM, TS); TW = VFMA(Th, T6, Tp); Tq = VFNMS(Th, T6, Tp); TF = VFNMS(T5, T6, T4); T7 = VFMA(T5, T6, T4); Tu = VFMA(Th, T3, Tt); TZ = VFNMS(Th, T3, Tt); TI = VFMA(T5, T3, Ta); Tb = VFNMS(T5, T3, Ta); Tm = VFMA(T5, Tf, Tl); TC = VFNMS(T5, Tf, Tl); T1i = VMUL(Ti, T6); T1e = VMUL(Ti, T3); T1C = VMUL(Tz, T6); T1y = VMUL(Tz, T3); T3z = LD(&(ii[0]), ms, &(ii[0])); } } T8 = LD(&(ri[WS(rs, 8)]), ms, &(ri[0])); Tc = LD(&(ii[WS(rs, 8)]), ms, &(ii[0])); T1u = LD(&(ri[WS(rs, 15)]), ms, &(ri[WS(rs, 1)])); T1j = VFNMS(Tm, T3, T1i); T1p = VFMA(Tm, T3, T1i); T1m = VFNMS(Tm, T6, T1e); T1f = VFMA(Tm, T6, T1e); T1D = VFNMS(TC, T3, T1C); T1O = VFMA(TC, T3, T1C); T1L = VFNMS(TC, T6, T1y); T1z = VFMA(TC, T6, T1y); T9 = VMUL(T7, T8); T3x = VMUL(T7, Tc); T1v = VMUL(TM, T1u); T1w = LD(&(ii[WS(rs, 15)]), ms, &(ii[WS(rs, 1)])); T1A = LD(&(ri[WS(rs, 7)]), ms, &(ri[WS(rs, 1)])); T1E = LD(&(ii[WS(rs, 7)]), ms, &(ii[WS(rs, 1)])); } { V T1x, T2x, T1F, T2z, T1N, T1P; { V T1H, T1J, T1I, T2E; { V Td, T3y, T2w, T1B, T2y; T1H = LD(&(ri[WS(rs, 3)]), ms, &(ri[WS(rs, 1)])); T1J = LD(&(ii[WS(rs, 3)]), ms, &(ii[WS(rs, 1)])); Td = VFMA(Tb, Tc, T9); T3y = VFNMS(Tb, T8, T3x); T1M = LD(&(ri[WS(rs, 11)]), ms, &(ri[WS(rs, 1)])); T1x = VFMA(TO, T1w, T1v); T2w = VMUL(TM, T1w); T1B = VMUL(T1z, T1A); T2y = VMUL(T1z, T1E); T1I = VMUL(Tf, T1H); T2E = VMUL(Tf, T1J); Te = VADD(T1, Td); T1U = VSUB(T1, Td); T3A = VADD(T3y, T3z); T3L = VSUB(T3z, T3y); T2x = VFNMS(TO, T1u, T2w); T1F = VFMA(T1D, T1E, T1B); T2z = VFNMS(T1D, T1A, T2y); T1N = VMUL(T1L, T1M); T1P = LD(&(ii[WS(rs, 11)]), ms, &(ii[WS(rs, 1)])); } T1K = VFMA(Th, T1J, T1I); T2F = VFNMS(Th, T1H, T2E); } Tj = LD(&(ri[WS(rs, 4)]), ms, &(ri[0])); Tn = LD(&(ii[WS(rs, 4)]), ms, &(ii[0])); T2D = VSUB(T1x, T1F); T1G = VADD(T1x, T1F); T3h = VADD(T2x, T2z); T2A = VSUB(T2x, T2z); T1Q = VFMA(T1O, T1P, T1N); T2G = VMUL(T1L, T1P); Tk = VMUL(Ti, Tj); T1V = VMUL(Ti, Tn); Tr = LD(&(ri[WS(rs, 12)]), ms, &(ri[0])); Tv = LD(&(ii[WS(rs, 12)]), ms, &(ii[0])); } } { V TE, T22, T15, T17, TK, T16, T2h, T24, T19, T1b; { V To, T1W, TG, TJ, Tw, T1Y, TH, T23; { V TA, TD, TB, T21, T2H, Ts, T1X; TA = LD(&(ri[WS(rs, 2)]), ms, &(ri[0])); TD = LD(&(ii[WS(rs, 2)]), ms, &(ii[0])); T2B = VSUB(T1K, T1Q); T1R = VADD(T1K, T1Q); T2H = VFNMS(T1O, T1M, T2G); To = VFMA(Tm, Tn, Tk); T1W = VFNMS(Tm, Tj, T1V); Ts = VMUL(Tq, Tr); T1X = VMUL(Tq, Tv); TB = VMUL(Tz, TA); T21 = VMUL(Tz, TD); TG = LD(&(ri[WS(rs, 10)]), ms, &(ri[0])); T3i = VADD(T2F, T2H); T2I = VSUB(T2F, T2H); TJ = LD(&(ii[WS(rs, 10)]), ms, &(ii[0])); Tw = VFMA(Tu, Tv, Ts); T1Y = VFNMS(Tu, Tr, T1X); TE = VFMA(TC, TD, TB); T22 = VFNMS(TC, TA, T21); TH = VMUL(TF, TG); } T15 = LD(&(ri[WS(rs, 1)]), ms, &(ri[WS(rs, 1)])); T17 = LD(&(ii[WS(rs, 1)]), ms, &(ii[WS(rs, 1)])); T23 = VMUL(TF, TJ); Tx = VADD(To, Tw); T3M = VSUB(To, Tw); T1Z = VSUB(T1W, T1Y); T3w = VADD(T1W, T1Y); TK = VFMA(TI, TJ, TH); T16 = VMUL(T2, T15); T2h = VMUL(T2, T17); T24 = VFNMS(TI, TG, T23); T19 = LD(&(ri[WS(rs, 9)]), ms, &(ri[WS(rs, 1)])); T1b = LD(&(ii[WS(rs, 9)]), ms, &(ii[WS(rs, 1)])); } { V T1g, T1k, T18, T2i, T1a, T2j, T1h, T2p, T1n, T1q; T1g = LD(&(ri[WS(rs, 5)]), ms, &(ri[WS(rs, 1)])); T1k = LD(&(ii[WS(rs, 5)]), ms, &(ii[WS(rs, 1)])); TL = VADD(TE, TK); T26 = VSUB(TE, TK); T18 = VFMA(T5, T17, T16); T2i = VFNMS(T5, T15, T2h); T25 = VSUB(T22, T24); T37 = VADD(T22, T24); T1a = VMUL(T3, T19); T2j = VMUL(T3, T1b); T1h = VMUL(T1f, T1g); T2p = VMUL(T1f, T1k); T1n = LD(&(ri[WS(rs, 13)]), ms, &(ri[WS(rs, 1)])); T1q = LD(&(ii[WS(rs, 13)]), ms, &(ii[WS(rs, 1)])); { V TQ, TU, TR, T29; { V T1c, T2k, T1o, T2r; TQ = LD(&(ri[WS(rs, 14)]), ms, &(ri[0])); TU = LD(&(ii[WS(rs, 14)]), ms, &(ii[0])); T1c = VFMA(T6, T1b, T1a); T2k = VFNMS(T6, T19, T2j); T1l = VFMA(T1j, T1k, T1h); T2q = VFNMS(T1j, T1g, T2p); T1o = VMUL(T1m, T1n); T2r = VMUL(T1m, T1q); TR = VMUL(TP, TQ); T29 = VMUL(TP, TU); T1d = VADD(T18, T1c); T2o = VSUB(T18, T1c); T2l = VSUB(T2i, T2k); T3c = VADD(T2i, T2k); T1r = VFMA(T1p, T1q, T1o); T2s = VFNMS(T1p, T1n, T2r); TX = LD(&(ri[WS(rs, 6)]), ms, &(ri[0])); T10 = LD(&(ii[WS(rs, 6)]), ms, &(ii[0])); } TV = VFMA(TT, TU, TR); T2a = VFNMS(TT, TQ, T29); } } } } { V T36, Ty, T3B, T3G, T1s, T2m, T2t, T3d, TY, T2b, T3g, T1S, T3s, T3j; T36 = VSUB(Te, Tx); Ty = VADD(Te, Tx); T3B = VADD(T3w, T3A); T3G = VSUB(T3A, T3w); T1s = VADD(T1l, T1r); T2m = VSUB(T1l, T1r); T2t = VSUB(T2q, T2s); T3d = VADD(T2q, T2s); TY = VMUL(TW, TX); T2b = VMUL(TW, T10); T3g = VSUB(T1G, T1R); T1S = VADD(T1G, T1R); T3s = VADD(T3h, T3i); T3j = VSUB(T3h, T3i); { V T3D, T1T, T3u, T3t, T28, T12, T38, T2d, T3n, T3f; { V T1t, T3b, T3e, T3r, T11, T2c; T1t = VADD(T1d, T1s); T3b = VSUB(T1d, T1s); T3e = VSUB(T3c, T3d); T3r = VADD(T3c, T3d); T11 = VFMA(TZ, T10, TY); T2c = VFNMS(TZ, TX, T2b); T3D = VSUB(T1S, T1t); T1T = VADD(T1t, T1S); T3u = VADD(T3r, T3s); T3t = VSUB(T3r, T3s); T28 = VSUB(TV, T11); T12 = VADD(TV, T11); T38 = VADD(T2a, T2c); T2d = VSUB(T2a, T2c); T3n = VSUB(T3e, T3b); T3f = VADD(T3b, T3e); } { V T2Q, T20, T3N, T3T, T2J, T2C, T2W, T2V, T3O, T2f, T3U, T2T; { V T2R, T27, T2e, T2S, T13, T3F; T2Q = VADD(T1U, T1Z); T20 = VSUB(T1U, T1Z); T3N = VSUB(T3L, T3M); T3T = VADD(T3M, T3L); T13 = VADD(TL, T12); T3F = VSUB(T12, TL); { V T3v, T39, T3o, T3k; T3v = VADD(T37, T38); T39 = VSUB(T37, T38); T3o = VADD(T3g, T3j); T3k = VSUB(T3g, T3j); { V T3H, T3J, T14, T3q; T3H = VADD(T3F, T3G); T3J = VSUB(T3G, T3F); T14 = VADD(Ty, T13); T3q = VSUB(Ty, T13); { V T3a, T3m, T3C, T3E; T3a = VADD(T36, T39); T3m = VSUB(T36, T39); T3C = VADD(T3v, T3B); T3E = VSUB(T3B, T3v); { V T3I, T3p, T3l, T3K; T3I = VADD(T3n, T3o); T3p = VSUB(T3n, T3o); T3l = VADD(T3f, T3k); T3K = VSUB(T3k, T3f); ST(&(ri[WS(rs, 4)]), VADD(T3q, T3t), ms, &(ri[0])); ST(&(ri[WS(rs, 12)]), VSUB(T3q, T3t), ms, &(ri[0])); ST(&(ri[0]), VADD(T14, T1T), ms, &(ri[0])); ST(&(ri[WS(rs, 8)]), VSUB(T14, T1T), ms, &(ri[0])); ST(&(ii[WS(rs, 4)]), VADD(T3D, T3E), ms, &(ii[0])); ST(&(ii[WS(rs, 12)]), VSUB(T3E, T3D), ms, &(ii[0])); ST(&(ii[0]), VADD(T3u, T3C), ms, &(ii[0])); ST(&(ii[WS(rs, 8)]), VSUB(T3C, T3u), ms, &(ii[0])); ST(&(ri[WS(rs, 6)]), VFMA(LDK(KP707106781), T3p, T3m), ms, &(ri[0])); ST(&(ri[WS(rs, 14)]), VFNMS(LDK(KP707106781), T3p, T3m), ms, &(ri[0])); ST(&(ii[WS(rs, 10)]), VFNMS(LDK(KP707106781), T3I, T3H), ms, &(ii[0])); ST(&(ii[WS(rs, 2)]), VFMA(LDK(KP707106781), T3I, T3H), ms, &(ii[0])); ST(&(ii[WS(rs, 14)]), VFNMS(LDK(KP707106781), T3K, T3J), ms, &(ii[0])); ST(&(ii[WS(rs, 6)]), VFMA(LDK(KP707106781), T3K, T3J), ms, &(ii[0])); ST(&(ri[WS(rs, 2)]), VFMA(LDK(KP707106781), T3l, T3a), ms, &(ri[0])); ST(&(ri[WS(rs, 10)]), VFNMS(LDK(KP707106781), T3l, T3a), ms, &(ri[0])); T2R = VADD(T26, T25); T27 = VSUB(T25, T26); T2e = VADD(T28, T2d); T2S = VSUB(T28, T2d); } } } } { V T2Y, T2Z, T2n, T2u; T2J = VSUB(T2D, T2I); T2Y = VADD(T2D, T2I); T2Z = VSUB(T2A, T2B); T2C = VADD(T2A, T2B); T2W = VSUB(T2l, T2m); T2n = VADD(T2l, T2m); T2u = VSUB(T2o, T2t); T2V = VADD(T2o, T2t); T3O = VADD(T27, T2e); T2f = VSUB(T27, T2e); T34 = VFMA(LDK(KP414213562), T2Y, T2Z); T30 = VFNMS(LDK(KP414213562), T2Z, T2Y); T3U = VSUB(T2S, T2R); T2T = VADD(T2R, T2S); T2N = VFNMS(LDK(KP414213562), T2n, T2u); T2v = VFMA(LDK(KP414213562), T2u, T2n); } } T2M = VFNMS(LDK(KP707106781), T2f, T20); T2g = VFMA(LDK(KP707106781), T2f, T20); T3V = VFMA(LDK(KP707106781), T3U, T3T); T3X = VFNMS(LDK(KP707106781), T3U, T3T); T32 = VFNMS(LDK(KP707106781), T2T, T2Q); T2U = VFMA(LDK(KP707106781), T2T, T2Q); T33 = VFNMS(LDK(KP414213562), T2V, T2W); T2X = VFMA(LDK(KP414213562), T2W, T2V); T2O = VFMA(LDK(KP414213562), T2C, T2J); T2K = VFNMS(LDK(KP414213562), T2J, T2C); T3P = VFMA(LDK(KP707106781), T3O, T3N); T3R = VFNMS(LDK(KP707106781), T3O, T3N); } } } } } { V T3Q, T35, T31, T3S; T3Q = VADD(T33, T34); T35 = VSUB(T33, T34); T31 = VADD(T2X, T30); T3S = VSUB(T30, T2X); { V T3W, T2P, T2L, T3Y; T3W = VSUB(T2O, T2N); T2P = VADD(T2N, T2O); T2L = VSUB(T2v, T2K); T3Y = VADD(T2v, T2K); ST(&(ri[WS(rs, 5)]), VFMA(LDK(KP923879532), T35, T32), ms, &(ri[WS(rs, 1)])); ST(&(ri[WS(rs, 13)]), VFNMS(LDK(KP923879532), T35, T32), ms, &(ri[WS(rs, 1)])); ST(&(ii[WS(rs, 9)]), VFNMS(LDK(KP923879532), T3Q, T3P), ms, &(ii[WS(rs, 1)])); ST(&(ii[WS(rs, 1)]), VFMA(LDK(KP923879532), T3Q, T3P), ms, &(ii[WS(rs, 1)])); ST(&(ii[WS(rs, 13)]), VFNMS(LDK(KP923879532), T3S, T3R), ms, &(ii[WS(rs, 1)])); ST(&(ii[WS(rs, 5)]), VFMA(LDK(KP923879532), T3S, T3R), ms, &(ii[WS(rs, 1)])); ST(&(ri[WS(rs, 1)]), VFMA(LDK(KP923879532), T31, T2U), ms, &(ri[WS(rs, 1)])); ST(&(ri[WS(rs, 9)]), VFNMS(LDK(KP923879532), T31, T2U), ms, &(ri[WS(rs, 1)])); ST(&(ri[WS(rs, 15)]), VFMA(LDK(KP923879532), T2P, T2M), ms, &(ri[WS(rs, 1)])); ST(&(ri[WS(rs, 7)]), VFNMS(LDK(KP923879532), T2P, T2M), ms, &(ri[WS(rs, 1)])); ST(&(ii[WS(rs, 11)]), VFNMS(LDK(KP923879532), T3W, T3V), ms, &(ii[WS(rs, 1)])); ST(&(ii[WS(rs, 3)]), VFMA(LDK(KP923879532), T3W, T3V), ms, &(ii[WS(rs, 1)])); ST(&(ii[WS(rs, 15)]), VFMA(LDK(KP923879532), T3Y, T3X), ms, &(ii[WS(rs, 1)])); ST(&(ii[WS(rs, 7)]), VFNMS(LDK(KP923879532), T3Y, T3X), ms, &(ii[WS(rs, 1)])); ST(&(ri[WS(rs, 3)]), VFMA(LDK(KP923879532), T2L, T2g), ms, &(ri[WS(rs, 1)])); ST(&(ri[WS(rs, 11)]), VFNMS(LDK(KP923879532), T2L, T2g), ms, &(ri[WS(rs, 1)])); } } } } VLEAVE(); } static const tw_instr twinstr[] = { VTW(0, 1), VTW(0, 3), VTW(0, 9), VTW(0, 15), {TW_NEXT, (2 * VL), 0} }; static const ct_desc desc = { 16, XSIMD_STRING("t2sv_16"), twinstr, &GENUS, {104, 42, 92, 0}, 0, 0, 0 }; void XSIMD(codelet_t2sv_16) (planner *p) { X(kdft_dit_register) (p, t2sv_16, &desc); } #else /* HAVE_FMA */ /* Generated by: ../../../genfft/gen_twiddle.native -simd -compact -variables 4 -pipeline-latency 8 -twiddle-log3 -precompute-twiddles -n 16 -name t2sv_16 -include ts.h */ /* * This function contains 196 FP additions, 108 FP multiplications, * (or, 156 additions, 68 multiplications, 40 fused multiply/add), * 82 stack variables, 3 constants, and 64 memory accesses */ #include "ts.h" static void t2sv_16(R *ri, R *ii, const R *W, stride rs, INT mb, INT me, INT ms) { DVK(KP382683432, +0.382683432365089771728459984030398866761344562); DVK(KP923879532, +0.923879532511286756128183189396788286822416626); DVK(KP707106781, +0.707106781186547524400844362104849039284835938); { INT m; for (m = mb, W = W + (mb * 8); m < me; m = m + (2 * VL), ri = ri + ((2 * VL) * ms), ii = ii + ((2 * VL) * ms), W = W + ((2 * VL) * 8), MAKE_VOLATILE_STRIDE(32, rs)) { V T2, T5, Tg, Ti, Tk, To, TE, TC, T6, T3, T8, TW, TJ, Tt, TU; V Tc, Tx, TH, TN, TO, TP, TR, T1f, T1k, T1b, T1i, T1y, T1H, T1u, T1F; { V T7, Tv, Ta, Ts, T4, Tw, Tb, Tr; { V Th, Tn, Tj, Tm; T2 = LDW(&(W[0])); T5 = LDW(&(W[TWVL * 1])); Tg = LDW(&(W[TWVL * 2])); Ti = LDW(&(W[TWVL * 3])); Th = VMUL(T2, Tg); Tn = VMUL(T5, Tg); Tj = VMUL(T5, Ti); Tm = VMUL(T2, Ti); Tk = VSUB(Th, Tj); To = VADD(Tm, Tn); TE = VSUB(Tm, Tn); TC = VADD(Th, Tj); T6 = LDW(&(W[TWVL * 5])); T7 = VMUL(T5, T6); Tv = VMUL(Tg, T6); Ta = VMUL(T2, T6); Ts = VMUL(Ti, T6); T3 = LDW(&(W[TWVL * 4])); T4 = VMUL(T2, T3); Tw = VMUL(Ti, T3); Tb = VMUL(T5, T3); Tr = VMUL(Tg, T3); } T8 = VADD(T4, T7); TW = VSUB(Tv, Tw); TJ = VADD(Ta, Tb); Tt = VSUB(Tr, Ts); TU = VADD(Tr, Ts); Tc = VSUB(Ta, Tb); Tx = VADD(Tv, Tw); TH = VSUB(T4, T7); TN = LDW(&(W[TWVL * 6])); TO = LDW(&(W[TWVL * 7])); TP = VFMA(T2, TN, VMUL(T5, TO)); TR = VFNMS(T5, TN, VMUL(T2, TO)); { V T1d, T1e, T19, T1a; T1d = VMUL(Tk, T6); T1e = VMUL(To, T3); T1f = VSUB(T1d, T1e); T1k = VADD(T1d, T1e); T19 = VMUL(Tk, T3); T1a = VMUL(To, T6); T1b = VADD(T19, T1a); T1i = VSUB(T19, T1a); } { V T1w, T1x, T1s, T1t; T1w = VMUL(TC, T6); T1x = VMUL(TE, T3); T1y = VSUB(T1w, T1x); T1H = VADD(T1w, T1x); T1s = VMUL(TC, T3); T1t = VMUL(TE, T6); T1u = VADD(T1s, T1t); T1F = VSUB(T1s, T1t); } } { V Tf, T3r, T1N, T3e, TA, T3s, T1Q, T3b, TM, T2M, T1W, T2w, TZ, T2N, T21; V T2x, T1B, T1K, T2V, T2W, T2X, T2Y, T2j, T2D, T2o, T2E, T18, T1n, T2Q, T2R; V T2S, T2T, T28, T2A, T2d, T2B; { V T1, T3d, Te, T3c, T9, Td; T1 = LD(&(ri[0]), ms, &(ri[0])); T3d = LD(&(ii[0]), ms, &(ii[0])); T9 = LD(&(ri[WS(rs, 8)]), ms, &(ri[0])); Td = LD(&(ii[WS(rs, 8)]), ms, &(ii[0])); Te = VFMA(T8, T9, VMUL(Tc, Td)); T3c = VFNMS(Tc, T9, VMUL(T8, Td)); Tf = VADD(T1, Te); T3r = VSUB(T3d, T3c); T1N = VSUB(T1, Te); T3e = VADD(T3c, T3d); } { V Tq, T1O, Tz, T1P; { V Tl, Tp, Tu, Ty; Tl = LD(&(ri[WS(rs, 4)]), ms, &(ri[0])); Tp = LD(&(ii[WS(rs, 4)]), ms, &(ii[0])); Tq = VFMA(Tk, Tl, VMUL(To, Tp)); T1O = VFNMS(To, Tl, VMUL(Tk, Tp)); Tu = LD(&(ri[WS(rs, 12)]), ms, &(ri[0])); Ty = LD(&(ii[WS(rs, 12)]), ms, &(ii[0])); Tz = VFMA(Tt, Tu, VMUL(Tx, Ty)); T1P = VFNMS(Tx, Tu, VMUL(Tt, Ty)); } TA = VADD(Tq, Tz); T3s = VSUB(Tq, Tz); T1Q = VSUB(T1O, T1P); T3b = VADD(T1O, T1P); } { V TG, T1S, TL, T1T, T1U, T1V; { V TD, TF, TI, TK; TD = LD(&(ri[WS(rs, 2)]), ms, &(ri[0])); TF = LD(&(ii[WS(rs, 2)]), ms, &(ii[0])); TG = VFMA(TC, TD, VMUL(TE, TF)); T1S = VFNMS(TE, TD, VMUL(TC, TF)); TI = LD(&(ri[WS(rs, 10)]), ms, &(ri[0])); TK = LD(&(ii[WS(rs, 10)]), ms, &(ii[0])); TL = VFMA(TH, TI, VMUL(TJ, TK)); T1T = VFNMS(TJ, TI, VMUL(TH, TK)); } TM = VADD(TG, TL); T2M = VADD(T1S, T1T); T1U = VSUB(T1S, T1T); T1V = VSUB(TG, TL); T1W = VSUB(T1U, T1V); T2w = VADD(T1V, T1U); } { V TT, T1Y, TY, T1Z, T1X, T20; { V TQ, TS, TV, TX; TQ = LD(&(ri[WS(rs, 14)]), ms, &(ri[0])); TS = LD(&(ii[WS(rs, 14)]), ms, &(ii[0])); TT = VFMA(TP, TQ, VMUL(TR, TS)); T1Y = VFNMS(TR, TQ, VMUL(TP, TS)); TV = LD(&(ri[WS(rs, 6)]), ms, &(ri[0])); TX = LD(&(ii[WS(rs, 6)]), ms, &(ii[0])); TY = VFMA(TU, TV, VMUL(TW, TX)); T1Z = VFNMS(TW, TV, VMUL(TU, TX)); } TZ = VADD(TT, TY); T2N = VADD(T1Y, T1Z); T1X = VSUB(TT, TY); T20 = VSUB(T1Y, T1Z); T21 = VADD(T1X, T20); T2x = VSUB(T1X, T20); } { V T1r, T2k, T1J, T2h, T1A, T2l, T1E, T2g; { V T1p, T1q, T1G, T1I; T1p = LD(&(ri[WS(rs, 15)]), ms, &(ri[WS(rs, 1)])); T1q = LD(&(ii[WS(rs, 15)]), ms, &(ii[WS(rs, 1)])); T1r = VFMA(TN, T1p, VMUL(TO, T1q)); T2k = VFNMS(TO, T1p, VMUL(TN, T1q)); T1G = LD(&(ri[WS(rs, 11)]), ms, &(ri[WS(rs, 1)])); T1I = LD(&(ii[WS(rs, 11)]), ms, &(ii[WS(rs, 1)])); T1J = VFMA(T1F, T1G, VMUL(T1H, T1I)); T2h = VFNMS(T1H, T1G, VMUL(T1F, T1I)); } { V T1v, T1z, T1C, T1D; T1v = LD(&(ri[WS(rs, 7)]), ms, &(ri[WS(rs, 1)])); T1z = LD(&(ii[WS(rs, 7)]), ms, &(ii[WS(rs, 1)])); T1A = VFMA(T1u, T1v, VMUL(T1y, T1z)); T2l = VFNMS(T1y, T1v, VMUL(T1u, T1z)); T1C = LD(&(ri[WS(rs, 3)]), ms, &(ri[WS(rs, 1)])); T1D = LD(&(ii[WS(rs, 3)]), ms, &(ii[WS(rs, 1)])); T1E = VFMA(Tg, T1C, VMUL(Ti, T1D)); T2g = VFNMS(Ti, T1C, VMUL(Tg, T1D)); } T1B = VADD(T1r, T1A); T1K = VADD(T1E, T1J); T2V = VSUB(T1B, T1K); T2W = VADD(T2k, T2l); T2X = VADD(T2g, T2h); T2Y = VSUB(T2W, T2X); { V T2f, T2i, T2m, T2n; T2f = VSUB(T1r, T1A); T2i = VSUB(T2g, T2h); T2j = VSUB(T2f, T2i); T2D = VADD(T2f, T2i); T2m = VSUB(T2k, T2l); T2n = VSUB(T1E, T1J); T2o = VADD(T2m, T2n); T2E = VSUB(T2m, T2n); } } { V T14, T24, T1m, T2b, T17, T25, T1h, T2a; { V T12, T13, T1j, T1l; T12 = LD(&(ri[WS(rs, 1)]), ms, &(ri[WS(rs, 1)])); T13 = LD(&(ii[WS(rs, 1)]), ms, &(ii[WS(rs, 1)])); T14 = VFMA(T2, T12, VMUL(T5, T13)); T24 = VFNMS(T5, T12, VMUL(T2, T13)); T1j = LD(&(ri[WS(rs, 13)]), ms, &(ri[WS(rs, 1)])); T1l = LD(&(ii[WS(rs, 13)]), ms, &(ii[WS(rs, 1)])); T1m = VFMA(T1i, T1j, VMUL(T1k, T1l)); T2b = VFNMS(T1k, T1j, VMUL(T1i, T1l)); } { V T15, T16, T1c, T1g; T15 = LD(&(ri[WS(rs, 9)]), ms, &(ri[WS(rs, 1)])); T16 = LD(&(ii[WS(rs, 9)]), ms, &(ii[WS(rs, 1)])); T17 = VFMA(T3, T15, VMUL(T6, T16)); T25 = VFNMS(T6, T15, VMUL(T3, T16)); T1c = LD(&(ri[WS(rs, 5)]), ms, &(ri[WS(rs, 1)])); T1g = LD(&(ii[WS(rs, 5)]), ms, &(ii[WS(rs, 1)])); T1h = VFMA(T1b, T1c, VMUL(T1f, T1g)); T2a = VFNMS(T1f, T1c, VMUL(T1b, T1g)); } T18 = VADD(T14, T17); T1n = VADD(T1h, T1m); T2Q = VSUB(T18, T1n); T2R = VADD(T24, T25); T2S = VADD(T2a, T2b); T2T = VSUB(T2R, T2S); { V T26, T27, T29, T2c; T26 = VSUB(T24, T25); T27 = VSUB(T1h, T1m); T28 = VADD(T26, T27); T2A = VSUB(T26, T27); T29 = VSUB(T14, T17); T2c = VSUB(T2a, T2b); T2d = VSUB(T29, T2c); T2B = VADD(T29, T2c); } } { V T23, T2r, T3A, T3C, T2q, T3B, T2u, T3x; { V T1R, T22, T3y, T3z; T1R = VSUB(T1N, T1Q); T22 = VMUL(LDK(KP707106781), VSUB(T1W, T21)); T23 = VADD(T1R, T22); T2r = VSUB(T1R, T22); T3y = VMUL(LDK(KP707106781), VSUB(T2x, T2w)); T3z = VADD(T3s, T3r); T3A = VADD(T3y, T3z); T3C = VSUB(T3z, T3y); } { V T2e, T2p, T2s, T2t; T2e = VFMA(LDK(KP923879532), T28, VMUL(LDK(KP382683432), T2d)); T2p = VFNMS(LDK(KP923879532), T2o, VMUL(LDK(KP382683432), T2j)); T2q = VADD(T2e, T2p); T3B = VSUB(T2p, T2e); T2s = VFNMS(LDK(KP923879532), T2d, VMUL(LDK(KP382683432), T28)); T2t = VFMA(LDK(KP382683432), T2o, VMUL(LDK(KP923879532), T2j)); T2u = VSUB(T2s, T2t); T3x = VADD(T2s, T2t); } ST(&(ri[WS(rs, 11)]), VSUB(T23, T2q), ms, &(ri[WS(rs, 1)])); ST(&(ii[WS(rs, 11)]), VSUB(T3A, T3x), ms, &(ii[WS(rs, 1)])); ST(&(ri[WS(rs, 3)]), VADD(T23, T2q), ms, &(ri[WS(rs, 1)])); ST(&(ii[WS(rs, 3)]), VADD(T3x, T3A), ms, &(ii[WS(rs, 1)])); ST(&(ri[WS(rs, 15)]), VSUB(T2r, T2u), ms, &(ri[WS(rs, 1)])); ST(&(ii[WS(rs, 15)]), VSUB(T3C, T3B), ms, &(ii[WS(rs, 1)])); ST(&(ri[WS(rs, 7)]), VADD(T2r, T2u), ms, &(ri[WS(rs, 1)])); ST(&(ii[WS(rs, 7)]), VADD(T3B, T3C), ms, &(ii[WS(rs, 1)])); } { V T2P, T31, T3m, T3o, T30, T3n, T34, T3j; { V T2L, T2O, T3k, T3l; T2L = VSUB(Tf, TA); T2O = VSUB(T2M, T2N); T2P = VADD(T2L, T2O); T31 = VSUB(T2L, T2O); T3k = VSUB(TZ, TM); T3l = VSUB(T3e, T3b); T3m = VADD(T3k, T3l); T3o = VSUB(T3l, T3k); } { V T2U, T2Z, T32, T33; T2U = VADD(T2Q, T2T); T2Z = VSUB(T2V, T2Y); T30 = VMUL(LDK(KP707106781), VADD(T2U, T2Z)); T3n = VMUL(LDK(KP707106781), VSUB(T2Z, T2U)); T32 = VSUB(T2T, T2Q); T33 = VADD(T2V, T2Y); T34 = VMUL(LDK(KP707106781), VSUB(T32, T33)); T3j = VMUL(LDK(KP707106781), VADD(T32, T33)); } ST(&(ri[WS(rs, 10)]), VSUB(T2P, T30), ms, &(ri[0])); ST(&(ii[WS(rs, 10)]), VSUB(T3m, T3j), ms, &(ii[0])); ST(&(ri[WS(rs, 2)]), VADD(T2P, T30), ms, &(ri[0])); ST(&(ii[WS(rs, 2)]), VADD(T3j, T3m), ms, &(ii[0])); ST(&(ri[WS(rs, 14)]), VSUB(T31, T34), ms, &(ri[0])); ST(&(ii[WS(rs, 14)]), VSUB(T3o, T3n), ms, &(ii[0])); ST(&(ri[WS(rs, 6)]), VADD(T31, T34), ms, &(ri[0])); ST(&(ii[WS(rs, 6)]), VADD(T3n, T3o), ms, &(ii[0])); } { V T2z, T2H, T3u, T3w, T2G, T3v, T2K, T3p; { V T2v, T2y, T3q, T3t; T2v = VADD(T1N, T1Q); T2y = VMUL(LDK(KP707106781), VADD(T2w, T2x)); T2z = VADD(T2v, T2y); T2H = VSUB(T2v, T2y); T3q = VMUL(LDK(KP707106781), VADD(T1W, T21)); T3t = VSUB(T3r, T3s); T3u = VADD(T3q, T3t); T3w = VSUB(T3t, T3q); } { V T2C, T2F, T2I, T2J; T2C = VFMA(LDK(KP382683432), T2A, VMUL(LDK(KP923879532), T2B)); T2F = VFNMS(LDK(KP382683432), T2E, VMUL(LDK(KP923879532), T2D)); T2G = VADD(T2C, T2F); T3v = VSUB(T2F, T2C); T2I = VFNMS(LDK(KP382683432), T2B, VMUL(LDK(KP923879532), T2A)); T2J = VFMA(LDK(KP923879532), T2E, VMUL(LDK(KP382683432), T2D)); T2K = VSUB(T2I, T2J); T3p = VADD(T2I, T2J); } ST(&(ri[WS(rs, 9)]), VSUB(T2z, T2G), ms, &(ri[WS(rs, 1)])); ST(&(ii[WS(rs, 9)]), VSUB(T3u, T3p), ms, &(ii[WS(rs, 1)])); ST(&(ri[WS(rs, 1)]), VADD(T2z, T2G), ms, &(ri[WS(rs, 1)])); ST(&(ii[WS(rs, 1)]), VADD(T3p, T3u), ms, &(ii[WS(rs, 1)])); ST(&(ri[WS(rs, 13)]), VSUB(T2H, T2K), ms, &(ri[WS(rs, 1)])); ST(&(ii[WS(rs, 13)]), VSUB(T3w, T3v), ms, &(ii[WS(rs, 1)])); ST(&(ri[WS(rs, 5)]), VADD(T2H, T2K), ms, &(ri[WS(rs, 1)])); ST(&(ii[WS(rs, 5)]), VADD(T3v, T3w), ms, &(ii[WS(rs, 1)])); } { V T11, T35, T3g, T3i, T1M, T3h, T38, T39; { V TB, T10, T3a, T3f; TB = VADD(Tf, TA); T10 = VADD(TM, TZ); T11 = VADD(TB, T10); T35 = VSUB(TB, T10); T3a = VADD(T2M, T2N); T3f = VADD(T3b, T3e); T3g = VADD(T3a, T3f); T3i = VSUB(T3f, T3a); } { V T1o, T1L, T36, T37; T1o = VADD(T18, T1n); T1L = VADD(T1B, T1K); T1M = VADD(T1o, T1L); T3h = VSUB(T1L, T1o); T36 = VADD(T2R, T2S); T37 = VADD(T2W, T2X); T38 = VSUB(T36, T37); T39 = VADD(T36, T37); } ST(&(ri[WS(rs, 8)]), VSUB(T11, T1M), ms, &(ri[0])); ST(&(ii[WS(rs, 8)]), VSUB(T3g, T39), ms, &(ii[0])); ST(&(ri[0]), VADD(T11, T1M), ms, &(ri[0])); ST(&(ii[0]), VADD(T39, T3g), ms, &(ii[0])); ST(&(ri[WS(rs, 12)]), VSUB(T35, T38), ms, &(ri[0])); ST(&(ii[WS(rs, 12)]), VSUB(T3i, T3h), ms, &(ii[0])); ST(&(ri[WS(rs, 4)]), VADD(T35, T38), ms, &(ri[0])); ST(&(ii[WS(rs, 4)]), VADD(T3h, T3i), ms, &(ii[0])); } } } } VLEAVE(); } static const tw_instr twinstr[] = { VTW(0, 1), VTW(0, 3), VTW(0, 9), VTW(0, 15), {TW_NEXT, (2 * VL), 0} }; static const ct_desc desc = { 16, XSIMD_STRING("t2sv_16"), twinstr, &GENUS, {156, 68, 40, 0}, 0, 0, 0 }; void XSIMD(codelet_t2sv_16) (planner *p) { X(kdft_dit_register) (p, t2sv_16, &desc); } #endif /* HAVE_FMA */ fftw-3.3.3/dft/simd/common/t2sv_32.c0000644000175000017500000017767012054410611013772 00000000000000/* * Copyright (c) 2003, 2007-11 Matteo Frigo * Copyright (c) 2003, 2007-11 Massachusetts Institute of Technology * * This program is free software; you can redistribute it and/or modify * it under the terms of the GNU General Public License as published by * the Free Software Foundation; either version 2 of the License, or * (at your option) any later version. * * This program is distributed in the hope that it will be useful, * but WITHOUT ANY WARRANTY; without even the implied warranty of * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the * GNU General Public License for more details. * * You should have received a copy of the GNU General Public License * along with this program; if not, write to the Free Software * Foundation, Inc., 51 Franklin Street, Fifth Floor, Boston, MA 02110-1301 USA * */ /* This file was automatically generated --- DO NOT EDIT */ /* Generated on Sun Nov 25 07:39:28 EST 2012 */ #include "codelet-dft.h" #ifdef HAVE_FMA /* Generated by: ../../../genfft/gen_twiddle.native -fma -reorder-insns -schedule-for-pipeline -simd -compact -variables 4 -pipeline-latency 8 -twiddle-log3 -precompute-twiddles -n 32 -name t2sv_32 -include ts.h */ /* * This function contains 488 FP additions, 350 FP multiplications, * (or, 236 additions, 98 multiplications, 252 fused multiply/add), * 204 stack variables, 7 constants, and 128 memory accesses */ #include "ts.h" static void t2sv_32(R *ri, R *ii, const R *W, stride rs, INT mb, INT me, INT ms) { DVK(KP831469612, +0.831469612302545237078788377617905756738560812); DVK(KP980785280, +0.980785280403230449126182236134239036973933731); DVK(KP668178637, +0.668178637919298919997757686523080761552472251); DVK(KP198912367, +0.198912367379658006911597622644676228597850501); DVK(KP923879532, +0.923879532511286756128183189396788286822416626); DVK(KP414213562, +0.414213562373095048801688724209698078569671875); DVK(KP707106781, +0.707106781186547524400844362104849039284835938); { INT m; for (m = mb, W = W + (mb * 8); m < me; m = m + (2 * VL), ri = ri + ((2 * VL) * ms), ii = ii + ((2 * VL) * ms), W = W + ((2 * VL) * 8), MAKE_VOLATILE_STRIDE(64, rs)) { V T6H, T74, T6U, T6E, T9r, T9t, T78, T7c, T6W, T6S, T73, T6K, T7a, T72, T9x; V T9z; { V T2, T8, T3, T6, Te, Ti, T5, Tc; T2 = LDW(&(W[0])); T8 = LDW(&(W[TWVL * 4])); T3 = LDW(&(W[TWVL * 2])); T6 = LDW(&(W[TWVL * 3])); Te = LDW(&(W[TWVL * 6])); Ti = LDW(&(W[TWVL * 7])); T5 = LDW(&(W[TWVL * 1])); Tc = LDW(&(W[TWVL * 5])); { V T2X, T2T, T34, T31, Tq, T46, T97, T8H, TH, T98, T4b, T8D, TZ, T7f, T1g; V T7g, T4j, T6t, T4q, T6u, T6x, T4z, T7m, T1J, T4G, T6y, T8d, T7l, T4O, T6A; V T2k, T7o, T6B, T4V, T7r, T8e, T5E, T6P, T3G, T7L, T6M, T61, T8n, T7I, T55; V T6I, T2N, T7A, T5s, T6F, T7x, T8i, T2R, T2U, T57, T3a, T5h, T62, T5L, T7J; V T43, T63, T5S, T8o, T7O, T2V, T2Y, T32, T35; { V T1w, T23, T1K, T1F, T1s, T1N, T26, T1z, T2w, T2s, T3Q, T3M, T3r, T3n, T2b; V T1U, T3C, T3j, T3z, T3f, T1R, T29, TR, Th, T2J, T2F, Td, TP, T1Z, T1V; V T2g, T2c, T1m, T4u, T1D, T1G, T1p, T1t, T1E, T4D, T1x, T1A, T1q, T4v; { V T1, Ts, T19, TJ, T7, TM, Tb, T11, T1C, T1o, TA, T15, TE, T1d, Tw; V T8G, Tk, Tn, Tj, TW, TS, To, Tt, Tx, TB, TF, Tl; { V T1Y, T1S, T2f, T2a; T1 = LD(&(ri[0]), ms, &(ri[0])); { V Tr, T18, T4, Ta; Tr = VMUL(T2, T8); T18 = VMUL(T3, T8); T4 = VMUL(T2, T3); Ta = VMUL(T2, T6); { V T10, T1n, Tz, T14; T10 = VMUL(T2, Te); T1n = VMUL(T8, Te); Tz = VMUL(T3, Te); T14 = VMUL(T2, Ti); { V T1r, TD, T1c, Tv; T1r = VMUL(T8, Ti); TD = VMUL(T3, Ti); T1c = VMUL(T3, Tc); Tv = VMUL(T2, Tc); T1w = VFNMS(T5, Tc, Tr); Ts = VFMA(T5, Tc, Tr); T19 = VFNMS(T6, Tc, T18); T23 = VFMA(T6, Tc, T18); TJ = VFNMS(T5, T6, T4); T7 = VFMA(T5, T6, T4); TM = VFMA(T5, T3, Ta); Tb = VFNMS(T5, T3, Ta); T11 = VFNMS(T5, Ti, T10); T1C = VFMA(T5, Ti, T10); T1o = VFMA(Tc, Ti, T1n); TA = VFMA(T6, Ti, Tz); T1K = VFNMS(T6, Ti, Tz); T1F = VFNMS(T5, Te, T14); T15 = VFMA(T5, Te, T14); T1s = VFNMS(Tc, Te, T1r); T1N = VFMA(T6, Te, TD); TE = VFNMS(T6, Te, TD); T26 = VFNMS(T6, T8, T1c); T1d = VFMA(T6, T8, T1c); T1z = VFMA(T5, T8, Tv); Tw = VFNMS(T5, T8, Tv); { V T2v, T2r, T3P, T3L; T2v = VMUL(T1w, Ti); T2r = VMUL(T1w, Te); T3P = VMUL(Ts, Ti); T3L = VMUL(Ts, Te); { V T3q, T3m, T2W, T2S; T3q = VMUL(T19, Ti); T3m = VMUL(T19, Te); T2W = VMUL(T23, Ti); T2S = VMUL(T23, Te); { V T1T, T3i, T3e, T1Q; T1T = VMUL(TJ, Tc); T3i = VMUL(TJ, Ti); T3e = VMUL(TJ, Te); T1Q = VMUL(TJ, T8); { V Tg, T2I, T2E, T9; Tg = VMUL(T7, Tc); T2I = VMUL(T7, Ti); T2E = VMUL(T7, Te); T9 = VMUL(T7, T8); T2w = VFNMS(T1z, Te, T2v); T2s = VFMA(T1z, Ti, T2r); T3Q = VFNMS(Tw, Te, T3P); T3M = VFMA(Tw, Ti, T3L); T3r = VFNMS(T1d, Te, T3q); T3n = VFMA(T1d, Ti, T3m); T2X = VFNMS(T26, Te, T2W); T2T = VFMA(T26, Ti, T2S); T2b = VFNMS(TM, T8, T1T); T1U = VFMA(TM, T8, T1T); T3C = VFNMS(TM, Te, T3i); T3j = VFMA(TM, Te, T3i); T3z = VFMA(TM, Ti, T3e); T3f = VFNMS(TM, Ti, T3e); T1R = VFNMS(TM, Tc, T1Q); T29 = VFMA(TM, Tc, T1Q); TR = VFNMS(Tb, T8, Tg); Th = VFMA(Tb, T8, Tg); T34 = VFMA(Tb, Te, T2I); T2J = VFNMS(Tb, Te, T2I); T31 = VFNMS(Tb, Ti, T2E); T2F = VFMA(Tb, Ti, T2E); Td = VFNMS(Tb, Tc, T9); TP = VFMA(Tb, Tc, T9); T1Y = VMUL(T1R, Ti); T1S = VMUL(T1R, Te); T2f = VMUL(T29, Ti); T2a = VMUL(T29, Te); T8G = LD(&(ii[0]), ms, &(ii[0])); } } } } } } } Tk = LD(&(ri[WS(rs, 16)]), ms, &(ri[0])); { V Tm, Tf, TV, TQ; Tm = VMUL(Td, Ti); Tf = VMUL(Td, Te); TV = VMUL(TP, Ti); TQ = VMUL(TP, Te); T1Z = VFNMS(T1U, Te, T1Y); T1V = VFMA(T1U, Ti, T1S); T2g = VFNMS(T2b, Te, T2f); T2c = VFMA(T2b, Ti, T2a); Tn = VFNMS(Th, Te, Tm); Tj = VFMA(Th, Ti, Tf); TW = VFNMS(TR, Te, TV); TS = VFMA(TR, Ti, TQ); } To = LD(&(ii[WS(rs, 16)]), ms, &(ii[0])); } Tt = LD(&(ri[WS(rs, 8)]), ms, &(ri[0])); Tx = LD(&(ii[WS(rs, 8)]), ms, &(ii[0])); TB = LD(&(ri[WS(rs, 24)]), ms, &(ri[0])); TF = LD(&(ii[WS(rs, 24)]), ms, &(ii[0])); Tl = VMUL(Tj, Tk); { V TO, T4f, TT, TX; { V Ty, T48, TG, T4a; { V TK, TN, T8E, Tu, T47, TC, T49, Tp, TL, T4e, T8F; TK = LD(&(ri[WS(rs, 4)]), ms, &(ri[0])); TN = LD(&(ii[WS(rs, 4)]), ms, &(ii[0])); T8E = VMUL(Tj, To); Tu = VMUL(Ts, Tt); T47 = VMUL(Ts, Tx); TC = VMUL(TA, TB); T49 = VMUL(TA, TF); Tp = VFMA(Tn, To, Tl); TL = VMUL(TJ, TK); T4e = VMUL(TJ, TN); T8F = VFNMS(Tn, Tk, T8E); Ty = VFMA(Tw, Tx, Tu); T48 = VFNMS(Tw, Tt, T47); TG = VFMA(TE, TF, TC); T4a = VFNMS(TE, TB, T49); Tq = VADD(T1, Tp); T46 = VSUB(T1, Tp); TO = VFMA(TM, TN, TL); T97 = VSUB(T8G, T8F); T8H = VADD(T8F, T8G); T4f = VFNMS(TM, TK, T4e); } TH = VADD(Ty, TG); T98 = VSUB(Ty, TG); T4b = VSUB(T48, T4a); T8D = VADD(T48, T4a); TT = LD(&(ri[WS(rs, 20)]), ms, &(ri[0])); TX = LD(&(ii[WS(rs, 20)]), ms, &(ii[0])); } { V T12, T16, T1a, T1e, T4k, T4p; T12 = LD(&(ri[WS(rs, 28)]), ms, &(ri[0])); T16 = LD(&(ii[WS(rs, 28)]), ms, &(ii[0])); T1a = LD(&(ri[WS(rs, 12)]), ms, &(ri[0])); T1e = LD(&(ii[WS(rs, 12)]), ms, &(ii[0])); { V TY, T4h, T17, T4m, T1f, T4o, T4d, T4i; { V T1j, T1l, TU, T4g, T13, T4l, T1b, T4n, T1k, T4t; T1j = LD(&(ri[WS(rs, 2)]), ms, &(ri[0])); T1l = LD(&(ii[WS(rs, 2)]), ms, &(ii[0])); TU = VMUL(TS, TT); T4g = VMUL(TS, TX); T13 = VMUL(T11, T12); T4l = VMUL(T11, T16); T1b = VMUL(T19, T1a); T4n = VMUL(T19, T1e); T1k = VMUL(T7, T1j); T4t = VMUL(T7, T1l); TY = VFMA(TW, TX, TU); T4h = VFNMS(TW, TT, T4g); T17 = VFMA(T15, T16, T13); T4m = VFNMS(T15, T12, T4l); T1f = VFMA(T1d, T1e, T1b); T4o = VFNMS(T1d, T1a, T4n); T1m = VFMA(Tb, T1l, T1k); T4u = VFNMS(Tb, T1j, T4t); } TZ = VADD(TO, TY); T4d = VSUB(TO, TY); T7f = VADD(T4f, T4h); T4i = VSUB(T4f, T4h); T1g = VADD(T17, T1f); T4k = VSUB(T17, T1f); T7g = VADD(T4m, T4o); T4p = VSUB(T4m, T4o); T1D = LD(&(ri[WS(rs, 26)]), ms, &(ri[0])); T1G = LD(&(ii[WS(rs, 26)]), ms, &(ii[0])); T4j = VADD(T4d, T4i); T6t = VSUB(T4i, T4d); } T1p = LD(&(ri[WS(rs, 18)]), ms, &(ri[0])); T1t = LD(&(ii[WS(rs, 18)]), ms, &(ii[0])); T4q = VSUB(T4k, T4p); T6u = VADD(T4k, T4p); T1E = VMUL(T1C, T1D); T4D = VMUL(T1C, T1G); T1x = LD(&(ri[WS(rs, 10)]), ms, &(ri[0])); T1A = LD(&(ii[WS(rs, 10)]), ms, &(ii[0])); T1q = VMUL(T1o, T1p); T4v = VMUL(T1o, T1t); } } } { V T3l, T5z, T3E, T5Z, T3v, T3x, T3w, T3t, T5B, T5W; { V T1P, T4J, T1W, T20, T2i, T4T, T1X, T4K, T24, T27; { V T2d, T2h, T1v, T4A, T7j, T4x, T2e, T4y, T1I, T4F, T7k, T4S; { V T1L, T1O, T1H, T4E, T1y, T4B, T1u, T4w, T1M, T4I, T1B, T4C; T1L = LD(&(ri[WS(rs, 30)]), ms, &(ri[0])); T1O = LD(&(ii[WS(rs, 30)]), ms, &(ii[0])); T1H = VFMA(T1F, T1G, T1E); T4E = VFNMS(T1F, T1D, T4D); T1y = VMUL(T1w, T1x); T4B = VMUL(T1w, T1A); T1u = VFMA(T1s, T1t, T1q); T4w = VFNMS(T1s, T1p, T4v); T1M = VMUL(T1K, T1L); T4I = VMUL(T1K, T1O); T2d = LD(&(ri[WS(rs, 22)]), ms, &(ri[0])); T2h = LD(&(ii[WS(rs, 22)]), ms, &(ii[0])); T1B = VFMA(T1z, T1A, T1y); T4C = VFNMS(T1z, T1x, T4B); T1v = VADD(T1m, T1u); T4A = VSUB(T1m, T1u); T7j = VADD(T4u, T4w); T4x = VSUB(T4u, T4w); T1P = VFMA(T1N, T1O, T1M); T4J = VFNMS(T1N, T1L, T4I); T2e = VMUL(T2c, T2d); T4y = VSUB(T1B, T1H); T1I = VADD(T1B, T1H); T4F = VSUB(T4C, T4E); T7k = VADD(T4C, T4E); T4S = VMUL(T2c, T2h); } T1W = LD(&(ri[WS(rs, 14)]), ms, &(ri[0])); T20 = LD(&(ii[WS(rs, 14)]), ms, &(ii[0])); T2i = VFMA(T2g, T2h, T2e); T6x = VADD(T4x, T4y); T4z = VSUB(T4x, T4y); T7m = VSUB(T1v, T1I); T1J = VADD(T1v, T1I); T4G = VADD(T4A, T4F); T6y = VSUB(T4A, T4F); T8d = VADD(T7j, T7k); T7l = VSUB(T7j, T7k); T4T = VFNMS(T2g, T2d, T4S); T1X = VMUL(T1V, T1W); T4K = VMUL(T1V, T20); T24 = LD(&(ri[WS(rs, 6)]), ms, &(ri[0])); T27 = LD(&(ii[WS(rs, 6)]), ms, &(ii[0])); } { V T22, T4P, T7p, T4M, T28, T4R, T3g, T3k; T3g = LD(&(ri[WS(rs, 31)]), ms, &(ri[WS(rs, 1)])); T3k = LD(&(ii[WS(rs, 31)]), ms, &(ii[WS(rs, 1)])); { V T3A, T3D, T21, T4L, T25, T4Q, T3h, T5y, T3B, T5Y; T3A = LD(&(ri[WS(rs, 23)]), ms, &(ri[WS(rs, 1)])); T3D = LD(&(ii[WS(rs, 23)]), ms, &(ii[WS(rs, 1)])); T21 = VFMA(T1Z, T20, T1X); T4L = VFNMS(T1Z, T1W, T4K); T25 = VMUL(T23, T24); T4Q = VMUL(T23, T27); T3h = VMUL(T3f, T3g); T5y = VMUL(T3f, T3k); T3B = VMUL(T3z, T3A); T5Y = VMUL(T3z, T3D); T22 = VADD(T1P, T21); T4P = VSUB(T1P, T21); T7p = VADD(T4J, T4L); T4M = VSUB(T4J, T4L); T28 = VFMA(T26, T27, T25); T4R = VFNMS(T26, T24, T4Q); T3l = VFMA(T3j, T3k, T3h); T5z = VFNMS(T3j, T3g, T5y); T3E = VFMA(T3C, T3D, T3B); T5Z = VFNMS(T3C, T3A, T5Y); } { V T3o, T3s, T2j, T4N, T7q, T4U, T3p, T5A; T3o = LD(&(ri[WS(rs, 15)]), ms, &(ri[WS(rs, 1)])); T3s = LD(&(ii[WS(rs, 15)]), ms, &(ii[WS(rs, 1)])); T2j = VADD(T28, T2i); T4N = VSUB(T28, T2i); T7q = VADD(T4R, T4T); T4U = VSUB(T4R, T4T); T3v = LD(&(ri[WS(rs, 7)]), ms, &(ri[WS(rs, 1)])); T3x = LD(&(ii[WS(rs, 7)]), ms, &(ii[WS(rs, 1)])); T3p = VMUL(T3n, T3o); T5A = VMUL(T3n, T3s); T4O = VSUB(T4M, T4N); T6A = VADD(T4M, T4N); T2k = VADD(T22, T2j); T7o = VSUB(T22, T2j); T6B = VSUB(T4P, T4U); T4V = VADD(T4P, T4U); T7r = VSUB(T7p, T7q); T8e = VADD(T7p, T7q); T3w = VMUL(TP, T3v); T3t = VFMA(T3r, T3s, T3p); T5B = VFNMS(T3r, T3o, T5A); T5W = VMUL(TP, T3x); } } } { V T2t, T2q, T50, T2L, T5q, T2u, T2x, T2A, T2C; { V T2n, T2p, T2G, T2K, T5V, T3u, T5C, T7G, T5X, T2o, T4Z, T2H, T5D, T3F, T5p; V T3y, T60, T7H; T2n = LD(&(ri[WS(rs, 1)]), ms, &(ri[WS(rs, 1)])); T2p = LD(&(ii[WS(rs, 1)]), ms, &(ii[WS(rs, 1)])); T2G = LD(&(ri[WS(rs, 25)]), ms, &(ri[WS(rs, 1)])); T2K = LD(&(ii[WS(rs, 25)]), ms, &(ii[WS(rs, 1)])); T3y = VFMA(TR, T3x, T3w); T5V = VSUB(T3l, T3t); T3u = VADD(T3l, T3t); T5C = VSUB(T5z, T5B); T7G = VADD(T5z, T5B); T5X = VFNMS(TR, T3v, T5W); T2o = VMUL(T2, T2n); T4Z = VMUL(T2, T2p); T2H = VMUL(T2F, T2G); T5D = VSUB(T3y, T3E); T3F = VADD(T3y, T3E); T5p = VMUL(T2F, T2K); T2t = LD(&(ri[WS(rs, 17)]), ms, &(ri[WS(rs, 1)])); T60 = VSUB(T5X, T5Z); T7H = VADD(T5X, T5Z); T2q = VFMA(T5, T2p, T2o); T50 = VFNMS(T5, T2n, T4Z); T2L = VFMA(T2J, T2K, T2H); T5E = VSUB(T5C, T5D); T6P = VADD(T5C, T5D); T3G = VADD(T3u, T3F); T7L = VSUB(T3u, T3F); T5q = VFNMS(T2J, T2G, T5p); T6M = VSUB(T5V, T60); T61 = VADD(T5V, T60); T8n = VADD(T7G, T7H); T7I = VSUB(T7G, T7H); T2u = VMUL(T2s, T2t); T2x = LD(&(ii[WS(rs, 17)]), ms, &(ii[WS(rs, 1)])); T2A = LD(&(ri[WS(rs, 9)]), ms, &(ri[WS(rs, 1)])); T2C = LD(&(ii[WS(rs, 9)]), ms, &(ii[WS(rs, 1)])); } { V T3N, T2z, T5m, T3K, T5G, T41, T5Q, T3O, T7v, T53, T2M, T54, T7w, T5r, T3R; V T3U, T3W; { V T3H, T3J, T3Y, T40, T52, T2D, T5o; T3H = LD(&(ri[WS(rs, 3)]), ms, &(ri[WS(rs, 1)])); T3J = LD(&(ii[WS(rs, 3)]), ms, &(ii[WS(rs, 1)])); T3Y = LD(&(ri[WS(rs, 11)]), ms, &(ri[WS(rs, 1)])); T40 = LD(&(ii[WS(rs, 11)]), ms, &(ii[WS(rs, 1)])); T3N = LD(&(ri[WS(rs, 19)]), ms, &(ri[WS(rs, 1)])); { V T2y, T51, T2B, T5n; T2y = VFMA(T2w, T2x, T2u); T51 = VMUL(T2s, T2x); T2B = VMUL(T8, T2A); T5n = VMUL(T8, T2C); { V T3I, T5F, T3Z, T5P; T3I = VMUL(T3, T3H); T5F = VMUL(T3, T3J); T3Z = VMUL(Td, T3Y); T5P = VMUL(Td, T40); T2z = VADD(T2q, T2y); T5m = VSUB(T2q, T2y); T52 = VFNMS(T2w, T2t, T51); T2D = VFMA(Tc, T2C, T2B); T5o = VFNMS(Tc, T2A, T5n); T3K = VFMA(T6, T3J, T3I); T5G = VFNMS(T6, T3H, T5F); T41 = VFMA(Th, T40, T3Z); T5Q = VFNMS(Th, T3Y, T5P); T3O = VMUL(T3M, T3N); } } T7v = VADD(T50, T52); T53 = VSUB(T50, T52); T2M = VADD(T2D, T2L); T54 = VSUB(T2D, T2L); T7w = VADD(T5o, T5q); T5r = VSUB(T5o, T5q); T3R = LD(&(ii[WS(rs, 19)]), ms, &(ii[WS(rs, 1)])); T3U = LD(&(ri[WS(rs, 27)]), ms, &(ri[WS(rs, 1)])); T3W = LD(&(ii[WS(rs, 27)]), ms, &(ii[WS(rs, 1)])); } { V T2O, T37, T39, T3T, T5K, T5I, T3X, T5O, T56, T38, T5g, T7M, T5J; { V T3S, T5H, T3V, T5N, T2P, T2Q; T2O = LD(&(ri[WS(rs, 5)]), ms, &(ri[WS(rs, 1)])); T55 = VSUB(T53, T54); T6I = VADD(T53, T54); T2N = VADD(T2z, T2M); T7A = VSUB(T2z, T2M); T5s = VADD(T5m, T5r); T6F = VSUB(T5m, T5r); T7x = VSUB(T7v, T7w); T8i = VADD(T7v, T7w); T3S = VFMA(T3Q, T3R, T3O); T5H = VMUL(T3M, T3R); T3V = VMUL(Te, T3U); T5N = VMUL(Te, T3W); T2P = VMUL(T29, T2O); T2Q = LD(&(ii[WS(rs, 5)]), ms, &(ii[WS(rs, 1)])); T37 = LD(&(ri[WS(rs, 13)]), ms, &(ri[WS(rs, 1)])); T39 = LD(&(ii[WS(rs, 13)]), ms, &(ii[WS(rs, 1)])); T3T = VADD(T3K, T3S); T5K = VSUB(T3K, T3S); T5I = VFNMS(T3Q, T3N, T5H); T3X = VFMA(Ti, T3W, T3V); T5O = VFNMS(Ti, T3U, T5N); T2R = VFMA(T2b, T2Q, T2P); T56 = VMUL(T29, T2Q); T38 = VMUL(T1R, T37); T5g = VMUL(T1R, T39); } T2U = LD(&(ri[WS(rs, 21)]), ms, &(ri[WS(rs, 1)])); T7M = VADD(T5G, T5I); T5J = VSUB(T5G, T5I); { V T42, T5M, T7N, T5R; T42 = VADD(T3X, T41); T5M = VSUB(T3X, T41); T7N = VADD(T5O, T5Q); T5R = VSUB(T5O, T5Q); T57 = VFNMS(T2b, T2O, T56); T3a = VFMA(T1U, T39, T38); T5h = VFNMS(T1U, T37, T5g); T62 = VADD(T5K, T5J); T5L = VSUB(T5J, T5K); T7J = VSUB(T42, T3T); T43 = VADD(T3T, T42); T63 = VSUB(T5M, T5R); T5S = VADD(T5M, T5R); T8o = VADD(T7M, T7N); T7O = VSUB(T7M, T7N); T2V = VMUL(T2T, T2U); } T2Y = LD(&(ii[WS(rs, 21)]), ms, &(ii[WS(rs, 1)])); T32 = LD(&(ri[WS(rs, 29)]), ms, &(ri[WS(rs, 1)])); T35 = LD(&(ii[WS(rs, 29)]), ms, &(ii[WS(rs, 1)])); } } } } } { V T5t, T5c, T5u, T5j, T8Z, T90; { V T7e, T8T, T8y, T7h, T8U, T8c, T8J, T44, T8u, T8q, T7y, T7D, T8w, T2m, T3d; V T8h, T8R, T8P, T8k, T8x, T8B, T8f; { V T1i, T8O, T8N, T2l, T3c, T8j; { V T8p, T5b, T30, T59, T36, T5f, TI, T1h, T8m, T5a, T7B; TI = VADD(Tq, TH); T7e = VSUB(Tq, TH); T8T = VSUB(T1g, TZ); T1h = VADD(TZ, T1g); T8y = VADD(T8n, T8o); T8p = VSUB(T8n, T8o); { V T8C, T8I, T2Z, T58, T33, T5e; T7h = VSUB(T7f, T7g); T8C = VADD(T7f, T7g); T8I = VADD(T8D, T8H); T8U = VSUB(T8H, T8D); T2Z = VFMA(T2X, T2Y, T2V); T58 = VMUL(T2T, T2Y); T33 = VMUL(T31, T32); T5e = VMUL(T31, T35); T1i = VADD(TI, T1h); T8c = VSUB(TI, T1h); T8O = VSUB(T8I, T8C); T8J = VADD(T8C, T8I); T5b = VSUB(T2R, T2Z); T30 = VADD(T2R, T2Z); T59 = VFNMS(T2X, T2U, T58); T36 = VFMA(T34, T35, T33); T5f = VFNMS(T34, T32, T5e); } T44 = VADD(T3G, T43); T8m = VSUB(T3G, T43); T5a = VSUB(T57, T59); T7B = VADD(T57, T59); { V T5d, T3b, T5i, T7C; T5d = VSUB(T36, T3a); T3b = VADD(T36, T3a); T5i = VSUB(T5f, T5h); T7C = VADD(T5f, T5h); T8N = VSUB(T2k, T1J); T2l = VADD(T1J, T2k); T8u = VADD(T8m, T8p); T8q = VSUB(T8m, T8p); T5t = VADD(T5b, T5a); T5c = VSUB(T5a, T5b); T7y = VSUB(T3b, T30); T3c = VADD(T30, T3b); T5u = VSUB(T5d, T5i); T5j = VADD(T5d, T5i); T8j = VADD(T7B, T7C); T7D = VSUB(T7B, T7C); } } T8w = VSUB(T1i, T2l); T2m = VADD(T1i, T2l); T3d = VADD(T2N, T3c); T8h = VSUB(T2N, T3c); T8R = VSUB(T8O, T8N); T8P = VADD(T8N, T8O); T8k = VSUB(T8i, T8j); T8x = VADD(T8i, T8j); T8B = VADD(T8d, T8e); T8f = VSUB(T8d, T8e); } { V T7P, T7K, T7X, T7Y, T82, T7z, T7W, T7i, T8a, T86, T91, T8V, T8W, T7t, T7E; V T81; { V T84, T85, T7n, T7s, T8L, T45; T8L = VSUB(T44, T3d); T45 = VADD(T3d, T44); { V T8t, T8l, T8A, T8z; T8t = VSUB(T8k, T8h); T8l = VADD(T8h, T8k); T8A = VADD(T8x, T8y); T8z = VSUB(T8x, T8y); { V T8M, T8K, T8s, T8g; T8M = VSUB(T8J, T8B); T8K = VADD(T8B, T8J); T8s = VSUB(T8c, T8f); T8g = VADD(T8c, T8f); ST(&(ri[0]), VADD(T2m, T45), ms, &(ri[0])); ST(&(ri[WS(rs, 16)]), VSUB(T2m, T45), ms, &(ri[0])); { V T8v, T8Q, T8S, T8r; T8v = VSUB(T8t, T8u); T8Q = VADD(T8t, T8u); T8S = VSUB(T8q, T8l); T8r = VADD(T8l, T8q); ST(&(ri[WS(rs, 8)]), VADD(T8w, T8z), ms, &(ri[0])); ST(&(ri[WS(rs, 24)]), VSUB(T8w, T8z), ms, &(ri[0])); ST(&(ii[WS(rs, 24)]), VSUB(T8M, T8L), ms, &(ii[0])); ST(&(ii[WS(rs, 8)]), VADD(T8L, T8M), ms, &(ii[0])); ST(&(ii[WS(rs, 16)]), VSUB(T8K, T8A), ms, &(ii[0])); ST(&(ii[0]), VADD(T8A, T8K), ms, &(ii[0])); ST(&(ri[WS(rs, 12)]), VFMA(LDK(KP707106781), T8v, T8s), ms, &(ri[0])); ST(&(ri[WS(rs, 28)]), VFNMS(LDK(KP707106781), T8v, T8s), ms, &(ri[0])); ST(&(ii[WS(rs, 20)]), VFNMS(LDK(KP707106781), T8Q, T8P), ms, &(ii[0])); ST(&(ii[WS(rs, 4)]), VFMA(LDK(KP707106781), T8Q, T8P), ms, &(ii[0])); ST(&(ii[WS(rs, 28)]), VFNMS(LDK(KP707106781), T8S, T8R), ms, &(ii[0])); ST(&(ii[WS(rs, 12)]), VFMA(LDK(KP707106781), T8S, T8R), ms, &(ii[0])); ST(&(ri[WS(rs, 4)]), VFMA(LDK(KP707106781), T8r, T8g), ms, &(ri[0])); ST(&(ri[WS(rs, 20)]), VFNMS(LDK(KP707106781), T8r, T8g), ms, &(ri[0])); } } } T7P = VSUB(T7L, T7O); T84 = VADD(T7L, T7O); T85 = VADD(T7I, T7J); T7K = VSUB(T7I, T7J); T7X = VADD(T7m, T7l); T7n = VSUB(T7l, T7m); T7s = VADD(T7o, T7r); T7Y = VSUB(T7o, T7r); T82 = VADD(T7x, T7y); T7z = VSUB(T7x, T7y); T7W = VADD(T7e, T7h); T7i = VSUB(T7e, T7h); T8a = VFMA(LDK(KP414213562), T84, T85); T86 = VFNMS(LDK(KP414213562), T85, T84); T91 = VSUB(T8U, T8T); T8V = VADD(T8T, T8U); T8W = VADD(T7n, T7s); T7t = VSUB(T7n, T7s); T7E = VSUB(T7A, T7D); T81 = VADD(T7A, T7D); } { V T7S, T7u, T7T, T7F, T92, T7Z, T89, T83, T7U, T7Q; T7S = VFNMS(LDK(KP707106781), T7t, T7i); T7u = VFMA(LDK(KP707106781), T7t, T7i); T7T = VFNMS(LDK(KP414213562), T7z, T7E); T7F = VFMA(LDK(KP414213562), T7E, T7z); T92 = VSUB(T7Y, T7X); T7Z = VADD(T7X, T7Y); T89 = VFNMS(LDK(KP414213562), T81, T82); T83 = VFMA(LDK(KP414213562), T82, T81); T7U = VFMA(LDK(KP414213562), T7K, T7P); T7Q = VFNMS(LDK(KP414213562), T7P, T7K); { V T8X, T95, T93, T80, T88, T87, T7V, T94, T96, T7R, T8Y, T8b; T8Z = VFNMS(LDK(KP707106781), T8W, T8V); T8X = VFMA(LDK(KP707106781), T8W, T8V); T95 = VFNMS(LDK(KP707106781), T92, T91); T93 = VFMA(LDK(KP707106781), T92, T91); T80 = VFMA(LDK(KP707106781), T7Z, T7W); T88 = VFNMS(LDK(KP707106781), T7Z, T7W); T90 = VSUB(T86, T83); T87 = VADD(T83, T86); T7V = VADD(T7T, T7U); T94 = VSUB(T7U, T7T); T96 = VADD(T7F, T7Q); T7R = VSUB(T7F, T7Q); T8Y = VADD(T89, T8a); T8b = VSUB(T89, T8a); ST(&(ri[WS(rs, 2)]), VFMA(LDK(KP923879532), T87, T80), ms, &(ri[0])); ST(&(ri[WS(rs, 18)]), VFNMS(LDK(KP923879532), T87, T80), ms, &(ri[0])); ST(&(ri[WS(rs, 30)]), VFMA(LDK(KP923879532), T7V, T7S), ms, &(ri[0])); ST(&(ri[WS(rs, 14)]), VFNMS(LDK(KP923879532), T7V, T7S), ms, &(ri[0])); ST(&(ii[WS(rs, 22)]), VFNMS(LDK(KP923879532), T94, T93), ms, &(ii[0])); ST(&(ii[WS(rs, 6)]), VFMA(LDK(KP923879532), T94, T93), ms, &(ii[0])); ST(&(ii[WS(rs, 30)]), VFMA(LDK(KP923879532), T96, T95), ms, &(ii[0])); ST(&(ii[WS(rs, 14)]), VFNMS(LDK(KP923879532), T96, T95), ms, &(ii[0])); ST(&(ri[WS(rs, 6)]), VFMA(LDK(KP923879532), T7R, T7u), ms, &(ri[0])); ST(&(ri[WS(rs, 22)]), VFNMS(LDK(KP923879532), T7R, T7u), ms, &(ri[0])); ST(&(ii[WS(rs, 18)]), VFNMS(LDK(KP923879532), T8Y, T8X), ms, &(ii[0])); ST(&(ii[WS(rs, 2)]), VFMA(LDK(KP923879532), T8Y, T8X), ms, &(ii[0])); ST(&(ri[WS(rs, 26)]), VFNMS(LDK(KP923879532), T8b, T88), ms, &(ri[0])); ST(&(ri[WS(rs, 10)]), VFMA(LDK(KP923879532), T8b, T88), ms, &(ri[0])); } } } } { V T6s, T9o, T9n, T6v, T6N, T6Q, T6G, T6J, T68, T4Y, T9f, T9d, T9l, T9j, T6g; V T6o, T6q, T6m, T66, T6a, T6p, T6j, T5x, T69; { V T6d, T6e, T6c, T4s, T9c, T4X, T9h, T9b, T5T, T64, T5k, T5v, T9i, T6f; { V T4c, T4r, T4H, T4W, T99, T9a; T6s = VSUB(T46, T4b); T4c = VADD(T46, T4b); T4r = VADD(T4j, T4q); T9o = VSUB(T4q, T4j); T6d = VFMA(LDK(KP414213562), T4z, T4G); T4H = VFNMS(LDK(KP414213562), T4G, T4z); T4W = VFMA(LDK(KP414213562), T4V, T4O); T6e = VFNMS(LDK(KP414213562), T4O, T4V); T9n = VADD(T98, T97); T99 = VSUB(T97, T98); T9a = VADD(T6t, T6u); T6v = VSUB(T6t, T6u); ST(&(ii[WS(rs, 26)]), VFNMS(LDK(KP923879532), T90, T8Z), ms, &(ii[0])); ST(&(ii[WS(rs, 10)]), VFMA(LDK(KP923879532), T90, T8Z), ms, &(ii[0])); T6c = VFMA(LDK(KP707106781), T4r, T4c); T4s = VFNMS(LDK(KP707106781), T4r, T4c); T9c = VADD(T4H, T4W); T4X = VSUB(T4H, T4W); T9h = VFNMS(LDK(KP707106781), T9a, T99); T9b = VFMA(LDK(KP707106781), T9a, T99); T6N = VSUB(T5S, T5L); T5T = VADD(T5L, T5S); T64 = VADD(T62, T63); T6Q = VSUB(T62, T63); T6G = VSUB(T5j, T5c); T5k = VADD(T5c, T5j); T5v = VADD(T5t, T5u); T6J = VSUB(T5t, T5u); } T68 = VFNMS(LDK(KP923879532), T4X, T4s); T4Y = VFMA(LDK(KP923879532), T4X, T4s); T9f = VFNMS(LDK(KP923879532), T9c, T9b); T9d = VFMA(LDK(KP923879532), T9c, T9b); T9i = VSUB(T6e, T6d); T6f = VADD(T6d, T6e); { V T6l, T5U, T6k, T65; T6l = VFMA(LDK(KP707106781), T5T, T5E); T5U = VFNMS(LDK(KP707106781), T5T, T5E); T6k = VFMA(LDK(KP707106781), T64, T61); T65 = VFNMS(LDK(KP707106781), T64, T61); { V T6i, T5l, T6h, T5w; T6i = VFMA(LDK(KP707106781), T5k, T55); T5l = VFNMS(LDK(KP707106781), T5k, T55); T6h = VFMA(LDK(KP707106781), T5v, T5s); T5w = VFNMS(LDK(KP707106781), T5v, T5s); T9l = VFNMS(LDK(KP923879532), T9i, T9h); T9j = VFMA(LDK(KP923879532), T9i, T9h); T6g = VFMA(LDK(KP923879532), T6f, T6c); T6o = VFNMS(LDK(KP923879532), T6f, T6c); T6q = VFMA(LDK(KP198912367), T6k, T6l); T6m = VFNMS(LDK(KP198912367), T6l, T6k); T66 = VFNMS(LDK(KP668178637), T65, T5U); T6a = VFMA(LDK(KP668178637), T5U, T65); T6p = VFNMS(LDK(KP198912367), T6h, T6i); T6j = VFMA(LDK(KP198912367), T6i, T6h); T5x = VFMA(LDK(KP668178637), T5w, T5l); T69 = VFNMS(LDK(KP668178637), T5l, T5w); } } } { V T6Y, T6w, T9w, T6D, T9v, T9p, T9q, T71, T77, T6O, T76, T6R; { V T6Z, T6z, T6C, T70; { V T6n, T9g, T9e, T6r; T6n = VADD(T6j, T6m); T9g = VSUB(T6m, T6j); T9e = VADD(T6p, T6q); T6r = VSUB(T6p, T6q); { V T9k, T6b, T67, T9m; T9k = VSUB(T6a, T69); T6b = VADD(T69, T6a); T67 = VSUB(T5x, T66); T9m = VADD(T5x, T66); ST(&(ii[WS(rs, 25)]), VFNMS(LDK(KP980785280), T9g, T9f), ms, &(ii[WS(rs, 1)])); ST(&(ii[WS(rs, 9)]), VFMA(LDK(KP980785280), T9g, T9f), ms, &(ii[WS(rs, 1)])); ST(&(ri[WS(rs, 1)]), VFMA(LDK(KP980785280), T6n, T6g), ms, &(ri[WS(rs, 1)])); ST(&(ri[WS(rs, 17)]), VFNMS(LDK(KP980785280), T6n, T6g), ms, &(ri[WS(rs, 1)])); ST(&(ri[WS(rs, 9)]), VFMA(LDK(KP980785280), T6r, T6o), ms, &(ri[WS(rs, 1)])); ST(&(ri[WS(rs, 25)]), VFNMS(LDK(KP980785280), T6r, T6o), ms, &(ri[WS(rs, 1)])); ST(&(ii[WS(rs, 17)]), VFNMS(LDK(KP980785280), T9e, T9d), ms, &(ii[WS(rs, 1)])); ST(&(ii[WS(rs, 1)]), VFMA(LDK(KP980785280), T9e, T9d), ms, &(ii[WS(rs, 1)])); ST(&(ri[WS(rs, 29)]), VFMA(LDK(KP831469612), T6b, T68), ms, &(ri[WS(rs, 1)])); ST(&(ri[WS(rs, 13)]), VFNMS(LDK(KP831469612), T6b, T68), ms, &(ri[WS(rs, 1)])); ST(&(ii[WS(rs, 21)]), VFNMS(LDK(KP831469612), T9k, T9j), ms, &(ii[WS(rs, 1)])); ST(&(ii[WS(rs, 5)]), VFMA(LDK(KP831469612), T9k, T9j), ms, &(ii[WS(rs, 1)])); ST(&(ii[WS(rs, 29)]), VFMA(LDK(KP831469612), T9m, T9l), ms, &(ii[WS(rs, 1)])); ST(&(ii[WS(rs, 13)]), VFNMS(LDK(KP831469612), T9m, T9l), ms, &(ii[WS(rs, 1)])); ST(&(ri[WS(rs, 5)]), VFMA(LDK(KP831469612), T67, T4Y), ms, &(ri[WS(rs, 1)])); ST(&(ri[WS(rs, 21)]), VFNMS(LDK(KP831469612), T67, T4Y), ms, &(ri[WS(rs, 1)])); T6Y = VFNMS(LDK(KP707106781), T6v, T6s); T6w = VFMA(LDK(KP707106781), T6v, T6s); } } T6Z = VFNMS(LDK(KP414213562), T6x, T6y); T6z = VFMA(LDK(KP414213562), T6y, T6x); T6C = VFNMS(LDK(KP414213562), T6B, T6A); T70 = VFMA(LDK(KP414213562), T6A, T6B); T9w = VADD(T6z, T6C); T6D = VSUB(T6z, T6C); T9v = VFNMS(LDK(KP707106781), T9o, T9n); T9p = VFMA(LDK(KP707106781), T9o, T9n); T9q = VSUB(T70, T6Z); T71 = VADD(T6Z, T70); T77 = VFMA(LDK(KP707106781), T6N, T6M); T6O = VFNMS(LDK(KP707106781), T6N, T6M); T76 = VFMA(LDK(KP707106781), T6Q, T6P); T6R = VFNMS(LDK(KP707106781), T6Q, T6P); T6H = VFNMS(LDK(KP707106781), T6G, T6F); T74 = VFMA(LDK(KP707106781), T6G, T6F); } T6U = VFNMS(LDK(KP923879532), T6D, T6w); T6E = VFMA(LDK(KP923879532), T6D, T6w); T9r = VFMA(LDK(KP923879532), T9q, T9p); T9t = VFNMS(LDK(KP923879532), T9q, T9p); T78 = VFNMS(LDK(KP198912367), T77, T76); T7c = VFMA(LDK(KP198912367), T76, T77); T6W = VFMA(LDK(KP668178637), T6O, T6R); T6S = VFNMS(LDK(KP668178637), T6R, T6O); T73 = VFMA(LDK(KP707106781), T6J, T6I); T6K = VFNMS(LDK(KP707106781), T6J, T6I); T7a = VFMA(LDK(KP923879532), T71, T6Y); T72 = VFNMS(LDK(KP923879532), T71, T6Y); T9x = VFNMS(LDK(KP923879532), T9w, T9v); T9z = VFMA(LDK(KP923879532), T9w, T9v); } } } } } { V T7b, T75, T6L, T6V; T7b = VFNMS(LDK(KP198912367), T73, T74); T75 = VFMA(LDK(KP198912367), T74, T73); T6L = VFMA(LDK(KP668178637), T6K, T6H); T6V = VFNMS(LDK(KP668178637), T6H, T6K); { V T79, T9A, T9y, T7d; T79 = VSUB(T75, T78); T9A = VADD(T75, T78); T9y = VSUB(T7c, T7b); T7d = VADD(T7b, T7c); { V T9s, T6X, T6T, T9u; T9s = VADD(T6V, T6W); T6X = VSUB(T6V, T6W); T6T = VADD(T6L, T6S); T9u = VSUB(T6S, T6L); ST(&(ii[WS(rs, 31)]), VFMA(LDK(KP980785280), T9A, T9z), ms, &(ii[WS(rs, 1)])); ST(&(ii[WS(rs, 15)]), VFNMS(LDK(KP980785280), T9A, T9z), ms, &(ii[WS(rs, 1)])); ST(&(ri[WS(rs, 7)]), VFMA(LDK(KP980785280), T79, T72), ms, &(ri[WS(rs, 1)])); ST(&(ri[WS(rs, 23)]), VFNMS(LDK(KP980785280), T79, T72), ms, &(ri[WS(rs, 1)])); ST(&(ri[WS(rs, 31)]), VFMA(LDK(KP980785280), T7d, T7a), ms, &(ri[WS(rs, 1)])); ST(&(ri[WS(rs, 15)]), VFNMS(LDK(KP980785280), T7d, T7a), ms, &(ri[WS(rs, 1)])); ST(&(ii[WS(rs, 23)]), VFNMS(LDK(KP980785280), T9y, T9x), ms, &(ii[WS(rs, 1)])); ST(&(ii[WS(rs, 7)]), VFMA(LDK(KP980785280), T9y, T9x), ms, &(ii[WS(rs, 1)])); ST(&(ri[WS(rs, 11)]), VFMA(LDK(KP831469612), T6X, T6U), ms, &(ri[WS(rs, 1)])); ST(&(ri[WS(rs, 27)]), VFNMS(LDK(KP831469612), T6X, T6U), ms, &(ri[WS(rs, 1)])); ST(&(ii[WS(rs, 19)]), VFNMS(LDK(KP831469612), T9s, T9r), ms, &(ii[WS(rs, 1)])); ST(&(ii[WS(rs, 3)]), VFMA(LDK(KP831469612), T9s, T9r), ms, &(ii[WS(rs, 1)])); ST(&(ii[WS(rs, 27)]), VFNMS(LDK(KP831469612), T9u, T9t), ms, &(ii[WS(rs, 1)])); ST(&(ii[WS(rs, 11)]), VFMA(LDK(KP831469612), T9u, T9t), ms, &(ii[WS(rs, 1)])); ST(&(ri[WS(rs, 3)]), VFMA(LDK(KP831469612), T6T, T6E), ms, &(ri[WS(rs, 1)])); ST(&(ri[WS(rs, 19)]), VFNMS(LDK(KP831469612), T6T, T6E), ms, &(ri[WS(rs, 1)])); } } } } } VLEAVE(); } static const tw_instr twinstr[] = { VTW(0, 1), VTW(0, 3), VTW(0, 9), VTW(0, 27), {TW_NEXT, (2 * VL), 0} }; static const ct_desc desc = { 32, XSIMD_STRING("t2sv_32"), twinstr, &GENUS, {236, 98, 252, 0}, 0, 0, 0 }; void XSIMD(codelet_t2sv_32) (planner *p) { X(kdft_dit_register) (p, t2sv_32, &desc); } #else /* HAVE_FMA */ /* Generated by: ../../../genfft/gen_twiddle.native -simd -compact -variables 4 -pipeline-latency 8 -twiddle-log3 -precompute-twiddles -n 32 -name t2sv_32 -include ts.h */ /* * This function contains 488 FP additions, 280 FP multiplications, * (or, 376 additions, 168 multiplications, 112 fused multiply/add), * 158 stack variables, 7 constants, and 128 memory accesses */ #include "ts.h" static void t2sv_32(R *ri, R *ii, const R *W, stride rs, INT mb, INT me, INT ms) { DVK(KP195090322, +0.195090322016128267848284868477022240927691618); DVK(KP980785280, +0.980785280403230449126182236134239036973933731); DVK(KP555570233, +0.555570233019602224742830813948532874374937191); DVK(KP831469612, +0.831469612302545237078788377617905756738560812); DVK(KP382683432, +0.382683432365089771728459984030398866761344562); DVK(KP923879532, +0.923879532511286756128183189396788286822416626); DVK(KP707106781, +0.707106781186547524400844362104849039284835938); { INT m; for (m = mb, W = W + (mb * 8); m < me; m = m + (2 * VL), ri = ri + ((2 * VL) * ms), ii = ii + ((2 * VL) * ms), W = W + ((2 * VL) * 8), MAKE_VOLATILE_STRIDE(64, rs)) { V T2, T5, T3, T6, T8, TM, TO, Td, T9, Te, Th, Tl, TD, TH, T1y; V T1H, T15, T1A, T11, T1F, T1n, T1p, T2q, T2I, T2u, T2K, T2V, T3b, T2Z, T3d; V Tu, Ty, T3l, T3n, T1t, T1v, T2f, T2h, T1a, T1e, T32, T34, T1W, T1Y, T2C; V T2E, Tg, TR, Tk, TS, Tm, TV, To, TT, T1M, T21, T1P, T22, T1Q, T25; V T1S, T23; { V Ts, T1d, Tx, T18, Tt, T1c, Tw, T19, TB, T14, TG, TZ, TC, T13, TF; V T10; { V T4, Tc, T7, Tb; T2 = LDW(&(W[0])); T5 = LDW(&(W[TWVL * 1])); T3 = LDW(&(W[TWVL * 2])); T6 = LDW(&(W[TWVL * 3])); T4 = VMUL(T2, T3); Tc = VMUL(T5, T3); T7 = VMUL(T5, T6); Tb = VMUL(T2, T6); T8 = VADD(T4, T7); TM = VSUB(T4, T7); TO = VADD(Tb, Tc); Td = VSUB(Tb, Tc); T9 = LDW(&(W[TWVL * 4])); Ts = VMUL(T2, T9); T1d = VMUL(T6, T9); Tx = VMUL(T5, T9); T18 = VMUL(T3, T9); Te = LDW(&(W[TWVL * 5])); Tt = VMUL(T5, Te); T1c = VMUL(T3, Te); Tw = VMUL(T2, Te); T19 = VMUL(T6, Te); Th = LDW(&(W[TWVL * 6])); TB = VMUL(T3, Th); T14 = VMUL(T5, Th); TG = VMUL(T6, Th); TZ = VMUL(T2, Th); Tl = LDW(&(W[TWVL * 7])); TC = VMUL(T6, Tl); T13 = VMUL(T2, Tl); TF = VMUL(T3, Tl); T10 = VMUL(T5, Tl); } TD = VADD(TB, TC); TH = VSUB(TF, TG); T1y = VADD(TZ, T10); T1H = VADD(TF, TG); T15 = VADD(T13, T14); T1A = VSUB(T13, T14); T11 = VSUB(TZ, T10); T1F = VSUB(TB, TC); T1n = VFMA(T9, Th, VMUL(Te, Tl)); T1p = VFNMS(Te, Th, VMUL(T9, Tl)); { V T2o, T2p, T2s, T2t; T2o = VMUL(T8, Th); T2p = VMUL(Td, Tl); T2q = VADD(T2o, T2p); T2I = VSUB(T2o, T2p); T2s = VMUL(T8, Tl); T2t = VMUL(Td, Th); T2u = VSUB(T2s, T2t); T2K = VADD(T2s, T2t); } { V T2T, T2U, T2X, T2Y; T2T = VMUL(TM, Th); T2U = VMUL(TO, Tl); T2V = VSUB(T2T, T2U); T3b = VADD(T2T, T2U); T2X = VMUL(TM, Tl); T2Y = VMUL(TO, Th); T2Z = VADD(T2X, T2Y); T3d = VSUB(T2X, T2Y); Tu = VADD(Ts, Tt); Ty = VSUB(Tw, Tx); T3l = VFMA(Tu, Th, VMUL(Ty, Tl)); T3n = VFNMS(Ty, Th, VMUL(Tu, Tl)); } T1t = VSUB(Ts, Tt); T1v = VADD(Tw, Tx); T2f = VFMA(T1t, Th, VMUL(T1v, Tl)); T2h = VFNMS(T1v, Th, VMUL(T1t, Tl)); T1a = VSUB(T18, T19); T1e = VADD(T1c, T1d); T32 = VFMA(T1a, Th, VMUL(T1e, Tl)); T34 = VFNMS(T1e, Th, VMUL(T1a, Tl)); T1W = VADD(T18, T19); T1Y = VSUB(T1c, T1d); T2C = VFMA(T1W, Th, VMUL(T1Y, Tl)); T2E = VFNMS(T1Y, Th, VMUL(T1W, Tl)); { V Ta, Tf, Ti, Tj; Ta = VMUL(T8, T9); Tf = VMUL(Td, Te); Tg = VSUB(Ta, Tf); TR = VADD(Ta, Tf); Ti = VMUL(T8, Te); Tj = VMUL(Td, T9); Tk = VADD(Ti, Tj); TS = VSUB(Ti, Tj); } Tm = VFMA(Tg, Th, VMUL(Tk, Tl)); TV = VFNMS(TS, Th, VMUL(TR, Tl)); To = VFNMS(Tk, Th, VMUL(Tg, Tl)); TT = VFMA(TR, Th, VMUL(TS, Tl)); { V T1K, T1L, T1N, T1O; T1K = VMUL(TM, T9); T1L = VMUL(TO, Te); T1M = VSUB(T1K, T1L); T21 = VADD(T1K, T1L); T1N = VMUL(TM, Te); T1O = VMUL(TO, T9); T1P = VADD(T1N, T1O); T22 = VSUB(T1N, T1O); } T1Q = VFMA(T1M, Th, VMUL(T1P, Tl)); T25 = VFNMS(T22, Th, VMUL(T21, Tl)); T1S = VFNMS(T1P, Th, VMUL(T1M, Tl)); T23 = VFMA(T21, Th, VMUL(T22, Tl)); } { V TL, T6f, T8c, T8q, T3F, T5t, T7I, T7W, T2y, T6B, T6y, T7j, T4k, T5J, T4B; V T5G, T3h, T6H, T6O, T7o, T4L, T5N, T52, T5Q, T1i, T7V, T6i, T7D, T3K, T5u; V T3P, T5v, T1E, T6n, T6m, T7e, T3W, T5y, T41, T5z, T29, T6p, T6s, T7f, T47; V T5B, T4c, T5C, T2R, T6z, T6E, T7k, T4v, T5H, T4E, T5K, T3y, T6P, T6K, T7p; V T4W, T5R, T55, T5O; { V T1, T7G, Tq, T7F, TA, T3C, TJ, T3D, Tn, Tp; T1 = LD(&(ri[0]), ms, &(ri[0])); T7G = LD(&(ii[0]), ms, &(ii[0])); Tn = LD(&(ri[WS(rs, 16)]), ms, &(ri[0])); Tp = LD(&(ii[WS(rs, 16)]), ms, &(ii[0])); Tq = VFMA(Tm, Tn, VMUL(To, Tp)); T7F = VFNMS(To, Tn, VMUL(Tm, Tp)); { V Tv, Tz, TE, TI; Tv = LD(&(ri[WS(rs, 8)]), ms, &(ri[0])); Tz = LD(&(ii[WS(rs, 8)]), ms, &(ii[0])); TA = VFMA(Tu, Tv, VMUL(Ty, Tz)); T3C = VFNMS(Ty, Tv, VMUL(Tu, Tz)); TE = LD(&(ri[WS(rs, 24)]), ms, &(ri[0])); TI = LD(&(ii[WS(rs, 24)]), ms, &(ii[0])); TJ = VFMA(TD, TE, VMUL(TH, TI)); T3D = VFNMS(TH, TE, VMUL(TD, TI)); } { V Tr, TK, T8a, T8b; Tr = VADD(T1, Tq); TK = VADD(TA, TJ); TL = VADD(Tr, TK); T6f = VSUB(Tr, TK); T8a = VSUB(T7G, T7F); T8b = VSUB(TA, TJ); T8c = VSUB(T8a, T8b); T8q = VADD(T8b, T8a); } { V T3B, T3E, T7E, T7H; T3B = VSUB(T1, Tq); T3E = VSUB(T3C, T3D); T3F = VSUB(T3B, T3E); T5t = VADD(T3B, T3E); T7E = VADD(T3C, T3D); T7H = VADD(T7F, T7G); T7I = VADD(T7E, T7H); T7W = VSUB(T7H, T7E); } } { V T2e, T4g, T2w, T4z, T2j, T4h, T2n, T4y; { V T2c, T2d, T2r, T2v; T2c = LD(&(ri[WS(rs, 1)]), ms, &(ri[WS(rs, 1)])); T2d = LD(&(ii[WS(rs, 1)]), ms, &(ii[WS(rs, 1)])); T2e = VFMA(T2, T2c, VMUL(T5, T2d)); T4g = VFNMS(T5, T2c, VMUL(T2, T2d)); T2r = LD(&(ri[WS(rs, 25)]), ms, &(ri[WS(rs, 1)])); T2v = LD(&(ii[WS(rs, 25)]), ms, &(ii[WS(rs, 1)])); T2w = VFMA(T2q, T2r, VMUL(T2u, T2v)); T4z = VFNMS(T2u, T2r, VMUL(T2q, T2v)); } { V T2g, T2i, T2l, T2m; T2g = LD(&(ri[WS(rs, 17)]), ms, &(ri[WS(rs, 1)])); T2i = LD(&(ii[WS(rs, 17)]), ms, &(ii[WS(rs, 1)])); T2j = VFMA(T2f, T2g, VMUL(T2h, T2i)); T4h = VFNMS(T2h, T2g, VMUL(T2f, T2i)); T2l = LD(&(ri[WS(rs, 9)]), ms, &(ri[WS(rs, 1)])); T2m = LD(&(ii[WS(rs, 9)]), ms, &(ii[WS(rs, 1)])); T2n = VFMA(T9, T2l, VMUL(Te, T2m)); T4y = VFNMS(Te, T2l, VMUL(T9, T2m)); } { V T2k, T2x, T6w, T6x; T2k = VADD(T2e, T2j); T2x = VADD(T2n, T2w); T2y = VADD(T2k, T2x); T6B = VSUB(T2k, T2x); T6w = VADD(T4g, T4h); T6x = VADD(T4y, T4z); T6y = VSUB(T6w, T6x); T7j = VADD(T6w, T6x); } { V T4i, T4j, T4x, T4A; T4i = VSUB(T4g, T4h); T4j = VSUB(T2n, T2w); T4k = VADD(T4i, T4j); T5J = VSUB(T4i, T4j); T4x = VSUB(T2e, T2j); T4A = VSUB(T4y, T4z); T4B = VSUB(T4x, T4A); T5G = VADD(T4x, T4A); } } { V T31, T4Y, T3f, T4J, T36, T4Z, T3a, T4I; { V T2W, T30, T3c, T3e; T2W = LD(&(ri[WS(rs, 31)]), ms, &(ri[WS(rs, 1)])); T30 = LD(&(ii[WS(rs, 31)]), ms, &(ii[WS(rs, 1)])); T31 = VFMA(T2V, T2W, VMUL(T2Z, T30)); T4Y = VFNMS(T2Z, T2W, VMUL(T2V, T30)); T3c = LD(&(ri[WS(rs, 23)]), ms, &(ri[WS(rs, 1)])); T3e = LD(&(ii[WS(rs, 23)]), ms, &(ii[WS(rs, 1)])); T3f = VFMA(T3b, T3c, VMUL(T3d, T3e)); T4J = VFNMS(T3d, T3c, VMUL(T3b, T3e)); } { V T33, T35, T38, T39; T33 = LD(&(ri[WS(rs, 15)]), ms, &(ri[WS(rs, 1)])); T35 = LD(&(ii[WS(rs, 15)]), ms, &(ii[WS(rs, 1)])); T36 = VFMA(T32, T33, VMUL(T34, T35)); T4Z = VFNMS(T34, T33, VMUL(T32, T35)); T38 = LD(&(ri[WS(rs, 7)]), ms, &(ri[WS(rs, 1)])); T39 = LD(&(ii[WS(rs, 7)]), ms, &(ii[WS(rs, 1)])); T3a = VFMA(TR, T38, VMUL(TS, T39)); T4I = VFNMS(TS, T38, VMUL(TR, T39)); } { V T37, T3g, T6M, T6N; T37 = VADD(T31, T36); T3g = VADD(T3a, T3f); T3h = VADD(T37, T3g); T6H = VSUB(T37, T3g); T6M = VADD(T4Y, T4Z); T6N = VADD(T4I, T4J); T6O = VSUB(T6M, T6N); T7o = VADD(T6M, T6N); } { V T4H, T4K, T50, T51; T4H = VSUB(T31, T36); T4K = VSUB(T4I, T4J); T4L = VSUB(T4H, T4K); T5N = VADD(T4H, T4K); T50 = VSUB(T4Y, T4Z); T51 = VSUB(T3a, T3f); T52 = VADD(T50, T51); T5Q = VSUB(T50, T51); } } { V TQ, T3G, T1g, T3N, TX, T3H, T17, T3M; { V TN, TP, T1b, T1f; TN = LD(&(ri[WS(rs, 4)]), ms, &(ri[0])); TP = LD(&(ii[WS(rs, 4)]), ms, &(ii[0])); TQ = VFMA(TM, TN, VMUL(TO, TP)); T3G = VFNMS(TO, TN, VMUL(TM, TP)); T1b = LD(&(ri[WS(rs, 12)]), ms, &(ri[0])); T1f = LD(&(ii[WS(rs, 12)]), ms, &(ii[0])); T1g = VFMA(T1a, T1b, VMUL(T1e, T1f)); T3N = VFNMS(T1e, T1b, VMUL(T1a, T1f)); } { V TU, TW, T12, T16; TU = LD(&(ri[WS(rs, 20)]), ms, &(ri[0])); TW = LD(&(ii[WS(rs, 20)]), ms, &(ii[0])); TX = VFMA(TT, TU, VMUL(TV, TW)); T3H = VFNMS(TV, TU, VMUL(TT, TW)); T12 = LD(&(ri[WS(rs, 28)]), ms, &(ri[0])); T16 = LD(&(ii[WS(rs, 28)]), ms, &(ii[0])); T17 = VFMA(T11, T12, VMUL(T15, T16)); T3M = VFNMS(T15, T12, VMUL(T11, T16)); } { V TY, T1h, T6g, T6h; TY = VADD(TQ, TX); T1h = VADD(T17, T1g); T1i = VADD(TY, T1h); T7V = VSUB(T1h, TY); T6g = VADD(T3G, T3H); T6h = VADD(T3M, T3N); T6i = VSUB(T6g, T6h); T7D = VADD(T6g, T6h); } { V T3I, T3J, T3L, T3O; T3I = VSUB(T3G, T3H); T3J = VSUB(TQ, TX); T3K = VSUB(T3I, T3J); T5u = VADD(T3J, T3I); T3L = VSUB(T17, T1g); T3O = VSUB(T3M, T3N); T3P = VADD(T3L, T3O); T5v = VSUB(T3L, T3O); } } { V T1m, T3S, T1C, T3Z, T1r, T3T, T1x, T3Y; { V T1k, T1l, T1z, T1B; T1k = LD(&(ri[WS(rs, 2)]), ms, &(ri[0])); T1l = LD(&(ii[WS(rs, 2)]), ms, &(ii[0])); T1m = VFMA(T8, T1k, VMUL(Td, T1l)); T3S = VFNMS(Td, T1k, VMUL(T8, T1l)); T1z = LD(&(ri[WS(rs, 26)]), ms, &(ri[0])); T1B = LD(&(ii[WS(rs, 26)]), ms, &(ii[0])); T1C = VFMA(T1y, T1z, VMUL(T1A, T1B)); T3Z = VFNMS(T1A, T1z, VMUL(T1y, T1B)); } { V T1o, T1q, T1u, T1w; T1o = LD(&(ri[WS(rs, 18)]), ms, &(ri[0])); T1q = LD(&(ii[WS(rs, 18)]), ms, &(ii[0])); T1r = VFMA(T1n, T1o, VMUL(T1p, T1q)); T3T = VFNMS(T1p, T1o, VMUL(T1n, T1q)); T1u = LD(&(ri[WS(rs, 10)]), ms, &(ri[0])); T1w = LD(&(ii[WS(rs, 10)]), ms, &(ii[0])); T1x = VFMA(T1t, T1u, VMUL(T1v, T1w)); T3Y = VFNMS(T1v, T1u, VMUL(T1t, T1w)); } { V T1s, T1D, T6k, T6l; T1s = VADD(T1m, T1r); T1D = VADD(T1x, T1C); T1E = VADD(T1s, T1D); T6n = VSUB(T1s, T1D); T6k = VADD(T3S, T3T); T6l = VADD(T3Y, T3Z); T6m = VSUB(T6k, T6l); T7e = VADD(T6k, T6l); } { V T3U, T3V, T3X, T40; T3U = VSUB(T3S, T3T); T3V = VSUB(T1x, T1C); T3W = VADD(T3U, T3V); T5y = VSUB(T3U, T3V); T3X = VSUB(T1m, T1r); T40 = VSUB(T3Y, T3Z); T41 = VSUB(T3X, T40); T5z = VADD(T3X, T40); } } { V T1J, T43, T27, T4a, T1U, T44, T20, T49; { V T1G, T1I, T24, T26; T1G = LD(&(ri[WS(rs, 30)]), ms, &(ri[0])); T1I = LD(&(ii[WS(rs, 30)]), ms, &(ii[0])); T1J = VFMA(T1F, T1G, VMUL(T1H, T1I)); T43 = VFNMS(T1H, T1G, VMUL(T1F, T1I)); T24 = LD(&(ri[WS(rs, 22)]), ms, &(ri[0])); T26 = LD(&(ii[WS(rs, 22)]), ms, &(ii[0])); T27 = VFMA(T23, T24, VMUL(T25, T26)); T4a = VFNMS(T25, T24, VMUL(T23, T26)); } { V T1R, T1T, T1X, T1Z; T1R = LD(&(ri[WS(rs, 14)]), ms, &(ri[0])); T1T = LD(&(ii[WS(rs, 14)]), ms, &(ii[0])); T1U = VFMA(T1Q, T1R, VMUL(T1S, T1T)); T44 = VFNMS(T1S, T1R, VMUL(T1Q, T1T)); T1X = LD(&(ri[WS(rs, 6)]), ms, &(ri[0])); T1Z = LD(&(ii[WS(rs, 6)]), ms, &(ii[0])); T20 = VFMA(T1W, T1X, VMUL(T1Y, T1Z)); T49 = VFNMS(T1Y, T1X, VMUL(T1W, T1Z)); } { V T1V, T28, T6q, T6r; T1V = VADD(T1J, T1U); T28 = VADD(T20, T27); T29 = VADD(T1V, T28); T6p = VSUB(T1V, T28); T6q = VADD(T43, T44); T6r = VADD(T49, T4a); T6s = VSUB(T6q, T6r); T7f = VADD(T6q, T6r); } { V T45, T46, T48, T4b; T45 = VSUB(T43, T44); T46 = VSUB(T20, T27); T47 = VADD(T45, T46); T5B = VSUB(T45, T46); T48 = VSUB(T1J, T1U); T4b = VSUB(T49, T4a); T4c = VSUB(T48, T4b); T5C = VADD(T48, T4b); } } { V T2B, T4r, T2G, T4s, T4q, T4t, T2M, T4m, T2P, T4n, T4l, T4o; { V T2z, T2A, T2D, T2F; T2z = LD(&(ri[WS(rs, 5)]), ms, &(ri[WS(rs, 1)])); T2A = LD(&(ii[WS(rs, 5)]), ms, &(ii[WS(rs, 1)])); T2B = VFMA(T21, T2z, VMUL(T22, T2A)); T4r = VFNMS(T22, T2z, VMUL(T21, T2A)); T2D = LD(&(ri[WS(rs, 21)]), ms, &(ri[WS(rs, 1)])); T2F = LD(&(ii[WS(rs, 21)]), ms, &(ii[WS(rs, 1)])); T2G = VFMA(T2C, T2D, VMUL(T2E, T2F)); T4s = VFNMS(T2E, T2D, VMUL(T2C, T2F)); } T4q = VSUB(T2B, T2G); T4t = VSUB(T4r, T4s); { V T2J, T2L, T2N, T2O; T2J = LD(&(ri[WS(rs, 29)]), ms, &(ri[WS(rs, 1)])); T2L = LD(&(ii[WS(rs, 29)]), ms, &(ii[WS(rs, 1)])); T2M = VFMA(T2I, T2J, VMUL(T2K, T2L)); T4m = VFNMS(T2K, T2J, VMUL(T2I, T2L)); T2N = LD(&(ri[WS(rs, 13)]), ms, &(ri[WS(rs, 1)])); T2O = LD(&(ii[WS(rs, 13)]), ms, &(ii[WS(rs, 1)])); T2P = VFMA(T1M, T2N, VMUL(T1P, T2O)); T4n = VFNMS(T1P, T2N, VMUL(T1M, T2O)); } T4l = VSUB(T2M, T2P); T4o = VSUB(T4m, T4n); { V T2H, T2Q, T6C, T6D; T2H = VADD(T2B, T2G); T2Q = VADD(T2M, T2P); T2R = VADD(T2H, T2Q); T6z = VSUB(T2Q, T2H); T6C = VADD(T4r, T4s); T6D = VADD(T4m, T4n); T6E = VSUB(T6C, T6D); T7k = VADD(T6C, T6D); } { V T4p, T4u, T4C, T4D; T4p = VSUB(T4l, T4o); T4u = VADD(T4q, T4t); T4v = VMUL(LDK(KP707106781), VSUB(T4p, T4u)); T5H = VMUL(LDK(KP707106781), VADD(T4u, T4p)); T4C = VSUB(T4t, T4q); T4D = VADD(T4l, T4o); T4E = VMUL(LDK(KP707106781), VSUB(T4C, T4D)); T5K = VMUL(LDK(KP707106781), VADD(T4C, T4D)); } } { V T3k, T4M, T3p, T4N, T4O, T4P, T3t, T4S, T3w, T4T, T4R, T4U; { V T3i, T3j, T3m, T3o; T3i = LD(&(ri[WS(rs, 3)]), ms, &(ri[WS(rs, 1)])); T3j = LD(&(ii[WS(rs, 3)]), ms, &(ii[WS(rs, 1)])); T3k = VFMA(T3, T3i, VMUL(T6, T3j)); T4M = VFNMS(T6, T3i, VMUL(T3, T3j)); T3m = LD(&(ri[WS(rs, 19)]), ms, &(ri[WS(rs, 1)])); T3o = LD(&(ii[WS(rs, 19)]), ms, &(ii[WS(rs, 1)])); T3p = VFMA(T3l, T3m, VMUL(T3n, T3o)); T4N = VFNMS(T3n, T3m, VMUL(T3l, T3o)); } T4O = VSUB(T4M, T4N); T4P = VSUB(T3k, T3p); { V T3r, T3s, T3u, T3v; T3r = LD(&(ri[WS(rs, 27)]), ms, &(ri[WS(rs, 1)])); T3s = LD(&(ii[WS(rs, 27)]), ms, &(ii[WS(rs, 1)])); T3t = VFMA(Th, T3r, VMUL(Tl, T3s)); T4S = VFNMS(Tl, T3r, VMUL(Th, T3s)); T3u = LD(&(ri[WS(rs, 11)]), ms, &(ri[WS(rs, 1)])); T3v = LD(&(ii[WS(rs, 11)]), ms, &(ii[WS(rs, 1)])); T3w = VFMA(Tg, T3u, VMUL(Tk, T3v)); T4T = VFNMS(Tk, T3u, VMUL(Tg, T3v)); } T4R = VSUB(T3t, T3w); T4U = VSUB(T4S, T4T); { V T3q, T3x, T6I, T6J; T3q = VADD(T3k, T3p); T3x = VADD(T3t, T3w); T3y = VADD(T3q, T3x); T6P = VSUB(T3x, T3q); T6I = VADD(T4M, T4N); T6J = VADD(T4S, T4T); T6K = VSUB(T6I, T6J); T7p = VADD(T6I, T6J); } { V T4Q, T4V, T53, T54; T4Q = VSUB(T4O, T4P); T4V = VADD(T4R, T4U); T4W = VMUL(LDK(KP707106781), VSUB(T4Q, T4V)); T5R = VMUL(LDK(KP707106781), VADD(T4Q, T4V)); T53 = VSUB(T4R, T4U); T54 = VADD(T4P, T4O); T55 = VMUL(LDK(KP707106781), VSUB(T53, T54)); T5O = VMUL(LDK(KP707106781), VADD(T54, T53)); } } { V T2b, T7x, T7K, T7M, T3A, T7L, T7A, T7B; { V T1j, T2a, T7C, T7J; T1j = VADD(TL, T1i); T2a = VADD(T1E, T29); T2b = VADD(T1j, T2a); T7x = VSUB(T1j, T2a); T7C = VADD(T7e, T7f); T7J = VADD(T7D, T7I); T7K = VADD(T7C, T7J); T7M = VSUB(T7J, T7C); } { V T2S, T3z, T7y, T7z; T2S = VADD(T2y, T2R); T3z = VADD(T3h, T3y); T3A = VADD(T2S, T3z); T7L = VSUB(T3z, T2S); T7y = VADD(T7j, T7k); T7z = VADD(T7o, T7p); T7A = VSUB(T7y, T7z); T7B = VADD(T7y, T7z); } ST(&(ri[WS(rs, 16)]), VSUB(T2b, T3A), ms, &(ri[0])); ST(&(ii[WS(rs, 16)]), VSUB(T7K, T7B), ms, &(ii[0])); ST(&(ri[0]), VADD(T2b, T3A), ms, &(ri[0])); ST(&(ii[0]), VADD(T7B, T7K), ms, &(ii[0])); ST(&(ri[WS(rs, 24)]), VSUB(T7x, T7A), ms, &(ri[0])); ST(&(ii[WS(rs, 24)]), VSUB(T7M, T7L), ms, &(ii[0])); ST(&(ri[WS(rs, 8)]), VADD(T7x, T7A), ms, &(ri[0])); ST(&(ii[WS(rs, 8)]), VADD(T7L, T7M), ms, &(ii[0])); } { V T7h, T7t, T7Q, T7S, T7m, T7u, T7r, T7v; { V T7d, T7g, T7O, T7P; T7d = VSUB(TL, T1i); T7g = VSUB(T7e, T7f); T7h = VADD(T7d, T7g); T7t = VSUB(T7d, T7g); T7O = VSUB(T29, T1E); T7P = VSUB(T7I, T7D); T7Q = VADD(T7O, T7P); T7S = VSUB(T7P, T7O); } { V T7i, T7l, T7n, T7q; T7i = VSUB(T2y, T2R); T7l = VSUB(T7j, T7k); T7m = VADD(T7i, T7l); T7u = VSUB(T7l, T7i); T7n = VSUB(T3h, T3y); T7q = VSUB(T7o, T7p); T7r = VSUB(T7n, T7q); T7v = VADD(T7n, T7q); } { V T7s, T7N, T7w, T7R; T7s = VMUL(LDK(KP707106781), VADD(T7m, T7r)); ST(&(ri[WS(rs, 20)]), VSUB(T7h, T7s), ms, &(ri[0])); ST(&(ri[WS(rs, 4)]), VADD(T7h, T7s), ms, &(ri[0])); T7N = VMUL(LDK(KP707106781), VADD(T7u, T7v)); ST(&(ii[WS(rs, 4)]), VADD(T7N, T7Q), ms, &(ii[0])); ST(&(ii[WS(rs, 20)]), VSUB(T7Q, T7N), ms, &(ii[0])); T7w = VMUL(LDK(KP707106781), VSUB(T7u, T7v)); ST(&(ri[WS(rs, 28)]), VSUB(T7t, T7w), ms, &(ri[0])); ST(&(ri[WS(rs, 12)]), VADD(T7t, T7w), ms, &(ri[0])); T7R = VMUL(LDK(KP707106781), VSUB(T7r, T7m)); ST(&(ii[WS(rs, 12)]), VADD(T7R, T7S), ms, &(ii[0])); ST(&(ii[WS(rs, 28)]), VSUB(T7S, T7R), ms, &(ii[0])); } } { V T6j, T7X, T83, T6X, T6u, T7U, T77, T7b, T70, T82, T6G, T6U, T74, T7a, T6R; V T6V; { V T6o, T6t, T6A, T6F; T6j = VSUB(T6f, T6i); T7X = VADD(T7V, T7W); T83 = VSUB(T7W, T7V); T6X = VADD(T6f, T6i); T6o = VSUB(T6m, T6n); T6t = VADD(T6p, T6s); T6u = VMUL(LDK(KP707106781), VSUB(T6o, T6t)); T7U = VMUL(LDK(KP707106781), VADD(T6o, T6t)); { V T75, T76, T6Y, T6Z; T75 = VADD(T6H, T6K); T76 = VADD(T6O, T6P); T77 = VFNMS(LDK(KP382683432), T76, VMUL(LDK(KP923879532), T75)); T7b = VFMA(LDK(KP923879532), T76, VMUL(LDK(KP382683432), T75)); T6Y = VADD(T6n, T6m); T6Z = VSUB(T6p, T6s); T70 = VMUL(LDK(KP707106781), VADD(T6Y, T6Z)); T82 = VMUL(LDK(KP707106781), VSUB(T6Z, T6Y)); } T6A = VSUB(T6y, T6z); T6F = VSUB(T6B, T6E); T6G = VFMA(LDK(KP923879532), T6A, VMUL(LDK(KP382683432), T6F)); T6U = VFNMS(LDK(KP923879532), T6F, VMUL(LDK(KP382683432), T6A)); { V T72, T73, T6L, T6Q; T72 = VADD(T6y, T6z); T73 = VADD(T6B, T6E); T74 = VFMA(LDK(KP382683432), T72, VMUL(LDK(KP923879532), T73)); T7a = VFNMS(LDK(KP382683432), T73, VMUL(LDK(KP923879532), T72)); T6L = VSUB(T6H, T6K); T6Q = VSUB(T6O, T6P); T6R = VFNMS(LDK(KP923879532), T6Q, VMUL(LDK(KP382683432), T6L)); T6V = VFMA(LDK(KP382683432), T6Q, VMUL(LDK(KP923879532), T6L)); } } { V T6v, T6S, T81, T84; T6v = VADD(T6j, T6u); T6S = VADD(T6G, T6R); ST(&(ri[WS(rs, 22)]), VSUB(T6v, T6S), ms, &(ri[0])); ST(&(ri[WS(rs, 6)]), VADD(T6v, T6S), ms, &(ri[0])); T81 = VADD(T6U, T6V); T84 = VADD(T82, T83); ST(&(ii[WS(rs, 6)]), VADD(T81, T84), ms, &(ii[0])); ST(&(ii[WS(rs, 22)]), VSUB(T84, T81), ms, &(ii[0])); } { V T6T, T6W, T85, T86; T6T = VSUB(T6j, T6u); T6W = VSUB(T6U, T6V); ST(&(ri[WS(rs, 30)]), VSUB(T6T, T6W), ms, &(ri[0])); ST(&(ri[WS(rs, 14)]), VADD(T6T, T6W), ms, &(ri[0])); T85 = VSUB(T6R, T6G); T86 = VSUB(T83, T82); ST(&(ii[WS(rs, 14)]), VADD(T85, T86), ms, &(ii[0])); ST(&(ii[WS(rs, 30)]), VSUB(T86, T85), ms, &(ii[0])); } { V T71, T78, T7T, T7Y; T71 = VADD(T6X, T70); T78 = VADD(T74, T77); ST(&(ri[WS(rs, 18)]), VSUB(T71, T78), ms, &(ri[0])); ST(&(ri[WS(rs, 2)]), VADD(T71, T78), ms, &(ri[0])); T7T = VADD(T7a, T7b); T7Y = VADD(T7U, T7X); ST(&(ii[WS(rs, 2)]), VADD(T7T, T7Y), ms, &(ii[0])); ST(&(ii[WS(rs, 18)]), VSUB(T7Y, T7T), ms, &(ii[0])); } { V T79, T7c, T7Z, T80; T79 = VSUB(T6X, T70); T7c = VSUB(T7a, T7b); ST(&(ri[WS(rs, 26)]), VSUB(T79, T7c), ms, &(ri[0])); ST(&(ri[WS(rs, 10)]), VADD(T79, T7c), ms, &(ri[0])); T7Z = VSUB(T77, T74); T80 = VSUB(T7X, T7U); ST(&(ii[WS(rs, 10)]), VADD(T7Z, T80), ms, &(ii[0])); ST(&(ii[WS(rs, 26)]), VSUB(T80, T7Z), ms, &(ii[0])); } } { V T3R, T5d, T8r, T8x, T4e, T8o, T5n, T5r, T4G, T5a, T5g, T8w, T5k, T5q, T57; V T5b, T3Q, T8p; T3Q = VMUL(LDK(KP707106781), VSUB(T3K, T3P)); T3R = VSUB(T3F, T3Q); T5d = VADD(T3F, T3Q); T8p = VMUL(LDK(KP707106781), VSUB(T5v, T5u)); T8r = VADD(T8p, T8q); T8x = VSUB(T8q, T8p); { V T42, T4d, T5l, T5m; T42 = VFNMS(LDK(KP923879532), T41, VMUL(LDK(KP382683432), T3W)); T4d = VFMA(LDK(KP382683432), T47, VMUL(LDK(KP923879532), T4c)); T4e = VSUB(T42, T4d); T8o = VADD(T42, T4d); T5l = VADD(T4L, T4W); T5m = VADD(T52, T55); T5n = VFNMS(LDK(KP555570233), T5m, VMUL(LDK(KP831469612), T5l)); T5r = VFMA(LDK(KP831469612), T5m, VMUL(LDK(KP555570233), T5l)); } { V T4w, T4F, T5e, T5f; T4w = VSUB(T4k, T4v); T4F = VSUB(T4B, T4E); T4G = VFMA(LDK(KP980785280), T4w, VMUL(LDK(KP195090322), T4F)); T5a = VFNMS(LDK(KP980785280), T4F, VMUL(LDK(KP195090322), T4w)); T5e = VFMA(LDK(KP923879532), T3W, VMUL(LDK(KP382683432), T41)); T5f = VFNMS(LDK(KP923879532), T47, VMUL(LDK(KP382683432), T4c)); T5g = VADD(T5e, T5f); T8w = VSUB(T5f, T5e); } { V T5i, T5j, T4X, T56; T5i = VADD(T4k, T4v); T5j = VADD(T4B, T4E); T5k = VFMA(LDK(KP555570233), T5i, VMUL(LDK(KP831469612), T5j)); T5q = VFNMS(LDK(KP555570233), T5j, VMUL(LDK(KP831469612), T5i)); T4X = VSUB(T4L, T4W); T56 = VSUB(T52, T55); T57 = VFNMS(LDK(KP980785280), T56, VMUL(LDK(KP195090322), T4X)); T5b = VFMA(LDK(KP195090322), T56, VMUL(LDK(KP980785280), T4X)); } { V T4f, T58, T8v, T8y; T4f = VADD(T3R, T4e); T58 = VADD(T4G, T57); ST(&(ri[WS(rs, 23)]), VSUB(T4f, T58), ms, &(ri[WS(rs, 1)])); ST(&(ri[WS(rs, 7)]), VADD(T4f, T58), ms, &(ri[WS(rs, 1)])); T8v = VADD(T5a, T5b); T8y = VADD(T8w, T8x); ST(&(ii[WS(rs, 7)]), VADD(T8v, T8y), ms, &(ii[WS(rs, 1)])); ST(&(ii[WS(rs, 23)]), VSUB(T8y, T8v), ms, &(ii[WS(rs, 1)])); } { V T59, T5c, T8z, T8A; T59 = VSUB(T3R, T4e); T5c = VSUB(T5a, T5b); ST(&(ri[WS(rs, 31)]), VSUB(T59, T5c), ms, &(ri[WS(rs, 1)])); ST(&(ri[WS(rs, 15)]), VADD(T59, T5c), ms, &(ri[WS(rs, 1)])); T8z = VSUB(T57, T4G); T8A = VSUB(T8x, T8w); ST(&(ii[WS(rs, 15)]), VADD(T8z, T8A), ms, &(ii[WS(rs, 1)])); ST(&(ii[WS(rs, 31)]), VSUB(T8A, T8z), ms, &(ii[WS(rs, 1)])); } { V T5h, T5o, T8n, T8s; T5h = VADD(T5d, T5g); T5o = VADD(T5k, T5n); ST(&(ri[WS(rs, 19)]), VSUB(T5h, T5o), ms, &(ri[WS(rs, 1)])); ST(&(ri[WS(rs, 3)]), VADD(T5h, T5o), ms, &(ri[WS(rs, 1)])); T8n = VADD(T5q, T5r); T8s = VADD(T8o, T8r); ST(&(ii[WS(rs, 3)]), VADD(T8n, T8s), ms, &(ii[WS(rs, 1)])); ST(&(ii[WS(rs, 19)]), VSUB(T8s, T8n), ms, &(ii[WS(rs, 1)])); } { V T5p, T5s, T8t, T8u; T5p = VSUB(T5d, T5g); T5s = VSUB(T5q, T5r); ST(&(ri[WS(rs, 27)]), VSUB(T5p, T5s), ms, &(ri[WS(rs, 1)])); ST(&(ri[WS(rs, 11)]), VADD(T5p, T5s), ms, &(ri[WS(rs, 1)])); T8t = VSUB(T5n, T5k); T8u = VSUB(T8r, T8o); ST(&(ii[WS(rs, 11)]), VADD(T8t, T8u), ms, &(ii[WS(rs, 1)])); ST(&(ii[WS(rs, 27)]), VSUB(T8u, T8t), ms, &(ii[WS(rs, 1)])); } } { V T5x, T5Z, T8d, T8j, T5E, T88, T69, T6d, T5M, T5W, T62, T8i, T66, T6c, T5T; V T5X, T5w, T89; T5w = VMUL(LDK(KP707106781), VADD(T5u, T5v)); T5x = VSUB(T5t, T5w); T5Z = VADD(T5t, T5w); T89 = VMUL(LDK(KP707106781), VADD(T3K, T3P)); T8d = VADD(T89, T8c); T8j = VSUB(T8c, T89); { V T5A, T5D, T67, T68; T5A = VFNMS(LDK(KP382683432), T5z, VMUL(LDK(KP923879532), T5y)); T5D = VFMA(LDK(KP923879532), T5B, VMUL(LDK(KP382683432), T5C)); T5E = VSUB(T5A, T5D); T88 = VADD(T5A, T5D); T67 = VADD(T5N, T5O); T68 = VADD(T5Q, T5R); T69 = VFNMS(LDK(KP195090322), T68, VMUL(LDK(KP980785280), T67)); T6d = VFMA(LDK(KP195090322), T67, VMUL(LDK(KP980785280), T68)); } { V T5I, T5L, T60, T61; T5I = VSUB(T5G, T5H); T5L = VSUB(T5J, T5K); T5M = VFMA(LDK(KP555570233), T5I, VMUL(LDK(KP831469612), T5L)); T5W = VFNMS(LDK(KP831469612), T5I, VMUL(LDK(KP555570233), T5L)); T60 = VFMA(LDK(KP382683432), T5y, VMUL(LDK(KP923879532), T5z)); T61 = VFNMS(LDK(KP382683432), T5B, VMUL(LDK(KP923879532), T5C)); T62 = VADD(T60, T61); T8i = VSUB(T61, T60); } { V T64, T65, T5P, T5S; T64 = VADD(T5G, T5H); T65 = VADD(T5J, T5K); T66 = VFMA(LDK(KP980785280), T64, VMUL(LDK(KP195090322), T65)); T6c = VFNMS(LDK(KP195090322), T64, VMUL(LDK(KP980785280), T65)); T5P = VSUB(T5N, T5O); T5S = VSUB(T5Q, T5R); T5T = VFNMS(LDK(KP831469612), T5S, VMUL(LDK(KP555570233), T5P)); T5X = VFMA(LDK(KP831469612), T5P, VMUL(LDK(KP555570233), T5S)); } { V T5F, T5U, T8h, T8k; T5F = VADD(T5x, T5E); T5U = VADD(T5M, T5T); ST(&(ri[WS(rs, 21)]), VSUB(T5F, T5U), ms, &(ri[WS(rs, 1)])); ST(&(ri[WS(rs, 5)]), VADD(T5F, T5U), ms, &(ri[WS(rs, 1)])); T8h = VADD(T5W, T5X); T8k = VADD(T8i, T8j); ST(&(ii[WS(rs, 5)]), VADD(T8h, T8k), ms, &(ii[WS(rs, 1)])); ST(&(ii[WS(rs, 21)]), VSUB(T8k, T8h), ms, &(ii[WS(rs, 1)])); } { V T5V, T5Y, T8l, T8m; T5V = VSUB(T5x, T5E); T5Y = VSUB(T5W, T5X); ST(&(ri[WS(rs, 29)]), VSUB(T5V, T5Y), ms, &(ri[WS(rs, 1)])); ST(&(ri[WS(rs, 13)]), VADD(T5V, T5Y), ms, &(ri[WS(rs, 1)])); T8l = VSUB(T5T, T5M); T8m = VSUB(T8j, T8i); ST(&(ii[WS(rs, 13)]), VADD(T8l, T8m), ms, &(ii[WS(rs, 1)])); ST(&(ii[WS(rs, 29)]), VSUB(T8m, T8l), ms, &(ii[WS(rs, 1)])); } { V T63, T6a, T87, T8e; T63 = VADD(T5Z, T62); T6a = VADD(T66, T69); ST(&(ri[WS(rs, 17)]), VSUB(T63, T6a), ms, &(ri[WS(rs, 1)])); ST(&(ri[WS(rs, 1)]), VADD(T63, T6a), ms, &(ri[WS(rs, 1)])); T87 = VADD(T6c, T6d); T8e = VADD(T88, T8d); ST(&(ii[WS(rs, 1)]), VADD(T87, T8e), ms, &(ii[WS(rs, 1)])); ST(&(ii[WS(rs, 17)]), VSUB(T8e, T87), ms, &(ii[WS(rs, 1)])); } { V T6b, T6e, T8f, T8g; T6b = VSUB(T5Z, T62); T6e = VSUB(T6c, T6d); ST(&(ri[WS(rs, 25)]), VSUB(T6b, T6e), ms, &(ri[WS(rs, 1)])); ST(&(ri[WS(rs, 9)]), VADD(T6b, T6e), ms, &(ri[WS(rs, 1)])); T8f = VSUB(T69, T66); T8g = VSUB(T8d, T88); ST(&(ii[WS(rs, 9)]), VADD(T8f, T8g), ms, &(ii[WS(rs, 1)])); ST(&(ii[WS(rs, 25)]), VSUB(T8g, T8f), ms, &(ii[WS(rs, 1)])); } } } } } VLEAVE(); } static const tw_instr twinstr[] = { VTW(0, 1), VTW(0, 3), VTW(0, 9), VTW(0, 27), {TW_NEXT, (2 * VL), 0} }; static const ct_desc desc = { 32, XSIMD_STRING("t2sv_32"), twinstr, &GENUS, {376, 168, 112, 0}, 0, 0, 0 }; void XSIMD(codelet_t2sv_32) (planner *p) { X(kdft_dit_register) (p, t2sv_32, &desc); } #endif /* HAVE_FMA */ fftw-3.3.3/dft/simd/common/q1fv_2.c0000644000175000017500000001012012054410602013636 00000000000000/* * Copyright (c) 2003, 2007-11 Matteo Frigo * Copyright (c) 2003, 2007-11 Massachusetts Institute of Technology * * This program is free software; you can redistribute it and/or modify * it under the terms of the GNU General Public License as published by * the Free Software Foundation; either version 2 of the License, or * (at your option) any later version. * * This program is distributed in the hope that it will be useful, * but WITHOUT ANY WARRANTY; without even the implied warranty of * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the * GNU General Public License for more details. * * You should have received a copy of the GNU General Public License * along with this program; if not, write to the Free Software * Foundation, Inc., 51 Franklin Street, Fifth Floor, Boston, MA 02110-1301 USA * */ /* This file was automatically generated --- DO NOT EDIT */ /* Generated on Sun Nov 25 07:39:30 EST 2012 */ #include "codelet-dft.h" #ifdef HAVE_FMA /* Generated by: ../../../genfft/gen_twidsq_c.native -fma -reorder-insns -schedule-for-pipeline -simd -compact -variables 4 -pipeline-latency 8 -n 2 -dif -name q1fv_2 -include q1f.h */ /* * This function contains 6 FP additions, 4 FP multiplications, * (or, 6 additions, 4 multiplications, 0 fused multiply/add), * 8 stack variables, 0 constants, and 8 memory accesses */ #include "q1f.h" static void q1fv_2(R *ri, R *ii, const R *W, stride rs, stride vs, INT mb, INT me, INT ms) { { INT m; R *x; x = ri; for (m = mb, W = W + (mb * ((TWVL / VL) * 2)); m < me; m = m + VL, x = x + (VL * ms), W = W + (TWVL * 2), MAKE_VOLATILE_STRIDE(4, rs), MAKE_VOLATILE_STRIDE(4, vs)) { V T1, T2, T4, T5, T3, T6; T1 = LD(&(x[0]), ms, &(x[0])); T2 = LD(&(x[WS(rs, 1)]), ms, &(x[WS(rs, 1)])); T4 = LD(&(x[WS(vs, 1)]), ms, &(x[WS(vs, 1)])); T5 = LD(&(x[WS(vs, 1) + WS(rs, 1)]), ms, &(x[WS(vs, 1) + WS(rs, 1)])); ST(&(x[0]), VADD(T1, T2), ms, &(x[0])); T3 = BYTWJ(&(W[0]), VSUB(T1, T2)); ST(&(x[WS(rs, 1)]), VADD(T4, T5), ms, &(x[WS(rs, 1)])); T6 = BYTWJ(&(W[0]), VSUB(T4, T5)); ST(&(x[WS(vs, 1)]), T3, ms, &(x[WS(vs, 1)])); ST(&(x[WS(vs, 1) + WS(rs, 1)]), T6, ms, &(x[WS(vs, 1) + WS(rs, 1)])); } } VLEAVE(); } static const tw_instr twinstr[] = { VTW(0, 1), {TW_NEXT, VL, 0} }; static const ct_desc desc = { 2, XSIMD_STRING("q1fv_2"), twinstr, &GENUS, {6, 4, 0, 0}, 0, 0, 0 }; void XSIMD(codelet_q1fv_2) (planner *p) { X(kdft_difsq_register) (p, q1fv_2, &desc); } #else /* HAVE_FMA */ /* Generated by: ../../../genfft/gen_twidsq_c.native -simd -compact -variables 4 -pipeline-latency 8 -n 2 -dif -name q1fv_2 -include q1f.h */ /* * This function contains 6 FP additions, 4 FP multiplications, * (or, 6 additions, 4 multiplications, 0 fused multiply/add), * 8 stack variables, 0 constants, and 8 memory accesses */ #include "q1f.h" static void q1fv_2(R *ri, R *ii, const R *W, stride rs, stride vs, INT mb, INT me, INT ms) { { INT m; R *x; x = ri; for (m = mb, W = W + (mb * ((TWVL / VL) * 2)); m < me; m = m + VL, x = x + (VL * ms), W = W + (TWVL * 2), MAKE_VOLATILE_STRIDE(4, rs), MAKE_VOLATILE_STRIDE(4, vs)) { V T1, T2, T3, T4, T5, T6; T1 = LD(&(x[0]), ms, &(x[0])); T2 = LD(&(x[WS(rs, 1)]), ms, &(x[WS(rs, 1)])); T3 = BYTWJ(&(W[0]), VSUB(T1, T2)); T4 = LD(&(x[WS(vs, 1)]), ms, &(x[WS(vs, 1)])); T5 = LD(&(x[WS(vs, 1) + WS(rs, 1)]), ms, &(x[WS(vs, 1) + WS(rs, 1)])); T6 = BYTWJ(&(W[0]), VSUB(T4, T5)); ST(&(x[WS(vs, 1)]), T3, ms, &(x[WS(vs, 1)])); ST(&(x[WS(vs, 1) + WS(rs, 1)]), T6, ms, &(x[WS(vs, 1) + WS(rs, 1)])); ST(&(x[0]), VADD(T1, T2), ms, &(x[0])); ST(&(x[WS(rs, 1)]), VADD(T4, T5), ms, &(x[WS(rs, 1)])); } } VLEAVE(); } static const tw_instr twinstr[] = { VTW(0, 1), {TW_NEXT, VL, 0} }; static const ct_desc desc = { 2, XSIMD_STRING("q1fv_2"), twinstr, &GENUS, {6, 4, 0, 0}, 0, 0, 0 }; void XSIMD(codelet_q1fv_2) (planner *p) { X(kdft_difsq_register) (p, q1fv_2, &desc); } #endif /* HAVE_FMA */ fftw-3.3.3/dft/simd/common/q1fv_4.c0000644000175000017500000002277312054410603013662 00000000000000/* * Copyright (c) 2003, 2007-11 Matteo Frigo * Copyright (c) 2003, 2007-11 Massachusetts Institute of Technology * * This program is free software; you can redistribute it and/or modify * it under the terms of the GNU General Public License as published by * the Free Software Foundation; either version 2 of the License, or * (at your option) any later version. * * This program is distributed in the hope that it will be useful, * but WITHOUT ANY WARRANTY; without even the implied warranty of * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the * GNU General Public License for more details. * * You should have received a copy of the GNU General Public License * along with this program; if not, write to the Free Software * Foundation, Inc., 51 Franklin Street, Fifth Floor, Boston, MA 02110-1301 USA * */ /* This file was automatically generated --- DO NOT EDIT */ /* Generated on Sun Nov 25 07:39:31 EST 2012 */ #include "codelet-dft.h" #ifdef HAVE_FMA /* Generated by: ../../../genfft/gen_twidsq_c.native -fma -reorder-insns -schedule-for-pipeline -simd -compact -variables 4 -pipeline-latency 8 -n 4 -dif -name q1fv_4 -include q1f.h */ /* * This function contains 44 FP additions, 32 FP multiplications, * (or, 36 additions, 24 multiplications, 8 fused multiply/add), * 38 stack variables, 0 constants, and 32 memory accesses */ #include "q1f.h" static void q1fv_4(R *ri, R *ii, const R *W, stride rs, stride vs, INT mb, INT me, INT ms) { { INT m; R *x; x = ri; for (m = mb, W = W + (mb * ((TWVL / VL) * 6)); m < me; m = m + VL, x = x + (VL * ms), W = W + (TWVL * 6), MAKE_VOLATILE_STRIDE(8, rs), MAKE_VOLATILE_STRIDE(8, vs)) { V Tb, Tm, Tx, TI; { V Tc, T9, T3, TG, TA, TH, TD, Ta, T6, Td, Tn, To, Tq, Tr, Tf; V Tg; { V T1, T2, Ty, Tz, TB, TC, T4, T5; T1 = LD(&(x[0]), ms, &(x[0])); T2 = LD(&(x[WS(rs, 2)]), ms, &(x[0])); Ty = LD(&(x[WS(vs, 3)]), ms, &(x[WS(vs, 3)])); Tz = LD(&(x[WS(vs, 3) + WS(rs, 2)]), ms, &(x[WS(vs, 3)])); TB = LD(&(x[WS(vs, 3) + WS(rs, 1)]), ms, &(x[WS(vs, 3) + WS(rs, 1)])); TC = LD(&(x[WS(vs, 3) + WS(rs, 3)]), ms, &(x[WS(vs, 3) + WS(rs, 1)])); T4 = LD(&(x[WS(rs, 1)]), ms, &(x[WS(rs, 1)])); T5 = LD(&(x[WS(rs, 3)]), ms, &(x[WS(rs, 1)])); Tc = LD(&(x[WS(vs, 1)]), ms, &(x[WS(vs, 1)])); T9 = VADD(T1, T2); T3 = VSUB(T1, T2); TG = VADD(Ty, Tz); TA = VSUB(Ty, Tz); TH = VADD(TB, TC); TD = VSUB(TB, TC); Ta = VADD(T4, T5); T6 = VSUB(T4, T5); Td = LD(&(x[WS(vs, 1) + WS(rs, 2)]), ms, &(x[WS(vs, 1)])); Tn = LD(&(x[WS(vs, 2)]), ms, &(x[WS(vs, 2)])); To = LD(&(x[WS(vs, 2) + WS(rs, 2)]), ms, &(x[WS(vs, 2)])); Tq = LD(&(x[WS(vs, 2) + WS(rs, 1)]), ms, &(x[WS(vs, 2) + WS(rs, 1)])); Tr = LD(&(x[WS(vs, 2) + WS(rs, 3)]), ms, &(x[WS(vs, 2) + WS(rs, 1)])); Tf = LD(&(x[WS(vs, 1) + WS(rs, 1)]), ms, &(x[WS(vs, 1) + WS(rs, 1)])); Tg = LD(&(x[WS(vs, 1) + WS(rs, 3)]), ms, &(x[WS(vs, 1) + WS(rs, 1)])); } { V Tk, Te, Tv, Tp, Tw, Ts, Tl, Th, T7, TE, Tu, TF; ST(&(x[0]), VADD(T9, Ta), ms, &(x[0])); Tk = VADD(Tc, Td); Te = VSUB(Tc, Td); Tv = VADD(Tn, To); Tp = VSUB(Tn, To); Tw = VADD(Tq, Tr); Ts = VSUB(Tq, Tr); Tl = VADD(Tf, Tg); Th = VSUB(Tf, Tg); ST(&(x[WS(rs, 3)]), VADD(TG, TH), ms, &(x[WS(rs, 1)])); T7 = BYTWJ(&(W[0]), VFNMSI(T6, T3)); TE = BYTWJ(&(W[0]), VFNMSI(TD, TA)); { V Tt, Ti, Tj, T8; T8 = BYTWJ(&(W[TWVL * 4]), VFMAI(T6, T3)); ST(&(x[WS(rs, 2)]), VADD(Tv, Tw), ms, &(x[0])); Tt = BYTWJ(&(W[0]), VFNMSI(Ts, Tp)); ST(&(x[WS(rs, 1)]), VADD(Tk, Tl), ms, &(x[WS(rs, 1)])); Ti = BYTWJ(&(W[0]), VFNMSI(Th, Te)); Tj = BYTWJ(&(W[TWVL * 4]), VFMAI(Th, Te)); ST(&(x[WS(vs, 1)]), T7, ms, &(x[WS(vs, 1)])); ST(&(x[WS(vs, 1) + WS(rs, 3)]), TE, ms, &(x[WS(vs, 1) + WS(rs, 1)])); ST(&(x[WS(vs, 3)]), T8, ms, &(x[WS(vs, 3)])); Tu = BYTWJ(&(W[TWVL * 4]), VFMAI(Ts, Tp)); ST(&(x[WS(vs, 1) + WS(rs, 2)]), Tt, ms, &(x[WS(vs, 1)])); TF = BYTWJ(&(W[TWVL * 4]), VFMAI(TD, TA)); ST(&(x[WS(vs, 1) + WS(rs, 1)]), Ti, ms, &(x[WS(vs, 1) + WS(rs, 1)])); ST(&(x[WS(vs, 3) + WS(rs, 1)]), Tj, ms, &(x[WS(vs, 3) + WS(rs, 1)])); } Tb = BYTWJ(&(W[TWVL * 2]), VSUB(T9, Ta)); Tm = BYTWJ(&(W[TWVL * 2]), VSUB(Tk, Tl)); Tx = BYTWJ(&(W[TWVL * 2]), VSUB(Tv, Tw)); ST(&(x[WS(vs, 3) + WS(rs, 2)]), Tu, ms, &(x[WS(vs, 3)])); TI = BYTWJ(&(W[TWVL * 2]), VSUB(TG, TH)); ST(&(x[WS(vs, 3) + WS(rs, 3)]), TF, ms, &(x[WS(vs, 3) + WS(rs, 1)])); } } ST(&(x[WS(vs, 2)]), Tb, ms, &(x[WS(vs, 2)])); ST(&(x[WS(vs, 2) + WS(rs, 1)]), Tm, ms, &(x[WS(vs, 2) + WS(rs, 1)])); ST(&(x[WS(vs, 2) + WS(rs, 2)]), Tx, ms, &(x[WS(vs, 2)])); ST(&(x[WS(vs, 2) + WS(rs, 3)]), TI, ms, &(x[WS(vs, 2) + WS(rs, 1)])); } } VLEAVE(); } static const tw_instr twinstr[] = { VTW(0, 1), VTW(0, 2), VTW(0, 3), {TW_NEXT, VL, 0} }; static const ct_desc desc = { 4, XSIMD_STRING("q1fv_4"), twinstr, &GENUS, {36, 24, 8, 0}, 0, 0, 0 }; void XSIMD(codelet_q1fv_4) (planner *p) { X(kdft_difsq_register) (p, q1fv_4, &desc); } #else /* HAVE_FMA */ /* Generated by: ../../../genfft/gen_twidsq_c.native -simd -compact -variables 4 -pipeline-latency 8 -n 4 -dif -name q1fv_4 -include q1f.h */ /* * This function contains 44 FP additions, 24 FP multiplications, * (or, 44 additions, 24 multiplications, 0 fused multiply/add), * 22 stack variables, 0 constants, and 32 memory accesses */ #include "q1f.h" static void q1fv_4(R *ri, R *ii, const R *W, stride rs, stride vs, INT mb, INT me, INT ms) { { INT m; R *x; x = ri; for (m = mb, W = W + (mb * ((TWVL / VL) * 6)); m < me; m = m + VL, x = x + (VL * ms), W = W + (TWVL * 6), MAKE_VOLATILE_STRIDE(8, rs), MAKE_VOLATILE_STRIDE(8, vs)) { V T3, T9, TA, TG, TD, TH, T6, Ta, Te, Tk, Tp, Tv, Ts, Tw, Th; V Tl; { V T1, T2, Ty, Tz; T1 = LD(&(x[0]), ms, &(x[0])); T2 = LD(&(x[WS(rs, 2)]), ms, &(x[0])); T3 = VSUB(T1, T2); T9 = VADD(T1, T2); Ty = LD(&(x[WS(vs, 3)]), ms, &(x[WS(vs, 3)])); Tz = LD(&(x[WS(vs, 3) + WS(rs, 2)]), ms, &(x[WS(vs, 3)])); TA = VSUB(Ty, Tz); TG = VADD(Ty, Tz); } { V TB, TC, T4, T5; TB = LD(&(x[WS(vs, 3) + WS(rs, 1)]), ms, &(x[WS(vs, 3) + WS(rs, 1)])); TC = LD(&(x[WS(vs, 3) + WS(rs, 3)]), ms, &(x[WS(vs, 3) + WS(rs, 1)])); TD = VBYI(VSUB(TB, TC)); TH = VADD(TB, TC); T4 = LD(&(x[WS(rs, 1)]), ms, &(x[WS(rs, 1)])); T5 = LD(&(x[WS(rs, 3)]), ms, &(x[WS(rs, 1)])); T6 = VBYI(VSUB(T4, T5)); Ta = VADD(T4, T5); } { V Tc, Td, Tn, To; Tc = LD(&(x[WS(vs, 1)]), ms, &(x[WS(vs, 1)])); Td = LD(&(x[WS(vs, 1) + WS(rs, 2)]), ms, &(x[WS(vs, 1)])); Te = VSUB(Tc, Td); Tk = VADD(Tc, Td); Tn = LD(&(x[WS(vs, 2)]), ms, &(x[WS(vs, 2)])); To = LD(&(x[WS(vs, 2) + WS(rs, 2)]), ms, &(x[WS(vs, 2)])); Tp = VSUB(Tn, To); Tv = VADD(Tn, To); } { V Tq, Tr, Tf, Tg; Tq = LD(&(x[WS(vs, 2) + WS(rs, 1)]), ms, &(x[WS(vs, 2) + WS(rs, 1)])); Tr = LD(&(x[WS(vs, 2) + WS(rs, 3)]), ms, &(x[WS(vs, 2) + WS(rs, 1)])); Ts = VBYI(VSUB(Tq, Tr)); Tw = VADD(Tq, Tr); Tf = LD(&(x[WS(vs, 1) + WS(rs, 1)]), ms, &(x[WS(vs, 1) + WS(rs, 1)])); Tg = LD(&(x[WS(vs, 1) + WS(rs, 3)]), ms, &(x[WS(vs, 1) + WS(rs, 1)])); Th = VBYI(VSUB(Tf, Tg)); Tl = VADD(Tf, Tg); } ST(&(x[0]), VADD(T9, Ta), ms, &(x[0])); ST(&(x[WS(rs, 1)]), VADD(Tk, Tl), ms, &(x[WS(rs, 1)])); ST(&(x[WS(rs, 2)]), VADD(Tv, Tw), ms, &(x[0])); ST(&(x[WS(rs, 3)]), VADD(TG, TH), ms, &(x[WS(rs, 1)])); { V T7, Ti, Tt, TE; T7 = BYTWJ(&(W[0]), VSUB(T3, T6)); ST(&(x[WS(vs, 1)]), T7, ms, &(x[WS(vs, 1)])); Ti = BYTWJ(&(W[0]), VSUB(Te, Th)); ST(&(x[WS(vs, 1) + WS(rs, 1)]), Ti, ms, &(x[WS(vs, 1) + WS(rs, 1)])); Tt = BYTWJ(&(W[0]), VSUB(Tp, Ts)); ST(&(x[WS(vs, 1) + WS(rs, 2)]), Tt, ms, &(x[WS(vs, 1)])); TE = BYTWJ(&(W[0]), VSUB(TA, TD)); ST(&(x[WS(vs, 1) + WS(rs, 3)]), TE, ms, &(x[WS(vs, 1) + WS(rs, 1)])); } { V T8, Tj, Tu, TF; T8 = BYTWJ(&(W[TWVL * 4]), VADD(T3, T6)); ST(&(x[WS(vs, 3)]), T8, ms, &(x[WS(vs, 3)])); Tj = BYTWJ(&(W[TWVL * 4]), VADD(Te, Th)); ST(&(x[WS(vs, 3) + WS(rs, 1)]), Tj, ms, &(x[WS(vs, 3) + WS(rs, 1)])); Tu = BYTWJ(&(W[TWVL * 4]), VADD(Tp, Ts)); ST(&(x[WS(vs, 3) + WS(rs, 2)]), Tu, ms, &(x[WS(vs, 3)])); TF = BYTWJ(&(W[TWVL * 4]), VADD(TA, TD)); ST(&(x[WS(vs, 3) + WS(rs, 3)]), TF, ms, &(x[WS(vs, 3) + WS(rs, 1)])); } { V Tb, Tm, Tx, TI; Tb = BYTWJ(&(W[TWVL * 2]), VSUB(T9, Ta)); ST(&(x[WS(vs, 2)]), Tb, ms, &(x[WS(vs, 2)])); Tm = BYTWJ(&(W[TWVL * 2]), VSUB(Tk, Tl)); ST(&(x[WS(vs, 2) + WS(rs, 1)]), Tm, ms, &(x[WS(vs, 2) + WS(rs, 1)])); Tx = BYTWJ(&(W[TWVL * 2]), VSUB(Tv, Tw)); ST(&(x[WS(vs, 2) + WS(rs, 2)]), Tx, ms, &(x[WS(vs, 2)])); TI = BYTWJ(&(W[TWVL * 2]), VSUB(TG, TH)); ST(&(x[WS(vs, 2) + WS(rs, 3)]), TI, ms, &(x[WS(vs, 2) + WS(rs, 1)])); } } } VLEAVE(); } static const tw_instr twinstr[] = { VTW(0, 1), VTW(0, 2), VTW(0, 3), {TW_NEXT, VL, 0} }; static const ct_desc desc = { 4, XSIMD_STRING("q1fv_4"), twinstr, &GENUS, {44, 24, 0, 0}, 0, 0, 0 }; void XSIMD(codelet_q1fv_4) (planner *p) { X(kdft_difsq_register) (p, q1fv_4, &desc); } #endif /* HAVE_FMA */ fftw-3.3.3/dft/simd/common/q1fv_5.c0000644000175000017500000004313312054410603013654 00000000000000/* * Copyright (c) 2003, 2007-11 Matteo Frigo * Copyright (c) 2003, 2007-11 Massachusetts Institute of Technology * * This program is free software; you can redistribute it and/or modify * it under the terms of the GNU General Public License as published by * the Free Software Foundation; either version 2 of the License, or * (at your option) any later version. * * This program is distributed in the hope that it will be useful, * but WITHOUT ANY WARRANTY; without even the implied warranty of * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the * GNU General Public License for more details. * * You should have received a copy of the GNU General Public License * along with this program; if not, write to the Free Software * Foundation, Inc., 51 Franklin Street, Fifth Floor, Boston, MA 02110-1301 USA * */ /* This file was automatically generated --- DO NOT EDIT */ /* Generated on Sun Nov 25 07:39:31 EST 2012 */ #include "codelet-dft.h" #ifdef HAVE_FMA /* Generated by: ../../../genfft/gen_twidsq_c.native -fma -reorder-insns -schedule-for-pipeline -simd -compact -variables 4 -pipeline-latency 8 -n 5 -dif -name q1fv_5 -include q1f.h */ /* * This function contains 100 FP additions, 95 FP multiplications, * (or, 55 additions, 50 multiplications, 45 fused multiply/add), * 69 stack variables, 4 constants, and 50 memory accesses */ #include "q1f.h" static void q1fv_5(R *ri, R *ii, const R *W, stride rs, stride vs, INT mb, INT me, INT ms) { DVK(KP559016994, +0.559016994374947424102293417182819058860154590); DVK(KP250000000, +0.250000000000000000000000000000000000000000000); DVK(KP618033988, +0.618033988749894848204586834365638117720309180); DVK(KP951056516, +0.951056516295153572116439333379382143405698634); { INT m; R *x; x = ri; for (m = mb, W = W + (mb * ((TWVL / VL) * 8)); m < me; m = m + VL, x = x + (VL * ms), W = W + (TWVL * 8), MAKE_VOLATILE_STRIDE(10, rs), MAKE_VOLATILE_STRIDE(10, vs)) { V Te, T1w, Ty, TS, TW, Tb, T1t, Tv, T1g, T1c, TP, TV, T1f, T19, TY; V TX; { V T1, T1j, Tl, Ti, Ta, T8, T1A, T1q, T1s, T9, TF, T1r, TZ, TR, TL; V TC, Ts, Tu, TQ, TI, T15, T1b, T10, T11, Tt; { V T1n, T1o, T1k, T1l, T7, Td, T4, Tc; { V T5, T6, T2, T3; T1 = LD(&(x[0]), ms, &(x[0])); T5 = LD(&(x[WS(rs, 2)]), ms, &(x[0])); T6 = LD(&(x[WS(rs, 3)]), ms, &(x[WS(rs, 1)])); T2 = LD(&(x[WS(rs, 1)]), ms, &(x[WS(rs, 1)])); T3 = LD(&(x[WS(rs, 4)]), ms, &(x[0])); T1j = LD(&(x[WS(vs, 4)]), ms, &(x[WS(vs, 4)])); T1n = LD(&(x[WS(vs, 4) + WS(rs, 2)]), ms, &(x[WS(vs, 4)])); T1o = LD(&(x[WS(vs, 4) + WS(rs, 3)]), ms, &(x[WS(vs, 4) + WS(rs, 1)])); T1k = LD(&(x[WS(vs, 4) + WS(rs, 1)]), ms, &(x[WS(vs, 4) + WS(rs, 1)])); T1l = LD(&(x[WS(vs, 4) + WS(rs, 4)]), ms, &(x[WS(vs, 4)])); T7 = VADD(T5, T6); Td = VSUB(T5, T6); T4 = VADD(T2, T3); Tc = VSUB(T2, T3); } { V Tm, Tn, Tr, Tx, T1v, T1p; Tl = LD(&(x[WS(vs, 1)]), ms, &(x[WS(vs, 1)])); T1v = VSUB(T1n, T1o); T1p = VADD(T1n, T1o); { V T1u, T1m, Tp, Tq; T1u = VSUB(T1k, T1l); T1m = VADD(T1k, T1l); Tp = LD(&(x[WS(vs, 1) + WS(rs, 2)]), ms, &(x[WS(vs, 1)])); Ti = VMUL(LDK(KP951056516), VFNMS(LDK(KP618033988), Tc, Td)); Te = VMUL(LDK(KP951056516), VFMA(LDK(KP618033988), Td, Tc)); Ta = VSUB(T4, T7); T8 = VADD(T4, T7); Tq = LD(&(x[WS(vs, 1) + WS(rs, 3)]), ms, &(x[WS(vs, 1) + WS(rs, 1)])); T1w = VMUL(LDK(KP951056516), VFMA(LDK(KP618033988), T1v, T1u)); T1A = VMUL(LDK(KP951056516), VFNMS(LDK(KP618033988), T1u, T1v)); T1q = VADD(T1m, T1p); T1s = VSUB(T1m, T1p); Tm = LD(&(x[WS(vs, 1) + WS(rs, 1)]), ms, &(x[WS(vs, 1) + WS(rs, 1)])); T9 = VFNMS(LDK(KP250000000), T8, T1); Tn = LD(&(x[WS(vs, 1) + WS(rs, 4)]), ms, &(x[WS(vs, 1)])); Tr = VADD(Tp, Tq); Tx = VSUB(Tp, Tq); } { V TJ, TK, TG, Tw, To, TH, T13, T14; TF = LD(&(x[WS(vs, 2)]), ms, &(x[WS(vs, 2)])); T1r = VFNMS(LDK(KP250000000), T1q, T1j); TJ = LD(&(x[WS(vs, 2) + WS(rs, 2)]), ms, &(x[WS(vs, 2)])); TK = LD(&(x[WS(vs, 2) + WS(rs, 3)]), ms, &(x[WS(vs, 2) + WS(rs, 1)])); TG = LD(&(x[WS(vs, 2) + WS(rs, 1)]), ms, &(x[WS(vs, 2) + WS(rs, 1)])); Tw = VSUB(Tm, Tn); To = VADD(Tm, Tn); TH = LD(&(x[WS(vs, 2) + WS(rs, 4)]), ms, &(x[WS(vs, 2)])); TZ = LD(&(x[WS(vs, 3)]), ms, &(x[WS(vs, 3)])); T13 = LD(&(x[WS(vs, 3) + WS(rs, 2)]), ms, &(x[WS(vs, 3)])); T14 = LD(&(x[WS(vs, 3) + WS(rs, 3)]), ms, &(x[WS(vs, 3) + WS(rs, 1)])); TR = VSUB(TJ, TK); TL = VADD(TJ, TK); Ty = VMUL(LDK(KP951056516), VFMA(LDK(KP618033988), Tx, Tw)); TC = VMUL(LDK(KP951056516), VFNMS(LDK(KP618033988), Tw, Tx)); Ts = VADD(To, Tr); Tu = VSUB(To, Tr); TQ = VSUB(TG, TH); TI = VADD(TG, TH); T15 = VADD(T13, T14); T1b = VSUB(T13, T14); T10 = LD(&(x[WS(vs, 3) + WS(rs, 1)]), ms, &(x[WS(vs, 3) + WS(rs, 1)])); T11 = LD(&(x[WS(vs, 3) + WS(rs, 4)]), ms, &(x[WS(vs, 3)])); Tt = VFNMS(LDK(KP250000000), Ts, Tl); } } } { V TO, T12, T1a, Th, T1z, TN, TM, T18, T17; ST(&(x[0]), VADD(T1, T8), ms, &(x[0])); TS = VMUL(LDK(KP951056516), VFMA(LDK(KP618033988), TR, TQ)); TW = VMUL(LDK(KP951056516), VFNMS(LDK(KP618033988), TQ, TR)); TM = VADD(TI, TL); TO = VSUB(TI, TL); ST(&(x[WS(rs, 4)]), VADD(T1j, T1q), ms, &(x[0])); T12 = VADD(T10, T11); T1a = VSUB(T10, T11); ST(&(x[WS(rs, 1)]), VADD(Tl, Ts), ms, &(x[WS(rs, 1)])); Th = VFNMS(LDK(KP559016994), Ta, T9); Tb = VFMA(LDK(KP559016994), Ta, T9); T1t = VFMA(LDK(KP559016994), T1s, T1r); T1z = VFNMS(LDK(KP559016994), T1s, T1r); ST(&(x[WS(rs, 2)]), VADD(TF, TM), ms, &(x[0])); TN = VFNMS(LDK(KP250000000), TM, TF); { V T16, Tk, Tj, T1C, T1B, TD, TE, TB; TB = VFNMS(LDK(KP559016994), Tu, Tt); Tv = VFMA(LDK(KP559016994), Tu, Tt); T1g = VMUL(LDK(KP951056516), VFNMS(LDK(KP618033988), T1a, T1b)); T1c = VMUL(LDK(KP951056516), VFMA(LDK(KP618033988), T1b, T1a)); T18 = VSUB(T12, T15); T16 = VADD(T12, T15); Tk = BYTWJ(&(W[TWVL * 4]), VFNMSI(Ti, Th)); Tj = BYTWJ(&(W[TWVL * 2]), VFMAI(Ti, Th)); T1C = BYTWJ(&(W[TWVL * 4]), VFNMSI(T1A, T1z)); T1B = BYTWJ(&(W[TWVL * 2]), VFMAI(T1A, T1z)); TD = BYTWJ(&(W[TWVL * 2]), VFMAI(TC, TB)); TE = BYTWJ(&(W[TWVL * 4]), VFNMSI(TC, TB)); ST(&(x[WS(rs, 3)]), VADD(TZ, T16), ms, &(x[WS(rs, 1)])); T17 = VFNMS(LDK(KP250000000), T16, TZ); ST(&(x[WS(vs, 3)]), Tk, ms, &(x[WS(vs, 3)])); ST(&(x[WS(vs, 2)]), Tj, ms, &(x[WS(vs, 2)])); ST(&(x[WS(vs, 3) + WS(rs, 4)]), T1C, ms, &(x[WS(vs, 3)])); ST(&(x[WS(vs, 2) + WS(rs, 4)]), T1B, ms, &(x[WS(vs, 2)])); ST(&(x[WS(vs, 2) + WS(rs, 1)]), TD, ms, &(x[WS(vs, 2) + WS(rs, 1)])); ST(&(x[WS(vs, 3) + WS(rs, 1)]), TE, ms, &(x[WS(vs, 3) + WS(rs, 1)])); } TP = VFMA(LDK(KP559016994), TO, TN); TV = VFNMS(LDK(KP559016994), TO, TN); T1f = VFNMS(LDK(KP559016994), T18, T17); T19 = VFMA(LDK(KP559016994), T18, T17); } } TY = BYTWJ(&(W[TWVL * 4]), VFNMSI(TW, TV)); TX = BYTWJ(&(W[TWVL * 2]), VFMAI(TW, TV)); { V T1i, T1h, TU, TT; T1i = BYTWJ(&(W[TWVL * 4]), VFNMSI(T1g, T1f)); T1h = BYTWJ(&(W[TWVL * 2]), VFMAI(T1g, T1f)); TU = BYTWJ(&(W[TWVL * 6]), VFMAI(TS, TP)); TT = BYTWJ(&(W[0]), VFNMSI(TS, TP)); { V Tg, Tf, TA, Tz; Tg = BYTWJ(&(W[TWVL * 6]), VFMAI(Te, Tb)); Tf = BYTWJ(&(W[0]), VFNMSI(Te, Tb)); TA = BYTWJ(&(W[TWVL * 6]), VFMAI(Ty, Tv)); Tz = BYTWJ(&(W[0]), VFNMSI(Ty, Tv)); { V T1e, T1d, T1y, T1x; T1e = BYTWJ(&(W[TWVL * 6]), VFMAI(T1c, T19)); T1d = BYTWJ(&(W[0]), VFNMSI(T1c, T19)); T1y = BYTWJ(&(W[TWVL * 6]), VFMAI(T1w, T1t)); T1x = BYTWJ(&(W[0]), VFNMSI(T1w, T1t)); ST(&(x[WS(vs, 3) + WS(rs, 2)]), TY, ms, &(x[WS(vs, 3)])); ST(&(x[WS(vs, 2) + WS(rs, 2)]), TX, ms, &(x[WS(vs, 2)])); ST(&(x[WS(vs, 3) + WS(rs, 3)]), T1i, ms, &(x[WS(vs, 3) + WS(rs, 1)])); ST(&(x[WS(vs, 2) + WS(rs, 3)]), T1h, ms, &(x[WS(vs, 2) + WS(rs, 1)])); ST(&(x[WS(vs, 4) + WS(rs, 2)]), TU, ms, &(x[WS(vs, 4)])); ST(&(x[WS(vs, 1) + WS(rs, 2)]), TT, ms, &(x[WS(vs, 1)])); ST(&(x[WS(vs, 4)]), Tg, ms, &(x[WS(vs, 4)])); ST(&(x[WS(vs, 1)]), Tf, ms, &(x[WS(vs, 1)])); ST(&(x[WS(vs, 4) + WS(rs, 1)]), TA, ms, &(x[WS(vs, 4) + WS(rs, 1)])); ST(&(x[WS(vs, 1) + WS(rs, 1)]), Tz, ms, &(x[WS(vs, 1) + WS(rs, 1)])); ST(&(x[WS(vs, 4) + WS(rs, 3)]), T1e, ms, &(x[WS(vs, 4) + WS(rs, 1)])); ST(&(x[WS(vs, 1) + WS(rs, 3)]), T1d, ms, &(x[WS(vs, 1) + WS(rs, 1)])); ST(&(x[WS(vs, 4) + WS(rs, 4)]), T1y, ms, &(x[WS(vs, 4)])); ST(&(x[WS(vs, 1) + WS(rs, 4)]), T1x, ms, &(x[WS(vs, 1)])); } } } } } VLEAVE(); } static const tw_instr twinstr[] = { VTW(0, 1), VTW(0, 2), VTW(0, 3), VTW(0, 4), {TW_NEXT, VL, 0} }; static const ct_desc desc = { 5, XSIMD_STRING("q1fv_5"), twinstr, &GENUS, {55, 50, 45, 0}, 0, 0, 0 }; void XSIMD(codelet_q1fv_5) (planner *p) { X(kdft_difsq_register) (p, q1fv_5, &desc); } #else /* HAVE_FMA */ /* Generated by: ../../../genfft/gen_twidsq_c.native -simd -compact -variables 4 -pipeline-latency 8 -n 5 -dif -name q1fv_5 -include q1f.h */ /* * This function contains 100 FP additions, 70 FP multiplications, * (or, 85 additions, 55 multiplications, 15 fused multiply/add), * 44 stack variables, 4 constants, and 50 memory accesses */ #include "q1f.h" static void q1fv_5(R *ri, R *ii, const R *W, stride rs, stride vs, INT mb, INT me, INT ms) { DVK(KP250000000, +0.250000000000000000000000000000000000000000000); DVK(KP587785252, +0.587785252292473129168705954639072768597652438); DVK(KP951056516, +0.951056516295153572116439333379382143405698634); DVK(KP559016994, +0.559016994374947424102293417182819058860154590); { INT m; R *x; x = ri; for (m = mb, W = W + (mb * ((TWVL / VL) * 8)); m < me; m = m + VL, x = x + (VL * ms), W = W + (TWVL * 8), MAKE_VOLATILE_STRIDE(10, rs), MAKE_VOLATILE_STRIDE(10, vs)) { V T8, T7, Th, Te, T9, Ta, T1q, T1p, T1z, T1w, T1r, T1s, Ts, Tr, TB; V Ty, Tt, Tu, TM, TL, TV, TS, TN, TO, T16, T15, T1f, T1c, T17, T18; { V T6, Td, T3, Tc; T8 = LD(&(x[0]), ms, &(x[0])); { V T4, T5, T1, T2; T4 = LD(&(x[WS(rs, 2)]), ms, &(x[0])); T5 = LD(&(x[WS(rs, 3)]), ms, &(x[WS(rs, 1)])); T6 = VADD(T4, T5); Td = VSUB(T4, T5); T1 = LD(&(x[WS(rs, 1)]), ms, &(x[WS(rs, 1)])); T2 = LD(&(x[WS(rs, 4)]), ms, &(x[0])); T3 = VADD(T1, T2); Tc = VSUB(T1, T2); } T7 = VMUL(LDK(KP559016994), VSUB(T3, T6)); Th = VBYI(VFNMS(LDK(KP587785252), Tc, VMUL(LDK(KP951056516), Td))); Te = VBYI(VFMA(LDK(KP951056516), Tc, VMUL(LDK(KP587785252), Td))); T9 = VADD(T3, T6); Ta = VFNMS(LDK(KP250000000), T9, T8); } { V T1o, T1v, T1l, T1u; T1q = LD(&(x[WS(vs, 4)]), ms, &(x[WS(vs, 4)])); { V T1m, T1n, T1j, T1k; T1m = LD(&(x[WS(vs, 4) + WS(rs, 2)]), ms, &(x[WS(vs, 4)])); T1n = LD(&(x[WS(vs, 4) + WS(rs, 3)]), ms, &(x[WS(vs, 4) + WS(rs, 1)])); T1o = VADD(T1m, T1n); T1v = VSUB(T1m, T1n); T1j = LD(&(x[WS(vs, 4) + WS(rs, 1)]), ms, &(x[WS(vs, 4) + WS(rs, 1)])); T1k = LD(&(x[WS(vs, 4) + WS(rs, 4)]), ms, &(x[WS(vs, 4)])); T1l = VADD(T1j, T1k); T1u = VSUB(T1j, T1k); } T1p = VMUL(LDK(KP559016994), VSUB(T1l, T1o)); T1z = VBYI(VFNMS(LDK(KP587785252), T1u, VMUL(LDK(KP951056516), T1v))); T1w = VBYI(VFMA(LDK(KP951056516), T1u, VMUL(LDK(KP587785252), T1v))); T1r = VADD(T1l, T1o); T1s = VFNMS(LDK(KP250000000), T1r, T1q); } { V Tq, Tx, Tn, Tw; Ts = LD(&(x[WS(vs, 1)]), ms, &(x[WS(vs, 1)])); { V To, Tp, Tl, Tm; To = LD(&(x[WS(vs, 1) + WS(rs, 2)]), ms, &(x[WS(vs, 1)])); Tp = LD(&(x[WS(vs, 1) + WS(rs, 3)]), ms, &(x[WS(vs, 1) + WS(rs, 1)])); Tq = VADD(To, Tp); Tx = VSUB(To, Tp); Tl = LD(&(x[WS(vs, 1) + WS(rs, 1)]), ms, &(x[WS(vs, 1) + WS(rs, 1)])); Tm = LD(&(x[WS(vs, 1) + WS(rs, 4)]), ms, &(x[WS(vs, 1)])); Tn = VADD(Tl, Tm); Tw = VSUB(Tl, Tm); } Tr = VMUL(LDK(KP559016994), VSUB(Tn, Tq)); TB = VBYI(VFNMS(LDK(KP587785252), Tw, VMUL(LDK(KP951056516), Tx))); Ty = VBYI(VFMA(LDK(KP951056516), Tw, VMUL(LDK(KP587785252), Tx))); Tt = VADD(Tn, Tq); Tu = VFNMS(LDK(KP250000000), Tt, Ts); } { V TK, TR, TH, TQ; TM = LD(&(x[WS(vs, 2)]), ms, &(x[WS(vs, 2)])); { V TI, TJ, TF, TG; TI = LD(&(x[WS(vs, 2) + WS(rs, 2)]), ms, &(x[WS(vs, 2)])); TJ = LD(&(x[WS(vs, 2) + WS(rs, 3)]), ms, &(x[WS(vs, 2) + WS(rs, 1)])); TK = VADD(TI, TJ); TR = VSUB(TI, TJ); TF = LD(&(x[WS(vs, 2) + WS(rs, 1)]), ms, &(x[WS(vs, 2) + WS(rs, 1)])); TG = LD(&(x[WS(vs, 2) + WS(rs, 4)]), ms, &(x[WS(vs, 2)])); TH = VADD(TF, TG); TQ = VSUB(TF, TG); } TL = VMUL(LDK(KP559016994), VSUB(TH, TK)); TV = VBYI(VFNMS(LDK(KP587785252), TQ, VMUL(LDK(KP951056516), TR))); TS = VBYI(VFMA(LDK(KP951056516), TQ, VMUL(LDK(KP587785252), TR))); TN = VADD(TH, TK); TO = VFNMS(LDK(KP250000000), TN, TM); } { V T14, T1b, T11, T1a; T16 = LD(&(x[WS(vs, 3)]), ms, &(x[WS(vs, 3)])); { V T12, T13, TZ, T10; T12 = LD(&(x[WS(vs, 3) + WS(rs, 2)]), ms, &(x[WS(vs, 3)])); T13 = LD(&(x[WS(vs, 3) + WS(rs, 3)]), ms, &(x[WS(vs, 3) + WS(rs, 1)])); T14 = VADD(T12, T13); T1b = VSUB(T12, T13); TZ = LD(&(x[WS(vs, 3) + WS(rs, 1)]), ms, &(x[WS(vs, 3) + WS(rs, 1)])); T10 = LD(&(x[WS(vs, 3) + WS(rs, 4)]), ms, &(x[WS(vs, 3)])); T11 = VADD(TZ, T10); T1a = VSUB(TZ, T10); } T15 = VMUL(LDK(KP559016994), VSUB(T11, T14)); T1f = VBYI(VFNMS(LDK(KP587785252), T1a, VMUL(LDK(KP951056516), T1b))); T1c = VBYI(VFMA(LDK(KP951056516), T1a, VMUL(LDK(KP587785252), T1b))); T17 = VADD(T11, T14); T18 = VFNMS(LDK(KP250000000), T17, T16); } ST(&(x[0]), VADD(T8, T9), ms, &(x[0])); ST(&(x[WS(rs, 4)]), VADD(T1q, T1r), ms, &(x[0])); ST(&(x[WS(rs, 2)]), VADD(TM, TN), ms, &(x[0])); ST(&(x[WS(rs, 3)]), VADD(T16, T17), ms, &(x[WS(rs, 1)])); ST(&(x[WS(rs, 1)]), VADD(Ts, Tt), ms, &(x[WS(rs, 1)])); { V Tj, Tk, Ti, T1B, T1C, T1A; Ti = VSUB(Ta, T7); Tj = BYTWJ(&(W[TWVL * 2]), VADD(Th, Ti)); Tk = BYTWJ(&(W[TWVL * 4]), VSUB(Ti, Th)); ST(&(x[WS(vs, 2)]), Tj, ms, &(x[WS(vs, 2)])); ST(&(x[WS(vs, 3)]), Tk, ms, &(x[WS(vs, 3)])); T1A = VSUB(T1s, T1p); T1B = BYTWJ(&(W[TWVL * 2]), VADD(T1z, T1A)); T1C = BYTWJ(&(W[TWVL * 4]), VSUB(T1A, T1z)); ST(&(x[WS(vs, 2) + WS(rs, 4)]), T1B, ms, &(x[WS(vs, 2)])); ST(&(x[WS(vs, 3) + WS(rs, 4)]), T1C, ms, &(x[WS(vs, 3)])); } { V T1h, T1i, T1g, TD, TE, TC; T1g = VSUB(T18, T15); T1h = BYTWJ(&(W[TWVL * 2]), VADD(T1f, T1g)); T1i = BYTWJ(&(W[TWVL * 4]), VSUB(T1g, T1f)); ST(&(x[WS(vs, 2) + WS(rs, 3)]), T1h, ms, &(x[WS(vs, 2) + WS(rs, 1)])); ST(&(x[WS(vs, 3) + WS(rs, 3)]), T1i, ms, &(x[WS(vs, 3) + WS(rs, 1)])); TC = VSUB(Tu, Tr); TD = BYTWJ(&(W[TWVL * 2]), VADD(TB, TC)); TE = BYTWJ(&(W[TWVL * 4]), VSUB(TC, TB)); ST(&(x[WS(vs, 2) + WS(rs, 1)]), TD, ms, &(x[WS(vs, 2) + WS(rs, 1)])); ST(&(x[WS(vs, 3) + WS(rs, 1)]), TE, ms, &(x[WS(vs, 3) + WS(rs, 1)])); } { V TX, TY, TW, TT, TU, TP; TW = VSUB(TO, TL); TX = BYTWJ(&(W[TWVL * 2]), VADD(TV, TW)); TY = BYTWJ(&(W[TWVL * 4]), VSUB(TW, TV)); ST(&(x[WS(vs, 2) + WS(rs, 2)]), TX, ms, &(x[WS(vs, 2)])); ST(&(x[WS(vs, 3) + WS(rs, 2)]), TY, ms, &(x[WS(vs, 3)])); TP = VADD(TL, TO); TT = BYTWJ(&(W[0]), VSUB(TP, TS)); TU = BYTWJ(&(W[TWVL * 6]), VADD(TS, TP)); ST(&(x[WS(vs, 1) + WS(rs, 2)]), TT, ms, &(x[WS(vs, 1)])); ST(&(x[WS(vs, 4) + WS(rs, 2)]), TU, ms, &(x[WS(vs, 4)])); } { V Tf, Tg, Tb, Tz, TA, Tv; Tb = VADD(T7, Ta); Tf = BYTWJ(&(W[0]), VSUB(Tb, Te)); Tg = BYTWJ(&(W[TWVL * 6]), VADD(Te, Tb)); ST(&(x[WS(vs, 1)]), Tf, ms, &(x[WS(vs, 1)])); ST(&(x[WS(vs, 4)]), Tg, ms, &(x[WS(vs, 4)])); Tv = VADD(Tr, Tu); Tz = BYTWJ(&(W[0]), VSUB(Tv, Ty)); TA = BYTWJ(&(W[TWVL * 6]), VADD(Ty, Tv)); ST(&(x[WS(vs, 1) + WS(rs, 1)]), Tz, ms, &(x[WS(vs, 1) + WS(rs, 1)])); ST(&(x[WS(vs, 4) + WS(rs, 1)]), TA, ms, &(x[WS(vs, 4) + WS(rs, 1)])); } { V T1d, T1e, T19, T1x, T1y, T1t; T19 = VADD(T15, T18); T1d = BYTWJ(&(W[0]), VSUB(T19, T1c)); T1e = BYTWJ(&(W[TWVL * 6]), VADD(T1c, T19)); ST(&(x[WS(vs, 1) + WS(rs, 3)]), T1d, ms, &(x[WS(vs, 1) + WS(rs, 1)])); ST(&(x[WS(vs, 4) + WS(rs, 3)]), T1e, ms, &(x[WS(vs, 4) + WS(rs, 1)])); T1t = VADD(T1p, T1s); T1x = BYTWJ(&(W[0]), VSUB(T1t, T1w)); T1y = BYTWJ(&(W[TWVL * 6]), VADD(T1w, T1t)); ST(&(x[WS(vs, 1) + WS(rs, 4)]), T1x, ms, &(x[WS(vs, 1)])); ST(&(x[WS(vs, 4) + WS(rs, 4)]), T1y, ms, &(x[WS(vs, 4)])); } } } VLEAVE(); } static const tw_instr twinstr[] = { VTW(0, 1), VTW(0, 2), VTW(0, 3), VTW(0, 4), {TW_NEXT, VL, 0} }; static const ct_desc desc = { 5, XSIMD_STRING("q1fv_5"), twinstr, &GENUS, {85, 55, 15, 0}, 0, 0, 0 }; void XSIMD(codelet_q1fv_5) (planner *p) { X(kdft_difsq_register) (p, q1fv_5, &desc); } #endif /* HAVE_FMA */ fftw-3.3.3/dft/simd/common/q1fv_8.c0000644000175000017500000011565712054410605013674 00000000000000/* * Copyright (c) 2003, 2007-11 Matteo Frigo * Copyright (c) 2003, 2007-11 Massachusetts Institute of Technology * * This program is free software; you can redistribute it and/or modify * it under the terms of the GNU General Public License as published by * the Free Software Foundation; either version 2 of the License, or * (at your option) any later version. * * This program is distributed in the hope that it will be useful, * but WITHOUT ANY WARRANTY; without even the implied warranty of * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the * GNU General Public License for more details. * * You should have received a copy of the GNU General Public License * along with this program; if not, write to the Free Software * Foundation, Inc., 51 Franklin Street, Fifth Floor, Boston, MA 02110-1301 USA * */ /* This file was automatically generated --- DO NOT EDIT */ /* Generated on Sun Nov 25 07:39:31 EST 2012 */ #include "codelet-dft.h" #ifdef HAVE_FMA /* Generated by: ../../../genfft/gen_twidsq_c.native -fma -reorder-insns -schedule-for-pipeline -simd -compact -variables 4 -pipeline-latency 8 -n 8 -dif -name q1fv_8 -include q1f.h */ /* * This function contains 264 FP additions, 192 FP multiplications, * (or, 184 additions, 112 multiplications, 80 fused multiply/add), * 117 stack variables, 1 constants, and 128 memory accesses */ #include "q1f.h" static void q1fv_8(R *ri, R *ii, const R *W, stride rs, stride vs, INT mb, INT me, INT ms) { DVK(KP707106781, +0.707106781186547524400844362104849039284835938); { INT m; R *x; x = ri; for (m = mb, W = W + (mb * ((TWVL / VL) * 14)); m < me; m = m + VL, x = x + (VL * ms), W = W + (TWVL * 14), MAKE_VOLATILE_STRIDE(16, rs), MAKE_VOLATILE_STRIDE(16, vs)) { V T42, T43, T1U, T1V, T2Y, T2Z, TT, TS; { V T3, Te, T1E, T1P, Tu, Tp, T25, T20, T2b, T2m, T3M, T2x, T2C, T3X, TA; V TL, T48, T4d, T17, T11, TW, T1i, T2I, T1y, T1t, T2T, T3f, T3q, T34, T39; V T3G, T3B, Ts, Tv, Tf, Ta, T23, T26, T1Q, T1L, T2A, T2D, T2n, T2i, T4b; V T4e, T3Y, T3T, TZ, T12, TM, TH, T35, T2L, T3j, T1w, T1z, T1j, T1e, T36; V T2O, T3C, T3i, T3k; { V T3d, T32, T3e, T3o, T3p, T33; { V T2v, T2w, T3V, T46, T3W; { V T1, T2, Tc, Td, T1C, T1D, T1N, T1O; T1 = LD(&(x[0]), ms, &(x[0])); T2 = LD(&(x[WS(rs, 4)]), ms, &(x[0])); Tc = LD(&(x[WS(rs, 2)]), ms, &(x[0])); Td = LD(&(x[WS(rs, 6)]), ms, &(x[0])); T1C = LD(&(x[WS(vs, 3)]), ms, &(x[WS(vs, 3)])); T1D = LD(&(x[WS(vs, 3) + WS(rs, 4)]), ms, &(x[WS(vs, 3)])); T1N = LD(&(x[WS(vs, 3) + WS(rs, 2)]), ms, &(x[WS(vs, 3)])); T1O = LD(&(x[WS(vs, 3) + WS(rs, 6)]), ms, &(x[WS(vs, 3)])); { V T29, T1Y, T1Z, T2a, T2k, T2l, Tn, To, T3K, T3L; T29 = LD(&(x[WS(vs, 4)]), ms, &(x[WS(vs, 4)])); T3 = VSUB(T1, T2); Tn = VADD(T1, T2); Te = VSUB(Tc, Td); To = VADD(Tc, Td); T1E = VSUB(T1C, T1D); T1Y = VADD(T1C, T1D); T1P = VSUB(T1N, T1O); T1Z = VADD(T1N, T1O); T2a = LD(&(x[WS(vs, 4) + WS(rs, 4)]), ms, &(x[WS(vs, 4)])); T2k = LD(&(x[WS(vs, 4) + WS(rs, 2)]), ms, &(x[WS(vs, 4)])); T2l = LD(&(x[WS(vs, 4) + WS(rs, 6)]), ms, &(x[WS(vs, 4)])); Tu = VSUB(Tn, To); Tp = VADD(Tn, To); T3K = LD(&(x[WS(vs, 7)]), ms, &(x[WS(vs, 7)])); T3L = LD(&(x[WS(vs, 7) + WS(rs, 4)]), ms, &(x[WS(vs, 7)])); T25 = VSUB(T1Y, T1Z); T20 = VADD(T1Y, T1Z); T2v = VADD(T29, T2a); T2b = VSUB(T29, T2a); T2w = VADD(T2k, T2l); T2m = VSUB(T2k, T2l); T3V = LD(&(x[WS(vs, 7) + WS(rs, 2)]), ms, &(x[WS(vs, 7)])); T46 = VADD(T3K, T3L); T3M = VSUB(T3K, T3L); T3W = LD(&(x[WS(vs, 7) + WS(rs, 6)]), ms, &(x[WS(vs, 7)])); } } { V T15, TU, T16, T1g, TV, T1h; { V Ty, Tz, TJ, TK, T47; Ty = LD(&(x[WS(vs, 1)]), ms, &(x[WS(vs, 1)])); Tz = LD(&(x[WS(vs, 1) + WS(rs, 4)]), ms, &(x[WS(vs, 1)])); TJ = LD(&(x[WS(vs, 1) + WS(rs, 2)]), ms, &(x[WS(vs, 1)])); T2x = VADD(T2v, T2w); T2C = VSUB(T2v, T2w); TK = LD(&(x[WS(vs, 1) + WS(rs, 6)]), ms, &(x[WS(vs, 1)])); T47 = VADD(T3V, T3W); T3X = VSUB(T3V, T3W); T15 = LD(&(x[WS(vs, 2)]), ms, &(x[WS(vs, 2)])); TA = VSUB(Ty, Tz); TU = VADD(Ty, Tz); T16 = LD(&(x[WS(vs, 2) + WS(rs, 4)]), ms, &(x[WS(vs, 2)])); T1g = LD(&(x[WS(vs, 2) + WS(rs, 2)]), ms, &(x[WS(vs, 2)])); TL = VSUB(TJ, TK); TV = VADD(TJ, TK); T48 = VADD(T46, T47); T4d = VSUB(T46, T47); T1h = LD(&(x[WS(vs, 2) + WS(rs, 6)]), ms, &(x[WS(vs, 2)])); } { V T2G, T1r, T2H, T2R, T1s, T2S; T2G = LD(&(x[WS(vs, 5)]), ms, &(x[WS(vs, 5)])); T17 = VSUB(T15, T16); T1r = VADD(T15, T16); T2H = LD(&(x[WS(vs, 5) + WS(rs, 4)]), ms, &(x[WS(vs, 5)])); T11 = VSUB(TU, TV); TW = VADD(TU, TV); T2R = LD(&(x[WS(vs, 5) + WS(rs, 2)]), ms, &(x[WS(vs, 5)])); T1i = VSUB(T1g, T1h); T1s = VADD(T1g, T1h); T2S = LD(&(x[WS(vs, 5) + WS(rs, 6)]), ms, &(x[WS(vs, 5)])); T3d = LD(&(x[WS(vs, 6)]), ms, &(x[WS(vs, 6)])); T2I = VSUB(T2G, T2H); T32 = VADD(T2G, T2H); T3e = LD(&(x[WS(vs, 6) + WS(rs, 4)]), ms, &(x[WS(vs, 6)])); T3o = LD(&(x[WS(vs, 6) + WS(rs, 2)]), ms, &(x[WS(vs, 6)])); T3p = LD(&(x[WS(vs, 6) + WS(rs, 6)]), ms, &(x[WS(vs, 6)])); T1y = VSUB(T1r, T1s); T1t = VADD(T1r, T1s); T33 = VADD(T2R, T2S); T2T = VSUB(T2R, T2S); } } } { V T2y, T2e, T3Q, T2z, T2h, T49, T3P, T3R; { V T6, Tq, T1I, Tr, T9, T21, T1H, T1J; { V T4, T3z, T3A, T5, T7, T8, T1F, T1G; T4 = LD(&(x[WS(rs, 1)]), ms, &(x[WS(rs, 1)])); T3f = VSUB(T3d, T3e); T3z = VADD(T3d, T3e); T3q = VSUB(T3o, T3p); T3A = VADD(T3o, T3p); T5 = LD(&(x[WS(rs, 5)]), ms, &(x[WS(rs, 1)])); T7 = LD(&(x[WS(rs, 7)]), ms, &(x[WS(rs, 1)])); T34 = VADD(T32, T33); T39 = VSUB(T32, T33); T8 = LD(&(x[WS(rs, 3)]), ms, &(x[WS(rs, 1)])); T1F = LD(&(x[WS(vs, 3) + WS(rs, 1)]), ms, &(x[WS(vs, 3) + WS(rs, 1)])); T1G = LD(&(x[WS(vs, 3) + WS(rs, 5)]), ms, &(x[WS(vs, 3) + WS(rs, 1)])); T3G = VSUB(T3z, T3A); T3B = VADD(T3z, T3A); T6 = VSUB(T4, T5); Tq = VADD(T4, T5); T1I = LD(&(x[WS(vs, 3) + WS(rs, 7)]), ms, &(x[WS(vs, 3) + WS(rs, 1)])); Tr = VADD(T7, T8); T9 = VSUB(T7, T8); T21 = VADD(T1F, T1G); T1H = VSUB(T1F, T1G); T1J = LD(&(x[WS(vs, 3) + WS(rs, 3)]), ms, &(x[WS(vs, 3) + WS(rs, 1)])); } { V T2f, T22, T1K, T2g, T2c, T2d, T3N, T3O; T2c = LD(&(x[WS(vs, 4) + WS(rs, 1)]), ms, &(x[WS(vs, 4) + WS(rs, 1)])); T2d = LD(&(x[WS(vs, 4) + WS(rs, 5)]), ms, &(x[WS(vs, 4) + WS(rs, 1)])); T2f = LD(&(x[WS(vs, 4) + WS(rs, 7)]), ms, &(x[WS(vs, 4) + WS(rs, 1)])); Ts = VADD(Tq, Tr); Tv = VSUB(Tr, Tq); Tf = VSUB(T9, T6); Ta = VADD(T6, T9); T22 = VADD(T1I, T1J); T1K = VSUB(T1I, T1J); T2y = VADD(T2c, T2d); T2e = VSUB(T2c, T2d); T2g = LD(&(x[WS(vs, 4) + WS(rs, 3)]), ms, &(x[WS(vs, 4) + WS(rs, 1)])); T3N = LD(&(x[WS(vs, 7) + WS(rs, 1)]), ms, &(x[WS(vs, 7) + WS(rs, 1)])); T3O = LD(&(x[WS(vs, 7) + WS(rs, 5)]), ms, &(x[WS(vs, 7) + WS(rs, 1)])); T3Q = LD(&(x[WS(vs, 7) + WS(rs, 7)]), ms, &(x[WS(vs, 7) + WS(rs, 1)])); T23 = VADD(T21, T22); T26 = VSUB(T22, T21); T1Q = VSUB(T1K, T1H); T1L = VADD(T1H, T1K); T2z = VADD(T2f, T2g); T2h = VSUB(T2f, T2g); T49 = VADD(T3N, T3O); T3P = VSUB(T3N, T3O); T3R = LD(&(x[WS(vs, 7) + WS(rs, 3)]), ms, &(x[WS(vs, 7) + WS(rs, 1)])); } } { V TX, TD, T1b, TY, TG, T1u, T1a, T1c; { V TE, T4a, T3S, TF, TB, TC, T18, T19; TB = LD(&(x[WS(vs, 1) + WS(rs, 1)]), ms, &(x[WS(vs, 1) + WS(rs, 1)])); TC = LD(&(x[WS(vs, 1) + WS(rs, 5)]), ms, &(x[WS(vs, 1) + WS(rs, 1)])); TE = LD(&(x[WS(vs, 1) + WS(rs, 7)]), ms, &(x[WS(vs, 1) + WS(rs, 1)])); T2A = VADD(T2y, T2z); T2D = VSUB(T2z, T2y); T2n = VSUB(T2h, T2e); T2i = VADD(T2e, T2h); T4a = VADD(T3Q, T3R); T3S = VSUB(T3Q, T3R); TX = VADD(TB, TC); TD = VSUB(TB, TC); TF = LD(&(x[WS(vs, 1) + WS(rs, 3)]), ms, &(x[WS(vs, 1) + WS(rs, 1)])); T18 = LD(&(x[WS(vs, 2) + WS(rs, 1)]), ms, &(x[WS(vs, 2) + WS(rs, 1)])); T19 = LD(&(x[WS(vs, 2) + WS(rs, 5)]), ms, &(x[WS(vs, 2) + WS(rs, 1)])); T1b = LD(&(x[WS(vs, 2) + WS(rs, 7)]), ms, &(x[WS(vs, 2) + WS(rs, 1)])); T4b = VADD(T49, T4a); T4e = VSUB(T4a, T49); T3Y = VSUB(T3S, T3P); T3T = VADD(T3P, T3S); TY = VADD(TE, TF); TG = VSUB(TE, TF); T1u = VADD(T18, T19); T1a = VSUB(T18, T19); T1c = LD(&(x[WS(vs, 2) + WS(rs, 3)]), ms, &(x[WS(vs, 2) + WS(rs, 1)])); } { V T2M, T1v, T1d, T2N, T2J, T2K, T3g, T3h; T2J = LD(&(x[WS(vs, 5) + WS(rs, 1)]), ms, &(x[WS(vs, 5) + WS(rs, 1)])); T2K = LD(&(x[WS(vs, 5) + WS(rs, 5)]), ms, &(x[WS(vs, 5) + WS(rs, 1)])); T2M = LD(&(x[WS(vs, 5) + WS(rs, 7)]), ms, &(x[WS(vs, 5) + WS(rs, 1)])); TZ = VADD(TX, TY); T12 = VSUB(TY, TX); TM = VSUB(TG, TD); TH = VADD(TD, TG); T1v = VADD(T1b, T1c); T1d = VSUB(T1b, T1c); T35 = VADD(T2J, T2K); T2L = VSUB(T2J, T2K); T2N = LD(&(x[WS(vs, 5) + WS(rs, 3)]), ms, &(x[WS(vs, 5) + WS(rs, 1)])); T3g = LD(&(x[WS(vs, 6) + WS(rs, 1)]), ms, &(x[WS(vs, 6) + WS(rs, 1)])); T3h = LD(&(x[WS(vs, 6) + WS(rs, 5)]), ms, &(x[WS(vs, 6) + WS(rs, 1)])); T3j = LD(&(x[WS(vs, 6) + WS(rs, 7)]), ms, &(x[WS(vs, 6) + WS(rs, 1)])); T1w = VADD(T1u, T1v); T1z = VSUB(T1v, T1u); T1j = VSUB(T1d, T1a); T1e = VADD(T1a, T1d); T36 = VADD(T2M, T2N); T2O = VSUB(T2M, T2N); T3C = VADD(T3g, T3h); T3i = VSUB(T3g, T3h); T3k = LD(&(x[WS(vs, 6) + WS(rs, 3)]), ms, &(x[WS(vs, 6) + WS(rs, 1)])); } } } } { V T3a, T2U, T2P, T3H, T3r, T3m, T13, T27, T3b, T4f; { V T37, T3E, T2B, T24; { V T3D, T3l, Tt, T4c; ST(&(x[0]), VADD(Tp, Ts), ms, &(x[0])); ST(&(x[WS(rs, 2)]), VADD(T1t, T1w), ms, &(x[0])); ST(&(x[WS(rs, 7)]), VADD(T48, T4b), ms, &(x[WS(rs, 1)])); T37 = VADD(T35, T36); T3a = VSUB(T36, T35); T2U = VSUB(T2O, T2L); T2P = VADD(T2L, T2O); T3D = VADD(T3j, T3k); T3l = VSUB(T3j, T3k); ST(&(x[WS(rs, 4)]), VADD(T2x, T2A), ms, &(x[0])); ST(&(x[WS(rs, 3)]), VADD(T20, T23), ms, &(x[WS(rs, 1)])); ST(&(x[WS(rs, 5)]), VADD(T34, T37), ms, &(x[WS(rs, 1)])); ST(&(x[WS(rs, 1)]), VADD(TW, TZ), ms, &(x[WS(rs, 1)])); Tt = BYTWJ(&(W[TWVL * 6]), VSUB(Tp, Ts)); T4c = BYTWJ(&(W[TWVL * 6]), VSUB(T48, T4b)); T3E = VADD(T3C, T3D); T3H = VSUB(T3D, T3C); T3r = VSUB(T3l, T3i); T3m = VADD(T3i, T3l); T2B = BYTWJ(&(W[TWVL * 6]), VSUB(T2x, T2A)); T24 = BYTWJ(&(W[TWVL * 6]), VSUB(T20, T23)); ST(&(x[WS(vs, 4)]), Tt, ms, &(x[WS(vs, 4)])); ST(&(x[WS(vs, 4) + WS(rs, 7)]), T4c, ms, &(x[WS(vs, 4) + WS(rs, 1)])); ST(&(x[WS(rs, 6)]), VADD(T3B, T3E), ms, &(x[0])); } { V T38, T1A, Tw, T10, T1x, T3F, T2E, T3I; T10 = BYTWJ(&(W[TWVL * 6]), VSUB(TW, TZ)); T1x = BYTWJ(&(W[TWVL * 6]), VSUB(T1t, T1w)); T3F = BYTWJ(&(W[TWVL * 6]), VSUB(T3B, T3E)); ST(&(x[WS(vs, 4) + WS(rs, 4)]), T2B, ms, &(x[WS(vs, 4)])); ST(&(x[WS(vs, 4) + WS(rs, 3)]), T24, ms, &(x[WS(vs, 4) + WS(rs, 1)])); T38 = BYTWJ(&(W[TWVL * 6]), VSUB(T34, T37)); T1A = BYTWJ(&(W[TWVL * 10]), VFNMSI(T1z, T1y)); Tw = BYTWJ(&(W[TWVL * 10]), VFNMSI(Tv, Tu)); ST(&(x[WS(vs, 4) + WS(rs, 1)]), T10, ms, &(x[WS(vs, 4) + WS(rs, 1)])); ST(&(x[WS(vs, 4) + WS(rs, 2)]), T1x, ms, &(x[WS(vs, 4)])); ST(&(x[WS(vs, 4) + WS(rs, 6)]), T3F, ms, &(x[WS(vs, 4)])); T2E = BYTWJ(&(W[TWVL * 10]), VFNMSI(T2D, T2C)); T3I = BYTWJ(&(W[TWVL * 10]), VFNMSI(T3H, T3G)); ST(&(x[WS(vs, 4) + WS(rs, 5)]), T38, ms, &(x[WS(vs, 4) + WS(rs, 1)])); ST(&(x[WS(vs, 6) + WS(rs, 2)]), T1A, ms, &(x[WS(vs, 6)])); ST(&(x[WS(vs, 6)]), Tw, ms, &(x[WS(vs, 6)])); T13 = BYTWJ(&(W[TWVL * 10]), VFNMSI(T12, T11)); T27 = BYTWJ(&(W[TWVL * 10]), VFNMSI(T26, T25)); T3b = BYTWJ(&(W[TWVL * 10]), VFNMSI(T3a, T39)); ST(&(x[WS(vs, 6) + WS(rs, 4)]), T2E, ms, &(x[WS(vs, 6)])); ST(&(x[WS(vs, 6) + WS(rs, 6)]), T3I, ms, &(x[WS(vs, 6)])); T4f = BYTWJ(&(W[TWVL * 10]), VFNMSI(T4e, T4d)); } } { V Tj, Tk, T2r, T2j, Ti, Th, T2o, T2s, T1M, T1R, T41, T40; { V T3c, T4g, T3J, T2F, Tx, T1B; Tx = BYTWJ(&(W[TWVL * 2]), VFMAI(Tv, Tu)); T1B = BYTWJ(&(W[TWVL * 2]), VFMAI(T1z, T1y)); ST(&(x[WS(vs, 6) + WS(rs, 1)]), T13, ms, &(x[WS(vs, 6) + WS(rs, 1)])); ST(&(x[WS(vs, 6) + WS(rs, 3)]), T27, ms, &(x[WS(vs, 6) + WS(rs, 1)])); ST(&(x[WS(vs, 6) + WS(rs, 5)]), T3b, ms, &(x[WS(vs, 6) + WS(rs, 1)])); T3c = BYTWJ(&(W[TWVL * 2]), VFMAI(T3a, T39)); T4g = BYTWJ(&(W[TWVL * 2]), VFMAI(T4e, T4d)); ST(&(x[WS(vs, 6) + WS(rs, 7)]), T4f, ms, &(x[WS(vs, 6) + WS(rs, 1)])); ST(&(x[WS(vs, 2)]), Tx, ms, &(x[WS(vs, 2)])); ST(&(x[WS(vs, 2) + WS(rs, 2)]), T1B, ms, &(x[WS(vs, 2)])); T3J = BYTWJ(&(W[TWVL * 2]), VFMAI(T3H, T3G)); T2F = BYTWJ(&(W[TWVL * 2]), VFMAI(T2D, T2C)); { V T14, Tb, Tg, T28, T3U, T3Z; T28 = BYTWJ(&(W[TWVL * 2]), VFMAI(T26, T25)); ST(&(x[WS(vs, 2) + WS(rs, 5)]), T3c, ms, &(x[WS(vs, 2) + WS(rs, 1)])); ST(&(x[WS(vs, 2) + WS(rs, 7)]), T4g, ms, &(x[WS(vs, 2) + WS(rs, 1)])); T14 = BYTWJ(&(W[TWVL * 2]), VFMAI(T12, T11)); Tj = VFNMS(LDK(KP707106781), Ta, T3); Tb = VFMA(LDK(KP707106781), Ta, T3); Tg = VFNMS(LDK(KP707106781), Tf, Te); Tk = VFMA(LDK(KP707106781), Tf, Te); ST(&(x[WS(vs, 2) + WS(rs, 6)]), T3J, ms, &(x[WS(vs, 2)])); ST(&(x[WS(vs, 2) + WS(rs, 4)]), T2F, ms, &(x[WS(vs, 2)])); ST(&(x[WS(vs, 2) + WS(rs, 3)]), T28, ms, &(x[WS(vs, 2) + WS(rs, 1)])); T3U = VFMA(LDK(KP707106781), T3T, T3M); T42 = VFNMS(LDK(KP707106781), T3T, T3M); T43 = VFMA(LDK(KP707106781), T3Y, T3X); T3Z = VFNMS(LDK(KP707106781), T3Y, T3X); ST(&(x[WS(vs, 2) + WS(rs, 1)]), T14, ms, &(x[WS(vs, 2) + WS(rs, 1)])); T2r = VFNMS(LDK(KP707106781), T2i, T2b); T2j = VFMA(LDK(KP707106781), T2i, T2b); Ti = BYTWJ(&(W[TWVL * 12]), VFMAI(Tg, Tb)); Th = BYTWJ(&(W[0]), VFNMSI(Tg, Tb)); T2o = VFNMS(LDK(KP707106781), T2n, T2m); T2s = VFMA(LDK(KP707106781), T2n, T2m); T1U = VFNMS(LDK(KP707106781), T1L, T1E); T1M = VFMA(LDK(KP707106781), T1L, T1E); T1R = VFNMS(LDK(KP707106781), T1Q, T1P); T1V = VFMA(LDK(KP707106781), T1Q, T1P); T41 = BYTWJ(&(W[TWVL * 12]), VFMAI(T3Z, T3U)); T40 = BYTWJ(&(W[0]), VFNMSI(T3Z, T3U)); } } { V TQ, TR, T1n, T1o, T3v, T3w; { V T1f, T1k, T3n, TP, TO, T3s, T2Q, T2V; { V TI, T2q, T2p, T1T, T1S, TN; TQ = VFNMS(LDK(KP707106781), TH, TA); TI = VFMA(LDK(KP707106781), TH, TA); ST(&(x[WS(vs, 7)]), Ti, ms, &(x[WS(vs, 7)])); ST(&(x[WS(vs, 1)]), Th, ms, &(x[WS(vs, 1)])); T2q = BYTWJ(&(W[TWVL * 12]), VFMAI(T2o, T2j)); T2p = BYTWJ(&(W[0]), VFNMSI(T2o, T2j)); T1T = BYTWJ(&(W[TWVL * 12]), VFMAI(T1R, T1M)); T1S = BYTWJ(&(W[0]), VFNMSI(T1R, T1M)); ST(&(x[WS(vs, 7) + WS(rs, 7)]), T41, ms, &(x[WS(vs, 7) + WS(rs, 1)])); ST(&(x[WS(vs, 1) + WS(rs, 7)]), T40, ms, &(x[WS(vs, 1) + WS(rs, 1)])); TN = VFNMS(LDK(KP707106781), TM, TL); TR = VFMA(LDK(KP707106781), TM, TL); T1n = VFNMS(LDK(KP707106781), T1e, T17); T1f = VFMA(LDK(KP707106781), T1e, T17); ST(&(x[WS(vs, 7) + WS(rs, 4)]), T2q, ms, &(x[WS(vs, 7)])); ST(&(x[WS(vs, 1) + WS(rs, 4)]), T2p, ms, &(x[WS(vs, 1)])); ST(&(x[WS(vs, 7) + WS(rs, 3)]), T1T, ms, &(x[WS(vs, 7) + WS(rs, 1)])); ST(&(x[WS(vs, 1) + WS(rs, 3)]), T1S, ms, &(x[WS(vs, 1) + WS(rs, 1)])); T1k = VFNMS(LDK(KP707106781), T1j, T1i); T1o = VFMA(LDK(KP707106781), T1j, T1i); T3v = VFNMS(LDK(KP707106781), T3m, T3f); T3n = VFMA(LDK(KP707106781), T3m, T3f); TP = BYTWJ(&(W[TWVL * 12]), VFMAI(TN, TI)); TO = BYTWJ(&(W[0]), VFNMSI(TN, TI)); T3s = VFNMS(LDK(KP707106781), T3r, T3q); T3w = VFMA(LDK(KP707106781), T3r, T3q); } T2Y = VFNMS(LDK(KP707106781), T2P, T2I); T2Q = VFMA(LDK(KP707106781), T2P, T2I); T2V = VFNMS(LDK(KP707106781), T2U, T2T); T2Z = VFMA(LDK(KP707106781), T2U, T2T); { V T3u, T3t, T2X, T2W, T1m, T1l; T1m = BYTWJ(&(W[TWVL * 12]), VFMAI(T1k, T1f)); T1l = BYTWJ(&(W[0]), VFNMSI(T1k, T1f)); ST(&(x[WS(vs, 7) + WS(rs, 1)]), TP, ms, &(x[WS(vs, 7) + WS(rs, 1)])); ST(&(x[WS(vs, 1) + WS(rs, 1)]), TO, ms, &(x[WS(vs, 1) + WS(rs, 1)])); T3u = BYTWJ(&(W[TWVL * 12]), VFMAI(T3s, T3n)); T3t = BYTWJ(&(W[0]), VFNMSI(T3s, T3n)); T2X = BYTWJ(&(W[TWVL * 12]), VFMAI(T2V, T2Q)); T2W = BYTWJ(&(W[0]), VFNMSI(T2V, T2Q)); ST(&(x[WS(vs, 7) + WS(rs, 2)]), T1m, ms, &(x[WS(vs, 7)])); ST(&(x[WS(vs, 1) + WS(rs, 2)]), T1l, ms, &(x[WS(vs, 1)])); ST(&(x[WS(vs, 7) + WS(rs, 6)]), T3u, ms, &(x[WS(vs, 7)])); ST(&(x[WS(vs, 1) + WS(rs, 6)]), T3t, ms, &(x[WS(vs, 1)])); ST(&(x[WS(vs, 7) + WS(rs, 5)]), T2X, ms, &(x[WS(vs, 7) + WS(rs, 1)])); ST(&(x[WS(vs, 1) + WS(rs, 5)]), T2W, ms, &(x[WS(vs, 1) + WS(rs, 1)])); } } { V T2u, T2t, T3y, T3x; { V T1q, T1p, Tm, Tl; T1q = BYTWJ(&(W[TWVL * 4]), VFMAI(T1o, T1n)); T1p = BYTWJ(&(W[TWVL * 8]), VFNMSI(T1o, T1n)); Tm = BYTWJ(&(W[TWVL * 4]), VFMAI(Tk, Tj)); Tl = BYTWJ(&(W[TWVL * 8]), VFNMSI(Tk, Tj)); ST(&(x[WS(vs, 3) + WS(rs, 2)]), T1q, ms, &(x[WS(vs, 3)])); ST(&(x[WS(vs, 5) + WS(rs, 2)]), T1p, ms, &(x[WS(vs, 5)])); T2u = BYTWJ(&(W[TWVL * 4]), VFMAI(T2s, T2r)); T2t = BYTWJ(&(W[TWVL * 8]), VFNMSI(T2s, T2r)); T3y = BYTWJ(&(W[TWVL * 4]), VFMAI(T3w, T3v)); T3x = BYTWJ(&(W[TWVL * 8]), VFNMSI(T3w, T3v)); ST(&(x[WS(vs, 3)]), Tm, ms, &(x[WS(vs, 3)])); ST(&(x[WS(vs, 5)]), Tl, ms, &(x[WS(vs, 5)])); } ST(&(x[WS(vs, 3) + WS(rs, 4)]), T2u, ms, &(x[WS(vs, 3)])); ST(&(x[WS(vs, 5) + WS(rs, 4)]), T2t, ms, &(x[WS(vs, 5)])); ST(&(x[WS(vs, 3) + WS(rs, 6)]), T3y, ms, &(x[WS(vs, 3)])); ST(&(x[WS(vs, 5) + WS(rs, 6)]), T3x, ms, &(x[WS(vs, 5)])); TT = BYTWJ(&(W[TWVL * 4]), VFMAI(TR, TQ)); TS = BYTWJ(&(W[TWVL * 8]), VFNMSI(TR, TQ)); } } } } } { V T31, T30, T45, T44, T1X, T1W; T1X = BYTWJ(&(W[TWVL * 4]), VFMAI(T1V, T1U)); T1W = BYTWJ(&(W[TWVL * 8]), VFNMSI(T1V, T1U)); ST(&(x[WS(vs, 3) + WS(rs, 1)]), TT, ms, &(x[WS(vs, 3) + WS(rs, 1)])); ST(&(x[WS(vs, 5) + WS(rs, 1)]), TS, ms, &(x[WS(vs, 5) + WS(rs, 1)])); T31 = BYTWJ(&(W[TWVL * 4]), VFMAI(T2Z, T2Y)); T30 = BYTWJ(&(W[TWVL * 8]), VFNMSI(T2Z, T2Y)); T45 = BYTWJ(&(W[TWVL * 4]), VFMAI(T43, T42)); T44 = BYTWJ(&(W[TWVL * 8]), VFNMSI(T43, T42)); ST(&(x[WS(vs, 3) + WS(rs, 3)]), T1X, ms, &(x[WS(vs, 3) + WS(rs, 1)])); ST(&(x[WS(vs, 5) + WS(rs, 3)]), T1W, ms, &(x[WS(vs, 5) + WS(rs, 1)])); ST(&(x[WS(vs, 3) + WS(rs, 5)]), T31, ms, &(x[WS(vs, 3) + WS(rs, 1)])); ST(&(x[WS(vs, 5) + WS(rs, 5)]), T30, ms, &(x[WS(vs, 5) + WS(rs, 1)])); ST(&(x[WS(vs, 3) + WS(rs, 7)]), T45, ms, &(x[WS(vs, 3) + WS(rs, 1)])); ST(&(x[WS(vs, 5) + WS(rs, 7)]), T44, ms, &(x[WS(vs, 5) + WS(rs, 1)])); } } } VLEAVE(); } static const tw_instr twinstr[] = { VTW(0, 1), VTW(0, 2), VTW(0, 3), VTW(0, 4), VTW(0, 5), VTW(0, 6), VTW(0, 7), {TW_NEXT, VL, 0} }; static const ct_desc desc = { 8, XSIMD_STRING("q1fv_8"), twinstr, &GENUS, {184, 112, 80, 0}, 0, 0, 0 }; void XSIMD(codelet_q1fv_8) (planner *p) { X(kdft_difsq_register) (p, q1fv_8, &desc); } #else /* HAVE_FMA */ /* Generated by: ../../../genfft/gen_twidsq_c.native -simd -compact -variables 4 -pipeline-latency 8 -n 8 -dif -name q1fv_8 -include q1f.h */ /* * This function contains 264 FP additions, 128 FP multiplications, * (or, 264 additions, 128 multiplications, 0 fused multiply/add), * 77 stack variables, 1 constants, and 128 memory accesses */ #include "q1f.h" static void q1fv_8(R *ri, R *ii, const R *W, stride rs, stride vs, INT mb, INT me, INT ms) { DVK(KP707106781, +0.707106781186547524400844362104849039284835938); { INT m; R *x; x = ri; for (m = mb, W = W + (mb * ((TWVL / VL) * 14)); m < me; m = m + VL, x = x + (VL * ms), W = W + (TWVL * 14), MAKE_VOLATILE_STRIDE(16, rs), MAKE_VOLATILE_STRIDE(16, vs)) { V T3, Tu, Tf, Tp, T1E, T25, T1Q, T20, T2b, T2C, T2n, T2x, T3M, T4d, T3Y; V T48, TA, T11, TM, TW, T17, T1y, T1j, T1t, T2I, T39, T2U, T34, T3f, T3G; V T3r, T3B, Ta, Tv, Tc, Ts, T1L, T26, T1N, T23, T2i, T2D, T2k, T2A, T3T; V T4e, T3V, T4b, TH, T12, TJ, TZ, T1e, T1z, T1g, T1w, T2P, T3a, T2R, T37; V T3m, T3H, T3o, T3E, T28, T14; { V T1, T2, Tn, Td, Te, To; T1 = LD(&(x[0]), ms, &(x[0])); T2 = LD(&(x[WS(rs, 4)]), ms, &(x[0])); Tn = VADD(T1, T2); Td = LD(&(x[WS(rs, 2)]), ms, &(x[0])); Te = LD(&(x[WS(rs, 6)]), ms, &(x[0])); To = VADD(Td, Te); T3 = VSUB(T1, T2); Tu = VSUB(Tn, To); Tf = VSUB(Td, Te); Tp = VADD(Tn, To); } { V T1C, T1D, T1Y, T1O, T1P, T1Z; T1C = LD(&(x[WS(vs, 3)]), ms, &(x[WS(vs, 3)])); T1D = LD(&(x[WS(vs, 3) + WS(rs, 4)]), ms, &(x[WS(vs, 3)])); T1Y = VADD(T1C, T1D); T1O = LD(&(x[WS(vs, 3) + WS(rs, 2)]), ms, &(x[WS(vs, 3)])); T1P = LD(&(x[WS(vs, 3) + WS(rs, 6)]), ms, &(x[WS(vs, 3)])); T1Z = VADD(T1O, T1P); T1E = VSUB(T1C, T1D); T25 = VSUB(T1Y, T1Z); T1Q = VSUB(T1O, T1P); T20 = VADD(T1Y, T1Z); } { V T29, T2a, T2v, T2l, T2m, T2w; T29 = LD(&(x[WS(vs, 4)]), ms, &(x[WS(vs, 4)])); T2a = LD(&(x[WS(vs, 4) + WS(rs, 4)]), ms, &(x[WS(vs, 4)])); T2v = VADD(T29, T2a); T2l = LD(&(x[WS(vs, 4) + WS(rs, 2)]), ms, &(x[WS(vs, 4)])); T2m = LD(&(x[WS(vs, 4) + WS(rs, 6)]), ms, &(x[WS(vs, 4)])); T2w = VADD(T2l, T2m); T2b = VSUB(T29, T2a); T2C = VSUB(T2v, T2w); T2n = VSUB(T2l, T2m); T2x = VADD(T2v, T2w); } { V T3K, T3L, T46, T3W, T3X, T47; T3K = LD(&(x[WS(vs, 7)]), ms, &(x[WS(vs, 7)])); T3L = LD(&(x[WS(vs, 7) + WS(rs, 4)]), ms, &(x[WS(vs, 7)])); T46 = VADD(T3K, T3L); T3W = LD(&(x[WS(vs, 7) + WS(rs, 2)]), ms, &(x[WS(vs, 7)])); T3X = LD(&(x[WS(vs, 7) + WS(rs, 6)]), ms, &(x[WS(vs, 7)])); T47 = VADD(T3W, T3X); T3M = VSUB(T3K, T3L); T4d = VSUB(T46, T47); T3Y = VSUB(T3W, T3X); T48 = VADD(T46, T47); } { V Ty, Tz, TU, TK, TL, TV; Ty = LD(&(x[WS(vs, 1)]), ms, &(x[WS(vs, 1)])); Tz = LD(&(x[WS(vs, 1) + WS(rs, 4)]), ms, &(x[WS(vs, 1)])); TU = VADD(Ty, Tz); TK = LD(&(x[WS(vs, 1) + WS(rs, 2)]), ms, &(x[WS(vs, 1)])); TL = LD(&(x[WS(vs, 1) + WS(rs, 6)]), ms, &(x[WS(vs, 1)])); TV = VADD(TK, TL); TA = VSUB(Ty, Tz); T11 = VSUB(TU, TV); TM = VSUB(TK, TL); TW = VADD(TU, TV); } { V T15, T16, T1r, T1h, T1i, T1s; T15 = LD(&(x[WS(vs, 2)]), ms, &(x[WS(vs, 2)])); T16 = LD(&(x[WS(vs, 2) + WS(rs, 4)]), ms, &(x[WS(vs, 2)])); T1r = VADD(T15, T16); T1h = LD(&(x[WS(vs, 2) + WS(rs, 2)]), ms, &(x[WS(vs, 2)])); T1i = LD(&(x[WS(vs, 2) + WS(rs, 6)]), ms, &(x[WS(vs, 2)])); T1s = VADD(T1h, T1i); T17 = VSUB(T15, T16); T1y = VSUB(T1r, T1s); T1j = VSUB(T1h, T1i); T1t = VADD(T1r, T1s); } { V T2G, T2H, T32, T2S, T2T, T33; T2G = LD(&(x[WS(vs, 5)]), ms, &(x[WS(vs, 5)])); T2H = LD(&(x[WS(vs, 5) + WS(rs, 4)]), ms, &(x[WS(vs, 5)])); T32 = VADD(T2G, T2H); T2S = LD(&(x[WS(vs, 5) + WS(rs, 2)]), ms, &(x[WS(vs, 5)])); T2T = LD(&(x[WS(vs, 5) + WS(rs, 6)]), ms, &(x[WS(vs, 5)])); T33 = VADD(T2S, T2T); T2I = VSUB(T2G, T2H); T39 = VSUB(T32, T33); T2U = VSUB(T2S, T2T); T34 = VADD(T32, T33); } { V T3d, T3e, T3z, T3p, T3q, T3A; T3d = LD(&(x[WS(vs, 6)]), ms, &(x[WS(vs, 6)])); T3e = LD(&(x[WS(vs, 6) + WS(rs, 4)]), ms, &(x[WS(vs, 6)])); T3z = VADD(T3d, T3e); T3p = LD(&(x[WS(vs, 6) + WS(rs, 2)]), ms, &(x[WS(vs, 6)])); T3q = LD(&(x[WS(vs, 6) + WS(rs, 6)]), ms, &(x[WS(vs, 6)])); T3A = VADD(T3p, T3q); T3f = VSUB(T3d, T3e); T3G = VSUB(T3z, T3A); T3r = VSUB(T3p, T3q); T3B = VADD(T3z, T3A); } { V T6, Tq, T9, Tr; { V T4, T5, T7, T8; T4 = LD(&(x[WS(rs, 1)]), ms, &(x[WS(rs, 1)])); T5 = LD(&(x[WS(rs, 5)]), ms, &(x[WS(rs, 1)])); T6 = VSUB(T4, T5); Tq = VADD(T4, T5); T7 = LD(&(x[WS(rs, 7)]), ms, &(x[WS(rs, 1)])); T8 = LD(&(x[WS(rs, 3)]), ms, &(x[WS(rs, 1)])); T9 = VSUB(T7, T8); Tr = VADD(T7, T8); } Ta = VMUL(LDK(KP707106781), VADD(T6, T9)); Tv = VBYI(VSUB(Tr, Tq)); Tc = VMUL(LDK(KP707106781), VSUB(T9, T6)); Ts = VADD(Tq, Tr); } { V T1H, T21, T1K, T22; { V T1F, T1G, T1I, T1J; T1F = LD(&(x[WS(vs, 3) + WS(rs, 1)]), ms, &(x[WS(vs, 3) + WS(rs, 1)])); T1G = LD(&(x[WS(vs, 3) + WS(rs, 5)]), ms, &(x[WS(vs, 3) + WS(rs, 1)])); T1H = VSUB(T1F, T1G); T21 = VADD(T1F, T1G); T1I = LD(&(x[WS(vs, 3) + WS(rs, 7)]), ms, &(x[WS(vs, 3) + WS(rs, 1)])); T1J = LD(&(x[WS(vs, 3) + WS(rs, 3)]), ms, &(x[WS(vs, 3) + WS(rs, 1)])); T1K = VSUB(T1I, T1J); T22 = VADD(T1I, T1J); } T1L = VMUL(LDK(KP707106781), VADD(T1H, T1K)); T26 = VBYI(VSUB(T22, T21)); T1N = VMUL(LDK(KP707106781), VSUB(T1K, T1H)); T23 = VADD(T21, T22); } { V T2e, T2y, T2h, T2z; { V T2c, T2d, T2f, T2g; T2c = LD(&(x[WS(vs, 4) + WS(rs, 1)]), ms, &(x[WS(vs, 4) + WS(rs, 1)])); T2d = LD(&(x[WS(vs, 4) + WS(rs, 5)]), ms, &(x[WS(vs, 4) + WS(rs, 1)])); T2e = VSUB(T2c, T2d); T2y = VADD(T2c, T2d); T2f = LD(&(x[WS(vs, 4) + WS(rs, 7)]), ms, &(x[WS(vs, 4) + WS(rs, 1)])); T2g = LD(&(x[WS(vs, 4) + WS(rs, 3)]), ms, &(x[WS(vs, 4) + WS(rs, 1)])); T2h = VSUB(T2f, T2g); T2z = VADD(T2f, T2g); } T2i = VMUL(LDK(KP707106781), VADD(T2e, T2h)); T2D = VBYI(VSUB(T2z, T2y)); T2k = VMUL(LDK(KP707106781), VSUB(T2h, T2e)); T2A = VADD(T2y, T2z); } { V T3P, T49, T3S, T4a; { V T3N, T3O, T3Q, T3R; T3N = LD(&(x[WS(vs, 7) + WS(rs, 1)]), ms, &(x[WS(vs, 7) + WS(rs, 1)])); T3O = LD(&(x[WS(vs, 7) + WS(rs, 5)]), ms, &(x[WS(vs, 7) + WS(rs, 1)])); T3P = VSUB(T3N, T3O); T49 = VADD(T3N, T3O); T3Q = LD(&(x[WS(vs, 7) + WS(rs, 7)]), ms, &(x[WS(vs, 7) + WS(rs, 1)])); T3R = LD(&(x[WS(vs, 7) + WS(rs, 3)]), ms, &(x[WS(vs, 7) + WS(rs, 1)])); T3S = VSUB(T3Q, T3R); T4a = VADD(T3Q, T3R); } T3T = VMUL(LDK(KP707106781), VADD(T3P, T3S)); T4e = VBYI(VSUB(T4a, T49)); T3V = VMUL(LDK(KP707106781), VSUB(T3S, T3P)); T4b = VADD(T49, T4a); } { V TD, TX, TG, TY; { V TB, TC, TE, TF; TB = LD(&(x[WS(vs, 1) + WS(rs, 1)]), ms, &(x[WS(vs, 1) + WS(rs, 1)])); TC = LD(&(x[WS(vs, 1) + WS(rs, 5)]), ms, &(x[WS(vs, 1) + WS(rs, 1)])); TD = VSUB(TB, TC); TX = VADD(TB, TC); TE = LD(&(x[WS(vs, 1) + WS(rs, 7)]), ms, &(x[WS(vs, 1) + WS(rs, 1)])); TF = LD(&(x[WS(vs, 1) + WS(rs, 3)]), ms, &(x[WS(vs, 1) + WS(rs, 1)])); TG = VSUB(TE, TF); TY = VADD(TE, TF); } TH = VMUL(LDK(KP707106781), VADD(TD, TG)); T12 = VBYI(VSUB(TY, TX)); TJ = VMUL(LDK(KP707106781), VSUB(TG, TD)); TZ = VADD(TX, TY); } { V T1a, T1u, T1d, T1v; { V T18, T19, T1b, T1c; T18 = LD(&(x[WS(vs, 2) + WS(rs, 1)]), ms, &(x[WS(vs, 2) + WS(rs, 1)])); T19 = LD(&(x[WS(vs, 2) + WS(rs, 5)]), ms, &(x[WS(vs, 2) + WS(rs, 1)])); T1a = VSUB(T18, T19); T1u = VADD(T18, T19); T1b = LD(&(x[WS(vs, 2) + WS(rs, 7)]), ms, &(x[WS(vs, 2) + WS(rs, 1)])); T1c = LD(&(x[WS(vs, 2) + WS(rs, 3)]), ms, &(x[WS(vs, 2) + WS(rs, 1)])); T1d = VSUB(T1b, T1c); T1v = VADD(T1b, T1c); } T1e = VMUL(LDK(KP707106781), VADD(T1a, T1d)); T1z = VBYI(VSUB(T1v, T1u)); T1g = VMUL(LDK(KP707106781), VSUB(T1d, T1a)); T1w = VADD(T1u, T1v); } { V T2L, T35, T2O, T36; { V T2J, T2K, T2M, T2N; T2J = LD(&(x[WS(vs, 5) + WS(rs, 1)]), ms, &(x[WS(vs, 5) + WS(rs, 1)])); T2K = LD(&(x[WS(vs, 5) + WS(rs, 5)]), ms, &(x[WS(vs, 5) + WS(rs, 1)])); T2L = VSUB(T2J, T2K); T35 = VADD(T2J, T2K); T2M = LD(&(x[WS(vs, 5) + WS(rs, 7)]), ms, &(x[WS(vs, 5) + WS(rs, 1)])); T2N = LD(&(x[WS(vs, 5) + WS(rs, 3)]), ms, &(x[WS(vs, 5) + WS(rs, 1)])); T2O = VSUB(T2M, T2N); T36 = VADD(T2M, T2N); } T2P = VMUL(LDK(KP707106781), VADD(T2L, T2O)); T3a = VBYI(VSUB(T36, T35)); T2R = VMUL(LDK(KP707106781), VSUB(T2O, T2L)); T37 = VADD(T35, T36); } { V T3i, T3C, T3l, T3D; { V T3g, T3h, T3j, T3k; T3g = LD(&(x[WS(vs, 6) + WS(rs, 1)]), ms, &(x[WS(vs, 6) + WS(rs, 1)])); T3h = LD(&(x[WS(vs, 6) + WS(rs, 5)]), ms, &(x[WS(vs, 6) + WS(rs, 1)])); T3i = VSUB(T3g, T3h); T3C = VADD(T3g, T3h); T3j = LD(&(x[WS(vs, 6) + WS(rs, 7)]), ms, &(x[WS(vs, 6) + WS(rs, 1)])); T3k = LD(&(x[WS(vs, 6) + WS(rs, 3)]), ms, &(x[WS(vs, 6) + WS(rs, 1)])); T3l = VSUB(T3j, T3k); T3D = VADD(T3j, T3k); } T3m = VMUL(LDK(KP707106781), VADD(T3i, T3l)); T3H = VBYI(VSUB(T3D, T3C)); T3o = VMUL(LDK(KP707106781), VSUB(T3l, T3i)); T3E = VADD(T3C, T3D); } ST(&(x[0]), VADD(Tp, Ts), ms, &(x[0])); ST(&(x[WS(rs, 2)]), VADD(T1t, T1w), ms, &(x[0])); ST(&(x[WS(rs, 5)]), VADD(T34, T37), ms, &(x[WS(rs, 1)])); ST(&(x[WS(rs, 7)]), VADD(T48, T4b), ms, &(x[WS(rs, 1)])); ST(&(x[WS(rs, 6)]), VADD(T3B, T3E), ms, &(x[0])); ST(&(x[WS(rs, 4)]), VADD(T2x, T2A), ms, &(x[0])); { V Tt, T4c, T2B, T24; ST(&(x[WS(rs, 3)]), VADD(T20, T23), ms, &(x[WS(rs, 1)])); ST(&(x[WS(rs, 1)]), VADD(TW, TZ), ms, &(x[WS(rs, 1)])); Tt = BYTWJ(&(W[TWVL * 6]), VSUB(Tp, Ts)); ST(&(x[WS(vs, 4)]), Tt, ms, &(x[WS(vs, 4)])); T4c = BYTWJ(&(W[TWVL * 6]), VSUB(T48, T4b)); ST(&(x[WS(vs, 4) + WS(rs, 7)]), T4c, ms, &(x[WS(vs, 4) + WS(rs, 1)])); T2B = BYTWJ(&(W[TWVL * 6]), VSUB(T2x, T2A)); ST(&(x[WS(vs, 4) + WS(rs, 4)]), T2B, ms, &(x[WS(vs, 4)])); T24 = BYTWJ(&(W[TWVL * 6]), VSUB(T20, T23)); ST(&(x[WS(vs, 4) + WS(rs, 3)]), T24, ms, &(x[WS(vs, 4) + WS(rs, 1)])); } { V T10, T1x, T3F, T38, T1A, Tw; T10 = BYTWJ(&(W[TWVL * 6]), VSUB(TW, TZ)); ST(&(x[WS(vs, 4) + WS(rs, 1)]), T10, ms, &(x[WS(vs, 4) + WS(rs, 1)])); T1x = BYTWJ(&(W[TWVL * 6]), VSUB(T1t, T1w)); ST(&(x[WS(vs, 4) + WS(rs, 2)]), T1x, ms, &(x[WS(vs, 4)])); T3F = BYTWJ(&(W[TWVL * 6]), VSUB(T3B, T3E)); ST(&(x[WS(vs, 4) + WS(rs, 6)]), T3F, ms, &(x[WS(vs, 4)])); T38 = BYTWJ(&(W[TWVL * 6]), VSUB(T34, T37)); ST(&(x[WS(vs, 4) + WS(rs, 5)]), T38, ms, &(x[WS(vs, 4) + WS(rs, 1)])); T1A = BYTWJ(&(W[TWVL * 10]), VSUB(T1y, T1z)); ST(&(x[WS(vs, 6) + WS(rs, 2)]), T1A, ms, &(x[WS(vs, 6)])); Tw = BYTWJ(&(W[TWVL * 10]), VSUB(Tu, Tv)); ST(&(x[WS(vs, 6)]), Tw, ms, &(x[WS(vs, 6)])); } { V T2E, T3I, T13, T27, T3b, T4f; T2E = BYTWJ(&(W[TWVL * 10]), VSUB(T2C, T2D)); ST(&(x[WS(vs, 6) + WS(rs, 4)]), T2E, ms, &(x[WS(vs, 6)])); T3I = BYTWJ(&(W[TWVL * 10]), VSUB(T3G, T3H)); ST(&(x[WS(vs, 6) + WS(rs, 6)]), T3I, ms, &(x[WS(vs, 6)])); T13 = BYTWJ(&(W[TWVL * 10]), VSUB(T11, T12)); ST(&(x[WS(vs, 6) + WS(rs, 1)]), T13, ms, &(x[WS(vs, 6) + WS(rs, 1)])); T27 = BYTWJ(&(W[TWVL * 10]), VSUB(T25, T26)); ST(&(x[WS(vs, 6) + WS(rs, 3)]), T27, ms, &(x[WS(vs, 6) + WS(rs, 1)])); T3b = BYTWJ(&(W[TWVL * 10]), VSUB(T39, T3a)); ST(&(x[WS(vs, 6) + WS(rs, 5)]), T3b, ms, &(x[WS(vs, 6) + WS(rs, 1)])); T4f = BYTWJ(&(W[TWVL * 10]), VSUB(T4d, T4e)); ST(&(x[WS(vs, 6) + WS(rs, 7)]), T4f, ms, &(x[WS(vs, 6) + WS(rs, 1)])); } { V Tx, T1B, T3c, T4g, T3J, T2F; Tx = BYTWJ(&(W[TWVL * 2]), VADD(Tu, Tv)); ST(&(x[WS(vs, 2)]), Tx, ms, &(x[WS(vs, 2)])); T1B = BYTWJ(&(W[TWVL * 2]), VADD(T1y, T1z)); ST(&(x[WS(vs, 2) + WS(rs, 2)]), T1B, ms, &(x[WS(vs, 2)])); T3c = BYTWJ(&(W[TWVL * 2]), VADD(T39, T3a)); ST(&(x[WS(vs, 2) + WS(rs, 5)]), T3c, ms, &(x[WS(vs, 2) + WS(rs, 1)])); T4g = BYTWJ(&(W[TWVL * 2]), VADD(T4d, T4e)); ST(&(x[WS(vs, 2) + WS(rs, 7)]), T4g, ms, &(x[WS(vs, 2) + WS(rs, 1)])); T3J = BYTWJ(&(W[TWVL * 2]), VADD(T3G, T3H)); ST(&(x[WS(vs, 2) + WS(rs, 6)]), T3J, ms, &(x[WS(vs, 2)])); T2F = BYTWJ(&(W[TWVL * 2]), VADD(T2C, T2D)); ST(&(x[WS(vs, 2) + WS(rs, 4)]), T2F, ms, &(x[WS(vs, 2)])); } T28 = BYTWJ(&(W[TWVL * 2]), VADD(T25, T26)); ST(&(x[WS(vs, 2) + WS(rs, 3)]), T28, ms, &(x[WS(vs, 2) + WS(rs, 1)])); T14 = BYTWJ(&(W[TWVL * 2]), VADD(T11, T12)); ST(&(x[WS(vs, 2) + WS(rs, 1)]), T14, ms, &(x[WS(vs, 2) + WS(rs, 1)])); { V Th, Ti, Tb, Tg; Tb = VADD(T3, Ta); Tg = VBYI(VSUB(Tc, Tf)); Th = BYTWJ(&(W[TWVL * 12]), VSUB(Tb, Tg)); Ti = BYTWJ(&(W[0]), VADD(Tb, Tg)); ST(&(x[WS(vs, 7)]), Th, ms, &(x[WS(vs, 7)])); ST(&(x[WS(vs, 1)]), Ti, ms, &(x[WS(vs, 1)])); } { V T40, T41, T3U, T3Z; T3U = VADD(T3M, T3T); T3Z = VBYI(VSUB(T3V, T3Y)); T40 = BYTWJ(&(W[TWVL * 12]), VSUB(T3U, T3Z)); T41 = BYTWJ(&(W[0]), VADD(T3U, T3Z)); ST(&(x[WS(vs, 7) + WS(rs, 7)]), T40, ms, &(x[WS(vs, 7) + WS(rs, 1)])); ST(&(x[WS(vs, 1) + WS(rs, 7)]), T41, ms, &(x[WS(vs, 1) + WS(rs, 1)])); } { V T2p, T2q, T2j, T2o; T2j = VADD(T2b, T2i); T2o = VBYI(VSUB(T2k, T2n)); T2p = BYTWJ(&(W[TWVL * 12]), VSUB(T2j, T2o)); T2q = BYTWJ(&(W[0]), VADD(T2j, T2o)); ST(&(x[WS(vs, 7) + WS(rs, 4)]), T2p, ms, &(x[WS(vs, 7)])); ST(&(x[WS(vs, 1) + WS(rs, 4)]), T2q, ms, &(x[WS(vs, 1)])); } { V T1S, T1T, T1M, T1R; T1M = VADD(T1E, T1L); T1R = VBYI(VSUB(T1N, T1Q)); T1S = BYTWJ(&(W[TWVL * 12]), VSUB(T1M, T1R)); T1T = BYTWJ(&(W[0]), VADD(T1M, T1R)); ST(&(x[WS(vs, 7) + WS(rs, 3)]), T1S, ms, &(x[WS(vs, 7) + WS(rs, 1)])); ST(&(x[WS(vs, 1) + WS(rs, 3)]), T1T, ms, &(x[WS(vs, 1) + WS(rs, 1)])); } { V TO, TP, TI, TN; TI = VADD(TA, TH); TN = VBYI(VSUB(TJ, TM)); TO = BYTWJ(&(W[TWVL * 12]), VSUB(TI, TN)); TP = BYTWJ(&(W[0]), VADD(TI, TN)); ST(&(x[WS(vs, 7) + WS(rs, 1)]), TO, ms, &(x[WS(vs, 7) + WS(rs, 1)])); ST(&(x[WS(vs, 1) + WS(rs, 1)]), TP, ms, &(x[WS(vs, 1) + WS(rs, 1)])); } { V T1l, T1m, T1f, T1k; T1f = VADD(T17, T1e); T1k = VBYI(VSUB(T1g, T1j)); T1l = BYTWJ(&(W[TWVL * 12]), VSUB(T1f, T1k)); T1m = BYTWJ(&(W[0]), VADD(T1f, T1k)); ST(&(x[WS(vs, 7) + WS(rs, 2)]), T1l, ms, &(x[WS(vs, 7)])); ST(&(x[WS(vs, 1) + WS(rs, 2)]), T1m, ms, &(x[WS(vs, 1)])); } { V T3t, T3u, T3n, T3s; T3n = VADD(T3f, T3m); T3s = VBYI(VSUB(T3o, T3r)); T3t = BYTWJ(&(W[TWVL * 12]), VSUB(T3n, T3s)); T3u = BYTWJ(&(W[0]), VADD(T3n, T3s)); ST(&(x[WS(vs, 7) + WS(rs, 6)]), T3t, ms, &(x[WS(vs, 7)])); ST(&(x[WS(vs, 1) + WS(rs, 6)]), T3u, ms, &(x[WS(vs, 1)])); } { V T2W, T2X, T2Q, T2V; T2Q = VADD(T2I, T2P); T2V = VBYI(VSUB(T2R, T2U)); T2W = BYTWJ(&(W[TWVL * 12]), VSUB(T2Q, T2V)); T2X = BYTWJ(&(W[0]), VADD(T2Q, T2V)); ST(&(x[WS(vs, 7) + WS(rs, 5)]), T2W, ms, &(x[WS(vs, 7) + WS(rs, 1)])); ST(&(x[WS(vs, 1) + WS(rs, 5)]), T2X, ms, &(x[WS(vs, 1) + WS(rs, 1)])); } { V T1p, T1q, T1n, T1o; T1n = VSUB(T17, T1e); T1o = VBYI(VADD(T1j, T1g)); T1p = BYTWJ(&(W[TWVL * 8]), VSUB(T1n, T1o)); T1q = BYTWJ(&(W[TWVL * 4]), VADD(T1n, T1o)); ST(&(x[WS(vs, 5) + WS(rs, 2)]), T1p, ms, &(x[WS(vs, 5)])); ST(&(x[WS(vs, 3) + WS(rs, 2)]), T1q, ms, &(x[WS(vs, 3)])); } { V Tl, Tm, Tj, Tk; Tj = VSUB(T3, Ta); Tk = VBYI(VADD(Tf, Tc)); Tl = BYTWJ(&(W[TWVL * 8]), VSUB(Tj, Tk)); Tm = BYTWJ(&(W[TWVL * 4]), VADD(Tj, Tk)); ST(&(x[WS(vs, 5)]), Tl, ms, &(x[WS(vs, 5)])); ST(&(x[WS(vs, 3)]), Tm, ms, &(x[WS(vs, 3)])); } { V T2t, T2u, T2r, T2s; T2r = VSUB(T2b, T2i); T2s = VBYI(VADD(T2n, T2k)); T2t = BYTWJ(&(W[TWVL * 8]), VSUB(T2r, T2s)); T2u = BYTWJ(&(W[TWVL * 4]), VADD(T2r, T2s)); ST(&(x[WS(vs, 5) + WS(rs, 4)]), T2t, ms, &(x[WS(vs, 5)])); ST(&(x[WS(vs, 3) + WS(rs, 4)]), T2u, ms, &(x[WS(vs, 3)])); } { V T3x, T3y, T3v, T3w; T3v = VSUB(T3f, T3m); T3w = VBYI(VADD(T3r, T3o)); T3x = BYTWJ(&(W[TWVL * 8]), VSUB(T3v, T3w)); T3y = BYTWJ(&(W[TWVL * 4]), VADD(T3v, T3w)); ST(&(x[WS(vs, 5) + WS(rs, 6)]), T3x, ms, &(x[WS(vs, 5)])); ST(&(x[WS(vs, 3) + WS(rs, 6)]), T3y, ms, &(x[WS(vs, 3)])); } { V TS, TT, TQ, TR; TQ = VSUB(TA, TH); TR = VBYI(VADD(TM, TJ)); TS = BYTWJ(&(W[TWVL * 8]), VSUB(TQ, TR)); TT = BYTWJ(&(W[TWVL * 4]), VADD(TQ, TR)); ST(&(x[WS(vs, 5) + WS(rs, 1)]), TS, ms, &(x[WS(vs, 5) + WS(rs, 1)])); ST(&(x[WS(vs, 3) + WS(rs, 1)]), TT, ms, &(x[WS(vs, 3) + WS(rs, 1)])); } { V T1W, T1X, T1U, T1V; T1U = VSUB(T1E, T1L); T1V = VBYI(VADD(T1Q, T1N)); T1W = BYTWJ(&(W[TWVL * 8]), VSUB(T1U, T1V)); T1X = BYTWJ(&(W[TWVL * 4]), VADD(T1U, T1V)); ST(&(x[WS(vs, 5) + WS(rs, 3)]), T1W, ms, &(x[WS(vs, 5) + WS(rs, 1)])); ST(&(x[WS(vs, 3) + WS(rs, 3)]), T1X, ms, &(x[WS(vs, 3) + WS(rs, 1)])); } { V T30, T31, T2Y, T2Z; T2Y = VSUB(T2I, T2P); T2Z = VBYI(VADD(T2U, T2R)); T30 = BYTWJ(&(W[TWVL * 8]), VSUB(T2Y, T2Z)); T31 = BYTWJ(&(W[TWVL * 4]), VADD(T2Y, T2Z)); ST(&(x[WS(vs, 5) + WS(rs, 5)]), T30, ms, &(x[WS(vs, 5) + WS(rs, 1)])); ST(&(x[WS(vs, 3) + WS(rs, 5)]), T31, ms, &(x[WS(vs, 3) + WS(rs, 1)])); } { V T44, T45, T42, T43; T42 = VSUB(T3M, T3T); T43 = VBYI(VADD(T3Y, T3V)); T44 = BYTWJ(&(W[TWVL * 8]), VSUB(T42, T43)); T45 = BYTWJ(&(W[TWVL * 4]), VADD(T42, T43)); ST(&(x[WS(vs, 5) + WS(rs, 7)]), T44, ms, &(x[WS(vs, 5) + WS(rs, 1)])); ST(&(x[WS(vs, 3) + WS(rs, 7)]), T45, ms, &(x[WS(vs, 3) + WS(rs, 1)])); } } } VLEAVE(); } static const tw_instr twinstr[] = { VTW(0, 1), VTW(0, 2), VTW(0, 3), VTW(0, 4), VTW(0, 5), VTW(0, 6), VTW(0, 7), {TW_NEXT, VL, 0} }; static const ct_desc desc = { 8, XSIMD_STRING("q1fv_8"), twinstr, &GENUS, {264, 128, 0, 0}, 0, 0, 0 }; void XSIMD(codelet_q1fv_8) (planner *p) { X(kdft_difsq_register) (p, q1fv_8, &desc); } #endif /* HAVE_FMA */ fftw-3.3.3/dft/simd/common/q1bv_2.c0000644000175000017500000001013412054410605013642 00000000000000/* * Copyright (c) 2003, 2007-11 Matteo Frigo * Copyright (c) 2003, 2007-11 Massachusetts Institute of Technology * * This program is free software; you can redistribute it and/or modify * it under the terms of the GNU General Public License as published by * the Free Software Foundation; either version 2 of the License, or * (at your option) any later version. * * This program is distributed in the hope that it will be useful, * but WITHOUT ANY WARRANTY; without even the implied warranty of * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the * GNU General Public License for more details. * * You should have received a copy of the GNU General Public License * along with this program; if not, write to the Free Software * Foundation, Inc., 51 Franklin Street, Fifth Floor, Boston, MA 02110-1301 USA * */ /* This file was automatically generated --- DO NOT EDIT */ /* Generated on Sun Nov 25 07:39:33 EST 2012 */ #include "codelet-dft.h" #ifdef HAVE_FMA /* Generated by: ../../../genfft/gen_twidsq_c.native -fma -reorder-insns -schedule-for-pipeline -simd -compact -variables 4 -pipeline-latency 8 -n 2 -dif -name q1bv_2 -include q1b.h -sign 1 */ /* * This function contains 6 FP additions, 4 FP multiplications, * (or, 6 additions, 4 multiplications, 0 fused multiply/add), * 8 stack variables, 0 constants, and 8 memory accesses */ #include "q1b.h" static void q1bv_2(R *ri, R *ii, const R *W, stride rs, stride vs, INT mb, INT me, INT ms) { { INT m; R *x; x = ii; for (m = mb, W = W + (mb * ((TWVL / VL) * 2)); m < me; m = m + VL, x = x + (VL * ms), W = W + (TWVL * 2), MAKE_VOLATILE_STRIDE(4, rs), MAKE_VOLATILE_STRIDE(4, vs)) { V T1, T2, T4, T5, T3, T6; T1 = LD(&(x[0]), ms, &(x[0])); T2 = LD(&(x[WS(rs, 1)]), ms, &(x[WS(rs, 1)])); T4 = LD(&(x[WS(vs, 1)]), ms, &(x[WS(vs, 1)])); T5 = LD(&(x[WS(vs, 1) + WS(rs, 1)]), ms, &(x[WS(vs, 1) + WS(rs, 1)])); ST(&(x[0]), VADD(T1, T2), ms, &(x[0])); T3 = BYTW(&(W[0]), VSUB(T1, T2)); ST(&(x[WS(rs, 1)]), VADD(T4, T5), ms, &(x[WS(rs, 1)])); T6 = BYTW(&(W[0]), VSUB(T4, T5)); ST(&(x[WS(vs, 1)]), T3, ms, &(x[WS(vs, 1)])); ST(&(x[WS(vs, 1) + WS(rs, 1)]), T6, ms, &(x[WS(vs, 1) + WS(rs, 1)])); } } VLEAVE(); } static const tw_instr twinstr[] = { VTW(0, 1), {TW_NEXT, VL, 0} }; static const ct_desc desc = { 2, XSIMD_STRING("q1bv_2"), twinstr, &GENUS, {6, 4, 0, 0}, 0, 0, 0 }; void XSIMD(codelet_q1bv_2) (planner *p) { X(kdft_difsq_register) (p, q1bv_2, &desc); } #else /* HAVE_FMA */ /* Generated by: ../../../genfft/gen_twidsq_c.native -simd -compact -variables 4 -pipeline-latency 8 -n 2 -dif -name q1bv_2 -include q1b.h -sign 1 */ /* * This function contains 6 FP additions, 4 FP multiplications, * (or, 6 additions, 4 multiplications, 0 fused multiply/add), * 8 stack variables, 0 constants, and 8 memory accesses */ #include "q1b.h" static void q1bv_2(R *ri, R *ii, const R *W, stride rs, stride vs, INT mb, INT me, INT ms) { { INT m; R *x; x = ii; for (m = mb, W = W + (mb * ((TWVL / VL) * 2)); m < me; m = m + VL, x = x + (VL * ms), W = W + (TWVL * 2), MAKE_VOLATILE_STRIDE(4, rs), MAKE_VOLATILE_STRIDE(4, vs)) { V T1, T2, T3, T4, T5, T6; T1 = LD(&(x[0]), ms, &(x[0])); T2 = LD(&(x[WS(rs, 1)]), ms, &(x[WS(rs, 1)])); T3 = BYTW(&(W[0]), VSUB(T1, T2)); T4 = LD(&(x[WS(vs, 1)]), ms, &(x[WS(vs, 1)])); T5 = LD(&(x[WS(vs, 1) + WS(rs, 1)]), ms, &(x[WS(vs, 1) + WS(rs, 1)])); T6 = BYTW(&(W[0]), VSUB(T4, T5)); ST(&(x[WS(vs, 1)]), T3, ms, &(x[WS(vs, 1)])); ST(&(x[WS(vs, 1) + WS(rs, 1)]), T6, ms, &(x[WS(vs, 1) + WS(rs, 1)])); ST(&(x[0]), VADD(T1, T2), ms, &(x[0])); ST(&(x[WS(rs, 1)]), VADD(T4, T5), ms, &(x[WS(rs, 1)])); } } VLEAVE(); } static const tw_instr twinstr[] = { VTW(0, 1), {TW_NEXT, VL, 0} }; static const ct_desc desc = { 2, XSIMD_STRING("q1bv_2"), twinstr, &GENUS, {6, 4, 0, 0}, 0, 0, 0 }; void XSIMD(codelet_q1bv_2) (planner *p) { X(kdft_difsq_register) (p, q1bv_2, &desc); } #endif /* HAVE_FMA */ fftw-3.3.3/dft/simd/common/q1bv_4.c0000644000175000017500000002276312054410605013657 00000000000000/* * Copyright (c) 2003, 2007-11 Matteo Frigo * Copyright (c) 2003, 2007-11 Massachusetts Institute of Technology * * This program is free software; you can redistribute it and/or modify * it under the terms of the GNU General Public License as published by * the Free Software Foundation; either version 2 of the License, or * (at your option) any later version. * * This program is distributed in the hope that it will be useful, * but WITHOUT ANY WARRANTY; without even the implied warranty of * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the * GNU General Public License for more details. * * You should have received a copy of the GNU General Public License * along with this program; if not, write to the Free Software * Foundation, Inc., 51 Franklin Street, Fifth Floor, Boston, MA 02110-1301 USA * */ /* This file was automatically generated --- DO NOT EDIT */ /* Generated on Sun Nov 25 07:39:33 EST 2012 */ #include "codelet-dft.h" #ifdef HAVE_FMA /* Generated by: ../../../genfft/gen_twidsq_c.native -fma -reorder-insns -schedule-for-pipeline -simd -compact -variables 4 -pipeline-latency 8 -n 4 -dif -name q1bv_4 -include q1b.h -sign 1 */ /* * This function contains 44 FP additions, 32 FP multiplications, * (or, 36 additions, 24 multiplications, 8 fused multiply/add), * 38 stack variables, 0 constants, and 32 memory accesses */ #include "q1b.h" static void q1bv_4(R *ri, R *ii, const R *W, stride rs, stride vs, INT mb, INT me, INT ms) { { INT m; R *x; x = ii; for (m = mb, W = W + (mb * ((TWVL / VL) * 6)); m < me; m = m + VL, x = x + (VL * ms), W = W + (TWVL * 6), MAKE_VOLATILE_STRIDE(8, rs), MAKE_VOLATILE_STRIDE(8, vs)) { V Tb, Tm, Tx, TI; { V Tc, T9, T3, TG, TA, TH, TD, Ta, T6, Td, Tn, To, Tq, Tr, Tf; V Tg; { V T1, T2, Ty, Tz, TB, TC, T4, T5; T1 = LD(&(x[0]), ms, &(x[0])); T2 = LD(&(x[WS(rs, 2)]), ms, &(x[0])); Ty = LD(&(x[WS(vs, 3)]), ms, &(x[WS(vs, 3)])); Tz = LD(&(x[WS(vs, 3) + WS(rs, 2)]), ms, &(x[WS(vs, 3)])); TB = LD(&(x[WS(vs, 3) + WS(rs, 1)]), ms, &(x[WS(vs, 3) + WS(rs, 1)])); TC = LD(&(x[WS(vs, 3) + WS(rs, 3)]), ms, &(x[WS(vs, 3) + WS(rs, 1)])); T4 = LD(&(x[WS(rs, 1)]), ms, &(x[WS(rs, 1)])); T5 = LD(&(x[WS(rs, 3)]), ms, &(x[WS(rs, 1)])); Tc = LD(&(x[WS(vs, 1)]), ms, &(x[WS(vs, 1)])); T9 = VADD(T1, T2); T3 = VSUB(T1, T2); TG = VADD(Ty, Tz); TA = VSUB(Ty, Tz); TH = VADD(TB, TC); TD = VSUB(TB, TC); Ta = VADD(T4, T5); T6 = VSUB(T4, T5); Td = LD(&(x[WS(vs, 1) + WS(rs, 2)]), ms, &(x[WS(vs, 1)])); Tn = LD(&(x[WS(vs, 2)]), ms, &(x[WS(vs, 2)])); To = LD(&(x[WS(vs, 2) + WS(rs, 2)]), ms, &(x[WS(vs, 2)])); Tq = LD(&(x[WS(vs, 2) + WS(rs, 1)]), ms, &(x[WS(vs, 2) + WS(rs, 1)])); Tr = LD(&(x[WS(vs, 2) + WS(rs, 3)]), ms, &(x[WS(vs, 2) + WS(rs, 1)])); Tf = LD(&(x[WS(vs, 1) + WS(rs, 1)]), ms, &(x[WS(vs, 1) + WS(rs, 1)])); Tg = LD(&(x[WS(vs, 1) + WS(rs, 3)]), ms, &(x[WS(vs, 1) + WS(rs, 1)])); } { V Tk, Te, Tv, Tp, Tw, Ts, Tl, Th, T7, TE, Tu, TF; ST(&(x[0]), VADD(T9, Ta), ms, &(x[0])); Tk = VADD(Tc, Td); Te = VSUB(Tc, Td); Tv = VADD(Tn, To); Tp = VSUB(Tn, To); Tw = VADD(Tq, Tr); Ts = VSUB(Tq, Tr); Tl = VADD(Tf, Tg); Th = VSUB(Tf, Tg); ST(&(x[WS(rs, 3)]), VADD(TG, TH), ms, &(x[WS(rs, 1)])); T7 = BYTW(&(W[TWVL * 4]), VFNMSI(T6, T3)); TE = BYTW(&(W[TWVL * 4]), VFNMSI(TD, TA)); { V Tt, Ti, Tj, T8; T8 = BYTW(&(W[0]), VFMAI(T6, T3)); ST(&(x[WS(rs, 2)]), VADD(Tv, Tw), ms, &(x[0])); Tt = BYTW(&(W[TWVL * 4]), VFNMSI(Ts, Tp)); ST(&(x[WS(rs, 1)]), VADD(Tk, Tl), ms, &(x[WS(rs, 1)])); Ti = BYTW(&(W[TWVL * 4]), VFNMSI(Th, Te)); Tj = BYTW(&(W[0]), VFMAI(Th, Te)); ST(&(x[WS(vs, 3)]), T7, ms, &(x[WS(vs, 3)])); ST(&(x[WS(vs, 3) + WS(rs, 3)]), TE, ms, &(x[WS(vs, 3) + WS(rs, 1)])); ST(&(x[WS(vs, 1)]), T8, ms, &(x[WS(vs, 1)])); Tu = BYTW(&(W[0]), VFMAI(Ts, Tp)); ST(&(x[WS(vs, 3) + WS(rs, 2)]), Tt, ms, &(x[WS(vs, 3)])); TF = BYTW(&(W[0]), VFMAI(TD, TA)); ST(&(x[WS(vs, 3) + WS(rs, 1)]), Ti, ms, &(x[WS(vs, 3) + WS(rs, 1)])); ST(&(x[WS(vs, 1) + WS(rs, 1)]), Tj, ms, &(x[WS(vs, 1) + WS(rs, 1)])); } Tb = BYTW(&(W[TWVL * 2]), VSUB(T9, Ta)); Tm = BYTW(&(W[TWVL * 2]), VSUB(Tk, Tl)); Tx = BYTW(&(W[TWVL * 2]), VSUB(Tv, Tw)); ST(&(x[WS(vs, 1) + WS(rs, 2)]), Tu, ms, &(x[WS(vs, 1)])); TI = BYTW(&(W[TWVL * 2]), VSUB(TG, TH)); ST(&(x[WS(vs, 1) + WS(rs, 3)]), TF, ms, &(x[WS(vs, 1) + WS(rs, 1)])); } } ST(&(x[WS(vs, 2)]), Tb, ms, &(x[WS(vs, 2)])); ST(&(x[WS(vs, 2) + WS(rs, 1)]), Tm, ms, &(x[WS(vs, 2) + WS(rs, 1)])); ST(&(x[WS(vs, 2) + WS(rs, 2)]), Tx, ms, &(x[WS(vs, 2)])); ST(&(x[WS(vs, 2) + WS(rs, 3)]), TI, ms, &(x[WS(vs, 2) + WS(rs, 1)])); } } VLEAVE(); } static const tw_instr twinstr[] = { VTW(0, 1), VTW(0, 2), VTW(0, 3), {TW_NEXT, VL, 0} }; static const ct_desc desc = { 4, XSIMD_STRING("q1bv_4"), twinstr, &GENUS, {36, 24, 8, 0}, 0, 0, 0 }; void XSIMD(codelet_q1bv_4) (planner *p) { X(kdft_difsq_register) (p, q1bv_4, &desc); } #else /* HAVE_FMA */ /* Generated by: ../../../genfft/gen_twidsq_c.native -simd -compact -variables 4 -pipeline-latency 8 -n 4 -dif -name q1bv_4 -include q1b.h -sign 1 */ /* * This function contains 44 FP additions, 24 FP multiplications, * (or, 44 additions, 24 multiplications, 0 fused multiply/add), * 22 stack variables, 0 constants, and 32 memory accesses */ #include "q1b.h" static void q1bv_4(R *ri, R *ii, const R *W, stride rs, stride vs, INT mb, INT me, INT ms) { { INT m; R *x; x = ii; for (m = mb, W = W + (mb * ((TWVL / VL) * 6)); m < me; m = m + VL, x = x + (VL * ms), W = W + (TWVL * 6), MAKE_VOLATILE_STRIDE(8, rs), MAKE_VOLATILE_STRIDE(8, vs)) { V T3, T9, TA, TG, TD, TH, T6, Ta, Te, Tk, Tp, Tv, Ts, Tw, Th; V Tl; { V T1, T2, Ty, Tz; T1 = LD(&(x[0]), ms, &(x[0])); T2 = LD(&(x[WS(rs, 2)]), ms, &(x[0])); T3 = VSUB(T1, T2); T9 = VADD(T1, T2); Ty = LD(&(x[WS(vs, 3)]), ms, &(x[WS(vs, 3)])); Tz = LD(&(x[WS(vs, 3) + WS(rs, 2)]), ms, &(x[WS(vs, 3)])); TA = VSUB(Ty, Tz); TG = VADD(Ty, Tz); } { V TB, TC, T4, T5; TB = LD(&(x[WS(vs, 3) + WS(rs, 1)]), ms, &(x[WS(vs, 3) + WS(rs, 1)])); TC = LD(&(x[WS(vs, 3) + WS(rs, 3)]), ms, &(x[WS(vs, 3) + WS(rs, 1)])); TD = VBYI(VSUB(TB, TC)); TH = VADD(TB, TC); T4 = LD(&(x[WS(rs, 1)]), ms, &(x[WS(rs, 1)])); T5 = LD(&(x[WS(rs, 3)]), ms, &(x[WS(rs, 1)])); T6 = VBYI(VSUB(T4, T5)); Ta = VADD(T4, T5); } { V Tc, Td, Tn, To; Tc = LD(&(x[WS(vs, 1)]), ms, &(x[WS(vs, 1)])); Td = LD(&(x[WS(vs, 1) + WS(rs, 2)]), ms, &(x[WS(vs, 1)])); Te = VSUB(Tc, Td); Tk = VADD(Tc, Td); Tn = LD(&(x[WS(vs, 2)]), ms, &(x[WS(vs, 2)])); To = LD(&(x[WS(vs, 2) + WS(rs, 2)]), ms, &(x[WS(vs, 2)])); Tp = VSUB(Tn, To); Tv = VADD(Tn, To); } { V Tq, Tr, Tf, Tg; Tq = LD(&(x[WS(vs, 2) + WS(rs, 1)]), ms, &(x[WS(vs, 2) + WS(rs, 1)])); Tr = LD(&(x[WS(vs, 2) + WS(rs, 3)]), ms, &(x[WS(vs, 2) + WS(rs, 1)])); Ts = VBYI(VSUB(Tq, Tr)); Tw = VADD(Tq, Tr); Tf = LD(&(x[WS(vs, 1) + WS(rs, 1)]), ms, &(x[WS(vs, 1) + WS(rs, 1)])); Tg = LD(&(x[WS(vs, 1) + WS(rs, 3)]), ms, &(x[WS(vs, 1) + WS(rs, 1)])); Th = VBYI(VSUB(Tf, Tg)); Tl = VADD(Tf, Tg); } ST(&(x[0]), VADD(T9, Ta), ms, &(x[0])); ST(&(x[WS(rs, 1)]), VADD(Tk, Tl), ms, &(x[WS(rs, 1)])); ST(&(x[WS(rs, 2)]), VADD(Tv, Tw), ms, &(x[0])); ST(&(x[WS(rs, 3)]), VADD(TG, TH), ms, &(x[WS(rs, 1)])); { V T7, Ti, Tt, TE; T7 = BYTW(&(W[TWVL * 4]), VSUB(T3, T6)); ST(&(x[WS(vs, 3)]), T7, ms, &(x[WS(vs, 3)])); Ti = BYTW(&(W[TWVL * 4]), VSUB(Te, Th)); ST(&(x[WS(vs, 3) + WS(rs, 1)]), Ti, ms, &(x[WS(vs, 3) + WS(rs, 1)])); Tt = BYTW(&(W[TWVL * 4]), VSUB(Tp, Ts)); ST(&(x[WS(vs, 3) + WS(rs, 2)]), Tt, ms, &(x[WS(vs, 3)])); TE = BYTW(&(W[TWVL * 4]), VSUB(TA, TD)); ST(&(x[WS(vs, 3) + WS(rs, 3)]), TE, ms, &(x[WS(vs, 3) + WS(rs, 1)])); } { V T8, Tj, Tu, TF; T8 = BYTW(&(W[0]), VADD(T3, T6)); ST(&(x[WS(vs, 1)]), T8, ms, &(x[WS(vs, 1)])); Tj = BYTW(&(W[0]), VADD(Te, Th)); ST(&(x[WS(vs, 1) + WS(rs, 1)]), Tj, ms, &(x[WS(vs, 1) + WS(rs, 1)])); Tu = BYTW(&(W[0]), VADD(Tp, Ts)); ST(&(x[WS(vs, 1) + WS(rs, 2)]), Tu, ms, &(x[WS(vs, 1)])); TF = BYTW(&(W[0]), VADD(TA, TD)); ST(&(x[WS(vs, 1) + WS(rs, 3)]), TF, ms, &(x[WS(vs, 1) + WS(rs, 1)])); } { V Tb, Tm, Tx, TI; Tb = BYTW(&(W[TWVL * 2]), VSUB(T9, Ta)); ST(&(x[WS(vs, 2)]), Tb, ms, &(x[WS(vs, 2)])); Tm = BYTW(&(W[TWVL * 2]), VSUB(Tk, Tl)); ST(&(x[WS(vs, 2) + WS(rs, 1)]), Tm, ms, &(x[WS(vs, 2) + WS(rs, 1)])); Tx = BYTW(&(W[TWVL * 2]), VSUB(Tv, Tw)); ST(&(x[WS(vs, 2) + WS(rs, 2)]), Tx, ms, &(x[WS(vs, 2)])); TI = BYTW(&(W[TWVL * 2]), VSUB(TG, TH)); ST(&(x[WS(vs, 2) + WS(rs, 3)]), TI, ms, &(x[WS(vs, 2) + WS(rs, 1)])); } } } VLEAVE(); } static const tw_instr twinstr[] = { VTW(0, 1), VTW(0, 2), VTW(0, 3), {TW_NEXT, VL, 0} }; static const ct_desc desc = { 4, XSIMD_STRING("q1bv_4"), twinstr, &GENUS, {44, 24, 0, 0}, 0, 0, 0 }; void XSIMD(codelet_q1bv_4) (planner *p) { X(kdft_difsq_register) (p, q1bv_4, &desc); } #endif /* HAVE_FMA */ fftw-3.3.3/dft/simd/common/q1bv_5.c0000644000175000017500000004310312054410606013650 00000000000000/* * Copyright (c) 2003, 2007-11 Matteo Frigo * Copyright (c) 2003, 2007-11 Massachusetts Institute of Technology * * This program is free software; you can redistribute it and/or modify * it under the terms of the GNU General Public License as published by * the Free Software Foundation; either version 2 of the License, or * (at your option) any later version. * * This program is distributed in the hope that it will be useful, * but WITHOUT ANY WARRANTY; without even the implied warranty of * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the * GNU General Public License for more details. * * You should have received a copy of the GNU General Public License * along with this program; if not, write to the Free Software * Foundation, Inc., 51 Franklin Street, Fifth Floor, Boston, MA 02110-1301 USA * */ /* This file was automatically generated --- DO NOT EDIT */ /* Generated on Sun Nov 25 07:39:33 EST 2012 */ #include "codelet-dft.h" #ifdef HAVE_FMA /* Generated by: ../../../genfft/gen_twidsq_c.native -fma -reorder-insns -schedule-for-pipeline -simd -compact -variables 4 -pipeline-latency 8 -n 5 -dif -name q1bv_5 -include q1b.h -sign 1 */ /* * This function contains 100 FP additions, 95 FP multiplications, * (or, 55 additions, 50 multiplications, 45 fused multiply/add), * 69 stack variables, 4 constants, and 50 memory accesses */ #include "q1b.h" static void q1bv_5(R *ri, R *ii, const R *W, stride rs, stride vs, INT mb, INT me, INT ms) { DVK(KP559016994, +0.559016994374947424102293417182819058860154590); DVK(KP250000000, +0.250000000000000000000000000000000000000000000); DVK(KP618033988, +0.618033988749894848204586834365638117720309180); DVK(KP951056516, +0.951056516295153572116439333379382143405698634); { INT m; R *x; x = ii; for (m = mb, W = W + (mb * ((TWVL / VL) * 8)); m < me; m = m + VL, x = x + (VL * ms), W = W + (TWVL * 8), MAKE_VOLATILE_STRIDE(10, rs), MAKE_VOLATILE_STRIDE(10, vs)) { V Te, T1w, Ty, TS, TW, Tb, T1t, Tv, T1g, T1c, TP, TV, T1f, T19, TY; V TX; { V T1, T1j, Tl, Ti, Ta, T8, T1A, T1q, T1s, T9, TF, T1r, TZ, TR, TL; V TC, Ts, Tu, TQ, TI, T15, T1b, T10, T11, Tt; { V T1n, T1o, T1k, T1l, T7, Td, T4, Tc; { V T5, T6, T2, T3; T1 = LD(&(x[0]), ms, &(x[0])); T5 = LD(&(x[WS(rs, 2)]), ms, &(x[0])); T6 = LD(&(x[WS(rs, 3)]), ms, &(x[WS(rs, 1)])); T2 = LD(&(x[WS(rs, 1)]), ms, &(x[WS(rs, 1)])); T3 = LD(&(x[WS(rs, 4)]), ms, &(x[0])); T1j = LD(&(x[WS(vs, 4)]), ms, &(x[WS(vs, 4)])); T1n = LD(&(x[WS(vs, 4) + WS(rs, 2)]), ms, &(x[WS(vs, 4)])); T1o = LD(&(x[WS(vs, 4) + WS(rs, 3)]), ms, &(x[WS(vs, 4) + WS(rs, 1)])); T1k = LD(&(x[WS(vs, 4) + WS(rs, 1)]), ms, &(x[WS(vs, 4) + WS(rs, 1)])); T1l = LD(&(x[WS(vs, 4) + WS(rs, 4)]), ms, &(x[WS(vs, 4)])); T7 = VADD(T5, T6); Td = VSUB(T5, T6); T4 = VADD(T2, T3); Tc = VSUB(T2, T3); } { V Tm, Tn, Tr, Tx, T1v, T1p; Tl = LD(&(x[WS(vs, 1)]), ms, &(x[WS(vs, 1)])); T1v = VSUB(T1n, T1o); T1p = VADD(T1n, T1o); { V T1u, T1m, Tp, Tq; T1u = VSUB(T1k, T1l); T1m = VADD(T1k, T1l); Tp = LD(&(x[WS(vs, 1) + WS(rs, 2)]), ms, &(x[WS(vs, 1)])); Ti = VMUL(LDK(KP951056516), VFNMS(LDK(KP618033988), Tc, Td)); Te = VMUL(LDK(KP951056516), VFMA(LDK(KP618033988), Td, Tc)); Ta = VSUB(T4, T7); T8 = VADD(T4, T7); Tq = LD(&(x[WS(vs, 1) + WS(rs, 3)]), ms, &(x[WS(vs, 1) + WS(rs, 1)])); T1w = VMUL(LDK(KP951056516), VFMA(LDK(KP618033988), T1v, T1u)); T1A = VMUL(LDK(KP951056516), VFNMS(LDK(KP618033988), T1u, T1v)); T1q = VADD(T1m, T1p); T1s = VSUB(T1m, T1p); Tm = LD(&(x[WS(vs, 1) + WS(rs, 1)]), ms, &(x[WS(vs, 1) + WS(rs, 1)])); T9 = VFNMS(LDK(KP250000000), T8, T1); Tn = LD(&(x[WS(vs, 1) + WS(rs, 4)]), ms, &(x[WS(vs, 1)])); Tr = VADD(Tp, Tq); Tx = VSUB(Tp, Tq); } { V TJ, TK, TG, Tw, To, TH, T13, T14; TF = LD(&(x[WS(vs, 2)]), ms, &(x[WS(vs, 2)])); T1r = VFNMS(LDK(KP250000000), T1q, T1j); TJ = LD(&(x[WS(vs, 2) + WS(rs, 2)]), ms, &(x[WS(vs, 2)])); TK = LD(&(x[WS(vs, 2) + WS(rs, 3)]), ms, &(x[WS(vs, 2) + WS(rs, 1)])); TG = LD(&(x[WS(vs, 2) + WS(rs, 1)]), ms, &(x[WS(vs, 2) + WS(rs, 1)])); Tw = VSUB(Tm, Tn); To = VADD(Tm, Tn); TH = LD(&(x[WS(vs, 2) + WS(rs, 4)]), ms, &(x[WS(vs, 2)])); TZ = LD(&(x[WS(vs, 3)]), ms, &(x[WS(vs, 3)])); T13 = LD(&(x[WS(vs, 3) + WS(rs, 2)]), ms, &(x[WS(vs, 3)])); T14 = LD(&(x[WS(vs, 3) + WS(rs, 3)]), ms, &(x[WS(vs, 3) + WS(rs, 1)])); TR = VSUB(TJ, TK); TL = VADD(TJ, TK); Ty = VMUL(LDK(KP951056516), VFMA(LDK(KP618033988), Tx, Tw)); TC = VMUL(LDK(KP951056516), VFNMS(LDK(KP618033988), Tw, Tx)); Ts = VADD(To, Tr); Tu = VSUB(To, Tr); TQ = VSUB(TG, TH); TI = VADD(TG, TH); T15 = VADD(T13, T14); T1b = VSUB(T13, T14); T10 = LD(&(x[WS(vs, 3) + WS(rs, 1)]), ms, &(x[WS(vs, 3) + WS(rs, 1)])); T11 = LD(&(x[WS(vs, 3) + WS(rs, 4)]), ms, &(x[WS(vs, 3)])); Tt = VFNMS(LDK(KP250000000), Ts, Tl); } } } { V TO, T12, T1a, Th, T1z, TN, TM, T18, T17; ST(&(x[0]), VADD(T1, T8), ms, &(x[0])); TS = VMUL(LDK(KP951056516), VFMA(LDK(KP618033988), TR, TQ)); TW = VMUL(LDK(KP951056516), VFNMS(LDK(KP618033988), TQ, TR)); TM = VADD(TI, TL); TO = VSUB(TI, TL); ST(&(x[WS(rs, 4)]), VADD(T1j, T1q), ms, &(x[0])); T12 = VADD(T10, T11); T1a = VSUB(T10, T11); ST(&(x[WS(rs, 1)]), VADD(Tl, Ts), ms, &(x[WS(rs, 1)])); Th = VFNMS(LDK(KP559016994), Ta, T9); Tb = VFMA(LDK(KP559016994), Ta, T9); T1t = VFMA(LDK(KP559016994), T1s, T1r); T1z = VFNMS(LDK(KP559016994), T1s, T1r); ST(&(x[WS(rs, 2)]), VADD(TF, TM), ms, &(x[0])); TN = VFNMS(LDK(KP250000000), TM, TF); { V T16, Tk, Tj, T1C, T1B, TD, TE, TB; TB = VFNMS(LDK(KP559016994), Tu, Tt); Tv = VFMA(LDK(KP559016994), Tu, Tt); T1g = VMUL(LDK(KP951056516), VFNMS(LDK(KP618033988), T1a, T1b)); T1c = VMUL(LDK(KP951056516), VFMA(LDK(KP618033988), T1b, T1a)); T18 = VSUB(T12, T15); T16 = VADD(T12, T15); Tk = BYTW(&(W[TWVL * 4]), VFMAI(Ti, Th)); Tj = BYTW(&(W[TWVL * 2]), VFNMSI(Ti, Th)); T1C = BYTW(&(W[TWVL * 4]), VFMAI(T1A, T1z)); T1B = BYTW(&(W[TWVL * 2]), VFNMSI(T1A, T1z)); TD = BYTW(&(W[TWVL * 2]), VFNMSI(TC, TB)); TE = BYTW(&(W[TWVL * 4]), VFMAI(TC, TB)); ST(&(x[WS(rs, 3)]), VADD(TZ, T16), ms, &(x[WS(rs, 1)])); T17 = VFNMS(LDK(KP250000000), T16, TZ); ST(&(x[WS(vs, 3)]), Tk, ms, &(x[WS(vs, 3)])); ST(&(x[WS(vs, 2)]), Tj, ms, &(x[WS(vs, 2)])); ST(&(x[WS(vs, 3) + WS(rs, 4)]), T1C, ms, &(x[WS(vs, 3)])); ST(&(x[WS(vs, 2) + WS(rs, 4)]), T1B, ms, &(x[WS(vs, 2)])); ST(&(x[WS(vs, 2) + WS(rs, 1)]), TD, ms, &(x[WS(vs, 2) + WS(rs, 1)])); ST(&(x[WS(vs, 3) + WS(rs, 1)]), TE, ms, &(x[WS(vs, 3) + WS(rs, 1)])); } TP = VFMA(LDK(KP559016994), TO, TN); TV = VFNMS(LDK(KP559016994), TO, TN); T1f = VFNMS(LDK(KP559016994), T18, T17); T19 = VFMA(LDK(KP559016994), T18, T17); } } TY = BYTW(&(W[TWVL * 4]), VFMAI(TW, TV)); TX = BYTW(&(W[TWVL * 2]), VFNMSI(TW, TV)); { V T1i, T1h, TU, TT; T1i = BYTW(&(W[TWVL * 4]), VFMAI(T1g, T1f)); T1h = BYTW(&(W[TWVL * 2]), VFNMSI(T1g, T1f)); TU = BYTW(&(W[TWVL * 6]), VFNMSI(TS, TP)); TT = BYTW(&(W[0]), VFMAI(TS, TP)); { V Tg, Tf, TA, Tz; Tg = BYTW(&(W[TWVL * 6]), VFNMSI(Te, Tb)); Tf = BYTW(&(W[0]), VFMAI(Te, Tb)); TA = BYTW(&(W[TWVL * 6]), VFNMSI(Ty, Tv)); Tz = BYTW(&(W[0]), VFMAI(Ty, Tv)); { V T1e, T1d, T1y, T1x; T1e = BYTW(&(W[TWVL * 6]), VFNMSI(T1c, T19)); T1d = BYTW(&(W[0]), VFMAI(T1c, T19)); T1y = BYTW(&(W[TWVL * 6]), VFNMSI(T1w, T1t)); T1x = BYTW(&(W[0]), VFMAI(T1w, T1t)); ST(&(x[WS(vs, 3) + WS(rs, 2)]), TY, ms, &(x[WS(vs, 3)])); ST(&(x[WS(vs, 2) + WS(rs, 2)]), TX, ms, &(x[WS(vs, 2)])); ST(&(x[WS(vs, 3) + WS(rs, 3)]), T1i, ms, &(x[WS(vs, 3) + WS(rs, 1)])); ST(&(x[WS(vs, 2) + WS(rs, 3)]), T1h, ms, &(x[WS(vs, 2) + WS(rs, 1)])); ST(&(x[WS(vs, 4) + WS(rs, 2)]), TU, ms, &(x[WS(vs, 4)])); ST(&(x[WS(vs, 1) + WS(rs, 2)]), TT, ms, &(x[WS(vs, 1)])); ST(&(x[WS(vs, 4)]), Tg, ms, &(x[WS(vs, 4)])); ST(&(x[WS(vs, 1)]), Tf, ms, &(x[WS(vs, 1)])); ST(&(x[WS(vs, 4) + WS(rs, 1)]), TA, ms, &(x[WS(vs, 4) + WS(rs, 1)])); ST(&(x[WS(vs, 1) + WS(rs, 1)]), Tz, ms, &(x[WS(vs, 1) + WS(rs, 1)])); ST(&(x[WS(vs, 4) + WS(rs, 3)]), T1e, ms, &(x[WS(vs, 4) + WS(rs, 1)])); ST(&(x[WS(vs, 1) + WS(rs, 3)]), T1d, ms, &(x[WS(vs, 1) + WS(rs, 1)])); ST(&(x[WS(vs, 4) + WS(rs, 4)]), T1y, ms, &(x[WS(vs, 4)])); ST(&(x[WS(vs, 1) + WS(rs, 4)]), T1x, ms, &(x[WS(vs, 1)])); } } } } } VLEAVE(); } static const tw_instr twinstr[] = { VTW(0, 1), VTW(0, 2), VTW(0, 3), VTW(0, 4), {TW_NEXT, VL, 0} }; static const ct_desc desc = { 5, XSIMD_STRING("q1bv_5"), twinstr, &GENUS, {55, 50, 45, 0}, 0, 0, 0 }; void XSIMD(codelet_q1bv_5) (planner *p) { X(kdft_difsq_register) (p, q1bv_5, &desc); } #else /* HAVE_FMA */ /* Generated by: ../../../genfft/gen_twidsq_c.native -simd -compact -variables 4 -pipeline-latency 8 -n 5 -dif -name q1bv_5 -include q1b.h -sign 1 */ /* * This function contains 100 FP additions, 70 FP multiplications, * (or, 85 additions, 55 multiplications, 15 fused multiply/add), * 44 stack variables, 4 constants, and 50 memory accesses */ #include "q1b.h" static void q1bv_5(R *ri, R *ii, const R *W, stride rs, stride vs, INT mb, INT me, INT ms) { DVK(KP250000000, +0.250000000000000000000000000000000000000000000); DVK(KP559016994, +0.559016994374947424102293417182819058860154590); DVK(KP587785252, +0.587785252292473129168705954639072768597652438); DVK(KP951056516, +0.951056516295153572116439333379382143405698634); { INT m; R *x; x = ii; for (m = mb, W = W + (mb * ((TWVL / VL) * 8)); m < me; m = m + VL, x = x + (VL * ms), W = W + (TWVL * 8), MAKE_VOLATILE_STRIDE(10, rs), MAKE_VOLATILE_STRIDE(10, vs)) { V Tb, T7, Th, Ta, Tc, Td, T1t, T1p, T1z, T1s, T1u, T1v, Tv, Tr, TB; V Tu, Tw, Tx, TP, TL, TV, TO, TQ, TR, T19, T15, T1f, T18, T1a, T1b; { V T6, T9, T3, T8; Tb = LD(&(x[0]), ms, &(x[0])); { V T4, T5, T1, T2; T4 = LD(&(x[WS(rs, 2)]), ms, &(x[0])); T5 = LD(&(x[WS(rs, 3)]), ms, &(x[WS(rs, 1)])); T6 = VSUB(T4, T5); T9 = VADD(T4, T5); T1 = LD(&(x[WS(rs, 1)]), ms, &(x[WS(rs, 1)])); T2 = LD(&(x[WS(rs, 4)]), ms, &(x[0])); T3 = VSUB(T1, T2); T8 = VADD(T1, T2); } T7 = VBYI(VFMA(LDK(KP951056516), T3, VMUL(LDK(KP587785252), T6))); Th = VBYI(VFNMS(LDK(KP951056516), T6, VMUL(LDK(KP587785252), T3))); Ta = VMUL(LDK(KP559016994), VSUB(T8, T9)); Tc = VADD(T8, T9); Td = VFNMS(LDK(KP250000000), Tc, Tb); } { V T1o, T1r, T1l, T1q; T1t = LD(&(x[WS(vs, 4)]), ms, &(x[WS(vs, 4)])); { V T1m, T1n, T1j, T1k; T1m = LD(&(x[WS(vs, 4) + WS(rs, 2)]), ms, &(x[WS(vs, 4)])); T1n = LD(&(x[WS(vs, 4) + WS(rs, 3)]), ms, &(x[WS(vs, 4) + WS(rs, 1)])); T1o = VSUB(T1m, T1n); T1r = VADD(T1m, T1n); T1j = LD(&(x[WS(vs, 4) + WS(rs, 1)]), ms, &(x[WS(vs, 4) + WS(rs, 1)])); T1k = LD(&(x[WS(vs, 4) + WS(rs, 4)]), ms, &(x[WS(vs, 4)])); T1l = VSUB(T1j, T1k); T1q = VADD(T1j, T1k); } T1p = VBYI(VFMA(LDK(KP951056516), T1l, VMUL(LDK(KP587785252), T1o))); T1z = VBYI(VFNMS(LDK(KP951056516), T1o, VMUL(LDK(KP587785252), T1l))); T1s = VMUL(LDK(KP559016994), VSUB(T1q, T1r)); T1u = VADD(T1q, T1r); T1v = VFNMS(LDK(KP250000000), T1u, T1t); } { V Tq, Tt, Tn, Ts; Tv = LD(&(x[WS(vs, 1)]), ms, &(x[WS(vs, 1)])); { V To, Tp, Tl, Tm; To = LD(&(x[WS(vs, 1) + WS(rs, 2)]), ms, &(x[WS(vs, 1)])); Tp = LD(&(x[WS(vs, 1) + WS(rs, 3)]), ms, &(x[WS(vs, 1) + WS(rs, 1)])); Tq = VSUB(To, Tp); Tt = VADD(To, Tp); Tl = LD(&(x[WS(vs, 1) + WS(rs, 1)]), ms, &(x[WS(vs, 1) + WS(rs, 1)])); Tm = LD(&(x[WS(vs, 1) + WS(rs, 4)]), ms, &(x[WS(vs, 1)])); Tn = VSUB(Tl, Tm); Ts = VADD(Tl, Tm); } Tr = VBYI(VFMA(LDK(KP951056516), Tn, VMUL(LDK(KP587785252), Tq))); TB = VBYI(VFNMS(LDK(KP951056516), Tq, VMUL(LDK(KP587785252), Tn))); Tu = VMUL(LDK(KP559016994), VSUB(Ts, Tt)); Tw = VADD(Ts, Tt); Tx = VFNMS(LDK(KP250000000), Tw, Tv); } { V TK, TN, TH, TM; TP = LD(&(x[WS(vs, 2)]), ms, &(x[WS(vs, 2)])); { V TI, TJ, TF, TG; TI = LD(&(x[WS(vs, 2) + WS(rs, 2)]), ms, &(x[WS(vs, 2)])); TJ = LD(&(x[WS(vs, 2) + WS(rs, 3)]), ms, &(x[WS(vs, 2) + WS(rs, 1)])); TK = VSUB(TI, TJ); TN = VADD(TI, TJ); TF = LD(&(x[WS(vs, 2) + WS(rs, 1)]), ms, &(x[WS(vs, 2) + WS(rs, 1)])); TG = LD(&(x[WS(vs, 2) + WS(rs, 4)]), ms, &(x[WS(vs, 2)])); TH = VSUB(TF, TG); TM = VADD(TF, TG); } TL = VBYI(VFMA(LDK(KP951056516), TH, VMUL(LDK(KP587785252), TK))); TV = VBYI(VFNMS(LDK(KP951056516), TK, VMUL(LDK(KP587785252), TH))); TO = VMUL(LDK(KP559016994), VSUB(TM, TN)); TQ = VADD(TM, TN); TR = VFNMS(LDK(KP250000000), TQ, TP); } { V T14, T17, T11, T16; T19 = LD(&(x[WS(vs, 3)]), ms, &(x[WS(vs, 3)])); { V T12, T13, TZ, T10; T12 = LD(&(x[WS(vs, 3) + WS(rs, 2)]), ms, &(x[WS(vs, 3)])); T13 = LD(&(x[WS(vs, 3) + WS(rs, 3)]), ms, &(x[WS(vs, 3) + WS(rs, 1)])); T14 = VSUB(T12, T13); T17 = VADD(T12, T13); TZ = LD(&(x[WS(vs, 3) + WS(rs, 1)]), ms, &(x[WS(vs, 3) + WS(rs, 1)])); T10 = LD(&(x[WS(vs, 3) + WS(rs, 4)]), ms, &(x[WS(vs, 3)])); T11 = VSUB(TZ, T10); T16 = VADD(TZ, T10); } T15 = VBYI(VFMA(LDK(KP951056516), T11, VMUL(LDK(KP587785252), T14))); T1f = VBYI(VFNMS(LDK(KP951056516), T14, VMUL(LDK(KP587785252), T11))); T18 = VMUL(LDK(KP559016994), VSUB(T16, T17)); T1a = VADD(T16, T17); T1b = VFNMS(LDK(KP250000000), T1a, T19); } ST(&(x[0]), VADD(Tb, Tc), ms, &(x[0])); ST(&(x[WS(rs, 4)]), VADD(T1t, T1u), ms, &(x[0])); ST(&(x[WS(rs, 2)]), VADD(TP, TQ), ms, &(x[0])); ST(&(x[WS(rs, 3)]), VADD(T19, T1a), ms, &(x[WS(rs, 1)])); ST(&(x[WS(rs, 1)]), VADD(Tv, Tw), ms, &(x[WS(rs, 1)])); { V Tj, Tk, Ti, T1B, T1C, T1A; Ti = VSUB(Td, Ta); Tj = BYTW(&(W[TWVL * 2]), VADD(Th, Ti)); Tk = BYTW(&(W[TWVL * 4]), VSUB(Ti, Th)); ST(&(x[WS(vs, 2)]), Tj, ms, &(x[WS(vs, 2)])); ST(&(x[WS(vs, 3)]), Tk, ms, &(x[WS(vs, 3)])); T1A = VSUB(T1v, T1s); T1B = BYTW(&(W[TWVL * 2]), VADD(T1z, T1A)); T1C = BYTW(&(W[TWVL * 4]), VSUB(T1A, T1z)); ST(&(x[WS(vs, 2) + WS(rs, 4)]), T1B, ms, &(x[WS(vs, 2)])); ST(&(x[WS(vs, 3) + WS(rs, 4)]), T1C, ms, &(x[WS(vs, 3)])); } { V T1h, T1i, T1g, TD, TE, TC; T1g = VSUB(T1b, T18); T1h = BYTW(&(W[TWVL * 2]), VADD(T1f, T1g)); T1i = BYTW(&(W[TWVL * 4]), VSUB(T1g, T1f)); ST(&(x[WS(vs, 2) + WS(rs, 3)]), T1h, ms, &(x[WS(vs, 2) + WS(rs, 1)])); ST(&(x[WS(vs, 3) + WS(rs, 3)]), T1i, ms, &(x[WS(vs, 3) + WS(rs, 1)])); TC = VSUB(Tx, Tu); TD = BYTW(&(W[TWVL * 2]), VADD(TB, TC)); TE = BYTW(&(W[TWVL * 4]), VSUB(TC, TB)); ST(&(x[WS(vs, 2) + WS(rs, 1)]), TD, ms, &(x[WS(vs, 2) + WS(rs, 1)])); ST(&(x[WS(vs, 3) + WS(rs, 1)]), TE, ms, &(x[WS(vs, 3) + WS(rs, 1)])); } { V TX, TY, TW, TT, TU, TS; TW = VSUB(TR, TO); TX = BYTW(&(W[TWVL * 2]), VADD(TV, TW)); TY = BYTW(&(W[TWVL * 4]), VSUB(TW, TV)); ST(&(x[WS(vs, 2) + WS(rs, 2)]), TX, ms, &(x[WS(vs, 2)])); ST(&(x[WS(vs, 3) + WS(rs, 2)]), TY, ms, &(x[WS(vs, 3)])); TS = VADD(TO, TR); TT = BYTW(&(W[0]), VADD(TL, TS)); TU = BYTW(&(W[TWVL * 6]), VSUB(TS, TL)); ST(&(x[WS(vs, 1) + WS(rs, 2)]), TT, ms, &(x[WS(vs, 1)])); ST(&(x[WS(vs, 4) + WS(rs, 2)]), TU, ms, &(x[WS(vs, 4)])); } { V Tf, Tg, Te, Tz, TA, Ty; Te = VADD(Ta, Td); Tf = BYTW(&(W[0]), VADD(T7, Te)); Tg = BYTW(&(W[TWVL * 6]), VSUB(Te, T7)); ST(&(x[WS(vs, 1)]), Tf, ms, &(x[WS(vs, 1)])); ST(&(x[WS(vs, 4)]), Tg, ms, &(x[WS(vs, 4)])); Ty = VADD(Tu, Tx); Tz = BYTW(&(W[0]), VADD(Tr, Ty)); TA = BYTW(&(W[TWVL * 6]), VSUB(Ty, Tr)); ST(&(x[WS(vs, 1) + WS(rs, 1)]), Tz, ms, &(x[WS(vs, 1) + WS(rs, 1)])); ST(&(x[WS(vs, 4) + WS(rs, 1)]), TA, ms, &(x[WS(vs, 4) + WS(rs, 1)])); } { V T1d, T1e, T1c, T1x, T1y, T1w; T1c = VADD(T18, T1b); T1d = BYTW(&(W[0]), VADD(T15, T1c)); T1e = BYTW(&(W[TWVL * 6]), VSUB(T1c, T15)); ST(&(x[WS(vs, 1) + WS(rs, 3)]), T1d, ms, &(x[WS(vs, 1) + WS(rs, 1)])); ST(&(x[WS(vs, 4) + WS(rs, 3)]), T1e, ms, &(x[WS(vs, 4) + WS(rs, 1)])); T1w = VADD(T1s, T1v); T1x = BYTW(&(W[0]), VADD(T1p, T1w)); T1y = BYTW(&(W[TWVL * 6]), VSUB(T1w, T1p)); ST(&(x[WS(vs, 1) + WS(rs, 4)]), T1x, ms, &(x[WS(vs, 1)])); ST(&(x[WS(vs, 4) + WS(rs, 4)]), T1y, ms, &(x[WS(vs, 4)])); } } } VLEAVE(); } static const tw_instr twinstr[] = { VTW(0, 1), VTW(0, 2), VTW(0, 3), VTW(0, 4), {TW_NEXT, VL, 0} }; static const ct_desc desc = { 5, XSIMD_STRING("q1bv_5"), twinstr, &GENUS, {85, 55, 15, 0}, 0, 0, 0 }; void XSIMD(codelet_q1bv_5) (planner *p) { X(kdft_difsq_register) (p, q1bv_5, &desc); } #endif /* HAVE_FMA */ fftw-3.3.3/dft/simd/common/q1bv_8.c0000644000175000017500000011556112054410607013664 00000000000000/* * Copyright (c) 2003, 2007-11 Matteo Frigo * Copyright (c) 2003, 2007-11 Massachusetts Institute of Technology * * This program is free software; you can redistribute it and/or modify * it under the terms of the GNU General Public License as published by * the Free Software Foundation; either version 2 of the License, or * (at your option) any later version. * * This program is distributed in the hope that it will be useful, * but WITHOUT ANY WARRANTY; without even the implied warranty of * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the * GNU General Public License for more details. * * You should have received a copy of the GNU General Public License * along with this program; if not, write to the Free Software * Foundation, Inc., 51 Franklin Street, Fifth Floor, Boston, MA 02110-1301 USA * */ /* This file was automatically generated --- DO NOT EDIT */ /* Generated on Sun Nov 25 07:39:33 EST 2012 */ #include "codelet-dft.h" #ifdef HAVE_FMA /* Generated by: ../../../genfft/gen_twidsq_c.native -fma -reorder-insns -schedule-for-pipeline -simd -compact -variables 4 -pipeline-latency 8 -n 8 -dif -name q1bv_8 -include q1b.h -sign 1 */ /* * This function contains 264 FP additions, 192 FP multiplications, * (or, 184 additions, 112 multiplications, 80 fused multiply/add), * 121 stack variables, 1 constants, and 128 memory accesses */ #include "q1b.h" static void q1bv_8(R *ri, R *ii, const R *W, stride rs, stride vs, INT mb, INT me, INT ms) { DVK(KP707106781, +0.707106781186547524400844362104849039284835938); { INT m; R *x; x = ii; for (m = mb, W = W + (mb * ((TWVL / VL) * 14)); m < me; m = m + VL, x = x + (VL * ms), W = W + (TWVL * 14), MAKE_VOLATILE_STRIDE(16, rs), MAKE_VOLATILE_STRIDE(16, vs)) { V T42, T43, T1U, T1V, T2Y, T2Z, TT, TS, T45, T44; { V T3, Te, T1E, T1P, Tv, Tp, T26, T20, T2b, T2m, T3M, T2x, T2D, T3X, TA; V TL, T48, T4e, T17, T12, TW, T1i, T2I, T1z, T1t, T2T, T3f, T3q, T34, T3a; V T3H, T3B, Ts, Tw, Tf, Ta, T23, T27, T1Q, T1L, T2A, T2E, T2n, T2i, T4b; V T4f, T3Y, T3T, TZ, T13, TM, TH, T35, T2L, T3j, T1w, T1A, T1j, T1e, T36; V T2O, T3C, T3i, T3k; { V T3d, T32, T3e, T3o, T3p, T33; { V T2v, T2w, T3V, T46, T3W; { V T1, T2, Tc, Td, T1C, T1D, T1N, T1O; T1 = LD(&(x[0]), ms, &(x[0])); T2 = LD(&(x[WS(rs, 4)]), ms, &(x[0])); Tc = LD(&(x[WS(rs, 2)]), ms, &(x[0])); Td = LD(&(x[WS(rs, 6)]), ms, &(x[0])); T1C = LD(&(x[WS(vs, 3)]), ms, &(x[WS(vs, 3)])); T1D = LD(&(x[WS(vs, 3) + WS(rs, 4)]), ms, &(x[WS(vs, 3)])); T1N = LD(&(x[WS(vs, 3) + WS(rs, 2)]), ms, &(x[WS(vs, 3)])); T1O = LD(&(x[WS(vs, 3) + WS(rs, 6)]), ms, &(x[WS(vs, 3)])); { V T29, T1Y, T1Z, T2a, T2k, T2l, Tn, To, T3K, T3L; T29 = LD(&(x[WS(vs, 4)]), ms, &(x[WS(vs, 4)])); T3 = VSUB(T1, T2); Tn = VADD(T1, T2); Te = VSUB(Tc, Td); To = VADD(Tc, Td); T1E = VSUB(T1C, T1D); T1Y = VADD(T1C, T1D); T1P = VSUB(T1N, T1O); T1Z = VADD(T1N, T1O); T2a = LD(&(x[WS(vs, 4) + WS(rs, 4)]), ms, &(x[WS(vs, 4)])); T2k = LD(&(x[WS(vs, 4) + WS(rs, 2)]), ms, &(x[WS(vs, 4)])); T2l = LD(&(x[WS(vs, 4) + WS(rs, 6)]), ms, &(x[WS(vs, 4)])); Tv = VADD(Tn, To); Tp = VSUB(Tn, To); T3K = LD(&(x[WS(vs, 7)]), ms, &(x[WS(vs, 7)])); T3L = LD(&(x[WS(vs, 7) + WS(rs, 4)]), ms, &(x[WS(vs, 7)])); T26 = VADD(T1Y, T1Z); T20 = VSUB(T1Y, T1Z); T2v = VADD(T29, T2a); T2b = VSUB(T29, T2a); T2w = VADD(T2k, T2l); T2m = VSUB(T2k, T2l); T3V = LD(&(x[WS(vs, 7) + WS(rs, 2)]), ms, &(x[WS(vs, 7)])); T46 = VADD(T3K, T3L); T3M = VSUB(T3K, T3L); T3W = LD(&(x[WS(vs, 7) + WS(rs, 6)]), ms, &(x[WS(vs, 7)])); } } { V T15, TU, T16, T1g, TV, T1h; { V Ty, Tz, TJ, TK, T47; Ty = LD(&(x[WS(vs, 1)]), ms, &(x[WS(vs, 1)])); Tz = LD(&(x[WS(vs, 1) + WS(rs, 4)]), ms, &(x[WS(vs, 1)])); TJ = LD(&(x[WS(vs, 1) + WS(rs, 2)]), ms, &(x[WS(vs, 1)])); T2x = VSUB(T2v, T2w); T2D = VADD(T2v, T2w); TK = LD(&(x[WS(vs, 1) + WS(rs, 6)]), ms, &(x[WS(vs, 1)])); T47 = VADD(T3V, T3W); T3X = VSUB(T3V, T3W); T15 = LD(&(x[WS(vs, 2)]), ms, &(x[WS(vs, 2)])); TA = VSUB(Ty, Tz); TU = VADD(Ty, Tz); T16 = LD(&(x[WS(vs, 2) + WS(rs, 4)]), ms, &(x[WS(vs, 2)])); T1g = LD(&(x[WS(vs, 2) + WS(rs, 2)]), ms, &(x[WS(vs, 2)])); TL = VSUB(TJ, TK); TV = VADD(TJ, TK); T48 = VSUB(T46, T47); T4e = VADD(T46, T47); T1h = LD(&(x[WS(vs, 2) + WS(rs, 6)]), ms, &(x[WS(vs, 2)])); } { V T2G, T1r, T2H, T2R, T1s, T2S; T2G = LD(&(x[WS(vs, 5)]), ms, &(x[WS(vs, 5)])); T17 = VSUB(T15, T16); T1r = VADD(T15, T16); T2H = LD(&(x[WS(vs, 5) + WS(rs, 4)]), ms, &(x[WS(vs, 5)])); T12 = VADD(TU, TV); TW = VSUB(TU, TV); T2R = LD(&(x[WS(vs, 5) + WS(rs, 2)]), ms, &(x[WS(vs, 5)])); T1i = VSUB(T1g, T1h); T1s = VADD(T1g, T1h); T2S = LD(&(x[WS(vs, 5) + WS(rs, 6)]), ms, &(x[WS(vs, 5)])); T3d = LD(&(x[WS(vs, 6)]), ms, &(x[WS(vs, 6)])); T2I = VSUB(T2G, T2H); T32 = VADD(T2G, T2H); T3e = LD(&(x[WS(vs, 6) + WS(rs, 4)]), ms, &(x[WS(vs, 6)])); T3o = LD(&(x[WS(vs, 6) + WS(rs, 2)]), ms, &(x[WS(vs, 6)])); T3p = LD(&(x[WS(vs, 6) + WS(rs, 6)]), ms, &(x[WS(vs, 6)])); T1z = VADD(T1r, T1s); T1t = VSUB(T1r, T1s); T33 = VADD(T2R, T2S); T2T = VSUB(T2R, T2S); } } } { V T2y, T2e, T3Q, T2z, T2h, T49, T3P, T3R; { V T6, Tq, T1I, Tr, T9, T21, T1H, T1J; { V T4, T3z, T3A, T5, T7, T8, T1F, T1G; T4 = LD(&(x[WS(rs, 1)]), ms, &(x[WS(rs, 1)])); T3f = VSUB(T3d, T3e); T3z = VADD(T3d, T3e); T3q = VSUB(T3o, T3p); T3A = VADD(T3o, T3p); T5 = LD(&(x[WS(rs, 5)]), ms, &(x[WS(rs, 1)])); T7 = LD(&(x[WS(rs, 7)]), ms, &(x[WS(rs, 1)])); T34 = VSUB(T32, T33); T3a = VADD(T32, T33); T8 = LD(&(x[WS(rs, 3)]), ms, &(x[WS(rs, 1)])); T1F = LD(&(x[WS(vs, 3) + WS(rs, 1)]), ms, &(x[WS(vs, 3) + WS(rs, 1)])); T1G = LD(&(x[WS(vs, 3) + WS(rs, 5)]), ms, &(x[WS(vs, 3) + WS(rs, 1)])); T3H = VADD(T3z, T3A); T3B = VSUB(T3z, T3A); T6 = VSUB(T4, T5); Tq = VADD(T4, T5); T1I = LD(&(x[WS(vs, 3) + WS(rs, 7)]), ms, &(x[WS(vs, 3) + WS(rs, 1)])); Tr = VADD(T7, T8); T9 = VSUB(T7, T8); T21 = VADD(T1F, T1G); T1H = VSUB(T1F, T1G); T1J = LD(&(x[WS(vs, 3) + WS(rs, 3)]), ms, &(x[WS(vs, 3) + WS(rs, 1)])); } { V T2f, T22, T1K, T2g, T2c, T2d, T3N, T3O; T2c = LD(&(x[WS(vs, 4) + WS(rs, 1)]), ms, &(x[WS(vs, 4) + WS(rs, 1)])); T2d = LD(&(x[WS(vs, 4) + WS(rs, 5)]), ms, &(x[WS(vs, 4) + WS(rs, 1)])); T2f = LD(&(x[WS(vs, 4) + WS(rs, 7)]), ms, &(x[WS(vs, 4) + WS(rs, 1)])); Ts = VSUB(Tq, Tr); Tw = VADD(Tq, Tr); Tf = VSUB(T6, T9); Ta = VADD(T6, T9); T22 = VADD(T1I, T1J); T1K = VSUB(T1I, T1J); T2y = VADD(T2c, T2d); T2e = VSUB(T2c, T2d); T2g = LD(&(x[WS(vs, 4) + WS(rs, 3)]), ms, &(x[WS(vs, 4) + WS(rs, 1)])); T3N = LD(&(x[WS(vs, 7) + WS(rs, 1)]), ms, &(x[WS(vs, 7) + WS(rs, 1)])); T3O = LD(&(x[WS(vs, 7) + WS(rs, 5)]), ms, &(x[WS(vs, 7) + WS(rs, 1)])); T3Q = LD(&(x[WS(vs, 7) + WS(rs, 7)]), ms, &(x[WS(vs, 7) + WS(rs, 1)])); T23 = VSUB(T21, T22); T27 = VADD(T21, T22); T1Q = VSUB(T1H, T1K); T1L = VADD(T1H, T1K); T2z = VADD(T2f, T2g); T2h = VSUB(T2f, T2g); T49 = VADD(T3N, T3O); T3P = VSUB(T3N, T3O); T3R = LD(&(x[WS(vs, 7) + WS(rs, 3)]), ms, &(x[WS(vs, 7) + WS(rs, 1)])); } } { V TX, TD, T1b, TY, TG, T1u, T1a, T1c; { V TE, T4a, T3S, TF, TB, TC, T18, T19; TB = LD(&(x[WS(vs, 1) + WS(rs, 1)]), ms, &(x[WS(vs, 1) + WS(rs, 1)])); TC = LD(&(x[WS(vs, 1) + WS(rs, 5)]), ms, &(x[WS(vs, 1) + WS(rs, 1)])); TE = LD(&(x[WS(vs, 1) + WS(rs, 7)]), ms, &(x[WS(vs, 1) + WS(rs, 1)])); T2A = VSUB(T2y, T2z); T2E = VADD(T2y, T2z); T2n = VSUB(T2e, T2h); T2i = VADD(T2e, T2h); T4a = VADD(T3Q, T3R); T3S = VSUB(T3Q, T3R); TX = VADD(TB, TC); TD = VSUB(TB, TC); TF = LD(&(x[WS(vs, 1) + WS(rs, 3)]), ms, &(x[WS(vs, 1) + WS(rs, 1)])); T18 = LD(&(x[WS(vs, 2) + WS(rs, 1)]), ms, &(x[WS(vs, 2) + WS(rs, 1)])); T19 = LD(&(x[WS(vs, 2) + WS(rs, 5)]), ms, &(x[WS(vs, 2) + WS(rs, 1)])); T1b = LD(&(x[WS(vs, 2) + WS(rs, 7)]), ms, &(x[WS(vs, 2) + WS(rs, 1)])); T4b = VSUB(T49, T4a); T4f = VADD(T49, T4a); T3Y = VSUB(T3P, T3S); T3T = VADD(T3P, T3S); TY = VADD(TE, TF); TG = VSUB(TE, TF); T1u = VADD(T18, T19); T1a = VSUB(T18, T19); T1c = LD(&(x[WS(vs, 2) + WS(rs, 3)]), ms, &(x[WS(vs, 2) + WS(rs, 1)])); } { V T2M, T1v, T1d, T2N, T2J, T2K, T3g, T3h; T2J = LD(&(x[WS(vs, 5) + WS(rs, 1)]), ms, &(x[WS(vs, 5) + WS(rs, 1)])); T2K = LD(&(x[WS(vs, 5) + WS(rs, 5)]), ms, &(x[WS(vs, 5) + WS(rs, 1)])); T2M = LD(&(x[WS(vs, 5) + WS(rs, 7)]), ms, &(x[WS(vs, 5) + WS(rs, 1)])); TZ = VSUB(TX, TY); T13 = VADD(TX, TY); TM = VSUB(TD, TG); TH = VADD(TD, TG); T1v = VADD(T1b, T1c); T1d = VSUB(T1b, T1c); T35 = VADD(T2J, T2K); T2L = VSUB(T2J, T2K); T2N = LD(&(x[WS(vs, 5) + WS(rs, 3)]), ms, &(x[WS(vs, 5) + WS(rs, 1)])); T3g = LD(&(x[WS(vs, 6) + WS(rs, 1)]), ms, &(x[WS(vs, 6) + WS(rs, 1)])); T3h = LD(&(x[WS(vs, 6) + WS(rs, 5)]), ms, &(x[WS(vs, 6) + WS(rs, 1)])); T3j = LD(&(x[WS(vs, 6) + WS(rs, 7)]), ms, &(x[WS(vs, 6) + WS(rs, 1)])); T1w = VSUB(T1u, T1v); T1A = VADD(T1u, T1v); T1j = VSUB(T1a, T1d); T1e = VADD(T1a, T1d); T36 = VADD(T2M, T2N); T2O = VSUB(T2M, T2N); T3C = VADD(T3g, T3h); T3i = VSUB(T3g, T3h); T3k = LD(&(x[WS(vs, 6) + WS(rs, 3)]), ms, &(x[WS(vs, 6) + WS(rs, 1)])); } } } } { V T3b, T2U, T2P, T3I, T3r, T3m, T11, T25, T39, T4d; { V T37, T3E, T2B, T24; { V T3D, T3l, Tt, T4c; ST(&(x[0]), VADD(Tv, Tw), ms, &(x[0])); ST(&(x[WS(rs, 2)]), VADD(T1z, T1A), ms, &(x[0])); ST(&(x[WS(rs, 7)]), VADD(T4e, T4f), ms, &(x[WS(rs, 1)])); T37 = VSUB(T35, T36); T3b = VADD(T35, T36); T2U = VSUB(T2L, T2O); T2P = VADD(T2L, T2O); T3D = VADD(T3j, T3k); T3l = VSUB(T3j, T3k); ST(&(x[WS(rs, 4)]), VADD(T2D, T2E), ms, &(x[0])); ST(&(x[WS(rs, 3)]), VADD(T26, T27), ms, &(x[WS(rs, 1)])); ST(&(x[WS(rs, 1)]), VADD(T12, T13), ms, &(x[WS(rs, 1)])); ST(&(x[WS(rs, 5)]), VADD(T3a, T3b), ms, &(x[WS(rs, 1)])); Tt = BYTW(&(W[TWVL * 10]), VFNMSI(Ts, Tp)); T4c = BYTW(&(W[TWVL * 10]), VFNMSI(T4b, T48)); T3E = VSUB(T3C, T3D); T3I = VADD(T3C, T3D); T3r = VSUB(T3i, T3l); T3m = VADD(T3i, T3l); T2B = BYTW(&(W[TWVL * 10]), VFNMSI(T2A, T2x)); T24 = BYTW(&(W[TWVL * 10]), VFNMSI(T23, T20)); ST(&(x[WS(vs, 6)]), Tt, ms, &(x[WS(vs, 6)])); ST(&(x[WS(vs, 6) + WS(rs, 7)]), T4c, ms, &(x[WS(vs, 6) + WS(rs, 1)])); } { V T38, T1y, Tu, T10, T1x, T3F, T2C, T3G; T10 = BYTW(&(W[TWVL * 10]), VFNMSI(TZ, TW)); ST(&(x[WS(rs, 6)]), VADD(T3H, T3I), ms, &(x[0])); T1x = BYTW(&(W[TWVL * 10]), VFNMSI(T1w, T1t)); T3F = BYTW(&(W[TWVL * 10]), VFNMSI(T3E, T3B)); ST(&(x[WS(vs, 6) + WS(rs, 4)]), T2B, ms, &(x[WS(vs, 6)])); ST(&(x[WS(vs, 6) + WS(rs, 3)]), T24, ms, &(x[WS(vs, 6) + WS(rs, 1)])); T38 = BYTW(&(W[TWVL * 10]), VFNMSI(T37, T34)); T1y = BYTW(&(W[TWVL * 2]), VFMAI(T1w, T1t)); ST(&(x[WS(vs, 6) + WS(rs, 1)]), T10, ms, &(x[WS(vs, 6) + WS(rs, 1)])); Tu = BYTW(&(W[TWVL * 2]), VFMAI(Ts, Tp)); ST(&(x[WS(vs, 6) + WS(rs, 2)]), T1x, ms, &(x[WS(vs, 6)])); ST(&(x[WS(vs, 6) + WS(rs, 6)]), T3F, ms, &(x[WS(vs, 6)])); T2C = BYTW(&(W[TWVL * 2]), VFMAI(T2A, T2x)); T3G = BYTW(&(W[TWVL * 2]), VFMAI(T3E, T3B)); ST(&(x[WS(vs, 6) + WS(rs, 5)]), T38, ms, &(x[WS(vs, 6) + WS(rs, 1)])); ST(&(x[WS(vs, 2) + WS(rs, 2)]), T1y, ms, &(x[WS(vs, 2)])); T11 = BYTW(&(W[TWVL * 2]), VFMAI(TZ, TW)); ST(&(x[WS(vs, 2)]), Tu, ms, &(x[WS(vs, 2)])); T25 = BYTW(&(W[TWVL * 2]), VFMAI(T23, T20)); T39 = BYTW(&(W[TWVL * 2]), VFMAI(T37, T34)); ST(&(x[WS(vs, 2) + WS(rs, 4)]), T2C, ms, &(x[WS(vs, 2)])); ST(&(x[WS(vs, 2) + WS(rs, 6)]), T3G, ms, &(x[WS(vs, 2)])); T4d = BYTW(&(W[TWVL * 2]), VFMAI(T4b, T48)); } } { V Tj, Tk, T2r, T2j, T2o, T2s, Ti, Th, T1M, T1R, T41, T40; { V T3c, T4g, T3J, T2F, Tx, T1B; Tx = BYTW(&(W[TWVL * 6]), VSUB(Tv, Tw)); ST(&(x[WS(vs, 2) + WS(rs, 1)]), T11, ms, &(x[WS(vs, 2) + WS(rs, 1)])); T1B = BYTW(&(W[TWVL * 6]), VSUB(T1z, T1A)); ST(&(x[WS(vs, 2) + WS(rs, 3)]), T25, ms, &(x[WS(vs, 2) + WS(rs, 1)])); ST(&(x[WS(vs, 2) + WS(rs, 5)]), T39, ms, &(x[WS(vs, 2) + WS(rs, 1)])); T3c = BYTW(&(W[TWVL * 6]), VSUB(T3a, T3b)); T4g = BYTW(&(W[TWVL * 6]), VSUB(T4e, T4f)); ST(&(x[WS(vs, 2) + WS(rs, 7)]), T4d, ms, &(x[WS(vs, 2) + WS(rs, 1)])); ST(&(x[WS(vs, 4)]), Tx, ms, &(x[WS(vs, 4)])); T3J = BYTW(&(W[TWVL * 6]), VSUB(T3H, T3I)); ST(&(x[WS(vs, 4) + WS(rs, 2)]), T1B, ms, &(x[WS(vs, 4)])); T2F = BYTW(&(W[TWVL * 6]), VSUB(T2D, T2E)); { V T14, Tb, Tg, T28, T3U, T3Z; T28 = BYTW(&(W[TWVL * 6]), VSUB(T26, T27)); ST(&(x[WS(vs, 4) + WS(rs, 5)]), T3c, ms, &(x[WS(vs, 4) + WS(rs, 1)])); ST(&(x[WS(vs, 4) + WS(rs, 7)]), T4g, ms, &(x[WS(vs, 4) + WS(rs, 1)])); T14 = BYTW(&(W[TWVL * 6]), VSUB(T12, T13)); Tj = VFMA(LDK(KP707106781), Ta, T3); Tb = VFNMS(LDK(KP707106781), Ta, T3); ST(&(x[WS(vs, 4) + WS(rs, 6)]), T3J, ms, &(x[WS(vs, 4)])); Tk = VFMA(LDK(KP707106781), Tf, Te); Tg = VFNMS(LDK(KP707106781), Tf, Te); ST(&(x[WS(vs, 4) + WS(rs, 4)]), T2F, ms, &(x[WS(vs, 4)])); ST(&(x[WS(vs, 4) + WS(rs, 3)]), T28, ms, &(x[WS(vs, 4) + WS(rs, 1)])); T3U = VFNMS(LDK(KP707106781), T3T, T3M); T42 = VFMA(LDK(KP707106781), T3T, T3M); T43 = VFMA(LDK(KP707106781), T3Y, T3X); T3Z = VFNMS(LDK(KP707106781), T3Y, T3X); ST(&(x[WS(vs, 4) + WS(rs, 1)]), T14, ms, &(x[WS(vs, 4) + WS(rs, 1)])); T2r = VFMA(LDK(KP707106781), T2i, T2b); T2j = VFNMS(LDK(KP707106781), T2i, T2b); T2o = VFNMS(LDK(KP707106781), T2n, T2m); T2s = VFMA(LDK(KP707106781), T2n, T2m); Ti = BYTW(&(W[TWVL * 8]), VFMAI(Tg, Tb)); Th = BYTW(&(W[TWVL * 4]), VFNMSI(Tg, Tb)); T1U = VFMA(LDK(KP707106781), T1L, T1E); T1M = VFNMS(LDK(KP707106781), T1L, T1E); T1R = VFNMS(LDK(KP707106781), T1Q, T1P); T1V = VFMA(LDK(KP707106781), T1Q, T1P); T41 = BYTW(&(W[TWVL * 8]), VFMAI(T3Z, T3U)); T40 = BYTW(&(W[TWVL * 4]), VFNMSI(T3Z, T3U)); } } { V TQ, TR, T1n, T1o, T3v, T3w; { V TI, TN, T1f, T1k, T3n, T3s; { V T1T, T1S, T2q, T2p; TQ = VFMA(LDK(KP707106781), TH, TA); TI = VFNMS(LDK(KP707106781), TH, TA); T2q = BYTW(&(W[TWVL * 8]), VFMAI(T2o, T2j)); T2p = BYTW(&(W[TWVL * 4]), VFNMSI(T2o, T2j)); ST(&(x[WS(vs, 5)]), Ti, ms, &(x[WS(vs, 5)])); ST(&(x[WS(vs, 3)]), Th, ms, &(x[WS(vs, 3)])); T1T = BYTW(&(W[TWVL * 8]), VFMAI(T1R, T1M)); T1S = BYTW(&(W[TWVL * 4]), VFNMSI(T1R, T1M)); ST(&(x[WS(vs, 5) + WS(rs, 7)]), T41, ms, &(x[WS(vs, 5) + WS(rs, 1)])); ST(&(x[WS(vs, 3) + WS(rs, 7)]), T40, ms, &(x[WS(vs, 3) + WS(rs, 1)])); ST(&(x[WS(vs, 5) + WS(rs, 4)]), T2q, ms, &(x[WS(vs, 5)])); ST(&(x[WS(vs, 3) + WS(rs, 4)]), T2p, ms, &(x[WS(vs, 3)])); TN = VFNMS(LDK(KP707106781), TM, TL); TR = VFMA(LDK(KP707106781), TM, TL); T1n = VFMA(LDK(KP707106781), T1e, T17); T1f = VFNMS(LDK(KP707106781), T1e, T17); ST(&(x[WS(vs, 5) + WS(rs, 3)]), T1T, ms, &(x[WS(vs, 5) + WS(rs, 1)])); ST(&(x[WS(vs, 3) + WS(rs, 3)]), T1S, ms, &(x[WS(vs, 3) + WS(rs, 1)])); T1k = VFNMS(LDK(KP707106781), T1j, T1i); T1o = VFMA(LDK(KP707106781), T1j, T1i); T3v = VFMA(LDK(KP707106781), T3m, T3f); T3n = VFNMS(LDK(KP707106781), T3m, T3f); T3s = VFNMS(LDK(KP707106781), T3r, T3q); T3w = VFMA(LDK(KP707106781), T3r, T3q); } { V T2Q, TP, TO, T2V, T2X, T2W; T2Y = VFMA(LDK(KP707106781), T2P, T2I); T2Q = VFNMS(LDK(KP707106781), T2P, T2I); TP = BYTW(&(W[TWVL * 8]), VFMAI(TN, TI)); TO = BYTW(&(W[TWVL * 4]), VFNMSI(TN, TI)); T2V = VFNMS(LDK(KP707106781), T2U, T2T); T2Z = VFMA(LDK(KP707106781), T2U, T2T); { V T1m, T1l, T3u, T3t; T1m = BYTW(&(W[TWVL * 8]), VFMAI(T1k, T1f)); T1l = BYTW(&(W[TWVL * 4]), VFNMSI(T1k, T1f)); T3u = BYTW(&(W[TWVL * 8]), VFMAI(T3s, T3n)); T3t = BYTW(&(W[TWVL * 4]), VFNMSI(T3s, T3n)); ST(&(x[WS(vs, 5) + WS(rs, 1)]), TP, ms, &(x[WS(vs, 5) + WS(rs, 1)])); ST(&(x[WS(vs, 3) + WS(rs, 1)]), TO, ms, &(x[WS(vs, 3) + WS(rs, 1)])); T2X = BYTW(&(W[TWVL * 8]), VFMAI(T2V, T2Q)); T2W = BYTW(&(W[TWVL * 4]), VFNMSI(T2V, T2Q)); ST(&(x[WS(vs, 5) + WS(rs, 2)]), T1m, ms, &(x[WS(vs, 5)])); ST(&(x[WS(vs, 3) + WS(rs, 2)]), T1l, ms, &(x[WS(vs, 3)])); ST(&(x[WS(vs, 5) + WS(rs, 6)]), T3u, ms, &(x[WS(vs, 5)])); ST(&(x[WS(vs, 3) + WS(rs, 6)]), T3t, ms, &(x[WS(vs, 3)])); } ST(&(x[WS(vs, 5) + WS(rs, 5)]), T2X, ms, &(x[WS(vs, 5) + WS(rs, 1)])); ST(&(x[WS(vs, 3) + WS(rs, 5)]), T2W, ms, &(x[WS(vs, 3) + WS(rs, 1)])); } } { V T3y, T3x, T1q, T1p; T1q = BYTW(&(W[TWVL * 12]), VFNMSI(T1o, T1n)); T1p = BYTW(&(W[0]), VFMAI(T1o, T1n)); { V Tm, Tl, T2u, T2t; Tm = BYTW(&(W[TWVL * 12]), VFNMSI(Tk, Tj)); Tl = BYTW(&(W[0]), VFMAI(Tk, Tj)); T2u = BYTW(&(W[TWVL * 12]), VFNMSI(T2s, T2r)); T2t = BYTW(&(W[0]), VFMAI(T2s, T2r)); ST(&(x[WS(vs, 7) + WS(rs, 2)]), T1q, ms, &(x[WS(vs, 7)])); ST(&(x[WS(vs, 1) + WS(rs, 2)]), T1p, ms, &(x[WS(vs, 1)])); T3y = BYTW(&(W[TWVL * 12]), VFNMSI(T3w, T3v)); T3x = BYTW(&(W[0]), VFMAI(T3w, T3v)); ST(&(x[WS(vs, 7)]), Tm, ms, &(x[WS(vs, 7)])); ST(&(x[WS(vs, 1)]), Tl, ms, &(x[WS(vs, 1)])); ST(&(x[WS(vs, 7) + WS(rs, 4)]), T2u, ms, &(x[WS(vs, 7)])); ST(&(x[WS(vs, 1) + WS(rs, 4)]), T2t, ms, &(x[WS(vs, 1)])); } ST(&(x[WS(vs, 7) + WS(rs, 6)]), T3y, ms, &(x[WS(vs, 7)])); ST(&(x[WS(vs, 1) + WS(rs, 6)]), T3x, ms, &(x[WS(vs, 1)])); TT = BYTW(&(W[TWVL * 12]), VFNMSI(TR, TQ)); TS = BYTW(&(W[0]), VFMAI(TR, TQ)); } } } } } { V T1X, T1W, T31, T30; T1X = BYTW(&(W[TWVL * 12]), VFNMSI(T1V, T1U)); T1W = BYTW(&(W[0]), VFMAI(T1V, T1U)); T31 = BYTW(&(W[TWVL * 12]), VFNMSI(T2Z, T2Y)); T30 = BYTW(&(W[0]), VFMAI(T2Z, T2Y)); ST(&(x[WS(vs, 7) + WS(rs, 1)]), TT, ms, &(x[WS(vs, 7) + WS(rs, 1)])); ST(&(x[WS(vs, 1) + WS(rs, 1)]), TS, ms, &(x[WS(vs, 1) + WS(rs, 1)])); T45 = BYTW(&(W[TWVL * 12]), VFNMSI(T43, T42)); T44 = BYTW(&(W[0]), VFMAI(T43, T42)); ST(&(x[WS(vs, 7) + WS(rs, 3)]), T1X, ms, &(x[WS(vs, 7) + WS(rs, 1)])); ST(&(x[WS(vs, 1) + WS(rs, 3)]), T1W, ms, &(x[WS(vs, 1) + WS(rs, 1)])); ST(&(x[WS(vs, 7) + WS(rs, 5)]), T31, ms, &(x[WS(vs, 7) + WS(rs, 1)])); ST(&(x[WS(vs, 1) + WS(rs, 5)]), T30, ms, &(x[WS(vs, 1) + WS(rs, 1)])); } ST(&(x[WS(vs, 7) + WS(rs, 7)]), T45, ms, &(x[WS(vs, 7) + WS(rs, 1)])); ST(&(x[WS(vs, 1) + WS(rs, 7)]), T44, ms, &(x[WS(vs, 1) + WS(rs, 1)])); } } VLEAVE(); } static const tw_instr twinstr[] = { VTW(0, 1), VTW(0, 2), VTW(0, 3), VTW(0, 4), VTW(0, 5), VTW(0, 6), VTW(0, 7), {TW_NEXT, VL, 0} }; static const ct_desc desc = { 8, XSIMD_STRING("q1bv_8"), twinstr, &GENUS, {184, 112, 80, 0}, 0, 0, 0 }; void XSIMD(codelet_q1bv_8) (planner *p) { X(kdft_difsq_register) (p, q1bv_8, &desc); } #else /* HAVE_FMA */ /* Generated by: ../../../genfft/gen_twidsq_c.native -simd -compact -variables 4 -pipeline-latency 8 -n 8 -dif -name q1bv_8 -include q1b.h -sign 1 */ /* * This function contains 264 FP additions, 128 FP multiplications, * (or, 264 additions, 128 multiplications, 0 fused multiply/add), * 77 stack variables, 1 constants, and 128 memory accesses */ #include "q1b.h" static void q1bv_8(R *ri, R *ii, const R *W, stride rs, stride vs, INT mb, INT me, INT ms) { DVK(KP707106781, +0.707106781186547524400844362104849039284835938); { INT m; R *x; x = ii; for (m = mb, W = W + (mb * ((TWVL / VL) * 14)); m < me; m = m + VL, x = x + (VL * ms), W = W + (TWVL * 14), MAKE_VOLATILE_STRIDE(16, rs), MAKE_VOLATILE_STRIDE(16, vs)) { V Ta, Tv, Te, Tp, T1L, T26, T1P, T20, T2i, T2D, T2m, T2x, T3T, T4e, T3X; V T48, TH, T12, TL, TW, T1e, T1z, T1i, T1t, T2P, T3a, T2T, T34, T3m, T3H; V T3q, T3B, T7, Tw, Tf, Ts, T1I, T27, T1Q, T23, T2f, T2E, T2n, T2A, T3Q; V T4f, T3Y, T4b, TE, T13, TM, TZ, T1b, T1A, T1j, T1w, T2M, T3b, T2U, T37; V T3j, T3I, T3r, T3E, T28, T14; { V T8, T9, To, Tc, Td, Tn; T8 = LD(&(x[WS(rs, 2)]), ms, &(x[0])); T9 = LD(&(x[WS(rs, 6)]), ms, &(x[0])); To = VADD(T8, T9); Tc = LD(&(x[0]), ms, &(x[0])); Td = LD(&(x[WS(rs, 4)]), ms, &(x[0])); Tn = VADD(Tc, Td); Ta = VSUB(T8, T9); Tv = VADD(Tn, To); Te = VSUB(Tc, Td); Tp = VSUB(Tn, To); } { V T1J, T1K, T1Z, T1N, T1O, T1Y; T1J = LD(&(x[WS(vs, 3) + WS(rs, 2)]), ms, &(x[WS(vs, 3)])); T1K = LD(&(x[WS(vs, 3) + WS(rs, 6)]), ms, &(x[WS(vs, 3)])); T1Z = VADD(T1J, T1K); T1N = LD(&(x[WS(vs, 3)]), ms, &(x[WS(vs, 3)])); T1O = LD(&(x[WS(vs, 3) + WS(rs, 4)]), ms, &(x[WS(vs, 3)])); T1Y = VADD(T1N, T1O); T1L = VSUB(T1J, T1K); T26 = VADD(T1Y, T1Z); T1P = VSUB(T1N, T1O); T20 = VSUB(T1Y, T1Z); } { V T2g, T2h, T2w, T2k, T2l, T2v; T2g = LD(&(x[WS(vs, 4) + WS(rs, 2)]), ms, &(x[WS(vs, 4)])); T2h = LD(&(x[WS(vs, 4) + WS(rs, 6)]), ms, &(x[WS(vs, 4)])); T2w = VADD(T2g, T2h); T2k = LD(&(x[WS(vs, 4)]), ms, &(x[WS(vs, 4)])); T2l = LD(&(x[WS(vs, 4) + WS(rs, 4)]), ms, &(x[WS(vs, 4)])); T2v = VADD(T2k, T2l); T2i = VSUB(T2g, T2h); T2D = VADD(T2v, T2w); T2m = VSUB(T2k, T2l); T2x = VSUB(T2v, T2w); } { V T3R, T3S, T47, T3V, T3W, T46; T3R = LD(&(x[WS(vs, 7) + WS(rs, 2)]), ms, &(x[WS(vs, 7)])); T3S = LD(&(x[WS(vs, 7) + WS(rs, 6)]), ms, &(x[WS(vs, 7)])); T47 = VADD(T3R, T3S); T3V = LD(&(x[WS(vs, 7)]), ms, &(x[WS(vs, 7)])); T3W = LD(&(x[WS(vs, 7) + WS(rs, 4)]), ms, &(x[WS(vs, 7)])); T46 = VADD(T3V, T3W); T3T = VSUB(T3R, T3S); T4e = VADD(T46, T47); T3X = VSUB(T3V, T3W); T48 = VSUB(T46, T47); } { V TF, TG, TV, TJ, TK, TU; TF = LD(&(x[WS(vs, 1) + WS(rs, 2)]), ms, &(x[WS(vs, 1)])); TG = LD(&(x[WS(vs, 1) + WS(rs, 6)]), ms, &(x[WS(vs, 1)])); TV = VADD(TF, TG); TJ = LD(&(x[WS(vs, 1)]), ms, &(x[WS(vs, 1)])); TK = LD(&(x[WS(vs, 1) + WS(rs, 4)]), ms, &(x[WS(vs, 1)])); TU = VADD(TJ, TK); TH = VSUB(TF, TG); T12 = VADD(TU, TV); TL = VSUB(TJ, TK); TW = VSUB(TU, TV); } { V T1c, T1d, T1s, T1g, T1h, T1r; T1c = LD(&(x[WS(vs, 2) + WS(rs, 2)]), ms, &(x[WS(vs, 2)])); T1d = LD(&(x[WS(vs, 2) + WS(rs, 6)]), ms, &(x[WS(vs, 2)])); T1s = VADD(T1c, T1d); T1g = LD(&(x[WS(vs, 2)]), ms, &(x[WS(vs, 2)])); T1h = LD(&(x[WS(vs, 2) + WS(rs, 4)]), ms, &(x[WS(vs, 2)])); T1r = VADD(T1g, T1h); T1e = VSUB(T1c, T1d); T1z = VADD(T1r, T1s); T1i = VSUB(T1g, T1h); T1t = VSUB(T1r, T1s); } { V T2N, T2O, T33, T2R, T2S, T32; T2N = LD(&(x[WS(vs, 5) + WS(rs, 2)]), ms, &(x[WS(vs, 5)])); T2O = LD(&(x[WS(vs, 5) + WS(rs, 6)]), ms, &(x[WS(vs, 5)])); T33 = VADD(T2N, T2O); T2R = LD(&(x[WS(vs, 5)]), ms, &(x[WS(vs, 5)])); T2S = LD(&(x[WS(vs, 5) + WS(rs, 4)]), ms, &(x[WS(vs, 5)])); T32 = VADD(T2R, T2S); T2P = VSUB(T2N, T2O); T3a = VADD(T32, T33); T2T = VSUB(T2R, T2S); T34 = VSUB(T32, T33); } { V T3k, T3l, T3A, T3o, T3p, T3z; T3k = LD(&(x[WS(vs, 6) + WS(rs, 2)]), ms, &(x[WS(vs, 6)])); T3l = LD(&(x[WS(vs, 6) + WS(rs, 6)]), ms, &(x[WS(vs, 6)])); T3A = VADD(T3k, T3l); T3o = LD(&(x[WS(vs, 6)]), ms, &(x[WS(vs, 6)])); T3p = LD(&(x[WS(vs, 6) + WS(rs, 4)]), ms, &(x[WS(vs, 6)])); T3z = VADD(T3o, T3p); T3m = VSUB(T3k, T3l); T3H = VADD(T3z, T3A); T3q = VSUB(T3o, T3p); T3B = VSUB(T3z, T3A); } { V T3, Tq, T6, Tr; { V T1, T2, T4, T5; T1 = LD(&(x[WS(rs, 1)]), ms, &(x[WS(rs, 1)])); T2 = LD(&(x[WS(rs, 5)]), ms, &(x[WS(rs, 1)])); T3 = VSUB(T1, T2); Tq = VADD(T1, T2); T4 = LD(&(x[WS(rs, 7)]), ms, &(x[WS(rs, 1)])); T5 = LD(&(x[WS(rs, 3)]), ms, &(x[WS(rs, 1)])); T6 = VSUB(T4, T5); Tr = VADD(T4, T5); } T7 = VMUL(LDK(KP707106781), VSUB(T3, T6)); Tw = VADD(Tq, Tr); Tf = VMUL(LDK(KP707106781), VADD(T3, T6)); Ts = VBYI(VSUB(Tq, Tr)); } { V T1E, T21, T1H, T22; { V T1C, T1D, T1F, T1G; T1C = LD(&(x[WS(vs, 3) + WS(rs, 1)]), ms, &(x[WS(vs, 3) + WS(rs, 1)])); T1D = LD(&(x[WS(vs, 3) + WS(rs, 5)]), ms, &(x[WS(vs, 3) + WS(rs, 1)])); T1E = VSUB(T1C, T1D); T21 = VADD(T1C, T1D); T1F = LD(&(x[WS(vs, 3) + WS(rs, 7)]), ms, &(x[WS(vs, 3) + WS(rs, 1)])); T1G = LD(&(x[WS(vs, 3) + WS(rs, 3)]), ms, &(x[WS(vs, 3) + WS(rs, 1)])); T1H = VSUB(T1F, T1G); T22 = VADD(T1F, T1G); } T1I = VMUL(LDK(KP707106781), VSUB(T1E, T1H)); T27 = VADD(T21, T22); T1Q = VMUL(LDK(KP707106781), VADD(T1E, T1H)); T23 = VBYI(VSUB(T21, T22)); } { V T2b, T2y, T2e, T2z; { V T29, T2a, T2c, T2d; T29 = LD(&(x[WS(vs, 4) + WS(rs, 1)]), ms, &(x[WS(vs, 4) + WS(rs, 1)])); T2a = LD(&(x[WS(vs, 4) + WS(rs, 5)]), ms, &(x[WS(vs, 4) + WS(rs, 1)])); T2b = VSUB(T29, T2a); T2y = VADD(T29, T2a); T2c = LD(&(x[WS(vs, 4) + WS(rs, 7)]), ms, &(x[WS(vs, 4) + WS(rs, 1)])); T2d = LD(&(x[WS(vs, 4) + WS(rs, 3)]), ms, &(x[WS(vs, 4) + WS(rs, 1)])); T2e = VSUB(T2c, T2d); T2z = VADD(T2c, T2d); } T2f = VMUL(LDK(KP707106781), VSUB(T2b, T2e)); T2E = VADD(T2y, T2z); T2n = VMUL(LDK(KP707106781), VADD(T2b, T2e)); T2A = VBYI(VSUB(T2y, T2z)); } { V T3M, T49, T3P, T4a; { V T3K, T3L, T3N, T3O; T3K = LD(&(x[WS(vs, 7) + WS(rs, 1)]), ms, &(x[WS(vs, 7) + WS(rs, 1)])); T3L = LD(&(x[WS(vs, 7) + WS(rs, 5)]), ms, &(x[WS(vs, 7) + WS(rs, 1)])); T3M = VSUB(T3K, T3L); T49 = VADD(T3K, T3L); T3N = LD(&(x[WS(vs, 7) + WS(rs, 7)]), ms, &(x[WS(vs, 7) + WS(rs, 1)])); T3O = LD(&(x[WS(vs, 7) + WS(rs, 3)]), ms, &(x[WS(vs, 7) + WS(rs, 1)])); T3P = VSUB(T3N, T3O); T4a = VADD(T3N, T3O); } T3Q = VMUL(LDK(KP707106781), VSUB(T3M, T3P)); T4f = VADD(T49, T4a); T3Y = VMUL(LDK(KP707106781), VADD(T3M, T3P)); T4b = VBYI(VSUB(T49, T4a)); } { V TA, TX, TD, TY; { V Ty, Tz, TB, TC; Ty = LD(&(x[WS(vs, 1) + WS(rs, 1)]), ms, &(x[WS(vs, 1) + WS(rs, 1)])); Tz = LD(&(x[WS(vs, 1) + WS(rs, 5)]), ms, &(x[WS(vs, 1) + WS(rs, 1)])); TA = VSUB(Ty, Tz); TX = VADD(Ty, Tz); TB = LD(&(x[WS(vs, 1) + WS(rs, 7)]), ms, &(x[WS(vs, 1) + WS(rs, 1)])); TC = LD(&(x[WS(vs, 1) + WS(rs, 3)]), ms, &(x[WS(vs, 1) + WS(rs, 1)])); TD = VSUB(TB, TC); TY = VADD(TB, TC); } TE = VMUL(LDK(KP707106781), VSUB(TA, TD)); T13 = VADD(TX, TY); TM = VMUL(LDK(KP707106781), VADD(TA, TD)); TZ = VBYI(VSUB(TX, TY)); } { V T17, T1u, T1a, T1v; { V T15, T16, T18, T19; T15 = LD(&(x[WS(vs, 2) + WS(rs, 1)]), ms, &(x[WS(vs, 2) + WS(rs, 1)])); T16 = LD(&(x[WS(vs, 2) + WS(rs, 5)]), ms, &(x[WS(vs, 2) + WS(rs, 1)])); T17 = VSUB(T15, T16); T1u = VADD(T15, T16); T18 = LD(&(x[WS(vs, 2) + WS(rs, 7)]), ms, &(x[WS(vs, 2) + WS(rs, 1)])); T19 = LD(&(x[WS(vs, 2) + WS(rs, 3)]), ms, &(x[WS(vs, 2) + WS(rs, 1)])); T1a = VSUB(T18, T19); T1v = VADD(T18, T19); } T1b = VMUL(LDK(KP707106781), VSUB(T17, T1a)); T1A = VADD(T1u, T1v); T1j = VMUL(LDK(KP707106781), VADD(T17, T1a)); T1w = VBYI(VSUB(T1u, T1v)); } { V T2I, T35, T2L, T36; { V T2G, T2H, T2J, T2K; T2G = LD(&(x[WS(vs, 5) + WS(rs, 1)]), ms, &(x[WS(vs, 5) + WS(rs, 1)])); T2H = LD(&(x[WS(vs, 5) + WS(rs, 5)]), ms, &(x[WS(vs, 5) + WS(rs, 1)])); T2I = VSUB(T2G, T2H); T35 = VADD(T2G, T2H); T2J = LD(&(x[WS(vs, 5) + WS(rs, 7)]), ms, &(x[WS(vs, 5) + WS(rs, 1)])); T2K = LD(&(x[WS(vs, 5) + WS(rs, 3)]), ms, &(x[WS(vs, 5) + WS(rs, 1)])); T2L = VSUB(T2J, T2K); T36 = VADD(T2J, T2K); } T2M = VMUL(LDK(KP707106781), VSUB(T2I, T2L)); T3b = VADD(T35, T36); T2U = VMUL(LDK(KP707106781), VADD(T2I, T2L)); T37 = VBYI(VSUB(T35, T36)); } { V T3f, T3C, T3i, T3D; { V T3d, T3e, T3g, T3h; T3d = LD(&(x[WS(vs, 6) + WS(rs, 1)]), ms, &(x[WS(vs, 6) + WS(rs, 1)])); T3e = LD(&(x[WS(vs, 6) + WS(rs, 5)]), ms, &(x[WS(vs, 6) + WS(rs, 1)])); T3f = VSUB(T3d, T3e); T3C = VADD(T3d, T3e); T3g = LD(&(x[WS(vs, 6) + WS(rs, 7)]), ms, &(x[WS(vs, 6) + WS(rs, 1)])); T3h = LD(&(x[WS(vs, 6) + WS(rs, 3)]), ms, &(x[WS(vs, 6) + WS(rs, 1)])); T3i = VSUB(T3g, T3h); T3D = VADD(T3g, T3h); } T3j = VMUL(LDK(KP707106781), VSUB(T3f, T3i)); T3I = VADD(T3C, T3D); T3r = VMUL(LDK(KP707106781), VADD(T3f, T3i)); T3E = VBYI(VSUB(T3C, T3D)); } ST(&(x[0]), VADD(Tv, Tw), ms, &(x[0])); ST(&(x[WS(rs, 2)]), VADD(T1z, T1A), ms, &(x[0])); ST(&(x[WS(rs, 5)]), VADD(T3a, T3b), ms, &(x[WS(rs, 1)])); ST(&(x[WS(rs, 7)]), VADD(T4e, T4f), ms, &(x[WS(rs, 1)])); ST(&(x[WS(rs, 6)]), VADD(T3H, T3I), ms, &(x[0])); ST(&(x[WS(rs, 4)]), VADD(T2D, T2E), ms, &(x[0])); { V Tt, T4c, T2B, T24; ST(&(x[WS(rs, 3)]), VADD(T26, T27), ms, &(x[WS(rs, 1)])); ST(&(x[WS(rs, 1)]), VADD(T12, T13), ms, &(x[WS(rs, 1)])); Tt = BYTW(&(W[TWVL * 10]), VSUB(Tp, Ts)); ST(&(x[WS(vs, 6)]), Tt, ms, &(x[WS(vs, 6)])); T4c = BYTW(&(W[TWVL * 10]), VSUB(T48, T4b)); ST(&(x[WS(vs, 6) + WS(rs, 7)]), T4c, ms, &(x[WS(vs, 6) + WS(rs, 1)])); T2B = BYTW(&(W[TWVL * 10]), VSUB(T2x, T2A)); ST(&(x[WS(vs, 6) + WS(rs, 4)]), T2B, ms, &(x[WS(vs, 6)])); T24 = BYTW(&(W[TWVL * 10]), VSUB(T20, T23)); ST(&(x[WS(vs, 6) + WS(rs, 3)]), T24, ms, &(x[WS(vs, 6) + WS(rs, 1)])); } { V T10, T1x, T3F, T38, T1y, Tu; T10 = BYTW(&(W[TWVL * 10]), VSUB(TW, TZ)); ST(&(x[WS(vs, 6) + WS(rs, 1)]), T10, ms, &(x[WS(vs, 6) + WS(rs, 1)])); T1x = BYTW(&(W[TWVL * 10]), VSUB(T1t, T1w)); ST(&(x[WS(vs, 6) + WS(rs, 2)]), T1x, ms, &(x[WS(vs, 6)])); T3F = BYTW(&(W[TWVL * 10]), VSUB(T3B, T3E)); ST(&(x[WS(vs, 6) + WS(rs, 6)]), T3F, ms, &(x[WS(vs, 6)])); T38 = BYTW(&(W[TWVL * 10]), VSUB(T34, T37)); ST(&(x[WS(vs, 6) + WS(rs, 5)]), T38, ms, &(x[WS(vs, 6) + WS(rs, 1)])); T1y = BYTW(&(W[TWVL * 2]), VADD(T1t, T1w)); ST(&(x[WS(vs, 2) + WS(rs, 2)]), T1y, ms, &(x[WS(vs, 2)])); Tu = BYTW(&(W[TWVL * 2]), VADD(Tp, Ts)); ST(&(x[WS(vs, 2)]), Tu, ms, &(x[WS(vs, 2)])); } { V T2C, T3G, T11, T25, T39, T4d; T2C = BYTW(&(W[TWVL * 2]), VADD(T2x, T2A)); ST(&(x[WS(vs, 2) + WS(rs, 4)]), T2C, ms, &(x[WS(vs, 2)])); T3G = BYTW(&(W[TWVL * 2]), VADD(T3B, T3E)); ST(&(x[WS(vs, 2) + WS(rs, 6)]), T3G, ms, &(x[WS(vs, 2)])); T11 = BYTW(&(W[TWVL * 2]), VADD(TW, TZ)); ST(&(x[WS(vs, 2) + WS(rs, 1)]), T11, ms, &(x[WS(vs, 2) + WS(rs, 1)])); T25 = BYTW(&(W[TWVL * 2]), VADD(T20, T23)); ST(&(x[WS(vs, 2) + WS(rs, 3)]), T25, ms, &(x[WS(vs, 2) + WS(rs, 1)])); T39 = BYTW(&(W[TWVL * 2]), VADD(T34, T37)); ST(&(x[WS(vs, 2) + WS(rs, 5)]), T39, ms, &(x[WS(vs, 2) + WS(rs, 1)])); T4d = BYTW(&(W[TWVL * 2]), VADD(T48, T4b)); ST(&(x[WS(vs, 2) + WS(rs, 7)]), T4d, ms, &(x[WS(vs, 2) + WS(rs, 1)])); } { V Tx, T1B, T3c, T4g, T3J, T2F; Tx = BYTW(&(W[TWVL * 6]), VSUB(Tv, Tw)); ST(&(x[WS(vs, 4)]), Tx, ms, &(x[WS(vs, 4)])); T1B = BYTW(&(W[TWVL * 6]), VSUB(T1z, T1A)); ST(&(x[WS(vs, 4) + WS(rs, 2)]), T1B, ms, &(x[WS(vs, 4)])); T3c = BYTW(&(W[TWVL * 6]), VSUB(T3a, T3b)); ST(&(x[WS(vs, 4) + WS(rs, 5)]), T3c, ms, &(x[WS(vs, 4) + WS(rs, 1)])); T4g = BYTW(&(W[TWVL * 6]), VSUB(T4e, T4f)); ST(&(x[WS(vs, 4) + WS(rs, 7)]), T4g, ms, &(x[WS(vs, 4) + WS(rs, 1)])); T3J = BYTW(&(W[TWVL * 6]), VSUB(T3H, T3I)); ST(&(x[WS(vs, 4) + WS(rs, 6)]), T3J, ms, &(x[WS(vs, 4)])); T2F = BYTW(&(W[TWVL * 6]), VSUB(T2D, T2E)); ST(&(x[WS(vs, 4) + WS(rs, 4)]), T2F, ms, &(x[WS(vs, 4)])); } T28 = BYTW(&(W[TWVL * 6]), VSUB(T26, T27)); ST(&(x[WS(vs, 4) + WS(rs, 3)]), T28, ms, &(x[WS(vs, 4) + WS(rs, 1)])); T14 = BYTW(&(W[TWVL * 6]), VSUB(T12, T13)); ST(&(x[WS(vs, 4) + WS(rs, 1)]), T14, ms, &(x[WS(vs, 4) + WS(rs, 1)])); { V Th, Ti, Tb, Tg; Tb = VBYI(VSUB(T7, Ta)); Tg = VSUB(Te, Tf); Th = BYTW(&(W[TWVL * 4]), VADD(Tb, Tg)); Ti = BYTW(&(W[TWVL * 8]), VSUB(Tg, Tb)); ST(&(x[WS(vs, 3)]), Th, ms, &(x[WS(vs, 3)])); ST(&(x[WS(vs, 5)]), Ti, ms, &(x[WS(vs, 5)])); } { V T40, T41, T3U, T3Z; T3U = VBYI(VSUB(T3Q, T3T)); T3Z = VSUB(T3X, T3Y); T40 = BYTW(&(W[TWVL * 4]), VADD(T3U, T3Z)); T41 = BYTW(&(W[TWVL * 8]), VSUB(T3Z, T3U)); ST(&(x[WS(vs, 3) + WS(rs, 7)]), T40, ms, &(x[WS(vs, 3) + WS(rs, 1)])); ST(&(x[WS(vs, 5) + WS(rs, 7)]), T41, ms, &(x[WS(vs, 5) + WS(rs, 1)])); } { V T2p, T2q, T2j, T2o; T2j = VBYI(VSUB(T2f, T2i)); T2o = VSUB(T2m, T2n); T2p = BYTW(&(W[TWVL * 4]), VADD(T2j, T2o)); T2q = BYTW(&(W[TWVL * 8]), VSUB(T2o, T2j)); ST(&(x[WS(vs, 3) + WS(rs, 4)]), T2p, ms, &(x[WS(vs, 3)])); ST(&(x[WS(vs, 5) + WS(rs, 4)]), T2q, ms, &(x[WS(vs, 5)])); } { V T1S, T1T, T1M, T1R; T1M = VBYI(VSUB(T1I, T1L)); T1R = VSUB(T1P, T1Q); T1S = BYTW(&(W[TWVL * 4]), VADD(T1M, T1R)); T1T = BYTW(&(W[TWVL * 8]), VSUB(T1R, T1M)); ST(&(x[WS(vs, 3) + WS(rs, 3)]), T1S, ms, &(x[WS(vs, 3) + WS(rs, 1)])); ST(&(x[WS(vs, 5) + WS(rs, 3)]), T1T, ms, &(x[WS(vs, 5) + WS(rs, 1)])); } { V TO, TP, TI, TN; TI = VBYI(VSUB(TE, TH)); TN = VSUB(TL, TM); TO = BYTW(&(W[TWVL * 4]), VADD(TI, TN)); TP = BYTW(&(W[TWVL * 8]), VSUB(TN, TI)); ST(&(x[WS(vs, 3) + WS(rs, 1)]), TO, ms, &(x[WS(vs, 3) + WS(rs, 1)])); ST(&(x[WS(vs, 5) + WS(rs, 1)]), TP, ms, &(x[WS(vs, 5) + WS(rs, 1)])); } { V T1l, T1m, T1f, T1k; T1f = VBYI(VSUB(T1b, T1e)); T1k = VSUB(T1i, T1j); T1l = BYTW(&(W[TWVL * 4]), VADD(T1f, T1k)); T1m = BYTW(&(W[TWVL * 8]), VSUB(T1k, T1f)); ST(&(x[WS(vs, 3) + WS(rs, 2)]), T1l, ms, &(x[WS(vs, 3)])); ST(&(x[WS(vs, 5) + WS(rs, 2)]), T1m, ms, &(x[WS(vs, 5)])); } { V T3t, T3u, T3n, T3s; T3n = VBYI(VSUB(T3j, T3m)); T3s = VSUB(T3q, T3r); T3t = BYTW(&(W[TWVL * 4]), VADD(T3n, T3s)); T3u = BYTW(&(W[TWVL * 8]), VSUB(T3s, T3n)); ST(&(x[WS(vs, 3) + WS(rs, 6)]), T3t, ms, &(x[WS(vs, 3)])); ST(&(x[WS(vs, 5) + WS(rs, 6)]), T3u, ms, &(x[WS(vs, 5)])); } { V T2W, T2X, T2Q, T2V; T2Q = VBYI(VSUB(T2M, T2P)); T2V = VSUB(T2T, T2U); T2W = BYTW(&(W[TWVL * 4]), VADD(T2Q, T2V)); T2X = BYTW(&(W[TWVL * 8]), VSUB(T2V, T2Q)); ST(&(x[WS(vs, 3) + WS(rs, 5)]), T2W, ms, &(x[WS(vs, 3) + WS(rs, 1)])); ST(&(x[WS(vs, 5) + WS(rs, 5)]), T2X, ms, &(x[WS(vs, 5) + WS(rs, 1)])); } { V T1p, T1q, T1n, T1o; T1n = VBYI(VADD(T1e, T1b)); T1o = VADD(T1i, T1j); T1p = BYTW(&(W[0]), VADD(T1n, T1o)); T1q = BYTW(&(W[TWVL * 12]), VSUB(T1o, T1n)); ST(&(x[WS(vs, 1) + WS(rs, 2)]), T1p, ms, &(x[WS(vs, 1)])); ST(&(x[WS(vs, 7) + WS(rs, 2)]), T1q, ms, &(x[WS(vs, 7)])); } { V Tl, Tm, Tj, Tk; Tj = VBYI(VADD(Ta, T7)); Tk = VADD(Te, Tf); Tl = BYTW(&(W[0]), VADD(Tj, Tk)); Tm = BYTW(&(W[TWVL * 12]), VSUB(Tk, Tj)); ST(&(x[WS(vs, 1)]), Tl, ms, &(x[WS(vs, 1)])); ST(&(x[WS(vs, 7)]), Tm, ms, &(x[WS(vs, 7)])); } { V T2t, T2u, T2r, T2s; T2r = VBYI(VADD(T2i, T2f)); T2s = VADD(T2m, T2n); T2t = BYTW(&(W[0]), VADD(T2r, T2s)); T2u = BYTW(&(W[TWVL * 12]), VSUB(T2s, T2r)); ST(&(x[WS(vs, 1) + WS(rs, 4)]), T2t, ms, &(x[WS(vs, 1)])); ST(&(x[WS(vs, 7) + WS(rs, 4)]), T2u, ms, &(x[WS(vs, 7)])); } { V T3x, T3y, T3v, T3w; T3v = VBYI(VADD(T3m, T3j)); T3w = VADD(T3q, T3r); T3x = BYTW(&(W[0]), VADD(T3v, T3w)); T3y = BYTW(&(W[TWVL * 12]), VSUB(T3w, T3v)); ST(&(x[WS(vs, 1) + WS(rs, 6)]), T3x, ms, &(x[WS(vs, 1)])); ST(&(x[WS(vs, 7) + WS(rs, 6)]), T3y, ms, &(x[WS(vs, 7)])); } { V TS, TT, TQ, TR; TQ = VBYI(VADD(TH, TE)); TR = VADD(TL, TM); TS = BYTW(&(W[0]), VADD(TQ, TR)); TT = BYTW(&(W[TWVL * 12]), VSUB(TR, TQ)); ST(&(x[WS(vs, 1) + WS(rs, 1)]), TS, ms, &(x[WS(vs, 1) + WS(rs, 1)])); ST(&(x[WS(vs, 7) + WS(rs, 1)]), TT, ms, &(x[WS(vs, 7) + WS(rs, 1)])); } { V T1W, T1X, T1U, T1V; T1U = VBYI(VADD(T1L, T1I)); T1V = VADD(T1P, T1Q); T1W = BYTW(&(W[0]), VADD(T1U, T1V)); T1X = BYTW(&(W[TWVL * 12]), VSUB(T1V, T1U)); ST(&(x[WS(vs, 1) + WS(rs, 3)]), T1W, ms, &(x[WS(vs, 1) + WS(rs, 1)])); ST(&(x[WS(vs, 7) + WS(rs, 3)]), T1X, ms, &(x[WS(vs, 7) + WS(rs, 1)])); } { V T30, T31, T2Y, T2Z; T2Y = VBYI(VADD(T2P, T2M)); T2Z = VADD(T2T, T2U); T30 = BYTW(&(W[0]), VADD(T2Y, T2Z)); T31 = BYTW(&(W[TWVL * 12]), VSUB(T2Z, T2Y)); ST(&(x[WS(vs, 1) + WS(rs, 5)]), T30, ms, &(x[WS(vs, 1) + WS(rs, 1)])); ST(&(x[WS(vs, 7) + WS(rs, 5)]), T31, ms, &(x[WS(vs, 7) + WS(rs, 1)])); } { V T44, T45, T42, T43; T42 = VBYI(VADD(T3T, T3Q)); T43 = VADD(T3X, T3Y); T44 = BYTW(&(W[0]), VADD(T42, T43)); T45 = BYTW(&(W[TWVL * 12]), VSUB(T43, T42)); ST(&(x[WS(vs, 1) + WS(rs, 7)]), T44, ms, &(x[WS(vs, 1) + WS(rs, 1)])); ST(&(x[WS(vs, 7) + WS(rs, 7)]), T45, ms, &(x[WS(vs, 7) + WS(rs, 1)])); } } } VLEAVE(); } static const tw_instr twinstr[] = { VTW(0, 1), VTW(0, 2), VTW(0, 3), VTW(0, 4), VTW(0, 5), VTW(0, 6), VTW(0, 7), {TW_NEXT, VL, 0} }; static const ct_desc desc = { 8, XSIMD_STRING("q1bv_8"), twinstr, &GENUS, {264, 128, 0, 0}, 0, 0, 0 }; void XSIMD(codelet_q1bv_8) (planner *p) { X(kdft_difsq_register) (p, q1bv_8, &desc); } #endif /* HAVE_FMA */ fftw-3.3.3/dft/simd/common/codlist.c0000644000175000017500000003344712054410606014226 00000000000000#include "ifftw.h" #include SIMD_HEADER extern void XSIMD(codelet_n1fv_2)(planner *); extern void XSIMD(codelet_n1fv_3)(planner *); extern void XSIMD(codelet_n1fv_4)(planner *); extern void XSIMD(codelet_n1fv_5)(planner *); extern void XSIMD(codelet_n1fv_6)(planner *); extern void XSIMD(codelet_n1fv_7)(planner *); extern void XSIMD(codelet_n1fv_8)(planner *); extern void XSIMD(codelet_n1fv_9)(planner *); extern void XSIMD(codelet_n1fv_10)(planner *); extern void XSIMD(codelet_n1fv_11)(planner *); extern void XSIMD(codelet_n1fv_12)(planner *); extern void XSIMD(codelet_n1fv_13)(planner *); extern void XSIMD(codelet_n1fv_14)(planner *); extern void XSIMD(codelet_n1fv_15)(planner *); extern void XSIMD(codelet_n1fv_16)(planner *); extern void XSIMD(codelet_n1fv_32)(planner *); extern void XSIMD(codelet_n1fv_64)(planner *); extern void XSIMD(codelet_n1fv_128)(planner *); extern void XSIMD(codelet_n1fv_20)(planner *); extern void XSIMD(codelet_n1fv_25)(planner *); extern void XSIMD(codelet_n1bv_2)(planner *); extern void XSIMD(codelet_n1bv_3)(planner *); extern void XSIMD(codelet_n1bv_4)(planner *); extern void XSIMD(codelet_n1bv_5)(planner *); extern void XSIMD(codelet_n1bv_6)(planner *); extern void XSIMD(codelet_n1bv_7)(planner *); extern void XSIMD(codelet_n1bv_8)(planner *); extern void XSIMD(codelet_n1bv_9)(planner *); extern void XSIMD(codelet_n1bv_10)(planner *); extern void XSIMD(codelet_n1bv_11)(planner *); extern void XSIMD(codelet_n1bv_12)(planner *); extern void XSIMD(codelet_n1bv_13)(planner *); extern void XSIMD(codelet_n1bv_14)(planner *); extern void XSIMD(codelet_n1bv_15)(planner *); extern void XSIMD(codelet_n1bv_16)(planner *); extern void XSIMD(codelet_n1bv_32)(planner *); extern void XSIMD(codelet_n1bv_64)(planner *); extern void XSIMD(codelet_n1bv_128)(planner *); extern void XSIMD(codelet_n1bv_20)(planner *); extern void XSIMD(codelet_n1bv_25)(planner *); extern void XSIMD(codelet_n2fv_2)(planner *); extern void XSIMD(codelet_n2fv_4)(planner *); extern void XSIMD(codelet_n2fv_6)(planner *); extern void XSIMD(codelet_n2fv_8)(planner *); extern void XSIMD(codelet_n2fv_10)(planner *); extern void XSIMD(codelet_n2fv_12)(planner *); extern void XSIMD(codelet_n2fv_14)(planner *); extern void XSIMD(codelet_n2fv_16)(planner *); extern void XSIMD(codelet_n2fv_32)(planner *); extern void XSIMD(codelet_n2fv_64)(planner *); extern void XSIMD(codelet_n2fv_20)(planner *); extern void XSIMD(codelet_n2bv_2)(planner *); extern void XSIMD(codelet_n2bv_4)(planner *); extern void XSIMD(codelet_n2bv_6)(planner *); extern void XSIMD(codelet_n2bv_8)(planner *); extern void XSIMD(codelet_n2bv_10)(planner *); extern void XSIMD(codelet_n2bv_12)(planner *); extern void XSIMD(codelet_n2bv_14)(planner *); extern void XSIMD(codelet_n2bv_16)(planner *); extern void XSIMD(codelet_n2bv_32)(planner *); extern void XSIMD(codelet_n2bv_64)(planner *); extern void XSIMD(codelet_n2bv_20)(planner *); extern void XSIMD(codelet_n2sv_4)(planner *); extern void XSIMD(codelet_n2sv_8)(planner *); extern void XSIMD(codelet_n2sv_16)(planner *); extern void XSIMD(codelet_n2sv_32)(planner *); extern void XSIMD(codelet_n2sv_64)(planner *); extern void XSIMD(codelet_t1fuv_2)(planner *); extern void XSIMD(codelet_t1fuv_3)(planner *); extern void XSIMD(codelet_t1fuv_4)(planner *); extern void XSIMD(codelet_t1fuv_5)(planner *); extern void XSIMD(codelet_t1fuv_6)(planner *); extern void XSIMD(codelet_t1fuv_7)(planner *); extern void XSIMD(codelet_t1fuv_8)(planner *); extern void XSIMD(codelet_t1fuv_9)(planner *); extern void XSIMD(codelet_t1fuv_10)(planner *); extern void XSIMD(codelet_t1fv_2)(planner *); extern void XSIMD(codelet_t1fv_3)(planner *); extern void XSIMD(codelet_t1fv_4)(planner *); extern void XSIMD(codelet_t1fv_5)(planner *); extern void XSIMD(codelet_t1fv_6)(planner *); extern void XSIMD(codelet_t1fv_7)(planner *); extern void XSIMD(codelet_t1fv_8)(planner *); extern void XSIMD(codelet_t1fv_9)(planner *); extern void XSIMD(codelet_t1fv_10)(planner *); extern void XSIMD(codelet_t1fv_12)(planner *); extern void XSIMD(codelet_t1fv_15)(planner *); extern void XSIMD(codelet_t1fv_16)(planner *); extern void XSIMD(codelet_t1fv_32)(planner *); extern void XSIMD(codelet_t1fv_64)(planner *); extern void XSIMD(codelet_t1fv_20)(planner *); extern void XSIMD(codelet_t1fv_25)(planner *); extern void XSIMD(codelet_t2fv_2)(planner *); extern void XSIMD(codelet_t2fv_4)(planner *); extern void XSIMD(codelet_t2fv_8)(planner *); extern void XSIMD(codelet_t2fv_16)(planner *); extern void XSIMD(codelet_t2fv_32)(planner *); extern void XSIMD(codelet_t2fv_64)(planner *); extern void XSIMD(codelet_t2fv_5)(planner *); extern void XSIMD(codelet_t2fv_10)(planner *); extern void XSIMD(codelet_t2fv_20)(planner *); extern void XSIMD(codelet_t2fv_25)(planner *); extern void XSIMD(codelet_t3fv_4)(planner *); extern void XSIMD(codelet_t3fv_8)(planner *); extern void XSIMD(codelet_t3fv_16)(planner *); extern void XSIMD(codelet_t3fv_32)(planner *); extern void XSIMD(codelet_t3fv_5)(planner *); extern void XSIMD(codelet_t3fv_10)(planner *); extern void XSIMD(codelet_t3fv_20)(planner *); extern void XSIMD(codelet_t3fv_25)(planner *); extern void XSIMD(codelet_t1buv_2)(planner *); extern void XSIMD(codelet_t1buv_3)(planner *); extern void XSIMD(codelet_t1buv_4)(planner *); extern void XSIMD(codelet_t1buv_5)(planner *); extern void XSIMD(codelet_t1buv_6)(planner *); extern void XSIMD(codelet_t1buv_7)(planner *); extern void XSIMD(codelet_t1buv_8)(planner *); extern void XSIMD(codelet_t1buv_9)(planner *); extern void XSIMD(codelet_t1buv_10)(planner *); extern void XSIMD(codelet_t1bv_2)(planner *); extern void XSIMD(codelet_t1bv_3)(planner *); extern void XSIMD(codelet_t1bv_4)(planner *); extern void XSIMD(codelet_t1bv_5)(planner *); extern void XSIMD(codelet_t1bv_6)(planner *); extern void XSIMD(codelet_t1bv_7)(planner *); extern void XSIMD(codelet_t1bv_8)(planner *); extern void XSIMD(codelet_t1bv_9)(planner *); extern void XSIMD(codelet_t1bv_10)(planner *); extern void XSIMD(codelet_t1bv_12)(planner *); extern void XSIMD(codelet_t1bv_15)(planner *); extern void XSIMD(codelet_t1bv_16)(planner *); extern void XSIMD(codelet_t1bv_32)(planner *); extern void XSIMD(codelet_t1bv_64)(planner *); extern void XSIMD(codelet_t1bv_20)(planner *); extern void XSIMD(codelet_t1bv_25)(planner *); extern void XSIMD(codelet_t2bv_2)(planner *); extern void XSIMD(codelet_t2bv_4)(planner *); extern void XSIMD(codelet_t2bv_8)(planner *); extern void XSIMD(codelet_t2bv_16)(planner *); extern void XSIMD(codelet_t2bv_32)(planner *); extern void XSIMD(codelet_t2bv_64)(planner *); extern void XSIMD(codelet_t2bv_5)(planner *); extern void XSIMD(codelet_t2bv_10)(planner *); extern void XSIMD(codelet_t2bv_20)(planner *); extern void XSIMD(codelet_t2bv_25)(planner *); extern void XSIMD(codelet_t3bv_4)(planner *); extern void XSIMD(codelet_t3bv_8)(planner *); extern void XSIMD(codelet_t3bv_16)(planner *); extern void XSIMD(codelet_t3bv_32)(planner *); extern void XSIMD(codelet_t3bv_5)(planner *); extern void XSIMD(codelet_t3bv_10)(planner *); extern void XSIMD(codelet_t3bv_20)(planner *); extern void XSIMD(codelet_t3bv_25)(planner *); extern void XSIMD(codelet_t1sv_2)(planner *); extern void XSIMD(codelet_t1sv_4)(planner *); extern void XSIMD(codelet_t1sv_8)(planner *); extern void XSIMD(codelet_t1sv_16)(planner *); extern void XSIMD(codelet_t1sv_32)(planner *); extern void XSIMD(codelet_t2sv_4)(planner *); extern void XSIMD(codelet_t2sv_8)(planner *); extern void XSIMD(codelet_t2sv_16)(planner *); extern void XSIMD(codelet_t2sv_32)(planner *); extern void XSIMD(codelet_q1fv_2)(planner *); extern void XSIMD(codelet_q1fv_4)(planner *); extern void XSIMD(codelet_q1fv_5)(planner *); extern void XSIMD(codelet_q1fv_8)(planner *); extern void XSIMD(codelet_q1bv_2)(planner *); extern void XSIMD(codelet_q1bv_4)(planner *); extern void XSIMD(codelet_q1bv_5)(planner *); extern void XSIMD(codelet_q1bv_8)(planner *); extern const solvtab XSIMD(solvtab_dft); const solvtab XSIMD(solvtab_dft) = { SOLVTAB(XSIMD(codelet_n1fv_2)), SOLVTAB(XSIMD(codelet_n1fv_3)), SOLVTAB(XSIMD(codelet_n1fv_4)), SOLVTAB(XSIMD(codelet_n1fv_5)), SOLVTAB(XSIMD(codelet_n1fv_6)), SOLVTAB(XSIMD(codelet_n1fv_7)), SOLVTAB(XSIMD(codelet_n1fv_8)), SOLVTAB(XSIMD(codelet_n1fv_9)), SOLVTAB(XSIMD(codelet_n1fv_10)), SOLVTAB(XSIMD(codelet_n1fv_11)), SOLVTAB(XSIMD(codelet_n1fv_12)), SOLVTAB(XSIMD(codelet_n1fv_13)), SOLVTAB(XSIMD(codelet_n1fv_14)), SOLVTAB(XSIMD(codelet_n1fv_15)), SOLVTAB(XSIMD(codelet_n1fv_16)), SOLVTAB(XSIMD(codelet_n1fv_32)), SOLVTAB(XSIMD(codelet_n1fv_64)), SOLVTAB(XSIMD(codelet_n1fv_128)), SOLVTAB(XSIMD(codelet_n1fv_20)), SOLVTAB(XSIMD(codelet_n1fv_25)), SOLVTAB(XSIMD(codelet_n1bv_2)), SOLVTAB(XSIMD(codelet_n1bv_3)), SOLVTAB(XSIMD(codelet_n1bv_4)), SOLVTAB(XSIMD(codelet_n1bv_5)), SOLVTAB(XSIMD(codelet_n1bv_6)), SOLVTAB(XSIMD(codelet_n1bv_7)), SOLVTAB(XSIMD(codelet_n1bv_8)), SOLVTAB(XSIMD(codelet_n1bv_9)), SOLVTAB(XSIMD(codelet_n1bv_10)), SOLVTAB(XSIMD(codelet_n1bv_11)), SOLVTAB(XSIMD(codelet_n1bv_12)), SOLVTAB(XSIMD(codelet_n1bv_13)), SOLVTAB(XSIMD(codelet_n1bv_14)), SOLVTAB(XSIMD(codelet_n1bv_15)), SOLVTAB(XSIMD(codelet_n1bv_16)), SOLVTAB(XSIMD(codelet_n1bv_32)), SOLVTAB(XSIMD(codelet_n1bv_64)), SOLVTAB(XSIMD(codelet_n1bv_128)), SOLVTAB(XSIMD(codelet_n1bv_20)), SOLVTAB(XSIMD(codelet_n1bv_25)), SOLVTAB(XSIMD(codelet_n2fv_2)), SOLVTAB(XSIMD(codelet_n2fv_4)), SOLVTAB(XSIMD(codelet_n2fv_6)), SOLVTAB(XSIMD(codelet_n2fv_8)), SOLVTAB(XSIMD(codelet_n2fv_10)), SOLVTAB(XSIMD(codelet_n2fv_12)), SOLVTAB(XSIMD(codelet_n2fv_14)), SOLVTAB(XSIMD(codelet_n2fv_16)), SOLVTAB(XSIMD(codelet_n2fv_32)), SOLVTAB(XSIMD(codelet_n2fv_64)), SOLVTAB(XSIMD(codelet_n2fv_20)), SOLVTAB(XSIMD(codelet_n2bv_2)), SOLVTAB(XSIMD(codelet_n2bv_4)), SOLVTAB(XSIMD(codelet_n2bv_6)), SOLVTAB(XSIMD(codelet_n2bv_8)), SOLVTAB(XSIMD(codelet_n2bv_10)), SOLVTAB(XSIMD(codelet_n2bv_12)), SOLVTAB(XSIMD(codelet_n2bv_14)), SOLVTAB(XSIMD(codelet_n2bv_16)), SOLVTAB(XSIMD(codelet_n2bv_32)), SOLVTAB(XSIMD(codelet_n2bv_64)), SOLVTAB(XSIMD(codelet_n2bv_20)), SOLVTAB(XSIMD(codelet_n2sv_4)), SOLVTAB(XSIMD(codelet_n2sv_8)), SOLVTAB(XSIMD(codelet_n2sv_16)), SOLVTAB(XSIMD(codelet_n2sv_32)), SOLVTAB(XSIMD(codelet_n2sv_64)), SOLVTAB(XSIMD(codelet_t1fuv_2)), SOLVTAB(XSIMD(codelet_t1fuv_3)), SOLVTAB(XSIMD(codelet_t1fuv_4)), SOLVTAB(XSIMD(codelet_t1fuv_5)), SOLVTAB(XSIMD(codelet_t1fuv_6)), SOLVTAB(XSIMD(codelet_t1fuv_7)), SOLVTAB(XSIMD(codelet_t1fuv_8)), SOLVTAB(XSIMD(codelet_t1fuv_9)), SOLVTAB(XSIMD(codelet_t1fuv_10)), SOLVTAB(XSIMD(codelet_t1fv_2)), SOLVTAB(XSIMD(codelet_t1fv_3)), SOLVTAB(XSIMD(codelet_t1fv_4)), SOLVTAB(XSIMD(codelet_t1fv_5)), SOLVTAB(XSIMD(codelet_t1fv_6)), SOLVTAB(XSIMD(codelet_t1fv_7)), SOLVTAB(XSIMD(codelet_t1fv_8)), SOLVTAB(XSIMD(codelet_t1fv_9)), SOLVTAB(XSIMD(codelet_t1fv_10)), SOLVTAB(XSIMD(codelet_t1fv_12)), SOLVTAB(XSIMD(codelet_t1fv_15)), SOLVTAB(XSIMD(codelet_t1fv_16)), SOLVTAB(XSIMD(codelet_t1fv_32)), SOLVTAB(XSIMD(codelet_t1fv_64)), SOLVTAB(XSIMD(codelet_t1fv_20)), SOLVTAB(XSIMD(codelet_t1fv_25)), SOLVTAB(XSIMD(codelet_t2fv_2)), SOLVTAB(XSIMD(codelet_t2fv_4)), SOLVTAB(XSIMD(codelet_t2fv_8)), SOLVTAB(XSIMD(codelet_t2fv_16)), SOLVTAB(XSIMD(codelet_t2fv_32)), SOLVTAB(XSIMD(codelet_t2fv_64)), SOLVTAB(XSIMD(codelet_t2fv_5)), SOLVTAB(XSIMD(codelet_t2fv_10)), SOLVTAB(XSIMD(codelet_t2fv_20)), SOLVTAB(XSIMD(codelet_t2fv_25)), SOLVTAB(XSIMD(codelet_t3fv_4)), SOLVTAB(XSIMD(codelet_t3fv_8)), SOLVTAB(XSIMD(codelet_t3fv_16)), SOLVTAB(XSIMD(codelet_t3fv_32)), SOLVTAB(XSIMD(codelet_t3fv_5)), SOLVTAB(XSIMD(codelet_t3fv_10)), SOLVTAB(XSIMD(codelet_t3fv_20)), SOLVTAB(XSIMD(codelet_t3fv_25)), SOLVTAB(XSIMD(codelet_t1buv_2)), SOLVTAB(XSIMD(codelet_t1buv_3)), SOLVTAB(XSIMD(codelet_t1buv_4)), SOLVTAB(XSIMD(codelet_t1buv_5)), SOLVTAB(XSIMD(codelet_t1buv_6)), SOLVTAB(XSIMD(codelet_t1buv_7)), SOLVTAB(XSIMD(codelet_t1buv_8)), SOLVTAB(XSIMD(codelet_t1buv_9)), SOLVTAB(XSIMD(codelet_t1buv_10)), SOLVTAB(XSIMD(codelet_t1bv_2)), SOLVTAB(XSIMD(codelet_t1bv_3)), SOLVTAB(XSIMD(codelet_t1bv_4)), SOLVTAB(XSIMD(codelet_t1bv_5)), SOLVTAB(XSIMD(codelet_t1bv_6)), SOLVTAB(XSIMD(codelet_t1bv_7)), SOLVTAB(XSIMD(codelet_t1bv_8)), SOLVTAB(XSIMD(codelet_t1bv_9)), SOLVTAB(XSIMD(codelet_t1bv_10)), SOLVTAB(XSIMD(codelet_t1bv_12)), SOLVTAB(XSIMD(codelet_t1bv_15)), SOLVTAB(XSIMD(codelet_t1bv_16)), SOLVTAB(XSIMD(codelet_t1bv_32)), SOLVTAB(XSIMD(codelet_t1bv_64)), SOLVTAB(XSIMD(codelet_t1bv_20)), SOLVTAB(XSIMD(codelet_t1bv_25)), SOLVTAB(XSIMD(codelet_t2bv_2)), SOLVTAB(XSIMD(codelet_t2bv_4)), SOLVTAB(XSIMD(codelet_t2bv_8)), SOLVTAB(XSIMD(codelet_t2bv_16)), SOLVTAB(XSIMD(codelet_t2bv_32)), SOLVTAB(XSIMD(codelet_t2bv_64)), SOLVTAB(XSIMD(codelet_t2bv_5)), SOLVTAB(XSIMD(codelet_t2bv_10)), SOLVTAB(XSIMD(codelet_t2bv_20)), SOLVTAB(XSIMD(codelet_t2bv_25)), SOLVTAB(XSIMD(codelet_t3bv_4)), SOLVTAB(XSIMD(codelet_t3bv_8)), SOLVTAB(XSIMD(codelet_t3bv_16)), SOLVTAB(XSIMD(codelet_t3bv_32)), SOLVTAB(XSIMD(codelet_t3bv_5)), SOLVTAB(XSIMD(codelet_t3bv_10)), SOLVTAB(XSIMD(codelet_t3bv_20)), SOLVTAB(XSIMD(codelet_t3bv_25)), SOLVTAB(XSIMD(codelet_t1sv_2)), SOLVTAB(XSIMD(codelet_t1sv_4)), SOLVTAB(XSIMD(codelet_t1sv_8)), SOLVTAB(XSIMD(codelet_t1sv_16)), SOLVTAB(XSIMD(codelet_t1sv_32)), SOLVTAB(XSIMD(codelet_t2sv_4)), SOLVTAB(XSIMD(codelet_t2sv_8)), SOLVTAB(XSIMD(codelet_t2sv_16)), SOLVTAB(XSIMD(codelet_t2sv_32)), SOLVTAB(XSIMD(codelet_q1fv_2)), SOLVTAB(XSIMD(codelet_q1fv_4)), SOLVTAB(XSIMD(codelet_q1fv_5)), SOLVTAB(XSIMD(codelet_q1fv_8)), SOLVTAB(XSIMD(codelet_q1bv_2)), SOLVTAB(XSIMD(codelet_q1bv_4)), SOLVTAB(XSIMD(codelet_q1bv_5)), SOLVTAB(XSIMD(codelet_q1bv_8)), SOLVTAB_END }; fftw-3.3.3/dft/simd/common/genus.c0000644000175000017500000002250312054410050013666 00000000000000/* * Copyright (c) 2003, 2007-11 Matteo Frigo * Copyright (c) 2003, 2007-11 Massachusetts Institute of Technology * * This program is free software; you can redistribute it and/or modify * it under the terms of the GNU General Public License as published by * the Free Software Foundation; either version 2 of the License, or * (at your option) any later version. * * This program is distributed in the hope that it will be useful, * but WITHOUT ANY WARRANTY; without even the implied warranty of * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the * GNU General Public License for more details. * * You should have received a copy of the GNU General Public License * along with this program; if not, write to the Free Software * Foundation, Inc., 51 Franklin Street, Fifth Floor, Boston, MA 02110-1301 USA * */ #include "codelet-dft.h" #include SIMD_HEADER #define EXTERN_CONST(t, x) extern const t x; const t x static int n1b_okp(const kdft_desc *d, const R *ri, const R *ii, const R *ro, const R *io, INT is, INT os, INT vl, INT ivs, INT ovs, const planner *plnr) { return (1 && ALIGNED(ii) && ALIGNED(io) && !NO_SIMDP(plnr) && SIMD_STRIDE_OK(is) && SIMD_STRIDE_OK(os) && SIMD_VSTRIDE_OK(ivs) && SIMD_VSTRIDE_OK(ovs) && ri == ii + 1 && ro == io + 1 && (vl % VL) == 0 && (!d->is || (d->is == is)) && (!d->os || (d->os == os)) && (!d->ivs || (d->ivs == ivs)) && (!d->ovs || (d->ovs == ovs)) ); } EXTERN_CONST(kdft_genus, XSIMD(dft_n1bsimd_genus)) = { n1b_okp, VL }; static int n1f_okp(const kdft_desc *d, const R *ri, const R *ii, const R *ro, const R *io, INT is, INT os, INT vl, INT ivs, INT ovs, const planner *plnr) { return (1 && ALIGNED(ri) && ALIGNED(ro) && !NO_SIMDP(plnr) && SIMD_STRIDE_OK(is) && SIMD_STRIDE_OK(os) && SIMD_VSTRIDE_OK(ivs) && SIMD_VSTRIDE_OK(ovs) && ii == ri + 1 && io == ro + 1 && (vl % VL) == 0 && (!d->is || (d->is == is)) && (!d->os || (d->os == os)) && (!d->ivs || (d->ivs == ivs)) && (!d->ovs || (d->ovs == ovs)) ); } EXTERN_CONST(kdft_genus, XSIMD(dft_n1fsimd_genus)) = { n1f_okp, VL }; static int n2b_okp(const kdft_desc *d, const R *ri, const R *ii, const R *ro, const R *io, INT is, INT os, INT vl, INT ivs, INT ovs, const planner *plnr) { return (1 && ALIGNEDA(ii) && ALIGNEDA(io) && !NO_SIMDP(plnr) && SIMD_STRIDE_OKA(is) && SIMD_VSTRIDE_OKA(ivs) && SIMD_VSTRIDE_OKA(os) /* os == 2 enforced by codelet */ && SIMD_STRIDE_OKPAIR(ovs) && ri == ii + 1 && ro == io + 1 && (vl % VL) == 0 && (!d->is || (d->is == is)) && (!d->os || (d->os == os)) && (!d->ivs || (d->ivs == ivs)) && (!d->ovs || (d->ovs == ovs)) ); } EXTERN_CONST(kdft_genus, XSIMD(dft_n2bsimd_genus)) = { n2b_okp, VL }; static int n2f_okp(const kdft_desc *d, const R *ri, const R *ii, const R *ro, const R *io, INT is, INT os, INT vl, INT ivs, INT ovs, const planner *plnr) { return (1 && ALIGNEDA(ri) && ALIGNEDA(ro) && !NO_SIMDP(plnr) && SIMD_STRIDE_OKA(is) && SIMD_VSTRIDE_OKA(ivs) && SIMD_VSTRIDE_OKA(os) /* os == 2 enforced by codelet */ && SIMD_STRIDE_OKPAIR(ovs) && ii == ri + 1 && io == ro + 1 && (vl % VL) == 0 && (!d->is || (d->is == is)) && (!d->os || (d->os == os)) && (!d->ivs || (d->ivs == ivs)) && (!d->ovs || (d->ovs == ovs)) ); } EXTERN_CONST(kdft_genus, XSIMD(dft_n2fsimd_genus)) = { n2f_okp, VL }; static int n2s_okp(const kdft_desc *d, const R *ri, const R *ii, const R *ro, const R *io, INT is, INT os, INT vl, INT ivs, INT ovs, const planner *plnr) { return (1 && !NO_SIMDP(plnr) && ALIGNEDA(ri) && ALIGNEDA(ii) && ALIGNEDA(ro) && ALIGNEDA(io) && SIMD_STRIDE_OKA(is) && ivs == 1 && os == 1 && SIMD_STRIDE_OKA(ovs) && (vl % (2 * VL)) == 0 && (!d->is || (d->is == is)) && (!d->os || (d->os == os)) && (!d->ivs || (d->ivs == ivs)) && (!d->ovs || (d->ovs == ovs)) ); } EXTERN_CONST(kdft_genus, XSIMD(dft_n2ssimd_genus)) = { n2s_okp, 2 * VL }; static int q1b_okp(const ct_desc *d, const R *rio, const R *iio, INT rs, INT vs, INT m, INT mb, INT me, INT ms, const planner *plnr) { return (1 && ALIGNED(iio) && !NO_SIMDP(plnr) && SIMD_STRIDE_OK(rs) && SIMD_STRIDE_OK(vs) && SIMD_VSTRIDE_OK(ms) && rio == iio + 1 && (m % VL) == 0 && (mb % VL) == 0 && (me % VL) == 0 && (!d->rs || (d->rs == rs)) && (!d->vs || (d->vs == vs)) && (!d->ms || (d->ms == ms)) ); } EXTERN_CONST(ct_genus, XSIMD(dft_q1bsimd_genus)) = { q1b_okp, VL }; static int q1f_okp(const ct_desc *d, const R *rio, const R *iio, INT rs, INT vs, INT m, INT mb, INT me, INT ms, const planner *plnr) { return (1 && ALIGNED(rio) && !NO_SIMDP(plnr) && SIMD_STRIDE_OK(rs) && SIMD_STRIDE_OK(vs) && SIMD_VSTRIDE_OK(ms) && iio == rio + 1 && (m % VL) == 0 && (mb % VL) == 0 && (me % VL) == 0 && (!d->rs || (d->rs == rs)) && (!d->vs || (d->vs == vs)) && (!d->ms || (d->ms == ms)) ); } EXTERN_CONST(ct_genus, XSIMD(dft_q1fsimd_genus)) = { q1f_okp, VL }; static int t_okp_common(const ct_desc *d, const R *rio, const R *iio, INT rs, INT vs, INT m, INT mb, INT me, INT ms, const planner *plnr) { UNUSED(rio); UNUSED(iio); return (1 && !NO_SIMDP(plnr) && SIMD_STRIDE_OKA(rs) && SIMD_VSTRIDE_OKA(ms) && (m % VL) == 0 && (mb % VL) == 0 && (me % VL) == 0 && (!d->rs || (d->rs == rs)) && (!d->vs || (d->vs == vs)) && (!d->ms || (d->ms == ms)) ); } static int t_okp_commonu(const ct_desc *d, const R *rio, const R *iio, INT rs, INT vs, INT m, INT mb, INT me, INT ms, const planner *plnr) { UNUSED(rio); UNUSED(iio); UNUSED(m); return (1 && !NO_SIMDP(plnr) && SIMD_STRIDE_OK(rs) && SIMD_VSTRIDE_OK(ms) && (mb % VL) == 0 && (me % VL) == 0 && (!d->rs || (d->rs == rs)) && (!d->vs || (d->vs == vs)) && (!d->ms || (d->ms == ms)) ); } static int t_okp_t1f(const ct_desc *d, const R *rio, const R *iio, INT rs, INT vs, INT m, INT mb, INT me, INT ms, const planner *plnr) { return t_okp_common(d, rio, iio, rs, vs, m, mb, me, ms, plnr) && iio == rio + 1 && ALIGNEDA(rio); } EXTERN_CONST(ct_genus, XSIMD(dft_t1fsimd_genus)) = { t_okp_t1f, VL }; static int t_okp_t1fu(const ct_desc *d, const R *rio, const R *iio, INT rs, INT vs, INT m, INT mb, INT me, INT ms, const planner *plnr) { return t_okp_commonu(d, rio, iio, rs, vs, m, mb, me, ms, plnr) && iio == rio + 1 && ALIGNED(rio); } EXTERN_CONST(ct_genus, XSIMD(dft_t1fusimd_genus)) = { t_okp_t1fu, VL }; static int t_okp_t1b(const ct_desc *d, const R *rio, const R *iio, INT rs, INT vs, INT m, INT mb, INT me, INT ms, const planner *plnr) { return t_okp_common(d, rio, iio, rs, vs, m, mb, me, ms, plnr) && rio == iio + 1 && ALIGNEDA(iio); } EXTERN_CONST(ct_genus, XSIMD(dft_t1bsimd_genus)) = { t_okp_t1b, VL }; static int t_okp_t1bu(const ct_desc *d, const R *rio, const R *iio, INT rs, INT vs, INT m, INT mb, INT me, INT ms, const planner *plnr) { return t_okp_commonu(d, rio, iio, rs, vs, m, mb, me, ms, plnr) && rio == iio + 1 && ALIGNED(iio); } EXTERN_CONST(ct_genus, XSIMD(dft_t1busimd_genus)) = { t_okp_t1bu, VL }; /* use t2* codelets only when n = m*radix is small, because t2* codelets use ~2n twiddle factors (instead of ~n) */ static int small_enough(const ct_desc *d, INT m) { return m * d->radix <= 16384; } static int t_okp_t2f(const ct_desc *d, const R *rio, const R *iio, INT rs, INT vs, INT m, INT mb, INT me, INT ms, const planner *plnr) { return t_okp_t1f(d, rio, iio, rs, vs, m, mb, me, ms, plnr) && small_enough(d, m); } EXTERN_CONST(ct_genus, XSIMD(dft_t2fsimd_genus)) = { t_okp_t2f, VL }; static int t_okp_t2b(const ct_desc *d, const R *rio, const R *iio, INT rs, INT vs, INT m, INT mb, INT me, INT ms, const planner *plnr) { return t_okp_t1b(d, rio, iio, rs, vs, m, mb, me, ms, plnr) && small_enough(d, m); } EXTERN_CONST(ct_genus, XSIMD(dft_t2bsimd_genus)) = { t_okp_t2b, VL }; static int ts_okp(const ct_desc *d, const R *rio, const R *iio, INT rs, INT vs, INT m, INT mb, INT me, INT ms, const planner *plnr) { UNUSED(rio); UNUSED(iio); return (1 && !NO_SIMDP(plnr) && ALIGNEDA(rio) && ALIGNEDA(iio) && SIMD_STRIDE_OKA(rs) && ms == 1 && (m % (2 * VL)) == 0 && (mb % (2 * VL)) == 0 && (me % (2 * VL)) == 0 && (!d->rs || (d->rs == rs)) && (!d->vs || (d->vs == vs)) && (!d->ms || (d->ms == ms)) ); } EXTERN_CONST(ct_genus, XSIMD(dft_tssimd_genus)) = { ts_okp, 2 * VL }; fftw-3.3.3/dft/simd/sse2/0000755000175000017500000000000012054411106012046 500000000000000fftw-3.3.3/dft/simd/sse2/Makefile.am0000644000175000017500000000043112054410050014015 00000000000000AM_CFLAGS = $(SSE2_CFLAGS) SIMD_HEADER=simd-sse2.h include $(top_srcdir)/dft/simd/codlist.mk include $(top_srcdir)/dft/simd/simd.mk if HAVE_SSE2 BUILT_SOURCES = $(EXTRA_DIST) noinst_LTLIBRARIES = libdft_sse2_codelets.la libdft_sse2_codelets_la_SOURCES = $(BUILT_SOURCES) endif fftw-3.3.3/dft/simd/sse2/Makefile.in0000644000175000017500000011115212054410200014026 00000000000000# Makefile.in generated by automake 1.11.6 from Makefile.am. # @configure_input@ # Copyright (C) 1994, 1995, 1996, 1997, 1998, 1999, 2000, 2001, 2002, # 2003, 2004, 2005, 2006, 2007, 2008, 2009, 2010, 2011 Free Software # Foundation, Inc. # This Makefile.in is free software; the Free Software Foundation # gives unlimited permission to copy and/or distribute it, # with or without modifications, as long as this notice is preserved. # This program is distributed in the hope that it will be useful, # but WITHOUT ANY WARRANTY, to the extent permitted by law; without # even the implied warranty of MERCHANTABILITY or FITNESS FOR A # PARTICULAR PURPOSE. @SET_MAKE@ # This file contains a standard list of DFT SIMD codelets. It is # included by common/Makefile to generate the C files with the actual # codelets in them. It is included by {sse,sse2,...}/Makefile to # generate and compile stub files that include common/*.c # You can customize FFTW for special needs, e.g. to handle certain # sizes more efficiently, by adding new codelets to the lists of those # included by default. If you change the list of codelets, any new # ones you added will be automatically generated when you run the # bootstrap script (see "Generating your own code" in the FFTW # manual). VPATH = @srcdir@ am__make_dryrun = \ { \ am__dry=no; \ case $$MAKEFLAGS in \ *\\[\ \ ]*) \ echo 'am--echo: ; @echo "AM" OK' | $(MAKE) -f - 2>/dev/null \ | grep '^AM OK$$' >/dev/null || am__dry=yes;; \ *) \ for am__flg in $$MAKEFLAGS; do \ case $$am__flg in \ *=*|--*) ;; \ *n*) am__dry=yes; break;; \ esac; \ done;; \ esac; \ test $$am__dry = yes; \ } pkgdatadir = $(datadir)/@PACKAGE@ pkgincludedir = $(includedir)/@PACKAGE@ pkglibdir = $(libdir)/@PACKAGE@ pkglibexecdir = $(libexecdir)/@PACKAGE@ am__cd = CDPATH="$${ZSH_VERSION+.}$(PATH_SEPARATOR)" && cd install_sh_DATA = $(install_sh) -c -m 644 install_sh_PROGRAM = $(install_sh) -c install_sh_SCRIPT = $(install_sh) -c INSTALL_HEADER = $(INSTALL_DATA) transform = $(program_transform_name) NORMAL_INSTALL = : PRE_INSTALL = : POST_INSTALL = : NORMAL_UNINSTALL = : PRE_UNINSTALL = : POST_UNINSTALL = : build_triplet = @build@ host_triplet = @host@ DIST_COMMON = $(srcdir)/Makefile.am $(srcdir)/Makefile.in \ $(top_srcdir)/dft/simd/codlist.mk \ $(top_srcdir)/dft/simd/simd.mk subdir = dft/simd/sse2 ACLOCAL_M4 = $(top_srcdir)/aclocal.m4 am__aclocal_m4_deps = $(top_srcdir)/m4/acx_mpi.m4 \ $(top_srcdir)/m4/acx_pthread.m4 \ $(top_srcdir)/m4/ax_cc_maxopt.m4 \ $(top_srcdir)/m4/ax_check_compiler_flags.m4 \ $(top_srcdir)/m4/ax_compiler_vendor.m4 \ $(top_srcdir)/m4/ax_gcc_aligns_stack.m4 \ $(top_srcdir)/m4/ax_gcc_version.m4 \ $(top_srcdir)/m4/ax_openmp.m4 $(top_srcdir)/m4/libtool.m4 \ $(top_srcdir)/m4/ltoptions.m4 $(top_srcdir)/m4/ltsugar.m4 \ $(top_srcdir)/m4/ltversion.m4 $(top_srcdir)/m4/lt~obsolete.m4 \ $(top_srcdir)/configure.ac am__configure_deps = $(am__aclocal_m4_deps) $(CONFIGURE_DEPENDENCIES) \ $(ACLOCAL_M4) mkinstalldirs = $(install_sh) -d CONFIG_HEADER = $(top_builddir)/config.h CONFIG_CLEAN_FILES = CONFIG_CLEAN_VPATH_FILES = LTLIBRARIES = $(noinst_LTLIBRARIES) libdft_sse2_codelets_la_LIBADD = am__libdft_sse2_codelets_la_SOURCES_DIST = n1fv_2.c n1fv_3.c n1fv_4.c \ n1fv_5.c n1fv_6.c n1fv_7.c n1fv_8.c n1fv_9.c n1fv_10.c \ n1fv_11.c n1fv_12.c n1fv_13.c n1fv_14.c n1fv_15.c n1fv_16.c \ n1fv_32.c n1fv_64.c n1fv_128.c n1fv_20.c n1fv_25.c n1bv_2.c \ n1bv_3.c n1bv_4.c n1bv_5.c n1bv_6.c n1bv_7.c n1bv_8.c n1bv_9.c \ n1bv_10.c n1bv_11.c n1bv_12.c n1bv_13.c n1bv_14.c n1bv_15.c \ n1bv_16.c n1bv_32.c n1bv_64.c n1bv_128.c n1bv_20.c n1bv_25.c \ n2fv_2.c n2fv_4.c n2fv_6.c n2fv_8.c n2fv_10.c n2fv_12.c \ n2fv_14.c n2fv_16.c n2fv_32.c n2fv_64.c n2fv_20.c n2bv_2.c \ n2bv_4.c n2bv_6.c n2bv_8.c n2bv_10.c n2bv_12.c n2bv_14.c \ n2bv_16.c n2bv_32.c n2bv_64.c n2bv_20.c n2sv_4.c n2sv_8.c \ n2sv_16.c n2sv_32.c n2sv_64.c t1fuv_2.c t1fuv_3.c t1fuv_4.c \ t1fuv_5.c t1fuv_6.c t1fuv_7.c t1fuv_8.c t1fuv_9.c t1fuv_10.c \ t1fv_2.c t1fv_3.c t1fv_4.c t1fv_5.c t1fv_6.c t1fv_7.c t1fv_8.c \ t1fv_9.c t1fv_10.c t1fv_12.c t1fv_15.c t1fv_16.c t1fv_32.c \ t1fv_64.c t1fv_20.c t1fv_25.c t2fv_2.c t2fv_4.c t2fv_8.c \ t2fv_16.c t2fv_32.c t2fv_64.c t2fv_5.c t2fv_10.c t2fv_20.c \ t2fv_25.c t3fv_4.c t3fv_8.c t3fv_16.c t3fv_32.c t3fv_5.c \ t3fv_10.c t3fv_20.c t3fv_25.c t1buv_2.c t1buv_3.c t1buv_4.c \ t1buv_5.c t1buv_6.c t1buv_7.c t1buv_8.c t1buv_9.c t1buv_10.c \ t1bv_2.c t1bv_3.c t1bv_4.c t1bv_5.c t1bv_6.c t1bv_7.c t1bv_8.c \ t1bv_9.c t1bv_10.c t1bv_12.c t1bv_15.c t1bv_16.c t1bv_32.c \ t1bv_64.c t1bv_20.c t1bv_25.c t2bv_2.c t2bv_4.c t2bv_8.c \ t2bv_16.c t2bv_32.c t2bv_64.c t2bv_5.c t2bv_10.c t2bv_20.c \ t2bv_25.c t3bv_4.c t3bv_8.c t3bv_16.c t3bv_32.c t3bv_5.c \ t3bv_10.c t3bv_20.c t3bv_25.c t1sv_2.c t1sv_4.c t1sv_8.c \ t1sv_16.c t1sv_32.c t2sv_4.c t2sv_8.c t2sv_16.c t2sv_32.c \ q1fv_2.c q1fv_4.c q1fv_5.c q1fv_8.c q1bv_2.c q1bv_4.c q1bv_5.c \ q1bv_8.c genus.c codlist.c am__objects_1 = n1fv_2.lo n1fv_3.lo n1fv_4.lo n1fv_5.lo n1fv_6.lo \ n1fv_7.lo n1fv_8.lo n1fv_9.lo n1fv_10.lo n1fv_11.lo n1fv_12.lo \ n1fv_13.lo n1fv_14.lo n1fv_15.lo n1fv_16.lo n1fv_32.lo \ n1fv_64.lo n1fv_128.lo n1fv_20.lo n1fv_25.lo am__objects_2 = n1bv_2.lo n1bv_3.lo n1bv_4.lo n1bv_5.lo n1bv_6.lo \ n1bv_7.lo n1bv_8.lo n1bv_9.lo n1bv_10.lo n1bv_11.lo n1bv_12.lo \ n1bv_13.lo n1bv_14.lo n1bv_15.lo n1bv_16.lo n1bv_32.lo \ n1bv_64.lo n1bv_128.lo n1bv_20.lo n1bv_25.lo am__objects_3 = n2fv_2.lo n2fv_4.lo n2fv_6.lo n2fv_8.lo n2fv_10.lo \ n2fv_12.lo n2fv_14.lo n2fv_16.lo n2fv_32.lo n2fv_64.lo \ n2fv_20.lo am__objects_4 = n2bv_2.lo n2bv_4.lo n2bv_6.lo n2bv_8.lo n2bv_10.lo \ n2bv_12.lo n2bv_14.lo n2bv_16.lo n2bv_32.lo n2bv_64.lo \ n2bv_20.lo am__objects_5 = n2sv_4.lo n2sv_8.lo n2sv_16.lo n2sv_32.lo n2sv_64.lo am__objects_6 = t1fuv_2.lo t1fuv_3.lo t1fuv_4.lo t1fuv_5.lo t1fuv_6.lo \ t1fuv_7.lo t1fuv_8.lo t1fuv_9.lo t1fuv_10.lo am__objects_7 = t1fv_2.lo t1fv_3.lo t1fv_4.lo t1fv_5.lo t1fv_6.lo \ t1fv_7.lo t1fv_8.lo t1fv_9.lo t1fv_10.lo t1fv_12.lo t1fv_15.lo \ t1fv_16.lo t1fv_32.lo t1fv_64.lo t1fv_20.lo t1fv_25.lo am__objects_8 = t2fv_2.lo t2fv_4.lo t2fv_8.lo t2fv_16.lo t2fv_32.lo \ t2fv_64.lo t2fv_5.lo t2fv_10.lo t2fv_20.lo t2fv_25.lo am__objects_9 = t3fv_4.lo t3fv_8.lo t3fv_16.lo t3fv_32.lo t3fv_5.lo \ t3fv_10.lo t3fv_20.lo t3fv_25.lo am__objects_10 = t1buv_2.lo t1buv_3.lo t1buv_4.lo t1buv_5.lo \ t1buv_6.lo t1buv_7.lo t1buv_8.lo t1buv_9.lo t1buv_10.lo am__objects_11 = t1bv_2.lo t1bv_3.lo t1bv_4.lo t1bv_5.lo t1bv_6.lo \ t1bv_7.lo t1bv_8.lo t1bv_9.lo t1bv_10.lo t1bv_12.lo t1bv_15.lo \ t1bv_16.lo t1bv_32.lo t1bv_64.lo t1bv_20.lo t1bv_25.lo am__objects_12 = t2bv_2.lo t2bv_4.lo t2bv_8.lo t2bv_16.lo t2bv_32.lo \ t2bv_64.lo t2bv_5.lo t2bv_10.lo t2bv_20.lo t2bv_25.lo am__objects_13 = t3bv_4.lo t3bv_8.lo t3bv_16.lo t3bv_32.lo t3bv_5.lo \ t3bv_10.lo t3bv_20.lo t3bv_25.lo am__objects_14 = t1sv_2.lo t1sv_4.lo t1sv_8.lo t1sv_16.lo t1sv_32.lo am__objects_15 = t2sv_4.lo t2sv_8.lo t2sv_16.lo t2sv_32.lo am__objects_16 = q1fv_2.lo q1fv_4.lo q1fv_5.lo q1fv_8.lo am__objects_17 = q1bv_2.lo q1bv_4.lo q1bv_5.lo q1bv_8.lo am__objects_18 = $(am__objects_1) $(am__objects_2) $(am__objects_3) \ $(am__objects_4) $(am__objects_5) $(am__objects_6) \ $(am__objects_7) $(am__objects_8) $(am__objects_9) \ $(am__objects_10) $(am__objects_11) $(am__objects_12) \ $(am__objects_13) $(am__objects_14) $(am__objects_15) \ $(am__objects_16) $(am__objects_17) am__objects_19 = $(am__objects_18) genus.lo codlist.lo @HAVE_SSE2_TRUE@am__objects_20 = $(am__objects_19) @HAVE_SSE2_TRUE@am_libdft_sse2_codelets_la_OBJECTS = \ @HAVE_SSE2_TRUE@ $(am__objects_20) libdft_sse2_codelets_la_OBJECTS = \ $(am_libdft_sse2_codelets_la_OBJECTS) @HAVE_SSE2_TRUE@am_libdft_sse2_codelets_la_rpath = DEFAULT_INCLUDES = -I.@am__isrc@ -I$(top_builddir) depcomp = $(SHELL) $(top_srcdir)/depcomp am__depfiles_maybe = depfiles am__mv = mv -f COMPILE = $(CC) $(DEFS) $(DEFAULT_INCLUDES) $(INCLUDES) $(AM_CPPFLAGS) \ $(CPPFLAGS) $(AM_CFLAGS) $(CFLAGS) LTCOMPILE = $(LIBTOOL) --tag=CC $(AM_LIBTOOLFLAGS) $(LIBTOOLFLAGS) \ --mode=compile $(CC) $(DEFS) $(DEFAULT_INCLUDES) $(INCLUDES) \ $(AM_CPPFLAGS) $(CPPFLAGS) $(AM_CFLAGS) $(CFLAGS) CCLD = $(CC) LINK = $(LIBTOOL) --tag=CC $(AM_LIBTOOLFLAGS) $(LIBTOOLFLAGS) \ --mode=link $(CCLD) $(AM_CFLAGS) $(CFLAGS) $(AM_LDFLAGS) \ $(LDFLAGS) -o $@ SOURCES = $(libdft_sse2_codelets_la_SOURCES) DIST_SOURCES = $(am__libdft_sse2_codelets_la_SOURCES_DIST) am__can_run_installinfo = \ case $$AM_UPDATE_INFO_DIR in \ n|no|NO) false;; \ *) (install-info --version) >/dev/null 2>&1;; \ esac ETAGS = etags CTAGS = ctags DISTFILES = $(DIST_COMMON) $(DIST_SOURCES) $(TEXINFOS) $(EXTRA_DIST) ACLOCAL = @ACLOCAL@ ALLOCA = @ALLOCA@ ALTIVEC_CFLAGS = @ALTIVEC_CFLAGS@ AMTAR = @AMTAR@ AR = @AR@ AS = @AS@ AUTOCONF = @AUTOCONF@ AUTOHEADER = @AUTOHEADER@ AUTOMAKE = @AUTOMAKE@ AVX_CFLAGS = @AVX_CFLAGS@ AWK = @AWK@ CC = @CC@ CCDEPMODE = @CCDEPMODE@ CFLAGS = @CFLAGS@ CHECK_PL_OPTS = @CHECK_PL_OPTS@ CPP = @CPP@ CPPFLAGS = @CPPFLAGS@ CYGPATH_W = @CYGPATH_W@ C_FFTW_R2R_KIND = @C_FFTW_R2R_KIND@ C_MPI_FINT = @C_MPI_FINT@ DEFS = @DEFS@ DEPDIR = @DEPDIR@ DLLTOOL = @DLLTOOL@ DSYMUTIL = @DSYMUTIL@ DUMPBIN = @DUMPBIN@ ECHO_C = @ECHO_C@ ECHO_N = @ECHO_N@ ECHO_T = @ECHO_T@ EGREP = @EGREP@ EXEEXT = @EXEEXT@ F77 = @F77@ FFLAGS = @FFLAGS@ FGREP = @FGREP@ FLIBS = @FLIBS@ GREP = @GREP@ INSTALL = @INSTALL@ INSTALL_DATA = @INSTALL_DATA@ INSTALL_PROGRAM = @INSTALL_PROGRAM@ INSTALL_SCRIPT = @INSTALL_SCRIPT@ INSTALL_STRIP_PROGRAM = @INSTALL_STRIP_PROGRAM@ LD = @LD@ LDFLAGS = @LDFLAGS@ LIBOBJS = @LIBOBJS@ LIBQUADMATH = @LIBQUADMATH@ LIBS = @LIBS@ LIBTOOL = @LIBTOOL@ LIPO = @LIPO@ LN_S = @LN_S@ LTLIBOBJS = @LTLIBOBJS@ MAINT = @MAINT@ MAKEINFO = @MAKEINFO@ MANIFEST_TOOL = @MANIFEST_TOOL@ MKDIR_P = @MKDIR_P@ MPICC = @MPICC@ MPILIBS = @MPILIBS@ MPIRUN = @MPIRUN@ NEON_CFLAGS = @NEON_CFLAGS@ NM = @NM@ NMEDIT = @NMEDIT@ OBJDUMP = @OBJDUMP@ OBJEXT = @OBJEXT@ OCAMLBUILD = @OCAMLBUILD@ OPENMP_CFLAGS = @OPENMP_CFLAGS@ OTOOL = @OTOOL@ OTOOL64 = @OTOOL64@ PACKAGE = @PACKAGE@ PACKAGE_BUGREPORT = @PACKAGE_BUGREPORT@ PACKAGE_NAME = @PACKAGE_NAME@ PACKAGE_STRING = @PACKAGE_STRING@ PACKAGE_TARNAME = @PACKAGE_TARNAME@ PACKAGE_URL = @PACKAGE_URL@ PACKAGE_VERSION = @PACKAGE_VERSION@ PATH_SEPARATOR = @PATH_SEPARATOR@ POW_LIB = @POW_LIB@ PRECISION = @PRECISION@ PREC_SUFFIX = @PREC_SUFFIX@ PTHREAD_CC = @PTHREAD_CC@ PTHREAD_CFLAGS = @PTHREAD_CFLAGS@ PTHREAD_LIBS = @PTHREAD_LIBS@ RANLIB = @RANLIB@ SED = @SED@ SET_MAKE = @SET_MAKE@ SHARED_VERSION_INFO = @SHARED_VERSION_INFO@ SHELL = @SHELL@ SSE2_CFLAGS = @SSE2_CFLAGS@ STACK_ALIGN_CFLAGS = @STACK_ALIGN_CFLAGS@ STRIP = @STRIP@ THREADLIBS = @THREADLIBS@ VERSION = @VERSION@ abs_builddir = @abs_builddir@ abs_srcdir = @abs_srcdir@ abs_top_builddir = @abs_top_builddir@ abs_top_srcdir = @abs_top_srcdir@ ac_ct_AR = @ac_ct_AR@ ac_ct_CC = @ac_ct_CC@ ac_ct_DUMPBIN = @ac_ct_DUMPBIN@ ac_ct_F77 = @ac_ct_F77@ acx_pthread_config = @acx_pthread_config@ am__include = @am__include@ am__leading_dot = @am__leading_dot@ am__quote = @am__quote@ am__tar = @am__tar@ am__untar = @am__untar@ bindir = @bindir@ build = @build@ build_alias = @build_alias@ build_cpu = @build_cpu@ build_os = @build_os@ build_vendor = @build_vendor@ builddir = @builddir@ datadir = @datadir@ datarootdir = @datarootdir@ docdir = @docdir@ dvidir = @dvidir@ exec_prefix = @exec_prefix@ host = @host@ host_alias = @host_alias@ host_cpu = @host_cpu@ host_os = @host_os@ host_vendor = @host_vendor@ htmldir = @htmldir@ includedir = @includedir@ infodir = @infodir@ install_sh = @install_sh@ libdir = @libdir@ libexecdir = @libexecdir@ localedir = @localedir@ localstatedir = @localstatedir@ mandir = @mandir@ mkdir_p = @mkdir_p@ oldincludedir = @oldincludedir@ pdfdir = @pdfdir@ prefix = @prefix@ program_transform_name = @program_transform_name@ psdir = @psdir@ sbindir = @sbindir@ sharedstatedir = @sharedstatedir@ srcdir = @srcdir@ sysconfdir = @sysconfdir@ target_alias = @target_alias@ top_build_prefix = @top_build_prefix@ top_builddir = @top_builddir@ top_srcdir = @top_srcdir@ AM_CFLAGS = $(SSE2_CFLAGS) SIMD_HEADER = simd-sse2.h ########################################################################### # n1fv_ is a hard-coded FFTW_FORWARD FFT of size , using SIMD N1F = n1fv_2.c n1fv_3.c n1fv_4.c n1fv_5.c n1fv_6.c n1fv_7.c n1fv_8.c \ n1fv_9.c n1fv_10.c n1fv_11.c n1fv_12.c n1fv_13.c n1fv_14.c n1fv_15.c \ n1fv_16.c n1fv_32.c n1fv_64.c n1fv_128.c n1fv_20.c n1fv_25.c # as above, with restricted input vector stride N2F = n2fv_2.c n2fv_4.c n2fv_6.c n2fv_8.c n2fv_10.c n2fv_12.c \ n2fv_14.c n2fv_16.c n2fv_32.c n2fv_64.c n2fv_20.c # as above, but FFTW_BACKWARD N1B = n1bv_2.c n1bv_3.c n1bv_4.c n1bv_5.c n1bv_6.c n1bv_7.c n1bv_8.c \ n1bv_9.c n1bv_10.c n1bv_11.c n1bv_12.c n1bv_13.c n1bv_14.c n1bv_15.c \ n1bv_16.c n1bv_32.c n1bv_64.c n1bv_128.c n1bv_20.c n1bv_25.c N2B = n2bv_2.c n2bv_4.c n2bv_6.c n2bv_8.c n2bv_10.c n2bv_12.c \ n2bv_14.c n2bv_16.c n2bv_32.c n2bv_64.c n2bv_20.c # split-complex codelets N2S = n2sv_4.c n2sv_8.c n2sv_16.c n2sv_32.c n2sv_64.c ########################################################################### # t1fv_ is a "twiddle" FFT of size , implementing a radix-r DIT step # for an FFTW_FORWARD transform, using SIMD T1F = t1fv_2.c t1fv_3.c t1fv_4.c t1fv_5.c t1fv_6.c t1fv_7.c t1fv_8.c \ t1fv_9.c t1fv_10.c t1fv_12.c t1fv_15.c t1fv_16.c t1fv_32.c t1fv_64.c \ t1fv_20.c t1fv_25.c # same as t1fv_*, but with different twiddle storage scheme T2F = t2fv_2.c t2fv_4.c t2fv_8.c t2fv_16.c t2fv_32.c t2fv_64.c \ t2fv_5.c t2fv_10.c t2fv_20.c t2fv_25.c T3F = t3fv_4.c t3fv_8.c t3fv_16.c t3fv_32.c t3fv_5.c t3fv_10.c \ t3fv_20.c t3fv_25.c T1FU = t1fuv_2.c t1fuv_3.c t1fuv_4.c t1fuv_5.c t1fuv_6.c t1fuv_7.c \ t1fuv_8.c t1fuv_9.c t1fuv_10.c # as above, but FFTW_BACKWARD T1B = t1bv_2.c t1bv_3.c t1bv_4.c t1bv_5.c t1bv_6.c t1bv_7.c t1bv_8.c \ t1bv_9.c t1bv_10.c t1bv_12.c t1bv_15.c t1bv_16.c t1bv_32.c t1bv_64.c \ t1bv_20.c t1bv_25.c # same as t1bv_*, but with different twiddle storage scheme T2B = t2bv_2.c t2bv_4.c t2bv_8.c t2bv_16.c t2bv_32.c t2bv_64.c \ t2bv_5.c t2bv_10.c t2bv_20.c t2bv_25.c T3B = t3bv_4.c t3bv_8.c t3bv_16.c t3bv_32.c t3bv_5.c t3bv_10.c \ t3bv_20.c t3bv_25.c T1BU = t1buv_2.c t1buv_3.c t1buv_4.c t1buv_5.c t1buv_6.c t1buv_7.c \ t1buv_8.c t1buv_9.c t1buv_10.c # split-complex codelets T1S = t1sv_2.c t1sv_4.c t1sv_8.c t1sv_16.c t1sv_32.c T2S = t2sv_4.c t2sv_8.c t2sv_16.c t2sv_32.c ########################################################################### # q1fv_ is twiddle FFTW_FORWARD FFTs of size (DIF step), # where the output is transposed, using SIMD. This is used for # in-place transposes in sizes that are divisible by ^2. These # codelets have size ~ ^2, so you should probably not use # bigger than 8 or so. Q1F = q1fv_2.c q1fv_4.c q1fv_5.c q1fv_8.c # as above, but FFTW_BACKWARD Q1B = q1bv_2.c q1bv_4.c q1bv_5.c q1bv_8.c ########################################################################### SIMD_CODELETS = $(N1F) $(N1B) $(N2F) $(N2B) $(N2S) $(T1FU) $(T1F) \ $(T2F) $(T3F) $(T1BU) $(T1B) $(T2B) $(T3B) $(T1S) $(T2S) $(Q1F) $(Q1B) AM_CPPFLAGS = -I$(top_srcdir)/kernel -I$(top_srcdir)/dft \ -I$(top_srcdir)/dft/simd -I$(top_srcdir)/simd-support EXTRA_DIST = $(SIMD_CODELETS) genus.c codlist.c @HAVE_SSE2_TRUE@BUILT_SOURCES = $(EXTRA_DIST) @HAVE_SSE2_TRUE@noinst_LTLIBRARIES = libdft_sse2_codelets.la @HAVE_SSE2_TRUE@libdft_sse2_codelets_la_SOURCES = $(BUILT_SOURCES) all: $(BUILT_SOURCES) $(MAKE) $(AM_MAKEFLAGS) all-am .SUFFIXES: .SUFFIXES: .c .lo .o .obj $(srcdir)/Makefile.in: @MAINTAINER_MODE_TRUE@ $(srcdir)/Makefile.am $(top_srcdir)/dft/simd/codlist.mk $(top_srcdir)/dft/simd/simd.mk $(am__configure_deps) @for dep in $?; do \ case '$(am__configure_deps)' in \ *$$dep*) \ ( cd $(top_builddir) && $(MAKE) $(AM_MAKEFLAGS) am--refresh ) \ && { if test -f $@; then exit 0; else break; fi; }; \ exit 1;; \ esac; \ done; \ echo ' cd $(top_srcdir) && $(AUTOMAKE) --gnu dft/simd/sse2/Makefile'; \ $(am__cd) $(top_srcdir) && \ $(AUTOMAKE) --gnu dft/simd/sse2/Makefile .PRECIOUS: Makefile Makefile: $(srcdir)/Makefile.in $(top_builddir)/config.status @case '$?' in \ *config.status*) \ cd $(top_builddir) && $(MAKE) $(AM_MAKEFLAGS) am--refresh;; \ *) \ echo ' cd $(top_builddir) && $(SHELL) ./config.status $(subdir)/$@ $(am__depfiles_maybe)'; \ cd $(top_builddir) && $(SHELL) ./config.status $(subdir)/$@ $(am__depfiles_maybe);; \ esac; $(top_srcdir)/dft/simd/codlist.mk $(top_srcdir)/dft/simd/simd.mk: $(top_builddir)/config.status: $(top_srcdir)/configure $(CONFIG_STATUS_DEPENDENCIES) cd $(top_builddir) && $(MAKE) $(AM_MAKEFLAGS) am--refresh $(top_srcdir)/configure: @MAINTAINER_MODE_TRUE@ $(am__configure_deps) cd $(top_builddir) && $(MAKE) $(AM_MAKEFLAGS) am--refresh $(ACLOCAL_M4): @MAINTAINER_MODE_TRUE@ $(am__aclocal_m4_deps) cd $(top_builddir) && $(MAKE) $(AM_MAKEFLAGS) am--refresh $(am__aclocal_m4_deps): clean-noinstLTLIBRARIES: -test -z "$(noinst_LTLIBRARIES)" || rm -f $(noinst_LTLIBRARIES) @list='$(noinst_LTLIBRARIES)'; for p in $$list; do \ dir="`echo $$p | sed -e 's|/[^/]*$$||'`"; \ test "$$dir" != "$$p" || dir=.; \ echo "rm -f \"$${dir}/so_locations\""; \ rm -f "$${dir}/so_locations"; \ done libdft_sse2_codelets.la: $(libdft_sse2_codelets_la_OBJECTS) $(libdft_sse2_codelets_la_DEPENDENCIES) $(EXTRA_libdft_sse2_codelets_la_DEPENDENCIES) $(LINK) $(am_libdft_sse2_codelets_la_rpath) $(libdft_sse2_codelets_la_OBJECTS) $(libdft_sse2_codelets_la_LIBADD) $(LIBS) mostlyclean-compile: -rm -f *.$(OBJEXT) distclean-compile: -rm -f *.tab.c @AMDEP_TRUE@@am__include@ @am__quote@./$(DEPDIR)/codlist.Plo@am__quote@ @AMDEP_TRUE@@am__include@ @am__quote@./$(DEPDIR)/genus.Plo@am__quote@ @AMDEP_TRUE@@am__include@ @am__quote@./$(DEPDIR)/n1bv_10.Plo@am__quote@ @AMDEP_TRUE@@am__include@ @am__quote@./$(DEPDIR)/n1bv_11.Plo@am__quote@ @AMDEP_TRUE@@am__include@ @am__quote@./$(DEPDIR)/n1bv_12.Plo@am__quote@ @AMDEP_TRUE@@am__include@ @am__quote@./$(DEPDIR)/n1bv_128.Plo@am__quote@ @AMDEP_TRUE@@am__include@ @am__quote@./$(DEPDIR)/n1bv_13.Plo@am__quote@ @AMDEP_TRUE@@am__include@ @am__quote@./$(DEPDIR)/n1bv_14.Plo@am__quote@ @AMDEP_TRUE@@am__include@ @am__quote@./$(DEPDIR)/n1bv_15.Plo@am__quote@ @AMDEP_TRUE@@am__include@ @am__quote@./$(DEPDIR)/n1bv_16.Plo@am__quote@ @AMDEP_TRUE@@am__include@ @am__quote@./$(DEPDIR)/n1bv_2.Plo@am__quote@ @AMDEP_TRUE@@am__include@ @am__quote@./$(DEPDIR)/n1bv_20.Plo@am__quote@ @AMDEP_TRUE@@am__include@ @am__quote@./$(DEPDIR)/n1bv_25.Plo@am__quote@ @AMDEP_TRUE@@am__include@ @am__quote@./$(DEPDIR)/n1bv_3.Plo@am__quote@ @AMDEP_TRUE@@am__include@ @am__quote@./$(DEPDIR)/n1bv_32.Plo@am__quote@ @AMDEP_TRUE@@am__include@ @am__quote@./$(DEPDIR)/n1bv_4.Plo@am__quote@ @AMDEP_TRUE@@am__include@ @am__quote@./$(DEPDIR)/n1bv_5.Plo@am__quote@ @AMDEP_TRUE@@am__include@ @am__quote@./$(DEPDIR)/n1bv_6.Plo@am__quote@ @AMDEP_TRUE@@am__include@ @am__quote@./$(DEPDIR)/n1bv_64.Plo@am__quote@ @AMDEP_TRUE@@am__include@ @am__quote@./$(DEPDIR)/n1bv_7.Plo@am__quote@ @AMDEP_TRUE@@am__include@ @am__quote@./$(DEPDIR)/n1bv_8.Plo@am__quote@ @AMDEP_TRUE@@am__include@ @am__quote@./$(DEPDIR)/n1bv_9.Plo@am__quote@ @AMDEP_TRUE@@am__include@ @am__quote@./$(DEPDIR)/n1fv_10.Plo@am__quote@ @AMDEP_TRUE@@am__include@ @am__quote@./$(DEPDIR)/n1fv_11.Plo@am__quote@ @AMDEP_TRUE@@am__include@ @am__quote@./$(DEPDIR)/n1fv_12.Plo@am__quote@ @AMDEP_TRUE@@am__include@ @am__quote@./$(DEPDIR)/n1fv_128.Plo@am__quote@ @AMDEP_TRUE@@am__include@ @am__quote@./$(DEPDIR)/n1fv_13.Plo@am__quote@ @AMDEP_TRUE@@am__include@ @am__quote@./$(DEPDIR)/n1fv_14.Plo@am__quote@ @AMDEP_TRUE@@am__include@ @am__quote@./$(DEPDIR)/n1fv_15.Plo@am__quote@ @AMDEP_TRUE@@am__include@ @am__quote@./$(DEPDIR)/n1fv_16.Plo@am__quote@ @AMDEP_TRUE@@am__include@ @am__quote@./$(DEPDIR)/n1fv_2.Plo@am__quote@ @AMDEP_TRUE@@am__include@ @am__quote@./$(DEPDIR)/n1fv_20.Plo@am__quote@ @AMDEP_TRUE@@am__include@ @am__quote@./$(DEPDIR)/n1fv_25.Plo@am__quote@ @AMDEP_TRUE@@am__include@ @am__quote@./$(DEPDIR)/n1fv_3.Plo@am__quote@ @AMDEP_TRUE@@am__include@ @am__quote@./$(DEPDIR)/n1fv_32.Plo@am__quote@ @AMDEP_TRUE@@am__include@ @am__quote@./$(DEPDIR)/n1fv_4.Plo@am__quote@ @AMDEP_TRUE@@am__include@ @am__quote@./$(DEPDIR)/n1fv_5.Plo@am__quote@ @AMDEP_TRUE@@am__include@ @am__quote@./$(DEPDIR)/n1fv_6.Plo@am__quote@ @AMDEP_TRUE@@am__include@ @am__quote@./$(DEPDIR)/n1fv_64.Plo@am__quote@ @AMDEP_TRUE@@am__include@ @am__quote@./$(DEPDIR)/n1fv_7.Plo@am__quote@ @AMDEP_TRUE@@am__include@ @am__quote@./$(DEPDIR)/n1fv_8.Plo@am__quote@ @AMDEP_TRUE@@am__include@ @am__quote@./$(DEPDIR)/n1fv_9.Plo@am__quote@ @AMDEP_TRUE@@am__include@ @am__quote@./$(DEPDIR)/n2bv_10.Plo@am__quote@ @AMDEP_TRUE@@am__include@ @am__quote@./$(DEPDIR)/n2bv_12.Plo@am__quote@ @AMDEP_TRUE@@am__include@ @am__quote@./$(DEPDIR)/n2bv_14.Plo@am__quote@ @AMDEP_TRUE@@am__include@ @am__quote@./$(DEPDIR)/n2bv_16.Plo@am__quote@ @AMDEP_TRUE@@am__include@ @am__quote@./$(DEPDIR)/n2bv_2.Plo@am__quote@ @AMDEP_TRUE@@am__include@ @am__quote@./$(DEPDIR)/n2bv_20.Plo@am__quote@ @AMDEP_TRUE@@am__include@ @am__quote@./$(DEPDIR)/n2bv_32.Plo@am__quote@ @AMDEP_TRUE@@am__include@ @am__quote@./$(DEPDIR)/n2bv_4.Plo@am__quote@ @AMDEP_TRUE@@am__include@ @am__quote@./$(DEPDIR)/n2bv_6.Plo@am__quote@ @AMDEP_TRUE@@am__include@ @am__quote@./$(DEPDIR)/n2bv_64.Plo@am__quote@ @AMDEP_TRUE@@am__include@ @am__quote@./$(DEPDIR)/n2bv_8.Plo@am__quote@ @AMDEP_TRUE@@am__include@ @am__quote@./$(DEPDIR)/n2fv_10.Plo@am__quote@ @AMDEP_TRUE@@am__include@ @am__quote@./$(DEPDIR)/n2fv_12.Plo@am__quote@ @AMDEP_TRUE@@am__include@ @am__quote@./$(DEPDIR)/n2fv_14.Plo@am__quote@ @AMDEP_TRUE@@am__include@ @am__quote@./$(DEPDIR)/n2fv_16.Plo@am__quote@ @AMDEP_TRUE@@am__include@ @am__quote@./$(DEPDIR)/n2fv_2.Plo@am__quote@ @AMDEP_TRUE@@am__include@ @am__quote@./$(DEPDIR)/n2fv_20.Plo@am__quote@ @AMDEP_TRUE@@am__include@ @am__quote@./$(DEPDIR)/n2fv_32.Plo@am__quote@ @AMDEP_TRUE@@am__include@ @am__quote@./$(DEPDIR)/n2fv_4.Plo@am__quote@ @AMDEP_TRUE@@am__include@ @am__quote@./$(DEPDIR)/n2fv_6.Plo@am__quote@ @AMDEP_TRUE@@am__include@ @am__quote@./$(DEPDIR)/n2fv_64.Plo@am__quote@ @AMDEP_TRUE@@am__include@ @am__quote@./$(DEPDIR)/n2fv_8.Plo@am__quote@ @AMDEP_TRUE@@am__include@ @am__quote@./$(DEPDIR)/n2sv_16.Plo@am__quote@ @AMDEP_TRUE@@am__include@ @am__quote@./$(DEPDIR)/n2sv_32.Plo@am__quote@ @AMDEP_TRUE@@am__include@ @am__quote@./$(DEPDIR)/n2sv_4.Plo@am__quote@ @AMDEP_TRUE@@am__include@ @am__quote@./$(DEPDIR)/n2sv_64.Plo@am__quote@ @AMDEP_TRUE@@am__include@ @am__quote@./$(DEPDIR)/n2sv_8.Plo@am__quote@ @AMDEP_TRUE@@am__include@ @am__quote@./$(DEPDIR)/q1bv_2.Plo@am__quote@ @AMDEP_TRUE@@am__include@ @am__quote@./$(DEPDIR)/q1bv_4.Plo@am__quote@ @AMDEP_TRUE@@am__include@ @am__quote@./$(DEPDIR)/q1bv_5.Plo@am__quote@ @AMDEP_TRUE@@am__include@ @am__quote@./$(DEPDIR)/q1bv_8.Plo@am__quote@ @AMDEP_TRUE@@am__include@ @am__quote@./$(DEPDIR)/q1fv_2.Plo@am__quote@ @AMDEP_TRUE@@am__include@ @am__quote@./$(DEPDIR)/q1fv_4.Plo@am__quote@ @AMDEP_TRUE@@am__include@ @am__quote@./$(DEPDIR)/q1fv_5.Plo@am__quote@ @AMDEP_TRUE@@am__include@ @am__quote@./$(DEPDIR)/q1fv_8.Plo@am__quote@ @AMDEP_TRUE@@am__include@ @am__quote@./$(DEPDIR)/t1buv_10.Plo@am__quote@ @AMDEP_TRUE@@am__include@ @am__quote@./$(DEPDIR)/t1buv_2.Plo@am__quote@ @AMDEP_TRUE@@am__include@ @am__quote@./$(DEPDIR)/t1buv_3.Plo@am__quote@ @AMDEP_TRUE@@am__include@ @am__quote@./$(DEPDIR)/t1buv_4.Plo@am__quote@ @AMDEP_TRUE@@am__include@ @am__quote@./$(DEPDIR)/t1buv_5.Plo@am__quote@ @AMDEP_TRUE@@am__include@ @am__quote@./$(DEPDIR)/t1buv_6.Plo@am__quote@ @AMDEP_TRUE@@am__include@ @am__quote@./$(DEPDIR)/t1buv_7.Plo@am__quote@ @AMDEP_TRUE@@am__include@ @am__quote@./$(DEPDIR)/t1buv_8.Plo@am__quote@ @AMDEP_TRUE@@am__include@ @am__quote@./$(DEPDIR)/t1buv_9.Plo@am__quote@ @AMDEP_TRUE@@am__include@ @am__quote@./$(DEPDIR)/t1bv_10.Plo@am__quote@ @AMDEP_TRUE@@am__include@ @am__quote@./$(DEPDIR)/t1bv_12.Plo@am__quote@ @AMDEP_TRUE@@am__include@ @am__quote@./$(DEPDIR)/t1bv_15.Plo@am__quote@ @AMDEP_TRUE@@am__include@ @am__quote@./$(DEPDIR)/t1bv_16.Plo@am__quote@ @AMDEP_TRUE@@am__include@ @am__quote@./$(DEPDIR)/t1bv_2.Plo@am__quote@ @AMDEP_TRUE@@am__include@ @am__quote@./$(DEPDIR)/t1bv_20.Plo@am__quote@ @AMDEP_TRUE@@am__include@ @am__quote@./$(DEPDIR)/t1bv_25.Plo@am__quote@ @AMDEP_TRUE@@am__include@ @am__quote@./$(DEPDIR)/t1bv_3.Plo@am__quote@ @AMDEP_TRUE@@am__include@ @am__quote@./$(DEPDIR)/t1bv_32.Plo@am__quote@ @AMDEP_TRUE@@am__include@ @am__quote@./$(DEPDIR)/t1bv_4.Plo@am__quote@ @AMDEP_TRUE@@am__include@ @am__quote@./$(DEPDIR)/t1bv_5.Plo@am__quote@ @AMDEP_TRUE@@am__include@ @am__quote@./$(DEPDIR)/t1bv_6.Plo@am__quote@ @AMDEP_TRUE@@am__include@ @am__quote@./$(DEPDIR)/t1bv_64.Plo@am__quote@ @AMDEP_TRUE@@am__include@ @am__quote@./$(DEPDIR)/t1bv_7.Plo@am__quote@ @AMDEP_TRUE@@am__include@ @am__quote@./$(DEPDIR)/t1bv_8.Plo@am__quote@ @AMDEP_TRUE@@am__include@ @am__quote@./$(DEPDIR)/t1bv_9.Plo@am__quote@ @AMDEP_TRUE@@am__include@ @am__quote@./$(DEPDIR)/t1fuv_10.Plo@am__quote@ @AMDEP_TRUE@@am__include@ @am__quote@./$(DEPDIR)/t1fuv_2.Plo@am__quote@ @AMDEP_TRUE@@am__include@ @am__quote@./$(DEPDIR)/t1fuv_3.Plo@am__quote@ @AMDEP_TRUE@@am__include@ @am__quote@./$(DEPDIR)/t1fuv_4.Plo@am__quote@ @AMDEP_TRUE@@am__include@ @am__quote@./$(DEPDIR)/t1fuv_5.Plo@am__quote@ @AMDEP_TRUE@@am__include@ @am__quote@./$(DEPDIR)/t1fuv_6.Plo@am__quote@ @AMDEP_TRUE@@am__include@ @am__quote@./$(DEPDIR)/t1fuv_7.Plo@am__quote@ @AMDEP_TRUE@@am__include@ @am__quote@./$(DEPDIR)/t1fuv_8.Plo@am__quote@ @AMDEP_TRUE@@am__include@ @am__quote@./$(DEPDIR)/t1fuv_9.Plo@am__quote@ @AMDEP_TRUE@@am__include@ @am__quote@./$(DEPDIR)/t1fv_10.Plo@am__quote@ @AMDEP_TRUE@@am__include@ @am__quote@./$(DEPDIR)/t1fv_12.Plo@am__quote@ @AMDEP_TRUE@@am__include@ @am__quote@./$(DEPDIR)/t1fv_15.Plo@am__quote@ @AMDEP_TRUE@@am__include@ @am__quote@./$(DEPDIR)/t1fv_16.Plo@am__quote@ @AMDEP_TRUE@@am__include@ @am__quote@./$(DEPDIR)/t1fv_2.Plo@am__quote@ @AMDEP_TRUE@@am__include@ @am__quote@./$(DEPDIR)/t1fv_20.Plo@am__quote@ @AMDEP_TRUE@@am__include@ @am__quote@./$(DEPDIR)/t1fv_25.Plo@am__quote@ @AMDEP_TRUE@@am__include@ @am__quote@./$(DEPDIR)/t1fv_3.Plo@am__quote@ @AMDEP_TRUE@@am__include@ @am__quote@./$(DEPDIR)/t1fv_32.Plo@am__quote@ @AMDEP_TRUE@@am__include@ @am__quote@./$(DEPDIR)/t1fv_4.Plo@am__quote@ @AMDEP_TRUE@@am__include@ @am__quote@./$(DEPDIR)/t1fv_5.Plo@am__quote@ @AMDEP_TRUE@@am__include@ @am__quote@./$(DEPDIR)/t1fv_6.Plo@am__quote@ @AMDEP_TRUE@@am__include@ @am__quote@./$(DEPDIR)/t1fv_64.Plo@am__quote@ @AMDEP_TRUE@@am__include@ @am__quote@./$(DEPDIR)/t1fv_7.Plo@am__quote@ @AMDEP_TRUE@@am__include@ @am__quote@./$(DEPDIR)/t1fv_8.Plo@am__quote@ @AMDEP_TRUE@@am__include@ @am__quote@./$(DEPDIR)/t1fv_9.Plo@am__quote@ @AMDEP_TRUE@@am__include@ @am__quote@./$(DEPDIR)/t1sv_16.Plo@am__quote@ @AMDEP_TRUE@@am__include@ @am__quote@./$(DEPDIR)/t1sv_2.Plo@am__quote@ @AMDEP_TRUE@@am__include@ @am__quote@./$(DEPDIR)/t1sv_32.Plo@am__quote@ @AMDEP_TRUE@@am__include@ @am__quote@./$(DEPDIR)/t1sv_4.Plo@am__quote@ @AMDEP_TRUE@@am__include@ @am__quote@./$(DEPDIR)/t1sv_8.Plo@am__quote@ @AMDEP_TRUE@@am__include@ @am__quote@./$(DEPDIR)/t2bv_10.Plo@am__quote@ @AMDEP_TRUE@@am__include@ @am__quote@./$(DEPDIR)/t2bv_16.Plo@am__quote@ @AMDEP_TRUE@@am__include@ @am__quote@./$(DEPDIR)/t2bv_2.Plo@am__quote@ @AMDEP_TRUE@@am__include@ @am__quote@./$(DEPDIR)/t2bv_20.Plo@am__quote@ @AMDEP_TRUE@@am__include@ @am__quote@./$(DEPDIR)/t2bv_25.Plo@am__quote@ @AMDEP_TRUE@@am__include@ @am__quote@./$(DEPDIR)/t2bv_32.Plo@am__quote@ @AMDEP_TRUE@@am__include@ @am__quote@./$(DEPDIR)/t2bv_4.Plo@am__quote@ @AMDEP_TRUE@@am__include@ @am__quote@./$(DEPDIR)/t2bv_5.Plo@am__quote@ @AMDEP_TRUE@@am__include@ @am__quote@./$(DEPDIR)/t2bv_64.Plo@am__quote@ @AMDEP_TRUE@@am__include@ @am__quote@./$(DEPDIR)/t2bv_8.Plo@am__quote@ @AMDEP_TRUE@@am__include@ @am__quote@./$(DEPDIR)/t2fv_10.Plo@am__quote@ @AMDEP_TRUE@@am__include@ @am__quote@./$(DEPDIR)/t2fv_16.Plo@am__quote@ @AMDEP_TRUE@@am__include@ @am__quote@./$(DEPDIR)/t2fv_2.Plo@am__quote@ @AMDEP_TRUE@@am__include@ @am__quote@./$(DEPDIR)/t2fv_20.Plo@am__quote@ @AMDEP_TRUE@@am__include@ @am__quote@./$(DEPDIR)/t2fv_25.Plo@am__quote@ @AMDEP_TRUE@@am__include@ @am__quote@./$(DEPDIR)/t2fv_32.Plo@am__quote@ @AMDEP_TRUE@@am__include@ @am__quote@./$(DEPDIR)/t2fv_4.Plo@am__quote@ @AMDEP_TRUE@@am__include@ @am__quote@./$(DEPDIR)/t2fv_5.Plo@am__quote@ @AMDEP_TRUE@@am__include@ @am__quote@./$(DEPDIR)/t2fv_64.Plo@am__quote@ @AMDEP_TRUE@@am__include@ @am__quote@./$(DEPDIR)/t2fv_8.Plo@am__quote@ @AMDEP_TRUE@@am__include@ @am__quote@./$(DEPDIR)/t2sv_16.Plo@am__quote@ @AMDEP_TRUE@@am__include@ @am__quote@./$(DEPDIR)/t2sv_32.Plo@am__quote@ @AMDEP_TRUE@@am__include@ @am__quote@./$(DEPDIR)/t2sv_4.Plo@am__quote@ @AMDEP_TRUE@@am__include@ @am__quote@./$(DEPDIR)/t2sv_8.Plo@am__quote@ @AMDEP_TRUE@@am__include@ @am__quote@./$(DEPDIR)/t3bv_10.Plo@am__quote@ @AMDEP_TRUE@@am__include@ @am__quote@./$(DEPDIR)/t3bv_16.Plo@am__quote@ @AMDEP_TRUE@@am__include@ @am__quote@./$(DEPDIR)/t3bv_20.Plo@am__quote@ @AMDEP_TRUE@@am__include@ @am__quote@./$(DEPDIR)/t3bv_25.Plo@am__quote@ @AMDEP_TRUE@@am__include@ @am__quote@./$(DEPDIR)/t3bv_32.Plo@am__quote@ @AMDEP_TRUE@@am__include@ @am__quote@./$(DEPDIR)/t3bv_4.Plo@am__quote@ @AMDEP_TRUE@@am__include@ @am__quote@./$(DEPDIR)/t3bv_5.Plo@am__quote@ @AMDEP_TRUE@@am__include@ @am__quote@./$(DEPDIR)/t3bv_8.Plo@am__quote@ @AMDEP_TRUE@@am__include@ @am__quote@./$(DEPDIR)/t3fv_10.Plo@am__quote@ @AMDEP_TRUE@@am__include@ @am__quote@./$(DEPDIR)/t3fv_16.Plo@am__quote@ @AMDEP_TRUE@@am__include@ @am__quote@./$(DEPDIR)/t3fv_20.Plo@am__quote@ @AMDEP_TRUE@@am__include@ @am__quote@./$(DEPDIR)/t3fv_25.Plo@am__quote@ @AMDEP_TRUE@@am__include@ @am__quote@./$(DEPDIR)/t3fv_32.Plo@am__quote@ @AMDEP_TRUE@@am__include@ @am__quote@./$(DEPDIR)/t3fv_4.Plo@am__quote@ @AMDEP_TRUE@@am__include@ @am__quote@./$(DEPDIR)/t3fv_5.Plo@am__quote@ @AMDEP_TRUE@@am__include@ @am__quote@./$(DEPDIR)/t3fv_8.Plo@am__quote@ .c.o: @am__fastdepCC_TRUE@ $(COMPILE) -MT $@ -MD -MP -MF $(DEPDIR)/$*.Tpo -c -o $@ $< @am__fastdepCC_TRUE@ $(am__mv) $(DEPDIR)/$*.Tpo $(DEPDIR)/$*.Po @AMDEP_TRUE@@am__fastdepCC_FALSE@ source='$<' object='$@' libtool=no @AMDEPBACKSLASH@ @AMDEP_TRUE@@am__fastdepCC_FALSE@ DEPDIR=$(DEPDIR) $(CCDEPMODE) $(depcomp) @AMDEPBACKSLASH@ @am__fastdepCC_FALSE@ $(COMPILE) -c $< .c.obj: @am__fastdepCC_TRUE@ $(COMPILE) -MT $@ -MD -MP -MF $(DEPDIR)/$*.Tpo -c -o $@ `$(CYGPATH_W) '$<'` @am__fastdepCC_TRUE@ $(am__mv) $(DEPDIR)/$*.Tpo $(DEPDIR)/$*.Po @AMDEP_TRUE@@am__fastdepCC_FALSE@ source='$<' object='$@' libtool=no @AMDEPBACKSLASH@ @AMDEP_TRUE@@am__fastdepCC_FALSE@ DEPDIR=$(DEPDIR) $(CCDEPMODE) $(depcomp) @AMDEPBACKSLASH@ @am__fastdepCC_FALSE@ $(COMPILE) -c `$(CYGPATH_W) '$<'` .c.lo: @am__fastdepCC_TRUE@ $(LTCOMPILE) -MT $@ -MD -MP -MF $(DEPDIR)/$*.Tpo -c -o $@ $< @am__fastdepCC_TRUE@ $(am__mv) $(DEPDIR)/$*.Tpo $(DEPDIR)/$*.Plo @AMDEP_TRUE@@am__fastdepCC_FALSE@ source='$<' object='$@' libtool=yes @AMDEPBACKSLASH@ @AMDEP_TRUE@@am__fastdepCC_FALSE@ DEPDIR=$(DEPDIR) $(CCDEPMODE) $(depcomp) @AMDEPBACKSLASH@ @am__fastdepCC_FALSE@ $(LTCOMPILE) -c -o $@ $< mostlyclean-libtool: -rm -f *.lo clean-libtool: -rm -rf .libs _libs ID: $(HEADERS) $(SOURCES) $(LISP) $(TAGS_FILES) list='$(SOURCES) $(HEADERS) $(LISP) $(TAGS_FILES)'; \ unique=`for i in $$list; do \ if test -f "$$i"; then echo $$i; else echo $(srcdir)/$$i; fi; \ done | \ $(AWK) '{ files[$$0] = 1; nonempty = 1; } \ END { if (nonempty) { for (i in files) print i; }; }'`; \ mkid -fID $$unique tags: TAGS TAGS: $(HEADERS) $(SOURCES) $(TAGS_DEPENDENCIES) \ $(TAGS_FILES) $(LISP) set x; \ here=`pwd`; \ list='$(SOURCES) $(HEADERS) $(LISP) $(TAGS_FILES)'; \ unique=`for i in $$list; do \ if test -f "$$i"; then echo $$i; else echo $(srcdir)/$$i; fi; \ done | \ $(AWK) '{ files[$$0] = 1; nonempty = 1; } \ END { if (nonempty) { for (i in files) print i; }; }'`; \ shift; \ if test -z "$(ETAGS_ARGS)$$*$$unique"; then :; else \ test -n "$$unique" || unique=$$empty_fix; \ if test $$# -gt 0; then \ $(ETAGS) $(ETAGSFLAGS) $(AM_ETAGSFLAGS) $(ETAGS_ARGS) \ "$$@" $$unique; \ else \ $(ETAGS) $(ETAGSFLAGS) $(AM_ETAGSFLAGS) $(ETAGS_ARGS) \ $$unique; \ fi; \ fi ctags: CTAGS CTAGS: $(HEADERS) $(SOURCES) $(TAGS_DEPENDENCIES) \ $(TAGS_FILES) $(LISP) list='$(SOURCES) $(HEADERS) $(LISP) $(TAGS_FILES)'; \ unique=`for i in $$list; do \ if test -f "$$i"; then echo $$i; else echo $(srcdir)/$$i; fi; \ done | \ $(AWK) '{ files[$$0] = 1; nonempty = 1; } \ END { if (nonempty) { for (i in files) print i; }; }'`; \ test -z "$(CTAGS_ARGS)$$unique" \ || $(CTAGS) $(CTAGSFLAGS) $(AM_CTAGSFLAGS) $(CTAGS_ARGS) \ $$unique GTAGS: here=`$(am__cd) $(top_builddir) && pwd` \ && $(am__cd) $(top_srcdir) \ && gtags -i $(GTAGS_ARGS) "$$here" distclean-tags: -rm -f TAGS ID GTAGS GRTAGS GSYMS GPATH tags distdir: $(DISTFILES) @srcdirstrip=`echo "$(srcdir)" | sed 's/[].[^$$\\*]/\\\\&/g'`; \ topsrcdirstrip=`echo "$(top_srcdir)" | sed 's/[].[^$$\\*]/\\\\&/g'`; \ list='$(DISTFILES)'; \ dist_files=`for file in $$list; do echo $$file; done | \ sed -e "s|^$$srcdirstrip/||;t" \ -e "s|^$$topsrcdirstrip/|$(top_builddir)/|;t"`; \ case $$dist_files in \ */*) $(MKDIR_P) `echo "$$dist_files" | \ sed '/\//!d;s|^|$(distdir)/|;s,/[^/]*$$,,' | \ sort -u` ;; \ esac; \ for file in $$dist_files; do \ if test -f $$file || test -d $$file; then d=.; else d=$(srcdir); fi; \ if test -d $$d/$$file; then \ dir=`echo "/$$file" | sed -e 's,/[^/]*$$,,'`; \ if test -d "$(distdir)/$$file"; then \ find "$(distdir)/$$file" -type d ! -perm -700 -exec chmod u+rwx {} \;; \ fi; \ if test -d $(srcdir)/$$file && test $$d != $(srcdir); then \ cp -fpR $(srcdir)/$$file "$(distdir)$$dir" || exit 1; \ find "$(distdir)/$$file" -type d ! -perm -700 -exec chmod u+rwx {} \;; \ fi; \ cp -fpR $$d/$$file "$(distdir)$$dir" || exit 1; \ else \ test -f "$(distdir)/$$file" \ || cp -p $$d/$$file "$(distdir)/$$file" \ || exit 1; \ fi; \ done check-am: all-am check: $(BUILT_SOURCES) $(MAKE) $(AM_MAKEFLAGS) check-am all-am: Makefile $(LTLIBRARIES) installdirs: install: $(BUILT_SOURCES) $(MAKE) $(AM_MAKEFLAGS) install-am install-exec: install-exec-am install-data: install-data-am uninstall: uninstall-am install-am: all-am @$(MAKE) $(AM_MAKEFLAGS) install-exec-am install-data-am installcheck: installcheck-am install-strip: if test -z '$(STRIP)'; then \ $(MAKE) $(AM_MAKEFLAGS) INSTALL_PROGRAM="$(INSTALL_STRIP_PROGRAM)" \ install_sh_PROGRAM="$(INSTALL_STRIP_PROGRAM)" INSTALL_STRIP_FLAG=-s \ install; \ else \ $(MAKE) $(AM_MAKEFLAGS) INSTALL_PROGRAM="$(INSTALL_STRIP_PROGRAM)" \ install_sh_PROGRAM="$(INSTALL_STRIP_PROGRAM)" INSTALL_STRIP_FLAG=-s \ "INSTALL_PROGRAM_ENV=STRIPPROG='$(STRIP)'" install; \ fi mostlyclean-generic: clean-generic: distclean-generic: -test -z "$(CONFIG_CLEAN_FILES)" || rm -f $(CONFIG_CLEAN_FILES) -test . = "$(srcdir)" || test -z "$(CONFIG_CLEAN_VPATH_FILES)" || rm -f $(CONFIG_CLEAN_VPATH_FILES) maintainer-clean-generic: @echo "This command is intended for maintainers to use" @echo "it deletes files that may require special tools to rebuild." -test -z "$(BUILT_SOURCES)" || rm -f $(BUILT_SOURCES) clean: clean-am clean-am: clean-generic clean-libtool clean-noinstLTLIBRARIES \ mostlyclean-am distclean: distclean-am -rm -rf ./$(DEPDIR) -rm -f Makefile distclean-am: clean-am distclean-compile distclean-generic \ distclean-tags dvi: dvi-am dvi-am: html: html-am html-am: info: info-am info-am: install-data-am: install-dvi: install-dvi-am install-dvi-am: install-exec-am: install-html: install-html-am install-html-am: install-info: install-info-am install-info-am: install-man: install-pdf: install-pdf-am install-pdf-am: install-ps: install-ps-am install-ps-am: installcheck-am: maintainer-clean: maintainer-clean-am -rm -rf ./$(DEPDIR) -rm -f Makefile maintainer-clean-am: distclean-am maintainer-clean-generic mostlyclean: mostlyclean-am mostlyclean-am: mostlyclean-compile mostlyclean-generic \ mostlyclean-libtool pdf: pdf-am pdf-am: ps: ps-am ps-am: uninstall-am: .MAKE: all check install install-am install-strip .PHONY: CTAGS GTAGS all all-am check check-am clean clean-generic \ clean-libtool clean-noinstLTLIBRARIES ctags distclean \ distclean-compile distclean-generic distclean-libtool \ distclean-tags distdir dvi dvi-am html html-am info info-am \ install install-am install-data install-data-am install-dvi \ install-dvi-am install-exec install-exec-am install-html \ install-html-am install-info install-info-am install-man \ install-pdf install-pdf-am install-ps install-ps-am \ install-strip installcheck installcheck-am installdirs \ maintainer-clean maintainer-clean-generic mostlyclean \ mostlyclean-compile mostlyclean-generic mostlyclean-libtool \ pdf pdf-am ps ps-am tags uninstall uninstall-am $(EXTRA_DIST): Makefile ( \ echo "/* Generated automatically. DO NOT EDIT! */"; \ echo "#define SIMD_HEADER \"$(SIMD_HEADER)\""; \ echo "#include \"../common/"$*".c\""; \ ) >$@ # Tell versions [3.59,3.63) of GNU make to not export all variables. # Otherwise a system limit (for SysV at least) may be exceeded. .NOEXPORT: fftw-3.3.3/dft/simd/sse2/n1fv_2.c0000644000175000017500000000015512054410611013226 00000000000000/* Generated automatically. DO NOT EDIT! */ #define SIMD_HEADER "simd-sse2.h" #include "../common/n1fv_2.c" fftw-3.3.3/dft/simd/sse2/n1fv_3.c0000644000175000017500000000015512054410611013227 00000000000000/* Generated automatically. DO NOT EDIT! */ #define SIMD_HEADER "simd-sse2.h" #include "../common/n1fv_3.c" fftw-3.3.3/dft/simd/sse2/n1fv_4.c0000644000175000017500000000015512054410611013230 00000000000000/* Generated automatically. DO NOT EDIT! */ #define SIMD_HEADER "simd-sse2.h" #include "../common/n1fv_4.c" fftw-3.3.3/dft/simd/sse2/n1fv_5.c0000644000175000017500000000015512054410611013231 00000000000000/* Generated automatically. DO NOT EDIT! */ #define SIMD_HEADER "simd-sse2.h" #include "../common/n1fv_5.c" fftw-3.3.3/dft/simd/sse2/n1fv_6.c0000644000175000017500000000015512054410611013232 00000000000000/* Generated automatically. DO NOT EDIT! */ #define SIMD_HEADER "simd-sse2.h" #include "../common/n1fv_6.c" fftw-3.3.3/dft/simd/sse2/n1fv_7.c0000644000175000017500000000015512054410611013233 00000000000000/* Generated automatically. DO NOT EDIT! */ #define SIMD_HEADER "simd-sse2.h" #include "../common/n1fv_7.c" fftw-3.3.3/dft/simd/sse2/n1fv_8.c0000644000175000017500000000015512054410611013234 00000000000000/* Generated automatically. DO NOT EDIT! */ #define SIMD_HEADER "simd-sse2.h" #include "../common/n1fv_8.c" fftw-3.3.3/dft/simd/sse2/n1fv_9.c0000644000175000017500000000015512054410611013235 00000000000000/* Generated automatically. DO NOT EDIT! */ #define SIMD_HEADER "simd-sse2.h" #include "../common/n1fv_9.c" fftw-3.3.3/dft/simd/sse2/n1fv_10.c0000644000175000017500000000015612054410611013306 00000000000000/* Generated automatically. DO NOT EDIT! */ #define SIMD_HEADER "simd-sse2.h" #include "../common/n1fv_10.c" fftw-3.3.3/dft/simd/sse2/n1fv_11.c0000644000175000017500000000015612054410611013307 00000000000000/* Generated automatically. DO NOT EDIT! */ #define SIMD_HEADER "simd-sse2.h" #include "../common/n1fv_11.c" fftw-3.3.3/dft/simd/sse2/n1fv_12.c0000644000175000017500000000015612054410611013310 00000000000000/* Generated automatically. DO NOT EDIT! */ #define SIMD_HEADER "simd-sse2.h" #include "../common/n1fv_12.c" fftw-3.3.3/dft/simd/sse2/n1fv_13.c0000644000175000017500000000015612054410611013311 00000000000000/* Generated automatically. DO NOT EDIT! */ #define SIMD_HEADER "simd-sse2.h" #include "../common/n1fv_13.c" fftw-3.3.3/dft/simd/sse2/n1fv_14.c0000644000175000017500000000015612054410611013312 00000000000000/* Generated automatically. DO NOT EDIT! */ #define SIMD_HEADER "simd-sse2.h" #include "../common/n1fv_14.c" fftw-3.3.3/dft/simd/sse2/n1fv_15.c0000644000175000017500000000015612054410611013313 00000000000000/* Generated automatically. DO NOT EDIT! */ #define SIMD_HEADER "simd-sse2.h" #include "../common/n1fv_15.c" fftw-3.3.3/dft/simd/sse2/n1fv_16.c0000644000175000017500000000015612054410611013314 00000000000000/* Generated automatically. DO NOT EDIT! */ #define SIMD_HEADER "simd-sse2.h" #include "../common/n1fv_16.c" fftw-3.3.3/dft/simd/sse2/n1fv_32.c0000644000175000017500000000015612054410611013312 00000000000000/* Generated automatically. DO NOT EDIT! */ #define SIMD_HEADER "simd-sse2.h" #include "../common/n1fv_32.c" fftw-3.3.3/dft/simd/sse2/n1fv_64.c0000644000175000017500000000015612054410611013317 00000000000000/* Generated automatically. DO NOT EDIT! */ #define SIMD_HEADER "simd-sse2.h" #include "../common/n1fv_64.c" fftw-3.3.3/dft/simd/sse2/n1fv_128.c0000644000175000017500000000015712054410611013401 00000000000000/* Generated automatically. DO NOT EDIT! */ #define SIMD_HEADER "simd-sse2.h" #include "../common/n1fv_128.c" fftw-3.3.3/dft/simd/sse2/n1fv_20.c0000644000175000017500000000015612054410611013307 00000000000000/* Generated automatically. DO NOT EDIT! */ #define SIMD_HEADER "simd-sse2.h" #include "../common/n1fv_20.c" fftw-3.3.3/dft/simd/sse2/n1fv_25.c0000644000175000017500000000015612054410611013314 00000000000000/* Generated automatically. DO NOT EDIT! */ #define SIMD_HEADER "simd-sse2.h" #include "../common/n1fv_25.c" fftw-3.3.3/dft/simd/sse2/n1bv_2.c0000644000175000017500000000015512054410611013222 00000000000000/* Generated automatically. DO NOT EDIT! */ #define SIMD_HEADER "simd-sse2.h" #include "../common/n1bv_2.c" fftw-3.3.3/dft/simd/sse2/n1bv_3.c0000644000175000017500000000015512054410611013223 00000000000000/* Generated automatically. DO NOT EDIT! */ #define SIMD_HEADER "simd-sse2.h" #include "../common/n1bv_3.c" fftw-3.3.3/dft/simd/sse2/n1bv_4.c0000644000175000017500000000015512054410611013224 00000000000000/* Generated automatically. DO NOT EDIT! */ #define SIMD_HEADER "simd-sse2.h" #include "../common/n1bv_4.c" fftw-3.3.3/dft/simd/sse2/n1bv_5.c0000644000175000017500000000015512054410611013225 00000000000000/* Generated automatically. DO NOT EDIT! */ #define SIMD_HEADER "simd-sse2.h" #include "../common/n1bv_5.c" fftw-3.3.3/dft/simd/sse2/n1bv_6.c0000644000175000017500000000015512054410611013226 00000000000000/* Generated automatically. DO NOT EDIT! */ #define SIMD_HEADER "simd-sse2.h" #include "../common/n1bv_6.c" fftw-3.3.3/dft/simd/sse2/n1bv_7.c0000644000175000017500000000015512054410611013227 00000000000000/* Generated automatically. DO NOT EDIT! */ #define SIMD_HEADER "simd-sse2.h" #include "../common/n1bv_7.c" fftw-3.3.3/dft/simd/sse2/n1bv_8.c0000644000175000017500000000015512054410611013230 00000000000000/* Generated automatically. DO NOT EDIT! */ #define SIMD_HEADER "simd-sse2.h" #include "../common/n1bv_8.c" fftw-3.3.3/dft/simd/sse2/n1bv_9.c0000644000175000017500000000015512054410611013231 00000000000000/* Generated automatically. DO NOT EDIT! */ #define SIMD_HEADER "simd-sse2.h" #include "../common/n1bv_9.c" fftw-3.3.3/dft/simd/sse2/n1bv_10.c0000644000175000017500000000015612054410611013302 00000000000000/* Generated automatically. DO NOT EDIT! */ #define SIMD_HEADER "simd-sse2.h" #include "../common/n1bv_10.c" fftw-3.3.3/dft/simd/sse2/n1bv_11.c0000644000175000017500000000015612054410611013303 00000000000000/* Generated automatically. DO NOT EDIT! */ #define SIMD_HEADER "simd-sse2.h" #include "../common/n1bv_11.c" fftw-3.3.3/dft/simd/sse2/n1bv_12.c0000644000175000017500000000015612054410611013304 00000000000000/* Generated automatically. DO NOT EDIT! */ #define SIMD_HEADER "simd-sse2.h" #include "../common/n1bv_12.c" fftw-3.3.3/dft/simd/sse2/n1bv_13.c0000644000175000017500000000015612054410611013305 00000000000000/* Generated automatically. DO NOT EDIT! */ #define SIMD_HEADER "simd-sse2.h" #include "../common/n1bv_13.c" fftw-3.3.3/dft/simd/sse2/n1bv_14.c0000644000175000017500000000015612054410611013306 00000000000000/* Generated automatically. DO NOT EDIT! */ #define SIMD_HEADER "simd-sse2.h" #include "../common/n1bv_14.c" fftw-3.3.3/dft/simd/sse2/n1bv_15.c0000644000175000017500000000015612054410611013307 00000000000000/* Generated automatically. DO NOT EDIT! */ #define SIMD_HEADER "simd-sse2.h" #include "../common/n1bv_15.c" fftw-3.3.3/dft/simd/sse2/n1bv_16.c0000644000175000017500000000015612054410611013310 00000000000000/* Generated automatically. DO NOT EDIT! */ #define SIMD_HEADER "simd-sse2.h" #include "../common/n1bv_16.c" fftw-3.3.3/dft/simd/sse2/n1bv_32.c0000644000175000017500000000015612054410611013306 00000000000000/* Generated automatically. DO NOT EDIT! */ #define SIMD_HEADER "simd-sse2.h" #include "../common/n1bv_32.c" fftw-3.3.3/dft/simd/sse2/n1bv_64.c0000644000175000017500000000015612054410611013313 00000000000000/* Generated automatically. DO NOT EDIT! */ #define SIMD_HEADER "simd-sse2.h" #include "../common/n1bv_64.c" fftw-3.3.3/dft/simd/sse2/n1bv_128.c0000644000175000017500000000015712054410611013375 00000000000000/* Generated automatically. DO NOT EDIT! */ #define SIMD_HEADER "simd-sse2.h" #include "../common/n1bv_128.c" fftw-3.3.3/dft/simd/sse2/n1bv_20.c0000644000175000017500000000015612054410611013303 00000000000000/* Generated automatically. DO NOT EDIT! */ #define SIMD_HEADER "simd-sse2.h" #include "../common/n1bv_20.c" fftw-3.3.3/dft/simd/sse2/n1bv_25.c0000644000175000017500000000015612054410611013310 00000000000000/* Generated automatically. DO NOT EDIT! */ #define SIMD_HEADER "simd-sse2.h" #include "../common/n1bv_25.c" fftw-3.3.3/dft/simd/sse2/n2fv_2.c0000644000175000017500000000015512054410611013227 00000000000000/* Generated automatically. DO NOT EDIT! */ #define SIMD_HEADER "simd-sse2.h" #include "../common/n2fv_2.c" fftw-3.3.3/dft/simd/sse2/n2fv_4.c0000644000175000017500000000015512054410611013231 00000000000000/* Generated automatically. DO NOT EDIT! */ #define SIMD_HEADER "simd-sse2.h" #include "../common/n2fv_4.c" fftw-3.3.3/dft/simd/sse2/n2fv_6.c0000644000175000017500000000015512054410611013233 00000000000000/* Generated automatically. DO NOT EDIT! */ #define SIMD_HEADER "simd-sse2.h" #include "../common/n2fv_6.c" fftw-3.3.3/dft/simd/sse2/n2fv_8.c0000644000175000017500000000015512054410611013235 00000000000000/* Generated automatically. DO NOT EDIT! */ #define SIMD_HEADER "simd-sse2.h" #include "../common/n2fv_8.c" fftw-3.3.3/dft/simd/sse2/n2fv_10.c0000644000175000017500000000015612054410611013307 00000000000000/* Generated automatically. DO NOT EDIT! */ #define SIMD_HEADER "simd-sse2.h" #include "../common/n2fv_10.c" fftw-3.3.3/dft/simd/sse2/n2fv_12.c0000644000175000017500000000015612054410611013311 00000000000000/* Generated automatically. DO NOT EDIT! */ #define SIMD_HEADER "simd-sse2.h" #include "../common/n2fv_12.c" fftw-3.3.3/dft/simd/sse2/n2fv_14.c0000644000175000017500000000015612054410611013313 00000000000000/* Generated automatically. DO NOT EDIT! */ #define SIMD_HEADER "simd-sse2.h" #include "../common/n2fv_14.c" fftw-3.3.3/dft/simd/sse2/n2fv_16.c0000644000175000017500000000015612054410611013315 00000000000000/* Generated automatically. DO NOT EDIT! */ #define SIMD_HEADER "simd-sse2.h" #include "../common/n2fv_16.c" fftw-3.3.3/dft/simd/sse2/n2fv_32.c0000644000175000017500000000015612054410611013313 00000000000000/* Generated automatically. DO NOT EDIT! */ #define SIMD_HEADER "simd-sse2.h" #include "../common/n2fv_32.c" fftw-3.3.3/dft/simd/sse2/n2fv_64.c0000644000175000017500000000015612054410611013320 00000000000000/* Generated automatically. DO NOT EDIT! */ #define SIMD_HEADER "simd-sse2.h" #include "../common/n2fv_64.c" fftw-3.3.3/dft/simd/sse2/n2fv_20.c0000644000175000017500000000015612054410611013310 00000000000000/* Generated automatically. DO NOT EDIT! */ #define SIMD_HEADER "simd-sse2.h" #include "../common/n2fv_20.c" fftw-3.3.3/dft/simd/sse2/n2bv_2.c0000644000175000017500000000015512054410611013223 00000000000000/* Generated automatically. DO NOT EDIT! */ #define SIMD_HEADER "simd-sse2.h" #include "../common/n2bv_2.c" fftw-3.3.3/dft/simd/sse2/n2bv_4.c0000644000175000017500000000015512054410611013225 00000000000000/* Generated automatically. DO NOT EDIT! */ #define SIMD_HEADER "simd-sse2.h" #include "../common/n2bv_4.c" fftw-3.3.3/dft/simd/sse2/n2bv_6.c0000644000175000017500000000015512054410611013227 00000000000000/* Generated automatically. DO NOT EDIT! */ #define SIMD_HEADER "simd-sse2.h" #include "../common/n2bv_6.c" fftw-3.3.3/dft/simd/sse2/n2bv_8.c0000644000175000017500000000015512054410611013231 00000000000000/* Generated automatically. DO NOT EDIT! */ #define SIMD_HEADER "simd-sse2.h" #include "../common/n2bv_8.c" fftw-3.3.3/dft/simd/sse2/n2bv_10.c0000644000175000017500000000015612054410611013303 00000000000000/* Generated automatically. DO NOT EDIT! */ #define SIMD_HEADER "simd-sse2.h" #include "../common/n2bv_10.c" fftw-3.3.3/dft/simd/sse2/n2bv_12.c0000644000175000017500000000015612054410611013305 00000000000000/* Generated automatically. DO NOT EDIT! */ #define SIMD_HEADER "simd-sse2.h" #include "../common/n2bv_12.c" fftw-3.3.3/dft/simd/sse2/n2bv_14.c0000644000175000017500000000015612054410611013307 00000000000000/* Generated automatically. DO NOT EDIT! */ #define SIMD_HEADER "simd-sse2.h" #include "../common/n2bv_14.c" fftw-3.3.3/dft/simd/sse2/n2bv_16.c0000644000175000017500000000015612054410611013311 00000000000000/* Generated automatically. DO NOT EDIT! */ #define SIMD_HEADER "simd-sse2.h" #include "../common/n2bv_16.c" fftw-3.3.3/dft/simd/sse2/n2bv_32.c0000644000175000017500000000015612054410611013307 00000000000000/* Generated automatically. DO NOT EDIT! */ #define SIMD_HEADER "simd-sse2.h" #include "../common/n2bv_32.c" fftw-3.3.3/dft/simd/sse2/n2bv_64.c0000644000175000017500000000015612054410611013314 00000000000000/* Generated automatically. DO NOT EDIT! */ #define SIMD_HEADER "simd-sse2.h" #include "../common/n2bv_64.c" fftw-3.3.3/dft/simd/sse2/n2bv_20.c0000644000175000017500000000015612054410611013304 00000000000000/* Generated automatically. DO NOT EDIT! */ #define SIMD_HEADER "simd-sse2.h" #include "../common/n2bv_20.c" fftw-3.3.3/dft/simd/sse2/n2sv_4.c0000644000175000017500000000015512054410611013246 00000000000000/* Generated automatically. DO NOT EDIT! */ #define SIMD_HEADER "simd-sse2.h" #include "../common/n2sv_4.c" fftw-3.3.3/dft/simd/sse2/n2sv_8.c0000644000175000017500000000015512054410611013252 00000000000000/* Generated automatically. DO NOT EDIT! */ #define SIMD_HEADER "simd-sse2.h" #include "../common/n2sv_8.c" fftw-3.3.3/dft/simd/sse2/n2sv_16.c0000644000175000017500000000015612054410611013332 00000000000000/* Generated automatically. DO NOT EDIT! */ #define SIMD_HEADER "simd-sse2.h" #include "../common/n2sv_16.c" fftw-3.3.3/dft/simd/sse2/n2sv_32.c0000644000175000017500000000015612054410611013330 00000000000000/* Generated automatically. DO NOT EDIT! */ #define SIMD_HEADER "simd-sse2.h" #include "../common/n2sv_32.c" fftw-3.3.3/dft/simd/sse2/n2sv_64.c0000644000175000017500000000015612054410611013335 00000000000000/* Generated automatically. DO NOT EDIT! */ #define SIMD_HEADER "simd-sse2.h" #include "../common/n2sv_64.c" fftw-3.3.3/dft/simd/sse2/t1fuv_2.c0000644000175000017500000000015612054410611013422 00000000000000/* Generated automatically. DO NOT EDIT! */ #define SIMD_HEADER "simd-sse2.h" #include "../common/t1fuv_2.c" fftw-3.3.3/dft/simd/sse2/t1fuv_3.c0000644000175000017500000000015612054410611013423 00000000000000/* Generated automatically. DO NOT EDIT! */ #define SIMD_HEADER "simd-sse2.h" #include "../common/t1fuv_3.c" fftw-3.3.3/dft/simd/sse2/t1fuv_4.c0000644000175000017500000000015612054410611013424 00000000000000/* Generated automatically. DO NOT EDIT! */ #define SIMD_HEADER "simd-sse2.h" #include "../common/t1fuv_4.c" fftw-3.3.3/dft/simd/sse2/t1fuv_5.c0000644000175000017500000000015612054410611013425 00000000000000/* Generated automatically. DO NOT EDIT! */ #define SIMD_HEADER "simd-sse2.h" #include "../common/t1fuv_5.c" fftw-3.3.3/dft/simd/sse2/t1fuv_6.c0000644000175000017500000000015612054410611013426 00000000000000/* Generated automatically. DO NOT EDIT! */ #define SIMD_HEADER "simd-sse2.h" #include "../common/t1fuv_6.c" fftw-3.3.3/dft/simd/sse2/t1fuv_7.c0000644000175000017500000000015612054410611013427 00000000000000/* Generated automatically. DO NOT EDIT! */ #define SIMD_HEADER "simd-sse2.h" #include "../common/t1fuv_7.c" fftw-3.3.3/dft/simd/sse2/t1fuv_8.c0000644000175000017500000000015612054410611013430 00000000000000/* Generated automatically. DO NOT EDIT! */ #define SIMD_HEADER "simd-sse2.h" #include "../common/t1fuv_8.c" fftw-3.3.3/dft/simd/sse2/t1fuv_9.c0000644000175000017500000000015612054410611013431 00000000000000/* Generated automatically. DO NOT EDIT! */ #define SIMD_HEADER "simd-sse2.h" #include "../common/t1fuv_9.c" fftw-3.3.3/dft/simd/sse2/t1fuv_10.c0000644000175000017500000000015712054410611013502 00000000000000/* Generated automatically. DO NOT EDIT! */ #define SIMD_HEADER "simd-sse2.h" #include "../common/t1fuv_10.c" fftw-3.3.3/dft/simd/sse2/t1fv_2.c0000644000175000017500000000015512054410611013234 00000000000000/* Generated automatically. DO NOT EDIT! */ #define SIMD_HEADER "simd-sse2.h" #include "../common/t1fv_2.c" fftw-3.3.3/dft/simd/sse2/t1fv_3.c0000644000175000017500000000015512054410611013235 00000000000000/* Generated automatically. DO NOT EDIT! */ #define SIMD_HEADER "simd-sse2.h" #include "../common/t1fv_3.c" fftw-3.3.3/dft/simd/sse2/t1fv_4.c0000644000175000017500000000015512054410611013236 00000000000000/* Generated automatically. DO NOT EDIT! */ #define SIMD_HEADER "simd-sse2.h" #include "../common/t1fv_4.c" fftw-3.3.3/dft/simd/sse2/t1fv_5.c0000644000175000017500000000015512054410611013237 00000000000000/* Generated automatically. DO NOT EDIT! */ #define SIMD_HEADER "simd-sse2.h" #include "../common/t1fv_5.c" fftw-3.3.3/dft/simd/sse2/t1fv_6.c0000644000175000017500000000015512054410611013240 00000000000000/* Generated automatically. DO NOT EDIT! */ #define SIMD_HEADER "simd-sse2.h" #include "../common/t1fv_6.c" fftw-3.3.3/dft/simd/sse2/t1fv_7.c0000644000175000017500000000015512054410611013241 00000000000000/* Generated automatically. DO NOT EDIT! */ #define SIMD_HEADER "simd-sse2.h" #include "../common/t1fv_7.c" fftw-3.3.3/dft/simd/sse2/t1fv_8.c0000644000175000017500000000015512054410611013242 00000000000000/* Generated automatically. DO NOT EDIT! */ #define SIMD_HEADER "simd-sse2.h" #include "../common/t1fv_8.c" fftw-3.3.3/dft/simd/sse2/t1fv_9.c0000644000175000017500000000015512054410611013243 00000000000000/* Generated automatically. DO NOT EDIT! */ #define SIMD_HEADER "simd-sse2.h" #include "../common/t1fv_9.c" fftw-3.3.3/dft/simd/sse2/t1fv_10.c0000644000175000017500000000015612054410611013314 00000000000000/* Generated automatically. DO NOT EDIT! */ #define SIMD_HEADER "simd-sse2.h" #include "../common/t1fv_10.c" fftw-3.3.3/dft/simd/sse2/t1fv_12.c0000644000175000017500000000015612054410611013316 00000000000000/* Generated automatically. DO NOT EDIT! */ #define SIMD_HEADER "simd-sse2.h" #include "../common/t1fv_12.c" fftw-3.3.3/dft/simd/sse2/t1fv_15.c0000644000175000017500000000015612054410611013321 00000000000000/* Generated automatically. DO NOT EDIT! */ #define SIMD_HEADER "simd-sse2.h" #include "../common/t1fv_15.c" fftw-3.3.3/dft/simd/sse2/t1fv_16.c0000644000175000017500000000015612054410611013322 00000000000000/* Generated automatically. DO NOT EDIT! */ #define SIMD_HEADER "simd-sse2.h" #include "../common/t1fv_16.c" fftw-3.3.3/dft/simd/sse2/t1fv_32.c0000644000175000017500000000015612054410611013320 00000000000000/* Generated automatically. DO NOT EDIT! */ #define SIMD_HEADER "simd-sse2.h" #include "../common/t1fv_32.c" fftw-3.3.3/dft/simd/sse2/t1fv_64.c0000644000175000017500000000015612054410611013325 00000000000000/* Generated automatically. DO NOT EDIT! */ #define SIMD_HEADER "simd-sse2.h" #include "../common/t1fv_64.c" fftw-3.3.3/dft/simd/sse2/t1fv_20.c0000644000175000017500000000015612054410611013315 00000000000000/* Generated automatically. DO NOT EDIT! */ #define SIMD_HEADER "simd-sse2.h" #include "../common/t1fv_20.c" fftw-3.3.3/dft/simd/sse2/t1fv_25.c0000644000175000017500000000015612054410611013322 00000000000000/* Generated automatically. DO NOT EDIT! */ #define SIMD_HEADER "simd-sse2.h" #include "../common/t1fv_25.c" fftw-3.3.3/dft/simd/sse2/t2fv_2.c0000644000175000017500000000015512054410611013235 00000000000000/* Generated automatically. DO NOT EDIT! */ #define SIMD_HEADER "simd-sse2.h" #include "../common/t2fv_2.c" fftw-3.3.3/dft/simd/sse2/t2fv_4.c0000644000175000017500000000015512054410611013237 00000000000000/* Generated automatically. DO NOT EDIT! */ #define SIMD_HEADER "simd-sse2.h" #include "../common/t2fv_4.c" fftw-3.3.3/dft/simd/sse2/t2fv_8.c0000644000175000017500000000015512054410611013243 00000000000000/* Generated automatically. DO NOT EDIT! */ #define SIMD_HEADER "simd-sse2.h" #include "../common/t2fv_8.c" fftw-3.3.3/dft/simd/sse2/t2fv_16.c0000644000175000017500000000015612054410611013323 00000000000000/* Generated automatically. DO NOT EDIT! */ #define SIMD_HEADER "simd-sse2.h" #include "../common/t2fv_16.c" fftw-3.3.3/dft/simd/sse2/t2fv_32.c0000644000175000017500000000015612054410611013321 00000000000000/* Generated automatically. DO NOT EDIT! */ #define SIMD_HEADER "simd-sse2.h" #include "../common/t2fv_32.c" fftw-3.3.3/dft/simd/sse2/t2fv_64.c0000644000175000017500000000015612054410611013326 00000000000000/* Generated automatically. DO NOT EDIT! */ #define SIMD_HEADER "simd-sse2.h" #include "../common/t2fv_64.c" fftw-3.3.3/dft/simd/sse2/t2fv_5.c0000644000175000017500000000015512054410611013240 00000000000000/* Generated automatically. DO NOT EDIT! */ #define SIMD_HEADER "simd-sse2.h" #include "../common/t2fv_5.c" fftw-3.3.3/dft/simd/sse2/t2fv_10.c0000644000175000017500000000015612054410611013315 00000000000000/* Generated automatically. DO NOT EDIT! */ #define SIMD_HEADER "simd-sse2.h" #include "../common/t2fv_10.c" fftw-3.3.3/dft/simd/sse2/t2fv_20.c0000644000175000017500000000015612054410611013316 00000000000000/* Generated automatically. DO NOT EDIT! */ #define SIMD_HEADER "simd-sse2.h" #include "../common/t2fv_20.c" fftw-3.3.3/dft/simd/sse2/t2fv_25.c0000644000175000017500000000015612054410611013323 00000000000000/* Generated automatically. DO NOT EDIT! */ #define SIMD_HEADER "simd-sse2.h" #include "../common/t2fv_25.c" fftw-3.3.3/dft/simd/sse2/t3fv_4.c0000644000175000017500000000015512054410611013240 00000000000000/* Generated automatically. DO NOT EDIT! */ #define SIMD_HEADER "simd-sse2.h" #include "../common/t3fv_4.c" fftw-3.3.3/dft/simd/sse2/t3fv_8.c0000644000175000017500000000015512054410611013244 00000000000000/* Generated automatically. DO NOT EDIT! */ #define SIMD_HEADER "simd-sse2.h" #include "../common/t3fv_8.c" fftw-3.3.3/dft/simd/sse2/t3fv_16.c0000644000175000017500000000015612054410611013324 00000000000000/* Generated automatically. DO NOT EDIT! */ #define SIMD_HEADER "simd-sse2.h" #include "../common/t3fv_16.c" fftw-3.3.3/dft/simd/sse2/t3fv_32.c0000644000175000017500000000015612054410611013322 00000000000000/* Generated automatically. DO NOT EDIT! */ #define SIMD_HEADER "simd-sse2.h" #include "../common/t3fv_32.c" fftw-3.3.3/dft/simd/sse2/t3fv_5.c0000644000175000017500000000015512054410611013241 00000000000000/* Generated automatically. DO NOT EDIT! */ #define SIMD_HEADER "simd-sse2.h" #include "../common/t3fv_5.c" fftw-3.3.3/dft/simd/sse2/t3fv_10.c0000644000175000017500000000015612054410611013316 00000000000000/* Generated automatically. DO NOT EDIT! */ #define SIMD_HEADER "simd-sse2.h" #include "../common/t3fv_10.c" fftw-3.3.3/dft/simd/sse2/t3fv_20.c0000644000175000017500000000015612054410611013317 00000000000000/* Generated automatically. DO NOT EDIT! */ #define SIMD_HEADER "simd-sse2.h" #include "../common/t3fv_20.c" fftw-3.3.3/dft/simd/sse2/t3fv_25.c0000644000175000017500000000015612054410611013324 00000000000000/* Generated automatically. DO NOT EDIT! */ #define SIMD_HEADER "simd-sse2.h" #include "../common/t3fv_25.c" fftw-3.3.3/dft/simd/sse2/t1buv_2.c0000644000175000017500000000015612054410611013416 00000000000000/* Generated automatically. DO NOT EDIT! */ #define SIMD_HEADER "simd-sse2.h" #include "../common/t1buv_2.c" fftw-3.3.3/dft/simd/sse2/t1buv_3.c0000644000175000017500000000015612054410611013417 00000000000000/* Generated automatically. DO NOT EDIT! */ #define SIMD_HEADER "simd-sse2.h" #include "../common/t1buv_3.c" fftw-3.3.3/dft/simd/sse2/t1buv_4.c0000644000175000017500000000015612054410611013420 00000000000000/* Generated automatically. DO NOT EDIT! */ #define SIMD_HEADER "simd-sse2.h" #include "../common/t1buv_4.c" fftw-3.3.3/dft/simd/sse2/t1buv_5.c0000644000175000017500000000015612054410611013421 00000000000000/* Generated automatically. DO NOT EDIT! */ #define SIMD_HEADER "simd-sse2.h" #include "../common/t1buv_5.c" fftw-3.3.3/dft/simd/sse2/t1buv_6.c0000644000175000017500000000015612054410611013422 00000000000000/* Generated automatically. DO NOT EDIT! */ #define SIMD_HEADER "simd-sse2.h" #include "../common/t1buv_6.c" fftw-3.3.3/dft/simd/sse2/t1buv_7.c0000644000175000017500000000015612054410611013423 00000000000000/* Generated automatically. DO NOT EDIT! */ #define SIMD_HEADER "simd-sse2.h" #include "../common/t1buv_7.c" fftw-3.3.3/dft/simd/sse2/t1buv_8.c0000644000175000017500000000015612054410611013424 00000000000000/* Generated automatically. DO NOT EDIT! */ #define SIMD_HEADER "simd-sse2.h" #include "../common/t1buv_8.c" fftw-3.3.3/dft/simd/sse2/t1buv_9.c0000644000175000017500000000015612054410611013425 00000000000000/* Generated automatically. DO NOT EDIT! */ #define SIMD_HEADER "simd-sse2.h" #include "../common/t1buv_9.c" fftw-3.3.3/dft/simd/sse2/t1buv_10.c0000644000175000017500000000015712054410611013476 00000000000000/* Generated automatically. DO NOT EDIT! */ #define SIMD_HEADER "simd-sse2.h" #include "../common/t1buv_10.c" fftw-3.3.3/dft/simd/sse2/t1bv_2.c0000644000175000017500000000015512054410611013230 00000000000000/* Generated automatically. DO NOT EDIT! */ #define SIMD_HEADER "simd-sse2.h" #include "../common/t1bv_2.c" fftw-3.3.3/dft/simd/sse2/t1bv_3.c0000644000175000017500000000015512054410611013231 00000000000000/* Generated automatically. DO NOT EDIT! */ #define SIMD_HEADER "simd-sse2.h" #include "../common/t1bv_3.c" fftw-3.3.3/dft/simd/sse2/t1bv_4.c0000644000175000017500000000015512054410611013232 00000000000000/* Generated automatically. DO NOT EDIT! */ #define SIMD_HEADER "simd-sse2.h" #include "../common/t1bv_4.c" fftw-3.3.3/dft/simd/sse2/t1bv_5.c0000644000175000017500000000015512054410611013233 00000000000000/* Generated automatically. DO NOT EDIT! */ #define SIMD_HEADER "simd-sse2.h" #include "../common/t1bv_5.c" fftw-3.3.3/dft/simd/sse2/t1bv_6.c0000644000175000017500000000015512054410611013234 00000000000000/* Generated automatically. DO NOT EDIT! */ #define SIMD_HEADER "simd-sse2.h" #include "../common/t1bv_6.c" fftw-3.3.3/dft/simd/sse2/t1bv_7.c0000644000175000017500000000015512054410611013235 00000000000000/* Generated automatically. DO NOT EDIT! */ #define SIMD_HEADER "simd-sse2.h" #include "../common/t1bv_7.c" fftw-3.3.3/dft/simd/sse2/t1bv_8.c0000644000175000017500000000015512054410611013236 00000000000000/* Generated automatically. DO NOT EDIT! */ #define SIMD_HEADER "simd-sse2.h" #include "../common/t1bv_8.c" fftw-3.3.3/dft/simd/sse2/t1bv_9.c0000644000175000017500000000015512054410611013237 00000000000000/* Generated automatically. DO NOT EDIT! */ #define SIMD_HEADER "simd-sse2.h" #include "../common/t1bv_9.c" fftw-3.3.3/dft/simd/sse2/t1bv_10.c0000644000175000017500000000015612054410611013310 00000000000000/* Generated automatically. DO NOT EDIT! */ #define SIMD_HEADER "simd-sse2.h" #include "../common/t1bv_10.c" fftw-3.3.3/dft/simd/sse2/t1bv_12.c0000644000175000017500000000015612054410611013312 00000000000000/* Generated automatically. DO NOT EDIT! */ #define SIMD_HEADER "simd-sse2.h" #include "../common/t1bv_12.c" fftw-3.3.3/dft/simd/sse2/t1bv_15.c0000644000175000017500000000015612054410611013315 00000000000000/* Generated automatically. DO NOT EDIT! */ #define SIMD_HEADER "simd-sse2.h" #include "../common/t1bv_15.c" fftw-3.3.3/dft/simd/sse2/t1bv_16.c0000644000175000017500000000015612054410611013316 00000000000000/* Generated automatically. DO NOT EDIT! */ #define SIMD_HEADER "simd-sse2.h" #include "../common/t1bv_16.c" fftw-3.3.3/dft/simd/sse2/t1bv_32.c0000644000175000017500000000015612054410611013314 00000000000000/* Generated automatically. DO NOT EDIT! */ #define SIMD_HEADER "simd-sse2.h" #include "../common/t1bv_32.c" fftw-3.3.3/dft/simd/sse2/t1bv_64.c0000644000175000017500000000015612054410611013321 00000000000000/* Generated automatically. DO NOT EDIT! */ #define SIMD_HEADER "simd-sse2.h" #include "../common/t1bv_64.c" fftw-3.3.3/dft/simd/sse2/t1bv_20.c0000644000175000017500000000015612054410611013311 00000000000000/* Generated automatically. DO NOT EDIT! */ #define SIMD_HEADER "simd-sse2.h" #include "../common/t1bv_20.c" fftw-3.3.3/dft/simd/sse2/t1bv_25.c0000644000175000017500000000015612054410611013316 00000000000000/* Generated automatically. DO NOT EDIT! */ #define SIMD_HEADER "simd-sse2.h" #include "../common/t1bv_25.c" fftw-3.3.3/dft/simd/sse2/t2bv_2.c0000644000175000017500000000015512054410611013231 00000000000000/* Generated automatically. DO NOT EDIT! */ #define SIMD_HEADER "simd-sse2.h" #include "../common/t2bv_2.c" fftw-3.3.3/dft/simd/sse2/t2bv_4.c0000644000175000017500000000015512054410611013233 00000000000000/* Generated automatically. DO NOT EDIT! */ #define SIMD_HEADER "simd-sse2.h" #include "../common/t2bv_4.c" fftw-3.3.3/dft/simd/sse2/t2bv_8.c0000644000175000017500000000015512054410611013237 00000000000000/* Generated automatically. DO NOT EDIT! */ #define SIMD_HEADER "simd-sse2.h" #include "../common/t2bv_8.c" fftw-3.3.3/dft/simd/sse2/t2bv_16.c0000644000175000017500000000015612054410611013317 00000000000000/* Generated automatically. DO NOT EDIT! */ #define SIMD_HEADER "simd-sse2.h" #include "../common/t2bv_16.c" fftw-3.3.3/dft/simd/sse2/t2bv_32.c0000644000175000017500000000015612054410611013315 00000000000000/* Generated automatically. DO NOT EDIT! */ #define SIMD_HEADER "simd-sse2.h" #include "../common/t2bv_32.c" fftw-3.3.3/dft/simd/sse2/t2bv_64.c0000644000175000017500000000015612054410611013322 00000000000000/* Generated automatically. DO NOT EDIT! */ #define SIMD_HEADER "simd-sse2.h" #include "../common/t2bv_64.c" fftw-3.3.3/dft/simd/sse2/t2bv_5.c0000644000175000017500000000015512054410611013234 00000000000000/* Generated automatically. DO NOT EDIT! */ #define SIMD_HEADER "simd-sse2.h" #include "../common/t2bv_5.c" fftw-3.3.3/dft/simd/sse2/t2bv_10.c0000644000175000017500000000015612054410611013311 00000000000000/* Generated automatically. DO NOT EDIT! */ #define SIMD_HEADER "simd-sse2.h" #include "../common/t2bv_10.c" fftw-3.3.3/dft/simd/sse2/t2bv_20.c0000644000175000017500000000015612054410611013312 00000000000000/* Generated automatically. DO NOT EDIT! */ #define SIMD_HEADER "simd-sse2.h" #include "../common/t2bv_20.c" fftw-3.3.3/dft/simd/sse2/t2bv_25.c0000644000175000017500000000015612054410611013317 00000000000000/* Generated automatically. DO NOT EDIT! */ #define SIMD_HEADER "simd-sse2.h" #include "../common/t2bv_25.c" fftw-3.3.3/dft/simd/sse2/t3bv_4.c0000644000175000017500000000015512054410611013234 00000000000000/* Generated automatically. DO NOT EDIT! */ #define SIMD_HEADER "simd-sse2.h" #include "../common/t3bv_4.c" fftw-3.3.3/dft/simd/sse2/t3bv_8.c0000644000175000017500000000015512054410611013240 00000000000000/* Generated automatically. DO NOT EDIT! */ #define SIMD_HEADER "simd-sse2.h" #include "../common/t3bv_8.c" fftw-3.3.3/dft/simd/sse2/t3bv_16.c0000644000175000017500000000015612054410611013320 00000000000000/* Generated automatically. DO NOT EDIT! */ #define SIMD_HEADER "simd-sse2.h" #include "../common/t3bv_16.c" fftw-3.3.3/dft/simd/sse2/t3bv_32.c0000644000175000017500000000015612054410611013316 00000000000000/* Generated automatically. DO NOT EDIT! */ #define SIMD_HEADER "simd-sse2.h" #include "../common/t3bv_32.c" fftw-3.3.3/dft/simd/sse2/t3bv_5.c0000644000175000017500000000015512054410611013235 00000000000000/* Generated automatically. DO NOT EDIT! */ #define SIMD_HEADER "simd-sse2.h" #include "../common/t3bv_5.c" fftw-3.3.3/dft/simd/sse2/t3bv_10.c0000644000175000017500000000015612054410611013312 00000000000000/* Generated automatically. DO NOT EDIT! */ #define SIMD_HEADER "simd-sse2.h" #include "../common/t3bv_10.c" fftw-3.3.3/dft/simd/sse2/t3bv_20.c0000644000175000017500000000015612054410611013313 00000000000000/* Generated automatically. DO NOT EDIT! */ #define SIMD_HEADER "simd-sse2.h" #include "../common/t3bv_20.c" fftw-3.3.3/dft/simd/sse2/t3bv_25.c0000644000175000017500000000015612054410611013320 00000000000000/* Generated automatically. DO NOT EDIT! */ #define SIMD_HEADER "simd-sse2.h" #include "../common/t3bv_25.c" fftw-3.3.3/dft/simd/sse2/t1sv_2.c0000644000175000017500000000015512054410611013251 00000000000000/* Generated automatically. DO NOT EDIT! */ #define SIMD_HEADER "simd-sse2.h" #include "../common/t1sv_2.c" fftw-3.3.3/dft/simd/sse2/t1sv_4.c0000644000175000017500000000015512054410611013253 00000000000000/* Generated automatically. DO NOT EDIT! */ #define SIMD_HEADER "simd-sse2.h" #include "../common/t1sv_4.c" fftw-3.3.3/dft/simd/sse2/t1sv_8.c0000644000175000017500000000015512054410611013257 00000000000000/* Generated automatically. DO NOT EDIT! */ #define SIMD_HEADER "simd-sse2.h" #include "../common/t1sv_8.c" fftw-3.3.3/dft/simd/sse2/t1sv_16.c0000644000175000017500000000015612054410611013337 00000000000000/* Generated automatically. DO NOT EDIT! */ #define SIMD_HEADER "simd-sse2.h" #include "../common/t1sv_16.c" fftw-3.3.3/dft/simd/sse2/t1sv_32.c0000644000175000017500000000015612054410611013335 00000000000000/* Generated automatically. DO NOT EDIT! */ #define SIMD_HEADER "simd-sse2.h" #include "../common/t1sv_32.c" fftw-3.3.3/dft/simd/sse2/t2sv_4.c0000644000175000017500000000015512054410611013254 00000000000000/* Generated automatically. DO NOT EDIT! */ #define SIMD_HEADER "simd-sse2.h" #include "../common/t2sv_4.c" fftw-3.3.3/dft/simd/sse2/t2sv_8.c0000644000175000017500000000015512054410611013260 00000000000000/* Generated automatically. DO NOT EDIT! */ #define SIMD_HEADER "simd-sse2.h" #include "../common/t2sv_8.c" fftw-3.3.3/dft/simd/sse2/t2sv_16.c0000644000175000017500000000015612054410611013340 00000000000000/* Generated automatically. DO NOT EDIT! */ #define SIMD_HEADER "simd-sse2.h" #include "../common/t2sv_16.c" fftw-3.3.3/dft/simd/sse2/t2sv_32.c0000644000175000017500000000015612054410611013336 00000000000000/* Generated automatically. DO NOT EDIT! */ #define SIMD_HEADER "simd-sse2.h" #include "../common/t2sv_32.c" fftw-3.3.3/dft/simd/sse2/q1fv_2.c0000644000175000017500000000015512054410611013231 00000000000000/* Generated automatically. DO NOT EDIT! */ #define SIMD_HEADER "simd-sse2.h" #include "../common/q1fv_2.c" fftw-3.3.3/dft/simd/sse2/q1fv_4.c0000644000175000017500000000015512054410611013233 00000000000000/* Generated automatically. DO NOT EDIT! */ #define SIMD_HEADER "simd-sse2.h" #include "../common/q1fv_4.c" fftw-3.3.3/dft/simd/sse2/q1fv_5.c0000644000175000017500000000015512054410611013234 00000000000000/* Generated automatically. DO NOT EDIT! */ #define SIMD_HEADER "simd-sse2.h" #include "../common/q1fv_5.c" fftw-3.3.3/dft/simd/sse2/q1fv_8.c0000644000175000017500000000015512054410611013237 00000000000000/* Generated automatically. DO NOT EDIT! */ #define SIMD_HEADER "simd-sse2.h" #include "../common/q1fv_8.c" fftw-3.3.3/dft/simd/sse2/q1bv_2.c0000644000175000017500000000015512054410611013225 00000000000000/* Generated automatically. DO NOT EDIT! */ #define SIMD_HEADER "simd-sse2.h" #include "../common/q1bv_2.c" fftw-3.3.3/dft/simd/sse2/q1bv_4.c0000644000175000017500000000015512054410611013227 00000000000000/* Generated automatically. DO NOT EDIT! */ #define SIMD_HEADER "simd-sse2.h" #include "../common/q1bv_4.c" fftw-3.3.3/dft/simd/sse2/q1bv_5.c0000644000175000017500000000015512054410611013230 00000000000000/* Generated automatically. DO NOT EDIT! */ #define SIMD_HEADER "simd-sse2.h" #include "../common/q1bv_5.c" fftw-3.3.3/dft/simd/sse2/q1bv_8.c0000644000175000017500000000015512054410611013233 00000000000000/* Generated automatically. DO NOT EDIT! */ #define SIMD_HEADER "simd-sse2.h" #include "../common/q1bv_8.c" fftw-3.3.3/dft/simd/sse2/genus.c0000644000175000017500000000015412054410611013253 00000000000000/* Generated automatically. DO NOT EDIT! */ #define SIMD_HEADER "simd-sse2.h" #include "../common/genus.c" fftw-3.3.3/dft/simd/sse2/codlist.c0000644000175000017500000000015612054410611013575 00000000000000/* Generated automatically. DO NOT EDIT! */ #define SIMD_HEADER "simd-sse2.h" #include "../common/codlist.c" fftw-3.3.3/dft/simd/avx/0000755000175000017500000000000012054411107011771 500000000000000fftw-3.3.3/dft/simd/avx/Makefile.am0000644000175000017500000000042412054410050013741 00000000000000AM_CFLAGS = $(AVX_CFLAGS) SIMD_HEADER=simd-avx.h include $(top_srcdir)/dft/simd/codlist.mk include $(top_srcdir)/dft/simd/simd.mk if HAVE_AVX BUILT_SOURCES = $(EXTRA_DIST) noinst_LTLIBRARIES = libdft_avx_codelets.la libdft_avx_codelets_la_SOURCES = $(BUILT_SOURCES) endif fftw-3.3.3/dft/simd/avx/Makefile.in0000644000175000017500000011106612054410177013771 00000000000000# Makefile.in generated by automake 1.11.6 from Makefile.am. # @configure_input@ # Copyright (C) 1994, 1995, 1996, 1997, 1998, 1999, 2000, 2001, 2002, # 2003, 2004, 2005, 2006, 2007, 2008, 2009, 2010, 2011 Free Software # Foundation, Inc. # This Makefile.in is free software; the Free Software Foundation # gives unlimited permission to copy and/or distribute it, # with or without modifications, as long as this notice is preserved. # This program is distributed in the hope that it will be useful, # but WITHOUT ANY WARRANTY, to the extent permitted by law; without # even the implied warranty of MERCHANTABILITY or FITNESS FOR A # PARTICULAR PURPOSE. @SET_MAKE@ # This file contains a standard list of DFT SIMD codelets. It is # included by common/Makefile to generate the C files with the actual # codelets in them. It is included by {sse,sse2,...}/Makefile to # generate and compile stub files that include common/*.c # You can customize FFTW for special needs, e.g. to handle certain # sizes more efficiently, by adding new codelets to the lists of those # included by default. If you change the list of codelets, any new # ones you added will be automatically generated when you run the # bootstrap script (see "Generating your own code" in the FFTW # manual). VPATH = @srcdir@ am__make_dryrun = \ { \ am__dry=no; \ case $$MAKEFLAGS in \ *\\[\ \ ]*) \ echo 'am--echo: ; @echo "AM" OK' | $(MAKE) -f - 2>/dev/null \ | grep '^AM OK$$' >/dev/null || am__dry=yes;; \ *) \ for am__flg in $$MAKEFLAGS; do \ case $$am__flg in \ *=*|--*) ;; \ *n*) am__dry=yes; break;; \ esac; \ done;; \ esac; \ test $$am__dry = yes; \ } pkgdatadir = $(datadir)/@PACKAGE@ pkgincludedir = $(includedir)/@PACKAGE@ pkglibdir = $(libdir)/@PACKAGE@ pkglibexecdir = $(libexecdir)/@PACKAGE@ am__cd = CDPATH="$${ZSH_VERSION+.}$(PATH_SEPARATOR)" && cd install_sh_DATA = $(install_sh) -c -m 644 install_sh_PROGRAM = $(install_sh) -c install_sh_SCRIPT = $(install_sh) -c INSTALL_HEADER = $(INSTALL_DATA) transform = $(program_transform_name) NORMAL_INSTALL = : PRE_INSTALL = : POST_INSTALL = : NORMAL_UNINSTALL = : PRE_UNINSTALL = : POST_UNINSTALL = : build_triplet = @build@ host_triplet = @host@ DIST_COMMON = $(srcdir)/Makefile.am $(srcdir)/Makefile.in \ $(top_srcdir)/dft/simd/codlist.mk \ $(top_srcdir)/dft/simd/simd.mk subdir = dft/simd/avx ACLOCAL_M4 = $(top_srcdir)/aclocal.m4 am__aclocal_m4_deps = $(top_srcdir)/m4/acx_mpi.m4 \ $(top_srcdir)/m4/acx_pthread.m4 \ $(top_srcdir)/m4/ax_cc_maxopt.m4 \ $(top_srcdir)/m4/ax_check_compiler_flags.m4 \ $(top_srcdir)/m4/ax_compiler_vendor.m4 \ $(top_srcdir)/m4/ax_gcc_aligns_stack.m4 \ $(top_srcdir)/m4/ax_gcc_version.m4 \ $(top_srcdir)/m4/ax_openmp.m4 $(top_srcdir)/m4/libtool.m4 \ $(top_srcdir)/m4/ltoptions.m4 $(top_srcdir)/m4/ltsugar.m4 \ $(top_srcdir)/m4/ltversion.m4 $(top_srcdir)/m4/lt~obsolete.m4 \ $(top_srcdir)/configure.ac am__configure_deps = $(am__aclocal_m4_deps) $(CONFIGURE_DEPENDENCIES) \ $(ACLOCAL_M4) mkinstalldirs = $(install_sh) -d CONFIG_HEADER = $(top_builddir)/config.h CONFIG_CLEAN_FILES = CONFIG_CLEAN_VPATH_FILES = LTLIBRARIES = $(noinst_LTLIBRARIES) libdft_avx_codelets_la_LIBADD = am__libdft_avx_codelets_la_SOURCES_DIST = n1fv_2.c n1fv_3.c n1fv_4.c \ n1fv_5.c n1fv_6.c n1fv_7.c n1fv_8.c n1fv_9.c n1fv_10.c \ n1fv_11.c n1fv_12.c n1fv_13.c n1fv_14.c n1fv_15.c n1fv_16.c \ n1fv_32.c n1fv_64.c n1fv_128.c n1fv_20.c n1fv_25.c n1bv_2.c \ n1bv_3.c n1bv_4.c n1bv_5.c n1bv_6.c n1bv_7.c n1bv_8.c n1bv_9.c \ n1bv_10.c n1bv_11.c n1bv_12.c n1bv_13.c n1bv_14.c n1bv_15.c \ n1bv_16.c n1bv_32.c n1bv_64.c n1bv_128.c n1bv_20.c n1bv_25.c \ n2fv_2.c n2fv_4.c n2fv_6.c n2fv_8.c n2fv_10.c n2fv_12.c \ n2fv_14.c n2fv_16.c n2fv_32.c n2fv_64.c n2fv_20.c n2bv_2.c \ n2bv_4.c n2bv_6.c n2bv_8.c n2bv_10.c n2bv_12.c n2bv_14.c \ n2bv_16.c n2bv_32.c n2bv_64.c n2bv_20.c n2sv_4.c n2sv_8.c \ n2sv_16.c n2sv_32.c n2sv_64.c t1fuv_2.c t1fuv_3.c t1fuv_4.c \ t1fuv_5.c t1fuv_6.c t1fuv_7.c t1fuv_8.c t1fuv_9.c t1fuv_10.c \ t1fv_2.c t1fv_3.c t1fv_4.c t1fv_5.c t1fv_6.c t1fv_7.c t1fv_8.c \ t1fv_9.c t1fv_10.c t1fv_12.c t1fv_15.c t1fv_16.c t1fv_32.c \ t1fv_64.c t1fv_20.c t1fv_25.c t2fv_2.c t2fv_4.c t2fv_8.c \ t2fv_16.c t2fv_32.c t2fv_64.c t2fv_5.c t2fv_10.c t2fv_20.c \ t2fv_25.c t3fv_4.c t3fv_8.c t3fv_16.c t3fv_32.c t3fv_5.c \ t3fv_10.c t3fv_20.c t3fv_25.c t1buv_2.c t1buv_3.c t1buv_4.c \ t1buv_5.c t1buv_6.c t1buv_7.c t1buv_8.c t1buv_9.c t1buv_10.c \ t1bv_2.c t1bv_3.c t1bv_4.c t1bv_5.c t1bv_6.c t1bv_7.c t1bv_8.c \ t1bv_9.c t1bv_10.c t1bv_12.c t1bv_15.c t1bv_16.c t1bv_32.c \ t1bv_64.c t1bv_20.c t1bv_25.c t2bv_2.c t2bv_4.c t2bv_8.c \ t2bv_16.c t2bv_32.c t2bv_64.c t2bv_5.c t2bv_10.c t2bv_20.c \ t2bv_25.c t3bv_4.c t3bv_8.c t3bv_16.c t3bv_32.c t3bv_5.c \ t3bv_10.c t3bv_20.c t3bv_25.c t1sv_2.c t1sv_4.c t1sv_8.c \ t1sv_16.c t1sv_32.c t2sv_4.c t2sv_8.c t2sv_16.c t2sv_32.c \ q1fv_2.c q1fv_4.c q1fv_5.c q1fv_8.c q1bv_2.c q1bv_4.c q1bv_5.c \ q1bv_8.c genus.c codlist.c am__objects_1 = n1fv_2.lo n1fv_3.lo n1fv_4.lo n1fv_5.lo n1fv_6.lo \ n1fv_7.lo n1fv_8.lo n1fv_9.lo n1fv_10.lo n1fv_11.lo n1fv_12.lo \ n1fv_13.lo n1fv_14.lo n1fv_15.lo n1fv_16.lo n1fv_32.lo \ n1fv_64.lo n1fv_128.lo n1fv_20.lo n1fv_25.lo am__objects_2 = n1bv_2.lo n1bv_3.lo n1bv_4.lo n1bv_5.lo n1bv_6.lo \ n1bv_7.lo n1bv_8.lo n1bv_9.lo n1bv_10.lo n1bv_11.lo n1bv_12.lo \ n1bv_13.lo n1bv_14.lo n1bv_15.lo n1bv_16.lo n1bv_32.lo \ n1bv_64.lo n1bv_128.lo n1bv_20.lo n1bv_25.lo am__objects_3 = n2fv_2.lo n2fv_4.lo n2fv_6.lo n2fv_8.lo n2fv_10.lo \ n2fv_12.lo n2fv_14.lo n2fv_16.lo n2fv_32.lo n2fv_64.lo \ n2fv_20.lo am__objects_4 = n2bv_2.lo n2bv_4.lo n2bv_6.lo n2bv_8.lo n2bv_10.lo \ n2bv_12.lo n2bv_14.lo n2bv_16.lo n2bv_32.lo n2bv_64.lo \ n2bv_20.lo am__objects_5 = n2sv_4.lo n2sv_8.lo n2sv_16.lo n2sv_32.lo n2sv_64.lo am__objects_6 = t1fuv_2.lo t1fuv_3.lo t1fuv_4.lo t1fuv_5.lo t1fuv_6.lo \ t1fuv_7.lo t1fuv_8.lo t1fuv_9.lo t1fuv_10.lo am__objects_7 = t1fv_2.lo t1fv_3.lo t1fv_4.lo t1fv_5.lo t1fv_6.lo \ t1fv_7.lo t1fv_8.lo t1fv_9.lo t1fv_10.lo t1fv_12.lo t1fv_15.lo \ t1fv_16.lo t1fv_32.lo t1fv_64.lo t1fv_20.lo t1fv_25.lo am__objects_8 = t2fv_2.lo t2fv_4.lo t2fv_8.lo t2fv_16.lo t2fv_32.lo \ t2fv_64.lo t2fv_5.lo t2fv_10.lo t2fv_20.lo t2fv_25.lo am__objects_9 = t3fv_4.lo t3fv_8.lo t3fv_16.lo t3fv_32.lo t3fv_5.lo \ t3fv_10.lo t3fv_20.lo t3fv_25.lo am__objects_10 = t1buv_2.lo t1buv_3.lo t1buv_4.lo t1buv_5.lo \ t1buv_6.lo t1buv_7.lo t1buv_8.lo t1buv_9.lo t1buv_10.lo am__objects_11 = t1bv_2.lo t1bv_3.lo t1bv_4.lo t1bv_5.lo t1bv_6.lo \ t1bv_7.lo t1bv_8.lo t1bv_9.lo t1bv_10.lo t1bv_12.lo t1bv_15.lo \ t1bv_16.lo t1bv_32.lo t1bv_64.lo t1bv_20.lo t1bv_25.lo am__objects_12 = t2bv_2.lo t2bv_4.lo t2bv_8.lo t2bv_16.lo t2bv_32.lo \ t2bv_64.lo t2bv_5.lo t2bv_10.lo t2bv_20.lo t2bv_25.lo am__objects_13 = t3bv_4.lo t3bv_8.lo t3bv_16.lo t3bv_32.lo t3bv_5.lo \ t3bv_10.lo t3bv_20.lo t3bv_25.lo am__objects_14 = t1sv_2.lo t1sv_4.lo t1sv_8.lo t1sv_16.lo t1sv_32.lo am__objects_15 = t2sv_4.lo t2sv_8.lo t2sv_16.lo t2sv_32.lo am__objects_16 = q1fv_2.lo q1fv_4.lo q1fv_5.lo q1fv_8.lo am__objects_17 = q1bv_2.lo q1bv_4.lo q1bv_5.lo q1bv_8.lo am__objects_18 = $(am__objects_1) $(am__objects_2) $(am__objects_3) \ $(am__objects_4) $(am__objects_5) $(am__objects_6) \ $(am__objects_7) $(am__objects_8) $(am__objects_9) \ $(am__objects_10) $(am__objects_11) $(am__objects_12) \ $(am__objects_13) $(am__objects_14) $(am__objects_15) \ $(am__objects_16) $(am__objects_17) am__objects_19 = $(am__objects_18) genus.lo codlist.lo @HAVE_AVX_TRUE@am__objects_20 = $(am__objects_19) @HAVE_AVX_TRUE@am_libdft_avx_codelets_la_OBJECTS = $(am__objects_20) libdft_avx_codelets_la_OBJECTS = $(am_libdft_avx_codelets_la_OBJECTS) @HAVE_AVX_TRUE@am_libdft_avx_codelets_la_rpath = DEFAULT_INCLUDES = -I.@am__isrc@ -I$(top_builddir) depcomp = $(SHELL) $(top_srcdir)/depcomp am__depfiles_maybe = depfiles am__mv = mv -f COMPILE = $(CC) $(DEFS) $(DEFAULT_INCLUDES) $(INCLUDES) $(AM_CPPFLAGS) \ $(CPPFLAGS) $(AM_CFLAGS) $(CFLAGS) LTCOMPILE = $(LIBTOOL) --tag=CC $(AM_LIBTOOLFLAGS) $(LIBTOOLFLAGS) \ --mode=compile $(CC) $(DEFS) $(DEFAULT_INCLUDES) $(INCLUDES) \ $(AM_CPPFLAGS) $(CPPFLAGS) $(AM_CFLAGS) $(CFLAGS) CCLD = $(CC) LINK = $(LIBTOOL) --tag=CC $(AM_LIBTOOLFLAGS) $(LIBTOOLFLAGS) \ --mode=link $(CCLD) $(AM_CFLAGS) $(CFLAGS) $(AM_LDFLAGS) \ $(LDFLAGS) -o $@ SOURCES = $(libdft_avx_codelets_la_SOURCES) DIST_SOURCES = $(am__libdft_avx_codelets_la_SOURCES_DIST) am__can_run_installinfo = \ case $$AM_UPDATE_INFO_DIR in \ n|no|NO) false;; \ *) (install-info --version) >/dev/null 2>&1;; \ esac ETAGS = etags CTAGS = ctags DISTFILES = $(DIST_COMMON) $(DIST_SOURCES) $(TEXINFOS) $(EXTRA_DIST) ACLOCAL = @ACLOCAL@ ALLOCA = @ALLOCA@ ALTIVEC_CFLAGS = @ALTIVEC_CFLAGS@ AMTAR = @AMTAR@ AR = @AR@ AS = @AS@ AUTOCONF = @AUTOCONF@ AUTOHEADER = @AUTOHEADER@ AUTOMAKE = @AUTOMAKE@ AVX_CFLAGS = @AVX_CFLAGS@ AWK = @AWK@ CC = @CC@ CCDEPMODE = @CCDEPMODE@ CFLAGS = @CFLAGS@ CHECK_PL_OPTS = @CHECK_PL_OPTS@ CPP = @CPP@ CPPFLAGS = @CPPFLAGS@ CYGPATH_W = @CYGPATH_W@ C_FFTW_R2R_KIND = @C_FFTW_R2R_KIND@ C_MPI_FINT = @C_MPI_FINT@ DEFS = @DEFS@ DEPDIR = @DEPDIR@ DLLTOOL = @DLLTOOL@ DSYMUTIL = @DSYMUTIL@ DUMPBIN = @DUMPBIN@ ECHO_C = @ECHO_C@ ECHO_N = @ECHO_N@ ECHO_T = @ECHO_T@ EGREP = @EGREP@ EXEEXT = @EXEEXT@ F77 = @F77@ FFLAGS = @FFLAGS@ FGREP = @FGREP@ FLIBS = @FLIBS@ GREP = @GREP@ INSTALL = @INSTALL@ INSTALL_DATA = @INSTALL_DATA@ INSTALL_PROGRAM = @INSTALL_PROGRAM@ INSTALL_SCRIPT = @INSTALL_SCRIPT@ INSTALL_STRIP_PROGRAM = @INSTALL_STRIP_PROGRAM@ LD = @LD@ LDFLAGS = @LDFLAGS@ LIBOBJS = @LIBOBJS@ LIBQUADMATH = @LIBQUADMATH@ LIBS = @LIBS@ LIBTOOL = @LIBTOOL@ LIPO = @LIPO@ LN_S = @LN_S@ LTLIBOBJS = @LTLIBOBJS@ MAINT = @MAINT@ MAKEINFO = @MAKEINFO@ MANIFEST_TOOL = @MANIFEST_TOOL@ MKDIR_P = @MKDIR_P@ MPICC = @MPICC@ MPILIBS = @MPILIBS@ MPIRUN = @MPIRUN@ NEON_CFLAGS = @NEON_CFLAGS@ NM = @NM@ NMEDIT = @NMEDIT@ OBJDUMP = @OBJDUMP@ OBJEXT = @OBJEXT@ OCAMLBUILD = @OCAMLBUILD@ OPENMP_CFLAGS = @OPENMP_CFLAGS@ OTOOL = @OTOOL@ OTOOL64 = @OTOOL64@ PACKAGE = @PACKAGE@ PACKAGE_BUGREPORT = @PACKAGE_BUGREPORT@ PACKAGE_NAME = @PACKAGE_NAME@ PACKAGE_STRING = @PACKAGE_STRING@ PACKAGE_TARNAME = @PACKAGE_TARNAME@ PACKAGE_URL = @PACKAGE_URL@ PACKAGE_VERSION = @PACKAGE_VERSION@ PATH_SEPARATOR = @PATH_SEPARATOR@ POW_LIB = @POW_LIB@ PRECISION = @PRECISION@ PREC_SUFFIX = @PREC_SUFFIX@ PTHREAD_CC = @PTHREAD_CC@ PTHREAD_CFLAGS = @PTHREAD_CFLAGS@ PTHREAD_LIBS = @PTHREAD_LIBS@ RANLIB = @RANLIB@ SED = @SED@ SET_MAKE = @SET_MAKE@ SHARED_VERSION_INFO = @SHARED_VERSION_INFO@ SHELL = @SHELL@ SSE2_CFLAGS = @SSE2_CFLAGS@ STACK_ALIGN_CFLAGS = @STACK_ALIGN_CFLAGS@ STRIP = @STRIP@ THREADLIBS = @THREADLIBS@ VERSION = @VERSION@ abs_builddir = @abs_builddir@ abs_srcdir = @abs_srcdir@ abs_top_builddir = @abs_top_builddir@ abs_top_srcdir = @abs_top_srcdir@ ac_ct_AR = @ac_ct_AR@ ac_ct_CC = @ac_ct_CC@ ac_ct_DUMPBIN = @ac_ct_DUMPBIN@ ac_ct_F77 = @ac_ct_F77@ acx_pthread_config = @acx_pthread_config@ am__include = @am__include@ am__leading_dot = @am__leading_dot@ am__quote = @am__quote@ am__tar = @am__tar@ am__untar = @am__untar@ bindir = @bindir@ build = @build@ build_alias = @build_alias@ build_cpu = @build_cpu@ build_os = @build_os@ build_vendor = @build_vendor@ builddir = @builddir@ datadir = @datadir@ datarootdir = @datarootdir@ docdir = @docdir@ dvidir = @dvidir@ exec_prefix = @exec_prefix@ host = @host@ host_alias = @host_alias@ host_cpu = @host_cpu@ host_os = @host_os@ host_vendor = @host_vendor@ htmldir = @htmldir@ includedir = @includedir@ infodir = @infodir@ install_sh = @install_sh@ libdir = @libdir@ libexecdir = @libexecdir@ localedir = @localedir@ localstatedir = @localstatedir@ mandir = @mandir@ mkdir_p = @mkdir_p@ oldincludedir = @oldincludedir@ pdfdir = @pdfdir@ prefix = @prefix@ program_transform_name = @program_transform_name@ psdir = @psdir@ sbindir = @sbindir@ sharedstatedir = @sharedstatedir@ srcdir = @srcdir@ sysconfdir = @sysconfdir@ target_alias = @target_alias@ top_build_prefix = @top_build_prefix@ top_builddir = @top_builddir@ top_srcdir = @top_srcdir@ AM_CFLAGS = $(AVX_CFLAGS) SIMD_HEADER = simd-avx.h ########################################################################### # n1fv_ is a hard-coded FFTW_FORWARD FFT of size , using SIMD N1F = n1fv_2.c n1fv_3.c n1fv_4.c n1fv_5.c n1fv_6.c n1fv_7.c n1fv_8.c \ n1fv_9.c n1fv_10.c n1fv_11.c n1fv_12.c n1fv_13.c n1fv_14.c n1fv_15.c \ n1fv_16.c n1fv_32.c n1fv_64.c n1fv_128.c n1fv_20.c n1fv_25.c # as above, with restricted input vector stride N2F = n2fv_2.c n2fv_4.c n2fv_6.c n2fv_8.c n2fv_10.c n2fv_12.c \ n2fv_14.c n2fv_16.c n2fv_32.c n2fv_64.c n2fv_20.c # as above, but FFTW_BACKWARD N1B = n1bv_2.c n1bv_3.c n1bv_4.c n1bv_5.c n1bv_6.c n1bv_7.c n1bv_8.c \ n1bv_9.c n1bv_10.c n1bv_11.c n1bv_12.c n1bv_13.c n1bv_14.c n1bv_15.c \ n1bv_16.c n1bv_32.c n1bv_64.c n1bv_128.c n1bv_20.c n1bv_25.c N2B = n2bv_2.c n2bv_4.c n2bv_6.c n2bv_8.c n2bv_10.c n2bv_12.c \ n2bv_14.c n2bv_16.c n2bv_32.c n2bv_64.c n2bv_20.c # split-complex codelets N2S = n2sv_4.c n2sv_8.c n2sv_16.c n2sv_32.c n2sv_64.c ########################################################################### # t1fv_ is a "twiddle" FFT of size , implementing a radix-r DIT step # for an FFTW_FORWARD transform, using SIMD T1F = t1fv_2.c t1fv_3.c t1fv_4.c t1fv_5.c t1fv_6.c t1fv_7.c t1fv_8.c \ t1fv_9.c t1fv_10.c t1fv_12.c t1fv_15.c t1fv_16.c t1fv_32.c t1fv_64.c \ t1fv_20.c t1fv_25.c # same as t1fv_*, but with different twiddle storage scheme T2F = t2fv_2.c t2fv_4.c t2fv_8.c t2fv_16.c t2fv_32.c t2fv_64.c \ t2fv_5.c t2fv_10.c t2fv_20.c t2fv_25.c T3F = t3fv_4.c t3fv_8.c t3fv_16.c t3fv_32.c t3fv_5.c t3fv_10.c \ t3fv_20.c t3fv_25.c T1FU = t1fuv_2.c t1fuv_3.c t1fuv_4.c t1fuv_5.c t1fuv_6.c t1fuv_7.c \ t1fuv_8.c t1fuv_9.c t1fuv_10.c # as above, but FFTW_BACKWARD T1B = t1bv_2.c t1bv_3.c t1bv_4.c t1bv_5.c t1bv_6.c t1bv_7.c t1bv_8.c \ t1bv_9.c t1bv_10.c t1bv_12.c t1bv_15.c t1bv_16.c t1bv_32.c t1bv_64.c \ t1bv_20.c t1bv_25.c # same as t1bv_*, but with different twiddle storage scheme T2B = t2bv_2.c t2bv_4.c t2bv_8.c t2bv_16.c t2bv_32.c t2bv_64.c \ t2bv_5.c t2bv_10.c t2bv_20.c t2bv_25.c T3B = t3bv_4.c t3bv_8.c t3bv_16.c t3bv_32.c t3bv_5.c t3bv_10.c \ t3bv_20.c t3bv_25.c T1BU = t1buv_2.c t1buv_3.c t1buv_4.c t1buv_5.c t1buv_6.c t1buv_7.c \ t1buv_8.c t1buv_9.c t1buv_10.c # split-complex codelets T1S = t1sv_2.c t1sv_4.c t1sv_8.c t1sv_16.c t1sv_32.c T2S = t2sv_4.c t2sv_8.c t2sv_16.c t2sv_32.c ########################################################################### # q1fv_ is twiddle FFTW_FORWARD FFTs of size (DIF step), # where the output is transposed, using SIMD. This is used for # in-place transposes in sizes that are divisible by ^2. These # codelets have size ~ ^2, so you should probably not use # bigger than 8 or so. Q1F = q1fv_2.c q1fv_4.c q1fv_5.c q1fv_8.c # as above, but FFTW_BACKWARD Q1B = q1bv_2.c q1bv_4.c q1bv_5.c q1bv_8.c ########################################################################### SIMD_CODELETS = $(N1F) $(N1B) $(N2F) $(N2B) $(N2S) $(T1FU) $(T1F) \ $(T2F) $(T3F) $(T1BU) $(T1B) $(T2B) $(T3B) $(T1S) $(T2S) $(Q1F) $(Q1B) AM_CPPFLAGS = -I$(top_srcdir)/kernel -I$(top_srcdir)/dft \ -I$(top_srcdir)/dft/simd -I$(top_srcdir)/simd-support EXTRA_DIST = $(SIMD_CODELETS) genus.c codlist.c @HAVE_AVX_TRUE@BUILT_SOURCES = $(EXTRA_DIST) @HAVE_AVX_TRUE@noinst_LTLIBRARIES = libdft_avx_codelets.la @HAVE_AVX_TRUE@libdft_avx_codelets_la_SOURCES = $(BUILT_SOURCES) all: $(BUILT_SOURCES) $(MAKE) $(AM_MAKEFLAGS) all-am .SUFFIXES: .SUFFIXES: .c .lo .o .obj $(srcdir)/Makefile.in: @MAINTAINER_MODE_TRUE@ $(srcdir)/Makefile.am $(top_srcdir)/dft/simd/codlist.mk $(top_srcdir)/dft/simd/simd.mk $(am__configure_deps) @for dep in $?; do \ case '$(am__configure_deps)' in \ *$$dep*) \ ( cd $(top_builddir) && $(MAKE) $(AM_MAKEFLAGS) am--refresh ) \ && { if test -f $@; then exit 0; else break; fi; }; \ exit 1;; \ esac; \ done; \ echo ' cd $(top_srcdir) && $(AUTOMAKE) --gnu dft/simd/avx/Makefile'; \ $(am__cd) $(top_srcdir) && \ $(AUTOMAKE) --gnu dft/simd/avx/Makefile .PRECIOUS: Makefile Makefile: $(srcdir)/Makefile.in $(top_builddir)/config.status @case '$?' in \ *config.status*) \ cd $(top_builddir) && $(MAKE) $(AM_MAKEFLAGS) am--refresh;; \ *) \ echo ' cd $(top_builddir) && $(SHELL) ./config.status $(subdir)/$@ $(am__depfiles_maybe)'; \ cd $(top_builddir) && $(SHELL) ./config.status $(subdir)/$@ $(am__depfiles_maybe);; \ esac; $(top_srcdir)/dft/simd/codlist.mk $(top_srcdir)/dft/simd/simd.mk: $(top_builddir)/config.status: $(top_srcdir)/configure $(CONFIG_STATUS_DEPENDENCIES) cd $(top_builddir) && $(MAKE) $(AM_MAKEFLAGS) am--refresh $(top_srcdir)/configure: @MAINTAINER_MODE_TRUE@ $(am__configure_deps) cd $(top_builddir) && $(MAKE) $(AM_MAKEFLAGS) am--refresh $(ACLOCAL_M4): @MAINTAINER_MODE_TRUE@ $(am__aclocal_m4_deps) cd $(top_builddir) && $(MAKE) $(AM_MAKEFLAGS) am--refresh $(am__aclocal_m4_deps): clean-noinstLTLIBRARIES: -test -z "$(noinst_LTLIBRARIES)" || rm -f $(noinst_LTLIBRARIES) @list='$(noinst_LTLIBRARIES)'; for p in $$list; do \ dir="`echo $$p | sed -e 's|/[^/]*$$||'`"; \ test "$$dir" != "$$p" || dir=.; \ echo "rm -f \"$${dir}/so_locations\""; \ rm -f "$${dir}/so_locations"; \ done libdft_avx_codelets.la: $(libdft_avx_codelets_la_OBJECTS) $(libdft_avx_codelets_la_DEPENDENCIES) $(EXTRA_libdft_avx_codelets_la_DEPENDENCIES) $(LINK) $(am_libdft_avx_codelets_la_rpath) $(libdft_avx_codelets_la_OBJECTS) $(libdft_avx_codelets_la_LIBADD) $(LIBS) mostlyclean-compile: -rm -f *.$(OBJEXT) distclean-compile: -rm -f *.tab.c @AMDEP_TRUE@@am__include@ @am__quote@./$(DEPDIR)/codlist.Plo@am__quote@ @AMDEP_TRUE@@am__include@ @am__quote@./$(DEPDIR)/genus.Plo@am__quote@ @AMDEP_TRUE@@am__include@ @am__quote@./$(DEPDIR)/n1bv_10.Plo@am__quote@ @AMDEP_TRUE@@am__include@ @am__quote@./$(DEPDIR)/n1bv_11.Plo@am__quote@ @AMDEP_TRUE@@am__include@ @am__quote@./$(DEPDIR)/n1bv_12.Plo@am__quote@ @AMDEP_TRUE@@am__include@ @am__quote@./$(DEPDIR)/n1bv_128.Plo@am__quote@ @AMDEP_TRUE@@am__include@ @am__quote@./$(DEPDIR)/n1bv_13.Plo@am__quote@ @AMDEP_TRUE@@am__include@ @am__quote@./$(DEPDIR)/n1bv_14.Plo@am__quote@ @AMDEP_TRUE@@am__include@ @am__quote@./$(DEPDIR)/n1bv_15.Plo@am__quote@ @AMDEP_TRUE@@am__include@ @am__quote@./$(DEPDIR)/n1bv_16.Plo@am__quote@ @AMDEP_TRUE@@am__include@ @am__quote@./$(DEPDIR)/n1bv_2.Plo@am__quote@ @AMDEP_TRUE@@am__include@ @am__quote@./$(DEPDIR)/n1bv_20.Plo@am__quote@ @AMDEP_TRUE@@am__include@ @am__quote@./$(DEPDIR)/n1bv_25.Plo@am__quote@ @AMDEP_TRUE@@am__include@ @am__quote@./$(DEPDIR)/n1bv_3.Plo@am__quote@ @AMDEP_TRUE@@am__include@ @am__quote@./$(DEPDIR)/n1bv_32.Plo@am__quote@ @AMDEP_TRUE@@am__include@ @am__quote@./$(DEPDIR)/n1bv_4.Plo@am__quote@ @AMDEP_TRUE@@am__include@ @am__quote@./$(DEPDIR)/n1bv_5.Plo@am__quote@ @AMDEP_TRUE@@am__include@ @am__quote@./$(DEPDIR)/n1bv_6.Plo@am__quote@ @AMDEP_TRUE@@am__include@ @am__quote@./$(DEPDIR)/n1bv_64.Plo@am__quote@ @AMDEP_TRUE@@am__include@ @am__quote@./$(DEPDIR)/n1bv_7.Plo@am__quote@ @AMDEP_TRUE@@am__include@ @am__quote@./$(DEPDIR)/n1bv_8.Plo@am__quote@ @AMDEP_TRUE@@am__include@ @am__quote@./$(DEPDIR)/n1bv_9.Plo@am__quote@ @AMDEP_TRUE@@am__include@ @am__quote@./$(DEPDIR)/n1fv_10.Plo@am__quote@ @AMDEP_TRUE@@am__include@ @am__quote@./$(DEPDIR)/n1fv_11.Plo@am__quote@ @AMDEP_TRUE@@am__include@ @am__quote@./$(DEPDIR)/n1fv_12.Plo@am__quote@ @AMDEP_TRUE@@am__include@ @am__quote@./$(DEPDIR)/n1fv_128.Plo@am__quote@ @AMDEP_TRUE@@am__include@ @am__quote@./$(DEPDIR)/n1fv_13.Plo@am__quote@ @AMDEP_TRUE@@am__include@ @am__quote@./$(DEPDIR)/n1fv_14.Plo@am__quote@ @AMDEP_TRUE@@am__include@ @am__quote@./$(DEPDIR)/n1fv_15.Plo@am__quote@ @AMDEP_TRUE@@am__include@ @am__quote@./$(DEPDIR)/n1fv_16.Plo@am__quote@ @AMDEP_TRUE@@am__include@ @am__quote@./$(DEPDIR)/n1fv_2.Plo@am__quote@ @AMDEP_TRUE@@am__include@ @am__quote@./$(DEPDIR)/n1fv_20.Plo@am__quote@ @AMDEP_TRUE@@am__include@ @am__quote@./$(DEPDIR)/n1fv_25.Plo@am__quote@ @AMDEP_TRUE@@am__include@ @am__quote@./$(DEPDIR)/n1fv_3.Plo@am__quote@ @AMDEP_TRUE@@am__include@ @am__quote@./$(DEPDIR)/n1fv_32.Plo@am__quote@ @AMDEP_TRUE@@am__include@ @am__quote@./$(DEPDIR)/n1fv_4.Plo@am__quote@ @AMDEP_TRUE@@am__include@ @am__quote@./$(DEPDIR)/n1fv_5.Plo@am__quote@ @AMDEP_TRUE@@am__include@ @am__quote@./$(DEPDIR)/n1fv_6.Plo@am__quote@ @AMDEP_TRUE@@am__include@ @am__quote@./$(DEPDIR)/n1fv_64.Plo@am__quote@ @AMDEP_TRUE@@am__include@ @am__quote@./$(DEPDIR)/n1fv_7.Plo@am__quote@ @AMDEP_TRUE@@am__include@ @am__quote@./$(DEPDIR)/n1fv_8.Plo@am__quote@ @AMDEP_TRUE@@am__include@ @am__quote@./$(DEPDIR)/n1fv_9.Plo@am__quote@ @AMDEP_TRUE@@am__include@ @am__quote@./$(DEPDIR)/n2bv_10.Plo@am__quote@ @AMDEP_TRUE@@am__include@ @am__quote@./$(DEPDIR)/n2bv_12.Plo@am__quote@ @AMDEP_TRUE@@am__include@ @am__quote@./$(DEPDIR)/n2bv_14.Plo@am__quote@ @AMDEP_TRUE@@am__include@ @am__quote@./$(DEPDIR)/n2bv_16.Plo@am__quote@ @AMDEP_TRUE@@am__include@ @am__quote@./$(DEPDIR)/n2bv_2.Plo@am__quote@ @AMDEP_TRUE@@am__include@ @am__quote@./$(DEPDIR)/n2bv_20.Plo@am__quote@ @AMDEP_TRUE@@am__include@ @am__quote@./$(DEPDIR)/n2bv_32.Plo@am__quote@ @AMDEP_TRUE@@am__include@ @am__quote@./$(DEPDIR)/n2bv_4.Plo@am__quote@ @AMDEP_TRUE@@am__include@ @am__quote@./$(DEPDIR)/n2bv_6.Plo@am__quote@ @AMDEP_TRUE@@am__include@ @am__quote@./$(DEPDIR)/n2bv_64.Plo@am__quote@ @AMDEP_TRUE@@am__include@ @am__quote@./$(DEPDIR)/n2bv_8.Plo@am__quote@ @AMDEP_TRUE@@am__include@ @am__quote@./$(DEPDIR)/n2fv_10.Plo@am__quote@ @AMDEP_TRUE@@am__include@ @am__quote@./$(DEPDIR)/n2fv_12.Plo@am__quote@ @AMDEP_TRUE@@am__include@ @am__quote@./$(DEPDIR)/n2fv_14.Plo@am__quote@ @AMDEP_TRUE@@am__include@ @am__quote@./$(DEPDIR)/n2fv_16.Plo@am__quote@ @AMDEP_TRUE@@am__include@ @am__quote@./$(DEPDIR)/n2fv_2.Plo@am__quote@ @AMDEP_TRUE@@am__include@ @am__quote@./$(DEPDIR)/n2fv_20.Plo@am__quote@ @AMDEP_TRUE@@am__include@ @am__quote@./$(DEPDIR)/n2fv_32.Plo@am__quote@ @AMDEP_TRUE@@am__include@ @am__quote@./$(DEPDIR)/n2fv_4.Plo@am__quote@ @AMDEP_TRUE@@am__include@ @am__quote@./$(DEPDIR)/n2fv_6.Plo@am__quote@ @AMDEP_TRUE@@am__include@ @am__quote@./$(DEPDIR)/n2fv_64.Plo@am__quote@ @AMDEP_TRUE@@am__include@ @am__quote@./$(DEPDIR)/n2fv_8.Plo@am__quote@ @AMDEP_TRUE@@am__include@ @am__quote@./$(DEPDIR)/n2sv_16.Plo@am__quote@ @AMDEP_TRUE@@am__include@ @am__quote@./$(DEPDIR)/n2sv_32.Plo@am__quote@ @AMDEP_TRUE@@am__include@ @am__quote@./$(DEPDIR)/n2sv_4.Plo@am__quote@ @AMDEP_TRUE@@am__include@ @am__quote@./$(DEPDIR)/n2sv_64.Plo@am__quote@ @AMDEP_TRUE@@am__include@ @am__quote@./$(DEPDIR)/n2sv_8.Plo@am__quote@ @AMDEP_TRUE@@am__include@ @am__quote@./$(DEPDIR)/q1bv_2.Plo@am__quote@ @AMDEP_TRUE@@am__include@ @am__quote@./$(DEPDIR)/q1bv_4.Plo@am__quote@ @AMDEP_TRUE@@am__include@ @am__quote@./$(DEPDIR)/q1bv_5.Plo@am__quote@ @AMDEP_TRUE@@am__include@ @am__quote@./$(DEPDIR)/q1bv_8.Plo@am__quote@ @AMDEP_TRUE@@am__include@ @am__quote@./$(DEPDIR)/q1fv_2.Plo@am__quote@ @AMDEP_TRUE@@am__include@ @am__quote@./$(DEPDIR)/q1fv_4.Plo@am__quote@ @AMDEP_TRUE@@am__include@ @am__quote@./$(DEPDIR)/q1fv_5.Plo@am__quote@ @AMDEP_TRUE@@am__include@ @am__quote@./$(DEPDIR)/q1fv_8.Plo@am__quote@ @AMDEP_TRUE@@am__include@ @am__quote@./$(DEPDIR)/t1buv_10.Plo@am__quote@ @AMDEP_TRUE@@am__include@ @am__quote@./$(DEPDIR)/t1buv_2.Plo@am__quote@ @AMDEP_TRUE@@am__include@ @am__quote@./$(DEPDIR)/t1buv_3.Plo@am__quote@ @AMDEP_TRUE@@am__include@ @am__quote@./$(DEPDIR)/t1buv_4.Plo@am__quote@ @AMDEP_TRUE@@am__include@ @am__quote@./$(DEPDIR)/t1buv_5.Plo@am__quote@ @AMDEP_TRUE@@am__include@ @am__quote@./$(DEPDIR)/t1buv_6.Plo@am__quote@ @AMDEP_TRUE@@am__include@ @am__quote@./$(DEPDIR)/t1buv_7.Plo@am__quote@ @AMDEP_TRUE@@am__include@ @am__quote@./$(DEPDIR)/t1buv_8.Plo@am__quote@ @AMDEP_TRUE@@am__include@ @am__quote@./$(DEPDIR)/t1buv_9.Plo@am__quote@ @AMDEP_TRUE@@am__include@ @am__quote@./$(DEPDIR)/t1bv_10.Plo@am__quote@ @AMDEP_TRUE@@am__include@ @am__quote@./$(DEPDIR)/t1bv_12.Plo@am__quote@ @AMDEP_TRUE@@am__include@ @am__quote@./$(DEPDIR)/t1bv_15.Plo@am__quote@ @AMDEP_TRUE@@am__include@ @am__quote@./$(DEPDIR)/t1bv_16.Plo@am__quote@ @AMDEP_TRUE@@am__include@ @am__quote@./$(DEPDIR)/t1bv_2.Plo@am__quote@ @AMDEP_TRUE@@am__include@ @am__quote@./$(DEPDIR)/t1bv_20.Plo@am__quote@ @AMDEP_TRUE@@am__include@ @am__quote@./$(DEPDIR)/t1bv_25.Plo@am__quote@ @AMDEP_TRUE@@am__include@ @am__quote@./$(DEPDIR)/t1bv_3.Plo@am__quote@ @AMDEP_TRUE@@am__include@ @am__quote@./$(DEPDIR)/t1bv_32.Plo@am__quote@ @AMDEP_TRUE@@am__include@ @am__quote@./$(DEPDIR)/t1bv_4.Plo@am__quote@ @AMDEP_TRUE@@am__include@ @am__quote@./$(DEPDIR)/t1bv_5.Plo@am__quote@ @AMDEP_TRUE@@am__include@ @am__quote@./$(DEPDIR)/t1bv_6.Plo@am__quote@ @AMDEP_TRUE@@am__include@ @am__quote@./$(DEPDIR)/t1bv_64.Plo@am__quote@ @AMDEP_TRUE@@am__include@ @am__quote@./$(DEPDIR)/t1bv_7.Plo@am__quote@ @AMDEP_TRUE@@am__include@ @am__quote@./$(DEPDIR)/t1bv_8.Plo@am__quote@ @AMDEP_TRUE@@am__include@ @am__quote@./$(DEPDIR)/t1bv_9.Plo@am__quote@ @AMDEP_TRUE@@am__include@ @am__quote@./$(DEPDIR)/t1fuv_10.Plo@am__quote@ @AMDEP_TRUE@@am__include@ @am__quote@./$(DEPDIR)/t1fuv_2.Plo@am__quote@ @AMDEP_TRUE@@am__include@ @am__quote@./$(DEPDIR)/t1fuv_3.Plo@am__quote@ @AMDEP_TRUE@@am__include@ @am__quote@./$(DEPDIR)/t1fuv_4.Plo@am__quote@ @AMDEP_TRUE@@am__include@ @am__quote@./$(DEPDIR)/t1fuv_5.Plo@am__quote@ @AMDEP_TRUE@@am__include@ @am__quote@./$(DEPDIR)/t1fuv_6.Plo@am__quote@ @AMDEP_TRUE@@am__include@ @am__quote@./$(DEPDIR)/t1fuv_7.Plo@am__quote@ @AMDEP_TRUE@@am__include@ @am__quote@./$(DEPDIR)/t1fuv_8.Plo@am__quote@ @AMDEP_TRUE@@am__include@ @am__quote@./$(DEPDIR)/t1fuv_9.Plo@am__quote@ @AMDEP_TRUE@@am__include@ @am__quote@./$(DEPDIR)/t1fv_10.Plo@am__quote@ @AMDEP_TRUE@@am__include@ @am__quote@./$(DEPDIR)/t1fv_12.Plo@am__quote@ @AMDEP_TRUE@@am__include@ @am__quote@./$(DEPDIR)/t1fv_15.Plo@am__quote@ @AMDEP_TRUE@@am__include@ @am__quote@./$(DEPDIR)/t1fv_16.Plo@am__quote@ @AMDEP_TRUE@@am__include@ @am__quote@./$(DEPDIR)/t1fv_2.Plo@am__quote@ @AMDEP_TRUE@@am__include@ @am__quote@./$(DEPDIR)/t1fv_20.Plo@am__quote@ @AMDEP_TRUE@@am__include@ @am__quote@./$(DEPDIR)/t1fv_25.Plo@am__quote@ @AMDEP_TRUE@@am__include@ @am__quote@./$(DEPDIR)/t1fv_3.Plo@am__quote@ @AMDEP_TRUE@@am__include@ @am__quote@./$(DEPDIR)/t1fv_32.Plo@am__quote@ @AMDEP_TRUE@@am__include@ @am__quote@./$(DEPDIR)/t1fv_4.Plo@am__quote@ @AMDEP_TRUE@@am__include@ @am__quote@./$(DEPDIR)/t1fv_5.Plo@am__quote@ @AMDEP_TRUE@@am__include@ @am__quote@./$(DEPDIR)/t1fv_6.Plo@am__quote@ @AMDEP_TRUE@@am__include@ @am__quote@./$(DEPDIR)/t1fv_64.Plo@am__quote@ @AMDEP_TRUE@@am__include@ @am__quote@./$(DEPDIR)/t1fv_7.Plo@am__quote@ @AMDEP_TRUE@@am__include@ @am__quote@./$(DEPDIR)/t1fv_8.Plo@am__quote@ @AMDEP_TRUE@@am__include@ @am__quote@./$(DEPDIR)/t1fv_9.Plo@am__quote@ @AMDEP_TRUE@@am__include@ @am__quote@./$(DEPDIR)/t1sv_16.Plo@am__quote@ @AMDEP_TRUE@@am__include@ @am__quote@./$(DEPDIR)/t1sv_2.Plo@am__quote@ @AMDEP_TRUE@@am__include@ @am__quote@./$(DEPDIR)/t1sv_32.Plo@am__quote@ @AMDEP_TRUE@@am__include@ @am__quote@./$(DEPDIR)/t1sv_4.Plo@am__quote@ @AMDEP_TRUE@@am__include@ @am__quote@./$(DEPDIR)/t1sv_8.Plo@am__quote@ @AMDEP_TRUE@@am__include@ @am__quote@./$(DEPDIR)/t2bv_10.Plo@am__quote@ @AMDEP_TRUE@@am__include@ @am__quote@./$(DEPDIR)/t2bv_16.Plo@am__quote@ @AMDEP_TRUE@@am__include@ @am__quote@./$(DEPDIR)/t2bv_2.Plo@am__quote@ @AMDEP_TRUE@@am__include@ @am__quote@./$(DEPDIR)/t2bv_20.Plo@am__quote@ @AMDEP_TRUE@@am__include@ @am__quote@./$(DEPDIR)/t2bv_25.Plo@am__quote@ @AMDEP_TRUE@@am__include@ @am__quote@./$(DEPDIR)/t2bv_32.Plo@am__quote@ @AMDEP_TRUE@@am__include@ @am__quote@./$(DEPDIR)/t2bv_4.Plo@am__quote@ @AMDEP_TRUE@@am__include@ @am__quote@./$(DEPDIR)/t2bv_5.Plo@am__quote@ @AMDEP_TRUE@@am__include@ @am__quote@./$(DEPDIR)/t2bv_64.Plo@am__quote@ @AMDEP_TRUE@@am__include@ @am__quote@./$(DEPDIR)/t2bv_8.Plo@am__quote@ @AMDEP_TRUE@@am__include@ @am__quote@./$(DEPDIR)/t2fv_10.Plo@am__quote@ @AMDEP_TRUE@@am__include@ @am__quote@./$(DEPDIR)/t2fv_16.Plo@am__quote@ @AMDEP_TRUE@@am__include@ @am__quote@./$(DEPDIR)/t2fv_2.Plo@am__quote@ @AMDEP_TRUE@@am__include@ @am__quote@./$(DEPDIR)/t2fv_20.Plo@am__quote@ @AMDEP_TRUE@@am__include@ @am__quote@./$(DEPDIR)/t2fv_25.Plo@am__quote@ @AMDEP_TRUE@@am__include@ @am__quote@./$(DEPDIR)/t2fv_32.Plo@am__quote@ @AMDEP_TRUE@@am__include@ @am__quote@./$(DEPDIR)/t2fv_4.Plo@am__quote@ @AMDEP_TRUE@@am__include@ @am__quote@./$(DEPDIR)/t2fv_5.Plo@am__quote@ @AMDEP_TRUE@@am__include@ @am__quote@./$(DEPDIR)/t2fv_64.Plo@am__quote@ @AMDEP_TRUE@@am__include@ @am__quote@./$(DEPDIR)/t2fv_8.Plo@am__quote@ @AMDEP_TRUE@@am__include@ @am__quote@./$(DEPDIR)/t2sv_16.Plo@am__quote@ @AMDEP_TRUE@@am__include@ @am__quote@./$(DEPDIR)/t2sv_32.Plo@am__quote@ @AMDEP_TRUE@@am__include@ @am__quote@./$(DEPDIR)/t2sv_4.Plo@am__quote@ @AMDEP_TRUE@@am__include@ @am__quote@./$(DEPDIR)/t2sv_8.Plo@am__quote@ @AMDEP_TRUE@@am__include@ @am__quote@./$(DEPDIR)/t3bv_10.Plo@am__quote@ @AMDEP_TRUE@@am__include@ @am__quote@./$(DEPDIR)/t3bv_16.Plo@am__quote@ @AMDEP_TRUE@@am__include@ @am__quote@./$(DEPDIR)/t3bv_20.Plo@am__quote@ @AMDEP_TRUE@@am__include@ @am__quote@./$(DEPDIR)/t3bv_25.Plo@am__quote@ @AMDEP_TRUE@@am__include@ @am__quote@./$(DEPDIR)/t3bv_32.Plo@am__quote@ @AMDEP_TRUE@@am__include@ @am__quote@./$(DEPDIR)/t3bv_4.Plo@am__quote@ @AMDEP_TRUE@@am__include@ @am__quote@./$(DEPDIR)/t3bv_5.Plo@am__quote@ @AMDEP_TRUE@@am__include@ @am__quote@./$(DEPDIR)/t3bv_8.Plo@am__quote@ @AMDEP_TRUE@@am__include@ @am__quote@./$(DEPDIR)/t3fv_10.Plo@am__quote@ @AMDEP_TRUE@@am__include@ @am__quote@./$(DEPDIR)/t3fv_16.Plo@am__quote@ @AMDEP_TRUE@@am__include@ @am__quote@./$(DEPDIR)/t3fv_20.Plo@am__quote@ @AMDEP_TRUE@@am__include@ @am__quote@./$(DEPDIR)/t3fv_25.Plo@am__quote@ @AMDEP_TRUE@@am__include@ @am__quote@./$(DEPDIR)/t3fv_32.Plo@am__quote@ @AMDEP_TRUE@@am__include@ @am__quote@./$(DEPDIR)/t3fv_4.Plo@am__quote@ @AMDEP_TRUE@@am__include@ @am__quote@./$(DEPDIR)/t3fv_5.Plo@am__quote@ @AMDEP_TRUE@@am__include@ @am__quote@./$(DEPDIR)/t3fv_8.Plo@am__quote@ .c.o: @am__fastdepCC_TRUE@ $(COMPILE) -MT $@ -MD -MP -MF $(DEPDIR)/$*.Tpo -c -o $@ $< @am__fastdepCC_TRUE@ $(am__mv) $(DEPDIR)/$*.Tpo $(DEPDIR)/$*.Po @AMDEP_TRUE@@am__fastdepCC_FALSE@ source='$<' object='$@' libtool=no @AMDEPBACKSLASH@ @AMDEP_TRUE@@am__fastdepCC_FALSE@ DEPDIR=$(DEPDIR) $(CCDEPMODE) $(depcomp) @AMDEPBACKSLASH@ @am__fastdepCC_FALSE@ $(COMPILE) -c $< .c.obj: @am__fastdepCC_TRUE@ $(COMPILE) -MT $@ -MD -MP -MF $(DEPDIR)/$*.Tpo -c -o $@ `$(CYGPATH_W) '$<'` @am__fastdepCC_TRUE@ $(am__mv) $(DEPDIR)/$*.Tpo $(DEPDIR)/$*.Po @AMDEP_TRUE@@am__fastdepCC_FALSE@ source='$<' object='$@' libtool=no @AMDEPBACKSLASH@ @AMDEP_TRUE@@am__fastdepCC_FALSE@ DEPDIR=$(DEPDIR) $(CCDEPMODE) $(depcomp) @AMDEPBACKSLASH@ @am__fastdepCC_FALSE@ $(COMPILE) -c `$(CYGPATH_W) '$<'` .c.lo: @am__fastdepCC_TRUE@ $(LTCOMPILE) -MT $@ -MD -MP -MF $(DEPDIR)/$*.Tpo -c -o $@ $< @am__fastdepCC_TRUE@ $(am__mv) $(DEPDIR)/$*.Tpo $(DEPDIR)/$*.Plo @AMDEP_TRUE@@am__fastdepCC_FALSE@ source='$<' object='$@' libtool=yes @AMDEPBACKSLASH@ @AMDEP_TRUE@@am__fastdepCC_FALSE@ DEPDIR=$(DEPDIR) $(CCDEPMODE) $(depcomp) @AMDEPBACKSLASH@ @am__fastdepCC_FALSE@ $(LTCOMPILE) -c -o $@ $< mostlyclean-libtool: -rm -f *.lo clean-libtool: -rm -rf .libs _libs ID: $(HEADERS) $(SOURCES) $(LISP) $(TAGS_FILES) list='$(SOURCES) $(HEADERS) $(LISP) $(TAGS_FILES)'; \ unique=`for i in $$list; do \ if test -f "$$i"; then echo $$i; else echo $(srcdir)/$$i; fi; \ done | \ $(AWK) '{ files[$$0] = 1; nonempty = 1; } \ END { if (nonempty) { for (i in files) print i; }; }'`; \ mkid -fID $$unique tags: TAGS TAGS: $(HEADERS) $(SOURCES) $(TAGS_DEPENDENCIES) \ $(TAGS_FILES) $(LISP) set x; \ here=`pwd`; \ list='$(SOURCES) $(HEADERS) $(LISP) $(TAGS_FILES)'; \ unique=`for i in $$list; do \ if test -f "$$i"; then echo $$i; else echo $(srcdir)/$$i; fi; \ done | \ $(AWK) '{ files[$$0] = 1; nonempty = 1; } \ END { if (nonempty) { for (i in files) print i; }; }'`; \ shift; \ if test -z "$(ETAGS_ARGS)$$*$$unique"; then :; else \ test -n "$$unique" || unique=$$empty_fix; \ if test $$# -gt 0; then \ $(ETAGS) $(ETAGSFLAGS) $(AM_ETAGSFLAGS) $(ETAGS_ARGS) \ "$$@" $$unique; \ else \ $(ETAGS) $(ETAGSFLAGS) $(AM_ETAGSFLAGS) $(ETAGS_ARGS) \ $$unique; \ fi; \ fi ctags: CTAGS CTAGS: $(HEADERS) $(SOURCES) $(TAGS_DEPENDENCIES) \ $(TAGS_FILES) $(LISP) list='$(SOURCES) $(HEADERS) $(LISP) $(TAGS_FILES)'; \ unique=`for i in $$list; do \ if test -f "$$i"; then echo $$i; else echo $(srcdir)/$$i; fi; \ done | \ $(AWK) '{ files[$$0] = 1; nonempty = 1; } \ END { if (nonempty) { for (i in files) print i; }; }'`; \ test -z "$(CTAGS_ARGS)$$unique" \ || $(CTAGS) $(CTAGSFLAGS) $(AM_CTAGSFLAGS) $(CTAGS_ARGS) \ $$unique GTAGS: here=`$(am__cd) $(top_builddir) && pwd` \ && $(am__cd) $(top_srcdir) \ && gtags -i $(GTAGS_ARGS) "$$here" distclean-tags: -rm -f TAGS ID GTAGS GRTAGS GSYMS GPATH tags distdir: $(DISTFILES) @srcdirstrip=`echo "$(srcdir)" | sed 's/[].[^$$\\*]/\\\\&/g'`; \ topsrcdirstrip=`echo "$(top_srcdir)" | sed 's/[].[^$$\\*]/\\\\&/g'`; \ list='$(DISTFILES)'; \ dist_files=`for file in $$list; do echo $$file; done | \ sed -e "s|^$$srcdirstrip/||;t" \ -e "s|^$$topsrcdirstrip/|$(top_builddir)/|;t"`; \ case $$dist_files in \ */*) $(MKDIR_P) `echo "$$dist_files" | \ sed '/\//!d;s|^|$(distdir)/|;s,/[^/]*$$,,' | \ sort -u` ;; \ esac; \ for file in $$dist_files; do \ if test -f $$file || test -d $$file; then d=.; else d=$(srcdir); fi; \ if test -d $$d/$$file; then \ dir=`echo "/$$file" | sed -e 's,/[^/]*$$,,'`; \ if test -d "$(distdir)/$$file"; then \ find "$(distdir)/$$file" -type d ! -perm -700 -exec chmod u+rwx {} \;; \ fi; \ if test -d $(srcdir)/$$file && test $$d != $(srcdir); then \ cp -fpR $(srcdir)/$$file "$(distdir)$$dir" || exit 1; \ find "$(distdir)/$$file" -type d ! -perm -700 -exec chmod u+rwx {} \;; \ fi; \ cp -fpR $$d/$$file "$(distdir)$$dir" || exit 1; \ else \ test -f "$(distdir)/$$file" \ || cp -p $$d/$$file "$(distdir)/$$file" \ || exit 1; \ fi; \ done check-am: all-am check: $(BUILT_SOURCES) $(MAKE) $(AM_MAKEFLAGS) check-am all-am: Makefile $(LTLIBRARIES) installdirs: install: $(BUILT_SOURCES) $(MAKE) $(AM_MAKEFLAGS) install-am install-exec: install-exec-am install-data: install-data-am uninstall: uninstall-am install-am: all-am @$(MAKE) $(AM_MAKEFLAGS) install-exec-am install-data-am installcheck: installcheck-am install-strip: if test -z '$(STRIP)'; then \ $(MAKE) $(AM_MAKEFLAGS) INSTALL_PROGRAM="$(INSTALL_STRIP_PROGRAM)" \ install_sh_PROGRAM="$(INSTALL_STRIP_PROGRAM)" INSTALL_STRIP_FLAG=-s \ install; \ else \ $(MAKE) $(AM_MAKEFLAGS) INSTALL_PROGRAM="$(INSTALL_STRIP_PROGRAM)" \ install_sh_PROGRAM="$(INSTALL_STRIP_PROGRAM)" INSTALL_STRIP_FLAG=-s \ "INSTALL_PROGRAM_ENV=STRIPPROG='$(STRIP)'" install; \ fi mostlyclean-generic: clean-generic: distclean-generic: -test -z "$(CONFIG_CLEAN_FILES)" || rm -f $(CONFIG_CLEAN_FILES) -test . = "$(srcdir)" || test -z "$(CONFIG_CLEAN_VPATH_FILES)" || rm -f $(CONFIG_CLEAN_VPATH_FILES) maintainer-clean-generic: @echo "This command is intended for maintainers to use" @echo "it deletes files that may require special tools to rebuild." -test -z "$(BUILT_SOURCES)" || rm -f $(BUILT_SOURCES) clean: clean-am clean-am: clean-generic clean-libtool clean-noinstLTLIBRARIES \ mostlyclean-am distclean: distclean-am -rm -rf ./$(DEPDIR) -rm -f Makefile distclean-am: clean-am distclean-compile distclean-generic \ distclean-tags dvi: dvi-am dvi-am: html: html-am html-am: info: info-am info-am: install-data-am: install-dvi: install-dvi-am install-dvi-am: install-exec-am: install-html: install-html-am install-html-am: install-info: install-info-am install-info-am: install-man: install-pdf: install-pdf-am install-pdf-am: install-ps: install-ps-am install-ps-am: installcheck-am: maintainer-clean: maintainer-clean-am -rm -rf ./$(DEPDIR) -rm -f Makefile maintainer-clean-am: distclean-am maintainer-clean-generic mostlyclean: mostlyclean-am mostlyclean-am: mostlyclean-compile mostlyclean-generic \ mostlyclean-libtool pdf: pdf-am pdf-am: ps: ps-am ps-am: uninstall-am: .MAKE: all check install install-am install-strip .PHONY: CTAGS GTAGS all all-am check check-am clean clean-generic \ clean-libtool clean-noinstLTLIBRARIES ctags distclean \ distclean-compile distclean-generic distclean-libtool \ distclean-tags distdir dvi dvi-am html html-am info info-am \ install install-am install-data install-data-am install-dvi \ install-dvi-am install-exec install-exec-am install-html \ install-html-am install-info install-info-am install-man \ install-pdf install-pdf-am install-ps install-ps-am \ install-strip installcheck installcheck-am installdirs \ maintainer-clean maintainer-clean-generic mostlyclean \ mostlyclean-compile mostlyclean-generic mostlyclean-libtool \ pdf pdf-am ps ps-am tags uninstall uninstall-am $(EXTRA_DIST): Makefile ( \ echo "/* Generated automatically. DO NOT EDIT! */"; \ echo "#define SIMD_HEADER \"$(SIMD_HEADER)\""; \ echo "#include \"../common/"$*".c\""; \ ) >$@ # Tell versions [3.59,3.63) of GNU make to not export all variables. # Otherwise a system limit (for SysV at least) may be exceeded. .NOEXPORT: fftw-3.3.3/dft/simd/avx/n1fv_2.c0000644000175000017500000000015412054411106013147 00000000000000/* Generated automatically. DO NOT EDIT! */ #define SIMD_HEADER "simd-avx.h" #include "../common/n1fv_2.c" fftw-3.3.3/dft/simd/avx/n1fv_3.c0000644000175000017500000000015412054411106013150 00000000000000/* Generated automatically. DO NOT EDIT! */ #define SIMD_HEADER "simd-avx.h" #include "../common/n1fv_3.c" fftw-3.3.3/dft/simd/avx/n1fv_4.c0000644000175000017500000000015412054411106013151 00000000000000/* Generated automatically. DO NOT EDIT! */ #define SIMD_HEADER "simd-avx.h" #include "../common/n1fv_4.c" fftw-3.3.3/dft/simd/avx/n1fv_5.c0000644000175000017500000000015412054411106013152 00000000000000/* Generated automatically. DO NOT EDIT! */ #define SIMD_HEADER "simd-avx.h" #include "../common/n1fv_5.c" fftw-3.3.3/dft/simd/avx/n1fv_6.c0000644000175000017500000000015412054411106013153 00000000000000/* Generated automatically. DO NOT EDIT! */ #define SIMD_HEADER "simd-avx.h" #include "../common/n1fv_6.c" fftw-3.3.3/dft/simd/avx/n1fv_7.c0000644000175000017500000000015412054411106013154 00000000000000/* Generated automatically. DO NOT EDIT! */ #define SIMD_HEADER "simd-avx.h" #include "../common/n1fv_7.c" fftw-3.3.3/dft/simd/avx/n1fv_8.c0000644000175000017500000000015412054411106013155 00000000000000/* Generated automatically. DO NOT EDIT! */ #define SIMD_HEADER "simd-avx.h" #include "../common/n1fv_8.c" fftw-3.3.3/dft/simd/avx/n1fv_9.c0000644000175000017500000000015412054411106013156 00000000000000/* Generated automatically. DO NOT EDIT! */ #define SIMD_HEADER "simd-avx.h" #include "../common/n1fv_9.c" fftw-3.3.3/dft/simd/avx/n1fv_10.c0000644000175000017500000000015512054411106013227 00000000000000/* Generated automatically. DO NOT EDIT! */ #define SIMD_HEADER "simd-avx.h" #include "../common/n1fv_10.c" fftw-3.3.3/dft/simd/avx/n1fv_11.c0000644000175000017500000000015512054411106013230 00000000000000/* Generated automatically. DO NOT EDIT! */ #define SIMD_HEADER "simd-avx.h" #include "../common/n1fv_11.c" fftw-3.3.3/dft/simd/avx/n1fv_12.c0000644000175000017500000000015512054411106013231 00000000000000/* Generated automatically. DO NOT EDIT! */ #define SIMD_HEADER "simd-avx.h" #include "../common/n1fv_12.c" fftw-3.3.3/dft/simd/avx/n1fv_13.c0000644000175000017500000000015512054411106013232 00000000000000/* Generated automatically. DO NOT EDIT! */ #define SIMD_HEADER "simd-avx.h" #include "../common/n1fv_13.c" fftw-3.3.3/dft/simd/avx/n1fv_14.c0000644000175000017500000000015512054411106013233 00000000000000/* Generated automatically. DO NOT EDIT! */ #define SIMD_HEADER "simd-avx.h" #include "../common/n1fv_14.c" fftw-3.3.3/dft/simd/avx/n1fv_15.c0000644000175000017500000000015512054411106013234 00000000000000/* Generated automatically. DO NOT EDIT! */ #define SIMD_HEADER "simd-avx.h" #include "../common/n1fv_15.c" fftw-3.3.3/dft/simd/avx/n1fv_16.c0000644000175000017500000000015512054411106013235 00000000000000/* Generated automatically. DO NOT EDIT! */ #define SIMD_HEADER "simd-avx.h" #include "../common/n1fv_16.c" fftw-3.3.3/dft/simd/avx/n1fv_32.c0000644000175000017500000000015512054411106013233 00000000000000/* Generated automatically. DO NOT EDIT! */ #define SIMD_HEADER "simd-avx.h" #include "../common/n1fv_32.c" fftw-3.3.3/dft/simd/avx/n1fv_64.c0000644000175000017500000000015512054411106013240 00000000000000/* Generated automatically. DO NOT EDIT! */ #define SIMD_HEADER "simd-avx.h" #include "../common/n1fv_64.c" fftw-3.3.3/dft/simd/avx/n1fv_128.c0000644000175000017500000000015612054411106013322 00000000000000/* Generated automatically. DO NOT EDIT! */ #define SIMD_HEADER "simd-avx.h" #include "../common/n1fv_128.c" fftw-3.3.3/dft/simd/avx/n1fv_20.c0000644000175000017500000000015512054411106013230 00000000000000/* Generated automatically. DO NOT EDIT! */ #define SIMD_HEADER "simd-avx.h" #include "../common/n1fv_20.c" fftw-3.3.3/dft/simd/avx/n1fv_25.c0000644000175000017500000000015512054411106013235 00000000000000/* Generated automatically. DO NOT EDIT! */ #define SIMD_HEADER "simd-avx.h" #include "../common/n1fv_25.c" fftw-3.3.3/dft/simd/avx/n1bv_2.c0000644000175000017500000000015412054411106013143 00000000000000/* Generated automatically. DO NOT EDIT! */ #define SIMD_HEADER "simd-avx.h" #include "../common/n1bv_2.c" fftw-3.3.3/dft/simd/avx/n1bv_3.c0000644000175000017500000000015412054411106013144 00000000000000/* Generated automatically. DO NOT EDIT! */ #define SIMD_HEADER "simd-avx.h" #include "../common/n1bv_3.c" fftw-3.3.3/dft/simd/avx/n1bv_4.c0000644000175000017500000000015412054411106013145 00000000000000/* Generated automatically. DO NOT EDIT! */ #define SIMD_HEADER "simd-avx.h" #include "../common/n1bv_4.c" fftw-3.3.3/dft/simd/avx/n1bv_5.c0000644000175000017500000000015412054411106013146 00000000000000/* Generated automatically. DO NOT EDIT! */ #define SIMD_HEADER "simd-avx.h" #include "../common/n1bv_5.c" fftw-3.3.3/dft/simd/avx/n1bv_6.c0000644000175000017500000000015412054411106013147 00000000000000/* Generated automatically. DO NOT EDIT! */ #define SIMD_HEADER "simd-avx.h" #include "../common/n1bv_6.c" fftw-3.3.3/dft/simd/avx/n1bv_7.c0000644000175000017500000000015412054411106013150 00000000000000/* Generated automatically. DO NOT EDIT! */ #define SIMD_HEADER "simd-avx.h" #include "../common/n1bv_7.c" fftw-3.3.3/dft/simd/avx/n1bv_8.c0000644000175000017500000000015412054411106013151 00000000000000/* Generated automatically. DO NOT EDIT! */ #define SIMD_HEADER "simd-avx.h" #include "../common/n1bv_8.c" fftw-3.3.3/dft/simd/avx/n1bv_9.c0000644000175000017500000000015412054411106013152 00000000000000/* Generated automatically. DO NOT EDIT! */ #define SIMD_HEADER "simd-avx.h" #include "../common/n1bv_9.c" fftw-3.3.3/dft/simd/avx/n1bv_10.c0000644000175000017500000000015512054411106013223 00000000000000/* Generated automatically. DO NOT EDIT! */ #define SIMD_HEADER "simd-avx.h" #include "../common/n1bv_10.c" fftw-3.3.3/dft/simd/avx/n1bv_11.c0000644000175000017500000000015512054411106013224 00000000000000/* Generated automatically. DO NOT EDIT! */ #define SIMD_HEADER "simd-avx.h" #include "../common/n1bv_11.c" fftw-3.3.3/dft/simd/avx/n1bv_12.c0000644000175000017500000000015512054411106013225 00000000000000/* Generated automatically. DO NOT EDIT! */ #define SIMD_HEADER "simd-avx.h" #include "../common/n1bv_12.c" fftw-3.3.3/dft/simd/avx/n1bv_13.c0000644000175000017500000000015512054411106013226 00000000000000/* Generated automatically. DO NOT EDIT! */ #define SIMD_HEADER "simd-avx.h" #include "../common/n1bv_13.c" fftw-3.3.3/dft/simd/avx/n1bv_14.c0000644000175000017500000000015512054411106013227 00000000000000/* Generated automatically. DO NOT EDIT! */ #define SIMD_HEADER "simd-avx.h" #include "../common/n1bv_14.c" fftw-3.3.3/dft/simd/avx/n1bv_15.c0000644000175000017500000000015512054411106013230 00000000000000/* Generated automatically. DO NOT EDIT! */ #define SIMD_HEADER "simd-avx.h" #include "../common/n1bv_15.c" fftw-3.3.3/dft/simd/avx/n1bv_16.c0000644000175000017500000000015512054411106013231 00000000000000/* Generated automatically. DO NOT EDIT! */ #define SIMD_HEADER "simd-avx.h" #include "../common/n1bv_16.c" fftw-3.3.3/dft/simd/avx/n1bv_32.c0000644000175000017500000000015512054411106013227 00000000000000/* Generated automatically. DO NOT EDIT! */ #define SIMD_HEADER "simd-avx.h" #include "../common/n1bv_32.c" fftw-3.3.3/dft/simd/avx/n1bv_64.c0000644000175000017500000000015512054411106013234 00000000000000/* Generated automatically. DO NOT EDIT! */ #define SIMD_HEADER "simd-avx.h" #include "../common/n1bv_64.c" fftw-3.3.3/dft/simd/avx/n1bv_128.c0000644000175000017500000000015612054411106013316 00000000000000/* Generated automatically. DO NOT EDIT! */ #define SIMD_HEADER "simd-avx.h" #include "../common/n1bv_128.c" fftw-3.3.3/dft/simd/avx/n1bv_20.c0000644000175000017500000000015512054411106013224 00000000000000/* Generated automatically. DO NOT EDIT! */ #define SIMD_HEADER "simd-avx.h" #include "../common/n1bv_20.c" fftw-3.3.3/dft/simd/avx/n1bv_25.c0000644000175000017500000000015512054411106013231 00000000000000/* Generated automatically. DO NOT EDIT! */ #define SIMD_HEADER "simd-avx.h" #include "../common/n1bv_25.c" fftw-3.3.3/dft/simd/avx/n2fv_2.c0000644000175000017500000000015412054411106013150 00000000000000/* Generated automatically. DO NOT EDIT! */ #define SIMD_HEADER "simd-avx.h" #include "../common/n2fv_2.c" fftw-3.3.3/dft/simd/avx/n2fv_4.c0000644000175000017500000000015412054411106013152 00000000000000/* Generated automatically. DO NOT EDIT! */ #define SIMD_HEADER "simd-avx.h" #include "../common/n2fv_4.c" fftw-3.3.3/dft/simd/avx/n2fv_6.c0000644000175000017500000000015412054411106013154 00000000000000/* Generated automatically. DO NOT EDIT! */ #define SIMD_HEADER "simd-avx.h" #include "../common/n2fv_6.c" fftw-3.3.3/dft/simd/avx/n2fv_8.c0000644000175000017500000000015412054411106013156 00000000000000/* Generated automatically. DO NOT EDIT! */ #define SIMD_HEADER "simd-avx.h" #include "../common/n2fv_8.c" fftw-3.3.3/dft/simd/avx/n2fv_10.c0000644000175000017500000000015512054411106013230 00000000000000/* Generated automatically. DO NOT EDIT! */ #define SIMD_HEADER "simd-avx.h" #include "../common/n2fv_10.c" fftw-3.3.3/dft/simd/avx/n2fv_12.c0000644000175000017500000000015512054411106013232 00000000000000/* Generated automatically. DO NOT EDIT! */ #define SIMD_HEADER "simd-avx.h" #include "../common/n2fv_12.c" fftw-3.3.3/dft/simd/avx/n2fv_14.c0000644000175000017500000000015512054411106013234 00000000000000/* Generated automatically. DO NOT EDIT! */ #define SIMD_HEADER "simd-avx.h" #include "../common/n2fv_14.c" fftw-3.3.3/dft/simd/avx/n2fv_16.c0000644000175000017500000000015512054411106013236 00000000000000/* Generated automatically. DO NOT EDIT! */ #define SIMD_HEADER "simd-avx.h" #include "../common/n2fv_16.c" fftw-3.3.3/dft/simd/avx/n2fv_32.c0000644000175000017500000000015512054411106013234 00000000000000/* Generated automatically. DO NOT EDIT! */ #define SIMD_HEADER "simd-avx.h" #include "../common/n2fv_32.c" fftw-3.3.3/dft/simd/avx/n2fv_64.c0000644000175000017500000000015512054411106013241 00000000000000/* Generated automatically. DO NOT EDIT! */ #define SIMD_HEADER "simd-avx.h" #include "../common/n2fv_64.c" fftw-3.3.3/dft/simd/avx/n2fv_20.c0000644000175000017500000000015512054411106013231 00000000000000/* Generated automatically. DO NOT EDIT! */ #define SIMD_HEADER "simd-avx.h" #include "../common/n2fv_20.c" fftw-3.3.3/dft/simd/avx/n2bv_2.c0000644000175000017500000000015412054411106013144 00000000000000/* Generated automatically. DO NOT EDIT! */ #define SIMD_HEADER "simd-avx.h" #include "../common/n2bv_2.c" fftw-3.3.3/dft/simd/avx/n2bv_4.c0000644000175000017500000000015412054411106013146 00000000000000/* Generated automatically. DO NOT EDIT! */ #define SIMD_HEADER "simd-avx.h" #include "../common/n2bv_4.c" fftw-3.3.3/dft/simd/avx/n2bv_6.c0000644000175000017500000000015412054411106013150 00000000000000/* Generated automatically. DO NOT EDIT! */ #define SIMD_HEADER "simd-avx.h" #include "../common/n2bv_6.c" fftw-3.3.3/dft/simd/avx/n2bv_8.c0000644000175000017500000000015412054411106013152 00000000000000/* Generated automatically. DO NOT EDIT! */ #define SIMD_HEADER "simd-avx.h" #include "../common/n2bv_8.c" fftw-3.3.3/dft/simd/avx/n2bv_10.c0000644000175000017500000000015512054411106013224 00000000000000/* Generated automatically. DO NOT EDIT! */ #define SIMD_HEADER "simd-avx.h" #include "../common/n2bv_10.c" fftw-3.3.3/dft/simd/avx/n2bv_12.c0000644000175000017500000000015512054411106013226 00000000000000/* Generated automatically. DO NOT EDIT! */ #define SIMD_HEADER "simd-avx.h" #include "../common/n2bv_12.c" fftw-3.3.3/dft/simd/avx/n2bv_14.c0000644000175000017500000000015512054411106013230 00000000000000/* Generated automatically. DO NOT EDIT! */ #define SIMD_HEADER "simd-avx.h" #include "../common/n2bv_14.c" fftw-3.3.3/dft/simd/avx/n2bv_16.c0000644000175000017500000000015512054411106013232 00000000000000/* Generated automatically. DO NOT EDIT! */ #define SIMD_HEADER "simd-avx.h" #include "../common/n2bv_16.c" fftw-3.3.3/dft/simd/avx/n2bv_32.c0000644000175000017500000000015512054411106013230 00000000000000/* Generated automatically. DO NOT EDIT! */ #define SIMD_HEADER "simd-avx.h" #include "../common/n2bv_32.c" fftw-3.3.3/dft/simd/avx/n2bv_64.c0000644000175000017500000000015512054411106013235 00000000000000/* Generated automatically. DO NOT EDIT! */ #define SIMD_HEADER "simd-avx.h" #include "../common/n2bv_64.c" fftw-3.3.3/dft/simd/avx/n2bv_20.c0000644000175000017500000000015512054411106013225 00000000000000/* Generated automatically. DO NOT EDIT! */ #define SIMD_HEADER "simd-avx.h" #include "../common/n2bv_20.c" fftw-3.3.3/dft/simd/avx/n2sv_4.c0000644000175000017500000000015412054411106013167 00000000000000/* Generated automatically. DO NOT EDIT! */ #define SIMD_HEADER "simd-avx.h" #include "../common/n2sv_4.c" fftw-3.3.3/dft/simd/avx/n2sv_8.c0000644000175000017500000000015412054411106013173 00000000000000/* Generated automatically. DO NOT EDIT! */ #define SIMD_HEADER "simd-avx.h" #include "../common/n2sv_8.c" fftw-3.3.3/dft/simd/avx/n2sv_16.c0000644000175000017500000000015512054411106013253 00000000000000/* Generated automatically. DO NOT EDIT! */ #define SIMD_HEADER "simd-avx.h" #include "../common/n2sv_16.c" fftw-3.3.3/dft/simd/avx/n2sv_32.c0000644000175000017500000000015512054411106013251 00000000000000/* Generated automatically. DO NOT EDIT! */ #define SIMD_HEADER "simd-avx.h" #include "../common/n2sv_32.c" fftw-3.3.3/dft/simd/avx/n2sv_64.c0000644000175000017500000000015512054411106013256 00000000000000/* Generated automatically. DO NOT EDIT! */ #define SIMD_HEADER "simd-avx.h" #include "../common/n2sv_64.c" fftw-3.3.3/dft/simd/avx/t1fuv_2.c0000644000175000017500000000015512054411106013343 00000000000000/* Generated automatically. DO NOT EDIT! */ #define SIMD_HEADER "simd-avx.h" #include "../common/t1fuv_2.c" fftw-3.3.3/dft/simd/avx/t1fuv_3.c0000644000175000017500000000015512054411106013344 00000000000000/* Generated automatically. DO NOT EDIT! */ #define SIMD_HEADER "simd-avx.h" #include "../common/t1fuv_3.c" fftw-3.3.3/dft/simd/avx/t1fuv_4.c0000644000175000017500000000015512054411106013345 00000000000000/* Generated automatically. DO NOT EDIT! */ #define SIMD_HEADER "simd-avx.h" #include "../common/t1fuv_4.c" fftw-3.3.3/dft/simd/avx/t1fuv_5.c0000644000175000017500000000015512054411106013346 00000000000000/* Generated automatically. DO NOT EDIT! */ #define SIMD_HEADER "simd-avx.h" #include "../common/t1fuv_5.c" fftw-3.3.3/dft/simd/avx/t1fuv_6.c0000644000175000017500000000015512054411106013347 00000000000000/* Generated automatically. DO NOT EDIT! */ #define SIMD_HEADER "simd-avx.h" #include "../common/t1fuv_6.c" fftw-3.3.3/dft/simd/avx/t1fuv_7.c0000644000175000017500000000015512054411106013350 00000000000000/* Generated automatically. DO NOT EDIT! */ #define SIMD_HEADER "simd-avx.h" #include "../common/t1fuv_7.c" fftw-3.3.3/dft/simd/avx/t1fuv_8.c0000644000175000017500000000015512054411106013351 00000000000000/* Generated automatically. DO NOT EDIT! */ #define SIMD_HEADER "simd-avx.h" #include "../common/t1fuv_8.c" fftw-3.3.3/dft/simd/avx/t1fuv_9.c0000644000175000017500000000015512054411106013352 00000000000000/* Generated automatically. DO NOT EDIT! */ #define SIMD_HEADER "simd-avx.h" #include "../common/t1fuv_9.c" fftw-3.3.3/dft/simd/avx/t1fuv_10.c0000644000175000017500000000015612054411106013423 00000000000000/* Generated automatically. DO NOT EDIT! */ #define SIMD_HEADER "simd-avx.h" #include "../common/t1fuv_10.c" fftw-3.3.3/dft/simd/avx/t1fv_2.c0000644000175000017500000000015412054411106013155 00000000000000/* Generated automatically. DO NOT EDIT! */ #define SIMD_HEADER "simd-avx.h" #include "../common/t1fv_2.c" fftw-3.3.3/dft/simd/avx/t1fv_3.c0000644000175000017500000000015412054411106013156 00000000000000/* Generated automatically. DO NOT EDIT! */ #define SIMD_HEADER "simd-avx.h" #include "../common/t1fv_3.c" fftw-3.3.3/dft/simd/avx/t1fv_4.c0000644000175000017500000000015412054411106013157 00000000000000/* Generated automatically. DO NOT EDIT! */ #define SIMD_HEADER "simd-avx.h" #include "../common/t1fv_4.c" fftw-3.3.3/dft/simd/avx/t1fv_5.c0000644000175000017500000000015412054411106013160 00000000000000/* Generated automatically. DO NOT EDIT! */ #define SIMD_HEADER "simd-avx.h" #include "../common/t1fv_5.c" fftw-3.3.3/dft/simd/avx/t1fv_6.c0000644000175000017500000000015412054411106013161 00000000000000/* Generated automatically. DO NOT EDIT! */ #define SIMD_HEADER "simd-avx.h" #include "../common/t1fv_6.c" fftw-3.3.3/dft/simd/avx/t1fv_7.c0000644000175000017500000000015412054411106013162 00000000000000/* Generated automatically. DO NOT EDIT! */ #define SIMD_HEADER "simd-avx.h" #include "../common/t1fv_7.c" fftw-3.3.3/dft/simd/avx/t1fv_8.c0000644000175000017500000000015412054411106013163 00000000000000/* Generated automatically. DO NOT EDIT! */ #define SIMD_HEADER "simd-avx.h" #include "../common/t1fv_8.c" fftw-3.3.3/dft/simd/avx/t1fv_9.c0000644000175000017500000000015412054411106013164 00000000000000/* Generated automatically. DO NOT EDIT! */ #define SIMD_HEADER "simd-avx.h" #include "../common/t1fv_9.c" fftw-3.3.3/dft/simd/avx/t1fv_10.c0000644000175000017500000000015512054411106013235 00000000000000/* Generated automatically. DO NOT EDIT! */ #define SIMD_HEADER "simd-avx.h" #include "../common/t1fv_10.c" fftw-3.3.3/dft/simd/avx/t1fv_12.c0000644000175000017500000000015512054411106013237 00000000000000/* Generated automatically. DO NOT EDIT! */ #define SIMD_HEADER "simd-avx.h" #include "../common/t1fv_12.c" fftw-3.3.3/dft/simd/avx/t1fv_15.c0000644000175000017500000000015512054411106013242 00000000000000/* Generated automatically. DO NOT EDIT! */ #define SIMD_HEADER "simd-avx.h" #include "../common/t1fv_15.c" fftw-3.3.3/dft/simd/avx/t1fv_16.c0000644000175000017500000000015512054411106013243 00000000000000/* Generated automatically. DO NOT EDIT! */ #define SIMD_HEADER "simd-avx.h" #include "../common/t1fv_16.c" fftw-3.3.3/dft/simd/avx/t1fv_32.c0000644000175000017500000000015512054411106013241 00000000000000/* Generated automatically. DO NOT EDIT! */ #define SIMD_HEADER "simd-avx.h" #include "../common/t1fv_32.c" fftw-3.3.3/dft/simd/avx/t1fv_64.c0000644000175000017500000000015512054411106013246 00000000000000/* Generated automatically. DO NOT EDIT! */ #define SIMD_HEADER "simd-avx.h" #include "../common/t1fv_64.c" fftw-3.3.3/dft/simd/avx/t1fv_20.c0000644000175000017500000000015512054411106013236 00000000000000/* Generated automatically. DO NOT EDIT! */ #define SIMD_HEADER "simd-avx.h" #include "../common/t1fv_20.c" fftw-3.3.3/dft/simd/avx/t1fv_25.c0000644000175000017500000000015512054411106013243 00000000000000/* Generated automatically. DO NOT EDIT! */ #define SIMD_HEADER "simd-avx.h" #include "../common/t1fv_25.c" fftw-3.3.3/dft/simd/avx/t2fv_2.c0000644000175000017500000000015412054411106013156 00000000000000/* Generated automatically. DO NOT EDIT! */ #define SIMD_HEADER "simd-avx.h" #include "../common/t2fv_2.c" fftw-3.3.3/dft/simd/avx/t2fv_4.c0000644000175000017500000000015412054411106013160 00000000000000/* Generated automatically. DO NOT EDIT! */ #define SIMD_HEADER "simd-avx.h" #include "../common/t2fv_4.c" fftw-3.3.3/dft/simd/avx/t2fv_8.c0000644000175000017500000000015412054411106013164 00000000000000/* Generated automatically. DO NOT EDIT! */ #define SIMD_HEADER "simd-avx.h" #include "../common/t2fv_8.c" fftw-3.3.3/dft/simd/avx/t2fv_16.c0000644000175000017500000000015512054411106013244 00000000000000/* Generated automatically. DO NOT EDIT! */ #define SIMD_HEADER "simd-avx.h" #include "../common/t2fv_16.c" fftw-3.3.3/dft/simd/avx/t2fv_32.c0000644000175000017500000000015512054411106013242 00000000000000/* Generated automatically. DO NOT EDIT! */ #define SIMD_HEADER "simd-avx.h" #include "../common/t2fv_32.c" fftw-3.3.3/dft/simd/avx/t2fv_64.c0000644000175000017500000000015512054411106013247 00000000000000/* Generated automatically. DO NOT EDIT! */ #define SIMD_HEADER "simd-avx.h" #include "../common/t2fv_64.c" fftw-3.3.3/dft/simd/avx/t2fv_5.c0000644000175000017500000000015412054411106013161 00000000000000/* Generated automatically. DO NOT EDIT! */ #define SIMD_HEADER "simd-avx.h" #include "../common/t2fv_5.c" fftw-3.3.3/dft/simd/avx/t2fv_10.c0000644000175000017500000000015512054411106013236 00000000000000/* Generated automatically. DO NOT EDIT! */ #define SIMD_HEADER "simd-avx.h" #include "../common/t2fv_10.c" fftw-3.3.3/dft/simd/avx/t2fv_20.c0000644000175000017500000000015512054411106013237 00000000000000/* Generated automatically. DO NOT EDIT! */ #define SIMD_HEADER "simd-avx.h" #include "../common/t2fv_20.c" fftw-3.3.3/dft/simd/avx/t2fv_25.c0000644000175000017500000000015512054411106013244 00000000000000/* Generated automatically. DO NOT EDIT! */ #define SIMD_HEADER "simd-avx.h" #include "../common/t2fv_25.c" fftw-3.3.3/dft/simd/avx/t3fv_4.c0000644000175000017500000000015412054411106013161 00000000000000/* Generated automatically. DO NOT EDIT! */ #define SIMD_HEADER "simd-avx.h" #include "../common/t3fv_4.c" fftw-3.3.3/dft/simd/avx/t3fv_8.c0000644000175000017500000000015412054411106013165 00000000000000/* Generated automatically. DO NOT EDIT! */ #define SIMD_HEADER "simd-avx.h" #include "../common/t3fv_8.c" fftw-3.3.3/dft/simd/avx/t3fv_16.c0000644000175000017500000000015512054411106013245 00000000000000/* Generated automatically. DO NOT EDIT! */ #define SIMD_HEADER "simd-avx.h" #include "../common/t3fv_16.c" fftw-3.3.3/dft/simd/avx/t3fv_32.c0000644000175000017500000000015512054411106013243 00000000000000/* Generated automatically. DO NOT EDIT! */ #define SIMD_HEADER "simd-avx.h" #include "../common/t3fv_32.c" fftw-3.3.3/dft/simd/avx/t3fv_5.c0000644000175000017500000000015412054411106013162 00000000000000/* Generated automatically. DO NOT EDIT! */ #define SIMD_HEADER "simd-avx.h" #include "../common/t3fv_5.c" fftw-3.3.3/dft/simd/avx/t3fv_10.c0000644000175000017500000000015512054411106013237 00000000000000/* Generated automatically. DO NOT EDIT! */ #define SIMD_HEADER "simd-avx.h" #include "../common/t3fv_10.c" fftw-3.3.3/dft/simd/avx/t3fv_20.c0000644000175000017500000000015512054411106013240 00000000000000/* Generated automatically. DO NOT EDIT! */ #define SIMD_HEADER "simd-avx.h" #include "../common/t3fv_20.c" fftw-3.3.3/dft/simd/avx/t3fv_25.c0000644000175000017500000000015512054411106013245 00000000000000/* Generated automatically. DO NOT EDIT! */ #define SIMD_HEADER "simd-avx.h" #include "../common/t3fv_25.c" fftw-3.3.3/dft/simd/avx/t1buv_2.c0000644000175000017500000000015512054411106013337 00000000000000/* Generated automatically. DO NOT EDIT! */ #define SIMD_HEADER "simd-avx.h" #include "../common/t1buv_2.c" fftw-3.3.3/dft/simd/avx/t1buv_3.c0000644000175000017500000000015512054411106013340 00000000000000/* Generated automatically. DO NOT EDIT! */ #define SIMD_HEADER "simd-avx.h" #include "../common/t1buv_3.c" fftw-3.3.3/dft/simd/avx/t1buv_4.c0000644000175000017500000000015512054411106013341 00000000000000/* Generated automatically. DO NOT EDIT! */ #define SIMD_HEADER "simd-avx.h" #include "../common/t1buv_4.c" fftw-3.3.3/dft/simd/avx/t1buv_5.c0000644000175000017500000000015512054411106013342 00000000000000/* Generated automatically. DO NOT EDIT! */ #define SIMD_HEADER "simd-avx.h" #include "../common/t1buv_5.c" fftw-3.3.3/dft/simd/avx/t1buv_6.c0000644000175000017500000000015512054411106013343 00000000000000/* Generated automatically. DO NOT EDIT! */ #define SIMD_HEADER "simd-avx.h" #include "../common/t1buv_6.c" fftw-3.3.3/dft/simd/avx/t1buv_7.c0000644000175000017500000000015512054411106013344 00000000000000/* Generated automatically. DO NOT EDIT! */ #define SIMD_HEADER "simd-avx.h" #include "../common/t1buv_7.c" fftw-3.3.3/dft/simd/avx/t1buv_8.c0000644000175000017500000000015512054411106013345 00000000000000/* Generated automatically. DO NOT EDIT! */ #define SIMD_HEADER "simd-avx.h" #include "../common/t1buv_8.c" fftw-3.3.3/dft/simd/avx/t1buv_9.c0000644000175000017500000000015512054411106013346 00000000000000/* Generated automatically. DO NOT EDIT! */ #define SIMD_HEADER "simd-avx.h" #include "../common/t1buv_9.c" fftw-3.3.3/dft/simd/avx/t1buv_10.c0000644000175000017500000000015612054411106013417 00000000000000/* Generated automatically. DO NOT EDIT! */ #define SIMD_HEADER "simd-avx.h" #include "../common/t1buv_10.c" fftw-3.3.3/dft/simd/avx/t1bv_2.c0000644000175000017500000000015412054411106013151 00000000000000/* Generated automatically. DO NOT EDIT! */ #define SIMD_HEADER "simd-avx.h" #include "../common/t1bv_2.c" fftw-3.3.3/dft/simd/avx/t1bv_3.c0000644000175000017500000000015412054411106013152 00000000000000/* Generated automatically. DO NOT EDIT! */ #define SIMD_HEADER "simd-avx.h" #include "../common/t1bv_3.c" fftw-3.3.3/dft/simd/avx/t1bv_4.c0000644000175000017500000000015412054411106013153 00000000000000/* Generated automatically. DO NOT EDIT! */ #define SIMD_HEADER "simd-avx.h" #include "../common/t1bv_4.c" fftw-3.3.3/dft/simd/avx/t1bv_5.c0000644000175000017500000000015412054411106013154 00000000000000/* Generated automatically. DO NOT EDIT! */ #define SIMD_HEADER "simd-avx.h" #include "../common/t1bv_5.c" fftw-3.3.3/dft/simd/avx/t1bv_6.c0000644000175000017500000000015412054411106013155 00000000000000/* Generated automatically. DO NOT EDIT! */ #define SIMD_HEADER "simd-avx.h" #include "../common/t1bv_6.c" fftw-3.3.3/dft/simd/avx/t1bv_7.c0000644000175000017500000000015412054411106013156 00000000000000/* Generated automatically. DO NOT EDIT! */ #define SIMD_HEADER "simd-avx.h" #include "../common/t1bv_7.c" fftw-3.3.3/dft/simd/avx/t1bv_8.c0000644000175000017500000000015412054411106013157 00000000000000/* Generated automatically. DO NOT EDIT! */ #define SIMD_HEADER "simd-avx.h" #include "../common/t1bv_8.c" fftw-3.3.3/dft/simd/avx/t1bv_9.c0000644000175000017500000000015412054411106013160 00000000000000/* Generated automatically. DO NOT EDIT! */ #define SIMD_HEADER "simd-avx.h" #include "../common/t1bv_9.c" fftw-3.3.3/dft/simd/avx/t1bv_10.c0000644000175000017500000000015512054411106013231 00000000000000/* Generated automatically. DO NOT EDIT! */ #define SIMD_HEADER "simd-avx.h" #include "../common/t1bv_10.c" fftw-3.3.3/dft/simd/avx/t1bv_12.c0000644000175000017500000000015512054411106013233 00000000000000/* Generated automatically. DO NOT EDIT! */ #define SIMD_HEADER "simd-avx.h" #include "../common/t1bv_12.c" fftw-3.3.3/dft/simd/avx/t1bv_15.c0000644000175000017500000000015512054411106013236 00000000000000/* Generated automatically. DO NOT EDIT! */ #define SIMD_HEADER "simd-avx.h" #include "../common/t1bv_15.c" fftw-3.3.3/dft/simd/avx/t1bv_16.c0000644000175000017500000000015512054411106013237 00000000000000/* Generated automatically. DO NOT EDIT! */ #define SIMD_HEADER "simd-avx.h" #include "../common/t1bv_16.c" fftw-3.3.3/dft/simd/avx/t1bv_32.c0000644000175000017500000000015512054411106013235 00000000000000/* Generated automatically. DO NOT EDIT! */ #define SIMD_HEADER "simd-avx.h" #include "../common/t1bv_32.c" fftw-3.3.3/dft/simd/avx/t1bv_64.c0000644000175000017500000000015512054411106013242 00000000000000/* Generated automatically. DO NOT EDIT! */ #define SIMD_HEADER "simd-avx.h" #include "../common/t1bv_64.c" fftw-3.3.3/dft/simd/avx/t1bv_20.c0000644000175000017500000000015512054411106013232 00000000000000/* Generated automatically. DO NOT EDIT! */ #define SIMD_HEADER "simd-avx.h" #include "../common/t1bv_20.c" fftw-3.3.3/dft/simd/avx/t1bv_25.c0000644000175000017500000000015512054411106013237 00000000000000/* Generated automatically. DO NOT EDIT! */ #define SIMD_HEADER "simd-avx.h" #include "../common/t1bv_25.c" fftw-3.3.3/dft/simd/avx/t2bv_2.c0000644000175000017500000000015412054411106013152 00000000000000/* Generated automatically. DO NOT EDIT! */ #define SIMD_HEADER "simd-avx.h" #include "../common/t2bv_2.c" fftw-3.3.3/dft/simd/avx/t2bv_4.c0000644000175000017500000000015412054411106013154 00000000000000/* Generated automatically. DO NOT EDIT! */ #define SIMD_HEADER "simd-avx.h" #include "../common/t2bv_4.c" fftw-3.3.3/dft/simd/avx/t2bv_8.c0000644000175000017500000000015412054411106013160 00000000000000/* Generated automatically. DO NOT EDIT! */ #define SIMD_HEADER "simd-avx.h" #include "../common/t2bv_8.c" fftw-3.3.3/dft/simd/avx/t2bv_16.c0000644000175000017500000000015512054411106013240 00000000000000/* Generated automatically. DO NOT EDIT! */ #define SIMD_HEADER "simd-avx.h" #include "../common/t2bv_16.c" fftw-3.3.3/dft/simd/avx/t2bv_32.c0000644000175000017500000000015512054411106013236 00000000000000/* Generated automatically. DO NOT EDIT! */ #define SIMD_HEADER "simd-avx.h" #include "../common/t2bv_32.c" fftw-3.3.3/dft/simd/avx/t2bv_64.c0000644000175000017500000000015512054411106013243 00000000000000/* Generated automatically. DO NOT EDIT! */ #define SIMD_HEADER "simd-avx.h" #include "../common/t2bv_64.c" fftw-3.3.3/dft/simd/avx/t2bv_5.c0000644000175000017500000000015412054411106013155 00000000000000/* Generated automatically. DO NOT EDIT! */ #define SIMD_HEADER "simd-avx.h" #include "../common/t2bv_5.c" fftw-3.3.3/dft/simd/avx/t2bv_10.c0000644000175000017500000000015512054411106013232 00000000000000/* Generated automatically. DO NOT EDIT! */ #define SIMD_HEADER "simd-avx.h" #include "../common/t2bv_10.c" fftw-3.3.3/dft/simd/avx/t2bv_20.c0000644000175000017500000000015512054411106013233 00000000000000/* Generated automatically. DO NOT EDIT! */ #define SIMD_HEADER "simd-avx.h" #include "../common/t2bv_20.c" fftw-3.3.3/dft/simd/avx/t2bv_25.c0000644000175000017500000000015512054411106013240 00000000000000/* Generated automatically. DO NOT EDIT! */ #define SIMD_HEADER "simd-avx.h" #include "../common/t2bv_25.c" fftw-3.3.3/dft/simd/avx/t3bv_4.c0000644000175000017500000000015412054411106013155 00000000000000/* Generated automatically. DO NOT EDIT! */ #define SIMD_HEADER "simd-avx.h" #include "../common/t3bv_4.c" fftw-3.3.3/dft/simd/avx/t3bv_8.c0000644000175000017500000000015412054411106013161 00000000000000/* Generated automatically. DO NOT EDIT! */ #define SIMD_HEADER "simd-avx.h" #include "../common/t3bv_8.c" fftw-3.3.3/dft/simd/avx/t3bv_16.c0000644000175000017500000000015512054411106013241 00000000000000/* Generated automatically. DO NOT EDIT! */ #define SIMD_HEADER "simd-avx.h" #include "../common/t3bv_16.c" fftw-3.3.3/dft/simd/avx/t3bv_32.c0000644000175000017500000000015512054411106013237 00000000000000/* Generated automatically. DO NOT EDIT! */ #define SIMD_HEADER "simd-avx.h" #include "../common/t3bv_32.c" fftw-3.3.3/dft/simd/avx/t3bv_5.c0000644000175000017500000000015412054411106013156 00000000000000/* Generated automatically. DO NOT EDIT! */ #define SIMD_HEADER "simd-avx.h" #include "../common/t3bv_5.c" fftw-3.3.3/dft/simd/avx/t3bv_10.c0000644000175000017500000000015512054411106013233 00000000000000/* Generated automatically. DO NOT EDIT! */ #define SIMD_HEADER "simd-avx.h" #include "../common/t3bv_10.c" fftw-3.3.3/dft/simd/avx/t3bv_20.c0000644000175000017500000000015512054411106013234 00000000000000/* Generated automatically. DO NOT EDIT! */ #define SIMD_HEADER "simd-avx.h" #include "../common/t3bv_20.c" fftw-3.3.3/dft/simd/avx/t3bv_25.c0000644000175000017500000000015512054411106013241 00000000000000/* Generated automatically. DO NOT EDIT! */ #define SIMD_HEADER "simd-avx.h" #include "../common/t3bv_25.c" fftw-3.3.3/dft/simd/avx/t1sv_2.c0000644000175000017500000000015412054411106013172 00000000000000/* Generated automatically. DO NOT EDIT! */ #define SIMD_HEADER "simd-avx.h" #include "../common/t1sv_2.c" fftw-3.3.3/dft/simd/avx/t1sv_4.c0000644000175000017500000000015412054411106013174 00000000000000/* Generated automatically. DO NOT EDIT! */ #define SIMD_HEADER "simd-avx.h" #include "../common/t1sv_4.c" fftw-3.3.3/dft/simd/avx/t1sv_8.c0000644000175000017500000000015412054411106013200 00000000000000/* Generated automatically. DO NOT EDIT! */ #define SIMD_HEADER "simd-avx.h" #include "../common/t1sv_8.c" fftw-3.3.3/dft/simd/avx/t1sv_16.c0000644000175000017500000000015512054411106013260 00000000000000/* Generated automatically. DO NOT EDIT! */ #define SIMD_HEADER "simd-avx.h" #include "../common/t1sv_16.c" fftw-3.3.3/dft/simd/avx/t1sv_32.c0000644000175000017500000000015512054411106013256 00000000000000/* Generated automatically. DO NOT EDIT! */ #define SIMD_HEADER "simd-avx.h" #include "../common/t1sv_32.c" fftw-3.3.3/dft/simd/avx/t2sv_4.c0000644000175000017500000000015412054411106013175 00000000000000/* Generated automatically. DO NOT EDIT! */ #define SIMD_HEADER "simd-avx.h" #include "../common/t2sv_4.c" fftw-3.3.3/dft/simd/avx/t2sv_8.c0000644000175000017500000000015412054411106013201 00000000000000/* Generated automatically. DO NOT EDIT! */ #define SIMD_HEADER "simd-avx.h" #include "../common/t2sv_8.c" fftw-3.3.3/dft/simd/avx/t2sv_16.c0000644000175000017500000000015512054411106013261 00000000000000/* Generated automatically. DO NOT EDIT! */ #define SIMD_HEADER "simd-avx.h" #include "../common/t2sv_16.c" fftw-3.3.3/dft/simd/avx/t2sv_32.c0000644000175000017500000000015512054411106013257 00000000000000/* Generated automatically. DO NOT EDIT! */ #define SIMD_HEADER "simd-avx.h" #include "../common/t2sv_32.c" fftw-3.3.3/dft/simd/avx/q1fv_2.c0000644000175000017500000000015412054411106013152 00000000000000/* Generated automatically. DO NOT EDIT! */ #define SIMD_HEADER "simd-avx.h" #include "../common/q1fv_2.c" fftw-3.3.3/dft/simd/avx/q1fv_4.c0000644000175000017500000000015412054411106013154 00000000000000/* Generated automatically. DO NOT EDIT! */ #define SIMD_HEADER "simd-avx.h" #include "../common/q1fv_4.c" fftw-3.3.3/dft/simd/avx/q1fv_5.c0000644000175000017500000000015412054411106013155 00000000000000/* Generated automatically. DO NOT EDIT! */ #define SIMD_HEADER "simd-avx.h" #include "../common/q1fv_5.c" fftw-3.3.3/dft/simd/avx/q1fv_8.c0000644000175000017500000000015412054411106013160 00000000000000/* Generated automatically. DO NOT EDIT! */ #define SIMD_HEADER "simd-avx.h" #include "../common/q1fv_8.c" fftw-3.3.3/dft/simd/avx/q1bv_2.c0000644000175000017500000000015412054411106013146 00000000000000/* Generated automatically. DO NOT EDIT! */ #define SIMD_HEADER "simd-avx.h" #include "../common/q1bv_2.c" fftw-3.3.3/dft/simd/avx/q1bv_4.c0000644000175000017500000000015412054411106013150 00000000000000/* Generated automatically. DO NOT EDIT! */ #define SIMD_HEADER "simd-avx.h" #include "../common/q1bv_4.c" fftw-3.3.3/dft/simd/avx/q1bv_5.c0000644000175000017500000000015412054411106013151 00000000000000/* Generated automatically. DO NOT EDIT! */ #define SIMD_HEADER "simd-avx.h" #include "../common/q1bv_5.c" fftw-3.3.3/dft/simd/avx/q1bv_8.c0000644000175000017500000000015412054411106013154 00000000000000/* Generated automatically. DO NOT EDIT! */ #define SIMD_HEADER "simd-avx.h" #include "../common/q1bv_8.c" fftw-3.3.3/dft/simd/avx/genus.c0000644000175000017500000000015312054411106013174 00000000000000/* Generated automatically. DO NOT EDIT! */ #define SIMD_HEADER "simd-avx.h" #include "../common/genus.c" fftw-3.3.3/dft/simd/avx/codlist.c0000644000175000017500000000015512054411106013516 00000000000000/* Generated automatically. DO NOT EDIT! */ #define SIMD_HEADER "simd-avx.h" #include "../common/codlist.c" fftw-3.3.3/dft/simd/altivec/0000755000175000017500000000000012054411107012622 500000000000000fftw-3.3.3/dft/simd/altivec/Makefile.am0000644000175000017500000000045012054410050014571 00000000000000AM_CFLAGS = $(ALTIVEC_CFLAGS) SIMD_HEADER=simd-altivec.h include $(top_srcdir)/dft/simd/codlist.mk include $(top_srcdir)/dft/simd/simd.mk if HAVE_ALTIVEC BUILT_SOURCES = $(EXTRA_DIST) noinst_LTLIBRARIES = libdft_altivec_codelets.la libdft_altivec_codelets_la_SOURCES = $(BUILT_SOURCES) endif fftw-3.3.3/dft/simd/altivec/Makefile.in0000644000175000017500000011130112054410177014612 00000000000000# Makefile.in generated by automake 1.11.6 from Makefile.am. # @configure_input@ # Copyright (C) 1994, 1995, 1996, 1997, 1998, 1999, 2000, 2001, 2002, # 2003, 2004, 2005, 2006, 2007, 2008, 2009, 2010, 2011 Free Software # Foundation, Inc. # This Makefile.in is free software; the Free Software Foundation # gives unlimited permission to copy and/or distribute it, # with or without modifications, as long as this notice is preserved. # This program is distributed in the hope that it will be useful, # but WITHOUT ANY WARRANTY, to the extent permitted by law; without # even the implied warranty of MERCHANTABILITY or FITNESS FOR A # PARTICULAR PURPOSE. @SET_MAKE@ # This file contains a standard list of DFT SIMD codelets. It is # included by common/Makefile to generate the C files with the actual # codelets in them. It is included by {sse,sse2,...}/Makefile to # generate and compile stub files that include common/*.c # You can customize FFTW for special needs, e.g. to handle certain # sizes more efficiently, by adding new codelets to the lists of those # included by default. If you change the list of codelets, any new # ones you added will be automatically generated when you run the # bootstrap script (see "Generating your own code" in the FFTW # manual). VPATH = @srcdir@ am__make_dryrun = \ { \ am__dry=no; \ case $$MAKEFLAGS in \ *\\[\ \ ]*) \ echo 'am--echo: ; @echo "AM" OK' | $(MAKE) -f - 2>/dev/null \ | grep '^AM OK$$' >/dev/null || am__dry=yes;; \ *) \ for am__flg in $$MAKEFLAGS; do \ case $$am__flg in \ *=*|--*) ;; \ *n*) am__dry=yes; break;; \ esac; \ done;; \ esac; \ test $$am__dry = yes; \ } pkgdatadir = $(datadir)/@PACKAGE@ pkgincludedir = $(includedir)/@PACKAGE@ pkglibdir = $(libdir)/@PACKAGE@ pkglibexecdir = $(libexecdir)/@PACKAGE@ am__cd = CDPATH="$${ZSH_VERSION+.}$(PATH_SEPARATOR)" && cd install_sh_DATA = $(install_sh) -c -m 644 install_sh_PROGRAM = $(install_sh) -c install_sh_SCRIPT = $(install_sh) -c INSTALL_HEADER = $(INSTALL_DATA) transform = $(program_transform_name) NORMAL_INSTALL = : PRE_INSTALL = : POST_INSTALL = : NORMAL_UNINSTALL = : PRE_UNINSTALL = : POST_UNINSTALL = : build_triplet = @build@ host_triplet = @host@ DIST_COMMON = $(srcdir)/Makefile.am $(srcdir)/Makefile.in \ $(top_srcdir)/dft/simd/codlist.mk \ $(top_srcdir)/dft/simd/simd.mk subdir = dft/simd/altivec ACLOCAL_M4 = $(top_srcdir)/aclocal.m4 am__aclocal_m4_deps = $(top_srcdir)/m4/acx_mpi.m4 \ $(top_srcdir)/m4/acx_pthread.m4 \ $(top_srcdir)/m4/ax_cc_maxopt.m4 \ $(top_srcdir)/m4/ax_check_compiler_flags.m4 \ $(top_srcdir)/m4/ax_compiler_vendor.m4 \ $(top_srcdir)/m4/ax_gcc_aligns_stack.m4 \ $(top_srcdir)/m4/ax_gcc_version.m4 \ $(top_srcdir)/m4/ax_openmp.m4 $(top_srcdir)/m4/libtool.m4 \ $(top_srcdir)/m4/ltoptions.m4 $(top_srcdir)/m4/ltsugar.m4 \ $(top_srcdir)/m4/ltversion.m4 $(top_srcdir)/m4/lt~obsolete.m4 \ $(top_srcdir)/configure.ac am__configure_deps = $(am__aclocal_m4_deps) $(CONFIGURE_DEPENDENCIES) \ $(ACLOCAL_M4) mkinstalldirs = $(install_sh) -d CONFIG_HEADER = $(top_builddir)/config.h CONFIG_CLEAN_FILES = CONFIG_CLEAN_VPATH_FILES = LTLIBRARIES = $(noinst_LTLIBRARIES) libdft_altivec_codelets_la_LIBADD = am__libdft_altivec_codelets_la_SOURCES_DIST = n1fv_2.c n1fv_3.c \ n1fv_4.c n1fv_5.c n1fv_6.c n1fv_7.c n1fv_8.c n1fv_9.c \ n1fv_10.c n1fv_11.c n1fv_12.c n1fv_13.c n1fv_14.c n1fv_15.c \ n1fv_16.c n1fv_32.c n1fv_64.c n1fv_128.c n1fv_20.c n1fv_25.c \ n1bv_2.c n1bv_3.c n1bv_4.c n1bv_5.c n1bv_6.c n1bv_7.c n1bv_8.c \ n1bv_9.c n1bv_10.c n1bv_11.c n1bv_12.c n1bv_13.c n1bv_14.c \ n1bv_15.c n1bv_16.c n1bv_32.c n1bv_64.c n1bv_128.c n1bv_20.c \ n1bv_25.c n2fv_2.c n2fv_4.c n2fv_6.c n2fv_8.c n2fv_10.c \ n2fv_12.c n2fv_14.c n2fv_16.c n2fv_32.c n2fv_64.c n2fv_20.c \ n2bv_2.c n2bv_4.c n2bv_6.c n2bv_8.c n2bv_10.c n2bv_12.c \ n2bv_14.c n2bv_16.c n2bv_32.c n2bv_64.c n2bv_20.c n2sv_4.c \ n2sv_8.c n2sv_16.c n2sv_32.c n2sv_64.c t1fuv_2.c t1fuv_3.c \ t1fuv_4.c t1fuv_5.c t1fuv_6.c t1fuv_7.c t1fuv_8.c t1fuv_9.c \ t1fuv_10.c t1fv_2.c t1fv_3.c t1fv_4.c t1fv_5.c t1fv_6.c \ t1fv_7.c t1fv_8.c t1fv_9.c t1fv_10.c t1fv_12.c t1fv_15.c \ t1fv_16.c t1fv_32.c t1fv_64.c t1fv_20.c t1fv_25.c t2fv_2.c \ t2fv_4.c t2fv_8.c t2fv_16.c t2fv_32.c t2fv_64.c t2fv_5.c \ t2fv_10.c t2fv_20.c t2fv_25.c t3fv_4.c t3fv_8.c t3fv_16.c \ t3fv_32.c t3fv_5.c t3fv_10.c t3fv_20.c t3fv_25.c t1buv_2.c \ t1buv_3.c t1buv_4.c t1buv_5.c t1buv_6.c t1buv_7.c t1buv_8.c \ t1buv_9.c t1buv_10.c t1bv_2.c t1bv_3.c t1bv_4.c t1bv_5.c \ t1bv_6.c t1bv_7.c t1bv_8.c t1bv_9.c t1bv_10.c t1bv_12.c \ t1bv_15.c t1bv_16.c t1bv_32.c t1bv_64.c t1bv_20.c t1bv_25.c \ t2bv_2.c t2bv_4.c t2bv_8.c t2bv_16.c t2bv_32.c t2bv_64.c \ t2bv_5.c t2bv_10.c t2bv_20.c t2bv_25.c t3bv_4.c t3bv_8.c \ t3bv_16.c t3bv_32.c t3bv_5.c t3bv_10.c t3bv_20.c t3bv_25.c \ t1sv_2.c t1sv_4.c t1sv_8.c t1sv_16.c t1sv_32.c t2sv_4.c \ t2sv_8.c t2sv_16.c t2sv_32.c q1fv_2.c q1fv_4.c q1fv_5.c \ q1fv_8.c q1bv_2.c q1bv_4.c q1bv_5.c q1bv_8.c genus.c codlist.c am__objects_1 = n1fv_2.lo n1fv_3.lo n1fv_4.lo n1fv_5.lo n1fv_6.lo \ n1fv_7.lo n1fv_8.lo n1fv_9.lo n1fv_10.lo n1fv_11.lo n1fv_12.lo \ n1fv_13.lo n1fv_14.lo n1fv_15.lo n1fv_16.lo n1fv_32.lo \ n1fv_64.lo n1fv_128.lo n1fv_20.lo n1fv_25.lo am__objects_2 = n1bv_2.lo n1bv_3.lo n1bv_4.lo n1bv_5.lo n1bv_6.lo \ n1bv_7.lo n1bv_8.lo n1bv_9.lo n1bv_10.lo n1bv_11.lo n1bv_12.lo \ n1bv_13.lo n1bv_14.lo n1bv_15.lo n1bv_16.lo n1bv_32.lo \ n1bv_64.lo n1bv_128.lo n1bv_20.lo n1bv_25.lo am__objects_3 = n2fv_2.lo n2fv_4.lo n2fv_6.lo n2fv_8.lo n2fv_10.lo \ n2fv_12.lo n2fv_14.lo n2fv_16.lo n2fv_32.lo n2fv_64.lo \ n2fv_20.lo am__objects_4 = n2bv_2.lo n2bv_4.lo n2bv_6.lo n2bv_8.lo n2bv_10.lo \ n2bv_12.lo n2bv_14.lo n2bv_16.lo n2bv_32.lo n2bv_64.lo \ n2bv_20.lo am__objects_5 = n2sv_4.lo n2sv_8.lo n2sv_16.lo n2sv_32.lo n2sv_64.lo am__objects_6 = t1fuv_2.lo t1fuv_3.lo t1fuv_4.lo t1fuv_5.lo t1fuv_6.lo \ t1fuv_7.lo t1fuv_8.lo t1fuv_9.lo t1fuv_10.lo am__objects_7 = t1fv_2.lo t1fv_3.lo t1fv_4.lo t1fv_5.lo t1fv_6.lo \ t1fv_7.lo t1fv_8.lo t1fv_9.lo t1fv_10.lo t1fv_12.lo t1fv_15.lo \ t1fv_16.lo t1fv_32.lo t1fv_64.lo t1fv_20.lo t1fv_25.lo am__objects_8 = t2fv_2.lo t2fv_4.lo t2fv_8.lo t2fv_16.lo t2fv_32.lo \ t2fv_64.lo t2fv_5.lo t2fv_10.lo t2fv_20.lo t2fv_25.lo am__objects_9 = t3fv_4.lo t3fv_8.lo t3fv_16.lo t3fv_32.lo t3fv_5.lo \ t3fv_10.lo t3fv_20.lo t3fv_25.lo am__objects_10 = t1buv_2.lo t1buv_3.lo t1buv_4.lo t1buv_5.lo \ t1buv_6.lo t1buv_7.lo t1buv_8.lo t1buv_9.lo t1buv_10.lo am__objects_11 = t1bv_2.lo t1bv_3.lo t1bv_4.lo t1bv_5.lo t1bv_6.lo \ t1bv_7.lo t1bv_8.lo t1bv_9.lo t1bv_10.lo t1bv_12.lo t1bv_15.lo \ t1bv_16.lo t1bv_32.lo t1bv_64.lo t1bv_20.lo t1bv_25.lo am__objects_12 = t2bv_2.lo t2bv_4.lo t2bv_8.lo t2bv_16.lo t2bv_32.lo \ t2bv_64.lo t2bv_5.lo t2bv_10.lo t2bv_20.lo t2bv_25.lo am__objects_13 = t3bv_4.lo t3bv_8.lo t3bv_16.lo t3bv_32.lo t3bv_5.lo \ t3bv_10.lo t3bv_20.lo t3bv_25.lo am__objects_14 = t1sv_2.lo t1sv_4.lo t1sv_8.lo t1sv_16.lo t1sv_32.lo am__objects_15 = t2sv_4.lo t2sv_8.lo t2sv_16.lo t2sv_32.lo am__objects_16 = q1fv_2.lo q1fv_4.lo q1fv_5.lo q1fv_8.lo am__objects_17 = q1bv_2.lo q1bv_4.lo q1bv_5.lo q1bv_8.lo am__objects_18 = $(am__objects_1) $(am__objects_2) $(am__objects_3) \ $(am__objects_4) $(am__objects_5) $(am__objects_6) \ $(am__objects_7) $(am__objects_8) $(am__objects_9) \ $(am__objects_10) $(am__objects_11) $(am__objects_12) \ $(am__objects_13) $(am__objects_14) $(am__objects_15) \ $(am__objects_16) $(am__objects_17) am__objects_19 = $(am__objects_18) genus.lo codlist.lo @HAVE_ALTIVEC_TRUE@am__objects_20 = $(am__objects_19) @HAVE_ALTIVEC_TRUE@am_libdft_altivec_codelets_la_OBJECTS = \ @HAVE_ALTIVEC_TRUE@ $(am__objects_20) libdft_altivec_codelets_la_OBJECTS = \ $(am_libdft_altivec_codelets_la_OBJECTS) @HAVE_ALTIVEC_TRUE@am_libdft_altivec_codelets_la_rpath = DEFAULT_INCLUDES = -I.@am__isrc@ -I$(top_builddir) depcomp = $(SHELL) $(top_srcdir)/depcomp am__depfiles_maybe = depfiles am__mv = mv -f COMPILE = $(CC) $(DEFS) $(DEFAULT_INCLUDES) $(INCLUDES) $(AM_CPPFLAGS) \ $(CPPFLAGS) $(AM_CFLAGS) $(CFLAGS) LTCOMPILE = $(LIBTOOL) --tag=CC $(AM_LIBTOOLFLAGS) $(LIBTOOLFLAGS) \ --mode=compile $(CC) $(DEFS) $(DEFAULT_INCLUDES) $(INCLUDES) \ $(AM_CPPFLAGS) $(CPPFLAGS) $(AM_CFLAGS) $(CFLAGS) CCLD = $(CC) LINK = $(LIBTOOL) --tag=CC $(AM_LIBTOOLFLAGS) $(LIBTOOLFLAGS) \ --mode=link $(CCLD) $(AM_CFLAGS) $(CFLAGS) $(AM_LDFLAGS) \ $(LDFLAGS) -o $@ SOURCES = $(libdft_altivec_codelets_la_SOURCES) DIST_SOURCES = $(am__libdft_altivec_codelets_la_SOURCES_DIST) am__can_run_installinfo = \ case $$AM_UPDATE_INFO_DIR in \ n|no|NO) false;; \ *) (install-info --version) >/dev/null 2>&1;; \ esac ETAGS = etags CTAGS = ctags DISTFILES = $(DIST_COMMON) $(DIST_SOURCES) $(TEXINFOS) $(EXTRA_DIST) ACLOCAL = @ACLOCAL@ ALLOCA = @ALLOCA@ ALTIVEC_CFLAGS = @ALTIVEC_CFLAGS@ AMTAR = @AMTAR@ AR = @AR@ AS = @AS@ AUTOCONF = @AUTOCONF@ AUTOHEADER = @AUTOHEADER@ AUTOMAKE = @AUTOMAKE@ AVX_CFLAGS = @AVX_CFLAGS@ AWK = @AWK@ CC = @CC@ CCDEPMODE = @CCDEPMODE@ CFLAGS = @CFLAGS@ CHECK_PL_OPTS = @CHECK_PL_OPTS@ CPP = @CPP@ CPPFLAGS = @CPPFLAGS@ CYGPATH_W = @CYGPATH_W@ C_FFTW_R2R_KIND = @C_FFTW_R2R_KIND@ C_MPI_FINT = @C_MPI_FINT@ DEFS = @DEFS@ DEPDIR = @DEPDIR@ DLLTOOL = @DLLTOOL@ DSYMUTIL = @DSYMUTIL@ DUMPBIN = @DUMPBIN@ ECHO_C = @ECHO_C@ ECHO_N = @ECHO_N@ ECHO_T = @ECHO_T@ EGREP = @EGREP@ EXEEXT = @EXEEXT@ F77 = @F77@ FFLAGS = @FFLAGS@ FGREP = @FGREP@ FLIBS = @FLIBS@ GREP = @GREP@ INSTALL = @INSTALL@ INSTALL_DATA = @INSTALL_DATA@ INSTALL_PROGRAM = @INSTALL_PROGRAM@ INSTALL_SCRIPT = @INSTALL_SCRIPT@ INSTALL_STRIP_PROGRAM = @INSTALL_STRIP_PROGRAM@ LD = @LD@ LDFLAGS = @LDFLAGS@ LIBOBJS = @LIBOBJS@ LIBQUADMATH = @LIBQUADMATH@ LIBS = @LIBS@ LIBTOOL = @LIBTOOL@ LIPO = @LIPO@ LN_S = @LN_S@ LTLIBOBJS = @LTLIBOBJS@ MAINT = @MAINT@ MAKEINFO = @MAKEINFO@ MANIFEST_TOOL = @MANIFEST_TOOL@ MKDIR_P = @MKDIR_P@ MPICC = @MPICC@ MPILIBS = @MPILIBS@ MPIRUN = @MPIRUN@ NEON_CFLAGS = @NEON_CFLAGS@ NM = @NM@ NMEDIT = @NMEDIT@ OBJDUMP = @OBJDUMP@ OBJEXT = @OBJEXT@ OCAMLBUILD = @OCAMLBUILD@ OPENMP_CFLAGS = @OPENMP_CFLAGS@ OTOOL = @OTOOL@ OTOOL64 = @OTOOL64@ PACKAGE = @PACKAGE@ PACKAGE_BUGREPORT = @PACKAGE_BUGREPORT@ PACKAGE_NAME = @PACKAGE_NAME@ PACKAGE_STRING = @PACKAGE_STRING@ PACKAGE_TARNAME = @PACKAGE_TARNAME@ PACKAGE_URL = @PACKAGE_URL@ PACKAGE_VERSION = @PACKAGE_VERSION@ PATH_SEPARATOR = @PATH_SEPARATOR@ POW_LIB = @POW_LIB@ PRECISION = @PRECISION@ PREC_SUFFIX = @PREC_SUFFIX@ PTHREAD_CC = @PTHREAD_CC@ PTHREAD_CFLAGS = @PTHREAD_CFLAGS@ PTHREAD_LIBS = @PTHREAD_LIBS@ RANLIB = @RANLIB@ SED = @SED@ SET_MAKE = @SET_MAKE@ SHARED_VERSION_INFO = @SHARED_VERSION_INFO@ SHELL = @SHELL@ SSE2_CFLAGS = @SSE2_CFLAGS@ STACK_ALIGN_CFLAGS = @STACK_ALIGN_CFLAGS@ STRIP = @STRIP@ THREADLIBS = @THREADLIBS@ VERSION = @VERSION@ abs_builddir = @abs_builddir@ abs_srcdir = @abs_srcdir@ abs_top_builddir = @abs_top_builddir@ abs_top_srcdir = @abs_top_srcdir@ ac_ct_AR = @ac_ct_AR@ ac_ct_CC = @ac_ct_CC@ ac_ct_DUMPBIN = @ac_ct_DUMPBIN@ ac_ct_F77 = @ac_ct_F77@ acx_pthread_config = @acx_pthread_config@ am__include = @am__include@ am__leading_dot = @am__leading_dot@ am__quote = @am__quote@ am__tar = @am__tar@ am__untar = @am__untar@ bindir = @bindir@ build = @build@ build_alias = @build_alias@ build_cpu = @build_cpu@ build_os = @build_os@ build_vendor = @build_vendor@ builddir = @builddir@ datadir = @datadir@ datarootdir = @datarootdir@ docdir = @docdir@ dvidir = @dvidir@ exec_prefix = @exec_prefix@ host = @host@ host_alias = @host_alias@ host_cpu = @host_cpu@ host_os = @host_os@ host_vendor = @host_vendor@ htmldir = @htmldir@ includedir = @includedir@ infodir = @infodir@ install_sh = @install_sh@ libdir = @libdir@ libexecdir = @libexecdir@ localedir = @localedir@ localstatedir = @localstatedir@ mandir = @mandir@ mkdir_p = @mkdir_p@ oldincludedir = @oldincludedir@ pdfdir = @pdfdir@ prefix = @prefix@ program_transform_name = @program_transform_name@ psdir = @psdir@ sbindir = @sbindir@ sharedstatedir = @sharedstatedir@ srcdir = @srcdir@ sysconfdir = @sysconfdir@ target_alias = @target_alias@ top_build_prefix = @top_build_prefix@ top_builddir = @top_builddir@ top_srcdir = @top_srcdir@ AM_CFLAGS = $(ALTIVEC_CFLAGS) SIMD_HEADER = simd-altivec.h ########################################################################### # n1fv_ is a hard-coded FFTW_FORWARD FFT of size , using SIMD N1F = n1fv_2.c n1fv_3.c n1fv_4.c n1fv_5.c n1fv_6.c n1fv_7.c n1fv_8.c \ n1fv_9.c n1fv_10.c n1fv_11.c n1fv_12.c n1fv_13.c n1fv_14.c n1fv_15.c \ n1fv_16.c n1fv_32.c n1fv_64.c n1fv_128.c n1fv_20.c n1fv_25.c # as above, with restricted input vector stride N2F = n2fv_2.c n2fv_4.c n2fv_6.c n2fv_8.c n2fv_10.c n2fv_12.c \ n2fv_14.c n2fv_16.c n2fv_32.c n2fv_64.c n2fv_20.c # as above, but FFTW_BACKWARD N1B = n1bv_2.c n1bv_3.c n1bv_4.c n1bv_5.c n1bv_6.c n1bv_7.c n1bv_8.c \ n1bv_9.c n1bv_10.c n1bv_11.c n1bv_12.c n1bv_13.c n1bv_14.c n1bv_15.c \ n1bv_16.c n1bv_32.c n1bv_64.c n1bv_128.c n1bv_20.c n1bv_25.c N2B = n2bv_2.c n2bv_4.c n2bv_6.c n2bv_8.c n2bv_10.c n2bv_12.c \ n2bv_14.c n2bv_16.c n2bv_32.c n2bv_64.c n2bv_20.c # split-complex codelets N2S = n2sv_4.c n2sv_8.c n2sv_16.c n2sv_32.c n2sv_64.c ########################################################################### # t1fv_ is a "twiddle" FFT of size , implementing a radix-r DIT step # for an FFTW_FORWARD transform, using SIMD T1F = t1fv_2.c t1fv_3.c t1fv_4.c t1fv_5.c t1fv_6.c t1fv_7.c t1fv_8.c \ t1fv_9.c t1fv_10.c t1fv_12.c t1fv_15.c t1fv_16.c t1fv_32.c t1fv_64.c \ t1fv_20.c t1fv_25.c # same as t1fv_*, but with different twiddle storage scheme T2F = t2fv_2.c t2fv_4.c t2fv_8.c t2fv_16.c t2fv_32.c t2fv_64.c \ t2fv_5.c t2fv_10.c t2fv_20.c t2fv_25.c T3F = t3fv_4.c t3fv_8.c t3fv_16.c t3fv_32.c t3fv_5.c t3fv_10.c \ t3fv_20.c t3fv_25.c T1FU = t1fuv_2.c t1fuv_3.c t1fuv_4.c t1fuv_5.c t1fuv_6.c t1fuv_7.c \ t1fuv_8.c t1fuv_9.c t1fuv_10.c # as above, but FFTW_BACKWARD T1B = t1bv_2.c t1bv_3.c t1bv_4.c t1bv_5.c t1bv_6.c t1bv_7.c t1bv_8.c \ t1bv_9.c t1bv_10.c t1bv_12.c t1bv_15.c t1bv_16.c t1bv_32.c t1bv_64.c \ t1bv_20.c t1bv_25.c # same as t1bv_*, but with different twiddle storage scheme T2B = t2bv_2.c t2bv_4.c t2bv_8.c t2bv_16.c t2bv_32.c t2bv_64.c \ t2bv_5.c t2bv_10.c t2bv_20.c t2bv_25.c T3B = t3bv_4.c t3bv_8.c t3bv_16.c t3bv_32.c t3bv_5.c t3bv_10.c \ t3bv_20.c t3bv_25.c T1BU = t1buv_2.c t1buv_3.c t1buv_4.c t1buv_5.c t1buv_6.c t1buv_7.c \ t1buv_8.c t1buv_9.c t1buv_10.c # split-complex codelets T1S = t1sv_2.c t1sv_4.c t1sv_8.c t1sv_16.c t1sv_32.c T2S = t2sv_4.c t2sv_8.c t2sv_16.c t2sv_32.c ########################################################################### # q1fv_ is twiddle FFTW_FORWARD FFTs of size (DIF step), # where the output is transposed, using SIMD. This is used for # in-place transposes in sizes that are divisible by ^2. These # codelets have size ~ ^2, so you should probably not use # bigger than 8 or so. Q1F = q1fv_2.c q1fv_4.c q1fv_5.c q1fv_8.c # as above, but FFTW_BACKWARD Q1B = q1bv_2.c q1bv_4.c q1bv_5.c q1bv_8.c ########################################################################### SIMD_CODELETS = $(N1F) $(N1B) $(N2F) $(N2B) $(N2S) $(T1FU) $(T1F) \ $(T2F) $(T3F) $(T1BU) $(T1B) $(T2B) $(T3B) $(T1S) $(T2S) $(Q1F) $(Q1B) AM_CPPFLAGS = -I$(top_srcdir)/kernel -I$(top_srcdir)/dft \ -I$(top_srcdir)/dft/simd -I$(top_srcdir)/simd-support EXTRA_DIST = $(SIMD_CODELETS) genus.c codlist.c @HAVE_ALTIVEC_TRUE@BUILT_SOURCES = $(EXTRA_DIST) @HAVE_ALTIVEC_TRUE@noinst_LTLIBRARIES = libdft_altivec_codelets.la @HAVE_ALTIVEC_TRUE@libdft_altivec_codelets_la_SOURCES = $(BUILT_SOURCES) all: $(BUILT_SOURCES) $(MAKE) $(AM_MAKEFLAGS) all-am .SUFFIXES: .SUFFIXES: .c .lo .o .obj $(srcdir)/Makefile.in: @MAINTAINER_MODE_TRUE@ $(srcdir)/Makefile.am $(top_srcdir)/dft/simd/codlist.mk $(top_srcdir)/dft/simd/simd.mk $(am__configure_deps) @for dep in $?; do \ case '$(am__configure_deps)' in \ *$$dep*) \ ( cd $(top_builddir) && $(MAKE) $(AM_MAKEFLAGS) am--refresh ) \ && { if test -f $@; then exit 0; else break; fi; }; \ exit 1;; \ esac; \ done; \ echo ' cd $(top_srcdir) && $(AUTOMAKE) --gnu dft/simd/altivec/Makefile'; \ $(am__cd) $(top_srcdir) && \ $(AUTOMAKE) --gnu dft/simd/altivec/Makefile .PRECIOUS: Makefile Makefile: $(srcdir)/Makefile.in $(top_builddir)/config.status @case '$?' in \ *config.status*) \ cd $(top_builddir) && $(MAKE) $(AM_MAKEFLAGS) am--refresh;; \ *) \ echo ' cd $(top_builddir) && $(SHELL) ./config.status $(subdir)/$@ $(am__depfiles_maybe)'; \ cd $(top_builddir) && $(SHELL) ./config.status $(subdir)/$@ $(am__depfiles_maybe);; \ esac; $(top_srcdir)/dft/simd/codlist.mk $(top_srcdir)/dft/simd/simd.mk: $(top_builddir)/config.status: $(top_srcdir)/configure $(CONFIG_STATUS_DEPENDENCIES) cd $(top_builddir) && $(MAKE) $(AM_MAKEFLAGS) am--refresh $(top_srcdir)/configure: @MAINTAINER_MODE_TRUE@ $(am__configure_deps) cd $(top_builddir) && $(MAKE) $(AM_MAKEFLAGS) am--refresh $(ACLOCAL_M4): @MAINTAINER_MODE_TRUE@ $(am__aclocal_m4_deps) cd $(top_builddir) && $(MAKE) $(AM_MAKEFLAGS) am--refresh $(am__aclocal_m4_deps): clean-noinstLTLIBRARIES: -test -z "$(noinst_LTLIBRARIES)" || rm -f $(noinst_LTLIBRARIES) @list='$(noinst_LTLIBRARIES)'; for p in $$list; do \ dir="`echo $$p | sed -e 's|/[^/]*$$||'`"; \ test "$$dir" != "$$p" || dir=.; \ echo "rm -f \"$${dir}/so_locations\""; \ rm -f "$${dir}/so_locations"; \ done libdft_altivec_codelets.la: $(libdft_altivec_codelets_la_OBJECTS) $(libdft_altivec_codelets_la_DEPENDENCIES) $(EXTRA_libdft_altivec_codelets_la_DEPENDENCIES) $(LINK) $(am_libdft_altivec_codelets_la_rpath) $(libdft_altivec_codelets_la_OBJECTS) $(libdft_altivec_codelets_la_LIBADD) $(LIBS) mostlyclean-compile: -rm -f *.$(OBJEXT) distclean-compile: -rm -f *.tab.c @AMDEP_TRUE@@am__include@ @am__quote@./$(DEPDIR)/codlist.Plo@am__quote@ @AMDEP_TRUE@@am__include@ @am__quote@./$(DEPDIR)/genus.Plo@am__quote@ @AMDEP_TRUE@@am__include@ @am__quote@./$(DEPDIR)/n1bv_10.Plo@am__quote@ @AMDEP_TRUE@@am__include@ @am__quote@./$(DEPDIR)/n1bv_11.Plo@am__quote@ @AMDEP_TRUE@@am__include@ @am__quote@./$(DEPDIR)/n1bv_12.Plo@am__quote@ @AMDEP_TRUE@@am__include@ @am__quote@./$(DEPDIR)/n1bv_128.Plo@am__quote@ @AMDEP_TRUE@@am__include@ @am__quote@./$(DEPDIR)/n1bv_13.Plo@am__quote@ @AMDEP_TRUE@@am__include@ @am__quote@./$(DEPDIR)/n1bv_14.Plo@am__quote@ @AMDEP_TRUE@@am__include@ @am__quote@./$(DEPDIR)/n1bv_15.Plo@am__quote@ @AMDEP_TRUE@@am__include@ @am__quote@./$(DEPDIR)/n1bv_16.Plo@am__quote@ @AMDEP_TRUE@@am__include@ @am__quote@./$(DEPDIR)/n1bv_2.Plo@am__quote@ @AMDEP_TRUE@@am__include@ @am__quote@./$(DEPDIR)/n1bv_20.Plo@am__quote@ @AMDEP_TRUE@@am__include@ @am__quote@./$(DEPDIR)/n1bv_25.Plo@am__quote@ @AMDEP_TRUE@@am__include@ @am__quote@./$(DEPDIR)/n1bv_3.Plo@am__quote@ @AMDEP_TRUE@@am__include@ @am__quote@./$(DEPDIR)/n1bv_32.Plo@am__quote@ @AMDEP_TRUE@@am__include@ @am__quote@./$(DEPDIR)/n1bv_4.Plo@am__quote@ @AMDEP_TRUE@@am__include@ @am__quote@./$(DEPDIR)/n1bv_5.Plo@am__quote@ @AMDEP_TRUE@@am__include@ @am__quote@./$(DEPDIR)/n1bv_6.Plo@am__quote@ @AMDEP_TRUE@@am__include@ @am__quote@./$(DEPDIR)/n1bv_64.Plo@am__quote@ @AMDEP_TRUE@@am__include@ @am__quote@./$(DEPDIR)/n1bv_7.Plo@am__quote@ @AMDEP_TRUE@@am__include@ @am__quote@./$(DEPDIR)/n1bv_8.Plo@am__quote@ @AMDEP_TRUE@@am__include@ @am__quote@./$(DEPDIR)/n1bv_9.Plo@am__quote@ @AMDEP_TRUE@@am__include@ @am__quote@./$(DEPDIR)/n1fv_10.Plo@am__quote@ @AMDEP_TRUE@@am__include@ @am__quote@./$(DEPDIR)/n1fv_11.Plo@am__quote@ @AMDEP_TRUE@@am__include@ @am__quote@./$(DEPDIR)/n1fv_12.Plo@am__quote@ @AMDEP_TRUE@@am__include@ @am__quote@./$(DEPDIR)/n1fv_128.Plo@am__quote@ @AMDEP_TRUE@@am__include@ @am__quote@./$(DEPDIR)/n1fv_13.Plo@am__quote@ @AMDEP_TRUE@@am__include@ @am__quote@./$(DEPDIR)/n1fv_14.Plo@am__quote@ @AMDEP_TRUE@@am__include@ @am__quote@./$(DEPDIR)/n1fv_15.Plo@am__quote@ @AMDEP_TRUE@@am__include@ @am__quote@./$(DEPDIR)/n1fv_16.Plo@am__quote@ @AMDEP_TRUE@@am__include@ @am__quote@./$(DEPDIR)/n1fv_2.Plo@am__quote@ @AMDEP_TRUE@@am__include@ @am__quote@./$(DEPDIR)/n1fv_20.Plo@am__quote@ @AMDEP_TRUE@@am__include@ @am__quote@./$(DEPDIR)/n1fv_25.Plo@am__quote@ @AMDEP_TRUE@@am__include@ @am__quote@./$(DEPDIR)/n1fv_3.Plo@am__quote@ @AMDEP_TRUE@@am__include@ @am__quote@./$(DEPDIR)/n1fv_32.Plo@am__quote@ @AMDEP_TRUE@@am__include@ @am__quote@./$(DEPDIR)/n1fv_4.Plo@am__quote@ @AMDEP_TRUE@@am__include@ @am__quote@./$(DEPDIR)/n1fv_5.Plo@am__quote@ @AMDEP_TRUE@@am__include@ @am__quote@./$(DEPDIR)/n1fv_6.Plo@am__quote@ @AMDEP_TRUE@@am__include@ @am__quote@./$(DEPDIR)/n1fv_64.Plo@am__quote@ @AMDEP_TRUE@@am__include@ @am__quote@./$(DEPDIR)/n1fv_7.Plo@am__quote@ @AMDEP_TRUE@@am__include@ @am__quote@./$(DEPDIR)/n1fv_8.Plo@am__quote@ @AMDEP_TRUE@@am__include@ @am__quote@./$(DEPDIR)/n1fv_9.Plo@am__quote@ @AMDEP_TRUE@@am__include@ @am__quote@./$(DEPDIR)/n2bv_10.Plo@am__quote@ @AMDEP_TRUE@@am__include@ @am__quote@./$(DEPDIR)/n2bv_12.Plo@am__quote@ @AMDEP_TRUE@@am__include@ @am__quote@./$(DEPDIR)/n2bv_14.Plo@am__quote@ @AMDEP_TRUE@@am__include@ @am__quote@./$(DEPDIR)/n2bv_16.Plo@am__quote@ @AMDEP_TRUE@@am__include@ @am__quote@./$(DEPDIR)/n2bv_2.Plo@am__quote@ @AMDEP_TRUE@@am__include@ @am__quote@./$(DEPDIR)/n2bv_20.Plo@am__quote@ @AMDEP_TRUE@@am__include@ @am__quote@./$(DEPDIR)/n2bv_32.Plo@am__quote@ @AMDEP_TRUE@@am__include@ @am__quote@./$(DEPDIR)/n2bv_4.Plo@am__quote@ @AMDEP_TRUE@@am__include@ @am__quote@./$(DEPDIR)/n2bv_6.Plo@am__quote@ @AMDEP_TRUE@@am__include@ @am__quote@./$(DEPDIR)/n2bv_64.Plo@am__quote@ @AMDEP_TRUE@@am__include@ @am__quote@./$(DEPDIR)/n2bv_8.Plo@am__quote@ @AMDEP_TRUE@@am__include@ @am__quote@./$(DEPDIR)/n2fv_10.Plo@am__quote@ @AMDEP_TRUE@@am__include@ @am__quote@./$(DEPDIR)/n2fv_12.Plo@am__quote@ @AMDEP_TRUE@@am__include@ @am__quote@./$(DEPDIR)/n2fv_14.Plo@am__quote@ @AMDEP_TRUE@@am__include@ @am__quote@./$(DEPDIR)/n2fv_16.Plo@am__quote@ @AMDEP_TRUE@@am__include@ @am__quote@./$(DEPDIR)/n2fv_2.Plo@am__quote@ @AMDEP_TRUE@@am__include@ @am__quote@./$(DEPDIR)/n2fv_20.Plo@am__quote@ @AMDEP_TRUE@@am__include@ @am__quote@./$(DEPDIR)/n2fv_32.Plo@am__quote@ @AMDEP_TRUE@@am__include@ @am__quote@./$(DEPDIR)/n2fv_4.Plo@am__quote@ @AMDEP_TRUE@@am__include@ @am__quote@./$(DEPDIR)/n2fv_6.Plo@am__quote@ @AMDEP_TRUE@@am__include@ @am__quote@./$(DEPDIR)/n2fv_64.Plo@am__quote@ @AMDEP_TRUE@@am__include@ @am__quote@./$(DEPDIR)/n2fv_8.Plo@am__quote@ @AMDEP_TRUE@@am__include@ @am__quote@./$(DEPDIR)/n2sv_16.Plo@am__quote@ @AMDEP_TRUE@@am__include@ @am__quote@./$(DEPDIR)/n2sv_32.Plo@am__quote@ @AMDEP_TRUE@@am__include@ @am__quote@./$(DEPDIR)/n2sv_4.Plo@am__quote@ @AMDEP_TRUE@@am__include@ @am__quote@./$(DEPDIR)/n2sv_64.Plo@am__quote@ @AMDEP_TRUE@@am__include@ @am__quote@./$(DEPDIR)/n2sv_8.Plo@am__quote@ @AMDEP_TRUE@@am__include@ @am__quote@./$(DEPDIR)/q1bv_2.Plo@am__quote@ @AMDEP_TRUE@@am__include@ @am__quote@./$(DEPDIR)/q1bv_4.Plo@am__quote@ @AMDEP_TRUE@@am__include@ @am__quote@./$(DEPDIR)/q1bv_5.Plo@am__quote@ @AMDEP_TRUE@@am__include@ @am__quote@./$(DEPDIR)/q1bv_8.Plo@am__quote@ @AMDEP_TRUE@@am__include@ @am__quote@./$(DEPDIR)/q1fv_2.Plo@am__quote@ @AMDEP_TRUE@@am__include@ @am__quote@./$(DEPDIR)/q1fv_4.Plo@am__quote@ @AMDEP_TRUE@@am__include@ @am__quote@./$(DEPDIR)/q1fv_5.Plo@am__quote@ @AMDEP_TRUE@@am__include@ @am__quote@./$(DEPDIR)/q1fv_8.Plo@am__quote@ @AMDEP_TRUE@@am__include@ @am__quote@./$(DEPDIR)/t1buv_10.Plo@am__quote@ @AMDEP_TRUE@@am__include@ @am__quote@./$(DEPDIR)/t1buv_2.Plo@am__quote@ @AMDEP_TRUE@@am__include@ @am__quote@./$(DEPDIR)/t1buv_3.Plo@am__quote@ @AMDEP_TRUE@@am__include@ @am__quote@./$(DEPDIR)/t1buv_4.Plo@am__quote@ @AMDEP_TRUE@@am__include@ @am__quote@./$(DEPDIR)/t1buv_5.Plo@am__quote@ @AMDEP_TRUE@@am__include@ @am__quote@./$(DEPDIR)/t1buv_6.Plo@am__quote@ @AMDEP_TRUE@@am__include@ @am__quote@./$(DEPDIR)/t1buv_7.Plo@am__quote@ @AMDEP_TRUE@@am__include@ @am__quote@./$(DEPDIR)/t1buv_8.Plo@am__quote@ @AMDEP_TRUE@@am__include@ @am__quote@./$(DEPDIR)/t1buv_9.Plo@am__quote@ @AMDEP_TRUE@@am__include@ @am__quote@./$(DEPDIR)/t1bv_10.Plo@am__quote@ @AMDEP_TRUE@@am__include@ @am__quote@./$(DEPDIR)/t1bv_12.Plo@am__quote@ @AMDEP_TRUE@@am__include@ @am__quote@./$(DEPDIR)/t1bv_15.Plo@am__quote@ @AMDEP_TRUE@@am__include@ @am__quote@./$(DEPDIR)/t1bv_16.Plo@am__quote@ @AMDEP_TRUE@@am__include@ @am__quote@./$(DEPDIR)/t1bv_2.Plo@am__quote@ @AMDEP_TRUE@@am__include@ @am__quote@./$(DEPDIR)/t1bv_20.Plo@am__quote@ @AMDEP_TRUE@@am__include@ @am__quote@./$(DEPDIR)/t1bv_25.Plo@am__quote@ @AMDEP_TRUE@@am__include@ @am__quote@./$(DEPDIR)/t1bv_3.Plo@am__quote@ @AMDEP_TRUE@@am__include@ @am__quote@./$(DEPDIR)/t1bv_32.Plo@am__quote@ @AMDEP_TRUE@@am__include@ @am__quote@./$(DEPDIR)/t1bv_4.Plo@am__quote@ @AMDEP_TRUE@@am__include@ @am__quote@./$(DEPDIR)/t1bv_5.Plo@am__quote@ @AMDEP_TRUE@@am__include@ @am__quote@./$(DEPDIR)/t1bv_6.Plo@am__quote@ @AMDEP_TRUE@@am__include@ @am__quote@./$(DEPDIR)/t1bv_64.Plo@am__quote@ @AMDEP_TRUE@@am__include@ @am__quote@./$(DEPDIR)/t1bv_7.Plo@am__quote@ @AMDEP_TRUE@@am__include@ @am__quote@./$(DEPDIR)/t1bv_8.Plo@am__quote@ @AMDEP_TRUE@@am__include@ @am__quote@./$(DEPDIR)/t1bv_9.Plo@am__quote@ @AMDEP_TRUE@@am__include@ @am__quote@./$(DEPDIR)/t1fuv_10.Plo@am__quote@ @AMDEP_TRUE@@am__include@ @am__quote@./$(DEPDIR)/t1fuv_2.Plo@am__quote@ @AMDEP_TRUE@@am__include@ @am__quote@./$(DEPDIR)/t1fuv_3.Plo@am__quote@ @AMDEP_TRUE@@am__include@ @am__quote@./$(DEPDIR)/t1fuv_4.Plo@am__quote@ @AMDEP_TRUE@@am__include@ @am__quote@./$(DEPDIR)/t1fuv_5.Plo@am__quote@ @AMDEP_TRUE@@am__include@ @am__quote@./$(DEPDIR)/t1fuv_6.Plo@am__quote@ @AMDEP_TRUE@@am__include@ @am__quote@./$(DEPDIR)/t1fuv_7.Plo@am__quote@ @AMDEP_TRUE@@am__include@ @am__quote@./$(DEPDIR)/t1fuv_8.Plo@am__quote@ @AMDEP_TRUE@@am__include@ @am__quote@./$(DEPDIR)/t1fuv_9.Plo@am__quote@ @AMDEP_TRUE@@am__include@ @am__quote@./$(DEPDIR)/t1fv_10.Plo@am__quote@ @AMDEP_TRUE@@am__include@ @am__quote@./$(DEPDIR)/t1fv_12.Plo@am__quote@ @AMDEP_TRUE@@am__include@ @am__quote@./$(DEPDIR)/t1fv_15.Plo@am__quote@ @AMDEP_TRUE@@am__include@ @am__quote@./$(DEPDIR)/t1fv_16.Plo@am__quote@ @AMDEP_TRUE@@am__include@ @am__quote@./$(DEPDIR)/t1fv_2.Plo@am__quote@ @AMDEP_TRUE@@am__include@ @am__quote@./$(DEPDIR)/t1fv_20.Plo@am__quote@ @AMDEP_TRUE@@am__include@ @am__quote@./$(DEPDIR)/t1fv_25.Plo@am__quote@ @AMDEP_TRUE@@am__include@ @am__quote@./$(DEPDIR)/t1fv_3.Plo@am__quote@ @AMDEP_TRUE@@am__include@ @am__quote@./$(DEPDIR)/t1fv_32.Plo@am__quote@ @AMDEP_TRUE@@am__include@ @am__quote@./$(DEPDIR)/t1fv_4.Plo@am__quote@ @AMDEP_TRUE@@am__include@ @am__quote@./$(DEPDIR)/t1fv_5.Plo@am__quote@ @AMDEP_TRUE@@am__include@ @am__quote@./$(DEPDIR)/t1fv_6.Plo@am__quote@ @AMDEP_TRUE@@am__include@ @am__quote@./$(DEPDIR)/t1fv_64.Plo@am__quote@ @AMDEP_TRUE@@am__include@ @am__quote@./$(DEPDIR)/t1fv_7.Plo@am__quote@ @AMDEP_TRUE@@am__include@ @am__quote@./$(DEPDIR)/t1fv_8.Plo@am__quote@ @AMDEP_TRUE@@am__include@ @am__quote@./$(DEPDIR)/t1fv_9.Plo@am__quote@ @AMDEP_TRUE@@am__include@ @am__quote@./$(DEPDIR)/t1sv_16.Plo@am__quote@ @AMDEP_TRUE@@am__include@ @am__quote@./$(DEPDIR)/t1sv_2.Plo@am__quote@ @AMDEP_TRUE@@am__include@ @am__quote@./$(DEPDIR)/t1sv_32.Plo@am__quote@ @AMDEP_TRUE@@am__include@ @am__quote@./$(DEPDIR)/t1sv_4.Plo@am__quote@ @AMDEP_TRUE@@am__include@ @am__quote@./$(DEPDIR)/t1sv_8.Plo@am__quote@ @AMDEP_TRUE@@am__include@ @am__quote@./$(DEPDIR)/t2bv_10.Plo@am__quote@ @AMDEP_TRUE@@am__include@ @am__quote@./$(DEPDIR)/t2bv_16.Plo@am__quote@ @AMDEP_TRUE@@am__include@ @am__quote@./$(DEPDIR)/t2bv_2.Plo@am__quote@ @AMDEP_TRUE@@am__include@ @am__quote@./$(DEPDIR)/t2bv_20.Plo@am__quote@ @AMDEP_TRUE@@am__include@ @am__quote@./$(DEPDIR)/t2bv_25.Plo@am__quote@ @AMDEP_TRUE@@am__include@ @am__quote@./$(DEPDIR)/t2bv_32.Plo@am__quote@ @AMDEP_TRUE@@am__include@ @am__quote@./$(DEPDIR)/t2bv_4.Plo@am__quote@ @AMDEP_TRUE@@am__include@ @am__quote@./$(DEPDIR)/t2bv_5.Plo@am__quote@ @AMDEP_TRUE@@am__include@ @am__quote@./$(DEPDIR)/t2bv_64.Plo@am__quote@ @AMDEP_TRUE@@am__include@ @am__quote@./$(DEPDIR)/t2bv_8.Plo@am__quote@ @AMDEP_TRUE@@am__include@ @am__quote@./$(DEPDIR)/t2fv_10.Plo@am__quote@ @AMDEP_TRUE@@am__include@ @am__quote@./$(DEPDIR)/t2fv_16.Plo@am__quote@ @AMDEP_TRUE@@am__include@ @am__quote@./$(DEPDIR)/t2fv_2.Plo@am__quote@ @AMDEP_TRUE@@am__include@ @am__quote@./$(DEPDIR)/t2fv_20.Plo@am__quote@ @AMDEP_TRUE@@am__include@ @am__quote@./$(DEPDIR)/t2fv_25.Plo@am__quote@ @AMDEP_TRUE@@am__include@ @am__quote@./$(DEPDIR)/t2fv_32.Plo@am__quote@ @AMDEP_TRUE@@am__include@ @am__quote@./$(DEPDIR)/t2fv_4.Plo@am__quote@ @AMDEP_TRUE@@am__include@ @am__quote@./$(DEPDIR)/t2fv_5.Plo@am__quote@ @AMDEP_TRUE@@am__include@ @am__quote@./$(DEPDIR)/t2fv_64.Plo@am__quote@ @AMDEP_TRUE@@am__include@ @am__quote@./$(DEPDIR)/t2fv_8.Plo@am__quote@ @AMDEP_TRUE@@am__include@ @am__quote@./$(DEPDIR)/t2sv_16.Plo@am__quote@ @AMDEP_TRUE@@am__include@ @am__quote@./$(DEPDIR)/t2sv_32.Plo@am__quote@ @AMDEP_TRUE@@am__include@ @am__quote@./$(DEPDIR)/t2sv_4.Plo@am__quote@ @AMDEP_TRUE@@am__include@ @am__quote@./$(DEPDIR)/t2sv_8.Plo@am__quote@ @AMDEP_TRUE@@am__include@ @am__quote@./$(DEPDIR)/t3bv_10.Plo@am__quote@ @AMDEP_TRUE@@am__include@ @am__quote@./$(DEPDIR)/t3bv_16.Plo@am__quote@ @AMDEP_TRUE@@am__include@ @am__quote@./$(DEPDIR)/t3bv_20.Plo@am__quote@ @AMDEP_TRUE@@am__include@ @am__quote@./$(DEPDIR)/t3bv_25.Plo@am__quote@ @AMDEP_TRUE@@am__include@ @am__quote@./$(DEPDIR)/t3bv_32.Plo@am__quote@ @AMDEP_TRUE@@am__include@ @am__quote@./$(DEPDIR)/t3bv_4.Plo@am__quote@ @AMDEP_TRUE@@am__include@ @am__quote@./$(DEPDIR)/t3bv_5.Plo@am__quote@ @AMDEP_TRUE@@am__include@ @am__quote@./$(DEPDIR)/t3bv_8.Plo@am__quote@ @AMDEP_TRUE@@am__include@ @am__quote@./$(DEPDIR)/t3fv_10.Plo@am__quote@ @AMDEP_TRUE@@am__include@ @am__quote@./$(DEPDIR)/t3fv_16.Plo@am__quote@ @AMDEP_TRUE@@am__include@ @am__quote@./$(DEPDIR)/t3fv_20.Plo@am__quote@ @AMDEP_TRUE@@am__include@ @am__quote@./$(DEPDIR)/t3fv_25.Plo@am__quote@ @AMDEP_TRUE@@am__include@ @am__quote@./$(DEPDIR)/t3fv_32.Plo@am__quote@ @AMDEP_TRUE@@am__include@ @am__quote@./$(DEPDIR)/t3fv_4.Plo@am__quote@ @AMDEP_TRUE@@am__include@ @am__quote@./$(DEPDIR)/t3fv_5.Plo@am__quote@ @AMDEP_TRUE@@am__include@ @am__quote@./$(DEPDIR)/t3fv_8.Plo@am__quote@ .c.o: @am__fastdepCC_TRUE@ $(COMPILE) -MT $@ -MD -MP -MF $(DEPDIR)/$*.Tpo -c -o $@ $< @am__fastdepCC_TRUE@ $(am__mv) $(DEPDIR)/$*.Tpo $(DEPDIR)/$*.Po @AMDEP_TRUE@@am__fastdepCC_FALSE@ source='$<' object='$@' libtool=no @AMDEPBACKSLASH@ @AMDEP_TRUE@@am__fastdepCC_FALSE@ DEPDIR=$(DEPDIR) $(CCDEPMODE) $(depcomp) @AMDEPBACKSLASH@ @am__fastdepCC_FALSE@ $(COMPILE) -c $< .c.obj: @am__fastdepCC_TRUE@ $(COMPILE) -MT $@ -MD -MP -MF $(DEPDIR)/$*.Tpo -c -o $@ `$(CYGPATH_W) '$<'` @am__fastdepCC_TRUE@ $(am__mv) $(DEPDIR)/$*.Tpo $(DEPDIR)/$*.Po @AMDEP_TRUE@@am__fastdepCC_FALSE@ source='$<' object='$@' libtool=no @AMDEPBACKSLASH@ @AMDEP_TRUE@@am__fastdepCC_FALSE@ DEPDIR=$(DEPDIR) $(CCDEPMODE) $(depcomp) @AMDEPBACKSLASH@ @am__fastdepCC_FALSE@ $(COMPILE) -c `$(CYGPATH_W) '$<'` .c.lo: @am__fastdepCC_TRUE@ $(LTCOMPILE) -MT $@ -MD -MP -MF $(DEPDIR)/$*.Tpo -c -o $@ $< @am__fastdepCC_TRUE@ $(am__mv) $(DEPDIR)/$*.Tpo $(DEPDIR)/$*.Plo @AMDEP_TRUE@@am__fastdepCC_FALSE@ source='$<' object='$@' libtool=yes @AMDEPBACKSLASH@ @AMDEP_TRUE@@am__fastdepCC_FALSE@ DEPDIR=$(DEPDIR) $(CCDEPMODE) $(depcomp) @AMDEPBACKSLASH@ @am__fastdepCC_FALSE@ $(LTCOMPILE) -c -o $@ $< mostlyclean-libtool: -rm -f *.lo clean-libtool: -rm -rf .libs _libs ID: $(HEADERS) $(SOURCES) $(LISP) $(TAGS_FILES) list='$(SOURCES) $(HEADERS) $(LISP) $(TAGS_FILES)'; \ unique=`for i in $$list; do \ if test -f "$$i"; then echo $$i; else echo $(srcdir)/$$i; fi; \ done | \ $(AWK) '{ files[$$0] = 1; nonempty = 1; } \ END { if (nonempty) { for (i in files) print i; }; }'`; \ mkid -fID $$unique tags: TAGS TAGS: $(HEADERS) $(SOURCES) $(TAGS_DEPENDENCIES) \ $(TAGS_FILES) $(LISP) set x; \ here=`pwd`; \ list='$(SOURCES) $(HEADERS) $(LISP) $(TAGS_FILES)'; \ unique=`for i in $$list; do \ if test -f "$$i"; then echo $$i; else echo $(srcdir)/$$i; fi; \ done | \ $(AWK) '{ files[$$0] = 1; nonempty = 1; } \ END { if (nonempty) { for (i in files) print i; }; }'`; \ shift; \ if test -z "$(ETAGS_ARGS)$$*$$unique"; then :; else \ test -n "$$unique" || unique=$$empty_fix; \ if test $$# -gt 0; then \ $(ETAGS) $(ETAGSFLAGS) $(AM_ETAGSFLAGS) $(ETAGS_ARGS) \ "$$@" $$unique; \ else \ $(ETAGS) $(ETAGSFLAGS) $(AM_ETAGSFLAGS) $(ETAGS_ARGS) \ $$unique; \ fi; \ fi ctags: CTAGS CTAGS: $(HEADERS) $(SOURCES) $(TAGS_DEPENDENCIES) \ $(TAGS_FILES) $(LISP) list='$(SOURCES) $(HEADERS) $(LISP) $(TAGS_FILES)'; \ unique=`for i in $$list; do \ if test -f "$$i"; then echo $$i; else echo $(srcdir)/$$i; fi; \ done | \ $(AWK) '{ files[$$0] = 1; nonempty = 1; } \ END { if (nonempty) { for (i in files) print i; }; }'`; \ test -z "$(CTAGS_ARGS)$$unique" \ || $(CTAGS) $(CTAGSFLAGS) $(AM_CTAGSFLAGS) $(CTAGS_ARGS) \ $$unique GTAGS: here=`$(am__cd) $(top_builddir) && pwd` \ && $(am__cd) $(top_srcdir) \ && gtags -i $(GTAGS_ARGS) "$$here" distclean-tags: -rm -f TAGS ID GTAGS GRTAGS GSYMS GPATH tags distdir: $(DISTFILES) @srcdirstrip=`echo "$(srcdir)" | sed 's/[].[^$$\\*]/\\\\&/g'`; \ topsrcdirstrip=`echo "$(top_srcdir)" | sed 's/[].[^$$\\*]/\\\\&/g'`; \ list='$(DISTFILES)'; \ dist_files=`for file in $$list; do echo $$file; done | \ sed -e "s|^$$srcdirstrip/||;t" \ -e "s|^$$topsrcdirstrip/|$(top_builddir)/|;t"`; \ case $$dist_files in \ */*) $(MKDIR_P) `echo "$$dist_files" | \ sed '/\//!d;s|^|$(distdir)/|;s,/[^/]*$$,,' | \ sort -u` ;; \ esac; \ for file in $$dist_files; do \ if test -f $$file || test -d $$file; then d=.; else d=$(srcdir); fi; \ if test -d $$d/$$file; then \ dir=`echo "/$$file" | sed -e 's,/[^/]*$$,,'`; \ if test -d "$(distdir)/$$file"; then \ find "$(distdir)/$$file" -type d ! -perm -700 -exec chmod u+rwx {} \;; \ fi; \ if test -d $(srcdir)/$$file && test $$d != $(srcdir); then \ cp -fpR $(srcdir)/$$file "$(distdir)$$dir" || exit 1; \ find "$(distdir)/$$file" -type d ! -perm -700 -exec chmod u+rwx {} \;; \ fi; \ cp -fpR $$d/$$file "$(distdir)$$dir" || exit 1; \ else \ test -f "$(distdir)/$$file" \ || cp -p $$d/$$file "$(distdir)/$$file" \ || exit 1; \ fi; \ done check-am: all-am check: $(BUILT_SOURCES) $(MAKE) $(AM_MAKEFLAGS) check-am all-am: Makefile $(LTLIBRARIES) installdirs: install: $(BUILT_SOURCES) $(MAKE) $(AM_MAKEFLAGS) install-am install-exec: install-exec-am install-data: install-data-am uninstall: uninstall-am install-am: all-am @$(MAKE) $(AM_MAKEFLAGS) install-exec-am install-data-am installcheck: installcheck-am install-strip: if test -z '$(STRIP)'; then \ $(MAKE) $(AM_MAKEFLAGS) INSTALL_PROGRAM="$(INSTALL_STRIP_PROGRAM)" \ install_sh_PROGRAM="$(INSTALL_STRIP_PROGRAM)" INSTALL_STRIP_FLAG=-s \ install; \ else \ $(MAKE) $(AM_MAKEFLAGS) INSTALL_PROGRAM="$(INSTALL_STRIP_PROGRAM)" \ install_sh_PROGRAM="$(INSTALL_STRIP_PROGRAM)" INSTALL_STRIP_FLAG=-s \ "INSTALL_PROGRAM_ENV=STRIPPROG='$(STRIP)'" install; \ fi mostlyclean-generic: clean-generic: distclean-generic: -test -z "$(CONFIG_CLEAN_FILES)" || rm -f $(CONFIG_CLEAN_FILES) -test . = "$(srcdir)" || test -z "$(CONFIG_CLEAN_VPATH_FILES)" || rm -f $(CONFIG_CLEAN_VPATH_FILES) maintainer-clean-generic: @echo "This command is intended for maintainers to use" @echo "it deletes files that may require special tools to rebuild." -test -z "$(BUILT_SOURCES)" || rm -f $(BUILT_SOURCES) clean: clean-am clean-am: clean-generic clean-libtool clean-noinstLTLIBRARIES \ mostlyclean-am distclean: distclean-am -rm -rf ./$(DEPDIR) -rm -f Makefile distclean-am: clean-am distclean-compile distclean-generic \ distclean-tags dvi: dvi-am dvi-am: html: html-am html-am: info: info-am info-am: install-data-am: install-dvi: install-dvi-am install-dvi-am: install-exec-am: install-html: install-html-am install-html-am: install-info: install-info-am install-info-am: install-man: install-pdf: install-pdf-am install-pdf-am: install-ps: install-ps-am install-ps-am: installcheck-am: maintainer-clean: maintainer-clean-am -rm -rf ./$(DEPDIR) -rm -f Makefile maintainer-clean-am: distclean-am maintainer-clean-generic mostlyclean: mostlyclean-am mostlyclean-am: mostlyclean-compile mostlyclean-generic \ mostlyclean-libtool pdf: pdf-am pdf-am: ps: ps-am ps-am: uninstall-am: .MAKE: all check install install-am install-strip .PHONY: CTAGS GTAGS all all-am check check-am clean clean-generic \ clean-libtool clean-noinstLTLIBRARIES ctags distclean \ distclean-compile distclean-generic distclean-libtool \ distclean-tags distdir dvi dvi-am html html-am info info-am \ install install-am install-data install-data-am install-dvi \ install-dvi-am install-exec install-exec-am install-html \ install-html-am install-info install-info-am install-man \ install-pdf install-pdf-am install-ps install-ps-am \ install-strip installcheck installcheck-am installdirs \ maintainer-clean maintainer-clean-generic mostlyclean \ mostlyclean-compile mostlyclean-generic mostlyclean-libtool \ pdf pdf-am ps ps-am tags uninstall uninstall-am $(EXTRA_DIST): Makefile ( \ echo "/* Generated automatically. DO NOT EDIT! */"; \ echo "#define SIMD_HEADER \"$(SIMD_HEADER)\""; \ echo "#include \"../common/"$*".c\""; \ ) >$@ # Tell versions [3.59,3.63) of GNU make to not export all variables. # Otherwise a system limit (for SysV at least) may be exceeded. .NOEXPORT: fftw-3.3.3/dft/simd/altivec/n1fv_2.c0000644000175000017500000000016012054411107013776 00000000000000/* Generated automatically. DO NOT EDIT! */ #define SIMD_HEADER "simd-altivec.h" #include "../common/n1fv_2.c" fftw-3.3.3/dft/simd/altivec/n1fv_3.c0000644000175000017500000000016012054411107013777 00000000000000/* Generated automatically. DO NOT EDIT! */ #define SIMD_HEADER "simd-altivec.h" #include "../common/n1fv_3.c" fftw-3.3.3/dft/simd/altivec/n1fv_4.c0000644000175000017500000000016012054411107014000 00000000000000/* Generated automatically. DO NOT EDIT! */ #define SIMD_HEADER "simd-altivec.h" #include "../common/n1fv_4.c" fftw-3.3.3/dft/simd/altivec/n1fv_5.c0000644000175000017500000000016012054411107014001 00000000000000/* Generated automatically. DO NOT EDIT! */ #define SIMD_HEADER "simd-altivec.h" #include "../common/n1fv_5.c" fftw-3.3.3/dft/simd/altivec/n1fv_6.c0000644000175000017500000000016012054411107014002 00000000000000/* Generated automatically. DO NOT EDIT! */ #define SIMD_HEADER "simd-altivec.h" #include "../common/n1fv_6.c" fftw-3.3.3/dft/simd/altivec/n1fv_7.c0000644000175000017500000000016012054411107014003 00000000000000/* Generated automatically. DO NOT EDIT! */ #define SIMD_HEADER "simd-altivec.h" #include "../common/n1fv_7.c" fftw-3.3.3/dft/simd/altivec/n1fv_8.c0000644000175000017500000000016012054411107014004 00000000000000/* Generated automatically. DO NOT EDIT! */ #define SIMD_HEADER "simd-altivec.h" #include "../common/n1fv_8.c" fftw-3.3.3/dft/simd/altivec/n1fv_9.c0000644000175000017500000000016012054411107014005 00000000000000/* Generated automatically. DO NOT EDIT! */ #define SIMD_HEADER "simd-altivec.h" #include "../common/n1fv_9.c" fftw-3.3.3/dft/simd/altivec/n1fv_10.c0000644000175000017500000000016112054411107014056 00000000000000/* Generated automatically. DO NOT EDIT! */ #define SIMD_HEADER "simd-altivec.h" #include "../common/n1fv_10.c" fftw-3.3.3/dft/simd/altivec/n1fv_11.c0000644000175000017500000000016112054411107014057 00000000000000/* Generated automatically. DO NOT EDIT! */ #define SIMD_HEADER "simd-altivec.h" #include "../common/n1fv_11.c" fftw-3.3.3/dft/simd/altivec/n1fv_12.c0000644000175000017500000000016112054411107014060 00000000000000/* Generated automatically. DO NOT EDIT! */ #define SIMD_HEADER "simd-altivec.h" #include "../common/n1fv_12.c" fftw-3.3.3/dft/simd/altivec/n1fv_13.c0000644000175000017500000000016112054411107014061 00000000000000/* Generated automatically. DO NOT EDIT! */ #define SIMD_HEADER "simd-altivec.h" #include "../common/n1fv_13.c" fftw-3.3.3/dft/simd/altivec/n1fv_14.c0000644000175000017500000000016112054411107014062 00000000000000/* Generated automatically. DO NOT EDIT! */ #define SIMD_HEADER "simd-altivec.h" #include "../common/n1fv_14.c" fftw-3.3.3/dft/simd/altivec/n1fv_15.c0000644000175000017500000000016112054411107014063 00000000000000/* Generated automatically. DO NOT EDIT! */ #define SIMD_HEADER "simd-altivec.h" #include "../common/n1fv_15.c" fftw-3.3.3/dft/simd/altivec/n1fv_16.c0000644000175000017500000000016112054411107014064 00000000000000/* Generated automatically. DO NOT EDIT! */ #define SIMD_HEADER "simd-altivec.h" #include "../common/n1fv_16.c" fftw-3.3.3/dft/simd/altivec/n1fv_32.c0000644000175000017500000000016112054411107014062 00000000000000/* Generated automatically. DO NOT EDIT! */ #define SIMD_HEADER "simd-altivec.h" #include "../common/n1fv_32.c" fftw-3.3.3/dft/simd/altivec/n1fv_64.c0000644000175000017500000000016112054411107014067 00000000000000/* Generated automatically. DO NOT EDIT! */ #define SIMD_HEADER "simd-altivec.h" #include "../common/n1fv_64.c" fftw-3.3.3/dft/simd/altivec/n1fv_128.c0000644000175000017500000000016212054411107014151 00000000000000/* Generated automatically. DO NOT EDIT! */ #define SIMD_HEADER "simd-altivec.h" #include "../common/n1fv_128.c" fftw-3.3.3/dft/simd/altivec/n1fv_20.c0000644000175000017500000000016112054411107014057 00000000000000/* Generated automatically. DO NOT EDIT! */ #define SIMD_HEADER "simd-altivec.h" #include "../common/n1fv_20.c" fftw-3.3.3/dft/simd/altivec/n1fv_25.c0000644000175000017500000000016112054411107014064 00000000000000/* Generated automatically. DO NOT EDIT! */ #define SIMD_HEADER "simd-altivec.h" #include "../common/n1fv_25.c" fftw-3.3.3/dft/simd/altivec/n1bv_2.c0000644000175000017500000000016012054411107013772 00000000000000/* Generated automatically. DO NOT EDIT! */ #define SIMD_HEADER "simd-altivec.h" #include "../common/n1bv_2.c" fftw-3.3.3/dft/simd/altivec/n1bv_3.c0000644000175000017500000000016012054411107013773 00000000000000/* Generated automatically. DO NOT EDIT! */ #define SIMD_HEADER "simd-altivec.h" #include "../common/n1bv_3.c" fftw-3.3.3/dft/simd/altivec/n1bv_4.c0000644000175000017500000000016012054411107013774 00000000000000/* Generated automatically. DO NOT EDIT! */ #define SIMD_HEADER "simd-altivec.h" #include "../common/n1bv_4.c" fftw-3.3.3/dft/simd/altivec/n1bv_5.c0000644000175000017500000000016012054411107013775 00000000000000/* Generated automatically. DO NOT EDIT! */ #define SIMD_HEADER "simd-altivec.h" #include "../common/n1bv_5.c" fftw-3.3.3/dft/simd/altivec/n1bv_6.c0000644000175000017500000000016012054411107013776 00000000000000/* Generated automatically. DO NOT EDIT! */ #define SIMD_HEADER "simd-altivec.h" #include "../common/n1bv_6.c" fftw-3.3.3/dft/simd/altivec/n1bv_7.c0000644000175000017500000000016012054411107013777 00000000000000/* Generated automatically. DO NOT EDIT! */ #define SIMD_HEADER "simd-altivec.h" #include "../common/n1bv_7.c" fftw-3.3.3/dft/simd/altivec/n1bv_8.c0000644000175000017500000000016012054411107014000 00000000000000/* Generated automatically. DO NOT EDIT! */ #define SIMD_HEADER "simd-altivec.h" #include "../common/n1bv_8.c" fftw-3.3.3/dft/simd/altivec/n1bv_9.c0000644000175000017500000000016012054411107014001 00000000000000/* Generated automatically. DO NOT EDIT! */ #define SIMD_HEADER "simd-altivec.h" #include "../common/n1bv_9.c" fftw-3.3.3/dft/simd/altivec/n1bv_10.c0000644000175000017500000000016112054411107014052 00000000000000/* Generated automatically. DO NOT EDIT! */ #define SIMD_HEADER "simd-altivec.h" #include "../common/n1bv_10.c" fftw-3.3.3/dft/simd/altivec/n1bv_11.c0000644000175000017500000000016112054411107014053 00000000000000/* Generated automatically. DO NOT EDIT! */ #define SIMD_HEADER "simd-altivec.h" #include "../common/n1bv_11.c" fftw-3.3.3/dft/simd/altivec/n1bv_12.c0000644000175000017500000000016112054411107014054 00000000000000/* Generated automatically. DO NOT EDIT! */ #define SIMD_HEADER "simd-altivec.h" #include "../common/n1bv_12.c" fftw-3.3.3/dft/simd/altivec/n1bv_13.c0000644000175000017500000000016112054411107014055 00000000000000/* Generated automatically. DO NOT EDIT! */ #define SIMD_HEADER "simd-altivec.h" #include "../common/n1bv_13.c" fftw-3.3.3/dft/simd/altivec/n1bv_14.c0000644000175000017500000000016112054411107014056 00000000000000/* Generated automatically. DO NOT EDIT! */ #define SIMD_HEADER "simd-altivec.h" #include "../common/n1bv_14.c" fftw-3.3.3/dft/simd/altivec/n1bv_15.c0000644000175000017500000000016112054411107014057 00000000000000/* Generated automatically. DO NOT EDIT! */ #define SIMD_HEADER "simd-altivec.h" #include "../common/n1bv_15.c" fftw-3.3.3/dft/simd/altivec/n1bv_16.c0000644000175000017500000000016112054411107014060 00000000000000/* Generated automatically. DO NOT EDIT! */ #define SIMD_HEADER "simd-altivec.h" #include "../common/n1bv_16.c" fftw-3.3.3/dft/simd/altivec/n1bv_32.c0000644000175000017500000000016112054411107014056 00000000000000/* Generated automatically. DO NOT EDIT! */ #define SIMD_HEADER "simd-altivec.h" #include "../common/n1bv_32.c" fftw-3.3.3/dft/simd/altivec/n1bv_64.c0000644000175000017500000000016112054411107014063 00000000000000/* Generated automatically. DO NOT EDIT! */ #define SIMD_HEADER "simd-altivec.h" #include "../common/n1bv_64.c" fftw-3.3.3/dft/simd/altivec/n1bv_128.c0000644000175000017500000000016212054411107014145 00000000000000/* Generated automatically. DO NOT EDIT! */ #define SIMD_HEADER "simd-altivec.h" #include "../common/n1bv_128.c" fftw-3.3.3/dft/simd/altivec/n1bv_20.c0000644000175000017500000000016112054411107014053 00000000000000/* Generated automatically. DO NOT EDIT! */ #define SIMD_HEADER "simd-altivec.h" #include "../common/n1bv_20.c" fftw-3.3.3/dft/simd/altivec/n1bv_25.c0000644000175000017500000000016112054411107014060 00000000000000/* Generated automatically. DO NOT EDIT! */ #define SIMD_HEADER "simd-altivec.h" #include "../common/n1bv_25.c" fftw-3.3.3/dft/simd/altivec/n2fv_2.c0000644000175000017500000000016012054411107013777 00000000000000/* Generated automatically. DO NOT EDIT! */ #define SIMD_HEADER "simd-altivec.h" #include "../common/n2fv_2.c" fftw-3.3.3/dft/simd/altivec/n2fv_4.c0000644000175000017500000000016012054411107014001 00000000000000/* Generated automatically. DO NOT EDIT! */ #define SIMD_HEADER "simd-altivec.h" #include "../common/n2fv_4.c" fftw-3.3.3/dft/simd/altivec/n2fv_6.c0000644000175000017500000000016012054411107014003 00000000000000/* Generated automatically. DO NOT EDIT! */ #define SIMD_HEADER "simd-altivec.h" #include "../common/n2fv_6.c" fftw-3.3.3/dft/simd/altivec/n2fv_8.c0000644000175000017500000000016012054411107014005 00000000000000/* Generated automatically. DO NOT EDIT! */ #define SIMD_HEADER "simd-altivec.h" #include "../common/n2fv_8.c" fftw-3.3.3/dft/simd/altivec/n2fv_10.c0000644000175000017500000000016112054411107014057 00000000000000/* Generated automatically. DO NOT EDIT! */ #define SIMD_HEADER "simd-altivec.h" #include "../common/n2fv_10.c" fftw-3.3.3/dft/simd/altivec/n2fv_12.c0000644000175000017500000000016112054411107014061 00000000000000/* Generated automatically. DO NOT EDIT! */ #define SIMD_HEADER "simd-altivec.h" #include "../common/n2fv_12.c" fftw-3.3.3/dft/simd/altivec/n2fv_14.c0000644000175000017500000000016112054411107014063 00000000000000/* Generated automatically. DO NOT EDIT! */ #define SIMD_HEADER "simd-altivec.h" #include "../common/n2fv_14.c" fftw-3.3.3/dft/simd/altivec/n2fv_16.c0000644000175000017500000000016112054411107014065 00000000000000/* Generated automatically. DO NOT EDIT! */ #define SIMD_HEADER "simd-altivec.h" #include "../common/n2fv_16.c" fftw-3.3.3/dft/simd/altivec/n2fv_32.c0000644000175000017500000000016112054411107014063 00000000000000/* Generated automatically. DO NOT EDIT! */ #define SIMD_HEADER "simd-altivec.h" #include "../common/n2fv_32.c" fftw-3.3.3/dft/simd/altivec/n2fv_64.c0000644000175000017500000000016112054411107014070 00000000000000/* Generated automatically. DO NOT EDIT! */ #define SIMD_HEADER "simd-altivec.h" #include "../common/n2fv_64.c" fftw-3.3.3/dft/simd/altivec/n2fv_20.c0000644000175000017500000000016112054411107014060 00000000000000/* Generated automatically. DO NOT EDIT! */ #define SIMD_HEADER "simd-altivec.h" #include "../common/n2fv_20.c" fftw-3.3.3/dft/simd/altivec/n2bv_2.c0000644000175000017500000000016012054411107013773 00000000000000/* Generated automatically. DO NOT EDIT! */ #define SIMD_HEADER "simd-altivec.h" #include "../common/n2bv_2.c" fftw-3.3.3/dft/simd/altivec/n2bv_4.c0000644000175000017500000000016012054411107013775 00000000000000/* Generated automatically. DO NOT EDIT! */ #define SIMD_HEADER "simd-altivec.h" #include "../common/n2bv_4.c" fftw-3.3.3/dft/simd/altivec/n2bv_6.c0000644000175000017500000000016012054411107013777 00000000000000/* Generated automatically. DO NOT EDIT! */ #define SIMD_HEADER "simd-altivec.h" #include "../common/n2bv_6.c" fftw-3.3.3/dft/simd/altivec/n2bv_8.c0000644000175000017500000000016012054411107014001 00000000000000/* Generated automatically. DO NOT EDIT! */ #define SIMD_HEADER "simd-altivec.h" #include "../common/n2bv_8.c" fftw-3.3.3/dft/simd/altivec/n2bv_10.c0000644000175000017500000000016112054411107014053 00000000000000/* Generated automatically. DO NOT EDIT! */ #define SIMD_HEADER "simd-altivec.h" #include "../common/n2bv_10.c" fftw-3.3.3/dft/simd/altivec/n2bv_12.c0000644000175000017500000000016112054411107014055 00000000000000/* Generated automatically. DO NOT EDIT! */ #define SIMD_HEADER "simd-altivec.h" #include "../common/n2bv_12.c" fftw-3.3.3/dft/simd/altivec/n2bv_14.c0000644000175000017500000000016112054411107014057 00000000000000/* Generated automatically. DO NOT EDIT! */ #define SIMD_HEADER "simd-altivec.h" #include "../common/n2bv_14.c" fftw-3.3.3/dft/simd/altivec/n2bv_16.c0000644000175000017500000000016112054411107014061 00000000000000/* Generated automatically. DO NOT EDIT! */ #define SIMD_HEADER "simd-altivec.h" #include "../common/n2bv_16.c" fftw-3.3.3/dft/simd/altivec/n2bv_32.c0000644000175000017500000000016112054411107014057 00000000000000/* Generated automatically. DO NOT EDIT! */ #define SIMD_HEADER "simd-altivec.h" #include "../common/n2bv_32.c" fftw-3.3.3/dft/simd/altivec/n2bv_64.c0000644000175000017500000000016112054411107014064 00000000000000/* Generated automatically. DO NOT EDIT! */ #define SIMD_HEADER "simd-altivec.h" #include "../common/n2bv_64.c" fftw-3.3.3/dft/simd/altivec/n2bv_20.c0000644000175000017500000000016112054411107014054 00000000000000/* Generated automatically. DO NOT EDIT! */ #define SIMD_HEADER "simd-altivec.h" #include "../common/n2bv_20.c" fftw-3.3.3/dft/simd/altivec/n2sv_4.c0000644000175000017500000000016012054411107014016 00000000000000/* Generated automatically. DO NOT EDIT! */ #define SIMD_HEADER "simd-altivec.h" #include "../common/n2sv_4.c" fftw-3.3.3/dft/simd/altivec/n2sv_8.c0000644000175000017500000000016012054411107014022 00000000000000/* Generated automatically. DO NOT EDIT! */ #define SIMD_HEADER "simd-altivec.h" #include "../common/n2sv_8.c" fftw-3.3.3/dft/simd/altivec/n2sv_16.c0000644000175000017500000000016112054411107014102 00000000000000/* Generated automatically. DO NOT EDIT! */ #define SIMD_HEADER "simd-altivec.h" #include "../common/n2sv_16.c" fftw-3.3.3/dft/simd/altivec/n2sv_32.c0000644000175000017500000000016112054411107014100 00000000000000/* Generated automatically. DO NOT EDIT! */ #define SIMD_HEADER "simd-altivec.h" #include "../common/n2sv_32.c" fftw-3.3.3/dft/simd/altivec/n2sv_64.c0000644000175000017500000000016112054411107014105 00000000000000/* Generated automatically. DO NOT EDIT! */ #define SIMD_HEADER "simd-altivec.h" #include "../common/n2sv_64.c" fftw-3.3.3/dft/simd/altivec/t1fuv_2.c0000644000175000017500000000016112054411107014172 00000000000000/* Generated automatically. DO NOT EDIT! */ #define SIMD_HEADER "simd-altivec.h" #include "../common/t1fuv_2.c" fftw-3.3.3/dft/simd/altivec/t1fuv_3.c0000644000175000017500000000016112054411107014173 00000000000000/* Generated automatically. DO NOT EDIT! */ #define SIMD_HEADER "simd-altivec.h" #include "../common/t1fuv_3.c" fftw-3.3.3/dft/simd/altivec/t1fuv_4.c0000644000175000017500000000016112054411107014174 00000000000000/* Generated automatically. DO NOT EDIT! */ #define SIMD_HEADER "simd-altivec.h" #include "../common/t1fuv_4.c" fftw-3.3.3/dft/simd/altivec/t1fuv_5.c0000644000175000017500000000016112054411107014175 00000000000000/* Generated automatically. DO NOT EDIT! */ #define SIMD_HEADER "simd-altivec.h" #include "../common/t1fuv_5.c" fftw-3.3.3/dft/simd/altivec/t1fuv_6.c0000644000175000017500000000016112054411107014176 00000000000000/* Generated automatically. DO NOT EDIT! */ #define SIMD_HEADER "simd-altivec.h" #include "../common/t1fuv_6.c" fftw-3.3.3/dft/simd/altivec/t1fuv_7.c0000644000175000017500000000016112054411107014177 00000000000000/* Generated automatically. DO NOT EDIT! */ #define SIMD_HEADER "simd-altivec.h" #include "../common/t1fuv_7.c" fftw-3.3.3/dft/simd/altivec/t1fuv_8.c0000644000175000017500000000016112054411107014200 00000000000000/* Generated automatically. DO NOT EDIT! */ #define SIMD_HEADER "simd-altivec.h" #include "../common/t1fuv_8.c" fftw-3.3.3/dft/simd/altivec/t1fuv_9.c0000644000175000017500000000016112054411107014201 00000000000000/* Generated automatically. DO NOT EDIT! */ #define SIMD_HEADER "simd-altivec.h" #include "../common/t1fuv_9.c" fftw-3.3.3/dft/simd/altivec/t1fuv_10.c0000644000175000017500000000016212054411107014252 00000000000000/* Generated automatically. DO NOT EDIT! */ #define SIMD_HEADER "simd-altivec.h" #include "../common/t1fuv_10.c" fftw-3.3.3/dft/simd/altivec/t1fv_2.c0000644000175000017500000000016012054411107014004 00000000000000/* Generated automatically. DO NOT EDIT! */ #define SIMD_HEADER "simd-altivec.h" #include "../common/t1fv_2.c" fftw-3.3.3/dft/simd/altivec/t1fv_3.c0000644000175000017500000000016012054411107014005 00000000000000/* Generated automatically. DO NOT EDIT! */ #define SIMD_HEADER "simd-altivec.h" #include "../common/t1fv_3.c" fftw-3.3.3/dft/simd/altivec/t1fv_4.c0000644000175000017500000000016012054411107014006 00000000000000/* Generated automatically. DO NOT EDIT! */ #define SIMD_HEADER "simd-altivec.h" #include "../common/t1fv_4.c" fftw-3.3.3/dft/simd/altivec/t1fv_5.c0000644000175000017500000000016012054411107014007 00000000000000/* Generated automatically. DO NOT EDIT! */ #define SIMD_HEADER "simd-altivec.h" #include "../common/t1fv_5.c" fftw-3.3.3/dft/simd/altivec/t1fv_6.c0000644000175000017500000000016012054411107014010 00000000000000/* Generated automatically. DO NOT EDIT! */ #define SIMD_HEADER "simd-altivec.h" #include "../common/t1fv_6.c" fftw-3.3.3/dft/simd/altivec/t1fv_7.c0000644000175000017500000000016012054411107014011 00000000000000/* Generated automatically. DO NOT EDIT! */ #define SIMD_HEADER "simd-altivec.h" #include "../common/t1fv_7.c" fftw-3.3.3/dft/simd/altivec/t1fv_8.c0000644000175000017500000000016012054411107014012 00000000000000/* Generated automatically. DO NOT EDIT! */ #define SIMD_HEADER "simd-altivec.h" #include "../common/t1fv_8.c" fftw-3.3.3/dft/simd/altivec/t1fv_9.c0000644000175000017500000000016012054411107014013 00000000000000/* Generated automatically. DO NOT EDIT! */ #define SIMD_HEADER "simd-altivec.h" #include "../common/t1fv_9.c" fftw-3.3.3/dft/simd/altivec/t1fv_10.c0000644000175000017500000000016112054411107014064 00000000000000/* Generated automatically. DO NOT EDIT! */ #define SIMD_HEADER "simd-altivec.h" #include "../common/t1fv_10.c" fftw-3.3.3/dft/simd/altivec/t1fv_12.c0000644000175000017500000000016112054411107014066 00000000000000/* Generated automatically. DO NOT EDIT! */ #define SIMD_HEADER "simd-altivec.h" #include "../common/t1fv_12.c" fftw-3.3.3/dft/simd/altivec/t1fv_15.c0000644000175000017500000000016112054411107014071 00000000000000/* Generated automatically. DO NOT EDIT! */ #define SIMD_HEADER "simd-altivec.h" #include "../common/t1fv_15.c" fftw-3.3.3/dft/simd/altivec/t1fv_16.c0000644000175000017500000000016112054411107014072 00000000000000/* Generated automatically. DO NOT EDIT! */ #define SIMD_HEADER "simd-altivec.h" #include "../common/t1fv_16.c" fftw-3.3.3/dft/simd/altivec/t1fv_32.c0000644000175000017500000000016112054411107014070 00000000000000/* Generated automatically. DO NOT EDIT! */ #define SIMD_HEADER "simd-altivec.h" #include "../common/t1fv_32.c" fftw-3.3.3/dft/simd/altivec/t1fv_64.c0000644000175000017500000000016112054411107014075 00000000000000/* Generated automatically. DO NOT EDIT! */ #define SIMD_HEADER "simd-altivec.h" #include "../common/t1fv_64.c" fftw-3.3.3/dft/simd/altivec/t1fv_20.c0000644000175000017500000000016112054411107014065 00000000000000/* Generated automatically. DO NOT EDIT! */ #define SIMD_HEADER "simd-altivec.h" #include "../common/t1fv_20.c" fftw-3.3.3/dft/simd/altivec/t1fv_25.c0000644000175000017500000000016112054411107014072 00000000000000/* Generated automatically. DO NOT EDIT! */ #define SIMD_HEADER "simd-altivec.h" #include "../common/t1fv_25.c" fftw-3.3.3/dft/simd/altivec/t2fv_2.c0000644000175000017500000000016012054411107014005 00000000000000/* Generated automatically. DO NOT EDIT! */ #define SIMD_HEADER "simd-altivec.h" #include "../common/t2fv_2.c" fftw-3.3.3/dft/simd/altivec/t2fv_4.c0000644000175000017500000000016012054411107014007 00000000000000/* Generated automatically. DO NOT EDIT! */ #define SIMD_HEADER "simd-altivec.h" #include "../common/t2fv_4.c" fftw-3.3.3/dft/simd/altivec/t2fv_8.c0000644000175000017500000000016012054411107014013 00000000000000/* Generated automatically. DO NOT EDIT! */ #define SIMD_HEADER "simd-altivec.h" #include "../common/t2fv_8.c" fftw-3.3.3/dft/simd/altivec/t2fv_16.c0000644000175000017500000000016112054411107014073 00000000000000/* Generated automatically. DO NOT EDIT! */ #define SIMD_HEADER "simd-altivec.h" #include "../common/t2fv_16.c" fftw-3.3.3/dft/simd/altivec/t2fv_32.c0000644000175000017500000000016112054411107014071 00000000000000/* Generated automatically. DO NOT EDIT! */ #define SIMD_HEADER "simd-altivec.h" #include "../common/t2fv_32.c" fftw-3.3.3/dft/simd/altivec/t2fv_64.c0000644000175000017500000000016112054411107014076 00000000000000/* Generated automatically. DO NOT EDIT! */ #define SIMD_HEADER "simd-altivec.h" #include "../common/t2fv_64.c" fftw-3.3.3/dft/simd/altivec/t2fv_5.c0000644000175000017500000000016012054411107014010 00000000000000/* Generated automatically. DO NOT EDIT! */ #define SIMD_HEADER "simd-altivec.h" #include "../common/t2fv_5.c" fftw-3.3.3/dft/simd/altivec/t2fv_10.c0000644000175000017500000000016112054411107014065 00000000000000/* Generated automatically. DO NOT EDIT! */ #define SIMD_HEADER "simd-altivec.h" #include "../common/t2fv_10.c" fftw-3.3.3/dft/simd/altivec/t2fv_20.c0000644000175000017500000000016112054411107014066 00000000000000/* Generated automatically. DO NOT EDIT! */ #define SIMD_HEADER "simd-altivec.h" #include "../common/t2fv_20.c" fftw-3.3.3/dft/simd/altivec/t2fv_25.c0000644000175000017500000000016112054411107014073 00000000000000/* Generated automatically. DO NOT EDIT! */ #define SIMD_HEADER "simd-altivec.h" #include "../common/t2fv_25.c" fftw-3.3.3/dft/simd/altivec/t3fv_4.c0000644000175000017500000000016012054411107014010 00000000000000/* Generated automatically. DO NOT EDIT! */ #define SIMD_HEADER "simd-altivec.h" #include "../common/t3fv_4.c" fftw-3.3.3/dft/simd/altivec/t3fv_8.c0000644000175000017500000000016012054411107014014 00000000000000/* Generated automatically. DO NOT EDIT! */ #define SIMD_HEADER "simd-altivec.h" #include "../common/t3fv_8.c" fftw-3.3.3/dft/simd/altivec/t3fv_16.c0000644000175000017500000000016112054411107014074 00000000000000/* Generated automatically. DO NOT EDIT! */ #define SIMD_HEADER "simd-altivec.h" #include "../common/t3fv_16.c" fftw-3.3.3/dft/simd/altivec/t3fv_32.c0000644000175000017500000000016112054411107014072 00000000000000/* Generated automatically. DO NOT EDIT! */ #define SIMD_HEADER "simd-altivec.h" #include "../common/t3fv_32.c" fftw-3.3.3/dft/simd/altivec/t3fv_5.c0000644000175000017500000000016012054411107014011 00000000000000/* Generated automatically. DO NOT EDIT! */ #define SIMD_HEADER "simd-altivec.h" #include "../common/t3fv_5.c" fftw-3.3.3/dft/simd/altivec/t3fv_10.c0000644000175000017500000000016112054411107014066 00000000000000/* Generated automatically. DO NOT EDIT! */ #define SIMD_HEADER "simd-altivec.h" #include "../common/t3fv_10.c" fftw-3.3.3/dft/simd/altivec/t3fv_20.c0000644000175000017500000000016112054411107014067 00000000000000/* Generated automatically. DO NOT EDIT! */ #define SIMD_HEADER "simd-altivec.h" #include "../common/t3fv_20.c" fftw-3.3.3/dft/simd/altivec/t3fv_25.c0000644000175000017500000000016112054411107014074 00000000000000/* Generated automatically. DO NOT EDIT! */ #define SIMD_HEADER "simd-altivec.h" #include "../common/t3fv_25.c" fftw-3.3.3/dft/simd/altivec/t1buv_2.c0000644000175000017500000000016112054411107014166 00000000000000/* Generated automatically. DO NOT EDIT! */ #define SIMD_HEADER "simd-altivec.h" #include "../common/t1buv_2.c" fftw-3.3.3/dft/simd/altivec/t1buv_3.c0000644000175000017500000000016112054411107014167 00000000000000/* Generated automatically. DO NOT EDIT! */ #define SIMD_HEADER "simd-altivec.h" #include "../common/t1buv_3.c" fftw-3.3.3/dft/simd/altivec/t1buv_4.c0000644000175000017500000000016112054411107014170 00000000000000/* Generated automatically. DO NOT EDIT! */ #define SIMD_HEADER "simd-altivec.h" #include "../common/t1buv_4.c" fftw-3.3.3/dft/simd/altivec/t1buv_5.c0000644000175000017500000000016112054411107014171 00000000000000/* Generated automatically. DO NOT EDIT! */ #define SIMD_HEADER "simd-altivec.h" #include "../common/t1buv_5.c" fftw-3.3.3/dft/simd/altivec/t1buv_6.c0000644000175000017500000000016112054411107014172 00000000000000/* Generated automatically. DO NOT EDIT! */ #define SIMD_HEADER "simd-altivec.h" #include "../common/t1buv_6.c" fftw-3.3.3/dft/simd/altivec/t1buv_7.c0000644000175000017500000000016112054411107014173 00000000000000/* Generated automatically. DO NOT EDIT! */ #define SIMD_HEADER "simd-altivec.h" #include "../common/t1buv_7.c" fftw-3.3.3/dft/simd/altivec/t1buv_8.c0000644000175000017500000000016112054411107014174 00000000000000/* Generated automatically. DO NOT EDIT! */ #define SIMD_HEADER "simd-altivec.h" #include "../common/t1buv_8.c" fftw-3.3.3/dft/simd/altivec/t1buv_9.c0000644000175000017500000000016112054411107014175 00000000000000/* Generated automatically. DO NOT EDIT! */ #define SIMD_HEADER "simd-altivec.h" #include "../common/t1buv_9.c" fftw-3.3.3/dft/simd/altivec/t1buv_10.c0000644000175000017500000000016212054411107014246 00000000000000/* Generated automatically. DO NOT EDIT! */ #define SIMD_HEADER "simd-altivec.h" #include "../common/t1buv_10.c" fftw-3.3.3/dft/simd/altivec/t1bv_2.c0000644000175000017500000000016012054411107014000 00000000000000/* Generated automatically. DO NOT EDIT! */ #define SIMD_HEADER "simd-altivec.h" #include "../common/t1bv_2.c" fftw-3.3.3/dft/simd/altivec/t1bv_3.c0000644000175000017500000000016012054411107014001 00000000000000/* Generated automatically. DO NOT EDIT! */ #define SIMD_HEADER "simd-altivec.h" #include "../common/t1bv_3.c" fftw-3.3.3/dft/simd/altivec/t1bv_4.c0000644000175000017500000000016012054411107014002 00000000000000/* Generated automatically. DO NOT EDIT! */ #define SIMD_HEADER "simd-altivec.h" #include "../common/t1bv_4.c" fftw-3.3.3/dft/simd/altivec/t1bv_5.c0000644000175000017500000000016012054411107014003 00000000000000/* Generated automatically. DO NOT EDIT! */ #define SIMD_HEADER "simd-altivec.h" #include "../common/t1bv_5.c" fftw-3.3.3/dft/simd/altivec/t1bv_6.c0000644000175000017500000000016012054411107014004 00000000000000/* Generated automatically. DO NOT EDIT! */ #define SIMD_HEADER "simd-altivec.h" #include "../common/t1bv_6.c" fftw-3.3.3/dft/simd/altivec/t1bv_7.c0000644000175000017500000000016012054411107014005 00000000000000/* Generated automatically. DO NOT EDIT! */ #define SIMD_HEADER "simd-altivec.h" #include "../common/t1bv_7.c" fftw-3.3.3/dft/simd/altivec/t1bv_8.c0000644000175000017500000000016012054411107014006 00000000000000/* Generated automatically. DO NOT EDIT! */ #define SIMD_HEADER "simd-altivec.h" #include "../common/t1bv_8.c" fftw-3.3.3/dft/simd/altivec/t1bv_9.c0000644000175000017500000000016012054411107014007 00000000000000/* Generated automatically. DO NOT EDIT! */ #define SIMD_HEADER "simd-altivec.h" #include "../common/t1bv_9.c" fftw-3.3.3/dft/simd/altivec/t1bv_10.c0000644000175000017500000000016112054411107014060 00000000000000/* Generated automatically. DO NOT EDIT! */ #define SIMD_HEADER "simd-altivec.h" #include "../common/t1bv_10.c" fftw-3.3.3/dft/simd/altivec/t1bv_12.c0000644000175000017500000000016112054411107014062 00000000000000/* Generated automatically. DO NOT EDIT! */ #define SIMD_HEADER "simd-altivec.h" #include "../common/t1bv_12.c" fftw-3.3.3/dft/simd/altivec/t1bv_15.c0000644000175000017500000000016112054411107014065 00000000000000/* Generated automatically. DO NOT EDIT! */ #define SIMD_HEADER "simd-altivec.h" #include "../common/t1bv_15.c" fftw-3.3.3/dft/simd/altivec/t1bv_16.c0000644000175000017500000000016112054411107014066 00000000000000/* Generated automatically. DO NOT EDIT! */ #define SIMD_HEADER "simd-altivec.h" #include "../common/t1bv_16.c" fftw-3.3.3/dft/simd/altivec/t1bv_32.c0000644000175000017500000000016112054411107014064 00000000000000/* Generated automatically. DO NOT EDIT! */ #define SIMD_HEADER "simd-altivec.h" #include "../common/t1bv_32.c" fftw-3.3.3/dft/simd/altivec/t1bv_64.c0000644000175000017500000000016112054411107014071 00000000000000/* Generated automatically. DO NOT EDIT! */ #define SIMD_HEADER "simd-altivec.h" #include "../common/t1bv_64.c" fftw-3.3.3/dft/simd/altivec/t1bv_20.c0000644000175000017500000000016112054411107014061 00000000000000/* Generated automatically. DO NOT EDIT! */ #define SIMD_HEADER "simd-altivec.h" #include "../common/t1bv_20.c" fftw-3.3.3/dft/simd/altivec/t1bv_25.c0000644000175000017500000000016112054411107014066 00000000000000/* Generated automatically. DO NOT EDIT! */ #define SIMD_HEADER "simd-altivec.h" #include "../common/t1bv_25.c" fftw-3.3.3/dft/simd/altivec/t2bv_2.c0000644000175000017500000000016012054411107014001 00000000000000/* Generated automatically. DO NOT EDIT! */ #define SIMD_HEADER "simd-altivec.h" #include "../common/t2bv_2.c" fftw-3.3.3/dft/simd/altivec/t2bv_4.c0000644000175000017500000000016012054411107014003 00000000000000/* Generated automatically. DO NOT EDIT! */ #define SIMD_HEADER "simd-altivec.h" #include "../common/t2bv_4.c" fftw-3.3.3/dft/simd/altivec/t2bv_8.c0000644000175000017500000000016012054411107014007 00000000000000/* Generated automatically. DO NOT EDIT! */ #define SIMD_HEADER "simd-altivec.h" #include "../common/t2bv_8.c" fftw-3.3.3/dft/simd/altivec/t2bv_16.c0000644000175000017500000000016112054411107014067 00000000000000/* Generated automatically. DO NOT EDIT! */ #define SIMD_HEADER "simd-altivec.h" #include "../common/t2bv_16.c" fftw-3.3.3/dft/simd/altivec/t2bv_32.c0000644000175000017500000000016112054411107014065 00000000000000/* Generated automatically. DO NOT EDIT! */ #define SIMD_HEADER "simd-altivec.h" #include "../common/t2bv_32.c" fftw-3.3.3/dft/simd/altivec/t2bv_64.c0000644000175000017500000000016112054411107014072 00000000000000/* Generated automatically. DO NOT EDIT! */ #define SIMD_HEADER "simd-altivec.h" #include "../common/t2bv_64.c" fftw-3.3.3/dft/simd/altivec/t2bv_5.c0000644000175000017500000000016012054411107014004 00000000000000/* Generated automatically. DO NOT EDIT! */ #define SIMD_HEADER "simd-altivec.h" #include "../common/t2bv_5.c" fftw-3.3.3/dft/simd/altivec/t2bv_10.c0000644000175000017500000000016112054411107014061 00000000000000/* Generated automatically. DO NOT EDIT! */ #define SIMD_HEADER "simd-altivec.h" #include "../common/t2bv_10.c" fftw-3.3.3/dft/simd/altivec/t2bv_20.c0000644000175000017500000000016112054411107014062 00000000000000/* Generated automatically. DO NOT EDIT! */ #define SIMD_HEADER "simd-altivec.h" #include "../common/t2bv_20.c" fftw-3.3.3/dft/simd/altivec/t2bv_25.c0000644000175000017500000000016112054411107014067 00000000000000/* Generated automatically. DO NOT EDIT! */ #define SIMD_HEADER "simd-altivec.h" #include "../common/t2bv_25.c" fftw-3.3.3/dft/simd/altivec/t3bv_4.c0000644000175000017500000000016012054411107014004 00000000000000/* Generated automatically. DO NOT EDIT! */ #define SIMD_HEADER "simd-altivec.h" #include "../common/t3bv_4.c" fftw-3.3.3/dft/simd/altivec/t3bv_8.c0000644000175000017500000000016012054411107014010 00000000000000/* Generated automatically. DO NOT EDIT! */ #define SIMD_HEADER "simd-altivec.h" #include "../common/t3bv_8.c" fftw-3.3.3/dft/simd/altivec/t3bv_16.c0000644000175000017500000000016112054411107014070 00000000000000/* Generated automatically. DO NOT EDIT! */ #define SIMD_HEADER "simd-altivec.h" #include "../common/t3bv_16.c" fftw-3.3.3/dft/simd/altivec/t3bv_32.c0000644000175000017500000000016112054411107014066 00000000000000/* Generated automatically. DO NOT EDIT! */ #define SIMD_HEADER "simd-altivec.h" #include "../common/t3bv_32.c" fftw-3.3.3/dft/simd/altivec/t3bv_5.c0000644000175000017500000000016012054411107014005 00000000000000/* Generated automatically. DO NOT EDIT! */ #define SIMD_HEADER "simd-altivec.h" #include "../common/t3bv_5.c" fftw-3.3.3/dft/simd/altivec/t3bv_10.c0000644000175000017500000000016112054411107014062 00000000000000/* Generated automatically. DO NOT EDIT! */ #define SIMD_HEADER "simd-altivec.h" #include "../common/t3bv_10.c" fftw-3.3.3/dft/simd/altivec/t3bv_20.c0000644000175000017500000000016112054411107014063 00000000000000/* Generated automatically. DO NOT EDIT! */ #define SIMD_HEADER "simd-altivec.h" #include "../common/t3bv_20.c" fftw-3.3.3/dft/simd/altivec/t3bv_25.c0000644000175000017500000000016112054411107014070 00000000000000/* Generated automatically. DO NOT EDIT! */ #define SIMD_HEADER "simd-altivec.h" #include "../common/t3bv_25.c" fftw-3.3.3/dft/simd/altivec/t1sv_2.c0000644000175000017500000000016012054411107014021 00000000000000/* Generated automatically. DO NOT EDIT! */ #define SIMD_HEADER "simd-altivec.h" #include "../common/t1sv_2.c" fftw-3.3.3/dft/simd/altivec/t1sv_4.c0000644000175000017500000000016012054411107014023 00000000000000/* Generated automatically. DO NOT EDIT! */ #define SIMD_HEADER "simd-altivec.h" #include "../common/t1sv_4.c" fftw-3.3.3/dft/simd/altivec/t1sv_8.c0000644000175000017500000000016012054411107014027 00000000000000/* Generated automatically. DO NOT EDIT! */ #define SIMD_HEADER "simd-altivec.h" #include "../common/t1sv_8.c" fftw-3.3.3/dft/simd/altivec/t1sv_16.c0000644000175000017500000000016112054411107014107 00000000000000/* Generated automatically. DO NOT EDIT! */ #define SIMD_HEADER "simd-altivec.h" #include "../common/t1sv_16.c" fftw-3.3.3/dft/simd/altivec/t1sv_32.c0000644000175000017500000000016112054411107014105 00000000000000/* Generated automatically. DO NOT EDIT! */ #define SIMD_HEADER "simd-altivec.h" #include "../common/t1sv_32.c" fftw-3.3.3/dft/simd/altivec/t2sv_4.c0000644000175000017500000000016012054411107014024 00000000000000/* Generated automatically. DO NOT EDIT! */ #define SIMD_HEADER "simd-altivec.h" #include "../common/t2sv_4.c" fftw-3.3.3/dft/simd/altivec/t2sv_8.c0000644000175000017500000000016012054411107014030 00000000000000/* Generated automatically. DO NOT EDIT! */ #define SIMD_HEADER "simd-altivec.h" #include "../common/t2sv_8.c" fftw-3.3.3/dft/simd/altivec/t2sv_16.c0000644000175000017500000000016112054411107014110 00000000000000/* Generated automatically. DO NOT EDIT! */ #define SIMD_HEADER "simd-altivec.h" #include "../common/t2sv_16.c" fftw-3.3.3/dft/simd/altivec/t2sv_32.c0000644000175000017500000000016112054411107014106 00000000000000/* Generated automatically. DO NOT EDIT! */ #define SIMD_HEADER "simd-altivec.h" #include "../common/t2sv_32.c" fftw-3.3.3/dft/simd/altivec/q1fv_2.c0000644000175000017500000000016012054411107014001 00000000000000/* Generated automatically. DO NOT EDIT! */ #define SIMD_HEADER "simd-altivec.h" #include "../common/q1fv_2.c" fftw-3.3.3/dft/simd/altivec/q1fv_4.c0000644000175000017500000000016012054411107014003 00000000000000/* Generated automatically. DO NOT EDIT! */ #define SIMD_HEADER "simd-altivec.h" #include "../common/q1fv_4.c" fftw-3.3.3/dft/simd/altivec/q1fv_5.c0000644000175000017500000000016012054411107014004 00000000000000/* Generated automatically. DO NOT EDIT! */ #define SIMD_HEADER "simd-altivec.h" #include "../common/q1fv_5.c" fftw-3.3.3/dft/simd/altivec/q1fv_8.c0000644000175000017500000000016012054411107014007 00000000000000/* Generated automatically. DO NOT EDIT! */ #define SIMD_HEADER "simd-altivec.h" #include "../common/q1fv_8.c" fftw-3.3.3/dft/simd/altivec/q1bv_2.c0000644000175000017500000000016012054411107013775 00000000000000/* Generated automatically. DO NOT EDIT! */ #define SIMD_HEADER "simd-altivec.h" #include "../common/q1bv_2.c" fftw-3.3.3/dft/simd/altivec/q1bv_4.c0000644000175000017500000000016012054411107013777 00000000000000/* Generated automatically. DO NOT EDIT! */ #define SIMD_HEADER "simd-altivec.h" #include "../common/q1bv_4.c" fftw-3.3.3/dft/simd/altivec/q1bv_5.c0000644000175000017500000000016012054411107014000 00000000000000/* Generated automatically. DO NOT EDIT! */ #define SIMD_HEADER "simd-altivec.h" #include "../common/q1bv_5.c" fftw-3.3.3/dft/simd/altivec/q1bv_8.c0000644000175000017500000000016012054411107014003 00000000000000/* Generated automatically. DO NOT EDIT! */ #define SIMD_HEADER "simd-altivec.h" #include "../common/q1bv_8.c" fftw-3.3.3/dft/simd/altivec/genus.c0000644000175000017500000000015712054411107014032 00000000000000/* Generated automatically. DO NOT EDIT! */ #define SIMD_HEADER "simd-altivec.h" #include "../common/genus.c" fftw-3.3.3/dft/simd/altivec/codlist.c0000644000175000017500000000016112054411107014345 00000000000000/* Generated automatically. DO NOT EDIT! */ #define SIMD_HEADER "simd-altivec.h" #include "../common/codlist.c" fftw-3.3.3/dft/simd/neon/0000755000175000017500000000000012054411107012132 500000000000000fftw-3.3.3/dft/simd/neon/Makefile.am0000644000175000017500000000043112054410050014100 00000000000000AM_CFLAGS = $(NEON_CFLAGS) SIMD_HEADER=simd-neon.h include $(top_srcdir)/dft/simd/codlist.mk include $(top_srcdir)/dft/simd/simd.mk if HAVE_NEON BUILT_SOURCES = $(EXTRA_DIST) noinst_LTLIBRARIES = libdft_neon_codelets.la libdft_neon_codelets_la_SOURCES = $(BUILT_SOURCES) endif fftw-3.3.3/dft/simd/neon/Makefile.in0000644000175000017500000011115212054410177014126 00000000000000# Makefile.in generated by automake 1.11.6 from Makefile.am. # @configure_input@ # Copyright (C) 1994, 1995, 1996, 1997, 1998, 1999, 2000, 2001, 2002, # 2003, 2004, 2005, 2006, 2007, 2008, 2009, 2010, 2011 Free Software # Foundation, Inc. # This Makefile.in is free software; the Free Software Foundation # gives unlimited permission to copy and/or distribute it, # with or without modifications, as long as this notice is preserved. # This program is distributed in the hope that it will be useful, # but WITHOUT ANY WARRANTY, to the extent permitted by law; without # even the implied warranty of MERCHANTABILITY or FITNESS FOR A # PARTICULAR PURPOSE. @SET_MAKE@ # This file contains a standard list of DFT SIMD codelets. It is # included by common/Makefile to generate the C files with the actual # codelets in them. It is included by {sse,sse2,...}/Makefile to # generate and compile stub files that include common/*.c # You can customize FFTW for special needs, e.g. to handle certain # sizes more efficiently, by adding new codelets to the lists of those # included by default. If you change the list of codelets, any new # ones you added will be automatically generated when you run the # bootstrap script (see "Generating your own code" in the FFTW # manual). VPATH = @srcdir@ am__make_dryrun = \ { \ am__dry=no; \ case $$MAKEFLAGS in \ *\\[\ \ ]*) \ echo 'am--echo: ; @echo "AM" OK' | $(MAKE) -f - 2>/dev/null \ | grep '^AM OK$$' >/dev/null || am__dry=yes;; \ *) \ for am__flg in $$MAKEFLAGS; do \ case $$am__flg in \ *=*|--*) ;; \ *n*) am__dry=yes; break;; \ esac; \ done;; \ esac; \ test $$am__dry = yes; \ } pkgdatadir = $(datadir)/@PACKAGE@ pkgincludedir = $(includedir)/@PACKAGE@ pkglibdir = $(libdir)/@PACKAGE@ pkglibexecdir = $(libexecdir)/@PACKAGE@ am__cd = CDPATH="$${ZSH_VERSION+.}$(PATH_SEPARATOR)" && cd install_sh_DATA = $(install_sh) -c -m 644 install_sh_PROGRAM = $(install_sh) -c install_sh_SCRIPT = $(install_sh) -c INSTALL_HEADER = $(INSTALL_DATA) transform = $(program_transform_name) NORMAL_INSTALL = : PRE_INSTALL = : POST_INSTALL = : NORMAL_UNINSTALL = : PRE_UNINSTALL = : POST_UNINSTALL = : build_triplet = @build@ host_triplet = @host@ DIST_COMMON = $(srcdir)/Makefile.am $(srcdir)/Makefile.in \ $(top_srcdir)/dft/simd/codlist.mk \ $(top_srcdir)/dft/simd/simd.mk subdir = dft/simd/neon ACLOCAL_M4 = $(top_srcdir)/aclocal.m4 am__aclocal_m4_deps = $(top_srcdir)/m4/acx_mpi.m4 \ $(top_srcdir)/m4/acx_pthread.m4 \ $(top_srcdir)/m4/ax_cc_maxopt.m4 \ $(top_srcdir)/m4/ax_check_compiler_flags.m4 \ $(top_srcdir)/m4/ax_compiler_vendor.m4 \ $(top_srcdir)/m4/ax_gcc_aligns_stack.m4 \ $(top_srcdir)/m4/ax_gcc_version.m4 \ $(top_srcdir)/m4/ax_openmp.m4 $(top_srcdir)/m4/libtool.m4 \ $(top_srcdir)/m4/ltoptions.m4 $(top_srcdir)/m4/ltsugar.m4 \ $(top_srcdir)/m4/ltversion.m4 $(top_srcdir)/m4/lt~obsolete.m4 \ $(top_srcdir)/configure.ac am__configure_deps = $(am__aclocal_m4_deps) $(CONFIGURE_DEPENDENCIES) \ $(ACLOCAL_M4) mkinstalldirs = $(install_sh) -d CONFIG_HEADER = $(top_builddir)/config.h CONFIG_CLEAN_FILES = CONFIG_CLEAN_VPATH_FILES = LTLIBRARIES = $(noinst_LTLIBRARIES) libdft_neon_codelets_la_LIBADD = am__libdft_neon_codelets_la_SOURCES_DIST = n1fv_2.c n1fv_3.c n1fv_4.c \ n1fv_5.c n1fv_6.c n1fv_7.c n1fv_8.c n1fv_9.c n1fv_10.c \ n1fv_11.c n1fv_12.c n1fv_13.c n1fv_14.c n1fv_15.c n1fv_16.c \ n1fv_32.c n1fv_64.c n1fv_128.c n1fv_20.c n1fv_25.c n1bv_2.c \ n1bv_3.c n1bv_4.c n1bv_5.c n1bv_6.c n1bv_7.c n1bv_8.c n1bv_9.c \ n1bv_10.c n1bv_11.c n1bv_12.c n1bv_13.c n1bv_14.c n1bv_15.c \ n1bv_16.c n1bv_32.c n1bv_64.c n1bv_128.c n1bv_20.c n1bv_25.c \ n2fv_2.c n2fv_4.c n2fv_6.c n2fv_8.c n2fv_10.c n2fv_12.c \ n2fv_14.c n2fv_16.c n2fv_32.c n2fv_64.c n2fv_20.c n2bv_2.c \ n2bv_4.c n2bv_6.c n2bv_8.c n2bv_10.c n2bv_12.c n2bv_14.c \ n2bv_16.c n2bv_32.c n2bv_64.c n2bv_20.c n2sv_4.c n2sv_8.c \ n2sv_16.c n2sv_32.c n2sv_64.c t1fuv_2.c t1fuv_3.c t1fuv_4.c \ t1fuv_5.c t1fuv_6.c t1fuv_7.c t1fuv_8.c t1fuv_9.c t1fuv_10.c \ t1fv_2.c t1fv_3.c t1fv_4.c t1fv_5.c t1fv_6.c t1fv_7.c t1fv_8.c \ t1fv_9.c t1fv_10.c t1fv_12.c t1fv_15.c t1fv_16.c t1fv_32.c \ t1fv_64.c t1fv_20.c t1fv_25.c t2fv_2.c t2fv_4.c t2fv_8.c \ t2fv_16.c t2fv_32.c t2fv_64.c t2fv_5.c t2fv_10.c t2fv_20.c \ t2fv_25.c t3fv_4.c t3fv_8.c t3fv_16.c t3fv_32.c t3fv_5.c \ t3fv_10.c t3fv_20.c t3fv_25.c t1buv_2.c t1buv_3.c t1buv_4.c \ t1buv_5.c t1buv_6.c t1buv_7.c t1buv_8.c t1buv_9.c t1buv_10.c \ t1bv_2.c t1bv_3.c t1bv_4.c t1bv_5.c t1bv_6.c t1bv_7.c t1bv_8.c \ t1bv_9.c t1bv_10.c t1bv_12.c t1bv_15.c t1bv_16.c t1bv_32.c \ t1bv_64.c t1bv_20.c t1bv_25.c t2bv_2.c t2bv_4.c t2bv_8.c \ t2bv_16.c t2bv_32.c t2bv_64.c t2bv_5.c t2bv_10.c t2bv_20.c \ t2bv_25.c t3bv_4.c t3bv_8.c t3bv_16.c t3bv_32.c t3bv_5.c \ t3bv_10.c t3bv_20.c t3bv_25.c t1sv_2.c t1sv_4.c t1sv_8.c \ t1sv_16.c t1sv_32.c t2sv_4.c t2sv_8.c t2sv_16.c t2sv_32.c \ q1fv_2.c q1fv_4.c q1fv_5.c q1fv_8.c q1bv_2.c q1bv_4.c q1bv_5.c \ q1bv_8.c genus.c codlist.c am__objects_1 = n1fv_2.lo n1fv_3.lo n1fv_4.lo n1fv_5.lo n1fv_6.lo \ n1fv_7.lo n1fv_8.lo n1fv_9.lo n1fv_10.lo n1fv_11.lo n1fv_12.lo \ n1fv_13.lo n1fv_14.lo n1fv_15.lo n1fv_16.lo n1fv_32.lo \ n1fv_64.lo n1fv_128.lo n1fv_20.lo n1fv_25.lo am__objects_2 = n1bv_2.lo n1bv_3.lo n1bv_4.lo n1bv_5.lo n1bv_6.lo \ n1bv_7.lo n1bv_8.lo n1bv_9.lo n1bv_10.lo n1bv_11.lo n1bv_12.lo \ n1bv_13.lo n1bv_14.lo n1bv_15.lo n1bv_16.lo n1bv_32.lo \ n1bv_64.lo n1bv_128.lo n1bv_20.lo n1bv_25.lo am__objects_3 = n2fv_2.lo n2fv_4.lo n2fv_6.lo n2fv_8.lo n2fv_10.lo \ n2fv_12.lo n2fv_14.lo n2fv_16.lo n2fv_32.lo n2fv_64.lo \ n2fv_20.lo am__objects_4 = n2bv_2.lo n2bv_4.lo n2bv_6.lo n2bv_8.lo n2bv_10.lo \ n2bv_12.lo n2bv_14.lo n2bv_16.lo n2bv_32.lo n2bv_64.lo \ n2bv_20.lo am__objects_5 = n2sv_4.lo n2sv_8.lo n2sv_16.lo n2sv_32.lo n2sv_64.lo am__objects_6 = t1fuv_2.lo t1fuv_3.lo t1fuv_4.lo t1fuv_5.lo t1fuv_6.lo \ t1fuv_7.lo t1fuv_8.lo t1fuv_9.lo t1fuv_10.lo am__objects_7 = t1fv_2.lo t1fv_3.lo t1fv_4.lo t1fv_5.lo t1fv_6.lo \ t1fv_7.lo t1fv_8.lo t1fv_9.lo t1fv_10.lo t1fv_12.lo t1fv_15.lo \ t1fv_16.lo t1fv_32.lo t1fv_64.lo t1fv_20.lo t1fv_25.lo am__objects_8 = t2fv_2.lo t2fv_4.lo t2fv_8.lo t2fv_16.lo t2fv_32.lo \ t2fv_64.lo t2fv_5.lo t2fv_10.lo t2fv_20.lo t2fv_25.lo am__objects_9 = t3fv_4.lo t3fv_8.lo t3fv_16.lo t3fv_32.lo t3fv_5.lo \ t3fv_10.lo t3fv_20.lo t3fv_25.lo am__objects_10 = t1buv_2.lo t1buv_3.lo t1buv_4.lo t1buv_5.lo \ t1buv_6.lo t1buv_7.lo t1buv_8.lo t1buv_9.lo t1buv_10.lo am__objects_11 = t1bv_2.lo t1bv_3.lo t1bv_4.lo t1bv_5.lo t1bv_6.lo \ t1bv_7.lo t1bv_8.lo t1bv_9.lo t1bv_10.lo t1bv_12.lo t1bv_15.lo \ t1bv_16.lo t1bv_32.lo t1bv_64.lo t1bv_20.lo t1bv_25.lo am__objects_12 = t2bv_2.lo t2bv_4.lo t2bv_8.lo t2bv_16.lo t2bv_32.lo \ t2bv_64.lo t2bv_5.lo t2bv_10.lo t2bv_20.lo t2bv_25.lo am__objects_13 = t3bv_4.lo t3bv_8.lo t3bv_16.lo t3bv_32.lo t3bv_5.lo \ t3bv_10.lo t3bv_20.lo t3bv_25.lo am__objects_14 = t1sv_2.lo t1sv_4.lo t1sv_8.lo t1sv_16.lo t1sv_32.lo am__objects_15 = t2sv_4.lo t2sv_8.lo t2sv_16.lo t2sv_32.lo am__objects_16 = q1fv_2.lo q1fv_4.lo q1fv_5.lo q1fv_8.lo am__objects_17 = q1bv_2.lo q1bv_4.lo q1bv_5.lo q1bv_8.lo am__objects_18 = $(am__objects_1) $(am__objects_2) $(am__objects_3) \ $(am__objects_4) $(am__objects_5) $(am__objects_6) \ $(am__objects_7) $(am__objects_8) $(am__objects_9) \ $(am__objects_10) $(am__objects_11) $(am__objects_12) \ $(am__objects_13) $(am__objects_14) $(am__objects_15) \ $(am__objects_16) $(am__objects_17) am__objects_19 = $(am__objects_18) genus.lo codlist.lo @HAVE_NEON_TRUE@am__objects_20 = $(am__objects_19) @HAVE_NEON_TRUE@am_libdft_neon_codelets_la_OBJECTS = \ @HAVE_NEON_TRUE@ $(am__objects_20) libdft_neon_codelets_la_OBJECTS = \ $(am_libdft_neon_codelets_la_OBJECTS) @HAVE_NEON_TRUE@am_libdft_neon_codelets_la_rpath = DEFAULT_INCLUDES = -I.@am__isrc@ -I$(top_builddir) depcomp = $(SHELL) $(top_srcdir)/depcomp am__depfiles_maybe = depfiles am__mv = mv -f COMPILE = $(CC) $(DEFS) $(DEFAULT_INCLUDES) $(INCLUDES) $(AM_CPPFLAGS) \ $(CPPFLAGS) $(AM_CFLAGS) $(CFLAGS) LTCOMPILE = $(LIBTOOL) --tag=CC $(AM_LIBTOOLFLAGS) $(LIBTOOLFLAGS) \ --mode=compile $(CC) $(DEFS) $(DEFAULT_INCLUDES) $(INCLUDES) \ $(AM_CPPFLAGS) $(CPPFLAGS) $(AM_CFLAGS) $(CFLAGS) CCLD = $(CC) LINK = $(LIBTOOL) --tag=CC $(AM_LIBTOOLFLAGS) $(LIBTOOLFLAGS) \ --mode=link $(CCLD) $(AM_CFLAGS) $(CFLAGS) $(AM_LDFLAGS) \ $(LDFLAGS) -o $@ SOURCES = $(libdft_neon_codelets_la_SOURCES) DIST_SOURCES = $(am__libdft_neon_codelets_la_SOURCES_DIST) am__can_run_installinfo = \ case $$AM_UPDATE_INFO_DIR in \ n|no|NO) false;; \ *) (install-info --version) >/dev/null 2>&1;; \ esac ETAGS = etags CTAGS = ctags DISTFILES = $(DIST_COMMON) $(DIST_SOURCES) $(TEXINFOS) $(EXTRA_DIST) ACLOCAL = @ACLOCAL@ ALLOCA = @ALLOCA@ ALTIVEC_CFLAGS = @ALTIVEC_CFLAGS@ AMTAR = @AMTAR@ AR = @AR@ AS = @AS@ AUTOCONF = @AUTOCONF@ AUTOHEADER = @AUTOHEADER@ AUTOMAKE = @AUTOMAKE@ AVX_CFLAGS = @AVX_CFLAGS@ AWK = @AWK@ CC = @CC@ CCDEPMODE = @CCDEPMODE@ CFLAGS = @CFLAGS@ CHECK_PL_OPTS = @CHECK_PL_OPTS@ CPP = @CPP@ CPPFLAGS = @CPPFLAGS@ CYGPATH_W = @CYGPATH_W@ C_FFTW_R2R_KIND = @C_FFTW_R2R_KIND@ C_MPI_FINT = @C_MPI_FINT@ DEFS = @DEFS@ DEPDIR = @DEPDIR@ DLLTOOL = @DLLTOOL@ DSYMUTIL = @DSYMUTIL@ DUMPBIN = @DUMPBIN@ ECHO_C = @ECHO_C@ ECHO_N = @ECHO_N@ ECHO_T = @ECHO_T@ EGREP = @EGREP@ EXEEXT = @EXEEXT@ F77 = @F77@ FFLAGS = @FFLAGS@ FGREP = @FGREP@ FLIBS = @FLIBS@ GREP = @GREP@ INSTALL = @INSTALL@ INSTALL_DATA = @INSTALL_DATA@ INSTALL_PROGRAM = @INSTALL_PROGRAM@ INSTALL_SCRIPT = @INSTALL_SCRIPT@ INSTALL_STRIP_PROGRAM = @INSTALL_STRIP_PROGRAM@ LD = @LD@ LDFLAGS = @LDFLAGS@ LIBOBJS = @LIBOBJS@ LIBQUADMATH = @LIBQUADMATH@ LIBS = @LIBS@ LIBTOOL = @LIBTOOL@ LIPO = @LIPO@ LN_S = @LN_S@ LTLIBOBJS = @LTLIBOBJS@ MAINT = @MAINT@ MAKEINFO = @MAKEINFO@ MANIFEST_TOOL = @MANIFEST_TOOL@ MKDIR_P = @MKDIR_P@ MPICC = @MPICC@ MPILIBS = @MPILIBS@ MPIRUN = @MPIRUN@ NEON_CFLAGS = @NEON_CFLAGS@ NM = @NM@ NMEDIT = @NMEDIT@ OBJDUMP = @OBJDUMP@ OBJEXT = @OBJEXT@ OCAMLBUILD = @OCAMLBUILD@ OPENMP_CFLAGS = @OPENMP_CFLAGS@ OTOOL = @OTOOL@ OTOOL64 = @OTOOL64@ PACKAGE = @PACKAGE@ PACKAGE_BUGREPORT = @PACKAGE_BUGREPORT@ PACKAGE_NAME = @PACKAGE_NAME@ PACKAGE_STRING = @PACKAGE_STRING@ PACKAGE_TARNAME = @PACKAGE_TARNAME@ PACKAGE_URL = @PACKAGE_URL@ PACKAGE_VERSION = @PACKAGE_VERSION@ PATH_SEPARATOR = @PATH_SEPARATOR@ POW_LIB = @POW_LIB@ PRECISION = @PRECISION@ PREC_SUFFIX = @PREC_SUFFIX@ PTHREAD_CC = @PTHREAD_CC@ PTHREAD_CFLAGS = @PTHREAD_CFLAGS@ PTHREAD_LIBS = @PTHREAD_LIBS@ RANLIB = @RANLIB@ SED = @SED@ SET_MAKE = @SET_MAKE@ SHARED_VERSION_INFO = @SHARED_VERSION_INFO@ SHELL = @SHELL@ SSE2_CFLAGS = @SSE2_CFLAGS@ STACK_ALIGN_CFLAGS = @STACK_ALIGN_CFLAGS@ STRIP = @STRIP@ THREADLIBS = @THREADLIBS@ VERSION = @VERSION@ abs_builddir = @abs_builddir@ abs_srcdir = @abs_srcdir@ abs_top_builddir = @abs_top_builddir@ abs_top_srcdir = @abs_top_srcdir@ ac_ct_AR = @ac_ct_AR@ ac_ct_CC = @ac_ct_CC@ ac_ct_DUMPBIN = @ac_ct_DUMPBIN@ ac_ct_F77 = @ac_ct_F77@ acx_pthread_config = @acx_pthread_config@ am__include = @am__include@ am__leading_dot = @am__leading_dot@ am__quote = @am__quote@ am__tar = @am__tar@ am__untar = @am__untar@ bindir = @bindir@ build = @build@ build_alias = @build_alias@ build_cpu = @build_cpu@ build_os = @build_os@ build_vendor = @build_vendor@ builddir = @builddir@ datadir = @datadir@ datarootdir = @datarootdir@ docdir = @docdir@ dvidir = @dvidir@ exec_prefix = @exec_prefix@ host = @host@ host_alias = @host_alias@ host_cpu = @host_cpu@ host_os = @host_os@ host_vendor = @host_vendor@ htmldir = @htmldir@ includedir = @includedir@ infodir = @infodir@ install_sh = @install_sh@ libdir = @libdir@ libexecdir = @libexecdir@ localedir = @localedir@ localstatedir = @localstatedir@ mandir = @mandir@ mkdir_p = @mkdir_p@ oldincludedir = @oldincludedir@ pdfdir = @pdfdir@ prefix = @prefix@ program_transform_name = @program_transform_name@ psdir = @psdir@ sbindir = @sbindir@ sharedstatedir = @sharedstatedir@ srcdir = @srcdir@ sysconfdir = @sysconfdir@ target_alias = @target_alias@ top_build_prefix = @top_build_prefix@ top_builddir = @top_builddir@ top_srcdir = @top_srcdir@ AM_CFLAGS = $(NEON_CFLAGS) SIMD_HEADER = simd-neon.h ########################################################################### # n1fv_ is a hard-coded FFTW_FORWARD FFT of size , using SIMD N1F = n1fv_2.c n1fv_3.c n1fv_4.c n1fv_5.c n1fv_6.c n1fv_7.c n1fv_8.c \ n1fv_9.c n1fv_10.c n1fv_11.c n1fv_12.c n1fv_13.c n1fv_14.c n1fv_15.c \ n1fv_16.c n1fv_32.c n1fv_64.c n1fv_128.c n1fv_20.c n1fv_25.c # as above, with restricted input vector stride N2F = n2fv_2.c n2fv_4.c n2fv_6.c n2fv_8.c n2fv_10.c n2fv_12.c \ n2fv_14.c n2fv_16.c n2fv_32.c n2fv_64.c n2fv_20.c # as above, but FFTW_BACKWARD N1B = n1bv_2.c n1bv_3.c n1bv_4.c n1bv_5.c n1bv_6.c n1bv_7.c n1bv_8.c \ n1bv_9.c n1bv_10.c n1bv_11.c n1bv_12.c n1bv_13.c n1bv_14.c n1bv_15.c \ n1bv_16.c n1bv_32.c n1bv_64.c n1bv_128.c n1bv_20.c n1bv_25.c N2B = n2bv_2.c n2bv_4.c n2bv_6.c n2bv_8.c n2bv_10.c n2bv_12.c \ n2bv_14.c n2bv_16.c n2bv_32.c n2bv_64.c n2bv_20.c # split-complex codelets N2S = n2sv_4.c n2sv_8.c n2sv_16.c n2sv_32.c n2sv_64.c ########################################################################### # t1fv_ is a "twiddle" FFT of size , implementing a radix-r DIT step # for an FFTW_FORWARD transform, using SIMD T1F = t1fv_2.c t1fv_3.c t1fv_4.c t1fv_5.c t1fv_6.c t1fv_7.c t1fv_8.c \ t1fv_9.c t1fv_10.c t1fv_12.c t1fv_15.c t1fv_16.c t1fv_32.c t1fv_64.c \ t1fv_20.c t1fv_25.c # same as t1fv_*, but with different twiddle storage scheme T2F = t2fv_2.c t2fv_4.c t2fv_8.c t2fv_16.c t2fv_32.c t2fv_64.c \ t2fv_5.c t2fv_10.c t2fv_20.c t2fv_25.c T3F = t3fv_4.c t3fv_8.c t3fv_16.c t3fv_32.c t3fv_5.c t3fv_10.c \ t3fv_20.c t3fv_25.c T1FU = t1fuv_2.c t1fuv_3.c t1fuv_4.c t1fuv_5.c t1fuv_6.c t1fuv_7.c \ t1fuv_8.c t1fuv_9.c t1fuv_10.c # as above, but FFTW_BACKWARD T1B = t1bv_2.c t1bv_3.c t1bv_4.c t1bv_5.c t1bv_6.c t1bv_7.c t1bv_8.c \ t1bv_9.c t1bv_10.c t1bv_12.c t1bv_15.c t1bv_16.c t1bv_32.c t1bv_64.c \ t1bv_20.c t1bv_25.c # same as t1bv_*, but with different twiddle storage scheme T2B = t2bv_2.c t2bv_4.c t2bv_8.c t2bv_16.c t2bv_32.c t2bv_64.c \ t2bv_5.c t2bv_10.c t2bv_20.c t2bv_25.c T3B = t3bv_4.c t3bv_8.c t3bv_16.c t3bv_32.c t3bv_5.c t3bv_10.c \ t3bv_20.c t3bv_25.c T1BU = t1buv_2.c t1buv_3.c t1buv_4.c t1buv_5.c t1buv_6.c t1buv_7.c \ t1buv_8.c t1buv_9.c t1buv_10.c # split-complex codelets T1S = t1sv_2.c t1sv_4.c t1sv_8.c t1sv_16.c t1sv_32.c T2S = t2sv_4.c t2sv_8.c t2sv_16.c t2sv_32.c ########################################################################### # q1fv_ is twiddle FFTW_FORWARD FFTs of size (DIF step), # where the output is transposed, using SIMD. This is used for # in-place transposes in sizes that are divisible by ^2. These # codelets have size ~ ^2, so you should probably not use # bigger than 8 or so. Q1F = q1fv_2.c q1fv_4.c q1fv_5.c q1fv_8.c # as above, but FFTW_BACKWARD Q1B = q1bv_2.c q1bv_4.c q1bv_5.c q1bv_8.c ########################################################################### SIMD_CODELETS = $(N1F) $(N1B) $(N2F) $(N2B) $(N2S) $(T1FU) $(T1F) \ $(T2F) $(T3F) $(T1BU) $(T1B) $(T2B) $(T3B) $(T1S) $(T2S) $(Q1F) $(Q1B) AM_CPPFLAGS = -I$(top_srcdir)/kernel -I$(top_srcdir)/dft \ -I$(top_srcdir)/dft/simd -I$(top_srcdir)/simd-support EXTRA_DIST = $(SIMD_CODELETS) genus.c codlist.c @HAVE_NEON_TRUE@BUILT_SOURCES = $(EXTRA_DIST) @HAVE_NEON_TRUE@noinst_LTLIBRARIES = libdft_neon_codelets.la @HAVE_NEON_TRUE@libdft_neon_codelets_la_SOURCES = $(BUILT_SOURCES) all: $(BUILT_SOURCES) $(MAKE) $(AM_MAKEFLAGS) all-am .SUFFIXES: .SUFFIXES: .c .lo .o .obj $(srcdir)/Makefile.in: @MAINTAINER_MODE_TRUE@ $(srcdir)/Makefile.am $(top_srcdir)/dft/simd/codlist.mk $(top_srcdir)/dft/simd/simd.mk $(am__configure_deps) @for dep in $?; do \ case '$(am__configure_deps)' in \ *$$dep*) \ ( cd $(top_builddir) && $(MAKE) $(AM_MAKEFLAGS) am--refresh ) \ && { if test -f $@; then exit 0; else break; fi; }; \ exit 1;; \ esac; \ done; \ echo ' cd $(top_srcdir) && $(AUTOMAKE) --gnu dft/simd/neon/Makefile'; \ $(am__cd) $(top_srcdir) && \ $(AUTOMAKE) --gnu dft/simd/neon/Makefile .PRECIOUS: Makefile Makefile: $(srcdir)/Makefile.in $(top_builddir)/config.status @case '$?' in \ *config.status*) \ cd $(top_builddir) && $(MAKE) $(AM_MAKEFLAGS) am--refresh;; \ *) \ echo ' cd $(top_builddir) && $(SHELL) ./config.status $(subdir)/$@ $(am__depfiles_maybe)'; \ cd $(top_builddir) && $(SHELL) ./config.status $(subdir)/$@ $(am__depfiles_maybe);; \ esac; $(top_srcdir)/dft/simd/codlist.mk $(top_srcdir)/dft/simd/simd.mk: $(top_builddir)/config.status: $(top_srcdir)/configure $(CONFIG_STATUS_DEPENDENCIES) cd $(top_builddir) && $(MAKE) $(AM_MAKEFLAGS) am--refresh $(top_srcdir)/configure: @MAINTAINER_MODE_TRUE@ $(am__configure_deps) cd $(top_builddir) && $(MAKE) $(AM_MAKEFLAGS) am--refresh $(ACLOCAL_M4): @MAINTAINER_MODE_TRUE@ $(am__aclocal_m4_deps) cd $(top_builddir) && $(MAKE) $(AM_MAKEFLAGS) am--refresh $(am__aclocal_m4_deps): clean-noinstLTLIBRARIES: -test -z "$(noinst_LTLIBRARIES)" || rm -f $(noinst_LTLIBRARIES) @list='$(noinst_LTLIBRARIES)'; for p in $$list; do \ dir="`echo $$p | sed -e 's|/[^/]*$$||'`"; \ test "$$dir" != "$$p" || dir=.; \ echo "rm -f \"$${dir}/so_locations\""; \ rm -f "$${dir}/so_locations"; \ done libdft_neon_codelets.la: $(libdft_neon_codelets_la_OBJECTS) $(libdft_neon_codelets_la_DEPENDENCIES) $(EXTRA_libdft_neon_codelets_la_DEPENDENCIES) $(LINK) $(am_libdft_neon_codelets_la_rpath) $(libdft_neon_codelets_la_OBJECTS) $(libdft_neon_codelets_la_LIBADD) $(LIBS) mostlyclean-compile: -rm -f *.$(OBJEXT) distclean-compile: -rm -f *.tab.c @AMDEP_TRUE@@am__include@ @am__quote@./$(DEPDIR)/codlist.Plo@am__quote@ @AMDEP_TRUE@@am__include@ @am__quote@./$(DEPDIR)/genus.Plo@am__quote@ @AMDEP_TRUE@@am__include@ @am__quote@./$(DEPDIR)/n1bv_10.Plo@am__quote@ @AMDEP_TRUE@@am__include@ @am__quote@./$(DEPDIR)/n1bv_11.Plo@am__quote@ @AMDEP_TRUE@@am__include@ @am__quote@./$(DEPDIR)/n1bv_12.Plo@am__quote@ @AMDEP_TRUE@@am__include@ @am__quote@./$(DEPDIR)/n1bv_128.Plo@am__quote@ @AMDEP_TRUE@@am__include@ @am__quote@./$(DEPDIR)/n1bv_13.Plo@am__quote@ @AMDEP_TRUE@@am__include@ @am__quote@./$(DEPDIR)/n1bv_14.Plo@am__quote@ @AMDEP_TRUE@@am__include@ @am__quote@./$(DEPDIR)/n1bv_15.Plo@am__quote@ @AMDEP_TRUE@@am__include@ @am__quote@./$(DEPDIR)/n1bv_16.Plo@am__quote@ @AMDEP_TRUE@@am__include@ @am__quote@./$(DEPDIR)/n1bv_2.Plo@am__quote@ @AMDEP_TRUE@@am__include@ @am__quote@./$(DEPDIR)/n1bv_20.Plo@am__quote@ @AMDEP_TRUE@@am__include@ @am__quote@./$(DEPDIR)/n1bv_25.Plo@am__quote@ @AMDEP_TRUE@@am__include@ @am__quote@./$(DEPDIR)/n1bv_3.Plo@am__quote@ @AMDEP_TRUE@@am__include@ @am__quote@./$(DEPDIR)/n1bv_32.Plo@am__quote@ @AMDEP_TRUE@@am__include@ @am__quote@./$(DEPDIR)/n1bv_4.Plo@am__quote@ @AMDEP_TRUE@@am__include@ @am__quote@./$(DEPDIR)/n1bv_5.Plo@am__quote@ @AMDEP_TRUE@@am__include@ @am__quote@./$(DEPDIR)/n1bv_6.Plo@am__quote@ @AMDEP_TRUE@@am__include@ @am__quote@./$(DEPDIR)/n1bv_64.Plo@am__quote@ @AMDEP_TRUE@@am__include@ @am__quote@./$(DEPDIR)/n1bv_7.Plo@am__quote@ @AMDEP_TRUE@@am__include@ @am__quote@./$(DEPDIR)/n1bv_8.Plo@am__quote@ @AMDEP_TRUE@@am__include@ @am__quote@./$(DEPDIR)/n1bv_9.Plo@am__quote@ @AMDEP_TRUE@@am__include@ @am__quote@./$(DEPDIR)/n1fv_10.Plo@am__quote@ @AMDEP_TRUE@@am__include@ @am__quote@./$(DEPDIR)/n1fv_11.Plo@am__quote@ @AMDEP_TRUE@@am__include@ @am__quote@./$(DEPDIR)/n1fv_12.Plo@am__quote@ @AMDEP_TRUE@@am__include@ @am__quote@./$(DEPDIR)/n1fv_128.Plo@am__quote@ @AMDEP_TRUE@@am__include@ @am__quote@./$(DEPDIR)/n1fv_13.Plo@am__quote@ @AMDEP_TRUE@@am__include@ @am__quote@./$(DEPDIR)/n1fv_14.Plo@am__quote@ @AMDEP_TRUE@@am__include@ @am__quote@./$(DEPDIR)/n1fv_15.Plo@am__quote@ @AMDEP_TRUE@@am__include@ @am__quote@./$(DEPDIR)/n1fv_16.Plo@am__quote@ @AMDEP_TRUE@@am__include@ @am__quote@./$(DEPDIR)/n1fv_2.Plo@am__quote@ @AMDEP_TRUE@@am__include@ @am__quote@./$(DEPDIR)/n1fv_20.Plo@am__quote@ @AMDEP_TRUE@@am__include@ @am__quote@./$(DEPDIR)/n1fv_25.Plo@am__quote@ @AMDEP_TRUE@@am__include@ @am__quote@./$(DEPDIR)/n1fv_3.Plo@am__quote@ @AMDEP_TRUE@@am__include@ @am__quote@./$(DEPDIR)/n1fv_32.Plo@am__quote@ @AMDEP_TRUE@@am__include@ @am__quote@./$(DEPDIR)/n1fv_4.Plo@am__quote@ @AMDEP_TRUE@@am__include@ @am__quote@./$(DEPDIR)/n1fv_5.Plo@am__quote@ @AMDEP_TRUE@@am__include@ @am__quote@./$(DEPDIR)/n1fv_6.Plo@am__quote@ @AMDEP_TRUE@@am__include@ @am__quote@./$(DEPDIR)/n1fv_64.Plo@am__quote@ @AMDEP_TRUE@@am__include@ @am__quote@./$(DEPDIR)/n1fv_7.Plo@am__quote@ @AMDEP_TRUE@@am__include@ @am__quote@./$(DEPDIR)/n1fv_8.Plo@am__quote@ @AMDEP_TRUE@@am__include@ @am__quote@./$(DEPDIR)/n1fv_9.Plo@am__quote@ @AMDEP_TRUE@@am__include@ @am__quote@./$(DEPDIR)/n2bv_10.Plo@am__quote@ @AMDEP_TRUE@@am__include@ @am__quote@./$(DEPDIR)/n2bv_12.Plo@am__quote@ @AMDEP_TRUE@@am__include@ @am__quote@./$(DEPDIR)/n2bv_14.Plo@am__quote@ @AMDEP_TRUE@@am__include@ @am__quote@./$(DEPDIR)/n2bv_16.Plo@am__quote@ @AMDEP_TRUE@@am__include@ @am__quote@./$(DEPDIR)/n2bv_2.Plo@am__quote@ @AMDEP_TRUE@@am__include@ @am__quote@./$(DEPDIR)/n2bv_20.Plo@am__quote@ @AMDEP_TRUE@@am__include@ @am__quote@./$(DEPDIR)/n2bv_32.Plo@am__quote@ @AMDEP_TRUE@@am__include@ @am__quote@./$(DEPDIR)/n2bv_4.Plo@am__quote@ @AMDEP_TRUE@@am__include@ @am__quote@./$(DEPDIR)/n2bv_6.Plo@am__quote@ @AMDEP_TRUE@@am__include@ @am__quote@./$(DEPDIR)/n2bv_64.Plo@am__quote@ @AMDEP_TRUE@@am__include@ @am__quote@./$(DEPDIR)/n2bv_8.Plo@am__quote@ @AMDEP_TRUE@@am__include@ @am__quote@./$(DEPDIR)/n2fv_10.Plo@am__quote@ @AMDEP_TRUE@@am__include@ @am__quote@./$(DEPDIR)/n2fv_12.Plo@am__quote@ @AMDEP_TRUE@@am__include@ @am__quote@./$(DEPDIR)/n2fv_14.Plo@am__quote@ @AMDEP_TRUE@@am__include@ @am__quote@./$(DEPDIR)/n2fv_16.Plo@am__quote@ @AMDEP_TRUE@@am__include@ @am__quote@./$(DEPDIR)/n2fv_2.Plo@am__quote@ @AMDEP_TRUE@@am__include@ @am__quote@./$(DEPDIR)/n2fv_20.Plo@am__quote@ @AMDEP_TRUE@@am__include@ @am__quote@./$(DEPDIR)/n2fv_32.Plo@am__quote@ @AMDEP_TRUE@@am__include@ @am__quote@./$(DEPDIR)/n2fv_4.Plo@am__quote@ @AMDEP_TRUE@@am__include@ @am__quote@./$(DEPDIR)/n2fv_6.Plo@am__quote@ @AMDEP_TRUE@@am__include@ @am__quote@./$(DEPDIR)/n2fv_64.Plo@am__quote@ @AMDEP_TRUE@@am__include@ @am__quote@./$(DEPDIR)/n2fv_8.Plo@am__quote@ @AMDEP_TRUE@@am__include@ @am__quote@./$(DEPDIR)/n2sv_16.Plo@am__quote@ @AMDEP_TRUE@@am__include@ @am__quote@./$(DEPDIR)/n2sv_32.Plo@am__quote@ @AMDEP_TRUE@@am__include@ @am__quote@./$(DEPDIR)/n2sv_4.Plo@am__quote@ @AMDEP_TRUE@@am__include@ @am__quote@./$(DEPDIR)/n2sv_64.Plo@am__quote@ @AMDEP_TRUE@@am__include@ @am__quote@./$(DEPDIR)/n2sv_8.Plo@am__quote@ @AMDEP_TRUE@@am__include@ @am__quote@./$(DEPDIR)/q1bv_2.Plo@am__quote@ @AMDEP_TRUE@@am__include@ @am__quote@./$(DEPDIR)/q1bv_4.Plo@am__quote@ @AMDEP_TRUE@@am__include@ @am__quote@./$(DEPDIR)/q1bv_5.Plo@am__quote@ @AMDEP_TRUE@@am__include@ @am__quote@./$(DEPDIR)/q1bv_8.Plo@am__quote@ @AMDEP_TRUE@@am__include@ @am__quote@./$(DEPDIR)/q1fv_2.Plo@am__quote@ @AMDEP_TRUE@@am__include@ @am__quote@./$(DEPDIR)/q1fv_4.Plo@am__quote@ @AMDEP_TRUE@@am__include@ @am__quote@./$(DEPDIR)/q1fv_5.Plo@am__quote@ @AMDEP_TRUE@@am__include@ @am__quote@./$(DEPDIR)/q1fv_8.Plo@am__quote@ @AMDEP_TRUE@@am__include@ @am__quote@./$(DEPDIR)/t1buv_10.Plo@am__quote@ @AMDEP_TRUE@@am__include@ @am__quote@./$(DEPDIR)/t1buv_2.Plo@am__quote@ @AMDEP_TRUE@@am__include@ @am__quote@./$(DEPDIR)/t1buv_3.Plo@am__quote@ @AMDEP_TRUE@@am__include@ @am__quote@./$(DEPDIR)/t1buv_4.Plo@am__quote@ @AMDEP_TRUE@@am__include@ @am__quote@./$(DEPDIR)/t1buv_5.Plo@am__quote@ @AMDEP_TRUE@@am__include@ @am__quote@./$(DEPDIR)/t1buv_6.Plo@am__quote@ @AMDEP_TRUE@@am__include@ @am__quote@./$(DEPDIR)/t1buv_7.Plo@am__quote@ @AMDEP_TRUE@@am__include@ @am__quote@./$(DEPDIR)/t1buv_8.Plo@am__quote@ @AMDEP_TRUE@@am__include@ @am__quote@./$(DEPDIR)/t1buv_9.Plo@am__quote@ @AMDEP_TRUE@@am__include@ @am__quote@./$(DEPDIR)/t1bv_10.Plo@am__quote@ @AMDEP_TRUE@@am__include@ @am__quote@./$(DEPDIR)/t1bv_12.Plo@am__quote@ @AMDEP_TRUE@@am__include@ @am__quote@./$(DEPDIR)/t1bv_15.Plo@am__quote@ @AMDEP_TRUE@@am__include@ @am__quote@./$(DEPDIR)/t1bv_16.Plo@am__quote@ @AMDEP_TRUE@@am__include@ @am__quote@./$(DEPDIR)/t1bv_2.Plo@am__quote@ @AMDEP_TRUE@@am__include@ @am__quote@./$(DEPDIR)/t1bv_20.Plo@am__quote@ @AMDEP_TRUE@@am__include@ @am__quote@./$(DEPDIR)/t1bv_25.Plo@am__quote@ @AMDEP_TRUE@@am__include@ @am__quote@./$(DEPDIR)/t1bv_3.Plo@am__quote@ @AMDEP_TRUE@@am__include@ @am__quote@./$(DEPDIR)/t1bv_32.Plo@am__quote@ @AMDEP_TRUE@@am__include@ @am__quote@./$(DEPDIR)/t1bv_4.Plo@am__quote@ @AMDEP_TRUE@@am__include@ @am__quote@./$(DEPDIR)/t1bv_5.Plo@am__quote@ @AMDEP_TRUE@@am__include@ @am__quote@./$(DEPDIR)/t1bv_6.Plo@am__quote@ @AMDEP_TRUE@@am__include@ @am__quote@./$(DEPDIR)/t1bv_64.Plo@am__quote@ @AMDEP_TRUE@@am__include@ @am__quote@./$(DEPDIR)/t1bv_7.Plo@am__quote@ @AMDEP_TRUE@@am__include@ @am__quote@./$(DEPDIR)/t1bv_8.Plo@am__quote@ @AMDEP_TRUE@@am__include@ @am__quote@./$(DEPDIR)/t1bv_9.Plo@am__quote@ @AMDEP_TRUE@@am__include@ @am__quote@./$(DEPDIR)/t1fuv_10.Plo@am__quote@ @AMDEP_TRUE@@am__include@ @am__quote@./$(DEPDIR)/t1fuv_2.Plo@am__quote@ @AMDEP_TRUE@@am__include@ @am__quote@./$(DEPDIR)/t1fuv_3.Plo@am__quote@ @AMDEP_TRUE@@am__include@ @am__quote@./$(DEPDIR)/t1fuv_4.Plo@am__quote@ @AMDEP_TRUE@@am__include@ @am__quote@./$(DEPDIR)/t1fuv_5.Plo@am__quote@ @AMDEP_TRUE@@am__include@ @am__quote@./$(DEPDIR)/t1fuv_6.Plo@am__quote@ @AMDEP_TRUE@@am__include@ @am__quote@./$(DEPDIR)/t1fuv_7.Plo@am__quote@ @AMDEP_TRUE@@am__include@ @am__quote@./$(DEPDIR)/t1fuv_8.Plo@am__quote@ @AMDEP_TRUE@@am__include@ @am__quote@./$(DEPDIR)/t1fuv_9.Plo@am__quote@ @AMDEP_TRUE@@am__include@ @am__quote@./$(DEPDIR)/t1fv_10.Plo@am__quote@ @AMDEP_TRUE@@am__include@ @am__quote@./$(DEPDIR)/t1fv_12.Plo@am__quote@ @AMDEP_TRUE@@am__include@ @am__quote@./$(DEPDIR)/t1fv_15.Plo@am__quote@ @AMDEP_TRUE@@am__include@ @am__quote@./$(DEPDIR)/t1fv_16.Plo@am__quote@ @AMDEP_TRUE@@am__include@ @am__quote@./$(DEPDIR)/t1fv_2.Plo@am__quote@ @AMDEP_TRUE@@am__include@ @am__quote@./$(DEPDIR)/t1fv_20.Plo@am__quote@ @AMDEP_TRUE@@am__include@ @am__quote@./$(DEPDIR)/t1fv_25.Plo@am__quote@ @AMDEP_TRUE@@am__include@ @am__quote@./$(DEPDIR)/t1fv_3.Plo@am__quote@ @AMDEP_TRUE@@am__include@ @am__quote@./$(DEPDIR)/t1fv_32.Plo@am__quote@ @AMDEP_TRUE@@am__include@ @am__quote@./$(DEPDIR)/t1fv_4.Plo@am__quote@ @AMDEP_TRUE@@am__include@ @am__quote@./$(DEPDIR)/t1fv_5.Plo@am__quote@ @AMDEP_TRUE@@am__include@ @am__quote@./$(DEPDIR)/t1fv_6.Plo@am__quote@ @AMDEP_TRUE@@am__include@ @am__quote@./$(DEPDIR)/t1fv_64.Plo@am__quote@ @AMDEP_TRUE@@am__include@ @am__quote@./$(DEPDIR)/t1fv_7.Plo@am__quote@ @AMDEP_TRUE@@am__include@ @am__quote@./$(DEPDIR)/t1fv_8.Plo@am__quote@ @AMDEP_TRUE@@am__include@ @am__quote@./$(DEPDIR)/t1fv_9.Plo@am__quote@ @AMDEP_TRUE@@am__include@ @am__quote@./$(DEPDIR)/t1sv_16.Plo@am__quote@ @AMDEP_TRUE@@am__include@ @am__quote@./$(DEPDIR)/t1sv_2.Plo@am__quote@ @AMDEP_TRUE@@am__include@ @am__quote@./$(DEPDIR)/t1sv_32.Plo@am__quote@ @AMDEP_TRUE@@am__include@ @am__quote@./$(DEPDIR)/t1sv_4.Plo@am__quote@ @AMDEP_TRUE@@am__include@ @am__quote@./$(DEPDIR)/t1sv_8.Plo@am__quote@ @AMDEP_TRUE@@am__include@ @am__quote@./$(DEPDIR)/t2bv_10.Plo@am__quote@ @AMDEP_TRUE@@am__include@ @am__quote@./$(DEPDIR)/t2bv_16.Plo@am__quote@ @AMDEP_TRUE@@am__include@ @am__quote@./$(DEPDIR)/t2bv_2.Plo@am__quote@ @AMDEP_TRUE@@am__include@ @am__quote@./$(DEPDIR)/t2bv_20.Plo@am__quote@ @AMDEP_TRUE@@am__include@ @am__quote@./$(DEPDIR)/t2bv_25.Plo@am__quote@ @AMDEP_TRUE@@am__include@ @am__quote@./$(DEPDIR)/t2bv_32.Plo@am__quote@ @AMDEP_TRUE@@am__include@ @am__quote@./$(DEPDIR)/t2bv_4.Plo@am__quote@ @AMDEP_TRUE@@am__include@ @am__quote@./$(DEPDIR)/t2bv_5.Plo@am__quote@ @AMDEP_TRUE@@am__include@ @am__quote@./$(DEPDIR)/t2bv_64.Plo@am__quote@ @AMDEP_TRUE@@am__include@ @am__quote@./$(DEPDIR)/t2bv_8.Plo@am__quote@ @AMDEP_TRUE@@am__include@ @am__quote@./$(DEPDIR)/t2fv_10.Plo@am__quote@ @AMDEP_TRUE@@am__include@ @am__quote@./$(DEPDIR)/t2fv_16.Plo@am__quote@ @AMDEP_TRUE@@am__include@ @am__quote@./$(DEPDIR)/t2fv_2.Plo@am__quote@ @AMDEP_TRUE@@am__include@ @am__quote@./$(DEPDIR)/t2fv_20.Plo@am__quote@ @AMDEP_TRUE@@am__include@ @am__quote@./$(DEPDIR)/t2fv_25.Plo@am__quote@ @AMDEP_TRUE@@am__include@ @am__quote@./$(DEPDIR)/t2fv_32.Plo@am__quote@ @AMDEP_TRUE@@am__include@ @am__quote@./$(DEPDIR)/t2fv_4.Plo@am__quote@ @AMDEP_TRUE@@am__include@ @am__quote@./$(DEPDIR)/t2fv_5.Plo@am__quote@ @AMDEP_TRUE@@am__include@ @am__quote@./$(DEPDIR)/t2fv_64.Plo@am__quote@ @AMDEP_TRUE@@am__include@ @am__quote@./$(DEPDIR)/t2fv_8.Plo@am__quote@ @AMDEP_TRUE@@am__include@ @am__quote@./$(DEPDIR)/t2sv_16.Plo@am__quote@ @AMDEP_TRUE@@am__include@ @am__quote@./$(DEPDIR)/t2sv_32.Plo@am__quote@ @AMDEP_TRUE@@am__include@ @am__quote@./$(DEPDIR)/t2sv_4.Plo@am__quote@ @AMDEP_TRUE@@am__include@ @am__quote@./$(DEPDIR)/t2sv_8.Plo@am__quote@ @AMDEP_TRUE@@am__include@ @am__quote@./$(DEPDIR)/t3bv_10.Plo@am__quote@ @AMDEP_TRUE@@am__include@ @am__quote@./$(DEPDIR)/t3bv_16.Plo@am__quote@ @AMDEP_TRUE@@am__include@ @am__quote@./$(DEPDIR)/t3bv_20.Plo@am__quote@ @AMDEP_TRUE@@am__include@ @am__quote@./$(DEPDIR)/t3bv_25.Plo@am__quote@ @AMDEP_TRUE@@am__include@ @am__quote@./$(DEPDIR)/t3bv_32.Plo@am__quote@ @AMDEP_TRUE@@am__include@ @am__quote@./$(DEPDIR)/t3bv_4.Plo@am__quote@ @AMDEP_TRUE@@am__include@ @am__quote@./$(DEPDIR)/t3bv_5.Plo@am__quote@ @AMDEP_TRUE@@am__include@ @am__quote@./$(DEPDIR)/t3bv_8.Plo@am__quote@ @AMDEP_TRUE@@am__include@ @am__quote@./$(DEPDIR)/t3fv_10.Plo@am__quote@ @AMDEP_TRUE@@am__include@ @am__quote@./$(DEPDIR)/t3fv_16.Plo@am__quote@ @AMDEP_TRUE@@am__include@ @am__quote@./$(DEPDIR)/t3fv_20.Plo@am__quote@ @AMDEP_TRUE@@am__include@ @am__quote@./$(DEPDIR)/t3fv_25.Plo@am__quote@ @AMDEP_TRUE@@am__include@ @am__quote@./$(DEPDIR)/t3fv_32.Plo@am__quote@ @AMDEP_TRUE@@am__include@ @am__quote@./$(DEPDIR)/t3fv_4.Plo@am__quote@ @AMDEP_TRUE@@am__include@ @am__quote@./$(DEPDIR)/t3fv_5.Plo@am__quote@ @AMDEP_TRUE@@am__include@ @am__quote@./$(DEPDIR)/t3fv_8.Plo@am__quote@ .c.o: @am__fastdepCC_TRUE@ $(COMPILE) -MT $@ -MD -MP -MF $(DEPDIR)/$*.Tpo -c -o $@ $< @am__fastdepCC_TRUE@ $(am__mv) $(DEPDIR)/$*.Tpo $(DEPDIR)/$*.Po @AMDEP_TRUE@@am__fastdepCC_FALSE@ source='$<' object='$@' libtool=no @AMDEPBACKSLASH@ @AMDEP_TRUE@@am__fastdepCC_FALSE@ DEPDIR=$(DEPDIR) $(CCDEPMODE) $(depcomp) @AMDEPBACKSLASH@ @am__fastdepCC_FALSE@ $(COMPILE) -c $< .c.obj: @am__fastdepCC_TRUE@ $(COMPILE) -MT $@ -MD -MP -MF $(DEPDIR)/$*.Tpo -c -o $@ `$(CYGPATH_W) '$<'` @am__fastdepCC_TRUE@ $(am__mv) $(DEPDIR)/$*.Tpo $(DEPDIR)/$*.Po @AMDEP_TRUE@@am__fastdepCC_FALSE@ source='$<' object='$@' libtool=no @AMDEPBACKSLASH@ @AMDEP_TRUE@@am__fastdepCC_FALSE@ DEPDIR=$(DEPDIR) $(CCDEPMODE) $(depcomp) @AMDEPBACKSLASH@ @am__fastdepCC_FALSE@ $(COMPILE) -c `$(CYGPATH_W) '$<'` .c.lo: @am__fastdepCC_TRUE@ $(LTCOMPILE) -MT $@ -MD -MP -MF $(DEPDIR)/$*.Tpo -c -o $@ $< @am__fastdepCC_TRUE@ $(am__mv) $(DEPDIR)/$*.Tpo $(DEPDIR)/$*.Plo @AMDEP_TRUE@@am__fastdepCC_FALSE@ source='$<' object='$@' libtool=yes @AMDEPBACKSLASH@ @AMDEP_TRUE@@am__fastdepCC_FALSE@ DEPDIR=$(DEPDIR) $(CCDEPMODE) $(depcomp) @AMDEPBACKSLASH@ @am__fastdepCC_FALSE@ $(LTCOMPILE) -c -o $@ $< mostlyclean-libtool: -rm -f *.lo clean-libtool: -rm -rf .libs _libs ID: $(HEADERS) $(SOURCES) $(LISP) $(TAGS_FILES) list='$(SOURCES) $(HEADERS) $(LISP) $(TAGS_FILES)'; \ unique=`for i in $$list; do \ if test -f "$$i"; then echo $$i; else echo $(srcdir)/$$i; fi; \ done | \ $(AWK) '{ files[$$0] = 1; nonempty = 1; } \ END { if (nonempty) { for (i in files) print i; }; }'`; \ mkid -fID $$unique tags: TAGS TAGS: $(HEADERS) $(SOURCES) $(TAGS_DEPENDENCIES) \ $(TAGS_FILES) $(LISP) set x; \ here=`pwd`; \ list='$(SOURCES) $(HEADERS) $(LISP) $(TAGS_FILES)'; \ unique=`for i in $$list; do \ if test -f "$$i"; then echo $$i; else echo $(srcdir)/$$i; fi; \ done | \ $(AWK) '{ files[$$0] = 1; nonempty = 1; } \ END { if (nonempty) { for (i in files) print i; }; }'`; \ shift; \ if test -z "$(ETAGS_ARGS)$$*$$unique"; then :; else \ test -n "$$unique" || unique=$$empty_fix; \ if test $$# -gt 0; then \ $(ETAGS) $(ETAGSFLAGS) $(AM_ETAGSFLAGS) $(ETAGS_ARGS) \ "$$@" $$unique; \ else \ $(ETAGS) $(ETAGSFLAGS) $(AM_ETAGSFLAGS) $(ETAGS_ARGS) \ $$unique; \ fi; \ fi ctags: CTAGS CTAGS: $(HEADERS) $(SOURCES) $(TAGS_DEPENDENCIES) \ $(TAGS_FILES) $(LISP) list='$(SOURCES) $(HEADERS) $(LISP) $(TAGS_FILES)'; \ unique=`for i in $$list; do \ if test -f "$$i"; then echo $$i; else echo $(srcdir)/$$i; fi; \ done | \ $(AWK) '{ files[$$0] = 1; nonempty = 1; } \ END { if (nonempty) { for (i in files) print i; }; }'`; \ test -z "$(CTAGS_ARGS)$$unique" \ || $(CTAGS) $(CTAGSFLAGS) $(AM_CTAGSFLAGS) $(CTAGS_ARGS) \ $$unique GTAGS: here=`$(am__cd) $(top_builddir) && pwd` \ && $(am__cd) $(top_srcdir) \ && gtags -i $(GTAGS_ARGS) "$$here" distclean-tags: -rm -f TAGS ID GTAGS GRTAGS GSYMS GPATH tags distdir: $(DISTFILES) @srcdirstrip=`echo "$(srcdir)" | sed 's/[].[^$$\\*]/\\\\&/g'`; \ topsrcdirstrip=`echo "$(top_srcdir)" | sed 's/[].[^$$\\*]/\\\\&/g'`; \ list='$(DISTFILES)'; \ dist_files=`for file in $$list; do echo $$file; done | \ sed -e "s|^$$srcdirstrip/||;t" \ -e "s|^$$topsrcdirstrip/|$(top_builddir)/|;t"`; \ case $$dist_files in \ */*) $(MKDIR_P) `echo "$$dist_files" | \ sed '/\//!d;s|^|$(distdir)/|;s,/[^/]*$$,,' | \ sort -u` ;; \ esac; \ for file in $$dist_files; do \ if test -f $$file || test -d $$file; then d=.; else d=$(srcdir); fi; \ if test -d $$d/$$file; then \ dir=`echo "/$$file" | sed -e 's,/[^/]*$$,,'`; \ if test -d "$(distdir)/$$file"; then \ find "$(distdir)/$$file" -type d ! -perm -700 -exec chmod u+rwx {} \;; \ fi; \ if test -d $(srcdir)/$$file && test $$d != $(srcdir); then \ cp -fpR $(srcdir)/$$file "$(distdir)$$dir" || exit 1; \ find "$(distdir)/$$file" -type d ! -perm -700 -exec chmod u+rwx {} \;; \ fi; \ cp -fpR $$d/$$file "$(distdir)$$dir" || exit 1; \ else \ test -f "$(distdir)/$$file" \ || cp -p $$d/$$file "$(distdir)/$$file" \ || exit 1; \ fi; \ done check-am: all-am check: $(BUILT_SOURCES) $(MAKE) $(AM_MAKEFLAGS) check-am all-am: Makefile $(LTLIBRARIES) installdirs: install: $(BUILT_SOURCES) $(MAKE) $(AM_MAKEFLAGS) install-am install-exec: install-exec-am install-data: install-data-am uninstall: uninstall-am install-am: all-am @$(MAKE) $(AM_MAKEFLAGS) install-exec-am install-data-am installcheck: installcheck-am install-strip: if test -z '$(STRIP)'; then \ $(MAKE) $(AM_MAKEFLAGS) INSTALL_PROGRAM="$(INSTALL_STRIP_PROGRAM)" \ install_sh_PROGRAM="$(INSTALL_STRIP_PROGRAM)" INSTALL_STRIP_FLAG=-s \ install; \ else \ $(MAKE) $(AM_MAKEFLAGS) INSTALL_PROGRAM="$(INSTALL_STRIP_PROGRAM)" \ install_sh_PROGRAM="$(INSTALL_STRIP_PROGRAM)" INSTALL_STRIP_FLAG=-s \ "INSTALL_PROGRAM_ENV=STRIPPROG='$(STRIP)'" install; \ fi mostlyclean-generic: clean-generic: distclean-generic: -test -z "$(CONFIG_CLEAN_FILES)" || rm -f $(CONFIG_CLEAN_FILES) -test . = "$(srcdir)" || test -z "$(CONFIG_CLEAN_VPATH_FILES)" || rm -f $(CONFIG_CLEAN_VPATH_FILES) maintainer-clean-generic: @echo "This command is intended for maintainers to use" @echo "it deletes files that may require special tools to rebuild." -test -z "$(BUILT_SOURCES)" || rm -f $(BUILT_SOURCES) clean: clean-am clean-am: clean-generic clean-libtool clean-noinstLTLIBRARIES \ mostlyclean-am distclean: distclean-am -rm -rf ./$(DEPDIR) -rm -f Makefile distclean-am: clean-am distclean-compile distclean-generic \ distclean-tags dvi: dvi-am dvi-am: html: html-am html-am: info: info-am info-am: install-data-am: install-dvi: install-dvi-am install-dvi-am: install-exec-am: install-html: install-html-am install-html-am: install-info: install-info-am install-info-am: install-man: install-pdf: install-pdf-am install-pdf-am: install-ps: install-ps-am install-ps-am: installcheck-am: maintainer-clean: maintainer-clean-am -rm -rf ./$(DEPDIR) -rm -f Makefile maintainer-clean-am: distclean-am maintainer-clean-generic mostlyclean: mostlyclean-am mostlyclean-am: mostlyclean-compile mostlyclean-generic \ mostlyclean-libtool pdf: pdf-am pdf-am: ps: ps-am ps-am: uninstall-am: .MAKE: all check install install-am install-strip .PHONY: CTAGS GTAGS all all-am check check-am clean clean-generic \ clean-libtool clean-noinstLTLIBRARIES ctags distclean \ distclean-compile distclean-generic distclean-libtool \ distclean-tags distdir dvi dvi-am html html-am info info-am \ install install-am install-data install-data-am install-dvi \ install-dvi-am install-exec install-exec-am install-html \ install-html-am install-info install-info-am install-man \ install-pdf install-pdf-am install-ps install-ps-am \ install-strip installcheck installcheck-am installdirs \ maintainer-clean maintainer-clean-generic mostlyclean \ mostlyclean-compile mostlyclean-generic mostlyclean-libtool \ pdf pdf-am ps ps-am tags uninstall uninstall-am $(EXTRA_DIST): Makefile ( \ echo "/* Generated automatically. DO NOT EDIT! */"; \ echo "#define SIMD_HEADER \"$(SIMD_HEADER)\""; \ echo "#include \"../common/"$*".c\""; \ ) >$@ # Tell versions [3.59,3.63) of GNU make to not export all variables. # Otherwise a system limit (for SysV at least) may be exceeded. .NOEXPORT: fftw-3.3.3/dft/simd/neon/n1fv_2.c0000644000175000017500000000015512054411107013312 00000000000000/* Generated automatically. DO NOT EDIT! */ #define SIMD_HEADER "simd-neon.h" #include "../common/n1fv_2.c" fftw-3.3.3/dft/simd/neon/n1fv_3.c0000644000175000017500000000015512054411107013313 00000000000000/* Generated automatically. DO NOT EDIT! */ #define SIMD_HEADER "simd-neon.h" #include "../common/n1fv_3.c" fftw-3.3.3/dft/simd/neon/n1fv_4.c0000644000175000017500000000015512054411107013314 00000000000000/* Generated automatically. DO NOT EDIT! */ #define SIMD_HEADER "simd-neon.h" #include "../common/n1fv_4.c" fftw-3.3.3/dft/simd/neon/n1fv_5.c0000644000175000017500000000015512054411107013315 00000000000000/* Generated automatically. DO NOT EDIT! */ #define SIMD_HEADER "simd-neon.h" #include "../common/n1fv_5.c" fftw-3.3.3/dft/simd/neon/n1fv_6.c0000644000175000017500000000015512054411107013316 00000000000000/* Generated automatically. DO NOT EDIT! */ #define SIMD_HEADER "simd-neon.h" #include "../common/n1fv_6.c" fftw-3.3.3/dft/simd/neon/n1fv_7.c0000644000175000017500000000015512054411107013317 00000000000000/* Generated automatically. DO NOT EDIT! */ #define SIMD_HEADER "simd-neon.h" #include "../common/n1fv_7.c" fftw-3.3.3/dft/simd/neon/n1fv_8.c0000644000175000017500000000015512054411107013320 00000000000000/* Generated automatically. DO NOT EDIT! */ #define SIMD_HEADER "simd-neon.h" #include "../common/n1fv_8.c" fftw-3.3.3/dft/simd/neon/n1fv_9.c0000644000175000017500000000015512054411107013321 00000000000000/* Generated automatically. DO NOT EDIT! */ #define SIMD_HEADER "simd-neon.h" #include "../common/n1fv_9.c" fftw-3.3.3/dft/simd/neon/n1fv_10.c0000644000175000017500000000015612054411107013372 00000000000000/* Generated automatically. DO NOT EDIT! */ #define SIMD_HEADER "simd-neon.h" #include "../common/n1fv_10.c" fftw-3.3.3/dft/simd/neon/n1fv_11.c0000644000175000017500000000015612054411107013373 00000000000000/* Generated automatically. DO NOT EDIT! */ #define SIMD_HEADER "simd-neon.h" #include "../common/n1fv_11.c" fftw-3.3.3/dft/simd/neon/n1fv_12.c0000644000175000017500000000015612054411107013374 00000000000000/* Generated automatically. DO NOT EDIT! */ #define SIMD_HEADER "simd-neon.h" #include "../common/n1fv_12.c" fftw-3.3.3/dft/simd/neon/n1fv_13.c0000644000175000017500000000015612054411107013375 00000000000000/* Generated automatically. DO NOT EDIT! */ #define SIMD_HEADER "simd-neon.h" #include "../common/n1fv_13.c" fftw-3.3.3/dft/simd/neon/n1fv_14.c0000644000175000017500000000015612054411107013376 00000000000000/* Generated automatically. DO NOT EDIT! */ #define SIMD_HEADER "simd-neon.h" #include "../common/n1fv_14.c" fftw-3.3.3/dft/simd/neon/n1fv_15.c0000644000175000017500000000015612054411107013377 00000000000000/* Generated automatically. DO NOT EDIT! */ #define SIMD_HEADER "simd-neon.h" #include "../common/n1fv_15.c" fftw-3.3.3/dft/simd/neon/n1fv_16.c0000644000175000017500000000015612054411107013400 00000000000000/* Generated automatically. DO NOT EDIT! */ #define SIMD_HEADER "simd-neon.h" #include "../common/n1fv_16.c" fftw-3.3.3/dft/simd/neon/n1fv_32.c0000644000175000017500000000015612054411107013376 00000000000000/* Generated automatically. DO NOT EDIT! */ #define SIMD_HEADER "simd-neon.h" #include "../common/n1fv_32.c" fftw-3.3.3/dft/simd/neon/n1fv_64.c0000644000175000017500000000015612054411107013403 00000000000000/* Generated automatically. DO NOT EDIT! */ #define SIMD_HEADER "simd-neon.h" #include "../common/n1fv_64.c" fftw-3.3.3/dft/simd/neon/n1fv_128.c0000644000175000017500000000015712054411107013465 00000000000000/* Generated automatically. DO NOT EDIT! */ #define SIMD_HEADER "simd-neon.h" #include "../common/n1fv_128.c" fftw-3.3.3/dft/simd/neon/n1fv_20.c0000644000175000017500000000015612054411107013373 00000000000000/* Generated automatically. DO NOT EDIT! */ #define SIMD_HEADER "simd-neon.h" #include "../common/n1fv_20.c" fftw-3.3.3/dft/simd/neon/n1fv_25.c0000644000175000017500000000015612054411107013400 00000000000000/* Generated automatically. DO NOT EDIT! */ #define SIMD_HEADER "simd-neon.h" #include "../common/n1fv_25.c" fftw-3.3.3/dft/simd/neon/n1bv_2.c0000644000175000017500000000015512054411107013306 00000000000000/* Generated automatically. DO NOT EDIT! */ #define SIMD_HEADER "simd-neon.h" #include "../common/n1bv_2.c" fftw-3.3.3/dft/simd/neon/n1bv_3.c0000644000175000017500000000015512054411107013307 00000000000000/* Generated automatically. DO NOT EDIT! */ #define SIMD_HEADER "simd-neon.h" #include "../common/n1bv_3.c" fftw-3.3.3/dft/simd/neon/n1bv_4.c0000644000175000017500000000015512054411107013310 00000000000000/* Generated automatically. DO NOT EDIT! */ #define SIMD_HEADER "simd-neon.h" #include "../common/n1bv_4.c" fftw-3.3.3/dft/simd/neon/n1bv_5.c0000644000175000017500000000015512054411107013311 00000000000000/* Generated automatically. DO NOT EDIT! */ #define SIMD_HEADER "simd-neon.h" #include "../common/n1bv_5.c" fftw-3.3.3/dft/simd/neon/n1bv_6.c0000644000175000017500000000015512054411107013312 00000000000000/* Generated automatically. DO NOT EDIT! */ #define SIMD_HEADER "simd-neon.h" #include "../common/n1bv_6.c" fftw-3.3.3/dft/simd/neon/n1bv_7.c0000644000175000017500000000015512054411107013313 00000000000000/* Generated automatically. DO NOT EDIT! */ #define SIMD_HEADER "simd-neon.h" #include "../common/n1bv_7.c" fftw-3.3.3/dft/simd/neon/n1bv_8.c0000644000175000017500000000015512054411107013314 00000000000000/* Generated automatically. DO NOT EDIT! */ #define SIMD_HEADER "simd-neon.h" #include "../common/n1bv_8.c" fftw-3.3.3/dft/simd/neon/n1bv_9.c0000644000175000017500000000015512054411107013315 00000000000000/* Generated automatically. DO NOT EDIT! */ #define SIMD_HEADER "simd-neon.h" #include "../common/n1bv_9.c" fftw-3.3.3/dft/simd/neon/n1bv_10.c0000644000175000017500000000015612054411107013366 00000000000000/* Generated automatically. DO NOT EDIT! */ #define SIMD_HEADER "simd-neon.h" #include "../common/n1bv_10.c" fftw-3.3.3/dft/simd/neon/n1bv_11.c0000644000175000017500000000015612054411107013367 00000000000000/* Generated automatically. DO NOT EDIT! */ #define SIMD_HEADER "simd-neon.h" #include "../common/n1bv_11.c" fftw-3.3.3/dft/simd/neon/n1bv_12.c0000644000175000017500000000015612054411107013370 00000000000000/* Generated automatically. DO NOT EDIT! */ #define SIMD_HEADER "simd-neon.h" #include "../common/n1bv_12.c" fftw-3.3.3/dft/simd/neon/n1bv_13.c0000644000175000017500000000015612054411107013371 00000000000000/* Generated automatically. DO NOT EDIT! */ #define SIMD_HEADER "simd-neon.h" #include "../common/n1bv_13.c" fftw-3.3.3/dft/simd/neon/n1bv_14.c0000644000175000017500000000015612054411107013372 00000000000000/* Generated automatically. DO NOT EDIT! */ #define SIMD_HEADER "simd-neon.h" #include "../common/n1bv_14.c" fftw-3.3.3/dft/simd/neon/n1bv_15.c0000644000175000017500000000015612054411107013373 00000000000000/* Generated automatically. DO NOT EDIT! */ #define SIMD_HEADER "simd-neon.h" #include "../common/n1bv_15.c" fftw-3.3.3/dft/simd/neon/n1bv_16.c0000644000175000017500000000015612054411107013374 00000000000000/* Generated automatically. DO NOT EDIT! */ #define SIMD_HEADER "simd-neon.h" #include "../common/n1bv_16.c" fftw-3.3.3/dft/simd/neon/n1bv_32.c0000644000175000017500000000015612054411107013372 00000000000000/* Generated automatically. DO NOT EDIT! */ #define SIMD_HEADER "simd-neon.h" #include "../common/n1bv_32.c" fftw-3.3.3/dft/simd/neon/n1bv_64.c0000644000175000017500000000015612054411107013377 00000000000000/* Generated automatically. DO NOT EDIT! */ #define SIMD_HEADER "simd-neon.h" #include "../common/n1bv_64.c" fftw-3.3.3/dft/simd/neon/n1bv_128.c0000644000175000017500000000015712054411107013461 00000000000000/* Generated automatically. DO NOT EDIT! */ #define SIMD_HEADER "simd-neon.h" #include "../common/n1bv_128.c" fftw-3.3.3/dft/simd/neon/n1bv_20.c0000644000175000017500000000015612054411107013367 00000000000000/* Generated automatically. DO NOT EDIT! */ #define SIMD_HEADER "simd-neon.h" #include "../common/n1bv_20.c" fftw-3.3.3/dft/simd/neon/n1bv_25.c0000644000175000017500000000015612054411107013374 00000000000000/* Generated automatically. DO NOT EDIT! */ #define SIMD_HEADER "simd-neon.h" #include "../common/n1bv_25.c" fftw-3.3.3/dft/simd/neon/n2fv_2.c0000644000175000017500000000015512054411107013313 00000000000000/* Generated automatically. DO NOT EDIT! */ #define SIMD_HEADER "simd-neon.h" #include "../common/n2fv_2.c" fftw-3.3.3/dft/simd/neon/n2fv_4.c0000644000175000017500000000015512054411107013315 00000000000000/* Generated automatically. DO NOT EDIT! */ #define SIMD_HEADER "simd-neon.h" #include "../common/n2fv_4.c" fftw-3.3.3/dft/simd/neon/n2fv_6.c0000644000175000017500000000015512054411107013317 00000000000000/* Generated automatically. DO NOT EDIT! */ #define SIMD_HEADER "simd-neon.h" #include "../common/n2fv_6.c" fftw-3.3.3/dft/simd/neon/n2fv_8.c0000644000175000017500000000015512054411107013321 00000000000000/* Generated automatically. DO NOT EDIT! */ #define SIMD_HEADER "simd-neon.h" #include "../common/n2fv_8.c" fftw-3.3.3/dft/simd/neon/n2fv_10.c0000644000175000017500000000015612054411107013373 00000000000000/* Generated automatically. DO NOT EDIT! */ #define SIMD_HEADER "simd-neon.h" #include "../common/n2fv_10.c" fftw-3.3.3/dft/simd/neon/n2fv_12.c0000644000175000017500000000015612054411107013375 00000000000000/* Generated automatically. DO NOT EDIT! */ #define SIMD_HEADER "simd-neon.h" #include "../common/n2fv_12.c" fftw-3.3.3/dft/simd/neon/n2fv_14.c0000644000175000017500000000015612054411107013377 00000000000000/* Generated automatically. DO NOT EDIT! */ #define SIMD_HEADER "simd-neon.h" #include "../common/n2fv_14.c" fftw-3.3.3/dft/simd/neon/n2fv_16.c0000644000175000017500000000015612054411107013401 00000000000000/* Generated automatically. DO NOT EDIT! */ #define SIMD_HEADER "simd-neon.h" #include "../common/n2fv_16.c" fftw-3.3.3/dft/simd/neon/n2fv_32.c0000644000175000017500000000015612054411107013377 00000000000000/* Generated automatically. DO NOT EDIT! */ #define SIMD_HEADER "simd-neon.h" #include "../common/n2fv_32.c" fftw-3.3.3/dft/simd/neon/n2fv_64.c0000644000175000017500000000015612054411107013404 00000000000000/* Generated automatically. DO NOT EDIT! */ #define SIMD_HEADER "simd-neon.h" #include "../common/n2fv_64.c" fftw-3.3.3/dft/simd/neon/n2fv_20.c0000644000175000017500000000015612054411107013374 00000000000000/* Generated automatically. DO NOT EDIT! */ #define SIMD_HEADER "simd-neon.h" #include "../common/n2fv_20.c" fftw-3.3.3/dft/simd/neon/n2bv_2.c0000644000175000017500000000015512054411107013307 00000000000000/* Generated automatically. DO NOT EDIT! */ #define SIMD_HEADER "simd-neon.h" #include "../common/n2bv_2.c" fftw-3.3.3/dft/simd/neon/n2bv_4.c0000644000175000017500000000015512054411107013311 00000000000000/* Generated automatically. DO NOT EDIT! */ #define SIMD_HEADER "simd-neon.h" #include "../common/n2bv_4.c" fftw-3.3.3/dft/simd/neon/n2bv_6.c0000644000175000017500000000015512054411107013313 00000000000000/* Generated automatically. DO NOT EDIT! */ #define SIMD_HEADER "simd-neon.h" #include "../common/n2bv_6.c" fftw-3.3.3/dft/simd/neon/n2bv_8.c0000644000175000017500000000015512054411107013315 00000000000000/* Generated automatically. DO NOT EDIT! */ #define SIMD_HEADER "simd-neon.h" #include "../common/n2bv_8.c" fftw-3.3.3/dft/simd/neon/n2bv_10.c0000644000175000017500000000015612054411107013367 00000000000000/* Generated automatically. DO NOT EDIT! */ #define SIMD_HEADER "simd-neon.h" #include "../common/n2bv_10.c" fftw-3.3.3/dft/simd/neon/n2bv_12.c0000644000175000017500000000015612054411107013371 00000000000000/* Generated automatically. DO NOT EDIT! */ #define SIMD_HEADER "simd-neon.h" #include "../common/n2bv_12.c" fftw-3.3.3/dft/simd/neon/n2bv_14.c0000644000175000017500000000015612054411107013373 00000000000000/* Generated automatically. DO NOT EDIT! */ #define SIMD_HEADER "simd-neon.h" #include "../common/n2bv_14.c" fftw-3.3.3/dft/simd/neon/n2bv_16.c0000644000175000017500000000015612054411107013375 00000000000000/* Generated automatically. DO NOT EDIT! */ #define SIMD_HEADER "simd-neon.h" #include "../common/n2bv_16.c" fftw-3.3.3/dft/simd/neon/n2bv_32.c0000644000175000017500000000015612054411107013373 00000000000000/* Generated automatically. DO NOT EDIT! */ #define SIMD_HEADER "simd-neon.h" #include "../common/n2bv_32.c" fftw-3.3.3/dft/simd/neon/n2bv_64.c0000644000175000017500000000015612054411107013400 00000000000000/* Generated automatically. DO NOT EDIT! */ #define SIMD_HEADER "simd-neon.h" #include "../common/n2bv_64.c" fftw-3.3.3/dft/simd/neon/n2bv_20.c0000644000175000017500000000015612054411107013370 00000000000000/* Generated automatically. DO NOT EDIT! */ #define SIMD_HEADER "simd-neon.h" #include "../common/n2bv_20.c" fftw-3.3.3/dft/simd/neon/n2sv_4.c0000644000175000017500000000015512054411107013332 00000000000000/* Generated automatically. DO NOT EDIT! */ #define SIMD_HEADER "simd-neon.h" #include "../common/n2sv_4.c" fftw-3.3.3/dft/simd/neon/n2sv_8.c0000644000175000017500000000015512054411107013336 00000000000000/* Generated automatically. DO NOT EDIT! */ #define SIMD_HEADER "simd-neon.h" #include "../common/n2sv_8.c" fftw-3.3.3/dft/simd/neon/n2sv_16.c0000644000175000017500000000015612054411107013416 00000000000000/* Generated automatically. DO NOT EDIT! */ #define SIMD_HEADER "simd-neon.h" #include "../common/n2sv_16.c" fftw-3.3.3/dft/simd/neon/n2sv_32.c0000644000175000017500000000015612054411107013414 00000000000000/* Generated automatically. DO NOT EDIT! */ #define SIMD_HEADER "simd-neon.h" #include "../common/n2sv_32.c" fftw-3.3.3/dft/simd/neon/n2sv_64.c0000644000175000017500000000015612054411107013421 00000000000000/* Generated automatically. DO NOT EDIT! */ #define SIMD_HEADER "simd-neon.h" #include "../common/n2sv_64.c" fftw-3.3.3/dft/simd/neon/t1fuv_2.c0000644000175000017500000000015612054411107013506 00000000000000/* Generated automatically. DO NOT EDIT! */ #define SIMD_HEADER "simd-neon.h" #include "../common/t1fuv_2.c" fftw-3.3.3/dft/simd/neon/t1fuv_3.c0000644000175000017500000000015612054411107013507 00000000000000/* Generated automatically. DO NOT EDIT! */ #define SIMD_HEADER "simd-neon.h" #include "../common/t1fuv_3.c" fftw-3.3.3/dft/simd/neon/t1fuv_4.c0000644000175000017500000000015612054411107013510 00000000000000/* Generated automatically. DO NOT EDIT! */ #define SIMD_HEADER "simd-neon.h" #include "../common/t1fuv_4.c" fftw-3.3.3/dft/simd/neon/t1fuv_5.c0000644000175000017500000000015612054411107013511 00000000000000/* Generated automatically. DO NOT EDIT! */ #define SIMD_HEADER "simd-neon.h" #include "../common/t1fuv_5.c" fftw-3.3.3/dft/simd/neon/t1fuv_6.c0000644000175000017500000000015612054411107013512 00000000000000/* Generated automatically. DO NOT EDIT! */ #define SIMD_HEADER "simd-neon.h" #include "../common/t1fuv_6.c" fftw-3.3.3/dft/simd/neon/t1fuv_7.c0000644000175000017500000000015612054411107013513 00000000000000/* Generated automatically. DO NOT EDIT! */ #define SIMD_HEADER "simd-neon.h" #include "../common/t1fuv_7.c" fftw-3.3.3/dft/simd/neon/t1fuv_8.c0000644000175000017500000000015612054411107013514 00000000000000/* Generated automatically. DO NOT EDIT! */ #define SIMD_HEADER "simd-neon.h" #include "../common/t1fuv_8.c" fftw-3.3.3/dft/simd/neon/t1fuv_9.c0000644000175000017500000000015612054411107013515 00000000000000/* Generated automatically. DO NOT EDIT! */ #define SIMD_HEADER "simd-neon.h" #include "../common/t1fuv_9.c" fftw-3.3.3/dft/simd/neon/t1fuv_10.c0000644000175000017500000000015712054411107013566 00000000000000/* Generated automatically. DO NOT EDIT! */ #define SIMD_HEADER "simd-neon.h" #include "../common/t1fuv_10.c" fftw-3.3.3/dft/simd/neon/t1fv_2.c0000644000175000017500000000015512054411107013320 00000000000000/* Generated automatically. DO NOT EDIT! */ #define SIMD_HEADER "simd-neon.h" #include "../common/t1fv_2.c" fftw-3.3.3/dft/simd/neon/t1fv_3.c0000644000175000017500000000015512054411107013321 00000000000000/* Generated automatically. DO NOT EDIT! */ #define SIMD_HEADER "simd-neon.h" #include "../common/t1fv_3.c" fftw-3.3.3/dft/simd/neon/t1fv_4.c0000644000175000017500000000015512054411107013322 00000000000000/* Generated automatically. DO NOT EDIT! */ #define SIMD_HEADER "simd-neon.h" #include "../common/t1fv_4.c" fftw-3.3.3/dft/simd/neon/t1fv_5.c0000644000175000017500000000015512054411107013323 00000000000000/* Generated automatically. DO NOT EDIT! */ #define SIMD_HEADER "simd-neon.h" #include "../common/t1fv_5.c" fftw-3.3.3/dft/simd/neon/t1fv_6.c0000644000175000017500000000015512054411107013324 00000000000000/* Generated automatically. DO NOT EDIT! */ #define SIMD_HEADER "simd-neon.h" #include "../common/t1fv_6.c" fftw-3.3.3/dft/simd/neon/t1fv_7.c0000644000175000017500000000015512054411107013325 00000000000000/* Generated automatically. DO NOT EDIT! */ #define SIMD_HEADER "simd-neon.h" #include "../common/t1fv_7.c" fftw-3.3.3/dft/simd/neon/t1fv_8.c0000644000175000017500000000015512054411107013326 00000000000000/* Generated automatically. DO NOT EDIT! */ #define SIMD_HEADER "simd-neon.h" #include "../common/t1fv_8.c" fftw-3.3.3/dft/simd/neon/t1fv_9.c0000644000175000017500000000015512054411107013327 00000000000000/* Generated automatically. DO NOT EDIT! */ #define SIMD_HEADER "simd-neon.h" #include "../common/t1fv_9.c" fftw-3.3.3/dft/simd/neon/t1fv_10.c0000644000175000017500000000015612054411107013400 00000000000000/* Generated automatically. DO NOT EDIT! */ #define SIMD_HEADER "simd-neon.h" #include "../common/t1fv_10.c" fftw-3.3.3/dft/simd/neon/t1fv_12.c0000644000175000017500000000015612054411107013402 00000000000000/* Generated automatically. DO NOT EDIT! */ #define SIMD_HEADER "simd-neon.h" #include "../common/t1fv_12.c" fftw-3.3.3/dft/simd/neon/t1fv_15.c0000644000175000017500000000015612054411107013405 00000000000000/* Generated automatically. DO NOT EDIT! */ #define SIMD_HEADER "simd-neon.h" #include "../common/t1fv_15.c" fftw-3.3.3/dft/simd/neon/t1fv_16.c0000644000175000017500000000015612054411107013406 00000000000000/* Generated automatically. DO NOT EDIT! */ #define SIMD_HEADER "simd-neon.h" #include "../common/t1fv_16.c" fftw-3.3.3/dft/simd/neon/t1fv_32.c0000644000175000017500000000015612054411107013404 00000000000000/* Generated automatically. DO NOT EDIT! */ #define SIMD_HEADER "simd-neon.h" #include "../common/t1fv_32.c" fftw-3.3.3/dft/simd/neon/t1fv_64.c0000644000175000017500000000015612054411107013411 00000000000000/* Generated automatically. DO NOT EDIT! */ #define SIMD_HEADER "simd-neon.h" #include "../common/t1fv_64.c" fftw-3.3.3/dft/simd/neon/t1fv_20.c0000644000175000017500000000015612054411107013401 00000000000000/* Generated automatically. DO NOT EDIT! */ #define SIMD_HEADER "simd-neon.h" #include "../common/t1fv_20.c" fftw-3.3.3/dft/simd/neon/t1fv_25.c0000644000175000017500000000015612054411107013406 00000000000000/* Generated automatically. DO NOT EDIT! */ #define SIMD_HEADER "simd-neon.h" #include "../common/t1fv_25.c" fftw-3.3.3/dft/simd/neon/t2fv_2.c0000644000175000017500000000015512054411107013321 00000000000000/* Generated automatically. DO NOT EDIT! */ #define SIMD_HEADER "simd-neon.h" #include "../common/t2fv_2.c" fftw-3.3.3/dft/simd/neon/t2fv_4.c0000644000175000017500000000015512054411107013323 00000000000000/* Generated automatically. DO NOT EDIT! */ #define SIMD_HEADER "simd-neon.h" #include "../common/t2fv_4.c" fftw-3.3.3/dft/simd/neon/t2fv_8.c0000644000175000017500000000015512054411107013327 00000000000000/* Generated automatically. DO NOT EDIT! */ #define SIMD_HEADER "simd-neon.h" #include "../common/t2fv_8.c" fftw-3.3.3/dft/simd/neon/t2fv_16.c0000644000175000017500000000015612054411107013407 00000000000000/* Generated automatically. DO NOT EDIT! */ #define SIMD_HEADER "simd-neon.h" #include "../common/t2fv_16.c" fftw-3.3.3/dft/simd/neon/t2fv_32.c0000644000175000017500000000015612054411107013405 00000000000000/* Generated automatically. DO NOT EDIT! */ #define SIMD_HEADER "simd-neon.h" #include "../common/t2fv_32.c" fftw-3.3.3/dft/simd/neon/t2fv_64.c0000644000175000017500000000015612054411107013412 00000000000000/* Generated automatically. DO NOT EDIT! */ #define SIMD_HEADER "simd-neon.h" #include "../common/t2fv_64.c" fftw-3.3.3/dft/simd/neon/t2fv_5.c0000644000175000017500000000015512054411107013324 00000000000000/* Generated automatically. DO NOT EDIT! */ #define SIMD_HEADER "simd-neon.h" #include "../common/t2fv_5.c" fftw-3.3.3/dft/simd/neon/t2fv_10.c0000644000175000017500000000015612054411107013401 00000000000000/* Generated automatically. DO NOT EDIT! */ #define SIMD_HEADER "simd-neon.h" #include "../common/t2fv_10.c" fftw-3.3.3/dft/simd/neon/t2fv_20.c0000644000175000017500000000015612054411107013402 00000000000000/* Generated automatically. DO NOT EDIT! */ #define SIMD_HEADER "simd-neon.h" #include "../common/t2fv_20.c" fftw-3.3.3/dft/simd/neon/t2fv_25.c0000644000175000017500000000015612054411107013407 00000000000000/* Generated automatically. DO NOT EDIT! */ #define SIMD_HEADER "simd-neon.h" #include "../common/t2fv_25.c" fftw-3.3.3/dft/simd/neon/t3fv_4.c0000644000175000017500000000015512054411107013324 00000000000000/* Generated automatically. DO NOT EDIT! */ #define SIMD_HEADER "simd-neon.h" #include "../common/t3fv_4.c" fftw-3.3.3/dft/simd/neon/t3fv_8.c0000644000175000017500000000015512054411107013330 00000000000000/* Generated automatically. DO NOT EDIT! */ #define SIMD_HEADER "simd-neon.h" #include "../common/t3fv_8.c" fftw-3.3.3/dft/simd/neon/t3fv_16.c0000644000175000017500000000015612054411107013410 00000000000000/* Generated automatically. DO NOT EDIT! */ #define SIMD_HEADER "simd-neon.h" #include "../common/t3fv_16.c" fftw-3.3.3/dft/simd/neon/t3fv_32.c0000644000175000017500000000015612054411107013406 00000000000000/* Generated automatically. DO NOT EDIT! */ #define SIMD_HEADER "simd-neon.h" #include "../common/t3fv_32.c" fftw-3.3.3/dft/simd/neon/t3fv_5.c0000644000175000017500000000015512054411107013325 00000000000000/* Generated automatically. DO NOT EDIT! */ #define SIMD_HEADER "simd-neon.h" #include "../common/t3fv_5.c" fftw-3.3.3/dft/simd/neon/t3fv_10.c0000644000175000017500000000015612054411107013402 00000000000000/* Generated automatically. DO NOT EDIT! */ #define SIMD_HEADER "simd-neon.h" #include "../common/t3fv_10.c" fftw-3.3.3/dft/simd/neon/t3fv_20.c0000644000175000017500000000015612054411107013403 00000000000000/* Generated automatically. DO NOT EDIT! */ #define SIMD_HEADER "simd-neon.h" #include "../common/t3fv_20.c" fftw-3.3.3/dft/simd/neon/t3fv_25.c0000644000175000017500000000015612054411107013410 00000000000000/* Generated automatically. DO NOT EDIT! */ #define SIMD_HEADER "simd-neon.h" #include "../common/t3fv_25.c" fftw-3.3.3/dft/simd/neon/t1buv_2.c0000644000175000017500000000015612054411107013502 00000000000000/* Generated automatically. DO NOT EDIT! */ #define SIMD_HEADER "simd-neon.h" #include "../common/t1buv_2.c" fftw-3.3.3/dft/simd/neon/t1buv_3.c0000644000175000017500000000015612054411107013503 00000000000000/* Generated automatically. DO NOT EDIT! */ #define SIMD_HEADER "simd-neon.h" #include "../common/t1buv_3.c" fftw-3.3.3/dft/simd/neon/t1buv_4.c0000644000175000017500000000015612054411107013504 00000000000000/* Generated automatically. DO NOT EDIT! */ #define SIMD_HEADER "simd-neon.h" #include "../common/t1buv_4.c" fftw-3.3.3/dft/simd/neon/t1buv_5.c0000644000175000017500000000015612054411107013505 00000000000000/* Generated automatically. DO NOT EDIT! */ #define SIMD_HEADER "simd-neon.h" #include "../common/t1buv_5.c" fftw-3.3.3/dft/simd/neon/t1buv_6.c0000644000175000017500000000015612054411107013506 00000000000000/* Generated automatically. DO NOT EDIT! */ #define SIMD_HEADER "simd-neon.h" #include "../common/t1buv_6.c" fftw-3.3.3/dft/simd/neon/t1buv_7.c0000644000175000017500000000015612054411107013507 00000000000000/* Generated automatically. DO NOT EDIT! */ #define SIMD_HEADER "simd-neon.h" #include "../common/t1buv_7.c" fftw-3.3.3/dft/simd/neon/t1buv_8.c0000644000175000017500000000015612054411107013510 00000000000000/* Generated automatically. DO NOT EDIT! */ #define SIMD_HEADER "simd-neon.h" #include "../common/t1buv_8.c" fftw-3.3.3/dft/simd/neon/t1buv_9.c0000644000175000017500000000015612054411107013511 00000000000000/* Generated automatically. DO NOT EDIT! */ #define SIMD_HEADER "simd-neon.h" #include "../common/t1buv_9.c" fftw-3.3.3/dft/simd/neon/t1buv_10.c0000644000175000017500000000015712054411107013562 00000000000000/* Generated automatically. DO NOT EDIT! */ #define SIMD_HEADER "simd-neon.h" #include "../common/t1buv_10.c" fftw-3.3.3/dft/simd/neon/t1bv_2.c0000644000175000017500000000015512054411107013314 00000000000000/* Generated automatically. DO NOT EDIT! */ #define SIMD_HEADER "simd-neon.h" #include "../common/t1bv_2.c" fftw-3.3.3/dft/simd/neon/t1bv_3.c0000644000175000017500000000015512054411107013315 00000000000000/* Generated automatically. DO NOT EDIT! */ #define SIMD_HEADER "simd-neon.h" #include "../common/t1bv_3.c" fftw-3.3.3/dft/simd/neon/t1bv_4.c0000644000175000017500000000015512054411107013316 00000000000000/* Generated automatically. DO NOT EDIT! */ #define SIMD_HEADER "simd-neon.h" #include "../common/t1bv_4.c" fftw-3.3.3/dft/simd/neon/t1bv_5.c0000644000175000017500000000015512054411107013317 00000000000000/* Generated automatically. DO NOT EDIT! */ #define SIMD_HEADER "simd-neon.h" #include "../common/t1bv_5.c" fftw-3.3.3/dft/simd/neon/t1bv_6.c0000644000175000017500000000015512054411107013320 00000000000000/* Generated automatically. DO NOT EDIT! */ #define SIMD_HEADER "simd-neon.h" #include "../common/t1bv_6.c" fftw-3.3.3/dft/simd/neon/t1bv_7.c0000644000175000017500000000015512054411107013321 00000000000000/* Generated automatically. DO NOT EDIT! */ #define SIMD_HEADER "simd-neon.h" #include "../common/t1bv_7.c" fftw-3.3.3/dft/simd/neon/t1bv_8.c0000644000175000017500000000015512054411107013322 00000000000000/* Generated automatically. DO NOT EDIT! */ #define SIMD_HEADER "simd-neon.h" #include "../common/t1bv_8.c" fftw-3.3.3/dft/simd/neon/t1bv_9.c0000644000175000017500000000015512054411107013323 00000000000000/* Generated automatically. DO NOT EDIT! */ #define SIMD_HEADER "simd-neon.h" #include "../common/t1bv_9.c" fftw-3.3.3/dft/simd/neon/t1bv_10.c0000644000175000017500000000015612054411107013374 00000000000000/* Generated automatically. DO NOT EDIT! */ #define SIMD_HEADER "simd-neon.h" #include "../common/t1bv_10.c" fftw-3.3.3/dft/simd/neon/t1bv_12.c0000644000175000017500000000015612054411107013376 00000000000000/* Generated automatically. DO NOT EDIT! */ #define SIMD_HEADER "simd-neon.h" #include "../common/t1bv_12.c" fftw-3.3.3/dft/simd/neon/t1bv_15.c0000644000175000017500000000015612054411107013401 00000000000000/* Generated automatically. DO NOT EDIT! */ #define SIMD_HEADER "simd-neon.h" #include "../common/t1bv_15.c" fftw-3.3.3/dft/simd/neon/t1bv_16.c0000644000175000017500000000015612054411107013402 00000000000000/* Generated automatically. DO NOT EDIT! */ #define SIMD_HEADER "simd-neon.h" #include "../common/t1bv_16.c" fftw-3.3.3/dft/simd/neon/t1bv_32.c0000644000175000017500000000015612054411107013400 00000000000000/* Generated automatically. DO NOT EDIT! */ #define SIMD_HEADER "simd-neon.h" #include "../common/t1bv_32.c" fftw-3.3.3/dft/simd/neon/t1bv_64.c0000644000175000017500000000015612054411107013405 00000000000000/* Generated automatically. DO NOT EDIT! */ #define SIMD_HEADER "simd-neon.h" #include "../common/t1bv_64.c" fftw-3.3.3/dft/simd/neon/t1bv_20.c0000644000175000017500000000015612054411107013375 00000000000000/* Generated automatically. DO NOT EDIT! */ #define SIMD_HEADER "simd-neon.h" #include "../common/t1bv_20.c" fftw-3.3.3/dft/simd/neon/t1bv_25.c0000644000175000017500000000015612054411107013402 00000000000000/* Generated automatically. DO NOT EDIT! */ #define SIMD_HEADER "simd-neon.h" #include "../common/t1bv_25.c" fftw-3.3.3/dft/simd/neon/t2bv_2.c0000644000175000017500000000015512054411107013315 00000000000000/* Generated automatically. DO NOT EDIT! */ #define SIMD_HEADER "simd-neon.h" #include "../common/t2bv_2.c" fftw-3.3.3/dft/simd/neon/t2bv_4.c0000644000175000017500000000015512054411107013317 00000000000000/* Generated automatically. DO NOT EDIT! */ #define SIMD_HEADER "simd-neon.h" #include "../common/t2bv_4.c" fftw-3.3.3/dft/simd/neon/t2bv_8.c0000644000175000017500000000015512054411107013323 00000000000000/* Generated automatically. DO NOT EDIT! */ #define SIMD_HEADER "simd-neon.h" #include "../common/t2bv_8.c" fftw-3.3.3/dft/simd/neon/t2bv_16.c0000644000175000017500000000015612054411107013403 00000000000000/* Generated automatically. DO NOT EDIT! */ #define SIMD_HEADER "simd-neon.h" #include "../common/t2bv_16.c" fftw-3.3.3/dft/simd/neon/t2bv_32.c0000644000175000017500000000015612054411107013401 00000000000000/* Generated automatically. DO NOT EDIT! */ #define SIMD_HEADER "simd-neon.h" #include "../common/t2bv_32.c" fftw-3.3.3/dft/simd/neon/t2bv_64.c0000644000175000017500000000015612054411107013406 00000000000000/* Generated automatically. DO NOT EDIT! */ #define SIMD_HEADER "simd-neon.h" #include "../common/t2bv_64.c" fftw-3.3.3/dft/simd/neon/t2bv_5.c0000644000175000017500000000015512054411107013320 00000000000000/* Generated automatically. DO NOT EDIT! */ #define SIMD_HEADER "simd-neon.h" #include "../common/t2bv_5.c" fftw-3.3.3/dft/simd/neon/t2bv_10.c0000644000175000017500000000015612054411107013375 00000000000000/* Generated automatically. DO NOT EDIT! */ #define SIMD_HEADER "simd-neon.h" #include "../common/t2bv_10.c" fftw-3.3.3/dft/simd/neon/t2bv_20.c0000644000175000017500000000015612054411107013376 00000000000000/* Generated automatically. DO NOT EDIT! */ #define SIMD_HEADER "simd-neon.h" #include "../common/t2bv_20.c" fftw-3.3.3/dft/simd/neon/t2bv_25.c0000644000175000017500000000015612054411107013403 00000000000000/* Generated automatically. DO NOT EDIT! */ #define SIMD_HEADER "simd-neon.h" #include "../common/t2bv_25.c" fftw-3.3.3/dft/simd/neon/t3bv_4.c0000644000175000017500000000015512054411107013320 00000000000000/* Generated automatically. DO NOT EDIT! */ #define SIMD_HEADER "simd-neon.h" #include "../common/t3bv_4.c" fftw-3.3.3/dft/simd/neon/t3bv_8.c0000644000175000017500000000015512054411107013324 00000000000000/* Generated automatically. DO NOT EDIT! */ #define SIMD_HEADER "simd-neon.h" #include "../common/t3bv_8.c" fftw-3.3.3/dft/simd/neon/t3bv_16.c0000644000175000017500000000015612054411107013404 00000000000000/* Generated automatically. DO NOT EDIT! */ #define SIMD_HEADER "simd-neon.h" #include "../common/t3bv_16.c" fftw-3.3.3/dft/simd/neon/t3bv_32.c0000644000175000017500000000015612054411107013402 00000000000000/* Generated automatically. DO NOT EDIT! */ #define SIMD_HEADER "simd-neon.h" #include "../common/t3bv_32.c" fftw-3.3.3/dft/simd/neon/t3bv_5.c0000644000175000017500000000015512054411107013321 00000000000000/* Generated automatically. DO NOT EDIT! */ #define SIMD_HEADER "simd-neon.h" #include "../common/t3bv_5.c" fftw-3.3.3/dft/simd/neon/t3bv_10.c0000644000175000017500000000015612054411107013376 00000000000000/* Generated automatically. DO NOT EDIT! */ #define SIMD_HEADER "simd-neon.h" #include "../common/t3bv_10.c" fftw-3.3.3/dft/simd/neon/t3bv_20.c0000644000175000017500000000015612054411107013377 00000000000000/* Generated automatically. DO NOT EDIT! */ #define SIMD_HEADER "simd-neon.h" #include "../common/t3bv_20.c" fftw-3.3.3/dft/simd/neon/t3bv_25.c0000644000175000017500000000015612054411107013404 00000000000000/* Generated automatically. DO NOT EDIT! */ #define SIMD_HEADER "simd-neon.h" #include "../common/t3bv_25.c" fftw-3.3.3/dft/simd/neon/t1sv_2.c0000644000175000017500000000015512054411107013335 00000000000000/* Generated automatically. DO NOT EDIT! */ #define SIMD_HEADER "simd-neon.h" #include "../common/t1sv_2.c" fftw-3.3.3/dft/simd/neon/t1sv_4.c0000644000175000017500000000015512054411107013337 00000000000000/* Generated automatically. DO NOT EDIT! */ #define SIMD_HEADER "simd-neon.h" #include "../common/t1sv_4.c" fftw-3.3.3/dft/simd/neon/t1sv_8.c0000644000175000017500000000015512054411107013343 00000000000000/* Generated automatically. DO NOT EDIT! */ #define SIMD_HEADER "simd-neon.h" #include "../common/t1sv_8.c" fftw-3.3.3/dft/simd/neon/t1sv_16.c0000644000175000017500000000015612054411107013423 00000000000000/* Generated automatically. DO NOT EDIT! */ #define SIMD_HEADER "simd-neon.h" #include "../common/t1sv_16.c" fftw-3.3.3/dft/simd/neon/t1sv_32.c0000644000175000017500000000015612054411107013421 00000000000000/* Generated automatically. DO NOT EDIT! */ #define SIMD_HEADER "simd-neon.h" #include "../common/t1sv_32.c" fftw-3.3.3/dft/simd/neon/t2sv_4.c0000644000175000017500000000015512054411107013340 00000000000000/* Generated automatically. DO NOT EDIT! */ #define SIMD_HEADER "simd-neon.h" #include "../common/t2sv_4.c" fftw-3.3.3/dft/simd/neon/t2sv_8.c0000644000175000017500000000015512054411107013344 00000000000000/* Generated automatically. DO NOT EDIT! */ #define SIMD_HEADER "simd-neon.h" #include "../common/t2sv_8.c" fftw-3.3.3/dft/simd/neon/t2sv_16.c0000644000175000017500000000015612054411107013424 00000000000000/* Generated automatically. DO NOT EDIT! */ #define SIMD_HEADER "simd-neon.h" #include "../common/t2sv_16.c" fftw-3.3.3/dft/simd/neon/t2sv_32.c0000644000175000017500000000015612054411107013422 00000000000000/* Generated automatically. DO NOT EDIT! */ #define SIMD_HEADER "simd-neon.h" #include "../common/t2sv_32.c" fftw-3.3.3/dft/simd/neon/q1fv_2.c0000644000175000017500000000015512054411107013315 00000000000000/* Generated automatically. DO NOT EDIT! */ #define SIMD_HEADER "simd-neon.h" #include "../common/q1fv_2.c" fftw-3.3.3/dft/simd/neon/q1fv_4.c0000644000175000017500000000015512054411107013317 00000000000000/* Generated automatically. DO NOT EDIT! */ #define SIMD_HEADER "simd-neon.h" #include "../common/q1fv_4.c" fftw-3.3.3/dft/simd/neon/q1fv_5.c0000644000175000017500000000015512054411107013320 00000000000000/* Generated automatically. DO NOT EDIT! */ #define SIMD_HEADER "simd-neon.h" #include "../common/q1fv_5.c" fftw-3.3.3/dft/simd/neon/q1fv_8.c0000644000175000017500000000015512054411107013323 00000000000000/* Generated automatically. DO NOT EDIT! */ #define SIMD_HEADER "simd-neon.h" #include "../common/q1fv_8.c" fftw-3.3.3/dft/simd/neon/q1bv_2.c0000644000175000017500000000015512054411107013311 00000000000000/* Generated automatically. DO NOT EDIT! */ #define SIMD_HEADER "simd-neon.h" #include "../common/q1bv_2.c" fftw-3.3.3/dft/simd/neon/q1bv_4.c0000644000175000017500000000015512054411107013313 00000000000000/* Generated automatically. DO NOT EDIT! */ #define SIMD_HEADER "simd-neon.h" #include "../common/q1bv_4.c" fftw-3.3.3/dft/simd/neon/q1bv_5.c0000644000175000017500000000015512054411107013314 00000000000000/* Generated automatically. DO NOT EDIT! */ #define SIMD_HEADER "simd-neon.h" #include "../common/q1bv_5.c" fftw-3.3.3/dft/simd/neon/q1bv_8.c0000644000175000017500000000015512054411107013317 00000000000000/* Generated automatically. DO NOT EDIT! */ #define SIMD_HEADER "simd-neon.h" #include "../common/q1bv_8.c" fftw-3.3.3/dft/simd/neon/genus.c0000644000175000017500000000015412054411107013337 00000000000000/* Generated automatically. DO NOT EDIT! */ #define SIMD_HEADER "simd-neon.h" #include "../common/genus.c" fftw-3.3.3/dft/simd/neon/codlist.c0000644000175000017500000000015612054411107013661 00000000000000/* Generated automatically. DO NOT EDIT! */ #define SIMD_HEADER "simd-neon.h" #include "../common/codlist.c" fftw-3.3.3/rdft/0000755000175000017500000000000012054411110010413 500000000000000fftw-3.3.3/rdft/Makefile.am0000644000175000017500000000141512054410050012372 00000000000000AM_CPPFLAGS = -I$(top_srcdir)/kernel -I$(top_srcdir)/dft SUBDIRS = scalar simd noinst_LTLIBRARIES = librdft.la # pkgincludedir = $(includedir)/fftw3@PREC_SUFFIX@ # pkginclude_HEADERS = codelet-rdft.h rdft.h RDFT2 = buffered2.c direct2.c nop2.c rank0-rdft2.c rank-geq2-rdft2.c \ plan2.c problem2.c solve2.c vrank-geq1-rdft2.c rdft2-rdft.c \ rdft2-tensor-max-index.c rdft2-inplace-strides.c rdft2-strides.c \ khc2c.c ct-hc2c.h ct-hc2c.c ct-hc2c-direct.c librdft_la_SOURCES = hc2hc.h hc2hc.c dft-r2hc.c dht-r2hc.c dht-rader.c \ buffered.c codelet-rdft.h conf.c direct-r2r.c direct-r2c.c generic.c \ hc2hc-direct.c hc2hc-generic.c khc2hc.c kr2c.c kr2r.c indirect.c nop.c \ plan.c problem.c rank0.c rank-geq2.c rdft.h rdft-dht.c solve.c \ vrank-geq1.c vrank3-transpose.c $(RDFT2) fftw-3.3.3/rdft/Makefile.in0000644000175000017500000006172412054410200012411 00000000000000# Makefile.in generated by automake 1.11.6 from Makefile.am. # @configure_input@ # Copyright (C) 1994, 1995, 1996, 1997, 1998, 1999, 2000, 2001, 2002, # 2003, 2004, 2005, 2006, 2007, 2008, 2009, 2010, 2011 Free Software # Foundation, Inc. # This Makefile.in is free software; the Free Software Foundation # gives unlimited permission to copy and/or distribute it, # with or without modifications, as long as this notice is preserved. # This program is distributed in the hope that it will be useful, # but WITHOUT ANY WARRANTY, to the extent permitted by law; without # even the implied warranty of MERCHANTABILITY or FITNESS FOR A # PARTICULAR PURPOSE. @SET_MAKE@ VPATH = @srcdir@ am__make_dryrun = \ { \ am__dry=no; \ case $$MAKEFLAGS in \ *\\[\ \ ]*) \ echo 'am--echo: ; @echo "AM" OK' | $(MAKE) -f - 2>/dev/null \ | grep '^AM OK$$' >/dev/null || am__dry=yes;; \ *) \ for am__flg in $$MAKEFLAGS; do \ case $$am__flg in \ *=*|--*) ;; \ *n*) am__dry=yes; break;; \ esac; \ done;; \ esac; \ test $$am__dry = yes; \ } pkgdatadir = $(datadir)/@PACKAGE@ pkgincludedir = $(includedir)/@PACKAGE@ pkglibdir = $(libdir)/@PACKAGE@ pkglibexecdir = $(libexecdir)/@PACKAGE@ am__cd = CDPATH="$${ZSH_VERSION+.}$(PATH_SEPARATOR)" && cd install_sh_DATA = $(install_sh) -c -m 644 install_sh_PROGRAM = $(install_sh) -c install_sh_SCRIPT = $(install_sh) -c INSTALL_HEADER = $(INSTALL_DATA) transform = $(program_transform_name) NORMAL_INSTALL = : PRE_INSTALL = : POST_INSTALL = : NORMAL_UNINSTALL = : PRE_UNINSTALL = : POST_UNINSTALL = : build_triplet = @build@ host_triplet = @host@ subdir = rdft DIST_COMMON = $(srcdir)/Makefile.am $(srcdir)/Makefile.in ACLOCAL_M4 = $(top_srcdir)/aclocal.m4 am__aclocal_m4_deps = $(top_srcdir)/m4/acx_mpi.m4 \ $(top_srcdir)/m4/acx_pthread.m4 \ $(top_srcdir)/m4/ax_cc_maxopt.m4 \ $(top_srcdir)/m4/ax_check_compiler_flags.m4 \ $(top_srcdir)/m4/ax_compiler_vendor.m4 \ $(top_srcdir)/m4/ax_gcc_aligns_stack.m4 \ $(top_srcdir)/m4/ax_gcc_version.m4 \ $(top_srcdir)/m4/ax_openmp.m4 $(top_srcdir)/m4/libtool.m4 \ $(top_srcdir)/m4/ltoptions.m4 $(top_srcdir)/m4/ltsugar.m4 \ $(top_srcdir)/m4/ltversion.m4 $(top_srcdir)/m4/lt~obsolete.m4 \ $(top_srcdir)/configure.ac am__configure_deps = $(am__aclocal_m4_deps) $(CONFIGURE_DEPENDENCIES) \ $(ACLOCAL_M4) mkinstalldirs = $(install_sh) -d CONFIG_HEADER = $(top_builddir)/config.h CONFIG_CLEAN_FILES = CONFIG_CLEAN_VPATH_FILES = LTLIBRARIES = $(noinst_LTLIBRARIES) librdft_la_LIBADD = am__objects_1 = buffered2.lo direct2.lo nop2.lo rank0-rdft2.lo \ rank-geq2-rdft2.lo plan2.lo problem2.lo solve2.lo \ vrank-geq1-rdft2.lo rdft2-rdft.lo rdft2-tensor-max-index.lo \ rdft2-inplace-strides.lo rdft2-strides.lo khc2c.lo ct-hc2c.lo \ ct-hc2c-direct.lo am_librdft_la_OBJECTS = hc2hc.lo dft-r2hc.lo dht-r2hc.lo dht-rader.lo \ buffered.lo conf.lo direct-r2r.lo direct-r2c.lo generic.lo \ hc2hc-direct.lo hc2hc-generic.lo khc2hc.lo kr2c.lo kr2r.lo \ indirect.lo nop.lo plan.lo problem.lo rank0.lo rank-geq2.lo \ rdft-dht.lo solve.lo vrank-geq1.lo vrank3-transpose.lo \ $(am__objects_1) librdft_la_OBJECTS = $(am_librdft_la_OBJECTS) DEFAULT_INCLUDES = -I.@am__isrc@ -I$(top_builddir) depcomp = $(SHELL) $(top_srcdir)/depcomp am__depfiles_maybe = depfiles am__mv = mv -f COMPILE = $(CC) $(DEFS) $(DEFAULT_INCLUDES) $(INCLUDES) $(AM_CPPFLAGS) \ $(CPPFLAGS) $(AM_CFLAGS) $(CFLAGS) LTCOMPILE = $(LIBTOOL) --tag=CC $(AM_LIBTOOLFLAGS) $(LIBTOOLFLAGS) \ --mode=compile $(CC) $(DEFS) $(DEFAULT_INCLUDES) $(INCLUDES) \ $(AM_CPPFLAGS) $(CPPFLAGS) $(AM_CFLAGS) $(CFLAGS) CCLD = $(CC) LINK = $(LIBTOOL) --tag=CC $(AM_LIBTOOLFLAGS) $(LIBTOOLFLAGS) \ --mode=link $(CCLD) $(AM_CFLAGS) $(CFLAGS) $(AM_LDFLAGS) \ $(LDFLAGS) -o $@ SOURCES = $(librdft_la_SOURCES) DIST_SOURCES = $(librdft_la_SOURCES) RECURSIVE_TARGETS = all-recursive check-recursive dvi-recursive \ html-recursive info-recursive install-data-recursive \ install-dvi-recursive install-exec-recursive \ install-html-recursive install-info-recursive \ install-pdf-recursive install-ps-recursive install-recursive \ installcheck-recursive installdirs-recursive pdf-recursive \ ps-recursive uninstall-recursive am__can_run_installinfo = \ case $$AM_UPDATE_INFO_DIR in \ n|no|NO) false;; \ *) (install-info --version) >/dev/null 2>&1;; \ esac RECURSIVE_CLEAN_TARGETS = mostlyclean-recursive clean-recursive \ distclean-recursive maintainer-clean-recursive AM_RECURSIVE_TARGETS = $(RECURSIVE_TARGETS:-recursive=) \ $(RECURSIVE_CLEAN_TARGETS:-recursive=) tags TAGS ctags CTAGS \ distdir ETAGS = etags CTAGS = ctags DIST_SUBDIRS = $(SUBDIRS) DISTFILES = $(DIST_COMMON) $(DIST_SOURCES) $(TEXINFOS) $(EXTRA_DIST) am__relativize = \ dir0=`pwd`; \ sed_first='s,^\([^/]*\)/.*$$,\1,'; \ sed_rest='s,^[^/]*/*,,'; \ sed_last='s,^.*/\([^/]*\)$$,\1,'; \ sed_butlast='s,/*[^/]*$$,,'; \ while test -n "$$dir1"; do \ first=`echo "$$dir1" | sed -e "$$sed_first"`; \ if test "$$first" != "."; then \ if test "$$first" = ".."; then \ dir2=`echo "$$dir0" | sed -e "$$sed_last"`/"$$dir2"; \ dir0=`echo "$$dir0" | sed -e "$$sed_butlast"`; \ else \ first2=`echo "$$dir2" | sed -e "$$sed_first"`; \ if test "$$first2" = "$$first"; then \ dir2=`echo "$$dir2" | sed -e "$$sed_rest"`; \ else \ dir2="../$$dir2"; \ fi; \ dir0="$$dir0"/"$$first"; \ fi; \ fi; \ dir1=`echo "$$dir1" | sed -e "$$sed_rest"`; \ done; \ reldir="$$dir2" ACLOCAL = @ACLOCAL@ ALLOCA = @ALLOCA@ ALTIVEC_CFLAGS = @ALTIVEC_CFLAGS@ AMTAR = @AMTAR@ AR = @AR@ AS = @AS@ AUTOCONF = @AUTOCONF@ AUTOHEADER = @AUTOHEADER@ AUTOMAKE = @AUTOMAKE@ AVX_CFLAGS = @AVX_CFLAGS@ AWK = @AWK@ CC = @CC@ CCDEPMODE = @CCDEPMODE@ CFLAGS = @CFLAGS@ CHECK_PL_OPTS = @CHECK_PL_OPTS@ CPP = @CPP@ CPPFLAGS = @CPPFLAGS@ CYGPATH_W = @CYGPATH_W@ C_FFTW_R2R_KIND = @C_FFTW_R2R_KIND@ C_MPI_FINT = @C_MPI_FINT@ DEFS = @DEFS@ DEPDIR = @DEPDIR@ DLLTOOL = @DLLTOOL@ DSYMUTIL = @DSYMUTIL@ DUMPBIN = @DUMPBIN@ ECHO_C = @ECHO_C@ ECHO_N = @ECHO_N@ ECHO_T = @ECHO_T@ EGREP = @EGREP@ EXEEXT = @EXEEXT@ F77 = @F77@ FFLAGS = @FFLAGS@ FGREP = @FGREP@ FLIBS = @FLIBS@ GREP = @GREP@ INSTALL = @INSTALL@ INSTALL_DATA = @INSTALL_DATA@ INSTALL_PROGRAM = @INSTALL_PROGRAM@ INSTALL_SCRIPT = @INSTALL_SCRIPT@ INSTALL_STRIP_PROGRAM = @INSTALL_STRIP_PROGRAM@ LD = @LD@ LDFLAGS = @LDFLAGS@ LIBOBJS = @LIBOBJS@ LIBQUADMATH = @LIBQUADMATH@ LIBS = @LIBS@ LIBTOOL = @LIBTOOL@ LIPO = @LIPO@ LN_S = @LN_S@ LTLIBOBJS = @LTLIBOBJS@ MAINT = @MAINT@ MAKEINFO = @MAKEINFO@ MANIFEST_TOOL = @MANIFEST_TOOL@ MKDIR_P = @MKDIR_P@ MPICC = @MPICC@ MPILIBS = @MPILIBS@ MPIRUN = @MPIRUN@ NEON_CFLAGS = @NEON_CFLAGS@ NM = @NM@ NMEDIT = @NMEDIT@ OBJDUMP = @OBJDUMP@ OBJEXT = @OBJEXT@ OCAMLBUILD = @OCAMLBUILD@ OPENMP_CFLAGS = @OPENMP_CFLAGS@ OTOOL = @OTOOL@ OTOOL64 = @OTOOL64@ PACKAGE = @PACKAGE@ PACKAGE_BUGREPORT = @PACKAGE_BUGREPORT@ PACKAGE_NAME = @PACKAGE_NAME@ PACKAGE_STRING = @PACKAGE_STRING@ PACKAGE_TARNAME = @PACKAGE_TARNAME@ PACKAGE_URL = @PACKAGE_URL@ PACKAGE_VERSION = @PACKAGE_VERSION@ PATH_SEPARATOR = @PATH_SEPARATOR@ POW_LIB = @POW_LIB@ PRECISION = @PRECISION@ PREC_SUFFIX = @PREC_SUFFIX@ PTHREAD_CC = @PTHREAD_CC@ PTHREAD_CFLAGS = @PTHREAD_CFLAGS@ PTHREAD_LIBS = @PTHREAD_LIBS@ RANLIB = @RANLIB@ SED = @SED@ SET_MAKE = @SET_MAKE@ SHARED_VERSION_INFO = @SHARED_VERSION_INFO@ SHELL = @SHELL@ SSE2_CFLAGS = @SSE2_CFLAGS@ STACK_ALIGN_CFLAGS = @STACK_ALIGN_CFLAGS@ STRIP = @STRIP@ THREADLIBS = @THREADLIBS@ VERSION = @VERSION@ abs_builddir = @abs_builddir@ abs_srcdir = @abs_srcdir@ abs_top_builddir = @abs_top_builddir@ abs_top_srcdir = @abs_top_srcdir@ ac_ct_AR = @ac_ct_AR@ ac_ct_CC = @ac_ct_CC@ ac_ct_DUMPBIN = @ac_ct_DUMPBIN@ ac_ct_F77 = @ac_ct_F77@ acx_pthread_config = @acx_pthread_config@ am__include = @am__include@ am__leading_dot = @am__leading_dot@ am__quote = @am__quote@ am__tar = @am__tar@ am__untar = @am__untar@ bindir = @bindir@ build = @build@ build_alias = @build_alias@ build_cpu = @build_cpu@ build_os = @build_os@ build_vendor = @build_vendor@ builddir = @builddir@ datadir = @datadir@ datarootdir = @datarootdir@ docdir = @docdir@ dvidir = @dvidir@ exec_prefix = @exec_prefix@ host = @host@ host_alias = @host_alias@ host_cpu = @host_cpu@ host_os = @host_os@ host_vendor = @host_vendor@ htmldir = @htmldir@ includedir = @includedir@ infodir = @infodir@ install_sh = @install_sh@ libdir = @libdir@ libexecdir = @libexecdir@ localedir = @localedir@ localstatedir = @localstatedir@ mandir = @mandir@ mkdir_p = @mkdir_p@ oldincludedir = @oldincludedir@ pdfdir = @pdfdir@ prefix = @prefix@ program_transform_name = @program_transform_name@ psdir = @psdir@ sbindir = @sbindir@ sharedstatedir = @sharedstatedir@ srcdir = @srcdir@ sysconfdir = @sysconfdir@ target_alias = @target_alias@ top_build_prefix = @top_build_prefix@ top_builddir = @top_builddir@ top_srcdir = @top_srcdir@ AM_CPPFLAGS = -I$(top_srcdir)/kernel -I$(top_srcdir)/dft SUBDIRS = scalar simd noinst_LTLIBRARIES = librdft.la # pkgincludedir = $(includedir)/fftw3@PREC_SUFFIX@ # pkginclude_HEADERS = codelet-rdft.h rdft.h RDFT2 = buffered2.c direct2.c nop2.c rank0-rdft2.c rank-geq2-rdft2.c \ plan2.c problem2.c solve2.c vrank-geq1-rdft2.c rdft2-rdft.c \ rdft2-tensor-max-index.c rdft2-inplace-strides.c rdft2-strides.c \ khc2c.c ct-hc2c.h ct-hc2c.c ct-hc2c-direct.c librdft_la_SOURCES = hc2hc.h hc2hc.c dft-r2hc.c dht-r2hc.c dht-rader.c \ buffered.c codelet-rdft.h conf.c direct-r2r.c direct-r2c.c generic.c \ hc2hc-direct.c hc2hc-generic.c khc2hc.c kr2c.c kr2r.c indirect.c nop.c \ plan.c problem.c rank0.c rank-geq2.c rdft.h rdft-dht.c solve.c \ vrank-geq1.c vrank3-transpose.c $(RDFT2) all: all-recursive .SUFFIXES: .SUFFIXES: .c .lo .o .obj $(srcdir)/Makefile.in: @MAINTAINER_MODE_TRUE@ $(srcdir)/Makefile.am $(am__configure_deps) @for dep in $?; do \ case '$(am__configure_deps)' in \ *$$dep*) \ ( cd $(top_builddir) && $(MAKE) $(AM_MAKEFLAGS) am--refresh ) \ && { if test -f $@; then exit 0; else break; fi; }; \ exit 1;; \ esac; \ done; \ echo ' cd $(top_srcdir) && $(AUTOMAKE) --gnu rdft/Makefile'; \ $(am__cd) $(top_srcdir) && \ $(AUTOMAKE) --gnu rdft/Makefile .PRECIOUS: Makefile Makefile: $(srcdir)/Makefile.in $(top_builddir)/config.status @case '$?' in \ *config.status*) \ cd $(top_builddir) && $(MAKE) $(AM_MAKEFLAGS) am--refresh;; \ *) \ echo ' cd $(top_builddir) && $(SHELL) ./config.status $(subdir)/$@ $(am__depfiles_maybe)'; \ cd $(top_builddir) && $(SHELL) ./config.status $(subdir)/$@ $(am__depfiles_maybe);; \ esac; $(top_builddir)/config.status: $(top_srcdir)/configure $(CONFIG_STATUS_DEPENDENCIES) cd $(top_builddir) && $(MAKE) $(AM_MAKEFLAGS) am--refresh $(top_srcdir)/configure: @MAINTAINER_MODE_TRUE@ $(am__configure_deps) cd $(top_builddir) && $(MAKE) $(AM_MAKEFLAGS) am--refresh $(ACLOCAL_M4): @MAINTAINER_MODE_TRUE@ $(am__aclocal_m4_deps) cd $(top_builddir) && $(MAKE) $(AM_MAKEFLAGS) am--refresh $(am__aclocal_m4_deps): clean-noinstLTLIBRARIES: -test -z "$(noinst_LTLIBRARIES)" || rm -f $(noinst_LTLIBRARIES) @list='$(noinst_LTLIBRARIES)'; for p in $$list; do \ dir="`echo $$p | sed -e 's|/[^/]*$$||'`"; \ test "$$dir" != "$$p" || dir=.; \ echo "rm -f \"$${dir}/so_locations\""; \ rm -f "$${dir}/so_locations"; \ done librdft.la: $(librdft_la_OBJECTS) $(librdft_la_DEPENDENCIES) $(EXTRA_librdft_la_DEPENDENCIES) $(LINK) $(librdft_la_OBJECTS) $(librdft_la_LIBADD) $(LIBS) mostlyclean-compile: -rm -f *.$(OBJEXT) distclean-compile: -rm -f *.tab.c @AMDEP_TRUE@@am__include@ @am__quote@./$(DEPDIR)/buffered.Plo@am__quote@ @AMDEP_TRUE@@am__include@ @am__quote@./$(DEPDIR)/buffered2.Plo@am__quote@ @AMDEP_TRUE@@am__include@ @am__quote@./$(DEPDIR)/conf.Plo@am__quote@ @AMDEP_TRUE@@am__include@ @am__quote@./$(DEPDIR)/ct-hc2c-direct.Plo@am__quote@ @AMDEP_TRUE@@am__include@ @am__quote@./$(DEPDIR)/ct-hc2c.Plo@am__quote@ @AMDEP_TRUE@@am__include@ @am__quote@./$(DEPDIR)/dft-r2hc.Plo@am__quote@ @AMDEP_TRUE@@am__include@ @am__quote@./$(DEPDIR)/dht-r2hc.Plo@am__quote@ @AMDEP_TRUE@@am__include@ @am__quote@./$(DEPDIR)/dht-rader.Plo@am__quote@ @AMDEP_TRUE@@am__include@ @am__quote@./$(DEPDIR)/direct-r2c.Plo@am__quote@ @AMDEP_TRUE@@am__include@ @am__quote@./$(DEPDIR)/direct-r2r.Plo@am__quote@ @AMDEP_TRUE@@am__include@ @am__quote@./$(DEPDIR)/direct2.Plo@am__quote@ @AMDEP_TRUE@@am__include@ @am__quote@./$(DEPDIR)/generic.Plo@am__quote@ @AMDEP_TRUE@@am__include@ @am__quote@./$(DEPDIR)/hc2hc-direct.Plo@am__quote@ @AMDEP_TRUE@@am__include@ @am__quote@./$(DEPDIR)/hc2hc-generic.Plo@am__quote@ @AMDEP_TRUE@@am__include@ @am__quote@./$(DEPDIR)/hc2hc.Plo@am__quote@ @AMDEP_TRUE@@am__include@ @am__quote@./$(DEPDIR)/indirect.Plo@am__quote@ @AMDEP_TRUE@@am__include@ @am__quote@./$(DEPDIR)/khc2c.Plo@am__quote@ @AMDEP_TRUE@@am__include@ @am__quote@./$(DEPDIR)/khc2hc.Plo@am__quote@ @AMDEP_TRUE@@am__include@ @am__quote@./$(DEPDIR)/kr2c.Plo@am__quote@ @AMDEP_TRUE@@am__include@ @am__quote@./$(DEPDIR)/kr2r.Plo@am__quote@ @AMDEP_TRUE@@am__include@ @am__quote@./$(DEPDIR)/nop.Plo@am__quote@ @AMDEP_TRUE@@am__include@ @am__quote@./$(DEPDIR)/nop2.Plo@am__quote@ @AMDEP_TRUE@@am__include@ @am__quote@./$(DEPDIR)/plan.Plo@am__quote@ @AMDEP_TRUE@@am__include@ @am__quote@./$(DEPDIR)/plan2.Plo@am__quote@ @AMDEP_TRUE@@am__include@ @am__quote@./$(DEPDIR)/problem.Plo@am__quote@ @AMDEP_TRUE@@am__include@ @am__quote@./$(DEPDIR)/problem2.Plo@am__quote@ @AMDEP_TRUE@@am__include@ @am__quote@./$(DEPDIR)/rank-geq2-rdft2.Plo@am__quote@ @AMDEP_TRUE@@am__include@ @am__quote@./$(DEPDIR)/rank-geq2.Plo@am__quote@ @AMDEP_TRUE@@am__include@ @am__quote@./$(DEPDIR)/rank0-rdft2.Plo@am__quote@ @AMDEP_TRUE@@am__include@ @am__quote@./$(DEPDIR)/rank0.Plo@am__quote@ @AMDEP_TRUE@@am__include@ @am__quote@./$(DEPDIR)/rdft-dht.Plo@am__quote@ @AMDEP_TRUE@@am__include@ @am__quote@./$(DEPDIR)/rdft2-inplace-strides.Plo@am__quote@ @AMDEP_TRUE@@am__include@ @am__quote@./$(DEPDIR)/rdft2-rdft.Plo@am__quote@ @AMDEP_TRUE@@am__include@ @am__quote@./$(DEPDIR)/rdft2-strides.Plo@am__quote@ @AMDEP_TRUE@@am__include@ @am__quote@./$(DEPDIR)/rdft2-tensor-max-index.Plo@am__quote@ @AMDEP_TRUE@@am__include@ @am__quote@./$(DEPDIR)/solve.Plo@am__quote@ @AMDEP_TRUE@@am__include@ @am__quote@./$(DEPDIR)/solve2.Plo@am__quote@ @AMDEP_TRUE@@am__include@ @am__quote@./$(DEPDIR)/vrank-geq1-rdft2.Plo@am__quote@ @AMDEP_TRUE@@am__include@ @am__quote@./$(DEPDIR)/vrank-geq1.Plo@am__quote@ @AMDEP_TRUE@@am__include@ @am__quote@./$(DEPDIR)/vrank3-transpose.Plo@am__quote@ .c.o: @am__fastdepCC_TRUE@ $(COMPILE) -MT $@ -MD -MP -MF $(DEPDIR)/$*.Tpo -c -o $@ $< @am__fastdepCC_TRUE@ $(am__mv) $(DEPDIR)/$*.Tpo $(DEPDIR)/$*.Po @AMDEP_TRUE@@am__fastdepCC_FALSE@ source='$<' object='$@' libtool=no @AMDEPBACKSLASH@ @AMDEP_TRUE@@am__fastdepCC_FALSE@ DEPDIR=$(DEPDIR) $(CCDEPMODE) $(depcomp) @AMDEPBACKSLASH@ @am__fastdepCC_FALSE@ $(COMPILE) -c $< .c.obj: @am__fastdepCC_TRUE@ $(COMPILE) -MT $@ -MD -MP -MF $(DEPDIR)/$*.Tpo -c -o $@ `$(CYGPATH_W) '$<'` @am__fastdepCC_TRUE@ $(am__mv) $(DEPDIR)/$*.Tpo $(DEPDIR)/$*.Po @AMDEP_TRUE@@am__fastdepCC_FALSE@ source='$<' object='$@' libtool=no @AMDEPBACKSLASH@ @AMDEP_TRUE@@am__fastdepCC_FALSE@ DEPDIR=$(DEPDIR) $(CCDEPMODE) $(depcomp) @AMDEPBACKSLASH@ @am__fastdepCC_FALSE@ $(COMPILE) -c `$(CYGPATH_W) '$<'` .c.lo: @am__fastdepCC_TRUE@ $(LTCOMPILE) -MT $@ -MD -MP -MF $(DEPDIR)/$*.Tpo -c -o $@ $< @am__fastdepCC_TRUE@ $(am__mv) $(DEPDIR)/$*.Tpo $(DEPDIR)/$*.Plo @AMDEP_TRUE@@am__fastdepCC_FALSE@ source='$<' object='$@' libtool=yes @AMDEPBACKSLASH@ @AMDEP_TRUE@@am__fastdepCC_FALSE@ DEPDIR=$(DEPDIR) $(CCDEPMODE) $(depcomp) @AMDEPBACKSLASH@ @am__fastdepCC_FALSE@ $(LTCOMPILE) -c -o $@ $< mostlyclean-libtool: -rm -f *.lo clean-libtool: -rm -rf .libs _libs # This directory's subdirectories are mostly independent; you can cd # into them and run `make' without going through this Makefile. # To change the values of `make' variables: instead of editing Makefiles, # (1) if the variable is set in `config.status', edit `config.status' # (which will cause the Makefiles to be regenerated when you run `make'); # (2) otherwise, pass the desired values on the `make' command line. $(RECURSIVE_TARGETS): @fail= failcom='exit 1'; \ for f in x $$MAKEFLAGS; do \ case $$f in \ *=* | --[!k]*);; \ *k*) failcom='fail=yes';; \ esac; \ done; \ dot_seen=no; \ target=`echo $@ | sed s/-recursive//`; \ list='$(SUBDIRS)'; for subdir in $$list; do \ echo "Making $$target in $$subdir"; \ if test "$$subdir" = "."; then \ dot_seen=yes; \ local_target="$$target-am"; \ else \ local_target="$$target"; \ fi; \ ($(am__cd) $$subdir && $(MAKE) $(AM_MAKEFLAGS) $$local_target) \ || eval $$failcom; \ done; \ if test "$$dot_seen" = "no"; then \ $(MAKE) $(AM_MAKEFLAGS) "$$target-am" || exit 1; \ fi; test -z "$$fail" $(RECURSIVE_CLEAN_TARGETS): @fail= failcom='exit 1'; \ for f in x $$MAKEFLAGS; do \ case $$f in \ *=* | --[!k]*);; \ *k*) failcom='fail=yes';; \ esac; \ done; \ dot_seen=no; \ case "$@" in \ distclean-* | maintainer-clean-*) list='$(DIST_SUBDIRS)' ;; \ *) list='$(SUBDIRS)' ;; \ esac; \ rev=''; for subdir in $$list; do \ if test "$$subdir" = "."; then :; else \ rev="$$subdir $$rev"; \ fi; \ done; \ rev="$$rev ."; \ target=`echo $@ | sed s/-recursive//`; \ for subdir in $$rev; do \ echo "Making $$target in $$subdir"; \ if test "$$subdir" = "."; then \ local_target="$$target-am"; \ else \ local_target="$$target"; \ fi; \ ($(am__cd) $$subdir && $(MAKE) $(AM_MAKEFLAGS) $$local_target) \ || eval $$failcom; \ done && test -z "$$fail" tags-recursive: list='$(SUBDIRS)'; for subdir in $$list; do \ test "$$subdir" = . || ($(am__cd) $$subdir && $(MAKE) $(AM_MAKEFLAGS) tags); \ done ctags-recursive: list='$(SUBDIRS)'; for subdir in $$list; do \ test "$$subdir" = . || ($(am__cd) $$subdir && $(MAKE) $(AM_MAKEFLAGS) ctags); \ done ID: $(HEADERS) $(SOURCES) $(LISP) $(TAGS_FILES) list='$(SOURCES) $(HEADERS) $(LISP) $(TAGS_FILES)'; \ unique=`for i in $$list; do \ if test -f "$$i"; then echo $$i; else echo $(srcdir)/$$i; fi; \ done | \ $(AWK) '{ files[$$0] = 1; nonempty = 1; } \ END { if (nonempty) { for (i in files) print i; }; }'`; \ mkid -fID $$unique tags: TAGS TAGS: tags-recursive $(HEADERS) $(SOURCES) $(TAGS_DEPENDENCIES) \ $(TAGS_FILES) $(LISP) set x; \ here=`pwd`; \ if ($(ETAGS) --etags-include --version) >/dev/null 2>&1; then \ include_option=--etags-include; \ empty_fix=.; \ else \ include_option=--include; \ empty_fix=; \ fi; \ list='$(SUBDIRS)'; for subdir in $$list; do \ if test "$$subdir" = .; then :; else \ test ! -f $$subdir/TAGS || \ set "$$@" "$$include_option=$$here/$$subdir/TAGS"; \ fi; \ done; \ list='$(SOURCES) $(HEADERS) $(LISP) $(TAGS_FILES)'; \ unique=`for i in $$list; do \ if test -f "$$i"; then echo $$i; else echo $(srcdir)/$$i; fi; \ done | \ $(AWK) '{ files[$$0] = 1; nonempty = 1; } \ END { if (nonempty) { for (i in files) print i; }; }'`; \ shift; \ if test -z "$(ETAGS_ARGS)$$*$$unique"; then :; else \ test -n "$$unique" || unique=$$empty_fix; \ if test $$# -gt 0; then \ $(ETAGS) $(ETAGSFLAGS) $(AM_ETAGSFLAGS) $(ETAGS_ARGS) \ "$$@" $$unique; \ else \ $(ETAGS) $(ETAGSFLAGS) $(AM_ETAGSFLAGS) $(ETAGS_ARGS) \ $$unique; \ fi; \ fi ctags: CTAGS CTAGS: ctags-recursive $(HEADERS) $(SOURCES) $(TAGS_DEPENDENCIES) \ $(TAGS_FILES) $(LISP) list='$(SOURCES) $(HEADERS) $(LISP) $(TAGS_FILES)'; \ unique=`for i in $$list; do \ if test -f "$$i"; then echo $$i; else echo $(srcdir)/$$i; fi; \ done | \ $(AWK) '{ files[$$0] = 1; nonempty = 1; } \ END { if (nonempty) { for (i in files) print i; }; }'`; \ test -z "$(CTAGS_ARGS)$$unique" \ || $(CTAGS) $(CTAGSFLAGS) $(AM_CTAGSFLAGS) $(CTAGS_ARGS) \ $$unique GTAGS: here=`$(am__cd) $(top_builddir) && pwd` \ && $(am__cd) $(top_srcdir) \ && gtags -i $(GTAGS_ARGS) "$$here" distclean-tags: -rm -f TAGS ID GTAGS GRTAGS GSYMS GPATH tags distdir: $(DISTFILES) @srcdirstrip=`echo "$(srcdir)" | sed 's/[].[^$$\\*]/\\\\&/g'`; \ topsrcdirstrip=`echo "$(top_srcdir)" | sed 's/[].[^$$\\*]/\\\\&/g'`; \ list='$(DISTFILES)'; \ dist_files=`for file in $$list; do echo $$file; done | \ sed -e "s|^$$srcdirstrip/||;t" \ -e "s|^$$topsrcdirstrip/|$(top_builddir)/|;t"`; \ case $$dist_files in \ */*) $(MKDIR_P) `echo "$$dist_files" | \ sed '/\//!d;s|^|$(distdir)/|;s,/[^/]*$$,,' | \ sort -u` ;; \ esac; \ for file in $$dist_files; do \ if test -f $$file || test -d $$file; then d=.; else d=$(srcdir); fi; \ if test -d $$d/$$file; then \ dir=`echo "/$$file" | sed -e 's,/[^/]*$$,,'`; \ if test -d "$(distdir)/$$file"; then \ find "$(distdir)/$$file" -type d ! -perm -700 -exec chmod u+rwx {} \;; \ fi; \ if test -d $(srcdir)/$$file && test $$d != $(srcdir); then \ cp -fpR $(srcdir)/$$file "$(distdir)$$dir" || exit 1; \ find "$(distdir)/$$file" -type d ! -perm -700 -exec chmod u+rwx {} \;; \ fi; \ cp -fpR $$d/$$file "$(distdir)$$dir" || exit 1; \ else \ test -f "$(distdir)/$$file" \ || cp -p $$d/$$file "$(distdir)/$$file" \ || exit 1; \ fi; \ done @list='$(DIST_SUBDIRS)'; for subdir in $$list; do \ if test "$$subdir" = .; then :; else \ $(am__make_dryrun) \ || test -d "$(distdir)/$$subdir" \ || $(MKDIR_P) "$(distdir)/$$subdir" \ || exit 1; \ dir1=$$subdir; dir2="$(distdir)/$$subdir"; \ $(am__relativize); \ new_distdir=$$reldir; \ dir1=$$subdir; dir2="$(top_distdir)"; \ $(am__relativize); \ new_top_distdir=$$reldir; \ echo " (cd $$subdir && $(MAKE) $(AM_MAKEFLAGS) top_distdir="$$new_top_distdir" distdir="$$new_distdir" \\"; \ echo " am__remove_distdir=: am__skip_length_check=: am__skip_mode_fix=: distdir)"; \ ($(am__cd) $$subdir && \ $(MAKE) $(AM_MAKEFLAGS) \ top_distdir="$$new_top_distdir" \ distdir="$$new_distdir" \ am__remove_distdir=: \ am__skip_length_check=: \ am__skip_mode_fix=: \ distdir) \ || exit 1; \ fi; \ done check-am: all-am check: check-recursive all-am: Makefile $(LTLIBRARIES) installdirs: installdirs-recursive installdirs-am: install: install-recursive install-exec: install-exec-recursive install-data: install-data-recursive uninstall: uninstall-recursive install-am: all-am @$(MAKE) $(AM_MAKEFLAGS) install-exec-am install-data-am installcheck: installcheck-recursive install-strip: if test -z '$(STRIP)'; then \ $(MAKE) $(AM_MAKEFLAGS) INSTALL_PROGRAM="$(INSTALL_STRIP_PROGRAM)" \ install_sh_PROGRAM="$(INSTALL_STRIP_PROGRAM)" INSTALL_STRIP_FLAG=-s \ install; \ else \ $(MAKE) $(AM_MAKEFLAGS) INSTALL_PROGRAM="$(INSTALL_STRIP_PROGRAM)" \ install_sh_PROGRAM="$(INSTALL_STRIP_PROGRAM)" INSTALL_STRIP_FLAG=-s \ "INSTALL_PROGRAM_ENV=STRIPPROG='$(STRIP)'" install; \ fi mostlyclean-generic: clean-generic: distclean-generic: -test -z "$(CONFIG_CLEAN_FILES)" || rm -f $(CONFIG_CLEAN_FILES) -test . = "$(srcdir)" || test -z "$(CONFIG_CLEAN_VPATH_FILES)" || rm -f $(CONFIG_CLEAN_VPATH_FILES) maintainer-clean-generic: @echo "This command is intended for maintainers to use" @echo "it deletes files that may require special tools to rebuild." clean: clean-recursive clean-am: clean-generic clean-libtool clean-noinstLTLIBRARIES \ mostlyclean-am distclean: distclean-recursive -rm -rf ./$(DEPDIR) -rm -f Makefile distclean-am: clean-am distclean-compile distclean-generic \ distclean-tags dvi: dvi-recursive dvi-am: html: html-recursive html-am: info: info-recursive info-am: install-data-am: install-dvi: install-dvi-recursive install-dvi-am: install-exec-am: install-html: install-html-recursive install-html-am: install-info: install-info-recursive install-info-am: install-man: install-pdf: install-pdf-recursive install-pdf-am: install-ps: install-ps-recursive install-ps-am: installcheck-am: maintainer-clean: maintainer-clean-recursive -rm -rf ./$(DEPDIR) -rm -f Makefile maintainer-clean-am: distclean-am maintainer-clean-generic mostlyclean: mostlyclean-recursive mostlyclean-am: mostlyclean-compile mostlyclean-generic \ mostlyclean-libtool pdf: pdf-recursive pdf-am: ps: ps-recursive ps-am: uninstall-am: .MAKE: $(RECURSIVE_CLEAN_TARGETS) $(RECURSIVE_TARGETS) ctags-recursive \ install-am install-strip tags-recursive .PHONY: $(RECURSIVE_CLEAN_TARGETS) $(RECURSIVE_TARGETS) CTAGS GTAGS \ all all-am check check-am clean clean-generic clean-libtool \ clean-noinstLTLIBRARIES ctags ctags-recursive distclean \ distclean-compile distclean-generic distclean-libtool \ distclean-tags distdir dvi dvi-am html html-am info info-am \ install install-am install-data install-data-am install-dvi \ install-dvi-am install-exec install-exec-am install-html \ install-html-am install-info install-info-am install-man \ install-pdf install-pdf-am install-ps install-ps-am \ install-strip installcheck installcheck-am installdirs \ installdirs-am maintainer-clean maintainer-clean-generic \ mostlyclean mostlyclean-compile mostlyclean-generic \ mostlyclean-libtool pdf pdf-am ps ps-am tags tags-recursive \ uninstall uninstall-am # Tell versions [3.59,3.63) of GNU make to not export all variables. # Otherwise a system limit (for SysV at least) may be exceeded. .NOEXPORT: fftw-3.3.3/rdft/hc2hc.h0000644000175000017500000000343712054410050011504 00000000000000/* * Copyright (c) 2003, 2007-11 Matteo Frigo * Copyright (c) 2003, 2007-11 Massachusetts Institute of Technology * * This program is free software; you can redistribute it and/or modify * it under the terms of the GNU General Public License as published by * the Free Software Foundation; either version 2 of the License, or * (at your option) any later version. * * This program is distributed in the hope that it will be useful, * but WITHOUT ANY WARRANTY; without even the implied warranty of * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the * GNU General Public License for more details. * * You should have received a copy of the GNU General Public License * along with this program; if not, write to the Free Software * Foundation, Inc., 51 Franklin Street, Fifth Floor, Boston, MA 02110-1301 USA * */ #include "rdft.h" typedef void (*hc2hcapply) (const plan *ego, R *IO); typedef struct hc2hc_solver_s hc2hc_solver; typedef plan *(*hc2hc_mkinferior)(const hc2hc_solver *ego, rdft_kind kind, INT r, INT m, INT s, INT vl, INT vs, INT mstart, INT mcount, R *IO, planner *plnr); typedef struct { plan super; hc2hcapply apply; } plan_hc2hc; extern plan *X(mkplan_hc2hc)(size_t size, const plan_adt *adt, hc2hcapply apply); #define MKPLAN_HC2HC(type, adt, apply) \ (type *)X(mkplan_hc2hc)(sizeof(type), adt, apply) struct hc2hc_solver_s { solver super; INT r; hc2hc_mkinferior mkcldw; }; hc2hc_solver *X(mksolver_hc2hc)(size_t size, INT r, hc2hc_mkinferior mkcldw); extern hc2hc_solver *(*X(mksolver_hc2hc_hook))(size_t, INT, hc2hc_mkinferior); void X(regsolver_hc2hc_direct)(planner *plnr, khc2hc codelet, const hc2hc_desc *desc); int X(hc2hc_applicable)(const hc2hc_solver *, const problem *, planner *); fftw-3.3.3/rdft/hc2hc.c0000644000175000017500000001256712054410050011503 00000000000000/* * Copyright (c) 2003, 2007-11 Matteo Frigo * Copyright (c) 2003, 2007-11 Massachusetts Institute of Technology * * This program is free software; you can redistribute it and/or modify * it under the terms of the GNU General Public License as published by * the Free Software Foundation; either version 2 of the License, or * (at your option) any later version. * * This program is distributed in the hope that it will be useful, * but WITHOUT ANY WARRANTY; without even the implied warranty of * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the * GNU General Public License for more details. * * You should have received a copy of the GNU General Public License * along with this program; if not, write to the Free Software * Foundation, Inc., 51 Franklin Street, Fifth Floor, Boston, MA 02110-1301 USA * */ #include "hc2hc.h" hc2hc_solver *(*X(mksolver_hc2hc_hook))(size_t, INT, hc2hc_mkinferior) = 0; typedef struct { plan_rdft super; plan *cld; plan *cldw; INT r; } P; static void apply_dit(const plan *ego_, R *I, R *O) { const P *ego = (const P *) ego_; plan_rdft *cld; plan_hc2hc *cldw; cld = (plan_rdft *) ego->cld; cld->apply(ego->cld, I, O); cldw = (plan_hc2hc *) ego->cldw; cldw->apply(ego->cldw, O); } static void apply_dif(const plan *ego_, R *I, R *O) { const P *ego = (const P *) ego_; plan_rdft *cld; plan_hc2hc *cldw; cldw = (plan_hc2hc *) ego->cldw; cldw->apply(ego->cldw, I); cld = (plan_rdft *) ego->cld; cld->apply(ego->cld, I, O); } static void awake(plan *ego_, enum wakefulness wakefulness) { P *ego = (P *) ego_; X(plan_awake)(ego->cld, wakefulness); X(plan_awake)(ego->cldw, wakefulness); } static void destroy(plan *ego_) { P *ego = (P *) ego_; X(plan_destroy_internal)(ego->cldw); X(plan_destroy_internal)(ego->cld); } static void print(const plan *ego_, printer *p) { const P *ego = (const P *) ego_; p->print(p, "(rdft-ct-%s/%D%(%p%)%(%p%))", ego->super.apply == apply_dit ? "dit" : "dif", ego->r, ego->cldw, ego->cld); } static int applicable0(const hc2hc_solver *ego, const problem *p_, planner *plnr) { const problem_rdft *p = (const problem_rdft *) p_; INT r; return (1 && p->sz->rnk == 1 && p->vecsz->rnk <= 1 && (/* either the problem is R2HC, which is solved by DIT */ (p->kind[0] == R2HC) || /* or the problem is HC2R, in which case it is solved by DIF, which destroys the input */ (p->kind[0] == HC2R && (p->I == p->O || !NO_DESTROY_INPUTP(plnr)))) && ((r = X(choose_radix)(ego->r, p->sz->dims[0].n)) > 0) && p->sz->dims[0].n > r); } int X(hc2hc_applicable)(const hc2hc_solver *ego, const problem *p_, planner *plnr) { const problem_rdft *p; if (!applicable0(ego, p_, plnr)) return 0; p = (const problem_rdft *) p_; return (0 || p->vecsz->rnk == 0 || !NO_VRECURSEP(plnr) ); } static plan *mkplan(const solver *ego_, const problem *p_, planner *plnr) { const hc2hc_solver *ego = (const hc2hc_solver *) ego_; const problem_rdft *p; P *pln = 0; plan *cld = 0, *cldw = 0; INT n, r, m, v, ivs, ovs; iodim *d; static const plan_adt padt = { X(rdft_solve), awake, print, destroy }; if (NO_NONTHREADEDP(plnr) || !X(hc2hc_applicable)(ego, p_, plnr)) return (plan *) 0; p = (const problem_rdft *) p_; d = p->sz->dims; n = d[0].n; r = X(choose_radix)(ego->r, n); m = n / r; X(tensor_tornk1)(p->vecsz, &v, &ivs, &ovs); switch (p->kind[0]) { case R2HC: cldw = ego->mkcldw(ego, R2HC, r, m, d[0].os, v, ovs, 0, (m+2)/2, p->O, plnr); if (!cldw) goto nada; cld = X(mkplan_d)(plnr, X(mkproblem_rdft_d)( X(mktensor_1d)(m, r * d[0].is, d[0].os), X(mktensor_2d)(r, d[0].is, m * d[0].os, v, ivs, ovs), p->I, p->O, p->kind) ); if (!cld) goto nada; pln = MKPLAN_RDFT(P, &padt, apply_dit); break; case HC2R: cldw = ego->mkcldw(ego, HC2R, r, m, d[0].is, v, ivs, 0, (m+2)/2, p->I, plnr); if (!cldw) goto nada; cld = X(mkplan_d)(plnr, X(mkproblem_rdft_d)( X(mktensor_1d)(m, d[0].is, r * d[0].os), X(mktensor_2d)(r, m * d[0].is, d[0].os, v, ivs, ovs), p->I, p->O, p->kind) ); if (!cld) goto nada; pln = MKPLAN_RDFT(P, &padt, apply_dif); break; default: A(0); } pln->cld = cld; pln->cldw = cldw; pln->r = r; X(ops_add)(&cld->ops, &cldw->ops, &pln->super.super.ops); /* inherit could_prune_now_p attribute from cldw */ pln->super.super.could_prune_now_p = cldw->could_prune_now_p; return &(pln->super.super); nada: X(plan_destroy_internal)(cldw); X(plan_destroy_internal)(cld); return (plan *) 0; } hc2hc_solver *X(mksolver_hc2hc)(size_t size, INT r, hc2hc_mkinferior mkcldw) { static const solver_adt sadt = { PROBLEM_RDFT, mkplan, 0 }; hc2hc_solver *slv = (hc2hc_solver *)X(mksolver)(size, &sadt); slv->r = r; slv->mkcldw = mkcldw; return slv; } plan *X(mkplan_hc2hc)(size_t size, const plan_adt *adt, hc2hcapply apply) { plan_hc2hc *ego; ego = (plan_hc2hc *) X(mkplan)(size, adt); ego->apply = apply; return &(ego->super); } fftw-3.3.3/rdft/dft-r2hc.c0000644000175000017500000001166712054410050012125 00000000000000/* * Copyright (c) 2003, 2007-11 Matteo Frigo * Copyright (c) 2003, 2007-11 Massachusetts Institute of Technology * * This program is free software; you can redistribute it and/or modify * it under the terms of the GNU General Public License as published by * the Free Software Foundation; either version 2 of the License, or * (at your option) any later version. * * This program is distributed in the hope that it will be useful, * but WITHOUT ANY WARRANTY; without even the implied warranty of * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the * GNU General Public License for more details. * * You should have received a copy of the GNU General Public License * along with this program; if not, write to the Free Software * Foundation, Inc., 51 Franklin Street, Fifth Floor, Boston, MA 02110-1301 USA * */ /* Compute the complex DFT by combining R2HC RDFTs on the real and imaginary parts. This could be useful for people just wanting to link to the real codelets and not the complex ones. It could also even be faster than the complex algorithms for split (as opposed to interleaved) real/imag complex data. */ #include "rdft.h" #include "dft.h" typedef struct { solver super; } S; typedef struct { plan_dft super; plan *cld; INT ishift, oshift; INT os; INT n; } P; static void apply(const plan *ego_, R *ri, R *ii, R *ro, R *io) { const P *ego = (const P *) ego_; INT n; UNUSED(ii); { /* transform vector of real & imag parts: */ plan_rdft *cld = (plan_rdft *) ego->cld; cld->apply((plan *) cld, ri + ego->ishift, ro + ego->oshift); } n = ego->n; if (n > 1) { INT i, os = ego->os; for (i = 1; i < (n + 1)/2; ++i) { E rop, iop, iom, rom; rop = ro[os * i]; iop = io[os * i]; rom = ro[os * (n - i)]; iom = io[os * (n - i)]; ro[os * i] = rop - iom; io[os * i] = iop + rom; ro[os * (n - i)] = rop + iom; io[os * (n - i)] = iop - rom; } } } static void awake(plan *ego_, enum wakefulness wakefulness) { P *ego = (P *) ego_; X(plan_awake)(ego->cld, wakefulness); } static void destroy(plan *ego_) { P *ego = (P *) ego_; X(plan_destroy_internal)(ego->cld); } static void print(const plan *ego_, printer *p) { const P *ego = (const P *) ego_; p->print(p, "(dft-r2hc-%D%(%p%))", ego->n, ego->cld); } static int applicable0(const problem *p_) { const problem_dft *p = (const problem_dft *) p_; return ((p->sz->rnk == 1 && p->vecsz->rnk == 0) || (p->sz->rnk == 0 && FINITE_RNK(p->vecsz->rnk)) ); } static int splitp(R *r, R *i, INT n, INT s) { return ((r > i ? (r - i) : (i - r)) >= n * (s > 0 ? s : 0-s)); } static int applicable(const problem *p_, const planner *plnr) { if (!applicable0(p_)) return 0; { const problem_dft *p = (const problem_dft *) p_; /* rank-0 problems are always OK */ if (p->sz->rnk == 0) return 1; /* this solver is ok for split arrays */ if (p->sz->rnk == 1 && splitp(p->ri, p->ii, p->sz->dims[0].n, p->sz->dims[0].is) && splitp(p->ro, p->io, p->sz->dims[0].n, p->sz->dims[0].os)) return 1; return !(NO_DFT_R2HCP(plnr)); } } static plan *mkplan(const solver *ego_, const problem *p_, planner *plnr) { P *pln; const problem_dft *p; plan *cld; INT ishift = 0, oshift = 0; static const plan_adt padt = { X(dft_solve), awake, print, destroy }; UNUSED(ego_); if (!applicable(p_, plnr)) return (plan *)0; p = (const problem_dft *) p_; { tensor *ri_vec = X(mktensor_1d)(2, p->ii - p->ri, p->io - p->ro); tensor *cld_vec = X(tensor_append)(ri_vec, p->vecsz); int i; for (i = 0; i < cld_vec->rnk; ++i) { /* make all istrides > 0 */ if (cld_vec->dims[i].is < 0) { INT nm1 = cld_vec->dims[i].n - 1; ishift -= nm1 * (cld_vec->dims[i].is *= -1); oshift -= nm1 * (cld_vec->dims[i].os *= -1); } } cld = X(mkplan_d)(plnr, X(mkproblem_rdft_1)(p->sz, cld_vec, p->ri + ishift, p->ro + oshift, R2HC)); X(tensor_destroy2)(ri_vec, cld_vec); } if (!cld) return (plan *)0; pln = MKPLAN_DFT(P, &padt, apply); if (p->sz->rnk == 0) { pln->n = 1; pln->os = 0; } else { pln->n = p->sz->dims[0].n; pln->os = p->sz->dims[0].os; } pln->ishift = ishift; pln->oshift = oshift; pln->cld = cld; pln->super.super.ops = cld->ops; pln->super.super.ops.other += 8 * ((pln->n - 1)/2); pln->super.super.ops.add += 4 * ((pln->n - 1)/2); pln->super.super.ops.other += 1; /* estimator hack for nop plans */ return &(pln->super.super); } /* constructor */ static solver *mksolver(void) { static const solver_adt sadt = { PROBLEM_DFT, mkplan, 0 }; S *slv = MKSOLVER(S, &sadt); return &(slv->super); } void X(dft_r2hc_register)(planner *p) { REGISTER_SOLVER(p, mksolver()); } fftw-3.3.3/rdft/dht-r2hc.c0000644000175000017500000000662412054410050012124 00000000000000/* * Copyright (c) 2003, 2007-11 Matteo Frigo * Copyright (c) 2003, 2007-11 Massachusetts Institute of Technology * * This program is free software; you can redistribute it and/or modify * it under the terms of the GNU General Public License as published by * the Free Software Foundation; either version 2 of the License, or * (at your option) any later version. * * This program is distributed in the hope that it will be useful, * but WITHOUT ANY WARRANTY; without even the implied warranty of * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the * GNU General Public License for more details. * * You should have received a copy of the GNU General Public License * along with this program; if not, write to the Free Software * Foundation, Inc., 51 Franklin Street, Fifth Floor, Boston, MA 02110-1301 USA * */ /* Solve a DHT problem (Discrete Hartley Transform) via post-processing of an R2HC problem. */ #include "rdft.h" typedef struct { solver super; } S; typedef struct { plan_rdft super; plan *cld; INT os; INT n; } P; static void apply(const plan *ego_, R *I, R *O) { const P *ego = (const P *) ego_; INT os = ego->os; INT i, n = ego->n; { plan_rdft *cld = (plan_rdft *) ego->cld; cld->apply((plan *) cld, I, O); } for (i = 1; i < n - i; ++i) { E a, b; a = O[os * i]; b = O[os * (n - i)]; #if FFT_SIGN == -1 O[os * i] = a - b; O[os * (n - i)] = a + b; #else O[os * i] = a + b; O[os * (n - i)] = a - b; #endif } } static void awake(plan *ego_, enum wakefulness wakefulness) { P *ego = (P *) ego_; X(plan_awake)(ego->cld, wakefulness); } static void destroy(plan *ego_) { P *ego = (P *) ego_; X(plan_destroy_internal)(ego->cld); } static void print(const plan *ego_, printer *p) { const P *ego = (const P *) ego_; p->print(p, "(dht-r2hc-%D%(%p%))", ego->n, ego->cld); } static int applicable0(const problem *p_, const planner *plnr) { const problem_rdft *p = (const problem_rdft *) p_; return (1 && !NO_DHT_R2HCP(plnr) && p->sz->rnk == 1 && p->vecsz->rnk == 0 && p->kind[0] == DHT ); } static int applicable(const solver *ego, const problem *p, const planner *plnr) { UNUSED(ego); return (!NO_SLOWP(plnr) && applicable0(p, plnr)); } static plan *mkplan(const solver *ego_, const problem *p_, planner *plnr) { P *pln; const problem_rdft *p; plan *cld; static const plan_adt padt = { X(rdft_solve), awake, print, destroy }; if (!applicable(ego_, p_, plnr)) return (plan *)0; p = (const problem_rdft *) p_; /* NO_DHT_R2HC stops infinite loops with rdft-dht.c */ cld = X(mkplan_f_d)(plnr, X(mkproblem_rdft_1)(p->sz, p->vecsz, p->I, p->O, R2HC), NO_DHT_R2HC, 0, 0); if (!cld) return (plan *)0; pln = MKPLAN_RDFT(P, &padt, apply); pln->n = p->sz->dims[0].n; pln->os = p->sz->dims[0].os; pln->cld = cld; pln->super.super.ops = cld->ops; pln->super.super.ops.other += 4 * ((pln->n - 1)/2); pln->super.super.ops.add += 2 * ((pln->n - 1)/2); return &(pln->super.super); } /* constructor */ static solver *mksolver(void) { static const solver_adt sadt = { PROBLEM_RDFT, mkplan, 0 }; S *slv = MKSOLVER(S, &sadt); return &(slv->super); } void X(dht_r2hc_register)(planner *p) { REGISTER_SOLVER(p, mksolver()); } fftw-3.3.3/rdft/dht-rader.c0000644000175000017500000002431012054410050012353 00000000000000/* * Copyright (c) 2003, 2007-11 Matteo Frigo * Copyright (c) 2003, 2007-11 Massachusetts Institute of Technology * * This program is free software; you can redistribute it and/or modify * it under the terms of the GNU General Public License as published by * the Free Software Foundation; either version 2 of the License, or * (at your option) any later version. * * This program is distributed in the hope that it will be useful, * but WITHOUT ANY WARRANTY; without even the implied warranty of * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the * GNU General Public License for more details. * * You should have received a copy of the GNU General Public License * along with this program; if not, write to the Free Software * Foundation, Inc., 51 Franklin Street, Fifth Floor, Boston, MA 02110-1301 USA * */ #include "rdft.h" /* * Compute DHTs of prime sizes using Rader's trick: turn them * into convolutions of size n - 1, which we then perform via a pair * of FFTs. (We can then do prime real FFTs via rdft-dht.c.) * * Optionally (determined by the "pad" field of the solver), we can * perform the (cyclic) convolution by zero-padding to a size * >= 2*(n-1) - 1. This is advantageous if n-1 has large prime factors. * */ typedef struct { solver super; int pad; } S; typedef struct { plan_rdft super; plan *cld1, *cld2; R *omega; INT n, npad, g, ginv; INT is, os; plan *cld_omega; } P; static rader_tl *omegas = 0; /***************************************************************************/ /* If R2HC_ONLY_CONV is 1, we use a trick to perform the convolution purely in terms of R2HC transforms, as opposed to R2HC followed by H2RC. This requires a few more operations, but allows us to share the same plan/codelets for both Rader children. */ #define R2HC_ONLY_CONV 1 static void apply(const plan *ego_, R *I, R *O) { const P *ego = (const P *) ego_; INT n = ego->n; /* prime */ INT npad = ego->npad; /* == n - 1 for unpadded Rader; always even */ INT is = ego->is, os; INT k, gpower, g; R *buf, *omega; R r0; buf = (R *) MALLOC(sizeof(R) * npad, BUFFERS); /* First, permute the input, storing in buf: */ g = ego->g; for (gpower = 1, k = 0; k < n - 1; ++k, gpower = MULMOD(gpower, g, n)) { buf[k] = I[gpower * is]; } /* gpower == g^(n-1) mod n == 1 */; A(n - 1 <= npad); for (k = n - 1; k < npad; ++k) /* optionally, zero-pad convolution */ buf[k] = 0; os = ego->os; /* compute RDFT of buf, storing in buf (i.e., in-place): */ { plan_rdft *cld = (plan_rdft *) ego->cld1; cld->apply((plan *) cld, buf, buf); } /* set output DC component: */ O[0] = (r0 = I[0]) + buf[0]; /* now, multiply by omega: */ omega = ego->omega; buf[0] *= omega[0]; for (k = 1; k < npad/2; ++k) { E rB, iB, rW, iW, a, b; rW = omega[k]; iW = omega[npad - k]; rB = buf[k]; iB = buf[npad - k]; a = rW * rB - iW * iB; b = rW * iB + iW * rB; #if R2HC_ONLY_CONV buf[k] = a + b; buf[npad - k] = a - b; #else buf[k] = a; buf[npad - k] = b; #endif } /* Nyquist component: */ A(k + k == npad); /* since npad is even */ buf[k] *= omega[k]; /* this will add input[0] to all of the outputs after the ifft */ buf[0] += r0; /* inverse FFT: */ { plan_rdft *cld = (plan_rdft *) ego->cld2; cld->apply((plan *) cld, buf, buf); } /* do inverse permutation to unshuffle the output: */ A(gpower == 1); #if R2HC_ONLY_CONV O[os] = buf[0]; gpower = g = ego->ginv; A(npad == n - 1 || npad/2 >= n - 1); if (npad == n - 1) { for (k = 1; k < npad/2; ++k, gpower = MULMOD(gpower, g, n)) { O[gpower * os] = buf[k] + buf[npad - k]; } O[gpower * os] = buf[k]; ++k, gpower = MULMOD(gpower, g, n); for (; k < npad; ++k, gpower = MULMOD(gpower, g, n)) { O[gpower * os] = buf[npad - k] - buf[k]; } } else { for (k = 1; k < n - 1; ++k, gpower = MULMOD(gpower, g, n)) { O[gpower * os] = buf[k] + buf[npad - k]; } } #else g = ego->ginv; for (k = 0; k < n - 1; ++k, gpower = MULMOD(gpower, g, n)) { O[gpower * os] = buf[k]; } #endif A(gpower == 1); X(ifree)(buf); } static R *mkomega(enum wakefulness wakefulness, plan *p_, INT n, INT npad, INT ginv) { plan_rdft *p = (plan_rdft *) p_; R *omega; INT i, gpower; trigreal scale; triggen *t; if ((omega = X(rader_tl_find)(n, npad + 1, ginv, omegas))) return omega; omega = (R *)MALLOC(sizeof(R) * npad, TWIDDLES); scale = npad; /* normalization for convolution */ t = X(mktriggen)(wakefulness, n); for (i = 0, gpower = 1; i < n-1; ++i, gpower = MULMOD(gpower, ginv, n)) { trigreal w[2]; t->cexpl(t, gpower, w); omega[i] = (w[0] + w[1]) / scale; } X(triggen_destroy)(t); A(gpower == 1); A(npad == n - 1 || npad >= 2*(n - 1) - 1); for (; i < npad; ++i) omega[i] = K(0.0); if (npad > n - 1) for (i = 1; i < n-1; ++i) omega[npad - i] = omega[n - 1 - i]; p->apply(p_, omega, omega); X(rader_tl_insert)(n, npad + 1, ginv, omega, &omegas); return omega; } static void free_omega(R *omega) { X(rader_tl_delete)(omega, &omegas); } /***************************************************************************/ static void awake(plan *ego_, enum wakefulness wakefulness) { P *ego = (P *) ego_; X(plan_awake)(ego->cld1, wakefulness); X(plan_awake)(ego->cld2, wakefulness); X(plan_awake)(ego->cld_omega, wakefulness); switch (wakefulness) { case SLEEPY: free_omega(ego->omega); ego->omega = 0; break; default: ego->g = X(find_generator)(ego->n); ego->ginv = X(power_mod)(ego->g, ego->n - 2, ego->n); A(MULMOD(ego->g, ego->ginv, ego->n) == 1); A(!ego->omega); ego->omega = mkomega(wakefulness, ego->cld_omega,ego->n,ego->npad,ego->ginv); break; } } static void destroy(plan *ego_) { P *ego = (P *) ego_; X(plan_destroy_internal)(ego->cld_omega); X(plan_destroy_internal)(ego->cld2); X(plan_destroy_internal)(ego->cld1); } static void print(const plan *ego_, printer *p) { const P *ego = (const P *) ego_; p->print(p, "(dht-rader-%D/%D%ois=%oos=%(%p%)", ego->n, ego->npad, ego->is, ego->os, ego->cld1); if (ego->cld2 != ego->cld1) p->print(p, "%(%p%)", ego->cld2); if (ego->cld_omega != ego->cld1 && ego->cld_omega != ego->cld2) p->print(p, "%(%p%)", ego->cld_omega); p->putchr(p, ')'); } static int applicable(const solver *ego, const problem *p_, const planner *plnr) { const problem_rdft *p = (const problem_rdft *) p_; UNUSED(ego); return (1 && p->sz->rnk == 1 && p->vecsz->rnk == 0 && p->kind[0] == DHT && X(is_prime)(p->sz->dims[0].n) && p->sz->dims[0].n > 2 && CIMPLIES(NO_SLOWP(plnr), p->sz->dims[0].n > RADER_MAX_SLOW) /* proclaim the solver SLOW if p-1 is not easily factorizable. Unlike in the complex case where Bluestein can solve the problem, in the DHT case we may have no other choice */ && CIMPLIES(NO_SLOWP(plnr), X(factors_into_small_primes)(p->sz->dims[0].n - 1)) ); } static INT choose_transform_size(INT minsz) { static const INT primes[] = { 2, 3, 5, 0 }; while (!X(factors_into)(minsz, primes) || minsz % 2) ++minsz; return minsz; } static plan *mkplan(const solver *ego_, const problem *p_, planner *plnr) { const S *ego = (const S *) ego_; const problem_rdft *p = (const problem_rdft *) p_; P *pln; INT n, npad; INT is, os; plan *cld1 = (plan *) 0; plan *cld2 = (plan *) 0; plan *cld_omega = (plan *) 0; R *buf = (R *) 0; problem *cldp; static const plan_adt padt = { X(rdft_solve), awake, print, destroy }; if (!applicable(ego_, p_, plnr)) return (plan *) 0; n = p->sz->dims[0].n; is = p->sz->dims[0].is; os = p->sz->dims[0].os; if (ego->pad) npad = choose_transform_size(2 * (n - 1) - 1); else npad = n - 1; /* initial allocation for the purpose of planning */ buf = (R *) MALLOC(sizeof(R) * npad, BUFFERS); cld1 = X(mkplan_f_d)(plnr, X(mkproblem_rdft_1_d)(X(mktensor_1d)(npad, 1, 1), X(mktensor_1d)(1, 0, 0), buf, buf, R2HC), NO_SLOW, 0, 0); if (!cld1) goto nada; cldp = X(mkproblem_rdft_1_d)( X(mktensor_1d)(npad, 1, 1), X(mktensor_1d)(1, 0, 0), buf, buf, #if R2HC_ONLY_CONV R2HC #else HC2R #endif ); if (!(cld2 = X(mkplan_f_d)(plnr, cldp, NO_SLOW, 0, 0))) goto nada; /* plan for omega */ cld_omega = X(mkplan_f_d)(plnr, X(mkproblem_rdft_1_d)( X(mktensor_1d)(npad, 1, 1), X(mktensor_1d)(1, 0, 0), buf, buf, R2HC), NO_SLOW, ESTIMATE, 0); if (!cld_omega) goto nada; /* deallocate buffers; let awake() or apply() allocate them for real */ X(ifree)(buf); buf = 0; pln = MKPLAN_RDFT(P, &padt, apply); pln->cld1 = cld1; pln->cld2 = cld2; pln->cld_omega = cld_omega; pln->omega = 0; pln->n = n; pln->npad = npad; pln->is = is; pln->os = os; X(ops_add)(&cld1->ops, &cld2->ops, &pln->super.super.ops); pln->super.super.ops.other += (npad/2-1)*6 + npad + n + (n-1) * ego->pad; pln->super.super.ops.add += (npad/2-1)*2 + 2 + (n-1) * ego->pad; pln->super.super.ops.mul += (npad/2-1)*4 + 2 + ego->pad; #if R2HC_ONLY_CONV pln->super.super.ops.other += n-2 - ego->pad; pln->super.super.ops.add += (npad/2-1)*2 + (n-2) - ego->pad; #endif return &(pln->super.super); nada: X(ifree0)(buf); X(plan_destroy_internal)(cld_omega); X(plan_destroy_internal)(cld2); X(plan_destroy_internal)(cld1); return 0; } /* constructors */ static solver *mksolver(int pad) { static const solver_adt sadt = { PROBLEM_RDFT, mkplan, 0 }; S *slv = MKSOLVER(S, &sadt); slv->pad = pad; return &(slv->super); } void X(dht_rader_register)(planner *p) { REGISTER_SOLVER(p, mksolver(0)); REGISTER_SOLVER(p, mksolver(1)); } fftw-3.3.3/rdft/buffered.c0000644000175000017500000002237712054410050012276 00000000000000/* * Copyright (c) 2003, 2007-11 Matteo Frigo * Copyright (c) 2003, 2007-11 Massachusetts Institute of Technology * * This program is free software; you can redistribute it and/or modify * it under the terms of the GNU General Public License as published by * the Free Software Foundation; either version 2 of the License, or * (at your option) any later version. * * This program is distributed in the hope that it will be useful, * but WITHOUT ANY WARRANTY; without even the implied warranty of * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the * GNU General Public License for more details. * * You should have received a copy of the GNU General Public License * along with this program; if not, write to the Free Software * Foundation, Inc., 51 Franklin Street, Fifth Floor, Boston, MA 02110-1301 USA * */ #include "rdft.h" typedef struct { solver super; int maxnbuf_ndx; } S; static const INT maxnbufs[] = { 8, 256 }; typedef struct { plan_rdft super; plan *cld, *cldcpy, *cldrest; INT n, vl, nbuf, bufdist; INT ivs_by_nbuf, ovs_by_nbuf; } P; /* transform a vector input with the help of bufs */ static void apply(const plan *ego_, R *I, R *O) { const P *ego = (const P *) ego_; plan_rdft *cld = (plan_rdft *) ego->cld; plan_rdft *cldcpy = (plan_rdft *) ego->cldcpy; plan_rdft *cldrest; INT i, vl = ego->vl, nbuf = ego->nbuf; INT ivs_by_nbuf = ego->ivs_by_nbuf, ovs_by_nbuf = ego->ovs_by_nbuf; R *bufs; bufs = (R *)MALLOC(sizeof(R) * nbuf * ego->bufdist, BUFFERS); for (i = nbuf; i <= vl; i += nbuf) { /* transform to bufs: */ cld->apply((plan *) cld, I, bufs); I += ivs_by_nbuf; /* copy back */ cldcpy->apply((plan *) cldcpy, bufs, O); O += ovs_by_nbuf; } X(ifree)(bufs); /* Do the remaining transforms, if any: */ cldrest = (plan_rdft *) ego->cldrest; cldrest->apply((plan *) cldrest, I, O); } /* for hc2r problems, copy the input into buffer, and then transform buffer->output, which allows for destruction of the buffer */ static void apply_hc2r(const plan *ego_, R *I, R *O) { const P *ego = (const P *) ego_; plan_rdft *cld = (plan_rdft *) ego->cld; plan_rdft *cldcpy = (plan_rdft *) ego->cldcpy; plan_rdft *cldrest; INT i, vl = ego->vl, nbuf = ego->nbuf; INT ivs_by_nbuf = ego->ivs_by_nbuf, ovs_by_nbuf = ego->ovs_by_nbuf; R *bufs; bufs = (R *)MALLOC(sizeof(R) * nbuf * ego->bufdist, BUFFERS); for (i = nbuf; i <= vl; i += nbuf) { /* copy input into bufs: */ cldcpy->apply((plan *) cldcpy, I, bufs); I += ivs_by_nbuf; /* transform to output */ cld->apply((plan *) cld, bufs, O); O += ovs_by_nbuf; } X(ifree)(bufs); /* Do the remaining transforms, if any: */ cldrest = (plan_rdft *) ego->cldrest; cldrest->apply((plan *) cldrest, I, O); } static void awake(plan *ego_, enum wakefulness wakefulness) { P *ego = (P *) ego_; X(plan_awake)(ego->cld, wakefulness); X(plan_awake)(ego->cldcpy, wakefulness); X(plan_awake)(ego->cldrest, wakefulness); } static void destroy(plan *ego_) { P *ego = (P *) ego_; X(plan_destroy_internal)(ego->cldrest); X(plan_destroy_internal)(ego->cldcpy); X(plan_destroy_internal)(ego->cld); } static void print(const plan *ego_, printer *p) { const P *ego = (const P *) ego_; p->print(p, "(rdft-buffered-%D%v/%D-%D%(%p%)%(%p%)%(%p%))", ego->n, ego->nbuf, ego->vl, ego->bufdist % ego->n, ego->cld, ego->cldcpy, ego->cldrest); } static int applicable0(const S *ego, const problem *p_, const planner *plnr) { const problem_rdft *p = (const problem_rdft *) p_; iodim *d = p->sz->dims; if (1 && p->vecsz->rnk <= 1 && p->sz->rnk == 1 ) { INT vl, ivs, ovs; X(tensor_tornk1)(p->vecsz, &vl, &ivs, &ovs); if (X(toobig)(d[0].n) && CONSERVE_MEMORYP(plnr)) return 0; /* if this solver is redundant, in the sense that a solver of lower index generates the same plan, then prune this solver */ if (X(nbuf_redundant)(d[0].n, vl, ego->maxnbuf_ndx, maxnbufs, NELEM(maxnbufs))) return 0; if (p->I != p->O) { if (p->kind[0] == HC2R) { /* Allow HC2R problems only if the input is to be preserved. This solver sets NO_DESTROY_INPUT, which prevents infinite loops */ return (NO_DESTROY_INPUTP(plnr)); } else { /* In principle, the buffered transforms might be useful when working out of place. However, in order to prevent infinite loops in the planner, we require that the output stride of the buffered transforms be greater than 1. */ return (d[0].os > 1); } } /* * If the problem is in place, the input/output strides must * be the same or the whole thing must fit in the buffer. */ if (X(tensor_inplace_strides2)(p->sz, p->vecsz)) return 1; if (/* fits into buffer: */ ((p->vecsz->rnk == 0) || (X(nbuf)(d[0].n, p->vecsz->dims[0].n, maxnbufs[ego->maxnbuf_ndx]) == p->vecsz->dims[0].n))) return 1; } return 0; } static int applicable(const S *ego, const problem *p_, const planner *plnr) { const problem_rdft *p; if (NO_BUFFERINGP(plnr)) return 0; if (!applicable0(ego, p_, plnr)) return 0; p = (const problem_rdft *) p_; if (p->kind[0] == HC2R) { if (NO_UGLYP(plnr)) { /* UGLY if in-place and too big, since the problem could be solved via transpositions */ if (p->I == p->O && X(toobig)(p->sz->dims[0].n)) return 0; } } else { if (NO_UGLYP(plnr)) { if (p->I != p->O) return 0; if (X(toobig)(p->sz->dims[0].n)) return 0; } } return 1; } static plan *mkplan(const solver *ego_, const problem *p_, planner *plnr) { P *pln; const S *ego = (const S *)ego_; plan *cld = (plan *) 0; plan *cldcpy = (plan *) 0; plan *cldrest = (plan *) 0; const problem_rdft *p = (const problem_rdft *) p_; R *bufs = (R *) 0; INT nbuf = 0, bufdist, n, vl; INT ivs, ovs; int hc2rp; static const plan_adt padt = { X(rdft_solve), awake, print, destroy }; if (!applicable(ego, p_, plnr)) goto nada; n = X(tensor_sz)(p->sz); X(tensor_tornk1)(p->vecsz, &vl, &ivs, &ovs); hc2rp = (p->kind[0] == HC2R); nbuf = X(nbuf)(n, vl, maxnbufs[ego->maxnbuf_ndx]); bufdist = X(bufdist)(n, vl); A(nbuf > 0); /* initial allocation for the purpose of planning */ bufs = (R *) MALLOC(sizeof(R) * nbuf * bufdist, BUFFERS); if (hc2rp) { /* allow destruction of buffer */ cld = X(mkplan_f_d)(plnr, X(mkproblem_rdft_d)( X(mktensor_1d)(n, 1, p->sz->dims[0].os), X(mktensor_1d)(nbuf, bufdist, ovs), bufs, TAINT(p->O, ovs * nbuf), p->kind), 0, 0, NO_DESTROY_INPUT); if (!cld) goto nada; /* copying input into buffer buffer is a rank-0 transform: */ cldcpy = X(mkplan_d)(plnr, X(mkproblem_rdft_0_d)( X(mktensor_2d)(nbuf, ivs, bufdist, n, p->sz->dims[0].is, 1), TAINT(p->I, ivs * nbuf), bufs)); if (!cldcpy) goto nada; } else { /* allow destruction of input if problem is in place */ cld = X(mkplan_f_d)(plnr, X(mkproblem_rdft_d)( X(mktensor_1d)(n, p->sz->dims[0].is, 1), X(mktensor_1d)(nbuf, ivs, bufdist), TAINT(p->I, ivs * nbuf), bufs, p->kind), 0, 0, (p->I == p->O) ? NO_DESTROY_INPUT : 0); if (!cld) goto nada; /* copying back from the buffer is a rank-0 transform: */ cldcpy = X(mkplan_d)(plnr, X(mkproblem_rdft_0_d)( X(mktensor_2d)(nbuf, bufdist, ovs, n, 1, p->sz->dims[0].os), bufs, TAINT(p->O, ovs * nbuf))); if (!cldcpy) goto nada; } /* deallocate buffers, let apply() allocate them for real */ X(ifree)(bufs); bufs = 0; /* plan the leftover transforms (cldrest): */ { INT id = ivs * (nbuf * (vl / nbuf)); INT od = ovs * (nbuf * (vl / nbuf)); cldrest = X(mkplan_d)(plnr, X(mkproblem_rdft_d)( X(tensor_copy)(p->sz), X(mktensor_1d)(vl % nbuf, ivs, ovs), p->I + id, p->O + od, p->kind)); } if (!cldrest) goto nada; pln = MKPLAN_RDFT(P, &padt, hc2rp ? apply_hc2r : apply); pln->cld = cld; pln->cldcpy = cldcpy; pln->cldrest = cldrest; pln->n = n; pln->vl = vl; pln->ivs_by_nbuf = ivs * nbuf; pln->ovs_by_nbuf = ovs * nbuf; pln->nbuf = nbuf; pln->bufdist = bufdist; { opcnt t; X(ops_add)(&cld->ops, &cldcpy->ops, &t); X(ops_madd)(vl / nbuf, &t, &cldrest->ops, &pln->super.super.ops); } return &(pln->super.super); nada: X(ifree0)(bufs); X(plan_destroy_internal)(cldrest); X(plan_destroy_internal)(cldcpy); X(plan_destroy_internal)(cld); return (plan *) 0; } static solver *mksolver(int maxnbuf_ndx) { static const solver_adt sadt = { PROBLEM_RDFT, mkplan, 0 }; S *slv = MKSOLVER(S, &sadt); slv->maxnbuf_ndx = maxnbuf_ndx; return &(slv->super); } void X(rdft_buffered_register)(planner *p) { size_t i; for (i = 0; i < NELEM(maxnbufs); ++i) REGISTER_SOLVER(p, mksolver(i)); } fftw-3.3.3/rdft/codelet-rdft.h0000644000175000017500000001133612054410050013066 00000000000000/* * Copyright (c) 2003, 2007-11 Matteo Frigo * Copyright (c) 2003, 2007-11 Massachusetts Institute of Technology * * This program is free software; you can redistribute it and/or modify * it under the terms of the GNU General Public License as published by * the Free Software Foundation; either version 2 of the License, or * (at your option) any later version. * * This program is distributed in the hope that it will be useful, * but WITHOUT ANY WARRANTY; without even the implied warranty of * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the * GNU General Public License for more details. * * You should have received a copy of the GNU General Public License * along with this program; if not, write to the Free Software * Foundation, Inc., 51 Franklin Street, Fifth Floor, Boston, MA 02110-1301 USA * */ /* * This header file must include every file or define every * type or macro which is required to compile a codelet. */ #ifndef __RDFT_CODELET_H__ #define __RDFT_CODELET_H__ #include "ifftw.h" /************************************************************** * types of codelets **************************************************************/ /* FOOab, with a,b in {0,1}, denotes the FOO transform where a/b say whether the input/output are shifted by half a sample/slot. */ typedef enum { R2HC00, R2HC01, R2HC10, R2HC11, HC2R00, HC2R01, HC2R10, HC2R11, DHT, REDFT00, REDFT01, REDFT10, REDFT11, /* real-even == DCT's */ RODFT00, RODFT01, RODFT10, RODFT11 /* real-odd == DST's */ } rdft_kind; /* standard R2HC/HC2R transforms are unshifted */ #define R2HC R2HC00 #define HC2R HC2R00 #define R2HCII R2HC01 #define HC2RIII HC2R10 /* (k) >= R2HC00 produces a warning under gcc because checking x >= 0 is superfluous for unsigned values...but it is needed because other compilers (e.g. icc) may define the enum to be a signed int...grrr. */ #define R2HC_KINDP(k) ((k) >= R2HC00 && (k) <= R2HC11) /* uses kr2hc_genus */ #define HC2R_KINDP(k) ((k) >= HC2R00 && (k) <= HC2R11) /* uses khc2r_genus */ #define R2R_KINDP(k) ((k) >= DHT) /* uses kr2r_genus */ #define REDFT_KINDP(k) ((k) >= REDFT00 && (k) <= REDFT11) #define RODFT_KINDP(k) ((k) >= RODFT00 && (k) <= RODFT11) #define REODFT_KINDP(k) ((k) >= REDFT00 && (k) <= RODFT11) /* codelets with real input (output) and complex output (input) */ typedef struct kr2c_desc_s kr2c_desc; typedef struct { rdft_kind kind; INT vl; } kr2c_genus; struct kr2c_desc_s { INT n; /* size of transform computed */ const char *nam; opcnt ops; const kr2c_genus *genus; }; typedef void (*kr2c) (R *R0, R *R1, R *Cr, R *Ci, stride rs, stride csr, stride csi, INT vl, INT ivs, INT ovs); void X(kr2c_register)(planner *p, kr2c codelet, const kr2c_desc *desc); /* half-complex to half-complex DIT/DIF codelets: */ typedef struct hc2hc_desc_s hc2hc_desc; typedef struct { rdft_kind kind; INT vl; } hc2hc_genus; struct hc2hc_desc_s { INT radix; const char *nam; const tw_instr *tw; const hc2hc_genus *genus; opcnt ops; }; typedef void (*khc2hc) (R *rioarray, R *iioarray, const R *W, stride rs, INT mb, INT me, INT ms); void X(khc2hc_register)(planner *p, khc2hc codelet, const hc2hc_desc *desc); /* half-complex to rdft2-complex DIT/DIF codelets: */ typedef struct hc2c_desc_s hc2c_desc; typedef enum { HC2C_VIA_RDFT, HC2C_VIA_DFT } hc2c_kind; typedef struct { int (*okp)( const R *Rp, const R *Ip, const R *Rm, const R *Im, INT rs, INT mb, INT me, INT ms, const planner *plnr); rdft_kind kind; INT vl; } hc2c_genus; struct hc2c_desc_s { INT radix; const char *nam; const tw_instr *tw; const hc2c_genus *genus; opcnt ops; }; typedef void (*khc2c) (R *Rp, R *Ip, R *Rm, R *Im, const R *W, stride rs, INT mb, INT me, INT ms); void X(khc2c_register)(planner *p, khc2c codelet, const hc2c_desc *desc, hc2c_kind hc2ckind); extern const solvtab X(solvtab_rdft_r2cf); extern const solvtab X(solvtab_rdft_r2cb); extern const solvtab X(solvtab_rdft_sse2); extern const solvtab X(solvtab_rdft_avx); extern const solvtab X(solvtab_rdft_altivec); extern const solvtab X(solvtab_rdft_neon); /* real-input & output DFT-like codelets (DHT, etc.) */ typedef struct kr2r_desc_s kr2r_desc; typedef struct { INT vl; } kr2r_genus; struct kr2r_desc_s { INT n; /* size of transform computed */ const char *nam; opcnt ops; const kr2r_genus *genus; rdft_kind kind; }; typedef void (*kr2r) (const R *I, R *O, stride is, stride os, INT vl, INT ivs, INT ovs); void X(kr2r_register)(planner *p, kr2r codelet, const kr2r_desc *desc); extern const solvtab X(solvtab_rdft_r2r); #endif /* __RDFT_CODELET_H__ */ fftw-3.3.3/rdft/conf.c0000644000175000017500000000435012054410050011430 00000000000000/* * Copyright (c) 2003, 2007-11 Matteo Frigo * Copyright (c) 2003, 2007-11 Massachusetts Institute of Technology * * This program is free software; you can redistribute it and/or modify * it under the terms of the GNU General Public License as published by * the Free Software Foundation; either version 2 of the License, or * (at your option) any later version. * * This program is distributed in the hope that it will be useful, * but WITHOUT ANY WARRANTY; without even the implied warranty of * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the * GNU General Public License for more details. * * You should have received a copy of the GNU General Public License * along with this program; if not, write to the Free Software * Foundation, Inc., 51 Franklin Street, Fifth Floor, Boston, MA 02110-1301 USA * */ #include "rdft.h" static const solvtab s = { SOLVTAB(X(rdft_indirect_register)), SOLVTAB(X(rdft_rank0_register)), SOLVTAB(X(rdft_vrank3_transpose_register)), SOLVTAB(X(rdft_vrank_geq1_register)), SOLVTAB(X(rdft_nop_register)), SOLVTAB(X(rdft_buffered_register)), SOLVTAB(X(rdft_generic_register)), SOLVTAB(X(rdft_rank_geq2_register)), SOLVTAB(X(dft_r2hc_register)), SOLVTAB(X(rdft_dht_register)), SOLVTAB(X(dht_r2hc_register)), SOLVTAB(X(dht_rader_register)), SOLVTAB(X(rdft2_vrank_geq1_register)), SOLVTAB(X(rdft2_nop_register)), SOLVTAB(X(rdft2_rank0_register)), SOLVTAB(X(rdft2_buffered_register)), SOLVTAB(X(rdft2_rank_geq2_register)), SOLVTAB(X(rdft2_rdft_register)), SOLVTAB(X(hc2hc_generic_register)), SOLVTAB_END }; void X(rdft_conf_standard)(planner *p) { X(solvtab_exec)(s, p); X(solvtab_exec)(X(solvtab_rdft_r2cf), p); X(solvtab_exec)(X(solvtab_rdft_r2cb), p); X(solvtab_exec)(X(solvtab_rdft_r2r), p); #if HAVE_SSE2 if (X(have_simd_sse2)()) X(solvtab_exec)(X(solvtab_rdft_sse2), p); #endif #if HAVE_AVX if (X(have_simd_avx)()) X(solvtab_exec)(X(solvtab_rdft_avx), p); #endif #if HAVE_ALTIVEC if (X(have_simd_altivec)()) X(solvtab_exec)(X(solvtab_rdft_altivec), p); #endif #if HAVE_NEON if (X(have_simd_neon)()) X(solvtab_exec)(X(solvtab_rdft_neon), p); #endif } fftw-3.3.3/rdft/direct-r2r.c0000644000175000017500000000663212054410050012465 00000000000000/* * Copyright (c) 2003, 2007-11 Matteo Frigo * Copyright (c) 2003, 2007-11 Massachusetts Institute of Technology * * This program is free software; you can redistribute it and/or modify * it under the terms of the GNU General Public License as published by * the Free Software Foundation; either version 2 of the License, or * (at your option) any later version. * * This program is distributed in the hope that it will be useful, * but WITHOUT ANY WARRANTY; without even the implied warranty of * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the * GNU General Public License for more details. * * You should have received a copy of the GNU General Public License * along with this program; if not, write to the Free Software * Foundation, Inc., 51 Franklin Street, Fifth Floor, Boston, MA 02110-1301 USA * */ /* direct RDFT solver, using r2r codelets */ #include "rdft.h" typedef struct { solver super; const kr2r_desc *desc; kr2r k; } S; typedef struct { plan_rdft super; INT vl, ivs, ovs; stride is, os; kr2r k; const S *slv; } P; static void apply(const plan *ego_, R *I, R *O) { const P *ego = (const P *) ego_; ASSERT_ALIGNED_DOUBLE; ego->k(I, O, ego->is, ego->os, ego->vl, ego->ivs, ego->ovs); } static void destroy(plan *ego_) { P *ego = (P *) ego_; X(stride_destroy)(ego->is); X(stride_destroy)(ego->os); } static void print(const plan *ego_, printer *p) { const P *ego = (const P *) ego_; const S *s = ego->slv; p->print(p, "(rdft-%s-direct-r2r-%D%v \"%s\")", X(rdft_kind_str)(s->desc->kind), s->desc->n, ego->vl, s->desc->nam); } static int applicable(const solver *ego_, const problem *p_) { const S *ego = (const S *) ego_; const problem_rdft *p = (const problem_rdft *) p_; INT vl; INT ivs, ovs; return ( 1 && p->sz->rnk == 1 && p->vecsz->rnk <= 1 && p->sz->dims[0].n == ego->desc->n && p->kind[0] == ego->desc->kind /* check strides etc */ && X(tensor_tornk1)(p->vecsz, &vl, &ivs, &ovs) && (0 /* can operate out-of-place */ || p->I != p->O /* computing one transform */ || vl == 1 /* can operate in-place as long as strides are the same */ || X(tensor_inplace_strides2)(p->sz, p->vecsz) ) ); } static plan *mkplan(const solver *ego_, const problem *p_, planner *plnr) { const S *ego = (const S *) ego_; P *pln; const problem_rdft *p; iodim *d; static const plan_adt padt = { X(rdft_solve), X(null_awake), print, destroy }; UNUSED(plnr); if (!applicable(ego_, p_)) return (plan *)0; p = (const problem_rdft *) p_; pln = MKPLAN_RDFT(P, &padt, apply); d = p->sz->dims; pln->k = ego->k; pln->is = X(mkstride)(d->n, d->is); pln->os = X(mkstride)(d->n, d->os); X(tensor_tornk1)(p->vecsz, &pln->vl, &pln->ivs, &pln->ovs); pln->slv = ego; X(ops_zero)(&pln->super.super.ops); X(ops_madd2)(pln->vl / ego->desc->genus->vl, &ego->desc->ops, &pln->super.super.ops); pln->super.super.could_prune_now_p = 1; return &(pln->super.super); } /* constructor */ solver *X(mksolver_rdft_r2r_direct)(kr2r k, const kr2r_desc *desc) { static const solver_adt sadt = { PROBLEM_RDFT, mkplan, 0 }; S *slv = MKSOLVER(S, &sadt); slv->k = k; slv->desc = desc; return &(slv->super); } fftw-3.3.3/rdft/direct-r2c.c0000644000175000017500000002140612054410050012442 00000000000000/* * Copyright (c) 2003, 2007-11 Matteo Frigo * Copyright (c) 2003, 2007-11 Massachusetts Institute of Technology * * This program is free software; you can redistribute it and/or modify * it under the terms of the GNU General Public License as published by * the Free Software Foundation; either version 2 of the License, or * (at your option) any later version. * * This program is distributed in the hope that it will be useful, * but WITHOUT ANY WARRANTY; without even the implied warranty of * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the * GNU General Public License for more details. * * You should have received a copy of the GNU General Public License * along with this program; if not, write to the Free Software * Foundation, Inc., 51 Franklin Street, Fifth Floor, Boston, MA 02110-1301 USA * */ /* direct RDFT solver, using r2c codelets */ #include "rdft.h" typedef struct { solver super; const kr2c_desc *desc; kr2c k; int bufferedp; } S; typedef struct { plan_rdft super; stride rs, csr, csi; stride brs, bcsr, bcsi; INT n, vl, rs0, ivs, ovs, ioffset, bioffset; kr2c k; const S *slv; } P; /************************************************************* Nonbuffered code *************************************************************/ static void apply_r2hc(const plan *ego_, R *I, R *O) { const P *ego = (const P *) ego_; ASSERT_ALIGNED_DOUBLE; ego->k(I, I + ego->rs0, O, O + ego->ioffset, ego->rs, ego->csr, ego->csi, ego->vl, ego->ivs, ego->ovs); } static void apply_hc2r(const plan *ego_, R *I, R *O) { const P *ego = (const P *) ego_; ASSERT_ALIGNED_DOUBLE; ego->k(O, O + ego->rs0, I, I + ego->ioffset, ego->rs, ego->csr, ego->csi, ego->vl, ego->ivs, ego->ovs); } /************************************************************* Buffered code *************************************************************/ /* should not be 2^k to avoid associativity conflicts */ static INT compute_batchsize(INT radix) { /* round up to multiple of 4 */ radix += 3; radix &= -4; return (radix + 2); } static void dobatch_r2hc(const P *ego, R *I, R *O, R *buf, INT batchsz) { X(cpy2d_ci)(I, buf, ego->n, ego->rs0, WS(ego->bcsr /* hack */, 1), batchsz, ego->ivs, 1, 1); if (IABS(WS(ego->csr, 1)) < IABS(ego->ovs)) { /* transform directly to output */ ego->k(buf, buf + WS(ego->bcsr /* hack */, 1), O, O + ego->ioffset, ego->brs, ego->csr, ego->csi, batchsz, 1, ego->ovs); } else { /* transform to buffer and copy back */ ego->k(buf, buf + WS(ego->bcsr /* hack */, 1), buf, buf + ego->bioffset, ego->brs, ego->bcsr, ego->bcsi, batchsz, 1, 1); X(cpy2d_co)(buf, O, ego->n, WS(ego->bcsr, 1), WS(ego->csr, 1), batchsz, 1, ego->ovs, 1); } } static void dobatch_hc2r(const P *ego, R *I, R *O, R *buf, INT batchsz) { if (IABS(WS(ego->csr, 1)) < IABS(ego->ivs)) { /* transform directly from input */ ego->k(buf, buf + WS(ego->bcsr /* hack */, 1), I, I + ego->ioffset, ego->brs, ego->csr, ego->csi, batchsz, ego->ivs, 1); } else { /* copy into buffer and transform in place */ X(cpy2d_ci)(I, buf, ego->n, WS(ego->csr, 1), WS(ego->bcsr, 1), batchsz, ego->ivs, 1, 1); ego->k(buf, buf + WS(ego->bcsr /* hack */, 1), buf, buf + ego->bioffset, ego->brs, ego->bcsr, ego->bcsi, batchsz, 1, 1); } X(cpy2d_co)(buf, O, ego->n, WS(ego->bcsr /* hack */, 1), ego->rs0, batchsz, 1, ego->ovs, 1); } static void iterate(const P *ego, R *I, R *O, void (*dobatch)(const P *ego, R *I, R *O, R *buf, INT batchsz)) { R *buf; INT vl = ego->vl; INT n = ego->n; INT i; INT batchsz = compute_batchsize(n); size_t bufsz = n * batchsz * sizeof(R); BUF_ALLOC(R *, buf, bufsz); for (i = 0; i < vl - batchsz; i += batchsz) { dobatch(ego, I, O, buf, batchsz); I += batchsz * ego->ivs; O += batchsz * ego->ovs; } dobatch(ego, I, O, buf, vl - i); BUF_FREE(buf, bufsz); } static void apply_buf_r2hc(const plan *ego_, R *I, R *O) { iterate((const P *) ego_, I, O, dobatch_r2hc); } static void apply_buf_hc2r(const plan *ego_, R *I, R *O) { iterate((const P *) ego_, I, O, dobatch_hc2r); } static void destroy(plan *ego_) { P *ego = (P *) ego_; X(stride_destroy)(ego->rs); X(stride_destroy)(ego->csr); X(stride_destroy)(ego->csi); X(stride_destroy)(ego->brs); X(stride_destroy)(ego->bcsr); X(stride_destroy)(ego->bcsi); } static void print(const plan *ego_, printer *p) { const P *ego = (const P *) ego_; const S *s = ego->slv; if (ego->slv->bufferedp) p->print(p, "(rdft-%s-directbuf/%D-r2c-%D%v \"%s\")", X(rdft_kind_str)(s->desc->genus->kind), /* hack */ WS(ego->bcsr, 1), ego->n, ego->vl, s->desc->nam); else p->print(p, "(rdft-%s-direct-r2c-%D%v \"%s\")", X(rdft_kind_str)(s->desc->genus->kind), ego->n, ego->vl, s->desc->nam); } static INT ioffset(rdft_kind kind, INT sz, INT s) { return(s * ((kind == R2HC || kind == HC2R) ? sz : (sz - 1))); } static int applicable(const solver *ego_, const problem *p_) { const S *ego = (const S *) ego_; const kr2c_desc *desc = ego->desc; const problem_rdft *p = (const problem_rdft *) p_; INT vl, ivs, ovs; return ( 1 && p->sz->rnk == 1 && p->vecsz->rnk <= 1 && p->sz->dims[0].n == desc->n && p->kind[0] == desc->genus->kind /* check strides etc */ && X(tensor_tornk1)(p->vecsz, &vl, &ivs, &ovs) && (0 /* can operate out-of-place */ || p->I != p->O /* computing one transform */ || vl == 1 /* can operate in-place as long as strides are the same */ || X(tensor_inplace_strides2)(p->sz, p->vecsz) ) ); } static int applicable_buf(const solver *ego_, const problem *p_) { const S *ego = (const S *) ego_; const kr2c_desc *desc = ego->desc; const problem_rdft *p = (const problem_rdft *) p_; INT vl, ivs, ovs, batchsz; return ( 1 && p->sz->rnk == 1 && p->vecsz->rnk <= 1 && p->sz->dims[0].n == desc->n && p->kind[0] == desc->genus->kind /* check strides etc */ && X(tensor_tornk1)(p->vecsz, &vl, &ivs, &ovs) && (batchsz = compute_batchsize(desc->n), 1) && (0 /* can operate out-of-place */ || p->I != p->O /* can operate in-place as long as strides are the same */ || X(tensor_inplace_strides2)(p->sz, p->vecsz) /* can do it if the problem fits in the buffer, no matter what the strides are */ || vl <= batchsz ) ); } static plan *mkplan(const solver *ego_, const problem *p_, planner *plnr) { const S *ego = (const S *) ego_; P *pln; const problem_rdft *p; iodim *d; INT rs, cs, b, n; static const plan_adt padt = { X(rdft_solve), X(null_awake), print, destroy }; UNUSED(plnr); if (ego->bufferedp) { if (!applicable_buf(ego_, p_)) return (plan *)0; } else { if (!applicable(ego_, p_)) return (plan *)0; } p = (const problem_rdft *) p_; if (R2HC_KINDP(p->kind[0])) { rs = p->sz->dims[0].is; cs = p->sz->dims[0].os; pln = MKPLAN_RDFT(P, &padt, ego->bufferedp ? apply_buf_r2hc : apply_r2hc); } else { rs = p->sz->dims[0].os; cs = p->sz->dims[0].is; pln = MKPLAN_RDFT(P, &padt, ego->bufferedp ? apply_buf_hc2r : apply_hc2r); } d = p->sz->dims; n = d[0].n; pln->k = ego->k; pln->n = n; pln->rs0 = rs; pln->rs = X(mkstride)(n, 2 * rs); pln->csr = X(mkstride)(n, cs); pln->csi = X(mkstride)(n, -cs); pln->ioffset = ioffset(p->kind[0], n, cs); b = compute_batchsize(n); pln->brs = X(mkstride)(n, 2 * b); pln->bcsr = X(mkstride)(n, b); pln->bcsi = X(mkstride)(n, -b); pln->bioffset = ioffset(p->kind[0], n, b); X(tensor_tornk1)(p->vecsz, &pln->vl, &pln->ivs, &pln->ovs); pln->slv = ego; X(ops_zero)(&pln->super.super.ops); X(ops_madd2)(pln->vl / ego->desc->genus->vl, &ego->desc->ops, &pln->super.super.ops); if (ego->bufferedp) pln->super.super.ops.other += 2 * n * pln->vl; pln->super.super.could_prune_now_p = !ego->bufferedp; return &(pln->super.super); } /* constructor */ static solver *mksolver(kr2c k, const kr2c_desc *desc, int bufferedp) { static const solver_adt sadt = { PROBLEM_RDFT, mkplan, 0 }; S *slv = MKSOLVER(S, &sadt); slv->k = k; slv->desc = desc; slv->bufferedp = bufferedp; return &(slv->super); } solver *X(mksolver_rdft_r2c_direct)(kr2c k, const kr2c_desc *desc) { return mksolver(k, desc, 0); } solver *X(mksolver_rdft_r2c_directbuf)(kr2c k, const kr2c_desc *desc) { return mksolver(k, desc, 1); } fftw-3.3.3/rdft/generic.c0000644000175000017500000001265412054410050012125 00000000000000/* * Copyright (c) 2003, 2007-11 Matteo Frigo * Copyright (c) 2003, 2007-11 Massachusetts Institute of Technology * * This program is free software; you can redistribute it and/or modify * it under the terms of the GNU General Public License as published by * the Free Software Foundation; either version 2 of the License, or * (at your option) any later version. * * This program is distributed in the hope that it will be useful, * but WITHOUT ANY WARRANTY; without even the implied warranty of * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the * GNU General Public License for more details. * * You should have received a copy of the GNU General Public License * along with this program; if not, write to the Free Software * Foundation, Inc., 51 Franklin Street, Fifth Floor, Boston, MA 02110-1301 USA * */ #include "rdft.h" typedef struct { solver super; rdft_kind kind; } S; typedef struct { plan_rdft super; twid *td; INT n, is, os; rdft_kind kind; } P; /***************************************************************************/ static void cdot_r2hc(INT n, const E *x, const R *w, R *or0, R *oi1) { INT i; E rr = x[0], ri = 0; x += 1; for (i = 1; i + i < n; ++i) { rr += x[0] * w[0]; ri += x[1] * w[1]; x += 2; w += 2; } *or0 = rr; *oi1 = ri; } static void hartley_r2hc(INT n, const R *xr, INT xs, E *o, R *pr) { INT i; E sr; o[0] = sr = xr[0]; o += 1; for (i = 1; i + i < n; ++i) { R a, b; a = xr[i * xs]; b = xr[(n - i) * xs]; sr += (o[0] = a + b); #if FFT_SIGN == -1 o[1] = b - a; #else o[1] = a - b; #endif o += 2; } *pr = sr; } static void apply_r2hc(const plan *ego_, R *I, R *O) { const P *ego = (const P *) ego_; INT i; INT n = ego->n, is = ego->is, os = ego->os; const R *W = ego->td->W; E *buf; size_t bufsz = n * sizeof(E); BUF_ALLOC(E *, buf, bufsz); hartley_r2hc(n, I, is, buf, O); for (i = 1; i + i < n; ++i) { cdot_r2hc(n, buf, W, O + i * os, O + (n - i) * os); W += n - 1; } BUF_FREE(buf, bufsz); } static void cdot_hc2r(INT n, const E *x, const R *w, R *or0, R *or1) { INT i; E rr = x[0], ii = 0; x += 1; for (i = 1; i + i < n; ++i) { rr += x[0] * w[0]; ii += x[1] * w[1]; x += 2; w += 2; } #if FFT_SIGN == -1 *or0 = rr - ii; *or1 = rr + ii; #else *or0 = rr + ii; *or1 = rr - ii; #endif } static void hartley_hc2r(INT n, const R *x, INT xs, E *o, R *pr) { INT i; E sr; o[0] = sr = x[0]; o += 1; for (i = 1; i + i < n; ++i) { sr += (o[0] = x[i * xs] + x[i * xs]); o[1] = x[(n - i) * xs] + x[(n - i) * xs]; o += 2; } *pr = sr; } static void apply_hc2r(const plan *ego_, R *I, R *O) { const P *ego = (const P *) ego_; INT i; INT n = ego->n, is = ego->is, os = ego->os; const R *W = ego->td->W; E *buf; size_t bufsz = n * sizeof(E); BUF_ALLOC(E *, buf, bufsz); hartley_hc2r(n, I, is, buf, O); for (i = 1; i + i < n; ++i) { cdot_hc2r(n, buf, W, O + i * os, O + (n - i) * os); W += n - 1; } BUF_FREE(buf, bufsz); } /***************************************************************************/ static void awake(plan *ego_, enum wakefulness wakefulness) { P *ego = (P *) ego_; static const tw_instr half_tw[] = { { TW_HALF, 1, 0 }, { TW_NEXT, 1, 0 } }; X(twiddle_awake)(wakefulness, &ego->td, half_tw, ego->n, ego->n, (ego->n - 1) / 2); } static void print(const plan *ego_, printer *p) { const P *ego = (const P *) ego_; p->print(p, "(rdft-generic-%s-%D)", ego->kind == R2HC ? "r2hc" : "hc2r", ego->n); } static int applicable(const S *ego, const problem *p_, const planner *plnr) { const problem_rdft *p = (const problem_rdft *) p_; return (1 && p->sz->rnk == 1 && p->vecsz->rnk == 0 && (p->sz->dims[0].n % 2) == 1 && CIMPLIES(NO_LARGE_GENERICP(plnr), p->sz->dims[0].n < GENERIC_MIN_BAD) && CIMPLIES(NO_SLOWP(plnr), p->sz->dims[0].n > GENERIC_MAX_SLOW) && X(is_prime)(p->sz->dims[0].n) && p->kind[0] == ego->kind ); } static plan *mkplan(const solver *ego_, const problem *p_, planner *plnr) { const S *ego = (const S *)ego_; const problem_rdft *p; P *pln; INT n; static const plan_adt padt = { X(rdft_solve), awake, print, X(plan_null_destroy) }; if (!applicable(ego, p_, plnr)) return (plan *)0; p = (const problem_rdft *) p_; pln = MKPLAN_RDFT(P, &padt, R2HC_KINDP(p->kind[0]) ? apply_r2hc : apply_hc2r); pln->n = n = p->sz->dims[0].n; pln->is = p->sz->dims[0].is; pln->os = p->sz->dims[0].os; pln->td = 0; pln->kind = ego->kind; pln->super.super.ops.add = (n-1) * 2.5; pln->super.super.ops.mul = 0; pln->super.super.ops.fma = 0.5 * (n-1) * (n-1) ; #if 0 /* these are nice pipelined sequential loads and should cost nothing */ pln->super.super.ops.other = (n-1)*(2 + 1 + (n-1)); /* approximate */ #endif return &(pln->super.super); } static solver *mksolver(rdft_kind kind) { static const solver_adt sadt = { PROBLEM_RDFT, mkplan, 0 }; S *slv = MKSOLVER(S, &sadt); slv->kind = kind; return &(slv->super); } void X(rdft_generic_register)(planner *p) { REGISTER_SOLVER(p, mksolver(R2HC)); REGISTER_SOLVER(p, mksolver(HC2R)); } fftw-3.3.3/rdft/hc2hc-direct.c0000644000175000017500000001747312054410050012754 00000000000000/* * Copyright (c) 2003, 2007-11 Matteo Frigo * Copyright (c) 2003, 2007-11 Massachusetts Institute of Technology * * This program is free software; you can redistribute it and/or modify * it under the terms of the GNU General Public License as published by * the Free Software Foundation; either version 2 of the License, or * (at your option) any later version. * * This program is distributed in the hope that it will be useful, * but WITHOUT ANY WARRANTY; without even the implied warranty of * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the * GNU General Public License for more details. * * You should have received a copy of the GNU General Public License * along with this program; if not, write to the Free Software * Foundation, Inc., 51 Franklin Street, Fifth Floor, Boston, MA 02110-1301 USA * */ #include "hc2hc.h" typedef struct { hc2hc_solver super; const hc2hc_desc *desc; khc2hc k; int bufferedp; } S; typedef struct { plan_hc2hc super; khc2hc k; plan *cld0, *cldm; /* children for 0th and middle butterflies */ INT r, m, v; INT ms, vs, mb, me; stride rs, brs; twid *td; const S *slv; } P; /************************************************************* Nonbuffered code *************************************************************/ static void apply(const plan *ego_, R *IO) { const P *ego = (const P *) ego_; plan_rdft *cld0 = (plan_rdft *) ego->cld0; plan_rdft *cldm = (plan_rdft *) ego->cldm; INT i, m = ego->m, v = ego->v; INT mb = ego->mb, me = ego->me; INT ms = ego->ms, vs = ego->vs; for (i = 0; i < v; ++i, IO += vs) { cld0->apply((plan *) cld0, IO, IO); ego->k(IO + ms * mb, IO + (m - mb) * ms, ego->td->W, ego->rs, mb, me, ms); cldm->apply((plan *) cldm, IO + (m/2) * ms, IO + (m/2) * ms); } } /************************************************************* Buffered code *************************************************************/ /* should not be 2^k to avoid associativity conflicts */ static INT compute_batchsize(INT radix) { /* round up to multiple of 4 */ radix += 3; radix &= -4; return (radix + 2); } static void dobatch(const P *ego, R *IOp, R *IOm, INT mb, INT me, R *bufp) { INT b = WS(ego->brs, 1); INT rs = WS(ego->rs, 1); INT r = ego->r; INT ms = ego->ms; R *bufm = bufp + b - 1; X(cpy2d_ci)(IOp + mb * ms, bufp, r, rs, b, me - mb, ms, 1, 1); X(cpy2d_ci)(IOm - mb * ms, bufm, r, rs, b, me - mb, -ms, -1, 1); ego->k(bufp, bufm, ego->td->W, ego->brs, mb, me, 1); X(cpy2d_co)(bufp, IOp + mb * ms, r, b, rs, me - mb, 1, ms, 1); X(cpy2d_co)(bufm, IOm - mb * ms, r, b, rs, me - mb, -1, -ms, 1); } static void apply_buf(const plan *ego_, R *IO) { const P *ego = (const P *) ego_; plan_rdft *cld0 = (plan_rdft *) ego->cld0; plan_rdft *cldm = (plan_rdft *) ego->cldm; INT i, j, m = ego->m, v = ego->v, r = ego->r; INT mb = ego->mb, me = ego->me, ms = ego->ms; INT batchsz = compute_batchsize(r); R *buf; size_t bufsz = r * batchsz * 2 * sizeof(R); BUF_ALLOC(R *, buf, bufsz); for (i = 0; i < v; ++i, IO += ego->vs) { R *IOp = IO; R *IOm = IO + m * ms; cld0->apply((plan *) cld0, IO, IO); for (j = mb; j + batchsz < me; j += batchsz) dobatch(ego, IOp, IOm, j, j + batchsz, buf); dobatch(ego, IOp, IOm, j, me, buf); cldm->apply((plan *) cldm, IO + ms * (m/2), IO + ms * (m/2)); } BUF_FREE(buf, bufsz); } static void awake(plan *ego_, enum wakefulness wakefulness) { P *ego = (P *) ego_; X(plan_awake)(ego->cld0, wakefulness); X(plan_awake)(ego->cldm, wakefulness); X(twiddle_awake)(wakefulness, &ego->td, ego->slv->desc->tw, ego->r * ego->m, ego->r, (ego->m - 1) / 2); } static void destroy(plan *ego_) { P *ego = (P *) ego_; X(plan_destroy_internal)(ego->cld0); X(plan_destroy_internal)(ego->cldm); X(stride_destroy)(ego->rs); X(stride_destroy)(ego->brs); } static void print(const plan *ego_, printer *p) { const P *ego = (const P *) ego_; const S *slv = ego->slv; const hc2hc_desc *e = slv->desc; INT batchsz = compute_batchsize(ego->r); if (slv->bufferedp) p->print(p, "(hc2hc-directbuf/%D-%D/%D%v \"%s\"%(%p%)%(%p%))", batchsz, ego->r, X(twiddle_length)(ego->r, e->tw), ego->v, e->nam, ego->cld0, ego->cldm); else p->print(p, "(hc2hc-direct-%D/%D%v \"%s\"%(%p%)%(%p%))", ego->r, X(twiddle_length)(ego->r, e->tw), ego->v, e->nam, ego->cld0, ego->cldm); } static int applicable0(const S *ego, rdft_kind kind, INT r) { const hc2hc_desc *e = ego->desc; return (1 && r == e->radix && kind == e->genus->kind ); } static int applicable(const S *ego, rdft_kind kind, INT r, INT m, INT v, const planner *plnr) { if (!applicable0(ego, kind, r)) return 0; if (NO_UGLYP(plnr) && X(ct_uglyp)((ego->bufferedp? (INT)512 : (INT)16), v, m * r, r)) return 0; return 1; } #define CLDMP(m, mstart, mcount) (2 * ((mstart) + (mcount)) == (m) + 2) #define CLD0P(mstart) ((mstart) == 0) static plan *mkcldw(const hc2hc_solver *ego_, rdft_kind kind, INT r, INT m, INT ms, INT v, INT vs, INT mstart, INT mcount, R *IO, planner *plnr) { const S *ego = (const S *) ego_; P *pln; const hc2hc_desc *e = ego->desc; plan *cld0 = 0, *cldm = 0; INT imid = (m / 2) * ms; INT rs = m * ms; static const plan_adt padt = { 0, awake, print, destroy }; if (!applicable(ego, kind, r, m, v, plnr)) return (plan *)0; cld0 = X(mkplan_d)( plnr, X(mkproblem_rdft_1_d)((CLD0P(mstart) ? X(mktensor_1d)(r, rs, rs) : X(mktensor_0d)()), X(mktensor_0d)(), TAINT(IO, vs), TAINT(IO, vs), kind)); if (!cld0) goto nada; cldm = X(mkplan_d)( plnr, X(mkproblem_rdft_1_d)((CLDMP(m, mstart, mcount) ? X(mktensor_1d)(r, rs, rs) : X(mktensor_0d)()), X(mktensor_0d)(), TAINT(IO + imid, vs), TAINT(IO + imid, vs), kind == R2HC ? R2HCII : HC2RIII)); if (!cldm) goto nada; pln = MKPLAN_HC2HC(P, &padt, ego->bufferedp ? apply_buf : apply); pln->k = ego->k; pln->td = 0; pln->r = r; pln->rs = X(mkstride)(r, rs); pln->m = m; pln->ms = ms; pln->v = v; pln->vs = vs; pln->slv = ego; pln->brs = X(mkstride)(r, 2 * compute_batchsize(r)); pln->cld0 = cld0; pln->cldm = cldm; pln->mb = mstart + CLD0P(mstart); pln->me = mstart + mcount - CLDMP(m, mstart, mcount); X(ops_zero)(&pln->super.super.ops); X(ops_madd2)(v * ((pln->me - pln->mb) / e->genus->vl), &e->ops, &pln->super.super.ops); X(ops_madd2)(v, &cld0->ops, &pln->super.super.ops); X(ops_madd2)(v, &cldm->ops, &pln->super.super.ops); if (ego->bufferedp) pln->super.super.ops.other += 4 * r * (pln->me - pln->mb) * v; pln->super.super.could_prune_now_p = (!ego->bufferedp && r >= 5 && r < 64 && m >= r); return &(pln->super.super); nada: X(plan_destroy_internal)(cld0); X(plan_destroy_internal)(cldm); return 0; } static void regone(planner *plnr, khc2hc codelet, const hc2hc_desc *desc, int bufferedp) { S *slv = (S *)X(mksolver_hc2hc)(sizeof(S), desc->radix, mkcldw); slv->k = codelet; slv->desc = desc; slv->bufferedp = bufferedp; REGISTER_SOLVER(plnr, &(slv->super.super)); if (X(mksolver_hc2hc_hook)) { slv = (S *)X(mksolver_hc2hc_hook)(sizeof(S), desc->radix, mkcldw); slv->k = codelet; slv->desc = desc; slv->bufferedp = bufferedp; REGISTER_SOLVER(plnr, &(slv->super.super)); } } void X(regsolver_hc2hc_direct)(planner *plnr, khc2hc codelet, const hc2hc_desc *desc) { regone(plnr, codelet, desc, /* bufferedp */0); regone(plnr, codelet, desc, /* bufferedp */1); } fftw-3.3.3/rdft/hc2hc-generic.c0000644000175000017500000002045312054410050013106 00000000000000/* * Copyright (c) 2003, 2007-11 Matteo Frigo * Copyright (c) 2003, 2007-11 Massachusetts Institute of Technology * * This program is free software; you can redistribute it and/or modify * it under the terms of the GNU General Public License as published by * the Free Software Foundation; either version 2 of the License, or * (at your option) any later version. * * This program is distributed in the hope that it will be useful, * but WITHOUT ANY WARRANTY; without even the implied warranty of * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the * GNU General Public License for more details. * * You should have received a copy of the GNU General Public License * along with this program; if not, write to the Free Software * Foundation, Inc., 51 Franklin Street, Fifth Floor, Boston, MA 02110-1301 USA * */ /* express a hc2hc problem in terms of rdft + multiplication by twiddle factors */ #include "hc2hc.h" typedef hc2hc_solver S; typedef struct { plan_hc2hc super; INT r, m, s, vl, vs, mstart1, mcount1; plan *cld0; plan *cld; twid *td; } P; /**************************************************************/ static void mktwiddle(P *ego, enum wakefulness wakefulness) { static const tw_instr tw[] = { { TW_HALF, 0, 0 }, { TW_NEXT, 1, 0 } }; /* note that R and M are swapped, to allow for sequential access both to data and twiddles */ X(twiddle_awake)(wakefulness, &ego->td, tw, ego->r * ego->m, ego->m, ego->r); } static void bytwiddle(const P *ego, R *IO, R sign) { INT i, j, k; INT r = ego->r, m = ego->m, s = ego->s, vl = ego->vl, vs = ego->vs; INT ms = m * s; INT mstart1 = ego->mstart1, mcount1 = ego->mcount1; INT wrem = 2 * ((m-1)/2 - mcount1); for (i = 0; i < vl; ++i, IO += vs) { const R *W = ego->td->W; A(m % 2 == 1); for (k = 1, W += (m - 1) + 2*(mstart1-1); k < r; ++k) { /* pr := IO + (j + mstart1) * s + k * ms */ R *pr = IO + mstart1 * s + k * ms; /* pi := IO + (m - j - mstart1) * s + k * ms */ R *pi = IO - mstart1 * s + (k + 1) * ms; for (j = 0; j < mcount1; ++j, pr += s, pi -= s) { E xr = *pr; E xi = *pi; E wr = W[0]; E wi = sign * W[1]; *pr = xr * wr - xi * wi; *pi = xi * wr + xr * wi; W += 2; } W += wrem; } } } static void swapri(R *IO, INT r, INT m, INT s, INT jstart, INT jend) { INT k; INT ms = m * s; INT js = jstart * s; for (k = 0; k + k < r; ++k) { /* pr := IO + (m - j) * s + k * ms */ R *pr = IO + (k + 1) * ms - js; /* pi := IO + (m - j) * s + (r - 1 - k) * ms */ R *pi = IO + (r - k) * ms - js; INT j; for (j = jstart; j < jend; j += 1, pr -= s, pi -= s) { R t = *pr; *pr = *pi; *pi = t; } } } static void reorder_dit(const P *ego, R *IO) { INT i, k; INT r = ego->r, m = ego->m, s = ego->s, vl = ego->vl, vs = ego->vs; INT ms = m * s; INT mstart1 = ego->mstart1, mend1 = mstart1 + ego->mcount1; for (i = 0; i < vl; ++i, IO += vs) { for (k = 1; k + k < r; ++k) { R *p0 = IO + k * ms; R *p1 = IO + (r - k) * ms; INT j; for (j = mstart1; j < mend1; ++j) { E rp, ip, im, rm; rp = p0[j * s]; im = p1[ms - j * s]; rm = p1[j * s]; ip = p0[ms - j * s]; p0[j * s] = rp - im; p1[ms - j * s] = rp + im; p1[j * s] = rm - ip; p0[ms - j * s] = ip + rm; } } swapri(IO, r, m, s, mstart1, mend1); } } static void reorder_dif(const P *ego, R *IO) { INT i, k; INT r = ego->r, m = ego->m, s = ego->s, vl = ego->vl, vs = ego->vs; INT ms = m * s; INT mstart1 = ego->mstart1, mend1 = mstart1 + ego->mcount1; for (i = 0; i < vl; ++i, IO += vs) { swapri(IO, r, m, s, mstart1, mend1); for (k = 1; k + k < r; ++k) { R *p0 = IO + k * ms; R *p1 = IO + (r - k) * ms; const R half = K(0.5); INT j; for (j = mstart1; j < mend1; ++j) { E rp, ip, im, rm; rp = half * p0[j * s]; im = half * p1[ms - j * s]; rm = half * p1[j * s]; ip = half * p0[ms - j * s]; p0[j * s] = rp + im; p1[ms - j * s] = im - rp; p1[j * s] = rm + ip; p0[ms - j * s] = ip - rm; } } } } static int applicable(rdft_kind kind, INT r, INT m, const planner *plnr) { return (1 && (kind == R2HC || kind == HC2R) && (m % 2) && (r % 2) && !NO_SLOWP(plnr) ); } /**************************************************************/ static void apply_dit(const plan *ego_, R *IO) { const P *ego = (const P *) ego_; INT start; plan_rdft *cld, *cld0; bytwiddle(ego, IO, K(-1.0)); cld0 = (plan_rdft *) ego->cld0; cld0->apply(ego->cld0, IO, IO); start = ego->mstart1 * ego->s; cld = (plan_rdft *) ego->cld; cld->apply(ego->cld, IO + start, IO + start); reorder_dit(ego, IO); } static void apply_dif(const plan *ego_, R *IO) { const P *ego = (const P *) ego_; INT start; plan_rdft *cld, *cld0; reorder_dif(ego, IO); cld0 = (plan_rdft *) ego->cld0; cld0->apply(ego->cld0, IO, IO); start = ego->mstart1 * ego->s; cld = (plan_rdft *) ego->cld; cld->apply(ego->cld, IO + start, IO + start); bytwiddle(ego, IO, K(1.0)); } static void awake(plan *ego_, enum wakefulness wakefulness) { P *ego = (P *) ego_; X(plan_awake)(ego->cld0, wakefulness); X(plan_awake)(ego->cld, wakefulness); mktwiddle(ego, wakefulness); } static void destroy(plan *ego_) { P *ego = (P *) ego_; X(plan_destroy_internal)(ego->cld); X(plan_destroy_internal)(ego->cld0); } static void print(const plan *ego_, printer *p) { const P *ego = (const P *) ego_; p->print(p, "(hc2hc-generic-%s-%D-%D%v%(%p%)%(%p%))", ego->super.apply == apply_dit ? "dit" : "dif", ego->r, ego->m, ego->vl, ego->cld0, ego->cld); } static plan *mkcldw(const hc2hc_solver *ego_, rdft_kind kind, INT r, INT m, INT s, INT vl, INT vs, INT mstart, INT mcount, R *IO, planner *plnr) { P *pln; plan *cld0 = 0, *cld = 0; INT mstart1, mcount1, mstride; static const plan_adt padt = { 0, awake, print, destroy }; UNUSED(ego_); A(mstart >= 0 && mcount > 0 && mstart + mcount <= (m+2)/2); if (!applicable(kind, r, m, plnr)) return (plan *)0; A(m % 2); mstart1 = mstart + (mstart == 0); mcount1 = mcount - (mstart == 0); mstride = m - (mstart + mcount - 1) - mstart1; /* 0th (DC) transform (vl of these), if mstart == 0 */ cld0 = X(mkplan_d)(plnr, X(mkproblem_rdft_1_d)( mstart == 0 ? X(mktensor_1d)(r, m * s, m * s) : X(mktensor_0d)(), X(mktensor_1d)(vl, vs, vs), IO, IO, kind) ); if (!cld0) goto nada; /* twiddle transforms: there are 2 x mcount1 x vl of these (where 2 corresponds to the real and imaginary parts) ... the 2 x mcount1 loops are combined if mstart=0 and mcount=(m+2)/2. */ cld = X(mkplan_d)(plnr, X(mkproblem_rdft_1_d)( X(mktensor_1d)(r, m * s, m * s), X(mktensor_3d)(2, mstride * s, mstride * s, mcount1, s, s, vl, vs, vs), IO + s * mstart1, IO + s * mstart1, kind) ); if (!cld) goto nada; pln = MKPLAN_HC2HC(P, &padt, (kind == R2HC) ? apply_dit : apply_dif); pln->cld = cld; pln->cld0 = cld0; pln->r = r; pln->m = m; pln->s = s; pln->vl = vl; pln->vs = vs; pln->td = 0; pln->mstart1 = mstart1; pln->mcount1 = mcount1; { double n0 = 0.5 * (r - 1) * (2 * mcount1) * vl; pln->super.super.ops = cld->ops; pln->super.super.ops.mul += (kind == R2HC ? 5.0 : 7.0) * n0; pln->super.super.ops.add += 4.0 * n0; pln->super.super.ops.other += 11.0 * n0; } return &(pln->super.super); nada: X(plan_destroy_internal)(cld); X(plan_destroy_internal)(cld0); return (plan *) 0; } static void regsolver(planner *plnr, INT r) { S *slv = (S *)X(mksolver_hc2hc)(sizeof(S), r, mkcldw); REGISTER_SOLVER(plnr, &(slv->super)); if (X(mksolver_hc2hc_hook)) { slv = (S *)X(mksolver_hc2hc_hook)(sizeof(S), r, mkcldw); REGISTER_SOLVER(plnr, &(slv->super)); } } void X(hc2hc_generic_register)(planner *p) { regsolver(p, 0); } fftw-3.3.3/rdft/khc2hc.c0000644000175000017500000000174412054410050011651 00000000000000/* * Copyright (c) 2003, 2007-11 Matteo Frigo * Copyright (c) 2003, 2007-11 Massachusetts Institute of Technology * * This program is free software; you can redistribute it and/or modify * it under the terms of the GNU General Public License as published by * the Free Software Foundation; either version 2 of the License, or * (at your option) any later version. * * This program is distributed in the hope that it will be useful, * but WITHOUT ANY WARRANTY; without even the implied warranty of * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the * GNU General Public License for more details. * * You should have received a copy of the GNU General Public License * along with this program; if not, write to the Free Software * Foundation, Inc., 51 Franklin Street, Fifth Floor, Boston, MA 02110-1301 USA * */ #include "hc2hc.h" void X(khc2hc_register)(planner *p, khc2hc codelet, const hc2hc_desc *desc) { X(regsolver_hc2hc_direct)(p, codelet, desc); } fftw-3.3.3/rdft/kr2c.c0000644000175000017500000000217312054410050011345 00000000000000/* * Copyright (c) 2003, 2007-11 Matteo Frigo * Copyright (c) 2003, 2007-11 Massachusetts Institute of Technology * * This program is free software; you can redistribute it and/or modify * it under the terms of the GNU General Public License as published by * the Free Software Foundation; either version 2 of the License, or * (at your option) any later version. * * This program is distributed in the hope that it will be useful, * but WITHOUT ANY WARRANTY; without even the implied warranty of * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the * GNU General Public License for more details. * * You should have received a copy of the GNU General Public License * along with this program; if not, write to the Free Software * Foundation, Inc., 51 Franklin Street, Fifth Floor, Boston, MA 02110-1301 USA * */ #include "rdft.h" void X(kr2c_register)(planner *p, kr2c codelet, const kr2c_desc *desc) { REGISTER_SOLVER(p, X(mksolver_rdft_r2c_direct)(codelet, desc)); REGISTER_SOLVER(p, X(mksolver_rdft_r2c_directbuf)(codelet, desc)); REGISTER_SOLVER(p, X(mksolver_rdft2_direct)(codelet, desc)); } fftw-3.3.3/rdft/kr2r.c0000644000175000017500000000176112054410050011366 00000000000000/* * Copyright (c) 2003, 2007-11 Matteo Frigo * Copyright (c) 2003, 2007-11 Massachusetts Institute of Technology * * This program is free software; you can redistribute it and/or modify * it under the terms of the GNU General Public License as published by * the Free Software Foundation; either version 2 of the License, or * (at your option) any later version. * * This program is distributed in the hope that it will be useful, * but WITHOUT ANY WARRANTY; without even the implied warranty of * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the * GNU General Public License for more details. * * You should have received a copy of the GNU General Public License * along with this program; if not, write to the Free Software * Foundation, Inc., 51 Franklin Street, Fifth Floor, Boston, MA 02110-1301 USA * */ #include "rdft.h" void X(kr2r_register)(planner *p, kr2r codelet, const kr2r_desc *desc) { REGISTER_SOLVER(p, X(mksolver_rdft_r2r_direct)(codelet, desc)); } fftw-3.3.3/rdft/indirect.c0000644000175000017500000001447612054410050012316 00000000000000/* * Copyright (c) 2003, 2007-11 Matteo Frigo * Copyright (c) 2003, 2007-11 Massachusetts Institute of Technology * * This program is free software; you can redistribute it and/or modify * it under the terms of the GNU General Public License as published by * the Free Software Foundation; either version 2 of the License, or * (at your option) any later version. * * This program is distributed in the hope that it will be useful, * but WITHOUT ANY WARRANTY; without even the implied warranty of * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the * GNU General Public License for more details. * * You should have received a copy of the GNU General Public License * along with this program; if not, write to the Free Software * Foundation, Inc., 51 Franklin Street, Fifth Floor, Boston, MA 02110-1301 USA * */ /* solvers/plans for vectors of small RDFT's that cannot be done in-place directly. Use a rank-0 plan to rearrange the data before or after the transform. Can also change an out-of-place plan into a copy + in-place (where the in-place transform is e.g. unit stride). */ /* FIXME: merge with rank-geq2.c(?), since this is just a special case of a rank split where the first/second transform has rank 0. */ #include "rdft.h" typedef problem *(*mkcld_t) (const problem_rdft *p); typedef struct { rdftapply apply; problem *(*mkcld)(const problem_rdft *p); const char *nam; } ndrct_adt; typedef struct { solver super; const ndrct_adt *adt; } S; typedef struct { plan_rdft super; plan *cldcpy, *cld; const S *slv; } P; /*-----------------------------------------------------------------------*/ /* first rearrange, then transform */ static void apply_before(const plan *ego_, R *I, R *O) { const P *ego = (const P *) ego_; { plan_rdft *cldcpy = (plan_rdft *) ego->cldcpy; cldcpy->apply(ego->cldcpy, I, O); } { plan_rdft *cld = (plan_rdft *) ego->cld; cld->apply(ego->cld, O, O); } } static problem *mkcld_before(const problem_rdft *p) { return X(mkproblem_rdft_d)(X(tensor_copy_inplace)(p->sz, INPLACE_OS), X(tensor_copy_inplace)(p->vecsz, INPLACE_OS), p->O, p->O, p->kind); } static const ndrct_adt adt_before = { apply_before, mkcld_before, "rdft-indirect-before" }; /*-----------------------------------------------------------------------*/ /* first transform, then rearrange */ static void apply_after(const plan *ego_, R *I, R *O) { const P *ego = (const P *) ego_; { plan_rdft *cld = (plan_rdft *) ego->cld; cld->apply(ego->cld, I, I); } { plan_rdft *cldcpy = (plan_rdft *) ego->cldcpy; cldcpy->apply(ego->cldcpy, I, O); } } static problem *mkcld_after(const problem_rdft *p) { return X(mkproblem_rdft_d)(X(tensor_copy_inplace)(p->sz, INPLACE_IS), X(tensor_copy_inplace)(p->vecsz, INPLACE_IS), p->I, p->I, p->kind); } static const ndrct_adt adt_after = { apply_after, mkcld_after, "rdft-indirect-after" }; /*-----------------------------------------------------------------------*/ static void destroy(plan *ego_) { P *ego = (P *) ego_; X(plan_destroy_internal)(ego->cld); X(plan_destroy_internal)(ego->cldcpy); } static void awake(plan *ego_, enum wakefulness wakefulness) { P *ego = (P *) ego_; X(plan_awake)(ego->cldcpy, wakefulness); X(plan_awake)(ego->cld, wakefulness); } static void print(const plan *ego_, printer *p) { const P *ego = (const P *) ego_; const S *s = ego->slv; p->print(p, "(%s%(%p%)%(%p%))", s->adt->nam, ego->cld, ego->cldcpy); } static int applicable0(const solver *ego_, const problem *p_, const planner *plnr) { const S *ego = (const S *) ego_; const problem_rdft *p = (const problem_rdft *) p_; return (1 && FINITE_RNK(p->vecsz->rnk) /* problem must be a nontrivial transform, not just a copy */ && p->sz->rnk > 0 && (0 /* problem must be in-place & require some rearrangement of the data */ || (p->I == p->O && !(X(tensor_inplace_strides2)(p->sz, p->vecsz))) /* or problem must be out of place, transforming from stride 1/2 to bigger stride, for apply_after */ || (p->I != p->O && ego->adt->apply == apply_after && !NO_DESTROY_INPUTP(plnr) && X(tensor_min_istride)(p->sz) <= 2 && X(tensor_min_ostride)(p->sz) > 2) /* or problem must be out of place, transforming to stride 1/2 from bigger stride, for apply_before */ || (p->I != p->O && ego->adt->apply == apply_before && X(tensor_min_ostride)(p->sz) <= 2 && X(tensor_min_istride)(p->sz) > 2) ) ); } static int applicable(const solver *ego_, const problem *p_, const planner *plnr) { if (!applicable0(ego_, p_, plnr)) return 0; if (NO_INDIRECT_OP_P(plnr)) { const problem_rdft *p = (const problem_rdft *)p_; if (p->I != p->O) return 0; } return 1; } static plan *mkplan(const solver *ego_, const problem *p_, planner *plnr) { const problem_rdft *p = (const problem_rdft *) p_; const S *ego = (const S *) ego_; P *pln; plan *cld = 0, *cldcpy = 0; static const plan_adt padt = { X(rdft_solve), awake, print, destroy }; if (!applicable(ego_, p_, plnr)) return (plan *) 0; cldcpy = X(mkplan_d)(plnr, X(mkproblem_rdft_0_d)( X(tensor_append)(p->vecsz, p->sz), p->I, p->O)); if (!cldcpy) goto nada; cld = X(mkplan_f_d)(plnr, ego->adt->mkcld(p), NO_BUFFERING, 0, 0); if (!cld) goto nada; pln = MKPLAN_RDFT(P, &padt, ego->adt->apply); pln->cld = cld; pln->cldcpy = cldcpy; pln->slv = ego; X(ops_add)(&cld->ops, &cldcpy->ops, &pln->super.super.ops); return &(pln->super.super); nada: X(plan_destroy_internal)(cld); X(plan_destroy_internal)(cldcpy); return (plan *)0; } static solver *mksolver(const ndrct_adt *adt) { static const solver_adt sadt = { PROBLEM_RDFT, mkplan, 0 }; S *slv = MKSOLVER(S, &sadt); slv->adt = adt; return &(slv->super); } void X(rdft_indirect_register)(planner *p) { unsigned i; static const ndrct_adt *const adts[] = { &adt_before, &adt_after }; for (i = 0; i < sizeof(adts) / sizeof(adts[0]); ++i) REGISTER_SOLVER(p, mksolver(adts[i])); } fftw-3.3.3/rdft/nop.c0000644000175000017500000000415112054410050011276 00000000000000/* * Copyright (c) 2003, 2007-11 Matteo Frigo * Copyright (c) 2003, 2007-11 Massachusetts Institute of Technology * * This program is free software; you can redistribute it and/or modify * it under the terms of the GNU General Public License as published by * the Free Software Foundation; either version 2 of the License, or * (at your option) any later version. * * This program is distributed in the hope that it will be useful, * but WITHOUT ANY WARRANTY; without even the implied warranty of * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the * GNU General Public License for more details. * * You should have received a copy of the GNU General Public License * along with this program; if not, write to the Free Software * Foundation, Inc., 51 Franklin Street, Fifth Floor, Boston, MA 02110-1301 USA * */ /* plans for vrank -infty RDFTs (nothing to do) */ #include "rdft.h" static void apply(const plan *ego_, R *I, R *O) { UNUSED(ego_); UNUSED(I); UNUSED(O); } static int applicable(const solver *ego_, const problem *p_) { const problem_rdft *p = (const problem_rdft *) p_; UNUSED(ego_); return 0 /* case 1 : -infty vector rank */ || (p->vecsz->rnk == RNK_MINFTY) /* case 2 : rank-0 in-place rdft */ || (1 && p->sz->rnk == 0 && FINITE_RNK(p->vecsz->rnk) && p->O == p->I && X(tensor_inplace_strides)(p->vecsz) ); } static void print(const plan *ego, printer *p) { UNUSED(ego); p->print(p, "(rdft-nop)"); } static plan *mkplan(const solver *ego, const problem *p, planner *plnr) { static const plan_adt padt = { X(rdft_solve), X(null_awake), print, X(plan_null_destroy) }; plan_rdft *pln; UNUSED(plnr); if (!applicable(ego, p)) return (plan *) 0; pln = MKPLAN_RDFT(plan_rdft, &padt, apply); X(ops_zero)(&pln->super.ops); return &(pln->super); } static solver *mksolver(void) { static const solver_adt sadt = { PROBLEM_RDFT, mkplan, 0 }; return MKSOLVER(solver, &sadt); } void X(rdft_nop_register)(planner *p) { REGISTER_SOLVER(p, mksolver()); } fftw-3.3.3/rdft/plan.c0000644000175000017500000000204712054410050011436 00000000000000/* * Copyright (c) 2003, 2007-11 Matteo Frigo * Copyright (c) 2003, 2007-11 Massachusetts Institute of Technology * * This program is free software; you can redistribute it and/or modify * it under the terms of the GNU General Public License as published by * the Free Software Foundation; either version 2 of the License, or * (at your option) any later version. * * This program is distributed in the hope that it will be useful, * but WITHOUT ANY WARRANTY; without even the implied warranty of * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the * GNU General Public License for more details. * * You should have received a copy of the GNU General Public License * along with this program; if not, write to the Free Software * Foundation, Inc., 51 Franklin Street, Fifth Floor, Boston, MA 02110-1301 USA * */ #include "rdft.h" plan *X(mkplan_rdft)(size_t size, const plan_adt *adt, rdftapply apply) { plan_rdft *ego; ego = (plan_rdft *) X(mkplan)(size, adt); ego->apply = apply; return &(ego->super); } fftw-3.3.3/rdft/problem.c0000644000175000017500000001516612054410050012152 00000000000000/* * Copyright (c) 2003, 2007-11 Matteo Frigo * Copyright (c) 2003, 2007-11 Massachusetts Institute of Technology * * This program is free software; you can redistribute it and/or modify * it under the terms of the GNU General Public License as published by * the Free Software Foundation; either version 2 of the License, or * (at your option) any later version. * * This program is distributed in the hope that it will be useful, * but WITHOUT ANY WARRANTY; without even the implied warranty of * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the * GNU General Public License for more details. * * You should have received a copy of the GNU General Public License * along with this program; if not, write to the Free Software * Foundation, Inc., 51 Franklin Street, Fifth Floor, Boston, MA 02110-1301 USA * */ #include "rdft.h" #include static void destroy(problem *ego_) { problem_rdft *ego = (problem_rdft *) ego_; #if !defined(STRUCT_HACK_C99) && !defined(STRUCT_HACK_KR) X(ifree0)(ego->kind); #endif X(tensor_destroy2)(ego->vecsz, ego->sz); X(ifree)(ego_); } static void kind_hash(md5 *m, const rdft_kind *kind, int rnk) { int i; for (i = 0; i < rnk; ++i) X(md5int)(m, kind[i]); } static void hash(const problem *p_, md5 *m) { const problem_rdft *p = (const problem_rdft *) p_; X(md5puts)(m, "rdft"); X(md5int)(m, p->I == p->O); kind_hash(m, p->kind, p->sz->rnk); X(md5int)(m, X(alignment_of)(p->I)); X(md5int)(m, X(alignment_of)(p->O)); X(tensor_md5)(m, p->sz); X(tensor_md5)(m, p->vecsz); } static void recur(const iodim *dims, int rnk, R *I) { if (rnk == RNK_MINFTY) return; else if (rnk == 0) I[0] = K(0.0); else if (rnk > 0) { INT i, n = dims[0].n, is = dims[0].is; if (rnk == 1) { /* this case is redundant but faster */ for (i = 0; i < n; ++i) I[i * is] = K(0.0); } else { for (i = 0; i < n; ++i) recur(dims + 1, rnk - 1, I + i * is); } } } void X(rdft_zerotens)(tensor *sz, R *I) { recur(sz->dims, sz->rnk, I); } #define KSTR_LEN 8 const char *X(rdft_kind_str)(rdft_kind kind) { static const char kstr[][KSTR_LEN] = { "r2hc", "r2hc01", "r2hc10", "r2hc11", "hc2r", "hc2r01", "hc2r10", "hc2r11", "dht", "redft00", "redft01", "redft10", "redft11", "rodft00", "rodft01", "rodft10", "rodft11" }; A(kind >= 0 && kind < sizeof(kstr) / KSTR_LEN); return kstr[kind]; } static void print(const problem *ego_, printer *p) { const problem_rdft *ego = (const problem_rdft *) ego_; int i; p->print(p, "(rdft %d %D %T %T", X(alignment_of)(ego->I), (INT)(ego->O - ego->I), ego->sz, ego->vecsz); for (i = 0; i < ego->sz->rnk; ++i) p->print(p, " %d", (int)ego->kind[i]); p->print(p, ")"); } static void zero(const problem *ego_) { const problem_rdft *ego = (const problem_rdft *) ego_; tensor *sz = X(tensor_append)(ego->vecsz, ego->sz); X(rdft_zerotens)(sz, UNTAINT(ego->I)); X(tensor_destroy)(sz); } static const problem_adt padt = { PROBLEM_RDFT, hash, zero, print, destroy }; /* Dimensions of size 1 that are not REDFT/RODFT are no-ops and can be eliminated. REDFT/RODFT unit dimensions often have factors of 2.0 and suchlike from normalization and phases, although in principle these constant factors from different dimensions could be combined. */ static int nontrivial(const iodim *d, rdft_kind kind) { return (d->n > 1 || kind == R2HC11 || kind == HC2R11 || (REODFT_KINDP(kind) && kind != REDFT01 && kind != RODFT01)); } problem *X(mkproblem_rdft)(const tensor *sz, const tensor *vecsz, R *I, R *O, const rdft_kind *kind) { problem_rdft *ego; int rnk = sz->rnk; int i; A(X(tensor_kosherp)(sz)); A(X(tensor_kosherp)(vecsz)); A(FINITE_RNK(sz->rnk)); if (UNTAINT(I) == UNTAINT(O)) I = O = JOIN_TAINT(I, O); if (I == O && !X(tensor_inplace_locations)(sz, vecsz)) return X(mkproblem_unsolvable)(); for (i = rnk = 0; i < sz->rnk; ++i) { A(sz->dims[i].n > 0); if (nontrivial(sz->dims + i, kind[i])) ++rnk; } #if defined(STRUCT_HACK_KR) ego = (problem_rdft *) X(mkproblem)(sizeof(problem_rdft) + sizeof(rdft_kind) * (rnk > 0 ? rnk - 1 : 0), &padt); #elif defined(STRUCT_HACK_C99) ego = (problem_rdft *) X(mkproblem)(sizeof(problem_rdft) + sizeof(rdft_kind) * rnk, &padt); #else ego = (problem_rdft *) X(mkproblem)(sizeof(problem_rdft), &padt); ego->kind = (rdft_kind *) MALLOC(sizeof(rdft_kind) * rnk, PROBLEMS); #endif /* do compression and sorting as in X(tensor_compress), but take transform kind into account (sigh) */ ego->sz = X(mktensor)(rnk); for (i = rnk = 0; i < sz->rnk; ++i) { if (nontrivial(sz->dims + i, kind[i])) { ego->kind[rnk] = kind[i]; ego->sz->dims[rnk++] = sz->dims[i]; } } for (i = 0; i + 1 < rnk; ++i) { int j; for (j = i + 1; j < rnk; ++j) if (X(dimcmp)(ego->sz->dims + i, ego->sz->dims + j) > 0) { iodim dswap; rdft_kind kswap; dswap = ego->sz->dims[i]; ego->sz->dims[i] = ego->sz->dims[j]; ego->sz->dims[j] = dswap; kswap = ego->kind[i]; ego->kind[i] = ego->kind[j]; ego->kind[j] = kswap; } } for (i = 0; i < rnk; ++i) if (ego->sz->dims[i].n == 2 && (ego->kind[i] == REDFT00 || ego->kind[i] == DHT || ego->kind[i] == HC2R)) ego->kind[i] = R2HC; /* size-2 transforms are equivalent */ ego->vecsz = X(tensor_compress_contiguous)(vecsz); ego->I = I; ego->O = O; A(FINITE_RNK(ego->sz->rnk)); return &(ego->super); } /* Same as X(mkproblem_rdft), but also destroy input tensors. */ problem *X(mkproblem_rdft_d)(tensor *sz, tensor *vecsz, R *I, R *O, const rdft_kind *kind) { problem *p = X(mkproblem_rdft)(sz, vecsz, I, O, kind); X(tensor_destroy2)(vecsz, sz); return p; } /* As above, but for rnk <= 1 only and takes a scalar kind parameter */ problem *X(mkproblem_rdft_1)(const tensor *sz, const tensor *vecsz, R *I, R *O, rdft_kind kind) { A(sz->rnk <= 1); return X(mkproblem_rdft)(sz, vecsz, I, O, &kind); } problem *X(mkproblem_rdft_1_d)(tensor *sz, tensor *vecsz, R *I, R *O, rdft_kind kind) { A(sz->rnk <= 1); return X(mkproblem_rdft_d)(sz, vecsz, I, O, &kind); } /* create a zero-dimensional problem */ problem *X(mkproblem_rdft_0_d)(tensor *vecsz, R *I, R *O) { return X(mkproblem_rdft_d)(X(mktensor_0d)(), vecsz, I, O, (const rdft_kind *)0); } fftw-3.3.3/rdft/rank0.c0000644000175000017500000002420312054410050011515 00000000000000/* * Copyright (c) 2003, 2007-11 Matteo Frigo * Copyright (c) 2003, 2007-11 Massachusetts Institute of Technology * * This program is free software; you can redistribute it and/or modify * it under the terms of the GNU General Public License as published by * the Free Software Foundation; either version 2 of the License, or * (at your option) any later version. * * This program is distributed in the hope that it will be useful, * but WITHOUT ANY WARRANTY; without even the implied warranty of * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the * GNU General Public License for more details. * * You should have received a copy of the GNU General Public License * along with this program; if not, write to the Free Software * Foundation, Inc., 51 Franklin Street, Fifth Floor, Boston, MA 02110-1301 USA * */ /* plans for rank-0 RDFTs (copy operations) */ #include "rdft.h" #ifdef HAVE_STRING_H #include /* for memcpy() */ #endif #define MAXRNK 32 /* FIXME: should malloc() */ typedef struct { plan_rdft super; INT vl; int rnk; iodim d[MAXRNK]; const char *nam; } P; typedef struct { solver super; rdftapply apply; int (*applicable)(const P *pln, const problem_rdft *p); const char *nam; } S; /* copy up to MAXRNK dimensions from problem into plan. If a contiguous dimension exists, save its length in pln->vl */ static int fill_iodim(P *pln, const problem_rdft *p) { int i; const tensor *vecsz = p->vecsz; pln->vl = 1; pln->rnk = 0; for (i = 0; i < vecsz->rnk; ++i) { /* extract contiguous dimensions */ if (pln->vl == 1 && vecsz->dims[i].is == 1 && vecsz->dims[i].os == 1) pln->vl = vecsz->dims[i].n; else if (pln->rnk == MAXRNK) return 0; else pln->d[pln->rnk++] = vecsz->dims[i]; } return 1; } /* generic higher-rank copy routine, calls cpy2d() to do the real work */ static void copy(const iodim *d, int rnk, INT vl, R *I, R *O, cpy2d_func cpy2d) { A(rnk >= 2); if (rnk == 2) cpy2d(I, O, d[0].n, d[0].is, d[0].os, d[1].n, d[1].is, d[1].os, vl); else { INT i; for (i = 0; i < d[0].n; ++i, I += d[0].is, O += d[0].os) copy(d + 1, rnk - 1, vl, I, O, cpy2d); } } /* FIXME: should be more general */ static int transposep(const P *pln) { int i; for (i = 0; i < pln->rnk - 2; ++i) if (pln->d[i].is != pln->d[i].os) return 0; return (pln->d[i].n == pln->d[i+1].n && pln->d[i].is == pln->d[i+1].os && pln->d[i].os == pln->d[i+1].is); } /* generic higher-rank transpose routine, calls transpose2d() to do * the real work */ static void transpose(const iodim *d, int rnk, INT vl, R *I, transpose_func transpose2d) { A(rnk >= 2); if (rnk == 2) transpose2d(I, d[0].n, d[0].is, d[0].os, vl); else { INT i; for (i = 0; i < d[0].n; ++i, I += d[0].is) transpose(d + 1, rnk - 1, vl, I, transpose2d); } } /**************************************************************/ /* rank 0,1,2, out of place, iterative */ static void apply_iter(const plan *ego_, R *I, R *O) { const P *ego = (const P *) ego_; switch (ego->rnk) { case 0: X(cpy1d)(I, O, ego->vl, 1, 1, 1); break; case 1: X(cpy1d)(I, O, ego->d[0].n, ego->d[0].is, ego->d[0].os, ego->vl); break; default: copy(ego->d, ego->rnk, ego->vl, I, O, X(cpy2d_ci)); break; } } static int applicable_iter(const P *pln, const problem_rdft *p) { UNUSED(pln); return (p->I != p->O); } /**************************************************************/ /* out of place, write contiguous output */ static void apply_cpy2dco(const plan *ego_, R *I, R *O) { const P *ego = (const P *) ego_; copy(ego->d, ego->rnk, ego->vl, I, O, X(cpy2d_co)); } static int applicable_cpy2dco(const P *pln, const problem_rdft *p) { int rnk = pln->rnk; return (1 && p->I != p->O && rnk >= 2 /* must not duplicate apply_iter */ && (X(iabs)(pln->d[rnk - 2].is) <= X(iabs)(pln->d[rnk - 1].is) || X(iabs)(pln->d[rnk - 2].os) <= X(iabs)(pln->d[rnk - 1].os)) ); } /**************************************************************/ /* out of place, tiled, no buffering */ static void apply_tiled(const plan *ego_, R *I, R *O) { const P *ego = (const P *) ego_; copy(ego->d, ego->rnk, ego->vl, I, O, X(cpy2d_tiled)); } static int applicable_tiled(const P *pln, const problem_rdft *p) { return (1 && p->I != p->O && pln->rnk >= 2 /* somewhat arbitrary */ && X(compute_tilesz)(pln->vl, 1) > 4 ); } /**************************************************************/ /* out of place, tiled, with buffer */ static void apply_tiledbuf(const plan *ego_, R *I, R *O) { const P *ego = (const P *) ego_; copy(ego->d, ego->rnk, ego->vl, I, O, X(cpy2d_tiledbuf)); } #define applicable_tiledbuf applicable_tiled /**************************************************************/ /* rank 0, out of place, using memcpy */ static void apply_memcpy(const plan *ego_, R *I, R *O) { const P *ego = (const P *) ego_; A(ego->rnk == 0); memcpy(O, I, ego->vl * sizeof(R)); } static int applicable_memcpy(const P *pln, const problem_rdft *p) { return (1 && p->I != p->O && pln->rnk == 0 && pln->vl > 2 /* do not bother memcpy-ing complex numbers */ ); } /**************************************************************/ /* rank > 0 vecloop, out of place, using memcpy (e.g. out-of-place transposes of vl-tuples ... for large vl it should be more efficient to use memcpy than the tiled stuff). */ static void memcpy_loop(INT cpysz, int rnk, const iodim *d, R *I, R *O) { INT i, n = d->n, is = d->is, os = d->os; if (rnk == 1) for (i = 0; i < n; ++i, I += is, O += os) memcpy(O, I, cpysz); else { --rnk; ++d; for (i = 0; i < n; ++i, I += is, O += os) memcpy_loop(cpysz, rnk, d, I, O); } } static void apply_memcpy_loop(const plan *ego_, R *I, R *O) { const P *ego = (const P *) ego_; memcpy_loop(ego->vl * sizeof(R), ego->rnk, ego->d, I, O); } static int applicable_memcpy_loop(const P *pln, const problem_rdft *p) { return (p->I != p->O && pln->rnk > 0 && pln->vl > 2 /* do not bother memcpy-ing complex numbers */); } /**************************************************************/ /* rank 2, in place, square transpose, iterative */ static void apply_ip_sq(const plan *ego_, R *I, R *O) { const P *ego = (const P *) ego_; UNUSED(O); transpose(ego->d, ego->rnk, ego->vl, I, X(transpose)); } static int applicable_ip_sq(const P *pln, const problem_rdft *p) { return (1 && p->I == p->O && pln->rnk >= 2 && transposep(pln)); } /**************************************************************/ /* rank 2, in place, square transpose, tiled */ static void apply_ip_sq_tiled(const plan *ego_, R *I, R *O) { const P *ego = (const P *) ego_; UNUSED(O); transpose(ego->d, ego->rnk, ego->vl, I, X(transpose_tiled)); } static int applicable_ip_sq_tiled(const P *pln, const problem_rdft *p) { return (1 && applicable_ip_sq(pln, p) /* somewhat arbitrary */ && X(compute_tilesz)(pln->vl, 2) > 4 ); } /**************************************************************/ /* rank 2, in place, square transpose, tiled, buffered */ static void apply_ip_sq_tiledbuf(const plan *ego_, R *I, R *O) { const P *ego = (const P *) ego_; UNUSED(O); transpose(ego->d, ego->rnk, ego->vl, I, X(transpose_tiledbuf)); } #define applicable_ip_sq_tiledbuf applicable_ip_sq_tiled /**************************************************************/ static int applicable(const S *ego, const problem *p_) { const problem_rdft *p = (const problem_rdft *) p_; P pln; return (1 && p->sz->rnk == 0 && FINITE_RNK(p->vecsz->rnk) && fill_iodim(&pln, p) && ego->applicable(&pln, p) ); } static void print(const plan *ego_, printer *p) { const P *ego = (const P *) ego_; int i; p->print(p, "(%s/%D", ego->nam, ego->vl); for (i = 0; i < ego->rnk; ++i) p->print(p, "%v", ego->d[i].n); p->print(p, ")"); } static plan *mkplan(const solver *ego_, const problem *p_, planner *plnr) { const problem_rdft *p; const S *ego = (const S *) ego_; P *pln; int retval; static const plan_adt padt = { X(rdft_solve), X(null_awake), print, X(plan_null_destroy) }; UNUSED(plnr); if (!applicable(ego, p_)) return (plan *) 0; p = (const problem_rdft *) p_; pln = MKPLAN_RDFT(P, &padt, ego->apply); retval = fill_iodim(pln, p); (void)retval; /* UNUSED unless DEBUG */ A(retval); A(pln->vl > 0); /* because FINITE_RNK(p->vecsz->rnk) holds */ pln->nam = ego->nam; /* X(tensor_sz)(p->vecsz) loads, X(tensor_sz)(p->vecsz) stores */ X(ops_other)(2 * X(tensor_sz)(p->vecsz), &pln->super.super.ops); return &(pln->super.super); } void X(rdft_rank0_register)(planner *p) { unsigned i; static struct { rdftapply apply; int (*applicable)(const P *, const problem_rdft *); const char *nam; } tab[] = { { apply_memcpy, applicable_memcpy, "rdft-rank0-memcpy" }, { apply_memcpy_loop, applicable_memcpy_loop, "rdft-rank0-memcpy-loop" }, { apply_iter, applicable_iter, "rdft-rank0-iter-ci" }, { apply_cpy2dco, applicable_cpy2dco, "rdft-rank0-iter-co" }, { apply_tiled, applicable_tiled, "rdft-rank0-tiled" }, { apply_tiledbuf, applicable_tiledbuf, "rdft-rank0-tiledbuf" }, { apply_ip_sq, applicable_ip_sq, "rdft-rank0-ip-sq" }, { apply_ip_sq_tiled, applicable_ip_sq_tiled, "rdft-rank0-ip-sq-tiled" }, { apply_ip_sq_tiledbuf, applicable_ip_sq_tiledbuf, "rdft-rank0-ip-sq-tiledbuf" }, }; for (i = 0; i < sizeof(tab) / sizeof(tab[0]); ++i) { static const solver_adt sadt = { PROBLEM_RDFT, mkplan, 0 }; S *slv = MKSOLVER(S, &sadt); slv->apply = tab[i].apply; slv->applicable = tab[i].applicable; slv->nam = tab[i].nam; REGISTER_SOLVER(p, &(slv->super)); } } fftw-3.3.3/rdft/rank-geq2.c0000644000175000017500000001337012054410050012274 00000000000000/* * Copyright (c) 2003, 2007-11 Matteo Frigo * Copyright (c) 2003, 2007-11 Massachusetts Institute of Technology * * This program is free software; you can redistribute it and/or modify * it under the terms of the GNU General Public License as published by * the Free Software Foundation; either version 2 of the License, or * (at your option) any later version. * * This program is distributed in the hope that it will be useful, * but WITHOUT ANY WARRANTY; without even the implied warranty of * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the * GNU General Public License for more details. * * You should have received a copy of the GNU General Public License * along with this program; if not, write to the Free Software * Foundation, Inc., 51 Franklin Street, Fifth Floor, Boston, MA 02110-1301 USA * */ /* plans for RDFT of rank >= 2 (multidimensional) */ /* FIXME: this solver cannot strictly be applied to multidimensional DHTs, since the latter are not separable...up to rnk-1 additional post-processing passes may be required. See also: R. N. Bracewell, O. Buneman, H. Hao, and J. Villasenor, "Fast two-dimensional Hartley transform," Proc. IEEE 74, 1282-1283 (1986). H. Hao and R. N. Bracewell, "A three-dimensional DFT algorithm using the fast Hartley transform," Proc. IEEE 75(2), 264-266 (1987). */ #include "rdft.h" typedef struct { solver super; int spltrnk; const int *buddies; int nbuddies; } S; typedef struct { plan_rdft super; plan *cld1, *cld2; const S *solver; } P; /* Compute multi-dimensional RDFT by applying the two cld plans (lower-rnk RDFTs). */ static void apply(const plan *ego_, R *I, R *O) { const P *ego = (const P *) ego_; plan_rdft *cld1, *cld2; cld1 = (plan_rdft *) ego->cld1; cld1->apply(ego->cld1, I, O); cld2 = (plan_rdft *) ego->cld2; cld2->apply(ego->cld2, O, O); } static void awake(plan *ego_, enum wakefulness wakefulness) { P *ego = (P *) ego_; X(plan_awake)(ego->cld1, wakefulness); X(plan_awake)(ego->cld2, wakefulness); } static void destroy(plan *ego_) { P *ego = (P *) ego_; X(plan_destroy_internal)(ego->cld2); X(plan_destroy_internal)(ego->cld1); } static void print(const plan *ego_, printer *p) { const P *ego = (const P *) ego_; const S *s = ego->solver; p->print(p, "(rdft-rank>=2/%d%(%p%)%(%p%))", s->spltrnk, ego->cld1, ego->cld2); } static int picksplit(const S *ego, const tensor *sz, int *rp) { A(sz->rnk > 1); /* cannot split rnk <= 1 */ if (!X(pickdim)(ego->spltrnk, ego->buddies, ego->nbuddies, sz, 1, rp)) return 0; *rp += 1; /* convert from dim. index to rank */ if (*rp >= sz->rnk) /* split must reduce rank */ return 0; return 1; } static int applicable0(const solver *ego_, const problem *p_, int *rp) { const problem_rdft *p = (const problem_rdft *) p_; const S *ego = (const S *)ego_; return (1 && FINITE_RNK(p->sz->rnk) && FINITE_RNK(p->vecsz->rnk) && p->sz->rnk >= 2 && picksplit(ego, p->sz, rp) ); } /* TODO: revise this. */ static int applicable(const solver *ego_, const problem *p_, const planner *plnr, int *rp) { const S *ego = (const S *)ego_; if (!applicable0(ego_, p_, rp)) return 0; if (NO_RANK_SPLITSP(plnr) && (ego->spltrnk != ego->buddies[0])) return 0; if (NO_UGLYP(plnr)) { /* Heuristic: if the vector stride is greater than the transform sz, don't use (prefer to do the vector loop first with a vrank-geq1 plan). */ const problem_rdft *p = (const problem_rdft *) p_; if (p->vecsz->rnk > 0 && X(tensor_min_stride)(p->vecsz) > X(tensor_max_index)(p->sz)) return 0; } return 1; } static plan *mkplan(const solver *ego_, const problem *p_, planner *plnr) { const S *ego = (const S *) ego_; const problem_rdft *p; P *pln; plan *cld1 = 0, *cld2 = 0; tensor *sz1, *sz2, *vecszi, *sz2i; int spltrnk; static const plan_adt padt = { X(rdft_solve), awake, print, destroy }; if (!applicable(ego_, p_, plnr, &spltrnk)) return (plan *) 0; p = (const problem_rdft *) p_; X(tensor_split)(p->sz, &sz1, spltrnk, &sz2); vecszi = X(tensor_copy_inplace)(p->vecsz, INPLACE_OS); sz2i = X(tensor_copy_inplace)(sz2, INPLACE_OS); cld1 = X(mkplan_d)(plnr, X(mkproblem_rdft_d)(X(tensor_copy)(sz2), X(tensor_append)(p->vecsz, sz1), p->I, p->O, p->kind + spltrnk)); if (!cld1) goto nada; cld2 = X(mkplan_d)(plnr, X(mkproblem_rdft_d)( X(tensor_copy_inplace)(sz1, INPLACE_OS), X(tensor_append)(vecszi, sz2i), p->O, p->O, p->kind)); if (!cld2) goto nada; pln = MKPLAN_RDFT(P, &padt, apply); pln->cld1 = cld1; pln->cld2 = cld2; pln->solver = ego; X(ops_add)(&cld1->ops, &cld2->ops, &pln->super.super.ops); X(tensor_destroy4)(sz2, sz1, vecszi, sz2i); return &(pln->super.super); nada: X(plan_destroy_internal)(cld2); X(plan_destroy_internal)(cld1); X(tensor_destroy4)(sz2, sz1, vecszi, sz2i); return (plan *) 0; } static solver *mksolver(int spltrnk, const int *buddies, int nbuddies) { static const solver_adt sadt = { PROBLEM_RDFT, mkplan, 0 }; S *slv = MKSOLVER(S, &sadt); slv->spltrnk = spltrnk; slv->buddies = buddies; slv->nbuddies = nbuddies; return &(slv->super); } void X(rdft_rank_geq2_register)(planner *p) { int i; static const int buddies[] = { 1, 0, -2 }; const int nbuddies = (int)(sizeof(buddies) / sizeof(buddies[0])); for (i = 0; i < nbuddies; ++i) REGISTER_SOLVER(p, mksolver(buddies[i], buddies, nbuddies)); /* FIXME: Should we try more buddies? See also dft/rank-geq2. */ } fftw-3.3.3/rdft/rdft.h0000644000175000017500000001305612054410050011452 00000000000000/* * Copyright (c) 2003, 2007-11 Matteo Frigo * Copyright (c) 2003, 2007-11 Massachusetts Institute of Technology * * This program is free software; you can redistribute it and/or modify * it under the terms of the GNU General Public License as published by * the Free Software Foundation; either version 2 of the License, or * (at your option) any later version. * * This program is distributed in the hope that it will be useful, * but WITHOUT ANY WARRANTY; without even the implied warranty of * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the * GNU General Public License for more details. * * You should have received a copy of the GNU General Public License * along with this program; if not, write to the Free Software * Foundation, Inc., 51 Franklin Street, Fifth Floor, Boston, MA 02110-1301 USA * */ #ifndef __RDFT_H__ #define __RDFT_H__ #include "ifftw.h" #include "codelet-rdft.h" #ifdef __cplusplus extern "C" { #endif /* __cplusplus */ /* problem.c: */ typedef struct { problem super; tensor *sz, *vecsz; R *I, *O; #if defined(STRUCT_HACK_KR) rdft_kind kind[1]; #elif defined(STRUCT_HACK_C99) rdft_kind kind[]; #else rdft_kind *kind; #endif } problem_rdft; void X(rdft_zerotens)(tensor *sz, R *I); problem *X(mkproblem_rdft)(const tensor *sz, const tensor *vecsz, R *I, R *O, const rdft_kind *kind); problem *X(mkproblem_rdft_d)(tensor *sz, tensor *vecsz, R *I, R *O, const rdft_kind *kind); problem *X(mkproblem_rdft_0_d)(tensor *vecsz, R *I, R *O); problem *X(mkproblem_rdft_1)(const tensor *sz, const tensor *vecsz, R *I, R *O, rdft_kind kind); problem *X(mkproblem_rdft_1_d)(tensor *sz, tensor *vecsz, R *I, R *O, rdft_kind kind); const char *X(rdft_kind_str)(rdft_kind kind); /* solve.c: */ void X(rdft_solve)(const plan *ego_, const problem *p_); /* plan.c: */ typedef void (*rdftapply) (const plan *ego, R *I, R *O); typedef struct { plan super; rdftapply apply; } plan_rdft; plan *X(mkplan_rdft)(size_t size, const plan_adt *adt, rdftapply apply); #define MKPLAN_RDFT(type, adt, apply) \ (type *)X(mkplan_rdft)(sizeof(type), adt, apply) /* various solvers */ solver *X(mksolver_rdft_r2c_direct)(kr2c k, const kr2c_desc *desc); solver *X(mksolver_rdft_r2c_directbuf)(kr2c k, const kr2c_desc *desc); solver *X(mksolver_rdft_r2r_direct)(kr2r k, const kr2r_desc *desc); void X(rdft_rank0_register)(planner *p); void X(rdft_vrank3_transpose_register)(planner *p); void X(rdft_rank_geq2_register)(planner *p); void X(rdft_indirect_register)(planner *p); void X(rdft_vrank_geq1_register)(planner *p); void X(rdft_buffered_register)(planner *p); void X(rdft_generic_register)(planner *p); void X(rdft_rader_hc2hc_register)(planner *p); void X(rdft_dht_register)(planner *p); void X(dht_r2hc_register)(planner *p); void X(dht_rader_register)(planner *p); void X(dft_r2hc_register)(planner *p); void X(rdft_nop_register)(planner *p); void X(hc2hc_generic_register)(planner *p); /****************************************************************************/ /* problem2.c: */ /* An RDFT2 problem transforms a 1d real array r[n] with stride is/os to/from an "unpacked" complex array {rio,iio}[n/2 + 1] with stride os/is. R0 points to the first even element of the real array. R1 points to the first odd element of the real array. Strides on the real side of the transform express distances between consecutive elements of the same array (even or odd). E.g., for a contiguous input R0 R1 R2 R3 ... the input stride would be 2, not 1. This convention is necessary for hc2c codelets to work, since they transpose even/odd with real/imag. Multidimensional transforms use complex DFTs for the noncontiguous dimensions. vecsz has the usual interpretation. */ typedef struct { problem super; tensor *sz; tensor *vecsz; R *r0, *r1; R *cr, *ci; rdft_kind kind; /* assert(kind < DHT) */ } problem_rdft2; problem *X(mkproblem_rdft2)(const tensor *sz, const tensor *vecsz, R *r0, R *r1, R *cr, R *ci, rdft_kind kind); problem *X(mkproblem_rdft2_d)(tensor *sz, tensor *vecsz, R *r0, R *r1, R *cr, R *ci, rdft_kind kind); problem *X(mkproblem_rdft2_d_3pointers)(tensor *sz, tensor *vecsz, R *r, R *cr, R *ci, rdft_kind kind); int X(rdft2_inplace_strides)(const problem_rdft2 *p, int vdim); INT X(rdft2_tensor_max_index)(const tensor *sz, rdft_kind k); void X(rdft2_strides)(rdft_kind kind, const iodim *d, INT *rs, INT *cs); INT X(rdft2_complex_n)(INT real_n, rdft_kind kind); /* verify.c: */ void X(rdft2_verify)(plan *pln, const problem_rdft2 *p, int rounds); /* solve.c: */ void X(rdft2_solve)(const plan *ego_, const problem *p_); /* plan.c: */ typedef void (*rdft2apply) (const plan *ego, R *r0, R *r1, R *cr, R *ci); typedef struct { plan super; rdft2apply apply; } plan_rdft2; plan *X(mkplan_rdft2)(size_t size, const plan_adt *adt, rdft2apply apply); #define MKPLAN_RDFT2(type, adt, apply) \ (type *)X(mkplan_rdft2)(sizeof(type), adt, apply) /* various solvers */ solver *X(mksolver_rdft2_direct)(kr2c k, const kr2c_desc *desc); void X(rdft2_vrank_geq1_register)(planner *p); void X(rdft2_buffered_register)(planner *p); void X(rdft2_rdft_register)(planner *p); void X(rdft2_nop_register)(planner *p); void X(rdft2_rank0_register)(planner *p); void X(rdft2_rank_geq2_register)(planner *p); /****************************************************************************/ /* configurations */ void X(rdft_conf_standard)(planner *p); #ifdef __cplusplus } /* extern "C" */ #endif /* __cplusplus */ #endif /* __RDFT_H__ */ fftw-3.3.3/rdft/rdft-dht.c0000644000175000017500000001263212054410050012221 00000000000000/* * Copyright (c) 2003, 2007-11 Matteo Frigo * Copyright (c) 2003, 2007-11 Massachusetts Institute of Technology * * This program is free software; you can redistribute it and/or modify * it under the terms of the GNU General Public License as published by * the Free Software Foundation; either version 2 of the License, or * (at your option) any later version. * * This program is distributed in the hope that it will be useful, * but WITHOUT ANY WARRANTY; without even the implied warranty of * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the * GNU General Public License for more details. * * You should have received a copy of the GNU General Public License * along with this program; if not, write to the Free Software * Foundation, Inc., 51 Franklin Street, Fifth Floor, Boston, MA 02110-1301 USA * */ /* Solve an R2HC/HC2R problem via post/pre processing of a DHT. This is mainly useful because we can use Rader to compute DHTs of prime sizes. It also allows us to express hc2r problems in terms of r2hc (via dht-r2hc), and to do hc2r problems without destroying the input. */ #include "rdft.h" typedef struct { solver super; } S; typedef struct { plan_rdft super; plan *cld; INT is, os; INT n; } P; static void apply_r2hc(const plan *ego_, R *I, R *O) { const P *ego = (const P *) ego_; INT os; INT i, n; { plan_rdft *cld = (plan_rdft *) ego->cld; cld->apply((plan *) cld, I, O); } n = ego->n; os = ego->os; for (i = 1; i < n - i; ++i) { E a, b; a = K(0.5) * O[os * i]; b = K(0.5) * O[os * (n - i)]; O[os * i] = a + b; #if FFT_SIGN == -1 O[os * (n - i)] = b - a; #else O[os * (n - i)] = a - b; #endif } } /* hc2r, destroying input as usual */ static void apply_hc2r(const plan *ego_, R *I, R *O) { const P *ego = (const P *) ego_; INT is = ego->is; INT i, n = ego->n; for (i = 1; i < n - i; ++i) { E a, b; a = I[is * i]; b = I[is * (n - i)]; #if FFT_SIGN == -1 I[is * i] = a - b; I[is * (n - i)] = a + b; #else I[is * i] = a + b; I[is * (n - i)] = a - b; #endif } { plan_rdft *cld = (plan_rdft *) ego->cld; cld->apply((plan *) cld, I, O); } } /* hc2r, without destroying input */ static void apply_hc2r_save(const plan *ego_, R *I, R *O) { const P *ego = (const P *) ego_; INT is = ego->is, os = ego->os; INT i, n = ego->n; O[0] = I[0]; for (i = 1; i < n - i; ++i) { E a, b; a = I[is * i]; b = I[is * (n - i)]; #if FFT_SIGN == -1 O[os * i] = a - b; O[os * (n - i)] = a + b; #else O[os * i] = a + b; O[os * (n - i)] = a - b; #endif } if (i == n - i) O[os * i] = I[is * i]; { plan_rdft *cld = (plan_rdft *) ego->cld; cld->apply((plan *) cld, O, O); } } static void awake(plan *ego_, enum wakefulness wakefulness) { P *ego = (P *) ego_; X(plan_awake)(ego->cld, wakefulness); } static void destroy(plan *ego_) { P *ego = (P *) ego_; X(plan_destroy_internal)(ego->cld); } static void print(const plan *ego_, printer *p) { const P *ego = (const P *) ego_; p->print(p, "(%s-dht-%D%(%p%))", ego->super.apply == apply_r2hc ? "r2hc" : "hc2r", ego->n, ego->cld); } static int applicable0(const solver *ego_, const problem *p_) { const problem_rdft *p = (const problem_rdft *) p_; UNUSED(ego_); return (1 && p->sz->rnk == 1 && p->vecsz->rnk == 0 && (p->kind[0] == R2HC || p->kind[0] == HC2R) /* hack: size-2 DHT etc. are defined as being equivalent to size-2 R2HC in problem.c, so we need this to prevent infinite loops for size 2 in EXHAUSTIVE mode: */ && p->sz->dims[0].n > 2 ); } static int applicable(const solver *ego, const problem *p_, const planner *plnr) { return (!NO_SLOWP(plnr) && applicable0(ego, p_)); } static plan *mkplan(const solver *ego_, const problem *p_, planner *plnr) { P *pln; const problem_rdft *p; problem *cldp; plan *cld; static const plan_adt padt = { X(rdft_solve), awake, print, destroy }; if (!applicable(ego_, p_, plnr)) return (plan *)0; p = (const problem_rdft *) p_; if (p->kind[0] == R2HC || !NO_DESTROY_INPUTP(plnr)) cldp = X(mkproblem_rdft_1)(p->sz, p->vecsz, p->I, p->O, DHT); else { tensor *sz = X(tensor_copy_inplace)(p->sz, INPLACE_OS); cldp = X(mkproblem_rdft_1)(sz, p->vecsz, p->O, p->O, DHT); X(tensor_destroy)(sz); } cld = X(mkplan_d)(plnr, cldp); if (!cld) return (plan *)0; pln = MKPLAN_RDFT(P, &padt, p->kind[0] == R2HC ? apply_r2hc : (NO_DESTROY_INPUTP(plnr) ? apply_hc2r_save : apply_hc2r)); pln->n = p->sz->dims[0].n; pln->is = p->sz->dims[0].is; pln->os = p->sz->dims[0].os; pln->cld = cld; pln->super.super.ops = cld->ops; pln->super.super.ops.other += 4 * ((pln->n - 1)/2); pln->super.super.ops.add += 2 * ((pln->n - 1)/2); if (p->kind[0] == R2HC) pln->super.super.ops.mul += 2 * ((pln->n - 1)/2); if (pln->super.apply == apply_hc2r_save) pln->super.super.ops.other += 2 + (pln->n % 2 ? 0 : 2); return &(pln->super.super); } /* constructor */ static solver *mksolver(void) { static const solver_adt sadt = { PROBLEM_RDFT, mkplan, 0 }; S *slv = MKSOLVER(S, &sadt); return &(slv->super); } void X(rdft_dht_register)(planner *p) { REGISTER_SOLVER(p, mksolver()); } fftw-3.3.3/rdft/solve.c0000644000175000017500000000216212054410050011632 00000000000000/* * Copyright (c) 2003, 2007-11 Matteo Frigo * Copyright (c) 2003, 2007-11 Massachusetts Institute of Technology * * This program is free software; you can redistribute it and/or modify * it under the terms of the GNU General Public License as published by * the Free Software Foundation; either version 2 of the License, or * (at your option) any later version. * * This program is distributed in the hope that it will be useful, * but WITHOUT ANY WARRANTY; without even the implied warranty of * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the * GNU General Public License for more details. * * You should have received a copy of the GNU General Public License * along with this program; if not, write to the Free Software * Foundation, Inc., 51 Franklin Street, Fifth Floor, Boston, MA 02110-1301 USA * */ #include "rdft.h" /* use the apply() operation for RDFT problems */ void X(rdft_solve)(const plan *ego_, const problem *p_) { const plan_rdft *ego = (const plan_rdft *) ego_; const problem_rdft *p = (const problem_rdft *) p_; ego->apply(ego_, UNTAINT(p->I), UNTAINT(p->O)); } fftw-3.3.3/rdft/vrank-geq1.c0000644000175000017500000001362612054410050012465 00000000000000/* * Copyright (c) 2003, 2007-11 Matteo Frigo * Copyright (c) 2003, 2007-11 Massachusetts Institute of Technology * * This program is free software; you can redistribute it and/or modify * it under the terms of the GNU General Public License as published by * the Free Software Foundation; either version 2 of the License, or * (at your option) any later version. * * This program is distributed in the hope that it will be useful, * but WITHOUT ANY WARRANTY; without even the implied warranty of * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the * GNU General Public License for more details. * * You should have received a copy of the GNU General Public License * along with this program; if not, write to the Free Software * Foundation, Inc., 51 Franklin Street, Fifth Floor, Boston, MA 02110-1301 USA * */ /* Plans for handling vector transform loops. These are *just* the loops, and rely on child plans for the actual RDFTs. They form a wrapper around solvers that don't have apply functions for non-null vectors. vrank-geq1 plans also recursively handle the case of multi-dimensional vectors, obviating the need for most solvers to deal with this. We can also play games here, such as reordering the vector loops. Each vrank-geq1 plan reduces the vector rank by 1, picking out a dimension determined by the vecloop_dim field of the solver. */ #include "rdft.h" typedef struct { solver super; int vecloop_dim; const int *buddies; int nbuddies; } S; typedef struct { plan_rdft super; plan *cld; INT vl; INT ivs, ovs; const S *solver; } P; static void apply(const plan *ego_, R *I, R *O) { const P *ego = (const P *) ego_; INT i, vl = ego->vl; INT ivs = ego->ivs, ovs = ego->ovs; rdftapply cldapply = ((plan_rdft *) ego->cld)->apply; for (i = 0; i < vl; ++i) { cldapply(ego->cld, I + i * ivs, O + i * ovs); } } static void awake(plan *ego_, enum wakefulness wakefulness) { P *ego = (P *) ego_; X(plan_awake)(ego->cld, wakefulness); } static void destroy(plan *ego_) { P *ego = (P *) ego_; X(plan_destroy_internal)(ego->cld); } static void print(const plan *ego_, printer *p) { const P *ego = (const P *) ego_; const S *s = ego->solver; p->print(p, "(rdft-vrank>=1-x%D/%d%(%p%))", ego->vl, s->vecloop_dim, ego->cld); } static int pickdim(const S *ego, const tensor *vecsz, int oop, int *dp) { return X(pickdim)(ego->vecloop_dim, ego->buddies, ego->nbuddies, vecsz, oop, dp); } static int applicable0(const solver *ego_, const problem *p_, int *dp) { const S *ego = (const S *) ego_; const problem_rdft *p = (const problem_rdft *) p_; return (1 && FINITE_RNK(p->vecsz->rnk) && p->vecsz->rnk > 0 && p->sz->rnk >= 0 && pickdim(ego, p->vecsz, p->I != p->O, dp) ); } static int applicable(const solver *ego_, const problem *p_, const planner *plnr, int *dp) { const S *ego = (const S *)ego_; const problem_rdft *p; if (!applicable0(ego_, p_, dp)) return 0; /* fftw2 behavior */ if (NO_VRANK_SPLITSP(plnr) && (ego->vecloop_dim != ego->buddies[0])) return 0; p = (const problem_rdft *) p_; if (NO_UGLYP(plnr)) { /* the rank-0 solver deals with the general case most of the time (an exception is loops of non-square transposes) */ if (NO_SLOWP(plnr) && p->sz->rnk == 0) return 0; /* Heuristic: if the transform is multi-dimensional, and the vector stride is less than the transform size, then we probably want to use a rank>=2 plan first in order to combine this vector with the transform-dimension vectors. */ { iodim *d = p->vecsz->dims + *dp; if (1 && p->sz->rnk > 1 && X(imin)(X(iabs)(d->is), X(iabs)(d->os)) < X(tensor_max_index)(p->sz) ) return 0; } /* prefer threaded version */ if (NO_NONTHREADEDP(plnr)) return 0; /* exploit built-in vecloops of (ugly) r{e,o}dft solvers */ if (p->vecsz->rnk == 1 && p->sz->rnk == 1 && REODFT_KINDP(p->kind[0])) return 0; } return 1; } static plan *mkplan(const solver *ego_, const problem *p_, planner *plnr) { const S *ego = (const S *) ego_; const problem_rdft *p; P *pln; plan *cld; int vdim; iodim *d; static const plan_adt padt = { X(rdft_solve), awake, print, destroy }; if (!applicable(ego_, p_, plnr, &vdim)) return (plan *) 0; p = (const problem_rdft *) p_; d = p->vecsz->dims + vdim; A(d->n > 1); cld = X(mkplan_d)(plnr, X(mkproblem_rdft_d)( X(tensor_copy)(p->sz), X(tensor_copy_except)(p->vecsz, vdim), TAINT(p->I, d->is), TAINT(p->O, d->os), p->kind)); if (!cld) return (plan *) 0; pln = MKPLAN_RDFT(P, &padt, apply); pln->cld = cld; pln->vl = d->n; pln->ivs = d->is; pln->ovs = d->os; pln->solver = ego; X(ops_zero)(&pln->super.super.ops); pln->super.super.ops.other = 3.14159; /* magic to prefer codelet loops */ X(ops_madd2)(pln->vl, &cld->ops, &pln->super.super.ops); if (p->sz->rnk != 1 || (p->sz->dims[0].n > 128)) pln->super.super.pcost = pln->vl * cld->pcost; return &(pln->super.super); } static solver *mksolver(int vecloop_dim, const int *buddies, int nbuddies) { static const solver_adt sadt = { PROBLEM_RDFT, mkplan, 0 }; S *slv = MKSOLVER(S, &sadt); slv->vecloop_dim = vecloop_dim; slv->buddies = buddies; slv->nbuddies = nbuddies; return &(slv->super); } void X(rdft_vrank_geq1_register)(planner *p) { int i; /* FIXME: Should we try other vecloop_dim values? */ static const int buddies[] = { 1, -1 }; const int nbuddies = (int)(sizeof(buddies) / sizeof(buddies[0])); for (i = 0; i < nbuddies; ++i) REGISTER_SOLVER(p, mksolver(buddies[i], buddies, nbuddies)); } fftw-3.3.3/rdft/vrank3-transpose.c0000644000175000017500000005416712054410050013736 00000000000000/* * Copyright (c) 2003, 2007-11 Matteo Frigo * Copyright (c) 2003, 2007-11 Massachusetts Institute of Technology * * This program is free software; you can redistribute it and/or modify * it under the terms of the GNU General Public License as published by * the Free Software Foundation; either version 2 of the License, or * (at your option) any later version. * * This program is distributed in the hope that it will be useful, * but WITHOUT ANY WARRANTY; without even the implied warranty of * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the * GNU General Public License for more details. * * You should have received a copy of the GNU General Public License * along with this program; if not, write to the Free Software * Foundation, Inc., 51 Franklin Street, Fifth Floor, Boston, MA 02110-1301 USA * */ /* rank-0, vector-rank-3, non-square in-place transposition (see rank0.c for square transposition) */ #include "rdft.h" #ifdef HAVE_STRING_H #include /* for memcpy() */ #endif struct P_s; typedef struct { rdftapply apply; int (*applicable)(const problem_rdft *p, planner *plnr, int dim0, int dim1, int dim2, INT *nbuf); int (*mkcldrn)(const problem_rdft *p, planner *plnr, struct P_s *ego); const char *nam; } transpose_adt; typedef struct { solver super; const transpose_adt *adt; } S; typedef struct P_s { plan_rdft super; INT n, m, vl; /* transpose n x m matrix of vl-tuples */ INT nbuf; /* buffer size */ INT nd, md, d; /* transpose-gcd params */ INT nc, mc; /* transpose-cut params */ plan *cld1, *cld2, *cld3; /* children, null if unused */ const S *slv; } P; /*************************************************************************/ /* some utilities for the solvers */ static INT gcd(INT a, INT b) { INT r; do { r = a % b; a = b; b = r; } while (r != 0); return a; } /* whether we can transpose with one of our routines expecting contiguous Ntuples */ static int Ntuple_transposable(const iodim *a, const iodim *b, INT vl, INT vs) { return (vs == 1 && b->is == vl && a->os == vl && ((a->n == b->n && a->is == b->os && a->is >= b->n && a->is % vl == 0) || (a->is == b->n * vl && b->os == a->n * vl))); } /* check whether a and b correspond to the first and second dimensions of a transpose of tuples with vector length = vl, stride = vs. */ static int transposable(const iodim *a, const iodim *b, INT vl, INT vs) { return ((a->n == b->n && a->os == b->is && a->is == b->os) || Ntuple_transposable(a, b, vl, vs)); } static int pickdim(const tensor *s, int *pdim0, int *pdim1, int *pdim2) { int dim0, dim1; for (dim0 = 0; dim0 < s->rnk; ++dim0) for (dim1 = 0; dim1 < s->rnk; ++dim1) { int dim2 = 3 - dim0 - dim1; if (dim0 == dim1) continue; if ((s->rnk == 2 || s->dims[dim2].is == s->dims[dim2].os) && transposable(s->dims + dim0, s->dims + dim1, s->rnk == 2 ? (INT)1 : s->dims[dim2].n, s->rnk == 2 ? (INT)1 : s->dims[dim2].is)) { *pdim0 = dim0; *pdim1 = dim1; *pdim2 = dim2; return 1; } } return 0; } #define MINBUFDIV 9 /* min factor by which buffer is smaller than data */ #define MAXBUF 65536 /* maximum non-ugly buffer */ /* generic applicability function */ static int applicable(const solver *ego_, const problem *p_, planner *plnr, int *dim0, int *dim1, int *dim2, INT *nbuf) { const S *ego = (const S *) ego_; const problem_rdft *p = (const problem_rdft *) p_; return (1 && p->I == p->O && p->sz->rnk == 0 && (p->vecsz->rnk == 2 || p->vecsz->rnk == 3) && pickdim(p->vecsz, dim0, dim1, dim2) /* UGLY if vecloop in wrong order for locality */ && (!NO_UGLYP(plnr) || p->vecsz->rnk == 2 || X(iabs)(p->vecsz->dims[*dim2].is) < X(imax)(X(iabs)(p->vecsz->dims[*dim0].is), X(iabs)(p->vecsz->dims[*dim0].os))) /* SLOW if non-square */ && (!NO_SLOWP(plnr) || p->vecsz->dims[*dim0].n == p->vecsz->dims[*dim1].n) && ego->adt->applicable(p, plnr, *dim0,*dim1,*dim2,nbuf) /* buffers too big are UGLY */ && ((!NO_UGLYP(plnr) && !CONSERVE_MEMORYP(plnr)) || *nbuf <= MAXBUF || *nbuf * MINBUFDIV <= X(tensor_sz)(p->vecsz)) ); } static void get_transpose_vec(const problem_rdft *p, int dim2, INT *vl,INT *vs) { if (p->vecsz->rnk == 2) { *vl = 1; *vs = 1; } else { *vl = p->vecsz->dims[dim2].n; *vs = p->vecsz->dims[dim2].is; /* == os */ } } /*************************************************************************/ /* Cache-oblivious in-place transpose of non-square matrices, based on transposes of blocks given by the gcd of the dimensions. This algorithm is related to algorithm V5 from Murray Dow, "Transposing a matrix on a vector computer," Parallel Computing 21 (12), 1997-2005 (1995), with the modification that we use cache-oblivious recursive transpose subroutines (and we derived it independently). For a p x q matrix, this requires scratch space equal to the size of the matrix divided by gcd(p,q). Alternatively, see also the "cut" algorithm below, if |p-q| * gcd(p,q) < max(p,q). */ static void apply_gcd(const plan *ego_, R *I, R *O) { const P *ego = (const P *) ego_; INT n = ego->nd, m = ego->md, d = ego->d; INT vl = ego->vl; R *buf = (R *)MALLOC(sizeof(R) * ego->nbuf, BUFFERS); INT i, num_el = n*m*d*vl; A(ego->n == n * d && ego->m == m * d); UNUSED(O); /* Transpose the matrix I in-place, where I is an (n*d) x (m*d) matrix of vl-tuples and buf contains n*m*d*vl elements. In general, to transpose a p x q matrix, you should call this routine with d = gcd(p, q), n = p/d, and m = q/d. */ A(n > 0 && m > 0 && vl > 0); A(d > 1); /* treat as (d x n) x (d' x m) matrix. (d' = d) */ /* First, transpose d x (n x d') x m to d x (d' x n) x m, using the buf matrix. This consists of d transposes of contiguous n x d' matrices of m-tuples. */ if (n > 1) { rdftapply cldapply = ((plan_rdft *) ego->cld1)->apply; for (i = 0; i < d; ++i) { cldapply(ego->cld1, I + i*num_el, buf); memcpy(I + i*num_el, buf, num_el*sizeof(R)); } } /* Now, transpose (d x d') x (n x m) to (d' x d) x (n x m), which is a square in-place transpose of n*m-tuples: */ { rdftapply cldapply = ((plan_rdft *) ego->cld2)->apply; cldapply(ego->cld2, I, I); } /* Finally, transpose d' x ((d x n) x m) to d' x (m x (d x n)), using the buf matrix. This consists of d' transposes of contiguous d*n x m matrices. */ if (m > 1) { rdftapply cldapply = ((plan_rdft *) ego->cld3)->apply; for (i = 0; i < d; ++i) { cldapply(ego->cld3, I + i*num_el, buf); memcpy(I + i*num_el, buf, num_el*sizeof(R)); } } X(ifree)(buf); } static int applicable_gcd(const problem_rdft *p, planner *plnr, int dim0, int dim1, int dim2, INT *nbuf) { INT n = p->vecsz->dims[dim0].n; INT m = p->vecsz->dims[dim1].n; INT d, vl, vs; get_transpose_vec(p, dim2, &vl, &vs); d = gcd(n, m); *nbuf = n * (m / d) * vl; return (!NO_SLOWP(plnr) /* FIXME: not really SLOW for large 1d ffts */ && n != m && d > 1 && Ntuple_transposable(p->vecsz->dims + dim0, p->vecsz->dims + dim1, vl, vs)); } static int mkcldrn_gcd(const problem_rdft *p, planner *plnr, P *ego) { INT n = ego->nd, m = ego->md, d = ego->d; INT vl = ego->vl; R *buf = (R *)MALLOC(sizeof(R) * ego->nbuf, BUFFERS); INT num_el = n*m*d*vl; if (n > 1) { ego->cld1 = X(mkplan_d)(plnr, X(mkproblem_rdft_0_d)( X(mktensor_3d)(n, d*m*vl, m*vl, d, m*vl, n*m*vl, m*vl, 1, 1), TAINT(p->I, num_el), buf)); if (!ego->cld1) goto nada; X(ops_madd)(d, &ego->cld1->ops, &ego->super.super.ops, &ego->super.super.ops); ego->super.super.ops.other += num_el * d * 2; } ego->cld2 = X(mkplan_d)(plnr, X(mkproblem_rdft_0_d)( X(mktensor_3d)(d, d*n*m*vl, n*m*vl, d, n*m*vl, d*n*m*vl, n*m*vl, 1, 1), p->I, p->I)); if (!ego->cld2) goto nada; X(ops_add2)(&ego->cld2->ops, &ego->super.super.ops); if (m > 1) { ego->cld3 = X(mkplan_d)(plnr, X(mkproblem_rdft_0_d)( X(mktensor_3d)(d*n, m*vl, vl, m, vl, d*n*vl, vl, 1, 1), TAINT(p->I, num_el), buf)); if (!ego->cld3) goto nada; X(ops_madd2)(d, &ego->cld3->ops, &ego->super.super.ops); ego->super.super.ops.other += num_el * d * 2; } X(ifree)(buf); return 1; nada: X(ifree)(buf); return 0; } static const transpose_adt adt_gcd = { apply_gcd, applicable_gcd, mkcldrn_gcd, "rdft-transpose-gcd" }; /*************************************************************************/ /* Cache-oblivious in-place transpose of non-square n x m matrices, based on transposing a sub-matrix first and then transposing the remainder(s) with the help of a buffer. See also transpose-gcd, above, if gcd(n,m) is large. This algorithm is related to algorithm V3 from Murray Dow, "Transposing a matrix on a vector computer," Parallel Computing 21 (12), 1997-2005 (1995), with the modifications that we use cache-oblivious recursive transpose subroutines and we have the generalization for large |n-m| below. The best case, and the one described by Dow, is for |n-m| small, in which case we transpose a square sub-matrix of size min(n,m), handling the remainder via a buffer. This requires scratch space equal to the size of the matrix times |n-m| / max(n,m). As a generalization when |n-m| is not small, we also support cutting *both* dimensions to an nc x mc matrix which is *not* necessarily square, but has a large gcd (and can therefore use transpose-gcd). */ static void apply_cut(const plan *ego_, R *I, R *O) { const P *ego = (const P *) ego_; INT n = ego->n, m = ego->m, nc = ego->nc, mc = ego->mc, vl = ego->vl; INT i; R *buf1 = (R *)MALLOC(sizeof(R) * ego->nbuf, BUFFERS); UNUSED(O); if (m > mc) { ((plan_rdft *) ego->cld1)->apply(ego->cld1, I + mc*vl, buf1); for (i = 0; i < nc; ++i) memmove(I + (mc*vl) * i, I + (m*vl) * i, sizeof(R) * (mc*vl)); } ((plan_rdft *) ego->cld2)->apply(ego->cld2, I, I); /* nc x mc transpose */ if (n > nc) { R *buf2 = buf1 + (m-mc)*(nc*vl); /* FIXME: force better alignment? */ memcpy(buf2, I + nc*(m*vl), (n-nc)*(m*vl)*sizeof(R)); for (i = mc-1; i >= 0; --i) memmove(I + (n*vl) * i, I + (nc*vl) * i, sizeof(R) * (n*vl)); ((plan_rdft *) ego->cld3)->apply(ego->cld3, buf2, I + nc*vl); } if (m > mc) { if (n > nc) for (i = mc; i < m; ++i) memcpy(I + i*(n*vl), buf1 + (i-mc)*(nc*vl), (nc*vl)*sizeof(R)); else memcpy(I + mc*(n*vl), buf1, (m-mc)*(n*vl)*sizeof(R)); } X(ifree)(buf1); } /* only cut one dimension if the resulting buffer is small enough */ static int cut1(INT n, INT m, INT vl) { return (X(imax)(n,m) >= X(iabs)(n-m) * MINBUFDIV || X(imin)(n,m) * X(iabs)(n-m) * vl <= MAXBUF); } #define CUT_NSRCH 32 /* range of sizes to search for possible cuts */ static int applicable_cut(const problem_rdft *p, planner *plnr, int dim0, int dim1, int dim2, INT *nbuf) { INT n = p->vecsz->dims[dim0].n; INT m = p->vecsz->dims[dim1].n; INT vl, vs; get_transpose_vec(p, dim2, &vl, &vs); *nbuf = 0; /* always small enough to be non-UGLY (?) */ A(MINBUFDIV <= CUT_NSRCH); /* assumed to avoid inf. loops below */ return (!NO_SLOWP(plnr) /* FIXME: not really SLOW for large 1d ffts? */ && n != m /* Don't call transpose-cut recursively (avoid inf. loops): the non-square sub-transpose produced when !cut1 should always have gcd(n,m) >= min(CUT_NSRCH,n,m), for which transpose-gcd is applicable */ && (cut1(n, m, vl) || gcd(n, m) < X(imin)(MINBUFDIV, X(imin)(n,m))) && Ntuple_transposable(p->vecsz->dims + dim0, p->vecsz->dims + dim1, vl, vs)); } static int mkcldrn_cut(const problem_rdft *p, planner *plnr, P *ego) { INT n = ego->n, m = ego->m, nc, mc; INT vl = ego->vl; R *buf; /* pick the "best" cut */ if (cut1(n, m, vl)) { nc = mc = X(imin)(n,m); } else { INT dc, ns, ms; dc = gcd(m, n); nc = n; mc = m; /* search for cut with largest gcd (TODO: different optimality criteria? different search range?) */ for (ms = m; ms > 0 && ms > m - CUT_NSRCH; --ms) { for (ns = n; ns > 0 && ns > n - CUT_NSRCH; --ns) { INT ds = gcd(ms, ns); if (ds > dc) { dc = ds; nc = ns; mc = ms; if (dc == X(imin)(ns, ms)) break; /* cannot get larger than this */ } } if (dc == X(imin)(n, ms)) break; /* cannot get larger than this */ } A(dc >= X(imin)(CUT_NSRCH, X(imin)(n, m))); } ego->nc = nc; ego->mc = mc; ego->nbuf = (m-mc)*(nc*vl) + (n-nc)*(m*vl); buf = (R *)MALLOC(sizeof(R) * ego->nbuf, BUFFERS); if (m > mc) { ego->cld1 = X(mkplan_d)(plnr, X(mkproblem_rdft_0_d)( X(mktensor_3d)(nc, m*vl, vl, m-mc, vl, nc*vl, vl, 1, 1), p->I + mc*vl, buf)); if (!ego->cld1) goto nada; X(ops_add2)(&ego->cld1->ops, &ego->super.super.ops); } ego->cld2 = X(mkplan_d)(plnr, X(mkproblem_rdft_0_d)( X(mktensor_3d)(nc, mc*vl, vl, mc, vl, nc*vl, vl, 1, 1), p->I, p->I)); if (!ego->cld2) goto nada; X(ops_add2)(&ego->cld2->ops, &ego->super.super.ops); if (n > nc) { ego->cld3 = X(mkplan_d)(plnr, X(mkproblem_rdft_0_d)( X(mktensor_3d)(n-nc, m*vl, vl, m, vl, n*vl, vl, 1, 1), buf + (m-mc)*(nc*vl), p->I + nc*vl)); if (!ego->cld3) goto nada; X(ops_add2)(&ego->cld3->ops, &ego->super.super.ops); } /* memcpy/memmove operations */ ego->super.super.ops.other += 2 * vl * (nc*mc * ((m > mc) + (n > nc)) + (n-nc)*m + (m-mc)*nc); X(ifree)(buf); return 1; nada: X(ifree)(buf); return 0; } static const transpose_adt adt_cut = { apply_cut, applicable_cut, mkcldrn_cut, "rdft-transpose-cut" }; /*************************************************************************/ /* In-place transpose routine from TOMS, which follows the cycles of the permutation so that it writes to each location only once. Because of cache-line and other issues, however, this routine is typically much slower than transpose-gcd or transpose-cut, even though the latter do some extra writes. On the other hand, if the vector length is large then the TOMS routine is best. The TOMS routine also has the advantage of requiring less buffer space for the case of gcd(nx,ny) small. However, in this case it has been superseded by the combination of the generalized transpose-cut method with the transpose-gcd method, which can always transpose with buffers a small fraction of the array size regardless of gcd(nx,ny). */ /* * TOMS Transpose. Algorithm 513 (Revised version of algorithm 380). * * These routines do in-place transposes of arrays. * * [ Cate, E.G. and Twigg, D.W., ACM Transactions on Mathematical Software, * vol. 3, no. 1, 104-110 (1977) ] * * C version by Steven G. Johnson (February 1997). */ /* * "a" is a 1D array of length ny*nx*N which constains the nx x ny * matrix of N-tuples to be transposed. "a" is stored in row-major * order (last index varies fastest). move is a 1D array of length * move_size used to store information to speed up the process. The * value move_size=(ny+nx)/2 is recommended. buf should be an array * of length 2*N. * */ static void transpose_toms513(R *a, INT nx, INT ny, INT N, char *move, INT move_size, R *buf) { INT i, im, mn; R *b, *c, *d; INT ncount; INT k; /* check arguments and initialize: */ A(ny > 0 && nx > 0 && N > 0 && move_size > 0); b = buf; /* Cate & Twigg have a special case for nx == ny, but we don't bother, since we already have special code for this case elsewhere. */ c = buf + N; ncount = 2; /* always at least 2 fixed points */ k = (mn = ny * nx) - 1; for (i = 0; i < move_size; ++i) move[i] = 0; if (ny >= 3 && nx >= 3) ncount += gcd(ny - 1, nx - 1) - 1; /* # fixed points */ i = 1; im = ny; while (1) { INT i1, i2, i1c, i2c; INT kmi; /** Rearrange the elements of a loop and its companion loop: **/ i1 = i; kmi = k - i; i1c = kmi; switch (N) { case 1: b[0] = a[i1]; c[0] = a[i1c]; break; case 2: b[0] = a[2*i1]; b[1] = a[2*i1+1]; c[0] = a[2*i1c]; c[1] = a[2*i1c+1]; break; default: memcpy(b, &a[N * i1], N * sizeof(R)); memcpy(c, &a[N * i1c], N * sizeof(R)); } while (1) { i2 = ny * i1 - k * (i1 / nx); i2c = k - i2; if (i1 < move_size) move[i1] = 1; if (i1c < move_size) move[i1c] = 1; ncount += 2; if (i2 == i) break; if (i2 == kmi) { d = b; b = c; c = d; break; } switch (N) { case 1: a[i1] = a[i2]; a[i1c] = a[i2c]; break; case 2: a[2*i1] = a[2*i2]; a[2*i1+1] = a[2*i2+1]; a[2*i1c] = a[2*i2c]; a[2*i1c+1] = a[2*i2c+1]; break; default: memcpy(&a[N * i1], &a[N * i2], N * sizeof(R)); memcpy(&a[N * i1c], &a[N * i2c], N * sizeof(R)); } i1 = i2; i1c = i2c; } switch (N) { case 1: a[i1] = b[0]; a[i1c] = c[0]; break; case 2: a[2*i1] = b[0]; a[2*i1+1] = b[1]; a[2*i1c] = c[0]; a[2*i1c+1] = c[1]; break; default: memcpy(&a[N * i1], b, N * sizeof(R)); memcpy(&a[N * i1c], c, N * sizeof(R)); } if (ncount >= mn) break; /* we've moved all elements */ /** Search for loops to rearrange: **/ while (1) { INT max = k - i; ++i; A(i <= max); im += ny; if (im > k) im -= k; i2 = im; if (i == i2) continue; if (i >= move_size) { while (i2 > i && i2 < max) { i1 = i2; i2 = ny * i1 - k * (i1 / nx); } if (i2 == i) break; } else if (!move[i]) break; } } } static void apply_toms513(const plan *ego_, R *I, R *O) { const P *ego = (const P *) ego_; INT n = ego->n, m = ego->m; INT vl = ego->vl; R *buf = (R *)MALLOC(sizeof(R) * ego->nbuf, BUFFERS); UNUSED(O); transpose_toms513(I, n, m, vl, (char *) (buf + 2*vl), (n+m)/2, buf); X(ifree)(buf); } static int applicable_toms513(const problem_rdft *p, planner *plnr, int dim0, int dim1, int dim2, INT *nbuf) { INT n = p->vecsz->dims[dim0].n; INT m = p->vecsz->dims[dim1].n; INT vl, vs; get_transpose_vec(p, dim2, &vl, &vs); *nbuf = 2*vl + ((n + m) / 2 * sizeof(char) + sizeof(R) - 1) / sizeof(R); return (!NO_SLOWP(plnr) && (vl > 8 || !NO_UGLYP(plnr)) /* UGLY for small vl */ && n != m && Ntuple_transposable(p->vecsz->dims + dim0, p->vecsz->dims + dim1, vl, vs)); } static int mkcldrn_toms513(const problem_rdft *p, planner *plnr, P *ego) { UNUSED(p); UNUSED(plnr); /* heuristic so that TOMS algorithm is last resort for small vl */ ego->super.super.ops.other += ego->n * ego->m * 2 * (ego->vl + 30); return 1; } static const transpose_adt adt_toms513 = { apply_toms513, applicable_toms513, mkcldrn_toms513, "rdft-transpose-toms513" }; /*-----------------------------------------------------------------------*/ /*-----------------------------------------------------------------------*/ /* generic stuff: */ static void awake(plan *ego_, enum wakefulness wakefulness) { P *ego = (P *) ego_; X(plan_awake)(ego->cld1, wakefulness); X(plan_awake)(ego->cld2, wakefulness); X(plan_awake)(ego->cld3, wakefulness); } static void print(const plan *ego_, printer *p) { const P *ego = (const P *) ego_; p->print(p, "(%s-%Dx%D%v", ego->slv->adt->nam, ego->n, ego->m, ego->vl); if (ego->cld1) p->print(p, "%(%p%)", ego->cld1); if (ego->cld2) p->print(p, "%(%p%)", ego->cld2); if (ego->cld3) p->print(p, "%(%p%)", ego->cld3); p->print(p, ")"); } static void destroy(plan *ego_) { P *ego = (P *) ego_; X(plan_destroy_internal)(ego->cld3); X(plan_destroy_internal)(ego->cld2); X(plan_destroy_internal)(ego->cld1); } static plan *mkplan(const solver *ego_, const problem *p_, planner *plnr) { const S *ego = (const S *) ego_; const problem_rdft *p; int dim0, dim1, dim2; INT nbuf, vs; P *pln; static const plan_adt padt = { X(rdft_solve), awake, print, destroy }; if (!applicable(ego_, p_, plnr, &dim0, &dim1, &dim2, &nbuf)) return (plan *) 0; p = (const problem_rdft *) p_; pln = MKPLAN_RDFT(P, &padt, ego->adt->apply); pln->n = p->vecsz->dims[dim0].n; pln->m = p->vecsz->dims[dim1].n; get_transpose_vec(p, dim2, &pln->vl, &vs); pln->nbuf = nbuf; pln->d = gcd(pln->n, pln->m); pln->nd = pln->n / pln->d; pln->md = pln->m / pln->d; pln->slv = ego; X(ops_zero)(&pln->super.super.ops); /* mkcldrn is responsible for ops */ pln->cld1 = pln->cld2 = pln->cld3 = 0; if (!ego->adt->mkcldrn(p, plnr, pln)) { X(plan_destroy_internal)(&(pln->super.super)); return 0; } return &(pln->super.super); } static solver *mksolver(const transpose_adt *adt) { static const solver_adt sadt = { PROBLEM_RDFT, mkplan, 0 }; S *slv = MKSOLVER(S, &sadt); slv->adt = adt; return &(slv->super); } void X(rdft_vrank3_transpose_register)(planner *p) { unsigned i; static const transpose_adt *const adts[] = { &adt_gcd, &adt_cut, &adt_toms513 }; for (i = 0; i < sizeof(adts) / sizeof(adts[0]); ++i) REGISTER_SOLVER(p, mksolver(adts[i])); } fftw-3.3.3/rdft/buffered2.c0000644000175000017500000002511212054410050012346 00000000000000/* * Copyright (c) 2003, 2007-11 Matteo Frigo * Copyright (c) 2003, 2007-11 Massachusetts Institute of Technology * * This program is free software; you can redistribute it and/or modify * it under the terms of the GNU General Public License as published by * the Free Software Foundation; either version 2 of the License, or * (at your option) any later version. * * This program is distributed in the hope that it will be useful, * but WITHOUT ANY WARRANTY; without even the implied warranty of * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the * GNU General Public License for more details. * * You should have received a copy of the GNU General Public License * along with this program; if not, write to the Free Software * Foundation, Inc., 51 Franklin Street, Fifth Floor, Boston, MA 02110-1301 USA * */ /* buffering of rdft2. We always buffer the complex array */ #include "rdft.h" #include "dft.h" typedef struct { solver super; int maxnbuf_ndx; } S; static const INT maxnbufs[] = { 8, 256 }; typedef struct { plan_rdft2 super; plan *cld, *cldcpy, *cldrest; INT n, vl, nbuf, bufdist; INT ivs_by_nbuf, ovs_by_nbuf; INT ioffset, roffset; } P; /* transform a vector input with the help of bufs */ static void apply_r2hc(const plan *ego_, R *r0, R *r1, R *cr, R *ci) { const P *ego = (const P *) ego_; plan_rdft2 *cld = (plan_rdft2 *) ego->cld; plan_dft *cldcpy = (plan_dft *) ego->cldcpy; INT i, vl = ego->vl, nbuf = ego->nbuf; INT ivs_by_nbuf = ego->ivs_by_nbuf, ovs_by_nbuf = ego->ovs_by_nbuf; R *bufs = (R *)MALLOC(sizeof(R) * nbuf * ego->bufdist, BUFFERS); R *bufr = bufs + ego->roffset; R *bufi = bufs + ego->ioffset; plan_rdft2 *cldrest; for (i = nbuf; i <= vl; i += nbuf) { /* transform to bufs: */ cld->apply((plan *) cld, r0, r1, bufr, bufi); r0 += ivs_by_nbuf; r1 += ivs_by_nbuf; /* copy back */ cldcpy->apply((plan *) cldcpy, bufr, bufi, cr, ci); cr += ovs_by_nbuf; ci += ovs_by_nbuf; } X(ifree)(bufs); /* Do the remaining transforms, if any: */ cldrest = (plan_rdft2 *) ego->cldrest; cldrest->apply((plan *) cldrest, r0, r1, cr, ci); } /* for hc2r problems, copy the input into buffer, and then transform buffer->output, which allows for destruction of the buffer */ static void apply_hc2r(const plan *ego_, R *r0, R *r1, R *cr, R *ci) { const P *ego = (const P *) ego_; plan_rdft2 *cld = (plan_rdft2 *) ego->cld; plan_dft *cldcpy = (plan_dft *) ego->cldcpy; INT i, vl = ego->vl, nbuf = ego->nbuf; INT ivs_by_nbuf = ego->ivs_by_nbuf, ovs_by_nbuf = ego->ovs_by_nbuf; R *bufs = (R *)MALLOC(sizeof(R) * nbuf * ego->bufdist, BUFFERS); R *bufr = bufs + ego->roffset; R *bufi = bufs + ego->ioffset; plan_rdft2 *cldrest; for (i = nbuf; i <= vl; i += nbuf) { /* copy input into bufs: */ cldcpy->apply((plan *) cldcpy, cr, ci, bufr, bufi); cr += ivs_by_nbuf; ci += ivs_by_nbuf; /* transform to output */ cld->apply((plan *) cld, r0, r1, bufr, bufi); r0 += ovs_by_nbuf; r1 += ovs_by_nbuf; } X(ifree)(bufs); /* Do the remaining transforms, if any: */ cldrest = (plan_rdft2 *) ego->cldrest; cldrest->apply((plan *) cldrest, r0, r1, cr, ci); } static void awake(plan *ego_, enum wakefulness wakefulness) { P *ego = (P *) ego_; X(plan_awake)(ego->cld, wakefulness); X(plan_awake)(ego->cldcpy, wakefulness); X(plan_awake)(ego->cldrest, wakefulness); } static void destroy(plan *ego_) { P *ego = (P *) ego_; X(plan_destroy_internal)(ego->cldrest); X(plan_destroy_internal)(ego->cldcpy); X(plan_destroy_internal)(ego->cld); } static void print(const plan *ego_, printer *p) { const P *ego = (const P *) ego_; p->print(p, "(rdft2-buffered-%D%v/%D-%D%(%p%)%(%p%)%(%p%))", ego->n, ego->nbuf, ego->vl, ego->bufdist % ego->n, ego->cld, ego->cldcpy, ego->cldrest); } static int applicable0(const S *ego, const problem *p_, const planner *plnr) { const problem_rdft2 *p = (const problem_rdft2 *) p_; iodim *d = p->sz->dims; if (1 && p->vecsz->rnk <= 1 && p->sz->rnk == 1 /* we assume even n throughout */ && (d[0].n % 2) == 0 /* and we only consider these two cases */ && (p->kind == R2HC || p->kind == HC2R) ) { INT vl, ivs, ovs; X(tensor_tornk1)(p->vecsz, &vl, &ivs, &ovs); if (X(toobig)(d[0].n) && CONSERVE_MEMORYP(plnr)) return 0; /* if this solver is redundant, in the sense that a solver of lower index generates the same plan, then prune this solver */ if (X(nbuf_redundant)(d[0].n, vl, ego->maxnbuf_ndx, maxnbufs, NELEM(maxnbufs))) return 0; if (p->r0 != p->cr) { if (p->kind == HC2R) { /* Allow HC2R problems only if the input is to be preserved. This solver sets NO_DESTROY_INPUT, which prevents infinite loops */ return (NO_DESTROY_INPUTP(plnr)); } else { /* In principle, the buffered transforms might be useful when working out of place. However, in order to prevent infinite loops in the planner, we require that the output stride of the buffered transforms be greater than 2. */ return (d[0].os > 2); } } /* * If the problem is in place, the input/output strides must * be the same or the whole thing must fit in the buffer. */ if (X(rdft2_inplace_strides(p, RNK_MINFTY))) return 1; if (/* fits into buffer: */ ((p->vecsz->rnk == 0) || (X(nbuf)(d[0].n, p->vecsz->dims[0].n, maxnbufs[ego->maxnbuf_ndx]) == p->vecsz->dims[0].n))) return 1; } return 0; } static int applicable(const S *ego, const problem *p_, const planner *plnr) { const problem_rdft2 *p; if (NO_BUFFERINGP(plnr)) return 0; if (!applicable0(ego, p_, plnr)) return 0; p = (const problem_rdft2 *) p_; if (p->kind == HC2R) { if (NO_UGLYP(plnr)) { /* UGLY if in-place and too big, since the problem could be solved via transpositions */ if (p->r0 == p->cr && X(toobig)(p->sz->dims[0].n)) return 0; } } else { if (NO_UGLYP(plnr)) { if (p->r0 != p->cr || X(toobig)(p->sz->dims[0].n)) return 0; } } return 1; } static plan *mkplan(const solver *ego_, const problem *p_, planner *plnr) { P *pln; const S *ego = (const S *)ego_; plan *cld = (plan *) 0; plan *cldcpy = (plan *) 0; plan *cldrest = (plan *) 0; const problem_rdft2 *p = (const problem_rdft2 *) p_; R *bufs = (R *) 0; INT nbuf = 0, bufdist, n, vl; INT ivs, ovs, ioffset, roffset, id, od; static const plan_adt padt = { X(rdft2_solve), awake, print, destroy }; if (!applicable(ego, p_, plnr)) goto nada; n = X(tensor_sz)(p->sz); X(tensor_tornk1)(p->vecsz, &vl, &ivs, &ovs); nbuf = X(nbuf)(n, vl, maxnbufs[ego->maxnbuf_ndx]); bufdist = X(bufdist)(n + 2, vl); /* complex-side rdft2 stores N+2 real numbers */ A(nbuf > 0); /* attempt to keep real and imaginary part in the same order, so as to allow optimizations in the the copy plan */ roffset = (p->cr - p->ci > 0) ? (INT)1 : (INT)0; ioffset = 1 - roffset; /* initial allocation for the purpose of planning */ bufs = (R *) MALLOC(sizeof(R) * nbuf * bufdist, BUFFERS); id = ivs * (nbuf * (vl / nbuf)); od = ovs * (nbuf * (vl / nbuf)); if (p->kind == R2HC) { /* allow destruction of input if problem is in place */ cld = X(mkplan_f_d)( plnr, X(mkproblem_rdft2_d)( X(mktensor_1d)(n, p->sz->dims[0].is, 2), X(mktensor_1d)(nbuf, ivs, bufdist), TAINT(p->r0, ivs * nbuf), TAINT(p->r1, ivs * nbuf), bufs + roffset, bufs + ioffset, p->kind), 0, 0, (p->r0 == p->cr) ? NO_DESTROY_INPUT : 0); if (!cld) goto nada; /* copying back from the buffer is a rank-0 DFT: */ cldcpy = X(mkplan_d)( plnr, X(mkproblem_dft_d)( X(mktensor_0d)(), X(mktensor_2d)(nbuf, bufdist, ovs, n/2+1, 2, p->sz->dims[0].os), bufs + roffset, bufs + ioffset, TAINT(p->cr, ovs * nbuf), TAINT(p->ci, ovs * nbuf) )); if (!cldcpy) goto nada; X(ifree)(bufs); bufs = 0; cldrest = X(mkplan_d)(plnr, X(mkproblem_rdft2_d)( X(tensor_copy)(p->sz), X(mktensor_1d)(vl % nbuf, ivs, ovs), p->r0 + id, p->r1 + id, p->cr + od, p->ci + od, p->kind)); if (!cldrest) goto nada; pln = MKPLAN_RDFT2(P, &padt, apply_r2hc); } else { /* allow destruction of buffer */ cld = X(mkplan_f_d)( plnr, X(mkproblem_rdft2_d)( X(mktensor_1d)(n, 2, p->sz->dims[0].os), X(mktensor_1d)(nbuf, bufdist, ovs), TAINT(p->r0, ovs * nbuf), TAINT(p->r1, ovs * nbuf), bufs + roffset, bufs + ioffset, p->kind), 0, 0, NO_DESTROY_INPUT); if (!cld) goto nada; /* copying input into buffer is a rank-0 DFT: */ cldcpy = X(mkplan_d)( plnr, X(mkproblem_dft_d)( X(mktensor_0d)(), X(mktensor_2d)(nbuf, ivs, bufdist, n/2+1, p->sz->dims[0].is, 2), TAINT(p->cr, ivs * nbuf), TAINT(p->ci, ivs * nbuf), bufs + roffset, bufs + ioffset)); if (!cldcpy) goto nada; X(ifree)(bufs); bufs = 0; cldrest = X(mkplan_d)(plnr, X(mkproblem_rdft2_d)( X(tensor_copy)(p->sz), X(mktensor_1d)(vl % nbuf, ivs, ovs), p->r0 + od, p->r1 + od, p->cr + id, p->ci + id, p->kind)); if (!cldrest) goto nada; pln = MKPLAN_RDFT2(P, &padt, apply_hc2r); } pln->cld = cld; pln->cldcpy = cldcpy; pln->cldrest = cldrest; pln->n = n; pln->vl = vl; pln->ivs_by_nbuf = ivs * nbuf; pln->ovs_by_nbuf = ovs * nbuf; pln->roffset = roffset; pln->ioffset = ioffset; pln->nbuf = nbuf; pln->bufdist = bufdist; { opcnt t; X(ops_add)(&cld->ops, &cldcpy->ops, &t); X(ops_madd)(vl / nbuf, &t, &cldrest->ops, &pln->super.super.ops); } return &(pln->super.super); nada: X(ifree0)(bufs); X(plan_destroy_internal)(cldrest); X(plan_destroy_internal)(cldcpy); X(plan_destroy_internal)(cld); return (plan *) 0; } static solver *mksolver(int maxnbuf_ndx) { static const solver_adt sadt = { PROBLEM_RDFT2, mkplan, 0 }; S *slv = MKSOLVER(S, &sadt); slv->maxnbuf_ndx = maxnbuf_ndx; return &(slv->super); } void X(rdft2_buffered_register)(planner *p) { size_t i; for (i = 0; i < NELEM(maxnbufs); ++i) REGISTER_SOLVER(p, mksolver(i)); } fftw-3.3.3/rdft/direct2.c0000644000175000017500000001043312054410050012036 00000000000000/* * Copyright (c) 2003, 2007-11 Matteo Frigo * Copyright (c) 2003, 2007-11 Massachusetts Institute of Technology * * This program is free software; you can redistribute it and/or modify * it under the terms of the GNU General Public License as published by * the Free Software Foundation; either version 2 of the License, or * (at your option) any later version. * * This program is distributed in the hope that it will be useful, * but WITHOUT ANY WARRANTY; without even the implied warranty of * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the * GNU General Public License for more details. * * You should have received a copy of the GNU General Public License * along with this program; if not, write to the Free Software * Foundation, Inc., 51 Franklin Street, Fifth Floor, Boston, MA 02110-1301 USA * */ /* direct RDFT2 R2HC/HC2R solver, if we have a codelet */ #include "rdft.h" typedef struct { solver super; const kr2c_desc *desc; kr2c k; } S; typedef struct { plan_rdft2 super; stride rs, cs; INT vl; INT ivs, ovs; kr2c k; const S *slv; INT ilast; } P; static void apply(const plan *ego_, R *r0, R *r1, R *cr, R *ci) { const P *ego = (const P *) ego_; ASSERT_ALIGNED_DOUBLE; ego->k(r0, r1, cr, ci, ego->rs, ego->cs, ego->cs, ego->vl, ego->ivs, ego->ovs); } static void apply_r2hc(const plan *ego_, R *r0, R *r1, R *cr, R *ci) { const P *ego = (const P *) ego_; INT i, vl = ego->vl, ovs = ego->ovs; ASSERT_ALIGNED_DOUBLE; ego->k(r0, r1, cr, ci, ego->rs, ego->cs, ego->cs, vl, ego->ivs, ovs); for (i = 0; i < vl; ++i, ci += ovs) ci[0] = ci[ego->ilast] = 0; } static void destroy(plan *ego_) { P *ego = (P *) ego_; X(stride_destroy)(ego->rs); X(stride_destroy)(ego->cs); } static void print(const plan *ego_, printer *p) { const P *ego = (const P *) ego_; const S *s = ego->slv; p->print(p, "(rdft2-%s-direct-%D%v \"%s\")", X(rdft_kind_str)(s->desc->genus->kind), s->desc->n, ego->vl, s->desc->nam); } static int applicable(const solver *ego_, const problem *p_) { const S *ego = (const S *) ego_; const kr2c_desc *desc = ego->desc; const problem_rdft2 *p = (const problem_rdft2 *) p_; INT vl; INT ivs, ovs; return ( 1 && p->sz->rnk == 1 && p->vecsz->rnk <= 1 && p->sz->dims[0].n == desc->n && p->kind == desc->genus->kind /* check strides etc */ && X(tensor_tornk1)(p->vecsz, &vl, &ivs, &ovs) && (0 /* can operate out-of-place */ || p->r0 != p->cr /* * can compute one transform in-place, no matter * what the strides are. */ || p->vecsz->rnk == 0 /* can operate in-place as long as strides are the same */ || X(rdft2_inplace_strides)(p, RNK_MINFTY) ) ); } static plan *mkplan(const solver *ego_, const problem *p_, planner *plnr) { const S *ego = (const S *) ego_; P *pln; const problem_rdft2 *p; iodim *d; int r2hc_kindp; static const plan_adt padt = { X(rdft2_solve), X(null_awake), print, destroy }; UNUSED(plnr); if (!applicable(ego_, p_)) return (plan *)0; p = (const problem_rdft2 *) p_; r2hc_kindp = R2HC_KINDP(p->kind); A(r2hc_kindp || HC2R_KINDP(p->kind)); pln = MKPLAN_RDFT2(P, &padt, p->kind == R2HC ? apply_r2hc : apply); d = p->sz->dims; pln->k = ego->k; pln->rs = X(mkstride)(d->n, r2hc_kindp ? d->is : d->os); pln->cs = X(mkstride)(d->n, r2hc_kindp ? d->os : d->is); X(tensor_tornk1)(p->vecsz, &pln->vl, &pln->ivs, &pln->ovs); /* Nyquist freq., if any */ pln->ilast = (d->n % 2) ? 0 : (d->n/2) * d->os; pln->slv = ego; X(ops_zero)(&pln->super.super.ops); X(ops_madd2)(pln->vl / ego->desc->genus->vl, &ego->desc->ops, &pln->super.super.ops); if (p->kind == R2HC) pln->super.super.ops.other += 2 * pln->vl; /* + 2 stores */ pln->super.super.could_prune_now_p = 1; return &(pln->super.super); } /* constructor */ solver *X(mksolver_rdft2_direct)(kr2c k, const kr2c_desc *desc) { static const solver_adt sadt = { PROBLEM_RDFT2, mkplan, 0 }; S *slv = MKSOLVER(S, &sadt); slv->k = k; slv->desc = desc; return &(slv->super); } fftw-3.3.3/rdft/nop2.c0000644000175000017500000000465512054410050011371 00000000000000/* * Copyright (c) 2003, 2007-11 Matteo Frigo * Copyright (c) 2003, 2007-11 Massachusetts Institute of Technology * * This program is free software; you can redistribute it and/or modify * it under the terms of the GNU General Public License as published by * the Free Software Foundation; either version 2 of the License, or * (at your option) any later version. * * This program is distributed in the hope that it will be useful, * but WITHOUT ANY WARRANTY; without even the implied warranty of * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the * GNU General Public License for more details. * * You should have received a copy of the GNU General Public License * along with this program; if not, write to the Free Software * Foundation, Inc., 51 Franklin Street, Fifth Floor, Boston, MA 02110-1301 USA * */ /* plans for vrank -infty RDFT2s (nothing to do), as well as in-place rank-0 HC2R. Note that in-place rank-0 R2HC is *not* a no-op, because we have to set the imaginary parts of the output to zero. */ #include "rdft.h" static void apply(const plan *ego_, R *r0, R *r1, R *cr, R *ci) { UNUSED(ego_); UNUSED(r0); UNUSED(r1); UNUSED(cr); UNUSED(ci); } static int applicable(const solver *ego_, const problem *p_) { const problem_rdft2 *p = (const problem_rdft2 *) p_; UNUSED(ego_); return(0 /* case 1 : -infty vector rank */ || (p->vecsz->rnk == RNK_MINFTY) /* case 2 : rank-0 in-place rdft, except that R2HC is not a no-op because it sets the imaginary part to 0 */ || (1 && p->kind != R2HC && p->sz->rnk == 0 && FINITE_RNK(p->vecsz->rnk) && (p->r0 == p->cr) && X(rdft2_inplace_strides)(p, RNK_MINFTY) )); } static void print(const plan *ego, printer *p) { UNUSED(ego); p->print(p, "(rdft2-nop)"); } static plan *mkplan(const solver *ego, const problem *p, planner *plnr) { static const plan_adt padt = { X(rdft2_solve), X(null_awake), print, X(plan_null_destroy) }; plan_rdft2 *pln; UNUSED(plnr); if (!applicable(ego, p)) return (plan *) 0; pln = MKPLAN_RDFT2(plan_rdft2, &padt, apply); X(ops_zero)(&pln->super.ops); return &(pln->super); } static solver *mksolver(void) { static const solver_adt sadt = { PROBLEM_RDFT2, mkplan, 0 }; return MKSOLVER(solver, &sadt); } void X(rdft2_nop_register)(planner *p) { REGISTER_SOLVER(p, mksolver()); } fftw-3.3.3/rdft/rank0-rdft2.c0000644000175000017500000001144312054410050012536 00000000000000/* * Copyright (c) 2003, 2007-11 Matteo Frigo * Copyright (c) 2003, 2007-11 Massachusetts Institute of Technology * * This program is free software; you can redistribute it and/or modify * it under the terms of the GNU General Public License as published by * the Free Software Foundation; either version 2 of the License, or * (at your option) any later version. * * This program is distributed in the hope that it will be useful, * but WITHOUT ANY WARRANTY; without even the implied warranty of * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the * GNU General Public License for more details. * * You should have received a copy of the GNU General Public License * along with this program; if not, write to the Free Software * Foundation, Inc., 51 Franklin Street, Fifth Floor, Boston, MA 02110-1301 USA * */ /* plans for rank-0 RDFT2 (copy operations, plus setting 0 imag. parts) */ #include "rdft.h" #ifdef HAVE_STRING_H #include /* for memcpy() */ #endif typedef struct { solver super; } S; typedef struct { plan_rdft super; INT vl; INT ivs, ovs; plan *cldcpy; } P; static int applicable(const problem *p_) { const problem_rdft2 *p = (const problem_rdft2 *) p_; return (1 && p->sz->rnk == 0 && (p->kind == HC2R || (1 && p->kind == R2HC && p->vecsz->rnk <= 1 && ((p->r0 != p->cr) || X(rdft2_inplace_strides)(p, RNK_MINFTY)) )) ); } static void apply_r2hc(const plan *ego_, R *r0, R *r1, R *cr, R *ci) { const P *ego = (const P *) ego_; INT i, vl = ego->vl; INT ivs = ego->ivs, ovs = ego->ovs; UNUSED(r1); /* rank-0 has no real odd-index elements */ for (i = 4; i <= vl; i += 4) { R x0, x1, x2, x3; x0 = *r0; r0 += ivs; x1 = *r0; r0 += ivs; x2 = *r0; r0 += ivs; x3 = *r0; r0 += ivs; *cr = x0; cr += ovs; *ci = K(0.0); ci += ovs; *cr = x1; cr += ovs; *ci = K(0.0); ci += ovs; *cr = x2; cr += ovs; *ci = K(0.0); ci += ovs; *cr = x3; cr += ovs; *ci = K(0.0); ci += ovs; } for (; i < vl + 4; ++i) { R x0; x0 = *r0; r0 += ivs; *cr = x0; cr += ovs; *ci = K(0.0); ci += ovs; } } /* in-place r2hc rank-0: set imaginary parts of output to 0 */ static void apply_r2hc_inplace(const plan *ego_, R *r0, R *r1, R *cr, R *ci) { const P *ego = (const P *) ego_; INT i, vl = ego->vl; INT ovs = ego->ovs; UNUSED(r0); UNUSED(r1); UNUSED(cr); for (i = 4; i <= vl; i += 4) { *ci = K(0.0); ci += ovs; *ci = K(0.0); ci += ovs; *ci = K(0.0); ci += ovs; *ci = K(0.0); ci += ovs; } for (; i < vl + 4; ++i) { *ci = K(0.0); ci += ovs; } } /* a rank-0 HC2R rdft2 problem is just a copy from cr to r0, so we can use a rank-0 rdft plan */ static void apply_hc2r(const plan *ego_, R *r0, R *r1, R *cr, R *ci) { const P *ego = (const P *) ego_; plan_rdft *cldcpy = (plan_rdft *) ego->cldcpy; UNUSED(ci); UNUSED(r1); cldcpy->apply((plan *) cldcpy, cr, r0); } static void awake(plan *ego_, enum wakefulness wakefulness) { P *ego = (P *) ego_; if (ego->cldcpy) X(plan_awake)(ego->cldcpy, wakefulness); } static void destroy(plan *ego_) { P *ego = (P *) ego_; if (ego->cldcpy) X(plan_destroy_internal)(ego->cldcpy); } static void print(const plan *ego_, printer *p) { const P *ego = (const P *) ego_; if (ego->cldcpy) p->print(p, "(rdft2-hc2r-rank0%(%p%))", ego->cldcpy); else p->print(p, "(rdft2-r2hc-rank0%v)", ego->vl); } static plan *mkplan(const solver *ego_, const problem *p_, planner *plnr) { const problem_rdft2 *p; plan *cldcpy = (plan *) 0; P *pln; static const plan_adt padt = { X(rdft2_solve), awake, print, destroy }; UNUSED(ego_); if (!applicable(p_)) return (plan *) 0; p = (const problem_rdft2 *) p_; if (p->kind == HC2R) { cldcpy = X(mkplan_d)(plnr, X(mkproblem_rdft_0_d)( X(tensor_copy)(p->vecsz), p->cr, p->r0)); if (!cldcpy) return (plan *) 0; } pln = MKPLAN_RDFT2(P, &padt, p->kind == R2HC ? (p->r0 == p->cr ? apply_r2hc_inplace : apply_r2hc) : apply_hc2r); if (p->kind == R2HC) X(tensor_tornk1)(p->vecsz, &pln->vl, &pln->ivs, &pln->ovs); pln->cldcpy = cldcpy; if (p->kind == R2HC) { /* vl loads, 2*vl stores */ X(ops_other)(3 * pln->vl, &pln->super.super.ops); } else { pln->super.super.ops = cldcpy->ops; } return &(pln->super.super); } static solver *mksolver(void) { static const solver_adt sadt = { PROBLEM_RDFT2, mkplan, 0 }; S *slv = MKSOLVER(S, &sadt); return &(slv->super); } void X(rdft2_rank0_register)(planner *p) { REGISTER_SOLVER(p, mksolver()); } fftw-3.3.3/rdft/rank-geq2-rdft2.c0000644000175000017500000001473512054410050013321 00000000000000/* * Copyright (c) 2003, 2007-11 Matteo Frigo * Copyright (c) 2003, 2007-11 Massachusetts Institute of Technology * * This program is free software; you can redistribute it and/or modify * it under the terms of the GNU General Public License as published by * the Free Software Foundation; either version 2 of the License, or * (at your option) any later version. * * This program is distributed in the hope that it will be useful, * but WITHOUT ANY WARRANTY; without even the implied warranty of * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the * GNU General Public License for more details. * * You should have received a copy of the GNU General Public License * along with this program; if not, write to the Free Software * Foundation, Inc., 51 Franklin Street, Fifth Floor, Boston, MA 02110-1301 USA * */ /* plans for RDFT2 of rank >= 2 (multidimensional) */ #include "rdft.h" #include "dft.h" typedef struct { solver super; int spltrnk; const int *buddies; int nbuddies; } S; typedef struct { plan_dft super; plan *cldr, *cldc; const S *solver; } P; static void apply_r2hc(const plan *ego_, R *r0, R *r1, R *cr, R *ci) { const P *ego = (const P *) ego_; { plan_rdft2 *cldr = (plan_rdft2 *) ego->cldr; cldr->apply((plan *) cldr, r0, r1, cr, ci); } { plan_dft *cldc = (plan_dft *) ego->cldc; cldc->apply((plan *) cldc, cr, ci, cr, ci); } } static void apply_hc2r(const plan *ego_, R *r0, R *r1, R *cr, R *ci) { const P *ego = (const P *) ego_; { plan_dft *cldc = (plan_dft *) ego->cldc; cldc->apply((plan *) cldc, ci, cr, ci, cr); } { plan_rdft2 *cldr = (plan_rdft2 *) ego->cldr; cldr->apply((plan *) cldr, r0, r1, cr, ci); } } static void awake(plan *ego_, enum wakefulness wakefulness) { P *ego = (P *) ego_; X(plan_awake)(ego->cldr, wakefulness); X(plan_awake)(ego->cldc, wakefulness); } static void destroy(plan *ego_) { P *ego = (P *) ego_; X(plan_destroy_internal)(ego->cldr); X(plan_destroy_internal)(ego->cldc); } static void print(const plan *ego_, printer *p) { const P *ego = (const P *) ego_; const S *s = ego->solver; p->print(p, "(rdft2-rank>=2/%d%(%p%)%(%p%))", s->spltrnk, ego->cldr, ego->cldc); } static int picksplit(const S *ego, const tensor *sz, int *rp) { A(sz->rnk > 1); /* cannot split rnk <= 1 */ if (!X(pickdim)(ego->spltrnk, ego->buddies, ego->nbuddies, sz, 1, rp)) return 0; *rp += 1; /* convert from dim. index to rank */ if (*rp >= sz->rnk) /* split must reduce rank */ return 0; return 1; } static int applicable0(const solver *ego_, const problem *p_, int *rp, const planner *plnr) { const problem_rdft2 *p = (const problem_rdft2 *) p_; const S *ego = (const S *)ego_; return (1 && FINITE_RNK(p->sz->rnk) && FINITE_RNK(p->vecsz->rnk) /* FIXME: multidimensional R2HCII ? */ && (p->kind == R2HC || p->kind == HC2R) && p->sz->rnk >= 2 && picksplit(ego, p->sz, rp) && (0 /* can work out-of-place, but HC2R destroys input */ || (p->r0 != p->cr && (p->kind == R2HC || !NO_DESTROY_INPUTP(plnr))) /* FIXME: what are sufficient conditions for inplace? */ || (p->r0 == p->cr)) ); } /* TODO: revise this. */ static int applicable(const solver *ego_, const problem *p_, const planner *plnr, int *rp) { const S *ego = (const S *)ego_; if (!applicable0(ego_, p_, rp, plnr)) return 0; if (NO_RANK_SPLITSP(plnr) && (ego->spltrnk != ego->buddies[0])) return 0; if (NO_UGLYP(plnr)) { const problem_rdft2 *p = (const problem_rdft2 *) p_; /* Heuristic: if the vector stride is greater than the transform size, don't use (prefer to do the vector loop first with a vrank-geq1 plan). */ if (p->vecsz->rnk > 0 && X(tensor_min_stride)(p->vecsz) > X(rdft2_tensor_max_index)(p->sz, p->kind)) return 0; } return 1; } static plan *mkplan(const solver *ego_, const problem *p_, planner *plnr) { const S *ego = (const S *) ego_; const problem_rdft2 *p; P *pln; plan *cldr = 0, *cldc = 0; tensor *sz1, *sz2, *vecszi, *sz2i; int spltrnk; inplace_kind k; problem *cldp; static const plan_adt padt = { X(rdft2_solve), awake, print, destroy }; if (!applicable(ego_, p_, plnr, &spltrnk)) return (plan *) 0; p = (const problem_rdft2 *) p_; X(tensor_split)(p->sz, &sz1, spltrnk, &sz2); k = p->kind == R2HC ? INPLACE_OS : INPLACE_IS; vecszi = X(tensor_copy_inplace)(p->vecsz, k); sz2i = X(tensor_copy_inplace)(sz2, k); /* complex data is ~half of real */ sz2i->dims[sz2i->rnk - 1].n = sz2i->dims[sz2i->rnk - 1].n/2 + 1; cldr = X(mkplan_d)(plnr, X(mkproblem_rdft2_d)(X(tensor_copy)(sz2), X(tensor_append)(p->vecsz, sz1), p->r0, p->r1, p->cr, p->ci, p->kind)); if (!cldr) goto nada; if (p->kind == R2HC) cldp = X(mkproblem_dft_d)(X(tensor_copy_inplace)(sz1, k), X(tensor_append)(vecszi, sz2i), p->cr, p->ci, p->cr, p->ci); else /* HC2R must swap re/im parts to get IDFT */ cldp = X(mkproblem_dft_d)(X(tensor_copy_inplace)(sz1, k), X(tensor_append)(vecszi, sz2i), p->ci, p->cr, p->ci, p->cr); cldc = X(mkplan_d)(plnr, cldp); if (!cldc) goto nada; pln = MKPLAN_RDFT2(P, &padt, p->kind == R2HC ? apply_r2hc : apply_hc2r); pln->cldr = cldr; pln->cldc = cldc; pln->solver = ego; X(ops_add)(&cldr->ops, &cldc->ops, &pln->super.super.ops); X(tensor_destroy4)(sz2i, vecszi, sz2, sz1); return &(pln->super.super); nada: X(plan_destroy_internal)(cldr); X(plan_destroy_internal)(cldc); X(tensor_destroy4)(sz2i, vecszi, sz2, sz1); return (plan *) 0; } static solver *mksolver(int spltrnk, const int *buddies, int nbuddies) { static const solver_adt sadt = { PROBLEM_RDFT2, mkplan, 0 }; S *slv = MKSOLVER(S, &sadt); slv->spltrnk = spltrnk; slv->buddies = buddies; slv->nbuddies = nbuddies; return &(slv->super); } void X(rdft2_rank_geq2_register)(planner *p) { int i; static const int buddies[] = { 1, 0, -2 }; const int nbuddies = (int)(sizeof(buddies) / sizeof(buddies[0])); for (i = 0; i < nbuddies; ++i) REGISTER_SOLVER(p, mksolver(buddies[i], buddies, nbuddies)); /* FIXME: Should we try more buddies? See also dft/rank-geq2. */ } fftw-3.3.3/rdft/plan2.c0000644000175000017500000000205312054410050011515 00000000000000/* * Copyright (c) 2003, 2007-11 Matteo Frigo * Copyright (c) 2003, 2007-11 Massachusetts Institute of Technology * * This program is free software; you can redistribute it and/or modify * it under the terms of the GNU General Public License as published by * the Free Software Foundation; either version 2 of the License, or * (at your option) any later version. * * This program is distributed in the hope that it will be useful, * but WITHOUT ANY WARRANTY; without even the implied warranty of * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the * GNU General Public License for more details. * * You should have received a copy of the GNU General Public License * along with this program; if not, write to the Free Software * Foundation, Inc., 51 Franklin Street, Fifth Floor, Boston, MA 02110-1301 USA * */ #include "rdft.h" plan *X(mkplan_rdft2)(size_t size, const plan_adt *adt, rdft2apply apply) { plan_rdft2 *ego; ego = (plan_rdft2 *) X(mkplan)(size, adt); ego->apply = apply; return &(ego->super); } fftw-3.3.3/rdft/problem2.c0000644000175000017500000001374512054410050012235 00000000000000/* * Copyright (c) 2003, 2007-11 Matteo Frigo * Copyright (c) 2003, 2007-11 Massachusetts Institute of Technology * * This program is free software; you can redistribute it and/or modify * it under the terms of the GNU General Public License as published by * the Free Software Foundation; either version 2 of the License, or * (at your option) any later version. * * This program is distributed in the hope that it will be useful, * but WITHOUT ANY WARRANTY; without even the implied warranty of * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the * GNU General Public License for more details. * * You should have received a copy of the GNU General Public License * along with this program; if not, write to the Free Software * Foundation, Inc., 51 Franklin Street, Fifth Floor, Boston, MA 02110-1301 USA * */ #include "dft.h" #include "rdft.h" #include static void destroy(problem *ego_) { problem_rdft2 *ego = (problem_rdft2 *) ego_; X(tensor_destroy2)(ego->vecsz, ego->sz); X(ifree)(ego_); } static void hash(const problem *p_, md5 *m) { const problem_rdft2 *p = (const problem_rdft2 *) p_; X(md5puts)(m, "rdft2"); X(md5int)(m, p->r0 == p->cr); X(md5INT)(m, p->r1 - p->r0); X(md5INT)(m, p->ci - p->cr); X(md5int)(m, X(alignment_of)(p->r0)); X(md5int)(m, X(alignment_of)(p->r1)); X(md5int)(m, X(alignment_of)(p->cr)); X(md5int)(m, X(alignment_of)(p->ci)); X(md5int)(m, p->kind); X(tensor_md5)(m, p->sz); X(tensor_md5)(m, p->vecsz); } static void print(const problem *ego_, printer *p) { const problem_rdft2 *ego = (const problem_rdft2 *) ego_; p->print(p, "(rdft2 %d %d %T %T)", (int)(ego->cr == ego->r0), (int)(ego->kind), ego->sz, ego->vecsz); } static void recur(const iodim *dims, int rnk, R *I0, R *I1) { if (rnk == RNK_MINFTY) return; else if (rnk == 0) I0[0] = K(0.0); else if (rnk > 0) { INT i, n = dims[0].n, is = dims[0].is; if (rnk == 1) { for (i = 0; i < n - 1; i += 2) { *I0 = *I1 = K(0.0); I0 += is; I1 += is; } if (i < n) *I0 = K(0.0); } else { for (i = 0; i < n; ++i) recur(dims + 1, rnk - 1, I0 + i * is, I1 + i * is); } } } static void vrecur(const iodim *vdims, int vrnk, const iodim *dims, int rnk, R *I0, R *I1) { if (vrnk == RNK_MINFTY) return; else if (vrnk == 0) recur(dims, rnk, I0, I1); else if (vrnk > 0) { INT i, n = vdims[0].n, is = vdims[0].is; for (i = 0; i < n; ++i) vrecur(vdims + 1, vrnk - 1, dims, rnk, I0 + i * is, I1 + i * is); } } INT X(rdft2_complex_n)(INT real_n, rdft_kind kind) { switch (kind) { case R2HC: case HC2R: return (real_n / 2) + 1; case R2HCII: case HC2RIII: return (real_n + 1) / 2; default: /* can't happen */ A(0); return 0; } } static void zero(const problem *ego_) { const problem_rdft2 *ego = (const problem_rdft2 *) ego_; if (R2HC_KINDP(ego->kind)) { /* FIXME: can we avoid the double recursion somehow? */ vrecur(ego->vecsz->dims, ego->vecsz->rnk, ego->sz->dims, ego->sz->rnk, UNTAINT(ego->r0), UNTAINT(ego->r1)); } else { tensor *sz; tensor *sz2 = X(tensor_copy)(ego->sz); int rnk = sz2->rnk; if (rnk > 0) /* ~half as many complex outputs */ sz2->dims[rnk-1].n = X(rdft2_complex_n)(sz2->dims[rnk-1].n, ego->kind); sz = X(tensor_append)(ego->vecsz, sz2); X(tensor_destroy)(sz2); X(dft_zerotens)(sz, UNTAINT(ego->cr), UNTAINT(ego->ci)); X(tensor_destroy)(sz); } } static const problem_adt padt = { PROBLEM_RDFT2, hash, zero, print, destroy }; problem *X(mkproblem_rdft2)(const tensor *sz, const tensor *vecsz, R *r0, R *r1, R *cr, R *ci, rdft_kind kind) { problem_rdft2 *ego; A(kind == R2HC || kind == R2HCII || kind == HC2R || kind == HC2RIII); A(X(tensor_kosherp)(sz)); A(X(tensor_kosherp)(vecsz)); A(FINITE_RNK(sz->rnk)); /* require in-place problems to use r0 == cr */ if (UNTAINT(r0) == UNTAINT(ci)) return X(mkproblem_unsolvable)(); /* FIXME: should check UNTAINT(r1) == UNTAINT(cr) but only if odd elements exist, which requires compressing the tensors first */ if (UNTAINT(r0) == UNTAINT(cr)) r0 = cr = JOIN_TAINT(r0, cr); ego = (problem_rdft2 *)X(mkproblem)(sizeof(problem_rdft2), &padt); if (sz->rnk > 1) { /* have to compress rnk-1 dims separately, ugh */ tensor *szc = X(tensor_copy_except)(sz, sz->rnk - 1); tensor *szr = X(tensor_copy_sub)(sz, sz->rnk - 1, 1); tensor *szcc = X(tensor_compress)(szc); if (szcc->rnk > 0) ego->sz = X(tensor_append)(szcc, szr); else ego->sz = X(tensor_compress)(szr); X(tensor_destroy2)(szc, szr); X(tensor_destroy)(szcc); } else { ego->sz = X(tensor_compress)(sz); } ego->vecsz = X(tensor_compress_contiguous)(vecsz); ego->r0 = r0; ego->r1 = r1; ego->cr = cr; ego->ci = ci; ego->kind = kind; A(FINITE_RNK(ego->sz->rnk)); return &(ego->super); } /* Same as X(mkproblem_rdft2), but also destroy input tensors. */ problem *X(mkproblem_rdft2_d)(tensor *sz, tensor *vecsz, R *r0, R *r1, R *cr, R *ci, rdft_kind kind) { problem *p = X(mkproblem_rdft2)(sz, vecsz, r0, r1, cr, ci, kind); X(tensor_destroy2)(vecsz, sz); return p; } /* Same as X(mkproblem_rdft2_d), but with only one R pointer. Used by the API. */ problem *X(mkproblem_rdft2_d_3pointers)(tensor *sz, tensor *vecsz, R *r0, R *cr, R *ci, rdft_kind kind) { problem *p; int rnk = sz->rnk; R *r1; if (rnk == 0) r1 = r0; else if (R2HC_KINDP(kind)) { r1 = r0 + sz->dims[rnk-1].is; sz->dims[rnk-1].is *= 2; } else { r1 = r0 + sz->dims[rnk-1].os; sz->dims[rnk-1].os *= 2; } p = X(mkproblem_rdft2)(sz, vecsz, r0, r1, cr, ci, kind); X(tensor_destroy2)(vecsz, sz); return p; } fftw-3.3.3/rdft/solve2.c0000644000175000017500000000223712054410050011717 00000000000000/* * Copyright (c) 2003, 2007-11 Matteo Frigo * Copyright (c) 2003, 2007-11 Massachusetts Institute of Technology * * This program is free software; you can redistribute it and/or modify * it under the terms of the GNU General Public License as published by * the Free Software Foundation; either version 2 of the License, or * (at your option) any later version. * * This program is distributed in the hope that it will be useful, * but WITHOUT ANY WARRANTY; without even the implied warranty of * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the * GNU General Public License for more details. * * You should have received a copy of the GNU General Public License * along with this program; if not, write to the Free Software * Foundation, Inc., 51 Franklin Street, Fifth Floor, Boston, MA 02110-1301 USA * */ #include "rdft.h" /* use the apply() operation for RDFT2 problems */ void X(rdft2_solve)(const plan *ego_, const problem *p_) { const plan_rdft2 *ego = (const plan_rdft2 *) ego_; const problem_rdft2 *p = (const problem_rdft2 *) p_; ego->apply(ego_, UNTAINT(p->r0), UNTAINT(p->r1), UNTAINT(p->cr), UNTAINT(p->ci)); } fftw-3.3.3/rdft/vrank-geq1-rdft2.c0000644000175000017500000001410112054410050013471 00000000000000/* * Copyright (c) 2003, 2007-11 Matteo Frigo * Copyright (c) 2003, 2007-11 Massachusetts Institute of Technology * * This program is free software; you can redistribute it and/or modify * it under the terms of the GNU General Public License as published by * the Free Software Foundation; either version 2 of the License, or * (at your option) any later version. * * This program is distributed in the hope that it will be useful, * but WITHOUT ANY WARRANTY; without even the implied warranty of * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the * GNU General Public License for more details. * * You should have received a copy of the GNU General Public License * along with this program; if not, write to the Free Software * Foundation, Inc., 51 Franklin Street, Fifth Floor, Boston, MA 02110-1301 USA * */ /* Plans for handling vector transform loops. These are *just* the loops, and rely on child plans for the actual RDFT2s. They form a wrapper around solvers that don't have apply functions for non-null vectors. vrank-geq1-rdft2 plans also recursively handle the case of multi-dimensional vectors, obviating the need for most solvers to deal with this. We can also play games here, such as reordering the vector loops. Each vrank-geq1-rdft2 plan reduces the vector rank by 1, picking out a dimension determined by the vecloop_dim field of the solver. */ #include "rdft.h" typedef struct { solver super; int vecloop_dim; const int *buddies; int nbuddies; } S; typedef struct { plan_rdft2 super; plan *cld; INT vl; INT rvs, cvs; const S *solver; } P; static void apply(const plan *ego_, R *r0, R *r1, R *cr, R *ci) { const P *ego = (const P *) ego_; INT i, vl = ego->vl; INT rvs = ego->rvs, cvs = ego->cvs; rdft2apply cldapply = ((plan_rdft2 *) ego->cld)->apply; for (i = 0; i < vl; ++i) { cldapply(ego->cld, r0 + i * rvs, r1 + i * rvs, cr + i * cvs, ci + i * cvs); } } static void awake(plan *ego_, enum wakefulness wakefulness) { P *ego = (P *) ego_; X(plan_awake)(ego->cld, wakefulness); } static void destroy(plan *ego_) { P *ego = (P *) ego_; X(plan_destroy_internal)(ego->cld); } static void print(const plan *ego_, printer *p) { const P *ego = (const P *) ego_; const S *s = ego->solver; p->print(p, "(rdft2-vrank>=1-x%D/%d%(%p%))", ego->vl, s->vecloop_dim, ego->cld); } static int pickdim(const S *ego, const tensor *vecsz, int oop, int *dp) { return X(pickdim)(ego->vecloop_dim, ego->buddies, ego->nbuddies, vecsz, oop, dp); } static int applicable0(const solver *ego_, const problem *p_, int *dp) { const S *ego = (const S *) ego_; const problem_rdft2 *p = (const problem_rdft2 *) p_; if (FINITE_RNK(p->vecsz->rnk) && p->vecsz->rnk > 0 && pickdim(ego, p->vecsz, p->r0 != p->cr, dp)) { if (p->r0 != p->cr) return 1; /* can always operate out-of-place */ return(X(rdft2_inplace_strides)(p, *dp)); } return 0; } static int applicable(const solver *ego_, const problem *p_, const planner *plnr, int *dp) { const S *ego = (const S *)ego_; if (!applicable0(ego_, p_, dp)) return 0; /* fftw2 behavior */ if (NO_VRANK_SPLITSP(plnr) && (ego->vecloop_dim != ego->buddies[0])) return 0; if (NO_UGLYP(plnr)) { const problem_rdft2 *p = (const problem_rdft2 *) p_; iodim *d = p->vecsz->dims + *dp; /* Heuristic: if the transform is multi-dimensional, and the vector stride is less than the transform size, then we probably want to use a rank>=2 plan first in order to combine this vector with the transform-dimension vectors. */ if (p->sz->rnk > 1 && X(imin)(X(iabs)(d->is), X(iabs)(d->os)) < X(rdft2_tensor_max_index)(p->sz, p->kind) ) return 0; /* Heuristic: don't use a vrank-geq1 for rank-0 vrank-1 transforms, since this case is better handled by rank-0 solvers. */ if (p->sz->rnk == 0 && p->vecsz->rnk == 1) return 0; if (NO_NONTHREADEDP(plnr)) return 0; /* prefer threaded version */ } return 1; } static plan *mkplan(const solver *ego_, const problem *p_, planner *plnr) { const S *ego = (const S *) ego_; const problem_rdft2 *p; P *pln; plan *cld; int vdim; iodim *d; INT rvs, cvs; static const plan_adt padt = { X(rdft2_solve), awake, print, destroy }; if (!applicable(ego_, p_, plnr, &vdim)) return (plan *) 0; p = (const problem_rdft2 *) p_; d = p->vecsz->dims + vdim; A(d->n > 1); /* or else, p->ri + d->is etc. are invalid */ X(rdft2_strides)(p->kind, d, &rvs, &cvs); cld = X(mkplan_d)(plnr, X(mkproblem_rdft2_d)( X(tensor_copy)(p->sz), X(tensor_copy_except)(p->vecsz, vdim), TAINT(p->r0, rvs), TAINT(p->r1, rvs), TAINT(p->cr, cvs), TAINT(p->ci, cvs), p->kind)); if (!cld) return (plan *) 0; pln = MKPLAN_RDFT2(P, &padt, apply); pln->cld = cld; pln->vl = d->n; pln->rvs = rvs; pln->cvs = cvs; pln->solver = ego; X(ops_zero)(&pln->super.super.ops); pln->super.super.ops.other = 3.14159; /* magic to prefer codelet loops */ X(ops_madd2)(pln->vl, &cld->ops, &pln->super.super.ops); if (p->sz->rnk != 1 || (p->sz->dims[0].n > 128)) pln->super.super.pcost = pln->vl * cld->pcost; return &(pln->super.super); } static solver *mksolver(int vecloop_dim, const int *buddies, int nbuddies) { static const solver_adt sadt = { PROBLEM_RDFT2, mkplan, 0 }; S *slv = MKSOLVER(S, &sadt); slv->vecloop_dim = vecloop_dim; slv->buddies = buddies; slv->nbuddies = nbuddies; return &(slv->super); } void X(rdft2_vrank_geq1_register)(planner *p) { int i; /* FIXME: Should we try other vecloop_dim values? */ static const int buddies[] = { 1, -1 }; const int nbuddies = (int)(sizeof(buddies) / sizeof(buddies[0])); for (i = 0; i < nbuddies; ++i) REGISTER_SOLVER(p, mksolver(buddies[i], buddies, nbuddies)); } fftw-3.3.3/rdft/rdft2-rdft.c0000644000175000017500000002167512054410050012472 00000000000000/* * Copyright (c) 2003, 2007-11 Matteo Frigo * Copyright (c) 2003, 2007-11 Massachusetts Institute of Technology * * This program is free software; you can redistribute it and/or modify * it under the terms of the GNU General Public License as published by * the Free Software Foundation; either version 2 of the License, or * (at your option) any later version. * * This program is distributed in the hope that it will be useful, * but WITHOUT ANY WARRANTY; without even the implied warranty of * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the * GNU General Public License for more details. * * You should have received a copy of the GNU General Public License * along with this program; if not, write to the Free Software * Foundation, Inc., 51 Franklin Street, Fifth Floor, Boston, MA 02110-1301 USA * */ #include "rdft.h" typedef struct { solver super; } S; typedef struct { plan_rdft2 super; plan *cld, *cldrest; INT n, vl, nbuf, bufdist; INT cs, ivs, ovs; } P; /***************************************************************************/ /* FIXME: have alternate copy functions that push a vector loop inside the n loops? */ /* copy halfcomplex array r (contiguous) to complex (strided) array rio/iio. */ static void hc2c(INT n, R *r, R *rio, R *iio, INT os) { INT i; rio[0] = r[0]; iio[0] = 0; for (i = 1; i + i < n; ++i) { rio[i * os] = r[i]; iio[i * os] = r[n - i]; } if (i + i == n) { /* store the Nyquist frequency */ rio[i * os] = r[i]; iio[i * os] = K(0.0); } } /* reverse of hc2c */ static void c2hc(INT n, R *rio, R *iio, INT is, R *r) { INT i; r[0] = rio[0]; for (i = 1; i + i < n; ++i) { r[i] = rio[i * is]; r[n - i] = iio[i * is]; } if (i + i == n) /* store the Nyquist frequency */ r[i] = rio[i * is]; } /***************************************************************************/ static void apply_r2hc(const plan *ego_, R *r0, R *r1, R *cr, R *ci) { const P *ego = (const P *) ego_; plan_rdft *cld = (plan_rdft *) ego->cld; INT i, j, vl = ego->vl, nbuf = ego->nbuf, bufdist = ego->bufdist; INT n = ego->n; INT ivs = ego->ivs, ovs = ego->ovs, os = ego->cs; R *bufs = (R *)MALLOC(sizeof(R) * nbuf * bufdist, BUFFERS); plan_rdft2 *cldrest; for (i = nbuf; i <= vl; i += nbuf) { /* transform to bufs: */ cld->apply((plan *) cld, r0, bufs); r0 += ivs * nbuf; r1 += ivs * nbuf; /* copy back */ for (j = 0; j < nbuf; ++j, cr += ovs, ci += ovs) hc2c(n, bufs + j*bufdist, cr, ci, os); } X(ifree)(bufs); /* Do the remaining transforms, if any: */ cldrest = (plan_rdft2 *) ego->cldrest; cldrest->apply((plan *) cldrest, r0, r1, cr, ci); } static void apply_hc2r(const plan *ego_, R *r0, R *r1, R *cr, R *ci) { const P *ego = (const P *) ego_; plan_rdft *cld = (plan_rdft *) ego->cld; INT i, j, vl = ego->vl, nbuf = ego->nbuf, bufdist = ego->bufdist; INT n = ego->n; INT ivs = ego->ivs, ovs = ego->ovs, is = ego->cs; R *bufs = (R *)MALLOC(sizeof(R) * nbuf * bufdist, BUFFERS); plan_rdft2 *cldrest; for (i = nbuf; i <= vl; i += nbuf) { /* copy to bufs */ for (j = 0; j < nbuf; ++j, cr += ivs, ci += ivs) c2hc(n, cr, ci, is, bufs + j*bufdist); /* transform back: */ cld->apply((plan *) cld, bufs, r0); r0 += ovs * nbuf; r1 += ovs * nbuf; } X(ifree)(bufs); /* Do the remaining transforms, if any: */ cldrest = (plan_rdft2 *) ego->cldrest; cldrest->apply((plan *) cldrest, r0, r1, cr, ci); } static void awake(plan *ego_, enum wakefulness wakefulness) { P *ego = (P *) ego_; X(plan_awake)(ego->cld, wakefulness); X(plan_awake)(ego->cldrest, wakefulness); } static void destroy(plan *ego_) { P *ego = (P *) ego_; X(plan_destroy_internal)(ego->cldrest); X(plan_destroy_internal)(ego->cld); } static void print(const plan *ego_, printer *p) { const P *ego = (const P *) ego_; p->print(p, "(rdft2-rdft-%s-%D%v/%D-%D%(%p%)%(%p%))", ego->super.apply == apply_r2hc ? "r2hc" : "hc2r", ego->n, ego->nbuf, ego->vl, ego->bufdist % ego->n, ego->cld, ego->cldrest); } static INT min_nbuf(const problem_rdft2 *p, INT n, INT vl) { INT is, os, ivs, ovs; if (p->r0 != p->cr) return 1; if (X(rdft2_inplace_strides(p, RNK_MINFTY))) return 1; A(p->vecsz->rnk == 1); /* rank 0 and MINFTY are inplace */ X(rdft2_strides)(p->kind, p->sz->dims, &is, &os); X(rdft2_strides)(p->kind, p->vecsz->dims, &ivs, &ovs); /* handle one potentially common case: "contiguous" real and complex arrays, which overlap because of the differing sizes. */ if (n * X(iabs)(is) <= X(iabs)(ivs) && (n/2 + 1) * X(iabs)(os) <= X(iabs)(ovs) && ( ((p->cr - p->ci) <= X(iabs)(os)) || ((p->ci - p->cr) <= X(iabs)(os)) ) && ivs > 0 && ovs > 0) { INT vsmin = X(imin)(ivs, ovs); INT vsmax = X(imax)(ivs, ovs); return(((vsmax - vsmin) * vl + vsmin - 1) / vsmin); } return vl; /* punt: just buffer the whole vector */ } static int applicable0(const problem *p_, const S *ego, const planner *plnr) { const problem_rdft2 *p = (const problem_rdft2 *) p_; UNUSED(ego); return(1 && p->vecsz->rnk <= 1 && p->sz->rnk == 1 /* FIXME: does it make sense to do R2HCII ? */ && (p->kind == R2HC || p->kind == HC2R) /* real strides must allow for reduction to rdft */ && (2 * (p->r1 - p->r0) == (((p->kind == R2HC) ? p->sz->dims[0].is : p->sz->dims[0].os))) && !(X(toobig)(p->sz->dims[0].n) && CONSERVE_MEMORYP(plnr)) ); } static int applicable(const problem *p_, const S *ego, const planner *plnr) { const problem_rdft2 *p; if (NO_BUFFERINGP(plnr)) return 0; if (!applicable0(p_, ego, plnr)) return 0; p = (const problem_rdft2 *) p_; if (NO_UGLYP(plnr)) { if (p->r0 != p->cr) return 0; if (X(toobig)(p->sz->dims[0].n)) return 0; } return 1; } static plan *mkplan(const solver *ego_, const problem *p_, planner *plnr) { const S *ego = (const S *) ego_; P *pln; plan *cld = (plan *) 0; plan *cldrest = (plan *) 0; const problem_rdft2 *p = (const problem_rdft2 *) p_; R *bufs = (R *) 0; INT nbuf = 0, bufdist, n, vl; INT ivs, ovs, rs, id, od; static const plan_adt padt = { X(rdft2_solve), awake, print, destroy }; if (!applicable(p_, ego, plnr)) goto nada; n = p->sz->dims[0].n; X(tensor_tornk1)(p->vecsz, &vl, &ivs, &ovs); nbuf = X(imax)(X(nbuf)(n, vl, 0), min_nbuf(p, n, vl)); bufdist = X(bufdist)(n, vl); A(nbuf > 0); /* initial allocation for the purpose of planning */ bufs = (R *) MALLOC(sizeof(R) * nbuf * bufdist, BUFFERS); id = ivs * (nbuf * (vl / nbuf)); od = ovs * (nbuf * (vl / nbuf)); if (p->kind == R2HC) { cld = X(mkplan_f_d)( plnr, X(mkproblem_rdft_d)( X(mktensor_1d)(n, p->sz->dims[0].is/2, 1), X(mktensor_1d)(nbuf, ivs, bufdist), TAINT(p->r0, ivs * nbuf), bufs, &p->kind), 0, 0, (p->r0 == p->cr) ? NO_DESTROY_INPUT : 0); if (!cld) goto nada; X(ifree)(bufs); bufs = 0; cldrest = X(mkplan_d)(plnr, X(mkproblem_rdft2_d)( X(tensor_copy)(p->sz), X(mktensor_1d)(vl % nbuf, ivs, ovs), p->r0 + id, p->r1 + id, p->cr + od, p->ci + od, p->kind)); if (!cldrest) goto nada; pln = MKPLAN_RDFT2(P, &padt, apply_r2hc); } else { A(p->kind == HC2R); cld = X(mkplan_f_d)( plnr, X(mkproblem_rdft_d)( X(mktensor_1d)(n, 1, p->sz->dims[0].os/2), X(mktensor_1d)(nbuf, bufdist, ovs), bufs, TAINT(p->r0, ovs * nbuf), &p->kind), 0, 0, NO_DESTROY_INPUT); /* always ok to destroy bufs */ if (!cld) goto nada; X(ifree)(bufs); bufs = 0; cldrest = X(mkplan_d)(plnr, X(mkproblem_rdft2_d)( X(tensor_copy)(p->sz), X(mktensor_1d)(vl % nbuf, ivs, ovs), p->r0 + od, p->r1 + od, p->cr + id, p->ci + id, p->kind)); if (!cldrest) goto nada; pln = MKPLAN_RDFT2(P, &padt, apply_hc2r); } pln->cld = cld; pln->cldrest = cldrest; pln->n = n; pln->vl = vl; pln->ivs = ivs; pln->ovs = ovs; X(rdft2_strides)(p->kind, &p->sz->dims[0], &rs, &pln->cs); pln->nbuf = nbuf; pln->bufdist = bufdist; X(ops_madd)(vl / nbuf, &cld->ops, &cldrest->ops, &pln->super.super.ops); pln->super.super.ops.other += (p->kind == R2HC ? (n + 2) : n) * vl; return &(pln->super.super); nada: X(ifree0)(bufs); X(plan_destroy_internal)(cldrest); X(plan_destroy_internal)(cld); return (plan *) 0; } static solver *mksolver(void) { static const solver_adt sadt = { PROBLEM_RDFT2, mkplan, 0 }; S *slv = MKSOLVER(S, &sadt); return &(slv->super); } void X(rdft2_rdft_register)(planner *p) { REGISTER_SOLVER(p, mksolver()); } fftw-3.3.3/rdft/rdft2-tensor-max-index.c0000644000175000017500000000272512054410050014730 00000000000000/* * Copyright (c) 2003, 2007-11 Matteo Frigo * Copyright (c) 2003, 2007-11 Massachusetts Institute of Technology * * This program is free software; you can redistribute it and/or modify * it under the terms of the GNU General Public License as published by * the Free Software Foundation; either version 2 of the License, or * (at your option) any later version. * * This program is distributed in the hope that it will be useful, * but WITHOUT ANY WARRANTY; without even the implied warranty of * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the * GNU General Public License for more details. * * You should have received a copy of the GNU General Public License * along with this program; if not, write to the Free Software * Foundation, Inc., 51 Franklin Street, Fifth Floor, Boston, MA 02110-1301 USA * */ #include "rdft.h" /* like X(tensor_max_index), but takes into account the special n/2+1 final dimension for the complex output/input of an R2HC/HC2R transform. */ INT X(rdft2_tensor_max_index)(const tensor *sz, rdft_kind k) { int i; INT n = 0; A(FINITE_RNK(sz->rnk)); for (i = 0; i + 1 < sz->rnk; ++i) { const iodim *p = sz->dims + i; n += (p->n - 1) * X(imax)(X(iabs)(p->is), X(iabs)(p->os)); } if (i < sz->rnk) { const iodim *p = sz->dims + i; INT is, os; X(rdft2_strides)(k, p, &is, &os); n += X(imax)((p->n - 1) * X(iabs)(is), (p->n/2) * X(iabs)(os)); } return n; } fftw-3.3.3/rdft/rdft2-inplace-strides.c0000644000175000017500000000446212054410050014614 00000000000000/* * Copyright (c) 2003, 2007-11 Matteo Frigo * Copyright (c) 2003, 2007-11 Massachusetts Institute of Technology * * This program is free software; you can redistribute it and/or modify * it under the terms of the GNU General Public License as published by * the Free Software Foundation; either version 2 of the License, or * (at your option) any later version. * * This program is distributed in the hope that it will be useful, * but WITHOUT ANY WARRANTY; without even the implied warranty of * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the * GNU General Public License for more details. * * You should have received a copy of the GNU General Public License * along with this program; if not, write to the Free Software * Foundation, Inc., 51 Franklin Street, Fifth Floor, Boston, MA 02110-1301 USA * */ #include "rdft.h" /* Check if the vecsz/sz strides are consistent with the problem being in-place for vecsz.dim[vdim], or for all dimensions if vdim == RNK_MINFTY. We can't just use tensor_inplace_strides because rdft transforms have the unfortunate property of differing input and output sizes. This routine is not exhaustive; we only return 1 for the most common case. */ int X(rdft2_inplace_strides)(const problem_rdft2 *p, int vdim) { INT N, Nc; INT rs, cs; int i; for (i = 0; i + 1 < p->sz->rnk; ++i) if (p->sz->dims[i].is != p->sz->dims[i].os) return 0; if (!FINITE_RNK(p->vecsz->rnk) || p->vecsz->rnk == 0) return 1; if (!FINITE_RNK(vdim)) { /* check all vector dimensions */ for (vdim = 0; vdim < p->vecsz->rnk; ++vdim) if (!X(rdft2_inplace_strides)(p, vdim)) return 0; return 1; } A(vdim < p->vecsz->rnk); if (p->sz->rnk == 0) return(p->vecsz->dims[vdim].is == p->vecsz->dims[vdim].os); N = X(tensor_sz)(p->sz); Nc = (N / p->sz->dims[p->sz->rnk-1].n) * (p->sz->dims[p->sz->rnk-1].n/2 + 1); X(rdft2_strides)(p->kind, p->sz->dims + p->sz->rnk - 1, &rs, &cs); /* the factor of 2 comes from the fact that RS is the stride of p->r0 and p->r1, which is twice as large as the strides in the r2r case */ return(p->vecsz->dims[vdim].is == p->vecsz->dims[vdim].os && (X(iabs)(2 * p->vecsz->dims[vdim].os) >= X(imax)(2 * Nc * X(iabs)(cs), N * X(iabs)(rs)))); } fftw-3.3.3/rdft/rdft2-strides.c0000644000175000017500000000243512054410050013201 00000000000000/* * Copyright (c) 2003, 2007-11 Matteo Frigo * Copyright (c) 2003, 2007-11 Massachusetts Institute of Technology * * This program is free software; you can redistribute it and/or modify * it under the terms of the GNU General Public License as published by * the Free Software Foundation; either version 2 of the License, or * (at your option) any later version. * * This program is distributed in the hope that it will be useful, * but WITHOUT ANY WARRANTY; without even the implied warranty of * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the * GNU General Public License for more details. * * You should have received a copy of the GNU General Public License * along with this program; if not, write to the Free Software * Foundation, Inc., 51 Franklin Street, Fifth Floor, Boston, MA 02110-1301 USA * */ #include "rdft.h" /* Deal with annoyance because the tensor (is,os) applies to (r,rio/iio) for R2HC and vice-versa for HC2R. We originally had (is,os) always apply to (r,rio/iio), but this causes other headaches with the tensor functions. */ void X(rdft2_strides)(rdft_kind kind, const iodim *d, INT *rs, INT *cs) { if (kind == R2HC) { *rs = d->is; *cs = d->os; } else { A(kind == HC2R); *rs = d->os; *cs = d->is; } } fftw-3.3.3/rdft/khc2c.c0000644000175000017500000000201112054410050011465 00000000000000/* * Copyright (c) 2003, 2007-11 Matteo Frigo * Copyright (c) 2003, 2007-11 Massachusetts Institute of Technology * * This program is free software; you can redistribute it and/or modify * it under the terms of the GNU General Public License as published by * the Free Software Foundation; either version 2 of the License, or * (at your option) any later version. * * This program is distributed in the hope that it will be useful, * but WITHOUT ANY WARRANTY; without even the implied warranty of * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the * GNU General Public License for more details. * * You should have received a copy of the GNU General Public License * along with this program; if not, write to the Free Software * Foundation, Inc., 51 Franklin Street, Fifth Floor, Boston, MA 02110-1301 USA * */ #include "ct-hc2c.h" void X(khc2c_register)(planner *p, khc2c codelet, const hc2c_desc *desc, hc2c_kind hc2ckind) { X(regsolver_hc2c_direct)(p, codelet, desc, hc2ckind); } fftw-3.3.3/rdft/ct-hc2c.h0000644000175000017500000000331512054410050011733 00000000000000/* * Copyright (c) 2003, 2007-11 Matteo Frigo * Copyright (c) 2003, 2007-11 Massachusetts Institute of Technology * * This program is free software; you can redistribute it and/or modify * it under the terms of the GNU General Public License as published by * the Free Software Foundation; either version 2 of the License, or * (at your option) any later version. * * This program is distributed in the hope that it will be useful, * but WITHOUT ANY WARRANTY; without even the implied warranty of * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the * GNU General Public License for more details. * * You should have received a copy of the GNU General Public License * along with this program; if not, write to the Free Software * Foundation, Inc., 51 Franklin Street, Fifth Floor, Boston, MA 02110-1301 USA * */ #include "rdft.h" typedef void (*hc2capply) (const plan *ego, R *cr, R *ci); typedef struct hc2c_solver_s hc2c_solver; typedef plan *(*hc2c_mkinferior)(const hc2c_solver *ego, rdft_kind kind, INT r, INT rs, INT m, INT ms, INT v, INT vs, R *cr, R *ci, planner *plnr); typedef struct { plan super; hc2capply apply; } plan_hc2c; extern plan *X(mkplan_hc2c)(size_t size, const plan_adt *adt, hc2capply apply); #define MKPLAN_HC2C(type, adt, apply) \ (type *)X(mkplan_hc2c)(sizeof(type), adt, apply) struct hc2c_solver_s { solver super; INT r; hc2c_mkinferior mkcldw; hc2c_kind hc2ckind; }; hc2c_solver *X(mksolver_hc2c)(size_t size, INT r, hc2c_kind hc2ckind, hc2c_mkinferior mkcldw); void X(regsolver_hc2c_direct)(planner *plnr, khc2c codelet, const hc2c_desc *desc, hc2c_kind hc2ckind); fftw-3.3.3/rdft/ct-hc2c.c0000644000175000017500000001623712054410050011735 00000000000000/* * Copyright (c) 2003, 2007-11 Matteo Frigo * Copyright (c) 2003, 2007-11 Massachusetts Institute of Technology * * This program is free software; you can redistribute it and/or modify * it under the terms of the GNU General Public License as published by * the Free Software Foundation; either version 2 of the License, or * (at your option) any later version. * * This program is distributed in the hope that it will be useful, * but WITHOUT ANY WARRANTY; without even the implied warranty of * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the * GNU General Public License for more details. * * You should have received a copy of the GNU General Public License * along with this program; if not, write to the Free Software * Foundation, Inc., 51 Franklin Street, Fifth Floor, Boston, MA 02110-1301 USA * */ #include "ct-hc2c.h" #include "dft.h" typedef struct { plan_rdft2 super; plan *cld; plan *cldw; INT r; } P; static void apply_dit(const plan *ego_, R *r0, R *r1, R *cr, R *ci) { const P *ego = (const P *) ego_; plan_rdft *cld; plan_hc2c *cldw; UNUSED(r1); cld = (plan_rdft *) ego->cld; cld->apply(ego->cld, r0, cr); cldw = (plan_hc2c *) ego->cldw; cldw->apply(ego->cldw, cr, ci); } static void apply_dif(const plan *ego_, R *r0, R *r1, R *cr, R *ci) { const P *ego = (const P *) ego_; plan_rdft *cld; plan_hc2c *cldw; UNUSED(r1); cldw = (plan_hc2c *) ego->cldw; cldw->apply(ego->cldw, cr, ci); cld = (plan_rdft *) ego->cld; cld->apply(ego->cld, cr, r0); } static void apply_dit_dft(const plan *ego_, R *r0, R *r1, R *cr, R *ci) { const P *ego = (const P *) ego_; plan_dft *cld; plan_hc2c *cldw; cld = (plan_dft *) ego->cld; cld->apply(ego->cld, r0, r1, cr, ci); cldw = (plan_hc2c *) ego->cldw; cldw->apply(ego->cldw, cr, ci); } static void apply_dif_dft(const plan *ego_, R *r0, R *r1, R *cr, R *ci) { const P *ego = (const P *) ego_; plan_dft *cld; plan_hc2c *cldw; cldw = (plan_hc2c *) ego->cldw; cldw->apply(ego->cldw, cr, ci); cld = (plan_dft *) ego->cld; cld->apply(ego->cld, ci, cr, r1, r0); } static void awake(plan *ego_, enum wakefulness wakefulness) { P *ego = (P *) ego_; X(plan_awake)(ego->cld, wakefulness); X(plan_awake)(ego->cldw, wakefulness); } static void destroy(plan *ego_) { P *ego = (P *) ego_; X(plan_destroy_internal)(ego->cldw); X(plan_destroy_internal)(ego->cld); } static void print(const plan *ego_, printer *p) { const P *ego = (const P *) ego_; p->print(p, "(rdft2-ct-%s/%D%(%p%)%(%p%))", (ego->super.apply == apply_dit || ego->super.apply == apply_dit_dft) ? "dit" : "dif", ego->r, ego->cldw, ego->cld); } static int applicable0(const hc2c_solver *ego, const problem *p_, planner *plnr) { const problem_rdft2 *p = (const problem_rdft2 *) p_; INT r; return (1 && p->sz->rnk == 1 && p->vecsz->rnk <= 1 && (/* either the problem is R2HC, which is solved by DIT */ (p->kind == R2HC) || /* or the problem is HC2R, in which case it is solved by DIF, which destroys the input */ (p->kind == HC2R && (p->r0 == p->cr || !NO_DESTROY_INPUTP(plnr)))) && ((r = X(choose_radix)(ego->r, p->sz->dims[0].n)) > 0) && p->sz->dims[0].n > r); } int X(hc2c_applicable)(const hc2c_solver *ego, const problem *p_, planner *plnr) { const problem_rdft2 *p; if (!applicable0(ego, p_, plnr)) return 0; p = (const problem_rdft2 *) p_; return (0 || p->vecsz->rnk == 0 || !NO_VRECURSEP(plnr) ); } static plan *mkplan(const solver *ego_, const problem *p_, planner *plnr) { const hc2c_solver *ego = (const hc2c_solver *) ego_; const problem_rdft2 *p; P *pln = 0; plan *cld = 0, *cldw = 0; INT n, r, m, v, ivs, ovs; iodim *d; static const plan_adt padt = { X(rdft2_solve), awake, print, destroy }; if (!X(hc2c_applicable)(ego, p_, plnr)) return (plan *) 0; p = (const problem_rdft2 *) p_; d = p->sz->dims; n = d[0].n; r = X(choose_radix)(ego->r, n); A((r % 2) == 0); m = n / r; X(tensor_tornk1)(p->vecsz, &v, &ivs, &ovs); switch (p->kind) { case R2HC: cldw = ego->mkcldw(ego, R2HC, r, m * d[0].os, m, d[0].os, v, ovs, p->cr, p->ci, plnr); if (!cldw) goto nada; switch (ego->hc2ckind) { case HC2C_VIA_RDFT: cld = X(mkplan_d)( plnr, X(mkproblem_rdft_1_d)( X(mktensor_1d)(m, (r/2)*d[0].is, d[0].os), X(mktensor_3d)( 2, p->r1 - p->r0, p->ci - p->cr, r / 2, d[0].is, m * d[0].os, v, ivs, ovs), p->r0, p->cr, R2HC) ); if (!cld) goto nada; pln = MKPLAN_RDFT2(P, &padt, apply_dit); break; case HC2C_VIA_DFT: cld = X(mkplan_d)( plnr, X(mkproblem_dft_d)( X(mktensor_1d)(m, (r/2)*d[0].is, d[0].os), X(mktensor_2d)( r / 2, d[0].is, m * d[0].os, v, ivs, ovs), p->r0, p->r1, p->cr, p->ci) ); if (!cld) goto nada; pln = MKPLAN_RDFT2(P, &padt, apply_dit_dft); break; } break; case HC2R: cldw = ego->mkcldw(ego, HC2R, r, m * d[0].is, m, d[0].is, v, ivs, p->cr, p->ci, plnr); if (!cldw) goto nada; switch (ego->hc2ckind) { case HC2C_VIA_RDFT: cld = X(mkplan_d)( plnr, X(mkproblem_rdft_1_d)( X(mktensor_1d)(m, d[0].is, (r/2)*d[0].os), X(mktensor_3d)( 2, p->ci - p->cr, p->r1 - p->r0, r / 2, m * d[0].is, d[0].os, v, ivs, ovs), p->cr, p->r0, HC2R) ); if (!cld) goto nada; pln = MKPLAN_RDFT2(P, &padt, apply_dif); break; case HC2C_VIA_DFT: cld = X(mkplan_d)( plnr, X(mkproblem_dft_d)( X(mktensor_1d)(m, d[0].is, (r/2)*d[0].os), X(mktensor_2d)( r / 2, m * d[0].is, d[0].os, v, ivs, ovs), p->ci, p->cr, p->r1, p->r0) ); if (!cld) goto nada; pln = MKPLAN_RDFT2(P, &padt, apply_dif_dft); break; } break; default: A(0); } pln->cld = cld; pln->cldw = cldw; pln->r = r; X(ops_add)(&cld->ops, &cldw->ops, &pln->super.super.ops); /* inherit could_prune_now_p attribute from cldw */ pln->super.super.could_prune_now_p = cldw->could_prune_now_p; return &(pln->super.super); nada: X(plan_destroy_internal)(cldw); X(plan_destroy_internal)(cld); return (plan *) 0; } hc2c_solver *X(mksolver_hc2c)(size_t size, INT r, hc2c_kind hc2ckind, hc2c_mkinferior mkcldw) { static const solver_adt sadt = { PROBLEM_RDFT2, mkplan, 0 }; hc2c_solver *slv = (hc2c_solver *)X(mksolver)(size, &sadt); slv->r = r; slv->hc2ckind = hc2ckind; slv->mkcldw = mkcldw; return slv; } plan *X(mkplan_hc2c)(size_t size, const plan_adt *adt, hc2capply apply) { plan_hc2c *ego; ego = (plan_hc2c *) X(mkplan)(size, adt); ego->apply = apply; return &(ego->super); } fftw-3.3.3/rdft/ct-hc2c-direct.c0000644000175000017500000002615012054410050013200 00000000000000/* * Copyright (c) 2003, 2007-11 Matteo Frigo * Copyright (c) 2003, 2007-11 Massachusetts Institute of Technology * * This program is free software; you can redistribute it and/or modify * it under the terms of the GNU General Public License as published by * the Free Software Foundation; either version 2 of the License, or * (at your option) any later version. * * This program is distributed in the hope that it will be useful, * but WITHOUT ANY WARRANTY; without even the implied warranty of * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the * GNU General Public License for more details. * * You should have received a copy of the GNU General Public License * along with this program; if not, write to the Free Software * Foundation, Inc., 51 Franklin Street, Fifth Floor, Boston, MA 02110-1301 USA * */ #include "ct-hc2c.h" typedef struct { hc2c_solver super; const hc2c_desc *desc; int bufferedp; khc2c k; } S; typedef struct { plan_hc2c super; khc2c k; plan *cld0, *cldm; /* children for 0th and middle butterflies */ INT r, m, v, extra_iter; INT ms, vs; stride rs, brs; twid *td; const S *slv; } P; /************************************************************* Nonbuffered code *************************************************************/ static void apply(const plan *ego_, R *cr, R *ci) { const P *ego = (const P *) ego_; plan_rdft2 *cld0 = (plan_rdft2 *) ego->cld0; plan_rdft2 *cldm = (plan_rdft2 *) ego->cldm; INT i, m = ego->m, v = ego->v; INT ms = ego->ms, vs = ego->vs; for (i = 0; i < v; ++i, cr += vs, ci += vs) { cld0->apply((plan *) cld0, cr, ci, cr, ci); ego->k(cr + ms, ci + ms, cr + (m-1)*ms, ci + (m-1)*ms, ego->td->W, ego->rs, 1, (m+1)/2, ms); cldm->apply((plan *) cldm, cr + (m/2)*ms, ci + (m/2)*ms, cr + (m/2)*ms, ci + (m/2)*ms); } } static void apply_extra_iter(const plan *ego_, R *cr, R *ci) { const P *ego = (const P *) ego_; plan_rdft2 *cld0 = (plan_rdft2 *) ego->cld0; plan_rdft2 *cldm = (plan_rdft2 *) ego->cldm; INT i, m = ego->m, v = ego->v; INT ms = ego->ms, vs = ego->vs; INT mm = (m-1)/2; for (i = 0; i < v; ++i, cr += vs, ci += vs) { cld0->apply((plan *) cld0, cr, ci, cr, ci); /* for 4-way SIMD when (m+1)/2-1 is odd: iterate over an even vector length MM-1, and then execute the last iteration as a 2-vector with vector stride 0. The twiddle factors of the second half of the last iteration are bogus, but we only store the results of the first half. */ ego->k(cr + ms, ci + ms, cr + (m-1)*ms, ci + (m-1)*ms, ego->td->W, ego->rs, 1, mm, ms); ego->k(cr + mm*ms, ci + mm*ms, cr + (m-mm)*ms, ci + (m-mm)*ms, ego->td->W, ego->rs, mm, mm+2, 0); cldm->apply((plan *) cldm, cr + (m/2)*ms, ci + (m/2)*ms, cr + (m/2)*ms, ci + (m/2)*ms); } } /************************************************************* Buffered code *************************************************************/ /* should not be 2^k to avoid associativity conflicts */ static INT compute_batchsize(INT radix) { /* round up to multiple of 4 */ radix += 3; radix &= -4; return (radix + 2); } static void dobatch(const P *ego, R *Rp, R *Ip, R *Rm, R *Im, INT mb, INT me, INT extra_iter, R *bufp) { INT b = WS(ego->brs, 1); INT rs = WS(ego->rs, 1); INT ms = ego->ms; R *bufm = bufp + b - 2; X(cpy2d_pair_ci)(Rp + mb * ms, Ip + mb * ms, bufp, bufp + 1, ego->r / 2, rs, b, me - mb, ms, 2); X(cpy2d_pair_ci)(Rm - mb * ms, Im - mb * ms, bufm, bufm + 1, ego->r / 2, rs, b, me - mb, -ms, -2); ego->k(bufp, bufp + 1, bufm, bufm + 1, ego->td->W, ego->brs, mb, me + extra_iter, 2); X(cpy2d_pair_co)(bufp, bufp + 1, Rp + mb * ms, Ip + mb * ms, ego->r / 2, b, rs, me - mb, 2, ms); X(cpy2d_pair_co)(bufm, bufm + 1, Rm - mb * ms, Im - mb * ms, ego->r / 2, b, rs, me - mb, -2, -ms); } static void apply_buf(const plan *ego_, R *cr, R *ci) { const P *ego = (const P *) ego_; plan_rdft2 *cld0 = (plan_rdft2 *) ego->cld0; plan_rdft2 *cldm = (plan_rdft2 *) ego->cldm; INT i, j, ms = ego->ms, v = ego->v; INT batchsz = compute_batchsize(ego->r); R *buf; INT mb = 1, me = (ego->m+1) / 2; size_t bufsz = ego->r * batchsz * 2 * sizeof(R); BUF_ALLOC(R *, buf, bufsz); for (i = 0; i < v; ++i, cr += ego->vs, ci += ego->vs) { R *Rp = cr; R *Ip = ci; R *Rm = cr + ego->m * ms; R *Im = ci + ego->m * ms; cld0->apply((plan *) cld0, Rp, Ip, Rp, Ip); for (j = mb; j + batchsz < me; j += batchsz) dobatch(ego, Rp, Ip, Rm, Im, j, j + batchsz, 0, buf); dobatch(ego, Rp, Ip, Rm, Im, j, me, ego->extra_iter, buf); cldm->apply((plan *) cldm, Rp + me * ms, Ip + me * ms, Rp + me * ms, Ip + me * ms); } BUF_FREE(buf, bufsz); } /************************************************************* common code *************************************************************/ static void awake(plan *ego_, enum wakefulness wakefulness) { P *ego = (P *) ego_; X(plan_awake)(ego->cld0, wakefulness); X(plan_awake)(ego->cldm, wakefulness); X(twiddle_awake)(wakefulness, &ego->td, ego->slv->desc->tw, ego->r * ego->m, ego->r, (ego->m - 1) / 2 + ego->extra_iter); } static void destroy(plan *ego_) { P *ego = (P *) ego_; X(plan_destroy_internal)(ego->cld0); X(plan_destroy_internal)(ego->cldm); X(stride_destroy)(ego->rs); X(stride_destroy)(ego->brs); } static void print(const plan *ego_, printer *p) { const P *ego = (const P *) ego_; const S *slv = ego->slv; const hc2c_desc *e = slv->desc; if (slv->bufferedp) p->print(p, "(hc2c-directbuf/%D-%D/%D/%D%v \"%s\"%(%p%)%(%p%))", compute_batchsize(ego->r), ego->r, X(twiddle_length)(ego->r, e->tw), ego->extra_iter, ego->v, e->nam, ego->cld0, ego->cldm); else p->print(p, "(hc2c-direct-%D/%D/%D%v \"%s\"%(%p%)%(%p%))", ego->r, X(twiddle_length)(ego->r, e->tw), ego->extra_iter, ego->v, e->nam, ego->cld0, ego->cldm); } static int applicable0(const S *ego, rdft_kind kind, INT r, INT rs, INT m, INT ms, INT v, INT vs, const R *cr, const R *ci, const planner *plnr, INT *extra_iter) { const hc2c_desc *e = ego->desc; UNUSED(v); return ( 1 && r == e->radix && kind == e->genus->kind /* first v-loop iteration */ && ((*extra_iter = 0, e->genus->okp(cr + ms, ci + ms, cr + (m-1)*ms, ci + (m-1)*ms, rs, 1, (m+1)/2, ms, plnr)) || (*extra_iter = 1, ((e->genus->okp(cr + ms, ci + ms, cr + (m-1)*ms, ci + (m-1)*ms, rs, 1, (m-1)/2, ms, plnr)) && (e->genus->okp(cr + ms, ci + ms, cr + (m-1)*ms, ci + (m-1)*ms, rs, (m-1)/2, (m-1)/2 + 2, 0, plnr))))) /* subsequent v-loop iterations */ && (cr += vs, ci += vs, 1) && e->genus->okp(cr + ms, ci + ms, cr + (m-1)*ms, ci + (m-1)*ms, rs, 1, (m+1)/2 - *extra_iter, ms, plnr) ); } static int applicable0_buf(const S *ego, rdft_kind kind, INT r, INT rs, INT m, INT ms, INT v, INT vs, const R *cr, const R *ci, const planner *plnr, INT *extra_iter) { const hc2c_desc *e = ego->desc; INT batchsz, brs; UNUSED(v); UNUSED(rs); UNUSED(ms); UNUSED(vs); return ( 1 && r == e->radix && kind == e->genus->kind /* ignore cr, ci, use buffer */ && (cr = (const R *)0, ci = cr + 1, batchsz = compute_batchsize(r), brs = 4 * batchsz, 1) && e->genus->okp(cr, ci, cr + brs - 2, ci + brs - 2, brs, 1, 1+batchsz, 2, plnr) && ((*extra_iter = 0, e->genus->okp(cr, ci, cr + brs - 2, ci + brs - 2, brs, 1, 1 + (((m-1)/2) % batchsz), 2, plnr)) || (*extra_iter = 1, e->genus->okp(cr, ci, cr + brs - 2, ci + brs - 2, brs, 1, 1 + 1 + (((m-1)/2) % batchsz), 2, plnr))) ); } static int applicable(const S *ego, rdft_kind kind, INT r, INT rs, INT m, INT ms, INT v, INT vs, R *cr, R *ci, const planner *plnr, INT *extra_iter) { if (ego->bufferedp) { if (!applicable0_buf(ego, kind, r, rs, m, ms, v, vs, cr, ci, plnr, extra_iter)) return 0; } else { if (!applicable0(ego, kind, r, rs, m, ms, v, vs, cr, ci, plnr, extra_iter)) return 0; } if (NO_UGLYP(plnr) && X(ct_uglyp)((ego->bufferedp? (INT)512 : (INT)16), v, m * r, r)) return 0; return 1; } static plan *mkcldw(const hc2c_solver *ego_, rdft_kind kind, INT r, INT rs, INT m, INT ms, INT v, INT vs, R *cr, R *ci, planner *plnr) { const S *ego = (const S *) ego_; P *pln; const hc2c_desc *e = ego->desc; plan *cld0 = 0, *cldm = 0; INT imid = (m / 2) * ms; INT extra_iter; static const plan_adt padt = { 0, awake, print, destroy }; if (!applicable(ego, kind, r, rs, m, ms, v, vs, cr, ci, plnr, &extra_iter)) return (plan *)0; cld0 = X(mkplan_d)( plnr, X(mkproblem_rdft2_d)(X(mktensor_1d)(r, rs, rs), X(mktensor_0d)(), TAINT(cr, vs), TAINT(ci, vs), TAINT(cr, vs), TAINT(ci, vs), kind)); if (!cld0) goto nada; cldm = X(mkplan_d)( plnr, X(mkproblem_rdft2_d)(((m % 2) ? X(mktensor_0d)() : X(mktensor_1d)(r, rs, rs) ), X(mktensor_0d)(), TAINT(cr + imid, vs), TAINT(ci + imid, vs), TAINT(cr + imid, vs), TAINT(ci + imid, vs), kind == R2HC ? R2HCII : HC2RIII)); if (!cldm) goto nada; if (ego->bufferedp) pln = MKPLAN_HC2C(P, &padt, apply_buf); else pln = MKPLAN_HC2C(P, &padt, extra_iter ? apply_extra_iter : apply); pln->k = ego->k; pln->td = 0; pln->r = r; pln->rs = X(mkstride)(r, rs); pln->m = m; pln->ms = ms; pln->v = v; pln->vs = vs; pln->slv = ego; pln->brs = X(mkstride)(r, 4 * compute_batchsize(r)); pln->cld0 = cld0; pln->cldm = cldm; pln->extra_iter = extra_iter; X(ops_zero)(&pln->super.super.ops); X(ops_madd2)(v * (((m - 1) / 2) / e->genus->vl), &e->ops, &pln->super.super.ops); X(ops_madd2)(v, &cld0->ops, &pln->super.super.ops); X(ops_madd2)(v, &cldm->ops, &pln->super.super.ops); if (ego->bufferedp) pln->super.super.ops.other += 4 * r * m * v; return &(pln->super.super); nada: X(plan_destroy_internal)(cld0); X(plan_destroy_internal)(cldm); return 0; } static void regone(planner *plnr, khc2c codelet, const hc2c_desc *desc, hc2c_kind hc2ckind, int bufferedp) { S *slv = (S *)X(mksolver_hc2c)(sizeof(S), desc->radix, hc2ckind, mkcldw); slv->k = codelet; slv->desc = desc; slv->bufferedp = bufferedp; REGISTER_SOLVER(plnr, &(slv->super.super)); } void X(regsolver_hc2c_direct)(planner *plnr, khc2c codelet, const hc2c_desc *desc, hc2c_kind hc2ckind) { regone(plnr, codelet, desc, hc2ckind, /* bufferedp */0); regone(plnr, codelet, desc, hc2ckind, /* bufferedp */1); } fftw-3.3.3/rdft/scalar/0000755000175000017500000000000012054411110011660 500000000000000fftw-3.3.3/rdft/scalar/Makefile.am0000644000175000017500000000036312054410050013640 00000000000000AM_CPPFLAGS = -I$(top_srcdir)/kernel -I$(top_srcdir)/rdft SUBDIRS = r2cf r2cb r2r noinst_LTLIBRARIES = librdft_scalar.la librdft_scalar_la_SOURCES = hb.h r2cb.h r2cbIII.h hf.h hfb.c r2c.c \ r2cf.h r2cfII.h r2r.c r2r.h hc2c.c hc2cf.h hc2cb.h fftw-3.3.3/rdft/scalar/Makefile.in0000644000175000017500000005260512054410200013654 00000000000000# Makefile.in generated by automake 1.11.6 from Makefile.am. # @configure_input@ # Copyright (C) 1994, 1995, 1996, 1997, 1998, 1999, 2000, 2001, 2002, # 2003, 2004, 2005, 2006, 2007, 2008, 2009, 2010, 2011 Free Software # Foundation, Inc. # This Makefile.in is free software; the Free Software Foundation # gives unlimited permission to copy and/or distribute it, # with or without modifications, as long as this notice is preserved. # This program is distributed in the hope that it will be useful, # but WITHOUT ANY WARRANTY, to the extent permitted by law; without # even the implied warranty of MERCHANTABILITY or FITNESS FOR A # PARTICULAR PURPOSE. @SET_MAKE@ VPATH = @srcdir@ am__make_dryrun = \ { \ am__dry=no; \ case $$MAKEFLAGS in \ *\\[\ \ ]*) \ echo 'am--echo: ; @echo "AM" OK' | $(MAKE) -f - 2>/dev/null \ | grep '^AM OK$$' >/dev/null || am__dry=yes;; \ *) \ for am__flg in $$MAKEFLAGS; do \ case $$am__flg in \ *=*|--*) ;; \ *n*) am__dry=yes; break;; \ esac; \ done;; \ esac; \ test $$am__dry = yes; \ } pkgdatadir = $(datadir)/@PACKAGE@ pkgincludedir = $(includedir)/@PACKAGE@ pkglibdir = $(libdir)/@PACKAGE@ pkglibexecdir = $(libexecdir)/@PACKAGE@ am__cd = CDPATH="$${ZSH_VERSION+.}$(PATH_SEPARATOR)" && cd install_sh_DATA = $(install_sh) -c -m 644 install_sh_PROGRAM = $(install_sh) -c install_sh_SCRIPT = $(install_sh) -c INSTALL_HEADER = $(INSTALL_DATA) transform = $(program_transform_name) NORMAL_INSTALL = : PRE_INSTALL = : POST_INSTALL = : NORMAL_UNINSTALL = : PRE_UNINSTALL = : POST_UNINSTALL = : build_triplet = @build@ host_triplet = @host@ subdir = rdft/scalar DIST_COMMON = $(srcdir)/Makefile.am $(srcdir)/Makefile.in ACLOCAL_M4 = $(top_srcdir)/aclocal.m4 am__aclocal_m4_deps = $(top_srcdir)/m4/acx_mpi.m4 \ $(top_srcdir)/m4/acx_pthread.m4 \ $(top_srcdir)/m4/ax_cc_maxopt.m4 \ $(top_srcdir)/m4/ax_check_compiler_flags.m4 \ $(top_srcdir)/m4/ax_compiler_vendor.m4 \ $(top_srcdir)/m4/ax_gcc_aligns_stack.m4 \ $(top_srcdir)/m4/ax_gcc_version.m4 \ $(top_srcdir)/m4/ax_openmp.m4 $(top_srcdir)/m4/libtool.m4 \ $(top_srcdir)/m4/ltoptions.m4 $(top_srcdir)/m4/ltsugar.m4 \ $(top_srcdir)/m4/ltversion.m4 $(top_srcdir)/m4/lt~obsolete.m4 \ $(top_srcdir)/configure.ac am__configure_deps = $(am__aclocal_m4_deps) $(CONFIGURE_DEPENDENCIES) \ $(ACLOCAL_M4) mkinstalldirs = $(install_sh) -d CONFIG_HEADER = $(top_builddir)/config.h CONFIG_CLEAN_FILES = CONFIG_CLEAN_VPATH_FILES = LTLIBRARIES = $(noinst_LTLIBRARIES) librdft_scalar_la_LIBADD = am_librdft_scalar_la_OBJECTS = hfb.lo r2c.lo r2r.lo hc2c.lo librdft_scalar_la_OBJECTS = $(am_librdft_scalar_la_OBJECTS) DEFAULT_INCLUDES = -I.@am__isrc@ -I$(top_builddir) depcomp = $(SHELL) $(top_srcdir)/depcomp am__depfiles_maybe = depfiles am__mv = mv -f COMPILE = $(CC) $(DEFS) $(DEFAULT_INCLUDES) $(INCLUDES) $(AM_CPPFLAGS) \ $(CPPFLAGS) $(AM_CFLAGS) $(CFLAGS) LTCOMPILE = $(LIBTOOL) --tag=CC $(AM_LIBTOOLFLAGS) $(LIBTOOLFLAGS) \ --mode=compile $(CC) $(DEFS) $(DEFAULT_INCLUDES) $(INCLUDES) \ $(AM_CPPFLAGS) $(CPPFLAGS) $(AM_CFLAGS) $(CFLAGS) CCLD = $(CC) LINK = $(LIBTOOL) --tag=CC $(AM_LIBTOOLFLAGS) $(LIBTOOLFLAGS) \ --mode=link $(CCLD) $(AM_CFLAGS) $(CFLAGS) $(AM_LDFLAGS) \ $(LDFLAGS) -o $@ SOURCES = $(librdft_scalar_la_SOURCES) DIST_SOURCES = $(librdft_scalar_la_SOURCES) RECURSIVE_TARGETS = all-recursive check-recursive dvi-recursive \ html-recursive info-recursive install-data-recursive \ install-dvi-recursive install-exec-recursive \ install-html-recursive install-info-recursive \ install-pdf-recursive install-ps-recursive install-recursive \ installcheck-recursive installdirs-recursive pdf-recursive \ ps-recursive uninstall-recursive am__can_run_installinfo = \ case $$AM_UPDATE_INFO_DIR in \ n|no|NO) false;; \ *) (install-info --version) >/dev/null 2>&1;; \ esac RECURSIVE_CLEAN_TARGETS = mostlyclean-recursive clean-recursive \ distclean-recursive maintainer-clean-recursive AM_RECURSIVE_TARGETS = $(RECURSIVE_TARGETS:-recursive=) \ $(RECURSIVE_CLEAN_TARGETS:-recursive=) tags TAGS ctags CTAGS \ distdir ETAGS = etags CTAGS = ctags DIST_SUBDIRS = $(SUBDIRS) DISTFILES = $(DIST_COMMON) $(DIST_SOURCES) $(TEXINFOS) $(EXTRA_DIST) am__relativize = \ dir0=`pwd`; \ sed_first='s,^\([^/]*\)/.*$$,\1,'; \ sed_rest='s,^[^/]*/*,,'; \ sed_last='s,^.*/\([^/]*\)$$,\1,'; \ sed_butlast='s,/*[^/]*$$,,'; \ while test -n "$$dir1"; do \ first=`echo "$$dir1" | sed -e "$$sed_first"`; \ if test "$$first" != "."; then \ if test "$$first" = ".."; then \ dir2=`echo "$$dir0" | sed -e "$$sed_last"`/"$$dir2"; \ dir0=`echo "$$dir0" | sed -e "$$sed_butlast"`; \ else \ first2=`echo "$$dir2" | sed -e "$$sed_first"`; \ if test "$$first2" = "$$first"; then \ dir2=`echo "$$dir2" | sed -e "$$sed_rest"`; \ else \ dir2="../$$dir2"; \ fi; \ dir0="$$dir0"/"$$first"; \ fi; \ fi; \ dir1=`echo "$$dir1" | sed -e "$$sed_rest"`; \ done; \ reldir="$$dir2" ACLOCAL = @ACLOCAL@ ALLOCA = @ALLOCA@ ALTIVEC_CFLAGS = @ALTIVEC_CFLAGS@ AMTAR = @AMTAR@ AR = @AR@ AS = @AS@ AUTOCONF = @AUTOCONF@ AUTOHEADER = @AUTOHEADER@ AUTOMAKE = @AUTOMAKE@ AVX_CFLAGS = @AVX_CFLAGS@ AWK = @AWK@ CC = @CC@ CCDEPMODE = @CCDEPMODE@ CFLAGS = @CFLAGS@ CHECK_PL_OPTS = @CHECK_PL_OPTS@ CPP = @CPP@ CPPFLAGS = @CPPFLAGS@ CYGPATH_W = @CYGPATH_W@ C_FFTW_R2R_KIND = @C_FFTW_R2R_KIND@ C_MPI_FINT = @C_MPI_FINT@ DEFS = @DEFS@ DEPDIR = @DEPDIR@ DLLTOOL = @DLLTOOL@ DSYMUTIL = @DSYMUTIL@ DUMPBIN = @DUMPBIN@ ECHO_C = @ECHO_C@ ECHO_N = @ECHO_N@ ECHO_T = @ECHO_T@ EGREP = @EGREP@ EXEEXT = @EXEEXT@ F77 = @F77@ FFLAGS = @FFLAGS@ FGREP = @FGREP@ FLIBS = @FLIBS@ GREP = @GREP@ INSTALL = @INSTALL@ INSTALL_DATA = @INSTALL_DATA@ INSTALL_PROGRAM = @INSTALL_PROGRAM@ INSTALL_SCRIPT = @INSTALL_SCRIPT@ INSTALL_STRIP_PROGRAM = @INSTALL_STRIP_PROGRAM@ LD = @LD@ LDFLAGS = @LDFLAGS@ LIBOBJS = @LIBOBJS@ LIBQUADMATH = @LIBQUADMATH@ LIBS = @LIBS@ LIBTOOL = @LIBTOOL@ LIPO = @LIPO@ LN_S = @LN_S@ LTLIBOBJS = @LTLIBOBJS@ MAINT = @MAINT@ MAKEINFO = @MAKEINFO@ MANIFEST_TOOL = @MANIFEST_TOOL@ MKDIR_P = @MKDIR_P@ MPICC = @MPICC@ MPILIBS = @MPILIBS@ MPIRUN = @MPIRUN@ NEON_CFLAGS = @NEON_CFLAGS@ NM = @NM@ NMEDIT = @NMEDIT@ OBJDUMP = @OBJDUMP@ OBJEXT = @OBJEXT@ OCAMLBUILD = @OCAMLBUILD@ OPENMP_CFLAGS = @OPENMP_CFLAGS@ OTOOL = @OTOOL@ OTOOL64 = @OTOOL64@ PACKAGE = @PACKAGE@ PACKAGE_BUGREPORT = @PACKAGE_BUGREPORT@ PACKAGE_NAME = @PACKAGE_NAME@ PACKAGE_STRING = @PACKAGE_STRING@ PACKAGE_TARNAME = @PACKAGE_TARNAME@ PACKAGE_URL = @PACKAGE_URL@ PACKAGE_VERSION = @PACKAGE_VERSION@ PATH_SEPARATOR = @PATH_SEPARATOR@ POW_LIB = @POW_LIB@ PRECISION = @PRECISION@ PREC_SUFFIX = @PREC_SUFFIX@ PTHREAD_CC = @PTHREAD_CC@ PTHREAD_CFLAGS = @PTHREAD_CFLAGS@ PTHREAD_LIBS = @PTHREAD_LIBS@ RANLIB = @RANLIB@ SED = @SED@ SET_MAKE = @SET_MAKE@ SHARED_VERSION_INFO = @SHARED_VERSION_INFO@ SHELL = @SHELL@ SSE2_CFLAGS = @SSE2_CFLAGS@ STACK_ALIGN_CFLAGS = @STACK_ALIGN_CFLAGS@ STRIP = @STRIP@ THREADLIBS = @THREADLIBS@ VERSION = @VERSION@ abs_builddir = @abs_builddir@ abs_srcdir = @abs_srcdir@ abs_top_builddir = @abs_top_builddir@ abs_top_srcdir = @abs_top_srcdir@ ac_ct_AR = @ac_ct_AR@ ac_ct_CC = @ac_ct_CC@ ac_ct_DUMPBIN = @ac_ct_DUMPBIN@ ac_ct_F77 = @ac_ct_F77@ acx_pthread_config = @acx_pthread_config@ am__include = @am__include@ am__leading_dot = @am__leading_dot@ am__quote = @am__quote@ am__tar = @am__tar@ am__untar = @am__untar@ bindir = @bindir@ build = @build@ build_alias = @build_alias@ build_cpu = @build_cpu@ build_os = @build_os@ build_vendor = @build_vendor@ builddir = @builddir@ datadir = @datadir@ datarootdir = @datarootdir@ docdir = @docdir@ dvidir = @dvidir@ exec_prefix = @exec_prefix@ host = @host@ host_alias = @host_alias@ host_cpu = @host_cpu@ host_os = @host_os@ host_vendor = @host_vendor@ htmldir = @htmldir@ includedir = @includedir@ infodir = @infodir@ install_sh = @install_sh@ libdir = @libdir@ libexecdir = @libexecdir@ localedir = @localedir@ localstatedir = @localstatedir@ mandir = @mandir@ mkdir_p = @mkdir_p@ oldincludedir = @oldincludedir@ pdfdir = @pdfdir@ prefix = @prefix@ program_transform_name = @program_transform_name@ psdir = @psdir@ sbindir = @sbindir@ sharedstatedir = @sharedstatedir@ srcdir = @srcdir@ sysconfdir = @sysconfdir@ target_alias = @target_alias@ top_build_prefix = @top_build_prefix@ top_builddir = @top_builddir@ top_srcdir = @top_srcdir@ AM_CPPFLAGS = -I$(top_srcdir)/kernel -I$(top_srcdir)/rdft SUBDIRS = r2cf r2cb r2r noinst_LTLIBRARIES = librdft_scalar.la librdft_scalar_la_SOURCES = hb.h r2cb.h r2cbIII.h hf.h hfb.c r2c.c \ r2cf.h r2cfII.h r2r.c r2r.h hc2c.c hc2cf.h hc2cb.h all: all-recursive .SUFFIXES: .SUFFIXES: .c .lo .o .obj $(srcdir)/Makefile.in: @MAINTAINER_MODE_TRUE@ $(srcdir)/Makefile.am $(am__configure_deps) @for dep in $?; do \ case '$(am__configure_deps)' in \ *$$dep*) \ ( cd $(top_builddir) && $(MAKE) $(AM_MAKEFLAGS) am--refresh ) \ && { if test -f $@; then exit 0; else break; fi; }; \ exit 1;; \ esac; \ done; \ echo ' cd $(top_srcdir) && $(AUTOMAKE) --gnu rdft/scalar/Makefile'; \ $(am__cd) $(top_srcdir) && \ $(AUTOMAKE) --gnu rdft/scalar/Makefile .PRECIOUS: Makefile Makefile: $(srcdir)/Makefile.in $(top_builddir)/config.status @case '$?' in \ *config.status*) \ cd $(top_builddir) && $(MAKE) $(AM_MAKEFLAGS) am--refresh;; \ *) \ echo ' cd $(top_builddir) && $(SHELL) ./config.status $(subdir)/$@ $(am__depfiles_maybe)'; \ cd $(top_builddir) && $(SHELL) ./config.status $(subdir)/$@ $(am__depfiles_maybe);; \ esac; $(top_builddir)/config.status: $(top_srcdir)/configure $(CONFIG_STATUS_DEPENDENCIES) cd $(top_builddir) && $(MAKE) $(AM_MAKEFLAGS) am--refresh $(top_srcdir)/configure: @MAINTAINER_MODE_TRUE@ $(am__configure_deps) cd $(top_builddir) && $(MAKE) $(AM_MAKEFLAGS) am--refresh $(ACLOCAL_M4): @MAINTAINER_MODE_TRUE@ $(am__aclocal_m4_deps) cd $(top_builddir) && $(MAKE) $(AM_MAKEFLAGS) am--refresh $(am__aclocal_m4_deps): clean-noinstLTLIBRARIES: -test -z "$(noinst_LTLIBRARIES)" || rm -f $(noinst_LTLIBRARIES) @list='$(noinst_LTLIBRARIES)'; for p in $$list; do \ dir="`echo $$p | sed -e 's|/[^/]*$$||'`"; \ test "$$dir" != "$$p" || dir=.; \ echo "rm -f \"$${dir}/so_locations\""; \ rm -f "$${dir}/so_locations"; \ done librdft_scalar.la: $(librdft_scalar_la_OBJECTS) $(librdft_scalar_la_DEPENDENCIES) $(EXTRA_librdft_scalar_la_DEPENDENCIES) $(LINK) $(librdft_scalar_la_OBJECTS) $(librdft_scalar_la_LIBADD) $(LIBS) mostlyclean-compile: -rm -f *.$(OBJEXT) distclean-compile: -rm -f *.tab.c @AMDEP_TRUE@@am__include@ @am__quote@./$(DEPDIR)/hc2c.Plo@am__quote@ @AMDEP_TRUE@@am__include@ @am__quote@./$(DEPDIR)/hfb.Plo@am__quote@ @AMDEP_TRUE@@am__include@ @am__quote@./$(DEPDIR)/r2c.Plo@am__quote@ @AMDEP_TRUE@@am__include@ @am__quote@./$(DEPDIR)/r2r.Plo@am__quote@ .c.o: @am__fastdepCC_TRUE@ $(COMPILE) -MT $@ -MD -MP -MF $(DEPDIR)/$*.Tpo -c -o $@ $< @am__fastdepCC_TRUE@ $(am__mv) $(DEPDIR)/$*.Tpo $(DEPDIR)/$*.Po @AMDEP_TRUE@@am__fastdepCC_FALSE@ source='$<' object='$@' libtool=no @AMDEPBACKSLASH@ @AMDEP_TRUE@@am__fastdepCC_FALSE@ DEPDIR=$(DEPDIR) $(CCDEPMODE) $(depcomp) @AMDEPBACKSLASH@ @am__fastdepCC_FALSE@ $(COMPILE) -c $< .c.obj: @am__fastdepCC_TRUE@ $(COMPILE) -MT $@ -MD -MP -MF $(DEPDIR)/$*.Tpo -c -o $@ `$(CYGPATH_W) '$<'` @am__fastdepCC_TRUE@ $(am__mv) $(DEPDIR)/$*.Tpo $(DEPDIR)/$*.Po @AMDEP_TRUE@@am__fastdepCC_FALSE@ source='$<' object='$@' libtool=no @AMDEPBACKSLASH@ @AMDEP_TRUE@@am__fastdepCC_FALSE@ DEPDIR=$(DEPDIR) $(CCDEPMODE) $(depcomp) @AMDEPBACKSLASH@ @am__fastdepCC_FALSE@ $(COMPILE) -c `$(CYGPATH_W) '$<'` .c.lo: @am__fastdepCC_TRUE@ $(LTCOMPILE) -MT $@ -MD -MP -MF $(DEPDIR)/$*.Tpo -c -o $@ $< @am__fastdepCC_TRUE@ $(am__mv) $(DEPDIR)/$*.Tpo $(DEPDIR)/$*.Plo @AMDEP_TRUE@@am__fastdepCC_FALSE@ source='$<' object='$@' libtool=yes @AMDEPBACKSLASH@ @AMDEP_TRUE@@am__fastdepCC_FALSE@ DEPDIR=$(DEPDIR) $(CCDEPMODE) $(depcomp) @AMDEPBACKSLASH@ @am__fastdepCC_FALSE@ $(LTCOMPILE) -c -o $@ $< mostlyclean-libtool: -rm -f *.lo clean-libtool: -rm -rf .libs _libs # This directory's subdirectories are mostly independent; you can cd # into them and run `make' without going through this Makefile. # To change the values of `make' variables: instead of editing Makefiles, # (1) if the variable is set in `config.status', edit `config.status' # (which will cause the Makefiles to be regenerated when you run `make'); # (2) otherwise, pass the desired values on the `make' command line. $(RECURSIVE_TARGETS): @fail= failcom='exit 1'; \ for f in x $$MAKEFLAGS; do \ case $$f in \ *=* | --[!k]*);; \ *k*) failcom='fail=yes';; \ esac; \ done; \ dot_seen=no; \ target=`echo $@ | sed s/-recursive//`; \ list='$(SUBDIRS)'; for subdir in $$list; do \ echo "Making $$target in $$subdir"; \ if test "$$subdir" = "."; then \ dot_seen=yes; \ local_target="$$target-am"; \ else \ local_target="$$target"; \ fi; \ ($(am__cd) $$subdir && $(MAKE) $(AM_MAKEFLAGS) $$local_target) \ || eval $$failcom; \ done; \ if test "$$dot_seen" = "no"; then \ $(MAKE) $(AM_MAKEFLAGS) "$$target-am" || exit 1; \ fi; test -z "$$fail" $(RECURSIVE_CLEAN_TARGETS): @fail= failcom='exit 1'; \ for f in x $$MAKEFLAGS; do \ case $$f in \ *=* | --[!k]*);; \ *k*) failcom='fail=yes';; \ esac; \ done; \ dot_seen=no; \ case "$@" in \ distclean-* | maintainer-clean-*) list='$(DIST_SUBDIRS)' ;; \ *) list='$(SUBDIRS)' ;; \ esac; \ rev=''; for subdir in $$list; do \ if test "$$subdir" = "."; then :; else \ rev="$$subdir $$rev"; \ fi; \ done; \ rev="$$rev ."; \ target=`echo $@ | sed s/-recursive//`; \ for subdir in $$rev; do \ echo "Making $$target in $$subdir"; \ if test "$$subdir" = "."; then \ local_target="$$target-am"; \ else \ local_target="$$target"; \ fi; \ ($(am__cd) $$subdir && $(MAKE) $(AM_MAKEFLAGS) $$local_target) \ || eval $$failcom; \ done && test -z "$$fail" tags-recursive: list='$(SUBDIRS)'; for subdir in $$list; do \ test "$$subdir" = . || ($(am__cd) $$subdir && $(MAKE) $(AM_MAKEFLAGS) tags); \ done ctags-recursive: list='$(SUBDIRS)'; for subdir in $$list; do \ test "$$subdir" = . || ($(am__cd) $$subdir && $(MAKE) $(AM_MAKEFLAGS) ctags); \ done ID: $(HEADERS) $(SOURCES) $(LISP) $(TAGS_FILES) list='$(SOURCES) $(HEADERS) $(LISP) $(TAGS_FILES)'; \ unique=`for i in $$list; do \ if test -f "$$i"; then echo $$i; else echo $(srcdir)/$$i; fi; \ done | \ $(AWK) '{ files[$$0] = 1; nonempty = 1; } \ END { if (nonempty) { for (i in files) print i; }; }'`; \ mkid -fID $$unique tags: TAGS TAGS: tags-recursive $(HEADERS) $(SOURCES) $(TAGS_DEPENDENCIES) \ $(TAGS_FILES) $(LISP) set x; \ here=`pwd`; \ if ($(ETAGS) --etags-include --version) >/dev/null 2>&1; then \ include_option=--etags-include; \ empty_fix=.; \ else \ include_option=--include; \ empty_fix=; \ fi; \ list='$(SUBDIRS)'; for subdir in $$list; do \ if test "$$subdir" = .; then :; else \ test ! -f $$subdir/TAGS || \ set "$$@" "$$include_option=$$here/$$subdir/TAGS"; \ fi; \ done; \ list='$(SOURCES) $(HEADERS) $(LISP) $(TAGS_FILES)'; \ unique=`for i in $$list; do \ if test -f "$$i"; then echo $$i; else echo $(srcdir)/$$i; fi; \ done | \ $(AWK) '{ files[$$0] = 1; nonempty = 1; } \ END { if (nonempty) { for (i in files) print i; }; }'`; \ shift; \ if test -z "$(ETAGS_ARGS)$$*$$unique"; then :; else \ test -n "$$unique" || unique=$$empty_fix; \ if test $$# -gt 0; then \ $(ETAGS) $(ETAGSFLAGS) $(AM_ETAGSFLAGS) $(ETAGS_ARGS) \ "$$@" $$unique; \ else \ $(ETAGS) $(ETAGSFLAGS) $(AM_ETAGSFLAGS) $(ETAGS_ARGS) \ $$unique; \ fi; \ fi ctags: CTAGS CTAGS: ctags-recursive $(HEADERS) $(SOURCES) $(TAGS_DEPENDENCIES) \ $(TAGS_FILES) $(LISP) list='$(SOURCES) $(HEADERS) $(LISP) $(TAGS_FILES)'; \ unique=`for i in $$list; do \ if test -f "$$i"; then echo $$i; else echo $(srcdir)/$$i; fi; \ done | \ $(AWK) '{ files[$$0] = 1; nonempty = 1; } \ END { if (nonempty) { for (i in files) print i; }; }'`; \ test -z "$(CTAGS_ARGS)$$unique" \ || $(CTAGS) $(CTAGSFLAGS) $(AM_CTAGSFLAGS) $(CTAGS_ARGS) \ $$unique GTAGS: here=`$(am__cd) $(top_builddir) && pwd` \ && $(am__cd) $(top_srcdir) \ && gtags -i $(GTAGS_ARGS) "$$here" distclean-tags: -rm -f TAGS ID GTAGS GRTAGS GSYMS GPATH tags distdir: $(DISTFILES) @srcdirstrip=`echo "$(srcdir)" | sed 's/[].[^$$\\*]/\\\\&/g'`; \ topsrcdirstrip=`echo "$(top_srcdir)" | sed 's/[].[^$$\\*]/\\\\&/g'`; \ list='$(DISTFILES)'; \ dist_files=`for file in $$list; do echo $$file; done | \ sed -e "s|^$$srcdirstrip/||;t" \ -e "s|^$$topsrcdirstrip/|$(top_builddir)/|;t"`; \ case $$dist_files in \ */*) $(MKDIR_P) `echo "$$dist_files" | \ sed '/\//!d;s|^|$(distdir)/|;s,/[^/]*$$,,' | \ sort -u` ;; \ esac; \ for file in $$dist_files; do \ if test -f $$file || test -d $$file; then d=.; else d=$(srcdir); fi; \ if test -d $$d/$$file; then \ dir=`echo "/$$file" | sed -e 's,/[^/]*$$,,'`; \ if test -d "$(distdir)/$$file"; then \ find "$(distdir)/$$file" -type d ! -perm -700 -exec chmod u+rwx {} \;; \ fi; \ if test -d $(srcdir)/$$file && test $$d != $(srcdir); then \ cp -fpR $(srcdir)/$$file "$(distdir)$$dir" || exit 1; \ find "$(distdir)/$$file" -type d ! -perm -700 -exec chmod u+rwx {} \;; \ fi; \ cp -fpR $$d/$$file "$(distdir)$$dir" || exit 1; \ else \ test -f "$(distdir)/$$file" \ || cp -p $$d/$$file "$(distdir)/$$file" \ || exit 1; \ fi; \ done @list='$(DIST_SUBDIRS)'; for subdir in $$list; do \ if test "$$subdir" = .; then :; else \ $(am__make_dryrun) \ || test -d "$(distdir)/$$subdir" \ || $(MKDIR_P) "$(distdir)/$$subdir" \ || exit 1; \ dir1=$$subdir; dir2="$(distdir)/$$subdir"; \ $(am__relativize); \ new_distdir=$$reldir; \ dir1=$$subdir; dir2="$(top_distdir)"; \ $(am__relativize); \ new_top_distdir=$$reldir; \ echo " (cd $$subdir && $(MAKE) $(AM_MAKEFLAGS) top_distdir="$$new_top_distdir" distdir="$$new_distdir" \\"; \ echo " am__remove_distdir=: am__skip_length_check=: am__skip_mode_fix=: distdir)"; \ ($(am__cd) $$subdir && \ $(MAKE) $(AM_MAKEFLAGS) \ top_distdir="$$new_top_distdir" \ distdir="$$new_distdir" \ am__remove_distdir=: \ am__skip_length_check=: \ am__skip_mode_fix=: \ distdir) \ || exit 1; \ fi; \ done check-am: all-am check: check-recursive all-am: Makefile $(LTLIBRARIES) installdirs: installdirs-recursive installdirs-am: install: install-recursive install-exec: install-exec-recursive install-data: install-data-recursive uninstall: uninstall-recursive install-am: all-am @$(MAKE) $(AM_MAKEFLAGS) install-exec-am install-data-am installcheck: installcheck-recursive install-strip: if test -z '$(STRIP)'; then \ $(MAKE) $(AM_MAKEFLAGS) INSTALL_PROGRAM="$(INSTALL_STRIP_PROGRAM)" \ install_sh_PROGRAM="$(INSTALL_STRIP_PROGRAM)" INSTALL_STRIP_FLAG=-s \ install; \ else \ $(MAKE) $(AM_MAKEFLAGS) INSTALL_PROGRAM="$(INSTALL_STRIP_PROGRAM)" \ install_sh_PROGRAM="$(INSTALL_STRIP_PROGRAM)" INSTALL_STRIP_FLAG=-s \ "INSTALL_PROGRAM_ENV=STRIPPROG='$(STRIP)'" install; \ fi mostlyclean-generic: clean-generic: distclean-generic: -test -z "$(CONFIG_CLEAN_FILES)" || rm -f $(CONFIG_CLEAN_FILES) -test . = "$(srcdir)" || test -z "$(CONFIG_CLEAN_VPATH_FILES)" || rm -f $(CONFIG_CLEAN_VPATH_FILES) maintainer-clean-generic: @echo "This command is intended for maintainers to use" @echo "it deletes files that may require special tools to rebuild." clean: clean-recursive clean-am: clean-generic clean-libtool clean-noinstLTLIBRARIES \ mostlyclean-am distclean: distclean-recursive -rm -rf ./$(DEPDIR) -rm -f Makefile distclean-am: clean-am distclean-compile distclean-generic \ distclean-tags dvi: dvi-recursive dvi-am: html: html-recursive html-am: info: info-recursive info-am: install-data-am: install-dvi: install-dvi-recursive install-dvi-am: install-exec-am: install-html: install-html-recursive install-html-am: install-info: install-info-recursive install-info-am: install-man: install-pdf: install-pdf-recursive install-pdf-am: install-ps: install-ps-recursive install-ps-am: installcheck-am: maintainer-clean: maintainer-clean-recursive -rm -rf ./$(DEPDIR) -rm -f Makefile maintainer-clean-am: distclean-am maintainer-clean-generic mostlyclean: mostlyclean-recursive mostlyclean-am: mostlyclean-compile mostlyclean-generic \ mostlyclean-libtool pdf: pdf-recursive pdf-am: ps: ps-recursive ps-am: uninstall-am: .MAKE: $(RECURSIVE_CLEAN_TARGETS) $(RECURSIVE_TARGETS) ctags-recursive \ install-am install-strip tags-recursive .PHONY: $(RECURSIVE_CLEAN_TARGETS) $(RECURSIVE_TARGETS) CTAGS GTAGS \ all all-am check check-am clean clean-generic clean-libtool \ clean-noinstLTLIBRARIES ctags ctags-recursive distclean \ distclean-compile distclean-generic distclean-libtool \ distclean-tags distdir dvi dvi-am html html-am info info-am \ install install-am install-data install-data-am install-dvi \ install-dvi-am install-exec install-exec-am install-html \ install-html-am install-info install-info-am install-man \ install-pdf install-pdf-am install-ps install-ps-am \ install-strip installcheck installcheck-am installdirs \ installdirs-am maintainer-clean maintainer-clean-generic \ mostlyclean mostlyclean-compile mostlyclean-generic \ mostlyclean-libtool pdf pdf-am ps ps-am tags tags-recursive \ uninstall uninstall-am # Tell versions [3.59,3.63) of GNU make to not export all variables. # Otherwise a system limit (for SysV at least) may be exceeded. .NOEXPORT: fftw-3.3.3/rdft/scalar/hb.h0000644000175000017500000000161512054410050012347 00000000000000/* * Copyright (c) 2003, 2007-11 Matteo Frigo * Copyright (c) 2003, 2007-11 Massachusetts Institute of Technology * * This program is free software; you can redistribute it and/or modify * it under the terms of the GNU General Public License as published by * the Free Software Foundation; either version 2 of the License, or * (at your option) any later version. * * This program is distributed in the hope that it will be useful, * but WITHOUT ANY WARRANTY; without even the implied warranty of * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the * GNU General Public License for more details. * * You should have received a copy of the GNU General Public License * along with this program; if not, write to the Free Software * Foundation, Inc., 51 Franklin Street, Fifth Floor, Boston, MA 02110-1301 USA * */ #define GENUS X(rdft_hb_genus) extern const hc2hc_genus GENUS; fftw-3.3.3/rdft/scalar/r2cb.h0000644000175000017500000000161612054410050012607 00000000000000/* * Copyright (c) 2003, 2007-11 Matteo Frigo * Copyright (c) 2003, 2007-11 Massachusetts Institute of Technology * * This program is free software; you can redistribute it and/or modify * it under the terms of the GNU General Public License as published by * the Free Software Foundation; either version 2 of the License, or * (at your option) any later version. * * This program is distributed in the hope that it will be useful, * but WITHOUT ANY WARRANTY; without even the implied warranty of * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the * GNU General Public License for more details. * * You should have received a copy of the GNU General Public License * along with this program; if not, write to the Free Software * Foundation, Inc., 51 Franklin Street, Fifth Floor, Boston, MA 02110-1301 USA * */ #define GENUS X(rdft_r2cb_genus) extern const kr2c_genus GENUS; fftw-3.3.3/rdft/scalar/r2cbIII.h0000644000175000017500000000162112054410050013136 00000000000000/* * Copyright (c) 2003, 2007-11 Matteo Frigo * Copyright (c) 2003, 2007-11 Massachusetts Institute of Technology * * This program is free software; you can redistribute it and/or modify * it under the terms of the GNU General Public License as published by * the Free Software Foundation; either version 2 of the License, or * (at your option) any later version. * * This program is distributed in the hope that it will be useful, * but WITHOUT ANY WARRANTY; without even the implied warranty of * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the * GNU General Public License for more details. * * You should have received a copy of the GNU General Public License * along with this program; if not, write to the Free Software * Foundation, Inc., 51 Franklin Street, Fifth Floor, Boston, MA 02110-1301 USA * */ #define GENUS X(rdft_r2cbIII_genus) extern const kr2c_genus GENUS; fftw-3.3.3/rdft/scalar/hf.h0000644000175000017500000000161512054410050012353 00000000000000/* * Copyright (c) 2003, 2007-11 Matteo Frigo * Copyright (c) 2003, 2007-11 Massachusetts Institute of Technology * * This program is free software; you can redistribute it and/or modify * it under the terms of the GNU General Public License as published by * the Free Software Foundation; either version 2 of the License, or * (at your option) any later version. * * This program is distributed in the hope that it will be useful, * but WITHOUT ANY WARRANTY; without even the implied warranty of * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the * GNU General Public License for more details. * * You should have received a copy of the GNU General Public License * along with this program; if not, write to the Free Software * Foundation, Inc., 51 Franklin Street, Fifth Floor, Boston, MA 02110-1301 USA * */ #define GENUS X(rdft_hf_genus) extern const hc2hc_genus GENUS; fftw-3.3.3/rdft/scalar/hfb.c0000644000175000017500000000174512054410050012514 00000000000000/* * Copyright (c) 2003, 2007-11 Matteo Frigo * Copyright (c) 2003, 2007-11 Massachusetts Institute of Technology * * This program is free software; you can redistribute it and/or modify * it under the terms of the GNU General Public License as published by * the Free Software Foundation; either version 2 of the License, or * (at your option) any later version. * * This program is distributed in the hope that it will be useful, * but WITHOUT ANY WARRANTY; without even the implied warranty of * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the * GNU General Public License for more details. * * You should have received a copy of the GNU General Public License * along with this program; if not, write to the Free Software * Foundation, Inc., 51 Franklin Street, Fifth Floor, Boston, MA 02110-1301 USA * */ #include "codelet-rdft.h" #include "hf.h" const hc2hc_genus GENUS = { R2HC, 1 }; #undef GENUS #include "hb.h" const hc2hc_genus GENUS = { HC2R, 1 }; fftw-3.3.3/rdft/scalar/r2c.c0000644000175000017500000000221112054410050012430 00000000000000/* * Copyright (c) 2003, 2007-11 Matteo Frigo * Copyright (c) 2003, 2007-11 Massachusetts Institute of Technology * * This program is free software; you can redistribute it and/or modify * it under the terms of the GNU General Public License as published by * the Free Software Foundation; either version 2 of the License, or * (at your option) any later version. * * This program is distributed in the hope that it will be useful, * but WITHOUT ANY WARRANTY; without even the implied warranty of * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the * GNU General Public License for more details. * * You should have received a copy of the GNU General Public License * along with this program; if not, write to the Free Software * Foundation, Inc., 51 Franklin Street, Fifth Floor, Boston, MA 02110-1301 USA * */ #include "codelet-rdft.h" #include "r2cf.h" const kr2c_genus GENUS = { R2HC, 1 }; #undef GENUS #include "r2cfII.h" const kr2c_genus GENUS = { R2HCII, 1 }; #undef GENUS #include "r2cb.h" const kr2c_genus GENUS = { HC2R, 1 }; #undef GENUS #include "r2cbIII.h" const kr2c_genus GENUS = { HC2RIII, 1 }; #undef GENUS fftw-3.3.3/rdft/scalar/r2cf.h0000644000175000017500000000161612054410050012613 00000000000000/* * Copyright (c) 2003, 2007-11 Matteo Frigo * Copyright (c) 2003, 2007-11 Massachusetts Institute of Technology * * This program is free software; you can redistribute it and/or modify * it under the terms of the GNU General Public License as published by * the Free Software Foundation; either version 2 of the License, or * (at your option) any later version. * * This program is distributed in the hope that it will be useful, * but WITHOUT ANY WARRANTY; without even the implied warranty of * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the * GNU General Public License for more details. * * You should have received a copy of the GNU General Public License * along with this program; if not, write to the Free Software * Foundation, Inc., 51 Franklin Street, Fifth Floor, Boston, MA 02110-1301 USA * */ #define GENUS X(rdft_r2cf_genus) extern const kr2c_genus GENUS; fftw-3.3.3/rdft/scalar/r2cfII.h0000644000175000017500000000162012054410050013030 00000000000000/* * Copyright (c) 2003, 2007-11 Matteo Frigo * Copyright (c) 2003, 2007-11 Massachusetts Institute of Technology * * This program is free software; you can redistribute it and/or modify * it under the terms of the GNU General Public License as published by * the Free Software Foundation; either version 2 of the License, or * (at your option) any later version. * * This program is distributed in the hope that it will be useful, * but WITHOUT ANY WARRANTY; without even the implied warranty of * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the * GNU General Public License for more details. * * You should have received a copy of the GNU General Public License * along with this program; if not, write to the Free Software * Foundation, Inc., 51 Franklin Street, Fifth Floor, Boston, MA 02110-1301 USA * */ #define GENUS X(rdft_r2cfII_genus) extern const kr2c_genus GENUS; fftw-3.3.3/rdft/scalar/r2r.c0000644000175000017500000000163112054410050012454 00000000000000/* * Copyright (c) 2003, 2007-11 Matteo Frigo * Copyright (c) 2003, 2007-11 Massachusetts Institute of Technology * * This program is free software; you can redistribute it and/or modify * it under the terms of the GNU General Public License as published by * the Free Software Foundation; either version 2 of the License, or * (at your option) any later version. * * This program is distributed in the hope that it will be useful, * but WITHOUT ANY WARRANTY; without even the implied warranty of * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the * GNU General Public License for more details. * * You should have received a copy of the GNU General Public License * along with this program; if not, write to the Free Software * Foundation, Inc., 51 Franklin Street, Fifth Floor, Boston, MA 02110-1301 USA * */ #include "codelet-rdft.h" #include "r2r.h" const kr2r_genus GENUS = { 1 }; fftw-3.3.3/rdft/scalar/r2r.h0000644000175000017500000000161512054410050012463 00000000000000/* * Copyright (c) 2003, 2007-11 Matteo Frigo * Copyright (c) 2003, 2007-11 Massachusetts Institute of Technology * * This program is free software; you can redistribute it and/or modify * it under the terms of the GNU General Public License as published by * the Free Software Foundation; either version 2 of the License, or * (at your option) any later version. * * This program is distributed in the hope that it will be useful, * but WITHOUT ANY WARRANTY; without even the implied warranty of * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the * GNU General Public License for more details. * * You should have received a copy of the GNU General Public License * along with this program; if not, write to the Free Software * Foundation, Inc., 51 Franklin Street, Fifth Floor, Boston, MA 02110-1301 USA * */ #define GENUS X(rdft_r2r_genus) extern const kr2r_genus GENUS; fftw-3.3.3/rdft/scalar/hc2c.c0000644000175000017500000000241212054410050012564 00000000000000/* * Copyright (c) 2003, 2007-11 Matteo Frigo * Copyright (c) 2003, 2007-11 Massachusetts Institute of Technology * * This program is free software; you can redistribute it and/or modify * it under the terms of the GNU General Public License as published by * the Free Software Foundation; either version 2 of the License, or * (at your option) any later version. * * This program is distributed in the hope that it will be useful, * but WITHOUT ANY WARRANTY; without even the implied warranty of * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the * GNU General Public License for more details. * * You should have received a copy of the GNU General Public License * along with this program; if not, write to the Free Software * Foundation, Inc., 51 Franklin Street, Fifth Floor, Boston, MA 02110-1301 USA * */ #include "codelet-rdft.h" #include "hc2cf.h" static int okp(const R *Rp, const R *Ip, const R *Rm, const R *Im, INT rs, INT mb, INT me, INT ms, const planner *plnr) { UNUSED(Rp); UNUSED(Ip); UNUSED(Rm); UNUSED(Im); UNUSED(rs); UNUSED(mb); UNUSED(me); UNUSED(ms); UNUSED(plnr); return 1; } const hc2c_genus GENUS = { okp, R2HC, 1 }; #undef GENUS #include "hc2cb.h" const hc2c_genus GENUS = { okp, HC2R, 1 }; fftw-3.3.3/rdft/scalar/hc2cf.h0000644000175000017500000000161712054410050012745 00000000000000/* * Copyright (c) 2003, 2007-11 Matteo Frigo * Copyright (c) 2003, 2007-11 Massachusetts Institute of Technology * * This program is free software; you can redistribute it and/or modify * it under the terms of the GNU General Public License as published by * the Free Software Foundation; either version 2 of the License, or * (at your option) any later version. * * This program is distributed in the hope that it will be useful, * but WITHOUT ANY WARRANTY; without even the implied warranty of * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the * GNU General Public License for more details. * * You should have received a copy of the GNU General Public License * along with this program; if not, write to the Free Software * Foundation, Inc., 51 Franklin Street, Fifth Floor, Boston, MA 02110-1301 USA * */ #define GENUS X(rdft_hc2cf_genus) extern const hc2c_genus GENUS; fftw-3.3.3/rdft/scalar/hc2cb.h0000644000175000017500000000161712054410050012741 00000000000000/* * Copyright (c) 2003, 2007-11 Matteo Frigo * Copyright (c) 2003, 2007-11 Massachusetts Institute of Technology * * This program is free software; you can redistribute it and/or modify * it under the terms of the GNU General Public License as published by * the Free Software Foundation; either version 2 of the License, or * (at your option) any later version. * * This program is distributed in the hope that it will be useful, * but WITHOUT ANY WARRANTY; without even the implied warranty of * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the * GNU General Public License for more details. * * You should have received a copy of the GNU General Public License * along with this program; if not, write to the Free Software * Foundation, Inc., 51 Franklin Street, Fifth Floor, Boston, MA 02110-1301 USA * */ #define GENUS X(rdft_hc2cb_genus) extern const hc2c_genus GENUS; fftw-3.3.3/rdft/scalar/r2cf/0000755000175000017500000000000012054411110012514 500000000000000fftw-3.3.3/rdft/scalar/r2cf/Makefile.am0000644000175000017500000001205112054410050014471 00000000000000# This Makefile.am specifies a set of codelets, efficient transforms # of small sizes, that are used as building blocks (kernels) by FFTW # to build up large transforms, as well as the options for generating # and compiling them. # You can customize FFTW for special needs, e.g. to handle certain # sizes more efficiently, by adding new codelets to the lists of those # included by default. If you change the list of codelets, any new # ones you added will be automatically generated when you run the # bootstrap script (see "Generating your own code" in the FFTW # manual). ########################################################################### AM_CPPFLAGS = -I$(top_srcdir)/kernel -I$(top_srcdir)/rdft \ -I$(top_srcdir)/rdft/scalar noinst_LTLIBRARIES = librdft_scalar_r2cf.la ########################################################################### # r2cf_ is a hard-coded real-to-complex FFT of size (base cases # of real-input FFT recursion) R2CF = r2cf_2.c r2cf_3.c r2cf_4.c r2cf_5.c r2cf_6.c r2cf_7.c r2cf_8.c \ r2cf_9.c r2cf_10.c r2cf_11.c r2cf_12.c r2cf_13.c r2cf_14.c r2cf_15.c \ r2cf_16.c r2cf_32.c r2cf_64.c r2cf_128.c \ r2cf_20.c r2cf_25.c # r2cf_30.c r2cf_40.c r2cf_50.c ########################################################################### # hf_ is a "twiddle" FFT of size , implementing a radix-r DIT # step for a real-input FFT. Every hf codelet must have a # corresponding r2cfII codelet (see below)! HF = hf_2.c hf_3.c hf_4.c hf_5.c hf_6.c hf_7.c hf_8.c hf_9.c \ hf_10.c hf_12.c hf_15.c hf_16.c hf_32.c hf_64.c \ hf_20.c hf_25.c # hf_30.c hf_40.c hf_50.c # like hf, but generates part of its trig table on the fly (good for large n) HF2 = hf2_4.c hf2_8.c hf2_16.c hf2_32.c \ hf2_5.c hf2_20.c hf2_25.c # an r2cf transform where the input is shifted by half a sample (output # is multiplied by a phase). This is needed as part of the DIT recursion; # every hf_ or hf2_ codelet should have a corresponding r2cfII_ R2CFII = r2cfII_2.c r2cfII_3.c r2cfII_4.c r2cfII_5.c r2cfII_6.c \ r2cfII_7.c r2cfII_8.c r2cfII_9.c r2cfII_10.c r2cfII_12.c r2cfII_15.c \ r2cfII_16.c r2cfII_32.c r2cfII_64.c \ r2cfII_20.c r2cfII_25.c # r2cfII_30.c r2cfII_40.c r2cfII_50.c ########################################################################### # hc2cf_ is a "twiddle" FFT of size , implementing a radix-r DIT # step for a real-input FFT with rdft2-style output. must be even. HC2CF = hc2cf_2.c hc2cf_4.c hc2cf_6.c hc2cf_8.c hc2cf_10.c hc2cf_12.c \ hc2cf_16.c hc2cf_32.c \ hc2cf_20.c # hc2cf_30.c HC2CFDFT = hc2cfdft_2.c hc2cfdft_4.c hc2cfdft_6.c hc2cfdft_8.c \ hc2cfdft_10.c hc2cfdft_12.c hc2cfdft_16.c hc2cfdft_32.c \ hc2cfdft_20.c # hc2cfdft_30.c # like hc2cf, but generates part of its trig table on the fly (good # for large n) HC2CF2 = hc2cf2_4.c hc2cf2_8.c hc2cf2_16.c hc2cf2_32.c \ hc2cf2_20.c # hc2cf2_30.c HC2CFDFT2 = hc2cfdft2_4.c hc2cfdft2_8.c hc2cfdft2_16.c hc2cfdft2_32.c \ hc2cfdft2_20.c # hc2cfdft2_30.c ########################################################################### ALL_CODELETS = $(R2CF) $(HF) $(HF2) $(R2CFII) $(HC2CF) $(HC2CF2) \ $(HC2CFDFT) $(HC2CFDFT2) BUILT_SOURCES= $(ALL_CODELETS) $(CODLIST) librdft_scalar_r2cf_la_SOURCES = $(BUILT_SOURCES) SOLVTAB_NAME = X(solvtab_rdft_r2cf) XRENAME=X # special rules for regenerating codelets. include $(top_srcdir)/support/Makefile.codelets if MAINTAINER_MODE FLAGS_R2CF=$(RDFT_FLAGS_COMMON) FLAGS_HF=$(RDFT_FLAGS_COMMON) FLAGS_HF2=$(RDFT_FLAGS_COMMON) -twiddle-log3 -precompute-twiddles FLAGS_HC2CF=$(RDFT_FLAGS_COMMON) FLAGS_HC2CF2=$(RDFT_FLAGS_COMMON) -twiddle-log3 -precompute-twiddles FLAGS_R2CFII=$(RDFT_FLAGS_COMMON) r2cf_%.c: $(CODELET_DEPS) $(GEN_R2CF) ($(PRELUDE_COMMANDS_RDFT); $(TWOVERS) $(GEN_R2CF) $(FLAGS_R2CF) -n $* -name r2cf_$* -include "r2cf.h") | $(ADD_DATE) | $(INDENT) >$@ hf_%.c: $(CODELET_DEPS) $(GEN_HC2HC) ($(PRELUDE_COMMANDS_RDFT); $(TWOVERS) $(GEN_HC2HC) $(FLAGS_HF) -n $* -dit -name hf_$* -include "hf.h") | $(ADD_DATE) | $(INDENT) >$@ hf2_%.c: $(CODELET_DEPS) $(GEN_HC2HC) ($(PRELUDE_COMMANDS_RDFT); $(TWOVERS) $(GEN_HC2HC) $(FLAGS_HF2) -n $* -dit -name hf2_$* -include "hf.h") | $(ADD_DATE) | $(INDENT) >$@ r2cfII_%.c: $(CODELET_DEPS) $(GEN_R2CF) ($(PRELUDE_COMMANDS_RDFT); $(TWOVERS) $(GEN_R2CF) $(FLAGS_R2CF) -n $* -name r2cfII_$* -dft-II -include "r2cfII.h") | $(ADD_DATE) | $(INDENT) >$@ hc2cf_%.c: $(CODELET_DEPS) $(GEN_HC2C) ($(PRELUDE_COMMANDS_RDFT); $(TWOVERS) $(GEN_HC2C) $(FLAGS_HC2CF) -n $* -dit -name hc2cf_$* -include "hc2cf.h") | $(ADD_DATE) | $(INDENT) >$@ hc2cf2_%.c: $(CODELET_DEPS) $(GEN_HC2C) ($(PRELUDE_COMMANDS_RDFT); $(TWOVERS) $(GEN_HC2C) $(FLAGS_HC2CF2) -n $* -dit -name hc2cf2_$* -include "hc2cf.h") | $(ADD_DATE) | $(INDENT) >$@ hc2cfdft_%.c: $(CODELET_DEPS) $(GEN_HC2CDFT) ($(PRELUDE_COMMANDS_RDFT); $(TWOVERS) $(GEN_HC2CDFT) $(FLAGS_HC2CF) -n $* -dit -name hc2cfdft_$* -include "hc2cf.h") | $(ADD_DATE) | $(INDENT) >$@ hc2cfdft2_%.c: $(CODELET_DEPS) $(GEN_HC2CDFT) ($(PRELUDE_COMMANDS_RDFT); $(TWOVERS) $(GEN_HC2CDFT) $(FLAGS_HC2CF2) -n $* -dit -name hc2cfdft2_$* -include "hc2cf.h") | $(ADD_DATE) | $(INDENT) >$@ endif # MAINTAINER_MODE fftw-3.3.3/rdft/scalar/r2cf/Makefile.in0000644000175000017500000010024012054410200014475 00000000000000# Makefile.in generated by automake 1.11.6 from Makefile.am. # @configure_input@ # Copyright (C) 1994, 1995, 1996, 1997, 1998, 1999, 2000, 2001, 2002, # 2003, 2004, 2005, 2006, 2007, 2008, 2009, 2010, 2011 Free Software # Foundation, Inc. # This Makefile.in is free software; the Free Software Foundation # gives unlimited permission to copy and/or distribute it, # with or without modifications, as long as this notice is preserved. # This program is distributed in the hope that it will be useful, # but WITHOUT ANY WARRANTY, to the extent permitted by law; without # even the implied warranty of MERCHANTABILITY or FITNESS FOR A # PARTICULAR PURPOSE. @SET_MAKE@ # This Makefile.am specifies a set of codelets, efficient transforms # of small sizes, that are used as building blocks (kernels) by FFTW # to build up large transforms, as well as the options for generating # and compiling them. # You can customize FFTW for special needs, e.g. to handle certain # sizes more efficiently, by adding new codelets to the lists of those # included by default. If you change the list of codelets, any new # ones you added will be automatically generated when you run the # bootstrap script (see "Generating your own code" in the FFTW # manual). # -*- makefile -*- # This file contains special make rules to generate codelets. # Most of this file requires GNU make . VPATH = @srcdir@ am__make_dryrun = \ { \ am__dry=no; \ case $$MAKEFLAGS in \ *\\[\ \ ]*) \ echo 'am--echo: ; @echo "AM" OK' | $(MAKE) -f - 2>/dev/null \ | grep '^AM OK$$' >/dev/null || am__dry=yes;; \ *) \ for am__flg in $$MAKEFLAGS; do \ case $$am__flg in \ *=*|--*) ;; \ *n*) am__dry=yes; break;; \ esac; \ done;; \ esac; \ test $$am__dry = yes; \ } pkgdatadir = $(datadir)/@PACKAGE@ pkgincludedir = $(includedir)/@PACKAGE@ pkglibdir = $(libdir)/@PACKAGE@ pkglibexecdir = $(libexecdir)/@PACKAGE@ am__cd = CDPATH="$${ZSH_VERSION+.}$(PATH_SEPARATOR)" && cd install_sh_DATA = $(install_sh) -c -m 644 install_sh_PROGRAM = $(install_sh) -c install_sh_SCRIPT = $(install_sh) -c INSTALL_HEADER = $(INSTALL_DATA) transform = $(program_transform_name) NORMAL_INSTALL = : PRE_INSTALL = : POST_INSTALL = : NORMAL_UNINSTALL = : PRE_UNINSTALL = : POST_UNINSTALL = : build_triplet = @build@ host_triplet = @host@ DIST_COMMON = $(srcdir)/Makefile.am $(srcdir)/Makefile.in \ $(top_srcdir)/support/Makefile.codelets subdir = rdft/scalar/r2cf ACLOCAL_M4 = $(top_srcdir)/aclocal.m4 am__aclocal_m4_deps = $(top_srcdir)/m4/acx_mpi.m4 \ $(top_srcdir)/m4/acx_pthread.m4 \ $(top_srcdir)/m4/ax_cc_maxopt.m4 \ $(top_srcdir)/m4/ax_check_compiler_flags.m4 \ $(top_srcdir)/m4/ax_compiler_vendor.m4 \ $(top_srcdir)/m4/ax_gcc_aligns_stack.m4 \ $(top_srcdir)/m4/ax_gcc_version.m4 \ $(top_srcdir)/m4/ax_openmp.m4 $(top_srcdir)/m4/libtool.m4 \ $(top_srcdir)/m4/ltoptions.m4 $(top_srcdir)/m4/ltsugar.m4 \ $(top_srcdir)/m4/ltversion.m4 $(top_srcdir)/m4/lt~obsolete.m4 \ $(top_srcdir)/configure.ac am__configure_deps = $(am__aclocal_m4_deps) $(CONFIGURE_DEPENDENCIES) \ $(ACLOCAL_M4) mkinstalldirs = $(install_sh) -d CONFIG_HEADER = $(top_builddir)/config.h CONFIG_CLEAN_FILES = CONFIG_CLEAN_VPATH_FILES = LTLIBRARIES = $(noinst_LTLIBRARIES) librdft_scalar_r2cf_la_LIBADD = am__objects_1 = r2cf_2.lo r2cf_3.lo r2cf_4.lo r2cf_5.lo r2cf_6.lo \ r2cf_7.lo r2cf_8.lo r2cf_9.lo r2cf_10.lo r2cf_11.lo r2cf_12.lo \ r2cf_13.lo r2cf_14.lo r2cf_15.lo r2cf_16.lo r2cf_32.lo \ r2cf_64.lo r2cf_128.lo r2cf_20.lo r2cf_25.lo am__objects_2 = hf_2.lo hf_3.lo hf_4.lo hf_5.lo hf_6.lo hf_7.lo \ hf_8.lo hf_9.lo hf_10.lo hf_12.lo hf_15.lo hf_16.lo hf_32.lo \ hf_64.lo hf_20.lo hf_25.lo am__objects_3 = hf2_4.lo hf2_8.lo hf2_16.lo hf2_32.lo hf2_5.lo \ hf2_20.lo hf2_25.lo am__objects_4 = r2cfII_2.lo r2cfII_3.lo r2cfII_4.lo r2cfII_5.lo \ r2cfII_6.lo r2cfII_7.lo r2cfII_8.lo r2cfII_9.lo r2cfII_10.lo \ r2cfII_12.lo r2cfII_15.lo r2cfII_16.lo r2cfII_32.lo \ r2cfII_64.lo r2cfII_20.lo r2cfII_25.lo am__objects_5 = hc2cf_2.lo hc2cf_4.lo hc2cf_6.lo hc2cf_8.lo \ hc2cf_10.lo hc2cf_12.lo hc2cf_16.lo hc2cf_32.lo hc2cf_20.lo am__objects_6 = hc2cf2_4.lo hc2cf2_8.lo hc2cf2_16.lo hc2cf2_32.lo \ hc2cf2_20.lo am__objects_7 = hc2cfdft_2.lo hc2cfdft_4.lo hc2cfdft_6.lo \ hc2cfdft_8.lo hc2cfdft_10.lo hc2cfdft_12.lo hc2cfdft_16.lo \ hc2cfdft_32.lo hc2cfdft_20.lo am__objects_8 = hc2cfdft2_4.lo hc2cfdft2_8.lo hc2cfdft2_16.lo \ hc2cfdft2_32.lo hc2cfdft2_20.lo am__objects_9 = $(am__objects_1) $(am__objects_2) $(am__objects_3) \ $(am__objects_4) $(am__objects_5) $(am__objects_6) \ $(am__objects_7) $(am__objects_8) am__objects_10 = codlist.lo am__objects_11 = $(am__objects_9) $(am__objects_10) am_librdft_scalar_r2cf_la_OBJECTS = $(am__objects_11) librdft_scalar_r2cf_la_OBJECTS = $(am_librdft_scalar_r2cf_la_OBJECTS) DEFAULT_INCLUDES = -I.@am__isrc@ -I$(top_builddir) depcomp = $(SHELL) $(top_srcdir)/depcomp am__depfiles_maybe = depfiles am__mv = mv -f COMPILE = $(CC) $(DEFS) $(DEFAULT_INCLUDES) $(INCLUDES) $(AM_CPPFLAGS) \ $(CPPFLAGS) $(AM_CFLAGS) $(CFLAGS) LTCOMPILE = $(LIBTOOL) --tag=CC $(AM_LIBTOOLFLAGS) $(LIBTOOLFLAGS) \ --mode=compile $(CC) $(DEFS) $(DEFAULT_INCLUDES) $(INCLUDES) \ $(AM_CPPFLAGS) $(CPPFLAGS) $(AM_CFLAGS) $(CFLAGS) CCLD = $(CC) LINK = $(LIBTOOL) --tag=CC $(AM_LIBTOOLFLAGS) $(LIBTOOLFLAGS) \ --mode=link $(CCLD) $(AM_CFLAGS) $(CFLAGS) $(AM_LDFLAGS) \ $(LDFLAGS) -o $@ SOURCES = $(librdft_scalar_r2cf_la_SOURCES) DIST_SOURCES = $(librdft_scalar_r2cf_la_SOURCES) am__can_run_installinfo = \ case $$AM_UPDATE_INFO_DIR in \ n|no|NO) false;; \ *) (install-info --version) >/dev/null 2>&1;; \ esac ETAGS = etags CTAGS = ctags DISTFILES = $(DIST_COMMON) $(DIST_SOURCES) $(TEXINFOS) $(EXTRA_DIST) ACLOCAL = @ACLOCAL@ ALLOCA = @ALLOCA@ ALTIVEC_CFLAGS = @ALTIVEC_CFLAGS@ AMTAR = @AMTAR@ AR = @AR@ AS = @AS@ AUTOCONF = @AUTOCONF@ AUTOHEADER = @AUTOHEADER@ AUTOMAKE = @AUTOMAKE@ AVX_CFLAGS = @AVX_CFLAGS@ AWK = @AWK@ CC = @CC@ CCDEPMODE = @CCDEPMODE@ CFLAGS = @CFLAGS@ CHECK_PL_OPTS = @CHECK_PL_OPTS@ CPP = @CPP@ CPPFLAGS = @CPPFLAGS@ CYGPATH_W = @CYGPATH_W@ C_FFTW_R2R_KIND = @C_FFTW_R2R_KIND@ C_MPI_FINT = @C_MPI_FINT@ DEFS = @DEFS@ DEPDIR = @DEPDIR@ DLLTOOL = @DLLTOOL@ DSYMUTIL = @DSYMUTIL@ DUMPBIN = @DUMPBIN@ ECHO_C = @ECHO_C@ ECHO_N = @ECHO_N@ ECHO_T = @ECHO_T@ EGREP = @EGREP@ EXEEXT = @EXEEXT@ F77 = @F77@ FFLAGS = @FFLAGS@ FGREP = @FGREP@ FLIBS = @FLIBS@ GREP = @GREP@ INSTALL = @INSTALL@ INSTALL_DATA = @INSTALL_DATA@ INSTALL_PROGRAM = @INSTALL_PROGRAM@ INSTALL_SCRIPT = @INSTALL_SCRIPT@ INSTALL_STRIP_PROGRAM = @INSTALL_STRIP_PROGRAM@ LD = @LD@ LDFLAGS = @LDFLAGS@ LIBOBJS = @LIBOBJS@ LIBQUADMATH = @LIBQUADMATH@ LIBS = @LIBS@ LIBTOOL = @LIBTOOL@ LIPO = @LIPO@ LN_S = @LN_S@ LTLIBOBJS = @LTLIBOBJS@ MAINT = @MAINT@ MAKEINFO = @MAKEINFO@ MANIFEST_TOOL = @MANIFEST_TOOL@ MKDIR_P = @MKDIR_P@ MPICC = @MPICC@ MPILIBS = @MPILIBS@ MPIRUN = @MPIRUN@ NEON_CFLAGS = @NEON_CFLAGS@ NM = @NM@ NMEDIT = @NMEDIT@ OBJDUMP = @OBJDUMP@ OBJEXT = @OBJEXT@ OCAMLBUILD = @OCAMLBUILD@ OPENMP_CFLAGS = @OPENMP_CFLAGS@ OTOOL = @OTOOL@ OTOOL64 = @OTOOL64@ PACKAGE = @PACKAGE@ PACKAGE_BUGREPORT = @PACKAGE_BUGREPORT@ PACKAGE_NAME = @PACKAGE_NAME@ PACKAGE_STRING = @PACKAGE_STRING@ PACKAGE_TARNAME = @PACKAGE_TARNAME@ PACKAGE_URL = @PACKAGE_URL@ PACKAGE_VERSION = @PACKAGE_VERSION@ PATH_SEPARATOR = @PATH_SEPARATOR@ POW_LIB = @POW_LIB@ PRECISION = @PRECISION@ PREC_SUFFIX = @PREC_SUFFIX@ PTHREAD_CC = @PTHREAD_CC@ PTHREAD_CFLAGS = @PTHREAD_CFLAGS@ PTHREAD_LIBS = @PTHREAD_LIBS@ RANLIB = @RANLIB@ SED = @SED@ SET_MAKE = @SET_MAKE@ SHARED_VERSION_INFO = @SHARED_VERSION_INFO@ SHELL = @SHELL@ SSE2_CFLAGS = @SSE2_CFLAGS@ STACK_ALIGN_CFLAGS = @STACK_ALIGN_CFLAGS@ STRIP = @STRIP@ THREADLIBS = @THREADLIBS@ VERSION = @VERSION@ abs_builddir = @abs_builddir@ abs_srcdir = @abs_srcdir@ abs_top_builddir = @abs_top_builddir@ abs_top_srcdir = @abs_top_srcdir@ ac_ct_AR = @ac_ct_AR@ ac_ct_CC = @ac_ct_CC@ ac_ct_DUMPBIN = @ac_ct_DUMPBIN@ ac_ct_F77 = @ac_ct_F77@ acx_pthread_config = @acx_pthread_config@ am__include = @am__include@ am__leading_dot = @am__leading_dot@ am__quote = @am__quote@ am__tar = @am__tar@ am__untar = @am__untar@ bindir = @bindir@ build = @build@ build_alias = @build_alias@ build_cpu = @build_cpu@ build_os = @build_os@ build_vendor = @build_vendor@ builddir = @builddir@ datadir = @datadir@ datarootdir = @datarootdir@ docdir = @docdir@ dvidir = @dvidir@ exec_prefix = @exec_prefix@ host = @host@ host_alias = @host_alias@ host_cpu = @host_cpu@ host_os = @host_os@ host_vendor = @host_vendor@ htmldir = @htmldir@ includedir = @includedir@ infodir = @infodir@ install_sh = @install_sh@ libdir = @libdir@ libexecdir = @libexecdir@ localedir = @localedir@ localstatedir = @localstatedir@ mandir = @mandir@ mkdir_p = @mkdir_p@ oldincludedir = @oldincludedir@ pdfdir = @pdfdir@ prefix = @prefix@ program_transform_name = @program_transform_name@ psdir = @psdir@ sbindir = @sbindir@ sharedstatedir = @sharedstatedir@ srcdir = @srcdir@ sysconfdir = @sysconfdir@ target_alias = @target_alias@ top_build_prefix = @top_build_prefix@ top_builddir = @top_builddir@ top_srcdir = @top_srcdir@ ########################################################################### AM_CPPFLAGS = -I$(top_srcdir)/kernel -I$(top_srcdir)/rdft \ -I$(top_srcdir)/rdft/scalar noinst_LTLIBRARIES = librdft_scalar_r2cf.la ########################################################################### # r2cf_ is a hard-coded real-to-complex FFT of size (base cases # of real-input FFT recursion) R2CF = r2cf_2.c r2cf_3.c r2cf_4.c r2cf_5.c r2cf_6.c r2cf_7.c r2cf_8.c \ r2cf_9.c r2cf_10.c r2cf_11.c r2cf_12.c r2cf_13.c r2cf_14.c r2cf_15.c \ r2cf_16.c r2cf_32.c r2cf_64.c r2cf_128.c \ r2cf_20.c r2cf_25.c # r2cf_30.c r2cf_40.c r2cf_50.c ########################################################################### # hf_ is a "twiddle" FFT of size , implementing a radix-r DIT # step for a real-input FFT. Every hf codelet must have a # corresponding r2cfII codelet (see below)! HF = hf_2.c hf_3.c hf_4.c hf_5.c hf_6.c hf_7.c hf_8.c hf_9.c \ hf_10.c hf_12.c hf_15.c hf_16.c hf_32.c hf_64.c \ hf_20.c hf_25.c # hf_30.c hf_40.c hf_50.c # like hf, but generates part of its trig table on the fly (good for large n) HF2 = hf2_4.c hf2_8.c hf2_16.c hf2_32.c \ hf2_5.c hf2_20.c hf2_25.c # an r2cf transform where the input is shifted by half a sample (output # is multiplied by a phase). This is needed as part of the DIT recursion; # every hf_ or hf2_ codelet should have a corresponding r2cfII_ R2CFII = r2cfII_2.c r2cfII_3.c r2cfII_4.c r2cfII_5.c r2cfII_6.c \ r2cfII_7.c r2cfII_8.c r2cfII_9.c r2cfII_10.c r2cfII_12.c r2cfII_15.c \ r2cfII_16.c r2cfII_32.c r2cfII_64.c \ r2cfII_20.c r2cfII_25.c # r2cfII_30.c r2cfII_40.c r2cfII_50.c ########################################################################### # hc2cf_ is a "twiddle" FFT of size , implementing a radix-r DIT # step for a real-input FFT with rdft2-style output. must be even. HC2CF = hc2cf_2.c hc2cf_4.c hc2cf_6.c hc2cf_8.c hc2cf_10.c hc2cf_12.c \ hc2cf_16.c hc2cf_32.c \ hc2cf_20.c # hc2cf_30.c HC2CFDFT = hc2cfdft_2.c hc2cfdft_4.c hc2cfdft_6.c hc2cfdft_8.c \ hc2cfdft_10.c hc2cfdft_12.c hc2cfdft_16.c hc2cfdft_32.c \ hc2cfdft_20.c # hc2cfdft_30.c # like hc2cf, but generates part of its trig table on the fly (good # for large n) HC2CF2 = hc2cf2_4.c hc2cf2_8.c hc2cf2_16.c hc2cf2_32.c \ hc2cf2_20.c # hc2cf2_30.c HC2CFDFT2 = hc2cfdft2_4.c hc2cfdft2_8.c hc2cfdft2_16.c hc2cfdft2_32.c \ hc2cfdft2_20.c # hc2cfdft2_30.c ########################################################################### ALL_CODELETS = $(R2CF) $(HF) $(HF2) $(R2CFII) $(HC2CF) $(HC2CF2) \ $(HC2CFDFT) $(HC2CFDFT2) BUILT_SOURCES = $(ALL_CODELETS) $(CODLIST) librdft_scalar_r2cf_la_SOURCES = $(BUILT_SOURCES) SOLVTAB_NAME = X(solvtab_rdft_r2cf) XRENAME = X CODLIST = codlist.c CODELET_NAME = codelet_ @MAINTAINER_MODE_TRUE@INDENT = indent -kr -cs -i5 -l800 -fca -nfc1 -sc -sob -cli4 -TR -Tplanner -TV @MAINTAINER_MODE_TRUE@TWOVERS = sh ${top_srcdir}/support/twovers.sh @MAINTAINER_MODE_TRUE@GENFFTDIR = ${top_builddir}/genfft @MAINTAINER_MODE_TRUE@GEN_NOTW = ${GENFFTDIR}/gen_notw.native @MAINTAINER_MODE_TRUE@GEN_NOTW_C = ${GENFFTDIR}/gen_notw_c.native @MAINTAINER_MODE_TRUE@GEN_TWIDDLE = ${GENFFTDIR}/gen_twiddle.native @MAINTAINER_MODE_TRUE@GEN_TWIDDLE_C = ${GENFFTDIR}/gen_twiddle_c.native @MAINTAINER_MODE_TRUE@GEN_TWIDSQ = ${GENFFTDIR}/gen_twidsq.native @MAINTAINER_MODE_TRUE@GEN_TWIDSQ_C = ${GENFFTDIR}/gen_twidsq_c.native @MAINTAINER_MODE_TRUE@GEN_R2CF = ${GENFFTDIR}/gen_r2cf.native @MAINTAINER_MODE_TRUE@GEN_R2CB = ${GENFFTDIR}/gen_r2cb.native @MAINTAINER_MODE_TRUE@GEN_HC2HC = ${GENFFTDIR}/gen_hc2hc.native @MAINTAINER_MODE_TRUE@GEN_HC2C = ${GENFFTDIR}/gen_hc2c.native @MAINTAINER_MODE_TRUE@GEN_HC2CDFT = ${GENFFTDIR}/gen_hc2cdft.native @MAINTAINER_MODE_TRUE@GEN_HC2CDFT_C = ${GENFFTDIR}/gen_hc2cdft_c.native @MAINTAINER_MODE_TRUE@GEN_R2R = ${GENFFTDIR}/gen_r2r.native @MAINTAINER_MODE_TRUE@PRELUDE_DFT = ${top_srcdir}/support/codelet_prelude.dft @MAINTAINER_MODE_TRUE@PRELUDE_RDFT = ${top_srcdir}/support/codelet_prelude.rdft @MAINTAINER_MODE_TRUE@ADD_DATE = sed -e s/@DATE@/"`date`"/ @MAINTAINER_MODE_TRUE@COPYRIGHT = ${top_srcdir}/COPYRIGHT @MAINTAINER_MODE_TRUE@CODELET_DEPS = $(COPYRIGHT) $(PRELUDE) @MAINTAINER_MODE_TRUE@PRELUDE_COMMANDS_DFT = cat $(COPYRIGHT) $(PRELUDE_DFT) @MAINTAINER_MODE_TRUE@PRELUDE_COMMANDS_RDFT = cat $(COPYRIGHT) $(PRELUDE_RDFT) @MAINTAINER_MODE_TRUE@FLAGS_COMMON = -compact -variables 4 @MAINTAINER_MODE_TRUE@DFT_FLAGS_COMMON = $(FLAGS_COMMON) -pipeline-latency 4 @MAINTAINER_MODE_TRUE@RDFT_FLAGS_COMMON = $(FLAGS_COMMON) -pipeline-latency 4 # special rules for regenerating codelets. @MAINTAINER_MODE_TRUE@FLAGS_R2CF = $(RDFT_FLAGS_COMMON) @MAINTAINER_MODE_TRUE@FLAGS_HF = $(RDFT_FLAGS_COMMON) @MAINTAINER_MODE_TRUE@FLAGS_HF2 = $(RDFT_FLAGS_COMMON) -twiddle-log3 -precompute-twiddles @MAINTAINER_MODE_TRUE@FLAGS_HC2CF = $(RDFT_FLAGS_COMMON) @MAINTAINER_MODE_TRUE@FLAGS_HC2CF2 = $(RDFT_FLAGS_COMMON) -twiddle-log3 -precompute-twiddles @MAINTAINER_MODE_TRUE@FLAGS_R2CFII = $(RDFT_FLAGS_COMMON) all: $(BUILT_SOURCES) $(MAKE) $(AM_MAKEFLAGS) all-am .SUFFIXES: .SUFFIXES: .c .lo .o .obj $(srcdir)/Makefile.in: @MAINTAINER_MODE_TRUE@ $(srcdir)/Makefile.am $(top_srcdir)/support/Makefile.codelets $(am__configure_deps) @for dep in $?; do \ case '$(am__configure_deps)' in \ *$$dep*) \ ( cd $(top_builddir) && $(MAKE) $(AM_MAKEFLAGS) am--refresh ) \ && { if test -f $@; then exit 0; else break; fi; }; \ exit 1;; \ esac; \ done; \ echo ' cd $(top_srcdir) && $(AUTOMAKE) --gnu rdft/scalar/r2cf/Makefile'; \ $(am__cd) $(top_srcdir) && \ $(AUTOMAKE) --gnu rdft/scalar/r2cf/Makefile .PRECIOUS: Makefile Makefile: $(srcdir)/Makefile.in $(top_builddir)/config.status @case '$?' in \ *config.status*) \ cd $(top_builddir) && $(MAKE) $(AM_MAKEFLAGS) am--refresh;; \ *) \ echo ' cd $(top_builddir) && $(SHELL) ./config.status $(subdir)/$@ $(am__depfiles_maybe)'; \ cd $(top_builddir) && $(SHELL) ./config.status $(subdir)/$@ $(am__depfiles_maybe);; \ esac; $(top_srcdir)/support/Makefile.codelets: $(top_builddir)/config.status: $(top_srcdir)/configure $(CONFIG_STATUS_DEPENDENCIES) cd $(top_builddir) && $(MAKE) $(AM_MAKEFLAGS) am--refresh $(top_srcdir)/configure: @MAINTAINER_MODE_TRUE@ $(am__configure_deps) cd $(top_builddir) && $(MAKE) $(AM_MAKEFLAGS) am--refresh $(ACLOCAL_M4): @MAINTAINER_MODE_TRUE@ $(am__aclocal_m4_deps) cd $(top_builddir) && $(MAKE) $(AM_MAKEFLAGS) am--refresh $(am__aclocal_m4_deps): clean-noinstLTLIBRARIES: -test -z "$(noinst_LTLIBRARIES)" || rm -f $(noinst_LTLIBRARIES) @list='$(noinst_LTLIBRARIES)'; for p in $$list; do \ dir="`echo $$p | sed -e 's|/[^/]*$$||'`"; \ test "$$dir" != "$$p" || dir=.; \ echo "rm -f \"$${dir}/so_locations\""; \ rm -f "$${dir}/so_locations"; \ done librdft_scalar_r2cf.la: $(librdft_scalar_r2cf_la_OBJECTS) $(librdft_scalar_r2cf_la_DEPENDENCIES) $(EXTRA_librdft_scalar_r2cf_la_DEPENDENCIES) $(LINK) $(librdft_scalar_r2cf_la_OBJECTS) $(librdft_scalar_r2cf_la_LIBADD) $(LIBS) mostlyclean-compile: -rm -f *.$(OBJEXT) distclean-compile: -rm -f *.tab.c @AMDEP_TRUE@@am__include@ @am__quote@./$(DEPDIR)/codlist.Plo@am__quote@ @AMDEP_TRUE@@am__include@ @am__quote@./$(DEPDIR)/hc2cf2_16.Plo@am__quote@ @AMDEP_TRUE@@am__include@ @am__quote@./$(DEPDIR)/hc2cf2_20.Plo@am__quote@ @AMDEP_TRUE@@am__include@ @am__quote@./$(DEPDIR)/hc2cf2_32.Plo@am__quote@ @AMDEP_TRUE@@am__include@ @am__quote@./$(DEPDIR)/hc2cf2_4.Plo@am__quote@ @AMDEP_TRUE@@am__include@ @am__quote@./$(DEPDIR)/hc2cf2_8.Plo@am__quote@ @AMDEP_TRUE@@am__include@ @am__quote@./$(DEPDIR)/hc2cf_10.Plo@am__quote@ @AMDEP_TRUE@@am__include@ @am__quote@./$(DEPDIR)/hc2cf_12.Plo@am__quote@ @AMDEP_TRUE@@am__include@ @am__quote@./$(DEPDIR)/hc2cf_16.Plo@am__quote@ @AMDEP_TRUE@@am__include@ @am__quote@./$(DEPDIR)/hc2cf_2.Plo@am__quote@ @AMDEP_TRUE@@am__include@ @am__quote@./$(DEPDIR)/hc2cf_20.Plo@am__quote@ @AMDEP_TRUE@@am__include@ @am__quote@./$(DEPDIR)/hc2cf_32.Plo@am__quote@ @AMDEP_TRUE@@am__include@ @am__quote@./$(DEPDIR)/hc2cf_4.Plo@am__quote@ @AMDEP_TRUE@@am__include@ @am__quote@./$(DEPDIR)/hc2cf_6.Plo@am__quote@ @AMDEP_TRUE@@am__include@ @am__quote@./$(DEPDIR)/hc2cf_8.Plo@am__quote@ @AMDEP_TRUE@@am__include@ @am__quote@./$(DEPDIR)/hc2cfdft2_16.Plo@am__quote@ @AMDEP_TRUE@@am__include@ @am__quote@./$(DEPDIR)/hc2cfdft2_20.Plo@am__quote@ @AMDEP_TRUE@@am__include@ @am__quote@./$(DEPDIR)/hc2cfdft2_32.Plo@am__quote@ @AMDEP_TRUE@@am__include@ @am__quote@./$(DEPDIR)/hc2cfdft2_4.Plo@am__quote@ @AMDEP_TRUE@@am__include@ @am__quote@./$(DEPDIR)/hc2cfdft2_8.Plo@am__quote@ @AMDEP_TRUE@@am__include@ @am__quote@./$(DEPDIR)/hc2cfdft_10.Plo@am__quote@ @AMDEP_TRUE@@am__include@ @am__quote@./$(DEPDIR)/hc2cfdft_12.Plo@am__quote@ @AMDEP_TRUE@@am__include@ @am__quote@./$(DEPDIR)/hc2cfdft_16.Plo@am__quote@ @AMDEP_TRUE@@am__include@ @am__quote@./$(DEPDIR)/hc2cfdft_2.Plo@am__quote@ @AMDEP_TRUE@@am__include@ @am__quote@./$(DEPDIR)/hc2cfdft_20.Plo@am__quote@ @AMDEP_TRUE@@am__include@ @am__quote@./$(DEPDIR)/hc2cfdft_32.Plo@am__quote@ @AMDEP_TRUE@@am__include@ @am__quote@./$(DEPDIR)/hc2cfdft_4.Plo@am__quote@ @AMDEP_TRUE@@am__include@ @am__quote@./$(DEPDIR)/hc2cfdft_6.Plo@am__quote@ @AMDEP_TRUE@@am__include@ @am__quote@./$(DEPDIR)/hc2cfdft_8.Plo@am__quote@ @AMDEP_TRUE@@am__include@ @am__quote@./$(DEPDIR)/hf2_16.Plo@am__quote@ @AMDEP_TRUE@@am__include@ @am__quote@./$(DEPDIR)/hf2_20.Plo@am__quote@ @AMDEP_TRUE@@am__include@ @am__quote@./$(DEPDIR)/hf2_25.Plo@am__quote@ @AMDEP_TRUE@@am__include@ @am__quote@./$(DEPDIR)/hf2_32.Plo@am__quote@ @AMDEP_TRUE@@am__include@ @am__quote@./$(DEPDIR)/hf2_4.Plo@am__quote@ @AMDEP_TRUE@@am__include@ @am__quote@./$(DEPDIR)/hf2_5.Plo@am__quote@ @AMDEP_TRUE@@am__include@ @am__quote@./$(DEPDIR)/hf2_8.Plo@am__quote@ @AMDEP_TRUE@@am__include@ @am__quote@./$(DEPDIR)/hf_10.Plo@am__quote@ @AMDEP_TRUE@@am__include@ @am__quote@./$(DEPDIR)/hf_12.Plo@am__quote@ @AMDEP_TRUE@@am__include@ @am__quote@./$(DEPDIR)/hf_15.Plo@am__quote@ @AMDEP_TRUE@@am__include@ @am__quote@./$(DEPDIR)/hf_16.Plo@am__quote@ @AMDEP_TRUE@@am__include@ @am__quote@./$(DEPDIR)/hf_2.Plo@am__quote@ @AMDEP_TRUE@@am__include@ @am__quote@./$(DEPDIR)/hf_20.Plo@am__quote@ @AMDEP_TRUE@@am__include@ @am__quote@./$(DEPDIR)/hf_25.Plo@am__quote@ @AMDEP_TRUE@@am__include@ @am__quote@./$(DEPDIR)/hf_3.Plo@am__quote@ @AMDEP_TRUE@@am__include@ @am__quote@./$(DEPDIR)/hf_32.Plo@am__quote@ @AMDEP_TRUE@@am__include@ @am__quote@./$(DEPDIR)/hf_4.Plo@am__quote@ @AMDEP_TRUE@@am__include@ @am__quote@./$(DEPDIR)/hf_5.Plo@am__quote@ @AMDEP_TRUE@@am__include@ @am__quote@./$(DEPDIR)/hf_6.Plo@am__quote@ @AMDEP_TRUE@@am__include@ @am__quote@./$(DEPDIR)/hf_64.Plo@am__quote@ @AMDEP_TRUE@@am__include@ @am__quote@./$(DEPDIR)/hf_7.Plo@am__quote@ @AMDEP_TRUE@@am__include@ @am__quote@./$(DEPDIR)/hf_8.Plo@am__quote@ @AMDEP_TRUE@@am__include@ @am__quote@./$(DEPDIR)/hf_9.Plo@am__quote@ @AMDEP_TRUE@@am__include@ @am__quote@./$(DEPDIR)/r2cfII_10.Plo@am__quote@ @AMDEP_TRUE@@am__include@ @am__quote@./$(DEPDIR)/r2cfII_12.Plo@am__quote@ @AMDEP_TRUE@@am__include@ @am__quote@./$(DEPDIR)/r2cfII_15.Plo@am__quote@ @AMDEP_TRUE@@am__include@ @am__quote@./$(DEPDIR)/r2cfII_16.Plo@am__quote@ @AMDEP_TRUE@@am__include@ @am__quote@./$(DEPDIR)/r2cfII_2.Plo@am__quote@ @AMDEP_TRUE@@am__include@ @am__quote@./$(DEPDIR)/r2cfII_20.Plo@am__quote@ @AMDEP_TRUE@@am__include@ @am__quote@./$(DEPDIR)/r2cfII_25.Plo@am__quote@ @AMDEP_TRUE@@am__include@ @am__quote@./$(DEPDIR)/r2cfII_3.Plo@am__quote@ @AMDEP_TRUE@@am__include@ @am__quote@./$(DEPDIR)/r2cfII_32.Plo@am__quote@ @AMDEP_TRUE@@am__include@ @am__quote@./$(DEPDIR)/r2cfII_4.Plo@am__quote@ @AMDEP_TRUE@@am__include@ @am__quote@./$(DEPDIR)/r2cfII_5.Plo@am__quote@ @AMDEP_TRUE@@am__include@ @am__quote@./$(DEPDIR)/r2cfII_6.Plo@am__quote@ @AMDEP_TRUE@@am__include@ @am__quote@./$(DEPDIR)/r2cfII_64.Plo@am__quote@ @AMDEP_TRUE@@am__include@ @am__quote@./$(DEPDIR)/r2cfII_7.Plo@am__quote@ @AMDEP_TRUE@@am__include@ @am__quote@./$(DEPDIR)/r2cfII_8.Plo@am__quote@ @AMDEP_TRUE@@am__include@ @am__quote@./$(DEPDIR)/r2cfII_9.Plo@am__quote@ @AMDEP_TRUE@@am__include@ @am__quote@./$(DEPDIR)/r2cf_10.Plo@am__quote@ @AMDEP_TRUE@@am__include@ @am__quote@./$(DEPDIR)/r2cf_11.Plo@am__quote@ @AMDEP_TRUE@@am__include@ @am__quote@./$(DEPDIR)/r2cf_12.Plo@am__quote@ @AMDEP_TRUE@@am__include@ @am__quote@./$(DEPDIR)/r2cf_128.Plo@am__quote@ @AMDEP_TRUE@@am__include@ @am__quote@./$(DEPDIR)/r2cf_13.Plo@am__quote@ @AMDEP_TRUE@@am__include@ @am__quote@./$(DEPDIR)/r2cf_14.Plo@am__quote@ @AMDEP_TRUE@@am__include@ @am__quote@./$(DEPDIR)/r2cf_15.Plo@am__quote@ @AMDEP_TRUE@@am__include@ @am__quote@./$(DEPDIR)/r2cf_16.Plo@am__quote@ @AMDEP_TRUE@@am__include@ @am__quote@./$(DEPDIR)/r2cf_2.Plo@am__quote@ @AMDEP_TRUE@@am__include@ @am__quote@./$(DEPDIR)/r2cf_20.Plo@am__quote@ @AMDEP_TRUE@@am__include@ @am__quote@./$(DEPDIR)/r2cf_25.Plo@am__quote@ @AMDEP_TRUE@@am__include@ @am__quote@./$(DEPDIR)/r2cf_3.Plo@am__quote@ @AMDEP_TRUE@@am__include@ @am__quote@./$(DEPDIR)/r2cf_32.Plo@am__quote@ @AMDEP_TRUE@@am__include@ @am__quote@./$(DEPDIR)/r2cf_4.Plo@am__quote@ @AMDEP_TRUE@@am__include@ @am__quote@./$(DEPDIR)/r2cf_5.Plo@am__quote@ @AMDEP_TRUE@@am__include@ @am__quote@./$(DEPDIR)/r2cf_6.Plo@am__quote@ @AMDEP_TRUE@@am__include@ @am__quote@./$(DEPDIR)/r2cf_64.Plo@am__quote@ @AMDEP_TRUE@@am__include@ @am__quote@./$(DEPDIR)/r2cf_7.Plo@am__quote@ @AMDEP_TRUE@@am__include@ @am__quote@./$(DEPDIR)/r2cf_8.Plo@am__quote@ @AMDEP_TRUE@@am__include@ @am__quote@./$(DEPDIR)/r2cf_9.Plo@am__quote@ .c.o: @am__fastdepCC_TRUE@ $(COMPILE) -MT $@ -MD -MP -MF $(DEPDIR)/$*.Tpo -c -o $@ $< @am__fastdepCC_TRUE@ $(am__mv) $(DEPDIR)/$*.Tpo $(DEPDIR)/$*.Po @AMDEP_TRUE@@am__fastdepCC_FALSE@ source='$<' object='$@' libtool=no @AMDEPBACKSLASH@ @AMDEP_TRUE@@am__fastdepCC_FALSE@ DEPDIR=$(DEPDIR) $(CCDEPMODE) $(depcomp) @AMDEPBACKSLASH@ @am__fastdepCC_FALSE@ $(COMPILE) -c $< .c.obj: @am__fastdepCC_TRUE@ $(COMPILE) -MT $@ -MD -MP -MF $(DEPDIR)/$*.Tpo -c -o $@ `$(CYGPATH_W) '$<'` @am__fastdepCC_TRUE@ $(am__mv) $(DEPDIR)/$*.Tpo $(DEPDIR)/$*.Po @AMDEP_TRUE@@am__fastdepCC_FALSE@ source='$<' object='$@' libtool=no @AMDEPBACKSLASH@ @AMDEP_TRUE@@am__fastdepCC_FALSE@ DEPDIR=$(DEPDIR) $(CCDEPMODE) $(depcomp) @AMDEPBACKSLASH@ @am__fastdepCC_FALSE@ $(COMPILE) -c `$(CYGPATH_W) '$<'` .c.lo: @am__fastdepCC_TRUE@ $(LTCOMPILE) -MT $@ -MD -MP -MF $(DEPDIR)/$*.Tpo -c -o $@ $< @am__fastdepCC_TRUE@ $(am__mv) $(DEPDIR)/$*.Tpo $(DEPDIR)/$*.Plo @AMDEP_TRUE@@am__fastdepCC_FALSE@ source='$<' object='$@' libtool=yes @AMDEPBACKSLASH@ @AMDEP_TRUE@@am__fastdepCC_FALSE@ DEPDIR=$(DEPDIR) $(CCDEPMODE) $(depcomp) @AMDEPBACKSLASH@ @am__fastdepCC_FALSE@ $(LTCOMPILE) -c -o $@ $< mostlyclean-libtool: -rm -f *.lo clean-libtool: -rm -rf .libs _libs ID: $(HEADERS) $(SOURCES) $(LISP) $(TAGS_FILES) list='$(SOURCES) $(HEADERS) $(LISP) $(TAGS_FILES)'; \ unique=`for i in $$list; do \ if test -f "$$i"; then echo $$i; else echo $(srcdir)/$$i; fi; \ done | \ $(AWK) '{ files[$$0] = 1; nonempty = 1; } \ END { if (nonempty) { for (i in files) print i; }; }'`; \ mkid -fID $$unique tags: TAGS TAGS: $(HEADERS) $(SOURCES) $(TAGS_DEPENDENCIES) \ $(TAGS_FILES) $(LISP) set x; \ here=`pwd`; \ list='$(SOURCES) $(HEADERS) $(LISP) $(TAGS_FILES)'; \ unique=`for i in $$list; do \ if test -f "$$i"; then echo $$i; else echo $(srcdir)/$$i; fi; \ done | \ $(AWK) '{ files[$$0] = 1; nonempty = 1; } \ END { if (nonempty) { for (i in files) print i; }; }'`; \ shift; \ if test -z "$(ETAGS_ARGS)$$*$$unique"; then :; else \ test -n "$$unique" || unique=$$empty_fix; \ if test $$# -gt 0; then \ $(ETAGS) $(ETAGSFLAGS) $(AM_ETAGSFLAGS) $(ETAGS_ARGS) \ "$$@" $$unique; \ else \ $(ETAGS) $(ETAGSFLAGS) $(AM_ETAGSFLAGS) $(ETAGS_ARGS) \ $$unique; \ fi; \ fi ctags: CTAGS CTAGS: $(HEADERS) $(SOURCES) $(TAGS_DEPENDENCIES) \ $(TAGS_FILES) $(LISP) list='$(SOURCES) $(HEADERS) $(LISP) $(TAGS_FILES)'; \ unique=`for i in $$list; do \ if test -f "$$i"; then echo $$i; else echo $(srcdir)/$$i; fi; \ done | \ $(AWK) '{ files[$$0] = 1; nonempty = 1; } \ END { if (nonempty) { for (i in files) print i; }; }'`; \ test -z "$(CTAGS_ARGS)$$unique" \ || $(CTAGS) $(CTAGSFLAGS) $(AM_CTAGSFLAGS) $(CTAGS_ARGS) \ $$unique GTAGS: here=`$(am__cd) $(top_builddir) && pwd` \ && $(am__cd) $(top_srcdir) \ && gtags -i $(GTAGS_ARGS) "$$here" distclean-tags: -rm -f TAGS ID GTAGS GRTAGS GSYMS GPATH tags distdir: $(DISTFILES) @srcdirstrip=`echo "$(srcdir)" | sed 's/[].[^$$\\*]/\\\\&/g'`; \ topsrcdirstrip=`echo "$(top_srcdir)" | sed 's/[].[^$$\\*]/\\\\&/g'`; \ list='$(DISTFILES)'; \ dist_files=`for file in $$list; do echo $$file; done | \ sed -e "s|^$$srcdirstrip/||;t" \ -e "s|^$$topsrcdirstrip/|$(top_builddir)/|;t"`; \ case $$dist_files in \ */*) $(MKDIR_P) `echo "$$dist_files" | \ sed '/\//!d;s|^|$(distdir)/|;s,/[^/]*$$,,' | \ sort -u` ;; \ esac; \ for file in $$dist_files; do \ if test -f $$file || test -d $$file; then d=.; else d=$(srcdir); fi; \ if test -d $$d/$$file; then \ dir=`echo "/$$file" | sed -e 's,/[^/]*$$,,'`; \ if test -d "$(distdir)/$$file"; then \ find "$(distdir)/$$file" -type d ! -perm -700 -exec chmod u+rwx {} \;; \ fi; \ if test -d $(srcdir)/$$file && test $$d != $(srcdir); then \ cp -fpR $(srcdir)/$$file "$(distdir)$$dir" || exit 1; \ find "$(distdir)/$$file" -type d ! -perm -700 -exec chmod u+rwx {} \;; \ fi; \ cp -fpR $$d/$$file "$(distdir)$$dir" || exit 1; \ else \ test -f "$(distdir)/$$file" \ || cp -p $$d/$$file "$(distdir)/$$file" \ || exit 1; \ fi; \ done check-am: all-am check: $(BUILT_SOURCES) $(MAKE) $(AM_MAKEFLAGS) check-am all-am: Makefile $(LTLIBRARIES) installdirs: install: $(BUILT_SOURCES) $(MAKE) $(AM_MAKEFLAGS) install-am install-exec: install-exec-am install-data: install-data-am uninstall: uninstall-am install-am: all-am @$(MAKE) $(AM_MAKEFLAGS) install-exec-am install-data-am installcheck: installcheck-am install-strip: if test -z '$(STRIP)'; then \ $(MAKE) $(AM_MAKEFLAGS) INSTALL_PROGRAM="$(INSTALL_STRIP_PROGRAM)" \ install_sh_PROGRAM="$(INSTALL_STRIP_PROGRAM)" INSTALL_STRIP_FLAG=-s \ install; \ else \ $(MAKE) $(AM_MAKEFLAGS) INSTALL_PROGRAM="$(INSTALL_STRIP_PROGRAM)" \ install_sh_PROGRAM="$(INSTALL_STRIP_PROGRAM)" INSTALL_STRIP_FLAG=-s \ "INSTALL_PROGRAM_ENV=STRIPPROG='$(STRIP)'" install; \ fi mostlyclean-generic: clean-generic: distclean-generic: -test -z "$(CONFIG_CLEAN_FILES)" || rm -f $(CONFIG_CLEAN_FILES) -test . = "$(srcdir)" || test -z "$(CONFIG_CLEAN_VPATH_FILES)" || rm -f $(CONFIG_CLEAN_VPATH_FILES) maintainer-clean-generic: @echo "This command is intended for maintainers to use" @echo "it deletes files that may require special tools to rebuild." -test -z "$(BUILT_SOURCES)" || rm -f $(BUILT_SOURCES) clean: clean-am clean-am: clean-generic clean-libtool clean-noinstLTLIBRARIES \ mostlyclean-am distclean: distclean-am -rm -rf ./$(DEPDIR) -rm -f Makefile distclean-am: clean-am distclean-compile distclean-generic \ distclean-tags dvi: dvi-am dvi-am: html: html-am html-am: info: info-am info-am: install-data-am: install-dvi: install-dvi-am install-dvi-am: install-exec-am: install-html: install-html-am install-html-am: install-info: install-info-am install-info-am: install-man: install-pdf: install-pdf-am install-pdf-am: install-ps: install-ps-am install-ps-am: installcheck-am: maintainer-clean: maintainer-clean-am -rm -rf ./$(DEPDIR) -rm -f Makefile maintainer-clean-am: distclean-am maintainer-clean-generic \ maintainer-clean-local mostlyclean: mostlyclean-am mostlyclean-am: mostlyclean-compile mostlyclean-generic \ mostlyclean-libtool pdf: pdf-am pdf-am: ps: ps-am ps-am: uninstall-am: .MAKE: all check install install-am install-strip .PHONY: CTAGS GTAGS all all-am check check-am clean clean-generic \ clean-libtool clean-noinstLTLIBRARIES ctags distclean \ distclean-compile distclean-generic distclean-libtool \ distclean-tags distdir dvi dvi-am html html-am info info-am \ install install-am install-data install-data-am install-dvi \ install-dvi-am install-exec install-exec-am install-html \ install-html-am install-info install-info-am install-man \ install-pdf install-pdf-am install-ps install-ps-am \ install-strip installcheck installcheck-am installdirs \ maintainer-clean maintainer-clean-generic \ maintainer-clean-local mostlyclean mostlyclean-compile \ mostlyclean-generic mostlyclean-libtool pdf pdf-am ps ps-am \ tags uninstall uninstall-am # rule to build codlist $(CODLIST): Makefile ( \ echo "#include \"ifftw.h\""; \ echo $(INCLUDE_SIMD_HEADER); \ echo; \ for i in $(ALL_CODELETS) NIL; do \ if test "$$i" != NIL; then \ j=`basename $$i | sed -e 's/[.][cS]$$//g'`; \ echo "extern void $(XRENAME)($(CODELET_NAME)$$j)(planner *);"; \ fi \ done; \ echo; \ echo; \ echo "extern const solvtab $(SOLVTAB_NAME);"; \ echo "const solvtab $(SOLVTAB_NAME) = {"; \ for i in $(ALL_CODELETS) NIL; do \ if test "$$i" != NIL; then \ j=`basename $$i | sed -e 's/[.][cS]$$//g'`; \ echo " SOLVTAB($(XRENAME)($(CODELET_NAME)$$j)),"; \ fi \ done; \ echo " SOLVTAB_END"; \ echo "};"; \ ) >$@ # only delete codlist.c in maintainer-mode, since it is included in the dist # FIXME: is there a way to delete in 'make clean' only when builddir != srcdir? maintainer-clean-local: rm -f $(CODLIST) # cancel the hideous builtin rules that cause an infinite loop @MAINTAINER_MODE_TRUE@%: %.o @MAINTAINER_MODE_TRUE@%: %.s @MAINTAINER_MODE_TRUE@%: %.c @MAINTAINER_MODE_TRUE@%: %.S @MAINTAINER_MODE_TRUE@r2cf_%.c: $(CODELET_DEPS) $(GEN_R2CF) @MAINTAINER_MODE_TRUE@ ($(PRELUDE_COMMANDS_RDFT); $(TWOVERS) $(GEN_R2CF) $(FLAGS_R2CF) -n $* -name r2cf_$* -include "r2cf.h") | $(ADD_DATE) | $(INDENT) >$@ @MAINTAINER_MODE_TRUE@hf_%.c: $(CODELET_DEPS) $(GEN_HC2HC) @MAINTAINER_MODE_TRUE@ ($(PRELUDE_COMMANDS_RDFT); $(TWOVERS) $(GEN_HC2HC) $(FLAGS_HF) -n $* -dit -name hf_$* -include "hf.h") | $(ADD_DATE) | $(INDENT) >$@ @MAINTAINER_MODE_TRUE@hf2_%.c: $(CODELET_DEPS) $(GEN_HC2HC) @MAINTAINER_MODE_TRUE@ ($(PRELUDE_COMMANDS_RDFT); $(TWOVERS) $(GEN_HC2HC) $(FLAGS_HF2) -n $* -dit -name hf2_$* -include "hf.h") | $(ADD_DATE) | $(INDENT) >$@ @MAINTAINER_MODE_TRUE@r2cfII_%.c: $(CODELET_DEPS) $(GEN_R2CF) @MAINTAINER_MODE_TRUE@ ($(PRELUDE_COMMANDS_RDFT); $(TWOVERS) $(GEN_R2CF) $(FLAGS_R2CF) -n $* -name r2cfII_$* -dft-II -include "r2cfII.h") | $(ADD_DATE) | $(INDENT) >$@ @MAINTAINER_MODE_TRUE@hc2cf_%.c: $(CODELET_DEPS) $(GEN_HC2C) @MAINTAINER_MODE_TRUE@ ($(PRELUDE_COMMANDS_RDFT); $(TWOVERS) $(GEN_HC2C) $(FLAGS_HC2CF) -n $* -dit -name hc2cf_$* -include "hc2cf.h") | $(ADD_DATE) | $(INDENT) >$@ @MAINTAINER_MODE_TRUE@hc2cf2_%.c: $(CODELET_DEPS) $(GEN_HC2C) @MAINTAINER_MODE_TRUE@ ($(PRELUDE_COMMANDS_RDFT); $(TWOVERS) $(GEN_HC2C) $(FLAGS_HC2CF2) -n $* -dit -name hc2cf2_$* -include "hc2cf.h") | $(ADD_DATE) | $(INDENT) >$@ @MAINTAINER_MODE_TRUE@hc2cfdft_%.c: $(CODELET_DEPS) $(GEN_HC2CDFT) @MAINTAINER_MODE_TRUE@ ($(PRELUDE_COMMANDS_RDFT); $(TWOVERS) $(GEN_HC2CDFT) $(FLAGS_HC2CF) -n $* -dit -name hc2cfdft_$* -include "hc2cf.h") | $(ADD_DATE) | $(INDENT) >$@ @MAINTAINER_MODE_TRUE@hc2cfdft2_%.c: $(CODELET_DEPS) $(GEN_HC2CDFT) @MAINTAINER_MODE_TRUE@ ($(PRELUDE_COMMANDS_RDFT); $(TWOVERS) $(GEN_HC2CDFT) $(FLAGS_HC2CF2) -n $* -dit -name hc2cfdft2_$* -include "hc2cf.h") | $(ADD_DATE) | $(INDENT) >$@ # Tell versions [3.59,3.63) of GNU make to not export all variables. # Otherwise a system limit (for SysV at least) may be exceeded. .NOEXPORT: fftw-3.3.3/rdft/scalar/r2cf/r2cf_2.c0000644000175000017500000000561712054410621013674 00000000000000/* * Copyright (c) 2003, 2007-11 Matteo Frigo * Copyright (c) 2003, 2007-11 Massachusetts Institute of Technology * * This program is free software; you can redistribute it and/or modify * it under the terms of the GNU General Public License as published by * the Free Software Foundation; either version 2 of the License, or * (at your option) any later version. * * This program is distributed in the hope that it will be useful, * but WITHOUT ANY WARRANTY; without even the implied warranty of * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the * GNU General Public License for more details. * * You should have received a copy of the GNU General Public License * along with this program; if not, write to the Free Software * Foundation, Inc., 51 Franklin Street, Fifth Floor, Boston, MA 02110-1301 USA * */ /* This file was automatically generated --- DO NOT EDIT */ /* Generated on Sun Nov 25 07:39:45 EST 2012 */ #include "codelet-rdft.h" #ifdef HAVE_FMA /* Generated by: ../../../genfft/gen_r2cf.native -fma -reorder-insns -schedule-for-pipeline -compact -variables 4 -pipeline-latency 4 -n 2 -name r2cf_2 -include r2cf.h */ /* * This function contains 2 FP additions, 0 FP multiplications, * (or, 2 additions, 0 multiplications, 0 fused multiply/add), * 3 stack variables, 0 constants, and 4 memory accesses */ #include "r2cf.h" static void r2cf_2(R *R0, R *R1, R *Cr, R *Ci, stride rs, stride csr, stride csi, INT v, INT ivs, INT ovs) { { INT i; for (i = v; i > 0; i = i - 1, R0 = R0 + ivs, R1 = R1 + ivs, Cr = Cr + ovs, Ci = Ci + ovs, MAKE_VOLATILE_STRIDE(8, rs), MAKE_VOLATILE_STRIDE(8, csr), MAKE_VOLATILE_STRIDE(8, csi)) { E T1, T2; T1 = R0[0]; T2 = R1[0]; Cr[0] = T1 + T2; Cr[WS(csr, 1)] = T1 - T2; } } } static const kr2c_desc desc = { 2, "r2cf_2", {2, 0, 0, 0}, &GENUS }; void X(codelet_r2cf_2) (planner *p) { X(kr2c_register) (p, r2cf_2, &desc); } #else /* HAVE_FMA */ /* Generated by: ../../../genfft/gen_r2cf.native -compact -variables 4 -pipeline-latency 4 -n 2 -name r2cf_2 -include r2cf.h */ /* * This function contains 2 FP additions, 0 FP multiplications, * (or, 2 additions, 0 multiplications, 0 fused multiply/add), * 3 stack variables, 0 constants, and 4 memory accesses */ #include "r2cf.h" static void r2cf_2(R *R0, R *R1, R *Cr, R *Ci, stride rs, stride csr, stride csi, INT v, INT ivs, INT ovs) { { INT i; for (i = v; i > 0; i = i - 1, R0 = R0 + ivs, R1 = R1 + ivs, Cr = Cr + ovs, Ci = Ci + ovs, MAKE_VOLATILE_STRIDE(8, rs), MAKE_VOLATILE_STRIDE(8, csr), MAKE_VOLATILE_STRIDE(8, csi)) { E T1, T2; T1 = R0[0]; T2 = R1[0]; Cr[WS(csr, 1)] = T1 - T2; Cr[0] = T1 + T2; } } } static const kr2c_desc desc = { 2, "r2cf_2", {2, 0, 0, 0}, &GENUS }; void X(codelet_r2cf_2) (planner *p) { X(kr2c_register) (p, r2cf_2, &desc); } #endif /* HAVE_FMA */ fftw-3.3.3/rdft/scalar/r2cf/r2cf_3.c0000644000175000017500000000666112054410621013675 00000000000000/* * Copyright (c) 2003, 2007-11 Matteo Frigo * Copyright (c) 2003, 2007-11 Massachusetts Institute of Technology * * This program is free software; you can redistribute it and/or modify * it under the terms of the GNU General Public License as published by * the Free Software Foundation; either version 2 of the License, or * (at your option) any later version. * * This program is distributed in the hope that it will be useful, * but WITHOUT ANY WARRANTY; without even the implied warranty of * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the * GNU General Public License for more details. * * You should have received a copy of the GNU General Public License * along with this program; if not, write to the Free Software * Foundation, Inc., 51 Franklin Street, Fifth Floor, Boston, MA 02110-1301 USA * */ /* This file was automatically generated --- DO NOT EDIT */ /* Generated on Sun Nov 25 07:39:45 EST 2012 */ #include "codelet-rdft.h" #ifdef HAVE_FMA /* Generated by: ../../../genfft/gen_r2cf.native -fma -reorder-insns -schedule-for-pipeline -compact -variables 4 -pipeline-latency 4 -n 3 -name r2cf_3 -include r2cf.h */ /* * This function contains 4 FP additions, 2 FP multiplications, * (or, 3 additions, 1 multiplications, 1 fused multiply/add), * 7 stack variables, 2 constants, and 6 memory accesses */ #include "r2cf.h" static void r2cf_3(R *R0, R *R1, R *Cr, R *Ci, stride rs, stride csr, stride csi, INT v, INT ivs, INT ovs) { DK(KP500000000, +0.500000000000000000000000000000000000000000000); DK(KP866025403, +0.866025403784438646763723170752936183471402627); { INT i; for (i = v; i > 0; i = i - 1, R0 = R0 + ivs, R1 = R1 + ivs, Cr = Cr + ovs, Ci = Ci + ovs, MAKE_VOLATILE_STRIDE(12, rs), MAKE_VOLATILE_STRIDE(12, csr), MAKE_VOLATILE_STRIDE(12, csi)) { E T1, T2, T3, T4; T1 = R0[0]; T2 = R1[0]; T3 = R0[WS(rs, 1)]; Ci[WS(csi, 1)] = KP866025403 * (T3 - T2); T4 = T2 + T3; Cr[0] = T1 + T4; Cr[WS(csr, 1)] = FNMS(KP500000000, T4, T1); } } } static const kr2c_desc desc = { 3, "r2cf_3", {3, 1, 1, 0}, &GENUS }; void X(codelet_r2cf_3) (planner *p) { X(kr2c_register) (p, r2cf_3, &desc); } #else /* HAVE_FMA */ /* Generated by: ../../../genfft/gen_r2cf.native -compact -variables 4 -pipeline-latency 4 -n 3 -name r2cf_3 -include r2cf.h */ /* * This function contains 4 FP additions, 2 FP multiplications, * (or, 3 additions, 1 multiplications, 1 fused multiply/add), * 7 stack variables, 2 constants, and 6 memory accesses */ #include "r2cf.h" static void r2cf_3(R *R0, R *R1, R *Cr, R *Ci, stride rs, stride csr, stride csi, INT v, INT ivs, INT ovs) { DK(KP866025403, +0.866025403784438646763723170752936183471402627); DK(KP500000000, +0.500000000000000000000000000000000000000000000); { INT i; for (i = v; i > 0; i = i - 1, R0 = R0 + ivs, R1 = R1 + ivs, Cr = Cr + ovs, Ci = Ci + ovs, MAKE_VOLATILE_STRIDE(12, rs), MAKE_VOLATILE_STRIDE(12, csr), MAKE_VOLATILE_STRIDE(12, csi)) { E T1, T2, T3, T4; T1 = R0[0]; T2 = R1[0]; T3 = R0[WS(rs, 1)]; T4 = T2 + T3; Cr[WS(csr, 1)] = FNMS(KP500000000, T4, T1); Ci[WS(csi, 1)] = KP866025403 * (T3 - T2); Cr[0] = T1 + T4; } } } static const kr2c_desc desc = { 3, "r2cf_3", {3, 1, 1, 0}, &GENUS }; void X(codelet_r2cf_3) (planner *p) { X(kr2c_register) (p, r2cf_3, &desc); } #endif /* HAVE_FMA */ fftw-3.3.3/rdft/scalar/r2cf/r2cf_4.c0000644000175000017500000000640512054410621013672 00000000000000/* * Copyright (c) 2003, 2007-11 Matteo Frigo * Copyright (c) 2003, 2007-11 Massachusetts Institute of Technology * * This program is free software; you can redistribute it and/or modify * it under the terms of the GNU General Public License as published by * the Free Software Foundation; either version 2 of the License, or * (at your option) any later version. * * This program is distributed in the hope that it will be useful, * but WITHOUT ANY WARRANTY; without even the implied warranty of * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the * GNU General Public License for more details. * * You should have received a copy of the GNU General Public License * along with this program; if not, write to the Free Software * Foundation, Inc., 51 Franklin Street, Fifth Floor, Boston, MA 02110-1301 USA * */ /* This file was automatically generated --- DO NOT EDIT */ /* Generated on Sun Nov 25 07:39:45 EST 2012 */ #include "codelet-rdft.h" #ifdef HAVE_FMA /* Generated by: ../../../genfft/gen_r2cf.native -fma -reorder-insns -schedule-for-pipeline -compact -variables 4 -pipeline-latency 4 -n 4 -name r2cf_4 -include r2cf.h */ /* * This function contains 6 FP additions, 0 FP multiplications, * (or, 6 additions, 0 multiplications, 0 fused multiply/add), * 7 stack variables, 0 constants, and 8 memory accesses */ #include "r2cf.h" static void r2cf_4(R *R0, R *R1, R *Cr, R *Ci, stride rs, stride csr, stride csi, INT v, INT ivs, INT ovs) { { INT i; for (i = v; i > 0; i = i - 1, R0 = R0 + ivs, R1 = R1 + ivs, Cr = Cr + ovs, Ci = Ci + ovs, MAKE_VOLATILE_STRIDE(16, rs), MAKE_VOLATILE_STRIDE(16, csr), MAKE_VOLATILE_STRIDE(16, csi)) { E T1, T2, T4, T5, T3, T6; T1 = R0[0]; T2 = R0[WS(rs, 1)]; T4 = R1[0]; T5 = R1[WS(rs, 1)]; Cr[WS(csr, 1)] = T1 - T2; T3 = T1 + T2; Ci[WS(csi, 1)] = T5 - T4; T6 = T4 + T5; Cr[0] = T3 + T6; Cr[WS(csr, 2)] = T3 - T6; } } } static const kr2c_desc desc = { 4, "r2cf_4", {6, 0, 0, 0}, &GENUS }; void X(codelet_r2cf_4) (planner *p) { X(kr2c_register) (p, r2cf_4, &desc); } #else /* HAVE_FMA */ /* Generated by: ../../../genfft/gen_r2cf.native -compact -variables 4 -pipeline-latency 4 -n 4 -name r2cf_4 -include r2cf.h */ /* * This function contains 6 FP additions, 0 FP multiplications, * (or, 6 additions, 0 multiplications, 0 fused multiply/add), * 7 stack variables, 0 constants, and 8 memory accesses */ #include "r2cf.h" static void r2cf_4(R *R0, R *R1, R *Cr, R *Ci, stride rs, stride csr, stride csi, INT v, INT ivs, INT ovs) { { INT i; for (i = v; i > 0; i = i - 1, R0 = R0 + ivs, R1 = R1 + ivs, Cr = Cr + ovs, Ci = Ci + ovs, MAKE_VOLATILE_STRIDE(16, rs), MAKE_VOLATILE_STRIDE(16, csr), MAKE_VOLATILE_STRIDE(16, csi)) { E T1, T2, T3, T4, T5, T6; T1 = R0[0]; T2 = R0[WS(rs, 1)]; T3 = T1 + T2; T4 = R1[0]; T5 = R1[WS(rs, 1)]; T6 = T4 + T5; Cr[WS(csr, 1)] = T1 - T2; Ci[WS(csi, 1)] = T5 - T4; Cr[WS(csr, 2)] = T3 - T6; Cr[0] = T3 + T6; } } } static const kr2c_desc desc = { 4, "r2cf_4", {6, 0, 0, 0}, &GENUS }; void X(codelet_r2cf_4) (planner *p) { X(kr2c_register) (p, r2cf_4, &desc); } #endif /* HAVE_FMA */ fftw-3.3.3/rdft/scalar/r2cf/r2cf_5.c0000644000175000017500000001067212054410621013674 00000000000000/* * Copyright (c) 2003, 2007-11 Matteo Frigo * Copyright (c) 2003, 2007-11 Massachusetts Institute of Technology * * This program is free software; you can redistribute it and/or modify * it under the terms of the GNU General Public License as published by * the Free Software Foundation; either version 2 of the License, or * (at your option) any later version. * * This program is distributed in the hope that it will be useful, * but WITHOUT ANY WARRANTY; without even the implied warranty of * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the * GNU General Public License for more details. * * You should have received a copy of the GNU General Public License * along with this program; if not, write to the Free Software * Foundation, Inc., 51 Franklin Street, Fifth Floor, Boston, MA 02110-1301 USA * */ /* This file was automatically generated --- DO NOT EDIT */ /* Generated on Sun Nov 25 07:39:45 EST 2012 */ #include "codelet-rdft.h" #ifdef HAVE_FMA /* Generated by: ../../../genfft/gen_r2cf.native -fma -reorder-insns -schedule-for-pipeline -compact -variables 4 -pipeline-latency 4 -n 5 -name r2cf_5 -include r2cf.h */ /* * This function contains 12 FP additions, 7 FP multiplications, * (or, 7 additions, 2 multiplications, 5 fused multiply/add), * 17 stack variables, 4 constants, and 10 memory accesses */ #include "r2cf.h" static void r2cf_5(R *R0, R *R1, R *Cr, R *Ci, stride rs, stride csr, stride csi, INT v, INT ivs, INT ovs) { DK(KP559016994, +0.559016994374947424102293417182819058860154590); DK(KP250000000, +0.250000000000000000000000000000000000000000000); DK(KP618033988, +0.618033988749894848204586834365638117720309180); DK(KP951056516, +0.951056516295153572116439333379382143405698634); { INT i; for (i = v; i > 0; i = i - 1, R0 = R0 + ivs, R1 = R1 + ivs, Cr = Cr + ovs, Ci = Ci + ovs, MAKE_VOLATILE_STRIDE(20, rs), MAKE_VOLATILE_STRIDE(20, csr), MAKE_VOLATILE_STRIDE(20, csi)) { E T7, T1, T2, T4, T5; T7 = R0[0]; T1 = R0[WS(rs, 2)]; T2 = R1[0]; T4 = R0[WS(rs, 1)]; T5 = R1[WS(rs, 1)]; { E T3, T8, T6, T9, Tc, Ta, Tb; T3 = T1 - T2; T8 = T2 + T1; T6 = T4 - T5; T9 = T4 + T5; Ci[WS(csi, 2)] = KP951056516 * (FMA(KP618033988, T3, T6)); Ci[WS(csi, 1)] = KP951056516 * (FNMS(KP618033988, T6, T3)); Tc = T8 - T9; Ta = T8 + T9; Tb = FNMS(KP250000000, Ta, T7); Cr[0] = T7 + Ta; Cr[WS(csr, 2)] = FNMS(KP559016994, Tc, Tb); Cr[WS(csr, 1)] = FMA(KP559016994, Tc, Tb); } } } } static const kr2c_desc desc = { 5, "r2cf_5", {7, 2, 5, 0}, &GENUS }; void X(codelet_r2cf_5) (planner *p) { X(kr2c_register) (p, r2cf_5, &desc); } #else /* HAVE_FMA */ /* Generated by: ../../../genfft/gen_r2cf.native -compact -variables 4 -pipeline-latency 4 -n 5 -name r2cf_5 -include r2cf.h */ /* * This function contains 12 FP additions, 6 FP multiplications, * (or, 9 additions, 3 multiplications, 3 fused multiply/add), * 17 stack variables, 4 constants, and 10 memory accesses */ #include "r2cf.h" static void r2cf_5(R *R0, R *R1, R *Cr, R *Ci, stride rs, stride csr, stride csi, INT v, INT ivs, INT ovs) { DK(KP250000000, +0.250000000000000000000000000000000000000000000); DK(KP559016994, +0.559016994374947424102293417182819058860154590); DK(KP587785252, +0.587785252292473129168705954639072768597652438); DK(KP951056516, +0.951056516295153572116439333379382143405698634); { INT i; for (i = v; i > 0; i = i - 1, R0 = R0 + ivs, R1 = R1 + ivs, Cr = Cr + ovs, Ci = Ci + ovs, MAKE_VOLATILE_STRIDE(20, rs), MAKE_VOLATILE_STRIDE(20, csr), MAKE_VOLATILE_STRIDE(20, csi)) { E Ta, T7, T8, T3, Tb, T6, T9, Tc; Ta = R0[0]; { E T1, T2, T4, T5; T1 = R0[WS(rs, 2)]; T2 = R1[0]; T7 = T2 + T1; T4 = R0[WS(rs, 1)]; T5 = R1[WS(rs, 1)]; T8 = T4 + T5; T3 = T1 - T2; Tb = T7 + T8; T6 = T4 - T5; } Ci[WS(csi, 1)] = FNMS(KP587785252, T6, KP951056516 * T3); Cr[0] = Ta + Tb; Ci[WS(csi, 2)] = FMA(KP587785252, T3, KP951056516 * T6); T9 = KP559016994 * (T7 - T8); Tc = FNMS(KP250000000, Tb, Ta); Cr[WS(csr, 1)] = T9 + Tc; Cr[WS(csr, 2)] = Tc - T9; } } } static const kr2c_desc desc = { 5, "r2cf_5", {9, 3, 3, 0}, &GENUS }; void X(codelet_r2cf_5) (planner *p) { X(kr2c_register) (p, r2cf_5, &desc); } #endif /* HAVE_FMA */ fftw-3.3.3/rdft/scalar/r2cf/r2cf_6.c0000644000175000017500000001042312054410621013667 00000000000000/* * Copyright (c) 2003, 2007-11 Matteo Frigo * Copyright (c) 2003, 2007-11 Massachusetts Institute of Technology * * This program is free software; you can redistribute it and/or modify * it under the terms of the GNU General Public License as published by * the Free Software Foundation; either version 2 of the License, or * (at your option) any later version. * * This program is distributed in the hope that it will be useful, * but WITHOUT ANY WARRANTY; without even the implied warranty of * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the * GNU General Public License for more details. * * You should have received a copy of the GNU General Public License * along with this program; if not, write to the Free Software * Foundation, Inc., 51 Franklin Street, Fifth Floor, Boston, MA 02110-1301 USA * */ /* This file was automatically generated --- DO NOT EDIT */ /* Generated on Sun Nov 25 07:39:45 EST 2012 */ #include "codelet-rdft.h" #ifdef HAVE_FMA /* Generated by: ../../../genfft/gen_r2cf.native -fma -reorder-insns -schedule-for-pipeline -compact -variables 4 -pipeline-latency 4 -n 6 -name r2cf_6 -include r2cf.h */ /* * This function contains 14 FP additions, 4 FP multiplications, * (or, 12 additions, 2 multiplications, 2 fused multiply/add), * 13 stack variables, 2 constants, and 12 memory accesses */ #include "r2cf.h" static void r2cf_6(R *R0, R *R1, R *Cr, R *Ci, stride rs, stride csr, stride csi, INT v, INT ivs, INT ovs) { DK(KP500000000, +0.500000000000000000000000000000000000000000000); DK(KP866025403, +0.866025403784438646763723170752936183471402627); { INT i; for (i = v; i > 0; i = i - 1, R0 = R0 + ivs, R1 = R1 + ivs, Cr = Cr + ovs, Ci = Ci + ovs, MAKE_VOLATILE_STRIDE(24, rs), MAKE_VOLATILE_STRIDE(24, csr), MAKE_VOLATILE_STRIDE(24, csi)) { E T4, Td, T3, Tc, T9, T5; { E T1, T2, T7, T8; T1 = R0[0]; T2 = R1[WS(rs, 1)]; T7 = R0[WS(rs, 2)]; T8 = R1[0]; T4 = R0[WS(rs, 1)]; Td = T1 + T2; T3 = T1 - T2; Tc = T7 + T8; T9 = T7 - T8; T5 = R1[WS(rs, 2)]; } { E T6, Tb, Te, Ta; T6 = T4 - T5; Tb = T4 + T5; Te = Tb + Tc; Ci[WS(csi, 2)] = KP866025403 * (Tb - Tc); Ta = T6 + T9; Ci[WS(csi, 1)] = KP866025403 * (T9 - T6); Cr[0] = Td + Te; Cr[WS(csr, 2)] = FNMS(KP500000000, Te, Td); Cr[WS(csr, 3)] = T3 + Ta; Cr[WS(csr, 1)] = FNMS(KP500000000, Ta, T3); } } } } static const kr2c_desc desc = { 6, "r2cf_6", {12, 2, 2, 0}, &GENUS }; void X(codelet_r2cf_6) (planner *p) { X(kr2c_register) (p, r2cf_6, &desc); } #else /* HAVE_FMA */ /* Generated by: ../../../genfft/gen_r2cf.native -compact -variables 4 -pipeline-latency 4 -n 6 -name r2cf_6 -include r2cf.h */ /* * This function contains 14 FP additions, 4 FP multiplications, * (or, 12 additions, 2 multiplications, 2 fused multiply/add), * 17 stack variables, 2 constants, and 12 memory accesses */ #include "r2cf.h" static void r2cf_6(R *R0, R *R1, R *Cr, R *Ci, stride rs, stride csr, stride csi, INT v, INT ivs, INT ovs) { DK(KP500000000, +0.500000000000000000000000000000000000000000000); DK(KP866025403, +0.866025403784438646763723170752936183471402627); { INT i; for (i = v; i > 0; i = i - 1, R0 = R0 + ivs, R1 = R1 + ivs, Cr = Cr + ovs, Ci = Ci + ovs, MAKE_VOLATILE_STRIDE(24, rs), MAKE_VOLATILE_STRIDE(24, csr), MAKE_VOLATILE_STRIDE(24, csi)) { E T3, Td, T9, Tc, T6, Tb, T1, T2, Ta, Te; T1 = R0[0]; T2 = R1[WS(rs, 1)]; T3 = T1 - T2; Td = T1 + T2; { E T7, T8, T4, T5; T7 = R0[WS(rs, 2)]; T8 = R1[0]; T9 = T7 - T8; Tc = T7 + T8; T4 = R0[WS(rs, 1)]; T5 = R1[WS(rs, 2)]; T6 = T4 - T5; Tb = T4 + T5; } Ci[WS(csi, 1)] = KP866025403 * (T9 - T6); Ta = T6 + T9; Cr[WS(csr, 1)] = FNMS(KP500000000, Ta, T3); Cr[WS(csr, 3)] = T3 + Ta; Ci[WS(csi, 2)] = KP866025403 * (Tb - Tc); Te = Tb + Tc; Cr[WS(csr, 2)] = FNMS(KP500000000, Te, Td); Cr[0] = Td + Te; } } } static const kr2c_desc desc = { 6, "r2cf_6", {12, 2, 2, 0}, &GENUS }; void X(codelet_r2cf_6) (planner *p) { X(kr2c_register) (p, r2cf_6, &desc); } #endif /* HAVE_FMA */ fftw-3.3.3/rdft/scalar/r2cf/r2cf_7.c0000644000175000017500000001311212054410622013667 00000000000000/* * Copyright (c) 2003, 2007-11 Matteo Frigo * Copyright (c) 2003, 2007-11 Massachusetts Institute of Technology * * This program is free software; you can redistribute it and/or modify * it under the terms of the GNU General Public License as published by * the Free Software Foundation; either version 2 of the License, or * (at your option) any later version. * * This program is distributed in the hope that it will be useful, * but WITHOUT ANY WARRANTY; without even the implied warranty of * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the * GNU General Public License for more details. * * You should have received a copy of the GNU General Public License * along with this program; if not, write to the Free Software * Foundation, Inc., 51 Franklin Street, Fifth Floor, Boston, MA 02110-1301 USA * */ /* This file was automatically generated --- DO NOT EDIT */ /* Generated on Sun Nov 25 07:39:45 EST 2012 */ #include "codelet-rdft.h" #ifdef HAVE_FMA /* Generated by: ../../../genfft/gen_r2cf.native -fma -reorder-insns -schedule-for-pipeline -compact -variables 4 -pipeline-latency 4 -n 7 -name r2cf_7 -include r2cf.h */ /* * This function contains 24 FP additions, 18 FP multiplications, * (or, 9 additions, 3 multiplications, 15 fused multiply/add), * 25 stack variables, 6 constants, and 14 memory accesses */ #include "r2cf.h" static void r2cf_7(R *R0, R *R1, R *Cr, R *Ci, stride rs, stride csr, stride csi, INT v, INT ivs, INT ovs) { DK(KP900968867, +0.900968867902419126236102319507445051165919162); DK(KP801937735, +0.801937735804838252472204639014890102331838324); DK(KP974927912, +0.974927912181823607018131682993931217232785801); DK(KP692021471, +0.692021471630095869627814897002069140197260599); DK(KP554958132, +0.554958132087371191422194871006410481067288862); DK(KP356895867, +0.356895867892209443894399510021300583399127187); { INT i; for (i = v; i > 0; i = i - 1, R0 = R0 + ivs, R1 = R1 + ivs, Cr = Cr + ovs, Ci = Ci + ovs, MAKE_VOLATILE_STRIDE(28, rs), MAKE_VOLATILE_STRIDE(28, csr), MAKE_VOLATILE_STRIDE(28, csi)) { E T1, Tg, Tc; { E Th, T4, Ti, Ta, Tj, T7, Td, T5, T6, Tl, Tk; T1 = R0[0]; { E T2, T3, T8, T9; T2 = R1[0]; T3 = R0[WS(rs, 3)]; T8 = R1[WS(rs, 1)]; T9 = R0[WS(rs, 2)]; T5 = R0[WS(rs, 1)]; Th = T3 - T2; T4 = T2 + T3; T6 = R1[WS(rs, 2)]; Ti = T9 - T8; Ta = T8 + T9; } Tj = T6 - T5; T7 = T5 + T6; Td = FNMS(KP356895867, T4, Ta); Tl = FMA(KP554958132, Ti, Th); Tk = FMA(KP554958132, Tj, Ti); { E Tm, Tf, Tb, Te; Tm = FNMS(KP554958132, Th, Tj); Cr[0] = T1 + T4 + T7 + Ta; Tf = FNMS(KP356895867, T7, T4); Tb = FNMS(KP356895867, Ta, T7); Te = FNMS(KP692021471, Td, T7); Ci[WS(csi, 2)] = KP974927912 * (FNMS(KP801937735, Tk, Th)); Ci[WS(csi, 3)] = KP974927912 * (FNMS(KP801937735, Tm, Ti)); Tg = FNMS(KP692021471, Tf, Ta); Tc = FNMS(KP692021471, Tb, T4); Cr[WS(csr, 2)] = FNMS(KP900968867, Te, T1); Ci[WS(csi, 1)] = KP974927912 * (FMA(KP801937735, Tl, Tj)); } } Cr[WS(csr, 1)] = FNMS(KP900968867, Tg, T1); Cr[WS(csr, 3)] = FNMS(KP900968867, Tc, T1); } } } static const kr2c_desc desc = { 7, "r2cf_7", {9, 3, 15, 0}, &GENUS }; void X(codelet_r2cf_7) (planner *p) { X(kr2c_register) (p, r2cf_7, &desc); } #else /* HAVE_FMA */ /* Generated by: ../../../genfft/gen_r2cf.native -compact -variables 4 -pipeline-latency 4 -n 7 -name r2cf_7 -include r2cf.h */ /* * This function contains 24 FP additions, 18 FP multiplications, * (or, 12 additions, 6 multiplications, 12 fused multiply/add), * 20 stack variables, 6 constants, and 14 memory accesses */ #include "r2cf.h" static void r2cf_7(R *R0, R *R1, R *Cr, R *Ci, stride rs, stride csr, stride csi, INT v, INT ivs, INT ovs) { DK(KP222520933, +0.222520933956314404288902564496794759466355569); DK(KP900968867, +0.900968867902419126236102319507445051165919162); DK(KP623489801, +0.623489801858733530525004884004239810632274731); DK(KP433883739, +0.433883739117558120475768332848358754609990728); DK(KP781831482, +0.781831482468029808708444526674057750232334519); DK(KP974927912, +0.974927912181823607018131682993931217232785801); { INT i; for (i = v; i > 0; i = i - 1, R0 = R0 + ivs, R1 = R1 + ivs, Cr = Cr + ovs, Ci = Ci + ovs, MAKE_VOLATILE_STRIDE(28, rs), MAKE_VOLATILE_STRIDE(28, csr), MAKE_VOLATILE_STRIDE(28, csi)) { E T1, Ta, Tb, T4, Td, T7, Tc, T8, T9; T1 = R0[0]; T8 = R1[0]; T9 = R0[WS(rs, 3)]; Ta = T8 + T9; Tb = T9 - T8; { E T2, T3, T5, T6; T2 = R0[WS(rs, 1)]; T3 = R1[WS(rs, 2)]; T4 = T2 + T3; Td = T3 - T2; T5 = R1[WS(rs, 1)]; T6 = R0[WS(rs, 2)]; T7 = T5 + T6; Tc = T6 - T5; } Ci[WS(csi, 2)] = FNMS(KP781831482, Tc, KP974927912 * Tb) - (KP433883739 * Td); Ci[WS(csi, 1)] = FMA(KP781831482, Tb, KP974927912 * Td) + (KP433883739 * Tc); Cr[WS(csr, 2)] = FMA(KP623489801, T7, T1) + FNMA(KP900968867, T4, KP222520933 * Ta); Ci[WS(csi, 3)] = FMA(KP433883739, Tb, KP974927912 * Tc) - (KP781831482 * Td); Cr[WS(csr, 3)] = FMA(KP623489801, T4, T1) + FNMA(KP222520933, T7, KP900968867 * Ta); Cr[WS(csr, 1)] = FMA(KP623489801, Ta, T1) + FNMA(KP900968867, T7, KP222520933 * T4); Cr[0] = T1 + Ta + T4 + T7; } } } static const kr2c_desc desc = { 7, "r2cf_7", {12, 6, 12, 0}, &GENUS }; void X(codelet_r2cf_7) (planner *p) { X(kr2c_register) (p, r2cf_7, &desc); } #endif /* HAVE_FMA */ fftw-3.3.3/rdft/scalar/r2cf/r2cf_8.c0000644000175000017500000001107212054410621013672 00000000000000/* * Copyright (c) 2003, 2007-11 Matteo Frigo * Copyright (c) 2003, 2007-11 Massachusetts Institute of Technology * * This program is free software; you can redistribute it and/or modify * it under the terms of the GNU General Public License as published by * the Free Software Foundation; either version 2 of the License, or * (at your option) any later version. * * This program is distributed in the hope that it will be useful, * but WITHOUT ANY WARRANTY; without even the implied warranty of * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the * GNU General Public License for more details. * * You should have received a copy of the GNU General Public License * along with this program; if not, write to the Free Software * Foundation, Inc., 51 Franklin Street, Fifth Floor, Boston, MA 02110-1301 USA * */ /* This file was automatically generated --- DO NOT EDIT */ /* Generated on Sun Nov 25 07:39:45 EST 2012 */ #include "codelet-rdft.h" #ifdef HAVE_FMA /* Generated by: ../../../genfft/gen_r2cf.native -fma -reorder-insns -schedule-for-pipeline -compact -variables 4 -pipeline-latency 4 -n 8 -name r2cf_8 -include r2cf.h */ /* * This function contains 20 FP additions, 4 FP multiplications, * (or, 16 additions, 0 multiplications, 4 fused multiply/add), * 18 stack variables, 1 constants, and 16 memory accesses */ #include "r2cf.h" static void r2cf_8(R *R0, R *R1, R *Cr, R *Ci, stride rs, stride csr, stride csi, INT v, INT ivs, INT ovs) { DK(KP707106781, +0.707106781186547524400844362104849039284835938); { INT i; for (i = v; i > 0; i = i - 1, R0 = R0 + ivs, R1 = R1 + ivs, Cr = Cr + ovs, Ci = Ci + ovs, MAKE_VOLATILE_STRIDE(32, rs), MAKE_VOLATILE_STRIDE(32, csr), MAKE_VOLATILE_STRIDE(32, csi)) { E T4, T7, T3, Tj, Td, T5, T8, T9; { E T1, T2, Tb, Tc; T1 = R0[0]; T2 = R0[WS(rs, 2)]; Tb = R1[WS(rs, 3)]; Tc = R1[WS(rs, 1)]; T4 = R0[WS(rs, 1)]; T7 = T1 - T2; T3 = T1 + T2; Tj = Tb + Tc; Td = Tb - Tc; T5 = R0[WS(rs, 3)]; T8 = R1[0]; T9 = R1[WS(rs, 2)]; } { E T6, Tf, Ta, Ti; T6 = T4 + T5; Tf = T4 - T5; Ta = T8 - T9; Ti = T8 + T9; { E Th, Tk, Te, Tg; Th = T3 + T6; Cr[WS(csr, 2)] = T3 - T6; Tk = Ti + Tj; Ci[WS(csi, 2)] = Tj - Ti; Te = Ta + Td; Tg = Td - Ta; Cr[0] = Th + Tk; Cr[WS(csr, 4)] = Th - Tk; Ci[WS(csi, 3)] = FMA(KP707106781, Tg, Tf); Ci[WS(csi, 1)] = FMS(KP707106781, Tg, Tf); Cr[WS(csr, 1)] = FMA(KP707106781, Te, T7); Cr[WS(csr, 3)] = FNMS(KP707106781, Te, T7); } } } } } static const kr2c_desc desc = { 8, "r2cf_8", {16, 0, 4, 0}, &GENUS }; void X(codelet_r2cf_8) (planner *p) { X(kr2c_register) (p, r2cf_8, &desc); } #else /* HAVE_FMA */ /* Generated by: ../../../genfft/gen_r2cf.native -compact -variables 4 -pipeline-latency 4 -n 8 -name r2cf_8 -include r2cf.h */ /* * This function contains 20 FP additions, 2 FP multiplications, * (or, 20 additions, 2 multiplications, 0 fused multiply/add), * 14 stack variables, 1 constants, and 16 memory accesses */ #include "r2cf.h" static void r2cf_8(R *R0, R *R1, R *Cr, R *Ci, stride rs, stride csr, stride csi, INT v, INT ivs, INT ovs) { DK(KP707106781, +0.707106781186547524400844362104849039284835938); { INT i; for (i = v; i > 0; i = i - 1, R0 = R0 + ivs, R1 = R1 + ivs, Cr = Cr + ovs, Ci = Ci + ovs, MAKE_VOLATILE_STRIDE(32, rs), MAKE_VOLATILE_STRIDE(32, csr), MAKE_VOLATILE_STRIDE(32, csi)) { E T3, T7, Td, Tj, T6, Tg, Ta, Ti; { E T1, T2, Tb, Tc; T1 = R0[0]; T2 = R0[WS(rs, 2)]; T3 = T1 + T2; T7 = T1 - T2; Tb = R1[WS(rs, 3)]; Tc = R1[WS(rs, 1)]; Td = Tb - Tc; Tj = Tb + Tc; } { E T4, T5, T8, T9; T4 = R0[WS(rs, 1)]; T5 = R0[WS(rs, 3)]; T6 = T4 + T5; Tg = T4 - T5; T8 = R1[0]; T9 = R1[WS(rs, 2)]; Ta = T8 - T9; Ti = T8 + T9; } Cr[WS(csr, 2)] = T3 - T6; Ci[WS(csi, 2)] = Tj - Ti; { E Te, Tf, Th, Tk; Te = KP707106781 * (Ta + Td); Cr[WS(csr, 3)] = T7 - Te; Cr[WS(csr, 1)] = T7 + Te; Tf = KP707106781 * (Td - Ta); Ci[WS(csi, 1)] = Tf - Tg; Ci[WS(csi, 3)] = Tg + Tf; Th = T3 + T6; Tk = Ti + Tj; Cr[WS(csr, 4)] = Th - Tk; Cr[0] = Th + Tk; } } } } static const kr2c_desc desc = { 8, "r2cf_8", {20, 2, 0, 0}, &GENUS }; void X(codelet_r2cf_8) (planner *p) { X(kr2c_register) (p, r2cf_8, &desc); } #endif /* HAVE_FMA */ fftw-3.3.3/rdft/scalar/r2cf/r2cf_9.c0000644000175000017500000002066212054410622013701 00000000000000/* * Copyright (c) 2003, 2007-11 Matteo Frigo * Copyright (c) 2003, 2007-11 Massachusetts Institute of Technology * * This program is free software; you can redistribute it and/or modify * it under the terms of the GNU General Public License as published by * the Free Software Foundation; either version 2 of the License, or * (at your option) any later version. * * This program is distributed in the hope that it will be useful, * but WITHOUT ANY WARRANTY; without even the implied warranty of * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the * GNU General Public License for more details. * * You should have received a copy of the GNU General Public License * along with this program; if not, write to the Free Software * Foundation, Inc., 51 Franklin Street, Fifth Floor, Boston, MA 02110-1301 USA * */ /* This file was automatically generated --- DO NOT EDIT */ /* Generated on Sun Nov 25 07:39:45 EST 2012 */ #include "codelet-rdft.h" #ifdef HAVE_FMA /* Generated by: ../../../genfft/gen_r2cf.native -fma -reorder-insns -schedule-for-pipeline -compact -variables 4 -pipeline-latency 4 -n 9 -name r2cf_9 -include r2cf.h */ /* * This function contains 38 FP additions, 30 FP multiplications, * (or, 12 additions, 4 multiplications, 26 fused multiply/add), * 57 stack variables, 18 constants, and 18 memory accesses */ #include "r2cf.h" static void r2cf_9(R *R0, R *R1, R *Cr, R *Ci, stride rs, stride csr, stride csi, INT v, INT ivs, INT ovs) { DK(KP907603734, +0.907603734547952313649323976213898122064543220); DK(KP852868531, +0.852868531952443209628250963940074071936020296); DK(KP347296355, +0.347296355333860697703433253538629592000751354); DK(KP666666666, +0.666666666666666666666666666666666666666666667); DK(KP879385241, +0.879385241571816768108218554649462939872416269); DK(KP984807753, +0.984807753012208059366743024589523013670643252); DK(KP673648177, +0.673648177666930348851716626769314796000375677); DK(KP898197570, +0.898197570222573798468955502359086394667167570); DK(KP939692620, +0.939692620785908384054109277324731469936208134); DK(KP866025403, +0.866025403784438646763723170752936183471402627); DK(KP203604859, +0.203604859554852403062088995281827210665664861); DK(KP152703644, +0.152703644666139302296566746461370407999248646); DK(KP394930843, +0.394930843634698457567117349190734585290304520); DK(KP968908795, +0.968908795874236621082202410917456709164223497); DK(KP726681596, +0.726681596905677465811651808188092531873167623); DK(KP586256827, +0.586256827714544512072145703099641959914944179); DK(KP184792530, +0.184792530904095372701352047572203755870913560); DK(KP500000000, +0.500000000000000000000000000000000000000000000); { INT i; for (i = v; i > 0; i = i - 1, R0 = R0 + ivs, R1 = R1 + ivs, Cr = Cr + ovs, Ci = Ci + ovs, MAKE_VOLATILE_STRIDE(36, rs), MAKE_VOLATILE_STRIDE(36, csr), MAKE_VOLATILE_STRIDE(36, csi)) { E Tp, Tz, Tw, Ts, TA; { E T1, T6, Tb, T7, T4, To, T8, Tc, Td, T2, T3; T1 = R0[0]; T2 = R1[WS(rs, 1)]; T3 = R0[WS(rs, 3)]; T6 = R1[0]; Tb = R0[WS(rs, 1)]; T7 = R0[WS(rs, 2)]; T4 = T2 + T3; To = T3 - T2; T8 = R1[WS(rs, 3)]; Tc = R1[WS(rs, 2)]; Td = R0[WS(rs, 4)]; { E T5, T9, Tk, Te, Ti; T5 = T1 + T4; Tp = FNMS(KP500000000, T4, T1); T9 = T7 + T8; Tk = T7 - T8; Te = Tc + Td; Ti = Td - Tc; { E Tl, Ta, Tu, Tf, Th; Tl = FMS(KP500000000, T9, T6); Ta = T6 + T9; Tu = FMA(KP184792530, Tk, Ti); Tf = Tb + Te; Th = FNMS(KP500000000, Te, Tb); { E Tq, Ty, Tm, Tt; Tq = FMA(KP586256827, Tl, Ti); Ty = FMA(KP726681596, Tk, Tl); Tm = FNMS(KP968908795, Tl, Tk); Tt = FMA(KP394930843, Th, To); { E Tj, Tx, Tg, Tv; Tj = FNMS(KP152703644, Ti, Th); Tx = FMA(KP203604859, Th, Ti); Tg = Ta + Tf; Ci[WS(csi, 3)] = KP866025403 * (Tf - Ta); Tv = FNMS(KP939692620, Tu, Tt); { E TB, Tn, TC, Tr; TB = FMA(KP898197570, Ty, Tx); Tz = FNMS(KP898197570, Ty, Tx); Tw = FNMS(KP673648177, Tm, Tj); Tn = FMA(KP673648177, Tm, Tj); Cr[0] = T5 + Tg; Cr[WS(csr, 3)] = FNMS(KP500000000, Tg, T5); Ci[WS(csi, 2)] = KP984807753 * (FNMS(KP879385241, Tv, Tl)); Ci[WS(csi, 1)] = -(KP984807753 * (FNMS(KP879385241, To, Tn))); TC = FMA(KP666666666, Tn, TB); Tr = FNMS(KP347296355, Tq, Tk); Ci[WS(csi, 4)] = KP866025403 * (FMA(KP852868531, TC, To)); Ts = FNMS(KP907603734, Tr, Th); } } } } } } Cr[WS(csr, 1)] = FMA(KP852868531, Tz, Tp); TA = FNMS(KP500000000, Tz, Tw); Cr[WS(csr, 2)] = FNMS(KP939692620, Ts, Tp); Cr[WS(csr, 4)] = FMA(KP852868531, TA, Tp); } } } static const kr2c_desc desc = { 9, "r2cf_9", {12, 4, 26, 0}, &GENUS }; void X(codelet_r2cf_9) (planner *p) { X(kr2c_register) (p, r2cf_9, &desc); } #else /* HAVE_FMA */ /* Generated by: ../../../genfft/gen_r2cf.native -compact -variables 4 -pipeline-latency 4 -n 9 -name r2cf_9 -include r2cf.h */ /* * This function contains 38 FP additions, 26 FP multiplications, * (or, 21 additions, 9 multiplications, 17 fused multiply/add), * 36 stack variables, 14 constants, and 18 memory accesses */ #include "r2cf.h" static void r2cf_9(R *R0, R *R1, R *Cr, R *Ci, stride rs, stride csr, stride csi, INT v, INT ivs, INT ovs) { DK(KP939692620, +0.939692620785908384054109277324731469936208134); DK(KP296198132, +0.296198132726023843175338011893050938967728390); DK(KP342020143, +0.342020143325668733044099614682259580763083368); DK(KP813797681, +0.813797681349373692844693217248393223289101568); DK(KP984807753, +0.984807753012208059366743024589523013670643252); DK(KP150383733, +0.150383733180435296639271897612501926072238258); DK(KP642787609, +0.642787609686539326322643409907263432907559884); DK(KP663413948, +0.663413948168938396205421319635891297216863310); DK(KP852868531, +0.852868531952443209628250963940074071936020296); DK(KP173648177, +0.173648177666930348851716626769314796000375677); DK(KP556670399, +0.556670399226419366452912952047023132968291906); DK(KP766044443, +0.766044443118978035202392650555416673935832457); DK(KP866025403, +0.866025403784438646763723170752936183471402627); DK(KP500000000, +0.500000000000000000000000000000000000000000000); { INT i; for (i = v; i > 0; i = i - 1, R0 = R0 + ivs, R1 = R1 + ivs, Cr = Cr + ovs, Ci = Ci + ovs, MAKE_VOLATILE_STRIDE(36, rs), MAKE_VOLATILE_STRIDE(36, csr), MAKE_VOLATILE_STRIDE(36, csi)) { E T1, T4, Tr, Ta, Tl, Ti, Tf, Tk, Tj, T2, T3, T5, Tg; T1 = R0[0]; T2 = R1[WS(rs, 1)]; T3 = R0[WS(rs, 3)]; T4 = T2 + T3; Tr = T3 - T2; { E T6, T7, T8, T9; T6 = R1[0]; T7 = R0[WS(rs, 2)]; T8 = R1[WS(rs, 3)]; T9 = T7 + T8; Ta = T6 + T9; Tl = T8 - T7; Ti = FNMS(KP500000000, T9, T6); } { E Tb, Tc, Td, Te; Tb = R0[WS(rs, 1)]; Tc = R1[WS(rs, 2)]; Td = R0[WS(rs, 4)]; Te = Tc + Td; Tf = Tb + Te; Tk = FNMS(KP500000000, Te, Tb); Tj = Td - Tc; } Ci[WS(csi, 3)] = KP866025403 * (Tf - Ta); T5 = T1 + T4; Tg = Ta + Tf; Cr[WS(csr, 3)] = FNMS(KP500000000, Tg, T5); Cr[0] = T5 + Tg; { E Tt, Th, Tm, Tn, To, Tp, Tq, Ts; Tt = KP866025403 * Tr; Th = FNMS(KP500000000, T4, T1); Tm = FMA(KP766044443, Ti, KP556670399 * Tl); Tn = FMA(KP173648177, Tk, KP852868531 * Tj); To = Tm + Tn; Tp = FNMS(KP642787609, Ti, KP663413948 * Tl); Tq = FNMS(KP984807753, Tk, KP150383733 * Tj); Ts = Tp + Tq; Cr[WS(csr, 1)] = Th + To; Ci[WS(csi, 1)] = Tt + Ts; Cr[WS(csr, 4)] = FMA(KP866025403, Tp - Tq, Th) - (KP500000000 * To); Ci[WS(csi, 4)] = FNMS(KP500000000, Ts, KP866025403 * (Tr + (Tn - Tm))); Ci[WS(csi, 2)] = FNMS(KP342020143, Tk, KP813797681 * Tj) + FNMA(KP150383733, Tl, KP984807753 * Ti) - Tt; Cr[WS(csr, 2)] = FMA(KP173648177, Ti, Th) + FNMA(KP296198132, Tj, KP939692620 * Tk) - (KP852868531 * Tl); } } } } static const kr2c_desc desc = { 9, "r2cf_9", {21, 9, 17, 0}, &GENUS }; void X(codelet_r2cf_9) (planner *p) { X(kr2c_register) (p, r2cf_9, &desc); } #endif /* HAVE_FMA */ fftw-3.3.3/rdft/scalar/r2cf/r2cf_10.c0000644000175000017500000001427012054410622013747 00000000000000/* * Copyright (c) 2003, 2007-11 Matteo Frigo * Copyright (c) 2003, 2007-11 Massachusetts Institute of Technology * * This program is free software; you can redistribute it and/or modify * it under the terms of the GNU General Public License as published by * the Free Software Foundation; either version 2 of the License, or * (at your option) any later version. * * This program is distributed in the hope that it will be useful, * but WITHOUT ANY WARRANTY; without even the implied warranty of * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the * GNU General Public License for more details. * * You should have received a copy of the GNU General Public License * along with this program; if not, write to the Free Software * Foundation, Inc., 51 Franklin Street, Fifth Floor, Boston, MA 02110-1301 USA * */ /* This file was automatically generated --- DO NOT EDIT */ /* Generated on Sun Nov 25 07:39:45 EST 2012 */ #include "codelet-rdft.h" #ifdef HAVE_FMA /* Generated by: ../../../genfft/gen_r2cf.native -fma -reorder-insns -schedule-for-pipeline -compact -variables 4 -pipeline-latency 4 -n 10 -name r2cf_10 -include r2cf.h */ /* * This function contains 34 FP additions, 14 FP multiplications, * (or, 24 additions, 4 multiplications, 10 fused multiply/add), * 29 stack variables, 4 constants, and 20 memory accesses */ #include "r2cf.h" static void r2cf_10(R *R0, R *R1, R *Cr, R *Ci, stride rs, stride csr, stride csi, INT v, INT ivs, INT ovs) { DK(KP559016994, +0.559016994374947424102293417182819058860154590); DK(KP250000000, +0.250000000000000000000000000000000000000000000); DK(KP618033988, +0.618033988749894848204586834365638117720309180); DK(KP951056516, +0.951056516295153572116439333379382143405698634); { INT i; for (i = v; i > 0; i = i - 1, R0 = R0 + ivs, R1 = R1 + ivs, Cr = Cr + ovs, Ci = Ci + ovs, MAKE_VOLATILE_STRIDE(40, rs), MAKE_VOLATILE_STRIDE(40, csr), MAKE_VOLATILE_STRIDE(40, csi)) { E Tt, T3, T7, Tq, T6, Tv, Tp, Tm, Th, T8, T1, T2, T9, Tr; T1 = R0[0]; T2 = R1[WS(rs, 2)]; { E Te, Tn, Td, Tf, Tb, Tc; Tb = R0[WS(rs, 2)]; Tc = R1[WS(rs, 4)]; Te = R0[WS(rs, 3)]; Tt = T1 + T2; T3 = T1 - T2; Tn = Tb + Tc; Td = Tb - Tc; Tf = R1[0]; { E T4, T5, To, Tg; T4 = R0[WS(rs, 1)]; T5 = R1[WS(rs, 3)]; T7 = R0[WS(rs, 4)]; To = Te + Tf; Tg = Te - Tf; Tq = T4 + T5; T6 = T4 - T5; Tv = Tn + To; Tp = Tn - To; Tm = Tg - Td; Th = Td + Tg; T8 = R1[WS(rs, 1)]; } } T9 = T7 - T8; Tr = T7 + T8; { E Ty, Tk, Tx, Tj, Tu, Ts; Tu = Tq + Tr; Ts = Tq - Tr; { E Ta, Tl, Tw, Ti; Ta = T6 + T9; Tl = T6 - T9; Ci[WS(csi, 4)] = KP951056516 * (FMA(KP618033988, Tp, Ts)); Ci[WS(csi, 2)] = KP951056516 * (FNMS(KP618033988, Ts, Tp)); Ty = Tu - Tv; Tw = Tu + Tv; Ci[WS(csi, 3)] = KP951056516 * (FMA(KP618033988, Tl, Tm)); Ci[WS(csi, 1)] = -(KP951056516 * (FNMS(KP618033988, Tm, Tl))); Tk = Ta - Th; Ti = Ta + Th; Cr[0] = Tt + Tw; Tx = FNMS(KP250000000, Tw, Tt); Cr[WS(csr, 5)] = T3 + Ti; Tj = FNMS(KP250000000, Ti, T3); } Cr[WS(csr, 4)] = FMA(KP559016994, Ty, Tx); Cr[WS(csr, 2)] = FNMS(KP559016994, Ty, Tx); Cr[WS(csr, 3)] = FNMS(KP559016994, Tk, Tj); Cr[WS(csr, 1)] = FMA(KP559016994, Tk, Tj); } } } } static const kr2c_desc desc = { 10, "r2cf_10", {24, 4, 10, 0}, &GENUS }; void X(codelet_r2cf_10) (planner *p) { X(kr2c_register) (p, r2cf_10, &desc); } #else /* HAVE_FMA */ /* Generated by: ../../../genfft/gen_r2cf.native -compact -variables 4 -pipeline-latency 4 -n 10 -name r2cf_10 -include r2cf.h */ /* * This function contains 34 FP additions, 12 FP multiplications, * (or, 28 additions, 6 multiplications, 6 fused multiply/add), * 26 stack variables, 4 constants, and 20 memory accesses */ #include "r2cf.h" static void r2cf_10(R *R0, R *R1, R *Cr, R *Ci, stride rs, stride csr, stride csi, INT v, INT ivs, INT ovs) { DK(KP250000000, +0.250000000000000000000000000000000000000000000); DK(KP559016994, +0.559016994374947424102293417182819058860154590); DK(KP951056516, +0.951056516295153572116439333379382143405698634); DK(KP587785252, +0.587785252292473129168705954639072768597652438); { INT i; for (i = v; i > 0; i = i - 1, R0 = R0 + ivs, R1 = R1 + ivs, Cr = Cr + ovs, Ci = Ci + ovs, MAKE_VOLATILE_STRIDE(40, rs), MAKE_VOLATILE_STRIDE(40, csr), MAKE_VOLATILE_STRIDE(40, csi)) { E Ti, Tt, Ta, Tn, Td, To, Te, Tv, T3, Tq, T6, Tr, T7, Tu, Tg; E Th; Tg = R0[0]; Th = R1[WS(rs, 2)]; Ti = Tg - Th; Tt = Tg + Th; { E T8, T9, Tb, Tc; T8 = R0[WS(rs, 2)]; T9 = R1[WS(rs, 4)]; Ta = T8 - T9; Tn = T8 + T9; Tb = R0[WS(rs, 3)]; Tc = R1[0]; Td = Tb - Tc; To = Tb + Tc; } Te = Ta + Td; Tv = Tn + To; { E T1, T2, T4, T5; T1 = R0[WS(rs, 1)]; T2 = R1[WS(rs, 3)]; T3 = T1 - T2; Tq = T1 + T2; T4 = R0[WS(rs, 4)]; T5 = R1[WS(rs, 1)]; T6 = T4 - T5; Tr = T4 + T5; } T7 = T3 + T6; Tu = Tq + Tr; { E Tl, Tm, Tf, Tj, Tk; Tl = Td - Ta; Tm = T3 - T6; Ci[WS(csi, 1)] = FNMS(KP951056516, Tm, KP587785252 * Tl); Ci[WS(csi, 3)] = FMA(KP587785252, Tm, KP951056516 * Tl); Tf = KP559016994 * (T7 - Te); Tj = T7 + Te; Tk = FNMS(KP250000000, Tj, Ti); Cr[WS(csr, 1)] = Tf + Tk; Cr[WS(csr, 5)] = Ti + Tj; Cr[WS(csr, 3)] = Tk - Tf; } { E Tp, Ts, Ty, Tw, Tx; Tp = Tn - To; Ts = Tq - Tr; Ci[WS(csi, 2)] = FNMS(KP587785252, Ts, KP951056516 * Tp); Ci[WS(csi, 4)] = FMA(KP951056516, Ts, KP587785252 * Tp); Ty = KP559016994 * (Tu - Tv); Tw = Tu + Tv; Tx = FNMS(KP250000000, Tw, Tt); Cr[WS(csr, 2)] = Tx - Ty; Cr[0] = Tt + Tw; Cr[WS(csr, 4)] = Ty + Tx; } } } } static const kr2c_desc desc = { 10, "r2cf_10", {28, 6, 6, 0}, &GENUS }; void X(codelet_r2cf_10) (planner *p) { X(kr2c_register) (p, r2cf_10, &desc); } #endif /* HAVE_FMA */ fftw-3.3.3/rdft/scalar/r2cf/r2cf_11.c0000644000175000017500000002162212054410624013751 00000000000000/* * Copyright (c) 2003, 2007-11 Matteo Frigo * Copyright (c) 2003, 2007-11 Massachusetts Institute of Technology * * This program is free software; you can redistribute it and/or modify * it under the terms of the GNU General Public License as published by * the Free Software Foundation; either version 2 of the License, or * (at your option) any later version. * * This program is distributed in the hope that it will be useful, * but WITHOUT ANY WARRANTY; without even the implied warranty of * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the * GNU General Public License for more details. * * You should have received a copy of the GNU General Public License * along with this program; if not, write to the Free Software * Foundation, Inc., 51 Franklin Street, Fifth Floor, Boston, MA 02110-1301 USA * */ /* This file was automatically generated --- DO NOT EDIT */ /* Generated on Sun Nov 25 07:39:45 EST 2012 */ #include "codelet-rdft.h" #ifdef HAVE_FMA /* Generated by: ../../../genfft/gen_r2cf.native -fma -reorder-insns -schedule-for-pipeline -compact -variables 4 -pipeline-latency 4 -n 11 -name r2cf_11 -include r2cf.h */ /* * This function contains 60 FP additions, 50 FP multiplications, * (or, 15 additions, 5 multiplications, 45 fused multiply/add), * 51 stack variables, 10 constants, and 22 memory accesses */ #include "r2cf.h" static void r2cf_11(R *R0, R *R1, R *Cr, R *Ci, stride rs, stride csr, stride csi, INT v, INT ivs, INT ovs) { DK(KP959492973, +0.959492973614497389890368057066327699062454848); DK(KP876768831, +0.876768831002589333891339807079336796764054852); DK(KP918985947, +0.918985947228994779780736114132655398124909697); DK(KP989821441, +0.989821441880932732376092037776718787376519372); DK(KP778434453, +0.778434453334651800608337670740821884709317477); DK(KP830830026, +0.830830026003772851058548298459246407048009821); DK(KP715370323, +0.715370323453429719112414662767260662417897278); DK(KP634356270, +0.634356270682424498893150776899916060542806975); DK(KP342584725, +0.342584725681637509502641509861112333758894680); DK(KP521108558, +0.521108558113202722944698153526659300680427422); { INT i; for (i = v; i > 0; i = i - 1, R0 = R0 + ivs, R1 = R1 + ivs, Cr = Cr + ovs, Ci = Ci + ovs, MAKE_VOLATILE_STRIDE(44, rs), MAKE_VOLATILE_STRIDE(44, csr), MAKE_VOLATILE_STRIDE(44, csi)) { E T1, Tg, TF, TB, TI, TL, Tz, TA; { E T4, TC, TE, T7, TD, Ta, TS, TG, TJ, Td, TP, TM, Ty, Tq, Th; E Tt, Tl; T1 = R0[0]; { E Tb, Tc, Tx, Tp; { E T2, T3, Te, Tf; T2 = R1[0]; T3 = R0[WS(rs, 5)]; Te = R1[WS(rs, 2)]; Tf = R0[WS(rs, 3)]; { E T5, T6, T8, T9; T5 = R0[WS(rs, 1)]; T4 = T2 + T3; TC = T3 - T2; Tg = Te + Tf; TE = Tf - Te; T6 = R1[WS(rs, 4)]; T8 = R1[WS(rs, 1)]; T9 = R0[WS(rs, 4)]; Tb = R0[WS(rs, 2)]; T7 = T5 + T6; TD = T5 - T6; Ta = T8 + T9; TF = T9 - T8; Tc = R1[WS(rs, 3)]; } } TS = FMA(KP521108558, TC, TD); TG = FMA(KP521108558, TF, TE); TJ = FMA(KP521108558, TE, TC); Td = Tb + Tc; TB = Tb - Tc; Tx = FNMS(KP342584725, Ta, T7); Tp = FNMS(KP342584725, T4, Ta); TP = FNMS(KP521108558, TB, TF); TM = FNMS(KP521108558, TD, TB); Ty = FNMS(KP634356270, Tx, Td); Tq = FNMS(KP634356270, Tp, Tg); Th = FNMS(KP342584725, Tg, Td); Tt = FNMS(KP342584725, Td, T4); Tl = FNMS(KP342584725, T7, Tg); } { E Tu, Ts, TN, Tv; { E Tm, TU, Tj, Ti, TT; TT = FMA(KP715370323, TS, TF); Ti = FNMS(KP634356270, Th, Ta); Tu = FNMS(KP634356270, Tt, T7); Tm = FNMS(KP634356270, Tl, T4); TU = FMA(KP830830026, TT, TB); Tj = FNMS(KP778434453, Ti, T7); { E Tk, TR, To, Tn, TQ, Tr; TQ = FMA(KP715370323, TP, TC); Tn = FNMS(KP778434453, Tm, Ta); Ci[WS(csi, 5)] = KP989821441 * (FMA(KP918985947, TU, TE)); Tk = FNMS(KP876768831, Tj, T4); TR = FNMS(KP830830026, TQ, TE); To = FNMS(KP876768831, Tn, Td); Tr = FNMS(KP778434453, Tq, Td); Cr[WS(csr, 5)] = FNMS(KP959492973, Tk, T1); Ci[WS(csi, 4)] = KP989821441 * (FNMS(KP918985947, TR, TD)); Cr[WS(csr, 4)] = FNMS(KP959492973, To, T1); Ts = FNMS(KP876768831, Tr, T7); } } TN = FNMS(KP715370323, TM, TE); Tv = FNMS(KP778434453, Tu, Tg); Cr[0] = T1 + T4 + T7 + Ta + Td + Tg; Cr[WS(csr, 3)] = FNMS(KP959492973, Ts, T1); { E TO, Tw, TH, TK; TO = FNMS(KP830830026, TN, TF); Tw = FNMS(KP876768831, Tv, Ta); TH = FMA(KP715370323, TG, TD); TK = FNMS(KP715370323, TJ, TB); Ci[WS(csi, 3)] = KP989821441 * (FNMS(KP918985947, TO, TC)); Cr[WS(csr, 2)] = FNMS(KP959492973, Tw, T1); TI = FNMS(KP830830026, TH, TC); TL = FMA(KP830830026, TK, TD); Tz = FNMS(KP778434453, Ty, T4); } } } Ci[WS(csi, 2)] = KP989821441 * (FMA(KP918985947, TI, TB)); Ci[WS(csi, 1)] = KP989821441 * (FNMS(KP918985947, TL, TF)); TA = FNMS(KP876768831, Tz, Tg); Cr[WS(csr, 1)] = FNMS(KP959492973, TA, T1); } } } static const kr2c_desc desc = { 11, "r2cf_11", {15, 5, 45, 0}, &GENUS }; void X(codelet_r2cf_11) (planner *p) { X(kr2c_register) (p, r2cf_11, &desc); } #else /* HAVE_FMA */ /* Generated by: ../../../genfft/gen_r2cf.native -compact -variables 4 -pipeline-latency 4 -n 11 -name r2cf_11 -include r2cf.h */ /* * This function contains 60 FP additions, 50 FP multiplications, * (or, 20 additions, 10 multiplications, 40 fused multiply/add), * 28 stack variables, 10 constants, and 22 memory accesses */ #include "r2cf.h" static void r2cf_11(R *R0, R *R1, R *Cr, R *Ci, stride rs, stride csr, stride csi, INT v, INT ivs, INT ovs) { DK(KP654860733, +0.654860733945285064056925072466293553183791199); DK(KP142314838, +0.142314838273285140443792668616369668791051361); DK(KP959492973, +0.959492973614497389890368057066327699062454848); DK(KP415415013, +0.415415013001886425529274149229623203524004910); DK(KP841253532, +0.841253532831181168861811648919367717513292498); DK(KP989821441, +0.989821441880932732376092037776718787376519372); DK(KP909631995, +0.909631995354518371411715383079028460060241051); DK(KP281732556, +0.281732556841429697711417915346616899035777899); DK(KP540640817, +0.540640817455597582107635954318691695431770608); DK(KP755749574, +0.755749574354258283774035843972344420179717445); { INT i; for (i = v; i > 0; i = i - 1, R0 = R0 + ivs, R1 = R1 + ivs, Cr = Cr + ovs, Ci = Ci + ovs, MAKE_VOLATILE_STRIDE(44, rs), MAKE_VOLATILE_STRIDE(44, csr), MAKE_VOLATILE_STRIDE(44, csi)) { E T1, T4, Tl, Tg, Th, Td, Ti, Ta, Tk, T7, Tj, Tb, Tc; T1 = R0[0]; { E T2, T3, Te, Tf; T2 = R0[WS(rs, 1)]; T3 = R1[WS(rs, 4)]; T4 = T2 + T3; Tl = T3 - T2; Te = R1[0]; Tf = R0[WS(rs, 5)]; Tg = Te + Tf; Th = Tf - Te; } Tb = R1[WS(rs, 1)]; Tc = R0[WS(rs, 4)]; Td = Tb + Tc; Ti = Tc - Tb; { E T8, T9, T5, T6; T8 = R1[WS(rs, 2)]; T9 = R0[WS(rs, 3)]; Ta = T8 + T9; Tk = T9 - T8; T5 = R0[WS(rs, 2)]; T6 = R1[WS(rs, 3)]; T7 = T5 + T6; Tj = T6 - T5; } Ci[WS(csi, 4)] = FMA(KP755749574, Th, KP540640817 * Ti) + FNMS(KP909631995, Tk, KP281732556 * Tj) - (KP989821441 * Tl); Cr[WS(csr, 4)] = FMA(KP841253532, Td, T1) + FNMS(KP959492973, T7, KP415415013 * Ta) + FNMA(KP142314838, T4, KP654860733 * Tg); Ci[WS(csi, 2)] = FMA(KP909631995, Th, KP755749574 * Tl) + FNMA(KP540640817, Tk, KP989821441 * Tj) - (KP281732556 * Ti); Ci[WS(csi, 5)] = FMA(KP281732556, Th, KP755749574 * Ti) + FNMS(KP909631995, Tj, KP989821441 * Tk) - (KP540640817 * Tl); Ci[WS(csi, 1)] = FMA(KP540640817, Th, KP909631995 * Tl) + FMA(KP989821441, Ti, KP755749574 * Tj) + (KP281732556 * Tk); Ci[WS(csi, 3)] = FMA(KP989821441, Th, KP540640817 * Tj) + FNMS(KP909631995, Ti, KP755749574 * Tk) - (KP281732556 * Tl); Cr[WS(csr, 3)] = FMA(KP415415013, Td, T1) + FNMS(KP654860733, Ta, KP841253532 * T7) + FNMA(KP959492973, T4, KP142314838 * Tg); Cr[WS(csr, 1)] = FMA(KP841253532, Tg, T1) + FNMS(KP959492973, Ta, KP415415013 * T4) + FNMA(KP654860733, T7, KP142314838 * Td); Cr[0] = T1 + Tg + T4 + Td + T7 + Ta; Cr[WS(csr, 2)] = FMA(KP415415013, Tg, T1) + FNMS(KP142314838, T7, KP841253532 * Ta) + FNMA(KP959492973, Td, KP654860733 * T4); Cr[WS(csr, 5)] = FMA(KP841253532, T4, T1) + FNMS(KP142314838, Ta, KP415415013 * T7) + FNMA(KP654860733, Td, KP959492973 * Tg); } } } static const kr2c_desc desc = { 11, "r2cf_11", {20, 10, 40, 0}, &GENUS }; void X(codelet_r2cf_11) (planner *p) { X(kr2c_register) (p, r2cf_11, &desc); } #endif /* HAVE_FMA */ fftw-3.3.3/rdft/scalar/r2cf/r2cf_12.c0000644000175000017500000001437012054410622013752 00000000000000/* * Copyright (c) 2003, 2007-11 Matteo Frigo * Copyright (c) 2003, 2007-11 Massachusetts Institute of Technology * * This program is free software; you can redistribute it and/or modify * it under the terms of the GNU General Public License as published by * the Free Software Foundation; either version 2 of the License, or * (at your option) any later version. * * This program is distributed in the hope that it will be useful, * but WITHOUT ANY WARRANTY; without even the implied warranty of * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the * GNU General Public License for more details. * * You should have received a copy of the GNU General Public License * along with this program; if not, write to the Free Software * Foundation, Inc., 51 Franklin Street, Fifth Floor, Boston, MA 02110-1301 USA * */ /* This file was automatically generated --- DO NOT EDIT */ /* Generated on Sun Nov 25 07:39:46 EST 2012 */ #include "codelet-rdft.h" #ifdef HAVE_FMA /* Generated by: ../../../genfft/gen_r2cf.native -fma -reorder-insns -schedule-for-pipeline -compact -variables 4 -pipeline-latency 4 -n 12 -name r2cf_12 -include r2cf.h */ /* * This function contains 38 FP additions, 10 FP multiplications, * (or, 30 additions, 2 multiplications, 8 fused multiply/add), * 31 stack variables, 2 constants, and 24 memory accesses */ #include "r2cf.h" static void r2cf_12(R *R0, R *R1, R *Cr, R *Ci, stride rs, stride csr, stride csi, INT v, INT ivs, INT ovs) { DK(KP866025403, +0.866025403784438646763723170752936183471402627); DK(KP500000000, +0.500000000000000000000000000000000000000000000); { INT i; for (i = v; i > 0; i = i - 1, R0 = R0 + ivs, R1 = R1 + ivs, Cr = Cr + ovs, Ci = Ci + ovs, MAKE_VOLATILE_STRIDE(48, rs), MAKE_VOLATILE_STRIDE(48, csr), MAKE_VOLATILE_STRIDE(48, csi)) { E Tm, T6, Ty, Tp, T5, Tk, Tt, Tb, Tc, Td, T9, Tn; { E T1, Tg, Th, Ti, T4, T2, T3, T7, T8, Tj; T1 = R0[0]; T2 = R0[WS(rs, 2)]; T3 = R0[WS(rs, 4)]; Tg = R1[WS(rs, 1)]; Th = R1[WS(rs, 3)]; Ti = R1[WS(rs, 5)]; T4 = T2 + T3; Tm = T3 - T2; T6 = R0[WS(rs, 3)]; Ty = Ti - Th; Tj = Th + Ti; Tp = FNMS(KP500000000, T4, T1); T5 = T1 + T4; T7 = R0[WS(rs, 5)]; Tk = FNMS(KP500000000, Tj, Tg); Tt = Tg + Tj; T8 = R0[WS(rs, 1)]; Tb = R1[WS(rs, 4)]; Tc = R1[0]; Td = R1[WS(rs, 2)]; T9 = T7 + T8; Tn = T8 - T7; } { E Te, Tz, To, TC; Te = Tc + Td; Tz = Td - Tc; To = Tm - Tn; TC = Tm + Tn; { E Ta, Tq, TA, TB; Ta = T6 + T9; Tq = FNMS(KP500000000, T9, T6); TA = Ty - Tz; TB = Ty + Tz; { E Tf, Tu, Tx, Tr; Tf = FNMS(KP500000000, Te, Tb); Tu = Tb + Te; Tx = Tp - Tq; Tr = Tp + Tq; { E Tv, Tw, Tl, Ts; Tv = T5 + Ta; Cr[WS(csr, 3)] = T5 - Ta; Ci[WS(csi, 4)] = KP866025403 * (TC + TB); Ci[WS(csi, 2)] = KP866025403 * (TB - TC); Tw = Tt + Tu; Ci[WS(csi, 3)] = Tt - Tu; Tl = Tf - Tk; Ts = Tk + Tf; Cr[WS(csr, 1)] = FMA(KP866025403, TA, Tx); Cr[WS(csr, 5)] = FNMS(KP866025403, TA, Tx); Cr[0] = Tv + Tw; Cr[WS(csr, 6)] = Tv - Tw; Cr[WS(csr, 4)] = Tr + Ts; Cr[WS(csr, 2)] = Tr - Ts; Ci[WS(csi, 5)] = FNMS(KP866025403, To, Tl); Ci[WS(csi, 1)] = FMA(KP866025403, To, Tl); } } } } } } } static const kr2c_desc desc = { 12, "r2cf_12", {30, 2, 8, 0}, &GENUS }; void X(codelet_r2cf_12) (planner *p) { X(kr2c_register) (p, r2cf_12, &desc); } #else /* HAVE_FMA */ /* Generated by: ../../../genfft/gen_r2cf.native -compact -variables 4 -pipeline-latency 4 -n 12 -name r2cf_12 -include r2cf.h */ /* * This function contains 38 FP additions, 8 FP multiplications, * (or, 34 additions, 4 multiplications, 4 fused multiply/add), * 21 stack variables, 2 constants, and 24 memory accesses */ #include "r2cf.h" static void r2cf_12(R *R0, R *R1, R *Cr, R *Ci, stride rs, stride csr, stride csi, INT v, INT ivs, INT ovs) { DK(KP866025403, +0.866025403784438646763723170752936183471402627); DK(KP500000000, +0.500000000000000000000000000000000000000000000); { INT i; for (i = v; i > 0; i = i - 1, R0 = R0 + ivs, R1 = R1 + ivs, Cr = Cr + ovs, Ci = Ci + ovs, MAKE_VOLATILE_STRIDE(48, rs), MAKE_VOLATILE_STRIDE(48, csr), MAKE_VOLATILE_STRIDE(48, csi)) { E T5, Tp, Tb, Tn, Ty, Tt, Ta, Tq, Tc, Ti, Tz, Tu, Td, To; { E T1, T2, T3, T4; T1 = R0[0]; T2 = R0[WS(rs, 2)]; T3 = R0[WS(rs, 4)]; T4 = T2 + T3; T5 = T1 + T4; Tp = FNMS(KP500000000, T4, T1); Tb = T3 - T2; } { E Tj, Tk, Tl, Tm; Tj = R1[WS(rs, 1)]; Tk = R1[WS(rs, 3)]; Tl = R1[WS(rs, 5)]; Tm = Tk + Tl; Tn = FNMS(KP500000000, Tm, Tj); Ty = Tl - Tk; Tt = Tj + Tm; } { E T6, T7, T8, T9; T6 = R0[WS(rs, 3)]; T7 = R0[WS(rs, 5)]; T8 = R0[WS(rs, 1)]; T9 = T7 + T8; Ta = T6 + T9; Tq = FNMS(KP500000000, T9, T6); Tc = T8 - T7; } { E Te, Tf, Tg, Th; Te = R1[WS(rs, 4)]; Tf = R1[0]; Tg = R1[WS(rs, 2)]; Th = Tf + Tg; Ti = FNMS(KP500000000, Th, Te); Tz = Tg - Tf; Tu = Te + Th; } Cr[WS(csr, 3)] = T5 - Ta; Ci[WS(csi, 3)] = Tt - Tu; Td = KP866025403 * (Tb - Tc); To = Ti - Tn; Ci[WS(csi, 1)] = Td + To; Ci[WS(csi, 5)] = To - Td; { E Tx, TA, Tv, Tw; Tx = Tp - Tq; TA = KP866025403 * (Ty - Tz); Cr[WS(csr, 5)] = Tx - TA; Cr[WS(csr, 1)] = Tx + TA; Tv = T5 + Ta; Tw = Tt + Tu; Cr[WS(csr, 6)] = Tv - Tw; Cr[0] = Tv + Tw; } { E Tr, Ts, TB, TC; Tr = Tp + Tq; Ts = Tn + Ti; Cr[WS(csr, 2)] = Tr - Ts; Cr[WS(csr, 4)] = Tr + Ts; TB = Ty + Tz; TC = Tb + Tc; Ci[WS(csi, 2)] = KP866025403 * (TB - TC); Ci[WS(csi, 4)] = KP866025403 * (TC + TB); } } } } static const kr2c_desc desc = { 12, "r2cf_12", {34, 4, 4, 0}, &GENUS }; void X(codelet_r2cf_12) (planner *p) { X(kr2c_register) (p, r2cf_12, &desc); } #endif /* HAVE_FMA */ fftw-3.3.3/rdft/scalar/r2cf/r2cf_13.c0000644000175000017500000003041112054410623013746 00000000000000/* * Copyright (c) 2003, 2007-11 Matteo Frigo * Copyright (c) 2003, 2007-11 Massachusetts Institute of Technology * * This program is free software; you can redistribute it and/or modify * it under the terms of the GNU General Public License as published by * the Free Software Foundation; either version 2 of the License, or * (at your option) any later version. * * This program is distributed in the hope that it will be useful, * but WITHOUT ANY WARRANTY; without even the implied warranty of * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the * GNU General Public License for more details. * * You should have received a copy of the GNU General Public License * along with this program; if not, write to the Free Software * Foundation, Inc., 51 Franklin Street, Fifth Floor, Boston, MA 02110-1301 USA * */ /* This file was automatically generated --- DO NOT EDIT */ /* Generated on Sun Nov 25 07:39:46 EST 2012 */ #include "codelet-rdft.h" #ifdef HAVE_FMA /* Generated by: ../../../genfft/gen_r2cf.native -fma -reorder-insns -schedule-for-pipeline -compact -variables 4 -pipeline-latency 4 -n 13 -name r2cf_13 -include r2cf.h */ /* * This function contains 76 FP additions, 51 FP multiplications, * (or, 31 additions, 6 multiplications, 45 fused multiply/add), * 68 stack variables, 23 constants, and 26 memory accesses */ #include "r2cf.h" static void r2cf_13(R *R0, R *R1, R *Cr, R *Ci, stride rs, stride csr, stride csi, INT v, INT ivs, INT ovs) { DK(KP516520780, +0.516520780623489722840901288569017135705033622); DK(KP300462606, +0.300462606288665774426601772289207995520941381); DK(KP581704778, +0.581704778510515730456870384989698884939833902); DK(KP859542535, +0.859542535098774820163672132761689612766401925); DK(KP769338817, +0.769338817572980603471413688209101117038278899); DK(KP686558370, +0.686558370781754340655719594850823015421401653); DK(KP514918778, +0.514918778086315755491789696138117261566051239); DK(KP251768516, +0.251768516431883313623436926934233488546674281); DK(KP503537032, +0.503537032863766627246873853868466977093348562); DK(KP904176221, +0.904176221990848204433795481776887926501523162); DK(KP575140729, +0.575140729474003121368385547455453388461001608); DK(KP957805992, +0.957805992594665126462521754605754580515587217); DK(KP600477271, +0.600477271932665282925769253334763009352012849); DK(KP522026385, +0.522026385161275033714027226654165028300441940); DK(KP301479260, +0.301479260047709873958013540496673347309208464); DK(KP226109445, +0.226109445035782405468510155372505010481906348); DK(KP853480001, +0.853480001859823990758994934970528322872359049); DK(KP083333333, +0.083333333333333333333333333333333333333333333); DK(KP612264650, +0.612264650376756543746494474777125408779395514); DK(KP038632954, +0.038632954644348171955506895830342264440241080); DK(KP302775637, +0.302775637731994646559610633735247973125648287); DK(KP866025403, +0.866025403784438646763723170752936183471402627); DK(KP500000000, +0.500000000000000000000000000000000000000000000); { INT i; for (i = v; i > 0; i = i - 1, R0 = R0 + ivs, R1 = R1 + ivs, Cr = Cr + ovs, Ci = Ci + ovs, MAKE_VOLATILE_STRIDE(52, rs), MAKE_VOLATILE_STRIDE(52, csr), MAKE_VOLATILE_STRIDE(52, csi)) { E T15, T1a, T11, T17, T14, T1b; { E TN, TD, TV, TA, Tb, TZ, T12, TS, Tx, Tu, Ti, TU; TN = R0[0]; { E T3, TP, Th, TB, Tp, Te, Tm, TC, Tr, T6, T9, Ts; { E Tn, Tf, Tg, T1, T2; T1 = R0[WS(rs, 4)]; T2 = R1[WS(rs, 2)]; Tn = R0[WS(rs, 6)]; Tf = R0[WS(rs, 5)]; Tg = R0[WS(rs, 2)]; T3 = T1 - T2; TP = T1 + T2; { E Tk, To, Tc, Td; Tk = R1[0]; Th = Tf - Tg; To = Tf + Tg; Tc = R1[WS(rs, 4)]; Td = R1[WS(rs, 1)]; { E T4, Tl, T5, T7, T8; T4 = R1[WS(rs, 5)]; TB = Tn + To; Tp = FMS(KP500000000, To, Tn); Tl = Td + Tc; Te = Tc - Td; T5 = R0[WS(rs, 3)]; T7 = R1[WS(rs, 3)]; T8 = R0[WS(rs, 1)]; Tm = FNMS(KP500000000, Tl, Tk); TC = Tk + Tl; Tr = T4 + T5; T6 = T4 - T5; T9 = T7 - T8; Ts = T7 + T8; } } } { E TO, Ta, Tt, TQ; TD = TB - TC; TO = TC + TB; Ta = T6 + T9; TV = T6 - T9; Tt = Tr - Ts; TQ = Tr + Ts; { E TX, Tq, TR, TY; TX = Tm - Tp; Tq = Tm + Tp; TA = T3 + Ta; Tb = FNMS(KP500000000, Ta, T3); TR = TP + TQ; TY = FNMS(KP500000000, TQ, TP); TZ = TX + TY; T12 = TX - TY; T15 = TO - TR; TS = TO + TR; Tx = FNMS(KP866025403, Tt, Tq); Tu = FMA(KP866025403, Tt, Tq); Ti = Te + Th; TU = Th - Te; } } } Cr[0] = TN + TS; { E Tw, Tj, T13, TW; Tw = FNMS(KP866025403, Ti, Tb); Tj = FMA(KP866025403, Ti, Tb); T13 = TU - TV; TW = TU + TV; { E TE, TI, Tv, TF, TG, Ty; TE = FMA(KP302775637, TD, TA); TI = FNMS(KP302775637, TA, TD); Tv = FMA(KP038632954, Tu, Tj); TF = FNMS(KP038632954, Tj, Tu); TG = FNMS(KP612264650, Tw, Tx); Ty = FMA(KP612264650, Tx, Tw); { E TT, Tz, TK, TH, TM, T10, TL, TJ; TT = FNMS(KP083333333, TS, TN); Tz = FNMS(KP853480001, Ty, Tv); TK = FMA(KP853480001, Ty, Tv); TH = FNMS(KP853480001, TG, TF); TM = FMA(KP853480001, TG, TF); T1a = FNMS(KP226109445, TW, TZ); T10 = FMA(KP301479260, TZ, TW); TL = FNMS(KP522026385, Tz, TE); Ci[WS(csi, 1)] = KP600477271 * (FMA(KP957805992, TE, Tz)); TJ = FMA(KP522026385, TH, TI); Ci[WS(csi, 5)] = -(KP600477271 * (FNMS(KP957805992, TI, TH))); Ci[WS(csi, 4)] = -(KP575140729 * (FMA(KP904176221, TM, TL))); Ci[WS(csi, 3)] = KP575140729 * (FNMS(KP904176221, TM, TL)); Ci[WS(csi, 6)] = KP575140729 * (FMA(KP904176221, TK, TJ)); Ci[WS(csi, 2)] = KP575140729 * (FNMS(KP904176221, TK, TJ)); T11 = FMA(KP503537032, T10, TT); T17 = FNMS(KP251768516, T10, TT); } T14 = FNMS(KP514918778, T13, T12); T1b = FMA(KP686558370, T12, T13); } } } { E T1e, T1c, T18, T16, T1d, T19; T1e = FMA(KP769338817, T1b, T1a); T1c = FNMS(KP769338817, T1b, T1a); T18 = FNMS(KP859542535, T14, T15); T16 = FMA(KP581704778, T15, T14); T1d = FNMS(KP300462606, T18, T17); T19 = FMA(KP300462606, T18, T17); Cr[WS(csr, 1)] = FMA(KP516520780, T16, T11); Cr[WS(csr, 5)] = FNMS(KP516520780, T16, T11); Cr[WS(csr, 2)] = FMA(KP503537032, T1e, T1d); Cr[WS(csr, 6)] = FNMS(KP503537032, T1e, T1d); Cr[WS(csr, 3)] = FMA(KP503537032, T1c, T19); Cr[WS(csr, 4)] = FNMS(KP503537032, T1c, T19); } } } } static const kr2c_desc desc = { 13, "r2cf_13", {31, 6, 45, 0}, &GENUS }; void X(codelet_r2cf_13) (planner *p) { X(kr2c_register) (p, r2cf_13, &desc); } #else /* HAVE_FMA */ /* Generated by: ../../../genfft/gen_r2cf.native -compact -variables 4 -pipeline-latency 4 -n 13 -name r2cf_13 -include r2cf.h */ /* * This function contains 76 FP additions, 34 FP multiplications, * (or, 57 additions, 15 multiplications, 19 fused multiply/add), * 55 stack variables, 20 constants, and 26 memory accesses */ #include "r2cf.h" static void r2cf_13(R *R0, R *R1, R *Cr, R *Ci, stride rs, stride csr, stride csi, INT v, INT ivs, INT ovs) { DK(KP083333333, +0.083333333333333333333333333333333333333333333); DK(KP075902986, +0.075902986037193865983102897245103540356428373); DK(KP251768516, +0.251768516431883313623436926934233488546674281); DK(KP503537032, +0.503537032863766627246873853868466977093348562); DK(KP113854479, +0.113854479055790798974654345867655310534642560); DK(KP265966249, +0.265966249214837287587521063842185948798330267); DK(KP387390585, +0.387390585467617292130675966426762851778775217); DK(KP300462606, +0.300462606288665774426601772289207995520941381); DK(KP132983124, +0.132983124607418643793760531921092974399165133); DK(KP258260390, +0.258260390311744861420450644284508567852516811); DK(KP2_000000000, +2.000000000000000000000000000000000000000000000); DK(KP1_732050807, +1.732050807568877293527446341505872366942805254); DK(KP300238635, +0.300238635966332641462884626667381504676006424); DK(KP011599105, +0.011599105605768290721655456654083252189827041); DK(KP156891391, +0.156891391051584611046832726756003269660212636); DK(KP256247671, +0.256247671582936600958684654061725059144125175); DK(KP174138601, +0.174138601152135905005660794929264742616964676); DK(KP575140729, +0.575140729474003121368385547455453388461001608); DK(KP866025403, +0.866025403784438646763723170752936183471402627); DK(KP500000000, +0.500000000000000000000000000000000000000000000); { INT i; for (i = v; i > 0; i = i - 1, R0 = R0 + ivs, R1 = R1 + ivs, Cr = Cr + ovs, Ci = Ci + ovs, MAKE_VOLATILE_STRIDE(52, rs), MAKE_VOLATILE_STRIDE(52, csr), MAKE_VOLATILE_STRIDE(52, csi)) { E T13, Tb, Tm, TW, TX, T14, TU, T10, Tz, TB, Tu, TC, TR, T11; T13 = R0[0]; { E Te, TO, Ta, Tv, To, T5, Tw, Tp, Th, Tr, Tk, Ts, Tl, TP, Tc; E Td; Tc = R0[WS(rs, 4)]; Td = R1[WS(rs, 2)]; Te = Tc - Td; TO = Tc + Td; { E T6, T7, T8, T9; T6 = R1[0]; T7 = R1[WS(rs, 1)]; T8 = R1[WS(rs, 4)]; T9 = T7 + T8; Ta = T6 + T9; Tv = T7 - T8; To = FNMS(KP500000000, T9, T6); } { E T1, T2, T3, T4; T1 = R0[WS(rs, 6)]; T2 = R0[WS(rs, 5)]; T3 = R0[WS(rs, 2)]; T4 = T2 + T3; T5 = T1 + T4; Tw = T2 - T3; Tp = FNMS(KP500000000, T4, T1); } { E Tf, Tg, Ti, Tj; Tf = R1[WS(rs, 5)]; Tg = R0[WS(rs, 3)]; Th = Tf - Tg; Tr = Tf + Tg; Ti = R1[WS(rs, 3)]; Tj = R0[WS(rs, 1)]; Tk = Ti - Tj; Ts = Ti + Tj; } Tl = Th + Tk; TP = Tr + Ts; Tb = T5 - Ta; Tm = Te + Tl; TW = Ta + T5; TX = TO + TP; T14 = TW + TX; { E TS, TT, Tx, Ty; TS = Tv + Tw; TT = Th - Tk; TU = TS - TT; T10 = TS + TT; Tx = KP866025403 * (Tv - Tw); Ty = FNMS(KP500000000, Tl, Te); Tz = Tx + Ty; TB = Ty - Tx; } { E Tq, Tt, TN, TQ; Tq = To - Tp; Tt = KP866025403 * (Tr - Ts); Tu = Tq - Tt; TC = Tq + Tt; TN = To + Tp; TQ = FNMS(KP500000000, TP, TO); TR = TN - TQ; T11 = TN + TQ; } } Cr[0] = T13 + T14; { E Tn, TG, TE, TF, TJ, TM, TK, TL; Tn = FNMS(KP174138601, Tm, KP575140729 * Tb); TG = FMA(KP174138601, Tb, KP575140729 * Tm); { E TA, TD, TH, TI; TA = FNMS(KP156891391, Tz, KP256247671 * Tu); TD = FNMS(KP300238635, TC, KP011599105 * TB); TE = TA + TD; TF = KP1_732050807 * (TD - TA); TH = FMA(KP300238635, TB, KP011599105 * TC); TI = FMA(KP256247671, Tz, KP156891391 * Tu); TJ = TH - TI; TM = KP1_732050807 * (TI + TH); } Ci[WS(csi, 5)] = FMA(KP2_000000000, TE, Tn); Ci[WS(csi, 1)] = FMA(KP2_000000000, TJ, TG); TK = TG - TJ; Ci[WS(csi, 4)] = TF - TK; Ci[WS(csi, 3)] = TF + TK; TL = Tn - TE; Ci[WS(csi, 2)] = TL - TM; Ci[WS(csi, 6)] = TL + TM; } { E TZ, T1b, T19, T1e, T16, T1a, TV, TY, T1c, T1d; TV = FNMS(KP132983124, TU, KP258260390 * TR); TY = KP300462606 * (TW - TX); TZ = FMA(KP2_000000000, TV, TY); T1b = TY - TV; { E T17, T18, T12, T15; T17 = FMA(KP387390585, TU, KP265966249 * TR); T18 = FNMS(KP503537032, T11, KP113854479 * T10); T19 = T17 - T18; T1e = T17 + T18; T12 = FMA(KP251768516, T10, KP075902986 * T11); T15 = FNMS(KP083333333, T14, T13); T16 = FMA(KP2_000000000, T12, T15); T1a = T15 - T12; } Cr[WS(csr, 1)] = TZ + T16; Cr[WS(csr, 5)] = T16 - TZ; T1c = T1a - T1b; Cr[WS(csr, 2)] = T19 + T1c; Cr[WS(csr, 6)] = T1c - T19; T1d = T1b + T1a; Cr[WS(csr, 3)] = T1d - T1e; Cr[WS(csr, 4)] = T1e + T1d; } } } } static const kr2c_desc desc = { 13, "r2cf_13", {57, 15, 19, 0}, &GENUS }; void X(codelet_r2cf_13) (planner *p) { X(kr2c_register) (p, r2cf_13, &desc); } #endif /* HAVE_FMA */ fftw-3.3.3/rdft/scalar/r2cf/r2cf_14.c0000644000175000017500000002117512054410623013756 00000000000000/* * Copyright (c) 2003, 2007-11 Matteo Frigo * Copyright (c) 2003, 2007-11 Massachusetts Institute of Technology * * This program is free software; you can redistribute it and/or modify * it under the terms of the GNU General Public License as published by * the Free Software Foundation; either version 2 of the License, or * (at your option) any later version. * * This program is distributed in the hope that it will be useful, * but WITHOUT ANY WARRANTY; without even the implied warranty of * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the * GNU General Public License for more details. * * You should have received a copy of the GNU General Public License * along with this program; if not, write to the Free Software * Foundation, Inc., 51 Franklin Street, Fifth Floor, Boston, MA 02110-1301 USA * */ /* This file was automatically generated --- DO NOT EDIT */ /* Generated on Sun Nov 25 07:39:46 EST 2012 */ #include "codelet-rdft.h" #ifdef HAVE_FMA /* Generated by: ../../../genfft/gen_r2cf.native -fma -reorder-insns -schedule-for-pipeline -compact -variables 4 -pipeline-latency 4 -n 14 -name r2cf_14 -include r2cf.h */ /* * This function contains 62 FP additions, 36 FP multiplications, * (or, 32 additions, 6 multiplications, 30 fused multiply/add), * 45 stack variables, 6 constants, and 28 memory accesses */ #include "r2cf.h" static void r2cf_14(R *R0, R *R1, R *Cr, R *Ci, stride rs, stride csr, stride csi, INT v, INT ivs, INT ovs) { DK(KP900968867, +0.900968867902419126236102319507445051165919162); DK(KP692021471, +0.692021471630095869627814897002069140197260599); DK(KP801937735, +0.801937735804838252472204639014890102331838324); DK(KP974927912, +0.974927912181823607018131682993931217232785801); DK(KP356895867, +0.356895867892209443894399510021300583399127187); DK(KP554958132, +0.554958132087371191422194871006410481067288862); { INT i; for (i = v; i > 0; i = i - 1, R0 = R0 + ivs, R1 = R1 + ivs, Cr = Cr + ovs, Ci = Ci + ovs, MAKE_VOLATILE_STRIDE(56, rs), MAKE_VOLATILE_STRIDE(56, csr), MAKE_VOLATILE_STRIDE(56, csi)) { E TN, T3, TG, TQ, Tx, To, TH, Td, TD, TO, Tw, Ta, TL, Ty, TT; E TI, Tg, Tr, Te, Tf, TP, TJ; { E Tl, TE, Tk, Tm; { E T1, T2, Ti, Tj; T1 = R0[0]; T2 = R1[WS(rs, 3)]; Ti = R0[WS(rs, 3)]; Tj = R1[WS(rs, 6)]; Tl = R0[WS(rs, 4)]; TN = T1 + T2; T3 = T1 - T2; TE = Ti + Tj; Tk = Ti - Tj; Tm = R1[0]; } { E T7, TC, T6, T8; { E T4, T5, TF, Tn; T4 = R0[WS(rs, 1)]; T5 = R1[WS(rs, 4)]; T7 = R0[WS(rs, 6)]; TF = Tl + Tm; Tn = Tl - Tm; TC = T4 + T5; T6 = T4 - T5; TG = TE - TF; TQ = TE + TF; Tx = Tn - Tk; To = Tk + Tn; T8 = R1[WS(rs, 2)]; } { E Tb, Tc, TB, T9; Tb = R0[WS(rs, 2)]; Tc = R1[WS(rs, 5)]; Te = R0[WS(rs, 5)]; TB = T7 + T8; T9 = T7 - T8; TH = Tb + Tc; Td = Tb - Tc; TD = TB - TC; TO = TC + TB; Tw = T6 - T9; Ta = T6 + T9; Tf = R1[WS(rs, 1)]; } } } TL = FNMS(KP554958132, TG, TD); Ty = FNMS(KP554958132, Tx, Tw); TT = FNMS(KP356895867, TO, TQ); TI = Te + Tf; Tg = Te - Tf; Tr = FNMS(KP356895867, Ta, To); TP = TH + TI; TJ = TH - TI; { E Th, Tv, TK, TM; Th = Td + Tg; Tv = Tg - Td; TK = FMA(KP554958132, TJ, TG); TM = FMA(KP554958132, TD, TJ); Ci[WS(csi, 6)] = KP974927912 * (FNMS(KP801937735, TL, TJ)); { E TR, TV, TU, Tz; TR = FNMS(KP356895867, TQ, TP); TV = FNMS(KP356895867, TP, TO); TU = FNMS(KP692021471, TT, TP); Cr[0] = TN + TO + TP + TQ; Tz = FMA(KP554958132, Tv, Tx); Ci[WS(csi, 1)] = KP974927912 * (FNMS(KP801937735, Ty, Tv)); { E TA, Ts, Tt, Tp; TA = FMA(KP554958132, Tw, Tv); Ts = FNMS(KP692021471, Tr, Th); Tt = FNMS(KP356895867, Th, Ta); Tp = FNMS(KP356895867, To, Th); Cr[WS(csr, 7)] = T3 + Ta + Th + To; Ci[WS(csi, 2)] = KP974927912 * (FMA(KP801937735, TK, TD)); Ci[WS(csi, 4)] = KP974927912 * (FNMS(KP801937735, TM, TG)); { E TS, TW, Tu, Tq; TS = FNMS(KP692021471, TR, TO); TW = FNMS(KP692021471, TV, TQ); Cr[WS(csr, 2)] = FNMS(KP900968867, TU, TN); Ci[WS(csi, 5)] = KP974927912 * (FMA(KP801937735, Tz, Tw)); Ci[WS(csi, 3)] = KP974927912 * (FNMS(KP801937735, TA, Tx)); Cr[WS(csr, 5)] = FNMS(KP900968867, Ts, T3); Tu = FNMS(KP692021471, Tt, To); Tq = FNMS(KP692021471, Tp, Ta); Cr[WS(csr, 4)] = FNMS(KP900968867, TS, TN); Cr[WS(csr, 6)] = FNMS(KP900968867, TW, TN); Cr[WS(csr, 1)] = FNMS(KP900968867, Tu, T3); Cr[WS(csr, 3)] = FNMS(KP900968867, Tq, T3); } } } } } } } static const kr2c_desc desc = { 14, "r2cf_14", {32, 6, 30, 0}, &GENUS }; void X(codelet_r2cf_14) (planner *p) { X(kr2c_register) (p, r2cf_14, &desc); } #else /* HAVE_FMA */ /* Generated by: ../../../genfft/gen_r2cf.native -compact -variables 4 -pipeline-latency 4 -n 14 -name r2cf_14 -include r2cf.h */ /* * This function contains 62 FP additions, 36 FP multiplications, * (or, 38 additions, 12 multiplications, 24 fused multiply/add), * 29 stack variables, 6 constants, and 28 memory accesses */ #include "r2cf.h" static void r2cf_14(R *R0, R *R1, R *Cr, R *Ci, stride rs, stride csr, stride csi, INT v, INT ivs, INT ovs) { DK(KP900968867, +0.900968867902419126236102319507445051165919162); DK(KP222520933, +0.222520933956314404288902564496794759466355569); DK(KP623489801, +0.623489801858733530525004884004239810632274731); DK(KP433883739, +0.433883739117558120475768332848358754609990728); DK(KP974927912, +0.974927912181823607018131682993931217232785801); DK(KP781831482, +0.781831482468029808708444526674057750232334519); { INT i; for (i = v; i > 0; i = i - 1, R0 = R0 + ivs, R1 = R1 + ivs, Cr = Cr + ovs, Ci = Ci + ovs, MAKE_VOLATILE_STRIDE(56, rs), MAKE_VOLATILE_STRIDE(56, csr), MAKE_VOLATILE_STRIDE(56, csi)) { E T3, TB, T6, Tv, Tn, Ts, Tk, Tt, Td, Ty, T9, Tw, Tg, Tz, T1; E T2; T1 = R0[0]; T2 = R1[WS(rs, 3)]; T3 = T1 - T2; TB = T1 + T2; { E T4, T5, Tl, Tm; T4 = R0[WS(rs, 2)]; T5 = R1[WS(rs, 5)]; T6 = T4 - T5; Tv = T4 + T5; Tl = R0[WS(rs, 6)]; Tm = R1[WS(rs, 2)]; Tn = Tl - Tm; Ts = Tl + Tm; } { E Ti, Tj, Tb, Tc; Ti = R0[WS(rs, 1)]; Tj = R1[WS(rs, 4)]; Tk = Ti - Tj; Tt = Ti + Tj; Tb = R0[WS(rs, 3)]; Tc = R1[WS(rs, 6)]; Td = Tb - Tc; Ty = Tb + Tc; } { E T7, T8, Te, Tf; T7 = R0[WS(rs, 5)]; T8 = R1[WS(rs, 1)]; T9 = T7 - T8; Tw = T7 + T8; Te = R0[WS(rs, 4)]; Tf = R1[0]; Tg = Te - Tf; Tz = Te + Tf; } { E Tp, Tr, Tq, Ta, To, Th; Tp = Tn - Tk; Tr = Tg - Td; Tq = T9 - T6; Ci[WS(csi, 1)] = FMA(KP781831482, Tp, KP974927912 * Tq) + (KP433883739 * Tr); Ci[WS(csi, 5)] = FMA(KP433883739, Tq, KP781831482 * Tr) - (KP974927912 * Tp); Ci[WS(csi, 3)] = FMA(KP433883739, Tp, KP974927912 * Tr) - (KP781831482 * Tq); Ta = T6 + T9; To = Tk + Tn; Th = Td + Tg; Cr[WS(csr, 3)] = FMA(KP623489801, Ta, T3) + FNMA(KP222520933, Th, KP900968867 * To); Cr[WS(csr, 7)] = T3 + To + Ta + Th; Cr[WS(csr, 1)] = FMA(KP623489801, To, T3) + FNMA(KP900968867, Th, KP222520933 * Ta); Cr[WS(csr, 5)] = FMA(KP623489801, Th, T3) + FNMA(KP900968867, Ta, KP222520933 * To); } { E Tu, TA, Tx, TC, TE, TD; Tu = Ts - Tt; TA = Ty - Tz; Tx = Tv - Tw; Ci[WS(csi, 2)] = FMA(KP974927912, Tu, KP433883739 * Tx) + (KP781831482 * TA); Ci[WS(csi, 6)] = FMA(KP974927912, Tx, KP433883739 * TA) - (KP781831482 * Tu); Ci[WS(csi, 4)] = FNMS(KP781831482, Tx, KP974927912 * TA) - (KP433883739 * Tu); TC = Tt + Ts; TE = Tv + Tw; TD = Ty + Tz; Cr[WS(csr, 6)] = FMA(KP623489801, TC, TB) + FNMA(KP900968867, TD, KP222520933 * TE); Cr[WS(csr, 2)] = FMA(KP623489801, TD, TB) + FNMA(KP900968867, TE, KP222520933 * TC); Cr[WS(csr, 4)] = FMA(KP623489801, TE, TB) + FNMA(KP222520933, TD, KP900968867 * TC); Cr[0] = TB + TC + TE + TD; } } } } static const kr2c_desc desc = { 14, "r2cf_14", {38, 12, 24, 0}, &GENUS }; void X(codelet_r2cf_14) (planner *p) { X(kr2c_register) (p, r2cf_14, &desc); } #endif /* HAVE_FMA */ fftw-3.3.3/rdft/scalar/r2cf/r2cf_15.c0000644000175000017500000002307612054410623013761 00000000000000/* * Copyright (c) 2003, 2007-11 Matteo Frigo * Copyright (c) 2003, 2007-11 Massachusetts Institute of Technology * * This program is free software; you can redistribute it and/or modify * it under the terms of the GNU General Public License as published by * the Free Software Foundation; either version 2 of the License, or * (at your option) any later version. * * This program is distributed in the hope that it will be useful, * but WITHOUT ANY WARRANTY; without even the implied warranty of * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the * GNU General Public License for more details. * * You should have received a copy of the GNU General Public License * along with this program; if not, write to the Free Software * Foundation, Inc., 51 Franklin Street, Fifth Floor, Boston, MA 02110-1301 USA * */ /* This file was automatically generated --- DO NOT EDIT */ /* Generated on Sun Nov 25 07:39:46 EST 2012 */ #include "codelet-rdft.h" #ifdef HAVE_FMA /* Generated by: ../../../genfft/gen_r2cf.native -fma -reorder-insns -schedule-for-pipeline -compact -variables 4 -pipeline-latency 4 -n 15 -name r2cf_15 -include r2cf.h */ /* * This function contains 64 FP additions, 35 FP multiplications, * (or, 36 additions, 7 multiplications, 28 fused multiply/add), * 50 stack variables, 8 constants, and 30 memory accesses */ #include "r2cf.h" static void r2cf_15(R *R0, R *R1, R *Cr, R *Ci, stride rs, stride csr, stride csi, INT v, INT ivs, INT ovs) { DK(KP910592997, +0.910592997310029334643087372129977886038870291); DK(KP823639103, +0.823639103546331925877420039278190003029660514); DK(KP559016994, +0.559016994374947424102293417182819058860154590); DK(KP951056516, +0.951056516295153572116439333379382143405698634); DK(KP250000000, +0.250000000000000000000000000000000000000000000); DK(KP618033988, +0.618033988749894848204586834365638117720309180); DK(KP866025403, +0.866025403784438646763723170752936183471402627); DK(KP500000000, +0.500000000000000000000000000000000000000000000); { INT i; for (i = v; i > 0; i = i - 1, R0 = R0 + ivs, R1 = R1 + ivs, Cr = Cr + ovs, Ci = Ci + ovs, MAKE_VOLATILE_STRIDE(60, rs), MAKE_VOLATILE_STRIDE(60, csr), MAKE_VOLATILE_STRIDE(60, csi)) { E Tw, Tz, Tp, Ty; { E Ti, TF, TR, TN, TX, T11, TM, TS, Tl, TH, Tf, To, TT, TD, Tg; E Th; TD = R0[0]; Tg = R0[WS(rs, 5)]; Th = R1[WS(rs, 2)]; { E Tj, Tq, Tt, Tm, T3, Tk, T4, Ta, Tr, Td, Tu, T5, TE; Tj = R1[WS(rs, 1)]; Tq = R0[WS(rs, 3)]; Tt = R1[WS(rs, 4)]; TE = Th + Tg; Ti = Tg - Th; Tm = R0[WS(rs, 6)]; { E T8, T9, T1, T2, Tb, Tc; T1 = R0[WS(rs, 4)]; T2 = R1[WS(rs, 6)]; TF = FNMS(KP500000000, TE, TD); TR = TD + TE; T8 = R1[WS(rs, 5)]; T3 = T1 - T2; Tk = T1 + T2; T9 = R1[0]; Tb = R0[WS(rs, 7)]; Tc = R0[WS(rs, 2)]; T4 = R0[WS(rs, 1)]; Ta = T8 - T9; Tr = T8 + T9; Td = Tb - Tc; Tu = Tb + Tc; T5 = R1[WS(rs, 3)]; } { E Ts, Tv, Te, Tn, T7, T6, TV, TW; TV = Tq + Tr; Ts = FNMS(KP500000000, Tr, Tq); Tv = FNMS(KP500000000, Tu, Tt); TW = Tt + Tu; Te = Ta + Td; TN = Td - Ta; Tn = T4 + T5; T6 = T4 - T5; TX = TV + TW; T11 = TW - TV; TM = T6 - T3; T7 = T3 + T6; TS = Tj + Tk; Tl = FNMS(KP500000000, Tk, Tj); TH = Ts + Tv; Tw = Ts - Tv; Tz = Te - T7; Tf = T7 + Te; To = FNMS(KP500000000, Tn, Tm); TT = Tm + Tn; } } { E TO, TQ, TU, T12, TK, TI, TG; Ci[WS(csi, 5)] = KP866025403 * (Tf - Ti); TG = Tl + To; Tp = Tl - To; TO = FMA(KP618033988, TN, TM); TQ = FNMS(KP618033988, TM, TN); TU = TS + TT; T12 = TS - TT; TK = TG - TH; TI = TG + TH; { E T10, TY, TL, TP, TJ, TZ; T10 = TU - TX; TY = TU + TX; Cr[WS(csr, 5)] = TF + TI; TJ = FNMS(KP250000000, TI, TF); Ci[WS(csi, 6)] = -(KP951056516 * (FNMS(KP618033988, T11, T12))); Ci[WS(csi, 3)] = KP951056516 * (FMA(KP618033988, T12, T11)); TL = FMA(KP559016994, TK, TJ); TP = FNMS(KP559016994, TK, TJ); Cr[0] = TR + TY; TZ = FNMS(KP250000000, TY, TR); Cr[WS(csr, 4)] = FNMS(KP823639103, TO, TL); Cr[WS(csr, 1)] = FMA(KP823639103, TO, TL); Cr[WS(csr, 7)] = FNMS(KP823639103, TQ, TP); Cr[WS(csr, 2)] = FMA(KP823639103, TQ, TP); Cr[WS(csr, 6)] = FMA(KP559016994, T10, TZ); Cr[WS(csr, 3)] = FNMS(KP559016994, T10, TZ); Ty = FMA(KP250000000, Tf, Ti); } } } { E TB, Tx, TC, TA; TB = FNMS(KP618033988, Tp, Tw); Tx = FMA(KP618033988, Tw, Tp); TC = FNMS(KP559016994, Tz, Ty); TA = FMA(KP559016994, Tz, Ty); Ci[WS(csi, 2)] = KP951056516 * (FNMS(KP910592997, TC, TB)); Ci[WS(csi, 7)] = KP951056516 * (FMA(KP910592997, TC, TB)); Ci[WS(csi, 4)] = KP951056516 * (FMA(KP910592997, TA, Tx)); Ci[WS(csi, 1)] = -(KP951056516 * (FNMS(KP910592997, TA, Tx))); } } } } static const kr2c_desc desc = { 15, "r2cf_15", {36, 7, 28, 0}, &GENUS }; void X(codelet_r2cf_15) (planner *p) { X(kr2c_register) (p, r2cf_15, &desc); } #else /* HAVE_FMA */ /* Generated by: ../../../genfft/gen_r2cf.native -compact -variables 4 -pipeline-latency 4 -n 15 -name r2cf_15 -include r2cf.h */ /* * This function contains 64 FP additions, 25 FP multiplications, * (or, 50 additions, 11 multiplications, 14 fused multiply/add), * 47 stack variables, 10 constants, and 30 memory accesses */ #include "r2cf.h" static void r2cf_15(R *R0, R *R1, R *Cr, R *Ci, stride rs, stride csr, stride csi, INT v, INT ivs, INT ovs) { DK(KP484122918, +0.484122918275927110647408174972799951354115213); DK(KP216506350, +0.216506350946109661690930792688234045867850657); DK(KP951056516, +0.951056516295153572116439333379382143405698634); DK(KP587785252, +0.587785252292473129168705954639072768597652438); DK(KP250000000, +0.250000000000000000000000000000000000000000000); DK(KP559016994, +0.559016994374947424102293417182819058860154590); DK(KP509036960, +0.509036960455127183450980863393907648510733164); DK(KP823639103, +0.823639103546331925877420039278190003029660514); DK(KP866025403, +0.866025403784438646763723170752936183471402627); DK(KP500000000, +0.500000000000000000000000000000000000000000000); { INT i; for (i = v; i > 0; i = i - 1, R0 = R0 + ivs, R1 = R1 + ivs, Cr = Cr + ovs, Ci = Ci + ovs, MAKE_VOLATILE_STRIDE(60, rs), MAKE_VOLATILE_STRIDE(60, csr), MAKE_VOLATILE_STRIDE(60, csi)) { E Ti, TR, TL, TD, TE, T7, Te, Tf, TV, TW, TX, Tv, Ty, TH, To; E Tr, TG, TS, TT, TU; { E TJ, Tg, Th, TK; TJ = R0[0]; Tg = R0[WS(rs, 5)]; Th = R1[WS(rs, 2)]; TK = Th + Tg; Ti = Tg - Th; TR = TJ + TK; TL = FNMS(KP500000000, TK, TJ); } { E Tm, Tt, Tw, Tp, T3, Tx, Ta, Tn, Td, Tq, T6, Tu; Tm = R1[WS(rs, 1)]; Tt = R0[WS(rs, 3)]; Tw = R1[WS(rs, 4)]; Tp = R0[WS(rs, 6)]; { E T1, T2, T8, T9; T1 = R0[WS(rs, 7)]; T2 = R0[WS(rs, 2)]; T3 = T1 - T2; Tx = T1 + T2; T8 = R1[WS(rs, 6)]; T9 = R0[WS(rs, 4)]; Ta = T8 - T9; Tn = T9 + T8; } { E Tb, Tc, T4, T5; Tb = R1[WS(rs, 3)]; Tc = R0[WS(rs, 1)]; Td = Tb - Tc; Tq = Tc + Tb; T4 = R1[0]; T5 = R1[WS(rs, 5)]; T6 = T4 - T5; Tu = T5 + T4; } TD = Ta - Td; TE = T6 + T3; T7 = T3 - T6; Te = Ta + Td; Tf = T7 - Te; TV = Tt + Tu; TW = Tw + Tx; TX = TV + TW; Tv = FNMS(KP500000000, Tu, Tt); Ty = FNMS(KP500000000, Tx, Tw); TH = Tv + Ty; To = FNMS(KP500000000, Tn, Tm); Tr = FNMS(KP500000000, Tq, Tp); TG = To + Tr; TS = Tm + Tn; TT = Tp + Tq; TU = TS + TT; } Ci[WS(csi, 5)] = KP866025403 * (Tf - Ti); { E TF, TP, TI, TM, TN, TQ, TO; TF = FMA(KP823639103, TD, KP509036960 * TE); TP = FNMS(KP509036960, TD, KP823639103 * TE); TI = KP559016994 * (TG - TH); TM = TG + TH; TN = FNMS(KP250000000, TM, TL); Cr[WS(csr, 5)] = TL + TM; TQ = TN - TI; Cr[WS(csr, 2)] = TP + TQ; Cr[WS(csr, 7)] = TQ - TP; TO = TI + TN; Cr[WS(csr, 1)] = TF + TO; Cr[WS(csr, 4)] = TO - TF; } { E T11, T12, T10, TY, TZ; T11 = TS - TT; T12 = TW - TV; Ci[WS(csi, 3)] = FMA(KP587785252, T11, KP951056516 * T12); Ci[WS(csi, 6)] = FNMS(KP951056516, T11, KP587785252 * T12); T10 = KP559016994 * (TU - TX); TY = TU + TX; TZ = FNMS(KP250000000, TY, TR); Cr[WS(csr, 3)] = TZ - T10; Cr[0] = TR + TY; Cr[WS(csr, 6)] = T10 + TZ; { E Tl, TB, TA, TC; { E Tj, Tk, Ts, Tz; Tj = FMA(KP866025403, Ti, KP216506350 * Tf); Tk = KP484122918 * (Te + T7); Tl = Tj + Tk; TB = Tk - Tj; Ts = To - Tr; Tz = Tv - Ty; TA = FMA(KP951056516, Ts, KP587785252 * Tz); TC = FNMS(KP587785252, Ts, KP951056516 * Tz); } Ci[WS(csi, 1)] = Tl - TA; Ci[WS(csi, 7)] = TC - TB; Ci[WS(csi, 4)] = Tl + TA; Ci[WS(csi, 2)] = TB + TC; } } } } } static const kr2c_desc desc = { 15, "r2cf_15", {50, 11, 14, 0}, &GENUS }; void X(codelet_r2cf_15) (planner *p) { X(kr2c_register) (p, r2cf_15, &desc); } #endif /* HAVE_FMA */ fftw-3.3.3/rdft/scalar/r2cf/r2cf_16.c0000644000175000017500000002010312054410623013746 00000000000000/* * Copyright (c) 2003, 2007-11 Matteo Frigo * Copyright (c) 2003, 2007-11 Massachusetts Institute of Technology * * This program is free software; you can redistribute it and/or modify * it under the terms of the GNU General Public License as published by * the Free Software Foundation; either version 2 of the License, or * (at your option) any later version. * * This program is distributed in the hope that it will be useful, * but WITHOUT ANY WARRANTY; without even the implied warranty of * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the * GNU General Public License for more details. * * You should have received a copy of the GNU General Public License * along with this program; if not, write to the Free Software * Foundation, Inc., 51 Franklin Street, Fifth Floor, Boston, MA 02110-1301 USA * */ /* This file was automatically generated --- DO NOT EDIT */ /* Generated on Sun Nov 25 07:39:47 EST 2012 */ #include "codelet-rdft.h" #ifdef HAVE_FMA /* Generated by: ../../../genfft/gen_r2cf.native -fma -reorder-insns -schedule-for-pipeline -compact -variables 4 -pipeline-latency 4 -n 16 -name r2cf_16 -include r2cf.h */ /* * This function contains 58 FP additions, 20 FP multiplications, * (or, 38 additions, 0 multiplications, 20 fused multiply/add), * 38 stack variables, 3 constants, and 32 memory accesses */ #include "r2cf.h" static void r2cf_16(R *R0, R *R1, R *Cr, R *Ci, stride rs, stride csr, stride csi, INT v, INT ivs, INT ovs) { DK(KP923879532, +0.923879532511286756128183189396788286822416626); DK(KP707106781, +0.707106781186547524400844362104849039284835938); DK(KP414213562, +0.414213562373095048801688724209698078569671875); { INT i; for (i = v; i > 0; i = i - 1, R0 = R0 + ivs, R1 = R1 + ivs, Cr = Cr + ovs, Ci = Ci + ovs, MAKE_VOLATILE_STRIDE(64, rs), MAKE_VOLATILE_STRIDE(64, csr), MAKE_VOLATILE_STRIDE(64, csi)) { E TQ, TP; { E TB, TN, Tf, T7, Te, Tv, TO, TE, Tq, TJ, Tp, TI, TT, Ty, Tm; E Tr, TK, Ts; { E TC, Ta, Td, TD; { E T1, T2, T4, T5; T1 = R0[0]; T2 = R0[WS(rs, 4)]; T4 = R0[WS(rs, 2)]; T5 = R0[WS(rs, 6)]; { E T8, T3, T6, T9, Tb, Tc; T8 = R0[WS(rs, 1)]; TB = T1 - T2; T3 = T1 + T2; TN = T4 - T5; T6 = T4 + T5; T9 = R0[WS(rs, 5)]; Tb = R0[WS(rs, 7)]; Tc = R0[WS(rs, 3)]; Tf = T3 - T6; T7 = T3 + T6; TC = T8 - T9; Ta = T8 + T9; Td = Tb + Tc; TD = Tb - Tc; } } { E TG, Ti, Tj, Tk, Tg, Th; Tg = R1[0]; Th = R1[WS(rs, 4)]; Te = Ta + Td; Tv = Td - Ta; TO = TD - TC; TE = TC + TD; TG = Tg - Th; Ti = Tg + Th; Tj = R1[WS(rs, 2)]; Tk = R1[WS(rs, 6)]; { E Tn, To, TH, Tl; Tn = R1[WS(rs, 7)]; To = R1[WS(rs, 3)]; Tq = R1[WS(rs, 1)]; TH = Tj - Tk; Tl = Tj + Tk; TJ = Tn - To; Tp = Tn + To; TI = FNMS(KP414213562, TH, TG); TT = FMA(KP414213562, TG, TH); Ty = Ti + Tl; Tm = Ti - Tl; Tr = R1[WS(rs, 5)]; } } } Cr[WS(csr, 4)] = T7 - Te; TK = Tr - Tq; Ts = Tq + Tr; { E Tx, TV, TF, TS, Tz, Tt, TM, TL; Tx = T7 + Te; TV = FNMS(KP707106781, TE, TB); TF = FMA(KP707106781, TE, TB); TL = FNMS(KP414213562, TK, TJ); TS = FMA(KP414213562, TJ, TK); Tz = Tp + Ts; Tt = Tp - Ts; TM = TI + TL; TQ = TL - TI; { E TR, TU, TW, TA, Tw, Tu; TP = FMA(KP707106781, TO, TN); TR = FNMS(KP707106781, TO, TN); TA = Ty + Tz; Ci[WS(csi, 4)] = Tz - Ty; Tw = Tt - Tm; Tu = Tm + Tt; Cr[WS(csr, 1)] = FMA(KP923879532, TM, TF); Cr[WS(csr, 7)] = FNMS(KP923879532, TM, TF); Cr[0] = Tx + TA; Cr[WS(csr, 8)] = Tx - TA; Ci[WS(csi, 6)] = FMS(KP707106781, Tw, Tv); Ci[WS(csi, 2)] = FMA(KP707106781, Tw, Tv); Cr[WS(csr, 2)] = FMA(KP707106781, Tu, Tf); Cr[WS(csr, 6)] = FNMS(KP707106781, Tu, Tf); TU = TS - TT; TW = TT + TS; Ci[WS(csi, 7)] = FMA(KP923879532, TU, TR); Ci[WS(csi, 1)] = FMS(KP923879532, TU, TR); Cr[WS(csr, 3)] = FMA(KP923879532, TW, TV); Cr[WS(csr, 5)] = FNMS(KP923879532, TW, TV); } } } Ci[WS(csi, 5)] = FMS(KP923879532, TQ, TP); Ci[WS(csi, 3)] = FMA(KP923879532, TQ, TP); } } } static const kr2c_desc desc = { 16, "r2cf_16", {38, 0, 20, 0}, &GENUS }; void X(codelet_r2cf_16) (planner *p) { X(kr2c_register) (p, r2cf_16, &desc); } #else /* HAVE_FMA */ /* Generated by: ../../../genfft/gen_r2cf.native -compact -variables 4 -pipeline-latency 4 -n 16 -name r2cf_16 -include r2cf.h */ /* * This function contains 58 FP additions, 12 FP multiplications, * (or, 54 additions, 8 multiplications, 4 fused multiply/add), * 34 stack variables, 3 constants, and 32 memory accesses */ #include "r2cf.h" static void r2cf_16(R *R0, R *R1, R *Cr, R *Ci, stride rs, stride csr, stride csi, INT v, INT ivs, INT ovs) { DK(KP923879532, +0.923879532511286756128183189396788286822416626); DK(KP382683432, +0.382683432365089771728459984030398866761344562); DK(KP707106781, +0.707106781186547524400844362104849039284835938); { INT i; for (i = v; i > 0; i = i - 1, R0 = R0 + ivs, R1 = R1 + ivs, Cr = Cr + ovs, Ci = Ci + ovs, MAKE_VOLATILE_STRIDE(64, rs), MAKE_VOLATILE_STRIDE(64, csr), MAKE_VOLATILE_STRIDE(64, csi)) { E T3, T6, T7, Tz, Ti, Ta, Td, Te, TA, Th, Tq, TV, TF, TP, Tx; E TU, TE, TM, Tg, Tf, TJ, TQ; { E T1, T2, T4, T5; T1 = R0[0]; T2 = R0[WS(rs, 4)]; T3 = T1 + T2; T4 = R0[WS(rs, 2)]; T5 = R0[WS(rs, 6)]; T6 = T4 + T5; T7 = T3 + T6; Tz = T1 - T2; Ti = T4 - T5; } { E T8, T9, Tb, Tc; T8 = R0[WS(rs, 1)]; T9 = R0[WS(rs, 5)]; Ta = T8 + T9; Tg = T8 - T9; Tb = R0[WS(rs, 7)]; Tc = R0[WS(rs, 3)]; Td = Tb + Tc; Tf = Tb - Tc; } Te = Ta + Td; TA = KP707106781 * (Tg + Tf); Th = KP707106781 * (Tf - Tg); { E Tm, TN, Tp, TO; { E Tk, Tl, Tn, To; Tk = R1[WS(rs, 7)]; Tl = R1[WS(rs, 3)]; Tm = Tk - Tl; TN = Tk + Tl; Tn = R1[WS(rs, 1)]; To = R1[WS(rs, 5)]; Tp = Tn - To; TO = Tn + To; } Tq = FNMS(KP923879532, Tp, KP382683432 * Tm); TV = TN + TO; TF = FMA(KP923879532, Tm, KP382683432 * Tp); TP = TN - TO; } { E Tt, TK, Tw, TL; { E Tr, Ts, Tu, Tv; Tr = R1[0]; Ts = R1[WS(rs, 4)]; Tt = Tr - Ts; TK = Tr + Ts; Tu = R1[WS(rs, 2)]; Tv = R1[WS(rs, 6)]; Tw = Tu - Tv; TL = Tu + Tv; } Tx = FMA(KP382683432, Tt, KP923879532 * Tw); TU = TK + TL; TE = FNMS(KP382683432, Tw, KP923879532 * Tt); TM = TK - TL; } Cr[WS(csr, 4)] = T7 - Te; Ci[WS(csi, 4)] = TV - TU; { E Tj, Ty, TD, TG; Tj = Th - Ti; Ty = Tq - Tx; Ci[WS(csi, 1)] = Tj + Ty; Ci[WS(csi, 7)] = Ty - Tj; TD = Tz + TA; TG = TE + TF; Cr[WS(csr, 7)] = TD - TG; Cr[WS(csr, 1)] = TD + TG; } { E TB, TC, TH, TI; TB = Tz - TA; TC = Tx + Tq; Cr[WS(csr, 5)] = TB - TC; Cr[WS(csr, 3)] = TB + TC; TH = Ti + Th; TI = TF - TE; Ci[WS(csi, 3)] = TH + TI; Ci[WS(csi, 5)] = TI - TH; } TJ = T3 - T6; TQ = KP707106781 * (TM + TP); Cr[WS(csr, 6)] = TJ - TQ; Cr[WS(csr, 2)] = TJ + TQ; { E TR, TS, TT, TW; TR = Td - Ta; TS = KP707106781 * (TP - TM); Ci[WS(csi, 2)] = TR + TS; Ci[WS(csi, 6)] = TS - TR; TT = T7 + Te; TW = TU + TV; Cr[WS(csr, 8)] = TT - TW; Cr[0] = TT + TW; } } } } static const kr2c_desc desc = { 16, "r2cf_16", {54, 8, 4, 0}, &GENUS }; void X(codelet_r2cf_16) (planner *p) { X(kr2c_register) (p, r2cf_16, &desc); } #endif /* HAVE_FMA */ fftw-3.3.3/rdft/scalar/r2cf/r2cf_32.c0000644000175000017500000004224312054410625013757 00000000000000/* * Copyright (c) 2003, 2007-11 Matteo Frigo * Copyright (c) 2003, 2007-11 Massachusetts Institute of Technology * * This program is free software; you can redistribute it and/or modify * it under the terms of the GNU General Public License as published by * the Free Software Foundation; either version 2 of the License, or * (at your option) any later version. * * This program is distributed in the hope that it will be useful, * but WITHOUT ANY WARRANTY; without even the implied warranty of * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the * GNU General Public License for more details. * * You should have received a copy of the GNU General Public License * along with this program; if not, write to the Free Software * Foundation, Inc., 51 Franklin Street, Fifth Floor, Boston, MA 02110-1301 USA * */ /* This file was automatically generated --- DO NOT EDIT */ /* Generated on Sun Nov 25 07:39:47 EST 2012 */ #include "codelet-rdft.h" #ifdef HAVE_FMA /* Generated by: ../../../genfft/gen_r2cf.native -fma -reorder-insns -schedule-for-pipeline -compact -variables 4 -pipeline-latency 4 -n 32 -name r2cf_32 -include r2cf.h */ /* * This function contains 156 FP additions, 68 FP multiplications, * (or, 88 additions, 0 multiplications, 68 fused multiply/add), * 89 stack variables, 7 constants, and 64 memory accesses */ #include "r2cf.h" static void r2cf_32(R *R0, R *R1, R *Cr, R *Ci, stride rs, stride csr, stride csi, INT v, INT ivs, INT ovs) { DK(KP831469612, +0.831469612302545237078788377617905756738560812); DK(KP668178637, +0.668178637919298919997757686523080761552472251); DK(KP980785280, +0.980785280403230449126182236134239036973933731); DK(KP198912367, +0.198912367379658006911597622644676228597850501); DK(KP923879532, +0.923879532511286756128183189396788286822416626); DK(KP707106781, +0.707106781186547524400844362104849039284835938); DK(KP414213562, +0.414213562373095048801688724209698078569671875); { INT i; for (i = v; i > 0; i = i - 1, R0 = R0 + ivs, R1 = R1 + ivs, Cr = Cr + ovs, Ci = Ci + ovs, MAKE_VOLATILE_STRIDE(128, rs), MAKE_VOLATILE_STRIDE(128, csr), MAKE_VOLATILE_STRIDE(128, csi)) { E T1x, T1M, T1I, T1E, T1J, T1H; { E Tv, T1h, T7, T2b, Te, T2n, Ty, T1i, T1l, TF, T2d, Tt, T1k, TC, T2c; E Tm, T2j, T1Z, T2k, T22, TK, T1B, T19, T1C, T1e, TO, TV, T1T, TN, TP; E T2g, T1S; { E TD, Tp, Tq, Tr; { E T1, T2, T4, T5; T1 = R0[0]; T2 = R0[WS(rs, 8)]; T4 = R0[WS(rs, 4)]; T5 = R0[WS(rs, 12)]; { E Ta, Tw, Tx, Td, Tn, To; { E T8, T3, T6, T9, Tb, Tc; T8 = R0[WS(rs, 2)]; Tv = T1 - T2; T3 = T1 + T2; T1h = T4 - T5; T6 = T4 + T5; T9 = R0[WS(rs, 10)]; Tb = R0[WS(rs, 14)]; Tc = R0[WS(rs, 6)]; T7 = T3 + T6; T2b = T3 - T6; Ta = T8 + T9; Tw = T8 - T9; Tx = Tb - Tc; Td = Tb + Tc; } Tn = R0[WS(rs, 15)]; To = R0[WS(rs, 7)]; Te = Ta + Td; T2n = Td - Ta; Ty = Tw + Tx; T1i = Tx - Tw; TD = Tn - To; Tp = Tn + To; Tq = R0[WS(rs, 3)]; Tr = R0[WS(rs, 11)]; } } { E Tj, TA, Ti, Tk; { E Tg, Th, TE, Ts; Tg = R0[WS(rs, 1)]; Th = R0[WS(rs, 9)]; Tj = R0[WS(rs, 5)]; TE = Tq - Tr; Ts = Tq + Tr; TA = Tg - Th; Ti = Tg + Th; T1l = FNMS(KP414213562, TD, TE); TF = FMA(KP414213562, TE, TD); T2d = Tp - Ts; Tt = Tp + Ts; Tk = R0[WS(rs, 13)]; } { E T11, T15, T1c, T20, T14, T16, T1X, T1Y, T1Q, T1R; { E T1a, T1b, T12, T13; { E TZ, T10, TB, Tl; TZ = R1[WS(rs, 15)]; T10 = R1[WS(rs, 7)]; T1a = R1[WS(rs, 11)]; TB = Tj - Tk; Tl = Tj + Tk; T1X = TZ + T10; T11 = TZ - T10; T1k = FMA(KP414213562, TA, TB); TC = FNMS(KP414213562, TB, TA); T2c = Ti - Tl; Tm = Ti + Tl; T1b = R1[WS(rs, 3)]; } T12 = R1[WS(rs, 1)]; T13 = R1[WS(rs, 9)]; T15 = R1[WS(rs, 13)]; T1Y = T1b + T1a; T1c = T1a - T1b; T20 = T12 + T13; T14 = T12 - T13; T16 = R1[WS(rs, 5)]; } T2j = T1X - T1Y; T1Z = T1X + T1Y; { E TT, TU, TL, TM; { E TI, T21, T17, TJ, T18, T1d; TI = R1[0]; T21 = T15 + T16; T17 = T15 - T16; TJ = R1[WS(rs, 8)]; TT = R1[WS(rs, 4)]; T2k = T21 - T20; T22 = T20 + T21; T18 = T14 + T17; T1d = T17 - T14; T1Q = TI + TJ; TK = TI - TJ; T1B = FNMS(KP707106781, T18, T11); T19 = FMA(KP707106781, T18, T11); T1C = FNMS(KP707106781, T1d, T1c); T1e = FMA(KP707106781, T1d, T1c); TU = R1[WS(rs, 12)]; } TL = R1[WS(rs, 2)]; TM = R1[WS(rs, 10)]; TO = R1[WS(rs, 14)]; T1R = TT + TU; TV = TT - TU; T1T = TL + TM; TN = TL - TM; TP = R1[WS(rs, 6)]; } T2g = T1Q - T1R; T1S = T1Q + T1R; } } } { E T1P, T25, T23, T2h, T1W, T1y, TS, T1z, TX, T27, T2a; { E Tf, Tu, T29, T28; { E T1U, TQ, T1V, TR, TW; T1P = T7 - Te; Tf = T7 + Te; T1U = TO + TP; TQ = TO - TP; Tu = Tm + Tt; T25 = Tt - Tm; T23 = T1Z - T22; T29 = T1Z + T22; T2h = T1U - T1T; T1V = T1T + T1U; TR = TN + TQ; TW = TN - TQ; T27 = Tf + Tu; T1W = T1S - T1V; T28 = T1S + T1V; T1y = FNMS(KP707106781, TR, TK); TS = FMA(KP707106781, TR, TK); T1z = FNMS(KP707106781, TW, TV); TX = FMA(KP707106781, TW, TV); T2a = T28 + T29; } Cr[WS(csr, 8)] = Tf - Tu; Ci[WS(csi, 8)] = T29 - T28; } Cr[0] = T27 + T2a; Cr[WS(csr, 16)] = T27 - T2a; { E T2s, T2i, T2v, T2f, T2r, T2p, T2l, T2t; { E T2o, T2e, T26, T24; T2o = T2d - T2c; T2e = T2c + T2d; T2s = FNMS(KP414213562, T2g, T2h); T2i = FMA(KP414213562, T2h, T2g); T26 = T23 - T1W; T24 = T1W + T23; T2v = FNMS(KP707106781, T2e, T2b); T2f = FMA(KP707106781, T2e, T2b); T2r = FMA(KP707106781, T2o, T2n); T2p = FNMS(KP707106781, T2o, T2n); Ci[WS(csi, 4)] = FMA(KP707106781, T26, T25); Ci[WS(csi, 12)] = FMS(KP707106781, T26, T25); Cr[WS(csr, 4)] = FMA(KP707106781, T24, T1P); Cr[WS(csr, 12)] = FNMS(KP707106781, T24, T1P); T2l = FNMS(KP414213562, T2k, T2j); T2t = FMA(KP414213562, T2j, T2k); } { E T1v, T1G, TH, T1s, T1F, T1w, T1o, T1g, T1p, T1n; { E T1f, TY, T1t, T1u, T1j, T1m; { E Tz, TG, T1q, T1r; T1v = FNMS(KP707106781, Ty, Tv); Tz = FMA(KP707106781, Ty, Tv); { E T2q, T2m, T2w, T2u; T2q = T2l - T2i; T2m = T2i + T2l; T2w = T2t - T2s; T2u = T2s + T2t; Ci[WS(csi, 10)] = FMA(KP923879532, T2q, T2p); Ci[WS(csi, 6)] = FMS(KP923879532, T2q, T2p); Cr[WS(csr, 2)] = FMA(KP923879532, T2m, T2f); Cr[WS(csr, 14)] = FNMS(KP923879532, T2m, T2f); Cr[WS(csr, 10)] = FNMS(KP923879532, T2w, T2v); Cr[WS(csr, 6)] = FMA(KP923879532, T2w, T2v); Ci[WS(csi, 2)] = FMA(KP923879532, T2u, T2r); Ci[WS(csi, 14)] = FMS(KP923879532, T2u, T2r); TG = TC + TF; T1G = TF - TC; } T1f = FNMS(KP198912367, T1e, T19); T1q = FMA(KP198912367, T19, T1e); T1r = FMA(KP198912367, TS, TX); TY = FNMS(KP198912367, TX, TS); T1t = FNMS(KP923879532, TG, Tz); TH = FMA(KP923879532, TG, Tz); T1u = T1r + T1q; T1s = T1q - T1r; T1F = FMA(KP707106781, T1i, T1h); T1j = FNMS(KP707106781, T1i, T1h); T1m = T1k + T1l; T1w = T1k - T1l; } Cr[WS(csr, 7)] = FMA(KP980785280, T1u, T1t); T1o = T1f - TY; T1g = TY + T1f; T1p = FMA(KP923879532, T1m, T1j); T1n = FNMS(KP923879532, T1m, T1j); Cr[WS(csr, 9)] = FNMS(KP980785280, T1u, T1t); } Cr[WS(csr, 1)] = FMA(KP980785280, T1g, TH); Cr[WS(csr, 15)] = FNMS(KP980785280, T1g, TH); Ci[WS(csi, 1)] = FMS(KP980785280, T1s, T1p); Ci[WS(csi, 15)] = FMA(KP980785280, T1s, T1p); Ci[WS(csi, 9)] = FMS(KP980785280, T1o, T1n); Ci[WS(csi, 7)] = FMA(KP980785280, T1o, T1n); { E T1A, T1D, T1N, T1O, T1K, T1L; T1A = FMA(KP668178637, T1z, T1y); T1K = FNMS(KP668178637, T1y, T1z); T1L = FNMS(KP668178637, T1B, T1C); T1D = FMA(KP668178637, T1C, T1B); T1N = FNMS(KP923879532, T1w, T1v); T1x = FMA(KP923879532, T1w, T1v); T1O = T1K + T1L; T1M = T1K - T1L; Cr[WS(csr, 5)] = FNMS(KP831469612, T1O, T1N); T1I = T1D - T1A; T1E = T1A + T1D; T1J = FMA(KP923879532, T1G, T1F); T1H = FNMS(KP923879532, T1G, T1F); Cr[WS(csr, 11)] = FMA(KP831469612, T1O, T1N); } } } } } Ci[WS(csi, 3)] = FMA(KP831469612, T1M, T1J); Cr[WS(csr, 3)] = FMA(KP831469612, T1E, T1x); Ci[WS(csi, 13)] = FMS(KP831469612, T1M, T1J); Cr[WS(csr, 13)] = FNMS(KP831469612, T1E, T1x); Ci[WS(csi, 11)] = FMA(KP831469612, T1I, T1H); Ci[WS(csi, 5)] = FMS(KP831469612, T1I, T1H); } } } static const kr2c_desc desc = { 32, "r2cf_32", {88, 0, 68, 0}, &GENUS }; void X(codelet_r2cf_32) (planner *p) { X(kr2c_register) (p, r2cf_32, &desc); } #else /* HAVE_FMA */ /* Generated by: ../../../genfft/gen_r2cf.native -compact -variables 4 -pipeline-latency 4 -n 32 -name r2cf_32 -include r2cf.h */ /* * This function contains 156 FP additions, 42 FP multiplications, * (or, 140 additions, 26 multiplications, 16 fused multiply/add), * 54 stack variables, 7 constants, and 64 memory accesses */ #include "r2cf.h" static void r2cf_32(R *R0, R *R1, R *Cr, R *Ci, stride rs, stride csr, stride csi, INT v, INT ivs, INT ovs) { DK(KP555570233, +0.555570233019602224742830813948532874374937191); DK(KP831469612, +0.831469612302545237078788377617905756738560812); DK(KP195090322, +0.195090322016128267848284868477022240927691618); DK(KP980785280, +0.980785280403230449126182236134239036973933731); DK(KP382683432, +0.382683432365089771728459984030398866761344562); DK(KP923879532, +0.923879532511286756128183189396788286822416626); DK(KP707106781, +0.707106781186547524400844362104849039284835938); { INT i; for (i = v; i > 0; i = i - 1, R0 = R0 + ivs, R1 = R1 + ivs, Cr = Cr + ovs, Ci = Ci + ovs, MAKE_VOLATILE_STRIDE(128, rs), MAKE_VOLATILE_STRIDE(128, csr), MAKE_VOLATILE_STRIDE(128, csi)) { E T7, T2b, Tv, T1l, Te, T2o, Ty, T1k, Tt, T2d, TF, T1h, Tm, T2c, TC; E T1i, T1Z, T22, T2k, T2j, T1e, T1C, T19, T1B, T1S, T1V, T2h, T2g, TX, T1z; E TS, T1y; { E T1, T2, T3, T4, T5, T6; T1 = R0[0]; T2 = R0[WS(rs, 8)]; T3 = T1 + T2; T4 = R0[WS(rs, 4)]; T5 = R0[WS(rs, 12)]; T6 = T4 + T5; T7 = T3 + T6; T2b = T3 - T6; Tv = T1 - T2; T1l = T4 - T5; } { E Ta, Tw, Td, Tx; { E T8, T9, Tb, Tc; T8 = R0[WS(rs, 2)]; T9 = R0[WS(rs, 10)]; Ta = T8 + T9; Tw = T8 - T9; Tb = R0[WS(rs, 14)]; Tc = R0[WS(rs, 6)]; Td = Tb + Tc; Tx = Tb - Tc; } Te = Ta + Td; T2o = Td - Ta; Ty = KP707106781 * (Tw + Tx); T1k = KP707106781 * (Tx - Tw); } { E Tp, TD, Ts, TE; { E Tn, To, Tq, Tr; Tn = R0[WS(rs, 15)]; To = R0[WS(rs, 7)]; Tp = Tn + To; TD = Tn - To; Tq = R0[WS(rs, 3)]; Tr = R0[WS(rs, 11)]; Ts = Tq + Tr; TE = Tq - Tr; } Tt = Tp + Ts; T2d = Tp - Ts; TF = FMA(KP923879532, TD, KP382683432 * TE); T1h = FNMS(KP923879532, TE, KP382683432 * TD); } { E Ti, TA, Tl, TB; { E Tg, Th, Tj, Tk; Tg = R0[WS(rs, 1)]; Th = R0[WS(rs, 9)]; Ti = Tg + Th; TA = Tg - Th; Tj = R0[WS(rs, 5)]; Tk = R0[WS(rs, 13)]; Tl = Tj + Tk; TB = Tj - Tk; } Tm = Ti + Tl; T2c = Ti - Tl; TC = FNMS(KP382683432, TB, KP923879532 * TA); T1i = FMA(KP382683432, TA, KP923879532 * TB); } { E T11, T1X, T1d, T1Y, T14, T20, T17, T21, T1a, T18; { E TZ, T10, T1b, T1c; TZ = R1[WS(rs, 15)]; T10 = R1[WS(rs, 7)]; T11 = TZ - T10; T1X = TZ + T10; T1b = R1[WS(rs, 3)]; T1c = R1[WS(rs, 11)]; T1d = T1b - T1c; T1Y = T1b + T1c; } { E T12, T13, T15, T16; T12 = R1[WS(rs, 1)]; T13 = R1[WS(rs, 9)]; T14 = T12 - T13; T20 = T12 + T13; T15 = R1[WS(rs, 13)]; T16 = R1[WS(rs, 5)]; T17 = T15 - T16; T21 = T15 + T16; } T1Z = T1X + T1Y; T22 = T20 + T21; T2k = T21 - T20; T2j = T1X - T1Y; T1a = KP707106781 * (T17 - T14); T1e = T1a - T1d; T1C = T1d + T1a; T18 = KP707106781 * (T14 + T17); T19 = T11 + T18; T1B = T11 - T18; } { E TK, T1Q, TW, T1R, TN, T1T, TQ, T1U, TT, TR; { E TI, TJ, TU, TV; TI = R1[0]; TJ = R1[WS(rs, 8)]; TK = TI - TJ; T1Q = TI + TJ; TU = R1[WS(rs, 4)]; TV = R1[WS(rs, 12)]; TW = TU - TV; T1R = TU + TV; } { E TL, TM, TO, TP; TL = R1[WS(rs, 2)]; TM = R1[WS(rs, 10)]; TN = TL - TM; T1T = TL + TM; TO = R1[WS(rs, 14)]; TP = R1[WS(rs, 6)]; TQ = TO - TP; T1U = TO + TP; } T1S = T1Q + T1R; T1V = T1T + T1U; T2h = T1U - T1T; T2g = T1Q - T1R; TT = KP707106781 * (TQ - TN); TX = TT - TW; T1z = TW + TT; TR = KP707106781 * (TN + TQ); TS = TK + TR; T1y = TK - TR; } { E Tf, Tu, T27, T28, T29, T2a; Tf = T7 + Te; Tu = Tm + Tt; T27 = Tf + Tu; T28 = T1S + T1V; T29 = T1Z + T22; T2a = T28 + T29; Cr[WS(csr, 8)] = Tf - Tu; Ci[WS(csi, 8)] = T29 - T28; Cr[WS(csr, 16)] = T27 - T2a; Cr[0] = T27 + T2a; } { E T1P, T25, T24, T26, T1W, T23; T1P = T7 - Te; T25 = Tt - Tm; T1W = T1S - T1V; T23 = T1Z - T22; T24 = KP707106781 * (T1W + T23); T26 = KP707106781 * (T23 - T1W); Cr[WS(csr, 12)] = T1P - T24; Ci[WS(csi, 12)] = T26 - T25; Cr[WS(csr, 4)] = T1P + T24; Ci[WS(csi, 4)] = T25 + T26; } { E T2f, T2v, T2p, T2r, T2m, T2q, T2u, T2w, T2e, T2n; T2e = KP707106781 * (T2c + T2d); T2f = T2b + T2e; T2v = T2b - T2e; T2n = KP707106781 * (T2d - T2c); T2p = T2n - T2o; T2r = T2o + T2n; { E T2i, T2l, T2s, T2t; T2i = FMA(KP923879532, T2g, KP382683432 * T2h); T2l = FNMS(KP382683432, T2k, KP923879532 * T2j); T2m = T2i + T2l; T2q = T2l - T2i; T2s = FNMS(KP382683432, T2g, KP923879532 * T2h); T2t = FMA(KP382683432, T2j, KP923879532 * T2k); T2u = T2s + T2t; T2w = T2t - T2s; } Cr[WS(csr, 14)] = T2f - T2m; Ci[WS(csi, 14)] = T2u - T2r; Cr[WS(csr, 2)] = T2f + T2m; Ci[WS(csi, 2)] = T2r + T2u; Ci[WS(csi, 6)] = T2p + T2q; Cr[WS(csr, 6)] = T2v + T2w; Ci[WS(csi, 10)] = T2q - T2p; Cr[WS(csr, 10)] = T2v - T2w; } { E TH, T1t, T1s, T1u, T1g, T1o, T1n, T1p; { E Tz, TG, T1q, T1r; Tz = Tv + Ty; TG = TC + TF; TH = Tz + TG; T1t = Tz - TG; T1q = FNMS(KP195090322, TS, KP980785280 * TX); T1r = FMA(KP195090322, T19, KP980785280 * T1e); T1s = T1q + T1r; T1u = T1r - T1q; } { E TY, T1f, T1j, T1m; TY = FMA(KP980785280, TS, KP195090322 * TX); T1f = FNMS(KP195090322, T1e, KP980785280 * T19); T1g = TY + T1f; T1o = T1f - TY; T1j = T1h - T1i; T1m = T1k - T1l; T1n = T1j - T1m; T1p = T1m + T1j; } Cr[WS(csr, 15)] = TH - T1g; Ci[WS(csi, 15)] = T1s - T1p; Cr[WS(csr, 1)] = TH + T1g; Ci[WS(csi, 1)] = T1p + T1s; Ci[WS(csi, 7)] = T1n + T1o; Cr[WS(csr, 7)] = T1t + T1u; Ci[WS(csi, 9)] = T1o - T1n; Cr[WS(csr, 9)] = T1t - T1u; } { E T1x, T1N, T1M, T1O, T1E, T1I, T1H, T1J; { E T1v, T1w, T1K, T1L; T1v = Tv - Ty; T1w = T1i + T1h; T1x = T1v + T1w; T1N = T1v - T1w; T1K = FNMS(KP555570233, T1y, KP831469612 * T1z); T1L = FMA(KP555570233, T1B, KP831469612 * T1C); T1M = T1K + T1L; T1O = T1L - T1K; } { E T1A, T1D, T1F, T1G; T1A = FMA(KP831469612, T1y, KP555570233 * T1z); T1D = FNMS(KP555570233, T1C, KP831469612 * T1B); T1E = T1A + T1D; T1I = T1D - T1A; T1F = TF - TC; T1G = T1l + T1k; T1H = T1F - T1G; T1J = T1G + T1F; } Cr[WS(csr, 13)] = T1x - T1E; Ci[WS(csi, 13)] = T1M - T1J; Cr[WS(csr, 3)] = T1x + T1E; Ci[WS(csi, 3)] = T1J + T1M; Ci[WS(csi, 5)] = T1H + T1I; Cr[WS(csr, 5)] = T1N + T1O; Ci[WS(csi, 11)] = T1I - T1H; Cr[WS(csr, 11)] = T1N - T1O; } } } } static const kr2c_desc desc = { 32, "r2cf_32", {140, 26, 16, 0}, &GENUS }; void X(codelet_r2cf_32) (planner *p) { X(kr2c_register) (p, r2cf_32, &desc); } #endif /* HAVE_FMA */ fftw-3.3.3/rdft/scalar/r2cf/r2cf_64.c0000644000175000017500000011615112054410640013761 00000000000000/* * Copyright (c) 2003, 2007-11 Matteo Frigo * Copyright (c) 2003, 2007-11 Massachusetts Institute of Technology * * This program is free software; you can redistribute it and/or modify * it under the terms of the GNU General Public License as published by * the Free Software Foundation; either version 2 of the License, or * (at your option) any later version. * * This program is distributed in the hope that it will be useful, * but WITHOUT ANY WARRANTY; without even the implied warranty of * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the * GNU General Public License for more details. * * You should have received a copy of the GNU General Public License * along with this program; if not, write to the Free Software * Foundation, Inc., 51 Franklin Street, Fifth Floor, Boston, MA 02110-1301 USA * */ /* This file was automatically generated --- DO NOT EDIT */ /* Generated on Sun Nov 25 07:39:47 EST 2012 */ #include "codelet-rdft.h" #ifdef HAVE_FMA /* Generated by: ../../../genfft/gen_r2cf.native -fma -reorder-insns -schedule-for-pipeline -compact -variables 4 -pipeline-latency 4 -n 64 -name r2cf_64 -include r2cf.h */ /* * This function contains 394 FP additions, 196 FP multiplications, * (or, 198 additions, 0 multiplications, 196 fused multiply/add), * 133 stack variables, 15 constants, and 128 memory accesses */ #include "r2cf.h" static void r2cf_64(R *R0, R *R1, R *Cr, R *Ci, stride rs, stride csr, stride csi, INT v, INT ivs, INT ovs) { DK(KP773010453, +0.773010453362736960810906609758469800971041293); DK(KP995184726, +0.995184726672196886244836953109479921575474869); DK(KP098491403, +0.098491403357164253077197521291327432293052451); DK(KP820678790, +0.820678790828660330972281985331011598767386482); DK(KP956940335, +0.956940335732208864935797886980269969482849206); DK(KP881921264, +0.881921264348355029712756863660388349508442621); DK(KP534511135, +0.534511135950791641089685961295362908582039528); DK(KP303346683, +0.303346683607342391675883946941299872384187453); DK(KP980785280, +0.980785280403230449126182236134239036973933731); DK(KP198912367, +0.198912367379658006911597622644676228597850501); DK(KP831469612, +0.831469612302545237078788377617905756738560812); DK(KP668178637, +0.668178637919298919997757686523080761552472251); DK(KP923879532, +0.923879532511286756128183189396788286822416626); DK(KP414213562, +0.414213562373095048801688724209698078569671875); DK(KP707106781, +0.707106781186547524400844362104849039284835938); { INT i; for (i = v; i > 0; i = i - 1, R0 = R0 + ivs, R1 = R1 + ivs, Cr = Cr + ovs, Ci = Ci + ovs, MAKE_VOLATILE_STRIDE(256, rs), MAKE_VOLATILE_STRIDE(256, csr), MAKE_VOLATILE_STRIDE(256, csi)) { E T5n, T5o; { E T11, T2j, T4P, T5P, T3D, T5p, T3d, Tf, T1k, T1H, T5D, T4l, T5A, T4a, T3i; E T2U, T1R, T2e, T5K, T4G, T5H, T4v, T3l, T31, T5s, T42, T5t, T3Z, T2n, T1b; E T3f, TZ, T5v, T3T, T5w, T3Q, T2m, T18, T3e, TK, T3K, T5Q, T4S, T5q, T14; E T2k, T3p, Tu, T4w, T1U, T5E, T4h, T5B, T4o, T3j, T2X, T1I, T1z, T1Z, T4A; E T24, T4x, T1X, T20; { E TN, T3V, TS, TX, T3X, TQ, T40, TT; { E T1g, T46, T1B, T1G, T47, T1j, T4j, T1C; { E T4, T3z, T3, T3B, Td, T5, T8, T9; { E T1, T2, Tb, Tc; T1 = R0[0]; T2 = R0[WS(rs, 16)]; Tb = R0[WS(rs, 28)]; Tc = R0[WS(rs, 12)]; T4 = R0[WS(rs, 8)]; T3z = T1 - T2; T3 = T1 + T2; T3B = Tb - Tc; Td = Tb + Tc; T5 = R0[WS(rs, 24)]; T8 = R0[WS(rs, 4)]; T9 = R0[WS(rs, 20)]; } { E T1E, T1F, T1h, T1i; { E T1e, T4N, T6, T3A, Ta, T1f; T1e = R1[0]; T4N = T4 - T5; T6 = T4 + T5; T3A = T8 - T9; Ta = T8 + T9; T1f = R1[WS(rs, 16)]; { E T7, T3C, T4O, Te; T11 = T3 - T6; T7 = T3 + T6; T3C = T3A + T3B; T4O = T3B - T3A; T2j = Td - Ta; Te = Ta + Td; T4P = FNMS(KP707106781, T4O, T4N); T5P = FMA(KP707106781, T4O, T4N); T3D = FMA(KP707106781, T3C, T3z); T5p = FNMS(KP707106781, T3C, T3z); T3d = T7 - Te; Tf = T7 + Te; T1g = T1e + T1f; T46 = T1e - T1f; } } T1E = R1[WS(rs, 4)]; T1F = R1[WS(rs, 20)]; T1h = R1[WS(rs, 8)]; T1i = R1[WS(rs, 24)]; T1B = R1[WS(rs, 28)]; T1G = T1E + T1F; T47 = T1E - T1F; T1j = T1h + T1i; T4j = T1h - T1i; T1C = R1[WS(rs, 12)]; } } { E T1N, T4r, T28, T2d, T4s, T1Q, T4E, T29; { E T2b, T2c, T1O, T1P; { E T2S, T48, T1D, T1L, T1M, T4k, T49, T2T; T1L = R1[WS(rs, 31)]; T1M = R1[WS(rs, 15)]; T2S = T1g + T1j; T1k = T1g - T1j; T48 = T1B - T1C; T1D = T1B + T1C; T1N = T1L + T1M; T4r = T1L - T1M; T4k = T47 - T48; T49 = T47 + T48; T2T = T1G + T1D; T1H = T1D - T1G; T5D = FNMS(KP707106781, T4k, T4j); T4l = FMA(KP707106781, T4k, T4j); T5A = FNMS(KP707106781, T49, T46); T4a = FMA(KP707106781, T49, T46); T3i = T2S - T2T; T2U = T2S + T2T; T2b = R1[WS(rs, 3)]; T2c = R1[WS(rs, 19)]; } T1O = R1[WS(rs, 7)]; T1P = R1[WS(rs, 23)]; T28 = R1[WS(rs, 27)]; T2d = T2b + T2c; T4s = T2b - T2c; T1Q = T1O + T1P; T4E = T1P - T1O; T29 = R1[WS(rs, 11)]; } { E TV, TW, TO, TP; { E T2Z, T4t, T2a, TL, TM, T4F, T4u, T30; TL = R0[WS(rs, 31)]; TM = R0[WS(rs, 15)]; T2Z = T1N + T1Q; T1R = T1N - T1Q; T4t = T28 - T29; T2a = T28 + T29; TN = TL + TM; T3V = TL - TM; T4F = T4t - T4s; T4u = T4s + T4t; T30 = T2d + T2a; T2e = T2a - T2d; T5K = FNMS(KP707106781, T4F, T4E); T4G = FMA(KP707106781, T4F, T4E); T5H = FNMS(KP707106781, T4u, T4r); T4v = FMA(KP707106781, T4u, T4r); T3l = T2Z - T30; T31 = T2Z + T30; TV = R0[WS(rs, 27)]; TW = R0[WS(rs, 11)]; } TO = R0[WS(rs, 7)]; TP = R0[WS(rs, 23)]; TS = R0[WS(rs, 3)]; TX = TV + TW; T3X = TV - TW; TQ = TO + TP; T40 = TO - TP; TT = R0[WS(rs, 19)]; } } } { E Ti, T3E, Tn, Ts, T3I, Tl, T3F, To; { E Ty, T3M, TD, TI, T3O, TB, T3R, TE; { E TG, TH, Tz, TA; { E T19, TR, T3W, TU, Tw, Tx; Tw = R0[WS(rs, 1)]; Tx = R0[WS(rs, 17)]; T19 = TN - TQ; TR = TN + TQ; T3W = TS - TT; TU = TS + TT; Ty = Tw + Tx; T3M = Tw - Tx; { E T41, T3Y, T1a, TY; T41 = T3W - T3X; T3Y = T3W + T3X; T1a = TX - TU; TY = TU + TX; T5s = FNMS(KP707106781, T41, T40); T42 = FMA(KP707106781, T41, T40); T5t = FNMS(KP707106781, T3Y, T3V); T3Z = FMA(KP707106781, T3Y, T3V); T2n = FMA(KP414213562, T19, T1a); T1b = FNMS(KP414213562, T1a, T19); T3f = TR - TY; TZ = TR + TY; TG = R0[WS(rs, 29)]; TH = R0[WS(rs, 13)]; } } Tz = R0[WS(rs, 9)]; TA = R0[WS(rs, 25)]; TD = R0[WS(rs, 5)]; TI = TG + TH; T3O = TG - TH; TB = Tz + TA; T3R = Tz - TA; TE = R0[WS(rs, 21)]; } { E Tq, Tr, Tj, Tk; { E T16, TC, T3N, TF, Tg, Th; Tg = R0[WS(rs, 2)]; Th = R0[WS(rs, 18)]; T16 = Ty - TB; TC = Ty + TB; T3N = TD - TE; TF = TD + TE; Ti = Tg + Th; T3E = Tg - Th; { E T3S, T3P, T17, TJ; T3S = T3N - T3O; T3P = T3N + T3O; T17 = TI - TF; TJ = TF + TI; T5v = FNMS(KP707106781, T3S, T3R); T3T = FMA(KP707106781, T3S, T3R); T5w = FNMS(KP707106781, T3P, T3M); T3Q = FMA(KP707106781, T3P, T3M); T2m = FNMS(KP414213562, T16, T17); T18 = FMA(KP414213562, T17, T16); T3e = TC - TJ; TK = TC + TJ; Tq = R0[WS(rs, 6)]; Tr = R0[WS(rs, 22)]; } } Tj = R0[WS(rs, 10)]; Tk = R0[WS(rs, 26)]; Tn = R0[WS(rs, 30)]; Ts = Tq + Tr; T3I = Tq - Tr; Tl = Tj + Tk; T3F = Tj - Tk; To = R0[WS(rs, 14)]; } } { E T1n, T4b, T1s, T4f, T1x, T4c, T1q, T1t; { E T1v, T1w, T1o, T1p; { E T1l, T4Q, T3G, Tm, T12, Tp, T3H, T1m; T1l = R1[WS(rs, 2)]; T4Q = FMA(KP414213562, T3E, T3F); T3G = FNMS(KP414213562, T3F, T3E); Tm = Ti + Tl; T12 = Ti - Tl; Tp = Tn + To; T3H = Tn - To; T1m = R1[WS(rs, 18)]; T1v = R1[WS(rs, 6)]; { E T4R, T3J, Tt, T13; T4R = FNMS(KP414213562, T3H, T3I); T3J = FMA(KP414213562, T3I, T3H); Tt = Tp + Ts; T13 = Tp - Ts; T1n = T1l + T1m; T4b = T1l - T1m; T3K = T3G + T3J; T5Q = T3J - T3G; T4S = T4Q + T4R; T5q = T4Q - T4R; T14 = T12 + T13; T2k = T13 - T12; T3p = Tt - Tm; Tu = Tm + Tt; T1w = R1[WS(rs, 22)]; } } T1o = R1[WS(rs, 10)]; T1p = R1[WS(rs, 26)]; T1s = R1[WS(rs, 30)]; T4f = T1v - T1w; T1x = T1v + T1w; T4c = T1o - T1p; T1q = T1o + T1p; T1t = R1[WS(rs, 14)]; } { E T22, T23, T1V, T1W; { E T1S, T4d, T4m, T2V, T1r, T4e, T1u, T1T; T1S = R1[WS(rs, 1)]; T4d = FNMS(KP414213562, T4c, T4b); T4m = FMA(KP414213562, T4b, T4c); T2V = T1n + T1q; T1r = T1n - T1q; T4e = T1s - T1t; T1u = T1s + T1t; T1T = R1[WS(rs, 17)]; T22 = R1[WS(rs, 5)]; { E T4g, T4n, T2W, T1y; T4g = FMA(KP414213562, T4f, T4e); T4n = FNMS(KP414213562, T4e, T4f); T2W = T1u + T1x; T1y = T1u - T1x; T4w = T1S - T1T; T1U = T1S + T1T; T5E = T4g - T4d; T4h = T4d + T4g; T5B = T4m - T4n; T4o = T4m + T4n; T3j = T2W - T2V; T2X = T2V + T2W; T1I = T1y - T1r; T1z = T1r + T1y; T23 = R1[WS(rs, 21)]; } } T1V = R1[WS(rs, 9)]; T1W = R1[WS(rs, 25)]; T1Z = R1[WS(rs, 29)]; T4A = T23 - T22; T24 = T22 + T23; T4x = T1W - T1V; T1X = T1V + T1W; T20 = R1[WS(rs, 13)]; } } } } { E T4C, T5L, T4J, T5I, T26, T2f, T3q, T3h, T3w, T3s, T3o, T3r, T3t; { E T2R, T37, T2Y, T3a, T39, T3m, T3b, T35, Tv, T10, T34, T3c, T3x, T3y; { E T4y, T4H, T32, T1Y, T4z, T21; T2R = Tf - Tu; Tv = Tf + Tu; T4y = FMA(KP414213562, T4x, T4w); T4H = FNMS(KP414213562, T4w, T4x); T32 = T1U + T1X; T1Y = T1U - T1X; T4z = T1Z - T20; T21 = T1Z + T20; T10 = TK + TZ; T37 = TZ - TK; T2Y = T2U - T2X; T3a = T2U + T2X; { E T4B, T4I, T33, T25; T4B = FNMS(KP414213562, T4A, T4z); T4I = FMA(KP414213562, T4z, T4A); T33 = T21 + T24; T25 = T21 - T24; T39 = Tv + T10; T4C = T4y + T4B; T5L = T4B - T4y; T4J = T4H + T4I; T5I = T4I - T4H; T34 = T32 + T33; T3m = T33 - T32; T26 = T1Y + T25; T2f = T25 - T1Y; } } Cr[WS(csr, 16)] = Tv - T10; T3b = T31 + T34; T35 = T31 - T34; Ci[WS(csi, 16)] = T3b - T3a; T3c = T3a + T3b; { E T3k, T3u, T3v, T3n, T36, T38, T3g; T3g = T3e + T3f; T3q = T3f - T3e; Cr[0] = T39 + T3c; Cr[WS(csr, 32)] = T39 - T3c; T36 = T2Y + T35; T38 = T35 - T2Y; T3x = FNMS(KP707106781, T3g, T3d); T3h = FMA(KP707106781, T3g, T3d); Ci[WS(csi, 8)] = FMA(KP707106781, T38, T37); Ci[WS(csi, 24)] = FMS(KP707106781, T38, T37); Cr[WS(csr, 8)] = FMA(KP707106781, T36, T2R); Cr[WS(csr, 24)] = FNMS(KP707106781, T36, T2R); T3k = FMA(KP414213562, T3j, T3i); T3u = FNMS(KP414213562, T3i, T3j); T3v = FMA(KP414213562, T3l, T3m); T3n = FNMS(KP414213562, T3m, T3l); T3y = T3v - T3u; T3w = T3u + T3v; T3s = T3n - T3k; T3o = T3k + T3n; } Cr[WS(csr, 12)] = FMA(KP923879532, T3y, T3x); Cr[WS(csr, 20)] = FNMS(KP923879532, T3y, T3x); } Cr[WS(csr, 4)] = FMA(KP923879532, T3o, T3h); Cr[WS(csr, 28)] = FNMS(KP923879532, T3o, T3h); T3r = FNMS(KP707106781, T3q, T3p); T3t = FMA(KP707106781, T3q, T3p); { E T27, T2g, T2v, T1d, T2r, T2p, T2s, T1K, T6l, T6m; { E T15, T2o, T2P, T2z, T2l, T1c, T1A, T1J, T2D, T2L, T2J, T2M, T2C, T2E, T2N; E T2F; { E T2H, T2I, T2x, T2y, T2A, T2B; T15 = FMA(KP707106781, T14, T11); T2x = FNMS(KP707106781, T14, T11); T2y = T2n - T2m; T2o = T2m + T2n; Ci[WS(csi, 4)] = FMA(KP923879532, T3w, T3t); Ci[WS(csi, 28)] = FMS(KP923879532, T3w, T3t); Ci[WS(csi, 20)] = FMA(KP923879532, T3s, T3r); Ci[WS(csi, 12)] = FMS(KP923879532, T3s, T3r); T2P = FNMS(KP923879532, T2y, T2x); T2z = FMA(KP923879532, T2y, T2x); T2l = FMA(KP707106781, T2k, T2j); T2H = FNMS(KP707106781, T2k, T2j); T2I = T1b - T18; T1c = T18 + T1b; T1A = FMA(KP707106781, T1z, T1k); T2A = FNMS(KP707106781, T1z, T1k); T2B = FNMS(KP707106781, T1I, T1H); T1J = FMA(KP707106781, T1I, T1H); T27 = FMA(KP707106781, T26, T1R); T2D = FNMS(KP707106781, T26, T1R); T2L = FNMS(KP923879532, T2I, T2H); T2J = FMA(KP923879532, T2I, T2H); T2M = FMA(KP668178637, T2A, T2B); T2C = FNMS(KP668178637, T2B, T2A); T2E = FNMS(KP707106781, T2f, T2e); T2g = FMA(KP707106781, T2f, T2e); } T2N = FNMS(KP668178637, T2D, T2E); T2F = FMA(KP668178637, T2E, T2D); T2v = FNMS(KP923879532, T1c, T15); T1d = FMA(KP923879532, T1c, T15); { E T2Q, T2O, T2K, T2G; T2Q = T2M - T2N; T2O = T2M + T2N; T2K = T2F - T2C; T2G = T2C + T2F; Cr[WS(csr, 10)] = FMA(KP831469612, T2Q, T2P); Cr[WS(csr, 22)] = FNMS(KP831469612, T2Q, T2P); Ci[WS(csi, 26)] = FNMS(KP831469612, T2O, T2L); Ci[WS(csi, 6)] = -(FMA(KP831469612, T2O, T2L)); Ci[WS(csi, 22)] = FMS(KP831469612, T2K, T2J); Ci[WS(csi, 10)] = FMA(KP831469612, T2K, T2J); Cr[WS(csr, 6)] = FMA(KP831469612, T2G, T2z); Cr[WS(csr, 26)] = FNMS(KP831469612, T2G, T2z); } T2r = FMA(KP923879532, T2o, T2l); T2p = FNMS(KP923879532, T2o, T2l); T2s = FNMS(KP198912367, T1A, T1J); T1K = FMA(KP198912367, T1J, T1A); } { E T63, T5r, T5R, T6d, T5J, T5M, T6e, T5y, T6j, T6b, T66, T67, T64, T5U, T5Z; E T5G; { E T5S, T5u, T5x, T5T, T2t, T2h; T63 = FMA(KP923879532, T5q, T5p); T5r = FNMS(KP923879532, T5q, T5p); T5R = FNMS(KP923879532, T5Q, T5P); T6d = FMA(KP923879532, T5Q, T5P); T2t = FMA(KP198912367, T27, T2g); T2h = FNMS(KP198912367, T2g, T27); T5S = FNMS(KP668178637, T5s, T5t); T5u = FMA(KP668178637, T5t, T5s); { E T2w, T2u, T2q, T2i; T2w = T2t - T2s; T2u = T2s + T2t; T2q = T2h - T1K; T2i = T1K + T2h; Cr[WS(csr, 14)] = FMA(KP980785280, T2w, T2v); Cr[WS(csr, 18)] = FNMS(KP980785280, T2w, T2v); Ci[WS(csi, 30)] = FMS(KP980785280, T2u, T2r); Ci[WS(csi, 2)] = FMA(KP980785280, T2u, T2r); Ci[WS(csi, 18)] = FMA(KP980785280, T2q, T2p); Ci[WS(csi, 14)] = FMS(KP980785280, T2q, T2p); Cr[WS(csr, 2)] = FMA(KP980785280, T2i, T1d); Cr[WS(csr, 30)] = FNMS(KP980785280, T2i, T1d); T5x = FNMS(KP668178637, T5w, T5v); T5T = FMA(KP668178637, T5v, T5w); } { E T69, T6a, T5C, T5F; T5J = FNMS(KP923879532, T5I, T5H); T69 = FMA(KP923879532, T5I, T5H); T6a = FNMS(KP923879532, T5L, T5K); T5M = FMA(KP923879532, T5L, T5K); T6e = T5x + T5u; T5y = T5u - T5x; T6j = FNMS(KP303346683, T69, T6a); T6b = FMA(KP303346683, T6a, T69); T66 = FMA(KP923879532, T5B, T5A); T5C = FNMS(KP923879532, T5B, T5A); T5F = FNMS(KP923879532, T5E, T5D); T67 = FMA(KP923879532, T5E, T5D); T64 = T5T + T5S; T5U = T5S - T5T; T5Z = FMA(KP534511135, T5C, T5F); T5G = FNMS(KP534511135, T5F, T5C); } } { E T61, T6i, T68, T62; { E T5z, T5Y, T5N, T5X, T5V, T60, T5W, T5O; T61 = FNMS(KP831469612, T5y, T5r); T5z = FMA(KP831469612, T5y, T5r); T6i = FNMS(KP303346683, T66, T67); T68 = FMA(KP303346683, T67, T66); T5Y = FMA(KP534511135, T5J, T5M); T5N = FNMS(KP534511135, T5M, T5J); T5X = FNMS(KP831469612, T5U, T5R); T5V = FMA(KP831469612, T5U, T5R); T60 = T5Y - T5Z; T62 = T5Z + T5Y; T5W = T5N - T5G; T5O = T5G + T5N; Ci[WS(csi, 27)] = FMA(KP881921264, T60, T5X); Ci[WS(csi, 5)] = FMS(KP881921264, T60, T5X); Cr[WS(csr, 5)] = FMA(KP881921264, T5O, T5z); Cr[WS(csr, 27)] = FNMS(KP881921264, T5O, T5z); Ci[WS(csi, 21)] = FMS(KP881921264, T5W, T5V); Ci[WS(csi, 11)] = FMA(KP881921264, T5W, T5V); } { E T6g, T6f, T6h, T6k, T65, T6c; T6l = FNMS(KP831469612, T64, T63); T65 = FMA(KP831469612, T64, T63); T6c = T68 + T6b; T6g = T6b - T68; T6f = FNMS(KP831469612, T6e, T6d); T6h = FMA(KP831469612, T6e, T6d); Cr[WS(csr, 11)] = FMA(KP881921264, T62, T61); Cr[WS(csr, 21)] = FNMS(KP881921264, T62, T61); Cr[WS(csr, 3)] = FMA(KP956940335, T6c, T65); Cr[WS(csr, 29)] = FNMS(KP956940335, T6c, T65); T6k = T6i - T6j; T6m = T6i + T6j; Ci[WS(csi, 29)] = FMS(KP956940335, T6k, T6h); Ci[WS(csi, 3)] = FMA(KP956940335, T6k, T6h); Ci[WS(csi, 19)] = FMA(KP956940335, T6g, T6f); Ci[WS(csi, 13)] = FMS(KP956940335, T6g, T6f); } } } { E T55, T3L, T4T, T5f, T4D, T4K, T5g, T44, T5l, T5d, T58, T59, T56, T4W, T51; E T4q; { E T4U, T3U, T43, T4V; T55 = FNMS(KP923879532, T3K, T3D); T3L = FMA(KP923879532, T3K, T3D); T4T = FMA(KP923879532, T4S, T4P); T5f = FNMS(KP923879532, T4S, T4P); Cr[WS(csr, 13)] = FNMS(KP956940335, T6m, T6l); Cr[WS(csr, 19)] = FMA(KP956940335, T6m, T6l); T4U = FMA(KP198912367, T3Q, T3T); T3U = FNMS(KP198912367, T3T, T3Q); T43 = FMA(KP198912367, T42, T3Z); T4V = FNMS(KP198912367, T3Z, T42); { E T5b, T5c, T4i, T4p; T4D = FMA(KP923879532, T4C, T4v); T5b = FNMS(KP923879532, T4C, T4v); T5c = FNMS(KP923879532, T4J, T4G); T4K = FMA(KP923879532, T4J, T4G); T5g = T43 - T3U; T44 = T3U + T43; T5l = FNMS(KP820678790, T5b, T5c); T5d = FMA(KP820678790, T5c, T5b); T58 = FNMS(KP923879532, T4h, T4a); T4i = FMA(KP923879532, T4h, T4a); T4p = FMA(KP923879532, T4o, T4l); T59 = FNMS(KP923879532, T4o, T4l); T56 = T4U - T4V; T4W = T4U + T4V; T51 = FMA(KP098491403, T4i, T4p); T4q = FNMS(KP098491403, T4p, T4i); } } { E T53, T5k, T5a, T54; { E T45, T50, T4L, T4Z, T4X, T52, T4Y, T4M; T53 = FNMS(KP980785280, T44, T3L); T45 = FMA(KP980785280, T44, T3L); T5k = FNMS(KP820678790, T58, T59); T5a = FMA(KP820678790, T59, T58); T50 = FMA(KP098491403, T4D, T4K); T4L = FNMS(KP098491403, T4K, T4D); T4Z = FMA(KP980785280, T4W, T4T); T4X = FNMS(KP980785280, T4W, T4T); T52 = T50 - T51; T54 = T51 + T50; T4Y = T4L - T4q; T4M = T4q + T4L; Ci[WS(csi, 31)] = FMA(KP995184726, T52, T4Z); Ci[WS(csi, 1)] = FMS(KP995184726, T52, T4Z); Cr[WS(csr, 1)] = FMA(KP995184726, T4M, T45); Cr[WS(csr, 31)] = FNMS(KP995184726, T4M, T45); Ci[WS(csi, 17)] = FMS(KP995184726, T4Y, T4X); Ci[WS(csi, 15)] = FMA(KP995184726, T4Y, T4X); } { E T5i, T5h, T5j, T5m, T57, T5e; T5n = FNMS(KP980785280, T56, T55); T57 = FMA(KP980785280, T56, T55); T5e = T5a + T5d; T5i = T5d - T5a; T5h = FNMS(KP980785280, T5g, T5f); T5j = FMA(KP980785280, T5g, T5f); Cr[WS(csr, 15)] = FMA(KP995184726, T54, T53); Cr[WS(csr, 17)] = FNMS(KP995184726, T54, T53); Cr[WS(csr, 7)] = FMA(KP773010453, T5e, T57); Cr[WS(csr, 25)] = FNMS(KP773010453, T5e, T57); T5m = T5k - T5l; T5o = T5k + T5l; Ci[WS(csi, 25)] = FMS(KP773010453, T5m, T5j); Ci[WS(csi, 7)] = FMA(KP773010453, T5m, T5j); Ci[WS(csi, 23)] = FMA(KP773010453, T5i, T5h); Ci[WS(csi, 9)] = FMS(KP773010453, T5i, T5h); } } } } } } Cr[WS(csr, 9)] = FNMS(KP773010453, T5o, T5n); Cr[WS(csr, 23)] = FMA(KP773010453, T5o, T5n); } } } static const kr2c_desc desc = { 64, "r2cf_64", {198, 0, 196, 0}, &GENUS }; void X(codelet_r2cf_64) (planner *p) { X(kr2c_register) (p, r2cf_64, &desc); } #else /* HAVE_FMA */ /* Generated by: ../../../genfft/gen_r2cf.native -compact -variables 4 -pipeline-latency 4 -n 64 -name r2cf_64 -include r2cf.h */ /* * This function contains 394 FP additions, 124 FP multiplications, * (or, 342 additions, 72 multiplications, 52 fused multiply/add), * 106 stack variables, 15 constants, and 128 memory accesses */ #include "r2cf.h" static void r2cf_64(R *R0, R *R1, R *Cr, R *Ci, stride rs, stride csr, stride csi, INT v, INT ivs, INT ovs) { DK(KP773010453, +0.773010453362736960810906609758469800971041293); DK(KP634393284, +0.634393284163645498215171613225493370675687095); DK(KP098017140, +0.098017140329560601994195563888641845861136673); DK(KP995184726, +0.995184726672196886244836953109479921575474869); DK(KP290284677, +0.290284677254462367636192375817395274691476278); DK(KP956940335, +0.956940335732208864935797886980269969482849206); DK(KP471396736, +0.471396736825997648556387625905254377657460319); DK(KP881921264, +0.881921264348355029712756863660388349508442621); DK(KP195090322, +0.195090322016128267848284868477022240927691618); DK(KP980785280, +0.980785280403230449126182236134239036973933731); DK(KP555570233, +0.555570233019602224742830813948532874374937191); DK(KP831469612, +0.831469612302545237078788377617905756738560812); DK(KP382683432, +0.382683432365089771728459984030398866761344562); DK(KP923879532, +0.923879532511286756128183189396788286822416626); DK(KP707106781, +0.707106781186547524400844362104849039284835938); { INT i; for (i = v; i > 0; i = i - 1, R0 = R0 + ivs, R1 = R1 + ivs, Cr = Cr + ovs, Ci = Ci + ovs, MAKE_VOLATILE_STRIDE(256, rs), MAKE_VOLATILE_STRIDE(256, csr), MAKE_VOLATILE_STRIDE(256, csi)) { E T4l, T5a, T15, T3n, T2T, T3Q, T7, Te, Tf, T4A, T4L, T1X, T3B, T23, T3y; E T5I, T66, T4R, T52, T2j, T3F, T2H, T3I, T5P, T69, T1i, T3t, T1l, T3u, TZ; E T63, T4v, T58, T1r, T3r, T1u, T3q, TK, T62, T4s, T57, Tm, Tt, Tu, T4o; E T5b, T1c, T3R, T2Q, T3o, T1M, T3z, T5L, T67, T26, T3C, T4H, T4M, T2y, T3J; E T5S, T6a, T2C, T3G, T4Y, T53; { E T3, T11, Td, T13, T6, T2S, Ta, T12, T14, T2R; { E T1, T2, Tb, Tc; T1 = R0[0]; T2 = R0[WS(rs, 16)]; T3 = T1 + T2; T11 = T1 - T2; Tb = R0[WS(rs, 28)]; Tc = R0[WS(rs, 12)]; Td = Tb + Tc; T13 = Tb - Tc; } { E T4, T5, T8, T9; T4 = R0[WS(rs, 8)]; T5 = R0[WS(rs, 24)]; T6 = T4 + T5; T2S = T4 - T5; T8 = R0[WS(rs, 4)]; T9 = R0[WS(rs, 20)]; Ta = T8 + T9; T12 = T8 - T9; } T4l = T3 - T6; T5a = Td - Ta; T14 = KP707106781 * (T12 + T13); T15 = T11 + T14; T3n = T11 - T14; T2R = KP707106781 * (T13 - T12); T2T = T2R - T2S; T3Q = T2S + T2R; T7 = T3 + T6; Te = Ta + Td; Tf = T7 + Te; } { E T1P, T4J, T21, T4y, T1S, T4K, T1W, T4z; { E T1N, T1O, T1Z, T20; T1N = R1[WS(rs, 28)]; T1O = R1[WS(rs, 12)]; T1P = T1N - T1O; T4J = T1N + T1O; T1Z = R1[0]; T20 = R1[WS(rs, 16)]; T21 = T1Z - T20; T4y = T1Z + T20; } { E T1Q, T1R, T1U, T1V; T1Q = R1[WS(rs, 4)]; T1R = R1[WS(rs, 20)]; T1S = T1Q - T1R; T4K = T1Q + T1R; T1U = R1[WS(rs, 8)]; T1V = R1[WS(rs, 24)]; T1W = T1U - T1V; T4z = T1U + T1V; } T4A = T4y - T4z; T4L = T4J - T4K; { E T1T, T22, T5G, T5H; T1T = KP707106781 * (T1P - T1S); T1X = T1T - T1W; T3B = T1W + T1T; T22 = KP707106781 * (T1S + T1P); T23 = T21 + T22; T3y = T21 - T22; T5G = T4y + T4z; T5H = T4K + T4J; T5I = T5G + T5H; T66 = T5G - T5H; } } { E T2b, T4P, T2G, T4Q, T2e, T51, T2h, T50; { E T29, T2a, T2E, T2F; T29 = R1[WS(rs, 31)]; T2a = R1[WS(rs, 15)]; T2b = T29 - T2a; T4P = T29 + T2a; T2E = R1[WS(rs, 7)]; T2F = R1[WS(rs, 23)]; T2G = T2E - T2F; T4Q = T2E + T2F; } { E T2c, T2d, T2f, T2g; T2c = R1[WS(rs, 3)]; T2d = R1[WS(rs, 19)]; T2e = T2c - T2d; T51 = T2c + T2d; T2f = R1[WS(rs, 27)]; T2g = R1[WS(rs, 11)]; T2h = T2f - T2g; T50 = T2f + T2g; } T4R = T4P - T4Q; T52 = T50 - T51; { E T2i, T2D, T5N, T5O; T2i = KP707106781 * (T2e + T2h); T2j = T2b + T2i; T3F = T2b - T2i; T2D = KP707106781 * (T2h - T2e); T2H = T2D - T2G; T3I = T2G + T2D; T5N = T4P + T4Q; T5O = T51 + T50; T5P = T5N + T5O; T69 = T5N - T5O; } } { E TN, T1e, TX, T1g, TQ, T1k, TU, T1f, T1h, T1j; { E TL, TM, TV, TW; TL = R0[WS(rs, 31)]; TM = R0[WS(rs, 15)]; TN = TL + TM; T1e = TL - TM; TV = R0[WS(rs, 27)]; TW = R0[WS(rs, 11)]; TX = TV + TW; T1g = TV - TW; } { E TO, TP, TS, TT; TO = R0[WS(rs, 7)]; TP = R0[WS(rs, 23)]; TQ = TO + TP; T1k = TO - TP; TS = R0[WS(rs, 3)]; TT = R0[WS(rs, 19)]; TU = TS + TT; T1f = TS - TT; } T1h = KP707106781 * (T1f + T1g); T1i = T1e + T1h; T3t = T1e - T1h; T1j = KP707106781 * (T1g - T1f); T1l = T1j - T1k; T3u = T1k + T1j; { E TR, TY, T4t, T4u; TR = TN + TQ; TY = TU + TX; TZ = TR + TY; T63 = TR - TY; T4t = TN - TQ; T4u = TX - TU; T4v = FNMS(KP382683432, T4u, KP923879532 * T4t); T58 = FMA(KP382683432, T4t, KP923879532 * T4u); } } { E Ty, T1s, TI, T1n, TB, T1q, TF, T1o, T1p, T1t; { E Tw, Tx, TG, TH; Tw = R0[WS(rs, 1)]; Tx = R0[WS(rs, 17)]; Ty = Tw + Tx; T1s = Tw - Tx; TG = R0[WS(rs, 29)]; TH = R0[WS(rs, 13)]; TI = TG + TH; T1n = TG - TH; } { E Tz, TA, TD, TE; Tz = R0[WS(rs, 9)]; TA = R0[WS(rs, 25)]; TB = Tz + TA; T1q = Tz - TA; TD = R0[WS(rs, 5)]; TE = R0[WS(rs, 21)]; TF = TD + TE; T1o = TD - TE; } T1p = KP707106781 * (T1n - T1o); T1r = T1p - T1q; T3r = T1q + T1p; T1t = KP707106781 * (T1o + T1n); T1u = T1s + T1t; T3q = T1s - T1t; { E TC, TJ, T4q, T4r; TC = Ty + TB; TJ = TF + TI; TK = TC + TJ; T62 = TC - TJ; T4q = Ty - TB; T4r = TI - TF; T4s = FMA(KP923879532, T4q, KP382683432 * T4r); T57 = FNMS(KP382683432, T4q, KP923879532 * T4r); } } { E Ti, T16, Ts, T1a, Tl, T17, Tp, T19, T4m, T4n; { E Tg, Th, Tq, Tr; Tg = R0[WS(rs, 2)]; Th = R0[WS(rs, 18)]; Ti = Tg + Th; T16 = Tg - Th; Tq = R0[WS(rs, 6)]; Tr = R0[WS(rs, 22)]; Ts = Tq + Tr; T1a = Tq - Tr; } { E Tj, Tk, Tn, To; Tj = R0[WS(rs, 10)]; Tk = R0[WS(rs, 26)]; Tl = Tj + Tk; T17 = Tj - Tk; Tn = R0[WS(rs, 30)]; To = R0[WS(rs, 14)]; Tp = Tn + To; T19 = Tn - To; } Tm = Ti + Tl; Tt = Tp + Ts; Tu = Tm + Tt; T4m = Ti - Tl; T4n = Tp - Ts; T4o = KP707106781 * (T4m + T4n); T5b = KP707106781 * (T4n - T4m); { E T18, T1b, T2O, T2P; T18 = FNMS(KP382683432, T17, KP923879532 * T16); T1b = FMA(KP923879532, T19, KP382683432 * T1a); T1c = T18 + T1b; T3R = T1b - T18; T2O = FNMS(KP923879532, T1a, KP382683432 * T19); T2P = FMA(KP382683432, T16, KP923879532 * T17); T2Q = T2O - T2P; T3o = T2P + T2O; } } { E T1A, T4E, T1K, T4C, T1D, T4F, T1H, T4B; { E T1y, T1z, T1I, T1J; T1y = R1[WS(rs, 30)]; T1z = R1[WS(rs, 14)]; T1A = T1y - T1z; T4E = T1y + T1z; T1I = R1[WS(rs, 10)]; T1J = R1[WS(rs, 26)]; T1K = T1I - T1J; T4C = T1I + T1J; } { E T1B, T1C, T1F, T1G; T1B = R1[WS(rs, 6)]; T1C = R1[WS(rs, 22)]; T1D = T1B - T1C; T4F = T1B + T1C; T1F = R1[WS(rs, 2)]; T1G = R1[WS(rs, 18)]; T1H = T1F - T1G; T4B = T1F + T1G; } { E T1E, T1L, T5J, T5K; T1E = FNMS(KP923879532, T1D, KP382683432 * T1A); T1L = FMA(KP382683432, T1H, KP923879532 * T1K); T1M = T1E - T1L; T3z = T1L + T1E; T5J = T4B + T4C; T5K = T4E + T4F; T5L = T5J + T5K; T67 = T5K - T5J; } { E T24, T25, T4D, T4G; T24 = FNMS(KP382683432, T1K, KP923879532 * T1H); T25 = FMA(KP923879532, T1A, KP382683432 * T1D); T26 = T24 + T25; T3C = T25 - T24; T4D = T4B - T4C; T4G = T4E - T4F; T4H = KP707106781 * (T4D + T4G); T4M = KP707106781 * (T4G - T4D); } } { E T2m, T4S, T2w, T4W, T2p, T4T, T2t, T4V; { E T2k, T2l, T2u, T2v; T2k = R1[WS(rs, 1)]; T2l = R1[WS(rs, 17)]; T2m = T2k - T2l; T4S = T2k + T2l; T2u = R1[WS(rs, 5)]; T2v = R1[WS(rs, 21)]; T2w = T2u - T2v; T4W = T2u + T2v; } { E T2n, T2o, T2r, T2s; T2n = R1[WS(rs, 9)]; T2o = R1[WS(rs, 25)]; T2p = T2n - T2o; T4T = T2n + T2o; T2r = R1[WS(rs, 29)]; T2s = R1[WS(rs, 13)]; T2t = T2r - T2s; T4V = T2r + T2s; } { E T2q, T2x, T5Q, T5R; T2q = FNMS(KP382683432, T2p, KP923879532 * T2m); T2x = FMA(KP923879532, T2t, KP382683432 * T2w); T2y = T2q + T2x; T3J = T2x - T2q; T5Q = T4S + T4T; T5R = T4V + T4W; T5S = T5Q + T5R; T6a = T5R - T5Q; } { E T2A, T2B, T4U, T4X; T2A = FNMS(KP923879532, T2w, KP382683432 * T2t); T2B = FMA(KP382683432, T2m, KP923879532 * T2p); T2C = T2A - T2B; T3G = T2B + T2A; T4U = T4S - T4T; T4X = T4V - T4W; T4Y = KP707106781 * (T4U + T4X); T53 = KP707106781 * (T4X - T4U); } } { E Tv, T10, T5X, T5Y, T5Z, T60; Tv = Tf + Tu; T10 = TK + TZ; T5X = Tv + T10; T5Y = T5I + T5L; T5Z = T5P + T5S; T60 = T5Y + T5Z; Cr[WS(csr, 16)] = Tv - T10; Ci[WS(csi, 16)] = T5Z - T5Y; Cr[WS(csr, 32)] = T5X - T60; Cr[0] = T5X + T60; } { E T5F, T5V, T5U, T5W, T5M, T5T; T5F = Tf - Tu; T5V = TZ - TK; T5M = T5I - T5L; T5T = T5P - T5S; T5U = KP707106781 * (T5M + T5T); T5W = KP707106781 * (T5T - T5M); Cr[WS(csr, 24)] = T5F - T5U; Ci[WS(csi, 24)] = T5W - T5V; Cr[WS(csr, 8)] = T5F + T5U; Ci[WS(csi, 8)] = T5V + T5W; } { E T65, T6l, T6k, T6m, T6c, T6g, T6f, T6h; { E T61, T64, T6i, T6j; T61 = T7 - Te; T64 = KP707106781 * (T62 + T63); T65 = T61 + T64; T6l = T61 - T64; T6i = FNMS(KP382683432, T66, KP923879532 * T67); T6j = FMA(KP382683432, T69, KP923879532 * T6a); T6k = T6i + T6j; T6m = T6j - T6i; } { E T68, T6b, T6d, T6e; T68 = FMA(KP923879532, T66, KP382683432 * T67); T6b = FNMS(KP382683432, T6a, KP923879532 * T69); T6c = T68 + T6b; T6g = T6b - T68; T6d = KP707106781 * (T63 - T62); T6e = Tt - Tm; T6f = T6d - T6e; T6h = T6e + T6d; } Cr[WS(csr, 28)] = T65 - T6c; Ci[WS(csi, 28)] = T6k - T6h; Cr[WS(csr, 4)] = T65 + T6c; Ci[WS(csi, 4)] = T6h + T6k; Ci[WS(csi, 12)] = T6f + T6g; Cr[WS(csr, 12)] = T6l + T6m; Ci[WS(csi, 20)] = T6g - T6f; Cr[WS(csr, 20)] = T6l - T6m; } { E T5n, T5D, T5x, T5z, T5q, T5A, T5t, T5B; { E T5l, T5m, T5v, T5w; T5l = T4l - T4o; T5m = T58 - T57; T5n = T5l + T5m; T5D = T5l - T5m; T5v = T4v - T4s; T5w = T5b - T5a; T5x = T5v - T5w; T5z = T5w + T5v; } { E T5o, T5p, T5r, T5s; T5o = T4A - T4H; T5p = T4M - T4L; T5q = FMA(KP831469612, T5o, KP555570233 * T5p); T5A = FNMS(KP555570233, T5o, KP831469612 * T5p); T5r = T4R - T4Y; T5s = T53 - T52; T5t = FNMS(KP555570233, T5s, KP831469612 * T5r); T5B = FMA(KP555570233, T5r, KP831469612 * T5s); } { E T5u, T5C, T5y, T5E; T5u = T5q + T5t; Cr[WS(csr, 26)] = T5n - T5u; Cr[WS(csr, 6)] = T5n + T5u; T5C = T5A + T5B; Ci[WS(csi, 6)] = T5z + T5C; Ci[WS(csi, 26)] = T5C - T5z; T5y = T5t - T5q; Ci[WS(csi, 10)] = T5x + T5y; Ci[WS(csi, 22)] = T5y - T5x; T5E = T5B - T5A; Cr[WS(csr, 22)] = T5D - T5E; Cr[WS(csr, 10)] = T5D + T5E; } } { E T4x, T5j, T5d, T5f, T4O, T5g, T55, T5h; { E T4p, T4w, T59, T5c; T4p = T4l + T4o; T4w = T4s + T4v; T4x = T4p + T4w; T5j = T4p - T4w; T59 = T57 + T58; T5c = T5a + T5b; T5d = T59 - T5c; T5f = T5c + T59; } { E T4I, T4N, T4Z, T54; T4I = T4A + T4H; T4N = T4L + T4M; T4O = FMA(KP980785280, T4I, KP195090322 * T4N); T5g = FNMS(KP195090322, T4I, KP980785280 * T4N); T4Z = T4R + T4Y; T54 = T52 + T53; T55 = FNMS(KP195090322, T54, KP980785280 * T4Z); T5h = FMA(KP195090322, T4Z, KP980785280 * T54); } { E T56, T5i, T5e, T5k; T56 = T4O + T55; Cr[WS(csr, 30)] = T4x - T56; Cr[WS(csr, 2)] = T4x + T56; T5i = T5g + T5h; Ci[WS(csi, 2)] = T5f + T5i; Ci[WS(csi, 30)] = T5i - T5f; T5e = T55 - T4O; Ci[WS(csi, 14)] = T5d + T5e; Ci[WS(csi, 18)] = T5e - T5d; T5k = T5h - T5g; Cr[WS(csr, 18)] = T5j - T5k; Cr[WS(csr, 14)] = T5j + T5k; } } { E T3p, T41, T4c, T3S, T3w, T4b, T49, T4h, T3P, T42, T3E, T3W, T46, T4g, T3L; E T3X; { E T3s, T3v, T3A, T3D; T3p = T3n + T3o; T41 = T3n - T3o; T4c = T3R - T3Q; T3S = T3Q + T3R; T3s = FMA(KP831469612, T3q, KP555570233 * T3r); T3v = FNMS(KP555570233, T3u, KP831469612 * T3t); T3w = T3s + T3v; T4b = T3v - T3s; { E T47, T48, T3N, T3O; T47 = T3F - T3G; T48 = T3J - T3I; T49 = FNMS(KP471396736, T48, KP881921264 * T47); T4h = FMA(KP471396736, T47, KP881921264 * T48); T3N = FNMS(KP555570233, T3q, KP831469612 * T3r); T3O = FMA(KP555570233, T3t, KP831469612 * T3u); T3P = T3N + T3O; T42 = T3O - T3N; } T3A = T3y + T3z; T3D = T3B + T3C; T3E = FMA(KP956940335, T3A, KP290284677 * T3D); T3W = FNMS(KP290284677, T3A, KP956940335 * T3D); { E T44, T45, T3H, T3K; T44 = T3y - T3z; T45 = T3C - T3B; T46 = FMA(KP881921264, T44, KP471396736 * T45); T4g = FNMS(KP471396736, T44, KP881921264 * T45); T3H = T3F + T3G; T3K = T3I + T3J; T3L = FNMS(KP290284677, T3K, KP956940335 * T3H); T3X = FMA(KP290284677, T3H, KP956940335 * T3K); } } { E T3x, T3M, T3V, T3Y; T3x = T3p + T3w; T3M = T3E + T3L; Cr[WS(csr, 29)] = T3x - T3M; Cr[WS(csr, 3)] = T3x + T3M; T3V = T3S + T3P; T3Y = T3W + T3X; Ci[WS(csi, 3)] = T3V + T3Y; Ci[WS(csi, 29)] = T3Y - T3V; } { E T3T, T3U, T3Z, T40; T3T = T3P - T3S; T3U = T3L - T3E; Ci[WS(csi, 13)] = T3T + T3U; Ci[WS(csi, 19)] = T3U - T3T; T3Z = T3p - T3w; T40 = T3X - T3W; Cr[WS(csr, 19)] = T3Z - T40; Cr[WS(csr, 13)] = T3Z + T40; } { E T43, T4a, T4f, T4i; T43 = T41 + T42; T4a = T46 + T49; Cr[WS(csr, 27)] = T43 - T4a; Cr[WS(csr, 5)] = T43 + T4a; T4f = T4c + T4b; T4i = T4g + T4h; Ci[WS(csi, 5)] = T4f + T4i; Ci[WS(csi, 27)] = T4i - T4f; } { E T4d, T4e, T4j, T4k; T4d = T4b - T4c; T4e = T49 - T46; Ci[WS(csi, 11)] = T4d + T4e; Ci[WS(csi, 21)] = T4e - T4d; T4j = T41 - T42; T4k = T4h - T4g; Cr[WS(csr, 21)] = T4j - T4k; Cr[WS(csr, 11)] = T4j + T4k; } } { E T1d, T33, T3e, T2U, T1w, T3d, T3b, T3j, T2N, T34, T28, T2Y, T38, T3i, T2J; E T2Z; { E T1m, T1v, T1Y, T27; T1d = T15 - T1c; T33 = T15 + T1c; T3e = T2T + T2Q; T2U = T2Q - T2T; T1m = FMA(KP195090322, T1i, KP980785280 * T1l); T1v = FNMS(KP195090322, T1u, KP980785280 * T1r); T1w = T1m - T1v; T3d = T1v + T1m; { E T39, T3a, T2L, T2M; T39 = T2j + T2y; T3a = T2H + T2C; T3b = FNMS(KP098017140, T3a, KP995184726 * T39); T3j = FMA(KP995184726, T3a, KP098017140 * T39); T2L = FNMS(KP195090322, T1l, KP980785280 * T1i); T2M = FMA(KP980785280, T1u, KP195090322 * T1r); T2N = T2L - T2M; T34 = T2M + T2L; } T1Y = T1M - T1X; T27 = T23 - T26; T28 = FMA(KP634393284, T1Y, KP773010453 * T27); T2Y = FNMS(KP634393284, T27, KP773010453 * T1Y); { E T36, T37, T2z, T2I; T36 = T1X + T1M; T37 = T23 + T26; T38 = FMA(KP098017140, T36, KP995184726 * T37); T3i = FNMS(KP098017140, T37, KP995184726 * T36); T2z = T2j - T2y; T2I = T2C - T2H; T2J = FNMS(KP634393284, T2I, KP773010453 * T2z); T2Z = FMA(KP773010453, T2I, KP634393284 * T2z); } } { E T1x, T2K, T2X, T30; T1x = T1d + T1w; T2K = T28 + T2J; Cr[WS(csr, 25)] = T1x - T2K; Cr[WS(csr, 7)] = T1x + T2K; T2X = T2U + T2N; T30 = T2Y + T2Z; Ci[WS(csi, 7)] = T2X + T30; Ci[WS(csi, 25)] = T30 - T2X; } { E T2V, T2W, T31, T32; T2V = T2N - T2U; T2W = T2J - T28; Ci[WS(csi, 9)] = T2V + T2W; Ci[WS(csi, 23)] = T2W - T2V; T31 = T1d - T1w; T32 = T2Z - T2Y; Cr[WS(csr, 23)] = T31 - T32; Cr[WS(csr, 9)] = T31 + T32; } { E T35, T3c, T3h, T3k; T35 = T33 + T34; T3c = T38 + T3b; Cr[WS(csr, 31)] = T35 - T3c; Cr[WS(csr, 1)] = T35 + T3c; T3h = T3e + T3d; T3k = T3i + T3j; Ci[WS(csi, 1)] = T3h + T3k; Ci[WS(csi, 31)] = T3k - T3h; } { E T3f, T3g, T3l, T3m; T3f = T3d - T3e; T3g = T3b - T38; Ci[WS(csi, 15)] = T3f + T3g; Ci[WS(csi, 17)] = T3g - T3f; T3l = T33 - T34; T3m = T3j - T3i; Cr[WS(csr, 17)] = T3l - T3m; Cr[WS(csr, 15)] = T3l + T3m; } } } } } static const kr2c_desc desc = { 64, "r2cf_64", {342, 72, 52, 0}, &GENUS }; void X(codelet_r2cf_64) (planner *p) { X(kr2c_register) (p, r2cf_64, &desc); } #endif /* HAVE_FMA */ fftw-3.3.3/rdft/scalar/r2cf/r2cf_128.c0000644000175000017500000027255012054410734014054 00000000000000/* * Copyright (c) 2003, 2007-11 Matteo Frigo * Copyright (c) 2003, 2007-11 Massachusetts Institute of Technology * * This program is free software; you can redistribute it and/or modify * it under the terms of the GNU General Public License as published by * the Free Software Foundation; either version 2 of the License, or * (at your option) any later version. * * This program is distributed in the hope that it will be useful, * but WITHOUT ANY WARRANTY; without even the implied warranty of * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the * GNU General Public License for more details. * * You should have received a copy of the GNU General Public License * along with this program; if not, write to the Free Software * Foundation, Inc., 51 Franklin Street, Fifth Floor, Boston, MA 02110-1301 USA * */ /* This file was automatically generated --- DO NOT EDIT */ /* Generated on Sun Nov 25 07:39:47 EST 2012 */ #include "codelet-rdft.h" #ifdef HAVE_FMA /* Generated by: ../../../genfft/gen_r2cf.native -fma -reorder-insns -schedule-for-pipeline -compact -variables 4 -pipeline-latency 4 -n 128 -name r2cf_128 -include r2cf.h */ /* * This function contains 956 FP additions, 516 FP multiplications, * (or, 440 additions, 0 multiplications, 516 fused multiply/add), * 229 stack variables, 31 constants, and 256 memory accesses */ #include "r2cf.h" static void r2cf_128(R *R0, R *R1, R *Cr, R *Ci, stride rs, stride csr, stride csi, INT v, INT ivs, INT ovs) { DK(KP989176509, +0.989176509964780973451673738016243063983689533); DK(KP803207531, +0.803207531480644909806676512963141923879569427); DK(KP148335987, +0.148335987538347428753676511486911367000625355); DK(KP741650546, +0.741650546272035369581266691172079863842265220); DK(KP998795456, +0.998795456205172392714771604759100694443203615); DK(KP740951125, +0.740951125354959091175616897495162729728955309); DK(KP049126849, +0.049126849769467254105343321271313617079695752); DK(KP906347169, +0.906347169019147157946142717268914412664134293); DK(KP857728610, +0.857728610000272069902269984284770137042490799); DK(KP970031253, +0.970031253194543992603984207286100251456865962); DK(KP599376933, +0.599376933681923766271389869014404232837890546); DK(KP250486960, +0.250486960191305461595702160124721208578685568); DK(KP941544065, +0.941544065183020778412509402599502357185589796); DK(KP903989293, +0.903989293123443331586200297230537048710132025); DK(KP472964775, +0.472964775891319928124438237972992463904131113); DK(KP357805721, +0.357805721314524104672487743774474392487532769); DK(KP773010453, +0.773010453362736960810906609758469800971041293); DK(KP995184726, +0.995184726672196886244836953109479921575474869); DK(KP098491403, +0.098491403357164253077197521291327432293052451); DK(KP820678790, +0.820678790828660330972281985331011598767386482); DK(KP956940335, +0.956940335732208864935797886980269969482849206); DK(KP881921264, +0.881921264348355029712756863660388349508442621); DK(KP534511135, +0.534511135950791641089685961295362908582039528); DK(KP303346683, +0.303346683607342391675883946941299872384187453); DK(KP831469612, +0.831469612302545237078788377617905756738560812); DK(KP980785280, +0.980785280403230449126182236134239036973933731); DK(KP923879532, +0.923879532511286756128183189396788286822416626); DK(KP198912367, +0.198912367379658006911597622644676228597850501); DK(KP668178637, +0.668178637919298919997757686523080761552472251); DK(KP414213562, +0.414213562373095048801688724209698078569671875); DK(KP707106781, +0.707106781186547524400844362104849039284835938); { INT i; for (i = v; i > 0; i = i - 1, R0 = R0 + ivs, R1 = R1 + ivs, Cr = Cr + ovs, Ci = Ci + ovs, MAKE_VOLATILE_STRIDE(512, rs), MAKE_VOLATILE_STRIDE(512, csr), MAKE_VOLATILE_STRIDE(512, csi)) { E T95, T96; { E TcD, TdR, T5P, T8v, T27, T7r, Tf, Ta5, T7s, T5S, T8w, T2e, TdS, TcG, Tbn; E Tu, TcK, TdU, TK, Ta6, T7w, T8y, T2o, T5U, TcN, TdV, TZ, Ta7, T7z, T8z; E T2x, T5V, T1g, Taa, Tab, T1v, Tew, TcX, Tex, TcU, T6A, T2M, T9b, T7E, T9a; E T7H, T6z, T2T, TeO, TdK, TeL, Tdz, T9p, T8d, T6O, T5G, T6L, T4X, Tc3, TaV; E Tc4, Tbi, T9s, T8o, TeH, Tdp, TeE, Tde, T9i, T7U, T6H, T4r, T6E, T3I, TbW; E Tao, TbX, TaL, T9l, T85, T1L, Tad, Tae, T20, Tez, Td6, TeA, Td3, T6x, T37; E T9e, T7L, T9d, T7O, T6w, T3e, TbZ, T3Z, T4s, Tc0, TeF, Tds, T4t, T4g, T87; E T80, TeI, Tdl, T86, T7X, TaM, TaD, Tb2, Tc6, T8e, T8f, T5e, T5H, Tb9, Tc7; E TeM, TdN, T5I, T5v, T8q, T8j, TeP, TdG; { E T7G, T2S, T2P, T7F; { E T28, Ti, Tn, T2c, Ts, T29, Tl, To; { E T4, T23, T3, T25, Td, T5, T8, T9; { E T1, T2, Tb, Tc; T1 = R0[0]; T2 = R0[WS(rs, 32)]; Tb = R0[WS(rs, 56)]; Tc = R0[WS(rs, 24)]; T4 = R0[WS(rs, 16)]; T23 = T1 - T2; T3 = T1 + T2; T25 = Tb - Tc; Td = Tb + Tc; T5 = R0[WS(rs, 48)]; T8 = R0[WS(rs, 8)]; T9 = R0[WS(rs, 40)]; } { E Tq, Tr, Tj, Tk; { E Tg, T5N, T6, T24, Ta, Th; Tg = R0[WS(rs, 4)]; T5N = T4 - T5; T6 = T4 + T5; T24 = T8 - T9; Ta = T8 + T9; Th = R0[WS(rs, 36)]; { E T7, T26, T5O, Te; TcD = T3 - T6; T7 = T3 + T6; T26 = T24 + T25; T5O = T25 - T24; TdR = Td - Ta; Te = Ta + Td; T5P = FNMS(KP707106781, T5O, T5N); T8v = FMA(KP707106781, T5O, T5N); T27 = FMA(KP707106781, T26, T23); T7r = FNMS(KP707106781, T26, T23); Tf = T7 + Te; Ta5 = T7 - Te; T28 = Tg - Th; Ti = Tg + Th; } } Tq = R0[WS(rs, 12)]; Tr = R0[WS(rs, 44)]; Tj = R0[WS(rs, 20)]; Tk = R0[WS(rs, 52)]; Tn = R0[WS(rs, 60)]; T2c = Tq - Tr; Ts = Tq + Tr; T29 = Tj - Tk; Tl = Tj + Tk; To = R0[WS(rs, 28)]; } } { E T2g, T2l, T2h, TF, TcI, TC, T2i, TI; { E Ty, TG, TB, TH; { E Tw, T5Q, T2a, TcE, Tm, T2b, Tp, Tx; Tw = R0[WS(rs, 2)]; T5Q = FMA(KP414213562, T28, T29); T2a = FNMS(KP414213562, T29, T28); TcE = Ti - Tl; Tm = Ti + Tl; T2b = Tn - To; Tp = Tn + To; Tx = R0[WS(rs, 34)]; { E Tz, TA, TD, TE; Tz = R0[WS(rs, 18)]; { E T5R, T2d, TcF, Tt; T5R = FNMS(KP414213562, T2b, T2c); T2d = FMA(KP414213562, T2c, T2b); TcF = Tp - Ts; Tt = Tp + Ts; T2g = Tw - Tx; Ty = Tw + Tx; T7s = T5Q - T5R; T5S = T5Q + T5R; T8w = T2d - T2a; T2e = T2a + T2d; TdS = TcF - TcE; TcG = TcE + TcF; Tbn = Tt - Tm; Tu = Tm + Tt; TA = R0[WS(rs, 50)]; } TD = R0[WS(rs, 10)]; TE = R0[WS(rs, 42)]; TG = R0[WS(rs, 58)]; T2l = Tz - TA; TB = Tz + TA; T2h = TD - TE; TF = TD + TE; TH = R0[WS(rs, 26)]; } } TcI = Ty - TB; TC = Ty + TB; T2i = TG - TH; TI = TG + TH; } { E T2p, T2u, T2q, TU, TcL, TR, T2r, TX; { E TN, TV, TQ, TW; { E T2k, T7u, T2n, T7v, TL, TM; TL = R0[WS(rs, 62)]; TM = R0[WS(rs, 30)]; { E TJ, TcJ, T2m, T2j; TJ = TF + TI; TcJ = TI - TF; T2m = T2h - T2i; T2j = T2h + T2i; TcK = FMA(KP414213562, TcJ, TcI); TdU = FNMS(KP414213562, TcI, TcJ); TK = TC + TJ; Ta6 = TC - TJ; T2k = FMA(KP707106781, T2j, T2g); T7u = FNMS(KP707106781, T2j, T2g); T2n = FMA(KP707106781, T2m, T2l); T7v = FNMS(KP707106781, T2m, T2l); T2p = TL - TM; TN = TL + TM; } T7w = FMA(KP668178637, T7v, T7u); T8y = FNMS(KP668178637, T7u, T7v); T2o = FNMS(KP198912367, T2n, T2k); T5U = FMA(KP198912367, T2k, T2n); { E TO, TP, TS, TT; TO = R0[WS(rs, 14)]; TP = R0[WS(rs, 46)]; TS = R0[WS(rs, 6)]; TT = R0[WS(rs, 38)]; TV = R0[WS(rs, 54)]; T2u = TO - TP; TQ = TO + TP; T2q = TS - TT; TU = TS + TT; TW = R0[WS(rs, 22)]; } } TcL = TN - TQ; TR = TN + TQ; T2r = TV - TW; TX = TV + TW; } { E T2A, T14, T2N, T17, T1b, T1e, T2D, T2O, T1r, T2I, T1q, T2Q, T2H, TcR, T1n; E T1s, T15, T16; { E T2t, T7x, T2w, T7y, T12, T13; T12 = R0[WS(rs, 1)]; T13 = R0[WS(rs, 33)]; { E TY, TcM, T2v, T2s; TY = TU + TX; TcM = TX - TU; T2v = T2q - T2r; T2s = T2q + T2r; TcN = FNMS(KP414213562, TcM, TcL); TdV = FMA(KP414213562, TcL, TcM); TZ = TR + TY; Ta7 = TR - TY; T2t = FMA(KP707106781, T2s, T2p); T7x = FNMS(KP707106781, T2s, T2p); T2w = FMA(KP707106781, T2v, T2u); T7y = FNMS(KP707106781, T2v, T2u); T2A = T12 - T13; T14 = T12 + T13; } T7z = FNMS(KP668178637, T7y, T7x); T8z = FMA(KP668178637, T7x, T7y); T2x = FMA(KP198912367, T2w, T2t); T5V = FNMS(KP198912367, T2t, T2w); T15 = R0[WS(rs, 17)]; T16 = R0[WS(rs, 49)]; } { E T1c, T2B, T1d, T19, T1a; T19 = R0[WS(rs, 9)]; T1a = R0[WS(rs, 41)]; T1c = R0[WS(rs, 57)]; T2N = T15 - T16; T17 = T15 + T16; T2B = T19 - T1a; T1b = T19 + T1a; T1d = R0[WS(rs, 25)]; { E T1k, T2F, T1j, T1l, T1h, T1i, T2C; T1h = R0[WS(rs, 5)]; T1i = R0[WS(rs, 37)]; T2C = T1c - T1d; T1e = T1c + T1d; T1k = R0[WS(rs, 21)]; T2F = T1h - T1i; T1j = T1h + T1i; T2D = T2B + T2C; T2O = T2B - T2C; T1l = R0[WS(rs, 53)]; { E T1o, T1p, T2G, T1m; T1o = R0[WS(rs, 61)]; T1p = R0[WS(rs, 29)]; T1r = R0[WS(rs, 13)]; T2G = T1k - T1l; T1m = T1k + T1l; T2I = T1o - T1p; T1q = T1o + T1p; T2Q = FMA(KP414213562, T2F, T2G); T2H = FNMS(KP414213562, T2G, T2F); TcR = T1j - T1m; T1n = T1j + T1m; T1s = R0[WS(rs, 45)]; } } } { E TcQ, TcV, T2K, T2R, T1u, TcT, TcW, TcS; { E T18, T1f, T1t, T2J; T18 = T14 + T17; TcQ = T14 - T17; TcV = T1e - T1b; T1f = T1b + T1e; T1t = T1r + T1s; T2J = T1r - T1s; T1g = T18 + T1f; Taa = T18 - T1f; T2K = FMA(KP414213562, T2J, T2I); T2R = FNMS(KP414213562, T2I, T2J); T1u = T1q + T1t; TcS = T1q - T1t; } TcT = TcR + TcS; TcW = TcS - TcR; { E T7C, T2E, T2L, T7D; T7C = FNMS(KP707106781, T2D, T2A); T2E = FMA(KP707106781, T2D, T2A); Tab = T1u - T1n; T1v = T1n + T1u; Tew = FNMS(KP707106781, TcW, TcV); TcX = FMA(KP707106781, TcW, TcV); Tex = FNMS(KP707106781, TcT, TcQ); TcU = FMA(KP707106781, TcT, TcQ); T2L = T2H + T2K; T7G = T2K - T2H; T7D = T2Q - T2R; T2S = T2Q + T2R; T2P = FMA(KP707106781, T2O, T2N); T7F = FNMS(KP707106781, T2O, T2N); T6A = FNMS(KP923879532, T2L, T2E); T2M = FMA(KP923879532, T2L, T2E); T9b = FNMS(KP923879532, T7D, T7C); T7E = FMA(KP923879532, T7D, T7C); } } } } } } { E T83, T84, T8m, T8n; { E TaP, T4z, TaQ, T5A, TaS, TaT, T4G, T5B, T4O, T5D, Tbh, Tdw, T4R, Tbc, T4S; E T4T; { E T4x, T4y, T5y, T5z; T4x = R1[WS(rs, 63)]; T9a = FNMS(KP923879532, T7G, T7F); T7H = FMA(KP923879532, T7G, T7F); T6z = FNMS(KP923879532, T2S, T2P); T2T = FMA(KP923879532, T2S, T2P); T4y = R1[WS(rs, 31)]; T5y = R1[WS(rs, 47)]; T5z = R1[WS(rs, 15)]; { E T4A, T4B, T4D, T4E; T4A = R1[WS(rs, 7)]; TaP = T4x + T4y; T4z = T4x - T4y; TaQ = T5z + T5y; T5A = T5y - T5z; T4B = R1[WS(rs, 39)]; T4D = R1[WS(rs, 55)]; T4E = R1[WS(rs, 23)]; { E T4K, Tbf, Tbg, T4N, T4P, T4Q; { E T4I, T4C, T4F, T4J, T4L, T4M; T4I = R1[WS(rs, 3)]; TaS = T4A + T4B; T4C = T4A - T4B; TaT = T4D + T4E; T4F = T4D - T4E; T4J = R1[WS(rs, 35)]; T4L = R1[WS(rs, 51)]; T4M = R1[WS(rs, 19)]; T4G = T4C + T4F; T5B = T4F - T4C; T4K = T4I - T4J; Tbf = T4I + T4J; Tbg = T4M + T4L; T4N = T4L - T4M; } T4P = R1[WS(rs, 59)]; T4Q = R1[WS(rs, 27)]; T4O = FMA(KP414213562, T4N, T4K); T5D = FNMS(KP414213562, T4K, T4N); Tbh = Tbf + Tbg; Tdw = Tbf - Tbg; T4R = T4P - T4Q; Tbc = T4P + T4Q; T4S = R1[WS(rs, 43)]; T4T = R1[WS(rs, 11)]; } } } { E T4H, T8b, TaR, Tdv, TdI, TaU, T4U, Tbd, T5C; T4H = FMA(KP707106781, T4G, T4z); T8b = FNMS(KP707106781, T4G, T4z); TaR = TaP + TaQ; Tdv = TaP - TaQ; TdI = TaT - TaS; TaU = TaS + TaT; T4U = T4S - T4T; Tbd = T4T + T4S; T8m = FNMS(KP707106781, T5B, T5A); T5C = FMA(KP707106781, T5B, T5A); { E Tbe, Tdx, T5E, T4V; Tbe = Tbc + Tbd; Tdx = Tbc - Tbd; T5E = FMA(KP414213562, T4R, T4U); T4V = FNMS(KP414213562, T4U, T4R); { E Tdy, TdJ, T5F, T8c, T4W; Tdy = Tdw + Tdx; TdJ = Tdx - Tdw; T5F = T5D + T5E; T8c = T5E - T5D; T8n = T4V - T4O; T4W = T4O + T4V; TeO = FNMS(KP707106781, TdJ, TdI); TdK = FMA(KP707106781, TdJ, TdI); TeL = FNMS(KP707106781, Tdy, Tdv); Tdz = FMA(KP707106781, Tdy, Tdv); T9p = FNMS(KP923879532, T8c, T8b); T8d = FMA(KP923879532, T8c, T8b); T6O = FNMS(KP923879532, T5F, T5C); T5G = FMA(KP923879532, T5F, T5C); T6L = FNMS(KP923879532, T4W, T4H); T4X = FMA(KP923879532, T4W, T4H); } Tc3 = TaR + TaU; TaV = TaR - TaU; Tc4 = Tbh + Tbe; Tbi = Tbe - Tbh; } } } { E Tai, T3k, Taj, T4l, Tal, Tam, T4m, T3r, T3D, TaF, T3C, Tdb, TaK, T3z, T4o; E T3E; { E T4j, T4k, T3i, T3j; T3i = R1[0]; T3j = R1[WS(rs, 32)]; T4j = R1[WS(rs, 16)]; T9s = FMA(KP923879532, T8n, T8m); T8o = FNMS(KP923879532, T8n, T8m); Tai = T3i + T3j; T3k = T3i - T3j; T4k = R1[WS(rs, 48)]; { E T3o, T3n, T3p, T3l, T3m; T3l = R1[WS(rs, 8)]; T3m = R1[WS(rs, 40)]; T3o = R1[WS(rs, 56)]; Taj = T4j + T4k; T4l = T4j - T4k; T3n = T3l - T3m; Tal = T3l + T3m; T3p = R1[WS(rs, 24)]; { E T3w, TaI, T3v, T3x, T3t, T3u, T3q; T3t = R1[WS(rs, 4)]; T3u = R1[WS(rs, 36)]; T3q = T3o - T3p; Tam = T3o + T3p; T3w = R1[WS(rs, 20)]; TaI = T3t + T3u; T3v = T3t - T3u; T4m = T3n - T3q; T3r = T3n + T3q; T3x = R1[WS(rs, 52)]; { E T3A, T3B, TaJ, T3y; T3A = R1[WS(rs, 60)]; T3B = R1[WS(rs, 28)]; T3D = R1[WS(rs, 12)]; TaJ = T3w + T3x; T3y = T3w - T3x; TaF = T3A + T3B; T3C = T3A - T3B; Tdb = TaI - TaJ; TaK = TaI + TaJ; T3z = FNMS(KP414213562, T3y, T3v); T4o = FMA(KP414213562, T3v, T3y); T3E = R1[WS(rs, 44)]; } } } } { E T3s, T7S, Tak, Tda, Tdn, Tan, T3F, TaG, T4n; T3s = FMA(KP707106781, T3r, T3k); T7S = FNMS(KP707106781, T3r, T3k); Tak = Tai + Taj; Tda = Tai - Taj; Tdn = Tam - Tal; Tan = Tal + Tam; T3F = T3D - T3E; TaG = T3D + T3E; T83 = FNMS(KP707106781, T4m, T4l); T4n = FMA(KP707106781, T4m, T4l); { E TaH, Tdc, T4p, T3G; TaH = TaF + TaG; Tdc = TaF - TaG; T4p = FNMS(KP414213562, T3C, T3F); T3G = FMA(KP414213562, T3F, T3C); { E Tdd, Tdo, T4q, T7T, T3H; Tdd = Tdb + Tdc; Tdo = Tdc - Tdb; T4q = T4o + T4p; T7T = T4o - T4p; T84 = T3G - T3z; T3H = T3z + T3G; TeH = FNMS(KP707106781, Tdo, Tdn); Tdp = FMA(KP707106781, Tdo, Tdn); TeE = FNMS(KP707106781, Tdd, Tda); Tde = FMA(KP707106781, Tdd, Tda); T9i = FNMS(KP923879532, T7T, T7S); T7U = FMA(KP923879532, T7T, T7S); T6H = FNMS(KP923879532, T4q, T4n); T4r = FMA(KP923879532, T4q, T4n); T6E = FNMS(KP923879532, T3H, T3s); T3I = FMA(KP923879532, T3H, T3s); } TbW = Tak + Tan; Tao = Tak - Tan; TbX = TaK + TaH; TaL = TaH - TaK; } } } { E T7N, T3d, T3a, T7M; { E T2V, T1z, T38, T1C, T1G, T1J, T2Y, T39, T1W, T33, T1V, T3b, T32, Td0, T1S; E T1X; { E T1A, T1B, T1x, T1y; T1x = R0[WS(rs, 63)]; T1y = R0[WS(rs, 31)]; T1A = R0[WS(rs, 15)]; T9l = FNMS(KP923879532, T84, T83); T85 = FMA(KP923879532, T84, T83); T2V = T1x - T1y; T1z = T1x + T1y; T1B = R0[WS(rs, 47)]; { E T1H, T2W, T1I, T1E, T1F; T1E = R0[WS(rs, 7)]; T1F = R0[WS(rs, 39)]; T1H = R0[WS(rs, 55)]; T38 = T1A - T1B; T1C = T1A + T1B; T2W = T1E - T1F; T1G = T1E + T1F; T1I = R0[WS(rs, 23)]; { E T1P, T30, T1O, T1Q, T1M, T1N, T2X; T1M = R0[WS(rs, 3)]; T1N = R0[WS(rs, 35)]; T2X = T1H - T1I; T1J = T1H + T1I; T1P = R0[WS(rs, 19)]; T30 = T1M - T1N; T1O = T1M + T1N; T2Y = T2W + T2X; T39 = T2W - T2X; T1Q = R0[WS(rs, 51)]; { E T1T, T1U, T31, T1R; T1T = R0[WS(rs, 59)]; T1U = R0[WS(rs, 27)]; T1W = R0[WS(rs, 11)]; T31 = T1P - T1Q; T1R = T1P + T1Q; T33 = T1T - T1U; T1V = T1T + T1U; T3b = FMA(KP414213562, T30, T31); T32 = FNMS(KP414213562, T31, T30); Td0 = T1O - T1R; T1S = T1O + T1R; T1X = R0[WS(rs, 43)]; } } } } { E TcZ, Td4, T35, T3c, T1Z, Td2, Td5, Td1; { E T1D, T1K, T1Y, T34; T1D = T1z + T1C; TcZ = T1z - T1C; Td4 = T1J - T1G; T1K = T1G + T1J; T1Y = T1W + T1X; T34 = T1W - T1X; T1L = T1D + T1K; Tad = T1D - T1K; T35 = FMA(KP414213562, T34, T33); T3c = FNMS(KP414213562, T33, T34); T1Z = T1V + T1Y; Td1 = T1V - T1Y; } Td2 = Td0 + Td1; Td5 = Td1 - Td0; { E T7J, T2Z, T36, T7K; T7J = FNMS(KP707106781, T2Y, T2V); T2Z = FMA(KP707106781, T2Y, T2V); Tae = T1Z - T1S; T20 = T1S + T1Z; Tez = FNMS(KP707106781, Td5, Td4); Td6 = FMA(KP707106781, Td5, Td4); TeA = FNMS(KP707106781, Td2, TcZ); Td3 = FMA(KP707106781, Td2, TcZ); T36 = T32 + T35; T7N = T35 - T32; T7K = T3b - T3c; T3d = T3b + T3c; T3a = FMA(KP707106781, T39, T38); T7M = FNMS(KP707106781, T39, T38); T6x = FNMS(KP923879532, T36, T2Z); T37 = FMA(KP923879532, T36, T2Z); T9e = FNMS(KP923879532, T7K, T7J); T7L = FMA(KP923879532, T7K, T7J); } } } { E Tav, T7V, T7W, TaC; { E T3L, T3W, Tdf, Tar, T42, T4d, Tay, Tdi, T46, Tau, Tdg, T3X, T3S, Taz, T45; E T47, Taw, Tax; { E T3J, T3K, T3U, T3V; T3J = R1[WS(rs, 2)]; T9d = FNMS(KP923879532, T7N, T7M); T7O = FMA(KP923879532, T7N, T7M); T6w = FNMS(KP923879532, T3d, T3a); T3e = FMA(KP923879532, T3d, T3a); T3K = R1[WS(rs, 34)]; T3U = R1[WS(rs, 18)]; T3V = R1[WS(rs, 50)]; { E T40, Tap, Taq, T41, T4b, T4c; T40 = R1[WS(rs, 62)]; T3L = T3J - T3K; Tap = T3J + T3K; T3W = T3U - T3V; Taq = T3U + T3V; T41 = R1[WS(rs, 30)]; T4b = R1[WS(rs, 14)]; T4c = R1[WS(rs, 46)]; Tdf = Tap - Taq; Tar = Tap + Taq; T42 = T40 - T41; Taw = T40 + T41; Tax = T4b + T4c; T4d = T4b - T4c; } } { E T3M, T3N, T3P, T3Q; T3M = R1[WS(rs, 10)]; Tay = Taw + Tax; Tdi = Taw - Tax; T3N = R1[WS(rs, 42)]; T3P = R1[WS(rs, 58)]; T3Q = R1[WS(rs, 26)]; { E T43, Tas, T3O, Tat, T3R, T44; T43 = R1[WS(rs, 6)]; Tas = T3M + T3N; T3O = T3M - T3N; Tat = T3P + T3Q; T3R = T3P - T3Q; T44 = R1[WS(rs, 38)]; T46 = R1[WS(rs, 54)]; Tau = Tas + Tat; Tdg = Tat - Tas; T3X = T3O - T3R; T3S = T3O + T3R; Taz = T43 + T44; T45 = T43 - T44; T47 = R1[WS(rs, 22)]; } } { E Tdq, Tdh, T49, T4e, Tdr, Tdk; Tav = Tar - Tau; TbZ = Tar + Tau; { E T3T, T3Y, TaA, T48, Tdj, TaB; T3T = FMA(KP707106781, T3S, T3L); T7V = FNMS(KP707106781, T3S, T3L); T7W = FNMS(KP707106781, T3X, T3W); T3Y = FMA(KP707106781, T3X, T3W); TaA = T46 + T47; T48 = T46 - T47; Tdq = FNMS(KP414213562, Tdf, Tdg); Tdh = FMA(KP414213562, Tdg, Tdf); T3Z = FNMS(KP198912367, T3Y, T3T); T4s = FMA(KP198912367, T3T, T3Y); Tdj = TaA - Taz; TaB = Taz + TaA; T49 = T45 + T48; T4e = T45 - T48; TaC = Tay - TaB; Tc0 = Tay + TaB; Tdr = FMA(KP414213562, Tdi, Tdj); Tdk = FNMS(KP414213562, Tdj, Tdi); } { E T7Z, T7Y, T4f, T4a; T7Z = FNMS(KP707106781, T4e, T4d); T4f = FMA(KP707106781, T4e, T4d); T4a = FMA(KP707106781, T49, T42); T7Y = FNMS(KP707106781, T49, T42); TeF = Tdr - Tdq; Tds = Tdq + Tdr; T4t = FNMS(KP198912367, T4a, T4f); T4g = FMA(KP198912367, T4f, T4a); T87 = FMA(KP668178637, T7Y, T7Z); T80 = FNMS(KP668178637, T7Z, T7Y); TeI = Tdh - Tdk; Tdl = Tdh + Tdk; } } } { E T50, T5b, TdA, TaY, T5h, T5s, Tb5, TdD, T5l, Tb1, TdB, T5c, T57, Tb6, T5k; E T5m, Tb3, Tb4; { E T4Y, T4Z, T59, T5a; T4Y = R1[WS(rs, 1)]; T86 = FNMS(KP668178637, T7V, T7W); T7X = FMA(KP668178637, T7W, T7V); TaM = TaC - Tav; TaD = Tav + TaC; T4Z = R1[WS(rs, 33)]; T59 = R1[WS(rs, 49)]; T5a = R1[WS(rs, 17)]; { E T5f, TaW, TaX, T5g, T5q, T5r; T5f = R1[WS(rs, 61)]; T50 = T4Y - T4Z; TaW = T4Y + T4Z; T5b = T59 - T5a; TaX = T5a + T59; T5g = R1[WS(rs, 29)]; T5q = R1[WS(rs, 45)]; T5r = R1[WS(rs, 13)]; TdA = TaW - TaX; TaY = TaW + TaX; T5h = T5f - T5g; Tb3 = T5f + T5g; Tb4 = T5r + T5q; T5s = T5q - T5r; } } { E T51, T52, T54, T55; T51 = R1[WS(rs, 9)]; Tb5 = Tb3 + Tb4; TdD = Tb3 - Tb4; T52 = R1[WS(rs, 41)]; T54 = R1[WS(rs, 57)]; T55 = R1[WS(rs, 25)]; { E T5i, TaZ, T53, Tb0, T56, T5j; T5i = R1[WS(rs, 5)]; TaZ = T51 + T52; T53 = T51 - T52; Tb0 = T54 + T55; T56 = T54 - T55; T5j = R1[WS(rs, 37)]; T5l = R1[WS(rs, 53)]; Tb1 = TaZ + Tb0; TdB = Tb0 - TaZ; T5c = T56 - T53; T57 = T53 + T56; Tb6 = T5i + T5j; T5k = T5i - T5j; T5m = R1[WS(rs, 21)]; } } { E TdL, TdC, T5o, T5t, TdM, TdF; Tb2 = TaY - Tb1; Tc6 = TaY + Tb1; { E T58, T5d, Tb7, T5n, TdE, Tb8; T58 = FMA(KP707106781, T57, T50); T8e = FNMS(KP707106781, T57, T50); T8f = FNMS(KP707106781, T5c, T5b); T5d = FMA(KP707106781, T5c, T5b); Tb7 = T5l + T5m; T5n = T5l - T5m; TdL = FNMS(KP414213562, TdA, TdB); TdC = FMA(KP414213562, TdB, TdA); T5e = FMA(KP198912367, T5d, T58); T5H = FNMS(KP198912367, T58, T5d); TdE = Tb7 - Tb6; Tb8 = Tb6 + Tb7; T5o = T5k + T5n; T5t = T5n - T5k; Tb9 = Tb5 - Tb8; Tc7 = Tb5 + Tb8; TdM = FMA(KP414213562, TdD, TdE); TdF = FNMS(KP414213562, TdE, TdD); } { E T8i, T8h, T5u, T5p; T8i = FNMS(KP707106781, T5t, T5s); T5u = FMA(KP707106781, T5t, T5s); T5p = FMA(KP707106781, T5o, T5h); T8h = FNMS(KP707106781, T5o, T5h); TeM = TdM - TdL; TdN = TdL + TdM; T5I = FMA(KP198912367, T5p, T5u); T5v = FNMS(KP198912367, T5u, T5p); T8q = FNMS(KP668178637, T8h, T8i); T8j = FMA(KP668178637, T8i, T8h); TeP = TdF - TdC; TdG = TdC + TdF; } } } } } } } { E T8p, T8g, TcH, TdW, TdT, TcO, Tfp, Tfk, Tfj, Tfq; { E Tbj, Tba, Tcy, Tco, TcB, Tcl, Tcx, Tcv, Tcz, Tcr; { E Tch, Tct, Tcp, Tcq, Tci, T1w, TbV, T11, Tcf, Tc9, T21, Tcj, Tcm, TbY, Tc1; E Tcn, Tcu, Tck; { E Tv, T10, Tc5, Tc8; Tch = Tf - Tu; Tv = Tf + Tu; T8p = FMA(KP668178637, T8e, T8f); T8g = FNMS(KP668178637, T8f, T8e); Tbj = Tb9 - Tb2; Tba = Tb2 + Tb9; T10 = TK + TZ; Tct = TZ - TK; Tcp = Tc3 - Tc4; Tc5 = Tc3 + Tc4; Tc8 = Tc6 + Tc7; Tcq = Tc7 - Tc6; Tci = T1g - T1v; T1w = T1g + T1v; TbV = Tv - T10; T11 = Tv + T10; Tcf = Tc5 + Tc8; Tc9 = Tc5 - Tc8; T21 = T1L + T20; Tcj = T1L - T20; Tcm = TbW - TbX; TbY = TbW + TbX; Tc1 = TbZ + Tc0; Tcn = Tc0 - TbZ; } { E Tcb, T22, Tce, Tc2; Tcb = T21 - T1w; T22 = T1w + T21; Tce = TbY + Tc1; Tc2 = TbY - Tc1; { E Tcd, Tcg, Tca, Tcc; Tcd = T11 + T22; Cr[WS(csr, 32)] = T11 - T22; Tcg = Tce + Tcf; Ci[WS(csi, 32)] = Tcf - Tce; Tca = Tc2 + Tc9; Tcc = Tc9 - Tc2; Cr[0] = Tcd + Tcg; Cr[WS(csr, 64)] = Tcd - Tcg; Ci[WS(csi, 48)] = FMS(KP707106781, Tcc, Tcb); Ci[WS(csi, 16)] = FMA(KP707106781, Tcc, Tcb); Cr[WS(csr, 16)] = FMA(KP707106781, Tca, TbV); Cr[WS(csr, 48)] = FNMS(KP707106781, Tca, TbV); Tcu = Tcj - Tci; Tck = Tci + Tcj; Tcy = FNMS(KP414213562, Tcm, Tcn); Tco = FMA(KP414213562, Tcn, Tcm); } } TcB = FNMS(KP707106781, Tck, Tch); Tcl = FMA(KP707106781, Tck, Tch); Tcx = FMA(KP707106781, Tcu, Tct); Tcv = FNMS(KP707106781, Tcu, Tct); Tcz = FMA(KP414213562, Tcp, Tcq); Tcr = FNMS(KP414213562, Tcq, Tcp); } { E TbT, TbO, TbN, TbU; { E Ta9, TbB, Tbb, TbL, Tbp, TbM, Tag, Tbk, TbR, TbJ, Tbw, TaO, TbC, Tbs, TbQ; E TbG; { E Tbq, Tbr, TbH, TbI; { E Tbo, Ta8, Tac, Taf; Tbo = Ta7 - Ta6; Ta8 = Ta6 + Ta7; { E TcC, TcA, Tcw, Tcs; TcC = Tcz - Tcy; TcA = Tcy + Tcz; Tcw = Tcr - Tco; Tcs = Tco + Tcr; Cr[WS(csr, 24)] = FMA(KP923879532, TcC, TcB); Cr[WS(csr, 40)] = FNMS(KP923879532, TcC, TcB); Ci[WS(csi, 56)] = FMS(KP923879532, TcA, Tcx); Ci[WS(csi, 8)] = FMA(KP923879532, TcA, Tcx); Ci[WS(csi, 40)] = FMA(KP923879532, Tcw, Tcv); Ci[WS(csi, 24)] = FMS(KP923879532, Tcw, Tcv); Cr[WS(csr, 8)] = FMA(KP923879532, Tcs, Tcl); Cr[WS(csr, 56)] = FNMS(KP923879532, Tcs, Tcl); Ta9 = FMA(KP707106781, Ta8, Ta5); TbB = FNMS(KP707106781, Ta8, Ta5); } Tbq = FNMS(KP414213562, Taa, Tab); Tac = FMA(KP414213562, Tab, Taa); Taf = FNMS(KP414213562, Tae, Tad); Tbr = FMA(KP414213562, Tad, Tae); Tbb = FMA(KP707106781, Tba, TaV); TbH = FNMS(KP707106781, Tba, TaV); TbL = FNMS(KP707106781, Tbo, Tbn); Tbp = FMA(KP707106781, Tbo, Tbn); TbM = Taf - Tac; Tag = Tac + Taf; TbI = FNMS(KP707106781, Tbj, Tbi); Tbk = FMA(KP707106781, Tbj, Tbi); } { E TbE, TbF, TaE, TaN; TbE = FNMS(KP707106781, TaD, Tao); TaE = FMA(KP707106781, TaD, Tao); TaN = FMA(KP707106781, TaM, TaL); TbF = FNMS(KP707106781, TaM, TaL); TbR = FNMS(KP668178637, TbH, TbI); TbJ = FMA(KP668178637, TbI, TbH); Tbw = FNMS(KP198912367, TaE, TaN); TaO = FMA(KP198912367, TaN, TaE); TbC = Tbr - Tbq; Tbs = Tbq + Tbr; TbQ = FMA(KP668178637, TbE, TbF); TbG = FNMS(KP668178637, TbF, TbE); } } { E Tbz, Tah, Tbv, Tbt, Tbx, Tbl; Tbz = FNMS(KP923879532, Tag, Ta9); Tah = FMA(KP923879532, Tag, Ta9); Tbv = FMA(KP923879532, Tbs, Tbp); Tbt = FNMS(KP923879532, Tbs, Tbp); Tbx = FMA(KP198912367, Tbb, Tbk); Tbl = FNMS(KP198912367, Tbk, Tbb); { E TbD, TbK, TbP, TbS; TbT = FNMS(KP923879532, TbC, TbB); TbD = FMA(KP923879532, TbC, TbB); { E TbA, Tby, Tbu, Tbm; TbA = Tbx - Tbw; Tby = Tbw + Tbx; Tbu = Tbl - TaO; Tbm = TaO + Tbl; Cr[WS(csr, 28)] = FMA(KP980785280, TbA, Tbz); Cr[WS(csr, 36)] = FNMS(KP980785280, TbA, Tbz); Ci[WS(csi, 60)] = FMS(KP980785280, Tby, Tbv); Ci[WS(csi, 4)] = FMA(KP980785280, Tby, Tbv); Ci[WS(csi, 36)] = FMA(KP980785280, Tbu, Tbt); Ci[WS(csi, 28)] = FMS(KP980785280, Tbu, Tbt); Cr[WS(csr, 4)] = FMA(KP980785280, Tbm, Tah); Cr[WS(csr, 60)] = FNMS(KP980785280, Tbm, Tah); TbK = TbG + TbJ; TbO = TbJ - TbG; } TbN = FMA(KP923879532, TbM, TbL); TbP = FNMS(KP923879532, TbM, TbL); TbS = TbQ + TbR; TbU = TbQ - TbR; Cr[WS(csr, 12)] = FMA(KP831469612, TbK, TbD); Cr[WS(csr, 52)] = FNMS(KP831469612, TbK, TbD); Ci[WS(csi, 52)] = FNMS(KP831469612, TbS, TbP); Ci[WS(csi, 12)] = -(FMA(KP831469612, TbS, TbP)); } } } { E TeN, Tf7, Tev, Tfm, Tfc, TeQ, TeX, TeW, Tfn, Tff, Tfi, TeC, Tf2, TeK, Tfh; E TeV, Tf8; { E TeG, TeJ, Tfd, Tfe, Tey, TeB, TeT, TeU; { E Tet, Teu, Tfa, Tfb; TcH = FMA(KP707106781, TcG, TcD); Tet = FNMS(KP707106781, TcG, TcD); Ci[WS(csi, 44)] = FMS(KP831469612, TbO, TbN); Ci[WS(csi, 20)] = FMA(KP831469612, TbO, TbN); Cr[WS(csr, 20)] = FMA(KP831469612, TbU, TbT); Cr[WS(csr, 44)] = FNMS(KP831469612, TbU, TbT); Teu = TdV - TdU; TdW = TdU + TdV; TeG = FNMS(KP923879532, TeF, TeE); Tfa = FMA(KP923879532, TeF, TeE); Tfb = FMA(KP923879532, TeI, TeH); TeJ = FNMS(KP923879532, TeI, TeH); TeN = FNMS(KP923879532, TeM, TeL); Tfd = FMA(KP923879532, TeM, TeL); Tf7 = FMA(KP923879532, Teu, Tet); Tev = FNMS(KP923879532, Teu, Tet); Tfm = FMA(KP303346683, Tfa, Tfb); Tfc = FNMS(KP303346683, Tfb, Tfa); Tfe = FNMS(KP923879532, TeP, TeO); TeQ = FMA(KP923879532, TeP, TeO); TeX = FNMS(KP668178637, Tew, Tex); Tey = FMA(KP668178637, Tex, Tew); TeB = FNMS(KP668178637, TeA, Tez); TeW = FMA(KP668178637, Tez, TeA); } Tfn = FNMS(KP303346683, Tfd, Tfe); Tff = FMA(KP303346683, Tfe, Tfd); Tfi = Tey + TeB; TeC = Tey - TeB; TdT = FMA(KP707106781, TdS, TdR); TeT = FNMS(KP707106781, TdS, TdR); TeU = TcN - TcK; TcO = TcK + TcN; Tf2 = FNMS(KP534511135, TeG, TeJ); TeK = FMA(KP534511135, TeJ, TeG); Tfh = FNMS(KP923879532, TeU, TeT); TeV = FMA(KP923879532, TeU, TeT); } { E Tf5, TeD, TeY, Tf3, TeR; Tf5 = FNMS(KP831469612, TeC, Tev); TeD = FMA(KP831469612, TeC, Tev); Tf8 = TeX + TeW; TeY = TeW - TeX; Tf3 = FMA(KP534511135, TeN, TeQ); TeR = FNMS(KP534511135, TeQ, TeN); { E Tf1, TeZ, Tf6, Tf4, Tf0, TeS; Tf1 = FMA(KP831469612, TeY, TeV); TeZ = FNMS(KP831469612, TeY, TeV); Tf6 = Tf3 - Tf2; Tf4 = Tf2 + Tf3; Tf0 = TeR - TeK; TeS = TeK + TeR; Ci[WS(csi, 54)] = FMS(KP881921264, Tf4, Tf1); Ci[WS(csi, 10)] = FMA(KP881921264, Tf4, Tf1); Ci[WS(csi, 42)] = FMA(KP881921264, Tf0, TeZ); Ci[WS(csi, 22)] = FMS(KP881921264, Tf0, TeZ); Cr[WS(csr, 10)] = FMA(KP881921264, TeS, TeD); Cr[WS(csr, 54)] = FNMS(KP881921264, TeS, TeD); Cr[WS(csr, 42)] = FNMS(KP881921264, Tf6, Tf5); Cr[WS(csr, 22)] = FMA(KP881921264, Tf6, Tf5); } } { E Tf9, Tfg, Tfl, Tfo; Tfp = FNMS(KP831469612, Tf8, Tf7); Tf9 = FMA(KP831469612, Tf8, Tf7); Tfg = Tfc + Tff; Tfk = Tff - Tfc; Tfj = FNMS(KP831469612, Tfi, Tfh); Tfl = FMA(KP831469612, Tfi, Tfh); Tfo = Tfm + Tfn; Tfq = Tfm - Tfn; Cr[WS(csr, 6)] = FMA(KP956940335, Tfg, Tf9); Cr[WS(csr, 58)] = FNMS(KP956940335, Tfg, Tf9); Ci[WS(csi, 58)] = FNMS(KP956940335, Tfo, Tfl); Ci[WS(csi, 6)] = -(FMA(KP956940335, Tfo, Tfl)); } } } } { E T2f, T5W, T5T, T2y, T5J, T5w, T4u, T4h, T7p, T7q; { E Ter, Tem, Tel, Tes; { E TdH, Te9, TcP, Teo, Tee, TdO, TdY, TdZ, Tep, Teh, Tek, Td8, Te4, Tdu, Tej; E TdX, Tea; { E Tdm, Tdt, Tef, Teg, TcY, Td7, Tec, Ted; Ci[WS(csi, 38)] = FMS(KP956940335, Tfk, Tfj); Ci[WS(csi, 26)] = FMA(KP956940335, Tfk, Tfj); Cr[WS(csr, 26)] = FMA(KP956940335, Tfq, Tfp); Cr[WS(csr, 38)] = FNMS(KP956940335, Tfq, Tfp); Tdm = FMA(KP923879532, Tdl, Tde); Tec = FNMS(KP923879532, Tdl, Tde); Ted = FNMS(KP923879532, Tds, Tdp); Tdt = FMA(KP923879532, Tds, Tdp); TdH = FMA(KP923879532, TdG, Tdz); Tef = FNMS(KP923879532, TdG, Tdz); Te9 = FNMS(KP923879532, TcO, TcH); TcP = FMA(KP923879532, TcO, TcH); Teo = FMA(KP820678790, Tec, Ted); Tee = FNMS(KP820678790, Ted, Tec); Teg = FNMS(KP923879532, TdN, TdK); TdO = FMA(KP923879532, TdN, TdK); TdY = FNMS(KP198912367, TcU, TcX); TcY = FMA(KP198912367, TcX, TcU); Td7 = FNMS(KP198912367, Td6, Td3); TdZ = FMA(KP198912367, Td3, Td6); Tep = FNMS(KP820678790, Tef, Teg); Teh = FMA(KP820678790, Teg, Tef); Tek = Td7 - TcY; Td8 = TcY + Td7; Te4 = FNMS(KP098491403, Tdm, Tdt); Tdu = FMA(KP098491403, Tdt, Tdm); Tej = FNMS(KP923879532, TdW, TdT); TdX = FMA(KP923879532, TdW, TdT); } { E Te7, Td9, Te0, Te5, TdP; Te7 = FNMS(KP980785280, Td8, TcP); Td9 = FMA(KP980785280, Td8, TcP); Tea = TdZ - TdY; Te0 = TdY + TdZ; Te5 = FMA(KP098491403, TdH, TdO); TdP = FNMS(KP098491403, TdO, TdH); { E Te3, Te1, Te8, Te6, Te2, TdQ; Te3 = FMA(KP980785280, Te0, TdX); Te1 = FNMS(KP980785280, Te0, TdX); Te8 = Te5 - Te4; Te6 = Te4 + Te5; Te2 = TdP - Tdu; TdQ = Tdu + TdP; Ci[WS(csi, 62)] = FMS(KP995184726, Te6, Te3); Ci[WS(csi, 2)] = FMA(KP995184726, Te6, Te3); Ci[WS(csi, 34)] = FMA(KP995184726, Te2, Te1); Ci[WS(csi, 30)] = FMS(KP995184726, Te2, Te1); Cr[WS(csr, 2)] = FMA(KP995184726, TdQ, Td9); Cr[WS(csr, 62)] = FNMS(KP995184726, TdQ, Td9); Cr[WS(csr, 34)] = FNMS(KP995184726, Te8, Te7); Cr[WS(csr, 30)] = FMA(KP995184726, Te8, Te7); } } { E Teb, Tei, Ten, Teq; Ter = FNMS(KP980785280, Tea, Te9); Teb = FMA(KP980785280, Tea, Te9); Tei = Tee + Teh; Tem = Teh - Tee; Tel = FMA(KP980785280, Tek, Tej); Ten = FNMS(KP980785280, Tek, Tej); Teq = Teo + Tep; Tes = Teo - Tep; Cr[WS(csr, 14)] = FMA(KP773010453, Tei, Teb); Cr[WS(csr, 50)] = FNMS(KP773010453, Tei, Teb); Ci[WS(csi, 50)] = FNMS(KP773010453, Teq, Ten); Ci[WS(csi, 14)] = -(FMA(KP773010453, Teq, Ten)); } } { E T77, T6v, T7i, T6C, T78, T6Y, T7h, T6V, T6N, T7d, T6P, T6F, T6I; { E T6W, T6X, T6T, T6U, T6M; { E T6t, T6u, T6y, T6B; T2f = FMA(KP923879532, T2e, T27); T6t = FNMS(KP923879532, T2e, T27); Ci[WS(csi, 46)] = FMS(KP773010453, Tem, Tel); Ci[WS(csi, 18)] = FMA(KP773010453, Tem, Tel); Cr[WS(csr, 18)] = FMA(KP773010453, Tes, Ter); Cr[WS(csr, 46)] = FNMS(KP773010453, Tes, Ter); T6u = T5U - T5V; T5W = T5U + T5V; T6W = FNMS(KP820678790, T6w, T6x); T6y = FMA(KP820678790, T6x, T6w); T6B = FNMS(KP820678790, T6A, T6z); T6X = FMA(KP820678790, T6z, T6A); T77 = FMA(KP980785280, T6u, T6t); T6v = FNMS(KP980785280, T6u, T6t); T7i = T6B + T6y; T6C = T6y - T6B; } T5T = FMA(KP923879532, T5S, T5P); T6T = FNMS(KP923879532, T5S, T5P); T6U = T2x - T2o; T2y = T2o + T2x; T5J = T5H + T5I; T6M = T5I - T5H; T78 = T6X + T6W; T6Y = T6W - T6X; T7h = FMA(KP980785280, T6U, T6T); T6V = FNMS(KP980785280, T6U, T6T); T6N = FNMS(KP980785280, T6M, T6L); T7d = FMA(KP980785280, T6M, T6L); T6P = T5v - T5e; T5w = T5e + T5v; T4u = T4s + T4t; T6F = T4s - T4t; T6I = T4g - T3Z; T4h = T3Z + T4g; } { E T75, T7f, T7n, T7c, T7m, T76; { E T6D, T72, T6R, T73, T6K, T71, T6Z, T7e, T6Q, T74, T70, T6S; T75 = FNMS(KP773010453, T6C, T6v); T6D = FMA(KP773010453, T6C, T6v); T7e = FNMS(KP980785280, T6P, T6O); T6Q = FMA(KP980785280, T6P, T6O); { E T7a, T6G, T7b, T6J; T7a = FMA(KP980785280, T6F, T6E); T6G = FNMS(KP980785280, T6F, T6E); T7b = FMA(KP980785280, T6I, T6H); T6J = FNMS(KP980785280, T6I, T6H); T7f = FMA(KP357805721, T7e, T7d); T7n = FNMS(KP357805721, T7d, T7e); T72 = FMA(KP472964775, T6N, T6Q); T6R = FNMS(KP472964775, T6Q, T6N); T7c = FMA(KP357805721, T7b, T7a); T7m = FNMS(KP357805721, T7a, T7b); T73 = FMA(KP472964775, T6G, T6J); T6K = FNMS(KP472964775, T6J, T6G); } T71 = FNMS(KP773010453, T6Y, T6V); T6Z = FMA(KP773010453, T6Y, T6V); T74 = T72 - T73; T76 = T73 + T72; T70 = T6R - T6K; T6S = T6K + T6R; Ci[WS(csi, 55)] = FMA(KP903989293, T74, T71); Ci[WS(csi, 9)] = FMS(KP903989293, T74, T71); Cr[WS(csr, 9)] = FMA(KP903989293, T6S, T6D); Cr[WS(csr, 55)] = FNMS(KP903989293, T6S, T6D); Ci[WS(csi, 41)] = FMS(KP903989293, T70, T6Z); Ci[WS(csi, 23)] = FMA(KP903989293, T70, T6Z); } { E T7k, T7j, T7l, T7o, T79, T7g; T7p = FNMS(KP773010453, T78, T77); T79 = FMA(KP773010453, T78, T77); T7g = T7c + T7f; T7k = T7f - T7c; T7j = FNMS(KP773010453, T7i, T7h); T7l = FMA(KP773010453, T7i, T7h); Cr[WS(csr, 23)] = FMA(KP903989293, T76, T75); Cr[WS(csr, 41)] = FNMS(KP903989293, T76, T75); Cr[WS(csr, 7)] = FMA(KP941544065, T7g, T79); Cr[WS(csr, 57)] = FNMS(KP941544065, T7g, T79); T7o = T7m - T7n; T7q = T7m + T7n; Ci[WS(csi, 57)] = FMS(KP941544065, T7o, T7l); Ci[WS(csi, 7)] = FMA(KP941544065, T7o, T7l); Ci[WS(csi, 39)] = FMA(KP941544065, T7k, T7j); Ci[WS(csi, 25)] = FMS(KP941544065, T7k, T7j); } } } } { E T7t, T8A, T8x, T7A, T8r, T8k, T88, T81, Ta3, Ta4, T6r, T6s; { E T9L, T99, T9W, T9g, T9M, T9C, T9V, T9z, T9k, T9O, T9T, Ta0, T9H, T9v, T9m; { E T9B, T9c, T9f, T9A, T97, T98; T7t = FMA(KP923879532, T7s, T7r); T97 = FNMS(KP923879532, T7s, T7r); T98 = T8z - T8y; T8A = T8y + T8z; T9B = FNMS(KP534511135, T9a, T9b); T9c = FMA(KP534511135, T9b, T9a); Cr[WS(csr, 25)] = FNMS(KP941544065, T7q, T7p); Cr[WS(csr, 39)] = FMA(KP941544065, T7q, T7p); T9L = FMA(KP831469612, T98, T97); T99 = FNMS(KP831469612, T98, T97); T9f = FNMS(KP534511135, T9e, T9d); T9A = FMA(KP534511135, T9d, T9e); { E T9x, T9y, T9q, T9t; T8x = FMA(KP923879532, T8w, T8v); T9x = FNMS(KP923879532, T8w, T8v); T9W = T9c + T9f; T9g = T9c - T9f; T9M = T9B + T9A; T9C = T9A - T9B; T9y = T7z - T7w; T7A = T7w + T7z; T8r = T8p + T8q; T9q = T8p - T8q; T9t = T8j - T8g; T8k = T8g + T8j; { E T9R, T9r, T9S, T9u, T9j; T88 = T86 + T87; T9j = T87 - T86; T9V = FNMS(KP831469612, T9y, T9x); T9z = FMA(KP831469612, T9y, T9x); T9R = FMA(KP831469612, T9q, T9p); T9r = FNMS(KP831469612, T9q, T9p); T9S = FMA(KP831469612, T9t, T9s); T9u = FNMS(KP831469612, T9t, T9s); T9k = FNMS(KP831469612, T9j, T9i); T9O = FMA(KP831469612, T9j, T9i); T9T = FNMS(KP250486960, T9S, T9R); Ta0 = FMA(KP250486960, T9R, T9S); T9H = FNMS(KP599376933, T9r, T9u); T9v = FMA(KP599376933, T9u, T9r); T9m = T7X - T80; T81 = T7X + T80; } } } { E T9J, T9h, T9F, T9D, T9P, T9n; T9J = FNMS(KP881921264, T9g, T99); T9h = FMA(KP881921264, T9g, T99); T9F = FMA(KP881921264, T9C, T9z); T9D = FNMS(KP881921264, T9C, T9z); T9P = FMA(KP831469612, T9m, T9l); T9n = FNMS(KP831469612, T9m, T9l); { E T9Y, T9X, T9Z, Ta2; { E T9N, Ta1, T9G, T9o, T9U, T9Q; Ta3 = FNMS(KP881921264, T9M, T9L); T9N = FMA(KP881921264, T9M, T9L); T9Q = FNMS(KP250486960, T9P, T9O); Ta1 = FMA(KP250486960, T9O, T9P); T9G = FNMS(KP599376933, T9k, T9n); T9o = FMA(KP599376933, T9n, T9k); T9U = T9Q + T9T; T9Y = T9T - T9Q; T9X = FNMS(KP881921264, T9W, T9V); T9Z = FMA(KP881921264, T9W, T9V); { E T9K, T9I, T9E, T9w; T9K = T9G + T9H; T9I = T9G - T9H; T9E = T9v - T9o; T9w = T9o + T9v; Cr[WS(csr, 5)] = FMA(KP970031253, T9U, T9N); Cr[WS(csr, 59)] = FNMS(KP970031253, T9U, T9N); Cr[WS(csr, 21)] = FNMS(KP857728610, T9K, T9J); Cr[WS(csr, 43)] = FMA(KP857728610, T9K, T9J); Ci[WS(csi, 53)] = FMS(KP857728610, T9I, T9F); Ci[WS(csi, 11)] = FMA(KP857728610, T9I, T9F); Ci[WS(csi, 43)] = FMA(KP857728610, T9E, T9D); Ci[WS(csi, 21)] = FMS(KP857728610, T9E, T9D); Cr[WS(csr, 11)] = FMA(KP857728610, T9w, T9h); Cr[WS(csr, 53)] = FNMS(KP857728610, T9w, T9h); Ta2 = Ta0 - Ta1; Ta4 = Ta1 + Ta0; } } Ci[WS(csi, 59)] = FMA(KP970031253, Ta2, T9Z); Ci[WS(csi, 5)] = FMS(KP970031253, Ta2, T9Z); Ci[WS(csi, 37)] = FMS(KP970031253, T9Y, T9X); Ci[WS(csi, 27)] = FMA(KP970031253, T9Y, T9X); } } } { E T69, T2z, T6k, T3g, T6a, T60, T6j, T5X, T4i, T6c, T6h, T6p, T64, T5L; { E T5Y, T2U, T3f, T5Z; T5Y = FMA(KP098491403, T2M, T2T); T2U = FNMS(KP098491403, T2T, T2M); Cr[WS(csr, 27)] = FMA(KP970031253, Ta4, Ta3); Cr[WS(csr, 37)] = FNMS(KP970031253, Ta4, Ta3); T69 = FNMS(KP980785280, T2y, T2f); T2z = FMA(KP980785280, T2y, T2f); T3f = FMA(KP098491403, T3e, T37); T5Z = FNMS(KP098491403, T37, T3e); T6k = T3f - T2U; T3g = T2U + T3f; T6a = T5Y - T5Z; T60 = T5Y + T5Z; { E T6f, T5x, T6g, T5K; T6j = FNMS(KP980785280, T5W, T5T); T5X = FMA(KP980785280, T5W, T5T); T6f = FNMS(KP980785280, T5w, T4X); T5x = FMA(KP980785280, T5w, T4X); T6g = FNMS(KP980785280, T5J, T5G); T5K = FMA(KP980785280, T5J, T5G); T4i = FMA(KP980785280, T4h, T3I); T6c = FNMS(KP980785280, T4h, T3I); T6h = FMA(KP906347169, T6g, T6f); T6p = FNMS(KP906347169, T6f, T6g); T64 = FMA(KP049126849, T5x, T5K); T5L = FNMS(KP049126849, T5K, T5x); } } { E T67, T3h, T63, T61, T6d, T4v; T67 = FNMS(KP995184726, T3g, T2z); T3h = FMA(KP995184726, T3g, T2z); T63 = FMA(KP995184726, T60, T5X); T61 = FNMS(KP995184726, T60, T5X); T6d = FNMS(KP980785280, T4u, T4r); T4v = FMA(KP980785280, T4u, T4r); { E T6m, T6l, T6n, T6q; { E T6b, T6o, T65, T4w, T6i, T6e; T6r = FNMS(KP995184726, T6a, T69); T6b = FMA(KP995184726, T6a, T69); T6e = FMA(KP906347169, T6d, T6c); T6o = FNMS(KP906347169, T6c, T6d); T65 = FMA(KP049126849, T4i, T4v); T4w = FNMS(KP049126849, T4v, T4i); T6i = T6e + T6h; T6m = T6h - T6e; T6l = FNMS(KP995184726, T6k, T6j); T6n = FMA(KP995184726, T6k, T6j); { E T68, T66, T62, T5M; T68 = T65 + T64; T66 = T64 - T65; T62 = T5L - T4w; T5M = T4w + T5L; Cr[WS(csr, 15)] = FMA(KP740951125, T6i, T6b); Cr[WS(csr, 49)] = FNMS(KP740951125, T6i, T6b); Cr[WS(csr, 31)] = FMA(KP998795456, T68, T67); Cr[WS(csr, 33)] = FNMS(KP998795456, T68, T67); Ci[WS(csi, 63)] = FMA(KP998795456, T66, T63); Ci[WS(csi, 1)] = FMS(KP998795456, T66, T63); Ci[WS(csi, 33)] = FMS(KP998795456, T62, T61); Ci[WS(csi, 31)] = FMA(KP998795456, T62, T61); Cr[WS(csr, 1)] = FMA(KP998795456, T5M, T3h); Cr[WS(csr, 63)] = FNMS(KP998795456, T5M, T3h); T6q = T6o - T6p; T6s = T6o + T6p; } } Ci[WS(csi, 49)] = FMS(KP740951125, T6q, T6n); Ci[WS(csi, 15)] = FMA(KP740951125, T6q, T6n); Ci[WS(csi, 47)] = FMA(KP740951125, T6m, T6l); Ci[WS(csi, 17)] = FMS(KP740951125, T6m, T6l); } } } { E T8N, T7B, T8Y, T7Q, T8O, T8E, T8X, T8B, T82, T8Q, T8V, T92, T8J, T8t; { E T8C, T7I, T7P, T8D; T8C = FNMS(KP303346683, T7E, T7H); T7I = FMA(KP303346683, T7H, T7E); Cr[WS(csr, 17)] = FNMS(KP740951125, T6s, T6r); Cr[WS(csr, 47)] = FMA(KP740951125, T6s, T6r); T8N = FNMS(KP831469612, T7A, T7t); T7B = FMA(KP831469612, T7A, T7t); T7P = FNMS(KP303346683, T7O, T7L); T8D = FMA(KP303346683, T7L, T7O); T8Y = T7P - T7I; T7Q = T7I + T7P; T8O = T8D - T8C; T8E = T8C + T8D; { E T8T, T8l, T8U, T8s; T8X = FNMS(KP831469612, T8A, T8x); T8B = FMA(KP831469612, T8A, T8x); T8T = FNMS(KP831469612, T8k, T8d); T8l = FMA(KP831469612, T8k, T8d); T8U = FNMS(KP831469612, T8r, T8o); T8s = FMA(KP831469612, T8r, T8o); T82 = FMA(KP831469612, T81, T7U); T8Q = FNMS(KP831469612, T81, T7U); T8V = FNMS(KP741650546, T8U, T8T); T92 = FMA(KP741650546, T8T, T8U); T8J = FNMS(KP148335987, T8l, T8s); T8t = FMA(KP148335987, T8s, T8l); } } { E T8L, T7R, T8H, T8F, T8R, T89; T8L = FNMS(KP956940335, T7Q, T7B); T7R = FMA(KP956940335, T7Q, T7B); T8H = FMA(KP956940335, T8E, T8B); T8F = FNMS(KP956940335, T8E, T8B); T8R = FNMS(KP831469612, T88, T85); T89 = FMA(KP831469612, T88, T85); { E T90, T8Z, T91, T94; { E T8P, T93, T8I, T8a, T8W, T8S; T95 = FNMS(KP956940335, T8O, T8N); T8P = FMA(KP956940335, T8O, T8N); T8S = FNMS(KP741650546, T8R, T8Q); T93 = FMA(KP741650546, T8Q, T8R); T8I = FNMS(KP148335987, T82, T89); T8a = FMA(KP148335987, T89, T82); T8W = T8S + T8V; T90 = T8V - T8S; T8Z = FMA(KP956940335, T8Y, T8X); T91 = FNMS(KP956940335, T8Y, T8X); { E T8M, T8K, T8G, T8u; T8M = T8I + T8J; T8K = T8I - T8J; T8G = T8t - T8a; T8u = T8a + T8t; Cr[WS(csr, 13)] = FMA(KP803207531, T8W, T8P); Cr[WS(csr, 51)] = FNMS(KP803207531, T8W, T8P); Cr[WS(csr, 29)] = FNMS(KP989176509, T8M, T8L); Cr[WS(csr, 35)] = FMA(KP989176509, T8M, T8L); Ci[WS(csi, 61)] = FMS(KP989176509, T8K, T8H); Ci[WS(csi, 3)] = FMA(KP989176509, T8K, T8H); Ci[WS(csi, 35)] = FMA(KP989176509, T8G, T8F); Ci[WS(csi, 29)] = FMS(KP989176509, T8G, T8F); Cr[WS(csr, 3)] = FMA(KP989176509, T8u, T7R); Cr[WS(csr, 61)] = FNMS(KP989176509, T8u, T7R); T94 = T92 - T93; T96 = T93 + T92; } } Ci[WS(csi, 51)] = FMA(KP803207531, T94, T91); Ci[WS(csi, 13)] = FMS(KP803207531, T94, T91); Ci[WS(csi, 45)] = FMS(KP803207531, T90, T8Z); Ci[WS(csi, 19)] = FMA(KP803207531, T90, T8Z); } } } } } } } Cr[WS(csr, 19)] = FMA(KP803207531, T96, T95); Cr[WS(csr, 45)] = FNMS(KP803207531, T96, T95); } } } static const kr2c_desc desc = { 128, "r2cf_128", {440, 0, 516, 0}, &GENUS }; void X(codelet_r2cf_128) (planner *p) { X(kr2c_register) (p, r2cf_128, &desc); } #else /* HAVE_FMA */ /* Generated by: ../../../genfft/gen_r2cf.native -compact -variables 4 -pipeline-latency 4 -n 128 -name r2cf_128 -include r2cf.h */ /* * This function contains 956 FP additions, 330 FP multiplications, * (or, 812 additions, 186 multiplications, 144 fused multiply/add), * 186 stack variables, 31 constants, and 256 memory accesses */ #include "r2cf.h" static void r2cf_128(R *R0, R *R1, R *Cr, R *Ci, stride rs, stride csr, stride csi, INT v, INT ivs, INT ovs) { DK(KP803207531, +0.803207531480644909806676512963141923879569427); DK(KP595699304, +0.595699304492433343467036528829969889511926338); DK(KP146730474, +0.146730474455361751658850129646717819706215317); DK(KP989176509, +0.989176509964780973451673738016243063983689533); DK(KP740951125, +0.740951125354959091175616897495162729728955309); DK(KP671558954, +0.671558954847018400625376850427421803228750632); DK(KP049067674, +0.049067674327418014254954976942682658314745363); DK(KP998795456, +0.998795456205172392714771604759100694443203615); DK(KP242980179, +0.242980179903263889948274162077471118320990783); DK(KP970031253, +0.970031253194543992603984207286100251456865962); DK(KP514102744, +0.514102744193221726593693838968815772608049120); DK(KP857728610, +0.857728610000272069902269984284770137042490799); DK(KP336889853, +0.336889853392220050689253212619147570477766780); DK(KP941544065, +0.941544065183020778412509402599502357185589796); DK(KP427555093, +0.427555093430282094320966856888798534304578629); DK(KP903989293, +0.903989293123443331586200297230537048710132025); DK(KP098017140, +0.098017140329560601994195563888641845861136673); DK(KP995184726, +0.995184726672196886244836953109479921575474869); DK(KP634393284, +0.634393284163645498215171613225493370675687095); DK(KP773010453, +0.773010453362736960810906609758469800971041293); DK(KP881921264, +0.881921264348355029712756863660388349508442621); DK(KP471396736, +0.471396736825997648556387625905254377657460319); DK(KP956940335, +0.956940335732208864935797886980269969482849206); DK(KP290284677, +0.290284677254462367636192375817395274691476278); DK(KP555570233, +0.555570233019602224742830813948532874374937191); DK(KP831469612, +0.831469612302545237078788377617905756738560812); DK(KP195090322, +0.195090322016128267848284868477022240927691618); DK(KP980785280, +0.980785280403230449126182236134239036973933731); DK(KP382683432, +0.382683432365089771728459984030398866761344562); DK(KP923879532, +0.923879532511286756128183189396788286822416626); DK(KP707106781, +0.707106781186547524400844362104849039284835938); { INT i; for (i = v; i > 0; i = i - 1, R0 = R0 + ivs, R1 = R1 + ivs, Cr = Cr + ovs, Ci = Ci + ovs, MAKE_VOLATILE_STRIDE(512, rs), MAKE_VOLATILE_STRIDE(512, csr), MAKE_VOLATILE_STRIDE(512, csi)) { E TcD, TdU, T27, T7r, T5S, T8y, Tf, Ta5, Tu, Tbq, TcG, TdV, T2e, T8z, T5V; E T7s, TK, Ta6, TcK, TdX, T2o, T5X, T7w, T8B, TZ, Ta7, TcN, TdY, T2x, T5Y; E T7z, T8C, T1g, Taa, TcU, TeA, TcX, Tez, T1v, Tab, T2M, T6z, T7E, T9e, T7H; E T9d, T2T, T6A, T4X, T6L, Tdz, TeL, TdK, TeP, T5G, T6P, T8d, T9p, TaV, Tc3; E Tbi, Tc4, T8o, T9t, T3I, T6H, Tde, TeH, Tdp, TeF, T4r, T6F, T7U, T9l, Tao; E TbW, TaL, TbX, T85, T9j, T1L, Tad, Td3, Tew, Td6, Tex, T20, Tae, T37, T6x; E T7L, T9a, T7O, T9b, T3e, T6w, TbZ, Tc0, T3Z, T4s, Tds, TeI, T4g, T4t, T80; E T87, Tdl, TeE, T7X, T86, TaD, TaM, Tc6, Tc7, T5e, T5H, TdN, TeM, T5v, T5I; E T8j, T8q, TdG, TeO, T8g, T8p, Tba, Tbj; { E T3, T23, Td, T25, T6, T5R, Ta, T24; { E T1, T2, Tb, Tc; T1 = R0[0]; T2 = R0[WS(rs, 32)]; T3 = T1 + T2; T23 = T1 - T2; Tb = R0[WS(rs, 56)]; Tc = R0[WS(rs, 24)]; Td = Tb + Tc; T25 = Tb - Tc; } { E T4, T5, T8, T9; T4 = R0[WS(rs, 16)]; T5 = R0[WS(rs, 48)]; T6 = T4 + T5; T5R = T4 - T5; T8 = R0[WS(rs, 8)]; T9 = R0[WS(rs, 40)]; Ta = T8 + T9; T24 = T8 - T9; } TcD = T3 - T6; TdU = Td - Ta; { E T26, T5Q, T7, Te; T26 = KP707106781 * (T24 + T25); T27 = T23 + T26; T7r = T23 - T26; T5Q = KP707106781 * (T25 - T24); T5S = T5Q - T5R; T8y = T5R + T5Q; T7 = T3 + T6; Te = Ta + Td; Tf = T7 + Te; Ta5 = T7 - Te; } } { E Ti, T28, Ts, T2c, Tl, T29, Tp, T2b; { E Tg, Th, Tq, Tr; Tg = R0[WS(rs, 4)]; Th = R0[WS(rs, 36)]; Ti = Tg + Th; T28 = Tg - Th; Tq = R0[WS(rs, 12)]; Tr = R0[WS(rs, 44)]; Ts = Tq + Tr; T2c = Tq - Tr; } { E Tj, Tk, Tn, To; Tj = R0[WS(rs, 20)]; Tk = R0[WS(rs, 52)]; Tl = Tj + Tk; T29 = Tj - Tk; Tn = R0[WS(rs, 60)]; To = R0[WS(rs, 28)]; Tp = Tn + To; T2b = Tn - To; } { E Tm, Tt, TcE, TcF; Tm = Ti + Tl; Tt = Tp + Ts; Tu = Tm + Tt; Tbq = Tt - Tm; TcE = Ti - Tl; TcF = Tp - Ts; TcG = KP707106781 * (TcE + TcF); TdV = KP707106781 * (TcF - TcE); } { E T2a, T2d, T5T, T5U; T2a = FNMS(KP382683432, T29, KP923879532 * T28); T2d = FMA(KP923879532, T2b, KP382683432 * T2c); T2e = T2a + T2d; T8z = T2d - T2a; T5T = FNMS(KP923879532, T2c, KP382683432 * T2b); T5U = FMA(KP382683432, T28, KP923879532 * T29); T5V = T5T - T5U; T7s = T5U + T5T; } } { E Ty, T2g, TB, T2m, TF, T2l, TI, T2j; { E Tw, Tx, Tz, TA; Tw = R0[WS(rs, 2)]; Tx = R0[WS(rs, 34)]; Ty = Tw + Tx; T2g = Tw - Tx; Tz = R0[WS(rs, 18)]; TA = R0[WS(rs, 50)]; TB = Tz + TA; T2m = Tz - TA; { E TD, TE, T2h, TG, TH, T2i; TD = R0[WS(rs, 10)]; TE = R0[WS(rs, 42)]; T2h = TD - TE; TG = R0[WS(rs, 58)]; TH = R0[WS(rs, 26)]; T2i = TG - TH; TF = TD + TE; T2l = KP707106781 * (T2i - T2h); TI = TG + TH; T2j = KP707106781 * (T2h + T2i); } } { E TC, TJ, TcI, TcJ; TC = Ty + TB; TJ = TF + TI; TK = TC + TJ; Ta6 = TC - TJ; TcI = Ty - TB; TcJ = TI - TF; TcK = FMA(KP923879532, TcI, KP382683432 * TcJ); TdX = FNMS(KP382683432, TcI, KP923879532 * TcJ); } { E T2k, T2n, T7u, T7v; T2k = T2g + T2j; T2n = T2l - T2m; T2o = FMA(KP980785280, T2k, KP195090322 * T2n); T5X = FNMS(KP195090322, T2k, KP980785280 * T2n); T7u = T2g - T2j; T7v = T2m + T2l; T7w = FMA(KP831469612, T7u, KP555570233 * T7v); T8B = FNMS(KP555570233, T7u, KP831469612 * T7v); } } { E TN, T2p, TQ, T2v, TU, T2u, TX, T2s; { E TL, TM, TO, TP; TL = R0[WS(rs, 62)]; TM = R0[WS(rs, 30)]; TN = TL + TM; T2p = TL - TM; TO = R0[WS(rs, 14)]; TP = R0[WS(rs, 46)]; TQ = TO + TP; T2v = TO - TP; { E TS, TT, T2q, TV, TW, T2r; TS = R0[WS(rs, 6)]; TT = R0[WS(rs, 38)]; T2q = TS - TT; TV = R0[WS(rs, 54)]; TW = R0[WS(rs, 22)]; T2r = TV - TW; TU = TS + TT; T2u = KP707106781 * (T2r - T2q); TX = TV + TW; T2s = KP707106781 * (T2q + T2r); } } { E TR, TY, TcL, TcM; TR = TN + TQ; TY = TU + TX; TZ = TR + TY; Ta7 = TR - TY; TcL = TN - TQ; TcM = TX - TU; TcN = FNMS(KP382683432, TcM, KP923879532 * TcL); TdY = FMA(KP382683432, TcL, KP923879532 * TcM); } { E T2t, T2w, T7x, T7y; T2t = T2p + T2s; T2w = T2u - T2v; T2x = FNMS(KP195090322, T2w, KP980785280 * T2t); T5Y = FMA(KP195090322, T2t, KP980785280 * T2w); T7x = T2p - T2s; T7y = T2v + T2u; T7z = FNMS(KP555570233, T7y, KP831469612 * T7x); T8C = FMA(KP555570233, T7x, KP831469612 * T7y); } } { E T14, T2N, T17, T2D, T1b, T2O, T1e, T2C, T1j, T1m, T2K, TcR, T2Q, T1q, T1t; E T2H, TcS, T2R; { E T12, T13, T15, T16; T12 = R0[WS(rs, 1)]; T13 = R0[WS(rs, 33)]; T14 = T12 + T13; T2N = T12 - T13; T15 = R0[WS(rs, 17)]; T16 = R0[WS(rs, 49)]; T17 = T15 + T16; T2D = T15 - T16; } { E T19, T1a, T2B, T1c, T1d, T2A; T19 = R0[WS(rs, 9)]; T1a = R0[WS(rs, 41)]; T2B = T19 - T1a; T1c = R0[WS(rs, 57)]; T1d = R0[WS(rs, 25)]; T2A = T1c - T1d; T1b = T19 + T1a; T2O = KP707106781 * (T2B + T2A); T1e = T1c + T1d; T2C = KP707106781 * (T2A - T2B); } { E T2I, T2J, T2F, T2G; { E T1h, T1i, T1k, T1l; T1h = R0[WS(rs, 5)]; T1i = R0[WS(rs, 37)]; T1j = T1h + T1i; T2I = T1h - T1i; T1k = R0[WS(rs, 21)]; T1l = R0[WS(rs, 53)]; T1m = T1k + T1l; T2J = T1k - T1l; } T2K = FMA(KP382683432, T2I, KP923879532 * T2J); TcR = T1j - T1m; T2Q = FNMS(KP382683432, T2J, KP923879532 * T2I); { E T1o, T1p, T1r, T1s; T1o = R0[WS(rs, 61)]; T1p = R0[WS(rs, 29)]; T1q = T1o + T1p; T2F = T1o - T1p; T1r = R0[WS(rs, 13)]; T1s = R0[WS(rs, 45)]; T1t = T1r + T1s; T2G = T1r - T1s; } T2H = FNMS(KP923879532, T2G, KP382683432 * T2F); TcS = T1q - T1t; T2R = FMA(KP923879532, T2F, KP382683432 * T2G); } { E T18, T1f, TcQ, TcT; T18 = T14 + T17; T1f = T1b + T1e; T1g = T18 + T1f; Taa = T18 - T1f; TcQ = T14 - T17; TcT = KP707106781 * (TcR + TcS); TcU = TcQ + TcT; TeA = TcQ - TcT; } { E TcV, TcW, T1n, T1u; TcV = T1e - T1b; TcW = KP707106781 * (TcS - TcR); TcX = TcV + TcW; Tez = TcW - TcV; T1n = T1j + T1m; T1u = T1q + T1t; T1v = T1n + T1u; Tab = T1u - T1n; } { E T2E, T2L, T7C, T7D; T2E = T2C - T2D; T2L = T2H - T2K; T2M = T2E + T2L; T6z = T2L - T2E; T7C = T2N - T2O; T7D = T2K + T2H; T7E = T7C + T7D; T9e = T7C - T7D; } { E T7F, T7G, T2P, T2S; T7F = T2D + T2C; T7G = T2R - T2Q; T7H = T7F + T7G; T9d = T7G - T7F; T2P = T2N + T2O; T2S = T2Q + T2R; T2T = T2P + T2S; T6A = T2P - T2S; } } { E T4z, TaP, T5B, TaQ, T4G, TaT, T5y, TaS, Tbf, Tbg, T4O, Tdw, T5E, Tbc, Tbd; E T4V, Tdx, T5D; { E T4x, T4y, T5z, T5A; T4x = R1[WS(rs, 63)]; T4y = R1[WS(rs, 31)]; T4z = T4x - T4y; TaP = T4x + T4y; T5z = R1[WS(rs, 15)]; T5A = R1[WS(rs, 47)]; T5B = T5z - T5A; TaQ = T5z + T5A; } { E T4A, T4B, T4C, T4D, T4E, T4F; T4A = R1[WS(rs, 7)]; T4B = R1[WS(rs, 39)]; T4C = T4A - T4B; T4D = R1[WS(rs, 55)]; T4E = R1[WS(rs, 23)]; T4F = T4D - T4E; T4G = KP707106781 * (T4C + T4F); TaT = T4D + T4E; T5y = KP707106781 * (T4F - T4C); TaS = T4A + T4B; } { E T4K, T4N, T4R, T4U; { E T4I, T4J, T4L, T4M; T4I = R1[WS(rs, 3)]; T4J = R1[WS(rs, 35)]; T4K = T4I - T4J; Tbf = T4I + T4J; T4L = R1[WS(rs, 19)]; T4M = R1[WS(rs, 51)]; T4N = T4L - T4M; Tbg = T4L + T4M; } T4O = FNMS(KP382683432, T4N, KP923879532 * T4K); Tdw = Tbf - Tbg; T5E = FMA(KP382683432, T4K, KP923879532 * T4N); { E T4P, T4Q, T4S, T4T; T4P = R1[WS(rs, 59)]; T4Q = R1[WS(rs, 27)]; T4R = T4P - T4Q; Tbc = T4P + T4Q; T4S = R1[WS(rs, 11)]; T4T = R1[WS(rs, 43)]; T4U = T4S - T4T; Tbd = T4S + T4T; } T4V = FMA(KP923879532, T4R, KP382683432 * T4U); Tdx = Tbc - Tbd; T5D = FNMS(KP923879532, T4U, KP382683432 * T4R); } { E T4H, T4W, Tdv, Tdy; T4H = T4z + T4G; T4W = T4O + T4V; T4X = T4H + T4W; T6L = T4H - T4W; Tdv = TaP - TaQ; Tdy = KP707106781 * (Tdw + Tdx); Tdz = Tdv + Tdy; TeL = Tdv - Tdy; } { E TdI, TdJ, T5C, T5F; TdI = TaT - TaS; TdJ = KP707106781 * (Tdx - Tdw); TdK = TdI + TdJ; TeP = TdJ - TdI; T5C = T5y - T5B; T5F = T5D - T5E; T5G = T5C + T5F; T6P = T5F - T5C; } { E T8b, T8c, TaR, TaU; T8b = T4z - T4G; T8c = T5E + T5D; T8d = T8b + T8c; T9p = T8b - T8c; TaR = TaP + TaQ; TaU = TaS + TaT; TaV = TaR - TaU; Tc3 = TaR + TaU; } { E Tbe, Tbh, T8m, T8n; Tbe = Tbc + Tbd; Tbh = Tbf + Tbg; Tbi = Tbe - Tbh; Tc4 = Tbh + Tbe; T8m = T5B + T5y; T8n = T4V - T4O; T8o = T8m + T8n; T9t = T8n - T8m; } } { E T3k, Tai, T4m, Taj, T3r, Tam, T4j, Tal, TaI, TaJ, T3z, Tdb, T4p, TaF, TaG; E T3G, Tdc, T4o; { E T3i, T3j, T4k, T4l; T3i = R1[0]; T3j = R1[WS(rs, 32)]; T3k = T3i - T3j; Tai = T3i + T3j; T4k = R1[WS(rs, 16)]; T4l = R1[WS(rs, 48)]; T4m = T4k - T4l; Taj = T4k + T4l; } { E T3l, T3m, T3n, T3o, T3p, T3q; T3l = R1[WS(rs, 8)]; T3m = R1[WS(rs, 40)]; T3n = T3l - T3m; T3o = R1[WS(rs, 56)]; T3p = R1[WS(rs, 24)]; T3q = T3o - T3p; T3r = KP707106781 * (T3n + T3q); Tam = T3o + T3p; T4j = KP707106781 * (T3q - T3n); Tal = T3l + T3m; } { E T3v, T3y, T3C, T3F; { E T3t, T3u, T3w, T3x; T3t = R1[WS(rs, 4)]; T3u = R1[WS(rs, 36)]; T3v = T3t - T3u; TaI = T3t + T3u; T3w = R1[WS(rs, 20)]; T3x = R1[WS(rs, 52)]; T3y = T3w - T3x; TaJ = T3w + T3x; } T3z = FNMS(KP382683432, T3y, KP923879532 * T3v); Tdb = TaI - TaJ; T4p = FMA(KP382683432, T3v, KP923879532 * T3y); { E T3A, T3B, T3D, T3E; T3A = R1[WS(rs, 60)]; T3B = R1[WS(rs, 28)]; T3C = T3A - T3B; TaF = T3A + T3B; T3D = R1[WS(rs, 12)]; T3E = R1[WS(rs, 44)]; T3F = T3D - T3E; TaG = T3D + T3E; } T3G = FMA(KP923879532, T3C, KP382683432 * T3F); Tdc = TaF - TaG; T4o = FNMS(KP923879532, T3F, KP382683432 * T3C); } { E T3s, T3H, Tda, Tdd; T3s = T3k + T3r; T3H = T3z + T3G; T3I = T3s + T3H; T6H = T3s - T3H; Tda = Tai - Taj; Tdd = KP707106781 * (Tdb + Tdc); Tde = Tda + Tdd; TeH = Tda - Tdd; } { E Tdn, Tdo, T4n, T4q; Tdn = Tam - Tal; Tdo = KP707106781 * (Tdc - Tdb); Tdp = Tdn + Tdo; TeF = Tdo - Tdn; T4n = T4j - T4m; T4q = T4o - T4p; T4r = T4n + T4q; T6F = T4q - T4n; } { E T7S, T7T, Tak, Tan; T7S = T3k - T3r; T7T = T4p + T4o; T7U = T7S + T7T; T9l = T7S - T7T; Tak = Tai + Taj; Tan = Tal + Tam; Tao = Tak - Tan; TbW = Tak + Tan; } { E TaH, TaK, T83, T84; TaH = TaF + TaG; TaK = TaI + TaJ; TaL = TaH - TaK; TbX = TaK + TaH; T83 = T4m + T4j; T84 = T3G - T3z; T85 = T83 + T84; T9j = T84 - T83; } } { E T1z, T2V, T1C, T39, T1G, T38, T1J, T2Y, T1O, T1R, T32, Td0, T3c, T1V, T1Y; E T35, Td1, T3b; { E T1x, T1y, T1A, T1B; T1x = R0[WS(rs, 63)]; T1y = R0[WS(rs, 31)]; T1z = T1x + T1y; T2V = T1x - T1y; T1A = R0[WS(rs, 15)]; T1B = R0[WS(rs, 47)]; T1C = T1A + T1B; T39 = T1A - T1B; } { E T1E, T1F, T2W, T1H, T1I, T2X; T1E = R0[WS(rs, 7)]; T1F = R0[WS(rs, 39)]; T2W = T1E - T1F; T1H = R0[WS(rs, 55)]; T1I = R0[WS(rs, 23)]; T2X = T1H - T1I; T1G = T1E + T1F; T38 = KP707106781 * (T2X - T2W); T1J = T1H + T1I; T2Y = KP707106781 * (T2W + T2X); } { E T30, T31, T33, T34; { E T1M, T1N, T1P, T1Q; T1M = R0[WS(rs, 3)]; T1N = R0[WS(rs, 35)]; T1O = T1M + T1N; T30 = T1M - T1N; T1P = R0[WS(rs, 19)]; T1Q = R0[WS(rs, 51)]; T1R = T1P + T1Q; T31 = T1P - T1Q; } T32 = FNMS(KP382683432, T31, KP923879532 * T30); Td0 = T1O - T1R; T3c = FMA(KP382683432, T30, KP923879532 * T31); { E T1T, T1U, T1W, T1X; T1T = R0[WS(rs, 59)]; T1U = R0[WS(rs, 27)]; T1V = T1T + T1U; T33 = T1T - T1U; T1W = R0[WS(rs, 11)]; T1X = R0[WS(rs, 43)]; T1Y = T1W + T1X; T34 = T1W - T1X; } T35 = FMA(KP923879532, T33, KP382683432 * T34); Td1 = T1V - T1Y; T3b = FNMS(KP923879532, T34, KP382683432 * T33); } { E T1D, T1K, TcZ, Td2; T1D = T1z + T1C; T1K = T1G + T1J; T1L = T1D + T1K; Tad = T1D - T1K; TcZ = T1z - T1C; Td2 = KP707106781 * (Td0 + Td1); Td3 = TcZ + Td2; Tew = TcZ - Td2; } { E Td4, Td5, T1S, T1Z; Td4 = T1J - T1G; Td5 = KP707106781 * (Td1 - Td0); Td6 = Td4 + Td5; Tex = Td5 - Td4; T1S = T1O + T1R; T1Z = T1V + T1Y; T20 = T1S + T1Z; Tae = T1Z - T1S; } { E T2Z, T36, T7J, T7K; T2Z = T2V + T2Y; T36 = T32 + T35; T37 = T2Z + T36; T6x = T2Z - T36; T7J = T2V - T2Y; T7K = T3c + T3b; T7L = T7J + T7K; T9a = T7J - T7K; } { E T7M, T7N, T3a, T3d; T7M = T39 + T38; T7N = T35 - T32; T7O = T7M + T7N; T9b = T7N - T7M; T3a = T38 - T39; T3d = T3b - T3c; T3e = T3a + T3d; T6w = T3d - T3a; } } { E T3L, Tdf, T3X, Tar, T42, Tdi, T4e, Tay, T3S, Tdg, T3U, Tau, T49, Tdj, T4b; E TaB, Tdh, Tdk; { E T3J, T3K, Tap, T3V, T3W, Taq; T3J = R1[WS(rs, 2)]; T3K = R1[WS(rs, 34)]; Tap = T3J + T3K; T3V = R1[WS(rs, 18)]; T3W = R1[WS(rs, 50)]; Taq = T3V + T3W; T3L = T3J - T3K; Tdf = Tap - Taq; T3X = T3V - T3W; Tar = Tap + Taq; } { E T40, T41, Taw, T4c, T4d, Tax; T40 = R1[WS(rs, 62)]; T41 = R1[WS(rs, 30)]; Taw = T40 + T41; T4c = R1[WS(rs, 14)]; T4d = R1[WS(rs, 46)]; Tax = T4c + T4d; T42 = T40 - T41; Tdi = Taw - Tax; T4e = T4c - T4d; Tay = Taw + Tax; } { E T3O, Tas, T3R, Tat; { E T3M, T3N, T3P, T3Q; T3M = R1[WS(rs, 10)]; T3N = R1[WS(rs, 42)]; T3O = T3M - T3N; Tas = T3M + T3N; T3P = R1[WS(rs, 58)]; T3Q = R1[WS(rs, 26)]; T3R = T3P - T3Q; Tat = T3P + T3Q; } T3S = KP707106781 * (T3O + T3R); Tdg = Tat - Tas; T3U = KP707106781 * (T3R - T3O); Tau = Tas + Tat; } { E T45, Taz, T48, TaA; { E T43, T44, T46, T47; T43 = R1[WS(rs, 6)]; T44 = R1[WS(rs, 38)]; T45 = T43 - T44; Taz = T43 + T44; T46 = R1[WS(rs, 54)]; T47 = R1[WS(rs, 22)]; T48 = T46 - T47; TaA = T46 + T47; } T49 = KP707106781 * (T45 + T48); Tdj = TaA - Taz; T4b = KP707106781 * (T48 - T45); TaB = Taz + TaA; } TbZ = Tar + Tau; Tc0 = Tay + TaB; { E T3T, T3Y, Tdq, Tdr; T3T = T3L + T3S; T3Y = T3U - T3X; T3Z = FMA(KP980785280, T3T, KP195090322 * T3Y); T4s = FNMS(KP195090322, T3T, KP980785280 * T3Y); Tdq = FNMS(KP382683432, Tdf, KP923879532 * Tdg); Tdr = FMA(KP382683432, Tdi, KP923879532 * Tdj); Tds = Tdq + Tdr; TeI = Tdr - Tdq; } { E T4a, T4f, T7Y, T7Z; T4a = T42 + T49; T4f = T4b - T4e; T4g = FNMS(KP195090322, T4f, KP980785280 * T4a); T4t = FMA(KP195090322, T4a, KP980785280 * T4f); T7Y = T42 - T49; T7Z = T4e + T4b; T80 = FNMS(KP555570233, T7Z, KP831469612 * T7Y); T87 = FMA(KP555570233, T7Y, KP831469612 * T7Z); } Tdh = FMA(KP923879532, Tdf, KP382683432 * Tdg); Tdk = FNMS(KP382683432, Tdj, KP923879532 * Tdi); Tdl = Tdh + Tdk; TeE = Tdk - Tdh; { E T7V, T7W, Tav, TaC; T7V = T3L - T3S; T7W = T3X + T3U; T7X = FMA(KP831469612, T7V, KP555570233 * T7W); T86 = FNMS(KP555570233, T7V, KP831469612 * T7W); Tav = Tar - Tau; TaC = Tay - TaB; TaD = KP707106781 * (Tav + TaC); TaM = KP707106781 * (TaC - Tav); } } { E T50, TdA, T5c, TaY, T5h, TdD, T5t, Tb5, T57, TdB, T59, Tb1, T5o, TdE, T5q; E Tb8, TdC, TdF; { E T4Y, T4Z, TaW, T5a, T5b, TaX; T4Y = R1[WS(rs, 1)]; T4Z = R1[WS(rs, 33)]; TaW = T4Y + T4Z; T5a = R1[WS(rs, 17)]; T5b = R1[WS(rs, 49)]; TaX = T5a + T5b; T50 = T4Y - T4Z; TdA = TaW - TaX; T5c = T5a - T5b; TaY = TaW + TaX; } { E T5f, T5g, Tb3, T5r, T5s, Tb4; T5f = R1[WS(rs, 61)]; T5g = R1[WS(rs, 29)]; Tb3 = T5f + T5g; T5r = R1[WS(rs, 13)]; T5s = R1[WS(rs, 45)]; Tb4 = T5r + T5s; T5h = T5f - T5g; TdD = Tb3 - Tb4; T5t = T5r - T5s; Tb5 = Tb3 + Tb4; } { E T53, TaZ, T56, Tb0; { E T51, T52, T54, T55; T51 = R1[WS(rs, 9)]; T52 = R1[WS(rs, 41)]; T53 = T51 - T52; TaZ = T51 + T52; T54 = R1[WS(rs, 57)]; T55 = R1[WS(rs, 25)]; T56 = T54 - T55; Tb0 = T54 + T55; } T57 = KP707106781 * (T53 + T56); TdB = Tb0 - TaZ; T59 = KP707106781 * (T56 - T53); Tb1 = TaZ + Tb0; } { E T5k, Tb6, T5n, Tb7; { E T5i, T5j, T5l, T5m; T5i = R1[WS(rs, 5)]; T5j = R1[WS(rs, 37)]; T5k = T5i - T5j; Tb6 = T5i + T5j; T5l = R1[WS(rs, 53)]; T5m = R1[WS(rs, 21)]; T5n = T5l - T5m; Tb7 = T5l + T5m; } T5o = KP707106781 * (T5k + T5n); TdE = Tb7 - Tb6; T5q = KP707106781 * (T5n - T5k); Tb8 = Tb6 + Tb7; } Tc6 = TaY + Tb1; Tc7 = Tb5 + Tb8; { E T58, T5d, TdL, TdM; T58 = T50 + T57; T5d = T59 - T5c; T5e = FMA(KP980785280, T58, KP195090322 * T5d); T5H = FNMS(KP195090322, T58, KP980785280 * T5d); TdL = FNMS(KP382683432, TdA, KP923879532 * TdB); TdM = FMA(KP382683432, TdD, KP923879532 * TdE); TdN = TdL + TdM; TeM = TdM - TdL; } { E T5p, T5u, T8h, T8i; T5p = T5h + T5o; T5u = T5q - T5t; T5v = FNMS(KP195090322, T5u, KP980785280 * T5p); T5I = FMA(KP195090322, T5p, KP980785280 * T5u); T8h = T5h - T5o; T8i = T5t + T5q; T8j = FNMS(KP555570233, T8i, KP831469612 * T8h); T8q = FMA(KP555570233, T8h, KP831469612 * T8i); } TdC = FMA(KP923879532, TdA, KP382683432 * TdB); TdF = FNMS(KP382683432, TdE, KP923879532 * TdD); TdG = TdC + TdF; TeO = TdF - TdC; { E T8e, T8f, Tb2, Tb9; T8e = T50 - T57; T8f = T5c + T59; T8g = FMA(KP831469612, T8e, KP555570233 * T8f); T8p = FNMS(KP555570233, T8e, KP831469612 * T8f); Tb2 = TaY - Tb1; Tb9 = Tb5 - Tb8; Tba = KP707106781 * (Tb2 + Tb9); Tbj = KP707106781 * (Tb9 - Tb2); } } { E T11, TbV, Tc9, Tcf, T22, Tcb, Tc2, Tce; { E Tv, T10, Tc5, Tc8; Tv = Tf + Tu; T10 = TK + TZ; T11 = Tv + T10; TbV = Tv - T10; Tc5 = Tc3 + Tc4; Tc8 = Tc6 + Tc7; Tc9 = Tc5 - Tc8; Tcf = Tc5 + Tc8; } { E T1w, T21, TbY, Tc1; T1w = T1g + T1v; T21 = T1L + T20; T22 = T1w + T21; Tcb = T21 - T1w; TbY = TbW + TbX; Tc1 = TbZ + Tc0; Tc2 = TbY - Tc1; Tce = TbY + Tc1; } Cr[WS(csr, 32)] = T11 - T22; Ci[WS(csi, 32)] = Tcf - Tce; { E Tca, Tcc, Tcd, Tcg; Tca = KP707106781 * (Tc2 + Tc9); Cr[WS(csr, 48)] = TbV - Tca; Cr[WS(csr, 16)] = TbV + Tca; Tcc = KP707106781 * (Tc9 - Tc2); Ci[WS(csi, 16)] = Tcb + Tcc; Ci[WS(csi, 48)] = Tcc - Tcb; Tcd = T11 + T22; Tcg = Tce + Tcf; Cr[WS(csr, 64)] = Tcd - Tcg; Cr[0] = Tcd + Tcg; } } { E Tch, Tcu, Tck, Tct, Tco, Tcy, Tcr, Tcz, Tci, Tcj; Tch = Tf - Tu; Tcu = TZ - TK; Tci = T1g - T1v; Tcj = T1L - T20; Tck = KP707106781 * (Tci + Tcj); Tct = KP707106781 * (Tcj - Tci); { E Tcm, Tcn, Tcp, Tcq; Tcm = TbW - TbX; Tcn = Tc0 - TbZ; Tco = FMA(KP923879532, Tcm, KP382683432 * Tcn); Tcy = FNMS(KP382683432, Tcm, KP923879532 * Tcn); Tcp = Tc3 - Tc4; Tcq = Tc7 - Tc6; Tcr = FNMS(KP382683432, Tcq, KP923879532 * Tcp); Tcz = FMA(KP382683432, Tcp, KP923879532 * Tcq); } { E Tcl, Tcs, Tcx, TcA; Tcl = Tch + Tck; Tcs = Tco + Tcr; Cr[WS(csr, 56)] = Tcl - Tcs; Cr[WS(csr, 8)] = Tcl + Tcs; Tcx = Tcu + Tct; TcA = Tcy + Tcz; Ci[WS(csi, 8)] = Tcx + TcA; Ci[WS(csi, 56)] = TcA - Tcx; } { E Tcv, Tcw, TcB, TcC; Tcv = Tct - Tcu; Tcw = Tcr - Tco; Ci[WS(csi, 24)] = Tcv + Tcw; Ci[WS(csi, 40)] = Tcw - Tcv; TcB = Tch - Tck; TcC = Tcz - Tcy; Cr[WS(csr, 40)] = TcB - TcC; Cr[WS(csr, 24)] = TcB + TcC; } } { E Ta9, TbB, Tbs, TbM, Tag, TbL, TbJ, TbR, TaO, Tbw, Tbp, TbC, TbG, TbQ, Tbl; E Tbx, Ta8, Tbr; Ta8 = KP707106781 * (Ta6 + Ta7); Ta9 = Ta5 + Ta8; TbB = Ta5 - Ta8; Tbr = KP707106781 * (Ta7 - Ta6); Tbs = Tbq + Tbr; TbM = Tbr - Tbq; { E Tac, Taf, TbH, TbI; Tac = FMA(KP923879532, Taa, KP382683432 * Tab); Taf = FNMS(KP382683432, Tae, KP923879532 * Tad); Tag = Tac + Taf; TbL = Taf - Tac; TbH = TaV - Tba; TbI = Tbj - Tbi; TbJ = FNMS(KP555570233, TbI, KP831469612 * TbH); TbR = FMA(KP555570233, TbH, KP831469612 * TbI); } { E TaE, TaN, Tbn, Tbo; TaE = Tao + TaD; TaN = TaL + TaM; TaO = FMA(KP980785280, TaE, KP195090322 * TaN); Tbw = FNMS(KP195090322, TaE, KP980785280 * TaN); Tbn = FNMS(KP382683432, Taa, KP923879532 * Tab); Tbo = FMA(KP382683432, Tad, KP923879532 * Tae); Tbp = Tbn + Tbo; TbC = Tbo - Tbn; } { E TbE, TbF, Tbb, Tbk; TbE = Tao - TaD; TbF = TaM - TaL; TbG = FMA(KP831469612, TbE, KP555570233 * TbF); TbQ = FNMS(KP555570233, TbE, KP831469612 * TbF); Tbb = TaV + Tba; Tbk = Tbi + Tbj; Tbl = FNMS(KP195090322, Tbk, KP980785280 * Tbb); Tbx = FMA(KP195090322, Tbb, KP980785280 * Tbk); } { E Tah, Tbm, Tbv, Tby; Tah = Ta9 + Tag; Tbm = TaO + Tbl; Cr[WS(csr, 60)] = Tah - Tbm; Cr[WS(csr, 4)] = Tah + Tbm; Tbv = Tbs + Tbp; Tby = Tbw + Tbx; Ci[WS(csi, 4)] = Tbv + Tby; Ci[WS(csi, 60)] = Tby - Tbv; } { E Tbt, Tbu, Tbz, TbA; Tbt = Tbp - Tbs; Tbu = Tbl - TaO; Ci[WS(csi, 28)] = Tbt + Tbu; Ci[WS(csi, 36)] = Tbu - Tbt; Tbz = Ta9 - Tag; TbA = Tbx - Tbw; Cr[WS(csr, 36)] = Tbz - TbA; Cr[WS(csr, 28)] = Tbz + TbA; } { E TbD, TbK, TbP, TbS; TbD = TbB + TbC; TbK = TbG + TbJ; Cr[WS(csr, 52)] = TbD - TbK; Cr[WS(csr, 12)] = TbD + TbK; TbP = TbM + TbL; TbS = TbQ + TbR; Ci[WS(csi, 12)] = TbP + TbS; Ci[WS(csi, 52)] = TbS - TbP; } { E TbN, TbO, TbT, TbU; TbN = TbL - TbM; TbO = TbJ - TbG; Ci[WS(csi, 20)] = TbN + TbO; Ci[WS(csi, 44)] = TbO - TbN; TbT = TbB - TbC; TbU = TbR - TbQ; Cr[WS(csr, 44)] = TbT - TbU; Cr[WS(csr, 20)] = TbT + TbU; } } { E Tev, Tf7, Tfc, Tfm, Tff, Tfn, TeC, Tfh, TeK, Tf2, TeV, Tf8, TeY, Tfi, TeR; E Tf3; { E Tet, Teu, Tfa, Tfb; Tet = TcD - TcG; Teu = TdY - TdX; Tev = Tet - Teu; Tf7 = Tet + Teu; Tfa = TeF + TeE; Tfb = TeH + TeI; Tfc = FMA(KP290284677, Tfa, KP956940335 * Tfb); Tfm = FNMS(KP290284677, Tfb, KP956940335 * Tfa); } { E Tfd, Tfe, Tey, TeB; Tfd = TeL + TeM; Tfe = TeP + TeO; Tff = FNMS(KP290284677, Tfe, KP956940335 * Tfd); Tfn = FMA(KP956940335, Tfe, KP290284677 * Tfd); Tey = FMA(KP555570233, Tew, KP831469612 * Tex); TeB = FNMS(KP555570233, TeA, KP831469612 * Tez); TeC = Tey - TeB; Tfh = TeB + Tey; } { E TeG, TeJ, TeT, TeU; TeG = TeE - TeF; TeJ = TeH - TeI; TeK = FMA(KP471396736, TeG, KP881921264 * TeJ); Tf2 = FNMS(KP471396736, TeJ, KP881921264 * TeG); TeT = FNMS(KP555570233, Tex, KP831469612 * Tew); TeU = FMA(KP831469612, TeA, KP555570233 * Tez); TeV = TeT - TeU; Tf8 = TeU + TeT; } { E TeW, TeX, TeN, TeQ; TeW = TcN - TcK; TeX = TdV - TdU; TeY = TeW - TeX; Tfi = TeX + TeW; TeN = TeL - TeM; TeQ = TeO - TeP; TeR = FNMS(KP471396736, TeQ, KP881921264 * TeN); Tf3 = FMA(KP881921264, TeQ, KP471396736 * TeN); } { E TeD, TeS, Tf1, Tf4; TeD = Tev + TeC; TeS = TeK + TeR; Cr[WS(csr, 54)] = TeD - TeS; Cr[WS(csr, 10)] = TeD + TeS; Tf1 = TeY + TeV; Tf4 = Tf2 + Tf3; Ci[WS(csi, 10)] = Tf1 + Tf4; Ci[WS(csi, 54)] = Tf4 - Tf1; } { E TeZ, Tf0, Tf5, Tf6; TeZ = TeV - TeY; Tf0 = TeR - TeK; Ci[WS(csi, 22)] = TeZ + Tf0; Ci[WS(csi, 42)] = Tf0 - TeZ; Tf5 = Tev - TeC; Tf6 = Tf3 - Tf2; Cr[WS(csr, 42)] = Tf5 - Tf6; Cr[WS(csr, 22)] = Tf5 + Tf6; } { E Tf9, Tfg, Tfl, Tfo; Tf9 = Tf7 + Tf8; Tfg = Tfc + Tff; Cr[WS(csr, 58)] = Tf9 - Tfg; Cr[WS(csr, 6)] = Tf9 + Tfg; Tfl = Tfi + Tfh; Tfo = Tfm + Tfn; Ci[WS(csi, 6)] = Tfl + Tfo; Ci[WS(csi, 58)] = Tfo - Tfl; } { E Tfj, Tfk, Tfp, Tfq; Tfj = Tfh - Tfi; Tfk = Tff - Tfc; Ci[WS(csi, 26)] = Tfj + Tfk; Ci[WS(csi, 38)] = Tfk - Tfj; Tfp = Tf7 - Tf8; Tfq = Tfn - Tfm; Cr[WS(csr, 38)] = Tfp - Tfq; Cr[WS(csr, 26)] = Tfp + Tfq; } } { E TcP, Te9, Tee, Teo, Teh, Tep, Td8, Tej, Tdu, Te4, TdT, Tea, Te0, Tek, TdP; E Te5; { E TcH, TcO, Tec, Ted; TcH = TcD + TcG; TcO = TcK + TcN; TcP = TcH + TcO; Te9 = TcH - TcO; Tec = Tde - Tdl; Ted = Tds - Tdp; Tee = FMA(KP773010453, Tec, KP634393284 * Ted); Teo = FNMS(KP634393284, Tec, KP773010453 * Ted); } { E Tef, Teg, TcY, Td7; Tef = Tdz - TdG; Teg = TdN - TdK; Teh = FNMS(KP634393284, Teg, KP773010453 * Tef); Tep = FMA(KP634393284, Tef, KP773010453 * Teg); TcY = FMA(KP980785280, TcU, KP195090322 * TcX); Td7 = FNMS(KP195090322, Td6, KP980785280 * Td3); Td8 = TcY + Td7; Tej = Td7 - TcY; } { E Tdm, Tdt, TdR, TdS; Tdm = Tde + Tdl; Tdt = Tdp + Tds; Tdu = FMA(KP995184726, Tdm, KP098017140 * Tdt); Te4 = FNMS(KP098017140, Tdm, KP995184726 * Tdt); TdR = FNMS(KP195090322, TcU, KP980785280 * TcX); TdS = FMA(KP195090322, Td3, KP980785280 * Td6); TdT = TdR + TdS; Tea = TdS - TdR; } { E TdW, TdZ, TdH, TdO; TdW = TdU + TdV; TdZ = TdX + TdY; Te0 = TdW + TdZ; Tek = TdZ - TdW; TdH = Tdz + TdG; TdO = TdK + TdN; TdP = FNMS(KP098017140, TdO, KP995184726 * TdH); Te5 = FMA(KP098017140, TdH, KP995184726 * TdO); } { E Td9, TdQ, Te3, Te6; Td9 = TcP + Td8; TdQ = Tdu + TdP; Cr[WS(csr, 62)] = Td9 - TdQ; Cr[WS(csr, 2)] = Td9 + TdQ; Te3 = Te0 + TdT; Te6 = Te4 + Te5; Ci[WS(csi, 2)] = Te3 + Te6; Ci[WS(csi, 62)] = Te6 - Te3; } { E Te1, Te2, Te7, Te8; Te1 = TdT - Te0; Te2 = TdP - Tdu; Ci[WS(csi, 30)] = Te1 + Te2; Ci[WS(csi, 34)] = Te2 - Te1; Te7 = TcP - Td8; Te8 = Te5 - Te4; Cr[WS(csr, 34)] = Te7 - Te8; Cr[WS(csr, 30)] = Te7 + Te8; } { E Teb, Tei, Ten, Teq; Teb = Te9 + Tea; Tei = Tee + Teh; Cr[WS(csr, 50)] = Teb - Tei; Cr[WS(csr, 14)] = Teb + Tei; Ten = Tek + Tej; Teq = Teo + Tep; Ci[WS(csi, 14)] = Ten + Teq; Ci[WS(csi, 50)] = Teq - Ten; } { E Tel, Tem, Ter, Tes; Tel = Tej - Tek; Tem = Teh - Tee; Ci[WS(csi, 18)] = Tel + Tem; Ci[WS(csi, 46)] = Tem - Tel; Ter = Te9 - Tea; Tes = Tep - Teo; Cr[WS(csr, 46)] = Ter - Tes; Cr[WS(csr, 18)] = Ter + Tes; } } { E T6v, T77, T6C, T7h, T6Y, T7i, T6V, T78, T6R, T7n, T73, T7f, T6K, T7m, T72; E T7c; { E T6t, T6u, T6T, T6U; T6t = T27 - T2e; T6u = T5Y - T5X; T6v = T6t - T6u; T77 = T6t + T6u; { E T6y, T6B, T6W, T6X; T6y = FMA(KP773010453, T6w, KP634393284 * T6x); T6B = FNMS(KP634393284, T6A, KP773010453 * T6z); T6C = T6y - T6B; T7h = T6B + T6y; T6W = T2x - T2o; T6X = T5V - T5S; T6Y = T6W - T6X; T7i = T6X + T6W; } T6T = FNMS(KP634393284, T6w, KP773010453 * T6x); T6U = FMA(KP634393284, T6z, KP773010453 * T6A); T6V = T6T - T6U; T78 = T6U + T6T; { E T6N, T7d, T6Q, T7e, T6M, T6O; T6M = T5I - T5H; T6N = T6L - T6M; T7d = T6L + T6M; T6O = T5v - T5e; T6Q = T6O - T6P; T7e = T6P + T6O; T6R = FNMS(KP427555093, T6Q, KP903989293 * T6N); T7n = FMA(KP941544065, T7e, KP336889853 * T7d); T73 = FMA(KP903989293, T6Q, KP427555093 * T6N); T7f = FNMS(KP336889853, T7e, KP941544065 * T7d); } { E T6G, T7a, T6J, T7b, T6E, T6I; T6E = T4g - T3Z; T6G = T6E - T6F; T7a = T6F + T6E; T6I = T4t - T4s; T6J = T6H - T6I; T7b = T6H + T6I; T6K = FMA(KP427555093, T6G, KP903989293 * T6J); T7m = FNMS(KP336889853, T7b, KP941544065 * T7a); T72 = FNMS(KP427555093, T6J, KP903989293 * T6G); T7c = FMA(KP336889853, T7a, KP941544065 * T7b); } } { E T6D, T6S, T71, T74; T6D = T6v + T6C; T6S = T6K + T6R; Cr[WS(csr, 55)] = T6D - T6S; Cr[WS(csr, 9)] = T6D + T6S; T71 = T6Y + T6V; T74 = T72 + T73; Ci[WS(csi, 9)] = T71 + T74; Ci[WS(csi, 55)] = T74 - T71; } { E T6Z, T70, T75, T76; T6Z = T6V - T6Y; T70 = T6R - T6K; Ci[WS(csi, 23)] = T6Z + T70; Ci[WS(csi, 41)] = T70 - T6Z; T75 = T6v - T6C; T76 = T73 - T72; Cr[WS(csr, 41)] = T75 - T76; Cr[WS(csr, 23)] = T75 + T76; } { E T79, T7g, T7l, T7o; T79 = T77 + T78; T7g = T7c + T7f; Cr[WS(csr, 57)] = T79 - T7g; Cr[WS(csr, 7)] = T79 + T7g; T7l = T7i + T7h; T7o = T7m + T7n; Ci[WS(csi, 7)] = T7l + T7o; Ci[WS(csi, 57)] = T7o - T7l; } { E T7j, T7k, T7p, T7q; T7j = T7h - T7i; T7k = T7f - T7c; Ci[WS(csi, 25)] = T7j + T7k; Ci[WS(csi, 39)] = T7k - T7j; T7p = T77 - T78; T7q = T7n - T7m; Cr[WS(csr, 39)] = T7p - T7q; Cr[WS(csr, 25)] = T7p + T7q; } } { E T99, T9L, T9g, T9V, T9C, T9W, T9z, T9M, T9v, Ta1, T9H, T9T, T9o, Ta0, T9G; E T9Q; { E T97, T98, T9x, T9y; T97 = T7r - T7s; T98 = T8C - T8B; T99 = T97 - T98; T9L = T97 + T98; { E T9c, T9f, T9A, T9B; T9c = FMA(KP471396736, T9a, KP881921264 * T9b); T9f = FNMS(KP471396736, T9e, KP881921264 * T9d); T9g = T9c - T9f; T9V = T9f + T9c; T9A = T7z - T7w; T9B = T8z - T8y; T9C = T9A - T9B; T9W = T9B + T9A; } T9x = FNMS(KP471396736, T9b, KP881921264 * T9a); T9y = FMA(KP881921264, T9e, KP471396736 * T9d); T9z = T9x - T9y; T9M = T9y + T9x; { E T9r, T9R, T9u, T9S, T9q, T9s; T9q = T8q - T8p; T9r = T9p - T9q; T9R = T9p + T9q; T9s = T8j - T8g; T9u = T9s - T9t; T9S = T9t + T9s; T9v = FNMS(KP514102744, T9u, KP857728610 * T9r); Ta1 = FMA(KP970031253, T9S, KP242980179 * T9R); T9H = FMA(KP857728610, T9u, KP514102744 * T9r); T9T = FNMS(KP242980179, T9S, KP970031253 * T9R); } { E T9k, T9O, T9n, T9P, T9i, T9m; T9i = T80 - T7X; T9k = T9i - T9j; T9O = T9j + T9i; T9m = T87 - T86; T9n = T9l - T9m; T9P = T9l + T9m; T9o = FMA(KP514102744, T9k, KP857728610 * T9n); Ta0 = FNMS(KP242980179, T9P, KP970031253 * T9O); T9G = FNMS(KP514102744, T9n, KP857728610 * T9k); T9Q = FMA(KP242980179, T9O, KP970031253 * T9P); } } { E T9h, T9w, T9F, T9I; T9h = T99 + T9g; T9w = T9o + T9v; Cr[WS(csr, 53)] = T9h - T9w; Cr[WS(csr, 11)] = T9h + T9w; T9F = T9C + T9z; T9I = T9G + T9H; Ci[WS(csi, 11)] = T9F + T9I; Ci[WS(csi, 53)] = T9I - T9F; } { E T9D, T9E, T9J, T9K; T9D = T9z - T9C; T9E = T9v - T9o; Ci[WS(csi, 21)] = T9D + T9E; Ci[WS(csi, 43)] = T9E - T9D; T9J = T99 - T9g; T9K = T9H - T9G; Cr[WS(csr, 43)] = T9J - T9K; Cr[WS(csr, 21)] = T9J + T9K; } { E T9N, T9U, T9Z, Ta2; T9N = T9L + T9M; T9U = T9Q + T9T; Cr[WS(csr, 59)] = T9N - T9U; Cr[WS(csr, 5)] = T9N + T9U; T9Z = T9W + T9V; Ta2 = Ta0 + Ta1; Ci[WS(csi, 5)] = T9Z + Ta2; Ci[WS(csi, 59)] = Ta2 - T9Z; } { E T9X, T9Y, Ta3, Ta4; T9X = T9V - T9W; T9Y = T9T - T9Q; Ci[WS(csi, 27)] = T9X + T9Y; Ci[WS(csi, 37)] = T9Y - T9X; Ta3 = T9L - T9M; Ta4 = Ta1 - Ta0; Cr[WS(csr, 37)] = Ta3 - Ta4; Cr[WS(csr, 27)] = Ta3 + Ta4; } } { E T2z, T69, T3g, T6j, T60, T6k, T5P, T6a, T5L, T6p, T65, T6h, T4w, T6o, T64; E T6e; { E T2f, T2y, T5N, T5O; T2f = T27 + T2e; T2y = T2o + T2x; T2z = T2f + T2y; T69 = T2f - T2y; { E T2U, T3f, T5W, T5Z; T2U = FMA(KP098017140, T2M, KP995184726 * T2T); T3f = FNMS(KP098017140, T3e, KP995184726 * T37); T3g = T2U + T3f; T6j = T3f - T2U; T5W = T5S + T5V; T5Z = T5X + T5Y; T60 = T5W + T5Z; T6k = T5Z - T5W; } T5N = FNMS(KP098017140, T2T, KP995184726 * T2M); T5O = FMA(KP995184726, T3e, KP098017140 * T37); T5P = T5N + T5O; T6a = T5O - T5N; { E T5x, T6f, T5K, T6g, T5w, T5J; T5w = T5e + T5v; T5x = T4X + T5w; T6f = T4X - T5w; T5J = T5H + T5I; T5K = T5G + T5J; T6g = T5J - T5G; T5L = FNMS(KP049067674, T5K, KP998795456 * T5x); T6p = FMA(KP671558954, T6f, KP740951125 * T6g); T65 = FMA(KP049067674, T5x, KP998795456 * T5K); T6h = FNMS(KP671558954, T6g, KP740951125 * T6f); } { E T4i, T6c, T4v, T6d, T4h, T4u; T4h = T3Z + T4g; T4i = T3I + T4h; T6c = T3I - T4h; T4u = T4s + T4t; T4v = T4r + T4u; T6d = T4u - T4r; T4w = FMA(KP998795456, T4i, KP049067674 * T4v); T6o = FNMS(KP671558954, T6c, KP740951125 * T6d); T64 = FNMS(KP049067674, T4i, KP998795456 * T4v); T6e = FMA(KP740951125, T6c, KP671558954 * T6d); } } { E T3h, T5M, T63, T66; T3h = T2z + T3g; T5M = T4w + T5L; Cr[WS(csr, 63)] = T3h - T5M; Cr[WS(csr, 1)] = T3h + T5M; T63 = T60 + T5P; T66 = T64 + T65; Ci[WS(csi, 1)] = T63 + T66; Ci[WS(csi, 63)] = T66 - T63; } { E T61, T62, T67, T68; T61 = T5P - T60; T62 = T5L - T4w; Ci[WS(csi, 31)] = T61 + T62; Ci[WS(csi, 33)] = T62 - T61; T67 = T2z - T3g; T68 = T65 - T64; Cr[WS(csr, 33)] = T67 - T68; Cr[WS(csr, 31)] = T67 + T68; } { E T6b, T6i, T6n, T6q; T6b = T69 + T6a; T6i = T6e + T6h; Cr[WS(csr, 49)] = T6b - T6i; Cr[WS(csr, 15)] = T6b + T6i; T6n = T6k + T6j; T6q = T6o + T6p; Ci[WS(csi, 15)] = T6n + T6q; Ci[WS(csi, 49)] = T6q - T6n; } { E T6l, T6m, T6r, T6s; T6l = T6j - T6k; T6m = T6h - T6e; Ci[WS(csi, 17)] = T6l + T6m; Ci[WS(csi, 47)] = T6m - T6l; T6r = T69 - T6a; T6s = T6p - T6o; Cr[WS(csr, 47)] = T6r - T6s; Cr[WS(csr, 17)] = T6r + T6s; } } { E T7B, T8N, T7Q, T8X, T8E, T8Y, T8x, T8O, T8t, T93, T8J, T8V, T8a, T92, T8I; E T8S; { E T7t, T7A, T8v, T8w; T7t = T7r + T7s; T7A = T7w + T7z; T7B = T7t + T7A; T8N = T7t - T7A; { E T7I, T7P, T8A, T8D; T7I = FMA(KP956940335, T7E, KP290284677 * T7H); T7P = FNMS(KP290284677, T7O, KP956940335 * T7L); T7Q = T7I + T7P; T8X = T7P - T7I; T8A = T8y + T8z; T8D = T8B + T8C; T8E = T8A + T8D; T8Y = T8D - T8A; } T8v = FNMS(KP290284677, T7E, KP956940335 * T7H); T8w = FMA(KP290284677, T7L, KP956940335 * T7O); T8x = T8v + T8w; T8O = T8w - T8v; { E T8l, T8T, T8s, T8U, T8k, T8r; T8k = T8g + T8j; T8l = T8d + T8k; T8T = T8d - T8k; T8r = T8p + T8q; T8s = T8o + T8r; T8U = T8r - T8o; T8t = FNMS(KP146730474, T8s, KP989176509 * T8l); T93 = FMA(KP595699304, T8T, KP803207531 * T8U); T8J = FMA(KP146730474, T8l, KP989176509 * T8s); T8V = FNMS(KP595699304, T8U, KP803207531 * T8T); } { E T82, T8Q, T89, T8R, T81, T88; T81 = T7X + T80; T82 = T7U + T81; T8Q = T7U - T81; T88 = T86 + T87; T89 = T85 + T88; T8R = T88 - T85; T8a = FMA(KP989176509, T82, KP146730474 * T89); T92 = FNMS(KP595699304, T8Q, KP803207531 * T8R); T8I = FNMS(KP146730474, T82, KP989176509 * T89); T8S = FMA(KP803207531, T8Q, KP595699304 * T8R); } } { E T7R, T8u, T8H, T8K; T7R = T7B + T7Q; T8u = T8a + T8t; Cr[WS(csr, 61)] = T7R - T8u; Cr[WS(csr, 3)] = T7R + T8u; T8H = T8E + T8x; T8K = T8I + T8J; Ci[WS(csi, 3)] = T8H + T8K; Ci[WS(csi, 61)] = T8K - T8H; } { E T8F, T8G, T8L, T8M; T8F = T8x - T8E; T8G = T8t - T8a; Ci[WS(csi, 29)] = T8F + T8G; Ci[WS(csi, 35)] = T8G - T8F; T8L = T7B - T7Q; T8M = T8J - T8I; Cr[WS(csr, 35)] = T8L - T8M; Cr[WS(csr, 29)] = T8L + T8M; } { E T8P, T8W, T91, T94; T8P = T8N + T8O; T8W = T8S + T8V; Cr[WS(csr, 51)] = T8P - T8W; Cr[WS(csr, 13)] = T8P + T8W; T91 = T8Y + T8X; T94 = T92 + T93; Ci[WS(csi, 13)] = T91 + T94; Ci[WS(csi, 51)] = T94 - T91; } { E T8Z, T90, T95, T96; T8Z = T8X - T8Y; T90 = T8V - T8S; Ci[WS(csi, 19)] = T8Z + T90; Ci[WS(csi, 45)] = T90 - T8Z; T95 = T8N - T8O; T96 = T93 - T92; Cr[WS(csr, 45)] = T95 - T96; Cr[WS(csr, 19)] = T95 + T96; } } } } } static const kr2c_desc desc = { 128, "r2cf_128", {812, 186, 144, 0}, &GENUS }; void X(codelet_r2cf_128) (planner *p) { X(kr2c_register) (p, r2cf_128, &desc); } #endif /* HAVE_FMA */ fftw-3.3.3/rdft/scalar/r2cf/r2cf_20.c0000644000175000017500000002442712054410624013757 00000000000000/* * Copyright (c) 2003, 2007-11 Matteo Frigo * Copyright (c) 2003, 2007-11 Massachusetts Institute of Technology * * This program is free software; you can redistribute it and/or modify * it under the terms of the GNU General Public License as published by * the Free Software Foundation; either version 2 of the License, or * (at your option) any later version. * * This program is distributed in the hope that it will be useful, * but WITHOUT ANY WARRANTY; without even the implied warranty of * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the * GNU General Public License for more details. * * You should have received a copy of the GNU General Public License * along with this program; if not, write to the Free Software * Foundation, Inc., 51 Franklin Street, Fifth Floor, Boston, MA 02110-1301 USA * */ /* This file was automatically generated --- DO NOT EDIT */ /* Generated on Sun Nov 25 07:39:48 EST 2012 */ #include "codelet-rdft.h" #ifdef HAVE_FMA /* Generated by: ../../../genfft/gen_r2cf.native -fma -reorder-insns -schedule-for-pipeline -compact -variables 4 -pipeline-latency 4 -n 20 -name r2cf_20 -include r2cf.h */ /* * This function contains 86 FP additions, 32 FP multiplications, * (or, 58 additions, 4 multiplications, 28 fused multiply/add), * 70 stack variables, 4 constants, and 40 memory accesses */ #include "r2cf.h" static void r2cf_20(R *R0, R *R1, R *Cr, R *Ci, stride rs, stride csr, stride csi, INT v, INT ivs, INT ovs) { DK(KP559016994, +0.559016994374947424102293417182819058860154590); DK(KP250000000, +0.250000000000000000000000000000000000000000000); DK(KP618033988, +0.618033988749894848204586834365638117720309180); DK(KP951056516, +0.951056516295153572116439333379382143405698634); { INT i; for (i = v; i > 0; i = i - 1, R0 = R0 + ivs, R1 = R1 + ivs, Cr = Cr + ovs, Ci = Ci + ovs, MAKE_VOLATILE_STRIDE(80, rs), MAKE_VOLATILE_STRIDE(80, csr), MAKE_VOLATILE_STRIDE(80, csi)) { E T1i, T1c, T1a, T1o, T1m, T1h, T1b, T13, T1j, T1n; { E T3, T1d, TJ, TV, T1k, T16, T19, T1l, Ty, Ti, T12, TD, T1g, TR, TX; E TK, Tt, TU, TW, TL, TE; { E T1, T2, TG, TH; T1 = R0[0]; T2 = R0[WS(rs, 5)]; TG = R1[WS(rs, 2)]; TH = R1[WS(rs, 7)]; { E T6, To, T17, Tx, T18, TC, Tj, T9, Tp, Tu, Td, T15, Tm, Tq, Te; E Tf; { E TA, TB, T7, T8; { E T4, TF, TI, T5, Tv, Tw; T4 = R0[WS(rs, 2)]; T3 = T1 - T2; TF = T1 + T2; T1d = TG - TH; TI = TG + TH; T5 = R0[WS(rs, 7)]; Tv = R1[WS(rs, 6)]; Tw = R1[WS(rs, 1)]; TJ = TF - TI; TV = TF + TI; T6 = T4 - T5; To = T4 + T5; T17 = Tw - Tv; Tx = Tv + Tw; } TA = R1[WS(rs, 8)]; TB = R1[WS(rs, 3)]; T7 = R0[WS(rs, 8)]; T8 = R0[WS(rs, 3)]; { E Tb, Tc, Tk, Tl; Tb = R0[WS(rs, 4)]; T18 = TB - TA; TC = TA + TB; Tj = T7 + T8; T9 = T7 - T8; Tc = R0[WS(rs, 9)]; Tk = R1[0]; Tl = R1[WS(rs, 5)]; Tp = R1[WS(rs, 4)]; Tu = Tb + Tc; Td = Tb - Tc; T15 = Tl - Tk; Tm = Tk + Tl; Tq = R1[WS(rs, 9)]; Te = R0[WS(rs, 6)]; Tf = R0[WS(rs, 1)]; } } { E Ta, Tr, Tz, T1e, T1f, Th, T14, Tg, TP, TQ; Ta = T6 + T9; T1k = T6 - T9; T14 = Tq - Tp; Tr = Tp + Tq; Tz = Te + Tf; Tg = Te - Tf; T16 = T14 - T15; T1e = T14 + T15; T1f = T17 + T18; T19 = T17 - T18; Th = Td + Tg; T1l = Td - Tg; Ty = Tu - Tx; TP = Tu + Tx; Ti = Ta + Th; T12 = Ta - Th; TD = Tz - TC; TQ = Tz + TC; T1g = T1e + T1f; T1i = T1e - T1f; { E TT, Tn, Ts, TS; TT = Tj + Tm; Tn = Tj - Tm; Ts = To - Tr; TS = To + Tr; TR = TP - TQ; TX = TP + TQ; TK = Ts + Tn; Tt = Tn - Ts; TU = TS - TT; TW = TS + TT; } } } } Cr[WS(csr, 5)] = T3 + Ti; Ci[WS(csi, 5)] = T1g - T1d; TL = Ty + TD; TE = Ty - TD; { E TY, T10, TM, TO, T11, TZ, TN; TY = TW + TX; T10 = TW - TX; Ci[WS(csi, 2)] = KP951056516 * (FMA(KP618033988, Tt, TE)); Ci[WS(csi, 6)] = KP951056516 * (FNMS(KP618033988, TE, Tt)); Ci[WS(csi, 4)] = KP951056516 * (FMA(KP618033988, TR, TU)); Ci[WS(csi, 8)] = -(KP951056516 * (FNMS(KP618033988, TU, TR))); TM = TK + TL; TO = TK - TL; T1c = FNMS(KP618033988, T16, T19); T1a = FMA(KP618033988, T19, T16); Cr[0] = TV + TY; TZ = FNMS(KP250000000, TY, TV); Cr[WS(csr, 10)] = TJ + TM; TN = FNMS(KP250000000, TM, TJ); Cr[WS(csr, 8)] = FNMS(KP559016994, T10, TZ); Cr[WS(csr, 4)] = FMA(KP559016994, T10, TZ); Cr[WS(csr, 6)] = FMA(KP559016994, TO, TN); Cr[WS(csr, 2)] = FNMS(KP559016994, TO, TN); T11 = FNMS(KP250000000, Ti, T3); T1o = FNMS(KP618033988, T1k, T1l); T1m = FMA(KP618033988, T1l, T1k); T1h = FMA(KP250000000, T1g, T1d); T1b = FNMS(KP559016994, T12, T11); T13 = FMA(KP559016994, T12, T11); } } Cr[WS(csr, 3)] = FNMS(KP951056516, T1c, T1b); Cr[WS(csr, 7)] = FMA(KP951056516, T1c, T1b); Cr[WS(csr, 1)] = FMA(KP951056516, T1a, T13); Cr[WS(csr, 9)] = FNMS(KP951056516, T1a, T13); T1j = FNMS(KP559016994, T1i, T1h); T1n = FMA(KP559016994, T1i, T1h); Ci[WS(csi, 3)] = FNMS(KP951056516, T1o, T1n); Ci[WS(csi, 7)] = FMA(KP951056516, T1o, T1n); Ci[WS(csi, 9)] = FMS(KP951056516, T1m, T1j); Ci[WS(csi, 1)] = -(FMA(KP951056516, T1m, T1j)); } } } static const kr2c_desc desc = { 20, "r2cf_20", {58, 4, 28, 0}, &GENUS }; void X(codelet_r2cf_20) (planner *p) { X(kr2c_register) (p, r2cf_20, &desc); } #else /* HAVE_FMA */ /* Generated by: ../../../genfft/gen_r2cf.native -compact -variables 4 -pipeline-latency 4 -n 20 -name r2cf_20 -include r2cf.h */ /* * This function contains 86 FP additions, 24 FP multiplications, * (or, 74 additions, 12 multiplications, 12 fused multiply/add), * 51 stack variables, 4 constants, and 40 memory accesses */ #include "r2cf.h" static void r2cf_20(R *R0, R *R1, R *Cr, R *Ci, stride rs, stride csr, stride csi, INT v, INT ivs, INT ovs) { DK(KP250000000, +0.250000000000000000000000000000000000000000000); DK(KP559016994, +0.559016994374947424102293417182819058860154590); DK(KP587785252, +0.587785252292473129168705954639072768597652438); DK(KP951056516, +0.951056516295153572116439333379382143405698634); { INT i; for (i = v; i > 0; i = i - 1, R0 = R0 + ivs, R1 = R1 + ivs, Cr = Cr + ovs, Ci = Ci + ovs, MAKE_VOLATILE_STRIDE(80, rs), MAKE_VOLATILE_STRIDE(80, csr), MAKE_VOLATILE_STRIDE(80, csi)) { E T3, T1m, TF, T17, Ts, TM, TN, Tz, Ta, Th, Ti, T1g, T1h, T1k, T10; E T13, T19, TG, TH, TI, T1d, T1e, T1j, TT, TW, T18; { E T1, T2, T15, TD, TE, T16; T1 = R0[0]; T2 = R0[WS(rs, 5)]; T15 = T1 + T2; TD = R1[WS(rs, 7)]; TE = R1[WS(rs, 2)]; T16 = TE + TD; T3 = T1 - T2; T1m = T15 + T16; TF = TD - TE; T17 = T15 - T16; } { E T6, TU, Tv, T12, Ty, TZ, T9, TR, Td, TY, To, TS, Tr, TV, Tg; E T11; { E T4, T5, Tt, Tu; T4 = R0[WS(rs, 2)]; T5 = R0[WS(rs, 7)]; T6 = T4 - T5; TU = T4 + T5; Tt = R1[WS(rs, 8)]; Tu = R1[WS(rs, 3)]; Tv = Tt - Tu; T12 = Tt + Tu; } { E Tw, Tx, T7, T8; Tw = R1[WS(rs, 6)]; Tx = R1[WS(rs, 1)]; Ty = Tw - Tx; TZ = Tw + Tx; T7 = R0[WS(rs, 8)]; T8 = R0[WS(rs, 3)]; T9 = T7 - T8; TR = T7 + T8; } { E Tb, Tc, Tm, Tn; Tb = R0[WS(rs, 4)]; Tc = R0[WS(rs, 9)]; Td = Tb - Tc; TY = Tb + Tc; Tm = R1[0]; Tn = R1[WS(rs, 5)]; To = Tm - Tn; TS = Tm + Tn; } { E Tp, Tq, Te, Tf; Tp = R1[WS(rs, 4)]; Tq = R1[WS(rs, 9)]; Tr = Tp - Tq; TV = Tp + Tq; Te = R0[WS(rs, 6)]; Tf = R0[WS(rs, 1)]; Tg = Te - Tf; T11 = Te + Tf; } Ts = To - Tr; TM = T6 - T9; TN = Td - Tg; Tz = Tv - Ty; Ta = T6 + T9; Th = Td + Tg; Ti = Ta + Th; T1g = TY + TZ; T1h = T11 + T12; T1k = T1g + T1h; T10 = TY - TZ; T13 = T11 - T12; T19 = T10 + T13; TG = Tr + To; TH = Ty + Tv; TI = TG + TH; T1d = TU + TV; T1e = TR + TS; T1j = T1d + T1e; TT = TR - TS; TW = TU - TV; T18 = TW + TT; } Cr[WS(csr, 5)] = T3 + Ti; Ci[WS(csi, 5)] = TF - TI; { E TX, T14, T1f, T1i; TX = TT - TW; T14 = T10 - T13; Ci[WS(csi, 6)] = FNMS(KP587785252, T14, KP951056516 * TX); Ci[WS(csi, 2)] = FMA(KP587785252, TX, KP951056516 * T14); T1f = T1d - T1e; T1i = T1g - T1h; Ci[WS(csi, 8)] = FNMS(KP951056516, T1i, KP587785252 * T1f); Ci[WS(csi, 4)] = FMA(KP951056516, T1f, KP587785252 * T1i); } { E T1l, T1n, T1o, T1c, T1a, T1b; T1l = KP559016994 * (T1j - T1k); T1n = T1j + T1k; T1o = FNMS(KP250000000, T1n, T1m); Cr[WS(csr, 4)] = T1l + T1o; Cr[0] = T1m + T1n; Cr[WS(csr, 8)] = T1o - T1l; T1c = KP559016994 * (T18 - T19); T1a = T18 + T19; T1b = FNMS(KP250000000, T1a, T17); Cr[WS(csr, 2)] = T1b - T1c; Cr[WS(csr, 10)] = T17 + T1a; Cr[WS(csr, 6)] = T1c + T1b; } { E TA, TC, Tl, TB, Tj, Tk; TA = FMA(KP951056516, Ts, KP587785252 * Tz); TC = FNMS(KP587785252, Ts, KP951056516 * Tz); Tj = KP559016994 * (Ta - Th); Tk = FNMS(KP250000000, Ti, T3); Tl = Tj + Tk; TB = Tk - Tj; Cr[WS(csr, 9)] = Tl - TA; Cr[WS(csr, 7)] = TB + TC; Cr[WS(csr, 1)] = Tl + TA; Cr[WS(csr, 3)] = TB - TC; } { E TO, TQ, TL, TP, TJ, TK; TO = FMA(KP951056516, TM, KP587785252 * TN); TQ = FNMS(KP587785252, TM, KP951056516 * TN); TJ = FMA(KP250000000, TI, TF); TK = KP559016994 * (TH - TG); TL = TJ + TK; TP = TK - TJ; Ci[WS(csi, 1)] = TL - TO; Ci[WS(csi, 7)] = TQ + TP; Ci[WS(csi, 9)] = TO + TL; Ci[WS(csi, 3)] = TP - TQ; } } } } static const kr2c_desc desc = { 20, "r2cf_20", {74, 12, 12, 0}, &GENUS }; void X(codelet_r2cf_20) (planner *p) { X(kr2c_register) (p, r2cf_20, &desc); } #endif /* HAVE_FMA */ fftw-3.3.3/rdft/scalar/r2cf/r2cf_25.c0000644000175000017500000006761412054410631013767 00000000000000/* * Copyright (c) 2003, 2007-11 Matteo Frigo * Copyright (c) 2003, 2007-11 Massachusetts Institute of Technology * * This program is free software; you can redistribute it and/or modify * it under the terms of the GNU General Public License as published by * the Free Software Foundation; either version 2 of the License, or * (at your option) any later version. * * This program is distributed in the hope that it will be useful, * but WITHOUT ANY WARRANTY; without even the implied warranty of * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the * GNU General Public License for more details. * * You should have received a copy of the GNU General Public License * along with this program; if not, write to the Free Software * Foundation, Inc., 51 Franklin Street, Fifth Floor, Boston, MA 02110-1301 USA * */ /* This file was automatically generated --- DO NOT EDIT */ /* Generated on Sun Nov 25 07:39:48 EST 2012 */ #include "codelet-rdft.h" #ifdef HAVE_FMA /* Generated by: ../../../genfft/gen_r2cf.native -fma -reorder-insns -schedule-for-pipeline -compact -variables 4 -pipeline-latency 4 -n 25 -name r2cf_25 -include r2cf.h */ /* * This function contains 200 FP additions, 168 FP multiplications, * (or, 44 additions, 12 multiplications, 156 fused multiply/add), * 157 stack variables, 66 constants, and 50 memory accesses */ #include "r2cf.h" static void r2cf_25(R *R0, R *R1, R *Cr, R *Ci, stride rs, stride csr, stride csi, INT v, INT ivs, INT ovs) { DK(KP792626838, +0.792626838241819413632131824093538848057784557); DK(KP876091699, +0.876091699473550838204498029706869638173524346); DK(KP809385824, +0.809385824416008241660603814668679683846476688); DK(KP860541664, +0.860541664367944677098261680920518816412804187); DK(KP681693190, +0.681693190061530575150324149145440022633095390); DK(KP560319534, +0.560319534973832390111614715371676131169633784); DK(KP997675361, +0.997675361079556513670859573984492383596555031); DK(KP237294955, +0.237294955877110315393888866460840817927895961); DK(KP897376177, +0.897376177523557693138608077137219684419427330); DK(KP923225144, +0.923225144846402650453449441572664695995209956); DK(KP956723877, +0.956723877038460305821989399535483155872969262); DK(KP949179823, +0.949179823508441261575555465843363271711583843); DK(KP669429328, +0.669429328479476605641803240971985825917022098); DK(KP570584518, +0.570584518783621657366766175430996792655723863); DK(KP262346850, +0.262346850930607871785420028382979691334784273); DK(KP876306680, +0.876306680043863587308115903922062583399064238); DK(KP906616052, +0.906616052148196230441134447086066874408359177); DK(KP683113946, +0.683113946453479238701949862233725244439656928); DK(KP559154169, +0.559154169276087864842202529084232643714075927); DK(KP921078979, +0.921078979742360627699756128143719920817673854); DK(KP904508497, +0.904508497187473712051146708591409529430077295); DK(KP999754674, +0.999754674276473633366203429228112409535557487); DK(KP968583161, +0.968583161128631119490168375464735813836012403); DK(KP242145790, +0.242145790282157779872542093866183953459003101); DK(KP904730450, +0.904730450839922351881287709692877908104763647); DK(KP845997307, +0.845997307939530944175097360758058292389769300); DK(KP855719849, +0.855719849902058969314654733608091555096772472); DK(KP982009705, +0.982009705009746369461829878184175962711969869); DK(KP916574801, +0.916574801383451584742370439148878693530976769); DK(KP690983005, +0.690983005625052575897706582817180941139845410); DK(KP952936919, +0.952936919628306576880750665357914584765951388); DK(KP998026728, +0.998026728428271561952336806863450553336905220); DK(KP831864738, +0.831864738706457140726048799369896829771167132); DK(KP803003575, +0.803003575438660414833440593570376004635464850); DK(KP522616830, +0.522616830205754336872861364785224694908468440); DK(KP829049696, +0.829049696159252993975487806364305442437946767); DK(KP999544308, +0.999544308746292983948881682379742149196758193); DK(KP772036680, +0.772036680810363904029489473607579825330539880); DK(KP763932022, +0.763932022500210303590826331268723764559381640); DK(KP992114701, +0.992114701314477831049793042785778521453036709); DK(KP447417479, +0.447417479732227551498980015410057305749330693); DK(KP734762448, +0.734762448793050413546343770063151342619912334); DK(KP894834959, +0.894834959464455102997960030820114611498661386); DK(KP867381224, +0.867381224396525206773171885031575671309956167); DK(KP958953096, +0.958953096729998668045963838399037225970891871); DK(KP912575812, +0.912575812670962425556968549836277086778922727); DK(KP951056516, +0.951056516295153572116439333379382143405698634); DK(KP244189809, +0.244189809627953270309879511234821255780225091); DK(KP269969613, +0.269969613759572083574752974412347470060951301); DK(KP522847744, +0.522847744331509716623755382187077770911012542); DK(KP578046249, +0.578046249379945007321754579646815604023525655); DK(KP603558818, +0.603558818296015001454675132653458027918768137); DK(KP667278218, +0.667278218140296670899089292254759909713898805); DK(KP447533225, +0.447533225982656890041886979663652563063114397); DK(KP494780565, +0.494780565770515410344588413655324772219443730); DK(KP987388751, +0.987388751065621252324603216482382109400433949); DK(KP893101515, +0.893101515366181661711202267938416198338079437); DK(KP132830569, +0.132830569247582714407653942074819768844536507); DK(KP120146378, +0.120146378570687701782758537356596213647956445); DK(KP059835404, +0.059835404262124915169548397419498386427871950); DK(KP066152395, +0.066152395967733048213034281011006031460903353); DK(KP786782374, +0.786782374965295178365099601674911834788448471); DK(KP869845200, +0.869845200362138853122720822420327157933056305); DK(KP559016994, +0.559016994374947424102293417182819058860154590); DK(KP250000000, +0.250000000000000000000000000000000000000000000); DK(KP618033988, +0.618033988749894848204586834365638117720309180); { INT i; for (i = v; i > 0; i = i - 1, R0 = R0 + ivs, R1 = R1 + ivs, Cr = Cr + ovs, Ci = Ci + ovs, MAKE_VOLATILE_STRIDE(100, rs), MAKE_VOLATILE_STRIDE(100, csr), MAKE_VOLATILE_STRIDE(100, csi)) { E T2H, T2w, T2x, T2A, T2C, T2v, T2M, T2y, T2B, T2N; { E T2u, TJ, T1O, T39, T2t, TB, T21, T1M, T2e, T26, T1B, T1r, T1k, T1c, T9; E T1X, T1R, T2k, T29, T1z, T1v, T1h, TX, Ti, T13, T2a, T2j, T1U, T1Y, TQ; E T1g, T1u, T1y, T12, Ts, T11, T1I; { E Tt, Tw, T16, Tx, Ty; { E T2p, TG, TH, TD, TE, TI, T2r; T2p = R0[0]; TG = R0[WS(rs, 5)]; TH = R1[WS(rs, 7)]; TD = R1[WS(rs, 2)]; TE = R0[WS(rs, 10)]; Tt = R1[WS(rs, 1)]; TI = TG - TH; T2r = TG + TH; { E TF, T2q, Tu, Tv, T2s; TF = TD - TE; T2q = TD + TE; Tu = R0[WS(rs, 4)]; Tv = R1[WS(rs, 11)]; T2u = T2q - T2r; T2s = T2q + T2r; TJ = FMA(KP618033988, TI, TF); T1O = FNMS(KP618033988, TF, TI); T39 = T2p + T2s; T2t = FNMS(KP250000000, T2s, T2p); Tw = Tu + Tv; T16 = Tv - Tu; Tx = R1[WS(rs, 6)]; Ty = R0[WS(rs, 9)]; } } { E T1P, TW, TS, TR; { E T1, T5, T1L, T18, T1a, TA, T4, TU, T6, T19; T1 = R0[WS(rs, 2)]; { E T2, T17, Tz, T3; T2 = R1[WS(rs, 4)]; T17 = Tx - Ty; Tz = Tx + Ty; T3 = R0[WS(rs, 12)]; T5 = R0[WS(rs, 7)]; T1L = FMA(KP618033988, T16, T17); T18 = FNMS(KP618033988, T17, T16); T1a = Tz - Tw; TA = Tw + Tz; T4 = T2 + T3; TU = T3 - T2; T6 = R1[WS(rs, 9)]; } TB = Tt + TA; T19 = FNMS(KP250000000, TA, Tt); { E T7, TV, T1b, T1K, T8; T7 = T5 + T6; TV = T5 - T6; T1b = FNMS(KP559016994, T1a, T19); T1K = FMA(KP559016994, T1a, T19); T1P = FMA(KP618033988, TU, TV); TW = FNMS(KP618033988, TV, TU); TS = T4 - T7; T8 = T4 + T7; T21 = FMA(KP869845200, T1K, T1L); T1M = FNMS(KP786782374, T1L, T1K); T2e = FMA(KP066152395, T1K, T1L); T26 = FNMS(KP059835404, T1L, T1K); T1B = FMA(KP120146378, T18, T1b); T1r = FNMS(KP132830569, T1b, T18); T1k = FMA(KP893101515, T18, T1b); T1c = FNMS(KP987388751, T1b, T18); T9 = T1 + T8; TR = FMS(KP250000000, T8, T1); } } { E Ta, Te, TK, Td, Tf; Ta = R1[0]; { E Tb, Tc, T1Q, TT; Tb = R0[WS(rs, 3)]; Tc = R1[WS(rs, 10)]; T1Q = FMA(KP559016994, TS, TR); TT = FNMS(KP559016994, TS, TR); Te = R1[WS(rs, 5)]; TK = Tb - Tc; Td = Tb + Tc; T1X = FNMS(KP120146378, T1P, T1Q); T1R = FMA(KP132830569, T1Q, T1P); T2k = FMA(KP494780565, T1Q, T1P); T29 = FNMS(KP447533225, T1P, T1Q); T1z = FMA(KP869845200, TT, TW); T1v = FNMS(KP786782374, TW, TT); T1h = FNMS(KP667278218, TT, TW); TX = FMA(KP603558818, TW, TT); Tf = R0[WS(rs, 8)]; } { E Tk, T1S, TM, TO, Tn, TZ, TN, T10, Tq, To, Th, Tp, TP, T1T, Tr; Tk = R0[WS(rs, 1)]; { E Tl, TL, Tg, Tm; Tl = R1[WS(rs, 3)]; TL = Tf - Te; Tg = Te + Tf; Tm = R0[WS(rs, 11)]; To = R0[WS(rs, 6)]; T1S = FMA(KP618033988, TK, TL); TM = FNMS(KP618033988, TL, TK); TO = Td - Tg; Th = Td + Tg; Tn = Tl + Tm; TZ = Tm - Tl; Tp = R1[WS(rs, 8)]; } Ti = Ta + Th; TN = FNMS(KP250000000, Th, Ta); T10 = Tp - To; Tq = To + Tp; TP = FMA(KP559016994, TO, TN); T1T = FNMS(KP559016994, TO, TN); Tr = Tn + Tq; T13 = Tn - Tq; T2a = FMA(KP578046249, T1T, T1S); T2j = FNMS(KP522847744, T1S, T1T); T1U = FNMS(KP987388751, T1T, T1S); T1Y = FMA(KP893101515, T1S, T1T); TQ = FMA(KP269969613, TP, TM); T1g = FNMS(KP244189809, TM, TP); T1u = FNMS(KP603558818, TM, TP); T1y = FMA(KP667278218, TP, TM); T12 = FMS(KP250000000, Tr, Tk); Ts = Tk + Tr; T11 = FMA(KP618033988, T10, TZ); T1I = FNMS(KP618033988, TZ, T10); } } } } { E T2f, T27, T1j, T15, T2K, T2J, T2I, T2T, T1Z, T2X, T1N, T1V, T2W, T2U, T22; E T1G; { E T3a, T3b, T20, T1J, T1C, T1s; { E Tj, TC, T1H, T14; T3a = T9 + Ti; Tj = T9 - Ti; TC = Ts - TB; T3b = Ts + TB; T1H = FMA(KP559016994, T13, T12); T14 = FNMS(KP559016994, T13, T12); Ci[WS(csi, 10)] = KP951056516 * (FMA(KP618033988, Tj, TC)); Ci[WS(csi, 5)] = KP951056516 * (FNMS(KP618033988, TC, Tj)); T20 = FNMS(KP066152395, T1H, T1I); T1J = FMA(KP059835404, T1I, T1H); T2f = FMA(KP667278218, T1H, T1I); T27 = FNMS(KP603558818, T1I, T1H); T1C = FNMS(KP494780565, T14, T11); T1s = FMA(KP447533225, T11, T14); T1j = FNMS(KP522847744, T11, T14); T15 = FMA(KP578046249, T14, T11); } { E T1A, T1t, T1w, T3c, T3e, T1D, T1x, T3d, T1E, T1F; T1A = FNMS(KP912575812, T1z, T1y); T2K = FMA(KP912575812, T1z, T1y); T2J = FNMS(KP958953096, T1s, T1r); T1t = FMA(KP958953096, T1s, T1r); T1w = FMA(KP912575812, T1v, T1u); T2H = FNMS(KP912575812, T1v, T1u); T3c = T3a + T3b; T3e = T3a - T3b; T2I = FMA(KP867381224, T1C, T1B); T1D = FNMS(KP867381224, T1C, T1B); T1x = FNMS(KP894834959, T1w, T1t); T2T = FMA(KP734762448, T1Y, T1X); T1Z = FNMS(KP734762448, T1Y, T1X); T3d = FNMS(KP250000000, T3c, T39); Cr[0] = T3c + T39; T1E = FMA(KP447417479, T1w, T1D); Ci[WS(csi, 4)] = KP951056516 * (FMA(KP992114701, T1x, TJ)); Cr[WS(csr, 10)] = FNMS(KP559016994, T3e, T3d); Cr[WS(csr, 5)] = FMA(KP559016994, T3e, T3d); T1F = FMA(KP763932022, T1E, T1t); T2X = FMA(KP772036680, T1M, T1J); T1N = FNMS(KP772036680, T1M, T1J); T1V = FMA(KP734762448, T1U, T1R); T2W = FNMS(KP734762448, T1U, T1R); T2U = FNMS(KP772036680, T21, T20); T22 = FMA(KP772036680, T21, T20); T1G = FMA(KP999544308, T1F, T1A); } } { E T1i, T1l, T2l, T2R, T2g, T2Q, T28, T32, T1f, T1n, T1p, T33, T2b; { E T24, TY, T1d, T1W, T23, T25, T1m, T1e; T2w = FMA(KP829049696, T1h, T1g); T1i = FNMS(KP829049696, T1h, T1g); T1W = FNMS(KP992114701, T1V, T1O); T23 = FNMS(KP522616830, T1V, T22); Ci[WS(csi, 9)] = KP951056516 * (FNMS(KP803003575, T1G, TJ)); T2x = FNMS(KP831864738, T1k, T1j); T1l = FMA(KP831864738, T1k, T1j); Ci[WS(csi, 3)] = KP998026728 * (FNMS(KP952936919, T1W, T1N)); T24 = FMA(KP690983005, T23, T1N); TY = FNMS(KP916574801, TX, TQ); T2A = FMA(KP916574801, TX, TQ); T2C = FNMS(KP831864738, T1c, T15); T1d = FMA(KP831864738, T1c, T15); T2l = FNMS(KP982009705, T2k, T2j); T2R = FMA(KP982009705, T2k, T2j); T25 = FNMS(KP855719849, T24, T1Z); T2g = FMA(KP845997307, T2f, T2e); T2Q = FNMS(KP845997307, T2f, T2e); T1m = FMA(KP904730450, T1d, TY); T1e = FNMS(KP904730450, T1d, TY); Ci[WS(csi, 8)] = -(KP951056516 * (FNMS(KP992114701, T25, T1O))); T28 = FNMS(KP845997307, T27, T26); T32 = FMA(KP845997307, T27, T26); T1f = FNMS(KP242145790, T1e, TJ); Ci[WS(csi, 1)] = -(KP951056516 * (FMA(KP968583161, T1e, TJ))); T1n = FNMS(KP999754674, T1m, T1l); T1p = FNMS(KP904508497, T1m, T1i); T33 = FMA(KP921078979, T2a, T29); T2b = FNMS(KP921078979, T2a, T29); } { E T2P, T2Z, T2V, T2O; { E T2d, T2n, T2i, T2Y, T2m, T2o; T2P = FNMS(KP559016994, T2u, T2t); T2v = FMA(KP559016994, T2u, T2t); { E T1o, T1q, T2h, T2c; T1o = FNMS(KP559154169, T1n, T1i); T1q = FMA(KP683113946, T1p, T1l); T2h = FMA(KP906616052, T2b, T28); T2c = FNMS(KP906616052, T2b, T28); Ci[WS(csi, 6)] = -(KP951056516 * (FMA(KP968583161, T1o, T1f))); Ci[WS(csi, 11)] = -(KP951056516 * (FMA(KP876306680, T1q, T1f))); T2d = FMA(KP262346850, T2c, T1O); Ci[WS(csi, 2)] = -(KP998026728 * (FNMS(KP952936919, T1O, T2c))); T2n = T2g + T2h; T2i = FMA(KP618033988, T2h, T2g); } T2m = FMA(KP570584518, T2l, T2i); T2o = FNMS(KP669429328, T2n, T2l); Ci[WS(csi, 12)] = KP951056516 * (FNMS(KP949179823, T2m, T2d)); Ci[WS(csi, 7)] = KP951056516 * (FNMS(KP876306680, T2o, T2d)); T2V = FMA(KP956723877, T2U, T2T); T2Y = FMA(KP522616830, T2T, T2X); T2Z = FNMS(KP763932022, T2Y, T2U); } Cr[WS(csr, 3)] = FMA(KP992114701, T2V, T2P); { E T30, T34, T2S, T31, T35; T30 = FMA(KP855719849, T2Z, T2W); T34 = FNMS(KP923225144, T2R, T2Q); T2S = FMA(KP923225144, T2R, T2Q); Cr[WS(csr, 8)] = FNMS(KP897376177, T30, T2P); T31 = FNMS(KP237294955, T2S, T2P); Cr[WS(csr, 2)] = FMA(KP949179823, T2S, T2P); T35 = FNMS(KP997675361, T34, T33); { E T37, T36, T38, T2L; T37 = FNMS(KP904508497, T34, T32); T36 = FMA(KP560319534, T35, T32); T38 = FNMS(KP681693190, T37, T33); Cr[WS(csr, 12)] = FNMS(KP949179823, T36, T31); Cr[WS(csr, 7)] = FNMS(KP860541664, T38, T31); T2O = FNMS(KP809385824, T2K, T2I); T2L = FNMS(KP447417479, T2K, T2J); T2M = FNMS(KP690983005, T2L, T2I); } } Cr[WS(csr, 4)] = FNMS(KP992114701, T2O, T2v); } } } } T2y = FNMS(KP904730450, T2x, T2w); T2B = FMA(KP904730450, T2x, T2w); T2N = FNMS(KP999544308, T2M, T2H); { E T2z, T2D, T2F, T2E, T2G; T2z = FNMS(KP242145790, T2y, T2v); Cr[WS(csr, 1)] = FMA(KP968583161, T2y, T2v); T2D = FMA(KP904730450, T2C, T2B); T2F = T2A + T2B; Cr[WS(csr, 9)] = FNMS(KP803003575, T2N, T2v); T2E = FNMS(KP618033988, T2D, T2A); T2G = FMA(KP683113946, T2F, T2C); Cr[WS(csr, 6)] = FNMS(KP876091699, T2E, T2z); Cr[WS(csr, 11)] = FNMS(KP792626838, T2G, T2z); } } } } static const kr2c_desc desc = { 25, "r2cf_25", {44, 12, 156, 0}, &GENUS }; void X(codelet_r2cf_25) (planner *p) { X(kr2c_register) (p, r2cf_25, &desc); } #else /* HAVE_FMA */ /* Generated by: ../../../genfft/gen_r2cf.native -compact -variables 4 -pipeline-latency 4 -n 25 -name r2cf_25 -include r2cf.h */ /* * This function contains 200 FP additions, 140 FP multiplications, * (or, 117 additions, 57 multiplications, 83 fused multiply/add), * 101 stack variables, 40 constants, and 50 memory accesses */ #include "r2cf.h" static void r2cf_25(R *R0, R *R1, R *Cr, R *Ci, stride rs, stride csr, stride csi, INT v, INT ivs, INT ovs) { DK(KP998026728, +0.998026728428271561952336806863450553336905220); DK(KP125581039, +0.125581039058626752152356449131262266244969664); DK(KP1_996053456, +1.996053456856543123904673613726901106673810439); DK(KP062790519, +0.062790519529313376076178224565631133122484832); DK(KP809016994, +0.809016994374947424102293417182819058860154590); DK(KP309016994, +0.309016994374947424102293417182819058860154590); DK(KP1_369094211, +1.369094211857377347464566715242418539779038465); DK(KP728968627, +0.728968627421411523146730319055259111372571664); DK(KP963507348, +0.963507348203430549974383005744259307057084020); DK(KP876306680, +0.876306680043863587308115903922062583399064238); DK(KP497379774, +0.497379774329709576484567492012895936835134813); DK(KP968583161, +0.968583161128631119490168375464735813836012403); DK(KP684547105, +0.684547105928688673732283357621209269889519233); DK(KP1_457937254, +1.457937254842823046293460638110518222745143328); DK(KP481753674, +0.481753674101715274987191502872129653528542010); DK(KP1_752613360, +1.752613360087727174616231807844125166798128477); DK(KP248689887, +0.248689887164854788242283746006447968417567406); DK(KP1_937166322, +1.937166322257262238980336750929471627672024806); DK(KP992114701, +0.992114701314477831049793042785778521453036709); DK(KP250666467, +0.250666467128608490746237519633017587885836494); DK(KP425779291, +0.425779291565072648862502445744251703979973042); DK(KP1_809654104, +1.809654104932039055427337295865395187940827822); DK(KP1_274847979, +1.274847979497379420353425623352032390869834596); DK(KP770513242, +0.770513242775789230803009636396177847271667672); DK(KP844327925, +0.844327925502015078548558063966681505381659241); DK(KP1_071653589, +1.071653589957993236542617535735279956127150691); DK(KP125333233, +0.125333233564304245373118759816508793942918247); DK(KP1_984229402, +1.984229402628955662099586085571557042906073418); DK(KP904827052, +0.904827052466019527713668647932697593970413911); DK(KP851558583, +0.851558583130145297725004891488503407959946084); DK(KP637423989, +0.637423989748689710176712811676016195434917298); DK(KP1_541026485, +1.541026485551578461606019272792355694543335344); DK(KP535826794, +0.535826794978996618271308767867639978063575346); DK(KP1_688655851, +1.688655851004030157097116127933363010763318483); DK(KP293892626, +0.293892626146236564584352977319536384298826219); DK(KP475528258, +0.475528258147576786058219666689691071702849317); DK(KP250000000, +0.250000000000000000000000000000000000000000000); DK(KP559016994, +0.559016994374947424102293417182819058860154590); DK(KP587785252, +0.587785252292473129168705954639072768597652438); DK(KP951056516, +0.951056516295153572116439333379382143405698634); { INT i; for (i = v; i > 0; i = i - 1, R0 = R0 + ivs, R1 = R1 + ivs, Cr = Cr + ovs, Ci = Ci + ovs, MAKE_VOLATILE_STRIDE(100, rs), MAKE_VOLATILE_STRIDE(100, csr), MAKE_VOLATILE_STRIDE(100, csi)) { E T8, T1j, T1V, T1l, T7, T9, Ta, T12, T2u, T1O, T19, T1P, Ti, T2r, T1K; E Tp, T1L, Tx, T2q, T1H, TE, T1I, TN, T2t, T1R, TU, T1S, T6, T1k, T3; E T2s, T2v; T8 = R0[0]; { E T4, T5, T1, T2; T4 = R0[WS(rs, 5)]; T5 = R1[WS(rs, 7)]; T6 = T4 + T5; T1k = T4 - T5; T1 = R1[WS(rs, 2)]; T2 = R0[WS(rs, 10)]; T3 = T1 + T2; T1j = T1 - T2; } T1V = KP951056516 * T1k; T1l = FMA(KP951056516, T1j, KP587785252 * T1k); T7 = KP559016994 * (T3 - T6); T9 = T3 + T6; Ta = FNMS(KP250000000, T9, T8); { E T16, T13, T14, TY, T17, T11, T15, T18; T16 = R1[WS(rs, 1)]; { E TW, TX, TZ, T10; TW = R0[WS(rs, 4)]; TX = R1[WS(rs, 11)]; T13 = TW + TX; TZ = R1[WS(rs, 6)]; T10 = R0[WS(rs, 9)]; T14 = TZ + T10; TY = TW - TX; T17 = T13 + T14; T11 = TZ - T10; } T12 = FMA(KP475528258, TY, KP293892626 * T11); T2u = T16 + T17; T1O = FNMS(KP293892626, TY, KP475528258 * T11); T15 = KP559016994 * (T13 - T14); T18 = FNMS(KP250000000, T17, T16); T19 = T15 + T18; T1P = T18 - T15; } { E Tm, Tj, Tk, Te, Tn, Th, Tl, To; Tm = R1[0]; { E Tc, Td, Tf, Tg; Tc = R0[WS(rs, 3)]; Td = R1[WS(rs, 10)]; Tj = Tc + Td; Tf = R1[WS(rs, 5)]; Tg = R0[WS(rs, 8)]; Tk = Tf + Tg; Te = Tc - Td; Tn = Tj + Tk; Th = Tf - Tg; } Ti = FMA(KP475528258, Te, KP293892626 * Th); T2r = Tm + Tn; T1K = FNMS(KP293892626, Te, KP475528258 * Th); Tl = KP559016994 * (Tj - Tk); To = FNMS(KP250000000, Tn, Tm); Tp = Tl + To; T1L = To - Tl; } { E TB, Ty, Tz, Tt, TC, Tw, TA, TD; TB = R0[WS(rs, 2)]; { E Tr, Ts, Tu, Tv; Tr = R1[WS(rs, 4)]; Ts = R0[WS(rs, 12)]; Ty = Tr + Ts; Tu = R0[WS(rs, 7)]; Tv = R1[WS(rs, 9)]; Tz = Tu + Tv; Tt = Tr - Ts; TC = Ty + Tz; Tw = Tu - Tv; } Tx = FMA(KP475528258, Tt, KP293892626 * Tw); T2q = TB + TC; T1H = FNMS(KP293892626, Tt, KP475528258 * Tw); TA = KP559016994 * (Ty - Tz); TD = FNMS(KP250000000, TC, TB); TE = TA + TD; T1I = TD - TA; } { E TR, TO, TP, TJ, TS, TM, TQ, TT; TR = R0[WS(rs, 1)]; { E TH, TI, TK, TL; TH = R1[WS(rs, 3)]; TI = R0[WS(rs, 11)]; TO = TH + TI; TK = R0[WS(rs, 6)]; TL = R1[WS(rs, 8)]; TP = TK + TL; TJ = TH - TI; TS = TO + TP; TM = TK - TL; } TN = FMA(KP475528258, TJ, KP293892626 * TM); T2t = TR + TS; T1R = FNMS(KP293892626, TJ, KP475528258 * TM); TQ = KP559016994 * (TO - TP); TT = FNMS(KP250000000, TS, TR); TU = TQ + TT; T1S = TT - TQ; } T2s = T2q - T2r; T2v = T2t - T2u; Ci[WS(csi, 5)] = FNMS(KP587785252, T2v, KP951056516 * T2s); Ci[WS(csi, 10)] = FMA(KP587785252, T2s, KP951056516 * T2v); { E T2z, T2y, T2A, T2w, T2x, T2B; T2z = T8 + T9; T2w = T2r + T2q; T2x = T2t + T2u; T2y = KP559016994 * (T2w - T2x); T2A = T2w + T2x; Cr[0] = T2z + T2A; T2B = FNMS(KP250000000, T2A, T2z); Cr[WS(csr, 5)] = T2y + T2B; Cr[WS(csr, 10)] = T2B - T2y; } { E Tb, Tq, TF, TG, T1E, T1F, T1G, T1B, T1C, T1D, TV, T1a, T1b, T1o, T1r; E T1s, T1z, T1x, T1e, T1h, T1i, T1u, T1t; Tb = T7 + Ta; Tq = FMA(KP1_688655851, Ti, KP535826794 * Tp); TF = FMA(KP1_541026485, Tx, KP637423989 * TE); TG = Tq - TF; T1E = FMA(KP851558583, TN, KP904827052 * TU); T1F = FMA(KP1_984229402, T12, KP125333233 * T19); T1G = T1E + T1F; T1B = FNMS(KP844327925, Tp, KP1_071653589 * Ti); T1C = FNMS(KP1_274847979, Tx, KP770513242 * TE); T1D = T1B + T1C; TV = FNMS(KP425779291, TU, KP1_809654104 * TN); T1a = FNMS(KP992114701, T19, KP250666467 * T12); T1b = TV + T1a; { E T1m, T1n, T1p, T1q; T1m = FMA(KP1_937166322, Ti, KP248689887 * Tp); T1n = FMA(KP1_071653589, Tx, KP844327925 * TE); T1o = T1m + T1n; T1p = FMA(KP1_752613360, TN, KP481753674 * TU); T1q = FMA(KP1_457937254, T12, KP684547105 * T19); T1r = T1p + T1q; T1s = T1o + T1r; T1z = T1q - T1p; T1x = T1n - T1m; } { E T1c, T1d, T1f, T1g; T1c = FNMS(KP497379774, Ti, KP968583161 * Tp); T1d = FNMS(KP1_688655851, Tx, KP535826794 * TE); T1e = T1c + T1d; T1f = FNMS(KP963507348, TN, KP876306680 * TU); T1g = FNMS(KP1_369094211, T12, KP728968627 * T19); T1h = T1f + T1g; T1i = T1e + T1h; T1u = T1f - T1g; T1t = T1d - T1c; } Cr[WS(csr, 1)] = Tb + T1i; Ci[WS(csi, 1)] = -(T1l + T1s); Cr[WS(csr, 4)] = Tb + TG + T1b; Ci[WS(csi, 4)] = T1l + T1D - T1G; Ci[WS(csi, 9)] = FMA(KP309016994, T1D, T1l) + FMA(KP587785252, T1a - TV, KP809016994 * T1G) - (KP951056516 * (Tq + TF)); Cr[WS(csr, 9)] = FMA(KP309016994, TG, Tb) + FMA(KP951056516, T1B - T1C, KP587785252 * (T1F - T1E)) - (KP809016994 * T1b); { E T1v, T1w, T1y, T1A; T1v = FMS(KP250000000, T1s, T1l); T1w = KP559016994 * (T1r - T1o); Ci[WS(csi, 11)] = FMA(KP587785252, T1t, KP951056516 * T1u) + T1v - T1w; Ci[WS(csi, 6)] = FMA(KP951056516, T1t, T1v) + FNMS(KP587785252, T1u, T1w); T1y = FNMS(KP250000000, T1i, Tb); T1A = KP559016994 * (T1e - T1h); Cr[WS(csr, 11)] = FMA(KP587785252, T1x, T1y) + FNMA(KP951056516, T1z, T1A); Cr[WS(csr, 6)] = FMA(KP951056516, T1x, T1A) + FMA(KP587785252, T1z, T1y); } } { E T1W, T1X, T1J, T1M, T1N, T21, T22, T23, T1Q, T1T, T1U, T1Y, T1Z, T20, T26; E T29, T2a, T2k, T2j, T2l, T2m, T2d, T2o, T2i; T1W = FNMS(KP587785252, T1j, T1V); T1X = Ta - T7; T1J = FNMS(KP125333233, T1I, KP1_984229402 * T1H); T1M = FMA(KP1_457937254, T1K, KP684547105 * T1L); T1N = T1J - T1M; T21 = FNMS(KP1_996053456, T1R, KP062790519 * T1S); T22 = FMA(KP1_541026485, T1O, KP637423989 * T1P); T23 = T21 - T22; T1Q = FNMS(KP770513242, T1P, KP1_274847979 * T1O); T1T = FMA(KP125581039, T1R, KP998026728 * T1S); T1U = T1Q - T1T; T1Y = FNMS(KP1_369094211, T1K, KP728968627 * T1L); T1Z = FMA(KP250666467, T1H, KP992114701 * T1I); T20 = T1Y - T1Z; { E T24, T25, T27, T28; T24 = FNMS(KP481753674, T1L, KP1_752613360 * T1K); T25 = FMA(KP851558583, T1H, KP904827052 * T1I); T26 = T24 - T25; T27 = FNMS(KP844327925, T1S, KP1_071653589 * T1R); T28 = FNMS(KP998026728, T1P, KP125581039 * T1O); T29 = T27 + T28; T2a = T26 + T29; T2k = T27 - T28; T2j = T24 + T25; } { E T2b, T2c, T2g, T2h; T2b = FNMS(KP425779291, T1I, KP1_809654104 * T1H); T2c = FMA(KP963507348, T1K, KP876306680 * T1L); T2l = T2c + T2b; T2g = FMA(KP1_688655851, T1R, KP535826794 * T1S); T2h = FMA(KP1_996053456, T1O, KP062790519 * T1P); T2m = T2g + T2h; T2d = T2b - T2c; T2o = T2l + T2m; T2i = T2g - T2h; } Ci[WS(csi, 2)] = T1W + T2a; Cr[WS(csr, 2)] = T1X + T2o; Ci[WS(csi, 3)] = T1N + T1U - T1W; Cr[WS(csr, 3)] = T1X + T20 + T23; Cr[WS(csr, 8)] = FMA(KP309016994, T20, T1X) + FNMA(KP809016994, T23, KP587785252 * (T1T + T1Q)) - (KP951056516 * (T1M + T1J)); Ci[WS(csi, 8)] = FNMS(KP587785252, T21 + T22, KP309016994 * T1N) + FNMA(KP809016994, T1U, KP951056516 * (T1Y + T1Z)) - T1W; { E T2e, T2f, T2n, T2p; T2e = KP559016994 * (T26 - T29); T2f = FNMS(KP250000000, T2a, T1W); Ci[WS(csi, 7)] = FMA(KP951056516, T2d, T2e) + FNMS(KP587785252, T2i, T2f); Ci[WS(csi, 12)] = FMA(KP587785252, T2d, T2f) + FMS(KP951056516, T2i, T2e); T2n = KP559016994 * (T2l - T2m); T2p = FNMS(KP250000000, T2o, T1X); Cr[WS(csr, 7)] = FMA(KP951056516, T2j, KP587785252 * T2k) + T2n + T2p; Cr[WS(csr, 12)] = FMA(KP587785252, T2j, T2p) + FNMA(KP951056516, T2k, T2n); } } } } } static const kr2c_desc desc = { 25, "r2cf_25", {117, 57, 83, 0}, &GENUS }; void X(codelet_r2cf_25) (planner *p) { X(kr2c_register) (p, r2cf_25, &desc); } #endif /* HAVE_FMA */ fftw-3.3.3/rdft/scalar/r2cf/hf_2.c0000644000175000017500000000671212054410625013436 00000000000000/* * Copyright (c) 2003, 2007-11 Matteo Frigo * Copyright (c) 2003, 2007-11 Massachusetts Institute of Technology * * This program is free software; you can redistribute it and/or modify * it under the terms of the GNU General Public License as published by * the Free Software Foundation; either version 2 of the License, or * (at your option) any later version. * * This program is distributed in the hope that it will be useful, * but WITHOUT ANY WARRANTY; without even the implied warranty of * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the * GNU General Public License for more details. * * You should have received a copy of the GNU General Public License * along with this program; if not, write to the Free Software * Foundation, Inc., 51 Franklin Street, Fifth Floor, Boston, MA 02110-1301 USA * */ /* This file was automatically generated --- DO NOT EDIT */ /* Generated on Sun Nov 25 07:39:49 EST 2012 */ #include "codelet-rdft.h" #ifdef HAVE_FMA /* Generated by: ../../../genfft/gen_hc2hc.native -fma -reorder-insns -schedule-for-pipeline -compact -variables 4 -pipeline-latency 4 -n 2 -dit -name hf_2 -include hf.h */ /* * This function contains 6 FP additions, 4 FP multiplications, * (or, 4 additions, 2 multiplications, 2 fused multiply/add), * 11 stack variables, 0 constants, and 8 memory accesses */ #include "hf.h" static void hf_2(R *cr, R *ci, const R *W, stride rs, INT mb, INT me, INT ms) { { INT m; for (m = mb, W = W + ((mb - 1) * 2); m < me; m = m + 1, cr = cr + ms, ci = ci - ms, W = W + 2, MAKE_VOLATILE_STRIDE(4, rs)) { E T1, Ta, T3, T6, T2, T5; T1 = cr[0]; Ta = ci[0]; T3 = cr[WS(rs, 1)]; T6 = ci[WS(rs, 1)]; T2 = W[0]; T5 = W[1]; { E T8, T4, T9, T7; T8 = T2 * T6; T4 = T2 * T3; T9 = FNMS(T5, T3, T8); T7 = FMA(T5, T6, T4); ci[WS(rs, 1)] = T9 + Ta; cr[WS(rs, 1)] = T9 - Ta; cr[0] = T1 + T7; ci[0] = T1 - T7; } } } } static const tw_instr twinstr[] = { {TW_FULL, 1, 2}, {TW_NEXT, 1, 0} }; static const hc2hc_desc desc = { 2, "hf_2", twinstr, &GENUS, {4, 2, 2, 0} }; void X(codelet_hf_2) (planner *p) { X(khc2hc_register) (p, hf_2, &desc); } #else /* HAVE_FMA */ /* Generated by: ../../../genfft/gen_hc2hc.native -compact -variables 4 -pipeline-latency 4 -n 2 -dit -name hf_2 -include hf.h */ /* * This function contains 6 FP additions, 4 FP multiplications, * (or, 4 additions, 2 multiplications, 2 fused multiply/add), * 9 stack variables, 0 constants, and 8 memory accesses */ #include "hf.h" static void hf_2(R *cr, R *ci, const R *W, stride rs, INT mb, INT me, INT ms) { { INT m; for (m = mb, W = W + ((mb - 1) * 2); m < me; m = m + 1, cr = cr + ms, ci = ci - ms, W = W + 2, MAKE_VOLATILE_STRIDE(4, rs)) { E T1, T8, T6, T7; T1 = cr[0]; T8 = ci[0]; { E T3, T5, T2, T4; T3 = cr[WS(rs, 1)]; T5 = ci[WS(rs, 1)]; T2 = W[0]; T4 = W[1]; T6 = FMA(T2, T3, T4 * T5); T7 = FNMS(T4, T3, T2 * T5); } ci[0] = T1 - T6; cr[0] = T1 + T6; cr[WS(rs, 1)] = T7 - T8; ci[WS(rs, 1)] = T7 + T8; } } } static const tw_instr twinstr[] = { {TW_FULL, 1, 2}, {TW_NEXT, 1, 0} }; static const hc2hc_desc desc = { 2, "hf_2", twinstr, &GENUS, {4, 2, 2, 0} }; void X(codelet_hf_2) (planner *p) { X(khc2hc_register) (p, hf_2, &desc); } #endif /* HAVE_FMA */ fftw-3.3.3/rdft/scalar/r2cf/hf_3.c0000644000175000017500000001142212054410626013432 00000000000000/* * Copyright (c) 2003, 2007-11 Matteo Frigo * Copyright (c) 2003, 2007-11 Massachusetts Institute of Technology * * This program is free software; you can redistribute it and/or modify * it under the terms of the GNU General Public License as published by * the Free Software Foundation; either version 2 of the License, or * (at your option) any later version. * * This program is distributed in the hope that it will be useful, * but WITHOUT ANY WARRANTY; without even the implied warranty of * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the * GNU General Public License for more details. * * You should have received a copy of the GNU General Public License * along with this program; if not, write to the Free Software * Foundation, Inc., 51 Franklin Street, Fifth Floor, Boston, MA 02110-1301 USA * */ /* This file was automatically generated --- DO NOT EDIT */ /* Generated on Sun Nov 25 07:39:49 EST 2012 */ #include "codelet-rdft.h" #ifdef HAVE_FMA /* Generated by: ../../../genfft/gen_hc2hc.native -fma -reorder-insns -schedule-for-pipeline -compact -variables 4 -pipeline-latency 4 -n 3 -dit -name hf_3 -include hf.h */ /* * This function contains 16 FP additions, 14 FP multiplications, * (or, 6 additions, 4 multiplications, 10 fused multiply/add), * 21 stack variables, 2 constants, and 12 memory accesses */ #include "hf.h" static void hf_3(R *cr, R *ci, const R *W, stride rs, INT mb, INT me, INT ms) { DK(KP866025403, +0.866025403784438646763723170752936183471402627); DK(KP500000000, +0.500000000000000000000000000000000000000000000); { INT m; for (m = mb, W = W + ((mb - 1) * 4); m < me; m = m + 1, cr = cr + ms, ci = ci - ms, W = W + 4, MAKE_VOLATILE_STRIDE(6, rs)) { E T1, Tl, T9, Tc, Tb, Th, T7, Ti, Ta, Tj, Td; T1 = cr[0]; Tl = ci[0]; { E T3, T6, T2, T5, Tg, T4, T8; T3 = cr[WS(rs, 1)]; T6 = ci[WS(rs, 1)]; T2 = W[0]; T5 = W[1]; T9 = cr[WS(rs, 2)]; Tc = ci[WS(rs, 2)]; Tg = T2 * T6; T4 = T2 * T3; T8 = W[2]; Tb = W[3]; Th = FNMS(T5, T3, Tg); T7 = FMA(T5, T6, T4); Ti = T8 * Tc; Ta = T8 * T9; } Tj = FNMS(Tb, T9, Ti); Td = FMA(Tb, Tc, Ta); { E Tk, Te, To, Tn, Tm, Tf; Tk = Th - Tj; Tm = Th + Tj; Te = T7 + Td; To = Td - T7; ci[WS(rs, 2)] = Tm + Tl; Tn = FNMS(KP500000000, Tm, Tl); cr[0] = T1 + Te; Tf = FNMS(KP500000000, Te, T1); ci[WS(rs, 1)] = FMA(KP866025403, To, Tn); cr[WS(rs, 2)] = FMS(KP866025403, To, Tn); cr[WS(rs, 1)] = FMA(KP866025403, Tk, Tf); ci[0] = FNMS(KP866025403, Tk, Tf); } } } } static const tw_instr twinstr[] = { {TW_FULL, 1, 3}, {TW_NEXT, 1, 0} }; static const hc2hc_desc desc = { 3, "hf_3", twinstr, &GENUS, {6, 4, 10, 0} }; void X(codelet_hf_3) (planner *p) { X(khc2hc_register) (p, hf_3, &desc); } #else /* HAVE_FMA */ /* Generated by: ../../../genfft/gen_hc2hc.native -compact -variables 4 -pipeline-latency 4 -n 3 -dit -name hf_3 -include hf.h */ /* * This function contains 16 FP additions, 12 FP multiplications, * (or, 10 additions, 6 multiplications, 6 fused multiply/add), * 15 stack variables, 2 constants, and 12 memory accesses */ #include "hf.h" static void hf_3(R *cr, R *ci, const R *W, stride rs, INT mb, INT me, INT ms) { DK(KP866025403, +0.866025403784438646763723170752936183471402627); DK(KP500000000, +0.500000000000000000000000000000000000000000000); { INT m; for (m = mb, W = W + ((mb - 1) * 4); m < me; m = m + 1, cr = cr + ms, ci = ci - ms, W = W + 4, MAKE_VOLATILE_STRIDE(6, rs)) { E T1, Ti, T6, Te, Tb, Tf, Tc, Tj; T1 = cr[0]; Ti = ci[0]; { E T3, T5, T2, T4; T3 = cr[WS(rs, 1)]; T5 = ci[WS(rs, 1)]; T2 = W[0]; T4 = W[1]; T6 = FMA(T2, T3, T4 * T5); Te = FNMS(T4, T3, T2 * T5); } { E T8, Ta, T7, T9; T8 = cr[WS(rs, 2)]; Ta = ci[WS(rs, 2)]; T7 = W[2]; T9 = W[3]; Tb = FMA(T7, T8, T9 * Ta); Tf = FNMS(T9, T8, T7 * Ta); } Tc = T6 + Tb; Tj = Te + Tf; { E Td, Tg, Th, Tk; cr[0] = T1 + Tc; Td = FNMS(KP500000000, Tc, T1); Tg = KP866025403 * (Te - Tf); ci[0] = Td - Tg; cr[WS(rs, 1)] = Td + Tg; ci[WS(rs, 2)] = Tj + Ti; Th = KP866025403 * (Tb - T6); Tk = FNMS(KP500000000, Tj, Ti); cr[WS(rs, 2)] = Th - Tk; ci[WS(rs, 1)] = Th + Tk; } } } } static const tw_instr twinstr[] = { {TW_FULL, 1, 3}, {TW_NEXT, 1, 0} }; static const hc2hc_desc desc = { 3, "hf_3", twinstr, &GENUS, {10, 6, 6, 0} }; void X(codelet_hf_3) (planner *p) { X(khc2hc_register) (p, hf_3, &desc); } #endif /* HAVE_FMA */ fftw-3.3.3/rdft/scalar/r2cf/hf_4.c0000644000175000017500000001205212054410626013433 00000000000000/* * Copyright (c) 2003, 2007-11 Matteo Frigo * Copyright (c) 2003, 2007-11 Massachusetts Institute of Technology * * This program is free software; you can redistribute it and/or modify * it under the terms of the GNU General Public License as published by * the Free Software Foundation; either version 2 of the License, or * (at your option) any later version. * * This program is distributed in the hope that it will be useful, * but WITHOUT ANY WARRANTY; without even the implied warranty of * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the * GNU General Public License for more details. * * You should have received a copy of the GNU General Public License * along with this program; if not, write to the Free Software * Foundation, Inc., 51 Franklin Street, Fifth Floor, Boston, MA 02110-1301 USA * */ /* This file was automatically generated --- DO NOT EDIT */ /* Generated on Sun Nov 25 07:39:50 EST 2012 */ #include "codelet-rdft.h" #ifdef HAVE_FMA /* Generated by: ../../../genfft/gen_hc2hc.native -fma -reorder-insns -schedule-for-pipeline -compact -variables 4 -pipeline-latency 4 -n 4 -dit -name hf_4 -include hf.h */ /* * This function contains 22 FP additions, 12 FP multiplications, * (or, 16 additions, 6 multiplications, 6 fused multiply/add), * 31 stack variables, 0 constants, and 16 memory accesses */ #include "hf.h" static void hf_4(R *cr, R *ci, const R *W, stride rs, INT mb, INT me, INT ms) { { INT m; for (m = mb, W = W + ((mb - 1) * 6); m < me; m = m + 1, cr = cr + ms, ci = ci - ms, W = W + 6, MAKE_VOLATILE_STRIDE(8, rs)) { E To, Te, Tm, T8, Ty, Tw, Tq, Tk; { E T1, Tv, Tu, T7, Tg, Tj, Tf, Ti, Tp, Th; T1 = cr[0]; Tv = ci[0]; { E T3, T6, T2, T5; T3 = cr[WS(rs, 2)]; T6 = ci[WS(rs, 2)]; T2 = W[2]; T5 = W[3]; { E Ta, Td, Tc, Tn, Tb, Tt, T4, T9; Ta = cr[WS(rs, 1)]; Td = ci[WS(rs, 1)]; Tt = T2 * T6; T4 = T2 * T3; T9 = W[0]; Tc = W[1]; Tu = FNMS(T5, T3, Tt); T7 = FMA(T5, T6, T4); Tn = T9 * Td; Tb = T9 * Ta; Tg = cr[WS(rs, 3)]; Tj = ci[WS(rs, 3)]; To = FNMS(Tc, Ta, Tn); Te = FMA(Tc, Td, Tb); Tf = W[4]; Ti = W[5]; } } Tm = T1 - T7; T8 = T1 + T7; Tp = Tf * Tj; Th = Tf * Tg; Ty = Tv - Tu; Tw = Tu + Tv; Tq = FNMS(Ti, Tg, Tp); Tk = FMA(Ti, Tj, Th); } { E Tr, Ts, Tl, Tx; Tr = To - Tq; Ts = To + Tq; Tl = Te + Tk; Tx = Tk - Te; ci[WS(rs, 3)] = Ts + Tw; cr[WS(rs, 2)] = Ts - Tw; cr[WS(rs, 1)] = Tm + Tr; ci[0] = Tm - Tr; ci[WS(rs, 2)] = Tx + Ty; cr[WS(rs, 3)] = Tx - Ty; cr[0] = T8 + Tl; ci[WS(rs, 1)] = T8 - Tl; } } } } static const tw_instr twinstr[] = { {TW_FULL, 1, 4}, {TW_NEXT, 1, 0} }; static const hc2hc_desc desc = { 4, "hf_4", twinstr, &GENUS, {16, 6, 6, 0} }; void X(codelet_hf_4) (planner *p) { X(khc2hc_register) (p, hf_4, &desc); } #else /* HAVE_FMA */ /* Generated by: ../../../genfft/gen_hc2hc.native -compact -variables 4 -pipeline-latency 4 -n 4 -dit -name hf_4 -include hf.h */ /* * This function contains 22 FP additions, 12 FP multiplications, * (or, 16 additions, 6 multiplications, 6 fused multiply/add), * 13 stack variables, 0 constants, and 16 memory accesses */ #include "hf.h" static void hf_4(R *cr, R *ci, const R *W, stride rs, INT mb, INT me, INT ms) { { INT m; for (m = mb, W = W + ((mb - 1) * 6); m < me; m = m + 1, cr = cr + ms, ci = ci - ms, W = W + 6, MAKE_VOLATILE_STRIDE(8, rs)) { E T1, Tp, T6, To, Tc, Tk, Th, Tl; T1 = cr[0]; Tp = ci[0]; { E T3, T5, T2, T4; T3 = cr[WS(rs, 2)]; T5 = ci[WS(rs, 2)]; T2 = W[2]; T4 = W[3]; T6 = FMA(T2, T3, T4 * T5); To = FNMS(T4, T3, T2 * T5); } { E T9, Tb, T8, Ta; T9 = cr[WS(rs, 1)]; Tb = ci[WS(rs, 1)]; T8 = W[0]; Ta = W[1]; Tc = FMA(T8, T9, Ta * Tb); Tk = FNMS(Ta, T9, T8 * Tb); } { E Te, Tg, Td, Tf; Te = cr[WS(rs, 3)]; Tg = ci[WS(rs, 3)]; Td = W[4]; Tf = W[5]; Th = FMA(Td, Te, Tf * Tg); Tl = FNMS(Tf, Te, Td * Tg); } { E T7, Ti, Tj, Tm; T7 = T1 + T6; Ti = Tc + Th; ci[WS(rs, 1)] = T7 - Ti; cr[0] = T7 + Ti; Tj = T1 - T6; Tm = Tk - Tl; ci[0] = Tj - Tm; cr[WS(rs, 1)] = Tj + Tm; } { E Tn, Tq, Tr, Ts; Tn = Tk + Tl; Tq = To + Tp; cr[WS(rs, 2)] = Tn - Tq; ci[WS(rs, 3)] = Tn + Tq; Tr = Th - Tc; Ts = Tp - To; cr[WS(rs, 3)] = Tr - Ts; ci[WS(rs, 2)] = Tr + Ts; } } } } static const tw_instr twinstr[] = { {TW_FULL, 1, 4}, {TW_NEXT, 1, 0} }; static const hc2hc_desc desc = { 4, "hf_4", twinstr, &GENUS, {16, 6, 6, 0} }; void X(codelet_hf_4) (planner *p) { X(khc2hc_register) (p, hf_4, &desc); } #endif /* HAVE_FMA */ fftw-3.3.3/rdft/scalar/r2cf/hf_5.c0000644000175000017500000001634012054410626013440 00000000000000/* * Copyright (c) 2003, 2007-11 Matteo Frigo * Copyright (c) 2003, 2007-11 Massachusetts Institute of Technology * * This program is free software; you can redistribute it and/or modify * it under the terms of the GNU General Public License as published by * the Free Software Foundation; either version 2 of the License, or * (at your option) any later version. * * This program is distributed in the hope that it will be useful, * but WITHOUT ANY WARRANTY; without even the implied warranty of * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the * GNU General Public License for more details. * * You should have received a copy of the GNU General Public License * along with this program; if not, write to the Free Software * Foundation, Inc., 51 Franklin Street, Fifth Floor, Boston, MA 02110-1301 USA * */ /* This file was automatically generated --- DO NOT EDIT */ /* Generated on Sun Nov 25 07:39:50 EST 2012 */ #include "codelet-rdft.h" #ifdef HAVE_FMA /* Generated by: ../../../genfft/gen_hc2hc.native -fma -reorder-insns -schedule-for-pipeline -compact -variables 4 -pipeline-latency 4 -n 5 -dit -name hf_5 -include hf.h */ /* * This function contains 40 FP additions, 34 FP multiplications, * (or, 14 additions, 8 multiplications, 26 fused multiply/add), * 43 stack variables, 4 constants, and 20 memory accesses */ #include "hf.h" static void hf_5(R *cr, R *ci, const R *W, stride rs, INT mb, INT me, INT ms) { DK(KP951056516, +0.951056516295153572116439333379382143405698634); DK(KP559016994, +0.559016994374947424102293417182819058860154590); DK(KP250000000, +0.250000000000000000000000000000000000000000000); DK(KP618033988, +0.618033988749894848204586834365638117720309180); { INT m; for (m = mb, W = W + ((mb - 1) * 8); m < me; m = m + 1, cr = cr + ms, ci = ci - ms, W = W + 8, MAKE_VOLATILE_STRIDE(10, rs)) { E T1, TJ, TK, TA, TR, Te, TC, Tk, TE, Tq; { E Tg, Tj, Tm, TB, Th, Tp, Tl, Ti, To, TD, Tn; T1 = cr[0]; TJ = ci[0]; { E T9, Tc, Ty, Ta, Tb, Tx, T7, Tf, Tz, Td; { E T3, T6, T8, Tw, T4, T2, T5; T3 = cr[WS(rs, 1)]; T6 = ci[WS(rs, 1)]; T2 = W[0]; T9 = cr[WS(rs, 4)]; Tc = ci[WS(rs, 4)]; T8 = W[6]; Tw = T2 * T6; T4 = T2 * T3; T5 = W[1]; Ty = T8 * Tc; Ta = T8 * T9; Tb = W[7]; Tx = FNMS(T5, T3, Tw); T7 = FMA(T5, T6, T4); } Tg = cr[WS(rs, 2)]; Tz = FNMS(Tb, T9, Ty); Td = FMA(Tb, Tc, Ta); Tj = ci[WS(rs, 2)]; Tf = W[2]; TK = Tx + Tz; TA = Tx - Tz; TR = Td - T7; Te = T7 + Td; Tm = cr[WS(rs, 3)]; TB = Tf * Tj; Th = Tf * Tg; Tp = ci[WS(rs, 3)]; Tl = W[4]; Ti = W[3]; To = W[5]; } TD = Tl * Tp; Tn = Tl * Tm; TC = FNMS(Ti, Tg, TB); Tk = FMA(Ti, Tj, Th); TE = FNMS(To, Tm, TD); Tq = FMA(To, Tp, Tn); } { E TG, TI, TO, TS, TU, Tu, TN, Tt, TL, TF; TL = TC + TE; TF = TC - TE; { E Tr, TQ, TM, Ts; Tr = Tk + Tq; TQ = Tk - Tq; TG = FMA(KP618033988, TF, TA); TI = FNMS(KP618033988, TA, TF); TO = TK - TL; TM = TK + TL; TS = FMA(KP618033988, TR, TQ); TU = FNMS(KP618033988, TQ, TR); Tu = Te - Tr; Ts = Te + Tr; ci[WS(rs, 4)] = TM + TJ; TN = FNMS(KP250000000, TM, TJ); cr[0] = T1 + Ts; Tt = FNMS(KP250000000, Ts, T1); } { E TT, TP, Tv, TH; TT = FMA(KP559016994, TO, TN); TP = FNMS(KP559016994, TO, TN); Tv = FMA(KP559016994, Tu, Tt); TH = FNMS(KP559016994, Tu, Tt); ci[WS(rs, 2)] = FMA(KP951056516, TS, TP); cr[WS(rs, 3)] = FMS(KP951056516, TS, TP); ci[WS(rs, 3)] = FMA(KP951056516, TU, TT); cr[WS(rs, 4)] = FMS(KP951056516, TU, TT); ci[WS(rs, 1)] = FMA(KP951056516, TI, TH); cr[WS(rs, 2)] = FNMS(KP951056516, TI, TH); cr[WS(rs, 1)] = FMA(KP951056516, TG, Tv); ci[0] = FNMS(KP951056516, TG, Tv); } } } } } static const tw_instr twinstr[] = { {TW_FULL, 1, 5}, {TW_NEXT, 1, 0} }; static const hc2hc_desc desc = { 5, "hf_5", twinstr, &GENUS, {14, 8, 26, 0} }; void X(codelet_hf_5) (planner *p) { X(khc2hc_register) (p, hf_5, &desc); } #else /* HAVE_FMA */ /* Generated by: ../../../genfft/gen_hc2hc.native -compact -variables 4 -pipeline-latency 4 -n 5 -dit -name hf_5 -include hf.h */ /* * This function contains 40 FP additions, 28 FP multiplications, * (or, 26 additions, 14 multiplications, 14 fused multiply/add), * 29 stack variables, 4 constants, and 20 memory accesses */ #include "hf.h" static void hf_5(R *cr, R *ci, const R *W, stride rs, INT mb, INT me, INT ms) { DK(KP250000000, +0.250000000000000000000000000000000000000000000); DK(KP559016994, +0.559016994374947424102293417182819058860154590); DK(KP587785252, +0.587785252292473129168705954639072768597652438); DK(KP951056516, +0.951056516295153572116439333379382143405698634); { INT m; for (m = mb, W = W + ((mb - 1) * 8); m < me; m = m + 1, cr = cr + ms, ci = ci - ms, W = W + 8, MAKE_VOLATILE_STRIDE(10, rs)) { E T1, TE, Tu, Tx, TC, TB, TF, TG, TH, Tc, Tn, To; T1 = cr[0]; TE = ci[0]; { E T6, Ts, Tm, Tw, Tb, Tt, Th, Tv; { E T3, T5, T2, T4; T3 = cr[WS(rs, 1)]; T5 = ci[WS(rs, 1)]; T2 = W[0]; T4 = W[1]; T6 = FMA(T2, T3, T4 * T5); Ts = FNMS(T4, T3, T2 * T5); } { E Tj, Tl, Ti, Tk; Tj = cr[WS(rs, 3)]; Tl = ci[WS(rs, 3)]; Ti = W[4]; Tk = W[5]; Tm = FMA(Ti, Tj, Tk * Tl); Tw = FNMS(Tk, Tj, Ti * Tl); } { E T8, Ta, T7, T9; T8 = cr[WS(rs, 4)]; Ta = ci[WS(rs, 4)]; T7 = W[6]; T9 = W[7]; Tb = FMA(T7, T8, T9 * Ta); Tt = FNMS(T9, T8, T7 * Ta); } { E Te, Tg, Td, Tf; Te = cr[WS(rs, 2)]; Tg = ci[WS(rs, 2)]; Td = W[2]; Tf = W[3]; Th = FMA(Td, Te, Tf * Tg); Tv = FNMS(Tf, Te, Td * Tg); } Tu = Ts - Tt; Tx = Tv - Tw; TC = Th - Tm; TB = Tb - T6; TF = Ts + Tt; TG = Tv + Tw; TH = TF + TG; Tc = T6 + Tb; Tn = Th + Tm; To = Tc + Tn; } cr[0] = T1 + To; { E Ty, TA, Tr, Tz, Tp, Tq; Ty = FMA(KP951056516, Tu, KP587785252 * Tx); TA = FNMS(KP587785252, Tu, KP951056516 * Tx); Tp = KP559016994 * (Tc - Tn); Tq = FNMS(KP250000000, To, T1); Tr = Tp + Tq; Tz = Tq - Tp; ci[0] = Tr - Ty; ci[WS(rs, 1)] = Tz + TA; cr[WS(rs, 1)] = Tr + Ty; cr[WS(rs, 2)] = Tz - TA; } ci[WS(rs, 4)] = TH + TE; { E TD, TL, TK, TM, TI, TJ; TD = FMA(KP587785252, TB, KP951056516 * TC); TL = FNMS(KP587785252, TC, KP951056516 * TB); TI = FNMS(KP250000000, TH, TE); TJ = KP559016994 * (TF - TG); TK = TI - TJ; TM = TJ + TI; cr[WS(rs, 3)] = TD - TK; ci[WS(rs, 3)] = TL + TM; ci[WS(rs, 2)] = TD + TK; cr[WS(rs, 4)] = TL - TM; } } } } static const tw_instr twinstr[] = { {TW_FULL, 1, 5}, {TW_NEXT, 1, 0} }; static const hc2hc_desc desc = { 5, "hf_5", twinstr, &GENUS, {26, 14, 14, 0} }; void X(codelet_hf_5) (planner *p) { X(khc2hc_register) (p, hf_5, &desc); } #endif /* HAVE_FMA */ fftw-3.3.3/rdft/scalar/r2cf/hf_6.c0000644000175000017500000001722512054410626013444 00000000000000/* * Copyright (c) 2003, 2007-11 Matteo Frigo * Copyright (c) 2003, 2007-11 Massachusetts Institute of Technology * * This program is free software; you can redistribute it and/or modify * it under the terms of the GNU General Public License as published by * the Free Software Foundation; either version 2 of the License, or * (at your option) any later version. * * This program is distributed in the hope that it will be useful, * but WITHOUT ANY WARRANTY; without even the implied warranty of * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the * GNU General Public License for more details. * * You should have received a copy of the GNU General Public License * along with this program; if not, write to the Free Software * Foundation, Inc., 51 Franklin Street, Fifth Floor, Boston, MA 02110-1301 USA * */ /* This file was automatically generated --- DO NOT EDIT */ /* Generated on Sun Nov 25 07:39:50 EST 2012 */ #include "codelet-rdft.h" #ifdef HAVE_FMA /* Generated by: ../../../genfft/gen_hc2hc.native -fma -reorder-insns -schedule-for-pipeline -compact -variables 4 -pipeline-latency 4 -n 6 -dit -name hf_6 -include hf.h */ /* * This function contains 46 FP additions, 32 FP multiplications, * (or, 24 additions, 10 multiplications, 22 fused multiply/add), * 47 stack variables, 2 constants, and 24 memory accesses */ #include "hf.h" static void hf_6(R *cr, R *ci, const R *W, stride rs, INT mb, INT me, INT ms) { DK(KP866025403, +0.866025403784438646763723170752936183471402627); DK(KP500000000, +0.500000000000000000000000000000000000000000000); { INT m; for (m = mb, W = W + ((mb - 1) * 10); m < me; m = m + 1, cr = cr + ms, ci = ci - ms, W = W + 10, MAKE_VOLATILE_STRIDE(12, rs)) { E T11, T12, T14, T13; { E T1, TV, TX, T7, Tn, Tq, TO, TR, TB, Tl, To, TH, Tt, Tw, Ts; E Tp, Tv; T1 = cr[0]; TV = ci[0]; { E T3, T6, T2, T5; T3 = cr[WS(rs, 3)]; T6 = ci[WS(rs, 3)]; T2 = W[4]; T5 = W[5]; { E Ta, Td, Tg, TM, Tb, Tj, Tf, Tc, Ti, TW, T4, T9; Ta = cr[WS(rs, 2)]; Td = ci[WS(rs, 2)]; TW = T2 * T6; T4 = T2 * T3; T9 = W[2]; Tg = cr[WS(rs, 5)]; TX = FNMS(T5, T3, TW); T7 = FMA(T5, T6, T4); TM = T9 * Td; Tb = T9 * Ta; Tj = ci[WS(rs, 5)]; Tf = W[8]; Tc = W[3]; Ti = W[9]; { E TN, Te, TL, Tk, TK, Th, Tm; Tn = cr[WS(rs, 4)]; TK = Tf * Tj; Th = Tf * Tg; TN = FNMS(Tc, Ta, TM); Te = FMA(Tc, Td, Tb); TL = FNMS(Ti, Tg, TK); Tk = FMA(Ti, Tj, Th); Tq = ci[WS(rs, 4)]; Tm = W[6]; TO = TL - TN; TR = TN + TL; TB = Te + Tk; Tl = Te - Tk; To = Tm * Tn; TH = Tm * Tq; } Tt = cr[WS(rs, 1)]; Tw = ci[WS(rs, 1)]; Ts = W[0]; Tp = W[7]; Tv = W[1]; } } { E TA, T8, TI, Tr, TG, Tx, TF, Tu; TA = T1 + T7; T8 = T1 - T7; TF = Ts * Tw; Tu = Ts * Tt; TI = FNMS(Tp, Tn, TH); Tr = FMA(Tp, Tq, To); TG = FNMS(Tv, Tt, TF); Tx = FMA(Tv, Tw, Tu); { E TY, TU, TP, TT, TD, T10, Tz, TZ, TQ, TE; T11 = TX + TV; TY = TV - TX; { E TJ, TS, TC, Ty; TJ = TG - TI; TS = TI + TG; TC = Tr + Tx; Ty = Tr - Tx; TU = TO + TJ; TP = TJ - TO; TT = TR - TS; T12 = TR + TS; T14 = TB - TC; TD = TB + TC; T10 = Ty - Tl; Tz = Tl + Ty; TZ = FMA(KP500000000, TU, TY); } cr[0] = TA + TD; TQ = FNMS(KP500000000, TD, TA); ci[WS(rs, 2)] = T8 + Tz; TE = FNMS(KP500000000, Tz, T8); cr[WS(rs, 3)] = TU - TY; cr[WS(rs, 2)] = FNMS(KP866025403, TT, TQ); ci[WS(rs, 1)] = FMA(KP866025403, TT, TQ); ci[0] = FNMS(KP866025403, TP, TE); cr[WS(rs, 1)] = FMA(KP866025403, TP, TE); ci[WS(rs, 4)] = FMA(KP866025403, T10, TZ); cr[WS(rs, 5)] = FMS(KP866025403, T10, TZ); } } } ci[WS(rs, 5)] = T12 + T11; T13 = FNMS(KP500000000, T12, T11); ci[WS(rs, 3)] = FMA(KP866025403, T14, T13); cr[WS(rs, 4)] = FMS(KP866025403, T14, T13); } } } static const tw_instr twinstr[] = { {TW_FULL, 1, 6}, {TW_NEXT, 1, 0} }; static const hc2hc_desc desc = { 6, "hf_6", twinstr, &GENUS, {24, 10, 22, 0} }; void X(codelet_hf_6) (planner *p) { X(khc2hc_register) (p, hf_6, &desc); } #else /* HAVE_FMA */ /* Generated by: ../../../genfft/gen_hc2hc.native -compact -variables 4 -pipeline-latency 4 -n 6 -dit -name hf_6 -include hf.h */ /* * This function contains 46 FP additions, 28 FP multiplications, * (or, 32 additions, 14 multiplications, 14 fused multiply/add), * 23 stack variables, 2 constants, and 24 memory accesses */ #include "hf.h" static void hf_6(R *cr, R *ci, const R *W, stride rs, INT mb, INT me, INT ms) { DK(KP500000000, +0.500000000000000000000000000000000000000000000); DK(KP866025403, +0.866025403784438646763723170752936183471402627); { INT m; for (m = mb, W = W + ((mb - 1) * 10); m < me; m = m + 1, cr = cr + ms, ci = ci - ms, W = W + 10, MAKE_VOLATILE_STRIDE(12, rs)) { E T7, TS, Tv, TO, Tt, TJ, Tx, TF, Ti, TI, Tw, TC; { E T1, TM, T6, TN; T1 = cr[0]; TM = ci[0]; { E T3, T5, T2, T4; T3 = cr[WS(rs, 3)]; T5 = ci[WS(rs, 3)]; T2 = W[4]; T4 = W[5]; T6 = FMA(T2, T3, T4 * T5); TN = FNMS(T4, T3, T2 * T5); } T7 = T1 - T6; TS = TN + TM; Tv = T1 + T6; TO = TM - TN; } { E Tn, TE, Ts, TD; { E Tk, Tm, Tj, Tl; Tk = cr[WS(rs, 4)]; Tm = ci[WS(rs, 4)]; Tj = W[6]; Tl = W[7]; Tn = FMA(Tj, Tk, Tl * Tm); TE = FNMS(Tl, Tk, Tj * Tm); } { E Tp, Tr, To, Tq; Tp = cr[WS(rs, 1)]; Tr = ci[WS(rs, 1)]; To = W[0]; Tq = W[1]; Ts = FMA(To, Tp, Tq * Tr); TD = FNMS(Tq, Tp, To * Tr); } Tt = Tn - Ts; TJ = TE + TD; Tx = Tn + Ts; TF = TD - TE; } { E Tc, TA, Th, TB; { E T9, Tb, T8, Ta; T9 = cr[WS(rs, 2)]; Tb = ci[WS(rs, 2)]; T8 = W[2]; Ta = W[3]; Tc = FMA(T8, T9, Ta * Tb); TA = FNMS(Ta, T9, T8 * Tb); } { E Te, Tg, Td, Tf; Te = cr[WS(rs, 5)]; Tg = ci[WS(rs, 5)]; Td = W[8]; Tf = W[9]; Th = FMA(Td, Te, Tf * Tg); TB = FNMS(Tf, Te, Td * Tg); } Ti = Tc - Th; TI = TA + TB; Tw = Tc + Th; TC = TA - TB; } { E TG, Tu, Tz, TK, Ty, TH; TG = KP866025403 * (TC + TF); Tu = Ti + Tt; Tz = FNMS(KP500000000, Tu, T7); ci[WS(rs, 2)] = T7 + Tu; cr[WS(rs, 1)] = Tz + TG; ci[0] = Tz - TG; TK = KP866025403 * (TI - TJ); Ty = Tw + Tx; TH = FNMS(KP500000000, Ty, Tv); cr[0] = Tv + Ty; ci[WS(rs, 1)] = TH + TK; cr[WS(rs, 2)] = TH - TK; } { E TP, TL, TQ, TR, TT, TU; TP = KP866025403 * (Tt - Ti); TL = TF - TC; TQ = FMA(KP500000000, TL, TO); cr[WS(rs, 3)] = TL - TO; ci[WS(rs, 4)] = TP + TQ; cr[WS(rs, 5)] = TP - TQ; TR = KP866025403 * (Tw - Tx); TT = TI + TJ; TU = FNMS(KP500000000, TT, TS); cr[WS(rs, 4)] = TR - TU; ci[WS(rs, 5)] = TT + TS; ci[WS(rs, 3)] = TR + TU; } } } } static const tw_instr twinstr[] = { {TW_FULL, 1, 6}, {TW_NEXT, 1, 0} }; static const hc2hc_desc desc = { 6, "hf_6", twinstr, &GENUS, {32, 14, 14, 0} }; void X(codelet_hf_6) (planner *p) { X(khc2hc_register) (p, hf_6, &desc); } #endif /* HAVE_FMA */ fftw-3.3.3/rdft/scalar/r2cf/hf_7.c0000644000175000017500000002504612054410627013446 00000000000000/* * Copyright (c) 2003, 2007-11 Matteo Frigo * Copyright (c) 2003, 2007-11 Massachusetts Institute of Technology * * This program is free software; you can redistribute it and/or modify * it under the terms of the GNU General Public License as published by * the Free Software Foundation; either version 2 of the License, or * (at your option) any later version. * * This program is distributed in the hope that it will be useful, * but WITHOUT ANY WARRANTY; without even the implied warranty of * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the * GNU General Public License for more details. * * You should have received a copy of the GNU General Public License * along with this program; if not, write to the Free Software * Foundation, Inc., 51 Franklin Street, Fifth Floor, Boston, MA 02110-1301 USA * */ /* This file was automatically generated --- DO NOT EDIT */ /* Generated on Sun Nov 25 07:39:50 EST 2012 */ #include "codelet-rdft.h" #ifdef HAVE_FMA /* Generated by: ../../../genfft/gen_hc2hc.native -fma -reorder-insns -schedule-for-pipeline -compact -variables 4 -pipeline-latency 4 -n 7 -dit -name hf_7 -include hf.h */ /* * This function contains 72 FP additions, 66 FP multiplications, * (or, 18 additions, 12 multiplications, 54 fused multiply/add), * 62 stack variables, 6 constants, and 28 memory accesses */ #include "hf.h" static void hf_7(R *cr, R *ci, const R *W, stride rs, INT mb, INT me, INT ms) { DK(KP974927912, +0.974927912181823607018131682993931217232785801); DK(KP801937735, +0.801937735804838252472204639014890102331838324); DK(KP900968867, +0.900968867902419126236102319507445051165919162); DK(KP692021471, +0.692021471630095869627814897002069140197260599); DK(KP554958132, +0.554958132087371191422194871006410481067288862); DK(KP356895867, +0.356895867892209443894399510021300583399127187); { INT m; for (m = mb, W = W + ((mb - 1) * 12); m < me; m = m + 1, cr = cr + ms, ci = ci - ms, W = W + 12, MAKE_VOLATILE_STRIDE(14, rs)) { E T1, TR, T18, T10, T12, T16, T11, T13; { E T19, T1a, T1i, Te, Tt, Tw, T1b, TM, T1h, Tr, Tu, TS, Tz, TC, Ty; E Tv, TB; T1 = cr[0]; T19 = ci[0]; { E T9, Tc, TP, Ta, Tb, TO, T7; { E T3, T6, T8, TN, T4, T2, T5; T3 = cr[WS(rs, 1)]; T6 = ci[WS(rs, 1)]; T2 = W[0]; T9 = cr[WS(rs, 6)]; Tc = ci[WS(rs, 6)]; T8 = W[10]; TN = T2 * T6; T4 = T2 * T3; T5 = W[1]; TP = T8 * Tc; Ta = T8 * T9; Tb = W[11]; TO = FNMS(T5, T3, TN); T7 = FMA(T5, T6, T4); } { E Tg, Tj, Th, TI, Tm, Tp, Tl, Ti, To, TQ, Td, Tf; Tg = cr[WS(rs, 2)]; TQ = FNMS(Tb, T9, TP); Td = FMA(Tb, Tc, Ta); Tj = ci[WS(rs, 2)]; Tf = W[2]; T1a = TO + TQ; TR = TO - TQ; T1i = Td - T7; Te = T7 + Td; Th = Tf * Tg; TI = Tf * Tj; Tm = cr[WS(rs, 5)]; Tp = ci[WS(rs, 5)]; Tl = W[8]; Ti = W[3]; To = W[9]; { E TJ, Tk, TL, Tq, TK, Tn, Ts; Tt = cr[WS(rs, 3)]; TK = Tl * Tp; Tn = Tl * Tm; TJ = FNMS(Ti, Tg, TI); Tk = FMA(Ti, Tj, Th); TL = FNMS(To, Tm, TK); Tq = FMA(To, Tp, Tn); Tw = ci[WS(rs, 3)]; Ts = W[4]; T1b = TJ + TL; TM = TJ - TL; T1h = Tq - Tk; Tr = Tk + Tq; Tu = Ts * Tt; TS = Ts * Tw; } Tz = cr[WS(rs, 4)]; TC = ci[WS(rs, 4)]; Ty = W[6]; Tv = W[5]; TB = W[7]; } } { E TF, TT, Tx, TV, TD, T1q, TU, TA; TF = FNMS(KP356895867, Tr, Te); TU = Ty * TC; TA = Ty * Tz; TT = FNMS(Tv, Tt, TS); Tx = FMA(Tv, Tw, Tu); TV = FNMS(TB, Tz, TU); TD = FMA(TB, TC, TA); T1q = FNMS(KP356895867, T1b, T1a); { E TW, TE, T1k, T1f; { E T1e, T1s, TY, T1p, T1u, TH, T1n, T1j, T1c, T1g; T1j = FNMS(KP554958132, T1i, T1h); T1c = TT + TV; TW = TT - TV; T1g = TD - Tx; TE = Tx + TD; { E T1d, T1l, T1r, TX; T1d = FNMS(KP356895867, T1c, T1b); T1l = FNMS(KP356895867, T1a, T1c); T1r = FNMS(KP692021471, T1q, T1c); ci[WS(rs, 6)] = T1a + T1b + T1c + T19; TX = FMA(KP554958132, TW, TR); { E T1o, T1t, TG, T1m; T1o = FMA(KP554958132, T1h, T1g); T1t = FMA(KP554958132, T1g, T1i); TG = FNMS(KP692021471, TF, TE); cr[0] = T1 + Te + Tr + TE; T1e = FNMS(KP692021471, T1d, T1a); T1m = FNMS(KP692021471, T1l, T1b); T1s = FNMS(KP900968867, T1r, T19); TY = FMA(KP801937735, TX, TM); T1p = FNMS(KP801937735, T1o, T1i); T1u = FMA(KP801937735, T1t, T1h); TH = FNMS(KP900968867, TG, T1); T1n = FNMS(KP900968867, T1m, T19); T1k = FNMS(KP801937735, T1j, T1g); } } ci[WS(rs, 5)] = FMA(KP974927912, T1u, T1s); cr[WS(rs, 6)] = FMS(KP974927912, T1u, T1s); cr[WS(rs, 1)] = FMA(KP974927912, TY, TH); ci[0] = FNMS(KP974927912, TY, TH); ci[WS(rs, 4)] = FMA(KP974927912, T1p, T1n); cr[WS(rs, 5)] = FMS(KP974927912, T1p, T1n); T1f = FNMS(KP900968867, T1e, T19); } { E T14, T17, T15, TZ; T14 = FNMS(KP356895867, TE, Tr); T17 = FNMS(KP554958132, TR, TM); TZ = FNMS(KP356895867, Te, TE); ci[WS(rs, 3)] = FMA(KP974927912, T1k, T1f); cr[WS(rs, 4)] = FMS(KP974927912, T1k, T1f); T15 = FNMS(KP692021471, T14, Te); T18 = FNMS(KP801937735, T17, TW); T10 = FNMS(KP692021471, TZ, Tr); T12 = FMA(KP554958132, TM, TW); T16 = FNMS(KP900968867, T15, T1); } } } } T11 = FNMS(KP900968867, T10, T1); T13 = FNMS(KP801937735, T12, TR); cr[WS(rs, 3)] = FMA(KP974927912, T18, T16); ci[WS(rs, 2)] = FNMS(KP974927912, T18, T16); cr[WS(rs, 2)] = FMA(KP974927912, T13, T11); ci[WS(rs, 1)] = FNMS(KP974927912, T13, T11); } } } static const tw_instr twinstr[] = { {TW_FULL, 1, 7}, {TW_NEXT, 1, 0} }; static const hc2hc_desc desc = { 7, "hf_7", twinstr, &GENUS, {18, 12, 54, 0} }; void X(codelet_hf_7) (planner *p) { X(khc2hc_register) (p, hf_7, &desc); } #else /* HAVE_FMA */ /* Generated by: ../../../genfft/gen_hc2hc.native -compact -variables 4 -pipeline-latency 4 -n 7 -dit -name hf_7 -include hf.h */ /* * This function contains 72 FP additions, 60 FP multiplications, * (or, 36 additions, 24 multiplications, 36 fused multiply/add), * 29 stack variables, 6 constants, and 28 memory accesses */ #include "hf.h" static void hf_7(R *cr, R *ci, const R *W, stride rs, INT mb, INT me, INT ms) { DK(KP222520933, +0.222520933956314404288902564496794759466355569); DK(KP900968867, +0.900968867902419126236102319507445051165919162); DK(KP623489801, +0.623489801858733530525004884004239810632274731); DK(KP433883739, +0.433883739117558120475768332848358754609990728); DK(KP974927912, +0.974927912181823607018131682993931217232785801); DK(KP781831482, +0.781831482468029808708444526674057750232334519); { INT m; for (m = mb, W = W + ((mb - 1) * 12); m < me; m = m + 1, cr = cr + ms, ci = ci - ms, W = W + 12, MAKE_VOLATILE_STRIDE(14, rs)) { E T1, TT, Tc, TV, TC, TO, Tn, TS, TI, TP, Ty, TU, TF, TQ; T1 = cr[0]; TT = ci[0]; { E T6, TA, Tb, TB; { E T3, T5, T2, T4; T3 = cr[WS(rs, 1)]; T5 = ci[WS(rs, 1)]; T2 = W[0]; T4 = W[1]; T6 = FMA(T2, T3, T4 * T5); TA = FNMS(T4, T3, T2 * T5); } { E T8, Ta, T7, T9; T8 = cr[WS(rs, 6)]; Ta = ci[WS(rs, 6)]; T7 = W[10]; T9 = W[11]; Tb = FMA(T7, T8, T9 * Ta); TB = FNMS(T9, T8, T7 * Ta); } Tc = T6 + Tb; TV = TA + TB; TC = TA - TB; TO = Tb - T6; } { E Th, TG, Tm, TH; { E Te, Tg, Td, Tf; Te = cr[WS(rs, 2)]; Tg = ci[WS(rs, 2)]; Td = W[2]; Tf = W[3]; Th = FMA(Td, Te, Tf * Tg); TG = FNMS(Tf, Te, Td * Tg); } { E Tj, Tl, Ti, Tk; Tj = cr[WS(rs, 5)]; Tl = ci[WS(rs, 5)]; Ti = W[8]; Tk = W[9]; Tm = FMA(Ti, Tj, Tk * Tl); TH = FNMS(Tk, Tj, Ti * Tl); } Tn = Th + Tm; TS = TG + TH; TI = TG - TH; TP = Th - Tm; } { E Ts, TD, Tx, TE; { E Tp, Tr, To, Tq; Tp = cr[WS(rs, 3)]; Tr = ci[WS(rs, 3)]; To = W[4]; Tq = W[5]; Ts = FMA(To, Tp, Tq * Tr); TD = FNMS(Tq, Tp, To * Tr); } { E Tu, Tw, Tt, Tv; Tu = cr[WS(rs, 4)]; Tw = ci[WS(rs, 4)]; Tt = W[6]; Tv = W[7]; Tx = FMA(Tt, Tu, Tv * Tw); TE = FNMS(Tv, Tu, Tt * Tw); } Ty = Ts + Tx; TU = TD + TE; TF = TD - TE; TQ = Tx - Ts; } { E TL, TK, TZ, T10; cr[0] = T1 + Tc + Tn + Ty; TL = FMA(KP781831482, TC, KP974927912 * TI) + (KP433883739 * TF); TK = FMA(KP623489801, Tc, T1) + FNMA(KP900968867, Ty, KP222520933 * Tn); ci[0] = TK - TL; cr[WS(rs, 1)] = TK + TL; ci[WS(rs, 6)] = TV + TS + TU + TT; TZ = FMA(KP781831482, TO, KP433883739 * TQ) - (KP974927912 * TP); T10 = FMA(KP623489801, TV, TT) + FNMA(KP900968867, TU, KP222520933 * TS); cr[WS(rs, 6)] = TZ - T10; ci[WS(rs, 5)] = TZ + T10; } { E TX, TY, TR, TW; TX = FMA(KP974927912, TO, KP433883739 * TP) - (KP781831482 * TQ); TY = FMA(KP623489801, TU, TT) + FNMA(KP900968867, TS, KP222520933 * TV); cr[WS(rs, 5)] = TX - TY; ci[WS(rs, 4)] = TX + TY; TR = FMA(KP433883739, TO, KP781831482 * TP) + (KP974927912 * TQ); TW = FMA(KP623489801, TS, TT) + FNMA(KP222520933, TU, KP900968867 * TV); cr[WS(rs, 4)] = TR - TW; ci[WS(rs, 3)] = TR + TW; } { E TN, TM, TJ, Tz; TN = FMA(KP433883739, TC, KP974927912 * TF) - (KP781831482 * TI); TM = FMA(KP623489801, Tn, T1) + FNMA(KP222520933, Ty, KP900968867 * Tc); ci[WS(rs, 2)] = TM - TN; cr[WS(rs, 3)] = TM + TN; TJ = FNMS(KP781831482, TF, KP974927912 * TC) - (KP433883739 * TI); Tz = FMA(KP623489801, Ty, T1) + FNMA(KP900968867, Tn, KP222520933 * Tc); ci[WS(rs, 1)] = Tz - TJ; cr[WS(rs, 2)] = Tz + TJ; } } } } static const tw_instr twinstr[] = { {TW_FULL, 1, 7}, {TW_NEXT, 1, 0} }; static const hc2hc_desc desc = { 7, "hf_7", twinstr, &GENUS, {36, 24, 36, 0} }; void X(codelet_hf_7) (planner *p) { X(khc2hc_register) (p, hf_7, &desc); } #endif /* HAVE_FMA */ fftw-3.3.3/rdft/scalar/r2cf/hf_8.c0000644000175000017500000002226612054410627013450 00000000000000/* * Copyright (c) 2003, 2007-11 Matteo Frigo * Copyright (c) 2003, 2007-11 Massachusetts Institute of Technology * * This program is free software; you can redistribute it and/or modify * it under the terms of the GNU General Public License as published by * the Free Software Foundation; either version 2 of the License, or * (at your option) any later version. * * This program is distributed in the hope that it will be useful, * but WITHOUT ANY WARRANTY; without even the implied warranty of * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the * GNU General Public License for more details. * * You should have received a copy of the GNU General Public License * along with this program; if not, write to the Free Software * Foundation, Inc., 51 Franklin Street, Fifth Floor, Boston, MA 02110-1301 USA * */ /* This file was automatically generated --- DO NOT EDIT */ /* Generated on Sun Nov 25 07:39:51 EST 2012 */ #include "codelet-rdft.h" #ifdef HAVE_FMA /* Generated by: ../../../genfft/gen_hc2hc.native -fma -reorder-insns -schedule-for-pipeline -compact -variables 4 -pipeline-latency 4 -n 8 -dit -name hf_8 -include hf.h */ /* * This function contains 66 FP additions, 36 FP multiplications, * (or, 44 additions, 14 multiplications, 22 fused multiply/add), * 61 stack variables, 1 constants, and 32 memory accesses */ #include "hf.h" static void hf_8(R *cr, R *ci, const R *W, stride rs, INT mb, INT me, INT ms) { DK(KP707106781, +0.707106781186547524400844362104849039284835938); { INT m; for (m = mb, W = W + ((mb - 1) * 14); m < me; m = m + 1, cr = cr + ms, ci = ci - ms, W = W + 14, MAKE_VOLATILE_STRIDE(16, rs)) { E T1f, T1g, T1e, Tm, T1q, T1o, T1p, TN, T1h, T1i; { E T1, T1m, T1l, T7, TS, Tk, TQ, Te, To, Tr, T17, TM, T12, Tu, TW; E Tp, Tx, Tt, Tq, Tw; { E T3, T6, T2, T5; T1 = cr[0]; T1m = ci[0]; T3 = cr[WS(rs, 4)]; T6 = ci[WS(rs, 4)]; T2 = W[6]; T5 = W[7]; { E Ta, Td, T9, Tc; { E Tg, Tj, Ti, TR, Th, T1k, T4, Tf; Tg = cr[WS(rs, 6)]; Tj = ci[WS(rs, 6)]; T1k = T2 * T6; T4 = T2 * T3; Tf = W[10]; Ti = W[11]; T1l = FNMS(T5, T3, T1k); T7 = FMA(T5, T6, T4); TR = Tf * Tj; Th = Tf * Tg; Ta = cr[WS(rs, 2)]; Td = ci[WS(rs, 2)]; TS = FNMS(Ti, Tg, TR); Tk = FMA(Ti, Tj, Th); T9 = W[2]; Tc = W[3]; } { E TB, TE, TH, T13, TC, TK, TG, TD, TJ, TP, Tb, TA, Tn; TB = cr[WS(rs, 7)]; TE = ci[WS(rs, 7)]; TP = T9 * Td; Tb = T9 * Ta; TA = W[12]; TH = cr[WS(rs, 3)]; TQ = FNMS(Tc, Ta, TP); Te = FMA(Tc, Td, Tb); T13 = TA * TE; TC = TA * TB; TK = ci[WS(rs, 3)]; TG = W[4]; TD = W[13]; TJ = W[5]; { E T14, TF, T16, TL, T15, TI; To = cr[WS(rs, 1)]; T15 = TG * TK; TI = TG * TH; T14 = FNMS(TD, TB, T13); TF = FMA(TD, TE, TC); T16 = FNMS(TJ, TH, T15); TL = FMA(TJ, TK, TI); Tr = ci[WS(rs, 1)]; Tn = W[0]; T17 = T14 - T16; T1f = T14 + T16; TM = TF + TL; T12 = TF - TL; } Tu = cr[WS(rs, 5)]; TW = Tn * Tr; Tp = Tn * To; Tx = ci[WS(rs, 5)]; Tt = W[8]; Tq = W[1]; Tw = W[9]; } } } { E T8, T1j, Tl, Tz, T1a, TU, T1n, T1b, T1c, T1v, T1t, T1u, T19, T1w, T1d; { E T1r, T10, TV, T1s, T11, T18; { E TO, TX, Ts, TZ, Ty, TT, TY, Tv; T8 = T1 + T7; TO = T1 - T7; TY = Tt * Tx; Tv = Tt * Tu; TX = FNMS(Tq, To, TW); Ts = FMA(Tq, Tr, Tp); TZ = FNMS(Tw, Tu, TY); Ty = FMA(Tw, Tx, Tv); TT = TQ - TS; T1j = TQ + TS; Tl = Te + Tk; T1r = Te - Tk; T10 = TX - TZ; T1g = TX + TZ; Tz = Ts + Ty; TV = Ts - Ty; T1a = TO - TT; TU = TO + TT; T1s = T1m - T1l; T1n = T1l + T1m; } T1b = TV - T10; T11 = TV + T10; T18 = T12 - T17; T1c = T12 + T17; T1v = T1s - T1r; T1t = T1r + T1s; T1u = T18 - T11; T19 = T11 + T18; } ci[WS(rs, 4)] = FMA(KP707106781, T1u, T1t); cr[WS(rs, 7)] = FMS(KP707106781, T1u, T1t); cr[WS(rs, 1)] = FMA(KP707106781, T19, TU); ci[WS(rs, 2)] = FNMS(KP707106781, T19, TU); T1w = T1c - T1b; T1d = T1b + T1c; ci[WS(rs, 6)] = FMA(KP707106781, T1w, T1v); cr[WS(rs, 5)] = FMS(KP707106781, T1w, T1v); ci[0] = FMA(KP707106781, T1d, T1a); cr[WS(rs, 3)] = FNMS(KP707106781, T1d, T1a); T1e = T8 - Tl; Tm = T8 + Tl; T1q = T1n - T1j; T1o = T1j + T1n; T1p = TM - Tz; TN = Tz + TM; } } ci[WS(rs, 5)] = T1p + T1q; cr[WS(rs, 6)] = T1p - T1q; cr[0] = Tm + TN; ci[WS(rs, 3)] = Tm - TN; T1h = T1f - T1g; T1i = T1g + T1f; ci[WS(rs, 7)] = T1i + T1o; cr[WS(rs, 4)] = T1i - T1o; ci[WS(rs, 1)] = T1e + T1h; cr[WS(rs, 2)] = T1e - T1h; } } } static const tw_instr twinstr[] = { {TW_FULL, 1, 8}, {TW_NEXT, 1, 0} }; static const hc2hc_desc desc = { 8, "hf_8", twinstr, &GENUS, {44, 14, 22, 0} }; void X(codelet_hf_8) (planner *p) { X(khc2hc_register) (p, hf_8, &desc); } #else /* HAVE_FMA */ /* Generated by: ../../../genfft/gen_hc2hc.native -compact -variables 4 -pipeline-latency 4 -n 8 -dit -name hf_8 -include hf.h */ /* * This function contains 66 FP additions, 32 FP multiplications, * (or, 52 additions, 18 multiplications, 14 fused multiply/add), * 28 stack variables, 1 constants, and 32 memory accesses */ #include "hf.h" static void hf_8(R *cr, R *ci, const R *W, stride rs, INT mb, INT me, INT ms) { DK(KP707106781, +0.707106781186547524400844362104849039284835938); { INT m; for (m = mb, W = W + ((mb - 1) * 14); m < me; m = m + 1, cr = cr + ms, ci = ci - ms, W = W + 14, MAKE_VOLATILE_STRIDE(16, rs)) { E T7, T1f, TH, T19, TF, T12, TR, TU, Ti, T1e, TK, T16, Tu, T13, TM; E TP; { E T1, T18, T6, T17; T1 = cr[0]; T18 = ci[0]; { E T3, T5, T2, T4; T3 = cr[WS(rs, 4)]; T5 = ci[WS(rs, 4)]; T2 = W[6]; T4 = W[7]; T6 = FMA(T2, T3, T4 * T5); T17 = FNMS(T4, T3, T2 * T5); } T7 = T1 + T6; T1f = T18 - T17; TH = T1 - T6; T19 = T17 + T18; } { E Tz, TS, TE, TT; { E Tw, Ty, Tv, Tx; Tw = cr[WS(rs, 7)]; Ty = ci[WS(rs, 7)]; Tv = W[12]; Tx = W[13]; Tz = FMA(Tv, Tw, Tx * Ty); TS = FNMS(Tx, Tw, Tv * Ty); } { E TB, TD, TA, TC; TB = cr[WS(rs, 3)]; TD = ci[WS(rs, 3)]; TA = W[4]; TC = W[5]; TE = FMA(TA, TB, TC * TD); TT = FNMS(TC, TB, TA * TD); } TF = Tz + TE; T12 = TS + TT; TR = Tz - TE; TU = TS - TT; } { E Tc, TI, Th, TJ; { E T9, Tb, T8, Ta; T9 = cr[WS(rs, 2)]; Tb = ci[WS(rs, 2)]; T8 = W[2]; Ta = W[3]; Tc = FMA(T8, T9, Ta * Tb); TI = FNMS(Ta, T9, T8 * Tb); } { E Te, Tg, Td, Tf; Te = cr[WS(rs, 6)]; Tg = ci[WS(rs, 6)]; Td = W[10]; Tf = W[11]; Th = FMA(Td, Te, Tf * Tg); TJ = FNMS(Tf, Te, Td * Tg); } Ti = Tc + Th; T1e = Tc - Th; TK = TI - TJ; T16 = TI + TJ; } { E To, TN, Tt, TO; { E Tl, Tn, Tk, Tm; Tl = cr[WS(rs, 1)]; Tn = ci[WS(rs, 1)]; Tk = W[0]; Tm = W[1]; To = FMA(Tk, Tl, Tm * Tn); TN = FNMS(Tm, Tl, Tk * Tn); } { E Tq, Ts, Tp, Tr; Tq = cr[WS(rs, 5)]; Ts = ci[WS(rs, 5)]; Tp = W[8]; Tr = W[9]; Tt = FMA(Tp, Tq, Tr * Ts); TO = FNMS(Tr, Tq, Tp * Ts); } Tu = To + Tt; T13 = TN + TO; TM = To - Tt; TP = TN - TO; } { E Tj, TG, T1b, T1c; Tj = T7 + Ti; TG = Tu + TF; ci[WS(rs, 3)] = Tj - TG; cr[0] = Tj + TG; T1b = TF - Tu; T1c = T19 - T16; cr[WS(rs, 6)] = T1b - T1c; ci[WS(rs, 5)] = T1b + T1c; { E TX, T1i, T10, T1h, TY, TZ; TX = TH - TK; T1i = T1f - T1e; TY = TM - TP; TZ = TR + TU; T10 = KP707106781 * (TY + TZ); T1h = KP707106781 * (TZ - TY); cr[WS(rs, 3)] = TX - T10; ci[WS(rs, 6)] = T1h + T1i; ci[0] = TX + T10; cr[WS(rs, 5)] = T1h - T1i; } } { E T15, T1a, T11, T14; T15 = T13 + T12; T1a = T16 + T19; cr[WS(rs, 4)] = T15 - T1a; ci[WS(rs, 7)] = T15 + T1a; T11 = T7 - Ti; T14 = T12 - T13; cr[WS(rs, 2)] = T11 - T14; ci[WS(rs, 1)] = T11 + T14; { E TL, T1g, TW, T1d, TQ, TV; TL = TH + TK; T1g = T1e + T1f; TQ = TM + TP; TV = TR - TU; TW = KP707106781 * (TQ + TV); T1d = KP707106781 * (TV - TQ); ci[WS(rs, 2)] = TL - TW; ci[WS(rs, 4)] = T1d + T1g; cr[WS(rs, 1)] = TL + TW; cr[WS(rs, 7)] = T1d - T1g; } } } } } static const tw_instr twinstr[] = { {TW_FULL, 1, 8}, {TW_NEXT, 1, 0} }; static const hc2hc_desc desc = { 8, "hf_8", twinstr, &GENUS, {52, 18, 14, 0} }; void X(codelet_hf_8) (planner *p) { X(khc2hc_register) (p, hf_8, &desc); } #endif /* HAVE_FMA */ fftw-3.3.3/rdft/scalar/r2cf/hf_9.c0000644000175000017500000003411312054410630013435 00000000000000/* * Copyright (c) 2003, 2007-11 Matteo Frigo * Copyright (c) 2003, 2007-11 Massachusetts Institute of Technology * * This program is free software; you can redistribute it and/or modify * it under the terms of the GNU General Public License as published by * the Free Software Foundation; either version 2 of the License, or * (at your option) any later version. * * This program is distributed in the hope that it will be useful, * but WITHOUT ANY WARRANTY; without even the implied warranty of * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the * GNU General Public License for more details. * * You should have received a copy of the GNU General Public License * along with this program; if not, write to the Free Software * Foundation, Inc., 51 Franklin Street, Fifth Floor, Boston, MA 02110-1301 USA * */ /* This file was automatically generated --- DO NOT EDIT */ /* Generated on Sun Nov 25 07:39:51 EST 2012 */ #include "codelet-rdft.h" #ifdef HAVE_FMA /* Generated by: ../../../genfft/gen_hc2hc.native -fma -reorder-insns -schedule-for-pipeline -compact -variables 4 -pipeline-latency 4 -n 9 -dit -name hf_9 -include hf.h */ /* * This function contains 96 FP additions, 88 FP multiplications, * (or, 24 additions, 16 multiplications, 72 fused multiply/add), * 69 stack variables, 10 constants, and 36 memory accesses */ #include "hf.h" static void hf_9(R *cr, R *ci, const R *W, stride rs, INT mb, INT me, INT ms) { DK(KP777861913, +0.777861913430206160028177977318626690410586096); DK(KP852868531, +0.852868531952443209628250963940074071936020296); DK(KP839099631, +0.839099631177280011763127298123181364687434283); DK(KP492403876, +0.492403876506104029683371512294761506835321626); DK(KP984807753, +0.984807753012208059366743024589523013670643252); DK(KP954188894, +0.954188894138671133499268364187245676532219158); DK(KP363970234, +0.363970234266202361351047882776834043890471784); DK(KP176326980, +0.176326980708464973471090386868618986121633062); DK(KP866025403, +0.866025403784438646763723170752936183471402627); DK(KP500000000, +0.500000000000000000000000000000000000000000000); { INT m; for (m = mb, W = W + ((mb - 1) * 16); m < me; m = m + 1, cr = cr + ms, ci = ci - ms, W = W + 16, MAKE_VOLATILE_STRIDE(18, rs)) { E T20, T1Z; { E T1, T1P, T1Q, T10, T1S, Te, TB, T1d, T1a, T19, T1M, TE, T1c, Tz, T1n; E TC, TH, TK, T1k, TR, TG, TJ, TD; T1 = cr[0]; T1P = ci[0]; { E T9, Tc, TY, Ta, Tb, TX, T7; { E T3, T6, T8, TW, T4, T2, T5; T3 = cr[WS(rs, 3)]; T6 = ci[WS(rs, 3)]; T2 = W[4]; T9 = cr[WS(rs, 6)]; Tc = ci[WS(rs, 6)]; T8 = W[10]; TW = T2 * T6; T4 = T2 * T3; T5 = W[5]; TY = T8 * Tc; Ta = T8 * T9; Tb = W[11]; TX = FNMS(T5, T3, TW); T7 = FMA(T5, T6, T4); } { E Th, Tk, Ti, T12, Tn, Tq, Tp, T17, Tx, T14, To, Tj, TZ, Td, Tg; E TA, Tl, Ty; Th = cr[WS(rs, 1)]; TZ = FNMS(Tb, T9, TY); Td = FMA(Tb, Tc, Ta); Tk = ci[WS(rs, 1)]; Tg = W[0]; T1Q = TX + TZ; T10 = TX - TZ; T1S = Td - T7; Te = T7 + Td; Ti = Tg * Th; T12 = Tg * Tk; { E Tt, Tw, Ts, Tv, T16, Tu, Tm; Tt = cr[WS(rs, 7)]; Tw = ci[WS(rs, 7)]; Ts = W[12]; Tv = W[13]; Tn = cr[WS(rs, 4)]; Tq = ci[WS(rs, 4)]; T16 = Ts * Tw; Tu = Ts * Tt; Tm = W[6]; Tp = W[7]; T17 = FNMS(Tv, Tt, T16); Tx = FMA(Tv, Tw, Tu); T14 = Tm * Tq; To = Tm * Tn; } Tj = W[1]; TB = cr[WS(rs, 2)]; { E T15, Tr, T13, T18; T15 = FNMS(Tp, Tn, T14); Tr = FMA(Tp, Tq, To); T13 = FNMS(Tj, Th, T12); Tl = FMA(Tj, Tk, Ti); T18 = T15 + T17; T1d = T15 - T17; Ty = Tr + Tx; T1a = Tr - Tx; T19 = FNMS(KP500000000, T18, T13); T1M = T13 + T18; TE = ci[WS(rs, 2)]; } T1c = FNMS(KP500000000, Ty, Tl); Tz = Tl + Ty; TA = W[2]; { E TN, TQ, TP, T1j, TO, TM; TN = cr[WS(rs, 8)]; TQ = ci[WS(rs, 8)]; TM = W[14]; T1n = TA * TE; TC = TA * TB; TP = W[15]; T1j = TM * TQ; TO = TM * TN; TH = cr[WS(rs, 5)]; TK = ci[WS(rs, 5)]; T1k = FNMS(TP, TN, T1j); TR = FMA(TP, TQ, TO); TG = W[8]; TJ = W[9]; } TD = W[3]; } } { E TV, Tf, T21, T1R, T1l, T1r, T1q, T1N, TT, T1g; { E T1o, TF, T1i, TL, T1h, TI, TS, T1p; TV = FNMS(KP500000000, Te, T1); Tf = T1 + Te; T1h = TG * TK; TI = TG * TH; T1o = FNMS(TD, TB, T1n); TF = FMA(TD, TE, TC); T1i = FNMS(TJ, TH, T1h); TL = FMA(TJ, TK, TI); T21 = T1Q + T1P; T1R = FNMS(KP500000000, T1Q, T1P); T1p = T1i + T1k; T1l = T1i - T1k; TS = TL + TR; T1r = TR - TL; T1q = FNMS(KP500000000, T1p, T1o); T1N = T1o + T1p; TT = TF + TS; T1g = FNMS(KP500000000, TS, TF); } { E T11, T1z, T1E, T1D, T1X, T1T, T1I, T1C, T1Y, T1y, T1u, T24, TU; T24 = TT - Tz; TU = Tz + TT; { E T22, T1O, T1L, T23; T22 = T1M + T1N; T1O = T1M - T1N; T11 = FNMS(KP866025403, T10, TV); T1z = FMA(KP866025403, T10, TV); T1L = FNMS(KP500000000, TU, Tf); cr[0] = Tf + TU; T23 = FNMS(KP500000000, T22, T21); ci[WS(rs, 8)] = T22 + T21; cr[WS(rs, 3)] = FMA(KP866025403, T1O, T1L); ci[WS(rs, 2)] = FNMS(KP866025403, T1O, T1L); ci[WS(rs, 5)] = FMA(KP866025403, T24, T23); cr[WS(rs, 6)] = FMS(KP866025403, T24, T23); } { E T1B, T1m, T1w, T1f, T1s, T1A, T1b, T1e, T1x, T1t; T1E = FNMS(KP866025403, T1a, T19); T1b = FMA(KP866025403, T1a, T19); T1e = FNMS(KP866025403, T1d, T1c); T1D = FMA(KP866025403, T1d, T1c); T1B = FMA(KP866025403, T1l, T1g); T1m = FNMS(KP866025403, T1l, T1g); T1X = FNMS(KP866025403, T1S, T1R); T1T = FMA(KP866025403, T1S, T1R); T1w = FNMS(KP176326980, T1b, T1e); T1f = FMA(KP176326980, T1e, T1b); T1s = FNMS(KP866025403, T1r, T1q); T1A = FMA(KP866025403, T1r, T1q); T1x = FMA(KP363970234, T1m, T1s); T1t = FNMS(KP363970234, T1s, T1m); T1I = FNMS(KP176326980, T1A, T1B); T1C = FMA(KP176326980, T1B, T1A); T1Y = FMA(KP954188894, T1x, T1w); T1y = FNMS(KP954188894, T1x, T1w); T20 = FMA(KP954188894, T1t, T1f); T1u = FNMS(KP954188894, T1t, T1f); } { E T1F, T1J, T1v, T1U, T1K; ci[WS(rs, 6)] = FNMS(KP984807753, T1Y, T1X); T1v = FNMS(KP492403876, T1u, T11); cr[WS(rs, 2)] = FMA(KP984807753, T1u, T11); T1F = FMA(KP839099631, T1E, T1D); T1J = FNMS(KP839099631, T1D, T1E); ci[WS(rs, 3)] = FNMS(KP852868531, T1y, T1v); ci[0] = FMA(KP852868531, T1y, T1v); T1U = FNMS(KP777861913, T1J, T1I); T1K = FMA(KP777861913, T1J, T1I); { E T1G, T1W, T1V, T1H; T1G = FMA(KP777861913, T1F, T1C); T1W = FNMS(KP777861913, T1F, T1C); T1Z = FMA(KP492403876, T1Y, T1X); T1V = FMA(KP492403876, T1U, T1T); ci[WS(rs, 7)] = FNMS(KP984807753, T1U, T1T); T1H = FNMS(KP492403876, T1G, T1z); cr[WS(rs, 1)] = FMA(KP984807753, T1G, T1z); ci[WS(rs, 4)] = FMA(KP852868531, T1W, T1V); cr[WS(rs, 7)] = FMS(KP852868531, T1W, T1V); cr[WS(rs, 4)] = FMA(KP852868531, T1K, T1H); ci[WS(rs, 1)] = FNMS(KP852868531, T1K, T1H); } } } } } cr[WS(rs, 8)] = -(FMA(KP852868531, T20, T1Z)); cr[WS(rs, 5)] = FMS(KP852868531, T20, T1Z); } } } static const tw_instr twinstr[] = { {TW_FULL, 1, 9}, {TW_NEXT, 1, 0} }; static const hc2hc_desc desc = { 9, "hf_9", twinstr, &GENUS, {24, 16, 72, 0} }; void X(codelet_hf_9) (planner *p) { X(khc2hc_register) (p, hf_9, &desc); } #else /* HAVE_FMA */ /* Generated by: ../../../genfft/gen_hc2hc.native -compact -variables 4 -pipeline-latency 4 -n 9 -dit -name hf_9 -include hf.h */ /* * This function contains 96 FP additions, 72 FP multiplications, * (or, 60 additions, 36 multiplications, 36 fused multiply/add), * 41 stack variables, 8 constants, and 36 memory accesses */ #include "hf.h" static void hf_9(R *cr, R *ci, const R *W, stride rs, INT mb, INT me, INT ms) { DK(KP642787609, +0.642787609686539326322643409907263432907559884); DK(KP766044443, +0.766044443118978035202392650555416673935832457); DK(KP939692620, +0.939692620785908384054109277324731469936208134); DK(KP342020143, +0.342020143325668733044099614682259580763083368); DK(KP984807753, +0.984807753012208059366743024589523013670643252); DK(KP173648177, +0.173648177666930348851716626769314796000375677); DK(KP500000000, +0.500000000000000000000000000000000000000000000); DK(KP866025403, +0.866025403784438646763723170752936183471402627); { INT m; for (m = mb, W = W + ((mb - 1) * 16); m < me; m = m + 1, cr = cr + ms, ci = ci - ms, W = W + 16, MAKE_VOLATILE_STRIDE(18, rs)) { E T1, T1B, TQ, T1A, Tc, TN, T1C, T1D, TL, T1x, T19, T1o, T1c, T1n, Tu; E T1w, TW, T1k, T11, T1l; { E T6, TO, Tb, TP; T1 = cr[0]; T1B = ci[0]; { E T3, T5, T2, T4; T3 = cr[WS(rs, 3)]; T5 = ci[WS(rs, 3)]; T2 = W[4]; T4 = W[5]; T6 = FMA(T2, T3, T4 * T5); TO = FNMS(T4, T3, T2 * T5); } { E T8, Ta, T7, T9; T8 = cr[WS(rs, 6)]; Ta = ci[WS(rs, 6)]; T7 = W[10]; T9 = W[11]; Tb = FMA(T7, T8, T9 * Ta); TP = FNMS(T9, T8, T7 * Ta); } TQ = KP866025403 * (TO - TP); T1A = KP866025403 * (Tb - T6); Tc = T6 + Tb; TN = FNMS(KP500000000, Tc, T1); T1C = TO + TP; T1D = FNMS(KP500000000, T1C, T1B); } { E Tz, T13, TE, T14, TJ, T15, TK, T16; { E Tw, Ty, Tv, Tx; Tw = cr[WS(rs, 2)]; Ty = ci[WS(rs, 2)]; Tv = W[2]; Tx = W[3]; Tz = FMA(Tv, Tw, Tx * Ty); T13 = FNMS(Tx, Tw, Tv * Ty); } { E TB, TD, TA, TC; TB = cr[WS(rs, 5)]; TD = ci[WS(rs, 5)]; TA = W[8]; TC = W[9]; TE = FMA(TA, TB, TC * TD); T14 = FNMS(TC, TB, TA * TD); } { E TG, TI, TF, TH; TG = cr[WS(rs, 8)]; TI = ci[WS(rs, 8)]; TF = W[14]; TH = W[15]; TJ = FMA(TF, TG, TH * TI); T15 = FNMS(TH, TG, TF * TI); } TK = TE + TJ; T16 = T14 + T15; TL = Tz + TK; T1x = T13 + T16; { E T17, T18, T1a, T1b; T17 = FNMS(KP500000000, T16, T13); T18 = KP866025403 * (TJ - TE); T19 = T17 - T18; T1o = T18 + T17; T1a = FNMS(KP500000000, TK, Tz); T1b = KP866025403 * (T14 - T15); T1c = T1a - T1b; T1n = T1a + T1b; } } { E Ti, TX, Tn, TT, Ts, TU, Tt, TY; { E Tf, Th, Te, Tg; Tf = cr[WS(rs, 1)]; Th = ci[WS(rs, 1)]; Te = W[0]; Tg = W[1]; Ti = FMA(Te, Tf, Tg * Th); TX = FNMS(Tg, Tf, Te * Th); } { E Tk, Tm, Tj, Tl; Tk = cr[WS(rs, 4)]; Tm = ci[WS(rs, 4)]; Tj = W[6]; Tl = W[7]; Tn = FMA(Tj, Tk, Tl * Tm); TT = FNMS(Tl, Tk, Tj * Tm); } { E Tp, Tr, To, Tq; Tp = cr[WS(rs, 7)]; Tr = ci[WS(rs, 7)]; To = W[12]; Tq = W[13]; Ts = FMA(To, Tp, Tq * Tr); TU = FNMS(Tq, Tp, To * Tr); } Tt = Tn + Ts; TY = TT + TU; Tu = Ti + Tt; T1w = TX + TY; { E TS, TV, TZ, T10; TS = FNMS(KP500000000, Tt, Ti); TV = KP866025403 * (TT - TU); TW = TS - TV; T1k = TS + TV; TZ = FNMS(KP500000000, TY, TX); T10 = KP866025403 * (Ts - Tn); T11 = TZ - T10; T1l = T10 + TZ; } } { E T1y, Td, TM, T1v; T1y = KP866025403 * (T1w - T1x); Td = T1 + Tc; TM = Tu + TL; T1v = FNMS(KP500000000, TM, Td); cr[0] = Td + TM; cr[WS(rs, 3)] = T1v + T1y; ci[WS(rs, 2)] = T1v - T1y; } { E TR, T1I, T1e, T1K, T1i, T1H, T1f, T1J; TR = TN - TQ; T1I = T1D - T1A; { E T12, T1d, T1g, T1h; T12 = FMA(KP173648177, TW, KP984807753 * T11); T1d = FNMS(KP939692620, T1c, KP342020143 * T19); T1e = T12 + T1d; T1K = KP866025403 * (T1d - T12); T1g = FNMS(KP984807753, TW, KP173648177 * T11); T1h = FMA(KP342020143, T1c, KP939692620 * T19); T1i = KP866025403 * (T1g + T1h); T1H = T1g - T1h; } cr[WS(rs, 2)] = TR + T1e; ci[WS(rs, 6)] = T1H + T1I; T1f = FNMS(KP500000000, T1e, TR); ci[0] = T1f - T1i; ci[WS(rs, 3)] = T1f + T1i; T1J = FMS(KP500000000, T1H, T1I); cr[WS(rs, 5)] = T1J - T1K; cr[WS(rs, 8)] = T1K + T1J; } { E T1L, T1M, T1N, T1O; T1L = KP866025403 * (TL - Tu); T1M = T1C + T1B; T1N = T1w + T1x; T1O = FNMS(KP500000000, T1N, T1M); cr[WS(rs, 6)] = T1L - T1O; ci[WS(rs, 8)] = T1N + T1M; ci[WS(rs, 5)] = T1L + T1O; } { E T1j, T1E, T1q, T1z, T1u, T1F, T1r, T1G; T1j = TN + TQ; T1E = T1A + T1D; { E T1m, T1p, T1s, T1t; T1m = FMA(KP766044443, T1k, KP642787609 * T1l); T1p = FMA(KP173648177, T1n, KP984807753 * T1o); T1q = T1m + T1p; T1z = KP866025403 * (T1p - T1m); T1s = FNMS(KP642787609, T1k, KP766044443 * T1l); T1t = FNMS(KP984807753, T1n, KP173648177 * T1o); T1u = KP866025403 * (T1s - T1t); T1F = T1s + T1t; } cr[WS(rs, 1)] = T1j + T1q; T1r = FNMS(KP500000000, T1q, T1j); ci[WS(rs, 1)] = T1r - T1u; cr[WS(rs, 4)] = T1r + T1u; ci[WS(rs, 7)] = T1F + T1E; T1G = FNMS(KP500000000, T1F, T1E); cr[WS(rs, 7)] = T1z - T1G; ci[WS(rs, 4)] = T1z + T1G; } } } } static const tw_instr twinstr[] = { {TW_FULL, 1, 9}, {TW_NEXT, 1, 0} }; static const hc2hc_desc desc = { 9, "hf_9", twinstr, &GENUS, {60, 36, 36, 0} }; void X(codelet_hf_9) (planner *p) { X(khc2hc_register) (p, hf_9, &desc); } #endif /* HAVE_FMA */ fftw-3.3.3/rdft/scalar/r2cf/hf_10.c0000644000175000017500000003311412054410630013505 00000000000000/* * Copyright (c) 2003, 2007-11 Matteo Frigo * Copyright (c) 2003, 2007-11 Massachusetts Institute of Technology * * This program is free software; you can redistribute it and/or modify * it under the terms of the GNU General Public License as published by * the Free Software Foundation; either version 2 of the License, or * (at your option) any later version. * * This program is distributed in the hope that it will be useful, * but WITHOUT ANY WARRANTY; without even the implied warranty of * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the * GNU General Public License for more details. * * You should have received a copy of the GNU General Public License * along with this program; if not, write to the Free Software * Foundation, Inc., 51 Franklin Street, Fifth Floor, Boston, MA 02110-1301 USA * */ /* This file was automatically generated --- DO NOT EDIT */ /* Generated on Sun Nov 25 07:39:52 EST 2012 */ #include "codelet-rdft.h" #ifdef HAVE_FMA /* Generated by: ../../../genfft/gen_hc2hc.native -fma -reorder-insns -schedule-for-pipeline -compact -variables 4 -pipeline-latency 4 -n 10 -dit -name hf_10 -include hf.h */ /* * This function contains 102 FP additions, 72 FP multiplications, * (or, 48 additions, 18 multiplications, 54 fused multiply/add), * 72 stack variables, 4 constants, and 40 memory accesses */ #include "hf.h" static void hf_10(R *cr, R *ci, const R *W, stride rs, INT mb, INT me, INT ms) { DK(KP951056516, +0.951056516295153572116439333379382143405698634); DK(KP559016994, +0.559016994374947424102293417182819058860154590); DK(KP250000000, +0.250000000000000000000000000000000000000000000); DK(KP618033988, +0.618033988749894848204586834365638117720309180); { INT m; for (m = mb, W = W + ((mb - 1) * 18); m < me; m = m + 1, cr = cr + ms, ci = ci - ms, W = W + 18, MAKE_VOLATILE_STRIDE(20, rs)) { E T29, T2d, T2c, T2e; { E T23, T1U, T8, T12, T1y, T1P, T25, T1H, T2b, T18, T10, T1Y, T1I, Tl, T13; E T1J, Ty, T14, T1n, T1O, T24, T1K; { E T1, T1R, T3, T6, T2, T5; T1 = cr[0]; T1R = ci[0]; T3 = cr[WS(rs, 5)]; T6 = ci[WS(rs, 5)]; T2 = W[8]; T5 = W[9]; { E T1p, TY, T1x, T1F, TM, T16, T1r, TS; { E TF, T1w, TO, TR, T1u, TL, TN, TQ, T1q, TP; { E TU, TX, TT, TW; { E TB, TE, T1S, T4, TA, TD; TB = cr[WS(rs, 4)]; TE = ci[WS(rs, 4)]; T1S = T2 * T6; T4 = T2 * T3; TA = W[6]; TD = W[7]; { E T1T, T7, T1v, TC; T1T = FNMS(T5, T3, T1S); T7 = FMA(T5, T6, T4); T1v = TA * TE; TC = TA * TB; T23 = T1T + T1R; T1U = T1R - T1T; T8 = T1 - T7; T12 = T1 + T7; TF = FMA(TD, TE, TC); T1w = FNMS(TD, TB, T1v); } } TU = cr[WS(rs, 1)]; TX = ci[WS(rs, 1)]; TT = W[0]; TW = W[1]; { E TH, TK, TJ, T1t, TI, T1o, TV, TG; TH = cr[WS(rs, 9)]; TK = ci[WS(rs, 9)]; T1o = TT * TX; TV = TT * TU; TG = W[16]; TJ = W[17]; T1p = FNMS(TW, TU, T1o); TY = FMA(TW, TX, TV); T1t = TG * TK; TI = TG * TH; TO = cr[WS(rs, 6)]; TR = ci[WS(rs, 6)]; T1u = FNMS(TJ, TH, T1t); TL = FMA(TJ, TK, TI); TN = W[10]; TQ = W[11]; } } T1x = T1u - T1w; T1F = T1w + T1u; TM = TF - TL; T16 = TF + TL; T1q = TN * TR; TP = TN * TO; T1r = FNMS(TQ, TO, T1q); TS = FMA(TQ, TR, TP); } { E T1l, Te, T1e, Tx, Tn, Tq, Tp, T1j, Tk, T1f, To; { E Tt, Tw, Tv, T1d, Tu; { E Ta, Td, T9, Tc, T1k, Tb, Ts; Ta = cr[WS(rs, 2)]; Td = ci[WS(rs, 2)]; { E T1G, T1s, TZ, T17; T1G = T1r + T1p; T1s = T1p - T1r; TZ = TS - TY; T17 = TS + TY; T1y = T1s - T1x; T1P = T1x + T1s; T25 = T1F + T1G; T1H = T1F - T1G; T2b = T16 - T17; T18 = T16 + T17; T10 = TM + TZ; T1Y = TZ - TM; T9 = W[2]; } Tc = W[3]; Tt = cr[WS(rs, 3)]; Tw = ci[WS(rs, 3)]; T1k = T9 * Td; Tb = T9 * Ta; Ts = W[4]; Tv = W[5]; T1l = FNMS(Tc, Ta, T1k); Te = FMA(Tc, Td, Tb); T1d = Ts * Tw; Tu = Ts * Tt; } { E Tg, Tj, Tf, Ti, T1i, Th, Tm; Tg = cr[WS(rs, 7)]; Tj = ci[WS(rs, 7)]; T1e = FNMS(Tv, Tt, T1d); Tx = FMA(Tv, Tw, Tu); Tf = W[12]; Ti = W[13]; Tn = cr[WS(rs, 8)]; Tq = ci[WS(rs, 8)]; T1i = Tf * Tj; Th = Tf * Tg; Tm = W[14]; Tp = W[15]; T1j = FNMS(Ti, Tg, T1i); Tk = FMA(Ti, Tj, Th); T1f = Tm * Tq; To = Tm * Tn; } } { E T1m, T1g, Tr, T1h; T1m = T1j - T1l; T1I = T1l + T1j; Tl = Te - Tk; T13 = Te + Tk; T1g = FNMS(Tp, Tn, T1f); Tr = FMA(Tp, Tq, To); T1J = T1g + T1e; T1h = T1e - T1g; Ty = Tr - Tx; T14 = Tr + Tx; T1n = T1h - T1m; T1O = T1m + T1h; } } } } T24 = T1I + T1J; T1K = T1I - T1J; { E T2a, T15, Tz, T1Z; T2a = T13 - T14; T15 = T13 + T14; Tz = Tl + Ty; T1Z = Ty - Tl; { E T1L, T1N, T1E, T1M; { E T19, T1D, T1C, T11, T1b; T19 = T15 + T18; T1D = T15 - T18; T11 = Tz + T10; T1b = Tz - T10; { E T1B, T1z, T1a, T1A, T1c; T1B = FNMS(KP618033988, T1n, T1y); T1z = FMA(KP618033988, T1y, T1n); ci[WS(rs, 4)] = T8 + T11; T1a = FNMS(KP250000000, T11, T8); T1A = FNMS(KP559016994, T1b, T1a); T1c = FMA(KP559016994, T1b, T1a); T1C = FNMS(KP250000000, T19, T12); T1L = FNMS(KP618033988, T1K, T1H); T1N = FMA(KP618033988, T1H, T1K); cr[WS(rs, 1)] = FMA(KP951056516, T1z, T1c); ci[0] = FNMS(KP951056516, T1z, T1c); cr[WS(rs, 3)] = FMA(KP951056516, T1B, T1A); ci[WS(rs, 2)] = FNMS(KP951056516, T1B, T1A); } cr[0] = T12 + T19; T1E = FNMS(KP559016994, T1D, T1C); T1M = FMA(KP559016994, T1D, T1C); } { E T1X, T21, T20, T22, T1Q, T1W, T1V, T26, T28, T27; T1Q = T1O + T1P; T1W = T1P - T1O; ci[WS(rs, 3)] = FMA(KP951056516, T1N, T1M); cr[WS(rs, 4)] = FNMS(KP951056516, T1N, T1M); ci[WS(rs, 1)] = FMA(KP951056516, T1L, T1E); cr[WS(rs, 2)] = FNMS(KP951056516, T1L, T1E); T1V = FMA(KP250000000, T1Q, T1U); cr[WS(rs, 5)] = T1Q - T1U; T1X = FNMS(KP559016994, T1W, T1V); T21 = FMA(KP559016994, T1W, T1V); T20 = FNMS(KP618033988, T1Z, T1Y); T22 = FMA(KP618033988, T1Y, T1Z); T26 = T24 + T25; T28 = T24 - T25; ci[WS(rs, 8)] = FMA(KP951056516, T22, T21); cr[WS(rs, 9)] = FMS(KP951056516, T22, T21); ci[WS(rs, 6)] = FMA(KP951056516, T20, T1X); cr[WS(rs, 7)] = FMS(KP951056516, T20, T1X); T27 = FNMS(KP250000000, T26, T23); ci[WS(rs, 9)] = T26 + T23; T29 = FMA(KP559016994, T28, T27); T2d = FNMS(KP559016994, T28, T27); T2c = FMA(KP618033988, T2b, T2a); T2e = FNMS(KP618033988, T2a, T2b); } } } } ci[WS(rs, 7)] = FMA(KP951056516, T2e, T2d); cr[WS(rs, 8)] = FMS(KP951056516, T2e, T2d); ci[WS(rs, 5)] = FMA(KP951056516, T2c, T29); cr[WS(rs, 6)] = FMS(KP951056516, T2c, T29); } } } static const tw_instr twinstr[] = { {TW_FULL, 1, 10}, {TW_NEXT, 1, 0} }; static const hc2hc_desc desc = { 10, "hf_10", twinstr, &GENUS, {48, 18, 54, 0} }; void X(codelet_hf_10) (planner *p) { X(khc2hc_register) (p, hf_10, &desc); } #else /* HAVE_FMA */ /* Generated by: ../../../genfft/gen_hc2hc.native -compact -variables 4 -pipeline-latency 4 -n 10 -dit -name hf_10 -include hf.h */ /* * This function contains 102 FP additions, 60 FP multiplications, * (or, 72 additions, 30 multiplications, 30 fused multiply/add), * 45 stack variables, 4 constants, and 40 memory accesses */ #include "hf.h" static void hf_10(R *cr, R *ci, const R *W, stride rs, INT mb, INT me, INT ms) { DK(KP587785252, +0.587785252292473129168705954639072768597652438); DK(KP951056516, +0.951056516295153572116439333379382143405698634); DK(KP250000000, +0.250000000000000000000000000000000000000000000); DK(KP559016994, +0.559016994374947424102293417182819058860154590); { INT m; for (m = mb, W = W + ((mb - 1) * 18); m < me; m = m + 1, cr = cr + ms, ci = ci - ms, W = W + 18, MAKE_VOLATILE_STRIDE(20, rs)) { E T7, T1R, TT, T1C, TF, TQ, TR, T1o, T1p, T1P, TX, TY, TZ, T1d, T1g; E T1x, Ti, Tt, Tu, T1r, T1s, T1O, TU, TV, TW, T16, T19, T1y; { E T1, T1A, T6, T1B; T1 = cr[0]; T1A = ci[0]; { E T3, T5, T2, T4; T3 = cr[WS(rs, 5)]; T5 = ci[WS(rs, 5)]; T2 = W[8]; T4 = W[9]; T6 = FMA(T2, T3, T4 * T5); T1B = FNMS(T4, T3, T2 * T5); } T7 = T1 - T6; T1R = T1B + T1A; TT = T1 + T6; T1C = T1A - T1B; } { E Tz, T1b, TP, T1e, TE, T1c, TK, T1f; { E Tw, Ty, Tv, Tx; Tw = cr[WS(rs, 4)]; Ty = ci[WS(rs, 4)]; Tv = W[6]; Tx = W[7]; Tz = FMA(Tv, Tw, Tx * Ty); T1b = FNMS(Tx, Tw, Tv * Ty); } { E TM, TO, TL, TN; TM = cr[WS(rs, 1)]; TO = ci[WS(rs, 1)]; TL = W[0]; TN = W[1]; TP = FMA(TL, TM, TN * TO); T1e = FNMS(TN, TM, TL * TO); } { E TB, TD, TA, TC; TB = cr[WS(rs, 9)]; TD = ci[WS(rs, 9)]; TA = W[16]; TC = W[17]; TE = FMA(TA, TB, TC * TD); T1c = FNMS(TC, TB, TA * TD); } { E TH, TJ, TG, TI; TH = cr[WS(rs, 6)]; TJ = ci[WS(rs, 6)]; TG = W[10]; TI = W[11]; TK = FMA(TG, TH, TI * TJ); T1f = FNMS(TI, TH, TG * TJ); } TF = Tz - TE; TQ = TK - TP; TR = TF + TQ; T1o = T1b + T1c; T1p = T1f + T1e; T1P = T1o + T1p; TX = Tz + TE; TY = TK + TP; TZ = TX + TY; T1d = T1b - T1c; T1g = T1e - T1f; T1x = T1g - T1d; } { E Tc, T14, Ts, T18, Th, T15, Tn, T17; { E T9, Tb, T8, Ta; T9 = cr[WS(rs, 2)]; Tb = ci[WS(rs, 2)]; T8 = W[2]; Ta = W[3]; Tc = FMA(T8, T9, Ta * Tb); T14 = FNMS(Ta, T9, T8 * Tb); } { E Tp, Tr, To, Tq; Tp = cr[WS(rs, 3)]; Tr = ci[WS(rs, 3)]; To = W[4]; Tq = W[5]; Ts = FMA(To, Tp, Tq * Tr); T18 = FNMS(Tq, Tp, To * Tr); } { E Te, Tg, Td, Tf; Te = cr[WS(rs, 7)]; Tg = ci[WS(rs, 7)]; Td = W[12]; Tf = W[13]; Th = FMA(Td, Te, Tf * Tg); T15 = FNMS(Tf, Te, Td * Tg); } { E Tk, Tm, Tj, Tl; Tk = cr[WS(rs, 8)]; Tm = ci[WS(rs, 8)]; Tj = W[14]; Tl = W[15]; Tn = FMA(Tj, Tk, Tl * Tm); T17 = FNMS(Tl, Tk, Tj * Tm); } Ti = Tc - Th; Tt = Tn - Ts; Tu = Ti + Tt; T1r = T14 + T15; T1s = T17 + T18; T1O = T1r + T1s; TU = Tc + Th; TV = Tn + Ts; TW = TU + TV; T16 = T14 - T15; T19 = T17 - T18; T1y = T16 + T19; } { E T11, TS, T12, T1i, T1k, T1a, T1h, T1j, T13; T11 = KP559016994 * (Tu - TR); TS = Tu + TR; T12 = FNMS(KP250000000, TS, T7); T1a = T16 - T19; T1h = T1d + T1g; T1i = FMA(KP951056516, T1a, KP587785252 * T1h); T1k = FNMS(KP587785252, T1a, KP951056516 * T1h); ci[WS(rs, 4)] = T7 + TS; T1j = T12 - T11; ci[WS(rs, 2)] = T1j - T1k; cr[WS(rs, 3)] = T1j + T1k; T13 = T11 + T12; ci[0] = T13 - T1i; cr[WS(rs, 1)] = T13 + T1i; } { E T1m, T10, T1l, T1u, T1w, T1q, T1t, T1v, T1n; T1m = KP559016994 * (TW - TZ); T10 = TW + TZ; T1l = FNMS(KP250000000, T10, TT); T1q = T1o - T1p; T1t = T1r - T1s; T1u = FNMS(KP587785252, T1t, KP951056516 * T1q); T1w = FMA(KP951056516, T1t, KP587785252 * T1q); cr[0] = TT + T10; T1v = T1m + T1l; cr[WS(rs, 4)] = T1v - T1w; ci[WS(rs, 3)] = T1v + T1w; T1n = T1l - T1m; cr[WS(rs, 2)] = T1n - T1u; ci[WS(rs, 1)] = T1n + T1u; } { E T1H, T1z, T1G, T1F, T1J, T1D, T1E, T1K, T1I; T1H = KP559016994 * (T1y + T1x); T1z = T1x - T1y; T1G = FMA(KP250000000, T1z, T1C); T1D = Ti - Tt; T1E = TQ - TF; T1F = FMA(KP587785252, T1D, KP951056516 * T1E); T1J = FNMS(KP951056516, T1D, KP587785252 * T1E); cr[WS(rs, 5)] = T1z - T1C; T1K = T1H + T1G; cr[WS(rs, 9)] = T1J - T1K; ci[WS(rs, 8)] = T1J + T1K; T1I = T1G - T1H; cr[WS(rs, 7)] = T1F - T1I; ci[WS(rs, 6)] = T1F + T1I; } { E T1Q, T1S, T1T, T1N, T1V, T1L, T1M, T1W, T1U; T1Q = KP559016994 * (T1O - T1P); T1S = T1O + T1P; T1T = FNMS(KP250000000, T1S, T1R); T1L = TU - TV; T1M = TX - TY; T1N = FMA(KP951056516, T1L, KP587785252 * T1M); T1V = FNMS(KP587785252, T1L, KP951056516 * T1M); ci[WS(rs, 9)] = T1S + T1R; T1W = T1T - T1Q; cr[WS(rs, 8)] = T1V - T1W; ci[WS(rs, 7)] = T1V + T1W; T1U = T1Q + T1T; cr[WS(rs, 6)] = T1N - T1U; ci[WS(rs, 5)] = T1N + T1U; } } } } static const tw_instr twinstr[] = { {TW_FULL, 1, 10}, {TW_NEXT, 1, 0} }; static const hc2hc_desc desc = { 10, "hf_10", twinstr, &GENUS, {72, 30, 30, 0} }; void X(codelet_hf_10) (planner *p) { X(khc2hc_register) (p, hf_10, &desc); } #endif /* HAVE_FMA */ fftw-3.3.3/rdft/scalar/r2cf/hf_12.c0000644000175000017500000003506012054410630013511 00000000000000/* * Copyright (c) 2003, 2007-11 Matteo Frigo * Copyright (c) 2003, 2007-11 Massachusetts Institute of Technology * * This program is free software; you can redistribute it and/or modify * it under the terms of the GNU General Public License as published by * the Free Software Foundation; either version 2 of the License, or * (at your option) any later version. * * This program is distributed in the hope that it will be useful, * but WITHOUT ANY WARRANTY; without even the implied warranty of * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the * GNU General Public License for more details. * * You should have received a copy of the GNU General Public License * along with this program; if not, write to the Free Software * Foundation, Inc., 51 Franklin Street, Fifth Floor, Boston, MA 02110-1301 USA * */ /* This file was automatically generated --- DO NOT EDIT */ /* Generated on Sun Nov 25 07:39:52 EST 2012 */ #include "codelet-rdft.h" #ifdef HAVE_FMA /* Generated by: ../../../genfft/gen_hc2hc.native -fma -reorder-insns -schedule-for-pipeline -compact -variables 4 -pipeline-latency 4 -n 12 -dit -name hf_12 -include hf.h */ /* * This function contains 118 FP additions, 68 FP multiplications, * (or, 72 additions, 22 multiplications, 46 fused multiply/add), * 84 stack variables, 2 constants, and 48 memory accesses */ #include "hf.h" static void hf_12(R *cr, R *ci, const R *W, stride rs, INT mb, INT me, INT ms) { DK(KP866025403, +0.866025403784438646763723170752936183471402627); DK(KP500000000, +0.500000000000000000000000000000000000000000000); { INT m; for (m = mb, W = W + ((mb - 1) * 22); m < me; m = m + 1, cr = cr + ms, ci = ci - ms, W = W + 22, MAKE_VOLATILE_STRIDE(24, rs)) { E T2u, T2n; { E T1, T2i, T2e, Tl, T1Y, T10, T1S, TG, T2f, T1s, T2s, Ty, T1Z, T1H, T21; E T1d, TI, TL, T2h, T1l, T2p, Te, TJ, T1w, TO, TR, TN, TK, TQ; { E TW, TZ, TY, T1X, TX; T1 = cr[0]; T2i = ci[0]; { E Th, Tk, Tg, Tj, T2d, Ti, TV; Th = cr[WS(rs, 6)]; Tk = ci[WS(rs, 6)]; Tg = W[10]; Tj = W[11]; TW = cr[WS(rs, 9)]; TZ = ci[WS(rs, 9)]; T2d = Tg * Tk; Ti = Tg * Th; TV = W[16]; TY = W[17]; T2e = FNMS(Tj, Th, T2d); Tl = FMA(Tj, Tk, Ti); T1X = TV * TZ; TX = TV * TW; } { E Tn, Tq, Tt, T1o, To, Tw, Ts, Tp, Tv; { E TC, TF, TB, TE, T1R, TD, Tm; TC = cr[WS(rs, 3)]; TF = ci[WS(rs, 3)]; T1Y = FNMS(TY, TW, T1X); T10 = FMA(TY, TZ, TX); TB = W[4]; TE = W[5]; Tn = cr[WS(rs, 10)]; Tq = ci[WS(rs, 10)]; T1R = TB * TF; TD = TB * TC; Tm = W[18]; Tt = cr[WS(rs, 2)]; T1S = FNMS(TE, TC, T1R); TG = FMA(TE, TF, TD); T1o = Tm * Tq; To = Tm * Tn; Tw = ci[WS(rs, 2)]; Ts = W[2]; Tp = W[19]; Tv = W[3]; } { E T12, T15, T13, T1D, T18, T1b, T17, T14, T1a; { E T1p, Tr, T1r, Tx, T1q, Tu, T11; T12 = cr[WS(rs, 1)]; T1q = Ts * Tw; Tu = Ts * Tt; T1p = FNMS(Tp, Tn, T1o); Tr = FMA(Tp, Tq, To); T1r = FNMS(Tv, Tt, T1q); Tx = FMA(Tv, Tw, Tu); T15 = ci[WS(rs, 1)]; T11 = W[0]; T2f = T1p + T1r; T1s = T1p - T1r; T2s = Tx - Tr; Ty = Tr + Tx; T13 = T11 * T12; T1D = T11 * T15; } T18 = cr[WS(rs, 5)]; T1b = ci[WS(rs, 5)]; T17 = W[8]; T14 = W[1]; T1a = W[9]; { E T3, T6, T4, T1h, T9, Tc, T8, T5, Tb; { E T1E, T16, T1G, T1c, T1F, T19, T2; T3 = cr[WS(rs, 4)]; T1F = T17 * T1b; T19 = T17 * T18; T1E = FNMS(T14, T12, T1D); T16 = FMA(T14, T15, T13); T1G = FNMS(T1a, T18, T1F); T1c = FMA(T1a, T1b, T19); T6 = ci[WS(rs, 4)]; T2 = W[6]; T1Z = T1E + T1G; T1H = T1E - T1G; T21 = T1c - T16; T1d = T16 + T1c; T4 = T2 * T3; T1h = T2 * T6; } T9 = cr[WS(rs, 8)]; Tc = ci[WS(rs, 8)]; T8 = W[14]; T5 = W[7]; Tb = W[15]; { E T1i, T7, T1k, Td, T1j, Ta, TH; TI = cr[WS(rs, 7)]; T1j = T8 * Tc; Ta = T8 * T9; T1i = FNMS(T5, T3, T1h); T7 = FMA(T5, T6, T4); T1k = FNMS(Tb, T9, T1j); Td = FMA(Tb, Tc, Ta); TL = ci[WS(rs, 7)]; TH = W[12]; T2h = T1i + T1k; T1l = T1i - T1k; T2p = Td - T7; Te = T7 + Td; TJ = TH * TI; T1w = TH * TL; } TO = cr[WS(rs, 11)]; TR = ci[WS(rs, 11)]; TN = W[20]; TK = W[13]; TQ = W[21]; } } } } { E T1g, T1n, T2r, T1A, T1V, T28, TA, T2o, T1v, T1C, T1U, T29, T2m, T2k, T2l; E T1f, T2a, T20; { E T2g, T1T, TT, T2j, TU, T1e; { E Tf, T1x, TM, T1z, TS, Tz, T1y, TP; T1g = FNMS(KP500000000, Te, T1); Tf = T1 + Te; T1y = TN * TR; TP = TN * TO; T1x = FNMS(TK, TI, T1w); TM = FMA(TK, TL, TJ); T1z = FNMS(TQ, TO, T1y); TS = FMA(TQ, TR, TP); Tz = Tl + Ty; T1n = FNMS(KP500000000, Ty, Tl); T2r = FNMS(KP500000000, T2f, T2e); T2g = T2e + T2f; T1T = T1x + T1z; T1A = T1x - T1z; T1V = TS - TM; TT = TM + TS; T28 = Tf - Tz; TA = Tf + Tz; T2j = T2h + T2i; T2o = FNMS(KP500000000, T2h, T2i); } T1v = FNMS(KP500000000, TT, TG); TU = TG + TT; T1e = T10 + T1d; T1C = FNMS(KP500000000, T1d, T10); T1U = FNMS(KP500000000, T1T, T1S); T29 = T1S + T1T; T2m = T2j - T2g; T2k = T2g + T2j; T2l = TU - T1e; T1f = TU + T1e; T2a = T1Y + T1Z; T20 = FNMS(KP500000000, T1Z, T1Y); } { E T1m, T1K, T2y, T2q, T2z, T2t, T1L, T1t, T1B, T1N, T2c, T2b; ci[WS(rs, 8)] = T2l + T2m; cr[WS(rs, 9)] = T2l - T2m; cr[0] = TA + T1f; ci[WS(rs, 5)] = TA - T1f; T2c = T29 + T2a; T2b = T29 - T2a; T1m = FNMS(KP866025403, T1l, T1g); T1K = FMA(KP866025403, T1l, T1g); ci[WS(rs, 11)] = T2c + T2k; cr[WS(rs, 6)] = T2c - T2k; ci[WS(rs, 2)] = T28 + T2b; cr[WS(rs, 3)] = T28 - T2b; T2y = FMA(KP866025403, T2p, T2o); T2q = FNMS(KP866025403, T2p, T2o); T2z = FMA(KP866025403, T2s, T2r); T2t = FNMS(KP866025403, T2s, T2r); T1L = FMA(KP866025403, T1s, T1n); T1t = FNMS(KP866025403, T1s, T1n); T1B = FNMS(KP866025403, T1A, T1v); T1N = FMA(KP866025403, T1A, T1v); { E T1Q, T23, T27, T2A, T1P, T2x, T24, T1M; { E T1u, T25, T26, T1O, T1I, T2w, T2v, T1W, T22, T2B, T1J, T2C; T1Q = T1m - T1t; T1u = T1m + T1t; T25 = FMA(KP866025403, T1V, T1U); T1W = FNMS(KP866025403, T1V, T1U); T26 = FMA(KP866025403, T21, T20); T22 = FNMS(KP866025403, T21, T20); T1O = FMA(KP866025403, T1H, T1C); T1I = FNMS(KP866025403, T1H, T1C); T2w = T2t + T2q; T2u = T2q - T2t; T23 = T1W - T22; T2v = T1W + T22; T2B = T25 + T26; T27 = T25 - T26; T2n = T1I - T1B; T1J = T1B + T1I; T2C = T2z + T2y; T2A = T2y - T2z; ci[WS(rs, 9)] = T2w - T2v; cr[WS(rs, 8)] = -(T2v + T2w); ci[WS(rs, 3)] = T1u + T1J; cr[WS(rs, 2)] = T1u - T1J; cr[WS(rs, 10)] = T2B - T2C; ci[WS(rs, 7)] = T2B + T2C; T1P = T1N + T1O; T2x = T1O - T1N; } T24 = T1K - T1L; T1M = T1K + T1L; ci[WS(rs, 10)] = T2x + T2A; cr[WS(rs, 7)] = T2x - T2A; cr[WS(rs, 4)] = T1M + T1P; ci[WS(rs, 1)] = T1M - T1P; cr[WS(rs, 1)] = T24 + T27; ci[WS(rs, 4)] = T24 - T27; cr[WS(rs, 5)] = T1Q + T23; ci[0] = T1Q - T23; } } } } ci[WS(rs, 6)] = T2n + T2u; cr[WS(rs, 11)] = T2n - T2u; } } } static const tw_instr twinstr[] = { {TW_FULL, 1, 12}, {TW_NEXT, 1, 0} }; static const hc2hc_desc desc = { 12, "hf_12", twinstr, &GENUS, {72, 22, 46, 0} }; void X(codelet_hf_12) (planner *p) { X(khc2hc_register) (p, hf_12, &desc); } #else /* HAVE_FMA */ /* Generated by: ../../../genfft/gen_hc2hc.native -compact -variables 4 -pipeline-latency 4 -n 12 -dit -name hf_12 -include hf.h */ /* * This function contains 118 FP additions, 60 FP multiplications, * (or, 88 additions, 30 multiplications, 30 fused multiply/add), * 47 stack variables, 2 constants, and 48 memory accesses */ #include "hf.h" static void hf_12(R *cr, R *ci, const R *W, stride rs, INT mb, INT me, INT ms) { DK(KP500000000, +0.500000000000000000000000000000000000000000000); DK(KP866025403, +0.866025403784438646763723170752936183471402627); { INT m; for (m = mb, W = W + ((mb - 1) * 22); m < me; m = m + 1, cr = cr + ms, ci = ci - ms, W = W + 22, MAKE_VOLATILE_STRIDE(24, rs)) { E T1, T1W, T18, T23, Tc, T15, T1V, T22, TR, T1E, T1o, T1D, T12, T1l, T1F; E T1G, Ti, T1S, T1d, T26, Tt, T1a, T1T, T25, TA, T1y, T1j, T1B, TL, T1g; E T1z, T1A; { E T6, T16, Tb, T17; T1 = cr[0]; T1W = ci[0]; { E T3, T5, T2, T4; T3 = cr[WS(rs, 4)]; T5 = ci[WS(rs, 4)]; T2 = W[6]; T4 = W[7]; T6 = FMA(T2, T3, T4 * T5); T16 = FNMS(T4, T3, T2 * T5); } { E T8, Ta, T7, T9; T8 = cr[WS(rs, 8)]; Ta = ci[WS(rs, 8)]; T7 = W[14]; T9 = W[15]; Tb = FMA(T7, T8, T9 * Ta); T17 = FNMS(T9, T8, T7 * Ta); } T18 = KP866025403 * (T16 - T17); T23 = KP866025403 * (Tb - T6); Tc = T6 + Tb; T15 = FNMS(KP500000000, Tc, T1); T1V = T16 + T17; T22 = FNMS(KP500000000, T1V, T1W); } { E T11, T1n, TW, T1m; { E TO, TQ, TN, TP; TO = cr[WS(rs, 9)]; TQ = ci[WS(rs, 9)]; TN = W[16]; TP = W[17]; TR = FMA(TN, TO, TP * TQ); T1E = FNMS(TP, TO, TN * TQ); } { E TY, T10, TX, TZ; TY = cr[WS(rs, 5)]; T10 = ci[WS(rs, 5)]; TX = W[8]; TZ = W[9]; T11 = FMA(TX, TY, TZ * T10); T1n = FNMS(TZ, TY, TX * T10); } { E TT, TV, TS, TU; TT = cr[WS(rs, 1)]; TV = ci[WS(rs, 1)]; TS = W[0]; TU = W[1]; TW = FMA(TS, TT, TU * TV); T1m = FNMS(TU, TT, TS * TV); } T1o = KP866025403 * (T1m - T1n); T1D = KP866025403 * (T11 - TW); T12 = TW + T11; T1l = FNMS(KP500000000, T12, TR); T1F = T1m + T1n; T1G = FNMS(KP500000000, T1F, T1E); } { E Ts, T1c, Tn, T1b; { E Tf, Th, Te, Tg; Tf = cr[WS(rs, 6)]; Th = ci[WS(rs, 6)]; Te = W[10]; Tg = W[11]; Ti = FMA(Te, Tf, Tg * Th); T1S = FNMS(Tg, Tf, Te * Th); } { E Tp, Tr, To, Tq; Tp = cr[WS(rs, 2)]; Tr = ci[WS(rs, 2)]; To = W[2]; Tq = W[3]; Ts = FMA(To, Tp, Tq * Tr); T1c = FNMS(Tq, Tp, To * Tr); } { E Tk, Tm, Tj, Tl; Tk = cr[WS(rs, 10)]; Tm = ci[WS(rs, 10)]; Tj = W[18]; Tl = W[19]; Tn = FMA(Tj, Tk, Tl * Tm); T1b = FNMS(Tl, Tk, Tj * Tm); } T1d = KP866025403 * (T1b - T1c); T26 = KP866025403 * (Ts - Tn); Tt = Tn + Ts; T1a = FNMS(KP500000000, Tt, Ti); T1T = T1b + T1c; T25 = FNMS(KP500000000, T1T, T1S); } { E TK, T1i, TF, T1h; { E Tx, Tz, Tw, Ty; Tx = cr[WS(rs, 3)]; Tz = ci[WS(rs, 3)]; Tw = W[4]; Ty = W[5]; TA = FMA(Tw, Tx, Ty * Tz); T1y = FNMS(Ty, Tx, Tw * Tz); } { E TH, TJ, TG, TI; TH = cr[WS(rs, 11)]; TJ = ci[WS(rs, 11)]; TG = W[20]; TI = W[21]; TK = FMA(TG, TH, TI * TJ); T1i = FNMS(TI, TH, TG * TJ); } { E TC, TE, TB, TD; TC = cr[WS(rs, 7)]; TE = ci[WS(rs, 7)]; TB = W[12]; TD = W[13]; TF = FMA(TB, TC, TD * TE); T1h = FNMS(TD, TC, TB * TE); } T1j = KP866025403 * (T1h - T1i); T1B = KP866025403 * (TK - TF); TL = TF + TK; T1g = FNMS(KP500000000, TL, TA); T1z = T1h + T1i; T1A = FNMS(KP500000000, T1z, T1y); } { E Tv, T1N, T1Y, T20, T14, T1Z, T1Q, T1R; { E Td, Tu, T1U, T1X; Td = T1 + Tc; Tu = Ti + Tt; Tv = Td + Tu; T1N = Td - Tu; T1U = T1S + T1T; T1X = T1V + T1W; T1Y = T1U + T1X; T20 = T1X - T1U; } { E TM, T13, T1O, T1P; TM = TA + TL; T13 = TR + T12; T14 = TM + T13; T1Z = TM - T13; T1O = T1y + T1z; T1P = T1E + T1F; T1Q = T1O - T1P; T1R = T1O + T1P; } ci[WS(rs, 5)] = Tv - T14; cr[WS(rs, 9)] = T1Z - T20; ci[WS(rs, 8)] = T1Z + T20; cr[0] = Tv + T14; cr[WS(rs, 3)] = T1N - T1Q; cr[WS(rs, 6)] = T1R - T1Y; ci[WS(rs, 11)] = T1R + T1Y; ci[WS(rs, 2)] = T1N + T1Q; } { E T1f, T1x, T28, T2a, T1q, T21, T1I, T29; { E T19, T1e, T24, T27; T19 = T15 - T18; T1e = T1a - T1d; T1f = T19 + T1e; T1x = T19 - T1e; T24 = T22 - T23; T27 = T25 - T26; T28 = T24 - T27; T2a = T27 + T24; } { E T1k, T1p, T1C, T1H; T1k = T1g - T1j; T1p = T1l - T1o; T1q = T1k + T1p; T21 = T1p - T1k; T1C = T1A - T1B; T1H = T1D - T1G; T1I = T1C + T1H; T29 = T1H - T1C; } cr[WS(rs, 2)] = T1f - T1q; cr[WS(rs, 8)] = T29 - T2a; ci[WS(rs, 9)] = T29 + T2a; ci[WS(rs, 3)] = T1f + T1q; ci[0] = T1x - T1I; cr[WS(rs, 11)] = T21 - T28; ci[WS(rs, 6)] = T21 + T28; cr[WS(rs, 5)] = T1x + T1I; } { E T1t, T1J, T2e, T2g, T1w, T2b, T1M, T2f; { E T1r, T1s, T2c, T2d; T1r = T15 + T18; T1s = T1a + T1d; T1t = T1r + T1s; T1J = T1r - T1s; T2c = T23 + T22; T2d = T26 + T25; T2e = T2c - T2d; T2g = T2d + T2c; } { E T1u, T1v, T1K, T1L; T1u = T1g + T1j; T1v = T1l + T1o; T1w = T1u + T1v; T2b = T1v - T1u; T1K = T1B + T1A; T1L = T1D + T1G; T1M = T1K - T1L; T2f = T1K + T1L; } ci[WS(rs, 1)] = T1t - T1w; cr[WS(rs, 1)] = T1J + T1M; cr[WS(rs, 4)] = T1t + T1w; ci[WS(rs, 4)] = T1J - T1M; cr[WS(rs, 7)] = T2b - T2e; ci[WS(rs, 7)] = T2f + T2g; ci[WS(rs, 10)] = T2b + T2e; cr[WS(rs, 10)] = T2f - T2g; } } } } static const tw_instr twinstr[] = { {TW_FULL, 1, 12}, {TW_NEXT, 1, 0} }; static const hc2hc_desc desc = { 12, "hf_12", twinstr, &GENUS, {88, 30, 30, 0} }; void X(codelet_hf_12) (planner *p) { X(khc2hc_register) (p, hf_12, &desc); } #endif /* HAVE_FMA */ fftw-3.3.3/rdft/scalar/r2cf/hf_15.c0000644000175000017500000005400612054410632013517 00000000000000/* * Copyright (c) 2003, 2007-11 Matteo Frigo * Copyright (c) 2003, 2007-11 Massachusetts Institute of Technology * * This program is free software; you can redistribute it and/or modify * it under the terms of the GNU General Public License as published by * the Free Software Foundation; either version 2 of the License, or * (at your option) any later version. * * This program is distributed in the hope that it will be useful, * but WITHOUT ANY WARRANTY; without even the implied warranty of * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the * GNU General Public License for more details. * * You should have received a copy of the GNU General Public License * along with this program; if not, write to the Free Software * Foundation, Inc., 51 Franklin Street, Fifth Floor, Boston, MA 02110-1301 USA * */ /* This file was automatically generated --- DO NOT EDIT */ /* Generated on Sun Nov 25 07:39:52 EST 2012 */ #include "codelet-rdft.h" #ifdef HAVE_FMA /* Generated by: ../../../genfft/gen_hc2hc.native -fma -reorder-insns -schedule-for-pipeline -compact -variables 4 -pipeline-latency 4 -n 15 -dit -name hf_15 -include hf.h */ /* * This function contains 184 FP additions, 140 FP multiplications, * (or, 72 additions, 28 multiplications, 112 fused multiply/add), * 97 stack variables, 6 constants, and 60 memory accesses */ #include "hf.h" static void hf_15(R *cr, R *ci, const R *W, stride rs, INT mb, INT me, INT ms) { DK(KP951056516, +0.951056516295153572116439333379382143405698634); DK(KP559016994, +0.559016994374947424102293417182819058860154590); DK(KP250000000, +0.250000000000000000000000000000000000000000000); DK(KP618033988, +0.618033988749894848204586834365638117720309180); DK(KP866025403, +0.866025403784438646763723170752936183471402627); DK(KP500000000, +0.500000000000000000000000000000000000000000000); { INT m; for (m = mb, W = W + ((mb - 1) * 28); m < me; m = m + 1, cr = cr + ms, ci = ci - ms, W = W + 28, MAKE_VOLATILE_STRIDE(30, rs)) { E T3v, T3E, T3G, T3A, T3y, T3z, T3F, T3B; { E T1G, T3l, T3H, T3k, T1B, Tf, T37, T1y, T2Y, T2M, T2a, T2i, T39, Tz, T2U; E T2t, T1O, T2e, T3a, TT, T10, T2V, T2z, T1V, T2f, T2C, T12, T15, T14, T21; E T1c, T1Y, T13; { E T2I, T1k, T1m, T1p, T1o, T28, T1w, T25, T1n; { E T1, T3i, T9, Tc, Tb, T1D, T7, T1E, Ta, T1j, T1i, T1h; T1 = cr[0]; T3i = ci[0]; { E T3, T6, T2, T5, T1C, T4, T8; T3 = cr[WS(rs, 5)]; T6 = ci[WS(rs, 5)]; T2 = W[8]; T5 = W[9]; T9 = cr[WS(rs, 10)]; Tc = ci[WS(rs, 10)]; T1C = T2 * T6; T4 = T2 * T3; T8 = W[18]; Tb = W[19]; T1D = FNMS(T5, T3, T1C); T7 = FMA(T5, T6, T4); T1E = T8 * Tc; Ta = T8 * T9; } { E T1g, T1F, Td, T1f, T3j, Te, T2H; T1g = cr[WS(rs, 9)]; T1j = ci[WS(rs, 9)]; T1F = FNMS(Tb, T9, T1E); Td = FMA(Tb, Tc, Ta); T1f = W[16]; T1i = W[17]; T1G = T1D - T1F; T3j = T1D + T1F; T3l = Td - T7; Te = T7 + Td; T2H = T1f * T1j; T1h = T1f * T1g; T3H = T3j + T3i; T3k = FNMS(KP500000000, T3j, T3i); T1B = FNMS(KP500000000, Te, T1); Tf = T1 + Te; T2I = FNMS(T1i, T1g, T2H); } T1k = FMA(T1i, T1j, T1h); { E T1s, T1v, T1r, T1u, T27, T1t, T1l; T1s = cr[WS(rs, 4)]; T1v = ci[WS(rs, 4)]; T1r = W[6]; T1u = W[7]; T1m = cr[WS(rs, 14)]; T1p = ci[WS(rs, 14)]; T27 = T1r * T1v; T1t = T1r * T1s; T1l = W[26]; T1o = W[27]; T28 = FNMS(T1u, T1s, T27); T1w = FMA(T1u, T1v, T1t); T25 = T1l * T1p; T1n = T1l * T1m; } } { E Tl, T2p, Tn, Tq, Tp, T1M, Tx, T1J, To; { E Th, Tk, T26, T1q, Tg, Tj; Th = cr[WS(rs, 3)]; Tk = ci[WS(rs, 3)]; T26 = FNMS(T1o, T1m, T25); T1q = FMA(T1o, T1p, T1n); Tg = W[4]; Tj = W[5]; { E T29, T2J, T1x, T2L; T29 = T26 - T28; T2J = T26 + T28; T1x = T1q + T1w; T2L = T1q - T1w; { E T2o, Ti, T2K, T24; T2o = Tg * Tk; Ti = Tg * Th; T2K = FNMS(KP500000000, T2J, T2I); T37 = T2I + T2J; T24 = FNMS(KP500000000, T1x, T1k); T1y = T1k + T1x; Tl = FMA(Tj, Tk, Ti); T2Y = FMA(KP866025403, T2L, T2K); T2M = FNMS(KP866025403, T2L, T2K); T2a = FNMS(KP866025403, T29, T24); T2i = FMA(KP866025403, T29, T24); T2p = FNMS(Tj, Th, T2o); } } } { E Tt, Tw, Ts, Tv, T1L, Tu, Tm; Tt = cr[WS(rs, 13)]; Tw = ci[WS(rs, 13)]; Ts = W[24]; Tv = W[25]; Tn = cr[WS(rs, 8)]; Tq = ci[WS(rs, 8)]; T1L = Ts * Tw; Tu = Ts * Tt; Tm = W[14]; Tp = W[15]; T1M = FNMS(Tv, Tt, T1L); Tx = FMA(Tv, Tw, Tu); T1J = Tm * Tq; To = Tm * Tn; } { E TF, T2v, TH, TK, TJ, T1T, TR, T1Q, TI; { E TB, TE, T1K, Tr, TA, TD; TB = cr[WS(rs, 12)]; TE = ci[WS(rs, 12)]; T1K = FNMS(Tp, Tn, T1J); Tr = FMA(Tp, Tq, To); TA = W[22]; TD = W[23]; { E T1N, T2q, Ty, T2s; T1N = T1K - T1M; T2q = T1K + T1M; Ty = Tr + Tx; T2s = Tr - Tx; { E T2u, TC, T2r, T1I; T2u = TA * TE; TC = TA * TB; T2r = FNMS(KP500000000, T2q, T2p); T39 = T2p + T2q; T1I = FNMS(KP500000000, Ty, Tl); Tz = Tl + Ty; TF = FMA(TD, TE, TC); T2U = FMA(KP866025403, T2s, T2r); T2t = FNMS(KP866025403, T2s, T2r); T1O = FNMS(KP866025403, T1N, T1I); T2e = FMA(KP866025403, T1N, T1I); T2v = FNMS(TD, TB, T2u); } } } { E TN, TQ, TM, TP, T1S, TO, TG; TN = cr[WS(rs, 7)]; TQ = ci[WS(rs, 7)]; TM = W[12]; TP = W[13]; TH = cr[WS(rs, 2)]; TK = ci[WS(rs, 2)]; T1S = TM * TQ; TO = TM * TN; TG = W[2]; TJ = W[3]; T1T = FNMS(TP, TN, T1S); TR = FMA(TP, TQ, TO); T1Q = TG * TK; TI = TG * TH; } { E TW, TZ, T1R, TL, TV, TY; TW = cr[WS(rs, 6)]; TZ = ci[WS(rs, 6)]; T1R = FNMS(TJ, TH, T1Q); TL = FMA(TJ, TK, TI); TV = W[10]; TY = W[11]; { E T1U, T2w, TS, T2y; T1U = T1R - T1T; T2w = T1R + T1T; TS = TL + TR; T2y = TL - TR; { E T2B, TX, T2x, T1P; T2B = TV * TZ; TX = TV * TW; T2x = FNMS(KP500000000, T2w, T2v); T3a = T2v + T2w; T1P = FNMS(KP500000000, TS, TF); TT = TF + TS; T10 = FMA(TY, TZ, TX); T2V = FMA(KP866025403, T2y, T2x); T2z = FNMS(KP866025403, T2y, T2x); T1V = FNMS(KP866025403, T1U, T1P); T2f = FMA(KP866025403, T1U, T1P); T2C = FNMS(TY, TW, T2B); } } } { E T18, T1b, T17, T1a, T20, T19, T11; T18 = cr[WS(rs, 1)]; T1b = ci[WS(rs, 1)]; T17 = W[0]; T1a = W[1]; T12 = cr[WS(rs, 11)]; T15 = ci[WS(rs, 11)]; T20 = T17 * T1b; T19 = T17 * T18; T11 = W[20]; T14 = W[21]; T21 = FNMS(T1a, T18, T20); T1c = FMA(T1a, T1b, T19); T1Y = T11 * T15; T13 = T11 * T12; } } } } { E T3I, T3O, T3w, T2d, T3J, T3P, T3x, T3C, T3D, T3f, T3g, T2Q, T2O, T3r, T3q; E T2k, T2m; { E T3b, T1Z, T16, TU; T3I = T39 + T3a; T3b = T39 - T3a; T1Z = FNMS(T14, T12, T1Y); T16 = FMA(T14, T15, T13); T3O = TT - Tz; TU = Tz + TT; { E T1H, T2G, T2h, T3e, T3c, T34, T1W, T32, T30, T33, T2b, T2S, T2R; { E T2W, T22, T1d, T2F, T2E, T36, T2D; T2W = T2U - T2V; T3w = T2U + T2V; T22 = T1Z - T21; T2D = T1Z + T21; T1d = T16 + T1c; T2F = T16 - T1c; T2E = FNMS(KP500000000, T2D, T2C); T36 = T2C + T2D; T2d = FMA(KP866025403, T1G, T1B); T1H = FNMS(KP866025403, T1G, T1B); { E T1e, T1X, T38, T2X; T1e = T10 + T1d; T1X = FNMS(KP500000000, T1d, T10); T38 = T36 - T37; T3J = T36 + T37; T2G = FNMS(KP866025403, T2F, T2E); T2X = FMA(KP866025403, T2F, T2E); { E T1z, T23, T2Z, T1A; T3P = T1y - T1e; T1z = T1e + T1y; T23 = FNMS(KP866025403, T22, T1X); T2h = FMA(KP866025403, T22, T1X); T3e = FMA(KP618033988, T38, T3b); T3c = FNMS(KP618033988, T3b, T38); T2Z = T2X - T2Y; T3x = T2X + T2Y; T1A = TU + T1z; T34 = TU - T1z; T3C = T1O - T1V; T1W = T1O + T1V; T32 = FNMS(KP618033988, T2W, T2Z); T30 = FMA(KP618033988, T2Z, T2W); cr[0] = Tf + T1A; T33 = FNMS(KP250000000, T1A, Tf); T2b = T23 + T2a; T3D = T23 - T2a; } } } { E T2A, T2N, T3d, T35, T2c; T3f = T2t + T2z; T2A = T2t - T2z; T2N = T2G - T2M; T3g = T2G + T2M; T3d = FMA(KP559016994, T34, T33); T35 = FNMS(KP559016994, T34, T33); T2c = T1W + T2b; T2S = T1W - T2b; cr[WS(rs, 3)] = FMA(KP951056516, T3c, T35); ci[WS(rs, 2)] = FNMS(KP951056516, T3c, T35); cr[WS(rs, 6)] = FMA(KP951056516, T3e, T3d); ci[WS(rs, 5)] = FNMS(KP951056516, T3e, T3d); cr[WS(rs, 5)] = T1H + T2c; T2R = FNMS(KP250000000, T2c, T1H); T2Q = FNMS(KP618033988, T2A, T2N); T2O = FMA(KP618033988, T2N, T2A); } { E T2T, T31, T2g, T2j; T2T = FMA(KP559016994, T2S, T2R); T31 = FNMS(KP559016994, T2S, T2R); T2g = T2e + T2f; T3r = T2e - T2f; T3q = T2h - T2i; T2j = T2h + T2i; ci[WS(rs, 3)] = FMA(KP951056516, T30, T2T); ci[0] = FNMS(KP951056516, T30, T2T); ci[WS(rs, 6)] = FMA(KP951056516, T32, T31); cr[WS(rs, 2)] = FNMS(KP951056516, T32, T31); T2k = T2g + T2j; T2m = T2g - T2j; } } } { E T3m, T3s, T3u, T3o, T3h, T2l, T2n, T2P; ci[WS(rs, 4)] = T2d + T2k; T2l = FNMS(KP250000000, T2k, T2d); T3m = FMA(KP866025403, T3l, T3k); T3v = FNMS(KP866025403, T3l, T3k); T3s = FNMS(KP618033988, T3r, T3q); T3u = FMA(KP618033988, T3q, T3r); T2n = FMA(KP559016994, T2m, T2l); T2P = FNMS(KP559016994, T2m, T2l); ci[WS(rs, 1)] = FMA(KP951056516, T2Q, T2P); cr[WS(rs, 7)] = FNMS(KP951056516, T2Q, T2P); cr[WS(rs, 1)] = FMA(KP951056516, T2O, T2n); cr[WS(rs, 4)] = FNMS(KP951056516, T2O, T2n); T3o = T3f - T3g; T3h = T3f + T3g; { E T3S, T3Q, T3K, T3M, T3n, T3p, T3t, T3L, T3R, T3N; cr[WS(rs, 10)] = -(T3h + T3m); T3n = FNMS(KP250000000, T3h, T3m); T3S = FNMS(KP618033988, T3O, T3P); T3Q = FMA(KP618033988, T3P, T3O); T3p = FNMS(KP559016994, T3o, T3n); T3t = FMA(KP559016994, T3o, T3n); ci[WS(rs, 7)] = FMA(KP951056516, T3s, T3p); cr[WS(rs, 13)] = FMS(KP951056516, T3s, T3p); ci[WS(rs, 13)] = FNMS(KP951056516, T3u, T3t); ci[WS(rs, 10)] = FMA(KP951056516, T3u, T3t); T3K = T3I + T3J; T3M = T3I - T3J; ci[WS(rs, 14)] = T3K + T3H; T3L = FNMS(KP250000000, T3K, T3H); T3E = FMA(KP618033988, T3D, T3C); T3G = FNMS(KP618033988, T3C, T3D); T3R = FNMS(KP559016994, T3M, T3L); T3N = FMA(KP559016994, T3M, T3L); ci[WS(rs, 8)] = FMA(KP951056516, T3Q, T3N); cr[WS(rs, 9)] = FMS(KP951056516, T3Q, T3N); ci[WS(rs, 11)] = FMA(KP951056516, T3S, T3R); cr[WS(rs, 12)] = FMS(KP951056516, T3S, T3R); T3A = T3x - T3w; T3y = T3w + T3x; } } } } ci[WS(rs, 9)] = T3y + T3v; T3z = FNMS(KP250000000, T3y, T3v); T3F = FMA(KP559016994, T3A, T3z); T3B = FNMS(KP559016994, T3A, T3z); cr[WS(rs, 14)] = -(FMA(KP951056516, T3E, T3B)); cr[WS(rs, 11)] = FMS(KP951056516, T3E, T3B); ci[WS(rs, 12)] = FMA(KP951056516, T3G, T3F); cr[WS(rs, 8)] = FMS(KP951056516, T3G, T3F); } } } static const tw_instr twinstr[] = { {TW_FULL, 1, 15}, {TW_NEXT, 1, 0} }; static const hc2hc_desc desc = { 15, "hf_15", twinstr, &GENUS, {72, 28, 112, 0} }; void X(codelet_hf_15) (planner *p) { X(khc2hc_register) (p, hf_15, &desc); } #else /* HAVE_FMA */ /* Generated by: ../../../genfft/gen_hc2hc.native -compact -variables 4 -pipeline-latency 4 -n 15 -dit -name hf_15 -include hf.h */ /* * This function contains 184 FP additions, 112 FP multiplications, * (or, 128 additions, 56 multiplications, 56 fused multiply/add), * 65 stack variables, 6 constants, and 60 memory accesses */ #include "hf.h" static void hf_15(R *cr, R *ci, const R *W, stride rs, INT mb, INT me, INT ms) { DK(KP587785252, +0.587785252292473129168705954639072768597652438); DK(KP951056516, +0.951056516295153572116439333379382143405698634); DK(KP250000000, +0.250000000000000000000000000000000000000000000); DK(KP559016994, +0.559016994374947424102293417182819058860154590); DK(KP500000000, +0.500000000000000000000000000000000000000000000); DK(KP866025403, +0.866025403784438646763723170752936183471402627); { INT m; for (m = mb, W = W + ((mb - 1) * 28); m < me; m = m + 1, cr = cr + ms, ci = ci - ms, W = W + 28, MAKE_VOLATILE_STRIDE(30, rs)) { E T1q, T2Q, Td, T1n, T2T, T3l, T13, T1k, T1l, T2E, T2F, T3j, T1H, T1T, T2k; E T2w, T2f, T2v, T1M, T1U, Tu, TL, TM, T2H, T2I, T3i, T1w, T1Q, T29, T2t; E T24, T2s, T1B, T1R; { E T1, T2R, T6, T1o, Tb, T1p, Tc, T2S; T1 = cr[0]; T2R = ci[0]; { E T3, T5, T2, T4; T3 = cr[WS(rs, 5)]; T5 = ci[WS(rs, 5)]; T2 = W[8]; T4 = W[9]; T6 = FMA(T2, T3, T4 * T5); T1o = FNMS(T4, T3, T2 * T5); } { E T8, Ta, T7, T9; T8 = cr[WS(rs, 10)]; Ta = ci[WS(rs, 10)]; T7 = W[18]; T9 = W[19]; Tb = FMA(T7, T8, T9 * Ta); T1p = FNMS(T9, T8, T7 * Ta); } T1q = KP866025403 * (T1o - T1p); T2Q = KP866025403 * (Tb - T6); Tc = T6 + Tb; Td = T1 + Tc; T1n = FNMS(KP500000000, Tc, T1); T2S = T1o + T1p; T2T = FNMS(KP500000000, T2S, T2R); T3l = T2S + T2R; } { E TR, T2c, T18, T2h, TW, T1E, T11, T1F, T12, T2d, T1d, T1J, T1i, T1K, T1j; E T2i; { E TO, TQ, TN, TP; TO = cr[WS(rs, 6)]; TQ = ci[WS(rs, 6)]; TN = W[10]; TP = W[11]; TR = FMA(TN, TO, TP * TQ); T2c = FNMS(TP, TO, TN * TQ); } { E T15, T17, T14, T16; T15 = cr[WS(rs, 9)]; T17 = ci[WS(rs, 9)]; T14 = W[16]; T16 = W[17]; T18 = FMA(T14, T15, T16 * T17); T2h = FNMS(T16, T15, T14 * T17); } { E TT, TV, TS, TU; TT = cr[WS(rs, 11)]; TV = ci[WS(rs, 11)]; TS = W[20]; TU = W[21]; TW = FMA(TS, TT, TU * TV); T1E = FNMS(TU, TT, TS * TV); } { E TY, T10, TX, TZ; TY = cr[WS(rs, 1)]; T10 = ci[WS(rs, 1)]; TX = W[0]; TZ = W[1]; T11 = FMA(TX, TY, TZ * T10); T1F = FNMS(TZ, TY, TX * T10); } T12 = TW + T11; T2d = T1E + T1F; { E T1a, T1c, T19, T1b; T1a = cr[WS(rs, 14)]; T1c = ci[WS(rs, 14)]; T19 = W[26]; T1b = W[27]; T1d = FMA(T19, T1a, T1b * T1c); T1J = FNMS(T1b, T1a, T19 * T1c); } { E T1f, T1h, T1e, T1g; T1f = cr[WS(rs, 4)]; T1h = ci[WS(rs, 4)]; T1e = W[6]; T1g = W[7]; T1i = FMA(T1e, T1f, T1g * T1h); T1K = FNMS(T1g, T1f, T1e * T1h); } T1j = T1d + T1i; T2i = T1J + T1K; { E T1D, T1G, T2g, T2j; T13 = TR + T12; T1k = T18 + T1j; T1l = T13 + T1k; T2E = T2c + T2d; T2F = T2h + T2i; T3j = T2E + T2F; T1D = FNMS(KP500000000, T12, TR); T1G = KP866025403 * (T1E - T1F); T1H = T1D - T1G; T1T = T1D + T1G; T2g = KP866025403 * (T1d - T1i); T2j = FNMS(KP500000000, T2i, T2h); T2k = T2g - T2j; T2w = T2g + T2j; { E T2b, T2e, T1I, T1L; T2b = KP866025403 * (T11 - TW); T2e = FNMS(KP500000000, T2d, T2c); T2f = T2b + T2e; T2v = T2e - T2b; T1I = FNMS(KP500000000, T1j, T18); T1L = KP866025403 * (T1J - T1K); T1M = T1I - T1L; T1U = T1I + T1L; } } } { E Ti, T21, Tz, T26, Tn, T1t, Ts, T1u, Tt, T22, TE, T1y, TJ, T1z, TK; E T27; { E Tf, Th, Te, Tg; Tf = cr[WS(rs, 3)]; Th = ci[WS(rs, 3)]; Te = W[4]; Tg = W[5]; Ti = FMA(Te, Tf, Tg * Th); T21 = FNMS(Tg, Tf, Te * Th); } { E Tw, Ty, Tv, Tx; Tw = cr[WS(rs, 12)]; Ty = ci[WS(rs, 12)]; Tv = W[22]; Tx = W[23]; Tz = FMA(Tv, Tw, Tx * Ty); T26 = FNMS(Tx, Tw, Tv * Ty); } { E Tk, Tm, Tj, Tl; Tk = cr[WS(rs, 8)]; Tm = ci[WS(rs, 8)]; Tj = W[14]; Tl = W[15]; Tn = FMA(Tj, Tk, Tl * Tm); T1t = FNMS(Tl, Tk, Tj * Tm); } { E Tp, Tr, To, Tq; Tp = cr[WS(rs, 13)]; Tr = ci[WS(rs, 13)]; To = W[24]; Tq = W[25]; Ts = FMA(To, Tp, Tq * Tr); T1u = FNMS(Tq, Tp, To * Tr); } Tt = Tn + Ts; T22 = T1t + T1u; { E TB, TD, TA, TC; TB = cr[WS(rs, 2)]; TD = ci[WS(rs, 2)]; TA = W[2]; TC = W[3]; TE = FMA(TA, TB, TC * TD); T1y = FNMS(TC, TB, TA * TD); } { E TG, TI, TF, TH; TG = cr[WS(rs, 7)]; TI = ci[WS(rs, 7)]; TF = W[12]; TH = W[13]; TJ = FMA(TF, TG, TH * TI); T1z = FNMS(TH, TG, TF * TI); } TK = TE + TJ; T27 = T1y + T1z; { E T1s, T1v, T25, T28; Tu = Ti + Tt; TL = Tz + TK; TM = Tu + TL; T2H = T21 + T22; T2I = T26 + T27; T3i = T2H + T2I; T1s = FNMS(KP500000000, Tt, Ti); T1v = KP866025403 * (T1t - T1u); T1w = T1s - T1v; T1Q = T1s + T1v; T25 = KP866025403 * (TJ - TE); T28 = FNMS(KP500000000, T27, T26); T29 = T25 + T28; T2t = T28 - T25; { E T20, T23, T1x, T1A; T20 = KP866025403 * (Ts - Tn); T23 = FNMS(KP500000000, T22, T21); T24 = T20 + T23; T2s = T23 - T20; T1x = FNMS(KP500000000, TK, Tz); T1A = KP866025403 * (T1y - T1z); T1B = T1x - T1A; T1R = T1x + T1A; } } } { E T2C, T1m, T2B, T2K, T2M, T2G, T2J, T2L, T2D; T2C = KP559016994 * (TM - T1l); T1m = TM + T1l; T2B = FNMS(KP250000000, T1m, Td); T2G = T2E - T2F; T2J = T2H - T2I; T2K = FNMS(KP587785252, T2J, KP951056516 * T2G); T2M = FMA(KP951056516, T2J, KP587785252 * T2G); cr[0] = Td + T1m; T2L = T2C + T2B; ci[WS(rs, 5)] = T2L - T2M; cr[WS(rs, 6)] = T2L + T2M; T2D = T2B - T2C; ci[WS(rs, 2)] = T2D - T2K; cr[WS(rs, 3)] = T2D + T2K; } { E T3k, T3m, T3n, T3h, T3p, T3f, T3g, T3q, T3o; T3k = KP559016994 * (T3i - T3j); T3m = T3i + T3j; T3n = FNMS(KP250000000, T3m, T3l); T3f = T1k - T13; T3g = Tu - TL; T3h = FNMS(KP951056516, T3g, KP587785252 * T3f); T3p = FMA(KP587785252, T3g, KP951056516 * T3f); ci[WS(rs, 14)] = T3m + T3l; T3q = T3n - T3k; cr[WS(rs, 12)] = T3p - T3q; ci[WS(rs, 11)] = T3p + T3q; T3o = T3k + T3n; cr[WS(rs, 9)] = T3h - T3o; ci[WS(rs, 8)] = T3h + T3o; } { E T2y, T2A, T1r, T1O, T2p, T2q, T2z, T2r; { E T2u, T2x, T1C, T1N; T2u = T2s - T2t; T2x = T2v - T2w; T2y = FMA(KP951056516, T2u, KP587785252 * T2x); T2A = FNMS(KP587785252, T2u, KP951056516 * T2x); T1r = T1n - T1q; T1C = T1w + T1B; T1N = T1H + T1M; T1O = T1C + T1N; T2p = KP559016994 * (T1C - T1N); T2q = FNMS(KP250000000, T1O, T1r); } cr[WS(rs, 5)] = T1r + T1O; T2z = T2q - T2p; cr[WS(rs, 2)] = T2z - T2A; ci[WS(rs, 6)] = T2z + T2A; T2r = T2p + T2q; ci[0] = T2r - T2y; ci[WS(rs, 3)] = T2r + T2y; } { E T35, T3d, T39, T3a, T38, T3b, T3e, T3c; { E T33, T34, T36, T37; T33 = T1w - T1B; T34 = T1H - T1M; T35 = FMA(KP951056516, T33, KP587785252 * T34); T3d = FNMS(KP587785252, T33, KP951056516 * T34); T39 = T2T - T2Q; T36 = T2v + T2w; T37 = T2s + T2t; T3a = T37 + T36; T38 = KP559016994 * (T36 - T37); T3b = FNMS(KP250000000, T3a, T39); } ci[WS(rs, 9)] = T3a + T39; T3e = T38 + T3b; cr[WS(rs, 8)] = T3d - T3e; ci[WS(rs, 12)] = T3d + T3e; T3c = T38 - T3b; cr[WS(rs, 11)] = T35 + T3c; cr[WS(rs, 14)] = T3c - T35; } { E T2X, T31, T2U, T2P, T2Y, T2Z, T32, T30; { E T2V, T2W, T2N, T2O; T2V = T1T - T1U; T2W = T1Q - T1R; T2X = FNMS(KP587785252, T2W, KP951056516 * T2V); T31 = FMA(KP951056516, T2W, KP587785252 * T2V); T2U = T2Q + T2T; T2N = T2k - T2f; T2O = T24 + T29; T2P = T2N - T2O; T2Y = FMA(KP250000000, T2P, T2U); T2Z = KP559016994 * (T2O + T2N); } cr[WS(rs, 10)] = T2P - T2U; T32 = T2Z + T2Y; ci[WS(rs, 10)] = T31 + T32; ci[WS(rs, 13)] = T32 - T31; T30 = T2Y - T2Z; cr[WS(rs, 13)] = T2X - T30; ci[WS(rs, 7)] = T2X + T30; } { E T2m, T2o, T1P, T1W, T1X, T1Y, T1Z, T2n; { E T2a, T2l, T1S, T1V; T2a = T24 - T29; T2l = T2f + T2k; T2m = FMA(KP951056516, T2a, KP587785252 * T2l); T2o = FNMS(KP587785252, T2a, KP951056516 * T2l); T1P = T1n + T1q; T1S = T1Q + T1R; T1V = T1T + T1U; T1W = T1S + T1V; T1X = KP559016994 * (T1S - T1V); T1Y = FNMS(KP250000000, T1W, T1P); } ci[WS(rs, 4)] = T1P + T1W; T1Z = T1X + T1Y; cr[WS(rs, 4)] = T1Z - T2m; cr[WS(rs, 1)] = T1Z + T2m; T2n = T1Y - T1X; cr[WS(rs, 7)] = T2n - T2o; ci[WS(rs, 1)] = T2n + T2o; } } } } static const tw_instr twinstr[] = { {TW_FULL, 1, 15}, {TW_NEXT, 1, 0} }; static const hc2hc_desc desc = { 15, "hf_15", twinstr, &GENUS, {128, 56, 56, 0} }; void X(codelet_hf_15) (planner *p) { X(khc2hc_register) (p, hf_15, &desc); } #endif /* HAVE_FMA */ fftw-3.3.3/rdft/scalar/r2cf/hf_16.c0000644000175000017500000004760012054410633013523 00000000000000/* * Copyright (c) 2003, 2007-11 Matteo Frigo * Copyright (c) 2003, 2007-11 Massachusetts Institute of Technology * * This program is free software; you can redistribute it and/or modify * it under the terms of the GNU General Public License as published by * the Free Software Foundation; either version 2 of the License, or * (at your option) any later version. * * This program is distributed in the hope that it will be useful, * but WITHOUT ANY WARRANTY; without even the implied warranty of * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the * GNU General Public License for more details. * * You should have received a copy of the GNU General Public License * along with this program; if not, write to the Free Software * Foundation, Inc., 51 Franklin Street, Fifth Floor, Boston, MA 02110-1301 USA * */ /* This file was automatically generated --- DO NOT EDIT */ /* Generated on Sun Nov 25 07:39:54 EST 2012 */ #include "codelet-rdft.h" #ifdef HAVE_FMA /* Generated by: ../../../genfft/gen_hc2hc.native -fma -reorder-insns -schedule-for-pipeline -compact -variables 4 -pipeline-latency 4 -n 16 -dit -name hf_16 -include hf.h */ /* * This function contains 174 FP additions, 100 FP multiplications, * (or, 104 additions, 30 multiplications, 70 fused multiply/add), * 95 stack variables, 3 constants, and 64 memory accesses */ #include "hf.h" static void hf_16(R *cr, R *ci, const R *W, stride rs, INT mb, INT me, INT ms) { DK(KP923879532, +0.923879532511286756128183189396788286822416626); DK(KP414213562, +0.414213562373095048801688724209698078569671875); DK(KP707106781, +0.707106781186547524400844362104849039284835938); { INT m; for (m = mb, W = W + ((mb - 1) * 30); m < me; m = m + 1, cr = cr + ms, ci = ci - ms, W = W + 30, MAKE_VOLATILE_STRIDE(32, rs)) { E T2T, T2Q; { E T3A, T3o, T8, T1I, T2w, T35, T2k, T1s, T2p, T36, T2r, T1F, T3k, T1N, T3z; E Tl, T1U, T2W, T1P, Tz, T2g, T30, T25, T11, TB, TE, T2a, T31, T2h, T1e; E TC, T1X, TH, TK, TG, TD, TJ; { E Ta, Td, Tb, T1J, Tg, Tj, Tf, Tc, Ti; { E T1h, T1k, T1n, T2s, T1i, T1q, T1m, T1j, T1p; { E T1, T3n, T3, T6, T2, T5; T1 = cr[0]; T3n = ci[0]; T3 = cr[WS(rs, 8)]; T6 = ci[WS(rs, 8)]; T2 = W[14]; T5 = W[15]; { E T3l, T4, T1g, T3m, T7; T1h = cr[WS(rs, 15)]; T1k = ci[WS(rs, 15)]; T3l = T2 * T6; T4 = T2 * T3; T1g = W[28]; T1n = cr[WS(rs, 7)]; T3m = FNMS(T5, T3, T3l); T7 = FMA(T5, T6, T4); T2s = T1g * T1k; T1i = T1g * T1h; T3A = T3n - T3m; T3o = T3m + T3n; T8 = T1 + T7; T1I = T1 - T7; T1q = ci[WS(rs, 7)]; T1m = W[12]; } T1j = W[29]; T1p = W[13]; } { E T1u, T1x, T1v, T2l, T1A, T1D, T1z, T1w, T1C; { E T2t, T1l, T2v, T1r, T2u, T1o, T1t; T1u = cr[WS(rs, 3)]; T2u = T1m * T1q; T1o = T1m * T1n; T2t = FNMS(T1j, T1h, T2s); T1l = FMA(T1j, T1k, T1i); T2v = FNMS(T1p, T1n, T2u); T1r = FMA(T1p, T1q, T1o); T1x = ci[WS(rs, 3)]; T1t = W[4]; T2w = T2t - T2v; T35 = T2t + T2v; T2k = T1l - T1r; T1s = T1l + T1r; T1v = T1t * T1u; T2l = T1t * T1x; } T1A = cr[WS(rs, 11)]; T1D = ci[WS(rs, 11)]; T1z = W[20]; T1w = W[5]; T1C = W[21]; { E T2m, T1y, T2o, T1E, T2n, T1B, T9; Ta = cr[WS(rs, 4)]; T2n = T1z * T1D; T1B = T1z * T1A; T2m = FNMS(T1w, T1u, T2l); T1y = FMA(T1w, T1x, T1v); T2o = FNMS(T1C, T1A, T2n); T1E = FMA(T1C, T1D, T1B); Td = ci[WS(rs, 4)]; T9 = W[6]; T2p = T2m - T2o; T36 = T2m + T2o; T2r = T1E - T1y; T1F = T1y + T1E; Tb = T9 * Ta; T1J = T9 * Td; } Tg = cr[WS(rs, 12)]; Tj = ci[WS(rs, 12)]; Tf = W[22]; Tc = W[7]; Ti = W[23]; } } { E TQ, TT, TR, T2c, TW, TZ, TV, TS, TY; { E To, Tr, Tp, T1Q, Tu, Tx, Tt, Tq, Tw; { E T1K, Te, T1M, Tk, T1L, Th, Tn; To = cr[WS(rs, 2)]; T1L = Tf * Tj; Th = Tf * Tg; T1K = FNMS(Tc, Ta, T1J); Te = FMA(Tc, Td, Tb); T1M = FNMS(Ti, Tg, T1L); Tk = FMA(Ti, Tj, Th); Tr = ci[WS(rs, 2)]; Tn = W[2]; T3k = T1K + T1M; T1N = T1K - T1M; T3z = Te - Tk; Tl = Te + Tk; Tp = Tn * To; T1Q = Tn * Tr; } Tu = cr[WS(rs, 10)]; Tx = ci[WS(rs, 10)]; Tt = W[18]; Tq = W[3]; Tw = W[19]; { E T1R, Ts, T1T, Ty, T1S, Tv, TP; TQ = cr[WS(rs, 1)]; T1S = Tt * Tx; Tv = Tt * Tu; T1R = FNMS(Tq, To, T1Q); Ts = FMA(Tq, Tr, Tp); T1T = FNMS(Tw, Tu, T1S); Ty = FMA(Tw, Tx, Tv); TT = ci[WS(rs, 1)]; TP = W[0]; T1U = T1R - T1T; T2W = T1R + T1T; T1P = Ts - Ty; Tz = Ts + Ty; TR = TP * TQ; T2c = TP * TT; } TW = cr[WS(rs, 9)]; TZ = ci[WS(rs, 9)]; TV = W[16]; TS = W[1]; TY = W[17]; } { E T13, T16, T14, T26, T19, T1c, T18, T15, T1b; { E T2d, TU, T2f, T10, T2e, TX, T12; T13 = cr[WS(rs, 5)]; T2e = TV * TZ; TX = TV * TW; T2d = FNMS(TS, TQ, T2c); TU = FMA(TS, TT, TR); T2f = FNMS(TY, TW, T2e); T10 = FMA(TY, TZ, TX); T16 = ci[WS(rs, 5)]; T12 = W[8]; T2g = T2d - T2f; T30 = T2d + T2f; T25 = TU - T10; T11 = TU + T10; T14 = T12 * T13; T26 = T12 * T16; } T19 = cr[WS(rs, 13)]; T1c = ci[WS(rs, 13)]; T18 = W[24]; T15 = W[9]; T1b = W[25]; { E T27, T17, T29, T1d, T28, T1a, TA; TB = cr[WS(rs, 14)]; T28 = T18 * T1c; T1a = T18 * T19; T27 = FNMS(T15, T13, T26); T17 = FMA(T15, T16, T14); T29 = FNMS(T1b, T19, T28); T1d = FMA(T1b, T1c, T1a); TE = ci[WS(rs, 14)]; TA = W[26]; T2a = T27 - T29; T31 = T27 + T29; T2h = T17 - T1d; T1e = T17 + T1d; TC = TA * TB; T1X = TA * TE; } TH = cr[WS(rs, 6)]; TK = ci[WS(rs, 6)]; TG = W[10]; TD = W[27]; TJ = W[11]; } } } { E T2U, T3u, T2Z, T21, T1W, T34, T2X, T37, T3t, T3q, T3e, T32, T3i, T3h; { E T3f, T3r, T1H, T3s, TO, T3g; { E Tm, T1Y, TF, T20, TL, T3p, T1Z, TI; T2U = T8 - Tl; Tm = T8 + Tl; T1Z = TG * TK; TI = TG * TH; T1Y = FNMS(TD, TB, T1X); TF = FMA(TD, TE, TC); T20 = FNMS(TJ, TH, T1Z); TL = FMA(TJ, TK, TI); T3p = T3k + T3o; T3u = T3o - T3k; { E T1f, TM, T1G, T3j, T2V, TN; T2Z = T11 - T1e; T1f = T11 + T1e; T21 = T1Y - T20; T2V = T1Y + T20; T1W = TF - TL; TM = TF + TL; T1G = T1s + T1F; T34 = T1s - T1F; T2X = T2V - T2W; T3j = T2W + T2V; T3f = T35 + T36; T37 = T35 - T36; T3t = Tz - TM; TN = Tz + TM; T3r = T1G - T1f; T1H = T1f + T1G; T3s = T3p - T3j; T3q = T3j + T3p; T3e = Tm - TN; TO = Tm + TN; T3g = T30 + T31; T32 = T30 - T31; } } cr[WS(rs, 12)] = T3r - T3s; ci[WS(rs, 11)] = T3r + T3s; ci[WS(rs, 7)] = TO - T1H; T3i = T3g + T3f; T3h = T3f - T3g; cr[0] = TO + T1H; } { E T3a, T2Y, T3x, T3v; ci[WS(rs, 15)] = T3i + T3q; cr[WS(rs, 8)] = T3i - T3q; ci[WS(rs, 3)] = T3e + T3h; cr[WS(rs, 4)] = T3e - T3h; T3a = T2U + T2X; T2Y = T2U - T2X; T3x = T3u - T3t; T3v = T3t + T3u; { E T2E, T1O, T3B, T3H, T2q, T2x, T3I, T23, T2R, T2O, T2J, T2K, T3C, T2H, T2B; E T2j; { E T2F, T1V, T22, T2G; { E T3b, T33, T3c, T38; T2E = T1I + T1N; T1O = T1I - T1N; T3b = T2Z - T32; T33 = T2Z + T32; T3c = T34 + T37; T38 = T34 - T37; T3B = T3z + T3A; T3H = T3A - T3z; { E T3d, T3y, T3w, T39; T3d = T3b + T3c; T3y = T3c - T3b; T3w = T38 - T33; T39 = T33 + T38; ci[WS(rs, 1)] = FMA(KP707106781, T3d, T3a); cr[WS(rs, 6)] = FNMS(KP707106781, T3d, T3a); ci[WS(rs, 13)] = FMA(KP707106781, T3y, T3x); cr[WS(rs, 10)] = FMS(KP707106781, T3y, T3x); ci[WS(rs, 9)] = FMA(KP707106781, T3w, T3v); cr[WS(rs, 14)] = FMS(KP707106781, T3w, T3v); cr[WS(rs, 2)] = FMA(KP707106781, T39, T2Y); ci[WS(rs, 5)] = FNMS(KP707106781, T39, T2Y); T2F = T1P + T1U; T1V = T1P - T1U; T22 = T1W + T21; T2G = T1W - T21; } } { E T2M, T2N, T2b, T2i; T2q = T2k - T2p; T2M = T2k + T2p; T2N = T2w + T2r; T2x = T2r - T2w; T3I = T22 - T1V; T23 = T1V + T22; T2R = FMA(KP414213562, T2M, T2N); T2O = FNMS(KP414213562, T2N, T2M); T2J = T25 + T2a; T2b = T25 - T2a; T2i = T2g + T2h; T2K = T2g - T2h; T3C = T2F - T2G; T2H = T2F + T2G; T2B = FMA(KP414213562, T2b, T2i); T2j = FNMS(KP414213562, T2i, T2b); } } { E T2A, T3G, T2P, T2D, T3E, T3F, T3D, T2I; { E T24, T2L, T2C, T2y, T3J, T3L, T3K, T2S, T2z, T3M; T2A = FNMS(KP707106781, T23, T1O); T24 = FMA(KP707106781, T23, T1O); T2S = FNMS(KP414213562, T2J, T2K); T2L = FMA(KP414213562, T2K, T2J); T2C = FMA(KP414213562, T2q, T2x); T2y = FNMS(KP414213562, T2x, T2q); T3J = FMA(KP707106781, T3I, T3H); T3L = FNMS(KP707106781, T3I, T3H); T2T = T2R - T2S; T3K = T2S + T2R; T3G = T2y - T2j; T2z = T2j + T2y; T3M = T2O - T2L; T2P = T2L + T2O; ci[WS(rs, 14)] = FMA(KP923879532, T3K, T3J); cr[WS(rs, 9)] = FMS(KP923879532, T3K, T3J); ci[0] = FMA(KP923879532, T2z, T24); cr[WS(rs, 7)] = FNMS(KP923879532, T2z, T24); cr[WS(rs, 13)] = FMS(KP923879532, T3M, T3L); ci[WS(rs, 10)] = FMA(KP923879532, T3M, T3L); T2D = T2B + T2C; T3E = T2C - T2B; } T2Q = FNMS(KP707106781, T2H, T2E); T2I = FMA(KP707106781, T2H, T2E); T3F = FNMS(KP707106781, T3C, T3B); T3D = FMA(KP707106781, T3C, T3B); cr[WS(rs, 3)] = FMA(KP923879532, T2D, T2A); ci[WS(rs, 4)] = FNMS(KP923879532, T2D, T2A); cr[WS(rs, 1)] = FMA(KP923879532, T2P, T2I); ci[WS(rs, 6)] = FNMS(KP923879532, T2P, T2I); ci[WS(rs, 8)] = FMA(KP923879532, T3E, T3D); cr[WS(rs, 15)] = FMS(KP923879532, T3E, T3D); ci[WS(rs, 12)] = FMA(KP923879532, T3G, T3F); cr[WS(rs, 11)] = FMS(KP923879532, T3G, T3F); } } } } } ci[WS(rs, 2)] = FMA(KP923879532, T2T, T2Q); cr[WS(rs, 5)] = FNMS(KP923879532, T2T, T2Q); } } } static const tw_instr twinstr[] = { {TW_FULL, 1, 16}, {TW_NEXT, 1, 0} }; static const hc2hc_desc desc = { 16, "hf_16", twinstr, &GENUS, {104, 30, 70, 0} }; void X(codelet_hf_16) (planner *p) { X(khc2hc_register) (p, hf_16, &desc); } #else /* HAVE_FMA */ /* Generated by: ../../../genfft/gen_hc2hc.native -compact -variables 4 -pipeline-latency 4 -n 16 -dit -name hf_16 -include hf.h */ /* * This function contains 174 FP additions, 84 FP multiplications, * (or, 136 additions, 46 multiplications, 38 fused multiply/add), * 52 stack variables, 3 constants, and 64 memory accesses */ #include "hf.h" static void hf_16(R *cr, R *ci, const R *W, stride rs, INT mb, INT me, INT ms) { DK(KP382683432, +0.382683432365089771728459984030398866761344562); DK(KP923879532, +0.923879532511286756128183189396788286822416626); DK(KP707106781, +0.707106781186547524400844362104849039284835938); { INT m; for (m = mb, W = W + ((mb - 1) * 30); m < me; m = m + 1, cr = cr + ms, ci = ci - ms, W = W + 30, MAKE_VOLATILE_STRIDE(32, rs)) { E T7, T38, T1t, T2U, Ti, T37, T1w, T2R, Tu, T2t, T1C, T2c, TF, T2s, T1H; E T2d, T1f, T1q, T2B, T2C, T2D, T2E, T1Z, T2k, T24, T2j, TS, T13, T2w, T2x; E T2y, T2z, T1O, T2h, T1T, T2g; { E T1, T2T, T6, T2S; T1 = cr[0]; T2T = ci[0]; { E T3, T5, T2, T4; T3 = cr[WS(rs, 8)]; T5 = ci[WS(rs, 8)]; T2 = W[14]; T4 = W[15]; T6 = FMA(T2, T3, T4 * T5); T2S = FNMS(T4, T3, T2 * T5); } T7 = T1 + T6; T38 = T2T - T2S; T1t = T1 - T6; T2U = T2S + T2T; } { E Tc, T1u, Th, T1v; { E T9, Tb, T8, Ta; T9 = cr[WS(rs, 4)]; Tb = ci[WS(rs, 4)]; T8 = W[6]; Ta = W[7]; Tc = FMA(T8, T9, Ta * Tb); T1u = FNMS(Ta, T9, T8 * Tb); } { E Te, Tg, Td, Tf; Te = cr[WS(rs, 12)]; Tg = ci[WS(rs, 12)]; Td = W[22]; Tf = W[23]; Th = FMA(Td, Te, Tf * Tg); T1v = FNMS(Tf, Te, Td * Tg); } Ti = Tc + Th; T37 = Tc - Th; T1w = T1u - T1v; T2R = T1u + T1v; } { E To, T1z, Tt, T1A, T1y, T1B; { E Tl, Tn, Tk, Tm; Tl = cr[WS(rs, 2)]; Tn = ci[WS(rs, 2)]; Tk = W[2]; Tm = W[3]; To = FMA(Tk, Tl, Tm * Tn); T1z = FNMS(Tm, Tl, Tk * Tn); } { E Tq, Ts, Tp, Tr; Tq = cr[WS(rs, 10)]; Ts = ci[WS(rs, 10)]; Tp = W[18]; Tr = W[19]; Tt = FMA(Tp, Tq, Tr * Ts); T1A = FNMS(Tr, Tq, Tp * Ts); } Tu = To + Tt; T2t = T1z + T1A; T1y = To - Tt; T1B = T1z - T1A; T1C = T1y - T1B; T2c = T1y + T1B; } { E Tz, T1E, TE, T1F, T1D, T1G; { E Tw, Ty, Tv, Tx; Tw = cr[WS(rs, 14)]; Ty = ci[WS(rs, 14)]; Tv = W[26]; Tx = W[27]; Tz = FMA(Tv, Tw, Tx * Ty); T1E = FNMS(Tx, Tw, Tv * Ty); } { E TB, TD, TA, TC; TB = cr[WS(rs, 6)]; TD = ci[WS(rs, 6)]; TA = W[10]; TC = W[11]; TE = FMA(TA, TB, TC * TD); T1F = FNMS(TC, TB, TA * TD); } TF = Tz + TE; T2s = T1E + T1F; T1D = Tz - TE; T1G = T1E - T1F; T1H = T1D + T1G; T2d = T1D - T1G; } { E T19, T1V, T1p, T22, T1e, T1W, T1k, T21; { E T16, T18, T15, T17; T16 = cr[WS(rs, 15)]; T18 = ci[WS(rs, 15)]; T15 = W[28]; T17 = W[29]; T19 = FMA(T15, T16, T17 * T18); T1V = FNMS(T17, T16, T15 * T18); } { E T1m, T1o, T1l, T1n; T1m = cr[WS(rs, 11)]; T1o = ci[WS(rs, 11)]; T1l = W[20]; T1n = W[21]; T1p = FMA(T1l, T1m, T1n * T1o); T22 = FNMS(T1n, T1m, T1l * T1o); } { E T1b, T1d, T1a, T1c; T1b = cr[WS(rs, 7)]; T1d = ci[WS(rs, 7)]; T1a = W[12]; T1c = W[13]; T1e = FMA(T1a, T1b, T1c * T1d); T1W = FNMS(T1c, T1b, T1a * T1d); } { E T1h, T1j, T1g, T1i; T1h = cr[WS(rs, 3)]; T1j = ci[WS(rs, 3)]; T1g = W[4]; T1i = W[5]; T1k = FMA(T1g, T1h, T1i * T1j); T21 = FNMS(T1i, T1h, T1g * T1j); } T1f = T19 + T1e; T1q = T1k + T1p; T2B = T1f - T1q; T2C = T1V + T1W; T2D = T21 + T22; T2E = T2C - T2D; { E T1X, T1Y, T20, T23; T1X = T1V - T1W; T1Y = T1k - T1p; T1Z = T1X + T1Y; T2k = T1X - T1Y; T20 = T19 - T1e; T23 = T21 - T22; T24 = T20 - T23; T2j = T20 + T23; } } { E TM, T1P, T12, T1M, TR, T1Q, TX, T1L; { E TJ, TL, TI, TK; TJ = cr[WS(rs, 1)]; TL = ci[WS(rs, 1)]; TI = W[0]; TK = W[1]; TM = FMA(TI, TJ, TK * TL); T1P = FNMS(TK, TJ, TI * TL); } { E TZ, T11, TY, T10; TZ = cr[WS(rs, 13)]; T11 = ci[WS(rs, 13)]; TY = W[24]; T10 = W[25]; T12 = FMA(TY, TZ, T10 * T11); T1M = FNMS(T10, TZ, TY * T11); } { E TO, TQ, TN, TP; TO = cr[WS(rs, 9)]; TQ = ci[WS(rs, 9)]; TN = W[16]; TP = W[17]; TR = FMA(TN, TO, TP * TQ); T1Q = FNMS(TP, TO, TN * TQ); } { E TU, TW, TT, TV; TU = cr[WS(rs, 5)]; TW = ci[WS(rs, 5)]; TT = W[8]; TV = W[9]; TX = FMA(TT, TU, TV * TW); T1L = FNMS(TV, TU, TT * TW); } TS = TM + TR; T13 = TX + T12; T2w = TS - T13; T2x = T1P + T1Q; T2y = T1L + T1M; T2z = T2x - T2y; { E T1K, T1N, T1R, T1S; T1K = TM - TR; T1N = T1L - T1M; T1O = T1K - T1N; T2h = T1K + T1N; T1R = T1P - T1Q; T1S = TX - T12; T1T = T1R + T1S; T2g = T1R - T1S; } } { E T1J, T27, T3a, T3c, T26, T3b, T2a, T35; { E T1x, T1I, T36, T39; T1x = T1t - T1w; T1I = KP707106781 * (T1C + T1H); T1J = T1x + T1I; T27 = T1x - T1I; T36 = KP707106781 * (T2c - T2d); T39 = T37 + T38; T3a = T36 + T39; T3c = T39 - T36; } { E T1U, T25, T28, T29; T1U = FNMS(KP382683432, T1T, KP923879532 * T1O); T25 = FMA(KP382683432, T1Z, KP923879532 * T24); T26 = T1U + T25; T3b = T25 - T1U; T28 = FMA(KP923879532, T1T, KP382683432 * T1O); T29 = FNMS(KP923879532, T1Z, KP382683432 * T24); T2a = T28 + T29; T35 = T29 - T28; } cr[WS(rs, 7)] = T1J - T26; cr[WS(rs, 11)] = T3b - T3c; ci[WS(rs, 12)] = T3b + T3c; ci[0] = T1J + T26; ci[WS(rs, 4)] = T27 - T2a; cr[WS(rs, 15)] = T35 - T3a; ci[WS(rs, 8)] = T35 + T3a; cr[WS(rs, 3)] = T27 + T2a; } { E TH, T2L, T2W, T2Y, T1s, T2X, T2O, T2P; { E Tj, TG, T2Q, T2V; Tj = T7 + Ti; TG = Tu + TF; TH = Tj + TG; T2L = Tj - TG; T2Q = T2t + T2s; T2V = T2R + T2U; T2W = T2Q + T2V; T2Y = T2V - T2Q; } { E T14, T1r, T2M, T2N; T14 = TS + T13; T1r = T1f + T1q; T1s = T14 + T1r; T2X = T1r - T14; T2M = T2C + T2D; T2N = T2x + T2y; T2O = T2M - T2N; T2P = T2N + T2M; } ci[WS(rs, 7)] = TH - T1s; cr[WS(rs, 12)] = T2X - T2Y; ci[WS(rs, 11)] = T2X + T2Y; cr[0] = TH + T1s; cr[WS(rs, 4)] = T2L - T2O; cr[WS(rs, 8)] = T2P - T2W; ci[WS(rs, 15)] = T2P + T2W; ci[WS(rs, 3)] = T2L + T2O; } { E T2f, T2n, T3g, T3i, T2m, T3h, T2q, T3d; { E T2b, T2e, T3e, T3f; T2b = T1t + T1w; T2e = KP707106781 * (T2c + T2d); T2f = T2b + T2e; T2n = T2b - T2e; T3e = KP707106781 * (T1H - T1C); T3f = T38 - T37; T3g = T3e + T3f; T3i = T3f - T3e; } { E T2i, T2l, T2o, T2p; T2i = FMA(KP382683432, T2g, KP923879532 * T2h); T2l = FNMS(KP382683432, T2k, KP923879532 * T2j); T2m = T2i + T2l; T3h = T2l - T2i; T2o = FNMS(KP923879532, T2g, KP382683432 * T2h); T2p = FMA(KP923879532, T2k, KP382683432 * T2j); T2q = T2o + T2p; T3d = T2p - T2o; } ci[WS(rs, 6)] = T2f - T2m; cr[WS(rs, 13)] = T3h - T3i; ci[WS(rs, 10)] = T3h + T3i; cr[WS(rs, 1)] = T2f + T2m; cr[WS(rs, 5)] = T2n - T2q; cr[WS(rs, 9)] = T3d - T3g; ci[WS(rs, 14)] = T3d + T3g; ci[WS(rs, 2)] = T2n + T2q; } { E T2v, T2H, T32, T34, T2G, T2Z, T2K, T33; { E T2r, T2u, T30, T31; T2r = T7 - Ti; T2u = T2s - T2t; T2v = T2r - T2u; T2H = T2r + T2u; T30 = Tu - TF; T31 = T2U - T2R; T32 = T30 + T31; T34 = T31 - T30; } { E T2A, T2F, T2I, T2J; T2A = T2w + T2z; T2F = T2B - T2E; T2G = KP707106781 * (T2A + T2F); T2Z = KP707106781 * (T2F - T2A); T2I = T2w - T2z; T2J = T2B + T2E; T2K = KP707106781 * (T2I + T2J); T33 = KP707106781 * (T2J - T2I); } ci[WS(rs, 5)] = T2v - T2G; cr[WS(rs, 10)] = T33 - T34; ci[WS(rs, 13)] = T33 + T34; cr[WS(rs, 2)] = T2v + T2G; cr[WS(rs, 6)] = T2H - T2K; cr[WS(rs, 14)] = T2Z - T32; ci[WS(rs, 9)] = T2Z + T32; ci[WS(rs, 1)] = T2H + T2K; } } } } static const tw_instr twinstr[] = { {TW_FULL, 1, 16}, {TW_NEXT, 1, 0} }; static const hc2hc_desc desc = { 16, "hf_16", twinstr, &GENUS, {136, 46, 38, 0} }; void X(codelet_hf_16) (planner *p) { X(khc2hc_register) (p, hf_16, &desc); } #endif /* HAVE_FMA */ fftw-3.3.3/rdft/scalar/r2cf/hf_32.c0000644000175000017500000013231712054410642013521 00000000000000/* * Copyright (c) 2003, 2007-11 Matteo Frigo * Copyright (c) 2003, 2007-11 Massachusetts Institute of Technology * * This program is free software; you can redistribute it and/or modify * it under the terms of the GNU General Public License as published by * the Free Software Foundation; either version 2 of the License, or * (at your option) any later version. * * This program is distributed in the hope that it will be useful, * but WITHOUT ANY WARRANTY; without even the implied warranty of * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the * GNU General Public License for more details. * * You should have received a copy of the GNU General Public License * along with this program; if not, write to the Free Software * Foundation, Inc., 51 Franklin Street, Fifth Floor, Boston, MA 02110-1301 USA * */ /* This file was automatically generated --- DO NOT EDIT */ /* Generated on Sun Nov 25 07:39:54 EST 2012 */ #include "codelet-rdft.h" #ifdef HAVE_FMA /* Generated by: ../../../genfft/gen_hc2hc.native -fma -reorder-insns -schedule-for-pipeline -compact -variables 4 -pipeline-latency 4 -n 32 -dit -name hf_32 -include hf.h */ /* * This function contains 434 FP additions, 260 FP multiplications, * (or, 236 additions, 62 multiplications, 198 fused multiply/add), * 135 stack variables, 7 constants, and 128 memory accesses */ #include "hf.h" static void hf_32(R *cr, R *ci, const R *W, stride rs, INT mb, INT me, INT ms) { DK(KP831469612, +0.831469612302545237078788377617905756738560812); DK(KP980785280, +0.980785280403230449126182236134239036973933731); DK(KP923879532, +0.923879532511286756128183189396788286822416626); DK(KP668178637, +0.668178637919298919997757686523080761552472251); DK(KP198912367, +0.198912367379658006911597622644676228597850501); DK(KP414213562, +0.414213562373095048801688724209698078569671875); DK(KP707106781, +0.707106781186547524400844362104849039284835938); { INT m; for (m = mb, W = W + ((mb - 1) * 62); m < me; m = m + 1, cr = cr + ms, ci = ci - ms, W = W + 62, MAKE_VOLATILE_STRIDE(64, rs)) { E T6D, T6A; { E T8y, T87, T8, T3w, T83, T3B, T8x, Tl, T6G, Tz, T3J, T5T, T6F, TM, T3Q; E T5U, T46, T5X, T7E, T6M, T5Y, T3Z, T6J, T1f, T7D, T6R, T61, T4e, T6O, T1G; E T60, T4l, T54, T6c, T7d, T7N, T32, T76, T6f, T5r, T4v, T65, T72, T7I, T29; E T6V, T68, T4S, T5t, T5b, T7O, T79, T7e, T3t, T5s, T5i, T4H, T2y, T4B, T6X; E T2m, T4w, T4F, T2s; { E T44, T1d, T3X, T6K, T11, T40, T42, T17, T5h, T5c; { E Ta, Td, Tg, T3x, Tb, Tj, Tf, Tc, Ti; { E T1, T86, T3, T6, T2, T5; T1 = cr[0]; T86 = ci[0]; T3 = cr[WS(rs, 16)]; T6 = ci[WS(rs, 16)]; T2 = W[30]; T5 = W[31]; { E T84, T4, T9, T85, T7; Ta = cr[WS(rs, 8)]; Td = ci[WS(rs, 8)]; T84 = T2 * T6; T4 = T2 * T3; T9 = W[14]; Tg = cr[WS(rs, 24)]; T85 = FNMS(T5, T3, T84); T7 = FMA(T5, T6, T4); T3x = T9 * Td; Tb = T9 * Ta; T8y = T86 - T85; T87 = T85 + T86; T8 = T1 + T7; T3w = T1 - T7; Tj = ci[WS(rs, 24)]; Tf = W[46]; } Tc = W[15]; Ti = W[47]; } { E Tu, Tx, T3F, Ts, Tw, T3G, Tv; { E To, Tr, Tp, T3E, Tq, Tt; { E T3y, Te, T3A, Tk, T3z, Th, Tn; To = cr[WS(rs, 4)]; T3z = Tf * Tj; Th = Tf * Tg; T3y = FNMS(Tc, Ta, T3x); Te = FMA(Tc, Td, Tb); T3A = FNMS(Ti, Tg, T3z); Tk = FMA(Ti, Tj, Th); Tr = ci[WS(rs, 4)]; Tn = W[6]; T83 = T3y + T3A; T3B = T3y - T3A; T8x = Te - Tk; Tl = Te + Tk; Tp = Tn * To; T3E = Tn * Tr; } Tq = W[7]; Tu = cr[WS(rs, 20)]; Tx = ci[WS(rs, 20)]; Tt = W[38]; T3F = FNMS(Tq, To, T3E); Ts = FMA(Tq, Tr, Tp); Tw = W[39]; T3G = Tt * Tx; Tv = Tt * Tu; } { E T3M, TF, TH, TK, TG, TJ, TE, TD, TC; { E TB, T3H, Ty, TA, T3I, T3D, T3L; TB = cr[WS(rs, 28)]; TE = ci[WS(rs, 28)]; T3H = FNMS(Tw, Tu, T3G); Ty = FMA(Tw, Tx, Tv); TA = W[54]; TD = W[55]; T6G = T3F + T3H; T3I = T3F - T3H; Tz = Ts + Ty; T3D = Ts - Ty; T3L = TA * TE; TC = TA * TB; T3J = T3D - T3I; T5T = T3D + T3I; T3M = FNMS(TD, TB, T3L); } TF = FMA(TD, TE, TC); TH = cr[WS(rs, 12)]; TK = ci[WS(rs, 12)]; TG = W[22]; TJ = W[23]; { E TU, T3U, T13, T16, T3W, T10, T12, T15, T41, T14; { E T19, T1c, T18, T1b, T3P, T3K; { E TQ, TT, T3N, TI, TP, TS; TQ = cr[WS(rs, 2)]; TT = ci[WS(rs, 2)]; T3N = TG * TK; TI = TG * TH; TP = W[2]; TS = W[3]; { E T3O, TL, T3T, TR; T3O = FNMS(TJ, TH, T3N); TL = FMA(TJ, TK, TI); T3T = TP * TT; TR = TP * TQ; T6F = T3M + T3O; T3P = T3M - T3O; TM = TF + TL; T3K = TF - TL; TU = FMA(TS, TT, TR); T3U = FNMS(TS, TQ, T3T); } } T3Q = T3K + T3P; T5U = T3K - T3P; T19 = cr[WS(rs, 26)]; T1c = ci[WS(rs, 26)]; T18 = W[50]; T1b = W[51]; { E TW, TZ, TY, T3V, TX, T43, T1a, TV; TW = cr[WS(rs, 18)]; TZ = ci[WS(rs, 18)]; T43 = T18 * T1c; T1a = T18 * T19; TV = W[34]; TY = W[35]; T44 = FNMS(T1b, T19, T43); T1d = FMA(T1b, T1c, T1a); T3V = TV * TZ; TX = TV * TW; T13 = cr[WS(rs, 10)]; T16 = ci[WS(rs, 10)]; T3W = FNMS(TY, TW, T3V); T10 = FMA(TY, TZ, TX); T12 = W[18]; T15 = W[19]; } } T3X = T3U - T3W; T6K = T3U + T3W; T11 = TU + T10; T40 = TU - T10; T41 = T12 * T16; T14 = T12 * T13; T42 = FNMS(T15, T13, T41); T17 = FMA(T15, T16, T14); } } } } { E T49, T1l, T4j, T1E, T1u, T1x, T1w, T4b, T1r, T4g, T1v; { E T1A, T1D, T1C, T4i, T1B; { E T1h, T1k, T1g, T1j, T48, T1i, T1z; T1h = cr[WS(rs, 30)]; T1k = ci[WS(rs, 30)]; { E T6L, T45, T1e, T3Y; T6L = T42 + T44; T45 = T42 - T44; T1e = T17 + T1d; T3Y = T17 - T1d; T46 = T40 - T45; T5X = T40 + T45; T7E = T6K + T6L; T6M = T6K - T6L; T5Y = T3X - T3Y; T3Z = T3X + T3Y; T6J = T11 - T1e; T1f = T11 + T1e; T1g = W[58]; } T1j = W[59]; T1A = cr[WS(rs, 22)]; T1D = ci[WS(rs, 22)]; T48 = T1g * T1k; T1i = T1g * T1h; T1z = W[42]; T1C = W[43]; T49 = FNMS(T1j, T1h, T48); T1l = FMA(T1j, T1k, T1i); T4i = T1z * T1D; T1B = T1z * T1A; } { E T1n, T1q, T1m, T1p, T4a, T1o, T1t; T1n = cr[WS(rs, 14)]; T1q = ci[WS(rs, 14)]; T4j = FNMS(T1C, T1A, T4i); T1E = FMA(T1C, T1D, T1B); T1m = W[26]; T1p = W[27]; T1u = cr[WS(rs, 6)]; T1x = ci[WS(rs, 6)]; T4a = T1m * T1q; T1o = T1m * T1n; T1t = W[10]; T1w = W[11]; T4b = FNMS(T1p, T1n, T4a); T1r = FMA(T1p, T1q, T1o); T4g = T1t * T1x; T1v = T1t * T1u; } } { E T4c, T6P, T1s, T4f, T4h, T1y; T4c = T49 - T4b; T6P = T49 + T4b; T1s = T1l + T1r; T4f = T1l - T1r; T4h = FNMS(T1w, T1u, T4g); T1y = FMA(T1w, T1x, T1v); { E T4k, T6Q, T4d, T1F; T4k = T4h - T4j; T6Q = T4h + T4j; T4d = T1y - T1E; T1F = T1y + T1E; T7D = T6P + T6Q; T6R = T6P - T6Q; T61 = T4c - T4d; T4e = T4c + T4d; T6O = T1s - T1F; T1G = T1s + T1F; T60 = T4f + T4k; T4l = T4f - T4k; } } } { E T5n, T2H, T52, T30, T2Q, T2T, T2S, T5p, T2N, T4Z, T2R; { E T2W, T2Z, T2Y, T51, T2X; { E T2D, T2G, T2C, T2F, T5m, T2E, T2V; T2D = cr[WS(rs, 31)]; T2G = ci[WS(rs, 31)]; T2C = W[60]; T2F = W[61]; T2W = cr[WS(rs, 23)]; T2Z = ci[WS(rs, 23)]; T5m = T2C * T2G; T2E = T2C * T2D; T2V = W[44]; T2Y = W[45]; T5n = FNMS(T2F, T2D, T5m); T2H = FMA(T2F, T2G, T2E); T51 = T2V * T2Z; T2X = T2V * T2W; } { E T2J, T2M, T2I, T2L, T5o, T2K, T2P; T2J = cr[WS(rs, 15)]; T2M = ci[WS(rs, 15)]; T52 = FNMS(T2Y, T2W, T51); T30 = FMA(T2Y, T2Z, T2X); T2I = W[28]; T2L = W[29]; T2Q = cr[WS(rs, 7)]; T2T = ci[WS(rs, 7)]; T5o = T2I * T2M; T2K = T2I * T2J; T2P = W[12]; T2S = W[13]; T5p = FNMS(T2L, T2J, T5o); T2N = FMA(T2L, T2M, T2K); T4Z = T2P * T2T; T2R = T2P * T2Q; } } { E T5q, T7b, T2O, T4Y, T50, T2U; T5q = T5n - T5p; T7b = T5n + T5p; T2O = T2H + T2N; T4Y = T2H - T2N; T50 = FNMS(T2S, T2Q, T4Z); T2U = FMA(T2S, T2T, T2R); { E T7c, T53, T31, T5l; T7c = T50 + T52; T53 = T50 - T52; T31 = T2U + T30; T5l = T30 - T2U; T54 = T4Y - T53; T6c = T4Y + T53; T7d = T7b - T7c; T7N = T7b + T7c; T32 = T2O + T31; T76 = T2O - T31; T6f = T5q + T5l; T5r = T5l - T5q; } } } { E T4N, T1O, T4t, T27, T1X, T20, T1Z, T4P, T1U, T4q, T1Y; { E T23, T26, T25, T4s, T24; { E T1K, T1N, T1J, T1M, T4M, T1L, T22; T1K = cr[WS(rs, 1)]; T1N = ci[WS(rs, 1)]; T1J = W[0]; T1M = W[1]; T23 = cr[WS(rs, 25)]; T26 = ci[WS(rs, 25)]; T4M = T1J * T1N; T1L = T1J * T1K; T22 = W[48]; T25 = W[49]; T4N = FNMS(T1M, T1K, T4M); T1O = FMA(T1M, T1N, T1L); T4s = T22 * T26; T24 = T22 * T23; } { E T1Q, T1T, T1P, T1S, T4O, T1R, T1W; T1Q = cr[WS(rs, 17)]; T1T = ci[WS(rs, 17)]; T4t = FNMS(T25, T23, T4s); T27 = FMA(T25, T26, T24); T1P = W[32]; T1S = W[33]; T1X = cr[WS(rs, 9)]; T20 = ci[WS(rs, 9)]; T4O = T1P * T1T; T1R = T1P * T1Q; T1W = W[16]; T1Z = W[17]; T4P = FNMS(T1S, T1Q, T4O); T1U = FMA(T1S, T1T, T1R); T4q = T1W * T20; T1Y = T1W * T1X; } } { E T4Q, T70, T1V, T4p, T4r, T21; T4Q = T4N - T4P; T70 = T4N + T4P; T1V = T1O + T1U; T4p = T1O - T1U; T4r = FNMS(T1Z, T1X, T4q); T21 = FMA(T1Z, T20, T1Y); { E T71, T4u, T4R, T28; T71 = T4r + T4t; T4u = T4r - T4t; T4R = T21 - T27; T28 = T21 + T27; T4v = T4p - T4u; T65 = T4p + T4u; T72 = T70 - T71; T7I = T70 + T71; T29 = T1V + T28; T6V = T1V - T28; T68 = T4Q - T4R; T4S = T4Q + T4R; } } } { E T57, T38, T5g, T3r, T3h, T3k, T3j, T59, T3e, T5d, T3i; { E T3n, T3q, T3p, T5f, T3o; { E T34, T37, T33, T36, T56, T35, T3m; T34 = cr[WS(rs, 3)]; T37 = ci[WS(rs, 3)]; T33 = W[4]; T36 = W[5]; T3n = cr[WS(rs, 11)]; T3q = ci[WS(rs, 11)]; T56 = T33 * T37; T35 = T33 * T34; T3m = W[20]; T3p = W[21]; T57 = FNMS(T36, T34, T56); T38 = FMA(T36, T37, T35); T5f = T3m * T3q; T3o = T3m * T3n; } { E T3a, T3d, T39, T3c, T58, T3b, T3g; T3a = cr[WS(rs, 19)]; T3d = ci[WS(rs, 19)]; T5g = FNMS(T3p, T3n, T5f); T3r = FMA(T3p, T3q, T3o); T39 = W[36]; T3c = W[37]; T3h = cr[WS(rs, 27)]; T3k = ci[WS(rs, 27)]; T58 = T39 * T3d; T3b = T39 * T3a; T3g = W[52]; T3j = W[53]; T59 = FNMS(T3c, T3a, T58); T3e = FMA(T3c, T3d, T3b); T5d = T3g * T3k; T3i = T3g * T3h; } } { E T5a, T78, T3f, T55, T5e, T3l, T77, T3s; T5a = T57 - T59; T78 = T57 + T59; T3f = T38 + T3e; T55 = T38 - T3e; T5e = FNMS(T3j, T3h, T5d); T3l = FMA(T3j, T3k, T3i); T5h = T5e - T5g; T77 = T5e + T5g; T3s = T3l + T3r; T5c = T3l - T3r; T5t = T55 + T5a; T5b = T55 - T5a; T7O = T78 + T77; T79 = T77 - T78; T7e = T3s - T3f; T3t = T3f + T3s; } } { E T4y, T2f, T2o, T2r, T4A, T2l, T2n, T2q, T4E, T2p; { E T2u, T2x, T2t, T2w; { E T2b, T2e, T2d, T4x, T2c, T2a; T2b = cr[WS(rs, 5)]; T2e = ci[WS(rs, 5)]; T2a = W[8]; T5s = T5c - T5h; T5i = T5c + T5h; T2d = W[9]; T4x = T2a * T2e; T2c = T2a * T2b; T2u = cr[WS(rs, 13)]; T2x = ci[WS(rs, 13)]; T4y = FNMS(T2d, T2b, T4x); T2f = FMA(T2d, T2e, T2c); T2t = W[24]; T2w = W[25]; } { E T2h, T2k, T2j, T4z, T2i, T4G, T2v, T2g; T2h = cr[WS(rs, 21)]; T2k = ci[WS(rs, 21)]; T4G = T2t * T2x; T2v = T2t * T2u; T2g = W[40]; T2j = W[41]; T4H = FNMS(T2w, T2u, T4G); T2y = FMA(T2w, T2x, T2v); T4z = T2g * T2k; T2i = T2g * T2h; T2o = cr[WS(rs, 29)]; T2r = ci[WS(rs, 29)]; T4A = FNMS(T2j, T2h, T4z); T2l = FMA(T2j, T2k, T2i); T2n = W[56]; T2q = W[57]; } } T4B = T4y - T4A; T6X = T4y + T4A; T2m = T2f + T2l; T4w = T2f - T2l; T4E = T2n * T2r; T2p = T2n * T2o; T4F = FNMS(T2q, T2o, T4E); T2s = FMA(T2q, T2r, T2p); } } { E T6E, T8j, T6Y, T73, T6H, T8k, T5S, T8O, T8N, T5V, T6g, T6d, T69, T66, T5O; E T5R; { E T4T, T4C, T4J, T4U, T7S, T7V; { E T7C, TO, T80, T7Z, T8e, T89, T8d, T1H, T8b, T3v, T7T, T7L, T7U, T7Q, T2A; E T7P, T7K, T7W, T1I; { E T7X, T7Y, T7J, T82, T88; { E Tm, T4I, T6W, T4D, T2z, TN; T6E = T8 - Tl; Tm = T8 + Tl; T4T = T4w + T4B; T4C = T4w - T4B; T4I = T4F - T4H; T6W = T4F + T4H; T4D = T2s - T2y; T2z = T2s + T2y; TN = Tz + TM; T8j = Tz - TM; T6Y = T6W - T6X; T7J = T6X + T6W; T4J = T4D + T4I; T4U = T4I - T4D; T2A = T2m + T2z; T73 = T2m - T2z; T7C = Tm - TN; TO = Tm + TN; } T7P = T7N - T7O; T7X = T7N + T7O; T7Y = T7I + T7J; T7K = T7I - T7J; T6H = T6F - T6G; T82 = T6G + T6F; T88 = T83 + T87; T8k = T87 - T83; T80 = T7Y + T7X; T7Z = T7X - T7Y; T8e = T88 - T82; T89 = T82 + T88; } { E T7H, T7M, T2B, T3u; T7H = T29 - T2A; T2B = T29 + T2A; T3u = T32 + T3t; T7M = T32 - T3t; T8d = T1f - T1G; T1H = T1f + T1G; T8b = T3u - T2B; T3v = T2B + T3u; T7T = T7H - T7K; T7L = T7H + T7K; T7U = T7M + T7P; T7Q = T7M - T7P; } T7W = TO - T1H; T1I = TO + T1H; { E T8g, T8h, T8f, T8i; { E T7R, T8c, T8a, T7G, T81, T7F; T8g = T7Q - T7L; T7R = T7L + T7Q; T81 = T7E + T7D; T7F = T7D - T7E; cr[0] = T1I + T3v; ci[WS(rs, 15)] = T1I - T3v; ci[WS(rs, 7)] = T7W + T7Z; cr[WS(rs, 8)] = T7W - T7Z; T8c = T89 - T81; T8a = T81 + T89; T7G = T7C - T7F; T7S = T7C + T7F; T8h = T8e - T8d; T8f = T8d + T8e; ci[WS(rs, 23)] = T8b + T8c; cr[WS(rs, 24)] = T8b - T8c; ci[WS(rs, 31)] = T80 + T8a; cr[WS(rs, 16)] = T80 - T8a; cr[WS(rs, 4)] = FMA(KP707106781, T7R, T7G); ci[WS(rs, 11)] = FNMS(KP707106781, T7R, T7G); } T8i = T7U - T7T; T7V = T7T + T7U; ci[WS(rs, 19)] = FMA(KP707106781, T8g, T8f); cr[WS(rs, 28)] = FMS(KP707106781, T8g, T8f); ci[WS(rs, 27)] = FMA(KP707106781, T8i, T8h); cr[WS(rs, 20)] = FMS(KP707106781, T8i, T8h); } } { E T5C, T3S, T8C, T4n, T8H, T8B, T8I, T5F, T4L, T5H, T5M, T5Q, T5A, T5w, T4V; { E T5D, T47, T4m, T5E, T8z, T8A, T3C, T3R, T5j, T5u; T5S = T3w + T3B; T3C = T3w - T3B; T3R = T3J + T3Q; T8O = T3Q - T3J; T5D = FNMS(KP414213562, T3Z, T46); T47 = FMA(KP414213562, T46, T3Z); ci[WS(rs, 3)] = FMA(KP707106781, T7V, T7S); cr[WS(rs, 12)] = FNMS(KP707106781, T7V, T7S); T5C = FMA(KP707106781, T3R, T3C); T3S = FNMS(KP707106781, T3R, T3C); T4m = FNMS(KP414213562, T4l, T4e); T5E = FMA(KP414213562, T4e, T4l); T8N = T8y - T8x; T8z = T8x + T8y; T8A = T5T - T5U; T5V = T5T + T5U; T8C = T47 + T4m; T4n = T47 - T4m; T8H = FNMS(KP707106781, T8A, T8z); T8B = FMA(KP707106781, T8A, T8z); T6g = T5i - T5b; T5j = T5b + T5i; T5u = T5s - T5t; T6d = T5t + T5s; { E T5K, T5k, T5L, T5v, T4K; T69 = T4J - T4C; T4K = T4C + T4J; T8I = T5E - T5D; T5F = T5D + T5E; T5K = FMA(KP707106781, T5j, T54); T5k = FNMS(KP707106781, T5j, T54); T5L = FMA(KP707106781, T5u, T5r); T5v = FNMS(KP707106781, T5u, T5r); T4L = FNMS(KP707106781, T4K, T4v); T5H = FMA(KP707106781, T4K, T4v); T5M = FNMS(KP198912367, T5L, T5K); T5Q = FMA(KP198912367, T5K, T5L); T5A = FNMS(KP668178637, T5k, T5v); T5w = FMA(KP668178637, T5v, T5k); T4V = T4T + T4U; T66 = T4T - T4U; } } { E T5y, T4o, T8J, T8L, T5I, T4W; T5y = FNMS(KP923879532, T4n, T3S); T4o = FMA(KP923879532, T4n, T3S); T8J = FMA(KP923879532, T8I, T8H); T8L = FNMS(KP923879532, T8I, T8H); T5I = FMA(KP707106781, T4V, T4S); T4W = FNMS(KP707106781, T4V, T4S); { E T8G, T8F, T8D, T8E; { E T5G, T5P, T5z, T4X, T5N, T5J; T5O = FNMS(KP923879532, T5F, T5C); T5G = FMA(KP923879532, T5F, T5C); T5J = FNMS(KP198912367, T5I, T5H); T5P = FMA(KP198912367, T5H, T5I); T5z = FNMS(KP668178637, T4L, T4W); T4X = FMA(KP668178637, T4W, T4L); T5N = T5J + T5M; T8G = T5M - T5J; T8F = FNMS(KP923879532, T8C, T8B); T8D = FMA(KP923879532, T8C, T8B); { E T5B, T8K, T8M, T5x; T5B = T5z + T5A; T8K = T5z - T5A; T8M = T5w - T4X; T5x = T4X + T5w; ci[0] = FMA(KP980785280, T5N, T5G); cr[WS(rs, 15)] = FNMS(KP980785280, T5N, T5G); ci[WS(rs, 4)] = FNMS(KP831469612, T5B, T5y); cr[WS(rs, 11)] = FMA(KP831469612, T5B, T5y); ci[WS(rs, 28)] = FMA(KP831469612, T8K, T8J); cr[WS(rs, 19)] = FMS(KP831469612, T8K, T8J); ci[WS(rs, 20)] = FMA(KP831469612, T8M, T8L); cr[WS(rs, 27)] = FMS(KP831469612, T8M, T8L); cr[WS(rs, 3)] = FMA(KP831469612, T5x, T4o); ci[WS(rs, 12)] = FNMS(KP831469612, T5x, T4o); T8E = T5Q - T5P; T5R = T5P + T5Q; } } ci[WS(rs, 16)] = FMA(KP980785280, T8E, T8D); cr[WS(rs, 31)] = FMS(KP980785280, T8E, T8D); ci[WS(rs, 24)] = FMA(KP980785280, T8G, T8F); cr[WS(rs, 23)] = FMS(KP980785280, T8G, T8F); } } } } { E T7y, T8q, T8p, T7B; { E T7a, T7m, T6I, T7f, T7A, T7w, T8r, T8l, T8m, T6T, T7k, T75, T8s, T7p, T7z; E T7t; { E T7n, T6N, T6S, T7o, T7u, T7v; T7a = T76 - T79; T7u = T76 + T79; cr[WS(rs, 7)] = FMA(KP980785280, T5R, T5O); ci[WS(rs, 8)] = FNMS(KP980785280, T5R, T5O); T7m = T6E + T6H; T6I = T6E - T6H; T7v = T7e - T7d; T7f = T7d + T7e; T7n = T6J - T6M; T6N = T6J + T6M; T7A = FMA(KP414213562, T7u, T7v); T7w = FNMS(KP414213562, T7v, T7u); T8r = T8k - T8j; T8l = T8j + T8k; T6S = T6O - T6R; T7o = T6O + T6R; { E T7r, T7s, T6Z, T74; T7r = T6V + T6Y; T6Z = T6V - T6Y; T74 = T72 - T73; T7s = T72 + T73; T8m = T6N - T6S; T6T = T6N + T6S; T7k = FNMS(KP414213562, T6Z, T74); T75 = FMA(KP414213562, T74, T6Z); T8s = T7o - T7n; T7p = T7n + T7o; T7z = FMA(KP414213562, T7r, T7s); T7t = FNMS(KP414213562, T7s, T7r); } } { E T7i, T6U, T8t, T8v, T7j, T7g; T7i = FNMS(KP707106781, T6T, T6I); T6U = FMA(KP707106781, T6T, T6I); T8t = FMA(KP707106781, T8s, T8r); T8v = FNMS(KP707106781, T8s, T8r); T7j = FMA(KP414213562, T7a, T7f); T7g = FNMS(KP414213562, T7f, T7a); { E T7q, T7x, T8n, T8o; T7y = FNMS(KP707106781, T7p, T7m); T7q = FMA(KP707106781, T7p, T7m); { E T7l, T8u, T8w, T7h; T7l = T7j - T7k; T8u = T7k + T7j; T8w = T7g - T75; T7h = T75 + T7g; ci[WS(rs, 5)] = FMA(KP923879532, T7l, T7i); cr[WS(rs, 10)] = FNMS(KP923879532, T7l, T7i); ci[WS(rs, 29)] = FMA(KP923879532, T8u, T8t); cr[WS(rs, 18)] = FMS(KP923879532, T8u, T8t); ci[WS(rs, 21)] = FMA(KP923879532, T8w, T8v); cr[WS(rs, 26)] = FMS(KP923879532, T8w, T8v); cr[WS(rs, 2)] = FMA(KP923879532, T7h, T6U); ci[WS(rs, 13)] = FNMS(KP923879532, T7h, T6U); T7x = T7t + T7w; T8q = T7w - T7t; } T8p = FNMS(KP707106781, T8m, T8l); T8n = FMA(KP707106781, T8m, T8l); T8o = T7A - T7z; T7B = T7z + T7A; ci[WS(rs, 1)] = FMA(KP923879532, T7x, T7q); cr[WS(rs, 14)] = FNMS(KP923879532, T7x, T7q); ci[WS(rs, 17)] = FMA(KP923879532, T8o, T8n); cr[WS(rs, 30)] = FMS(KP923879532, T8o, T8n); } } } { E T6o, T5W, T8W, T63, T8V, T8P, T8Q, T6r, T6e, T6w; { E T6q, T6p, T5Z, T62; ci[WS(rs, 25)] = FMA(KP923879532, T8q, T8p); cr[WS(rs, 22)] = FMS(KP923879532, T8q, T8p); cr[WS(rs, 6)] = FMA(KP923879532, T7B, T7y); ci[WS(rs, 9)] = FNMS(KP923879532, T7B, T7y); T6q = FNMS(KP414213562, T5X, T5Y); T5Z = FMA(KP414213562, T5Y, T5X); T62 = FNMS(KP414213562, T61, T60); T6p = FMA(KP414213562, T60, T61); T6o = FNMS(KP707106781, T5V, T5S); T5W = FMA(KP707106781, T5V, T5S); T8W = T5Z - T62; T63 = T5Z + T62; T8V = FNMS(KP707106781, T8O, T8N); T8P = FMA(KP707106781, T8O, T8N); T8Q = T6q + T6p; T6r = T6p - T6q; T6e = FMA(KP707106781, T6d, T6c); T6w = FNMS(KP707106781, T6d, T6c); } { E T6k, T8U, T6z, T6n, T8S, T8T, T8R, T6s; { E T64, T6y, T6l, T6i, T6v, T6m, T6b, T8X, T8Z, T8Y, T6j, T90; { E T6C, T6B, T6x, T6h; T6k = FNMS(KP923879532, T63, T5W); T64 = FMA(KP923879532, T63, T5W); T6x = FNMS(KP707106781, T6g, T6f); T6h = FMA(KP707106781, T6g, T6f); { E T6t, T67, T6u, T6a; T6t = FNMS(KP707106781, T66, T65); T67 = FMA(KP707106781, T66, T65); T6u = FNMS(KP707106781, T69, T68); T6a = FMA(KP707106781, T69, T68); T6y = FMA(KP668178637, T6x, T6w); T6C = FNMS(KP668178637, T6w, T6x); T6l = FMA(KP198912367, T6e, T6h); T6i = FNMS(KP198912367, T6h, T6e); T6v = FNMS(KP668178637, T6u, T6t); T6B = FMA(KP668178637, T6t, T6u); T6m = FNMS(KP198912367, T67, T6a); T6b = FMA(KP198912367, T6a, T67); } T8X = FMA(KP923879532, T8W, T8V); T8Z = FNMS(KP923879532, T8W, T8V); T6D = T6B - T6C; T8Y = T6B + T6C; } T8U = T6i - T6b; T6j = T6b + T6i; T90 = T6y - T6v; T6z = T6v + T6y; ci[WS(rs, 18)] = FNMS(KP831469612, T8Y, T8X); cr[WS(rs, 29)] = -(FMA(KP831469612, T8Y, T8X)); cr[WS(rs, 1)] = FMA(KP980785280, T6j, T64); ci[WS(rs, 14)] = FNMS(KP980785280, T6j, T64); cr[WS(rs, 21)] = FMS(KP831469612, T90, T8Z); ci[WS(rs, 26)] = FMA(KP831469612, T90, T8Z); T6n = T6l - T6m; T8S = T6m + T6l; } T6A = FNMS(KP923879532, T6r, T6o); T6s = FMA(KP923879532, T6r, T6o); T8T = FNMS(KP923879532, T8Q, T8P); T8R = FMA(KP923879532, T8Q, T8P); ci[WS(rs, 6)] = FMA(KP980785280, T6n, T6k); cr[WS(rs, 9)] = FNMS(KP980785280, T6n, T6k); ci[WS(rs, 2)] = FMA(KP831469612, T6z, T6s); cr[WS(rs, 13)] = FNMS(KP831469612, T6z, T6s); ci[WS(rs, 30)] = FMA(KP980785280, T8S, T8R); cr[WS(rs, 17)] = FMS(KP980785280, T8S, T8R); ci[WS(rs, 22)] = FMA(KP980785280, T8U, T8T); cr[WS(rs, 25)] = FMS(KP980785280, T8U, T8T); } } } } } cr[WS(rs, 5)] = FMA(KP831469612, T6D, T6A); ci[WS(rs, 10)] = FNMS(KP831469612, T6D, T6A); } } } static const tw_instr twinstr[] = { {TW_FULL, 1, 32}, {TW_NEXT, 1, 0} }; static const hc2hc_desc desc = { 32, "hf_32", twinstr, &GENUS, {236, 62, 198, 0} }; void X(codelet_hf_32) (planner *p) { X(khc2hc_register) (p, hf_32, &desc); } #else /* HAVE_FMA */ /* Generated by: ../../../genfft/gen_hc2hc.native -compact -variables 4 -pipeline-latency 4 -n 32 -dit -name hf_32 -include hf.h */ /* * This function contains 434 FP additions, 208 FP multiplications, * (or, 340 additions, 114 multiplications, 94 fused multiply/add), * 96 stack variables, 7 constants, and 128 memory accesses */ #include "hf.h" static void hf_32(R *cr, R *ci, const R *W, stride rs, INT mb, INT me, INT ms) { DK(KP555570233, +0.555570233019602224742830813948532874374937191); DK(KP831469612, +0.831469612302545237078788377617905756738560812); DK(KP980785280, +0.980785280403230449126182236134239036973933731); DK(KP195090322, +0.195090322016128267848284868477022240927691618); DK(KP382683432, +0.382683432365089771728459984030398866761344562); DK(KP923879532, +0.923879532511286756128183189396788286822416626); DK(KP707106781, +0.707106781186547524400844362104849039284835938); { INT m; for (m = mb, W = W + ((mb - 1) * 62); m < me; m = m + 1, cr = cr + ms, ci = ci - ms, W = W + 62, MAKE_VOLATILE_STRIDE(64, rs)) { E Tj, T5F, T7C, T7Q, T35, T4T, T78, T7m, T1Q, T61, T5Y, T6J, T3K, T56, T41; E T59, T2B, T67, T6e, T6O, T4b, T5g, T4s, T5d, TG, T7l, T5I, T73, T3a, T4U; E T3f, T4V, T14, T5K, T5N, T6F, T3m, T4Z, T3r, T4Y, T1r, T5P, T5S, T6E, T3x; E T52, T3C, T51, T2d, T5Z, T64, T6K, T3V, T5a, T44, T57, T2Y, T6f, T6a, T6P; E T4m, T5e, T4v, T5h; { E T1, T76, T6, T75, Tc, T32, Th, T33; T1 = cr[0]; T76 = ci[0]; { E T3, T5, T2, T4; T3 = cr[WS(rs, 16)]; T5 = ci[WS(rs, 16)]; T2 = W[30]; T4 = W[31]; T6 = FMA(T2, T3, T4 * T5); T75 = FNMS(T4, T3, T2 * T5); } { E T9, Tb, T8, Ta; T9 = cr[WS(rs, 8)]; Tb = ci[WS(rs, 8)]; T8 = W[14]; Ta = W[15]; Tc = FMA(T8, T9, Ta * Tb); T32 = FNMS(Ta, T9, T8 * Tb); } { E Te, Tg, Td, Tf; Te = cr[WS(rs, 24)]; Tg = ci[WS(rs, 24)]; Td = W[46]; Tf = W[47]; Th = FMA(Td, Te, Tf * Tg); T33 = FNMS(Tf, Te, Td * Tg); } { E T7, Ti, T7A, T7B; T7 = T1 + T6; Ti = Tc + Th; Tj = T7 + Ti; T5F = T7 - Ti; T7A = Tc - Th; T7B = T76 - T75; T7C = T7A + T7B; T7Q = T7B - T7A; } { E T31, T34, T74, T77; T31 = T1 - T6; T34 = T32 - T33; T35 = T31 + T34; T4T = T31 - T34; T74 = T32 + T33; T77 = T75 + T76; T78 = T74 + T77; T7m = T77 - T74; } } { E T1y, T3X, T1O, T3I, T1D, T3Y, T1J, T3H; { E T1v, T1x, T1u, T1w; T1v = cr[WS(rs, 1)]; T1x = ci[WS(rs, 1)]; T1u = W[0]; T1w = W[1]; T1y = FMA(T1u, T1v, T1w * T1x); T3X = FNMS(T1w, T1v, T1u * T1x); } { E T1L, T1N, T1K, T1M; T1L = cr[WS(rs, 25)]; T1N = ci[WS(rs, 25)]; T1K = W[48]; T1M = W[49]; T1O = FMA(T1K, T1L, T1M * T1N); T3I = FNMS(T1M, T1L, T1K * T1N); } { E T1A, T1C, T1z, T1B; T1A = cr[WS(rs, 17)]; T1C = ci[WS(rs, 17)]; T1z = W[32]; T1B = W[33]; T1D = FMA(T1z, T1A, T1B * T1C); T3Y = FNMS(T1B, T1A, T1z * T1C); } { E T1G, T1I, T1F, T1H; T1G = cr[WS(rs, 9)]; T1I = ci[WS(rs, 9)]; T1F = W[16]; T1H = W[17]; T1J = FMA(T1F, T1G, T1H * T1I); T3H = FNMS(T1H, T1G, T1F * T1I); } { E T1E, T1P, T5W, T5X; T1E = T1y + T1D; T1P = T1J + T1O; T1Q = T1E + T1P; T61 = T1E - T1P; T5W = T3X + T3Y; T5X = T3H + T3I; T5Y = T5W - T5X; T6J = T5W + T5X; } { E T3G, T3J, T3Z, T40; T3G = T1y - T1D; T3J = T3H - T3I; T3K = T3G + T3J; T56 = T3G - T3J; T3Z = T3X - T3Y; T40 = T1J - T1O; T41 = T3Z - T40; T59 = T3Z + T40; } } { E T2j, T47, T2z, T4q, T2o, T48, T2u, T4p; { E T2g, T2i, T2f, T2h; T2g = cr[WS(rs, 31)]; T2i = ci[WS(rs, 31)]; T2f = W[60]; T2h = W[61]; T2j = FMA(T2f, T2g, T2h * T2i); T47 = FNMS(T2h, T2g, T2f * T2i); } { E T2w, T2y, T2v, T2x; T2w = cr[WS(rs, 23)]; T2y = ci[WS(rs, 23)]; T2v = W[44]; T2x = W[45]; T2z = FMA(T2v, T2w, T2x * T2y); T4q = FNMS(T2x, T2w, T2v * T2y); } { E T2l, T2n, T2k, T2m; T2l = cr[WS(rs, 15)]; T2n = ci[WS(rs, 15)]; T2k = W[28]; T2m = W[29]; T2o = FMA(T2k, T2l, T2m * T2n); T48 = FNMS(T2m, T2l, T2k * T2n); } { E T2r, T2t, T2q, T2s; T2r = cr[WS(rs, 7)]; T2t = ci[WS(rs, 7)]; T2q = W[12]; T2s = W[13]; T2u = FMA(T2q, T2r, T2s * T2t); T4p = FNMS(T2s, T2r, T2q * T2t); } { E T2p, T2A, T6c, T6d; T2p = T2j + T2o; T2A = T2u + T2z; T2B = T2p + T2A; T67 = T2p - T2A; T6c = T47 + T48; T6d = T4p + T4q; T6e = T6c - T6d; T6O = T6c + T6d; } { E T49, T4a, T4o, T4r; T49 = T47 - T48; T4a = T2u - T2z; T4b = T49 - T4a; T5g = T49 + T4a; T4o = T2j - T2o; T4r = T4p - T4q; T4s = T4o + T4r; T5d = T4o - T4r; } } { E To, T37, TE, T3d, Tt, T38, Tz, T3c; { E Tl, Tn, Tk, Tm; Tl = cr[WS(rs, 4)]; Tn = ci[WS(rs, 4)]; Tk = W[6]; Tm = W[7]; To = FMA(Tk, Tl, Tm * Tn); T37 = FNMS(Tm, Tl, Tk * Tn); } { E TB, TD, TA, TC; TB = cr[WS(rs, 12)]; TD = ci[WS(rs, 12)]; TA = W[22]; TC = W[23]; TE = FMA(TA, TB, TC * TD); T3d = FNMS(TC, TB, TA * TD); } { E Tq, Ts, Tp, Tr; Tq = cr[WS(rs, 20)]; Ts = ci[WS(rs, 20)]; Tp = W[38]; Tr = W[39]; Tt = FMA(Tp, Tq, Tr * Ts); T38 = FNMS(Tr, Tq, Tp * Ts); } { E Tw, Ty, Tv, Tx; Tw = cr[WS(rs, 28)]; Ty = ci[WS(rs, 28)]; Tv = W[54]; Tx = W[55]; Tz = FMA(Tv, Tw, Tx * Ty); T3c = FNMS(Tx, Tw, Tv * Ty); } { E Tu, TF, T5G, T5H; Tu = To + Tt; TF = Tz + TE; TG = Tu + TF; T7l = Tu - TF; T5G = T3c + T3d; T5H = T37 + T38; T5I = T5G - T5H; T73 = T5H + T5G; } { E T36, T39, T3b, T3e; T36 = To - Tt; T39 = T37 - T38; T3a = T36 + T39; T4U = T36 - T39; T3b = Tz - TE; T3e = T3c - T3d; T3f = T3b - T3e; T4V = T3b + T3e; } } { E TM, T3n, T12, T3k, TR, T3o, TX, T3j; { E TJ, TL, TI, TK; TJ = cr[WS(rs, 2)]; TL = ci[WS(rs, 2)]; TI = W[2]; TK = W[3]; TM = FMA(TI, TJ, TK * TL); T3n = FNMS(TK, TJ, TI * TL); } { E TZ, T11, TY, T10; TZ = cr[WS(rs, 26)]; T11 = ci[WS(rs, 26)]; TY = W[50]; T10 = W[51]; T12 = FMA(TY, TZ, T10 * T11); T3k = FNMS(T10, TZ, TY * T11); } { E TO, TQ, TN, TP; TO = cr[WS(rs, 18)]; TQ = ci[WS(rs, 18)]; TN = W[34]; TP = W[35]; TR = FMA(TN, TO, TP * TQ); T3o = FNMS(TP, TO, TN * TQ); } { E TU, TW, TT, TV; TU = cr[WS(rs, 10)]; TW = ci[WS(rs, 10)]; TT = W[18]; TV = W[19]; TX = FMA(TT, TU, TV * TW); T3j = FNMS(TV, TU, TT * TW); } { E TS, T13, T5L, T5M; TS = TM + TR; T13 = TX + T12; T14 = TS + T13; T5K = TS - T13; T5L = T3n + T3o; T5M = T3j + T3k; T5N = T5L - T5M; T6F = T5L + T5M; } { E T3i, T3l, T3p, T3q; T3i = TM - TR; T3l = T3j - T3k; T3m = T3i + T3l; T4Z = T3i - T3l; T3p = T3n - T3o; T3q = TX - T12; T3r = T3p - T3q; T4Y = T3p + T3q; } } { E T19, T3t, T1p, T3A, T1e, T3u, T1k, T3z; { E T16, T18, T15, T17; T16 = cr[WS(rs, 30)]; T18 = ci[WS(rs, 30)]; T15 = W[58]; T17 = W[59]; T19 = FMA(T15, T16, T17 * T18); T3t = FNMS(T17, T16, T15 * T18); } { E T1m, T1o, T1l, T1n; T1m = cr[WS(rs, 22)]; T1o = ci[WS(rs, 22)]; T1l = W[42]; T1n = W[43]; T1p = FMA(T1l, T1m, T1n * T1o); T3A = FNMS(T1n, T1m, T1l * T1o); } { E T1b, T1d, T1a, T1c; T1b = cr[WS(rs, 14)]; T1d = ci[WS(rs, 14)]; T1a = W[26]; T1c = W[27]; T1e = FMA(T1a, T1b, T1c * T1d); T3u = FNMS(T1c, T1b, T1a * T1d); } { E T1h, T1j, T1g, T1i; T1h = cr[WS(rs, 6)]; T1j = ci[WS(rs, 6)]; T1g = W[10]; T1i = W[11]; T1k = FMA(T1g, T1h, T1i * T1j); T3z = FNMS(T1i, T1h, T1g * T1j); } { E T1f, T1q, T5Q, T5R; T1f = T19 + T1e; T1q = T1k + T1p; T1r = T1f + T1q; T5P = T1f - T1q; T5Q = T3t + T3u; T5R = T3z + T3A; T5S = T5Q - T5R; T6E = T5Q + T5R; } { E T3v, T3w, T3y, T3B; T3v = T3t - T3u; T3w = T1k - T1p; T3x = T3v - T3w; T52 = T3v + T3w; T3y = T19 - T1e; T3B = T3z - T3A; T3C = T3y + T3B; T51 = T3y - T3B; } } { E T1V, T3M, T20, T3N, T3L, T3O, T26, T3Q, T2b, T3R, T3S, T3T; { E T1S, T1U, T1R, T1T; T1S = cr[WS(rs, 5)]; T1U = ci[WS(rs, 5)]; T1R = W[8]; T1T = W[9]; T1V = FMA(T1R, T1S, T1T * T1U); T3M = FNMS(T1T, T1S, T1R * T1U); } { E T1X, T1Z, T1W, T1Y; T1X = cr[WS(rs, 21)]; T1Z = ci[WS(rs, 21)]; T1W = W[40]; T1Y = W[41]; T20 = FMA(T1W, T1X, T1Y * T1Z); T3N = FNMS(T1Y, T1X, T1W * T1Z); } T3L = T1V - T20; T3O = T3M - T3N; { E T23, T25, T22, T24; T23 = cr[WS(rs, 29)]; T25 = ci[WS(rs, 29)]; T22 = W[56]; T24 = W[57]; T26 = FMA(T22, T23, T24 * T25); T3Q = FNMS(T24, T23, T22 * T25); } { E T28, T2a, T27, T29; T28 = cr[WS(rs, 13)]; T2a = ci[WS(rs, 13)]; T27 = W[24]; T29 = W[25]; T2b = FMA(T27, T28, T29 * T2a); T3R = FNMS(T29, T28, T27 * T2a); } T3S = T3Q - T3R; T3T = T26 - T2b; { E T21, T2c, T62, T63; T21 = T1V + T20; T2c = T26 + T2b; T2d = T21 + T2c; T5Z = T21 - T2c; T62 = T3Q + T3R; T63 = T3M + T3N; T64 = T62 - T63; T6K = T63 + T62; } { E T3P, T3U, T42, T43; T3P = T3L + T3O; T3U = T3S - T3T; T3V = KP707106781 * (T3P - T3U); T5a = KP707106781 * (T3P + T3U); T42 = T3T + T3S; T43 = T3L - T3O; T44 = KP707106781 * (T42 - T43); T57 = KP707106781 * (T43 + T42); } } { E T2G, T4i, T2L, T4j, T4h, T4k, T2R, T4d, T2W, T4e, T4c, T4f; { E T2D, T2F, T2C, T2E; T2D = cr[WS(rs, 3)]; T2F = ci[WS(rs, 3)]; T2C = W[4]; T2E = W[5]; T2G = FMA(T2C, T2D, T2E * T2F); T4i = FNMS(T2E, T2D, T2C * T2F); } { E T2I, T2K, T2H, T2J; T2I = cr[WS(rs, 19)]; T2K = ci[WS(rs, 19)]; T2H = W[36]; T2J = W[37]; T2L = FMA(T2H, T2I, T2J * T2K); T4j = FNMS(T2J, T2I, T2H * T2K); } T4h = T2G - T2L; T4k = T4i - T4j; { E T2O, T2Q, T2N, T2P; T2O = cr[WS(rs, 27)]; T2Q = ci[WS(rs, 27)]; T2N = W[52]; T2P = W[53]; T2R = FMA(T2N, T2O, T2P * T2Q); T4d = FNMS(T2P, T2O, T2N * T2Q); } { E T2T, T2V, T2S, T2U; T2T = cr[WS(rs, 11)]; T2V = ci[WS(rs, 11)]; T2S = W[20]; T2U = W[21]; T2W = FMA(T2S, T2T, T2U * T2V); T4e = FNMS(T2U, T2T, T2S * T2V); } T4c = T2R - T2W; T4f = T4d - T4e; { E T2M, T2X, T68, T69; T2M = T2G + T2L; T2X = T2R + T2W; T2Y = T2M + T2X; T6f = T2M - T2X; T68 = T4d + T4e; T69 = T4i + T4j; T6a = T68 - T69; T6P = T69 + T68; } { E T4g, T4l, T4t, T4u; T4g = T4c + T4f; T4l = T4h - T4k; T4m = KP707106781 * (T4g - T4l); T5e = KP707106781 * (T4l + T4g); T4t = T4h + T4k; T4u = T4f - T4c; T4v = KP707106781 * (T4t - T4u); T5h = KP707106781 * (T4t + T4u); } } { E T1t, T6X, T7a, T7c, T30, T7b, T70, T71; { E TH, T1s, T72, T79; TH = Tj + TG; T1s = T14 + T1r; T1t = TH + T1s; T6X = TH - T1s; T72 = T6F + T6E; T79 = T73 + T78; T7a = T72 + T79; T7c = T79 - T72; } { E T2e, T2Z, T6Y, T6Z; T2e = T1Q + T2d; T2Z = T2B + T2Y; T30 = T2e + T2Z; T7b = T2Z - T2e; T6Y = T6O + T6P; T6Z = T6J + T6K; T70 = T6Y - T6Z; T71 = T6Z + T6Y; } ci[WS(rs, 15)] = T1t - T30; cr[WS(rs, 24)] = T7b - T7c; ci[WS(rs, 23)] = T7b + T7c; cr[0] = T1t + T30; cr[WS(rs, 8)] = T6X - T70; cr[WS(rs, 16)] = T71 - T7a; ci[WS(rs, 31)] = T71 + T7a; ci[WS(rs, 7)] = T6X + T70; } { E T4X, T5p, T7D, T7J, T54, T7y, T5z, T5D, T5c, T5m, T5s, T7I, T5w, T5C, T5j; E T5n, T4W, T7z; T4W = KP707106781 * (T4U + T4V); T4X = T4T - T4W; T5p = T4T + T4W; T7z = KP707106781 * (T3a - T3f); T7D = T7z + T7C; T7J = T7C - T7z; { E T50, T53, T5x, T5y; T50 = FMA(KP923879532, T4Y, KP382683432 * T4Z); T53 = FNMS(KP923879532, T52, KP382683432 * T51); T54 = T50 + T53; T7y = T50 - T53; T5x = T5d + T5e; T5y = T5g + T5h; T5z = FNMS(KP980785280, T5y, KP195090322 * T5x); T5D = FMA(KP980785280, T5x, KP195090322 * T5y); } { E T58, T5b, T5q, T5r; T58 = T56 - T57; T5b = T59 - T5a; T5c = FMA(KP831469612, T58, KP555570233 * T5b); T5m = FNMS(KP831469612, T5b, KP555570233 * T58); T5q = FNMS(KP382683432, T4Y, KP923879532 * T4Z); T5r = FMA(KP382683432, T52, KP923879532 * T51); T5s = T5q + T5r; T7I = T5r - T5q; } { E T5u, T5v, T5f, T5i; T5u = T56 + T57; T5v = T59 + T5a; T5w = FMA(KP195090322, T5u, KP980785280 * T5v); T5C = FNMS(KP195090322, T5v, KP980785280 * T5u); T5f = T5d - T5e; T5i = T5g - T5h; T5j = FNMS(KP555570233, T5i, KP831469612 * T5f); T5n = FMA(KP555570233, T5f, KP831469612 * T5i); } { E T55, T5k, T7H, T7K; T55 = T4X + T54; T5k = T5c + T5j; ci[WS(rs, 12)] = T55 - T5k; cr[WS(rs, 3)] = T55 + T5k; T7H = T5n - T5m; T7K = T7I + T7J; cr[WS(rs, 19)] = T7H - T7K; ci[WS(rs, 28)] = T7H + T7K; } { E T7L, T7M, T5l, T5o; T7L = T5j - T5c; T7M = T7J - T7I; cr[WS(rs, 27)] = T7L - T7M; ci[WS(rs, 20)] = T7L + T7M; T5l = T4X - T54; T5o = T5m + T5n; cr[WS(rs, 11)] = T5l - T5o; ci[WS(rs, 4)] = T5l + T5o; } { E T5t, T5A, T7x, T7E; T5t = T5p - T5s; T5A = T5w + T5z; ci[WS(rs, 8)] = T5t - T5A; cr[WS(rs, 7)] = T5t + T5A; T7x = T5z - T5w; T7E = T7y + T7D; cr[WS(rs, 31)] = T7x - T7E; ci[WS(rs, 16)] = T7x + T7E; } { E T7F, T7G, T5B, T5E; T7F = T5D - T5C; T7G = T7D - T7y; cr[WS(rs, 23)] = T7F - T7G; ci[WS(rs, 24)] = T7F + T7G; T5B = T5p + T5s; T5E = T5C + T5D; cr[WS(rs, 15)] = T5B - T5E; ci[0] = T5B + T5E; } } { E T6H, T6T, T7g, T7i, T6M, T6U, T6R, T6V; { E T6D, T6G, T7e, T7f; T6D = Tj - TG; T6G = T6E - T6F; T6H = T6D - T6G; T6T = T6D + T6G; T7e = T14 - T1r; T7f = T78 - T73; T7g = T7e + T7f; T7i = T7f - T7e; } { E T6I, T6L, T6N, T6Q; T6I = T1Q - T2d; T6L = T6J - T6K; T6M = T6I + T6L; T6U = T6I - T6L; T6N = T2B - T2Y; T6Q = T6O - T6P; T6R = T6N - T6Q; T6V = T6N + T6Q; } { E T6S, T7h, T6W, T7d; T6S = KP707106781 * (T6M + T6R); ci[WS(rs, 11)] = T6H - T6S; cr[WS(rs, 4)] = T6H + T6S; T7h = KP707106781 * (T6V - T6U); cr[WS(rs, 20)] = T7h - T7i; ci[WS(rs, 27)] = T7h + T7i; T6W = KP707106781 * (T6U + T6V); cr[WS(rs, 12)] = T6T - T6W; ci[WS(rs, 3)] = T6T + T6W; T7d = KP707106781 * (T6R - T6M); cr[WS(rs, 28)] = T7d - T7g; ci[WS(rs, 19)] = T7d + T7g; } } { E T5J, T7n, T7t, T6n, T5U, T7k, T6x, T6B, T6q, T7s, T66, T6k, T6u, T6A, T6h; E T6l; { E T5O, T5T, T60, T65; T5J = T5F - T5I; T7n = T7l + T7m; T7t = T7m - T7l; T6n = T5F + T5I; T5O = T5K + T5N; T5T = T5P - T5S; T5U = KP707106781 * (T5O + T5T); T7k = KP707106781 * (T5O - T5T); { E T6v, T6w, T6o, T6p; T6v = T6e + T6f; T6w = T67 + T6a; T6x = FMA(KP382683432, T6v, KP923879532 * T6w); T6B = FNMS(KP923879532, T6v, KP382683432 * T6w); T6o = T5K - T5N; T6p = T5P + T5S; T6q = KP707106781 * (T6o + T6p); T7s = KP707106781 * (T6p - T6o); } T60 = T5Y - T5Z; T65 = T61 - T64; T66 = FMA(KP382683432, T60, KP923879532 * T65); T6k = FNMS(KP923879532, T60, KP382683432 * T65); { E T6s, T6t, T6b, T6g; T6s = T61 + T64; T6t = T5Y + T5Z; T6u = FNMS(KP382683432, T6t, KP923879532 * T6s); T6A = FMA(KP923879532, T6t, KP382683432 * T6s); T6b = T67 - T6a; T6g = T6e - T6f; T6h = FNMS(KP382683432, T6g, KP923879532 * T6b); T6l = FMA(KP923879532, T6g, KP382683432 * T6b); } } { E T5V, T6i, T7r, T7u; T5V = T5J + T5U; T6i = T66 + T6h; ci[WS(rs, 13)] = T5V - T6i; cr[WS(rs, 2)] = T5V + T6i; T7r = T6l - T6k; T7u = T7s + T7t; cr[WS(rs, 18)] = T7r - T7u; ci[WS(rs, 29)] = T7r + T7u; } { E T7v, T7w, T6j, T6m; T7v = T6h - T66; T7w = T7t - T7s; cr[WS(rs, 26)] = T7v - T7w; ci[WS(rs, 21)] = T7v + T7w; T6j = T5J - T5U; T6m = T6k + T6l; cr[WS(rs, 10)] = T6j - T6m; ci[WS(rs, 5)] = T6j + T6m; } { E T6r, T6y, T7j, T7o; T6r = T6n + T6q; T6y = T6u + T6x; cr[WS(rs, 14)] = T6r - T6y; ci[WS(rs, 1)] = T6r + T6y; T7j = T6B - T6A; T7o = T7k + T7n; cr[WS(rs, 30)] = T7j - T7o; ci[WS(rs, 17)] = T7j + T7o; } { E T7p, T7q, T6z, T6C; T7p = T6x - T6u; T7q = T7n - T7k; cr[WS(rs, 22)] = T7p - T7q; ci[WS(rs, 25)] = T7p + T7q; T6z = T6n - T6q; T6C = T6A + T6B; ci[WS(rs, 9)] = T6z - T6C; cr[WS(rs, 6)] = T6z + T6C; } } { E T3h, T4D, T7R, T7X, T3E, T7O, T4N, T4R, T46, T4A, T4G, T7W, T4K, T4Q, T4x; E T4B, T3g, T7P; T3g = KP707106781 * (T3a + T3f); T3h = T35 - T3g; T4D = T35 + T3g; T7P = KP707106781 * (T4V - T4U); T7R = T7P + T7Q; T7X = T7Q - T7P; { E T3s, T3D, T4L, T4M; T3s = FNMS(KP923879532, T3r, KP382683432 * T3m); T3D = FMA(KP923879532, T3x, KP382683432 * T3C); T3E = T3s + T3D; T7O = T3D - T3s; T4L = T4s + T4v; T4M = T4b + T4m; T4N = FNMS(KP195090322, T4M, KP980785280 * T4L); T4R = FMA(KP980785280, T4M, KP195090322 * T4L); } { E T3W, T45, T4E, T4F; T3W = T3K - T3V; T45 = T41 - T44; T46 = FNMS(KP555570233, T45, KP831469612 * T3W); T4A = FMA(KP831469612, T45, KP555570233 * T3W); T4E = FMA(KP382683432, T3r, KP923879532 * T3m); T4F = FNMS(KP382683432, T3x, KP923879532 * T3C); T4G = T4E + T4F; T7W = T4E - T4F; } { E T4I, T4J, T4n, T4w; T4I = T41 + T44; T4J = T3K + T3V; T4K = FMA(KP195090322, T4I, KP980785280 * T4J); T4Q = FNMS(KP980785280, T4I, KP195090322 * T4J); T4n = T4b - T4m; T4w = T4s - T4v; T4x = FMA(KP555570233, T4n, KP831469612 * T4w); T4B = FNMS(KP831469612, T4n, KP555570233 * T4w); } { E T3F, T4y, T7V, T7Y; T3F = T3h + T3E; T4y = T46 + T4x; cr[WS(rs, 13)] = T3F - T4y; ci[WS(rs, 2)] = T3F + T4y; T7V = T4B - T4A; T7Y = T7W + T7X; cr[WS(rs, 29)] = T7V - T7Y; ci[WS(rs, 18)] = T7V + T7Y; } { E T7Z, T80, T4z, T4C; T7Z = T4x - T46; T80 = T7X - T7W; cr[WS(rs, 21)] = T7Z - T80; ci[WS(rs, 26)] = T7Z + T80; T4z = T3h - T3E; T4C = T4A + T4B; ci[WS(rs, 10)] = T4z - T4C; cr[WS(rs, 5)] = T4z + T4C; } { E T4H, T4O, T7N, T7S; T4H = T4D + T4G; T4O = T4K + T4N; ci[WS(rs, 14)] = T4H - T4O; cr[WS(rs, 1)] = T4H + T4O; T7N = T4R - T4Q; T7S = T7O + T7R; cr[WS(rs, 17)] = T7N - T7S; ci[WS(rs, 30)] = T7N + T7S; } { E T7T, T7U, T4P, T4S; T7T = T4N - T4K; T7U = T7R - T7O; cr[WS(rs, 25)] = T7T - T7U; ci[WS(rs, 22)] = T7T + T7U; T4P = T4D - T4G; T4S = T4Q + T4R; cr[WS(rs, 9)] = T4P - T4S; ci[WS(rs, 6)] = T4P + T4S; } } } } } static const tw_instr twinstr[] = { {TW_FULL, 1, 32}, {TW_NEXT, 1, 0} }; static const hc2hc_desc desc = { 32, "hf_32", twinstr, &GENUS, {340, 114, 94, 0} }; void X(codelet_hf_32) (planner *p) { X(khc2hc_register) (p, hf_32, &desc); } #endif /* HAVE_FMA */ fftw-3.3.3/rdft/scalar/r2cf/hf_64.c0000644000175000017500000032402112054410712013517 00000000000000/* * Copyright (c) 2003, 2007-11 Matteo Frigo * Copyright (c) 2003, 2007-11 Massachusetts Institute of Technology * * This program is free software; you can redistribute it and/or modify * it under the terms of the GNU General Public License as published by * the Free Software Foundation; either version 2 of the License, or * (at your option) any later version. * * This program is distributed in the hope that it will be useful, * but WITHOUT ANY WARRANTY; without even the implied warranty of * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the * GNU General Public License for more details. * * You should have received a copy of the GNU General Public License * along with this program; if not, write to the Free Software * Foundation, Inc., 51 Franklin Street, Fifth Floor, Boston, MA 02110-1301 USA * */ /* This file was automatically generated --- DO NOT EDIT */ /* Generated on Sun Nov 25 07:39:55 EST 2012 */ #include "codelet-rdft.h" #ifdef HAVE_FMA /* Generated by: ../../../genfft/gen_hc2hc.native -fma -reorder-insns -schedule-for-pipeline -compact -variables 4 -pipeline-latency 4 -n 64 -dit -name hf_64 -include hf.h */ /* * This function contains 1038 FP additions, 644 FP multiplications, * (or, 520 additions, 126 multiplications, 518 fused multiply/add), * 246 stack variables, 15 constants, and 256 memory accesses */ #include "hf.h" static void hf_64(R *cr, R *ci, const R *W, stride rs, INT mb, INT me, INT ms) { DK(KP881921264, +0.881921264348355029712756863660388349508442621); DK(KP956940335, +0.956940335732208864935797886980269969482849206); DK(KP773010453, +0.773010453362736960810906609758469800971041293); DK(KP995184726, +0.995184726672196886244836953109479921575474869); DK(KP831469612, +0.831469612302545237078788377617905756738560812); DK(KP980785280, +0.980785280403230449126182236134239036973933731); DK(KP668178637, +0.668178637919298919997757686523080761552472251); DK(KP534511135, +0.534511135950791641089685961295362908582039528); DK(KP303346683, +0.303346683607342391675883946941299872384187453); DK(KP098491403, +0.098491403357164253077197521291327432293052451); DK(KP820678790, +0.820678790828660330972281985331011598767386482); DK(KP923879532, +0.923879532511286756128183189396788286822416626); DK(KP198912367, +0.198912367379658006911597622644676228597850501); DK(KP707106781, +0.707106781186547524400844362104849039284835938); DK(KP414213562, +0.414213562373095048801688724209698078569671875); { INT m; for (m = mb, W = W + ((mb - 1) * 126); m < me; m = m + 1, cr = cr + ms, ci = ci - ms, W = W + 126, MAKE_VOLATILE_STRIDE(128, rs)) { E Tku, Tky, Tkt, Tkx; { E TiV, Tjm, T7e, TcA, TjR, Tkl, Tm, TeM, T7Q, TcI, TeZ, Thr, T1G, TeW, TcJ; E T7X, T87, TcN, Tf5, Thw, T29, Tf8, TcQ, T8u, Taq, Tdm, Tgc, ThX, T5K, TfS; E Tdx, Tbj, TcB, T7l, TiP, TeP, Tjl, TN, TcC, T7s, T7B, TcF, TeU, Ths, T7I; E TcG, T1f, TeR, T8G, TcU, Tfg, ThB, T32, Tfj, TcX, T93, T9h, Td3, TfK, ThM; E T3X, Tfr, Tde, Taa, Thx, Tfb, Tf6, T2A, T8x, TcO, T8m, TcR, Tfm, ThC, T3t; E Tfh, T96, TcV, T8V, TcY, ThN, Tfu, TfL, T4o, Tad, Td4, T9w, Tdf, TfV, ThY; E T6b, Tg9, Tbm, Tdn, TaF, Tdy, ThJ, T4Q, TfN, TfA, Taf, T9M, Td8, Tdh, ThI; E T5h, TfO, TfF, Tag, Ta1, Tdb, Tdi, ThU, T6D, Tgf, Tg1, Tbo, TaV, Tdr, TdA; E Tb2, Tds, Tg5, ThT, Tg2, T74, Tdt, Tb9; { E T7a, Te, T78, T8, TjQ, TiU, T7c, Tk; { E T1, TiT, TiS, T7, Tg, Tj, Tf, Ti, T7b, Th; T1 = cr[0]; TiT = ci[0]; { E T3, T6, T2, T5; T3 = cr[WS(rs, 32)]; T6 = ci[WS(rs, 32)]; T2 = W[62]; T5 = W[63]; { E Ta, Td, Tc, T79, Tb, TiR, T4, T9; Ta = cr[WS(rs, 16)]; Td = ci[WS(rs, 16)]; TiR = T2 * T6; T4 = T2 * T3; T9 = W[30]; Tc = W[31]; TiS = FNMS(T5, T3, TiR); T7 = FMA(T5, T6, T4); T79 = T9 * Td; Tb = T9 * Ta; Tg = cr[WS(rs, 48)]; Tj = ci[WS(rs, 48)]; T7a = FNMS(Tc, Ta, T79); Te = FMA(Tc, Td, Tb); Tf = W[94]; Ti = W[95]; } } T78 = T1 - T7; T8 = T1 + T7; TjQ = TiT - TiS; TiU = TiS + TiT; T7b = Tf * Tj; Th = Tf * Tg; T7c = FNMS(Ti, Tg, T7b); Tk = FMA(Ti, Tj, Th); } { E T7S, T1l, T7O, T1E, T1u, T1x, T1w, T7U, T1r, T7L, T1v; { E T1A, T1D, T1C, T7N, T1B; { E T1h, T1k, T1g, T1j, T7R, T1i, T1z; T1h = cr[WS(rs, 60)]; T1k = ci[WS(rs, 60)]; { E T7d, TiQ, Tl, TjP; T7d = T7a - T7c; TiQ = T7a + T7c; Tl = Te + Tk; TjP = Te - Tk; TiV = TiQ + TiU; Tjm = TiU - TiQ; T7e = T78 - T7d; TcA = T78 + T7d; TjR = TjP + TjQ; Tkl = TjQ - TjP; Tm = T8 + Tl; TeM = T8 - Tl; T1g = W[118]; } T1j = W[119]; T1A = cr[WS(rs, 44)]; T1D = ci[WS(rs, 44)]; T7R = T1g * T1k; T1i = T1g * T1h; T1z = W[86]; T1C = W[87]; T7S = FNMS(T1j, T1h, T7R); T1l = FMA(T1j, T1k, T1i); T7N = T1z * T1D; T1B = T1z * T1A; } { E T1n, T1q, T1m, T1p, T7T, T1o, T1t; T1n = cr[WS(rs, 28)]; T1q = ci[WS(rs, 28)]; T7O = FNMS(T1C, T1A, T7N); T1E = FMA(T1C, T1D, T1B); T1m = W[54]; T1p = W[55]; T1u = cr[WS(rs, 12)]; T1x = ci[WS(rs, 12)]; T7T = T1m * T1q; T1o = T1m * T1n; T1t = W[22]; T1w = W[23]; T7U = FNMS(T1p, T1n, T7T); T1r = FMA(T1p, T1q, T1o); T7L = T1t * T1x; T1v = T1t * T1u; } } { E T7V, TeX, T1s, T7K, T7M, T1y; T7V = T7S - T7U; TeX = T7S + T7U; T1s = T1l + T1r; T7K = T1l - T1r; T7M = FNMS(T1w, T1u, T7L); T1y = FMA(T1w, T1x, T1v); { E TeY, T7P, T7W, T1F; TeY = T7M + T7O; T7P = T7M - T7O; T7W = T1y - T1E; T1F = T1y + T1E; T7Q = T7K - T7P; TcI = T7K + T7P; TeZ = TeX - TeY; Thr = TeX + TeY; T1G = T1s + T1F; TeW = T1s - T1F; TcJ = T7V - T7W; T7X = T7V + T7W; } } } } { E T8p, T1O, T85, T27, T1X, T20, T1Z, T8r, T1U, T82, T1Y; { E T23, T26, T25, T84, T24; { E T1K, T1N, T1J, T1M, T8o, T1L, T22; T1K = cr[WS(rs, 2)]; T1N = ci[WS(rs, 2)]; T1J = W[2]; T1M = W[3]; T23 = cr[WS(rs, 50)]; T26 = ci[WS(rs, 50)]; T8o = T1J * T1N; T1L = T1J * T1K; T22 = W[98]; T25 = W[99]; T8p = FNMS(T1M, T1K, T8o); T1O = FMA(T1M, T1N, T1L); T84 = T22 * T26; T24 = T22 * T23; } { E T1Q, T1T, T1P, T1S, T8q, T1R, T1W; T1Q = cr[WS(rs, 34)]; T1T = ci[WS(rs, 34)]; T85 = FNMS(T25, T23, T84); T27 = FMA(T25, T26, T24); T1P = W[66]; T1S = W[67]; T1X = cr[WS(rs, 18)]; T20 = ci[WS(rs, 18)]; T8q = T1P * T1T; T1R = T1P * T1Q; T1W = W[34]; T1Z = W[35]; T8r = FNMS(T1S, T1Q, T8q); T1U = FMA(T1S, T1T, T1R); T82 = T1W * T20; T1Y = T1W * T1X; } } { E T8s, Tf3, T1V, T81, T83, T21; T8s = T8p - T8r; Tf3 = T8p + T8r; T1V = T1O + T1U; T81 = T1O - T1U; T83 = FNMS(T1Z, T1X, T82); T21 = FMA(T1Z, T20, T1Y); { E Tf4, T86, T8t, T28; Tf4 = T83 + T85; T86 = T83 - T85; T8t = T21 - T27; T28 = T21 + T27; T87 = T81 - T86; TcN = T81 + T86; Tf5 = Tf3 - Tf4; Thw = Tf3 + Tf4; T29 = T1V + T28; Tf8 = T1V - T28; TcQ = T8s - T8t; T8u = T8s + T8t; } } } { E Tbf, T5p, Tao, T5I, T5y, T5B, T5A, Tbh, T5v, Tal, T5z; { E T5E, T5H, T5G, Tan, T5F; { E T5l, T5o, T5k, T5n, Tbe, T5m, T5D; T5l = cr[WS(rs, 63)]; T5o = ci[WS(rs, 63)]; T5k = W[124]; T5n = W[125]; T5E = cr[WS(rs, 47)]; T5H = ci[WS(rs, 47)]; Tbe = T5k * T5o; T5m = T5k * T5l; T5D = W[92]; T5G = W[93]; Tbf = FNMS(T5n, T5l, Tbe); T5p = FMA(T5n, T5o, T5m); Tan = T5D * T5H; T5F = T5D * T5E; } { E T5r, T5u, T5q, T5t, Tbg, T5s, T5x; T5r = cr[WS(rs, 31)]; T5u = ci[WS(rs, 31)]; Tao = FNMS(T5G, T5E, Tan); T5I = FMA(T5G, T5H, T5F); T5q = W[60]; T5t = W[61]; T5y = cr[WS(rs, 15)]; T5B = ci[WS(rs, 15)]; Tbg = T5q * T5u; T5s = T5q * T5r; T5x = W[28]; T5A = W[29]; Tbh = FNMS(T5t, T5r, Tbg); T5v = FMA(T5t, T5u, T5s); Tal = T5x * T5B; T5z = T5x * T5y; } } { E Tbi, Tga, T5w, Tak, Tam, T5C; Tbi = Tbf - Tbh; Tga = Tbf + Tbh; T5w = T5p + T5v; Tak = T5p - T5v; Tam = FNMS(T5A, T5y, Tal); T5C = FMA(T5A, T5B, T5z); { E Tgb, Tap, T5J, Tbd; Tgb = Tam + Tao; Tap = Tam - Tao; T5J = T5C + T5I; Tbd = T5I - T5C; Taq = Tak - Tap; Tdm = Tak + Tap; Tgc = Tga - Tgb; ThX = Tga + Tgb; T5K = T5w + T5J; TfS = T5w - T5J; Tdx = Tbi + Tbd; Tbj = Tbd - Tbi; } } } { E T7z, T1d, T7G, TeS, T11, T7v, T7x, T17, T7r, T7m; { E T7h, Ts, T7q, TL, TB, TE, TD, T7j, Ty, T7n, TC; { E TH, TK, TJ, T7p, TI; { E To, Tr, Tn, Tq, T7g, Tp, TG; To = cr[WS(rs, 8)]; Tr = ci[WS(rs, 8)]; Tn = W[14]; Tq = W[15]; TH = cr[WS(rs, 24)]; TK = ci[WS(rs, 24)]; T7g = Tn * Tr; Tp = Tn * To; TG = W[46]; TJ = W[47]; T7h = FNMS(Tq, To, T7g); Ts = FMA(Tq, Tr, Tp); T7p = TG * TK; TI = TG * TH; } { E Tu, Tx, Tt, Tw, T7i, Tv, TA; Tu = cr[WS(rs, 40)]; Tx = ci[WS(rs, 40)]; T7q = FNMS(TJ, TH, T7p); TL = FMA(TJ, TK, TI); Tt = W[78]; Tw = W[79]; TB = cr[WS(rs, 56)]; TE = ci[WS(rs, 56)]; T7i = Tt * Tx; Tv = Tt * Tu; TA = W[110]; TD = W[111]; T7j = FNMS(Tw, Tu, T7i); Ty = FMA(Tw, Tx, Tv); T7n = TA * TE; TC = TA * TB; } } { E T7k, TeO, Tz, T7f, T7o, TF, TeN, TM; T7k = T7h - T7j; TeO = T7h + T7j; Tz = Ts + Ty; T7f = Ts - Ty; T7o = FNMS(TD, TB, T7n); TF = FMA(TD, TE, TC); T7r = T7o - T7q; TeN = T7o + T7q; TM = TF + TL; T7m = TF - TL; TcB = T7f + T7k; T7l = T7f - T7k; TiP = TeO + TeN; TeP = TeN - TeO; Tjl = Tz - TM; TN = Tz + TM; } } { E T7D, TU, T13, T16, T7F, T10, T12, T15, T7w, T14; { E T19, T1c, T18, T1b; { E TQ, TT, TS, T7C, TR, TP; TQ = cr[WS(rs, 4)]; TT = ci[WS(rs, 4)]; TP = W[6]; TcC = T7m - T7r; T7s = T7m + T7r; TS = W[7]; T7C = TP * TT; TR = TP * TQ; T19 = cr[WS(rs, 52)]; T1c = ci[WS(rs, 52)]; T7D = FNMS(TS, TQ, T7C); TU = FMA(TS, TT, TR); T18 = W[102]; T1b = W[103]; } { E TW, TZ, TY, T7E, TX, T7y, T1a, TV; TW = cr[WS(rs, 36)]; TZ = ci[WS(rs, 36)]; T7y = T18 * T1c; T1a = T18 * T19; TV = W[70]; TY = W[71]; T7z = FNMS(T1b, T19, T7y); T1d = FMA(T1b, T1c, T1a); T7E = TV * TZ; TX = TV * TW; T13 = cr[WS(rs, 20)]; T16 = ci[WS(rs, 20)]; T7F = FNMS(TY, TW, T7E); T10 = FMA(TY, TZ, TX); T12 = W[38]; T15 = W[39]; } } T7G = T7D - T7F; TeS = T7D + T7F; T11 = TU + T10; T7v = TU - T10; T7w = T12 * T16; T14 = T12 * T13; T7x = FNMS(T15, T13, T7w); T17 = FMA(T15, T16, T14); } { E T8Y, T2H, T8E, T30, T2Q, T2T, T2S, T90, T2N, T8B, T2R; { E T2W, T2Z, T2Y, T8D, T2X; { E T2D, T2G, T2C, T2F, T8X, T2E, T2V; T2D = cr[WS(rs, 62)]; T2G = ci[WS(rs, 62)]; { E TeT, T7A, T1e, T7H; TeT = T7x + T7z; T7A = T7x - T7z; T1e = T17 + T1d; T7H = T17 - T1d; T7B = T7v - T7A; TcF = T7v + T7A; TeU = TeS - TeT; Ths = TeS + TeT; T7I = T7G + T7H; TcG = T7G - T7H; T1f = T11 + T1e; TeR = T11 - T1e; T2C = W[122]; } T2F = W[123]; T2W = cr[WS(rs, 46)]; T2Z = ci[WS(rs, 46)]; T8X = T2C * T2G; T2E = T2C * T2D; T2V = W[90]; T2Y = W[91]; T8Y = FNMS(T2F, T2D, T8X); T2H = FMA(T2F, T2G, T2E); T8D = T2V * T2Z; T2X = T2V * T2W; } { E T2J, T2M, T2I, T2L, T8Z, T2K, T2P; T2J = cr[WS(rs, 30)]; T2M = ci[WS(rs, 30)]; T8E = FNMS(T2Y, T2W, T8D); T30 = FMA(T2Y, T2Z, T2X); T2I = W[58]; T2L = W[59]; T2Q = cr[WS(rs, 14)]; T2T = ci[WS(rs, 14)]; T8Z = T2I * T2M; T2K = T2I * T2J; T2P = W[26]; T2S = W[27]; T90 = FNMS(T2L, T2J, T8Z); T2N = FMA(T2L, T2M, T2K); T8B = T2P * T2T; T2R = T2P * T2Q; } } { E T91, Tfe, T2O, T8A, T8C, T2U; T91 = T8Y - T90; Tfe = T8Y + T90; T2O = T2H + T2N; T8A = T2H - T2N; T8C = FNMS(T2S, T2Q, T8B); T2U = FMA(T2S, T2T, T2R); { E Tff, T8F, T92, T31; Tff = T8C + T8E; T8F = T8C - T8E; T92 = T2U - T30; T31 = T2U + T30; T8G = T8A - T8F; TcU = T8A + T8F; Tfg = Tfe - Tff; ThB = Tfe + Tff; T32 = T2O + T31; Tfj = T2O - T31; TcX = T91 - T92; T93 = T91 + T92; } } } { E Ta5, T3C, T9f, T3V, T3L, T3O, T3N, Ta7, T3I, T9c, T3M; { E T3R, T3U, T3T, T9e, T3S; { E T3y, T3B, T3x, T3A, Ta4, T3z, T3Q; T3y = cr[WS(rs, 1)]; T3B = ci[WS(rs, 1)]; T3x = W[0]; T3A = W[1]; T3R = cr[WS(rs, 49)]; T3U = ci[WS(rs, 49)]; Ta4 = T3x * T3B; T3z = T3x * T3y; T3Q = W[96]; T3T = W[97]; Ta5 = FNMS(T3A, T3y, Ta4); T3C = FMA(T3A, T3B, T3z); T9e = T3Q * T3U; T3S = T3Q * T3R; } { E T3E, T3H, T3D, T3G, Ta6, T3F, T3K; T3E = cr[WS(rs, 33)]; T3H = ci[WS(rs, 33)]; T9f = FNMS(T3T, T3R, T9e); T3V = FMA(T3T, T3U, T3S); T3D = W[64]; T3G = W[65]; T3L = cr[WS(rs, 17)]; T3O = ci[WS(rs, 17)]; Ta6 = T3D * T3H; T3F = T3D * T3E; T3K = W[32]; T3N = W[33]; Ta7 = FNMS(T3G, T3E, Ta6); T3I = FMA(T3G, T3H, T3F); T9c = T3K * T3O; T3M = T3K * T3L; } } { E Ta8, TfI, T3J, T9b, T9d, T3P; Ta8 = Ta5 - Ta7; TfI = Ta5 + Ta7; T3J = T3C + T3I; T9b = T3C - T3I; T9d = FNMS(T3N, T3L, T9c); T3P = FMA(T3N, T3O, T3M); { E TfJ, T9g, Ta9, T3W; TfJ = T9d + T9f; T9g = T9d - T9f; Ta9 = T3P - T3V; T3W = T3P + T3V; T9h = T9b - T9g; Td3 = T9b + T9g; TfK = TfI - TfJ; ThM = TfI + TfJ; T3X = T3J + T3W; Tfr = T3J - T3W; Tde = Ta8 - Ta9; Taa = Ta8 + Ta9; } } } } { E TaC, T69, Taw, TfU, T5X, Tar, TaA, T63; { E T8S, T3r, T8M, Tfl, T3f, T8H, T8Q, T3l; { E T8k, T8f, T8v, T8e; { E T8a, T2f, T8j, T2y, T2o, T2r, T2q, T8c, T2l, T8g, T2p; { E T2u, T2x, T2w, T8i, T2v; { E T2b, T2e, T2a, T2d, T89, T2c, T2t; T2b = cr[WS(rs, 10)]; T2e = ci[WS(rs, 10)]; T2a = W[18]; T2d = W[19]; T2u = cr[WS(rs, 26)]; T2x = ci[WS(rs, 26)]; T89 = T2a * T2e; T2c = T2a * T2b; T2t = W[50]; T2w = W[51]; T8a = FNMS(T2d, T2b, T89); T2f = FMA(T2d, T2e, T2c); T8i = T2t * T2x; T2v = T2t * T2u; } { E T2h, T2k, T2g, T2j, T8b, T2i, T2n; T2h = cr[WS(rs, 42)]; T2k = ci[WS(rs, 42)]; T8j = FNMS(T2w, T2u, T8i); T2y = FMA(T2w, T2x, T2v); T2g = W[82]; T2j = W[83]; T2o = cr[WS(rs, 58)]; T2r = ci[WS(rs, 58)]; T8b = T2g * T2k; T2i = T2g * T2h; T2n = W[114]; T2q = W[115]; T8c = FNMS(T2j, T2h, T8b); T2l = FMA(T2j, T2k, T2i); T8g = T2n * T2r; T2p = T2n * T2o; } } { E T8d, Tfa, T2m, T88, T8h, T2s, Tf9, T2z; T8d = T8a - T8c; Tfa = T8a + T8c; T2m = T2f + T2l; T88 = T2f - T2l; T8h = FNMS(T2q, T2o, T8g); T2s = FMA(T2q, T2r, T2p); T8k = T8h - T8j; Tf9 = T8h + T8j; T2z = T2s + T2y; T8f = T2s - T2y; T8v = T88 + T8d; T8e = T88 - T8d; Thx = Tfa + Tf9; Tfb = Tf9 - Tfa; Tf6 = T2m - T2z; T2A = T2m + T2z; } } { E T38, T8J, T3h, T3k, T8L, T3e, T3g, T3j, T8P, T3i; { E T3n, T3q, T3m, T3p; { E T34, T37, T33, T8w, T8l, T36, T8I, T35; T34 = cr[WS(rs, 6)]; T37 = ci[WS(rs, 6)]; T33 = W[10]; T8w = T8k - T8f; T8l = T8f + T8k; T36 = W[11]; T8I = T33 * T37; T35 = T33 * T34; T8x = T8v + T8w; TcO = T8v - T8w; T8m = T8e + T8l; TcR = T8l - T8e; T38 = FMA(T36, T37, T35); T8J = FNMS(T36, T34, T8I); } T3n = cr[WS(rs, 22)]; T3q = ci[WS(rs, 22)]; T3m = W[42]; T3p = W[43]; { E T3a, T3d, T3c, T8K, T3b, T8R, T3o, T39; T3a = cr[WS(rs, 38)]; T3d = ci[WS(rs, 38)]; T8R = T3m * T3q; T3o = T3m * T3n; T39 = W[74]; T3c = W[75]; T8S = FNMS(T3p, T3n, T8R); T3r = FMA(T3p, T3q, T3o); T8K = T39 * T3d; T3b = T39 * T3a; T3h = cr[WS(rs, 54)]; T3k = ci[WS(rs, 54)]; T8L = FNMS(T3c, T3a, T8K); T3e = FMA(T3c, T3d, T3b); T3g = W[106]; T3j = W[107]; } } T8M = T8J - T8L; Tfl = T8J + T8L; T3f = T38 + T3e; T8H = T38 - T3e; T8P = T3g * T3k; T3i = T3g * T3h; T8Q = FNMS(T3j, T3h, T8P); T3l = FMA(T3j, T3k, T3i); } } { E T9u, T9p, Tab, T9o; { E T9k, T43, T9t, T4m, T4c, T4f, T4e, T9m, T49, T9q, T4d; { E T4i, T4l, T4k, T9s, T4j; { E T3Z, T42, T3Y, T41, T9j, T40, T4h; { E T94, T8N, T8T, Tfk, T8O, T3s, T8U, T95; T3Z = cr[WS(rs, 9)]; T94 = T8H + T8M; T8N = T8H - T8M; T8T = T8Q - T8S; Tfk = T8Q + T8S; T8O = T3l - T3r; T3s = T3l + T3r; T42 = ci[WS(rs, 9)]; Tfm = Tfk - Tfl; ThC = Tfl + Tfk; T8U = T8O + T8T; T95 = T8T - T8O; T3t = T3f + T3s; Tfh = T3f - T3s; T96 = T94 + T95; TcV = T94 - T95; T8V = T8N + T8U; TcY = T8U - T8N; T3Y = W[16]; } T41 = W[17]; T4i = cr[WS(rs, 25)]; T4l = ci[WS(rs, 25)]; T9j = T3Y * T42; T40 = T3Y * T3Z; T4h = W[48]; T4k = W[49]; T9k = FNMS(T41, T3Z, T9j); T43 = FMA(T41, T42, T40); T9s = T4h * T4l; T4j = T4h * T4i; } { E T45, T48, T44, T47, T9l, T46, T4b; T45 = cr[WS(rs, 41)]; T48 = ci[WS(rs, 41)]; T9t = FNMS(T4k, T4i, T9s); T4m = FMA(T4k, T4l, T4j); T44 = W[80]; T47 = W[81]; T4c = cr[WS(rs, 57)]; T4f = ci[WS(rs, 57)]; T9l = T44 * T48; T46 = T44 * T45; T4b = W[112]; T4e = W[113]; T9m = FNMS(T47, T45, T9l); T49 = FMA(T47, T48, T46); T9q = T4b * T4f; T4d = T4b * T4c; } } { E T9n, Tft, T4a, T9i, T9r, T4g, Tfs, T4n; T9n = T9k - T9m; Tft = T9k + T9m; T4a = T43 + T49; T9i = T43 - T49; T9r = FNMS(T4e, T4c, T9q); T4g = FMA(T4e, T4f, T4d); T9u = T9r - T9t; Tfs = T9r + T9t; T4n = T4g + T4m; T9p = T4g - T4m; Tab = T9i + T9n; T9o = T9i - T9n; ThN = Tft + Tfs; Tfu = Tfs - Tft; TfL = T4a - T4n; T4o = T4a + T4n; } } { E T5Q, Tat, T5Z, T62, Tav, T5W, T5Y, T61, Taz, T60; { E T65, T68, T64, T67; { E T5M, T5P, T5L, Tac, T9v, T5O, Tas, T5N; T5M = cr[WS(rs, 7)]; T5P = ci[WS(rs, 7)]; T5L = W[12]; Tac = T9u - T9p; T9v = T9p + T9u; T5O = W[13]; Tas = T5L * T5P; T5N = T5L * T5M; Tad = Tab + Tac; Td4 = Tab - Tac; T9w = T9o + T9v; Tdf = T9v - T9o; T5Q = FMA(T5O, T5P, T5N); Tat = FNMS(T5O, T5M, Tas); } T65 = cr[WS(rs, 23)]; T68 = ci[WS(rs, 23)]; T64 = W[44]; T67 = W[45]; { E T5S, T5V, T5U, Tau, T5T, TaB, T66, T5R; T5S = cr[WS(rs, 39)]; T5V = ci[WS(rs, 39)]; TaB = T64 * T68; T66 = T64 * T65; T5R = W[76]; T5U = W[77]; TaC = FNMS(T67, T65, TaB); T69 = FMA(T67, T68, T66); Tau = T5R * T5V; T5T = T5R * T5S; T5Z = cr[WS(rs, 55)]; T62 = ci[WS(rs, 55)]; Tav = FNMS(T5U, T5S, Tau); T5W = FMA(T5U, T5V, T5T); T5Y = W[108]; T61 = W[109]; } } Taw = Tat - Tav; TfU = Tat + Tav; T5X = T5Q + T5W; Tar = T5Q - T5W; Taz = T5Y * T62; T60 = T5Y * T5Z; TaA = FNMS(T61, T5Z, Taz); T63 = FMA(T61, T62, T60); } } } { E T9T, Td9, TfE, TfB, Tda, Ta0; { E T9E, Td6, Tfz, Tfw, Td7, T9L; { E T9G, T4v, T9C, T4O, T4E, T4H, T4G, T9I, T4B, T9z, T4F; { E T4K, T4N, T4M, T9B, T4L; { E T4r, T4u, T4q, T4t, T9F, T4s, T4J; { E Tbl, Tax, TaD, TfT, Tay, T6a, TaE, Tbk; T4r = cr[WS(rs, 5)]; Tbl = Tar + Taw; Tax = Tar - Taw; TaD = TaA - TaC; TfT = TaA + TaC; Tay = T63 - T69; T6a = T63 + T69; T4u = ci[WS(rs, 5)]; TfV = TfT - TfU; ThY = TfU + TfT; TaE = Tay + TaD; Tbk = Tay - TaD; T6b = T5X + T6a; Tg9 = T6a - T5X; Tbm = Tbk - Tbl; Tdn = Tbl + Tbk; TaF = Tax + TaE; Tdy = TaE - Tax; T4q = W[8]; } T4t = W[9]; T4K = cr[WS(rs, 53)]; T4N = ci[WS(rs, 53)]; T9F = T4q * T4u; T4s = T4q * T4r; T4J = W[104]; T4M = W[105]; T9G = FNMS(T4t, T4r, T9F); T4v = FMA(T4t, T4u, T4s); T9B = T4J * T4N; T4L = T4J * T4K; } { E T4x, T4A, T4w, T4z, T9H, T4y, T4D; T4x = cr[WS(rs, 37)]; T4A = ci[WS(rs, 37)]; T9C = FNMS(T4M, T4K, T9B); T4O = FMA(T4M, T4N, T4L); T4w = W[72]; T4z = W[73]; T4E = cr[WS(rs, 21)]; T4H = ci[WS(rs, 21)]; T9H = T4w * T4A; T4y = T4w * T4x; T4D = W[40]; T4G = W[41]; T9I = FNMS(T4z, T4x, T9H); T4B = FMA(T4z, T4A, T4y); T9z = T4D * T4H; T4F = T4D * T4E; } } { E T9J, Tfx, T4C, T9y, T9A, T4I; T9J = T9G - T9I; Tfx = T9G + T9I; T4C = T4v + T4B; T9y = T4v - T4B; T9A = FNMS(T4G, T4E, T9z); T4I = FMA(T4G, T4H, T4F); { E Tfy, T9D, T9K, T4P; Tfy = T9A + T9C; T9D = T9A - T9C; T9K = T4I - T4O; T4P = T4I + T4O; T9E = T9y - T9D; Td6 = T9y + T9D; Tfz = Tfx - Tfy; ThJ = Tfx + Tfy; Tfw = T4C - T4P; T4Q = T4C + T4P; Td7 = T9J - T9K; T9L = T9J + T9K; } } } { E T9V, T4W, T9R, T5f, T55, T58, T57, T9X, T52, T9O, T56; { E T5b, T5e, T5d, T9Q, T5c; { E T4S, T4V, T4R, T4U, T9U, T4T, T5a; T4S = cr[WS(rs, 61)]; TfN = Tfw + Tfz; TfA = Tfw - Tfz; Taf = FMA(KP414213562, T9E, T9L); T9M = FNMS(KP414213562, T9L, T9E); Td8 = FMA(KP414213562, Td7, Td6); Tdh = FNMS(KP414213562, Td6, Td7); T4V = ci[WS(rs, 61)]; T4R = W[120]; T4U = W[121]; T5b = cr[WS(rs, 45)]; T5e = ci[WS(rs, 45)]; T9U = T4R * T4V; T4T = T4R * T4S; T5a = W[88]; T5d = W[89]; T9V = FNMS(T4U, T4S, T9U); T4W = FMA(T4U, T4V, T4T); T9Q = T5a * T5e; T5c = T5a * T5b; } { E T4Y, T51, T4X, T50, T9W, T4Z, T54; T4Y = cr[WS(rs, 29)]; T51 = ci[WS(rs, 29)]; T9R = FNMS(T5d, T5b, T9Q); T5f = FMA(T5d, T5e, T5c); T4X = W[56]; T50 = W[57]; T55 = cr[WS(rs, 13)]; T58 = ci[WS(rs, 13)]; T9W = T4X * T51; T4Z = T4X * T4Y; T54 = W[24]; T57 = W[25]; T9X = FNMS(T50, T4Y, T9W); T52 = FMA(T50, T51, T4Z); T9O = T54 * T58; T56 = T54 * T55; } } { E T9Y, TfC, T53, T9N, T9P, T59; T9Y = T9V - T9X; TfC = T9V + T9X; T53 = T4W + T52; T9N = T4W - T52; T9P = FNMS(T57, T55, T9O); T59 = FMA(T57, T58, T56); { E TfD, T9S, T9Z, T5g; TfD = T9P + T9R; T9S = T9P - T9R; T9Z = T59 - T5f; T5g = T59 + T5f; T9T = T9N - T9S; Td9 = T9N + T9S; TfE = TfC - TfD; ThI = TfC + TfD; TfB = T53 - T5g; T5h = T53 + T5g; Tda = T9Y - T9Z; Ta0 = T9Y + T9Z; } } } } { E TaN, Tdp, Tg0, TfX, Tdq, TaU; { E TaQ, T6i, TaL, T6B, T6r, T6u, T6t, TaS, T6o, TaI, T6s; { E T6x, T6A, T6z, TaK, T6y; { E T6e, T6h, T6d, T6g, TaP, T6f, T6w; T6e = cr[WS(rs, 3)]; TfO = TfE - TfB; TfF = TfB + TfE; Tag = FNMS(KP414213562, T9T, Ta0); Ta1 = FMA(KP414213562, Ta0, T9T); Tdb = FNMS(KP414213562, Tda, Td9); Tdi = FMA(KP414213562, Td9, Tda); T6h = ci[WS(rs, 3)]; T6d = W[4]; T6g = W[5]; T6x = cr[WS(rs, 51)]; T6A = ci[WS(rs, 51)]; TaP = T6d * T6h; T6f = T6d * T6e; T6w = W[100]; T6z = W[101]; TaQ = FNMS(T6g, T6e, TaP); T6i = FMA(T6g, T6h, T6f); TaK = T6w * T6A; T6y = T6w * T6x; } { E T6k, T6n, T6j, T6m, TaR, T6l, T6q; T6k = cr[WS(rs, 35)]; T6n = ci[WS(rs, 35)]; TaL = FNMS(T6z, T6x, TaK); T6B = FMA(T6z, T6A, T6y); T6j = W[68]; T6m = W[69]; T6r = cr[WS(rs, 19)]; T6u = ci[WS(rs, 19)]; TaR = T6j * T6n; T6l = T6j * T6k; T6q = W[36]; T6t = W[37]; TaS = FNMS(T6m, T6k, TaR); T6o = FMA(T6m, T6n, T6l); TaI = T6q * T6u; T6s = T6q * T6r; } } { E TaT, TfY, T6p, TaH, TaJ, T6v; TaT = TaQ - TaS; TfY = TaQ + TaS; T6p = T6i + T6o; TaH = T6i - T6o; TaJ = FNMS(T6t, T6r, TaI); T6v = FMA(T6t, T6u, T6s); { E TfZ, TaM, T6C, TaO; TfZ = TaJ + TaL; TaM = TaJ - TaL; T6C = T6v + T6B; TaO = T6B - T6v; TaN = TaH - TaM; Tdp = TaH + TaM; Tg0 = TfY - TfZ; ThU = TfY + TfZ; TfX = T6p - T6C; T6D = T6p + T6C; Tdq = TaT + TaO; TaU = TaO - TaT; } } } { E Tb5, T6J, Tb0, T72, T6S, T6V, T6U, Tb7, T6P, TaX, T6T; { E T6Y, T71, T70, TaZ, T6Z; { E T6F, T6I, T6E, T6H, Tb4, T6G, T6X; T6F = cr[WS(rs, 59)]; Tgf = TfX + Tg0; Tg1 = TfX - Tg0; Tbo = FNMS(KP414213562, TaN, TaU); TaV = FMA(KP414213562, TaU, TaN); Tdr = FMA(KP414213562, Tdq, Tdp); TdA = FNMS(KP414213562, Tdp, Tdq); T6I = ci[WS(rs, 59)]; T6E = W[116]; T6H = W[117]; T6Y = cr[WS(rs, 43)]; T71 = ci[WS(rs, 43)]; Tb4 = T6E * T6I; T6G = T6E * T6F; T6X = W[84]; T70 = W[85]; Tb5 = FNMS(T6H, T6F, Tb4); T6J = FMA(T6H, T6I, T6G); TaZ = T6X * T71; T6Z = T6X * T6Y; } { E T6L, T6O, T6K, T6N, Tb6, T6M, T6R; T6L = cr[WS(rs, 27)]; T6O = ci[WS(rs, 27)]; Tb0 = FNMS(T70, T6Y, TaZ); T72 = FMA(T70, T71, T6Z); T6K = W[52]; T6N = W[53]; T6S = cr[WS(rs, 11)]; T6V = ci[WS(rs, 11)]; Tb6 = T6K * T6O; T6M = T6K * T6L; T6R = W[20]; T6U = W[21]; Tb7 = FNMS(T6N, T6L, Tb6); T6P = FMA(T6N, T6O, T6M); TaX = T6R * T6V; T6T = T6R * T6S; } } { E Tb8, Tg3, T6Q, TaW, TaY, T6W; Tb8 = Tb5 - Tb7; Tg3 = Tb5 + Tb7; T6Q = T6J + T6P; TaW = T6J - T6P; TaY = FNMS(T6U, T6S, TaX); T6W = FMA(T6U, T6V, T6T); { E Tg4, Tb1, T73, Tb3; Tg4 = TaY + Tb0; Tb1 = TaY - Tb0; T73 = T6W + T72; Tb3 = T72 - T6W; Tb2 = TaW - Tb1; Tds = TaW + Tb1; Tg5 = Tg3 - Tg4; ThT = Tg3 + Tg4; Tg2 = T6Q - T73; T74 = T6Q + T73; Tdt = Tb8 + Tb3; Tb9 = Tb3 - Tb8; } } } } } } { E Thq, Tge, Tg6, Tdu, TdB, Tj7, Thv, ThA, Tht, Tj8, ThD, Thy, ThS, Ti0, ThZ; E ThV, ThH, ThP, ThO, ThK, Tkm, TcD, Tk0, Tk4, TjZ, Tk3, Tik, Tin; { E Tbp, Tba, TiI, TiL; { E Tio, T1I, Tj1, T3v, Tj2, TiX, TiN, Tir, T76, TiJ, TiC, TiG, T5j, Tit, Tiw; E TiK; { E TiO, TiW, Tip, Tiq; { E TO, T1H, T2B, T3u; Thq = Tm - TN; TO = Tm + TN; Tge = Tg2 - Tg5; Tg6 = Tg2 + Tg5; Tbp = FMA(KP414213562, Tb2, Tb9); Tba = FNMS(KP414213562, Tb9, Tb2); Tdu = FNMS(KP414213562, Tdt, Tds); TdB = FMA(KP414213562, Tds, Tdt); T1H = T1f + T1G; Tj7 = T1f - T1G; Thv = T29 - T2A; T2B = T29 + T2A; T3u = T32 + T3t; ThA = T32 - T3t; Tht = Thr - Ths; TiO = Ths + Thr; Tio = TO - T1H; T1I = TO + T1H; Tj1 = T2B - T3u; T3v = T2B + T3u; TiW = TiP + TiV; Tj8 = TiV - TiP; } ThD = ThB - ThC; Tip = ThB + ThC; Tiq = Thw + Thx; Thy = Thw - Thx; { E T6c, T75, Tiz, TiA; ThS = T5K - T6b; T6c = T5K + T6b; Tj2 = TiW - TiO; TiX = TiO + TiW; TiN = Tiq + Tip; Tir = Tip - Tiq; T75 = T6D + T74; Ti0 = T74 - T6D; ThZ = ThX - ThY; Tiz = ThX + ThY; TiA = ThU + ThT; ThV = ThT - ThU; { E T4p, Tiy, TiB, T5i, Tiu, Tiv; ThH = T3X - T4o; T4p = T3X + T4o; T76 = T6c + T75; Tiy = T6c - T75; TiJ = Tiz + TiA; TiB = Tiz - TiA; T5i = T4Q + T5h; ThP = T4Q - T5h; ThO = ThM - ThN; Tiu = ThM + ThN; Tiv = ThJ + ThI; ThK = ThI - ThJ; TiC = Tiy - TiB; TiG = Tiy + TiB; T5j = T4p + T5i; Tit = T4p - T5i; Tiw = Tiu - Tiv; TiK = Tiu + Tiv; } } } { E TiZ, TiD, TiH, TiE, Tis, TiM, TiY, Tj0; { E T3w, TiF, Tix, T77, Tj5, Tj3, Tj6, Tj4; TiI = T1I - T3v; T3w = T1I + T3v; TiF = Tit - Tiw; Tix = Tit + Tiw; T77 = T5j + T76; TiZ = T76 - T5j; Tj5 = Tj2 - Tj1; Tj3 = Tj1 + Tj2; TiD = Tix + TiC; Tj4 = TiC - Tix; cr[0] = T3w + T77; ci[WS(rs, 31)] = T3w - T77; Tj6 = TiG - TiF; TiH = TiF + TiG; ci[WS(rs, 39)] = FMA(KP707106781, Tj4, Tj3); cr[WS(rs, 56)] = FMS(KP707106781, Tj4, Tj3); TiE = Tio + Tir; Tis = Tio - Tir; ci[WS(rs, 55)] = FMA(KP707106781, Tj6, Tj5); cr[WS(rs, 40)] = FMS(KP707106781, Tj6, Tj5); } TiL = TiJ - TiK; TiM = TiK + TiJ; cr[WS(rs, 8)] = FMA(KP707106781, TiD, Tis); ci[WS(rs, 23)] = FNMS(KP707106781, TiD, Tis); ci[WS(rs, 7)] = FMA(KP707106781, TiH, TiE); cr[WS(rs, 24)] = FNMS(KP707106781, TiH, TiE); TiY = TiN + TiX; Tj0 = TiX - TiN; ci[WS(rs, 63)] = TiM + TiY; cr[WS(rs, 32)] = TiM - TiY; ci[WS(rs, 47)] = TiZ + Tj0; cr[WS(rs, 48)] = TiZ - Tj0; } } { E TjW, TbB, Tk2, T99, TbF, TbL, Tbv, Taj, Tcu, Tcy, Tci, Tce, Tcr, Tcx, Tch; E Tc7, Tcn, Tkg, Tka, TbZ, TbP, T7J, TbO, T7u, Tk7, TjT, TbI, TbM, Tbw, Tbs; E T7Y, TbQ; { E TbX, TbW, TbU, TbT, Tc1, Tc5, Tc4, Tc2, TaG, Tbq, Tbn, Tcb, Tcs, Tca, Tcc; E Tbb, Tcm, TbV; { E T8W, Tbz, T8z, T97, T8n, T8y; TbX = FNMS(KP707106781, T8m, T87); T8n = FMA(KP707106781, T8m, T87); T8y = FMA(KP707106781, T8x, T8u); TbW = FNMS(KP707106781, T8x, T8u); TbU = FNMS(KP707106781, T8V, T8G); T8W = FMA(KP707106781, T8V, T8G); ci[WS(rs, 15)] = TiI + TiL; cr[WS(rs, 16)] = TiI - TiL; Tbz = FMA(KP198912367, T8n, T8y); T8z = FNMS(KP198912367, T8y, T8n); T97 = FMA(KP707106781, T96, T93); TbT = FNMS(KP707106781, T96, T93); { E Tae, TbD, Ta3, Tah; { E T9x, Ta2, TbA, T98; Tc1 = FNMS(KP707106781, T9w, T9h); T9x = FMA(KP707106781, T9w, T9h); Ta2 = T9M + Ta1; Tc5 = Ta1 - T9M; Tc4 = FNMS(KP707106781, Tad, Taa); Tae = FMA(KP707106781, Tad, Taa); TbA = FNMS(KP198912367, T8W, T97); T98 = FMA(KP198912367, T97, T8W); TbD = FNMS(KP923879532, Ta2, T9x); Ta3 = FMA(KP923879532, Ta2, T9x); TjW = Tbz + TbA; TbB = Tbz - TbA; Tk2 = T98 - T8z; T99 = T8z + T98; Tah = Taf + Tag; Tc2 = Taf - Tag; } { E Tc8, Tc9, TbE, Tai; TaG = FMA(KP707106781, TaF, Taq); Tc8 = FNMS(KP707106781, TaF, Taq); Tc9 = Tbp - Tbo; Tbq = Tbo + Tbp; Tbn = FMA(KP707106781, Tbm, Tbj); Tcb = FNMS(KP707106781, Tbm, Tbj); TbE = FNMS(KP923879532, Tah, Tae); Tai = FMA(KP923879532, Tah, Tae); Tcs = FMA(KP923879532, Tc9, Tc8); Tca = FNMS(KP923879532, Tc9, Tc8); TbF = FMA(KP820678790, TbE, TbD); TbL = FNMS(KP820678790, TbD, TbE); Tbv = FMA(KP098491403, Ta3, Tai); Taj = FNMS(KP098491403, Tai, Ta3); Tcc = Tba - TaV; Tbb = TaV + Tba; } } } { E Tcp, Tc3, Tct, Tcd, Tcq, Tc6; Tct = FNMS(KP923879532, Tcc, Tcb); Tcd = FMA(KP923879532, Tcc, Tcb); Tcp = FMA(KP923879532, Tc2, Tc1); Tc3 = FNMS(KP923879532, Tc2, Tc1); Tcu = FMA(KP303346683, Tct, Tcs); Tcy = FNMS(KP303346683, Tcs, Tct); Tci = FMA(KP534511135, Tca, Tcd); Tce = FNMS(KP534511135, Tcd, Tca); Tcq = FMA(KP923879532, Tc5, Tc4); Tc6 = FNMS(KP923879532, Tc5, Tc4); Tcm = FNMS(KP668178637, TbT, TbU); TbV = FMA(KP668178637, TbU, TbT); Tcr = FMA(KP303346683, Tcq, Tcp); Tcx = FNMS(KP303346683, Tcp, Tcq); Tch = FMA(KP534511135, Tc3, Tc6); Tc7 = FNMS(KP534511135, Tc6, Tc3); } { E TbG, Tbc, Tcl, TbY; Tcl = FMA(KP668178637, TbW, TbX); TbY = FNMS(KP668178637, TbX, TbW); TbG = FNMS(KP923879532, Tbb, TaG); Tbc = FMA(KP923879532, Tbb, TaG); Tcn = Tcl + Tcm; Tkg = Tcl - Tcm; Tka = TbY + TbV; TbZ = TbV - TbY; { E T7t, TjS, TbH, Tbr; Tkm = T7s - T7l; T7t = T7l + T7s; TjS = TcB - TcC; TcD = TcB + TcC; TbP = FMA(KP414213562, T7B, T7I); T7J = FNMS(KP414213562, T7I, T7B); TbH = FNMS(KP923879532, Tbq, Tbn); Tbr = FMA(KP923879532, Tbq, Tbn); TbO = FNMS(KP707106781, T7t, T7e); T7u = FMA(KP707106781, T7t, T7e); Tk7 = FNMS(KP707106781, TjS, TjR); TjT = FMA(KP707106781, TjS, TjR); TbI = FMA(KP820678790, TbH, TbG); TbM = FNMS(KP820678790, TbG, TbH); Tbw = FMA(KP098491403, Tbc, Tbr); Tbs = FNMS(KP098491403, Tbr, Tbc); T7Y = FMA(KP414213562, T7X, T7Q); TbQ = FNMS(KP414213562, T7Q, T7X); } } } { E Tk1, TjV, Tck, TbS, Tkd, Tcz, Tkh, Tcf, TjY, Tk6, Tke, Tcv, Tki, Tcj; { E Tbu, TbC, Tkb, Tkc, Tkj, Tkk, Tbx, TbJ; { E Tbt, Tkf, Tk9, T9a, TbK, TbN, Tby; Tk0 = Tbs - Taj; Tbt = Taj + Tbs; { E Tk8, T7Z, TjU, TbR, T80; Tk8 = T7Y - T7J; T7Z = T7J + T7Y; TjU = TbP + TbQ; TbR = TbP - TbQ; Tkf = FNMS(KP923879532, Tk8, Tk7); Tk9 = FMA(KP923879532, Tk8, Tk7); Tby = FNMS(KP923879532, T7Z, T7u); T80 = FMA(KP923879532, T7Z, T7u); Tk1 = FNMS(KP923879532, TjU, TjT); TjV = FMA(KP923879532, TjU, TjT); Tck = FMA(KP923879532, TbR, TbO); TbS = FNMS(KP923879532, TbR, TbO); T9a = FMA(KP980785280, T99, T80); Tbu = FNMS(KP980785280, T99, T80); } TbC = FMA(KP980785280, TbB, Tby); TbK = FNMS(KP980785280, TbB, Tby); TbN = TbL + TbM; Tk4 = TbL - TbM; Tkd = FNMS(KP831469612, Tka, Tk9); Tkb = FMA(KP831469612, Tka, Tk9); ci[0] = FMA(KP995184726, Tbt, T9a); cr[WS(rs, 31)] = FNMS(KP995184726, Tbt, T9a); ci[WS(rs, 8)] = FNMS(KP773010453, TbN, TbK); cr[WS(rs, 23)] = FMA(KP773010453, TbN, TbK); Tkc = Tcx - Tcy; Tcz = Tcx + Tcy; Tkh = FMA(KP831469612, Tkg, Tkf); Tkj = FNMS(KP831469612, Tkg, Tkf); Tkk = Tce - Tc7; Tcf = Tc7 + Tce; } ci[WS(rs, 60)] = FMA(KP956940335, Tkc, Tkb); cr[WS(rs, 35)] = FMS(KP956940335, Tkc, Tkb); ci[WS(rs, 52)] = FMA(KP881921264, Tkk, Tkj); cr[WS(rs, 43)] = FMS(KP881921264, Tkk, Tkj); Tbx = Tbv + Tbw; TjY = Tbw - Tbv; TbJ = TbF + TbI; Tk6 = TbI - TbF; cr[WS(rs, 15)] = FMA(KP995184726, Tbx, Tbu); ci[WS(rs, 16)] = FNMS(KP995184726, Tbx, Tbu); cr[WS(rs, 7)] = FMA(KP773010453, TbJ, TbC); ci[WS(rs, 24)] = FNMS(KP773010453, TbJ, TbC); Tke = Tcu - Tcr; Tcv = Tcr + Tcu; Tki = Tci - Tch; Tcj = Tch + Tci; } { E Tcg, Tco, TjX, Tk5, Tc0, Tcw; Tcg = FNMS(KP831469612, TbZ, TbS); Tc0 = FMA(KP831469612, TbZ, TbS); ci[WS(rs, 44)] = FMA(KP956940335, Tke, Tkd); cr[WS(rs, 51)] = FMS(KP956940335, Tke, Tkd); ci[WS(rs, 36)] = FMA(KP881921264, Tki, Tkh); cr[WS(rs, 59)] = FMS(KP881921264, Tki, Tkh); Tco = FMA(KP831469612, Tcn, Tck); Tcw = FNMS(KP831469612, Tcn, Tck); TjZ = FNMS(KP980785280, TjW, TjV); TjX = FMA(KP980785280, TjW, TjV); ci[WS(rs, 4)] = FMA(KP881921264, Tcf, Tc0); cr[WS(rs, 27)] = FNMS(KP881921264, Tcf, Tc0); ci[WS(rs, 12)] = FNMS(KP956940335, Tcz, Tcw); cr[WS(rs, 19)] = FMA(KP956940335, Tcz, Tcw); Tk3 = FMA(KP980785280, Tk2, Tk1); Tk5 = FNMS(KP980785280, Tk2, Tk1); ci[WS(rs, 32)] = FMA(KP995184726, TjY, TjX); cr[WS(rs, 63)] = FMS(KP995184726, TjY, TjX); ci[WS(rs, 40)] = FMA(KP773010453, Tk6, Tk5); cr[WS(rs, 55)] = FMS(KP773010453, Tk6, Tk5); cr[WS(rs, 11)] = FMA(KP881921264, Tcj, Tcg); ci[WS(rs, 20)] = FNMS(KP881921264, Tcj, Tcg); cr[WS(rs, 3)] = FMA(KP956940335, Tcv, Tco); ci[WS(rs, 28)] = FNMS(KP956940335, Tcv, Tco); } } } } { E Ti8, Thu, Tjf, Tj9, Tib, Tjg, Tja, ThF, Tig, ThW, Tif, Til, Ti6, ThR; ci[WS(rs, 48)] = FMA(KP995184726, Tk0, TjZ); cr[WS(rs, 47)] = FMS(KP995184726, Tk0, TjZ); ci[WS(rs, 56)] = FMA(KP773010453, Tk4, Tk3); cr[WS(rs, 39)] = FMS(KP773010453, Tk4, Tk3); Ti8 = Thq + Tht; Thu = Thq - Tht; Tjf = Tj8 - Tj7; Tj9 = Tj7 + Tj8; { E Tid, ThL, Tie, ThQ; { E Ti9, Thz, Tia, ThE; Ti9 = Thv - Thy; Thz = Thv + Thy; Tia = ThA + ThD; ThE = ThA - ThD; Tib = Ti9 + Tia; Tjg = Tia - Ti9; Tja = Thz - ThE; ThF = Thz + ThE; Tid = ThH + ThK; ThL = ThH - ThK; } Tie = ThO + ThP; ThQ = ThO - ThP; Tig = ThS + ThV; ThW = ThS - ThV; Tif = FNMS(KP414213562, Tie, Tid); Til = FMA(KP414213562, Tid, Tie); Ti6 = FNMS(KP414213562, ThL, ThQ); ThR = FMA(KP414213562, ThQ, ThL); } { E Ti4, ThG, Tjh, Tjj, Tih, Ti1; Ti4 = FNMS(KP707106781, ThF, Thu); ThG = FMA(KP707106781, ThF, Thu); Tjh = FMA(KP707106781, Tjg, Tjf); Tjj = FNMS(KP707106781, Tjg, Tjf); Tih = Ti0 - ThZ; Ti1 = ThZ + Ti0; { E Tje, Tjd, Tjb, Tjc; { E Tic, Tim, Ti5, Ti2, Tij, Tii; Tik = FNMS(KP707106781, Tib, Ti8); Tic = FMA(KP707106781, Tib, Ti8); Tii = FNMS(KP414213562, Tih, Tig); Tim = FMA(KP414213562, Tig, Tih); Ti5 = FMA(KP414213562, ThW, Ti1); Ti2 = FNMS(KP414213562, Ti1, ThW); Tij = Tif + Tii; Tje = Tii - Tif; Tjd = FNMS(KP707106781, Tja, Tj9); Tjb = FMA(KP707106781, Tja, Tj9); { E Ti7, Tji, Tjk, Ti3; Ti7 = Ti5 - Ti6; Tji = Ti6 + Ti5; Tjk = Ti2 - ThR; Ti3 = ThR + Ti2; ci[WS(rs, 3)] = FMA(KP923879532, Tij, Tic); cr[WS(rs, 28)] = FNMS(KP923879532, Tij, Tic); ci[WS(rs, 11)] = FMA(KP923879532, Ti7, Ti4); cr[WS(rs, 20)] = FNMS(KP923879532, Ti7, Ti4); ci[WS(rs, 59)] = FMA(KP923879532, Tji, Tjh); cr[WS(rs, 36)] = FMS(KP923879532, Tji, Tjh); ci[WS(rs, 43)] = FMA(KP923879532, Tjk, Tjj); cr[WS(rs, 52)] = FMS(KP923879532, Tjk, Tjj); cr[WS(rs, 4)] = FMA(KP923879532, Ti3, ThG); ci[WS(rs, 27)] = FNMS(KP923879532, Ti3, ThG); Tjc = Tim - Til; Tin = Til + Tim; } } ci[WS(rs, 35)] = FMA(KP923879532, Tjc, Tjb); cr[WS(rs, 60)] = FMS(KP923879532, Tjc, Tjb); ci[WS(rs, 51)] = FMA(KP923879532, Tje, Tjd); cr[WS(rs, 44)] = FMS(KP923879532, Tje, Tjd); } } } { E Tjy, Tju, Tjt, Tjx; { E TjD, TjJ, Tgo, Tf2, Tjp, Tjv, Tha, TgI, Tgg, Tgd, Tgr, Tjw, Tjq, Tfp, Thk; E Tho, Th7, Th4, Tgv, TgB, Tgl, TfR, TjE, Thd, TjK, TgP, Tgw, Tg8, Thh, Thn; E Th8, TgX; { E TgK, TgJ, TgN, TgM, TfW, Th1, Thi, Th0, Th2, Tg7; { E TgE, TeQ, TjB, Tjn, TgF, TgG, TjC, Tf1, TeV, Tf0; TgE = TeM - TeP; TeQ = TeM + TeP; TjB = Tjm - Tjl; Tjn = Tjl + Tjm; TgF = TeR + TeU; TeV = TeR - TeU; cr[WS(rs, 12)] = FMA(KP923879532, Tin, Tik); ci[WS(rs, 19)] = FNMS(KP923879532, Tin, Tik); Tf0 = TeW + TeZ; TgG = TeW - TeZ; TjC = Tf0 - TeV; Tf1 = TeV + Tf0; { E Tfi, Tgp, Tfd, Tfn; { E Tf7, Tjo, TgH, Tfc; TgK = Tf5 - Tf6; Tf7 = Tf5 + Tf6; TjD = FMA(KP707106781, TjC, TjB); TjJ = FNMS(KP707106781, TjC, TjB); Tgo = FMA(KP707106781, Tf1, TeQ); Tf2 = FNMS(KP707106781, Tf1, TeQ); Tjo = TgF - TgG; TgH = TgF + TgG; Tfc = Tf8 + Tfb; TgJ = Tf8 - Tfb; TgN = Tfg - Tfh; Tfi = Tfg + Tfh; Tjp = FMA(KP707106781, Tjo, Tjn); Tjv = FNMS(KP707106781, Tjo, Tjn); Tha = FNMS(KP707106781, TgH, TgE); TgI = FMA(KP707106781, TgH, TgE); Tgp = FNMS(KP414213562, Tf7, Tfc); Tfd = FMA(KP414213562, Tfc, Tf7); Tfn = Tfj + Tfm; TgM = Tfj - Tfm; } { E TgY, TgZ, Tgq, Tfo; TfW = TfS + TfV; TgY = TfS - TfV; TgZ = Tgf + Tge; Tgg = Tge - Tgf; Tgd = Tg9 - Tgc; Th1 = Tgc + Tg9; Tgq = FMA(KP414213562, Tfi, Tfn); Tfo = FNMS(KP414213562, Tfn, Tfi); Thi = FNMS(KP707106781, TgZ, TgY); Th0 = FMA(KP707106781, TgZ, TgY); Tgr = Tgp + Tgq; Tjw = Tgq - Tgp; Tjq = Tfd + Tfo; Tfp = Tfd - Tfo; Th2 = Tg6 - Tg1; Tg7 = Tg1 + Tg6; } } } { E TgR, TgV, TgU, TgS, Thc, TgL; { E TfM, Tgt, TfH, TfP, Tgu, TfQ; { E Tfv, TfG, Thj, Th3; TgR = Tfr - Tfu; Tfv = Tfr + Tfu; TfG = TfA + TfF; TgV = TfF - TfA; TgU = TfK - TfL; TfM = TfK + TfL; Thj = FNMS(KP707106781, Th2, Th1); Th3 = FMA(KP707106781, Th2, Th1); Tgt = FMA(KP707106781, TfG, Tfv); TfH = FNMS(KP707106781, TfG, Tfv); Thk = FMA(KP668178637, Thj, Thi); Tho = FNMS(KP668178637, Thi, Thj); Th7 = FMA(KP198912367, Th0, Th3); Th4 = FNMS(KP198912367, Th3, Th0); TfP = TfN + TfO; TgS = TfN - TfO; } Tgu = FMA(KP707106781, TfP, TfM); TfQ = FNMS(KP707106781, TfP, TfM); Thc = FNMS(KP414213562, TgJ, TgK); TgL = FMA(KP414213562, TgK, TgJ); Tgv = FNMS(KP198912367, Tgu, Tgt); TgB = FMA(KP198912367, Tgt, Tgu); Tgl = FNMS(KP668178637, TfH, TfQ); TfR = FMA(KP668178637, TfQ, TfH); } { E Thf, TgT, Thb, TgO, Thg, TgW; Thb = FMA(KP414213562, TgM, TgN); TgO = FNMS(KP414213562, TgN, TgM); Thf = FNMS(KP707106781, TgS, TgR); TgT = FMA(KP707106781, TgS, TgR); TjE = Thc + Thb; Thd = Thb - Thc; TjK = TgL - TgO; TgP = TgL + TgO; Thg = FNMS(KP707106781, TgV, TgU); TgW = FMA(KP707106781, TgV, TgU); Tgw = FMA(KP707106781, Tg7, TfW); Tg8 = FNMS(KP707106781, Tg7, TfW); Thh = FNMS(KP668178637, Thg, Thf); Thn = FMA(KP668178637, Thf, Thg); Th8 = FNMS(KP198912367, TgT, TgW); TgX = FMA(KP198912367, TgW, TgT); } } } { E TjH, Th9, TjL, Tjs, TjA, Thl, TjI, Th5, TjM, Thp; { E Tgk, Tfq, TgA, Tgs, TjN, Tgy, Tgm, TgD, Tgj, TjO, Tgn, Tgz; Tgk = FNMS(KP923879532, Tfp, Tf2); Tfq = FMA(KP923879532, Tfp, Tf2); TgA = FNMS(KP923879532, Tgr, Tgo); Tgs = FMA(KP923879532, Tgr, Tgo); { E TjF, Tgx, Tgh, TjG, TgC, Tgi; TjH = FNMS(KP923879532, TjE, TjD); TjF = FMA(KP923879532, TjE, TjD); Tgx = FMA(KP707106781, Tgg, Tgd); Tgh = FNMS(KP707106781, Tgg, Tgd); TjG = Th8 + Th7; Th9 = Th7 - Th8; TjL = FMA(KP923879532, TjK, TjJ); TjN = FNMS(KP923879532, TjK, TjJ); Tgy = FNMS(KP198912367, Tgx, Tgw); TgC = FMA(KP198912367, Tgw, Tgx); Tgm = FNMS(KP668178637, Tg8, Tgh); Tgi = FMA(KP668178637, Tgh, Tg8); ci[WS(rs, 61)] = FMA(KP980785280, TjG, TjF); cr[WS(rs, 34)] = FMS(KP980785280, TjG, TjF); TgD = TgB + TgC; Tjs = TgC - TgB; TjA = Tgi - TfR; Tgj = TfR + Tgi; TjO = Thk - Thh; Thl = Thh + Thk; } cr[WS(rs, 14)] = FMA(KP980785280, TgD, TgA); ci[WS(rs, 17)] = FNMS(KP980785280, TgD, TgA); cr[WS(rs, 6)] = FMA(KP831469612, Tgj, Tfq); ci[WS(rs, 25)] = FNMS(KP831469612, Tgj, Tfq); ci[WS(rs, 53)] = FMA(KP831469612, TjO, TjN); cr[WS(rs, 42)] = FMS(KP831469612, TjO, TjN); Tgn = Tgl + Tgm; Tjy = Tgl - Tgm; Tgz = Tgv + Tgy; Tju = Tgy - Tgv; ci[WS(rs, 9)] = FNMS(KP831469612, Tgn, Tgk); cr[WS(rs, 22)] = FMA(KP831469612, Tgn, Tgk); ci[WS(rs, 1)] = FMA(KP980785280, Tgz, Tgs); cr[WS(rs, 30)] = FNMS(KP980785280, Tgz, Tgs); TjI = Th4 - TgX; Th5 = TgX + Th4; TjM = Thn + Tho; Thp = Thn - Tho; } { E Th6, The, Tjr, Tjz, TgQ, Thm; Th6 = FNMS(KP923879532, TgP, TgI); TgQ = FMA(KP923879532, TgP, TgI); ci[WS(rs, 45)] = FMA(KP980785280, TjI, TjH); cr[WS(rs, 50)] = FMS(KP980785280, TjI, TjH); ci[WS(rs, 37)] = FNMS(KP831469612, TjM, TjL); cr[WS(rs, 58)] = -(FMA(KP831469612, TjM, TjL)); The = FMA(KP923879532, Thd, Tha); Thm = FNMS(KP923879532, Thd, Tha); Tjt = FNMS(KP923879532, Tjq, Tjp); Tjr = FMA(KP923879532, Tjq, Tjp); cr[WS(rs, 2)] = FMA(KP980785280, Th5, TgQ); ci[WS(rs, 29)] = FNMS(KP980785280, Th5, TgQ); cr[WS(rs, 10)] = FMA(KP831469612, Thp, Thm); ci[WS(rs, 21)] = FNMS(KP831469612, Thp, Thm); Tjx = FMA(KP923879532, Tjw, Tjv); Tjz = FNMS(KP923879532, Tjw, Tjv); ci[WS(rs, 33)] = FMA(KP980785280, Tjs, Tjr); cr[WS(rs, 62)] = FMS(KP980785280, Tjs, Tjr); ci[WS(rs, 41)] = FMA(KP831469612, TjA, Tjz); cr[WS(rs, 54)] = FMS(KP831469612, TjA, Tjz); ci[WS(rs, 13)] = FMA(KP980785280, Th9, Th6); cr[WS(rs, 18)] = FNMS(KP980785280, Th9, Th6); ci[WS(rs, 5)] = FMA(KP831469612, Thl, The); cr[WS(rs, 26)] = FNMS(KP831469612, Thl, The); } } } { E Tkq, TdN, Tkw, Td1, TdR, TdX, TdI, Tdl, TeG, TeK, Tet, Teq, TeD, TeJ, Teu; E Tej, Tez, TkK, TkE, Teb, Te2, TcH, Te0, TcE, TkB, Tkn, TdU, TdY, TdH, TdE; E TcK, Te1; { E Te6, Te5, Te9, Te8, Ted, Teh, Teg, Tee, Tdo, TdC, Tdz, Ten, TeE, Tem, Teo; E Tdv, Tex, Te7; { E TcP, TcS, TcW, TcZ; Te6 = FNMS(KP707106781, TcO, TcN); TcP = FMA(KP707106781, TcO, TcN); ci[WS(rs, 49)] = FMA(KP980785280, Tju, Tjt); cr[WS(rs, 46)] = FMS(KP980785280, Tju, Tjt); ci[WS(rs, 57)] = FMA(KP831469612, Tjy, Tjx); cr[WS(rs, 38)] = FMS(KP831469612, Tjy, Tjx); TcS = FMA(KP707106781, TcR, TcQ); Te5 = FNMS(KP707106781, TcR, TcQ); Te9 = FNMS(KP707106781, TcV, TcU); TcW = FMA(KP707106781, TcV, TcU); TcZ = FMA(KP707106781, TcY, TcX); Te8 = FNMS(KP707106781, TcY, TcX); { E Tdg, TdP, Tdd, Tdj; { E Td5, TdM, TcT, TdL, Td0, Tdc; Ted = FNMS(KP707106781, Td4, Td3); Td5 = FMA(KP707106781, Td4, Td3); TdM = FNMS(KP198912367, TcP, TcS); TcT = FMA(KP198912367, TcS, TcP); TdL = FMA(KP198912367, TcW, TcZ); Td0 = FNMS(KP198912367, TcZ, TcW); Tdc = Td8 + Tdb; Teh = Td8 - Tdb; Teg = FNMS(KP707106781, Tdf, Tde); Tdg = FMA(KP707106781, Tdf, Tde); Tkq = TdM + TdL; TdN = TdL - TdM; Tkw = TcT - Td0; Td1 = TcT + Td0; TdP = FNMS(KP923879532, Tdc, Td5); Tdd = FMA(KP923879532, Tdc, Td5); Tdj = Tdh + Tdi; Tee = Tdi - Tdh; } { E Tek, Tel, TdQ, Tdk; Tdo = FMA(KP707106781, Tdn, Tdm); Tek = FNMS(KP707106781, Tdn, Tdm); Tel = TdB - TdA; TdC = TdA + TdB; Tdz = FMA(KP707106781, Tdy, Tdx); Ten = FNMS(KP707106781, Tdy, Tdx); TdQ = FNMS(KP923879532, Tdj, Tdg); Tdk = FMA(KP923879532, Tdj, Tdg); TeE = FMA(KP923879532, Tel, Tek); Tem = FNMS(KP923879532, Tel, Tek); TdR = FNMS(KP820678790, TdQ, TdP); TdX = FMA(KP820678790, TdP, TdQ); TdI = FNMS(KP098491403, Tdd, Tdk); Tdl = FMA(KP098491403, Tdk, Tdd); Teo = Tdu - Tdr; Tdv = Tdr + Tdu; } } } { E TeB, Tef, TeF, Tep, TeC, Tei; TeF = FNMS(KP923879532, Teo, Ten); Tep = FMA(KP923879532, Teo, Ten); TeB = FMA(KP923879532, Tee, Ted); Tef = FNMS(KP923879532, Tee, Ted); TeG = FMA(KP303346683, TeF, TeE); TeK = FNMS(KP303346683, TeE, TeF); Tet = FMA(KP534511135, Tem, Tep); Teq = FNMS(KP534511135, Tep, Tem); TeC = FMA(KP923879532, Teh, Teg); Tei = FNMS(KP923879532, Teh, Teg); Tex = FNMS(KP668178637, Te5, Te6); Te7 = FMA(KP668178637, Te6, Te5); TeD = FNMS(KP303346683, TeC, TeB); TeJ = FMA(KP303346683, TeB, TeC); Teu = FNMS(KP534511135, Tef, Tei); Tej = FMA(KP534511135, Tei, Tef); } { E TdS, Tdw, Tey, Tea, TdT, TdD; Tey = FMA(KP668178637, Te8, Te9); Tea = FNMS(KP668178637, Te9, Te8); TdS = FNMS(KP923879532, Tdv, Tdo); Tdw = FMA(KP923879532, Tdv, Tdo); Tez = Tex + Tey; TkK = Tey - Tex; TkE = Te7 + Tea; Teb = Te7 - Tea; Te2 = FNMS(KP414213562, TcF, TcG); TcH = FMA(KP414213562, TcG, TcF); TdT = FNMS(KP923879532, TdC, Tdz); TdD = FMA(KP923879532, TdC, Tdz); Te0 = FNMS(KP707106781, TcD, TcA); TcE = FMA(KP707106781, TcD, TcA); TkB = FNMS(KP707106781, Tkm, Tkl); Tkn = FMA(KP707106781, Tkm, Tkl); TdU = FMA(KP820678790, TdT, TdS); TdY = FNMS(KP820678790, TdS, TdT); TdH = FMA(KP098491403, Tdw, TdD); TdE = FNMS(KP098491403, TdD, Tdw); TcK = FNMS(KP414213562, TcJ, TcI); Te1 = FMA(KP414213562, TcI, TcJ); } } { E Tkv, Tkp, Tew, Te4, TkH, TeL, TkL, Ter, Tks, TkA, TkI, TeH, TkM, Tev; { E TdG, TdO, TkF, TkG, TkN, TkO, TdJ, TdV; { E TdF, TkJ, TkD, Td2, TdW, TdZ, TdK; Tku = TdE - Tdl; TdF = Tdl + TdE; { E TkC, TcL, Tko, Te3, TcM; TkC = TcH - TcK; TcL = TcH + TcK; Tko = Te2 + Te1; Te3 = Te1 - Te2; TkJ = FNMS(KP923879532, TkC, TkB); TkD = FMA(KP923879532, TkC, TkB); TdK = FNMS(KP923879532, TcL, TcE); TcM = FMA(KP923879532, TcL, TcE); Tkv = FNMS(KP923879532, Tko, Tkn); Tkp = FMA(KP923879532, Tko, Tkn); Tew = FMA(KP923879532, Te3, Te0); Te4 = FNMS(KP923879532, Te3, Te0); Td2 = FMA(KP980785280, Td1, TcM); TdG = FNMS(KP980785280, Td1, TcM); } TdO = FMA(KP980785280, TdN, TdK); TdW = FNMS(KP980785280, TdN, TdK); TdZ = TdX - TdY; Tky = TdX + TdY; TkH = FNMS(KP831469612, TkE, TkD); TkF = FMA(KP831469612, TkE, TkD); cr[WS(rs, 1)] = FMA(KP995184726, TdF, Td2); ci[WS(rs, 30)] = FNMS(KP995184726, TdF, Td2); cr[WS(rs, 9)] = FMA(KP773010453, TdZ, TdW); ci[WS(rs, 22)] = FNMS(KP773010453, TdZ, TdW); TkG = TeJ + TeK; TeL = TeJ - TeK; TkL = FMA(KP831469612, TkK, TkJ); TkN = FNMS(KP831469612, TkK, TkJ); TkO = Teq - Tej; Ter = Tej + Teq; } ci[WS(rs, 34)] = FNMS(KP956940335, TkG, TkF); cr[WS(rs, 61)] = -(FMA(KP956940335, TkG, TkF)); ci[WS(rs, 42)] = FMA(KP881921264, TkO, TkN); cr[WS(rs, 53)] = FMS(KP881921264, TkO, TkN); TdJ = TdH - TdI; Tks = TdI + TdH; TdV = TdR + TdU; TkA = TdU - TdR; ci[WS(rs, 14)] = FMA(KP995184726, TdJ, TdG); cr[WS(rs, 17)] = FNMS(KP995184726, TdJ, TdG); ci[WS(rs, 6)] = FMA(KP773010453, TdV, TdO); cr[WS(rs, 25)] = FNMS(KP773010453, TdV, TdO); TkI = TeG - TeD; TeH = TeD + TeG; TkM = Teu + Tet; Tev = Tet - Teu; } { E Tes, TeA, Tkr, Tkz, Tec, TeI; Tes = FNMS(KP831469612, Teb, Te4); Tec = FMA(KP831469612, Teb, Te4); ci[WS(rs, 50)] = FMA(KP956940335, TkI, TkH); cr[WS(rs, 45)] = FMS(KP956940335, TkI, TkH); ci[WS(rs, 58)] = FMA(KP881921264, TkM, TkL); cr[WS(rs, 37)] = FMS(KP881921264, TkM, TkL); TeA = FMA(KP831469612, Tez, Tew); TeI = FNMS(KP831469612, Tez, Tew); Tkt = FNMS(KP980785280, Tkq, Tkp); Tkr = FMA(KP980785280, Tkq, Tkp); cr[WS(rs, 5)] = FMA(KP881921264, Ter, Tec); ci[WS(rs, 26)] = FNMS(KP881921264, Ter, Tec); cr[WS(rs, 13)] = FMA(KP956940335, TeL, TeI); ci[WS(rs, 18)] = FNMS(KP956940335, TeL, TeI); Tkx = FMA(KP980785280, Tkw, Tkv); Tkz = FNMS(KP980785280, Tkw, Tkv); ci[WS(rs, 62)] = FMA(KP995184726, Tks, Tkr); cr[WS(rs, 33)] = FMS(KP995184726, Tks, Tkr); ci[WS(rs, 54)] = FMA(KP773010453, TkA, Tkz); cr[WS(rs, 41)] = FMS(KP773010453, TkA, Tkz); ci[WS(rs, 10)] = FMA(KP881921264, Tev, Tes); cr[WS(rs, 21)] = FNMS(KP881921264, Tev, Tes); ci[WS(rs, 2)] = FMA(KP956940335, TeH, TeA); cr[WS(rs, 29)] = FNMS(KP956940335, TeH, TeA); } } } } } } ci[WS(rs, 46)] = FMA(KP995184726, Tku, Tkt); cr[WS(rs, 49)] = FMS(KP995184726, Tku, Tkt); ci[WS(rs, 38)] = FNMS(KP773010453, Tky, Tkx); cr[WS(rs, 57)] = -(FMA(KP773010453, Tky, Tkx)); } } } static const tw_instr twinstr[] = { {TW_FULL, 1, 64}, {TW_NEXT, 1, 0} }; static const hc2hc_desc desc = { 64, "hf_64", twinstr, &GENUS, {520, 126, 518, 0} }; void X(codelet_hf_64) (planner *p) { X(khc2hc_register) (p, hf_64, &desc); } #else /* HAVE_FMA */ /* Generated by: ../../../genfft/gen_hc2hc.native -compact -variables 4 -pipeline-latency 4 -n 64 -dit -name hf_64 -include hf.h */ /* * This function contains 1038 FP additions, 500 FP multiplications, * (or, 808 additions, 270 multiplications, 230 fused multiply/add), * 176 stack variables, 15 constants, and 256 memory accesses */ #include "hf.h" static void hf_64(R *cr, R *ci, const R *W, stride rs, INT mb, INT me, INT ms) { DK(KP290284677, +0.290284677254462367636192375817395274691476278); DK(KP956940335, +0.956940335732208864935797886980269969482849206); DK(KP881921264, +0.881921264348355029712756863660388349508442621); DK(KP471396736, +0.471396736825997648556387625905254377657460319); DK(KP555570233, +0.555570233019602224742830813948532874374937191); DK(KP831469612, +0.831469612302545237078788377617905756738560812); DK(KP098017140, +0.098017140329560601994195563888641845861136673); DK(KP995184726, +0.995184726672196886244836953109479921575474869); DK(KP773010453, +0.773010453362736960810906609758469800971041293); DK(KP634393284, +0.634393284163645498215171613225493370675687095); DK(KP980785280, +0.980785280403230449126182236134239036973933731); DK(KP195090322, +0.195090322016128267848284868477022240927691618); DK(KP382683432, +0.382683432365089771728459984030398866761344562); DK(KP923879532, +0.923879532511286756128183189396788286822416626); DK(KP707106781, +0.707106781186547524400844362104849039284835938); { INT m; for (m = mb, W = W + ((mb - 1) * 126); m < me; m = m + 1, cr = cr + ms, ci = ci - ms, W = W + 126, MAKE_VOLATILE_STRIDE(128, rs)) { E Tj, TcL, ThT, Tin, T6b, Taz, TgT, Thn, TG, Thm, TcO, TgO, T6m, Tim, TaC; E ThQ, T14, Tfr, T6y, T9O, TaG, Tc0, TcU, TeE, T1r, Tfq, T6J, T9P, TaJ, Tc1; E TcZ, TeF, T1Q, T2d, Tfu, Tfv, Tfw, Tfx, T6Q, TaM, Tdb, TeI, T71, TaQ, T7a; E TaN, Td6, TeJ, T77, TaP, T2B, T2Y, Tfz, TfA, TfB, TfC, T7h, TaW, Tdm, TeL; E T7s, TaU, T7B, TaX, Tdh, TeM, T7y, TaT, T5j, TfR, Tec, TeX, TfY, Tgy, T8D; E Tbl, T8O, Tbx, T9l, Tbm, TdV, Tf0, T9i, Tbw, T3M, TfL, TdL, TeT, TfI, Tgt; E T7K, Tbd, T7V, Tb3, T8s, Tbe, Tdu, TeQ, T8p, Tb2, T4x, TfJ, TdE, TdM, TfO; E Tgu, T87, T8u, T8i, T8v, Tba, Tbh, Tdz, TdN, Tb7, Tbg, T64, TfZ, Te5, Ted; E TfU, Tgz, T90, T9n, T9b, T9o, Tbt, TbA, Te0, Tee, Tbq, Tbz; { E T1, TgR, T6, TgQ, Tc, T68, Th, T69; T1 = cr[0]; TgR = ci[0]; { E T3, T5, T2, T4; T3 = cr[WS(rs, 32)]; T5 = ci[WS(rs, 32)]; T2 = W[62]; T4 = W[63]; T6 = FMA(T2, T3, T4 * T5); TgQ = FNMS(T4, T3, T2 * T5); } { E T9, Tb, T8, Ta; T9 = cr[WS(rs, 16)]; Tb = ci[WS(rs, 16)]; T8 = W[30]; Ta = W[31]; Tc = FMA(T8, T9, Ta * Tb); T68 = FNMS(Ta, T9, T8 * Tb); } { E Te, Tg, Td, Tf; Te = cr[WS(rs, 48)]; Tg = ci[WS(rs, 48)]; Td = W[94]; Tf = W[95]; Th = FMA(Td, Te, Tf * Tg); T69 = FNMS(Tf, Te, Td * Tg); } { E T7, Ti, ThR, ThS; T7 = T1 + T6; Ti = Tc + Th; Tj = T7 + Ti; TcL = T7 - Ti; ThR = Tc - Th; ThS = TgR - TgQ; ThT = ThR + ThS; Tin = ThS - ThR; } { E T67, T6a, TgP, TgS; T67 = T1 - T6; T6a = T68 - T69; T6b = T67 - T6a; Taz = T67 + T6a; TgP = T68 + T69; TgS = TgQ + TgR; TgT = TgP + TgS; Thn = TgS - TgP; } } { E To, T6d, Tt, T6e, T6c, T6f, Tz, T6i, TE, T6j, T6h, T6k; { E Tl, Tn, Tk, Tm; Tl = cr[WS(rs, 8)]; Tn = ci[WS(rs, 8)]; Tk = W[14]; Tm = W[15]; To = FMA(Tk, Tl, Tm * Tn); T6d = FNMS(Tm, Tl, Tk * Tn); } { E Tq, Ts, Tp, Tr; Tq = cr[WS(rs, 40)]; Ts = ci[WS(rs, 40)]; Tp = W[78]; Tr = W[79]; Tt = FMA(Tp, Tq, Tr * Ts); T6e = FNMS(Tr, Tq, Tp * Ts); } T6c = To - Tt; T6f = T6d - T6e; { E Tw, Ty, Tv, Tx; Tw = cr[WS(rs, 56)]; Ty = ci[WS(rs, 56)]; Tv = W[110]; Tx = W[111]; Tz = FMA(Tv, Tw, Tx * Ty); T6i = FNMS(Tx, Tw, Tv * Ty); } { E TB, TD, TA, TC; TB = cr[WS(rs, 24)]; TD = ci[WS(rs, 24)]; TA = W[46]; TC = W[47]; TE = FMA(TA, TB, TC * TD); T6j = FNMS(TC, TB, TA * TD); } T6h = Tz - TE; T6k = T6i - T6j; { E Tu, TF, TcM, TcN; Tu = To + Tt; TF = Tz + TE; TG = Tu + TF; Thm = Tu - TF; TcM = T6i + T6j; TcN = T6d + T6e; TcO = TcM - TcN; TgO = TcN + TcM; } { E T6g, T6l, TaA, TaB; T6g = T6c - T6f; T6l = T6h + T6k; T6m = KP707106781 * (T6g + T6l); Tim = KP707106781 * (T6l - T6g); TaA = T6c + T6f; TaB = T6h - T6k; TaC = KP707106781 * (TaA + TaB); ThQ = KP707106781 * (TaA - TaB); } } { E TS, TcR, T6o, T6v, T13, TcS, T6r, T6w, T6s, T6x; { E TM, T6t, TR, T6u; { E TJ, TL, TI, TK; TJ = cr[WS(rs, 4)]; TL = ci[WS(rs, 4)]; TI = W[6]; TK = W[7]; TM = FMA(TI, TJ, TK * TL); T6t = FNMS(TK, TJ, TI * TL); } { E TO, TQ, TN, TP; TO = cr[WS(rs, 36)]; TQ = ci[WS(rs, 36)]; TN = W[70]; TP = W[71]; TR = FMA(TN, TO, TP * TQ); T6u = FNMS(TP, TO, TN * TQ); } TS = TM + TR; TcR = T6t + T6u; T6o = TM - TR; T6v = T6t - T6u; } { E TX, T6p, T12, T6q; { E TU, TW, TT, TV; TU = cr[WS(rs, 20)]; TW = ci[WS(rs, 20)]; TT = W[38]; TV = W[39]; TX = FMA(TT, TU, TV * TW); T6p = FNMS(TV, TU, TT * TW); } { E TZ, T11, TY, T10; TZ = cr[WS(rs, 52)]; T11 = ci[WS(rs, 52)]; TY = W[102]; T10 = W[103]; T12 = FMA(TY, TZ, T10 * T11); T6q = FNMS(T10, TZ, TY * T11); } T13 = TX + T12; TcS = T6p + T6q; T6r = T6p - T6q; T6w = TX - T12; } T14 = TS + T13; Tfr = TcR + TcS; T6s = T6o - T6r; T6x = T6v + T6w; T6y = FNMS(KP382683432, T6x, KP923879532 * T6s); T9O = FMA(KP923879532, T6x, KP382683432 * T6s); { E TaE, TaF, TcQ, TcT; TaE = T6v - T6w; TaF = T6o + T6r; TaG = FMA(KP382683432, TaE, KP923879532 * TaF); Tc0 = FNMS(KP923879532, TaE, KP382683432 * TaF); TcQ = TS - T13; TcT = TcR - TcS; TcU = TcQ + TcT; TeE = TcQ - TcT; } } { E T1f, TcW, T6B, T6E, T1q, TcX, T6C, T6H, T6D, T6I; { E T19, T6z, T1e, T6A; { E T16, T18, T15, T17; T16 = cr[WS(rs, 60)]; T18 = ci[WS(rs, 60)]; T15 = W[118]; T17 = W[119]; T19 = FMA(T15, T16, T17 * T18); T6z = FNMS(T17, T16, T15 * T18); } { E T1b, T1d, T1a, T1c; T1b = cr[WS(rs, 28)]; T1d = ci[WS(rs, 28)]; T1a = W[54]; T1c = W[55]; T1e = FMA(T1a, T1b, T1c * T1d); T6A = FNMS(T1c, T1b, T1a * T1d); } T1f = T19 + T1e; TcW = T6z + T6A; T6B = T6z - T6A; T6E = T19 - T1e; } { E T1k, T6F, T1p, T6G; { E T1h, T1j, T1g, T1i; T1h = cr[WS(rs, 12)]; T1j = ci[WS(rs, 12)]; T1g = W[22]; T1i = W[23]; T1k = FMA(T1g, T1h, T1i * T1j); T6F = FNMS(T1i, T1h, T1g * T1j); } { E T1m, T1o, T1l, T1n; T1m = cr[WS(rs, 44)]; T1o = ci[WS(rs, 44)]; T1l = W[86]; T1n = W[87]; T1p = FMA(T1l, T1m, T1n * T1o); T6G = FNMS(T1n, T1m, T1l * T1o); } T1q = T1k + T1p; TcX = T6F + T6G; T6C = T1k - T1p; T6H = T6F - T6G; } T1r = T1f + T1q; Tfq = TcW + TcX; T6D = T6B + T6C; T6I = T6E - T6H; T6J = FMA(KP382683432, T6D, KP923879532 * T6I); T9P = FNMS(KP923879532, T6D, KP382683432 * T6I); { E TaH, TaI, TcV, TcY; TaH = T6E + T6H; TaI = T6B - T6C; TaJ = FNMS(KP382683432, TaI, KP923879532 * TaH); Tc1 = FMA(KP923879532, TaI, KP382683432 * TaH); TcV = T1f - T1q; TcY = TcW - TcX; TcZ = TcV - TcY; TeF = TcV + TcY; } } { E T1y, T73, T1D, T74, T1E, Td7, T1J, T6N, T1O, T6O, T1P, Td8, T21, Td4, T6R; E T6U, T2c, Td3, T6W, T6Z; { E T1v, T1x, T1u, T1w; T1v = cr[WS(rs, 2)]; T1x = ci[WS(rs, 2)]; T1u = W[2]; T1w = W[3]; T1y = FMA(T1u, T1v, T1w * T1x); T73 = FNMS(T1w, T1v, T1u * T1x); } { E T1A, T1C, T1z, T1B; T1A = cr[WS(rs, 34)]; T1C = ci[WS(rs, 34)]; T1z = W[66]; T1B = W[67]; T1D = FMA(T1z, T1A, T1B * T1C); T74 = FNMS(T1B, T1A, T1z * T1C); } T1E = T1y + T1D; Td7 = T73 + T74; { E T1G, T1I, T1F, T1H; T1G = cr[WS(rs, 18)]; T1I = ci[WS(rs, 18)]; T1F = W[34]; T1H = W[35]; T1J = FMA(T1F, T1G, T1H * T1I); T6N = FNMS(T1H, T1G, T1F * T1I); } { E T1L, T1N, T1K, T1M; T1L = cr[WS(rs, 50)]; T1N = ci[WS(rs, 50)]; T1K = W[98]; T1M = W[99]; T1O = FMA(T1K, T1L, T1M * T1N); T6O = FNMS(T1M, T1L, T1K * T1N); } T1P = T1J + T1O; Td8 = T6N + T6O; { E T1V, T6S, T20, T6T; { E T1S, T1U, T1R, T1T; T1S = cr[WS(rs, 10)]; T1U = ci[WS(rs, 10)]; T1R = W[18]; T1T = W[19]; T1V = FMA(T1R, T1S, T1T * T1U); T6S = FNMS(T1T, T1S, T1R * T1U); } { E T1X, T1Z, T1W, T1Y; T1X = cr[WS(rs, 42)]; T1Z = ci[WS(rs, 42)]; T1W = W[82]; T1Y = W[83]; T20 = FMA(T1W, T1X, T1Y * T1Z); T6T = FNMS(T1Y, T1X, T1W * T1Z); } T21 = T1V + T20; Td4 = T6S + T6T; T6R = T1V - T20; T6U = T6S - T6T; } { E T26, T6X, T2b, T6Y; { E T23, T25, T22, T24; T23 = cr[WS(rs, 58)]; T25 = ci[WS(rs, 58)]; T22 = W[114]; T24 = W[115]; T26 = FMA(T22, T23, T24 * T25); T6X = FNMS(T24, T23, T22 * T25); } { E T28, T2a, T27, T29; T28 = cr[WS(rs, 26)]; T2a = ci[WS(rs, 26)]; T27 = W[50]; T29 = W[51]; T2b = FMA(T27, T28, T29 * T2a); T6Y = FNMS(T29, T28, T27 * T2a); } T2c = T26 + T2b; Td3 = T6X + T6Y; T6W = T26 - T2b; T6Z = T6X - T6Y; } T1Q = T1E + T1P; T2d = T21 + T2c; Tfu = T1Q - T2d; Tfv = Td7 + Td8; Tfw = Td4 + Td3; Tfx = Tfv - Tfw; { E T6M, T6P, Td9, Tda; T6M = T1y - T1D; T6P = T6N - T6O; T6Q = T6M - T6P; TaM = T6M + T6P; Td9 = Td7 - Td8; Tda = T21 - T2c; Tdb = Td9 - Tda; TeI = Td9 + Tda; } { E T6V, T70, T78, T79; T6V = T6R - T6U; T70 = T6W + T6Z; T71 = KP707106781 * (T6V + T70); TaQ = KP707106781 * (T70 - T6V); T78 = T6R + T6U; T79 = T6Z - T6W; T7a = KP707106781 * (T78 + T79); TaN = KP707106781 * (T78 - T79); } { E Td2, Td5, T75, T76; Td2 = T1E - T1P; Td5 = Td3 - Td4; Td6 = Td2 - Td5; TeJ = Td2 + Td5; T75 = T73 - T74; T76 = T1J - T1O; T77 = T75 + T76; TaP = T75 - T76; } } { E T2j, T7u, T2o, T7v, T2p, Tdd, T2u, T7e, T2z, T7f, T2A, Tde, T2M, Tdk, T7i; E T7l, T2X, Tdj, T7n, T7q; { E T2g, T2i, T2f, T2h; T2g = cr[WS(rs, 62)]; T2i = ci[WS(rs, 62)]; T2f = W[122]; T2h = W[123]; T2j = FMA(T2f, T2g, T2h * T2i); T7u = FNMS(T2h, T2g, T2f * T2i); } { E T2l, T2n, T2k, T2m; T2l = cr[WS(rs, 30)]; T2n = ci[WS(rs, 30)]; T2k = W[58]; T2m = W[59]; T2o = FMA(T2k, T2l, T2m * T2n); T7v = FNMS(T2m, T2l, T2k * T2n); } T2p = T2j + T2o; Tdd = T7u + T7v; { E T2r, T2t, T2q, T2s; T2r = cr[WS(rs, 14)]; T2t = ci[WS(rs, 14)]; T2q = W[26]; T2s = W[27]; T2u = FMA(T2q, T2r, T2s * T2t); T7e = FNMS(T2s, T2r, T2q * T2t); } { E T2w, T2y, T2v, T2x; T2w = cr[WS(rs, 46)]; T2y = ci[WS(rs, 46)]; T2v = W[90]; T2x = W[91]; T2z = FMA(T2v, T2w, T2x * T2y); T7f = FNMS(T2x, T2w, T2v * T2y); } T2A = T2u + T2z; Tde = T7e + T7f; { E T2G, T7j, T2L, T7k; { E T2D, T2F, T2C, T2E; T2D = cr[WS(rs, 6)]; T2F = ci[WS(rs, 6)]; T2C = W[10]; T2E = W[11]; T2G = FMA(T2C, T2D, T2E * T2F); T7j = FNMS(T2E, T2D, T2C * T2F); } { E T2I, T2K, T2H, T2J; T2I = cr[WS(rs, 38)]; T2K = ci[WS(rs, 38)]; T2H = W[74]; T2J = W[75]; T2L = FMA(T2H, T2I, T2J * T2K); T7k = FNMS(T2J, T2I, T2H * T2K); } T2M = T2G + T2L; Tdk = T7j + T7k; T7i = T2G - T2L; T7l = T7j - T7k; } { E T2R, T7o, T2W, T7p; { E T2O, T2Q, T2N, T2P; T2O = cr[WS(rs, 54)]; T2Q = ci[WS(rs, 54)]; T2N = W[106]; T2P = W[107]; T2R = FMA(T2N, T2O, T2P * T2Q); T7o = FNMS(T2P, T2O, T2N * T2Q); } { E T2T, T2V, T2S, T2U; T2T = cr[WS(rs, 22)]; T2V = ci[WS(rs, 22)]; T2S = W[42]; T2U = W[43]; T2W = FMA(T2S, T2T, T2U * T2V); T7p = FNMS(T2U, T2T, T2S * T2V); } T2X = T2R + T2W; Tdj = T7o + T7p; T7n = T2R - T2W; T7q = T7o - T7p; } T2B = T2p + T2A; T2Y = T2M + T2X; Tfz = T2B - T2Y; TfA = Tdd + Tde; TfB = Tdk + Tdj; TfC = TfA - TfB; { E T7d, T7g, Tdi, Tdl; T7d = T2j - T2o; T7g = T7e - T7f; T7h = T7d - T7g; TaW = T7d + T7g; Tdi = T2p - T2A; Tdl = Tdj - Tdk; Tdm = Tdi - Tdl; TeL = Tdi + Tdl; } { E T7m, T7r, T7z, T7A; T7m = T7i - T7l; T7r = T7n + T7q; T7s = KP707106781 * (T7m + T7r); TaU = KP707106781 * (T7r - T7m); T7z = T7i + T7l; T7A = T7q - T7n; T7B = KP707106781 * (T7z + T7A); TaX = KP707106781 * (T7z - T7A); } { E Tdf, Tdg, T7w, T7x; Tdf = Tdd - Tde; Tdg = T2M - T2X; Tdh = Tdf - Tdg; TeM = Tdf + Tdg; T7w = T7u - T7v; T7x = T2u - T2z; T7y = T7w + T7x; TaT = T7w - T7x; } } { E T4D, T9e, T4I, T9f, T4J, TdR, T4O, T8A, T4T, T8B, T4U, TdS, T56, Tea, T8E; E T8H, T5h, Te9, T8J, T8M; { E T4A, T4C, T4z, T4B; T4A = cr[WS(rs, 63)]; T4C = ci[WS(rs, 63)]; T4z = W[124]; T4B = W[125]; T4D = FMA(T4z, T4A, T4B * T4C); T9e = FNMS(T4B, T4A, T4z * T4C); } { E T4F, T4H, T4E, T4G; T4F = cr[WS(rs, 31)]; T4H = ci[WS(rs, 31)]; T4E = W[60]; T4G = W[61]; T4I = FMA(T4E, T4F, T4G * T4H); T9f = FNMS(T4G, T4F, T4E * T4H); } T4J = T4D + T4I; TdR = T9e + T9f; { E T4L, T4N, T4K, T4M; T4L = cr[WS(rs, 15)]; T4N = ci[WS(rs, 15)]; T4K = W[28]; T4M = W[29]; T4O = FMA(T4K, T4L, T4M * T4N); T8A = FNMS(T4M, T4L, T4K * T4N); } { E T4Q, T4S, T4P, T4R; T4Q = cr[WS(rs, 47)]; T4S = ci[WS(rs, 47)]; T4P = W[92]; T4R = W[93]; T4T = FMA(T4P, T4Q, T4R * T4S); T8B = FNMS(T4R, T4Q, T4P * T4S); } T4U = T4O + T4T; TdS = T8A + T8B; { E T50, T8F, T55, T8G; { E T4X, T4Z, T4W, T4Y; T4X = cr[WS(rs, 7)]; T4Z = ci[WS(rs, 7)]; T4W = W[12]; T4Y = W[13]; T50 = FMA(T4W, T4X, T4Y * T4Z); T8F = FNMS(T4Y, T4X, T4W * T4Z); } { E T52, T54, T51, T53; T52 = cr[WS(rs, 39)]; T54 = ci[WS(rs, 39)]; T51 = W[76]; T53 = W[77]; T55 = FMA(T51, T52, T53 * T54); T8G = FNMS(T53, T52, T51 * T54); } T56 = T50 + T55; Tea = T8F + T8G; T8E = T50 - T55; T8H = T8F - T8G; } { E T5b, T8K, T5g, T8L; { E T58, T5a, T57, T59; T58 = cr[WS(rs, 55)]; T5a = ci[WS(rs, 55)]; T57 = W[108]; T59 = W[109]; T5b = FMA(T57, T58, T59 * T5a); T8K = FNMS(T59, T58, T57 * T5a); } { E T5d, T5f, T5c, T5e; T5d = cr[WS(rs, 23)]; T5f = ci[WS(rs, 23)]; T5c = W[44]; T5e = W[45]; T5g = FMA(T5c, T5d, T5e * T5f); T8L = FNMS(T5e, T5d, T5c * T5f); } T5h = T5b + T5g; Te9 = T8K + T8L; T8J = T5b - T5g; T8M = T8K - T8L; } { E T4V, T5i, Te8, Teb; T4V = T4J + T4U; T5i = T56 + T5h; T5j = T4V + T5i; TfR = T4V - T5i; Te8 = T4J - T4U; Teb = Te9 - Tea; Tec = Te8 - Teb; TeX = Te8 + Teb; } { E TfW, TfX, T8z, T8C; TfW = TdR + TdS; TfX = Tea + Te9; TfY = TfW - TfX; Tgy = TfW + TfX; T8z = T4D - T4I; T8C = T8A - T8B; T8D = T8z - T8C; Tbl = T8z + T8C; } { E T8I, T8N, T9j, T9k; T8I = T8E - T8H; T8N = T8J + T8M; T8O = KP707106781 * (T8I + T8N); Tbx = KP707106781 * (T8N - T8I); T9j = T8E + T8H; T9k = T8M - T8J; T9l = KP707106781 * (T9j + T9k); Tbm = KP707106781 * (T9j - T9k); } { E TdT, TdU, T9g, T9h; TdT = TdR - TdS; TdU = T56 - T5h; TdV = TdT - TdU; Tf0 = TdT + TdU; T9g = T9e - T9f; T9h = T4O - T4T; T9i = T9g + T9h; Tbw = T9g - T9h; } } { E T36, T7G, T3b, T7H, T3c, TdH, T3h, T8m, T3m, T8n, T3n, TdI, T3z, Tds, T7L; E T7O, T3K, Tdr, T7S, T7T; { E T33, T35, T32, T34; T33 = cr[WS(rs, 1)]; T35 = ci[WS(rs, 1)]; T32 = W[0]; T34 = W[1]; T36 = FMA(T32, T33, T34 * T35); T7G = FNMS(T34, T33, T32 * T35); } { E T38, T3a, T37, T39; T38 = cr[WS(rs, 33)]; T3a = ci[WS(rs, 33)]; T37 = W[64]; T39 = W[65]; T3b = FMA(T37, T38, T39 * T3a); T7H = FNMS(T39, T38, T37 * T3a); } T3c = T36 + T3b; TdH = T7G + T7H; { E T3e, T3g, T3d, T3f; T3e = cr[WS(rs, 17)]; T3g = ci[WS(rs, 17)]; T3d = W[32]; T3f = W[33]; T3h = FMA(T3d, T3e, T3f * T3g); T8m = FNMS(T3f, T3e, T3d * T3g); } { E T3j, T3l, T3i, T3k; T3j = cr[WS(rs, 49)]; T3l = ci[WS(rs, 49)]; T3i = W[96]; T3k = W[97]; T3m = FMA(T3i, T3j, T3k * T3l); T8n = FNMS(T3k, T3j, T3i * T3l); } T3n = T3h + T3m; TdI = T8m + T8n; { E T3t, T7M, T3y, T7N; { E T3q, T3s, T3p, T3r; T3q = cr[WS(rs, 9)]; T3s = ci[WS(rs, 9)]; T3p = W[16]; T3r = W[17]; T3t = FMA(T3p, T3q, T3r * T3s); T7M = FNMS(T3r, T3q, T3p * T3s); } { E T3v, T3x, T3u, T3w; T3v = cr[WS(rs, 41)]; T3x = ci[WS(rs, 41)]; T3u = W[80]; T3w = W[81]; T3y = FMA(T3u, T3v, T3w * T3x); T7N = FNMS(T3w, T3v, T3u * T3x); } T3z = T3t + T3y; Tds = T7M + T7N; T7L = T3t - T3y; T7O = T7M - T7N; } { E T3E, T7Q, T3J, T7R; { E T3B, T3D, T3A, T3C; T3B = cr[WS(rs, 57)]; T3D = ci[WS(rs, 57)]; T3A = W[112]; T3C = W[113]; T3E = FMA(T3A, T3B, T3C * T3D); T7Q = FNMS(T3C, T3B, T3A * T3D); } { E T3G, T3I, T3F, T3H; T3G = cr[WS(rs, 25)]; T3I = ci[WS(rs, 25)]; T3F = W[48]; T3H = W[49]; T3J = FMA(T3F, T3G, T3H * T3I); T7R = FNMS(T3H, T3G, T3F * T3I); } T3K = T3E + T3J; Tdr = T7Q + T7R; T7S = T7Q - T7R; T7T = T3E - T3J; } { E T3o, T3L, TdJ, TdK; T3o = T3c + T3n; T3L = T3z + T3K; T3M = T3o + T3L; TfL = T3o - T3L; TdJ = TdH - TdI; TdK = T3z - T3K; TdL = TdJ - TdK; TeT = TdJ + TdK; } { E TfG, TfH, T7I, T7J; TfG = TdH + TdI; TfH = Tds + Tdr; TfI = TfG - TfH; Tgt = TfG + TfH; T7I = T7G - T7H; T7J = T3h - T3m; T7K = T7I + T7J; Tbd = T7I - T7J; } { E T7P, T7U, T8q, T8r; T7P = T7L + T7O; T7U = T7S - T7T; T7V = KP707106781 * (T7P + T7U); Tb3 = KP707106781 * (T7P - T7U); T8q = T7L - T7O; T8r = T7T + T7S; T8s = KP707106781 * (T8q + T8r); Tbe = KP707106781 * (T8r - T8q); } { E Tdq, Tdt, T8l, T8o; Tdq = T3c - T3n; Tdt = Tdr - Tds; Tdu = Tdq - Tdt; TeQ = Tdq + Tdt; T8l = T36 - T3b; T8o = T8m - T8n; T8p = T8l - T8o; Tb2 = T8l + T8o; } } { E T3X, Tdw, T7Z, T82, T4v, TdB, T8b, T8g, T48, Tdx, T80, T85, T4k, TdA, T8a; E T8d; { E T3R, T7X, T3W, T7Y; { E T3O, T3Q, T3N, T3P; T3O = cr[WS(rs, 5)]; T3Q = ci[WS(rs, 5)]; T3N = W[8]; T3P = W[9]; T3R = FMA(T3N, T3O, T3P * T3Q); T7X = FNMS(T3P, T3O, T3N * T3Q); } { E T3T, T3V, T3S, T3U; T3T = cr[WS(rs, 37)]; T3V = ci[WS(rs, 37)]; T3S = W[72]; T3U = W[73]; T3W = FMA(T3S, T3T, T3U * T3V); T7Y = FNMS(T3U, T3T, T3S * T3V); } T3X = T3R + T3W; Tdw = T7X + T7Y; T7Z = T7X - T7Y; T82 = T3R - T3W; } { E T4p, T8e, T4u, T8f; { E T4m, T4o, T4l, T4n; T4m = cr[WS(rs, 13)]; T4o = ci[WS(rs, 13)]; T4l = W[24]; T4n = W[25]; T4p = FMA(T4l, T4m, T4n * T4o); T8e = FNMS(T4n, T4m, T4l * T4o); } { E T4r, T4t, T4q, T4s; T4r = cr[WS(rs, 45)]; T4t = ci[WS(rs, 45)]; T4q = W[88]; T4s = W[89]; T4u = FMA(T4q, T4r, T4s * T4t); T8f = FNMS(T4s, T4r, T4q * T4t); } T4v = T4p + T4u; TdB = T8e + T8f; T8b = T4p - T4u; T8g = T8e - T8f; } { E T42, T83, T47, T84; { E T3Z, T41, T3Y, T40; T3Z = cr[WS(rs, 21)]; T41 = ci[WS(rs, 21)]; T3Y = W[40]; T40 = W[41]; T42 = FMA(T3Y, T3Z, T40 * T41); T83 = FNMS(T40, T3Z, T3Y * T41); } { E T44, T46, T43, T45; T44 = cr[WS(rs, 53)]; T46 = ci[WS(rs, 53)]; T43 = W[104]; T45 = W[105]; T47 = FMA(T43, T44, T45 * T46); T84 = FNMS(T45, T44, T43 * T46); } T48 = T42 + T47; Tdx = T83 + T84; T80 = T42 - T47; T85 = T83 - T84; } { E T4e, T88, T4j, T89; { E T4b, T4d, T4a, T4c; T4b = cr[WS(rs, 61)]; T4d = ci[WS(rs, 61)]; T4a = W[120]; T4c = W[121]; T4e = FMA(T4a, T4b, T4c * T4d); T88 = FNMS(T4c, T4b, T4a * T4d); } { E T4g, T4i, T4f, T4h; T4g = cr[WS(rs, 29)]; T4i = ci[WS(rs, 29)]; T4f = W[56]; T4h = W[57]; T4j = FMA(T4f, T4g, T4h * T4i); T89 = FNMS(T4h, T4g, T4f * T4i); } T4k = T4e + T4j; TdA = T88 + T89; T8a = T88 - T89; T8d = T4e - T4j; } { E T49, T4w, TdC, TdD; T49 = T3X + T48; T4w = T4k + T4v; T4x = T49 + T4w; TfJ = T49 - T4w; TdC = TdA - TdB; TdD = T4k - T4v; TdE = TdC - TdD; TdM = TdD + TdC; } { E TfM, TfN, T81, T86; TfM = TdA + TdB; TfN = Tdw + Tdx; TfO = TfM - TfN; Tgu = TfN + TfM; T81 = T7Z + T80; T86 = T82 - T85; T87 = FMA(KP923879532, T81, KP382683432 * T86); T8u = FNMS(KP382683432, T81, KP923879532 * T86); } { E T8c, T8h, Tb8, Tb9; T8c = T8a + T8b; T8h = T8d - T8g; T8i = FNMS(KP382683432, T8h, KP923879532 * T8c); T8v = FMA(KP382683432, T8c, KP923879532 * T8h); Tb8 = T8d + T8g; Tb9 = T8a - T8b; Tba = FNMS(KP382683432, Tb9, KP923879532 * Tb8); Tbh = FMA(KP923879532, Tb9, KP382683432 * Tb8); } { E Tdv, Tdy, Tb5, Tb6; Tdv = T3X - T48; Tdy = Tdw - Tdx; Tdz = Tdv + Tdy; TdN = Tdv - Tdy; Tb5 = T7Z - T80; Tb6 = T82 + T85; Tb7 = FMA(KP382683432, Tb5, KP923879532 * Tb6); Tbg = FNMS(KP382683432, Tb6, KP923879532 * Tb5); } } { E T5u, Te2, T8Q, T8X, T62, TdY, T94, T99, T5F, Te3, T8T, T8Y, T5R, TdX, T93; E T96; { E T5o, T8V, T5t, T8W; { E T5l, T5n, T5k, T5m; T5l = cr[WS(rs, 3)]; T5n = ci[WS(rs, 3)]; T5k = W[4]; T5m = W[5]; T5o = FMA(T5k, T5l, T5m * T5n); T8V = FNMS(T5m, T5l, T5k * T5n); } { E T5q, T5s, T5p, T5r; T5q = cr[WS(rs, 35)]; T5s = ci[WS(rs, 35)]; T5p = W[68]; T5r = W[69]; T5t = FMA(T5p, T5q, T5r * T5s); T8W = FNMS(T5r, T5q, T5p * T5s); } T5u = T5o + T5t; Te2 = T8V + T8W; T8Q = T5o - T5t; T8X = T8V - T8W; } { E T5W, T97, T61, T98; { E T5T, T5V, T5S, T5U; T5T = cr[WS(rs, 11)]; T5V = ci[WS(rs, 11)]; T5S = W[20]; T5U = W[21]; T5W = FMA(T5S, T5T, T5U * T5V); T97 = FNMS(T5U, T5T, T5S * T5V); } { E T5Y, T60, T5X, T5Z; T5Y = cr[WS(rs, 43)]; T60 = ci[WS(rs, 43)]; T5X = W[84]; T5Z = W[85]; T61 = FMA(T5X, T5Y, T5Z * T60); T98 = FNMS(T5Z, T5Y, T5X * T60); } T62 = T5W + T61; TdY = T97 + T98; T94 = T5W - T61; T99 = T97 - T98; } { E T5z, T8R, T5E, T8S; { E T5w, T5y, T5v, T5x; T5w = cr[WS(rs, 19)]; T5y = ci[WS(rs, 19)]; T5v = W[36]; T5x = W[37]; T5z = FMA(T5v, T5w, T5x * T5y); T8R = FNMS(T5x, T5w, T5v * T5y); } { E T5B, T5D, T5A, T5C; T5B = cr[WS(rs, 51)]; T5D = ci[WS(rs, 51)]; T5A = W[100]; T5C = W[101]; T5E = FMA(T5A, T5B, T5C * T5D); T8S = FNMS(T5C, T5B, T5A * T5D); } T5F = T5z + T5E; Te3 = T8R + T8S; T8T = T8R - T8S; T8Y = T5z - T5E; } { E T5L, T91, T5Q, T92; { E T5I, T5K, T5H, T5J; T5I = cr[WS(rs, 59)]; T5K = ci[WS(rs, 59)]; T5H = W[116]; T5J = W[117]; T5L = FMA(T5H, T5I, T5J * T5K); T91 = FNMS(T5J, T5I, T5H * T5K); } { E T5N, T5P, T5M, T5O; T5N = cr[WS(rs, 27)]; T5P = ci[WS(rs, 27)]; T5M = W[52]; T5O = W[53]; T5Q = FMA(T5M, T5N, T5O * T5P); T92 = FNMS(T5O, T5N, T5M * T5P); } T5R = T5L + T5Q; TdX = T91 + T92; T93 = T91 - T92; T96 = T5L - T5Q; } { E T5G, T63, Te1, Te4; T5G = T5u + T5F; T63 = T5R + T62; T64 = T5G + T63; TfZ = T5G - T63; Te1 = T5u - T5F; Te4 = Te2 - Te3; Te5 = Te1 - Te4; Ted = Te1 + Te4; } { E TfS, TfT, T8U, T8Z; TfS = TdX + TdY; TfT = Te2 + Te3; TfU = TfS - TfT; Tgz = TfT + TfS; T8U = T8Q - T8T; T8Z = T8X + T8Y; T90 = FNMS(KP382683432, T8Z, KP923879532 * T8U); T9n = FMA(KP923879532, T8Z, KP382683432 * T8U); } { E T95, T9a, Tbr, Tbs; T95 = T93 + T94; T9a = T96 - T99; T9b = FMA(KP382683432, T95, KP923879532 * T9a); T9o = FNMS(KP382683432, T9a, KP923879532 * T95); Tbr = T96 + T99; Tbs = T93 - T94; Tbt = FNMS(KP382683432, Tbs, KP923879532 * Tbr); TbA = FMA(KP923879532, Tbs, KP382683432 * Tbr); } { E TdW, TdZ, Tbo, Tbp; TdW = T5R - T62; TdZ = TdX - TdY; Te0 = TdW + TdZ; Tee = TdZ - TdW; Tbo = T8X - T8Y; Tbp = T8Q + T8T; Tbq = FMA(KP382683432, Tbo, KP923879532 * Tbp); Tbz = FNMS(KP382683432, Tbp, KP923879532 * Tbo); } } { E T1t, Tgn, TgK, TgL, TgV, Th1, T30, Th0, T66, TgX, Tgw, TgE, TgB, TgF, Tgq; E TgM; { E TH, T1s, TgI, TgJ; TH = Tj + TG; T1s = T14 + T1r; T1t = TH + T1s; Tgn = TH - T1s; TgI = Tgy + Tgz; TgJ = Tgt + Tgu; TgK = TgI - TgJ; TgL = TgJ + TgI; } { E TgN, TgU, T2e, T2Z; TgN = Tfr + Tfq; TgU = TgO + TgT; TgV = TgN + TgU; Th1 = TgU - TgN; T2e = T1Q + T2d; T2Z = T2B + T2Y; T30 = T2e + T2Z; Th0 = T2e - T2Z; } { E T4y, T65, Tgs, Tgv; T4y = T3M + T4x; T65 = T5j + T64; T66 = T4y + T65; TgX = T65 - T4y; Tgs = T3M - T4x; Tgv = Tgt - Tgu; Tgw = Tgs + Tgv; TgE = Tgs - Tgv; } { E Tgx, TgA, Tgo, Tgp; Tgx = T5j - T64; TgA = Tgy - Tgz; TgB = Tgx - TgA; TgF = Tgx + TgA; Tgo = TfA + TfB; Tgp = Tfv + Tfw; Tgq = Tgo - Tgp; TgM = Tgp + Tgo; } { E T31, TgW, TgY, TgH; T31 = T1t + T30; ci[WS(rs, 31)] = T31 - T66; cr[0] = T31 + T66; TgW = TgM + TgV; cr[WS(rs, 32)] = TgL - TgW; ci[WS(rs, 63)] = TgL + TgW; TgY = TgV - TgM; cr[WS(rs, 48)] = TgX - TgY; ci[WS(rs, 47)] = TgX + TgY; TgH = T1t - T30; cr[WS(rs, 16)] = TgH - TgK; ci[WS(rs, 15)] = TgH + TgK; } { E Tgr, TgC, TgZ, Th2; Tgr = Tgn - Tgq; TgC = KP707106781 * (Tgw + TgB); ci[WS(rs, 23)] = Tgr - TgC; cr[WS(rs, 8)] = Tgr + TgC; TgZ = KP707106781 * (TgB - Tgw); Th2 = Th0 + Th1; cr[WS(rs, 56)] = TgZ - Th2; ci[WS(rs, 39)] = TgZ + Th2; } { E Th3, Th4, TgD, TgG; Th3 = KP707106781 * (TgF - TgE); Th4 = Th1 - Th0; cr[WS(rs, 40)] = Th3 - Th4; ci[WS(rs, 55)] = Th3 + Th4; TgD = Tgn + Tgq; TgG = KP707106781 * (TgE + TgF); cr[WS(rs, 24)] = TgD - TgG; ci[WS(rs, 7)] = TgD + TgG; } } { E T6L, T9x, ThV, Ti1, T7E, Ti0, T9A, ThO, T8y, T9K, T9u, T9E, T9r, T9L, T9v; E T9H; { E T6n, T6K, ThP, ThU; T6n = T6b + T6m; T6K = T6y + T6J; T6L = T6n - T6K; T9x = T6n + T6K; ThP = T9O - T9P; ThU = ThQ + ThT; ThV = ThP + ThU; Ti1 = ThU - ThP; } { E T7c, T9y, T7D, T9z; { E T72, T7b, T7t, T7C; T72 = T6Q + T71; T7b = T77 + T7a; T7c = FMA(KP195090322, T72, KP980785280 * T7b); T9y = FNMS(KP195090322, T7b, KP980785280 * T72); T7t = T7h + T7s; T7C = T7y + T7B; T7D = FNMS(KP980785280, T7C, KP195090322 * T7t); T9z = FMA(KP980785280, T7t, KP195090322 * T7C); } T7E = T7c + T7D; Ti0 = T9z - T9y; T9A = T9y + T9z; ThO = T7c - T7D; } { E T8k, T9D, T8x, T9C; { E T7W, T8j, T8t, T8w; T7W = T7K + T7V; T8j = T87 + T8i; T8k = T7W - T8j; T9D = T7W + T8j; T8t = T8p + T8s; T8w = T8u + T8v; T8x = T8t - T8w; T9C = T8t + T8w; } T8y = FMA(KP634393284, T8k, KP773010453 * T8x); T9K = FMA(KP995184726, T9D, KP098017140 * T9C); T9u = FNMS(KP773010453, T8k, KP634393284 * T8x); T9E = FNMS(KP098017140, T9D, KP995184726 * T9C); } { E T9d, T9G, T9q, T9F; { E T8P, T9c, T9m, T9p; T8P = T8D + T8O; T9c = T90 + T9b; T9d = T8P - T9c; T9G = T8P + T9c; T9m = T9i + T9l; T9p = T9n + T9o; T9q = T9m - T9p; T9F = T9m + T9p; } T9r = FNMS(KP634393284, T9q, KP773010453 * T9d); T9L = FNMS(KP995184726, T9F, KP098017140 * T9G); T9v = FMA(KP773010453, T9q, KP634393284 * T9d); T9H = FMA(KP098017140, T9F, KP995184726 * T9G); } { E T7F, T9s, ThZ, Ti2; T7F = T6L + T7E; T9s = T8y + T9r; ci[WS(rs, 24)] = T7F - T9s; cr[WS(rs, 7)] = T7F + T9s; ThZ = T9v - T9u; Ti2 = Ti0 + Ti1; cr[WS(rs, 39)] = ThZ - Ti2; ci[WS(rs, 56)] = ThZ + Ti2; } { E Ti3, Ti4, T9t, T9w; Ti3 = T9r - T8y; Ti4 = Ti1 - Ti0; cr[WS(rs, 55)] = Ti3 - Ti4; ci[WS(rs, 40)] = Ti3 + Ti4; T9t = T6L - T7E; T9w = T9u + T9v; cr[WS(rs, 23)] = T9t - T9w; ci[WS(rs, 8)] = T9t + T9w; } { E T9B, T9I, ThN, ThW; T9B = T9x + T9A; T9I = T9E + T9H; cr[WS(rs, 31)] = T9B - T9I; ci[0] = T9B + T9I; ThN = T9L - T9K; ThW = ThO + ThV; cr[WS(rs, 63)] = ThN - ThW; ci[WS(rs, 32)] = ThN + ThW; } { E ThX, ThY, T9J, T9M; ThX = T9H - T9E; ThY = ThV - ThO; cr[WS(rs, 47)] = ThX - ThY; ci[WS(rs, 48)] = ThX + ThY; T9J = T9x - T9A; T9M = T9K + T9L; ci[WS(rs, 16)] = T9J - T9M; cr[WS(rs, 15)] = T9J + T9M; } } { E Tft, Tg7, Tgh, Tgl, Th9, Thf, TfE, Th6, TfQ, Tg4, Tga, The, Tge, Tgk, Tg1; E Tg5; { E Tfp, Tfs, Tgf, Tgg; Tfp = Tj - TG; Tfs = Tfq - Tfr; Tft = Tfp - Tfs; Tg7 = Tfp + Tfs; Tgf = TfY + TfZ; Tgg = TfR + TfU; Tgh = FMA(KP382683432, Tgf, KP923879532 * Tgg); Tgl = FNMS(KP923879532, Tgf, KP382683432 * Tgg); } { E Th7, Th8, Tfy, TfD; Th7 = T14 - T1r; Th8 = TgT - TgO; Th9 = Th7 + Th8; Thf = Th8 - Th7; Tfy = Tfu + Tfx; TfD = Tfz - TfC; TfE = KP707106781 * (Tfy + TfD); Th6 = KP707106781 * (Tfy - TfD); } { E TfK, TfP, Tg8, Tg9; TfK = TfI - TfJ; TfP = TfL - TfO; TfQ = FMA(KP382683432, TfK, KP923879532 * TfP); Tg4 = FNMS(KP923879532, TfK, KP382683432 * TfP); Tg8 = Tfu - Tfx; Tg9 = Tfz + TfC; Tga = KP707106781 * (Tg8 + Tg9); The = KP707106781 * (Tg9 - Tg8); } { E Tgc, Tgd, TfV, Tg0; Tgc = TfL + TfO; Tgd = TfI + TfJ; Tge = FNMS(KP382683432, Tgd, KP923879532 * Tgc); Tgk = FMA(KP923879532, Tgd, KP382683432 * Tgc); TfV = TfR - TfU; Tg0 = TfY - TfZ; Tg1 = FNMS(KP382683432, Tg0, KP923879532 * TfV); Tg5 = FMA(KP923879532, Tg0, KP382683432 * TfV); } { E TfF, Tg2, Thd, Thg; TfF = Tft + TfE; Tg2 = TfQ + Tg1; ci[WS(rs, 27)] = TfF - Tg2; cr[WS(rs, 4)] = TfF + Tg2; Thd = Tg5 - Tg4; Thg = The + Thf; cr[WS(rs, 36)] = Thd - Thg; ci[WS(rs, 59)] = Thd + Thg; } { E Thh, Thi, Tg3, Tg6; Thh = Tg1 - TfQ; Thi = Thf - The; cr[WS(rs, 52)] = Thh - Thi; ci[WS(rs, 43)] = Thh + Thi; Tg3 = Tft - TfE; Tg6 = Tg4 + Tg5; cr[WS(rs, 20)] = Tg3 - Tg6; ci[WS(rs, 11)] = Tg3 + Tg6; } { E Tgb, Tgi, Th5, Tha; Tgb = Tg7 + Tga; Tgi = Tge + Tgh; cr[WS(rs, 28)] = Tgb - Tgi; ci[WS(rs, 3)] = Tgb + Tgi; Th5 = Tgl - Tgk; Tha = Th6 + Th9; cr[WS(rs, 60)] = Th5 - Tha; ci[WS(rs, 35)] = Th5 + Tha; } { E Thb, Thc, Tgj, Tgm; Thb = Tgh - Tge; Thc = Th9 - Th6; cr[WS(rs, 44)] = Thb - Thc; ci[WS(rs, 51)] = Thb + Thc; Tgj = Tg7 - Tga; Tgm = Tgk + Tgl; ci[WS(rs, 19)] = Tgj - Tgm; cr[WS(rs, 12)] = Tgj + Tgm; } } { E TeH, Tf9, TeO, Thk, Thp, Thv, Tfc, Thu, Tf3, Tfn, Tf7, Tfj, TeW, Tfm, Tf6; E Tfg; { E TeD, TeG, Tfa, Tfb; TeD = TcL + TcO; TeG = KP707106781 * (TeE + TeF); TeH = TeD - TeG; Tf9 = TeD + TeG; { E TeK, TeN, Thl, Tho; TeK = FMA(KP923879532, TeI, KP382683432 * TeJ); TeN = FNMS(KP923879532, TeM, KP382683432 * TeL); TeO = TeK + TeN; Thk = TeK - TeN; Thl = KP707106781 * (TcU - TcZ); Tho = Thm + Thn; Thp = Thl + Tho; Thv = Tho - Thl; } Tfa = FNMS(KP382683432, TeI, KP923879532 * TeJ); Tfb = FMA(KP382683432, TeM, KP923879532 * TeL); Tfc = Tfa + Tfb; Thu = Tfb - Tfa; { E TeZ, Tfh, Tf2, Tfi, TeY, Tf1; TeY = KP707106781 * (Te5 + Te0); TeZ = TeX - TeY; Tfh = TeX + TeY; Tf1 = KP707106781 * (Ted + Tee); Tf2 = Tf0 - Tf1; Tfi = Tf0 + Tf1; Tf3 = FNMS(KP555570233, Tf2, KP831469612 * TeZ); Tfn = FMA(KP980785280, Tfh, KP195090322 * Tfi); Tf7 = FMA(KP555570233, TeZ, KP831469612 * Tf2); Tfj = FNMS(KP980785280, Tfi, KP195090322 * Tfh); } { E TeS, Tfe, TeV, Tff, TeR, TeU; TeR = KP707106781 * (TdN + TdM); TeS = TeQ - TeR; Tfe = TeQ + TeR; TeU = KP707106781 * (Tdz + TdE); TeV = TeT - TeU; Tff = TeT + TeU; TeW = FMA(KP831469612, TeS, KP555570233 * TeV); Tfm = FNMS(KP195090322, Tff, KP980785280 * Tfe); Tf6 = FNMS(KP831469612, TeV, KP555570233 * TeS); Tfg = FMA(KP195090322, Tfe, KP980785280 * Tff); } } { E TeP, Tf4, Tht, Thw; TeP = TeH + TeO; Tf4 = TeW + Tf3; ci[WS(rs, 25)] = TeP - Tf4; cr[WS(rs, 6)] = TeP + Tf4; Tht = Tf7 - Tf6; Thw = Thu + Thv; cr[WS(rs, 38)] = Tht - Thw; ci[WS(rs, 57)] = Tht + Thw; } { E Thx, Thy, Tf5, Tf8; Thx = Tf3 - TeW; Thy = Thv - Thu; cr[WS(rs, 54)] = Thx - Thy; ci[WS(rs, 41)] = Thx + Thy; Tf5 = TeH - TeO; Tf8 = Tf6 + Tf7; cr[WS(rs, 22)] = Tf5 - Tf8; ci[WS(rs, 9)] = Tf5 + Tf8; } { E Tfd, Tfk, Thj, Thq; Tfd = Tf9 - Tfc; Tfk = Tfg + Tfj; ci[WS(rs, 17)] = Tfd - Tfk; cr[WS(rs, 14)] = Tfd + Tfk; Thj = Tfj - Tfg; Thq = Thk + Thp; cr[WS(rs, 62)] = Thj - Thq; ci[WS(rs, 33)] = Thj + Thq; } { E Thr, Ths, Tfl, Tfo; Thr = Tfn - Tfm; Ths = Thp - Thk; cr[WS(rs, 46)] = Thr - Ths; ci[WS(rs, 49)] = Thr + Ths; Tfl = Tf9 + Tfc; Tfo = Tfm + Tfn; cr[WS(rs, 30)] = Tfl - Tfo; ci[WS(rs, 1)] = Tfl + Tfo; } } { E Td1, Ten, Tdo, ThA, ThD, ThJ, Teq, ThI, Teh, TeB, Tel, Tex, TdQ, TeA, Tek; E Teu; { E TcP, Td0, Teo, Tep; TcP = TcL - TcO; Td0 = KP707106781 * (TcU + TcZ); Td1 = TcP - Td0; Ten = TcP + Td0; { E Tdc, Tdn, ThB, ThC; Tdc = FNMS(KP923879532, Tdb, KP382683432 * Td6); Tdn = FMA(KP923879532, Tdh, KP382683432 * Tdm); Tdo = Tdc + Tdn; ThA = Tdn - Tdc; ThB = KP707106781 * (TeF - TeE); ThC = Thn - Thm; ThD = ThB + ThC; ThJ = ThC - ThB; } Teo = FMA(KP382683432, Tdb, KP923879532 * Td6); Tep = FNMS(KP382683432, Tdh, KP923879532 * Tdm); Teq = Teo + Tep; ThI = Teo - Tep; { E Te7, Tew, Teg, Tev, Te6, Tef; Te6 = KP707106781 * (Te0 - Te5); Te7 = TdV - Te6; Tew = TdV + Te6; Tef = KP707106781 * (Ted - Tee); Teg = Tec - Tef; Tev = Tec + Tef; Teh = FMA(KP555570233, Te7, KP831469612 * Teg); TeB = FMA(KP980785280, Tew, KP195090322 * Tev); Tel = FNMS(KP831469612, Te7, KP555570233 * Teg); Tex = FNMS(KP195090322, Tew, KP980785280 * Tev); } { E TdG, Tet, TdP, Tes, TdF, TdO; TdF = KP707106781 * (Tdz - TdE); TdG = Tdu - TdF; Tet = Tdu + TdF; TdO = KP707106781 * (TdM - TdN); TdP = TdL - TdO; Tes = TdL + TdO; TdQ = FNMS(KP555570233, TdP, KP831469612 * TdG); TeA = FNMS(KP980785280, Tes, KP195090322 * Tet); Tek = FMA(KP831469612, TdP, KP555570233 * TdG); Teu = FMA(KP195090322, Tes, KP980785280 * Tet); } } { E Tdp, Tei, ThH, ThK; Tdp = Td1 + Tdo; Tei = TdQ + Teh; cr[WS(rs, 26)] = Tdp - Tei; ci[WS(rs, 5)] = Tdp + Tei; ThH = Tel - Tek; ThK = ThI + ThJ; cr[WS(rs, 58)] = ThH - ThK; ci[WS(rs, 37)] = ThH + ThK; } { E ThL, ThM, Tej, Tem; ThL = Teh - TdQ; ThM = ThJ - ThI; cr[WS(rs, 42)] = ThL - ThM; ci[WS(rs, 53)] = ThL + ThM; Tej = Td1 - Tdo; Tem = Tek + Tel; ci[WS(rs, 21)] = Tej - Tem; cr[WS(rs, 10)] = Tej + Tem; } { E Ter, Tey, Thz, ThE; Ter = Ten + Teq; Tey = Teu + Tex; ci[WS(rs, 29)] = Ter - Tey; cr[WS(rs, 2)] = Ter + Tey; Thz = TeB - TeA; ThE = ThA + ThD; cr[WS(rs, 34)] = Thz - ThE; ci[WS(rs, 61)] = Thz + ThE; } { E ThF, ThG, Tez, TeC; ThF = Tex - Teu; ThG = ThD - ThA; cr[WS(rs, 50)] = ThF - ThG; ci[WS(rs, 45)] = ThF + ThG; Tez = Ten - Teq; TeC = TeA + TeB; cr[WS(rs, 18)] = Tez - TeC; ci[WS(rs, 13)] = Tez + TeC; } } { E Tc3, Tcv, TiD, TiJ, Tca, TiI, Tcy, TiA, Tci, TcI, Tcs, TcC, Tcp, TcJ, Tct; E TcF; { E TbZ, Tc2, TiB, TiC; TbZ = Taz - TaC; Tc2 = Tc0 + Tc1; Tc3 = TbZ - Tc2; Tcv = TbZ + Tc2; TiB = TaG - TaJ; TiC = Tin - Tim; TiD = TiB + TiC; TiJ = TiC - TiB; } { E Tc6, Tcw, Tc9, Tcx; { E Tc4, Tc5, Tc7, Tc8; Tc4 = TaP - TaQ; Tc5 = TaM - TaN; Tc6 = FMA(KP831469612, Tc4, KP555570233 * Tc5); Tcw = FNMS(KP555570233, Tc4, KP831469612 * Tc5); Tc7 = TaW - TaX; Tc8 = TaT - TaU; Tc9 = FNMS(KP831469612, Tc8, KP555570233 * Tc7); Tcx = FMA(KP555570233, Tc8, KP831469612 * Tc7); } Tca = Tc6 + Tc9; TiI = Tcx - Tcw; Tcy = Tcw + Tcx; TiA = Tc6 - Tc9; } { E Tce, TcB, Tch, TcA; { E Tcc, Tcd, Tcf, Tcg; Tcc = Tbd - Tbe; Tcd = Tb7 - Tba; Tce = Tcc - Tcd; TcB = Tcc + Tcd; Tcf = Tb2 - Tb3; Tcg = Tbh - Tbg; Tch = Tcf - Tcg; TcA = Tcf + Tcg; } Tci = FMA(KP471396736, Tce, KP881921264 * Tch); TcI = FMA(KP956940335, TcB, KP290284677 * TcA); Tcs = FNMS(KP881921264, Tce, KP471396736 * Tch); TcC = FNMS(KP290284677, TcB, KP956940335 * TcA); } { E Tcl, TcE, Tco, TcD; { E Tcj, Tck, Tcm, Tcn; Tcj = Tbl - Tbm; Tck = TbA - Tbz; Tcl = Tcj - Tck; TcE = Tcj + Tck; Tcm = Tbw - Tbx; Tcn = Tbq - Tbt; Tco = Tcm - Tcn; TcD = Tcm + Tcn; } Tcp = FNMS(KP471396736, Tco, KP881921264 * Tcl); TcJ = FNMS(KP956940335, TcD, KP290284677 * TcE); Tct = FMA(KP881921264, Tco, KP471396736 * Tcl); TcF = FMA(KP290284677, TcD, KP956940335 * TcE); } { E Tcb, Tcq, TiH, TiK; Tcb = Tc3 + Tca; Tcq = Tci + Tcp; ci[WS(rs, 26)] = Tcb - Tcq; cr[WS(rs, 5)] = Tcb + Tcq; TiH = Tct - Tcs; TiK = TiI + TiJ; cr[WS(rs, 37)] = TiH - TiK; ci[WS(rs, 58)] = TiH + TiK; } { E TiL, TiM, Tcr, Tcu; TiL = Tcp - Tci; TiM = TiJ - TiI; cr[WS(rs, 53)] = TiL - TiM; ci[WS(rs, 42)] = TiL + TiM; Tcr = Tc3 - Tca; Tcu = Tcs + Tct; cr[WS(rs, 21)] = Tcr - Tcu; ci[WS(rs, 10)] = Tcr + Tcu; } { E Tcz, TcG, Tiz, TiE; Tcz = Tcv + Tcy; TcG = TcC + TcF; cr[WS(rs, 29)] = Tcz - TcG; ci[WS(rs, 2)] = Tcz + TcG; Tiz = TcJ - TcI; TiE = TiA + TiD; cr[WS(rs, 61)] = Tiz - TiE; ci[WS(rs, 34)] = Tiz + TiE; } { E TiF, TiG, TcH, TcK; TiF = TcF - TcC; TiG = TiD - TiA; cr[WS(rs, 45)] = TiF - TiG; ci[WS(rs, 50)] = TiF + TiG; TcH = Tcv - Tcy; TcK = TcI + TcJ; ci[WS(rs, 18)] = TcH - TcK; cr[WS(rs, 13)] = TcH + TcK; } } { E TaL, TbJ, Tip, Tiv, Tb0, Tiu, TbM, Tik, Tbk, TbW, TbG, TbQ, TbD, TbX, TbH; E TbT; { E TaD, TaK, Til, Tio; TaD = Taz + TaC; TaK = TaG + TaJ; TaL = TaD - TaK; TbJ = TaD + TaK; Til = Tc1 - Tc0; Tio = Tim + Tin; Tip = Til + Tio; Tiv = Tio - Til; } { E TaS, TbK, TaZ, TbL; { E TaO, TaR, TaV, TaY; TaO = TaM + TaN; TaR = TaP + TaQ; TaS = FNMS(KP980785280, TaR, KP195090322 * TaO); TbK = FMA(KP195090322, TaR, KP980785280 * TaO); TaV = TaT + TaU; TaY = TaW + TaX; TaZ = FMA(KP980785280, TaV, KP195090322 * TaY); TbL = FNMS(KP195090322, TaV, KP980785280 * TaY); } Tb0 = TaS + TaZ; Tiu = TbK - TbL; TbM = TbK + TbL; Tik = TaZ - TaS; } { E Tbc, TbO, Tbj, TbP; { E Tb4, Tbb, Tbf, Tbi; Tb4 = Tb2 + Tb3; Tbb = Tb7 + Tba; Tbc = Tb4 - Tbb; TbO = Tb4 + Tbb; Tbf = Tbd + Tbe; Tbi = Tbg + Tbh; Tbj = Tbf - Tbi; TbP = Tbf + Tbi; } Tbk = FMA(KP634393284, Tbc, KP773010453 * Tbj); TbW = FNMS(KP995184726, TbP, KP098017140 * TbO); TbG = FNMS(KP634393284, Tbj, KP773010453 * Tbc); TbQ = FMA(KP995184726, TbO, KP098017140 * TbP); } { E Tbv, TbR, TbC, TbS; { E Tbn, Tbu, Tby, TbB; Tbn = Tbl + Tbm; Tbu = Tbq + Tbt; Tbv = Tbn - Tbu; TbR = Tbn + Tbu; Tby = Tbw + Tbx; TbB = Tbz + TbA; TbC = Tby - TbB; TbS = Tby + TbB; } TbD = FNMS(KP773010453, TbC, KP634393284 * Tbv); TbX = FMA(KP098017140, TbR, KP995184726 * TbS); TbH = FMA(KP773010453, Tbv, KP634393284 * TbC); TbT = FNMS(KP098017140, TbS, KP995184726 * TbR); } { E Tb1, TbE, Tit, Tiw; Tb1 = TaL - Tb0; TbE = Tbk + TbD; ci[WS(rs, 22)] = Tb1 - TbE; cr[WS(rs, 9)] = Tb1 + TbE; Tit = TbD - Tbk; Tiw = Tiu + Tiv; cr[WS(rs, 57)] = Tit - Tiw; ci[WS(rs, 38)] = Tit + Tiw; } { E Tix, Tiy, TbF, TbI; Tix = TbH - TbG; Tiy = Tiv - Tiu; cr[WS(rs, 41)] = Tix - Tiy; ci[WS(rs, 54)] = Tix + Tiy; TbF = TaL + Tb0; TbI = TbG + TbH; cr[WS(rs, 25)] = TbF - TbI; ci[WS(rs, 6)] = TbF + TbI; } { E TbN, TbU, Tij, Tiq; TbN = TbJ + TbM; TbU = TbQ + TbT; ci[WS(rs, 30)] = TbN - TbU; cr[WS(rs, 1)] = TbN + TbU; Tij = TbX - TbW; Tiq = Tik + Tip; cr[WS(rs, 33)] = Tij - Tiq; ci[WS(rs, 62)] = Tij + Tiq; } { E Tir, Tis, TbV, TbY; Tir = TbT - TbQ; Tis = Tip - Tik; cr[WS(rs, 49)] = Tir - Tis; ci[WS(rs, 46)] = Tir + Tis; TbV = TbJ - TbM; TbY = TbW + TbX; cr[WS(rs, 17)] = TbV - TbY; ci[WS(rs, 14)] = TbV + TbY; } } { E T9R, Taj, Ti9, Tif, T9Y, Tie, Tam, Ti6, Ta6, Taw, Tag, Taq, Tad, Tax, Tah; E Tat; { E T9N, T9Q, Ti7, Ti8; T9N = T6b - T6m; T9Q = T9O + T9P; T9R = T9N - T9Q; Taj = T9N + T9Q; Ti7 = T6J - T6y; Ti8 = ThT - ThQ; Ti9 = Ti7 + Ti8; Tif = Ti8 - Ti7; } { E T9U, Tak, T9X, Tal; { E T9S, T9T, T9V, T9W; T9S = T6Q - T71; T9T = T77 - T7a; T9U = FNMS(KP831469612, T9T, KP555570233 * T9S); Tak = FMA(KP831469612, T9S, KP555570233 * T9T); T9V = T7h - T7s; T9W = T7y - T7B; T9X = FMA(KP555570233, T9V, KP831469612 * T9W); Tal = FNMS(KP555570233, T9W, KP831469612 * T9V); } T9Y = T9U + T9X; Tie = Tak - Tal; Tam = Tak + Tal; Ti6 = T9X - T9U; } { E Ta2, Tao, Ta5, Tap; { E Ta0, Ta1, Ta3, Ta4; Ta0 = T8p - T8s; Ta1 = T87 - T8i; Ta2 = Ta0 - Ta1; Tao = Ta0 + Ta1; Ta3 = T7K - T7V; Ta4 = T8v - T8u; Ta5 = Ta3 - Ta4; Tap = Ta3 + Ta4; } Ta6 = FMA(KP471396736, Ta2, KP881921264 * Ta5); Taw = FNMS(KP956940335, Tap, KP290284677 * Tao); Tag = FNMS(KP471396736, Ta5, KP881921264 * Ta2); Taq = FMA(KP956940335, Tao, KP290284677 * Tap); } { E Ta9, Tar, Tac, Tas; { E Ta7, Ta8, Taa, Tab; Ta7 = T8D - T8O; Ta8 = T9n - T9o; Ta9 = Ta7 - Ta8; Tar = Ta7 + Ta8; Taa = T9i - T9l; Tab = T9b - T90; Tac = Taa - Tab; Tas = Taa + Tab; } Tad = FNMS(KP881921264, Tac, KP471396736 * Ta9); Tax = FMA(KP290284677, Tar, KP956940335 * Tas); Tah = FMA(KP881921264, Ta9, KP471396736 * Tac); Tat = FNMS(KP290284677, Tas, KP956940335 * Tar); } { E T9Z, Tae, Tid, Tig; T9Z = T9R - T9Y; Tae = Ta6 + Tad; ci[WS(rs, 20)] = T9Z - Tae; cr[WS(rs, 11)] = T9Z + Tae; Tid = Tad - Ta6; Tig = Tie + Tif; cr[WS(rs, 59)] = Tid - Tig; ci[WS(rs, 36)] = Tid + Tig; } { E Tih, Tii, Taf, Tai; Tih = Tah - Tag; Tii = Tif - Tie; cr[WS(rs, 43)] = Tih - Tii; ci[WS(rs, 52)] = Tih + Tii; Taf = T9R + T9Y; Tai = Tag + Tah; cr[WS(rs, 27)] = Taf - Tai; ci[WS(rs, 4)] = Taf + Tai; } { E Tan, Tau, Ti5, Tia; Tan = Taj + Tam; Tau = Taq + Tat; ci[WS(rs, 28)] = Tan - Tau; cr[WS(rs, 3)] = Tan + Tau; Ti5 = Tax - Taw; Tia = Ti6 + Ti9; cr[WS(rs, 35)] = Ti5 - Tia; ci[WS(rs, 60)] = Ti5 + Tia; } { E Tib, Tic, Tav, Tay; Tib = Tat - Taq; Tic = Ti9 - Ti6; cr[WS(rs, 51)] = Tib - Tic; ci[WS(rs, 44)] = Tib + Tic; Tav = Taj - Tam; Tay = Taw + Tax; cr[WS(rs, 19)] = Tav - Tay; ci[WS(rs, 12)] = Tav + Tay; } } } } } static const tw_instr twinstr[] = { {TW_FULL, 1, 64}, {TW_NEXT, 1, 0} }; static const hc2hc_desc desc = { 64, "hf_64", twinstr, &GENUS, {808, 270, 230, 0} }; void X(codelet_hf_64) (planner *p) { X(khc2hc_register) (p, hf_64, &desc); } #endif /* HAVE_FMA */ fftw-3.3.3/rdft/scalar/r2cf/hf_20.c0000644000175000017500000006673012054410641013522 00000000000000/* * Copyright (c) 2003, 2007-11 Matteo Frigo * Copyright (c) 2003, 2007-11 Massachusetts Institute of Technology * * This program is free software; you can redistribute it and/or modify * it under the terms of the GNU General Public License as published by * the Free Software Foundation; either version 2 of the License, or * (at your option) any later version. * * This program is distributed in the hope that it will be useful, * but WITHOUT ANY WARRANTY; without even the implied warranty of * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the * GNU General Public License for more details. * * You should have received a copy of the GNU General Public License * along with this program; if not, write to the Free Software * Foundation, Inc., 51 Franklin Street, Fifth Floor, Boston, MA 02110-1301 USA * */ /* This file was automatically generated --- DO NOT EDIT */ /* Generated on Sun Nov 25 07:40:00 EST 2012 */ #include "codelet-rdft.h" #ifdef HAVE_FMA /* Generated by: ../../../genfft/gen_hc2hc.native -fma -reorder-insns -schedule-for-pipeline -compact -variables 4 -pipeline-latency 4 -n 20 -dit -name hf_20 -include hf.h */ /* * This function contains 246 FP additions, 148 FP multiplications, * (or, 136 additions, 38 multiplications, 110 fused multiply/add), * 100 stack variables, 4 constants, and 80 memory accesses */ #include "hf.h" static void hf_20(R *cr, R *ci, const R *W, stride rs, INT mb, INT me, INT ms) { DK(KP951056516, +0.951056516295153572116439333379382143405698634); DK(KP559016994, +0.559016994374947424102293417182819058860154590); DK(KP250000000, +0.250000000000000000000000000000000000000000000); DK(KP618033988, +0.618033988749894848204586834365638117720309180); { INT m; for (m = mb, W = W + ((mb - 1) * 38); m < me; m = m + 1, cr = cr + ms, ci = ci - ms, W = W + 38, MAKE_VOLATILE_STRIDE(40, rs)) { E T54, T5a, T5c, T56, T53, T55, T5b, T57; { E T4N, T4q, T8, T2i, T4r, T2n, T4O, Tl, T2v, T3v, T43, T4b, TN, T2b, T3F; E T3a, T2R, T3z, T3T, T4f, T27, T2f, T3J, T3i, T2K, T3y, T3W, T4e, T1G, T2e; E T3I, T3p, T2C, T3w, T40, T4c, T1e, T2c, T3G, T33; { E T1, T4p, T3, T6, T2, T5; T1 = cr[0]; T4p = ci[0]; T3 = cr[WS(rs, 10)]; T6 = ci[WS(rs, 10)]; T2 = W[18]; T5 = W[19]; { E Ta, Td, Tg, T2j, Tb, Tj, Tf, Tc, Ti; { E T4n, T4, T9, T4o, T7; Ta = cr[WS(rs, 5)]; Td = ci[WS(rs, 5)]; T4n = T2 * T6; T4 = T2 * T3; T9 = W[8]; Tg = cr[WS(rs, 15)]; T4o = FNMS(T5, T3, T4n); T7 = FMA(T5, T6, T4); T2j = T9 * Td; Tb = T9 * Ta; T4N = T4p - T4o; T4q = T4o + T4p; T8 = T1 + T7; T2i = T1 - T7; Tj = ci[WS(rs, 15)]; Tf = W[28]; } Tc = W[9]; Ti = W[29]; { E T36, Ts, T2t, TL, TB, TE, TD, T38, Ty, T2q, TC; { E TH, TK, TJ, T2s, TI; { E To, Tr, Tp, T35, Tq, TG; { E T2k, Te, T2m, Tk, T2l, Th, Tn; To = cr[WS(rs, 4)]; T2l = Tf * Tj; Th = Tf * Tg; T2k = FNMS(Tc, Ta, T2j); Te = FMA(Tc, Td, Tb); T2m = FNMS(Ti, Tg, T2l); Tk = FMA(Ti, Tj, Th); Tr = ci[WS(rs, 4)]; Tn = W[6]; T4r = T2k + T2m; T2n = T2k - T2m; T4O = Te - Tk; Tl = Te + Tk; Tp = Tn * To; T35 = Tn * Tr; } Tq = W[7]; TH = cr[WS(rs, 19)]; TK = ci[WS(rs, 19)]; TG = W[36]; T36 = FNMS(Tq, To, T35); Ts = FMA(Tq, Tr, Tp); TJ = W[37]; T2s = TG * TK; TI = TG * TH; } { E Tu, Tx, Tt, Tw, T37, Tv, TA; Tu = cr[WS(rs, 14)]; Tx = ci[WS(rs, 14)]; T2t = FNMS(TJ, TH, T2s); TL = FMA(TJ, TK, TI); Tt = W[26]; Tw = W[27]; TB = cr[WS(rs, 9)]; TE = ci[WS(rs, 9)]; T37 = Tt * Tx; Tv = Tt * Tu; TA = W[16]; TD = W[17]; T38 = FNMS(Tw, Tu, T37); Ty = FMA(Tw, Tx, Tv); T2q = TA * TE; TC = TA * TB; } } { E T39, T42, Tz, T2p, T2r, TF; T39 = T36 - T38; T42 = T36 + T38; Tz = Ts + Ty; T2p = Ts - Ty; T2r = FNMS(TD, TB, T2q); TF = FMA(TD, TE, TC); { E T41, T2u, TM, T34; T41 = T2r + T2t; T2u = T2r - T2t; TM = TF + TL; T34 = TL - TF; T2v = T2p - T2u; T3v = T2p + T2u; T43 = T41 - T42; T4b = T42 + T41; TN = Tz - TM; T2b = Tz + TM; T3F = T39 + T34; T3a = T34 - T39; } } } } } { E T3e, T1M, T2P, T25, T1V, T1Y, T1X, T3g, T1S, T2M, T1W; { E T21, T24, T23, T2O, T22; { E T1I, T1L, T1H, T1K, T3d, T1J, T20; T1I = cr[WS(rs, 12)]; T1L = ci[WS(rs, 12)]; T1H = W[22]; T1K = W[23]; T21 = cr[WS(rs, 7)]; T24 = ci[WS(rs, 7)]; T3d = T1H * T1L; T1J = T1H * T1I; T20 = W[12]; T23 = W[13]; T3e = FNMS(T1K, T1I, T3d); T1M = FMA(T1K, T1L, T1J); T2O = T20 * T24; T22 = T20 * T21; } { E T1O, T1R, T1N, T1Q, T3f, T1P, T1U; T1O = cr[WS(rs, 2)]; T1R = ci[WS(rs, 2)]; T2P = FNMS(T23, T21, T2O); T25 = FMA(T23, T24, T22); T1N = W[2]; T1Q = W[3]; T1V = cr[WS(rs, 17)]; T1Y = ci[WS(rs, 17)]; T3f = T1N * T1R; T1P = T1N * T1O; T1U = W[32]; T1X = W[33]; T3g = FNMS(T1Q, T1O, T3f); T1S = FMA(T1Q, T1R, T1P); T2M = T1U * T1Y; T1W = T1U * T1V; } } { E T3h, T3S, T1T, T2L, T2N, T1Z; T3h = T3e - T3g; T3S = T3e + T3g; T1T = T1M + T1S; T2L = T1M - T1S; T2N = FNMS(T1X, T1V, T2M); T1Z = FMA(T1X, T1Y, T1W); { E T3R, T2Q, T26, T3c; T3R = T2N + T2P; T2Q = T2N - T2P; T26 = T1Z + T25; T3c = T25 - T1Z; T2R = T2L - T2Q; T3z = T2L + T2Q; T3T = T3R - T3S; T4f = T3S + T3R; T27 = T1T - T26; T2f = T1T + T26; T3J = T3h + T3c; T3i = T3c - T3h; } } } { E T3l, T1l, T2I, T1E, T1u, T1x, T1w, T3n, T1r, T2F, T1v; { E T1A, T1D, T1C, T2H, T1B; { E T1h, T1k, T1g, T1j, T3k, T1i, T1z; T1h = cr[WS(rs, 8)]; T1k = ci[WS(rs, 8)]; T1g = W[14]; T1j = W[15]; T1A = cr[WS(rs, 3)]; T1D = ci[WS(rs, 3)]; T3k = T1g * T1k; T1i = T1g * T1h; T1z = W[4]; T1C = W[5]; T3l = FNMS(T1j, T1h, T3k); T1l = FMA(T1j, T1k, T1i); T2H = T1z * T1D; T1B = T1z * T1A; } { E T1n, T1q, T1m, T1p, T3m, T1o, T1t; T1n = cr[WS(rs, 18)]; T1q = ci[WS(rs, 18)]; T2I = FNMS(T1C, T1A, T2H); T1E = FMA(T1C, T1D, T1B); T1m = W[34]; T1p = W[35]; T1u = cr[WS(rs, 13)]; T1x = ci[WS(rs, 13)]; T3m = T1m * T1q; T1o = T1m * T1n; T1t = W[24]; T1w = W[25]; T3n = FNMS(T1p, T1n, T3m); T1r = FMA(T1p, T1q, T1o); T2F = T1t * T1x; T1v = T1t * T1u; } } { E T3o, T3V, T1s, T2E, T2G, T1y; T3o = T3l - T3n; T3V = T3l + T3n; T1s = T1l + T1r; T2E = T1l - T1r; T2G = FNMS(T1w, T1u, T2F); T1y = FMA(T1w, T1x, T1v); { E T3U, T2J, T1F, T3j; T3U = T2G + T2I; T2J = T2G - T2I; T1F = T1y + T1E; T3j = T1E - T1y; T2K = T2E - T2J; T3y = T2E + T2J; T3W = T3U - T3V; T4e = T3V + T3U; T1G = T1s - T1F; T2e = T1s + T1F; T3I = T3o + T3j; T3p = T3j - T3o; } } } { E T2Z, TT, T2A, T1c, T12, T15, T14, T31, TZ, T2x, T13; { E T18, T1b, T1a, T2z, T19; { E TP, TS, TO, TR, T2Y, TQ, T17; TP = cr[WS(rs, 16)]; TS = ci[WS(rs, 16)]; TO = W[30]; TR = W[31]; T18 = cr[WS(rs, 11)]; T1b = ci[WS(rs, 11)]; T2Y = TO * TS; TQ = TO * TP; T17 = W[20]; T1a = W[21]; T2Z = FNMS(TR, TP, T2Y); TT = FMA(TR, TS, TQ); T2z = T17 * T1b; T19 = T17 * T18; } { E TV, TY, TU, TX, T30, TW, T11; TV = cr[WS(rs, 6)]; TY = ci[WS(rs, 6)]; T2A = FNMS(T1a, T18, T2z); T1c = FMA(T1a, T1b, T19); TU = W[10]; TX = W[11]; T12 = cr[WS(rs, 1)]; T15 = ci[WS(rs, 1)]; T30 = TU * TY; TW = TU * TV; T11 = W[0]; T14 = W[1]; T31 = FNMS(TX, TV, T30); TZ = FMA(TX, TY, TW); T2x = T11 * T15; T13 = T11 * T12; } } { E T32, T3Z, T10, T2w, T2y, T16; T32 = T2Z - T31; T3Z = T2Z + T31; T10 = TT + TZ; T2w = TT - TZ; T2y = FNMS(T14, T12, T2x); T16 = FMA(T14, T15, T13); { E T3Y, T2B, T1d, T2X; T3Y = T2y + T2A; T2B = T2y - T2A; T1d = T16 + T1c; T2X = T1c - T16; T2C = T2w - T2B; T3w = T2w + T2B; T40 = T3Y - T3Z; T4c = T3Z + T3Y; T1e = T10 - T1d; T2c = T10 + T1d; T3G = T32 + T2X; T33 = T2X - T32; } } } { E T4l, T4k, T4w, T4x, T4Q, T4R, T2o, T4X, T4W, T4C, T4D, T4J, T4h, T4j, T4I; E T51, T52, T49, T3r, T3t, T58, T2D, T48, T2S, T59; { E T2a, T47, T45, T3u, T3x, T3N, T3L, T3A, T46, T3Q; { E Tm, T1f, T28, T3X, T44; T4l = T3W + T3T; T3X = T3T - T3W; T44 = T40 - T43; T4k = T43 + T40; T2a = T8 + Tl; Tm = T8 - Tl; T1f = TN + T1e; T4w = T1e - TN; T4x = T1G - T27; T28 = T1G + T27; T47 = FMA(KP618033988, T3X, T44); T45 = FNMS(KP618033988, T44, T3X); { E T3H, T29, T3P, T3K, T3O; T3H = T3F - T3G; T4Q = T3F + T3G; T29 = T1f + T28; T3P = T1f - T28; T4R = T3I + T3J; T3K = T3I - T3J; T3u = T2i + T2n; T2o = T2i - T2n; T4X = T3v - T3w; T3x = T3v + T3w; ci[WS(rs, 9)] = Tm + T29; T3O = FNMS(KP250000000, T29, Tm); T3N = FNMS(KP618033988, T3H, T3K); T3L = FMA(KP618033988, T3K, T3H); T3A = T3y + T3z; T4W = T3y - T3z; T46 = FMA(KP559016994, T3P, T3O); T3Q = FNMS(KP559016994, T3P, T3O); } } { E T2d, T2g, T3b, T3q, T2h; { E T4d, T3D, T3C, T4g, T3B, T3M, T3E; T4C = T4b + T4c; T4d = T4b - T4c; T3D = T3x - T3A; T3B = T3x + T3A; ci[WS(rs, 1)] = FMA(KP951056516, T45, T3Q); cr[WS(rs, 2)] = FNMS(KP951056516, T45, T3Q); cr[WS(rs, 6)] = FMA(KP951056516, T47, T46); ci[WS(rs, 5)] = FNMS(KP951056516, T47, T46); cr[WS(rs, 5)] = T3u + T3B; T3C = FNMS(KP250000000, T3B, T3u); T4g = T4e - T4f; T4D = T4e + T4f; T2d = T2b + T2c; T4J = T2b - T2c; T3M = FNMS(KP559016994, T3D, T3C); T3E = FMA(KP559016994, T3D, T3C); T4h = FMA(KP618033988, T4g, T4d); T4j = FNMS(KP618033988, T4d, T4g); cr[WS(rs, 9)] = FNMS(KP951056516, T3L, T3E); cr[WS(rs, 1)] = FMA(KP951056516, T3L, T3E); ci[WS(rs, 6)] = FMA(KP951056516, T3N, T3M); ci[WS(rs, 2)] = FNMS(KP951056516, T3N, T3M); T4I = T2f - T2e; T2g = T2e + T2f; } T3b = T33 - T3a; T51 = T3a + T33; T52 = T3p + T3i; T3q = T3i - T3p; T2h = T2d + T2g; T49 = T2d - T2g; T3r = FMA(KP618033988, T3q, T3b); T3t = FNMS(KP618033988, T3b, T3q); T58 = T2v - T2C; T2D = T2v + T2C; cr[0] = T2a + T2h; T48 = FNMS(KP250000000, T2h, T2a); T2S = T2K + T2R; T59 = T2K - T2R; } } { E T4B, T4P, T4Y, T50, T4U, T4S; { E T4A, T4y, T4s, T4m, T4u, T4t, T4z, T4v; { E T2V, T2U, T4i, T4a, T2T, T2W, T3s; T4i = FNMS(KP559016994, T49, T48); T4a = FMA(KP559016994, T49, T48); T2T = T2D + T2S; T2V = T2D - T2S; ci[WS(rs, 3)] = FMA(KP951056516, T4h, T4a); cr[WS(rs, 4)] = FNMS(KP951056516, T4h, T4a); cr[WS(rs, 8)] = FMA(KP951056516, T4j, T4i); ci[WS(rs, 7)] = FNMS(KP951056516, T4j, T4i); ci[WS(rs, 4)] = T2o + T2T; T2U = FNMS(KP250000000, T2T, T2o); T4A = FMA(KP618033988, T4w, T4x); T4y = FNMS(KP618033988, T4x, T4w); T4B = T4r + T4q; T4s = T4q - T4r; T2W = FMA(KP559016994, T2V, T2U); T3s = FNMS(KP559016994, T2V, T2U); ci[WS(rs, 8)] = FMA(KP951056516, T3r, T2W); ci[0] = FNMS(KP951056516, T3r, T2W); cr[WS(rs, 7)] = FNMS(KP951056516, T3t, T3s); cr[WS(rs, 3)] = FMA(KP951056516, T3t, T3s); T4m = T4k + T4l; T4u = T4l - T4k; } cr[WS(rs, 10)] = T4m - T4s; T4t = FMA(KP250000000, T4m, T4s); T4P = T4N - T4O; T54 = T4O + T4N; T4Y = FNMS(KP618033988, T4X, T4W); T50 = FMA(KP618033988, T4W, T4X); T4z = FNMS(KP559016994, T4u, T4t); T4v = FMA(KP559016994, T4u, T4t); ci[WS(rs, 13)] = FMA(KP951056516, T4y, T4v); cr[WS(rs, 14)] = FMS(KP951056516, T4y, T4v); ci[WS(rs, 17)] = FMA(KP951056516, T4A, T4z); cr[WS(rs, 18)] = FMS(KP951056516, T4A, T4z); T4U = T4Q - T4R; T4S = T4Q + T4R; } { E T4M, T4K, T4E, T4G, T4T, T4V, T4Z, T4F, T4L, T4H; ci[WS(rs, 14)] = T4S + T4P; T4T = FNMS(KP250000000, T4S, T4P); T4M = FNMS(KP618033988, T4I, T4J); T4K = FMA(KP618033988, T4J, T4I); T4V = FNMS(KP559016994, T4U, T4T); T4Z = FMA(KP559016994, T4U, T4T); cr[WS(rs, 17)] = -(FMA(KP951056516, T4Y, T4V)); cr[WS(rs, 13)] = FMS(KP951056516, T4Y, T4V); ci[WS(rs, 18)] = FNMS(KP951056516, T50, T4Z); ci[WS(rs, 10)] = FMA(KP951056516, T50, T4Z); T4E = T4C + T4D; T4G = T4C - T4D; ci[WS(rs, 19)] = T4E + T4B; T4F = FNMS(KP250000000, T4E, T4B); T5a = FMA(KP618033988, T59, T58); T5c = FNMS(KP618033988, T58, T59); T4L = FMA(KP559016994, T4G, T4F); T4H = FNMS(KP559016994, T4G, T4F); ci[WS(rs, 11)] = FMA(KP951056516, T4K, T4H); cr[WS(rs, 12)] = FMS(KP951056516, T4K, T4H); ci[WS(rs, 15)] = FMA(KP951056516, T4M, T4L); cr[WS(rs, 16)] = FMS(KP951056516, T4M, T4L); T56 = T51 - T52; T53 = T51 + T52; } } } } cr[WS(rs, 15)] = T53 - T54; T55 = FMA(KP250000000, T53, T54); T5b = FMA(KP559016994, T56, T55); T57 = FNMS(KP559016994, T56, T55); cr[WS(rs, 19)] = -(FMA(KP951056516, T5a, T57)); cr[WS(rs, 11)] = FMS(KP951056516, T5a, T57); ci[WS(rs, 16)] = FNMS(KP951056516, T5c, T5b); ci[WS(rs, 12)] = FMA(KP951056516, T5c, T5b); } } } static const tw_instr twinstr[] = { {TW_FULL, 1, 20}, {TW_NEXT, 1, 0} }; static const hc2hc_desc desc = { 20, "hf_20", twinstr, &GENUS, {136, 38, 110, 0} }; void X(codelet_hf_20) (planner *p) { X(khc2hc_register) (p, hf_20, &desc); } #else /* HAVE_FMA */ /* Generated by: ../../../genfft/gen_hc2hc.native -compact -variables 4 -pipeline-latency 4 -n 20 -dit -name hf_20 -include hf.h */ /* * This function contains 246 FP additions, 124 FP multiplications, * (or, 184 additions, 62 multiplications, 62 fused multiply/add), * 85 stack variables, 4 constants, and 80 memory accesses */ #include "hf.h" static void hf_20(R *cr, R *ci, const R *W, stride rs, INT mb, INT me, INT ms) { DK(KP587785252, +0.587785252292473129168705954639072768597652438); DK(KP951056516, +0.951056516295153572116439333379382143405698634); DK(KP250000000, +0.250000000000000000000000000000000000000000000); DK(KP559016994, +0.559016994374947424102293417182819058860154590); { INT m; for (m = mb, W = W + ((mb - 1) * 38); m < me; m = m + 1, cr = cr + ms, ci = ci - ms, W = W + 38, MAKE_VOLATILE_STRIDE(40, rs)) { E Tj, T1R, T4j, T4s, T2q, T37, T3Q, T42, T1r, T1O, T1P, T3i, T3l, T3J, T3D; E T3E, T44, T1V, T1W, T1X, T2e, T2j, T2k, T2W, T2X, T4f, T33, T34, T35, T2J; E T2O, T4q, TG, T13, T14, T3p, T3s, T3K, T3A, T3B, T43, T1S, T1T, T1U, T23; E T28, T29, T2T, T2U, T4e, T30, T31, T32, T2y, T2D, T4p; { E T1, T3N, T6, T3M, Tc, T2n, Th, T2o; T1 = cr[0]; T3N = ci[0]; { E T3, T5, T2, T4; T3 = cr[WS(rs, 10)]; T5 = ci[WS(rs, 10)]; T2 = W[18]; T4 = W[19]; T6 = FMA(T2, T3, T4 * T5); T3M = FNMS(T4, T3, T2 * T5); } { E T9, Tb, T8, Ta; T9 = cr[WS(rs, 5)]; Tb = ci[WS(rs, 5)]; T8 = W[8]; Ta = W[9]; Tc = FMA(T8, T9, Ta * Tb); T2n = FNMS(Ta, T9, T8 * Tb); } { E Te, Tg, Td, Tf; Te = cr[WS(rs, 15)]; Tg = ci[WS(rs, 15)]; Td = W[28]; Tf = W[29]; Th = FMA(Td, Te, Tf * Tg); T2o = FNMS(Tf, Te, Td * Tg); } { E T7, Ti, T4h, T4i; T7 = T1 + T6; Ti = Tc + Th; Tj = T7 - Ti; T1R = T7 + Ti; T4h = T3N - T3M; T4i = Tc - Th; T4j = T4h - T4i; T4s = T4i + T4h; } { E T2m, T2p, T3O, T3P; T2m = T1 - T6; T2p = T2n - T2o; T2q = T2m - T2p; T37 = T2m + T2p; T3O = T3M + T3N; T3P = T2n + T2o; T3Q = T3O - T3P; T42 = T3P + T3O; } } { E T1f, T3g, T2a, T2H, T1N, T3j, T2i, T2N, T1q, T3h, T2d, T2I, T1C, T3k, T2f; E T2M; { E T19, T2F, T1e, T2G; { E T16, T18, T15, T17; T16 = cr[WS(rs, 8)]; T18 = ci[WS(rs, 8)]; T15 = W[14]; T17 = W[15]; T19 = FMA(T15, T16, T17 * T18); T2F = FNMS(T17, T16, T15 * T18); } { E T1b, T1d, T1a, T1c; T1b = cr[WS(rs, 18)]; T1d = ci[WS(rs, 18)]; T1a = W[34]; T1c = W[35]; T1e = FMA(T1a, T1b, T1c * T1d); T2G = FNMS(T1c, T1b, T1a * T1d); } T1f = T19 + T1e; T3g = T2F + T2G; T2a = T19 - T1e; T2H = T2F - T2G; } { E T1H, T2g, T1M, T2h; { E T1E, T1G, T1D, T1F; T1E = cr[WS(rs, 17)]; T1G = ci[WS(rs, 17)]; T1D = W[32]; T1F = W[33]; T1H = FMA(T1D, T1E, T1F * T1G); T2g = FNMS(T1F, T1E, T1D * T1G); } { E T1J, T1L, T1I, T1K; T1J = cr[WS(rs, 7)]; T1L = ci[WS(rs, 7)]; T1I = W[12]; T1K = W[13]; T1M = FMA(T1I, T1J, T1K * T1L); T2h = FNMS(T1K, T1J, T1I * T1L); } T1N = T1H + T1M; T3j = T2g + T2h; T2i = T2g - T2h; T2N = T1H - T1M; } { E T1k, T2b, T1p, T2c; { E T1h, T1j, T1g, T1i; T1h = cr[WS(rs, 13)]; T1j = ci[WS(rs, 13)]; T1g = W[24]; T1i = W[25]; T1k = FMA(T1g, T1h, T1i * T1j); T2b = FNMS(T1i, T1h, T1g * T1j); } { E T1m, T1o, T1l, T1n; T1m = cr[WS(rs, 3)]; T1o = ci[WS(rs, 3)]; T1l = W[4]; T1n = W[5]; T1p = FMA(T1l, T1m, T1n * T1o); T2c = FNMS(T1n, T1m, T1l * T1o); } T1q = T1k + T1p; T3h = T2b + T2c; T2d = T2b - T2c; T2I = T1k - T1p; } { E T1w, T2K, T1B, T2L; { E T1t, T1v, T1s, T1u; T1t = cr[WS(rs, 12)]; T1v = ci[WS(rs, 12)]; T1s = W[22]; T1u = W[23]; T1w = FMA(T1s, T1t, T1u * T1v); T2K = FNMS(T1u, T1t, T1s * T1v); } { E T1y, T1A, T1x, T1z; T1y = cr[WS(rs, 2)]; T1A = ci[WS(rs, 2)]; T1x = W[2]; T1z = W[3]; T1B = FMA(T1x, T1y, T1z * T1A); T2L = FNMS(T1z, T1y, T1x * T1A); } T1C = T1w + T1B; T3k = T2K + T2L; T2f = T1w - T1B; T2M = T2K - T2L; } T1r = T1f - T1q; T1O = T1C - T1N; T1P = T1r + T1O; T3i = T3g - T3h; T3l = T3j - T3k; T3J = T3l - T3i; T3D = T3g + T3h; T3E = T3k + T3j; T44 = T3D + T3E; T1V = T1f + T1q; T1W = T1C + T1N; T1X = T1V + T1W; T2e = T2a - T2d; T2j = T2f - T2i; T2k = T2e + T2j; T2W = T2H - T2I; T2X = T2M - T2N; T4f = T2W + T2X; T33 = T2a + T2d; T34 = T2f + T2i; T35 = T33 + T34; T2J = T2H + T2I; T2O = T2M + T2N; T4q = T2J + T2O; } { E Tu, T3n, T1Z, T2w, T12, T3r, T27, T2z, TF, T3o, T22, T2x, TR, T3q, T24; E T2C; { E To, T2u, Tt, T2v; { E Tl, Tn, Tk, Tm; Tl = cr[WS(rs, 4)]; Tn = ci[WS(rs, 4)]; Tk = W[6]; Tm = W[7]; To = FMA(Tk, Tl, Tm * Tn); T2u = FNMS(Tm, Tl, Tk * Tn); } { E Tq, Ts, Tp, Tr; Tq = cr[WS(rs, 14)]; Ts = ci[WS(rs, 14)]; Tp = W[26]; Tr = W[27]; Tt = FMA(Tp, Tq, Tr * Ts); T2v = FNMS(Tr, Tq, Tp * Ts); } Tu = To + Tt; T3n = T2u + T2v; T1Z = To - Tt; T2w = T2u - T2v; } { E TW, T25, T11, T26; { E TT, TV, TS, TU; TT = cr[WS(rs, 1)]; TV = ci[WS(rs, 1)]; TS = W[0]; TU = W[1]; TW = FMA(TS, TT, TU * TV); T25 = FNMS(TU, TT, TS * TV); } { E TY, T10, TX, TZ; TY = cr[WS(rs, 11)]; T10 = ci[WS(rs, 11)]; TX = W[20]; TZ = W[21]; T11 = FMA(TX, TY, TZ * T10); T26 = FNMS(TZ, TY, TX * T10); } T12 = TW + T11; T3r = T25 + T26; T27 = T25 - T26; T2z = T11 - TW; } { E Tz, T20, TE, T21; { E Tw, Ty, Tv, Tx; Tw = cr[WS(rs, 9)]; Ty = ci[WS(rs, 9)]; Tv = W[16]; Tx = W[17]; Tz = FMA(Tv, Tw, Tx * Ty); T20 = FNMS(Tx, Tw, Tv * Ty); } { E TB, TD, TA, TC; TB = cr[WS(rs, 19)]; TD = ci[WS(rs, 19)]; TA = W[36]; TC = W[37]; TE = FMA(TA, TB, TC * TD); T21 = FNMS(TC, TB, TA * TD); } TF = Tz + TE; T3o = T20 + T21; T22 = T20 - T21; T2x = Tz - TE; } { E TL, T2A, TQ, T2B; { E TI, TK, TH, TJ; TI = cr[WS(rs, 16)]; TK = ci[WS(rs, 16)]; TH = W[30]; TJ = W[31]; TL = FMA(TH, TI, TJ * TK); T2A = FNMS(TJ, TI, TH * TK); } { E TN, TP, TM, TO; TN = cr[WS(rs, 6)]; TP = ci[WS(rs, 6)]; TM = W[10]; TO = W[11]; TQ = FMA(TM, TN, TO * TP); T2B = FNMS(TO, TN, TM * TP); } TR = TL + TQ; T3q = T2A + T2B; T24 = TL - TQ; T2C = T2A - T2B; } TG = Tu - TF; T13 = TR - T12; T14 = TG + T13; T3p = T3n - T3o; T3s = T3q - T3r; T3K = T3p + T3s; T3A = T3n + T3o; T3B = T3q + T3r; T43 = T3A + T3B; T1S = Tu + TF; T1T = TR + T12; T1U = T1S + T1T; T23 = T1Z - T22; T28 = T24 - T27; T29 = T23 + T28; T2T = T2w - T2x; T2U = T2C + T2z; T4e = T2T + T2U; T30 = T1Z + T22; T31 = T24 + T27; T32 = T30 + T31; T2y = T2w + T2x; T2D = T2z - T2C; T4p = T2D - T2y; } { E T3e, T1Q, T3d, T3u, T3w, T3m, T3t, T3v, T3f; T3e = KP559016994 * (T14 - T1P); T1Q = T14 + T1P; T3d = FNMS(KP250000000, T1Q, Tj); T3m = T3i + T3l; T3t = T3p - T3s; T3u = FNMS(KP587785252, T3t, KP951056516 * T3m); T3w = FMA(KP951056516, T3t, KP587785252 * T3m); ci[WS(rs, 9)] = Tj + T1Q; T3v = T3e + T3d; ci[WS(rs, 5)] = T3v - T3w; cr[WS(rs, 6)] = T3v + T3w; T3f = T3d - T3e; cr[WS(rs, 2)] = T3f - T3u; ci[WS(rs, 1)] = T3f + T3u; } { E T36, T38, T39, T2Z, T3c, T2V, T2Y, T3b, T3a; T36 = KP559016994 * (T32 - T35); T38 = T32 + T35; T39 = FNMS(KP250000000, T38, T37); T2V = T2T - T2U; T2Y = T2W - T2X; T2Z = FMA(KP951056516, T2V, KP587785252 * T2Y); T3c = FNMS(KP587785252, T2V, KP951056516 * T2Y); cr[WS(rs, 5)] = T37 + T38; T3b = T39 - T36; ci[WS(rs, 2)] = T3b - T3c; ci[WS(rs, 6)] = T3c + T3b; T3a = T36 + T39; cr[WS(rs, 1)] = T2Z + T3a; cr[WS(rs, 9)] = T3a - T2Z; } { E T3x, T1Y, T3y, T3G, T3I, T3C, T3F, T3H, T3z; T3x = KP559016994 * (T1U - T1X); T1Y = T1U + T1X; T3y = FNMS(KP250000000, T1Y, T1R); T3C = T3A - T3B; T3F = T3D - T3E; T3G = FMA(KP951056516, T3C, KP587785252 * T3F); T3I = FNMS(KP587785252, T3C, KP951056516 * T3F); cr[0] = T1R + T1Y; T3H = T3y - T3x; ci[WS(rs, 7)] = T3H - T3I; cr[WS(rs, 8)] = T3H + T3I; T3z = T3x + T3y; cr[WS(rs, 4)] = T3z - T3G; ci[WS(rs, 3)] = T3z + T3G; } { E T2l, T2r, T2s, T2Q, T2R, T2E, T2P, T2S, T2t; T2l = KP559016994 * (T29 - T2k); T2r = T29 + T2k; T2s = FNMS(KP250000000, T2r, T2q); T2E = T2y + T2D; T2P = T2J - T2O; T2Q = FMA(KP951056516, T2E, KP587785252 * T2P); T2R = FNMS(KP587785252, T2E, KP951056516 * T2P); ci[WS(rs, 4)] = T2q + T2r; T2S = T2s - T2l; cr[WS(rs, 3)] = T2R + T2S; cr[WS(rs, 7)] = T2S - T2R; T2t = T2l + T2s; ci[0] = T2t - T2Q; ci[WS(rs, 8)] = T2Q + T2t; } { E T3U, T3L, T3V, T3T, T3X, T3R, T3S, T3Y, T3W; T3U = KP559016994 * (T3K + T3J); T3L = T3J - T3K; T3V = FMA(KP250000000, T3L, T3Q); T3R = T13 - TG; T3S = T1r - T1O; T3T = FNMS(KP587785252, T3S, KP951056516 * T3R); T3X = FMA(KP587785252, T3R, KP951056516 * T3S); cr[WS(rs, 10)] = T3L - T3Q; T3Y = T3V - T3U; cr[WS(rs, 18)] = T3X - T3Y; ci[WS(rs, 17)] = T3X + T3Y; T3W = T3U + T3V; cr[WS(rs, 14)] = T3T - T3W; ci[WS(rs, 13)] = T3T + T3W; } { E T4g, T4k, T4l, T4d, T4n, T4b, T4c, T4o, T4m; T4g = KP559016994 * (T4e - T4f); T4k = T4e + T4f; T4l = FNMS(KP250000000, T4k, T4j); T4b = T33 - T34; T4c = T30 - T31; T4d = FNMS(KP587785252, T4c, KP951056516 * T4b); T4n = FMA(KP951056516, T4c, KP587785252 * T4b); ci[WS(rs, 14)] = T4k + T4j; T4o = T4g + T4l; ci[WS(rs, 10)] = T4n + T4o; ci[WS(rs, 18)] = T4o - T4n; T4m = T4g - T4l; cr[WS(rs, 13)] = T4d + T4m; cr[WS(rs, 17)] = T4m - T4d; } { E T47, T45, T46, T41, T49, T3Z, T40, T4a, T48; T47 = KP559016994 * (T43 - T44); T45 = T43 + T44; T46 = FNMS(KP250000000, T45, T42); T3Z = T1S - T1T; T40 = T1V - T1W; T41 = FNMS(KP951056516, T40, KP587785252 * T3Z); T49 = FMA(KP951056516, T3Z, KP587785252 * T40); ci[WS(rs, 19)] = T45 + T42; T4a = T47 + T46; cr[WS(rs, 16)] = T49 - T4a; ci[WS(rs, 15)] = T49 + T4a; T48 = T46 - T47; cr[WS(rs, 12)] = T41 - T48; ci[WS(rs, 11)] = T41 + T48; } { E T4w, T4r, T4x, T4v, T4z, T4t, T4u, T4A, T4y; T4w = KP559016994 * (T4p + T4q); T4r = T4p - T4q; T4x = FMA(KP250000000, T4r, T4s); T4t = T23 - T28; T4u = T2e - T2j; T4v = FMA(KP951056516, T4t, KP587785252 * T4u); T4z = FNMS(KP587785252, T4t, KP951056516 * T4u); cr[WS(rs, 15)] = T4r - T4s; T4A = T4w + T4x; ci[WS(rs, 12)] = T4z + T4A; ci[WS(rs, 16)] = T4A - T4z; T4y = T4w - T4x; cr[WS(rs, 11)] = T4v + T4y; cr[WS(rs, 19)] = T4y - T4v; } } } } static const tw_instr twinstr[] = { {TW_FULL, 1, 20}, {TW_NEXT, 1, 0} }; static const hc2hc_desc desc = { 20, "hf_20", twinstr, &GENUS, {184, 62, 62, 0} }; void X(codelet_hf_20) (planner *p) { X(khc2hc_register) (p, hf_20, &desc); } #endif /* HAVE_FMA */ fftw-3.3.3/rdft/scalar/r2cf/hf_25.c0000644000175000017500000014145212054410647013530 00000000000000/* * Copyright (c) 2003, 2007-11 Matteo Frigo * Copyright (c) 2003, 2007-11 Massachusetts Institute of Technology * * This program is free software; you can redistribute it and/or modify * it under the terms of the GNU General Public License as published by * the Free Software Foundation; either version 2 of the License, or * (at your option) any later version. * * This program is distributed in the hope that it will be useful, * but WITHOUT ANY WARRANTY; without even the implied warranty of * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the * GNU General Public License for more details. * * You should have received a copy of the GNU General Public License * along with this program; if not, write to the Free Software * Foundation, Inc., 51 Franklin Street, Fifth Floor, Boston, MA 02110-1301 USA * */ /* This file was automatically generated --- DO NOT EDIT */ /* Generated on Sun Nov 25 07:40:01 EST 2012 */ #include "codelet-rdft.h" #ifdef HAVE_FMA /* Generated by: ../../../genfft/gen_hc2hc.native -fma -reorder-insns -schedule-for-pipeline -compact -variables 4 -pipeline-latency 4 -n 25 -dit -name hf_25 -include hf.h */ /* * This function contains 400 FP additions, 364 FP multiplications, * (or, 84 additions, 48 multiplications, 316 fused multiply/add), * 178 stack variables, 47 constants, and 100 memory accesses */ #include "hf.h" static void hf_25(R *cr, R *ci, const R *W, stride rs, INT mb, INT me, INT ms) { DK(KP949179823, +0.949179823508441261575555465843363271711583843); DK(KP860541664, +0.860541664367944677098261680920518816412804187); DK(KP621716863, +0.621716863012209892444754556304102309693593202); DK(KP614372930, +0.614372930789563808870829930444362096004872855); DK(KP557913902, +0.557913902031834264187699648465567037992437152); DK(KP249506682, +0.249506682107067890488084201715862638334226305); DK(KP560319534, +0.560319534973832390111614715371676131169633784); DK(KP681693190, +0.681693190061530575150324149145440022633095390); DK(KP906616052, +0.906616052148196230441134447086066874408359177); DK(KP968479752, +0.968479752739016373193524836781420152702090879); DK(KP845997307, +0.845997307939530944175097360758058292389769300); DK(KP998026728, +0.998026728428271561952336806863450553336905220); DK(KP994076283, +0.994076283785401014123185814696322018529298887); DK(KP734762448, +0.734762448793050413546343770063151342619912334); DK(KP772036680, +0.772036680810363904029489473607579825330539880); DK(KP062914667, +0.062914667253649757225485955897349402364686947); DK(KP833417178, +0.833417178328688677408962550243238843138996060); DK(KP921177326, +0.921177326965143320250447435415066029359282231); DK(KP541454447, +0.541454447536312777046285590082819509052033189); DK(KP803003575, +0.803003575438660414833440593570376004635464850); DK(KP943557151, +0.943557151597354104399655195398983005179443399); DK(KP554608978, +0.554608978404018097464974850792216217022558774); DK(KP242145790, +0.242145790282157779872542093866183953459003101); DK(KP559154169, +0.559154169276087864842202529084232643714075927); DK(KP683113946, +0.683113946453479238701949862233725244439656928); DK(KP248028675, +0.248028675328619457762448260696444630363259177); DK(KP968583161, +0.968583161128631119490168375464735813836012403); DK(KP525970792, +0.525970792408939708442463226536226366643874659); DK(KP726211448, +0.726211448929902658173535992263577167607493062); DK(KP904730450, +0.904730450839922351881287709692877908104763647); DK(KP831864738, +0.831864738706457140726048799369896829771167132); DK(KP871714437, +0.871714437527667770979999223229522602943903653); DK(KP549754652, +0.549754652192770074288023275540779861653779767); DK(KP992114701, +0.992114701314477831049793042785778521453036709); DK(KP939062505, +0.939062505817492352556001843133229685779824606); DK(KP256756360, +0.256756360367726783319498520922669048172391148); DK(KP851038619, +0.851038619207379630836264138867114231259902550); DK(KP912575812, +0.912575812670962425556968549836277086778922727); DK(KP912018591, +0.912018591466481957908415381764119056233607330); DK(KP634619297, +0.634619297544148100711287640319130485732531031); DK(KP470564281, +0.470564281212251493087595091036643380879947982); DK(KP827271945, +0.827271945972475634034355757144307982555673741); DK(KP126329378, +0.126329378446108174786050455341811215027378105); DK(KP951056516, +0.951056516295153572116439333379382143405698634); DK(KP559016994, +0.559016994374947424102293417182819058860154590); DK(KP250000000, +0.250000000000000000000000000000000000000000000); DK(KP618033988, +0.618033988749894848204586834365638117720309180); { INT m; for (m = mb, W = W + ((mb - 1) * 48); m < me; m = m + 1, cr = cr + ms, ci = ci - ms, W = W + 48, MAKE_VOLATILE_STRIDE(50, rs)) { E T7i, T6o, T6m, T7o, T7m, T7h, T6n, T6f, T7j, T7n; { E T6W, T5G, T3Y, T3M, T7q, T70, T6V, T7P, Tt, T3L, T5T, T45, T5Q, T4c, T3G; E T2G, T5P, T49, T5S, T42, T65, T4H, T68, T4A, T2Z, T11, T67, T4x, T64, T4E; E T5Y, T4W, T61, T4P, T3d, T1z, T60, T4M, T5X, T4T, T3g, T1G, T3q, T4q, T4j; E T26, T3i, T1M, T3k, T1S; { E T3u, T2e, T3E, T44, T4b, T2E, T3w, T2k, T3y, T2q; { E T1, T6R, T3P, T7, T3W, Tq, T9, Tc, Tb, T3U, Tk, T3Q, Ta; { E T3, T6, T2, T5; T1 = cr[0]; T6R = ci[0]; T3 = cr[WS(rs, 5)]; T6 = ci[WS(rs, 5)]; T2 = W[8]; T5 = W[9]; { E Tm, Tp, To, T3V, Tn, T3O, T4, Tl; Tm = cr[WS(rs, 15)]; Tp = ci[WS(rs, 15)]; T3O = T2 * T6; T4 = T2 * T3; Tl = W[28]; To = W[29]; T3P = FNMS(T5, T3, T3O); T7 = FMA(T5, T6, T4); T3V = Tl * Tp; Tn = Tl * Tm; { E Tg, Tj, Tf, Ti, T3T, Th, T8; Tg = cr[WS(rs, 10)]; Tj = ci[WS(rs, 10)]; T3W = FNMS(To, Tm, T3V); Tq = FMA(To, Tp, Tn); Tf = W[18]; Ti = W[19]; T9 = cr[WS(rs, 20)]; Tc = ci[WS(rs, 20)]; T3T = Tf * Tj; Th = Tf * Tg; T8 = W[38]; Tb = W[39]; T3U = FNMS(Ti, Tg, T3T); Tk = FMA(Ti, Tj, Th); T3Q = T8 * Tc; Ta = T8 * T9; } } } { E T6T, T3X, T6Y, Tr, T3R, Td; T6T = T3U + T3W; T3X = T3U - T3W; T6Y = Tk - Tq; Tr = Tk + Tq; T3R = FNMS(Tb, T9, T3Q); Td = FMA(Tb, Tc, Ta); { E T3S, T6Z, Te, T6U, T6S, Ts; T3S = T3P - T3R; T6S = T3P + T3R; T6Z = T7 - Td; Te = T7 + Td; T6W = T6S - T6T; T6U = T6S + T6T; T5G = FNMS(KP618033988, T3S, T3X); T3Y = FMA(KP618033988, T3X, T3S); T3M = Te - Tr; Ts = Te + Tr; T7q = FMA(KP618033988, T6Y, T6Z); T70 = FNMS(KP618033988, T6Z, T6Y); T6V = FNMS(KP250000000, T6U, T6R); T7P = T6U + T6R; Tt = T1 + Ts; T3L = FNMS(KP250000000, Ts, T1); } } } { E T2g, T2j, T2m, T3v, T2h, T2p, T2l, T2i, T2o, T3x, T2n; { E T2a, T2d, T29, T2c; T2a = cr[WS(rs, 3)]; T2d = ci[WS(rs, 3)]; T29 = W[4]; T2c = W[5]; { E T2t, T2w, T2z, T3A, T2u, T2C, T2y, T2v, T2B, T3t, T2b, T2s, T2f; T2t = cr[WS(rs, 13)]; T2w = ci[WS(rs, 13)]; T3t = T29 * T2d; T2b = T29 * T2a; T2s = W[24]; T2z = cr[WS(rs, 18)]; T3u = FNMS(T2c, T2a, T3t); T2e = FMA(T2c, T2d, T2b); T3A = T2s * T2w; T2u = T2s * T2t; T2C = ci[WS(rs, 18)]; T2y = W[34]; T2v = W[25]; T2B = W[35]; { E T3B, T2x, T3D, T2D, T3C, T2A; T2g = cr[WS(rs, 8)]; T3C = T2y * T2C; T2A = T2y * T2z; T3B = FNMS(T2v, T2t, T3A); T2x = FMA(T2v, T2w, T2u); T3D = FNMS(T2B, T2z, T3C); T2D = FMA(T2B, T2C, T2A); T2j = ci[WS(rs, 8)]; T2f = W[14]; T3E = T3B + T3D; T44 = T3D - T3B; T4b = T2x - T2D; T2E = T2x + T2D; } T2m = cr[WS(rs, 23)]; T3v = T2f * T2j; T2h = T2f * T2g; T2p = ci[WS(rs, 23)]; T2l = W[44]; T2i = W[15]; T2o = W[45]; } } T3x = T2l * T2p; T2n = T2l * T2m; T3w = FNMS(T2i, T2g, T3v); T2k = FMA(T2i, T2j, T2h); T3y = FNMS(T2o, T2m, T3x); T2q = FMA(T2o, T2p, T2n); } { E T2N, Tz, T2X, T4G, T4z, TZ, T2P, TF, T2R, TL; { E TB, TE, TH, T2O, TC, TK, TG, TD, TJ, T2Q, TI; { E Tv, Ty, Tu, Tx; { E T48, T41, T47, T40, T43, T3z; Tv = cr[WS(rs, 1)]; T43 = T3y - T3w; T3z = T3w + T3y; { E T4a, T2r, T3F, T2F; T4a = T2k - T2q; T2r = T2k + T2q; T5T = FNMS(KP618033988, T43, T44); T45 = FMA(KP618033988, T44, T43); T3F = T3z + T3E; T48 = T3E - T3z; T5Q = FNMS(KP618033988, T4a, T4b); T4c = FMA(KP618033988, T4b, T4a); T2F = T2r + T2E; T41 = T2E - T2r; T3G = T3u + T3F; T47 = FNMS(KP250000000, T3F, T3u); T2G = T2e + T2F; T40 = FNMS(KP250000000, T2F, T2e); Ty = ci[WS(rs, 1)]; } T5P = FMA(KP559016994, T48, T47); T49 = FNMS(KP559016994, T48, T47); T5S = FMA(KP559016994, T41, T40); T42 = FNMS(KP559016994, T41, T40); Tu = W[0]; } Tx = W[1]; { E TO, TR, TU, T2T, TP, TX, TT, TQ, TW, T2M, Tw, TN, TA; TO = cr[WS(rs, 11)]; TR = ci[WS(rs, 11)]; T2M = Tu * Ty; Tw = Tu * Tv; TN = W[20]; TU = cr[WS(rs, 16)]; T2N = FNMS(Tx, Tv, T2M); Tz = FMA(Tx, Ty, Tw); T2T = TN * TR; TP = TN * TO; TX = ci[WS(rs, 16)]; TT = W[30]; TQ = W[21]; TW = W[31]; { E T2U, TS, T2W, TY, T2V, TV; TB = cr[WS(rs, 6)]; T2V = TT * TX; TV = TT * TU; T2U = FNMS(TQ, TO, T2T); TS = FMA(TQ, TR, TP); T2W = FNMS(TW, TU, T2V); TY = FMA(TW, TX, TV); TE = ci[WS(rs, 6)]; TA = W[10]; T2X = T2U + T2W; T4G = T2W - T2U; T4z = TY - TS; TZ = TS + TY; } TH = cr[WS(rs, 21)]; T2O = TA * TE; TC = TA * TB; TK = ci[WS(rs, 21)]; TG = W[40]; TD = W[11]; TJ = W[41]; } } T2Q = TG * TK; TI = TG * TH; T2P = FNMS(TD, TB, T2O); TF = FMA(TD, TE, TC); T2R = FNMS(TJ, TH, T2Q); TL = FMA(TJ, TK, TI); } { E T31, T17, T3b, T4V, T4O, T1x, T33, T1d, T35, T1j; { E T19, T1c, T1f, T32, T1a, T1i, T1e, T1b, T1h, T34, T1g; { E T13, T16, T12, T15; { E T4w, T4D, T4v, T4C, T4F, T2S; T13 = cr[WS(rs, 4)]; T4F = T2P - T2R; T2S = T2P + T2R; { E T4y, TM, T2Y, T10; T4y = TL - TF; TM = TF + TL; T65 = FMA(KP618033988, T4F, T4G); T4H = FNMS(KP618033988, T4G, T4F); T2Y = T2S + T2X; T4w = T2S - T2X; T68 = FNMS(KP618033988, T4y, T4z); T4A = FMA(KP618033988, T4z, T4y); T10 = TM + TZ; T4D = TM - TZ; T2Z = T2N + T2Y; T4v = FNMS(KP250000000, T2Y, T2N); T11 = Tz + T10; T4C = FNMS(KP250000000, T10, Tz); T16 = ci[WS(rs, 4)]; } T67 = FNMS(KP559016994, T4w, T4v); T4x = FMA(KP559016994, T4w, T4v); T64 = FNMS(KP559016994, T4D, T4C); T4E = FMA(KP559016994, T4D, T4C); T12 = W[6]; } T15 = W[7]; { E T1m, T1p, T1s, T37, T1n, T1v, T1r, T1o, T1u, T30, T14, T1l, T18; T1m = cr[WS(rs, 14)]; T1p = ci[WS(rs, 14)]; T30 = T12 * T16; T14 = T12 * T13; T1l = W[26]; T1s = cr[WS(rs, 19)]; T31 = FNMS(T15, T13, T30); T17 = FMA(T15, T16, T14); T37 = T1l * T1p; T1n = T1l * T1m; T1v = ci[WS(rs, 19)]; T1r = W[36]; T1o = W[27]; T1u = W[37]; { E T38, T1q, T3a, T1w, T39, T1t; T19 = cr[WS(rs, 9)]; T39 = T1r * T1v; T1t = T1r * T1s; T38 = FNMS(T1o, T1m, T37); T1q = FMA(T1o, T1p, T1n); T3a = FNMS(T1u, T1s, T39); T1w = FMA(T1u, T1v, T1t); T1c = ci[WS(rs, 9)]; T18 = W[16]; T3b = T38 + T3a; T4V = T3a - T38; T4O = T1w - T1q; T1x = T1q + T1w; } T1f = cr[WS(rs, 24)]; T32 = T18 * T1c; T1a = T18 * T19; T1i = ci[WS(rs, 24)]; T1e = W[46]; T1b = W[17]; T1h = W[47]; } } T34 = T1e * T1i; T1g = T1e * T1f; T33 = FNMS(T1b, T19, T32); T1d = FMA(T1b, T1c, T1a); T35 = FNMS(T1h, T1f, T34); T1j = FMA(T1h, T1i, T1g); } { E T1I, T1L, T1O, T3h, T1J, T1R, T1N, T1K, T1Q, T3j, T1P; { E T1C, T1F, T1B, T1E; { E T4L, T4S, T4K, T4R, T4U, T36; T1C = cr[WS(rs, 2)]; T4U = T35 - T33; T36 = T33 + T35; { E T4N, T1k, T3c, T1y; T4N = T1j - T1d; T1k = T1d + T1j; T5Y = FNMS(KP618033988, T4U, T4V); T4W = FMA(KP618033988, T4V, T4U); T3c = T36 + T3b; T4L = T3b - T36; T61 = FNMS(KP618033988, T4N, T4O); T4P = FMA(KP618033988, T4O, T4N); T1y = T1k + T1x; T4S = T1k - T1x; T3d = T31 + T3c; T4K = FNMS(KP250000000, T3c, T31); T1z = T17 + T1y; T4R = FNMS(KP250000000, T1y, T17); T1F = ci[WS(rs, 2)]; } T60 = FMA(KP559016994, T4L, T4K); T4M = FNMS(KP559016994, T4L, T4K); T5X = FNMS(KP559016994, T4S, T4R); T4T = FMA(KP559016994, T4S, T4R); T1B = W[2]; } T1E = W[3]; { E T1V, T1Y, T21, T3m, T1W, T24, T20, T1X, T23, T3f, T1D, T1U, T1H; T1V = cr[WS(rs, 12)]; T1Y = ci[WS(rs, 12)]; T3f = T1B * T1F; T1D = T1B * T1C; T1U = W[22]; T21 = cr[WS(rs, 17)]; T3g = FNMS(T1E, T1C, T3f); T1G = FMA(T1E, T1F, T1D); T3m = T1U * T1Y; T1W = T1U * T1V; T24 = ci[WS(rs, 17)]; T20 = W[32]; T1X = W[23]; T23 = W[33]; { E T3n, T1Z, T3p, T25, T3o, T22; T1I = cr[WS(rs, 7)]; T3o = T20 * T24; T22 = T20 * T21; T3n = FNMS(T1X, T1V, T3m); T1Z = FMA(T1X, T1Y, T1W); T3p = FNMS(T23, T21, T3o); T25 = FMA(T23, T24, T22); T1L = ci[WS(rs, 7)]; T1H = W[12]; T3q = T3n + T3p; T4q = T3n - T3p; T4j = T25 - T1Z; T26 = T1Z + T25; } T1O = cr[WS(rs, 22)]; T3h = T1H * T1L; T1J = T1H * T1I; T1R = ci[WS(rs, 22)]; T1N = W[42]; T1K = W[13]; T1Q = W[43]; } } T3j = T1N * T1R; T1P = T1N * T1O; T3i = FNMS(T1K, T1I, T3h); T1M = FMA(T1K, T1L, T1J); T3k = FNMS(T1Q, T1O, T3j); T1S = FMA(T1Q, T1R, T1P); } } } } { E T7Q, T5M, T5J, T7R, T5I, T5L, T7X, T7W, T5F, T6X, T5u, T7M, T7O, T5C, T5E; E T5t, T7J, T7N; { E T4r, T4k, T4h, T4o, T3K, T3I, T1A, T2H, T28; { E T3e, T4g, T4n, T4f, T4m, T3H, T4p, T3l; T7Q = T2Z + T3d; T3e = T2Z - T3d; T4p = T3k - T3i; T3l = T3i + T3k; { E T4i, T1T, T3r, T27, T3s; T4i = T1S - T1M; T1T = T1M + T1S; T5M = FMA(KP618033988, T4p, T4q); T4r = FNMS(KP618033988, T4q, T4p); T3r = T3l + T3q; T4g = T3q - T3l; T5J = FNMS(KP618033988, T4i, T4j); T4k = FMA(KP618033988, T4j, T4i); T27 = T1T + T26; T4n = T26 - T1T; T3s = T3g + T3r; T4f = FNMS(KP250000000, T3r, T3g); T28 = T1G + T27; T4m = FNMS(KP250000000, T27, T1G); T3H = T3s - T3G; T7R = T3s + T3G; } T5I = FMA(KP559016994, T4g, T4f); T4h = FNMS(KP559016994, T4g, T4f); T5L = FMA(KP559016994, T4n, T4m); T4o = FNMS(KP559016994, T4n, T4m); T3K = FNMS(KP618033988, T3e, T3H); T3I = FMA(KP618033988, T3H, T3e); } T1A = T11 + T1z; T7X = T1z - T11; T7W = T28 - T2G; T2H = T28 + T2G; { E T3Z, T5d, T7r, T7D, T5h, T5i, T5m, T5l, T59, T7K, T56, T7L, T7I, T7G, T52; E T50, T5w, T5g, T5q, T5A, T3N, T7p; T3N = FMA(KP559016994, T3M, T3L); T5F = FNMS(KP559016994, T3M, T3L); T6X = FNMS(KP559016994, T6W, T6V); T7p = FMA(KP559016994, T6W, T6V); { E T5o, T5p, T57, T4e, T4Y, T55, T4l, T4s, T4B, T5f, T5e, T4I; { E T46, T2K, T2J, T4d, T2I; T46 = FMA(KP951056516, T45, T42); T5o = FNMS(KP951056516, T45, T42); T2I = T1A + T2H; T2K = T1A - T2H; T3Z = FNMS(KP951056516, T3Y, T3N); T5d = FMA(KP951056516, T3Y, T3N); T7r = FNMS(KP951056516, T7q, T7p); T7D = FMA(KP951056516, T7q, T7p); cr[0] = Tt + T2I; T2J = FNMS(KP250000000, T2I, Tt); T5p = FNMS(KP951056516, T4c, T49); T4d = FMA(KP951056516, T4c, T49); { E T4Q, T4X, T2L, T3J; T4Q = FNMS(KP951056516, T4P, T4M); T5h = FMA(KP951056516, T4P, T4M); T5i = FNMS(KP951056516, T4W, T4T); T4X = FMA(KP951056516, T4W, T4T); T2L = FMA(KP559016994, T2K, T2J); T3J = FNMS(KP559016994, T2K, T2J); T57 = FMA(KP126329378, T46, T4d); T4e = FNMS(KP126329378, T4d, T46); cr[WS(rs, 5)] = FMA(KP951056516, T3I, T2L); ci[WS(rs, 4)] = FNMS(KP951056516, T3I, T2L); ci[WS(rs, 9)] = FMA(KP951056516, T3K, T3J); cr[WS(rs, 10)] = FNMS(KP951056516, T3K, T3J); T4Y = FMA(KP827271945, T4X, T4Q); T55 = FNMS(KP827271945, T4Q, T4X); } } T4l = FNMS(KP951056516, T4k, T4h); T5m = FMA(KP951056516, T4k, T4h); T5l = FNMS(KP951056516, T4r, T4o); T4s = FMA(KP951056516, T4r, T4o); T4B = FNMS(KP951056516, T4A, T4x); T5f = FMA(KP951056516, T4A, T4x); T5e = FMA(KP951056516, T4H, T4E); T4I = FNMS(KP951056516, T4H, T4E); { E T4u, T4Z, T4t, T58; T4t = FNMS(KP470564281, T4s, T4l); T58 = FMA(KP470564281, T4l, T4s); { E T4J, T54, T7E, T7F; T4J = FMA(KP634619297, T4I, T4B); T54 = FNMS(KP634619297, T4B, T4I); T59 = FNMS(KP912018591, T58, T57); T7E = FMA(KP912018591, T58, T57); T7K = FMA(KP912018591, T4t, T4e); T4u = FNMS(KP912018591, T4t, T4e); T56 = FMA(KP912575812, T55, T54); T7F = FNMS(KP912575812, T55, T54); T7L = FMA(KP912575812, T4Y, T4J); T4Z = FNMS(KP912575812, T4Y, T4J); T7I = FNMS(KP851038619, T7F, T7E); T7G = FMA(KP851038619, T7F, T7E); } T52 = FMA(KP851038619, T4Z, T4u); T50 = FNMS(KP851038619, T4Z, T4u); } T5w = FNMS(KP256756360, T5e, T5f); T5g = FMA(KP256756360, T5f, T5e); T5q = FMA(KP939062505, T5p, T5o); T5A = FNMS(KP939062505, T5o, T5p); } { E T5y, T7z, T5B, T7y, T7w, T7u, T5s; { E T5k, T5r, T5j, T5x; cr[WS(rs, 4)] = FNMS(KP992114701, T50, T3Z); T5j = FMA(KP634619297, T5i, T5h); T5x = FNMS(KP634619297, T5h, T5i); { E T5n, T5z, T7s, T7t; T5n = FMA(KP549754652, T5m, T5l); T5z = FNMS(KP549754652, T5l, T5m); T5y = FMA(KP871714437, T5x, T5w); T7s = FNMS(KP871714437, T5x, T5w); T7z = FNMS(KP871714437, T5j, T5g); T5k = FMA(KP871714437, T5j, T5g); T5B = FNMS(KP831864738, T5A, T5z); T7t = FMA(KP831864738, T5A, T5z); T7y = FNMS(KP831864738, T5q, T5n); T5r = FMA(KP831864738, T5q, T5n); T7w = FNMS(KP904730450, T7t, T7s); T7u = FMA(KP904730450, T7t, T7s); } ci[WS(rs, 20)] = FNMS(KP992114701, T7G, T7D); T5u = FNMS(KP904730450, T5r, T5k); T5s = FMA(KP904730450, T5r, T5k); } { E T5a, T5c, T7A, T7C, T7v, T53, T5b, T51, T7H, T7x, T7B; T5a = FNMS(KP726211448, T59, T56); T5c = FMA(KP525970792, T56, T59); ci[WS(rs, 23)] = FMA(KP968583161, T7u, T7r); cr[WS(rs, 1)] = FMA(KP968583161, T5s, T5d); T51 = FMA(KP248028675, T50, T3Z); T7A = FNMS(KP683113946, T7z, T7y); T7C = FMA(KP559154169, T7y, T7z); T7v = FNMS(KP242145790, T7u, T7r); T53 = FMA(KP554608978, T52, T51); T5b = FNMS(KP554608978, T52, T51); T7M = FNMS(KP525970792, T7L, T7K); T7O = FMA(KP726211448, T7K, T7L); ci[WS(rs, 10)] = FNMS(KP943557151, T5c, T5b); ci[WS(rs, 5)] = FMA(KP943557151, T5c, T5b); ci[0] = FMA(KP803003575, T5a, T53); cr[WS(rs, 9)] = FNMS(KP803003575, T5a, T53); T7x = FNMS(KP541454447, T7w, T7v); T7B = FMA(KP541454447, T7w, T7v); T7H = FMA(KP248028675, T7G, T7D); cr[WS(rs, 21)] = -(FMA(KP921177326, T7C, T7B)); ci[WS(rs, 18)] = FNMS(KP921177326, T7C, T7B); ci[WS(rs, 13)] = FMA(KP833417178, T7A, T7x); cr[WS(rs, 16)] = FMS(KP833417178, T7A, T7x); T5C = FMA(KP559154169, T5B, T5y); T5E = FNMS(KP683113946, T5y, T5B); T5t = FNMS(KP242145790, T5s, T5d); T7J = FNMS(KP554608978, T7I, T7H); T7N = FMA(KP554608978, T7I, T7H); } } } } { E T7Y, T80, T5v, T5D; cr[WS(rs, 24)] = -(FMA(KP803003575, T7O, T7N)); ci[WS(rs, 15)] = FNMS(KP803003575, T7O, T7N); cr[WS(rs, 19)] = FMS(KP943557151, T7M, T7J); cr[WS(rs, 14)] = -(FMA(KP943557151, T7M, T7J)); T5v = FMA(KP541454447, T5u, T5t); T5D = FNMS(KP541454447, T5u, T5t); cr[WS(rs, 11)] = FNMS(KP833417178, T5E, T5D); ci[WS(rs, 8)] = FMA(KP833417178, T5E, T5D); cr[WS(rs, 6)] = FMA(KP921177326, T5C, T5v); ci[WS(rs, 3)] = FNMS(KP921177326, T5C, T5v); T7Y = FMA(KP618033988, T7X, T7W); T80 = FNMS(KP618033988, T7W, T7X); { E T6t, T6p, T5H, T7d, T71, T6u, T6y, T6x, T6l, T7k, T6i, T7l, T7g, T6c, T6e; E T6s, T6L, T6J, T6C; { E T6A, T6B, T5O, T6j, T6h, T6a, T6q, T5R, T5U, T6r, T5Z, T62; { E T5K, T7U, T7T, T5N, T7S; T6t = FNMS(KP951056516, T5J, T5I); T5K = FMA(KP951056516, T5J, T5I); T7U = T7Q - T7R; T7S = T7Q + T7R; T6p = FNMS(KP951056516, T5G, T5F); T5H = FMA(KP951056516, T5G, T5F); T7d = FNMS(KP951056516, T70, T6X); T71 = FMA(KP951056516, T70, T6X); ci[WS(rs, 24)] = T7S + T7P; T7T = FNMS(KP250000000, T7S, T7P); T5N = FMA(KP951056516, T5M, T5L); T6u = FNMS(KP951056516, T5M, T5L); { E T66, T69, T7Z, T7V; T6A = FMA(KP951056516, T65, T64); T66 = FNMS(KP951056516, T65, T64); T69 = FMA(KP951056516, T68, T67); T6B = FNMS(KP951056516, T68, T67); T7Z = FMA(KP559016994, T7U, T7T); T7V = FNMS(KP559016994, T7U, T7T); T5O = FMA(KP062914667, T5N, T5K); T6j = FNMS(KP062914667, T5K, T5N); ci[WS(rs, 14)] = FMA(KP951056516, T7Y, T7V); cr[WS(rs, 15)] = FMS(KP951056516, T7Y, T7V); ci[WS(rs, 19)] = FMA(KP951056516, T80, T7Z); cr[WS(rs, 20)] = FMS(KP951056516, T80, T7Z); T6h = FNMS(KP939062505, T66, T69); T6a = FMA(KP939062505, T69, T66); } } T6q = FMA(KP951056516, T5Q, T5P); T5R = FNMS(KP951056516, T5Q, T5P); T5U = FNMS(KP951056516, T5T, T5S); T6r = FMA(KP951056516, T5T, T5S); T6y = FMA(KP951056516, T5Y, T5X); T5Z = FNMS(KP951056516, T5Y, T5X); T62 = FMA(KP951056516, T61, T60); T6x = FNMS(KP951056516, T61, T60); { E T5W, T6b, T6k, T5V; T6k = FMA(KP827271945, T5R, T5U); T5V = FNMS(KP827271945, T5U, T5R); { E T6g, T63, T7e, T7f; T6g = FMA(KP126329378, T5Z, T62); T63 = FNMS(KP126329378, T62, T5Z); T7e = FMA(KP772036680, T6k, T6j); T6l = FNMS(KP772036680, T6k, T6j); T5W = FMA(KP772036680, T5V, T5O); T7k = FNMS(KP772036680, T5V, T5O); T7f = FNMS(KP734762448, T6h, T6g); T6i = FMA(KP734762448, T6h, T6g); T6b = FNMS(KP734762448, T6a, T63); T7l = FMA(KP734762448, T6a, T63); T7g = FMA(KP994076283, T7f, T7e); T7i = FNMS(KP994076283, T7f, T7e); } T6c = FNMS(KP994076283, T6b, T5W); T6e = FMA(KP994076283, T6b, T5W); } T6s = FMA(KP062914667, T6r, T6q); T6L = FNMS(KP062914667, T6q, T6r); T6J = FNMS(KP549754652, T6A, T6B); T6C = FMA(KP549754652, T6B, T6A); } { E T6N, T78, T6K, T79, T74, T76, T6E, T6G; { E T6w, T6D, T6M, T6v; cr[WS(rs, 3)] = FMA(KP998026728, T6c, T5H); T6M = FNMS(KP634619297, T6t, T6u); T6v = FMA(KP634619297, T6u, T6t); { E T6I, T6z, T72, T73; T6I = FMA(KP470564281, T6x, T6y); T6z = FNMS(KP470564281, T6y, T6x); T72 = FMA(KP845997307, T6M, T6L); T6N = FNMS(KP845997307, T6M, T6L); T6w = FMA(KP845997307, T6v, T6s); T78 = FNMS(KP845997307, T6v, T6s); T73 = FNMS(KP968479752, T6J, T6I); T6K = FMA(KP968479752, T6J, T6I); T6D = FMA(KP968479752, T6C, T6z); T79 = FNMS(KP968479752, T6C, T6z); T74 = FMA(KP906616052, T73, T72); T76 = FNMS(KP906616052, T73, T72); } ci[WS(rs, 21)] = FNMS(KP998026728, T7g, T7d); T6E = FMA(KP906616052, T6D, T6w); T6G = FNMS(KP906616052, T6D, T6w); } { E T7c, T7a, T6Q, T6O, T6F, T7b, T77, T75, T6d, T6P, T6H; T7c = FMA(KP681693190, T78, T79); T7a = FNMS(KP560319534, T79, T78); ci[WS(rs, 22)] = FNMS(KP998026728, T74, T71); cr[WS(rs, 2)] = FMA(KP998026728, T6E, T6p); T75 = FMA(KP249506682, T74, T71); T6Q = FNMS(KP560319534, T6K, T6N); T6O = FMA(KP681693190, T6N, T6K); T6F = FNMS(KP249506682, T6E, T6p); T7b = FMA(KP557913902, T76, T75); T77 = FNMS(KP557913902, T76, T75); T6o = FMA(KP614372930, T6i, T6l); T6m = FNMS(KP621716863, T6l, T6i); cr[WS(rs, 22)] = FMS(KP860541664, T7c, T7b); ci[WS(rs, 17)] = FMA(KP860541664, T7c, T7b); ci[WS(rs, 12)] = FNMS(KP949179823, T7a, T77); cr[WS(rs, 17)] = -(FMA(KP949179823, T7a, T77)); T6P = FMA(KP557913902, T6G, T6F); T6H = FNMS(KP557913902, T6G, T6F); T6d = FNMS(KP249506682, T6c, T5H); ci[WS(rs, 7)] = FMA(KP949179823, T6Q, T6P); cr[WS(rs, 12)] = FNMS(KP949179823, T6Q, T6P); cr[WS(rs, 7)] = FMA(KP860541664, T6O, T6H); ci[WS(rs, 2)] = FNMS(KP860541664, T6O, T6H); T7o = FMA(KP621716863, T7k, T7l); T7m = FNMS(KP614372930, T7l, T7k); T7h = FMA(KP249506682, T7g, T7d); T6n = FMA(KP557913902, T6e, T6d); T6f = FNMS(KP557913902, T6e, T6d); } } } } } } ci[WS(rs, 6)] = FNMS(KP949179823, T6o, T6n); ci[WS(rs, 11)] = FMA(KP949179823, T6o, T6n); cr[WS(rs, 8)] = FMA(KP943557151, T6m, T6f); ci[WS(rs, 1)] = FNMS(KP943557151, T6m, T6f); T7j = FNMS(KP557913902, T7i, T7h); T7n = FMA(KP557913902, T7i, T7h); cr[WS(rs, 23)] = -(FMA(KP943557151, T7o, T7n)); ci[WS(rs, 16)] = FNMS(KP943557151, T7o, T7n); cr[WS(rs, 18)] = FMS(KP949179823, T7m, T7j); cr[WS(rs, 13)] = -(FMA(KP949179823, T7m, T7j)); } } } static const tw_instr twinstr[] = { {TW_FULL, 1, 25}, {TW_NEXT, 1, 0} }; static const hc2hc_desc desc = { 25, "hf_25", twinstr, &GENUS, {84, 48, 316, 0} }; void X(codelet_hf_25) (planner *p) { X(khc2hc_register) (p, hf_25, &desc); } #else /* HAVE_FMA */ /* Generated by: ../../../genfft/gen_hc2hc.native -compact -variables 4 -pipeline-latency 4 -n 25 -dit -name hf_25 -include hf.h */ /* * This function contains 400 FP additions, 280 FP multiplications, * (or, 260 additions, 140 multiplications, 140 fused multiply/add), * 101 stack variables, 20 constants, and 100 memory accesses */ #include "hf.h" static void hf_25(R *cr, R *ci, const R *W, stride rs, INT mb, INT me, INT ms) { DK(KP998026728, +0.998026728428271561952336806863450553336905220); DK(KP062790519, +0.062790519529313376076178224565631133122484832); DK(KP684547105, +0.684547105928688673732283357621209269889519233); DK(KP728968627, +0.728968627421411523146730319055259111372571664); DK(KP481753674, +0.481753674101715274987191502872129653528542010); DK(KP876306680, +0.876306680043863587308115903922062583399064238); DK(KP248689887, +0.248689887164854788242283746006447968417567406); DK(KP968583161, +0.968583161128631119490168375464735813836012403); DK(KP992114701, +0.992114701314477831049793042785778521453036709); DK(KP125333233, +0.125333233564304245373118759816508793942918247); DK(KP425779291, +0.425779291565072648862502445744251703979973042); DK(KP904827052, +0.904827052466019527713668647932697593970413911); DK(KP637423989, +0.637423989748689710176712811676016195434917298); DK(KP770513242, +0.770513242775789230803009636396177847271667672); DK(KP844327925, +0.844327925502015078548558063966681505381659241); DK(KP535826794, +0.535826794978996618271308767867639978063575346); DK(KP587785252, +0.587785252292473129168705954639072768597652438); DK(KP951056516, +0.951056516295153572116439333379382143405698634); DK(KP250000000, +0.250000000000000000000000000000000000000000000); DK(KP559016994, +0.559016994374947424102293417182819058860154590); { INT m; for (m = mb, W = W + ((mb - 1) * 48); m < me; m = m + 1, cr = cr + ms, ci = ci - ms, W = W + 48, MAKE_VOLATILE_STRIDE(50, rs)) { E T1, T6b, T2l, T6g, To, T2m, T6e, T6f, T6a, T6H, T2u, T4I, T2i, T60, T3S; E T5D, T4r, T58, T3Z, T5C, T4q, T5b, TS, T5W, T2G, T5s, T4g, T4M, T2R, T5t; E T4h, T4P, T1l, T5X, T37, T5v, T4k, T4T, T3e, T5w, T4j, T4W, T1P, T5Z, T3v; E T5A, T4o, T54, T3C, T5z, T4n, T51; { E T6, T2o, Tb, T2p, Tc, T6c, Th, T2r, Tm, T2s, Tn, T6d; T1 = cr[0]; T6b = ci[0]; { E T3, T5, T2, T4; T3 = cr[WS(rs, 5)]; T5 = ci[WS(rs, 5)]; T2 = W[8]; T4 = W[9]; T6 = FMA(T2, T3, T4 * T5); T2o = FNMS(T4, T3, T2 * T5); } { E T8, Ta, T7, T9; T8 = cr[WS(rs, 20)]; Ta = ci[WS(rs, 20)]; T7 = W[38]; T9 = W[39]; Tb = FMA(T7, T8, T9 * Ta); T2p = FNMS(T9, T8, T7 * Ta); } Tc = T6 + Tb; T6c = T2o + T2p; { E Te, Tg, Td, Tf; Te = cr[WS(rs, 10)]; Tg = ci[WS(rs, 10)]; Td = W[18]; Tf = W[19]; Th = FMA(Td, Te, Tf * Tg); T2r = FNMS(Tf, Te, Td * Tg); } { E Tj, Tl, Ti, Tk; Tj = cr[WS(rs, 15)]; Tl = ci[WS(rs, 15)]; Ti = W[28]; Tk = W[29]; Tm = FMA(Ti, Tj, Tk * Tl); T2s = FNMS(Tk, Tj, Ti * Tl); } Tn = Th + Tm; T6d = T2r + T2s; T2l = KP559016994 * (Tc - Tn); T6g = KP559016994 * (T6c - T6d); To = Tc + Tn; T2m = FNMS(KP250000000, To, T1); T6e = T6c + T6d; T6f = FNMS(KP250000000, T6e, T6b); { E T68, T69, T2q, T2t; T68 = Th - Tm; T69 = T6 - Tb; T6a = FNMS(KP587785252, T69, KP951056516 * T68); T6H = FMA(KP951056516, T69, KP587785252 * T68); T2q = T2o - T2p; T2t = T2r - T2s; T2u = FMA(KP951056516, T2q, KP587785252 * T2t); T4I = FNMS(KP587785252, T2q, KP951056516 * T2t); } } { E T1U, T3O, T3E, T3F, T3X, T3W, T3J, T3M, T3P, T25, T2g, T2h; { E T1R, T1T, T1Q, T1S; T1R = cr[WS(rs, 3)]; T1T = ci[WS(rs, 3)]; T1Q = W[4]; T1S = W[5]; T1U = FMA(T1Q, T1R, T1S * T1T); T3O = FNMS(T1S, T1R, T1Q * T1T); } { E T1Z, T3H, T2f, T3L, T24, T3I, T2a, T3K; { E T1W, T1Y, T1V, T1X; T1W = cr[WS(rs, 8)]; T1Y = ci[WS(rs, 8)]; T1V = W[14]; T1X = W[15]; T1Z = FMA(T1V, T1W, T1X * T1Y); T3H = FNMS(T1X, T1W, T1V * T1Y); } { E T2c, T2e, T2b, T2d; T2c = cr[WS(rs, 18)]; T2e = ci[WS(rs, 18)]; T2b = W[34]; T2d = W[35]; T2f = FMA(T2b, T2c, T2d * T2e); T3L = FNMS(T2d, T2c, T2b * T2e); } { E T21, T23, T20, T22; T21 = cr[WS(rs, 23)]; T23 = ci[WS(rs, 23)]; T20 = W[44]; T22 = W[45]; T24 = FMA(T20, T21, T22 * T23); T3I = FNMS(T22, T21, T20 * T23); } { E T27, T29, T26, T28; T27 = cr[WS(rs, 13)]; T29 = ci[WS(rs, 13)]; T26 = W[24]; T28 = W[25]; T2a = FMA(T26, T27, T28 * T29); T3K = FNMS(T28, T27, T26 * T29); } T3E = T1Z - T24; T3F = T2a - T2f; T3X = T3K - T3L; T3W = T3H - T3I; T3J = T3H + T3I; T3M = T3K + T3L; T3P = T3J + T3M; T25 = T1Z + T24; T2g = T2a + T2f; T2h = T25 + T2g; } T2i = T1U + T2h; T60 = T3O + T3P; { E T3G, T57, T3R, T56, T3N, T3Q; T3G = FMA(KP951056516, T3E, KP587785252 * T3F); T57 = FNMS(KP587785252, T3E, KP951056516 * T3F); T3N = KP559016994 * (T3J - T3M); T3Q = FNMS(KP250000000, T3P, T3O); T3R = T3N + T3Q; T56 = T3Q - T3N; T3S = T3G + T3R; T5D = T57 + T56; T4r = T3R - T3G; T58 = T56 - T57; } { E T3Y, T5a, T3V, T59, T3T, T3U; T3Y = FMA(KP951056516, T3W, KP587785252 * T3X); T5a = FNMS(KP587785252, T3W, KP951056516 * T3X); T3T = KP559016994 * (T25 - T2g); T3U = FNMS(KP250000000, T2h, T1U); T3V = T3T + T3U; T59 = T3U - T3T; T3Z = T3V - T3Y; T5C = T59 - T5a; T4q = T3V + T3Y; T5b = T59 + T5a; } } { E Tu, T2N, T2B, T2E, T2I, T2H, T2K, T2L, T2O, TF, TQ, TR; { E Tr, Tt, Tq, Ts; Tr = cr[WS(rs, 1)]; Tt = ci[WS(rs, 1)]; Tq = W[0]; Ts = W[1]; Tu = FMA(Tq, Tr, Ts * Tt); T2N = FNMS(Ts, Tr, Tq * Tt); } { E Tz, T2z, TP, T2D, TE, T2A, TK, T2C; { E Tw, Ty, Tv, Tx; Tw = cr[WS(rs, 6)]; Ty = ci[WS(rs, 6)]; Tv = W[10]; Tx = W[11]; Tz = FMA(Tv, Tw, Tx * Ty); T2z = FNMS(Tx, Tw, Tv * Ty); } { E TM, TO, TL, TN; TM = cr[WS(rs, 16)]; TO = ci[WS(rs, 16)]; TL = W[30]; TN = W[31]; TP = FMA(TL, TM, TN * TO); T2D = FNMS(TN, TM, TL * TO); } { E TB, TD, TA, TC; TB = cr[WS(rs, 21)]; TD = ci[WS(rs, 21)]; TA = W[40]; TC = W[41]; TE = FMA(TA, TB, TC * TD); T2A = FNMS(TC, TB, TA * TD); } { E TH, TJ, TG, TI; TH = cr[WS(rs, 11)]; TJ = ci[WS(rs, 11)]; TG = W[20]; TI = W[21]; TK = FMA(TG, TH, TI * TJ); T2C = FNMS(TI, TH, TG * TJ); } T2B = T2z - T2A; T2E = T2C - T2D; T2I = TK - TP; T2H = Tz - TE; T2K = T2z + T2A; T2L = T2C + T2D; T2O = T2K + T2L; TF = Tz + TE; TQ = TK + TP; TR = TF + TQ; } TS = Tu + TR; T5W = T2N + T2O; { E T2F, T4L, T2y, T4K, T2w, T2x; T2F = FMA(KP951056516, T2B, KP587785252 * T2E); T4L = FNMS(KP587785252, T2B, KP951056516 * T2E); T2w = KP559016994 * (TF - TQ); T2x = FNMS(KP250000000, TR, Tu); T2y = T2w + T2x; T4K = T2x - T2w; T2G = T2y - T2F; T5s = T4K - T4L; T4g = T2y + T2F; T4M = T4K + T4L; } { E T2J, T4O, T2Q, T4N, T2M, T2P; T2J = FMA(KP951056516, T2H, KP587785252 * T2I); T4O = FNMS(KP587785252, T2H, KP951056516 * T2I); T2M = KP559016994 * (T2K - T2L); T2P = FNMS(KP250000000, T2O, T2N); T2Q = T2M + T2P; T4N = T2P - T2M; T2R = T2J + T2Q; T5t = T4O + T4N; T4h = T2Q - T2J; T4P = T4N - T4O; } } { E TX, T33, T2T, T2U, T3c, T3b, T2Y, T31, T34, T18, T1j, T1k; { E TU, TW, TT, TV; TU = cr[WS(rs, 4)]; TW = ci[WS(rs, 4)]; TT = W[6]; TV = W[7]; TX = FMA(TT, TU, TV * TW); T33 = FNMS(TV, TU, TT * TW); } { E T12, T2W, T1i, T30, T17, T2X, T1d, T2Z; { E TZ, T11, TY, T10; TZ = cr[WS(rs, 9)]; T11 = ci[WS(rs, 9)]; TY = W[16]; T10 = W[17]; T12 = FMA(TY, TZ, T10 * T11); T2W = FNMS(T10, TZ, TY * T11); } { E T1f, T1h, T1e, T1g; T1f = cr[WS(rs, 19)]; T1h = ci[WS(rs, 19)]; T1e = W[36]; T1g = W[37]; T1i = FMA(T1e, T1f, T1g * T1h); T30 = FNMS(T1g, T1f, T1e * T1h); } { E T14, T16, T13, T15; T14 = cr[WS(rs, 24)]; T16 = ci[WS(rs, 24)]; T13 = W[46]; T15 = W[47]; T17 = FMA(T13, T14, T15 * T16); T2X = FNMS(T15, T14, T13 * T16); } { E T1a, T1c, T19, T1b; T1a = cr[WS(rs, 14)]; T1c = ci[WS(rs, 14)]; T19 = W[26]; T1b = W[27]; T1d = FMA(T19, T1a, T1b * T1c); T2Z = FNMS(T1b, T1a, T19 * T1c); } T2T = T17 - T12; T2U = T1d - T1i; T3c = T2Z - T30; T3b = T2W - T2X; T2Y = T2W + T2X; T31 = T2Z + T30; T34 = T2Y + T31; T18 = T12 + T17; T1j = T1d + T1i; T1k = T18 + T1j; } T1l = TX + T1k; T5X = T33 + T34; { E T2V, T4S, T36, T4R, T32, T35; T2V = FNMS(KP587785252, T2U, KP951056516 * T2T); T4S = FMA(KP587785252, T2T, KP951056516 * T2U); T32 = KP559016994 * (T2Y - T31); T35 = FNMS(KP250000000, T34, T33); T36 = T32 + T35; T4R = T35 - T32; T37 = T2V - T36; T5v = T4S + T4R; T4k = T2V + T36; T4T = T4R - T4S; } { E T3d, T4V, T3a, T4U, T38, T39; T3d = FMA(KP951056516, T3b, KP587785252 * T3c); T4V = FNMS(KP587785252, T3b, KP951056516 * T3c); T38 = KP559016994 * (T18 - T1j); T39 = FNMS(KP250000000, T1k, TX); T3a = T38 + T39; T4U = T39 - T38; T3e = T3a - T3d; T5w = T4U - T4V; T4j = T3a + T3d; T4W = T4U + T4V; } } { E T1r, T3r, T3h, T3i, T3A, T3z, T3m, T3p, T3s, T1C, T1N, T1O; { E T1o, T1q, T1n, T1p; T1o = cr[WS(rs, 2)]; T1q = ci[WS(rs, 2)]; T1n = W[2]; T1p = W[3]; T1r = FMA(T1n, T1o, T1p * T1q); T3r = FNMS(T1p, T1o, T1n * T1q); } { E T1w, T3k, T1M, T3o, T1B, T3l, T1H, T3n; { E T1t, T1v, T1s, T1u; T1t = cr[WS(rs, 7)]; T1v = ci[WS(rs, 7)]; T1s = W[12]; T1u = W[13]; T1w = FMA(T1s, T1t, T1u * T1v); T3k = FNMS(T1u, T1t, T1s * T1v); } { E T1J, T1L, T1I, T1K; T1J = cr[WS(rs, 17)]; T1L = ci[WS(rs, 17)]; T1I = W[32]; T1K = W[33]; T1M = FMA(T1I, T1J, T1K * T1L); T3o = FNMS(T1K, T1J, T1I * T1L); } { E T1y, T1A, T1x, T1z; T1y = cr[WS(rs, 22)]; T1A = ci[WS(rs, 22)]; T1x = W[42]; T1z = W[43]; T1B = FMA(T1x, T1y, T1z * T1A); T3l = FNMS(T1z, T1y, T1x * T1A); } { E T1E, T1G, T1D, T1F; T1E = cr[WS(rs, 12)]; T1G = ci[WS(rs, 12)]; T1D = W[22]; T1F = W[23]; T1H = FMA(T1D, T1E, T1F * T1G); T3n = FNMS(T1F, T1E, T1D * T1G); } T3h = T1w - T1B; T3i = T1H - T1M; T3A = T3n - T3o; T3z = T3k - T3l; T3m = T3k + T3l; T3p = T3n + T3o; T3s = T3m + T3p; T1C = T1w + T1B; T1N = T1H + T1M; T1O = T1C + T1N; } T1P = T1r + T1O; T5Z = T3r + T3s; { E T3j, T53, T3u, T52, T3q, T3t; T3j = FMA(KP951056516, T3h, KP587785252 * T3i); T53 = FNMS(KP587785252, T3h, KP951056516 * T3i); T3q = KP559016994 * (T3m - T3p); T3t = FNMS(KP250000000, T3s, T3r); T3u = T3q + T3t; T52 = T3t - T3q; T3v = T3j + T3u; T5A = T53 + T52; T4o = T3u - T3j; T54 = T52 - T53; } { E T3B, T50, T3y, T4Z, T3w, T3x; T3B = FMA(KP951056516, T3z, KP587785252 * T3A); T50 = FNMS(KP587785252, T3z, KP951056516 * T3A); T3w = KP559016994 * (T1C - T1N); T3x = FNMS(KP250000000, T1O, T1r); T3y = T3w + T3x; T4Z = T3x - T3w; T3C = T3y - T3B; T5z = T4Z - T50; T4n = T3y + T3B; T51 = T4Z + T50; } } { E T62, T64, Tp, T2k, T5T, T5U, T63, T5V; { E T5Y, T61, T1m, T2j; T5Y = T5W - T5X; T61 = T5Z - T60; T62 = FMA(KP951056516, T5Y, KP587785252 * T61); T64 = FNMS(KP587785252, T5Y, KP951056516 * T61); Tp = T1 + To; T1m = TS + T1l; T2j = T1P + T2i; T2k = T1m + T2j; T5T = KP559016994 * (T1m - T2j); T5U = FNMS(KP250000000, T2k, Tp); } cr[0] = Tp + T2k; T63 = T5U - T5T; cr[WS(rs, 10)] = T63 - T64; ci[WS(rs, 9)] = T63 + T64; T5V = T5T + T5U; ci[WS(rs, 4)] = T5V - T62; cr[WS(rs, 5)] = T5V + T62; } { E T2v, T4f, T6I, T6U, T42, T6Z, T43, T6Y, T4A, T6N, T4D, T6L, T4u, T6E, T4v; E T6D, T48, T6V, T4b, T6T, T2n, T6G; T2n = T2l + T2m; T2v = T2n - T2u; T4f = T2n + T2u; T6G = T6g + T6f; T6I = T6G - T6H; T6U = T6H + T6G; { E T2S, T3f, T3g, T3D, T40, T41; T2S = FMA(KP535826794, T2G, KP844327925 * T2R); T3f = FNMS(KP637423989, T3e, KP770513242 * T37); T3g = T2S + T3f; T3D = FNMS(KP425779291, T3C, KP904827052 * T3v); T40 = FNMS(KP992114701, T3Z, KP125333233 * T3S); T41 = T3D + T40; T42 = T3g + T41; T6Z = T3D - T40; T43 = KP559016994 * (T3g - T41); T6Y = T3f - T2S; } { E T4y, T4z, T6J, T4B, T4C, T6K; T4y = FNMS(KP248689887, T4g, KP968583161 * T4h); T4z = FNMS(KP844327925, T4j, KP535826794 * T4k); T6J = T4y + T4z; T4B = FNMS(KP481753674, T4n, KP876306680 * T4o); T4C = FNMS(KP684547105, T4q, KP728968627 * T4r); T6K = T4B + T4C; T4A = T4y - T4z; T6N = KP559016994 * (T6J - T6K); T4D = T4B - T4C; T6L = T6J + T6K; } { E T4i, T4l, T4m, T4p, T4s, T4t; T4i = FMA(KP968583161, T4g, KP248689887 * T4h); T4l = FMA(KP535826794, T4j, KP844327925 * T4k); T4m = T4i + T4l; T4p = FMA(KP876306680, T4n, KP481753674 * T4o); T4s = FMA(KP728968627, T4q, KP684547105 * T4r); T4t = T4p + T4s; T4u = T4m + T4t; T6E = T4p - T4s; T4v = KP559016994 * (T4m - T4t); T6D = T4l - T4i; } { E T46, T47, T6R, T49, T4a, T6S; T46 = FNMS(KP844327925, T2G, KP535826794 * T2R); T47 = FMA(KP770513242, T3e, KP637423989 * T37); T6R = T46 + T47; T49 = FMA(KP125333233, T3Z, KP992114701 * T3S); T4a = FMA(KP904827052, T3C, KP425779291 * T3v); T6S = T4a + T49; T48 = T46 - T47; T6V = T6R - T6S; T4b = T49 - T4a; T6T = KP559016994 * (T6R + T6S); } cr[WS(rs, 4)] = T2v + T42; ci[WS(rs, 23)] = T6L + T6I; ci[WS(rs, 20)] = T6V + T6U; cr[WS(rs, 1)] = T4f + T4u; { E T4c, T4e, T45, T4d, T44; T4c = FMA(KP951056516, T48, KP587785252 * T4b); T4e = FNMS(KP587785252, T48, KP951056516 * T4b); T44 = FNMS(KP250000000, T42, T2v); T45 = T43 + T44; T4d = T44 - T43; ci[0] = T45 - T4c; ci[WS(rs, 5)] = T4d + T4e; cr[WS(rs, 9)] = T45 + T4c; ci[WS(rs, 10)] = T4d - T4e; } { E T6F, T6P, T6O, T6Q, T6M; T6F = FMA(KP587785252, T6D, KP951056516 * T6E); T6P = FNMS(KP587785252, T6E, KP951056516 * T6D); T6M = FNMS(KP250000000, T6L, T6I); T6O = T6M - T6N; T6Q = T6N + T6M; cr[WS(rs, 16)] = T6F - T6O; ci[WS(rs, 18)] = T6P + T6Q; ci[WS(rs, 13)] = T6F + T6O; cr[WS(rs, 21)] = T6P - T6Q; } { E T70, T71, T6X, T72, T6W; T70 = FMA(KP587785252, T6Y, KP951056516 * T6Z); T71 = FNMS(KP587785252, T6Z, KP951056516 * T6Y); T6W = FNMS(KP250000000, T6V, T6U); T6X = T6T - T6W; T72 = T6T + T6W; cr[WS(rs, 14)] = T6X - T70; ci[WS(rs, 15)] = T71 + T72; cr[WS(rs, 19)] = T70 + T6X; cr[WS(rs, 24)] = T71 - T72; } { E T4E, T4G, T4x, T4F, T4w; T4E = FMA(KP951056516, T4A, KP587785252 * T4D); T4G = FNMS(KP587785252, T4A, KP951056516 * T4D); T4w = FNMS(KP250000000, T4u, T4f); T4x = T4v + T4w; T4F = T4w - T4v; ci[WS(rs, 3)] = T4x - T4E; ci[WS(rs, 8)] = T4F + T4G; cr[WS(rs, 6)] = T4x + T4E; cr[WS(rs, 11)] = T4F - T4G; } } { E T75, T7d, T76, T79, T7a, T7b, T7e, T7c; { E T73, T74, T77, T78; T73 = T1l - TS; T74 = T1P - T2i; T75 = FMA(KP587785252, T73, KP951056516 * T74); T7d = FNMS(KP587785252, T74, KP951056516 * T73); T76 = T6e + T6b; T77 = T5W + T5X; T78 = T5Z + T60; T79 = T77 + T78; T7a = FNMS(KP250000000, T79, T76); T7b = KP559016994 * (T77 - T78); } ci[WS(rs, 24)] = T79 + T76; T7e = T7b + T7a; cr[WS(rs, 20)] = T7d - T7e; ci[WS(rs, 19)] = T7d + T7e; T7c = T7a - T7b; cr[WS(rs, 15)] = T75 - T7c; ci[WS(rs, 14)] = T75 + T7c; } { E T4J, T5r, T6i, T6u, T5e, T6z, T5f, T6y, T5M, T6n, T5P, T6l, T5G, T66, T5H; E T65, T5k, T6v, T5n, T6t, T4H, T6h; T4H = T2m - T2l; T4J = T4H + T4I; T5r = T4H - T4I; T6h = T6f - T6g; T6i = T6a + T6h; T6u = T6h - T6a; { E T4Q, T4X, T4Y, T55, T5c, T5d; T4Q = FMA(KP728968627, T4M, KP684547105 * T4P); T4X = FNMS(KP992114701, T4W, KP125333233 * T4T); T4Y = T4Q + T4X; T55 = FMA(KP062790519, T51, KP998026728 * T54); T5c = FNMS(KP637423989, T5b, KP770513242 * T58); T5d = T55 + T5c; T5e = T4Y + T5d; T6z = T55 - T5c; T5f = KP559016994 * (T4Y - T5d); T6y = T4X - T4Q; } { E T5K, T5L, T6j, T5N, T5O, T6k; T5K = FNMS(KP481753674, T5s, KP876306680 * T5t); T5L = FMA(KP904827052, T5w, KP425779291 * T5v); T6j = T5K - T5L; T5N = FNMS(KP844327925, T5z, KP535826794 * T5A); T5O = FNMS(KP998026728, T5C, KP062790519 * T5D); T6k = T5N + T5O; T5M = T5K + T5L; T6n = KP559016994 * (T6j - T6k); T5P = T5N - T5O; T6l = T6j + T6k; } { E T5u, T5x, T5y, T5B, T5E, T5F; T5u = FMA(KP876306680, T5s, KP481753674 * T5t); T5x = FNMS(KP425779291, T5w, KP904827052 * T5v); T5y = T5u + T5x; T5B = FMA(KP535826794, T5z, KP844327925 * T5A); T5E = FMA(KP062790519, T5C, KP998026728 * T5D); T5F = T5B + T5E; T5G = T5y + T5F; T66 = T5B - T5E; T5H = KP559016994 * (T5y - T5F); T65 = T5x - T5u; } { E T5i, T5j, T6r, T5l, T5m, T6s; T5i = FNMS(KP684547105, T4M, KP728968627 * T4P); T5j = FMA(KP125333233, T4W, KP992114701 * T4T); T6r = T5i - T5j; T5l = FNMS(KP998026728, T51, KP062790519 * T54); T5m = FMA(KP770513242, T5b, KP637423989 * T58); T6s = T5l - T5m; T5k = T5i + T5j; T6v = T6r + T6s; T5n = T5l + T5m; T6t = KP559016994 * (T6r - T6s); } cr[WS(rs, 3)] = T4J + T5e; ci[WS(rs, 22)] = T6l + T6i; ci[WS(rs, 21)] = T6v + T6u; cr[WS(rs, 2)] = T5r + T5G; { E T67, T6p, T6o, T6q, T6m; T67 = FMA(KP587785252, T65, KP951056516 * T66); T6p = FNMS(KP587785252, T66, KP951056516 * T65); T6m = FNMS(KP250000000, T6l, T6i); T6o = T6m - T6n; T6q = T6n + T6m; cr[WS(rs, 17)] = T67 - T6o; ci[WS(rs, 17)] = T6p + T6q; ci[WS(rs, 12)] = T67 + T6o; cr[WS(rs, 22)] = T6p - T6q; } { E T5Q, T5S, T5J, T5R, T5I; T5Q = FMA(KP951056516, T5M, KP587785252 * T5P); T5S = FNMS(KP587785252, T5M, KP951056516 * T5P); T5I = FNMS(KP250000000, T5G, T5r); T5J = T5H + T5I; T5R = T5I - T5H; ci[WS(rs, 2)] = T5J - T5Q; ci[WS(rs, 7)] = T5R + T5S; cr[WS(rs, 7)] = T5J + T5Q; cr[WS(rs, 12)] = T5R - T5S; } { E T5o, T5q, T5h, T5p, T5g; T5o = FMA(KP951056516, T5k, KP587785252 * T5n); T5q = FNMS(KP587785252, T5k, KP951056516 * T5n); T5g = FNMS(KP250000000, T5e, T4J); T5h = T5f + T5g; T5p = T5g - T5f; ci[WS(rs, 1)] = T5h - T5o; ci[WS(rs, 6)] = T5p + T5q; cr[WS(rs, 8)] = T5h + T5o; ci[WS(rs, 11)] = T5p - T5q; } { E T6A, T6B, T6x, T6C, T6w; T6A = FMA(KP587785252, T6y, KP951056516 * T6z); T6B = FNMS(KP587785252, T6z, KP951056516 * T6y); T6w = FNMS(KP250000000, T6v, T6u); T6x = T6t - T6w; T6C = T6t + T6w; cr[WS(rs, 13)] = T6x - T6A; ci[WS(rs, 16)] = T6B + T6C; cr[WS(rs, 18)] = T6A + T6x; cr[WS(rs, 23)] = T6B - T6C; } } } } } static const tw_instr twinstr[] = { {TW_FULL, 1, 25}, {TW_NEXT, 1, 0} }; static const hc2hc_desc desc = { 25, "hf_25", twinstr, &GENUS, {260, 140, 140, 0} }; void X(codelet_hf_25) (planner *p) { X(khc2hc_register) (p, hf_25, &desc); } #endif /* HAVE_FMA */ fftw-3.3.3/rdft/scalar/r2cf/hf2_4.c0000644000175000017500000001226312054410642013517 00000000000000/* * Copyright (c) 2003, 2007-11 Matteo Frigo * Copyright (c) 2003, 2007-11 Massachusetts Institute of Technology * * This program is free software; you can redistribute it and/or modify * it under the terms of the GNU General Public License as published by * the Free Software Foundation; either version 2 of the License, or * (at your option) any later version. * * This program is distributed in the hope that it will be useful, * but WITHOUT ANY WARRANTY; without even the implied warranty of * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the * GNU General Public License for more details. * * You should have received a copy of the GNU General Public License * along with this program; if not, write to the Free Software * Foundation, Inc., 51 Franklin Street, Fifth Floor, Boston, MA 02110-1301 USA * */ /* This file was automatically generated --- DO NOT EDIT */ /* Generated on Sun Nov 25 07:40:02 EST 2012 */ #include "codelet-rdft.h" #ifdef HAVE_FMA /* Generated by: ../../../genfft/gen_hc2hc.native -fma -reorder-insns -schedule-for-pipeline -compact -variables 4 -pipeline-latency 4 -twiddle-log3 -precompute-twiddles -n 4 -dit -name hf2_4 -include hf.h */ /* * This function contains 24 FP additions, 16 FP multiplications, * (or, 16 additions, 8 multiplications, 8 fused multiply/add), * 33 stack variables, 0 constants, and 16 memory accesses */ #include "hf.h" static void hf2_4(R *cr, R *ci, const R *W, stride rs, INT mb, INT me, INT ms) { { INT m; for (m = mb, W = W + ((mb - 1) * 4); m < me; m = m + 1, cr = cr + ms, ci = ci - ms, W = W + 4, MAKE_VOLATILE_STRIDE(8, rs)) { E Ti, Tq, To, Te, TA, Ty, Tm, Ts; { E T2, T6, T3, T5; T2 = W[0]; T6 = W[3]; T3 = W[2]; T5 = W[1]; { E T1, Tx, Td, Tw, Tj, Tl, Ta, T4, Tk, Tr; T1 = cr[0]; Ta = T2 * T6; T4 = T2 * T3; Tx = ci[0]; { E T8, Tb, T7, Tc; T8 = cr[WS(rs, 2)]; Tb = FNMS(T5, T3, Ta); T7 = FMA(T5, T6, T4); Tc = ci[WS(rs, 2)]; { E Tf, Th, T9, Tv, Tg, Tp; Tf = cr[WS(rs, 1)]; Th = ci[WS(rs, 1)]; T9 = T7 * T8; Tv = T7 * Tc; Tg = T2 * Tf; Tp = T2 * Th; Td = FMA(Tb, Tc, T9); Tw = FNMS(Tb, T8, Tv); Ti = FMA(T5, Th, Tg); Tq = FNMS(T5, Tf, Tp); } Tj = cr[WS(rs, 3)]; Tl = ci[WS(rs, 3)]; } To = T1 - Td; Te = T1 + Td; Tk = T3 * Tj; Tr = T3 * Tl; TA = Tx - Tw; Ty = Tw + Tx; Tm = FMA(T6, Tl, Tk); Ts = FNMS(T6, Tj, Tr); } } { E Tn, Tz, Tt, Tu; Tn = Ti + Tm; Tz = Tm - Ti; Tt = Tq - Ts; Tu = Tq + Ts; ci[WS(rs, 2)] = Tz + TA; cr[WS(rs, 3)] = Tz - TA; cr[0] = Te + Tn; ci[WS(rs, 1)] = Te - Tn; ci[WS(rs, 3)] = Tu + Ty; cr[WS(rs, 2)] = Tu - Ty; cr[WS(rs, 1)] = To + Tt; ci[0] = To - Tt; } } } } static const tw_instr twinstr[] = { {TW_CEXP, 1, 1}, {TW_CEXP, 1, 3}, {TW_NEXT, 1, 0} }; static const hc2hc_desc desc = { 4, "hf2_4", twinstr, &GENUS, {16, 8, 8, 0} }; void X(codelet_hf2_4) (planner *p) { X(khc2hc_register) (p, hf2_4, &desc); } #else /* HAVE_FMA */ /* Generated by: ../../../genfft/gen_hc2hc.native -compact -variables 4 -pipeline-latency 4 -twiddle-log3 -precompute-twiddles -n 4 -dit -name hf2_4 -include hf.h */ /* * This function contains 24 FP additions, 16 FP multiplications, * (or, 16 additions, 8 multiplications, 8 fused multiply/add), * 21 stack variables, 0 constants, and 16 memory accesses */ #include "hf.h" static void hf2_4(R *cr, R *ci, const R *W, stride rs, INT mb, INT me, INT ms) { { INT m; for (m = mb, W = W + ((mb - 1) * 4); m < me; m = m + 1, cr = cr + ms, ci = ci - ms, W = W + 4, MAKE_VOLATILE_STRIDE(8, rs)) { E T2, T4, T3, T5, T6, T8; T2 = W[0]; T4 = W[1]; T3 = W[2]; T5 = W[3]; T6 = FMA(T2, T3, T4 * T5); T8 = FNMS(T4, T3, T2 * T5); { E T1, Tp, Ta, To, Te, Tk, Th, Tl, T7, T9; T1 = cr[0]; Tp = ci[0]; T7 = cr[WS(rs, 2)]; T9 = ci[WS(rs, 2)]; Ta = FMA(T6, T7, T8 * T9); To = FNMS(T8, T7, T6 * T9); { E Tc, Td, Tf, Tg; Tc = cr[WS(rs, 1)]; Td = ci[WS(rs, 1)]; Te = FMA(T2, Tc, T4 * Td); Tk = FNMS(T4, Tc, T2 * Td); Tf = cr[WS(rs, 3)]; Tg = ci[WS(rs, 3)]; Th = FMA(T3, Tf, T5 * Tg); Tl = FNMS(T5, Tf, T3 * Tg); } { E Tb, Ti, Tj, Tm; Tb = T1 + Ta; Ti = Te + Th; ci[WS(rs, 1)] = Tb - Ti; cr[0] = Tb + Ti; Tj = T1 - Ta; Tm = Tk - Tl; ci[0] = Tj - Tm; cr[WS(rs, 1)] = Tj + Tm; } { E Tn, Tq, Tr, Ts; Tn = Tk + Tl; Tq = To + Tp; cr[WS(rs, 2)] = Tn - Tq; ci[WS(rs, 3)] = Tn + Tq; Tr = Th - Te; Ts = Tp - To; cr[WS(rs, 3)] = Tr - Ts; ci[WS(rs, 2)] = Tr + Ts; } } } } } static const tw_instr twinstr[] = { {TW_CEXP, 1, 1}, {TW_CEXP, 1, 3}, {TW_NEXT, 1, 0} }; static const hc2hc_desc desc = { 4, "hf2_4", twinstr, &GENUS, {16, 8, 8, 0} }; void X(codelet_hf2_4) (planner *p) { X(khc2hc_register) (p, hf2_4, &desc); } #endif /* HAVE_FMA */ fftw-3.3.3/rdft/scalar/r2cf/hf2_8.c0000644000175000017500000002415112054410642013522 00000000000000/* * Copyright (c) 2003, 2007-11 Matteo Frigo * Copyright (c) 2003, 2007-11 Massachusetts Institute of Technology * * This program is free software; you can redistribute it and/or modify * it under the terms of the GNU General Public License as published by * the Free Software Foundation; either version 2 of the License, or * (at your option) any later version. * * This program is distributed in the hope that it will be useful, * but WITHOUT ANY WARRANTY; without even the implied warranty of * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the * GNU General Public License for more details. * * You should have received a copy of the GNU General Public License * along with this program; if not, write to the Free Software * Foundation, Inc., 51 Franklin Street, Fifth Floor, Boston, MA 02110-1301 USA * */ /* This file was automatically generated --- DO NOT EDIT */ /* Generated on Sun Nov 25 07:40:02 EST 2012 */ #include "codelet-rdft.h" #ifdef HAVE_FMA /* Generated by: ../../../genfft/gen_hc2hc.native -fma -reorder-insns -schedule-for-pipeline -compact -variables 4 -pipeline-latency 4 -twiddle-log3 -precompute-twiddles -n 8 -dit -name hf2_8 -include hf.h */ /* * This function contains 74 FP additions, 50 FP multiplications, * (or, 44 additions, 20 multiplications, 30 fused multiply/add), * 64 stack variables, 1 constants, and 32 memory accesses */ #include "hf.h" static void hf2_8(R *cr, R *ci, const R *W, stride rs, INT mb, INT me, INT ms) { DK(KP707106781, +0.707106781186547524400844362104849039284835938); { INT m; for (m = mb, W = W + ((mb - 1) * 6); m < me; m = m + 1, cr = cr + ms, ci = ci - ms, W = W + 6, MAKE_VOLATILE_STRIDE(16, rs)) { E TS, T1l, TJ, T1m, T1k, Tw, T1w, T1u; { E T2, T3, Tl, Tn, T5, T4, Tm, Tr, T6; T2 = W[0]; T3 = W[2]; Tl = W[4]; Tn = W[5]; T5 = W[1]; T4 = T2 * T3; Tm = T2 * Tl; Tr = T2 * Tn; T6 = W[3]; { E T1, T1s, TG, Td, T1r, Tu, TY, Tk, TW, T18, T1d, TD, TH, TA, T13; E TE, T14; { E To, Ts, Tf, T7, T8, Ti, Tb, T9, Tc, TC, Ta, TF, TB, Tg, Th; E Tj; T1 = cr[0]; To = FMA(T5, Tn, Tm); Ts = FNMS(T5, Tl, Tr); Tf = FMA(T5, T6, T4); T7 = FNMS(T5, T6, T4); Ta = T2 * T6; T1s = ci[0]; T8 = cr[WS(rs, 4)]; TF = Tf * Tn; TB = Tf * Tl; Ti = FNMS(T5, T3, Ta); Tb = FMA(T5, T3, Ta); T9 = T7 * T8; Tc = ci[WS(rs, 4)]; TG = FNMS(Ti, Tl, TF); TC = FMA(Ti, Tn, TB); { E Tp, T1q, Tt, Tq, TX; Tp = cr[WS(rs, 6)]; Td = FMA(Tb, Tc, T9); T1q = T7 * Tc; Tt = ci[WS(rs, 6)]; Tq = To * Tp; Tg = cr[WS(rs, 2)]; T1r = FNMS(Tb, T8, T1q); TX = To * Tt; Tu = FMA(Ts, Tt, Tq); Th = Tf * Tg; Tj = ci[WS(rs, 2)]; TY = FNMS(Ts, Tp, TX); } { E TO, TQ, TN, TP, T1a, T1b; { E TK, TM, TL, T19, TV; TK = cr[WS(rs, 7)]; TM = ci[WS(rs, 7)]; Tk = FMA(Ti, Tj, Th); TV = Tf * Tj; TL = Tl * TK; T19 = Tl * TM; TO = cr[WS(rs, 3)]; TW = FNMS(Ti, Tg, TV); TQ = ci[WS(rs, 3)]; TN = FMA(Tn, TM, TL); TP = T3 * TO; T1a = FNMS(Tn, TK, T19); T1b = T3 * TQ; } { E Tx, Tz, Ty, T12, T1c, TR; Tx = cr[WS(rs, 1)]; TR = FMA(T6, TQ, TP); Tz = ci[WS(rs, 1)]; T1c = FNMS(T6, TO, T1b); Ty = T2 * Tx; T18 = TN - TR; TS = TN + TR; T12 = T2 * Tz; T1d = T1a - T1c; T1l = T1a + T1c; TD = cr[WS(rs, 5)]; TH = ci[WS(rs, 5)]; TA = FMA(T5, Tz, Ty); T13 = FNMS(T5, Tx, T12); TE = TC * TD; T14 = TC * TH; } } } { E Te, T1p, Tv, T1t; { E T1g, T10, T1z, T1B, T1C, T1j, T1A, T1f; { E T1x, T11, T16, T1y; { E TU, TZ, TI, T15; Te = T1 + Td; TU = T1 - Td; TZ = TW - TY; T1p = TW + TY; TI = FMA(TG, TH, TE); T15 = FNMS(TG, TD, T14); Tv = Tk + Tu; T1x = Tk - Tu; T1g = TU - TZ; T10 = TU + TZ; T11 = TA - TI; TJ = TA + TI; T1m = T13 + T15; T16 = T13 - T15; T1y = T1s - T1r; T1t = T1r + T1s; } { E T1i, T1e, T17, T1h; T1i = T18 + T1d; T1e = T18 - T1d; T17 = T11 + T16; T1h = T11 - T16; T1z = T1x + T1y; T1B = T1y - T1x; T1C = T1i - T1h; T1j = T1h + T1i; T1A = T1e - T17; T1f = T17 + T1e; } } cr[WS(rs, 3)] = FNMS(KP707106781, T1j, T1g); cr[WS(rs, 7)] = FMS(KP707106781, T1A, T1z); cr[WS(rs, 1)] = FMA(KP707106781, T1f, T10); ci[WS(rs, 2)] = FNMS(KP707106781, T1f, T10); ci[WS(rs, 6)] = FMA(KP707106781, T1C, T1B); cr[WS(rs, 5)] = FMS(KP707106781, T1C, T1B); ci[WS(rs, 4)] = FMA(KP707106781, T1A, T1z); ci[0] = FMA(KP707106781, T1j, T1g); } T1k = Te - Tv; Tw = Te + Tv; T1w = T1t - T1p; T1u = T1p + T1t; } } } { E TT, T1v, T1n, T1o; TT = TJ + TS; T1v = TS - TJ; T1n = T1l - T1m; T1o = T1m + T1l; ci[WS(rs, 5)] = T1v + T1w; cr[WS(rs, 6)] = T1v - T1w; cr[0] = Tw + TT; ci[WS(rs, 3)] = Tw - TT; ci[WS(rs, 7)] = T1o + T1u; cr[WS(rs, 4)] = T1o - T1u; ci[WS(rs, 1)] = T1k + T1n; cr[WS(rs, 2)] = T1k - T1n; } } } } static const tw_instr twinstr[] = { {TW_CEXP, 1, 1}, {TW_CEXP, 1, 3}, {TW_CEXP, 1, 7}, {TW_NEXT, 1, 0} }; static const hc2hc_desc desc = { 8, "hf2_8", twinstr, &GENUS, {44, 20, 30, 0} }; void X(codelet_hf2_8) (planner *p) { X(khc2hc_register) (p, hf2_8, &desc); } #else /* HAVE_FMA */ /* Generated by: ../../../genfft/gen_hc2hc.native -compact -variables 4 -pipeline-latency 4 -twiddle-log3 -precompute-twiddles -n 8 -dit -name hf2_8 -include hf.h */ /* * This function contains 74 FP additions, 44 FP multiplications, * (or, 56 additions, 26 multiplications, 18 fused multiply/add), * 42 stack variables, 1 constants, and 32 memory accesses */ #include "hf.h" static void hf2_8(R *cr, R *ci, const R *W, stride rs, INT mb, INT me, INT ms) { DK(KP707106781, +0.707106781186547524400844362104849039284835938); { INT m; for (m = mb, W = W + ((mb - 1) * 6); m < me; m = m + 1, cr = cr + ms, ci = ci - ms, W = W + 6, MAKE_VOLATILE_STRIDE(16, rs)) { E T2, T5, T3, T6, T8, Tc, Tg, Ti, Tl, Tm, Tn, Tz, Tp, Tx; { E T4, Tb, T7, Ta; T2 = W[0]; T5 = W[1]; T3 = W[2]; T6 = W[3]; T4 = T2 * T3; Tb = T5 * T3; T7 = T5 * T6; Ta = T2 * T6; T8 = T4 - T7; Tc = Ta + Tb; Tg = T4 + T7; Ti = Ta - Tb; Tl = W[4]; Tm = W[5]; Tn = FMA(T2, Tl, T5 * Tm); Tz = FNMS(Ti, Tl, Tg * Tm); Tp = FNMS(T5, Tl, T2 * Tm); Tx = FMA(Tg, Tl, Ti * Tm); } { E Tf, T1j, TL, T1d, TJ, T16, TV, TY, Ts, T1i, TO, T1a, TC, T17, TQ; E TT; { E T1, T1c, Te, T1b, T9, Td; T1 = cr[0]; T1c = ci[0]; T9 = cr[WS(rs, 4)]; Td = ci[WS(rs, 4)]; Te = FMA(T8, T9, Tc * Td); T1b = FNMS(Tc, T9, T8 * Td); Tf = T1 + Te; T1j = T1c - T1b; TL = T1 - Te; T1d = T1b + T1c; } { E TF, TW, TI, TX; { E TD, TE, TG, TH; TD = cr[WS(rs, 7)]; TE = ci[WS(rs, 7)]; TF = FMA(Tl, TD, Tm * TE); TW = FNMS(Tm, TD, Tl * TE); TG = cr[WS(rs, 3)]; TH = ci[WS(rs, 3)]; TI = FMA(T3, TG, T6 * TH); TX = FNMS(T6, TG, T3 * TH); } TJ = TF + TI; T16 = TW + TX; TV = TF - TI; TY = TW - TX; } { E Tk, TM, Tr, TN; { E Th, Tj, To, Tq; Th = cr[WS(rs, 2)]; Tj = ci[WS(rs, 2)]; Tk = FMA(Tg, Th, Ti * Tj); TM = FNMS(Ti, Th, Tg * Tj); To = cr[WS(rs, 6)]; Tq = ci[WS(rs, 6)]; Tr = FMA(Tn, To, Tp * Tq); TN = FNMS(Tp, To, Tn * Tq); } Ts = Tk + Tr; T1i = Tk - Tr; TO = TM - TN; T1a = TM + TN; } { E Tw, TR, TB, TS; { E Tu, Tv, Ty, TA; Tu = cr[WS(rs, 1)]; Tv = ci[WS(rs, 1)]; Tw = FMA(T2, Tu, T5 * Tv); TR = FNMS(T5, Tu, T2 * Tv); Ty = cr[WS(rs, 5)]; TA = ci[WS(rs, 5)]; TB = FMA(Tx, Ty, Tz * TA); TS = FNMS(Tz, Ty, Tx * TA); } TC = Tw + TB; T17 = TR + TS; TQ = Tw - TB; TT = TR - TS; } { E Tt, TK, T1f, T1g; Tt = Tf + Ts; TK = TC + TJ; ci[WS(rs, 3)] = Tt - TK; cr[0] = Tt + TK; T1f = TJ - TC; T1g = T1d - T1a; cr[WS(rs, 6)] = T1f - T1g; ci[WS(rs, 5)] = T1f + T1g; { E T11, T1m, T14, T1l, T12, T13; T11 = TL - TO; T1m = T1j - T1i; T12 = TQ - TT; T13 = TV + TY; T14 = KP707106781 * (T12 + T13); T1l = KP707106781 * (T13 - T12); cr[WS(rs, 3)] = T11 - T14; ci[WS(rs, 6)] = T1l + T1m; ci[0] = T11 + T14; cr[WS(rs, 5)] = T1l - T1m; } } { E T19, T1e, T15, T18; T19 = T17 + T16; T1e = T1a + T1d; cr[WS(rs, 4)] = T19 - T1e; ci[WS(rs, 7)] = T19 + T1e; T15 = Tf - Ts; T18 = T16 - T17; cr[WS(rs, 2)] = T15 - T18; ci[WS(rs, 1)] = T15 + T18; { E TP, T1k, T10, T1h, TU, TZ; TP = TL + TO; T1k = T1i + T1j; TU = TQ + TT; TZ = TV - TY; T10 = KP707106781 * (TU + TZ); T1h = KP707106781 * (TZ - TU); ci[WS(rs, 2)] = TP - T10; ci[WS(rs, 4)] = T1h + T1k; cr[WS(rs, 1)] = TP + T10; cr[WS(rs, 7)] = T1h - T1k; } } } } } } static const tw_instr twinstr[] = { {TW_CEXP, 1, 1}, {TW_CEXP, 1, 3}, {TW_CEXP, 1, 7}, {TW_NEXT, 1, 0} }; static const hc2hc_desc desc = { 8, "hf2_8", twinstr, &GENUS, {56, 26, 18, 0} }; void X(codelet_hf2_8) (planner *p) { X(khc2hc_register) (p, hf2_8, &desc); } #endif /* HAVE_FMA */ fftw-3.3.3/rdft/scalar/r2cf/hf2_16.c0000644000175000017500000005463512054410644013615 00000000000000/* * Copyright (c) 2003, 2007-11 Matteo Frigo * Copyright (c) 2003, 2007-11 Massachusetts Institute of Technology * * This program is free software; you can redistribute it and/or modify * it under the terms of the GNU General Public License as published by * the Free Software Foundation; either version 2 of the License, or * (at your option) any later version. * * This program is distributed in the hope that it will be useful, * but WITHOUT ANY WARRANTY; without even the implied warranty of * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the * GNU General Public License for more details. * * You should have received a copy of the GNU General Public License * along with this program; if not, write to the Free Software * Foundation, Inc., 51 Franklin Street, Fifth Floor, Boston, MA 02110-1301 USA * */ /* This file was automatically generated --- DO NOT EDIT */ /* Generated on Sun Nov 25 07:40:02 EST 2012 */ #include "codelet-rdft.h" #ifdef HAVE_FMA /* Generated by: ../../../genfft/gen_hc2hc.native -fma -reorder-insns -schedule-for-pipeline -compact -variables 4 -pipeline-latency 4 -twiddle-log3 -precompute-twiddles -n 16 -dit -name hf2_16 -include hf.h */ /* * This function contains 196 FP additions, 134 FP multiplications, * (or, 104 additions, 42 multiplications, 92 fused multiply/add), * 106 stack variables, 3 constants, and 64 memory accesses */ #include "hf.h" static void hf2_16(R *cr, R *ci, const R *W, stride rs, INT mb, INT me, INT ms) { DK(KP923879532, +0.923879532511286756128183189396788286822416626); DK(KP414213562, +0.414213562373095048801688724209698078569671875); DK(KP707106781, +0.707106781186547524400844362104849039284835938); { INT m; for (m = mb, W = W + ((mb - 1) * 8); m < me; m = m + 1, cr = cr + ms, ci = ci - ms, W = W + 8, MAKE_VOLATILE_STRIDE(32, rs)) { E T35, T32; { E T2, Tf, TM, TO, T3, Tg, TN, TS, T4, Tp, T6, T5, Th; T2 = W[0]; Tf = W[2]; TM = W[6]; TO = W[7]; T3 = W[4]; Tg = T2 * Tf; TN = T2 * TM; TS = T2 * TO; T4 = T2 * T3; Tp = Tf * T3; T6 = W[5]; T5 = W[1]; Th = W[3]; { E TZ, Te, T1U, T3A, T3M, T2w, T1G, T2I, T3h, T1R, T2D, T2B, T3i, Tx, T3L; E T1Z, T3w, TL, T21, T26, T38, T1d, T2h, T2s, T3c, T1s, T2t, T2m, T3d, TX; E T10, TV, T2a, TY, T2b; { E TF, TP, TT, Tq, TW, Tz, Tu, TI, TC, T1m, T1f, T1p, T1j, Tr, Ts; E Tv, To, T1W; { E Ti, Tm, T1L, T1O, T1D, T1A, T1x, T2G, T1F, T2F; { E T1, T7, Tb, T3z, T8, T1z, T9, Tc; { E T1i, T1e, T1C, T1y, Tt, Ta, Tl; T1 = cr[0]; Tt = Tf * T6; Ta = T2 * T6; T7 = FMA(T5, T6, T4); TF = FNMS(T5, T6, T4); TP = FMA(T5, TO, TN); TT = FNMS(T5, TM, TS); Tq = FNMS(Th, T6, Tp); TW = FMA(Th, T6, Tp); Tz = FMA(T5, Th, Tg); Ti = FNMS(T5, Th, Tg); Tl = T2 * Th; Tu = FMA(Th, T3, Tt); TZ = FNMS(Th, T3, Tt); TI = FMA(T5, T3, Ta); Tb = FNMS(T5, T3, Ta); T1i = Ti * T6; T1e = Ti * T3; T1C = Tz * T6; T1y = Tz * T3; Tm = FMA(T5, Tf, Tl); TC = FNMS(T5, Tf, Tl); T3z = ci[0]; T8 = cr[WS(rs, 8)]; T1m = FNMS(Tm, T6, T1e); T1f = FMA(Tm, T6, T1e); T1p = FMA(Tm, T3, T1i); T1j = FNMS(Tm, T3, T1i); T1L = FNMS(TC, T6, T1y); T1z = FMA(TC, T6, T1y); T1O = FMA(TC, T3, T1C); T1D = FNMS(TC, T3, T1C); T9 = T7 * T8; Tc = ci[WS(rs, 8)]; } { E T1u, T1w, T1v, T2E, T3y, T1B, T1E, Td, T3x; T1u = cr[WS(rs, 15)]; T1w = ci[WS(rs, 15)]; T1A = cr[WS(rs, 7)]; Td = FMA(Tb, Tc, T9); T3x = T7 * Tc; T1v = TM * T1u; T2E = TM * T1w; Te = T1 + Td; T1U = T1 - Td; T3y = FNMS(Tb, T8, T3x); T1B = T1z * T1A; T1E = ci[WS(rs, 7)]; T1x = FMA(TO, T1w, T1v); T3A = T3y + T3z; T3M = T3z - T3y; T2G = T1z * T1E; T1F = FMA(T1D, T1E, T1B); T2F = FNMS(TO, T1u, T2E); } } { E T1H, T1I, T1J, T1M, T1P, T2H; T1H = cr[WS(rs, 3)]; T2H = FNMS(T1D, T1A, T2G); T2w = T1x - T1F; T1G = T1x + T1F; T1I = Tf * T1H; T2I = T2F - T2H; T3h = T2F + T2H; T1J = ci[WS(rs, 3)]; T1M = cr[WS(rs, 11)]; T1P = ci[WS(rs, 11)]; { E Tj, Tk, Tn, T1V; { E T1K, T2y, T1Q, T2A, T2x, T1N, T2z; Tj = cr[WS(rs, 4)]; T1K = FMA(Th, T1J, T1I); T2x = Tf * T1J; T1N = T1L * T1M; T2z = T1L * T1P; Tk = Ti * Tj; T2y = FNMS(Th, T1H, T2x); T1Q = FMA(T1O, T1P, T1N); T2A = FNMS(T1O, T1M, T2z); Tn = ci[WS(rs, 4)]; Tr = cr[WS(rs, 12)]; T1R = T1K + T1Q; T2D = T1Q - T1K; T2B = T2y - T2A; T3i = T2y + T2A; T1V = Ti * Tn; Ts = Tq * Tr; Tv = ci[WS(rs, 12)]; } To = FMA(Tm, Tn, Tk); T1W = FNMS(Tm, Tj, T1V); } } } { E T19, T1b, T18, T2p, T1a, T2q; { E TE, T23, TK, T25; { E TA, TD, TB, T22, TG, TJ, TH, T24, T1Y, Tw, T1X; TA = cr[WS(rs, 2)]; Tw = FMA(Tu, Tv, Ts); T1X = Tq * Tv; TD = ci[WS(rs, 2)]; TB = Tz * TA; Tx = To + Tw; T3L = To - Tw; T1Y = FNMS(Tu, Tr, T1X); T22 = Tz * TD; TG = cr[WS(rs, 10)]; TJ = ci[WS(rs, 10)]; T1Z = T1W - T1Y; T3w = T1W + T1Y; TH = TF * TG; T24 = TF * TJ; TE = FMA(TC, TD, TB); T23 = FNMS(TC, TA, T22); TK = FMA(TI, TJ, TH); T25 = FNMS(TI, TG, T24); } { E T15, T17, T16, T2o; T15 = cr[WS(rs, 1)]; T17 = ci[WS(rs, 1)]; TL = TE + TK; T21 = TE - TK; T26 = T23 - T25; T38 = T23 + T25; T16 = T2 * T15; T2o = T2 * T17; T19 = cr[WS(rs, 9)]; T1b = ci[WS(rs, 9)]; T18 = FMA(T5, T17, T16); T2p = FNMS(T5, T15, T2o); T1a = T3 * T19; T2q = T3 * T1b; } } { E T1n, T1q, T1l, T2j, T1o, T2k; { E T1g, T1k, T1h, T2i, T1c, T2r; T1g = cr[WS(rs, 5)]; T1k = ci[WS(rs, 5)]; T1c = FMA(T6, T1b, T1a); T2r = FNMS(T6, T19, T2q); T1h = T1f * T1g; T2i = T1f * T1k; T1d = T18 + T1c; T2h = T18 - T1c; T2s = T2p - T2r; T3c = T2p + T2r; T1n = cr[WS(rs, 13)]; T1q = ci[WS(rs, 13)]; T1l = FMA(T1j, T1k, T1h); T2j = FNMS(T1j, T1g, T2i); T1o = T1m * T1n; T2k = T1m * T1q; } { E TQ, TU, TR, T29, T1r, T2l; TQ = cr[WS(rs, 14)]; TU = ci[WS(rs, 14)]; T1r = FMA(T1p, T1q, T1o); T2l = FNMS(T1p, T1n, T2k); TR = TP * TQ; T29 = TP * TU; T1s = T1l + T1r; T2t = T1l - T1r; T2m = T2j - T2l; T3d = T2j + T2l; TX = cr[WS(rs, 6)]; T10 = ci[WS(rs, 6)]; TV = FMA(TT, TU, TR); T2a = FNMS(TT, TQ, T29); TY = TW * TX; T2b = TW * T10; } } } } { E T36, T3G, T3b, T3g, T28, T2d, T3F, T39, T3j, T3q, T3C, T3e, T3u, T3t; { E T3D, T1T, T3r, T14, T3E, T3s; { E Ty, T3B, T11, T2c, T13, T3v; T36 = Te - Tx; Ty = Te + Tx; T3B = T3w + T3A; T3G = T3A - T3w; T11 = FMA(TZ, T10, TY); T2c = FNMS(TZ, TX, T2b); { E T1t, T1S, T12, T37; T3b = T1d - T1s; T1t = T1d + T1s; T1S = T1G + T1R; T3g = T1G - T1R; T12 = TV + T11; T28 = TV - T11; T2d = T2a - T2c; T37 = T2a + T2c; T3D = T1S - T1t; T1T = T1t + T1S; T13 = TL + T12; T3F = TL - T12; T39 = T37 - T38; T3v = T38 + T37; } T3j = T3h - T3i; T3r = T3h + T3i; T3q = Ty - T13; T14 = Ty + T13; T3E = T3B - T3v; T3C = T3v + T3B; T3s = T3c + T3d; T3e = T3c - T3d; } ci[WS(rs, 7)] = T14 - T1T; cr[WS(rs, 12)] = T3D - T3E; ci[WS(rs, 11)] = T3D + T3E; T3u = T3s + T3r; T3t = T3r - T3s; cr[0] = T14 + T1T; } { E T3m, T3a, T3J, T3H; ci[WS(rs, 15)] = T3u + T3C; cr[WS(rs, 8)] = T3u - T3C; ci[WS(rs, 3)] = T3q + T3t; cr[WS(rs, 4)] = T3q - T3t; T3m = T36 + T39; T3a = T36 - T39; T3J = T3G - T3F; T3H = T3F + T3G; { E T2Q, T20, T3N, T3T, T2C, T2J, T3U, T2f, T33, T30, T2V, T2W, T3O, T2T, T2N; E T2v; { E T2R, T27, T2e, T2S; { E T3n, T3f, T3o, T3k; T2Q = T1U + T1Z; T20 = T1U - T1Z; T3n = T3b - T3e; T3f = T3b + T3e; T3o = T3g + T3j; T3k = T3g - T3j; T3N = T3L + T3M; T3T = T3M - T3L; { E T3p, T3K, T3I, T3l; T3p = T3n + T3o; T3K = T3o - T3n; T3I = T3k - T3f; T3l = T3f + T3k; ci[WS(rs, 1)] = FMA(KP707106781, T3p, T3m); cr[WS(rs, 6)] = FNMS(KP707106781, T3p, T3m); ci[WS(rs, 13)] = FMA(KP707106781, T3K, T3J); cr[WS(rs, 10)] = FMS(KP707106781, T3K, T3J); ci[WS(rs, 9)] = FMA(KP707106781, T3I, T3H); cr[WS(rs, 14)] = FMS(KP707106781, T3I, T3H); cr[WS(rs, 2)] = FMA(KP707106781, T3l, T3a); ci[WS(rs, 5)] = FNMS(KP707106781, T3l, T3a); T2R = T21 + T26; T27 = T21 - T26; T2e = T28 + T2d; T2S = T28 - T2d; } } { E T2Y, T2Z, T2n, T2u; T2C = T2w - T2B; T2Y = T2w + T2B; T2Z = T2I + T2D; T2J = T2D - T2I; T3U = T2e - T27; T2f = T27 + T2e; T33 = FMA(KP414213562, T2Y, T2Z); T30 = FNMS(KP414213562, T2Z, T2Y); T2V = T2h + T2m; T2n = T2h - T2m; T2u = T2s + T2t; T2W = T2s - T2t; T3O = T2R - T2S; T2T = T2R + T2S; T2N = FMA(KP414213562, T2n, T2u); T2v = FNMS(KP414213562, T2u, T2n); } } { E T2M, T3S, T31, T2P, T3Q, T3R, T3P, T2U; { E T2g, T2X, T2O, T2K, T3V, T3X, T3W, T34, T2L, T3Y; T2M = FNMS(KP707106781, T2f, T20); T2g = FMA(KP707106781, T2f, T20); T34 = FNMS(KP414213562, T2V, T2W); T2X = FMA(KP414213562, T2W, T2V); T2O = FMA(KP414213562, T2C, T2J); T2K = FNMS(KP414213562, T2J, T2C); T3V = FMA(KP707106781, T3U, T3T); T3X = FNMS(KP707106781, T3U, T3T); T35 = T33 - T34; T3W = T34 + T33; T3S = T2K - T2v; T2L = T2v + T2K; T3Y = T30 - T2X; T31 = T2X + T30; ci[WS(rs, 14)] = FMA(KP923879532, T3W, T3V); cr[WS(rs, 9)] = FMS(KP923879532, T3W, T3V); ci[0] = FMA(KP923879532, T2L, T2g); cr[WS(rs, 7)] = FNMS(KP923879532, T2L, T2g); cr[WS(rs, 13)] = FMS(KP923879532, T3Y, T3X); ci[WS(rs, 10)] = FMA(KP923879532, T3Y, T3X); T2P = T2N + T2O; T3Q = T2O - T2N; } T32 = FNMS(KP707106781, T2T, T2Q); T2U = FMA(KP707106781, T2T, T2Q); T3R = FNMS(KP707106781, T3O, T3N); T3P = FMA(KP707106781, T3O, T3N); cr[WS(rs, 3)] = FMA(KP923879532, T2P, T2M); ci[WS(rs, 4)] = FNMS(KP923879532, T2P, T2M); cr[WS(rs, 1)] = FMA(KP923879532, T31, T2U); ci[WS(rs, 6)] = FNMS(KP923879532, T31, T2U); ci[WS(rs, 8)] = FMA(KP923879532, T3Q, T3P); cr[WS(rs, 15)] = FMS(KP923879532, T3Q, T3P); ci[WS(rs, 12)] = FMA(KP923879532, T3S, T3R); cr[WS(rs, 11)] = FMS(KP923879532, T3S, T3R); } } } } } } ci[WS(rs, 2)] = FMA(KP923879532, T35, T32); cr[WS(rs, 5)] = FNMS(KP923879532, T35, T32); } } } static const tw_instr twinstr[] = { {TW_CEXP, 1, 1}, {TW_CEXP, 1, 3}, {TW_CEXP, 1, 9}, {TW_CEXP, 1, 15}, {TW_NEXT, 1, 0} }; static const hc2hc_desc desc = { 16, "hf2_16", twinstr, &GENUS, {104, 42, 92, 0} }; void X(codelet_hf2_16) (planner *p) { X(khc2hc_register) (p, hf2_16, &desc); } #else /* HAVE_FMA */ /* Generated by: ../../../genfft/gen_hc2hc.native -compact -variables 4 -pipeline-latency 4 -twiddle-log3 -precompute-twiddles -n 16 -dit -name hf2_16 -include hf.h */ /* * This function contains 196 FP additions, 108 FP multiplications, * (or, 156 additions, 68 multiplications, 40 fused multiply/add), * 82 stack variables, 3 constants, and 64 memory accesses */ #include "hf.h" static void hf2_16(R *cr, R *ci, const R *W, stride rs, INT mb, INT me, INT ms) { DK(KP382683432, +0.382683432365089771728459984030398866761344562); DK(KP923879532, +0.923879532511286756128183189396788286822416626); DK(KP707106781, +0.707106781186547524400844362104849039284835938); { INT m; for (m = mb, W = W + ((mb - 1) * 8); m < me; m = m + 1, cr = cr + ms, ci = ci - ms, W = W + 8, MAKE_VOLATILE_STRIDE(32, rs)) { E T2, T5, Tg, Ti, Tk, To, TE, TC, T6, T3, T8, TW, TJ, Tt, TU; E Tc, Tx, TH, TN, TO, TP, TR, T1f, T1k, T1b, T1i, T1y, T1H, T1u, T1F; { E T7, Tv, Ta, Ts, T4, Tw, Tb, Tr; { E Th, Tn, Tj, Tm; T2 = W[0]; T5 = W[1]; Tg = W[2]; Ti = W[3]; Th = T2 * Tg; Tn = T5 * Tg; Tj = T5 * Ti; Tm = T2 * Ti; Tk = Th - Tj; To = Tm + Tn; TE = Tm - Tn; TC = Th + Tj; T6 = W[5]; T7 = T5 * T6; Tv = Tg * T6; Ta = T2 * T6; Ts = Ti * T6; T3 = W[4]; T4 = T2 * T3; Tw = Ti * T3; Tb = T5 * T3; Tr = Tg * T3; } T8 = T4 + T7; TW = Tv - Tw; TJ = Ta + Tb; Tt = Tr - Ts; TU = Tr + Ts; Tc = Ta - Tb; Tx = Tv + Tw; TH = T4 - T7; TN = W[6]; TO = W[7]; TP = FMA(T2, TN, T5 * TO); TR = FNMS(T5, TN, T2 * TO); { E T1d, T1e, T19, T1a; T1d = Tk * T6; T1e = To * T3; T1f = T1d - T1e; T1k = T1d + T1e; T19 = Tk * T3; T1a = To * T6; T1b = T19 + T1a; T1i = T19 - T1a; } { E T1w, T1x, T1s, T1t; T1w = TC * T6; T1x = TE * T3; T1y = T1w - T1x; T1H = T1w + T1x; T1s = TC * T3; T1t = TE * T6; T1u = T1s + T1t; T1F = T1s - T1t; } } { E Tf, T3s, T1N, T3e, TA, T3r, T1Q, T3b, TM, T2N, T1W, T2w, TZ, T2M, T21; E T2x, T1B, T1K, T2V, T2W, T2X, T2Y, T2j, T2E, T2o, T2D, T18, T1n, T2Q, T2R; E T2S, T2T, T28, T2B, T2d, T2A; { E T1, T3d, Te, T3c, T9, Td; T1 = cr[0]; T3d = ci[0]; T9 = cr[WS(rs, 8)]; Td = ci[WS(rs, 8)]; Te = FMA(T8, T9, Tc * Td); T3c = FNMS(Tc, T9, T8 * Td); Tf = T1 + Te; T3s = T3d - T3c; T1N = T1 - Te; T3e = T3c + T3d; } { E Tq, T1O, Tz, T1P; { E Tl, Tp, Tu, Ty; Tl = cr[WS(rs, 4)]; Tp = ci[WS(rs, 4)]; Tq = FMA(Tk, Tl, To * Tp); T1O = FNMS(To, Tl, Tk * Tp); Tu = cr[WS(rs, 12)]; Ty = ci[WS(rs, 12)]; Tz = FMA(Tt, Tu, Tx * Ty); T1P = FNMS(Tx, Tu, Tt * Ty); } TA = Tq + Tz; T3r = Tq - Tz; T1Q = T1O - T1P; T3b = T1O + T1P; } { E TG, T1T, TL, T1U, T1S, T1V; { E TD, TF, TI, TK; TD = cr[WS(rs, 2)]; TF = ci[WS(rs, 2)]; TG = FMA(TC, TD, TE * TF); T1T = FNMS(TE, TD, TC * TF); TI = cr[WS(rs, 10)]; TK = ci[WS(rs, 10)]; TL = FMA(TH, TI, TJ * TK); T1U = FNMS(TJ, TI, TH * TK); } TM = TG + TL; T2N = T1T + T1U; T1S = TG - TL; T1V = T1T - T1U; T1W = T1S - T1V; T2w = T1S + T1V; } { E TT, T1Y, TY, T1Z, T1X, T20; { E TQ, TS, TV, TX; TQ = cr[WS(rs, 14)]; TS = ci[WS(rs, 14)]; TT = FMA(TP, TQ, TR * TS); T1Y = FNMS(TR, TQ, TP * TS); TV = cr[WS(rs, 6)]; TX = ci[WS(rs, 6)]; TY = FMA(TU, TV, TW * TX); T1Z = FNMS(TW, TV, TU * TX); } TZ = TT + TY; T2M = T1Y + T1Z; T1X = TT - TY; T20 = T1Y - T1Z; T21 = T1X + T20; T2x = T1X - T20; } { E T1r, T2f, T1J, T2m, T1A, T2g, T1E, T2l; { E T1p, T1q, T1G, T1I; T1p = cr[WS(rs, 15)]; T1q = ci[WS(rs, 15)]; T1r = FMA(TN, T1p, TO * T1q); T2f = FNMS(TO, T1p, TN * T1q); T1G = cr[WS(rs, 11)]; T1I = ci[WS(rs, 11)]; T1J = FMA(T1F, T1G, T1H * T1I); T2m = FNMS(T1H, T1G, T1F * T1I); } { E T1v, T1z, T1C, T1D; T1v = cr[WS(rs, 7)]; T1z = ci[WS(rs, 7)]; T1A = FMA(T1u, T1v, T1y * T1z); T2g = FNMS(T1y, T1v, T1u * T1z); T1C = cr[WS(rs, 3)]; T1D = ci[WS(rs, 3)]; T1E = FMA(Tg, T1C, Ti * T1D); T2l = FNMS(Ti, T1C, Tg * T1D); } T1B = T1r + T1A; T1K = T1E + T1J; T2V = T1B - T1K; T2W = T2f + T2g; T2X = T2l + T2m; T2Y = T2W - T2X; { E T2h, T2i, T2k, T2n; T2h = T2f - T2g; T2i = T1E - T1J; T2j = T2h + T2i; T2E = T2h - T2i; T2k = T1r - T1A; T2n = T2l - T2m; T2o = T2k - T2n; T2D = T2k + T2n; } } { E T14, T29, T1m, T26, T17, T2a, T1h, T25; { E T12, T13, T1j, T1l; T12 = cr[WS(rs, 1)]; T13 = ci[WS(rs, 1)]; T14 = FMA(T2, T12, T5 * T13); T29 = FNMS(T5, T12, T2 * T13); T1j = cr[WS(rs, 13)]; T1l = ci[WS(rs, 13)]; T1m = FMA(T1i, T1j, T1k * T1l); T26 = FNMS(T1k, T1j, T1i * T1l); } { E T15, T16, T1c, T1g; T15 = cr[WS(rs, 9)]; T16 = ci[WS(rs, 9)]; T17 = FMA(T3, T15, T6 * T16); T2a = FNMS(T6, T15, T3 * T16); T1c = cr[WS(rs, 5)]; T1g = ci[WS(rs, 5)]; T1h = FMA(T1b, T1c, T1f * T1g); T25 = FNMS(T1f, T1c, T1b * T1g); } T18 = T14 + T17; T1n = T1h + T1m; T2Q = T18 - T1n; T2R = T29 + T2a; T2S = T25 + T26; T2T = T2R - T2S; { E T24, T27, T2b, T2c; T24 = T14 - T17; T27 = T25 - T26; T28 = T24 - T27; T2B = T24 + T27; T2b = T29 - T2a; T2c = T1h - T1m; T2d = T2b + T2c; T2A = T2b - T2c; } } { E T23, T2r, T3u, T3w, T2q, T3v, T2u, T3p; { E T1R, T22, T3q, T3t; T1R = T1N - T1Q; T22 = KP707106781 * (T1W + T21); T23 = T1R + T22; T2r = T1R - T22; T3q = KP707106781 * (T2w - T2x); T3t = T3r + T3s; T3u = T3q + T3t; T3w = T3t - T3q; } { E T2e, T2p, T2s, T2t; T2e = FNMS(KP382683432, T2d, KP923879532 * T28); T2p = FMA(KP382683432, T2j, KP923879532 * T2o); T2q = T2e + T2p; T3v = T2p - T2e; T2s = FMA(KP923879532, T2d, KP382683432 * T28); T2t = FNMS(KP923879532, T2j, KP382683432 * T2o); T2u = T2s + T2t; T3p = T2t - T2s; } cr[WS(rs, 7)] = T23 - T2q; cr[WS(rs, 11)] = T3v - T3w; ci[WS(rs, 12)] = T3v + T3w; ci[0] = T23 + T2q; ci[WS(rs, 4)] = T2r - T2u; cr[WS(rs, 15)] = T3p - T3u; ci[WS(rs, 8)] = T3p + T3u; cr[WS(rs, 3)] = T2r + T2u; } { E T11, T35, T3g, T3i, T1M, T3h, T38, T39; { E TB, T10, T3a, T3f; TB = Tf + TA; T10 = TM + TZ; T11 = TB + T10; T35 = TB - T10; T3a = T2N + T2M; T3f = T3b + T3e; T3g = T3a + T3f; T3i = T3f - T3a; } { E T1o, T1L, T36, T37; T1o = T18 + T1n; T1L = T1B + T1K; T1M = T1o + T1L; T3h = T1L - T1o; T36 = T2W + T2X; T37 = T2R + T2S; T38 = T36 - T37; T39 = T37 + T36; } ci[WS(rs, 7)] = T11 - T1M; cr[WS(rs, 12)] = T3h - T3i; ci[WS(rs, 11)] = T3h + T3i; cr[0] = T11 + T1M; cr[WS(rs, 4)] = T35 - T38; cr[WS(rs, 8)] = T39 - T3g; ci[WS(rs, 15)] = T39 + T3g; ci[WS(rs, 3)] = T35 + T38; } { E T2z, T2H, T3A, T3C, T2G, T3B, T2K, T3x; { E T2v, T2y, T3y, T3z; T2v = T1N + T1Q; T2y = KP707106781 * (T2w + T2x); T2z = T2v + T2y; T2H = T2v - T2y; T3y = KP707106781 * (T21 - T1W); T3z = T3s - T3r; T3A = T3y + T3z; T3C = T3z - T3y; } { E T2C, T2F, T2I, T2J; T2C = FMA(KP382683432, T2A, KP923879532 * T2B); T2F = FNMS(KP382683432, T2E, KP923879532 * T2D); T2G = T2C + T2F; T3B = T2F - T2C; T2I = FNMS(KP923879532, T2A, KP382683432 * T2B); T2J = FMA(KP923879532, T2E, KP382683432 * T2D); T2K = T2I + T2J; T3x = T2J - T2I; } ci[WS(rs, 6)] = T2z - T2G; cr[WS(rs, 13)] = T3B - T3C; ci[WS(rs, 10)] = T3B + T3C; cr[WS(rs, 1)] = T2z + T2G; cr[WS(rs, 5)] = T2H - T2K; cr[WS(rs, 9)] = T3x - T3A; ci[WS(rs, 14)] = T3x + T3A; ci[WS(rs, 2)] = T2H + T2K; } { E T2P, T31, T3m, T3o, T30, T3j, T34, T3n; { E T2L, T2O, T3k, T3l; T2L = Tf - TA; T2O = T2M - T2N; T2P = T2L - T2O; T31 = T2L + T2O; T3k = TM - TZ; T3l = T3e - T3b; T3m = T3k + T3l; T3o = T3l - T3k; } { E T2U, T2Z, T32, T33; T2U = T2Q + T2T; T2Z = T2V - T2Y; T30 = KP707106781 * (T2U + T2Z); T3j = KP707106781 * (T2Z - T2U); T32 = T2Q - T2T; T33 = T2V + T2Y; T34 = KP707106781 * (T32 + T33); T3n = KP707106781 * (T33 - T32); } ci[WS(rs, 5)] = T2P - T30; cr[WS(rs, 10)] = T3n - T3o; ci[WS(rs, 13)] = T3n + T3o; cr[WS(rs, 2)] = T2P + T30; cr[WS(rs, 6)] = T31 - T34; cr[WS(rs, 14)] = T3j - T3m; ci[WS(rs, 9)] = T3j + T3m; ci[WS(rs, 1)] = T31 + T34; } } } } } static const tw_instr twinstr[] = { {TW_CEXP, 1, 1}, {TW_CEXP, 1, 3}, {TW_CEXP, 1, 9}, {TW_CEXP, 1, 15}, {TW_NEXT, 1, 0} }; static const hc2hc_desc desc = { 16, "hf2_16", twinstr, &GENUS, {156, 68, 40, 0} }; void X(codelet_hf2_16) (planner *p) { X(khc2hc_register) (p, hf2_16, &desc); } #endif /* HAVE_FMA */ fftw-3.3.3/rdft/scalar/r2cf/hf2_32.c0000644000175000017500000015064612054410655013614 00000000000000/* * Copyright (c) 2003, 2007-11 Matteo Frigo * Copyright (c) 2003, 2007-11 Massachusetts Institute of Technology * * This program is free software; you can redistribute it and/or modify * it under the terms of the GNU General Public License as published by * the Free Software Foundation; either version 2 of the License, or * (at your option) any later version. * * This program is distributed in the hope that it will be useful, * but WITHOUT ANY WARRANTY; without even the implied warranty of * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the * GNU General Public License for more details. * * You should have received a copy of the GNU General Public License * along with this program; if not, write to the Free Software * Foundation, Inc., 51 Franklin Street, Fifth Floor, Boston, MA 02110-1301 USA * */ /* This file was automatically generated --- DO NOT EDIT */ /* Generated on Sun Nov 25 07:40:04 EST 2012 */ #include "codelet-rdft.h" #ifdef HAVE_FMA /* Generated by: ../../../genfft/gen_hc2hc.native -fma -reorder-insns -schedule-for-pipeline -compact -variables 4 -pipeline-latency 4 -twiddle-log3 -precompute-twiddles -n 32 -dit -name hf2_32 -include hf.h */ /* * This function contains 488 FP additions, 350 FP multiplications, * (or, 236 additions, 98 multiplications, 252 fused multiply/add), * 181 stack variables, 7 constants, and 128 memory accesses */ #include "hf.h" static void hf2_32(R *cr, R *ci, const R *W, stride rs, INT mb, INT me, INT ms) { DK(KP831469612, +0.831469612302545237078788377617905756738560812); DK(KP980785280, +0.980785280403230449126182236134239036973933731); DK(KP923879532, +0.923879532511286756128183189396788286822416626); DK(KP668178637, +0.668178637919298919997757686523080761552472251); DK(KP198912367, +0.198912367379658006911597622644676228597850501); DK(KP414213562, +0.414213562373095048801688724209698078569671875); DK(KP707106781, +0.707106781186547524400844362104849039284835938); { INT m; for (m = mb, W = W + ((mb - 1) * 8); m < me; m = m + 1, cr = cr + ms, ci = ci - ms, W = W + 8, MAKE_VOLATILE_STRIDE(64, rs)) { E T7d, T7a; { E T2, T8, T3, T6, Te, Tr, T18, T4, Ta, Tz, T1n, T10, Ti, T5, Tc; T2 = W[0]; T8 = W[4]; T3 = W[2]; T6 = W[3]; Te = W[6]; Tr = T2 * T8; T18 = T3 * T8; T4 = T2 * T3; Ta = T2 * T6; Tz = T3 * Te; T1n = T8 * Te; T10 = T2 * Te; Ti = W[7]; T5 = W[1]; Tc = W[5]; { E T34, T31, T2X, T2T, Tq, T46, T8H, T98, TH, T97, T4b, T8D, TZ, T7g, T4j; E T6t, T1g, T7f, T4q, T6u, T4z, T6y, T1J, T7j, T7m, T8e, T6x, T4G, T2k, T7o; E T7r, T8d, T6B, T4O, T6A, T4V, T6P, T61, T7G, T3G, T6M, T5E, T8n, T7N, T6I; E T5s, T7v, T2N, T6F, T55, T8i, T7C, T5L, T63, T43, T7O, T5S, T62, T7J, T8o; E T2U, T2R, T2V, T58, T3a, T5h, T2Y, T32, T35; { E T1K, T23, T1N, T26, T2b, T1U, T3C, T3j, T3z, T3f, T1R, T29, TR, Th, T2J; E T2F, Td, TP, T3r, T3n, T2w, T2s, T3Q, T3M, T1Z, T1V, T2g, T2c; { E T11, T1C, TM, Tb, TJ, T7, T1o, T19, T1w, T1F, T15, T1s, T1d, T1z, TW; E TS, Ty, T48, TG, T4a; { E T1, TA, Ts, TE, Tw, Tn, Tj, T8G, Tk, To, T14; T1 = cr[0]; TA = FMA(T6, Ti, Tz); T1K = FNMS(T6, Ti, Tz); T14 = T2 * Ti; { E T1r, TD, T1c, Tv; T1r = T8 * Ti; TD = T3 * Ti; T11 = FNMS(T5, Ti, T10); T1C = FMA(T5, Ti, T10); TM = FMA(T5, T3, Ta); Tb = FNMS(T5, T3, Ta); TJ = FNMS(T5, T6, T4); T7 = FMA(T5, T6, T4); T1o = FMA(Tc, Ti, T1n); T23 = FMA(T6, Tc, T18); T19 = FNMS(T6, Tc, T18); T1w = FNMS(T5, Tc, Tr); Ts = FMA(T5, Tc, Tr); T1c = T3 * Tc; Tv = T2 * Tc; T1F = FNMS(T5, Te, T14); T15 = FMA(T5, Te, T14); T1s = FNMS(Tc, Te, T1r); T1N = FMA(T6, Te, TD); TE = FNMS(T6, Te, TD); { E T1T, T3i, T3e, T1Q; T1T = TJ * Tc; T3i = TJ * Ti; T3e = TJ * Te; T1Q = TJ * T8; { E Tg, T2I, T2E, T9; Tg = T7 * Tc; T2I = T7 * Ti; T2E = T7 * Te; T9 = T7 * T8; { E T3q, T3m, T2v, T2r; T3q = T19 * Ti; T3m = T19 * Te; T2v = T1w * Ti; T2r = T1w * Te; { E T2W, T2S, T3P, T3L; T2W = T23 * Ti; T2S = T23 * Te; T3P = Ts * Ti; T3L = Ts * Te; T26 = FNMS(T6, T8, T1c); T1d = FMA(T6, T8, T1c); T1z = FMA(T5, T8, Tv); Tw = FNMS(T5, T8, Tv); T2b = FNMS(TM, T8, T1T); T1U = FMA(TM, T8, T1T); T3C = FNMS(TM, Te, T3i); T3j = FMA(TM, Te, T3i); T3z = FMA(TM, Ti, T3e); T3f = FNMS(TM, Ti, T3e); T1R = FNMS(TM, Tc, T1Q); T29 = FMA(TM, Tc, T1Q); TR = FNMS(Tb, T8, Tg); Th = FMA(Tb, T8, Tg); T34 = FMA(Tb, Te, T2I); T2J = FNMS(Tb, Te, T2I); T31 = FNMS(Tb, Ti, T2E); T2F = FMA(Tb, Ti, T2E); Td = FNMS(Tb, Tc, T9); TP = FMA(Tb, Tc, T9); T2X = FNMS(T26, Te, T2W); T2T = FMA(T26, Ti, T2S); T3r = FNMS(T1d, Te, T3q); T3n = FMA(T1d, Ti, T3m); T2w = FNMS(T1z, Te, T2v); T2s = FMA(T1z, Ti, T2r); T3Q = FNMS(Tw, Te, T3P); T3M = FMA(Tw, Ti, T3L); { E T1Y, T1S, T2f, T2a; T1Y = T1R * Ti; T1S = T1R * Te; T2f = T29 * Ti; T2a = T29 * Te; { E Tm, Tf, TV, TQ; Tm = Td * Ti; Tf = Td * Te; TV = TP * Ti; TQ = TP * Te; T1Z = FNMS(T1U, Te, T1Y); T1V = FMA(T1U, Ti, T1S); T2g = FNMS(T2b, Te, T2f); T2c = FMA(T2b, Ti, T2a); Tn = FNMS(Th, Te, Tm); Tj = FMA(Th, Ti, Tf); TW = FNMS(TR, Te, TV); TS = FMA(TR, Ti, TQ); T8G = ci[0]; } } } } } } } Tk = cr[WS(rs, 16)]; To = ci[WS(rs, 16)]; { E Tt, Tx, Tu, T47, TB, TF, TC, T49; { E Tl, T8E, Tp, T8F; Tt = cr[WS(rs, 8)]; Tx = ci[WS(rs, 8)]; Tl = Tj * Tk; T8E = Tj * To; Tu = Ts * Tt; T47 = Ts * Tx; Tp = FMA(Tn, To, Tl); T8F = FNMS(Tn, Tk, T8E); TB = cr[WS(rs, 24)]; TF = ci[WS(rs, 24)]; Tq = T1 + Tp; T46 = T1 - Tp; T8H = T8F + T8G; T98 = T8G - T8F; TC = TA * TB; T49 = TA * TF; } Ty = FMA(Tw, Tx, Tu); T48 = FNMS(Tw, Tt, T47); TG = FMA(TE, TF, TC); T4a = FNMS(TE, TB, T49); } } { E TT, TX, TO, T4f, TU, T4g; { E TK, TN, TL, T4e; TK = cr[WS(rs, 4)]; TN = ci[WS(rs, 4)]; TH = Ty + TG; T97 = Ty - TG; T4b = T48 - T4a; T8D = T48 + T4a; TL = TJ * TK; T4e = TJ * TN; TT = cr[WS(rs, 20)]; TX = ci[WS(rs, 20)]; TO = FMA(TM, TN, TL); T4f = FNMS(TM, TK, T4e); TU = TS * TT; T4g = TS * TX; } { E T17, T4m, T1a, T1e, T4d, T4i; { E T12, T16, TY, T4h, T13, T4l; T12 = cr[WS(rs, 28)]; T16 = ci[WS(rs, 28)]; TY = FMA(TW, TX, TU); T4h = FNMS(TW, TT, T4g); T13 = T11 * T12; T4l = T11 * T16; TZ = TO + TY; T4d = TO - TY; T7g = T4f + T4h; T4i = T4f - T4h; T17 = FMA(T15, T16, T13); T4m = FNMS(T15, T12, T4l); } T4j = T4d - T4i; T6t = T4d + T4i; T1a = cr[WS(rs, 12)]; T1e = ci[WS(rs, 12)]; { E T1m, T4u, T1H, T4E, T1x, T1A, T1u, T4w, T1y, T4B; { E T1D, T1G, T1E, T4D; { E T1f, T4o, T4k, T4p; { E T1j, T1l, T1b, T4n, T1k, T4t; T1j = cr[WS(rs, 2)]; T1l = ci[WS(rs, 2)]; T1b = T19 * T1a; T4n = T19 * T1e; T1k = T7 * T1j; T4t = T7 * T1l; T1f = FMA(T1d, T1e, T1b); T4o = FNMS(T1d, T1a, T4n); T1m = FMA(Tb, T1l, T1k); T4u = FNMS(Tb, T1j, T4t); } T1g = T17 + T1f; T4k = T17 - T1f; T7f = T4m + T4o; T4p = T4m - T4o; T1D = cr[WS(rs, 26)]; T1G = ci[WS(rs, 26)]; T4q = T4k + T4p; T6u = T4k - T4p; T1E = T1C * T1D; T4D = T1C * T1G; } { E T1p, T1t, T1q, T4v; T1p = cr[WS(rs, 18)]; T1t = ci[WS(rs, 18)]; T1H = FMA(T1F, T1G, T1E); T4E = FNMS(T1F, T1D, T4D); T1q = T1o * T1p; T4v = T1o * T1t; T1x = cr[WS(rs, 10)]; T1A = ci[WS(rs, 10)]; T1u = FMA(T1s, T1t, T1q); T4w = FNMS(T1s, T1p, T4v); T1y = T1w * T1x; T4B = T1w * T1A; } } { E T4A, T1v, T7k, T4x, T1B, T4C; T4A = T1m - T1u; T1v = T1m + T1u; T7k = T4u + T4w; T4x = T4u - T4w; T1B = FMA(T1z, T1A, T1y); T4C = FNMS(T1z, T1x, T4B); { E T1I, T4y, T4F, T7l; T1I = T1B + T1H; T4y = T1B - T1H; T4F = T4C - T4E; T7l = T4C + T4E; T4z = T4x + T4y; T6y = T4x - T4y; T1J = T1v + T1I; T7j = T1v - T1I; T7m = T7k - T7l; T8e = T7k + T7l; T6x = T4A + T4F; T4G = T4A - T4F; } } } } } } { E T5C, T3u, T5y, T7L, T60, T5V, T3F, T5A, T4P, T4U; { E T1P, T4J, T2i, T4T, T21, T4L, T28, T4R; { E T1L, T1O, T1W, T20; T1L = cr[WS(rs, 30)]; T1O = ci[WS(rs, 30)]; { E T2d, T2h, T1M, T4I, T2e, T4S; T2d = cr[WS(rs, 22)]; T2h = ci[WS(rs, 22)]; T1M = T1K * T1L; T4I = T1K * T1O; T2e = T2c * T2d; T4S = T2c * T2h; T1P = FMA(T1N, T1O, T1M); T4J = FNMS(T1N, T1L, T4I); T2i = FMA(T2g, T2h, T2e); T4T = FNMS(T2g, T2d, T4S); } T1W = cr[WS(rs, 14)]; T20 = ci[WS(rs, 14)]; { E T24, T27, T1X, T4K, T25, T4Q; T24 = cr[WS(rs, 6)]; T27 = ci[WS(rs, 6)]; T1X = T1V * T1W; T4K = T1V * T20; T25 = T23 * T24; T4Q = T23 * T27; T21 = FMA(T1Z, T20, T1X); T4L = FNMS(T1Z, T1W, T4K); T28 = FMA(T26, T27, T25); T4R = FNMS(T26, T24, T4Q); } } { E T22, T7p, T4M, T4N, T2j, T7q; T4P = T1P - T21; T22 = T1P + T21; T7p = T4J + T4L; T4M = T4J - T4L; T4N = T28 - T2i; T2j = T28 + T2i; T7q = T4R + T4T; T4U = T4R - T4T; T2k = T22 + T2j; T7o = T22 - T2j; T7r = T7p - T7q; T8d = T7p + T7q; T6B = T4M - T4N; T4O = T4M + T4N; } } { E T3l, T5X, T3E, T3v, T3t, T3w, T3x, T5Z, T3A, T3B, T3D, T3y, T5z; { E T3g, T3k, T3h, T5W; T3g = cr[WS(rs, 31)]; T3k = ci[WS(rs, 31)]; T3A = cr[WS(rs, 23)]; T6A = T4P + T4U; T4V = T4P - T4U; T3h = T3f * T3g; T5W = T3f * T3k; T3B = T3z * T3A; T3D = ci[WS(rs, 23)]; T3l = FMA(T3j, T3k, T3h); T5X = FNMS(T3j, T3g, T5W); } { E T3o, T5B, T3s, T3p, T5Y; T3o = cr[WS(rs, 15)]; T3E = FMA(T3C, T3D, T3B); T5B = T3z * T3D; T3s = ci[WS(rs, 15)]; T3p = T3n * T3o; T3v = cr[WS(rs, 7)]; T5C = FNMS(T3C, T3A, T5B); T5Y = T3n * T3s; T3t = FMA(T3r, T3s, T3p); T3w = TP * T3v; T3x = ci[WS(rs, 7)]; T5Z = FNMS(T3r, T3o, T5Y); } T3u = T3l + T3t; T5y = T3l - T3t; T3y = FMA(TR, T3x, T3w); T5z = TP * T3x; T7L = T5X + T5Z; T60 = T5X - T5Z; T5V = T3E - T3y; T3F = T3y + T3E; T5A = FNMS(TR, T3v, T5z); } { E T2L, T53, T4Z, T2z, T7A, T5q, T2D, T51; { E T2q, T5n, T2y, T2A, T2C, T5p, T2B, T50; { E T2G, T2K, T2n, T5m, T2t, T5o; { E T2o, T2p, T5D, T7M; T2n = cr[WS(rs, 1)]; T6P = T60 + T5V; T61 = T5V - T60; T7G = T3u - T3F; T3G = T3u + T3F; T5D = T5A - T5C; T7M = T5A + T5C; T2o = T2 * T2n; T2p = ci[WS(rs, 1)]; T6M = T5y + T5D; T5E = T5y - T5D; T8n = T7L + T7M; T7N = T7L - T7M; T5m = T2 * T2p; T2q = FMA(T5, T2p, T2o); } T2G = cr[WS(rs, 25)]; T2K = ci[WS(rs, 25)]; T5n = FNMS(T5, T2n, T5m); { E T2x, T2u, T2H, T52; T2t = cr[WS(rs, 17)]; T2H = T2F * T2G; T52 = T2F * T2K; T2x = ci[WS(rs, 17)]; T2u = T2s * T2t; T2L = FMA(T2J, T2K, T2H); T53 = FNMS(T2J, T2G, T52); T5o = T2s * T2x; T2y = FMA(T2w, T2x, T2u); } T2A = cr[WS(rs, 9)]; T2C = ci[WS(rs, 9)]; T5p = FNMS(T2w, T2t, T5o); } T4Z = T2q - T2y; T2z = T2q + T2y; T2B = T8 * T2A; T50 = T8 * T2C; T7A = T5n + T5p; T5q = T5n - T5p; T2D = FMA(Tc, T2C, T2B); T51 = FNMS(Tc, T2A, T50); } { E T3N, T3K, T3O, T5H, T41, T5Q, T3R, T3U, T3W; { E T3H, T3I, T3J, T3Y, T40, T5G, T3Z, T5P; T3H = cr[WS(rs, 3)]; { E T5r, T2M, T54, T7B; T5r = T2D - T2L; T2M = T2D + T2L; T54 = T51 - T53; T7B = T51 + T53; T6I = T5q - T5r; T5s = T5q + T5r; T7v = T2z - T2M; T2N = T2z + T2M; T6F = T4Z + T54; T55 = T4Z - T54; T8i = T7A + T7B; T7C = T7A - T7B; T3I = T3 * T3H; } T3J = ci[WS(rs, 3)]; T3Y = cr[WS(rs, 11)]; T40 = ci[WS(rs, 11)]; T3N = cr[WS(rs, 19)]; T3K = FMA(T6, T3J, T3I); T5G = T3 * T3J; T3Z = Td * T3Y; T5P = Td * T40; T3O = T3M * T3N; T5H = FNMS(T6, T3H, T5G); T41 = FMA(Th, T40, T3Z); T5Q = FNMS(Th, T3Y, T5P); T3R = ci[WS(rs, 19)]; T3U = cr[WS(rs, 27)]; T3W = ci[WS(rs, 27)]; } { E T2O, T2P, T2Q, T37, T39, T57, T38, T5g; { E T3T, T5F, T5J, T3X, T5O, T7I, T5K; T2O = cr[WS(rs, 5)]; { E T3S, T5I, T3V, T5N; T3S = FMA(T3Q, T3R, T3O); T5I = T3M * T3R; T3V = Te * T3U; T5N = Te * T3W; T3T = T3K + T3S; T5F = T3K - T3S; T5J = FNMS(T3Q, T3N, T5I); T3X = FMA(Ti, T3W, T3V); T5O = FNMS(Ti, T3U, T5N); T2P = T29 * T2O; } T7I = T5H + T5J; T5K = T5H - T5J; { E T42, T5M, T7H, T5R; T42 = T3X + T41; T5M = T3X - T41; T7H = T5O + T5Q; T5R = T5O - T5Q; T5L = T5F - T5K; T63 = T5F + T5K; T43 = T3T + T42; T7O = T42 - T3T; T5S = T5M + T5R; T62 = T5M - T5R; T7J = T7H - T7I; T8o = T7I + T7H; T2Q = ci[WS(rs, 5)]; } } T37 = cr[WS(rs, 13)]; T39 = ci[WS(rs, 13)]; T2U = cr[WS(rs, 21)]; T2R = FMA(T2b, T2Q, T2P); T57 = T29 * T2Q; T38 = T1R * T37; T5g = T1R * T39; T2V = T2T * T2U; T58 = FNMS(T2b, T2O, T57); T3a = FMA(T1U, T39, T38); T5h = FNMS(T1U, T37, T5g); T2Y = ci[WS(rs, 21)]; T32 = cr[WS(rs, 29)]; T35 = ci[WS(rs, 29)]; } } } } } { E T7e, T8T, T7D, T7y, T7h, T8U, T6s, T9o, T9n, T6v, T6Q, T6N, T6J, T6G, T6o; E T6r; { E T5c, T5t, T5j, T5u, T8s, T8v; { E T8c, T1i, T8A, T8z, T8O, T8J, T8N, T2l, T8L, T45, T8t, T8l, T8u, T8q, T3c; E T8p, T8k, T8w, T2m; { E T8x, T8y, T8j, T8C, T8I; { E TI, T30, T56, T5a, T36, T5f, T1h, T7x, T5b; TI = Tq + TH; T7e = Tq - TH; { E T2Z, T59, T33, T5e; T2Z = FMA(T2X, T2Y, T2V); T59 = T2T * T2Y; T33 = T31 * T32; T5e = T31 * T35; T30 = T2R + T2Z; T56 = T2R - T2Z; T5a = FNMS(T2X, T2U, T59); T36 = FMA(T34, T35, T33); T5f = FNMS(T34, T32, T5e); T1h = TZ + T1g; T8T = TZ - T1g; } T7x = T58 + T5a; T5b = T58 - T5a; { E T3b, T5d, T7w, T5i; T3b = T36 + T3a; T5d = T36 - T3a; T7w = T5f + T5h; T5i = T5f - T5h; T5c = T56 - T5b; T5t = T56 + T5b; T3c = T30 + T3b; T7D = T30 - T3b; T5j = T5d + T5i; T5u = T5i - T5d; T7y = T7w - T7x; T8j = T7x + T7w; T8c = TI - T1h; T1i = TI + T1h; } } T8p = T8n - T8o; T8x = T8n + T8o; T8y = T8i + T8j; T8k = T8i - T8j; T7h = T7f - T7g; T8C = T7g + T7f; T8I = T8D + T8H; T8U = T8H - T8D; T8A = T8y + T8x; T8z = T8x - T8y; T8O = T8I - T8C; T8J = T8C + T8I; } { E T8h, T8m, T3d, T44; T8h = T2N - T3c; T3d = T2N + T3c; T44 = T3G + T43; T8m = T3G - T43; T8N = T1J - T2k; T2l = T1J + T2k; T8L = T44 - T3d; T45 = T3d + T44; T8t = T8h - T8k; T8l = T8h + T8k; T8u = T8m + T8p; T8q = T8m - T8p; } T8w = T1i - T2l; T2m = T1i + T2l; { E T8Q, T8R, T8P, T8S; { E T8r, T8M, T8K, T8g, T8B, T8f; T8Q = T8q - T8l; T8r = T8l + T8q; T8B = T8e + T8d; T8f = T8d - T8e; cr[0] = T2m + T45; ci[WS(rs, 15)] = T2m - T45; ci[WS(rs, 7)] = T8w + T8z; cr[WS(rs, 8)] = T8w - T8z; T8M = T8J - T8B; T8K = T8B + T8J; T8g = T8c - T8f; T8s = T8c + T8f; T8R = T8O - T8N; T8P = T8N + T8O; ci[WS(rs, 23)] = T8L + T8M; cr[WS(rs, 24)] = T8L - T8M; ci[WS(rs, 31)] = T8A + T8K; cr[WS(rs, 16)] = T8A - T8K; cr[WS(rs, 4)] = FMA(KP707106781, T8r, T8g); ci[WS(rs, 11)] = FNMS(KP707106781, T8r, T8g); } T8S = T8u - T8t; T8v = T8t + T8u; ci[WS(rs, 19)] = FMA(KP707106781, T8Q, T8P); cr[WS(rs, 28)] = FMS(KP707106781, T8Q, T8P); ci[WS(rs, 27)] = FMA(KP707106781, T8S, T8R); cr[WS(rs, 20)] = FMS(KP707106781, T8S, T8R); } } { E T6c, T4s, T9c, T4X, T9h, T9b, T9i, T6f, T5l, T6h, T6m, T6q, T6a, T66, T5v; { E T6d, T4H, T4W, T6e, T99, T9a, T4c, T4r, T5T, T64; T6s = T46 + T4b; T4c = T46 - T4b; T4r = T4j + T4q; T9o = T4q - T4j; T6d = FNMS(KP414213562, T4z, T4G); T4H = FMA(KP414213562, T4G, T4z); ci[WS(rs, 3)] = FMA(KP707106781, T8v, T8s); cr[WS(rs, 12)] = FNMS(KP707106781, T8v, T8s); T6c = FMA(KP707106781, T4r, T4c); T4s = FNMS(KP707106781, T4r, T4c); T4W = FNMS(KP414213562, T4V, T4O); T6e = FMA(KP414213562, T4O, T4V); T9n = T98 - T97; T99 = T97 + T98; T9a = T6t - T6u; T6v = T6t + T6u; T9c = T4H + T4W; T4X = T4H - T4W; T9h = FNMS(KP707106781, T9a, T99); T9b = FMA(KP707106781, T9a, T99); T6Q = T5S - T5L; T5T = T5L + T5S; T64 = T62 - T63; T6N = T63 + T62; { E T6k, T5U, T6l, T65, T5k; T6J = T5j - T5c; T5k = T5c + T5j; T9i = T6e - T6d; T6f = T6d + T6e; T6k = FMA(KP707106781, T5T, T5E); T5U = FNMS(KP707106781, T5T, T5E); T6l = FMA(KP707106781, T64, T61); T65 = FNMS(KP707106781, T64, T61); T5l = FNMS(KP707106781, T5k, T55); T6h = FMA(KP707106781, T5k, T55); T6m = FNMS(KP198912367, T6l, T6k); T6q = FMA(KP198912367, T6k, T6l); T6a = FNMS(KP668178637, T5U, T65); T66 = FMA(KP668178637, T65, T5U); T5v = T5t + T5u; T6G = T5t - T5u; } } { E T68, T4Y, T9j, T9l, T6i, T5w; T68 = FNMS(KP923879532, T4X, T4s); T4Y = FMA(KP923879532, T4X, T4s); T9j = FMA(KP923879532, T9i, T9h); T9l = FNMS(KP923879532, T9i, T9h); T6i = FMA(KP707106781, T5v, T5s); T5w = FNMS(KP707106781, T5v, T5s); { E T9g, T9f, T9d, T9e; { E T6g, T6p, T69, T5x, T6n, T6j; T6o = FNMS(KP923879532, T6f, T6c); T6g = FMA(KP923879532, T6f, T6c); T6j = FNMS(KP198912367, T6i, T6h); T6p = FMA(KP198912367, T6h, T6i); T69 = FNMS(KP668178637, T5l, T5w); T5x = FMA(KP668178637, T5w, T5l); T6n = T6j + T6m; T9g = T6m - T6j; T9f = FNMS(KP923879532, T9c, T9b); T9d = FMA(KP923879532, T9c, T9b); { E T6b, T9k, T9m, T67; T6b = T69 + T6a; T9k = T69 - T6a; T9m = T66 - T5x; T67 = T5x + T66; ci[0] = FMA(KP980785280, T6n, T6g); cr[WS(rs, 15)] = FNMS(KP980785280, T6n, T6g); ci[WS(rs, 4)] = FNMS(KP831469612, T6b, T68); cr[WS(rs, 11)] = FMA(KP831469612, T6b, T68); ci[WS(rs, 28)] = FMA(KP831469612, T9k, T9j); cr[WS(rs, 19)] = FMS(KP831469612, T9k, T9j); ci[WS(rs, 20)] = FMA(KP831469612, T9m, T9l); cr[WS(rs, 27)] = FMS(KP831469612, T9m, T9l); cr[WS(rs, 3)] = FMA(KP831469612, T67, T4Y); ci[WS(rs, 12)] = FNMS(KP831469612, T67, T4Y); T9e = T6q - T6p; T6r = T6p + T6q; } } ci[WS(rs, 16)] = FMA(KP980785280, T9e, T9d); cr[WS(rs, 31)] = FMS(KP980785280, T9e, T9d); ci[WS(rs, 24)] = FMA(KP980785280, T9g, T9f); cr[WS(rs, 23)] = FMS(KP980785280, T9g, T9f); } } } } { E T88, T90, T8Z, T8b; { E T7K, T7W, T7i, T7P, T8a, T86, T91, T8V, T8W, T7t, T7U, T7F, T92, T7Z, T89; E T83; { E T7X, T7n, T7s, T7Y, T84, T85; T7K = T7G - T7J; T84 = T7G + T7J; cr[WS(rs, 7)] = FMA(KP980785280, T6r, T6o); ci[WS(rs, 8)] = FNMS(KP980785280, T6r, T6o); T7W = T7e + T7h; T7i = T7e - T7h; T85 = T7O - T7N; T7P = T7N + T7O; T7X = T7j - T7m; T7n = T7j + T7m; T8a = FMA(KP414213562, T84, T85); T86 = FNMS(KP414213562, T85, T84); T91 = T8U - T8T; T8V = T8T + T8U; T7s = T7o - T7r; T7Y = T7o + T7r; { E T81, T82, T7z, T7E; T81 = T7v + T7y; T7z = T7v - T7y; T7E = T7C - T7D; T82 = T7C + T7D; T8W = T7n - T7s; T7t = T7n + T7s; T7U = FNMS(KP414213562, T7z, T7E); T7F = FMA(KP414213562, T7E, T7z); T92 = T7Y - T7X; T7Z = T7X + T7Y; T89 = FMA(KP414213562, T81, T82); T83 = FNMS(KP414213562, T82, T81); } } { E T7S, T7u, T93, T95, T7T, T7Q; T7S = FNMS(KP707106781, T7t, T7i); T7u = FMA(KP707106781, T7t, T7i); T93 = FMA(KP707106781, T92, T91); T95 = FNMS(KP707106781, T92, T91); T7T = FMA(KP414213562, T7K, T7P); T7Q = FNMS(KP414213562, T7P, T7K); { E T80, T87, T8X, T8Y; T88 = FNMS(KP707106781, T7Z, T7W); T80 = FMA(KP707106781, T7Z, T7W); { E T7V, T94, T96, T7R; T7V = T7T - T7U; T94 = T7U + T7T; T96 = T7Q - T7F; T7R = T7F + T7Q; ci[WS(rs, 5)] = FMA(KP923879532, T7V, T7S); cr[WS(rs, 10)] = FNMS(KP923879532, T7V, T7S); ci[WS(rs, 29)] = FMA(KP923879532, T94, T93); cr[WS(rs, 18)] = FMS(KP923879532, T94, T93); ci[WS(rs, 21)] = FMA(KP923879532, T96, T95); cr[WS(rs, 26)] = FMS(KP923879532, T96, T95); cr[WS(rs, 2)] = FMA(KP923879532, T7R, T7u); ci[WS(rs, 13)] = FNMS(KP923879532, T7R, T7u); T87 = T83 + T86; T90 = T86 - T83; } T8Z = FNMS(KP707106781, T8W, T8V); T8X = FMA(KP707106781, T8W, T8V); T8Y = T8a - T89; T8b = T89 + T8a; ci[WS(rs, 1)] = FMA(KP923879532, T87, T80); cr[WS(rs, 14)] = FNMS(KP923879532, T87, T80); ci[WS(rs, 17)] = FMA(KP923879532, T8Y, T8X); cr[WS(rs, 30)] = FMS(KP923879532, T8Y, T8X); } } } { E T6Y, T6w, T9w, T6D, T9v, T9p, T9q, T71, T6O, T76; { E T70, T6Z, T6z, T6C; ci[WS(rs, 25)] = FMA(KP923879532, T90, T8Z); cr[WS(rs, 22)] = FMS(KP923879532, T90, T8Z); cr[WS(rs, 6)] = FMA(KP923879532, T8b, T88); ci[WS(rs, 9)] = FNMS(KP923879532, T8b, T88); T70 = FNMS(KP414213562, T6x, T6y); T6z = FMA(KP414213562, T6y, T6x); T6C = FNMS(KP414213562, T6B, T6A); T6Z = FMA(KP414213562, T6A, T6B); T6Y = FNMS(KP707106781, T6v, T6s); T6w = FMA(KP707106781, T6v, T6s); T9w = T6z - T6C; T6D = T6z + T6C; T9v = FNMS(KP707106781, T9o, T9n); T9p = FMA(KP707106781, T9o, T9n); T9q = T70 + T6Z; T71 = T6Z - T70; T6O = FMA(KP707106781, T6N, T6M); T76 = FNMS(KP707106781, T6N, T6M); } { E T6U, T9u, T79, T6X, T9s, T9t, T9r, T72; { E T6E, T78, T6V, T6S, T75, T6W, T6L, T9x, T9z, T9y, T6T, T9A; { E T7c, T7b, T77, T6R; T6U = FNMS(KP923879532, T6D, T6w); T6E = FMA(KP923879532, T6D, T6w); T77 = FNMS(KP707106781, T6Q, T6P); T6R = FMA(KP707106781, T6Q, T6P); { E T73, T6H, T74, T6K; T73 = FNMS(KP707106781, T6G, T6F); T6H = FMA(KP707106781, T6G, T6F); T74 = FNMS(KP707106781, T6J, T6I); T6K = FMA(KP707106781, T6J, T6I); T78 = FMA(KP668178637, T77, T76); T7c = FNMS(KP668178637, T76, T77); T6V = FMA(KP198912367, T6O, T6R); T6S = FNMS(KP198912367, T6R, T6O); T75 = FNMS(KP668178637, T74, T73); T7b = FMA(KP668178637, T73, T74); T6W = FNMS(KP198912367, T6H, T6K); T6L = FMA(KP198912367, T6K, T6H); } T9x = FMA(KP923879532, T9w, T9v); T9z = FNMS(KP923879532, T9w, T9v); T7d = T7b - T7c; T9y = T7b + T7c; } T9u = T6S - T6L; T6T = T6L + T6S; T9A = T78 - T75; T79 = T75 + T78; ci[WS(rs, 18)] = FNMS(KP831469612, T9y, T9x); cr[WS(rs, 29)] = -(FMA(KP831469612, T9y, T9x)); cr[WS(rs, 1)] = FMA(KP980785280, T6T, T6E); ci[WS(rs, 14)] = FNMS(KP980785280, T6T, T6E); cr[WS(rs, 21)] = FMS(KP831469612, T9A, T9z); ci[WS(rs, 26)] = FMA(KP831469612, T9A, T9z); T6X = T6V - T6W; T9s = T6W + T6V; } T7a = FNMS(KP923879532, T71, T6Y); T72 = FMA(KP923879532, T71, T6Y); T9t = FNMS(KP923879532, T9q, T9p); T9r = FMA(KP923879532, T9q, T9p); ci[WS(rs, 6)] = FMA(KP980785280, T6X, T6U); cr[WS(rs, 9)] = FNMS(KP980785280, T6X, T6U); ci[WS(rs, 2)] = FMA(KP831469612, T79, T72); cr[WS(rs, 13)] = FNMS(KP831469612, T79, T72); ci[WS(rs, 30)] = FMA(KP980785280, T9s, T9r); cr[WS(rs, 17)] = FMS(KP980785280, T9s, T9r); ci[WS(rs, 22)] = FMA(KP980785280, T9u, T9t); cr[WS(rs, 25)] = FMS(KP980785280, T9u, T9t); } } } } } } cr[WS(rs, 5)] = FMA(KP831469612, T7d, T7a); ci[WS(rs, 10)] = FNMS(KP831469612, T7d, T7a); } } } static const tw_instr twinstr[] = { {TW_CEXP, 1, 1}, {TW_CEXP, 1, 3}, {TW_CEXP, 1, 9}, {TW_CEXP, 1, 27}, {TW_NEXT, 1, 0} }; static const hc2hc_desc desc = { 32, "hf2_32", twinstr, &GENUS, {236, 98, 252, 0} }; void X(codelet_hf2_32) (planner *p) { X(khc2hc_register) (p, hf2_32, &desc); } #else /* HAVE_FMA */ /* Generated by: ../../../genfft/gen_hc2hc.native -compact -variables 4 -pipeline-latency 4 -twiddle-log3 -precompute-twiddles -n 32 -dit -name hf2_32 -include hf.h */ /* * This function contains 488 FP additions, 280 FP multiplications, * (or, 376 additions, 168 multiplications, 112 fused multiply/add), * 158 stack variables, 7 constants, and 128 memory accesses */ #include "hf.h" static void hf2_32(R *cr, R *ci, const R *W, stride rs, INT mb, INT me, INT ms) { DK(KP555570233, +0.555570233019602224742830813948532874374937191); DK(KP831469612, +0.831469612302545237078788377617905756738560812); DK(KP980785280, +0.980785280403230449126182236134239036973933731); DK(KP195090322, +0.195090322016128267848284868477022240927691618); DK(KP382683432, +0.382683432365089771728459984030398866761344562); DK(KP923879532, +0.923879532511286756128183189396788286822416626); DK(KP707106781, +0.707106781186547524400844362104849039284835938); { INT m; for (m = mb, W = W + ((mb - 1) * 8); m < me; m = m + 1, cr = cr + ms, ci = ci - ms, W = W + 8, MAKE_VOLATILE_STRIDE(64, rs)) { E T2, T5, T3, T6, T8, TM, TO, Td, T9, Te, Th, Tl, TD, TH, T1y; E T1H, T15, T1A, T11, T1F, T1n, T1p, T2q, T2I, T2u, T2K, T2V, T3b, T2Z, T3d; E Tu, Ty, T3l, T3n, T1t, T1v, T2f, T2h, T1a, T1e, T32, T34, T1W, T1Y, T2C; E T2E, Tg, TR, Tk, TS, Tm, TV, To, TT, T1M, T21, T1P, T22, T1Q, T25; E T1S, T23; { E Ts, T1d, Tx, T18, Tt, T1c, Tw, T19, TB, T14, TG, TZ, TC, T13, TF; E T10; { E T4, Tc, T7, Tb; T2 = W[0]; T5 = W[1]; T3 = W[2]; T6 = W[3]; T4 = T2 * T3; Tc = T5 * T3; T7 = T5 * T6; Tb = T2 * T6; T8 = T4 + T7; TM = T4 - T7; TO = Tb + Tc; Td = Tb - Tc; T9 = W[4]; Ts = T2 * T9; T1d = T6 * T9; Tx = T5 * T9; T18 = T3 * T9; Te = W[5]; Tt = T5 * Te; T1c = T3 * Te; Tw = T2 * Te; T19 = T6 * Te; Th = W[6]; TB = T3 * Th; T14 = T5 * Th; TG = T6 * Th; TZ = T2 * Th; Tl = W[7]; TC = T6 * Tl; T13 = T2 * Tl; TF = T3 * Tl; T10 = T5 * Tl; } TD = TB + TC; TH = TF - TG; T1y = TZ + T10; T1H = TF + TG; T15 = T13 + T14; T1A = T13 - T14; T11 = TZ - T10; T1F = TB - TC; T1n = FMA(T9, Th, Te * Tl); T1p = FNMS(Te, Th, T9 * Tl); { E T2o, T2p, T2s, T2t; T2o = T8 * Th; T2p = Td * Tl; T2q = T2o + T2p; T2I = T2o - T2p; T2s = T8 * Tl; T2t = Td * Th; T2u = T2s - T2t; T2K = T2s + T2t; } { E T2T, T2U, T2X, T2Y; T2T = TM * Th; T2U = TO * Tl; T2V = T2T - T2U; T3b = T2T + T2U; T2X = TM * Tl; T2Y = TO * Th; T2Z = T2X + T2Y; T3d = T2X - T2Y; Tu = Ts + Tt; Ty = Tw - Tx; T3l = FMA(Tu, Th, Ty * Tl); T3n = FNMS(Ty, Th, Tu * Tl); } T1t = Ts - Tt; T1v = Tw + Tx; T2f = FMA(T1t, Th, T1v * Tl); T2h = FNMS(T1v, Th, T1t * Tl); T1a = T18 - T19; T1e = T1c + T1d; T32 = FMA(T1a, Th, T1e * Tl); T34 = FNMS(T1e, Th, T1a * Tl); T1W = T18 + T19; T1Y = T1c - T1d; T2C = FMA(T1W, Th, T1Y * Tl); T2E = FNMS(T1Y, Th, T1W * Tl); { E Ta, Tf, Ti, Tj; Ta = T8 * T9; Tf = Td * Te; Tg = Ta - Tf; TR = Ta + Tf; Ti = T8 * Te; Tj = Td * T9; Tk = Ti + Tj; TS = Ti - Tj; } Tm = FMA(Tg, Th, Tk * Tl); TV = FNMS(TS, Th, TR * Tl); To = FNMS(Tk, Th, Tg * Tl); TT = FMA(TR, Th, TS * Tl); { E T1K, T1L, T1N, T1O; T1K = TM * T9; T1L = TO * Te; T1M = T1K - T1L; T21 = T1K + T1L; T1N = TM * Te; T1O = TO * T9; T1P = T1N + T1O; T22 = T1N - T1O; } T1Q = FMA(T1M, Th, T1P * Tl); T25 = FNMS(T22, Th, T21 * Tl); T1S = FNMS(T1P, Th, T1M * Tl); T23 = FMA(T21, Th, T22 * Tl); } { E TL, T6f, T8c, T8q, T3F, T5t, T7I, T7W, T2y, T6B, T6y, T7j, T4k, T5G, T4B; E T5J, T3h, T6H, T6O, T7o, T4L, T5Q, T52, T5N, T1i, T7V, T6i, T7D, T3K, T5u; E T3P, T5v, T1E, T6k, T6n, T7f, T3W, T5z, T41, T5y, T29, T6p, T6s, T7e, T47; E T5C, T4c, T5B, T2R, T6z, T6E, T7k, T4v, T5K, T4E, T5H, T3y, T6P, T6K, T7p; E T4W, T5O, T55, T5R; { E T1, T7G, Tq, T7F, TA, T3C, TJ, T3D, Tn, Tp; T1 = cr[0]; T7G = ci[0]; Tn = cr[WS(rs, 16)]; Tp = ci[WS(rs, 16)]; Tq = FMA(Tm, Tn, To * Tp); T7F = FNMS(To, Tn, Tm * Tp); { E Tv, Tz, TE, TI; Tv = cr[WS(rs, 8)]; Tz = ci[WS(rs, 8)]; TA = FMA(Tu, Tv, Ty * Tz); T3C = FNMS(Ty, Tv, Tu * Tz); TE = cr[WS(rs, 24)]; TI = ci[WS(rs, 24)]; TJ = FMA(TD, TE, TH * TI); T3D = FNMS(TH, TE, TD * TI); } { E Tr, TK, T8a, T8b; Tr = T1 + Tq; TK = TA + TJ; TL = Tr + TK; T6f = Tr - TK; T8a = TA - TJ; T8b = T7G - T7F; T8c = T8a + T8b; T8q = T8b - T8a; } { E T3B, T3E, T7E, T7H; T3B = T1 - Tq; T3E = T3C - T3D; T3F = T3B + T3E; T5t = T3B - T3E; T7E = T3C + T3D; T7H = T7F + T7G; T7I = T7E + T7H; T7W = T7H - T7E; } } { E T2e, T4x, T2w, T4i, T2j, T4y, T2n, T4h; { E T2c, T2d, T2r, T2v; T2c = cr[WS(rs, 1)]; T2d = ci[WS(rs, 1)]; T2e = FMA(T2, T2c, T5 * T2d); T4x = FNMS(T5, T2c, T2 * T2d); T2r = cr[WS(rs, 25)]; T2v = ci[WS(rs, 25)]; T2w = FMA(T2q, T2r, T2u * T2v); T4i = FNMS(T2u, T2r, T2q * T2v); } { E T2g, T2i, T2l, T2m; T2g = cr[WS(rs, 17)]; T2i = ci[WS(rs, 17)]; T2j = FMA(T2f, T2g, T2h * T2i); T4y = FNMS(T2h, T2g, T2f * T2i); T2l = cr[WS(rs, 9)]; T2m = ci[WS(rs, 9)]; T2n = FMA(T9, T2l, Te * T2m); T4h = FNMS(Te, T2l, T9 * T2m); } { E T2k, T2x, T6w, T6x; T2k = T2e + T2j; T2x = T2n + T2w; T2y = T2k + T2x; T6B = T2k - T2x; T6w = T4x + T4y; T6x = T4h + T4i; T6y = T6w - T6x; T7j = T6w + T6x; } { E T4g, T4j, T4z, T4A; T4g = T2e - T2j; T4j = T4h - T4i; T4k = T4g + T4j; T5G = T4g - T4j; T4z = T4x - T4y; T4A = T2n - T2w; T4B = T4z - T4A; T5J = T4z + T4A; } } { E T31, T4H, T3f, T50, T36, T4I, T3a, T4Z; { E T2W, T30, T3c, T3e; T2W = cr[WS(rs, 31)]; T30 = ci[WS(rs, 31)]; T31 = FMA(T2V, T2W, T2Z * T30); T4H = FNMS(T2Z, T2W, T2V * T30); T3c = cr[WS(rs, 23)]; T3e = ci[WS(rs, 23)]; T3f = FMA(T3b, T3c, T3d * T3e); T50 = FNMS(T3d, T3c, T3b * T3e); } { E T33, T35, T38, T39; T33 = cr[WS(rs, 15)]; T35 = ci[WS(rs, 15)]; T36 = FMA(T32, T33, T34 * T35); T4I = FNMS(T34, T33, T32 * T35); T38 = cr[WS(rs, 7)]; T39 = ci[WS(rs, 7)]; T3a = FMA(TR, T38, TS * T39); T4Z = FNMS(TS, T38, TR * T39); } { E T37, T3g, T6M, T6N; T37 = T31 + T36; T3g = T3a + T3f; T3h = T37 + T3g; T6H = T37 - T3g; T6M = T4H + T4I; T6N = T4Z + T50; T6O = T6M - T6N; T7o = T6M + T6N; } { E T4J, T4K, T4Y, T51; T4J = T4H - T4I; T4K = T3a - T3f; T4L = T4J - T4K; T5Q = T4J + T4K; T4Y = T31 - T36; T51 = T4Z - T50; T52 = T4Y + T51; T5N = T4Y - T51; } } { E TQ, T3H, T1g, T3N, TX, T3I, T17, T3M; { E TN, TP, T1b, T1f; TN = cr[WS(rs, 4)]; TP = ci[WS(rs, 4)]; TQ = FMA(TM, TN, TO * TP); T3H = FNMS(TO, TN, TM * TP); T1b = cr[WS(rs, 12)]; T1f = ci[WS(rs, 12)]; T1g = FMA(T1a, T1b, T1e * T1f); T3N = FNMS(T1e, T1b, T1a * T1f); } { E TU, TW, T12, T16; TU = cr[WS(rs, 20)]; TW = ci[WS(rs, 20)]; TX = FMA(TT, TU, TV * TW); T3I = FNMS(TV, TU, TT * TW); T12 = cr[WS(rs, 28)]; T16 = ci[WS(rs, 28)]; T17 = FMA(T11, T12, T15 * T16); T3M = FNMS(T15, T12, T11 * T16); } { E TY, T1h, T6g, T6h; TY = TQ + TX; T1h = T17 + T1g; T1i = TY + T1h; T7V = TY - T1h; T6g = T3M + T3N; T6h = T3H + T3I; T6i = T6g - T6h; T7D = T6h + T6g; } { E T3G, T3J, T3L, T3O; T3G = TQ - TX; T3J = T3H - T3I; T3K = T3G + T3J; T5u = T3G - T3J; T3L = T17 - T1g; T3O = T3M - T3N; T3P = T3L - T3O; T5v = T3L + T3O; } } { E T1m, T3X, T1C, T3U, T1r, T3Y, T1x, T3T; { E T1k, T1l, T1z, T1B; T1k = cr[WS(rs, 2)]; T1l = ci[WS(rs, 2)]; T1m = FMA(T8, T1k, Td * T1l); T3X = FNMS(Td, T1k, T8 * T1l); T1z = cr[WS(rs, 26)]; T1B = ci[WS(rs, 26)]; T1C = FMA(T1y, T1z, T1A * T1B); T3U = FNMS(T1A, T1z, T1y * T1B); } { E T1o, T1q, T1u, T1w; T1o = cr[WS(rs, 18)]; T1q = ci[WS(rs, 18)]; T1r = FMA(T1n, T1o, T1p * T1q); T3Y = FNMS(T1p, T1o, T1n * T1q); T1u = cr[WS(rs, 10)]; T1w = ci[WS(rs, 10)]; T1x = FMA(T1t, T1u, T1v * T1w); T3T = FNMS(T1v, T1u, T1t * T1w); } { E T1s, T1D, T6l, T6m; T1s = T1m + T1r; T1D = T1x + T1C; T1E = T1s + T1D; T6k = T1s - T1D; T6l = T3X + T3Y; T6m = T3T + T3U; T6n = T6l - T6m; T7f = T6l + T6m; } { E T3S, T3V, T3Z, T40; T3S = T1m - T1r; T3V = T3T - T3U; T3W = T3S + T3V; T5z = T3S - T3V; T3Z = T3X - T3Y; T40 = T1x - T1C; T41 = T3Z - T40; T5y = T3Z + T40; } } { E T1J, T43, T27, T4a, T1U, T44, T20, T49; { E T1G, T1I, T24, T26; T1G = cr[WS(rs, 30)]; T1I = ci[WS(rs, 30)]; T1J = FMA(T1F, T1G, T1H * T1I); T43 = FNMS(T1H, T1G, T1F * T1I); T24 = cr[WS(rs, 22)]; T26 = ci[WS(rs, 22)]; T27 = FMA(T23, T24, T25 * T26); T4a = FNMS(T25, T24, T23 * T26); } { E T1R, T1T, T1X, T1Z; T1R = cr[WS(rs, 14)]; T1T = ci[WS(rs, 14)]; T1U = FMA(T1Q, T1R, T1S * T1T); T44 = FNMS(T1S, T1R, T1Q * T1T); T1X = cr[WS(rs, 6)]; T1Z = ci[WS(rs, 6)]; T20 = FMA(T1W, T1X, T1Y * T1Z); T49 = FNMS(T1Y, T1X, T1W * T1Z); } { E T1V, T28, T6q, T6r; T1V = T1J + T1U; T28 = T20 + T27; T29 = T1V + T28; T6p = T1V - T28; T6q = T43 + T44; T6r = T49 + T4a; T6s = T6q - T6r; T7e = T6q + T6r; } { E T45, T46, T48, T4b; T45 = T43 - T44; T46 = T20 - T27; T47 = T45 - T46; T5C = T45 + T46; T48 = T1J - T1U; T4b = T49 - T4a; T4c = T48 + T4b; T5B = T48 - T4b; } } { E T2B, T4m, T2G, T4n, T4l, T4o, T2M, T4q, T2P, T4r, T4s, T4t; { E T2z, T2A, T2D, T2F; T2z = cr[WS(rs, 5)]; T2A = ci[WS(rs, 5)]; T2B = FMA(T21, T2z, T22 * T2A); T4m = FNMS(T22, T2z, T21 * T2A); T2D = cr[WS(rs, 21)]; T2F = ci[WS(rs, 21)]; T2G = FMA(T2C, T2D, T2E * T2F); T4n = FNMS(T2E, T2D, T2C * T2F); } T4l = T2B - T2G; T4o = T4m - T4n; { E T2J, T2L, T2N, T2O; T2J = cr[WS(rs, 29)]; T2L = ci[WS(rs, 29)]; T2M = FMA(T2I, T2J, T2K * T2L); T4q = FNMS(T2K, T2J, T2I * T2L); T2N = cr[WS(rs, 13)]; T2O = ci[WS(rs, 13)]; T2P = FMA(T1M, T2N, T1P * T2O); T4r = FNMS(T1P, T2N, T1M * T2O); } T4s = T4q - T4r; T4t = T2M - T2P; { E T2H, T2Q, T6C, T6D; T2H = T2B + T2G; T2Q = T2M + T2P; T2R = T2H + T2Q; T6z = T2H - T2Q; T6C = T4q + T4r; T6D = T4m + T4n; T6E = T6C - T6D; T7k = T6D + T6C; } { E T4p, T4u, T4C, T4D; T4p = T4l + T4o; T4u = T4s - T4t; T4v = KP707106781 * (T4p - T4u); T5K = KP707106781 * (T4p + T4u); T4C = T4t + T4s; T4D = T4l - T4o; T4E = KP707106781 * (T4C - T4D); T5H = KP707106781 * (T4D + T4C); } } { E T3k, T4S, T3p, T4T, T4R, T4U, T3t, T4N, T3w, T4O, T4M, T4P; { E T3i, T3j, T3m, T3o; T3i = cr[WS(rs, 3)]; T3j = ci[WS(rs, 3)]; T3k = FMA(T3, T3i, T6 * T3j); T4S = FNMS(T6, T3i, T3 * T3j); T3m = cr[WS(rs, 19)]; T3o = ci[WS(rs, 19)]; T3p = FMA(T3l, T3m, T3n * T3o); T4T = FNMS(T3n, T3m, T3l * T3o); } T4R = T3k - T3p; T4U = T4S - T4T; { E T3r, T3s, T3u, T3v; T3r = cr[WS(rs, 27)]; T3s = ci[WS(rs, 27)]; T3t = FMA(Th, T3r, Tl * T3s); T4N = FNMS(Tl, T3r, Th * T3s); T3u = cr[WS(rs, 11)]; T3v = ci[WS(rs, 11)]; T3w = FMA(Tg, T3u, Tk * T3v); T4O = FNMS(Tk, T3u, Tg * T3v); } T4M = T3t - T3w; T4P = T4N - T4O; { E T3q, T3x, T6I, T6J; T3q = T3k + T3p; T3x = T3t + T3w; T3y = T3q + T3x; T6P = T3q - T3x; T6I = T4N + T4O; T6J = T4S + T4T; T6K = T6I - T6J; T7p = T6J + T6I; } { E T4Q, T4V, T53, T54; T4Q = T4M + T4P; T4V = T4R - T4U; T4W = KP707106781 * (T4Q - T4V); T5O = KP707106781 * (T4V + T4Q); T53 = T4R + T4U; T54 = T4P - T4M; T55 = KP707106781 * (T53 - T54); T5R = KP707106781 * (T53 + T54); } } { E T2b, T7x, T7K, T7M, T3A, T7L, T7A, T7B; { E T1j, T2a, T7C, T7J; T1j = TL + T1i; T2a = T1E + T29; T2b = T1j + T2a; T7x = T1j - T2a; T7C = T7f + T7e; T7J = T7D + T7I; T7K = T7C + T7J; T7M = T7J - T7C; } { E T2S, T3z, T7y, T7z; T2S = T2y + T2R; T3z = T3h + T3y; T3A = T2S + T3z; T7L = T3z - T2S; T7y = T7o + T7p; T7z = T7j + T7k; T7A = T7y - T7z; T7B = T7z + T7y; } ci[WS(rs, 15)] = T2b - T3A; cr[WS(rs, 24)] = T7L - T7M; ci[WS(rs, 23)] = T7L + T7M; cr[0] = T2b + T3A; cr[WS(rs, 8)] = T7x - T7A; cr[WS(rs, 16)] = T7B - T7K; ci[WS(rs, 31)] = T7B + T7K; ci[WS(rs, 7)] = T7x + T7A; } { E T5x, T5Z, T8d, T8j, T5E, T88, T69, T6d, T5M, T5W, T62, T8i, T66, T6c, T5T; E T5X, T5w, T89; T5w = KP707106781 * (T5u + T5v); T5x = T5t - T5w; T5Z = T5t + T5w; T89 = KP707106781 * (T3K - T3P); T8d = T89 + T8c; T8j = T8c - T89; { E T5A, T5D, T67, T68; T5A = FMA(KP923879532, T5y, KP382683432 * T5z); T5D = FNMS(KP923879532, T5C, KP382683432 * T5B); T5E = T5A + T5D; T88 = T5A - T5D; T67 = T5N + T5O; T68 = T5Q + T5R; T69 = FNMS(KP980785280, T68, KP195090322 * T67); T6d = FMA(KP980785280, T67, KP195090322 * T68); } { E T5I, T5L, T60, T61; T5I = T5G - T5H; T5L = T5J - T5K; T5M = FMA(KP831469612, T5I, KP555570233 * T5L); T5W = FNMS(KP831469612, T5L, KP555570233 * T5I); T60 = FNMS(KP382683432, T5y, KP923879532 * T5z); T61 = FMA(KP382683432, T5C, KP923879532 * T5B); T62 = T60 + T61; T8i = T61 - T60; } { E T64, T65, T5P, T5S; T64 = T5G + T5H; T65 = T5J + T5K; T66 = FMA(KP195090322, T64, KP980785280 * T65); T6c = FNMS(KP195090322, T65, KP980785280 * T64); T5P = T5N - T5O; T5S = T5Q - T5R; T5T = FNMS(KP555570233, T5S, KP831469612 * T5P); T5X = FMA(KP555570233, T5P, KP831469612 * T5S); } { E T5F, T5U, T8h, T8k; T5F = T5x + T5E; T5U = T5M + T5T; ci[WS(rs, 12)] = T5F - T5U; cr[WS(rs, 3)] = T5F + T5U; T8h = T5X - T5W; T8k = T8i + T8j; cr[WS(rs, 19)] = T8h - T8k; ci[WS(rs, 28)] = T8h + T8k; } { E T8l, T8m, T5V, T5Y; T8l = T5T - T5M; T8m = T8j - T8i; cr[WS(rs, 27)] = T8l - T8m; ci[WS(rs, 20)] = T8l + T8m; T5V = T5x - T5E; T5Y = T5W + T5X; cr[WS(rs, 11)] = T5V - T5Y; ci[WS(rs, 4)] = T5V + T5Y; } { E T63, T6a, T87, T8e; T63 = T5Z - T62; T6a = T66 + T69; ci[WS(rs, 8)] = T63 - T6a; cr[WS(rs, 7)] = T63 + T6a; T87 = T69 - T66; T8e = T88 + T8d; cr[WS(rs, 31)] = T87 - T8e; ci[WS(rs, 16)] = T87 + T8e; } { E T8f, T8g, T6b, T6e; T8f = T6d - T6c; T8g = T8d - T88; cr[WS(rs, 23)] = T8f - T8g; ci[WS(rs, 24)] = T8f + T8g; T6b = T5Z + T62; T6e = T6c + T6d; cr[WS(rs, 15)] = T6b - T6e; ci[0] = T6b + T6e; } } { E T7h, T7t, T7Q, T7S, T7m, T7u, T7r, T7v; { E T7d, T7g, T7O, T7P; T7d = TL - T1i; T7g = T7e - T7f; T7h = T7d - T7g; T7t = T7d + T7g; T7O = T1E - T29; T7P = T7I - T7D; T7Q = T7O + T7P; T7S = T7P - T7O; } { E T7i, T7l, T7n, T7q; T7i = T2y - T2R; T7l = T7j - T7k; T7m = T7i + T7l; T7u = T7i - T7l; T7n = T3h - T3y; T7q = T7o - T7p; T7r = T7n - T7q; T7v = T7n + T7q; } { E T7s, T7R, T7w, T7N; T7s = KP707106781 * (T7m + T7r); ci[WS(rs, 11)] = T7h - T7s; cr[WS(rs, 4)] = T7h + T7s; T7R = KP707106781 * (T7v - T7u); cr[WS(rs, 20)] = T7R - T7S; ci[WS(rs, 27)] = T7R + T7S; T7w = KP707106781 * (T7u + T7v); cr[WS(rs, 12)] = T7t - T7w; ci[WS(rs, 3)] = T7t + T7w; T7N = KP707106781 * (T7r - T7m); cr[WS(rs, 28)] = T7N - T7Q; ci[WS(rs, 19)] = T7N + T7Q; } } { E T6j, T7X, T83, T6X, T6u, T7U, T77, T7b, T70, T82, T6G, T6U, T74, T7a, T6R; E T6V; { E T6o, T6t, T6A, T6F; T6j = T6f - T6i; T7X = T7V + T7W; T83 = T7W - T7V; T6X = T6f + T6i; T6o = T6k + T6n; T6t = T6p - T6s; T6u = KP707106781 * (T6o + T6t); T7U = KP707106781 * (T6o - T6t); { E T75, T76, T6Y, T6Z; T75 = T6O + T6P; T76 = T6H + T6K; T77 = FMA(KP382683432, T75, KP923879532 * T76); T7b = FNMS(KP923879532, T75, KP382683432 * T76); T6Y = T6k - T6n; T6Z = T6p + T6s; T70 = KP707106781 * (T6Y + T6Z); T82 = KP707106781 * (T6Z - T6Y); } T6A = T6y - T6z; T6F = T6B - T6E; T6G = FMA(KP382683432, T6A, KP923879532 * T6F); T6U = FNMS(KP923879532, T6A, KP382683432 * T6F); { E T72, T73, T6L, T6Q; T72 = T6B + T6E; T73 = T6y + T6z; T74 = FNMS(KP382683432, T73, KP923879532 * T72); T7a = FMA(KP923879532, T73, KP382683432 * T72); T6L = T6H - T6K; T6Q = T6O - T6P; T6R = FNMS(KP382683432, T6Q, KP923879532 * T6L); T6V = FMA(KP923879532, T6Q, KP382683432 * T6L); } } { E T6v, T6S, T81, T84; T6v = T6j + T6u; T6S = T6G + T6R; ci[WS(rs, 13)] = T6v - T6S; cr[WS(rs, 2)] = T6v + T6S; T81 = T6V - T6U; T84 = T82 + T83; cr[WS(rs, 18)] = T81 - T84; ci[WS(rs, 29)] = T81 + T84; } { E T85, T86, T6T, T6W; T85 = T6R - T6G; T86 = T83 - T82; cr[WS(rs, 26)] = T85 - T86; ci[WS(rs, 21)] = T85 + T86; T6T = T6j - T6u; T6W = T6U + T6V; cr[WS(rs, 10)] = T6T - T6W; ci[WS(rs, 5)] = T6T + T6W; } { E T71, T78, T7T, T7Y; T71 = T6X + T70; T78 = T74 + T77; cr[WS(rs, 14)] = T71 - T78; ci[WS(rs, 1)] = T71 + T78; T7T = T7b - T7a; T7Y = T7U + T7X; cr[WS(rs, 30)] = T7T - T7Y; ci[WS(rs, 17)] = T7T + T7Y; } { E T7Z, T80, T79, T7c; T7Z = T77 - T74; T80 = T7X - T7U; cr[WS(rs, 22)] = T7Z - T80; ci[WS(rs, 25)] = T7Z + T80; T79 = T6X - T70; T7c = T7a + T7b; ci[WS(rs, 9)] = T79 - T7c; cr[WS(rs, 6)] = T79 + T7c; } } { E T3R, T5d, T8r, T8x, T4e, T8o, T5n, T5r, T4G, T5a, T5g, T8w, T5k, T5q, T57; E T5b, T3Q, T8p; T3Q = KP707106781 * (T3K + T3P); T3R = T3F - T3Q; T5d = T3F + T3Q; T8p = KP707106781 * (T5v - T5u); T8r = T8p + T8q; T8x = T8q - T8p; { E T42, T4d, T5l, T5m; T42 = FNMS(KP923879532, T41, KP382683432 * T3W); T4d = FMA(KP923879532, T47, KP382683432 * T4c); T4e = T42 + T4d; T8o = T4d - T42; T5l = T52 + T55; T5m = T4L + T4W; T5n = FNMS(KP195090322, T5m, KP980785280 * T5l); T5r = FMA(KP980785280, T5m, KP195090322 * T5l); } { E T4w, T4F, T5e, T5f; T4w = T4k - T4v; T4F = T4B - T4E; T4G = FNMS(KP555570233, T4F, KP831469612 * T4w); T5a = FMA(KP831469612, T4F, KP555570233 * T4w); T5e = FMA(KP382683432, T41, KP923879532 * T3W); T5f = FNMS(KP382683432, T47, KP923879532 * T4c); T5g = T5e + T5f; T8w = T5e - T5f; } { E T5i, T5j, T4X, T56; T5i = T4B + T4E; T5j = T4k + T4v; T5k = FMA(KP195090322, T5i, KP980785280 * T5j); T5q = FNMS(KP980785280, T5i, KP195090322 * T5j); T4X = T4L - T4W; T56 = T52 - T55; T57 = FMA(KP555570233, T4X, KP831469612 * T56); T5b = FNMS(KP831469612, T4X, KP555570233 * T56); } { E T4f, T58, T8v, T8y; T4f = T3R + T4e; T58 = T4G + T57; cr[WS(rs, 13)] = T4f - T58; ci[WS(rs, 2)] = T4f + T58; T8v = T5b - T5a; T8y = T8w + T8x; cr[WS(rs, 29)] = T8v - T8y; ci[WS(rs, 18)] = T8v + T8y; } { E T8z, T8A, T59, T5c; T8z = T57 - T4G; T8A = T8x - T8w; cr[WS(rs, 21)] = T8z - T8A; ci[WS(rs, 26)] = T8z + T8A; T59 = T3R - T4e; T5c = T5a + T5b; ci[WS(rs, 10)] = T59 - T5c; cr[WS(rs, 5)] = T59 + T5c; } { E T5h, T5o, T8n, T8s; T5h = T5d + T5g; T5o = T5k + T5n; ci[WS(rs, 14)] = T5h - T5o; cr[WS(rs, 1)] = T5h + T5o; T8n = T5r - T5q; T8s = T8o + T8r; cr[WS(rs, 17)] = T8n - T8s; ci[WS(rs, 30)] = T8n + T8s; } { E T8t, T8u, T5p, T5s; T8t = T5n - T5k; T8u = T8r - T8o; cr[WS(rs, 25)] = T8t - T8u; ci[WS(rs, 22)] = T8t + T8u; T5p = T5d - T5g; T5s = T5q + T5r; cr[WS(rs, 9)] = T5p - T5s; ci[WS(rs, 6)] = T5p + T5s; } } } } } } static const tw_instr twinstr[] = { {TW_CEXP, 1, 1}, {TW_CEXP, 1, 3}, {TW_CEXP, 1, 9}, {TW_CEXP, 1, 27}, {TW_NEXT, 1, 0} }; static const hc2hc_desc desc = { 32, "hf2_32", twinstr, &GENUS, {376, 168, 112, 0} }; void X(codelet_hf2_32) (planner *p) { X(khc2hc_register) (p, hf2_32, &desc); } #endif /* HAVE_FMA */ fftw-3.3.3/rdft/scalar/r2cf/hf2_5.c0000644000175000017500000001730012054410647013522 00000000000000/* * Copyright (c) 2003, 2007-11 Matteo Frigo * Copyright (c) 2003, 2007-11 Massachusetts Institute of Technology * * This program is free software; you can redistribute it and/or modify * it under the terms of the GNU General Public License as published by * the Free Software Foundation; either version 2 of the License, or * (at your option) any later version. * * This program is distributed in the hope that it will be useful, * but WITHOUT ANY WARRANTY; without even the implied warranty of * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the * GNU General Public License for more details. * * You should have received a copy of the GNU General Public License * along with this program; if not, write to the Free Software * Foundation, Inc., 51 Franklin Street, Fifth Floor, Boston, MA 02110-1301 USA * */ /* This file was automatically generated --- DO NOT EDIT */ /* Generated on Sun Nov 25 07:40:07 EST 2012 */ #include "codelet-rdft.h" #ifdef HAVE_FMA /* Generated by: ../../../genfft/gen_hc2hc.native -fma -reorder-insns -schedule-for-pipeline -compact -variables 4 -pipeline-latency 4 -twiddle-log3 -precompute-twiddles -n 5 -dit -name hf2_5 -include hf.h */ /* * This function contains 44 FP additions, 40 FP multiplications, * (or, 14 additions, 10 multiplications, 30 fused multiply/add), * 47 stack variables, 4 constants, and 20 memory accesses */ #include "hf.h" static void hf2_5(R *cr, R *ci, const R *W, stride rs, INT mb, INT me, INT ms) { DK(KP951056516, +0.951056516295153572116439333379382143405698634); DK(KP559016994, +0.559016994374947424102293417182819058860154590); DK(KP250000000, +0.250000000000000000000000000000000000000000000); DK(KP618033988, +0.618033988749894848204586834365638117720309180); { INT m; for (m = mb, W = W + ((mb - 1) * 4); m < me; m = m + 1, cr = cr + ms, ci = ci - ms, W = W + 4, MAKE_VOLATILE_STRIDE(10, rs)) { E Ta, T1, TL, Tp, TT, Ti, TM, TC, To, TE, Ts, TF, T2, T8, T5; E TS, Tt, TG; T2 = W[0]; Ta = W[3]; T8 = W[2]; T5 = W[1]; { E Tq, Tr, Te, T9; T1 = cr[0]; Te = T2 * Ta; T9 = T2 * T8; TL = ci[0]; { E T3, Tf, Tm, Tj, Tb, T4, T6, Tc, Tg; T3 = cr[WS(rs, 1)]; Tf = FMA(T5, T8, Te); Tm = FNMS(T5, T8, Te); Tj = FMA(T5, Ta, T9); Tb = FNMS(T5, Ta, T9); T4 = T2 * T3; T6 = ci[WS(rs, 1)]; Tc = cr[WS(rs, 4)]; Tg = ci[WS(rs, 4)]; { E Tk, Tl, Tn, TD; { E T7, Tz, Th, TB, Ty, Td, TA; Tk = cr[WS(rs, 2)]; T7 = FMA(T5, T6, T4); Ty = T2 * T6; Td = Tb * Tc; TA = Tb * Tg; Tl = Tj * Tk; Tz = FNMS(T5, T3, Ty); Th = FMA(Tf, Tg, Td); TB = FNMS(Tf, Tc, TA); Tn = ci[WS(rs, 2)]; Tp = cr[WS(rs, 3)]; TT = Th - T7; Ti = T7 + Th; TM = Tz + TB; TC = Tz - TB; TD = Tj * Tn; Tq = T8 * Tp; Tr = ci[WS(rs, 3)]; } To = FMA(Tm, Tn, Tl); TE = FNMS(Tm, Tk, TD); } } Ts = FMA(Ta, Tr, Tq); TF = T8 * Tr; } TS = To - Ts; Tt = To + Ts; TG = FNMS(Ta, Tp, TF); { E TU, TW, TV, TR, Tw, Tu; TU = FMA(KP618033988, TT, TS); TW = FNMS(KP618033988, TS, TT); Tw = Ti - Tt; Tu = Ti + Tt; { E TN, TH, Tv, TI, TK; TN = TE + TG; TH = TE - TG; cr[0] = T1 + Tu; Tv = FNMS(KP250000000, Tu, T1); TI = FMA(KP618033988, TH, TC); TK = FNMS(KP618033988, TC, TH); { E TQ, TO, Tx, TJ, TP; TQ = TM - TN; TO = TM + TN; Tx = FMA(KP559016994, Tw, Tv); TJ = FNMS(KP559016994, Tw, Tv); ci[WS(rs, 4)] = TO + TL; TP = FNMS(KP250000000, TO, TL); ci[WS(rs, 1)] = FMA(KP951056516, TK, TJ); cr[WS(rs, 2)] = FNMS(KP951056516, TK, TJ); cr[WS(rs, 1)] = FMA(KP951056516, TI, Tx); ci[0] = FNMS(KP951056516, TI, Tx); TV = FMA(KP559016994, TQ, TP); TR = FNMS(KP559016994, TQ, TP); } } ci[WS(rs, 2)] = FMA(KP951056516, TU, TR); cr[WS(rs, 3)] = FMS(KP951056516, TU, TR); ci[WS(rs, 3)] = FMA(KP951056516, TW, TV); cr[WS(rs, 4)] = FMS(KP951056516, TW, TV); } } } } static const tw_instr twinstr[] = { {TW_CEXP, 1, 1}, {TW_CEXP, 1, 3}, {TW_NEXT, 1, 0} }; static const hc2hc_desc desc = { 5, "hf2_5", twinstr, &GENUS, {14, 10, 30, 0} }; void X(codelet_hf2_5) (planner *p) { X(khc2hc_register) (p, hf2_5, &desc); } #else /* HAVE_FMA */ /* Generated by: ../../../genfft/gen_hc2hc.native -compact -variables 4 -pipeline-latency 4 -twiddle-log3 -precompute-twiddles -n 5 -dit -name hf2_5 -include hf.h */ /* * This function contains 44 FP additions, 32 FP multiplications, * (or, 30 additions, 18 multiplications, 14 fused multiply/add), * 37 stack variables, 4 constants, and 20 memory accesses */ #include "hf.h" static void hf2_5(R *cr, R *ci, const R *W, stride rs, INT mb, INT me, INT ms) { DK(KP250000000, +0.250000000000000000000000000000000000000000000); DK(KP559016994, +0.559016994374947424102293417182819058860154590); DK(KP587785252, +0.587785252292473129168705954639072768597652438); DK(KP951056516, +0.951056516295153572116439333379382143405698634); { INT m; for (m = mb, W = W + ((mb - 1) * 4); m < me; m = m + 1, cr = cr + ms, ci = ci - ms, W = W + 4, MAKE_VOLATILE_STRIDE(10, rs)) { E T2, T4, T7, T9, Tb, Tl, Tf, Tj; { E T8, Te, Ta, Td; T2 = W[0]; T4 = W[1]; T7 = W[2]; T9 = W[3]; T8 = T2 * T7; Te = T4 * T7; Ta = T4 * T9; Td = T2 * T9; Tb = T8 - Ta; Tl = Td - Te; Tf = Td + Te; Tj = T8 + Ta; } { E T1, TI, Ty, TB, TG, TF, TJ, TK, TL, Ti, Tr, Ts; T1 = cr[0]; TI = ci[0]; { E T6, Tw, Tq, TA, Th, Tx, Tn, Tz; { E T3, T5, To, Tp; T3 = cr[WS(rs, 1)]; T5 = ci[WS(rs, 1)]; T6 = FMA(T2, T3, T4 * T5); Tw = FNMS(T4, T3, T2 * T5); To = cr[WS(rs, 3)]; Tp = ci[WS(rs, 3)]; Tq = FMA(T7, To, T9 * Tp); TA = FNMS(T9, To, T7 * Tp); } { E Tc, Tg, Tk, Tm; Tc = cr[WS(rs, 4)]; Tg = ci[WS(rs, 4)]; Th = FMA(Tb, Tc, Tf * Tg); Tx = FNMS(Tf, Tc, Tb * Tg); Tk = cr[WS(rs, 2)]; Tm = ci[WS(rs, 2)]; Tn = FMA(Tj, Tk, Tl * Tm); Tz = FNMS(Tl, Tk, Tj * Tm); } Ty = Tw - Tx; TB = Tz - TA; TG = Tn - Tq; TF = Th - T6; TJ = Tw + Tx; TK = Tz + TA; TL = TJ + TK; Ti = T6 + Th; Tr = Tn + Tq; Ts = Ti + Tr; } cr[0] = T1 + Ts; { E TC, TE, Tv, TD, Tt, Tu; TC = FMA(KP951056516, Ty, KP587785252 * TB); TE = FNMS(KP587785252, Ty, KP951056516 * TB); Tt = KP559016994 * (Ti - Tr); Tu = FNMS(KP250000000, Ts, T1); Tv = Tt + Tu; TD = Tu - Tt; ci[0] = Tv - TC; ci[WS(rs, 1)] = TD + TE; cr[WS(rs, 1)] = Tv + TC; cr[WS(rs, 2)] = TD - TE; } ci[WS(rs, 4)] = TL + TI; { E TH, TP, TO, TQ, TM, TN; TH = FMA(KP587785252, TF, KP951056516 * TG); TP = FNMS(KP587785252, TG, KP951056516 * TF); TM = FNMS(KP250000000, TL, TI); TN = KP559016994 * (TJ - TK); TO = TM - TN; TQ = TN + TM; cr[WS(rs, 3)] = TH - TO; ci[WS(rs, 3)] = TP + TQ; ci[WS(rs, 2)] = TH + TO; cr[WS(rs, 4)] = TP - TQ; } } } } } static const tw_instr twinstr[] = { {TW_CEXP, 1, 1}, {TW_CEXP, 1, 3}, {TW_NEXT, 1, 0} }; static const hc2hc_desc desc = { 5, "hf2_5", twinstr, &GENUS, {30, 18, 14, 0} }; void X(codelet_hf2_5) (planner *p) { X(khc2hc_register) (p, hf2_5, &desc); } #endif /* HAVE_FMA */ fftw-3.3.3/rdft/scalar/r2cf/hf2_20.c0000644000175000017500000007321612054410651013602 00000000000000/* * Copyright (c) 2003, 2007-11 Matteo Frigo * Copyright (c) 2003, 2007-11 Massachusetts Institute of Technology * * This program is free software; you can redistribute it and/or modify * it under the terms of the GNU General Public License as published by * the Free Software Foundation; either version 2 of the License, or * (at your option) any later version. * * This program is distributed in the hope that it will be useful, * but WITHOUT ANY WARRANTY; without even the implied warranty of * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the * GNU General Public License for more details. * * You should have received a copy of the GNU General Public License * along with this program; if not, write to the Free Software * Foundation, Inc., 51 Franklin Street, Fifth Floor, Boston, MA 02110-1301 USA * */ /* This file was automatically generated --- DO NOT EDIT */ /* Generated on Sun Nov 25 07:40:07 EST 2012 */ #include "codelet-rdft.h" #ifdef HAVE_FMA /* Generated by: ../../../genfft/gen_hc2hc.native -fma -reorder-insns -schedule-for-pipeline -compact -variables 4 -pipeline-latency 4 -twiddle-log3 -precompute-twiddles -n 20 -dit -name hf2_20 -include hf.h */ /* * This function contains 276 FP additions, 198 FP multiplications, * (or, 136 additions, 58 multiplications, 140 fused multiply/add), * 146 stack variables, 4 constants, and 80 memory accesses */ #include "hf.h" static void hf2_20(R *cr, R *ci, const R *W, stride rs, INT mb, INT me, INT ms) { DK(KP951056516, +0.951056516295153572116439333379382143405698634); DK(KP559016994, +0.559016994374947424102293417182819058860154590); DK(KP250000000, +0.250000000000000000000000000000000000000000000); DK(KP618033988, +0.618033988749894848204586834365638117720309180); { INT m; for (m = mb, W = W + ((mb - 1) * 8); m < me; m = m + 1, cr = cr + ms, ci = ci - ms, W = W + 8, MAKE_VOLATILE_STRIDE(40, rs)) { E T5o, T5u, T5w, T5q, T5n, T5p, T5v, T5r; { E T2, Th, Tf, T6, T5, Tl, T1p, T1n, Ti, T3, Tt, Tv, T24, T1f, T1D; E Tb, T1P, Tm, T21, T1b, T7, T1A, Tw, T1H, T13, TA, T1L, T17, T1S, Tq; E T1o, T2g, T1t, T2c, TO, TK; { E T1e, Ta, Tk, Tg; T2 = W[0]; Th = W[3]; Tf = W[2]; T6 = W[5]; T5 = W[1]; Tk = T2 * Th; Tg = T2 * Tf; T1e = Tf * T6; Ta = T2 * T6; Tl = FMA(T5, Tf, Tk); T1p = FNMS(T5, Tf, Tk); T1n = FMA(T5, Th, Tg); Ti = FNMS(T5, Th, Tg); T3 = W[4]; Tt = W[6]; Tv = W[7]; { E Tp, Tj, TN, TJ; Tp = Ti * T6; T24 = FMA(Th, T3, T1e); T1f = FNMS(Th, T3, T1e); T1D = FNMS(T5, T3, Ta); Tb = FMA(T5, T3, Ta); Tj = Ti * T3; { E T1a, T4, Tu, T1G; T1a = Tf * T3; T4 = T2 * T3; Tu = Ti * Tt; T1G = T2 * Tt; { E T12, Tz, T1K, T16; T12 = Tf * Tt; Tz = Ti * Tv; T1K = T2 * Tv; T16 = Tf * Tv; T1P = FNMS(Tl, T6, Tj); Tm = FMA(Tl, T6, Tj); T21 = FNMS(Th, T6, T1a); T1b = FMA(Th, T6, T1a); T7 = FNMS(T5, T6, T4); T1A = FMA(T5, T6, T4); Tw = FMA(Tl, Tv, Tu); T1H = FMA(T5, Tv, T1G); T13 = FMA(Th, Tv, T12); TA = FNMS(Tl, Tt, Tz); T1L = FNMS(T5, Tt, T1K); T17 = FNMS(Th, Tt, T16); T1S = FMA(Tl, T3, Tp); Tq = FNMS(Tl, T3, Tp); } } T1o = T1n * T3; T2g = T1n * Tv; TN = Tm * Tv; TJ = Tm * Tt; T1t = T1n * T6; T2c = T1n * Tt; TO = FNMS(Tq, Tt, TN); TK = FMA(Tq, Tv, TJ); } } { E Te, T2C, T4K, T57, T58, TD, T2H, T4L, T3u, T3Z, T11, T2v, T2P, T3P, T4n; E T4v, T3C, T43, T2r, T2z, T3b, T3T, T4d, T4z, T3J, T42, T20, T2y, T34, T3S; E T4g, T4y, T1c, T19, T1d, T3j, T1w, T2U, T1g, T1j, T1l; { E T2d, T2h, T2k, T1q, T1u, T2n, TL, TI, TM, T3q, TZ, T2N, TP, TS, TU; { E T1, T4J, T8, T9, Tc; T1 = cr[0]; T4J = ci[0]; T8 = cr[WS(rs, 10)]; T2d = FMA(T1p, Tv, T2c); T2h = FNMS(T1p, Tt, T2g); T2k = FMA(T1p, T6, T1o); T1q = FNMS(T1p, T6, T1o); T1u = FMA(T1p, T3, T1t); T2n = FNMS(T1p, T3, T1t); T9 = T7 * T8; Tc = ci[WS(rs, 10)]; { E Tx, Ts, T2F, TC, T2E; { E Tn, Tr, To, T2D, T4I, Ty, TB, Td, T4H; Tn = cr[WS(rs, 5)]; Tr = ci[WS(rs, 5)]; Tx = cr[WS(rs, 15)]; Td = FMA(Tb, Tc, T9); T4H = T7 * Tc; To = Tm * Tn; T2D = Tm * Tr; Te = T1 + Td; T2C = T1 - Td; T4I = FNMS(Tb, T8, T4H); Ty = Tw * Tx; TB = ci[WS(rs, 15)]; Ts = FMA(Tq, Tr, To); T4K = T4I + T4J; T57 = T4J - T4I; T2F = Tw * TB; TC = FMA(TA, TB, Ty); T2E = FNMS(Tq, Tn, T2D); } { E TF, TG, TH, TW, TY, T2G, T3p, TX, T2M; TF = cr[WS(rs, 4)]; T2G = FNMS(TA, Tx, T2F); T58 = Ts - TC; TD = Ts + TC; TG = Ti * TF; T2H = T2E - T2G; T4L = T2E + T2G; TH = ci[WS(rs, 4)]; TW = cr[WS(rs, 19)]; TY = ci[WS(rs, 19)]; TL = cr[WS(rs, 14)]; TI = FMA(Tl, TH, TG); T3p = Ti * TH; TX = Tt * TW; T2M = Tt * TY; TM = TK * TL; T3q = FNMS(Tl, TF, T3p); TZ = FMA(Tv, TY, TX); T2N = FNMS(Tv, TW, T2M); TP = ci[WS(rs, 14)]; TS = cr[WS(rs, 9)]; TU = ci[WS(rs, 9)]; } } } { E T27, T26, T28, T3y, T2p, T39, T29, T2e, T2i; { E T22, T23, T25, T2l, T2o, T3x, T2m, T38; { E TR, T2J, T3s, TV, T2L, T4m, T3t; T22 = cr[WS(rs, 12)]; { E TQ, T3r, TT, T2K; TQ = FMA(TO, TP, TM); T3r = TK * TP; TT = T3 * TS; T2K = T3 * TU; TR = TI + TQ; T2J = TI - TQ; T3s = FNMS(TO, TL, T3r); TV = FMA(T6, TU, TT); T2L = FNMS(T6, TS, T2K); T23 = T21 * T22; } T4m = T3q + T3s; T3t = T3q - T3s; { E T10, T3o, T4l, T2O; T10 = TV + TZ; T3o = TZ - TV; T4l = T2L + T2N; T2O = T2L - T2N; T3u = T3o - T3t; T3Z = T3t + T3o; T11 = TR - T10; T2v = TR + T10; T2P = T2J - T2O; T3P = T2J + T2O; T4n = T4l - T4m; T4v = T4m + T4l; T25 = ci[WS(rs, 12)]; } } T2l = cr[WS(rs, 7)]; T2o = ci[WS(rs, 7)]; T27 = cr[WS(rs, 2)]; T26 = FMA(T24, T25, T23); T3x = T21 * T25; T2m = T2k * T2l; T38 = T2k * T2o; T28 = T1n * T27; T3y = FNMS(T24, T22, T3x); T2p = FMA(T2n, T2o, T2m); T39 = FNMS(T2n, T2l, T38); T29 = ci[WS(rs, 2)]; T2e = cr[WS(rs, 17)]; T2i = ci[WS(rs, 17)]; } { E T1I, T1F, T1J, T3F, T1Y, T32, T1M, T1Q, T1T; { E T1B, T1C, T1E, T1V, T1X, T3E, T1W, T31; { E T2b, T35, T3A, T2j, T37, T4c, T3B; T1B = cr[WS(rs, 8)]; { E T2a, T3z, T2f, T36; T2a = FMA(T1p, T29, T28); T3z = T1n * T29; T2f = T2d * T2e; T36 = T2d * T2i; T2b = T26 + T2a; T35 = T26 - T2a; T3A = FNMS(T1p, T27, T3z); T2j = FMA(T2h, T2i, T2f); T37 = FNMS(T2h, T2e, T36); T1C = T1A * T1B; } T4c = T3y + T3A; T3B = T3y - T3A; { E T2q, T3w, T4b, T3a; T2q = T2j + T2p; T3w = T2p - T2j; T4b = T37 + T39; T3a = T37 - T39; T3C = T3w - T3B; T43 = T3B + T3w; T2r = T2b - T2q; T2z = T2b + T2q; T3b = T35 - T3a; T3T = T35 + T3a; T4d = T4b - T4c; T4z = T4c + T4b; T1E = ci[WS(rs, 8)]; } } T1V = cr[WS(rs, 3)]; T1X = ci[WS(rs, 3)]; T1I = cr[WS(rs, 18)]; T1F = FMA(T1D, T1E, T1C); T3E = T1A * T1E; T1W = Tf * T1V; T31 = Tf * T1X; T1J = T1H * T1I; T3F = FNMS(T1D, T1B, T3E); T1Y = FMA(Th, T1X, T1W); T32 = FNMS(Th, T1V, T31); T1M = ci[WS(rs, 18)]; T1Q = cr[WS(rs, 13)]; T1T = ci[WS(rs, 13)]; } { E T14, T15, T18, T1r, T1v, T3i, T1s, T2T; { E T1O, T2Y, T3H, T1U, T30, T4f, T3I; T14 = cr[WS(rs, 16)]; { E T1N, T3G, T1R, T2Z; T1N = FMA(T1L, T1M, T1J); T3G = T1H * T1M; T1R = T1P * T1Q; T2Z = T1P * T1T; T1O = T1F + T1N; T2Y = T1F - T1N; T3H = FNMS(T1L, T1I, T3G); T1U = FMA(T1S, T1T, T1R); T30 = FNMS(T1S, T1Q, T2Z); T15 = T13 * T14; } T4f = T3F + T3H; T3I = T3F - T3H; { E T1Z, T3D, T4e, T33; T1Z = T1U + T1Y; T3D = T1Y - T1U; T4e = T30 + T32; T33 = T30 - T32; T3J = T3D - T3I; T42 = T3I + T3D; T20 = T1O - T1Z; T2y = T1O + T1Z; T34 = T2Y - T33; T3S = T2Y + T33; T4g = T4e - T4f; T4y = T4f + T4e; T18 = ci[WS(rs, 16)]; } } T1r = cr[WS(rs, 11)]; T1v = ci[WS(rs, 11)]; T1c = cr[WS(rs, 6)]; T19 = FMA(T17, T18, T15); T3i = T13 * T18; T1s = T1q * T1r; T2T = T1q * T1v; T1d = T1b * T1c; T3j = FNMS(T17, T14, T3i); T1w = FMA(T1u, T1v, T1s); T2U = FNMS(T1u, T1r, T2T); T1g = ci[WS(rs, 6)]; T1j = cr[WS(rs, 1)]; T1l = ci[WS(rs, 1)]; } } } } { E T4F, T4Q, T4R, T5a, T4E, T5b, T2I, T5h, T5g, T4W, T4X, T53, T52, T5l, T5m; E T5s, T2X, T3N, T3L, T3c, T5t; { E T2u, T3n, T2w, T2W, T4w, T4r, T4p, T45, T47, T3O, T3R, T4a, T4q, T3U; { E T4h, TE, T40, T3Q, T4k, T1z, T2s, T49, T48; { E T1i, T2Q, T3l, T1m, T2S, T4j, T3m; T4h = T4d - T4g; T4F = T4g + T4d; { E T1h, T3k, T1k, T2R; T1h = FMA(T1f, T1g, T1d); T3k = T1b * T1g; T1k = T2 * T1j; T2R = T2 * T1l; T1i = T19 + T1h; T2Q = T19 - T1h; T3l = FNMS(T1f, T1c, T3k); T1m = FMA(T5, T1l, T1k); T2S = FNMS(T5, T1j, T2R); } TE = Te - TD; T2u = Te + TD; T4j = T3j + T3l; T3m = T3j - T3l; { E T1x, T3h, T4i, T2V, T1y; T1x = T1m + T1w; T3h = T1w - T1m; T4i = T2S + T2U; T2V = T2S - T2U; T3n = T3h - T3m; T40 = T3m + T3h; T1y = T1i - T1x; T2w = T1i + T1x; T2W = T2Q - T2V; T3Q = T2Q + T2V; T4k = T4i - T4j; T4w = T4j + T4i; T4Q = T1y - T11; T1z = T11 + T1y; T2s = T20 + T2r; T4R = T20 - T2r; } } { E T41, T4o, T44, T2t; T5a = T3Z + T40; T41 = T3Z - T40; T4o = T4k - T4n; T4E = T4n + T4k; T5b = T42 + T43; T44 = T42 - T43; T49 = T1z - T2s; T2t = T1z + T2s; T4r = FMA(KP618033988, T4h, T4o); T4p = FNMS(KP618033988, T4o, T4h); T45 = FMA(KP618033988, T44, T41); T47 = FNMS(KP618033988, T41, T44); ci[WS(rs, 9)] = TE + T2t; T48 = FNMS(KP250000000, T2t, TE); } T3O = T2C + T2H; T2I = T2C - T2H; T5h = T3P - T3Q; T3R = T3P + T3Q; T4a = FNMS(KP559016994, T49, T48); T4q = FMA(KP559016994, T49, T48); T3U = T3S + T3T; T5g = T3S - T3T; } { E T2x, T4B, T4D, T2A, T3Y, T46; { E T4x, T3X, T3V, T4A, T3W; T4W = T4v + T4w; T4x = T4v - T4w; ci[WS(rs, 1)] = FMA(KP951056516, T4p, T4a); cr[WS(rs, 2)] = FNMS(KP951056516, T4p, T4a); cr[WS(rs, 6)] = FMA(KP951056516, T4r, T4q); ci[WS(rs, 5)] = FNMS(KP951056516, T4r, T4q); T3X = T3R - T3U; T3V = T3R + T3U; T4A = T4y - T4z; T4X = T4y + T4z; T2x = T2v + T2w; T53 = T2v - T2w; cr[WS(rs, 5)] = T3O + T3V; T3W = FNMS(KP250000000, T3V, T3O); T4B = FMA(KP618033988, T4A, T4x); T4D = FNMS(KP618033988, T4x, T4A); T52 = T2z - T2y; T2A = T2y + T2z; T3Y = FMA(KP559016994, T3X, T3W); T46 = FNMS(KP559016994, T3X, T3W); } { E T3v, T4t, T4s, T3K, T2B, T4u, T4C; T3v = T3n - T3u; T5l = T3u + T3n; T2B = T2x + T2A; T4t = T2x - T2A; cr[WS(rs, 9)] = FNMS(KP951056516, T45, T3Y); cr[WS(rs, 1)] = FMA(KP951056516, T45, T3Y); ci[WS(rs, 6)] = FMA(KP951056516, T47, T46); ci[WS(rs, 2)] = FNMS(KP951056516, T47, T46); cr[0] = T2u + T2B; T4s = FNMS(KP250000000, T2B, T2u); T5m = T3J + T3C; T3K = T3C - T3J; T5s = T2P - T2W; T2X = T2P + T2W; T4u = FMA(KP559016994, T4t, T4s); T4C = FNMS(KP559016994, T4t, T4s); T3N = FNMS(KP618033988, T3v, T3K); T3L = FMA(KP618033988, T3K, T3v); ci[WS(rs, 3)] = FMA(KP951056516, T4B, T4u); cr[WS(rs, 4)] = FNMS(KP951056516, T4B, T4u); cr[WS(rs, 8)] = FMA(KP951056516, T4D, T4C); ci[WS(rs, 7)] = FNMS(KP951056516, T4D, T4C); T3c = T34 + T3b; T5t = T34 - T3b; } } } { E T4V, T5i, T5k, T59, T5e, T5c; { E T4M, T3f, T4U, T4S, T3e, T3d; T4V = T4L + T4K; T4M = T4K - T4L; T3f = T2X - T3c; T3d = T2X + T3c; T4U = FMA(KP618033988, T4Q, T4R); T4S = FNMS(KP618033988, T4R, T4Q); ci[WS(rs, 4)] = T2I + T3d; T3e = FNMS(KP250000000, T3d, T2I); { E T4O, T4N, T3g, T3M, T4G, T4T, T4P; T3g = FMA(KP559016994, T3f, T3e); T3M = FNMS(KP559016994, T3f, T3e); T4O = T4F - T4E; T4G = T4E + T4F; ci[WS(rs, 8)] = FMA(KP951056516, T3L, T3g); ci[0] = FNMS(KP951056516, T3L, T3g); cr[WS(rs, 7)] = FNMS(KP951056516, T3N, T3M); cr[WS(rs, 3)] = FMA(KP951056516, T3N, T3M); cr[WS(rs, 10)] = T4G - T4M; T4N = FMA(KP250000000, T4G, T4M); T5i = FNMS(KP618033988, T5h, T5g); T5k = FMA(KP618033988, T5g, T5h); T59 = T57 - T58; T5o = T58 + T57; T4T = FNMS(KP559016994, T4O, T4N); T4P = FMA(KP559016994, T4O, T4N); ci[WS(rs, 13)] = FMA(KP951056516, T4S, T4P); cr[WS(rs, 14)] = FMS(KP951056516, T4S, T4P); ci[WS(rs, 17)] = FMA(KP951056516, T4U, T4T); cr[WS(rs, 18)] = FMS(KP951056516, T4U, T4T); T5e = T5a - T5b; T5c = T5a + T5b; } } { E T56, T54, T4Y, T50, T5d, T5f, T5j, T4Z, T55, T51; ci[WS(rs, 14)] = T5c + T59; T5d = FNMS(KP250000000, T5c, T59); T56 = FNMS(KP618033988, T52, T53); T54 = FMA(KP618033988, T53, T52); T5f = FNMS(KP559016994, T5e, T5d); T5j = FMA(KP559016994, T5e, T5d); cr[WS(rs, 17)] = -(FMA(KP951056516, T5i, T5f)); cr[WS(rs, 13)] = FMS(KP951056516, T5i, T5f); ci[WS(rs, 18)] = FNMS(KP951056516, T5k, T5j); ci[WS(rs, 10)] = FMA(KP951056516, T5k, T5j); T4Y = T4W + T4X; T50 = T4W - T4X; ci[WS(rs, 19)] = T4Y + T4V; T4Z = FNMS(KP250000000, T4Y, T4V); T5u = FMA(KP618033988, T5t, T5s); T5w = FNMS(KP618033988, T5s, T5t); T55 = FMA(KP559016994, T50, T4Z); T51 = FNMS(KP559016994, T50, T4Z); ci[WS(rs, 11)] = FMA(KP951056516, T54, T51); cr[WS(rs, 12)] = FMS(KP951056516, T54, T51); ci[WS(rs, 15)] = FMA(KP951056516, T56, T55); cr[WS(rs, 16)] = FMS(KP951056516, T56, T55); T5q = T5l - T5m; T5n = T5l + T5m; } } } } } cr[WS(rs, 15)] = T5n - T5o; T5p = FMA(KP250000000, T5n, T5o); T5v = FMA(KP559016994, T5q, T5p); T5r = FNMS(KP559016994, T5q, T5p); cr[WS(rs, 19)] = -(FMA(KP951056516, T5u, T5r)); cr[WS(rs, 11)] = FMS(KP951056516, T5u, T5r); ci[WS(rs, 16)] = FNMS(KP951056516, T5w, T5v); ci[WS(rs, 12)] = FMA(KP951056516, T5w, T5v); } } } static const tw_instr twinstr[] = { {TW_CEXP, 1, 1}, {TW_CEXP, 1, 3}, {TW_CEXP, 1, 9}, {TW_CEXP, 1, 19}, {TW_NEXT, 1, 0} }; static const hc2hc_desc desc = { 20, "hf2_20", twinstr, &GENUS, {136, 58, 140, 0} }; void X(codelet_hf2_20) (planner *p) { X(khc2hc_register) (p, hf2_20, &desc); } #else /* HAVE_FMA */ /* Generated by: ../../../genfft/gen_hc2hc.native -compact -variables 4 -pipeline-latency 4 -twiddle-log3 -precompute-twiddles -n 20 -dit -name hf2_20 -include hf.h */ /* * This function contains 276 FP additions, 164 FP multiplications, * (or, 204 additions, 92 multiplications, 72 fused multiply/add), * 123 stack variables, 4 constants, and 80 memory accesses */ #include "hf.h" static void hf2_20(R *cr, R *ci, const R *W, stride rs, INT mb, INT me, INT ms) { DK(KP587785252, +0.587785252292473129168705954639072768597652438); DK(KP951056516, +0.951056516295153572116439333379382143405698634); DK(KP250000000, +0.250000000000000000000000000000000000000000000); DK(KP559016994, +0.559016994374947424102293417182819058860154590); { INT m; for (m = mb, W = W + ((mb - 1) * 8); m < me; m = m + 1, cr = cr + ms, ci = ci - ms, W = W + 8, MAKE_VOLATILE_STRIDE(40, rs)) { E T2, T5, Tg, Ti, Tk, To, T1h, T1f, T6, T3, T8, T14, T1Q, Tc, T1O; E T1v, T18, T1t, T1n, T24, T1j, T22, Tq, Tu, T1E, T1G, Tx, Ty, Tz, TJ; E T1Z, TB, T1X, T1A, TZ, TL, T1y, TX; { E T7, T16, Ta, T13, T4, T17, Tb, T12; { E Th, Tn, Tj, Tm; T2 = W[0]; T5 = W[1]; Tg = W[2]; Ti = W[3]; Th = T2 * Tg; Tn = T5 * Tg; Tj = T5 * Ti; Tm = T2 * Ti; Tk = Th - Tj; To = Tm + Tn; T1h = Tm - Tn; T1f = Th + Tj; T6 = W[5]; T7 = T5 * T6; T16 = Tg * T6; Ta = T2 * T6; T13 = Ti * T6; T3 = W[4]; T4 = T2 * T3; T17 = Ti * T3; Tb = T5 * T3; T12 = Tg * T3; } T8 = T4 - T7; T14 = T12 + T13; T1Q = T16 + T17; Tc = Ta + Tb; T1O = T12 - T13; T1v = Ta - Tb; T18 = T16 - T17; T1t = T4 + T7; { E T1l, T1m, T1g, T1i; T1l = T1f * T6; T1m = T1h * T3; T1n = T1l + T1m; T24 = T1l - T1m; T1g = T1f * T3; T1i = T1h * T6; T1j = T1g - T1i; T22 = T1g + T1i; { E Tl, Tp, Ts, Tt; Tl = Tk * T3; Tp = To * T6; Tq = Tl + Tp; Ts = Tk * T6; Tt = To * T3; Tu = Ts - Tt; T1E = Tl - Tp; T1G = Ts + Tt; Tx = W[6]; Ty = W[7]; Tz = FMA(Tk, Tx, To * Ty); TJ = FMA(Tq, Tx, Tu * Ty); T1Z = FNMS(T1h, Tx, T1f * Ty); TB = FNMS(To, Tx, Tk * Ty); T1X = FMA(T1f, Tx, T1h * Ty); T1A = FNMS(T5, Tx, T2 * Ty); TZ = FNMS(Ti, Tx, Tg * Ty); TL = FNMS(Tu, Tx, Tq * Ty); T1y = FMA(T2, Tx, T5 * Ty); TX = FMA(Tg, Tx, Ti * Ty); } } } { E TF, T2b, T4D, T4M, T2K, T3r, T4a, T4m, T1N, T28, T29, T3C, T3F, T43, T3X; E T3Y, T4o, T2f, T2g, T2h, T2y, T2D, T2E, T3g, T3h, T4z, T3n, T3o, T3p, T33; E T38, T4K, TW, T1r, T1s, T3J, T3M, T44, T3U, T3V, T4n, T2c, T2d, T2e, T2n; E T2s, T2t, T3d, T3e, T4y, T3k, T3l, T3m, T2S, T2X, T4J; { E T1, T47, Te, T46, Tw, T2H, TD, T2I, T9, Td; T1 = cr[0]; T47 = ci[0]; T9 = cr[WS(rs, 10)]; Td = ci[WS(rs, 10)]; Te = FMA(T8, T9, Tc * Td); T46 = FNMS(Tc, T9, T8 * Td); { E Tr, Tv, TA, TC; Tr = cr[WS(rs, 5)]; Tv = ci[WS(rs, 5)]; Tw = FMA(Tq, Tr, Tu * Tv); T2H = FNMS(Tu, Tr, Tq * Tv); TA = cr[WS(rs, 15)]; TC = ci[WS(rs, 15)]; TD = FMA(Tz, TA, TB * TC); T2I = FNMS(TB, TA, Tz * TC); } { E Tf, TE, T4B, T4C; Tf = T1 + Te; TE = Tw + TD; TF = Tf - TE; T2b = Tf + TE; T4B = T47 - T46; T4C = Tw - TD; T4D = T4B - T4C; T4M = T4C + T4B; } { E T2G, T2J, T48, T49; T2G = T1 - Te; T2J = T2H - T2I; T2K = T2G - T2J; T3r = T2G + T2J; T48 = T46 + T47; T49 = T2H + T2I; T4a = T48 - T49; T4m = T49 + T48; } } { E T1D, T3A, T2u, T31, T27, T3D, T2C, T37, T1M, T3B, T2x, T32, T1W, T3E, T2z; E T36; { E T1x, T2Z, T1C, T30; { E T1u, T1w, T1z, T1B; T1u = cr[WS(rs, 8)]; T1w = ci[WS(rs, 8)]; T1x = FMA(T1t, T1u, T1v * T1w); T2Z = FNMS(T1v, T1u, T1t * T1w); T1z = cr[WS(rs, 18)]; T1B = ci[WS(rs, 18)]; T1C = FMA(T1y, T1z, T1A * T1B); T30 = FNMS(T1A, T1z, T1y * T1B); } T1D = T1x + T1C; T3A = T2Z + T30; T2u = T1x - T1C; T31 = T2Z - T30; } { E T21, T2A, T26, T2B; { E T1Y, T20, T23, T25; T1Y = cr[WS(rs, 17)]; T20 = ci[WS(rs, 17)]; T21 = FMA(T1X, T1Y, T1Z * T20); T2A = FNMS(T1Z, T1Y, T1X * T20); T23 = cr[WS(rs, 7)]; T25 = ci[WS(rs, 7)]; T26 = FMA(T22, T23, T24 * T25); T2B = FNMS(T24, T23, T22 * T25); } T27 = T21 + T26; T3D = T2A + T2B; T2C = T2A - T2B; T37 = T21 - T26; } { E T1I, T2v, T1L, T2w; { E T1F, T1H, T1J, T1K; T1F = cr[WS(rs, 13)]; T1H = ci[WS(rs, 13)]; T1I = FMA(T1E, T1F, T1G * T1H); T2v = FNMS(T1G, T1F, T1E * T1H); T1J = cr[WS(rs, 3)]; T1K = ci[WS(rs, 3)]; T1L = FMA(Tg, T1J, Ti * T1K); T2w = FNMS(Ti, T1J, Tg * T1K); } T1M = T1I + T1L; T3B = T2v + T2w; T2x = T2v - T2w; T32 = T1I - T1L; } { E T1S, T34, T1V, T35; { E T1P, T1R, T1T, T1U; T1P = cr[WS(rs, 12)]; T1R = ci[WS(rs, 12)]; T1S = FMA(T1O, T1P, T1Q * T1R); T34 = FNMS(T1Q, T1P, T1O * T1R); T1T = cr[WS(rs, 2)]; T1U = ci[WS(rs, 2)]; T1V = FMA(T1f, T1T, T1h * T1U); T35 = FNMS(T1h, T1T, T1f * T1U); } T1W = T1S + T1V; T3E = T34 + T35; T2z = T1S - T1V; T36 = T34 - T35; } T1N = T1D - T1M; T28 = T1W - T27; T29 = T1N + T28; T3C = T3A - T3B; T3F = T3D - T3E; T43 = T3F - T3C; T3X = T3A + T3B; T3Y = T3E + T3D; T4o = T3X + T3Y; T2f = T1D + T1M; T2g = T1W + T27; T2h = T2f + T2g; T2y = T2u - T2x; T2D = T2z - T2C; T2E = T2y + T2D; T3g = T31 - T32; T3h = T36 - T37; T4z = T3g + T3h; T3n = T2u + T2x; T3o = T2z + T2C; T3p = T3n + T3o; T33 = T31 + T32; T38 = T36 + T37; T4K = T33 + T38; } { E TO, T3H, T2j, T2Q, T1q, T3L, T2r, T2T, TV, T3I, T2m, T2R, T1b, T3K, T2o; E T2W; { E TI, T2O, TN, T2P; { E TG, TH, TK, TM; TG = cr[WS(rs, 4)]; TH = ci[WS(rs, 4)]; TI = FMA(Tk, TG, To * TH); T2O = FNMS(To, TG, Tk * TH); TK = cr[WS(rs, 14)]; TM = ci[WS(rs, 14)]; TN = FMA(TJ, TK, TL * TM); T2P = FNMS(TL, TK, TJ * TM); } TO = TI + TN; T3H = T2O + T2P; T2j = TI - TN; T2Q = T2O - T2P; } { E T1e, T2p, T1p, T2q; { E T1c, T1d, T1k, T1o; T1c = cr[WS(rs, 1)]; T1d = ci[WS(rs, 1)]; T1e = FMA(T2, T1c, T5 * T1d); T2p = FNMS(T5, T1c, T2 * T1d); T1k = cr[WS(rs, 11)]; T1o = ci[WS(rs, 11)]; T1p = FMA(T1j, T1k, T1n * T1o); T2q = FNMS(T1n, T1k, T1j * T1o); } T1q = T1e + T1p; T3L = T2p + T2q; T2r = T2p - T2q; T2T = T1p - T1e; } { E TR, T2k, TU, T2l; { E TP, TQ, TS, TT; TP = cr[WS(rs, 9)]; TQ = ci[WS(rs, 9)]; TR = FMA(T3, TP, T6 * TQ); T2k = FNMS(T6, TP, T3 * TQ); TS = cr[WS(rs, 19)]; TT = ci[WS(rs, 19)]; TU = FMA(Tx, TS, Ty * TT); T2l = FNMS(Ty, TS, Tx * TT); } TV = TR + TU; T3I = T2k + T2l; T2m = T2k - T2l; T2R = TR - TU; } { E T11, T2U, T1a, T2V; { E TY, T10, T15, T19; TY = cr[WS(rs, 16)]; T10 = ci[WS(rs, 16)]; T11 = FMA(TX, TY, TZ * T10); T2U = FNMS(TZ, TY, TX * T10); T15 = cr[WS(rs, 6)]; T19 = ci[WS(rs, 6)]; T1a = FMA(T14, T15, T18 * T19); T2V = FNMS(T18, T15, T14 * T19); } T1b = T11 + T1a; T3K = T2U + T2V; T2o = T11 - T1a; T2W = T2U - T2V; } TW = TO - TV; T1r = T1b - T1q; T1s = TW + T1r; T3J = T3H - T3I; T3M = T3K - T3L; T44 = T3J + T3M; T3U = T3H + T3I; T3V = T3K + T3L; T4n = T3U + T3V; T2c = TO + TV; T2d = T1b + T1q; T2e = T2c + T2d; T2n = T2j - T2m; T2s = T2o - T2r; T2t = T2n + T2s; T3d = T2Q - T2R; T3e = T2W + T2T; T4y = T3d + T3e; T3k = T2j + T2m; T3l = T2o + T2r; T3m = T3k + T3l; T2S = T2Q + T2R; T2X = T2T - T2W; T4J = T2X - T2S; } { E T3y, T2a, T3x, T3O, T3Q, T3G, T3N, T3P, T3z; T3y = KP559016994 * (T1s - T29); T2a = T1s + T29; T3x = FNMS(KP250000000, T2a, TF); T3G = T3C + T3F; T3N = T3J - T3M; T3O = FNMS(KP587785252, T3N, KP951056516 * T3G); T3Q = FMA(KP951056516, T3N, KP587785252 * T3G); ci[WS(rs, 9)] = TF + T2a; T3P = T3y + T3x; ci[WS(rs, 5)] = T3P - T3Q; cr[WS(rs, 6)] = T3P + T3Q; T3z = T3x - T3y; cr[WS(rs, 2)] = T3z - T3O; ci[WS(rs, 1)] = T3z + T3O; } { E T3q, T3s, T3t, T3j, T3w, T3f, T3i, T3v, T3u; T3q = KP559016994 * (T3m - T3p); T3s = T3m + T3p; T3t = FNMS(KP250000000, T3s, T3r); T3f = T3d - T3e; T3i = T3g - T3h; T3j = FMA(KP951056516, T3f, KP587785252 * T3i); T3w = FNMS(KP587785252, T3f, KP951056516 * T3i); cr[WS(rs, 5)] = T3r + T3s; T3v = T3t - T3q; ci[WS(rs, 2)] = T3v - T3w; ci[WS(rs, 6)] = T3w + T3v; T3u = T3q + T3t; cr[WS(rs, 1)] = T3j + T3u; cr[WS(rs, 9)] = T3u - T3j; } { E T3R, T2i, T3S, T40, T42, T3W, T3Z, T41, T3T; T3R = KP559016994 * (T2e - T2h); T2i = T2e + T2h; T3S = FNMS(KP250000000, T2i, T2b); T3W = T3U - T3V; T3Z = T3X - T3Y; T40 = FMA(KP951056516, T3W, KP587785252 * T3Z); T42 = FNMS(KP587785252, T3W, KP951056516 * T3Z); cr[0] = T2b + T2i; T41 = T3S - T3R; ci[WS(rs, 7)] = T41 - T42; cr[WS(rs, 8)] = T41 + T42; T3T = T3R + T3S; cr[WS(rs, 4)] = T3T - T40; ci[WS(rs, 3)] = T3T + T40; } { E T2F, T2L, T2M, T3a, T3b, T2Y, T39, T3c, T2N; T2F = KP559016994 * (T2t - T2E); T2L = T2t + T2E; T2M = FNMS(KP250000000, T2L, T2K); T2Y = T2S + T2X; T39 = T33 - T38; T3a = FMA(KP951056516, T2Y, KP587785252 * T39); T3b = FNMS(KP587785252, T2Y, KP951056516 * T39); ci[WS(rs, 4)] = T2K + T2L; T3c = T2M - T2F; cr[WS(rs, 3)] = T3b + T3c; cr[WS(rs, 7)] = T3c - T3b; T2N = T2F + T2M; ci[0] = T2N - T3a; ci[WS(rs, 8)] = T3a + T2N; } { E T4e, T45, T4f, T4d, T4h, T4b, T4c, T4i, T4g; T4e = KP559016994 * (T44 + T43); T45 = T43 - T44; T4f = FMA(KP250000000, T45, T4a); T4b = T1r - TW; T4c = T1N - T28; T4d = FNMS(KP587785252, T4c, KP951056516 * T4b); T4h = FMA(KP587785252, T4b, KP951056516 * T4c); cr[WS(rs, 10)] = T45 - T4a; T4i = T4f - T4e; cr[WS(rs, 18)] = T4h - T4i; ci[WS(rs, 17)] = T4h + T4i; T4g = T4e + T4f; cr[WS(rs, 14)] = T4d - T4g; ci[WS(rs, 13)] = T4d + T4g; } { E T4A, T4E, T4F, T4x, T4H, T4v, T4w, T4I, T4G; T4A = KP559016994 * (T4y - T4z); T4E = T4y + T4z; T4F = FNMS(KP250000000, T4E, T4D); T4v = T3n - T3o; T4w = T3k - T3l; T4x = FNMS(KP587785252, T4w, KP951056516 * T4v); T4H = FMA(KP951056516, T4w, KP587785252 * T4v); ci[WS(rs, 14)] = T4E + T4D; T4I = T4A + T4F; ci[WS(rs, 10)] = T4H + T4I; ci[WS(rs, 18)] = T4I - T4H; T4G = T4A - T4F; cr[WS(rs, 13)] = T4x + T4G; cr[WS(rs, 17)] = T4G - T4x; } { E T4r, T4p, T4q, T4l, T4t, T4j, T4k, T4u, T4s; T4r = KP559016994 * (T4n - T4o); T4p = T4n + T4o; T4q = FNMS(KP250000000, T4p, T4m); T4j = T2c - T2d; T4k = T2f - T2g; T4l = FNMS(KP951056516, T4k, KP587785252 * T4j); T4t = FMA(KP951056516, T4j, KP587785252 * T4k); ci[WS(rs, 19)] = T4p + T4m; T4u = T4r + T4q; cr[WS(rs, 16)] = T4t - T4u; ci[WS(rs, 15)] = T4t + T4u; T4s = T4q - T4r; cr[WS(rs, 12)] = T4l - T4s; ci[WS(rs, 11)] = T4l + T4s; } { E T4Q, T4L, T4R, T4P, T4T, T4N, T4O, T4U, T4S; T4Q = KP559016994 * (T4J + T4K); T4L = T4J - T4K; T4R = FMA(KP250000000, T4L, T4M); T4N = T2n - T2s; T4O = T2y - T2D; T4P = FMA(KP951056516, T4N, KP587785252 * T4O); T4T = FNMS(KP587785252, T4N, KP951056516 * T4O); cr[WS(rs, 15)] = T4L - T4M; T4U = T4Q + T4R; ci[WS(rs, 12)] = T4T + T4U; ci[WS(rs, 16)] = T4U - T4T; T4S = T4Q - T4R; cr[WS(rs, 11)] = T4P + T4S; cr[WS(rs, 19)] = T4S - T4P; } } } } } static const tw_instr twinstr[] = { {TW_CEXP, 1, 1}, {TW_CEXP, 1, 3}, {TW_CEXP, 1, 9}, {TW_CEXP, 1, 19}, {TW_NEXT, 1, 0} }; static const hc2hc_desc desc = { 20, "hf2_20", twinstr, &GENUS, {204, 92, 72, 0} }; void X(codelet_hf2_20) (planner *p) { X(khc2hc_register) (p, hf2_20, &desc); } #endif /* HAVE_FMA */ fftw-3.3.3/rdft/scalar/r2cf/hf2_25.c0000644000175000017500000015327212054410657013616 00000000000000/* * Copyright (c) 2003, 2007-11 Matteo Frigo * Copyright (c) 2003, 2007-11 Massachusetts Institute of Technology * * This program is free software; you can redistribute it and/or modify * it under the terms of the GNU General Public License as published by * the Free Software Foundation; either version 2 of the License, or * (at your option) any later version. * * This program is distributed in the hope that it will be useful, * but WITHOUT ANY WARRANTY; without even the implied warranty of * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the * GNU General Public License for more details. * * You should have received a copy of the GNU General Public License * along with this program; if not, write to the Free Software * Foundation, Inc., 51 Franklin Street, Fifth Floor, Boston, MA 02110-1301 USA * */ /* This file was automatically generated --- DO NOT EDIT */ /* Generated on Sun Nov 25 07:40:09 EST 2012 */ #include "codelet-rdft.h" #ifdef HAVE_FMA /* Generated by: ../../../genfft/gen_hc2hc.native -fma -reorder-insns -schedule-for-pipeline -compact -variables 4 -pipeline-latency 4 -twiddle-log3 -precompute-twiddles -n 25 -dit -name hf2_25 -include hf.h */ /* * This function contains 440 FP additions, 434 FP multiplications, * (or, 84 additions, 78 multiplications, 356 fused multiply/add), * 215 stack variables, 47 constants, and 100 memory accesses */ #include "hf.h" static void hf2_25(R *cr, R *ci, const R *W, stride rs, INT mb, INT me, INT ms) { DK(KP949179823, +0.949179823508441261575555465843363271711583843); DK(KP860541664, +0.860541664367944677098261680920518816412804187); DK(KP621716863, +0.621716863012209892444754556304102309693593202); DK(KP614372930, +0.614372930789563808870829930444362096004872855); DK(KP557913902, +0.557913902031834264187699648465567037992437152); DK(KP249506682, +0.249506682107067890488084201715862638334226305); DK(KP560319534, +0.560319534973832390111614715371676131169633784); DK(KP681693190, +0.681693190061530575150324149145440022633095390); DK(KP906616052, +0.906616052148196230441134447086066874408359177); DK(KP968479752, +0.968479752739016373193524836781420152702090879); DK(KP845997307, +0.845997307939530944175097360758058292389769300); DK(KP998026728, +0.998026728428271561952336806863450553336905220); DK(KP994076283, +0.994076283785401014123185814696322018529298887); DK(KP734762448, +0.734762448793050413546343770063151342619912334); DK(KP772036680, +0.772036680810363904029489473607579825330539880); DK(KP062914667, +0.062914667253649757225485955897349402364686947); DK(KP921177326, +0.921177326965143320250447435415066029359282231); DK(KP833417178, +0.833417178328688677408962550243238843138996060); DK(KP541454447, +0.541454447536312777046285590082819509052033189); DK(KP803003575, +0.803003575438660414833440593570376004635464850); DK(KP943557151, +0.943557151597354104399655195398983005179443399); DK(KP242145790, +0.242145790282157779872542093866183953459003101); DK(KP554608978, +0.554608978404018097464974850792216217022558774); DK(KP559154169, +0.559154169276087864842202529084232643714075927); DK(KP683113946, +0.683113946453479238701949862233725244439656928); DK(KP968583161, +0.968583161128631119490168375464735813836012403); DK(KP248028675, +0.248028675328619457762448260696444630363259177); DK(KP904730450, +0.904730450839922351881287709692877908104763647); DK(KP831864738, +0.831864738706457140726048799369896829771167132); DK(KP525970792, +0.525970792408939708442463226536226366643874659); DK(KP726211448, +0.726211448929902658173535992263577167607493062); DK(KP549754652, +0.549754652192770074288023275540779861653779767); DK(KP871714437, +0.871714437527667770979999223229522602943903653); DK(KP992114701, +0.992114701314477831049793042785778521453036709); DK(KP939062505, +0.939062505817492352556001843133229685779824606); DK(KP851038619, +0.851038619207379630836264138867114231259902550); DK(KP256756360, +0.256756360367726783319498520922669048172391148); DK(KP912575812, +0.912575812670962425556968549836277086778922727); DK(KP634619297, +0.634619297544148100711287640319130485732531031); DK(KP912018591, +0.912018591466481957908415381764119056233607330); DK(KP470564281, +0.470564281212251493087595091036643380879947982); DK(KP827271945, +0.827271945972475634034355757144307982555673741); DK(KP126329378, +0.126329378446108174786050455341811215027378105); DK(KP951056516, +0.951056516295153572116439333379382143405698634); DK(KP559016994, +0.559016994374947424102293417182819058860154590); DK(KP250000000, +0.250000000000000000000000000000000000000000000); DK(KP618033988, +0.618033988749894848204586834365638117720309180); { INT m; for (m = mb, W = W + ((mb - 1) * 8); m < me; m = m + 1, cr = cr + ms, ci = ci - ms, W = W + 8, MAKE_VOLATILE_STRIDE(50, rs)) { E T7M, T6S, T6Q, T7S, T7Q, T7L, T6R, T6J, T7N, T7R; { E T2, T8, T3, T6, Tk, Tv, TS, T4, Ta, TD, T2L, T10, Tm, T5, Tc; T2 = W[0]; T8 = W[4]; T3 = W[2]; T6 = W[3]; Tk = W[6]; Tv = T2 * T8; TS = T3 * T8; T4 = T2 * T3; Ta = T2 * T6; TD = T8 * Tk; T2L = T2 * Tk; T10 = T3 * Tk; Tm = W[7]; T5 = W[1]; Tc = W[5]; { E T7u, T7U, T4s, T6a, T4g, TN, T4f, T7q, T8j, T7p, T4G, T6k, T3a, T4z, T6n; E T6m, T4w, T4a, T4D, T6j, T6C, T54, T6z, T5b, T1v, T3t, T6y, T58, T6B, T51; E T6v, T5j, T6s, T5q, T21, T3H, T6r, T5n, T6u, T5g, T26, T3K, T4N, T2A, T3U; E T4U, T2c, T3M, T2k, T3O; { E T11, T1b, Tb, T19, T7, T2m, TT, T15, T2Q, TX, T2p, T1g, T2a, T2e, T2i; E T27, T1c, T1O, T1K, T1q, T1m, T2x, T2t, T1W, T1S, T2G, T3Y, T2N, T4F, T38; E T48, T4y, T2K, T40, T2S, T41; { E T2M, T1j, T1l, T2X, T2U, T35, T31, T7l, T7n, T7m, T2O, T2R; { E T1, Tj, T4j, TK, T4q, TC, T4o, Tt, T4l; { E TE, Tw, TI, TA, Th, Tr, Tn, Td, Te, Ti, T14, T2P, TH, Tx, TB; T1 = cr[0]; T11 = FMA(T6, Tm, T10); T14 = T3 * Tm; T2P = T2 * Tm; TH = T8 * Tm; T2M = FMA(T5, Tm, T2L); T1b = FNMS(T5, T3, Ta); Tb = FMA(T5, T3, Ta); T19 = FMA(T5, T6, T4); T7 = FNMS(T5, T6, T4); T2m = FNMS(T6, Tc, TS); TT = FMA(T6, Tc, TS); TE = FMA(Tc, Tm, TD); T1j = FMA(T5, Tc, Tv); Tw = FNMS(T5, Tc, Tv); { E TW, Tz, T1f, T2d; TW = T3 * Tc; Tz = T2 * Tc; T15 = FNMS(T6, Tk, T14); T2Q = FNMS(T5, Tk, T2P); TI = FNMS(Tc, Tk, TH); T1f = T19 * Tc; T2d = T19 * Tk; { E T2h, T1a, Tg, Tq; T2h = T19 * Tm; T1a = T19 * T8; Tg = T7 * Tc; Tq = T7 * Tm; { E Tl, T9, T1p, T1k; Tl = T7 * Tk; T9 = T7 * T8; T1p = T1j * Tm; T1k = T1j * Tk; { E T34, T30, T1N, T1J; T34 = TT * Tm; T30 = TT * Tk; T1N = Tw * Tm; T1J = Tw * Tk; TX = FNMS(T6, T8, TW); T2p = FMA(T6, T8, TW); TA = FMA(T5, T8, Tz); T1l = FNMS(T5, T8, Tz); T1g = FMA(T1b, T8, T1f); T2a = FNMS(T1b, T8, T1f); T2e = FMA(T1b, Tm, T2d); T2i = FNMS(T1b, Tk, T2h); T27 = FMA(T1b, Tc, T1a); T1c = FNMS(T1b, Tc, T1a); T2X = FMA(Tb, T8, Tg); Th = FNMS(Tb, T8, Tg); Tr = FNMS(Tb, Tk, Tq); Tn = FMA(Tb, Tm, Tl); Td = FMA(Tb, Tc, T9); T2U = FNMS(Tb, Tc, T9); T35 = FNMS(TX, Tk, T34); T31 = FMA(TX, Tm, T30); T1O = FNMS(TA, Tk, T1N); T1K = FMA(TA, Tm, T1J); T1q = FNMS(T1l, Tk, T1p); T1m = FMA(T1l, Tm, T1k); { E T2w, T2s, T1V, T1R; T2w = T27 * Tm; T2s = T27 * Tk; T1V = Td * Tm; T1R = Td * Tk; T2x = FNMS(T2a, Tk, T2w); T2t = FMA(T2a, Tm, T2s); T1W = FNMS(Th, Tk, T1V); T1S = FMA(Th, Tm, T1R); T7l = ci[0]; Te = cr[WS(rs, 5)]; Ti = ci[WS(rs, 5)]; } } } } } { E TF, TJ, Tf, T4i, TG, T4p; TF = cr[WS(rs, 15)]; TJ = ci[WS(rs, 15)]; Tf = Td * Te; T4i = Td * Ti; TG = TE * TF; T4p = TE * TJ; Tj = FMA(Th, Ti, Tf); T4j = FNMS(Th, Te, T4i); TK = FMA(TI, TJ, TG); T4q = FNMS(TI, TF, T4p); } Tx = cr[WS(rs, 10)]; TB = ci[WS(rs, 10)]; { E To, Ts, Ty, T4n, Tp, T4k; To = cr[WS(rs, 20)]; Ts = ci[WS(rs, 20)]; Ty = Tw * Tx; T4n = Tw * TB; Tp = Tn * To; T4k = Tn * Ts; TC = FMA(TA, TB, Ty); T4o = FNMS(TA, Tx, T4n); Tt = FMA(Tr, Ts, Tp); T4l = FNMS(Tr, To, T4k); } } { E TL, T7s, T4r, Tu, T7t, T4m, TM; TL = TC + TK; T7s = TC - TK; T4r = T4o - T4q; T7n = T4o + T4q; Tu = Tj + Tt; T7t = Tj - Tt; T4m = T4j - T4l; T7m = T4j + T4l; T7u = FNMS(KP618033988, T7t, T7s); T7U = FMA(KP618033988, T7s, T7t); T4s = FMA(KP618033988, T4r, T4m); T6a = FNMS(KP618033988, T4m, T4r); T4g = Tu - TL; TM = Tu + TL; TN = T1 + TM; T4f = FNMS(KP250000000, TM, T1); } } { E T2D, T2F, T7o, T2E, T3X; T2D = cr[WS(rs, 3)]; T2F = ci[WS(rs, 3)]; T7q = T7m - T7n; T7o = T7m + T7n; T2E = T3 * T2D; T3X = T3 * T2F; { E T2V, T2W, T2Y, T32, T36; T2V = cr[WS(rs, 13)]; T8j = T7o + T7l; T7p = FNMS(KP250000000, T7o, T7l); T2G = FMA(T6, T2F, T2E); T3Y = FNMS(T6, T2D, T3X); T2W = T2U * T2V; T2Y = ci[WS(rs, 13)]; T32 = cr[WS(rs, 18)]; T36 = ci[WS(rs, 18)]; { E T2H, T2I, T2J, T3Z; { E T2Z, T45, T37, T47, T44, T33, T46; T2H = cr[WS(rs, 8)]; T2Z = FMA(T2X, T2Y, T2W); T44 = T2U * T2Y; T33 = T31 * T32; T46 = T31 * T36; T2I = T1j * T2H; T45 = FNMS(T2X, T2V, T44); T37 = FMA(T35, T36, T33); T47 = FNMS(T35, T32, T46); T2J = ci[WS(rs, 8)]; T2N = cr[WS(rs, 23)]; T4F = T2Z - T37; T38 = T2Z + T37; T48 = T45 + T47; T4y = T47 - T45; T3Z = T1j * T2J; T2O = T2M * T2N; T2R = ci[WS(rs, 23)]; } T2K = FMA(T1l, T2J, T2I); T40 = FNMS(T1l, T2H, T3Z); } } } T2S = FMA(T2Q, T2R, T2O); T41 = T2M * T2R; } { E TR, T3h, T1t, T53, T3r, T5a, TZ, T3j, T17, T3l; { E T12, T16, T13, T3k; { E TO, TP, T4C, T4B, TQ; { E T2T, T4E, T42, T4v, T39; TO = cr[WS(rs, 1)]; T2T = T2K + T2S; T4E = T2K - T2S; T42 = FNMS(T2Q, T2N, T41); TP = T2 * TO; T4G = FMA(KP618033988, T4F, T4E); T6k = FNMS(KP618033988, T4E, T4F); T4v = T38 - T2T; T39 = T2T + T38; { E T43, T4x, T4u, T49; T43 = T40 + T42; T4x = T42 - T40; T4u = FNMS(KP250000000, T39, T2G); T3a = T2G + T39; T4z = FMA(KP618033988, T4y, T4x); T6n = FNMS(KP618033988, T4x, T4y); T4C = T48 - T43; T49 = T43 + T48; T6m = FMA(KP559016994, T4v, T4u); T4w = FNMS(KP559016994, T4v, T4u); T4B = FNMS(KP250000000, T49, T3Y); T4a = T3Y + T49; TQ = ci[WS(rs, 1)]; } } { E T1n, T1r, T1i, T1o, T3o, T3p; { E T1d, T1h, T1e, T3n, T3g; T1d = cr[WS(rs, 11)]; T1h = ci[WS(rs, 11)]; T4D = FNMS(KP559016994, T4C, T4B); T6j = FMA(KP559016994, T4C, T4B); TR = FMA(T5, TQ, TP); T3g = T2 * TQ; T1e = T1c * T1d; T3n = T1c * T1h; T1n = cr[WS(rs, 16)]; T3h = FNMS(T5, TO, T3g); T1r = ci[WS(rs, 16)]; T1i = FMA(T1g, T1h, T1e); T1o = T1m * T1n; T3o = FNMS(T1g, T1d, T3n); T3p = T1m * T1r; } { E TU, TY, TV, T3i, T3q, T1s; TU = cr[WS(rs, 6)]; T1s = FMA(T1q, T1r, T1o); TY = ci[WS(rs, 6)]; T3q = FNMS(T1q, T1n, T3p); TV = TT * TU; T1t = T1i + T1s; T53 = T1s - T1i; T3i = TT * TY; T3r = T3o + T3q; T5a = T3q - T3o; T12 = cr[WS(rs, 21)]; T16 = ci[WS(rs, 21)]; TZ = FMA(TX, TY, TV); T3j = FNMS(TX, TU, T3i); T13 = T11 * T12; T3k = T11 * T16; } } } T17 = FMA(T15, T16, T13); T3l = FNMS(T15, T12, T3k); } { E T1z, T3v, T5i, T1Z, T3F, T5p, T1D, T3x, T1H, T3z; { E T1E, T1G, T1F, T3y; { E T1w, T1y, T1x, T57, T50, T56, T4Z, T3u, T18, T52; T1w = cr[WS(rs, 4)]; T1y = ci[WS(rs, 4)]; T18 = TZ + T17; T52 = T17 - TZ; { E T3m, T59, T1u, T3s; T3m = T3j + T3l; T59 = T3j - T3l; T1x = T7 * T1w; T6C = FNMS(KP618033988, T52, T53); T54 = FMA(KP618033988, T53, T52); T1u = T18 + T1t; T57 = T18 - T1t; T6z = FMA(KP618033988, T59, T5a); T5b = FNMS(KP618033988, T5a, T59); T3s = T3m + T3r; T50 = T3m - T3r; T1v = TR + T1u; T56 = FNMS(KP250000000, T1u, TR); T3t = T3h + T3s; T4Z = FNMS(KP250000000, T3s, T3h); T3u = T7 * T1y; } T6y = FNMS(KP559016994, T57, T56); T58 = FMA(KP559016994, T57, T56); T6B = FNMS(KP559016994, T50, T4Z); T51 = FMA(KP559016994, T50, T4Z); T1z = FMA(Tb, T1y, T1x); T3v = FNMS(Tb, T1w, T3u); } { E T1Q, T3C, T1Y, T3E; { E T1L, T1P, T1T, T1X, T1M, T3B, T1U, T3D; T1L = cr[WS(rs, 14)]; T1P = ci[WS(rs, 14)]; T1T = cr[WS(rs, 19)]; T1X = ci[WS(rs, 19)]; T1M = T1K * T1L; T3B = T1K * T1P; T1U = T1S * T1T; T3D = T1S * T1X; T1Q = FMA(T1O, T1P, T1M); T3C = FNMS(T1O, T1L, T3B); T1Y = FMA(T1W, T1X, T1U); T3E = FNMS(T1W, T1T, T3D); } { E T1A, T1C, T1B, T3w; T1A = cr[WS(rs, 9)]; T1C = ci[WS(rs, 9)]; T5i = T1Y - T1Q; T1Z = T1Q + T1Y; T3F = T3C + T3E; T5p = T3E - T3C; T1B = T8 * T1A; T3w = T8 * T1C; T1E = cr[WS(rs, 24)]; T1G = ci[WS(rs, 24)]; T1D = FMA(Tc, T1C, T1B); T3x = FNMS(Tc, T1A, T3w); T1F = Tk * T1E; T3y = Tk * T1G; } } T1H = FMA(Tm, T1G, T1F); T3z = FNMS(Tm, T1E, T3y); } { E T2f, T2j, T2g, T3N; { E T23, T25, T24, T5m, T5f, T5l, T5e, T3J, T1I, T5h; T23 = cr[WS(rs, 2)]; T25 = ci[WS(rs, 2)]; T1I = T1D + T1H; T5h = T1H - T1D; { E T3A, T5o, T20, T3G; T3A = T3x + T3z; T5o = T3z - T3x; T24 = T19 * T23; T6v = FNMS(KP618033988, T5h, T5i); T5j = FMA(KP618033988, T5i, T5h); T20 = T1I + T1Z; T5m = T1I - T1Z; T6s = FNMS(KP618033988, T5o, T5p); T5q = FMA(KP618033988, T5p, T5o); T3G = T3A + T3F; T5f = T3F - T3A; T21 = T1z + T20; T5l = FNMS(KP250000000, T20, T1z); T3H = T3v + T3G; T5e = FNMS(KP250000000, T3G, T3v); T3J = T19 * T25; } T6r = FNMS(KP559016994, T5m, T5l); T5n = FMA(KP559016994, T5m, T5l); T6u = FMA(KP559016994, T5f, T5e); T5g = FNMS(KP559016994, T5f, T5e); T26 = FMA(T1b, T25, T24); T3K = FNMS(T1b, T23, T3J); } { E T2r, T3R, T2z, T3T; { E T2n, T2q, T2u, T2y, T2o, T3Q, T2v, T3S; T2n = cr[WS(rs, 12)]; T2q = ci[WS(rs, 12)]; T2u = cr[WS(rs, 17)]; T2y = ci[WS(rs, 17)]; T2o = T2m * T2n; T3Q = T2m * T2q; T2v = T2t * T2u; T3S = T2t * T2y; T2r = FMA(T2p, T2q, T2o); T3R = FNMS(T2p, T2n, T3Q); T2z = FMA(T2x, T2y, T2v); T3T = FNMS(T2x, T2u, T3S); } { E T28, T2b, T29, T3L; T28 = cr[WS(rs, 7)]; T2b = ci[WS(rs, 7)]; T4N = T2z - T2r; T2A = T2r + T2z; T3U = T3R + T3T; T4U = T3R - T3T; T29 = T27 * T28; T3L = T27 * T2b; T2f = cr[WS(rs, 22)]; T2j = ci[WS(rs, 22)]; T2c = FMA(T2a, T2b, T29); T3M = FNMS(T2a, T28, T3L); T2g = T2e * T2f; T3N = T2e * T2j; } } T2k = FMA(T2i, T2j, T2g); T3O = FNMS(T2i, T2f, T3N); } } } } { E T8k, T6d, T6g, T8r, T6f, T8l, T6c, T8q, T69, T7r, T5Y, T8g, T8i, T66, T68; E T5X, T8d, T8h; { E T4O, T4V, T22, T4S, T4L, T3b, T4e, T4c, T3I; T8k = T3t + T3H; T3I = T3t - T3H; { E T2l, T4M, T3P, T4T; T2l = T2c + T2k; T4M = T2k - T2c; T3P = T3M + T3O; T4T = T3O - T3M; T4O = FMA(KP618033988, T4N, T4M); T6d = FNMS(KP618033988, T4M, T4N); { E T4R, T2B, T4K, T3V; T4R = T2A - T2l; T2B = T2l + T2A; T4V = FNMS(KP618033988, T4U, T4T); T6g = FMA(KP618033988, T4T, T4U); T4K = T3U - T3P; T3V = T3P + T3U; { E T4Q, T2C, T4J, T3W, T4b; T4Q = FNMS(KP250000000, T2B, T26); T2C = T26 + T2B; T4J = FNMS(KP250000000, T3V, T3K); T3W = T3K + T3V; T8r = T21 - T1v; T22 = T1v + T21; T4S = FNMS(KP559016994, T4R, T4Q); T6f = FMA(KP559016994, T4R, T4Q); T4b = T3W - T4a; T8l = T3W + T4a; T6c = FMA(KP559016994, T4K, T4J); T4L = FNMS(KP559016994, T4K, T4J); T8q = T2C - T3a; T3b = T2C + T3a; T4e = FNMS(KP618033988, T3I, T4b); T4c = FMA(KP618033988, T4b, T3I); } } } { E T5H, T4t, T7V, T87, T5Q, T5P, T5D, T8e, T5A, T8f, T5K, T60, T8c, T8a, T5u; E T5w, T5U, T64, T5N, T61; { E T3e, T3d, T4h, T3c, T7T; T4h = FMA(KP559016994, T4g, T4f); T69 = FNMS(KP559016994, T4g, T4f); T3c = T22 + T3b; T3e = T22 - T3b; T7r = FNMS(KP559016994, T7q, T7p); T7T = FMA(KP559016994, T7q, T7p); T5H = FMA(KP951056516, T4s, T4h); T4t = FNMS(KP951056516, T4s, T4h); cr[0] = TN + T3c; T3d = FNMS(KP250000000, T3c, TN); T7V = FNMS(KP951056516, T7U, T7T); T87 = FMA(KP951056516, T7U, T7T); { E T5S, T5T, T5L, T4I, T5B, T5M, T55, T5J, T5s, T5z, T4X, T5C, T5I, T5c; { E T5k, T5r, T4P, T4W; { E T4A, T4d, T3f, T4H; T4A = FMA(KP951056516, T4z, T4w); T5S = FNMS(KP951056516, T4z, T4w); T4d = FNMS(KP559016994, T3e, T3d); T3f = FMA(KP559016994, T3e, T3d); T5T = FNMS(KP951056516, T4G, T4D); T4H = FMA(KP951056516, T4G, T4D); T5k = FNMS(KP951056516, T5j, T5g); T5L = FMA(KP951056516, T5j, T5g); cr[WS(rs, 5)] = FMA(KP951056516, T4c, T3f); ci[WS(rs, 4)] = FNMS(KP951056516, T4c, T3f); ci[WS(rs, 9)] = FMA(KP951056516, T4e, T4d); cr[WS(rs, 10)] = FNMS(KP951056516, T4e, T4d); T4I = FNMS(KP126329378, T4H, T4A); T5B = FMA(KP126329378, T4A, T4H); T5M = FNMS(KP951056516, T5q, T5n); T5r = FMA(KP951056516, T5q, T5n); } T4P = FNMS(KP951056516, T4O, T4L); T5Q = FMA(KP951056516, T4O, T4L); T5P = FNMS(KP951056516, T4V, T4S); T4W = FMA(KP951056516, T4V, T4S); T55 = FNMS(KP951056516, T54, T51); T5J = FMA(KP951056516, T54, T51); T5s = FMA(KP827271945, T5r, T5k); T5z = FNMS(KP827271945, T5k, T5r); T4X = FNMS(KP470564281, T4W, T4P); T5C = FMA(KP470564281, T4P, T4W); T5I = FMA(KP951056516, T5b, T58); T5c = FNMS(KP951056516, T5b, T58); } { E T88, T4Y, T5d, T5y, T89, T5t; T5D = FNMS(KP912018591, T5C, T5B); T88 = FMA(KP912018591, T5C, T5B); T8e = FMA(KP912018591, T4X, T4I); T4Y = FNMS(KP912018591, T4X, T4I); T5d = FMA(KP634619297, T5c, T55); T5y = FNMS(KP634619297, T55, T5c); T5A = FMA(KP912575812, T5z, T5y); T89 = FNMS(KP912575812, T5z, T5y); T8f = FMA(KP912575812, T5s, T5d); T5t = FNMS(KP912575812, T5s, T5d); T5K = FMA(KP256756360, T5J, T5I); T60 = FNMS(KP256756360, T5I, T5J); T8c = FNMS(KP851038619, T89, T88); T8a = FMA(KP851038619, T89, T88); T5u = FNMS(KP851038619, T5t, T4Y); T5w = FMA(KP851038619, T5t, T4Y); } T5U = FMA(KP939062505, T5T, T5S); T64 = FNMS(KP939062505, T5S, T5T); T5N = FMA(KP634619297, T5M, T5L); T61 = FNMS(KP634619297, T5L, T5M); } } { E T62, T7W, T83, T5O, T5R, T63; cr[WS(rs, 4)] = FNMS(KP992114701, T5u, T4t); T62 = FMA(KP871714437, T61, T60); T7W = FNMS(KP871714437, T61, T60); T83 = FNMS(KP871714437, T5N, T5K); T5O = FMA(KP871714437, T5N, T5K); T5R = FMA(KP549754652, T5Q, T5P); T63 = FNMS(KP549754652, T5P, T5Q); ci[WS(rs, 20)] = FNMS(KP992114701, T8a, T87); { E T65, T5W, T84, T86, T81, T85, T8b; { E T5E, T5G, T82, T80, T7Y, T5v, T7X, T5V, T5F, T5x, T7Z; T5E = FNMS(KP726211448, T5D, T5A); T5G = FMA(KP525970792, T5A, T5D); T65 = FNMS(KP831864738, T64, T63); T7X = FMA(KP831864738, T64, T63); T82 = FNMS(KP831864738, T5U, T5R); T5V = FMA(KP831864738, T5U, T5R); T80 = FNMS(KP904730450, T7X, T7W); T7Y = FMA(KP904730450, T7X, T7W); T5Y = FNMS(KP904730450, T5V, T5O); T5W = FMA(KP904730450, T5V, T5O); T5v = FMA(KP248028675, T5u, T4t); ci[WS(rs, 23)] = FMA(KP968583161, T7Y, T7V); cr[WS(rs, 1)] = FMA(KP968583161, T5W, T5H); T84 = FNMS(KP683113946, T83, T82); T86 = FMA(KP559154169, T82, T83); T5F = FNMS(KP554608978, T5w, T5v); T5x = FMA(KP554608978, T5w, T5v); T7Z = FNMS(KP242145790, T7Y, T7V); ci[WS(rs, 10)] = FNMS(KP943557151, T5G, T5F); ci[WS(rs, 5)] = FMA(KP943557151, T5G, T5F); ci[0] = FMA(KP803003575, T5E, T5x); cr[WS(rs, 9)] = FNMS(KP803003575, T5E, T5x); T81 = FNMS(KP541454447, T80, T7Z); T85 = FMA(KP541454447, T80, T7Z); } T8g = FNMS(KP525970792, T8f, T8e); T8i = FMA(KP726211448, T8e, T8f); ci[WS(rs, 13)] = FMA(KP833417178, T84, T81); cr[WS(rs, 16)] = FMS(KP833417178, T84, T81); cr[WS(rs, 21)] = -(FMA(KP921177326, T86, T85)); ci[WS(rs, 18)] = FNMS(KP921177326, T86, T85); T8b = FMA(KP248028675, T8a, T87); T66 = FMA(KP559154169, T65, T62); T68 = FNMS(KP683113946, T62, T65); T5X = FNMS(KP242145790, T5W, T5H); T8d = FNMS(KP554608978, T8c, T8b); T8h = FMA(KP554608978, T8c, T8b); } } } } { E T8s, T8u, T5Z, T67; cr[WS(rs, 24)] = -(FMA(KP803003575, T8i, T8h)); ci[WS(rs, 15)] = FNMS(KP803003575, T8i, T8h); cr[WS(rs, 19)] = FMS(KP943557151, T8g, T8d); cr[WS(rs, 14)] = -(FMA(KP943557151, T8g, T8d)); T5Z = FMA(KP541454447, T5Y, T5X); T67 = FNMS(KP541454447, T5Y, T5X); cr[WS(rs, 11)] = FNMS(KP833417178, T68, T67); ci[WS(rs, 8)] = FMA(KP833417178, T68, T67); cr[WS(rs, 6)] = FMA(KP921177326, T66, T5Z); ci[WS(rs, 3)] = FNMS(KP921177326, T66, T5Z); T8s = FMA(KP618033988, T8r, T8q); T8u = FNMS(KP618033988, T8q, T8r); { E T6X, T6T, T6b, T7H, T7v, T6Y, T72, T71, T6P, T7O, T6M, T7P, T7K, T6G, T6I; E T6W, T7f, T7d, T76; { E T74, T75, T6i, T6N, T6L, T6E, T6U, T6l, T6o, T6V, T6t, T6w; { E T6e, T8o, T8n, T6h, T8m; T6X = FNMS(KP951056516, T6d, T6c); T6e = FMA(KP951056516, T6d, T6c); T8o = T8k - T8l; T8m = T8k + T8l; T6T = FNMS(KP951056516, T6a, T69); T6b = FMA(KP951056516, T6a, T69); T7H = FNMS(KP951056516, T7u, T7r); T7v = FMA(KP951056516, T7u, T7r); ci[WS(rs, 24)] = T8m + T8j; T8n = FNMS(KP250000000, T8m, T8j); T6h = FMA(KP951056516, T6g, T6f); T6Y = FNMS(KP951056516, T6g, T6f); { E T6A, T6D, T8t, T8p; T74 = FMA(KP951056516, T6z, T6y); T6A = FNMS(KP951056516, T6z, T6y); T6D = FMA(KP951056516, T6C, T6B); T75 = FNMS(KP951056516, T6C, T6B); T8t = FMA(KP559016994, T8o, T8n); T8p = FNMS(KP559016994, T8o, T8n); T6i = FMA(KP062914667, T6h, T6e); T6N = FNMS(KP062914667, T6e, T6h); ci[WS(rs, 14)] = FMA(KP951056516, T8s, T8p); cr[WS(rs, 15)] = FMS(KP951056516, T8s, T8p); ci[WS(rs, 19)] = FMA(KP951056516, T8u, T8t); cr[WS(rs, 20)] = FMS(KP951056516, T8u, T8t); T6L = FNMS(KP939062505, T6A, T6D); T6E = FMA(KP939062505, T6D, T6A); } } T6U = FMA(KP951056516, T6k, T6j); T6l = FNMS(KP951056516, T6k, T6j); T6o = FNMS(KP951056516, T6n, T6m); T6V = FMA(KP951056516, T6n, T6m); T72 = FMA(KP951056516, T6s, T6r); T6t = FNMS(KP951056516, T6s, T6r); T6w = FMA(KP951056516, T6v, T6u); T71 = FNMS(KP951056516, T6v, T6u); { E T6q, T6F, T6O, T6p; T6O = FMA(KP827271945, T6l, T6o); T6p = FNMS(KP827271945, T6o, T6l); { E T6K, T6x, T7I, T7J; T6K = FMA(KP126329378, T6t, T6w); T6x = FNMS(KP126329378, T6w, T6t); T7I = FMA(KP772036680, T6O, T6N); T6P = FNMS(KP772036680, T6O, T6N); T6q = FMA(KP772036680, T6p, T6i); T7O = FNMS(KP772036680, T6p, T6i); T7J = FNMS(KP734762448, T6L, T6K); T6M = FMA(KP734762448, T6L, T6K); T6F = FNMS(KP734762448, T6E, T6x); T7P = FMA(KP734762448, T6E, T6x); T7K = FMA(KP994076283, T7J, T7I); T7M = FNMS(KP994076283, T7J, T7I); } T6G = FNMS(KP994076283, T6F, T6q); T6I = FMA(KP994076283, T6F, T6q); } T6W = FMA(KP062914667, T6V, T6U); T7f = FNMS(KP062914667, T6U, T6V); T7d = FNMS(KP549754652, T74, T75); T76 = FMA(KP549754652, T75, T74); } { E T7h, T7C, T7e, T7D, T7y, T7A, T78, T7a; { E T70, T77, T7g, T6Z; cr[WS(rs, 3)] = FMA(KP998026728, T6G, T6b); T7g = FNMS(KP634619297, T6X, T6Y); T6Z = FMA(KP634619297, T6Y, T6X); { E T7c, T73, T7w, T7x; T7c = FMA(KP470564281, T71, T72); T73 = FNMS(KP470564281, T72, T71); T7w = FMA(KP845997307, T7g, T7f); T7h = FNMS(KP845997307, T7g, T7f); T70 = FMA(KP845997307, T6Z, T6W); T7C = FNMS(KP845997307, T6Z, T6W); T7x = FNMS(KP968479752, T7d, T7c); T7e = FMA(KP968479752, T7d, T7c); T77 = FMA(KP968479752, T76, T73); T7D = FNMS(KP968479752, T76, T73); T7y = FMA(KP906616052, T7x, T7w); T7A = FNMS(KP906616052, T7x, T7w); } ci[WS(rs, 21)] = FNMS(KP998026728, T7K, T7H); T78 = FMA(KP906616052, T77, T70); T7a = FNMS(KP906616052, T77, T70); } { E T7G, T7E, T7k, T7i, T79, T7F, T7B, T7z, T6H, T7j, T7b; T7G = FMA(KP681693190, T7C, T7D); T7E = FNMS(KP560319534, T7D, T7C); ci[WS(rs, 22)] = FNMS(KP998026728, T7y, T7v); cr[WS(rs, 2)] = FMA(KP998026728, T78, T6T); T7z = FMA(KP249506682, T7y, T7v); T7k = FNMS(KP560319534, T7e, T7h); T7i = FMA(KP681693190, T7h, T7e); T79 = FNMS(KP249506682, T78, T6T); T7F = FMA(KP557913902, T7A, T7z); T7B = FNMS(KP557913902, T7A, T7z); T6S = FMA(KP614372930, T6M, T6P); T6Q = FNMS(KP621716863, T6P, T6M); cr[WS(rs, 22)] = FMS(KP860541664, T7G, T7F); ci[WS(rs, 17)] = FMA(KP860541664, T7G, T7F); ci[WS(rs, 12)] = FNMS(KP949179823, T7E, T7B); cr[WS(rs, 17)] = -(FMA(KP949179823, T7E, T7B)); T7j = FMA(KP557913902, T7a, T79); T7b = FNMS(KP557913902, T7a, T79); T6H = FNMS(KP249506682, T6G, T6b); ci[WS(rs, 7)] = FMA(KP949179823, T7k, T7j); cr[WS(rs, 12)] = FNMS(KP949179823, T7k, T7j); cr[WS(rs, 7)] = FMA(KP860541664, T7i, T7b); ci[WS(rs, 2)] = FNMS(KP860541664, T7i, T7b); T7S = FMA(KP621716863, T7O, T7P); T7Q = FNMS(KP614372930, T7P, T7O); T7L = FMA(KP249506682, T7K, T7H); T6R = FMA(KP557913902, T6I, T6H); T6J = FNMS(KP557913902, T6I, T6H); } } } } } } } ci[WS(rs, 6)] = FNMS(KP949179823, T6S, T6R); ci[WS(rs, 11)] = FMA(KP949179823, T6S, T6R); cr[WS(rs, 8)] = FMA(KP943557151, T6Q, T6J); ci[WS(rs, 1)] = FNMS(KP943557151, T6Q, T6J); T7N = FNMS(KP557913902, T7M, T7L); T7R = FMA(KP557913902, T7M, T7L); cr[WS(rs, 23)] = -(FMA(KP943557151, T7S, T7R)); ci[WS(rs, 16)] = FNMS(KP943557151, T7S, T7R); cr[WS(rs, 18)] = FMS(KP949179823, T7Q, T7N); cr[WS(rs, 13)] = -(FMA(KP949179823, T7Q, T7N)); } } } static const tw_instr twinstr[] = { {TW_CEXP, 1, 1}, {TW_CEXP, 1, 3}, {TW_CEXP, 1, 9}, {TW_CEXP, 1, 24}, {TW_NEXT, 1, 0} }; static const hc2hc_desc desc = { 25, "hf2_25", twinstr, &GENUS, {84, 78, 356, 0} }; void X(codelet_hf2_25) (planner *p) { X(khc2hc_register) (p, hf2_25, &desc); } #else /* HAVE_FMA */ /* Generated by: ../../../genfft/gen_hc2hc.native -compact -variables 4 -pipeline-latency 4 -twiddle-log3 -precompute-twiddles -n 25 -dit -name hf2_25 -include hf.h */ /* * This function contains 440 FP additions, 340 FP multiplications, * (or, 280 additions, 180 multiplications, 160 fused multiply/add), * 149 stack variables, 20 constants, and 100 memory accesses */ #include "hf.h" static void hf2_25(R *cr, R *ci, const R *W, stride rs, INT mb, INT me, INT ms) { DK(KP998026728, +0.998026728428271561952336806863450553336905220); DK(KP062790519, +0.062790519529313376076178224565631133122484832); DK(KP684547105, +0.684547105928688673732283357621209269889519233); DK(KP728968627, +0.728968627421411523146730319055259111372571664); DK(KP481753674, +0.481753674101715274987191502872129653528542010); DK(KP876306680, +0.876306680043863587308115903922062583399064238); DK(KP248689887, +0.248689887164854788242283746006447968417567406); DK(KP968583161, +0.968583161128631119490168375464735813836012403); DK(KP992114701, +0.992114701314477831049793042785778521453036709); DK(KP125333233, +0.125333233564304245373118759816508793942918247); DK(KP425779291, +0.425779291565072648862502445744251703979973042); DK(KP904827052, +0.904827052466019527713668647932697593970413911); DK(KP637423989, +0.637423989748689710176712811676016195434917298); DK(KP770513242, +0.770513242775789230803009636396177847271667672); DK(KP844327925, +0.844327925502015078548558063966681505381659241); DK(KP535826794, +0.535826794978996618271308767867639978063575346); DK(KP587785252, +0.587785252292473129168705954639072768597652438); DK(KP951056516, +0.951056516295153572116439333379382143405698634); DK(KP250000000, +0.250000000000000000000000000000000000000000000); DK(KP559016994, +0.559016994374947424102293417182819058860154590); { INT m; for (m = mb, W = W + ((mb - 1) * 8); m < me; m = m + 1, cr = cr + ms, ci = ci - ms, W = W + 8, MAKE_VOLATILE_STRIDE(50, rs)) { E T2, T5, T3, T6, T8, Td, T16, T14, Te, T9, T21, T23, Tx, TR, T1g; E TB, T1f, TV, T1Q, Tg, T1S, Tk, T18, T2s, T1c, T2q, Tn, To, Tp, Tr; E T28, T2x, TY, T2k, T2m, T2v, TG, TE, T10, T1h, T1E, T26, T1B, T1G, T1V; E T1X, T1z, T1j; { E Tw, TT, Tz, TQ, Tv, TU, TA, TP; { E T4, Tc, T7, Tb; T2 = W[0]; T5 = W[1]; T3 = W[2]; T6 = W[3]; T4 = T2 * T3; Tc = T5 * T3; T7 = T5 * T6; Tb = T2 * T6; T8 = T4 - T7; Td = Tb + Tc; T16 = Tb - Tc; T14 = T4 + T7; Te = W[5]; Tw = T5 * Te; TT = T3 * Te; Tz = T2 * Te; TQ = T6 * Te; T9 = W[4]; Tv = T2 * T9; TU = T6 * T9; TA = T5 * T9; TP = T3 * T9; } T21 = TP - TQ; T23 = TT + TU; { E T15, T17, Ta, Tf, T1a, T1b, Ti, Tj; Tx = Tv - Tw; TR = TP + TQ; T1g = Tz - TA; TB = Tz + TA; T1f = Tv + Tw; TV = TT - TU; T15 = T14 * T9; T17 = T16 * Te; T1Q = T15 + T17; Ta = T8 * T9; Tf = Td * Te; Tg = Ta + Tf; T1a = T14 * Te; T1b = T16 * T9; T1S = T1a - T1b; Ti = T8 * Te; Tj = Td * T9; Tk = Ti - Tj; T18 = T15 - T17; T2s = Ti + Tj; T1c = T1a + T1b; T2q = Ta - Tf; Tn = W[6]; To = W[7]; Tp = FMA(T8, Tn, Td * To); Tr = FNMS(Td, Tn, T8 * To); T28 = FNMS(T1S, Tn, T1Q * To); T2x = FNMS(TV, Tn, TR * To); TY = FMA(T3, Tn, T6 * To); T2k = FMA(T2, Tn, T5 * To); T2m = FNMS(T5, Tn, T2 * To); T2v = FMA(TR, Tn, TV * To); TG = FNMS(Te, Tn, T9 * To); TE = FMA(T9, Tn, Te * To); T10 = FNMS(T6, Tn, T3 * To); T1h = FMA(T1f, Tn, T1g * To); T1E = FMA(Tg, Tn, Tk * To); T26 = FMA(T1Q, Tn, T1S * To); T1B = FNMS(TB, Tn, Tx * To); T1G = FNMS(Tk, Tn, Tg * To); T1V = FMA(T14, Tn, T16 * To); T1X = FNMS(T16, Tn, T14 * To); T1z = FMA(Tx, Tn, TB * To); T1j = FNMS(T1g, Tn, T1f * To); } } { E T1, T6v, T2F, T6A, TK, T2G, T6y, T6z, T6u, T71, T2O, T52, T2C, T6k, T4c; E T5X, T4L, T5s, T4j, T5W, T4K, T5v, T1o, T6g, T30, T5M, T4A, T56, T3b, T5N; E T4B, T59, T1L, T6h, T3r, T5P, T4E, T5d, T3y, T5Q, T4D, T5g, T2d, T6j, T3P; E T5U, T4I, T5o, T3W, T5T, T4H, T5l; { E Tm, T2I, Tt, T2J, Tu, T6w, TD, T2L, TI, T2M, TJ, T6x; T1 = cr[0]; T6v = ci[0]; { E Th, Tl, Tq, Ts; Th = cr[WS(rs, 5)]; Tl = ci[WS(rs, 5)]; Tm = FMA(Tg, Th, Tk * Tl); T2I = FNMS(Tk, Th, Tg * Tl); Tq = cr[WS(rs, 20)]; Ts = ci[WS(rs, 20)]; Tt = FMA(Tp, Tq, Tr * Ts); T2J = FNMS(Tr, Tq, Tp * Ts); } Tu = Tm + Tt; T6w = T2I + T2J; { E Ty, TC, TF, TH; Ty = cr[WS(rs, 10)]; TC = ci[WS(rs, 10)]; TD = FMA(Tx, Ty, TB * TC); T2L = FNMS(TB, Ty, Tx * TC); TF = cr[WS(rs, 15)]; TH = ci[WS(rs, 15)]; TI = FMA(TE, TF, TG * TH); T2M = FNMS(TG, TF, TE * TH); } TJ = TD + TI; T6x = T2L + T2M; T2F = KP559016994 * (Tu - TJ); T6A = KP559016994 * (T6w - T6x); TK = Tu + TJ; T2G = FNMS(KP250000000, TK, T1); T6y = T6w + T6x; T6z = FNMS(KP250000000, T6y, T6v); { E T6s, T6t, T2K, T2N; T6s = TD - TI; T6t = Tm - Tt; T6u = FNMS(KP587785252, T6t, KP951056516 * T6s); T71 = FMA(KP951056516, T6t, KP587785252 * T6s); T2K = T2I - T2J; T2N = T2L - T2M; T2O = FMA(KP951056516, T2K, KP587785252 * T2N); T52 = FNMS(KP587785252, T2K, KP951056516 * T2N); } } { E T2g, T48, T3Y, T3Z, T4h, T4g, T43, T46, T49, T2p, T2A, T2B, T2e, T2f; T2e = cr[WS(rs, 3)]; T2f = ci[WS(rs, 3)]; T2g = FMA(T3, T2e, T6 * T2f); T48 = FNMS(T6, T2e, T3 * T2f); { E T2j, T41, T2z, T45, T2o, T42, T2u, T44; { E T2h, T2i, T2w, T2y; T2h = cr[WS(rs, 8)]; T2i = ci[WS(rs, 8)]; T2j = FMA(T1f, T2h, T1g * T2i); T41 = FNMS(T1g, T2h, T1f * T2i); T2w = cr[WS(rs, 18)]; T2y = ci[WS(rs, 18)]; T2z = FMA(T2v, T2w, T2x * T2y); T45 = FNMS(T2x, T2w, T2v * T2y); } { E T2l, T2n, T2r, T2t; T2l = cr[WS(rs, 23)]; T2n = ci[WS(rs, 23)]; T2o = FMA(T2k, T2l, T2m * T2n); T42 = FNMS(T2m, T2l, T2k * T2n); T2r = cr[WS(rs, 13)]; T2t = ci[WS(rs, 13)]; T2u = FMA(T2q, T2r, T2s * T2t); T44 = FNMS(T2s, T2r, T2q * T2t); } T3Y = T2j - T2o; T3Z = T2u - T2z; T4h = T44 - T45; T4g = T41 - T42; T43 = T41 + T42; T46 = T44 + T45; T49 = T43 + T46; T2p = T2j + T2o; T2A = T2u + T2z; T2B = T2p + T2A; } T2C = T2g + T2B; T6k = T48 + T49; { E T40, T5r, T4b, T5q, T47, T4a; T40 = FMA(KP951056516, T3Y, KP587785252 * T3Z); T5r = FNMS(KP587785252, T3Y, KP951056516 * T3Z); T47 = KP559016994 * (T43 - T46); T4a = FNMS(KP250000000, T49, T48); T4b = T47 + T4a; T5q = T4a - T47; T4c = T40 + T4b; T5X = T5r + T5q; T4L = T4b - T40; T5s = T5q - T5r; } { E T4i, T5u, T4f, T5t, T4d, T4e; T4i = FMA(KP951056516, T4g, KP587785252 * T4h); T5u = FNMS(KP587785252, T4g, KP951056516 * T4h); T4d = KP559016994 * (T2p - T2A); T4e = FNMS(KP250000000, T2B, T2g); T4f = T4d + T4e; T5t = T4e - T4d; T4j = T4f - T4i; T5W = T5t - T5u; T4K = T4f + T4i; T5v = T5t + T5u; } } { E TO, T37, T2V, T2Y, T32, T31, T34, T35, T38, T13, T1m, T1n, TM, TN; TM = cr[WS(rs, 1)]; TN = ci[WS(rs, 1)]; TO = FMA(T2, TM, T5 * TN); T37 = FNMS(T5, TM, T2 * TN); { E TX, T2T, T1l, T2X, T12, T2U, T1e, T2W; { E TS, TW, T1i, T1k; TS = cr[WS(rs, 6)]; TW = ci[WS(rs, 6)]; TX = FMA(TR, TS, TV * TW); T2T = FNMS(TV, TS, TR * TW); T1i = cr[WS(rs, 16)]; T1k = ci[WS(rs, 16)]; T1l = FMA(T1h, T1i, T1j * T1k); T2X = FNMS(T1j, T1i, T1h * T1k); } { E TZ, T11, T19, T1d; TZ = cr[WS(rs, 21)]; T11 = ci[WS(rs, 21)]; T12 = FMA(TY, TZ, T10 * T11); T2U = FNMS(T10, TZ, TY * T11); T19 = cr[WS(rs, 11)]; T1d = ci[WS(rs, 11)]; T1e = FMA(T18, T19, T1c * T1d); T2W = FNMS(T1c, T19, T18 * T1d); } T2V = T2T - T2U; T2Y = T2W - T2X; T32 = T1e - T1l; T31 = TX - T12; T34 = T2T + T2U; T35 = T2W + T2X; T38 = T34 + T35; T13 = TX + T12; T1m = T1e + T1l; T1n = T13 + T1m; } T1o = TO + T1n; T6g = T37 + T38; { E T2Z, T55, T2S, T54, T2Q, T2R; T2Z = FMA(KP951056516, T2V, KP587785252 * T2Y); T55 = FNMS(KP587785252, T2V, KP951056516 * T2Y); T2Q = KP559016994 * (T13 - T1m); T2R = FNMS(KP250000000, T1n, TO); T2S = T2Q + T2R; T54 = T2R - T2Q; T30 = T2S - T2Z; T5M = T54 - T55; T4A = T2S + T2Z; T56 = T54 + T55; } { E T33, T58, T3a, T57, T36, T39; T33 = FMA(KP951056516, T31, KP587785252 * T32); T58 = FNMS(KP587785252, T31, KP951056516 * T32); T36 = KP559016994 * (T34 - T35); T39 = FNMS(KP250000000, T38, T37); T3a = T36 + T39; T57 = T39 - T36; T3b = T33 + T3a; T5N = T58 + T57; T4B = T3a - T33; T59 = T57 - T58; } } { E T1r, T3n, T3d, T3e, T3w, T3v, T3i, T3l, T3o, T1y, T1J, T1K, T1p, T1q; T1p = cr[WS(rs, 4)]; T1q = ci[WS(rs, 4)]; T1r = FMA(T8, T1p, Td * T1q); T3n = FNMS(Td, T1p, T8 * T1q); { E T1u, T3g, T1I, T3k, T1x, T3h, T1D, T3j; { E T1s, T1t, T1F, T1H; T1s = cr[WS(rs, 9)]; T1t = ci[WS(rs, 9)]; T1u = FMA(T9, T1s, Te * T1t); T3g = FNMS(Te, T1s, T9 * T1t); T1F = cr[WS(rs, 19)]; T1H = ci[WS(rs, 19)]; T1I = FMA(T1E, T1F, T1G * T1H); T3k = FNMS(T1G, T1F, T1E * T1H); } { E T1v, T1w, T1A, T1C; T1v = cr[WS(rs, 24)]; T1w = ci[WS(rs, 24)]; T1x = FMA(Tn, T1v, To * T1w); T3h = FNMS(To, T1v, Tn * T1w); T1A = cr[WS(rs, 14)]; T1C = ci[WS(rs, 14)]; T1D = FMA(T1z, T1A, T1B * T1C); T3j = FNMS(T1B, T1A, T1z * T1C); } T3d = T1x - T1u; T3e = T1D - T1I; T3w = T3j - T3k; T3v = T3g - T3h; T3i = T3g + T3h; T3l = T3j + T3k; T3o = T3i + T3l; T1y = T1u + T1x; T1J = T1D + T1I; T1K = T1y + T1J; } T1L = T1r + T1K; T6h = T3n + T3o; { E T3f, T5c, T3q, T5b, T3m, T3p; T3f = FNMS(KP587785252, T3e, KP951056516 * T3d); T5c = FMA(KP587785252, T3d, KP951056516 * T3e); T3m = KP559016994 * (T3i - T3l); T3p = FNMS(KP250000000, T3o, T3n); T3q = T3m + T3p; T5b = T3p - T3m; T3r = T3f - T3q; T5P = T5c + T5b; T4E = T3f + T3q; T5d = T5b - T5c; } { E T3x, T5f, T3u, T5e, T3s, T3t; T3x = FMA(KP951056516, T3v, KP587785252 * T3w); T5f = FNMS(KP587785252, T3v, KP951056516 * T3w); T3s = KP559016994 * (T1y - T1J); T3t = FNMS(KP250000000, T1K, T1r); T3u = T3s + T3t; T5e = T3t - T3s; T3y = T3u - T3x; T5Q = T5e - T5f; T4D = T3u + T3x; T5g = T5e + T5f; } } { E T1P, T3L, T3B, T3C, T3U, T3T, T3G, T3J, T3M, T20, T2b, T2c, T1N, T1O; T1N = cr[WS(rs, 2)]; T1O = ci[WS(rs, 2)]; T1P = FMA(T14, T1N, T16 * T1O); T3L = FNMS(T16, T1N, T14 * T1O); { E T1U, T3E, T2a, T3I, T1Z, T3F, T25, T3H; { E T1R, T1T, T27, T29; T1R = cr[WS(rs, 7)]; T1T = ci[WS(rs, 7)]; T1U = FMA(T1Q, T1R, T1S * T1T); T3E = FNMS(T1S, T1R, T1Q * T1T); T27 = cr[WS(rs, 17)]; T29 = ci[WS(rs, 17)]; T2a = FMA(T26, T27, T28 * T29); T3I = FNMS(T28, T27, T26 * T29); } { E T1W, T1Y, T22, T24; T1W = cr[WS(rs, 22)]; T1Y = ci[WS(rs, 22)]; T1Z = FMA(T1V, T1W, T1X * T1Y); T3F = FNMS(T1X, T1W, T1V * T1Y); T22 = cr[WS(rs, 12)]; T24 = ci[WS(rs, 12)]; T25 = FMA(T21, T22, T23 * T24); T3H = FNMS(T23, T22, T21 * T24); } T3B = T1U - T1Z; T3C = T25 - T2a; T3U = T3H - T3I; T3T = T3E - T3F; T3G = T3E + T3F; T3J = T3H + T3I; T3M = T3G + T3J; T20 = T1U + T1Z; T2b = T25 + T2a; T2c = T20 + T2b; } T2d = T1P + T2c; T6j = T3L + T3M; { E T3D, T5n, T3O, T5m, T3K, T3N; T3D = FMA(KP951056516, T3B, KP587785252 * T3C); T5n = FNMS(KP587785252, T3B, KP951056516 * T3C); T3K = KP559016994 * (T3G - T3J); T3N = FNMS(KP250000000, T3M, T3L); T3O = T3K + T3N; T5m = T3N - T3K; T3P = T3D + T3O; T5U = T5n + T5m; T4I = T3O - T3D; T5o = T5m - T5n; } { E T3V, T5k, T3S, T5j, T3Q, T3R; T3V = FMA(KP951056516, T3T, KP587785252 * T3U); T5k = FNMS(KP587785252, T3T, KP951056516 * T3U); T3Q = KP559016994 * (T20 - T2b); T3R = FNMS(KP250000000, T2c, T1P); T3S = T3Q + T3R; T5j = T3R - T3Q; T3W = T3S - T3V; T5T = T5j - T5k; T4H = T3S + T3V; T5l = T5j + T5k; } } { E T6m, T6o, TL, T2E, T6d, T6e, T6n, T6f; { E T6i, T6l, T1M, T2D; T6i = T6g - T6h; T6l = T6j - T6k; T6m = FMA(KP951056516, T6i, KP587785252 * T6l); T6o = FNMS(KP587785252, T6i, KP951056516 * T6l); TL = T1 + TK; T1M = T1o + T1L; T2D = T2d + T2C; T2E = T1M + T2D; T6d = KP559016994 * (T1M - T2D); T6e = FNMS(KP250000000, T2E, TL); } cr[0] = TL + T2E; T6n = T6e - T6d; cr[WS(rs, 10)] = T6n - T6o; ci[WS(rs, 9)] = T6n + T6o; T6f = T6d + T6e; ci[WS(rs, 4)] = T6f - T6m; cr[WS(rs, 5)] = T6f + T6m; } { E T2P, T4z, T72, T7e, T4m, T7j, T4n, T7i, T4U, T77, T4X, T75, T4O, T6Y, T4P; E T6X, T4s, T7f, T4v, T7d, T2H, T70; T2H = T2F + T2G; T2P = T2H - T2O; T4z = T2H + T2O; T70 = T6A + T6z; T72 = T70 - T71; T7e = T71 + T70; { E T3c, T3z, T3A, T3X, T4k, T4l; T3c = FMA(KP535826794, T30, KP844327925 * T3b); T3z = FNMS(KP637423989, T3y, KP770513242 * T3r); T3A = T3c + T3z; T3X = FNMS(KP425779291, T3W, KP904827052 * T3P); T4k = FNMS(KP992114701, T4j, KP125333233 * T4c); T4l = T3X + T4k; T4m = T3A + T4l; T7j = T3X - T4k; T4n = KP559016994 * (T3A - T4l); T7i = T3z - T3c; } { E T4S, T4T, T73, T4V, T4W, T74; T4S = FNMS(KP248689887, T4A, KP968583161 * T4B); T4T = FNMS(KP844327925, T4D, KP535826794 * T4E); T73 = T4S + T4T; T4V = FNMS(KP481753674, T4H, KP876306680 * T4I); T4W = FNMS(KP684547105, T4K, KP728968627 * T4L); T74 = T4V + T4W; T4U = T4S - T4T; T77 = KP559016994 * (T73 - T74); T4X = T4V - T4W; T75 = T73 + T74; } { E T4C, T4F, T4G, T4J, T4M, T4N; T4C = FMA(KP968583161, T4A, KP248689887 * T4B); T4F = FMA(KP535826794, T4D, KP844327925 * T4E); T4G = T4C + T4F; T4J = FMA(KP876306680, T4H, KP481753674 * T4I); T4M = FMA(KP728968627, T4K, KP684547105 * T4L); T4N = T4J + T4M; T4O = T4G + T4N; T6Y = T4J - T4M; T4P = KP559016994 * (T4G - T4N); T6X = T4F - T4C; } { E T4q, T4r, T7b, T4t, T4u, T7c; T4q = FNMS(KP844327925, T30, KP535826794 * T3b); T4r = FMA(KP770513242, T3y, KP637423989 * T3r); T7b = T4q + T4r; T4t = FMA(KP125333233, T4j, KP992114701 * T4c); T4u = FMA(KP904827052, T3W, KP425779291 * T3P); T7c = T4u + T4t; T4s = T4q - T4r; T7f = T7b - T7c; T4v = T4t - T4u; T7d = KP559016994 * (T7b + T7c); } cr[WS(rs, 4)] = T2P + T4m; ci[WS(rs, 23)] = T75 + T72; ci[WS(rs, 20)] = T7f + T7e; cr[WS(rs, 1)] = T4z + T4O; { E T4w, T4y, T4p, T4x, T4o; T4w = FMA(KP951056516, T4s, KP587785252 * T4v); T4y = FNMS(KP587785252, T4s, KP951056516 * T4v); T4o = FNMS(KP250000000, T4m, T2P); T4p = T4n + T4o; T4x = T4o - T4n; ci[0] = T4p - T4w; ci[WS(rs, 5)] = T4x + T4y; cr[WS(rs, 9)] = T4p + T4w; ci[WS(rs, 10)] = T4x - T4y; } { E T6Z, T79, T78, T7a, T76; T6Z = FMA(KP587785252, T6X, KP951056516 * T6Y); T79 = FNMS(KP587785252, T6Y, KP951056516 * T6X); T76 = FNMS(KP250000000, T75, T72); T78 = T76 - T77; T7a = T77 + T76; cr[WS(rs, 16)] = T6Z - T78; ci[WS(rs, 18)] = T79 + T7a; ci[WS(rs, 13)] = T6Z + T78; cr[WS(rs, 21)] = T79 - T7a; } { E T7k, T7l, T7h, T7m, T7g; T7k = FMA(KP587785252, T7i, KP951056516 * T7j); T7l = FNMS(KP587785252, T7j, KP951056516 * T7i); T7g = FNMS(KP250000000, T7f, T7e); T7h = T7d - T7g; T7m = T7d + T7g; cr[WS(rs, 14)] = T7h - T7k; ci[WS(rs, 15)] = T7l + T7m; cr[WS(rs, 19)] = T7k + T7h; cr[WS(rs, 24)] = T7l - T7m; } { E T4Y, T50, T4R, T4Z, T4Q; T4Y = FMA(KP951056516, T4U, KP587785252 * T4X); T50 = FNMS(KP587785252, T4U, KP951056516 * T4X); T4Q = FNMS(KP250000000, T4O, T4z); T4R = T4P + T4Q; T4Z = T4Q - T4P; ci[WS(rs, 3)] = T4R - T4Y; ci[WS(rs, 8)] = T4Z + T50; cr[WS(rs, 6)] = T4R + T4Y; cr[WS(rs, 11)] = T4Z - T50; } } { E T7p, T7x, T7q, T7t, T7u, T7v, T7y, T7w; { E T7n, T7o, T7r, T7s; T7n = T1L - T1o; T7o = T2d - T2C; T7p = FMA(KP587785252, T7n, KP951056516 * T7o); T7x = FNMS(KP587785252, T7o, KP951056516 * T7n); T7q = T6y + T6v; T7r = T6g + T6h; T7s = T6j + T6k; T7t = T7r + T7s; T7u = FNMS(KP250000000, T7t, T7q); T7v = KP559016994 * (T7r - T7s); } ci[WS(rs, 24)] = T7t + T7q; T7y = T7v + T7u; cr[WS(rs, 20)] = T7x - T7y; ci[WS(rs, 19)] = T7x + T7y; T7w = T7u - T7v; cr[WS(rs, 15)] = T7p - T7w; ci[WS(rs, 14)] = T7p + T7w; } { E T53, T5L, T6C, T6O, T5y, T6T, T5z, T6S, T66, T6H, T69, T6F, T60, T6q, T61; E T6p, T5E, T6P, T5H, T6N, T51, T6B; T51 = T2G - T2F; T53 = T51 + T52; T5L = T51 - T52; T6B = T6z - T6A; T6C = T6u + T6B; T6O = T6B - T6u; { E T5a, T5h, T5i, T5p, T5w, T5x; T5a = FMA(KP728968627, T56, KP684547105 * T59); T5h = FNMS(KP992114701, T5g, KP125333233 * T5d); T5i = T5a + T5h; T5p = FMA(KP062790519, T5l, KP998026728 * T5o); T5w = FNMS(KP637423989, T5v, KP770513242 * T5s); T5x = T5p + T5w; T5y = T5i + T5x; T6T = T5p - T5w; T5z = KP559016994 * (T5i - T5x); T6S = T5h - T5a; } { E T64, T65, T6D, T67, T68, T6E; T64 = FNMS(KP481753674, T5M, KP876306680 * T5N); T65 = FMA(KP904827052, T5Q, KP425779291 * T5P); T6D = T64 - T65; T67 = FNMS(KP844327925, T5T, KP535826794 * T5U); T68 = FNMS(KP998026728, T5W, KP062790519 * T5X); T6E = T67 + T68; T66 = T64 + T65; T6H = KP559016994 * (T6D - T6E); T69 = T67 - T68; T6F = T6D + T6E; } { E T5O, T5R, T5S, T5V, T5Y, T5Z; T5O = FMA(KP876306680, T5M, KP481753674 * T5N); T5R = FNMS(KP425779291, T5Q, KP904827052 * T5P); T5S = T5O + T5R; T5V = FMA(KP535826794, T5T, KP844327925 * T5U); T5Y = FMA(KP062790519, T5W, KP998026728 * T5X); T5Z = T5V + T5Y; T60 = T5S + T5Z; T6q = T5V - T5Y; T61 = KP559016994 * (T5S - T5Z); T6p = T5R - T5O; } { E T5C, T5D, T6L, T5F, T5G, T6M; T5C = FNMS(KP684547105, T56, KP728968627 * T59); T5D = FMA(KP125333233, T5g, KP992114701 * T5d); T6L = T5C - T5D; T5F = FNMS(KP998026728, T5l, KP062790519 * T5o); T5G = FMA(KP770513242, T5v, KP637423989 * T5s); T6M = T5F - T5G; T5E = T5C + T5D; T6P = T6L + T6M; T5H = T5F + T5G; T6N = KP559016994 * (T6L - T6M); } cr[WS(rs, 3)] = T53 + T5y; ci[WS(rs, 22)] = T6F + T6C; ci[WS(rs, 21)] = T6P + T6O; cr[WS(rs, 2)] = T5L + T60; { E T6r, T6J, T6I, T6K, T6G; T6r = FMA(KP587785252, T6p, KP951056516 * T6q); T6J = FNMS(KP587785252, T6q, KP951056516 * T6p); T6G = FNMS(KP250000000, T6F, T6C); T6I = T6G - T6H; T6K = T6H + T6G; cr[WS(rs, 17)] = T6r - T6I; ci[WS(rs, 17)] = T6J + T6K; ci[WS(rs, 12)] = T6r + T6I; cr[WS(rs, 22)] = T6J - T6K; } { E T6a, T6c, T63, T6b, T62; T6a = FMA(KP951056516, T66, KP587785252 * T69); T6c = FNMS(KP587785252, T66, KP951056516 * T69); T62 = FNMS(KP250000000, T60, T5L); T63 = T61 + T62; T6b = T62 - T61; ci[WS(rs, 2)] = T63 - T6a; ci[WS(rs, 7)] = T6b + T6c; cr[WS(rs, 7)] = T63 + T6a; cr[WS(rs, 12)] = T6b - T6c; } { E T5I, T5K, T5B, T5J, T5A; T5I = FMA(KP951056516, T5E, KP587785252 * T5H); T5K = FNMS(KP587785252, T5E, KP951056516 * T5H); T5A = FNMS(KP250000000, T5y, T53); T5B = T5z + T5A; T5J = T5A - T5z; ci[WS(rs, 1)] = T5B - T5I; ci[WS(rs, 6)] = T5J + T5K; cr[WS(rs, 8)] = T5B + T5I; ci[WS(rs, 11)] = T5J - T5K; } { E T6U, T6V, T6R, T6W, T6Q; T6U = FMA(KP587785252, T6S, KP951056516 * T6T); T6V = FNMS(KP587785252, T6T, KP951056516 * T6S); T6Q = FNMS(KP250000000, T6P, T6O); T6R = T6N - T6Q; T6W = T6N + T6Q; cr[WS(rs, 13)] = T6R - T6U; ci[WS(rs, 16)] = T6V + T6W; cr[WS(rs, 18)] = T6U + T6R; cr[WS(rs, 23)] = T6V - T6W; } } } } } } static const tw_instr twinstr[] = { {TW_CEXP, 1, 1}, {TW_CEXP, 1, 3}, {TW_CEXP, 1, 9}, {TW_CEXP, 1, 24}, {TW_NEXT, 1, 0} }; static const hc2hc_desc desc = { 25, "hf2_25", twinstr, &GENUS, {280, 180, 160, 0} }; void X(codelet_hf2_25) (planner *p) { X(khc2hc_register) (p, hf2_25, &desc); } #endif /* HAVE_FMA */ fftw-3.3.3/rdft/scalar/r2cf/r2cfII_2.c0000644000175000017500000000562712054410655014126 00000000000000/* * Copyright (c) 2003, 2007-11 Matteo Frigo * Copyright (c) 2003, 2007-11 Massachusetts Institute of Technology * * This program is free software; you can redistribute it and/or modify * it under the terms of the GNU General Public License as published by * the Free Software Foundation; either version 2 of the License, or * (at your option) any later version. * * This program is distributed in the hope that it will be useful, * but WITHOUT ANY WARRANTY; without even the implied warranty of * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the * GNU General Public License for more details. * * You should have received a copy of the GNU General Public License * along with this program; if not, write to the Free Software * Foundation, Inc., 51 Franklin Street, Fifth Floor, Boston, MA 02110-1301 USA * */ /* This file was automatically generated --- DO NOT EDIT */ /* Generated on Sun Nov 25 07:40:13 EST 2012 */ #include "codelet-rdft.h" #ifdef HAVE_FMA /* Generated by: ../../../genfft/gen_r2cf.native -fma -reorder-insns -schedule-for-pipeline -compact -variables 4 -pipeline-latency 4 -n 2 -name r2cfII_2 -dft-II -include r2cfII.h */ /* * This function contains 0 FP additions, 0 FP multiplications, * (or, 0 additions, 0 multiplications, 0 fused multiply/add), * 3 stack variables, 0 constants, and 4 memory accesses */ #include "r2cfII.h" static void r2cfII_2(R *R0, R *R1, R *Cr, R *Ci, stride rs, stride csr, stride csi, INT v, INT ivs, INT ovs) { { INT i; for (i = v; i > 0; i = i - 1, R0 = R0 + ivs, R1 = R1 + ivs, Cr = Cr + ovs, Ci = Ci + ovs, MAKE_VOLATILE_STRIDE(8, rs), MAKE_VOLATILE_STRIDE(8, csr), MAKE_VOLATILE_STRIDE(8, csi)) { E T1, T2; T1 = R0[0]; T2 = R1[0]; Cr[0] = T1; Ci[0] = -T2; } } } static const kr2c_desc desc = { 2, "r2cfII_2", {0, 0, 0, 0}, &GENUS }; void X(codelet_r2cfII_2) (planner *p) { X(kr2c_register) (p, r2cfII_2, &desc); } #else /* HAVE_FMA */ /* Generated by: ../../../genfft/gen_r2cf.native -compact -variables 4 -pipeline-latency 4 -n 2 -name r2cfII_2 -dft-II -include r2cfII.h */ /* * This function contains 0 FP additions, 0 FP multiplications, * (or, 0 additions, 0 multiplications, 0 fused multiply/add), * 3 stack variables, 0 constants, and 4 memory accesses */ #include "r2cfII.h" static void r2cfII_2(R *R0, R *R1, R *Cr, R *Ci, stride rs, stride csr, stride csi, INT v, INT ivs, INT ovs) { { INT i; for (i = v; i > 0; i = i - 1, R0 = R0 + ivs, R1 = R1 + ivs, Cr = Cr + ovs, Ci = Ci + ovs, MAKE_VOLATILE_STRIDE(8, rs), MAKE_VOLATILE_STRIDE(8, csr), MAKE_VOLATILE_STRIDE(8, csi)) { E T1, T2; T1 = R0[0]; T2 = R1[0]; Cr[0] = T1; Ci[0] = -T2; } } } static const kr2c_desc desc = { 2, "r2cfII_2", {0, 0, 0, 0}, &GENUS }; void X(codelet_r2cfII_2) (planner *p) { X(kr2c_register) (p, r2cfII_2, &desc); } #endif /* HAVE_FMA */ fftw-3.3.3/rdft/scalar/r2cf/r2cfII_3.c0000644000175000017500000000672012054410655014122 00000000000000/* * Copyright (c) 2003, 2007-11 Matteo Frigo * Copyright (c) 2003, 2007-11 Massachusetts Institute of Technology * * This program is free software; you can redistribute it and/or modify * it under the terms of the GNU General Public License as published by * the Free Software Foundation; either version 2 of the License, or * (at your option) any later version. * * This program is distributed in the hope that it will be useful, * but WITHOUT ANY WARRANTY; without even the implied warranty of * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the * GNU General Public License for more details. * * You should have received a copy of the GNU General Public License * along with this program; if not, write to the Free Software * Foundation, Inc., 51 Franklin Street, Fifth Floor, Boston, MA 02110-1301 USA * */ /* This file was automatically generated --- DO NOT EDIT */ /* Generated on Sun Nov 25 07:40:13 EST 2012 */ #include "codelet-rdft.h" #ifdef HAVE_FMA /* Generated by: ../../../genfft/gen_r2cf.native -fma -reorder-insns -schedule-for-pipeline -compact -variables 4 -pipeline-latency 4 -n 3 -name r2cfII_3 -dft-II -include r2cfII.h */ /* * This function contains 4 FP additions, 2 FP multiplications, * (or, 3 additions, 1 multiplications, 1 fused multiply/add), * 7 stack variables, 2 constants, and 6 memory accesses */ #include "r2cfII.h" static void r2cfII_3(R *R0, R *R1, R *Cr, R *Ci, stride rs, stride csr, stride csi, INT v, INT ivs, INT ovs) { DK(KP500000000, +0.500000000000000000000000000000000000000000000); DK(KP866025403, +0.866025403784438646763723170752936183471402627); { INT i; for (i = v; i > 0; i = i - 1, R0 = R0 + ivs, R1 = R1 + ivs, Cr = Cr + ovs, Ci = Ci + ovs, MAKE_VOLATILE_STRIDE(12, rs), MAKE_VOLATILE_STRIDE(12, csr), MAKE_VOLATILE_STRIDE(12, csi)) { E T3, T1, T2, T4; T3 = R0[0]; T1 = R1[0]; T2 = R0[WS(rs, 1)]; Ci[0] = -(KP866025403 * (T1 + T2)); T4 = T2 - T1; Cr[WS(csr, 1)] = T3 + T4; Cr[0] = FNMS(KP500000000, T4, T3); } } } static const kr2c_desc desc = { 3, "r2cfII_3", {3, 1, 1, 0}, &GENUS }; void X(codelet_r2cfII_3) (planner *p) { X(kr2c_register) (p, r2cfII_3, &desc); } #else /* HAVE_FMA */ /* Generated by: ../../../genfft/gen_r2cf.native -compact -variables 4 -pipeline-latency 4 -n 3 -name r2cfII_3 -dft-II -include r2cfII.h */ /* * This function contains 4 FP additions, 2 FP multiplications, * (or, 3 additions, 1 multiplications, 1 fused multiply/add), * 7 stack variables, 2 constants, and 6 memory accesses */ #include "r2cfII.h" static void r2cfII_3(R *R0, R *R1, R *Cr, R *Ci, stride rs, stride csr, stride csi, INT v, INT ivs, INT ovs) { DK(KP500000000, +0.500000000000000000000000000000000000000000000); DK(KP866025403, +0.866025403784438646763723170752936183471402627); { INT i; for (i = v; i > 0; i = i - 1, R0 = R0 + ivs, R1 = R1 + ivs, Cr = Cr + ovs, Ci = Ci + ovs, MAKE_VOLATILE_STRIDE(12, rs), MAKE_VOLATILE_STRIDE(12, csr), MAKE_VOLATILE_STRIDE(12, csi)) { E T1, T2, T3, T4; T1 = R0[0]; T2 = R1[0]; T3 = R0[WS(rs, 1)]; T4 = T2 - T3; Cr[WS(csr, 1)] = T1 - T4; Ci[0] = -(KP866025403 * (T2 + T3)); Cr[0] = FMA(KP500000000, T4, T1); } } } static const kr2c_desc desc = { 3, "r2cfII_3", {3, 1, 1, 0}, &GENUS }; void X(codelet_r2cfII_3) (planner *p) { X(kr2c_register) (p, r2cfII_3, &desc); } #endif /* HAVE_FMA */ fftw-3.3.3/rdft/scalar/r2cf/r2cfII_4.c0000644000175000017500000000703312054410655014121 00000000000000/* * Copyright (c) 2003, 2007-11 Matteo Frigo * Copyright (c) 2003, 2007-11 Massachusetts Institute of Technology * * This program is free software; you can redistribute it and/or modify * it under the terms of the GNU General Public License as published by * the Free Software Foundation; either version 2 of the License, or * (at your option) any later version. * * This program is distributed in the hope that it will be useful, * but WITHOUT ANY WARRANTY; without even the implied warranty of * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the * GNU General Public License for more details. * * You should have received a copy of the GNU General Public License * along with this program; if not, write to the Free Software * Foundation, Inc., 51 Franklin Street, Fifth Floor, Boston, MA 02110-1301 USA * */ /* This file was automatically generated --- DO NOT EDIT */ /* Generated on Sun Nov 25 07:40:13 EST 2012 */ #include "codelet-rdft.h" #ifdef HAVE_FMA /* Generated by: ../../../genfft/gen_r2cf.native -fma -reorder-insns -schedule-for-pipeline -compact -variables 4 -pipeline-latency 4 -n 4 -name r2cfII_4 -dft-II -include r2cfII.h */ /* * This function contains 6 FP additions, 4 FP multiplications, * (or, 2 additions, 0 multiplications, 4 fused multiply/add), * 8 stack variables, 1 constants, and 8 memory accesses */ #include "r2cfII.h" static void r2cfII_4(R *R0, R *R1, R *Cr, R *Ci, stride rs, stride csr, stride csi, INT v, INT ivs, INT ovs) { DK(KP707106781, +0.707106781186547524400844362104849039284835938); { INT i; for (i = v; i > 0; i = i - 1, R0 = R0 + ivs, R1 = R1 + ivs, Cr = Cr + ovs, Ci = Ci + ovs, MAKE_VOLATILE_STRIDE(16, rs), MAKE_VOLATILE_STRIDE(16, csr), MAKE_VOLATILE_STRIDE(16, csi)) { E T1, T5, T2, T3, T4, T6; T1 = R0[0]; T5 = R0[WS(rs, 1)]; T2 = R1[0]; T3 = R1[WS(rs, 1)]; T4 = T2 - T3; T6 = T2 + T3; Ci[0] = -(FMA(KP707106781, T6, T5)); Ci[WS(csi, 1)] = FNMS(KP707106781, T6, T5); Cr[0] = FMA(KP707106781, T4, T1); Cr[WS(csr, 1)] = FNMS(KP707106781, T4, T1); } } } static const kr2c_desc desc = { 4, "r2cfII_4", {2, 0, 4, 0}, &GENUS }; void X(codelet_r2cfII_4) (planner *p) { X(kr2c_register) (p, r2cfII_4, &desc); } #else /* HAVE_FMA */ /* Generated by: ../../../genfft/gen_r2cf.native -compact -variables 4 -pipeline-latency 4 -n 4 -name r2cfII_4 -dft-II -include r2cfII.h */ /* * This function contains 6 FP additions, 2 FP multiplications, * (or, 6 additions, 2 multiplications, 0 fused multiply/add), * 8 stack variables, 1 constants, and 8 memory accesses */ #include "r2cfII.h" static void r2cfII_4(R *R0, R *R1, R *Cr, R *Ci, stride rs, stride csr, stride csi, INT v, INT ivs, INT ovs) { DK(KP707106781, +0.707106781186547524400844362104849039284835938); { INT i; for (i = v; i > 0; i = i - 1, R0 = R0 + ivs, R1 = R1 + ivs, Cr = Cr + ovs, Ci = Ci + ovs, MAKE_VOLATILE_STRIDE(16, rs), MAKE_VOLATILE_STRIDE(16, csr), MAKE_VOLATILE_STRIDE(16, csi)) { E T1, T6, T4, T5, T2, T3; T1 = R0[0]; T6 = R0[WS(rs, 1)]; T2 = R1[0]; T3 = R1[WS(rs, 1)]; T4 = KP707106781 * (T2 - T3); T5 = KP707106781 * (T2 + T3); Cr[WS(csr, 1)] = T1 - T4; Ci[WS(csi, 1)] = T6 - T5; Cr[0] = T1 + T4; Ci[0] = -(T5 + T6); } } } static const kr2c_desc desc = { 4, "r2cfII_4", {6, 2, 0, 0}, &GENUS }; void X(codelet_r2cfII_4) (planner *p) { X(kr2c_register) (p, r2cfII_4, &desc); } #endif /* HAVE_FMA */ fftw-3.3.3/rdft/scalar/r2cf/r2cfII_5.c0000644000175000017500000001073512054410655014125 00000000000000/* * Copyright (c) 2003, 2007-11 Matteo Frigo * Copyright (c) 2003, 2007-11 Massachusetts Institute of Technology * * This program is free software; you can redistribute it and/or modify * it under the terms of the GNU General Public License as published by * the Free Software Foundation; either version 2 of the License, or * (at your option) any later version. * * This program is distributed in the hope that it will be useful, * but WITHOUT ANY WARRANTY; without even the implied warranty of * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the * GNU General Public License for more details. * * You should have received a copy of the GNU General Public License * along with this program; if not, write to the Free Software * Foundation, Inc., 51 Franklin Street, Fifth Floor, Boston, MA 02110-1301 USA * */ /* This file was automatically generated --- DO NOT EDIT */ /* Generated on Sun Nov 25 07:40:13 EST 2012 */ #include "codelet-rdft.h" #ifdef HAVE_FMA /* Generated by: ../../../genfft/gen_r2cf.native -fma -reorder-insns -schedule-for-pipeline -compact -variables 4 -pipeline-latency 4 -n 5 -name r2cfII_5 -dft-II -include r2cfII.h */ /* * This function contains 12 FP additions, 7 FP multiplications, * (or, 7 additions, 2 multiplications, 5 fused multiply/add), * 17 stack variables, 4 constants, and 10 memory accesses */ #include "r2cfII.h" static void r2cfII_5(R *R0, R *R1, R *Cr, R *Ci, stride rs, stride csr, stride csi, INT v, INT ivs, INT ovs) { DK(KP559016994, +0.559016994374947424102293417182819058860154590); DK(KP250000000, +0.250000000000000000000000000000000000000000000); DK(KP618033988, +0.618033988749894848204586834365638117720309180); DK(KP951056516, +0.951056516295153572116439333379382143405698634); { INT i; for (i = v; i > 0; i = i - 1, R0 = R0 + ivs, R1 = R1 + ivs, Cr = Cr + ovs, Ci = Ci + ovs, MAKE_VOLATILE_STRIDE(20, rs), MAKE_VOLATILE_STRIDE(20, csr), MAKE_VOLATILE_STRIDE(20, csi)) { E T1, T2, T3, T5, T6; T1 = R0[0]; T2 = R0[WS(rs, 1)]; T3 = R1[WS(rs, 1)]; T5 = R0[WS(rs, 2)]; T6 = R1[0]; { E Tb, T4, Tc, T7, Ta, T8, T9; Tb = T2 + T3; T4 = T2 - T3; Tc = T5 + T6; T7 = T5 - T6; Ci[0] = -(KP951056516 * (FMA(KP618033988, Tc, Tb))); Ci[WS(csi, 1)] = -(KP951056516 * (FNMS(KP618033988, Tb, Tc))); Ta = T4 - T7; T8 = T4 + T7; T9 = FNMS(KP250000000, T8, T1); Cr[WS(csr, 2)] = T1 + T8; Cr[WS(csr, 1)] = FNMS(KP559016994, Ta, T9); Cr[0] = FMA(KP559016994, Ta, T9); } } } } static const kr2c_desc desc = { 5, "r2cfII_5", {7, 2, 5, 0}, &GENUS }; void X(codelet_r2cfII_5) (planner *p) { X(kr2c_register) (p, r2cfII_5, &desc); } #else /* HAVE_FMA */ /* Generated by: ../../../genfft/gen_r2cf.native -compact -variables 4 -pipeline-latency 4 -n 5 -name r2cfII_5 -dft-II -include r2cfII.h */ /* * This function contains 12 FP additions, 6 FP multiplications, * (or, 9 additions, 3 multiplications, 3 fused multiply/add), * 17 stack variables, 4 constants, and 10 memory accesses */ #include "r2cfII.h" static void r2cfII_5(R *R0, R *R1, R *Cr, R *Ci, stride rs, stride csr, stride csi, INT v, INT ivs, INT ovs) { DK(KP250000000, +0.250000000000000000000000000000000000000000000); DK(KP559016994, +0.559016994374947424102293417182819058860154590); DK(KP951056516, +0.951056516295153572116439333379382143405698634); DK(KP587785252, +0.587785252292473129168705954639072768597652438); { INT i; for (i = v; i > 0; i = i - 1, R0 = R0 + ivs, R1 = R1 + ivs, Cr = Cr + ovs, Ci = Ci + ovs, MAKE_VOLATILE_STRIDE(20, rs), MAKE_VOLATILE_STRIDE(20, csr), MAKE_VOLATILE_STRIDE(20, csi)) { E T8, T3, T6, T9, Tc, Tb, T7, Ta; T8 = R0[0]; { E T1, T2, T4, T5; T1 = R0[WS(rs, 1)]; T2 = R1[WS(rs, 1)]; T3 = T1 - T2; T4 = R0[WS(rs, 2)]; T5 = R1[0]; T6 = T4 - T5; T9 = T3 + T6; Tc = T4 + T5; Tb = T1 + T2; } Cr[WS(csr, 2)] = T8 + T9; Ci[WS(csi, 1)] = FNMS(KP951056516, Tc, KP587785252 * Tb); Ci[0] = -(FMA(KP951056516, Tb, KP587785252 * Tc)); T7 = KP559016994 * (T3 - T6); Ta = FNMS(KP250000000, T9, T8); Cr[0] = T7 + Ta; Cr[WS(csr, 1)] = Ta - T7; } } } static const kr2c_desc desc = { 5, "r2cfII_5", {9, 3, 3, 0}, &GENUS }; void X(codelet_r2cfII_5) (planner *p) { X(kr2c_register) (p, r2cfII_5, &desc); } #endif /* HAVE_FMA */ fftw-3.3.3/rdft/scalar/r2cf/r2cfII_6.c0000644000175000017500000001030712054410655014121 00000000000000/* * Copyright (c) 2003, 2007-11 Matteo Frigo * Copyright (c) 2003, 2007-11 Massachusetts Institute of Technology * * This program is free software; you can redistribute it and/or modify * it under the terms of the GNU General Public License as published by * the Free Software Foundation; either version 2 of the License, or * (at your option) any later version. * * This program is distributed in the hope that it will be useful, * but WITHOUT ANY WARRANTY; without even the implied warranty of * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the * GNU General Public License for more details. * * You should have received a copy of the GNU General Public License * along with this program; if not, write to the Free Software * Foundation, Inc., 51 Franklin Street, Fifth Floor, Boston, MA 02110-1301 USA * */ /* This file was automatically generated --- DO NOT EDIT */ /* Generated on Sun Nov 25 07:40:13 EST 2012 */ #include "codelet-rdft.h" #ifdef HAVE_FMA /* Generated by: ../../../genfft/gen_r2cf.native -fma -reorder-insns -schedule-for-pipeline -compact -variables 4 -pipeline-latency 4 -n 6 -name r2cfII_6 -dft-II -include r2cfII.h */ /* * This function contains 13 FP additions, 6 FP multiplications, * (or, 7 additions, 0 multiplications, 6 fused multiply/add), * 15 stack variables, 2 constants, and 12 memory accesses */ #include "r2cfII.h" static void r2cfII_6(R *R0, R *R1, R *Cr, R *Ci, stride rs, stride csr, stride csi, INT v, INT ivs, INT ovs) { DK(KP866025403, +0.866025403784438646763723170752936183471402627); DK(KP500000000, +0.500000000000000000000000000000000000000000000); { INT i; for (i = v; i > 0; i = i - 1, R0 = R0 + ivs, R1 = R1 + ivs, Cr = Cr + ovs, Ci = Ci + ovs, MAKE_VOLATILE_STRIDE(24, rs), MAKE_VOLATILE_STRIDE(24, csr), MAKE_VOLATILE_STRIDE(24, csi)) { E T1, T9, T2, T3, T6, T7; T1 = R0[0]; T9 = R1[WS(rs, 1)]; T2 = R0[WS(rs, 2)]; T3 = R0[WS(rs, 1)]; T6 = R1[WS(rs, 2)]; T7 = R1[0]; { E Tc, T4, Ta, T8, T5, Tb; Cr[WS(csr, 1)] = T1 + T2 - T3; Tc = T2 + T3; T4 = T3 - T2; Ta = T6 + T7; T8 = T6 - T7; T5 = FMA(KP500000000, T4, T1); Tb = FMA(KP500000000, Ta, T9); Ci[WS(csi, 1)] = T9 - Ta; Cr[WS(csr, 2)] = FMA(KP866025403, T8, T5); Cr[0] = FNMS(KP866025403, T8, T5); Ci[WS(csi, 2)] = FMS(KP866025403, Tc, Tb); Ci[0] = -(FMA(KP866025403, Tc, Tb)); } } } } static const kr2c_desc desc = { 6, "r2cfII_6", {7, 0, 6, 0}, &GENUS }; void X(codelet_r2cfII_6) (planner *p) { X(kr2c_register) (p, r2cfII_6, &desc); } #else /* HAVE_FMA */ /* Generated by: ../../../genfft/gen_r2cf.native -compact -variables 4 -pipeline-latency 4 -n 6 -name r2cfII_6 -dft-II -include r2cfII.h */ /* * This function contains 13 FP additions, 4 FP multiplications, * (or, 11 additions, 2 multiplications, 2 fused multiply/add), * 14 stack variables, 2 constants, and 12 memory accesses */ #include "r2cfII.h" static void r2cfII_6(R *R0, R *R1, R *Cr, R *Ci, stride rs, stride csr, stride csi, INT v, INT ivs, INT ovs) { DK(KP500000000, +0.500000000000000000000000000000000000000000000); DK(KP866025403, +0.866025403784438646763723170752936183471402627); { INT i; for (i = v; i > 0; i = i - 1, R0 = R0 + ivs, R1 = R1 + ivs, Cr = Cr + ovs, Ci = Ci + ovs, MAKE_VOLATILE_STRIDE(24, rs), MAKE_VOLATILE_STRIDE(24, csr), MAKE_VOLATILE_STRIDE(24, csi)) { E Ta, T7, T9, T1, T3, T2, T8, T4, T5, T6, Tb; Ta = R1[WS(rs, 1)]; T5 = R1[WS(rs, 2)]; T6 = R1[0]; T7 = KP866025403 * (T5 - T6); T9 = T5 + T6; T1 = R0[0]; T3 = R0[WS(rs, 1)]; T2 = R0[WS(rs, 2)]; T8 = KP866025403 * (T2 + T3); T4 = FMA(KP500000000, T3 - T2, T1); Cr[0] = T4 - T7; Cr[WS(csr, 2)] = T4 + T7; Ci[WS(csi, 1)] = Ta - T9; Cr[WS(csr, 1)] = T1 + T2 - T3; Tb = FMA(KP500000000, T9, Ta); Ci[0] = -(T8 + Tb); Ci[WS(csi, 2)] = T8 - Tb; } } } static const kr2c_desc desc = { 6, "r2cfII_6", {11, 2, 2, 0}, &GENUS }; void X(codelet_r2cfII_6) (planner *p) { X(kr2c_register) (p, r2cfII_6, &desc); } #endif /* HAVE_FMA */ fftw-3.3.3/rdft/scalar/r2cf/r2cfII_7.c0000644000175000017500000001314712054410656014130 00000000000000/* * Copyright (c) 2003, 2007-11 Matteo Frigo * Copyright (c) 2003, 2007-11 Massachusetts Institute of Technology * * This program is free software; you can redistribute it and/or modify * it under the terms of the GNU General Public License as published by * the Free Software Foundation; either version 2 of the License, or * (at your option) any later version. * * This program is distributed in the hope that it will be useful, * but WITHOUT ANY WARRANTY; without even the implied warranty of * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the * GNU General Public License for more details. * * You should have received a copy of the GNU General Public License * along with this program; if not, write to the Free Software * Foundation, Inc., 51 Franklin Street, Fifth Floor, Boston, MA 02110-1301 USA * */ /* This file was automatically generated --- DO NOT EDIT */ /* Generated on Sun Nov 25 07:40:13 EST 2012 */ #include "codelet-rdft.h" #ifdef HAVE_FMA /* Generated by: ../../../genfft/gen_r2cf.native -fma -reorder-insns -schedule-for-pipeline -compact -variables 4 -pipeline-latency 4 -n 7 -name r2cfII_7 -dft-II -include r2cfII.h */ /* * This function contains 24 FP additions, 18 FP multiplications, * (or, 9 additions, 3 multiplications, 15 fused multiply/add), * 25 stack variables, 6 constants, and 14 memory accesses */ #include "r2cfII.h" static void r2cfII_7(R *R0, R *R1, R *Cr, R *Ci, stride rs, stride csr, stride csi, INT v, INT ivs, INT ovs) { DK(KP900968867, +0.900968867902419126236102319507445051165919162); DK(KP692021471, +0.692021471630095869627814897002069140197260599); DK(KP801937735, +0.801937735804838252472204639014890102331838324); DK(KP974927912, +0.974927912181823607018131682993931217232785801); DK(KP554958132, +0.554958132087371191422194871006410481067288862); DK(KP356895867, +0.356895867892209443894399510021300583399127187); { INT i; for (i = v; i > 0; i = i - 1, R0 = R0 + ivs, R1 = R1 + ivs, Cr = Cr + ovs, Ci = Ci + ovs, MAKE_VOLATILE_STRIDE(28, rs), MAKE_VOLATILE_STRIDE(28, csr), MAKE_VOLATILE_STRIDE(28, csi)) { E Td, Tk; { E T4, T3, Te, T5, T9, Tf, T6, Tg, Tj; Td = R0[0]; { E T1, T2, T7, T8; T1 = R0[WS(rs, 1)]; T2 = R1[WS(rs, 2)]; T7 = R1[WS(rs, 1)]; T8 = R0[WS(rs, 2)]; T4 = R1[0]; T3 = T1 + T2; Te = T1 - T2; T5 = R0[WS(rs, 3)]; T9 = T7 + T8; Tf = T8 - T7; } T6 = T4 + T5; Tg = T5 - T4; Tj = FNMS(KP356895867, Tf, Te); { E Ta, Th, Tl, Tb, Ti, Tm, Tc; Tb = FNMS(KP554958132, T3, T9); Ta = FMA(KP554958132, T9, T6); Th = FNMS(KP356895867, Tg, Tf); Tl = FNMS(KP356895867, Te, Tg); Ci[WS(csi, 1)] = -(KP974927912 * (FNMS(KP801937735, Tb, T6))); Ci[WS(csi, 2)] = KP974927912 * (FNMS(KP801937735, Ta, T3)); Ti = FNMS(KP692021471, Th, Te); Tm = FNMS(KP692021471, Tl, Tf); Cr[WS(csr, 3)] = Te + Tg + Tf + Td; Tc = FMA(KP554958132, T6, T3); Cr[WS(csr, 1)] = FNMS(KP900968867, Ti, Td); Cr[WS(csr, 2)] = FNMS(KP900968867, Tm, Td); Tk = FNMS(KP692021471, Tj, Tg); Ci[0] = -(KP974927912 * (FMA(KP801937735, Tc, T9))); } } Cr[0] = FNMS(KP900968867, Tk, Td); } } } static const kr2c_desc desc = { 7, "r2cfII_7", {9, 3, 15, 0}, &GENUS }; void X(codelet_r2cfII_7) (planner *p) { X(kr2c_register) (p, r2cfII_7, &desc); } #else /* HAVE_FMA */ /* Generated by: ../../../genfft/gen_r2cf.native -compact -variables 4 -pipeline-latency 4 -n 7 -name r2cfII_7 -dft-II -include r2cfII.h */ /* * This function contains 24 FP additions, 18 FP multiplications, * (or, 12 additions, 6 multiplications, 12 fused multiply/add), * 20 stack variables, 6 constants, and 14 memory accesses */ #include "r2cfII.h" static void r2cfII_7(R *R0, R *R1, R *Cr, R *Ci, stride rs, stride csr, stride csi, INT v, INT ivs, INT ovs) { DK(KP900968867, +0.900968867902419126236102319507445051165919162); DK(KP222520933, +0.222520933956314404288902564496794759466355569); DK(KP623489801, +0.623489801858733530525004884004239810632274731); DK(KP433883739, +0.433883739117558120475768332848358754609990728); DK(KP974927912, +0.974927912181823607018131682993931217232785801); DK(KP781831482, +0.781831482468029808708444526674057750232334519); { INT i; for (i = v; i > 0; i = i - 1, R0 = R0 + ivs, R1 = R1 + ivs, Cr = Cr + ovs, Ci = Ci + ovs, MAKE_VOLATILE_STRIDE(28, rs), MAKE_VOLATILE_STRIDE(28, csr), MAKE_VOLATILE_STRIDE(28, csi)) { E T1, Ta, Td, T4, Tb, T7, Tc, T8, T9; T1 = R0[0]; T8 = R1[0]; T9 = R0[WS(rs, 3)]; Ta = T8 - T9; Td = T8 + T9; { E T2, T3, T5, T6; T2 = R0[WS(rs, 1)]; T3 = R1[WS(rs, 2)]; T4 = T2 - T3; Tb = T2 + T3; T5 = R1[WS(rs, 1)]; T6 = R0[WS(rs, 2)]; T7 = T5 - T6; Tc = T5 + T6; } Ci[0] = -(FMA(KP781831482, Tb, KP974927912 * Tc) + (KP433883739 * Td)); Ci[WS(csi, 1)] = FNMS(KP974927912, Td, KP781831482 * Tc) - (KP433883739 * Tb); Cr[0] = FMA(KP623489801, T4, T1) + FMA(KP222520933, T7, KP900968867 * Ta); Ci[WS(csi, 2)] = FNMS(KP781831482, Td, KP974927912 * Tb) - (KP433883739 * Tc); Cr[WS(csr, 2)] = FMA(KP900968867, T7, T1) + FNMA(KP623489801, Ta, KP222520933 * T4); Cr[WS(csr, 1)] = FMA(KP222520933, Ta, T1) + FNMA(KP623489801, T7, KP900968867 * T4); Cr[WS(csr, 3)] = T1 + T4 - (T7 + Ta); } } } static const kr2c_desc desc = { 7, "r2cfII_7", {12, 6, 12, 0}, &GENUS }; void X(codelet_r2cfII_7) (planner *p) { X(kr2c_register) (p, r2cfII_7, &desc); } #endif /* HAVE_FMA */ fftw-3.3.3/rdft/scalar/r2cf/r2cfII_8.c0000644000175000017500000001246712054410656014135 00000000000000/* * Copyright (c) 2003, 2007-11 Matteo Frigo * Copyright (c) 2003, 2007-11 Massachusetts Institute of Technology * * This program is free software; you can redistribute it and/or modify * it under the terms of the GNU General Public License as published by * the Free Software Foundation; either version 2 of the License, or * (at your option) any later version. * * This program is distributed in the hope that it will be useful, * but WITHOUT ANY WARRANTY; without even the implied warranty of * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the * GNU General Public License for more details. * * You should have received a copy of the GNU General Public License * along with this program; if not, write to the Free Software * Foundation, Inc., 51 Franklin Street, Fifth Floor, Boston, MA 02110-1301 USA * */ /* This file was automatically generated --- DO NOT EDIT */ /* Generated on Sun Nov 25 07:40:14 EST 2012 */ #include "codelet-rdft.h" #ifdef HAVE_FMA /* Generated by: ../../../genfft/gen_r2cf.native -fma -reorder-insns -schedule-for-pipeline -compact -variables 4 -pipeline-latency 4 -n 8 -name r2cfII_8 -dft-II -include r2cfII.h */ /* * This function contains 22 FP additions, 16 FP multiplications, * (or, 6 additions, 0 multiplications, 16 fused multiply/add), * 22 stack variables, 3 constants, and 16 memory accesses */ #include "r2cfII.h" static void r2cfII_8(R *R0, R *R1, R *Cr, R *Ci, stride rs, stride csr, stride csi, INT v, INT ivs, INT ovs) { DK(KP923879532, +0.923879532511286756128183189396788286822416626); DK(KP707106781, +0.707106781186547524400844362104849039284835938); DK(KP414213562, +0.414213562373095048801688724209698078569671875); { INT i; for (i = v; i > 0; i = i - 1, R0 = R0 + ivs, R1 = R1 + ivs, Cr = Cr + ovs, Ci = Ci + ovs, MAKE_VOLATILE_STRIDE(32, rs), MAKE_VOLATILE_STRIDE(32, csr), MAKE_VOLATILE_STRIDE(32, csi)) { E Te, T8, Td, T5, Tj, Tl, Tf, Tb; { E T1, Th, T9, Ti, T4, Ta; T1 = R0[0]; Th = R0[WS(rs, 2)]; { E T2, T3, T6, T7; T2 = R0[WS(rs, 1)]; T3 = R0[WS(rs, 3)]; T6 = R1[0]; T7 = R1[WS(rs, 2)]; T9 = R1[WS(rs, 3)]; Ti = T2 + T3; T4 = T2 - T3; Te = FMA(KP414213562, T6, T7); T8 = FNMS(KP414213562, T7, T6); Ta = R1[WS(rs, 1)]; } Td = FNMS(KP707106781, T4, T1); T5 = FMA(KP707106781, T4, T1); Tj = FMA(KP707106781, Ti, Th); Tl = FNMS(KP707106781, Ti, Th); Tf = FMA(KP414213562, T9, Ta); Tb = FMS(KP414213562, Ta, T9); } { E Tk, Tg, Tc, Tm; Tk = Te + Tf; Tg = Te - Tf; Tc = T8 + Tb; Tm = Tb - T8; Cr[WS(csr, 1)] = FMA(KP923879532, Tg, Td); Cr[WS(csr, 2)] = FNMS(KP923879532, Tg, Td); Ci[WS(csi, 3)] = FNMS(KP923879532, Tk, Tj); Ci[0] = -(FMA(KP923879532, Tk, Tj)); Ci[WS(csi, 1)] = FMA(KP923879532, Tm, Tl); Ci[WS(csi, 2)] = FMS(KP923879532, Tm, Tl); Cr[0] = FMA(KP923879532, Tc, T5); Cr[WS(csr, 3)] = FNMS(KP923879532, Tc, T5); } } } } static const kr2c_desc desc = { 8, "r2cfII_8", {6, 0, 16, 0}, &GENUS }; void X(codelet_r2cfII_8) (planner *p) { X(kr2c_register) (p, r2cfII_8, &desc); } #else /* HAVE_FMA */ /* Generated by: ../../../genfft/gen_r2cf.native -compact -variables 4 -pipeline-latency 4 -n 8 -name r2cfII_8 -dft-II -include r2cfII.h */ /* * This function contains 22 FP additions, 10 FP multiplications, * (or, 18 additions, 6 multiplications, 4 fused multiply/add), * 18 stack variables, 3 constants, and 16 memory accesses */ #include "r2cfII.h" static void r2cfII_8(R *R0, R *R1, R *Cr, R *Ci, stride rs, stride csr, stride csi, INT v, INT ivs, INT ovs) { DK(KP382683432, +0.382683432365089771728459984030398866761344562); DK(KP923879532, +0.923879532511286756128183189396788286822416626); DK(KP707106781, +0.707106781186547524400844362104849039284835938); { INT i; for (i = v; i > 0; i = i - 1, R0 = R0 + ivs, R1 = R1 + ivs, Cr = Cr + ovs, Ci = Ci + ovs, MAKE_VOLATILE_STRIDE(32, rs), MAKE_VOLATILE_STRIDE(32, csr), MAKE_VOLATILE_STRIDE(32, csi)) { E T1, Tj, T4, Ti, T8, Te, Tb, Tf, T2, T3; T1 = R0[0]; Tj = R0[WS(rs, 2)]; T2 = R0[WS(rs, 1)]; T3 = R0[WS(rs, 3)]; T4 = KP707106781 * (T2 - T3); Ti = KP707106781 * (T2 + T3); { E T6, T7, T9, Ta; T6 = R1[0]; T7 = R1[WS(rs, 2)]; T8 = FNMS(KP382683432, T7, KP923879532 * T6); Te = FMA(KP382683432, T6, KP923879532 * T7); T9 = R1[WS(rs, 1)]; Ta = R1[WS(rs, 3)]; Tb = FNMS(KP923879532, Ta, KP382683432 * T9); Tf = FMA(KP923879532, T9, KP382683432 * Ta); } { E T5, Tc, Th, Tk; T5 = T1 + T4; Tc = T8 + Tb; Cr[WS(csr, 3)] = T5 - Tc; Cr[0] = T5 + Tc; Th = Te + Tf; Tk = Ti + Tj; Ci[0] = -(Th + Tk); Ci[WS(csi, 3)] = Tk - Th; } { E Td, Tg, Tl, Tm; Td = T1 - T4; Tg = Te - Tf; Cr[WS(csr, 2)] = Td - Tg; Cr[WS(csr, 1)] = Td + Tg; Tl = Tb - T8; Tm = Tj - Ti; Ci[WS(csi, 2)] = Tl - Tm; Ci[WS(csi, 1)] = Tl + Tm; } } } } static const kr2c_desc desc = { 8, "r2cfII_8", {18, 6, 4, 0}, &GENUS }; void X(codelet_r2cfII_8) (planner *p) { X(kr2c_register) (p, r2cfII_8, &desc); } #endif /* HAVE_FMA */ fftw-3.3.3/rdft/scalar/r2cf/r2cfII_9.c0000644000175000017500000002120212054410657014122 00000000000000/* * Copyright (c) 2003, 2007-11 Matteo Frigo * Copyright (c) 2003, 2007-11 Massachusetts Institute of Technology * * This program is free software; you can redistribute it and/or modify * it under the terms of the GNU General Public License as published by * the Free Software Foundation; either version 2 of the License, or * (at your option) any later version. * * This program is distributed in the hope that it will be useful, * but WITHOUT ANY WARRANTY; without even the implied warranty of * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the * GNU General Public License for more details. * * You should have received a copy of the GNU General Public License * along with this program; if not, write to the Free Software * Foundation, Inc., 51 Franklin Street, Fifth Floor, Boston, MA 02110-1301 USA * */ /* This file was automatically generated --- DO NOT EDIT */ /* Generated on Sun Nov 25 07:40:14 EST 2012 */ #include "codelet-rdft.h" #ifdef HAVE_FMA /* Generated by: ../../../genfft/gen_r2cf.native -fma -reorder-insns -schedule-for-pipeline -compact -variables 4 -pipeline-latency 4 -n 9 -name r2cfII_9 -dft-II -include r2cfII.h */ /* * This function contains 42 FP additions, 34 FP multiplications, * (or, 12 additions, 4 multiplications, 30 fused multiply/add), * 46 stack variables, 17 constants, and 18 memory accesses */ #include "r2cfII.h" static void r2cfII_9(R *R0, R *R1, R *Cr, R *Ci, stride rs, stride csr, stride csi, INT v, INT ivs, INT ovs) { DK(KP939692620, +0.939692620785908384054109277324731469936208134); DK(KP879385241, +0.879385241571816768108218554649462939872416269); DK(KP984807753, +0.984807753012208059366743024589523013670643252); DK(KP852868531, +0.852868531952443209628250963940074071936020296); DK(KP666666666, +0.666666666666666666666666666666666666666666667); DK(KP673648177, +0.673648177666930348851716626769314796000375677); DK(KP898197570, +0.898197570222573798468955502359086394667167570); DK(KP826351822, +0.826351822333069651148283373230685203999624323); DK(KP907603734, +0.907603734547952313649323976213898122064543220); DK(KP866025403, +0.866025403784438646763723170752936183471402627); DK(KP420276625, +0.420276625461206169731530603237061658838781920); DK(KP315207469, +0.315207469095904627298647952427796244129086440); DK(KP203604859, +0.203604859554852403062088995281827210665664861); DK(KP152703644, +0.152703644666139302296566746461370407999248646); DK(KP726681596, +0.726681596905677465811651808188092531873167623); DK(KP968908795, +0.968908795874236621082202410917456709164223497); DK(KP500000000, +0.500000000000000000000000000000000000000000000); { INT i; for (i = v; i > 0; i = i - 1, R0 = R0 + ivs, R1 = R1 + ivs, Cr = Cr + ovs, Ci = Ci + ovs, MAKE_VOLATILE_STRIDE(36, rs), MAKE_VOLATILE_STRIDE(36, csr), MAKE_VOLATILE_STRIDE(36, csi)) { E To, T5, Tp, Ta, Ti, Tm, TB, Tq, Tt, Tf, Th; { E T1, T6, T4, Tb, Tk, T9, Tc, Td, Tl, Te; { E T2, T3, T7, T8; T1 = R0[0]; T2 = R0[WS(rs, 3)]; T3 = R1[WS(rs, 1)]; T6 = R0[WS(rs, 1)]; T7 = R0[WS(rs, 4)]; T8 = R1[WS(rs, 2)]; T4 = T2 - T3; To = T2 + T3; Tb = R0[WS(rs, 2)]; Tk = T7 + T8; T9 = T7 - T8; Tc = R1[0]; Td = R1[WS(rs, 3)]; } T5 = T1 + T4; Tp = FNMS(KP500000000, T4, T1); Ta = T6 + T9; Tl = FNMS(KP500000000, T9, T6); Te = Tc + Td; Ti = Tc - Td; Tm = FMA(KP968908795, Tl, Tk); TB = FNMS(KP726681596, Tk, Tl); Tq = FNMS(KP152703644, Tk, Tl); Tt = FMA(KP203604859, Tl, Tk); Tf = Tb - Te; Th = FMA(KP500000000, Te, Tb); } { E Ts, Tr, TA, Tj, Tg; Ts = FMA(KP315207469, Ti, Th); Tr = FNMS(KP420276625, Th, Ti); TA = FMA(KP203604859, Th, Ti); Tj = FNMS(KP152703644, Ti, Th); Tg = Ta + Tf; Ci[WS(csi, 1)] = KP866025403 * (Tf - Ta); { E Tu, Tx, TF, TC; Tu = FNMS(KP907603734, Tt, Ts); Tx = FNMS(KP826351822, Tr, Tq); TF = FMA(KP898197570, TB, TA); TC = FNMS(KP898197570, TB, TA); { E TE, Tn, Tv, Ty; TE = FNMS(KP673648177, Tm, Tj); Tn = FMA(KP673648177, Tm, Tj); Cr[WS(csr, 4)] = T5 + Tg; Cr[WS(csr, 1)] = FNMS(KP500000000, Tg, T5); Tv = FNMS(KP666666666, Tu, Tr); Ty = FNMS(KP666666666, Tx, Tt); Cr[0] = FMA(KP852868531, TF, Tp); { E TG, TD, Tw, Tz; TG = FMA(KP500000000, TF, TE); Ci[0] = -(KP984807753 * (FMA(KP879385241, To, Tn))); TD = FNMS(KP666666666, Tn, TC); Tw = FMA(KP826351822, Tv, Tq); Tz = FMA(KP907603734, Ty, Ts); Cr[WS(csr, 3)] = FNMS(KP852868531, TG, Tp); Ci[WS(csi, 3)] = -(KP866025403 * (FMA(KP852868531, TD, To))); Cr[WS(csr, 2)] = FNMS(KP852868531, Tw, Tp); Ci[WS(csi, 2)] = KP866025403 * (FNMS(KP939692620, Tz, To)); } } } } } } } static const kr2c_desc desc = { 9, "r2cfII_9", {12, 4, 30, 0}, &GENUS }; void X(codelet_r2cfII_9) (planner *p) { X(kr2c_register) (p, r2cfII_9, &desc); } #else /* HAVE_FMA */ /* Generated by: ../../../genfft/gen_r2cf.native -compact -variables 4 -pipeline-latency 4 -n 9 -name r2cfII_9 -dft-II -include r2cfII.h */ /* * This function contains 42 FP additions, 30 FP multiplications, * (or, 25 additions, 13 multiplications, 17 fused multiply/add), * 39 stack variables, 14 constants, and 18 memory accesses */ #include "r2cfII.h" static void r2cfII_9(R *R0, R *R1, R *Cr, R *Ci, stride rs, stride csr, stride csi, INT v, INT ivs, INT ovs) { DK(KP663413948, +0.663413948168938396205421319635891297216863310); DK(KP642787609, +0.642787609686539326322643409907263432907559884); DK(KP556670399, +0.556670399226419366452912952047023132968291906); DK(KP766044443, +0.766044443118978035202392650555416673935832457); DK(KP852868531, +0.852868531952443209628250963940074071936020296); DK(KP173648177, +0.173648177666930348851716626769314796000375677); DK(KP984807753, +0.984807753012208059366743024589523013670643252); DK(KP150383733, +0.150383733180435296639271897612501926072238258); DK(KP813797681, +0.813797681349373692844693217248393223289101568); DK(KP342020143, +0.342020143325668733044099614682259580763083368); DK(KP939692620, +0.939692620785908384054109277324731469936208134); DK(KP296198132, +0.296198132726023843175338011893050938967728390); DK(KP866025403, +0.866025403784438646763723170752936183471402627); DK(KP500000000, +0.500000000000000000000000000000000000000000000); { INT i; for (i = v; i > 0; i = i - 1, R0 = R0 + ivs, R1 = R1 + ivs, Cr = Cr + ovs, Ci = Ci + ovs, MAKE_VOLATILE_STRIDE(36, rs), MAKE_VOLATILE_STRIDE(36, csr), MAKE_VOLATILE_STRIDE(36, csi)) { E T1, T4, To, Ta, Tl, Tk, Tf, Ti, Th, T2, T3, T5, Tg; T1 = R0[0]; T2 = R1[WS(rs, 1)]; T3 = R0[WS(rs, 3)]; T4 = T2 - T3; To = T2 + T3; { E T6, T7, T8, T9; T6 = R0[WS(rs, 1)]; T7 = R1[WS(rs, 2)]; T8 = R0[WS(rs, 4)]; T9 = T7 - T8; Ta = T6 - T9; Tl = T7 + T8; Tk = FMA(KP500000000, T9, T6); } { E Tb, Tc, Td, Te; Tb = R0[WS(rs, 2)]; Tc = R1[0]; Td = R1[WS(rs, 3)]; Te = Tc + Td; Tf = Tb - Te; Ti = FMA(KP500000000, Te, Tb); Th = Tc - Td; } Ci[WS(csi, 1)] = KP866025403 * (Tf - Ta); T5 = T1 - T4; Tg = Ta + Tf; Cr[WS(csr, 1)] = FNMS(KP500000000, Tg, T5); Cr[WS(csr, 4)] = T5 + Tg; { E Tr, Tt, Tw, Tv, Tu, Tp, Tq, Ts, Tj, Tm, Tn; Tr = FMA(KP500000000, T4, T1); Tt = FMA(KP296198132, Th, KP939692620 * Ti); Tw = FNMS(KP813797681, Th, KP342020143 * Ti); Tv = FNMS(KP984807753, Tk, KP150383733 * Tl); Tu = FMA(KP173648177, Tk, KP852868531 * Tl); Tp = FNMS(KP556670399, Tl, KP766044443 * Tk); Tq = FMA(KP852868531, Th, KP173648177 * Ti); Ts = Tp + Tq; Tj = FNMS(KP984807753, Ti, KP150383733 * Th); Tm = FMA(KP642787609, Tk, KP663413948 * Tl); Tn = Tj - Tm; Ci[0] = FNMS(KP866025403, To, Tn); Cr[0] = Tr + Ts; Ci[WS(csi, 3)] = FNMS(KP500000000, Tn, KP866025403 * ((Tp - Tq) - To)); Cr[WS(csr, 3)] = FMA(KP866025403, Tm + Tj, Tr) - (KP500000000 * Ts); Ci[WS(csi, 2)] = FMA(KP866025403, To - (Tu + Tt), KP500000000 * (Tw - Tv)); Cr[WS(csr, 2)] = FMA(KP500000000, Tt - Tu, Tr) + (KP866025403 * (Tv + Tw)); } } } } static const kr2c_desc desc = { 9, "r2cfII_9", {25, 13, 17, 0}, &GENUS }; void X(codelet_r2cfII_9) (planner *p) { X(kr2c_register) (p, r2cfII_9, &desc); } #endif /* HAVE_FMA */ fftw-3.3.3/rdft/scalar/r2cf/r2cfII_10.c0000644000175000017500000001436312054410657014204 00000000000000/* * Copyright (c) 2003, 2007-11 Matteo Frigo * Copyright (c) 2003, 2007-11 Massachusetts Institute of Technology * * This program is free software; you can redistribute it and/or modify * it under the terms of the GNU General Public License as published by * the Free Software Foundation; either version 2 of the License, or * (at your option) any later version. * * This program is distributed in the hope that it will be useful, * but WITHOUT ANY WARRANTY; without even the implied warranty of * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the * GNU General Public License for more details. * * You should have received a copy of the GNU General Public License * along with this program; if not, write to the Free Software * Foundation, Inc., 51 Franklin Street, Fifth Floor, Boston, MA 02110-1301 USA * */ /* This file was automatically generated --- DO NOT EDIT */ /* Generated on Sun Nov 25 07:40:15 EST 2012 */ #include "codelet-rdft.h" #ifdef HAVE_FMA /* Generated by: ../../../genfft/gen_r2cf.native -fma -reorder-insns -schedule-for-pipeline -compact -variables 4 -pipeline-latency 4 -n 10 -name r2cfII_10 -dft-II -include r2cfII.h */ /* * This function contains 32 FP additions, 18 FP multiplications, * (or, 14 additions, 0 multiplications, 18 fused multiply/add), * 37 stack variables, 4 constants, and 20 memory accesses */ #include "r2cfII.h" static void r2cfII_10(R *R0, R *R1, R *Cr, R *Ci, stride rs, stride csr, stride csi, INT v, INT ivs, INT ovs) { DK(KP951056516, +0.951056516295153572116439333379382143405698634); DK(KP559016994, +0.559016994374947424102293417182819058860154590); DK(KP250000000, +0.250000000000000000000000000000000000000000000); DK(KP618033988, +0.618033988749894848204586834365638117720309180); { INT i; for (i = v; i > 0; i = i - 1, R0 = R0 + ivs, R1 = R1 + ivs, Cr = Cr + ovs, Ci = Ci + ovs, MAKE_VOLATILE_STRIDE(40, rs), MAKE_VOLATILE_STRIDE(40, csr), MAKE_VOLATILE_STRIDE(40, csi)) { E Tq, Ti, Tk, Tu, Tw, Tp, Tb, Tj, Tr, Tv; { E T1, To, Ts, Tt, T8, Ta, Te, Tm, Tl, Th, Tn, T9; T1 = R0[0]; To = R1[WS(rs, 2)]; { E T2, T3, T5, T6; T2 = R0[WS(rs, 2)]; T3 = R0[WS(rs, 3)]; T5 = R0[WS(rs, 4)]; T6 = R0[WS(rs, 1)]; { E Tc, T4, T7, Td, Tf, Tg; Tc = R1[0]; Ts = T2 + T3; T4 = T2 - T3; Tt = T5 + T6; T7 = T5 - T6; Td = R1[WS(rs, 4)]; Tf = R1[WS(rs, 1)]; Tg = R1[WS(rs, 3)]; T8 = T4 + T7; Ta = T4 - T7; Te = Tc - Td; Tm = Tc + Td; Tl = Tf + Tg; Th = Tf - Tg; } } Cr[WS(csr, 2)] = T1 + T8; Tn = Tl - Tm; Tq = Tm + Tl; Ti = FMA(KP618033988, Th, Te); Tk = FNMS(KP618033988, Te, Th); Ci[WS(csi, 2)] = Tn - To; T9 = FNMS(KP250000000, T8, T1); Tu = FMA(KP618033988, Tt, Ts); Tw = FNMS(KP618033988, Ts, Tt); Tp = FMA(KP250000000, Tn, To); Tb = FMA(KP559016994, Ta, T9); Tj = FNMS(KP559016994, Ta, T9); } Tr = FMA(KP559016994, Tq, Tp); Tv = FNMS(KP559016994, Tq, Tp); Cr[WS(csr, 1)] = FNMS(KP951056516, Tk, Tj); Cr[WS(csr, 3)] = FMA(KP951056516, Tk, Tj); Cr[0] = FMA(KP951056516, Ti, Tb); Cr[WS(csr, 4)] = FNMS(KP951056516, Ti, Tb); Ci[WS(csi, 1)] = FNMS(KP951056516, Tw, Tv); Ci[WS(csi, 3)] = FMA(KP951056516, Tw, Tv); Ci[WS(csi, 4)] = FMS(KP951056516, Tu, Tr); Ci[0] = -(FMA(KP951056516, Tu, Tr)); } } } static const kr2c_desc desc = { 10, "r2cfII_10", {14, 0, 18, 0}, &GENUS }; void X(codelet_r2cfII_10) (planner *p) { X(kr2c_register) (p, r2cfII_10, &desc); } #else /* HAVE_FMA */ /* Generated by: ../../../genfft/gen_r2cf.native -compact -variables 4 -pipeline-latency 4 -n 10 -name r2cfII_10 -dft-II -include r2cfII.h */ /* * This function contains 32 FP additions, 12 FP multiplications, * (or, 26 additions, 6 multiplications, 6 fused multiply/add), * 21 stack variables, 4 constants, and 20 memory accesses */ #include "r2cfII.h" static void r2cfII_10(R *R0, R *R1, R *Cr, R *Ci, stride rs, stride csr, stride csi, INT v, INT ivs, INT ovs) { DK(KP250000000, +0.250000000000000000000000000000000000000000000); DK(KP587785252, +0.587785252292473129168705954639072768597652438); DK(KP951056516, +0.951056516295153572116439333379382143405698634); DK(KP559016994, +0.559016994374947424102293417182819058860154590); { INT i; for (i = v; i > 0; i = i - 1, R0 = R0 + ivs, R1 = R1 + ivs, Cr = Cr + ovs, Ci = Ci + ovs, MAKE_VOLATILE_STRIDE(40, rs), MAKE_VOLATILE_STRIDE(40, csr), MAKE_VOLATILE_STRIDE(40, csi)) { E T1, To, T8, Tq, T9, Tp, Te, Ts, Th, Tn; T1 = R0[0]; To = R1[WS(rs, 2)]; { E T2, T3, T4, T5, T6, T7; T2 = R0[WS(rs, 2)]; T3 = R0[WS(rs, 3)]; T4 = T2 - T3; T5 = R0[WS(rs, 4)]; T6 = R0[WS(rs, 1)]; T7 = T5 - T6; T8 = T4 + T7; Tq = T5 + T6; T9 = KP559016994 * (T4 - T7); Tp = T2 + T3; } { E Tc, Td, Tm, Tf, Tg, Tl; Tc = R1[0]; Td = R1[WS(rs, 4)]; Tm = Tc + Td; Tf = R1[WS(rs, 1)]; Tg = R1[WS(rs, 3)]; Tl = Tf + Tg; Te = Tc - Td; Ts = KP559016994 * (Tm + Tl); Th = Tf - Tg; Tn = Tl - Tm; } Cr[WS(csr, 2)] = T1 + T8; Ci[WS(csi, 2)] = Tn - To; { E Ti, Tk, Tb, Tj, Ta; Ti = FMA(KP951056516, Te, KP587785252 * Th); Tk = FNMS(KP587785252, Te, KP951056516 * Th); Ta = FNMS(KP250000000, T8, T1); Tb = T9 + Ta; Tj = Ta - T9; Cr[WS(csr, 4)] = Tb - Ti; Cr[WS(csr, 3)] = Tj + Tk; Cr[0] = Tb + Ti; Cr[WS(csr, 1)] = Tj - Tk; } { E Tr, Tw, Tu, Tv, Tt; Tr = FMA(KP951056516, Tp, KP587785252 * Tq); Tw = FNMS(KP587785252, Tp, KP951056516 * Tq); Tt = FMA(KP250000000, Tn, To); Tu = Ts + Tt; Tv = Tt - Ts; Ci[0] = -(Tr + Tu); Ci[WS(csi, 3)] = Tw + Tv; Ci[WS(csi, 4)] = Tr - Tu; Ci[WS(csi, 1)] = Tv - Tw; } } } } static const kr2c_desc desc = { 10, "r2cfII_10", {26, 6, 6, 0}, &GENUS }; void X(codelet_r2cfII_10) (planner *p) { X(kr2c_register) (p, r2cfII_10, &desc); } #endif /* HAVE_FMA */ fftw-3.3.3/rdft/scalar/r2cf/r2cfII_12.c0000644000175000017500000001574512054410657014213 00000000000000/* * Copyright (c) 2003, 2007-11 Matteo Frigo * Copyright (c) 2003, 2007-11 Massachusetts Institute of Technology * * This program is free software; you can redistribute it and/or modify * it under the terms of the GNU General Public License as published by * the Free Software Foundation; either version 2 of the License, or * (at your option) any later version. * * This program is distributed in the hope that it will be useful, * but WITHOUT ANY WARRANTY; without even the implied warranty of * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the * GNU General Public License for more details. * * You should have received a copy of the GNU General Public License * along with this program; if not, write to the Free Software * Foundation, Inc., 51 Franklin Street, Fifth Floor, Boston, MA 02110-1301 USA * */ /* This file was automatically generated --- DO NOT EDIT */ /* Generated on Sun Nov 25 07:40:15 EST 2012 */ #include "codelet-rdft.h" #ifdef HAVE_FMA /* Generated by: ../../../genfft/gen_r2cf.native -fma -reorder-insns -schedule-for-pipeline -compact -variables 4 -pipeline-latency 4 -n 12 -name r2cfII_12 -dft-II -include r2cfII.h */ /* * This function contains 45 FP additions, 24 FP multiplications, * (or, 21 additions, 0 multiplications, 24 fused multiply/add), * 37 stack variables, 3 constants, and 24 memory accesses */ #include "r2cfII.h" static void r2cfII_12(R *R0, R *R1, R *Cr, R *Ci, stride rs, stride csr, stride csi, INT v, INT ivs, INT ovs) { DK(KP707106781, +0.707106781186547524400844362104849039284835938); DK(KP866025403, +0.866025403784438646763723170752936183471402627); DK(KP500000000, +0.500000000000000000000000000000000000000000000); { INT i; for (i = v; i > 0; i = i - 1, R0 = R0 + ivs, R1 = R1 + ivs, Cr = Cr + ovs, Ci = Ci + ovs, MAKE_VOLATILE_STRIDE(48, rs), MAKE_VOLATILE_STRIDE(48, csr), MAKE_VOLATILE_STRIDE(48, csi)) { E TD, TB, Tp, T9, Tq, Tr, TE, To, Ts, TC; { E T8, T1, Tv, Tm, TF, Tz, Tl, Ta, Tb, Tt, TA, T4, Tc; { E Tx, Th, Ti, Tj, Ty, T6, T7, T2, T3, Tk; Tx = R0[WS(rs, 3)]; T6 = R0[WS(rs, 5)]; T7 = R0[WS(rs, 1)]; Th = R1[WS(rs, 4)]; Ti = R1[WS(rs, 2)]; Tj = R1[0]; Ty = T6 + T7; T8 = T6 - T7; T1 = R0[0]; Tv = Ti - Tj - Th; Tk = Ti - Tj; Tm = Ti + Tj; TF = Tx - Ty; Tz = FMA(KP500000000, Ty, Tx); T2 = R0[WS(rs, 2)]; T3 = R0[WS(rs, 4)]; Tl = FMA(KP500000000, Tk, Th); Ta = R1[WS(rs, 1)]; Tb = R1[WS(rs, 3)]; Tt = T1 + T3 - T2; TA = T3 + T2; T4 = T2 - T3; Tc = R1[WS(rs, 5)]; } { E Tn, Tg, T5, Tu; TD = FNMS(KP866025403, TA, Tz); TB = FMA(KP866025403, TA, Tz); T5 = FMA(KP500000000, T4, T1); Tu = Ta + Tc - Tb; { E Td, Tf, TG, Tw, Te; Td = Tb - Tc; Tf = Tc + Tb; Tp = FMA(KP866025403, T8, T5); T9 = FNMS(KP866025403, T8, T5); TG = Tv - Tu; Tw = Tu + Tv; Te = FMA(KP500000000, Td, Ta); Tq = FMA(KP866025403, Tm, Tl); Tn = FNMS(KP866025403, Tm, Tl); Ci[WS(csi, 1)] = FMA(KP707106781, TG, TF); Ci[WS(csi, 4)] = FMS(KP707106781, TG, TF); Cr[WS(csr, 4)] = FMA(KP707106781, Tw, Tt); Cr[WS(csr, 1)] = FNMS(KP707106781, Tw, Tt); Tg = FNMS(KP866025403, Tf, Te); Tr = FMA(KP866025403, Tf, Te); } TE = Tg + Tn; To = Tg - Tn; } } Ci[WS(csi, 2)] = FMS(KP707106781, TE, TD); Ci[WS(csi, 3)] = FMA(KP707106781, TE, TD); Cr[0] = FMA(KP707106781, To, T9); Cr[WS(csr, 5)] = FNMS(KP707106781, To, T9); Ts = Tq - Tr; TC = Tr + Tq; Ci[0] = -(FMA(KP707106781, TC, TB)); Ci[WS(csi, 5)] = FNMS(KP707106781, TC, TB); Cr[WS(csr, 2)] = FMA(KP707106781, Ts, Tp); Cr[WS(csr, 3)] = FNMS(KP707106781, Ts, Tp); } } } static const kr2c_desc desc = { 12, "r2cfII_12", {21, 0, 24, 0}, &GENUS }; void X(codelet_r2cfII_12) (planner *p) { X(kr2c_register) (p, r2cfII_12, &desc); } #else /* HAVE_FMA */ /* Generated by: ../../../genfft/gen_r2cf.native -compact -variables 4 -pipeline-latency 4 -n 12 -name r2cfII_12 -dft-II -include r2cfII.h */ /* * This function contains 43 FP additions, 12 FP multiplications, * (or, 39 additions, 8 multiplications, 4 fused multiply/add), * 28 stack variables, 5 constants, and 24 memory accesses */ #include "r2cfII.h" static void r2cfII_12(R *R0, R *R1, R *Cr, R *Ci, stride rs, stride csr, stride csi, INT v, INT ivs, INT ovs) { DK(KP353553390, +0.353553390593273762200422181052424519642417969); DK(KP707106781, +0.707106781186547524400844362104849039284835938); DK(KP612372435, +0.612372435695794524549321018676472847991486870); DK(KP500000000, +0.500000000000000000000000000000000000000000000); DK(KP866025403, +0.866025403784438646763723170752936183471402627); { INT i; for (i = v; i > 0; i = i - 1, R0 = R0 + ivs, R1 = R1 + ivs, Cr = Cr + ovs, Ci = Ci + ovs, MAKE_VOLATILE_STRIDE(48, rs), MAKE_VOLATILE_STRIDE(48, csr), MAKE_VOLATILE_STRIDE(48, csi)) { E Tx, Tg, T4, Tz, Ty, Tj, TA, T9, Tm, Tl, Te, Tp, To, Tf, TE; E TF; { E T1, T3, T2, Th, Ti; T1 = R0[0]; T3 = R0[WS(rs, 2)]; T2 = R0[WS(rs, 4)]; Tx = KP866025403 * (T2 + T3); Tg = FMA(KP500000000, T3 - T2, T1); T4 = T1 + T2 - T3; Tz = R0[WS(rs, 3)]; Th = R0[WS(rs, 5)]; Ti = R0[WS(rs, 1)]; Ty = Th + Ti; Tj = KP866025403 * (Th - Ti); TA = FMA(KP500000000, Ty, Tz); } { E T5, T6, T7, T8; T5 = R1[WS(rs, 1)]; T6 = R1[WS(rs, 5)]; T7 = R1[WS(rs, 3)]; T8 = T6 - T7; T9 = T5 + T8; Tm = KP612372435 * (T6 + T7); Tl = FNMS(KP353553390, T8, KP707106781 * T5); } { E Td, Ta, Tb, Tc; Td = R1[WS(rs, 4)]; Ta = R1[WS(rs, 2)]; Tb = R1[0]; Tc = Ta - Tb; Te = Tc - Td; Tp = FMA(KP353553390, Tc, KP707106781 * Td); To = KP612372435 * (Ta + Tb); } Tf = KP707106781 * (T9 + Te); Cr[WS(csr, 1)] = T4 - Tf; Cr[WS(csr, 4)] = T4 + Tf; TE = KP707106781 * (Te - T9); TF = Tz - Ty; Ci[WS(csi, 4)] = TE - TF; Ci[WS(csi, 1)] = TE + TF; { E Tk, TB, Tr, Tw, Tn, Tq; Tk = Tg - Tj; TB = Tx - TA; Tn = Tl - Tm; Tq = To - Tp; Tr = Tn + Tq; Tw = Tn - Tq; Cr[WS(csr, 5)] = Tk - Tr; Ci[WS(csi, 2)] = Tw + TB; Cr[0] = Tk + Tr; Ci[WS(csi, 3)] = Tw - TB; } { E Ts, TD, Tv, TC, Tt, Tu; Ts = Tg + Tj; TD = Tx + TA; Tt = To + Tp; Tu = Tm + Tl; Tv = Tt - Tu; TC = Tu + Tt; Cr[WS(csr, 3)] = Ts - Tv; Ci[WS(csi, 5)] = TD - TC; Cr[WS(csr, 2)] = Ts + Tv; Ci[0] = -(TC + TD); } } } } static const kr2c_desc desc = { 12, "r2cfII_12", {39, 8, 4, 0}, &GENUS }; void X(codelet_r2cfII_12) (planner *p) { X(kr2c_register) (p, r2cfII_12, &desc); } #endif /* HAVE_FMA */ fftw-3.3.3/rdft/scalar/r2cf/r2cfII_15.c0000644000175000017500000002403012054410660014173 00000000000000/* * Copyright (c) 2003, 2007-11 Matteo Frigo * Copyright (c) 2003, 2007-11 Massachusetts Institute of Technology * * This program is free software; you can redistribute it and/or modify * it under the terms of the GNU General Public License as published by * the Free Software Foundation; either version 2 of the License, or * (at your option) any later version. * * This program is distributed in the hope that it will be useful, * but WITHOUT ANY WARRANTY; without even the implied warranty of * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the * GNU General Public License for more details. * * You should have received a copy of the GNU General Public License * along with this program; if not, write to the Free Software * Foundation, Inc., 51 Franklin Street, Fifth Floor, Boston, MA 02110-1301 USA * */ /* This file was automatically generated --- DO NOT EDIT */ /* Generated on Sun Nov 25 07:40:15 EST 2012 */ #include "codelet-rdft.h" #ifdef HAVE_FMA /* Generated by: ../../../genfft/gen_r2cf.native -fma -reorder-insns -schedule-for-pipeline -compact -variables 4 -pipeline-latency 4 -n 15 -name r2cfII_15 -dft-II -include r2cfII.h */ /* * This function contains 72 FP additions, 41 FP multiplications, * (or, 38 additions, 7 multiplications, 34 fused multiply/add), * 57 stack variables, 12 constants, and 30 memory accesses */ #include "r2cfII.h" static void r2cfII_15(R *R0, R *R1, R *Cr, R *Ci, stride rs, stride csr, stride csi, INT v, INT ivs, INT ovs) { DK(KP823639103, +0.823639103546331925877420039278190003029660514); DK(KP910592997, +0.910592997310029334643087372129977886038870291); DK(KP951056516, +0.951056516295153572116439333379382143405698634); DK(KP866025403, +0.866025403784438646763723170752936183471402627); DK(KP500000000, +0.500000000000000000000000000000000000000000000); DK(KP559016994, +0.559016994374947424102293417182819058860154590); DK(KP690983005, +0.690983005625052575897706582817180941139845410); DK(KP552786404, +0.552786404500042060718165266253744752911876328); DK(KP447213595, +0.447213595499957939281834733746255247088123672); DK(KP809016994, +0.809016994374947424102293417182819058860154590); DK(KP618033988, +0.618033988749894848204586834365638117720309180); DK(KP250000000, +0.250000000000000000000000000000000000000000000); { INT i; for (i = v; i > 0; i = i - 1, R0 = R0 + ivs, R1 = R1 + ivs, Cr = Cr + ovs, Ci = Ci + ovs, MAKE_VOLATILE_STRIDE(60, rs), MAKE_VOLATILE_STRIDE(60, csr), MAKE_VOLATILE_STRIDE(60, csi)) { E T9, TQ, TV, TW, Tw, TJ; { E Ta, Tl, Tg, T8, T7, TF, TX, TT, Tm, Th, TM, TZ, Tr, Tn, Tj; E Tz, To, TN, TH, Tp, TO; Ta = R0[WS(rs, 5)]; Tl = R1[WS(rs, 2)]; { E T1, T2, T5, T3, T4; T1 = R0[0]; T2 = R0[WS(rs, 3)]; T5 = R1[WS(rs, 4)]; T3 = R0[WS(rs, 6)]; T4 = R1[WS(rs, 1)]; { E Tb, TL, Te, TK, TR, Tf, Ti, Ty; Tb = R1[0]; TR = T2 + T5; Tg = R0[WS(rs, 2)]; { E T6, TS, Tc, Td; T6 = T2 + T3 - T4 - T5; T8 = (T3 + T5 - T2) - T4; TS = T3 + T4; Tc = R1[WS(rs, 3)]; Td = R1[WS(rs, 6)]; T7 = FNMS(KP250000000, T6, T1); TF = T1 + T6; TX = FNMS(KP618033988, TR, TS); TT = FMA(KP618033988, TS, TR); TL = Tc - Td; Te = Tc + Td; } TK = Tg + Tb; Tm = R0[WS(rs, 7)]; Tf = Tb - Te; Th = Tb + Te; TM = FMA(KP618033988, TL, TK); TZ = FNMS(KP618033988, TK, TL); Ti = FMA(KP809016994, Th, Tg); Ty = FMA(KP447213595, Th, Tf); Tr = R1[WS(rs, 5)]; Tn = R0[WS(rs, 1)]; Tj = FNMS(KP552786404, Ti, Tf); Tz = FNMS(KP690983005, Ty, Tg); To = R0[WS(rs, 4)]; TN = Tr + Tm; } } TH = Ta + Tg - Th; Tp = Tn + To; TO = To - Tn; { E Tx, TA, TP, T14, T11, Tu, TD; { E T10, TI, TC, TY; T9 = FNMS(KP559016994, T8, T7); Tx = FMA(KP559016994, T8, T7); TA = FNMS(KP809016994, Tz, Ta); TP = FMA(KP618033988, TO, TN); TY = FNMS(KP618033988, TN, TO); { E Tq, Ts, TG, Tt, TB; Tq = Tm - Tp; Ts = Tm + Tp; T14 = TZ - TY; T10 = TY + TZ; TG = Ts - Tr - Tl; Tt = FMA(KP809016994, Ts, Tr); TB = FMA(KP447213595, Ts, Tq); T11 = FMA(KP500000000, T10, TX); Ci[WS(csi, 2)] = KP866025403 * (TH - TG); TI = TG + TH; Tu = FNMS(KP552786404, Tt, Tq); TC = FNMS(KP690983005, TB, Tr); } Ci[WS(csi, 1)] = KP951056516 * (T10 - TX); Cr[WS(csr, 7)] = TF + TI; Cr[WS(csr, 2)] = FNMS(KP500000000, TI, TF); TD = FNMS(KP809016994, TC, Tl); } { E TU, Tk, T13, Tv, T12, TE; TQ = TM - TP; TU = TP + TM; T12 = TD + TA; TE = TA - TD; Tk = FNMS(KP559016994, Tj, Ta); TV = FMA(KP500000000, TU, TT); Ci[WS(csi, 6)] = -(KP951056516 * (FMA(KP910592997, T12, T11))); Ci[WS(csi, 3)] = KP951056516 * (FNMS(KP910592997, T12, T11)); T13 = FNMS(KP500000000, TE, Tx); Cr[WS(csr, 1)] = Tx + TE; Tv = FNMS(KP559016994, Tu, Tl); Ci[WS(csi, 4)] = KP951056516 * (TT - TU); Cr[WS(csr, 6)] = FMA(KP823639103, T14, T13); Cr[WS(csr, 3)] = FNMS(KP823639103, T14, T13); TW = Tv + Tk; Tw = Tk - Tv; } } } Ci[WS(csi, 5)] = -(KP951056516 * (FNMS(KP910592997, TW, TV))); Ci[0] = -(KP951056516 * (FMA(KP910592997, TW, TV))); TJ = FNMS(KP500000000, Tw, T9); Cr[WS(csr, 4)] = T9 + Tw; Cr[0] = FMA(KP823639103, TQ, TJ); Cr[WS(csr, 5)] = FNMS(KP823639103, TQ, TJ); } } } static const kr2c_desc desc = { 15, "r2cfII_15", {38, 7, 34, 0}, &GENUS }; void X(codelet_r2cfII_15) (planner *p) { X(kr2c_register) (p, r2cfII_15, &desc); } #else /* HAVE_FMA */ /* Generated by: ../../../genfft/gen_r2cf.native -compact -variables 4 -pipeline-latency 4 -n 15 -name r2cfII_15 -dft-II -include r2cfII.h */ /* * This function contains 72 FP additions, 33 FP multiplications, * (or, 54 additions, 15 multiplications, 18 fused multiply/add), * 37 stack variables, 8 constants, and 30 memory accesses */ #include "r2cfII.h" static void r2cfII_15(R *R0, R *R1, R *Cr, R *Ci, stride rs, stride csr, stride csi, INT v, INT ivs, INT ovs) { DK(KP500000000, +0.500000000000000000000000000000000000000000000); DK(KP866025403, +0.866025403784438646763723170752936183471402627); DK(KP809016994, +0.809016994374947424102293417182819058860154590); DK(KP309016994, +0.309016994374947424102293417182819058860154590); DK(KP250000000, +0.250000000000000000000000000000000000000000000); DK(KP559016994, +0.559016994374947424102293417182819058860154590); DK(KP587785252, +0.587785252292473129168705954639072768597652438); DK(KP951056516, +0.951056516295153572116439333379382143405698634); { INT i; for (i = v; i > 0; i = i - 1, R0 = R0 + ivs, R1 = R1 + ivs, Cr = Cr + ovs, Ci = Ci + ovs, MAKE_VOLATILE_STRIDE(60, rs), MAKE_VOLATILE_STRIDE(60, csr), MAKE_VOLATILE_STRIDE(60, csi)) { E T1, T2, Tx, TR, TE, T7, TD, Th, Tm, Tr, TQ, TA, TB, Tf, Te; E Tu, TS, Td, TH, TO; T1 = R0[WS(rs, 5)]; { E T3, Tv, T6, Tw, T4, T5; T2 = R0[WS(rs, 2)]; T3 = R1[0]; Tv = T2 + T3; T4 = R1[WS(rs, 3)]; T5 = R1[WS(rs, 6)]; T6 = T4 + T5; Tw = T4 - T5; Tx = FMA(KP951056516, Tv, KP587785252 * Tw); TR = FNMS(KP587785252, Tv, KP951056516 * Tw); TE = KP559016994 * (T3 - T6); T7 = T3 + T6; TD = KP250000000 * T7; } { E Ti, Tl, Tj, Tk, Tp, Tq; Th = R0[0]; Ti = R1[WS(rs, 4)]; Tl = R0[WS(rs, 6)]; Tj = R1[WS(rs, 1)]; Tk = R0[WS(rs, 3)]; Tp = Tk + Ti; Tq = Tl + Tj; Tm = Ti + Tj - (Tk + Tl); Tr = FMA(KP951056516, Tp, KP587785252 * Tq); TQ = FNMS(KP951056516, Tq, KP587785252 * Tp); TA = FMA(KP250000000, Tm, Th); TB = KP559016994 * (Tl + Ti - (Tk + Tj)); } { E T9, Tt, Tc, Ts, Ta, Tb, TG; Tf = R1[WS(rs, 2)]; T9 = R0[WS(rs, 7)]; Te = R1[WS(rs, 5)]; Tt = T9 + Te; Ta = R0[WS(rs, 1)]; Tb = R0[WS(rs, 4)]; Tc = Ta + Tb; Ts = Ta - Tb; Tu = FNMS(KP951056516, Tt, KP587785252 * Ts); TS = FMA(KP951056516, Ts, KP587785252 * Tt); Td = T9 + Tc; TG = KP559016994 * (T9 - Tc); TH = FNMS(KP309016994, Te, TG) + FNMA(KP250000000, Td, Tf); TO = FMS(KP809016994, Te, Tf) + FNMA(KP250000000, Td, TG); } { E Tn, T8, Tg, To; Tn = Th - Tm; T8 = T1 + T2 - T7; Tg = Td - Te - Tf; To = T8 + Tg; Ci[WS(csi, 2)] = KP866025403 * (T8 - Tg); Cr[WS(csr, 2)] = FNMS(KP500000000, To, Tn); Cr[WS(csr, 7)] = Tn + To; } { E TM, TX, TT, TV, TP, TU, TN, TW; TM = TB + TA; TX = KP866025403 * (TR + TS); TT = TR - TS; TV = FMS(KP500000000, TT, TQ); TN = T1 + TE + FNMS(KP809016994, T2, TD); TP = TN + TO; TU = KP866025403 * (TO - TN); Cr[WS(csr, 1)] = TM + TP; Ci[WS(csi, 1)] = TQ + TT; Ci[WS(csi, 6)] = TU - TV; Ci[WS(csi, 3)] = TU + TV; TW = FNMS(KP500000000, TP, TM); Cr[WS(csr, 3)] = TW - TX; Cr[WS(csr, 6)] = TW + TX; } { E Tz, TC, Ty, TK, TI, TL, TF, TJ; Tz = KP866025403 * (Tx + Tu); TC = TA - TB; Ty = Tu - Tx; TK = FMS(KP500000000, Ty, Tr); TF = FMA(KP309016994, T2, T1) + TD - TE; TI = TF + TH; TL = KP866025403 * (TH - TF); Ci[WS(csi, 4)] = Tr + Ty; Cr[WS(csr, 4)] = TC + TI; Ci[WS(csi, 5)] = TK - TL; Ci[0] = TK + TL; TJ = FNMS(KP500000000, TI, TC); Cr[0] = Tz + TJ; Cr[WS(csr, 5)] = TJ - Tz; } } } } static const kr2c_desc desc = { 15, "r2cfII_15", {54, 15, 18, 0}, &GENUS }; void X(codelet_r2cfII_15) (planner *p) { X(kr2c_register) (p, r2cfII_15, &desc); } #endif /* HAVE_FMA */ fftw-3.3.3/rdft/scalar/r2cf/r2cfII_16.c0000644000175000017500000002343012054410661014200 00000000000000/* * Copyright (c) 2003, 2007-11 Matteo Frigo * Copyright (c) 2003, 2007-11 Massachusetts Institute of Technology * * This program is free software; you can redistribute it and/or modify * it under the terms of the GNU General Public License as published by * the Free Software Foundation; either version 2 of the License, or * (at your option) any later version. * * This program is distributed in the hope that it will be useful, * but WITHOUT ANY WARRANTY; without even the implied warranty of * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the * GNU General Public License for more details. * * You should have received a copy of the GNU General Public License * along with this program; if not, write to the Free Software * Foundation, Inc., 51 Franklin Street, Fifth Floor, Boston, MA 02110-1301 USA * */ /* This file was automatically generated --- DO NOT EDIT */ /* Generated on Sun Nov 25 07:40:15 EST 2012 */ #include "codelet-rdft.h" #ifdef HAVE_FMA /* Generated by: ../../../genfft/gen_r2cf.native -fma -reorder-insns -schedule-for-pipeline -compact -variables 4 -pipeline-latency 4 -n 16 -name r2cfII_16 -dft-II -include r2cfII.h */ /* * This function contains 66 FP additions, 48 FP multiplications, * (or, 18 additions, 0 multiplications, 48 fused multiply/add), * 54 stack variables, 7 constants, and 32 memory accesses */ #include "r2cfII.h" static void r2cfII_16(R *R0, R *R1, R *Cr, R *Ci, stride rs, stride csr, stride csi, INT v, INT ivs, INT ovs) { DK(KP980785280, +0.980785280403230449126182236134239036973933731); DK(KP198912367, +0.198912367379658006911597622644676228597850501); DK(KP831469612, +0.831469612302545237078788377617905756738560812); DK(KP923879532, +0.923879532511286756128183189396788286822416626); DK(KP668178637, +0.668178637919298919997757686523080761552472251); DK(KP414213562, +0.414213562373095048801688724209698078569671875); DK(KP707106781, +0.707106781186547524400844362104849039284835938); { INT i; for (i = v; i > 0; i = i - 1, R0 = R0 + ivs, R1 = R1 + ivs, Cr = Cr + ovs, Ci = Ci + ovs, MAKE_VOLATILE_STRIDE(64, rs), MAKE_VOLATILE_STRIDE(64, csr), MAKE_VOLATILE_STRIDE(64, csi)) { E TN, TF, TX, TV, TO, TP, TY, TM, TQ, TW; { E TT, TZ, TB, T5, Tu, TK, TJ, Tr, T9, TC, T8, Tl, TH, TG, Ti; E Ta; { E T1, TR, Tn, Ts, To, TS, T4, Tp, T2, T3; T1 = R0[0]; TR = R0[WS(rs, 4)]; T2 = R0[WS(rs, 2)]; T3 = R0[WS(rs, 6)]; Tn = R1[WS(rs, 7)]; Ts = R1[WS(rs, 3)]; To = R1[WS(rs, 1)]; TS = T2 + T3; T4 = T2 - T3; Tp = R1[WS(rs, 5)]; { E Te, Tj, Tf, Tg, Tt, Tq; Te = R1[0]; TT = FMA(KP707106781, TS, TR); TZ = FNMS(KP707106781, TS, TR); TB = FMA(KP707106781, T4, T1); T5 = FNMS(KP707106781, T4, T1); Tt = To + Tp; Tq = To - Tp; Tj = R1[WS(rs, 4)]; Tf = R1[WS(rs, 2)]; Tu = FNMS(KP707106781, Tt, Ts); TK = FMA(KP707106781, Tt, Ts); TJ = FMS(KP707106781, Tq, Tn); Tr = FMA(KP707106781, Tq, Tn); Tg = R1[WS(rs, 6)]; { E T6, T7, Tk, Th; T6 = R0[WS(rs, 5)]; T7 = R0[WS(rs, 1)]; T9 = R0[WS(rs, 3)]; Tk = Tf + Tg; Th = Tf - Tg; TC = FNMS(KP414213562, T6, T7); T8 = FMA(KP414213562, T7, T6); Tl = FNMS(KP707106781, Tk, Tj); TH = FMA(KP707106781, Tk, Tj); TG = FMA(KP707106781, Th, Te); Ti = FNMS(KP707106781, Th, Te); Ta = R0[WS(rs, 7)]; } } } { E TE, TU, Ty, Tv, TI, TL; Ty = FNMS(KP668178637, Tr, Tu); Tv = FMA(KP668178637, Tu, Tr); { E Tw, T14, T12, TA, T11, T13, Tx, Td; { E Tz, Tm, TD, Tb, T10, Tc; Tz = FNMS(KP668178637, Ti, Tl); Tm = FMA(KP668178637, Tl, Ti); TD = FMS(KP414213562, T9, Ta); Tb = FMA(KP414213562, Ta, T9); Tw = Tm - Tv; T14 = Tm + Tv; T10 = TD - TC; TE = TC + TD; Tc = T8 - Tb; TU = T8 + Tb; T12 = Tz + Ty; TA = Ty - Tz; T11 = FMA(KP923879532, T10, TZ); T13 = FNMS(KP923879532, T10, TZ); Tx = FNMS(KP923879532, Tc, T5); Td = FMA(KP923879532, Tc, T5); } Ci[WS(csi, 2)] = -(FMA(KP831469612, T14, T13)); Ci[WS(csi, 5)] = FNMS(KP831469612, T14, T13); Cr[WS(csr, 1)] = FMA(KP831469612, Tw, Td); Cr[WS(csr, 6)] = FNMS(KP831469612, Tw, Td); Cr[WS(csr, 5)] = FNMS(KP831469612, TA, Tx); Ci[WS(csi, 1)] = FMA(KP831469612, T12, T11); Cr[WS(csr, 2)] = FMA(KP831469612, TA, Tx); Ci[WS(csi, 6)] = FMS(KP831469612, T12, T11); } TN = FNMS(KP923879532, TE, TB); TF = FMA(KP923879532, TE, TB); TX = FNMS(KP923879532, TU, TT); TV = FMA(KP923879532, TU, TT); TO = FMA(KP198912367, TG, TH); TI = FNMS(KP198912367, TH, TG); TL = FMA(KP198912367, TK, TJ); TP = FNMS(KP198912367, TJ, TK); TY = TL - TI; TM = TI + TL; } } Ci[WS(csi, 4)] = FMS(KP980785280, TY, TX); Ci[WS(csi, 3)] = FMA(KP980785280, TY, TX); Cr[0] = FMA(KP980785280, TM, TF); Cr[WS(csr, 7)] = FNMS(KP980785280, TM, TF); TQ = TO - TP; TW = TO + TP; Ci[0] = -(FMA(KP980785280, TW, TV)); Ci[WS(csi, 7)] = FNMS(KP980785280, TW, TV); Cr[WS(csr, 3)] = FMA(KP980785280, TQ, TN); Cr[WS(csr, 4)] = FNMS(KP980785280, TQ, TN); } } } static const kr2c_desc desc = { 16, "r2cfII_16", {18, 0, 48, 0}, &GENUS }; void X(codelet_r2cfII_16) (planner *p) { X(kr2c_register) (p, r2cfII_16, &desc); } #else /* HAVE_FMA */ /* Generated by: ../../../genfft/gen_r2cf.native -compact -variables 4 -pipeline-latency 4 -n 16 -name r2cfII_16 -dft-II -include r2cfII.h */ /* * This function contains 66 FP additions, 30 FP multiplications, * (or, 54 additions, 18 multiplications, 12 fused multiply/add), * 32 stack variables, 7 constants, and 32 memory accesses */ #include "r2cfII.h" static void r2cfII_16(R *R0, R *R1, R *Cr, R *Ci, stride rs, stride csr, stride csi, INT v, INT ivs, INT ovs) { DK(KP555570233, +0.555570233019602224742830813948532874374937191); DK(KP831469612, +0.831469612302545237078788377617905756738560812); DK(KP980785280, +0.980785280403230449126182236134239036973933731); DK(KP195090322, +0.195090322016128267848284868477022240927691618); DK(KP382683432, +0.382683432365089771728459984030398866761344562); DK(KP923879532, +0.923879532511286756128183189396788286822416626); DK(KP707106781, +0.707106781186547524400844362104849039284835938); { INT i; for (i = v; i > 0; i = i - 1, R0 = R0 + ivs, R1 = R1 + ivs, Cr = Cr + ovs, Ci = Ci + ovs, MAKE_VOLATILE_STRIDE(64, rs), MAKE_VOLATILE_STRIDE(64, csr), MAKE_VOLATILE_STRIDE(64, csi)) { E T5, T11, TB, TV, Tr, TK, Tu, TJ, Ti, TH, Tl, TG, Tc, T10, TE; E TS; { E T1, TU, T4, TT, T2, T3; T1 = R0[0]; TU = R0[WS(rs, 4)]; T2 = R0[WS(rs, 2)]; T3 = R0[WS(rs, 6)]; T4 = KP707106781 * (T2 - T3); TT = KP707106781 * (T2 + T3); T5 = T1 + T4; T11 = TU - TT; TB = T1 - T4; TV = TT + TU; } { E Tq, Tt, Tp, Ts, Tn, To; Tq = R1[WS(rs, 7)]; Tt = R1[WS(rs, 3)]; Tn = R1[WS(rs, 1)]; To = R1[WS(rs, 5)]; Tp = KP707106781 * (Tn - To); Ts = KP707106781 * (Tn + To); Tr = Tp - Tq; TK = Tt - Ts; Tu = Ts + Tt; TJ = Tp + Tq; } { E Te, Tk, Th, Tj, Tf, Tg; Te = R1[0]; Tk = R1[WS(rs, 4)]; Tf = R1[WS(rs, 2)]; Tg = R1[WS(rs, 6)]; Th = KP707106781 * (Tf - Tg); Tj = KP707106781 * (Tf + Tg); Ti = Te + Th; TH = Tk - Tj; Tl = Tj + Tk; TG = Te - Th; } { E T8, TC, Tb, TD; { E T6, T7, T9, Ta; T6 = R0[WS(rs, 1)]; T7 = R0[WS(rs, 5)]; T8 = FNMS(KP382683432, T7, KP923879532 * T6); TC = FMA(KP382683432, T6, KP923879532 * T7); T9 = R0[WS(rs, 3)]; Ta = R0[WS(rs, 7)]; Tb = FNMS(KP923879532, Ta, KP382683432 * T9); TD = FMA(KP923879532, T9, KP382683432 * Ta); } Tc = T8 + Tb; T10 = Tb - T8; TE = TC - TD; TS = TC + TD; } { E Td, TW, Tw, TR, Tm, Tv; Td = T5 - Tc; TW = TS + TV; Tm = FMA(KP195090322, Ti, KP980785280 * Tl); Tv = FNMS(KP980785280, Tu, KP195090322 * Tr); Tw = Tm + Tv; TR = Tv - Tm; Cr[WS(csr, 4)] = Td - Tw; Ci[WS(csi, 7)] = TR + TW; Cr[WS(csr, 3)] = Td + Tw; Ci[0] = TR - TW; } { E Tx, TY, TA, TX, Ty, Tz; Tx = T5 + Tc; TY = TV - TS; Ty = FNMS(KP195090322, Tl, KP980785280 * Ti); Tz = FMA(KP980785280, Tr, KP195090322 * Tu); TA = Ty + Tz; TX = Tz - Ty; Cr[WS(csr, 7)] = Tx - TA; Ci[WS(csi, 3)] = TX + TY; Cr[0] = Tx + TA; Ci[WS(csi, 4)] = TX - TY; } { E TF, T12, TM, TZ, TI, TL; TF = TB + TE; T12 = T10 - T11; TI = FMA(KP831469612, TG, KP555570233 * TH); TL = FMA(KP831469612, TJ, KP555570233 * TK); TM = TI - TL; TZ = TI + TL; Cr[WS(csr, 6)] = TF - TM; Ci[WS(csi, 2)] = T12 - TZ; Cr[WS(csr, 1)] = TF + TM; Ci[WS(csi, 5)] = -(TZ + T12); } { E TN, T14, TQ, T13, TO, TP; TN = TB - TE; T14 = T10 + T11; TO = FNMS(KP555570233, TJ, KP831469612 * TK); TP = FNMS(KP555570233, TG, KP831469612 * TH); TQ = TO - TP; T13 = TP + TO; Cr[WS(csr, 5)] = TN - TQ; Ci[WS(csi, 1)] = T13 + T14; Cr[WS(csr, 2)] = TN + TQ; Ci[WS(csi, 6)] = T13 - T14; } } } } static const kr2c_desc desc = { 16, "r2cfII_16", {54, 18, 12, 0}, &GENUS }; void X(codelet_r2cfII_16) (planner *p) { X(kr2c_register) (p, r2cfII_16, &desc); } #endif /* HAVE_FMA */ fftw-3.3.3/rdft/scalar/r2cf/r2cfII_32.c0000644000175000017500000005145712054410667014216 00000000000000/* * Copyright (c) 2003, 2007-11 Matteo Frigo * Copyright (c) 2003, 2007-11 Massachusetts Institute of Technology * * This program is free software; you can redistribute it and/or modify * it under the terms of the GNU General Public License as published by * the Free Software Foundation; either version 2 of the License, or * (at your option) any later version. * * This program is distributed in the hope that it will be useful, * but WITHOUT ANY WARRANTY; without even the implied warranty of * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the * GNU General Public License for more details. * * You should have received a copy of the GNU General Public License * along with this program; if not, write to the Free Software * Foundation, Inc., 51 Franklin Street, Fifth Floor, Boston, MA 02110-1301 USA * */ /* This file was automatically generated --- DO NOT EDIT */ /* Generated on Sun Nov 25 07:40:16 EST 2012 */ #include "codelet-rdft.h" #ifdef HAVE_FMA /* Generated by: ../../../genfft/gen_r2cf.native -fma -reorder-insns -schedule-for-pipeline -compact -variables 4 -pipeline-latency 4 -n 32 -name r2cfII_32 -dft-II -include r2cfII.h */ /* * This function contains 174 FP additions, 128 FP multiplications, * (or, 46 additions, 0 multiplications, 128 fused multiply/add), * 96 stack variables, 15 constants, and 64 memory accesses */ #include "r2cfII.h" static void r2cfII_32(R *R0, R *R1, R *Cr, R *Ci, stride rs, stride csr, stride csi, INT v, INT ivs, INT ovs) { DK(KP773010453, +0.773010453362736960810906609758469800971041293); DK(KP820678790, +0.820678790828660330972281985331011598767386482); DK(KP956940335, +0.956940335732208864935797886980269969482849206); DK(KP303346683, +0.303346683607342391675883946941299872384187453); DK(KP995184726, +0.995184726672196886244836953109479921575474869); DK(KP980785280, +0.980785280403230449126182236134239036973933731); DK(KP098491403, +0.098491403357164253077197521291327432293052451); DK(KP881921264, +0.881921264348355029712756863660388349508442621); DK(KP831469612, +0.831469612302545237078788377617905756738560812); DK(KP534511135, +0.534511135950791641089685961295362908582039528); DK(KP923879532, +0.923879532511286756128183189396788286822416626); DK(KP668178637, +0.668178637919298919997757686523080761552472251); DK(KP198912367, +0.198912367379658006911597622644676228597850501); DK(KP707106781, +0.707106781186547524400844362104849039284835938); DK(KP414213562, +0.414213562373095048801688724209698078569671875); { INT i; for (i = v; i > 0; i = i - 1, R0 = R0 + ivs, R1 = R1 + ivs, Cr = Cr + ovs, Ci = Ci + ovs, MAKE_VOLATILE_STRIDE(128, rs), MAKE_VOLATILE_STRIDE(128, csr), MAKE_VOLATILE_STRIDE(128, csi)) { E T23, T1S, T21, T1L, T2z, T2x, T1Z, T22; { E T2n, T2B, T1z, T5, T1C, T2C, T2o, Tc, T27, T1J, T1l, Tm, T26, T1G, T1k; E Tv, T1s, T1c, T2e, T1Y, T1r, T15, T2d, T1V, TP, TF, T1M, TC, T1P, TN; E TO, TI; { E T1A, T8, Te, Tj, Tf, T1B, Tb, Tg; { E T1, T2l, T2, T3, T9, Ta; T1 = R0[0]; T2l = R0[WS(rs, 8)]; T2 = R0[WS(rs, 4)]; T3 = R0[WS(rs, 12)]; { E T6, T7, T2m, T4; T6 = R0[WS(rs, 10)]; T7 = R0[WS(rs, 2)]; T9 = R0[WS(rs, 6)]; T2m = T2 + T3; T4 = T2 - T3; T1A = FNMS(KP414213562, T6, T7); T8 = FMA(KP414213562, T7, T6); T2n = FMA(KP707106781, T2m, T2l); T2B = FNMS(KP707106781, T2m, T2l); T1z = FMA(KP707106781, T4, T1); T5 = FNMS(KP707106781, T4, T1); Ta = R0[WS(rs, 14)]; } Te = R0[WS(rs, 7)]; Tj = R0[WS(rs, 15)]; Tf = R0[WS(rs, 3)]; T1B = FMS(KP414213562, T9, Ta); Tb = FMA(KP414213562, Ta, T9); Tg = R0[WS(rs, 11)]; } { E Tn, Ts, To, T1I, Tl, T1H, Ti, Tp, Tk, Th, T1T, T1U; Tn = R0[WS(rs, 9)]; T1C = T1A + T1B; T2C = T1B - T1A; T2o = T8 + Tb; Tc = T8 - Tb; Tk = Tg - Tf; Th = Tf + Tg; Ts = R0[WS(rs, 1)]; To = R0[WS(rs, 5)]; T1I = FMA(KP707106781, Tk, Tj); Tl = FNMS(KP707106781, Tk, Tj); T1H = FMA(KP707106781, Th, Te); Ti = FNMS(KP707106781, Th, Te); Tp = R0[WS(rs, 13)]; { E TT, T16, TY, T17, TW, TZ, T11, T12, Tt, Tq; TT = R1[WS(rs, 15)]; T27 = FNMS(KP198912367, T1H, T1I); T1J = FMA(KP198912367, T1I, T1H); T1l = FMA(KP668178637, Ti, Tl); Tm = FNMS(KP668178637, Tl, Ti); Tt = To - Tp; Tq = To + Tp; T16 = R1[WS(rs, 7)]; { E TU, T1F, Tu, T1E, Tr, TV; TU = R1[WS(rs, 3)]; T1F = FMA(KP707106781, Tt, Ts); Tu = FNMS(KP707106781, Tt, Ts); T1E = FMA(KP707106781, Tq, Tn); Tr = FNMS(KP707106781, Tq, Tn); TV = R1[WS(rs, 11)]; TY = R1[WS(rs, 9)]; T26 = FNMS(KP198912367, T1E, T1F); T1G = FMA(KP198912367, T1F, T1E); T1k = FMA(KP668178637, Tr, Tu); Tv = FNMS(KP668178637, Tu, Tr); T17 = TU + TV; TW = TU - TV; TZ = R1[WS(rs, 1)]; T11 = R1[WS(rs, 5)]; T12 = R1[WS(rs, 13)]; } { E TX, T1a, T10, T19, T13, T1W, T18, T1b, T14, T1X; T1T = FMS(KP707106781, TW, TT); TX = FMA(KP707106781, TW, TT); T1a = FNMS(KP414213562, TY, TZ); T10 = FMA(KP414213562, TZ, TY); T19 = FMS(KP414213562, T11, T12); T13 = FMA(KP414213562, T12, T11); T1W = FMA(KP707106781, T17, T16); T18 = FNMS(KP707106781, T17, T16); T1b = T19 - T1a; T1U = T1a + T19; T14 = T10 - T13; T1X = T10 + T13; T1s = FMA(KP923879532, T1b, T18); T1c = FNMS(KP923879532, T1b, T18); T2e = FMA(KP923879532, T1X, T1W); T1Y = FNMS(KP923879532, T1X, T1W); T1r = FNMS(KP923879532, T14, TX); T15 = FMA(KP923879532, T14, TX); } } { E Ty, TL, TG, TM, TB, TH; Ty = R1[0]; TL = R1[WS(rs, 8)]; { E Tz, TA, TD, TE; Tz = R1[WS(rs, 4)]; T2d = FMA(KP923879532, T1U, T1T); T1V = FNMS(KP923879532, T1U, T1T); TA = R1[WS(rs, 12)]; TD = R1[WS(rs, 10)]; TE = R1[WS(rs, 2)]; TG = R1[WS(rs, 6)]; TM = Tz + TA; TB = Tz - TA; TP = FNMS(KP414213562, TD, TE); TF = FMA(KP414213562, TE, TD); TH = R1[WS(rs, 14)]; } T1M = FMA(KP707106781, TB, Ty); TC = FNMS(KP707106781, TB, Ty); T1P = FMA(KP707106781, TM, TL); TN = FNMS(KP707106781, TM, TL); TO = FMS(KP414213562, TG, TH); TI = FMA(KP414213562, TH, TG); } } } { E T1j, T1O, T1p, T1R, T1o, T2E, T2D, T1m, T1D, T2w, T2v, T1K, T2i, T2c, T2h; E T29, T2t, T2r, T2f, T2j; { E T2a, T2b, T1g, TS, T1f, Tx, T2N, T2L, T1d, T1h; { E Td, TR, TK, Tw, T2J, T2K; T1j = FMA(KP923879532, Tc, T5); Td = FNMS(KP923879532, Tc, T5); { E T1N, TQ, T1Q, TJ; T1N = TP + TO; TQ = TO - TP; T1Q = TF + TI; TJ = TF - TI; T2a = FMA(KP923879532, T1N, T1M); T1O = FNMS(KP923879532, T1N, T1M); T1p = FMA(KP923879532, TQ, TN); TR = FNMS(KP923879532, TQ, TN); T2b = FMA(KP923879532, T1Q, T1P); T1R = FNMS(KP923879532, T1Q, T1P); T1o = FMA(KP923879532, TJ, TC); TK = FNMS(KP923879532, TJ, TC); Tw = Tm - Tv; T2E = Tv + Tm; } T2D = FMA(KP923879532, T2C, T2B); T2J = FNMS(KP923879532, T2C, T2B); T2K = T1k + T1l; T1m = T1k - T1l; T1g = FMA(KP534511135, TK, TR); TS = FNMS(KP534511135, TR, TK); T1f = FNMS(KP831469612, Tw, Td); Tx = FMA(KP831469612, Tw, Td); T2N = FNMS(KP831469612, T2K, T2J); T2L = FMA(KP831469612, T2K, T2J); T1d = FNMS(KP534511135, T1c, T15); T1h = FMA(KP534511135, T15, T1c); } { E T25, T28, T2p, T2q; T1D = FNMS(KP923879532, T1C, T1z); T25 = FMA(KP923879532, T1C, T1z); { E T2O, T1e, T2M, T1i; T2O = TS + T1d; T1e = TS - T1d; T2M = T1g + T1h; T1i = T1g - T1h; Ci[WS(csi, 5)] = FNMS(KP881921264, T2O, T2N); Ci[WS(csi, 10)] = -(FMA(KP881921264, T2O, T2N)); Cr[WS(csr, 2)] = FMA(KP881921264, T1e, Tx); Cr[WS(csr, 13)] = FNMS(KP881921264, T1e, Tx); Ci[WS(csi, 2)] = -(FMA(KP881921264, T2M, T2L)); Ci[WS(csi, 13)] = FNMS(KP881921264, T2M, T2L); Cr[WS(csr, 5)] = FMA(KP881921264, T1i, T1f); Cr[WS(csr, 10)] = FNMS(KP881921264, T1i, T1f); T28 = T26 - T27; T2w = T26 + T27; } T2v = FNMS(KP923879532, T2o, T2n); T2p = FMA(KP923879532, T2o, T2n); T2q = T1G + T1J; T1K = T1G - T1J; T2i = FMA(KP098491403, T2a, T2b); T2c = FNMS(KP098491403, T2b, T2a); T2h = FNMS(KP980785280, T28, T25); T29 = FMA(KP980785280, T28, T25); T2t = FNMS(KP980785280, T2q, T2p); T2r = FMA(KP980785280, T2q, T2p); T2f = FMA(KP098491403, T2e, T2d); T2j = FNMS(KP098491403, T2d, T2e); } } { E T1x, T1q, T1v, T1n, T2H, T2F, T1t, T1w; { E T2u, T2g, T2s, T2k; T2u = T2f - T2c; T2g = T2c + T2f; T2s = T2i + T2j; T2k = T2i - T2j; Ci[WS(csi, 7)] = FMA(KP995184726, T2u, T2t); Ci[WS(csi, 8)] = FMS(KP995184726, T2u, T2t); Cr[0] = FMA(KP995184726, T2g, T29); Cr[WS(csr, 15)] = FNMS(KP995184726, T2g, T29); Ci[0] = -(FMA(KP995184726, T2s, T2r)); Ci[WS(csi, 15)] = FNMS(KP995184726, T2s, T2r); Cr[WS(csr, 7)] = FMA(KP995184726, T2k, T2h); Cr[WS(csr, 8)] = FNMS(KP995184726, T2k, T2h); } T1x = FNMS(KP303346683, T1o, T1p); T1q = FMA(KP303346683, T1p, T1o); T1v = FNMS(KP831469612, T1m, T1j); T1n = FMA(KP831469612, T1m, T1j); T2H = FNMS(KP831469612, T2E, T2D); T2F = FMA(KP831469612, T2E, T2D); T1t = FMA(KP303346683, T1s, T1r); T1w = FNMS(KP303346683, T1r, T1s); { E T2I, T1u, T2G, T1y; T2I = T1q + T1t; T1u = T1q - T1t; T2G = T1x + T1w; T1y = T1w - T1x; Ci[WS(csi, 6)] = -(FMA(KP956940335, T2I, T2H)); Ci[WS(csi, 9)] = FNMS(KP956940335, T2I, T2H); Cr[WS(csr, 1)] = FMA(KP956940335, T1u, T1n); Cr[WS(csr, 14)] = FNMS(KP956940335, T1u, T1n); Ci[WS(csi, 1)] = FMA(KP956940335, T2G, T2F); Ci[WS(csi, 14)] = FMS(KP956940335, T2G, T2F); Cr[WS(csr, 6)] = FMA(KP956940335, T1y, T1v); Cr[WS(csr, 9)] = FNMS(KP956940335, T1y, T1v); } T23 = FNMS(KP820678790, T1O, T1R); T1S = FMA(KP820678790, T1R, T1O); T21 = FNMS(KP980785280, T1K, T1D); T1L = FMA(KP980785280, T1K, T1D); T2z = FMA(KP980785280, T2w, T2v); T2x = FNMS(KP980785280, T2w, T2v); T1Z = FNMS(KP820678790, T1Y, T1V); T22 = FMA(KP820678790, T1V, T1Y); } } } { E T20, T2A, T24, T2y; T20 = T1S + T1Z; T2A = T1Z - T1S; T24 = T22 - T23; T2y = T23 + T22; Ci[WS(csi, 4)] = FMS(KP773010453, T2A, T2z); Ci[WS(csi, 11)] = FMA(KP773010453, T2A, T2z); Cr[WS(csr, 3)] = FMA(KP773010453, T20, T1L); Cr[WS(csr, 12)] = FNMS(KP773010453, T20, T1L); Ci[WS(csi, 3)] = FMA(KP773010453, T2y, T2x); Ci[WS(csi, 12)] = FMS(KP773010453, T2y, T2x); Cr[WS(csr, 4)] = FMA(KP773010453, T24, T21); Cr[WS(csr, 11)] = FNMS(KP773010453, T24, T21); } } } } static const kr2c_desc desc = { 32, "r2cfII_32", {46, 0, 128, 0}, &GENUS }; void X(codelet_r2cfII_32) (planner *p) { X(kr2c_register) (p, r2cfII_32, &desc); } #else /* HAVE_FMA */ /* Generated by: ../../../genfft/gen_r2cf.native -compact -variables 4 -pipeline-latency 4 -n 32 -name r2cfII_32 -dft-II -include r2cfII.h */ /* * This function contains 174 FP additions, 82 FP multiplications, * (or, 138 additions, 46 multiplications, 36 fused multiply/add), * 62 stack variables, 15 constants, and 64 memory accesses */ #include "r2cfII.h" static void r2cfII_32(R *R0, R *R1, R *Cr, R *Ci, stride rs, stride csr, stride csi, INT v, INT ivs, INT ovs) { DK(KP471396736, +0.471396736825997648556387625905254377657460319); DK(KP881921264, +0.881921264348355029712756863660388349508442621); DK(KP634393284, +0.634393284163645498215171613225493370675687095); DK(KP773010453, +0.773010453362736960810906609758469800971041293); DK(KP290284677, +0.290284677254462367636192375817395274691476278); DK(KP956940335, +0.956940335732208864935797886980269969482849206); DK(KP995184726, +0.995184726672196886244836953109479921575474869); DK(KP098017140, +0.098017140329560601994195563888641845861136673); DK(KP555570233, +0.555570233019602224742830813948532874374937191); DK(KP831469612, +0.831469612302545237078788377617905756738560812); DK(KP195090322, +0.195090322016128267848284868477022240927691618); DK(KP980785280, +0.980785280403230449126182236134239036973933731); DK(KP382683432, +0.382683432365089771728459984030398866761344562); DK(KP923879532, +0.923879532511286756128183189396788286822416626); DK(KP707106781, +0.707106781186547524400844362104849039284835938); { INT i; for (i = v; i > 0; i = i - 1, R0 = R0 + ivs, R1 = R1 + ivs, Cr = Cr + ovs, Ci = Ci + ovs, MAKE_VOLATILE_STRIDE(128, rs), MAKE_VOLATILE_STRIDE(128, csr), MAKE_VOLATILE_STRIDE(128, csi)) { E T5, T2D, T1z, T2q, Tc, T2C, T1C, T2n, Tm, T1k, T1J, T26, Tv, T1l, T1G; E T27, T15, T1r, T1Y, T2e, T1c, T1s, T1V, T2d, TK, T1o, T1R, T2b, TR, T1p; E T1O, T2a; { E T1, T2p, T4, T2o, T2, T3; T1 = R0[0]; T2p = R0[WS(rs, 8)]; T2 = R0[WS(rs, 4)]; T3 = R0[WS(rs, 12)]; T4 = KP707106781 * (T2 - T3); T2o = KP707106781 * (T2 + T3); T5 = T1 + T4; T2D = T2p - T2o; T1z = T1 - T4; T2q = T2o + T2p; } { E T8, T1A, Tb, T1B; { E T6, T7, T9, Ta; T6 = R0[WS(rs, 2)]; T7 = R0[WS(rs, 10)]; T8 = FNMS(KP382683432, T7, KP923879532 * T6); T1A = FMA(KP382683432, T6, KP923879532 * T7); T9 = R0[WS(rs, 6)]; Ta = R0[WS(rs, 14)]; Tb = FNMS(KP923879532, Ta, KP382683432 * T9); T1B = FMA(KP923879532, T9, KP382683432 * Ta); } Tc = T8 + Tb; T2C = Tb - T8; T1C = T1A - T1B; T2n = T1A + T1B; } { E Te, Tk, Th, Tj, Tf, Tg; Te = R0[WS(rs, 1)]; Tk = R0[WS(rs, 9)]; Tf = R0[WS(rs, 5)]; Tg = R0[WS(rs, 13)]; Th = KP707106781 * (Tf - Tg); Tj = KP707106781 * (Tf + Tg); { E Ti, Tl, T1H, T1I; Ti = Te + Th; Tl = Tj + Tk; Tm = FNMS(KP195090322, Tl, KP980785280 * Ti); T1k = FMA(KP195090322, Ti, KP980785280 * Tl); T1H = Tk - Tj; T1I = Te - Th; T1J = FNMS(KP555570233, T1I, KP831469612 * T1H); T26 = FMA(KP831469612, T1I, KP555570233 * T1H); } } { E Tq, Tt, Tp, Ts, Tn, To; Tq = R0[WS(rs, 15)]; Tt = R0[WS(rs, 7)]; Tn = R0[WS(rs, 3)]; To = R0[WS(rs, 11)]; Tp = KP707106781 * (Tn - To); Ts = KP707106781 * (Tn + To); { E Tr, Tu, T1E, T1F; Tr = Tp - Tq; Tu = Ts + Tt; Tv = FMA(KP980785280, Tr, KP195090322 * Tu); T1l = FNMS(KP980785280, Tu, KP195090322 * Tr); T1E = Tt - Ts; T1F = Tp + Tq; T1G = FNMS(KP555570233, T1F, KP831469612 * T1E); T27 = FMA(KP831469612, T1F, KP555570233 * T1E); } } { E TW, T1a, TV, T19, T10, T16, T13, T17, TT, TU; TW = R1[WS(rs, 15)]; T1a = R1[WS(rs, 7)]; TT = R1[WS(rs, 3)]; TU = R1[WS(rs, 11)]; TV = KP707106781 * (TT - TU); T19 = KP707106781 * (TT + TU); { E TY, TZ, T11, T12; TY = R1[WS(rs, 1)]; TZ = R1[WS(rs, 9)]; T10 = FNMS(KP382683432, TZ, KP923879532 * TY); T16 = FMA(KP382683432, TY, KP923879532 * TZ); T11 = R1[WS(rs, 5)]; T12 = R1[WS(rs, 13)]; T13 = FNMS(KP923879532, T12, KP382683432 * T11); T17 = FMA(KP923879532, T11, KP382683432 * T12); } { E TX, T14, T1W, T1X; TX = TV - TW; T14 = T10 + T13; T15 = TX + T14; T1r = TX - T14; T1W = T13 - T10; T1X = T1a - T19; T1Y = T1W - T1X; T2e = T1W + T1X; } { E T18, T1b, T1T, T1U; T18 = T16 + T17; T1b = T19 + T1a; T1c = T18 + T1b; T1s = T1b - T18; T1T = TV + TW; T1U = T16 - T17; T1V = T1T + T1U; T2d = T1U - T1T; } } { E Ty, TP, TB, TO, TF, TL, TI, TM, Tz, TA; Ty = R1[0]; TP = R1[WS(rs, 8)]; Tz = R1[WS(rs, 4)]; TA = R1[WS(rs, 12)]; TB = KP707106781 * (Tz - TA); TO = KP707106781 * (Tz + TA); { E TD, TE, TG, TH; TD = R1[WS(rs, 2)]; TE = R1[WS(rs, 10)]; TF = FNMS(KP382683432, TE, KP923879532 * TD); TL = FMA(KP382683432, TD, KP923879532 * TE); TG = R1[WS(rs, 6)]; TH = R1[WS(rs, 14)]; TI = FNMS(KP923879532, TH, KP382683432 * TG); TM = FMA(KP923879532, TG, KP382683432 * TH); } { E TC, TJ, T1P, T1Q; TC = Ty + TB; TJ = TF + TI; TK = TC + TJ; T1o = TC - TJ; T1P = TI - TF; T1Q = TP - TO; T1R = T1P - T1Q; T2b = T1P + T1Q; } { E TN, TQ, T1M, T1N; TN = TL + TM; TQ = TO + TP; TR = TN + TQ; T1p = TQ - TN; T1M = Ty - TB; T1N = TL - TM; T1O = T1M - T1N; T2a = T1M + T1N; } } { E Tx, T1f, T2s, T2u, T1e, T2l, T1i, T2t; { E Td, Tw, T2m, T2r; Td = T5 + Tc; Tw = Tm + Tv; Tx = Td - Tw; T1f = Td + Tw; T2m = T1l - T1k; T2r = T2n + T2q; T2s = T2m - T2r; T2u = T2m + T2r; } { E TS, T1d, T1g, T1h; TS = FMA(KP098017140, TK, KP995184726 * TR); T1d = FNMS(KP995184726, T1c, KP098017140 * T15); T1e = TS + T1d; T2l = T1d - TS; T1g = FNMS(KP098017140, TR, KP995184726 * TK); T1h = FMA(KP995184726, T15, KP098017140 * T1c); T1i = T1g + T1h; T2t = T1h - T1g; } Cr[WS(csr, 8)] = Tx - T1e; Ci[WS(csi, 8)] = T2t - T2u; Cr[WS(csr, 7)] = Tx + T1e; Ci[WS(csi, 7)] = T2t + T2u; Cr[WS(csr, 15)] = T1f - T1i; Ci[WS(csi, 15)] = T2l - T2s; Cr[0] = T1f + T1i; Ci[0] = T2l + T2s; } { E T29, T2h, T2M, T2O, T2g, T2J, T2k, T2N; { E T25, T28, T2K, T2L; T25 = T1z + T1C; T28 = T26 - T27; T29 = T25 + T28; T2h = T25 - T28; T2K = T1J + T1G; T2L = T2C + T2D; T2M = T2K - T2L; T2O = T2K + T2L; } { E T2c, T2f, T2i, T2j; T2c = FMA(KP956940335, T2a, KP290284677 * T2b); T2f = FNMS(KP290284677, T2e, KP956940335 * T2d); T2g = T2c + T2f; T2J = T2f - T2c; T2i = FMA(KP290284677, T2d, KP956940335 * T2e); T2j = FNMS(KP290284677, T2a, KP956940335 * T2b); T2k = T2i - T2j; T2N = T2j + T2i; } Cr[WS(csr, 14)] = T29 - T2g; Ci[WS(csi, 14)] = T2N - T2O; Cr[WS(csr, 1)] = T29 + T2g; Ci[WS(csi, 1)] = T2N + T2O; Cr[WS(csr, 9)] = T2h - T2k; Ci[WS(csi, 9)] = T2J - T2M; Cr[WS(csr, 6)] = T2h + T2k; Ci[WS(csi, 6)] = T2J + T2M; } { E T1n, T1v, T2y, T2A, T1u, T2v, T1y, T2z; { E T1j, T1m, T2w, T2x; T1j = T5 - Tc; T1m = T1k + T1l; T1n = T1j + T1m; T1v = T1j - T1m; T2w = Tv - Tm; T2x = T2q - T2n; T2y = T2w - T2x; T2A = T2w + T2x; } { E T1q, T1t, T1w, T1x; T1q = FMA(KP773010453, T1o, KP634393284 * T1p); T1t = FNMS(KP634393284, T1s, KP773010453 * T1r); T1u = T1q + T1t; T2v = T1t - T1q; T1w = FMA(KP634393284, T1r, KP773010453 * T1s); T1x = FNMS(KP634393284, T1o, KP773010453 * T1p); T1y = T1w - T1x; T2z = T1x + T1w; } Cr[WS(csr, 12)] = T1n - T1u; Ci[WS(csi, 12)] = T2z - T2A; Cr[WS(csr, 3)] = T1n + T1u; Ci[WS(csi, 3)] = T2z + T2A; Cr[WS(csr, 11)] = T1v - T1y; Ci[WS(csi, 11)] = T2v - T2y; Cr[WS(csr, 4)] = T1v + T1y; Ci[WS(csi, 4)] = T2v + T2y; } { E T1L, T21, T2G, T2I, T20, T2H, T24, T2B; { E T1D, T1K, T2E, T2F; T1D = T1z - T1C; T1K = T1G - T1J; T1L = T1D + T1K; T21 = T1D - T1K; T2E = T2C - T2D; T2F = T26 + T27; T2G = T2E - T2F; T2I = T2F + T2E; } { E T1S, T1Z, T22, T23; T1S = FMA(KP881921264, T1O, KP471396736 * T1R); T1Z = FMA(KP881921264, T1V, KP471396736 * T1Y); T20 = T1S - T1Z; T2H = T1S + T1Z; T22 = FNMS(KP471396736, T1V, KP881921264 * T1Y); T23 = FNMS(KP471396736, T1O, KP881921264 * T1R); T24 = T22 - T23; T2B = T23 + T22; } Cr[WS(csr, 13)] = T1L - T20; Ci[WS(csi, 13)] = T2B - T2G; Cr[WS(csr, 2)] = T1L + T20; Ci[WS(csi, 2)] = T2B + T2G; Cr[WS(csr, 10)] = T21 - T24; Ci[WS(csi, 10)] = T2I - T2H; Cr[WS(csr, 5)] = T21 + T24; Ci[WS(csi, 5)] = -(T2H + T2I); } } } } static const kr2c_desc desc = { 32, "r2cfII_32", {138, 46, 36, 0}, &GENUS }; void X(codelet_r2cfII_32) (planner *p) { X(kr2c_register) (p, r2cfII_32, &desc); } #endif /* HAVE_FMA */ fftw-3.3.3/rdft/scalar/r2cf/r2cfII_64.c0000644000175000017500000013701412054410714014206 00000000000000/* * Copyright (c) 2003, 2007-11 Matteo Frigo * Copyright (c) 2003, 2007-11 Massachusetts Institute of Technology * * This program is free software; you can redistribute it and/or modify * it under the terms of the GNU General Public License as published by * the Free Software Foundation; either version 2 of the License, or * (at your option) any later version. * * This program is distributed in the hope that it will be useful, * but WITHOUT ANY WARRANTY; without even the implied warranty of * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the * GNU General Public License for more details. * * You should have received a copy of the GNU General Public License * along with this program; if not, write to the Free Software * Foundation, Inc., 51 Franklin Street, Fifth Floor, Boston, MA 02110-1301 USA * */ /* This file was automatically generated --- DO NOT EDIT */ /* Generated on Sun Nov 25 07:40:17 EST 2012 */ #include "codelet-rdft.h" #ifdef HAVE_FMA /* Generated by: ../../../genfft/gen_r2cf.native -fma -reorder-insns -schedule-for-pipeline -compact -variables 4 -pipeline-latency 4 -n 64 -name r2cfII_64 -dft-II -include r2cfII.h */ /* * This function contains 434 FP additions, 320 FP multiplications, * (or, 114 additions, 0 multiplications, 320 fused multiply/add), * 158 stack variables, 31 constants, and 128 memory accesses */ #include "r2cfII.h" static void r2cfII_64(R *R0, R *R1, R *Cr, R *Ci, stride rs, stride csr, stride csi, INT v, INT ivs, INT ovs) { DK(KP941544065, +0.941544065183020778412509402599502357185589796); DK(KP903989293, +0.903989293123443331586200297230537048710132025); DK(KP773010453, +0.773010453362736960810906609758469800971041293); DK(KP472964775, +0.472964775891319928124438237972992463904131113); DK(KP357805721, +0.357805721314524104672487743774474392487532769); DK(KP820678790, +0.820678790828660330972281985331011598767386482); DK(KP989176509, +0.989176509964780973451673738016243063983689533); DK(KP803207531, +0.803207531480644909806676512963141923879569427); DK(KP956940335, +0.956940335732208864935797886980269969482849206); DK(KP741650546, +0.741650546272035369581266691172079863842265220); DK(KP148335987, +0.148335987538347428753676511486911367000625355); DK(KP303346683, +0.303346683607342391675883946941299872384187453); DK(KP998795456, +0.998795456205172392714771604759100694443203615); DK(KP740951125, +0.740951125354959091175616897495162729728955309); DK(KP995184726, +0.995184726672196886244836953109479921575474869); DK(KP906347169, +0.906347169019147157946142717268914412664134293); DK(KP049126849, +0.049126849769467254105343321271313617079695752); DK(KP098491403, +0.098491403357164253077197521291327432293052451); DK(KP980785280, +0.980785280403230449126182236134239036973933731); DK(KP970031253, +0.970031253194543992603984207286100251456865962); DK(KP857728610, +0.857728610000272069902269984284770137042490799); DK(KP881921264, +0.881921264348355029712756863660388349508442621); DK(KP599376933, +0.599376933681923766271389869014404232837890546); DK(KP250486960, +0.250486960191305461595702160124721208578685568); DK(KP534511135, +0.534511135950791641089685961295362908582039528); DK(KP831469612, +0.831469612302545237078788377617905756738560812); DK(KP923879532, +0.923879532511286756128183189396788286822416626); DK(KP414213562, +0.414213562373095048801688724209698078569671875); DK(KP668178637, +0.668178637919298919997757686523080761552472251); DK(KP198912367, +0.198912367379658006911597622644676228597850501); DK(KP707106781, +0.707106781186547524400844362104849039284835938); { INT i; for (i = v; i > 0; i = i - 1, R0 = R0 + ivs, R1 = R1 + ivs, Cr = Cr + ovs, Ci = Ci + ovs, MAKE_VOLATILE_STRIDE(256, rs), MAKE_VOLATILE_STRIDE(256, csr), MAKE_VOLATILE_STRIDE(256, csi)) { E T5b, T6q, T6p, T5e; { E T5h, T3Z, T35, Tm, T5g, T3W, T34, Tv, T5f, T3T, T6N, T6z, T6j, T65, T33; E Td, T5z, T4D, T3q, T2C, T5C, T4O, T3n, T2b, T5k, T4b, T3c, TR, T5l, T4e; E T3b, TK, T5n, T44, T39, T1c, T5o, T47, T38, T15, T5s, T4k, T3j, T1T, T5v; E T4v, T3g, T1s, T1t, T1y, T5D, T4K, T5A, T4R, T3o, T2F, T3r, T2u, T1C, T1H; E T1D, T1z, T1w, T1E; { E T2A, T26, T4B, T23, T4M, T2y, T2z, T29; { E Te, Tj, Tn, Ts, To, Tk, Th, Tp, Tf, Tg; Te = R0[WS(rs, 14)]; Tj = R0[WS(rs, 30)]; Tf = R0[WS(rs, 6)]; Tg = R0[WS(rs, 22)]; Tn = R0[WS(rs, 18)]; Ts = R0[WS(rs, 2)]; To = R0[WS(rs, 10)]; Tk = Tg - Tf; Th = Tf + Tg; Tp = R0[WS(rs, 26)]; { E T3Q, T8, T3P, T5, T6x, T63, T3R, Tb; { E T1, T61, T9, T62, T4, Ta; { E T3V, Tu, T3U, Tr, T3Y, Tl; T1 = R0[0]; T3Y = FMA(KP707106781, Tk, Tj); Tl = FNMS(KP707106781, Tk, Tj); { E T3X, Ti, Tt, Tq; T3X = FMA(KP707106781, Th, Te); Ti = FNMS(KP707106781, Th, Te); Tt = To - Tp; Tq = To + Tp; T5h = FNMS(KP198912367, T3X, T3Y); T3Z = FMA(KP198912367, T3Y, T3X); T35 = FMA(KP668178637, Ti, Tl); Tm = FNMS(KP668178637, Tl, Ti); T3V = FMA(KP707106781, Tt, Ts); Tu = FNMS(KP707106781, Tt, Ts); T3U = FMA(KP707106781, Tq, Tn); Tr = FNMS(KP707106781, Tq, Tn); T61 = R0[WS(rs, 16)]; } { E T2, T3, T6, T7; T2 = R0[WS(rs, 8)]; T5g = FNMS(KP198912367, T3U, T3V); T3W = FMA(KP198912367, T3V, T3U); T34 = FMA(KP668178637, Tr, Tu); Tv = FNMS(KP668178637, Tu, Tr); T3 = R0[WS(rs, 24)]; T6 = R0[WS(rs, 20)]; T7 = R0[WS(rs, 4)]; T9 = R0[WS(rs, 12)]; T62 = T2 + T3; T4 = T2 - T3; T3Q = FNMS(KP414213562, T6, T7); T8 = FMA(KP414213562, T7, T6); Ta = R0[WS(rs, 28)]; } } T3P = FMA(KP707106781, T4, T1); T5 = FNMS(KP707106781, T4, T1); T6x = FNMS(KP707106781, T62, T61); T63 = FMA(KP707106781, T62, T61); T3R = FMS(KP414213562, T9, Ta); Tb = FMA(KP414213562, Ta, T9); } { E T1Z, T2w, T27, T2x, T22, T28; T1Z = R1[WS(rs, 31)]; { E T3S, T6y, T64, Tc; T3S = T3Q + T3R; T6y = T3R - T3Q; T64 = T8 + Tb; Tc = T8 - Tb; T5f = FMA(KP923879532, T3S, T3P); T3T = FNMS(KP923879532, T3S, T3P); T6N = FNMS(KP923879532, T6y, T6x); T6z = FMA(KP923879532, T6y, T6x); T6j = FNMS(KP923879532, T64, T63); T65 = FMA(KP923879532, T64, T63); T33 = FMA(KP923879532, Tc, T5); Td = FNMS(KP923879532, Tc, T5); T2w = R1[WS(rs, 15)]; } { E T20, T21, T24, T25; T20 = R1[WS(rs, 7)]; T21 = R1[WS(rs, 23)]; T24 = R1[WS(rs, 19)]; T25 = R1[WS(rs, 3)]; T27 = R1[WS(rs, 11)]; T2x = T20 + T21; T22 = T20 - T21; T2A = FNMS(KP414213562, T24, T25); T26 = FMA(KP414213562, T25, T24); T28 = R1[WS(rs, 27)]; } T4B = FMS(KP707106781, T22, T1Z); T23 = FMA(KP707106781, T22, T1Z); T4M = FMA(KP707106781, T2x, T2w); T2y = FNMS(KP707106781, T2x, T2w); T2z = FMS(KP414213562, T27, T28); T29 = FMA(KP414213562, T28, T27); } } } { E T1a, T10, T42, TX, T45, T18, T19, T13; { E TP, TF, T49, TC, T4c, TN, TO, TI; { E Ty, TL, TG, TM, TB, TH; Ty = R0[WS(rs, 17)]; { E T4C, T2B, T4N, T2a; T4C = T2A + T2z; T2B = T2z - T2A; T4N = T26 + T29; T2a = T26 - T29; T5z = FMA(KP923879532, T4C, T4B); T4D = FNMS(KP923879532, T4C, T4B); T3q = FMA(KP923879532, T2B, T2y); T2C = FNMS(KP923879532, T2B, T2y); T5C = FMA(KP923879532, T4N, T4M); T4O = FNMS(KP923879532, T4N, T4M); T3n = FNMS(KP923879532, T2a, T23); T2b = FMA(KP923879532, T2a, T23); TL = R0[WS(rs, 1)]; } { E Tz, TA, TD, TE; Tz = R0[WS(rs, 9)]; TA = R0[WS(rs, 25)]; TD = R0[WS(rs, 29)]; TE = R0[WS(rs, 13)]; TG = R0[WS(rs, 5)]; TM = Tz - TA; TB = Tz + TA; TP = FMA(KP414213562, TD, TE); TF = FMS(KP414213562, TE, TD); TH = R0[WS(rs, 21)]; } T49 = FMA(KP707106781, TB, Ty); TC = FNMS(KP707106781, TB, Ty); T4c = FMA(KP707106781, TM, TL); TN = FNMS(KP707106781, TM, TL); TO = FMA(KP414213562, TG, TH); TI = FNMS(KP414213562, TH, TG); } { E TT, T16, T11, T17, TW, T12; TT = R0[WS(rs, 15)]; { E T4a, TQ, T4d, TJ; T4a = TO + TP; TQ = TO - TP; T4d = TI + TF; TJ = TF - TI; T5k = FMA(KP923879532, T4a, T49); T4b = FNMS(KP923879532, T4a, T49); T3c = FMA(KP923879532, TQ, TN); TR = FNMS(KP923879532, TQ, TN); T5l = FMA(KP923879532, T4d, T4c); T4e = FNMS(KP923879532, T4d, T4c); T3b = FMA(KP923879532, TJ, TC); TK = FNMS(KP923879532, TJ, TC); T16 = R0[WS(rs, 31)]; } { E TU, TV, TY, TZ; TU = R0[WS(rs, 7)]; TV = R0[WS(rs, 23)]; TY = R0[WS(rs, 3)]; TZ = R0[WS(rs, 19)]; T11 = R0[WS(rs, 27)]; T17 = TV - TU; TW = TU + TV; T1a = FMA(KP414213562, TY, TZ); T10 = FMS(KP414213562, TZ, TY); T12 = R0[WS(rs, 11)]; } T42 = FMA(KP707106781, TW, TT); TX = FNMS(KP707106781, TW, TT); T45 = FMA(KP707106781, T17, T16); T18 = FNMS(KP707106781, T17, T16); T19 = FMA(KP414213562, T11, T12); T13 = FNMS(KP414213562, T12, T11); } } { E T1R, T1n, T4i, T1k, T4t, T1P, T1Q, T1q; { E T1g, T1N, T1o, T1O, T1j, T1p; T1g = R1[0]; { E T43, T1b, T46, T14; T43 = T1a + T19; T1b = T19 - T1a; T46 = T10 + T13; T14 = T10 - T13; T5n = FMA(KP923879532, T43, T42); T44 = FNMS(KP923879532, T43, T42); T39 = FMA(KP923879532, T1b, T18); T1c = FNMS(KP923879532, T1b, T18); T5o = FMA(KP923879532, T46, T45); T47 = FNMS(KP923879532, T46, T45); T38 = FMA(KP923879532, T14, TX); T15 = FNMS(KP923879532, T14, TX); T1N = R1[WS(rs, 16)]; } { E T1h, T1i, T1l, T1m; T1h = R1[WS(rs, 8)]; T1i = R1[WS(rs, 24)]; T1l = R1[WS(rs, 20)]; T1m = R1[WS(rs, 4)]; T1o = R1[WS(rs, 12)]; T1O = T1h + T1i; T1j = T1h - T1i; T1R = FNMS(KP414213562, T1l, T1m); T1n = FMA(KP414213562, T1m, T1l); T1p = R1[WS(rs, 28)]; } T4i = FMA(KP707106781, T1j, T1g); T1k = FNMS(KP707106781, T1j, T1g); T4t = FMA(KP707106781, T1O, T1N); T1P = FNMS(KP707106781, T1O, T1N); T1Q = FMS(KP414213562, T1o, T1p); T1q = FMA(KP414213562, T1p, T1o); } { E T2c, T2h, T2l, T2q, T2m, T2i, T2f, T2n, T2d, T2e; T2c = R1[WS(rs, 13)]; { E T4j, T1S, T4u, T1r; T4j = T1R + T1Q; T1S = T1Q - T1R; T4u = T1n + T1q; T1r = T1n - T1q; T5s = FMA(KP923879532, T4j, T4i); T4k = FNMS(KP923879532, T4j, T4i); T3j = FMA(KP923879532, T1S, T1P); T1T = FNMS(KP923879532, T1S, T1P); T5v = FMA(KP923879532, T4u, T4t); T4v = FNMS(KP923879532, T4u, T4t); T3g = FMA(KP923879532, T1r, T1k); T1s = FNMS(KP923879532, T1r, T1k); T2h = R1[WS(rs, 29)]; T2d = R1[WS(rs, 5)]; T2e = R1[WS(rs, 21)]; } T2l = R1[WS(rs, 17)]; T2q = R1[WS(rs, 1)]; T2m = R1[WS(rs, 9)]; T2i = T2d - T2e; T2f = T2d + T2e; T2n = R1[WS(rs, 25)]; { E T1u, T1v, T2j, T4I; T1t = R1[WS(rs, 14)]; T2j = FMA(KP707106781, T2i, T2h); T4I = FMS(KP707106781, T2i, T2h); { E T4H, T2g, T2r, T2o; T4H = FMA(KP707106781, T2f, T2c); T2g = FNMS(KP707106781, T2f, T2c); T2r = T2m - T2n; T2o = T2m + T2n; { E T4J, T4P, T2E, T2k; T4J = FNMS(KP198912367, T4I, T4H); T4P = FMA(KP198912367, T4H, T4I); T2E = FMA(KP668178637, T2g, T2j); T2k = FNMS(KP668178637, T2j, T2g); { E T2s, T4F, T4E, T2p; T2s = FNMS(KP707106781, T2r, T2q); T4F = FMA(KP707106781, T2r, T2q); T4E = FMA(KP707106781, T2o, T2l); T2p = FNMS(KP707106781, T2o, T2l); T1y = R1[WS(rs, 30)]; T1u = R1[WS(rs, 6)]; { E T4G, T4Q, T2D, T2t; T4G = FMA(KP198912367, T4F, T4E); T4Q = FNMS(KP198912367, T4E, T4F); T2D = FMA(KP668178637, T2p, T2s); T2t = FNMS(KP668178637, T2s, T2p); T5D = T4G + T4J; T4K = T4G - T4J; T5A = T4Q + T4P; T4R = T4P - T4Q; T3o = T2D - T2E; T2F = T2D + T2E; T3r = T2t + T2k; T2u = T2k - T2t; T1v = R1[WS(rs, 22)]; } } } } T1C = R1[WS(rs, 18)]; T1H = R1[WS(rs, 2)]; T1D = R1[WS(rs, 10)]; T1z = T1u - T1v; T1w = T1u + T1v; T1E = R1[WS(rs, 26)]; } } } } } { E T6A, T4r, T4y, T3h, T3k, T36, T6k, T40, T5X, T6c, T6b, T60; { E T5w, T5t, T2Z, T6U, T6T, T32; { E Tx, T2N, T2v, T6V, T6P, T6Q, T1e, T2G, T31, T2X, T2L, T1Y, T6W, T2Q, T30; E T2U; { E T1W, T1L, T2O, T2P, T2V, T2W, T6O, TS, T1d; { E T4q, T4w, T1V, T1B, T1J, T4m, T4l, T1G, Tw, T1A, T4p; T6A = Tv + Tm; Tw = Tm - Tv; T1A = FMA(KP707106781, T1z, T1y); T4p = FMS(KP707106781, T1z, T1y); { E T4o, T1x, T1I, T1F; T4o = FMA(KP707106781, T1w, T1t); T1x = FNMS(KP707106781, T1w, T1t); T1I = T1D - T1E; T1F = T1D + T1E; T4q = FNMS(KP198912367, T4p, T4o); T4w = FMA(KP198912367, T4o, T4p); T1V = FMA(KP668178637, T1x, T1A); T1B = FNMS(KP668178637, T1A, T1x); T1J = FNMS(KP707106781, T1I, T1H); T4m = FMA(KP707106781, T1I, T1H); T4l = FMA(KP707106781, T1F, T1C); T1G = FNMS(KP707106781, T1F, T1C); Tx = FNMS(KP831469612, Tw, Td); T2N = FMA(KP831469612, Tw, Td); } { E T4n, T4x, T1U, T1K; T4n = FMA(KP198912367, T4m, T4l); T4x = FNMS(KP198912367, T4l, T4m); T1U = FMA(KP668178637, T1G, T1J); T1K = FNMS(KP668178637, T1J, T1G); T5w = T4n + T4q; T4r = T4n - T4q; T5t = T4x + T4w; T4y = T4w - T4x; T3h = T1U - T1V; T1W = T1U + T1V; T3k = T1K + T1B; T1L = T1B - T1K; T6O = T34 + T35; T36 = T34 - T35; } } T2O = FNMS(KP534511135, TK, TR); TS = FMA(KP534511135, TR, TK); T1d = FMA(KP534511135, T1c, T15); T2P = FNMS(KP534511135, T15, T1c); T2v = FMA(KP831469612, T2u, T2b); T2V = FNMS(KP831469612, T2u, T2b); T6V = FNMS(KP831469612, T6O, T6N); T6P = FMA(KP831469612, T6O, T6N); T6Q = TS + T1d; T1e = TS - T1d; T2W = FMA(KP831469612, T2F, T2C); T2G = FNMS(KP831469612, T2F, T2C); { E T2S, T2T, T1M, T1X; T2S = FMA(KP831469612, T1L, T1s); T1M = FNMS(KP831469612, T1L, T1s); T1X = FNMS(KP831469612, T1W, T1T); T2T = FMA(KP831469612, T1W, T1T); T31 = FMA(KP250486960, T2V, T2W); T2X = FNMS(KP250486960, T2W, T2V); T2L = FNMS(KP599376933, T1M, T1X); T1Y = FMA(KP599376933, T1X, T1M); T6W = T2O + T2P; T2Q = T2O - T2P; T30 = FMA(KP250486960, T2S, T2T); T2U = FNMS(KP250486960, T2T, T2S); } } { E T2J, T1f, T6X, T6Z, T2K, T2H; T2J = FNMS(KP881921264, T1e, Tx); T1f = FMA(KP881921264, T1e, Tx); T6X = FNMS(KP881921264, T6W, T6V); T6Z = FMA(KP881921264, T6W, T6V); T2K = FNMS(KP599376933, T2v, T2G); T2H = FMA(KP599376933, T2G, T2v); { E T2R, T2Y, T6R, T6S; T2Z = FNMS(KP881921264, T2Q, T2N); T2R = FMA(KP881921264, T2Q, T2N); { E T2M, T6Y, T70, T2I; T2M = T2K - T2L; T6Y = T2L + T2K; T70 = T1Y + T2H; T2I = T1Y - T2H; Cr[WS(csr, 10)] = FMA(KP857728610, T2M, T2J); Cr[WS(csr, 21)] = FNMS(KP857728610, T2M, T2J); Ci[WS(csi, 5)] = FMA(KP857728610, T6Y, T6X); Ci[WS(csi, 26)] = FMS(KP857728610, T6Y, T6X); Ci[WS(csi, 21)] = FNMS(KP857728610, T70, T6Z); Ci[WS(csi, 10)] = -(FMA(KP857728610, T70, T6Z)); Cr[WS(csr, 5)] = FMA(KP857728610, T2I, T1f); Cr[WS(csr, 26)] = FNMS(KP857728610, T2I, T1f); T2Y = T2U - T2X; T6U = T2U + T2X; } T6T = FNMS(KP881921264, T6Q, T6P); T6R = FMA(KP881921264, T6Q, T6P); T6S = T30 + T31; T32 = T30 - T31; Cr[WS(csr, 2)] = FMA(KP970031253, T2Y, T2R); Cr[WS(csr, 29)] = FNMS(KP970031253, T2Y, T2R); Ci[WS(csi, 29)] = FNMS(KP970031253, T6S, T6R); Ci[WS(csi, 2)] = -(FMA(KP970031253, T6S, T6R)); } } } { E T5j, T5L, T5B, T6d, T67, T68, T5q, T5E, T5Z, T5V, T5J, T5y, T6e, T5O, T5Y; E T5S; { E T5M, T5N, T5T, T5U; { E T66, T5i, T5m, T5p; T6k = T5g + T5h; T5i = T5g - T5h; Cr[WS(csr, 13)] = FMA(KP970031253, T32, T2Z); Cr[WS(csr, 18)] = FNMS(KP970031253, T32, T2Z); Ci[WS(csi, 13)] = FNMS(KP970031253, T6U, T6T); Ci[WS(csi, 18)] = -(FMA(KP970031253, T6U, T6T)); T5j = FNMS(KP980785280, T5i, T5f); T5L = FMA(KP980785280, T5i, T5f); T66 = T3W + T3Z; T40 = T3W - T3Z; T5M = FNMS(KP098491403, T5k, T5l); T5m = FMA(KP098491403, T5l, T5k); T5p = FMA(KP098491403, T5o, T5n); T5N = FNMS(KP098491403, T5n, T5o); T5B = FNMS(KP980785280, T5A, T5z); T5T = FMA(KP980785280, T5A, T5z); T6d = FNMS(KP980785280, T66, T65); T67 = FMA(KP980785280, T66, T65); T68 = T5m + T5p; T5q = T5m - T5p; T5U = FMA(KP980785280, T5D, T5C); T5E = FNMS(KP980785280, T5D, T5C); } { E T5Q, T5R, T5u, T5x; T5Q = FMA(KP980785280, T5t, T5s); T5u = FNMS(KP980785280, T5t, T5s); T5x = FNMS(KP980785280, T5w, T5v); T5R = FMA(KP980785280, T5w, T5v); T5Z = FNMS(KP049126849, T5T, T5U); T5V = FMA(KP049126849, T5U, T5T); T5J = FNMS(KP906347169, T5u, T5x); T5y = FMA(KP906347169, T5x, T5u); T6e = T5M + T5N; T5O = T5M - T5N; T5Y = FMA(KP049126849, T5Q, T5R); T5S = FNMS(KP049126849, T5R, T5Q); } } { E T5H, T5r, T6f, T6h, T5I, T5F; T5H = FNMS(KP995184726, T5q, T5j); T5r = FMA(KP995184726, T5q, T5j); T6f = FNMS(KP995184726, T6e, T6d); T6h = FMA(KP995184726, T6e, T6d); T5I = FMA(KP906347169, T5B, T5E); T5F = FNMS(KP906347169, T5E, T5B); { E T5P, T5W, T69, T6a; T5X = FNMS(KP995184726, T5O, T5L); T5P = FMA(KP995184726, T5O, T5L); { E T5K, T6g, T6i, T5G; T5K = T5I - T5J; T6g = T5J + T5I; T6i = T5F - T5y; T5G = T5y + T5F; Cr[WS(csr, 8)] = FMA(KP740951125, T5K, T5H); Cr[WS(csr, 23)] = FNMS(KP740951125, T5K, T5H); Ci[WS(csi, 7)] = FMA(KP740951125, T6g, T6f); Ci[WS(csi, 24)] = FMS(KP740951125, T6g, T6f); Ci[WS(csi, 23)] = FMA(KP740951125, T6i, T6h); Ci[WS(csi, 8)] = FMS(KP740951125, T6i, T6h); Cr[WS(csr, 7)] = FMA(KP740951125, T5G, T5r); Cr[WS(csr, 24)] = FNMS(KP740951125, T5G, T5r); T5W = T5S + T5V; T6c = T5V - T5S; } T6b = FNMS(KP995184726, T68, T67); T69 = FMA(KP995184726, T68, T67); T6a = T5Y + T5Z; T60 = T5Y - T5Z; Cr[0] = FMA(KP998795456, T5W, T5P); Cr[WS(csr, 31)] = FNMS(KP998795456, T5W, T5P); Ci[WS(csi, 31)] = FNMS(KP998795456, T6a, T69); Ci[0] = -(FMA(KP998795456, T6a, T69)); } } } } { E T3L, T6G, T6F, T3O; { E T37, T3z, T3p, T6H, T6B, T6C, T3e, T3s, T3M, T3J, T3w, T3m, T6I, T3C, T3N; E T3G; { E T3B, T3A, T3H, T3I, T3a, T3d; Cr[WS(csr, 15)] = FMA(KP998795456, T60, T5X); Cr[WS(csr, 16)] = FNMS(KP998795456, T60, T5X); Ci[WS(csi, 15)] = FMA(KP998795456, T6c, T6b); Ci[WS(csi, 16)] = FMS(KP998795456, T6c, T6b); T37 = FNMS(KP831469612, T36, T33); T3z = FMA(KP831469612, T36, T33); T3B = FMA(KP303346683, T38, T39); T3a = FNMS(KP303346683, T39, T38); T3d = FNMS(KP303346683, T3c, T3b); T3A = FMA(KP303346683, T3b, T3c); T3p = FMA(KP831469612, T3o, T3n); T3H = FNMS(KP831469612, T3o, T3n); T6H = FNMS(KP831469612, T6A, T6z); T6B = FMA(KP831469612, T6A, T6z); T6C = T3d + T3a; T3e = T3a - T3d; T3I = FMA(KP831469612, T3r, T3q); T3s = FNMS(KP831469612, T3r, T3q); { E T3E, T3F, T3i, T3l; T3E = FMA(KP831469612, T3h, T3g); T3i = FNMS(KP831469612, T3h, T3g); T3l = FNMS(KP831469612, T3k, T3j); T3F = FMA(KP831469612, T3k, T3j); T3M = FNMS(KP148335987, T3H, T3I); T3J = FMA(KP148335987, T3I, T3H); T3w = FMA(KP741650546, T3i, T3l); T3m = FNMS(KP741650546, T3l, T3i); T6I = T3A + T3B; T3C = T3A - T3B; T3N = FNMS(KP148335987, T3E, T3F); T3G = FMA(KP148335987, T3F, T3E); } } { E T3v, T3f, T6J, T6L, T3x, T3t; T3v = FNMS(KP956940335, T3e, T37); T3f = FMA(KP956940335, T3e, T37); T6J = FMA(KP956940335, T6I, T6H); T6L = FNMS(KP956940335, T6I, T6H); T3x = FMA(KP741650546, T3p, T3s); T3t = FNMS(KP741650546, T3s, T3p); { E T3D, T3K, T6D, T6E; T3L = FNMS(KP956940335, T3C, T3z); T3D = FMA(KP956940335, T3C, T3z); { E T3y, T6K, T6M, T3u; T3y = T3w - T3x; T6K = T3w + T3x; T6M = T3m + T3t; T3u = T3m - T3t; Cr[WS(csr, 9)] = FMA(KP803207531, T3y, T3v); Cr[WS(csr, 22)] = FNMS(KP803207531, T3y, T3v); Ci[WS(csi, 25)] = FNMS(KP803207531, T6K, T6J); Ci[WS(csi, 6)] = -(FMA(KP803207531, T6K, T6J)); Ci[WS(csi, 9)] = FNMS(KP803207531, T6M, T6L); Ci[WS(csi, 22)] = -(FMA(KP803207531, T6M, T6L)); Cr[WS(csr, 6)] = FMA(KP803207531, T3u, T3f); Cr[WS(csr, 25)] = FNMS(KP803207531, T3u, T3f); T3K = T3G - T3J; T6G = T3G + T3J; } T6F = FNMS(KP956940335, T6C, T6B); T6D = FMA(KP956940335, T6C, T6B); T6E = T3N + T3M; T3O = T3M - T3N; Cr[WS(csr, 1)] = FMA(KP989176509, T3K, T3D); Cr[WS(csr, 30)] = FNMS(KP989176509, T3K, T3D); Ci[WS(csi, 1)] = FMA(KP989176509, T6E, T6D); Ci[WS(csi, 30)] = FMS(KP989176509, T6E, T6D); } } } { E T41, T4Z, T4L, T6r, T6l, T6m, T4g, T4S, T5c, T59, T4W, T4A, T6s, T52, T5d; E T56; { E T51, T50, T57, T58, T48, T4f; Cr[WS(csr, 14)] = FMA(KP989176509, T3O, T3L); Cr[WS(csr, 17)] = FNMS(KP989176509, T3O, T3L); Ci[WS(csi, 17)] = FNMS(KP989176509, T6G, T6F); Ci[WS(csi, 14)] = -(FMA(KP989176509, T6G, T6F)); T41 = FNMS(KP980785280, T40, T3T); T4Z = FMA(KP980785280, T40, T3T); T51 = FMA(KP820678790, T44, T47); T48 = FNMS(KP820678790, T47, T44); T4f = FNMS(KP820678790, T4e, T4b); T50 = FMA(KP820678790, T4b, T4e); T4L = FNMS(KP980785280, T4K, T4D); T57 = FMA(KP980785280, T4K, T4D); T6r = FMA(KP980785280, T6k, T6j); T6l = FNMS(KP980785280, T6k, T6j); T6m = T4f + T48; T4g = T48 - T4f; T58 = FMA(KP980785280, T4R, T4O); T4S = FNMS(KP980785280, T4R, T4O); { E T54, T55, T4s, T4z; T54 = FMA(KP980785280, T4r, T4k); T4s = FNMS(KP980785280, T4r, T4k); T4z = FNMS(KP980785280, T4y, T4v); T55 = FMA(KP980785280, T4y, T4v); T5c = FMA(KP357805721, T57, T58); T59 = FNMS(KP357805721, T58, T57); T4W = FMA(KP472964775, T4s, T4z); T4A = FNMS(KP472964775, T4z, T4s); T6s = T50 + T51; T52 = T50 - T51; T5d = FNMS(KP357805721, T54, T55); T56 = FMA(KP357805721, T55, T54); } } { E T4V, T4h, T6t, T6v, T4X, T4T; T4V = FNMS(KP773010453, T4g, T41); T4h = FMA(KP773010453, T4g, T41); T6t = FMA(KP773010453, T6s, T6r); T6v = FNMS(KP773010453, T6s, T6r); T4X = FNMS(KP472964775, T4L, T4S); T4T = FMA(KP472964775, T4S, T4L); { E T53, T5a, T6n, T6o; T5b = FNMS(KP773010453, T52, T4Z); T53 = FMA(KP773010453, T52, T4Z); { E T4Y, T6u, T6w, T4U; T4Y = T4W - T4X; T6u = T4W + T4X; T6w = T4T - T4A; T4U = T4A + T4T; Cr[WS(csr, 11)] = FMA(KP903989293, T4Y, T4V); Cr[WS(csr, 20)] = FNMS(KP903989293, T4Y, T4V); Ci[WS(csi, 27)] = FNMS(KP903989293, T6u, T6t); Ci[WS(csi, 4)] = -(FMA(KP903989293, T6u, T6t)); Ci[WS(csi, 11)] = FMA(KP903989293, T6w, T6v); Ci[WS(csi, 20)] = FMS(KP903989293, T6w, T6v); Cr[WS(csr, 4)] = FMA(KP903989293, T4U, T4h); Cr[WS(csr, 27)] = FNMS(KP903989293, T4U, T4h); T5a = T56 + T59; T6q = T59 - T56; } T6p = FNMS(KP773010453, T6m, T6l); T6n = FMA(KP773010453, T6m, T6l); T6o = T5d + T5c; T5e = T5c - T5d; Cr[WS(csr, 3)] = FMA(KP941544065, T5a, T53); Cr[WS(csr, 28)] = FNMS(KP941544065, T5a, T53); Ci[WS(csi, 3)] = FMA(KP941544065, T6o, T6n); Ci[WS(csi, 28)] = FMS(KP941544065, T6o, T6n); } } } } } } Cr[WS(csr, 12)] = FMA(KP941544065, T5e, T5b); Cr[WS(csr, 19)] = FNMS(KP941544065, T5e, T5b); Ci[WS(csi, 19)] = FMA(KP941544065, T6q, T6p); Ci[WS(csi, 12)] = FMS(KP941544065, T6q, T6p); } } } static const kr2c_desc desc = { 64, "r2cfII_64", {114, 0, 320, 0}, &GENUS }; void X(codelet_r2cfII_64) (planner *p) { X(kr2c_register) (p, r2cfII_64, &desc); } #else /* HAVE_FMA */ /* Generated by: ../../../genfft/gen_r2cf.native -compact -variables 4 -pipeline-latency 4 -n 64 -name r2cfII_64 -dft-II -include r2cfII.h */ /* * This function contains 434 FP additions, 206 FP multiplications, * (or, 342 additions, 114 multiplications, 92 fused multiply/add), * 118 stack variables, 31 constants, and 128 memory accesses */ #include "r2cfII.h" static void r2cfII_64(R *R0, R *R1, R *Cr, R *Ci, stride rs, stride csr, stride csi, INT v, INT ivs, INT ovs) { DK(KP242980179, +0.242980179903263889948274162077471118320990783); DK(KP970031253, +0.970031253194543992603984207286100251456865962); DK(KP857728610, +0.857728610000272069902269984284770137042490799); DK(KP514102744, +0.514102744193221726593693838968815772608049120); DK(KP471396736, +0.471396736825997648556387625905254377657460319); DK(KP881921264, +0.881921264348355029712756863660388349508442621); DK(KP427555093, +0.427555093430282094320966856888798534304578629); DK(KP903989293, +0.903989293123443331586200297230537048710132025); DK(KP336889853, +0.336889853392220050689253212619147570477766780); DK(KP941544065, +0.941544065183020778412509402599502357185589796); DK(KP773010453, +0.773010453362736960810906609758469800971041293); DK(KP634393284, +0.634393284163645498215171613225493370675687095); DK(KP595699304, +0.595699304492433343467036528829969889511926338); DK(KP803207531, +0.803207531480644909806676512963141923879569427); DK(KP146730474, +0.146730474455361751658850129646717819706215317); DK(KP989176509, +0.989176509964780973451673738016243063983689533); DK(KP956940335, +0.956940335732208864935797886980269969482849206); DK(KP290284677, +0.290284677254462367636192375817395274691476278); DK(KP049067674, +0.049067674327418014254954976942682658314745363); DK(KP998795456, +0.998795456205172392714771604759100694443203615); DK(KP671558954, +0.671558954847018400625376850427421803228750632); DK(KP740951125, +0.740951125354959091175616897495162729728955309); DK(KP098017140, +0.098017140329560601994195563888641845861136673); DK(KP995184726, +0.995184726672196886244836953109479921575474869); DK(KP382683432, +0.382683432365089771728459984030398866761344562); DK(KP923879532, +0.923879532511286756128183189396788286822416626); DK(KP555570233, +0.555570233019602224742830813948532874374937191); DK(KP831469612, +0.831469612302545237078788377617905756738560812); DK(KP195090322, +0.195090322016128267848284868477022240927691618); DK(KP980785280, +0.980785280403230449126182236134239036973933731); DK(KP707106781, +0.707106781186547524400844362104849039284835938); { INT i; for (i = v; i > 0; i = i - 1, R0 = R0 + ivs, R1 = R1 + ivs, Cr = Cr + ovs, Ci = Ci + ovs, MAKE_VOLATILE_STRIDE(256, rs), MAKE_VOLATILE_STRIDE(256, csr), MAKE_VOLATILE_STRIDE(256, csi)) { E Tm, T34, T3Z, T5g, Tv, T35, T3W, T5h, Td, T33, T6B, T6Q, T3T, T5f, T68; E T6m, T2b, T3n, T4O, T5D, T2F, T3r, T4K, T5z, TK, T3c, T47, T5n, TR, T3b; E T44, T5o, T15, T38, T4e, T5l, T1c, T39, T4b, T5k, T1s, T3g, T4v, T5w, T1W; E T3k, T4k, T5s, T2u, T3q, T4R, T5A, T2y, T3o, T4H, T5C, T1L, T3j, T4y, T5t; E T1P, T3h, T4r, T5v; { E Te, Tk, Th, Tj, Tf, Tg; Te = R0[WS(rs, 2)]; Tk = R0[WS(rs, 18)]; Tf = R0[WS(rs, 10)]; Tg = R0[WS(rs, 26)]; Th = KP707106781 * (Tf - Tg); Tj = KP707106781 * (Tf + Tg); { E Ti, Tl, T3X, T3Y; Ti = Te + Th; Tl = Tj + Tk; Tm = FNMS(KP195090322, Tl, KP980785280 * Ti); T34 = FMA(KP195090322, Ti, KP980785280 * Tl); T3X = Tk - Tj; T3Y = Te - Th; T3Z = FNMS(KP555570233, T3Y, KP831469612 * T3X); T5g = FMA(KP831469612, T3Y, KP555570233 * T3X); } } { E Tq, Tt, Tp, Ts, Tn, To; Tq = R0[WS(rs, 30)]; Tt = R0[WS(rs, 14)]; Tn = R0[WS(rs, 6)]; To = R0[WS(rs, 22)]; Tp = KP707106781 * (Tn - To); Ts = KP707106781 * (Tn + To); { E Tr, Tu, T3U, T3V; Tr = Tp - Tq; Tu = Ts + Tt; Tv = FMA(KP980785280, Tr, KP195090322 * Tu); T35 = FNMS(KP980785280, Tu, KP195090322 * Tr); T3U = Tt - Ts; T3V = Tp + Tq; T3W = FNMS(KP555570233, T3V, KP831469612 * T3U); T5h = FMA(KP831469612, T3V, KP555570233 * T3U); } } { E T1, T66, T4, T65, T8, T3Q, Tb, T3R, T2, T3; T1 = R0[0]; T66 = R0[WS(rs, 16)]; T2 = R0[WS(rs, 8)]; T3 = R0[WS(rs, 24)]; T4 = KP707106781 * (T2 - T3); T65 = KP707106781 * (T2 + T3); { E T6, T7, T9, Ta; T6 = R0[WS(rs, 4)]; T7 = R0[WS(rs, 20)]; T8 = FNMS(KP382683432, T7, KP923879532 * T6); T3Q = FMA(KP382683432, T6, KP923879532 * T7); T9 = R0[WS(rs, 12)]; Ta = R0[WS(rs, 28)]; Tb = FNMS(KP923879532, Ta, KP382683432 * T9); T3R = FMA(KP923879532, T9, KP382683432 * Ta); } { E T5, Tc, T6z, T6A; T5 = T1 + T4; Tc = T8 + Tb; Td = T5 + Tc; T33 = T5 - Tc; T6z = Tb - T8; T6A = T66 - T65; T6B = T6z - T6A; T6Q = T6z + T6A; } { E T3P, T3S, T64, T67; T3P = T1 - T4; T3S = T3Q - T3R; T3T = T3P - T3S; T5f = T3P + T3S; T64 = T3Q + T3R; T67 = T65 + T66; T68 = T64 + T67; T6m = T67 - T64; } } { E T22, T2D, T21, T2C, T26, T2z, T29, T2A, T1Z, T20; T22 = R1[WS(rs, 31)]; T2D = R1[WS(rs, 15)]; T1Z = R1[WS(rs, 7)]; T20 = R1[WS(rs, 23)]; T21 = KP707106781 * (T1Z - T20); T2C = KP707106781 * (T1Z + T20); { E T24, T25, T27, T28; T24 = R1[WS(rs, 3)]; T25 = R1[WS(rs, 19)]; T26 = FNMS(KP382683432, T25, KP923879532 * T24); T2z = FMA(KP382683432, T24, KP923879532 * T25); T27 = R1[WS(rs, 11)]; T28 = R1[WS(rs, 27)]; T29 = FNMS(KP923879532, T28, KP382683432 * T27); T2A = FMA(KP923879532, T27, KP382683432 * T28); } { E T23, T2a, T4M, T4N; T23 = T21 - T22; T2a = T26 + T29; T2b = T23 + T2a; T3n = T23 - T2a; T4M = T29 - T26; T4N = T2D - T2C; T4O = T4M - T4N; T5D = T4M + T4N; } { E T2B, T2E, T4I, T4J; T2B = T2z + T2A; T2E = T2C + T2D; T2F = T2B + T2E; T3r = T2E - T2B; T4I = T21 + T22; T4J = T2z - T2A; T4K = T4I + T4J; T5z = T4J - T4I; } } { E Ty, TP, TB, TO, TF, TL, TI, TM, Tz, TA; Ty = R0[WS(rs, 1)]; TP = R0[WS(rs, 17)]; Tz = R0[WS(rs, 9)]; TA = R0[WS(rs, 25)]; TB = KP707106781 * (Tz - TA); TO = KP707106781 * (Tz + TA); { E TD, TE, TG, TH; TD = R0[WS(rs, 5)]; TE = R0[WS(rs, 21)]; TF = FNMS(KP382683432, TE, KP923879532 * TD); TL = FMA(KP382683432, TD, KP923879532 * TE); TG = R0[WS(rs, 13)]; TH = R0[WS(rs, 29)]; TI = FNMS(KP923879532, TH, KP382683432 * TG); TM = FMA(KP923879532, TG, KP382683432 * TH); } { E TC, TJ, T45, T46; TC = Ty + TB; TJ = TF + TI; TK = TC + TJ; T3c = TC - TJ; T45 = TI - TF; T46 = TP - TO; T47 = T45 - T46; T5n = T45 + T46; } { E TN, TQ, T42, T43; TN = TL + TM; TQ = TO + TP; TR = TN + TQ; T3b = TQ - TN; T42 = Ty - TB; T43 = TL - TM; T44 = T42 - T43; T5o = T42 + T43; } } { E TW, T1a, TV, T19, T10, T16, T13, T17, TT, TU; TW = R0[WS(rs, 31)]; T1a = R0[WS(rs, 15)]; TT = R0[WS(rs, 7)]; TU = R0[WS(rs, 23)]; TV = KP707106781 * (TT - TU); T19 = KP707106781 * (TT + TU); { E TY, TZ, T11, T12; TY = R0[WS(rs, 3)]; TZ = R0[WS(rs, 19)]; T10 = FNMS(KP382683432, TZ, KP923879532 * TY); T16 = FMA(KP382683432, TY, KP923879532 * TZ); T11 = R0[WS(rs, 11)]; T12 = R0[WS(rs, 27)]; T13 = FNMS(KP923879532, T12, KP382683432 * T11); T17 = FMA(KP923879532, T11, KP382683432 * T12); } { E TX, T14, T4c, T4d; TX = TV - TW; T14 = T10 + T13; T15 = TX + T14; T38 = TX - T14; T4c = T13 - T10; T4d = T1a - T19; T4e = T4c - T4d; T5l = T4c + T4d; } { E T18, T1b, T49, T4a; T18 = T16 + T17; T1b = T19 + T1a; T1c = T18 + T1b; T39 = T1b - T18; T49 = TV + TW; T4a = T16 - T17; T4b = T49 + T4a; T5k = T4a - T49; } } { E T1g, T1U, T1j, T1T, T1n, T1Q, T1q, T1R, T1h, T1i; T1g = R1[0]; T1U = R1[WS(rs, 16)]; T1h = R1[WS(rs, 8)]; T1i = R1[WS(rs, 24)]; T1j = KP707106781 * (T1h - T1i); T1T = KP707106781 * (T1h + T1i); { E T1l, T1m, T1o, T1p; T1l = R1[WS(rs, 4)]; T1m = R1[WS(rs, 20)]; T1n = FNMS(KP382683432, T1m, KP923879532 * T1l); T1Q = FMA(KP382683432, T1l, KP923879532 * T1m); T1o = R1[WS(rs, 12)]; T1p = R1[WS(rs, 28)]; T1q = FNMS(KP923879532, T1p, KP382683432 * T1o); T1R = FMA(KP923879532, T1o, KP382683432 * T1p); } { E T1k, T1r, T4t, T4u; T1k = T1g + T1j; T1r = T1n + T1q; T1s = T1k + T1r; T3g = T1k - T1r; T4t = T1q - T1n; T4u = T1U - T1T; T4v = T4t - T4u; T5w = T4t + T4u; } { E T1S, T1V, T4i, T4j; T1S = T1Q + T1R; T1V = T1T + T1U; T1W = T1S + T1V; T3k = T1V - T1S; T4i = T1g - T1j; T4j = T1Q - T1R; T4k = T4i - T4j; T5s = T4i + T4j; } } { E T2g, T4F, T2j, T4E, T2p, T4C, T2s, T4B; { E T2c, T2i, T2f, T2h, T2d, T2e; T2c = R1[WS(rs, 1)]; T2i = R1[WS(rs, 17)]; T2d = R1[WS(rs, 9)]; T2e = R1[WS(rs, 25)]; T2f = KP707106781 * (T2d - T2e); T2h = KP707106781 * (T2d + T2e); T2g = T2c + T2f; T4F = T2c - T2f; T2j = T2h + T2i; T4E = T2i - T2h; } { E T2o, T2r, T2n, T2q, T2l, T2m; T2o = R1[WS(rs, 29)]; T2r = R1[WS(rs, 13)]; T2l = R1[WS(rs, 5)]; T2m = R1[WS(rs, 21)]; T2n = KP707106781 * (T2l - T2m); T2q = KP707106781 * (T2l + T2m); T2p = T2n - T2o; T4C = T2n + T2o; T2s = T2q + T2r; T4B = T2r - T2q; } { E T2k, T2t, T4P, T4Q; T2k = FNMS(KP195090322, T2j, KP980785280 * T2g); T2t = FMA(KP980785280, T2p, KP195090322 * T2s); T2u = T2k + T2t; T3q = T2t - T2k; T4P = FMA(KP831469612, T4F, KP555570233 * T4E); T4Q = FMA(KP831469612, T4C, KP555570233 * T4B); T4R = T4P + T4Q; T5A = T4P - T4Q; } { E T2w, T2x, T4D, T4G; T2w = FNMS(KP980785280, T2s, KP195090322 * T2p); T2x = FMA(KP195090322, T2g, KP980785280 * T2j); T2y = T2w - T2x; T3o = T2x + T2w; T4D = FNMS(KP555570233, T4C, KP831469612 * T4B); T4G = FNMS(KP555570233, T4F, KP831469612 * T4E); T4H = T4D - T4G; T5C = T4G + T4D; } } { E T1x, T4p, T1A, T4o, T1G, T4m, T1J, T4l; { E T1t, T1z, T1w, T1y, T1u, T1v; T1t = R1[WS(rs, 2)]; T1z = R1[WS(rs, 18)]; T1u = R1[WS(rs, 10)]; T1v = R1[WS(rs, 26)]; T1w = KP707106781 * (T1u - T1v); T1y = KP707106781 * (T1u + T1v); T1x = T1t + T1w; T4p = T1t - T1w; T1A = T1y + T1z; T4o = T1z - T1y; } { E T1F, T1I, T1E, T1H, T1C, T1D; T1F = R1[WS(rs, 30)]; T1I = R1[WS(rs, 14)]; T1C = R1[WS(rs, 6)]; T1D = R1[WS(rs, 22)]; T1E = KP707106781 * (T1C - T1D); T1H = KP707106781 * (T1C + T1D); T1G = T1E - T1F; T4m = T1E + T1F; T1J = T1H + T1I; T4l = T1I - T1H; } { E T1B, T1K, T4w, T4x; T1B = FNMS(KP195090322, T1A, KP980785280 * T1x); T1K = FMA(KP980785280, T1G, KP195090322 * T1J); T1L = T1B + T1K; T3j = T1K - T1B; T4w = FMA(KP831469612, T4p, KP555570233 * T4o); T4x = FMA(KP831469612, T4m, KP555570233 * T4l); T4y = T4w + T4x; T5t = T4w - T4x; } { E T1N, T1O, T4n, T4q; T1N = FNMS(KP980785280, T1J, KP195090322 * T1G); T1O = FMA(KP195090322, T1x, KP980785280 * T1A); T1P = T1N - T1O; T3h = T1O + T1N; T4n = FNMS(KP555570233, T4m, KP831469612 * T4l); T4q = FNMS(KP555570233, T4p, KP831469612 * T4o); T4r = T4n - T4q; T5v = T4q + T4n; } } { E Tx, T2N, T69, T6f, T1e, T6e, T2X, T30, T1Y, T2L, T2Q, T62, T2U, T31, T2H; E T2K, Tw, T63; Tw = Tm + Tv; Tx = Td + Tw; T2N = Td - Tw; T63 = T35 - T34; T69 = T63 - T68; T6f = T63 + T68; { E TS, T1d, T2V, T2W; TS = FNMS(KP098017140, TR, KP995184726 * TK); T1d = FMA(KP995184726, T15, KP098017140 * T1c); T1e = TS + T1d; T6e = T1d - TS; T2V = T2b - T2u; T2W = T2y + T2F; T2X = FNMS(KP671558954, T2W, KP740951125 * T2V); T30 = FMA(KP671558954, T2V, KP740951125 * T2W); } { E T1M, T1X, T2O, T2P; T1M = T1s + T1L; T1X = T1P - T1W; T1Y = FMA(KP998795456, T1M, KP049067674 * T1X); T2L = FNMS(KP049067674, T1M, KP998795456 * T1X); T2O = FMA(KP098017140, TK, KP995184726 * TR); T2P = FNMS(KP995184726, T1c, KP098017140 * T15); T2Q = T2O + T2P; T62 = T2P - T2O; } { E T2S, T2T, T2v, T2G; T2S = T1s - T1L; T2T = T1P + T1W; T2U = FMA(KP740951125, T2S, KP671558954 * T2T); T31 = FNMS(KP671558954, T2S, KP740951125 * T2T); T2v = T2b + T2u; T2G = T2y - T2F; T2H = FNMS(KP049067674, T2G, KP998795456 * T2v); T2K = FMA(KP049067674, T2v, KP998795456 * T2G); } { E T1f, T2I, T6b, T6c; T1f = Tx + T1e; T2I = T1Y + T2H; Cr[WS(csr, 31)] = T1f - T2I; Cr[0] = T1f + T2I; T6b = T2L + T2K; T6c = T62 + T69; Ci[WS(csi, 31)] = T6b - T6c; Ci[0] = T6b + T6c; } { E T2J, T2M, T61, T6a; T2J = Tx - T1e; T2M = T2K - T2L; Cr[WS(csr, 16)] = T2J - T2M; Cr[WS(csr, 15)] = T2J + T2M; T61 = T2H - T1Y; T6a = T62 - T69; Ci[WS(csi, 16)] = T61 - T6a; Ci[WS(csi, 15)] = T61 + T6a; } { E T2R, T2Y, T6h, T6i; T2R = T2N + T2Q; T2Y = T2U + T2X; Cr[WS(csr, 24)] = T2R - T2Y; Cr[WS(csr, 7)] = T2R + T2Y; T6h = T31 + T30; T6i = T6e + T6f; Ci[WS(csi, 24)] = T6h - T6i; Ci[WS(csi, 7)] = T6h + T6i; } { E T2Z, T32, T6d, T6g; T2Z = T2N - T2Q; T32 = T30 - T31; Cr[WS(csr, 23)] = T2Z - T32; Cr[WS(csr, 8)] = T2Z + T32; T6d = T2X - T2U; T6g = T6e - T6f; Ci[WS(csi, 23)] = T6d - T6g; Ci[WS(csi, 8)] = T6d + T6g; } } { E T5j, T5L, T6R, T6X, T5q, T6W, T5V, T5Y, T5y, T5J, T5O, T6O, T5S, T5Z, T5F; E T5I, T5i, T6P; T5i = T5g - T5h; T5j = T5f - T5i; T5L = T5f + T5i; T6P = T3Z + T3W; T6R = T6P - T6Q; T6X = T6P + T6Q; { E T5m, T5p, T5T, T5U; T5m = FMA(KP290284677, T5k, KP956940335 * T5l); T5p = FNMS(KP290284677, T5o, KP956940335 * T5n); T5q = T5m - T5p; T6W = T5p + T5m; T5T = T5z + T5A; T5U = T5C + T5D; T5V = FNMS(KP146730474, T5U, KP989176509 * T5T); T5Y = FMA(KP146730474, T5T, KP989176509 * T5U); } { E T5u, T5x, T5M, T5N; T5u = T5s - T5t; T5x = T5v - T5w; T5y = FMA(KP803207531, T5u, KP595699304 * T5x); T5J = FNMS(KP595699304, T5u, KP803207531 * T5x); T5M = FMA(KP956940335, T5o, KP290284677 * T5n); T5N = FNMS(KP290284677, T5l, KP956940335 * T5k); T5O = T5M + T5N; T6O = T5N - T5M; } { E T5Q, T5R, T5B, T5E; T5Q = T5s + T5t; T5R = T5v + T5w; T5S = FMA(KP989176509, T5Q, KP146730474 * T5R); T5Z = FNMS(KP146730474, T5Q, KP989176509 * T5R); T5B = T5z - T5A; T5E = T5C - T5D; T5F = FNMS(KP595699304, T5E, KP803207531 * T5B); T5I = FMA(KP595699304, T5B, KP803207531 * T5E); } { E T5r, T5G, T6T, T6U; T5r = T5j + T5q; T5G = T5y + T5F; Cr[WS(csr, 25)] = T5r - T5G; Cr[WS(csr, 6)] = T5r + T5G; T6T = T5J + T5I; T6U = T6O + T6R; Ci[WS(csi, 25)] = T6T - T6U; Ci[WS(csi, 6)] = T6T + T6U; } { E T5H, T5K, T6N, T6S; T5H = T5j - T5q; T5K = T5I - T5J; Cr[WS(csr, 22)] = T5H - T5K; Cr[WS(csr, 9)] = T5H + T5K; T6N = T5F - T5y; T6S = T6O - T6R; Ci[WS(csi, 22)] = T6N - T6S; Ci[WS(csi, 9)] = T6N + T6S; } { E T5P, T5W, T6Z, T70; T5P = T5L + T5O; T5W = T5S + T5V; Cr[WS(csr, 30)] = T5P - T5W; Cr[WS(csr, 1)] = T5P + T5W; T6Z = T5Z + T5Y; T70 = T6W + T6X; Ci[WS(csi, 30)] = T6Z - T70; Ci[WS(csi, 1)] = T6Z + T70; } { E T5X, T60, T6V, T6Y; T5X = T5L - T5O; T60 = T5Y - T5Z; Cr[WS(csr, 17)] = T5X - T60; Cr[WS(csr, 14)] = T5X + T60; T6V = T5V - T5S; T6Y = T6W - T6X; Ci[WS(csi, 17)] = T6V - T6Y; Ci[WS(csi, 14)] = T6V + T6Y; } } { E T37, T3z, T6n, T6t, T3e, T6s, T3J, T3M, T3m, T3x, T3C, T6k, T3G, T3N, T3t; E T3w, T36, T6l; T36 = T34 + T35; T37 = T33 - T36; T3z = T33 + T36; T6l = Tv - Tm; T6n = T6l - T6m; T6t = T6l + T6m; { E T3a, T3d, T3H, T3I; T3a = FMA(KP634393284, T38, KP773010453 * T39); T3d = FNMS(KP634393284, T3c, KP773010453 * T3b); T3e = T3a - T3d; T6s = T3d + T3a; T3H = T3n + T3o; T3I = T3q + T3r; T3J = FNMS(KP336889853, T3I, KP941544065 * T3H); T3M = FMA(KP336889853, T3H, KP941544065 * T3I); } { E T3i, T3l, T3A, T3B; T3i = T3g - T3h; T3l = T3j - T3k; T3m = FMA(KP903989293, T3i, KP427555093 * T3l); T3x = FNMS(KP427555093, T3i, KP903989293 * T3l); T3A = FMA(KP773010453, T3c, KP634393284 * T3b); T3B = FNMS(KP634393284, T39, KP773010453 * T38); T3C = T3A + T3B; T6k = T3B - T3A; } { E T3E, T3F, T3p, T3s; T3E = T3g + T3h; T3F = T3j + T3k; T3G = FMA(KP941544065, T3E, KP336889853 * T3F); T3N = FNMS(KP336889853, T3E, KP941544065 * T3F); T3p = T3n - T3o; T3s = T3q - T3r; T3t = FNMS(KP427555093, T3s, KP903989293 * T3p); T3w = FMA(KP427555093, T3p, KP903989293 * T3s); } { E T3f, T3u, T6p, T6q; T3f = T37 + T3e; T3u = T3m + T3t; Cr[WS(csr, 27)] = T3f - T3u; Cr[WS(csr, 4)] = T3f + T3u; T6p = T3x + T3w; T6q = T6k + T6n; Ci[WS(csi, 27)] = T6p - T6q; Ci[WS(csi, 4)] = T6p + T6q; } { E T3v, T3y, T6j, T6o; T3v = T37 - T3e; T3y = T3w - T3x; Cr[WS(csr, 20)] = T3v - T3y; Cr[WS(csr, 11)] = T3v + T3y; T6j = T3t - T3m; T6o = T6k - T6n; Ci[WS(csi, 20)] = T6j - T6o; Ci[WS(csi, 11)] = T6j + T6o; } { E T3D, T3K, T6v, T6w; T3D = T3z + T3C; T3K = T3G + T3J; Cr[WS(csr, 28)] = T3D - T3K; Cr[WS(csr, 3)] = T3D + T3K; T6v = T3N + T3M; T6w = T6s + T6t; Ci[WS(csi, 28)] = T6v - T6w; Ci[WS(csi, 3)] = T6v + T6w; } { E T3L, T3O, T6r, T6u; T3L = T3z - T3C; T3O = T3M - T3N; Cr[WS(csr, 19)] = T3L - T3O; Cr[WS(csr, 12)] = T3L + T3O; T6r = T3J - T3G; T6u = T6s - T6t; Ci[WS(csi, 19)] = T6r - T6u; Ci[WS(csi, 12)] = T6r + T6u; } } { E T41, T4Z, T6D, T6J, T4g, T6I, T59, T5d, T4A, T4X, T52, T6y, T56, T5c, T4T; E T4W, T40, T6C; T40 = T3W - T3Z; T41 = T3T + T40; T4Z = T3T - T40; T6C = T5g + T5h; T6D = T6B - T6C; T6J = T6C + T6B; { E T48, T4f, T57, T58; T48 = FMA(KP881921264, T44, KP471396736 * T47); T4f = FMA(KP881921264, T4b, KP471396736 * T4e); T4g = T48 - T4f; T6I = T48 + T4f; T57 = T4K + T4H; T58 = T4R + T4O; T59 = FMA(KP514102744, T57, KP857728610 * T58); T5d = FNMS(KP857728610, T57, KP514102744 * T58); } { E T4s, T4z, T50, T51; T4s = T4k + T4r; T4z = T4v - T4y; T4A = FMA(KP970031253, T4s, KP242980179 * T4z); T4X = FNMS(KP242980179, T4s, KP970031253 * T4z); T50 = FNMS(KP471396736, T4b, KP881921264 * T4e); T51 = FNMS(KP471396736, T44, KP881921264 * T47); T52 = T50 - T51; T6y = T51 + T50; } { E T54, T55, T4L, T4S; T54 = T4k - T4r; T55 = T4y + T4v; T56 = FMA(KP514102744, T54, KP857728610 * T55); T5c = FNMS(KP514102744, T55, KP857728610 * T54); T4L = T4H - T4K; T4S = T4O - T4R; T4T = FNMS(KP242980179, T4S, KP970031253 * T4L); T4W = FMA(KP242980179, T4L, KP970031253 * T4S); } { E T4h, T4U, T6F, T6G; T4h = T41 + T4g; T4U = T4A + T4T; Cr[WS(csr, 29)] = T4h - T4U; Cr[WS(csr, 2)] = T4h + T4U; T6F = T4X + T4W; T6G = T6y + T6D; Ci[WS(csi, 29)] = T6F - T6G; Ci[WS(csi, 2)] = T6F + T6G; } { E T4V, T4Y, T6x, T6E; T4V = T41 - T4g; T4Y = T4W - T4X; Cr[WS(csr, 18)] = T4V - T4Y; Cr[WS(csr, 13)] = T4V + T4Y; T6x = T4T - T4A; T6E = T6y - T6D; Ci[WS(csi, 18)] = T6x - T6E; Ci[WS(csi, 13)] = T6x + T6E; } { E T53, T5a, T6L, T6M; T53 = T4Z - T52; T5a = T56 - T59; Cr[WS(csr, 21)] = T53 - T5a; Cr[WS(csr, 10)] = T53 + T5a; T6L = T5d - T5c; T6M = T6J - T6I; Ci[WS(csi, 21)] = T6L - T6M; Ci[WS(csi, 10)] = T6L + T6M; } { E T5b, T5e, T6H, T6K; T5b = T4Z + T52; T5e = T5c + T5d; Cr[WS(csr, 26)] = T5b - T5e; Cr[WS(csr, 5)] = T5b + T5e; T6H = T56 + T59; T6K = T6I + T6J; Ci[WS(csi, 5)] = -(T6H + T6K); Ci[WS(csi, 26)] = T6K - T6H; } } } } } static const kr2c_desc desc = { 64, "r2cfII_64", {342, 114, 92, 0}, &GENUS }; void X(codelet_r2cfII_64) (planner *p) { X(kr2c_register) (p, r2cfII_64, &desc); } #endif /* HAVE_FMA */ fftw-3.3.3/rdft/scalar/r2cf/r2cfII_20.c0000644000175000017500000003123112054410670014171 00000000000000/* * Copyright (c) 2003, 2007-11 Matteo Frigo * Copyright (c) 2003, 2007-11 Massachusetts Institute of Technology * * This program is free software; you can redistribute it and/or modify * it under the terms of the GNU General Public License as published by * the Free Software Foundation; either version 2 of the License, or * (at your option) any later version. * * This program is distributed in the hope that it will be useful, * but WITHOUT ANY WARRANTY; without even the implied warranty of * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the * GNU General Public License for more details. * * You should have received a copy of the GNU General Public License * along with this program; if not, write to the Free Software * Foundation, Inc., 51 Franklin Street, Fifth Floor, Boston, MA 02110-1301 USA * */ /* This file was automatically generated --- DO NOT EDIT */ /* Generated on Sun Nov 25 07:40:23 EST 2012 */ #include "codelet-rdft.h" #ifdef HAVE_FMA /* Generated by: ../../../genfft/gen_r2cf.native -fma -reorder-insns -schedule-for-pipeline -compact -variables 4 -pipeline-latency 4 -n 20 -name r2cfII_20 -dft-II -include r2cfII.h */ /* * This function contains 102 FP additions, 63 FP multiplications, * (or, 39 additions, 0 multiplications, 63 fused multiply/add), * 67 stack variables, 10 constants, and 40 memory accesses */ #include "r2cfII.h" static void r2cfII_20(R *R0, R *R1, R *Cr, R *Ci, stride rs, stride csr, stride csi, INT v, INT ivs, INT ovs) { DK(KP707106781, +0.707106781186547524400844362104849039284835938); DK(KP951056516, +0.951056516295153572116439333379382143405698634); DK(KP559016994, +0.559016994374947424102293417182819058860154590); DK(KP690983005, +0.690983005625052575897706582817180941139845410); DK(KP552786404, +0.552786404500042060718165266253744752911876328); DK(KP447213595, +0.447213595499957939281834733746255247088123672); DK(KP809016994, +0.809016994374947424102293417182819058860154590); DK(KP250000000, +0.250000000000000000000000000000000000000000000); DK(KP618033988, +0.618033988749894848204586834365638117720309180); DK(KP381966011, +0.381966011250105151795413165634361882279690820); { INT i; for (i = v; i > 0; i = i - 1, R0 = R0 + ivs, R1 = R1 + ivs, Cr = Cr + ovs, Ci = Ci + ovs, MAKE_VOLATILE_STRIDE(80, rs), MAKE_VOLATILE_STRIDE(80, csr), MAKE_VOLATILE_STRIDE(80, csi)) { E Tv, TK, TN, Th, T1l, T1n, Ts, TH; { E Ti, T1d, T1f, T1e, T1g, T1p, TS, Tg, To, T8, T7, T19, T1r, T1k, Tx; E Tp, TX, Ty, TF, Tr, TV, Tz, TA, TI; { E Ta, Tb, Td, Te; Ti = R1[WS(rs, 2)]; T1d = R0[WS(rs, 5)]; Ta = R0[WS(rs, 9)]; Tb = R0[WS(rs, 1)]; Td = R0[WS(rs, 3)]; Te = R0[WS(rs, 7)]; { E T1, T2, T5, T3, T4, T1i, Tc, Tf; T1 = R0[0]; T1f = Ta + Tb; Tc = Ta - Tb; T1e = Td + Te; Tf = Td - Te; T2 = R0[WS(rs, 4)]; T5 = R0[WS(rs, 6)]; T1g = FMA(KP381966011, T1f, T1e); T1p = FMA(KP381966011, T1e, T1f); TS = FMA(KP618033988, Tc, Tf); Tg = FNMS(KP618033988, Tf, Tc); T3 = R0[WS(rs, 8)]; T4 = R0[WS(rs, 2)]; T1i = T2 + T5; { E Tj, Tu, Tm, Tt, Tn, Tq, TU; Tj = R1[WS(rs, 8)]; To = R1[WS(rs, 6)]; { E T6, T1j, Tk, Tl; T6 = T2 + T3 - T4 - T5; T8 = (T3 + T5 - T2) - T4; T1j = T3 + T4; Tk = R1[0]; Tl = R1[WS(rs, 4)]; T7 = FNMS(KP250000000, T6, T1); T19 = T1 + T6; T1r = FNMS(KP618033988, T1i, T1j); T1k = FMA(KP618033988, T1j, T1i); Tu = Tk - Tl; Tm = Tk + Tl; } Tt = To + Tj; Tx = R1[WS(rs, 7)]; Tn = Tj - Tm; Tp = Tj + Tm; Tv = FNMS(KP618033988, Tu, Tt); TX = FMA(KP618033988, Tt, Tu); Tq = FMA(KP809016994, Tp, To); TU = FMA(KP447213595, Tp, Tn); Ty = R1[WS(rs, 1)]; TF = R1[WS(rs, 3)]; Tr = FNMS(KP552786404, Tq, Tn); TV = FNMS(KP690983005, TU, To); Tz = R1[WS(rs, 5)]; TA = R1[WS(rs, 9)]; TI = TF + Ty; } } } { E T1w, TJ, TB, T1a; T1w = T1f + T1d - T1e; TJ = Tz - TA; TB = Tz + TA; T1a = Ti + To - Tp; { E T9, T12, TT, T15, TG, TD, T1s, T1u, TW, T11, T10, T1h; { E TE, TC, TR, T1b; T9 = FNMS(KP559016994, T8, T7); TR = FMA(KP559016994, T8, T7); TK = FMA(KP618033988, TJ, TI); T12 = FNMS(KP618033988, TI, TJ); TE = Ty - TB; TC = Ty + TB; TT = FMA(KP951056516, TS, TR); T15 = FNMS(KP951056516, TS, TR); TG = FNMS(KP552786404, TF, TE); T1b = TC - TF - Tx; { E TZ, T1q, T1c, T1x; TZ = FMA(KP447213595, TC, TE); TD = FMA(KP250000000, TC, Tx); T1q = FNMS(KP809016994, T1p, T1d); T1c = T1a + T1b; T1x = T1a - T1b; T10 = FNMS(KP690983005, TZ, TF); T1s = FNMS(KP951056516, T1r, T1q); T1u = FMA(KP951056516, T1r, T1q); Ci[WS(csi, 7)] = FMA(KP707106781, T1x, T1w); Ci[WS(csi, 2)] = FMS(KP707106781, T1x, T1w); Cr[WS(csr, 7)] = FMA(KP707106781, T1c, T19); Cr[WS(csr, 2)] = FNMS(KP707106781, T1c, T19); } } TW = FNMS(KP809016994, TV, Ti); T11 = FNMS(KP809016994, T10, Tx); T1h = FMA(KP809016994, T1g, T1d); { E T17, TY, T16, T13; T17 = FNMS(KP951056516, TX, TW); TY = FMA(KP951056516, TX, TW); T16 = FMA(KP951056516, T12, T11); T13 = FNMS(KP951056516, T12, T11); TN = FMA(KP951056516, Tg, T9); Th = FNMS(KP951056516, Tg, T9); { E T18, T1v, T1t, T14; T18 = T16 - T17; T1v = T17 + T16; T1t = TY + T13; T14 = TY - T13; Cr[WS(csr, 1)] = FMA(KP707106781, T18, T15); Cr[WS(csr, 8)] = FNMS(KP707106781, T18, T15); Ci[WS(csi, 3)] = FMA(KP707106781, T1v, T1u); Ci[WS(csi, 6)] = FMS(KP707106781, T1v, T1u); Ci[WS(csi, 1)] = FNMS(KP707106781, T1t, T1s); Ci[WS(csi, 8)] = -(FMA(KP707106781, T1t, T1s)); Cr[WS(csr, 3)] = FMA(KP707106781, T14, TT); Cr[WS(csr, 6)] = FNMS(KP707106781, T14, TT); T1l = FMA(KP951056516, T1k, T1h); T1n = FNMS(KP951056516, T1k, T1h); } } Ts = FNMS(KP559016994, Tr, Ti); TH = FNMS(KP559016994, TG, TD); } } } { E TO, Tw, TP, TL; TO = FMA(KP951056516, Tv, Ts); Tw = FNMS(KP951056516, Tv, Ts); TP = FMA(KP951056516, TK, TH); TL = FNMS(KP951056516, TK, TH); { E TQ, T1m, T1o, TM; TQ = TO - TP; T1m = TO + TP; T1o = Tw + TL; TM = Tw - TL; Cr[WS(csr, 4)] = FMA(KP707106781, TQ, TN); Cr[WS(csr, 5)] = FNMS(KP707106781, TQ, TN); Ci[WS(csi, 9)] = FNMS(KP707106781, T1m, T1l); Ci[0] = -(FMA(KP707106781, T1m, T1l)); Ci[WS(csi, 5)] = FNMS(KP707106781, T1o, T1n); Ci[WS(csi, 4)] = -(FMA(KP707106781, T1o, T1n)); Cr[0] = FMA(KP707106781, TM, Th); Cr[WS(csr, 9)] = FNMS(KP707106781, TM, Th); } } } } } static const kr2c_desc desc = { 20, "r2cfII_20", {39, 0, 63, 0}, &GENUS }; void X(codelet_r2cfII_20) (planner *p) { X(kr2c_register) (p, r2cfII_20, &desc); } #else /* HAVE_FMA */ /* Generated by: ../../../genfft/gen_r2cf.native -compact -variables 4 -pipeline-latency 4 -n 20 -name r2cfII_20 -dft-II -include r2cfII.h */ /* * This function contains 102 FP additions, 34 FP multiplications, * (or, 86 additions, 18 multiplications, 16 fused multiply/add), * 60 stack variables, 13 constants, and 40 memory accesses */ #include "r2cfII.h" static void r2cfII_20(R *R0, R *R1, R *Cr, R *Ci, stride rs, stride csr, stride csi, INT v, INT ivs, INT ovs) { DK(KP572061402, +0.572061402817684297600072783580302076536153377); DK(KP218508012, +0.218508012224410535399650602527877556893735408); DK(KP309016994, +0.309016994374947424102293417182819058860154590); DK(KP809016994, +0.809016994374947424102293417182819058860154590); DK(KP559016994, +0.559016994374947424102293417182819058860154590); DK(KP951056516, +0.951056516295153572116439333379382143405698634); DK(KP587785252, +0.587785252292473129168705954639072768597652438); DK(KP250000000, +0.250000000000000000000000000000000000000000000); DK(KP176776695, +0.176776695296636881100211090526212259821208984); DK(KP395284707, +0.395284707521047416499861693054089816714944392); DK(KP672498511, +0.672498511963957326960058968885748755876783111); DK(KP415626937, +0.415626937777453428589967464113135184222253485); DK(KP707106781, +0.707106781186547524400844362104849039284835938); { INT i; for (i = v; i > 0; i = i - 1, R0 = R0 + ivs, R1 = R1 + ivs, Cr = Cr + ovs, Ci = Ci + ovs, MAKE_VOLATILE_STRIDE(80, rs), MAKE_VOLATILE_STRIDE(80, csr), MAKE_VOLATILE_STRIDE(80, csi)) { E T8, TD, Tm, TN, T9, TC, TY, TE, Te, TF, Tl, TK, T12, TL, Tk; E TM, T1, T6, Tq, T1l, T1c, Tp, T1f, T1e, T1d, Ty, TW, T1g, T1m, Tx; E Tu; T8 = R1[WS(rs, 2)]; TD = KP707106781 * T8; Tm = R1[WS(rs, 7)]; TN = KP707106781 * Tm; { E Ta, TA, Td, TB, Tb, Tc; T9 = R1[WS(rs, 6)]; Ta = R1[WS(rs, 8)]; TA = T9 + Ta; Tb = R1[0]; Tc = R1[WS(rs, 4)]; Td = Tb + Tc; TB = Tb - Tc; TC = FMA(KP415626937, TA, KP672498511 * TB); TY = FNMS(KP415626937, TB, KP672498511 * TA); TE = KP395284707 * (Ta - Td); Te = Ta + Td; TF = KP176776695 * Te; } { E Tg, TJ, Tj, TI, Th, Ti; Tg = R1[WS(rs, 1)]; Tl = R1[WS(rs, 3)]; TJ = Tg + Tl; Th = R1[WS(rs, 5)]; Ti = R1[WS(rs, 9)]; Tj = Th + Ti; TI = Th - Ti; TK = FNMS(KP415626937, TJ, KP672498511 * TI); T12 = FMA(KP415626937, TI, KP672498511 * TJ); TL = KP395284707 * (Tg - Tj); Tk = Tg + Tj; TM = KP176776695 * Tk; } { E T2, T5, T3, T4, T1a, T1b; T1 = R0[0]; T2 = R0[WS(rs, 6)]; T5 = R0[WS(rs, 8)]; T3 = R0[WS(rs, 2)]; T4 = R0[WS(rs, 4)]; T1a = T4 + T2; T1b = T5 + T3; T6 = T2 + T3 - (T4 + T5); Tq = FMA(KP250000000, T6, T1); T1l = FNMS(KP951056516, T1b, KP587785252 * T1a); T1c = FMA(KP951056516, T1a, KP587785252 * T1b); Tp = KP559016994 * (T5 + T2 - (T4 + T3)); } T1f = R0[WS(rs, 5)]; { E Tv, Tw, Ts, Tt; Tv = R0[WS(rs, 9)]; Tw = R0[WS(rs, 1)]; Tx = Tv - Tw; T1e = Tv + Tw; Ts = R0[WS(rs, 3)]; Tt = R0[WS(rs, 7)]; Tu = Ts - Tt; T1d = Ts + Tt; } Ty = FMA(KP951056516, Tu, KP587785252 * Tx); TW = FNMS(KP951056516, Tx, KP587785252 * Tu); T1g = FMA(KP809016994, T1d, KP309016994 * T1e) + T1f; T1m = FNMS(KP809016994, T1e, T1f) - (KP309016994 * T1d); { E T7, T1r, To, T1q, Tf, Tn; T7 = T1 - T6; T1r = T1e + T1f - T1d; Tf = T8 + (T9 - Te); Tn = (Tk - Tl) - Tm; To = KP707106781 * (Tf + Tn); T1q = KP707106781 * (Tf - Tn); Cr[WS(csr, 2)] = T7 - To; Ci[WS(csi, 2)] = T1q - T1r; Cr[WS(csr, 7)] = T7 + To; Ci[WS(csi, 7)] = T1q + T1r; } { E T1h, T1j, TX, T15, T10, T16, T13, T17, TV, TZ, T11; T1h = T1c - T1g; T1j = T1c + T1g; TV = Tq - Tp; TX = TV - TW; T15 = TV + TW; TZ = FMA(KP218508012, T9, TD) + TF - TE; T10 = TY + TZ; T16 = TZ - TY; T11 = FNMS(KP218508012, Tl, TL) - (TM + TN); T13 = T11 - T12; T17 = T11 + T12; { E T14, T19, T18, T1i; T14 = T10 + T13; Cr[WS(csr, 5)] = TX - T14; Cr[WS(csr, 4)] = TX + T14; T19 = T17 - T16; Ci[WS(csi, 5)] = T19 - T1h; Ci[WS(csi, 4)] = T19 + T1h; T18 = T16 + T17; Cr[WS(csr, 9)] = T15 - T18; Cr[0] = T15 + T18; T1i = T13 - T10; Ci[0] = T1i - T1j; Ci[WS(csi, 9)] = T1i + T1j; } } { E T1n, T1p, Tz, TR, TH, TS, TP, TT, Tr, TG, TO; T1n = T1l + T1m; T1p = T1m - T1l; Tr = Tp + Tq; Tz = Tr + Ty; TR = Tr - Ty; TG = TD + TE + FNMS(KP572061402, T9, TF); TH = TC + TG; TS = TC - TG; TO = TL + TM + FNMS(KP572061402, Tl, TN); TP = TK - TO; TT = TK + TO; { E TQ, T1o, TU, T1k; TQ = TH + TP; Cr[WS(csr, 6)] = Tz - TQ; Cr[WS(csr, 3)] = Tz + TQ; T1o = TT - TS; Ci[WS(csi, 6)] = T1o - T1p; Ci[WS(csi, 3)] = T1o + T1p; TU = TS + TT; Cr[WS(csr, 8)] = TR - TU; Cr[WS(csr, 1)] = TR + TU; T1k = TP - TH; Ci[WS(csi, 8)] = T1k - T1n; Ci[WS(csi, 1)] = T1k + T1n; } } } } } static const kr2c_desc desc = { 20, "r2cfII_20", {86, 18, 16, 0}, &GENUS }; void X(codelet_r2cfII_20) (planner *p) { X(kr2c_register) (p, r2cfII_20, &desc); } #endif /* HAVE_FMA */ fftw-3.3.3/rdft/scalar/r2cf/r2cfII_25.c0000644000175000017500000007176012054410676014217 00000000000000/* * Copyright (c) 2003, 2007-11 Matteo Frigo * Copyright (c) 2003, 2007-11 Massachusetts Institute of Technology * * This program is free software; you can redistribute it and/or modify * it under the terms of the GNU General Public License as published by * the Free Software Foundation; either version 2 of the License, or * (at your option) any later version. * * This program is distributed in the hope that it will be useful, * but WITHOUT ANY WARRANTY; without even the implied warranty of * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the * GNU General Public License for more details. * * You should have received a copy of the GNU General Public License * along with this program; if not, write to the Free Software * Foundation, Inc., 51 Franklin Street, Fifth Floor, Boston, MA 02110-1301 USA * */ /* This file was automatically generated --- DO NOT EDIT */ /* Generated on Sun Nov 25 07:40:24 EST 2012 */ #include "codelet-rdft.h" #ifdef HAVE_FMA /* Generated by: ../../../genfft/gen_r2cf.native -fma -reorder-insns -schedule-for-pipeline -compact -variables 4 -pipeline-latency 4 -n 25 -name r2cfII_25 -dft-II -include r2cfII.h */ /* * This function contains 212 FP additions, 177 FP multiplications, * (or, 47 additions, 12 multiplications, 165 fused multiply/add), * 163 stack variables, 67 constants, and 50 memory accesses */ #include "r2cfII.h" static void r2cfII_25(R *R0, R *R1, R *Cr, R *Ci, stride rs, stride csr, stride csi, INT v, INT ivs, INT ovs) { DK(KP876091699, +0.876091699473550838204498029706869638173524346); DK(KP792626838, +0.792626838241819413632131824093538848057784557); DK(KP690668130, +0.690668130712929053565177988380887884042527623); DK(KP809385824, +0.809385824416008241660603814668679683846476688); DK(KP860541664, +0.860541664367944677098261680920518816412804187); DK(KP681693190, +0.681693190061530575150324149145440022633095390); DK(KP560319534, +0.560319534973832390111614715371676131169633784); DK(KP237294955, +0.237294955877110315393888866460840817927895961); DK(KP897376177, +0.897376177523557693138608077137219684419427330); DK(KP997675361, +0.997675361079556513670859573984492383596555031); DK(KP584303379, +0.584303379262766050358567120694562180043261496); DK(KP653711795, +0.653711795629256296299985401753308353544378892); DK(KP591287873, +0.591287873858343558732323717242372865934480959); DK(KP645989928, +0.645989928319777763844272876603899665178054552); DK(KP956723877, +0.956723877038460305821989399535483155872969262); DK(KP952936919, +0.952936919628306576880750665357914584765951388); DK(KP998026728, +0.998026728428271561952336806863450553336905220); DK(KP945422727, +0.945422727388575946270360266328811958657216298); DK(KP559154169, +0.559154169276087864842202529084232643714075927); DK(KP683113946, +0.683113946453479238701949862233725244439656928); DK(KP999754674, +0.999754674276473633366203429228112409535557487); DK(KP968583161, +0.968583161128631119490168375464735813836012403); DK(KP242145790, +0.242145790282157779872542093866183953459003101); DK(KP734762448, +0.734762448793050413546343770063151342619912334); DK(KP904730450, +0.904730450839922351881287709692877908104763647); DK(KP876306680, +0.876306680043863587308115903922062583399064238); DK(KP949179823, +0.949179823508441261575555465843363271711583843); DK(KP772036680, +0.772036680810363904029489473607579825330539880); DK(KP669429328, +0.669429328479476605641803240971985825917022098); DK(KP916574801, +0.916574801383451584742370439148878693530976769); DK(KP829049696, +0.829049696159252993975487806364305442437946767); DK(KP923225144, +0.923225144846402650453449441572664695995209956); DK(KP262346850, +0.262346850930607871785420028382979691334784273); DK(KP992114701, +0.992114701314477831049793042785778521453036709); DK(KP803003575, +0.803003575438660414833440593570376004635464850); DK(KP906616052, +0.906616052148196230441134447086066874408359177); DK(KP831864738, +0.831864738706457140726048799369896829771167132); DK(KP763583905, +0.763583905359130246362948588764067237776594106); DK(KP921078979, +0.921078979742360627699756128143719920817673854); DK(KP904508497, +0.904508497187473712051146708591409529430077295); DK(KP248028675, +0.248028675328619457762448260696444630363259177); DK(KP894834959, +0.894834959464455102997960030820114611498661386); DK(KP982009705, +0.982009705009746369461829878184175962711969869); DK(KP845997307, +0.845997307939530944175097360758058292389769300); DK(KP958953096, +0.958953096729998668045963838399037225970891871); DK(KP867381224, +0.867381224396525206773171885031575671309956167); DK(KP912575812, +0.912575812670962425556968549836277086778922727); DK(KP951056516, +0.951056516295153572116439333379382143405698634); DK(KP869845200, +0.869845200362138853122720822420327157933056305); DK(KP786782374, +0.786782374965295178365099601674911834788448471); DK(KP120146378, +0.120146378570687701782758537356596213647956445); DK(KP132830569, +0.132830569247582714407653942074819768844536507); DK(KP269969613, +0.269969613759572083574752974412347470060951301); DK(KP244189809, +0.244189809627953270309879511234821255780225091); DK(KP987388751, +0.987388751065621252324603216482382109400433949); DK(KP893101515, +0.893101515366181661711202267938416198338079437); DK(KP494780565, +0.494780565770515410344588413655324772219443730); DK(KP447533225, +0.447533225982656890041886979663652563063114397); DK(KP522847744, +0.522847744331509716623755382187077770911012542); DK(KP578046249, +0.578046249379945007321754579646815604023525655); DK(KP066152395, +0.066152395967733048213034281011006031460903353); DK(KP059835404, +0.059835404262124915169548397419498386427871950); DK(KP667278218, +0.667278218140296670899089292254759909713898805); DK(KP603558818, +0.603558818296015001454675132653458027918768137); DK(KP559016994, +0.559016994374947424102293417182819058860154590); DK(KP250000000, +0.250000000000000000000000000000000000000000000); DK(KP618033988, +0.618033988749894848204586834365638117720309180); { INT i; for (i = v; i > 0; i = i - 1, R0 = R0 + ivs, R1 = R1 + ivs, Cr = Cr + ovs, Ci = Ci + ovs, MAKE_VOLATILE_STRIDE(100, rs), MAKE_VOLATILE_STRIDE(100, csr), MAKE_VOLATILE_STRIDE(100, csi)) { E T2R, T2T, T2D, T2C, T2H, T2G, T2B, T2P, T2S; { E T2A, TJ, T1K, T3l, T2z, TB, T2d, T2l, T1N, T21, T15, T1g, T1s, T1D, T9; E T25, T1X, T2o, T2g, T1z, T1u, T1j, TQ, Ti, T1a, T2f, T2p, T1U, T24, TX; E T1k, T1v, T1A, T19, Ts, T18, T1P; { E Tt, Tw, TZ, Tx, Ty; { E T2v, TG, TH, TD, TE, TI, T2x; T2v = R0[0]; TG = R0[WS(rs, 10)]; TH = R1[WS(rs, 2)]; TD = R0[WS(rs, 5)]; TE = R1[WS(rs, 7)]; Tt = R0[WS(rs, 2)]; TI = TG + TH; T2x = TG - TH; { E TF, T2w, Tu, Tv, T2y; TF = TD + TE; T2w = TD - TE; Tu = R0[WS(rs, 7)]; Tv = R1[WS(rs, 9)]; T2A = T2w - T2x; T2y = T2w + T2x; TJ = FMA(KP618033988, TI, TF); T1K = FNMS(KP618033988, TF, TI); T3l = T2v + T2y; T2z = FNMS(KP250000000, T2y, T2v); Tw = Tu - Tv; TZ = Tu + Tv; Tx = R0[WS(rs, 12)]; Ty = R1[WS(rs, 4)]; } } { E TO, TN, TM, T1V; { E T1, T1M, T11, T13, T4, TK, T12, TL, T7, T5, TA, T6, T14, T1L, T8; T1 = R0[WS(rs, 1)]; { E T2, T10, Tz, T3; T2 = R0[WS(rs, 6)]; T10 = Tx + Ty; Tz = Tx - Ty; T3 = R1[WS(rs, 8)]; T5 = R0[WS(rs, 11)]; T1M = FNMS(KP618033988, TZ, T10); T11 = FMA(KP618033988, T10, TZ); T13 = Tz - Tw; TA = Tw + Tz; T4 = T2 - T3; TK = T2 + T3; T6 = R1[WS(rs, 3)]; } TB = Tt + TA; T12 = FNMS(KP250000000, TA, Tt); TL = T5 + T6; T7 = T5 - T6; T14 = FNMS(KP559016994, T13, T12); T1L = FMA(KP559016994, T13, T12); T8 = T4 + T7; TO = T4 - T7; T2d = FNMS(KP603558818, T1M, T1L); T2l = FMA(KP667278218, T1L, T1M); T1N = FMA(KP059835404, T1M, T1L); T21 = FNMS(KP066152395, T1L, T1M); T15 = FMA(KP578046249, T14, T11); T1g = FNMS(KP522847744, T11, T14); T1s = FMA(KP447533225, T11, T14); T1D = FNMS(KP494780565, T14, T11); TN = FNMS(KP250000000, T8, T1); T9 = T1 + T8; TM = FMA(KP618033988, TL, TK); T1V = FNMS(KP618033988, TK, TL); } { E Th, Td, TU, Tc, Te; Th = R0[WS(rs, 4)]; { E Ta, Tb, T1W, TP; Ta = R0[WS(rs, 9)]; Tb = R1[WS(rs, 11)]; T1W = FNMS(KP559016994, TO, TN); TP = FMA(KP559016994, TO, TN); Td = R1[WS(rs, 6)]; TU = Ta + Tb; Tc = Ta - Tb; T25 = FNMS(KP893101515, T1V, T1W); T1X = FMA(KP987388751, T1W, T1V); T2o = FMA(KP522847744, T1V, T1W); T2g = FNMS(KP578046249, T1W, T1V); T1z = FMA(KP667278218, TP, TM); T1u = FNMS(KP603558818, TM, TP); T1j = FNMS(KP244189809, TM, TP); TQ = FMA(KP269969613, TP, TM); Te = R1[WS(rs, 1)]; } { E Tk, T1S, TW, TS, Tn, T16, TR, T17, Tq, To, Tg, Tp, TT, T1T, Tr; Tk = R0[WS(rs, 3)]; { E Tl, TV, Tf, Tm; Tl = R0[WS(rs, 8)]; TV = Te - Td; Tf = Td + Te; Tm = R1[WS(rs, 10)]; To = R1[0]; T1S = FMA(KP618033988, TU, TV); TW = FNMS(KP618033988, TV, TU); TS = Tc + Tf; Tg = Tc - Tf; Tn = Tl - Tm; T16 = Tl + Tm; Tp = R1[WS(rs, 5)]; } Ti = Tg + Th; TR = FNMS(KP250000000, Tg, Th); T17 = Tp - To; Tq = To + Tp; TT = FMA(KP559016994, TS, TR); T1T = FNMS(KP559016994, TS, TR); Tr = Tn - Tq; T1a = Tn + Tq; T2f = FNMS(KP447533225, T1S, T1T); T2p = FMA(KP494780565, T1T, T1S); T1U = FMA(KP132830569, T1T, T1S); T24 = FNMS(KP120146378, T1S, T1T); TX = FMA(KP603558818, TW, TT); T1k = FNMS(KP667278218, TT, TW); T1v = FNMS(KP786782374, TW, TT); T1A = FMA(KP869845200, TT, TW); T19 = FNMS(KP250000000, Tr, Tk); Ts = Tk + Tr; T18 = FMA(KP618033988, T17, T16); T1P = FNMS(KP618033988, T16, T17); } } } } { E T22, T1Q, T1h, T1c, T2O, T2N, T2m, T3a, T3b, T2q, T1y, T3f, T2e, T2h, T3e; E T1H, T1J; { E T3m, T3n, T2k, T2c, T1C, T1r; { E Tj, TC, T1O, T1b; T3m = T9 + Ti; Tj = T9 - Ti; TC = Ts - TB; T3n = TB + Ts; T1O = FNMS(KP559016994, T1a, T19); T1b = FMA(KP559016994, T1a, T19); Ci[WS(csi, 7)] = KP951056516 * (FMA(KP618033988, Tj, TC)); Ci[WS(csi, 2)] = -(KP951056516 * (FNMS(KP618033988, TC, Tj))); T22 = FMA(KP869845200, T1O, T1P); T1Q = FNMS(KP786782374, T1P, T1O); T2k = FMA(KP066152395, T1O, T1P); T2c = FNMS(KP059835404, T1P, T1O); T1C = FNMS(KP120146378, T18, T1b); T1r = FMA(KP132830569, T1b, T18); T1h = FNMS(KP893101515, T18, T1b); T1c = FMA(KP987388751, T1b, T18); } { E T1B, T1E, T1t, T3o, T3q, T1w, T3p; T1B = FMA(KP912575812, T1A, T1z); T2O = FNMS(KP912575812, T1A, T1z); T2N = FNMS(KP867381224, T1D, T1C); T1E = FMA(KP867381224, T1D, T1C); T1t = FMA(KP958953096, T1s, T1r); T2R = FNMS(KP958953096, T1s, T1r); T3o = T3m + T3n; T3q = T3m - T3n; T2T = FMA(KP912575812, T1v, T1u); T1w = FNMS(KP912575812, T1v, T1u); T2m = FNMS(KP845997307, T2l, T2k); T3a = FMA(KP845997307, T2l, T2k); T3b = FNMS(KP982009705, T2p, T2o); T2q = FMA(KP982009705, T2p, T2o); T3p = FNMS(KP250000000, T3o, T3l); Cr[WS(csr, 12)] = T3o + T3l; { E T1x, T1F, T1G, T1I; T1x = FMA(KP894834959, T1w, T1t); T1F = FNMS(KP894834959, T1w, T1t); Cr[WS(csr, 7)] = FNMS(KP559016994, T3q, T3p); Cr[WS(csr, 2)] = FMA(KP559016994, T3q, T3p); T1y = FMA(KP248028675, T1x, TJ); T1G = FNMS(KP904508497, T1F, T1E); T1I = FNMS(KP894834959, T1B, T1F); T3f = FNMS(KP845997307, T2d, T2c); T2e = FMA(KP845997307, T2d, T2c); T2h = FNMS(KP921078979, T2g, T2f); T3e = FMA(KP921078979, T2g, T2f); T1H = FMA(KP763583905, T1G, T1B); T1J = FMA(KP559016994, T1I, T1E); } } } { E T1i, T1l, T23, T30, T2Z, T26, T1R, T33, T1f, T1n, T1p, T34, T1Y, T3d, T3k; E T3i; { E T2j, TY, T2s, T2u, T1d, T1m, T1e; T2D = FMA(KP831864738, T1h, T1g); T1i = FNMS(KP831864738, T1h, T1g); { E T2i, T2n, T2r, T2t; T2i = FMA(KP906616052, T2h, T2e); T2n = FNMS(KP906616052, T2h, T2e); Ci[WS(csi, 4)] = KP951056516 * (FNMS(KP803003575, T1H, T1y)); Ci[WS(csi, 9)] = KP951056516 * (FNMS(KP992114701, T1J, T1y)); T2j = FMA(KP262346850, T2i, T1K); T2r = FNMS(KP923225144, T2q, T2n); T2t = T2m + T2n; T2C = FNMS(KP829049696, T1k, T1j); T1l = FMA(KP829049696, T1k, T1j); TY = FMA(KP916574801, TX, TQ); T2H = FNMS(KP916574801, TX, TQ); T2s = FNMS(KP618033988, T2r, T2m); T2u = FNMS(KP669429328, T2t, T2q); T2G = FNMS(KP831864738, T1c, T15); T1d = FMA(KP831864738, T1c, T15); } T23 = FNMS(KP772036680, T22, T21); T30 = FMA(KP772036680, T22, T21); Ci[WS(csi, 8)] = KP951056516 * (FMA(KP949179823, T2s, T2j)); Ci[WS(csi, 3)] = KP951056516 * (FNMS(KP876306680, T2u, T2j)); T1m = FNMS(KP904730450, T1d, TY); T1e = FMA(KP904730450, T1d, TY); T2Z = FNMS(KP734762448, T25, T24); T26 = FMA(KP734762448, T25, T24); T1R = FMA(KP772036680, T1Q, T1N); T33 = FNMS(KP772036680, T1Q, T1N); T1f = FNMS(KP242145790, T1e, TJ); Ci[0] = -(KP951056516 * (FMA(KP968583161, T1e, TJ))); T1n = FNMS(KP904508497, T1m, T1l); T1p = FNMS(KP999754674, T1m, T1i); T34 = FNMS(KP734762448, T1X, T1U); T1Y = FMA(KP734762448, T1X, T1U); } { E T2Y, T31, T38, T36, T3c, T3g; { E T20, T28, T2a, T29, T2b, T35; T2Y = FNMS(KP559016994, T2A, T2z); T2B = FMA(KP559016994, T2A, T2z); { E T1o, T1q, T27, T1Z; T1o = FNMS(KP683113946, T1n, T1i); T1q = FMA(KP559154169, T1p, T1l); T27 = FNMS(KP945422727, T1Y, T1R); T1Z = FMA(KP945422727, T1Y, T1R); Ci[WS(csi, 5)] = -(KP951056516 * (FNMS(KP876306680, T1o, T1f))); Ci[WS(csi, 10)] = -(KP951056516 * (FNMS(KP968583161, T1q, T1f))); T20 = FNMS(KP262346850, T1Z, T1K); Ci[WS(csi, 1)] = -(KP998026728 * (FMA(KP952936919, T1K, T1Z))); T28 = FMA(KP956723877, T27, T26); T2a = T27 - T23; } T29 = FMA(KP645989928, T28, T23); T2b = FMA(KP591287873, T2a, T26); Ci[WS(csi, 6)] = -(KP951056516 * (FMA(KP949179823, T29, T20))); Ci[WS(csi, 11)] = -(KP951056516 * (FNMS(KP992114701, T2b, T20))); T31 = FMA(KP956723877, T30, T2Z); T35 = FNMS(KP956723877, T30, T2Z); T38 = FMA(KP618033988, T35, T34); T36 = T34 + T35; } Cr[WS(csr, 1)] = FNMS(KP992114701, T31, T2Y); T3c = FMA(KP923225144, T3b, T3a); T3g = FNMS(KP923225144, T3b, T3a); { E T32, T37, T3h, T3j, T39; T32 = FMA(KP248028675, T31, T2Y); T39 = FNMS(KP653711795, T33, T38); T37 = FMA(KP584303379, T36, T33); T3h = FNMS(KP904508497, T3g, T3f); T3j = FNMS(KP997675361, T3g, T3e); Cr[WS(csr, 11)] = FNMS(KP897376177, T39, T32); Cr[WS(csr, 6)] = FMA(KP949179823, T37, T32); T3d = FNMS(KP237294955, T3c, T2Y); T3k = FNMS(KP560319534, T3j, T3f); T3i = FMA(KP681693190, T3h, T3e); } } Cr[WS(csr, 8)] = FMA(KP949179823, T3k, T3d); Cr[WS(csr, 3)] = FMA(KP860541664, T3i, T3d); T2P = FNMS(KP809385824, T2O, T2N); T2S = FMA(KP809385824, T2O, T2N); } } } { E T2F, T2K, T2M, T2Q; T2Q = FMA(KP248028675, T2P, T2B); { E T2U, T2W, T2E, T2I; T2U = FNMS(KP894834959, T2T, T2S); T2W = T2R + T2S; T2E = FMA(KP904730450, T2D, T2C); T2I = FNMS(KP904730450, T2D, T2C); { E T2V, T2X, T2J, T2L; T2V = FNMS(KP618033988, T2U, T2R); T2X = FNMS(KP690668130, T2W, T2T); T2F = FNMS(KP242145790, T2E, T2B); Cr[0] = FMA(KP968583161, T2E, T2B); T2J = T2H + T2I; T2L = FMA(KP904730450, T2G, T2I); Cr[WS(csr, 9)] = FMA(KP897376177, T2V, T2Q); Cr[WS(csr, 4)] = FNMS(KP803003575, T2X, T2Q); T2K = FNMS(KP683113946, T2J, T2G); T2M = FMA(KP618033988, T2L, T2H); } } Cr[WS(csr, 5)] = FMA(KP792626838, T2K, T2F); Cr[WS(csr, 10)] = FMA(KP876091699, T2M, T2F); } } } } static const kr2c_desc desc = { 25, "r2cfII_25", {47, 12, 165, 0}, &GENUS }; void X(codelet_r2cfII_25) (planner *p) { X(kr2c_register) (p, r2cfII_25, &desc); } #else /* HAVE_FMA */ /* Generated by: ../../../genfft/gen_r2cf.native -compact -variables 4 -pipeline-latency 4 -n 25 -name r2cfII_25 -dft-II -include r2cfII.h */ /* * This function contains 213 FP additions, 148 FP multiplications, * (or, 126 additions, 61 multiplications, 87 fused multiply/add), * 94 stack variables, 38 constants, and 50 memory accesses */ #include "r2cfII.h" static void r2cfII_25(R *R0, R *R1, R *Cr, R *Ci, stride rs, stride csr, stride csi, INT v, INT ivs, INT ovs) { DK(KP1_996053456, +1.996053456856543123904673613726901106673810439); DK(KP062790519, +0.062790519529313376076178224565631133122484832); DK(KP125581039, +0.125581039058626752152356449131262266244969664); DK(KP998026728, +0.998026728428271561952336806863450553336905220); DK(KP1_369094211, +1.369094211857377347464566715242418539779038465); DK(KP728968627, +0.728968627421411523146730319055259111372571664); DK(KP963507348, +0.963507348203430549974383005744259307057084020); DK(KP876306680, +0.876306680043863587308115903922062583399064238); DK(KP497379774, +0.497379774329709576484567492012895936835134813); DK(KP968583161, +0.968583161128631119490168375464735813836012403); DK(KP1_457937254, +1.457937254842823046293460638110518222745143328); DK(KP684547105, +0.684547105928688673732283357621209269889519233); DK(KP1_752613360, +1.752613360087727174616231807844125166798128477); DK(KP481753674, +0.481753674101715274987191502872129653528542010); DK(KP1_937166322, +1.937166322257262238980336750929471627672024806); DK(KP248689887, +0.248689887164854788242283746006447968417567406); DK(KP992114701, +0.992114701314477831049793042785778521453036709); DK(KP250666467, +0.250666467128608490746237519633017587885836494); DK(KP1_809654104, +1.809654104932039055427337295865395187940827822); DK(KP425779291, +0.425779291565072648862502445744251703979973042); DK(KP1_541026485, +1.541026485551578461606019272792355694543335344); DK(KP637423989, +0.637423989748689710176712811676016195434917298); DK(KP1_688655851, +1.688655851004030157097116127933363010763318483); DK(KP535826794, +0.535826794978996618271308767867639978063575346); DK(KP851558583, +0.851558583130145297725004891488503407959946084); DK(KP904827052, +0.904827052466019527713668647932697593970413911); DK(KP1_984229402, +1.984229402628955662099586085571557042906073418); DK(KP125333233, +0.125333233564304245373118759816508793942918247); DK(KP1_274847979, +1.274847979497379420353425623352032390869834596); DK(KP770513242, +0.770513242775789230803009636396177847271667672); DK(KP844327925, +0.844327925502015078548558063966681505381659241); DK(KP1_071653589, +1.071653589957993236542617535735279956127150691); DK(KP293892626, +0.293892626146236564584352977319536384298826219); DK(KP475528258, +0.475528258147576786058219666689691071702849317); DK(KP250000000, +0.250000000000000000000000000000000000000000000); DK(KP587785252, +0.587785252292473129168705954639072768597652438); DK(KP951056516, +0.951056516295153572116439333379382143405698634); DK(KP559016994, +0.559016994374947424102293417182819058860154590); { INT i; for (i = v; i > 0; i = i - 1, R0 = R0 + ivs, R1 = R1 + ivs, Cr = Cr + ovs, Ci = Ci + ovs, MAKE_VOLATILE_STRIDE(100, rs), MAKE_VOLATILE_STRIDE(100, csr), MAKE_VOLATILE_STRIDE(100, csi)) { E TE, TR, T2i, T1z, TL, TS, TB, T2d, T1l, T1i, T2c, T9, T23, TZ, TW; E T22, Ti, T26, T16, T13, T25, Ts, T2a, T1e, T1b, T29, TP, TQ; { E TK, T1y, TH, T1x; TE = R0[0]; { E TI, TJ, TF, TG; TI = R0[WS(rs, 10)]; TJ = R1[WS(rs, 2)]; TK = TI - TJ; T1y = TI + TJ; TF = R0[WS(rs, 5)]; TG = R1[WS(rs, 7)]; TH = TF - TG; T1x = TF + TG; } TR = KP559016994 * (TH - TK); T2i = FNMS(KP587785252, T1x, KP951056516 * T1y); T1z = FMA(KP951056516, T1x, KP587785252 * T1y); TL = TH + TK; TS = FNMS(KP250000000, TL, TE); } { E Tt, Tw, Tz, TA, T1k, T1j, T1g, T1h; Tt = R0[WS(rs, 3)]; { E Tu, Tv, Tx, Ty; Tu = R0[WS(rs, 8)]; Tv = R1[WS(rs, 10)]; Tw = Tu - Tv; Tx = R1[0]; Ty = R1[WS(rs, 5)]; Tz = Tx + Ty; TA = Tw - Tz; T1k = Ty - Tx; T1j = Tu + Tv; } TB = Tt + TA; T2d = FNMS(KP293892626, T1j, KP475528258 * T1k); T1l = FMA(KP475528258, T1j, KP293892626 * T1k); T1g = FNMS(KP250000000, TA, Tt); T1h = KP559016994 * (Tw + Tz); T1i = T1g + T1h; T2c = T1g - T1h; } { E T1, T4, T7, T8, TY, TX, TU, TV; T1 = R0[WS(rs, 1)]; { E T2, T3, T5, T6; T2 = R0[WS(rs, 6)]; T3 = R1[WS(rs, 8)]; T4 = T2 - T3; T5 = R0[WS(rs, 11)]; T6 = R1[WS(rs, 3)]; T7 = T5 - T6; T8 = T4 + T7; TY = T5 + T6; TX = T2 + T3; } T9 = T1 + T8; T23 = FNMS(KP293892626, TX, KP475528258 * TY); TZ = FMA(KP475528258, TX, KP293892626 * TY); TU = KP559016994 * (T4 - T7); TV = FNMS(KP250000000, T8, T1); TW = TU + TV; T22 = TV - TU; } { E Ta, Td, Tg, Th, T15, T14, T11, T12; Ta = R0[WS(rs, 4)]; { E Tb, Tc, Te, Tf; Tb = R0[WS(rs, 9)]; Tc = R1[WS(rs, 11)]; Td = Tb - Tc; Te = R1[WS(rs, 1)]; Tf = R1[WS(rs, 6)]; Tg = Te + Tf; Th = Td - Tg; T15 = Tf - Te; T14 = Tb + Tc; } Ti = Ta + Th; T26 = FNMS(KP293892626, T14, KP475528258 * T15); T16 = FMA(KP475528258, T14, KP293892626 * T15); T11 = FNMS(KP250000000, Th, Ta); T12 = KP559016994 * (Td + Tg); T13 = T11 + T12; T25 = T11 - T12; } { E Tk, Tn, Tq, Tr, T1d, T1c, T19, T1a; Tk = R0[WS(rs, 2)]; { E Tl, Tm, To, Tp; Tl = R0[WS(rs, 7)]; Tm = R1[WS(rs, 9)]; Tn = Tl - Tm; To = R0[WS(rs, 12)]; Tp = R1[WS(rs, 4)]; Tq = To - Tp; Tr = Tn + Tq; T1d = To + Tp; T1c = Tl + Tm; } Ts = Tk + Tr; T2a = FNMS(KP293892626, T1c, KP475528258 * T1d); T1e = FMA(KP475528258, T1c, KP293892626 * T1d); T19 = KP559016994 * (Tn - Tq); T1a = FNMS(KP250000000, Tr, Tk); T1b = T19 + T1a; T29 = T1a - T19; } TP = TB - Ts; TQ = T9 - Ti; Ci[WS(csi, 2)] = FNMS(KP951056516, TQ, KP587785252 * TP); Ci[WS(csi, 7)] = FMA(KP587785252, TQ, KP951056516 * TP); { E TM, TD, TN, Tj, TC, TO; TM = TE + TL; Tj = T9 + Ti; TC = Ts + TB; TD = KP559016994 * (Tj - TC); TN = Tj + TC; Cr[WS(csr, 12)] = TM + TN; TO = FNMS(KP250000000, TN, TM); Cr[WS(csr, 2)] = TD + TO; Cr[WS(csr, 7)] = TO - TD; } { E TT, T1J, T1Y, T1U, T1X, T1P, T1V, T1M, T1W, T1A, T1B, T1r, T1C, T1v, T18; E T1n, T1o, T1G, T1D; TT = TR + TS; { E T1H, T1I, T1S, T1T; T1H = FNMS(KP844327925, TW, KP1_071653589 * TZ); T1I = FNMS(KP1_274847979, T16, KP770513242 * T13); T1J = T1H - T1I; T1Y = T1H + T1I; T1S = FMA(KP125333233, T1i, KP1_984229402 * T1l); T1T = FMA(KP904827052, T1b, KP851558583 * T1e); T1U = T1S - T1T; T1X = T1T + T1S; } { E T1N, T1O, T1K, T1L; T1N = FMA(KP535826794, TW, KP1_688655851 * TZ); T1O = FMA(KP637423989, T13, KP1_541026485 * T16); T1P = T1N - T1O; T1V = T1N + T1O; T1K = FNMS(KP1_809654104, T1e, KP425779291 * T1b); T1L = FNMS(KP992114701, T1i, KP250666467 * T1l); T1M = T1K - T1L; T1W = T1K + T1L; } { E T1p, T1q, T1t, T1u; T1p = FMA(KP844327925, T13, KP1_071653589 * T16); T1q = FMA(KP248689887, TW, KP1_937166322 * TZ); T1A = T1q + T1p; T1t = FMA(KP481753674, T1b, KP1_752613360 * T1e); T1u = FMA(KP684547105, T1i, KP1_457937254 * T1l); T1B = T1t + T1u; T1r = T1p - T1q; T1C = T1A + T1B; T1v = T1t - T1u; } { E T10, T17, T1f, T1m; T10 = FNMS(KP497379774, TZ, KP968583161 * TW); T17 = FNMS(KP1_688655851, T16, KP535826794 * T13); T18 = T10 + T17; T1f = FNMS(KP963507348, T1e, KP876306680 * T1b); T1m = FNMS(KP1_369094211, T1l, KP728968627 * T1i); T1n = T1f + T1m; T1o = T18 + T1n; T1G = T10 - T17; T1D = T1f - T1m; } { E T1R, T1Q, T20, T1Z; Cr[0] = TT + T1o; Ci[0] = -(T1z + T1C); T1R = KP559016994 * (T1P + T1M); T1Q = FMA(KP250000000, T1M - T1P, TT); Cr[WS(csr, 4)] = FMA(KP951056516, T1J, T1Q) + FMA(KP587785252, T1U, T1R); Cr[WS(csr, 9)] = FMA(KP951056516, T1U, T1Q) + FNMA(KP587785252, T1J, T1R); T20 = KP559016994 * (T1Y + T1X); T1Z = FMA(KP250000000, T1X - T1Y, T1z); Ci[WS(csi, 9)] = FMA(KP587785252, T1V, KP951056516 * T1W) + T1Z - T20; Ci[WS(csi, 4)] = FMA(KP587785252, T1W, T1Z) + FNMS(KP951056516, T1V, T20); { E T1E, T1F, T1s, T1w; T1E = FMS(KP250000000, T1C, T1z); T1F = KP559016994 * (T1B - T1A); Ci[WS(csi, 5)] = FMA(KP951056516, T1D, T1E) + FNMA(KP587785252, T1G, T1F); Ci[WS(csi, 10)] = FMA(KP951056516, T1G, KP587785252 * T1D) + T1E + T1F; T1s = FNMS(KP250000000, T1o, TT); T1w = KP559016994 * (T18 - T1n); Cr[WS(csr, 5)] = FMA(KP587785252, T1r, T1s) + FMS(KP951056516, T1v, T1w); Cr[WS(csr, 10)] = T1w + FMA(KP587785252, T1v, T1s) - (KP951056516 * T1r); } } } { E T21, T2z, T2L, T2K, T2M, T2F, T2P, T2C, T2Q, T2l, T2o, T2p, T2w, T2u, T28; E T2f, T2g, T2s, T2h; T21 = TS - TR; { E T2x, T2y, T2I, T2J; T2x = FNMS(KP844327925, T29, KP1_071653589 * T2a); T2y = FNMS(KP125581039, T2d, KP998026728 * T2c); T2z = T2x + T2y; T2L = T2y - T2x; T2I = FNMS(KP481753674, T22, KP1_752613360 * T23); T2J = FMA(KP904827052, T25, KP851558583 * T26); T2K = T2I + T2J; T2M = T2I - T2J; } { E T2D, T2E, T2A, T2B; T2D = FMA(KP535826794, T29, KP1_688655851 * T2a); T2E = FMA(KP062790519, T2c, KP1_996053456 * T2d); T2F = T2D + T2E; T2P = T2E - T2D; T2A = FMA(KP876306680, T22, KP963507348 * T23); T2B = FNMS(KP425779291, T25, KP1_809654104 * T26); T2C = T2A + T2B; T2Q = T2A - T2B; } { E T2j, T2k, T2m, T2n; T2j = FNMS(KP125333233, T25, KP1_984229402 * T26); T2k = FMA(KP684547105, T22, KP1_457937254 * T23); T2l = T2j - T2k; T2m = FNMS(KP770513242, T2c, KP1_274847979 * T2d); T2n = FMA(KP998026728, T29, KP125581039 * T2a); T2o = T2m - T2n; T2p = T2l + T2o; T2w = T2k + T2j; T2u = T2n + T2m; } { E T24, T27, T2b, T2e; T24 = FNMS(KP1_369094211, T23, KP728968627 * T22); T27 = FMA(KP992114701, T25, KP250666467 * T26); T28 = T24 - T27; T2b = FNMS(KP1_996053456, T2a, KP062790519 * T29); T2e = FMA(KP637423989, T2c, KP1_541026485 * T2d); T2f = T2b - T2e; T2g = T28 + T2f; T2s = T24 + T27; T2h = T2b + T2e; } { E T2H, T2G, T2O, T2N; Cr[WS(csr, 1)] = T21 + T2g; Ci[WS(csi, 1)] = T2p - T2i; T2H = KP559016994 * (T2C - T2F); T2G = FNMS(KP250000000, T2C + T2F, T21); Cr[WS(csr, 8)] = FMA(KP951056516, T2z, T2G) + FNMA(KP587785252, T2K, T2H); Cr[WS(csr, 3)] = FMA(KP951056516, T2K, KP587785252 * T2z) + T2G + T2H; T2O = KP559016994 * (T2M + T2L); T2N = FMA(KP250000000, T2L - T2M, T2i); Ci[WS(csi, 3)] = T2N + FMA(KP587785252, T2P, T2O) - (KP951056516 * T2Q); Ci[WS(csi, 8)] = FMA(KP587785252, T2Q, T2N) + FMS(KP951056516, T2P, T2O); { E T2t, T2v, T2q, T2r; T2t = FNMS(KP250000000, T2g, T21); T2v = KP559016994 * (T28 - T2f); Cr[WS(csr, 6)] = FMA(KP951056516, T2u, T2t) + FNMA(KP587785252, T2w, T2v); Cr[WS(csr, 11)] = FMA(KP951056516, T2w, T2v) + FMA(KP587785252, T2u, T2t); T2q = KP250000000 * T2p; T2r = KP559016994 * (T2l - T2o); Ci[WS(csi, 6)] = FMS(KP951056516, T2h, T2i + T2q) + FNMA(KP587785252, T2s, T2r); Ci[WS(csi, 11)] = FMA(KP951056516, T2s, KP587785252 * T2h) + T2r - (T2i + T2q); } } } } } } static const kr2c_desc desc = { 25, "r2cfII_25", {126, 61, 87, 0}, &GENUS }; void X(codelet_r2cfII_25) (planner *p) { X(kr2c_register) (p, r2cfII_25, &desc); } #endif /* HAVE_FMA */ fftw-3.3.3/rdft/scalar/r2cf/hc2cf_2.c0000644000175000017500000000704012054410676014027 00000000000000/* * Copyright (c) 2003, 2007-11 Matteo Frigo * Copyright (c) 2003, 2007-11 Massachusetts Institute of Technology * * This program is free software; you can redistribute it and/or modify * it under the terms of the GNU General Public License as published by * the Free Software Foundation; either version 2 of the License, or * (at your option) any later version. * * This program is distributed in the hope that it will be useful, * but WITHOUT ANY WARRANTY; without even the implied warranty of * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the * GNU General Public License for more details. * * You should have received a copy of the GNU General Public License * along with this program; if not, write to the Free Software * Foundation, Inc., 51 Franklin Street, Fifth Floor, Boston, MA 02110-1301 USA * */ /* This file was automatically generated --- DO NOT EDIT */ /* Generated on Sun Nov 25 07:40:30 EST 2012 */ #include "codelet-rdft.h" #ifdef HAVE_FMA /* Generated by: ../../../genfft/gen_hc2c.native -fma -reorder-insns -schedule-for-pipeline -compact -variables 4 -pipeline-latency 4 -n 2 -dit -name hc2cf_2 -include hc2cf.h */ /* * This function contains 6 FP additions, 4 FP multiplications, * (or, 4 additions, 2 multiplications, 2 fused multiply/add), * 11 stack variables, 0 constants, and 8 memory accesses */ #include "hc2cf.h" static void hc2cf_2(R *Rp, R *Ip, R *Rm, R *Im, const R *W, stride rs, INT mb, INT me, INT ms) { { INT m; for (m = mb, W = W + ((mb - 1) * 2); m < me; m = m + 1, Rp = Rp + ms, Ip = Ip + ms, Rm = Rm - ms, Im = Im - ms, W = W + 2, MAKE_VOLATILE_STRIDE(8, rs)) { E T1, Ta, T3, T6, T2, T5; T1 = Rp[0]; Ta = Rm[0]; T3 = Ip[0]; T6 = Im[0]; T2 = W[0]; T5 = W[1]; { E T8, T4, T9, T7; T8 = T2 * T6; T4 = T2 * T3; T9 = FNMS(T5, T3, T8); T7 = FMA(T5, T6, T4); Ip[0] = T9 + Ta; Im[0] = T9 - Ta; Rp[0] = T1 + T7; Rm[0] = T1 - T7; } } } } static const tw_instr twinstr[] = { {TW_FULL, 1, 2}, {TW_NEXT, 1, 0} }; static const hc2c_desc desc = { 2, "hc2cf_2", twinstr, &GENUS, {4, 2, 2, 0} }; void X(codelet_hc2cf_2) (planner *p) { X(khc2c_register) (p, hc2cf_2, &desc, HC2C_VIA_RDFT); } #else /* HAVE_FMA */ /* Generated by: ../../../genfft/gen_hc2c.native -compact -variables 4 -pipeline-latency 4 -n 2 -dit -name hc2cf_2 -include hc2cf.h */ /* * This function contains 6 FP additions, 4 FP multiplications, * (or, 4 additions, 2 multiplications, 2 fused multiply/add), * 9 stack variables, 0 constants, and 8 memory accesses */ #include "hc2cf.h" static void hc2cf_2(R *Rp, R *Ip, R *Rm, R *Im, const R *W, stride rs, INT mb, INT me, INT ms) { { INT m; for (m = mb, W = W + ((mb - 1) * 2); m < me; m = m + 1, Rp = Rp + ms, Ip = Ip + ms, Rm = Rm - ms, Im = Im - ms, W = W + 2, MAKE_VOLATILE_STRIDE(8, rs)) { E T1, T8, T6, T7; T1 = Rp[0]; T8 = Rm[0]; { E T3, T5, T2, T4; T3 = Ip[0]; T5 = Im[0]; T2 = W[0]; T4 = W[1]; T6 = FMA(T2, T3, T4 * T5); T7 = FNMS(T4, T3, T2 * T5); } Rm[0] = T1 - T6; Im[0] = T7 - T8; Rp[0] = T1 + T6; Ip[0] = T7 + T8; } } } static const tw_instr twinstr[] = { {TW_FULL, 1, 2}, {TW_NEXT, 1, 0} }; static const hc2c_desc desc = { 2, "hc2cf_2", twinstr, &GENUS, {4, 2, 2, 0} }; void X(codelet_hc2cf_2) (planner *p) { X(khc2c_register) (p, hc2cf_2, &desc, HC2C_VIA_RDFT); } #endif /* HAVE_FMA */ fftw-3.3.3/rdft/scalar/r2cf/hc2cf_4.c0000644000175000017500000001220212054410676014025 00000000000000/* * Copyright (c) 2003, 2007-11 Matteo Frigo * Copyright (c) 2003, 2007-11 Massachusetts Institute of Technology * * This program is free software; you can redistribute it and/or modify * it under the terms of the GNU General Public License as published by * the Free Software Foundation; either version 2 of the License, or * (at your option) any later version. * * This program is distributed in the hope that it will be useful, * but WITHOUT ANY WARRANTY; without even the implied warranty of * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the * GNU General Public License for more details. * * You should have received a copy of the GNU General Public License * along with this program; if not, write to the Free Software * Foundation, Inc., 51 Franklin Street, Fifth Floor, Boston, MA 02110-1301 USA * */ /* This file was automatically generated --- DO NOT EDIT */ /* Generated on Sun Nov 25 07:40:30 EST 2012 */ #include "codelet-rdft.h" #ifdef HAVE_FMA /* Generated by: ../../../genfft/gen_hc2c.native -fma -reorder-insns -schedule-for-pipeline -compact -variables 4 -pipeline-latency 4 -n 4 -dit -name hc2cf_4 -include hc2cf.h */ /* * This function contains 22 FP additions, 12 FP multiplications, * (or, 16 additions, 6 multiplications, 6 fused multiply/add), * 31 stack variables, 0 constants, and 16 memory accesses */ #include "hc2cf.h" static void hc2cf_4(R *Rp, R *Ip, R *Rm, R *Im, const R *W, stride rs, INT mb, INT me, INT ms) { { INT m; for (m = mb, W = W + ((mb - 1) * 6); m < me; m = m + 1, Rp = Rp + ms, Ip = Ip + ms, Rm = Rm - ms, Im = Im - ms, W = W + 6, MAKE_VOLATILE_STRIDE(16, rs)) { E To, Te, Tm, T8, Tw, Ty, Tq, Tk; { E T1, Tv, Tu, T7, Tg, Tj, Tf, Ti, Tp, Th; T1 = Rp[0]; Tv = Rm[0]; { E T3, T6, T2, T5; T3 = Rp[WS(rs, 1)]; T6 = Rm[WS(rs, 1)]; T2 = W[2]; T5 = W[3]; { E Ta, Td, Tc, Tn, Tb, Tt, T4, T9; Ta = Ip[0]; Td = Im[0]; Tt = T2 * T6; T4 = T2 * T3; T9 = W[0]; Tc = W[1]; Tu = FNMS(T5, T3, Tt); T7 = FMA(T5, T6, T4); Tn = T9 * Td; Tb = T9 * Ta; Tg = Ip[WS(rs, 1)]; Tj = Im[WS(rs, 1)]; To = FNMS(Tc, Ta, Tn); Te = FMA(Tc, Td, Tb); Tf = W[4]; Ti = W[5]; } } Tm = T1 - T7; T8 = T1 + T7; Tw = Tu + Tv; Ty = Tv - Tu; Tp = Tf * Tj; Th = Tf * Tg; Tq = FNMS(Ti, Tg, Tp); Tk = FMA(Ti, Tj, Th); } { E Ts, Tr, Tl, Tx; Ts = To + Tq; Tr = To - Tq; Tl = Te + Tk; Tx = Tk - Te; Rp[WS(rs, 1)] = Tm + Tr; Rm[0] = Tm - Tr; Ip[0] = Ts + Tw; Im[WS(rs, 1)] = Ts - Tw; Ip[WS(rs, 1)] = Tx + Ty; Im[0] = Tx - Ty; Rp[0] = T8 + Tl; Rm[WS(rs, 1)] = T8 - Tl; } } } } static const tw_instr twinstr[] = { {TW_FULL, 1, 4}, {TW_NEXT, 1, 0} }; static const hc2c_desc desc = { 4, "hc2cf_4", twinstr, &GENUS, {16, 6, 6, 0} }; void X(codelet_hc2cf_4) (planner *p) { X(khc2c_register) (p, hc2cf_4, &desc, HC2C_VIA_RDFT); } #else /* HAVE_FMA */ /* Generated by: ../../../genfft/gen_hc2c.native -compact -variables 4 -pipeline-latency 4 -n 4 -dit -name hc2cf_4 -include hc2cf.h */ /* * This function contains 22 FP additions, 12 FP multiplications, * (or, 16 additions, 6 multiplications, 6 fused multiply/add), * 13 stack variables, 0 constants, and 16 memory accesses */ #include "hc2cf.h" static void hc2cf_4(R *Rp, R *Ip, R *Rm, R *Im, const R *W, stride rs, INT mb, INT me, INT ms) { { INT m; for (m = mb, W = W + ((mb - 1) * 6); m < me; m = m + 1, Rp = Rp + ms, Ip = Ip + ms, Rm = Rm - ms, Im = Im - ms, W = W + 6, MAKE_VOLATILE_STRIDE(16, rs)) { E T1, Tp, T6, To, Tc, Tk, Th, Tl; T1 = Rp[0]; Tp = Rm[0]; { E T3, T5, T2, T4; T3 = Rp[WS(rs, 1)]; T5 = Rm[WS(rs, 1)]; T2 = W[2]; T4 = W[3]; T6 = FMA(T2, T3, T4 * T5); To = FNMS(T4, T3, T2 * T5); } { E T9, Tb, T8, Ta; T9 = Ip[0]; Tb = Im[0]; T8 = W[0]; Ta = W[1]; Tc = FMA(T8, T9, Ta * Tb); Tk = FNMS(Ta, T9, T8 * Tb); } { E Te, Tg, Td, Tf; Te = Ip[WS(rs, 1)]; Tg = Im[WS(rs, 1)]; Td = W[4]; Tf = W[5]; Th = FMA(Td, Te, Tf * Tg); Tl = FNMS(Tf, Te, Td * Tg); } { E T7, Ti, Tn, Tq; T7 = T1 + T6; Ti = Tc + Th; Rm[WS(rs, 1)] = T7 - Ti; Rp[0] = T7 + Ti; Tn = Tk + Tl; Tq = To + Tp; Im[WS(rs, 1)] = Tn - Tq; Ip[0] = Tn + Tq; } { E Tj, Tm, Tr, Ts; Tj = T1 - T6; Tm = Tk - Tl; Rm[0] = Tj - Tm; Rp[WS(rs, 1)] = Tj + Tm; Tr = Th - Tc; Ts = Tp - To; Im[0] = Tr - Ts; Ip[WS(rs, 1)] = Tr + Ts; } } } } static const tw_instr twinstr[] = { {TW_FULL, 1, 4}, {TW_NEXT, 1, 0} }; static const hc2c_desc desc = { 4, "hc2cf_4", twinstr, &GENUS, {16, 6, 6, 0} }; void X(codelet_hc2cf_4) (planner *p) { X(khc2c_register) (p, hc2cf_4, &desc, HC2C_VIA_RDFT); } #endif /* HAVE_FMA */ fftw-3.3.3/rdft/scalar/r2cf/hc2cf_6.c0000644000175000017500000001733712054410676014045 00000000000000/* * Copyright (c) 2003, 2007-11 Matteo Frigo * Copyright (c) 2003, 2007-11 Massachusetts Institute of Technology * * This program is free software; you can redistribute it and/or modify * it under the terms of the GNU General Public License as published by * the Free Software Foundation; either version 2 of the License, or * (at your option) any later version. * * This program is distributed in the hope that it will be useful, * but WITHOUT ANY WARRANTY; without even the implied warranty of * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the * GNU General Public License for more details. * * You should have received a copy of the GNU General Public License * along with this program; if not, write to the Free Software * Foundation, Inc., 51 Franklin Street, Fifth Floor, Boston, MA 02110-1301 USA * */ /* This file was automatically generated --- DO NOT EDIT */ /* Generated on Sun Nov 25 07:40:30 EST 2012 */ #include "codelet-rdft.h" #ifdef HAVE_FMA /* Generated by: ../../../genfft/gen_hc2c.native -fma -reorder-insns -schedule-for-pipeline -compact -variables 4 -pipeline-latency 4 -n 6 -dit -name hc2cf_6 -include hc2cf.h */ /* * This function contains 46 FP additions, 32 FP multiplications, * (or, 24 additions, 10 multiplications, 22 fused multiply/add), * 47 stack variables, 2 constants, and 24 memory accesses */ #include "hc2cf.h" static void hc2cf_6(R *Rp, R *Ip, R *Rm, R *Im, const R *W, stride rs, INT mb, INT me, INT ms) { DK(KP866025403, +0.866025403784438646763723170752936183471402627); DK(KP500000000, +0.500000000000000000000000000000000000000000000); { INT m; for (m = mb, W = W + ((mb - 1) * 10); m < me; m = m + 1, Rp = Rp + ms, Ip = Ip + ms, Rm = Rm - ms, Im = Im - ms, W = W + 10, MAKE_VOLATILE_STRIDE(24, rs)) { E TY, TU, T10, TZ; { E T1, TX, TW, T7, Tn, Tq, TJ, TS, TB, Tl, To, TK, Tt, Tw, Ts; E Tp, Tv; T1 = Rp[0]; TX = Rm[0]; { E T3, T6, T2, T5; T3 = Ip[WS(rs, 1)]; T6 = Im[WS(rs, 1)]; T2 = W[4]; T5 = W[5]; { E Ta, Td, Tg, TF, Tb, Tj, Tf, Tc, Ti, TV, T4, T9; Ta = Rp[WS(rs, 1)]; Td = Rm[WS(rs, 1)]; TV = T2 * T6; T4 = T2 * T3; T9 = W[2]; Tg = Ip[WS(rs, 2)]; TW = FNMS(T5, T3, TV); T7 = FMA(T5, T6, T4); TF = T9 * Td; Tb = T9 * Ta; Tj = Im[WS(rs, 2)]; Tf = W[8]; Tc = W[3]; Ti = W[9]; { E TG, Te, TI, Tk, TH, Th, Tm; Tn = Rp[WS(rs, 2)]; TH = Tf * Tj; Th = Tf * Tg; TG = FNMS(Tc, Ta, TF); Te = FMA(Tc, Td, Tb); TI = FNMS(Ti, Tg, TH); Tk = FMA(Ti, Tj, Th); Tq = Rm[WS(rs, 2)]; Tm = W[6]; TJ = TG + TI; TS = TI - TG; TB = Te + Tk; Tl = Te - Tk; To = Tm * Tn; TK = Tm * Tq; } Tt = Ip[0]; Tw = Im[0]; Ts = W[0]; Tp = W[7]; Tv = W[1]; } } { E TA, T8, TL, Tr, TN, Tx, T12, TM, Tu; TA = T1 + T7; T8 = T1 - T7; TM = Ts * Tw; Tu = Ts * Tt; TL = FNMS(Tp, Tn, TK); Tr = FMA(Tp, Tq, To); TN = FNMS(Tv, Tt, TM); Tx = FMA(Tv, Tw, Tu); T12 = TX - TW; TY = TW + TX; { E TP, TT, TD, TQ, TE, Tz, T14, T13; { E TO, TR, TC, Ty, T11; TO = TL + TN; TR = TN - TL; TC = Tr + Tx; Ty = Tr - Tx; TP = TJ - TO; TU = TJ + TO; TT = TR - TS; T11 = TS + TR; Tz = Tl + Ty; T14 = Ty - Tl; Im[WS(rs, 2)] = T11 - T12; T13 = FMA(KP500000000, T11, T12); T10 = TB - TC; TD = TB + TC; } Rm[WS(rs, 2)] = T8 + Tz; TQ = FNMS(KP500000000, Tz, T8); Im[0] = FMS(KP866025403, T14, T13); Ip[WS(rs, 1)] = FMA(KP866025403, T14, T13); TE = FNMS(KP500000000, TD, TA); Rm[0] = FNMS(KP866025403, TT, TQ); Rp[WS(rs, 1)] = FMA(KP866025403, TT, TQ); Rp[0] = TA + TD; Rm[WS(rs, 1)] = FMA(KP866025403, TP, TE); Rp[WS(rs, 2)] = FNMS(KP866025403, TP, TE); } } } Ip[0] = TU + TY; TZ = FNMS(KP500000000, TU, TY); Im[WS(rs, 1)] = FMS(KP866025403, T10, TZ); Ip[WS(rs, 2)] = FMA(KP866025403, T10, TZ); } } } static const tw_instr twinstr[] = { {TW_FULL, 1, 6}, {TW_NEXT, 1, 0} }; static const hc2c_desc desc = { 6, "hc2cf_6", twinstr, &GENUS, {24, 10, 22, 0} }; void X(codelet_hc2cf_6) (planner *p) { X(khc2c_register) (p, hc2cf_6, &desc, HC2C_VIA_RDFT); } #else /* HAVE_FMA */ /* Generated by: ../../../genfft/gen_hc2c.native -compact -variables 4 -pipeline-latency 4 -n 6 -dit -name hc2cf_6 -include hc2cf.h */ /* * This function contains 46 FP additions, 28 FP multiplications, * (or, 32 additions, 14 multiplications, 14 fused multiply/add), * 23 stack variables, 2 constants, and 24 memory accesses */ #include "hc2cf.h" static void hc2cf_6(R *Rp, R *Ip, R *Rm, R *Im, const R *W, stride rs, INT mb, INT me, INT ms) { DK(KP500000000, +0.500000000000000000000000000000000000000000000); DK(KP866025403, +0.866025403784438646763723170752936183471402627); { INT m; for (m = mb, W = W + ((mb - 1) * 10); m < me; m = m + 1, Rp = Rp + ms, Ip = Ip + ms, Rm = Rm - ms, Im = Im - ms, W = W + 10, MAKE_VOLATILE_STRIDE(24, rs)) { E T7, TS, Tv, TO, Tt, TJ, Tx, TF, Ti, TI, Tw, TC; { E T1, TN, T6, TM; T1 = Rp[0]; TN = Rm[0]; { E T3, T5, T2, T4; T3 = Ip[WS(rs, 1)]; T5 = Im[WS(rs, 1)]; T2 = W[4]; T4 = W[5]; T6 = FMA(T2, T3, T4 * T5); TM = FNMS(T4, T3, T2 * T5); } T7 = T1 - T6; TS = TN - TM; Tv = T1 + T6; TO = TM + TN; } { E Tn, TD, Ts, TE; { E Tk, Tm, Tj, Tl; Tk = Rp[WS(rs, 2)]; Tm = Rm[WS(rs, 2)]; Tj = W[6]; Tl = W[7]; Tn = FMA(Tj, Tk, Tl * Tm); TD = FNMS(Tl, Tk, Tj * Tm); } { E Tp, Tr, To, Tq; Tp = Ip[0]; Tr = Im[0]; To = W[0]; Tq = W[1]; Ts = FMA(To, Tp, Tq * Tr); TE = FNMS(Tq, Tp, To * Tr); } Tt = Tn - Ts; TJ = TE - TD; Tx = Tn + Ts; TF = TD + TE; } { E Tc, TA, Th, TB; { E T9, Tb, T8, Ta; T9 = Rp[WS(rs, 1)]; Tb = Rm[WS(rs, 1)]; T8 = W[2]; Ta = W[3]; Tc = FMA(T8, T9, Ta * Tb); TA = FNMS(Ta, T9, T8 * Tb); } { E Te, Tg, Td, Tf; Te = Ip[WS(rs, 2)]; Tg = Im[WS(rs, 2)]; Td = W[8]; Tf = W[9]; Th = FMA(Td, Te, Tf * Tg); TB = FNMS(Tf, Te, Td * Tg); } Ti = Tc - Th; TI = TA - TB; Tw = Tc + Th; TC = TA + TB; } { E TK, Tu, TH, TT, TR, TU; TK = KP866025403 * (TI + TJ); Tu = Ti + Tt; TH = FNMS(KP500000000, Tu, T7); Rm[WS(rs, 2)] = T7 + Tu; Rp[WS(rs, 1)] = TH + TK; Rm[0] = TH - TK; TT = KP866025403 * (Tt - Ti); TR = TJ - TI; TU = FMA(KP500000000, TR, TS); Im[WS(rs, 2)] = TR - TS; Ip[WS(rs, 1)] = TT + TU; Im[0] = TT - TU; } { E TG, Ty, Tz, TP, TL, TQ; TG = KP866025403 * (TC - TF); Ty = Tw + Tx; Tz = FNMS(KP500000000, Ty, Tv); Rp[0] = Tv + Ty; Rm[WS(rs, 1)] = Tz + TG; Rp[WS(rs, 2)] = Tz - TG; TP = KP866025403 * (Tw - Tx); TL = TC + TF; TQ = FNMS(KP500000000, TL, TO); Ip[0] = TL + TO; Ip[WS(rs, 2)] = TP + TQ; Im[WS(rs, 1)] = TP - TQ; } } } } static const tw_instr twinstr[] = { {TW_FULL, 1, 6}, {TW_NEXT, 1, 0} }; static const hc2c_desc desc = { 6, "hc2cf_6", twinstr, &GENUS, {32, 14, 14, 0} }; void X(codelet_hc2cf_6) (planner *p) { X(khc2c_register) (p, hc2cf_6, &desc, HC2C_VIA_RDFT); } #endif /* HAVE_FMA */ fftw-3.3.3/rdft/scalar/r2cf/hc2cf_8.c0000644000175000017500000002236612054410676014045 00000000000000/* * Copyright (c) 2003, 2007-11 Matteo Frigo * Copyright (c) 2003, 2007-11 Massachusetts Institute of Technology * * This program is free software; you can redistribute it and/or modify * it under the terms of the GNU General Public License as published by * the Free Software Foundation; either version 2 of the License, or * (at your option) any later version. * * This program is distributed in the hope that it will be useful, * but WITHOUT ANY WARRANTY; without even the implied warranty of * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the * GNU General Public License for more details. * * You should have received a copy of the GNU General Public License * along with this program; if not, write to the Free Software * Foundation, Inc., 51 Franklin Street, Fifth Floor, Boston, MA 02110-1301 USA * */ /* This file was automatically generated --- DO NOT EDIT */ /* Generated on Sun Nov 25 07:40:30 EST 2012 */ #include "codelet-rdft.h" #ifdef HAVE_FMA /* Generated by: ../../../genfft/gen_hc2c.native -fma -reorder-insns -schedule-for-pipeline -compact -variables 4 -pipeline-latency 4 -n 8 -dit -name hc2cf_8 -include hc2cf.h */ /* * This function contains 66 FP additions, 36 FP multiplications, * (or, 44 additions, 14 multiplications, 22 fused multiply/add), * 61 stack variables, 1 constants, and 32 memory accesses */ #include "hc2cf.h" static void hc2cf_8(R *Rp, R *Ip, R *Rm, R *Im, const R *W, stride rs, INT mb, INT me, INT ms) { DK(KP707106781, +0.707106781186547524400844362104849039284835938); { INT m; for (m = mb, W = W + ((mb - 1) * 14); m < me; m = m + 1, Rp = Rp + ms, Ip = Ip + ms, Rm = Rm - ms, Im = Im - ms, W = W + 14, MAKE_VOLATILE_STRIDE(32, rs)) { E T1g, T1f, T1e, Tm, T1q, T1o, T1p, TN, T1h, T1i; { E T1, T1m, T1l, T7, TS, Tk, TQ, Te, To, Tr, T17, TM, T12, Tu, TW; E Tp, Tx, Tt, Tq, Tw; { E T3, T6, T2, T5; T1 = Rp[0]; T1m = Rm[0]; T3 = Rp[WS(rs, 2)]; T6 = Rm[WS(rs, 2)]; T2 = W[6]; T5 = W[7]; { E Ta, Td, T9, Tc; { E Tg, Tj, Ti, TR, Th, T1k, T4, Tf; Tg = Rp[WS(rs, 3)]; Tj = Rm[WS(rs, 3)]; T1k = T2 * T6; T4 = T2 * T3; Tf = W[10]; Ti = W[11]; T1l = FNMS(T5, T3, T1k); T7 = FMA(T5, T6, T4); TR = Tf * Tj; Th = Tf * Tg; Ta = Rp[WS(rs, 1)]; Td = Rm[WS(rs, 1)]; TS = FNMS(Ti, Tg, TR); Tk = FMA(Ti, Tj, Th); T9 = W[2]; Tc = W[3]; } { E TB, TE, TH, T13, TC, TK, TG, TD, TJ, TP, Tb, TA, Tn; TB = Ip[WS(rs, 3)]; TE = Im[WS(rs, 3)]; TP = T9 * Td; Tb = T9 * Ta; TA = W[12]; TH = Ip[WS(rs, 1)]; TQ = FNMS(Tc, Ta, TP); Te = FMA(Tc, Td, Tb); T13 = TA * TE; TC = TA * TB; TK = Im[WS(rs, 1)]; TG = W[4]; TD = W[13]; TJ = W[5]; { E T14, TF, T16, TL, T15, TI; To = Ip[0]; T15 = TG * TK; TI = TG * TH; T14 = FNMS(TD, TB, T13); TF = FMA(TD, TE, TC); T16 = FNMS(TJ, TH, T15); TL = FMA(TJ, TK, TI); Tr = Im[0]; Tn = W[0]; T17 = T14 - T16; T1g = T14 + T16; TM = TF + TL; T12 = TF - TL; } Tu = Ip[WS(rs, 2)]; TW = Tn * Tr; Tp = Tn * To; Tx = Im[WS(rs, 2)]; Tt = W[8]; Tq = W[1]; Tw = W[9]; } } } { E T8, T1j, T1n, Tz, T1a, TU, Tl, T1b, T1c, T1v, T1t, T1w, T19, T1u, T1d; { E T1r, T10, TV, T1s, T11, T18; { E TO, TX, Ts, TZ, Ty, TT, TY, Tv; T8 = T1 + T7; TO = T1 - T7; TY = Tt * Tx; Tv = Tt * Tu; TX = FNMS(Tq, To, TW); Ts = FMA(Tq, Tr, Tp); TZ = FNMS(Tw, Tu, TY); Ty = FMA(Tw, Tx, Tv); TT = TQ - TS; T1j = TQ + TS; T1n = T1l + T1m; T1r = T1m - T1l; T10 = TX - TZ; T1f = TX + TZ; Tz = Ts + Ty; TV = Ts - Ty; T1a = TO - TT; TU = TO + TT; T1s = Te - Tk; Tl = Te + Tk; } T1b = T10 - TV; T11 = TV + T10; T18 = T12 - T17; T1c = T12 + T17; T1v = T1s + T1r; T1t = T1r - T1s; T1w = T18 - T11; T19 = T11 + T18; } Ip[WS(rs, 3)] = FMA(KP707106781, T1w, T1v); Im[0] = FMS(KP707106781, T1w, T1v); Rp[WS(rs, 1)] = FMA(KP707106781, T19, TU); Rm[WS(rs, 2)] = FNMS(KP707106781, T19, TU); T1u = T1b + T1c; T1d = T1b - T1c; Ip[WS(rs, 1)] = FMA(KP707106781, T1u, T1t); Im[WS(rs, 2)] = FMS(KP707106781, T1u, T1t); Rp[WS(rs, 3)] = FMA(KP707106781, T1d, T1a); Rm[0] = FNMS(KP707106781, T1d, T1a); T1e = T8 - Tl; Tm = T8 + Tl; T1q = T1n - T1j; T1o = T1j + T1n; T1p = TM - Tz; TN = Tz + TM; } } Ip[WS(rs, 2)] = T1p + T1q; Im[WS(rs, 1)] = T1p - T1q; Rp[0] = Tm + TN; Rm[WS(rs, 3)] = Tm - TN; T1h = T1f - T1g; T1i = T1f + T1g; Ip[0] = T1i + T1o; Im[WS(rs, 3)] = T1i - T1o; Rp[WS(rs, 2)] = T1e + T1h; Rm[WS(rs, 1)] = T1e - T1h; } } } static const tw_instr twinstr[] = { {TW_FULL, 1, 8}, {TW_NEXT, 1, 0} }; static const hc2c_desc desc = { 8, "hc2cf_8", twinstr, &GENUS, {44, 14, 22, 0} }; void X(codelet_hc2cf_8) (planner *p) { X(khc2c_register) (p, hc2cf_8, &desc, HC2C_VIA_RDFT); } #else /* HAVE_FMA */ /* Generated by: ../../../genfft/gen_hc2c.native -compact -variables 4 -pipeline-latency 4 -n 8 -dit -name hc2cf_8 -include hc2cf.h */ /* * This function contains 66 FP additions, 32 FP multiplications, * (or, 52 additions, 18 multiplications, 14 fused multiply/add), * 28 stack variables, 1 constants, and 32 memory accesses */ #include "hc2cf.h" static void hc2cf_8(R *Rp, R *Ip, R *Rm, R *Im, const R *W, stride rs, INT mb, INT me, INT ms) { DK(KP707106781, +0.707106781186547524400844362104849039284835938); { INT m; for (m = mb, W = W + ((mb - 1) * 14); m < me; m = m + 1, Rp = Rp + ms, Ip = Ip + ms, Rm = Rm - ms, Im = Im - ms, W = W + 14, MAKE_VOLATILE_STRIDE(32, rs)) { E T7, T1e, TH, T19, TF, T13, TR, TU, Ti, T1f, TK, T16, Tu, T12, TM; E TP; { E T1, T18, T6, T17; T1 = Rp[0]; T18 = Rm[0]; { E T3, T5, T2, T4; T3 = Rp[WS(rs, 2)]; T5 = Rm[WS(rs, 2)]; T2 = W[6]; T4 = W[7]; T6 = FMA(T2, T3, T4 * T5); T17 = FNMS(T4, T3, T2 * T5); } T7 = T1 + T6; T1e = T18 - T17; TH = T1 - T6; T19 = T17 + T18; } { E Tz, TS, TE, TT; { E Tw, Ty, Tv, Tx; Tw = Ip[WS(rs, 3)]; Ty = Im[WS(rs, 3)]; Tv = W[12]; Tx = W[13]; Tz = FMA(Tv, Tw, Tx * Ty); TS = FNMS(Tx, Tw, Tv * Ty); } { E TB, TD, TA, TC; TB = Ip[WS(rs, 1)]; TD = Im[WS(rs, 1)]; TA = W[4]; TC = W[5]; TE = FMA(TA, TB, TC * TD); TT = FNMS(TC, TB, TA * TD); } TF = Tz + TE; T13 = TS + TT; TR = Tz - TE; TU = TS - TT; } { E Tc, TI, Th, TJ; { E T9, Tb, T8, Ta; T9 = Rp[WS(rs, 1)]; Tb = Rm[WS(rs, 1)]; T8 = W[2]; Ta = W[3]; Tc = FMA(T8, T9, Ta * Tb); TI = FNMS(Ta, T9, T8 * Tb); } { E Te, Tg, Td, Tf; Te = Rp[WS(rs, 3)]; Tg = Rm[WS(rs, 3)]; Td = W[10]; Tf = W[11]; Th = FMA(Td, Te, Tf * Tg); TJ = FNMS(Tf, Te, Td * Tg); } Ti = Tc + Th; T1f = Tc - Th; TK = TI - TJ; T16 = TI + TJ; } { E To, TN, Tt, TO; { E Tl, Tn, Tk, Tm; Tl = Ip[0]; Tn = Im[0]; Tk = W[0]; Tm = W[1]; To = FMA(Tk, Tl, Tm * Tn); TN = FNMS(Tm, Tl, Tk * Tn); } { E Tq, Ts, Tp, Tr; Tq = Ip[WS(rs, 2)]; Ts = Im[WS(rs, 2)]; Tp = W[8]; Tr = W[9]; Tt = FMA(Tp, Tq, Tr * Ts); TO = FNMS(Tr, Tq, Tp * Ts); } Tu = To + Tt; T12 = TN + TO; TM = To - Tt; TP = TN - TO; } { E Tj, TG, T1b, T1c; Tj = T7 + Ti; TG = Tu + TF; Rm[WS(rs, 3)] = Tj - TG; Rp[0] = Tj + TG; { E T15, T1a, T11, T14; T15 = T12 + T13; T1a = T16 + T19; Im[WS(rs, 3)] = T15 - T1a; Ip[0] = T15 + T1a; T11 = T7 - Ti; T14 = T12 - T13; Rm[WS(rs, 1)] = T11 - T14; Rp[WS(rs, 2)] = T11 + T14; } T1b = TF - Tu; T1c = T19 - T16; Im[WS(rs, 1)] = T1b - T1c; Ip[WS(rs, 2)] = T1b + T1c; { E TX, T1g, T10, T1d, TY, TZ; TX = TH - TK; T1g = T1e - T1f; TY = TP - TM; TZ = TR + TU; T10 = KP707106781 * (TY - TZ); T1d = KP707106781 * (TY + TZ); Rm[0] = TX - T10; Ip[WS(rs, 1)] = T1d + T1g; Rp[WS(rs, 3)] = TX + T10; Im[WS(rs, 2)] = T1d - T1g; } { E TL, T1i, TW, T1h, TQ, TV; TL = TH + TK; T1i = T1f + T1e; TQ = TM + TP; TV = TR - TU; TW = KP707106781 * (TQ + TV); T1h = KP707106781 * (TV - TQ); Rm[WS(rs, 2)] = TL - TW; Ip[WS(rs, 3)] = T1h + T1i; Rp[WS(rs, 1)] = TL + TW; Im[0] = T1h - T1i; } } } } } static const tw_instr twinstr[] = { {TW_FULL, 1, 8}, {TW_NEXT, 1, 0} }; static const hc2c_desc desc = { 8, "hc2cf_8", twinstr, &GENUS, {52, 18, 14, 0} }; void X(codelet_hc2cf_8) (planner *p) { X(khc2c_register) (p, hc2cf_8, &desc, HC2C_VIA_RDFT); } #endif /* HAVE_FMA */ fftw-3.3.3/rdft/scalar/r2cf/hc2cf_10.c0000644000175000017500000003346112054410677014115 00000000000000/* * Copyright (c) 2003, 2007-11 Matteo Frigo * Copyright (c) 2003, 2007-11 Massachusetts Institute of Technology * * This program is free software; you can redistribute it and/or modify * it under the terms of the GNU General Public License as published by * the Free Software Foundation; either version 2 of the License, or * (at your option) any later version. * * This program is distributed in the hope that it will be useful, * but WITHOUT ANY WARRANTY; without even the implied warranty of * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the * GNU General Public License for more details. * * You should have received a copy of the GNU General Public License * along with this program; if not, write to the Free Software * Foundation, Inc., 51 Franklin Street, Fifth Floor, Boston, MA 02110-1301 USA * */ /* This file was automatically generated --- DO NOT EDIT */ /* Generated on Sun Nov 25 07:40:30 EST 2012 */ #include "codelet-rdft.h" #ifdef HAVE_FMA /* Generated by: ../../../genfft/gen_hc2c.native -fma -reorder-insns -schedule-for-pipeline -compact -variables 4 -pipeline-latency 4 -n 10 -dit -name hc2cf_10 -include hc2cf.h */ /* * This function contains 102 FP additions, 72 FP multiplications, * (or, 48 additions, 18 multiplications, 54 fused multiply/add), * 70 stack variables, 4 constants, and 40 memory accesses */ #include "hc2cf.h" static void hc2cf_10(R *Rp, R *Ip, R *Rm, R *Im, const R *W, stride rs, INT mb, INT me, INT ms) { DK(KP951056516, +0.951056516295153572116439333379382143405698634); DK(KP559016994, +0.559016994374947424102293417182819058860154590); DK(KP250000000, +0.250000000000000000000000000000000000000000000); DK(KP618033988, +0.618033988749894848204586834365638117720309180); { INT m; for (m = mb, W = W + ((mb - 1) * 18); m < me; m = m + 1, Rp = Rp + ms, Ip = Ip + ms, Rm = Rm - ms, Im = Im - ms, W = W + 18, MAKE_VOLATILE_STRIDE(40, rs)) { E T1X, T21, T20, T22; { E T26, T1U, T8, T12, T1n, T1P, T24, T1K, T1Y, T18, T10, T2b, T1H, T23, T15; E T1Z, T2a, Tz, T1O, T1y; { E T1, T1T, T3, T6, T2, T5; T1 = Rp[0]; T1T = Rm[0]; T3 = Ip[WS(rs, 2)]; T6 = Im[WS(rs, 2)]; T2 = W[8]; T5 = W[9]; { E T1l, TY, T1h, T1J, TM, T16, T1j, TS; { E TF, T1e, TO, TR, T1g, TL, TN, TQ, T1i, TP; { E TU, TX, TT, TW; { E TB, TE, T1R, T4, TA, TD; TB = Rp[WS(rs, 2)]; TE = Rm[WS(rs, 2)]; T1R = T2 * T6; T4 = T2 * T3; TA = W[6]; TD = W[7]; { E T1S, T7, T1d, TC; T1S = FNMS(T5, T3, T1R); T7 = FMA(T5, T6, T4); T1d = TA * TE; TC = TA * TB; T26 = T1T - T1S; T1U = T1S + T1T; T8 = T1 - T7; T12 = T1 + T7; TF = FMA(TD, TE, TC); T1e = FNMS(TD, TB, T1d); } } TU = Ip[0]; TX = Im[0]; TT = W[0]; TW = W[1]; { E TH, TK, TJ, T1f, TI, T1k, TV, TG; TH = Ip[WS(rs, 4)]; TK = Im[WS(rs, 4)]; T1k = TT * TX; TV = TT * TU; TG = W[16]; TJ = W[17]; T1l = FNMS(TW, TU, T1k); TY = FMA(TW, TX, TV); T1f = TG * TK; TI = TG * TH; TO = Rp[WS(rs, 3)]; TR = Rm[WS(rs, 3)]; T1g = FNMS(TJ, TH, T1f); TL = FMA(TJ, TK, TI); TN = W[10]; TQ = W[11]; } } T1h = T1e + T1g; T1J = T1g - T1e; TM = TF - TL; T16 = TF + TL; T1i = TN * TR; TP = TN * TO; T1j = FNMS(TQ, TO, T1i); TS = FMA(TQ, TR, TP); } { E T1p, Te, T1w, Tx, Tn, Tq, Tp, T1r, Tk, T1t, To; { E Tt, Tw, Tv, T1v, Tu; { E Ta, Td, T9, Tc, T1o, Tb, Ts; Ta = Rp[WS(rs, 1)]; Td = Rm[WS(rs, 1)]; { E T1I, T1m, TZ, T17; T1I = T1l - T1j; T1m = T1j + T1l; TZ = TS - TY; T17 = TS + TY; T1n = T1h - T1m; T1P = T1h + T1m; T24 = T1J + T1I; T1K = T1I - T1J; T1Y = T16 - T17; T18 = T16 + T17; T10 = TM + TZ; T2b = TZ - TM; T9 = W[2]; } Tc = W[3]; Tt = Ip[WS(rs, 1)]; Tw = Im[WS(rs, 1)]; T1o = T9 * Td; Tb = T9 * Ta; Ts = W[4]; Tv = W[5]; T1p = FNMS(Tc, Ta, T1o); Te = FMA(Tc, Td, Tb); T1v = Ts * Tw; Tu = Ts * Tt; } { E Tg, Tj, Tf, Ti, T1q, Th, Tm; Tg = Ip[WS(rs, 3)]; Tj = Im[WS(rs, 3)]; T1w = FNMS(Tv, Tt, T1v); Tx = FMA(Tv, Tw, Tu); Tf = W[12]; Ti = W[13]; Tn = Rp[WS(rs, 4)]; Tq = Rm[WS(rs, 4)]; T1q = Tf * Tj; Th = Tf * Tg; Tm = W[14]; Tp = W[15]; T1r = FNMS(Ti, Tg, T1q); Tk = FMA(Ti, Tj, Th); T1t = Tm * Tq; To = Tm * Tn; } } { E T1s, T1G, Tl, T13, T1u, Tr; T1s = T1p + T1r; T1G = T1r - T1p; Tl = Te - Tk; T13 = Te + Tk; T1u = FNMS(Tp, Tn, T1t); Tr = FMA(Tp, Tq, To); { E T1x, T1F, T14, Ty; T1x = T1u + T1w; T1F = T1w - T1u; T14 = Tr + Tx; Ty = Tr - Tx; T1H = T1F - T1G; T23 = T1G + T1F; T15 = T13 + T14; T1Z = T13 - T14; T2a = Ty - Tl; Tz = Tl + Ty; T1O = T1s + T1x; T1y = T1s - T1x; } } } } } { E T2c, T2e, T29, T2d; { E T1D, T11, T25, T28, T27; T1D = Tz - T10; T11 = Tz + T10; T25 = T23 + T24; T28 = T24 - T23; { E T1N, T1L, T1C, T1M, T1E; T1N = FNMS(KP618033988, T1H, T1K); T1L = FMA(KP618033988, T1K, T1H); Rm[WS(rs, 4)] = T8 + T11; T1C = FNMS(KP250000000, T11, T8); T1M = FNMS(KP559016994, T1D, T1C); T1E = FMA(KP559016994, T1D, T1C); T27 = FMA(KP250000000, T25, T26); T2c = FMA(KP618033988, T2b, T2a); T2e = FNMS(KP618033988, T2a, T2b); Rp[WS(rs, 1)] = FMA(KP951056516, T1L, T1E); Rm[0] = FNMS(KP951056516, T1L, T1E); Rp[WS(rs, 3)] = FMA(KP951056516, T1N, T1M); Rm[WS(rs, 2)] = FNMS(KP951056516, T1N, T1M); } Im[WS(rs, 4)] = T25 - T26; T29 = FMA(KP559016994, T28, T27); T2d = FNMS(KP559016994, T28, T27); } { E T1c, T1A, T1z, T1B, T19, T1b, T1a, T1Q, T1W, T1V; T19 = T15 + T18; T1b = T15 - T18; Ip[WS(rs, 3)] = FMA(KP951056516, T2e, T2d); Im[WS(rs, 2)] = FMS(KP951056516, T2e, T2d); Ip[WS(rs, 1)] = FMA(KP951056516, T2c, T29); Im[0] = FMS(KP951056516, T2c, T29); T1a = FNMS(KP250000000, T19, T12); Rp[0] = T12 + T19; T1c = FNMS(KP559016994, T1b, T1a); T1A = FMA(KP559016994, T1b, T1a); T1z = FNMS(KP618033988, T1y, T1n); T1B = FMA(KP618033988, T1n, T1y); T1Q = T1O + T1P; T1W = T1O - T1P; Rm[WS(rs, 3)] = FMA(KP951056516, T1B, T1A); Rp[WS(rs, 4)] = FNMS(KP951056516, T1B, T1A); Rm[WS(rs, 1)] = FMA(KP951056516, T1z, T1c); Rp[WS(rs, 2)] = FNMS(KP951056516, T1z, T1c); T1V = FNMS(KP250000000, T1Q, T1U); Ip[0] = T1Q + T1U; T1X = FNMS(KP559016994, T1W, T1V); T21 = FMA(KP559016994, T1W, T1V); T20 = FNMS(KP618033988, T1Z, T1Y); T22 = FMA(KP618033988, T1Y, T1Z); } } } Ip[WS(rs, 4)] = FMA(KP951056516, T22, T21); Im[WS(rs, 3)] = FMS(KP951056516, T22, T21); Ip[WS(rs, 2)] = FMA(KP951056516, T20, T1X); Im[WS(rs, 1)] = FMS(KP951056516, T20, T1X); } } } static const tw_instr twinstr[] = { {TW_FULL, 1, 10}, {TW_NEXT, 1, 0} }; static const hc2c_desc desc = { 10, "hc2cf_10", twinstr, &GENUS, {48, 18, 54, 0} }; void X(codelet_hc2cf_10) (planner *p) { X(khc2c_register) (p, hc2cf_10, &desc, HC2C_VIA_RDFT); } #else /* HAVE_FMA */ /* Generated by: ../../../genfft/gen_hc2c.native -compact -variables 4 -pipeline-latency 4 -n 10 -dit -name hc2cf_10 -include hc2cf.h */ /* * This function contains 102 FP additions, 60 FP multiplications, * (or, 72 additions, 30 multiplications, 30 fused multiply/add), * 45 stack variables, 4 constants, and 40 memory accesses */ #include "hc2cf.h" static void hc2cf_10(R *Rp, R *Ip, R *Rm, R *Im, const R *W, stride rs, INT mb, INT me, INT ms) { DK(KP587785252, +0.587785252292473129168705954639072768597652438); DK(KP951056516, +0.951056516295153572116439333379382143405698634); DK(KP250000000, +0.250000000000000000000000000000000000000000000); DK(KP559016994, +0.559016994374947424102293417182819058860154590); { INT m; for (m = mb, W = W + ((mb - 1) * 18); m < me; m = m + 1, Rp = Rp + ms, Ip = Ip + ms, Rm = Rm - ms, Im = Im - ms, W = W + 18, MAKE_VOLATILE_STRIDE(40, rs)) { E T7, T1O, TT, T1C, TF, TQ, TR, T1r, T1s, T1L, TX, TY, TZ, T16, T19; E T1y, Ti, Tt, Tu, T1o, T1p, T1M, TU, TV, TW, T1d, T1g, T1x; { E T1, T1B, T6, T1A; T1 = Rp[0]; T1B = Rm[0]; { E T3, T5, T2, T4; T3 = Ip[WS(rs, 2)]; T5 = Im[WS(rs, 2)]; T2 = W[8]; T4 = W[9]; T6 = FMA(T2, T3, T4 * T5); T1A = FNMS(T4, T3, T2 * T5); } T7 = T1 - T6; T1O = T1B - T1A; TT = T1 + T6; T1C = T1A + T1B; } { E Tz, T14, TP, T18, TE, T15, TK, T17; { E Tw, Ty, Tv, Tx; Tw = Rp[WS(rs, 2)]; Ty = Rm[WS(rs, 2)]; Tv = W[6]; Tx = W[7]; Tz = FMA(Tv, Tw, Tx * Ty); T14 = FNMS(Tx, Tw, Tv * Ty); } { E TM, TO, TL, TN; TM = Ip[0]; TO = Im[0]; TL = W[0]; TN = W[1]; TP = FMA(TL, TM, TN * TO); T18 = FNMS(TN, TM, TL * TO); } { E TB, TD, TA, TC; TB = Ip[WS(rs, 4)]; TD = Im[WS(rs, 4)]; TA = W[16]; TC = W[17]; TE = FMA(TA, TB, TC * TD); T15 = FNMS(TC, TB, TA * TD); } { E TH, TJ, TG, TI; TH = Rp[WS(rs, 3)]; TJ = Rm[WS(rs, 3)]; TG = W[10]; TI = W[11]; TK = FMA(TG, TH, TI * TJ); T17 = FNMS(TI, TH, TG * TJ); } TF = Tz - TE; TQ = TK - TP; TR = TF + TQ; T1r = T14 - T15; T1s = T18 - T17; T1L = T1s - T1r; TX = Tz + TE; TY = TK + TP; TZ = TX + TY; T16 = T14 + T15; T19 = T17 + T18; T1y = T16 + T19; } { E Tc, T1b, Ts, T1f, Th, T1c, Tn, T1e; { E T9, Tb, T8, Ta; T9 = Rp[WS(rs, 1)]; Tb = Rm[WS(rs, 1)]; T8 = W[2]; Ta = W[3]; Tc = FMA(T8, T9, Ta * Tb); T1b = FNMS(Ta, T9, T8 * Tb); } { E Tp, Tr, To, Tq; Tp = Ip[WS(rs, 1)]; Tr = Im[WS(rs, 1)]; To = W[4]; Tq = W[5]; Ts = FMA(To, Tp, Tq * Tr); T1f = FNMS(Tq, Tp, To * Tr); } { E Te, Tg, Td, Tf; Te = Ip[WS(rs, 3)]; Tg = Im[WS(rs, 3)]; Td = W[12]; Tf = W[13]; Th = FMA(Td, Te, Tf * Tg); T1c = FNMS(Tf, Te, Td * Tg); } { E Tk, Tm, Tj, Tl; Tk = Rp[WS(rs, 4)]; Tm = Rm[WS(rs, 4)]; Tj = W[14]; Tl = W[15]; Tn = FMA(Tj, Tk, Tl * Tm); T1e = FNMS(Tl, Tk, Tj * Tm); } Ti = Tc - Th; Tt = Tn - Ts; Tu = Ti + Tt; T1o = T1b - T1c; T1p = T1e - T1f; T1M = T1o + T1p; TU = Tc + Th; TV = Tn + Ts; TW = TU + TV; T1d = T1b + T1c; T1g = T1e + T1f; T1x = T1d + T1g; } { E T1l, TS, T1m, T1u, T1w, T1q, T1t, T1v, T1n; T1l = KP559016994 * (Tu - TR); TS = Tu + TR; T1m = FNMS(KP250000000, TS, T7); T1q = T1o - T1p; T1t = T1r + T1s; T1u = FMA(KP951056516, T1q, KP587785252 * T1t); T1w = FNMS(KP587785252, T1q, KP951056516 * T1t); Rm[WS(rs, 4)] = T7 + TS; T1v = T1m - T1l; Rm[WS(rs, 2)] = T1v - T1w; Rp[WS(rs, 3)] = T1v + T1w; T1n = T1l + T1m; Rm[0] = T1n - T1u; Rp[WS(rs, 1)] = T1n + T1u; } { E T1S, T1N, T1T, T1R, T1V, T1P, T1Q, T1W, T1U; T1S = KP559016994 * (T1M + T1L); T1N = T1L - T1M; T1T = FMA(KP250000000, T1N, T1O); T1P = TQ - TF; T1Q = Ti - Tt; T1R = FNMS(KP951056516, T1Q, KP587785252 * T1P); T1V = FMA(KP587785252, T1Q, KP951056516 * T1P); Im[WS(rs, 4)] = T1N - T1O; T1W = T1T - T1S; Im[WS(rs, 2)] = T1V - T1W; Ip[WS(rs, 3)] = T1V + T1W; T1U = T1S + T1T; Im[0] = T1R - T1U; Ip[WS(rs, 1)] = T1R + T1U; } { E T12, T10, T11, T1i, T1k, T1a, T1h, T1j, T13; T12 = KP559016994 * (TW - TZ); T10 = TW + TZ; T11 = FNMS(KP250000000, T10, TT); T1a = T16 - T19; T1h = T1d - T1g; T1i = FNMS(KP587785252, T1h, KP951056516 * T1a); T1k = FMA(KP951056516, T1h, KP587785252 * T1a); Rp[0] = TT + T10; T1j = T12 + T11; Rp[WS(rs, 4)] = T1j - T1k; Rm[WS(rs, 3)] = T1j + T1k; T13 = T11 - T12; Rp[WS(rs, 2)] = T13 - T1i; Rm[WS(rs, 1)] = T13 + T1i; } { E T1H, T1z, T1G, T1F, T1J, T1D, T1E, T1K, T1I; T1H = KP559016994 * (T1x - T1y); T1z = T1x + T1y; T1G = FNMS(KP250000000, T1z, T1C); T1D = TX - TY; T1E = TU - TV; T1F = FNMS(KP587785252, T1E, KP951056516 * T1D); T1J = FMA(KP951056516, T1E, KP587785252 * T1D); Ip[0] = T1z + T1C; T1K = T1H + T1G; Im[WS(rs, 3)] = T1J - T1K; Ip[WS(rs, 4)] = T1J + T1K; T1I = T1G - T1H; Im[WS(rs, 1)] = T1F - T1I; Ip[WS(rs, 2)] = T1F + T1I; } } } } static const tw_instr twinstr[] = { {TW_FULL, 1, 10}, {TW_NEXT, 1, 0} }; static const hc2c_desc desc = { 10, "hc2cf_10", twinstr, &GENUS, {72, 30, 30, 0} }; void X(codelet_hc2cf_10) (planner *p) { X(khc2c_register) (p, hc2cf_10, &desc, HC2C_VIA_RDFT); } #endif /* HAVE_FMA */ fftw-3.3.3/rdft/scalar/r2cf/hc2cf_12.c0000644000175000017500000003516212054410677014117 00000000000000/* * Copyright (c) 2003, 2007-11 Matteo Frigo * Copyright (c) 2003, 2007-11 Massachusetts Institute of Technology * * This program is free software; you can redistribute it and/or modify * it under the terms of the GNU General Public License as published by * the Free Software Foundation; either version 2 of the License, or * (at your option) any later version. * * This program is distributed in the hope that it will be useful, * but WITHOUT ANY WARRANTY; without even the implied warranty of * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the * GNU General Public License for more details. * * You should have received a copy of the GNU General Public License * along with this program; if not, write to the Free Software * Foundation, Inc., 51 Franklin Street, Fifth Floor, Boston, MA 02110-1301 USA * */ /* This file was automatically generated --- DO NOT EDIT */ /* Generated on Sun Nov 25 07:40:31 EST 2012 */ #include "codelet-rdft.h" #ifdef HAVE_FMA /* Generated by: ../../../genfft/gen_hc2c.native -fma -reorder-insns -schedule-for-pipeline -compact -variables 4 -pipeline-latency 4 -n 12 -dit -name hc2cf_12 -include hc2cf.h */ /* * This function contains 118 FP additions, 68 FP multiplications, * (or, 72 additions, 22 multiplications, 46 fused multiply/add), * 84 stack variables, 2 constants, and 48 memory accesses */ #include "hc2cf.h" static void hc2cf_12(R *Rp, R *Ip, R *Rm, R *Im, const R *W, stride rs, INT mb, INT me, INT ms) { DK(KP866025403, +0.866025403784438646763723170752936183471402627); DK(KP500000000, +0.500000000000000000000000000000000000000000000); { INT m; for (m = mb, W = W + ((mb - 1) * 22); m < me; m = m + 1, Rp = Rp + ms, Ip = Ip + ms, Rm = Rm - ms, Im = Im - ms, W = W + 22, MAKE_VOLATILE_STRIDE(48, rs)) { E T2n, T2u; { E T1, T2i, T2e, Tl, T1Y, T10, T1S, TG, T2f, T1s, T2s, Ty, T1Z, T1H, T21; E T1d, TI, TL, T2h, T1l, T2p, Te, TJ, T1w, TO, TR, TN, TK, TQ; { E TW, TZ, TY, T1X, TX; T1 = Rp[0]; T2i = Rm[0]; { E Th, Tk, Tg, Tj, T2d, Ti, TV; Th = Rp[WS(rs, 3)]; Tk = Rm[WS(rs, 3)]; Tg = W[10]; Tj = W[11]; TW = Ip[WS(rs, 4)]; TZ = Im[WS(rs, 4)]; T2d = Tg * Tk; Ti = Tg * Th; TV = W[16]; TY = W[17]; T2e = FNMS(Tj, Th, T2d); Tl = FMA(Tj, Tk, Ti); T1X = TV * TZ; TX = TV * TW; } { E Tn, Tq, Tt, T1o, To, Tw, Ts, Tp, Tv; { E TC, TF, TB, TE, T1R, TD, Tm; TC = Ip[WS(rs, 1)]; TF = Im[WS(rs, 1)]; T1Y = FNMS(TY, TW, T1X); T10 = FMA(TY, TZ, TX); TB = W[4]; TE = W[5]; Tn = Rp[WS(rs, 5)]; Tq = Rm[WS(rs, 5)]; T1R = TB * TF; TD = TB * TC; Tm = W[18]; Tt = Rp[WS(rs, 1)]; T1S = FNMS(TE, TC, T1R); TG = FMA(TE, TF, TD); T1o = Tm * Tq; To = Tm * Tn; Tw = Rm[WS(rs, 1)]; Ts = W[2]; Tp = W[19]; Tv = W[3]; } { E T12, T15, T13, T1D, T18, T1b, T17, T14, T1a; { E T1p, Tr, T1r, Tx, T1q, Tu, T11; T12 = Ip[0]; T1q = Ts * Tw; Tu = Ts * Tt; T1p = FNMS(Tp, Tn, T1o); Tr = FMA(Tp, Tq, To); T1r = FNMS(Tv, Tt, T1q); Tx = FMA(Tv, Tw, Tu); T15 = Im[0]; T11 = W[0]; T2f = T1p + T1r; T1s = T1p - T1r; T2s = Tx - Tr; Ty = Tr + Tx; T13 = T11 * T12; T1D = T11 * T15; } T18 = Ip[WS(rs, 2)]; T1b = Im[WS(rs, 2)]; T17 = W[8]; T14 = W[1]; T1a = W[9]; { E T3, T6, T4, T1h, T9, Tc, T8, T5, Tb; { E T1E, T16, T1G, T1c, T1F, T19, T2; T3 = Rp[WS(rs, 2)]; T1F = T17 * T1b; T19 = T17 * T18; T1E = FNMS(T14, T12, T1D); T16 = FMA(T14, T15, T13); T1G = FNMS(T1a, T18, T1F); T1c = FMA(T1a, T1b, T19); T6 = Rm[WS(rs, 2)]; T2 = W[6]; T1Z = T1E + T1G; T1H = T1E - T1G; T21 = T1c - T16; T1d = T16 + T1c; T4 = T2 * T3; T1h = T2 * T6; } T9 = Rp[WS(rs, 4)]; Tc = Rm[WS(rs, 4)]; T8 = W[14]; T5 = W[7]; Tb = W[15]; { E T1i, T7, T1k, Td, T1j, Ta, TH; TI = Ip[WS(rs, 3)]; T1j = T8 * Tc; Ta = T8 * T9; T1i = FNMS(T5, T3, T1h); T7 = FMA(T5, T6, T4); T1k = FNMS(Tb, T9, T1j); Td = FMA(Tb, Tc, Ta); TL = Im[WS(rs, 3)]; TH = W[12]; T2h = T1i + T1k; T1l = T1i - T1k; T2p = Td - T7; Te = T7 + Td; TJ = TH * TI; T1w = TH * TL; } TO = Ip[WS(rs, 5)]; TR = Im[WS(rs, 5)]; TN = W[20]; TK = W[13]; TQ = W[21]; } } } } { E T1g, T1n, T2r, T1A, T1V, T28, TA, T2o, T1v, T1C, T1U, T29, T2m, T2k, T2l; E T1f, T2a, T20; { E T2g, T1T, TT, T2j, TU, T1e; { E Tf, T1x, TM, T1z, TS, Tz, T1y, TP; T1g = FNMS(KP500000000, Te, T1); Tf = T1 + Te; T1y = TN * TR; TP = TN * TO; T1x = FNMS(TK, TI, T1w); TM = FMA(TK, TL, TJ); T1z = FNMS(TQ, TO, T1y); TS = FMA(TQ, TR, TP); Tz = Tl + Ty; T1n = FNMS(KP500000000, Ty, Tl); T2r = FNMS(KP500000000, T2f, T2e); T2g = T2e + T2f; T1T = T1x + T1z; T1A = T1x - T1z; T1V = TS - TM; TT = TM + TS; T28 = Tf - Tz; TA = Tf + Tz; T2j = T2h + T2i; T2o = FNMS(KP500000000, T2h, T2i); } T1v = FNMS(KP500000000, TT, TG); TU = TG + TT; T1e = T10 + T1d; T1C = FNMS(KP500000000, T1d, T10); T1U = FNMS(KP500000000, T1T, T1S); T29 = T1S + T1T; T2m = T2j - T2g; T2k = T2g + T2j; T2l = TU - T1e; T1f = TU + T1e; T2a = T1Y + T1Z; T20 = FNMS(KP500000000, T1Z, T1Y); } { E T1m, T1K, T2z, T2q, T2y, T2t, T1L, T1t, T1B, T1N, T2c, T2b; Im[WS(rs, 2)] = T2l - T2m; Ip[WS(rs, 3)] = T2l + T2m; Rp[0] = TA + T1f; Rm[WS(rs, 5)] = TA - T1f; T2c = T29 + T2a; T2b = T29 - T2a; T1m = FNMS(KP866025403, T1l, T1g); T1K = FMA(KP866025403, T1l, T1g); Ip[0] = T2c + T2k; Im[WS(rs, 5)] = T2c - T2k; Rm[WS(rs, 2)] = T28 + T2b; Rp[WS(rs, 3)] = T28 - T2b; T2z = FNMS(KP866025403, T2p, T2o); T2q = FMA(KP866025403, T2p, T2o); T2y = FNMS(KP866025403, T2s, T2r); T2t = FMA(KP866025403, T2s, T2r); T1L = FMA(KP866025403, T1s, T1n); T1t = FNMS(KP866025403, T1s, T1n); T1B = FNMS(KP866025403, T1A, T1v); T1N = FMA(KP866025403, T1A, T1v); { E T1Q, T2C, T23, T24, T2B, T27, T2v, T2w; { E T1u, T25, T26, T1O, T1I, T2A, T2x, T1W, T22, T1M, T1J, T1P; T1Q = T1m - T1t; T1u = T1m + T1t; T25 = FMA(KP866025403, T1V, T1U); T1W = FNMS(KP866025403, T1V, T1U); T26 = FMA(KP866025403, T21, T20); T22 = FNMS(KP866025403, T21, T20); T1O = FMA(KP866025403, T1H, T1C); T1I = FNMS(KP866025403, T1H, T1C); T2A = T2y + T2z; T2C = T2z - T2y; T23 = T1W - T22; T2x = T1W + T22; T1M = T1K + T1L; T24 = T1K - T1L; T2B = T1I - T1B; T1J = T1B + T1I; T1P = T1N + T1O; T2n = T1O - T1N; Ip[WS(rs, 2)] = T2A - T2x; Im[WS(rs, 3)] = -(T2x + T2A); Rm[WS(rs, 3)] = T1u + T1J; Rp[WS(rs, 2)] = T1u - T1J; Rm[WS(rs, 1)] = T1M - T1P; Rp[WS(rs, 4)] = T1M + T1P; T27 = T25 - T26; T2v = T25 + T26; T2w = T2t + T2q; T2u = T2q - T2t; } Ip[WS(rs, 4)] = T2v + T2w; Im[WS(rs, 1)] = T2v - T2w; Rp[WS(rs, 5)] = T1Q + T23; Rm[0] = T1Q - T23; Ip[WS(rs, 5)] = T2B + T2C; Im[0] = T2B - T2C; Rp[WS(rs, 1)] = T24 + T27; Rm[WS(rs, 4)] = T24 - T27; } } } } Ip[WS(rs, 1)] = T2n + T2u; Im[WS(rs, 4)] = T2n - T2u; } } } static const tw_instr twinstr[] = { {TW_FULL, 1, 12}, {TW_NEXT, 1, 0} }; static const hc2c_desc desc = { 12, "hc2cf_12", twinstr, &GENUS, {72, 22, 46, 0} }; void X(codelet_hc2cf_12) (planner *p) { X(khc2c_register) (p, hc2cf_12, &desc, HC2C_VIA_RDFT); } #else /* HAVE_FMA */ /* Generated by: ../../../genfft/gen_hc2c.native -compact -variables 4 -pipeline-latency 4 -n 12 -dit -name hc2cf_12 -include hc2cf.h */ /* * This function contains 118 FP additions, 60 FP multiplications, * (or, 88 additions, 30 multiplications, 30 fused multiply/add), * 47 stack variables, 2 constants, and 48 memory accesses */ #include "hc2cf.h" static void hc2cf_12(R *Rp, R *Ip, R *Rm, R *Im, const R *W, stride rs, INT mb, INT me, INT ms) { DK(KP500000000, +0.500000000000000000000000000000000000000000000); DK(KP866025403, +0.866025403784438646763723170752936183471402627); { INT m; for (m = mb, W = W + ((mb - 1) * 22); m < me; m = m + 1, Rp = Rp + ms, Ip = Ip + ms, Rm = Rm - ms, Im = Im - ms, W = W + 22, MAKE_VOLATILE_STRIDE(48, rs)) { E T1, T1W, T18, T22, Tc, T15, T1V, T23, TR, T1E, T1o, T1D, T12, T1l, T1F; E T1G, Ti, T1S, T1d, T25, Tt, T1a, T1T, T26, TA, T1y, T1j, T1B, TL, T1g; E T1z, T1A; { E T6, T16, Tb, T17; T1 = Rp[0]; T1W = Rm[0]; { E T3, T5, T2, T4; T3 = Rp[WS(rs, 2)]; T5 = Rm[WS(rs, 2)]; T2 = W[6]; T4 = W[7]; T6 = FMA(T2, T3, T4 * T5); T16 = FNMS(T4, T3, T2 * T5); } { E T8, Ta, T7, T9; T8 = Rp[WS(rs, 4)]; Ta = Rm[WS(rs, 4)]; T7 = W[14]; T9 = W[15]; Tb = FMA(T7, T8, T9 * Ta); T17 = FNMS(T9, T8, T7 * Ta); } T18 = KP866025403 * (T16 - T17); T22 = KP866025403 * (Tb - T6); Tc = T6 + Tb; T15 = FNMS(KP500000000, Tc, T1); T1V = T16 + T17; T23 = FNMS(KP500000000, T1V, T1W); } { E T11, T1n, TW, T1m; { E TO, TQ, TN, TP; TO = Ip[WS(rs, 4)]; TQ = Im[WS(rs, 4)]; TN = W[16]; TP = W[17]; TR = FMA(TN, TO, TP * TQ); T1E = FNMS(TP, TO, TN * TQ); } { E TY, T10, TX, TZ; TY = Ip[WS(rs, 2)]; T10 = Im[WS(rs, 2)]; TX = W[8]; TZ = W[9]; T11 = FMA(TX, TY, TZ * T10); T1n = FNMS(TZ, TY, TX * T10); } { E TT, TV, TS, TU; TT = Ip[0]; TV = Im[0]; TS = W[0]; TU = W[1]; TW = FMA(TS, TT, TU * TV); T1m = FNMS(TU, TT, TS * TV); } T1o = KP866025403 * (T1m - T1n); T1D = KP866025403 * (T11 - TW); T12 = TW + T11; T1l = FNMS(KP500000000, T12, TR); T1F = T1m + T1n; T1G = FNMS(KP500000000, T1F, T1E); } { E Ts, T1c, Tn, T1b; { E Tf, Th, Te, Tg; Tf = Rp[WS(rs, 3)]; Th = Rm[WS(rs, 3)]; Te = W[10]; Tg = W[11]; Ti = FMA(Te, Tf, Tg * Th); T1S = FNMS(Tg, Tf, Te * Th); } { E Tp, Tr, To, Tq; Tp = Rp[WS(rs, 1)]; Tr = Rm[WS(rs, 1)]; To = W[2]; Tq = W[3]; Ts = FMA(To, Tp, Tq * Tr); T1c = FNMS(Tq, Tp, To * Tr); } { E Tk, Tm, Tj, Tl; Tk = Rp[WS(rs, 5)]; Tm = Rm[WS(rs, 5)]; Tj = W[18]; Tl = W[19]; Tn = FMA(Tj, Tk, Tl * Tm); T1b = FNMS(Tl, Tk, Tj * Tm); } T1d = KP866025403 * (T1b - T1c); T25 = KP866025403 * (Ts - Tn); Tt = Tn + Ts; T1a = FNMS(KP500000000, Tt, Ti); T1T = T1b + T1c; T26 = FNMS(KP500000000, T1T, T1S); } { E TK, T1i, TF, T1h; { E Tx, Tz, Tw, Ty; Tx = Ip[WS(rs, 1)]; Tz = Im[WS(rs, 1)]; Tw = W[4]; Ty = W[5]; TA = FMA(Tw, Tx, Ty * Tz); T1y = FNMS(Ty, Tx, Tw * Tz); } { E TH, TJ, TG, TI; TH = Ip[WS(rs, 5)]; TJ = Im[WS(rs, 5)]; TG = W[20]; TI = W[21]; TK = FMA(TG, TH, TI * TJ); T1i = FNMS(TI, TH, TG * TJ); } { E TC, TE, TB, TD; TC = Ip[WS(rs, 3)]; TE = Im[WS(rs, 3)]; TB = W[12]; TD = W[13]; TF = FMA(TB, TC, TD * TE); T1h = FNMS(TD, TC, TB * TE); } T1j = KP866025403 * (T1h - T1i); T1B = KP866025403 * (TK - TF); TL = TF + TK; T1g = FNMS(KP500000000, TL, TA); T1z = T1h + T1i; T1A = FNMS(KP500000000, T1z, T1y); } { E Tv, T1N, T1Y, T20, T14, T1Z, T1Q, T1R; { E Td, Tu, T1U, T1X; Td = T1 + Tc; Tu = Ti + Tt; Tv = Td + Tu; T1N = Td - Tu; T1U = T1S + T1T; T1X = T1V + T1W; T1Y = T1U + T1X; T20 = T1X - T1U; } { E TM, T13, T1O, T1P; TM = TA + TL; T13 = TR + T12; T14 = TM + T13; T1Z = TM - T13; T1O = T1y + T1z; T1P = T1E + T1F; T1Q = T1O - T1P; T1R = T1O + T1P; } Rm[WS(rs, 5)] = Tv - T14; Im[WS(rs, 5)] = T1R - T1Y; Rp[0] = Tv + T14; Ip[0] = T1R + T1Y; Rp[WS(rs, 3)] = T1N - T1Q; Ip[WS(rs, 3)] = T1Z + T20; Rm[WS(rs, 2)] = T1N + T1Q; Im[WS(rs, 2)] = T1Z - T20; } { E T1t, T1J, T28, T2a, T1w, T21, T1M, T29; { E T1r, T1s, T24, T27; T1r = T15 + T18; T1s = T1a + T1d; T1t = T1r + T1s; T1J = T1r - T1s; T24 = T22 + T23; T27 = T25 + T26; T28 = T24 - T27; T2a = T27 + T24; } { E T1u, T1v, T1K, T1L; T1u = T1g + T1j; T1v = T1l + T1o; T1w = T1u + T1v; T21 = T1v - T1u; T1K = T1B + T1A; T1L = T1D + T1G; T1M = T1K - T1L; T29 = T1K + T1L; } Rm[WS(rs, 1)] = T1t - T1w; Im[WS(rs, 1)] = T29 - T2a; Rp[WS(rs, 4)] = T1t + T1w; Ip[WS(rs, 4)] = T29 + T2a; Rm[WS(rs, 4)] = T1J - T1M; Im[WS(rs, 4)] = T21 - T28; Rp[WS(rs, 1)] = T1J + T1M; Ip[WS(rs, 1)] = T21 + T28; } { E T1f, T1x, T2e, T2g, T1q, T2f, T1I, T2b; { E T19, T1e, T2c, T2d; T19 = T15 - T18; T1e = T1a - T1d; T1f = T19 + T1e; T1x = T19 - T1e; T2c = T26 - T25; T2d = T23 - T22; T2e = T2c + T2d; T2g = T2d - T2c; } { E T1k, T1p, T1C, T1H; T1k = T1g - T1j; T1p = T1l - T1o; T1q = T1k + T1p; T2f = T1p - T1k; T1C = T1A - T1B; T1H = T1D - T1G; T1I = T1C + T1H; T2b = T1H - T1C; } Rp[WS(rs, 2)] = T1f - T1q; Ip[WS(rs, 2)] = T2b + T2e; Rm[WS(rs, 3)] = T1f + T1q; Im[WS(rs, 3)] = T2b - T2e; Rm[0] = T1x - T1I; Im[0] = T2f - T2g; Rp[WS(rs, 5)] = T1x + T1I; Ip[WS(rs, 5)] = T2f + T2g; } } } } static const tw_instr twinstr[] = { {TW_FULL, 1, 12}, {TW_NEXT, 1, 0} }; static const hc2c_desc desc = { 12, "hc2cf_12", twinstr, &GENUS, {88, 30, 30, 0} }; void X(codelet_hc2cf_12) (planner *p) { X(khc2c_register) (p, hc2cf_12, &desc, HC2C_VIA_RDFT); } #endif /* HAVE_FMA */ fftw-3.3.3/rdft/scalar/r2cf/hc2cf_16.c0000644000175000017500000004773712054410701014122 00000000000000/* * Copyright (c) 2003, 2007-11 Matteo Frigo * Copyright (c) 2003, 2007-11 Massachusetts Institute of Technology * * This program is free software; you can redistribute it and/or modify * it under the terms of the GNU General Public License as published by * the Free Software Foundation; either version 2 of the License, or * (at your option) any later version. * * This program is distributed in the hope that it will be useful, * but WITHOUT ANY WARRANTY; without even the implied warranty of * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the * GNU General Public License for more details. * * You should have received a copy of the GNU General Public License * along with this program; if not, write to the Free Software * Foundation, Inc., 51 Franklin Street, Fifth Floor, Boston, MA 02110-1301 USA * */ /* This file was automatically generated --- DO NOT EDIT */ /* Generated on Sun Nov 25 07:40:31 EST 2012 */ #include "codelet-rdft.h" #ifdef HAVE_FMA /* Generated by: ../../../genfft/gen_hc2c.native -fma -reorder-insns -schedule-for-pipeline -compact -variables 4 -pipeline-latency 4 -n 16 -dit -name hc2cf_16 -include hc2cf.h */ /* * This function contains 174 FP additions, 100 FP multiplications, * (or, 104 additions, 30 multiplications, 70 fused multiply/add), * 97 stack variables, 3 constants, and 64 memory accesses */ #include "hc2cf.h" static void hc2cf_16(R *Rp, R *Ip, R *Rm, R *Im, const R *W, stride rs, INT mb, INT me, INT ms) { DK(KP923879532, +0.923879532511286756128183189396788286822416626); DK(KP414213562, +0.414213562373095048801688724209698078569671875); DK(KP707106781, +0.707106781186547524400844362104849039284835938); { INT m; for (m = mb, W = W + ((mb - 1) * 30); m < me; m = m + 1, Rp = Rp + ms, Ip = Ip + ms, Rm = Rm - ms, Im = Im - ms, W = W + 30, MAKE_VOLATILE_STRIDE(64, rs)) { E T3G, T3F; { E T3z, T3o, T8, T1I, T2p, T35, T2r, T1s, T2w, T36, T2k, T1F, T3k, T1N, T3A; E Tl, T1T, T2V, T1U, Tz, T29, T30, T2c, T11, TB, TE, T2h, T31, T2a, T1e; E TC, T1X, TH, TK, TG, TD, TJ; { E Ta, Td, Tb, T1J, Tg, Tj, Tf, Tc, Ti; { E T1h, T1k, T1n, T2l, T1i, T1q, T1m, T1j, T1p; { E T1, T3n, T3, T6, T2, T5; T1 = Rp[0]; T3n = Rm[0]; T3 = Rp[WS(rs, 4)]; T6 = Rm[WS(rs, 4)]; T2 = W[14]; T5 = W[15]; { E T3l, T4, T1g, T3m, T7; T1h = Ip[WS(rs, 7)]; T1k = Im[WS(rs, 7)]; T3l = T2 * T6; T4 = T2 * T3; T1g = W[28]; T1n = Ip[WS(rs, 3)]; T3m = FNMS(T5, T3, T3l); T7 = FMA(T5, T6, T4); T2l = T1g * T1k; T1i = T1g * T1h; T3z = T3n - T3m; T3o = T3m + T3n; T8 = T1 + T7; T1I = T1 - T7; T1q = Im[WS(rs, 3)]; T1m = W[12]; } T1j = W[29]; T1p = W[13]; } { E T1u, T1x, T1v, T2s, T1A, T1D, T1z, T1w, T1C; { E T2m, T1l, T2o, T1r, T2n, T1o, T1t; T1u = Ip[WS(rs, 1)]; T2n = T1m * T1q; T1o = T1m * T1n; T2m = FNMS(T1j, T1h, T2l); T1l = FMA(T1j, T1k, T1i); T2o = FNMS(T1p, T1n, T2n); T1r = FMA(T1p, T1q, T1o); T1x = Im[WS(rs, 1)]; T1t = W[4]; T2p = T2m - T2o; T35 = T2m + T2o; T2r = T1l - T1r; T1s = T1l + T1r; T1v = T1t * T1u; T2s = T1t * T1x; } T1A = Ip[WS(rs, 5)]; T1D = Im[WS(rs, 5)]; T1z = W[20]; T1w = W[5]; T1C = W[21]; { E T2t, T1y, T2v, T1E, T2u, T1B, T9; Ta = Rp[WS(rs, 2)]; T2u = T1z * T1D; T1B = T1z * T1A; T2t = FNMS(T1w, T1u, T2s); T1y = FMA(T1w, T1x, T1v); T2v = FNMS(T1C, T1A, T2u); T1E = FMA(T1C, T1D, T1B); Td = Rm[WS(rs, 2)]; T9 = W[6]; T2w = T2t - T2v; T36 = T2t + T2v; T2k = T1E - T1y; T1F = T1y + T1E; Tb = T9 * Ta; T1J = T9 * Td; } Tg = Rp[WS(rs, 6)]; Tj = Rm[WS(rs, 6)]; Tf = W[22]; Tc = W[7]; Ti = W[23]; } } { E TQ, TT, TR, T25, TW, TZ, TV, TS, TY; { E To, Tr, Tp, T1P, Tu, Tx, Tt, Tq, Tw; { E T1K, Te, T1M, Tk, T1L, Th, Tn; To = Rp[WS(rs, 1)]; T1L = Tf * Tj; Th = Tf * Tg; T1K = FNMS(Tc, Ta, T1J); Te = FMA(Tc, Td, Tb); T1M = FNMS(Ti, Tg, T1L); Tk = FMA(Ti, Tj, Th); Tr = Rm[WS(rs, 1)]; Tn = W[2]; T3k = T1K + T1M; T1N = T1K - T1M; T3A = Te - Tk; Tl = Te + Tk; Tp = Tn * To; T1P = Tn * Tr; } Tu = Rp[WS(rs, 5)]; Tx = Rm[WS(rs, 5)]; Tt = W[18]; Tq = W[3]; Tw = W[19]; { E T1Q, Ts, T1S, Ty, T1R, Tv, TP; TQ = Ip[0]; T1R = Tt * Tx; Tv = Tt * Tu; T1Q = FNMS(Tq, To, T1P); Ts = FMA(Tq, Tr, Tp); T1S = FNMS(Tw, Tu, T1R); Ty = FMA(Tw, Tx, Tv); TT = Im[0]; TP = W[0]; T1T = T1Q - T1S; T2V = T1Q + T1S; T1U = Ts - Ty; Tz = Ts + Ty; TR = TP * TQ; T25 = TP * TT; } TW = Ip[WS(rs, 4)]; TZ = Im[WS(rs, 4)]; TV = W[16]; TS = W[1]; TY = W[17]; } { E T13, T16, T14, T2d, T19, T1c, T18, T15, T1b; { E T26, TU, T28, T10, T27, TX, T12; T13 = Ip[WS(rs, 2)]; T27 = TV * TZ; TX = TV * TW; T26 = FNMS(TS, TQ, T25); TU = FMA(TS, TT, TR); T28 = FNMS(TY, TW, T27); T10 = FMA(TY, TZ, TX); T16 = Im[WS(rs, 2)]; T12 = W[8]; T29 = T26 - T28; T30 = T26 + T28; T2c = TU - T10; T11 = TU + T10; T14 = T12 * T13; T2d = T12 * T16; } T19 = Ip[WS(rs, 6)]; T1c = Im[WS(rs, 6)]; T18 = W[24]; T15 = W[9]; T1b = W[25]; { E T2e, T17, T2g, T1d, T2f, T1a, TA; TB = Rp[WS(rs, 7)]; T2f = T18 * T1c; T1a = T18 * T19; T2e = FNMS(T15, T13, T2d); T17 = FMA(T15, T16, T14); T2g = FNMS(T1b, T19, T2f); T1d = FMA(T1b, T1c, T1a); TE = Rm[WS(rs, 7)]; TA = W[26]; T2h = T2e - T2g; T31 = T2e + T2g; T2a = T17 - T1d; T1e = T17 + T1d; TC = TA * TB; T1X = TA * TE; } TH = Rp[WS(rs, 3)]; TK = Rm[WS(rs, 3)]; TG = W[10]; TD = W[27]; TJ = W[11]; } } } { E T2U, T3u, T2Z, T21, T1W, T34, T2X, T3f, T32, T3t, T1H, T3q, T3e, TO, T3g; E T37, T3r, T3s, T3h, T3i; { E Tm, T1Y, TF, T20, TL, T3p, T1Z, TI; T2U = T8 - Tl; Tm = T8 + Tl; T1Z = TG * TK; TI = TG * TH; T1Y = FNMS(TD, TB, T1X); TF = FMA(TD, TE, TC); T20 = FNMS(TJ, TH, T1Z); TL = FMA(TJ, TK, TI); T3p = T3k + T3o; T3u = T3o - T3k; { E T1f, TM, T1G, T3j, T2W, TN; T2Z = T11 - T1e; T1f = T11 + T1e; T21 = T1Y - T20; T2W = T1Y + T20; T1W = TF - TL; TM = TF + TL; T1G = T1s + T1F; T34 = T1s - T1F; T2X = T2V - T2W; T3j = T2V + T2W; T3f = T30 + T31; T32 = T30 - T31; T3t = TM - Tz; TN = Tz + TM; T3r = T1G - T1f; T1H = T1f + T1G; T3s = T3p - T3j; T3q = T3j + T3p; T3e = Tm - TN; TO = Tm + TN; T3g = T35 + T36; T37 = T35 - T36; } } Im[WS(rs, 3)] = T3r - T3s; Ip[WS(rs, 4)] = T3r + T3s; Rp[0] = TO + T1H; Rm[WS(rs, 7)] = TO - T1H; T3h = T3f - T3g; T3i = T3f + T3g; { E T3a, T2Y, T3x, T3v, T3b, T33; Ip[0] = T3i + T3q; Im[WS(rs, 7)] = T3i - T3q; Rp[WS(rs, 4)] = T3e + T3h; Rm[WS(rs, 3)] = T3e - T3h; T3a = T2U - T2X; T2Y = T2U + T2X; T3x = T3u - T3t; T3v = T3t + T3u; T3b = T32 - T2Z; T33 = T2Z + T32; { E T2E, T1O, T3B, T3H, T2x, T2q, T3C, T23, T2S, T2O, T2K, T2J, T3I, T2H, T2B; E T2j; { E T2F, T1V, T22, T2G, T3c, T38; T2E = T1I + T1N; T1O = T1I - T1N; T3B = T3z - T3A; T3H = T3A + T3z; T3c = T34 + T37; T38 = T34 - T37; T2F = T1U + T1T; T1V = T1T - T1U; { E T3d, T3w, T3y, T39; T3d = T3b - T3c; T3w = T3b + T3c; T3y = T38 - T33; T39 = T33 + T38; Rp[WS(rs, 6)] = FMA(KP707106781, T3d, T3a); Rm[WS(rs, 1)] = FNMS(KP707106781, T3d, T3a); Ip[WS(rs, 2)] = FMA(KP707106781, T3w, T3v); Im[WS(rs, 5)] = FMS(KP707106781, T3w, T3v); Ip[WS(rs, 6)] = FMA(KP707106781, T3y, T3x); Im[WS(rs, 1)] = FMS(KP707106781, T3y, T3x); Rp[WS(rs, 2)] = FMA(KP707106781, T39, T2Y); Rm[WS(rs, 5)] = FNMS(KP707106781, T39, T2Y); T22 = T1W + T21; T2G = T1W - T21; } { E T2M, T2N, T2b, T2i; T2x = T2r - T2w; T2M = T2r + T2w; T2N = T2p + T2k; T2q = T2k - T2p; T3C = T1V + T22; T23 = T1V - T22; T2S = FMA(KP414213562, T2M, T2N); T2O = FNMS(KP414213562, T2N, T2M); T2K = T29 - T2a; T2b = T29 + T2a; T2i = T2c - T2h; T2J = T2c + T2h; T3I = T2G - T2F; T2H = T2F + T2G; T2B = FNMS(KP414213562, T2b, T2i); T2j = FMA(KP414213562, T2i, T2b); } } { E T2R, T2L, T3L, T3M; { E T2A, T24, T2C, T2y, T3J, T3K, T2D, T2z; T2A = FNMS(KP707106781, T23, T1O); T24 = FMA(KP707106781, T23, T1O); T2R = FNMS(KP414213562, T2J, T2K); T2L = FMA(KP414213562, T2K, T2J); T2C = FNMS(KP414213562, T2q, T2x); T2y = FMA(KP414213562, T2x, T2q); T3J = FMA(KP707106781, T3I, T3H); T3L = FNMS(KP707106781, T3I, T3H); T3K = T2C - T2B; T2D = T2B + T2C; T3M = T2y - T2j; T2z = T2j + T2y; Ip[WS(rs, 3)] = FMA(KP923879532, T3K, T3J); Im[WS(rs, 4)] = FMS(KP923879532, T3K, T3J); Rp[WS(rs, 3)] = FMA(KP923879532, T2z, T24); Rm[WS(rs, 4)] = FNMS(KP923879532, T2z, T24); Rm[0] = FMA(KP923879532, T2D, T2A); Rp[WS(rs, 7)] = FNMS(KP923879532, T2D, T2A); } { E T2Q, T3D, T3E, T2T, T2I, T2P; T2Q = FNMS(KP707106781, T2H, T2E); T2I = FMA(KP707106781, T2H, T2E); T2P = T2L + T2O; T3G = T2O - T2L; T3F = FNMS(KP707106781, T3C, T3B); T3D = FMA(KP707106781, T3C, T3B); Ip[WS(rs, 7)] = FMA(KP923879532, T3M, T3L); Im[0] = FMS(KP923879532, T3M, T3L); Rp[WS(rs, 1)] = FMA(KP923879532, T2P, T2I); Rm[WS(rs, 6)] = FNMS(KP923879532, T2P, T2I); T3E = T2R + T2S; T2T = T2R - T2S; Ip[WS(rs, 1)] = FMA(KP923879532, T3E, T3D); Im[WS(rs, 6)] = FMS(KP923879532, T3E, T3D); Rp[WS(rs, 5)] = FMA(KP923879532, T2T, T2Q); Rm[WS(rs, 2)] = FNMS(KP923879532, T2T, T2Q); } } } } } } Ip[WS(rs, 5)] = FMA(KP923879532, T3G, T3F); Im[WS(rs, 2)] = FMS(KP923879532, T3G, T3F); } } } static const tw_instr twinstr[] = { {TW_FULL, 1, 16}, {TW_NEXT, 1, 0} }; static const hc2c_desc desc = { 16, "hc2cf_16", twinstr, &GENUS, {104, 30, 70, 0} }; void X(codelet_hc2cf_16) (planner *p) { X(khc2c_register) (p, hc2cf_16, &desc, HC2C_VIA_RDFT); } #else /* HAVE_FMA */ /* Generated by: ../../../genfft/gen_hc2c.native -compact -variables 4 -pipeline-latency 4 -n 16 -dit -name hc2cf_16 -include hc2cf.h */ /* * This function contains 174 FP additions, 84 FP multiplications, * (or, 136 additions, 46 multiplications, 38 fused multiply/add), * 52 stack variables, 3 constants, and 64 memory accesses */ #include "hc2cf.h" static void hc2cf_16(R *Rp, R *Ip, R *Rm, R *Im, const R *W, stride rs, INT mb, INT me, INT ms) { DK(KP382683432, +0.382683432365089771728459984030398866761344562); DK(KP923879532, +0.923879532511286756128183189396788286822416626); DK(KP707106781, +0.707106781186547524400844362104849039284835938); { INT m; for (m = mb, W = W + ((mb - 1) * 30); m < me; m = m + 1, Rp = Rp + ms, Ip = Ip + ms, Rm = Rm - ms, Im = Im - ms, W = W + 30, MAKE_VOLATILE_STRIDE(64, rs)) { E T7, T37, T1t, T2U, Ti, T38, T1w, T2R, Tu, T2s, T1C, T2c, TF, T2t, T1H; E T2d, T1f, T1q, T2B, T2C, T2D, T2E, T1Z, T2j, T24, T2k, TS, T13, T2w, T2x; E T2y, T2z, T1O, T2g, T1T, T2h; { E T1, T2T, T6, T2S; T1 = Rp[0]; T2T = Rm[0]; { E T3, T5, T2, T4; T3 = Rp[WS(rs, 4)]; T5 = Rm[WS(rs, 4)]; T2 = W[14]; T4 = W[15]; T6 = FMA(T2, T3, T4 * T5); T2S = FNMS(T4, T3, T2 * T5); } T7 = T1 + T6; T37 = T2T - T2S; T1t = T1 - T6; T2U = T2S + T2T; } { E Tc, T1u, Th, T1v; { E T9, Tb, T8, Ta; T9 = Rp[WS(rs, 2)]; Tb = Rm[WS(rs, 2)]; T8 = W[6]; Ta = W[7]; Tc = FMA(T8, T9, Ta * Tb); T1u = FNMS(Ta, T9, T8 * Tb); } { E Te, Tg, Td, Tf; Te = Rp[WS(rs, 6)]; Tg = Rm[WS(rs, 6)]; Td = W[22]; Tf = W[23]; Th = FMA(Td, Te, Tf * Tg); T1v = FNMS(Tf, Te, Td * Tg); } Ti = Tc + Th; T38 = Tc - Th; T1w = T1u - T1v; T2R = T1u + T1v; } { E To, T1y, Tt, T1z, T1A, T1B; { E Tl, Tn, Tk, Tm; Tl = Rp[WS(rs, 1)]; Tn = Rm[WS(rs, 1)]; Tk = W[2]; Tm = W[3]; To = FMA(Tk, Tl, Tm * Tn); T1y = FNMS(Tm, Tl, Tk * Tn); } { E Tq, Ts, Tp, Tr; Tq = Rp[WS(rs, 5)]; Ts = Rm[WS(rs, 5)]; Tp = W[18]; Tr = W[19]; Tt = FMA(Tp, Tq, Tr * Ts); T1z = FNMS(Tr, Tq, Tp * Ts); } Tu = To + Tt; T2s = T1y + T1z; T1A = T1y - T1z; T1B = To - Tt; T1C = T1A - T1B; T2c = T1B + T1A; } { E Tz, T1E, TE, T1F, T1D, T1G; { E Tw, Ty, Tv, Tx; Tw = Rp[WS(rs, 7)]; Ty = Rm[WS(rs, 7)]; Tv = W[26]; Tx = W[27]; Tz = FMA(Tv, Tw, Tx * Ty); T1E = FNMS(Tx, Tw, Tv * Ty); } { E TB, TD, TA, TC; TB = Rp[WS(rs, 3)]; TD = Rm[WS(rs, 3)]; TA = W[10]; TC = W[11]; TE = FMA(TA, TB, TC * TD); T1F = FNMS(TC, TB, TA * TD); } TF = Tz + TE; T2t = T1E + T1F; T1D = Tz - TE; T1G = T1E - T1F; T1H = T1D + T1G; T2d = T1D - T1G; } { E T19, T20, T1p, T1X, T1e, T21, T1k, T1W; { E T16, T18, T15, T17; T16 = Ip[WS(rs, 7)]; T18 = Im[WS(rs, 7)]; T15 = W[28]; T17 = W[29]; T19 = FMA(T15, T16, T17 * T18); T20 = FNMS(T17, T16, T15 * T18); } { E T1m, T1o, T1l, T1n; T1m = Ip[WS(rs, 5)]; T1o = Im[WS(rs, 5)]; T1l = W[20]; T1n = W[21]; T1p = FMA(T1l, T1m, T1n * T1o); T1X = FNMS(T1n, T1m, T1l * T1o); } { E T1b, T1d, T1a, T1c; T1b = Ip[WS(rs, 3)]; T1d = Im[WS(rs, 3)]; T1a = W[12]; T1c = W[13]; T1e = FMA(T1a, T1b, T1c * T1d); T21 = FNMS(T1c, T1b, T1a * T1d); } { E T1h, T1j, T1g, T1i; T1h = Ip[WS(rs, 1)]; T1j = Im[WS(rs, 1)]; T1g = W[4]; T1i = W[5]; T1k = FMA(T1g, T1h, T1i * T1j); T1W = FNMS(T1i, T1h, T1g * T1j); } T1f = T19 + T1e; T1q = T1k + T1p; T2B = T1f - T1q; T2C = T20 + T21; T2D = T1W + T1X; T2E = T2C - T2D; { E T1V, T1Y, T22, T23; T1V = T19 - T1e; T1Y = T1W - T1X; T1Z = T1V - T1Y; T2j = T1V + T1Y; T22 = T20 - T21; T23 = T1k - T1p; T24 = T22 + T23; T2k = T22 - T23; } } { E TM, T1K, T12, T1R, TR, T1L, TX, T1Q; { E TJ, TL, TI, TK; TJ = Ip[0]; TL = Im[0]; TI = W[0]; TK = W[1]; TM = FMA(TI, TJ, TK * TL); T1K = FNMS(TK, TJ, TI * TL); } { E TZ, T11, TY, T10; TZ = Ip[WS(rs, 6)]; T11 = Im[WS(rs, 6)]; TY = W[24]; T10 = W[25]; T12 = FMA(TY, TZ, T10 * T11); T1R = FNMS(T10, TZ, TY * T11); } { E TO, TQ, TN, TP; TO = Ip[WS(rs, 4)]; TQ = Im[WS(rs, 4)]; TN = W[16]; TP = W[17]; TR = FMA(TN, TO, TP * TQ); T1L = FNMS(TP, TO, TN * TQ); } { E TU, TW, TT, TV; TU = Ip[WS(rs, 2)]; TW = Im[WS(rs, 2)]; TT = W[8]; TV = W[9]; TX = FMA(TT, TU, TV * TW); T1Q = FNMS(TV, TU, TT * TW); } TS = TM + TR; T13 = TX + T12; T2w = TS - T13; T2x = T1K + T1L; T2y = T1Q + T1R; T2z = T2x - T2y; { E T1M, T1N, T1P, T1S; T1M = T1K - T1L; T1N = TX - T12; T1O = T1M + T1N; T2g = T1M - T1N; T1P = TM - TR; T1S = T1Q - T1R; T1T = T1P - T1S; T2h = T1P + T1S; } } { E T1J, T27, T3g, T3i, T26, T3h, T2a, T3d; { E T1x, T1I, T3e, T3f; T1x = T1t - T1w; T1I = KP707106781 * (T1C - T1H); T1J = T1x + T1I; T27 = T1x - T1I; T3e = KP707106781 * (T2d - T2c); T3f = T38 + T37; T3g = T3e + T3f; T3i = T3f - T3e; } { E T1U, T25, T28, T29; T1U = FMA(KP923879532, T1O, KP382683432 * T1T); T25 = FNMS(KP923879532, T24, KP382683432 * T1Z); T26 = T1U + T25; T3h = T25 - T1U; T28 = FNMS(KP923879532, T1T, KP382683432 * T1O); T29 = FMA(KP382683432, T24, KP923879532 * T1Z); T2a = T28 - T29; T3d = T28 + T29; } Rm[WS(rs, 4)] = T1J - T26; Im[WS(rs, 4)] = T3d - T3g; Rp[WS(rs, 3)] = T1J + T26; Ip[WS(rs, 3)] = T3d + T3g; Rm[0] = T27 - T2a; Im[0] = T3h - T3i; Rp[WS(rs, 7)] = T27 + T2a; Ip[WS(rs, 7)] = T3h + T3i; } { E T2v, T2H, T32, T34, T2G, T33, T2K, T2Z; { E T2r, T2u, T30, T31; T2r = T7 - Ti; T2u = T2s - T2t; T2v = T2r + T2u; T2H = T2r - T2u; T30 = TF - Tu; T31 = T2U - T2R; T32 = T30 + T31; T34 = T31 - T30; } { E T2A, T2F, T2I, T2J; T2A = T2w + T2z; T2F = T2B - T2E; T2G = KP707106781 * (T2A + T2F); T33 = KP707106781 * (T2F - T2A); T2I = T2z - T2w; T2J = T2B + T2E; T2K = KP707106781 * (T2I - T2J); T2Z = KP707106781 * (T2I + T2J); } Rm[WS(rs, 5)] = T2v - T2G; Im[WS(rs, 5)] = T2Z - T32; Rp[WS(rs, 2)] = T2v + T2G; Ip[WS(rs, 2)] = T2Z + T32; Rm[WS(rs, 1)] = T2H - T2K; Im[WS(rs, 1)] = T33 - T34; Rp[WS(rs, 6)] = T2H + T2K; Ip[WS(rs, 6)] = T33 + T34; } { E T2f, T2n, T3a, T3c, T2m, T3b, T2q, T35; { E T2b, T2e, T36, T39; T2b = T1t + T1w; T2e = KP707106781 * (T2c + T2d); T2f = T2b + T2e; T2n = T2b - T2e; T36 = KP707106781 * (T1C + T1H); T39 = T37 - T38; T3a = T36 + T39; T3c = T39 - T36; } { E T2i, T2l, T2o, T2p; T2i = FMA(KP382683432, T2g, KP923879532 * T2h); T2l = FNMS(KP382683432, T2k, KP923879532 * T2j); T2m = T2i + T2l; T3b = T2l - T2i; T2o = FNMS(KP382683432, T2h, KP923879532 * T2g); T2p = FMA(KP923879532, T2k, KP382683432 * T2j); T2q = T2o - T2p; T35 = T2o + T2p; } Rm[WS(rs, 6)] = T2f - T2m; Im[WS(rs, 6)] = T35 - T3a; Rp[WS(rs, 1)] = T2f + T2m; Ip[WS(rs, 1)] = T35 + T3a; Rm[WS(rs, 2)] = T2n - T2q; Im[WS(rs, 2)] = T3b - T3c; Rp[WS(rs, 5)] = T2n + T2q; Ip[WS(rs, 5)] = T3b + T3c; } { E TH, T2L, T2W, T2Y, T1s, T2X, T2O, T2P; { E Tj, TG, T2Q, T2V; Tj = T7 + Ti; TG = Tu + TF; TH = Tj + TG; T2L = Tj - TG; T2Q = T2s + T2t; T2V = T2R + T2U; T2W = T2Q + T2V; T2Y = T2V - T2Q; } { E T14, T1r, T2M, T2N; T14 = TS + T13; T1r = T1f + T1q; T1s = T14 + T1r; T2X = T1r - T14; T2M = T2x + T2y; T2N = T2C + T2D; T2O = T2M - T2N; T2P = T2M + T2N; } Rm[WS(rs, 7)] = TH - T1s; Im[WS(rs, 7)] = T2P - T2W; Rp[0] = TH + T1s; Ip[0] = T2P + T2W; Rm[WS(rs, 3)] = T2L - T2O; Im[WS(rs, 3)] = T2X - T2Y; Rp[WS(rs, 4)] = T2L + T2O; Ip[WS(rs, 4)] = T2X + T2Y; } } } } static const tw_instr twinstr[] = { {TW_FULL, 1, 16}, {TW_NEXT, 1, 0} }; static const hc2c_desc desc = { 16, "hc2cf_16", twinstr, &GENUS, {136, 46, 38, 0} }; void X(codelet_hc2cf_16) (planner *p) { X(khc2c_register) (p, hc2cf_16, &desc, HC2C_VIA_RDFT); } #endif /* HAVE_FMA */ fftw-3.3.3/rdft/scalar/r2cf/hc2cf_32.c0000644000175000017500000013251412054410711014105 00000000000000/* * Copyright (c) 2003, 2007-11 Matteo Frigo * Copyright (c) 2003, 2007-11 Massachusetts Institute of Technology * * This program is free software; you can redistribute it and/or modify * it under the terms of the GNU General Public License as published by * the Free Software Foundation; either version 2 of the License, or * (at your option) any later version. * * This program is distributed in the hope that it will be useful, * but WITHOUT ANY WARRANTY; without even the implied warranty of * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the * GNU General Public License for more details. * * You should have received a copy of the GNU General Public License * along with this program; if not, write to the Free Software * Foundation, Inc., 51 Franklin Street, Fifth Floor, Boston, MA 02110-1301 USA * */ /* This file was automatically generated --- DO NOT EDIT */ /* Generated on Sun Nov 25 07:40:33 EST 2012 */ #include "codelet-rdft.h" #ifdef HAVE_FMA /* Generated by: ../../../genfft/gen_hc2c.native -fma -reorder-insns -schedule-for-pipeline -compact -variables 4 -pipeline-latency 4 -n 32 -dit -name hc2cf_32 -include hc2cf.h */ /* * This function contains 434 FP additions, 260 FP multiplications, * (or, 236 additions, 62 multiplications, 198 fused multiply/add), * 135 stack variables, 7 constants, and 128 memory accesses */ #include "hc2cf.h" static void hc2cf_32(R *Rp, R *Ip, R *Rm, R *Im, const R *W, stride rs, INT mb, INT me, INT ms) { DK(KP831469612, +0.831469612302545237078788377617905756738560812); DK(KP980785280, +0.980785280403230449126182236134239036973933731); DK(KP198912367, +0.198912367379658006911597622644676228597850501); DK(KP668178637, +0.668178637919298919997757686523080761552472251); DK(KP923879532, +0.923879532511286756128183189396788286822416626); DK(KP414213562, +0.414213562373095048801688724209698078569671875); DK(KP707106781, +0.707106781186547524400844362104849039284835938); { INT m; for (m = mb, W = W + ((mb - 1) * 62); m < me; m = m + 1, Rp = Rp + ms, Ip = Ip + ms, Rm = Rm - ms, Im = Im - ms, W = W + 62, MAKE_VOLATILE_STRIDE(128, rs)) { E T90, T8Z; { E T8x, T87, T8, T3w, T83, T3B, T8y, Tl, T6F, Tz, T3J, T5T, T6G, TM, T3Q; E T5U, T3Z, T5Y, T7D, T6L, T5X, T46, T6M, T1f, T4e, T61, T7E, T6R, T6O, T1G; E T60, T4l, T54, T6c, T79, T7N, T32, T7b, T6f, T5r, T4v, T65, T6X, T7I, T29; E T70, T68, T4S, T5s, T5b, T7O, T7e, T76, T3t, T5t, T5i, T4H, T2y, T4B, T71; E T2m, T4w, T4F, T2s; { E T3X, T1d, T44, T6J, T11, T3T, T3V, T17, T5h, T5c; { E Ta, Td, Tg, T3x, Tb, Tj, Tf, Tc, Ti; { E T1, T86, T3, T6, T2, T5; T1 = Rp[0]; T86 = Rm[0]; T3 = Rp[WS(rs, 8)]; T6 = Rm[WS(rs, 8)]; T2 = W[30]; T5 = W[31]; { E T84, T4, T9, T85, T7; Ta = Rp[WS(rs, 4)]; Td = Rm[WS(rs, 4)]; T84 = T2 * T6; T4 = T2 * T3; T9 = W[14]; Tg = Rp[WS(rs, 12)]; T85 = FNMS(T5, T3, T84); T7 = FMA(T5, T6, T4); T3x = T9 * Td; Tb = T9 * Ta; T8x = T86 - T85; T87 = T85 + T86; T8 = T1 + T7; T3w = T1 - T7; Tj = Rm[WS(rs, 12)]; Tf = W[46]; } Tc = W[15]; Ti = W[47]; } { E Tu, Tx, T3F, Ts, Tw, T3G, Tv; { E To, Tr, Tp, T3E, Tq, Tt; { E T3y, Te, T3A, Tk, T3z, Th, Tn; To = Rp[WS(rs, 2)]; T3z = Tf * Tj; Th = Tf * Tg; T3y = FNMS(Tc, Ta, T3x); Te = FMA(Tc, Td, Tb); T3A = FNMS(Ti, Tg, T3z); Tk = FMA(Ti, Tj, Th); Tr = Rm[WS(rs, 2)]; Tn = W[6]; T83 = T3y + T3A; T3B = T3y - T3A; T8y = Te - Tk; Tl = Te + Tk; Tp = Tn * To; T3E = Tn * Tr; } Tq = W[7]; Tu = Rp[WS(rs, 10)]; Tx = Rm[WS(rs, 10)]; Tt = W[38]; T3F = FNMS(Tq, To, T3E); Ts = FMA(Tq, Tr, Tp); Tw = W[39]; T3G = Tt * Tx; Tv = Tt * Tu; } { E T3M, TF, TH, TK, TG, TJ, TE, TD, TC; { E TB, T3H, Ty, TA, T3I, T3D, T3L; TB = Rp[WS(rs, 14)]; TE = Rm[WS(rs, 14)]; T3H = FNMS(Tw, Tu, T3G); Ty = FMA(Tw, Tx, Tv); TA = W[54]; TD = W[55]; T6F = T3F + T3H; T3I = T3F - T3H; Tz = Ts + Ty; T3D = Ts - Ty; T3L = TA * TE; TC = TA * TB; T3J = T3D + T3I; T5T = T3I - T3D; T3M = FNMS(TD, TB, T3L); } TF = FMA(TD, TE, TC); TH = Rp[WS(rs, 6)]; TK = Rm[WS(rs, 6)]; TG = W[22]; TJ = W[23]; { E TU, T41, T13, T16, T43, T10, T12, T15, T3U, T14; { E T19, T1c, T18, T1b, T3P, T3K; { E TQ, TT, T3N, TI, TP, TS; TQ = Rp[WS(rs, 1)]; TT = Rm[WS(rs, 1)]; T3N = TG * TK; TI = TG * TH; TP = W[2]; TS = W[3]; { E T3O, TL, T40, TR; T3O = FNMS(TJ, TH, T3N); TL = FMA(TJ, TK, TI); T40 = TP * TT; TR = TP * TQ; T6G = T3M + T3O; T3P = T3M - T3O; TM = TF + TL; T3K = TF - TL; TU = FMA(TS, TT, TR); T41 = FNMS(TS, TQ, T40); } } T3Q = T3K - T3P; T5U = T3K + T3P; T19 = Rp[WS(rs, 13)]; T1c = Rm[WS(rs, 13)]; T18 = W[50]; T1b = W[51]; { E TW, TZ, TY, T42, TX, T3W, T1a, TV; TW = Rp[WS(rs, 9)]; TZ = Rm[WS(rs, 9)]; T3W = T18 * T1c; T1a = T18 * T19; TV = W[34]; TY = W[35]; T3X = FNMS(T1b, T19, T3W); T1d = FMA(T1b, T1c, T1a); T42 = TV * TZ; TX = TV * TW; T13 = Rp[WS(rs, 5)]; T16 = Rm[WS(rs, 5)]; T43 = FNMS(TY, TW, T42); T10 = FMA(TY, TZ, TX); T12 = W[18]; T15 = W[19]; } } T44 = T41 - T43; T6J = T41 + T43; T11 = TU + T10; T3T = TU - T10; T3U = T12 * T16; T14 = T12 * T13; T3V = FNMS(T15, T13, T3U); T17 = FMA(T15, T16, T14); } } } } { E T4g, T1l, T4c, T1E, T1u, T1x, T1w, T4i, T1r, T49, T1v; { E T1A, T1D, T1C, T4b, T1B; { E T1h, T1k, T1g, T1j, T4f, T1i, T1z; T1h = Rp[WS(rs, 15)]; T1k = Rm[WS(rs, 15)]; { E T6K, T3Y, T1e, T45; T6K = T3V + T3X; T3Y = T3V - T3X; T1e = T17 + T1d; T45 = T17 - T1d; T3Z = T3T + T3Y; T5Y = T3T - T3Y; T7D = T6J + T6K; T6L = T6J - T6K; T5X = T44 + T45; T46 = T44 - T45; T6M = T11 - T1e; T1f = T11 + T1e; T1g = W[58]; } T1j = W[59]; T1A = Rp[WS(rs, 11)]; T1D = Rm[WS(rs, 11)]; T4f = T1g * T1k; T1i = T1g * T1h; T1z = W[42]; T1C = W[43]; T4g = FNMS(T1j, T1h, T4f); T1l = FMA(T1j, T1k, T1i); T4b = T1z * T1D; T1B = T1z * T1A; } { E T1n, T1q, T1m, T1p, T4h, T1o, T1t; T1n = Rp[WS(rs, 7)]; T1q = Rm[WS(rs, 7)]; T4c = FNMS(T1C, T1A, T4b); T1E = FMA(T1C, T1D, T1B); T1m = W[26]; T1p = W[27]; T1u = Rp[WS(rs, 3)]; T1x = Rm[WS(rs, 3)]; T4h = T1m * T1q; T1o = T1m * T1n; T1t = W[10]; T1w = W[11]; T4i = FNMS(T1p, T1n, T4h); T1r = FMA(T1p, T1q, T1o); T49 = T1t * T1x; T1v = T1t * T1u; } } { E T4j, T6P, T1s, T48, T4a, T1y; T4j = T4g - T4i; T6P = T4g + T4i; T1s = T1l + T1r; T48 = T1l - T1r; T4a = FNMS(T1w, T1u, T49); T1y = FMA(T1w, T1x, T1v); { E T6Q, T4d, T4k, T1F; T6Q = T4a + T4c; T4d = T4a - T4c; T4k = T1y - T1E; T1F = T1y + T1E; T4e = T48 + T4d; T61 = T48 - T4d; T7E = T6P + T6Q; T6R = T6P - T6Q; T6O = T1s - T1F; T1G = T1s + T1F; T60 = T4j + T4k; T4l = T4j - T4k; } } } { E T5m, T2H, T52, T30, T2Q, T2T, T2S, T5o, T2N, T4Z, T2R; { E T2W, T2Z, T2Y, T51, T2X; { E T2D, T2G, T2C, T2F, T5l, T2E, T2V; T2D = Ip[WS(rs, 15)]; T2G = Im[WS(rs, 15)]; T2C = W[60]; T2F = W[61]; T2W = Ip[WS(rs, 11)]; T2Z = Im[WS(rs, 11)]; T5l = T2C * T2G; T2E = T2C * T2D; T2V = W[44]; T2Y = W[45]; T5m = FNMS(T2F, T2D, T5l); T2H = FMA(T2F, T2G, T2E); T51 = T2V * T2Z; T2X = T2V * T2W; } { E T2J, T2M, T2I, T2L, T5n, T2K, T2P; T2J = Ip[WS(rs, 7)]; T2M = Im[WS(rs, 7)]; T52 = FNMS(T2Y, T2W, T51); T30 = FMA(T2Y, T2Z, T2X); T2I = W[28]; T2L = W[29]; T2Q = Ip[WS(rs, 3)]; T2T = Im[WS(rs, 3)]; T5n = T2I * T2M; T2K = T2I * T2J; T2P = W[12]; T2S = W[13]; T5o = FNMS(T2L, T2J, T5n); T2N = FMA(T2L, T2M, T2K); T4Z = T2P * T2T; T2R = T2P * T2Q; } } { E T5p, T77, T2O, T4Y, T50, T2U; T5p = T5m - T5o; T77 = T5m + T5o; T2O = T2H + T2N; T4Y = T2H - T2N; T50 = FNMS(T2S, T2Q, T4Z); T2U = FMA(T2S, T2T, T2R); { E T78, T53, T5q, T31; T78 = T50 + T52; T53 = T50 - T52; T5q = T30 - T2U; T31 = T2U + T30; T54 = T4Y + T53; T6c = T4Y - T53; T79 = T77 - T78; T7N = T77 + T78; T32 = T2O + T31; T7b = T2O - T31; T6f = T5q - T5p; T5r = T5p + T5q; } } } { E T4N, T1O, T4t, T27, T1X, T20, T1Z, T4P, T1U, T4q, T1Y; { E T23, T26, T25, T4s, T24; { E T1K, T1N, T1J, T1M, T4M, T1L, T22; T1K = Ip[0]; T1N = Im[0]; T1J = W[0]; T1M = W[1]; T23 = Ip[WS(rs, 12)]; T26 = Im[WS(rs, 12)]; T4M = T1J * T1N; T1L = T1J * T1K; T22 = W[48]; T25 = W[49]; T4N = FNMS(T1M, T1K, T4M); T1O = FMA(T1M, T1N, T1L); T4s = T22 * T26; T24 = T22 * T23; } { E T1Q, T1T, T1P, T1S, T4O, T1R, T1W; T1Q = Ip[WS(rs, 8)]; T1T = Im[WS(rs, 8)]; T4t = FNMS(T25, T23, T4s); T27 = FMA(T25, T26, T24); T1P = W[32]; T1S = W[33]; T1X = Ip[WS(rs, 4)]; T20 = Im[WS(rs, 4)]; T4O = T1P * T1T; T1R = T1P * T1Q; T1W = W[16]; T1Z = W[17]; T4P = FNMS(T1S, T1Q, T4O); T1U = FMA(T1S, T1T, T1R); T4q = T1W * T20; T1Y = T1W * T1X; } } { E T4Q, T6V, T1V, T4p, T4r, T21; T4Q = T4N - T4P; T6V = T4N + T4P; T1V = T1O + T1U; T4p = T1O - T1U; T4r = FNMS(T1Z, T1X, T4q); T21 = FMA(T1Z, T20, T1Y); { E T6W, T4u, T4R, T28; T6W = T4r + T4t; T4u = T4r - T4t; T4R = T21 - T27; T28 = T21 + T27; T4v = T4p + T4u; T65 = T4p - T4u; T6X = T6V - T6W; T7I = T6V + T6W; T29 = T1V + T28; T70 = T1V - T28; T68 = T4Q + T4R; T4S = T4Q - T4R; } } } { E T57, T38, T5g, T3r, T3h, T3k, T3j, T59, T3e, T5d, T3i; { E T3n, T3q, T3p, T5f, T3o; { E T34, T37, T33, T36, T56, T35, T3m; T34 = Ip[WS(rs, 1)]; T37 = Im[WS(rs, 1)]; T33 = W[4]; T36 = W[5]; T3n = Ip[WS(rs, 5)]; T3q = Im[WS(rs, 5)]; T56 = T33 * T37; T35 = T33 * T34; T3m = W[20]; T3p = W[21]; T57 = FNMS(T36, T34, T56); T38 = FMA(T36, T37, T35); T5f = T3m * T3q; T3o = T3m * T3n; } { E T3a, T3d, T39, T3c, T58, T3b, T3g; T3a = Ip[WS(rs, 9)]; T3d = Im[WS(rs, 9)]; T5g = FNMS(T3p, T3n, T5f); T3r = FMA(T3p, T3q, T3o); T39 = W[36]; T3c = W[37]; T3h = Ip[WS(rs, 13)]; T3k = Im[WS(rs, 13)]; T58 = T39 * T3d; T3b = T39 * T3a; T3g = W[52]; T3j = W[53]; T59 = FNMS(T3c, T3a, T58); T3e = FMA(T3c, T3d, T3b); T5d = T3g * T3k; T3i = T3g * T3h; } } { E T5a, T7c, T3f, T55, T5e, T3l, T7d, T3s; T5a = T57 - T59; T7c = T57 + T59; T3f = T38 + T3e; T55 = T38 - T3e; T5e = FNMS(T3j, T3h, T5d); T3l = FMA(T3j, T3k, T3i); T5h = T5e - T5g; T7d = T5e + T5g; T3s = T3l + T3r; T5c = T3l - T3r; T5s = T5a - T55; T5b = T55 + T5a; T7O = T7c + T7d; T7e = T7c - T7d; T76 = T3s - T3f; T3t = T3f + T3s; } } { E T4y, T2f, T2o, T2r, T4A, T2l, T2n, T2q, T4E, T2p; { E T2u, T2x, T2t, T2w; { E T2b, T2e, T2d, T4x, T2c, T2a; T2b = Ip[WS(rs, 2)]; T2e = Im[WS(rs, 2)]; T2a = W[8]; T5t = T5c + T5h; T5i = T5c - T5h; T2d = W[9]; T4x = T2a * T2e; T2c = T2a * T2b; T2u = Ip[WS(rs, 6)]; T2x = Im[WS(rs, 6)]; T4y = FNMS(T2d, T2b, T4x); T2f = FMA(T2d, T2e, T2c); T2t = W[24]; T2w = W[25]; } { E T2h, T2k, T2j, T4z, T2i, T4G, T2v, T2g; T2h = Ip[WS(rs, 10)]; T2k = Im[WS(rs, 10)]; T4G = T2t * T2x; T2v = T2t * T2u; T2g = W[40]; T2j = W[41]; T4H = FNMS(T2w, T2u, T4G); T2y = FMA(T2w, T2x, T2v); T4z = T2g * T2k; T2i = T2g * T2h; T2o = Ip[WS(rs, 14)]; T2r = Im[WS(rs, 14)]; T4A = FNMS(T2j, T2h, T4z); T2l = FMA(T2j, T2k, T2i); T2n = W[56]; T2q = W[57]; } } T4B = T4y - T4A; T71 = T4y + T4A; T2m = T2f + T2l; T4w = T2f - T2l; T4E = T2n * T2r; T2p = T2n * T2o; T4F = FNMS(T2q, T2o, T4E); T2s = FMA(T2q, T2r, T2p); } } { E T4T, T4C, T4J, T4U, T7y, T8q, T8p, T7B; { E T6E, T8j, T73, T6Y, T6H, T8k, T8i, T8h; { E T7C, TO, T80, T7Z, T8e, T89, T8d, T1H, T8b, T3v, T7T, T7L, T7U, T7Q, T2A; E T7K, T7P, T7W, T1I; { E T7X, T7Y, T7J, T82, T88; { E Tm, T4I, T72, T4D, T2z, TN; T6E = T8 - Tl; Tm = T8 + Tl; T4T = T4B - T4w; T4C = T4w + T4B; T4I = T4F - T4H; T72 = T4F + T4H; T4D = T2s - T2y; T2z = T2s + T2y; TN = Tz + TM; T8j = TM - Tz; T73 = T71 - T72; T7J = T71 + T72; T4J = T4D - T4I; T4U = T4D + T4I; T2A = T2m + T2z; T6Y = T2z - T2m; T7C = Tm - TN; TO = Tm + TN; } T7K = T7I - T7J; T7X = T7I + T7J; T7Y = T7N + T7O; T7P = T7N - T7O; T6H = T6F - T6G; T82 = T6F + T6G; T88 = T83 + T87; T8k = T87 - T83; T80 = T7X + T7Y; T7Z = T7X - T7Y; T8e = T88 - T82; T89 = T82 + T88; } { E T7H, T7M, T2B, T3u; T7H = T29 - T2A; T2B = T29 + T2A; T3u = T32 + T3t; T7M = T32 - T3t; T8d = T1G - T1f; T1H = T1f + T1G; T8b = T3u - T2B; T3v = T2B + T3u; T7T = T7K - T7H; T7L = T7H + T7K; T7U = T7M + T7P; T7Q = T7M - T7P; } T7W = TO - T1H; T1I = TO + T1H; { E T7S, T8f, T8g, T7V; { E T7R, T8c, T8a, T7G, T81, T7F; T8i = T7Q - T7L; T7R = T7L + T7Q; T81 = T7D + T7E; T7F = T7D - T7E; Rp[0] = T1I + T3v; Rm[WS(rs, 15)] = T1I - T3v; Rp[WS(rs, 8)] = T7W + T7Z; Rm[WS(rs, 7)] = T7W - T7Z; T8c = T89 - T81; T8a = T81 + T89; T7G = T7C + T7F; T7S = T7C - T7F; T8h = T8e - T8d; T8f = T8d + T8e; Ip[WS(rs, 8)] = T8b + T8c; Im[WS(rs, 7)] = T8b - T8c; Ip[0] = T80 + T8a; Im[WS(rs, 15)] = T80 - T8a; Rp[WS(rs, 4)] = FMA(KP707106781, T7R, T7G); Rm[WS(rs, 11)] = FNMS(KP707106781, T7R, T7G); T8g = T7T + T7U; T7V = T7T - T7U; } Ip[WS(rs, 4)] = FMA(KP707106781, T8g, T8f); Im[WS(rs, 11)] = FMS(KP707106781, T8g, T8f); Rp[WS(rs, 12)] = FMA(KP707106781, T7V, T7S); Rm[WS(rs, 3)] = FNMS(KP707106781, T7V, T7S); } } { E T7f, T7m, T6I, T7a, T7A, T7w, T8r, T8l, T8m, T6T, T7j, T75, T8s, T7p, T7z; E T7t; { E T7n, T6N, T6S, T7o, T7u, T7v; T7f = T7b - T7e; T7u = T7b + T7e; Ip[WS(rs, 12)] = FMA(KP707106781, T8i, T8h); Im[WS(rs, 3)] = FMS(KP707106781, T8i, T8h); T7m = T6E + T6H; T6I = T6E - T6H; T7v = T79 + T76; T7a = T76 - T79; T7n = T6M + T6L; T6N = T6L - T6M; T7A = FMA(KP414213562, T7u, T7v); T7w = FNMS(KP414213562, T7v, T7u); T8r = T8k - T8j; T8l = T8j + T8k; T6S = T6O + T6R; T7o = T6O - T6R; { E T7s, T7r, T6Z, T74; T7s = T6X + T6Y; T6Z = T6X - T6Y; T74 = T70 - T73; T7r = T70 + T73; T8m = T6N + T6S; T6T = T6N - T6S; T7j = FNMS(KP414213562, T6Z, T74); T75 = FMA(KP414213562, T74, T6Z); T8s = T7o - T7n; T7p = T7n + T7o; T7z = FNMS(KP414213562, T7r, T7s); T7t = FMA(KP414213562, T7s, T7r); } } { E T7i, T6U, T8t, T8v, T7k, T7g; T7i = FNMS(KP707106781, T6T, T6I); T6U = FMA(KP707106781, T6T, T6I); T8t = FMA(KP707106781, T8s, T8r); T8v = FNMS(KP707106781, T8s, T8r); T7k = FNMS(KP414213562, T7a, T7f); T7g = FMA(KP414213562, T7f, T7a); { E T7q, T7x, T8n, T8o; T7y = FNMS(KP707106781, T7p, T7m); T7q = FMA(KP707106781, T7p, T7m); { E T7l, T8u, T8w, T7h; T7l = T7j + T7k; T8u = T7k - T7j; T8w = T7g - T75; T7h = T75 + T7g; Rm[WS(rs, 1)] = FMA(KP923879532, T7l, T7i); Rp[WS(rs, 14)] = FNMS(KP923879532, T7l, T7i); Ip[WS(rs, 6)] = FMA(KP923879532, T8u, T8t); Im[WS(rs, 9)] = FMS(KP923879532, T8u, T8t); Ip[WS(rs, 14)] = FMA(KP923879532, T8w, T8v); Im[WS(rs, 1)] = FMS(KP923879532, T8w, T8v); Rp[WS(rs, 6)] = FMA(KP923879532, T7h, T6U); Rm[WS(rs, 9)] = FNMS(KP923879532, T7h, T6U); T7x = T7t + T7w; T8q = T7w - T7t; } T8p = FNMS(KP707106781, T8m, T8l); T8n = FMA(KP707106781, T8m, T8l); T8o = T7z + T7A; T7B = T7z - T7A; Rp[WS(rs, 2)] = FMA(KP923879532, T7x, T7q); Rm[WS(rs, 13)] = FNMS(KP923879532, T7x, T7q); Ip[WS(rs, 2)] = FMA(KP923879532, T8o, T8n); Im[WS(rs, 13)] = FMS(KP923879532, T8o, T8n); } } } } { E T5S, T8O, T8N, T5V, T6g, T6d, T69, T66, T8K, T8J; { E T5C, T3S, T8I, T4n, T8H, T8B, T8C, T5F, T5k, T5K, T5u, T4K, T4V; { E T5D, T5E, T8z, T8A, T5j; { E T3C, T3R, T47, T4m; T5S = T3w - T3B; T3C = T3w + T3B; Rp[WS(rs, 10)] = FMA(KP923879532, T7B, T7y); Rm[WS(rs, 5)] = FNMS(KP923879532, T7B, T7y); Ip[WS(rs, 10)] = FMA(KP923879532, T8q, T8p); Im[WS(rs, 5)] = FMS(KP923879532, T8q, T8p); T3R = T3J + T3Q; T8O = T3Q - T3J; T5D = FNMS(KP414213562, T3Z, T46); T47 = FMA(KP414213562, T46, T3Z); T4m = FNMS(KP414213562, T4l, T4e); T5E = FMA(KP414213562, T4e, T4l); T8N = T8y + T8x; T8z = T8x - T8y; T5C = FNMS(KP707106781, T3R, T3C); T3S = FMA(KP707106781, T3R, T3C); T8I = T4m - T47; T4n = T47 + T4m; T8A = T5T + T5U; T5V = T5T - T5U; } T6g = T5i - T5b; T5j = T5b + T5i; T8H = FNMS(KP707106781, T8A, T8z); T8B = FMA(KP707106781, T8A, T8z); T8C = T5D + T5E; T5F = T5D - T5E; T5k = FMA(KP707106781, T5j, T54); T5K = FNMS(KP707106781, T5j, T54); T5u = T5s + T5t; T6d = T5t - T5s; T69 = T4C - T4J; T4K = T4C + T4J; T4V = T4T + T4U; T66 = T4U - T4T; } { E T5M, T5Q, T5J, T5P, T8F, T8G; { E T5y, T4o, T5A, T5w, T5z, T4X, T8D, T5L, T5v, T8E, T5B, T5x; T5y = FNMS(KP923879532, T4n, T3S); T4o = FMA(KP923879532, T4n, T3S); T5L = FNMS(KP707106781, T5u, T5r); T5v = FMA(KP707106781, T5u, T5r); { E T5H, T4L, T5I, T4W; T5H = FNMS(KP707106781, T4K, T4v); T4L = FMA(KP707106781, T4K, T4v); T5I = FNMS(KP707106781, T4V, T4S); T4W = FMA(KP707106781, T4V, T4S); T5M = FMA(KP668178637, T5L, T5K); T5Q = FNMS(KP668178637, T5K, T5L); T5A = FMA(KP198912367, T5k, T5v); T5w = FNMS(KP198912367, T5v, T5k); T5J = FNMS(KP668178637, T5I, T5H); T5P = FMA(KP668178637, T5H, T5I); T5z = FNMS(KP198912367, T4L, T4W); T4X = FMA(KP198912367, T4W, T4L); } T8D = FMA(KP923879532, T8C, T8B); T8F = FNMS(KP923879532, T8C, T8B); T8E = T5z + T5A; T5B = T5z - T5A; T8G = T5w - T4X; T5x = T4X + T5w; Ip[WS(rs, 1)] = FMA(KP980785280, T8E, T8D); Im[WS(rs, 14)] = FMS(KP980785280, T8E, T8D); Rp[WS(rs, 1)] = FMA(KP980785280, T5x, T4o); Rm[WS(rs, 14)] = FNMS(KP980785280, T5x, T4o); Rp[WS(rs, 9)] = FMA(KP980785280, T5B, T5y); Rm[WS(rs, 6)] = FNMS(KP980785280, T5B, T5y); } { E T5O, T8L, T8M, T5R, T5G, T5N; T5O = FMA(KP923879532, T5F, T5C); T5G = FNMS(KP923879532, T5F, T5C); T5N = T5J + T5M; T8K = T5M - T5J; T8J = FMA(KP923879532, T8I, T8H); T8L = FNMS(KP923879532, T8I, T8H); Ip[WS(rs, 9)] = FMA(KP980785280, T8G, T8F); Im[WS(rs, 6)] = FMS(KP980785280, T8G, T8F); Rm[WS(rs, 2)] = FMA(KP831469612, T5N, T5G); Rp[WS(rs, 13)] = FNMS(KP831469612, T5N, T5G); T8M = T5P + T5Q; T5R = T5P - T5Q; Ip[WS(rs, 13)] = FNMS(KP831469612, T8M, T8L); Im[WS(rs, 2)] = -(FMA(KP831469612, T8M, T8L)); Rp[WS(rs, 5)] = FMA(KP831469612, T5R, T5O); Rm[WS(rs, 10)] = FNMS(KP831469612, T5R, T5O); } } } { E T6o, T5W, T8W, T63, T8V, T8P, T8Q, T6r, T67, T6u, T6y, T6C, T6m, T6i; { E T6p, T5Z, T62, T6q; T6p = FNMS(KP414213562, T5X, T5Y); T5Z = FMA(KP414213562, T5Y, T5X); Ip[WS(rs, 5)] = FMA(KP831469612, T8K, T8J); Im[WS(rs, 10)] = FMS(KP831469612, T8K, T8J); T6o = FNMS(KP707106781, T5V, T5S); T5W = FMA(KP707106781, T5V, T5S); T62 = FNMS(KP414213562, T61, T60); T6q = FMA(KP414213562, T60, T61); T8W = T5Z + T62; T63 = T5Z - T62; T8V = FNMS(KP707106781, T8O, T8N); T8P = FMA(KP707106781, T8O, T8N); { E T6x, T6e, T6w, T6h; T8Q = T6q - T6p; T6r = T6p + T6q; T6x = FMA(KP707106781, T6d, T6c); T6e = FNMS(KP707106781, T6d, T6c); T6w = FMA(KP707106781, T6g, T6f); T6h = FNMS(KP707106781, T6g, T6f); T67 = FNMS(KP707106781, T66, T65); T6u = FMA(KP707106781, T66, T65); T6y = FMA(KP198912367, T6x, T6w); T6C = FNMS(KP198912367, T6w, T6x); T6m = FNMS(KP668178637, T6e, T6h); T6i = FMA(KP668178637, T6h, T6e); } } { E T6k, T64, T8R, T8T, T6t, T6a; T6k = FNMS(KP923879532, T63, T5W); T64 = FMA(KP923879532, T63, T5W); T8R = FMA(KP923879532, T8Q, T8P); T8T = FNMS(KP923879532, T8Q, T8P); T6t = FMA(KP707106781, T69, T68); T6a = FNMS(KP707106781, T69, T68); { E T6A, T8X, T8Y, T6D; { E T6s, T6B, T6l, T6b, T6z, T6v; T6A = FMA(KP923879532, T6r, T6o); T6s = FNMS(KP923879532, T6r, T6o); T6v = FMA(KP198912367, T6u, T6t); T6B = FNMS(KP198912367, T6t, T6u); T6l = FNMS(KP668178637, T67, T6a); T6b = FMA(KP668178637, T6a, T67); T6z = T6v + T6y; T90 = T6y - T6v; T8Z = FMA(KP923879532, T8W, T8V); T8X = FNMS(KP923879532, T8W, T8V); { E T6n, T8S, T8U, T6j; T6n = T6l + T6m; T8S = T6l - T6m; T8U = T6i - T6b; T6j = T6b + T6i; Rp[WS(rs, 7)] = FMA(KP980785280, T6z, T6s); Rm[WS(rs, 8)] = FNMS(KP980785280, T6z, T6s); Rp[WS(rs, 11)] = FMA(KP831469612, T6n, T6k); Rm[WS(rs, 4)] = FNMS(KP831469612, T6n, T6k); Ip[WS(rs, 3)] = FMA(KP831469612, T8S, T8R); Im[WS(rs, 12)] = FMS(KP831469612, T8S, T8R); Ip[WS(rs, 11)] = FMA(KP831469612, T8U, T8T); Im[WS(rs, 4)] = FMS(KP831469612, T8U, T8T); Rp[WS(rs, 3)] = FMA(KP831469612, T6j, T64); Rm[WS(rs, 12)] = FNMS(KP831469612, T6j, T64); T8Y = T6C - T6B; T6D = T6B + T6C; } } Ip[WS(rs, 7)] = FMA(KP980785280, T8Y, T8X); Im[WS(rs, 8)] = FMS(KP980785280, T8Y, T8X); Rm[0] = FMA(KP980785280, T6D, T6A); Rp[WS(rs, 15)] = FNMS(KP980785280, T6D, T6A); } } } } } } Ip[WS(rs, 15)] = FMA(KP980785280, T90, T8Z); Im[0] = FMS(KP980785280, T90, T8Z); } } } static const tw_instr twinstr[] = { {TW_FULL, 1, 32}, {TW_NEXT, 1, 0} }; static const hc2c_desc desc = { 32, "hc2cf_32", twinstr, &GENUS, {236, 62, 198, 0} }; void X(codelet_hc2cf_32) (planner *p) { X(khc2c_register) (p, hc2cf_32, &desc, HC2C_VIA_RDFT); } #else /* HAVE_FMA */ /* Generated by: ../../../genfft/gen_hc2c.native -compact -variables 4 -pipeline-latency 4 -n 32 -dit -name hc2cf_32 -include hc2cf.h */ /* * This function contains 434 FP additions, 208 FP multiplications, * (or, 340 additions, 114 multiplications, 94 fused multiply/add), * 96 stack variables, 7 constants, and 128 memory accesses */ #include "hc2cf.h" static void hc2cf_32(R *Rp, R *Ip, R *Rm, R *Im, const R *W, stride rs, INT mb, INT me, INT ms) { DK(KP195090322, +0.195090322016128267848284868477022240927691618); DK(KP980785280, +0.980785280403230449126182236134239036973933731); DK(KP555570233, +0.555570233019602224742830813948532874374937191); DK(KP831469612, +0.831469612302545237078788377617905756738560812); DK(KP382683432, +0.382683432365089771728459984030398866761344562); DK(KP923879532, +0.923879532511286756128183189396788286822416626); DK(KP707106781, +0.707106781186547524400844362104849039284835938); { INT m; for (m = mb, W = W + ((mb - 1) * 62); m < me; m = m + 1, Rp = Rp + ms, Ip = Ip + ms, Rm = Rm - ms, Im = Im - ms, W = W + 62, MAKE_VOLATILE_STRIDE(128, rs)) { E Tj, T5F, T7C, T7Q, T35, T4T, T78, T7m, T1Q, T61, T5Y, T6J, T3K, T59, T41; E T56, T2B, T67, T6e, T6O, T4b, T5d, T4s, T5g, TG, T7l, T5I, T73, T3a, T4U; E T3f, T4V, T14, T5N, T5M, T6E, T3m, T4Y, T3r, T4Z, T1r, T5P, T5S, T6F, T3x; E T51, T3C, T52, T2d, T5Z, T64, T6K, T3V, T57, T44, T5a, T2Y, T6f, T6a, T6P; E T4m, T5h, T4v, T5e; { E T1, T76, T6, T75, Tc, T32, Th, T33; T1 = Rp[0]; T76 = Rm[0]; { E T3, T5, T2, T4; T3 = Rp[WS(rs, 8)]; T5 = Rm[WS(rs, 8)]; T2 = W[30]; T4 = W[31]; T6 = FMA(T2, T3, T4 * T5); T75 = FNMS(T4, T3, T2 * T5); } { E T9, Tb, T8, Ta; T9 = Rp[WS(rs, 4)]; Tb = Rm[WS(rs, 4)]; T8 = W[14]; Ta = W[15]; Tc = FMA(T8, T9, Ta * Tb); T32 = FNMS(Ta, T9, T8 * Tb); } { E Te, Tg, Td, Tf; Te = Rp[WS(rs, 12)]; Tg = Rm[WS(rs, 12)]; Td = W[46]; Tf = W[47]; Th = FMA(Td, Te, Tf * Tg); T33 = FNMS(Tf, Te, Td * Tg); } { E T7, Ti, T7A, T7B; T7 = T1 + T6; Ti = Tc + Th; Tj = T7 + Ti; T5F = T7 - Ti; T7A = T76 - T75; T7B = Tc - Th; T7C = T7A - T7B; T7Q = T7B + T7A; } { E T31, T34, T74, T77; T31 = T1 - T6; T34 = T32 - T33; T35 = T31 - T34; T4T = T31 + T34; T74 = T32 + T33; T77 = T75 + T76; T78 = T74 + T77; T7m = T77 - T74; } } { E T1y, T3G, T1O, T3Z, T1D, T3H, T1J, T3Y; { E T1v, T1x, T1u, T1w; T1v = Ip[0]; T1x = Im[0]; T1u = W[0]; T1w = W[1]; T1y = FMA(T1u, T1v, T1w * T1x); T3G = FNMS(T1w, T1v, T1u * T1x); } { E T1L, T1N, T1K, T1M; T1L = Ip[WS(rs, 12)]; T1N = Im[WS(rs, 12)]; T1K = W[48]; T1M = W[49]; T1O = FMA(T1K, T1L, T1M * T1N); T3Z = FNMS(T1M, T1L, T1K * T1N); } { E T1A, T1C, T1z, T1B; T1A = Ip[WS(rs, 8)]; T1C = Im[WS(rs, 8)]; T1z = W[32]; T1B = W[33]; T1D = FMA(T1z, T1A, T1B * T1C); T3H = FNMS(T1B, T1A, T1z * T1C); } { E T1G, T1I, T1F, T1H; T1G = Ip[WS(rs, 4)]; T1I = Im[WS(rs, 4)]; T1F = W[16]; T1H = W[17]; T1J = FMA(T1F, T1G, T1H * T1I); T3Y = FNMS(T1H, T1G, T1F * T1I); } { E T1E, T1P, T5W, T5X; T1E = T1y + T1D; T1P = T1J + T1O; T1Q = T1E + T1P; T61 = T1E - T1P; T5W = T3G + T3H; T5X = T3Y + T3Z; T5Y = T5W - T5X; T6J = T5W + T5X; } { E T3I, T3J, T3X, T40; T3I = T3G - T3H; T3J = T1J - T1O; T3K = T3I + T3J; T59 = T3I - T3J; T3X = T1y - T1D; T40 = T3Y - T3Z; T41 = T3X - T40; T56 = T3X + T40; } } { E T2j, T4o, T2z, T49, T2o, T4p, T2u, T48; { E T2g, T2i, T2f, T2h; T2g = Ip[WS(rs, 15)]; T2i = Im[WS(rs, 15)]; T2f = W[60]; T2h = W[61]; T2j = FMA(T2f, T2g, T2h * T2i); T4o = FNMS(T2h, T2g, T2f * T2i); } { E T2w, T2y, T2v, T2x; T2w = Ip[WS(rs, 11)]; T2y = Im[WS(rs, 11)]; T2v = W[44]; T2x = W[45]; T2z = FMA(T2v, T2w, T2x * T2y); T49 = FNMS(T2x, T2w, T2v * T2y); } { E T2l, T2n, T2k, T2m; T2l = Ip[WS(rs, 7)]; T2n = Im[WS(rs, 7)]; T2k = W[28]; T2m = W[29]; T2o = FMA(T2k, T2l, T2m * T2n); T4p = FNMS(T2m, T2l, T2k * T2n); } { E T2r, T2t, T2q, T2s; T2r = Ip[WS(rs, 3)]; T2t = Im[WS(rs, 3)]; T2q = W[12]; T2s = W[13]; T2u = FMA(T2q, T2r, T2s * T2t); T48 = FNMS(T2s, T2r, T2q * T2t); } { E T2p, T2A, T6c, T6d; T2p = T2j + T2o; T2A = T2u + T2z; T2B = T2p + T2A; T67 = T2p - T2A; T6c = T4o + T4p; T6d = T48 + T49; T6e = T6c - T6d; T6O = T6c + T6d; } { E T47, T4a, T4q, T4r; T47 = T2j - T2o; T4a = T48 - T49; T4b = T47 - T4a; T5d = T47 + T4a; T4q = T4o - T4p; T4r = T2u - T2z; T4s = T4q + T4r; T5g = T4q - T4r; } } { E To, T36, TE, T3d, Tt, T37, Tz, T3c; { E Tl, Tn, Tk, Tm; Tl = Rp[WS(rs, 2)]; Tn = Rm[WS(rs, 2)]; Tk = W[6]; Tm = W[7]; To = FMA(Tk, Tl, Tm * Tn); T36 = FNMS(Tm, Tl, Tk * Tn); } { E TB, TD, TA, TC; TB = Rp[WS(rs, 6)]; TD = Rm[WS(rs, 6)]; TA = W[22]; TC = W[23]; TE = FMA(TA, TB, TC * TD); T3d = FNMS(TC, TB, TA * TD); } { E Tq, Ts, Tp, Tr; Tq = Rp[WS(rs, 10)]; Ts = Rm[WS(rs, 10)]; Tp = W[38]; Tr = W[39]; Tt = FMA(Tp, Tq, Tr * Ts); T37 = FNMS(Tr, Tq, Tp * Ts); } { E Tw, Ty, Tv, Tx; Tw = Rp[WS(rs, 14)]; Ty = Rm[WS(rs, 14)]; Tv = W[54]; Tx = W[55]; Tz = FMA(Tv, Tw, Tx * Ty); T3c = FNMS(Tx, Tw, Tv * Ty); } { E Tu, TF, T5G, T5H; Tu = To + Tt; TF = Tz + TE; TG = Tu + TF; T7l = TF - Tu; T5G = T36 + T37; T5H = T3c + T3d; T5I = T5G - T5H; T73 = T5G + T5H; } { E T38, T39, T3b, T3e; T38 = T36 - T37; T39 = To - Tt; T3a = T38 - T39; T4U = T39 + T38; T3b = Tz - TE; T3e = T3c - T3d; T3f = T3b + T3e; T4V = T3b - T3e; } } { E TM, T3i, T12, T3p, TR, T3j, TX, T3o; { E TJ, TL, TI, TK; TJ = Rp[WS(rs, 1)]; TL = Rm[WS(rs, 1)]; TI = W[2]; TK = W[3]; TM = FMA(TI, TJ, TK * TL); T3i = FNMS(TK, TJ, TI * TL); } { E TZ, T11, TY, T10; TZ = Rp[WS(rs, 13)]; T11 = Rm[WS(rs, 13)]; TY = W[50]; T10 = W[51]; T12 = FMA(TY, TZ, T10 * T11); T3p = FNMS(T10, TZ, TY * T11); } { E TO, TQ, TN, TP; TO = Rp[WS(rs, 9)]; TQ = Rm[WS(rs, 9)]; TN = W[34]; TP = W[35]; TR = FMA(TN, TO, TP * TQ); T3j = FNMS(TP, TO, TN * TQ); } { E TU, TW, TT, TV; TU = Rp[WS(rs, 5)]; TW = Rm[WS(rs, 5)]; TT = W[18]; TV = W[19]; TX = FMA(TT, TU, TV * TW); T3o = FNMS(TV, TU, TT * TW); } { E TS, T13, T5K, T5L; TS = TM + TR; T13 = TX + T12; T14 = TS + T13; T5N = TS - T13; T5K = T3i + T3j; T5L = T3o + T3p; T5M = T5K - T5L; T6E = T5K + T5L; } { E T3k, T3l, T3n, T3q; T3k = T3i - T3j; T3l = TX - T12; T3m = T3k + T3l; T4Y = T3k - T3l; T3n = TM - TR; T3q = T3o - T3p; T3r = T3n - T3q; T4Z = T3n + T3q; } } { E T19, T3t, T1p, T3A, T1e, T3u, T1k, T3z; { E T16, T18, T15, T17; T16 = Rp[WS(rs, 15)]; T18 = Rm[WS(rs, 15)]; T15 = W[58]; T17 = W[59]; T19 = FMA(T15, T16, T17 * T18); T3t = FNMS(T17, T16, T15 * T18); } { E T1m, T1o, T1l, T1n; T1m = Rp[WS(rs, 11)]; T1o = Rm[WS(rs, 11)]; T1l = W[42]; T1n = W[43]; T1p = FMA(T1l, T1m, T1n * T1o); T3A = FNMS(T1n, T1m, T1l * T1o); } { E T1b, T1d, T1a, T1c; T1b = Rp[WS(rs, 7)]; T1d = Rm[WS(rs, 7)]; T1a = W[26]; T1c = W[27]; T1e = FMA(T1a, T1b, T1c * T1d); T3u = FNMS(T1c, T1b, T1a * T1d); } { E T1h, T1j, T1g, T1i; T1h = Rp[WS(rs, 3)]; T1j = Rm[WS(rs, 3)]; T1g = W[10]; T1i = W[11]; T1k = FMA(T1g, T1h, T1i * T1j); T3z = FNMS(T1i, T1h, T1g * T1j); } { E T1f, T1q, T5Q, T5R; T1f = T19 + T1e; T1q = T1k + T1p; T1r = T1f + T1q; T5P = T1f - T1q; T5Q = T3t + T3u; T5R = T3z + T3A; T5S = T5Q - T5R; T6F = T5Q + T5R; } { E T3v, T3w, T3y, T3B; T3v = T3t - T3u; T3w = T1k - T1p; T3x = T3v + T3w; T51 = T3v - T3w; T3y = T19 - T1e; T3B = T3z - T3A; T3C = T3y - T3B; T52 = T3y + T3B; } } { E T1V, T3R, T20, T3S, T3Q, T3T, T26, T3M, T2b, T3N, T3L, T3O; { E T1S, T1U, T1R, T1T; T1S = Ip[WS(rs, 2)]; T1U = Im[WS(rs, 2)]; T1R = W[8]; T1T = W[9]; T1V = FMA(T1R, T1S, T1T * T1U); T3R = FNMS(T1T, T1S, T1R * T1U); } { E T1X, T1Z, T1W, T1Y; T1X = Ip[WS(rs, 10)]; T1Z = Im[WS(rs, 10)]; T1W = W[40]; T1Y = W[41]; T20 = FMA(T1W, T1X, T1Y * T1Z); T3S = FNMS(T1Y, T1X, T1W * T1Z); } T3Q = T1V - T20; T3T = T3R - T3S; { E T23, T25, T22, T24; T23 = Ip[WS(rs, 14)]; T25 = Im[WS(rs, 14)]; T22 = W[56]; T24 = W[57]; T26 = FMA(T22, T23, T24 * T25); T3M = FNMS(T24, T23, T22 * T25); } { E T28, T2a, T27, T29; T28 = Ip[WS(rs, 6)]; T2a = Im[WS(rs, 6)]; T27 = W[24]; T29 = W[25]; T2b = FMA(T27, T28, T29 * T2a); T3N = FNMS(T29, T28, T27 * T2a); } T3L = T26 - T2b; T3O = T3M - T3N; { E T21, T2c, T62, T63; T21 = T1V + T20; T2c = T26 + T2b; T2d = T21 + T2c; T5Z = T2c - T21; T62 = T3R + T3S; T63 = T3M + T3N; T64 = T62 - T63; T6K = T62 + T63; } { E T3P, T3U, T42, T43; T3P = T3L - T3O; T3U = T3Q + T3T; T3V = KP707106781 * (T3P - T3U); T57 = KP707106781 * (T3U + T3P); T42 = T3T - T3Q; T43 = T3L + T3O; T44 = KP707106781 * (T42 - T43); T5a = KP707106781 * (T42 + T43); } } { E T2G, T4c, T2L, T4d, T4e, T4f, T2R, T4i, T2W, T4j, T4h, T4k; { E T2D, T2F, T2C, T2E; T2D = Ip[WS(rs, 1)]; T2F = Im[WS(rs, 1)]; T2C = W[4]; T2E = W[5]; T2G = FMA(T2C, T2D, T2E * T2F); T4c = FNMS(T2E, T2D, T2C * T2F); } { E T2I, T2K, T2H, T2J; T2I = Ip[WS(rs, 9)]; T2K = Im[WS(rs, 9)]; T2H = W[36]; T2J = W[37]; T2L = FMA(T2H, T2I, T2J * T2K); T4d = FNMS(T2J, T2I, T2H * T2K); } T4e = T4c - T4d; T4f = T2G - T2L; { E T2O, T2Q, T2N, T2P; T2O = Ip[WS(rs, 13)]; T2Q = Im[WS(rs, 13)]; T2N = W[52]; T2P = W[53]; T2R = FMA(T2N, T2O, T2P * T2Q); T4i = FNMS(T2P, T2O, T2N * T2Q); } { E T2T, T2V, T2S, T2U; T2T = Ip[WS(rs, 5)]; T2V = Im[WS(rs, 5)]; T2S = W[20]; T2U = W[21]; T2W = FMA(T2S, T2T, T2U * T2V); T4j = FNMS(T2U, T2T, T2S * T2V); } T4h = T2R - T2W; T4k = T4i - T4j; { E T2M, T2X, T68, T69; T2M = T2G + T2L; T2X = T2R + T2W; T2Y = T2M + T2X; T6f = T2X - T2M; T68 = T4c + T4d; T69 = T4i + T4j; T6a = T68 - T69; T6P = T68 + T69; } { E T4g, T4l, T4t, T4u; T4g = T4e - T4f; T4l = T4h + T4k; T4m = KP707106781 * (T4g - T4l); T5h = KP707106781 * (T4g + T4l); T4t = T4h - T4k; T4u = T4f + T4e; T4v = KP707106781 * (T4t - T4u); T5e = KP707106781 * (T4u + T4t); } } { E T1t, T6X, T7a, T7c, T30, T7b, T70, T71; { E TH, T1s, T72, T79; TH = Tj + TG; T1s = T14 + T1r; T1t = TH + T1s; T6X = TH - T1s; T72 = T6E + T6F; T79 = T73 + T78; T7a = T72 + T79; T7c = T79 - T72; } { E T2e, T2Z, T6Y, T6Z; T2e = T1Q + T2d; T2Z = T2B + T2Y; T30 = T2e + T2Z; T7b = T2Z - T2e; T6Y = T6J + T6K; T6Z = T6O + T6P; T70 = T6Y - T6Z; T71 = T6Y + T6Z; } Rm[WS(rs, 15)] = T1t - T30; Im[WS(rs, 15)] = T71 - T7a; Rp[0] = T1t + T30; Ip[0] = T71 + T7a; Rm[WS(rs, 7)] = T6X - T70; Im[WS(rs, 7)] = T7b - T7c; Rp[WS(rs, 8)] = T6X + T70; Ip[WS(rs, 8)] = T7b + T7c; } { E T6H, T6T, T7g, T7i, T6M, T6U, T6R, T6V; { E T6D, T6G, T7e, T7f; T6D = Tj - TG; T6G = T6E - T6F; T6H = T6D + T6G; T6T = T6D - T6G; T7e = T1r - T14; T7f = T78 - T73; T7g = T7e + T7f; T7i = T7f - T7e; } { E T6I, T6L, T6N, T6Q; T6I = T1Q - T2d; T6L = T6J - T6K; T6M = T6I + T6L; T6U = T6L - T6I; T6N = T2B - T2Y; T6Q = T6O - T6P; T6R = T6N - T6Q; T6V = T6N + T6Q; } { E T6S, T7d, T6W, T7h; T6S = KP707106781 * (T6M + T6R); Rm[WS(rs, 11)] = T6H - T6S; Rp[WS(rs, 4)] = T6H + T6S; T7d = KP707106781 * (T6U + T6V); Im[WS(rs, 11)] = T7d - T7g; Ip[WS(rs, 4)] = T7d + T7g; T6W = KP707106781 * (T6U - T6V); Rm[WS(rs, 3)] = T6T - T6W; Rp[WS(rs, 12)] = T6T + T6W; T7h = KP707106781 * (T6R - T6M); Im[WS(rs, 3)] = T7h - T7i; Ip[WS(rs, 12)] = T7h + T7i; } } { E T5J, T7n, T7t, T6n, T5U, T7k, T6x, T6B, T6q, T7s, T66, T6k, T6u, T6A, T6h; E T6l; { E T5O, T5T, T60, T65; T5J = T5F - T5I; T7n = T7l + T7m; T7t = T7m - T7l; T6n = T5F + T5I; T5O = T5M - T5N; T5T = T5P + T5S; T5U = KP707106781 * (T5O - T5T); T7k = KP707106781 * (T5O + T5T); { E T6v, T6w, T6o, T6p; T6v = T67 + T6a; T6w = T6e + T6f; T6x = FNMS(KP382683432, T6w, KP923879532 * T6v); T6B = FMA(KP923879532, T6w, KP382683432 * T6v); T6o = T5N + T5M; T6p = T5P - T5S; T6q = KP707106781 * (T6o + T6p); T7s = KP707106781 * (T6p - T6o); } T60 = T5Y - T5Z; T65 = T61 - T64; T66 = FMA(KP923879532, T60, KP382683432 * T65); T6k = FNMS(KP923879532, T65, KP382683432 * T60); { E T6s, T6t, T6b, T6g; T6s = T5Y + T5Z; T6t = T61 + T64; T6u = FMA(KP382683432, T6s, KP923879532 * T6t); T6A = FNMS(KP382683432, T6t, KP923879532 * T6s); T6b = T67 - T6a; T6g = T6e - T6f; T6h = FNMS(KP923879532, T6g, KP382683432 * T6b); T6l = FMA(KP382683432, T6g, KP923879532 * T6b); } } { E T5V, T6i, T7r, T7u; T5V = T5J + T5U; T6i = T66 + T6h; Rm[WS(rs, 9)] = T5V - T6i; Rp[WS(rs, 6)] = T5V + T6i; T7r = T6k + T6l; T7u = T7s + T7t; Im[WS(rs, 9)] = T7r - T7u; Ip[WS(rs, 6)] = T7r + T7u; } { E T6j, T6m, T7v, T7w; T6j = T5J - T5U; T6m = T6k - T6l; Rm[WS(rs, 1)] = T6j - T6m; Rp[WS(rs, 14)] = T6j + T6m; T7v = T6h - T66; T7w = T7t - T7s; Im[WS(rs, 1)] = T7v - T7w; Ip[WS(rs, 14)] = T7v + T7w; } { E T6r, T6y, T7j, T7o; T6r = T6n + T6q; T6y = T6u + T6x; Rm[WS(rs, 13)] = T6r - T6y; Rp[WS(rs, 2)] = T6r + T6y; T7j = T6A + T6B; T7o = T7k + T7n; Im[WS(rs, 13)] = T7j - T7o; Ip[WS(rs, 2)] = T7j + T7o; } { E T6z, T6C, T7p, T7q; T6z = T6n - T6q; T6C = T6A - T6B; Rm[WS(rs, 5)] = T6z - T6C; Rp[WS(rs, 10)] = T6z + T6C; T7p = T6x - T6u; T7q = T7n - T7k; Im[WS(rs, 5)] = T7p - T7q; Ip[WS(rs, 10)] = T7p + T7q; } } { E T3h, T4D, T7R, T7X, T3E, T7O, T4N, T4R, T46, T4A, T4G, T7W, T4K, T4Q, T4x; E T4B, T3g, T7P; T3g = KP707106781 * (T3a - T3f); T3h = T35 - T3g; T4D = T35 + T3g; T7P = KP707106781 * (T4V - T4U); T7R = T7P + T7Q; T7X = T7Q - T7P; { E T3s, T3D, T4L, T4M; T3s = FNMS(KP923879532, T3r, KP382683432 * T3m); T3D = FMA(KP382683432, T3x, KP923879532 * T3C); T3E = T3s - T3D; T7O = T3s + T3D; T4L = T4b + T4m; T4M = T4s + T4v; T4N = FNMS(KP555570233, T4M, KP831469612 * T4L); T4R = FMA(KP831469612, T4M, KP555570233 * T4L); } { E T3W, T45, T4E, T4F; T3W = T3K - T3V; T45 = T41 - T44; T46 = FMA(KP980785280, T3W, KP195090322 * T45); T4A = FNMS(KP980785280, T45, KP195090322 * T3W); T4E = FMA(KP923879532, T3m, KP382683432 * T3r); T4F = FNMS(KP923879532, T3x, KP382683432 * T3C); T4G = T4E + T4F; T7W = T4F - T4E; } { E T4I, T4J, T4n, T4w; T4I = T3K + T3V; T4J = T41 + T44; T4K = FMA(KP555570233, T4I, KP831469612 * T4J); T4Q = FNMS(KP555570233, T4J, KP831469612 * T4I); T4n = T4b - T4m; T4w = T4s - T4v; T4x = FNMS(KP980785280, T4w, KP195090322 * T4n); T4B = FMA(KP195090322, T4w, KP980785280 * T4n); } { E T3F, T4y, T7V, T7Y; T3F = T3h + T3E; T4y = T46 + T4x; Rm[WS(rs, 8)] = T3F - T4y; Rp[WS(rs, 7)] = T3F + T4y; T7V = T4A + T4B; T7Y = T7W + T7X; Im[WS(rs, 8)] = T7V - T7Y; Ip[WS(rs, 7)] = T7V + T7Y; } { E T4z, T4C, T7Z, T80; T4z = T3h - T3E; T4C = T4A - T4B; Rm[0] = T4z - T4C; Rp[WS(rs, 15)] = T4z + T4C; T7Z = T4x - T46; T80 = T7X - T7W; Im[0] = T7Z - T80; Ip[WS(rs, 15)] = T7Z + T80; } { E T4H, T4O, T7N, T7S; T4H = T4D + T4G; T4O = T4K + T4N; Rm[WS(rs, 12)] = T4H - T4O; Rp[WS(rs, 3)] = T4H + T4O; T7N = T4Q + T4R; T7S = T7O + T7R; Im[WS(rs, 12)] = T7N - T7S; Ip[WS(rs, 3)] = T7N + T7S; } { E T4P, T4S, T7T, T7U; T4P = T4D - T4G; T4S = T4Q - T4R; Rm[WS(rs, 4)] = T4P - T4S; Rp[WS(rs, 11)] = T4P + T4S; T7T = T4N - T4K; T7U = T7R - T7O; Im[WS(rs, 4)] = T7T - T7U; Ip[WS(rs, 11)] = T7T + T7U; } } { E T4X, T5p, T7D, T7J, T54, T7y, T5z, T5D, T5c, T5m, T5s, T7I, T5w, T5C, T5j; E T5n, T4W, T7z; T4W = KP707106781 * (T4U + T4V); T4X = T4T - T4W; T5p = T4T + T4W; T7z = KP707106781 * (T3a + T3f); T7D = T7z + T7C; T7J = T7C - T7z; { E T50, T53, T5x, T5y; T50 = FNMS(KP382683432, T4Z, KP923879532 * T4Y); T53 = FMA(KP923879532, T51, KP382683432 * T52); T54 = T50 - T53; T7y = T50 + T53; T5x = T5d + T5e; T5y = T5g + T5h; T5z = FNMS(KP195090322, T5y, KP980785280 * T5x); T5D = FMA(KP195090322, T5x, KP980785280 * T5y); } { E T58, T5b, T5q, T5r; T58 = T56 - T57; T5b = T59 - T5a; T5c = FMA(KP555570233, T58, KP831469612 * T5b); T5m = FNMS(KP831469612, T58, KP555570233 * T5b); T5q = FMA(KP382683432, T4Y, KP923879532 * T4Z); T5r = FNMS(KP382683432, T51, KP923879532 * T52); T5s = T5q + T5r; T7I = T5r - T5q; } { E T5u, T5v, T5f, T5i; T5u = T56 + T57; T5v = T59 + T5a; T5w = FMA(KP980785280, T5u, KP195090322 * T5v); T5C = FNMS(KP195090322, T5u, KP980785280 * T5v); T5f = T5d - T5e; T5i = T5g - T5h; T5j = FNMS(KP831469612, T5i, KP555570233 * T5f); T5n = FMA(KP831469612, T5f, KP555570233 * T5i); } { E T55, T5k, T7H, T7K; T55 = T4X + T54; T5k = T5c + T5j; Rm[WS(rs, 10)] = T55 - T5k; Rp[WS(rs, 5)] = T55 + T5k; T7H = T5m + T5n; T7K = T7I + T7J; Im[WS(rs, 10)] = T7H - T7K; Ip[WS(rs, 5)] = T7H + T7K; } { E T5l, T5o, T7L, T7M; T5l = T4X - T54; T5o = T5m - T5n; Rm[WS(rs, 2)] = T5l - T5o; Rp[WS(rs, 13)] = T5l + T5o; T7L = T5j - T5c; T7M = T7J - T7I; Im[WS(rs, 2)] = T7L - T7M; Ip[WS(rs, 13)] = T7L + T7M; } { E T5t, T5A, T7x, T7E; T5t = T5p + T5s; T5A = T5w + T5z; Rm[WS(rs, 14)] = T5t - T5A; Rp[WS(rs, 1)] = T5t + T5A; T7x = T5C + T5D; T7E = T7y + T7D; Im[WS(rs, 14)] = T7x - T7E; Ip[WS(rs, 1)] = T7x + T7E; } { E T5B, T5E, T7F, T7G; T5B = T5p - T5s; T5E = T5C - T5D; Rm[WS(rs, 6)] = T5B - T5E; Rp[WS(rs, 9)] = T5B + T5E; T7F = T5z - T5w; T7G = T7D - T7y; Im[WS(rs, 6)] = T7F - T7G; Ip[WS(rs, 9)] = T7F + T7G; } } } } } static const tw_instr twinstr[] = { {TW_FULL, 1, 32}, {TW_NEXT, 1, 0} }; static const hc2c_desc desc = { 32, "hc2cf_32", twinstr, &GENUS, {340, 114, 94, 0} }; void X(codelet_hc2cf_32) (planner *p) { X(khc2c_register) (p, hc2cf_32, &desc, HC2C_VIA_RDFT); } #endif /* HAVE_FMA */ fftw-3.3.3/rdft/scalar/r2cf/hc2cf_20.c0000644000175000017500000006676012054410712014114 00000000000000/* * Copyright (c) 2003, 2007-11 Matteo Frigo * Copyright (c) 2003, 2007-11 Massachusetts Institute of Technology * * This program is free software; you can redistribute it and/or modify * it under the terms of the GNU General Public License as published by * the Free Software Foundation; either version 2 of the License, or * (at your option) any later version. * * This program is distributed in the hope that it will be useful, * but WITHOUT ANY WARRANTY; without even the implied warranty of * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the * GNU General Public License for more details. * * You should have received a copy of the GNU General Public License * along with this program; if not, write to the Free Software * Foundation, Inc., 51 Franklin Street, Fifth Floor, Boston, MA 02110-1301 USA * */ /* This file was automatically generated --- DO NOT EDIT */ /* Generated on Sun Nov 25 07:40:41 EST 2012 */ #include "codelet-rdft.h" #ifdef HAVE_FMA /* Generated by: ../../../genfft/gen_hc2c.native -fma -reorder-insns -schedule-for-pipeline -compact -variables 4 -pipeline-latency 4 -n 20 -dit -name hc2cf_20 -include hc2cf.h */ /* * This function contains 246 FP additions, 148 FP multiplications, * (or, 136 additions, 38 multiplications, 110 fused multiply/add), * 97 stack variables, 4 constants, and 80 memory accesses */ #include "hc2cf.h" static void hc2cf_20(R *Rp, R *Ip, R *Rm, R *Im, const R *W, stride rs, INT mb, INT me, INT ms) { DK(KP951056516, +0.951056516295153572116439333379382143405698634); DK(KP559016994, +0.559016994374947424102293417182819058860154590); DK(KP250000000, +0.250000000000000000000000000000000000000000000); DK(KP618033988, +0.618033988749894848204586834365638117720309180); { INT m; for (m = mb, W = W + ((mb - 1) * 38); m < me; m = m + 1, Rp = Rp + ms, Ip = Ip + ms, Rm = Rm - ms, Im = Im - ms, W = W + 38, MAKE_VOLATILE_STRIDE(80, rs)) { E T4P, T4Y, T50, T4U, T4S, T4T, T4Z, T4V; { E T4N, T4r, T8, T2i, T4n, T2n, T4O, Tl, T2v, T3v, T3T, T4f, TN, T2b, T3F; E T3p, T2R, T3z, T43, T4b, T27, T2f, T3J, T33, T2K, T3y, T40, T4c, T1G, T2e; E T3I, T3a, T2C, T3w, T3W, T4e, T1e, T2c, T3G, T3i; { E T1, T4q, T3, T6, T2, T5; T1 = Rp[0]; T4q = Rm[0]; T3 = Rp[WS(rs, 5)]; T6 = Rm[WS(rs, 5)]; T2 = W[18]; T5 = W[19]; { E Ta, Td, Tg, T2j, Tb, Tj, Tf, Tc, Ti; { E T4o, T4, T9, T4p, T7; Ta = Ip[WS(rs, 2)]; Td = Im[WS(rs, 2)]; T4o = T2 * T6; T4 = T2 * T3; T9 = W[8]; Tg = Ip[WS(rs, 7)]; T4p = FNMS(T5, T3, T4o); T7 = FMA(T5, T6, T4); T2j = T9 * Td; Tb = T9 * Ta; T4N = T4q - T4p; T4r = T4p + T4q; T8 = T1 + T7; T2i = T1 - T7; Tj = Im[WS(rs, 7)]; Tf = W[28]; } Tc = W[9]; Ti = W[29]; { E T3l, Ts, T2t, TL, TB, TE, TD, T3n, Ty, T2q, TC; { E TH, TK, TJ, T2s, TI; { E To, Tr, Tp, T3k, Tq, TG; { E T2k, Te, T2m, Tk, T2l, Th, Tn; To = Rp[WS(rs, 2)]; T2l = Tf * Tj; Th = Tf * Tg; T2k = FNMS(Tc, Ta, T2j); Te = FMA(Tc, Td, Tb); T2m = FNMS(Ti, Tg, T2l); Tk = FMA(Ti, Tj, Th); Tr = Rm[WS(rs, 2)]; Tn = W[6]; T4n = T2k + T2m; T2n = T2k - T2m; T4O = Te - Tk; Tl = Te + Tk; Tp = Tn * To; T3k = Tn * Tr; } Tq = W[7]; TH = Ip[WS(rs, 9)]; TK = Im[WS(rs, 9)]; TG = W[36]; T3l = FNMS(Tq, To, T3k); Ts = FMA(Tq, Tr, Tp); TJ = W[37]; T2s = TG * TK; TI = TG * TH; } { E Tu, Tx, Tt, Tw, T3m, Tv, TA; Tu = Rp[WS(rs, 7)]; Tx = Rm[WS(rs, 7)]; T2t = FNMS(TJ, TH, T2s); TL = FMA(TJ, TK, TI); Tt = W[26]; Tw = W[27]; TB = Ip[WS(rs, 4)]; TE = Im[WS(rs, 4)]; T3m = Tt * Tx; Tv = Tt * Tu; TA = W[16]; TD = W[17]; T3n = FNMS(Tw, Tu, T3m); Ty = FMA(Tw, Tx, Tv); T2q = TA * TE; TC = TA * TB; } } { E T3o, T3R, Tz, T2p, T2r, TF; T3o = T3l - T3n; T3R = T3l + T3n; Tz = Ts + Ty; T2p = Ts - Ty; T2r = FNMS(TD, TB, T2q); TF = FMA(TD, TE, TC); { E T3S, T2u, TM, T3j; T3S = T2r + T2t; T2u = T2r - T2t; TM = TF + TL; T3j = TL - TF; T2v = T2p - T2u; T3v = T2p + T2u; T3T = T3R + T3S; T4f = T3S - T3R; TN = Tz - TM; T2b = Tz + TM; T3F = T3o + T3j; T3p = T3j - T3o; } } } } } { E T2Z, T1M, T2P, T25, T1V, T1Y, T1X, T31, T1S, T2M, T1W; { E T21, T24, T23, T2O, T22; { E T1I, T1L, T1H, T1K, T2Y, T1J, T20; T1I = Rp[WS(rs, 6)]; T1L = Rm[WS(rs, 6)]; T1H = W[22]; T1K = W[23]; T21 = Ip[WS(rs, 3)]; T24 = Im[WS(rs, 3)]; T2Y = T1H * T1L; T1J = T1H * T1I; T20 = W[12]; T23 = W[13]; T2Z = FNMS(T1K, T1I, T2Y); T1M = FMA(T1K, T1L, T1J); T2O = T20 * T24; T22 = T20 * T21; } { E T1O, T1R, T1N, T1Q, T30, T1P, T1U; T1O = Rp[WS(rs, 1)]; T1R = Rm[WS(rs, 1)]; T2P = FNMS(T23, T21, T2O); T25 = FMA(T23, T24, T22); T1N = W[2]; T1Q = W[3]; T1V = Ip[WS(rs, 8)]; T1Y = Im[WS(rs, 8)]; T30 = T1N * T1R; T1P = T1N * T1O; T1U = W[32]; T1X = W[33]; T31 = FNMS(T1Q, T1O, T30); T1S = FMA(T1Q, T1R, T1P); T2M = T1U * T1Y; T1W = T1U * T1V; } } { E T32, T41, T1T, T2L, T2N, T1Z; T32 = T2Z - T31; T41 = T2Z + T31; T1T = T1M + T1S; T2L = T1M - T1S; T2N = FNMS(T1X, T1V, T2M); T1Z = FMA(T1X, T1Y, T1W); { E T42, T2Q, T26, T2X; T42 = T2N + T2P; T2Q = T2N - T2P; T26 = T1Z + T25; T2X = T25 - T1Z; T2R = T2L - T2Q; T3z = T2L + T2Q; T43 = T41 + T42; T4b = T42 - T41; T27 = T1T - T26; T2f = T1T + T26; T3J = T32 + T2X; T33 = T2X - T32; } } } { E T36, T1l, T2I, T1E, T1u, T1x, T1w, T38, T1r, T2F, T1v; { E T1A, T1D, T1C, T2H, T1B; { E T1h, T1k, T1g, T1j, T35, T1i, T1z; T1h = Rp[WS(rs, 4)]; T1k = Rm[WS(rs, 4)]; T1g = W[14]; T1j = W[15]; T1A = Ip[WS(rs, 1)]; T1D = Im[WS(rs, 1)]; T35 = T1g * T1k; T1i = T1g * T1h; T1z = W[4]; T1C = W[5]; T36 = FNMS(T1j, T1h, T35); T1l = FMA(T1j, T1k, T1i); T2H = T1z * T1D; T1B = T1z * T1A; } { E T1n, T1q, T1m, T1p, T37, T1o, T1t; T1n = Rp[WS(rs, 9)]; T1q = Rm[WS(rs, 9)]; T2I = FNMS(T1C, T1A, T2H); T1E = FMA(T1C, T1D, T1B); T1m = W[34]; T1p = W[35]; T1u = Ip[WS(rs, 6)]; T1x = Im[WS(rs, 6)]; T37 = T1m * T1q; T1o = T1m * T1n; T1t = W[24]; T1w = W[25]; T38 = FNMS(T1p, T1n, T37); T1r = FMA(T1p, T1q, T1o); T2F = T1t * T1x; T1v = T1t * T1u; } } { E T39, T3Y, T1s, T2E, T2G, T1y; T39 = T36 - T38; T3Y = T36 + T38; T1s = T1l + T1r; T2E = T1l - T1r; T2G = FNMS(T1w, T1u, T2F); T1y = FMA(T1w, T1x, T1v); { E T3Z, T2J, T1F, T34; T3Z = T2G + T2I; T2J = T2G - T2I; T1F = T1y + T1E; T34 = T1E - T1y; T2K = T2E - T2J; T3y = T2E + T2J; T40 = T3Y + T3Z; T4c = T3Z - T3Y; T1G = T1s - T1F; T2e = T1s + T1F; T3I = T39 + T34; T3a = T34 - T39; } } } { E T3e, TT, T2A, T1c, T12, T15, T14, T3g, TZ, T2x, T13; { E T18, T1b, T1a, T2z, T19; { E TP, TS, TO, TR, T3d, TQ, T17; TP = Rp[WS(rs, 8)]; TS = Rm[WS(rs, 8)]; TO = W[30]; TR = W[31]; T18 = Ip[WS(rs, 5)]; T1b = Im[WS(rs, 5)]; T3d = TO * TS; TQ = TO * TP; T17 = W[20]; T1a = W[21]; T3e = FNMS(TR, TP, T3d); TT = FMA(TR, TS, TQ); T2z = T17 * T1b; T19 = T17 * T18; } { E TV, TY, TU, TX, T3f, TW, T11; TV = Rp[WS(rs, 3)]; TY = Rm[WS(rs, 3)]; T2A = FNMS(T1a, T18, T2z); T1c = FMA(T1a, T1b, T19); TU = W[10]; TX = W[11]; T12 = Ip[0]; T15 = Im[0]; T3f = TU * TY; TW = TU * TV; T11 = W[0]; T14 = W[1]; T3g = FNMS(TX, TV, T3f); TZ = FMA(TX, TY, TW); T2x = T11 * T15; T13 = T11 * T12; } } { E T3h, T3U, T10, T2w, T2y, T16; T3h = T3e - T3g; T3U = T3e + T3g; T10 = TT + TZ; T2w = TT - TZ; T2y = FNMS(T14, T12, T2x); T16 = FMA(T14, T15, T13); { E T3V, T2B, T1d, T3c; T3V = T2y + T2A; T2B = T2y - T2A; T1d = T16 + T1c; T3c = T1c - T16; T2C = T2w - T2B; T3w = T2w + T2B; T3W = T3U + T3V; T4e = T3V - T3U; T1e = T10 - T1d; T2c = T10 + T1d; T3G = T3h + T3c; T3i = T3c - T3h; } } } { E T4s, T4k, T4l, T45, T47, T3P, T4y, T4A, T3O; { E T4C, T4B, T2a, T4j, T4h, T4E, T4M, T4K, T4i, T4a; { E Tm, T1f, T4J, T4I, T28, T4d, T4g, T29, T49, T48; T4C = T4c + T4b; T4d = T4b - T4c; T4g = T4e - T4f; T4B = T4f + T4e; T2a = T8 + Tl; Tm = T8 - Tl; T1f = TN + T1e; T4J = T1e - TN; T4I = T1G - T27; T28 = T1G + T27; T4j = FMA(KP618033988, T4d, T4g); T4h = FNMS(KP618033988, T4g, T4d); T29 = T1f + T28; T49 = T1f - T28; T4E = T4r - T4n; T4s = T4n + T4r; Rm[WS(rs, 9)] = Tm + T29; T48 = FNMS(KP250000000, T29, Tm); T4M = FNMS(KP618033988, T4I, T4J); T4K = FMA(KP618033988, T4J, T4I); T4i = FMA(KP559016994, T49, T48); T4a = FNMS(KP559016994, T49, T48); } { E T2d, T4w, T4x, T2g, T2h; { E T3X, T4G, T4F, T44, T4D, T4L, T4H; T4k = T3T + T3W; T3X = T3T - T3W; T4G = T4C - T4B; T4D = T4B + T4C; Rm[WS(rs, 1)] = FMA(KP951056516, T4h, T4a); Rp[WS(rs, 2)] = FNMS(KP951056516, T4h, T4a); Rp[WS(rs, 6)] = FMA(KP951056516, T4j, T4i); Rm[WS(rs, 5)] = FNMS(KP951056516, T4j, T4i); Im[WS(rs, 9)] = T4D - T4E; T4F = FMA(KP250000000, T4D, T4E); T44 = T40 - T43; T4l = T40 + T43; T2d = T2b + T2c; T4w = T2b - T2c; T4L = FMA(KP559016994, T4G, T4F); T4H = FNMS(KP559016994, T4G, T4F); T45 = FMA(KP618033988, T44, T3X); T47 = FNMS(KP618033988, T3X, T44); Ip[WS(rs, 2)] = FMA(KP951056516, T4K, T4H); Im[WS(rs, 1)] = FMS(KP951056516, T4K, T4H); Ip[WS(rs, 6)] = FMA(KP951056516, T4M, T4L); Im[WS(rs, 5)] = FMS(KP951056516, T4M, T4L); T4x = T2f - T2e; T2g = T2e + T2f; } T2h = T2d + T2g; T3P = T2d - T2g; T4y = FNMS(KP618033988, T4x, T4w); T4A = FMA(KP618033988, T4w, T4x); Rp[0] = T2a + T2h; T3O = FNMS(KP250000000, T2h, T2a); } } { E T3u, T54, T5a, T5c, T56, T53; { E T52, T51, T3t, T3r, T2o, T58, T59, T2T, T2V, T4u, T4t, T2U, T3s, T2W; { E T3b, T3q, T46, T3Q, T4m; T52 = T3a + T33; T3b = T33 - T3a; T3q = T3i - T3p; T51 = T3p + T3i; T46 = FNMS(KP559016994, T3P, T3O); T3Q = FMA(KP559016994, T3P, T3O); T4m = T4k + T4l; T4u = T4k - T4l; Rm[WS(rs, 3)] = FMA(KP951056516, T45, T3Q); Rp[WS(rs, 4)] = FNMS(KP951056516, T45, T3Q); Rp[WS(rs, 8)] = FMA(KP951056516, T47, T46); Rm[WS(rs, 7)] = FNMS(KP951056516, T47, T46); Ip[0] = T4m + T4s; T4t = FNMS(KP250000000, T4m, T4s); T3t = FMA(KP618033988, T3b, T3q); T3r = FNMS(KP618033988, T3q, T3b); } T3u = T2i + T2n; T2o = T2i - T2n; { E T4v, T4z, T2D, T2S; T4v = FMA(KP559016994, T4u, T4t); T4z = FNMS(KP559016994, T4u, T4t); T2D = T2v + T2C; T58 = T2v - T2C; T59 = T2K - T2R; T2S = T2K + T2R; Ip[WS(rs, 4)] = FMA(KP951056516, T4y, T4v); Im[WS(rs, 3)] = FMS(KP951056516, T4y, T4v); Ip[WS(rs, 8)] = FMA(KP951056516, T4A, T4z); Im[WS(rs, 7)] = FMS(KP951056516, T4A, T4z); T2T = T2D + T2S; T2V = T2D - T2S; } Rm[WS(rs, 4)] = T2o + T2T; T2U = FNMS(KP250000000, T2T, T2o); T54 = T4O + T4N; T4P = T4N - T4O; T5a = FMA(KP618033988, T59, T58); T5c = FNMS(KP618033988, T58, T59); T3s = FMA(KP559016994, T2V, T2U); T2W = FNMS(KP559016994, T2V, T2U); Rp[WS(rs, 7)] = FNMS(KP951056516, T3r, T2W); Rp[WS(rs, 3)] = FMA(KP951056516, T3r, T2W); Rm[0] = FNMS(KP951056516, T3t, T3s); Rm[WS(rs, 8)] = FMA(KP951056516, T3t, T3s); T56 = T51 - T52; T53 = T51 + T52; } { E T4Q, T4R, T3N, T3L, T4X, T4W, T3B, T3D, T3H, T3K, T55, T3C, T3M, T3E; T4Q = T3F + T3G; T3H = T3F - T3G; T3K = T3I - T3J; T4R = T3I + T3J; Im[WS(rs, 4)] = T53 - T54; T55 = FMA(KP250000000, T53, T54); T3N = FNMS(KP618033988, T3H, T3K); T3L = FMA(KP618033988, T3K, T3H); { E T57, T5b, T3x, T3A; T57 = FNMS(KP559016994, T56, T55); T5b = FMA(KP559016994, T56, T55); T3x = T3v + T3w; T4X = T3v - T3w; T4W = T3y - T3z; T3A = T3y + T3z; Im[0] = -(FMA(KP951056516, T5a, T57)); Im[WS(rs, 8)] = FMS(KP951056516, T5a, T57); Ip[WS(rs, 7)] = FMA(KP951056516, T5c, T5b); Ip[WS(rs, 3)] = FNMS(KP951056516, T5c, T5b); T3B = T3x + T3A; T3D = T3x - T3A; } Rp[WS(rs, 5)] = T3u + T3B; T3C = FNMS(KP250000000, T3B, T3u); T4Y = FNMS(KP618033988, T4X, T4W); T50 = FMA(KP618033988, T4W, T4X); T3M = FNMS(KP559016994, T3D, T3C); T3E = FMA(KP559016994, T3D, T3C); Rp[WS(rs, 9)] = FNMS(KP951056516, T3L, T3E); Rp[WS(rs, 1)] = FMA(KP951056516, T3L, T3E); Rm[WS(rs, 2)] = FNMS(KP951056516, T3N, T3M); Rm[WS(rs, 6)] = FMA(KP951056516, T3N, T3M); T4U = T4Q - T4R; T4S = T4Q + T4R; } } } } Ip[WS(rs, 5)] = T4S + T4P; T4T = FNMS(KP250000000, T4S, T4P); T4Z = FMA(KP559016994, T4U, T4T); T4V = FNMS(KP559016994, T4U, T4T); Im[WS(rs, 2)] = -(FMA(KP951056516, T4Y, T4V)); Im[WS(rs, 6)] = FMS(KP951056516, T4Y, T4V); Ip[WS(rs, 9)] = FMA(KP951056516, T50, T4Z); Ip[WS(rs, 1)] = FNMS(KP951056516, T50, T4Z); } } } static const tw_instr twinstr[] = { {TW_FULL, 1, 20}, {TW_NEXT, 1, 0} }; static const hc2c_desc desc = { 20, "hc2cf_20", twinstr, &GENUS, {136, 38, 110, 0} }; void X(codelet_hc2cf_20) (planner *p) { X(khc2c_register) (p, hc2cf_20, &desc, HC2C_VIA_RDFT); } #else /* HAVE_FMA */ /* Generated by: ../../../genfft/gen_hc2c.native -compact -variables 4 -pipeline-latency 4 -n 20 -dit -name hc2cf_20 -include hc2cf.h */ /* * This function contains 246 FP additions, 124 FP multiplications, * (or, 184 additions, 62 multiplications, 62 fused multiply/add), * 85 stack variables, 4 constants, and 80 memory accesses */ #include "hc2cf.h" static void hc2cf_20(R *Rp, R *Ip, R *Rm, R *Im, const R *W, stride rs, INT mb, INT me, INT ms) { DK(KP587785252, +0.587785252292473129168705954639072768597652438); DK(KP951056516, +0.951056516295153572116439333379382143405698634); DK(KP250000000, +0.250000000000000000000000000000000000000000000); DK(KP559016994, +0.559016994374947424102293417182819058860154590); { INT m; for (m = mb, W = W + ((mb - 1) * 38); m < me; m = m + 1, Rp = Rp + ms, Ip = Ip + ms, Rm = Rm - ms, Im = Im - ms, W = W + 38, MAKE_VOLATILE_STRIDE(80, rs)) { E Tj, T1R, T4j, T4s, T2q, T37, T3Q, T42, T1r, T1O, T1P, T3p, T3s, T3K, T3A; E T3B, T3Z, T1V, T1W, T1X, T23, T28, T4q, T2W, T2X, T4f, T33, T34, T35, T2G; E T2L, T2M, TG, T13, T14, T3i, T3l, T3J, T3D, T3E, T40, T1S, T1T, T1U, T2e; E T2j, T4p, T2T, T2U, T4e, T30, T31, T32, T2v, T2A, T2B; { E T1, T3O, T6, T3N, Tc, T2n, Th, T2o; T1 = Rp[0]; T3O = Rm[0]; { E T3, T5, T2, T4; T3 = Rp[WS(rs, 5)]; T5 = Rm[WS(rs, 5)]; T2 = W[18]; T4 = W[19]; T6 = FMA(T2, T3, T4 * T5); T3N = FNMS(T4, T3, T2 * T5); } { E T9, Tb, T8, Ta; T9 = Ip[WS(rs, 2)]; Tb = Im[WS(rs, 2)]; T8 = W[8]; Ta = W[9]; Tc = FMA(T8, T9, Ta * Tb); T2n = FNMS(Ta, T9, T8 * Tb); } { E Te, Tg, Td, Tf; Te = Ip[WS(rs, 7)]; Tg = Im[WS(rs, 7)]; Td = W[28]; Tf = W[29]; Th = FMA(Td, Te, Tf * Tg); T2o = FNMS(Tf, Te, Td * Tg); } { E T7, Ti, T4h, T4i; T7 = T1 + T6; Ti = Tc + Th; Tj = T7 - Ti; T1R = T7 + Ti; T4h = T3O - T3N; T4i = Tc - Th; T4j = T4h - T4i; T4s = T4i + T4h; } { E T2m, T2p, T3M, T3P; T2m = T1 - T6; T2p = T2n - T2o; T2q = T2m - T2p; T37 = T2m + T2p; T3M = T2n + T2o; T3P = T3N + T3O; T3Q = T3M + T3P; T42 = T3P - T3M; } } { E T1f, T3n, T21, T2C, T1N, T3r, T27, T2K, T1q, T3o, T22, T2F, T1C, T3q, T26; E T2H; { E T19, T1Z, T1e, T20; { E T16, T18, T15, T17; T16 = Rp[WS(rs, 4)]; T18 = Rm[WS(rs, 4)]; T15 = W[14]; T17 = W[15]; T19 = FMA(T15, T16, T17 * T18); T1Z = FNMS(T17, T16, T15 * T18); } { E T1b, T1d, T1a, T1c; T1b = Rp[WS(rs, 9)]; T1d = Rm[WS(rs, 9)]; T1a = W[34]; T1c = W[35]; T1e = FMA(T1a, T1b, T1c * T1d); T20 = FNMS(T1c, T1b, T1a * T1d); } T1f = T19 + T1e; T3n = T1Z + T20; T21 = T1Z - T20; T2C = T19 - T1e; } { E T1H, T2I, T1M, T2J; { E T1E, T1G, T1D, T1F; T1E = Ip[WS(rs, 8)]; T1G = Im[WS(rs, 8)]; T1D = W[32]; T1F = W[33]; T1H = FMA(T1D, T1E, T1F * T1G); T2I = FNMS(T1F, T1E, T1D * T1G); } { E T1J, T1L, T1I, T1K; T1J = Ip[WS(rs, 3)]; T1L = Im[WS(rs, 3)]; T1I = W[12]; T1K = W[13]; T1M = FMA(T1I, T1J, T1K * T1L); T2J = FNMS(T1K, T1J, T1I * T1L); } T1N = T1H + T1M; T3r = T2I + T2J; T27 = T1H - T1M; T2K = T2I - T2J; } { E T1k, T2D, T1p, T2E; { E T1h, T1j, T1g, T1i; T1h = Ip[WS(rs, 6)]; T1j = Im[WS(rs, 6)]; T1g = W[24]; T1i = W[25]; T1k = FMA(T1g, T1h, T1i * T1j); T2D = FNMS(T1i, T1h, T1g * T1j); } { E T1m, T1o, T1l, T1n; T1m = Ip[WS(rs, 1)]; T1o = Im[WS(rs, 1)]; T1l = W[4]; T1n = W[5]; T1p = FMA(T1l, T1m, T1n * T1o); T2E = FNMS(T1n, T1m, T1l * T1o); } T1q = T1k + T1p; T3o = T2D + T2E; T22 = T1k - T1p; T2F = T2D - T2E; } { E T1w, T24, T1B, T25; { E T1t, T1v, T1s, T1u; T1t = Rp[WS(rs, 6)]; T1v = Rm[WS(rs, 6)]; T1s = W[22]; T1u = W[23]; T1w = FMA(T1s, T1t, T1u * T1v); T24 = FNMS(T1u, T1t, T1s * T1v); } { E T1y, T1A, T1x, T1z; T1y = Rp[WS(rs, 1)]; T1A = Rm[WS(rs, 1)]; T1x = W[2]; T1z = W[3]; T1B = FMA(T1x, T1y, T1z * T1A); T25 = FNMS(T1z, T1y, T1x * T1A); } T1C = T1w + T1B; T3q = T24 + T25; T26 = T24 - T25; T2H = T1w - T1B; } T1r = T1f - T1q; T1O = T1C - T1N; T1P = T1r + T1O; T3p = T3n + T3o; T3s = T3q + T3r; T3K = T3p + T3s; T3A = T3n - T3o; T3B = T3r - T3q; T3Z = T3B - T3A; T1V = T1f + T1q; T1W = T1C + T1N; T1X = T1V + T1W; T23 = T21 + T22; T28 = T26 + T27; T4q = T23 + T28; T2W = T21 - T22; T2X = T26 - T27; T4f = T2W + T2X; T33 = T2C + T2F; T34 = T2H + T2K; T35 = T33 + T34; T2G = T2C - T2F; T2L = T2H - T2K; T2M = T2G + T2L; } { E Tu, T3g, T2c, T2r, T12, T3k, T2f, T2z, TF, T3h, T2d, T2u, TR, T3j, T2i; E T2w; { E To, T2a, Tt, T2b; { E Tl, Tn, Tk, Tm; Tl = Rp[WS(rs, 2)]; Tn = Rm[WS(rs, 2)]; Tk = W[6]; Tm = W[7]; To = FMA(Tk, Tl, Tm * Tn); T2a = FNMS(Tm, Tl, Tk * Tn); } { E Tq, Ts, Tp, Tr; Tq = Rp[WS(rs, 7)]; Ts = Rm[WS(rs, 7)]; Tp = W[26]; Tr = W[27]; Tt = FMA(Tp, Tq, Tr * Ts); T2b = FNMS(Tr, Tq, Tp * Ts); } Tu = To + Tt; T3g = T2a + T2b; T2c = T2a - T2b; T2r = To - Tt; } { E TW, T2x, T11, T2y; { E TT, TV, TS, TU; TT = Ip[0]; TV = Im[0]; TS = W[0]; TU = W[1]; TW = FMA(TS, TT, TU * TV); T2x = FNMS(TU, TT, TS * TV); } { E TY, T10, TX, TZ; TY = Ip[WS(rs, 5)]; T10 = Im[WS(rs, 5)]; TX = W[20]; TZ = W[21]; T11 = FMA(TX, TY, TZ * T10); T2y = FNMS(TZ, TY, TX * T10); } T12 = TW + T11; T3k = T2x + T2y; T2f = T11 - TW; T2z = T2x - T2y; } { E Tz, T2s, TE, T2t; { E Tw, Ty, Tv, Tx; Tw = Ip[WS(rs, 4)]; Ty = Im[WS(rs, 4)]; Tv = W[16]; Tx = W[17]; Tz = FMA(Tv, Tw, Tx * Ty); T2s = FNMS(Tx, Tw, Tv * Ty); } { E TB, TD, TA, TC; TB = Ip[WS(rs, 9)]; TD = Im[WS(rs, 9)]; TA = W[36]; TC = W[37]; TE = FMA(TA, TB, TC * TD); T2t = FNMS(TC, TB, TA * TD); } TF = Tz + TE; T3h = T2s + T2t; T2d = Tz - TE; T2u = T2s - T2t; } { E TL, T2g, TQ, T2h; { E TI, TK, TH, TJ; TI = Rp[WS(rs, 8)]; TK = Rm[WS(rs, 8)]; TH = W[30]; TJ = W[31]; TL = FMA(TH, TI, TJ * TK); T2g = FNMS(TJ, TI, TH * TK); } { E TN, TP, TM, TO; TN = Rp[WS(rs, 3)]; TP = Rm[WS(rs, 3)]; TM = W[10]; TO = W[11]; TQ = FMA(TM, TN, TO * TP); T2h = FNMS(TO, TN, TM * TP); } TR = TL + TQ; T3j = T2g + T2h; T2i = T2g - T2h; T2w = TL - TQ; } TG = Tu - TF; T13 = TR - T12; T14 = TG + T13; T3i = T3g + T3h; T3l = T3j + T3k; T3J = T3i + T3l; T3D = T3g - T3h; T3E = T3j - T3k; T40 = T3D + T3E; T1S = Tu + TF; T1T = TR + T12; T1U = T1S + T1T; T2e = T2c + T2d; T2j = T2f - T2i; T4p = T2j - T2e; T2T = T2c - T2d; T2U = T2i + T2f; T4e = T2T + T2U; T30 = T2r + T2u; T31 = T2w + T2z; T32 = T30 + T31; T2v = T2r - T2u; T2A = T2w - T2z; T2B = T2v + T2A; } { E T3y, T1Q, T3x, T3G, T3I, T3C, T3F, T3H, T3z; T3y = KP559016994 * (T14 - T1P); T1Q = T14 + T1P; T3x = FNMS(KP250000000, T1Q, Tj); T3C = T3A + T3B; T3F = T3D - T3E; T3G = FNMS(KP587785252, T3F, KP951056516 * T3C); T3I = FMA(KP951056516, T3F, KP587785252 * T3C); Rm[WS(rs, 9)] = Tj + T1Q; T3H = T3y + T3x; Rm[WS(rs, 5)] = T3H - T3I; Rp[WS(rs, 6)] = T3H + T3I; T3z = T3x - T3y; Rp[WS(rs, 2)] = T3z - T3G; Rm[WS(rs, 1)] = T3z + T3G; } { E T47, T41, T46, T45, T49, T43, T44, T4a, T48; T47 = KP559016994 * (T40 + T3Z); T41 = T3Z - T40; T46 = FMA(KP250000000, T41, T42); T43 = T13 - TG; T44 = T1r - T1O; T45 = FMA(KP587785252, T43, KP951056516 * T44); T49 = FNMS(KP587785252, T44, KP951056516 * T43); Im[WS(rs, 9)] = T41 - T42; T4a = T47 + T46; Im[WS(rs, 5)] = T49 - T4a; Ip[WS(rs, 6)] = T49 + T4a; T48 = T46 - T47; Im[WS(rs, 1)] = T45 - T48; Ip[WS(rs, 2)] = T45 + T48; } { E T3d, T1Y, T3e, T3u, T3w, T3m, T3t, T3v, T3f; T3d = KP559016994 * (T1U - T1X); T1Y = T1U + T1X; T3e = FNMS(KP250000000, T1Y, T1R); T3m = T3i - T3l; T3t = T3p - T3s; T3u = FMA(KP951056516, T3m, KP587785252 * T3t); T3w = FNMS(KP587785252, T3m, KP951056516 * T3t); Rp[0] = T1R + T1Y; T3v = T3e - T3d; Rm[WS(rs, 7)] = T3v - T3w; Rp[WS(rs, 8)] = T3v + T3w; T3f = T3d + T3e; Rp[WS(rs, 4)] = T3f - T3u; Rm[WS(rs, 3)] = T3f + T3u; } { E T3U, T3L, T3V, T3T, T3X, T3R, T3S, T3Y, T3W; T3U = KP559016994 * (T3J - T3K); T3L = T3J + T3K; T3V = FNMS(KP250000000, T3L, T3Q); T3R = T1S - T1T; T3S = T1V - T1W; T3T = FMA(KP951056516, T3R, KP587785252 * T3S); T3X = FNMS(KP951056516, T3S, KP587785252 * T3R); Ip[0] = T3L + T3Q; T3Y = T3V - T3U; Im[WS(rs, 7)] = T3X - T3Y; Ip[WS(rs, 8)] = T3X + T3Y; T3W = T3U + T3V; Im[WS(rs, 3)] = T3T - T3W; Ip[WS(rs, 4)] = T3T + T3W; } { E T2P, T2N, T2O, T2l, T2R, T29, T2k, T2S, T2Q; T2P = KP559016994 * (T2B - T2M); T2N = T2B + T2M; T2O = FNMS(KP250000000, T2N, T2q); T29 = T23 - T28; T2k = T2e + T2j; T2l = FNMS(KP587785252, T2k, KP951056516 * T29); T2R = FMA(KP951056516, T2k, KP587785252 * T29); Rm[WS(rs, 4)] = T2q + T2N; T2S = T2P + T2O; Rm[WS(rs, 8)] = T2R + T2S; Rm[0] = T2S - T2R; T2Q = T2O - T2P; Rp[WS(rs, 3)] = T2l + T2Q; Rp[WS(rs, 7)] = T2Q - T2l; } { E T4w, T4r, T4x, T4v, T4A, T4t, T4u, T4z, T4y; T4w = KP559016994 * (T4p + T4q); T4r = T4p - T4q; T4x = FMA(KP250000000, T4r, T4s); T4t = T2v - T2A; T4u = T2G - T2L; T4v = FMA(KP951056516, T4t, KP587785252 * T4u); T4A = FNMS(KP587785252, T4t, KP951056516 * T4u); Im[WS(rs, 4)] = T4r - T4s; T4z = T4w + T4x; Ip[WS(rs, 3)] = T4z - T4A; Ip[WS(rs, 7)] = T4A + T4z; T4y = T4w - T4x; Im[WS(rs, 8)] = T4v + T4y; Im[0] = T4y - T4v; } { E T36, T38, T39, T2Z, T3b, T2V, T2Y, T3c, T3a; T36 = KP559016994 * (T32 - T35); T38 = T32 + T35; T39 = FNMS(KP250000000, T38, T37); T2V = T2T - T2U; T2Y = T2W - T2X; T2Z = FMA(KP951056516, T2V, KP587785252 * T2Y); T3b = FNMS(KP587785252, T2V, KP951056516 * T2Y); Rp[WS(rs, 5)] = T37 + T38; T3c = T39 - T36; Rm[WS(rs, 6)] = T3b + T3c; Rm[WS(rs, 2)] = T3c - T3b; T3a = T36 + T39; Rp[WS(rs, 1)] = T2Z + T3a; Rp[WS(rs, 9)] = T3a - T2Z; } { E T4g, T4k, T4l, T4d, T4o, T4b, T4c, T4n, T4m; T4g = KP559016994 * (T4e - T4f); T4k = T4e + T4f; T4l = FNMS(KP250000000, T4k, T4j); T4b = T33 - T34; T4c = T30 - T31; T4d = FNMS(KP587785252, T4c, KP951056516 * T4b); T4o = FMA(KP951056516, T4c, KP587785252 * T4b); Ip[WS(rs, 5)] = T4k + T4j; T4n = T4g + T4l; Ip[WS(rs, 1)] = T4n - T4o; Ip[WS(rs, 9)] = T4o + T4n; T4m = T4g - T4l; Im[WS(rs, 6)] = T4d + T4m; Im[WS(rs, 2)] = T4m - T4d; } } } } static const tw_instr twinstr[] = { {TW_FULL, 1, 20}, {TW_NEXT, 1, 0} }; static const hc2c_desc desc = { 20, "hc2cf_20", twinstr, &GENUS, {184, 62, 62, 0} }; void X(codelet_hc2cf_20) (planner *p) { X(khc2c_register) (p, hc2cf_20, &desc, HC2C_VIA_RDFT); } #endif /* HAVE_FMA */ fftw-3.3.3/rdft/scalar/r2cf/hc2cf2_4.c0000644000175000017500000001241312054410712014102 00000000000000/* * Copyright (c) 2003, 2007-11 Matteo Frigo * Copyright (c) 2003, 2007-11 Massachusetts Institute of Technology * * This program is free software; you can redistribute it and/or modify * it under the terms of the GNU General Public License as published by * the Free Software Foundation; either version 2 of the License, or * (at your option) any later version. * * This program is distributed in the hope that it will be useful, * but WITHOUT ANY WARRANTY; without even the implied warranty of * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the * GNU General Public License for more details. * * You should have received a copy of the GNU General Public License * along with this program; if not, write to the Free Software * Foundation, Inc., 51 Franklin Street, Fifth Floor, Boston, MA 02110-1301 USA * */ /* This file was automatically generated --- DO NOT EDIT */ /* Generated on Sun Nov 25 07:40:42 EST 2012 */ #include "codelet-rdft.h" #ifdef HAVE_FMA /* Generated by: ../../../genfft/gen_hc2c.native -fma -reorder-insns -schedule-for-pipeline -compact -variables 4 -pipeline-latency 4 -twiddle-log3 -precompute-twiddles -n 4 -dit -name hc2cf2_4 -include hc2cf.h */ /* * This function contains 24 FP additions, 16 FP multiplications, * (or, 16 additions, 8 multiplications, 8 fused multiply/add), * 33 stack variables, 0 constants, and 16 memory accesses */ #include "hc2cf.h" static void hc2cf2_4(R *Rp, R *Ip, R *Rm, R *Im, const R *W, stride rs, INT mb, INT me, INT ms) { { INT m; for (m = mb, W = W + ((mb - 1) * 4); m < me; m = m + 1, Rp = Rp + ms, Ip = Ip + ms, Rm = Rm - ms, Im = Im - ms, W = W + 4, MAKE_VOLATILE_STRIDE(16, rs)) { E Ti, Tq, To, Te, Ty, TA, Tm, Ts; { E T2, T6, T3, T5; T2 = W[0]; T6 = W[3]; T3 = W[2]; T5 = W[1]; { E T1, Tx, Td, Tw, Tj, Tl, Ta, T4, Tk, Tr; T1 = Rp[0]; Ta = T2 * T6; T4 = T2 * T3; Tx = Rm[0]; { E T8, Tb, T7, Tc; T8 = Rp[WS(rs, 1)]; Tb = FNMS(T5, T3, Ta); T7 = FMA(T5, T6, T4); Tc = Rm[WS(rs, 1)]; { E Tf, Th, T9, Tv, Tg, Tp; Tf = Ip[0]; Th = Im[0]; T9 = T7 * T8; Tv = T7 * Tc; Tg = T2 * Tf; Tp = T2 * Th; Td = FMA(Tb, Tc, T9); Tw = FNMS(Tb, T8, Tv); Ti = FMA(T5, Th, Tg); Tq = FNMS(T5, Tf, Tp); } Tj = Ip[WS(rs, 1)]; Tl = Im[WS(rs, 1)]; } To = T1 - Td; Te = T1 + Td; Ty = Tw + Tx; TA = Tx - Tw; Tk = T3 * Tj; Tr = T3 * Tl; Tm = FMA(T6, Tl, Tk); Ts = FNMS(T6, Tj, Tr); } } { E Tn, Tz, Tu, Tt; Tn = Ti + Tm; Tz = Tm - Ti; Tu = Tq + Ts; Tt = Tq - Ts; Ip[WS(rs, 1)] = Tz + TA; Im[0] = Tz - TA; Rp[0] = Te + Tn; Rm[WS(rs, 1)] = Te - Tn; Rp[WS(rs, 1)] = To + Tt; Rm[0] = To - Tt; Ip[0] = Tu + Ty; Im[WS(rs, 1)] = Tu - Ty; } } } } static const tw_instr twinstr[] = { {TW_CEXP, 1, 1}, {TW_CEXP, 1, 3}, {TW_NEXT, 1, 0} }; static const hc2c_desc desc = { 4, "hc2cf2_4", twinstr, &GENUS, {16, 8, 8, 0} }; void X(codelet_hc2cf2_4) (planner *p) { X(khc2c_register) (p, hc2cf2_4, &desc, HC2C_VIA_RDFT); } #else /* HAVE_FMA */ /* Generated by: ../../../genfft/gen_hc2c.native -compact -variables 4 -pipeline-latency 4 -twiddle-log3 -precompute-twiddles -n 4 -dit -name hc2cf2_4 -include hc2cf.h */ /* * This function contains 24 FP additions, 16 FP multiplications, * (or, 16 additions, 8 multiplications, 8 fused multiply/add), * 21 stack variables, 0 constants, and 16 memory accesses */ #include "hc2cf.h" static void hc2cf2_4(R *Rp, R *Ip, R *Rm, R *Im, const R *W, stride rs, INT mb, INT me, INT ms) { { INT m; for (m = mb, W = W + ((mb - 1) * 4); m < me; m = m + 1, Rp = Rp + ms, Ip = Ip + ms, Rm = Rm - ms, Im = Im - ms, W = W + 4, MAKE_VOLATILE_STRIDE(16, rs)) { E T2, T4, T3, T5, T6, T8; T2 = W[0]; T4 = W[1]; T3 = W[2]; T5 = W[3]; T6 = FMA(T2, T3, T4 * T5); T8 = FNMS(T4, T3, T2 * T5); { E T1, Tp, Ta, To, Te, Tk, Th, Tl, T7, T9; T1 = Rp[0]; Tp = Rm[0]; T7 = Rp[WS(rs, 1)]; T9 = Rm[WS(rs, 1)]; Ta = FMA(T6, T7, T8 * T9); To = FNMS(T8, T7, T6 * T9); { E Tc, Td, Tf, Tg; Tc = Ip[0]; Td = Im[0]; Te = FMA(T2, Tc, T4 * Td); Tk = FNMS(T4, Tc, T2 * Td); Tf = Ip[WS(rs, 1)]; Tg = Im[WS(rs, 1)]; Th = FMA(T3, Tf, T5 * Tg); Tl = FNMS(T5, Tf, T3 * Tg); } { E Tb, Ti, Tn, Tq; Tb = T1 + Ta; Ti = Te + Th; Rm[WS(rs, 1)] = Tb - Ti; Rp[0] = Tb + Ti; Tn = Tk + Tl; Tq = To + Tp; Im[WS(rs, 1)] = Tn - Tq; Ip[0] = Tn + Tq; } { E Tj, Tm, Tr, Ts; Tj = T1 - Ta; Tm = Tk - Tl; Rm[0] = Tj - Tm; Rp[WS(rs, 1)] = Tj + Tm; Tr = Th - Te; Ts = Tp - To; Im[0] = Tr - Ts; Ip[WS(rs, 1)] = Tr + Ts; } } } } } static const tw_instr twinstr[] = { {TW_CEXP, 1, 1}, {TW_CEXP, 1, 3}, {TW_NEXT, 1, 0} }; static const hc2c_desc desc = { 4, "hc2cf2_4", twinstr, &GENUS, {16, 8, 8, 0} }; void X(codelet_hc2cf2_4) (planner *p) { X(khc2c_register) (p, hc2cf2_4, &desc, HC2C_VIA_RDFT); } #endif /* HAVE_FMA */ fftw-3.3.3/rdft/scalar/r2cf/hc2cf2_8.c0000644000175000017500000002435012054410712014111 00000000000000/* * Copyright (c) 2003, 2007-11 Matteo Frigo * Copyright (c) 2003, 2007-11 Massachusetts Institute of Technology * * This program is free software; you can redistribute it and/or modify * it under the terms of the GNU General Public License as published by * the Free Software Foundation; either version 2 of the License, or * (at your option) any later version. * * This program is distributed in the hope that it will be useful, * but WITHOUT ANY WARRANTY; without even the implied warranty of * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the * GNU General Public License for more details. * * You should have received a copy of the GNU General Public License * along with this program; if not, write to the Free Software * Foundation, Inc., 51 Franklin Street, Fifth Floor, Boston, MA 02110-1301 USA * */ /* This file was automatically generated --- DO NOT EDIT */ /* Generated on Sun Nov 25 07:40:42 EST 2012 */ #include "codelet-rdft.h" #ifdef HAVE_FMA /* Generated by: ../../../genfft/gen_hc2c.native -fma -reorder-insns -schedule-for-pipeline -compact -variables 4 -pipeline-latency 4 -twiddle-log3 -precompute-twiddles -n 8 -dit -name hc2cf2_8 -include hc2cf.h */ /* * This function contains 74 FP additions, 50 FP multiplications, * (or, 44 additions, 20 multiplications, 30 fused multiply/add), * 64 stack variables, 1 constants, and 32 memory accesses */ #include "hc2cf.h" static void hc2cf2_8(R *Rp, R *Ip, R *Rm, R *Im, const R *W, stride rs, INT mb, INT me, INT ms) { DK(KP707106781, +0.707106781186547524400844362104849039284835938); { INT m; for (m = mb, W = W + ((mb - 1) * 6); m < me; m = m + 1, Rp = Rp + ms, Ip = Ip + ms, Rm = Rm - ms, Im = Im - ms, W = W + 6, MAKE_VOLATILE_STRIDE(32, rs)) { E TS, T1m, TJ, T1l, T1k, Tw, T1w, T1u; { E T2, T3, Tl, Tn, T5, T4, Tm, Tr, T6; T2 = W[0]; T3 = W[2]; Tl = W[4]; Tn = W[5]; T5 = W[1]; T4 = T2 * T3; Tm = T2 * Tl; Tr = T2 * Tn; T6 = W[3]; { E T1, T1s, TG, Td, T1r, Tu, TY, Tk, TW, T18, T1d, TD, TH, TA, T13; E TE, T14; { E To, Ts, Tf, T7, T8, Ti, Tb, T9, Tc, TC, Ta, TF, TB, Tg, Th; E Tj; T1 = Rp[0]; To = FMA(T5, Tn, Tm); Ts = FNMS(T5, Tl, Tr); Tf = FMA(T5, T6, T4); T7 = FNMS(T5, T6, T4); Ta = T2 * T6; T1s = Rm[0]; T8 = Rp[WS(rs, 2)]; TF = Tf * Tn; TB = Tf * Tl; Ti = FNMS(T5, T3, Ta); Tb = FMA(T5, T3, Ta); T9 = T7 * T8; Tc = Rm[WS(rs, 2)]; TG = FNMS(Ti, Tl, TF); TC = FMA(Ti, Tn, TB); { E Tp, T1q, Tt, Tq, TX; Tp = Rp[WS(rs, 3)]; Td = FMA(Tb, Tc, T9); T1q = T7 * Tc; Tt = Rm[WS(rs, 3)]; Tq = To * Tp; Tg = Rp[WS(rs, 1)]; T1r = FNMS(Tb, T8, T1q); TX = To * Tt; Tu = FMA(Ts, Tt, Tq); Th = Tf * Tg; Tj = Rm[WS(rs, 1)]; TY = FNMS(Ts, Tp, TX); } { E TO, TQ, TN, TP, T1a, T1b; { E TK, TM, TL, T19, TV; TK = Ip[WS(rs, 3)]; TM = Im[WS(rs, 3)]; Tk = FMA(Ti, Tj, Th); TV = Tf * Tj; TL = Tl * TK; T19 = Tl * TM; TO = Ip[WS(rs, 1)]; TW = FNMS(Ti, Tg, TV); TQ = Im[WS(rs, 1)]; TN = FMA(Tn, TM, TL); TP = T3 * TO; T1a = FNMS(Tn, TK, T19); T1b = T3 * TQ; } { E Tx, Tz, Ty, T12, T1c, TR; Tx = Ip[0]; TR = FMA(T6, TQ, TP); Tz = Im[0]; T1c = FNMS(T6, TO, T1b); Ty = T2 * Tx; T18 = TN - TR; TS = TN + TR; T12 = T2 * Tz; T1d = T1a - T1c; T1m = T1a + T1c; TD = Ip[WS(rs, 2)]; TH = Im[WS(rs, 2)]; TA = FMA(T5, Tz, Ty); T13 = FNMS(T5, Tx, T12); TE = TC * TD; T14 = TC * TH; } } } { E Te, T1p, T1t, Tv; { E T1g, T10, T1z, T1B, T1A, T1j, T1C, T1f; { E T1x, T11, T16, T1y; { E TU, TZ, TI, T15; Te = T1 + Td; TU = T1 - Td; TZ = TW - TY; T1p = TW + TY; TI = FMA(TG, TH, TE); T15 = FNMS(TG, TD, T14); T1t = T1r + T1s; T1x = T1s - T1r; T1g = TU - TZ; T10 = TU + TZ; T11 = TA - TI; TJ = TA + TI; T1l = T13 + T15; T16 = T13 - T15; T1y = Tk - Tu; Tv = Tk + Tu; } { E T1i, T1e, T17, T1h; T1i = T18 + T1d; T1e = T18 - T1d; T17 = T11 + T16; T1h = T16 - T11; T1z = T1x - T1y; T1B = T1y + T1x; T1A = T1h + T1i; T1j = T1h - T1i; T1C = T1e - T17; T1f = T17 + T1e; } } Rm[0] = FNMS(KP707106781, T1j, T1g); Im[0] = FMS(KP707106781, T1C, T1B); Rp[WS(rs, 1)] = FMA(KP707106781, T1f, T10); Rm[WS(rs, 2)] = FNMS(KP707106781, T1f, T10); Ip[WS(rs, 1)] = FMA(KP707106781, T1A, T1z); Im[WS(rs, 2)] = FMS(KP707106781, T1A, T1z); Rp[WS(rs, 3)] = FMA(KP707106781, T1j, T1g); Ip[WS(rs, 3)] = FMA(KP707106781, T1C, T1B); } T1k = Te - Tv; Tw = Te + Tv; T1w = T1t - T1p; T1u = T1p + T1t; } } } { E TT, T1v, T1n, T1o; TT = TJ + TS; T1v = TS - TJ; T1n = T1l - T1m; T1o = T1l + T1m; Ip[WS(rs, 2)] = T1v + T1w; Im[WS(rs, 1)] = T1v - T1w; Rp[0] = Tw + TT; Rm[WS(rs, 3)] = Tw - TT; Ip[0] = T1o + T1u; Im[WS(rs, 3)] = T1o - T1u; Rp[WS(rs, 2)] = T1k + T1n; Rm[WS(rs, 1)] = T1k - T1n; } } } } static const tw_instr twinstr[] = { {TW_CEXP, 1, 1}, {TW_CEXP, 1, 3}, {TW_CEXP, 1, 7}, {TW_NEXT, 1, 0} }; static const hc2c_desc desc = { 8, "hc2cf2_8", twinstr, &GENUS, {44, 20, 30, 0} }; void X(codelet_hc2cf2_8) (planner *p) { X(khc2c_register) (p, hc2cf2_8, &desc, HC2C_VIA_RDFT); } #else /* HAVE_FMA */ /* Generated by: ../../../genfft/gen_hc2c.native -compact -variables 4 -pipeline-latency 4 -twiddle-log3 -precompute-twiddles -n 8 -dit -name hc2cf2_8 -include hc2cf.h */ /* * This function contains 74 FP additions, 44 FP multiplications, * (or, 56 additions, 26 multiplications, 18 fused multiply/add), * 42 stack variables, 1 constants, and 32 memory accesses */ #include "hc2cf.h" static void hc2cf2_8(R *Rp, R *Ip, R *Rm, R *Im, const R *W, stride rs, INT mb, INT me, INT ms) { DK(KP707106781, +0.707106781186547524400844362104849039284835938); { INT m; for (m = mb, W = W + ((mb - 1) * 6); m < me; m = m + 1, Rp = Rp + ms, Ip = Ip + ms, Rm = Rm - ms, Im = Im - ms, W = W + 6, MAKE_VOLATILE_STRIDE(32, rs)) { E T2, T5, T3, T6, T8, Tc, Tg, Ti, Tl, Tm, Tn, Tz, Tp, Tx; { E T4, Tb, T7, Ta; T2 = W[0]; T5 = W[1]; T3 = W[2]; T6 = W[3]; T4 = T2 * T3; Tb = T5 * T3; T7 = T5 * T6; Ta = T2 * T6; T8 = T4 - T7; Tc = Ta + Tb; Tg = T4 + T7; Ti = Ta - Tb; Tl = W[4]; Tm = W[5]; Tn = FMA(T2, Tl, T5 * Tm); Tz = FNMS(Ti, Tl, Tg * Tm); Tp = FNMS(T5, Tl, T2 * Tm); Tx = FMA(Tg, Tl, Ti * Tm); } { E Tf, T1i, TL, T1d, TJ, T17, TV, TY, Ts, T1j, TO, T1a, TC, T16, TQ; E TT; { E T1, T1c, Te, T1b, T9, Td; T1 = Rp[0]; T1c = Rm[0]; T9 = Rp[WS(rs, 2)]; Td = Rm[WS(rs, 2)]; Te = FMA(T8, T9, Tc * Td); T1b = FNMS(Tc, T9, T8 * Td); Tf = T1 + Te; T1i = T1c - T1b; TL = T1 - Te; T1d = T1b + T1c; } { E TF, TW, TI, TX; { E TD, TE, TG, TH; TD = Ip[WS(rs, 3)]; TE = Im[WS(rs, 3)]; TF = FMA(Tl, TD, Tm * TE); TW = FNMS(Tm, TD, Tl * TE); TG = Ip[WS(rs, 1)]; TH = Im[WS(rs, 1)]; TI = FMA(T3, TG, T6 * TH); TX = FNMS(T6, TG, T3 * TH); } TJ = TF + TI; T17 = TW + TX; TV = TF - TI; TY = TW - TX; } { E Tk, TM, Tr, TN; { E Th, Tj, To, Tq; Th = Rp[WS(rs, 1)]; Tj = Rm[WS(rs, 1)]; Tk = FMA(Tg, Th, Ti * Tj); TM = FNMS(Ti, Th, Tg * Tj); To = Rp[WS(rs, 3)]; Tq = Rm[WS(rs, 3)]; Tr = FMA(Tn, To, Tp * Tq); TN = FNMS(Tp, To, Tn * Tq); } Ts = Tk + Tr; T1j = Tk - Tr; TO = TM - TN; T1a = TM + TN; } { E Tw, TR, TB, TS; { E Tu, Tv, Ty, TA; Tu = Ip[0]; Tv = Im[0]; Tw = FMA(T2, Tu, T5 * Tv); TR = FNMS(T5, Tu, T2 * Tv); Ty = Ip[WS(rs, 2)]; TA = Im[WS(rs, 2)]; TB = FMA(Tx, Ty, Tz * TA); TS = FNMS(Tz, Ty, Tx * TA); } TC = Tw + TB; T16 = TR + TS; TQ = Tw - TB; TT = TR - TS; } { E Tt, TK, T1f, T1g; Tt = Tf + Ts; TK = TC + TJ; Rm[WS(rs, 3)] = Tt - TK; Rp[0] = Tt + TK; { E T19, T1e, T15, T18; T19 = T16 + T17; T1e = T1a + T1d; Im[WS(rs, 3)] = T19 - T1e; Ip[0] = T19 + T1e; T15 = Tf - Ts; T18 = T16 - T17; Rm[WS(rs, 1)] = T15 - T18; Rp[WS(rs, 2)] = T15 + T18; } T1f = TJ - TC; T1g = T1d - T1a; Im[WS(rs, 1)] = T1f - T1g; Ip[WS(rs, 2)] = T1f + T1g; { E T11, T1k, T14, T1h, T12, T13; T11 = TL - TO; T1k = T1i - T1j; T12 = TT - TQ; T13 = TV + TY; T14 = KP707106781 * (T12 - T13); T1h = KP707106781 * (T12 + T13); Rm[0] = T11 - T14; Ip[WS(rs, 1)] = T1h + T1k; Rp[WS(rs, 3)] = T11 + T14; Im[WS(rs, 2)] = T1h - T1k; } { E TP, T1m, T10, T1l, TU, TZ; TP = TL + TO; T1m = T1j + T1i; TU = TQ + TT; TZ = TV - TY; T10 = KP707106781 * (TU + TZ); T1l = KP707106781 * (TZ - TU); Rm[WS(rs, 2)] = TP - T10; Ip[WS(rs, 3)] = T1l + T1m; Rp[WS(rs, 1)] = TP + T10; Im[0] = T1l - T1m; } } } } } } static const tw_instr twinstr[] = { {TW_CEXP, 1, 1}, {TW_CEXP, 1, 3}, {TW_CEXP, 1, 7}, {TW_NEXT, 1, 0} }; static const hc2c_desc desc = { 8, "hc2cf2_8", twinstr, &GENUS, {56, 26, 18, 0} }; void X(codelet_hc2cf2_8) (planner *p) { X(khc2c_register) (p, hc2cf2_8, &desc, HC2C_VIA_RDFT); } #endif /* HAVE_FMA */ fftw-3.3.3/rdft/scalar/r2cf/hc2cf2_16.c0000644000175000017500000005471512054410714014202 00000000000000/* * Copyright (c) 2003, 2007-11 Matteo Frigo * Copyright (c) 2003, 2007-11 Massachusetts Institute of Technology * * This program is free software; you can redistribute it and/or modify * it under the terms of the GNU General Public License as published by * the Free Software Foundation; either version 2 of the License, or * (at your option) any later version. * * This program is distributed in the hope that it will be useful, * but WITHOUT ANY WARRANTY; without even the implied warranty of * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the * GNU General Public License for more details. * * You should have received a copy of the GNU General Public License * along with this program; if not, write to the Free Software * Foundation, Inc., 51 Franklin Street, Fifth Floor, Boston, MA 02110-1301 USA * */ /* This file was automatically generated --- DO NOT EDIT */ /* Generated on Sun Nov 25 07:40:42 EST 2012 */ #include "codelet-rdft.h" #ifdef HAVE_FMA /* Generated by: ../../../genfft/gen_hc2c.native -fma -reorder-insns -schedule-for-pipeline -compact -variables 4 -pipeline-latency 4 -twiddle-log3 -precompute-twiddles -n 16 -dit -name hc2cf2_16 -include hc2cf.h */ /* * This function contains 196 FP additions, 134 FP multiplications, * (or, 104 additions, 42 multiplications, 92 fused multiply/add), * 100 stack variables, 3 constants, and 64 memory accesses */ #include "hc2cf.h" static void hc2cf2_16(R *Rp, R *Ip, R *Rm, R *Im, const R *W, stride rs, INT mb, INT me, INT ms) { DK(KP923879532, +0.923879532511286756128183189396788286822416626); DK(KP414213562, +0.414213562373095048801688724209698078569671875); DK(KP707106781, +0.707106781186547524400844362104849039284835938); { INT m; for (m = mb, W = W + ((mb - 1) * 8); m < me; m = m + 1, Rp = Rp + ms, Ip = Ip + ms, Rm = Rm - ms, Im = Im - ms, W = W + 8, MAKE_VOLATILE_STRIDE(64, rs)) { E T3S, T3R; { E T2, Tf, TM, TO, T3, Tg, TN, TS, T4, Tp, T6, T5, Th; T2 = W[0]; Tf = W[2]; TM = W[6]; TO = W[7]; T3 = W[4]; Tg = T2 * Tf; TN = T2 * TM; TS = T2 * TO; T4 = T2 * T3; Tp = Tf * T3; T6 = W[5]; T5 = W[1]; Th = W[3]; { E TZ, Te, T1U, T3A, T3L, T2D, T1G, T2B, T3h, T1R, T2w, T2I, T3i, Tx, T3M; E T1Z, T3w, TL, T26, T25, T37, T1d, T2o, T2l, T3c, T1s, T2m, T2t, T3d, TX; E T10, TV, T2a, TY, T2b; { E TF, TP, TT, Tq, TW, Tz, Tu, TI, TC, T1m, T1f, T1p, T1j, Tr, Ts; E Tv, To, T1W; { E Ti, Tm, T1L, T1O, T1D, T1A, T1x, T2z, T1F, T2y; { E T1, T7, Tb, T3z, T8, T1z, T9, Tc; { E T1i, T1e, T1C, T1y, Tt, Ta, Tl; T1 = Rp[0]; Tt = Tf * T6; Ta = T2 * T6; T7 = FMA(T5, T6, T4); TF = FNMS(T5, T6, T4); TP = FMA(T5, TO, TN); TT = FNMS(T5, TM, TS); Tq = FNMS(Th, T6, Tp); TW = FMA(Th, T6, Tp); Tz = FMA(T5, Th, Tg); Ti = FNMS(T5, Th, Tg); Tl = T2 * Th; Tu = FMA(Th, T3, Tt); TZ = FNMS(Th, T3, Tt); TI = FMA(T5, T3, Ta); Tb = FNMS(T5, T3, Ta); T1i = Ti * T6; T1e = Ti * T3; T1C = Tz * T6; T1y = Tz * T3; Tm = FMA(T5, Tf, Tl); TC = FNMS(T5, Tf, Tl); T3z = Rm[0]; T8 = Rp[WS(rs, 4)]; T1m = FNMS(Tm, T6, T1e); T1f = FMA(Tm, T6, T1e); T1p = FMA(Tm, T3, T1i); T1j = FNMS(Tm, T3, T1i); T1L = FNMS(TC, T6, T1y); T1z = FMA(TC, T6, T1y); T1O = FMA(TC, T3, T1C); T1D = FNMS(TC, T3, T1C); T9 = T7 * T8; Tc = Rm[WS(rs, 4)]; } { E T1u, T1w, T1v, T2x, T3y, T1B, T1E, Td, T3x; T1u = Ip[WS(rs, 7)]; T1w = Im[WS(rs, 7)]; T1A = Ip[WS(rs, 3)]; Td = FMA(Tb, Tc, T9); T3x = T7 * Tc; T1v = TM * T1u; T2x = TM * T1w; Te = T1 + Td; T1U = T1 - Td; T3y = FNMS(Tb, T8, T3x); T1B = T1z * T1A; T1E = Im[WS(rs, 3)]; T1x = FMA(TO, T1w, T1v); T3A = T3y + T3z; T3L = T3z - T3y; T2z = T1z * T1E; T1F = FMA(T1D, T1E, T1B); T2y = FNMS(TO, T1u, T2x); } } { E T1H, T1I, T1J, T1M, T1P, T2A; T1H = Ip[WS(rs, 1)]; T2A = FNMS(T1D, T1A, T2z); T2D = T1x - T1F; T1G = T1x + T1F; T1I = Tf * T1H; T2B = T2y - T2A; T3h = T2y + T2A; T1J = Im[WS(rs, 1)]; T1M = Ip[WS(rs, 5)]; T1P = Im[WS(rs, 5)]; { E Tj, Tk, Tn, T1V; { E T1K, T2F, T1Q, T2H, T2E, T1N, T2G; Tj = Rp[WS(rs, 2)]; T1K = FMA(Th, T1J, T1I); T2E = Tf * T1J; T1N = T1L * T1M; T2G = T1L * T1P; Tk = Ti * Tj; T2F = FNMS(Th, T1H, T2E); T1Q = FMA(T1O, T1P, T1N); T2H = FNMS(T1O, T1M, T2G); Tn = Rm[WS(rs, 2)]; Tr = Rp[WS(rs, 6)]; T1R = T1K + T1Q; T2w = T1Q - T1K; T2I = T2F - T2H; T3i = T2F + T2H; T1V = Ti * Tn; Ts = Tq * Tr; Tv = Rm[WS(rs, 6)]; } To = FMA(Tm, Tn, Tk); T1W = FNMS(Tm, Tj, T1V); } } } { E T19, T1b, T18, T2i, T1a, T2j; { E TE, T22, TK, T24; { E TA, TD, TB, T21, TG, TJ, TH, T23, T1Y, Tw, T1X; TA = Rp[WS(rs, 1)]; Tw = FMA(Tu, Tv, Ts); T1X = Tq * Tv; TD = Rm[WS(rs, 1)]; TB = Tz * TA; Tx = To + Tw; T3M = To - Tw; T1Y = FNMS(Tu, Tr, T1X); T21 = Tz * TD; TG = Rp[WS(rs, 5)]; TJ = Rm[WS(rs, 5)]; T1Z = T1W - T1Y; T3w = T1W + T1Y; TH = TF * TG; T23 = TF * TJ; TE = FMA(TC, TD, TB); T22 = FNMS(TC, TA, T21); TK = FMA(TI, TJ, TH); T24 = FNMS(TI, TG, T23); } { E T15, T17, T16, T2h; T15 = Ip[0]; T17 = Im[0]; TL = TE + TK; T26 = TE - TK; T25 = T22 - T24; T37 = T22 + T24; T16 = T2 * T15; T2h = T2 * T17; T19 = Ip[WS(rs, 4)]; T1b = Im[WS(rs, 4)]; T18 = FMA(T5, T17, T16); T2i = FNMS(T5, T15, T2h); T1a = T3 * T19; T2j = T3 * T1b; } } { E T1n, T1q, T1l, T2q, T1o, T2r; { E T1g, T1k, T1h, T2p, T1c, T2k; T1g = Ip[WS(rs, 2)]; T1k = Im[WS(rs, 2)]; T1c = FMA(T6, T1b, T1a); T2k = FNMS(T6, T19, T2j); T1h = T1f * T1g; T2p = T1f * T1k; T1d = T18 + T1c; T2o = T18 - T1c; T2l = T2i - T2k; T3c = T2i + T2k; T1n = Ip[WS(rs, 6)]; T1q = Im[WS(rs, 6)]; T1l = FMA(T1j, T1k, T1h); T2q = FNMS(T1j, T1g, T2p); T1o = T1m * T1n; T2r = T1m * T1q; } { E TQ, TU, TR, T29, T1r, T2s; TQ = Rp[WS(rs, 7)]; TU = Rm[WS(rs, 7)]; T1r = FMA(T1p, T1q, T1o); T2s = FNMS(T1p, T1n, T2r); TR = TP * TQ; T29 = TP * TU; T1s = T1l + T1r; T2m = T1l - T1r; T2t = T2q - T2s; T3d = T2q + T2s; TX = Rp[WS(rs, 3)]; T10 = Rm[WS(rs, 3)]; TV = FMA(TT, TU, TR); T2a = FNMS(TT, TQ, T29); TY = TW * TX; T2b = TW * T10; } } } } { E T36, T3G, T3b, T3g, T28, T2d, T3F, T39, T3e, T3q, T3C, T3j, T3u, T3t; { E T3D, T1T, T3r, T14, T3E, T3s; { E Ty, T3B, T11, T2c, T13, T3v; T36 = Te - Tx; Ty = Te + Tx; T3B = T3w + T3A; T3G = T3A - T3w; T11 = FMA(TZ, T10, TY); T2c = FNMS(TZ, TX, T2b); { E T1t, T1S, T12, T38; T3b = T1d - T1s; T1t = T1d + T1s; T1S = T1G + T1R; T3g = T1G - T1R; T12 = TV + T11; T28 = TV - T11; T2d = T2a - T2c; T38 = T2a + T2c; T3D = T1S - T1t; T1T = T1t + T1S; T13 = TL + T12; T3F = T12 - TL; T39 = T37 - T38; T3v = T37 + T38; } T3e = T3c - T3d; T3r = T3c + T3d; T3q = Ty - T13; T14 = Ty + T13; T3E = T3B - T3v; T3C = T3v + T3B; T3s = T3h + T3i; T3j = T3h - T3i; } Rm[WS(rs, 7)] = T14 - T1T; Rp[0] = T14 + T1T; Im[WS(rs, 3)] = T3D - T3E; T3u = T3r + T3s; T3t = T3r - T3s; Ip[WS(rs, 4)] = T3D + T3E; } { E T3m, T3a, T3J, T3H; Ip[0] = T3u + T3C; Im[WS(rs, 7)] = T3u - T3C; Rp[WS(rs, 4)] = T3q + T3t; Rm[WS(rs, 3)] = T3q - T3t; T3m = T36 - T39; T3a = T36 + T39; T3J = T3G - T3F; T3H = T3F + T3G; { E T2Q, T20, T3N, T3T, T2J, T2C, T3O, T2f, T34, T30, T2W, T2V, T3U, T2T, T2N; E T2v; { E T2R, T27, T2e, T2S; { E T3n, T3f, T3o, T3k; T2Q = T1U + T1Z; T20 = T1U - T1Z; T3n = T3e - T3b; T3f = T3b + T3e; T3o = T3g + T3j; T3k = T3g - T3j; T3N = T3L - T3M; T3T = T3M + T3L; { E T3p, T3I, T3K, T3l; T3p = T3n - T3o; T3I = T3n + T3o; T3K = T3k - T3f; T3l = T3f + T3k; Rp[WS(rs, 6)] = FMA(KP707106781, T3p, T3m); Rm[WS(rs, 1)] = FNMS(KP707106781, T3p, T3m); Ip[WS(rs, 2)] = FMA(KP707106781, T3I, T3H); Im[WS(rs, 5)] = FMS(KP707106781, T3I, T3H); Ip[WS(rs, 6)] = FMA(KP707106781, T3K, T3J); Im[WS(rs, 1)] = FMS(KP707106781, T3K, T3J); Rp[WS(rs, 2)] = FMA(KP707106781, T3l, T3a); Rm[WS(rs, 5)] = FNMS(KP707106781, T3l, T3a); T2R = T26 + T25; T27 = T25 - T26; T2e = T28 + T2d; T2S = T28 - T2d; } } { E T2Y, T2Z, T2n, T2u; T2J = T2D - T2I; T2Y = T2D + T2I; T2Z = T2B + T2w; T2C = T2w - T2B; T3O = T27 + T2e; T2f = T27 - T2e; T34 = FMA(KP414213562, T2Y, T2Z); T30 = FNMS(KP414213562, T2Z, T2Y); T2W = T2l - T2m; T2n = T2l + T2m; T2u = T2o - T2t; T2V = T2o + T2t; T3U = T2S - T2R; T2T = T2R + T2S; T2N = FNMS(KP414213562, T2n, T2u); T2v = FMA(KP414213562, T2u, T2n); } } { E T33, T2X, T3X, T3Y; { E T2M, T2g, T2O, T2K, T3V, T3W, T2P, T2L; T2M = FNMS(KP707106781, T2f, T20); T2g = FMA(KP707106781, T2f, T20); T33 = FNMS(KP414213562, T2V, T2W); T2X = FMA(KP414213562, T2W, T2V); T2O = FNMS(KP414213562, T2C, T2J); T2K = FMA(KP414213562, T2J, T2C); T3V = FMA(KP707106781, T3U, T3T); T3X = FNMS(KP707106781, T3U, T3T); T3W = T2O - T2N; T2P = T2N + T2O; T3Y = T2K - T2v; T2L = T2v + T2K; Ip[WS(rs, 3)] = FMA(KP923879532, T3W, T3V); Im[WS(rs, 4)] = FMS(KP923879532, T3W, T3V); Rp[WS(rs, 3)] = FMA(KP923879532, T2L, T2g); Rm[WS(rs, 4)] = FNMS(KP923879532, T2L, T2g); Rm[0] = FMA(KP923879532, T2P, T2M); Rp[WS(rs, 7)] = FNMS(KP923879532, T2P, T2M); } { E T32, T3P, T3Q, T35, T2U, T31; T32 = FNMS(KP707106781, T2T, T2Q); T2U = FMA(KP707106781, T2T, T2Q); T31 = T2X + T30; T3S = T30 - T2X; T3R = FNMS(KP707106781, T3O, T3N); T3P = FMA(KP707106781, T3O, T3N); Ip[WS(rs, 7)] = FMA(KP923879532, T3Y, T3X); Im[0] = FMS(KP923879532, T3Y, T3X); Rp[WS(rs, 1)] = FMA(KP923879532, T31, T2U); Rm[WS(rs, 6)] = FNMS(KP923879532, T31, T2U); T3Q = T33 + T34; T35 = T33 - T34; Ip[WS(rs, 1)] = FMA(KP923879532, T3Q, T3P); Im[WS(rs, 6)] = FMS(KP923879532, T3Q, T3P); Rp[WS(rs, 5)] = FMA(KP923879532, T35, T32); Rm[WS(rs, 2)] = FNMS(KP923879532, T35, T32); } } } } } } } Ip[WS(rs, 5)] = FMA(KP923879532, T3S, T3R); Im[WS(rs, 2)] = FMS(KP923879532, T3S, T3R); } } } static const tw_instr twinstr[] = { {TW_CEXP, 1, 1}, {TW_CEXP, 1, 3}, {TW_CEXP, 1, 9}, {TW_CEXP, 1, 15}, {TW_NEXT, 1, 0} }; static const hc2c_desc desc = { 16, "hc2cf2_16", twinstr, &GENUS, {104, 42, 92, 0} }; void X(codelet_hc2cf2_16) (planner *p) { X(khc2c_register) (p, hc2cf2_16, &desc, HC2C_VIA_RDFT); } #else /* HAVE_FMA */ /* Generated by: ../../../genfft/gen_hc2c.native -compact -variables 4 -pipeline-latency 4 -twiddle-log3 -precompute-twiddles -n 16 -dit -name hc2cf2_16 -include hc2cf.h */ /* * This function contains 196 FP additions, 108 FP multiplications, * (or, 156 additions, 68 multiplications, 40 fused multiply/add), * 82 stack variables, 3 constants, and 64 memory accesses */ #include "hc2cf.h" static void hc2cf2_16(R *Rp, R *Ip, R *Rm, R *Im, const R *W, stride rs, INT mb, INT me, INT ms) { DK(KP382683432, +0.382683432365089771728459984030398866761344562); DK(KP923879532, +0.923879532511286756128183189396788286822416626); DK(KP707106781, +0.707106781186547524400844362104849039284835938); { INT m; for (m = mb, W = W + ((mb - 1) * 8); m < me; m = m + 1, Rp = Rp + ms, Ip = Ip + ms, Rm = Rm - ms, Im = Im - ms, W = W + 8, MAKE_VOLATILE_STRIDE(64, rs)) { E T2, T5, Tg, Ti, Tk, To, TE, TC, T6, T3, T8, TW, TJ, Tt, TU; E Tc, Tx, TH, TN, TO, TP, TR, T1f, T1k, T1b, T1i, T1y, T1H, T1u, T1F; { E T7, Tv, Ta, Ts, T4, Tw, Tb, Tr; { E Th, Tn, Tj, Tm; T2 = W[0]; T5 = W[1]; Tg = W[2]; Ti = W[3]; Th = T2 * Tg; Tn = T5 * Tg; Tj = T5 * Ti; Tm = T2 * Ti; Tk = Th - Tj; To = Tm + Tn; TE = Tm - Tn; TC = Th + Tj; T6 = W[5]; T7 = T5 * T6; Tv = Tg * T6; Ta = T2 * T6; Ts = Ti * T6; T3 = W[4]; T4 = T2 * T3; Tw = Ti * T3; Tb = T5 * T3; Tr = Tg * T3; } T8 = T4 + T7; TW = Tv - Tw; TJ = Ta + Tb; Tt = Tr - Ts; TU = Tr + Ts; Tc = Ta - Tb; Tx = Tv + Tw; TH = T4 - T7; TN = W[6]; TO = W[7]; TP = FMA(T2, TN, T5 * TO); TR = FNMS(T5, TN, T2 * TO); { E T1d, T1e, T19, T1a; T1d = Tk * T6; T1e = To * T3; T1f = T1d - T1e; T1k = T1d + T1e; T19 = Tk * T3; T1a = To * T6; T1b = T19 + T1a; T1i = T19 - T1a; } { E T1w, T1x, T1s, T1t; T1w = TC * T6; T1x = TE * T3; T1y = T1w - T1x; T1H = T1w + T1x; T1s = TC * T3; T1t = TE * T6; T1u = T1s + T1t; T1F = T1s - T1t; } } { E Tf, T3r, T1N, T3e, TA, T3s, T1Q, T3b, TM, T2M, T1W, T2w, TZ, T2N, T21; E T2x, T1B, T1K, T2V, T2W, T2X, T2Y, T2j, T2D, T2o, T2E, T18, T1n, T2Q, T2R; E T2S, T2T, T28, T2A, T2d, T2B; { E T1, T3d, Te, T3c, T9, Td; T1 = Rp[0]; T3d = Rm[0]; T9 = Rp[WS(rs, 4)]; Td = Rm[WS(rs, 4)]; Te = FMA(T8, T9, Tc * Td); T3c = FNMS(Tc, T9, T8 * Td); Tf = T1 + Te; T3r = T3d - T3c; T1N = T1 - Te; T3e = T3c + T3d; } { E Tq, T1O, Tz, T1P; { E Tl, Tp, Tu, Ty; Tl = Rp[WS(rs, 2)]; Tp = Rm[WS(rs, 2)]; Tq = FMA(Tk, Tl, To * Tp); T1O = FNMS(To, Tl, Tk * Tp); Tu = Rp[WS(rs, 6)]; Ty = Rm[WS(rs, 6)]; Tz = FMA(Tt, Tu, Tx * Ty); T1P = FNMS(Tx, Tu, Tt * Ty); } TA = Tq + Tz; T3s = Tq - Tz; T1Q = T1O - T1P; T3b = T1O + T1P; } { E TG, T1S, TL, T1T, T1U, T1V; { E TD, TF, TI, TK; TD = Rp[WS(rs, 1)]; TF = Rm[WS(rs, 1)]; TG = FMA(TC, TD, TE * TF); T1S = FNMS(TE, TD, TC * TF); TI = Rp[WS(rs, 5)]; TK = Rm[WS(rs, 5)]; TL = FMA(TH, TI, TJ * TK); T1T = FNMS(TJ, TI, TH * TK); } TM = TG + TL; T2M = T1S + T1T; T1U = T1S - T1T; T1V = TG - TL; T1W = T1U - T1V; T2w = T1V + T1U; } { E TT, T1Y, TY, T1Z, T1X, T20; { E TQ, TS, TV, TX; TQ = Rp[WS(rs, 7)]; TS = Rm[WS(rs, 7)]; TT = FMA(TP, TQ, TR * TS); T1Y = FNMS(TR, TQ, TP * TS); TV = Rp[WS(rs, 3)]; TX = Rm[WS(rs, 3)]; TY = FMA(TU, TV, TW * TX); T1Z = FNMS(TW, TV, TU * TX); } TZ = TT + TY; T2N = T1Y + T1Z; T1X = TT - TY; T20 = T1Y - T1Z; T21 = T1X + T20; T2x = T1X - T20; } { E T1r, T2k, T1J, T2h, T1A, T2l, T1E, T2g; { E T1p, T1q, T1G, T1I; T1p = Ip[WS(rs, 7)]; T1q = Im[WS(rs, 7)]; T1r = FMA(TN, T1p, TO * T1q); T2k = FNMS(TO, T1p, TN * T1q); T1G = Ip[WS(rs, 5)]; T1I = Im[WS(rs, 5)]; T1J = FMA(T1F, T1G, T1H * T1I); T2h = FNMS(T1H, T1G, T1F * T1I); } { E T1v, T1z, T1C, T1D; T1v = Ip[WS(rs, 3)]; T1z = Im[WS(rs, 3)]; T1A = FMA(T1u, T1v, T1y * T1z); T2l = FNMS(T1y, T1v, T1u * T1z); T1C = Ip[WS(rs, 1)]; T1D = Im[WS(rs, 1)]; T1E = FMA(Tg, T1C, Ti * T1D); T2g = FNMS(Ti, T1C, Tg * T1D); } T1B = T1r + T1A; T1K = T1E + T1J; T2V = T1B - T1K; T2W = T2k + T2l; T2X = T2g + T2h; T2Y = T2W - T2X; { E T2f, T2i, T2m, T2n; T2f = T1r - T1A; T2i = T2g - T2h; T2j = T2f - T2i; T2D = T2f + T2i; T2m = T2k - T2l; T2n = T1E - T1J; T2o = T2m + T2n; T2E = T2m - T2n; } } { E T14, T24, T1m, T2b, T17, T25, T1h, T2a; { E T12, T13, T1j, T1l; T12 = Ip[0]; T13 = Im[0]; T14 = FMA(T2, T12, T5 * T13); T24 = FNMS(T5, T12, T2 * T13); T1j = Ip[WS(rs, 6)]; T1l = Im[WS(rs, 6)]; T1m = FMA(T1i, T1j, T1k * T1l); T2b = FNMS(T1k, T1j, T1i * T1l); } { E T15, T16, T1c, T1g; T15 = Ip[WS(rs, 4)]; T16 = Im[WS(rs, 4)]; T17 = FMA(T3, T15, T6 * T16); T25 = FNMS(T6, T15, T3 * T16); T1c = Ip[WS(rs, 2)]; T1g = Im[WS(rs, 2)]; T1h = FMA(T1b, T1c, T1f * T1g); T2a = FNMS(T1f, T1c, T1b * T1g); } T18 = T14 + T17; T1n = T1h + T1m; T2Q = T18 - T1n; T2R = T24 + T25; T2S = T2a + T2b; T2T = T2R - T2S; { E T26, T27, T29, T2c; T26 = T24 - T25; T27 = T1h - T1m; T28 = T26 + T27; T2A = T26 - T27; T29 = T14 - T17; T2c = T2a - T2b; T2d = T29 - T2c; T2B = T29 + T2c; } } { E T23, T2r, T3A, T3C, T2q, T3B, T2u, T3x; { E T1R, T22, T3y, T3z; T1R = T1N - T1Q; T22 = KP707106781 * (T1W - T21); T23 = T1R + T22; T2r = T1R - T22; T3y = KP707106781 * (T2x - T2w); T3z = T3s + T3r; T3A = T3y + T3z; T3C = T3z - T3y; } { E T2e, T2p, T2s, T2t; T2e = FMA(KP923879532, T28, KP382683432 * T2d); T2p = FNMS(KP923879532, T2o, KP382683432 * T2j); T2q = T2e + T2p; T3B = T2p - T2e; T2s = FNMS(KP923879532, T2d, KP382683432 * T28); T2t = FMA(KP382683432, T2o, KP923879532 * T2j); T2u = T2s - T2t; T3x = T2s + T2t; } Rm[WS(rs, 4)] = T23 - T2q; Im[WS(rs, 4)] = T3x - T3A; Rp[WS(rs, 3)] = T23 + T2q; Ip[WS(rs, 3)] = T3x + T3A; Rm[0] = T2r - T2u; Im[0] = T3B - T3C; Rp[WS(rs, 7)] = T2r + T2u; Ip[WS(rs, 7)] = T3B + T3C; } { E T2P, T31, T3m, T3o, T30, T3n, T34, T3j; { E T2L, T2O, T3k, T3l; T2L = Tf - TA; T2O = T2M - T2N; T2P = T2L + T2O; T31 = T2L - T2O; T3k = TZ - TM; T3l = T3e - T3b; T3m = T3k + T3l; T3o = T3l - T3k; } { E T2U, T2Z, T32, T33; T2U = T2Q + T2T; T2Z = T2V - T2Y; T30 = KP707106781 * (T2U + T2Z); T3n = KP707106781 * (T2Z - T2U); T32 = T2T - T2Q; T33 = T2V + T2Y; T34 = KP707106781 * (T32 - T33); T3j = KP707106781 * (T32 + T33); } Rm[WS(rs, 5)] = T2P - T30; Im[WS(rs, 5)] = T3j - T3m; Rp[WS(rs, 2)] = T2P + T30; Ip[WS(rs, 2)] = T3j + T3m; Rm[WS(rs, 1)] = T31 - T34; Im[WS(rs, 1)] = T3n - T3o; Rp[WS(rs, 6)] = T31 + T34; Ip[WS(rs, 6)] = T3n + T3o; } { E T2z, T2H, T3u, T3w, T2G, T3v, T2K, T3p; { E T2v, T2y, T3q, T3t; T2v = T1N + T1Q; T2y = KP707106781 * (T2w + T2x); T2z = T2v + T2y; T2H = T2v - T2y; T3q = KP707106781 * (T1W + T21); T3t = T3r - T3s; T3u = T3q + T3t; T3w = T3t - T3q; } { E T2C, T2F, T2I, T2J; T2C = FMA(KP382683432, T2A, KP923879532 * T2B); T2F = FNMS(KP382683432, T2E, KP923879532 * T2D); T2G = T2C + T2F; T3v = T2F - T2C; T2I = FNMS(KP382683432, T2B, KP923879532 * T2A); T2J = FMA(KP923879532, T2E, KP382683432 * T2D); T2K = T2I - T2J; T3p = T2I + T2J; } Rm[WS(rs, 6)] = T2z - T2G; Im[WS(rs, 6)] = T3p - T3u; Rp[WS(rs, 1)] = T2z + T2G; Ip[WS(rs, 1)] = T3p + T3u; Rm[WS(rs, 2)] = T2H - T2K; Im[WS(rs, 2)] = T3v - T3w; Rp[WS(rs, 5)] = T2H + T2K; Ip[WS(rs, 5)] = T3v + T3w; } { E T11, T35, T3g, T3i, T1M, T3h, T38, T39; { E TB, T10, T3a, T3f; TB = Tf + TA; T10 = TM + TZ; T11 = TB + T10; T35 = TB - T10; T3a = T2M + T2N; T3f = T3b + T3e; T3g = T3a + T3f; T3i = T3f - T3a; } { E T1o, T1L, T36, T37; T1o = T18 + T1n; T1L = T1B + T1K; T1M = T1o + T1L; T3h = T1L - T1o; T36 = T2R + T2S; T37 = T2W + T2X; T38 = T36 - T37; T39 = T36 + T37; } Rm[WS(rs, 7)] = T11 - T1M; Im[WS(rs, 7)] = T39 - T3g; Rp[0] = T11 + T1M; Ip[0] = T39 + T3g; Rm[WS(rs, 3)] = T35 - T38; Im[WS(rs, 3)] = T3h - T3i; Rp[WS(rs, 4)] = T35 + T38; Ip[WS(rs, 4)] = T3h + T3i; } } } } } static const tw_instr twinstr[] = { {TW_CEXP, 1, 1}, {TW_CEXP, 1, 3}, {TW_CEXP, 1, 9}, {TW_CEXP, 1, 15}, {TW_NEXT, 1, 0} }; static const hc2c_desc desc = { 16, "hc2cf2_16", twinstr, &GENUS, {156, 68, 40, 0} }; void X(codelet_hc2cf2_16) (planner *p) { X(khc2c_register) (p, hc2cf2_16, &desc, HC2C_VIA_RDFT); } #endif /* HAVE_FMA */ fftw-3.3.3/rdft/scalar/r2cf/hc2cf2_32.c0000644000175000017500000015043712054410724014177 00000000000000/* * Copyright (c) 2003, 2007-11 Matteo Frigo * Copyright (c) 2003, 2007-11 Massachusetts Institute of Technology * * This program is free software; you can redistribute it and/or modify * it under the terms of the GNU General Public License as published by * the Free Software Foundation; either version 2 of the License, or * (at your option) any later version. * * This program is distributed in the hope that it will be useful, * but WITHOUT ANY WARRANTY; without even the implied warranty of * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the * GNU General Public License for more details. * * You should have received a copy of the GNU General Public License * along with this program; if not, write to the Free Software * Foundation, Inc., 51 Franklin Street, Fifth Floor, Boston, MA 02110-1301 USA * */ /* This file was automatically generated --- DO NOT EDIT */ /* Generated on Sun Nov 25 07:40:43 EST 2012 */ #include "codelet-rdft.h" #ifdef HAVE_FMA /* Generated by: ../../../genfft/gen_hc2c.native -fma -reorder-insns -schedule-for-pipeline -compact -variables 4 -pipeline-latency 4 -twiddle-log3 -precompute-twiddles -n 32 -dit -name hc2cf2_32 -include hc2cf.h */ /* * This function contains 488 FP additions, 350 FP multiplications, * (or, 236 additions, 98 multiplications, 252 fused multiply/add), * 181 stack variables, 7 constants, and 128 memory accesses */ #include "hc2cf.h" static void hc2cf2_32(R *Rp, R *Ip, R *Rm, R *Im, const R *W, stride rs, INT mb, INT me, INT ms) { DK(KP831469612, +0.831469612302545237078788377617905756738560812); DK(KP980785280, +0.980785280403230449126182236134239036973933731); DK(KP198912367, +0.198912367379658006911597622644676228597850501); DK(KP668178637, +0.668178637919298919997757686523080761552472251); DK(KP923879532, +0.923879532511286756128183189396788286822416626); DK(KP414213562, +0.414213562373095048801688724209698078569671875); DK(KP707106781, +0.707106781186547524400844362104849039284835938); { INT m; for (m = mb, W = W + ((mb - 1) * 8); m < me; m = m + 1, Rp = Rp + ms, Ip = Ip + ms, Rm = Rm - ms, Im = Im - ms, W = W + 8, MAKE_VOLATILE_STRIDE(128, rs)) { E T9A, T9z; { E T2, T8, T3, T6, Te, Tr, T18, T4, Ta, Tz, T1n, T10, Ti, T5, Tc; T2 = W[0]; T8 = W[4]; T3 = W[2]; T6 = W[3]; Te = W[6]; Tr = T2 * T8; T18 = T3 * T8; T4 = T2 * T3; Ta = T2 * T6; Tz = T3 * Te; T1n = T8 * Te; T10 = T2 * Te; Ti = W[7]; T5 = W[1]; Tc = W[5]; { E T34, T31, T2X, T2T, Tq, T46, T8H, T97, TH, T98, T4b, T8D, TZ, T7f, T4j; E T6t, T1g, T7g, T4q, T6u, T1J, T7m, T6y, T4z, T7l, T8d, T6x, T4G, T2k, T7o; E T7r, T8e, T6B, T4O, T6A, T4V, T7L, T3G, T6P, T61, T6M, T5E, T8n, T7J, T5s; E T6I, T2N, T7A, T55, T6F, T7x, T8i, T5L, T62, T43, T7G, T5S, T63, T7O, T8o; E T2U, T2R, T2V, T58, T3a, T5h, T2Y, T32, T35; { E T1K, T23, T1N, T26, T2b, T1U, T3C, T3j, T3z, T3f, T1R, T29, TR, Th, T2J; E T2F, Td, TP, T3r, T3n, T2w, T2s, T3Q, T3M, T1Z, T1V, T2g, T2c; { E T11, T1C, TM, Tb, TJ, T7, T1o, T19, T1w, T1F, T15, T1s, T1d, T1z, TW; E TS, Ty, T48, TG, T4a; { E T1, TA, Ts, TE, Tw, Tn, Tj, T8G, Tk, To, T14; T1 = Rp[0]; TA = FMA(T6, Ti, Tz); T1K = FNMS(T6, Ti, Tz); T14 = T2 * Ti; { E T1r, TD, T1c, Tv; T1r = T8 * Ti; TD = T3 * Ti; T11 = FNMS(T5, Ti, T10); T1C = FMA(T5, Ti, T10); TM = FMA(T5, T3, Ta); Tb = FNMS(T5, T3, Ta); TJ = FNMS(T5, T6, T4); T7 = FMA(T5, T6, T4); T1o = FMA(Tc, Ti, T1n); T23 = FMA(T6, Tc, T18); T19 = FNMS(T6, Tc, T18); T1w = FNMS(T5, Tc, Tr); Ts = FMA(T5, Tc, Tr); T1c = T3 * Tc; Tv = T2 * Tc; T1F = FNMS(T5, Te, T14); T15 = FMA(T5, Te, T14); T1s = FNMS(Tc, Te, T1r); T1N = FMA(T6, Te, TD); TE = FNMS(T6, Te, TD); { E T1T, T3i, T3e, T1Q; T1T = TJ * Tc; T3i = TJ * Ti; T3e = TJ * Te; T1Q = TJ * T8; { E Tg, T2I, T2E, T9; Tg = T7 * Tc; T2I = T7 * Ti; T2E = T7 * Te; T9 = T7 * T8; { E T3q, T3m, T2v, T2r; T3q = T19 * Ti; T3m = T19 * Te; T2v = T1w * Ti; T2r = T1w * Te; { E T2W, T2S, T3P, T3L; T2W = T23 * Ti; T2S = T23 * Te; T3P = Ts * Ti; T3L = Ts * Te; T26 = FNMS(T6, T8, T1c); T1d = FMA(T6, T8, T1c); T1z = FMA(T5, T8, Tv); Tw = FNMS(T5, T8, Tv); T2b = FNMS(TM, T8, T1T); T1U = FMA(TM, T8, T1T); T3C = FNMS(TM, Te, T3i); T3j = FMA(TM, Te, T3i); T3z = FMA(TM, Ti, T3e); T3f = FNMS(TM, Ti, T3e); T1R = FNMS(TM, Tc, T1Q); T29 = FMA(TM, Tc, T1Q); TR = FNMS(Tb, T8, Tg); Th = FMA(Tb, T8, Tg); T34 = FMA(Tb, Te, T2I); T2J = FNMS(Tb, Te, T2I); T31 = FNMS(Tb, Ti, T2E); T2F = FMA(Tb, Ti, T2E); Td = FNMS(Tb, Tc, T9); TP = FMA(Tb, Tc, T9); T2X = FNMS(T26, Te, T2W); T2T = FMA(T26, Ti, T2S); T3r = FNMS(T1d, Te, T3q); T3n = FMA(T1d, Ti, T3m); T2w = FNMS(T1z, Te, T2v); T2s = FMA(T1z, Ti, T2r); T3Q = FNMS(Tw, Te, T3P); T3M = FMA(Tw, Ti, T3L); { E T1Y, T1S, T2f, T2a; T1Y = T1R * Ti; T1S = T1R * Te; T2f = T29 * Ti; T2a = T29 * Te; { E Tm, Tf, TV, TQ; Tm = Td * Ti; Tf = Td * Te; TV = TP * Ti; TQ = TP * Te; T1Z = FNMS(T1U, Te, T1Y); T1V = FMA(T1U, Ti, T1S); T2g = FNMS(T2b, Te, T2f); T2c = FMA(T2b, Ti, T2a); Tn = FNMS(Th, Te, Tm); Tj = FMA(Th, Ti, Tf); TW = FNMS(TR, Te, TV); TS = FMA(TR, Ti, TQ); T8G = Rm[0]; } } } } } } } Tk = Rp[WS(rs, 8)]; To = Rm[WS(rs, 8)]; { E Tt, Tx, Tu, T47, TB, TF, TC, T49; { E Tl, T8E, Tp, T8F; Tt = Rp[WS(rs, 4)]; Tx = Rm[WS(rs, 4)]; Tl = Tj * Tk; T8E = Tj * To; Tu = Ts * Tt; T47 = Ts * Tx; Tp = FMA(Tn, To, Tl); T8F = FNMS(Tn, Tk, T8E); TB = Rp[WS(rs, 12)]; TF = Rm[WS(rs, 12)]; Tq = T1 + Tp; T46 = T1 - Tp; T8H = T8F + T8G; T97 = T8G - T8F; TC = TA * TB; T49 = TA * TF; } Ty = FMA(Tw, Tx, Tu); T48 = FNMS(Tw, Tt, T47); TG = FMA(TE, TF, TC); T4a = FNMS(TE, TB, T49); } } { E TT, TX, TO, T4f, TU, T4g; { E TK, TN, TL, T4e; TK = Rp[WS(rs, 2)]; TN = Rm[WS(rs, 2)]; TH = Ty + TG; T98 = Ty - TG; T4b = T48 - T4a; T8D = T48 + T4a; TL = TJ * TK; T4e = TJ * TN; TT = Rp[WS(rs, 10)]; TX = Rm[WS(rs, 10)]; TO = FMA(TM, TN, TL); T4f = FNMS(TM, TK, T4e); TU = TS * TT; T4g = TS * TX; } { E T17, T4m, T1a, T1e, T4d, T4i; { E T12, T16, TY, T4h, T13, T4l; T12 = Rp[WS(rs, 14)]; T16 = Rm[WS(rs, 14)]; TY = FMA(TW, TX, TU); T4h = FNMS(TW, TT, T4g); T13 = T11 * T12; T4l = T11 * T16; TZ = TO + TY; T4d = TO - TY; T7f = T4f + T4h; T4i = T4f - T4h; T17 = FMA(T15, T16, T13); T4m = FNMS(T15, T12, T4l); } T4j = T4d + T4i; T6t = T4i - T4d; T1a = Rp[WS(rs, 6)]; T1e = Rm[WS(rs, 6)]; { E T1m, T4B, T1H, T4x, T1x, T1A, T1u, T4D, T1y, T4u; { E T1D, T1G, T1E, T4w; { E T1f, T4o, T4k, T4p; { E T1j, T1l, T1b, T4n, T1k, T4A; T1j = Rp[WS(rs, 1)]; T1l = Rm[WS(rs, 1)]; T1b = T19 * T1a; T4n = T19 * T1e; T1k = T7 * T1j; T4A = T7 * T1l; T1f = FMA(T1d, T1e, T1b); T4o = FNMS(T1d, T1a, T4n); T1m = FMA(Tb, T1l, T1k); T4B = FNMS(Tb, T1j, T4A); } T1g = T17 + T1f; T4k = T17 - T1f; T7g = T4m + T4o; T4p = T4m - T4o; T1D = Rp[WS(rs, 13)]; T1G = Rm[WS(rs, 13)]; T4q = T4k - T4p; T6u = T4k + T4p; T1E = T1C * T1D; T4w = T1C * T1G; } { E T1p, T1t, T1q, T4C; T1p = Rp[WS(rs, 9)]; T1t = Rm[WS(rs, 9)]; T1H = FMA(T1F, T1G, T1E); T4x = FNMS(T1F, T1D, T4w); T1q = T1o * T1p; T4C = T1o * T1t; T1x = Rp[WS(rs, 5)]; T1A = Rm[WS(rs, 5)]; T1u = FMA(T1s, T1t, T1q); T4D = FNMS(T1s, T1p, T4C); T1y = T1w * T1x; T4u = T1w * T1A; } } { E T4t, T1v, T7j, T4E, T1B, T4v; T4t = T1m - T1u; T1v = T1m + T1u; T7j = T4B + T4D; T4E = T4B - T4D; T1B = FMA(T1z, T1A, T1y); T4v = FNMS(T1z, T1x, T4u); { E T4F, T1I, T4y, T7k; T4F = T1B - T1H; T1I = T1B + T1H; T4y = T4v - T4x; T7k = T4v + T4x; T1J = T1v + T1I; T7m = T1v - T1I; T6y = T4t - T4y; T4z = T4t + T4y; T7l = T7j - T7k; T8d = T7j + T7k; T6x = T4E + T4F; T4G = T4E - T4F; } } } } } } { E T5C, T3u, T5y, T7H, T5Z, T3F, T60, T5A, T4T, T4U; { E T1P, T4Q, T2i, T4M, T21, T4S, T28, T4K; { E T1L, T1O, T1W, T20; T1L = Rp[WS(rs, 15)]; T1O = Rm[WS(rs, 15)]; { E T2d, T2h, T1M, T4P, T2e, T4L; T2d = Rp[WS(rs, 11)]; T2h = Rm[WS(rs, 11)]; T1M = T1K * T1L; T4P = T1K * T1O; T2e = T2c * T2d; T4L = T2c * T2h; T1P = FMA(T1N, T1O, T1M); T4Q = FNMS(T1N, T1L, T4P); T2i = FMA(T2g, T2h, T2e); T4M = FNMS(T2g, T2d, T4L); } T1W = Rp[WS(rs, 7)]; T20 = Rm[WS(rs, 7)]; { E T24, T27, T1X, T4R, T25, T4J; T24 = Rp[WS(rs, 3)]; T27 = Rm[WS(rs, 3)]; T1X = T1V * T1W; T4R = T1V * T20; T25 = T23 * T24; T4J = T23 * T27; T21 = FMA(T1Z, T20, T1X); T4S = FNMS(T1Z, T1W, T4R); T28 = FMA(T26, T27, T25); T4K = FNMS(T26, T24, T4J); } } { E T4I, T22, T7p, T2j, T7q, T4N; T4I = T1P - T21; T22 = T1P + T21; T7p = T4Q + T4S; T4T = T4Q - T4S; T4U = T28 - T2i; T2j = T28 + T2i; T7q = T4K + T4M; T4N = T4K - T4M; T2k = T22 + T2j; T7o = T22 - T2j; T7r = T7p - T7q; T8e = T7p + T7q; T6B = T4I - T4N; T4O = T4I + T4N; } } { E T3l, T5W, T3E, T3v, T3t, T3w, T3x, T5Y, T3A, T3B, T3D, T3y, T5z; { E T3g, T3k, T3h, T5V; T3g = Ip[WS(rs, 15)]; T3k = Im[WS(rs, 15)]; T3A = Ip[WS(rs, 11)]; T6A = T4T + T4U; T4V = T4T - T4U; T3h = T3f * T3g; T5V = T3f * T3k; T3B = T3z * T3A; T3D = Im[WS(rs, 11)]; T3l = FMA(T3j, T3k, T3h); T5W = FNMS(T3j, T3g, T5V); } { E T3o, T5B, T3s, T3p, T5X; T3o = Ip[WS(rs, 7)]; T3E = FMA(T3C, T3D, T3B); T5B = T3z * T3D; T3s = Im[WS(rs, 7)]; T3p = T3n * T3o; T3v = Ip[WS(rs, 3)]; T5C = FNMS(T3C, T3A, T5B); T5X = T3n * T3s; T3t = FMA(T3r, T3s, T3p); T3w = TP * T3v; T3x = Im[WS(rs, 3)]; T5Y = FNMS(T3r, T3o, T5X); } T3u = T3l + T3t; T5y = T3l - T3t; T3y = FMA(TR, T3x, T3w); T5z = TP * T3x; T7H = T5W + T5Y; T5Z = T5W - T5Y; T3F = T3y + T3E; T60 = T3E - T3y; T5A = FNMS(TR, T3v, T5z); } { E T2t, T2q, T2u, T5n, T2L, T53, T2x, T2A, T2C; { E T2n, T2o, T2p, T2G, T2K, T5D, T7I, T5m, T2H, T52; T2n = Ip[0]; T7L = T3u - T3F; T3G = T3u + T3F; T5D = T5A - T5C; T7I = T5A + T5C; T6P = T60 - T5Z; T61 = T5Z + T60; T6M = T5y - T5D; T5E = T5y + T5D; T8n = T7H + T7I; T7J = T7H - T7I; T2o = T2 * T2n; T2p = Im[0]; T2G = Ip[WS(rs, 12)]; T2K = Im[WS(rs, 12)]; T2t = Ip[WS(rs, 8)]; T2q = FMA(T5, T2p, T2o); T5m = T2 * T2p; T2H = T2F * T2G; T52 = T2F * T2K; T2u = T2s * T2t; T5n = FNMS(T5, T2n, T5m); T2L = FMA(T2J, T2K, T2H); T53 = FNMS(T2J, T2G, T52); T2x = Im[WS(rs, 8)]; T2A = Ip[WS(rs, 4)]; T2C = Im[WS(rs, 4)]; } { E T3N, T3K, T3O, T5H, T41, T5Q, T3R, T3U, T3W; { E T3H, T3I, T3J, T3Y, T40, T5G, T3Z, T5P; { E T2z, T4Z, T5p, T2D, T51, T7v, T5q; T3H = Ip[WS(rs, 1)]; { E T2y, T5o, T2B, T50; T2y = FMA(T2w, T2x, T2u); T5o = T2s * T2x; T2B = T8 * T2A; T50 = T8 * T2C; T2z = T2q + T2y; T4Z = T2q - T2y; T5p = FNMS(T2w, T2t, T5o); T2D = FMA(Tc, T2C, T2B); T51 = FNMS(Tc, T2A, T50); T3I = T3 * T3H; } T7v = T5n + T5p; T5q = T5n - T5p; { E T2M, T5r, T7w, T54; T2M = T2D + T2L; T5r = T2D - T2L; T7w = T51 + T53; T54 = T51 - T53; T5s = T5q - T5r; T6I = T5q + T5r; T2N = T2z + T2M; T7A = T2z - T2M; T55 = T4Z + T54; T6F = T4Z - T54; T7x = T7v - T7w; T8i = T7v + T7w; T3J = Im[WS(rs, 1)]; } } T3Y = Ip[WS(rs, 5)]; T40 = Im[WS(rs, 5)]; T3N = Ip[WS(rs, 9)]; T3K = FMA(T6, T3J, T3I); T5G = T3 * T3J; T3Z = Td * T3Y; T5P = Td * T40; T3O = T3M * T3N; T5H = FNMS(T6, T3H, T5G); T41 = FMA(Th, T40, T3Z); T5Q = FNMS(Th, T3Y, T5P); T3R = Im[WS(rs, 9)]; T3U = Ip[WS(rs, 13)]; T3W = Im[WS(rs, 13)]; } { E T2O, T2P, T2Q, T37, T39, T57, T38, T5g; { E T3T, T5F, T5J, T3X, T5O, T7M, T5K; T2O = Ip[WS(rs, 2)]; { E T3S, T5I, T3V, T5N; T3S = FMA(T3Q, T3R, T3O); T5I = T3M * T3R; T3V = Te * T3U; T5N = Te * T3W; T3T = T3K + T3S; T5F = T3K - T3S; T5J = FNMS(T3Q, T3N, T5I); T3X = FMA(Ti, T3W, T3V); T5O = FNMS(Ti, T3U, T5N); T2P = T29 * T2O; } T7M = T5H + T5J; T5K = T5H - T5J; { E T42, T5M, T7N, T5R; T42 = T3X + T41; T5M = T3X - T41; T7N = T5O + T5Q; T5R = T5O - T5Q; T5L = T5F + T5K; T62 = T5K - T5F; T43 = T3T + T42; T7G = T42 - T3T; T5S = T5M - T5R; T63 = T5M + T5R; T7O = T7M - T7N; T8o = T7M + T7N; T2Q = Im[WS(rs, 2)]; } } T37 = Ip[WS(rs, 6)]; T39 = Im[WS(rs, 6)]; T2U = Ip[WS(rs, 10)]; T2R = FMA(T2b, T2Q, T2P); T57 = T29 * T2Q; T38 = T1R * T37; T5g = T1R * T39; T2V = T2T * T2U; T58 = FNMS(T2b, T2O, T57); T3a = FMA(T1U, T39, T38); T5h = FNMS(T1U, T37, T5g); T2Y = Im[WS(rs, 10)]; T32 = Ip[WS(rs, 14)]; T35 = Im[WS(rs, 14)]; } } } } } { E T5c, T5t, T5j, T5u, T88, T90, T8Z, T8b; { E T7e, T8T, T7y, T7D, T7h, T8U, T8S, T8R; { E T8c, T1i, T8A, T8z, T8O, T8J, T8N, T2l, T8L, T45, T8t, T8l, T8u, T8q, T3c; E T8k, T8p, T8w, T2m; { E T8x, T8y, T8j, T8C, T8I; { E TI, T30, T56, T5a, T36, T5f, T1h, T7B, T5b; TI = Tq + TH; T7e = Tq - TH; { E T2Z, T59, T33, T5e; T2Z = FMA(T2X, T2Y, T2V); T59 = T2T * T2Y; T33 = T31 * T32; T5e = T31 * T35; T30 = T2R + T2Z; T56 = T2R - T2Z; T5a = FNMS(T2X, T2U, T59); T36 = FMA(T34, T35, T33); T5f = FNMS(T34, T32, T5e); T1h = TZ + T1g; T8T = T1g - TZ; } T7B = T58 + T5a; T5b = T58 - T5a; { E T3b, T5d, T7C, T5i; T3b = T36 + T3a; T5d = T36 - T3a; T7C = T5f + T5h; T5i = T5f - T5h; T5c = T56 + T5b; T5t = T5b - T56; T3c = T30 + T3b; T7y = T3b - T30; T5j = T5d - T5i; T5u = T5d + T5i; T7D = T7B - T7C; T8j = T7B + T7C; T8c = TI - T1h; T1i = TI + T1h; } } T8k = T8i - T8j; T8x = T8i + T8j; T8y = T8n + T8o; T8p = T8n - T8o; T7h = T7f - T7g; T8C = T7f + T7g; T8I = T8D + T8H; T8U = T8H - T8D; T8A = T8x + T8y; T8z = T8x - T8y; T8O = T8I - T8C; T8J = T8C + T8I; } { E T8h, T8m, T3d, T44; T8h = T2N - T3c; T3d = T2N + T3c; T44 = T3G + T43; T8m = T3G - T43; T8N = T2k - T1J; T2l = T1J + T2k; T8L = T44 - T3d; T45 = T3d + T44; T8t = T8k - T8h; T8l = T8h + T8k; T8u = T8m + T8p; T8q = T8m - T8p; } T8w = T1i - T2l; T2m = T1i + T2l; { E T8s, T8P, T8Q, T8v; { E T8r, T8M, T8K, T8g, T8B, T8f; T8S = T8q - T8l; T8r = T8l + T8q; T8B = T8d + T8e; T8f = T8d - T8e; Rp[0] = T2m + T45; Rm[WS(rs, 15)] = T2m - T45; Rp[WS(rs, 8)] = T8w + T8z; Rm[WS(rs, 7)] = T8w - T8z; T8M = T8J - T8B; T8K = T8B + T8J; T8g = T8c + T8f; T8s = T8c - T8f; T8R = T8O - T8N; T8P = T8N + T8O; Ip[WS(rs, 8)] = T8L + T8M; Im[WS(rs, 7)] = T8L - T8M; Ip[0] = T8A + T8K; Im[WS(rs, 15)] = T8A - T8K; Rp[WS(rs, 4)] = FMA(KP707106781, T8r, T8g); Rm[WS(rs, 11)] = FNMS(KP707106781, T8r, T8g); T8Q = T8t + T8u; T8v = T8t - T8u; } Ip[WS(rs, 4)] = FMA(KP707106781, T8Q, T8P); Im[WS(rs, 11)] = FMS(KP707106781, T8Q, T8P); Rp[WS(rs, 12)] = FMA(KP707106781, T8v, T8s); Rm[WS(rs, 3)] = FNMS(KP707106781, T8v, T8s); } } { E T7P, T7W, T7i, T7K, T8a, T86, T91, T8V, T8W, T7t, T7T, T7F, T92, T7Z, T89; E T83; { E T7X, T7n, T7s, T7Y, T84, T85; T7P = T7L - T7O; T84 = T7L + T7O; Ip[WS(rs, 12)] = FMA(KP707106781, T8S, T8R); Im[WS(rs, 3)] = FMS(KP707106781, T8S, T8R); T7W = T7e + T7h; T7i = T7e - T7h; T85 = T7J + T7G; T7K = T7G - T7J; T7X = T7m + T7l; T7n = T7l - T7m; T8a = FMA(KP414213562, T84, T85); T86 = FNMS(KP414213562, T85, T84); T91 = T8U - T8T; T8V = T8T + T8U; T7s = T7o + T7r; T7Y = T7o - T7r; { E T82, T81, T7z, T7E; T82 = T7x + T7y; T7z = T7x - T7y; T7E = T7A - T7D; T81 = T7A + T7D; T8W = T7n + T7s; T7t = T7n - T7s; T7T = FNMS(KP414213562, T7z, T7E); T7F = FMA(KP414213562, T7E, T7z); T92 = T7Y - T7X; T7Z = T7X + T7Y; T89 = FNMS(KP414213562, T81, T82); T83 = FMA(KP414213562, T82, T81); } } { E T7S, T7u, T93, T95, T7U, T7Q; T7S = FNMS(KP707106781, T7t, T7i); T7u = FMA(KP707106781, T7t, T7i); T93 = FMA(KP707106781, T92, T91); T95 = FNMS(KP707106781, T92, T91); T7U = FNMS(KP414213562, T7K, T7P); T7Q = FMA(KP414213562, T7P, T7K); { E T80, T87, T8X, T8Y; T88 = FNMS(KP707106781, T7Z, T7W); T80 = FMA(KP707106781, T7Z, T7W); { E T7V, T94, T96, T7R; T7V = T7T + T7U; T94 = T7U - T7T; T96 = T7Q - T7F; T7R = T7F + T7Q; Rm[WS(rs, 1)] = FMA(KP923879532, T7V, T7S); Rp[WS(rs, 14)] = FNMS(KP923879532, T7V, T7S); Ip[WS(rs, 6)] = FMA(KP923879532, T94, T93); Im[WS(rs, 9)] = FMS(KP923879532, T94, T93); Ip[WS(rs, 14)] = FMA(KP923879532, T96, T95); Im[WS(rs, 1)] = FMS(KP923879532, T96, T95); Rp[WS(rs, 6)] = FMA(KP923879532, T7R, T7u); Rm[WS(rs, 9)] = FNMS(KP923879532, T7R, T7u); T87 = T83 + T86; T90 = T86 - T83; } T8Z = FNMS(KP707106781, T8W, T8V); T8X = FMA(KP707106781, T8W, T8V); T8Y = T89 + T8a; T8b = T89 - T8a; Rp[WS(rs, 2)] = FMA(KP923879532, T87, T80); Rm[WS(rs, 13)] = FNMS(KP923879532, T87, T80); Ip[WS(rs, 2)] = FMA(KP923879532, T8Y, T8X); Im[WS(rs, 13)] = FMS(KP923879532, T8Y, T8X); } } } } { E T6s, T9o, T9n, T6v, T6Q, T6N, T6J, T6G, T9k, T9j; { E T6c, T4s, T9i, T4X, T9h, T9b, T9c, T6f, T5U, T6k, T64, T5k, T5v; { E T6d, T6e, T99, T9a, T5T; { E T4c, T4r, T4H, T4W; T6s = T46 - T4b; T4c = T46 + T4b; Rp[WS(rs, 10)] = FMA(KP923879532, T8b, T88); Rm[WS(rs, 5)] = FNMS(KP923879532, T8b, T88); Ip[WS(rs, 10)] = FMA(KP923879532, T90, T8Z); Im[WS(rs, 5)] = FMS(KP923879532, T90, T8Z); T4r = T4j + T4q; T9o = T4q - T4j; T6d = FNMS(KP414213562, T4z, T4G); T4H = FMA(KP414213562, T4G, T4z); T4W = FNMS(KP414213562, T4V, T4O); T6e = FMA(KP414213562, T4O, T4V); T9n = T98 + T97; T99 = T97 - T98; T6c = FNMS(KP707106781, T4r, T4c); T4s = FMA(KP707106781, T4r, T4c); T9i = T4W - T4H; T4X = T4H + T4W; T9a = T6t + T6u; T6v = T6t - T6u; } T6Q = T5S - T5L; T5T = T5L + T5S; T9h = FNMS(KP707106781, T9a, T99); T9b = FMA(KP707106781, T9a, T99); T9c = T6d + T6e; T6f = T6d - T6e; T5U = FMA(KP707106781, T5T, T5E); T6k = FNMS(KP707106781, T5T, T5E); T64 = T62 + T63; T6N = T63 - T62; T6J = T5c - T5j; T5k = T5c + T5j; T5v = T5t + T5u; T6G = T5u - T5t; } { E T6m, T6q, T6j, T6p, T9f, T9g; { E T68, T4Y, T6a, T66, T69, T5x, T9d, T6l, T65, T9e, T6b, T67; T68 = FNMS(KP923879532, T4X, T4s); T4Y = FMA(KP923879532, T4X, T4s); T6l = FNMS(KP707106781, T64, T61); T65 = FMA(KP707106781, T64, T61); { E T6h, T5l, T6i, T5w; T6h = FNMS(KP707106781, T5k, T55); T5l = FMA(KP707106781, T5k, T55); T6i = FNMS(KP707106781, T5v, T5s); T5w = FMA(KP707106781, T5v, T5s); T6m = FMA(KP668178637, T6l, T6k); T6q = FNMS(KP668178637, T6k, T6l); T6a = FMA(KP198912367, T5U, T65); T66 = FNMS(KP198912367, T65, T5U); T6j = FNMS(KP668178637, T6i, T6h); T6p = FMA(KP668178637, T6h, T6i); T69 = FNMS(KP198912367, T5l, T5w); T5x = FMA(KP198912367, T5w, T5l); } T9d = FMA(KP923879532, T9c, T9b); T9f = FNMS(KP923879532, T9c, T9b); T9e = T69 + T6a; T6b = T69 - T6a; T9g = T66 - T5x; T67 = T5x + T66; Ip[WS(rs, 1)] = FMA(KP980785280, T9e, T9d); Im[WS(rs, 14)] = FMS(KP980785280, T9e, T9d); Rp[WS(rs, 1)] = FMA(KP980785280, T67, T4Y); Rm[WS(rs, 14)] = FNMS(KP980785280, T67, T4Y); Rp[WS(rs, 9)] = FMA(KP980785280, T6b, T68); Rm[WS(rs, 6)] = FNMS(KP980785280, T6b, T68); } { E T6o, T9l, T9m, T6r, T6g, T6n; T6o = FMA(KP923879532, T6f, T6c); T6g = FNMS(KP923879532, T6f, T6c); T6n = T6j + T6m; T9k = T6m - T6j; T9j = FMA(KP923879532, T9i, T9h); T9l = FNMS(KP923879532, T9i, T9h); Ip[WS(rs, 9)] = FMA(KP980785280, T9g, T9f); Im[WS(rs, 6)] = FMS(KP980785280, T9g, T9f); Rm[WS(rs, 2)] = FMA(KP831469612, T6n, T6g); Rp[WS(rs, 13)] = FNMS(KP831469612, T6n, T6g); T9m = T6p + T6q; T6r = T6p - T6q; Ip[WS(rs, 13)] = FNMS(KP831469612, T9m, T9l); Im[WS(rs, 2)] = -(FMA(KP831469612, T9m, T9l)); Rp[WS(rs, 5)] = FMA(KP831469612, T6r, T6o); Rm[WS(rs, 10)] = FNMS(KP831469612, T6r, T6o); } } } { E T6Y, T6w, T9w, T6D, T9v, T9p, T9q, T71, T6H, T74, T78, T7c, T6W, T6S; { E T6Z, T6z, T6C, T70; T6Z = FNMS(KP414213562, T6x, T6y); T6z = FMA(KP414213562, T6y, T6x); Ip[WS(rs, 5)] = FMA(KP831469612, T9k, T9j); Im[WS(rs, 10)] = FMS(KP831469612, T9k, T9j); T6Y = FNMS(KP707106781, T6v, T6s); T6w = FMA(KP707106781, T6v, T6s); T6C = FNMS(KP414213562, T6B, T6A); T70 = FMA(KP414213562, T6A, T6B); T9w = T6z + T6C; T6D = T6z - T6C; T9v = FNMS(KP707106781, T9o, T9n); T9p = FMA(KP707106781, T9o, T9n); { E T77, T6O, T76, T6R; T9q = T70 - T6Z; T71 = T6Z + T70; T77 = FMA(KP707106781, T6N, T6M); T6O = FNMS(KP707106781, T6N, T6M); T76 = FMA(KP707106781, T6Q, T6P); T6R = FNMS(KP707106781, T6Q, T6P); T6H = FNMS(KP707106781, T6G, T6F); T74 = FMA(KP707106781, T6G, T6F); T78 = FMA(KP198912367, T77, T76); T7c = FNMS(KP198912367, T76, T77); T6W = FNMS(KP668178637, T6O, T6R); T6S = FMA(KP668178637, T6R, T6O); } } { E T6U, T6E, T9r, T9t, T73, T6K; T6U = FNMS(KP923879532, T6D, T6w); T6E = FMA(KP923879532, T6D, T6w); T9r = FMA(KP923879532, T9q, T9p); T9t = FNMS(KP923879532, T9q, T9p); T73 = FMA(KP707106781, T6J, T6I); T6K = FNMS(KP707106781, T6J, T6I); { E T7a, T9x, T9y, T7d; { E T72, T7b, T6V, T6L, T79, T75; T7a = FMA(KP923879532, T71, T6Y); T72 = FNMS(KP923879532, T71, T6Y); T75 = FMA(KP198912367, T74, T73); T7b = FNMS(KP198912367, T73, T74); T6V = FNMS(KP668178637, T6H, T6K); T6L = FMA(KP668178637, T6K, T6H); T79 = T75 + T78; T9A = T78 - T75; T9z = FMA(KP923879532, T9w, T9v); T9x = FNMS(KP923879532, T9w, T9v); { E T6X, T9s, T9u, T6T; T6X = T6V + T6W; T9s = T6V - T6W; T9u = T6S - T6L; T6T = T6L + T6S; Rp[WS(rs, 7)] = FMA(KP980785280, T79, T72); Rm[WS(rs, 8)] = FNMS(KP980785280, T79, T72); Rp[WS(rs, 11)] = FMA(KP831469612, T6X, T6U); Rm[WS(rs, 4)] = FNMS(KP831469612, T6X, T6U); Ip[WS(rs, 3)] = FMA(KP831469612, T9s, T9r); Im[WS(rs, 12)] = FMS(KP831469612, T9s, T9r); Ip[WS(rs, 11)] = FMA(KP831469612, T9u, T9t); Im[WS(rs, 4)] = FMS(KP831469612, T9u, T9t); Rp[WS(rs, 3)] = FMA(KP831469612, T6T, T6E); Rm[WS(rs, 12)] = FNMS(KP831469612, T6T, T6E); T9y = T7c - T7b; T7d = T7b + T7c; } } Ip[WS(rs, 7)] = FMA(KP980785280, T9y, T9x); Im[WS(rs, 8)] = FMS(KP980785280, T9y, T9x); Rm[0] = FMA(KP980785280, T7d, T7a); Rp[WS(rs, 15)] = FNMS(KP980785280, T7d, T7a); } } } } } } } Ip[WS(rs, 15)] = FMA(KP980785280, T9A, T9z); Im[0] = FMS(KP980785280, T9A, T9z); } } } static const tw_instr twinstr[] = { {TW_CEXP, 1, 1}, {TW_CEXP, 1, 3}, {TW_CEXP, 1, 9}, {TW_CEXP, 1, 27}, {TW_NEXT, 1, 0} }; static const hc2c_desc desc = { 32, "hc2cf2_32", twinstr, &GENUS, {236, 98, 252, 0} }; void X(codelet_hc2cf2_32) (planner *p) { X(khc2c_register) (p, hc2cf2_32, &desc, HC2C_VIA_RDFT); } #else /* HAVE_FMA */ /* Generated by: ../../../genfft/gen_hc2c.native -compact -variables 4 -pipeline-latency 4 -twiddle-log3 -precompute-twiddles -n 32 -dit -name hc2cf2_32 -include hc2cf.h */ /* * This function contains 488 FP additions, 280 FP multiplications, * (or, 376 additions, 168 multiplications, 112 fused multiply/add), * 158 stack variables, 7 constants, and 128 memory accesses */ #include "hc2cf.h" static void hc2cf2_32(R *Rp, R *Ip, R *Rm, R *Im, const R *W, stride rs, INT mb, INT me, INT ms) { DK(KP195090322, +0.195090322016128267848284868477022240927691618); DK(KP980785280, +0.980785280403230449126182236134239036973933731); DK(KP555570233, +0.555570233019602224742830813948532874374937191); DK(KP831469612, +0.831469612302545237078788377617905756738560812); DK(KP382683432, +0.382683432365089771728459984030398866761344562); DK(KP923879532, +0.923879532511286756128183189396788286822416626); DK(KP707106781, +0.707106781186547524400844362104849039284835938); { INT m; for (m = mb, W = W + ((mb - 1) * 8); m < me; m = m + 1, Rp = Rp + ms, Ip = Ip + ms, Rm = Rm - ms, Im = Im - ms, W = W + 8, MAKE_VOLATILE_STRIDE(128, rs)) { E T2, T5, T3, T6, T8, TM, TO, Td, T9, Te, Th, Tl, TD, TH, T1y; E T1H, T15, T1A, T11, T1F, T1n, T1p, T2q, T2I, T2u, T2K, T2V, T3b, T2Z, T3d; E Tu, Ty, T3l, T3n, T1t, T1v, T2f, T2h, T1a, T1e, T32, T34, T1W, T1Y, T2C; E T2E, Tg, TR, Tk, TS, Tm, TV, To, TT, T1M, T21, T1P, T22, T1Q, T25; E T1S, T23; { E Ts, T1d, Tx, T18, Tt, T1c, Tw, T19, TB, T14, TG, TZ, TC, T13, TF; E T10; { E T4, Tc, T7, Tb; T2 = W[0]; T5 = W[1]; T3 = W[2]; T6 = W[3]; T4 = T2 * T3; Tc = T5 * T3; T7 = T5 * T6; Tb = T2 * T6; T8 = T4 + T7; TM = T4 - T7; TO = Tb + Tc; Td = Tb - Tc; T9 = W[4]; Ts = T2 * T9; T1d = T6 * T9; Tx = T5 * T9; T18 = T3 * T9; Te = W[5]; Tt = T5 * Te; T1c = T3 * Te; Tw = T2 * Te; T19 = T6 * Te; Th = W[6]; TB = T3 * Th; T14 = T5 * Th; TG = T6 * Th; TZ = T2 * Th; Tl = W[7]; TC = T6 * Tl; T13 = T2 * Tl; TF = T3 * Tl; T10 = T5 * Tl; } TD = TB + TC; TH = TF - TG; T1y = TZ + T10; T1H = TF + TG; T15 = T13 + T14; T1A = T13 - T14; T11 = TZ - T10; T1F = TB - TC; T1n = FMA(T9, Th, Te * Tl); T1p = FNMS(Te, Th, T9 * Tl); { E T2o, T2p, T2s, T2t; T2o = T8 * Th; T2p = Td * Tl; T2q = T2o + T2p; T2I = T2o - T2p; T2s = T8 * Tl; T2t = Td * Th; T2u = T2s - T2t; T2K = T2s + T2t; } { E T2T, T2U, T2X, T2Y; T2T = TM * Th; T2U = TO * Tl; T2V = T2T - T2U; T3b = T2T + T2U; T2X = TM * Tl; T2Y = TO * Th; T2Z = T2X + T2Y; T3d = T2X - T2Y; Tu = Ts + Tt; Ty = Tw - Tx; T3l = FMA(Tu, Th, Ty * Tl); T3n = FNMS(Ty, Th, Tu * Tl); } T1t = Ts - Tt; T1v = Tw + Tx; T2f = FMA(T1t, Th, T1v * Tl); T2h = FNMS(T1v, Th, T1t * Tl); T1a = T18 - T19; T1e = T1c + T1d; T32 = FMA(T1a, Th, T1e * Tl); T34 = FNMS(T1e, Th, T1a * Tl); T1W = T18 + T19; T1Y = T1c - T1d; T2C = FMA(T1W, Th, T1Y * Tl); T2E = FNMS(T1Y, Th, T1W * Tl); { E Ta, Tf, Ti, Tj; Ta = T8 * T9; Tf = Td * Te; Tg = Ta - Tf; TR = Ta + Tf; Ti = T8 * Te; Tj = Td * T9; Tk = Ti + Tj; TS = Ti - Tj; } Tm = FMA(Tg, Th, Tk * Tl); TV = FNMS(TS, Th, TR * Tl); To = FNMS(Tk, Th, Tg * Tl); TT = FMA(TR, Th, TS * Tl); { E T1K, T1L, T1N, T1O; T1K = TM * T9; T1L = TO * Te; T1M = T1K - T1L; T21 = T1K + T1L; T1N = TM * Te; T1O = TO * T9; T1P = T1N + T1O; T22 = T1N - T1O; } T1Q = FMA(T1M, Th, T1P * Tl); T25 = FNMS(T22, Th, T21 * Tl); T1S = FNMS(T1P, Th, T1M * Tl); T23 = FMA(T21, Th, T22 * Tl); } { E TL, T6f, T8c, T8q, T3F, T5t, T7I, T7W, T2y, T6B, T6y, T7j, T4k, T5J, T4B; E T5G, T3h, T6H, T6O, T7o, T4L, T5N, T52, T5Q, T1i, T7V, T6i, T7D, T3K, T5u; E T3P, T5v, T1E, T6n, T6m, T7e, T3W, T5y, T41, T5z, T29, T6p, T6s, T7f, T47; E T5B, T4c, T5C, T2R, T6z, T6E, T7k, T4v, T5H, T4E, T5K, T3y, T6P, T6K, T7p; E T4W, T5R, T55, T5O; { E T1, T7G, Tq, T7F, TA, T3C, TJ, T3D, Tn, Tp; T1 = Rp[0]; T7G = Rm[0]; Tn = Rp[WS(rs, 8)]; Tp = Rm[WS(rs, 8)]; Tq = FMA(Tm, Tn, To * Tp); T7F = FNMS(To, Tn, Tm * Tp); { E Tv, Tz, TE, TI; Tv = Rp[WS(rs, 4)]; Tz = Rm[WS(rs, 4)]; TA = FMA(Tu, Tv, Ty * Tz); T3C = FNMS(Ty, Tv, Tu * Tz); TE = Rp[WS(rs, 12)]; TI = Rm[WS(rs, 12)]; TJ = FMA(TD, TE, TH * TI); T3D = FNMS(TH, TE, TD * TI); } { E Tr, TK, T8a, T8b; Tr = T1 + Tq; TK = TA + TJ; TL = Tr + TK; T6f = Tr - TK; T8a = T7G - T7F; T8b = TA - TJ; T8c = T8a - T8b; T8q = T8b + T8a; } { E T3B, T3E, T7E, T7H; T3B = T1 - Tq; T3E = T3C - T3D; T3F = T3B - T3E; T5t = T3B + T3E; T7E = T3C + T3D; T7H = T7F + T7G; T7I = T7E + T7H; T7W = T7H - T7E; } } { E T2e, T4g, T2w, T4z, T2j, T4h, T2n, T4y; { E T2c, T2d, T2r, T2v; T2c = Ip[0]; T2d = Im[0]; T2e = FMA(T2, T2c, T5 * T2d); T4g = FNMS(T5, T2c, T2 * T2d); T2r = Ip[WS(rs, 12)]; T2v = Im[WS(rs, 12)]; T2w = FMA(T2q, T2r, T2u * T2v); T4z = FNMS(T2u, T2r, T2q * T2v); } { E T2g, T2i, T2l, T2m; T2g = Ip[WS(rs, 8)]; T2i = Im[WS(rs, 8)]; T2j = FMA(T2f, T2g, T2h * T2i); T4h = FNMS(T2h, T2g, T2f * T2i); T2l = Ip[WS(rs, 4)]; T2m = Im[WS(rs, 4)]; T2n = FMA(T9, T2l, Te * T2m); T4y = FNMS(Te, T2l, T9 * T2m); } { E T2k, T2x, T6w, T6x; T2k = T2e + T2j; T2x = T2n + T2w; T2y = T2k + T2x; T6B = T2k - T2x; T6w = T4g + T4h; T6x = T4y + T4z; T6y = T6w - T6x; T7j = T6w + T6x; } { E T4i, T4j, T4x, T4A; T4i = T4g - T4h; T4j = T2n - T2w; T4k = T4i + T4j; T5J = T4i - T4j; T4x = T2e - T2j; T4A = T4y - T4z; T4B = T4x - T4A; T5G = T4x + T4A; } } { E T31, T4Y, T3f, T4J, T36, T4Z, T3a, T4I; { E T2W, T30, T3c, T3e; T2W = Ip[WS(rs, 15)]; T30 = Im[WS(rs, 15)]; T31 = FMA(T2V, T2W, T2Z * T30); T4Y = FNMS(T2Z, T2W, T2V * T30); T3c = Ip[WS(rs, 11)]; T3e = Im[WS(rs, 11)]; T3f = FMA(T3b, T3c, T3d * T3e); T4J = FNMS(T3d, T3c, T3b * T3e); } { E T33, T35, T38, T39; T33 = Ip[WS(rs, 7)]; T35 = Im[WS(rs, 7)]; T36 = FMA(T32, T33, T34 * T35); T4Z = FNMS(T34, T33, T32 * T35); T38 = Ip[WS(rs, 3)]; T39 = Im[WS(rs, 3)]; T3a = FMA(TR, T38, TS * T39); T4I = FNMS(TS, T38, TR * T39); } { E T37, T3g, T6M, T6N; T37 = T31 + T36; T3g = T3a + T3f; T3h = T37 + T3g; T6H = T37 - T3g; T6M = T4Y + T4Z; T6N = T4I + T4J; T6O = T6M - T6N; T7o = T6M + T6N; } { E T4H, T4K, T50, T51; T4H = T31 - T36; T4K = T4I - T4J; T4L = T4H - T4K; T5N = T4H + T4K; T50 = T4Y - T4Z; T51 = T3a - T3f; T52 = T50 + T51; T5Q = T50 - T51; } } { E TQ, T3G, T1g, T3N, TX, T3H, T17, T3M; { E TN, TP, T1b, T1f; TN = Rp[WS(rs, 2)]; TP = Rm[WS(rs, 2)]; TQ = FMA(TM, TN, TO * TP); T3G = FNMS(TO, TN, TM * TP); T1b = Rp[WS(rs, 6)]; T1f = Rm[WS(rs, 6)]; T1g = FMA(T1a, T1b, T1e * T1f); T3N = FNMS(T1e, T1b, T1a * T1f); } { E TU, TW, T12, T16; TU = Rp[WS(rs, 10)]; TW = Rm[WS(rs, 10)]; TX = FMA(TT, TU, TV * TW); T3H = FNMS(TV, TU, TT * TW); T12 = Rp[WS(rs, 14)]; T16 = Rm[WS(rs, 14)]; T17 = FMA(T11, T12, T15 * T16); T3M = FNMS(T15, T12, T11 * T16); } { E TY, T1h, T6g, T6h; TY = TQ + TX; T1h = T17 + T1g; T1i = TY + T1h; T7V = T1h - TY; T6g = T3G + T3H; T6h = T3M + T3N; T6i = T6g - T6h; T7D = T6g + T6h; } { E T3I, T3J, T3L, T3O; T3I = T3G - T3H; T3J = TQ - TX; T3K = T3I - T3J; T5u = T3J + T3I; T3L = T17 - T1g; T3O = T3M - T3N; T3P = T3L + T3O; T5v = T3L - T3O; } } { E T1m, T3S, T1C, T3Z, T1r, T3T, T1x, T3Y; { E T1k, T1l, T1z, T1B; T1k = Rp[WS(rs, 1)]; T1l = Rm[WS(rs, 1)]; T1m = FMA(T8, T1k, Td * T1l); T3S = FNMS(Td, T1k, T8 * T1l); T1z = Rp[WS(rs, 13)]; T1B = Rm[WS(rs, 13)]; T1C = FMA(T1y, T1z, T1A * T1B); T3Z = FNMS(T1A, T1z, T1y * T1B); } { E T1o, T1q, T1u, T1w; T1o = Rp[WS(rs, 9)]; T1q = Rm[WS(rs, 9)]; T1r = FMA(T1n, T1o, T1p * T1q); T3T = FNMS(T1p, T1o, T1n * T1q); T1u = Rp[WS(rs, 5)]; T1w = Rm[WS(rs, 5)]; T1x = FMA(T1t, T1u, T1v * T1w); T3Y = FNMS(T1v, T1u, T1t * T1w); } { E T1s, T1D, T6k, T6l; T1s = T1m + T1r; T1D = T1x + T1C; T1E = T1s + T1D; T6n = T1s - T1D; T6k = T3S + T3T; T6l = T3Y + T3Z; T6m = T6k - T6l; T7e = T6k + T6l; } { E T3U, T3V, T3X, T40; T3U = T3S - T3T; T3V = T1x - T1C; T3W = T3U + T3V; T5y = T3U - T3V; T3X = T1m - T1r; T40 = T3Y - T3Z; T41 = T3X - T40; T5z = T3X + T40; } } { E T1J, T43, T27, T4a, T1U, T44, T20, T49; { E T1G, T1I, T24, T26; T1G = Rp[WS(rs, 15)]; T1I = Rm[WS(rs, 15)]; T1J = FMA(T1F, T1G, T1H * T1I); T43 = FNMS(T1H, T1G, T1F * T1I); T24 = Rp[WS(rs, 11)]; T26 = Rm[WS(rs, 11)]; T27 = FMA(T23, T24, T25 * T26); T4a = FNMS(T25, T24, T23 * T26); } { E T1R, T1T, T1X, T1Z; T1R = Rp[WS(rs, 7)]; T1T = Rm[WS(rs, 7)]; T1U = FMA(T1Q, T1R, T1S * T1T); T44 = FNMS(T1S, T1R, T1Q * T1T); T1X = Rp[WS(rs, 3)]; T1Z = Rm[WS(rs, 3)]; T20 = FMA(T1W, T1X, T1Y * T1Z); T49 = FNMS(T1Y, T1X, T1W * T1Z); } { E T1V, T28, T6q, T6r; T1V = T1J + T1U; T28 = T20 + T27; T29 = T1V + T28; T6p = T1V - T28; T6q = T43 + T44; T6r = T49 + T4a; T6s = T6q - T6r; T7f = T6q + T6r; } { E T45, T46, T48, T4b; T45 = T43 - T44; T46 = T20 - T27; T47 = T45 + T46; T5B = T45 - T46; T48 = T1J - T1U; T4b = T49 - T4a; T4c = T48 - T4b; T5C = T48 + T4b; } } { E T2B, T4r, T2G, T4s, T4q, T4t, T2M, T4m, T2P, T4n, T4l, T4o; { E T2z, T2A, T2D, T2F; T2z = Ip[WS(rs, 2)]; T2A = Im[WS(rs, 2)]; T2B = FMA(T21, T2z, T22 * T2A); T4r = FNMS(T22, T2z, T21 * T2A); T2D = Ip[WS(rs, 10)]; T2F = Im[WS(rs, 10)]; T2G = FMA(T2C, T2D, T2E * T2F); T4s = FNMS(T2E, T2D, T2C * T2F); } T4q = T2B - T2G; T4t = T4r - T4s; { E T2J, T2L, T2N, T2O; T2J = Ip[WS(rs, 14)]; T2L = Im[WS(rs, 14)]; T2M = FMA(T2I, T2J, T2K * T2L); T4m = FNMS(T2K, T2J, T2I * T2L); T2N = Ip[WS(rs, 6)]; T2O = Im[WS(rs, 6)]; T2P = FMA(T1M, T2N, T1P * T2O); T4n = FNMS(T1P, T2N, T1M * T2O); } T4l = T2M - T2P; T4o = T4m - T4n; { E T2H, T2Q, T6C, T6D; T2H = T2B + T2G; T2Q = T2M + T2P; T2R = T2H + T2Q; T6z = T2Q - T2H; T6C = T4r + T4s; T6D = T4m + T4n; T6E = T6C - T6D; T7k = T6C + T6D; } { E T4p, T4u, T4C, T4D; T4p = T4l - T4o; T4u = T4q + T4t; T4v = KP707106781 * (T4p - T4u); T5H = KP707106781 * (T4u + T4p); T4C = T4t - T4q; T4D = T4l + T4o; T4E = KP707106781 * (T4C - T4D); T5K = KP707106781 * (T4C + T4D); } } { E T3k, T4M, T3p, T4N, T4O, T4P, T3t, T4S, T3w, T4T, T4R, T4U; { E T3i, T3j, T3m, T3o; T3i = Ip[WS(rs, 1)]; T3j = Im[WS(rs, 1)]; T3k = FMA(T3, T3i, T6 * T3j); T4M = FNMS(T6, T3i, T3 * T3j); T3m = Ip[WS(rs, 9)]; T3o = Im[WS(rs, 9)]; T3p = FMA(T3l, T3m, T3n * T3o); T4N = FNMS(T3n, T3m, T3l * T3o); } T4O = T4M - T4N; T4P = T3k - T3p; { E T3r, T3s, T3u, T3v; T3r = Ip[WS(rs, 13)]; T3s = Im[WS(rs, 13)]; T3t = FMA(Th, T3r, Tl * T3s); T4S = FNMS(Tl, T3r, Th * T3s); T3u = Ip[WS(rs, 5)]; T3v = Im[WS(rs, 5)]; T3w = FMA(Tg, T3u, Tk * T3v); T4T = FNMS(Tk, T3u, Tg * T3v); } T4R = T3t - T3w; T4U = T4S - T4T; { E T3q, T3x, T6I, T6J; T3q = T3k + T3p; T3x = T3t + T3w; T3y = T3q + T3x; T6P = T3x - T3q; T6I = T4M + T4N; T6J = T4S + T4T; T6K = T6I - T6J; T7p = T6I + T6J; } { E T4Q, T4V, T53, T54; T4Q = T4O - T4P; T4V = T4R + T4U; T4W = KP707106781 * (T4Q - T4V); T5R = KP707106781 * (T4Q + T4V); T53 = T4R - T4U; T54 = T4P + T4O; T55 = KP707106781 * (T53 - T54); T5O = KP707106781 * (T54 + T53); } } { E T2b, T7x, T7K, T7M, T3A, T7L, T7A, T7B; { E T1j, T2a, T7C, T7J; T1j = TL + T1i; T2a = T1E + T29; T2b = T1j + T2a; T7x = T1j - T2a; T7C = T7e + T7f; T7J = T7D + T7I; T7K = T7C + T7J; T7M = T7J - T7C; } { E T2S, T3z, T7y, T7z; T2S = T2y + T2R; T3z = T3h + T3y; T3A = T2S + T3z; T7L = T3z - T2S; T7y = T7j + T7k; T7z = T7o + T7p; T7A = T7y - T7z; T7B = T7y + T7z; } Rm[WS(rs, 15)] = T2b - T3A; Im[WS(rs, 15)] = T7B - T7K; Rp[0] = T2b + T3A; Ip[0] = T7B + T7K; Rm[WS(rs, 7)] = T7x - T7A; Im[WS(rs, 7)] = T7L - T7M; Rp[WS(rs, 8)] = T7x + T7A; Ip[WS(rs, 8)] = T7L + T7M; } { E T7h, T7t, T7Q, T7S, T7m, T7u, T7r, T7v; { E T7d, T7g, T7O, T7P; T7d = TL - T1i; T7g = T7e - T7f; T7h = T7d + T7g; T7t = T7d - T7g; T7O = T29 - T1E; T7P = T7I - T7D; T7Q = T7O + T7P; T7S = T7P - T7O; } { E T7i, T7l, T7n, T7q; T7i = T2y - T2R; T7l = T7j - T7k; T7m = T7i + T7l; T7u = T7l - T7i; T7n = T3h - T3y; T7q = T7o - T7p; T7r = T7n - T7q; T7v = T7n + T7q; } { E T7s, T7N, T7w, T7R; T7s = KP707106781 * (T7m + T7r); Rm[WS(rs, 11)] = T7h - T7s; Rp[WS(rs, 4)] = T7h + T7s; T7N = KP707106781 * (T7u + T7v); Im[WS(rs, 11)] = T7N - T7Q; Ip[WS(rs, 4)] = T7N + T7Q; T7w = KP707106781 * (T7u - T7v); Rm[WS(rs, 3)] = T7t - T7w; Rp[WS(rs, 12)] = T7t + T7w; T7R = KP707106781 * (T7r - T7m); Im[WS(rs, 3)] = T7R - T7S; Ip[WS(rs, 12)] = T7R + T7S; } } { E T6j, T7X, T83, T6X, T6u, T7U, T77, T7b, T70, T82, T6G, T6U, T74, T7a, T6R; E T6V; { E T6o, T6t, T6A, T6F; T6j = T6f - T6i; T7X = T7V + T7W; T83 = T7W - T7V; T6X = T6f + T6i; T6o = T6m - T6n; T6t = T6p + T6s; T6u = KP707106781 * (T6o - T6t); T7U = KP707106781 * (T6o + T6t); { E T75, T76, T6Y, T6Z; T75 = T6H + T6K; T76 = T6O + T6P; T77 = FNMS(KP382683432, T76, KP923879532 * T75); T7b = FMA(KP923879532, T76, KP382683432 * T75); T6Y = T6n + T6m; T6Z = T6p - T6s; T70 = KP707106781 * (T6Y + T6Z); T82 = KP707106781 * (T6Z - T6Y); } T6A = T6y - T6z; T6F = T6B - T6E; T6G = FMA(KP923879532, T6A, KP382683432 * T6F); T6U = FNMS(KP923879532, T6F, KP382683432 * T6A); { E T72, T73, T6L, T6Q; T72 = T6y + T6z; T73 = T6B + T6E; T74 = FMA(KP382683432, T72, KP923879532 * T73); T7a = FNMS(KP382683432, T73, KP923879532 * T72); T6L = T6H - T6K; T6Q = T6O - T6P; T6R = FNMS(KP923879532, T6Q, KP382683432 * T6L); T6V = FMA(KP382683432, T6Q, KP923879532 * T6L); } } { E T6v, T6S, T81, T84; T6v = T6j + T6u; T6S = T6G + T6R; Rm[WS(rs, 9)] = T6v - T6S; Rp[WS(rs, 6)] = T6v + T6S; T81 = T6U + T6V; T84 = T82 + T83; Im[WS(rs, 9)] = T81 - T84; Ip[WS(rs, 6)] = T81 + T84; } { E T6T, T6W, T85, T86; T6T = T6j - T6u; T6W = T6U - T6V; Rm[WS(rs, 1)] = T6T - T6W; Rp[WS(rs, 14)] = T6T + T6W; T85 = T6R - T6G; T86 = T83 - T82; Im[WS(rs, 1)] = T85 - T86; Ip[WS(rs, 14)] = T85 + T86; } { E T71, T78, T7T, T7Y; T71 = T6X + T70; T78 = T74 + T77; Rm[WS(rs, 13)] = T71 - T78; Rp[WS(rs, 2)] = T71 + T78; T7T = T7a + T7b; T7Y = T7U + T7X; Im[WS(rs, 13)] = T7T - T7Y; Ip[WS(rs, 2)] = T7T + T7Y; } { E T79, T7c, T7Z, T80; T79 = T6X - T70; T7c = T7a - T7b; Rm[WS(rs, 5)] = T79 - T7c; Rp[WS(rs, 10)] = T79 + T7c; T7Z = T77 - T74; T80 = T7X - T7U; Im[WS(rs, 5)] = T7Z - T80; Ip[WS(rs, 10)] = T7Z + T80; } } { E T3R, T5d, T8r, T8x, T4e, T8o, T5n, T5r, T4G, T5a, T5g, T8w, T5k, T5q, T57; E T5b, T3Q, T8p; T3Q = KP707106781 * (T3K - T3P); T3R = T3F - T3Q; T5d = T3F + T3Q; T8p = KP707106781 * (T5v - T5u); T8r = T8p + T8q; T8x = T8q - T8p; { E T42, T4d, T5l, T5m; T42 = FNMS(KP923879532, T41, KP382683432 * T3W); T4d = FMA(KP382683432, T47, KP923879532 * T4c); T4e = T42 - T4d; T8o = T42 + T4d; T5l = T4L + T4W; T5m = T52 + T55; T5n = FNMS(KP555570233, T5m, KP831469612 * T5l); T5r = FMA(KP831469612, T5m, KP555570233 * T5l); } { E T4w, T4F, T5e, T5f; T4w = T4k - T4v; T4F = T4B - T4E; T4G = FMA(KP980785280, T4w, KP195090322 * T4F); T5a = FNMS(KP980785280, T4F, KP195090322 * T4w); T5e = FMA(KP923879532, T3W, KP382683432 * T41); T5f = FNMS(KP923879532, T47, KP382683432 * T4c); T5g = T5e + T5f; T8w = T5f - T5e; } { E T5i, T5j, T4X, T56; T5i = T4k + T4v; T5j = T4B + T4E; T5k = FMA(KP555570233, T5i, KP831469612 * T5j); T5q = FNMS(KP555570233, T5j, KP831469612 * T5i); T4X = T4L - T4W; T56 = T52 - T55; T57 = FNMS(KP980785280, T56, KP195090322 * T4X); T5b = FMA(KP195090322, T56, KP980785280 * T4X); } { E T4f, T58, T8v, T8y; T4f = T3R + T4e; T58 = T4G + T57; Rm[WS(rs, 8)] = T4f - T58; Rp[WS(rs, 7)] = T4f + T58; T8v = T5a + T5b; T8y = T8w + T8x; Im[WS(rs, 8)] = T8v - T8y; Ip[WS(rs, 7)] = T8v + T8y; } { E T59, T5c, T8z, T8A; T59 = T3R - T4e; T5c = T5a - T5b; Rm[0] = T59 - T5c; Rp[WS(rs, 15)] = T59 + T5c; T8z = T57 - T4G; T8A = T8x - T8w; Im[0] = T8z - T8A; Ip[WS(rs, 15)] = T8z + T8A; } { E T5h, T5o, T8n, T8s; T5h = T5d + T5g; T5o = T5k + T5n; Rm[WS(rs, 12)] = T5h - T5o; Rp[WS(rs, 3)] = T5h + T5o; T8n = T5q + T5r; T8s = T8o + T8r; Im[WS(rs, 12)] = T8n - T8s; Ip[WS(rs, 3)] = T8n + T8s; } { E T5p, T5s, T8t, T8u; T5p = T5d - T5g; T5s = T5q - T5r; Rm[WS(rs, 4)] = T5p - T5s; Rp[WS(rs, 11)] = T5p + T5s; T8t = T5n - T5k; T8u = T8r - T8o; Im[WS(rs, 4)] = T8t - T8u; Ip[WS(rs, 11)] = T8t + T8u; } } { E T5x, T5Z, T8d, T8j, T5E, T88, T69, T6d, T5M, T5W, T62, T8i, T66, T6c, T5T; E T5X, T5w, T89; T5w = KP707106781 * (T5u + T5v); T5x = T5t - T5w; T5Z = T5t + T5w; T89 = KP707106781 * (T3K + T3P); T8d = T89 + T8c; T8j = T8c - T89; { E T5A, T5D, T67, T68; T5A = FNMS(KP382683432, T5z, KP923879532 * T5y); T5D = FMA(KP923879532, T5B, KP382683432 * T5C); T5E = T5A - T5D; T88 = T5A + T5D; T67 = T5N + T5O; T68 = T5Q + T5R; T69 = FNMS(KP195090322, T68, KP980785280 * T67); T6d = FMA(KP195090322, T67, KP980785280 * T68); } { E T5I, T5L, T60, T61; T5I = T5G - T5H; T5L = T5J - T5K; T5M = FMA(KP555570233, T5I, KP831469612 * T5L); T5W = FNMS(KP831469612, T5I, KP555570233 * T5L); T60 = FMA(KP382683432, T5y, KP923879532 * T5z); T61 = FNMS(KP382683432, T5B, KP923879532 * T5C); T62 = T60 + T61; T8i = T61 - T60; } { E T64, T65, T5P, T5S; T64 = T5G + T5H; T65 = T5J + T5K; T66 = FMA(KP980785280, T64, KP195090322 * T65); T6c = FNMS(KP195090322, T64, KP980785280 * T65); T5P = T5N - T5O; T5S = T5Q - T5R; T5T = FNMS(KP831469612, T5S, KP555570233 * T5P); T5X = FMA(KP831469612, T5P, KP555570233 * T5S); } { E T5F, T5U, T8h, T8k; T5F = T5x + T5E; T5U = T5M + T5T; Rm[WS(rs, 10)] = T5F - T5U; Rp[WS(rs, 5)] = T5F + T5U; T8h = T5W + T5X; T8k = T8i + T8j; Im[WS(rs, 10)] = T8h - T8k; Ip[WS(rs, 5)] = T8h + T8k; } { E T5V, T5Y, T8l, T8m; T5V = T5x - T5E; T5Y = T5W - T5X; Rm[WS(rs, 2)] = T5V - T5Y; Rp[WS(rs, 13)] = T5V + T5Y; T8l = T5T - T5M; T8m = T8j - T8i; Im[WS(rs, 2)] = T8l - T8m; Ip[WS(rs, 13)] = T8l + T8m; } { E T63, T6a, T87, T8e; T63 = T5Z + T62; T6a = T66 + T69; Rm[WS(rs, 14)] = T63 - T6a; Rp[WS(rs, 1)] = T63 + T6a; T87 = T6c + T6d; T8e = T88 + T8d; Im[WS(rs, 14)] = T87 - T8e; Ip[WS(rs, 1)] = T87 + T8e; } { E T6b, T6e, T8f, T8g; T6b = T5Z - T62; T6e = T6c - T6d; Rm[WS(rs, 6)] = T6b - T6e; Rp[WS(rs, 9)] = T6b + T6e; T8f = T69 - T66; T8g = T8d - T88; Im[WS(rs, 6)] = T8f - T8g; Ip[WS(rs, 9)] = T8f + T8g; } } } } } } static const tw_instr twinstr[] = { {TW_CEXP, 1, 1}, {TW_CEXP, 1, 3}, {TW_CEXP, 1, 9}, {TW_CEXP, 1, 27}, {TW_NEXT, 1, 0} }; static const hc2c_desc desc = { 32, "hc2cf2_32", twinstr, &GENUS, {376, 168, 112, 0} }; void X(codelet_hc2cf2_32) (planner *p) { X(khc2c_register) (p, hc2cf2_32, &desc, HC2C_VIA_RDFT); } #endif /* HAVE_FMA */ fftw-3.3.3/rdft/scalar/r2cf/hc2cf2_20.c0000644000175000017500000007331412054410716014173 00000000000000/* * Copyright (c) 2003, 2007-11 Matteo Frigo * Copyright (c) 2003, 2007-11 Massachusetts Institute of Technology * * This program is free software; you can redistribute it and/or modify * it under the terms of the GNU General Public License as published by * the Free Software Foundation; either version 2 of the License, or * (at your option) any later version. * * This program is distributed in the hope that it will be useful, * but WITHOUT ANY WARRANTY; without even the implied warranty of * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the * GNU General Public License for more details. * * You should have received a copy of the GNU General Public License * along with this program; if not, write to the Free Software * Foundation, Inc., 51 Franklin Street, Fifth Floor, Boston, MA 02110-1301 USA * */ /* This file was automatically generated --- DO NOT EDIT */ /* Generated on Sun Nov 25 07:40:44 EST 2012 */ #include "codelet-rdft.h" #ifdef HAVE_FMA /* Generated by: ../../../genfft/gen_hc2c.native -fma -reorder-insns -schedule-for-pipeline -compact -variables 4 -pipeline-latency 4 -twiddle-log3 -precompute-twiddles -n 20 -dit -name hc2cf2_20 -include hc2cf.h */ /* * This function contains 276 FP additions, 198 FP multiplications, * (or, 136 additions, 58 multiplications, 140 fused multiply/add), * 142 stack variables, 4 constants, and 80 memory accesses */ #include "hc2cf.h" static void hc2cf2_20(R *Rp, R *Ip, R *Rm, R *Im, const R *W, stride rs, INT mb, INT me, INT ms) { DK(KP951056516, +0.951056516295153572116439333379382143405698634); DK(KP559016994, +0.559016994374947424102293417182819058860154590); DK(KP250000000, +0.250000000000000000000000000000000000000000000); DK(KP618033988, +0.618033988749894848204586834365638117720309180); { INT m; for (m = mb, W = W + ((mb - 1) * 8); m < me; m = m + 1, Rp = Rp + ms, Ip = Ip + ms, Rm = Rm - ms, Im = Im - ms, W = W + 8, MAKE_VOLATILE_STRIDE(80, rs)) { E T59, T5i, T5k, T5e, T5c, T5d, T5j, T5f; { E T2, Th, Tf, T6, T5, Tl, T1p, T1n, Ti, T3, Tt, Tv, T24, T1f, T1D; E Tb, T1P, Tm, T21, T1b, T7, T1A, Tw, T1H, T13, TA, T1L, T17, T1S, Tq; E T1o, T2g, T1t, T2c, TO, TK; { E T1e, Ta, Tk, Tg; T2 = W[0]; Th = W[3]; Tf = W[2]; T6 = W[5]; T5 = W[1]; Tk = T2 * Th; Tg = T2 * Tf; T1e = Tf * T6; Ta = T2 * T6; Tl = FMA(T5, Tf, Tk); T1p = FNMS(T5, Tf, Tk); T1n = FMA(T5, Th, Tg); Ti = FNMS(T5, Th, Tg); T3 = W[4]; Tt = W[6]; Tv = W[7]; { E Tp, Tj, TN, TJ; Tp = Ti * T6; T24 = FMA(Th, T3, T1e); T1f = FNMS(Th, T3, T1e); T1D = FNMS(T5, T3, Ta); Tb = FMA(T5, T3, Ta); Tj = Ti * T3; { E T1a, T4, Tu, T1G; T1a = Tf * T3; T4 = T2 * T3; Tu = Ti * Tt; T1G = T2 * Tt; { E T12, Tz, T1K, T16; T12 = Tf * Tt; Tz = Ti * Tv; T1K = T2 * Tv; T16 = Tf * Tv; T1P = FNMS(Tl, T6, Tj); Tm = FMA(Tl, T6, Tj); T21 = FNMS(Th, T6, T1a); T1b = FMA(Th, T6, T1a); T7 = FNMS(T5, T6, T4); T1A = FMA(T5, T6, T4); Tw = FMA(Tl, Tv, Tu); T1H = FMA(T5, Tv, T1G); T13 = FMA(Th, Tv, T12); TA = FNMS(Tl, Tt, Tz); T1L = FNMS(T5, Tt, T1K); T17 = FNMS(Th, Tt, T16); T1S = FMA(Tl, T3, Tp); Tq = FNMS(Tl, T3, Tp); } } T1o = T1n * T3; T2g = T1n * Tv; TN = Tm * Tv; TJ = Tm * Tt; T1t = T1n * T6; T2c = T1n * Tt; TO = FNMS(Tq, Tt, TN); TK = FMA(Tq, Tv, TJ); } } { E Te, T2C, T4L, T57, T58, TD, T2H, T4H, T3J, T3Z, T11, T2v, T2P, T3P, T4d; E T4z, T3n, T43, T2r, T2z, T3b, T3T, T4n, T4v, T3u, T42, T20, T2y, T34, T3S; E T4k, T4w, T1c, T19, T1d, T3y, T1w, T2U, T1g, T1j, T1l; { E T2d, T2h, T2k, T1q, T1u, T2n, TL, TI, TM, T3F, TZ, T2N, TP, TS, TU; { E T1, T4K, T8, T9, Tc; T1 = Rp[0]; T4K = Rm[0]; T8 = Rp[WS(rs, 5)]; T2d = FMA(T1p, Tv, T2c); T2h = FNMS(T1p, Tt, T2g); T2k = FMA(T1p, T6, T1o); T1q = FNMS(T1p, T6, T1o); T1u = FMA(T1p, T3, T1t); T2n = FNMS(T1p, T3, T1t); T9 = T7 * T8; Tc = Rm[WS(rs, 5)]; { E Tx, Ts, T2F, TC, T2E; { E Tn, Tr, To, T2D, T4J, Ty, TB, Td, T4I; Tn = Ip[WS(rs, 2)]; Tr = Im[WS(rs, 2)]; Tx = Ip[WS(rs, 7)]; Td = FMA(Tb, Tc, T9); T4I = T7 * Tc; To = Tm * Tn; T2D = Tm * Tr; Te = T1 + Td; T2C = T1 - Td; T4J = FNMS(Tb, T8, T4I); Ty = Tw * Tx; TB = Im[WS(rs, 7)]; Ts = FMA(Tq, Tr, To); T4L = T4J + T4K; T57 = T4K - T4J; T2F = Tw * TB; TC = FMA(TA, TB, Ty); T2E = FNMS(Tq, Tn, T2D); } { E TF, TG, TH, TW, TY, T2G, T3E, TX, T2M; TF = Rp[WS(rs, 2)]; T2G = FNMS(TA, Tx, T2F); T58 = Ts - TC; TD = Ts + TC; TG = Ti * TF; T2H = T2E - T2G; T4H = T2E + T2G; TH = Rm[WS(rs, 2)]; TW = Ip[WS(rs, 9)]; TY = Im[WS(rs, 9)]; TL = Rp[WS(rs, 7)]; TI = FMA(Tl, TH, TG); T3E = Ti * TH; TX = Tt * TW; T2M = Tt * TY; TM = TK * TL; T3F = FNMS(Tl, TF, T3E); TZ = FMA(Tv, TY, TX); T2N = FNMS(Tv, TW, T2M); TP = Rm[WS(rs, 7)]; TS = Ip[WS(rs, 4)]; TU = Im[WS(rs, 4)]; } } } { E T27, T26, T28, T3j, T2p, T39, T29, T2e, T2i; { E T22, T23, T25, T2l, T2o, T3i, T2m, T38; { E TR, T2J, T3H, TV, T2L, T4b, T3I; T22 = Rp[WS(rs, 6)]; { E TQ, T3G, TT, T2K; TQ = FMA(TO, TP, TM); T3G = TK * TP; TT = T3 * TS; T2K = T3 * TU; TR = TI + TQ; T2J = TI - TQ; T3H = FNMS(TO, TL, T3G); TV = FMA(T6, TU, TT); T2L = FNMS(T6, TS, T2K); T23 = T21 * T22; } T4b = T3F + T3H; T3I = T3F - T3H; { E T10, T3D, T4c, T2O; T10 = TV + TZ; T3D = TZ - TV; T4c = T2L + T2N; T2O = T2L - T2N; T3J = T3D - T3I; T3Z = T3I + T3D; T11 = TR - T10; T2v = TR + T10; T2P = T2J - T2O; T3P = T2J + T2O; T4d = T4b + T4c; T4z = T4c - T4b; T25 = Rm[WS(rs, 6)]; } } T2l = Ip[WS(rs, 3)]; T2o = Im[WS(rs, 3)]; T27 = Rp[WS(rs, 1)]; T26 = FMA(T24, T25, T23); T3i = T21 * T25; T2m = T2k * T2l; T38 = T2k * T2o; T28 = T1n * T27; T3j = FNMS(T24, T22, T3i); T2p = FMA(T2n, T2o, T2m); T39 = FNMS(T2n, T2l, T38); T29 = Rm[WS(rs, 1)]; T2e = Ip[WS(rs, 8)]; T2i = Im[WS(rs, 8)]; } { E T1I, T1F, T1J, T3q, T1Y, T32, T1M, T1Q, T1T; { E T1B, T1C, T1E, T1V, T1X, T3p, T1W, T31; { E T2b, T35, T3l, T2j, T37, T4l, T3m; T1B = Rp[WS(rs, 4)]; { E T2a, T3k, T2f, T36; T2a = FMA(T1p, T29, T28); T3k = T1n * T29; T2f = T2d * T2e; T36 = T2d * T2i; T2b = T26 + T2a; T35 = T26 - T2a; T3l = FNMS(T1p, T27, T3k); T2j = FMA(T2h, T2i, T2f); T37 = FNMS(T2h, T2e, T36); T1C = T1A * T1B; } T4l = T3j + T3l; T3m = T3j - T3l; { E T2q, T3h, T4m, T3a; T2q = T2j + T2p; T3h = T2p - T2j; T4m = T37 + T39; T3a = T37 - T39; T3n = T3h - T3m; T43 = T3m + T3h; T2r = T2b - T2q; T2z = T2b + T2q; T3b = T35 - T3a; T3T = T35 + T3a; T4n = T4l + T4m; T4v = T4m - T4l; T1E = Rm[WS(rs, 4)]; } } T1V = Ip[WS(rs, 1)]; T1X = Im[WS(rs, 1)]; T1I = Rp[WS(rs, 9)]; T1F = FMA(T1D, T1E, T1C); T3p = T1A * T1E; T1W = Tf * T1V; T31 = Tf * T1X; T1J = T1H * T1I; T3q = FNMS(T1D, T1B, T3p); T1Y = FMA(Th, T1X, T1W); T32 = FNMS(Th, T1V, T31); T1M = Rm[WS(rs, 9)]; T1Q = Ip[WS(rs, 6)]; T1T = Im[WS(rs, 6)]; } { E T14, T15, T18, T1r, T1v, T3x, T1s, T2T; { E T1O, T2Y, T3s, T1U, T30, T4i, T3t; T14 = Rp[WS(rs, 8)]; { E T1N, T3r, T1R, T2Z; T1N = FMA(T1L, T1M, T1J); T3r = T1H * T1M; T1R = T1P * T1Q; T2Z = T1P * T1T; T1O = T1F + T1N; T2Y = T1F - T1N; T3s = FNMS(T1L, T1I, T3r); T1U = FMA(T1S, T1T, T1R); T30 = FNMS(T1S, T1Q, T2Z); T15 = T13 * T14; } T4i = T3q + T3s; T3t = T3q - T3s; { E T1Z, T3o, T4j, T33; T1Z = T1U + T1Y; T3o = T1Y - T1U; T4j = T30 + T32; T33 = T30 - T32; T3u = T3o - T3t; T42 = T3t + T3o; T20 = T1O - T1Z; T2y = T1O + T1Z; T34 = T2Y - T33; T3S = T2Y + T33; T4k = T4i + T4j; T4w = T4j - T4i; T18 = Rm[WS(rs, 8)]; } } T1r = Ip[WS(rs, 5)]; T1v = Im[WS(rs, 5)]; T1c = Rp[WS(rs, 3)]; T19 = FMA(T17, T18, T15); T3x = T13 * T18; T1s = T1q * T1r; T2T = T1q * T1v; T1d = T1b * T1c; T3y = FNMS(T17, T14, T3x); T1w = FMA(T1u, T1v, T1s); T2U = FNMS(T1u, T1r, T2T); T1g = Rm[WS(rs, 3)]; T1j = Ip[0]; T1l = Im[0]; } } } } { E T3C, T40, T2W, T3Q, T4M, T4E, T4F, T4U, T4S; { E T4W, T2u, T2w, T4g, T4V, T4D, T4B, T54, T56, T4Y, T4u, T4C; { E T4x, TE, T53, T1z, T2s, T52, T4A, T4t, T4s, T2t; { E T1i, T2Q, T3A, T1m, T2S; T4x = T4v - T4w; T4W = T4w + T4v; { E T1h, T3z, T1k, T2R; T1h = FMA(T1f, T1g, T1d); T3z = T1b * T1g; T1k = T2 * T1j; T2R = T2 * T1l; T1i = T19 + T1h; T2Q = T19 - T1h; T3A = FNMS(T1f, T1c, T3z); T1m = FMA(T5, T1l, T1k); T2S = FNMS(T5, T1j, T2R); } TE = Te - TD; T2u = Te + TD; { E T4e, T3B, T1x, T3w; T4e = T3y + T3A; T3B = T3y - T3A; T1x = T1m + T1w; T3w = T1w - T1m; { E T4f, T2V, T1y, T4y; T4f = T2S + T2U; T2V = T2S - T2U; T3C = T3w - T3B; T40 = T3B + T3w; T1y = T1i - T1x; T2w = T1i + T1x; T2W = T2Q - T2V; T3Q = T2Q + T2V; T4g = T4e + T4f; T4y = T4f - T4e; T53 = T1y - T11; T1z = T11 + T1y; T2s = T20 + T2r; T52 = T20 - T2r; T4V = T4z + T4y; T4A = T4y - T4z; } } } T4t = T1z - T2s; T2t = T1z + T2s; T4D = FMA(KP618033988, T4x, T4A); T4B = FNMS(KP618033988, T4A, T4x); T54 = FMA(KP618033988, T53, T52); T56 = FNMS(KP618033988, T52, T53); Rm[WS(rs, 9)] = TE + T2t; T4s = FNMS(KP250000000, T2t, TE); T4Y = T4L - T4H; T4M = T4H + T4L; T4u = FNMS(KP559016994, T4t, T4s); T4C = FMA(KP559016994, T4t, T4s); } { E T2x, T4Q, T4p, T4r, T4R, T2A, T51, T55; { E T4h, T50, T4X, T4o, T4Z; T4E = T4d + T4g; T4h = T4d - T4g; Rm[WS(rs, 1)] = FMA(KP951056516, T4B, T4u); Rp[WS(rs, 2)] = FNMS(KP951056516, T4B, T4u); Rp[WS(rs, 6)] = FMA(KP951056516, T4D, T4C); Rm[WS(rs, 5)] = FNMS(KP951056516, T4D, T4C); T50 = T4W - T4V; T4X = T4V + T4W; T4o = T4k - T4n; T4F = T4k + T4n; T2x = T2v + T2w; T4Q = T2v - T2w; Im[WS(rs, 9)] = T4X - T4Y; T4Z = FMA(KP250000000, T4X, T4Y); T4p = FMA(KP618033988, T4o, T4h); T4r = FNMS(KP618033988, T4h, T4o); T4R = T2z - T2y; T2A = T2y + T2z; T51 = FNMS(KP559016994, T50, T4Z); T55 = FMA(KP559016994, T50, T4Z); } { E T49, T48, T2B, T4a, T4q; T2B = T2x + T2A; T49 = T2x - T2A; Ip[WS(rs, 2)] = FMA(KP951056516, T54, T51); Im[WS(rs, 1)] = FMS(KP951056516, T54, T51); Ip[WS(rs, 6)] = FMA(KP951056516, T56, T55); Im[WS(rs, 5)] = FMS(KP951056516, T56, T55); Rp[0] = T2u + T2B; T48 = FNMS(KP250000000, T2B, T2u); T4a = FMA(KP559016994, T49, T48); T4q = FNMS(KP559016994, T49, T48); T4U = FMA(KP618033988, T4Q, T4R); T4S = FNMS(KP618033988, T4R, T4Q); Rm[WS(rs, 3)] = FMA(KP951056516, T4p, T4a); Rp[WS(rs, 4)] = FNMS(KP951056516, T4p, T4a); Rp[WS(rs, 8)] = FMA(KP951056516, T4r, T4q); Rm[WS(rs, 7)] = FNMS(KP951056516, T4r, T4q); } } } { E T3O, T5u, T5w, T5o, T5q, T5n; { E T5m, T5l, T2I, T4O, T3N, T3L, T2X, T5s, T4N, T5t, T3c, T3v, T3K, T4G; T5m = T3u + T3n; T3v = T3n - T3u; T3K = T3C - T3J; T5l = T3J + T3C; T3O = T2C + T2H; T2I = T2C - T2H; T4O = T4E - T4F; T4G = T4E + T4F; T3N = FMA(KP618033988, T3v, T3K); T3L = FNMS(KP618033988, T3K, T3v); T2X = T2P + T2W; T5s = T2P - T2W; Ip[0] = T4G + T4M; T4N = FNMS(KP250000000, T4G, T4M); T5t = T34 - T3b; T3c = T34 + T3b; { E T3f, T3e, T4P, T4T, T3d, T3M, T3g; T4P = FMA(KP559016994, T4O, T4N); T4T = FNMS(KP559016994, T4O, T4N); T3f = T2X - T3c; T3d = T2X + T3c; Ip[WS(rs, 4)] = FMA(KP951056516, T4S, T4P); Im[WS(rs, 3)] = FMS(KP951056516, T4S, T4P); Ip[WS(rs, 8)] = FMA(KP951056516, T4U, T4T); Im[WS(rs, 7)] = FMS(KP951056516, T4U, T4T); Rm[WS(rs, 4)] = T2I + T3d; T3e = FNMS(KP250000000, T3d, T2I); T5u = FMA(KP618033988, T5t, T5s); T5w = FNMS(KP618033988, T5s, T5t); T5o = T58 + T57; T59 = T57 - T58; T3M = FMA(KP559016994, T3f, T3e); T3g = FNMS(KP559016994, T3f, T3e); Rp[WS(rs, 7)] = FNMS(KP951056516, T3L, T3g); Rp[WS(rs, 3)] = FMA(KP951056516, T3L, T3g); Rm[0] = FNMS(KP951056516, T3N, T3M); Rm[WS(rs, 8)] = FMA(KP951056516, T3N, T3M); T5q = T5l - T5m; T5n = T5l + T5m; } } { E T5a, T5b, T47, T45, T5h, T5g, T3V, T3X, T41, T44, T5p, T3W, T46, T3Y; T5a = T3Z + T40; T41 = T3Z - T40; T44 = T42 - T43; T5b = T42 + T43; Im[WS(rs, 4)] = T5n - T5o; T5p = FMA(KP250000000, T5n, T5o); T47 = FNMS(KP618033988, T41, T44); T45 = FMA(KP618033988, T44, T41); { E T5r, T5v, T3R, T3U; T5r = FNMS(KP559016994, T5q, T5p); T5v = FMA(KP559016994, T5q, T5p); T3R = T3P + T3Q; T5h = T3P - T3Q; T5g = T3S - T3T; T3U = T3S + T3T; Im[0] = -(FMA(KP951056516, T5u, T5r)); Im[WS(rs, 8)] = FMS(KP951056516, T5u, T5r); Ip[WS(rs, 7)] = FMA(KP951056516, T5w, T5v); Ip[WS(rs, 3)] = FNMS(KP951056516, T5w, T5v); T3V = T3R + T3U; T3X = T3R - T3U; } Rp[WS(rs, 5)] = T3O + T3V; T3W = FNMS(KP250000000, T3V, T3O); T5i = FNMS(KP618033988, T5h, T5g); T5k = FMA(KP618033988, T5g, T5h); T46 = FNMS(KP559016994, T3X, T3W); T3Y = FMA(KP559016994, T3X, T3W); Rp[WS(rs, 9)] = FNMS(KP951056516, T45, T3Y); Rp[WS(rs, 1)] = FMA(KP951056516, T45, T3Y); Rm[WS(rs, 2)] = FNMS(KP951056516, T47, T46); Rm[WS(rs, 6)] = FMA(KP951056516, T47, T46); T5e = T5a - T5b; T5c = T5a + T5b; } } } } } Ip[WS(rs, 5)] = T5c + T59; T5d = FNMS(KP250000000, T5c, T59); T5j = FMA(KP559016994, T5e, T5d); T5f = FNMS(KP559016994, T5e, T5d); Im[WS(rs, 2)] = -(FMA(KP951056516, T5i, T5f)); Im[WS(rs, 6)] = FMS(KP951056516, T5i, T5f); Ip[WS(rs, 9)] = FMA(KP951056516, T5k, T5j); Ip[WS(rs, 1)] = FNMS(KP951056516, T5k, T5j); } } } static const tw_instr twinstr[] = { {TW_CEXP, 1, 1}, {TW_CEXP, 1, 3}, {TW_CEXP, 1, 9}, {TW_CEXP, 1, 19}, {TW_NEXT, 1, 0} }; static const hc2c_desc desc = { 20, "hc2cf2_20", twinstr, &GENUS, {136, 58, 140, 0} }; void X(codelet_hc2cf2_20) (planner *p) { X(khc2c_register) (p, hc2cf2_20, &desc, HC2C_VIA_RDFT); } #else /* HAVE_FMA */ /* Generated by: ../../../genfft/gen_hc2c.native -compact -variables 4 -pipeline-latency 4 -twiddle-log3 -precompute-twiddles -n 20 -dit -name hc2cf2_20 -include hc2cf.h */ /* * This function contains 276 FP additions, 164 FP multiplications, * (or, 204 additions, 92 multiplications, 72 fused multiply/add), * 123 stack variables, 4 constants, and 80 memory accesses */ #include "hc2cf.h" static void hc2cf2_20(R *Rp, R *Ip, R *Rm, R *Im, const R *W, stride rs, INT mb, INT me, INT ms) { DK(KP587785252, +0.587785252292473129168705954639072768597652438); DK(KP951056516, +0.951056516295153572116439333379382143405698634); DK(KP250000000, +0.250000000000000000000000000000000000000000000); DK(KP559016994, +0.559016994374947424102293417182819058860154590); { INT m; for (m = mb, W = W + ((mb - 1) * 8); m < me; m = m + 1, Rp = Rp + ms, Ip = Ip + ms, Rm = Rm - ms, Im = Im - ms, W = W + 8, MAKE_VOLATILE_STRIDE(80, rs)) { E T2, T5, Tg, Ti, Tk, To, T1h, T1f, T6, T3, T8, T14, T1Q, Tc, T1O; E T1v, T18, T1t, T1n, T24, T1j, T22, Tq, Tu, T1E, T1G, Tx, Ty, Tz, TJ; E T1Z, TB, T1X, T1A, TZ, TL, T1y, TX; { E T7, T16, Ta, T13, T4, T17, Tb, T12; { E Th, Tn, Tj, Tm; T2 = W[0]; T5 = W[1]; Tg = W[2]; Ti = W[3]; Th = T2 * Tg; Tn = T5 * Tg; Tj = T5 * Ti; Tm = T2 * Ti; Tk = Th - Tj; To = Tm + Tn; T1h = Tm - Tn; T1f = Th + Tj; T6 = W[5]; T7 = T5 * T6; T16 = Tg * T6; Ta = T2 * T6; T13 = Ti * T6; T3 = W[4]; T4 = T2 * T3; T17 = Ti * T3; Tb = T5 * T3; T12 = Tg * T3; } T8 = T4 - T7; T14 = T12 + T13; T1Q = T16 + T17; Tc = Ta + Tb; T1O = T12 - T13; T1v = Ta - Tb; T18 = T16 - T17; T1t = T4 + T7; { E T1l, T1m, T1g, T1i; T1l = T1f * T6; T1m = T1h * T3; T1n = T1l + T1m; T24 = T1l - T1m; T1g = T1f * T3; T1i = T1h * T6; T1j = T1g - T1i; T22 = T1g + T1i; { E Tl, Tp, Ts, Tt; Tl = Tk * T3; Tp = To * T6; Tq = Tl + Tp; Ts = Tk * T6; Tt = To * T3; Tu = Ts - Tt; T1E = Tl - Tp; T1G = Ts + Tt; Tx = W[6]; Ty = W[7]; Tz = FMA(Tk, Tx, To * Ty); TJ = FMA(Tq, Tx, Tu * Ty); T1Z = FNMS(T1h, Tx, T1f * Ty); TB = FNMS(To, Tx, Tk * Ty); T1X = FMA(T1f, Tx, T1h * Ty); T1A = FNMS(T5, Tx, T2 * Ty); TZ = FNMS(Ti, Tx, Tg * Ty); TL = FNMS(Tu, Tx, Tq * Ty); T1y = FMA(T2, Tx, T5 * Ty); TX = FMA(Tg, Tx, Ti * Ty); } } } { E TF, T2b, T4D, T4M, T2K, T3r, T4a, T4m, T1N, T28, T29, T3J, T3M, T44, T3U; E T3V, T4j, T2f, T2g, T2h, T2n, T2s, T4K, T3g, T3h, T4z, T3n, T3o, T3p, T30; E T35, T36, TW, T1r, T1s, T3C, T3F, T43, T3X, T3Y, T4k, T2c, T2d, T2e, T2y; E T2D, T4J, T3d, T3e, T4y, T3k, T3l, T3m, T2P, T2U, T2V; { E T1, T48, Te, T47, Tw, T2H, TD, T2I, T9, Td; T1 = Rp[0]; T48 = Rm[0]; T9 = Rp[WS(rs, 5)]; Td = Rm[WS(rs, 5)]; Te = FMA(T8, T9, Tc * Td); T47 = FNMS(Tc, T9, T8 * Td); { E Tr, Tv, TA, TC; Tr = Ip[WS(rs, 2)]; Tv = Im[WS(rs, 2)]; Tw = FMA(Tq, Tr, Tu * Tv); T2H = FNMS(Tu, Tr, Tq * Tv); TA = Ip[WS(rs, 7)]; TC = Im[WS(rs, 7)]; TD = FMA(Tz, TA, TB * TC); T2I = FNMS(TB, TA, Tz * TC); } { E Tf, TE, T4B, T4C; Tf = T1 + Te; TE = Tw + TD; TF = Tf - TE; T2b = Tf + TE; T4B = T48 - T47; T4C = Tw - TD; T4D = T4B - T4C; T4M = T4C + T4B; } { E T2G, T2J, T46, T49; T2G = T1 - Te; T2J = T2H - T2I; T2K = T2G - T2J; T3r = T2G + T2J; T46 = T2H + T2I; T49 = T47 + T48; T4a = T46 + T49; T4m = T49 - T46; } } { E T1D, T3H, T2l, T2W, T27, T3L, T2r, T34, T1M, T3I, T2m, T2Z, T1W, T3K, T2q; E T31; { E T1x, T2j, T1C, T2k; { E T1u, T1w, T1z, T1B; T1u = Rp[WS(rs, 4)]; T1w = Rm[WS(rs, 4)]; T1x = FMA(T1t, T1u, T1v * T1w); T2j = FNMS(T1v, T1u, T1t * T1w); T1z = Rp[WS(rs, 9)]; T1B = Rm[WS(rs, 9)]; T1C = FMA(T1y, T1z, T1A * T1B); T2k = FNMS(T1A, T1z, T1y * T1B); } T1D = T1x + T1C; T3H = T2j + T2k; T2l = T2j - T2k; T2W = T1x - T1C; } { E T21, T32, T26, T33; { E T1Y, T20, T23, T25; T1Y = Ip[WS(rs, 8)]; T20 = Im[WS(rs, 8)]; T21 = FMA(T1X, T1Y, T1Z * T20); T32 = FNMS(T1Z, T1Y, T1X * T20); T23 = Ip[WS(rs, 3)]; T25 = Im[WS(rs, 3)]; T26 = FMA(T22, T23, T24 * T25); T33 = FNMS(T24, T23, T22 * T25); } T27 = T21 + T26; T3L = T32 + T33; T2r = T21 - T26; T34 = T32 - T33; } { E T1I, T2X, T1L, T2Y; { E T1F, T1H, T1J, T1K; T1F = Ip[WS(rs, 6)]; T1H = Im[WS(rs, 6)]; T1I = FMA(T1E, T1F, T1G * T1H); T2X = FNMS(T1G, T1F, T1E * T1H); T1J = Ip[WS(rs, 1)]; T1K = Im[WS(rs, 1)]; T1L = FMA(Tg, T1J, Ti * T1K); T2Y = FNMS(Ti, T1J, Tg * T1K); } T1M = T1I + T1L; T3I = T2X + T2Y; T2m = T1I - T1L; T2Z = T2X - T2Y; } { E T1S, T2o, T1V, T2p; { E T1P, T1R, T1T, T1U; T1P = Rp[WS(rs, 6)]; T1R = Rm[WS(rs, 6)]; T1S = FMA(T1O, T1P, T1Q * T1R); T2o = FNMS(T1Q, T1P, T1O * T1R); T1T = Rp[WS(rs, 1)]; T1U = Rm[WS(rs, 1)]; T1V = FMA(T1f, T1T, T1h * T1U); T2p = FNMS(T1h, T1T, T1f * T1U); } T1W = T1S + T1V; T3K = T2o + T2p; T2q = T2o - T2p; T31 = T1S - T1V; } T1N = T1D - T1M; T28 = T1W - T27; T29 = T1N + T28; T3J = T3H + T3I; T3M = T3K + T3L; T44 = T3J + T3M; T3U = T3H - T3I; T3V = T3L - T3K; T4j = T3V - T3U; T2f = T1D + T1M; T2g = T1W + T27; T2h = T2f + T2g; T2n = T2l + T2m; T2s = T2q + T2r; T4K = T2n + T2s; T3g = T2l - T2m; T3h = T2q - T2r; T4z = T3g + T3h; T3n = T2W + T2Z; T3o = T31 + T34; T3p = T3n + T3o; T30 = T2W - T2Z; T35 = T31 - T34; T36 = T30 + T35; } { E TO, T3A, T2w, T2L, T1q, T3E, T2z, T2T, TV, T3B, T2x, T2O, T1b, T3D, T2C; E T2Q; { E TI, T2u, TN, T2v; { E TG, TH, TK, TM; TG = Rp[WS(rs, 2)]; TH = Rm[WS(rs, 2)]; TI = FMA(Tk, TG, To * TH); T2u = FNMS(To, TG, Tk * TH); TK = Rp[WS(rs, 7)]; TM = Rm[WS(rs, 7)]; TN = FMA(TJ, TK, TL * TM); T2v = FNMS(TL, TK, TJ * TM); } TO = TI + TN; T3A = T2u + T2v; T2w = T2u - T2v; T2L = TI - TN; } { E T1e, T2R, T1p, T2S; { E T1c, T1d, T1k, T1o; T1c = Ip[0]; T1d = Im[0]; T1e = FMA(T2, T1c, T5 * T1d); T2R = FNMS(T5, T1c, T2 * T1d); T1k = Ip[WS(rs, 5)]; T1o = Im[WS(rs, 5)]; T1p = FMA(T1j, T1k, T1n * T1o); T2S = FNMS(T1n, T1k, T1j * T1o); } T1q = T1e + T1p; T3E = T2R + T2S; T2z = T1p - T1e; T2T = T2R - T2S; } { E TR, T2M, TU, T2N; { E TP, TQ, TS, TT; TP = Ip[WS(rs, 4)]; TQ = Im[WS(rs, 4)]; TR = FMA(T3, TP, T6 * TQ); T2M = FNMS(T6, TP, T3 * TQ); TS = Ip[WS(rs, 9)]; TT = Im[WS(rs, 9)]; TU = FMA(Tx, TS, Ty * TT); T2N = FNMS(Ty, TS, Tx * TT); } TV = TR + TU; T3B = T2M + T2N; T2x = TR - TU; T2O = T2M - T2N; } { E T11, T2A, T1a, T2B; { E TY, T10, T15, T19; TY = Rp[WS(rs, 8)]; T10 = Rm[WS(rs, 8)]; T11 = FMA(TX, TY, TZ * T10); T2A = FNMS(TZ, TY, TX * T10); T15 = Rp[WS(rs, 3)]; T19 = Rm[WS(rs, 3)]; T1a = FMA(T14, T15, T18 * T19); T2B = FNMS(T18, T15, T14 * T19); } T1b = T11 + T1a; T3D = T2A + T2B; T2C = T2A - T2B; T2Q = T11 - T1a; } TW = TO - TV; T1r = T1b - T1q; T1s = TW + T1r; T3C = T3A + T3B; T3F = T3D + T3E; T43 = T3C + T3F; T3X = T3A - T3B; T3Y = T3D - T3E; T4k = T3X + T3Y; T2c = TO + TV; T2d = T1b + T1q; T2e = T2c + T2d; T2y = T2w + T2x; T2D = T2z - T2C; T4J = T2D - T2y; T3d = T2w - T2x; T3e = T2C + T2z; T4y = T3d + T3e; T3k = T2L + T2O; T3l = T2Q + T2T; T3m = T3k + T3l; T2P = T2L - T2O; T2U = T2Q - T2T; T2V = T2P + T2U; } { E T3S, T2a, T3R, T40, T42, T3W, T3Z, T41, T3T; T3S = KP559016994 * (T1s - T29); T2a = T1s + T29; T3R = FNMS(KP250000000, T2a, TF); T3W = T3U + T3V; T3Z = T3X - T3Y; T40 = FNMS(KP587785252, T3Z, KP951056516 * T3W); T42 = FMA(KP951056516, T3Z, KP587785252 * T3W); Rm[WS(rs, 9)] = TF + T2a; T41 = T3S + T3R; Rm[WS(rs, 5)] = T41 - T42; Rp[WS(rs, 6)] = T41 + T42; T3T = T3R - T3S; Rp[WS(rs, 2)] = T3T - T40; Rm[WS(rs, 1)] = T3T + T40; } { E T4r, T4l, T4q, T4p, T4t, T4n, T4o, T4u, T4s; T4r = KP559016994 * (T4k + T4j); T4l = T4j - T4k; T4q = FMA(KP250000000, T4l, T4m); T4n = T1r - TW; T4o = T1N - T28; T4p = FMA(KP587785252, T4n, KP951056516 * T4o); T4t = FNMS(KP587785252, T4o, KP951056516 * T4n); Im[WS(rs, 9)] = T4l - T4m; T4u = T4r + T4q; Im[WS(rs, 5)] = T4t - T4u; Ip[WS(rs, 6)] = T4t + T4u; T4s = T4q - T4r; Im[WS(rs, 1)] = T4p - T4s; Ip[WS(rs, 2)] = T4p + T4s; } { E T3x, T2i, T3y, T3O, T3Q, T3G, T3N, T3P, T3z; T3x = KP559016994 * (T2e - T2h); T2i = T2e + T2h; T3y = FNMS(KP250000000, T2i, T2b); T3G = T3C - T3F; T3N = T3J - T3M; T3O = FMA(KP951056516, T3G, KP587785252 * T3N); T3Q = FNMS(KP587785252, T3G, KP951056516 * T3N); Rp[0] = T2b + T2i; T3P = T3y - T3x; Rm[WS(rs, 7)] = T3P - T3Q; Rp[WS(rs, 8)] = T3P + T3Q; T3z = T3x + T3y; Rp[WS(rs, 4)] = T3z - T3O; Rm[WS(rs, 3)] = T3z + T3O; } { E T4e, T45, T4f, T4d, T4h, T4b, T4c, T4i, T4g; T4e = KP559016994 * (T43 - T44); T45 = T43 + T44; T4f = FNMS(KP250000000, T45, T4a); T4b = T2c - T2d; T4c = T2f - T2g; T4d = FMA(KP951056516, T4b, KP587785252 * T4c); T4h = FNMS(KP951056516, T4c, KP587785252 * T4b); Ip[0] = T45 + T4a; T4i = T4f - T4e; Im[WS(rs, 7)] = T4h - T4i; Ip[WS(rs, 8)] = T4h + T4i; T4g = T4e + T4f; Im[WS(rs, 3)] = T4d - T4g; Ip[WS(rs, 4)] = T4d + T4g; } { E T39, T37, T38, T2F, T3b, T2t, T2E, T3c, T3a; T39 = KP559016994 * (T2V - T36); T37 = T2V + T36; T38 = FNMS(KP250000000, T37, T2K); T2t = T2n - T2s; T2E = T2y + T2D; T2F = FNMS(KP587785252, T2E, KP951056516 * T2t); T3b = FMA(KP951056516, T2E, KP587785252 * T2t); Rm[WS(rs, 4)] = T2K + T37; T3c = T39 + T38; Rm[WS(rs, 8)] = T3b + T3c; Rm[0] = T3c - T3b; T3a = T38 - T39; Rp[WS(rs, 3)] = T2F + T3a; Rp[WS(rs, 7)] = T3a - T2F; } { E T4Q, T4L, T4R, T4P, T4U, T4N, T4O, T4T, T4S; T4Q = KP559016994 * (T4J + T4K); T4L = T4J - T4K; T4R = FMA(KP250000000, T4L, T4M); T4N = T2P - T2U; T4O = T30 - T35; T4P = FMA(KP951056516, T4N, KP587785252 * T4O); T4U = FNMS(KP587785252, T4N, KP951056516 * T4O); Im[WS(rs, 4)] = T4L - T4M; T4T = T4Q + T4R; Ip[WS(rs, 3)] = T4T - T4U; Ip[WS(rs, 7)] = T4U + T4T; T4S = T4Q - T4R; Im[WS(rs, 8)] = T4P + T4S; Im[0] = T4S - T4P; } { E T3q, T3s, T3t, T3j, T3v, T3f, T3i, T3w, T3u; T3q = KP559016994 * (T3m - T3p); T3s = T3m + T3p; T3t = FNMS(KP250000000, T3s, T3r); T3f = T3d - T3e; T3i = T3g - T3h; T3j = FMA(KP951056516, T3f, KP587785252 * T3i); T3v = FNMS(KP587785252, T3f, KP951056516 * T3i); Rp[WS(rs, 5)] = T3r + T3s; T3w = T3t - T3q; Rm[WS(rs, 6)] = T3v + T3w; Rm[WS(rs, 2)] = T3w - T3v; T3u = T3q + T3t; Rp[WS(rs, 1)] = T3j + T3u; Rp[WS(rs, 9)] = T3u - T3j; } { E T4A, T4E, T4F, T4x, T4I, T4v, T4w, T4H, T4G; T4A = KP559016994 * (T4y - T4z); T4E = T4y + T4z; T4F = FNMS(KP250000000, T4E, T4D); T4v = T3n - T3o; T4w = T3k - T3l; T4x = FNMS(KP587785252, T4w, KP951056516 * T4v); T4I = FMA(KP951056516, T4w, KP587785252 * T4v); Ip[WS(rs, 5)] = T4E + T4D; T4H = T4A + T4F; Ip[WS(rs, 1)] = T4H - T4I; Ip[WS(rs, 9)] = T4I + T4H; T4G = T4A - T4F; Im[WS(rs, 6)] = T4x + T4G; Im[WS(rs, 2)] = T4G - T4x; } } } } } static const tw_instr twinstr[] = { {TW_CEXP, 1, 1}, {TW_CEXP, 1, 3}, {TW_CEXP, 1, 9}, {TW_CEXP, 1, 19}, {TW_NEXT, 1, 0} }; static const hc2c_desc desc = { 20, "hc2cf2_20", twinstr, &GENUS, {204, 92, 72, 0} }; void X(codelet_hc2cf2_20) (planner *p) { X(khc2c_register) (p, hc2cf2_20, &desc, HC2C_VIA_RDFT); } #endif /* HAVE_FMA */ fftw-3.3.3/rdft/scalar/r2cf/hc2cfdft_2.c0000644000175000017500000001006512054410714014517 00000000000000/* * Copyright (c) 2003, 2007-11 Matteo Frigo * Copyright (c) 2003, 2007-11 Massachusetts Institute of Technology * * This program is free software; you can redistribute it and/or modify * it under the terms of the GNU General Public License as published by * the Free Software Foundation; either version 2 of the License, or * (at your option) any later version. * * This program is distributed in the hope that it will be useful, * but WITHOUT ANY WARRANTY; without even the implied warranty of * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the * GNU General Public License for more details. * * You should have received a copy of the GNU General Public License * along with this program; if not, write to the Free Software * Foundation, Inc., 51 Franklin Street, Fifth Floor, Boston, MA 02110-1301 USA * */ /* This file was automatically generated --- DO NOT EDIT */ /* Generated on Sun Nov 25 07:40:44 EST 2012 */ #include "codelet-rdft.h" #ifdef HAVE_FMA /* Generated by: ../../../genfft/gen_hc2cdft.native -fma -reorder-insns -schedule-for-pipeline -compact -variables 4 -pipeline-latency 4 -n 2 -dit -name hc2cfdft_2 -include hc2cf.h */ /* * This function contains 10 FP additions, 8 FP multiplications, * (or, 8 additions, 6 multiplications, 2 fused multiply/add), * 12 stack variables, 1 constants, and 8 memory accesses */ #include "hc2cf.h" static void hc2cfdft_2(R *Rp, R *Ip, R *Rm, R *Im, const R *W, stride rs, INT mb, INT me, INT ms) { DK(KP500000000, +0.500000000000000000000000000000000000000000000); { INT m; for (m = mb, W = W + ((mb - 1) * 2); m < me; m = m + 1, Rp = Rp + ms, Ip = Ip + ms, Rm = Rm - ms, Im = Im - ms, W = W + 2, MAKE_VOLATILE_STRIDE(8, rs)) { E T9, Ta, T3, Tc, T7, T4; { E T1, T2, T5, T6; T1 = Ip[0]; T2 = Im[0]; T5 = Rm[0]; T6 = Rp[0]; T9 = W[1]; Ta = T1 + T2; T3 = T1 - T2; Tc = T6 + T5; T7 = T5 - T6; T4 = W[0]; } { E Td, T8, Te, Tb; Td = T9 * T7; T8 = T4 * T7; Te = FMA(T4, Ta, Td); Tb = FNMS(T9, Ta, T8); Rp[0] = KP500000000 * (Tc + Te); Rm[0] = KP500000000 * (Tc - Te); Im[0] = KP500000000 * (Tb - T3); Ip[0] = KP500000000 * (T3 + Tb); } } } } static const tw_instr twinstr[] = { {TW_FULL, 1, 2}, {TW_NEXT, 1, 0} }; static const hc2c_desc desc = { 2, "hc2cfdft_2", twinstr, &GENUS, {8, 6, 2, 0} }; void X(codelet_hc2cfdft_2) (planner *p) { X(khc2c_register) (p, hc2cfdft_2, &desc, HC2C_VIA_DFT); } #else /* HAVE_FMA */ /* Generated by: ../../../genfft/gen_hc2cdft.native -compact -variables 4 -pipeline-latency 4 -n 2 -dit -name hc2cfdft_2 -include hc2cf.h */ /* * This function contains 10 FP additions, 8 FP multiplications, * (or, 8 additions, 6 multiplications, 2 fused multiply/add), * 10 stack variables, 1 constants, and 8 memory accesses */ #include "hc2cf.h" static void hc2cfdft_2(R *Rp, R *Ip, R *Rm, R *Im, const R *W, stride rs, INT mb, INT me, INT ms) { DK(KP500000000, +0.500000000000000000000000000000000000000000000); { INT m; for (m = mb, W = W + ((mb - 1) * 2); m < me; m = m + 1, Rp = Rp + ms, Ip = Ip + ms, Rm = Rm - ms, Im = Im - ms, W = W + 2, MAKE_VOLATILE_STRIDE(8, rs)) { E T3, T9, T7, Tb; { E T1, T2, T5, T6; T1 = Ip[0]; T2 = Im[0]; T3 = T1 - T2; T9 = T1 + T2; T5 = Rm[0]; T6 = Rp[0]; T7 = T5 - T6; Tb = T6 + T5; } { E Ta, Tc, T4, T8; T4 = W[0]; T8 = W[1]; Ta = FNMS(T8, T9, T4 * T7); Tc = FMA(T8, T7, T4 * T9); Ip[0] = KP500000000 * (T3 + Ta); Rp[0] = KP500000000 * (Tb + Tc); Im[0] = KP500000000 * (Ta - T3); Rm[0] = KP500000000 * (Tb - Tc); } } } } static const tw_instr twinstr[] = { {TW_FULL, 1, 2}, {TW_NEXT, 1, 0} }; static const hc2c_desc desc = { 2, "hc2cfdft_2", twinstr, &GENUS, {8, 6, 2, 0} }; void X(codelet_hc2cfdft_2) (planner *p) { X(khc2c_register) (p, hc2cfdft_2, &desc, HC2C_VIA_DFT); } #endif /* HAVE_FMA */ fftw-3.3.3/rdft/scalar/r2cf/hc2cfdft_4.c0000644000175000017500000001361212054410714014522 00000000000000/* * Copyright (c) 2003, 2007-11 Matteo Frigo * Copyright (c) 2003, 2007-11 Massachusetts Institute of Technology * * This program is free software; you can redistribute it and/or modify * it under the terms of the GNU General Public License as published by * the Free Software Foundation; either version 2 of the License, or * (at your option) any later version. * * This program is distributed in the hope that it will be useful, * but WITHOUT ANY WARRANTY; without even the implied warranty of * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the * GNU General Public License for more details. * * You should have received a copy of the GNU General Public License * along with this program; if not, write to the Free Software * Foundation, Inc., 51 Franklin Street, Fifth Floor, Boston, MA 02110-1301 USA * */ /* This file was automatically generated --- DO NOT EDIT */ /* Generated on Sun Nov 25 07:40:44 EST 2012 */ #include "codelet-rdft.h" #ifdef HAVE_FMA /* Generated by: ../../../genfft/gen_hc2cdft.native -fma -reorder-insns -schedule-for-pipeline -compact -variables 4 -pipeline-latency 4 -n 4 -dit -name hc2cfdft_4 -include hc2cf.h */ /* * This function contains 30 FP additions, 20 FP multiplications, * (or, 24 additions, 14 multiplications, 6 fused multiply/add), * 32 stack variables, 1 constants, and 16 memory accesses */ #include "hc2cf.h" static void hc2cfdft_4(R *Rp, R *Ip, R *Rm, R *Im, const R *W, stride rs, INT mb, INT me, INT ms) { DK(KP500000000, +0.500000000000000000000000000000000000000000000); { INT m; for (m = mb, W = W + ((mb - 1) * 6); m < me; m = m + 1, Rp = Rp + ms, Ip = Ip + ms, Rm = Rm - ms, Im = Im - ms, W = W + 6, MAKE_VOLATILE_STRIDE(16, rs)) { E Td, Tu, Tr, T4, Tm, To, T9, T5, TA, Tp, Tv, TD, T6, Tq; { E Tk, Tl, Tf, TC, Tj, T7, T8, T1, Tn, Tb, Tc; Tb = Ip[0]; Tc = Im[0]; { E Ti, Tg, Th, T2, T3; Tg = Rm[0]; Th = Rp[0]; Tk = W[1]; Tl = Tb + Tc; Td = Tb - Tc; Tu = Th + Tg; Ti = Tg - Th; Tf = W[0]; T2 = Ip[WS(rs, 1)]; T3 = Im[WS(rs, 1)]; TC = Tk * Ti; Tj = Tf * Ti; T7 = Rp[WS(rs, 1)]; Tr = T2 + T3; T4 = T2 - T3; T8 = Rm[WS(rs, 1)]; T1 = W[2]; Tn = W[4]; } Tm = FNMS(Tk, Tl, Tj); To = T7 - T8; T9 = T7 + T8; T5 = T1 * T4; TA = Tn * Tr; Tp = Tn * To; Tv = T1 * T9; TD = FMA(Tf, Tl, TC); T6 = W[3]; Tq = W[5]; } { E Tw, Ta, TB, Ts; Tw = FMA(T6, T4, Tv); Ta = FNMS(T6, T9, T5); TB = FNMS(Tq, To, TA); Ts = FMA(Tq, Tr, Tp); { E TF, Tx, Te, Tz; TF = Tu + Tw; Tx = Tu - Tw; Te = Ta + Td; Tz = Td - Ta; { E TG, TE, Tt, Ty; TG = TB + TD; TE = TB - TD; Tt = Tm - Ts; Ty = Ts + Tm; Im[0] = KP500000000 * (TE - Tz); Ip[WS(rs, 1)] = KP500000000 * (Tz + TE); Rp[0] = KP500000000 * (TF + TG); Rm[WS(rs, 1)] = KP500000000 * (TF - TG); Rp[WS(rs, 1)] = KP500000000 * (Tx + Ty); Rm[0] = KP500000000 * (Tx - Ty); Im[WS(rs, 1)] = KP500000000 * (Tt - Te); Ip[0] = KP500000000 * (Te + Tt); } } } } } } static const tw_instr twinstr[] = { {TW_FULL, 1, 4}, {TW_NEXT, 1, 0} }; static const hc2c_desc desc = { 4, "hc2cfdft_4", twinstr, &GENUS, {24, 14, 6, 0} }; void X(codelet_hc2cfdft_4) (planner *p) { X(khc2c_register) (p, hc2cfdft_4, &desc, HC2C_VIA_DFT); } #else /* HAVE_FMA */ /* Generated by: ../../../genfft/gen_hc2cdft.native -compact -variables 4 -pipeline-latency 4 -n 4 -dit -name hc2cfdft_4 -include hc2cf.h */ /* * This function contains 30 FP additions, 20 FP multiplications, * (or, 24 additions, 14 multiplications, 6 fused multiply/add), * 18 stack variables, 1 constants, and 16 memory accesses */ #include "hc2cf.h" static void hc2cfdft_4(R *Rp, R *Ip, R *Rm, R *Im, const R *W, stride rs, INT mb, INT me, INT ms) { DK(KP500000000, +0.500000000000000000000000000000000000000000000); { INT m; for (m = mb, W = W + ((mb - 1) * 6); m < me; m = m + 1, Rp = Rp + ms, Ip = Ip + ms, Rm = Rm - ms, Im = Im - ms, W = W + 6, MAKE_VOLATILE_STRIDE(16, rs)) { E Tc, Tr, Tk, Tx, T9, Ts, Tp, Tw; { E Ta, Tb, Tj, Tf, Tg, Th, Te, Ti; Ta = Ip[0]; Tb = Im[0]; Tj = Ta + Tb; Tf = Rm[0]; Tg = Rp[0]; Th = Tf - Tg; Tc = Ta - Tb; Tr = Tg + Tf; Te = W[0]; Ti = W[1]; Tk = FNMS(Ti, Tj, Te * Th); Tx = FMA(Ti, Th, Te * Tj); } { E T4, To, T8, Tm; { E T2, T3, T6, T7; T2 = Ip[WS(rs, 1)]; T3 = Im[WS(rs, 1)]; T4 = T2 - T3; To = T2 + T3; T6 = Rp[WS(rs, 1)]; T7 = Rm[WS(rs, 1)]; T8 = T6 + T7; Tm = T6 - T7; } { E T1, T5, Tl, Tn; T1 = W[2]; T5 = W[3]; T9 = FNMS(T5, T8, T1 * T4); Ts = FMA(T1, T8, T5 * T4); Tl = W[4]; Tn = W[5]; Tp = FMA(Tl, Tm, Tn * To); Tw = FNMS(Tn, Tm, Tl * To); } } { E Td, Tq, Tz, TA; Td = T9 + Tc; Tq = Tk - Tp; Ip[0] = KP500000000 * (Td + Tq); Im[WS(rs, 1)] = KP500000000 * (Tq - Td); Tz = Tr + Ts; TA = Tw + Tx; Rm[WS(rs, 1)] = KP500000000 * (Tz - TA); Rp[0] = KP500000000 * (Tz + TA); } { E Tt, Tu, Tv, Ty; Tt = Tr - Ts; Tu = Tp + Tk; Rm[0] = KP500000000 * (Tt - Tu); Rp[WS(rs, 1)] = KP500000000 * (Tt + Tu); Tv = Tc - T9; Ty = Tw - Tx; Ip[WS(rs, 1)] = KP500000000 * (Tv + Ty); Im[0] = KP500000000 * (Ty - Tv); } } } } static const tw_instr twinstr[] = { {TW_FULL, 1, 4}, {TW_NEXT, 1, 0} }; static const hc2c_desc desc = { 4, "hc2cfdft_4", twinstr, &GENUS, {24, 14, 6, 0} }; void X(codelet_hc2cfdft_4) (planner *p) { X(khc2c_register) (p, hc2cfdft_4, &desc, HC2C_VIA_DFT); } #endif /* HAVE_FMA */ fftw-3.3.3/rdft/scalar/r2cf/hc2cfdft_6.c0000644000175000017500000002175412054410714014532 00000000000000/* * Copyright (c) 2003, 2007-11 Matteo Frigo * Copyright (c) 2003, 2007-11 Massachusetts Institute of Technology * * This program is free software; you can redistribute it and/or modify * it under the terms of the GNU General Public License as published by * the Free Software Foundation; either version 2 of the License, or * (at your option) any later version. * * This program is distributed in the hope that it will be useful, * but WITHOUT ANY WARRANTY; without even the implied warranty of * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the * GNU General Public License for more details. * * You should have received a copy of the GNU General Public License * along with this program; if not, write to the Free Software * Foundation, Inc., 51 Franklin Street, Fifth Floor, Boston, MA 02110-1301 USA * */ /* This file was automatically generated --- DO NOT EDIT */ /* Generated on Sun Nov 25 07:40:44 EST 2012 */ #include "codelet-rdft.h" #ifdef HAVE_FMA /* Generated by: ../../../genfft/gen_hc2cdft.native -fma -reorder-insns -schedule-for-pipeline -compact -variables 4 -pipeline-latency 4 -n 6 -dit -name hc2cfdft_6 -include hc2cf.h */ /* * This function contains 58 FP additions, 44 FP multiplications, * (or, 36 additions, 22 multiplications, 22 fused multiply/add), * 42 stack variables, 2 constants, and 24 memory accesses */ #include "hc2cf.h" static void hc2cfdft_6(R *Rp, R *Ip, R *Rm, R *Im, const R *W, stride rs, INT mb, INT me, INT ms) { DK(KP866025403, +0.866025403784438646763723170752936183471402627); DK(KP500000000, +0.500000000000000000000000000000000000000000000); { INT m; for (m = mb, W = W + ((mb - 1) * 10); m < me; m = m + 1, Rp = Rp + ms, Ip = Ip + ms, Rm = Rm - ms, Im = Im - ms, W = W + 10, MAKE_VOLATILE_STRIDE(24, rs)) { E TP, TT, TN, TM, TY, T13; { E T3, TQ, TJ, T12, Tu, TB, TX, T10, Tj, Tf, Ti, Td, Th, TU, TS; { E TC, TI, TF, TH, TA, Tw, TZ; { E T1, T2, TD, TE; T1 = Ip[0]; T2 = Im[0]; TD = Rm[0]; TE = Rp[0]; TC = W[0]; T3 = T1 - T2; TI = T1 + T2; TQ = TE + TD; TF = TD - TE; TH = W[1]; } { E Tr, To, Ts, Tl, Tq; { E Tm, Tn, TG, T11; Tm = Rm[WS(rs, 2)]; Tn = Rp[WS(rs, 2)]; TG = TC * TF; T11 = TH * TF; Tr = Ip[WS(rs, 2)]; TA = Tn + Tm; To = Tm - Tn; TJ = FNMS(TH, TI, TG); T12 = FMA(TC, TI, T11); Ts = Im[WS(rs, 2)]; } Tl = W[8]; Tq = W[9]; { E Tz, Ty, TW, Tx, Tt, Tp; Tw = W[6]; Tx = Tr - Ts; Tt = Tr + Ts; Tp = Tl * To; Tz = W[7]; Ty = Tw * Tx; TW = Tl * Tt; Tu = FNMS(Tq, Tt, Tp); TZ = Tz * Tx; TB = FNMS(Tz, TA, Ty); TX = FMA(Tq, To, TW); } } { E T5, T6, Ta, Tb; T5 = Ip[WS(rs, 1)]; T10 = FMA(Tw, TA, TZ); T6 = Im[WS(rs, 1)]; Ta = Rp[WS(rs, 1)]; Tb = Rm[WS(rs, 1)]; { E T4, Tg, T7, Tc, T9, T8, TR; T4 = W[5]; Tg = T5 - T6; T7 = T5 + T6; Tj = Ta + Tb; Tc = Ta - Tb; T9 = W[4]; T8 = T4 * T7; Tf = W[2]; Ti = W[3]; TR = T9 * T7; Td = FMA(T9, Tc, T8); Th = Tf * Tg; TU = Ti * Tg; TS = FNMS(T4, Tc, TR); } } } { E Te, T1d, TK, Tv, T1a, T1b, Tk, TV; TP = Td + T3; Te = T3 - Td; Tk = FNMS(Ti, Tj, Th); TV = FMA(Tf, Tj, TU); T1d = TQ + TS; TT = TQ - TS; TN = TJ - TB; TK = TB + TJ; Tv = Tk + Tu; TM = Tu - Tk; TY = TV - TX; T1a = TV + TX; T1b = T10 + T12; T13 = T10 - T12; { E T1g, TL, T1e, T1c, T19, T1f; T1g = Tv - TK; TL = Tv + TK; T1e = T1a + T1b; T1c = T1a - T1b; T19 = FNMS(KP500000000, TL, Te); Ip[0] = KP500000000 * (Te + TL); T1f = FNMS(KP500000000, T1e, T1d); Rp[0] = KP500000000 * (T1d + T1e); Im[WS(rs, 1)] = -(KP500000000 * (FNMS(KP866025403, T1c, T19))); Ip[WS(rs, 2)] = KP500000000 * (FMA(KP866025403, T1c, T19)); Rm[WS(rs, 1)] = KP500000000 * (FMA(KP866025403, T1g, T1f)); Rp[WS(rs, 2)] = KP500000000 * (FNMS(KP866025403, T1g, T1f)); } } } { E TO, T16, T14, T18, T17, T15; TO = TM + TN; T16 = TN - TM; T14 = TY + T13; T18 = T13 - TY; T17 = FMA(KP500000000, TO, TP); Im[WS(rs, 2)] = KP500000000 * (TO - TP); T15 = FNMS(KP500000000, T14, TT); Rm[WS(rs, 2)] = KP500000000 * (TT + T14); Im[0] = -(KP500000000 * (FNMS(KP866025403, T18, T17))); Ip[WS(rs, 1)] = KP500000000 * (FMA(KP866025403, T18, T17)); Rm[0] = KP500000000 * (FNMS(KP866025403, T16, T15)); Rp[WS(rs, 1)] = KP500000000 * (FMA(KP866025403, T16, T15)); } } } } static const tw_instr twinstr[] = { {TW_FULL, 1, 6}, {TW_NEXT, 1, 0} }; static const hc2c_desc desc = { 6, "hc2cfdft_6", twinstr, &GENUS, {36, 22, 22, 0} }; void X(codelet_hc2cfdft_6) (planner *p) { X(khc2c_register) (p, hc2cfdft_6, &desc, HC2C_VIA_DFT); } #else /* HAVE_FMA */ /* Generated by: ../../../genfft/gen_hc2cdft.native -compact -variables 4 -pipeline-latency 4 -n 6 -dit -name hc2cfdft_6 -include hc2cf.h */ /* * This function contains 58 FP additions, 36 FP multiplications, * (or, 44 additions, 22 multiplications, 14 fused multiply/add), * 40 stack variables, 3 constants, and 24 memory accesses */ #include "hc2cf.h" static void hc2cfdft_6(R *Rp, R *Ip, R *Rm, R *Im, const R *W, stride rs, INT mb, INT me, INT ms) { DK(KP250000000, +0.250000000000000000000000000000000000000000000); DK(KP500000000, +0.500000000000000000000000000000000000000000000); DK(KP433012701, +0.433012701892219323381861585376468091735701313); { INT m; for (m = mb, W = W + ((mb - 1) * 10); m < me; m = m + 1, Rp = Rp + ms, Ip = Ip + ms, Rm = Rm - ms, Im = Im - ms, W = W + 10, MAKE_VOLATILE_STRIDE(24, rs)) { E T3, TM, Tc, TN, Ts, T10, TI, TR, TF, T11, TH, TU; { E T1, T2, TD, Tz, TA, TB, T7, Tf, Tb, Th, Tq, Tw, Tm, Tu, T4; E T8; { E T5, T6, T9, Ta; T1 = Ip[0]; T2 = Im[0]; TD = T1 + T2; Tz = Rm[0]; TA = Rp[0]; TB = Tz - TA; T5 = Ip[WS(rs, 1)]; T6 = Im[WS(rs, 1)]; T7 = T5 + T6; Tf = T5 - T6; T9 = Rp[WS(rs, 1)]; Ta = Rm[WS(rs, 1)]; Tb = T9 - Ta; Th = T9 + Ta; { E To, Tp, Tk, Tl; To = Rp[WS(rs, 2)]; Tp = Rm[WS(rs, 2)]; Tq = To - Tp; Tw = To + Tp; Tk = Ip[WS(rs, 2)]; Tl = Im[WS(rs, 2)]; Tm = Tk + Tl; Tu = Tk - Tl; } } T3 = T1 - T2; TM = TA + Tz; T4 = W[5]; T8 = W[4]; Tc = FMA(T4, T7, T8 * Tb); TN = FNMS(T4, Tb, T8 * T7); { E Ti, TP, Tr, TQ; { E Te, Tg, Tj, Tn; Te = W[2]; Tg = W[3]; Ti = FNMS(Tg, Th, Te * Tf); TP = FMA(Tg, Tf, Te * Th); Tj = W[9]; Tn = W[8]; Tr = FMA(Tj, Tm, Tn * Tq); TQ = FNMS(Tj, Tq, Tn * Tm); } Ts = Ti - Tr; T10 = TP + TQ; TI = Ti + Tr; TR = TP - TQ; } { E Tx, TS, TE, TT; { E Tt, Tv, Ty, TC; Tt = W[6]; Tv = W[7]; Tx = FNMS(Tv, Tw, Tt * Tu); TS = FMA(Tv, Tu, Tt * Tw); Ty = W[0]; TC = W[1]; TE = FNMS(TC, TD, Ty * TB); TT = FMA(TC, TB, Ty * TD); } TF = Tx + TE; T11 = TS + TT; TH = TE - Tx; TU = TS - TT; } } { E T12, Td, TG, TZ; T12 = KP433012701 * (T10 - T11); Td = T3 - Tc; TG = Ts + TF; TZ = FNMS(KP250000000, TG, KP500000000 * Td); Ip[0] = KP500000000 * (Td + TG); Im[WS(rs, 1)] = T12 - TZ; Ip[WS(rs, 2)] = TZ + T12; } { E T16, T13, T14, T15; T16 = KP433012701 * (Ts - TF); T13 = TM + TN; T14 = T10 + T11; T15 = FNMS(KP250000000, T14, KP500000000 * T13); Rp[WS(rs, 2)] = T15 - T16; Rp[0] = KP500000000 * (T13 + T14); Rm[WS(rs, 1)] = T16 + T15; } { E TY, TJ, TK, TX; TY = KP433012701 * (TU - TR); TJ = TH - TI; TK = Tc + T3; TX = FMA(KP500000000, TK, KP250000000 * TJ); Im[WS(rs, 2)] = KP500000000 * (TJ - TK); Im[0] = TY - TX; Ip[WS(rs, 1)] = TX + TY; } { E TL, TO, TV, TW; TL = KP433012701 * (TI + TH); TO = TM - TN; TV = TR + TU; TW = FNMS(KP250000000, TV, KP500000000 * TO); Rp[WS(rs, 1)] = TL + TW; Rm[WS(rs, 2)] = KP500000000 * (TO + TV); Rm[0] = TW - TL; } } } } static const tw_instr twinstr[] = { {TW_FULL, 1, 6}, {TW_NEXT, 1, 0} }; static const hc2c_desc desc = { 6, "hc2cfdft_6", twinstr, &GENUS, {44, 22, 14, 0} }; void X(codelet_hc2cfdft_6) (planner *p) { X(khc2c_register) (p, hc2cfdft_6, &desc, HC2C_VIA_DFT); } #endif /* HAVE_FMA */ fftw-3.3.3/rdft/scalar/r2cf/hc2cfdft_8.c0000644000175000017500000002573112054410715014534 00000000000000/* * Copyright (c) 2003, 2007-11 Matteo Frigo * Copyright (c) 2003, 2007-11 Massachusetts Institute of Technology * * This program is free software; you can redistribute it and/or modify * it under the terms of the GNU General Public License as published by * the Free Software Foundation; either version 2 of the License, or * (at your option) any later version. * * This program is distributed in the hope that it will be useful, * but WITHOUT ANY WARRANTY; without even the implied warranty of * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the * GNU General Public License for more details. * * You should have received a copy of the GNU General Public License * along with this program; if not, write to the Free Software * Foundation, Inc., 51 Franklin Street, Fifth Floor, Boston, MA 02110-1301 USA * */ /* This file was automatically generated --- DO NOT EDIT */ /* Generated on Sun Nov 25 07:40:44 EST 2012 */ #include "codelet-rdft.h" #ifdef HAVE_FMA /* Generated by: ../../../genfft/gen_hc2cdft.native -fma -reorder-insns -schedule-for-pipeline -compact -variables 4 -pipeline-latency 4 -n 8 -dit -name hc2cfdft_8 -include hc2cf.h */ /* * This function contains 82 FP additions, 52 FP multiplications, * (or, 60 additions, 30 multiplications, 22 fused multiply/add), * 55 stack variables, 2 constants, and 32 memory accesses */ #include "hc2cf.h" static void hc2cfdft_8(R *Rp, R *Ip, R *Rm, R *Im, const R *W, stride rs, INT mb, INT me, INT ms) { DK(KP707106781, +0.707106781186547524400844362104849039284835938); DK(KP500000000, +0.500000000000000000000000000000000000000000000); { INT m; for (m = mb, W = W + ((mb - 1) * 14); m < me; m = m + 1, Rp = Rp + ms, Ip = Ip + ms, Rm = Rm - ms, Im = Im - ms, W = W + 14, MAKE_VOLATILE_STRIDE(32, rs)) { E T1A, T1w, T1z, T1x, T1H, T1v, T1L, T1F; { E Ty, T14, TO, T1o, Tv, TG, T16, T1m, Ta, T19, T1h, TV, T10, TX, TZ; E Tk, T1i, TY, T1b, TF, TB, T1l; { E TH, TN, TK, TM; { E Tw, Tx, TI, TJ; Tw = Ip[0]; Tx = Im[0]; TI = Rm[0]; TJ = Rp[0]; TH = W[0]; Ty = Tw - Tx; TN = Tw + Tx; T14 = TJ + TI; TK = TI - TJ; TM = W[1]; } { E Ts, Tp, Tt, Tm, Tr; { E Tn, To, TL, T1n; Tn = Ip[WS(rs, 2)]; To = Im[WS(rs, 2)]; TL = TH * TK; T1n = TM * TK; Ts = Rp[WS(rs, 2)]; TF = Tn + To; Tp = Tn - To; TO = FNMS(TM, TN, TL); T1o = FMA(TH, TN, T1n); Tt = Rm[WS(rs, 2)]; } Tm = W[6]; Tr = W[7]; { E TE, TD, T15, TC, Tu, Tq; TB = W[8]; TC = Tt - Ts; Tu = Ts + Tt; Tq = Tm * Tp; TE = W[9]; TD = TB * TC; T15 = Tm * Tu; Tv = FNMS(Tr, Tu, Tq); T1l = TE * TC; TG = FNMS(TE, TF, TD); T16 = FMA(Tr, Tp, T15); } } } { E TU, TR, TT, T1g, TS; { E T2, T3, T7, T8; T2 = Ip[WS(rs, 1)]; T1m = FMA(TB, TF, T1l); T3 = Im[WS(rs, 1)]; T7 = Rp[WS(rs, 1)]; T8 = Rm[WS(rs, 1)]; { E T1, T4, T9, T6, T5, TQ, T18; T1 = W[2]; TU = T2 + T3; T4 = T2 - T3; TR = T7 - T8; T9 = T7 + T8; T6 = W[3]; T5 = T1 * T4; TQ = W[4]; T18 = T1 * T9; TT = W[5]; Ta = FNMS(T6, T9, T5); T1g = TQ * TU; TS = TQ * TR; T19 = FMA(T6, T4, T18); } } { E Tc, Td, Th, Ti; Tc = Ip[WS(rs, 3)]; T1h = FNMS(TT, TR, T1g); TV = FMA(TT, TU, TS); Td = Im[WS(rs, 3)]; Th = Rp[WS(rs, 3)]; Ti = Rm[WS(rs, 3)]; { E Tb, Te, Tj, Tg, Tf, TW, T1a; Tb = W[10]; T10 = Tc + Td; Te = Tc - Td; TX = Th - Ti; Tj = Th + Ti; Tg = W[11]; Tf = Tb * Te; TW = W[12]; T1a = Tb * Tj; TZ = W[13]; Tk = FNMS(Tg, Tj, Tf); T1i = TW * T10; TY = TW * TX; T1b = FMA(Tg, Te, T1a); } } } { E T1E, T1t, TA, T1s, T1D, T1u, T1e, T13, T1r, T1d; { E TP, T1f, T1q, T12, T17, T1c; { E Tl, T11, Tz, T1p, T1k, T1j; T1E = Ta - Tk; Tl = Ta + Tk; T1j = FNMS(TZ, TX, T1i); T11 = FMA(TZ, T10, TY); Tz = Tv + Ty; T1t = Ty - Tv; T1A = T1o - T1m; T1p = T1m + T1o; T1k = T1h + T1j; T1w = T1j - T1h; T1z = TO - TG; TP = TG + TO; T1f = Tz - Tl; TA = Tl + Tz; T1s = T1k + T1p; T1q = T1k - T1p; T12 = TV + T11; T1x = TV - T11; T1D = T14 - T16; T17 = T14 + T16; T1c = T19 + T1b; T1u = T19 - T1b; } Im[WS(rs, 1)] = KP500000000 * (T1q - T1f); T1e = T12 + TP; T13 = TP - T12; T1r = T17 + T1c; T1d = T17 - T1c; Ip[WS(rs, 2)] = KP500000000 * (T1f + T1q); } Im[WS(rs, 3)] = KP500000000 * (T13 - TA); Ip[0] = KP500000000 * (TA + T13); Rm[WS(rs, 3)] = KP500000000 * (T1r - T1s); Rp[0] = KP500000000 * (T1r + T1s); Rp[WS(rs, 2)] = KP500000000 * (T1d + T1e); Rm[WS(rs, 1)] = KP500000000 * (T1d - T1e); T1H = T1u + T1t; T1v = T1t - T1u; T1L = T1D + T1E; T1F = T1D - T1E; } } { E T1y, T1I, T1B, T1J; T1y = T1w + T1x; T1I = T1w - T1x; T1B = T1z - T1A; T1J = T1z + T1A; { E T1M, T1K, T1C, T1G; T1M = T1I + T1J; T1K = T1I - T1J; T1C = T1y + T1B; T1G = T1B - T1y; Im[0] = -(KP500000000 * (FNMS(KP707106781, T1K, T1H))); Ip[WS(rs, 3)] = KP500000000 * (FMA(KP707106781, T1K, T1H)); Rp[WS(rs, 1)] = KP500000000 * (FMA(KP707106781, T1M, T1L)); Rm[WS(rs, 2)] = KP500000000 * (FNMS(KP707106781, T1M, T1L)); Rp[WS(rs, 3)] = KP500000000 * (FMA(KP707106781, T1G, T1F)); Rm[0] = KP500000000 * (FNMS(KP707106781, T1G, T1F)); Im[WS(rs, 2)] = -(KP500000000 * (FNMS(KP707106781, T1C, T1v))); Ip[WS(rs, 1)] = KP500000000 * (FMA(KP707106781, T1C, T1v)); } } } } } static const tw_instr twinstr[] = { {TW_FULL, 1, 8}, {TW_NEXT, 1, 0} }; static const hc2c_desc desc = { 8, "hc2cfdft_8", twinstr, &GENUS, {60, 30, 22, 0} }; void X(codelet_hc2cfdft_8) (planner *p) { X(khc2c_register) (p, hc2cfdft_8, &desc, HC2C_VIA_DFT); } #else /* HAVE_FMA */ /* Generated by: ../../../genfft/gen_hc2cdft.native -compact -variables 4 -pipeline-latency 4 -n 8 -dit -name hc2cfdft_8 -include hc2cf.h */ /* * This function contains 82 FP additions, 44 FP multiplications, * (or, 68 additions, 30 multiplications, 14 fused multiply/add), * 39 stack variables, 2 constants, and 32 memory accesses */ #include "hc2cf.h" static void hc2cfdft_8(R *Rp, R *Ip, R *Rm, R *Im, const R *W, stride rs, INT mb, INT me, INT ms) { DK(KP353553390, +0.353553390593273762200422181052424519642417969); DK(KP500000000, +0.500000000000000000000000000000000000000000000); { INT m; for (m = mb, W = W + ((mb - 1) * 14); m < me; m = m + 1, Rp = Rp + ms, Ip = Ip + ms, Rm = Rm - ms, Im = Im - ms, W = W + 14, MAKE_VOLATILE_STRIDE(32, rs)) { E Tv, TX, Ts, TY, TE, T1a, TJ, T19, T1l, T1m, T9, T10, Ti, T11, TP; E T16, TU, T17, T1i, T1j; { E Tt, Tu, TD, Tz, TA, TB, Tn, TI, Tr, TG, Tk, To; Tt = Ip[0]; Tu = Im[0]; TD = Tt + Tu; Tz = Rm[0]; TA = Rp[0]; TB = Tz - TA; { E Tl, Tm, Tp, Tq; Tl = Ip[WS(rs, 2)]; Tm = Im[WS(rs, 2)]; Tn = Tl - Tm; TI = Tl + Tm; Tp = Rp[WS(rs, 2)]; Tq = Rm[WS(rs, 2)]; Tr = Tp + Tq; TG = Tp - Tq; } Tv = Tt - Tu; TX = TA + Tz; Tk = W[6]; To = W[7]; Ts = FNMS(To, Tr, Tk * Tn); TY = FMA(Tk, Tr, To * Tn); { E Ty, TC, TF, TH; Ty = W[0]; TC = W[1]; TE = FNMS(TC, TD, Ty * TB); T1a = FMA(TC, TB, Ty * TD); TF = W[8]; TH = W[9]; TJ = FMA(TF, TG, TH * TI); T19 = FNMS(TH, TG, TF * TI); } T1l = TJ + TE; T1m = T1a - T19; } { E T4, TO, T8, TM, Td, TT, Th, TR; { E T2, T3, T6, T7; T2 = Ip[WS(rs, 1)]; T3 = Im[WS(rs, 1)]; T4 = T2 - T3; TO = T2 + T3; T6 = Rp[WS(rs, 1)]; T7 = Rm[WS(rs, 1)]; T8 = T6 + T7; TM = T6 - T7; } { E Tb, Tc, Tf, Tg; Tb = Ip[WS(rs, 3)]; Tc = Im[WS(rs, 3)]; Td = Tb - Tc; TT = Tb + Tc; Tf = Rp[WS(rs, 3)]; Tg = Rm[WS(rs, 3)]; Th = Tf + Tg; TR = Tf - Tg; } { E T1, T5, Ta, Te; T1 = W[2]; T5 = W[3]; T9 = FNMS(T5, T8, T1 * T4); T10 = FMA(T1, T8, T5 * T4); Ta = W[10]; Te = W[11]; Ti = FNMS(Te, Th, Ta * Td); T11 = FMA(Ta, Th, Te * Td); { E TL, TN, TQ, TS; TL = W[4]; TN = W[5]; TP = FMA(TL, TM, TN * TO); T16 = FNMS(TN, TM, TL * TO); TQ = W[12]; TS = W[13]; TU = FMA(TQ, TR, TS * TT); T17 = FNMS(TS, TR, TQ * TT); } T1i = T17 - T16; T1j = TP - TU; } } { E T1h, T1t, T1w, T1y, T1o, T1s, T1r, T1x; { E T1f, T1g, T1u, T1v; T1f = Tv - Ts; T1g = T10 - T11; T1h = KP500000000 * (T1f - T1g); T1t = KP500000000 * (T1g + T1f); T1u = T1i - T1j; T1v = T1l + T1m; T1w = KP353553390 * (T1u - T1v); T1y = KP353553390 * (T1u + T1v); } { E T1k, T1n, T1p, T1q; T1k = T1i + T1j; T1n = T1l - T1m; T1o = KP353553390 * (T1k + T1n); T1s = KP353553390 * (T1n - T1k); T1p = TX - TY; T1q = T9 - Ti; T1r = KP500000000 * (T1p - T1q); T1x = KP500000000 * (T1p + T1q); } Ip[WS(rs, 1)] = T1h + T1o; Rp[WS(rs, 1)] = T1x + T1y; Im[WS(rs, 2)] = T1o - T1h; Rm[WS(rs, 2)] = T1x - T1y; Rm[0] = T1r - T1s; Im[0] = T1w - T1t; Rp[WS(rs, 3)] = T1r + T1s; Ip[WS(rs, 3)] = T1t + T1w; } { E Tx, T15, T1c, T1e, TW, T14, T13, T1d; { E Tj, Tw, T18, T1b; Tj = T9 + Ti; Tw = Ts + Tv; Tx = Tj + Tw; T15 = Tw - Tj; T18 = T16 + T17; T1b = T19 + T1a; T1c = T18 - T1b; T1e = T18 + T1b; } { E TK, TV, TZ, T12; TK = TE - TJ; TV = TP + TU; TW = TK - TV; T14 = TV + TK; TZ = TX + TY; T12 = T10 + T11; T13 = TZ - T12; T1d = TZ + T12; } Ip[0] = KP500000000 * (Tx + TW); Rp[0] = KP500000000 * (T1d + T1e); Im[WS(rs, 3)] = KP500000000 * (TW - Tx); Rm[WS(rs, 3)] = KP500000000 * (T1d - T1e); Rm[WS(rs, 1)] = KP500000000 * (T13 - T14); Im[WS(rs, 1)] = KP500000000 * (T1c - T15); Rp[WS(rs, 2)] = KP500000000 * (T13 + T14); Ip[WS(rs, 2)] = KP500000000 * (T15 + T1c); } } } } static const tw_instr twinstr[] = { {TW_FULL, 1, 8}, {TW_NEXT, 1, 0} }; static const hc2c_desc desc = { 8, "hc2cfdft_8", twinstr, &GENUS, {68, 30, 14, 0} }; void X(codelet_hc2cfdft_8) (planner *p) { X(khc2c_register) (p, hc2cfdft_8, &desc, HC2C_VIA_DFT); } #endif /* HAVE_FMA */ fftw-3.3.3/rdft/scalar/r2cf/hc2cfdft_10.c0000644000175000017500000003710412054410715014602 00000000000000/* * Copyright (c) 2003, 2007-11 Matteo Frigo * Copyright (c) 2003, 2007-11 Massachusetts Institute of Technology * * This program is free software; you can redistribute it and/or modify * it under the terms of the GNU General Public License as published by * the Free Software Foundation; either version 2 of the License, or * (at your option) any later version. * * This program is distributed in the hope that it will be useful, * but WITHOUT ANY WARRANTY; without even the implied warranty of * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the * GNU General Public License for more details. * * You should have received a copy of the GNU General Public License * along with this program; if not, write to the Free Software * Foundation, Inc., 51 Franklin Street, Fifth Floor, Boston, MA 02110-1301 USA * */ /* This file was automatically generated --- DO NOT EDIT */ /* Generated on Sun Nov 25 07:40:45 EST 2012 */ #include "codelet-rdft.h" #ifdef HAVE_FMA /* Generated by: ../../../genfft/gen_hc2cdft.native -fma -reorder-insns -schedule-for-pipeline -compact -variables 4 -pipeline-latency 4 -n 10 -dit -name hc2cfdft_10 -include hc2cf.h */ /* * This function contains 122 FP additions, 92 FP multiplications, * (or, 68 additions, 38 multiplications, 54 fused multiply/add), * 94 stack variables, 5 constants, and 40 memory accesses */ #include "hc2cf.h" static void hc2cfdft_10(R *Rp, R *Ip, R *Rm, R *Im, const R *W, stride rs, INT mb, INT me, INT ms) { DK(KP951056516, +0.951056516295153572116439333379382143405698634); DK(KP559016994, +0.559016994374947424102293417182819058860154590); DK(KP250000000, +0.250000000000000000000000000000000000000000000); DK(KP500000000, +0.500000000000000000000000000000000000000000000); DK(KP618033988, +0.618033988749894848204586834365638117720309180); { INT m; for (m = mb, W = W + ((mb - 1) * 18); m < me; m = m + 1, Rp = Rp + ms, Ip = Ip + ms, Rm = Rm - ms, Im = Im - ms, W = W + 18, MAKE_VOLATILE_STRIDE(40, rs)) { E T1x, T1I, T1T, T22, T20; { E T3, T1u, T1S, T2f, Td, T1w, T14, T1p, T1j, T1q, T1N, T2e, T1z, To, T2i; E T1H, TQ, T1n, Ty, T1B; { E T1h, TW, Tc, T1b, T1g, T1f, T1Q, TV, T7, TS, T1J, TU, Ts, T19, T18; E T15, Tx, T17, T1O, T1A, Tt, TD, Ti, TE, Tn, TA, T1F, TC, T1y, Tj; E T11, T12, TJ, TZ, TO, TY, TG, T1L, T1e, T1, T2; T1 = Ip[0]; T2 = Im[0]; { E Ta, Tb, T1c, T1d; Ta = Rp[WS(rs, 2)]; Tb = Rm[WS(rs, 2)]; T1c = Rm[0]; T1h = T1 + T2; T3 = T1 - T2; T1d = Rp[0]; TW = Ta + Tb; Tc = Ta - Tb; T1b = W[0]; T1u = T1d + T1c; T1e = T1c - T1d; T1g = W[1]; } { E T16, Tp, TT, T5, T6, TB, Tf; T5 = Ip[WS(rs, 2)]; T6 = Im[WS(rs, 2)]; T1f = T1b * T1e; T1Q = T1g * T1e; TV = W[7]; T7 = T5 + T6; TT = T5 - T6; TS = W[6]; { E Tv, Tw, Tq, Tr; Tq = Rm[WS(rs, 3)]; Tr = Rp[WS(rs, 3)]; T1J = TV * TT; TU = TS * TT; Tv = Ip[WS(rs, 3)]; Ts = Tq - Tr; T19 = Tr + Tq; Tw = Im[WS(rs, 3)]; T18 = W[11]; T15 = W[10]; Tx = Tv + Tw; T16 = Tv - Tw; Tp = W[12]; } { E Tg, Th, Tl, Tm; Tg = Ip[WS(rs, 1)]; T17 = T15 * T16; T1O = T18 * T16; T1A = Tp * Tx; Tt = Tp * Ts; Th = Im[WS(rs, 1)]; Tl = Rp[WS(rs, 1)]; Tm = Rm[WS(rs, 1)]; TD = W[5]; Ti = Tg - Th; TE = Tg + Th; Tn = Tl + Tm; TB = Tm - Tl; TA = W[4]; Tf = W[2]; T1F = TD * TB; } { E TH, TI, TM, TN; TH = Ip[WS(rs, 4)]; TC = TA * TB; T1y = Tf * Tn; Tj = Tf * Ti; TI = Im[WS(rs, 4)]; TM = Rp[WS(rs, 4)]; TN = Rm[WS(rs, 4)]; T11 = W[17]; T12 = TH + TI; TJ = TH - TI; TZ = TN - TM; TO = TM + TN; TY = W[16]; TG = W[14]; T1L = T11 * TZ; } } { E T10, T1D, TK, T4, T9, T1P, T1R, T8, T1v; T10 = TY * TZ; T1D = TG * TO; TK = TG * TJ; T4 = W[9]; T9 = W[8]; T1P = FMA(T15, T19, T1O); T1R = FMA(T1b, T1h, T1Q); T8 = T4 * T7; T1v = T9 * T7; { E TX, T13, T1a, T1i; TX = FNMS(TV, TW, TU); T1S = T1P - T1R; T2f = T1P + T1R; Td = FMA(T9, Tc, T8); T1w = FNMS(T4, Tc, T1v); T13 = FNMS(T11, T12, T10); T1a = FNMS(T18, T19, T17); T1i = FNMS(T1g, T1h, T1f); { E T1K, T1M, TF, T1G, TL; T1K = FMA(TS, TW, T1J); T14 = TX + T13; T1p = T13 - TX; T1j = T1a + T1i; T1q = T1i - T1a; T1M = FMA(TY, T12, T1L); TF = FNMS(TD, TE, TC); T1G = FMA(TA, TE, T1F); TL = W[15]; T1N = T1K - T1M; T2e = T1K + T1M; { E Tk, T1E, TP, Tu; Tk = W[3]; T1E = FMA(TL, TJ, T1D); TP = FNMS(TL, TO, TK); Tu = W[13]; T1z = FMA(Tk, Ti, T1y); To = FNMS(Tk, Tn, Tj); T2i = T1G + T1E; T1H = T1E - T1G; TQ = TF + TP; T1n = TF - TP; Ty = FNMS(Tu, Tx, Tt); T1B = FMA(Tu, Ts, T1A); } } } } } { E T2p, T1t, T1m, T1C, T2o, T2m, T2k, T2w, T2y, T2n, T2d, T2l; { E T2g, Te, T2h, T2u, T1k, TR, T2v, Tz; T2p = T2e + T2f; T2g = T2e - T2f; Te = T3 - Td; T1t = Td + T3; Tz = To + Ty; T1m = Ty - To; T2h = T1z + T1B; T1C = T1z - T1B; T2u = T14 - T1j; T1k = T14 + T1j; TR = Tz + TQ; T2v = Tz - TQ; { E T2c, T2b, T2j, T1l; T2j = T2h - T2i; T2o = T2h + T2i; T2c = TR - T1k; T1l = TR + T1k; T2m = FMA(KP618033988, T2g, T2j); T2k = FNMS(KP618033988, T2j, T2g); T2w = FNMS(KP618033988, T2v, T2u); T2y = FMA(KP618033988, T2u, T2v); Ip[0] = KP500000000 * (Te + T1l); T2b = FNMS(KP250000000, T1l, Te); T2n = T1u + T1w; T1x = T1u - T1w; T2d = FNMS(KP559016994, T2c, T2b); T2l = FMA(KP559016994, T2c, T2b); } } { E T1o, T1Y, T28, T2a, T1Z, T1r, T2t, T2x; { E T26, T2s, T2q, T27, T2r; T1I = T1C + T1H; T26 = T1H - T1C; Im[WS(rs, 1)] = -(KP500000000 * (FNMS(KP951056516, T2k, T2d))); Ip[WS(rs, 2)] = KP500000000 * (FMA(KP951056516, T2k, T2d)); Im[WS(rs, 3)] = -(KP500000000 * (FNMS(KP951056516, T2m, T2l))); Ip[WS(rs, 4)] = KP500000000 * (FMA(KP951056516, T2m, T2l)); T2s = T2o - T2p; T2q = T2o + T2p; T27 = T1S - T1N; T1T = T1N + T1S; T1o = T1m + T1n; T1Y = T1n - T1m; Rp[0] = KP500000000 * (T2n + T2q); T2r = FNMS(KP250000000, T2q, T2n); T28 = FMA(KP618033988, T27, T26); T2a = FNMS(KP618033988, T26, T27); T1Z = T1q - T1p; T1r = T1p + T1q; T2t = FNMS(KP559016994, T2s, T2r); T2x = FMA(KP559016994, T2s, T2r); } { E T24, T23, T1s, T25, T29; T1s = T1o + T1r; T24 = T1r - T1o; Rm[WS(rs, 1)] = KP500000000 * (FMA(KP951056516, T2w, T2t)); Rp[WS(rs, 2)] = KP500000000 * (FNMS(KP951056516, T2w, T2t)); Rm[WS(rs, 3)] = KP500000000 * (FMA(KP951056516, T2y, T2x)); Rp[WS(rs, 4)] = KP500000000 * (FNMS(KP951056516, T2y, T2x)); Im[WS(rs, 4)] = KP500000000 * (T1s - T1t); T23 = FMA(KP250000000, T1s, T1t); T25 = FMA(KP559016994, T24, T23); T29 = FNMS(KP559016994, T24, T23); T22 = FNMS(KP618033988, T1Y, T1Z); T20 = FMA(KP618033988, T1Z, T1Y); Im[0] = -(KP500000000 * (FNMS(KP951056516, T28, T25))); Ip[WS(rs, 1)] = KP500000000 * (FMA(KP951056516, T28, T25)); Im[WS(rs, 2)] = -(KP500000000 * (FNMS(KP951056516, T2a, T29))); Ip[WS(rs, 3)] = KP500000000 * (FMA(KP951056516, T2a, T29)); } } } } { E T1U, T1W, T1V, T21, T1X; T1U = T1I + T1T; T1W = T1I - T1T; Rm[WS(rs, 4)] = KP500000000 * (T1x + T1U); T1V = FNMS(KP250000000, T1U, T1x); T21 = FNMS(KP559016994, T1W, T1V); T1X = FMA(KP559016994, T1W, T1V); Rm[0] = KP500000000 * (FNMS(KP951056516, T20, T1X)); Rp[WS(rs, 1)] = KP500000000 * (FMA(KP951056516, T20, T1X)); Rm[WS(rs, 2)] = KP500000000 * (FNMS(KP951056516, T22, T21)); Rp[WS(rs, 3)] = KP500000000 * (FMA(KP951056516, T22, T21)); } } } } static const tw_instr twinstr[] = { {TW_FULL, 1, 10}, {TW_NEXT, 1, 0} }; static const hc2c_desc desc = { 10, "hc2cfdft_10", twinstr, &GENUS, {68, 38, 54, 0} }; void X(codelet_hc2cfdft_10) (planner *p) { X(khc2c_register) (p, hc2cfdft_10, &desc, HC2C_VIA_DFT); } #else /* HAVE_FMA */ /* Generated by: ../../../genfft/gen_hc2cdft.native -compact -variables 4 -pipeline-latency 4 -n 10 -dit -name hc2cfdft_10 -include hc2cf.h */ /* * This function contains 122 FP additions, 68 FP multiplications, * (or, 92 additions, 38 multiplications, 30 fused multiply/add), * 62 stack variables, 5 constants, and 40 memory accesses */ #include "hc2cf.h" static void hc2cfdft_10(R *Rp, R *Ip, R *Rm, R *Im, const R *W, stride rs, INT mb, INT me, INT ms) { DK(KP293892626, +0.293892626146236564584352977319536384298826219); DK(KP475528258, +0.475528258147576786058219666689691071702849317); DK(KP125000000, +0.125000000000000000000000000000000000000000000); DK(KP500000000, +0.500000000000000000000000000000000000000000000); DK(KP279508497, +0.279508497187473712051146708591409529430077295); { INT m; for (m = mb, W = W + ((mb - 1) * 18); m < me; m = m + 1, Rp = Rp + ms, Ip = Ip + ms, Rm = Rm - ms, Im = Im - ms, W = W + 18, MAKE_VOLATILE_STRIDE(40, rs)) { E Tw, TL, TM, T1W, T1X, T27, T1Z, T20, T26, TX, T1a, T1b, T1d, T1e, T1f; E T1q, T1t, T1u, T1x, T1A, T1B, T1g, T1h, T1i, Td, T25, T1k, T1F; { E T3, T1D, T19, T1z, T7, Tb, TR, T1v, Tm, T1o, TK, T1s, Tv, T1p, T12; E T1y, TF, T1r, TW, T1w; { E T1, T2, T18, T14, T15, T16, T13, T17; T1 = Ip[0]; T2 = Im[0]; T18 = T1 + T2; T14 = Rm[0]; T15 = Rp[0]; T16 = T14 - T15; T3 = T1 - T2; T1D = T15 + T14; T13 = W[0]; T17 = W[1]; T19 = FNMS(T17, T18, T13 * T16); T1z = FMA(T17, T16, T13 * T18); } { E T5, T6, TO, T9, Ta, TQ, TN, TP; T5 = Ip[WS(rs, 2)]; T6 = Im[WS(rs, 2)]; TO = T5 - T6; T9 = Rp[WS(rs, 2)]; Ta = Rm[WS(rs, 2)]; TQ = T9 + Ta; T7 = T5 + T6; Tb = T9 - Ta; TN = W[6]; TP = W[7]; TR = FNMS(TP, TQ, TN * TO); T1v = FMA(TP, TO, TN * TQ); } { E Th, TJ, Tl, TH; { E Tf, Tg, Tj, Tk; Tf = Ip[WS(rs, 1)]; Tg = Im[WS(rs, 1)]; Th = Tf - Tg; TJ = Tf + Tg; Tj = Rp[WS(rs, 1)]; Tk = Rm[WS(rs, 1)]; Tl = Tj + Tk; TH = Tj - Tk; } { E Te, Ti, TG, TI; Te = W[2]; Ti = W[3]; Tm = FNMS(Ti, Tl, Te * Th); T1o = FMA(Te, Tl, Ti * Th); TG = W[4]; TI = W[5]; TK = FMA(TG, TH, TI * TJ); T1s = FNMS(TI, TH, TG * TJ); } } { E Tq, TZ, Tu, T11; { E To, Tp, Ts, Tt; To = Ip[WS(rs, 3)]; Tp = Im[WS(rs, 3)]; Tq = To + Tp; TZ = To - Tp; Ts = Rp[WS(rs, 3)]; Tt = Rm[WS(rs, 3)]; Tu = Ts - Tt; T11 = Ts + Tt; } { E Tn, Tr, TY, T10; Tn = W[13]; Tr = W[12]; Tv = FMA(Tn, Tq, Tr * Tu); T1p = FNMS(Tn, Tu, Tr * Tq); TY = W[10]; T10 = W[11]; T12 = FNMS(T10, T11, TY * TZ); T1y = FMA(T10, TZ, TY * T11); } } { E TA, TV, TE, TT; { E Ty, Tz, TC, TD; Ty = Ip[WS(rs, 4)]; Tz = Im[WS(rs, 4)]; TA = Ty - Tz; TV = Ty + Tz; TC = Rp[WS(rs, 4)]; TD = Rm[WS(rs, 4)]; TE = TC + TD; TT = TC - TD; } { E Tx, TB, TS, TU; Tx = W[14]; TB = W[15]; TF = FNMS(TB, TE, Tx * TA); T1r = FMA(Tx, TE, TB * TA); TS = W[16]; TU = W[17]; TW = FMA(TS, TT, TU * TV); T1w = FNMS(TU, TT, TS * TV); } } Tw = Tm - Tv; TL = TF - TK; TM = Tw + TL; T1W = T1v + T1w; T1X = T1y + T1z; T27 = T1W + T1X; T1Z = T1o + T1p; T20 = T1s + T1r; T26 = T1Z + T20; TX = TR - TW; T1a = T12 + T19; T1b = TX + T1a; T1d = T19 - T12; T1e = TR + TW; T1f = T1d - T1e; T1q = T1o - T1p; T1t = T1r - T1s; T1u = T1q + T1t; T1x = T1v - T1w; T1A = T1y - T1z; T1B = T1x + T1A; T1g = Tm + Tv; T1h = TK + TF; T1i = T1g + T1h; { E Tc, T1E, T4, T8; T4 = W[9]; T8 = W[8]; Tc = FMA(T4, T7, T8 * Tb); T1E = FNMS(T4, Tb, T8 * T7); Td = T3 - Tc; T25 = T1D + T1E; T1k = Tc + T3; T1F = T1D - T1E; } } { E T1U, T1c, T1T, T22, T24, T1Y, T21, T23, T1V; T1U = KP279508497 * (TM - T1b); T1c = TM + T1b; T1T = FNMS(KP125000000, T1c, KP500000000 * Td); T1Y = T1W - T1X; T21 = T1Z - T20; T22 = FNMS(KP293892626, T21, KP475528258 * T1Y); T24 = FMA(KP475528258, T21, KP293892626 * T1Y); Ip[0] = KP500000000 * (Td + T1c); T23 = T1U + T1T; Ip[WS(rs, 4)] = T23 + T24; Im[WS(rs, 3)] = T24 - T23; T1V = T1T - T1U; Ip[WS(rs, 2)] = T1V + T22; Im[WS(rs, 1)] = T22 - T1V; } { E T2a, T28, T29, T2e, T2g, T2c, T2d, T2f, T2b; T2a = KP279508497 * (T26 - T27); T28 = T26 + T27; T29 = FNMS(KP125000000, T28, KP500000000 * T25); T2c = TX - T1a; T2d = Tw - TL; T2e = FNMS(KP293892626, T2d, KP475528258 * T2c); T2g = FMA(KP475528258, T2d, KP293892626 * T2c); Rp[0] = KP500000000 * (T25 + T28); T2f = T2a + T29; Rp[WS(rs, 4)] = T2f - T2g; Rm[WS(rs, 3)] = T2g + T2f; T2b = T29 - T2a; Rp[WS(rs, 2)] = T2b - T2e; Rm[WS(rs, 1)] = T2e + T2b; } { E T1M, T1j, T1L, T1Q, T1S, T1O, T1P, T1R, T1N; T1M = KP279508497 * (T1i + T1f); T1j = T1f - T1i; T1L = FMA(KP500000000, T1k, KP125000000 * T1j); T1O = T1A - T1x; T1P = T1q - T1t; T1Q = FNMS(KP475528258, T1P, KP293892626 * T1O); T1S = FMA(KP293892626, T1P, KP475528258 * T1O); Im[WS(rs, 4)] = KP500000000 * (T1j - T1k); T1R = T1L - T1M; Ip[WS(rs, 3)] = T1R + T1S; Im[WS(rs, 2)] = T1S - T1R; T1N = T1L + T1M; Ip[WS(rs, 1)] = T1N + T1Q; Im[0] = T1Q - T1N; } { E T1C, T1G, T1H, T1n, T1J, T1l, T1m, T1K, T1I; T1C = KP279508497 * (T1u - T1B); T1G = T1u + T1B; T1H = FNMS(KP125000000, T1G, KP500000000 * T1F); T1l = T1g - T1h; T1m = T1e + T1d; T1n = FMA(KP475528258, T1l, KP293892626 * T1m); T1J = FNMS(KP293892626, T1l, KP475528258 * T1m); Rm[WS(rs, 4)] = KP500000000 * (T1F + T1G); T1K = T1H - T1C; Rp[WS(rs, 3)] = T1J + T1K; Rm[WS(rs, 2)] = T1K - T1J; T1I = T1C + T1H; Rp[WS(rs, 1)] = T1n + T1I; Rm[0] = T1I - T1n; } } } } static const tw_instr twinstr[] = { {TW_FULL, 1, 10}, {TW_NEXT, 1, 0} }; static const hc2c_desc desc = { 10, "hc2cfdft_10", twinstr, &GENUS, {92, 38, 30, 0} }; void X(codelet_hc2cfdft_10) (planner *p) { X(khc2c_register) (p, hc2cfdft_10, &desc, HC2C_VIA_DFT); } #endif /* HAVE_FMA */ fftw-3.3.3/rdft/scalar/r2cf/hc2cfdft_12.c0000644000175000017500000004173712054410716014614 00000000000000/* * Copyright (c) 2003, 2007-11 Matteo Frigo * Copyright (c) 2003, 2007-11 Massachusetts Institute of Technology * * This program is free software; you can redistribute it and/or modify * it under the terms of the GNU General Public License as published by * the Free Software Foundation; either version 2 of the License, or * (at your option) any later version. * * This program is distributed in the hope that it will be useful, * but WITHOUT ANY WARRANTY; without even the implied warranty of * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the * GNU General Public License for more details. * * You should have received a copy of the GNU General Public License * along with this program; if not, write to the Free Software * Foundation, Inc., 51 Franklin Street, Fifth Floor, Boston, MA 02110-1301 USA * */ /* This file was automatically generated --- DO NOT EDIT */ /* Generated on Sun Nov 25 07:40:45 EST 2012 */ #include "codelet-rdft.h" #ifdef HAVE_FMA /* Generated by: ../../../genfft/gen_hc2cdft.native -fma -reorder-insns -schedule-for-pipeline -compact -variables 4 -pipeline-latency 4 -n 12 -dit -name hc2cfdft_12 -include hc2cf.h */ /* * This function contains 142 FP additions, 92 FP multiplications, * (or, 96 additions, 46 multiplications, 46 fused multiply/add), * 71 stack variables, 2 constants, and 48 memory accesses */ #include "hc2cf.h" static void hc2cfdft_12(R *Rp, R *Ip, R *Rm, R *Im, const R *W, stride rs, INT mb, INT me, INT ms) { DK(KP866025403, +0.866025403784438646763723170752936183471402627); DK(KP500000000, +0.500000000000000000000000000000000000000000000); { INT m; for (m = mb, W = W + ((mb - 1) * 22); m < me; m = m + 1, Rp = Rp + ms, Ip = Ip + ms, Rm = Rm - ms, Im = Im - ms, W = W + 22, MAKE_VOLATILE_STRIDE(48, rs)) { E T2z, T2M; { E To, T1E, T2H, T1m, T1W, Tl, T1J, T2i, T2K, T1B, T2I, T2e, T19, T2E, T2C; E T27, T1M, Tz, T2B, T1f, T1O, TJ, TT, T1Q; { E T2b, T1s, T1A, T2d; { E T1u, T1z, T1v, T2c, T1i, Te, T1l, Tj, Tf, T1H, T4, T1o, T1, T1r, T9; E T1n, T5; { E T1x, T1y, T1t, Tm, Tn; Tm = Ip[0]; Tn = Im[0]; T1x = Rp[0]; T1y = Rm[0]; T1t = W[0]; T1u = Tm + Tn; To = Tm - Tn; { E Th, Ti, Tb, Tc, Td; Tc = Ip[WS(rs, 4)]; T1z = T1x - T1y; T1E = T1x + T1y; Td = Im[WS(rs, 4)]; T1v = T1t * T1u; Th = Rp[WS(rs, 4)]; T2c = T1t * T1z; T1i = Tc + Td; Te = Tc - Td; Ti = Rm[WS(rs, 4)]; Tb = W[14]; { E T7, T8, T2, T3; T2 = Ip[WS(rs, 2)]; T1l = Th - Ti; Tj = Th + Ti; Tf = Tb * Te; T3 = Im[WS(rs, 2)]; T7 = Rp[WS(rs, 2)]; T1H = Tb * Tj; T8 = Rm[WS(rs, 2)]; T4 = T2 - T3; T1o = T2 + T3; T1 = W[6]; T1r = T7 - T8; T9 = T7 + T8; T1n = W[8]; T5 = T1 * T4; } } } { E T1F, T2a, T1p, T1h, T1k; T1F = T1 * T9; T2a = T1n * T1r; T1p = T1n * T1o; T1h = W[16]; T1k = W[17]; { E T1G, Ta, Tk, T1I, T1q, T1w; { E T6, Tg, T2G, T1j; T6 = W[7]; Tg = W[15]; T2G = T1h * T1l; T1j = T1h * T1i; T1G = FMA(T6, T4, T1F); Ta = FNMS(T6, T9, T5); T2H = FMA(T1k, T1i, T2G); T1m = FNMS(T1k, T1l, T1j); Tk = FNMS(Tg, Tj, Tf); T1I = FMA(Tg, Te, T1H); } T1q = W[9]; T1w = W[1]; T1W = Ta - Tk; Tl = Ta + Tk; T1J = T1G + T1I; T2i = T1I - T1G; T2b = FMA(T1q, T1o, T2a); T1s = FNMS(T1q, T1r, T1p); T1A = FNMS(T1w, T1z, T1v); T2d = FMA(T1w, T1u, T2c); } } } { E T11, Tt, T10, TX, Ty, TZ, T23, T1b, TN, TS, T1e, T1P, TO, T17, TD; E T16, T13, T14, TI, TA; { E Tw, Tx, Tr, Ts, TK; Tr = Ip[WS(rs, 3)]; Ts = Im[WS(rs, 3)]; T2K = T1s - T1A; T1B = T1s + T1A; T2I = T2b + T2d; T2e = T2b - T2d; Tw = Rp[WS(rs, 3)]; T11 = Tr + Ts; Tt = Tr - Ts; Tx = Rm[WS(rs, 3)]; T10 = W[12]; TX = W[13]; { E TL, TY, TM, TQ, TR; TL = Ip[WS(rs, 1)]; Ty = Tw + Tx; TY = Tx - Tw; TM = Im[WS(rs, 1)]; TQ = Rp[WS(rs, 1)]; TR = Rm[WS(rs, 1)]; TZ = TX * TY; T23 = T10 * TY; T1b = TL + TM; TN = TL - TM; TS = TQ + TR; T1e = TQ - TR; } TK = W[2]; { E TG, TH, TB, TC; TB = Ip[WS(rs, 5)]; TC = Im[WS(rs, 5)]; TG = Rp[WS(rs, 5)]; T1P = TK * TS; TO = TK * TN; T17 = TB + TC; TD = TB - TC; TH = Rm[WS(rs, 5)]; T16 = W[20]; T13 = W[21]; T14 = TH - TG; TI = TG + TH; TA = W[18]; } } { E T12, T1N, TE, T18, T24, T26, T25, T15; T12 = FMA(T10, T11, TZ); T15 = T13 * T14; T25 = T16 * T14; T1N = TA * TI; TE = TA * TD; T18 = FMA(T16, T17, T15); T24 = FNMS(TX, T11, T23); T26 = FNMS(T13, T17, T25); { E Tv, T1L, Tu, Tq; Tq = W[10]; T19 = T12 + T18; T2E = T18 - T12; Tv = W[11]; T2C = T24 + T26; T27 = T24 - T26; T1L = Tq * Ty; Tu = Tq * Tt; { E T1d, T2A, T1c, T1a, TF, TP; T1a = W[4]; T1d = W[5]; T1M = FMA(Tv, Tt, T1L); Tz = FNMS(Tv, Ty, Tu); T2A = T1a * T1e; T1c = T1a * T1b; TF = W[19]; TP = W[3]; T2B = FMA(T1d, T1b, T2A); T1f = FNMS(T1d, T1e, T1c); T1O = FMA(TF, TD, T1N); TJ = FNMS(TF, TI, TE); TT = FNMS(TP, TS, TO); T1Q = FMA(TP, TN, T1P); } } } } } { E T2h, T2D, T1Z, T2l, T2J, T22, T2k, T29, T30, T1U, T1V, T1Y, T2Z, T1T; { E T2Y, TW, T2V, T1D, T1K, T1S; { E Tp, T2W, TU, T1R, T2X, T1g, TV, T1C; T2h = FNMS(KP500000000, Tl, To); Tp = Tl + To; T2W = T2C - T2B; T2D = FMA(KP500000000, T2C, T2B); T1Z = TJ - TT; TU = TJ + TT; T1R = T1O + T1Q; T2l = T1Q - T1O; T2J = FNMS(KP500000000, T2I, T2H); T2X = T2H + T2I; T1g = T19 + T1f; T22 = FNMS(KP500000000, T19, T1f); T2k = FNMS(KP500000000, TU, Tz); TV = Tz + TU; T1C = T1m + T1B; T29 = FNMS(KP500000000, T1B, T1m); T2Y = T2W - T2X; T30 = T2W + T2X; TW = Tp - TV; T2V = TV + Tp; T1U = T1g + T1C; T1D = T1g - T1C; T1V = FNMS(KP500000000, T1J, T1E); T1K = T1E + T1J; T1S = T1M + T1R; T1Y = FNMS(KP500000000, T1R, T1M); } Ip[WS(rs, 3)] = KP500000000 * (TW + T1D); Im[WS(rs, 2)] = KP500000000 * (T1D - TW); Im[WS(rs, 5)] = KP500000000 * (T2Y - T2V); T2Z = T1K - T1S; T1T = T1K + T1S; Ip[0] = KP500000000 * (T2V + T2Y); } { E T2v, T1X, T2Q, T2F, T2R, T2L, T2w, T20, T2t, T28, T2p, T2j; Rm[WS(rs, 2)] = KP500000000 * (T2Z + T30); Rp[WS(rs, 3)] = KP500000000 * (T2Z - T30); Rp[0] = KP500000000 * (T1T + T1U); Rm[WS(rs, 5)] = KP500000000 * (T1T - T1U); T2v = FMA(KP866025403, T1W, T1V); T1X = FNMS(KP866025403, T1W, T1V); T2Q = FMA(KP866025403, T2E, T2D); T2F = FNMS(KP866025403, T2E, T2D); T2R = FMA(KP866025403, T2K, T2J); T2L = FNMS(KP866025403, T2K, T2J); T2w = FMA(KP866025403, T1Z, T1Y); T20 = FNMS(KP866025403, T1Z, T1Y); T2t = FMA(KP866025403, T27, T22); T28 = FNMS(KP866025403, T27, T22); T2p = FMA(KP866025403, T2i, T2h); T2j = FNMS(KP866025403, T2i, T2h); { E T2T, T2q, T2s, T2U; { E T21, T2f, T2S, T2n, T2P, T2m, T2o, T2g; T2T = T1X - T20; T21 = T1X + T20; T2q = FMA(KP866025403, T2l, T2k); T2m = FNMS(KP866025403, T2l, T2k); T2s = FMA(KP866025403, T2e, T29); T2f = FNMS(KP866025403, T2e, T29); T2S = T2Q + T2R; T2U = T2R - T2Q; T2n = T2j - T2m; T2P = T2m + T2j; T2o = T2f - T28; T2g = T28 + T2f; Im[WS(rs, 3)] = KP500000000 * (T2S - T2P); Ip[WS(rs, 2)] = KP500000000 * (T2P + T2S); Rm[WS(rs, 3)] = KP500000000 * (T21 + T2g); Rp[WS(rs, 2)] = KP500000000 * (T21 - T2g); Ip[WS(rs, 5)] = KP500000000 * (T2n + T2o); Im[0] = KP500000000 * (T2o - T2n); } { E T2y, T2x, T2N, T2O, T2r, T2u; T2z = T2q + T2p; T2r = T2p - T2q; T2u = T2s - T2t; T2y = T2t + T2s; T2x = T2v + T2w; T2N = T2v - T2w; Rp[WS(rs, 5)] = KP500000000 * (T2T + T2U); Rm[0] = KP500000000 * (T2T - T2U); Im[WS(rs, 4)] = KP500000000 * (T2u - T2r); Ip[WS(rs, 1)] = KP500000000 * (T2r + T2u); T2O = T2L - T2F; T2M = T2F + T2L; Rp[WS(rs, 1)] = KP500000000 * (T2N + T2O); Rm[WS(rs, 4)] = KP500000000 * (T2N - T2O); Rp[WS(rs, 4)] = KP500000000 * (T2x + T2y); Rm[WS(rs, 1)] = KP500000000 * (T2x - T2y); } } } } } Im[WS(rs, 1)] = -(KP500000000 * (T2z + T2M)); Ip[WS(rs, 4)] = KP500000000 * (T2z - T2M); } } } static const tw_instr twinstr[] = { {TW_FULL, 1, 12}, {TW_NEXT, 1, 0} }; static const hc2c_desc desc = { 12, "hc2cfdft_12", twinstr, &GENUS, {96, 46, 46, 0} }; void X(codelet_hc2cfdft_12) (planner *p) { X(khc2c_register) (p, hc2cfdft_12, &desc, HC2C_VIA_DFT); } #else /* HAVE_FMA */ /* Generated by: ../../../genfft/gen_hc2cdft.native -compact -variables 4 -pipeline-latency 4 -n 12 -dit -name hc2cfdft_12 -include hc2cf.h */ /* * This function contains 142 FP additions, 76 FP multiplications, * (or, 112 additions, 46 multiplications, 30 fused multiply/add), * 52 stack variables, 3 constants, and 48 memory accesses */ #include "hc2cf.h" static void hc2cfdft_12(R *Rp, R *Ip, R *Rm, R *Im, const R *W, stride rs, INT mb, INT me, INT ms) { DK(KP250000000, +0.250000000000000000000000000000000000000000000); DK(KP500000000, +0.500000000000000000000000000000000000000000000); DK(KP433012701, +0.433012701892219323381861585376468091735701313); { INT m; for (m = mb, W = W + ((mb - 1) * 22); m < me; m = m + 1, Rp = Rp + ms, Ip = Ip + ms, Rm = Rm - ms, Im = Im - ms, W = W + 22, MAKE_VOLATILE_STRIDE(48, rs)) { E Tm, T1t, T1d, T2j, Tj, T1Y, T1w, T1G, T1q, T2q, T1U, T2k, Tw, T1y, T17; E T2g, TP, T21, T1B, T1J, T12, T2u, T1P, T2h; { E Tk, Tl, T1k, T1m, T1n, T1o, T4, T1f, T8, T1h, Th, T1c, Td, T1a, T19; E T1b; { E T2, T3, T6, T7; Tk = Ip[0]; Tl = Im[0]; T1k = Tk + Tl; T1m = Rp[0]; T1n = Rm[0]; T1o = T1m - T1n; T2 = Ip[WS(rs, 2)]; T3 = Im[WS(rs, 2)]; T4 = T2 - T3; T1f = T2 + T3; T6 = Rp[WS(rs, 2)]; T7 = Rm[WS(rs, 2)]; T8 = T6 + T7; T1h = T6 - T7; { E Tf, Tg, Tb, Tc; Tf = Rp[WS(rs, 4)]; Tg = Rm[WS(rs, 4)]; Th = Tf + Tg; T1c = Tf - Tg; Tb = Ip[WS(rs, 4)]; Tc = Im[WS(rs, 4)]; Td = Tb - Tc; T1a = Tb + Tc; } } Tm = Tk - Tl; T1t = T1m + T1n; T19 = W[16]; T1b = W[17]; T1d = FNMS(T1b, T1c, T19 * T1a); T2j = FMA(T19, T1c, T1b * T1a); { E T9, T1u, Ti, T1v; { E T1, T5, Ta, Te; T1 = W[6]; T5 = W[7]; T9 = FNMS(T5, T8, T1 * T4); T1u = FMA(T1, T8, T5 * T4); Ta = W[14]; Te = W[15]; Ti = FNMS(Te, Th, Ta * Td); T1v = FMA(Ta, Th, Te * Td); } Tj = T9 + Ti; T1Y = KP433012701 * (T1v - T1u); T1w = T1u + T1v; T1G = KP433012701 * (T9 - Ti); } { E T1i, T1S, T1p, T1T; { E T1e, T1g, T1j, T1l; T1e = W[8]; T1g = W[9]; T1i = FNMS(T1g, T1h, T1e * T1f); T1S = FMA(T1e, T1h, T1g * T1f); T1j = W[0]; T1l = W[1]; T1p = FNMS(T1l, T1o, T1j * T1k); T1T = FMA(T1j, T1o, T1l * T1k); } T1q = T1i + T1p; T2q = KP433012701 * (T1i - T1p); T1U = KP433012701 * (T1S - T1T); T2k = T1S + T1T; } } { E Tr, TT, Tv, TV, TA, TY, TE, T10, TN, T14, TJ, T16; { E Tp, Tq, TC, TD; Tp = Ip[WS(rs, 3)]; Tq = Im[WS(rs, 3)]; Tr = Tp - Tq; TT = Tp + Tq; { E Tt, Tu, Ty, Tz; Tt = Rp[WS(rs, 3)]; Tu = Rm[WS(rs, 3)]; Tv = Tt + Tu; TV = Tt - Tu; Ty = Ip[WS(rs, 5)]; Tz = Im[WS(rs, 5)]; TA = Ty - Tz; TY = Ty + Tz; } TC = Rp[WS(rs, 5)]; TD = Rm[WS(rs, 5)]; TE = TC + TD; T10 = TC - TD; { E TL, TM, TH, TI; TL = Rp[WS(rs, 1)]; TM = Rm[WS(rs, 1)]; TN = TL + TM; T14 = TM - TL; TH = Ip[WS(rs, 1)]; TI = Im[WS(rs, 1)]; TJ = TH - TI; T16 = TH + TI; } } { E To, Ts, T13, T15; To = W[10]; Ts = W[11]; Tw = FNMS(Ts, Tv, To * Tr); T1y = FMA(To, Tv, Ts * Tr); T13 = W[5]; T15 = W[4]; T17 = FMA(T13, T14, T15 * T16); T2g = FNMS(T13, T16, T15 * T14); } { E TF, T1z, TO, T1A; { E Tx, TB, TG, TK; Tx = W[18]; TB = W[19]; TF = FNMS(TB, TE, Tx * TA); T1z = FMA(Tx, TE, TB * TA); TG = W[2]; TK = W[3]; TO = FNMS(TK, TN, TG * TJ); T1A = FMA(TG, TN, TK * TJ); } TP = TF + TO; T21 = KP433012701 * (T1A - T1z); T1B = T1z + T1A; T1J = KP433012701 * (TF - TO); } { E TW, T1O, T11, T1N; { E TS, TU, TX, TZ; TS = W[12]; TU = W[13]; TW = FNMS(TU, TV, TS * TT); T1O = FMA(TS, TV, TU * TT); TX = W[20]; TZ = W[21]; T11 = FNMS(TZ, T10, TX * TY); T1N = FMA(TX, T10, TZ * TY); } T12 = TW + T11; T2u = KP433012701 * (T11 - TW); T1P = KP433012701 * (T1N - T1O); T2h = T1O + T1N; } } { E TR, T2f, T2m, T2o, T1s, T1E, T1D, T2n; { E Tn, TQ, T2i, T2l; Tn = Tj + Tm; TQ = Tw + TP; TR = Tn - TQ; T2f = TQ + Tn; T2i = T2g - T2h; T2l = T2j + T2k; T2m = T2i - T2l; T2o = T2i + T2l; } { E T18, T1r, T1x, T1C; T18 = T12 + T17; T1r = T1d + T1q; T1s = T18 - T1r; T1E = T18 + T1r; T1x = T1t + T1w; T1C = T1y + T1B; T1D = T1x + T1C; T2n = T1x - T1C; } Ip[WS(rs, 3)] = KP500000000 * (TR + T1s); Rp[WS(rs, 3)] = KP500000000 * (T2n - T2o); Im[WS(rs, 2)] = KP500000000 * (T1s - TR); Rm[WS(rs, 2)] = KP500000000 * (T2n + T2o); Rm[WS(rs, 5)] = KP500000000 * (T1D - T1E); Im[WS(rs, 5)] = KP500000000 * (T2m - T2f); Rp[0] = KP500000000 * (T1D + T1E); Ip[0] = KP500000000 * (T2f + T2m); } { E T1H, T2b, T2s, T2B, T2v, T2A, T1K, T2c, T1Q, T29, T1Z, T25, T22, T26, T1V; E T28; { E T1F, T2r, T2t, T1I; T1F = FNMS(KP250000000, T1w, KP500000000 * T1t); T1H = T1F - T1G; T2b = T1F + T1G; T2r = FNMS(KP500000000, T2j, KP250000000 * T2k); T2s = T2q - T2r; T2B = T2q + T2r; T2t = FMA(KP250000000, T2h, KP500000000 * T2g); T2v = T2t - T2u; T2A = T2u + T2t; T1I = FNMS(KP250000000, T1B, KP500000000 * T1y); T1K = T1I - T1J; T2c = T1I + T1J; } { E T1M, T1X, T20, T1R; T1M = FNMS(KP250000000, T12, KP500000000 * T17); T1Q = T1M - T1P; T29 = T1P + T1M; T1X = FNMS(KP250000000, Tj, KP500000000 * Tm); T1Z = T1X - T1Y; T25 = T1Y + T1X; T20 = FNMS(KP250000000, TP, KP500000000 * Tw); T22 = T20 - T21; T26 = T21 + T20; T1R = FNMS(KP250000000, T1q, KP500000000 * T1d); T1V = T1R - T1U; T28 = T1R + T1U; } { E T1L, T1W, T2p, T2w; T1L = T1H + T1K; T1W = T1Q + T1V; Rp[WS(rs, 2)] = T1L - T1W; Rm[WS(rs, 3)] = T1L + T1W; T2p = T22 + T1Z; T2w = T2s - T2v; Ip[WS(rs, 2)] = T2p + T2w; Im[WS(rs, 3)] = T2w - T2p; } { E T23, T24, T2x, T2y; T23 = T1Z - T22; T24 = T1V - T1Q; Ip[WS(rs, 5)] = T23 + T24; Im[0] = T24 - T23; T2x = T1H - T1K; T2y = T2v + T2s; Rm[0] = T2x - T2y; Rp[WS(rs, 5)] = T2x + T2y; } { E T27, T2a, T2z, T2C; T27 = T25 - T26; T2a = T28 - T29; Ip[WS(rs, 1)] = T27 + T2a; Im[WS(rs, 4)] = T2a - T27; T2z = T2b - T2c; T2C = T2A - T2B; Rm[WS(rs, 4)] = T2z - T2C; Rp[WS(rs, 1)] = T2z + T2C; } { E T2d, T2e, T2D, T2E; T2d = T2b + T2c; T2e = T29 + T28; Rm[WS(rs, 1)] = T2d - T2e; Rp[WS(rs, 4)] = T2d + T2e; T2D = T26 + T25; T2E = T2A + T2B; Ip[WS(rs, 4)] = T2D + T2E; Im[WS(rs, 1)] = T2E - T2D; } } } } } static const tw_instr twinstr[] = { {TW_FULL, 1, 12}, {TW_NEXT, 1, 0} }; static const hc2c_desc desc = { 12, "hc2cfdft_12", twinstr, &GENUS, {112, 46, 30, 0} }; void X(codelet_hc2cfdft_12) (planner *p) { X(khc2c_register) (p, hc2cfdft_12, &desc, HC2C_VIA_DFT); } #endif /* HAVE_FMA */ fftw-3.3.3/rdft/scalar/r2cf/hc2cfdft_16.c0000644000175000017500000005606012054410720014606 00000000000000/* * Copyright (c) 2003, 2007-11 Matteo Frigo * Copyright (c) 2003, 2007-11 Massachusetts Institute of Technology * * This program is free software; you can redistribute it and/or modify * it under the terms of the GNU General Public License as published by * the Free Software Foundation; either version 2 of the License, or * (at your option) any later version. * * This program is distributed in the hope that it will be useful, * but WITHOUT ANY WARRANTY; without even the implied warranty of * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the * GNU General Public License for more details. * * You should have received a copy of the GNU General Public License * along with this program; if not, write to the Free Software * Foundation, Inc., 51 Franklin Street, Fifth Floor, Boston, MA 02110-1301 USA * */ /* This file was automatically generated --- DO NOT EDIT */ /* Generated on Sun Nov 25 07:40:46 EST 2012 */ #include "codelet-rdft.h" #ifdef HAVE_FMA /* Generated by: ../../../genfft/gen_hc2cdft.native -fma -reorder-insns -schedule-for-pipeline -compact -variables 4 -pipeline-latency 4 -n 16 -dit -name hc2cfdft_16 -include hc2cf.h */ /* * This function contains 206 FP additions, 132 FP multiplications, * (or, 136 additions, 62 multiplications, 70 fused multiply/add), * 96 stack variables, 4 constants, and 64 memory accesses */ #include "hc2cf.h" static void hc2cfdft_16(R *Rp, R *Ip, R *Rm, R *Im, const R *W, stride rs, INT mb, INT me, INT ms) { DK(KP923879532, +0.923879532511286756128183189396788286822416626); DK(KP414213562, +0.414213562373095048801688724209698078569671875); DK(KP707106781, +0.707106781186547524400844362104849039284835938); DK(KP500000000, +0.500000000000000000000000000000000000000000000); { INT m; for (m = mb, W = W + ((mb - 1) * 30); m < me; m = m + 1, Rp = Rp + ms, Ip = Ip + ms, Rm = Rm - ms, Im = Im - ms, W = W + 30, MAKE_VOLATILE_STRIDE(64, rs)) { E T4d, T4g; { E T1f, T2e, T3D, T1K, T2g, T1c, T3H, T2W, T2j, TR, T3E, T2R, T2l, T11, T3G; E T1v, T3p, T2s, Tl, T3o, T3w, T2G, T3z, T1Y, T23, T20, T2H, T21, T29, Tz; E T26, TE, TA, T2v, T2J, T27, Tv, T2u, TB, T22, T28; { E T1o, T1u, T2T, T2V; { E T1I, T1A, T16, T1C, T1H, T1G, T2U, T1z, T1b, T1x, T1w; { E T1d, T1e, T14, T15; T1d = Ip[0]; T1e = Im[0]; T14 = Ip[WS(rs, 4)]; T15 = Im[WS(rs, 4)]; { E T1F, T1D, T1E, T19, T1a; T1D = Rm[0]; T1I = T1d + T1e; T1f = T1d - T1e; T1E = Rp[0]; T1A = T14 + T15; T16 = T14 - T15; T1C = W[0]; T2e = T1E + T1D; T1F = T1D - T1E; T1H = W[1]; T19 = Rp[WS(rs, 4)]; T1a = Rm[WS(rs, 4)]; T1G = T1C * T1F; T2U = T1H * T1F; T1z = W[17]; T1b = T19 + T1a; T1x = T1a - T19; T1w = W[16]; } } { E T2S, T1y, T13, T18; T2S = T1z * T1x; T1y = T1w * T1x; T13 = W[14]; T18 = W[15]; { E T1J, T1B, T2f, T17; T1J = FNMS(T1H, T1I, T1G); T1B = FNMS(T1z, T1A, T1y); T2f = T13 * T1b; T17 = T13 * T16; T2T = FMA(T1w, T1A, T2S); T3D = T1J - T1B; T1K = T1B + T1J; T2g = FMA(T18, T16, T2f); T1c = FNMS(T18, T1b, T17); T2V = FMA(T1C, T1I, T2U); } } } { E T1n, TL, T1m, T1j, TQ, T1l, T2N, TV, T1t, T10, T1q, T1s, T1p, T1r, T2O; E T2Q; { E TO, TP, TJ, TK; TJ = Ip[WS(rs, 2)]; TK = Im[WS(rs, 2)]; TO = Rp[WS(rs, 2)]; T3H = T2V - T2T; T2W = T2T + T2V; T1n = TJ + TK; TL = TJ - TK; TP = Rm[WS(rs, 2)]; T1m = W[9]; T1j = W[8]; { E TT, T1k, TU, TY, TZ; TT = Ip[WS(rs, 6)]; TQ = TO + TP; T1k = TP - TO; TU = Im[WS(rs, 6)]; TY = Rp[WS(rs, 6)]; TZ = Rm[WS(rs, 6)]; T1l = T1j * T1k; T2N = T1m * T1k; TV = TT - TU; T1t = TT + TU; T10 = TY + TZ; T1q = TZ - TY; T1s = W[25]; T1p = W[24]; } } { E TN, T2P, T2i, TM, TI; TI = W[6]; TN = W[7]; T2P = T1s * T1q; T1r = T1p * T1q; T2i = TI * TQ; TM = TI * TL; T2O = FMA(T1j, T1n, T2N); T2Q = FMA(T1p, T1t, T2P); T2j = FMA(TN, TL, T2i); TR = FNMS(TN, TQ, TM); } { E TX, T2k, TW, TS; TS = W[22]; T3E = T2O - T2Q; T2R = T2O + T2Q; TX = W[23]; T2k = TS * T10; TW = TS * TV; T1o = FNMS(T1m, T1n, T1l); T1u = FNMS(T1s, T1t, T1r); T2l = FMA(TX, TV, T2k); T11 = FNMS(TX, T10, TW); } } { E T1Q, T1N, T2C, T1O, T1W, Te, T1T, Tj, Tf, T2q, T2E, T1U, Ta, T2p, Tg; E T1P, T1V; { E T4, T9, T5, T2o, Tb, T1S, T1, T1M, T6; { E T2, T3, T7, T8; T2 = Ip[WS(rs, 1)]; T3G = T1o - T1u; T1v = T1o + T1u; T3 = Im[WS(rs, 1)]; T7 = Rp[WS(rs, 1)]; T8 = Rm[WS(rs, 1)]; T1 = W[2]; T1Q = T2 + T3; T4 = T2 - T3; T1N = T7 - T8; T9 = T7 + T8; T1M = W[4]; T5 = T1 * T4; } { E Tc, Td, Th, Ti; Tc = Ip[WS(rs, 5)]; T2o = T1 * T9; T2C = T1M * T1Q; T1O = T1M * T1N; Td = Im[WS(rs, 5)]; Th = Rp[WS(rs, 5)]; Ti = Rm[WS(rs, 5)]; Tb = W[18]; T1W = Tc + Td; Te = Tc - Td; T1T = Th - Ti; Tj = Th + Ti; T1S = W[20]; Tf = Tb * Te; } T6 = W[3]; T2q = Tb * Tj; T2E = T1S * T1W; T1U = T1S * T1T; Ta = FNMS(T6, T9, T5); T2p = FMA(T6, T4, T2o); Tg = W[19]; T1P = W[5]; T1V = W[21]; } { E Tp, Tu, Tq, T2t, Tw, T25, Tm, T1Z, Tr; { E Tn, To, Ts, Tt, T2r, Tk; Tn = Ip[WS(rs, 7)]; T2r = FMA(Tg, Te, T2q); Tk = FNMS(Tg, Tj, Tf); { E T2D, T1R, T2F, T1X; T2D = FNMS(T1P, T1N, T2C); T1R = FMA(T1P, T1Q, T1O); T2F = FNMS(T1V, T1T, T2E); T1X = FMA(T1V, T1W, T1U); T3p = T2p - T2r; T2s = T2p + T2r; Tl = Ta + Tk; T3o = Ta - Tk; T3w = T2F - T2D; T2G = T2D + T2F; T3z = T1X - T1R; T1Y = T1R + T1X; To = Im[WS(rs, 7)]; } Ts = Rp[WS(rs, 7)]; Tt = Rm[WS(rs, 7)]; Tm = W[26]; T23 = Tn + To; Tp = Tn - To; T20 = Ts - Tt; Tu = Ts + Tt; T1Z = W[28]; Tq = Tm * Tp; } { E Tx, Ty, TC, TD; Tx = Ip[WS(rs, 3)]; T2t = Tm * Tu; T2H = T1Z * T23; T21 = T1Z * T20; Ty = Im[WS(rs, 3)]; TC = Rp[WS(rs, 3)]; TD = Rm[WS(rs, 3)]; Tw = W[10]; T29 = Tx + Ty; Tz = Tx - Ty; T26 = TC - TD; TE = TC + TD; T25 = W[12]; TA = Tw * Tz; } Tr = W[27]; T2v = Tw * TE; T2J = T25 * T29; T27 = T25 * T26; Tv = FNMS(Tr, Tu, Tq); T2u = FMA(Tr, Tp, T2t); TB = W[11]; T22 = W[29]; T28 = W[13]; } } } { E T3r, T3s, T3A, T3x, T3M, T3l, T3L, T3m, T3f, T3i; { E T3c, TH, T36, T3g, T3h, T39, T32, T1h, T2A, T2d, T2h, T31, T2y, T30, T2Y; E T2m, T2B, T1i; { E T2x, T2M, T1L, T2c, T2X, T12, T1g; { E TG, T2b, T34, T2L, T2w, TF, T37, T38, T35; T2w = FMA(TB, Tz, T2v); TF = FNMS(TB, TE, TA); { E T2I, T24, T2K, T2a; T2I = FNMS(T22, T20, T2H); T24 = FMA(T22, T23, T21); T2K = FNMS(T28, T26, T2J); T2a = FMA(T28, T29, T27); T3r = T2u - T2w; T2x = T2u + T2w; TG = Tv + TF; T3s = Tv - TF; T2L = T2I + T2K; T3A = T2I - T2K; T3x = T2a - T24; T2b = T24 + T2a; } T2M = T2G + T2L; T34 = T2L - T2G; T37 = T1K - T1v; T1L = T1v + T1K; T2c = T1Y + T2b; T35 = T1Y - T2b; T3c = Tl - TG; TH = Tl + TG; T38 = T2W - T2R; T2X = T2R + T2W; T36 = T34 + T35; T3g = T34 - T35; T3M = TR - T11; T12 = TR + T11; T3h = T37 + T38; T39 = T37 - T38; T1g = T1c + T1f; T3l = T1f - T1c; } T32 = T1g - T12; T1h = T12 + T1g; T2A = T2c + T1L; T2d = T1L - T2c; T3L = T2e - T2g; T2h = T2e + T2g; T31 = T2x - T2s; T2y = T2s + T2x; T30 = T2M + T2X; T2Y = T2M - T2X; T2m = T2j + T2l; T3m = T2j - T2l; } T2B = T1h - TH; T1i = TH + T1h; { E T3e, T3d, T3j, T3k; { E T33, T3b, T2z, T2Z, T3a, T2n; T3f = T32 - T31; T33 = T31 + T32; T3b = T2h - T2m; T2n = T2h + T2m; Im[WS(rs, 7)] = KP500000000 * (T2d - T1i); Ip[0] = KP500000000 * (T1i + T2d); Im[WS(rs, 3)] = KP500000000 * (T2Y - T2B); Ip[WS(rs, 4)] = KP500000000 * (T2B + T2Y); T2z = T2n - T2y; T2Z = T2n + T2y; T3a = T36 + T39; T3e = T39 - T36; T3d = T3b - T3c; T3j = T3b + T3c; Rp[WS(rs, 4)] = KP500000000 * (T2z + T2A); Rm[WS(rs, 3)] = KP500000000 * (T2z - T2A); Rp[0] = KP500000000 * (T2Z + T30); Rm[WS(rs, 7)] = KP500000000 * (T2Z - T30); Im[WS(rs, 5)] = -(KP500000000 * (FNMS(KP707106781, T3a, T33))); Ip[WS(rs, 2)] = KP500000000 * (FMA(KP707106781, T3a, T33)); T3k = T3g + T3h; T3i = T3g - T3h; } Rp[WS(rs, 2)] = KP500000000 * (FMA(KP707106781, T3k, T3j)); Rm[WS(rs, 5)] = KP500000000 * (FNMS(KP707106781, T3k, T3j)); Rp[WS(rs, 6)] = KP500000000 * (FMA(KP707106781, T3e, T3d)); Rm[WS(rs, 1)] = KP500000000 * (FNMS(KP707106781, T3e, T3d)); } } { E T3Z, T3n, T3F, T3I, T4e, T44, T4f, T47, T4a, T3u, T3U, T3C, T49, T3N, T40; E T3Q; { E T3y, T3B, T3O, T3q, T3t, T3P; { E T42, T43, T45, T46; T3y = T3w + T3x; T42 = T3w - T3x; Im[WS(rs, 1)] = -(KP500000000 * (FNMS(KP707106781, T3i, T3f))); Ip[WS(rs, 6)] = KP500000000 * (FMA(KP707106781, T3i, T3f)); T3Z = T3m + T3l; T3n = T3l - T3m; T43 = T3A - T3z; T3B = T3z + T3A; T3F = T3D - T3E; T45 = T3E + T3D; T46 = T3H - T3G; T3I = T3G + T3H; T3O = T3p + T3o; T3q = T3o - T3p; T4e = FNMS(KP414213562, T42, T43); T44 = FMA(KP414213562, T43, T42); T4f = FNMS(KP414213562, T45, T46); T47 = FMA(KP414213562, T46, T45); T3t = T3r + T3s; T3P = T3r - T3s; } T4a = T3q - T3t; T3u = T3q + T3t; T3U = FNMS(KP414213562, T3y, T3B); T3C = FMA(KP414213562, T3B, T3y); T49 = T3L - T3M; T3N = T3L + T3M; T40 = T3P - T3O; T3Q = T3O + T3P; } { E T3T, T3v, T3X, T3R, T3J, T3V; T3T = FNMS(KP707106781, T3u, T3n); T3v = FMA(KP707106781, T3u, T3n); T3X = FMA(KP707106781, T3Q, T3N); T3R = FNMS(KP707106781, T3Q, T3N); T3J = FNMS(KP414213562, T3I, T3F); T3V = FMA(KP414213562, T3F, T3I); { E T4c, T4b, T4h, T4i, T41, T48; T4d = FMA(KP707106781, T40, T3Z); T41 = FNMS(KP707106781, T40, T3Z); T48 = T44 - T47; T4c = T44 + T47; { E T3Y, T3W, T3K, T3S; T3Y = T3U + T3V; T3W = T3U - T3V; T3K = T3C + T3J; T3S = T3J - T3C; Im[WS(rs, 2)] = -(KP500000000 * (FNMS(KP923879532, T3W, T3T))); Ip[WS(rs, 5)] = KP500000000 * (FMA(KP923879532, T3W, T3T)); Rp[WS(rs, 1)] = KP500000000 * (FMA(KP923879532, T3Y, T3X)); Rm[WS(rs, 6)] = KP500000000 * (FNMS(KP923879532, T3Y, T3X)); Rp[WS(rs, 5)] = KP500000000 * (FMA(KP923879532, T3S, T3R)); Rm[WS(rs, 2)] = KP500000000 * (FNMS(KP923879532, T3S, T3R)); Im[WS(rs, 6)] = -(KP500000000 * (FNMS(KP923879532, T3K, T3v))); Ip[WS(rs, 1)] = KP500000000 * (FMA(KP923879532, T3K, T3v)); Ip[WS(rs, 7)] = KP500000000 * (FMA(KP923879532, T48, T41)); Im[0] = -(KP500000000 * (FNMS(KP923879532, T48, T41))); } T4b = FMA(KP707106781, T4a, T49); T4h = FNMS(KP707106781, T4a, T49); T4i = T4e + T4f; T4g = T4e - T4f; Rm[0] = KP500000000 * (FMA(KP923879532, T4i, T4h)); Rp[WS(rs, 7)] = KP500000000 * (FNMS(KP923879532, T4i, T4h)); Rp[WS(rs, 3)] = KP500000000 * (FMA(KP923879532, T4c, T4b)); Rm[WS(rs, 4)] = KP500000000 * (FNMS(KP923879532, T4c, T4b)); } } } } } Im[WS(rs, 4)] = -(KP500000000 * (FNMS(KP923879532, T4g, T4d))); Ip[WS(rs, 3)] = KP500000000 * (FMA(KP923879532, T4g, T4d)); } } } static const tw_instr twinstr[] = { {TW_FULL, 1, 16}, {TW_NEXT, 1, 0} }; static const hc2c_desc desc = { 16, "hc2cfdft_16", twinstr, &GENUS, {136, 62, 70, 0} }; void X(codelet_hc2cfdft_16) (planner *p) { X(khc2c_register) (p, hc2cfdft_16, &desc, HC2C_VIA_DFT); } #else /* HAVE_FMA */ /* Generated by: ../../../genfft/gen_hc2cdft.native -compact -variables 4 -pipeline-latency 4 -n 16 -dit -name hc2cfdft_16 -include hc2cf.h */ /* * This function contains 206 FP additions, 100 FP multiplications, * (or, 168 additions, 62 multiplications, 38 fused multiply/add), * 61 stack variables, 4 constants, and 64 memory accesses */ #include "hc2cf.h" static void hc2cfdft_16(R *Rp, R *Ip, R *Rm, R *Im, const R *W, stride rs, INT mb, INT me, INT ms) { DK(KP461939766, +0.461939766255643378064091594698394143411208313); DK(KP191341716, +0.191341716182544885864229992015199433380672281); DK(KP353553390, +0.353553390593273762200422181052424519642417969); DK(KP500000000, +0.500000000000000000000000000000000000000000000); { INT m; for (m = mb, W = W + ((mb - 1) * 30); m < me; m = m + 1, Rp = Rp + ms, Ip = Ip + ms, Rm = Rm - ms, Im = Im - ms, W = W + 30, MAKE_VOLATILE_STRIDE(64, rs)) { E T19, T3h, T21, T2Y, T1o, T3d, T2s, T39, TW, T3i, T24, T2Z, T1z, T3c, T2p; E T3a, Tj, T2S, T28, T2R, T1L, T36, T2i, T32, TC, T2V, T2b, T2U, T1W, T35; E T2l, T33; { E T10, T1m, T14, T1k, T18, T1h, T1f, T1Z; { E TY, TZ, T12, T13; TY = Ip[WS(rs, 4)]; TZ = Im[WS(rs, 4)]; T10 = TY - TZ; T1m = TY + TZ; T12 = Rp[WS(rs, 4)]; T13 = Rm[WS(rs, 4)]; T14 = T12 + T13; T1k = T12 - T13; } { E T16, T17, T1d, T1e; T16 = Ip[0]; T17 = Im[0]; T18 = T16 - T17; T1h = T16 + T17; T1d = Rm[0]; T1e = Rp[0]; T1f = T1d - T1e; T1Z = T1e + T1d; } { E T15, T20, TX, T11; TX = W[14]; T11 = W[15]; T15 = FNMS(T11, T14, TX * T10); T20 = FMA(TX, T14, T11 * T10); T19 = T15 + T18; T3h = T1Z - T20; T21 = T1Z + T20; T2Y = T18 - T15; } { E T1i, T2r, T1n, T2q; { E T1c, T1g, T1j, T1l; T1c = W[0]; T1g = W[1]; T1i = FNMS(T1g, T1h, T1c * T1f); T2r = FMA(T1g, T1f, T1c * T1h); T1j = W[16]; T1l = W[17]; T1n = FMA(T1j, T1k, T1l * T1m); T2q = FNMS(T1l, T1k, T1j * T1m); } T1o = T1i - T1n; T3d = T2r - T2q; T2s = T2q + T2r; T39 = T1n + T1i; } } { E TH, T1s, TL, T1q, TQ, T1x, TU, T1v; { E TF, TG, TJ, TK; TF = Ip[WS(rs, 2)]; TG = Im[WS(rs, 2)]; TH = TF - TG; T1s = TF + TG; TJ = Rp[WS(rs, 2)]; TK = Rm[WS(rs, 2)]; TL = TJ + TK; T1q = TJ - TK; } { E TO, TP, TS, TT; TO = Ip[WS(rs, 6)]; TP = Im[WS(rs, 6)]; TQ = TO - TP; T1x = TO + TP; TS = Rp[WS(rs, 6)]; TT = Rm[WS(rs, 6)]; TU = TS + TT; T1v = TS - TT; } { E TM, T22, TV, T23; { E TE, TI, TN, TR; TE = W[6]; TI = W[7]; TM = FNMS(TI, TL, TE * TH); T22 = FMA(TE, TL, TI * TH); TN = W[22]; TR = W[23]; TV = FNMS(TR, TU, TN * TQ); T23 = FMA(TN, TU, TR * TQ); } TW = TM + TV; T3i = TM - TV; T24 = T22 + T23; T2Z = T22 - T23; } { E T1t, T2n, T1y, T2o; { E T1p, T1r, T1u, T1w; T1p = W[8]; T1r = W[9]; T1t = FMA(T1p, T1q, T1r * T1s); T2n = FNMS(T1r, T1q, T1p * T1s); T1u = W[24]; T1w = W[25]; T1y = FMA(T1u, T1v, T1w * T1x); T2o = FNMS(T1w, T1v, T1u * T1x); } T1z = T1t + T1y; T3c = T1y - T1t; T2p = T2n + T2o; T3a = T2n - T2o; } } { E T4, T1E, T8, T1C, Td, T1J, Th, T1H; { E T2, T3, T6, T7; T2 = Ip[WS(rs, 1)]; T3 = Im[WS(rs, 1)]; T4 = T2 - T3; T1E = T2 + T3; T6 = Rp[WS(rs, 1)]; T7 = Rm[WS(rs, 1)]; T8 = T6 + T7; T1C = T6 - T7; } { E Tb, Tc, Tf, Tg; Tb = Ip[WS(rs, 5)]; Tc = Im[WS(rs, 5)]; Td = Tb - Tc; T1J = Tb + Tc; Tf = Rp[WS(rs, 5)]; Tg = Rm[WS(rs, 5)]; Th = Tf + Tg; T1H = Tf - Tg; } { E T9, T26, Ti, T27; { E T1, T5, Ta, Te; T1 = W[2]; T5 = W[3]; T9 = FNMS(T5, T8, T1 * T4); T26 = FMA(T1, T8, T5 * T4); Ta = W[18]; Te = W[19]; Ti = FNMS(Te, Th, Ta * Td); T27 = FMA(Ta, Th, Te * Td); } Tj = T9 + Ti; T2S = T26 - T27; T28 = T26 + T27; T2R = T9 - Ti; } { E T1F, T2g, T1K, T2h; { E T1B, T1D, T1G, T1I; T1B = W[4]; T1D = W[5]; T1F = FMA(T1B, T1C, T1D * T1E); T2g = FNMS(T1D, T1C, T1B * T1E); T1G = W[20]; T1I = W[21]; T1K = FMA(T1G, T1H, T1I * T1J); T2h = FNMS(T1I, T1H, T1G * T1J); } T1L = T1F + T1K; T36 = T2g - T2h; T2i = T2g + T2h; T32 = T1K - T1F; } } { E Tn, T1P, Tr, T1N, Tw, T1U, TA, T1S; { E Tl, Tm, Tp, Tq; Tl = Ip[WS(rs, 7)]; Tm = Im[WS(rs, 7)]; Tn = Tl - Tm; T1P = Tl + Tm; Tp = Rp[WS(rs, 7)]; Tq = Rm[WS(rs, 7)]; Tr = Tp + Tq; T1N = Tp - Tq; } { E Tu, Tv, Ty, Tz; Tu = Ip[WS(rs, 3)]; Tv = Im[WS(rs, 3)]; Tw = Tu - Tv; T1U = Tu + Tv; Ty = Rp[WS(rs, 3)]; Tz = Rm[WS(rs, 3)]; TA = Ty + Tz; T1S = Ty - Tz; } { E Ts, T29, TB, T2a; { E Tk, To, Tt, Tx; Tk = W[26]; To = W[27]; Ts = FNMS(To, Tr, Tk * Tn); T29 = FMA(Tk, Tr, To * Tn); Tt = W[10]; Tx = W[11]; TB = FNMS(Tx, TA, Tt * Tw); T2a = FMA(Tt, TA, Tx * Tw); } TC = Ts + TB; T2V = Ts - TB; T2b = T29 + T2a; T2U = T29 - T2a; } { E T1Q, T2j, T1V, T2k; { E T1M, T1O, T1R, T1T; T1M = W[28]; T1O = W[29]; T1Q = FMA(T1M, T1N, T1O * T1P); T2j = FNMS(T1O, T1N, T1M * T1P); T1R = W[12]; T1T = W[13]; T1V = FMA(T1R, T1S, T1T * T1U); T2k = FNMS(T1T, T1S, T1R * T1U); } T1W = T1Q + T1V; T35 = T1V - T1Q; T2l = T2j + T2k; T33 = T2j - T2k; } } { E T1b, T2f, T2u, T2w, T1Y, T2e, T2d, T2v; { E TD, T1a, T2m, T2t; TD = Tj + TC; T1a = TW + T19; T1b = TD + T1a; T2f = T1a - TD; T2m = T2i + T2l; T2t = T2p + T2s; T2u = T2m - T2t; T2w = T2m + T2t; } { E T1A, T1X, T25, T2c; T1A = T1o - T1z; T1X = T1L + T1W; T1Y = T1A - T1X; T2e = T1X + T1A; T25 = T21 + T24; T2c = T28 + T2b; T2d = T25 - T2c; T2v = T25 + T2c; } Ip[0] = KP500000000 * (T1b + T1Y); Rp[0] = KP500000000 * (T2v + T2w); Im[WS(rs, 7)] = KP500000000 * (T1Y - T1b); Rm[WS(rs, 7)] = KP500000000 * (T2v - T2w); Rm[WS(rs, 3)] = KP500000000 * (T2d - T2e); Im[WS(rs, 3)] = KP500000000 * (T2u - T2f); Rp[WS(rs, 4)] = KP500000000 * (T2d + T2e); Ip[WS(rs, 4)] = KP500000000 * (T2f + T2u); } { E T2z, T2L, T2J, T2P, T2C, T2M, T2F, T2N; { E T2x, T2y, T2H, T2I; T2x = T2b - T28; T2y = T19 - TW; T2z = KP500000000 * (T2x + T2y); T2L = KP500000000 * (T2y - T2x); T2H = T21 - T24; T2I = Tj - TC; T2J = KP500000000 * (T2H - T2I); T2P = KP500000000 * (T2H + T2I); } { E T2A, T2B, T2D, T2E; T2A = T2l - T2i; T2B = T1L - T1W; T2C = T2A + T2B; T2M = T2A - T2B; T2D = T1z + T1o; T2E = T2s - T2p; T2F = T2D - T2E; T2N = T2D + T2E; } { E T2G, T2Q, T2K, T2O; T2G = KP353553390 * (T2C + T2F); Ip[WS(rs, 2)] = T2z + T2G; Im[WS(rs, 5)] = T2G - T2z; T2Q = KP353553390 * (T2M + T2N); Rm[WS(rs, 5)] = T2P - T2Q; Rp[WS(rs, 2)] = T2P + T2Q; T2K = KP353553390 * (T2F - T2C); Rm[WS(rs, 1)] = T2J - T2K; Rp[WS(rs, 6)] = T2J + T2K; T2O = KP353553390 * (T2M - T2N); Ip[WS(rs, 6)] = T2L + T2O; Im[WS(rs, 1)] = T2O - T2L; } } { E T30, T3w, T3F, T3j, T2X, T3G, T3D, T3L, T3m, T3v, T38, T3q, T3A, T3K, T3f; E T3r; { E T2T, T2W, T34, T37; T30 = KP500000000 * (T2Y - T2Z); T3w = KP500000000 * (T2Z + T2Y); T3F = KP500000000 * (T3h - T3i); T3j = KP500000000 * (T3h + T3i); T2T = T2R - T2S; T2W = T2U + T2V; T2X = KP353553390 * (T2T + T2W); T3G = KP353553390 * (T2T - T2W); { E T3B, T3C, T3k, T3l; T3B = T3a + T39; T3C = T3d - T3c; T3D = FNMS(KP461939766, T3C, KP191341716 * T3B); T3L = FMA(KP461939766, T3B, KP191341716 * T3C); T3k = T2S + T2R; T3l = T2U - T2V; T3m = KP353553390 * (T3k + T3l); T3v = KP353553390 * (T3l - T3k); } T34 = T32 + T33; T37 = T35 - T36; T38 = FMA(KP191341716, T34, KP461939766 * T37); T3q = FNMS(KP191341716, T37, KP461939766 * T34); { E T3y, T3z, T3b, T3e; T3y = T33 - T32; T3z = T36 + T35; T3A = FMA(KP461939766, T3y, KP191341716 * T3z); T3K = FNMS(KP461939766, T3z, KP191341716 * T3y); T3b = T39 - T3a; T3e = T3c + T3d; T3f = FNMS(KP191341716, T3e, KP461939766 * T3b); T3r = FMA(KP191341716, T3b, KP461939766 * T3e); } } { E T31, T3g, T3t, T3u; T31 = T2X + T30; T3g = T38 + T3f; Ip[WS(rs, 1)] = T31 + T3g; Im[WS(rs, 6)] = T3g - T31; T3t = T3j + T3m; T3u = T3q + T3r; Rm[WS(rs, 6)] = T3t - T3u; Rp[WS(rs, 1)] = T3t + T3u; } { E T3n, T3o, T3p, T3s; T3n = T3j - T3m; T3o = T3f - T38; Rm[WS(rs, 2)] = T3n - T3o; Rp[WS(rs, 5)] = T3n + T3o; T3p = T30 - T2X; T3s = T3q - T3r; Ip[WS(rs, 5)] = T3p + T3s; Im[WS(rs, 2)] = T3s - T3p; } { E T3x, T3E, T3N, T3O; T3x = T3v + T3w; T3E = T3A + T3D; Ip[WS(rs, 3)] = T3x + T3E; Im[WS(rs, 4)] = T3E - T3x; T3N = T3F + T3G; T3O = T3K + T3L; Rm[WS(rs, 4)] = T3N - T3O; Rp[WS(rs, 3)] = T3N + T3O; } { E T3H, T3I, T3J, T3M; T3H = T3F - T3G; T3I = T3D - T3A; Rm[0] = T3H - T3I; Rp[WS(rs, 7)] = T3H + T3I; T3J = T3w - T3v; T3M = T3K - T3L; Ip[WS(rs, 7)] = T3J + T3M; Im[0] = T3M - T3J; } } } } } static const tw_instr twinstr[] = { {TW_FULL, 1, 16}, {TW_NEXT, 1, 0} }; static const hc2c_desc desc = { 16, "hc2cfdft_16", twinstr, &GENUS, {168, 62, 38, 0} }; void X(codelet_hc2cfdft_16) (planner *p) { X(khc2c_register) (p, hc2cfdft_16, &desc, HC2C_VIA_DFT); } #endif /* HAVE_FMA */ fftw-3.3.3/rdft/scalar/r2cf/hc2cfdft_32.c0000644000175000017500000015037112054410731014606 00000000000000/* * Copyright (c) 2003, 2007-11 Matteo Frigo * Copyright (c) 2003, 2007-11 Massachusetts Institute of Technology * * This program is free software; you can redistribute it and/or modify * it under the terms of the GNU General Public License as published by * the Free Software Foundation; either version 2 of the License, or * (at your option) any later version. * * This program is distributed in the hope that it will be useful, * but WITHOUT ANY WARRANTY; without even the implied warranty of * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the * GNU General Public License for more details. * * You should have received a copy of the GNU General Public License * along with this program; if not, write to the Free Software * Foundation, Inc., 51 Franklin Street, Fifth Floor, Boston, MA 02110-1301 USA * */ /* This file was automatically generated --- DO NOT EDIT */ /* Generated on Sun Nov 25 07:40:46 EST 2012 */ #include "codelet-rdft.h" #ifdef HAVE_FMA /* Generated by: ../../../genfft/gen_hc2cdft.native -fma -reorder-insns -schedule-for-pipeline -compact -variables 4 -pipeline-latency 4 -n 32 -dit -name hc2cfdft_32 -include hc2cf.h */ /* * This function contains 498 FP additions, 324 FP multiplications, * (or, 300 additions, 126 multiplications, 198 fused multiply/add), * 172 stack variables, 8 constants, and 128 memory accesses */ #include "hc2cf.h" static void hc2cfdft_32(R *Rp, R *Ip, R *Rm, R *Im, const R *W, stride rs, INT mb, INT me, INT ms) { DK(KP980785280, +0.980785280403230449126182236134239036973933731); DK(KP831469612, +0.831469612302545237078788377617905756738560812); DK(KP668178637, +0.668178637919298919997757686523080761552472251); DK(KP198912367, +0.198912367379658006911597622644676228597850501); DK(KP923879532, +0.923879532511286756128183189396788286822416626); DK(KP414213562, +0.414213562373095048801688724209698078569671875); DK(KP707106781, +0.707106781186547524400844362104849039284835938); DK(KP500000000, +0.500000000000000000000000000000000000000000000); { INT m; for (m = mb, W = W + ((mb - 1) * 62); m < me; m = m + 1, Rp = Rp + ms, Ip = Ip + ms, Rm = Rm - ms, Im = Im - ms, W = W + 62, MAKE_VOLATILE_STRIDE(128, rs)) { E T9X, Ta0; { E T3B, T89, T61, T8l, T2F, T7p, T8t, T4B, T7I, T5e, T7L, T1n, T7R, T5E, T82; E T4u, T3m, T8k, T5W, T8a, T2r, T8u, T4G, T7q, T59, T7K, T7H, T12, T5z, T81; E T7Q, T4h, T4Y, T7D, T7A, Tl, T5o, T3Q, T84, T7V, T2V, T4M, T7t, T7s, T1K; E T5L, T8e, T8n, T38, T7v, T4R, T7w, T25, T5Q, T8h, T8o, T3V, T3S, T5p, T3T; E T41, Tz, T3Y, TE, TA, T51, T5r, T3Z, Tv, T50, TB, T3U, T40; { E T49, T46, T5v, T47, T4f, TV, T4c, T10, TW, T57, T5x, T4d, TR, T56, TX; E T48, T4e; { E T4m, T4j, T5A, T4k, T4s, T1g, T4p, T1l, T1h, T5c, T5C, T4q, T1c, T5b, T1i; E T4l, T4r; { E T2E, T4y, T2B, T4A; { E T3y, T3z, T3t, T5Z, T3x, T2v, T3r, T3q, T3n, T2A, T3o, T2s; { E T2C, T2D, T3w, T3u, T3v; T2C = Ip[0]; T2D = Im[0]; T3u = Rm[0]; T3v = Rp[0]; T3y = W[1]; T3z = T2C + T2D; T2E = T2C - T2D; T4y = T3v + T3u; T3w = T3u - T3v; T3t = W[0]; { E T2y, T2z, T2t, T2u; T2t = Ip[WS(rs, 8)]; T2u = Im[WS(rs, 8)]; T5Z = T3y * T3w; T3x = T3t * T3w; T2y = Rp[WS(rs, 8)]; T2v = T2t - T2u; T3r = T2t + T2u; T2z = Rm[WS(rs, 8)]; T3q = W[33]; T3n = W[32]; T2A = T2y + T2z; T3o = T2z - T2y; T2s = W[30]; } } { E T3A, T5X, T4z, T2w, T3s, T3p, T5Y, T60, T2x; T3A = FNMS(T3y, T3z, T3x); T3p = T3n * T3o; T5X = T3q * T3o; T4z = T2s * T2A; T2w = T2s * T2v; T3s = FNMS(T3q, T3r, T3p); T5Y = FMA(T3n, T3r, T5X); T60 = FMA(T3t, T3z, T5Z); T2x = W[31]; T3B = T3s + T3A; T89 = T3A - T3s; T61 = T5Y + T60; T8l = T60 - T5Y; T2B = FNMS(T2x, T2A, T2w); T4A = FMA(T2x, T2v, T4z); } } { E T16, T1b, T17, T5a, T1d, T4o, T18; { E T19, T1a, T13, T4i, T14, T15; T14 = Ip[WS(rs, 3)]; T15 = Im[WS(rs, 3)]; T2F = T2B + T2E; T7p = T2E - T2B; T8t = T4y - T4A; T4B = T4y + T4A; T4m = T14 + T15; T16 = T14 - T15; T19 = Rp[WS(rs, 3)]; T1a = Rm[WS(rs, 3)]; T13 = W[10]; T4i = W[12]; { E T1e, T1f, T1j, T1k; T1e = Ip[WS(rs, 11)]; T4j = T19 - T1a; T1b = T19 + T1a; T17 = T13 * T16; T5A = T4i * T4m; T4k = T4i * T4j; T5a = T13 * T1b; T1f = Im[WS(rs, 11)]; T1j = Rp[WS(rs, 11)]; T1k = Rm[WS(rs, 11)]; T1d = W[42]; T4s = T1e + T1f; T1g = T1e - T1f; T4p = T1j - T1k; T1l = T1j + T1k; T4o = W[44]; T1h = T1d * T1g; } } T18 = W[11]; T5c = T1d * T1l; T5C = T4o * T4s; T4q = T4o * T4p; T1c = FNMS(T18, T1b, T17); T5b = FMA(T18, T16, T5a); T1i = W[43]; T4l = W[13]; T4r = W[45]; } } { E T4D, T2g, T2q, T4F; { E T3d, T3e, T2a, T2f, T3a, T5S, T3c, T4C, T2b, T3j, T2k, T3k, T2p, T3h, T3g; E T2h, T5U, T3b, T27; { E T28, T29, T2d, T2e, T5d, T1m; T28 = Ip[WS(rs, 4)]; T5d = FMA(T1i, T1g, T5c); T1m = FNMS(T1i, T1l, T1h); { E T5B, T4n, T5D, T4t; T5B = FNMS(T4l, T4j, T5A); T4n = FMA(T4l, T4m, T4k); T5D = FNMS(T4r, T4p, T5C); T4t = FMA(T4r, T4s, T4q); T7I = T5b - T5d; T5e = T5b + T5d; T7L = T1c - T1m; T1n = T1c + T1m; T7R = T5D - T5B; T5E = T5B + T5D; T82 = T4t - T4n; T4u = T4n + T4t; T29 = Im[WS(rs, 4)]; } T2d = Rp[WS(rs, 4)]; T2e = Rm[WS(rs, 4)]; T3d = W[17]; T3e = T28 + T29; T2a = T28 - T29; T3b = T2e - T2d; T2f = T2d + T2e; T3a = W[16]; T27 = W[14]; T5S = T3d * T3b; } { E T2i, T2j, T2n, T2o; T2i = Ip[WS(rs, 12)]; T3c = T3a * T3b; T4C = T27 * T2f; T2b = T27 * T2a; T2j = Im[WS(rs, 12)]; T2n = Rp[WS(rs, 12)]; T2o = Rm[WS(rs, 12)]; T3j = W[49]; T2k = T2i - T2j; T3k = T2i + T2j; T2p = T2n + T2o; T3h = T2o - T2n; T3g = W[48]; T2h = W[46]; T5U = T3j * T3h; } { E T3f, T3i, T4E, T2l; T3f = FNMS(T3d, T3e, T3c); T3i = T3g * T3h; T4E = T2h * T2p; T2l = T2h * T2k; { E T5T, T3l, T5V, T2c, T2m; T5T = FMA(T3a, T3e, T5S); T3l = FNMS(T3j, T3k, T3i); T5V = FMA(T3g, T3k, T5U); T2c = W[15]; T2m = W[47]; T3m = T3f + T3l; T8k = T3f - T3l; T5W = T5T + T5V; T8a = T5T - T5V; T4D = FMA(T2c, T2a, T4C); T2g = FNMS(T2c, T2f, T2b); T2q = FNMS(T2m, T2p, T2l); T4F = FMA(T2m, T2k, T4E); } } } { E TL, TQ, TM, T55, TS, T4b, TN; { E TO, TP, TI, T45, TJ, TK; TJ = Ip[WS(rs, 15)]; TK = Im[WS(rs, 15)]; T2r = T2g + T2q; T8u = T2g - T2q; T4G = T4D + T4F; T7q = T4D - T4F; T49 = TJ + TK; TL = TJ - TK; TO = Rp[WS(rs, 15)]; TP = Rm[WS(rs, 15)]; TI = W[58]; T45 = W[60]; { E TT, TU, TY, TZ; TT = Ip[WS(rs, 7)]; T46 = TO - TP; TQ = TO + TP; TM = TI * TL; T5v = T45 * T49; T47 = T45 * T46; T55 = TI * TQ; TU = Im[WS(rs, 7)]; TY = Rp[WS(rs, 7)]; TZ = Rm[WS(rs, 7)]; TS = W[26]; T4f = TT + TU; TV = TT - TU; T4c = TY - TZ; T10 = TY + TZ; T4b = W[28]; TW = TS * TV; } } TN = W[59]; T57 = TS * T10; T5x = T4b * T4f; T4d = T4b * T4c; TR = FNMS(TN, TQ, TM); T56 = FMA(TN, TL, T55); TX = W[27]; T48 = W[61]; T4e = W[29]; } } } { E T8c, T8d, T8f, T8g; { E T3I, T3F, T5k, T3G, T3O, Te, T3L, Tj, Tf, T4W, T5m, T3M, Ta, T4V, Tg; E T3H, T3N; { E T4, T9, T5, T4U, Tb, T3K, T1, T3E, T6; { E T2, T3, T7, T8, T58, T11; T2 = Ip[WS(rs, 1)]; T58 = FMA(TX, TV, T57); T11 = FNMS(TX, T10, TW); { E T5w, T4a, T5y, T4g; T5w = FNMS(T48, T46, T5v); T4a = FMA(T48, T49, T47); T5y = FNMS(T4e, T4c, T5x); T4g = FMA(T4e, T4f, T4d); T59 = T56 + T58; T7K = T56 - T58; T7H = TR - T11; T12 = TR + T11; T5z = T5w + T5y; T81 = T5w - T5y; T7Q = T4g - T4a; T4h = T4a + T4g; T3 = Im[WS(rs, 1)]; } T7 = Rp[WS(rs, 1)]; T8 = Rm[WS(rs, 1)]; T1 = W[2]; T3I = T2 + T3; T4 = T2 - T3; T3F = T7 - T8; T9 = T7 + T8; T3E = W[4]; T5 = T1 * T4; } { E Tc, Td, Th, Ti; Tc = Ip[WS(rs, 9)]; T4U = T1 * T9; T5k = T3E * T3I; T3G = T3E * T3F; Td = Im[WS(rs, 9)]; Th = Rp[WS(rs, 9)]; Ti = Rm[WS(rs, 9)]; Tb = W[34]; T3O = Tc + Td; Te = Tc - Td; T3L = Th - Ti; Tj = Th + Ti; T3K = W[36]; Tf = Tb * Te; } T6 = W[3]; T4W = Tb * Tj; T5m = T3K * T3O; T3M = T3K * T3L; Ta = FNMS(T6, T9, T5); T4V = FMA(T6, T4, T4U); Tg = W[35]; T3H = W[5]; T3N = W[37]; } { E T1t, T2N, T2M, T2J, T1y, T2L, T5H, T4I, T1u, T2S, T1D, T2T, T1I, T2Q, T2P; E T1A, T5J; { E T2K, T1q, T1w, T1x; { E T1r, T7U, T7T, T1s, T4X, Tk; T1r = Ip[WS(rs, 2)]; T4X = FMA(Tg, Te, T4W); Tk = FNMS(Tg, Tj, Tf); { E T5l, T3J, T5n, T3P; T5l = FNMS(T3H, T3F, T5k); T3J = FMA(T3H, T3I, T3G); T5n = FNMS(T3N, T3L, T5m); T3P = FMA(T3N, T3O, T3M); T4Y = T4V + T4X; T7D = T4V - T4X; T7A = Ta - Tk; Tl = Ta + Tk; T7U = T5l - T5n; T5o = T5l + T5n; T7T = T3P - T3J; T3Q = T3J + T3P; T1s = Im[WS(rs, 2)]; } T1w = Rp[WS(rs, 2)]; T84 = T7U + T7T; T7V = T7T - T7U; T1t = T1r - T1s; T2N = T1r + T1s; T1x = Rm[WS(rs, 2)]; } T2M = W[9]; T2J = W[8]; T1y = T1w + T1x; T2K = T1x - T1w; T1q = W[6]; { E T1B, T1C, T1G, T1H; T1B = Ip[WS(rs, 10)]; T2L = T2J * T2K; T5H = T2M * T2K; T4I = T1q * T1y; T1u = T1q * T1t; T1C = Im[WS(rs, 10)]; T1G = Rp[WS(rs, 10)]; T1H = Rm[WS(rs, 10)]; T2S = W[41]; T1D = T1B - T1C; T2T = T1B + T1C; T1I = T1G + T1H; T2Q = T1H - T1G; T2P = W[40]; T1A = W[38]; T5J = T2S * T2Q; } } { E T2R, T4K, T1E, T1z, T4J, T1F, T1v, T2O, T2U; T1v = W[7]; T2R = T2P * T2Q; T4K = T1A * T1I; T1E = T1A * T1D; T1z = FNMS(T1v, T1y, T1u); T4J = FMA(T1v, T1t, T4I); T1F = W[39]; T2O = FNMS(T2M, T2N, T2L); T2U = FNMS(T2S, T2T, T2R); { E T5I, T4L, T1J, T5K; T5I = FMA(T2J, T2N, T5H); T4L = FMA(T1F, T1D, T4K); T1J = FNMS(T1F, T1I, T1E); T8c = T2O - T2U; T2V = T2O + T2U; T5K = FMA(T2P, T2T, T5J); T4M = T4J + T4L; T7t = T4J - T4L; T7s = T1z - T1J; T1K = T1z + T1J; T8d = T5I - T5K; T5L = T5I + T5K; } } } } { E T2Z, T30, T1O, T1T, T2W, T5M, T2Y, T4N, T1P, T35, T1Y, T36, T23, T33, T32; E T1V, T5O, T2X, T1L; { E T1M, T1N, T1R, T1S; T1M = Ip[WS(rs, 14)]; T8e = T8c - T8d; T8n = T8c + T8d; T1N = Im[WS(rs, 14)]; T1R = Rp[WS(rs, 14)]; T1S = Rm[WS(rs, 14)]; T2Z = W[57]; T30 = T1M + T1N; T1O = T1M - T1N; T2X = T1S - T1R; T1T = T1R + T1S; T2W = W[56]; T1L = W[54]; T5M = T2Z * T2X; } { E T1W, T1X, T21, T22; T1W = Ip[WS(rs, 6)]; T2Y = T2W * T2X; T4N = T1L * T1T; T1P = T1L * T1O; T1X = Im[WS(rs, 6)]; T21 = Rp[WS(rs, 6)]; T22 = Rm[WS(rs, 6)]; T35 = W[25]; T1Y = T1W - T1X; T36 = T1W + T1X; T23 = T21 + T22; T33 = T22 - T21; T32 = W[24]; T1V = W[22]; T5O = T35 * T33; } { E T34, T4P, T1Z, T1U, T4O, T20, T1Q, T31, T37; T1Q = W[55]; T34 = T32 * T33; T4P = T1V * T23; T1Z = T1V * T1Y; T1U = FNMS(T1Q, T1T, T1P); T4O = FMA(T1Q, T1O, T4N); T20 = W[23]; T31 = FNMS(T2Z, T30, T2Y); T37 = FNMS(T35, T36, T34); { E T5N, T4Q, T24, T5P; T5N = FMA(T2W, T30, T5M); T4Q = FMA(T20, T1Y, T4P); T24 = FNMS(T20, T23, T1Z); T8f = T31 - T37; T38 = T31 + T37; T5P = FMA(T32, T36, T5O); T7v = T4O - T4Q; T4R = T4O + T4Q; T7w = T1U - T24; T25 = T1U + T24; T8g = T5N - T5P; T5Q = T5N + T5P; } } } { E Tp, Tu, Tq, T4Z, Tw, T3X, Tm, T3R, Tr; { E Tn, To, Ts, Tt; Tn = Ip[WS(rs, 5)]; T8h = T8f + T8g; T8o = T8g - T8f; To = Im[WS(rs, 5)]; Ts = Rp[WS(rs, 5)]; Tt = Rm[WS(rs, 5)]; Tm = W[18]; T3V = Tn + To; Tp = Tn - To; T3S = Ts - Tt; Tu = Ts + Tt; T3R = W[20]; Tq = Tm * Tp; } { E Tx, Ty, TC, TD; Tx = Ip[WS(rs, 13)]; T4Z = Tm * Tu; T5p = T3R * T3V; T3T = T3R * T3S; Ty = Im[WS(rs, 13)]; TC = Rp[WS(rs, 13)]; TD = Rm[WS(rs, 13)]; Tw = W[50]; T41 = Tx + Ty; Tz = Tx - Ty; T3Y = TC - TD; TE = TC + TD; T3X = W[52]; TA = Tw * Tz; } Tr = W[19]; T51 = Tw * TE; T5r = T3X * T41; T3Z = T3X * T3Y; Tv = FNMS(Tr, Tu, Tq); T50 = FMA(Tr, Tp, T4Z); TB = W[51]; T3U = W[21]; T40 = W[53]; } } } { E T6y, T7B, T7E, T6u, T6S, T85, T7Y, T6s, T6v, T6x, T6R, T6r, T6F, T6D, T6C; E T6G, T6M, T6K, T6J, T6N, T6l, T6o, T7j, T7m; { E T6i, T1p, T68, T2H, T67, T5g, T6h, T4T, T4w, T5G, T6d, T3D, T6c, T6m, T63; E T6e; { E T5t, T43, T26, T2G, T54, T5f, T4H, T4S; { E T1o, T53, T7W, T7X, TH, T52, TF, T5q; T6y = T12 - T1n; T1o = T12 + T1n; T52 = FMA(TB, Tz, T51); TF = FNMS(TB, TE, TA); T5q = FNMS(T3U, T3S, T5p); { E T3W, T5s, T42, TG; T3W = FMA(T3U, T3V, T3T); T5s = FNMS(T40, T3Y, T5r); T42 = FMA(T40, T41, T3Z); T7B = T50 - T52; T53 = T50 + T52; T7E = Tv - TF; TG = Tv + TF; T7W = T5s - T5q; T5t = T5q + T5s; T7X = T3W - T42; T43 = T3W + T42; TH = Tl + TG; T6u = Tl - TG; } T6S = T1K - T25; T26 = T1K + T25; T85 = T7W - T7X; T7Y = T7W + T7X; T6i = TH - T1o; T1p = TH + T1o; T2G = T2r + T2F; T6s = T2F - T2r; T6v = T4Y - T53; T54 = T4Y + T53; T5f = T59 + T5e; T6x = T59 - T5e; } T6R = T4B - T4G; T4H = T4B + T4G; T68 = T2G - T26; T2H = T26 + T2G; T67 = T5f - T54; T5g = T54 + T5f; T4S = T4M + T4R; T6r = T4R - T4M; { E T5u, T6b, T5F, T44, T4v; T6F = T43 - T3Q; T44 = T3Q + T43; T4v = T4h + T4u; T6D = T4u - T4h; T6C = T5t - T5o; T5u = T5o + T5t; T6h = T4H - T4S; T4T = T4H + T4S; T6b = T44 - T4v; T4w = T44 + T4v; T6G = T5z - T5E; T5F = T5z + T5E; { E T5R, T62, T39, T3C, T6a; T6M = T2V - T38; T39 = T2V + T38; T3C = T3m + T3B; T6K = T3B - T3m; T6a = T5F - T5u; T5G = T5u + T5F; T6J = T5Q - T5L; T5R = T5L + T5Q; T6d = T3C - T39; T3D = T39 + T3C; T6N = T61 - T5W; T62 = T5W + T61; T6c = T6a + T6b; T6m = T6a - T6b; T63 = T5R + T62; T6e = T62 - T5R; } } } { E T5j, T6n, T6f, T64; { E T5i, T5h, T65, T66, T2I, T4x; T5j = T2H - T1p; T2I = T1p + T2H; T4x = T3D - T4w; T5i = T4w + T3D; T6n = T6d + T6e; T6f = T6d - T6e; T5h = T4T - T5g; T65 = T4T + T5g; Im[WS(rs, 15)] = KP500000000 * (T4x - T2I); Ip[0] = KP500000000 * (T2I + T4x); T66 = T5G + T63; T64 = T5G - T63; Rp[0] = KP500000000 * (T65 + T66); Rm[WS(rs, 15)] = KP500000000 * (T65 - T66); Rp[WS(rs, 8)] = KP500000000 * (T5h + T5i); Rm[WS(rs, 7)] = KP500000000 * (T5h - T5i); } { E T6k, T6j, T6p, T6q, T69, T6g; T6l = T68 - T67; T69 = T67 + T68; T6g = T6c + T6f; T6k = T6f - T6c; T6j = T6h - T6i; T6p = T6h + T6i; Im[WS(rs, 7)] = KP500000000 * (T64 - T5j); Ip[WS(rs, 8)] = KP500000000 * (T5j + T64); Im[WS(rs, 11)] = -(KP500000000 * (FNMS(KP707106781, T6g, T69))); Ip[WS(rs, 4)] = KP500000000 * (FMA(KP707106781, T6g, T69)); T6q = T6m + T6n; T6o = T6m - T6n; Rp[WS(rs, 4)] = KP500000000 * (FMA(KP707106781, T6q, T6p)); Rm[WS(rs, 11)] = KP500000000 * (FNMS(KP707106781, T6q, T6p)); Rp[WS(rs, 12)] = KP500000000 * (FMA(KP707106781, T6k, T6j)); Rm[WS(rs, 3)] = KP500000000 * (FNMS(KP707106781, T6k, T6j)); } } } { E T75, T6t, T7f, T6T, T76, T6W, T7g, T6A, T7b, T6L, T7a, T7k, T70, T6I, T6U; E T6w; Im[WS(rs, 3)] = -(KP500000000 * (FNMS(KP707106781, T6o, T6l))); Ip[WS(rs, 12)] = KP500000000 * (FMA(KP707106781, T6o, T6l)); T75 = T6s - T6r; T6t = T6r + T6s; T7f = T6R - T6S; T6T = T6R + T6S; T6U = T6v + T6u; T6w = T6u - T6v; { E T78, T6E, T6V, T6z, T79, T6H; T6V = T6x - T6y; T6z = T6x + T6y; T78 = T6C - T6D; T6E = T6C + T6D; T76 = T6V - T6U; T6W = T6U + T6V; T7g = T6w - T6z; T6A = T6w + T6z; T79 = T6G - T6F; T6H = T6F + T6G; T7b = T6K - T6J; T6L = T6J + T6K; T7a = FMA(KP414213562, T79, T78); T7k = FNMS(KP414213562, T78, T79); T70 = FNMS(KP414213562, T6E, T6H); T6I = FMA(KP414213562, T6H, T6E); } { E T6Z, T6B, T73, T6X, T7c, T6O; T6Z = FNMS(KP707106781, T6A, T6t); T6B = FMA(KP707106781, T6A, T6t); T73 = FMA(KP707106781, T6W, T6T); T6X = FNMS(KP707106781, T6W, T6T); T7c = T6N - T6M; T6O = T6M + T6N; { E T7i, T7h, T7n, T7o; { E T77, T7l, T71, T6P, T7e, T7d; T7j = FMA(KP707106781, T76, T75); T77 = FNMS(KP707106781, T76, T75); T7d = FMA(KP414213562, T7c, T7b); T7l = FNMS(KP414213562, T7b, T7c); T71 = FMA(KP414213562, T6L, T6O); T6P = FNMS(KP414213562, T6O, T6L); T7e = T7a - T7d; T7i = T7a + T7d; T7h = FMA(KP707106781, T7g, T7f); T7n = FNMS(KP707106781, T7g, T7f); { E T72, T74, T6Y, T6Q; T72 = T70 - T71; T74 = T70 + T71; T6Y = T6P - T6I; T6Q = T6I + T6P; Im[WS(rs, 1)] = -(KP500000000 * (FNMS(KP923879532, T7e, T77))); Ip[WS(rs, 14)] = KP500000000 * (FMA(KP923879532, T7e, T77)); Im[WS(rs, 5)] = -(KP500000000 * (FNMS(KP923879532, T72, T6Z))); Ip[WS(rs, 10)] = KP500000000 * (FMA(KP923879532, T72, T6Z)); Rp[WS(rs, 2)] = KP500000000 * (FMA(KP923879532, T74, T73)); Rm[WS(rs, 13)] = KP500000000 * (FNMS(KP923879532, T74, T73)); Rp[WS(rs, 10)] = KP500000000 * (FMA(KP923879532, T6Y, T6X)); Rm[WS(rs, 5)] = KP500000000 * (FNMS(KP923879532, T6Y, T6X)); Im[WS(rs, 13)] = -(KP500000000 * (FNMS(KP923879532, T6Q, T6B))); Ip[WS(rs, 2)] = KP500000000 * (FMA(KP923879532, T6Q, T6B)); T7o = T7k + T7l; T7m = T7k - T7l; } } Rm[WS(rs, 1)] = KP500000000 * (FMA(KP923879532, T7o, T7n)); Rp[WS(rs, 14)] = KP500000000 * (FNMS(KP923879532, T7o, T7n)); Rp[WS(rs, 6)] = KP500000000 * (FMA(KP923879532, T7i, T7h)); Rm[WS(rs, 9)] = KP500000000 * (FNMS(KP923879532, T7i, T7h)); } } } { E T9x, T9T, T8L, T7z, T97, T9J, T8V, T8z, T8M, T8C, T8W, T7O, T9O, T9Y, T9E; E T9t, T8Q, T90, T8G, T88, T8p, T8m, T9K, T9A, T9U, T9e, T8R, T8j, T9R, T9Z; E T9F, T9m; { E T9c, T9b, T99, T98, T7S, T86, T83, T9q, T9M, T9p, T9r, T7Z, T9z, T9a; { E T95, T7r, T9v, T8v, T8w, T8x, T9w, T7y, T7u, T7x; T95 = T7q + T7p; T7r = T7p - T7q; T9v = T8t - T8u; T8v = T8t + T8u; T8w = T7t + T7s; T7u = T7s - T7t; Im[WS(rs, 9)] = -(KP500000000 * (FNMS(KP923879532, T7m, T7j))); Ip[WS(rs, 6)] = KP500000000 * (FMA(KP923879532, T7m, T7j)); T7x = T7v + T7w; T8x = T7v - T7w; T9w = T7u - T7x; T7y = T7u + T7x; { E T7J, T8A, T7G, T7M; { E T7C, T96, T8y, T7F; T9c = T7A + T7B; T7C = T7A - T7B; T9x = FMA(KP707106781, T9w, T9v); T9T = FNMS(KP707106781, T9w, T9v); T8L = FNMS(KP707106781, T7y, T7r); T7z = FMA(KP707106781, T7y, T7r); T96 = T8x - T8w; T8y = T8w + T8x; T7F = T7D + T7E; T9b = T7D - T7E; T99 = T7H + T7I; T7J = T7H - T7I; T97 = FMA(KP707106781, T96, T95); T9J = FNMS(KP707106781, T96, T95); T8V = FNMS(KP707106781, T8y, T8v); T8z = FMA(KP707106781, T8y, T8v); T8A = FMA(KP414213562, T7C, T7F); T7G = FNMS(KP414213562, T7F, T7C); T7M = T7K + T7L; T98 = T7K - T7L; } { E T9n, T9o, T8B, T7N; T7S = T7Q + T7R; T9n = T7R - T7Q; T9o = T85 - T84; T86 = T84 + T85; T83 = T81 + T82; T9q = T81 - T82; T8B = FNMS(KP414213562, T7J, T7M); T7N = FMA(KP414213562, T7M, T7J); T9M = FMA(KP707106781, T9o, T9n); T9p = FNMS(KP707106781, T9o, T9n); T8M = T8B - T8A; T8C = T8A + T8B; T8W = T7G - T7N; T7O = T7G + T7N; T9r = T7Y - T7V; T7Z = T7V + T7Y; } } } { E T8O, T80, T9N, T9s, T8P, T87; T9N = FMA(KP707106781, T9r, T9q); T9s = FNMS(KP707106781, T9r, T9q); T8O = FNMS(KP707106781, T7Z, T7S); T80 = FMA(KP707106781, T7Z, T7S); T9O = FMA(KP198912367, T9N, T9M); T9Y = FNMS(KP198912367, T9M, T9N); T9E = FMA(KP668178637, T9p, T9s); T9t = FNMS(KP668178637, T9s, T9p); T8P = FNMS(KP707106781, T86, T83); T87 = FMA(KP707106781, T86, T83); T9z = FNMS(KP414213562, T98, T99); T9a = FMA(KP414213562, T99, T98); T8Q = FNMS(KP668178637, T8P, T8O); T90 = FMA(KP668178637, T8O, T8P); T8G = FNMS(KP198912367, T80, T87); T88 = FMA(KP198912367, T87, T80); } { E T8b, T9j, T9P, T9i, T9k, T8i, T9Q, T9l; { E T9g, T9h, T9y, T9d; T8b = T89 - T8a; T9g = T8a + T89; T9h = T8n - T8o; T8p = T8n + T8o; T8m = T8k + T8l; T9j = T8l - T8k; T9y = FMA(KP414213562, T9b, T9c); T9d = FNMS(KP414213562, T9c, T9b); T9P = FMA(KP707106781, T9h, T9g); T9i = FNMS(KP707106781, T9h, T9g); T9K = T9y + T9z; T9A = T9y - T9z; T9U = T9d + T9a; T9e = T9a - T9d; T9k = T8h - T8e; T8i = T8e + T8h; } T9Q = FMA(KP707106781, T9k, T9j); T9l = FNMS(KP707106781, T9k, T9j); T8R = FNMS(KP707106781, T8i, T8b); T8j = FMA(KP707106781, T8i, T8b); T9R = FMA(KP198912367, T9Q, T9P); T9Z = FNMS(KP198912367, T9P, T9Q); T9F = FMA(KP668178637, T9i, T9l); T9m = FNMS(KP668178637, T9l, T9i); } } { E T8Z, T92, T9D, T9G; { E T8F, T7P, T8J, T8D, T8S, T8q; T8F = FNMS(KP923879532, T7O, T7z); T7P = FMA(KP923879532, T7O, T7z); T8J = FMA(KP923879532, T8C, T8z); T8D = FNMS(KP923879532, T8C, T8z); T8S = FNMS(KP707106781, T8p, T8m); T8q = FMA(KP707106781, T8p, T8m); { E T8Y, T8X, T93, T94; { E T8N, T91, T8H, T8r, T8U, T8T; T8Z = FMA(KP923879532, T8M, T8L); T8N = FNMS(KP923879532, T8M, T8L); T8T = FMA(KP668178637, T8S, T8R); T91 = FNMS(KP668178637, T8R, T8S); T8H = FMA(KP198912367, T8j, T8q); T8r = FNMS(KP198912367, T8q, T8j); T8U = T8Q + T8T; T8Y = T8T - T8Q; T8X = FMA(KP923879532, T8W, T8V); T93 = FNMS(KP923879532, T8W, T8V); { E T8I, T8K, T8E, T8s; T8I = T8G - T8H; T8K = T8G + T8H; T8E = T8r - T88; T8s = T88 + T8r; Im[WS(rs, 2)] = -(KP500000000 * (FMA(KP831469612, T8U, T8N))); Ip[WS(rs, 13)] = KP500000000 * (FNMS(KP831469612, T8U, T8N)); Im[WS(rs, 6)] = -(KP500000000 * (FNMS(KP980785280, T8I, T8F))); Ip[WS(rs, 9)] = KP500000000 * (FMA(KP980785280, T8I, T8F)); Rp[WS(rs, 1)] = KP500000000 * (FMA(KP980785280, T8K, T8J)); Rm[WS(rs, 14)] = KP500000000 * (FNMS(KP980785280, T8K, T8J)); Rp[WS(rs, 9)] = KP500000000 * (FMA(KP980785280, T8E, T8D)); Rm[WS(rs, 6)] = KP500000000 * (FNMS(KP980785280, T8E, T8D)); Im[WS(rs, 14)] = -(KP500000000 * (FNMS(KP980785280, T8s, T7P))); Ip[WS(rs, 1)] = KP500000000 * (FMA(KP980785280, T8s, T7P)); T94 = T90 + T91; T92 = T90 - T91; } } Rm[WS(rs, 2)] = KP500000000 * (FMA(KP831469612, T94, T93)); Rp[WS(rs, 13)] = KP500000000 * (FNMS(KP831469612, T94, T93)); Rp[WS(rs, 5)] = KP500000000 * (FMA(KP831469612, T8Y, T8X)); Rm[WS(rs, 10)] = KP500000000 * (FNMS(KP831469612, T8Y, T8X)); } } { E T9C, T9B, T9H, T9I, T9f, T9u; T9D = FNMS(KP923879532, T9e, T97); T9f = FMA(KP923879532, T9e, T97); T9u = T9m - T9t; T9C = T9t + T9m; T9B = FNMS(KP923879532, T9A, T9x); T9H = FMA(KP923879532, T9A, T9x); Im[WS(rs, 10)] = -(KP500000000 * (FNMS(KP831469612, T92, T8Z))); Ip[WS(rs, 5)] = KP500000000 * (FMA(KP831469612, T92, T8Z)); Im[WS(rs, 12)] = -(KP500000000 * (FNMS(KP831469612, T9u, T9f))); Ip[WS(rs, 3)] = KP500000000 * (FMA(KP831469612, T9u, T9f)); T9I = T9E + T9F; T9G = T9E - T9F; Rp[WS(rs, 3)] = KP500000000 * (FMA(KP831469612, T9I, T9H)); Rm[WS(rs, 12)] = KP500000000 * (FNMS(KP831469612, T9I, T9H)); Rp[WS(rs, 11)] = KP500000000 * (FMA(KP831469612, T9C, T9B)); Rm[WS(rs, 4)] = KP500000000 * (FNMS(KP831469612, T9C, T9B)); } { E T9W, T9V, Ta1, Ta2, T9L, T9S; T9X = FNMS(KP923879532, T9K, T9J); T9L = FMA(KP923879532, T9K, T9J); T9S = T9O - T9R; T9W = T9O + T9R; T9V = FNMS(KP923879532, T9U, T9T); Ta1 = FMA(KP923879532, T9U, T9T); Im[WS(rs, 4)] = -(KP500000000 * (FNMS(KP831469612, T9G, T9D))); Ip[WS(rs, 11)] = KP500000000 * (FMA(KP831469612, T9G, T9D)); Im[0] = -(KP500000000 * (FNMS(KP980785280, T9S, T9L))); Ip[WS(rs, 15)] = KP500000000 * (FMA(KP980785280, T9S, T9L)); Ta2 = T9Y + T9Z; Ta0 = T9Y - T9Z; Rm[0] = KP500000000 * (FMA(KP980785280, Ta2, Ta1)); Rp[WS(rs, 15)] = KP500000000 * (FNMS(KP980785280, Ta2, Ta1)); Rp[WS(rs, 7)] = KP500000000 * (FMA(KP980785280, T9W, T9V)); Rm[WS(rs, 8)] = KP500000000 * (FNMS(KP980785280, T9W, T9V)); } } } } } Im[WS(rs, 8)] = -(KP500000000 * (FNMS(KP980785280, Ta0, T9X))); Ip[WS(rs, 7)] = KP500000000 * (FMA(KP980785280, Ta0, T9X)); } } } static const tw_instr twinstr[] = { {TW_FULL, 1, 32}, {TW_NEXT, 1, 0} }; static const hc2c_desc desc = { 32, "hc2cfdft_32", twinstr, &GENUS, {300, 126, 198, 0} }; void X(codelet_hc2cfdft_32) (planner *p) { X(khc2c_register) (p, hc2cfdft_32, &desc, HC2C_VIA_DFT); } #else /* HAVE_FMA */ /* Generated by: ../../../genfft/gen_hc2cdft.native -compact -variables 4 -pipeline-latency 4 -n 32 -dit -name hc2cfdft_32 -include hc2cf.h */ /* * This function contains 498 FP additions, 228 FP multiplications, * (or, 404 additions, 134 multiplications, 94 fused multiply/add), * 106 stack variables, 9 constants, and 128 memory accesses */ #include "hc2cf.h" static void hc2cfdft_32(R *Rp, R *Ip, R *Rm, R *Im, const R *W, stride rs, INT mb, INT me, INT ms) { DK(KP277785116, +0.277785116509801112371415406974266437187468595); DK(KP415734806, +0.415734806151272618539394188808952878369280406); DK(KP097545161, +0.097545161008064133924142434238511120463845809); DK(KP490392640, +0.490392640201615224563091118067119518486966865); DK(KP707106781, +0.707106781186547524400844362104849039284835938); DK(KP191341716, +0.191341716182544885864229992015199433380672281); DK(KP461939766, +0.461939766255643378064091594698394143411208313); DK(KP353553390, +0.353553390593273762200422181052424519642417969); DK(KP500000000, +0.500000000000000000000000000000000000000000000); { INT m; for (m = mb, W = W + ((mb - 1) * 62); m < me; m = m + 1, Rp = Rp + ms, Ip = Ip + ms, Rm = Rm - ms, Im = Im - ms, W = W + 62, MAKE_VOLATILE_STRIDE(128, rs)) { E T2S, T5K, T52, T5N, T7p, T8r, T7i, T8o, T2q, T7t, T45, T6L, T2d, T7u, T48; E T6M, T1A, T4c, T4f, T1T, T3f, T5M, T7e, T7l, T6J, T7x, T4V, T5J, T7b, T7k; E T6G, T7w, Tj, TC, T5r, T4k, T4n, T5s, T3D, T5C, T6V, T72, T4G, T5F, T6u; E T86, T6S, T71, T6r, T85, TW, T1f, T5v, T4r, T4u, T5u, T40, T5G, T76, T8k; E T4N, T5D, T6B, T89, T6Z, T8h, T6y, T88; { E T1Y, T22, T2L, T4W, T2p, T43, T2A, T50, T27, T2b, T2Q, T4X, T2h, T2l, T2F; E T4Z; { E T1W, T1X, T2K, T20, T21, T2I, T2H, T2J; T1W = Ip[WS(rs, 4)]; T1X = Im[WS(rs, 4)]; T2K = T1W + T1X; T20 = Rp[WS(rs, 4)]; T21 = Rm[WS(rs, 4)]; T2I = T20 - T21; T1Y = T1W - T1X; T22 = T20 + T21; T2H = W[16]; T2J = W[17]; T2L = FMA(T2H, T2I, T2J * T2K); T4W = FNMS(T2J, T2I, T2H * T2K); } { E T2n, T2o, T2z, T2v, T2w, T2x, T2u, T2y; T2n = Ip[0]; T2o = Im[0]; T2z = T2n + T2o; T2v = Rm[0]; T2w = Rp[0]; T2x = T2v - T2w; T2p = T2n - T2o; T43 = T2w + T2v; T2u = W[0]; T2y = W[1]; T2A = FNMS(T2y, T2z, T2u * T2x); T50 = FMA(T2y, T2x, T2u * T2z); } { E T25, T26, T2P, T29, T2a, T2N, T2M, T2O; T25 = Ip[WS(rs, 12)]; T26 = Im[WS(rs, 12)]; T2P = T25 + T26; T29 = Rp[WS(rs, 12)]; T2a = Rm[WS(rs, 12)]; T2N = T29 - T2a; T27 = T25 - T26; T2b = T29 + T2a; T2M = W[48]; T2O = W[49]; T2Q = FMA(T2M, T2N, T2O * T2P); T4X = FNMS(T2O, T2N, T2M * T2P); } { E T2f, T2g, T2E, T2j, T2k, T2C, T2B, T2D; T2f = Ip[WS(rs, 8)]; T2g = Im[WS(rs, 8)]; T2E = T2f + T2g; T2j = Rp[WS(rs, 8)]; T2k = Rm[WS(rs, 8)]; T2C = T2j - T2k; T2h = T2f - T2g; T2l = T2j + T2k; T2B = W[32]; T2D = W[33]; T2F = FMA(T2B, T2C, T2D * T2E); T4Z = FNMS(T2D, T2C, T2B * T2E); } { E T2G, T2R, T7g, T7h; T2G = T2A - T2F; T2R = T2L + T2Q; T2S = T2G - T2R; T5K = T2R + T2G; { E T4Y, T51, T7n, T7o; T4Y = T4W + T4X; T51 = T4Z + T50; T52 = T4Y + T51; T5N = T51 - T4Y; T7n = T2Q - T2L; T7o = T50 - T4Z; T7p = T7n + T7o; T8r = T7o - T7n; } T7g = T2F + T2A; T7h = T4W - T4X; T7i = T7g - T7h; T8o = T7h + T7g; { E T2m, T44, T2e, T2i; T2e = W[30]; T2i = W[31]; T2m = FNMS(T2i, T2l, T2e * T2h); T44 = FMA(T2e, T2l, T2i * T2h); T2q = T2m + T2p; T7t = T43 - T44; T45 = T43 + T44; T6L = T2p - T2m; } { E T23, T46, T2c, T47; { E T1V, T1Z, T24, T28; T1V = W[14]; T1Z = W[15]; T23 = FNMS(T1Z, T22, T1V * T1Y); T46 = FMA(T1V, T22, T1Z * T1Y); T24 = W[46]; T28 = W[47]; T2c = FNMS(T28, T2b, T24 * T27); T47 = FMA(T24, T2b, T28 * T27); } T2d = T23 + T2c; T7u = T23 - T2c; T48 = T46 + T47; T6M = T46 - T47; } } } { E T1q, T4a, T2X, T4P, T1S, T4e, T3d, T4T, T1z, T4b, T32, T4Q, T1J, T4d, T38; E T4S; { E T1l, T2W, T1p, T2U; { E T1j, T1k, T1n, T1o; T1j = Ip[WS(rs, 2)]; T1k = Im[WS(rs, 2)]; T1l = T1j - T1k; T2W = T1j + T1k; T1n = Rp[WS(rs, 2)]; T1o = Rm[WS(rs, 2)]; T1p = T1n + T1o; T2U = T1n - T1o; } { E T1i, T1m, T2T, T2V; T1i = W[6]; T1m = W[7]; T1q = FNMS(T1m, T1p, T1i * T1l); T4a = FMA(T1i, T1p, T1m * T1l); T2T = W[8]; T2V = W[9]; T2X = FMA(T2T, T2U, T2V * T2W); T4P = FNMS(T2V, T2U, T2T * T2W); } } { E T1N, T3c, T1R, T3a; { E T1L, T1M, T1P, T1Q; T1L = Ip[WS(rs, 6)]; T1M = Im[WS(rs, 6)]; T1N = T1L - T1M; T3c = T1L + T1M; T1P = Rp[WS(rs, 6)]; T1Q = Rm[WS(rs, 6)]; T1R = T1P + T1Q; T3a = T1P - T1Q; } { E T1K, T1O, T39, T3b; T1K = W[22]; T1O = W[23]; T1S = FNMS(T1O, T1R, T1K * T1N); T4e = FMA(T1K, T1R, T1O * T1N); T39 = W[24]; T3b = W[25]; T3d = FMA(T39, T3a, T3b * T3c); T4T = FNMS(T3b, T3a, T39 * T3c); } } { E T1u, T31, T1y, T2Z; { E T1s, T1t, T1w, T1x; T1s = Ip[WS(rs, 10)]; T1t = Im[WS(rs, 10)]; T1u = T1s - T1t; T31 = T1s + T1t; T1w = Rp[WS(rs, 10)]; T1x = Rm[WS(rs, 10)]; T1y = T1w + T1x; T2Z = T1w - T1x; } { E T1r, T1v, T2Y, T30; T1r = W[38]; T1v = W[39]; T1z = FNMS(T1v, T1y, T1r * T1u); T4b = FMA(T1r, T1y, T1v * T1u); T2Y = W[40]; T30 = W[41]; T32 = FMA(T2Y, T2Z, T30 * T31); T4Q = FNMS(T30, T2Z, T2Y * T31); } } { E T1E, T37, T1I, T35; { E T1C, T1D, T1G, T1H; T1C = Ip[WS(rs, 14)]; T1D = Im[WS(rs, 14)]; T1E = T1C - T1D; T37 = T1C + T1D; T1G = Rp[WS(rs, 14)]; T1H = Rm[WS(rs, 14)]; T1I = T1G + T1H; T35 = T1G - T1H; } { E T1B, T1F, T34, T36; T1B = W[54]; T1F = W[55]; T1J = FNMS(T1F, T1I, T1B * T1E); T4d = FMA(T1B, T1I, T1F * T1E); T34 = W[56]; T36 = W[57]; T38 = FMA(T34, T35, T36 * T37); T4S = FNMS(T36, T35, T34 * T37); } } { E T33, T3e, T4R, T4U; T1A = T1q + T1z; T4c = T4a + T4b; T4f = T4d + T4e; T1T = T1J + T1S; T33 = T2X + T32; T3e = T38 + T3d; T3f = T33 + T3e; T5M = T3e - T33; { E T7c, T7d, T6H, T6I; T7c = T4S - T4T; T7d = T3d - T38; T7e = T7c + T7d; T7l = T7c - T7d; T6H = T4d - T4e; T6I = T1J - T1S; T6J = T6H + T6I; T7x = T6H - T6I; } T4R = T4P + T4Q; T4U = T4S + T4T; T4V = T4R + T4U; T5J = T4U - T4R; { E T79, T7a, T6E, T6F; T79 = T32 - T2X; T7a = T4P - T4Q; T7b = T79 - T7a; T7k = T7a + T79; T6E = T1q - T1z; T6F = T4a - T4b; T6G = T6E - T6F; T7w = T6F + T6E; } } } { E T9, T4i, T3l, T4A, TB, T4m, T3B, T4E, Ti, T4j, T3q, T4B, Ts, T4l, T3w; E T4D; { E T4, T3k, T8, T3i; { E T2, T3, T6, T7; T2 = Ip[WS(rs, 1)]; T3 = Im[WS(rs, 1)]; T4 = T2 - T3; T3k = T2 + T3; T6 = Rp[WS(rs, 1)]; T7 = Rm[WS(rs, 1)]; T8 = T6 + T7; T3i = T6 - T7; } { E T1, T5, T3h, T3j; T1 = W[2]; T5 = W[3]; T9 = FNMS(T5, T8, T1 * T4); T4i = FMA(T1, T8, T5 * T4); T3h = W[4]; T3j = W[5]; T3l = FMA(T3h, T3i, T3j * T3k); T4A = FNMS(T3j, T3i, T3h * T3k); } } { E Tw, T3A, TA, T3y; { E Tu, Tv, Ty, Tz; Tu = Ip[WS(rs, 13)]; Tv = Im[WS(rs, 13)]; Tw = Tu - Tv; T3A = Tu + Tv; Ty = Rp[WS(rs, 13)]; Tz = Rm[WS(rs, 13)]; TA = Ty + Tz; T3y = Ty - Tz; } { E Tt, Tx, T3x, T3z; Tt = W[50]; Tx = W[51]; TB = FNMS(Tx, TA, Tt * Tw); T4m = FMA(Tt, TA, Tx * Tw); T3x = W[52]; T3z = W[53]; T3B = FMA(T3x, T3y, T3z * T3A); T4E = FNMS(T3z, T3y, T3x * T3A); } } { E Td, T3p, Th, T3n; { E Tb, Tc, Tf, Tg; Tb = Ip[WS(rs, 9)]; Tc = Im[WS(rs, 9)]; Td = Tb - Tc; T3p = Tb + Tc; Tf = Rp[WS(rs, 9)]; Tg = Rm[WS(rs, 9)]; Th = Tf + Tg; T3n = Tf - Tg; } { E Ta, Te, T3m, T3o; Ta = W[34]; Te = W[35]; Ti = FNMS(Te, Th, Ta * Td); T4j = FMA(Ta, Th, Te * Td); T3m = W[36]; T3o = W[37]; T3q = FMA(T3m, T3n, T3o * T3p); T4B = FNMS(T3o, T3n, T3m * T3p); } } { E Tn, T3v, Tr, T3t; { E Tl, Tm, Tp, Tq; Tl = Ip[WS(rs, 5)]; Tm = Im[WS(rs, 5)]; Tn = Tl - Tm; T3v = Tl + Tm; Tp = Rp[WS(rs, 5)]; Tq = Rm[WS(rs, 5)]; Tr = Tp + Tq; T3t = Tp - Tq; } { E Tk, To, T3s, T3u; Tk = W[18]; To = W[19]; Ts = FNMS(To, Tr, Tk * Tn); T4l = FMA(Tk, Tr, To * Tn); T3s = W[20]; T3u = W[21]; T3w = FMA(T3s, T3t, T3u * T3v); T4D = FNMS(T3u, T3t, T3s * T3v); } } Tj = T9 + Ti; TC = Ts + TB; T5r = Tj - TC; T4k = T4i + T4j; T4n = T4l + T4m; T5s = T4k - T4n; { E T3r, T3C, T6T, T6U; T3r = T3l + T3q; T3C = T3w + T3B; T3D = T3r + T3C; T5C = T3C - T3r; T6T = T4E - T4D; T6U = T3w - T3B; T6V = T6T + T6U; T72 = T6T - T6U; } { E T4C, T4F, T6s, T6t; T4C = T4A + T4B; T4F = T4D + T4E; T4G = T4C + T4F; T5F = T4F - T4C; T6s = T4i - T4j; T6t = Ts - TB; T6u = T6s + T6t; T86 = T6s - T6t; } { E T6Q, T6R, T6p, T6q; T6Q = T3q - T3l; T6R = T4A - T4B; T6S = T6Q - T6R; T71 = T6R + T6Q; T6p = T9 - Ti; T6q = T4l - T4m; T6r = T6p - T6q; T85 = T6p + T6q; } } { E TM, T4p, T3I, T4H, T1e, T4t, T3Y, T4L, TV, T4q, T3N, T4I, T15, T4s, T3T; E T4K; { E TH, T3H, TL, T3F; { E TF, TG, TJ, TK; TF = Ip[WS(rs, 15)]; TG = Im[WS(rs, 15)]; TH = TF - TG; T3H = TF + TG; TJ = Rp[WS(rs, 15)]; TK = Rm[WS(rs, 15)]; TL = TJ + TK; T3F = TJ - TK; } { E TE, TI, T3E, T3G; TE = W[58]; TI = W[59]; TM = FNMS(TI, TL, TE * TH); T4p = FMA(TE, TL, TI * TH); T3E = W[60]; T3G = W[61]; T3I = FMA(T3E, T3F, T3G * T3H); T4H = FNMS(T3G, T3F, T3E * T3H); } } { E T19, T3X, T1d, T3V; { E T17, T18, T1b, T1c; T17 = Ip[WS(rs, 11)]; T18 = Im[WS(rs, 11)]; T19 = T17 - T18; T3X = T17 + T18; T1b = Rp[WS(rs, 11)]; T1c = Rm[WS(rs, 11)]; T1d = T1b + T1c; T3V = T1b - T1c; } { E T16, T1a, T3U, T3W; T16 = W[42]; T1a = W[43]; T1e = FNMS(T1a, T1d, T16 * T19); T4t = FMA(T16, T1d, T1a * T19); T3U = W[44]; T3W = W[45]; T3Y = FMA(T3U, T3V, T3W * T3X); T4L = FNMS(T3W, T3V, T3U * T3X); } } { E TQ, T3M, TU, T3K; { E TO, TP, TS, TT; TO = Ip[WS(rs, 7)]; TP = Im[WS(rs, 7)]; TQ = TO - TP; T3M = TO + TP; TS = Rp[WS(rs, 7)]; TT = Rm[WS(rs, 7)]; TU = TS + TT; T3K = TS - TT; } { E TN, TR, T3J, T3L; TN = W[26]; TR = W[27]; TV = FNMS(TR, TU, TN * TQ); T4q = FMA(TN, TU, TR * TQ); T3J = W[28]; T3L = W[29]; T3N = FMA(T3J, T3K, T3L * T3M); T4I = FNMS(T3L, T3K, T3J * T3M); } } { E T10, T3S, T14, T3Q; { E TY, TZ, T12, T13; TY = Ip[WS(rs, 3)]; TZ = Im[WS(rs, 3)]; T10 = TY - TZ; T3S = TY + TZ; T12 = Rp[WS(rs, 3)]; T13 = Rm[WS(rs, 3)]; T14 = T12 + T13; T3Q = T12 - T13; } { E TX, T11, T3P, T3R; TX = W[10]; T11 = W[11]; T15 = FNMS(T11, T14, TX * T10); T4s = FMA(TX, T14, T11 * T10); T3P = W[12]; T3R = W[13]; T3T = FMA(T3P, T3Q, T3R * T3S); T4K = FNMS(T3R, T3Q, T3P * T3S); } } TW = TM + TV; T1f = T15 + T1e; T5v = TW - T1f; T4r = T4p + T4q; T4u = T4s + T4t; T5u = T4r - T4u; { E T3O, T3Z, T74, T75; T3O = T3I + T3N; T3Z = T3T + T3Y; T40 = T3O + T3Z; T5G = T3Z - T3O; T74 = T4H - T4I; T75 = T3Y - T3T; T76 = T74 + T75; T8k = T74 - T75; } { E T4J, T4M, T6z, T6A; T4J = T4H + T4I; T4M = T4K + T4L; T4N = T4J + T4M; T5D = T4J - T4M; T6z = T4p - T4q; T6A = T15 - T1e; T6B = T6z + T6A; T89 = T6z - T6A; } { E T6X, T6Y, T6w, T6x; T6X = T3N - T3I; T6Y = T4K - T4L; T6Z = T6X - T6Y; T8h = T6X + T6Y; T6w = TM - TV; T6x = T4s - T4t; T6y = T6w - T6x; T88 = T6w + T6x; } } { E T1h, T5i, T5c, T5m, T5f, T5n, T2s, T58, T42, T4y, T4w, T57, T54, T56, T4h; E T5h; { E TD, T1g, T5a, T5b; TD = Tj + TC; T1g = TW + T1f; T1h = TD + T1g; T5i = TD - T1g; T5a = T4N - T4G; T5b = T3D - T40; T5c = T5a + T5b; T5m = T5a - T5b; } { E T5d, T5e, T1U, T2r; T5d = T3f + T2S; T5e = T52 - T4V; T5f = T5d - T5e; T5n = T5d + T5e; T1U = T1A + T1T; T2r = T2d + T2q; T2s = T1U + T2r; T58 = T2r - T1U; } { E T3g, T41, T4o, T4v; T3g = T2S - T3f; T41 = T3D + T40; T42 = T3g - T41; T4y = T41 + T3g; T4o = T4k + T4n; T4v = T4r + T4u; T4w = T4o + T4v; T57 = T4v - T4o; } { E T4O, T53, T49, T4g; T4O = T4G + T4N; T53 = T4V + T52; T54 = T4O - T53; T56 = T4O + T53; T49 = T45 + T48; T4g = T4c + T4f; T4h = T49 + T4g; T5h = T49 - T4g; } { E T2t, T55, T4x, T4z; T2t = T1h + T2s; Ip[0] = KP500000000 * (T2t + T42); Im[WS(rs, 15)] = KP500000000 * (T42 - T2t); T55 = T4h + T4w; Rm[WS(rs, 15)] = KP500000000 * (T55 - T56); Rp[0] = KP500000000 * (T55 + T56); T4x = T4h - T4w; Rm[WS(rs, 7)] = KP500000000 * (T4x - T4y); Rp[WS(rs, 8)] = KP500000000 * (T4x + T4y); T4z = T2s - T1h; Ip[WS(rs, 8)] = KP500000000 * (T4z + T54); Im[WS(rs, 7)] = KP500000000 * (T54 - T4z); } { E T59, T5g, T5p, T5q; T59 = KP500000000 * (T57 + T58); T5g = KP353553390 * (T5c + T5f); Ip[WS(rs, 4)] = T59 + T5g; Im[WS(rs, 11)] = T5g - T59; T5p = KP500000000 * (T5h + T5i); T5q = KP353553390 * (T5m + T5n); Rm[WS(rs, 11)] = T5p - T5q; Rp[WS(rs, 4)] = T5p + T5q; } { E T5j, T5k, T5l, T5o; T5j = KP500000000 * (T5h - T5i); T5k = KP353553390 * (T5f - T5c); Rm[WS(rs, 3)] = T5j - T5k; Rp[WS(rs, 12)] = T5j + T5k; T5l = KP500000000 * (T58 - T57); T5o = KP353553390 * (T5m - T5n); Ip[WS(rs, 12)] = T5l + T5o; Im[WS(rs, 3)] = T5o - T5l; } } { E T5x, T6g, T6a, T6k, T6d, T6l, T5A, T66, T5I, T60, T5T, T6f, T5W, T65, T5P; E T61; { E T5t, T5w, T68, T69; T5t = T5r - T5s; T5w = T5u + T5v; T5x = KP353553390 * (T5t + T5w); T6g = KP353553390 * (T5t - T5w); T68 = T5D - T5C; T69 = T5G - T5F; T6a = FMA(KP461939766, T68, KP191341716 * T69); T6k = FNMS(KP461939766, T69, KP191341716 * T68); } { E T6b, T6c, T5y, T5z; T6b = T5K - T5J; T6c = T5N - T5M; T6d = FNMS(KP461939766, T6c, KP191341716 * T6b); T6l = FMA(KP461939766, T6b, KP191341716 * T6c); T5y = T4f - T4c; T5z = T2q - T2d; T5A = KP500000000 * (T5y + T5z); T66 = KP500000000 * (T5z - T5y); } { E T5E, T5H, T5R, T5S; T5E = T5C + T5D; T5H = T5F + T5G; T5I = FMA(KP191341716, T5E, KP461939766 * T5H); T60 = FNMS(KP191341716, T5H, KP461939766 * T5E); T5R = T45 - T48; T5S = T1A - T1T; T5T = KP500000000 * (T5R + T5S); T6f = KP500000000 * (T5R - T5S); } { E T5U, T5V, T5L, T5O; T5U = T5s + T5r; T5V = T5u - T5v; T5W = KP353553390 * (T5U + T5V); T65 = KP353553390 * (T5V - T5U); T5L = T5J + T5K; T5O = T5M + T5N; T5P = FNMS(KP191341716, T5O, KP461939766 * T5L); T61 = FMA(KP191341716, T5L, KP461939766 * T5O); } { E T5B, T5Q, T63, T64; T5B = T5x + T5A; T5Q = T5I + T5P; Ip[WS(rs, 2)] = T5B + T5Q; Im[WS(rs, 13)] = T5Q - T5B; T63 = T5T + T5W; T64 = T60 + T61; Rm[WS(rs, 13)] = T63 - T64; Rp[WS(rs, 2)] = T63 + T64; } { E T5X, T5Y, T5Z, T62; T5X = T5T - T5W; T5Y = T5P - T5I; Rm[WS(rs, 5)] = T5X - T5Y; Rp[WS(rs, 10)] = T5X + T5Y; T5Z = T5A - T5x; T62 = T60 - T61; Ip[WS(rs, 10)] = T5Z + T62; Im[WS(rs, 5)] = T62 - T5Z; } { E T67, T6e, T6n, T6o; T67 = T65 + T66; T6e = T6a + T6d; Ip[WS(rs, 6)] = T67 + T6e; Im[WS(rs, 9)] = T6e - T67; T6n = T6f + T6g; T6o = T6k + T6l; Rm[WS(rs, 9)] = T6n - T6o; Rp[WS(rs, 6)] = T6n + T6o; } { E T6h, T6i, T6j, T6m; T6h = T6f - T6g; T6i = T6d - T6a; Rm[WS(rs, 1)] = T6h - T6i; Rp[WS(rs, 14)] = T6h + T6i; T6j = T66 - T65; T6m = T6k - T6l; Ip[WS(rs, 14)] = T6j + T6m; Im[WS(rs, 1)] = T6m - T6j; } } { E T6D, T7W, T6O, T7M, T7C, T7L, T7z, T7V, T7r, T81, T7H, T7T, T78, T80, T7G; E T7Q; { E T6v, T6C, T7v, T7y; T6v = FNMS(KP191341716, T6u, KP461939766 * T6r); T6C = FMA(KP461939766, T6y, KP191341716 * T6B); T6D = T6v + T6C; T7W = T6v - T6C; { E T6K, T6N, T7A, T7B; T6K = KP353553390 * (T6G + T6J); T6N = KP500000000 * (T6L - T6M); T6O = T6K + T6N; T7M = T6N - T6K; T7A = FMA(KP191341716, T6r, KP461939766 * T6u); T7B = FNMS(KP191341716, T6y, KP461939766 * T6B); T7C = T7A + T7B; T7L = T7B - T7A; } T7v = KP500000000 * (T7t + T7u); T7y = KP353553390 * (T7w + T7x); T7z = T7v + T7y; T7V = T7v - T7y; { E T7j, T7R, T7q, T7S, T7f, T7m; T7f = KP707106781 * (T7b + T7e); T7j = T7f + T7i; T7R = T7i - T7f; T7m = KP707106781 * (T7k + T7l); T7q = T7m + T7p; T7S = T7p - T7m; T7r = FNMS(KP097545161, T7q, KP490392640 * T7j); T81 = FMA(KP415734806, T7R, KP277785116 * T7S); T7H = FMA(KP097545161, T7j, KP490392640 * T7q); T7T = FNMS(KP415734806, T7S, KP277785116 * T7R); } { E T70, T7O, T77, T7P, T6W, T73; T6W = KP707106781 * (T6S + T6V); T70 = T6W + T6Z; T7O = T6Z - T6W; T73 = KP707106781 * (T71 + T72); T77 = T73 + T76; T7P = T76 - T73; T78 = FMA(KP490392640, T70, KP097545161 * T77); T80 = FNMS(KP415734806, T7O, KP277785116 * T7P); T7G = FNMS(KP097545161, T70, KP490392640 * T77); T7Q = FMA(KP277785116, T7O, KP415734806 * T7P); } } { E T6P, T7s, T7J, T7K; T6P = T6D + T6O; T7s = T78 + T7r; Ip[WS(rs, 1)] = T6P + T7s; Im[WS(rs, 14)] = T7s - T6P; T7J = T7z + T7C; T7K = T7G + T7H; Rm[WS(rs, 14)] = T7J - T7K; Rp[WS(rs, 1)] = T7J + T7K; } { E T7D, T7E, T7F, T7I; T7D = T7z - T7C; T7E = T7r - T78; Rm[WS(rs, 6)] = T7D - T7E; Rp[WS(rs, 9)] = T7D + T7E; T7F = T6O - T6D; T7I = T7G - T7H; Ip[WS(rs, 9)] = T7F + T7I; Im[WS(rs, 6)] = T7I - T7F; } { E T7N, T7U, T83, T84; T7N = T7L + T7M; T7U = T7Q + T7T; Ip[WS(rs, 5)] = T7N + T7U; Im[WS(rs, 10)] = T7U - T7N; T83 = T7V + T7W; T84 = T80 + T81; Rm[WS(rs, 10)] = T83 - T84; Rp[WS(rs, 5)] = T83 + T84; } { E T7X, T7Y, T7Z, T82; T7X = T7V - T7W; T7Y = T7T - T7Q; Rm[WS(rs, 2)] = T7X - T7Y; Rp[WS(rs, 13)] = T7X + T7Y; T7Z = T7M - T7L; T82 = T80 - T81; Ip[WS(rs, 13)] = T7Z + T82; Im[WS(rs, 2)] = T82 - T7Z; } } { E T8b, T8U, T8e, T8K, T8A, T8J, T8x, T8T, T8t, T8Z, T8F, T8R, T8m, T8Y, T8E; E T8O; { E T87, T8a, T8v, T8w; T87 = FNMS(KP461939766, T86, KP191341716 * T85); T8a = FMA(KP191341716, T88, KP461939766 * T89); T8b = T87 + T8a; T8U = T87 - T8a; { E T8c, T8d, T8y, T8z; T8c = KP353553390 * (T7x - T7w); T8d = KP500000000 * (T6M + T6L); T8e = T8c + T8d; T8K = T8d - T8c; T8y = FMA(KP461939766, T85, KP191341716 * T86); T8z = FNMS(KP461939766, T88, KP191341716 * T89); T8A = T8y + T8z; T8J = T8z - T8y; } T8v = KP500000000 * (T7t - T7u); T8w = KP353553390 * (T6G - T6J); T8x = T8v + T8w; T8T = T8v - T8w; { E T8p, T8P, T8s, T8Q, T8n, T8q; T8n = KP707106781 * (T7l - T7k); T8p = T8n + T8o; T8P = T8o - T8n; T8q = KP707106781 * (T7b - T7e); T8s = T8q + T8r; T8Q = T8r - T8q; T8t = FNMS(KP277785116, T8s, KP415734806 * T8p); T8Z = FMA(KP490392640, T8P, KP097545161 * T8Q); T8F = FMA(KP277785116, T8p, KP415734806 * T8s); T8R = FNMS(KP490392640, T8Q, KP097545161 * T8P); } { E T8i, T8M, T8l, T8N, T8g, T8j; T8g = KP707106781 * (T72 - T71); T8i = T8g + T8h; T8M = T8h - T8g; T8j = KP707106781 * (T6S - T6V); T8l = T8j + T8k; T8N = T8k - T8j; T8m = FMA(KP415734806, T8i, KP277785116 * T8l); T8Y = FNMS(KP490392640, T8M, KP097545161 * T8N); T8E = FNMS(KP277785116, T8i, KP415734806 * T8l); T8O = FMA(KP097545161, T8M, KP490392640 * T8N); } } { E T8f, T8u, T8H, T8I; T8f = T8b + T8e; T8u = T8m + T8t; Ip[WS(rs, 3)] = T8f + T8u; Im[WS(rs, 12)] = T8u - T8f; T8H = T8x + T8A; T8I = T8E + T8F; Rm[WS(rs, 12)] = T8H - T8I; Rp[WS(rs, 3)] = T8H + T8I; } { E T8B, T8C, T8D, T8G; T8B = T8x - T8A; T8C = T8t - T8m; Rm[WS(rs, 4)] = T8B - T8C; Rp[WS(rs, 11)] = T8B + T8C; T8D = T8e - T8b; T8G = T8E - T8F; Ip[WS(rs, 11)] = T8D + T8G; Im[WS(rs, 4)] = T8G - T8D; } { E T8L, T8S, T91, T92; T8L = T8J + T8K; T8S = T8O + T8R; Ip[WS(rs, 7)] = T8L + T8S; Im[WS(rs, 8)] = T8S - T8L; T91 = T8T + T8U; T92 = T8Y + T8Z; Rm[WS(rs, 8)] = T91 - T92; Rp[WS(rs, 7)] = T91 + T92; } { E T8V, T8W, T8X, T90; T8V = T8T - T8U; T8W = T8R - T8O; Rm[0] = T8V - T8W; Rp[WS(rs, 15)] = T8V + T8W; T8X = T8K - T8J; T90 = T8Y - T8Z; Ip[WS(rs, 15)] = T8X + T90; Im[0] = T90 - T8X; } } } } } static const tw_instr twinstr[] = { {TW_FULL, 1, 32}, {TW_NEXT, 1, 0} }; static const hc2c_desc desc = { 32, "hc2cfdft_32", twinstr, &GENUS, {404, 134, 94, 0} }; void X(codelet_hc2cfdft_32) (planner *p) { X(khc2c_register) (p, hc2cfdft_32, &desc, HC2C_VIA_DFT); } #endif /* HAVE_FMA */ fftw-3.3.3/rdft/scalar/r2cf/hc2cfdft_20.c0000644000175000017500000007562112054410722014607 00000000000000/* * Copyright (c) 2003, 2007-11 Matteo Frigo * Copyright (c) 2003, 2007-11 Massachusetts Institute of Technology * * This program is free software; you can redistribute it and/or modify * it under the terms of the GNU General Public License as published by * the Free Software Foundation; either version 2 of the License, or * (at your option) any later version. * * This program is distributed in the hope that it will be useful, * but WITHOUT ANY WARRANTY; without even the implied warranty of * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the * GNU General Public License for more details. * * You should have received a copy of the GNU General Public License * along with this program; if not, write to the Free Software * Foundation, Inc., 51 Franklin Street, Fifth Floor, Boston, MA 02110-1301 USA * */ /* This file was automatically generated --- DO NOT EDIT */ /* Generated on Sun Nov 25 07:40:48 EST 2012 */ #include "codelet-rdft.h" #ifdef HAVE_FMA /* Generated by: ../../../genfft/gen_hc2cdft.native -fma -reorder-insns -schedule-for-pipeline -compact -variables 4 -pipeline-latency 4 -n 20 -dit -name hc2cfdft_20 -include hc2cf.h */ /* * This function contains 286 FP additions, 188 FP multiplications, * (or, 176 additions, 78 multiplications, 110 fused multiply/add), * 174 stack variables, 5 constants, and 80 memory accesses */ #include "hc2cf.h" static void hc2cfdft_20(R *Rp, R *Ip, R *Rm, R *Im, const R *W, stride rs, INT mb, INT me, INT ms) { DK(KP951056516, +0.951056516295153572116439333379382143405698634); DK(KP559016994, +0.559016994374947424102293417182819058860154590); DK(KP250000000, +0.250000000000000000000000000000000000000000000); DK(KP500000000, +0.500000000000000000000000000000000000000000000); DK(KP618033988, +0.618033988749894848204586834365638117720309180); { INT m; for (m = mb, W = W + ((mb - 1) * 38); m < me; m = m + 1, Rp = Rp + ms, Ip = Ip + ms, Rm = Rm - ms, Im = Im - ms, W = W + 38, MAKE_VOLATILE_STRIDE(80, rs)) { E T4X, T5i, T5k, T5e, T5c, T5d, T5j, T5f; { E T2E, T4W, T3v, T4k, T2M, T3w, T4V, T4j, T2p, T2T, T5a, T5A, T3D, T3o, T4b; E T4B, T1Y, T2S, T5z, T57, T3h, T3C, T4A, T44, TH, T2P, T50, T5x, T3z, T32; E T3P, T4D, T3V, T3U, T5w, T53, T2Q, T1o, T3A, T39; { E T1V, T9, T2w, Tu, T1, T6, T1R, T1U, T1T, T2Y, T5, T40, T2F, T10, T2C; E TE, TX, T2m, T1y, T4g, TS, T33, TW, Tw, TB, T2y, T2B, TA, T3L, T2A; E T3t, T1q, T1v, T2i, T2l, T2k, T3d, T1u, T48, Tm, Tr, T2s, T2v, T2u, T3J; E Tq, T3r, T20, T1g, T23, T1l, T1h, T3S, T3k, T21, T2H, TL, T2K, TQ, TM; E T35, T4h, T2I, T2f, T2g, T1I, T1D, T2c, T46, T2e, T3b, T1E, T28, T16, T29; E T1b, T25, T3i, T27, T3Q, T17, T1O, T1P, Tj, T1M, Te, T1L, Tb, T3Y, TV; E T1d, T1Z; { E T1S, T4, T7, T8; T7 = Rp[WS(rs, 9)]; T8 = Rm[WS(rs, 9)]; { E Ts, Tt, T2, T3; Ts = Rp[WS(rs, 2)]; Tt = Rm[WS(rs, 2)]; T2 = Ip[WS(rs, 9)]; T1V = T7 + T8; T9 = T7 - T8; T2w = Ts - Tt; Tu = Ts + Tt; T3 = Im[WS(rs, 9)]; T1 = W[36]; T6 = W[37]; T1R = W[34]; T1S = T2 - T3; T4 = T2 + T3; T1U = W[35]; } { E TY, TZ, TC, TD; TY = Ip[0]; T1T = T1R * T1S; T2Y = T6 * T4; T5 = T1 * T4; T40 = T1U * T1S; TZ = Im[0]; TC = Rp[WS(rs, 7)]; TD = Rm[WS(rs, 7)]; { E T1w, T1x, TT, TU; T1w = Rp[WS(rs, 1)]; T2F = TY - TZ; T10 = TY + TZ; T2C = TC - TD; TE = TC + TD; T1x = Rm[WS(rs, 1)]; TT = Rm[0]; TU = Rp[0]; TX = W[0]; T2m = T1w + T1x; T1y = T1w - T1x; T4g = TU + TT; TV = TT - TU; TS = W[1]; } } } { E T2j, T1t, T1r, T1s; { E Tx, Ty, T2z, Tz; Tx = Ip[WS(rs, 7)]; Ty = Im[WS(rs, 7)]; T33 = TX * TV; TW = TS * TV; Tw = W[26]; T2z = Tx + Ty; Tz = Tx - Ty; TB = W[27]; T2y = W[28]; T2B = W[29]; TA = Tw * Tz; T3L = TB * Tz; T2A = T2y * T2z; T3t = T2B * T2z; } T1r = Ip[WS(rs, 1)]; T1s = Im[WS(rs, 1)]; T1q = W[4]; T1v = W[5]; T2i = W[2]; T2j = T1r - T1s; T1t = T1r + T1s; T2l = W[3]; { E T2t, Tp, Tn, To; Tn = Ip[WS(rs, 2)]; T2k = T2i * T2j; T3d = T1v * T1t; T1u = T1q * T1t; T48 = T2l * T2j; To = Im[WS(rs, 2)]; Tm = W[6]; Tr = W[7]; T2s = W[8]; T2t = Tn + To; Tp = Tn - To; T2v = W[9]; { E T1e, T1f, T1j, T1k; T1e = Ip[WS(rs, 3)]; T2u = T2s * T2t; T3J = Tr * Tp; Tq = Tm * Tp; T3r = T2v * T2t; T1f = Im[WS(rs, 3)]; T1j = Rp[WS(rs, 3)]; T1k = Rm[WS(rs, 3)]; T1d = W[10]; T20 = T1e + T1f; T1g = T1e - T1f; T23 = T1j - T1k; T1l = T1j + T1k; T1Z = W[12]; T1h = T1d * T1g; } } } { E T2d, T1A, TI, T2G, T26, T13; { E TJ, TK, TO, TP; TJ = Ip[WS(rs, 5)]; T3S = T1d * T1l; T3k = T1Z * T23; T21 = T1Z * T20; TK = Im[WS(rs, 5)]; TO = Rp[WS(rs, 5)]; TP = Rm[WS(rs, 5)]; TI = W[20]; T2H = TJ - TK; TL = TJ + TK; T2K = TO + TP; TQ = TO - TP; T2G = W[18]; TM = TI * TL; } { E T1G, T1H, T1B, T1C; T1G = Rm[WS(rs, 6)]; T35 = TI * TQ; T4h = T2G * T2K; T2I = T2G * T2H; T1H = Rp[WS(rs, 6)]; T1B = Ip[WS(rs, 6)]; T1C = Im[WS(rs, 6)]; T2f = W[23]; T2g = T1H + T1G; T1I = T1G - T1H; T2d = T1B - T1C; T1D = T1B + T1C; T2c = W[22]; T1A = W[24]; T46 = T2f * T2d; } { E T14, T15, T19, T1a; T14 = Ip[WS(rs, 8)]; T2e = T2c * T2d; T3b = T1A * T1I; T1E = T1A * T1D; T15 = Im[WS(rs, 8)]; T19 = Rp[WS(rs, 8)]; T1a = Rm[WS(rs, 8)]; T28 = W[32]; T16 = T14 - T15; T29 = T14 + T15; T1b = T19 + T1a; T26 = T1a - T19; T25 = W[33]; T13 = W[30]; T3i = T28 * T26; } { E Th, Ti, Tc, Td; Th = Rm[WS(rs, 4)]; T27 = T25 * T26; T3Q = T13 * T1b; T17 = T13 * T16; Ti = Rp[WS(rs, 4)]; Tc = Ip[WS(rs, 4)]; Td = Im[WS(rs, 4)]; T1O = W[15]; T1P = Ti + Th; Tj = Th - Ti; T1M = Tc - Td; Te = Tc + Td; T1L = W[14]; Tb = W[16]; T3Y = T1O * T1M; } } { E T1N, T2W, Tf, T2L, T4i; { E T2x, T2D, T3s, T3u, T2J; T2x = FNMS(T2v, T2w, T2u); T1N = T1L * T1M; T2W = Tb * Tj; Tf = Tb * Te; T2D = FNMS(T2B, T2C, T2A); T3s = FMA(T2s, T2w, T3r); T3u = FMA(T2y, T2C, T3t); T2J = W[19]; T2E = T2x - T2D; T4W = T2x + T2D; T3v = T3s + T3u; T4k = T3u - T3s; T2L = FNMS(T2J, T2K, T2I); T4i = FMA(T2J, T2H, T4h); } { E T42, T43, T45, T4a, T3O, T3N; { E T2a, T3j, T47, T3l, T24, T2o, T3n, T49, T22, T2h, T2n; T2a = FMA(T28, T29, T27); T3j = FNMS(T25, T29, T3i); T2M = T2F - T2L; T3w = T2L + T2F; T4V = T4g + T4i; T4j = T4g - T4i; T22 = W[13]; T2h = FNMS(T2f, T2g, T2e); T2n = FNMS(T2l, T2m, T2k); T47 = FMA(T2c, T2g, T46); T3l = FMA(T22, T20, T3k); T24 = FNMS(T22, T23, T21); T2o = T2h - T2n; T3n = T2h + T2n; T49 = FMA(T2i, T2m, T48); { E T2b, T58, T3m, T59; T2b = T24 - T2a; T58 = T2a + T24; T3m = T3j - T3l; T45 = T3j + T3l; T4a = T47 - T49; T59 = T47 + T49; T2p = T2b - T2o; T2T = T2b + T2o; T5a = T58 + T59; T5A = T59 - T58; T3D = T3m + T3n; T3o = T3m - T3n; } } { E T1z, T3e, T1Q, T3c, T1J, T1W, T3Z, T41, T1F; T1z = FNMS(T1v, T1y, T1u); T3e = FMA(T1q, T1y, T3d); T1F = W[25]; T4b = T45 + T4a; T4B = T4a - T45; T1Q = FNMS(T1O, T1P, T1N); T3c = FNMS(T1F, T1D, T3b); T1J = FMA(T1F, T1I, T1E); T1W = FNMS(T1U, T1V, T1T); T3Z = FMA(T1L, T1P, T3Y); T41 = FMA(T1R, T1V, T40); { E T56, T3g, T55, T1K, T1X, T3f; T56 = T1J + T1z; T1K = T1z - T1J; T3g = T1Q + T1W; T1X = T1Q - T1W; T55 = T3Z + T41; T42 = T3Z - T41; T1Y = T1K - T1X; T2S = T1X + T1K; T43 = T3c + T3e; T3f = T3c - T3e; T5z = T55 - T56; T57 = T55 + T56; T3h = T3f - T3g; T3C = T3g + T3f; } } { E Ta, T2Z, T3K, T2X, Tk, TG, T31, T3M, Tg, Tv, TF; Ta = FNMS(T6, T9, T5); T4A = T42 - T43; T44 = T42 + T43; T2Z = FMA(T1, T9, T2Y); Tg = W[17]; Tv = FNMS(Tr, Tu, Tq); TF = FNMS(TB, TE, TA); T3K = FMA(Tm, Tu, T3J); T2X = FNMS(Tg, Te, T2W); Tk = FMA(Tg, Tj, Tf); TG = Tv - TF; T31 = Tv + TF; T3M = FMA(Tw, TE, T3L); { E Tl, T4Z, T30, T4Y; Tl = Ta - Tk; T4Z = Tk + Ta; T30 = T2X - T2Z; T3O = T2X + T2Z; T3N = T3K - T3M; T4Y = T3K + T3M; TH = Tl - TG; T2P = TG + Tl; T50 = T4Y + T4Z; T5x = T4Y - T4Z; T3z = T31 + T30; T32 = T30 - T31; } } { E T11, T34, T36, TR, T1i, T3R, T1c, TN, T18; T11 = FMA(TX, T10, TW); T34 = FNMS(TS, T10, T33); TN = W[21]; T3P = T3N + T3O; T4D = T3N - T3O; T18 = W[31]; T36 = FMA(TN, TL, T35); TR = FNMS(TN, TQ, TM); T1i = W[11]; T3R = FMA(T18, T16, T3Q); T1c = FNMS(T18, T1b, T17); { E T52, T12, T3T, T1m; T52 = TR + T11; T12 = TR - T11; T3T = FMA(T1i, T1g, T3S); T1m = FNMS(T1i, T1l, T1h); { E T37, T51, T38, T1n; T3V = T36 + T34; T37 = T34 - T36; T51 = T3R + T3T; T3U = T3R - T3T; T38 = T1c + T1m; T1n = T1c - T1m; T5w = T51 - T52; T53 = T51 + T52; T2Q = T1n + T12; T1o = T12 - T1n; T3A = T38 + T37; T39 = T37 - T38; } } } } } } { E T4l, T4m, T4n, T4w, T4u; { E T4L, T2O, T3W, T4K, T4I, T4G, T4S, T4U, T4J, T4z, T4H; { E T4C, T2N, T4R, T1p, T4E, T2q, T4Q; T4L = T4A + T4B; T4C = T4A - T4B; T2N = T2E + T2M; T2O = T2M - T2E; T4R = T1o - TH; T1p = TH + T1o; T4E = T3U - T3V; T3W = T3U + T3V; T2q = T1Y + T2p; T4Q = T2p - T1Y; { E T4y, T4x, T4F, T2r; T4F = T4D - T4E; T4K = T4D + T4E; T4y = T1p - T2q; T2r = T1p + T2q; T4I = FMA(KP618033988, T4C, T4F); T4G = FNMS(KP618033988, T4F, T4C); T4S = FNMS(KP618033988, T4R, T4Q); T4U = FMA(KP618033988, T4Q, T4R); Im[WS(rs, 4)] = KP500000000 * (T2r - T2N); T4x = FMA(KP250000000, T2r, T2N); T4J = T4j - T4k; T4l = T4j + T4k; T4z = FMA(KP559016994, T4y, T4x); T4H = FNMS(KP559016994, T4y, T4x); } } { E T2R, T4s, T4d, T4f, T4t, T2U, T4P, T4T; { E T3X, T4O, T4M, T4c, T4N; T4m = T3P + T3W; T3X = T3P - T3W; Ip[WS(rs, 7)] = KP500000000 * (FMA(KP951056516, T4G, T4z)); Ip[WS(rs, 3)] = KP500000000 * (FNMS(KP951056516, T4G, T4z)); Im[WS(rs, 8)] = -(KP500000000 * (FNMS(KP951056516, T4I, T4H))); Im[0] = -(KP500000000 * (FMA(KP951056516, T4I, T4H))); T4O = T4K - T4L; T4M = T4K + T4L; T4c = T44 - T4b; T4n = T44 + T4b; T2R = T2P + T2Q; T4s = T2P - T2Q; Rm[WS(rs, 4)] = KP500000000 * (T4J + T4M); T4N = FNMS(KP250000000, T4M, T4J); T4d = FMA(KP618033988, T4c, T3X); T4f = FNMS(KP618033988, T3X, T4c); T4t = T2S - T2T; T2U = T2S + T2T; T4P = FNMS(KP559016994, T4O, T4N); T4T = FMA(KP559016994, T4O, T4N); } { E T3H, T3G, T2V, T3I, T4e; T2V = T2R + T2U; T3H = T2R - T2U; Rp[WS(rs, 7)] = KP500000000 * (FNMS(KP951056516, T4S, T4P)); Rp[WS(rs, 3)] = KP500000000 * (FMA(KP951056516, T4S, T4P)); Rm[0] = KP500000000 * (FNMS(KP951056516, T4U, T4T)); Rm[WS(rs, 8)] = KP500000000 * (FMA(KP951056516, T4U, T4T)); Ip[WS(rs, 5)] = KP500000000 * (T2O + T2V); T3G = FNMS(KP250000000, T2V, T2O); T3I = FMA(KP559016994, T3H, T3G); T4e = FNMS(KP559016994, T3H, T3G); T4w = FNMS(KP618033988, T4s, T4t); T4u = FMA(KP618033988, T4t, T4s); Ip[WS(rs, 9)] = KP500000000 * (FMA(KP951056516, T4d, T3I)); Ip[WS(rs, 1)] = KP500000000 * (FNMS(KP951056516, T4d, T3I)); Im[WS(rs, 6)] = -(KP500000000 * (FNMS(KP951056516, T4f, T4e))); Im[WS(rs, 2)] = -(KP500000000 * (FMA(KP951056516, T4f, T4e))); } } } { E T3y, T5O, T5Q, T5F, T5K, T5I; { E T5G, T5H, T3x, T4q, T5E, T5C, T3a, T5N, T4p, T5M, T3p, T5y, T5B, T4o; T5G = T5x + T5w; T5y = T5w - T5x; T5B = T5z - T5A; T5H = T5z + T5A; T3y = T3w - T3v; T3x = T3v + T3w; T4q = T4m - T4n; T4o = T4m + T4n; T5E = FMA(KP618033988, T5y, T5B); T5C = FNMS(KP618033988, T5B, T5y); T3a = T32 + T39; T5N = T39 - T32; Rp[WS(rs, 5)] = KP500000000 * (T4l + T4o); T4p = FNMS(KP250000000, T4o, T4l); T5M = T3o - T3h; T3p = T3h + T3o; { E T5u, T5t, T4r, T4v, T3q, T5D, T5v; T4r = FMA(KP559016994, T4q, T4p); T4v = FNMS(KP559016994, T4q, T4p); T5u = T3p - T3a; T3q = T3a + T3p; Rp[WS(rs, 9)] = KP500000000 * (FNMS(KP951056516, T4u, T4r)); Rp[WS(rs, 1)] = KP500000000 * (FMA(KP951056516, T4u, T4r)); Rm[WS(rs, 2)] = KP500000000 * (FNMS(KP951056516, T4w, T4v)); Rm[WS(rs, 6)] = KP500000000 * (FMA(KP951056516, T4w, T4v)); Im[WS(rs, 9)] = KP500000000 * (T3q - T3x); T5t = FMA(KP250000000, T3q, T3x); T5O = FNMS(KP618033988, T5N, T5M); T5Q = FMA(KP618033988, T5M, T5N); T5F = T4V - T4W; T4X = T4V + T4W; T5D = FNMS(KP559016994, T5u, T5t); T5v = FMA(KP559016994, T5u, T5t); Im[WS(rs, 5)] = -(KP500000000 * (FNMS(KP951056516, T5C, T5v))); Ip[WS(rs, 6)] = KP500000000 * (FMA(KP951056516, T5C, T5v)); Im[WS(rs, 1)] = -(KP500000000 * (FNMS(KP951056516, T5E, T5D))); Ip[WS(rs, 2)] = KP500000000 * (FMA(KP951056516, T5E, T5D)); T5K = T5G - T5H; T5I = T5G + T5H; } } { E T54, T5b, T5s, T5q, T5g, T5h, T3F, T5m, T5o, T5p, T5J, T5l, T5r, T5n; T54 = T50 + T53; T5o = T50 - T53; T5p = T5a - T57; T5b = T57 + T5a; Rm[WS(rs, 9)] = KP500000000 * (T5F + T5I); T5J = FNMS(KP250000000, T5I, T5F); T5s = FMA(KP618033988, T5o, T5p); T5q = FNMS(KP618033988, T5p, T5o); { E T5L, T5P, T3B, T3E; T5L = FNMS(KP559016994, T5K, T5J); T5P = FMA(KP559016994, T5K, T5J); T3B = T3z + T3A; T5g = T3z - T3A; T5h = T3C - T3D; T3E = T3C + T3D; Rm[WS(rs, 1)] = KP500000000 * (FMA(KP951056516, T5O, T5L)); Rp[WS(rs, 2)] = KP500000000 * (FNMS(KP951056516, T5O, T5L)); Rm[WS(rs, 5)] = KP500000000 * (FNMS(KP951056516, T5Q, T5P)); Rp[WS(rs, 6)] = KP500000000 * (FMA(KP951056516, T5Q, T5P)); T3F = T3B + T3E; T5m = T3B - T3E; } Ip[0] = KP500000000 * (T3y + T3F); T5l = FNMS(KP250000000, T3F, T3y); T5i = FMA(KP618033988, T5h, T5g); T5k = FNMS(KP618033988, T5g, T5h); T5r = FNMS(KP559016994, T5m, T5l); T5n = FMA(KP559016994, T5m, T5l); Im[WS(rs, 3)] = -(KP500000000 * (FNMS(KP951056516, T5q, T5n))); Ip[WS(rs, 4)] = KP500000000 * (FMA(KP951056516, T5q, T5n)); Im[WS(rs, 7)] = -(KP500000000 * (FNMS(KP951056516, T5s, T5r))); Ip[WS(rs, 8)] = KP500000000 * (FMA(KP951056516, T5s, T5r)); T5e = T54 - T5b; T5c = T54 + T5b; } } } } Rp[0] = KP500000000 * (T4X + T5c); T5d = FNMS(KP250000000, T5c, T4X); T5j = FNMS(KP559016994, T5e, T5d); T5f = FMA(KP559016994, T5e, T5d); Rm[WS(rs, 3)] = KP500000000 * (FMA(KP951056516, T5i, T5f)); Rp[WS(rs, 4)] = KP500000000 * (FNMS(KP951056516, T5i, T5f)); Rm[WS(rs, 7)] = KP500000000 * (FNMS(KP951056516, T5k, T5j)); Rp[WS(rs, 8)] = KP500000000 * (FMA(KP951056516, T5k, T5j)); } } } static const tw_instr twinstr[] = { {TW_FULL, 1, 20}, {TW_NEXT, 1, 0} }; static const hc2c_desc desc = { 20, "hc2cfdft_20", twinstr, &GENUS, {176, 78, 110, 0} }; void X(codelet_hc2cfdft_20) (planner *p) { X(khc2c_register) (p, hc2cfdft_20, &desc, HC2C_VIA_DFT); } #else /* HAVE_FMA */ /* Generated by: ../../../genfft/gen_hc2cdft.native -compact -variables 4 -pipeline-latency 4 -n 20 -dit -name hc2cfdft_20 -include hc2cf.h */ /* * This function contains 286 FP additions, 140 FP multiplications, * (or, 224 additions, 78 multiplications, 62 fused multiply/add), * 98 stack variables, 5 constants, and 80 memory accesses */ #include "hc2cf.h" static void hc2cfdft_20(R *Rp, R *Ip, R *Rm, R *Im, const R *W, stride rs, INT mb, INT me, INT ms) { DK(KP125000000, +0.125000000000000000000000000000000000000000000); DK(KP500000000, +0.500000000000000000000000000000000000000000000); DK(KP279508497, +0.279508497187473712051146708591409529430077295); DK(KP293892626, +0.293892626146236564584352977319536384298826219); DK(KP475528258, +0.475528258147576786058219666689691071702849317); { INT m; for (m = mb, W = W + ((mb - 1) * 38); m < me; m = m + 1, Rp = Rp + ms, Ip = Ip + ms, Rm = Rm - ms, Im = Im - ms, W = W + 38, MAKE_VOLATILE_STRIDE(80, rs)) { E T12, T2w, T4o, T4V, T2H, T3a, T4y, T4Y, T1z, T2v, T25, T2y, T2s, T2z, T4v; E T4X, T4r, T4U, T3A, T3Z, T2X, T37, T3k, T41, T2M, T39, T3v, T3Y, T2S, T36; E T3p, T42, Td, T4G, T33, T3N, Tw, T4H, T32, T3O; { E T3, T3L, T1x, T2V, Th, Tl, TC, T3g, Tq, Tu, TH, T3h, T7, Tb, T1q; E T2U, TR, T2P, T1F, T3r, T23, T2K, T2f, T3y, T1k, T3m, T2q, T2E, T10, T2Q; E T1K, T3s, T1U, T2J, T2a, T3x, T1b, T3l, T2l, T2D; { E T1, T2, T1s, T1u, T1v, T1w, T1r, T1t; T1 = Ip[0]; T2 = Im[0]; T1s = T1 + T2; T1u = Rp[0]; T1v = Rm[0]; T1w = T1u - T1v; T3 = T1 - T2; T3L = T1u + T1v; T1r = W[0]; T1t = W[1]; T1x = FNMS(T1t, T1w, T1r * T1s); T2V = FMA(T1r, T1w, T1t * T1s); } { E Tf, Tg, Tz, Tj, Tk, TB, Ty, TA; Tf = Ip[WS(rs, 2)]; Tg = Im[WS(rs, 2)]; Tz = Tf - Tg; Tj = Rp[WS(rs, 2)]; Tk = Rm[WS(rs, 2)]; TB = Tj + Tk; Th = Tf + Tg; Tl = Tj - Tk; Ty = W[6]; TA = W[7]; TC = FNMS(TA, TB, Ty * Tz); T3g = FMA(TA, Tz, Ty * TB); } { E To, Tp, TE, Ts, Tt, TG, TD, TF; To = Ip[WS(rs, 7)]; Tp = Im[WS(rs, 7)]; TE = To - Tp; Ts = Rp[WS(rs, 7)]; Tt = Rm[WS(rs, 7)]; TG = Ts + Tt; Tq = To + Tp; Tu = Ts - Tt; TD = W[26]; TF = W[27]; TH = FNMS(TF, TG, TD * TE); T3h = FMA(TF, TE, TD * TG); } { E T5, T6, T1n, T9, Ta, T1p, T1m, T1o; T5 = Ip[WS(rs, 5)]; T6 = Im[WS(rs, 5)]; T1n = T5 + T6; T9 = Rp[WS(rs, 5)]; Ta = Rm[WS(rs, 5)]; T1p = T9 - Ta; T7 = T5 - T6; Tb = T9 + Ta; T1m = W[20]; T1o = W[21]; T1q = FNMS(T1o, T1p, T1m * T1n); T2U = FMA(T1m, T1p, T1o * T1n); } { E TM, T1C, TQ, T1E; { E TK, TL, TO, TP; TK = Ip[WS(rs, 4)]; TL = Im[WS(rs, 4)]; TM = TK + TL; T1C = TK - TL; TO = Rp[WS(rs, 4)]; TP = Rm[WS(rs, 4)]; TQ = TO - TP; T1E = TO + TP; } { E TJ, TN, T1B, T1D; TJ = W[16]; TN = W[17]; TR = FNMS(TN, TQ, TJ * TM); T2P = FMA(TN, TM, TJ * TQ); T1B = W[14]; T1D = W[15]; T1F = FNMS(T1D, T1E, T1B * T1C); T3r = FMA(T1D, T1C, T1B * T1E); } } { E T1Y, T2c, T22, T2e; { E T1W, T1X, T20, T21; T1W = Ip[WS(rs, 1)]; T1X = Im[WS(rs, 1)]; T1Y = T1W + T1X; T2c = T1W - T1X; T20 = Rp[WS(rs, 1)]; T21 = Rm[WS(rs, 1)]; T22 = T20 - T21; T2e = T20 + T21; } { E T1V, T1Z, T2b, T2d; T1V = W[4]; T1Z = W[5]; T23 = FNMS(T1Z, T22, T1V * T1Y); T2K = FMA(T1Z, T1Y, T1V * T22); T2b = W[2]; T2d = W[3]; T2f = FNMS(T2d, T2e, T2b * T2c); T3y = FMA(T2d, T2c, T2b * T2e); } } { E T1f, T2n, T1j, T2p; { E T1d, T1e, T1h, T1i; T1d = Ip[WS(rs, 3)]; T1e = Im[WS(rs, 3)]; T1f = T1d - T1e; T2n = T1d + T1e; T1h = Rp[WS(rs, 3)]; T1i = Rm[WS(rs, 3)]; T1j = T1h + T1i; T2p = T1h - T1i; } { E T1c, T1g, T2m, T2o; T1c = W[10]; T1g = W[11]; T1k = FNMS(T1g, T1j, T1c * T1f); T3m = FMA(T1c, T1j, T1g * T1f); T2m = W[12]; T2o = W[13]; T2q = FNMS(T2o, T2p, T2m * T2n); T2E = FMA(T2m, T2p, T2o * T2n); } } { E TV, T1H, TZ, T1J; { E TT, TU, TX, TY; TT = Ip[WS(rs, 9)]; TU = Im[WS(rs, 9)]; TV = TT + TU; T1H = TT - TU; TX = Rp[WS(rs, 9)]; TY = Rm[WS(rs, 9)]; TZ = TX - TY; T1J = TX + TY; } { E TS, TW, T1G, T1I; TS = W[36]; TW = W[37]; T10 = FNMS(TW, TZ, TS * TV); T2Q = FMA(TW, TV, TS * TZ); T1G = W[34]; T1I = W[35]; T1K = FNMS(T1I, T1J, T1G * T1H); T3s = FMA(T1I, T1H, T1G * T1J); } } { E T1P, T27, T1T, T29; { E T1N, T1O, T1R, T1S; T1N = Ip[WS(rs, 6)]; T1O = Im[WS(rs, 6)]; T1P = T1N + T1O; T27 = T1N - T1O; T1R = Rp[WS(rs, 6)]; T1S = Rm[WS(rs, 6)]; T1T = T1R - T1S; T29 = T1R + T1S; } { E T1M, T1Q, T26, T28; T1M = W[24]; T1Q = W[25]; T1U = FNMS(T1Q, T1T, T1M * T1P); T2J = FMA(T1Q, T1P, T1M * T1T); T26 = W[22]; T28 = W[23]; T2a = FNMS(T28, T29, T26 * T27); T3x = FMA(T28, T27, T26 * T29); } } { E T16, T2k, T1a, T2i; { E T14, T15, T18, T19; T14 = Ip[WS(rs, 8)]; T15 = Im[WS(rs, 8)]; T16 = T14 - T15; T2k = T14 + T15; T18 = Rp[WS(rs, 8)]; T19 = Rm[WS(rs, 8)]; T1a = T18 + T19; T2i = T19 - T18; } { E T13, T17, T2h, T2j; T13 = W[30]; T17 = W[31]; T1b = FNMS(T17, T1a, T13 * T16); T3l = FMA(T13, T1a, T17 * T16); T2h = W[33]; T2j = W[32]; T2l = FMA(T2h, T2i, T2j * T2k); T2D = FNMS(T2h, T2k, T2j * T2i); } } { E T2g, T2r, T3n, T3o; { E TI, T11, T4m, T4n; TI = TC - TH; T11 = TR - T10; T12 = TI - T11; T2w = TI + T11; T4m = T3g + T3h; T4n = TR + T10; T4o = T4m + T4n; T4V = T4m - T4n; } { E T2F, T2G, T4w, T4x; T2F = T2D - T2E; T2G = T2a + T2f; T2H = T2F - T2G; T3a = T2F + T2G; T4w = T2l + T2q; T4x = T3x + T3y; T4y = T4w + T4x; T4Y = T4x - T4w; } { E T1l, T1y, T1L, T24; T1l = T1b - T1k; T1y = T1q - T1x; T1z = T1l + T1y; T2v = T1y - T1l; T1L = T1F - T1K; T24 = T1U - T23; T25 = T1L - T24; T2y = T1L + T24; } T2g = T2a - T2f; T2r = T2l - T2q; T2s = T2g - T2r; T2z = T2r + T2g; { E T4t, T4u, T4p, T4q; T4t = T3r + T3s; T4u = T1U + T23; T4v = T4t + T4u; T4X = T4t - T4u; T4p = T3l + T3m; T4q = T1q + T1x; T4r = T4p + T4q; T4U = T4p - T4q; } { E T3w, T3z, T2T, T2W; T3w = T2D + T2E; T3z = T3x - T3y; T3A = T3w + T3z; T3Z = T3z - T3w; T2T = T1b + T1k; T2W = T2U + T2V; T2X = T2T + T2W; T37 = T2T - T2W; } { E T3i, T3j, T2I, T2L; T3i = T3g - T3h; T3j = T2Q - T2P; T3k = T3i + T3j; T41 = T3i - T3j; T2I = T1F + T1K; T2L = T2J + T2K; T2M = T2I + T2L; T39 = T2I - T2L; } { E T3t, T3u, T2O, T2R; T3t = T3r - T3s; T3u = T2K - T2J; T3v = T3t + T3u; T3Y = T3t - T3u; T2O = TC + TH; T2R = T2P + T2Q; T2S = T2O + T2R; T36 = T2O - T2R; } T3n = T3l - T3m; T3o = T2U - T2V; T3p = T3n + T3o; T42 = T3n - T3o; { E Tc, T3M, T4, T8; T4 = W[18]; T8 = W[19]; Tc = FNMS(T8, Tb, T4 * T7); T3M = FMA(T4, Tb, T8 * T7); Td = T3 - Tc; T4G = T3L + T3M; T33 = Tc + T3; T3N = T3L - T3M; } { E Tm, T30, Tv, T31; { E Te, Ti, Tn, Tr; Te = W[8]; Ti = W[9]; Tm = FNMS(Ti, Tl, Te * Th); T30 = FMA(Ti, Th, Te * Tl); Tn = W[28]; Tr = W[29]; Tv = FNMS(Tr, Tu, Tn * Tq); T31 = FMA(Tr, Tq, Tn * Tu); } Tw = Tm - Tv; T4H = Tm + Tv; T32 = T30 + T31; T3O = T31 - T30; } } } { E T3C, T3E, Tx, T2u, T3d, T3e, T3D, T3f; { E T3q, T3B, T1A, T2t; T3q = T3k - T3p; T3B = T3v - T3A; T3C = FMA(KP475528258, T3q, KP293892626 * T3B); T3E = FNMS(KP293892626, T3q, KP475528258 * T3B); Tx = Td - Tw; T1A = T12 + T1z; T2t = T25 + T2s; T2u = T1A + T2t; T3d = KP279508497 * (T1A - T2t); T3e = FNMS(KP125000000, T2u, KP500000000 * Tx); } Ip[WS(rs, 5)] = KP500000000 * (Tx + T2u); T3D = T3d - T3e; Im[WS(rs, 2)] = T3D - T3E; Im[WS(rs, 6)] = T3D + T3E; T3f = T3d + T3e; Ip[WS(rs, 1)] = T3f - T3C; Ip[WS(rs, 9)] = T3f + T3C; } { E T3H, T3T, T3P, T3Q, T3K, T3R, T3U, T3S; { E T3F, T3G, T3I, T3J; T3F = T12 - T1z; T3G = T25 - T2s; T3H = FMA(KP475528258, T3F, KP293892626 * T3G); T3T = FNMS(KP293892626, T3F, KP475528258 * T3G); T3P = T3N + T3O; T3I = T3k + T3p; T3J = T3v + T3A; T3Q = T3I + T3J; T3K = KP279508497 * (T3I - T3J); T3R = FNMS(KP125000000, T3Q, KP500000000 * T3P); } Rp[WS(rs, 5)] = KP500000000 * (T3P + T3Q); T3U = T3R - T3K; Rm[WS(rs, 6)] = T3T + T3U; Rm[WS(rs, 2)] = T3U - T3T; T3S = T3K + T3R; Rp[WS(rs, 1)] = T3H + T3S; Rp[WS(rs, 9)] = T3S - T3H; } { E T44, T46, T2C, T2B, T3V, T3W, T45, T3X; { E T40, T43, T2x, T2A; T40 = T3Y - T3Z; T43 = T41 - T42; T44 = FNMS(KP293892626, T43, KP475528258 * T40); T46 = FMA(KP475528258, T43, KP293892626 * T40); T2C = Tw + Td; T2x = T2v - T2w; T2A = T2y + T2z; T2B = T2x - T2A; T3V = FMA(KP500000000, T2C, KP125000000 * T2B); T3W = KP279508497 * (T2x + T2A); } Im[WS(rs, 4)] = KP500000000 * (T2B - T2C); T45 = T3W - T3V; Im[0] = T45 - T46; Im[WS(rs, 8)] = T45 + T46; T3X = T3V + T3W; Ip[WS(rs, 3)] = T3X - T44; Ip[WS(rs, 7)] = T3X + T44; } { E T49, T4h, T4a, T4d, T4e, T4f, T4i, T4g; { E T47, T48, T4b, T4c; T47 = T2y - T2z; T48 = T2w + T2v; T49 = FNMS(KP293892626, T48, KP475528258 * T47); T4h = FMA(KP475528258, T48, KP293892626 * T47); T4a = T3N - T3O; T4b = T41 + T42; T4c = T3Y + T3Z; T4d = T4b + T4c; T4e = FNMS(KP125000000, T4d, KP500000000 * T4a); T4f = KP279508497 * (T4b - T4c); } Rm[WS(rs, 4)] = KP500000000 * (T4a + T4d); T4i = T4f + T4e; Rm[WS(rs, 8)] = T4h + T4i; Rm[0] = T4i - T4h; T4g = T4e - T4f; Rp[WS(rs, 3)] = T49 + T4g; Rp[WS(rs, 7)] = T4g - T49; } { E T50, T52, T34, T2Z, T4R, T4S, T51, T4T; { E T4W, T4Z, T2N, T2Y; T4W = T4U - T4V; T4Z = T4X - T4Y; T50 = FNMS(KP293892626, T4Z, KP475528258 * T4W); T52 = FMA(KP293892626, T4W, KP475528258 * T4Z); T34 = T32 + T33; T2N = T2H - T2M; T2Y = T2S + T2X; T2Z = T2N - T2Y; T4R = FMA(KP500000000, T34, KP125000000 * T2Z); T4S = KP279508497 * (T2Y + T2N); } Im[WS(rs, 9)] = KP500000000 * (T2Z - T34); T51 = T4R - T4S; Ip[WS(rs, 2)] = T51 + T52; Im[WS(rs, 1)] = T52 - T51; T4T = T4R + T4S; Ip[WS(rs, 6)] = T4T + T50; Im[WS(rs, 5)] = T50 - T4T; } { E T5c, T5d, T53, T56, T57, T58, T5e, T59; { E T5a, T5b, T54, T55; T5a = T2M + T2H; T5b = T2S - T2X; T5c = FNMS(KP293892626, T5b, KP475528258 * T5a); T5d = FMA(KP475528258, T5b, KP293892626 * T5a); T53 = T4G - T4H; T54 = T4V + T4U; T55 = T4X + T4Y; T56 = T54 + T55; T57 = FNMS(KP125000000, T56, KP500000000 * T53); T58 = KP279508497 * (T54 - T55); } Rm[WS(rs, 9)] = KP500000000 * (T53 + T56); T5e = T58 + T57; Rp[WS(rs, 6)] = T5d + T5e; Rm[WS(rs, 5)] = T5e - T5d; T59 = T57 - T58; Rp[WS(rs, 2)] = T59 - T5c; Rm[WS(rs, 1)] = T5c + T59; } { E T4A, T4C, T35, T3c, T4j, T4k, T4B, T4l; { E T4s, T4z, T38, T3b; T4s = T4o - T4r; T4z = T4v - T4y; T4A = FNMS(KP475528258, T4z, KP293892626 * T4s); T4C = FMA(KP475528258, T4s, KP293892626 * T4z); T35 = T33 - T32; T38 = T36 + T37; T3b = T39 + T3a; T3c = T38 + T3b; T4j = FNMS(KP125000000, T3c, KP500000000 * T35); T4k = KP279508497 * (T38 - T3b); } Ip[0] = KP500000000 * (T35 + T3c); T4B = T4k + T4j; Ip[WS(rs, 4)] = T4B + T4C; Im[WS(rs, 3)] = T4C - T4B; T4l = T4j - T4k; Ip[WS(rs, 8)] = T4l + T4A; Im[WS(rs, 7)] = T4A - T4l; } { E T4O, T4P, T4I, T4J, T4F, T4K, T4Q, T4L; { E T4M, T4N, T4D, T4E; T4M = T36 - T37; T4N = T39 - T3a; T4O = FMA(KP475528258, T4M, KP293892626 * T4N); T4P = FNMS(KP293892626, T4M, KP475528258 * T4N); T4I = T4G + T4H; T4D = T4o + T4r; T4E = T4v + T4y; T4J = T4D + T4E; T4F = KP279508497 * (T4D - T4E); T4K = FNMS(KP125000000, T4J, KP500000000 * T4I); } Rp[0] = KP500000000 * (T4I + T4J); T4Q = T4K - T4F; Rp[WS(rs, 8)] = T4P + T4Q; Rm[WS(rs, 7)] = T4Q - T4P; T4L = T4F + T4K; Rp[WS(rs, 4)] = T4L - T4O; Rm[WS(rs, 3)] = T4O + T4L; } } } } static const tw_instr twinstr[] = { {TW_FULL, 1, 20}, {TW_NEXT, 1, 0} }; static const hc2c_desc desc = { 20, "hc2cfdft_20", twinstr, &GENUS, {224, 78, 62, 0} }; void X(codelet_hc2cfdft_20) (planner *p) { X(khc2c_register) (p, hc2cfdft_20, &desc, HC2C_VIA_DFT); } #endif /* HAVE_FMA */ fftw-3.3.3/rdft/scalar/r2cf/hc2cfdft2_4.c0000644000175000017500000001422412054410722014603 00000000000000/* * Copyright (c) 2003, 2007-11 Matteo Frigo * Copyright (c) 2003, 2007-11 Massachusetts Institute of Technology * * This program is free software; you can redistribute it and/or modify * it under the terms of the GNU General Public License as published by * the Free Software Foundation; either version 2 of the License, or * (at your option) any later version. * * This program is distributed in the hope that it will be useful, * but WITHOUT ANY WARRANTY; without even the implied warranty of * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the * GNU General Public License for more details. * * You should have received a copy of the GNU General Public License * along with this program; if not, write to the Free Software * Foundation, Inc., 51 Franklin Street, Fifth Floor, Boston, MA 02110-1301 USA * */ /* This file was automatically generated --- DO NOT EDIT */ /* Generated on Sun Nov 25 07:40:50 EST 2012 */ #include "codelet-rdft.h" #ifdef HAVE_FMA /* Generated by: ../../../genfft/gen_hc2cdft.native -fma -reorder-insns -schedule-for-pipeline -compact -variables 4 -pipeline-latency 4 -twiddle-log3 -precompute-twiddles -n 4 -dit -name hc2cfdft2_4 -include hc2cf.h */ /* * This function contains 32 FP additions, 24 FP multiplications, * (or, 24 additions, 16 multiplications, 8 fused multiply/add), * 33 stack variables, 1 constants, and 16 memory accesses */ #include "hc2cf.h" static void hc2cfdft2_4(R *Rp, R *Ip, R *Rm, R *Im, const R *W, stride rs, INT mb, INT me, INT ms) { DK(KP500000000, +0.500000000000000000000000000000000000000000000); { INT m; for (m = mb, W = W + ((mb - 1) * 4); m < me; m = m + 1, Rp = Rp + ms, Ip = Ip + ms, Rm = Rm - ms, Im = Im - ms, W = W + 4, MAKE_VOLATILE_STRIDE(16, rs)) { E T1, T5, T2, T4; T1 = W[0]; T5 = W[3]; T2 = W[2]; T4 = W[1]; { E Tc, T6, Tp, Tj, Tw, Tt, T9, TE, To, TC, Ta, Tr, Tf, Tl, Tm; { E Th, Tb, T3, Ti; Th = Ip[0]; Tb = T1 * T5; T3 = T1 * T2; Ti = Im[0]; Tl = Rm[0]; Tc = FNMS(T4, T2, Tb); T6 = FMA(T4, T5, T3); Tp = Th + Ti; Tj = Th - Ti; Tm = Rp[0]; } { E T7, T8, Td, Tn, Te; T7 = Ip[WS(rs, 1)]; T8 = Im[WS(rs, 1)]; Td = Rp[WS(rs, 1)]; Tw = Tm + Tl; Tn = Tl - Tm; Tt = T7 + T8; T9 = T7 - T8; Te = Rm[WS(rs, 1)]; TE = T4 * Tn; To = T1 * Tn; TC = T2 * Tt; Ta = T6 * T9; Tr = Td - Te; Tf = Td + Te; } { E Tq, Tk, TB, Ty, Tu, TI, TG, TF; Tq = FNMS(T4, Tp, To); TF = FMA(T1, Tp, TE); { E Tg, Tx, TD, Ts; Tg = FNMS(Tc, Tf, Ta); Tx = T6 * Tf; TD = FNMS(T5, Tr, TC); Ts = T2 * Tr; Tk = Tg + Tj; TB = Tj - Tg; Ty = FMA(Tc, T9, Tx); Tu = FMA(T5, Tt, Ts); TI = TD + TF; TG = TD - TF; } { E Tz, TH, Tv, TA; Tz = Tw - Ty; TH = Tw + Ty; Tv = Tq - Tu; TA = Tu + Tq; Rp[0] = KP500000000 * (TH + TI); Rm[WS(rs, 1)] = KP500000000 * (TH - TI); Rm[0] = KP500000000 * (Tz - TA); Im[WS(rs, 1)] = KP500000000 * (Tv - Tk); Ip[0] = KP500000000 * (Tk + Tv); Im[0] = KP500000000 * (TG - TB); Rp[WS(rs, 1)] = KP500000000 * (Tz + TA); Ip[WS(rs, 1)] = KP500000000 * (TB + TG); } } } } } } static const tw_instr twinstr[] = { {TW_CEXP, 1, 1}, {TW_CEXP, 1, 3}, {TW_NEXT, 1, 0} }; static const hc2c_desc desc = { 4, "hc2cfdft2_4", twinstr, &GENUS, {24, 16, 8, 0} }; void X(codelet_hc2cfdft2_4) (planner *p) { X(khc2c_register) (p, hc2cfdft2_4, &desc, HC2C_VIA_DFT); } #else /* HAVE_FMA */ /* Generated by: ../../../genfft/gen_hc2cdft.native -compact -variables 4 -pipeline-latency 4 -twiddle-log3 -precompute-twiddles -n 4 -dit -name hc2cfdft2_4 -include hc2cf.h */ /* * This function contains 32 FP additions, 24 FP multiplications, * (or, 24 additions, 16 multiplications, 8 fused multiply/add), * 24 stack variables, 1 constants, and 16 memory accesses */ #include "hc2cf.h" static void hc2cfdft2_4(R *Rp, R *Ip, R *Rm, R *Im, const R *W, stride rs, INT mb, INT me, INT ms) { DK(KP500000000, +0.500000000000000000000000000000000000000000000); { INT m; for (m = mb, W = W + ((mb - 1) * 4); m < me; m = m + 1, Rp = Rp + ms, Ip = Ip + ms, Rm = Rm - ms, Im = Im - ms, W = W + 4, MAKE_VOLATILE_STRIDE(16, rs)) { E T1, T3, T2, T4, T5, T9; T1 = W[0]; T3 = W[1]; T2 = W[2]; T4 = W[3]; T5 = FMA(T1, T2, T3 * T4); T9 = FNMS(T3, T2, T1 * T4); { E Tg, Tr, Tm, Tx, Td, Tw, Tp, Ts; { E Te, Tf, Tl, Ti, Tj, Tk; Te = Ip[0]; Tf = Im[0]; Tl = Te + Tf; Ti = Rm[0]; Tj = Rp[0]; Tk = Ti - Tj; Tg = Te - Tf; Tr = Tj + Ti; Tm = FNMS(T3, Tl, T1 * Tk); Tx = FMA(T3, Tk, T1 * Tl); } { E T8, To, Tc, Tn; { E T6, T7, Ta, Tb; T6 = Ip[WS(rs, 1)]; T7 = Im[WS(rs, 1)]; T8 = T6 - T7; To = T6 + T7; Ta = Rp[WS(rs, 1)]; Tb = Rm[WS(rs, 1)]; Tc = Ta + Tb; Tn = Ta - Tb; } Td = FNMS(T9, Tc, T5 * T8); Tw = FNMS(T4, Tn, T2 * To); Tp = FMA(T2, Tn, T4 * To); Ts = FMA(T5, Tc, T9 * T8); } { E Th, Tq, Tz, TA; Th = Td + Tg; Tq = Tm - Tp; Ip[0] = KP500000000 * (Th + Tq); Im[WS(rs, 1)] = KP500000000 * (Tq - Th); Tz = Tr + Ts; TA = Tw + Tx; Rm[WS(rs, 1)] = KP500000000 * (Tz - TA); Rp[0] = KP500000000 * (Tz + TA); } { E Tt, Tu, Tv, Ty; Tt = Tr - Ts; Tu = Tp + Tm; Rm[0] = KP500000000 * (Tt - Tu); Rp[WS(rs, 1)] = KP500000000 * (Tt + Tu); Tv = Tg - Td; Ty = Tw - Tx; Ip[WS(rs, 1)] = KP500000000 * (Tv + Ty); Im[0] = KP500000000 * (Ty - Tv); } } } } } static const tw_instr twinstr[] = { {TW_CEXP, 1, 1}, {TW_CEXP, 1, 3}, {TW_NEXT, 1, 0} }; static const hc2c_desc desc = { 4, "hc2cfdft2_4", twinstr, &GENUS, {24, 16, 8, 0} }; void X(codelet_hc2cfdft2_4) (planner *p) { X(khc2c_register) (p, hc2cfdft2_4, &desc, HC2C_VIA_DFT); } #endif /* HAVE_FMA */ fftw-3.3.3/rdft/scalar/r2cf/hc2cfdft2_8.c0000644000175000017500000002720212054410722014607 00000000000000/* * Copyright (c) 2003, 2007-11 Matteo Frigo * Copyright (c) 2003, 2007-11 Massachusetts Institute of Technology * * This program is free software; you can redistribute it and/or modify * it under the terms of the GNU General Public License as published by * the Free Software Foundation; either version 2 of the License, or * (at your option) any later version. * * This program is distributed in the hope that it will be useful, * but WITHOUT ANY WARRANTY; without even the implied warranty of * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the * GNU General Public License for more details. * * You should have received a copy of the GNU General Public License * along with this program; if not, write to the Free Software * Foundation, Inc., 51 Franklin Street, Fifth Floor, Boston, MA 02110-1301 USA * */ /* This file was automatically generated --- DO NOT EDIT */ /* Generated on Sun Nov 25 07:40:50 EST 2012 */ #include "codelet-rdft.h" #ifdef HAVE_FMA /* Generated by: ../../../genfft/gen_hc2cdft.native -fma -reorder-insns -schedule-for-pipeline -compact -variables 4 -pipeline-latency 4 -twiddle-log3 -precompute-twiddles -n 8 -dit -name hc2cfdft2_8 -include hc2cf.h */ /* * This function contains 90 FP additions, 66 FP multiplications, * (or, 60 additions, 36 multiplications, 30 fused multiply/add), * 68 stack variables, 2 constants, and 32 memory accesses */ #include "hc2cf.h" static void hc2cfdft2_8(R *Rp, R *Ip, R *Rm, R *Im, const R *W, stride rs, INT mb, INT me, INT ms) { DK(KP707106781, +0.707106781186547524400844362104849039284835938); DK(KP500000000, +0.500000000000000000000000000000000000000000000); { INT m; for (m = mb, W = W + ((mb - 1) * 6); m < me; m = m + 1, Rp = Rp + ms, Ip = Ip + ms, Rm = Rm - ms, Im = Im - ms, W = W + 6, MAKE_VOLATILE_STRIDE(32, rs)) { E T1G, T1F, T1C, T1D, T1N, T1B, T1R, T1L; { E T1, T2, Th, Tj, T4, T3, Ti, Tp, T5; T1 = W[0]; T2 = W[2]; Th = W[4]; Tj = W[5]; T4 = W[1]; T3 = T1 * T2; Ti = T1 * Th; Tp = T1 * Tj; T5 = W[3]; { E Tk, Tq, TI, T1a, T1u, TY, TF, TS, T1s, T1c, Tr, T1n, Tg, T16, Tn; E T13, T1f, Ts, To, T1o; { E T6, Tw, Tc, TB, TQ, TM, TC, TR, Tz, TD, TA; { E TX, TV, TT, TU; { E TG, Tb, TH, TP, TL; TG = Ip[0]; Tk = FMA(T4, Tj, Ti); Tq = FNMS(T4, Th, Tp); T6 = FMA(T4, T5, T3); Tw = FNMS(T4, T5, T3); Tb = T1 * T5; TH = Im[0]; TT = Rm[0]; TP = T6 * Tj; TL = T6 * Th; Tc = FNMS(T4, T2, Tb); TB = FMA(T4, T2, Tb); TX = TG + TH; TI = TG - TH; TU = Rp[0]; TQ = FNMS(Tc, Th, TP); TM = FMA(Tc, Tj, TL); } T1a = TU + TT; TV = TT - TU; { E Tx, Ty, T1t, TW; Tx = Ip[WS(rs, 2)]; Ty = Im[WS(rs, 2)]; T1t = T4 * TV; TW = T1 * TV; TC = Rp[WS(rs, 2)]; TR = Tx + Ty; Tz = Tx - Ty; T1u = FMA(T1, TX, T1t); TY = FNMS(T4, TX, TW); TD = Rm[WS(rs, 2)]; } TA = Tw * Tz; } { E Td, T9, T12, Te, Ta, T1m; { E T7, T8, TN, TE, TO, T1r, T1b; T7 = Ip[WS(rs, 1)]; T8 = Im[WS(rs, 1)]; TN = TD - TC; TE = TC + TD; Td = Rp[WS(rs, 1)]; T9 = T7 - T8; T12 = T7 + T8; TO = TM * TN; T1r = TQ * TN; T1b = Tw * TE; TF = FNMS(TB, TE, TA); TS = FNMS(TQ, TR, TO); T1s = FMA(TM, TR, T1r); T1c = FMA(TB, Tz, T1b); Te = Rm[WS(rs, 1)]; } Ta = T6 * T9; T1m = T2 * T12; { E Tl, T10, Tf, Tm, T11, T1e; Tl = Ip[WS(rs, 3)]; T10 = Td - Te; Tf = Td + Te; Tm = Im[WS(rs, 3)]; Tr = Rp[WS(rs, 3)]; T11 = T2 * T10; T1n = FNMS(T5, T10, T1m); T1e = T6 * Tf; Tg = FNMS(Tc, Tf, Ta); T16 = Tl + Tm; Tn = Tl - Tm; T13 = FMA(T5, T12, T11); T1f = FMA(Tc, T9, T1e); Ts = Rm[WS(rs, 3)]; } To = Tk * Tn; T1o = Th * T16; } } { E T1z, T1K, T1y, T1k, T1J, T1A, T1x, T1j; { E T1w, TK, T1l, T19, T1d, T1i; { E TJ, T14, Tt, T1v, T1h; T1z = TI - TF; TJ = TF + TI; T14 = Tr - Ts; Tt = Tr + Ts; T1v = T1s + T1u; T1G = T1u - T1s; { E TZ, T1q, Tv, T18, T15; T1F = TY - TS; TZ = TS + TY; T15 = Th * T14; { E T1p, T1g, Tu, T17; T1p = FNMS(Tj, T14, T1o); T1g = Tk * Tt; Tu = FNMS(Tq, Tt, To); T17 = FMA(Tj, T16, T15); T1C = T1p - T1n; T1q = T1n + T1p; T1h = FMA(Tq, Tn, T1g); T1K = Tg - Tu; Tv = Tg + Tu; T18 = T13 + T17; T1D = T13 - T17; } T1w = T1q - T1v; T1y = T1q + T1v; TK = Tv + TJ; T1l = TJ - Tv; T1k = T18 + TZ; T19 = TZ - T18; } T1J = T1a - T1c; T1d = T1a + T1c; T1i = T1f + T1h; T1A = T1f - T1h; } Ip[0] = KP500000000 * (TK + T19); Im[WS(rs, 3)] = KP500000000 * (T19 - TK); Im[WS(rs, 1)] = KP500000000 * (T1w - T1l); T1x = T1d + T1i; T1j = T1d - T1i; Ip[WS(rs, 2)] = KP500000000 * (T1l + T1w); } Rm[WS(rs, 3)] = KP500000000 * (T1x - T1y); Rp[0] = KP500000000 * (T1x + T1y); Rp[WS(rs, 2)] = KP500000000 * (T1j + T1k); Rm[WS(rs, 1)] = KP500000000 * (T1j - T1k); T1N = T1A + T1z; T1B = T1z - T1A; T1R = T1J + T1K; T1L = T1J - T1K; } } } { E T1E, T1O, T1H, T1P; T1E = T1C + T1D; T1O = T1C - T1D; T1H = T1F - T1G; T1P = T1F + T1G; { E T1S, T1Q, T1I, T1M; T1S = T1O + T1P; T1Q = T1O - T1P; T1I = T1E + T1H; T1M = T1H - T1E; Im[0] = -(KP500000000 * (FNMS(KP707106781, T1Q, T1N))); Ip[WS(rs, 3)] = KP500000000 * (FMA(KP707106781, T1Q, T1N)); Rp[WS(rs, 1)] = KP500000000 * (FMA(KP707106781, T1S, T1R)); Rm[WS(rs, 2)] = KP500000000 * (FNMS(KP707106781, T1S, T1R)); Rp[WS(rs, 3)] = KP500000000 * (FMA(KP707106781, T1M, T1L)); Rm[0] = KP500000000 * (FNMS(KP707106781, T1M, T1L)); Im[WS(rs, 2)] = -(KP500000000 * (FNMS(KP707106781, T1I, T1B))); Ip[WS(rs, 1)] = KP500000000 * (FMA(KP707106781, T1I, T1B)); } } } } } static const tw_instr twinstr[] = { {TW_CEXP, 1, 1}, {TW_CEXP, 1, 3}, {TW_CEXP, 1, 7}, {TW_NEXT, 1, 0} }; static const hc2c_desc desc = { 8, "hc2cfdft2_8", twinstr, &GENUS, {60, 36, 30, 0} }; void X(codelet_hc2cfdft2_8) (planner *p) { X(khc2c_register) (p, hc2cfdft2_8, &desc, HC2C_VIA_DFT); } #else /* HAVE_FMA */ /* Generated by: ../../../genfft/gen_hc2cdft.native -compact -variables 4 -pipeline-latency 4 -twiddle-log3 -precompute-twiddles -n 8 -dit -name hc2cfdft2_8 -include hc2cf.h */ /* * This function contains 90 FP additions, 56 FP multiplications, * (or, 72 additions, 38 multiplications, 18 fused multiply/add), * 51 stack variables, 2 constants, and 32 memory accesses */ #include "hc2cf.h" static void hc2cfdft2_8(R *Rp, R *Ip, R *Rm, R *Im, const R *W, stride rs, INT mb, INT me, INT ms) { DK(KP353553390, +0.353553390593273762200422181052424519642417969); DK(KP500000000, +0.500000000000000000000000000000000000000000000); { INT m; for (m = mb, W = W + ((mb - 1) * 6); m < me; m = m + 1, Rp = Rp + ms, Ip = Ip + ms, Rm = Rm - ms, Im = Im - ms, W = W + 6, MAKE_VOLATILE_STRIDE(32, rs)) { E T1, T4, T2, T5, Tu, Ty, T7, Td, Ti, Tj, Tk, TP, To, TN; { E T3, Tc, T6, Tb; T1 = W[0]; T4 = W[1]; T2 = W[2]; T5 = W[3]; T3 = T1 * T2; Tc = T4 * T2; T6 = T4 * T5; Tb = T1 * T5; Tu = T3 - T6; Ty = Tb + Tc; T7 = T3 + T6; Td = Tb - Tc; Ti = W[4]; Tj = W[5]; Tk = FMA(T1, Ti, T4 * Tj); TP = FNMS(Td, Ti, T7 * Tj); To = FNMS(T4, Ti, T1 * Tj); TN = FMA(T7, Ti, Td * Tj); } { E TF, T11, TC, T12, T1d, T1e, T1q, TM, TR, T1p, Th, Ts, T15, T14, T1a; E T1b, T1m, TV, TY, T1n; { E TD, TE, TL, TI, TJ, TK, Tx, TQ, TB, TO; TD = Ip[0]; TE = Im[0]; TL = TD + TE; TI = Rm[0]; TJ = Rp[0]; TK = TI - TJ; { E Tv, Tw, Tz, TA; Tv = Ip[WS(rs, 2)]; Tw = Im[WS(rs, 2)]; Tx = Tv - Tw; TQ = Tv + Tw; Tz = Rp[WS(rs, 2)]; TA = Rm[WS(rs, 2)]; TB = Tz + TA; TO = Tz - TA; } TF = TD - TE; T11 = TJ + TI; TC = FNMS(Ty, TB, Tu * Tx); T12 = FMA(Tu, TB, Ty * Tx); T1d = FNMS(TP, TO, TN * TQ); T1e = FMA(T4, TK, T1 * TL); T1q = T1e - T1d; TM = FNMS(T4, TL, T1 * TK); TR = FMA(TN, TO, TP * TQ); T1p = TR + TM; } { E Ta, TU, Tg, TT, Tn, TX, Tr, TW; { E T8, T9, Te, Tf; T8 = Ip[WS(rs, 1)]; T9 = Im[WS(rs, 1)]; Ta = T8 - T9; TU = T8 + T9; Te = Rp[WS(rs, 1)]; Tf = Rm[WS(rs, 1)]; Tg = Te + Tf; TT = Te - Tf; } { E Tl, Tm, Tp, Tq; Tl = Ip[WS(rs, 3)]; Tm = Im[WS(rs, 3)]; Tn = Tl - Tm; TX = Tl + Tm; Tp = Rp[WS(rs, 3)]; Tq = Rm[WS(rs, 3)]; Tr = Tp + Tq; TW = Tp - Tq; } Th = FNMS(Td, Tg, T7 * Ta); Ts = FNMS(To, Tr, Tk * Tn); T15 = FMA(Tk, Tr, To * Tn); T14 = FMA(T7, Tg, Td * Ta); T1a = FNMS(T5, TT, T2 * TU); T1b = FNMS(Tj, TW, Ti * TX); T1m = T1b - T1a; TV = FMA(T2, TT, T5 * TU); TY = FMA(Ti, TW, Tj * TX); T1n = TV - TY; } { E T1l, T1x, T1A, T1C, T1s, T1w, T1v, T1B; { E T1j, T1k, T1y, T1z; T1j = TF - TC; T1k = T14 - T15; T1l = KP500000000 * (T1j - T1k); T1x = KP500000000 * (T1k + T1j); T1y = T1m - T1n; T1z = T1p + T1q; T1A = KP353553390 * (T1y - T1z); T1C = KP353553390 * (T1y + T1z); } { E T1o, T1r, T1t, T1u; T1o = T1m + T1n; T1r = T1p - T1q; T1s = KP353553390 * (T1o + T1r); T1w = KP353553390 * (T1r - T1o); T1t = T11 - T12; T1u = Th - Ts; T1v = KP500000000 * (T1t - T1u); T1B = KP500000000 * (T1t + T1u); } Ip[WS(rs, 1)] = T1l + T1s; Rp[WS(rs, 1)] = T1B + T1C; Im[WS(rs, 2)] = T1s - T1l; Rm[WS(rs, 2)] = T1B - T1C; Rm[0] = T1v - T1w; Im[0] = T1A - T1x; Rp[WS(rs, 3)] = T1v + T1w; Ip[WS(rs, 3)] = T1x + T1A; } { E TH, T19, T1g, T1i, T10, T18, T17, T1h; { E Tt, TG, T1c, T1f; Tt = Th + Ts; TG = TC + TF; TH = Tt + TG; T19 = TG - Tt; T1c = T1a + T1b; T1f = T1d + T1e; T1g = T1c - T1f; T1i = T1c + T1f; } { E TS, TZ, T13, T16; TS = TM - TR; TZ = TV + TY; T10 = TS - TZ; T18 = TZ + TS; T13 = T11 + T12; T16 = T14 + T15; T17 = T13 - T16; T1h = T13 + T16; } Ip[0] = KP500000000 * (TH + T10); Rp[0] = KP500000000 * (T1h + T1i); Im[WS(rs, 3)] = KP500000000 * (T10 - TH); Rm[WS(rs, 3)] = KP500000000 * (T1h - T1i); Rm[WS(rs, 1)] = KP500000000 * (T17 - T18); Im[WS(rs, 1)] = KP500000000 * (T1g - T19); Rp[WS(rs, 2)] = KP500000000 * (T17 + T18); Ip[WS(rs, 2)] = KP500000000 * (T19 + T1g); } } } } } static const tw_instr twinstr[] = { {TW_CEXP, 1, 1}, {TW_CEXP, 1, 3}, {TW_CEXP, 1, 7}, {TW_NEXT, 1, 0} }; static const hc2c_desc desc = { 8, "hc2cfdft2_8", twinstr, &GENUS, {72, 38, 18, 0} }; void X(codelet_hc2cfdft2_8) (planner *p) { X(khc2c_register) (p, hc2cfdft2_8, &desc, HC2C_VIA_DFT); } #endif /* HAVE_FMA */ fftw-3.3.3/rdft/scalar/r2cf/hc2cfdft2_16.c0000644000175000017500000006230412054410725014673 00000000000000/* * Copyright (c) 2003, 2007-11 Matteo Frigo * Copyright (c) 2003, 2007-11 Massachusetts Institute of Technology * * This program is free software; you can redistribute it and/or modify * it under the terms of the GNU General Public License as published by * the Free Software Foundation; either version 2 of the License, or * (at your option) any later version. * * This program is distributed in the hope that it will be useful, * but WITHOUT ANY WARRANTY; without even the implied warranty of * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the * GNU General Public License for more details. * * You should have received a copy of the GNU General Public License * along with this program; if not, write to the Free Software * Foundation, Inc., 51 Franklin Street, Fifth Floor, Boston, MA 02110-1301 USA * */ /* This file was automatically generated --- DO NOT EDIT */ /* Generated on Sun Nov 25 07:40:50 EST 2012 */ #include "codelet-rdft.h" #ifdef HAVE_FMA /* Generated by: ../../../genfft/gen_hc2cdft.native -fma -reorder-insns -schedule-for-pipeline -compact -variables 4 -pipeline-latency 4 -twiddle-log3 -precompute-twiddles -n 16 -dit -name hc2cfdft2_16 -include hc2cf.h */ /* * This function contains 228 FP additions, 166 FP multiplications, * (or, 136 additions, 74 multiplications, 92 fused multiply/add), * 103 stack variables, 4 constants, and 64 memory accesses */ #include "hc2cf.h" static void hc2cfdft2_16(R *Rp, R *Ip, R *Rm, R *Im, const R *W, stride rs, INT mb, INT me, INT ms) { DK(KP923879532, +0.923879532511286756128183189396788286822416626); DK(KP414213562, +0.414213562373095048801688724209698078569671875); DK(KP707106781, +0.707106781186547524400844362104849039284835938); DK(KP500000000, +0.500000000000000000000000000000000000000000000); { INT m; for (m = mb, W = W + ((mb - 1) * 8); m < me; m = m + 1, Rp = Rp + ms, Ip = Ip + ms, Rm = Rm - ms, Im = Im - ms, W = W + 8, MAKE_VOLATILE_STRIDE(64, rs)) { E T4p, T4o, T4n, T4s; { E T1, T2, Tw, Ty, Th, T3, Tx, TE, Ti, TK, Tj, T4, T5; T1 = W[0]; T2 = W[2]; Tw = W[6]; Ty = W[7]; Th = W[4]; T3 = T1 * T2; Tx = T1 * Tw; TE = T1 * Ty; Ti = T1 * Th; TK = T2 * Th; Tj = W[5]; T4 = W[1]; T5 = W[3]; { E T1v, T2q, T1s, T2s, T38, T3T, T1Y, T3P, T17, T1h, T2x, T2v, T33, T3Q, T3S; E T1N, Tv, T3A, T2E, T3B, T3L, T2c, T3I, T2S, TW, T3E, T3J, T2n, T3D, T2J; E T3M, T2X; { E TF, Tk, Tz, TL, T6, TR, Tq, Tc, T2h, T25, T2k, T29, T1G, T1M, T2P; E T2R; { E T18, TY, T1d, T13, T1H, T1A, T1K, T1E, T37, T1R, T35, T1X; { E T1j, T1o, T1W, T1p, T1m, T1Q, T1U, T1q; { E T1k, T1l, T1S, T1T; { E T1t, T28, T24, T1D, T1z, T1u, TQ, Tp, Tb; T1t = Ip[0]; TQ = T2 * Tj; Tp = T1 * Tj; TF = FNMS(T4, Tw, TE); T1j = FMA(T4, Tj, Ti); Tk = FNMS(T4, Tj, Ti); Tz = FMA(T4, Ty, Tx); T18 = FNMS(T5, Tj, TK); TL = FMA(T5, Tj, TK); TY = FNMS(T4, T5, T3); T6 = FMA(T4, T5, T3); Tb = T1 * T5; TR = FNMS(T5, Th, TQ); T1d = FMA(T5, Th, TQ); Tq = FMA(T4, Th, Tp); T1o = FNMS(T4, Th, Tp); T28 = T6 * Tj; T24 = T6 * Th; T1D = TY * Tj; T1z = TY * Th; Tc = FNMS(T4, T2, Tb); T13 = FMA(T4, T2, Tb); T1u = Im[0]; T1k = Ip[WS(rs, 4)]; T2h = FMA(Tc, Tj, T24); T25 = FNMS(Tc, Tj, T24); T2k = FNMS(Tc, Th, T28); T29 = FMA(Tc, Th, T28); T1H = FNMS(T13, Tj, T1z); T1A = FMA(T13, Tj, T1z); T1K = FMA(T13, Th, T1D); T1E = FNMS(T13, Th, T1D); T1W = T1t + T1u; T1v = T1t - T1u; T1l = Im[WS(rs, 4)]; } T1S = Rm[0]; T1T = Rp[0]; T1p = Rp[WS(rs, 4)]; T1m = T1k - T1l; T1Q = T1k + T1l; T2q = T1T + T1S; T1U = T1S - T1T; T1q = Rm[WS(rs, 4)]; } { E T36, T1V, T1O, T1r, T1n, T1P, T34, T2r; T36 = T4 * T1U; T1V = T1 * T1U; T1O = T1q - T1p; T1r = T1p + T1q; T1n = T1j * T1m; T37 = FMA(T1, T1W, T36); T2r = T1j * T1r; T1P = Th * T1O; T34 = Tj * T1O; T1s = FNMS(T1o, T1r, T1n); T2s = FMA(T1o, T1m, T2r); T1R = FNMS(Tj, T1Q, T1P); T35 = FMA(Th, T1Q, T34); T1X = FNMS(T4, T1W, T1V); } } { E T1F, T11, T1e, T16, T1L, T1b, T1f, T1C, T2Z; { E T14, T15, TZ, T10, T19, T1a, T1B; TZ = Ip[WS(rs, 2)]; T10 = Im[WS(rs, 2)]; T38 = T35 + T37; T3T = T37 - T35; T1Y = T1R + T1X; T3P = T1X - T1R; T1F = TZ + T10; T11 = TZ - T10; T14 = Rp[WS(rs, 2)]; T15 = Rm[WS(rs, 2)]; T19 = Ip[WS(rs, 6)]; T1a = Im[WS(rs, 6)]; T1e = Rp[WS(rs, 6)]; T16 = T14 + T15; T1B = T15 - T14; T1L = T19 + T1a; T1b = T19 - T1a; T1f = Rm[WS(rs, 6)]; T1C = T1A * T1B; T2Z = T1E * T1B; } { E T1J, T31, T2u, T30, T32; { E T12, T1g, T1I, T1c, T2w; T12 = TY * T11; T1g = T1e + T1f; T1I = T1f - T1e; T1c = T18 * T1b; T17 = FNMS(T13, T16, T12); T2w = T18 * T1g; T1J = T1H * T1I; T31 = T1K * T1I; T1h = FNMS(T1d, T1g, T1c); T2x = FMA(T1d, T1b, T2w); } T2u = TY * T16; T30 = FMA(T1A, T1F, T2Z); T32 = FMA(T1H, T1L, T31); T1G = FNMS(T1E, T1F, T1C); T2v = FMA(T13, T11, T2u); T1M = FNMS(T1K, T1L, T1J); T33 = T30 + T32; T3Q = T30 - T32; } } } { E Tl, T22, T9, T20, Tf, T2O, Ta, T21, T2A, Tm, Tr, Ts; { E T7, T8, Td, Te; T7 = Ip[WS(rs, 1)]; T3S = T1G - T1M; T1N = T1G + T1M; T8 = Im[WS(rs, 1)]; Td = Rp[WS(rs, 1)]; Te = Rm[WS(rs, 1)]; Tl = Ip[WS(rs, 5)]; T22 = T7 + T8; T9 = T7 - T8; T20 = Td - Te; Tf = Td + Te; T2O = T2 * T22; Ta = T6 * T9; T21 = T2 * T20; T2A = T6 * Tf; Tm = Im[WS(rs, 5)]; Tr = Rp[WS(rs, 5)]; Ts = Rm[WS(rs, 5)]; } { E Tg, T2a, Tn, T26, T2Q, T27, T2C, T2B, Tu, Tt, To, T23, T2D, T2b; Tg = FNMS(Tc, Tf, Ta); T2a = Tl + Tm; Tn = Tl - Tm; T26 = Tr - Ts; Tt = Tr + Ts; T2Q = T25 * T2a; To = Tk * Tn; T27 = T25 * T26; T2C = Tk * Tt; T2B = FMA(Tc, T9, T2A); Tu = FNMS(Tq, Tt, To); T23 = FMA(T5, T22, T21); T2D = FMA(Tq, Tn, T2C); T2b = FMA(T29, T2a, T27); Tv = Tg + Tu; T3A = Tg - Tu; T2P = FNMS(T5, T20, T2O); T2E = T2B + T2D; T3B = T2B - T2D; T3L = T2b - T23; T2c = T23 + T2b; T2R = FNMS(T29, T26, T2Q); } } { E T2f, TC, T2T, TD, T2d, TI, TS, T2e, T2F, T2l, TO, TT; { E TG, TH, TA, TB, TM, TN; TA = Ip[WS(rs, 7)]; TB = Im[WS(rs, 7)]; TG = Rp[WS(rs, 7)]; T3I = T2R - T2P; T2S = T2P + T2R; T2f = TA + TB; TC = TA - TB; TH = Rm[WS(rs, 7)]; TM = Ip[WS(rs, 3)]; T2T = Tw * T2f; TD = Tz * TC; T2d = TG - TH; TI = TG + TH; TN = Im[WS(rs, 3)]; TS = Rp[WS(rs, 3)]; T2e = Tw * T2d; T2F = Tz * TI; T2l = TM + TN; TO = TM - TN; TT = Rm[WS(rs, 3)]; } { E TJ, T2V, TP, T2i, TU, T2G; TJ = FNMS(TF, TI, TD); T2V = T2h * T2l; TP = TL * TO; T2i = TS - TT; TU = TS + TT; T2G = FMA(TF, TC, T2F); { E T2g, T2j, TV, T2H; T2g = FMA(Ty, T2f, T2e); T2j = T2h * T2i; TV = FNMS(TR, TU, TP); T2H = TL * TU; { E T2U, T2m, T2I, T2W; T2U = FNMS(Ty, T2d, T2T); T2m = FMA(T2k, T2l, T2j); TW = TJ + TV; T3E = TJ - TV; T2I = FMA(TR, TO, T2H); T2W = FNMS(T2k, T2i, T2V); T3J = T2m - T2g; T2n = T2g + T2m; T3D = T2G - T2I; T2J = T2G + T2I; T3M = T2U - T2W; T2X = T2U + T2W; } } } } } { E T3Y, T3x, T3X, T3y, T3r, T3q, T3p, T3u; { E T2Y, T3o, TX, T3s, T3i, T39, T3t, T3l, T3e, T1x, T2M, T2p, T3d, T2K, T2t; E T2y; { E T2o, T1Z, T3j, T3k, T1i, T1w, T3g, T3h; T2Y = T2S + T2X; T3g = T2X - T2S; T3h = T2c - T2n; T2o = T2c + T2n; T1Z = T1N + T1Y; T3j = T1Y - T1N; T3o = Tv - TW; TX = Tv + TW; T3s = T3g - T3h; T3i = T3g + T3h; T3k = T38 - T33; T39 = T33 + T38; T3Y = T17 - T1h; T1i = T17 + T1h; T1w = T1s + T1v; T3x = T1v - T1s; T3t = T3j + T3k; T3l = T3j - T3k; T3e = T1w - T1i; T1x = T1i + T1w; T2M = T2o + T1Z; T2p = T1Z - T2o; T3d = T2J - T2E; T2K = T2E + T2J; T3X = T2q - T2s; T2t = T2q + T2s; T2y = T2v + T2x; T3y = T2v - T2x; } { E T2N, T3c, T3a, T3n, T3b, T2L, T2z, T1y; T2N = T1x - TX; T1y = TX + T1x; T3c = T2Y + T39; T3a = T2Y - T39; T3n = T2t - T2y; T2z = T2t + T2y; Ip[0] = KP500000000 * (T1y + T2p); Im[WS(rs, 7)] = KP500000000 * (T2p - T1y); T3b = T2z + T2K; T2L = T2z - T2K; { E T3f, T3m, T3v, T3w; T3r = T3e - T3d; T3f = T3d + T3e; Im[WS(rs, 3)] = KP500000000 * (T3a - T2N); Ip[WS(rs, 4)] = KP500000000 * (T2N + T3a); Rp[WS(rs, 4)] = KP500000000 * (T2L + T2M); Rm[WS(rs, 3)] = KP500000000 * (T2L - T2M); Rp[0] = KP500000000 * (T3b + T3c); Rm[WS(rs, 7)] = KP500000000 * (T3b - T3c); T3m = T3i + T3l; T3q = T3l - T3i; T3p = T3n - T3o; T3v = T3n + T3o; T3w = T3s + T3t; T3u = T3s - T3t; Im[WS(rs, 5)] = -(KP500000000 * (FNMS(KP707106781, T3m, T3f))); Ip[WS(rs, 2)] = KP500000000 * (FMA(KP707106781, T3m, T3f)); Rp[WS(rs, 2)] = KP500000000 * (FMA(KP707106781, T3w, T3v)); Rm[WS(rs, 5)] = KP500000000 * (FNMS(KP707106781, T3w, T3v)); } } } { E T3R, T4b, T3z, T4q, T4g, T3U, T40, T41, T4r, T4j, T4m, T3G, T46, T3O, T4l; E T3Z, T4c; { E T3K, T3N, T4h, T4i, T3C, T3F, T4e, T4f; Rp[WS(rs, 6)] = KP500000000 * (FMA(KP707106781, T3q, T3p)); Rm[WS(rs, 1)] = KP500000000 * (FNMS(KP707106781, T3q, T3p)); Im[WS(rs, 1)] = -(KP500000000 * (FNMS(KP707106781, T3u, T3r))); Ip[WS(rs, 6)] = KP500000000 * (FMA(KP707106781, T3u, T3r)); T3K = T3I + T3J; T4e = T3I - T3J; T4f = T3M - T3L; T3N = T3L + T3M; T3R = T3P - T3Q; T4h = T3Q + T3P; T4b = T3y + T3x; T3z = T3x - T3y; T4q = FNMS(KP414213562, T4e, T4f); T4g = FMA(KP414213562, T4f, T4e); T4i = T3T - T3S; T3U = T3S + T3T; T40 = T3B + T3A; T3C = T3A - T3B; T3F = T3D + T3E; T41 = T3D - T3E; T4r = FNMS(KP414213562, T4h, T4i); T4j = FMA(KP414213562, T4i, T4h); T4m = T3C - T3F; T3G = T3C + T3F; T46 = FNMS(KP414213562, T3K, T3N); T3O = FMA(KP414213562, T3N, T3K); T4l = T3X - T3Y; T3Z = T3X + T3Y; } { E T45, T3H, T42, T47, T3V; T45 = FNMS(KP707106781, T3G, T3z); T3H = FMA(KP707106781, T3G, T3z); T4c = T41 - T40; T42 = T40 + T41; T47 = FMA(KP414213562, T3R, T3U); T3V = FNMS(KP414213562, T3U, T3R); { E T49, T43, T48, T4a, T44, T3W; T49 = FMA(KP707106781, T42, T3Z); T43 = FNMS(KP707106781, T42, T3Z); T48 = T46 - T47; T4a = T46 + T47; T44 = T3V - T3O; T3W = T3O + T3V; Rp[WS(rs, 1)] = KP500000000 * (FMA(KP923879532, T4a, T49)); Rm[WS(rs, 6)] = KP500000000 * (FNMS(KP923879532, T4a, T49)); Rp[WS(rs, 5)] = KP500000000 * (FMA(KP923879532, T44, T43)); Rm[WS(rs, 2)] = KP500000000 * (FNMS(KP923879532, T44, T43)); Im[WS(rs, 6)] = -(KP500000000 * (FNMS(KP923879532, T3W, T3H))); Ip[WS(rs, 1)] = KP500000000 * (FMA(KP923879532, T3W, T3H)); Ip[WS(rs, 5)] = KP500000000 * (FMA(KP923879532, T48, T45)); Im[WS(rs, 2)] = -(KP500000000 * (FNMS(KP923879532, T48, T45))); } } { E T4d, T4k, T4t, T4u; T4p = FMA(KP707106781, T4c, T4b); T4d = FNMS(KP707106781, T4c, T4b); T4k = T4g - T4j; T4o = T4g + T4j; T4n = FMA(KP707106781, T4m, T4l); T4t = FNMS(KP707106781, T4m, T4l); T4u = T4q + T4r; T4s = T4q - T4r; Im[0] = -(KP500000000 * (FNMS(KP923879532, T4k, T4d))); Ip[WS(rs, 7)] = KP500000000 * (FMA(KP923879532, T4k, T4d)); Rm[0] = KP500000000 * (FMA(KP923879532, T4u, T4t)); Rp[WS(rs, 7)] = KP500000000 * (FNMS(KP923879532, T4u, T4t)); } } } } } Rp[WS(rs, 3)] = KP500000000 * (FMA(KP923879532, T4o, T4n)); Rm[WS(rs, 4)] = KP500000000 * (FNMS(KP923879532, T4o, T4n)); Im[WS(rs, 4)] = -(KP500000000 * (FNMS(KP923879532, T4s, T4p))); Ip[WS(rs, 3)] = KP500000000 * (FMA(KP923879532, T4s, T4p)); } } } static const tw_instr twinstr[] = { {TW_CEXP, 1, 1}, {TW_CEXP, 1, 3}, {TW_CEXP, 1, 9}, {TW_CEXP, 1, 15}, {TW_NEXT, 1, 0} }; static const hc2c_desc desc = { 16, "hc2cfdft2_16", twinstr, &GENUS, {136, 74, 92, 0} }; void X(codelet_hc2cfdft2_16) (planner *p) { X(khc2c_register) (p, hc2cfdft2_16, &desc, HC2C_VIA_DFT); } #else /* HAVE_FMA */ /* Generated by: ../../../genfft/gen_hc2cdft.native -compact -variables 4 -pipeline-latency 4 -twiddle-log3 -precompute-twiddles -n 16 -dit -name hc2cfdft2_16 -include hc2cf.h */ /* * This function contains 228 FP additions, 124 FP multiplications, * (or, 188 additions, 84 multiplications, 40 fused multiply/add), * 91 stack variables, 4 constants, and 64 memory accesses */ #include "hc2cf.h" static void hc2cfdft2_16(R *Rp, R *Ip, R *Rm, R *Im, const R *W, stride rs, INT mb, INT me, INT ms) { DK(KP461939766, +0.461939766255643378064091594698394143411208313); DK(KP191341716, +0.191341716182544885864229992015199433380672281); DK(KP353553390, +0.353553390593273762200422181052424519642417969); DK(KP500000000, +0.500000000000000000000000000000000000000000000); { INT m; for (m = mb, W = W + ((mb - 1) * 8); m < me; m = m + 1, Rp = Rp + ms, Ip = Ip + ms, Rm = Rm - ms, Im = Im - ms, W = W + 8, MAKE_VOLATILE_STRIDE(64, rs)) { E T1, T4, T2, T5, T7, Td, T12, TY, Tk, Ti, Tm, T1l, T1b, TL, T1h; E Ts, TR, T17, Ty, Tz, TA, TE, T1L, T1Q, T1H, T1O, T24, T2d, T20, T2b; { E Tl, TP, Tq, TK, Tj, TQ, Tr, TJ; { E T3, Tc, T6, Tb; T1 = W[0]; T4 = W[1]; T2 = W[2]; T5 = W[3]; T3 = T1 * T2; Tc = T4 * T2; T6 = T4 * T5; Tb = T1 * T5; T7 = T3 + T6; Td = Tb - Tc; T12 = Tb + Tc; TY = T3 - T6; Tk = W[5]; Tl = T4 * Tk; TP = T2 * Tk; Tq = T1 * Tk; TK = T5 * Tk; Ti = W[4]; Tj = T1 * Ti; TQ = T5 * Ti; Tr = T4 * Ti; TJ = T2 * Ti; } Tm = Tj - Tl; T1l = Tq - Tr; T1b = TP + TQ; TL = TJ + TK; T1h = Tj + Tl; Ts = Tq + Tr; TR = TP - TQ; T17 = TJ - TK; Ty = W[6]; Tz = W[7]; TA = FMA(T1, Ty, T4 * Tz); TE = FNMS(T4, Ty, T1 * Tz); { E T1J, T1K, T1F, T1G; T1J = TY * Tk; T1K = T12 * Ti; T1L = T1J - T1K; T1Q = T1J + T1K; T1F = TY * Ti; T1G = T12 * Tk; T1H = T1F + T1G; T1O = T1F - T1G; } { E T22, T23, T1Y, T1Z; T22 = T7 * Tk; T23 = Td * Ti; T24 = T22 + T23; T2d = T22 - T23; T1Y = T7 * Ti; T1Z = Td * Tk; T20 = T1Y - T1Z; T2b = T1Y + T1Z; } } { E T1t, T3i, T2l, T3B, T1E, T3t, T2M, T3x, T1g, T3C, T2J, T3u, T1T, T3w, T2o; E T3j, Tx, T3b, T2C, T3q, T27, T3m, T2s, T3c, TW, T3f, T2F, T3n, T2g, T3p; E T2v, T3e; { E T1k, T1C, T1o, T1B, T1s, T1z, T1y, T2j, T1p, T2k; { E T1i, T1j, T1m, T1n; T1i = Ip[WS(rs, 4)]; T1j = Im[WS(rs, 4)]; T1k = T1i - T1j; T1C = T1i + T1j; T1m = Rp[WS(rs, 4)]; T1n = Rm[WS(rs, 4)]; T1o = T1m + T1n; T1B = T1m - T1n; } { E T1q, T1r, T1w, T1x; T1q = Ip[0]; T1r = Im[0]; T1s = T1q - T1r; T1z = T1q + T1r; T1w = Rm[0]; T1x = Rp[0]; T1y = T1w - T1x; T2j = T1x + T1w; } T1p = FNMS(T1l, T1o, T1h * T1k); T1t = T1p + T1s; T3i = T1s - T1p; T2k = FMA(T1h, T1o, T1l * T1k); T2l = T2j + T2k; T3B = T2j - T2k; { E T1A, T1D, T2K, T2L; T1A = FNMS(T4, T1z, T1 * T1y); T1D = FMA(Ti, T1B, Tk * T1C); T1E = T1A - T1D; T3t = T1D + T1A; T2K = FNMS(Tk, T1B, Ti * T1C); T2L = FMA(T4, T1y, T1 * T1z); T2M = T2K + T2L; T3x = T2L - T2K; } } { E T11, T1M, T15, T1I, T1a, T1R, T1e, T1P; { E TZ, T10, T13, T14; TZ = Ip[WS(rs, 2)]; T10 = Im[WS(rs, 2)]; T11 = TZ - T10; T1M = TZ + T10; T13 = Rp[WS(rs, 2)]; T14 = Rm[WS(rs, 2)]; T15 = T13 + T14; T1I = T13 - T14; } { E T18, T19, T1c, T1d; T18 = Ip[WS(rs, 6)]; T19 = Im[WS(rs, 6)]; T1a = T18 - T19; T1R = T18 + T19; T1c = Rp[WS(rs, 6)]; T1d = Rm[WS(rs, 6)]; T1e = T1c + T1d; T1P = T1c - T1d; } { E T16, T1f, T2H, T2I; T16 = FNMS(T12, T15, TY * T11); T1f = FNMS(T1b, T1e, T17 * T1a); T1g = T16 + T1f; T3C = T16 - T1f; T2H = FNMS(T1L, T1I, T1H * T1M); T2I = FNMS(T1Q, T1P, T1O * T1R); T2J = T2H + T2I; T3u = T2H - T2I; } { E T1N, T1S, T2m, T2n; T1N = FMA(T1H, T1I, T1L * T1M); T1S = FMA(T1O, T1P, T1Q * T1R); T1T = T1N + T1S; T3w = T1S - T1N; T2m = FMA(TY, T15, T12 * T11); T2n = FMA(T17, T1e, T1b * T1a); T2o = T2m + T2n; T3j = T2m - T2n; } } { E Ta, T1W, Tg, T1V, Tp, T25, Tv, T21; { E T8, T9, Te, Tf; T8 = Ip[WS(rs, 1)]; T9 = Im[WS(rs, 1)]; Ta = T8 - T9; T1W = T8 + T9; Te = Rp[WS(rs, 1)]; Tf = Rm[WS(rs, 1)]; Tg = Te + Tf; T1V = Te - Tf; } { E Tn, To, Tt, Tu; Tn = Ip[WS(rs, 5)]; To = Im[WS(rs, 5)]; Tp = Tn - To; T25 = Tn + To; Tt = Rp[WS(rs, 5)]; Tu = Rm[WS(rs, 5)]; Tv = Tt + Tu; T21 = Tt - Tu; } { E Th, Tw, T2A, T2B; Th = FNMS(Td, Tg, T7 * Ta); Tw = FNMS(Ts, Tv, Tm * Tp); Tx = Th + Tw; T3b = Th - Tw; T2A = FNMS(T5, T1V, T2 * T1W); T2B = FNMS(T24, T21, T20 * T25); T2C = T2A + T2B; T3q = T2A - T2B; } { E T1X, T26, T2q, T2r; T1X = FMA(T2, T1V, T5 * T1W); T26 = FMA(T20, T21, T24 * T25); T27 = T1X + T26; T3m = T26 - T1X; T2q = FMA(T7, Tg, Td * Ta); T2r = FMA(Tm, Tv, Ts * Tp); T2s = T2q + T2r; T3c = T2q - T2r; } } { E TD, T29, TH, T28, TO, T2e, TU, T2c; { E TB, TC, TF, TG; TB = Ip[WS(rs, 7)]; TC = Im[WS(rs, 7)]; TD = TB - TC; T29 = TB + TC; TF = Rp[WS(rs, 7)]; TG = Rm[WS(rs, 7)]; TH = TF + TG; T28 = TF - TG; } { E TM, TN, TS, TT; TM = Ip[WS(rs, 3)]; TN = Im[WS(rs, 3)]; TO = TM - TN; T2e = TM + TN; TS = Rp[WS(rs, 3)]; TT = Rm[WS(rs, 3)]; TU = TS + TT; T2c = TS - TT; } { E TI, TV, T2D, T2E; TI = FNMS(TE, TH, TA * TD); TV = FNMS(TR, TU, TL * TO); TW = TI + TV; T3f = TI - TV; T2D = FNMS(Tz, T28, Ty * T29); T2E = FNMS(T2d, T2c, T2b * T2e); T2F = T2D + T2E; T3n = T2D - T2E; } { E T2a, T2f, T2t, T2u; T2a = FMA(Ty, T28, Tz * T29); T2f = FMA(T2b, T2c, T2d * T2e); T2g = T2a + T2f; T3p = T2f - T2a; T2t = FMA(TA, TH, TE * TD); T2u = FMA(TL, TU, TR * TO); T2v = T2t + T2u; T3e = T2t - T2u; } } { E T1v, T2z, T2O, T2Q, T2i, T2y, T2x, T2P; { E TX, T1u, T2G, T2N; TX = Tx + TW; T1u = T1g + T1t; T1v = TX + T1u; T2z = T1u - TX; T2G = T2C + T2F; T2N = T2J + T2M; T2O = T2G - T2N; T2Q = T2G + T2N; } { E T1U, T2h, T2p, T2w; T1U = T1E - T1T; T2h = T27 + T2g; T2i = T1U - T2h; T2y = T2h + T1U; T2p = T2l + T2o; T2w = T2s + T2v; T2x = T2p - T2w; T2P = T2p + T2w; } Ip[0] = KP500000000 * (T1v + T2i); Rp[0] = KP500000000 * (T2P + T2Q); Im[WS(rs, 7)] = KP500000000 * (T2i - T1v); Rm[WS(rs, 7)] = KP500000000 * (T2P - T2Q); Rm[WS(rs, 3)] = KP500000000 * (T2x - T2y); Im[WS(rs, 3)] = KP500000000 * (T2O - T2z); Rp[WS(rs, 4)] = KP500000000 * (T2x + T2y); Ip[WS(rs, 4)] = KP500000000 * (T2z + T2O); } { E T2T, T35, T33, T39, T2W, T36, T2Z, T37; { E T2R, T2S, T31, T32; T2R = T2v - T2s; T2S = T1t - T1g; T2T = KP500000000 * (T2R + T2S); T35 = KP500000000 * (T2S - T2R); T31 = T2l - T2o; T32 = Tx - TW; T33 = KP500000000 * (T31 - T32); T39 = KP500000000 * (T31 + T32); } { E T2U, T2V, T2X, T2Y; T2U = T2F - T2C; T2V = T27 - T2g; T2W = T2U + T2V; T36 = T2U - T2V; T2X = T1T + T1E; T2Y = T2M - T2J; T2Z = T2X - T2Y; T37 = T2X + T2Y; } { E T30, T3a, T34, T38; T30 = KP353553390 * (T2W + T2Z); Ip[WS(rs, 2)] = T2T + T30; Im[WS(rs, 5)] = T30 - T2T; T3a = KP353553390 * (T36 + T37); Rm[WS(rs, 5)] = T39 - T3a; Rp[WS(rs, 2)] = T39 + T3a; T34 = KP353553390 * (T2Z - T2W); Rm[WS(rs, 1)] = T33 - T34; Rp[WS(rs, 6)] = T33 + T34; T38 = KP353553390 * (T36 - T37); Ip[WS(rs, 6)] = T35 + T38; Im[WS(rs, 1)] = T38 - T35; } } { E T3k, T3Q, T3Z, T3D, T3h, T40, T3X, T45, T3G, T3P, T3s, T3K, T3U, T44, T3z; E T3L; { E T3d, T3g, T3o, T3r; T3k = KP500000000 * (T3i - T3j); T3Q = KP500000000 * (T3j + T3i); T3Z = KP500000000 * (T3B - T3C); T3D = KP500000000 * (T3B + T3C); T3d = T3b - T3c; T3g = T3e + T3f; T3h = KP353553390 * (T3d + T3g); T40 = KP353553390 * (T3d - T3g); { E T3V, T3W, T3E, T3F; T3V = T3u + T3t; T3W = T3x - T3w; T3X = FNMS(KP461939766, T3W, KP191341716 * T3V); T45 = FMA(KP461939766, T3V, KP191341716 * T3W); T3E = T3c + T3b; T3F = T3e - T3f; T3G = KP353553390 * (T3E + T3F); T3P = KP353553390 * (T3F - T3E); } T3o = T3m + T3n; T3r = T3p - T3q; T3s = FMA(KP191341716, T3o, KP461939766 * T3r); T3K = FNMS(KP191341716, T3r, KP461939766 * T3o); { E T3S, T3T, T3v, T3y; T3S = T3n - T3m; T3T = T3q + T3p; T3U = FMA(KP461939766, T3S, KP191341716 * T3T); T44 = FNMS(KP461939766, T3T, KP191341716 * T3S); T3v = T3t - T3u; T3y = T3w + T3x; T3z = FNMS(KP191341716, T3y, KP461939766 * T3v); T3L = FMA(KP191341716, T3v, KP461939766 * T3y); } } { E T3l, T3A, T3N, T3O; T3l = T3h + T3k; T3A = T3s + T3z; Ip[WS(rs, 1)] = T3l + T3A; Im[WS(rs, 6)] = T3A - T3l; T3N = T3D + T3G; T3O = T3K + T3L; Rm[WS(rs, 6)] = T3N - T3O; Rp[WS(rs, 1)] = T3N + T3O; } { E T3H, T3I, T3J, T3M; T3H = T3D - T3G; T3I = T3z - T3s; Rm[WS(rs, 2)] = T3H - T3I; Rp[WS(rs, 5)] = T3H + T3I; T3J = T3k - T3h; T3M = T3K - T3L; Ip[WS(rs, 5)] = T3J + T3M; Im[WS(rs, 2)] = T3M - T3J; } { E T3R, T3Y, T47, T48; T3R = T3P + T3Q; T3Y = T3U + T3X; Ip[WS(rs, 3)] = T3R + T3Y; Im[WS(rs, 4)] = T3Y - T3R; T47 = T3Z + T40; T48 = T44 + T45; Rm[WS(rs, 4)] = T47 - T48; Rp[WS(rs, 3)] = T47 + T48; } { E T41, T42, T43, T46; T41 = T3Z - T40; T42 = T3X - T3U; Rm[0] = T41 - T42; Rp[WS(rs, 7)] = T41 + T42; T43 = T3Q - T3P; T46 = T44 - T45; Ip[WS(rs, 7)] = T43 + T46; Im[0] = T46 - T43; } } } } } } static const tw_instr twinstr[] = { {TW_CEXP, 1, 1}, {TW_CEXP, 1, 3}, {TW_CEXP, 1, 9}, {TW_CEXP, 1, 15}, {TW_NEXT, 1, 0} }; static const hc2c_desc desc = { 16, "hc2cfdft2_16", twinstr, &GENUS, {188, 84, 40, 0} }; void X(codelet_hc2cfdft2_16) (planner *p) { X(khc2c_register) (p, hc2cfdft2_16, &desc, HC2C_VIA_DFT); } #endif /* HAVE_FMA */ fftw-3.3.3/rdft/scalar/r2cf/hc2cfdft2_32.c0000644000175000017500000016412312054410740014670 00000000000000/* * Copyright (c) 2003, 2007-11 Matteo Frigo * Copyright (c) 2003, 2007-11 Massachusetts Institute of Technology * * This program is free software; you can redistribute it and/or modify * it under the terms of the GNU General Public License as published by * the Free Software Foundation; either version 2 of the License, or * (at your option) any later version. * * This program is distributed in the hope that it will be useful, * but WITHOUT ANY WARRANTY; without even the implied warranty of * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the * GNU General Public License for more details. * * You should have received a copy of the GNU General Public License * along with this program; if not, write to the Free Software * Foundation, Inc., 51 Franklin Street, Fifth Floor, Boston, MA 02110-1301 USA * */ /* This file was automatically generated --- DO NOT EDIT */ /* Generated on Sun Nov 25 07:40:52 EST 2012 */ #include "codelet-rdft.h" #ifdef HAVE_FMA /* Generated by: ../../../genfft/gen_hc2cdft.native -fma -reorder-insns -schedule-for-pipeline -compact -variables 4 -pipeline-latency 4 -twiddle-log3 -precompute-twiddles -n 32 -dit -name hc2cfdft2_32 -include hc2cf.h */ /* * This function contains 552 FP additions, 414 FP multiplications, * (or, 300 additions, 162 multiplications, 252 fused multiply/add), * 196 stack variables, 8 constants, and 128 memory accesses */ #include "hc2cf.h" static void hc2cfdft2_32(R *Rp, R *Ip, R *Rm, R *Im, const R *W, stride rs, INT mb, INT me, INT ms) { DK(KP980785280, +0.980785280403230449126182236134239036973933731); DK(KP831469612, +0.831469612302545237078788377617905756738560812); DK(KP668178637, +0.668178637919298919997757686523080761552472251); DK(KP198912367, +0.198912367379658006911597622644676228597850501); DK(KP923879532, +0.923879532511286756128183189396788286822416626); DK(KP414213562, +0.414213562373095048801688724209698078569671875); DK(KP707106781, +0.707106781186547524400844362104849039284835938); DK(KP500000000, +0.500000000000000000000000000000000000000000000); { INT m; for (m = mb, W = W + ((mb - 1) * 8); m < me; m = m + 1, Rp = Rp + ms, Ip = Ip + ms, Rm = Rm - ms, Im = Im - ms, W = W + 8, MAKE_VOLATILE_STRIDE(128, rs)) { E Tax, TaA; { E T1, Th, T2, T5, Ti, Ty, T1t, T3, Tb, Tj, TY, TK, Tl, T4, Tk; T1 = W[0]; Th = W[4]; T2 = W[2]; T5 = W[3]; Ti = W[6]; Ty = T1 * Th; T1t = T2 * Th; T3 = T1 * T2; Tb = T1 * T5; Tj = Th * Ti; TY = T2 * Ti; TK = T1 * Ti; Tl = W[7]; T4 = W[1]; Tk = W[5]; { E T3j, T7Z, T5b, T93, T6B, T8V, T4d, T8J, T8r, T6e, T8l, T1T, T8C, T54, T8i; E T5O, T94, T31, T8K, T6w, T8U, T3Y, T80, T5g, T8B, T69, T8h, T1s, T8q, T4T; E T8k, T5J, Tx, T8a, T5y, T8d, T4s, T5Y, T8v, T8E, T2k, T82, T6l, T3z, T83; E T5m, T8X, T8O, T2F, T86, T6q, T3M, T85, T5r, T8Y, T8R, TW, T8e, T8x, T4B; E T5D, T8b, T63, T8w; { E TL, T2l, T1c, Tc, T1a, T6, Tm, T2v, Tz, T2q, TR, Ts, T2A, TF, T1H; E T1g, T1d, T1F, T34, T3F, T3B, T32, T3w, T3s, T4p, T4l, T2f, T29, T4K, T4S; E T5G, T5I; { E TZ, T2R, T2H, T15, T2W, T2M, T4I, T4E, T3V, T3S, T4Q, T4M, T1n, T1h, T4X; E T53, T5L, T5N, T5d, T5f; { E T1u, T1A, T51, T4Y, T28, T25, T44, T40, T1O, T1I, T3b, T35, T4b, T3i, T45; E T38, T39, T58, T49, T3e, T41; { E T3g, T3h, T36, T37, TQ; T3g = Ip[0]; TZ = FNMS(T5, Tl, TY); T2R = FMA(T5, Tl, TY); TQ = T1 * Tl; { E T14, Tr, T1z, TE; T14 = T2 * Tl; Tr = Th * Tl; TL = FMA(T4, Tl, TK); T2l = FNMS(T4, Tl, TK); T1c = FMA(T4, T2, Tb); Tc = FNMS(T4, T2, Tb); T1a = FNMS(T4, T5, T3); T6 = FMA(T4, T5, T3); Tm = FMA(Tk, Tl, Tj); T2v = FNMS(T5, Tk, T1t); T1u = FMA(T5, Tk, T1t); Tz = FNMS(T4, Tk, Ty); T2H = FMA(T4, Tk, Ty); T1z = T2 * Tk; TE = T1 * Tk; T2q = FMA(T4, Ti, TQ); TR = FNMS(T4, Ti, TQ); T15 = FMA(T5, Ti, T14); T2W = FNMS(T5, Ti, T14); Ts = FNMS(Tk, Ti, Tr); { E T1f, T4H, T4D, T1b; T1f = T1a * Tk; T4H = T1a * Tl; T4D = T1a * Ti; T1b = T1a * Th; { E T27, T3E, T3A, T24; T27 = T6 * Tk; T3E = T6 * Tl; T3A = T6 * Ti; T24 = T6 * Th; { E T3v, T3r, T4P, T4L; T3v = T1u * Tl; T3r = T1u * Ti; T4P = T2v * Tl; T4L = T2v * Ti; { E T4o, T4k, T43, T3Z; T4o = T2H * Tl; T4k = T2H * Ti; T43 = Tz * Tl; T3Z = Tz * Ti; T1A = FNMS(T5, Th, T1z); T2A = FMA(T5, Th, T1z); T2M = FNMS(T4, Th, TE); TF = FMA(T4, Th, TE); T1H = FNMS(T1c, Th, T1f); T1g = FMA(T1c, Th, T1f); T51 = FNMS(T1c, Ti, T4H); T4I = FMA(T1c, Ti, T4H); T4Y = FMA(T1c, Tl, T4D); T4E = FNMS(T1c, Tl, T4D); T1d = FNMS(T1c, Tk, T1b); T1F = FMA(T1c, Tk, T1b); T34 = FMA(Tc, Th, T27); T28 = FNMS(Tc, Th, T27); T3V = FNMS(Tc, Ti, T3E); T3F = FMA(Tc, Ti, T3E); T3S = FMA(Tc, Tl, T3A); T3B = FNMS(Tc, Tl, T3A); T25 = FMA(Tc, Tk, T24); T32 = FNMS(Tc, Tk, T24); T3w = FNMS(T1A, Ti, T3v); T3s = FMA(T1A, Tl, T3r); T4Q = FNMS(T2A, Ti, T4P); T4M = FMA(T2A, Tl, T4L); T4p = FNMS(T2M, Ti, T4o); T4l = FMA(T2M, Tl, T4k); T44 = FNMS(TF, Ti, T43); T40 = FMA(TF, Tl, T3Z); { E T1m, T1e, T1N, T1G; T1m = T1d * Tl; T1e = T1d * Ti; T1N = T1F * Tl; T1G = T1F * Ti; { E T2e, T26, T3a, T33; T2e = T25 * Tl; T26 = T25 * Ti; T3a = T32 * Tl; T33 = T32 * Ti; T1n = FNMS(T1g, Ti, T1m); T1h = FMA(T1g, Tl, T1e); T1O = FNMS(T1H, Ti, T1N); T1I = FMA(T1H, Tl, T1G); T2f = FNMS(T28, Ti, T2e); T29 = FMA(T28, Tl, T26); T3b = FNMS(T34, Ti, T3a); T35 = FMA(T34, Tl, T33); T3h = Im[0]; } } } } } } } T36 = Ip[WS(rs, 8)]; T37 = Im[WS(rs, 8)]; { E T47, T48, T3c, T3d; T47 = Rm[0]; T4b = T3g + T3h; T3i = T3g - T3h; T45 = T36 + T37; T38 = T36 - T37; T48 = Rp[0]; T3c = Rp[WS(rs, 8)]; T3d = Rm[WS(rs, 8)]; T39 = T35 * T38; T58 = T48 + T47; T49 = T47 - T48; T3e = T3c + T3d; T41 = T3d - T3c; } } { E T4W, T1x, T1y, T6a, T4U, T1D, T1P, T4V, T5K, T52, T1L, T1Q; { E T1B, T1C, T1J, T1K; { E T1v, T6A, T4c, T5a, T6y, T46, T1w, T6z, T4a; T1v = Ip[WS(rs, 3)]; T6z = T4 * T49; T4a = T1 * T49; { E T3f, T59, T6x, T42; T3f = FNMS(T3b, T3e, T39); T59 = T35 * T3e; T6x = T44 * T41; T42 = T40 * T41; T6A = FMA(T1, T4b, T6z); T4c = FNMS(T4, T4b, T4a); T3j = T3f + T3i; T7Z = T3i - T3f; T5a = FMA(T3b, T38, T59); T6y = FMA(T40, T45, T6x); T46 = FNMS(T44, T45, T42); T1w = Im[WS(rs, 3)]; } T5b = T58 + T5a; T93 = T58 - T5a; T6B = T6y + T6A; T8V = T6A - T6y; T4d = T46 + T4c; T8J = T4c - T46; T4W = T1v + T1w; T1x = T1v - T1w; } T1B = Rp[WS(rs, 3)]; T1C = Rm[WS(rs, 3)]; T1y = T1u * T1x; T6a = T25 * T4W; T1J = Ip[WS(rs, 11)]; T4U = T1B - T1C; T1D = T1B + T1C; T1K = Im[WS(rs, 11)]; T1P = Rp[WS(rs, 11)]; T4V = T25 * T4U; T5K = T1u * T1D; T52 = T1J + T1K; T1L = T1J - T1K; T1Q = Rm[WS(rs, 11)]; } { E T1E, T6c, T1M, T4Z, T1R, T6b; T1E = FNMS(T1A, T1D, T1y); T6c = T4Y * T52; T1M = T1I * T1L; T4Z = T1P - T1Q; T1R = T1P + T1Q; T6b = FNMS(T28, T4U, T6a); { E T5M, T6d, T50, T1S; T4X = FMA(T28, T4W, T4V); T6d = FNMS(T51, T4Z, T6c); T50 = T4Y * T4Z; T1S = FNMS(T1O, T1R, T1M); T5M = T1I * T1R; T8r = T6d - T6b; T6e = T6b + T6d; T8l = T1E - T1S; T1T = T1E + T1S; T53 = FMA(T51, T52, T50); T5L = FMA(T1A, T1x, T5K); T5N = FMA(T1O, T1L, T5M); } } } } { E T3Q, T2K, T2P, T2L, T6s, T3P, T5c, T3W, T2U, T2X, T2Y, T2V; { E T2I, T2J, T2N, T2O, T2S, T3O, T2T; T2I = Ip[WS(rs, 4)]; T8C = T53 - T4X; T54 = T4X + T53; T8i = T5L - T5N; T5O = T5L + T5N; T2J = Im[WS(rs, 4)]; T2N = Rp[WS(rs, 4)]; T2O = Rm[WS(rs, 4)]; T2S = Ip[WS(rs, 12)]; T3Q = T2I + T2J; T2K = T2I - T2J; T3O = T2O - T2N; T2P = T2N + T2O; T2T = Im[WS(rs, 12)]; T2L = T2H * T2K; T6s = Tk * T3O; T3P = Th * T3O; T5c = T2H * T2P; T3W = T2S + T2T; T2U = T2S - T2T; T2X = Rp[WS(rs, 12)]; T2Y = Rm[WS(rs, 12)]; T2V = T2R * T2U; } { E T2Q, T6t, T3T, T2Z, T3R, T6u, T3U; T2Q = FNMS(T2M, T2P, T2L); T6t = FMA(Th, T3Q, T6s); T3T = T2Y - T2X; T2Z = T2X + T2Y; T3R = FNMS(Tk, T3Q, T3P); T5d = FMA(T2M, T2K, T5c); T6u = T3V * T3T; T3U = T3S * T3T; { E T30, T5e, T6v, T3X; T30 = FNMS(T2W, T2Z, T2V); T5e = T2R * T2Z; T6v = FMA(T3S, T3W, T6u); T3X = FNMS(T3V, T3W, T3U); T94 = T2Q - T30; T31 = T2Q + T30; T8K = T6t - T6v; T6w = T6t + T6v; T8U = T3R - T3X; T3Y = T3R + T3X; T5f = FMA(T2W, T2U, T5e); } } } { E T4J, T12, T65, T13, T4F, T18, T1o, T4G, T5F, T4R, T1k, T1p; { E T16, T17, T10, T11, T1i, T1j; T10 = Ip[WS(rs, 15)]; T11 = Im[WS(rs, 15)]; T16 = Rp[WS(rs, 15)]; T80 = T5d - T5f; T5g = T5d + T5f; T4J = T10 + T11; T12 = T10 - T11; T17 = Rm[WS(rs, 15)]; T1i = Ip[WS(rs, 7)]; T65 = T4E * T4J; T13 = TZ * T12; T4F = T16 - T17; T18 = T16 + T17; T1j = Im[WS(rs, 7)]; T1o = Rp[WS(rs, 7)]; T4G = T4E * T4F; T5F = TZ * T18; T4R = T1i + T1j; T1k = T1i - T1j; T1p = Rm[WS(rs, 7)]; } { E T19, T67, T1l, T4N, T1q, T66; T19 = FNMS(T15, T18, T13); T67 = T4M * T4R; T1l = T1h * T1k; T4N = T1o - T1p; T1q = T1o + T1p; T66 = FNMS(T4I, T4F, T65); { E T5H, T68, T4O, T1r; T4K = FMA(T4I, T4J, T4G); T68 = FNMS(T4Q, T4N, T67); T4O = T4M * T4N; T1r = FNMS(T1n, T1q, T1l); T5H = T1h * T1q; T8B = T66 - T68; T69 = T66 + T68; T8h = T19 - T1r; T1s = T19 + T1r; T4S = FMA(T4Q, T4R, T4O); T5G = FMA(T15, T12, T5F); T5I = FMA(T1n, T1k, T5H); } } } } { E T2c, T3x, T2d, T23, T5j, T3q, T2i, T3t, T6i, T8t, T5V, T5X; { E Tn, T4i, T9, T4g, Tf, T5U, Ta, T4h, T5u, To, Tt, Tu; { E T7, T8, Td, Te; T7 = Ip[WS(rs, 1)]; T8q = T4S - T4K; T4T = T4K + T4S; T8k = T5G - T5I; T5J = T5G + T5I; T8 = Im[WS(rs, 1)]; Td = Rp[WS(rs, 1)]; Te = Rm[WS(rs, 1)]; Tn = Ip[WS(rs, 9)]; T4i = T7 + T8; T9 = T7 - T8; T4g = Td - Te; Tf = Td + Te; T5U = T2 * T4i; Ta = T6 * T9; T4h = T2 * T4g; T5u = T6 * Tf; To = Im[WS(rs, 9)]; Tt = Rp[WS(rs, 9)]; Tu = Rm[WS(rs, 9)]; } { E Tg, T4q, Tp, T4m, Tv, T5W, Tq, T4n, T5w; Tg = FNMS(Tc, Tf, Ta); T4q = Tn + To; Tp = Tn - To; T4m = Tt - Tu; Tv = Tt + Tu; T5W = T4l * T4q; Tq = Tm * Tp; T4n = T4l * T4m; T5w = Tm * Tv; { E T5v, Tw, T4j, T5x, T4r; T5v = FMA(Tc, T9, T5u); Tw = FNMS(Ts, Tv, Tq); T4j = FMA(T5, T4i, T4h); T5x = FMA(Ts, Tp, T5w); T4r = FMA(T4p, T4q, T4n); Tx = Tg + Tw; T8a = Tg - Tw; T5y = T5v + T5x; T8d = T5v - T5x; T4s = T4j + T4r; T8t = T4r - T4j; T5V = FNMS(T5, T4g, T5U); T5X = FNMS(T4p, T4m, T5W); } } } { E T3p, T1Y, T1Z, T22, T2g, T6h, T3o, T5i, T2h; { E T20, T21, T1W, T1X, T8u, T2a, T2b, T3n; T1W = Ip[WS(rs, 2)]; T1X = Im[WS(rs, 2)]; T8u = T5V - T5X; T5Y = T5V + T5X; T20 = Rp[WS(rs, 2)]; T3p = T1W + T1X; T1Y = T1W - T1X; T8v = T8t - T8u; T8E = T8u + T8t; T21 = Rm[WS(rs, 2)]; T1Z = T1a * T1Y; T2a = Ip[WS(rs, 10)]; T2b = Im[WS(rs, 10)]; T3n = T21 - T20; T22 = T20 + T21; T2g = Rp[WS(rs, 10)]; T2c = T2a - T2b; T3x = T2a + T2b; T6h = T1H * T3n; T3o = T1F * T3n; T5i = T1a * T22; T2d = T29 * T2c; T2h = Rm[WS(rs, 10)]; } T23 = FNMS(T1c, T22, T1Z); T5j = FMA(T1c, T1Y, T5i); T3q = FNMS(T1H, T3p, T3o); T2i = T2g + T2h; T3t = T2h - T2g; T6i = FMA(T1F, T3p, T6h); } { E T2y, T3K, T2z, T2u, T5o, T3H, T2D, T3I, T6n; { E T3G, T2o, T2p, T2t, T6m, T3D, T5n, T2B, T2C; { E T2r, T2s, T2m, T2n, T3C, T2w, T2x; { E T8N, T8M, T6j, T3u, T2j; T2m = Ip[WS(rs, 14)]; T6j = T3w * T3t; T3u = T3s * T3t; T2j = FNMS(T2f, T2i, T2d); { E T5k, T6k, T3y, T5l; T5k = T29 * T2i; T6k = FMA(T3s, T3x, T6j); T3y = FNMS(T3w, T3x, T3u); T2k = T23 + T2j; T82 = T23 - T2j; T5l = FMA(T2f, T2c, T5k); T6l = T6i + T6k; T8N = T6i - T6k; T3z = T3q + T3y; T8M = T3q - T3y; T83 = T5j - T5l; T5m = T5j + T5l; T2n = Im[WS(rs, 14)]; } T8X = T8M + T8N; T8O = T8M - T8N; } T2r = Rp[WS(rs, 14)]; T3G = T2m + T2n; T2o = T2m - T2n; T2s = Rm[WS(rs, 14)]; T2w = Ip[WS(rs, 6)]; T2x = Im[WS(rs, 6)]; T2p = T2l * T2o; T3C = T2s - T2r; T2t = T2r + T2s; T2y = T2w - T2x; T3K = T2w + T2x; T6m = T3F * T3C; T3D = T3B * T3C; T5n = T2l * T2t; T2z = T2v * T2y; T2B = Rp[WS(rs, 6)]; T2C = Rm[WS(rs, 6)]; } T2u = FNMS(T2q, T2t, T2p); T5o = FMA(T2q, T2o, T5n); T3H = FNMS(T3F, T3G, T3D); T2D = T2B + T2C; T3I = T2C - T2B; T6n = FMA(T3B, T3G, T6m); } { E T4v, TC, T5Z, TD, T4t, TI, TS, T4u, T5z, T4z, TO, TT; { E TG, TH, TA, TB, TM, TN; { E T8Q, T8P, T6o, T3J, T2E; TA = Ip[WS(rs, 5)]; T6o = T1g * T3I; T3J = T1d * T3I; T2E = FNMS(T2A, T2D, T2z); { E T5p, T6p, T3L, T5q; T5p = T2v * T2D; T6p = FMA(T1d, T3K, T6o); T3L = FNMS(T1g, T3K, T3J); T2F = T2u + T2E; T86 = T2u - T2E; T5q = FMA(T2A, T2y, T5p); T6q = T6n + T6p; T8Q = T6n - T6p; T3M = T3H + T3L; T8P = T3H - T3L; T85 = T5o - T5q; T5r = T5o + T5q; TB = Im[WS(rs, 5)]; } T8Y = T8Q - T8P; T8R = T8P + T8Q; } TG = Rp[WS(rs, 5)]; T4v = TA + TB; TC = TA - TB; TH = Rm[WS(rs, 5)]; TM = Ip[WS(rs, 13)]; T5Z = T32 * T4v; TD = Tz * TC; T4t = TG - TH; TI = TG + TH; TN = Im[WS(rs, 13)]; TS = Rp[WS(rs, 13)]; T4u = T32 * T4t; T5z = Tz * TI; T4z = TM + TN; TO = TM - TN; TT = Rm[WS(rs, 13)]; } { E TJ, T61, TP, T4x, TU; TJ = FNMS(TF, TI, TD); T61 = Ti * T4z; TP = TL * TO; T4x = TS - TT; TU = TS + TT; { E T5A, T60, T5C, T62; T5A = FMA(TF, TC, T5z); { E T4w, T4y, TV, T5B, T4A; T4w = FMA(T34, T4v, T4u); T4y = Ti * T4x; TV = FNMS(TR, TU, TP); T5B = TL * TU; T60 = FNMS(T34, T4t, T5Z); T4A = FMA(Tl, T4z, T4y); TW = TJ + TV; T8e = TJ - TV; T5C = FMA(TR, TO, T5B); T8x = T4w - T4A; T4B = T4w + T4A; T62 = FNMS(Tl, T4x, T61); } T5D = T5A + T5C; T8b = T5A - T5C; T63 = T60 + T62; T8w = T62 - T60; } } } } } } { E T74, T78, T8F, T8y, T7s, T72, T75, T77, T7r, T71, T7f, T7d, T7c, T7g, T7m; E T7k, T7j, T7n, T6V, T6Y, T7T, T7W; { E T6S, T1V, T6I, T3l, T6H, T5Q, T6R, T5t, T56, T6g, T6N, T4f, T6M, T6W, T6D; E T6O; { E T2G, T3k, T5E, T5P, TX, T1U, T5h, T5s; T74 = Tx - TW; TX = Tx + TW; T1U = T1s + T1T; T78 = T1s - T1T; T8F = T8w - T8x; T8y = T8w + T8x; T7s = T2k - T2F; T2G = T2k + T2F; T6S = TX - T1U; T1V = TX + T1U; T3k = T31 + T3j; T72 = T3j - T31; T75 = T5y - T5D; T5E = T5y + T5D; T5P = T5J + T5O; T77 = T5J - T5O; T7r = T5b - T5g; T5h = T5b + T5g; T6I = T3k - T2G; T3l = T2G + T3k; T6H = T5P - T5E; T5Q = T5E + T5P; T5s = T5m + T5r; T71 = T5r - T5m; { E T64, T6L, T6f, T4C, T55; T7f = T4B - T4s; T4C = T4s + T4B; T55 = T4T + T54; T7d = T54 - T4T; T7c = T63 - T5Y; T64 = T5Y + T63; T6R = T5h - T5s; T5t = T5h + T5s; T6L = T4C - T55; T56 = T4C + T55; T7g = T69 - T6e; T6f = T69 + T6e; { E T6r, T6C, T3N, T4e, T6K; T7m = T3z - T3M; T3N = T3z + T3M; T4e = T3Y + T4d; T7k = T4d - T3Y; T6K = T6f - T64; T6g = T64 + T6f; T7j = T6q - T6l; T6r = T6l + T6q; T6N = T4e - T3N; T4f = T3N + T4e; T7n = T6B - T6w; T6C = T6w + T6B; T6M = T6K + T6L; T6W = T6K - T6L; T6D = T6r + T6C; T6O = T6C - T6r; } } } { E T5T, T6X, T6P, T6E; { E T5S, T5R, T6F, T6G, T3m, T57; T5T = T3l - T1V; T3m = T1V + T3l; T57 = T4f - T56; T5S = T56 + T4f; T6X = T6N + T6O; T6P = T6N - T6O; T5R = T5t - T5Q; T6F = T5t + T5Q; Im[WS(rs, 15)] = KP500000000 * (T57 - T3m); Ip[0] = KP500000000 * (T3m + T57); T6G = T6g + T6D; T6E = T6g - T6D; Rp[0] = KP500000000 * (T6F + T6G); Rm[WS(rs, 15)] = KP500000000 * (T6F - T6G); Rp[WS(rs, 8)] = KP500000000 * (T5R + T5S); Rm[WS(rs, 7)] = KP500000000 * (T5R - T5S); } { E T6U, T6T, T6Z, T70, T6J, T6Q; T6V = T6I - T6H; T6J = T6H + T6I; T6Q = T6M + T6P; T6U = T6P - T6M; T6T = T6R - T6S; T6Z = T6R + T6S; Im[WS(rs, 7)] = KP500000000 * (T6E - T5T); Ip[WS(rs, 8)] = KP500000000 * (T5T + T6E); Im[WS(rs, 11)] = -(KP500000000 * (FNMS(KP707106781, T6Q, T6J))); Ip[WS(rs, 4)] = KP500000000 * (FMA(KP707106781, T6Q, T6J)); T70 = T6W + T6X; T6Y = T6W - T6X; Rp[WS(rs, 4)] = KP500000000 * (FMA(KP707106781, T70, T6Z)); Rm[WS(rs, 11)] = KP500000000 * (FNMS(KP707106781, T70, T6Z)); Rp[WS(rs, 12)] = KP500000000 * (FMA(KP707106781, T6U, T6T)); Rm[WS(rs, 3)] = KP500000000 * (FNMS(KP707106781, T6U, T6T)); } } } { E T7F, T73, T7P, T7t, T7G, T7w, T7Q, T7a, T7L, T7l, T7K, T7U, T7A, T7i, T7u; E T76; Im[WS(rs, 3)] = -(KP500000000 * (FNMS(KP707106781, T6Y, T6V))); Ip[WS(rs, 12)] = KP500000000 * (FMA(KP707106781, T6Y, T6V)); T7F = T72 - T71; T73 = T71 + T72; T7P = T7r - T7s; T7t = T7r + T7s; T7u = T75 + T74; T76 = T74 - T75; { E T7I, T7e, T7v, T79, T7J, T7h; T7v = T77 - T78; T79 = T77 + T78; T7I = T7c - T7d; T7e = T7c + T7d; T7G = T7v - T7u; T7w = T7u + T7v; T7Q = T76 - T79; T7a = T76 + T79; T7J = T7g - T7f; T7h = T7f + T7g; T7L = T7k - T7j; T7l = T7j + T7k; T7K = FMA(KP414213562, T7J, T7I); T7U = FNMS(KP414213562, T7I, T7J); T7A = FNMS(KP414213562, T7e, T7h); T7i = FMA(KP414213562, T7h, T7e); } { E T7z, T7b, T7D, T7x, T7M, T7o; T7z = FNMS(KP707106781, T7a, T73); T7b = FMA(KP707106781, T7a, T73); T7D = FMA(KP707106781, T7w, T7t); T7x = FNMS(KP707106781, T7w, T7t); T7M = T7n - T7m; T7o = T7m + T7n; { E T7S, T7R, T7X, T7Y; { E T7H, T7V, T7B, T7p, T7O, T7N; T7T = FMA(KP707106781, T7G, T7F); T7H = FNMS(KP707106781, T7G, T7F); T7N = FMA(KP414213562, T7M, T7L); T7V = FNMS(KP414213562, T7L, T7M); T7B = FMA(KP414213562, T7l, T7o); T7p = FNMS(KP414213562, T7o, T7l); T7O = T7K - T7N; T7S = T7K + T7N; T7R = FMA(KP707106781, T7Q, T7P); T7X = FNMS(KP707106781, T7Q, T7P); { E T7C, T7E, T7y, T7q; T7C = T7A - T7B; T7E = T7A + T7B; T7y = T7p - T7i; T7q = T7i + T7p; Im[WS(rs, 1)] = -(KP500000000 * (FNMS(KP923879532, T7O, T7H))); Ip[WS(rs, 14)] = KP500000000 * (FMA(KP923879532, T7O, T7H)); Im[WS(rs, 5)] = -(KP500000000 * (FNMS(KP923879532, T7C, T7z))); Ip[WS(rs, 10)] = KP500000000 * (FMA(KP923879532, T7C, T7z)); Rp[WS(rs, 2)] = KP500000000 * (FMA(KP923879532, T7E, T7D)); Rm[WS(rs, 13)] = KP500000000 * (FNMS(KP923879532, T7E, T7D)); Rp[WS(rs, 10)] = KP500000000 * (FMA(KP923879532, T7y, T7x)); Rm[WS(rs, 5)] = KP500000000 * (FNMS(KP923879532, T7y, T7x)); Im[WS(rs, 13)] = -(KP500000000 * (FNMS(KP923879532, T7q, T7b))); Ip[WS(rs, 2)] = KP500000000 * (FMA(KP923879532, T7q, T7b)); T7Y = T7U + T7V; T7W = T7U - T7V; } } Rm[WS(rs, 1)] = KP500000000 * (FMA(KP923879532, T7Y, T7X)); Rp[WS(rs, 14)] = KP500000000 * (FNMS(KP923879532, T7Y, T7X)); Rp[WS(rs, 6)] = KP500000000 * (FMA(KP923879532, T7S, T7R)); Rm[WS(rs, 9)] = KP500000000 * (FNMS(KP923879532, T7S, T7R)); } } } { E Ta7, Tat, T9l, T89, T9H, Taj, T9v, T99, T9m, T9c, T9w, T8o, Tao, Tay, Tae; E Ta3, T9q, T9A, T9g, T8I, T8Z, T8W, Tak, Taa, Tau, T9O, T9r, T8T, Tar, Taz; E Taf, T9W; { E T9M, T9L, T9J, T9I, T8s, T8G, T8D, Ta0, Tam, T9Z, Ta1, T8z, Ta9, T9K; { E T9F, T81, Ta5, T95, T96, T97, Ta6, T88, T84, T87; T9F = T80 + T7Z; T81 = T7Z - T80; Ta5 = T93 - T94; T95 = T93 + T94; T96 = T83 + T82; T84 = T82 - T83; Im[WS(rs, 9)] = -(KP500000000 * (FNMS(KP923879532, T7W, T7T))); Ip[WS(rs, 6)] = KP500000000 * (FMA(KP923879532, T7W, T7T)); T87 = T85 + T86; T97 = T85 - T86; Ta6 = T84 - T87; T88 = T84 + T87; { E T8j, T9a, T8g, T8m; { E T8c, T9G, T98, T8f; T9M = T8a + T8b; T8c = T8a - T8b; Ta7 = FMA(KP707106781, Ta6, Ta5); Tat = FNMS(KP707106781, Ta6, Ta5); T9l = FNMS(KP707106781, T88, T81); T89 = FMA(KP707106781, T88, T81); T9G = T97 - T96; T98 = T96 + T97; T8f = T8d + T8e; T9L = T8d - T8e; T9J = T8h + T8i; T8j = T8h - T8i; T9H = FMA(KP707106781, T9G, T9F); Taj = FNMS(KP707106781, T9G, T9F); T9v = FNMS(KP707106781, T98, T95); T99 = FMA(KP707106781, T98, T95); T9a = FMA(KP414213562, T8c, T8f); T8g = FNMS(KP414213562, T8f, T8c); T8m = T8k + T8l; T9I = T8k - T8l; } { E T9X, T9Y, T9b, T8n; T8s = T8q + T8r; T9X = T8r - T8q; T9Y = T8F - T8E; T8G = T8E + T8F; T8D = T8B + T8C; Ta0 = T8B - T8C; T9b = FNMS(KP414213562, T8j, T8m); T8n = FMA(KP414213562, T8m, T8j); Tam = FMA(KP707106781, T9Y, T9X); T9Z = FNMS(KP707106781, T9Y, T9X); T9m = T9b - T9a; T9c = T9a + T9b; T9w = T8g - T8n; T8o = T8g + T8n; Ta1 = T8y - T8v; T8z = T8v + T8y; } } } { E T9o, T8A, Tan, Ta2, T9p, T8H; Tan = FMA(KP707106781, Ta1, Ta0); Ta2 = FNMS(KP707106781, Ta1, Ta0); T9o = FNMS(KP707106781, T8z, T8s); T8A = FMA(KP707106781, T8z, T8s); Tao = FMA(KP198912367, Tan, Tam); Tay = FNMS(KP198912367, Tam, Tan); Tae = FMA(KP668178637, T9Z, Ta2); Ta3 = FNMS(KP668178637, Ta2, T9Z); T9p = FNMS(KP707106781, T8G, T8D); T8H = FMA(KP707106781, T8G, T8D); Ta9 = FNMS(KP414213562, T9I, T9J); T9K = FMA(KP414213562, T9J, T9I); T9q = FNMS(KP668178637, T9p, T9o); T9A = FMA(KP668178637, T9o, T9p); T9g = FNMS(KP198912367, T8A, T8H); T8I = FMA(KP198912367, T8H, T8A); } { E T8L, T9T, Tap, T9S, T9U, T8S, Taq, T9V; { E T9Q, T9R, Ta8, T9N; T8L = T8J - T8K; T9Q = T8K + T8J; T9R = T8X - T8Y; T8Z = T8X + T8Y; T8W = T8U + T8V; T9T = T8V - T8U; Ta8 = FMA(KP414213562, T9L, T9M); T9N = FNMS(KP414213562, T9M, T9L); Tap = FMA(KP707106781, T9R, T9Q); T9S = FNMS(KP707106781, T9R, T9Q); Tak = Ta8 + Ta9; Taa = Ta8 - Ta9; Tau = T9N + T9K; T9O = T9K - T9N; T9U = T8R - T8O; T8S = T8O + T8R; } Taq = FMA(KP707106781, T9U, T9T); T9V = FNMS(KP707106781, T9U, T9T); T9r = FNMS(KP707106781, T8S, T8L); T8T = FMA(KP707106781, T8S, T8L); Tar = FMA(KP198912367, Taq, Tap); Taz = FNMS(KP198912367, Tap, Taq); Taf = FMA(KP668178637, T9S, T9V); T9W = FNMS(KP668178637, T9V, T9S); } } { E T9z, T9C, Tad, Tag; { E T9f, T8p, T9j, T9d, T9s, T90; T9f = FNMS(KP923879532, T8o, T89); T8p = FMA(KP923879532, T8o, T89); T9j = FMA(KP923879532, T9c, T99); T9d = FNMS(KP923879532, T9c, T99); T9s = FNMS(KP707106781, T8Z, T8W); T90 = FMA(KP707106781, T8Z, T8W); { E T9y, T9x, T9D, T9E; { E T9n, T9B, T9h, T91, T9u, T9t; T9z = FMA(KP923879532, T9m, T9l); T9n = FNMS(KP923879532, T9m, T9l); T9t = FMA(KP668178637, T9s, T9r); T9B = FNMS(KP668178637, T9r, T9s); T9h = FMA(KP198912367, T8T, T90); T91 = FNMS(KP198912367, T90, T8T); T9u = T9q + T9t; T9y = T9t - T9q; T9x = FMA(KP923879532, T9w, T9v); T9D = FNMS(KP923879532, T9w, T9v); { E T9i, T9k, T9e, T92; T9i = T9g - T9h; T9k = T9g + T9h; T9e = T91 - T8I; T92 = T8I + T91; Im[WS(rs, 2)] = -(KP500000000 * (FMA(KP831469612, T9u, T9n))); Ip[WS(rs, 13)] = KP500000000 * (FNMS(KP831469612, T9u, T9n)); Im[WS(rs, 6)] = -(KP500000000 * (FNMS(KP980785280, T9i, T9f))); Ip[WS(rs, 9)] = KP500000000 * (FMA(KP980785280, T9i, T9f)); Rp[WS(rs, 1)] = KP500000000 * (FMA(KP980785280, T9k, T9j)); Rm[WS(rs, 14)] = KP500000000 * (FNMS(KP980785280, T9k, T9j)); Rp[WS(rs, 9)] = KP500000000 * (FMA(KP980785280, T9e, T9d)); Rm[WS(rs, 6)] = KP500000000 * (FNMS(KP980785280, T9e, T9d)); Im[WS(rs, 14)] = -(KP500000000 * (FNMS(KP980785280, T92, T8p))); Ip[WS(rs, 1)] = KP500000000 * (FMA(KP980785280, T92, T8p)); T9E = T9A + T9B; T9C = T9A - T9B; } } Rm[WS(rs, 2)] = KP500000000 * (FMA(KP831469612, T9E, T9D)); Rp[WS(rs, 13)] = KP500000000 * (FNMS(KP831469612, T9E, T9D)); Rp[WS(rs, 5)] = KP500000000 * (FMA(KP831469612, T9y, T9x)); Rm[WS(rs, 10)] = KP500000000 * (FNMS(KP831469612, T9y, T9x)); } } { E Tac, Tab, Tah, Tai, T9P, Ta4; Tad = FNMS(KP923879532, T9O, T9H); T9P = FMA(KP923879532, T9O, T9H); Ta4 = T9W - Ta3; Tac = Ta3 + T9W; Tab = FNMS(KP923879532, Taa, Ta7); Tah = FMA(KP923879532, Taa, Ta7); Im[WS(rs, 10)] = -(KP500000000 * (FNMS(KP831469612, T9C, T9z))); Ip[WS(rs, 5)] = KP500000000 * (FMA(KP831469612, T9C, T9z)); Im[WS(rs, 12)] = -(KP500000000 * (FNMS(KP831469612, Ta4, T9P))); Ip[WS(rs, 3)] = KP500000000 * (FMA(KP831469612, Ta4, T9P)); Tai = Tae + Taf; Tag = Tae - Taf; Rp[WS(rs, 3)] = KP500000000 * (FMA(KP831469612, Tai, Tah)); Rm[WS(rs, 12)] = KP500000000 * (FNMS(KP831469612, Tai, Tah)); Rp[WS(rs, 11)] = KP500000000 * (FMA(KP831469612, Tac, Tab)); Rm[WS(rs, 4)] = KP500000000 * (FNMS(KP831469612, Tac, Tab)); } { E Taw, Tav, TaB, TaC, Tal, Tas; Tax = FNMS(KP923879532, Tak, Taj); Tal = FMA(KP923879532, Tak, Taj); Tas = Tao - Tar; Taw = Tao + Tar; Tav = FNMS(KP923879532, Tau, Tat); TaB = FMA(KP923879532, Tau, Tat); Im[WS(rs, 4)] = -(KP500000000 * (FNMS(KP831469612, Tag, Tad))); Ip[WS(rs, 11)] = KP500000000 * (FMA(KP831469612, Tag, Tad)); Im[0] = -(KP500000000 * (FNMS(KP980785280, Tas, Tal))); Ip[WS(rs, 15)] = KP500000000 * (FMA(KP980785280, Tas, Tal)); TaC = Tay + Taz; TaA = Tay - Taz; Rm[0] = KP500000000 * (FMA(KP980785280, TaC, TaB)); Rp[WS(rs, 15)] = KP500000000 * (FNMS(KP980785280, TaC, TaB)); Rp[WS(rs, 7)] = KP500000000 * (FMA(KP980785280, Taw, Tav)); Rm[WS(rs, 8)] = KP500000000 * (FNMS(KP980785280, Taw, Tav)); } } } } } } Im[WS(rs, 8)] = -(KP500000000 * (FNMS(KP980785280, TaA, Tax))); Ip[WS(rs, 7)] = KP500000000 * (FMA(KP980785280, TaA, Tax)); } } } static const tw_instr twinstr[] = { {TW_CEXP, 1, 1}, {TW_CEXP, 1, 3}, {TW_CEXP, 1, 9}, {TW_CEXP, 1, 27}, {TW_NEXT, 1, 0} }; static const hc2c_desc desc = { 32, "hc2cfdft2_32", twinstr, &GENUS, {300, 162, 252, 0} }; void X(codelet_hc2cfdft2_32) (planner *p) { X(khc2c_register) (p, hc2cfdft2_32, &desc, HC2C_VIA_DFT); } #else /* HAVE_FMA */ /* Generated by: ../../../genfft/gen_hc2cdft.native -compact -variables 4 -pipeline-latency 4 -twiddle-log3 -precompute-twiddles -n 32 -dit -name hc2cfdft2_32 -include hc2cf.h */ /* * This function contains 552 FP additions, 300 FP multiplications, * (or, 440 additions, 188 multiplications, 112 fused multiply/add), * 166 stack variables, 9 constants, and 128 memory accesses */ #include "hc2cf.h" static void hc2cfdft2_32(R *Rp, R *Ip, R *Rm, R *Im, const R *W, stride rs, INT mb, INT me, INT ms) { DK(KP277785116, +0.277785116509801112371415406974266437187468595); DK(KP415734806, +0.415734806151272618539394188808952878369280406); DK(KP097545161, +0.097545161008064133924142434238511120463845809); DK(KP490392640, +0.490392640201615224563091118067119518486966865); DK(KP707106781, +0.707106781186547524400844362104849039284835938); DK(KP191341716, +0.191341716182544885864229992015199433380672281); DK(KP461939766, +0.461939766255643378064091594698394143411208313); DK(KP353553390, +0.353553390593273762200422181052424519642417969); DK(KP500000000, +0.500000000000000000000000000000000000000000000); { INT m; for (m = mb, W = W + ((mb - 1) * 8); m < me; m = m + 1, Rp = Rp + ms, Ip = Ip + ms, Rm = Rm - ms, Im = Im - ms, W = W + 8, MAKE_VOLATILE_STRIDE(128, rs)) { E T1, T4, T2, T5, T7, T1b, T1d, Td, Ti, Tk, Tj, Tl, TL, TR, T2h; E T2O, T16, T2l, T10, T2K, Tm, Tq, T3s, T3K, T3w, T3M, T4e, T4u, T4i, T4w; E Ty, TE, T3h, T3j, T2q, T2u, T4l, T4n, T1v, T1B, T3E, T3G, T2B, T2F, T3Y; E T40, T1f, T1G, T1i, T1H, T1j, T1M, T1n, T1I, T23, T2U, T26, T2V, T27, T30; E T2b, T2W; { E Tw, T1A, TD, T1t, Tx, T1z, TC, T1u, TJ, T15, TQ, TY, TK, T14, TP; E TZ; { E T3, Tc, T6, Tb; T1 = W[0]; T4 = W[1]; T2 = W[2]; T5 = W[3]; T3 = T1 * T2; Tc = T4 * T2; T6 = T4 * T5; Tb = T1 * T5; T7 = T3 + T6; T1b = T3 - T6; T1d = Tb + Tc; Td = Tb - Tc; Ti = W[4]; Tw = T1 * Ti; T1A = T5 * Ti; TD = T4 * Ti; T1t = T2 * Ti; Tk = W[5]; Tx = T4 * Tk; T1z = T2 * Tk; TC = T1 * Tk; T1u = T5 * Tk; Tj = W[6]; TJ = T1 * Tj; T15 = T5 * Tj; TQ = T4 * Tj; TY = T2 * Tj; Tl = W[7]; TK = T4 * Tl; T14 = T2 * Tl; TP = T1 * Tl; TZ = T5 * Tl; } TL = TJ + TK; TR = TP - TQ; T2h = TJ - TK; T2O = T14 - T15; T16 = T14 + T15; T2l = TP + TQ; T10 = TY - TZ; T2K = TY + TZ; Tm = FMA(Ti, Tj, Tk * Tl); Tq = FNMS(Tk, Tj, Ti * Tl); { E T3q, T3r, T3u, T3v; T3q = T7 * Tj; T3r = Td * Tl; T3s = T3q + T3r; T3K = T3q - T3r; T3u = T7 * Tl; T3v = Td * Tj; T3w = T3u - T3v; T3M = T3u + T3v; } { E T4c, T4d, T4g, T4h; T4c = T1b * Tj; T4d = T1d * Tl; T4e = T4c - T4d; T4u = T4c + T4d; T4g = T1b * Tl; T4h = T1d * Tj; T4i = T4g + T4h; T4w = T4g - T4h; Ty = Tw - Tx; TE = TC + TD; T3h = FMA(Ty, Tj, TE * Tl); T3j = FNMS(TE, Tj, Ty * Tl); } T2q = T1t - T1u; T2u = T1z + T1A; T4l = FMA(T2q, Tj, T2u * Tl); T4n = FNMS(T2u, Tj, T2q * Tl); T1v = T1t + T1u; T1B = T1z - T1A; T3E = FMA(T1v, Tj, T1B * Tl); T3G = FNMS(T1B, Tj, T1v * Tl); T2B = Tw + Tx; T2F = TC - TD; T3Y = FMA(T2B, Tj, T2F * Tl); T40 = FNMS(T2F, Tj, T2B * Tl); { E T1c, T1e, T1g, T1h; T1c = T1b * Ti; T1e = T1d * Tk; T1f = T1c - T1e; T1G = T1c + T1e; T1g = T1b * Tk; T1h = T1d * Ti; T1i = T1g + T1h; T1H = T1g - T1h; } T1j = FMA(T1f, Tj, T1i * Tl); T1M = FNMS(T1H, Tj, T1G * Tl); T1n = FNMS(T1i, Tj, T1f * Tl); T1I = FMA(T1G, Tj, T1H * Tl); { E T21, T22, T24, T25; T21 = T7 * Ti; T22 = Td * Tk; T23 = T21 + T22; T2U = T21 - T22; T24 = T7 * Tk; T25 = Td * Ti; T26 = T24 - T25; T2V = T24 + T25; } T27 = FMA(T23, Tj, T26 * Tl); T30 = FNMS(T2V, Tj, T2U * Tl); T2b = FNMS(T26, Tj, T23 * Tl); T2W = FMA(T2U, Tj, T2V * Tl); } { E T38, T7l, T7S, T8Y, T7Z, T91, T3A, T6k, T4F, T83, T5C, T6n, T2T, T84, T4I; E T7m, T2g, T4M, T4P, T2z, T3T, T6m, T7O, T7V, T7j, T87, T5v, T6j, T7L, T7U; E T7g, T86, Tv, TW, T61, T4U, T4X, T62, T4b, T6c, T7v, T7C, T5g, T6f, T74; E T8G, T7s, T7B, T71, T8F, T1s, T1R, T65, T51, T54, T64, T4A, T6g, T7G, T8U; E T5n, T6d, T7b, T8J, T7z, T8R, T78, T8I; { E T2E, T2I, T3p, T5w, T37, T4D, T3g, T5A, T2N, T2R, T3y, T5x, T2Z, T33, T3l; E T5z; { E T2C, T2D, T3o, T2G, T2H, T3n; T2C = Ip[WS(rs, 4)]; T2D = Im[WS(rs, 4)]; T3o = T2C + T2D; T2G = Rp[WS(rs, 4)]; T2H = Rm[WS(rs, 4)]; T3n = T2G - T2H; T2E = T2C - T2D; T2I = T2G + T2H; T3p = FMA(Ti, T3n, Tk * T3o); T5w = FNMS(Tk, T3n, Ti * T3o); } { E T35, T36, T3f, T3c, T3d, T3e; T35 = Ip[0]; T36 = Im[0]; T3f = T35 + T36; T3c = Rm[0]; T3d = Rp[0]; T3e = T3c - T3d; T37 = T35 - T36; T4D = T3d + T3c; T3g = FNMS(T4, T3f, T1 * T3e); T5A = FMA(T4, T3e, T1 * T3f); } { E T2L, T2M, T3x, T2P, T2Q, T3t; T2L = Ip[WS(rs, 12)]; T2M = Im[WS(rs, 12)]; T3x = T2L + T2M; T2P = Rp[WS(rs, 12)]; T2Q = Rm[WS(rs, 12)]; T3t = T2P - T2Q; T2N = T2L - T2M; T2R = T2P + T2Q; T3y = FMA(T3s, T3t, T3w * T3x); T5x = FNMS(T3w, T3t, T3s * T3x); } { E T2X, T2Y, T3k, T31, T32, T3i; T2X = Ip[WS(rs, 8)]; T2Y = Im[WS(rs, 8)]; T3k = T2X + T2Y; T31 = Rp[WS(rs, 8)]; T32 = Rm[WS(rs, 8)]; T3i = T31 - T32; T2Z = T2X - T2Y; T33 = T31 + T32; T3l = FMA(T3h, T3i, T3j * T3k); T5z = FNMS(T3j, T3i, T3h * T3k); } { E T34, T7Q, T7R, T4E, T5y, T5B; T34 = FNMS(T30, T33, T2W * T2Z); T38 = T34 + T37; T7l = T37 - T34; T7Q = T3l + T3g; T7R = T5w - T5x; T7S = T7Q - T7R; T8Y = T7R + T7Q; { E T7X, T7Y, T3m, T3z; T7X = T3y - T3p; T7Y = T5A - T5z; T7Z = T7X + T7Y; T91 = T7Y - T7X; T3m = T3g - T3l; T3z = T3p + T3y; T3A = T3m - T3z; T6k = T3z + T3m; } T4E = FMA(T2W, T33, T30 * T2Z); T4F = T4D + T4E; T83 = T4D - T4E; T5y = T5w + T5x; T5B = T5z + T5A; T5C = T5y + T5B; T6n = T5B - T5y; { E T2J, T2S, T4G, T4H; T2J = FNMS(T2F, T2I, T2B * T2E); T2S = FNMS(T2O, T2R, T2K * T2N); T2T = T2J + T2S; T84 = T2J - T2S; T4G = FMA(T2B, T2I, T2F * T2E); T4H = FMA(T2K, T2R, T2O * T2N); T4I = T4G + T4H; T7m = T4G - T4H; } } } { E T20, T5p, T3D, T4K, T2y, T5t, T3R, T4O, T2f, T5q, T3I, T4L, T2p, T5s, T3O; E T4N; { E T1W, T3C, T1Z, T3B; { E T1U, T1V, T1X, T1Y; T1U = Ip[WS(rs, 2)]; T1V = Im[WS(rs, 2)]; T1W = T1U - T1V; T3C = T1U + T1V; T1X = Rp[WS(rs, 2)]; T1Y = Rm[WS(rs, 2)]; T1Z = T1X + T1Y; T3B = T1X - T1Y; } T20 = FNMS(T1d, T1Z, T1b * T1W); T5p = FNMS(T1H, T3B, T1G * T3C); T3D = FMA(T1G, T3B, T1H * T3C); T4K = FMA(T1b, T1Z, T1d * T1W); } { E T2t, T3Q, T2x, T3P; { E T2r, T2s, T2v, T2w; T2r = Ip[WS(rs, 6)]; T2s = Im[WS(rs, 6)]; T2t = T2r - T2s; T3Q = T2r + T2s; T2v = Rp[WS(rs, 6)]; T2w = Rm[WS(rs, 6)]; T2x = T2v + T2w; T3P = T2v - T2w; } T2y = FNMS(T2u, T2x, T2q * T2t); T5t = FNMS(T1i, T3P, T1f * T3Q); T3R = FMA(T1f, T3P, T1i * T3Q); T4O = FMA(T2q, T2x, T2u * T2t); } { E T2a, T3H, T2e, T3F; { E T28, T29, T2c, T2d; T28 = Ip[WS(rs, 10)]; T29 = Im[WS(rs, 10)]; T2a = T28 - T29; T3H = T28 + T29; T2c = Rp[WS(rs, 10)]; T2d = Rm[WS(rs, 10)]; T2e = T2c + T2d; T3F = T2c - T2d; } T2f = FNMS(T2b, T2e, T27 * T2a); T5q = FNMS(T3G, T3F, T3E * T3H); T3I = FMA(T3E, T3F, T3G * T3H); T4L = FMA(T27, T2e, T2b * T2a); } { E T2k, T3N, T2o, T3L; { E T2i, T2j, T2m, T2n; T2i = Ip[WS(rs, 14)]; T2j = Im[WS(rs, 14)]; T2k = T2i - T2j; T3N = T2i + T2j; T2m = Rp[WS(rs, 14)]; T2n = Rm[WS(rs, 14)]; T2o = T2m + T2n; T3L = T2m - T2n; } T2p = FNMS(T2l, T2o, T2h * T2k); T5s = FNMS(T3M, T3L, T3K * T3N); T3O = FMA(T3K, T3L, T3M * T3N); T4N = FMA(T2h, T2o, T2l * T2k); } { E T3J, T3S, T5r, T5u; T2g = T20 + T2f; T4M = T4K + T4L; T4P = T4N + T4O; T2z = T2p + T2y; T3J = T3D + T3I; T3S = T3O + T3R; T3T = T3J + T3S; T6m = T3S - T3J; { E T7M, T7N, T7h, T7i; T7M = T5s - T5t; T7N = T3R - T3O; T7O = T7M + T7N; T7V = T7M - T7N; T7h = T4N - T4O; T7i = T2p - T2y; T7j = T7h + T7i; T87 = T7h - T7i; } T5r = T5p + T5q; T5u = T5s + T5t; T5v = T5r + T5u; T6j = T5u - T5r; { E T7J, T7K, T7e, T7f; T7J = T3I - T3D; T7K = T5p - T5q; T7L = T7J - T7K; T7U = T7K + T7J; T7e = T20 - T2f; T7f = T4K - T4L; T7g = T7e - T7f; T86 = T7f + T7e; } } } { E Th, T5a, T3X, T4S, TV, T5e, T49, T4W, Tu, T5b, T42, T4T, TI, T5d, T46; E T4V; { E Ta, T3W, Tg, T3V; { E T8, T9, Te, Tf; T8 = Ip[WS(rs, 1)]; T9 = Im[WS(rs, 1)]; Ta = T8 - T9; T3W = T8 + T9; Te = Rp[WS(rs, 1)]; Tf = Rm[WS(rs, 1)]; Tg = Te + Tf; T3V = Te - Tf; } Th = FNMS(Td, Tg, T7 * Ta); T5a = FNMS(T5, T3V, T2 * T3W); T3X = FMA(T2, T3V, T5 * T3W); T4S = FMA(T7, Tg, Td * Ta); } { E TO, T48, TU, T47; { E TM, TN, TS, TT; TM = Ip[WS(rs, 13)]; TN = Im[WS(rs, 13)]; TO = TM - TN; T48 = TM + TN; TS = Rp[WS(rs, 13)]; TT = Rm[WS(rs, 13)]; TU = TS + TT; T47 = TS - TT; } TV = FNMS(TR, TU, TL * TO); T5e = FNMS(Tl, T47, Tj * T48); T49 = FMA(Tj, T47, Tl * T48); T4W = FMA(TL, TU, TR * TO); } { E Tp, T41, Tt, T3Z; { E Tn, To, Tr, Ts; Tn = Ip[WS(rs, 9)]; To = Im[WS(rs, 9)]; Tp = Tn - To; T41 = Tn + To; Tr = Rp[WS(rs, 9)]; Ts = Rm[WS(rs, 9)]; Tt = Tr + Ts; T3Z = Tr - Ts; } Tu = FNMS(Tq, Tt, Tm * Tp); T5b = FNMS(T40, T3Z, T3Y * T41); T42 = FMA(T3Y, T3Z, T40 * T41); T4T = FMA(Tm, Tt, Tq * Tp); } { E TB, T45, TH, T44; { E Tz, TA, TF, TG; Tz = Ip[WS(rs, 5)]; TA = Im[WS(rs, 5)]; TB = Tz - TA; T45 = Tz + TA; TF = Rp[WS(rs, 5)]; TG = Rm[WS(rs, 5)]; TH = TF + TG; T44 = TF - TG; } TI = FNMS(TE, TH, Ty * TB); T5d = FNMS(T2V, T44, T2U * T45); T46 = FMA(T2U, T44, T2V * T45); T4V = FMA(Ty, TH, TE * TB); } Tv = Th + Tu; TW = TI + TV; T61 = Tv - TW; T4U = T4S + T4T; T4X = T4V + T4W; T62 = T4U - T4X; { E T43, T4a, T7t, T7u; T43 = T3X + T42; T4a = T46 + T49; T4b = T43 + T4a; T6c = T4a - T43; T7t = T5e - T5d; T7u = T46 - T49; T7v = T7t + T7u; T7C = T7t - T7u; } { E T5c, T5f, T72, T73; T5c = T5a + T5b; T5f = T5d + T5e; T5g = T5c + T5f; T6f = T5f - T5c; T72 = T4S - T4T; T73 = TI - TV; T74 = T72 + T73; T8G = T72 - T73; } { E T7q, T7r, T6Z, T70; T7q = T42 - T3X; T7r = T5a - T5b; T7s = T7q - T7r; T7B = T7r + T7q; T6Z = Th - Tu; T70 = T4V - T4W; T71 = T6Z - T70; T8F = T6Z + T70; } } { E T1a, T5h, T4k, T4Z, T1Q, T5l, T4y, T53, T1r, T5i, T4p, T50, T1F, T5k, T4t; E T52; { E T13, T4j, T19, T4f; { E T11, T12, T17, T18; T11 = Ip[WS(rs, 15)]; T12 = Im[WS(rs, 15)]; T13 = T11 - T12; T4j = T11 + T12; T17 = Rp[WS(rs, 15)]; T18 = Rm[WS(rs, 15)]; T19 = T17 + T18; T4f = T17 - T18; } T1a = FNMS(T16, T19, T10 * T13); T5h = FNMS(T4i, T4f, T4e * T4j); T4k = FMA(T4e, T4f, T4i * T4j); T4Z = FMA(T10, T19, T16 * T13); } { E T1L, T4x, T1P, T4v; { E T1J, T1K, T1N, T1O; T1J = Ip[WS(rs, 11)]; T1K = Im[WS(rs, 11)]; T1L = T1J - T1K; T4x = T1J + T1K; T1N = Rp[WS(rs, 11)]; T1O = Rm[WS(rs, 11)]; T1P = T1N + T1O; T4v = T1N - T1O; } T1Q = FNMS(T1M, T1P, T1I * T1L); T5l = FNMS(T4w, T4v, T4u * T4x); T4y = FMA(T4u, T4v, T4w * T4x); T53 = FMA(T1I, T1P, T1M * T1L); } { E T1m, T4o, T1q, T4m; { E T1k, T1l, T1o, T1p; T1k = Ip[WS(rs, 7)]; T1l = Im[WS(rs, 7)]; T1m = T1k - T1l; T4o = T1k + T1l; T1o = Rp[WS(rs, 7)]; T1p = Rm[WS(rs, 7)]; T1q = T1o + T1p; T4m = T1o - T1p; } T1r = FNMS(T1n, T1q, T1j * T1m); T5i = FNMS(T4n, T4m, T4l * T4o); T4p = FMA(T4l, T4m, T4n * T4o); T50 = FMA(T1j, T1q, T1n * T1m); } { E T1y, T4s, T1E, T4r; { E T1w, T1x, T1C, T1D; T1w = Ip[WS(rs, 3)]; T1x = Im[WS(rs, 3)]; T1y = T1w - T1x; T4s = T1w + T1x; T1C = Rp[WS(rs, 3)]; T1D = Rm[WS(rs, 3)]; T1E = T1C + T1D; T4r = T1C - T1D; } T1F = FNMS(T1B, T1E, T1v * T1y); T5k = FNMS(T26, T4r, T23 * T4s); T4t = FMA(T23, T4r, T26 * T4s); T52 = FMA(T1v, T1E, T1B * T1y); } T1s = T1a + T1r; T1R = T1F + T1Q; T65 = T1s - T1R; T51 = T4Z + T50; T54 = T52 + T53; T64 = T51 - T54; { E T4q, T4z, T7E, T7F; T4q = T4k + T4p; T4z = T4t + T4y; T4A = T4q + T4z; T6g = T4z - T4q; T7E = T5h - T5i; T7F = T4y - T4t; T7G = T7E + T7F; T8U = T7E - T7F; } { E T5j, T5m, T79, T7a; T5j = T5h + T5i; T5m = T5k + T5l; T5n = T5j + T5m; T6d = T5j - T5m; T79 = T4Z - T50; T7a = T1F - T1Q; T7b = T79 + T7a; T8J = T79 - T7a; } { E T7x, T7y, T76, T77; T7x = T4p - T4k; T7y = T5k - T5l; T7z = T7x - T7y; T8R = T7x + T7y; T76 = T1a - T1r; T77 = T52 - T53; T78 = T76 - T77; T8I = T76 + T77; } } { E T1T, T5S, T5M, T5W, T5P, T5X, T3a, T5I, T4C, T58, T56, T5H, T5E, T5G, T4R; E T5R; { E TX, T1S, T5K, T5L; TX = Tv + TW; T1S = T1s + T1R; T1T = TX + T1S; T5S = TX - T1S; T5K = T5n - T5g; T5L = T4b - T4A; T5M = T5K + T5L; T5W = T5K - T5L; } { E T5N, T5O, T2A, T39; T5N = T3T + T3A; T5O = T5C - T5v; T5P = T5N - T5O; T5X = T5N + T5O; T2A = T2g + T2z; T39 = T2T + T38; T3a = T2A + T39; T5I = T39 - T2A; } { E T3U, T4B, T4Y, T55; T3U = T3A - T3T; T4B = T4b + T4A; T4C = T3U - T4B; T58 = T4B + T3U; T4Y = T4U + T4X; T55 = T51 + T54; T56 = T4Y + T55; T5H = T55 - T4Y; } { E T5o, T5D, T4J, T4Q; T5o = T5g + T5n; T5D = T5v + T5C; T5E = T5o - T5D; T5G = T5o + T5D; T4J = T4F + T4I; T4Q = T4M + T4P; T4R = T4J + T4Q; T5R = T4J - T4Q; } { E T3b, T5F, T57, T59; T3b = T1T + T3a; Ip[0] = KP500000000 * (T3b + T4C); Im[WS(rs, 15)] = KP500000000 * (T4C - T3b); T5F = T4R + T56; Rm[WS(rs, 15)] = KP500000000 * (T5F - T5G); Rp[0] = KP500000000 * (T5F + T5G); T57 = T4R - T56; Rm[WS(rs, 7)] = KP500000000 * (T57 - T58); Rp[WS(rs, 8)] = KP500000000 * (T57 + T58); T59 = T3a - T1T; Ip[WS(rs, 8)] = KP500000000 * (T59 + T5E); Im[WS(rs, 7)] = KP500000000 * (T5E - T59); } { E T5J, T5Q, T5Z, T60; T5J = KP500000000 * (T5H + T5I); T5Q = KP353553390 * (T5M + T5P); Ip[WS(rs, 4)] = T5J + T5Q; Im[WS(rs, 11)] = T5Q - T5J; T5Z = KP500000000 * (T5R + T5S); T60 = KP353553390 * (T5W + T5X); Rm[WS(rs, 11)] = T5Z - T60; Rp[WS(rs, 4)] = T5Z + T60; } { E T5T, T5U, T5V, T5Y; T5T = KP500000000 * (T5R - T5S); T5U = KP353553390 * (T5P - T5M); Rm[WS(rs, 3)] = T5T - T5U; Rp[WS(rs, 12)] = T5T + T5U; T5V = KP500000000 * (T5I - T5H); T5Y = KP353553390 * (T5W - T5X); Ip[WS(rs, 12)] = T5V + T5Y; Im[WS(rs, 3)] = T5Y - T5V; } } { E T67, T6Q, T6K, T6U, T6N, T6V, T6a, T6G, T6i, T6A, T6t, T6P, T6w, T6F, T6p; E T6B; { E T63, T66, T6I, T6J; T63 = T61 - T62; T66 = T64 + T65; T67 = KP353553390 * (T63 + T66); T6Q = KP353553390 * (T63 - T66); T6I = T6d - T6c; T6J = T6g - T6f; T6K = FMA(KP461939766, T6I, KP191341716 * T6J); T6U = FNMS(KP461939766, T6J, KP191341716 * T6I); } { E T6L, T6M, T68, T69; T6L = T6k - T6j; T6M = T6n - T6m; T6N = FNMS(KP461939766, T6M, KP191341716 * T6L); T6V = FMA(KP461939766, T6L, KP191341716 * T6M); T68 = T4P - T4M; T69 = T38 - T2T; T6a = KP500000000 * (T68 + T69); T6G = KP500000000 * (T69 - T68); } { E T6e, T6h, T6r, T6s; T6e = T6c + T6d; T6h = T6f + T6g; T6i = FMA(KP191341716, T6e, KP461939766 * T6h); T6A = FNMS(KP191341716, T6h, KP461939766 * T6e); T6r = T4F - T4I; T6s = T2g - T2z; T6t = KP500000000 * (T6r + T6s); T6P = KP500000000 * (T6r - T6s); } { E T6u, T6v, T6l, T6o; T6u = T62 + T61; T6v = T64 - T65; T6w = KP353553390 * (T6u + T6v); T6F = KP353553390 * (T6v - T6u); T6l = T6j + T6k; T6o = T6m + T6n; T6p = FNMS(KP191341716, T6o, KP461939766 * T6l); T6B = FMA(KP191341716, T6l, KP461939766 * T6o); } { E T6b, T6q, T6D, T6E; T6b = T67 + T6a; T6q = T6i + T6p; Ip[WS(rs, 2)] = T6b + T6q; Im[WS(rs, 13)] = T6q - T6b; T6D = T6t + T6w; T6E = T6A + T6B; Rm[WS(rs, 13)] = T6D - T6E; Rp[WS(rs, 2)] = T6D + T6E; } { E T6x, T6y, T6z, T6C; T6x = T6t - T6w; T6y = T6p - T6i; Rm[WS(rs, 5)] = T6x - T6y; Rp[WS(rs, 10)] = T6x + T6y; T6z = T6a - T67; T6C = T6A - T6B; Ip[WS(rs, 10)] = T6z + T6C; Im[WS(rs, 5)] = T6C - T6z; } { E T6H, T6O, T6X, T6Y; T6H = T6F + T6G; T6O = T6K + T6N; Ip[WS(rs, 6)] = T6H + T6O; Im[WS(rs, 9)] = T6O - T6H; T6X = T6P + T6Q; T6Y = T6U + T6V; Rm[WS(rs, 9)] = T6X - T6Y; Rp[WS(rs, 6)] = T6X + T6Y; } { E T6R, T6S, T6T, T6W; T6R = T6P - T6Q; T6S = T6N - T6K; Rm[WS(rs, 1)] = T6R - T6S; Rp[WS(rs, 14)] = T6R + T6S; T6T = T6G - T6F; T6W = T6U - T6V; Ip[WS(rs, 14)] = T6T + T6W; Im[WS(rs, 1)] = T6W - T6T; } } { E T7d, T8w, T7o, T8m, T8c, T8l, T89, T8v, T81, T8B, T8h, T8t, T7I, T8A, T8g; E T8q; { E T75, T7c, T85, T88; T75 = FNMS(KP191341716, T74, KP461939766 * T71); T7c = FMA(KP461939766, T78, KP191341716 * T7b); T7d = T75 + T7c; T8w = T75 - T7c; { E T7k, T7n, T8a, T8b; T7k = KP353553390 * (T7g + T7j); T7n = KP500000000 * (T7l - T7m); T7o = T7k + T7n; T8m = T7n - T7k; T8a = FMA(KP191341716, T71, KP461939766 * T74); T8b = FNMS(KP191341716, T78, KP461939766 * T7b); T8c = T8a + T8b; T8l = T8b - T8a; } T85 = KP500000000 * (T83 + T84); T88 = KP353553390 * (T86 + T87); T89 = T85 + T88; T8v = T85 - T88; { E T7T, T8r, T80, T8s, T7P, T7W; T7P = KP707106781 * (T7L + T7O); T7T = T7P + T7S; T8r = T7S - T7P; T7W = KP707106781 * (T7U + T7V); T80 = T7W + T7Z; T8s = T7Z - T7W; T81 = FNMS(KP097545161, T80, KP490392640 * T7T); T8B = FMA(KP415734806, T8r, KP277785116 * T8s); T8h = FMA(KP097545161, T7T, KP490392640 * T80); T8t = FNMS(KP415734806, T8s, KP277785116 * T8r); } { E T7A, T8o, T7H, T8p, T7w, T7D; T7w = KP707106781 * (T7s + T7v); T7A = T7w + T7z; T8o = T7z - T7w; T7D = KP707106781 * (T7B + T7C); T7H = T7D + T7G; T8p = T7G - T7D; T7I = FMA(KP490392640, T7A, KP097545161 * T7H); T8A = FNMS(KP415734806, T8o, KP277785116 * T8p); T8g = FNMS(KP097545161, T7A, KP490392640 * T7H); T8q = FMA(KP277785116, T8o, KP415734806 * T8p); } } { E T7p, T82, T8j, T8k; T7p = T7d + T7o; T82 = T7I + T81; Ip[WS(rs, 1)] = T7p + T82; Im[WS(rs, 14)] = T82 - T7p; T8j = T89 + T8c; T8k = T8g + T8h; Rm[WS(rs, 14)] = T8j - T8k; Rp[WS(rs, 1)] = T8j + T8k; } { E T8d, T8e, T8f, T8i; T8d = T89 - T8c; T8e = T81 - T7I; Rm[WS(rs, 6)] = T8d - T8e; Rp[WS(rs, 9)] = T8d + T8e; T8f = T7o - T7d; T8i = T8g - T8h; Ip[WS(rs, 9)] = T8f + T8i; Im[WS(rs, 6)] = T8i - T8f; } { E T8n, T8u, T8D, T8E; T8n = T8l + T8m; T8u = T8q + T8t; Ip[WS(rs, 5)] = T8n + T8u; Im[WS(rs, 10)] = T8u - T8n; T8D = T8v + T8w; T8E = T8A + T8B; Rm[WS(rs, 10)] = T8D - T8E; Rp[WS(rs, 5)] = T8D + T8E; } { E T8x, T8y, T8z, T8C; T8x = T8v - T8w; T8y = T8t - T8q; Rm[WS(rs, 2)] = T8x - T8y; Rp[WS(rs, 13)] = T8x + T8y; T8z = T8m - T8l; T8C = T8A - T8B; Ip[WS(rs, 13)] = T8z + T8C; Im[WS(rs, 2)] = T8C - T8z; } } { E T8L, T9u, T8O, T9k, T9a, T9j, T97, T9t, T93, T9z, T9f, T9r, T8W, T9y, T9e; E T9o; { E T8H, T8K, T95, T96; T8H = FNMS(KP461939766, T8G, KP191341716 * T8F); T8K = FMA(KP191341716, T8I, KP461939766 * T8J); T8L = T8H + T8K; T9u = T8H - T8K; { E T8M, T8N, T98, T99; T8M = KP353553390 * (T87 - T86); T8N = KP500000000 * (T7m + T7l); T8O = T8M + T8N; T9k = T8N - T8M; T98 = FMA(KP461939766, T8F, KP191341716 * T8G); T99 = FNMS(KP461939766, T8I, KP191341716 * T8J); T9a = T98 + T99; T9j = T99 - T98; } T95 = KP500000000 * (T83 - T84); T96 = KP353553390 * (T7g - T7j); T97 = T95 + T96; T9t = T95 - T96; { E T8Z, T9p, T92, T9q, T8X, T90; T8X = KP707106781 * (T7V - T7U); T8Z = T8X + T8Y; T9p = T8Y - T8X; T90 = KP707106781 * (T7L - T7O); T92 = T90 + T91; T9q = T91 - T90; T93 = FNMS(KP277785116, T92, KP415734806 * T8Z); T9z = FMA(KP490392640, T9p, KP097545161 * T9q); T9f = FMA(KP277785116, T8Z, KP415734806 * T92); T9r = FNMS(KP490392640, T9q, KP097545161 * T9p); } { E T8S, T9m, T8V, T9n, T8Q, T8T; T8Q = KP707106781 * (T7C - T7B); T8S = T8Q + T8R; T9m = T8R - T8Q; T8T = KP707106781 * (T7s - T7v); T8V = T8T + T8U; T9n = T8U - T8T; T8W = FMA(KP415734806, T8S, KP277785116 * T8V); T9y = FNMS(KP490392640, T9m, KP097545161 * T9n); T9e = FNMS(KP277785116, T8S, KP415734806 * T8V); T9o = FMA(KP097545161, T9m, KP490392640 * T9n); } } { E T8P, T94, T9h, T9i; T8P = T8L + T8O; T94 = T8W + T93; Ip[WS(rs, 3)] = T8P + T94; Im[WS(rs, 12)] = T94 - T8P; T9h = T97 + T9a; T9i = T9e + T9f; Rm[WS(rs, 12)] = T9h - T9i; Rp[WS(rs, 3)] = T9h + T9i; } { E T9b, T9c, T9d, T9g; T9b = T97 - T9a; T9c = T93 - T8W; Rm[WS(rs, 4)] = T9b - T9c; Rp[WS(rs, 11)] = T9b + T9c; T9d = T8O - T8L; T9g = T9e - T9f; Ip[WS(rs, 11)] = T9d + T9g; Im[WS(rs, 4)] = T9g - T9d; } { E T9l, T9s, T9B, T9C; T9l = T9j + T9k; T9s = T9o + T9r; Ip[WS(rs, 7)] = T9l + T9s; Im[WS(rs, 8)] = T9s - T9l; T9B = T9t + T9u; T9C = T9y + T9z; Rm[WS(rs, 8)] = T9B - T9C; Rp[WS(rs, 7)] = T9B + T9C; } { E T9v, T9w, T9x, T9A; T9v = T9t - T9u; T9w = T9r - T9o; Rm[0] = T9v - T9w; Rp[WS(rs, 15)] = T9v + T9w; T9x = T9k - T9j; T9A = T9y - T9z; Ip[WS(rs, 15)] = T9x + T9A; Im[0] = T9A - T9x; } } } } } } static const tw_instr twinstr[] = { {TW_CEXP, 1, 1}, {TW_CEXP, 1, 3}, {TW_CEXP, 1, 9}, {TW_CEXP, 1, 27}, {TW_NEXT, 1, 0} }; static const hc2c_desc desc = { 32, "hc2cfdft2_32", twinstr, &GENUS, {440, 188, 112, 0} }; void X(codelet_hc2cfdft2_32) (planner *p) { X(khc2c_register) (p, hc2cfdft2_32, &desc, HC2C_VIA_DFT); } #endif /* HAVE_FMA */ fftw-3.3.3/rdft/scalar/r2cf/hc2cfdft2_20.c0000644000175000017500000010361512054410727014671 00000000000000/* * Copyright (c) 2003, 2007-11 Matteo Frigo * Copyright (c) 2003, 2007-11 Massachusetts Institute of Technology * * This program is free software; you can redistribute it and/or modify * it under the terms of the GNU General Public License as published by * the Free Software Foundation; either version 2 of the License, or * (at your option) any later version. * * This program is distributed in the hope that it will be useful, * but WITHOUT ANY WARRANTY; without even the implied warranty of * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the * GNU General Public License for more details. * * You should have received a copy of the GNU General Public License * along with this program; if not, write to the Free Software * Foundation, Inc., 51 Franklin Street, Fifth Floor, Boston, MA 02110-1301 USA * */ /* This file was automatically generated --- DO NOT EDIT */ /* Generated on Sun Nov 25 07:40:53 EST 2012 */ #include "codelet-rdft.h" #ifdef HAVE_FMA /* Generated by: ../../../genfft/gen_hc2cdft.native -fma -reorder-insns -schedule-for-pipeline -compact -variables 4 -pipeline-latency 4 -twiddle-log3 -precompute-twiddles -n 20 -dit -name hc2cfdft2_20 -include hc2cf.h */ /* * This function contains 316 FP additions, 238 FP multiplications, * (or, 176 additions, 98 multiplications, 140 fused multiply/add), * 180 stack variables, 5 constants, and 80 memory accesses */ #include "hc2cf.h" static void hc2cfdft2_20(R *Rp, R *Ip, R *Rm, R *Im, const R *W, stride rs, INT mb, INT me, INT ms) { DK(KP951056516, +0.951056516295153572116439333379382143405698634); DK(KP559016994, +0.559016994374947424102293417182819058860154590); DK(KP250000000, +0.250000000000000000000000000000000000000000000); DK(KP500000000, +0.500000000000000000000000000000000000000000000); DK(KP618033988, +0.618033988749894848204586834365638117720309180); { INT m; for (m = mb, W = W + ((mb - 1) * 8); m < me; m = m + 1, Rp = Rp + ms, Ip = Ip + ms, Rm = Rm - ms, Im = Im - ms, W = W + 8, MAKE_VOLATILE_STRIDE(80, rs)) { E T5h, T5C, T5E, T5y, T5w, T5x, T5D, T5z; { E Tm, Tq, Tn, T1, T6, Tg, Tp, Tb, T1i, TU, Tr, TW, Tx, T2B, T1A; E T1u, T2y, T33, T26, T1o, T30, T22, TD, T1Q, T2a, T2e, T2V, T2R, TG, T1V; E TV, TH, TN, T2t, T12, T2p; { E Tw, To, T29, T1h, T1n, T2d, TC, T2U; Tm = W[0]; Tq = W[3]; Tn = W[2]; T1 = W[6]; T6 = W[7]; Tw = Tm * Tq; To = Tm * Tn; T29 = Tm * T1; T1h = Tn * T1; T1n = Tn * T6; T2d = Tm * T6; Tg = W[5]; Tp = W[1]; Tb = W[4]; { E T21, T25, T1t, T1z; T1i = FMA(Tq, T6, T1h); T25 = Tm * Tg; T1z = Tn * Tg; TU = FMA(Tp, Tq, To); Tr = FNMS(Tp, Tq, To); TW = FNMS(Tp, Tn, Tw); Tx = FMA(Tp, Tn, Tw); T1t = Tn * Tb; T21 = Tm * Tb; T2B = FMA(Tq, Tb, T1z); T1A = FNMS(Tq, Tb, T1z); TC = Tr * Tb; T1u = FMA(Tq, Tg, T1t); T2y = FNMS(Tq, Tg, T1t); T33 = FMA(Tp, Tb, T25); T26 = FNMS(Tp, Tb, T25); T1o = FNMS(Tq, T1, T1n); T30 = FNMS(Tp, Tg, T21); T22 = FMA(Tp, Tg, T21); } TD = FMA(Tx, Tg, TC); T1Q = FNMS(Tx, Tg, TC); T2a = FMA(Tp, T6, T29); T2e = FNMS(Tp, T1, T2d); T2U = Tr * T6; { E T2Q, TE, TM, TF; T2Q = Tr * T1; TF = Tr * Tg; T2V = FNMS(Tx, T1, T2U); T2R = FMA(Tx, T6, T2Q); TG = FNMS(Tx, Tb, TF); T1V = FMA(Tx, Tb, TF); TE = TD * T1; TM = TD * T6; TV = TU * Tb; TH = FMA(TG, T6, TE); TN = FNMS(TG, T1, TM); T2t = TU * T1; T12 = TU * Tg; T2p = TU * T6; } } { E T36, T3Q, T5f, T4D, T5g, T2Y, T4E, T3P, T5R, T5k, T39, TT, T3T, T3m, T49; E T4X, T5T, T5r, T3c, T2i, T3W, T3B, T4o, T4U, T5U, T5u, T3d, T2J, T3X, T3I; E T4v, T4V, T5Q, T5n, T3a, T1G, T3U, T3t, T4g, T4Y; { E T13, T2m, T2q, T2u, T2f, T9, T2O, TA, T2c, T4k, T3i, T5, T2Z, T1e, T2G; E T1O, T2W, TQ, T2C, T1Y, T3v, T27, Tj, T1l, T2v, T3g, T1m, T1D, T2n, T1x; E T2k, T3E, T4c, T2l, T1y, T10, T31, T16, T34, T32, T11, T4B, T3p, T4A, T1T; E T3n, T1b, T2A, T4q, T1U, Te, Tf, T24, T4i, T1r, T4a, T3C, T2s, T43, Tv; E T3L, T2N, T45, TL, T3N, T2T, T2E, T1K; { E T2j, TX, T1B, T1C; { E T1c, T1d, T1M, T1N; { E T2, T3, T7, T8; T7 = Rp[WS(rs, 9)]; T8 = Rm[WS(rs, 9)]; T2 = Ip[WS(rs, 9)]; T2j = FMA(TW, Tg, TV); TX = FNMS(TW, Tg, TV); T13 = FMA(TW, Tb, T12); T2m = FNMS(TW, Tb, T12); T2q = FNMS(TW, T1, T2p); T2u = FMA(TW, T6, T2t); T2f = T7 + T8; T9 = T7 - T8; T3 = Im[WS(rs, 9)]; { E Ty, Tz, T2b, T4; Ty = Rp[WS(rs, 2)]; Tz = Rm[WS(rs, 2)]; T1c = Ip[0]; T2b = T2 - T3; T4 = T2 + T3; T2O = Ty - Tz; TA = Ty + Tz; T2c = T2a * T2b; T4k = T2e * T2b; T3i = T6 * T4; T5 = T1 * T4; T1d = Im[0]; T1M = Rp[WS(rs, 1)]; T1N = Rm[WS(rs, 1)]; } } { E TO, TP, T1W, T1X; TO = Rp[WS(rs, 7)]; T2Z = T1c - T1d; T1e = T1c + T1d; T2G = T1M + T1N; T1O = T1M - T1N; TP = Rm[WS(rs, 7)]; T1W = Rm[WS(rs, 6)]; T1X = Rp[WS(rs, 6)]; { E Th, Ti, T1j, T1k; Th = Rm[WS(rs, 4)]; T2W = TO - TP; TQ = TO + TP; T2C = T1X + T1W; T1Y = T1W - T1X; Ti = Rp[WS(rs, 4)]; T1j = Ip[WS(rs, 8)]; T1k = Im[WS(rs, 8)]; T3v = T1Q * T1Y; T27 = Ti + Th; Tj = Th - Ti; T1l = T1j - T1k; T2v = T1j + T1k; T1B = Rp[WS(rs, 3)]; T3g = Tb * Tj; T1m = T1i * T1l; T1C = Rm[WS(rs, 3)]; } } } { E T18, T19, T1R, T1S; { E TY, TZ, T1v, T1w, T14, T15; T1v = Ip[WS(rs, 3)]; T1w = Im[WS(rs, 3)]; TY = Ip[WS(rs, 5)]; T1D = T1B + T1C; T2n = T1B - T1C; T1x = T1v - T1w; T2k = T1v + T1w; T3E = T2j * T2n; T4c = T1u * T1D; T2l = T2j * T2k; T1y = T1u * T1x; TZ = Im[WS(rs, 5)]; T14 = Rp[WS(rs, 5)]; T15 = Rm[WS(rs, 5)]; T18 = Rm[0]; T10 = TY + TZ; T31 = TY - TZ; T16 = T14 - T15; T34 = T14 + T15; T32 = T30 * T31; T11 = TX * T10; T4B = T30 * T34; T3p = TX * T16; T19 = Rp[0]; T1R = Ip[WS(rs, 6)]; T1S = Im[WS(rs, 6)]; } { E T2r, T23, T1p, T1q; { E Tc, T1a, T2z, Td; Tc = Ip[WS(rs, 4)]; T1a = T18 - T19; T4A = T19 + T18; T1T = T1R + T1S; T2z = T1R - T1S; Td = Im[WS(rs, 4)]; T3n = Tm * T1a; T1b = Tp * T1a; T2A = T2y * T2z; T4q = T2B * T2z; T1U = T1Q * T1T; T23 = Tc - Td; Te = Tc + Td; } T1p = Rp[WS(rs, 8)]; T1q = Rm[WS(rs, 8)]; Tf = Tb * Te; T24 = T22 * T23; T4i = T26 * T23; T1r = T1p + T1q; T2r = T1q - T1p; { E T2M, Tu, Ts, Tt; Ts = Ip[WS(rs, 2)]; Tt = Im[WS(rs, 2)]; T4a = T1i * T1r; T3C = T2u * T2r; T2s = T2q * T2r; T2M = Ts + Tt; Tu = Ts - Tt; { E T2S, TK, TI, TJ, T1I, T1J; TI = Ip[WS(rs, 7)]; TJ = Im[WS(rs, 7)]; T43 = Tx * Tu; Tv = Tr * Tu; T3L = TG * T2M; T2N = TD * T2M; T2S = TI + TJ; TK = TI - TJ; T1I = Ip[WS(rs, 1)]; T1J = Im[WS(rs, 1)]; T45 = TN * TK; TL = TH * TK; T3N = T2V * T2S; T2T = T2R * T2S; T2E = T1I - T1J; T1K = T1I + T1J; } } } } } { E T3x, T1L, T2F, T4s, T2P, T2X, T3M, T3O, T35, T4C; T35 = FNMS(T33, T34, T32); T4C = FMA(T33, T31, T4B); T3x = Tq * T1K; T1L = Tn * T1K; T2F = TU * T2E; T4s = TW * T2E; T36 = T2Z - T35; T3Q = T35 + T2Z; T5f = T4A + T4C; T4D = T4A - T4C; T2P = FNMS(TG, T2O, T2N); T2X = FNMS(T2V, T2W, T2T); T3M = FMA(TD, T2O, T3L); T3O = FMA(T2R, T2W, T3N); { E TB, T5j, Tl, T5i, T47, TR, T3h, T3j; { E Ta, Tk, T44, T46; Ta = FNMS(T6, T9, T5); T5g = T2P + T2X; T2Y = T2P - T2X; T4E = T3O - T3M; T3P = T3M + T3O; Tk = FMA(Tg, Tj, Tf); T44 = FMA(Tr, TA, T43); T46 = FMA(TH, TQ, T45); TB = FNMS(Tx, TA, Tv); T5j = Tk + Ta; Tl = Ta - Tk; T5i = T44 + T46; T47 = T44 - T46; TR = FNMS(TN, TQ, TL); T3h = FNMS(Tg, Te, T3g); T3j = FMA(T1, T9, T3i); } { E T3l, T48, T3k, TS; T5R = T5i - T5j; T5k = T5i + T5j; T3l = TB + TR; TS = TB - TR; T48 = T3h + T3j; T3k = T3h - T3j; T39 = TS + Tl; TT = Tl - TS; T3T = T3l + T3k; T3m = T3k - T3l; T49 = T47 + T48; T4X = T47 - T48; } } { E T28, T5q, T20, T5p, T4m, T2g, T3w, T3y; { E T1P, T1Z, T4j, T4l; T1P = FNMS(Tq, T1O, T1L); T1Z = FMA(T1V, T1Y, T1U); T4j = FMA(T22, T27, T4i); T4l = FMA(T2a, T2f, T4k); T28 = FNMS(T26, T27, T24); T5q = T1Z + T1P; T20 = T1P - T1Z; T5p = T4j + T4l; T4m = T4j - T4l; T2g = FNMS(T2e, T2f, T2c); T3w = FNMS(T1V, T1T, T3v); T3y = FMA(Tn, T1O, T3x); } { E T3A, T4n, T3z, T2h; T5T = T5p - T5q; T5r = T5p + T5q; T3A = T28 + T2g; T2h = T28 - T2g; T4n = T3w + T3y; T3z = T3w - T3y; T3c = T2h + T20; T2i = T20 - T2h; T3W = T3A + T3z; T3B = T3z - T3A; T4o = T4m + T4n; T4U = T4m - T4n; } } { E T2D, T5s, T2x, T5t, T4u, T2H, T3D, T3F; { E T2o, T2w, T4r, T4t; T2o = FNMS(T2m, T2n, T2l); T2w = FMA(T2u, T2v, T2s); T4r = FMA(T2y, T2C, T4q); T4t = FMA(TU, T2G, T4s); T2D = FNMS(T2B, T2C, T2A); T5s = T2w + T2o; T2x = T2o - T2w; T5t = T4r + T4t; T4u = T4r - T4t; T2H = FNMS(TW, T2G, T2F); T3D = FNMS(T2q, T2v, T3C); T3F = FMA(T2m, T2k, T3E); } { E T3H, T4p, T3G, T2I; T5U = T5t - T5s; T5u = T5s + T5t; T3H = T2D + T2H; T2I = T2D - T2H; T4p = T3D + T3F; T3G = T3D - T3F; T3d = T2x + T2I; T2J = T2x - T2I; T3X = T3G + T3H; T3I = T3G - T3H; T4v = T4p + T4u; T4V = T4u - T4p; } } { E T1s, T5m, T1g, T5l, T4e, T1E, T3o, T3q; { E T17, T1f, T4b, T4d; T17 = FNMS(T13, T16, T11); T1f = FMA(Tm, T1e, T1b); T4b = FMA(T1o, T1l, T4a); T4d = FMA(T1A, T1x, T4c); T1s = FNMS(T1o, T1r, T1m); T5m = T17 + T1f; T1g = T17 - T1f; T5l = T4b + T4d; T4e = T4b - T4d; T1E = FNMS(T1A, T1D, T1y); T3o = FNMS(Tp, T1e, T3n); T3q = FMA(T13, T10, T3p); } { E T3s, T4f, T3r, T1F; T5Q = T5l - T5m; T5n = T5l + T5m; T3s = T1s + T1E; T1F = T1s - T1E; T4f = T3q + T3o; T3r = T3o - T3q; T3a = T1F + T1g; T1G = T1g - T1F; T3U = T3s + T3r; T3t = T3r - T3s; T4g = T4e + T4f; T4Y = T4e - T4f; } } } } { E T4F, T4G, T4H, T4x, T4z, T41, T4O, T4Q, T40; { E T55, T38, T54, T50, T52, T53, T5e, T5c, T51, T4T; { E T4W, T37, T4Z, T1H, T5b, T5a, T2K, T2L, T4S, T4R; T55 = T4U + T4V; T4W = T4U - T4V; T37 = T2Y + T36; T38 = T36 - T2Y; T54 = T4X + T4Y; T4Z = T4X - T4Y; T1H = TT + T1G; T5b = T1G - TT; T5a = T2J - T2i; T2K = T2i + T2J; T50 = FNMS(KP618033988, T4Z, T4W); T52 = FMA(KP618033988, T4W, T4Z); T2L = T1H + T2K; T4S = T1H - T2K; T53 = T4D - T4E; T4F = T4D + T4E; Im[WS(rs, 4)] = KP500000000 * (T2L - T37); T4R = FMA(KP250000000, T2L, T37); T5e = FMA(KP618033988, T5a, T5b); T5c = FNMS(KP618033988, T5b, T5a); T51 = FNMS(KP559016994, T4S, T4R); T4T = FMA(KP559016994, T4S, T4R); } { E T3b, T4M, T4N, T3e, T3f; { E T4h, T58, T57, T4w, T56, T5d, T59; T4G = T49 + T4g; T4h = T49 - T4g; T58 = T54 - T55; T56 = T54 + T55; Ip[WS(rs, 7)] = KP500000000 * (FMA(KP951056516, T50, T4T)); Ip[WS(rs, 3)] = KP500000000 * (FNMS(KP951056516, T50, T4T)); Im[WS(rs, 8)] = -(KP500000000 * (FNMS(KP951056516, T52, T51))); Im[0] = -(KP500000000 * (FMA(KP951056516, T52, T51))); Rm[WS(rs, 4)] = KP500000000 * (T53 + T56); T57 = FNMS(KP250000000, T56, T53); T4w = T4o - T4v; T4H = T4o + T4v; T3b = T39 + T3a; T4M = T39 - T3a; T5d = FMA(KP559016994, T58, T57); T59 = FNMS(KP559016994, T58, T57); T4x = FMA(KP618033988, T4w, T4h); T4z = FNMS(KP618033988, T4h, T4w); Rp[WS(rs, 7)] = KP500000000 * (FNMS(KP951056516, T5c, T59)); Rp[WS(rs, 3)] = KP500000000 * (FMA(KP951056516, T5c, T59)); Rm[0] = KP500000000 * (FNMS(KP951056516, T5e, T5d)); Rm[WS(rs, 8)] = KP500000000 * (FMA(KP951056516, T5e, T5d)); T4N = T3c - T3d; T3e = T3c + T3d; } T3f = T3b + T3e; T41 = T3b - T3e; T4O = FMA(KP618033988, T4N, T4M); T4Q = FNMS(KP618033988, T4M, T4N); Ip[WS(rs, 5)] = KP500000000 * (T38 + T3f); T40 = FNMS(KP250000000, T3f, T38); } } { E T3S, T5Z, T68, T6a, T64, T62; { E T60, T61, T5Y, T5W, T3R, T67, T66, T3K, T5O, T4K, T4J, T5N, T5X, T5P; { E T5S, T5V, T4y, T42, T4I; T60 = T5R + T5Q; T5S = T5Q - T5R; T5V = T5T - T5U; T61 = T5T + T5U; T4y = FNMS(KP559016994, T41, T40); T42 = FMA(KP559016994, T41, T40); T4I = T4G + T4H; T4K = T4G - T4H; Ip[WS(rs, 9)] = KP500000000 * (FMA(KP951056516, T4x, T42)); Ip[WS(rs, 1)] = KP500000000 * (FNMS(KP951056516, T4x, T42)); Im[WS(rs, 6)] = -(KP500000000 * (FNMS(KP951056516, T4z, T4y))); Im[WS(rs, 2)] = -(KP500000000 * (FMA(KP951056516, T4z, T4y))); Rp[WS(rs, 5)] = KP500000000 * (T4F + T4I); T4J = FNMS(KP250000000, T4I, T4F); T5Y = FMA(KP618033988, T5S, T5V); T5W = FNMS(KP618033988, T5V, T5S); } T3S = T3Q - T3P; T3R = T3P + T3Q; { E T4L, T4P, T3u, T3J; T4L = FMA(KP559016994, T4K, T4J); T4P = FNMS(KP559016994, T4K, T4J); T3u = T3m + T3t; T67 = T3t - T3m; T66 = T3I - T3B; T3J = T3B + T3I; Rp[WS(rs, 9)] = KP500000000 * (FNMS(KP951056516, T4O, T4L)); Rp[WS(rs, 1)] = KP500000000 * (FMA(KP951056516, T4O, T4L)); Rm[WS(rs, 2)] = KP500000000 * (FNMS(KP951056516, T4Q, T4P)); Rm[WS(rs, 6)] = KP500000000 * (FMA(KP951056516, T4Q, T4P)); T3K = T3u + T3J; T5O = T3J - T3u; } Im[WS(rs, 9)] = KP500000000 * (T3K - T3R); T5N = FMA(KP250000000, T3K, T3R); T5Z = T5f - T5g; T5h = T5f + T5g; T68 = FNMS(KP618033988, T67, T66); T6a = FMA(KP618033988, T66, T67); T5X = FNMS(KP559016994, T5O, T5N); T5P = FMA(KP559016994, T5O, T5N); Im[WS(rs, 5)] = -(KP500000000 * (FNMS(KP951056516, T5W, T5P))); Ip[WS(rs, 6)] = KP500000000 * (FMA(KP951056516, T5W, T5P)); Im[WS(rs, 1)] = -(KP500000000 * (FNMS(KP951056516, T5Y, T5X))); Ip[WS(rs, 2)] = KP500000000 * (FMA(KP951056516, T5Y, T5X)); T64 = T60 - T61; T62 = T60 + T61; } { E T5o, T5v, T5M, T5K, T5A, T5B, T3Z, T5G, T5I, T5J, T63, T5F, T5L, T5H; T5o = T5k + T5n; T5I = T5k - T5n; T5J = T5u - T5r; T5v = T5r + T5u; Rm[WS(rs, 9)] = KP500000000 * (T5Z + T62); T63 = FNMS(KP250000000, T62, T5Z); T5M = FMA(KP618033988, T5I, T5J); T5K = FNMS(KP618033988, T5J, T5I); { E T65, T69, T3V, T3Y; T65 = FNMS(KP559016994, T64, T63); T69 = FMA(KP559016994, T64, T63); T3V = T3T + T3U; T5A = T3T - T3U; T5B = T3W - T3X; T3Y = T3W + T3X; Rm[WS(rs, 1)] = KP500000000 * (FMA(KP951056516, T68, T65)); Rp[WS(rs, 2)] = KP500000000 * (FNMS(KP951056516, T68, T65)); Rm[WS(rs, 5)] = KP500000000 * (FNMS(KP951056516, T6a, T69)); Rp[WS(rs, 6)] = KP500000000 * (FMA(KP951056516, T6a, T69)); T3Z = T3V + T3Y; T5G = T3V - T3Y; } Ip[0] = KP500000000 * (T3S + T3Z); T5F = FNMS(KP250000000, T3Z, T3S); T5C = FMA(KP618033988, T5B, T5A); T5E = FNMS(KP618033988, T5A, T5B); T5L = FNMS(KP559016994, T5G, T5F); T5H = FMA(KP559016994, T5G, T5F); Im[WS(rs, 3)] = -(KP500000000 * (FNMS(KP951056516, T5K, T5H))); Ip[WS(rs, 4)] = KP500000000 * (FMA(KP951056516, T5K, T5H)); Im[WS(rs, 7)] = -(KP500000000 * (FNMS(KP951056516, T5M, T5L))); Ip[WS(rs, 8)] = KP500000000 * (FMA(KP951056516, T5M, T5L)); T5y = T5o - T5v; T5w = T5o + T5v; } } } } } Rp[0] = KP500000000 * (T5h + T5w); T5x = FNMS(KP250000000, T5w, T5h); T5D = FNMS(KP559016994, T5y, T5x); T5z = FMA(KP559016994, T5y, T5x); Rm[WS(rs, 3)] = KP500000000 * (FMA(KP951056516, T5C, T5z)); Rp[WS(rs, 4)] = KP500000000 * (FNMS(KP951056516, T5C, T5z)); Rm[WS(rs, 7)] = KP500000000 * (FNMS(KP951056516, T5E, T5D)); Rp[WS(rs, 8)] = KP500000000 * (FMA(KP951056516, T5E, T5D)); } } } static const tw_instr twinstr[] = { {TW_CEXP, 1, 1}, {TW_CEXP, 1, 3}, {TW_CEXP, 1, 9}, {TW_CEXP, 1, 19}, {TW_NEXT, 1, 0} }; static const hc2c_desc desc = { 20, "hc2cfdft2_20", twinstr, &GENUS, {176, 98, 140, 0} }; void X(codelet_hc2cfdft2_20) (planner *p) { X(khc2c_register) (p, hc2cfdft2_20, &desc, HC2C_VIA_DFT); } #else /* HAVE_FMA */ /* Generated by: ../../../genfft/gen_hc2cdft.native -compact -variables 4 -pipeline-latency 4 -twiddle-log3 -precompute-twiddles -n 20 -dit -name hc2cfdft2_20 -include hc2cf.h */ /* * This function contains 316 FP additions, 180 FP multiplications, * (or, 244 additions, 108 multiplications, 72 fused multiply/add), * 134 stack variables, 5 constants, and 80 memory accesses */ #include "hc2cf.h" static void hc2cfdft2_20(R *Rp, R *Ip, R *Rm, R *Im, const R *W, stride rs, INT mb, INT me, INT ms) { DK(KP125000000, +0.125000000000000000000000000000000000000000000); DK(KP500000000, +0.500000000000000000000000000000000000000000000); DK(KP279508497, +0.279508497187473712051146708591409529430077295); DK(KP293892626, +0.293892626146236564584352977319536384298826219); DK(KP475528258, +0.475528258147576786058219666689691071702849317); { INT m; for (m = mb, W = W + ((mb - 1) * 8); m < me; m = m + 1, Rp = Rp + ms, Ip = Ip + ms, Rm = Rm - ms, Im = Im - ms, W = W + 8, MAKE_VOLATILE_STRIDE(80, rs)) { E T4, T7, Tm, To, Tq, Tu, T1I, T1G, T8, T5, Ta, T1u, T2u, Tg, T2s; E T21, T1A, T1Z, T1O, T2I, T1K, T2G, Tw, TC, T2a, T2e, TH, TI, TJ, TX; E T2D, TN, T2B, T26, T1n, TZ, T24, T1j; { E T9, T1y, Te, T1t, T6, T1z, Tf, T1s; { E Tn, Tt, Tp, Ts; T4 = W[0]; T7 = W[1]; Tm = W[2]; To = W[3]; Tn = T4 * Tm; Tt = T7 * Tm; Tp = T7 * To; Ts = T4 * To; Tq = Tn - Tp; Tu = Ts + Tt; T1I = Ts - Tt; T1G = Tn + Tp; T8 = W[5]; T9 = T7 * T8; T1y = Tm * T8; Te = T4 * T8; T1t = To * T8; T5 = W[4]; T6 = T4 * T5; T1z = To * T5; Tf = T7 * T5; T1s = Tm * T5; } Ta = T6 - T9; T1u = T1s + T1t; T2u = T1y + T1z; Tg = Te + Tf; T2s = T1s - T1t; T21 = Te - Tf; T1A = T1y - T1z; T1Z = T6 + T9; { E T1M, T1N, T1H, T1J; T1M = T1G * T8; T1N = T1I * T5; T1O = T1M + T1N; T2I = T1M - T1N; T1H = T1G * T5; T1J = T1I * T8; T1K = T1H - T1J; T2G = T1H + T1J; { E Tr, Tv, TA, TB; Tr = Tq * T5; Tv = Tu * T8; Tw = Tr + Tv; TA = Tq * T8; TB = Tu * T5; TC = TA - TB; T2a = Tr - Tv; T2e = TA + TB; TH = W[6]; TI = W[7]; TJ = FMA(Tq, TH, Tu * TI); TX = FMA(Tw, TH, TC * TI); T2D = FMA(T1G, TH, T1I * TI); TN = FNMS(Tu, TH, Tq * TI); T2B = FNMS(T1I, TH, T1G * TI); T26 = FNMS(T7, TH, T4 * TI); T1n = FNMS(To, TH, Tm * TI); TZ = FNMS(TC, TH, Tw * TI); T24 = FMA(T4, TH, T7 * TI); T1j = FMA(Tm, TH, To * TI); } } } { E Tl, T3n, T1i, T2Q, T47, T50, T4S, T5i, T2M, T2T, T4I, T5f, T4L, T5e, T4P; E T5h, T2r, T2S, T1X, T2P, T31, T3u, T36, T3t, T3E, T4l, T3U, T4j, T3h, T3r; E T3J, T4m, T3c, T3q, T3P, T4i, TS, T51, T3m, T48; { E T3, T45, T1V, T3f, Tz, TF, TW, T3A, TM, TQ, T11, T3B, Td, Tj, T1Q; E T3e, T19, T3L, T23, T39, T2p, T3S, T2z, T34, T1E, T3G, T2K, T2Y, T1g, T3M; E T28, T3a, T2i, T3R, T2w, T33, T1r, T3F, T2F, T2X, T4N, T4O; { E T1, T2, T1R, T1S, T1T, T1U; T1 = Ip[0]; T2 = Im[0]; T1R = T1 + T2; T1S = Rp[0]; T1T = Rm[0]; T1U = T1S - T1T; T3 = T1 - T2; T45 = T1S + T1T; T1V = FNMS(T7, T1U, T4 * T1R); T3f = FMA(T4, T1U, T7 * T1R); } { E Tx, Ty, TU, TD, TE, TV; Tx = Ip[WS(rs, 2)]; Ty = Im[WS(rs, 2)]; TU = Tx - Ty; TD = Rp[WS(rs, 2)]; TE = Rm[WS(rs, 2)]; TV = TD + TE; Tz = Tx + Ty; TF = TD - TE; TW = FNMS(Tu, TV, Tq * TU); T3A = FMA(Tu, TU, Tq * TV); } { E TK, TL, TY, TO, TP, T10; TK = Ip[WS(rs, 7)]; TL = Im[WS(rs, 7)]; TY = TK - TL; TO = Rp[WS(rs, 7)]; TP = Rm[WS(rs, 7)]; T10 = TO + TP; TM = TK + TL; TQ = TO - TP; T11 = FNMS(TZ, T10, TX * TY); T3B = FMA(TZ, TY, TX * T10); } { E Tb, Tc, T1L, Th, Ti, T1P; Tb = Ip[WS(rs, 5)]; Tc = Im[WS(rs, 5)]; T1L = Tb + Tc; Th = Rp[WS(rs, 5)]; Ti = Rm[WS(rs, 5)]; T1P = Th - Ti; Td = Tb - Tc; Tj = Th + Ti; T1Q = FNMS(T1O, T1P, T1K * T1L); T3e = FMA(T1K, T1P, T1O * T1L); } { E T15, T20, T18, T22; { E T13, T14, T16, T17; T13 = Ip[WS(rs, 4)]; T14 = Im[WS(rs, 4)]; T15 = T13 + T14; T20 = T13 - T14; T16 = Rp[WS(rs, 4)]; T17 = Rm[WS(rs, 4)]; T18 = T16 - T17; T22 = T16 + T17; } T19 = FNMS(T8, T18, T5 * T15); T3L = FMA(T21, T20, T1Z * T22); T23 = FNMS(T21, T22, T1Z * T20); T39 = FMA(T8, T15, T5 * T18); } { E T2l, T2x, T2o, T2y; { E T2j, T2k, T2m, T2n; T2j = Ip[WS(rs, 1)]; T2k = Im[WS(rs, 1)]; T2l = T2j + T2k; T2x = T2j - T2k; T2m = Rp[WS(rs, 1)]; T2n = Rm[WS(rs, 1)]; T2o = T2m - T2n; T2y = T2m + T2n; } T2p = FNMS(To, T2o, Tm * T2l); T3S = FMA(T1I, T2x, T1G * T2y); T2z = FNMS(T1I, T2y, T1G * T2x); T34 = FMA(To, T2l, Tm * T2o); } { E T1x, T2H, T1D, T2J; { E T1v, T1w, T1B, T1C; T1v = Ip[WS(rs, 3)]; T1w = Im[WS(rs, 3)]; T1x = T1v - T1w; T2H = T1v + T1w; T1B = Rp[WS(rs, 3)]; T1C = Rm[WS(rs, 3)]; T1D = T1B + T1C; T2J = T1B - T1C; } T1E = FNMS(T1A, T1D, T1u * T1x); T3G = FMA(T1u, T1D, T1A * T1x); T2K = FNMS(T2I, T2J, T2G * T2H); T2Y = FMA(T2G, T2J, T2I * T2H); } { E T1c, T25, T1f, T27; { E T1a, T1b, T1d, T1e; T1a = Ip[WS(rs, 9)]; T1b = Im[WS(rs, 9)]; T1c = T1a + T1b; T25 = T1a - T1b; T1d = Rp[WS(rs, 9)]; T1e = Rm[WS(rs, 9)]; T1f = T1d - T1e; T27 = T1d + T1e; } T1g = FNMS(TI, T1f, TH * T1c); T3M = FMA(T26, T25, T24 * T27); T28 = FNMS(T26, T27, T24 * T25); T3a = FMA(TI, T1c, TH * T1f); } { E T2d, T2t, T2h, T2v; { E T2b, T2c, T2f, T2g; T2b = Ip[WS(rs, 6)]; T2c = Im[WS(rs, 6)]; T2d = T2b + T2c; T2t = T2b - T2c; T2f = Rp[WS(rs, 6)]; T2g = Rm[WS(rs, 6)]; T2h = T2f - T2g; T2v = T2f + T2g; } T2i = FNMS(T2e, T2h, T2a * T2d); T3R = FMA(T2u, T2t, T2s * T2v); T2w = FNMS(T2u, T2v, T2s * T2t); T33 = FMA(T2e, T2d, T2a * T2h); } { E T1m, T2E, T1q, T2C; { E T1k, T1l, T1o, T1p; T1k = Ip[WS(rs, 8)]; T1l = Im[WS(rs, 8)]; T1m = T1k - T1l; T2E = T1k + T1l; T1o = Rp[WS(rs, 8)]; T1p = Rm[WS(rs, 8)]; T1q = T1o + T1p; T2C = T1p - T1o; } T1r = FNMS(T1n, T1q, T1j * T1m); T3F = FMA(T1j, T1q, T1n * T1m); T2F = FMA(T2B, T2C, T2D * T2E); T2X = FNMS(T2B, T2E, T2D * T2C); } { E Tk, T12, T1h, T46; Tk = FNMS(Tg, Tj, Ta * Td); Tl = T3 - Tk; T3n = Tk + T3; T12 = TW - T11; T1h = T19 - T1g; T1i = T12 - T1h; T2Q = T12 + T1h; T46 = FMA(Ta, Tj, Tg * Td); T47 = T45 - T46; T50 = T45 + T46; { E T4Q, T4R, T2A, T2L; T4Q = T2F + T2K; T4R = T3R + T3S; T4S = T4Q + T4R; T5i = T4R - T4Q; T2A = T2w - T2z; T2L = T2F - T2K; T2M = T2A - T2L; T2T = T2L + T2A; } } { E T4G, T4H, T4J, T4K; T4G = T3A + T3B; T4H = T19 + T1g; T4I = T4G + T4H; T5f = T4G - T4H; T4J = T3F + T3G; T4K = T1Q + T1V; T4L = T4J + T4K; T5e = T4J - T4K; } T4N = T3L + T3M; T4O = T2i + T2p; T4P = T4N + T4O; T5h = T4N - T4O; { E T29, T2q, T1F, T1W; T29 = T23 - T28; T2q = T2i - T2p; T2r = T29 - T2q; T2S = T29 + T2q; T1F = T1r - T1E; T1W = T1Q - T1V; T1X = T1F + T1W; T2P = T1W - T1F; } { E T3C, T3D, T3N, T3O; { E T2Z, T30, T32, T35; T2Z = T2X - T2Y; T30 = T2w + T2z; T31 = T2Z - T30; T3u = T2Z + T30; T32 = T23 + T28; T35 = T33 + T34; T36 = T32 + T35; T3t = T32 - T35; } T3C = T3A - T3B; T3D = T3a - T39; T3E = T3C + T3D; T4l = T3C - T3D; { E T3Q, T3T, T3d, T3g; T3Q = T2X + T2Y; T3T = T3R - T3S; T3U = T3Q + T3T; T4j = T3T - T3Q; T3d = T1r + T1E; T3g = T3e + T3f; T3h = T3d + T3g; T3r = T3d - T3g; } { E T3H, T3I, T38, T3b; T3H = T3F - T3G; T3I = T3e - T3f; T3J = T3H + T3I; T4m = T3H - T3I; T38 = TW + T11; T3b = T39 + T3a; T3c = T38 + T3b; T3q = T38 - T3b; } T3N = T3L - T3M; T3O = T34 - T33; T3P = T3N + T3O; T4i = T3N - T3O; { E TG, TR, T3k, T3l; TG = FNMS(TC, TF, Tw * Tz); TR = FNMS(TN, TQ, TJ * TM); TS = TG - TR; T51 = TG + TR; T3k = FMA(TC, Tz, Tw * TF); T3l = FMA(TN, TM, TJ * TQ); T3m = T3k + T3l; T48 = T3l - T3k; } } } { E T3W, T3Y, TT, T2O, T3x, T3y, T3X, T3z; { E T3K, T3V, T1Y, T2N; T3K = T3E - T3J; T3V = T3P - T3U; T3W = FMA(KP475528258, T3K, KP293892626 * T3V); T3Y = FNMS(KP293892626, T3K, KP475528258 * T3V); TT = Tl - TS; T1Y = T1i + T1X; T2N = T2r + T2M; T2O = T1Y + T2N; T3x = KP279508497 * (T1Y - T2N); T3y = FNMS(KP125000000, T2O, KP500000000 * TT); } Ip[WS(rs, 5)] = KP500000000 * (TT + T2O); T3X = T3x - T3y; Im[WS(rs, 2)] = T3X - T3Y; Im[WS(rs, 6)] = T3X + T3Y; T3z = T3x + T3y; Ip[WS(rs, 1)] = T3z - T3W; Ip[WS(rs, 9)] = T3z + T3W; } { E T41, T4d, T49, T4a, T44, T4b, T4e, T4c; { E T3Z, T40, T42, T43; T3Z = T1i - T1X; T40 = T2r - T2M; T41 = FMA(KP475528258, T3Z, KP293892626 * T40); T4d = FNMS(KP293892626, T3Z, KP475528258 * T40); T49 = T47 + T48; T42 = T3E + T3J; T43 = T3P + T3U; T4a = T42 + T43; T44 = KP279508497 * (T42 - T43); T4b = FNMS(KP125000000, T4a, KP500000000 * T49); } Rp[WS(rs, 5)] = KP500000000 * (T49 + T4a); T4e = T4b - T44; Rm[WS(rs, 6)] = T4d + T4e; Rm[WS(rs, 2)] = T4e - T4d; T4c = T44 + T4b; Rp[WS(rs, 1)] = T41 + T4c; Rp[WS(rs, 9)] = T4c - T41; } { E T4o, T4q, T2W, T2V, T4f, T4g, T4p, T4h; { E T4k, T4n, T2R, T2U; T4k = T4i - T4j; T4n = T4l - T4m; T4o = FNMS(KP293892626, T4n, KP475528258 * T4k); T4q = FMA(KP475528258, T4n, KP293892626 * T4k); T2W = TS + Tl; T2R = T2P - T2Q; T2U = T2S + T2T; T2V = T2R - T2U; T4f = FMA(KP500000000, T2W, KP125000000 * T2V); T4g = KP279508497 * (T2R + T2U); } Im[WS(rs, 4)] = KP500000000 * (T2V - T2W); T4p = T4g - T4f; Im[0] = T4p - T4q; Im[WS(rs, 8)] = T4p + T4q; T4h = T4f + T4g; Ip[WS(rs, 3)] = T4h - T4o; Ip[WS(rs, 7)] = T4h + T4o; } { E T4t, T4B, T4u, T4x, T4y, T4z, T4C, T4A; { E T4r, T4s, T4v, T4w; T4r = T2S - T2T; T4s = T2Q + T2P; T4t = FNMS(KP293892626, T4s, KP475528258 * T4r); T4B = FMA(KP475528258, T4s, KP293892626 * T4r); T4u = T47 - T48; T4v = T4l + T4m; T4w = T4i + T4j; T4x = T4v + T4w; T4y = FNMS(KP125000000, T4x, KP500000000 * T4u); T4z = KP279508497 * (T4v - T4w); } Rm[WS(rs, 4)] = KP500000000 * (T4u + T4x); T4C = T4z + T4y; Rm[WS(rs, 8)] = T4B + T4C; Rm[0] = T4C - T4B; T4A = T4y - T4z; Rp[WS(rs, 3)] = T4t + T4A; Rp[WS(rs, 7)] = T4A - T4t; } { E T5k, T5m, T3o, T3j, T5b, T5c, T5l, T5d; { E T5g, T5j, T37, T3i; T5g = T5e - T5f; T5j = T5h - T5i; T5k = FNMS(KP293892626, T5j, KP475528258 * T5g); T5m = FMA(KP293892626, T5g, KP475528258 * T5j); T3o = T3m + T3n; T37 = T31 - T36; T3i = T3c + T3h; T3j = T37 - T3i; T5b = FMA(KP500000000, T3o, KP125000000 * T3j); T5c = KP279508497 * (T3i + T37); } Im[WS(rs, 9)] = KP500000000 * (T3j - T3o); T5l = T5b - T5c; Ip[WS(rs, 2)] = T5l + T5m; Im[WS(rs, 1)] = T5m - T5l; T5d = T5b + T5c; Ip[WS(rs, 6)] = T5d + T5k; Im[WS(rs, 5)] = T5k - T5d; } { E T5w, T5x, T5n, T5q, T5r, T5s, T5y, T5t; { E T5u, T5v, T5o, T5p; T5u = T36 + T31; T5v = T3c - T3h; T5w = FNMS(KP293892626, T5v, KP475528258 * T5u); T5x = FMA(KP475528258, T5v, KP293892626 * T5u); T5n = T50 - T51; T5o = T5f + T5e; T5p = T5h + T5i; T5q = T5o + T5p; T5r = FNMS(KP125000000, T5q, KP500000000 * T5n); T5s = KP279508497 * (T5o - T5p); } Rm[WS(rs, 9)] = KP500000000 * (T5n + T5q); T5y = T5s + T5r; Rp[WS(rs, 6)] = T5x + T5y; Rm[WS(rs, 5)] = T5y - T5x; T5t = T5r - T5s; Rp[WS(rs, 2)] = T5t - T5w; Rm[WS(rs, 1)] = T5w + T5t; } { E T4U, T4W, T3p, T3w, T4D, T4E, T4V, T4F; { E T4M, T4T, T3s, T3v; T4M = T4I - T4L; T4T = T4P - T4S; T4U = FNMS(KP475528258, T4T, KP293892626 * T4M); T4W = FMA(KP475528258, T4M, KP293892626 * T4T); T3p = T3n - T3m; T3s = T3q + T3r; T3v = T3t + T3u; T3w = T3s + T3v; T4D = FNMS(KP125000000, T3w, KP500000000 * T3p); T4E = KP279508497 * (T3s - T3v); } Ip[0] = KP500000000 * (T3p + T3w); T4V = T4E + T4D; Ip[WS(rs, 4)] = T4V + T4W; Im[WS(rs, 3)] = T4W - T4V; T4F = T4D - T4E; Ip[WS(rs, 8)] = T4F + T4U; Im[WS(rs, 7)] = T4U - T4F; } { E T58, T59, T52, T53, T4Z, T54, T5a, T55; { E T56, T57, T4X, T4Y; T56 = T3q - T3r; T57 = T3t - T3u; T58 = FMA(KP475528258, T56, KP293892626 * T57); T59 = FNMS(KP293892626, T56, KP475528258 * T57); T52 = T50 + T51; T4X = T4I + T4L; T4Y = T4P + T4S; T53 = T4X + T4Y; T4Z = KP279508497 * (T4X - T4Y); T54 = FNMS(KP125000000, T53, KP500000000 * T52); } Rp[0] = KP500000000 * (T52 + T53); T5a = T54 - T4Z; Rp[WS(rs, 8)] = T59 + T5a; Rm[WS(rs, 7)] = T5a - T59; T55 = T4Z + T54; Rp[WS(rs, 4)] = T55 - T58; Rm[WS(rs, 3)] = T58 + T55; } } } } } static const tw_instr twinstr[] = { {TW_CEXP, 1, 1}, {TW_CEXP, 1, 3}, {TW_CEXP, 1, 9}, {TW_CEXP, 1, 19}, {TW_NEXT, 1, 0} }; static const hc2c_desc desc = { 20, "hc2cfdft2_20", twinstr, &GENUS, {244, 108, 72, 0} }; void X(codelet_hc2cfdft2_20) (planner *p) { X(khc2c_register) (p, hc2cfdft2_20, &desc, HC2C_VIA_DFT); } #endif /* HAVE_FMA */ fftw-3.3.3/rdft/scalar/r2cf/codlist.c0000644000175000017500000001507612054410727014267 00000000000000#include "ifftw.h" extern void X(codelet_r2cf_2)(planner *); extern void X(codelet_r2cf_3)(planner *); extern void X(codelet_r2cf_4)(planner *); extern void X(codelet_r2cf_5)(planner *); extern void X(codelet_r2cf_6)(planner *); extern void X(codelet_r2cf_7)(planner *); extern void X(codelet_r2cf_8)(planner *); extern void X(codelet_r2cf_9)(planner *); extern void X(codelet_r2cf_10)(planner *); extern void X(codelet_r2cf_11)(planner *); extern void X(codelet_r2cf_12)(planner *); extern void X(codelet_r2cf_13)(planner *); extern void X(codelet_r2cf_14)(planner *); extern void X(codelet_r2cf_15)(planner *); extern void X(codelet_r2cf_16)(planner *); extern void X(codelet_r2cf_32)(planner *); extern void X(codelet_r2cf_64)(planner *); extern void X(codelet_r2cf_128)(planner *); extern void X(codelet_r2cf_20)(planner *); extern void X(codelet_r2cf_25)(planner *); extern void X(codelet_hf_2)(planner *); extern void X(codelet_hf_3)(planner *); extern void X(codelet_hf_4)(planner *); extern void X(codelet_hf_5)(planner *); extern void X(codelet_hf_6)(planner *); extern void X(codelet_hf_7)(planner *); extern void X(codelet_hf_8)(planner *); extern void X(codelet_hf_9)(planner *); extern void X(codelet_hf_10)(planner *); extern void X(codelet_hf_12)(planner *); extern void X(codelet_hf_15)(planner *); extern void X(codelet_hf_16)(planner *); extern void X(codelet_hf_32)(planner *); extern void X(codelet_hf_64)(planner *); extern void X(codelet_hf_20)(planner *); extern void X(codelet_hf_25)(planner *); extern void X(codelet_hf2_4)(planner *); extern void X(codelet_hf2_8)(planner *); extern void X(codelet_hf2_16)(planner *); extern void X(codelet_hf2_32)(planner *); extern void X(codelet_hf2_5)(planner *); extern void X(codelet_hf2_20)(planner *); extern void X(codelet_hf2_25)(planner *); extern void X(codelet_r2cfII_2)(planner *); extern void X(codelet_r2cfII_3)(planner *); extern void X(codelet_r2cfII_4)(planner *); extern void X(codelet_r2cfII_5)(planner *); extern void X(codelet_r2cfII_6)(planner *); extern void X(codelet_r2cfII_7)(planner *); extern void X(codelet_r2cfII_8)(planner *); extern void X(codelet_r2cfII_9)(planner *); extern void X(codelet_r2cfII_10)(planner *); extern void X(codelet_r2cfII_12)(planner *); extern void X(codelet_r2cfII_15)(planner *); extern void X(codelet_r2cfII_16)(planner *); extern void X(codelet_r2cfII_32)(planner *); extern void X(codelet_r2cfII_64)(planner *); extern void X(codelet_r2cfII_20)(planner *); extern void X(codelet_r2cfII_25)(planner *); extern void X(codelet_hc2cf_2)(planner *); extern void X(codelet_hc2cf_4)(planner *); extern void X(codelet_hc2cf_6)(planner *); extern void X(codelet_hc2cf_8)(planner *); extern void X(codelet_hc2cf_10)(planner *); extern void X(codelet_hc2cf_12)(planner *); extern void X(codelet_hc2cf_16)(planner *); extern void X(codelet_hc2cf_32)(planner *); extern void X(codelet_hc2cf_20)(planner *); extern void X(codelet_hc2cf2_4)(planner *); extern void X(codelet_hc2cf2_8)(planner *); extern void X(codelet_hc2cf2_16)(planner *); extern void X(codelet_hc2cf2_32)(planner *); extern void X(codelet_hc2cf2_20)(planner *); extern void X(codelet_hc2cfdft_2)(planner *); extern void X(codelet_hc2cfdft_4)(planner *); extern void X(codelet_hc2cfdft_6)(planner *); extern void X(codelet_hc2cfdft_8)(planner *); extern void X(codelet_hc2cfdft_10)(planner *); extern void X(codelet_hc2cfdft_12)(planner *); extern void X(codelet_hc2cfdft_16)(planner *); extern void X(codelet_hc2cfdft_32)(planner *); extern void X(codelet_hc2cfdft_20)(planner *); extern void X(codelet_hc2cfdft2_4)(planner *); extern void X(codelet_hc2cfdft2_8)(planner *); extern void X(codelet_hc2cfdft2_16)(planner *); extern void X(codelet_hc2cfdft2_32)(planner *); extern void X(codelet_hc2cfdft2_20)(planner *); extern const solvtab X(solvtab_rdft_r2cf); const solvtab X(solvtab_rdft_r2cf) = { SOLVTAB(X(codelet_r2cf_2)), SOLVTAB(X(codelet_r2cf_3)), SOLVTAB(X(codelet_r2cf_4)), SOLVTAB(X(codelet_r2cf_5)), SOLVTAB(X(codelet_r2cf_6)), SOLVTAB(X(codelet_r2cf_7)), SOLVTAB(X(codelet_r2cf_8)), SOLVTAB(X(codelet_r2cf_9)), SOLVTAB(X(codelet_r2cf_10)), SOLVTAB(X(codelet_r2cf_11)), SOLVTAB(X(codelet_r2cf_12)), SOLVTAB(X(codelet_r2cf_13)), SOLVTAB(X(codelet_r2cf_14)), SOLVTAB(X(codelet_r2cf_15)), SOLVTAB(X(codelet_r2cf_16)), SOLVTAB(X(codelet_r2cf_32)), SOLVTAB(X(codelet_r2cf_64)), SOLVTAB(X(codelet_r2cf_128)), SOLVTAB(X(codelet_r2cf_20)), SOLVTAB(X(codelet_r2cf_25)), SOLVTAB(X(codelet_hf_2)), SOLVTAB(X(codelet_hf_3)), SOLVTAB(X(codelet_hf_4)), SOLVTAB(X(codelet_hf_5)), SOLVTAB(X(codelet_hf_6)), SOLVTAB(X(codelet_hf_7)), SOLVTAB(X(codelet_hf_8)), SOLVTAB(X(codelet_hf_9)), SOLVTAB(X(codelet_hf_10)), SOLVTAB(X(codelet_hf_12)), SOLVTAB(X(codelet_hf_15)), SOLVTAB(X(codelet_hf_16)), SOLVTAB(X(codelet_hf_32)), SOLVTAB(X(codelet_hf_64)), SOLVTAB(X(codelet_hf_20)), SOLVTAB(X(codelet_hf_25)), SOLVTAB(X(codelet_hf2_4)), SOLVTAB(X(codelet_hf2_8)), SOLVTAB(X(codelet_hf2_16)), SOLVTAB(X(codelet_hf2_32)), SOLVTAB(X(codelet_hf2_5)), SOLVTAB(X(codelet_hf2_20)), SOLVTAB(X(codelet_hf2_25)), SOLVTAB(X(codelet_r2cfII_2)), SOLVTAB(X(codelet_r2cfII_3)), SOLVTAB(X(codelet_r2cfII_4)), SOLVTAB(X(codelet_r2cfII_5)), SOLVTAB(X(codelet_r2cfII_6)), SOLVTAB(X(codelet_r2cfII_7)), SOLVTAB(X(codelet_r2cfII_8)), SOLVTAB(X(codelet_r2cfII_9)), SOLVTAB(X(codelet_r2cfII_10)), SOLVTAB(X(codelet_r2cfII_12)), SOLVTAB(X(codelet_r2cfII_15)), SOLVTAB(X(codelet_r2cfII_16)), SOLVTAB(X(codelet_r2cfII_32)), SOLVTAB(X(codelet_r2cfII_64)), SOLVTAB(X(codelet_r2cfII_20)), SOLVTAB(X(codelet_r2cfII_25)), SOLVTAB(X(codelet_hc2cf_2)), SOLVTAB(X(codelet_hc2cf_4)), SOLVTAB(X(codelet_hc2cf_6)), SOLVTAB(X(codelet_hc2cf_8)), SOLVTAB(X(codelet_hc2cf_10)), SOLVTAB(X(codelet_hc2cf_12)), SOLVTAB(X(codelet_hc2cf_16)), SOLVTAB(X(codelet_hc2cf_32)), SOLVTAB(X(codelet_hc2cf_20)), SOLVTAB(X(codelet_hc2cf2_4)), SOLVTAB(X(codelet_hc2cf2_8)), SOLVTAB(X(codelet_hc2cf2_16)), SOLVTAB(X(codelet_hc2cf2_32)), SOLVTAB(X(codelet_hc2cf2_20)), SOLVTAB(X(codelet_hc2cfdft_2)), SOLVTAB(X(codelet_hc2cfdft_4)), SOLVTAB(X(codelet_hc2cfdft_6)), SOLVTAB(X(codelet_hc2cfdft_8)), SOLVTAB(X(codelet_hc2cfdft_10)), SOLVTAB(X(codelet_hc2cfdft_12)), SOLVTAB(X(codelet_hc2cfdft_16)), SOLVTAB(X(codelet_hc2cfdft_32)), SOLVTAB(X(codelet_hc2cfdft_20)), SOLVTAB(X(codelet_hc2cfdft2_4)), SOLVTAB(X(codelet_hc2cfdft2_8)), SOLVTAB(X(codelet_hc2cfdft2_16)), SOLVTAB(X(codelet_hc2cfdft2_32)), SOLVTAB(X(codelet_hc2cfdft2_20)), SOLVTAB_END }; fftw-3.3.3/rdft/scalar/r2cb/0000755000175000017500000000000012054411110012510 500000000000000fftw-3.3.3/rdft/scalar/r2cb/Makefile.am0000644000175000017500000001216412054410050014472 00000000000000# This Makefile.am specifies a set of codelets, efficient transforms # of small sizes, that are used as building blocks (kernels) by FFTW # to build up large transforms, as well as the options for generating # and compiling them. # You can customize FFTW for special needs, e.g. to handle certain # sizes more efficiently, by adding new codelets to the lists of those # included by default. If you change the list of codelets, any new # ones you added will be automatically generated when you run the # bootstrap script (see "Generating your own code" in the FFTW # manual). ########################################################################### AM_CPPFLAGS = -I$(top_srcdir)/kernel -I$(top_srcdir)/rdft \ -I$(top_srcdir)/rdft/scalar noinst_LTLIBRARIES = librdft_scalar_r2cb.la ########################################################################### # r2cb_ is a hard-coded complex-to-real FFT of size (base cases # of real-output FFT recursion) R2CB = r2cb_2.c r2cb_3.c r2cb_4.c r2cb_5.c r2cb_6.c r2cb_7.c r2cb_8.c \ r2cb_9.c r2cb_10.c r2cb_11.c r2cb_12.c r2cb_13.c r2cb_14.c r2cb_15.c \ r2cb_16.c r2cb_32.c r2cb_64.c r2cb_128.c r2cb_20.c r2cb_25.c # r2cb_30.c r2cb_40.c r2cb_50.c ########################################################################### # hb_ is a "twiddle" FFT of size , implementing a radix-r DIF # step for a real-output FFT. Every hb codelet must have a # corresponding r2cbIII codelet (see below)! HB = hb_2.c hb_3.c hb_4.c hb_5.c hb_6.c hb_7.c hb_8.c hb_9.c \ hb_10.c hb_12.c hb_15.c hb_16.c hb_32.c hb_64.c \ hb_20.c hb_25.c # hb_30.c hb_40.c hb_50.c # like hb, but generates part of its trig table on the fly (good for large n) HB2 = hb2_4.c hb2_8.c hb2_16.c hb2_32.c \ hb2_5.c hb2_20.c hb2_25.c # an r2cb transform where the output is shifted by half a sample (input # is multiplied by a phase). This is needed as part of the DIF recursion; # every hb_ or hb2_ codelet should have a corresponding r2cbIII_ R2CBIII = r2cbIII_2.c r2cbIII_3.c r2cbIII_4.c r2cbIII_5.c r2cbIII_6.c \ r2cbIII_7.c r2cbIII_8.c r2cbIII_9.c r2cbIII_10.c r2cbIII_12.c \ r2cbIII_15.c r2cbIII_16.c r2cbIII_32.c r2cbIII_64.c \ r2cbIII_20.c r2cbIII_25.c # r2cbIII_30.c r2cbIII_40.c r2cbIII_50.c ########################################################################### # hc2cb_ is a "twiddle" FFT of size , implementing a radix-r DIF # step for a real-input FFT with rdft2-style output. must be even. HC2CB = hc2cb_2.c hc2cb_4.c hc2cb_6.c hc2cb_8.c hc2cb_10.c hc2cb_12.c \ hc2cb_16.c hc2cb_32.c \ hc2cb_20.c # hc2cb_30.c HC2CBDFT = hc2cbdft_2.c hc2cbdft_4.c hc2cbdft_6.c hc2cbdft_8.c \ hc2cbdft_10.c hc2cbdft_12.c hc2cbdft_16.c hc2cbdft_32.c \ hc2cbdft_20.c # hc2cbdft_30.c # like hc2cb, but generates part of its trig table on the fly (good # for large n) HC2CB2 = hc2cb2_4.c hc2cb2_8.c hc2cb2_16.c hc2cb2_32.c \ hc2cb2_20.c # hc2cb2_30.c HC2CBDFT2 = hc2cbdft2_4.c hc2cbdft2_8.c hc2cbdft2_16.c hc2cbdft2_32.c \ hc2cbdft2_20.c # hc2cbdft2_30.c ########################################################################### ALL_CODELETS = $(R2CB) $(HB) $(HB2) $(R2CBIII) $(HC2CB) $(HC2CB2) \ $(HC2CBDFT) $(HC2CBDFT2) BUILT_SOURCES= $(ALL_CODELETS) $(CODLIST) librdft_scalar_r2cb_la_SOURCES = $(BUILT_SOURCES) SOLVTAB_NAME = X(solvtab_rdft_r2cb) XRENAME=X # special rules for regenerating codelets. include $(top_srcdir)/support/Makefile.codelets if MAINTAINER_MODE FLAGS_R2CB=$(RDFT_FLAGS_COMMON) -sign 1 FLAGS_HB=$(RDFT_FLAGS_COMMON) -sign 1 FLAGS_HB2=$(RDFT_FLAGS_COMMON) -sign 1 -twiddle-log3 -precompute-twiddles FLAGS_HC2CB=$(RDFT_FLAGS_COMMON) -sign 1 FLAGS_HC2CB2=$(RDFT_FLAGS_COMMON) -sign 1 -twiddle-log3 -precompute-twiddles FLAGS_R2CBIII=$(RDFT_FLAGS_COMMON) -sign 1 r2cb_%.c: $(CODELET_DEPS) $(GEN_R2CB) ($(PRELUDE_COMMANDS_RDFT); $(TWOVERS) $(GEN_R2CB) $(FLAGS_R2CB) -n $* -name r2cb_$* -include "r2cb.h") | $(ADD_DATE) | $(INDENT) >$@ hb_%.c: $(CODELET_DEPS) $(GEN_HC2HC) ($(PRELUDE_COMMANDS_RDFT); $(TWOVERS) $(GEN_HC2HC) $(FLAGS_HB) -n $* -dif -name hb_$* -include "hb.h") | $(ADD_DATE) | $(INDENT) >$@ hb2_%.c: $(CODELET_DEPS) $(GEN_HC2HC) ($(PRELUDE_COMMANDS_RDFT); $(TWOVERS) $(GEN_HC2HC) $(FLAGS_HB2) -n $* -dif -name hb2_$* -include "hb.h") | $(ADD_DATE) | $(INDENT) >$@ r2cbIII_%.c: $(CODELET_DEPS) $(GEN_R2CB) ($(PRELUDE_COMMANDS_RDFT); $(TWOVERS) $(GEN_R2CB) $(FLAGS_R2CB) -n $* -name r2cbIII_$* -dft-III -include "r2cbIII.h") | $(ADD_DATE) | $(INDENT) >$@ hc2cb_%.c: $(CODELET_DEPS) $(GEN_HC2C) ($(PRELUDE_COMMANDS_RDFT); $(TWOVERS) $(GEN_HC2C) $(FLAGS_HC2CB) -n $* -dif -name hc2cb_$* -include "hc2cb.h") | $(ADD_DATE) | $(INDENT) >$@ hc2cb2_%.c: $(CODELET_DEPS) $(GEN_HC2C) ($(PRELUDE_COMMANDS_RDFT); $(TWOVERS) $(GEN_HC2C) $(FLAGS_HC2CB2) -n $* -dif -name hc2cb2_$* -include "hc2cb.h") | $(ADD_DATE) | $(INDENT) >$@ hc2cbdft_%.c: $(CODELET_DEPS) $(GEN_HC2CDFT) ($(PRELUDE_COMMANDS_RDFT); $(TWOVERS) $(GEN_HC2CDFT) $(FLAGS_HC2CB) -n $* -dif -name hc2cbdft_$* -include "hc2cb.h") | $(ADD_DATE) | $(INDENT) >$@ hc2cbdft2_%.c: $(CODELET_DEPS) $(GEN_HC2CDFT) ($(PRELUDE_COMMANDS_RDFT); $(TWOVERS) $(GEN_HC2CDFT) $(FLAGS_HC2CB) -n $* -dif -name hc2cbdft2_$* -include "hc2cb.h") | $(ADD_DATE) | $(INDENT) >$@ endif # MAINTAINER_MODE fftw-3.3.3/rdft/scalar/r2cb/Makefile.in0000644000175000017500000010041312054410200014473 00000000000000# Makefile.in generated by automake 1.11.6 from Makefile.am. # @configure_input@ # Copyright (C) 1994, 1995, 1996, 1997, 1998, 1999, 2000, 2001, 2002, # 2003, 2004, 2005, 2006, 2007, 2008, 2009, 2010, 2011 Free Software # Foundation, Inc. # This Makefile.in is free software; the Free Software Foundation # gives unlimited permission to copy and/or distribute it, # with or without modifications, as long as this notice is preserved. # This program is distributed in the hope that it will be useful, # but WITHOUT ANY WARRANTY, to the extent permitted by law; without # even the implied warranty of MERCHANTABILITY or FITNESS FOR A # PARTICULAR PURPOSE. @SET_MAKE@ # This Makefile.am specifies a set of codelets, efficient transforms # of small sizes, that are used as building blocks (kernels) by FFTW # to build up large transforms, as well as the options for generating # and compiling them. # You can customize FFTW for special needs, e.g. to handle certain # sizes more efficiently, by adding new codelets to the lists of those # included by default. If you change the list of codelets, any new # ones you added will be automatically generated when you run the # bootstrap script (see "Generating your own code" in the FFTW # manual). # -*- makefile -*- # This file contains special make rules to generate codelets. # Most of this file requires GNU make . VPATH = @srcdir@ am__make_dryrun = \ { \ am__dry=no; \ case $$MAKEFLAGS in \ *\\[\ \ ]*) \ echo 'am--echo: ; @echo "AM" OK' | $(MAKE) -f - 2>/dev/null \ | grep '^AM OK$$' >/dev/null || am__dry=yes;; \ *) \ for am__flg in $$MAKEFLAGS; do \ case $$am__flg in \ *=*|--*) ;; \ *n*) am__dry=yes; break;; \ esac; \ done;; \ esac; \ test $$am__dry = yes; \ } pkgdatadir = $(datadir)/@PACKAGE@ pkgincludedir = $(includedir)/@PACKAGE@ pkglibdir = $(libdir)/@PACKAGE@ pkglibexecdir = $(libexecdir)/@PACKAGE@ am__cd = CDPATH="$${ZSH_VERSION+.}$(PATH_SEPARATOR)" && cd install_sh_DATA = $(install_sh) -c -m 644 install_sh_PROGRAM = $(install_sh) -c install_sh_SCRIPT = $(install_sh) -c INSTALL_HEADER = $(INSTALL_DATA) transform = $(program_transform_name) NORMAL_INSTALL = : PRE_INSTALL = : POST_INSTALL = : NORMAL_UNINSTALL = : PRE_UNINSTALL = : POST_UNINSTALL = : build_triplet = @build@ host_triplet = @host@ DIST_COMMON = $(srcdir)/Makefile.am $(srcdir)/Makefile.in \ $(top_srcdir)/support/Makefile.codelets subdir = rdft/scalar/r2cb ACLOCAL_M4 = $(top_srcdir)/aclocal.m4 am__aclocal_m4_deps = $(top_srcdir)/m4/acx_mpi.m4 \ $(top_srcdir)/m4/acx_pthread.m4 \ $(top_srcdir)/m4/ax_cc_maxopt.m4 \ $(top_srcdir)/m4/ax_check_compiler_flags.m4 \ $(top_srcdir)/m4/ax_compiler_vendor.m4 \ $(top_srcdir)/m4/ax_gcc_aligns_stack.m4 \ $(top_srcdir)/m4/ax_gcc_version.m4 \ $(top_srcdir)/m4/ax_openmp.m4 $(top_srcdir)/m4/libtool.m4 \ $(top_srcdir)/m4/ltoptions.m4 $(top_srcdir)/m4/ltsugar.m4 \ $(top_srcdir)/m4/ltversion.m4 $(top_srcdir)/m4/lt~obsolete.m4 \ $(top_srcdir)/configure.ac am__configure_deps = $(am__aclocal_m4_deps) $(CONFIGURE_DEPENDENCIES) \ $(ACLOCAL_M4) mkinstalldirs = $(install_sh) -d CONFIG_HEADER = $(top_builddir)/config.h CONFIG_CLEAN_FILES = CONFIG_CLEAN_VPATH_FILES = LTLIBRARIES = $(noinst_LTLIBRARIES) librdft_scalar_r2cb_la_LIBADD = am__objects_1 = r2cb_2.lo r2cb_3.lo r2cb_4.lo r2cb_5.lo r2cb_6.lo \ r2cb_7.lo r2cb_8.lo r2cb_9.lo r2cb_10.lo r2cb_11.lo r2cb_12.lo \ r2cb_13.lo r2cb_14.lo r2cb_15.lo r2cb_16.lo r2cb_32.lo \ r2cb_64.lo r2cb_128.lo r2cb_20.lo r2cb_25.lo am__objects_2 = hb_2.lo hb_3.lo hb_4.lo hb_5.lo hb_6.lo hb_7.lo \ hb_8.lo hb_9.lo hb_10.lo hb_12.lo hb_15.lo hb_16.lo hb_32.lo \ hb_64.lo hb_20.lo hb_25.lo am__objects_3 = hb2_4.lo hb2_8.lo hb2_16.lo hb2_32.lo hb2_5.lo \ hb2_20.lo hb2_25.lo am__objects_4 = r2cbIII_2.lo r2cbIII_3.lo r2cbIII_4.lo r2cbIII_5.lo \ r2cbIII_6.lo r2cbIII_7.lo r2cbIII_8.lo r2cbIII_9.lo \ r2cbIII_10.lo r2cbIII_12.lo r2cbIII_15.lo r2cbIII_16.lo \ r2cbIII_32.lo r2cbIII_64.lo r2cbIII_20.lo r2cbIII_25.lo am__objects_5 = hc2cb_2.lo hc2cb_4.lo hc2cb_6.lo hc2cb_8.lo \ hc2cb_10.lo hc2cb_12.lo hc2cb_16.lo hc2cb_32.lo hc2cb_20.lo am__objects_6 = hc2cb2_4.lo hc2cb2_8.lo hc2cb2_16.lo hc2cb2_32.lo \ hc2cb2_20.lo am__objects_7 = hc2cbdft_2.lo hc2cbdft_4.lo hc2cbdft_6.lo \ hc2cbdft_8.lo hc2cbdft_10.lo hc2cbdft_12.lo hc2cbdft_16.lo \ hc2cbdft_32.lo hc2cbdft_20.lo am__objects_8 = hc2cbdft2_4.lo hc2cbdft2_8.lo hc2cbdft2_16.lo \ hc2cbdft2_32.lo hc2cbdft2_20.lo am__objects_9 = $(am__objects_1) $(am__objects_2) $(am__objects_3) \ $(am__objects_4) $(am__objects_5) $(am__objects_6) \ $(am__objects_7) $(am__objects_8) am__objects_10 = codlist.lo am__objects_11 = $(am__objects_9) $(am__objects_10) am_librdft_scalar_r2cb_la_OBJECTS = $(am__objects_11) librdft_scalar_r2cb_la_OBJECTS = $(am_librdft_scalar_r2cb_la_OBJECTS) DEFAULT_INCLUDES = -I.@am__isrc@ -I$(top_builddir) depcomp = $(SHELL) $(top_srcdir)/depcomp am__depfiles_maybe = depfiles am__mv = mv -f COMPILE = $(CC) $(DEFS) $(DEFAULT_INCLUDES) $(INCLUDES) $(AM_CPPFLAGS) \ $(CPPFLAGS) $(AM_CFLAGS) $(CFLAGS) LTCOMPILE = $(LIBTOOL) --tag=CC $(AM_LIBTOOLFLAGS) $(LIBTOOLFLAGS) \ --mode=compile $(CC) $(DEFS) $(DEFAULT_INCLUDES) $(INCLUDES) \ $(AM_CPPFLAGS) $(CPPFLAGS) $(AM_CFLAGS) $(CFLAGS) CCLD = $(CC) LINK = $(LIBTOOL) --tag=CC $(AM_LIBTOOLFLAGS) $(LIBTOOLFLAGS) \ --mode=link $(CCLD) $(AM_CFLAGS) $(CFLAGS) $(AM_LDFLAGS) \ $(LDFLAGS) -o $@ SOURCES = $(librdft_scalar_r2cb_la_SOURCES) DIST_SOURCES = $(librdft_scalar_r2cb_la_SOURCES) am__can_run_installinfo = \ case $$AM_UPDATE_INFO_DIR in \ n|no|NO) false;; \ *) (install-info --version) >/dev/null 2>&1;; \ esac ETAGS = etags CTAGS = ctags DISTFILES = $(DIST_COMMON) $(DIST_SOURCES) $(TEXINFOS) $(EXTRA_DIST) ACLOCAL = @ACLOCAL@ ALLOCA = @ALLOCA@ ALTIVEC_CFLAGS = @ALTIVEC_CFLAGS@ AMTAR = @AMTAR@ AR = @AR@ AS = @AS@ AUTOCONF = @AUTOCONF@ AUTOHEADER = @AUTOHEADER@ AUTOMAKE = @AUTOMAKE@ AVX_CFLAGS = @AVX_CFLAGS@ AWK = @AWK@ CC = @CC@ CCDEPMODE = @CCDEPMODE@ CFLAGS = @CFLAGS@ CHECK_PL_OPTS = @CHECK_PL_OPTS@ CPP = @CPP@ CPPFLAGS = @CPPFLAGS@ CYGPATH_W = @CYGPATH_W@ C_FFTW_R2R_KIND = @C_FFTW_R2R_KIND@ C_MPI_FINT = @C_MPI_FINT@ DEFS = @DEFS@ DEPDIR = @DEPDIR@ DLLTOOL = @DLLTOOL@ DSYMUTIL = @DSYMUTIL@ DUMPBIN = @DUMPBIN@ ECHO_C = @ECHO_C@ ECHO_N = @ECHO_N@ ECHO_T = @ECHO_T@ EGREP = @EGREP@ EXEEXT = @EXEEXT@ F77 = @F77@ FFLAGS = @FFLAGS@ FGREP = @FGREP@ FLIBS = @FLIBS@ GREP = @GREP@ INSTALL = @INSTALL@ INSTALL_DATA = @INSTALL_DATA@ INSTALL_PROGRAM = @INSTALL_PROGRAM@ INSTALL_SCRIPT = @INSTALL_SCRIPT@ INSTALL_STRIP_PROGRAM = @INSTALL_STRIP_PROGRAM@ LD = @LD@ LDFLAGS = @LDFLAGS@ LIBOBJS = @LIBOBJS@ LIBQUADMATH = @LIBQUADMATH@ LIBS = @LIBS@ LIBTOOL = @LIBTOOL@ LIPO = @LIPO@ LN_S = @LN_S@ LTLIBOBJS = @LTLIBOBJS@ MAINT = @MAINT@ MAKEINFO = @MAKEINFO@ MANIFEST_TOOL = @MANIFEST_TOOL@ MKDIR_P = @MKDIR_P@ MPICC = @MPICC@ MPILIBS = @MPILIBS@ MPIRUN = @MPIRUN@ NEON_CFLAGS = @NEON_CFLAGS@ NM = @NM@ NMEDIT = @NMEDIT@ OBJDUMP = @OBJDUMP@ OBJEXT = @OBJEXT@ OCAMLBUILD = @OCAMLBUILD@ OPENMP_CFLAGS = @OPENMP_CFLAGS@ OTOOL = @OTOOL@ OTOOL64 = @OTOOL64@ PACKAGE = @PACKAGE@ PACKAGE_BUGREPORT = @PACKAGE_BUGREPORT@ PACKAGE_NAME = @PACKAGE_NAME@ PACKAGE_STRING = @PACKAGE_STRING@ PACKAGE_TARNAME = @PACKAGE_TARNAME@ PACKAGE_URL = @PACKAGE_URL@ PACKAGE_VERSION = @PACKAGE_VERSION@ PATH_SEPARATOR = @PATH_SEPARATOR@ POW_LIB = @POW_LIB@ PRECISION = @PRECISION@ PREC_SUFFIX = @PREC_SUFFIX@ PTHREAD_CC = @PTHREAD_CC@ PTHREAD_CFLAGS = @PTHREAD_CFLAGS@ PTHREAD_LIBS = @PTHREAD_LIBS@ RANLIB = @RANLIB@ SED = @SED@ SET_MAKE = @SET_MAKE@ SHARED_VERSION_INFO = @SHARED_VERSION_INFO@ SHELL = @SHELL@ SSE2_CFLAGS = @SSE2_CFLAGS@ STACK_ALIGN_CFLAGS = @STACK_ALIGN_CFLAGS@ STRIP = @STRIP@ THREADLIBS = @THREADLIBS@ VERSION = @VERSION@ abs_builddir = @abs_builddir@ abs_srcdir = @abs_srcdir@ abs_top_builddir = @abs_top_builddir@ abs_top_srcdir = @abs_top_srcdir@ ac_ct_AR = @ac_ct_AR@ ac_ct_CC = @ac_ct_CC@ ac_ct_DUMPBIN = @ac_ct_DUMPBIN@ ac_ct_F77 = @ac_ct_F77@ acx_pthread_config = @acx_pthread_config@ am__include = @am__include@ am__leading_dot = @am__leading_dot@ am__quote = @am__quote@ am__tar = @am__tar@ am__untar = @am__untar@ bindir = @bindir@ build = @build@ build_alias = @build_alias@ build_cpu = @build_cpu@ build_os = @build_os@ build_vendor = @build_vendor@ builddir = @builddir@ datadir = @datadir@ datarootdir = @datarootdir@ docdir = @docdir@ dvidir = @dvidir@ exec_prefix = @exec_prefix@ host = @host@ host_alias = @host_alias@ host_cpu = @host_cpu@ host_os = @host_os@ host_vendor = @host_vendor@ htmldir = @htmldir@ includedir = @includedir@ infodir = @infodir@ install_sh = @install_sh@ libdir = @libdir@ libexecdir = @libexecdir@ localedir = @localedir@ localstatedir = @localstatedir@ mandir = @mandir@ mkdir_p = @mkdir_p@ oldincludedir = @oldincludedir@ pdfdir = @pdfdir@ prefix = @prefix@ program_transform_name = @program_transform_name@ psdir = @psdir@ sbindir = @sbindir@ sharedstatedir = @sharedstatedir@ srcdir = @srcdir@ sysconfdir = @sysconfdir@ target_alias = @target_alias@ top_build_prefix = @top_build_prefix@ top_builddir = @top_builddir@ top_srcdir = @top_srcdir@ ########################################################################### AM_CPPFLAGS = -I$(top_srcdir)/kernel -I$(top_srcdir)/rdft \ -I$(top_srcdir)/rdft/scalar noinst_LTLIBRARIES = librdft_scalar_r2cb.la ########################################################################### # r2cb_ is a hard-coded complex-to-real FFT of size (base cases # of real-output FFT recursion) R2CB = r2cb_2.c r2cb_3.c r2cb_4.c r2cb_5.c r2cb_6.c r2cb_7.c r2cb_8.c \ r2cb_9.c r2cb_10.c r2cb_11.c r2cb_12.c r2cb_13.c r2cb_14.c r2cb_15.c \ r2cb_16.c r2cb_32.c r2cb_64.c r2cb_128.c r2cb_20.c r2cb_25.c # r2cb_30.c r2cb_40.c r2cb_50.c ########################################################################### # hb_ is a "twiddle" FFT of size , implementing a radix-r DIF # step for a real-output FFT. Every hb codelet must have a # corresponding r2cbIII codelet (see below)! HB = hb_2.c hb_3.c hb_4.c hb_5.c hb_6.c hb_7.c hb_8.c hb_9.c \ hb_10.c hb_12.c hb_15.c hb_16.c hb_32.c hb_64.c \ hb_20.c hb_25.c # hb_30.c hb_40.c hb_50.c # like hb, but generates part of its trig table on the fly (good for large n) HB2 = hb2_4.c hb2_8.c hb2_16.c hb2_32.c \ hb2_5.c hb2_20.c hb2_25.c # an r2cb transform where the output is shifted by half a sample (input # is multiplied by a phase). This is needed as part of the DIF recursion; # every hb_ or hb2_ codelet should have a corresponding r2cbIII_ R2CBIII = r2cbIII_2.c r2cbIII_3.c r2cbIII_4.c r2cbIII_5.c r2cbIII_6.c \ r2cbIII_7.c r2cbIII_8.c r2cbIII_9.c r2cbIII_10.c r2cbIII_12.c \ r2cbIII_15.c r2cbIII_16.c r2cbIII_32.c r2cbIII_64.c \ r2cbIII_20.c r2cbIII_25.c # r2cbIII_30.c r2cbIII_40.c r2cbIII_50.c ########################################################################### # hc2cb_ is a "twiddle" FFT of size , implementing a radix-r DIF # step for a real-input FFT with rdft2-style output. must be even. HC2CB = hc2cb_2.c hc2cb_4.c hc2cb_6.c hc2cb_8.c hc2cb_10.c hc2cb_12.c \ hc2cb_16.c hc2cb_32.c \ hc2cb_20.c # hc2cb_30.c HC2CBDFT = hc2cbdft_2.c hc2cbdft_4.c hc2cbdft_6.c hc2cbdft_8.c \ hc2cbdft_10.c hc2cbdft_12.c hc2cbdft_16.c hc2cbdft_32.c \ hc2cbdft_20.c # hc2cbdft_30.c # like hc2cb, but generates part of its trig table on the fly (good # for large n) HC2CB2 = hc2cb2_4.c hc2cb2_8.c hc2cb2_16.c hc2cb2_32.c \ hc2cb2_20.c # hc2cb2_30.c HC2CBDFT2 = hc2cbdft2_4.c hc2cbdft2_8.c hc2cbdft2_16.c hc2cbdft2_32.c \ hc2cbdft2_20.c # hc2cbdft2_30.c ########################################################################### ALL_CODELETS = $(R2CB) $(HB) $(HB2) $(R2CBIII) $(HC2CB) $(HC2CB2) \ $(HC2CBDFT) $(HC2CBDFT2) BUILT_SOURCES = $(ALL_CODELETS) $(CODLIST) librdft_scalar_r2cb_la_SOURCES = $(BUILT_SOURCES) SOLVTAB_NAME = X(solvtab_rdft_r2cb) XRENAME = X CODLIST = codlist.c CODELET_NAME = codelet_ @MAINTAINER_MODE_TRUE@INDENT = indent -kr -cs -i5 -l800 -fca -nfc1 -sc -sob -cli4 -TR -Tplanner -TV @MAINTAINER_MODE_TRUE@TWOVERS = sh ${top_srcdir}/support/twovers.sh @MAINTAINER_MODE_TRUE@GENFFTDIR = ${top_builddir}/genfft @MAINTAINER_MODE_TRUE@GEN_NOTW = ${GENFFTDIR}/gen_notw.native @MAINTAINER_MODE_TRUE@GEN_NOTW_C = ${GENFFTDIR}/gen_notw_c.native @MAINTAINER_MODE_TRUE@GEN_TWIDDLE = ${GENFFTDIR}/gen_twiddle.native @MAINTAINER_MODE_TRUE@GEN_TWIDDLE_C = ${GENFFTDIR}/gen_twiddle_c.native @MAINTAINER_MODE_TRUE@GEN_TWIDSQ = ${GENFFTDIR}/gen_twidsq.native @MAINTAINER_MODE_TRUE@GEN_TWIDSQ_C = ${GENFFTDIR}/gen_twidsq_c.native @MAINTAINER_MODE_TRUE@GEN_R2CF = ${GENFFTDIR}/gen_r2cf.native @MAINTAINER_MODE_TRUE@GEN_R2CB = ${GENFFTDIR}/gen_r2cb.native @MAINTAINER_MODE_TRUE@GEN_HC2HC = ${GENFFTDIR}/gen_hc2hc.native @MAINTAINER_MODE_TRUE@GEN_HC2C = ${GENFFTDIR}/gen_hc2c.native @MAINTAINER_MODE_TRUE@GEN_HC2CDFT = ${GENFFTDIR}/gen_hc2cdft.native @MAINTAINER_MODE_TRUE@GEN_HC2CDFT_C = ${GENFFTDIR}/gen_hc2cdft_c.native @MAINTAINER_MODE_TRUE@GEN_R2R = ${GENFFTDIR}/gen_r2r.native @MAINTAINER_MODE_TRUE@PRELUDE_DFT = ${top_srcdir}/support/codelet_prelude.dft @MAINTAINER_MODE_TRUE@PRELUDE_RDFT = ${top_srcdir}/support/codelet_prelude.rdft @MAINTAINER_MODE_TRUE@ADD_DATE = sed -e s/@DATE@/"`date`"/ @MAINTAINER_MODE_TRUE@COPYRIGHT = ${top_srcdir}/COPYRIGHT @MAINTAINER_MODE_TRUE@CODELET_DEPS = $(COPYRIGHT) $(PRELUDE) @MAINTAINER_MODE_TRUE@PRELUDE_COMMANDS_DFT = cat $(COPYRIGHT) $(PRELUDE_DFT) @MAINTAINER_MODE_TRUE@PRELUDE_COMMANDS_RDFT = cat $(COPYRIGHT) $(PRELUDE_RDFT) @MAINTAINER_MODE_TRUE@FLAGS_COMMON = -compact -variables 4 @MAINTAINER_MODE_TRUE@DFT_FLAGS_COMMON = $(FLAGS_COMMON) -pipeline-latency 4 @MAINTAINER_MODE_TRUE@RDFT_FLAGS_COMMON = $(FLAGS_COMMON) -pipeline-latency 4 # special rules for regenerating codelets. @MAINTAINER_MODE_TRUE@FLAGS_R2CB = $(RDFT_FLAGS_COMMON) -sign 1 @MAINTAINER_MODE_TRUE@FLAGS_HB = $(RDFT_FLAGS_COMMON) -sign 1 @MAINTAINER_MODE_TRUE@FLAGS_HB2 = $(RDFT_FLAGS_COMMON) -sign 1 -twiddle-log3 -precompute-twiddles @MAINTAINER_MODE_TRUE@FLAGS_HC2CB = $(RDFT_FLAGS_COMMON) -sign 1 @MAINTAINER_MODE_TRUE@FLAGS_HC2CB2 = $(RDFT_FLAGS_COMMON) -sign 1 -twiddle-log3 -precompute-twiddles @MAINTAINER_MODE_TRUE@FLAGS_R2CBIII = $(RDFT_FLAGS_COMMON) -sign 1 all: $(BUILT_SOURCES) $(MAKE) $(AM_MAKEFLAGS) all-am .SUFFIXES: .SUFFIXES: .c .lo .o .obj $(srcdir)/Makefile.in: @MAINTAINER_MODE_TRUE@ $(srcdir)/Makefile.am $(top_srcdir)/support/Makefile.codelets $(am__configure_deps) @for dep in $?; do \ case '$(am__configure_deps)' in \ *$$dep*) \ ( cd $(top_builddir) && $(MAKE) $(AM_MAKEFLAGS) am--refresh ) \ && { if test -f $@; then exit 0; else break; fi; }; \ exit 1;; \ esac; \ done; \ echo ' cd $(top_srcdir) && $(AUTOMAKE) --gnu rdft/scalar/r2cb/Makefile'; \ $(am__cd) $(top_srcdir) && \ $(AUTOMAKE) --gnu rdft/scalar/r2cb/Makefile .PRECIOUS: Makefile Makefile: $(srcdir)/Makefile.in $(top_builddir)/config.status @case '$?' in \ *config.status*) \ cd $(top_builddir) && $(MAKE) $(AM_MAKEFLAGS) am--refresh;; \ *) \ echo ' cd $(top_builddir) && $(SHELL) ./config.status $(subdir)/$@ $(am__depfiles_maybe)'; \ cd $(top_builddir) && $(SHELL) ./config.status $(subdir)/$@ $(am__depfiles_maybe);; \ esac; $(top_srcdir)/support/Makefile.codelets: $(top_builddir)/config.status: $(top_srcdir)/configure $(CONFIG_STATUS_DEPENDENCIES) cd $(top_builddir) && $(MAKE) $(AM_MAKEFLAGS) am--refresh $(top_srcdir)/configure: @MAINTAINER_MODE_TRUE@ $(am__configure_deps) cd $(top_builddir) && $(MAKE) $(AM_MAKEFLAGS) am--refresh $(ACLOCAL_M4): @MAINTAINER_MODE_TRUE@ $(am__aclocal_m4_deps) cd $(top_builddir) && $(MAKE) $(AM_MAKEFLAGS) am--refresh $(am__aclocal_m4_deps): clean-noinstLTLIBRARIES: -test -z "$(noinst_LTLIBRARIES)" || rm -f $(noinst_LTLIBRARIES) @list='$(noinst_LTLIBRARIES)'; for p in $$list; do \ dir="`echo $$p | sed -e 's|/[^/]*$$||'`"; \ test "$$dir" != "$$p" || dir=.; \ echo "rm -f \"$${dir}/so_locations\""; \ rm -f "$${dir}/so_locations"; \ done librdft_scalar_r2cb.la: $(librdft_scalar_r2cb_la_OBJECTS) $(librdft_scalar_r2cb_la_DEPENDENCIES) $(EXTRA_librdft_scalar_r2cb_la_DEPENDENCIES) $(LINK) $(librdft_scalar_r2cb_la_OBJECTS) $(librdft_scalar_r2cb_la_LIBADD) $(LIBS) mostlyclean-compile: -rm -f *.$(OBJEXT) distclean-compile: -rm -f *.tab.c @AMDEP_TRUE@@am__include@ @am__quote@./$(DEPDIR)/codlist.Plo@am__quote@ @AMDEP_TRUE@@am__include@ @am__quote@./$(DEPDIR)/hb2_16.Plo@am__quote@ @AMDEP_TRUE@@am__include@ @am__quote@./$(DEPDIR)/hb2_20.Plo@am__quote@ @AMDEP_TRUE@@am__include@ @am__quote@./$(DEPDIR)/hb2_25.Plo@am__quote@ @AMDEP_TRUE@@am__include@ @am__quote@./$(DEPDIR)/hb2_32.Plo@am__quote@ @AMDEP_TRUE@@am__include@ @am__quote@./$(DEPDIR)/hb2_4.Plo@am__quote@ @AMDEP_TRUE@@am__include@ @am__quote@./$(DEPDIR)/hb2_5.Plo@am__quote@ @AMDEP_TRUE@@am__include@ @am__quote@./$(DEPDIR)/hb2_8.Plo@am__quote@ @AMDEP_TRUE@@am__include@ @am__quote@./$(DEPDIR)/hb_10.Plo@am__quote@ @AMDEP_TRUE@@am__include@ @am__quote@./$(DEPDIR)/hb_12.Plo@am__quote@ @AMDEP_TRUE@@am__include@ @am__quote@./$(DEPDIR)/hb_15.Plo@am__quote@ @AMDEP_TRUE@@am__include@ @am__quote@./$(DEPDIR)/hb_16.Plo@am__quote@ @AMDEP_TRUE@@am__include@ @am__quote@./$(DEPDIR)/hb_2.Plo@am__quote@ @AMDEP_TRUE@@am__include@ @am__quote@./$(DEPDIR)/hb_20.Plo@am__quote@ @AMDEP_TRUE@@am__include@ @am__quote@./$(DEPDIR)/hb_25.Plo@am__quote@ @AMDEP_TRUE@@am__include@ @am__quote@./$(DEPDIR)/hb_3.Plo@am__quote@ @AMDEP_TRUE@@am__include@ @am__quote@./$(DEPDIR)/hb_32.Plo@am__quote@ @AMDEP_TRUE@@am__include@ @am__quote@./$(DEPDIR)/hb_4.Plo@am__quote@ @AMDEP_TRUE@@am__include@ @am__quote@./$(DEPDIR)/hb_5.Plo@am__quote@ @AMDEP_TRUE@@am__include@ @am__quote@./$(DEPDIR)/hb_6.Plo@am__quote@ @AMDEP_TRUE@@am__include@ @am__quote@./$(DEPDIR)/hb_64.Plo@am__quote@ @AMDEP_TRUE@@am__include@ @am__quote@./$(DEPDIR)/hb_7.Plo@am__quote@ @AMDEP_TRUE@@am__include@ @am__quote@./$(DEPDIR)/hb_8.Plo@am__quote@ @AMDEP_TRUE@@am__include@ @am__quote@./$(DEPDIR)/hb_9.Plo@am__quote@ @AMDEP_TRUE@@am__include@ @am__quote@./$(DEPDIR)/hc2cb2_16.Plo@am__quote@ @AMDEP_TRUE@@am__include@ @am__quote@./$(DEPDIR)/hc2cb2_20.Plo@am__quote@ @AMDEP_TRUE@@am__include@ @am__quote@./$(DEPDIR)/hc2cb2_32.Plo@am__quote@ @AMDEP_TRUE@@am__include@ @am__quote@./$(DEPDIR)/hc2cb2_4.Plo@am__quote@ @AMDEP_TRUE@@am__include@ @am__quote@./$(DEPDIR)/hc2cb2_8.Plo@am__quote@ @AMDEP_TRUE@@am__include@ @am__quote@./$(DEPDIR)/hc2cb_10.Plo@am__quote@ @AMDEP_TRUE@@am__include@ @am__quote@./$(DEPDIR)/hc2cb_12.Plo@am__quote@ @AMDEP_TRUE@@am__include@ @am__quote@./$(DEPDIR)/hc2cb_16.Plo@am__quote@ @AMDEP_TRUE@@am__include@ @am__quote@./$(DEPDIR)/hc2cb_2.Plo@am__quote@ @AMDEP_TRUE@@am__include@ @am__quote@./$(DEPDIR)/hc2cb_20.Plo@am__quote@ @AMDEP_TRUE@@am__include@ @am__quote@./$(DEPDIR)/hc2cb_32.Plo@am__quote@ @AMDEP_TRUE@@am__include@ @am__quote@./$(DEPDIR)/hc2cb_4.Plo@am__quote@ @AMDEP_TRUE@@am__include@ @am__quote@./$(DEPDIR)/hc2cb_6.Plo@am__quote@ @AMDEP_TRUE@@am__include@ @am__quote@./$(DEPDIR)/hc2cb_8.Plo@am__quote@ @AMDEP_TRUE@@am__include@ @am__quote@./$(DEPDIR)/hc2cbdft2_16.Plo@am__quote@ @AMDEP_TRUE@@am__include@ @am__quote@./$(DEPDIR)/hc2cbdft2_20.Plo@am__quote@ @AMDEP_TRUE@@am__include@ @am__quote@./$(DEPDIR)/hc2cbdft2_32.Plo@am__quote@ @AMDEP_TRUE@@am__include@ @am__quote@./$(DEPDIR)/hc2cbdft2_4.Plo@am__quote@ @AMDEP_TRUE@@am__include@ @am__quote@./$(DEPDIR)/hc2cbdft2_8.Plo@am__quote@ @AMDEP_TRUE@@am__include@ @am__quote@./$(DEPDIR)/hc2cbdft_10.Plo@am__quote@ @AMDEP_TRUE@@am__include@ @am__quote@./$(DEPDIR)/hc2cbdft_12.Plo@am__quote@ @AMDEP_TRUE@@am__include@ @am__quote@./$(DEPDIR)/hc2cbdft_16.Plo@am__quote@ @AMDEP_TRUE@@am__include@ @am__quote@./$(DEPDIR)/hc2cbdft_2.Plo@am__quote@ @AMDEP_TRUE@@am__include@ @am__quote@./$(DEPDIR)/hc2cbdft_20.Plo@am__quote@ @AMDEP_TRUE@@am__include@ @am__quote@./$(DEPDIR)/hc2cbdft_32.Plo@am__quote@ @AMDEP_TRUE@@am__include@ @am__quote@./$(DEPDIR)/hc2cbdft_4.Plo@am__quote@ @AMDEP_TRUE@@am__include@ @am__quote@./$(DEPDIR)/hc2cbdft_6.Plo@am__quote@ @AMDEP_TRUE@@am__include@ @am__quote@./$(DEPDIR)/hc2cbdft_8.Plo@am__quote@ @AMDEP_TRUE@@am__include@ @am__quote@./$(DEPDIR)/r2cbIII_10.Plo@am__quote@ @AMDEP_TRUE@@am__include@ @am__quote@./$(DEPDIR)/r2cbIII_12.Plo@am__quote@ @AMDEP_TRUE@@am__include@ @am__quote@./$(DEPDIR)/r2cbIII_15.Plo@am__quote@ @AMDEP_TRUE@@am__include@ @am__quote@./$(DEPDIR)/r2cbIII_16.Plo@am__quote@ @AMDEP_TRUE@@am__include@ @am__quote@./$(DEPDIR)/r2cbIII_2.Plo@am__quote@ @AMDEP_TRUE@@am__include@ @am__quote@./$(DEPDIR)/r2cbIII_20.Plo@am__quote@ @AMDEP_TRUE@@am__include@ @am__quote@./$(DEPDIR)/r2cbIII_25.Plo@am__quote@ @AMDEP_TRUE@@am__include@ @am__quote@./$(DEPDIR)/r2cbIII_3.Plo@am__quote@ @AMDEP_TRUE@@am__include@ @am__quote@./$(DEPDIR)/r2cbIII_32.Plo@am__quote@ @AMDEP_TRUE@@am__include@ @am__quote@./$(DEPDIR)/r2cbIII_4.Plo@am__quote@ @AMDEP_TRUE@@am__include@ @am__quote@./$(DEPDIR)/r2cbIII_5.Plo@am__quote@ @AMDEP_TRUE@@am__include@ @am__quote@./$(DEPDIR)/r2cbIII_6.Plo@am__quote@ @AMDEP_TRUE@@am__include@ @am__quote@./$(DEPDIR)/r2cbIII_64.Plo@am__quote@ @AMDEP_TRUE@@am__include@ @am__quote@./$(DEPDIR)/r2cbIII_7.Plo@am__quote@ @AMDEP_TRUE@@am__include@ @am__quote@./$(DEPDIR)/r2cbIII_8.Plo@am__quote@ @AMDEP_TRUE@@am__include@ @am__quote@./$(DEPDIR)/r2cbIII_9.Plo@am__quote@ @AMDEP_TRUE@@am__include@ @am__quote@./$(DEPDIR)/r2cb_10.Plo@am__quote@ @AMDEP_TRUE@@am__include@ @am__quote@./$(DEPDIR)/r2cb_11.Plo@am__quote@ @AMDEP_TRUE@@am__include@ @am__quote@./$(DEPDIR)/r2cb_12.Plo@am__quote@ @AMDEP_TRUE@@am__include@ @am__quote@./$(DEPDIR)/r2cb_128.Plo@am__quote@ @AMDEP_TRUE@@am__include@ @am__quote@./$(DEPDIR)/r2cb_13.Plo@am__quote@ @AMDEP_TRUE@@am__include@ @am__quote@./$(DEPDIR)/r2cb_14.Plo@am__quote@ @AMDEP_TRUE@@am__include@ @am__quote@./$(DEPDIR)/r2cb_15.Plo@am__quote@ @AMDEP_TRUE@@am__include@ @am__quote@./$(DEPDIR)/r2cb_16.Plo@am__quote@ @AMDEP_TRUE@@am__include@ @am__quote@./$(DEPDIR)/r2cb_2.Plo@am__quote@ @AMDEP_TRUE@@am__include@ @am__quote@./$(DEPDIR)/r2cb_20.Plo@am__quote@ @AMDEP_TRUE@@am__include@ @am__quote@./$(DEPDIR)/r2cb_25.Plo@am__quote@ @AMDEP_TRUE@@am__include@ @am__quote@./$(DEPDIR)/r2cb_3.Plo@am__quote@ @AMDEP_TRUE@@am__include@ @am__quote@./$(DEPDIR)/r2cb_32.Plo@am__quote@ @AMDEP_TRUE@@am__include@ @am__quote@./$(DEPDIR)/r2cb_4.Plo@am__quote@ @AMDEP_TRUE@@am__include@ @am__quote@./$(DEPDIR)/r2cb_5.Plo@am__quote@ @AMDEP_TRUE@@am__include@ @am__quote@./$(DEPDIR)/r2cb_6.Plo@am__quote@ @AMDEP_TRUE@@am__include@ @am__quote@./$(DEPDIR)/r2cb_64.Plo@am__quote@ @AMDEP_TRUE@@am__include@ @am__quote@./$(DEPDIR)/r2cb_7.Plo@am__quote@ @AMDEP_TRUE@@am__include@ @am__quote@./$(DEPDIR)/r2cb_8.Plo@am__quote@ @AMDEP_TRUE@@am__include@ @am__quote@./$(DEPDIR)/r2cb_9.Plo@am__quote@ .c.o: @am__fastdepCC_TRUE@ $(COMPILE) -MT $@ -MD -MP -MF $(DEPDIR)/$*.Tpo -c -o $@ $< @am__fastdepCC_TRUE@ $(am__mv) $(DEPDIR)/$*.Tpo $(DEPDIR)/$*.Po @AMDEP_TRUE@@am__fastdepCC_FALSE@ source='$<' object='$@' libtool=no @AMDEPBACKSLASH@ @AMDEP_TRUE@@am__fastdepCC_FALSE@ DEPDIR=$(DEPDIR) $(CCDEPMODE) $(depcomp) @AMDEPBACKSLASH@ @am__fastdepCC_FALSE@ $(COMPILE) -c $< .c.obj: @am__fastdepCC_TRUE@ $(COMPILE) -MT $@ -MD -MP -MF $(DEPDIR)/$*.Tpo -c -o $@ `$(CYGPATH_W) '$<'` @am__fastdepCC_TRUE@ $(am__mv) $(DEPDIR)/$*.Tpo $(DEPDIR)/$*.Po @AMDEP_TRUE@@am__fastdepCC_FALSE@ source='$<' object='$@' libtool=no @AMDEPBACKSLASH@ @AMDEP_TRUE@@am__fastdepCC_FALSE@ DEPDIR=$(DEPDIR) $(CCDEPMODE) $(depcomp) @AMDEPBACKSLASH@ @am__fastdepCC_FALSE@ $(COMPILE) -c `$(CYGPATH_W) '$<'` .c.lo: @am__fastdepCC_TRUE@ $(LTCOMPILE) -MT $@ -MD -MP -MF $(DEPDIR)/$*.Tpo -c -o $@ $< @am__fastdepCC_TRUE@ $(am__mv) $(DEPDIR)/$*.Tpo $(DEPDIR)/$*.Plo @AMDEP_TRUE@@am__fastdepCC_FALSE@ source='$<' object='$@' libtool=yes @AMDEPBACKSLASH@ @AMDEP_TRUE@@am__fastdepCC_FALSE@ DEPDIR=$(DEPDIR) $(CCDEPMODE) $(depcomp) @AMDEPBACKSLASH@ @am__fastdepCC_FALSE@ $(LTCOMPILE) -c -o $@ $< mostlyclean-libtool: -rm -f *.lo clean-libtool: -rm -rf .libs _libs ID: $(HEADERS) $(SOURCES) $(LISP) $(TAGS_FILES) list='$(SOURCES) $(HEADERS) $(LISP) $(TAGS_FILES)'; \ unique=`for i in $$list; do \ if test -f "$$i"; then echo $$i; else echo $(srcdir)/$$i; fi; \ done | \ $(AWK) '{ files[$$0] = 1; nonempty = 1; } \ END { if (nonempty) { for (i in files) print i; }; }'`; \ mkid -fID $$unique tags: TAGS TAGS: $(HEADERS) $(SOURCES) $(TAGS_DEPENDENCIES) \ $(TAGS_FILES) $(LISP) set x; \ here=`pwd`; \ list='$(SOURCES) $(HEADERS) $(LISP) $(TAGS_FILES)'; \ unique=`for i in $$list; do \ if test -f "$$i"; then echo $$i; else echo $(srcdir)/$$i; fi; \ done | \ $(AWK) '{ files[$$0] = 1; nonempty = 1; } \ END { if (nonempty) { for (i in files) print i; }; }'`; \ shift; \ if test -z "$(ETAGS_ARGS)$$*$$unique"; then :; else \ test -n "$$unique" || unique=$$empty_fix; \ if test $$# -gt 0; then \ $(ETAGS) $(ETAGSFLAGS) $(AM_ETAGSFLAGS) $(ETAGS_ARGS) \ "$$@" $$unique; \ else \ $(ETAGS) $(ETAGSFLAGS) $(AM_ETAGSFLAGS) $(ETAGS_ARGS) \ $$unique; \ fi; \ fi ctags: CTAGS CTAGS: $(HEADERS) $(SOURCES) $(TAGS_DEPENDENCIES) \ $(TAGS_FILES) $(LISP) list='$(SOURCES) $(HEADERS) $(LISP) $(TAGS_FILES)'; \ unique=`for i in $$list; do \ if test -f "$$i"; then echo $$i; else echo $(srcdir)/$$i; fi; \ done | \ $(AWK) '{ files[$$0] = 1; nonempty = 1; } \ END { if (nonempty) { for (i in files) print i; }; }'`; \ test -z "$(CTAGS_ARGS)$$unique" \ || $(CTAGS) $(CTAGSFLAGS) $(AM_CTAGSFLAGS) $(CTAGS_ARGS) \ $$unique GTAGS: here=`$(am__cd) $(top_builddir) && pwd` \ && $(am__cd) $(top_srcdir) \ && gtags -i $(GTAGS_ARGS) "$$here" distclean-tags: -rm -f TAGS ID GTAGS GRTAGS GSYMS GPATH tags distdir: $(DISTFILES) @srcdirstrip=`echo "$(srcdir)" | sed 's/[].[^$$\\*]/\\\\&/g'`; \ topsrcdirstrip=`echo "$(top_srcdir)" | sed 's/[].[^$$\\*]/\\\\&/g'`; \ list='$(DISTFILES)'; \ dist_files=`for file in $$list; do echo $$file; done | \ sed -e "s|^$$srcdirstrip/||;t" \ -e "s|^$$topsrcdirstrip/|$(top_builddir)/|;t"`; \ case $$dist_files in \ */*) $(MKDIR_P) `echo "$$dist_files" | \ sed '/\//!d;s|^|$(distdir)/|;s,/[^/]*$$,,' | \ sort -u` ;; \ esac; \ for file in $$dist_files; do \ if test -f $$file || test -d $$file; then d=.; else d=$(srcdir); fi; \ if test -d $$d/$$file; then \ dir=`echo "/$$file" | sed -e 's,/[^/]*$$,,'`; \ if test -d "$(distdir)/$$file"; then \ find "$(distdir)/$$file" -type d ! -perm -700 -exec chmod u+rwx {} \;; \ fi; \ if test -d $(srcdir)/$$file && test $$d != $(srcdir); then \ cp -fpR $(srcdir)/$$file "$(distdir)$$dir" || exit 1; \ find "$(distdir)/$$file" -type d ! -perm -700 -exec chmod u+rwx {} \;; \ fi; \ cp -fpR $$d/$$file "$(distdir)$$dir" || exit 1; \ else \ test -f "$(distdir)/$$file" \ || cp -p $$d/$$file "$(distdir)/$$file" \ || exit 1; \ fi; \ done check-am: all-am check: $(BUILT_SOURCES) $(MAKE) $(AM_MAKEFLAGS) check-am all-am: Makefile $(LTLIBRARIES) installdirs: install: $(BUILT_SOURCES) $(MAKE) $(AM_MAKEFLAGS) install-am install-exec: install-exec-am install-data: install-data-am uninstall: uninstall-am install-am: all-am @$(MAKE) $(AM_MAKEFLAGS) install-exec-am install-data-am installcheck: installcheck-am install-strip: if test -z '$(STRIP)'; then \ $(MAKE) $(AM_MAKEFLAGS) INSTALL_PROGRAM="$(INSTALL_STRIP_PROGRAM)" \ install_sh_PROGRAM="$(INSTALL_STRIP_PROGRAM)" INSTALL_STRIP_FLAG=-s \ install; \ else \ $(MAKE) $(AM_MAKEFLAGS) INSTALL_PROGRAM="$(INSTALL_STRIP_PROGRAM)" \ install_sh_PROGRAM="$(INSTALL_STRIP_PROGRAM)" INSTALL_STRIP_FLAG=-s \ "INSTALL_PROGRAM_ENV=STRIPPROG='$(STRIP)'" install; \ fi mostlyclean-generic: clean-generic: distclean-generic: -test -z "$(CONFIG_CLEAN_FILES)" || rm -f $(CONFIG_CLEAN_FILES) -test . = "$(srcdir)" || test -z "$(CONFIG_CLEAN_VPATH_FILES)" || rm -f $(CONFIG_CLEAN_VPATH_FILES) maintainer-clean-generic: @echo "This command is intended for maintainers to use" @echo "it deletes files that may require special tools to rebuild." -test -z "$(BUILT_SOURCES)" || rm -f $(BUILT_SOURCES) clean: clean-am clean-am: clean-generic clean-libtool clean-noinstLTLIBRARIES \ mostlyclean-am distclean: distclean-am -rm -rf ./$(DEPDIR) -rm -f Makefile distclean-am: clean-am distclean-compile distclean-generic \ distclean-tags dvi: dvi-am dvi-am: html: html-am html-am: info: info-am info-am: install-data-am: install-dvi: install-dvi-am install-dvi-am: install-exec-am: install-html: install-html-am install-html-am: install-info: install-info-am install-info-am: install-man: install-pdf: install-pdf-am install-pdf-am: install-ps: install-ps-am install-ps-am: installcheck-am: maintainer-clean: maintainer-clean-am -rm -rf ./$(DEPDIR) -rm -f Makefile maintainer-clean-am: distclean-am maintainer-clean-generic \ maintainer-clean-local mostlyclean: mostlyclean-am mostlyclean-am: mostlyclean-compile mostlyclean-generic \ mostlyclean-libtool pdf: pdf-am pdf-am: ps: ps-am ps-am: uninstall-am: .MAKE: all check install install-am install-strip .PHONY: CTAGS GTAGS all all-am check check-am clean clean-generic \ clean-libtool clean-noinstLTLIBRARIES ctags distclean \ distclean-compile distclean-generic distclean-libtool \ distclean-tags distdir dvi dvi-am html html-am info info-am \ install install-am install-data install-data-am install-dvi \ install-dvi-am install-exec install-exec-am install-html \ install-html-am install-info install-info-am install-man \ install-pdf install-pdf-am install-ps install-ps-am \ install-strip installcheck installcheck-am installdirs \ maintainer-clean maintainer-clean-generic \ maintainer-clean-local mostlyclean mostlyclean-compile \ mostlyclean-generic mostlyclean-libtool pdf pdf-am ps ps-am \ tags uninstall uninstall-am # rule to build codlist $(CODLIST): Makefile ( \ echo "#include \"ifftw.h\""; \ echo $(INCLUDE_SIMD_HEADER); \ echo; \ for i in $(ALL_CODELETS) NIL; do \ if test "$$i" != NIL; then \ j=`basename $$i | sed -e 's/[.][cS]$$//g'`; \ echo "extern void $(XRENAME)($(CODELET_NAME)$$j)(planner *);"; \ fi \ done; \ echo; \ echo; \ echo "extern const solvtab $(SOLVTAB_NAME);"; \ echo "const solvtab $(SOLVTAB_NAME) = {"; \ for i in $(ALL_CODELETS) NIL; do \ if test "$$i" != NIL; then \ j=`basename $$i | sed -e 's/[.][cS]$$//g'`; \ echo " SOLVTAB($(XRENAME)($(CODELET_NAME)$$j)),"; \ fi \ done; \ echo " SOLVTAB_END"; \ echo "};"; \ ) >$@ # only delete codlist.c in maintainer-mode, since it is included in the dist # FIXME: is there a way to delete in 'make clean' only when builddir != srcdir? maintainer-clean-local: rm -f $(CODLIST) # cancel the hideous builtin rules that cause an infinite loop @MAINTAINER_MODE_TRUE@%: %.o @MAINTAINER_MODE_TRUE@%: %.s @MAINTAINER_MODE_TRUE@%: %.c @MAINTAINER_MODE_TRUE@%: %.S @MAINTAINER_MODE_TRUE@r2cb_%.c: $(CODELET_DEPS) $(GEN_R2CB) @MAINTAINER_MODE_TRUE@ ($(PRELUDE_COMMANDS_RDFT); $(TWOVERS) $(GEN_R2CB) $(FLAGS_R2CB) -n $* -name r2cb_$* -include "r2cb.h") | $(ADD_DATE) | $(INDENT) >$@ @MAINTAINER_MODE_TRUE@hb_%.c: $(CODELET_DEPS) $(GEN_HC2HC) @MAINTAINER_MODE_TRUE@ ($(PRELUDE_COMMANDS_RDFT); $(TWOVERS) $(GEN_HC2HC) $(FLAGS_HB) -n $* -dif -name hb_$* -include "hb.h") | $(ADD_DATE) | $(INDENT) >$@ @MAINTAINER_MODE_TRUE@hb2_%.c: $(CODELET_DEPS) $(GEN_HC2HC) @MAINTAINER_MODE_TRUE@ ($(PRELUDE_COMMANDS_RDFT); $(TWOVERS) $(GEN_HC2HC) $(FLAGS_HB2) -n $* -dif -name hb2_$* -include "hb.h") | $(ADD_DATE) | $(INDENT) >$@ @MAINTAINER_MODE_TRUE@r2cbIII_%.c: $(CODELET_DEPS) $(GEN_R2CB) @MAINTAINER_MODE_TRUE@ ($(PRELUDE_COMMANDS_RDFT); $(TWOVERS) $(GEN_R2CB) $(FLAGS_R2CB) -n $* -name r2cbIII_$* -dft-III -include "r2cbIII.h") | $(ADD_DATE) | $(INDENT) >$@ @MAINTAINER_MODE_TRUE@hc2cb_%.c: $(CODELET_DEPS) $(GEN_HC2C) @MAINTAINER_MODE_TRUE@ ($(PRELUDE_COMMANDS_RDFT); $(TWOVERS) $(GEN_HC2C) $(FLAGS_HC2CB) -n $* -dif -name hc2cb_$* -include "hc2cb.h") | $(ADD_DATE) | $(INDENT) >$@ @MAINTAINER_MODE_TRUE@hc2cb2_%.c: $(CODELET_DEPS) $(GEN_HC2C) @MAINTAINER_MODE_TRUE@ ($(PRELUDE_COMMANDS_RDFT); $(TWOVERS) $(GEN_HC2C) $(FLAGS_HC2CB2) -n $* -dif -name hc2cb2_$* -include "hc2cb.h") | $(ADD_DATE) | $(INDENT) >$@ @MAINTAINER_MODE_TRUE@hc2cbdft_%.c: $(CODELET_DEPS) $(GEN_HC2CDFT) @MAINTAINER_MODE_TRUE@ ($(PRELUDE_COMMANDS_RDFT); $(TWOVERS) $(GEN_HC2CDFT) $(FLAGS_HC2CB) -n $* -dif -name hc2cbdft_$* -include "hc2cb.h") | $(ADD_DATE) | $(INDENT) >$@ @MAINTAINER_MODE_TRUE@hc2cbdft2_%.c: $(CODELET_DEPS) $(GEN_HC2CDFT) @MAINTAINER_MODE_TRUE@ ($(PRELUDE_COMMANDS_RDFT); $(TWOVERS) $(GEN_HC2CDFT) $(FLAGS_HC2CB) -n $* -dif -name hc2cbdft2_$* -include "hc2cb.h") | $(ADD_DATE) | $(INDENT) >$@ # Tell versions [3.59,3.63) of GNU make to not export all variables. # Otherwise a system limit (for SysV at least) may be exceeded. .NOEXPORT: fftw-3.3.3/rdft/scalar/r2cb/r2cb_2.c0000644000175000017500000000563712054410743013673 00000000000000/* * Copyright (c) 2003, 2007-11 Matteo Frigo * Copyright (c) 2003, 2007-11 Massachusetts Institute of Technology * * This program is free software; you can redistribute it and/or modify * it under the terms of the GNU General Public License as published by * the Free Software Foundation; either version 2 of the License, or * (at your option) any later version. * * This program is distributed in the hope that it will be useful, * but WITHOUT ANY WARRANTY; without even the implied warranty of * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the * GNU General Public License for more details. * * You should have received a copy of the GNU General Public License * along with this program; if not, write to the Free Software * Foundation, Inc., 51 Franklin Street, Fifth Floor, Boston, MA 02110-1301 USA * */ /* This file was automatically generated --- DO NOT EDIT */ /* Generated on Sun Nov 25 07:41:07 EST 2012 */ #include "codelet-rdft.h" #ifdef HAVE_FMA /* Generated by: ../../../genfft/gen_r2cb.native -fma -reorder-insns -schedule-for-pipeline -compact -variables 4 -pipeline-latency 4 -sign 1 -n 2 -name r2cb_2 -include r2cb.h */ /* * This function contains 2 FP additions, 0 FP multiplications, * (or, 2 additions, 0 multiplications, 0 fused multiply/add), * 3 stack variables, 0 constants, and 4 memory accesses */ #include "r2cb.h" static void r2cb_2(R *R0, R *R1, R *Cr, R *Ci, stride rs, stride csr, stride csi, INT v, INT ivs, INT ovs) { { INT i; for (i = v; i > 0; i = i - 1, R0 = R0 + ovs, R1 = R1 + ovs, Cr = Cr + ivs, Ci = Ci + ivs, MAKE_VOLATILE_STRIDE(8, rs), MAKE_VOLATILE_STRIDE(8, csr), MAKE_VOLATILE_STRIDE(8, csi)) { E T1, T2; T1 = Cr[0]; T2 = Cr[WS(csr, 1)]; R0[0] = T1 + T2; R1[0] = T1 - T2; } } } static const kr2c_desc desc = { 2, "r2cb_2", {2, 0, 0, 0}, &GENUS }; void X(codelet_r2cb_2) (planner *p) { X(kr2c_register) (p, r2cb_2, &desc); } #else /* HAVE_FMA */ /* Generated by: ../../../genfft/gen_r2cb.native -compact -variables 4 -pipeline-latency 4 -sign 1 -n 2 -name r2cb_2 -include r2cb.h */ /* * This function contains 2 FP additions, 0 FP multiplications, * (or, 2 additions, 0 multiplications, 0 fused multiply/add), * 3 stack variables, 0 constants, and 4 memory accesses */ #include "r2cb.h" static void r2cb_2(R *R0, R *R1, R *Cr, R *Ci, stride rs, stride csr, stride csi, INT v, INT ivs, INT ovs) { { INT i; for (i = v; i > 0; i = i - 1, R0 = R0 + ovs, R1 = R1 + ovs, Cr = Cr + ivs, Ci = Ci + ivs, MAKE_VOLATILE_STRIDE(8, rs), MAKE_VOLATILE_STRIDE(8, csr), MAKE_VOLATILE_STRIDE(8, csi)) { E T1, T2; T1 = Cr[0]; T2 = Cr[WS(csr, 1)]; R1[0] = T1 - T2; R0[0] = T1 + T2; } } } static const kr2c_desc desc = { 2, "r2cb_2", {2, 0, 0, 0}, &GENUS }; void X(codelet_r2cb_2) (planner *p) { X(kr2c_register) (p, r2cb_2, &desc); } #endif /* HAVE_FMA */ fftw-3.3.3/rdft/scalar/r2cb/r2cb_3.c0000644000175000017500000000676712054410743013701 00000000000000/* * Copyright (c) 2003, 2007-11 Matteo Frigo * Copyright (c) 2003, 2007-11 Massachusetts Institute of Technology * * This program is free software; you can redistribute it and/or modify * it under the terms of the GNU General Public License as published by * the Free Software Foundation; either version 2 of the License, or * (at your option) any later version. * * This program is distributed in the hope that it will be useful, * but WITHOUT ANY WARRANTY; without even the implied warranty of * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the * GNU General Public License for more details. * * You should have received a copy of the GNU General Public License * along with this program; if not, write to the Free Software * Foundation, Inc., 51 Franklin Street, Fifth Floor, Boston, MA 02110-1301 USA * */ /* This file was automatically generated --- DO NOT EDIT */ /* Generated on Sun Nov 25 07:41:07 EST 2012 */ #include "codelet-rdft.h" #ifdef HAVE_FMA /* Generated by: ../../../genfft/gen_r2cb.native -fma -reorder-insns -schedule-for-pipeline -compact -variables 4 -pipeline-latency 4 -sign 1 -n 3 -name r2cb_3 -include r2cb.h */ /* * This function contains 4 FP additions, 3 FP multiplications, * (or, 1 additions, 0 multiplications, 3 fused multiply/add), * 7 stack variables, 2 constants, and 6 memory accesses */ #include "r2cb.h" static void r2cb_3(R *R0, R *R1, R *Cr, R *Ci, stride rs, stride csr, stride csi, INT v, INT ivs, INT ovs) { DK(KP1_732050807, +1.732050807568877293527446341505872366942805254); DK(KP2_000000000, +2.000000000000000000000000000000000000000000000); { INT i; for (i = v; i > 0; i = i - 1, R0 = R0 + ovs, R1 = R1 + ovs, Cr = Cr + ivs, Ci = Ci + ivs, MAKE_VOLATILE_STRIDE(12, rs), MAKE_VOLATILE_STRIDE(12, csr), MAKE_VOLATILE_STRIDE(12, csi)) { E T4, T1, T2, T3; T4 = Ci[WS(csi, 1)]; T1 = Cr[0]; T2 = Cr[WS(csr, 1)]; R0[0] = FMA(KP2_000000000, T2, T1); T3 = T1 - T2; R1[0] = FNMS(KP1_732050807, T4, T3); R0[WS(rs, 1)] = FMA(KP1_732050807, T4, T3); } } } static const kr2c_desc desc = { 3, "r2cb_3", {1, 0, 3, 0}, &GENUS }; void X(codelet_r2cb_3) (planner *p) { X(kr2c_register) (p, r2cb_3, &desc); } #else /* HAVE_FMA */ /* Generated by: ../../../genfft/gen_r2cb.native -compact -variables 4 -pipeline-latency 4 -sign 1 -n 3 -name r2cb_3 -include r2cb.h */ /* * This function contains 4 FP additions, 2 FP multiplications, * (or, 3 additions, 1 multiplications, 1 fused multiply/add), * 8 stack variables, 2 constants, and 6 memory accesses */ #include "r2cb.h" static void r2cb_3(R *R0, R *R1, R *Cr, R *Ci, stride rs, stride csr, stride csi, INT v, INT ivs, INT ovs) { DK(KP2_000000000, +2.000000000000000000000000000000000000000000000); DK(KP1_732050807, +1.732050807568877293527446341505872366942805254); { INT i; for (i = v; i > 0; i = i - 1, R0 = R0 + ovs, R1 = R1 + ovs, Cr = Cr + ivs, Ci = Ci + ivs, MAKE_VOLATILE_STRIDE(12, rs), MAKE_VOLATILE_STRIDE(12, csr), MAKE_VOLATILE_STRIDE(12, csi)) { E T5, T1, T2, T3, T4; T4 = Ci[WS(csi, 1)]; T5 = KP1_732050807 * T4; T1 = Cr[0]; T2 = Cr[WS(csr, 1)]; T3 = T1 - T2; R0[0] = FMA(KP2_000000000, T2, T1); R0[WS(rs, 1)] = T3 + T5; R1[0] = T3 - T5; } } } static const kr2c_desc desc = { 3, "r2cb_3", {3, 1, 1, 0}, &GENUS }; void X(codelet_r2cb_3) (planner *p) { X(kr2c_register) (p, r2cb_3, &desc); } #endif /* HAVE_FMA */ fftw-3.3.3/rdft/scalar/r2cb/r2cb_4.c0000644000175000017500000000711612054410743013667 00000000000000/* * Copyright (c) 2003, 2007-11 Matteo Frigo * Copyright (c) 2003, 2007-11 Massachusetts Institute of Technology * * This program is free software; you can redistribute it and/or modify * it under the terms of the GNU General Public License as published by * the Free Software Foundation; either version 2 of the License, or * (at your option) any later version. * * This program is distributed in the hope that it will be useful, * but WITHOUT ANY WARRANTY; without even the implied warranty of * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the * GNU General Public License for more details. * * You should have received a copy of the GNU General Public License * along with this program; if not, write to the Free Software * Foundation, Inc., 51 Franklin Street, Fifth Floor, Boston, MA 02110-1301 USA * */ /* This file was automatically generated --- DO NOT EDIT */ /* Generated on Sun Nov 25 07:41:07 EST 2012 */ #include "codelet-rdft.h" #ifdef HAVE_FMA /* Generated by: ../../../genfft/gen_r2cb.native -fma -reorder-insns -schedule-for-pipeline -compact -variables 4 -pipeline-latency 4 -sign 1 -n 4 -name r2cb_4 -include r2cb.h */ /* * This function contains 6 FP additions, 4 FP multiplications, * (or, 2 additions, 0 multiplications, 4 fused multiply/add), * 8 stack variables, 1 constants, and 8 memory accesses */ #include "r2cb.h" static void r2cb_4(R *R0, R *R1, R *Cr, R *Ci, stride rs, stride csr, stride csi, INT v, INT ivs, INT ovs) { DK(KP2_000000000, +2.000000000000000000000000000000000000000000000); { INT i; for (i = v; i > 0; i = i - 1, R0 = R0 + ovs, R1 = R1 + ovs, Cr = Cr + ivs, Ci = Ci + ivs, MAKE_VOLATILE_STRIDE(16, rs), MAKE_VOLATILE_STRIDE(16, csr), MAKE_VOLATILE_STRIDE(16, csi)) { E T4, T6, T1, T2, T3, T5; T4 = Cr[WS(csr, 1)]; T6 = Ci[WS(csi, 1)]; T1 = Cr[0]; T2 = Cr[WS(csr, 2)]; T3 = T1 + T2; T5 = T1 - T2; R1[0] = FNMS(KP2_000000000, T6, T5); R1[WS(rs, 1)] = FMA(KP2_000000000, T6, T5); R0[0] = FMA(KP2_000000000, T4, T3); R0[WS(rs, 1)] = FNMS(KP2_000000000, T4, T3); } } } static const kr2c_desc desc = { 4, "r2cb_4", {2, 0, 4, 0}, &GENUS }; void X(codelet_r2cb_4) (planner *p) { X(kr2c_register) (p, r2cb_4, &desc); } #else /* HAVE_FMA */ /* Generated by: ../../../genfft/gen_r2cb.native -compact -variables 4 -pipeline-latency 4 -sign 1 -n 4 -name r2cb_4 -include r2cb.h */ /* * This function contains 6 FP additions, 2 FP multiplications, * (or, 6 additions, 2 multiplications, 0 fused multiply/add), * 10 stack variables, 1 constants, and 8 memory accesses */ #include "r2cb.h" static void r2cb_4(R *R0, R *R1, R *Cr, R *Ci, stride rs, stride csr, stride csi, INT v, INT ivs, INT ovs) { DK(KP2_000000000, +2.000000000000000000000000000000000000000000000); { INT i; for (i = v; i > 0; i = i - 1, R0 = R0 + ovs, R1 = R1 + ovs, Cr = Cr + ivs, Ci = Ci + ivs, MAKE_VOLATILE_STRIDE(16, rs), MAKE_VOLATILE_STRIDE(16, csr), MAKE_VOLATILE_STRIDE(16, csi)) { E T5, T8, T3, T6; { E T4, T7, T1, T2; T4 = Cr[WS(csr, 1)]; T5 = KP2_000000000 * T4; T7 = Ci[WS(csi, 1)]; T8 = KP2_000000000 * T7; T1 = Cr[0]; T2 = Cr[WS(csr, 2)]; T3 = T1 + T2; T6 = T1 - T2; } R0[WS(rs, 1)] = T3 - T5; R1[WS(rs, 1)] = T6 + T8; R0[0] = T3 + T5; R1[0] = T6 - T8; } } } static const kr2c_desc desc = { 4, "r2cb_4", {6, 2, 0, 0}, &GENUS }; void X(codelet_r2cb_4) (planner *p) { X(kr2c_register) (p, r2cb_4, &desc); } #endif /* HAVE_FMA */ fftw-3.3.3/rdft/scalar/r2cb/r2cb_5.c0000644000175000017500000001131612054410743013665 00000000000000/* * Copyright (c) 2003, 2007-11 Matteo Frigo * Copyright (c) 2003, 2007-11 Massachusetts Institute of Technology * * This program is free software; you can redistribute it and/or modify * it under the terms of the GNU General Public License as published by * the Free Software Foundation; either version 2 of the License, or * (at your option) any later version. * * This program is distributed in the hope that it will be useful, * but WITHOUT ANY WARRANTY; without even the implied warranty of * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the * GNU General Public License for more details. * * You should have received a copy of the GNU General Public License * along with this program; if not, write to the Free Software * Foundation, Inc., 51 Franklin Street, Fifth Floor, Boston, MA 02110-1301 USA * */ /* This file was automatically generated --- DO NOT EDIT */ /* Generated on Sun Nov 25 07:41:07 EST 2012 */ #include "codelet-rdft.h" #ifdef HAVE_FMA /* Generated by: ../../../genfft/gen_r2cb.native -fma -reorder-insns -schedule-for-pipeline -compact -variables 4 -pipeline-latency 4 -sign 1 -n 5 -name r2cb_5 -include r2cb.h */ /* * This function contains 12 FP additions, 10 FP multiplications, * (or, 2 additions, 0 multiplications, 10 fused multiply/add), * 18 stack variables, 5 constants, and 10 memory accesses */ #include "r2cb.h" static void r2cb_5(R *R0, R *R1, R *Cr, R *Ci, stride rs, stride csr, stride csi, INT v, INT ivs, INT ovs) { DK(KP1_902113032, +1.902113032590307144232878666758764286811397268); DK(KP1_118033988, +1.118033988749894848204586834365638117720309180); DK(KP500000000, +0.500000000000000000000000000000000000000000000); DK(KP2_000000000, +2.000000000000000000000000000000000000000000000); DK(KP618033988, +0.618033988749894848204586834365638117720309180); { INT i; for (i = v; i > 0; i = i - 1, R0 = R0 + ovs, R1 = R1 + ovs, Cr = Cr + ivs, Ci = Ci + ivs, MAKE_VOLATILE_STRIDE(20, rs), MAKE_VOLATILE_STRIDE(20, csr), MAKE_VOLATILE_STRIDE(20, csi)) { E T1, T2, T3, Tc, Ta, T8, T9; T8 = Ci[WS(csi, 1)]; T9 = Ci[WS(csi, 2)]; T1 = Cr[0]; T2 = Cr[WS(csr, 1)]; T3 = Cr[WS(csr, 2)]; Tc = FMS(KP618033988, T8, T9); Ta = FMA(KP618033988, T9, T8); { E T6, T4, T5, T7, Tb; T6 = T2 - T3; T4 = T2 + T3; R0[0] = FMA(KP2_000000000, T4, T1); T5 = FNMS(KP500000000, T4, T1); T7 = FMA(KP1_118033988, T6, T5); Tb = FNMS(KP1_118033988, T6, T5); R0[WS(rs, 2)] = FMA(KP1_902113032, Ta, T7); R1[0] = FNMS(KP1_902113032, Ta, T7); R1[WS(rs, 1)] = FMA(KP1_902113032, Tc, Tb); R0[WS(rs, 1)] = FNMS(KP1_902113032, Tc, Tb); } } } } static const kr2c_desc desc = { 5, "r2cb_5", {2, 0, 10, 0}, &GENUS }; void X(codelet_r2cb_5) (planner *p) { X(kr2c_register) (p, r2cb_5, &desc); } #else /* HAVE_FMA */ /* Generated by: ../../../genfft/gen_r2cb.native -compact -variables 4 -pipeline-latency 4 -sign 1 -n 5 -name r2cb_5 -include r2cb.h */ /* * This function contains 12 FP additions, 7 FP multiplications, * (or, 8 additions, 3 multiplications, 4 fused multiply/add), * 18 stack variables, 5 constants, and 10 memory accesses */ #include "r2cb.h" static void r2cb_5(R *R0, R *R1, R *Cr, R *Ci, stride rs, stride csr, stride csi, INT v, INT ivs, INT ovs) { DK(KP2_000000000, +2.000000000000000000000000000000000000000000000); DK(KP1_118033988, +1.118033988749894848204586834365638117720309180); DK(KP500000000, +0.500000000000000000000000000000000000000000000); DK(KP1_902113032, +1.902113032590307144232878666758764286811397268); DK(KP1_175570504, +1.175570504584946258337411909278145537195304875); { INT i; for (i = v; i > 0; i = i - 1, R0 = R0 + ovs, R1 = R1 + ovs, Cr = Cr + ivs, Ci = Ci + ivs, MAKE_VOLATILE_STRIDE(20, rs), MAKE_VOLATILE_STRIDE(20, csr), MAKE_VOLATILE_STRIDE(20, csi)) { E Ta, Tc, T1, T4, T5, T6, Tb, T7; { E T8, T9, T2, T3; T8 = Ci[WS(csi, 1)]; T9 = Ci[WS(csi, 2)]; Ta = FNMS(KP1_902113032, T9, KP1_175570504 * T8); Tc = FMA(KP1_902113032, T8, KP1_175570504 * T9); T1 = Cr[0]; T2 = Cr[WS(csr, 1)]; T3 = Cr[WS(csr, 2)]; T4 = T2 + T3; T5 = FNMS(KP500000000, T4, T1); T6 = KP1_118033988 * (T2 - T3); } R0[0] = FMA(KP2_000000000, T4, T1); Tb = T6 + T5; R1[0] = Tb - Tc; R0[WS(rs, 2)] = Tb + Tc; T7 = T5 - T6; R0[WS(rs, 1)] = T7 - Ta; R1[WS(rs, 1)] = T7 + Ta; } } } static const kr2c_desc desc = { 5, "r2cb_5", {8, 3, 4, 0}, &GENUS }; void X(codelet_r2cb_5) (planner *p) { X(kr2c_register) (p, r2cb_5, &desc); } #endif /* HAVE_FMA */ fftw-3.3.3/rdft/scalar/r2cb/r2cb_6.c0000644000175000017500000001054012054410743013664 00000000000000/* * Copyright (c) 2003, 2007-11 Matteo Frigo * Copyright (c) 2003, 2007-11 Massachusetts Institute of Technology * * This program is free software; you can redistribute it and/or modify * it under the terms of the GNU General Public License as published by * the Free Software Foundation; either version 2 of the License, or * (at your option) any later version. * * This program is distributed in the hope that it will be useful, * but WITHOUT ANY WARRANTY; without even the implied warranty of * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the * GNU General Public License for more details. * * You should have received a copy of the GNU General Public License * along with this program; if not, write to the Free Software * Foundation, Inc., 51 Franklin Street, Fifth Floor, Boston, MA 02110-1301 USA * */ /* This file was automatically generated --- DO NOT EDIT */ /* Generated on Sun Nov 25 07:41:07 EST 2012 */ #include "codelet-rdft.h" #ifdef HAVE_FMA /* Generated by: ../../../genfft/gen_r2cb.native -fma -reorder-insns -schedule-for-pipeline -compact -variables 4 -pipeline-latency 4 -sign 1 -n 6 -name r2cb_6 -include r2cb.h */ /* * This function contains 14 FP additions, 6 FP multiplications, * (or, 8 additions, 0 multiplications, 6 fused multiply/add), * 13 stack variables, 2 constants, and 12 memory accesses */ #include "r2cb.h" static void r2cb_6(R *R0, R *R1, R *Cr, R *Ci, stride rs, stride csr, stride csi, INT v, INT ivs, INT ovs) { DK(KP1_732050807, +1.732050807568877293527446341505872366942805254); DK(KP2_000000000, +2.000000000000000000000000000000000000000000000); { INT i; for (i = v; i > 0; i = i - 1, R0 = R0 + ovs, R1 = R1 + ovs, Cr = Cr + ivs, Ci = Ci + ivs, MAKE_VOLATILE_STRIDE(24, rs), MAKE_VOLATILE_STRIDE(24, csr), MAKE_VOLATILE_STRIDE(24, csi)) { E T4, T7, T3, Te, Tc, T5; { E T1, T2, Ta, Tb; T1 = Cr[0]; T2 = Cr[WS(csr, 3)]; Ta = Ci[WS(csi, 2)]; Tb = Ci[WS(csi, 1)]; T4 = Cr[WS(csr, 2)]; T7 = T1 - T2; T3 = T1 + T2; Te = Ta + Tb; Tc = Ta - Tb; T5 = Cr[WS(csr, 1)]; } { E T6, T8, Td, T9; T6 = T4 + T5; T8 = T5 - T4; Td = T7 + T8; R1[WS(rs, 1)] = FNMS(KP2_000000000, T8, T7); T9 = T3 - T6; R0[0] = FMA(KP2_000000000, T6, T3); R1[WS(rs, 2)] = FMA(KP1_732050807, Te, Td); R1[0] = FNMS(KP1_732050807, Te, Td); R0[WS(rs, 1)] = FMA(KP1_732050807, Tc, T9); R0[WS(rs, 2)] = FNMS(KP1_732050807, Tc, T9); } } } } static const kr2c_desc desc = { 6, "r2cb_6", {8, 0, 6, 0}, &GENUS }; void X(codelet_r2cb_6) (planner *p) { X(kr2c_register) (p, r2cb_6, &desc); } #else /* HAVE_FMA */ /* Generated by: ../../../genfft/gen_r2cb.native -compact -variables 4 -pipeline-latency 4 -sign 1 -n 6 -name r2cb_6 -include r2cb.h */ /* * This function contains 14 FP additions, 4 FP multiplications, * (or, 12 additions, 2 multiplications, 2 fused multiply/add), * 17 stack variables, 2 constants, and 12 memory accesses */ #include "r2cb.h" static void r2cb_6(R *R0, R *R1, R *Cr, R *Ci, stride rs, stride csr, stride csi, INT v, INT ivs, INT ovs) { DK(KP2_000000000, +2.000000000000000000000000000000000000000000000); DK(KP1_732050807, +1.732050807568877293527446341505872366942805254); { INT i; for (i = v; i > 0; i = i - 1, R0 = R0 + ovs, R1 = R1 + ovs, Cr = Cr + ivs, Ci = Ci + ivs, MAKE_VOLATILE_STRIDE(24, rs), MAKE_VOLATILE_STRIDE(24, csr), MAKE_VOLATILE_STRIDE(24, csi)) { E T3, T7, Tc, Te, T6, T8, T1, T2, T9, Td; T1 = Cr[0]; T2 = Cr[WS(csr, 3)]; T3 = T1 - T2; T7 = T1 + T2; { E Ta, Tb, T4, T5; Ta = Ci[WS(csi, 2)]; Tb = Ci[WS(csi, 1)]; Tc = KP1_732050807 * (Ta - Tb); Te = KP1_732050807 * (Ta + Tb); T4 = Cr[WS(csr, 2)]; T5 = Cr[WS(csr, 1)]; T6 = T4 - T5; T8 = T4 + T5; } R1[WS(rs, 1)] = FMA(KP2_000000000, T6, T3); R0[0] = FMA(KP2_000000000, T8, T7); T9 = T7 - T8; R0[WS(rs, 2)] = T9 - Tc; R0[WS(rs, 1)] = T9 + Tc; Td = T3 - T6; R1[0] = Td - Te; R1[WS(rs, 2)] = Td + Te; } } } static const kr2c_desc desc = { 6, "r2cb_6", {12, 2, 2, 0}, &GENUS }; void X(codelet_r2cb_6) (planner *p) { X(kr2c_register) (p, r2cb_6, &desc); } #endif /* HAVE_FMA */ fftw-3.3.3/rdft/scalar/r2cb/r2cb_7.c0000644000175000017500000001361412054410744013673 00000000000000/* * Copyright (c) 2003, 2007-11 Matteo Frigo * Copyright (c) 2003, 2007-11 Massachusetts Institute of Technology * * This program is free software; you can redistribute it and/or modify * it under the terms of the GNU General Public License as published by * the Free Software Foundation; either version 2 of the License, or * (at your option) any later version. * * This program is distributed in the hope that it will be useful, * but WITHOUT ANY WARRANTY; without even the implied warranty of * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the * GNU General Public License for more details. * * You should have received a copy of the GNU General Public License * along with this program; if not, write to the Free Software * Foundation, Inc., 51 Franklin Street, Fifth Floor, Boston, MA 02110-1301 USA * */ /* This file was automatically generated --- DO NOT EDIT */ /* Generated on Sun Nov 25 07:41:07 EST 2012 */ #include "codelet-rdft.h" #ifdef HAVE_FMA /* Generated by: ../../../genfft/gen_r2cb.native -fma -reorder-insns -schedule-for-pipeline -compact -variables 4 -pipeline-latency 4 -sign 1 -n 7 -name r2cb_7 -include r2cb.h */ /* * This function contains 24 FP additions, 22 FP multiplications, * (or, 2 additions, 0 multiplications, 22 fused multiply/add), * 31 stack variables, 7 constants, and 14 memory accesses */ #include "r2cb.h" static void r2cb_7(R *R0, R *R1, R *Cr, R *Ci, stride rs, stride csr, stride csi, INT v, INT ivs, INT ovs) { DK(KP1_949855824, +1.949855824363647214036263365987862434465571601); DK(KP1_801937735, +1.801937735804838252472204639014890102331838324); DK(KP2_000000000, +2.000000000000000000000000000000000000000000000); DK(KP692021471, +0.692021471630095869627814897002069140197260599); DK(KP801937735, +0.801937735804838252472204639014890102331838324); DK(KP356895867, +0.356895867892209443894399510021300583399127187); DK(KP554958132, +0.554958132087371191422194871006410481067288862); { INT i; for (i = v; i > 0; i = i - 1, R0 = R0 + ovs, R1 = R1 + ovs, Cr = Cr + ivs, Ci = Ci + ivs, MAKE_VOLATILE_STRIDE(28, rs), MAKE_VOLATILE_STRIDE(28, csr), MAKE_VOLATILE_STRIDE(28, csi)) { E Tn, Td, Tg, Ti, Tl, T8; { E T1, T9, Tb, Ta, T2, T4, Th, Tm, Tc, T3, Te; T1 = Cr[0]; T9 = Ci[WS(csi, 2)]; Tb = Ci[WS(csi, 3)]; Ta = Ci[WS(csi, 1)]; T2 = Cr[WS(csr, 1)]; T4 = Cr[WS(csr, 3)]; Th = FMA(KP554958132, T9, Tb); Tm = FMS(KP554958132, Ta, T9); Tc = FMA(KP554958132, Tb, Ta); T3 = Cr[WS(csr, 2)]; Te = FNMS(KP356895867, T2, T4); Tn = FMA(KP801937735, Tm, Tb); { E Tf, Tk, T7, T5, Tj, T6; Td = FMA(KP801937735, Tc, T9); T5 = T2 + T3 + T4; Tj = FNMS(KP356895867, T4, T3); T6 = FNMS(KP356895867, T3, T2); Tf = FNMS(KP692021471, Te, T3); R0[0] = FMA(KP2_000000000, T5, T1); Tk = FNMS(KP692021471, Tj, T2); T7 = FNMS(KP692021471, T6, T4); Tg = FNMS(KP1_801937735, Tf, T1); Ti = FNMS(KP801937735, Th, Ta); Tl = FNMS(KP1_801937735, Tk, T1); T8 = FNMS(KP1_801937735, T7, T1); } } R1[WS(rs, 2)] = FMA(KP1_949855824, Ti, Tg); R0[WS(rs, 1)] = FNMS(KP1_949855824, Ti, Tg); R0[WS(rs, 2)] = FMA(KP1_949855824, Tn, Tl); R1[WS(rs, 1)] = FNMS(KP1_949855824, Tn, Tl); R0[WS(rs, 3)] = FMA(KP1_949855824, Td, T8); R1[0] = FNMS(KP1_949855824, Td, T8); } } } static const kr2c_desc desc = { 7, "r2cb_7", {2, 0, 22, 0}, &GENUS }; void X(codelet_r2cb_7) (planner *p) { X(kr2c_register) (p, r2cb_7, &desc); } #else /* HAVE_FMA */ /* Generated by: ../../../genfft/gen_r2cb.native -compact -variables 4 -pipeline-latency 4 -sign 1 -n 7 -name r2cb_7 -include r2cb.h */ /* * This function contains 24 FP additions, 19 FP multiplications, * (or, 11 additions, 6 multiplications, 13 fused multiply/add), * 21 stack variables, 7 constants, and 14 memory accesses */ #include "r2cb.h" static void r2cb_7(R *R0, R *R1, R *Cr, R *Ci, stride rs, stride csr, stride csi, INT v, INT ivs, INT ovs) { DK(KP2_000000000, +2.000000000000000000000000000000000000000000000); DK(KP1_801937735, +1.801937735804838252472204639014890102331838324); DK(KP445041867, +0.445041867912628808577805128993589518932711138); DK(KP1_246979603, +1.246979603717467061050009768008479621264549462); DK(KP867767478, +0.867767478235116240951536665696717509219981456); DK(KP1_949855824, +1.949855824363647214036263365987862434465571601); DK(KP1_563662964, +1.563662964936059617416889053348115500464669037); { INT i; for (i = v; i > 0; i = i - 1, R0 = R0 + ovs, R1 = R1 + ovs, Cr = Cr + ivs, Ci = Ci + ivs, MAKE_VOLATILE_STRIDE(28, rs), MAKE_VOLATILE_STRIDE(28, csr), MAKE_VOLATILE_STRIDE(28, csi)) { E T9, Td, Tb, T1, T4, T2, T3, T5, Tc, Ta, T6, T8, T7; T6 = Ci[WS(csi, 2)]; T8 = Ci[WS(csi, 1)]; T7 = Ci[WS(csi, 3)]; T9 = FNMS(KP1_949855824, T7, KP1_563662964 * T6) - (KP867767478 * T8); Td = FMA(KP867767478, T6, KP1_563662964 * T7) - (KP1_949855824 * T8); Tb = FMA(KP1_563662964, T8, KP1_949855824 * T6) + (KP867767478 * T7); T1 = Cr[0]; T4 = Cr[WS(csr, 3)]; T2 = Cr[WS(csr, 1)]; T3 = Cr[WS(csr, 2)]; T5 = FMA(KP1_246979603, T3, T1) + FNMA(KP445041867, T4, KP1_801937735 * T2); Tc = FMA(KP1_246979603, T4, T1) + FNMA(KP1_801937735, T3, KP445041867 * T2); Ta = FMA(KP1_246979603, T2, T1) + FNMA(KP1_801937735, T4, KP445041867 * T3); R0[WS(rs, 2)] = T5 - T9; R1[WS(rs, 1)] = T5 + T9; R0[WS(rs, 1)] = Tc + Td; R1[WS(rs, 2)] = Tc - Td; R0[WS(rs, 3)] = Ta + Tb; R1[0] = Ta - Tb; R0[0] = FMA(KP2_000000000, T2 + T3 + T4, T1); } } } static const kr2c_desc desc = { 7, "r2cb_7", {11, 6, 13, 0}, &GENUS }; void X(codelet_r2cb_7) (planner *p) { X(kr2c_register) (p, r2cb_7, &desc); } #endif /* HAVE_FMA */ fftw-3.3.3/rdft/scalar/r2cb/r2cb_8.c0000644000175000017500000001172612054410743013675 00000000000000/* * Copyright (c) 2003, 2007-11 Matteo Frigo * Copyright (c) 2003, 2007-11 Massachusetts Institute of Technology * * This program is free software; you can redistribute it and/or modify * it under the terms of the GNU General Public License as published by * the Free Software Foundation; either version 2 of the License, or * (at your option) any later version. * * This program is distributed in the hope that it will be useful, * but WITHOUT ANY WARRANTY; without even the implied warranty of * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the * GNU General Public License for more details. * * You should have received a copy of the GNU General Public License * along with this program; if not, write to the Free Software * Foundation, Inc., 51 Franklin Street, Fifth Floor, Boston, MA 02110-1301 USA * */ /* This file was automatically generated --- DO NOT EDIT */ /* Generated on Sun Nov 25 07:41:07 EST 2012 */ #include "codelet-rdft.h" #ifdef HAVE_FMA /* Generated by: ../../../genfft/gen_r2cb.native -fma -reorder-insns -schedule-for-pipeline -compact -variables 4 -pipeline-latency 4 -sign 1 -n 8 -name r2cb_8 -include r2cb.h */ /* * This function contains 20 FP additions, 12 FP multiplications, * (or, 8 additions, 0 multiplications, 12 fused multiply/add), * 19 stack variables, 2 constants, and 16 memory accesses */ #include "r2cb.h" static void r2cb_8(R *R0, R *R1, R *Cr, R *Ci, stride rs, stride csr, stride csi, INT v, INT ivs, INT ovs) { DK(KP1_414213562, +1.414213562373095048801688724209698078569671875); DK(KP2_000000000, +2.000000000000000000000000000000000000000000000); { INT i; for (i = v; i > 0; i = i - 1, R0 = R0 + ovs, R1 = R1 + ovs, Cr = Cr + ivs, Ci = Ci + ivs, MAKE_VOLATILE_STRIDE(32, rs), MAKE_VOLATILE_STRIDE(32, csr), MAKE_VOLATILE_STRIDE(32, csi)) { E Th, Tb, Tg, Ti; { E T4, Ta, Td, T9, T3, Tc, T8, Te; T4 = Cr[WS(csr, 2)]; Ta = Ci[WS(csi, 2)]; { E T1, T2, T6, T7; T1 = Cr[0]; T2 = Cr[WS(csr, 4)]; T6 = Cr[WS(csr, 1)]; T7 = Cr[WS(csr, 3)]; Td = Ci[WS(csi, 1)]; T9 = T1 - T2; T3 = T1 + T2; Tc = T6 - T7; T8 = T6 + T7; Te = Ci[WS(csi, 3)]; } { E Tj, T5, Tk, Tf; Tj = FNMS(KP2_000000000, T4, T3); T5 = FMA(KP2_000000000, T4, T3); Th = FMA(KP2_000000000, Ta, T9); Tb = FNMS(KP2_000000000, Ta, T9); Tk = Td - Te; Tf = Td + Te; R0[0] = FMA(KP2_000000000, T8, T5); R0[WS(rs, 2)] = FNMS(KP2_000000000, T8, T5); R0[WS(rs, 3)] = FMA(KP2_000000000, Tk, Tj); R0[WS(rs, 1)] = FNMS(KP2_000000000, Tk, Tj); Tg = Tc - Tf; Ti = Tc + Tf; } } R1[0] = FMA(KP1_414213562, Tg, Tb); R1[WS(rs, 2)] = FNMS(KP1_414213562, Tg, Tb); R1[WS(rs, 3)] = FMA(KP1_414213562, Ti, Th); R1[WS(rs, 1)] = FNMS(KP1_414213562, Ti, Th); } } } static const kr2c_desc desc = { 8, "r2cb_8", {8, 0, 12, 0}, &GENUS }; void X(codelet_r2cb_8) (planner *p) { X(kr2c_register) (p, r2cb_8, &desc); } #else /* HAVE_FMA */ /* Generated by: ../../../genfft/gen_r2cb.native -compact -variables 4 -pipeline-latency 4 -sign 1 -n 8 -name r2cb_8 -include r2cb.h */ /* * This function contains 20 FP additions, 6 FP multiplications, * (or, 20 additions, 6 multiplications, 0 fused multiply/add), * 21 stack variables, 2 constants, and 16 memory accesses */ #include "r2cb.h" static void r2cb_8(R *R0, R *R1, R *Cr, R *Ci, stride rs, stride csr, stride csi, INT v, INT ivs, INT ovs) { DK(KP1_414213562, +1.414213562373095048801688724209698078569671875); DK(KP2_000000000, +2.000000000000000000000000000000000000000000000); { INT i; for (i = v; i > 0; i = i - 1, R0 = R0 + ovs, R1 = R1 + ovs, Cr = Cr + ivs, Ci = Ci + ivs, MAKE_VOLATILE_STRIDE(32, rs), MAKE_VOLATILE_STRIDE(32, csr), MAKE_VOLATILE_STRIDE(32, csi)) { E T5, Tg, T3, Te, T9, Ti, Td, Tj, T6, Ta; { E T4, Tf, T1, T2; T4 = Cr[WS(csr, 2)]; T5 = KP2_000000000 * T4; Tf = Ci[WS(csi, 2)]; Tg = KP2_000000000 * Tf; T1 = Cr[0]; T2 = Cr[WS(csr, 4)]; T3 = T1 + T2; Te = T1 - T2; { E T7, T8, Tb, Tc; T7 = Cr[WS(csr, 1)]; T8 = Cr[WS(csr, 3)]; T9 = KP2_000000000 * (T7 + T8); Ti = T7 - T8; Tb = Ci[WS(csi, 1)]; Tc = Ci[WS(csi, 3)]; Td = KP2_000000000 * (Tb - Tc); Tj = Tb + Tc; } } T6 = T3 + T5; R0[WS(rs, 2)] = T6 - T9; R0[0] = T6 + T9; Ta = T3 - T5; R0[WS(rs, 1)] = Ta - Td; R0[WS(rs, 3)] = Ta + Td; { E Th, Tk, Tl, Tm; Th = Te - Tg; Tk = KP1_414213562 * (Ti - Tj); R1[WS(rs, 2)] = Th - Tk; R1[0] = Th + Tk; Tl = Te + Tg; Tm = KP1_414213562 * (Ti + Tj); R1[WS(rs, 1)] = Tl - Tm; R1[WS(rs, 3)] = Tl + Tm; } } } } static const kr2c_desc desc = { 8, "r2cb_8", {20, 6, 0, 0}, &GENUS }; void X(codelet_r2cb_8) (planner *p) { X(kr2c_register) (p, r2cb_8, &desc); } #endif /* HAVE_FMA */ fftw-3.3.3/rdft/scalar/r2cb/r2cb_9.c0000644000175000017500000001703612054410743013676 00000000000000/* * Copyright (c) 2003, 2007-11 Matteo Frigo * Copyright (c) 2003, 2007-11 Massachusetts Institute of Technology * * This program is free software; you can redistribute it and/or modify * it under the terms of the GNU General Public License as published by * the Free Software Foundation; either version 2 of the License, or * (at your option) any later version. * * This program is distributed in the hope that it will be useful, * but WITHOUT ANY WARRANTY; without even the implied warranty of * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the * GNU General Public License for more details. * * You should have received a copy of the GNU General Public License * along with this program; if not, write to the Free Software * Foundation, Inc., 51 Franklin Street, Fifth Floor, Boston, MA 02110-1301 USA * */ /* This file was automatically generated --- DO NOT EDIT */ /* Generated on Sun Nov 25 07:41:07 EST 2012 */ #include "codelet-rdft.h" #ifdef HAVE_FMA /* Generated by: ../../../genfft/gen_r2cb.native -fma -reorder-insns -schedule-for-pipeline -compact -variables 4 -pipeline-latency 4 -sign 1 -n 9 -name r2cb_9 -include r2cb.h */ /* * This function contains 32 FP additions, 24 FP multiplications, * (or, 8 additions, 0 multiplications, 24 fused multiply/add), * 40 stack variables, 12 constants, and 18 memory accesses */ #include "r2cb.h" static void r2cb_9(R *R0, R *R1, R *Cr, R *Ci, stride rs, stride csr, stride csi, INT v, INT ivs, INT ovs) { DK(KP1_326827896, +1.326827896337876792410842639271782594433726619); DK(KP1_705737063, +1.705737063904886419256501927880148143872040591); DK(KP766044443, +0.766044443118978035202392650555416673935832457); DK(KP1_532088886, +1.532088886237956070404785301110833347871664914); DK(KP984807753, +0.984807753012208059366743024589523013670643252); DK(KP1_969615506, +1.969615506024416118733486049179046027341286503); DK(KP839099631, +0.839099631177280011763127298123181364687434283); DK(KP176326980, +0.176326980708464973471090386868618986121633062); DK(KP866025403, +0.866025403784438646763723170752936183471402627); DK(KP500000000, +0.500000000000000000000000000000000000000000000); DK(KP1_732050807, +1.732050807568877293527446341505872366942805254); DK(KP2_000000000, +2.000000000000000000000000000000000000000000000); { INT i; for (i = v; i > 0; i = i - 1, R0 = R0 + ovs, R1 = R1 + ovs, Cr = Cr + ivs, Ci = Ci + ivs, MAKE_VOLATILE_STRIDE(36, rs), MAKE_VOLATILE_STRIDE(36, csr), MAKE_VOLATILE_STRIDE(36, csi)) { E T4, Th, T3, Tb, Tp, Tk, T7, Tf, Ti, Ta, T1, T2; Ta = Ci[WS(csi, 3)]; T1 = Cr[0]; T2 = Cr[WS(csr, 3)]; T4 = Cr[WS(csr, 1)]; Th = Ci[WS(csi, 1)]; { E T5, T9, T6, Td, Te; T5 = Cr[WS(csr, 4)]; T9 = T1 - T2; T3 = FMA(KP2_000000000, T2, T1); T6 = Cr[WS(csr, 2)]; Td = Ci[WS(csi, 4)]; Te = Ci[WS(csi, 2)]; Tb = FNMS(KP1_732050807, Ta, T9); Tp = FMA(KP1_732050807, Ta, T9); Tk = T6 - T5; T7 = T5 + T6; Tf = Td + Te; Ti = Td - Te; } { E Tu, To, Tt, Tn, Tc, T8; Tc = FNMS(KP500000000, T7, T4); T8 = T4 + T7; { E Tw, Tj, Tr, Tg, Tv; Tw = Ti + Th; Tj = FNMS(KP500000000, Ti, Th); Tr = FMA(KP866025403, Tf, Tc); Tg = FNMS(KP866025403, Tf, Tc); Tv = T3 - T8; R0[0] = FMA(KP2_000000000, T8, T3); { E Tq, Tl, Ts, Tm; Tq = FMA(KP866025403, Tk, Tj); Tl = FNMS(KP866025403, Tk, Tj); R0[WS(rs, 3)] = FMA(KP1_732050807, Tw, Tv); R1[WS(rs, 1)] = FNMS(KP1_732050807, Tw, Tv); Ts = FNMS(KP176326980, Tr, Tq); Tu = FMA(KP176326980, Tq, Tr); Tm = FNMS(KP839099631, Tl, Tg); To = FMA(KP839099631, Tg, Tl); R0[WS(rs, 1)] = FNMS(KP1_969615506, Ts, Tp); Tt = FMA(KP984807753, Ts, Tp); R1[0] = FMA(KP1_532088886, Tm, Tb); Tn = FNMS(KP766044443, Tm, Tb); } } R1[WS(rs, 2)] = FNMS(KP1_705737063, Tu, Tt); R0[WS(rs, 4)] = FMA(KP1_705737063, Tu, Tt); R0[WS(rs, 2)] = FNMS(KP1_326827896, To, Tn); R1[WS(rs, 3)] = FMA(KP1_326827896, To, Tn); } } } } static const kr2c_desc desc = { 9, "r2cb_9", {8, 0, 24, 0}, &GENUS }; void X(codelet_r2cb_9) (planner *p) { X(kr2c_register) (p, r2cb_9, &desc); } #else /* HAVE_FMA */ /* Generated by: ../../../genfft/gen_r2cb.native -compact -variables 4 -pipeline-latency 4 -sign 1 -n 9 -name r2cb_9 -include r2cb.h */ /* * This function contains 32 FP additions, 18 FP multiplications, * (or, 22 additions, 8 multiplications, 10 fused multiply/add), * 35 stack variables, 12 constants, and 18 memory accesses */ #include "r2cb.h" static void r2cb_9(R *R0, R *R1, R *Cr, R *Ci, stride rs, stride csr, stride csi, INT v, INT ivs, INT ovs) { DK(KP984807753, +0.984807753012208059366743024589523013670643252); DK(KP173648177, +0.173648177666930348851716626769314796000375677); DK(KP300767466, +0.300767466360870593278543795225003852144476517); DK(KP1_705737063, +1.705737063904886419256501927880148143872040591); DK(KP642787609, +0.642787609686539326322643409907263432907559884); DK(KP766044443, +0.766044443118978035202392650555416673935832457); DK(KP1_326827896, +1.326827896337876792410842639271782594433726619); DK(KP1_113340798, +1.113340798452838732905825904094046265936583811); DK(KP500000000, +0.500000000000000000000000000000000000000000000); DK(KP866025403, +0.866025403784438646763723170752936183471402627); DK(KP2_000000000, +2.000000000000000000000000000000000000000000000); DK(KP1_732050807, +1.732050807568877293527446341505872366942805254); { INT i; for (i = v; i > 0; i = i - 1, R0 = R0 + ovs, R1 = R1 + ovs, Cr = Cr + ivs, Ci = Ci + ivs, MAKE_VOLATILE_STRIDE(36, rs), MAKE_VOLATILE_STRIDE(36, csr), MAKE_VOLATILE_STRIDE(36, csi)) { E T3, Tq, Tc, Tk, Tj, T8, Tm, Ts, Th, Tr, Tw, Tx; { E Tb, T1, T2, T9, Ta; Ta = Ci[WS(csi, 3)]; Tb = KP1_732050807 * Ta; T1 = Cr[0]; T2 = Cr[WS(csr, 3)]; T9 = T1 - T2; T3 = FMA(KP2_000000000, T2, T1); Tq = T9 + Tb; Tc = T9 - Tb; } { E T4, T7, Ti, Tg, Tl, Td; T4 = Cr[WS(csr, 1)]; Tk = Ci[WS(csi, 1)]; { E T5, T6, Te, Tf; T5 = Cr[WS(csr, 4)]; T6 = Cr[WS(csr, 2)]; T7 = T5 + T6; Ti = KP866025403 * (T5 - T6); Te = Ci[WS(csi, 4)]; Tf = Ci[WS(csi, 2)]; Tg = KP866025403 * (Te + Tf); Tj = Tf - Te; } T8 = T4 + T7; Tl = FMA(KP500000000, Tj, Tk); Tm = Ti + Tl; Ts = Tl - Ti; Td = FNMS(KP500000000, T7, T4); Th = Td - Tg; Tr = Td + Tg; } R0[0] = FMA(KP2_000000000, T8, T3); Tw = T3 - T8; Tx = KP1_732050807 * (Tk - Tj); R1[WS(rs, 1)] = Tw - Tx; R0[WS(rs, 3)] = Tw + Tx; { E Tp, Tn, To, Tv, Tt, Tu; Tp = FMA(KP1_113340798, Th, KP1_326827896 * Tm); Tn = FNMS(KP642787609, Tm, KP766044443 * Th); To = Tc - Tn; R1[0] = FMA(KP2_000000000, Tn, Tc); R1[WS(rs, 3)] = To + Tp; R0[WS(rs, 2)] = To - Tp; Tv = FMA(KP1_705737063, Tr, KP300767466 * Ts); Tt = FNMS(KP984807753, Ts, KP173648177 * Tr); Tu = Tq - Tt; R0[WS(rs, 1)] = FMA(KP2_000000000, Tt, Tq); R0[WS(rs, 4)] = Tu + Tv; R1[WS(rs, 2)] = Tu - Tv; } } } } static const kr2c_desc desc = { 9, "r2cb_9", {22, 8, 10, 0}, &GENUS }; void X(codelet_r2cb_9) (planner *p) { X(kr2c_register) (p, r2cb_9, &desc); } #endif /* HAVE_FMA */ fftw-3.3.3/rdft/scalar/r2cb/r2cb_10.c0000644000175000017500000001515412054410743013745 00000000000000/* * Copyright (c) 2003, 2007-11 Matteo Frigo * Copyright (c) 2003, 2007-11 Massachusetts Institute of Technology * * This program is free software; you can redistribute it and/or modify * it under the terms of the GNU General Public License as published by * the Free Software Foundation; either version 2 of the License, or * (at your option) any later version. * * This program is distributed in the hope that it will be useful, * but WITHOUT ANY WARRANTY; without even the implied warranty of * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the * GNU General Public License for more details. * * You should have received a copy of the GNU General Public License * along with this program; if not, write to the Free Software * Foundation, Inc., 51 Franklin Street, Fifth Floor, Boston, MA 02110-1301 USA * */ /* This file was automatically generated --- DO NOT EDIT */ /* Generated on Sun Nov 25 07:41:07 EST 2012 */ #include "codelet-rdft.h" #ifdef HAVE_FMA /* Generated by: ../../../genfft/gen_r2cb.native -fma -reorder-insns -schedule-for-pipeline -compact -variables 4 -pipeline-latency 4 -sign 1 -n 10 -name r2cb_10 -include r2cb.h */ /* * This function contains 34 FP additions, 20 FP multiplications, * (or, 14 additions, 0 multiplications, 20 fused multiply/add), * 30 stack variables, 5 constants, and 20 memory accesses */ #include "r2cb.h" static void r2cb_10(R *R0, R *R1, R *Cr, R *Ci, stride rs, stride csr, stride csi, INT v, INT ivs, INT ovs) { DK(KP1_902113032, +1.902113032590307144232878666758764286811397268); DK(KP1_118033988, +1.118033988749894848204586834365638117720309180); DK(KP2_000000000, +2.000000000000000000000000000000000000000000000); DK(KP500000000, +0.500000000000000000000000000000000000000000000); DK(KP618033988, +0.618033988749894848204586834365638117720309180); { INT i; for (i = v; i > 0; i = i - 1, R0 = R0 + ovs, R1 = R1 + ovs, Cr = Cr + ivs, Ci = Ci + ivs, MAKE_VOLATILE_STRIDE(40, rs), MAKE_VOLATILE_STRIDE(40, csr), MAKE_VOLATILE_STRIDE(40, csi)) { E Tb, T3, Tc, T6, Tq, To, Ty, Tw, Td, T9; { E Tu, Tn, T7, Tv, Tk, T8; { E T1, T2, Tl, Tm; T1 = Cr[0]; T2 = Cr[WS(csr, 5)]; Tl = Ci[WS(csi, 2)]; Tm = Ci[WS(csi, 3)]; { E Ti, Tj, T4, T5; Ti = Ci[WS(csi, 4)]; Tb = T1 + T2; T3 = T1 - T2; Tu = Tl + Tm; Tn = Tl - Tm; Tj = Ci[WS(csi, 1)]; T4 = Cr[WS(csr, 2)]; T5 = Cr[WS(csr, 3)]; T7 = Cr[WS(csr, 4)]; Tv = Ti + Tj; Tk = Ti - Tj; Tc = T4 + T5; T6 = T4 - T5; T8 = Cr[WS(csr, 1)]; } } Tq = FMA(KP618033988, Tk, Tn); To = FNMS(KP618033988, Tn, Tk); Ty = FNMS(KP618033988, Tu, Tv); Tw = FMA(KP618033988, Tv, Tu); Td = T7 + T8; T9 = T7 - T8; } { E Te, Tg, Ta, Ts, Tf, Tr; Te = Tc + Td; Tg = Tc - Td; Ta = T6 + T9; Ts = T6 - T9; Tf = FNMS(KP500000000, Te, Tb); R0[0] = FMA(KP2_000000000, Te, Tb); Tr = FNMS(KP500000000, Ta, T3); R1[WS(rs, 2)] = FMA(KP2_000000000, Ta, T3); { E Th, Tp, Tt, Tx; Th = FNMS(KP1_118033988, Tg, Tf); Tp = FMA(KP1_118033988, Tg, Tf); Tt = FMA(KP1_118033988, Ts, Tr); Tx = FNMS(KP1_118033988, Ts, Tr); R0[WS(rs, 3)] = FNMS(KP1_902113032, Tq, Tp); R0[WS(rs, 2)] = FMA(KP1_902113032, Tq, Tp); R0[WS(rs, 1)] = FMA(KP1_902113032, To, Th); R0[WS(rs, 4)] = FNMS(KP1_902113032, To, Th); R1[WS(rs, 1)] = FNMS(KP1_902113032, Ty, Tx); R1[WS(rs, 3)] = FMA(KP1_902113032, Ty, Tx); R1[WS(rs, 4)] = FMA(KP1_902113032, Tw, Tt); R1[0] = FNMS(KP1_902113032, Tw, Tt); } } } } } static const kr2c_desc desc = { 10, "r2cb_10", {14, 0, 20, 0}, &GENUS }; void X(codelet_r2cb_10) (planner *p) { X(kr2c_register) (p, r2cb_10, &desc); } #else /* HAVE_FMA */ /* Generated by: ../../../genfft/gen_r2cb.native -compact -variables 4 -pipeline-latency 4 -sign 1 -n 10 -name r2cb_10 -include r2cb.h */ /* * This function contains 34 FP additions, 14 FP multiplications, * (or, 26 additions, 6 multiplications, 8 fused multiply/add), * 26 stack variables, 5 constants, and 20 memory accesses */ #include "r2cb.h" static void r2cb_10(R *R0, R *R1, R *Cr, R *Ci, stride rs, stride csr, stride csi, INT v, INT ivs, INT ovs) { DK(KP500000000, +0.500000000000000000000000000000000000000000000); DK(KP1_902113032, +1.902113032590307144232878666758764286811397268); DK(KP1_175570504, +1.175570504584946258337411909278145537195304875); DK(KP2_000000000, +2.000000000000000000000000000000000000000000000); DK(KP1_118033988, +1.118033988749894848204586834365638117720309180); { INT i; for (i = v; i > 0; i = i - 1, R0 = R0 + ovs, R1 = R1 + ovs, Cr = Cr + ivs, Ci = Ci + ivs, MAKE_VOLATILE_STRIDE(40, rs), MAKE_VOLATILE_STRIDE(40, csr), MAKE_VOLATILE_STRIDE(40, csi)) { E T3, Tb, Tn, Tv, Tk, Tu, Ta, Ts, Te, Tg, Ti, Tj; { E T1, T2, Tl, Tm; T1 = Cr[0]; T2 = Cr[WS(csr, 5)]; T3 = T1 - T2; Tb = T1 + T2; Tl = Ci[WS(csi, 4)]; Tm = Ci[WS(csi, 1)]; Tn = Tl - Tm; Tv = Tl + Tm; } Ti = Ci[WS(csi, 2)]; Tj = Ci[WS(csi, 3)]; Tk = Ti - Tj; Tu = Ti + Tj; { E T6, Tc, T9, Td; { E T4, T5, T7, T8; T4 = Cr[WS(csr, 2)]; T5 = Cr[WS(csr, 3)]; T6 = T4 - T5; Tc = T4 + T5; T7 = Cr[WS(csr, 4)]; T8 = Cr[WS(csr, 1)]; T9 = T7 - T8; Td = T7 + T8; } Ta = T6 + T9; Ts = KP1_118033988 * (T6 - T9); Te = Tc + Td; Tg = KP1_118033988 * (Tc - Td); } R1[WS(rs, 2)] = FMA(KP2_000000000, Ta, T3); R0[0] = FMA(KP2_000000000, Te, Tb); { E To, Tq, Th, Tp, Tf; To = FNMS(KP1_902113032, Tn, KP1_175570504 * Tk); Tq = FMA(KP1_902113032, Tk, KP1_175570504 * Tn); Tf = FNMS(KP500000000, Te, Tb); Th = Tf - Tg; Tp = Tg + Tf; R0[WS(rs, 1)] = Th - To; R0[WS(rs, 2)] = Tp + Tq; R0[WS(rs, 4)] = Th + To; R0[WS(rs, 3)] = Tp - Tq; } { E Tw, Ty, Tt, Tx, Tr; Tw = FNMS(KP1_902113032, Tv, KP1_175570504 * Tu); Ty = FMA(KP1_902113032, Tu, KP1_175570504 * Tv); Tr = FNMS(KP500000000, Ta, T3); Tt = Tr - Ts; Tx = Ts + Tr; R1[WS(rs, 3)] = Tt - Tw; R1[WS(rs, 4)] = Tx + Ty; R1[WS(rs, 1)] = Tt + Tw; R1[0] = Tx - Ty; } } } } static const kr2c_desc desc = { 10, "r2cb_10", {26, 6, 8, 0}, &GENUS }; void X(codelet_r2cb_10) (planner *p) { X(kr2c_register) (p, r2cb_10, &desc); } #endif /* HAVE_FMA */ fftw-3.3.3/rdft/scalar/r2cb/r2cb_11.c0000644000175000017500000002246012054410745013746 00000000000000/* * Copyright (c) 2003, 2007-11 Matteo Frigo * Copyright (c) 2003, 2007-11 Massachusetts Institute of Technology * * This program is free software; you can redistribute it and/or modify * it under the terms of the GNU General Public License as published by * the Free Software Foundation; either version 2 of the License, or * (at your option) any later version. * * This program is distributed in the hope that it will be useful, * but WITHOUT ANY WARRANTY; without even the implied warranty of * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the * GNU General Public License for more details. * * You should have received a copy of the GNU General Public License * along with this program; if not, write to the Free Software * Foundation, Inc., 51 Franklin Street, Fifth Floor, Boston, MA 02110-1301 USA * */ /* This file was automatically generated --- DO NOT EDIT */ /* Generated on Sun Nov 25 07:41:07 EST 2012 */ #include "codelet-rdft.h" #ifdef HAVE_FMA /* Generated by: ../../../genfft/gen_r2cb.native -fma -reorder-insns -schedule-for-pipeline -compact -variables 4 -pipeline-latency 4 -sign 1 -n 11 -name r2cb_11 -include r2cb.h */ /* * This function contains 60 FP additions, 56 FP multiplications, * (or, 4 additions, 0 multiplications, 56 fused multiply/add), * 53 stack variables, 11 constants, and 22 memory accesses */ #include "r2cb.h" static void r2cb_11(R *R0, R *R1, R *Cr, R *Ci, stride rs, stride csr, stride csi, INT v, INT ivs, INT ovs) { DK(KP1_979642883, +1.979642883761865464752184075553437574753038744); DK(KP1_918985947, +1.918985947228994779780736114132655398124909697); DK(KP876768831, +0.876768831002589333891339807079336796764054852); DK(KP918985947, +0.918985947228994779780736114132655398124909697); DK(KP778434453, +0.778434453334651800608337670740821884709317477); DK(KP2_000000000, +2.000000000000000000000000000000000000000000000); DK(KP634356270, +0.634356270682424498893150776899916060542806975); DK(KP342584725, +0.342584725681637509502641509861112333758894680); DK(KP830830026, +0.830830026003772851058548298459246407048009821); DK(KP715370323, +0.715370323453429719112414662767260662417897278); DK(KP521108558, +0.521108558113202722944698153526659300680427422); { INT i; for (i = v; i > 0; i = i - 1, R0 = R0 + ovs, R1 = R1 + ovs, Cr = Cr + ivs, Ci = Ci + ivs, MAKE_VOLATILE_STRIDE(44, rs), MAKE_VOLATILE_STRIDE(44, csr), MAKE_VOLATILE_STRIDE(44, csi)) { E Tf, Tq, Tt, Tu; { E T1, Td, Th, Te, Tg, T2, Ts, TK, TB, TT, Tj, T6, T3, T4, T5; E Tr; T1 = Cr[0]; Td = Ci[WS(csi, 3)]; Th = Ci[WS(csi, 5)]; Te = Ci[WS(csi, 2)]; Tf = Ci[WS(csi, 4)]; Tg = Ci[WS(csi, 1)]; Tr = FMA(KP521108558, Td, Th); T2 = Cr[WS(csr, 1)]; { E TJ, TA, TS, Ti; TJ = FMA(KP521108558, Tf, Td); TA = FNMS(KP521108558, Te, Tf); TS = FMS(KP521108558, Tg, Te); Ti = FMA(KP521108558, Th, Tg); Ts = FNMS(KP715370323, Tr, Te); TK = FMA(KP715370323, TJ, Tg); TB = FMA(KP715370323, TA, Th); TT = FMA(KP715370323, TS, Td); Tj = FMA(KP715370323, Ti, Tf); T6 = Cr[WS(csr, 5)]; } T3 = Cr[WS(csr, 2)]; T4 = Cr[WS(csr, 3)]; T5 = Cr[WS(csr, 4)]; { E TG, Tx, To, Tl, Tb, TU, TQ, TP, Ta; { E Tk, TE, Tv, T8; Tk = FMA(KP830830026, Tj, Te); TE = FNMS(KP342584725, T3, T6); Tv = FNMS(KP342584725, T2, T4); T8 = FNMS(KP342584725, T4, T3); { E T7, Tm, TN, TF; T7 = T2 + T3 + T4 + T5 + T6; Tm = FNMS(KP342584725, T5, T2); TN = FNMS(KP342584725, T6, T5); TF = FNMS(KP634356270, TE, T2); { E Tw, T9, Tn, TO; Tw = FNMS(KP634356270, Tv, T6); T9 = FNMS(KP634356270, T8, T5); R0[0] = FMA(KP2_000000000, T7, T1); Tn = FNMS(KP634356270, Tm, T3); TO = FNMS(KP634356270, TN, T4); TG = FNMS(KP778434453, TF, T4); Tx = FNMS(KP778434453, Tw, T5); Ta = FNMS(KP778434453, T9, T2); To = FNMS(KP778434453, Tn, T6); TP = FNMS(KP778434453, TO, T3); Tl = FMA(KP918985947, Tk, Td); } } } Tb = FNMS(KP876768831, Ta, T6); TU = FNMS(KP830830026, TT, Tf); TQ = FNMS(KP876768831, TP, T2); { E TI, TL, Ty, TC; { E Tc, TV, TR, TH; TH = FNMS(KP876768831, TG, T5); Tc = FNMS(KP1_918985947, Tb, T1); TV = FNMS(KP918985947, TU, Th); TR = FNMS(KP1_918985947, TQ, T1); TI = FNMS(KP1_918985947, TH, T1); R0[WS(rs, 5)] = FMA(KP1_979642883, Tl, Tc); R1[0] = FNMS(KP1_979642883, Tl, Tc); R0[WS(rs, 3)] = FMA(KP1_979642883, TV, TR); R1[WS(rs, 2)] = FNMS(KP1_979642883, TV, TR); TL = FNMS(KP830830026, TK, Th); } Ty = FNMS(KP876768831, Tx, T3); TC = FNMS(KP830830026, TB, Td); { E TM, Tz, TD, Tp; Tp = FNMS(KP876768831, To, T4); TM = FMA(KP918985947, TL, Te); Tz = FNMS(KP1_918985947, Ty, T1); TD = FNMS(KP918985947, TC, Tg); Tq = FNMS(KP1_918985947, Tp, T1); R0[WS(rs, 2)] = FMA(KP1_979642883, TM, TI); R1[WS(rs, 3)] = FNMS(KP1_979642883, TM, TI); R0[WS(rs, 4)] = FMA(KP1_979642883, TD, Tz); R1[WS(rs, 1)] = FNMS(KP1_979642883, TD, Tz); Tt = FMA(KP830830026, Ts, Tg); } } } } Tu = FNMS(KP918985947, Tt, Tf); R0[WS(rs, 1)] = FMA(KP1_979642883, Tu, Tq); R1[WS(rs, 4)] = FNMS(KP1_979642883, Tu, Tq); } } } static const kr2c_desc desc = { 11, "r2cb_11", {4, 0, 56, 0}, &GENUS }; void X(codelet_r2cb_11) (planner *p) { X(kr2c_register) (p, r2cb_11, &desc); } #else /* HAVE_FMA */ /* Generated by: ../../../genfft/gen_r2cb.native -compact -variables 4 -pipeline-latency 4 -sign 1 -n 11 -name r2cb_11 -include r2cb.h */ /* * This function contains 60 FP additions, 51 FP multiplications, * (or, 19 additions, 10 multiplications, 41 fused multiply/add), * 33 stack variables, 11 constants, and 22 memory accesses */ #include "r2cb.h" static void r2cb_11(R *R0, R *R1, R *Cr, R *Ci, stride rs, stride csr, stride csi, INT v, INT ivs, INT ovs) { DK(KP2_000000000, +2.000000000000000000000000000000000000000000000); DK(KP1_918985947, +1.918985947228994779780736114132655398124909697); DK(KP1_309721467, +1.309721467890570128113850144932587106367582399); DK(KP284629676, +0.284629676546570280887585337232739337582102722); DK(KP830830026, +0.830830026003772851058548298459246407048009821); DK(KP1_682507065, +1.682507065662362337723623297838735435026584997); DK(KP563465113, +0.563465113682859395422835830693233798071555798); DK(KP1_511499148, +1.511499148708516567548071687944688840359434890); DK(KP1_979642883, +1.979642883761865464752184075553437574753038744); DK(KP1_819263990, +1.819263990709036742823430766158056920120482102); DK(KP1_081281634, +1.081281634911195164215271908637383390863541216); { INT i; for (i = v; i > 0; i = i - 1, R0 = R0 + ovs, R1 = R1 + ovs, Cr = Cr + ivs, Ci = Ci + ivs, MAKE_VOLATILE_STRIDE(44, rs), MAKE_VOLATILE_STRIDE(44, csr), MAKE_VOLATILE_STRIDE(44, csi)) { E Td, Tl, Tf, Th, Tj, T1, T2, T6, T5, T4, T3, T7, Tk, Te, Tg; E Ti; { E T8, Tc, T9, Ta, Tb; T8 = Ci[WS(csi, 2)]; Tc = Ci[WS(csi, 1)]; T9 = Ci[WS(csi, 4)]; Ta = Ci[WS(csi, 5)]; Tb = Ci[WS(csi, 3)]; Td = FMA(KP1_081281634, T8, KP1_819263990 * T9) + FNMA(KP1_979642883, Ta, KP1_511499148 * Tb) - (KP563465113 * Tc); Tl = FMA(KP1_979642883, T8, KP1_819263990 * Ta) + FNMA(KP563465113, T9, KP1_081281634 * Tb) - (KP1_511499148 * Tc); Tf = FMA(KP563465113, T8, KP1_819263990 * Tb) + FNMA(KP1_511499148, Ta, KP1_081281634 * T9) - (KP1_979642883 * Tc); Th = FMA(KP1_081281634, Tc, KP1_819263990 * T8) + FMA(KP1_979642883, Tb, KP1_511499148 * T9) + (KP563465113 * Ta); Tj = FMA(KP563465113, Tb, KP1_979642883 * T9) + FNMS(KP1_511499148, T8, KP1_081281634 * Ta) - (KP1_819263990 * Tc); } T1 = Cr[0]; T2 = Cr[WS(csr, 1)]; T6 = Cr[WS(csr, 5)]; T5 = Cr[WS(csr, 4)]; T4 = Cr[WS(csr, 3)]; T3 = Cr[WS(csr, 2)]; T7 = FMA(KP1_682507065, T3, T1) + FNMS(KP284629676, T6, KP830830026 * T5) + FNMA(KP1_309721467, T4, KP1_918985947 * T2); Tk = FMA(KP1_682507065, T4, T1) + FNMS(KP1_918985947, T5, KP830830026 * T6) + FNMA(KP284629676, T3, KP1_309721467 * T2); Te = FMA(KP830830026, T4, T1) + FNMS(KP1_309721467, T6, KP1_682507065 * T5) + FNMA(KP1_918985947, T3, KP284629676 * T2); Tg = FMA(KP1_682507065, T2, T1) + FNMS(KP1_918985947, T6, KP830830026 * T3) + FNMA(KP1_309721467, T5, KP284629676 * T4); Ti = FMA(KP830830026, T2, T1) + FNMS(KP284629676, T5, KP1_682507065 * T6) + FNMA(KP1_918985947, T4, KP1_309721467 * T3); R0[WS(rs, 3)] = T7 - Td; R0[WS(rs, 4)] = Te - Tf; R0[WS(rs, 2)] = Tk + Tl; R1[WS(rs, 2)] = T7 + Td; R1[WS(rs, 3)] = Tk - Tl; R0[WS(rs, 1)] = Ti + Tj; R1[WS(rs, 1)] = Te + Tf; R0[WS(rs, 5)] = Tg + Th; R1[0] = Tg - Th; R1[WS(rs, 4)] = Ti - Tj; R0[0] = FMA(KP2_000000000, T2 + T3 + T4 + T5 + T6, T1); } } } static const kr2c_desc desc = { 11, "r2cb_11", {19, 10, 41, 0}, &GENUS }; void X(codelet_r2cb_11) (planner *p) { X(kr2c_register) (p, r2cb_11, &desc); } #endif /* HAVE_FMA */ fftw-3.3.3/rdft/scalar/r2cb/r2cb_12.c0000644000175000017500000001457312054410744013754 00000000000000/* * Copyright (c) 2003, 2007-11 Matteo Frigo * Copyright (c) 2003, 2007-11 Massachusetts Institute of Technology * * This program is free software; you can redistribute it and/or modify * it under the terms of the GNU General Public License as published by * the Free Software Foundation; either version 2 of the License, or * (at your option) any later version. * * This program is distributed in the hope that it will be useful, * but WITHOUT ANY WARRANTY; without even the implied warranty of * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the * GNU General Public License for more details. * * You should have received a copy of the GNU General Public License * along with this program; if not, write to the Free Software * Foundation, Inc., 51 Franklin Street, Fifth Floor, Boston, MA 02110-1301 USA * */ /* This file was automatically generated --- DO NOT EDIT */ /* Generated on Sun Nov 25 07:41:07 EST 2012 */ #include "codelet-rdft.h" #ifdef HAVE_FMA /* Generated by: ../../../genfft/gen_r2cb.native -fma -reorder-insns -schedule-for-pipeline -compact -variables 4 -pipeline-latency 4 -sign 1 -n 12 -name r2cb_12 -include r2cb.h */ /* * This function contains 38 FP additions, 16 FP multiplications, * (or, 22 additions, 0 multiplications, 16 fused multiply/add), * 31 stack variables, 2 constants, and 24 memory accesses */ #include "r2cb.h" static void r2cb_12(R *R0, R *R1, R *Cr, R *Ci, stride rs, stride csr, stride csi, INT v, INT ivs, INT ovs) { DK(KP1_732050807, +1.732050807568877293527446341505872366942805254); DK(KP2_000000000, +2.000000000000000000000000000000000000000000000); { INT i; for (i = v; i > 0; i = i - 1, R0 = R0 + ovs, R1 = R1 + ovs, Cr = Cr + ivs, Ci = Ci + ivs, MAKE_VOLATILE_STRIDE(48, rs), MAKE_VOLATILE_STRIDE(48, csr), MAKE_VOLATILE_STRIDE(48, csi)) { E Ts, Tr; { E Tz, Te, Tn, Tk, Tc, Tw, Ty, Th, T4, T3, Td, T5; { E T8, Tu, Tl, Tm, Tb, T9, Ta, T1, T2, Tv; T8 = Cr[WS(csr, 3)]; T9 = Cr[WS(csr, 5)]; Ta = Cr[WS(csr, 1)]; Tu = Ci[WS(csi, 3)]; Tl = Ci[WS(csi, 5)]; Tm = Ci[WS(csi, 1)]; Tb = T9 + Ta; Tz = T9 - Ta; Te = Ci[WS(csi, 4)]; Tn = Tl - Tm; Tv = Tl + Tm; Tk = FNMS(KP2_000000000, T8, Tb); Tc = T8 + Tb; T1 = Cr[0]; T2 = Cr[WS(csr, 4)]; Tw = Tu - Tv; Ty = FMA(KP2_000000000, Tu, Tv); Th = Ci[WS(csi, 2)]; T4 = Cr[WS(csr, 6)]; T3 = FMA(KP2_000000000, T2, T1); Td = T1 - T2; T5 = Cr[WS(csr, 2)]; } { E To, Tp, Tf, Tg, T6, TA, TC; To = FMA(KP1_732050807, Tn, Tk); Ts = FNMS(KP1_732050807, Tn, Tk); Tp = FNMS(KP1_732050807, Te, Td); Tf = FMA(KP1_732050807, Te, Td); Tg = T4 - T5; T6 = FMA(KP2_000000000, T5, T4); TA = FMA(KP1_732050807, Tz, Ty); TC = FNMS(KP1_732050807, Tz, Ty); { E Tt, T7, Ti, Tq, Tj, TB, Tx; Tt = T3 - T6; T7 = T3 + T6; Ti = FNMS(KP1_732050807, Th, Tg); Tq = FMA(KP1_732050807, Th, Tg); R0[0] = FMA(KP2_000000000, Tc, T7); R0[WS(rs, 3)] = FNMS(KP2_000000000, Tc, T7); Tj = Tf + Ti; TB = Tf - Ti; Tr = Tp + Tq; Tx = Tp - Tq; R1[WS(rs, 5)] = TB + TC; R1[WS(rs, 2)] = TB - TC; R0[WS(rs, 4)] = Tj - To; R0[WS(rs, 1)] = Tj + To; R1[WS(rs, 3)] = Tx + TA; R1[0] = Tx - TA; R1[WS(rs, 4)] = FNMS(KP2_000000000, Tw, Tt); R1[WS(rs, 1)] = FMA(KP2_000000000, Tw, Tt); } } } R0[WS(rs, 2)] = Tr - Ts; R0[WS(rs, 5)] = Tr + Ts; } } } static const kr2c_desc desc = { 12, "r2cb_12", {22, 0, 16, 0}, &GENUS }; void X(codelet_r2cb_12) (planner *p) { X(kr2c_register) (p, r2cb_12, &desc); } #else /* HAVE_FMA */ /* Generated by: ../../../genfft/gen_r2cb.native -compact -variables 4 -pipeline-latency 4 -sign 1 -n 12 -name r2cb_12 -include r2cb.h */ /* * This function contains 38 FP additions, 10 FP multiplications, * (or, 34 additions, 6 multiplications, 4 fused multiply/add), * 25 stack variables, 2 constants, and 24 memory accesses */ #include "r2cb.h" static void r2cb_12(R *R0, R *R1, R *Cr, R *Ci, stride rs, stride csr, stride csi, INT v, INT ivs, INT ovs) { DK(KP1_732050807, +1.732050807568877293527446341505872366942805254); DK(KP2_000000000, +2.000000000000000000000000000000000000000000000); { INT i; for (i = v; i > 0; i = i - 1, R0 = R0 + ovs, R1 = R1 + ovs, Cr = Cr + ivs, Ci = Ci + ivs, MAKE_VOLATILE_STRIDE(48, rs), MAKE_VOLATILE_STRIDE(48, csr), MAKE_VOLATILE_STRIDE(48, csi)) { E T8, Tb, Tm, TA, Tw, Tx, Tp, TB, T3, Tr, Tg, T6, Ts, Tk; { E T9, Ta, Tn, To; T8 = Cr[WS(csr, 3)]; T9 = Cr[WS(csr, 5)]; Ta = Cr[WS(csr, 1)]; Tb = T9 + Ta; Tm = FMS(KP2_000000000, T8, Tb); TA = KP1_732050807 * (T9 - Ta); Tw = Ci[WS(csi, 3)]; Tn = Ci[WS(csi, 5)]; To = Ci[WS(csi, 1)]; Tx = Tn + To; Tp = KP1_732050807 * (Tn - To); TB = FMA(KP2_000000000, Tw, Tx); } { E Tf, T1, T2, Td, Te; Te = Ci[WS(csi, 4)]; Tf = KP1_732050807 * Te; T1 = Cr[0]; T2 = Cr[WS(csr, 4)]; Td = T1 - T2; T3 = FMA(KP2_000000000, T2, T1); Tr = Td - Tf; Tg = Td + Tf; } { E Tj, T4, T5, Th, Ti; Ti = Ci[WS(csi, 2)]; Tj = KP1_732050807 * Ti; T4 = Cr[WS(csr, 6)]; T5 = Cr[WS(csr, 2)]; Th = T4 - T5; T6 = FMA(KP2_000000000, T5, T4); Ts = Th + Tj; Tk = Th - Tj; } { E T7, Tc, Tz, TC; T7 = T3 + T6; Tc = KP2_000000000 * (T8 + Tb); R0[WS(rs, 3)] = T7 - Tc; R0[0] = T7 + Tc; { E Tl, Tq, TD, TE; Tl = Tg + Tk; Tq = Tm - Tp; R0[WS(rs, 1)] = Tl - Tq; R0[WS(rs, 4)] = Tl + Tq; TD = Tg - Tk; TE = TB - TA; R1[WS(rs, 2)] = TD - TE; R1[WS(rs, 5)] = TD + TE; } Tz = Tr - Ts; TC = TA + TB; R1[0] = Tz - TC; R1[WS(rs, 3)] = Tz + TC; { E Tv, Ty, Tt, Tu; Tv = T3 - T6; Ty = KP2_000000000 * (Tw - Tx); R1[WS(rs, 4)] = Tv - Ty; R1[WS(rs, 1)] = Tv + Ty; Tt = Tr + Ts; Tu = Tm + Tp; R0[WS(rs, 5)] = Tt - Tu; R0[WS(rs, 2)] = Tt + Tu; } } } } } static const kr2c_desc desc = { 12, "r2cb_12", {34, 6, 4, 0}, &GENUS }; void X(codelet_r2cb_12) (planner *p) { X(kr2c_register) (p, r2cb_12, &desc); } #endif /* HAVE_FMA */ fftw-3.3.3/rdft/scalar/r2cb/r2cb_13.c0000644000175000017500000003117112054410745013747 00000000000000/* * Copyright (c) 2003, 2007-11 Matteo Frigo * Copyright (c) 2003, 2007-11 Massachusetts Institute of Technology * * This program is free software; you can redistribute it and/or modify * it under the terms of the GNU General Public License as published by * the Free Software Foundation; either version 2 of the License, or * (at your option) any later version. * * This program is distributed in the hope that it will be useful, * but WITHOUT ANY WARRANTY; without even the implied warranty of * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the * GNU General Public License for more details. * * You should have received a copy of the GNU General Public License * along with this program; if not, write to the Free Software * Foundation, Inc., 51 Franklin Street, Fifth Floor, Boston, MA 02110-1301 USA * */ /* This file was automatically generated --- DO NOT EDIT */ /* Generated on Sun Nov 25 07:41:08 EST 2012 */ #include "codelet-rdft.h" #ifdef HAVE_FMA /* Generated by: ../../../genfft/gen_r2cb.native -fma -reorder-insns -schedule-for-pipeline -compact -variables 4 -pipeline-latency 4 -sign 1 -n 13 -name r2cb_13 -include r2cb.h */ /* * This function contains 76 FP additions, 58 FP multiplications, * (or, 18 additions, 0 multiplications, 58 fused multiply/add), * 76 stack variables, 26 constants, and 26 memory accesses */ #include "r2cb.h" static void r2cb_13(R *R0, R *R1, R *Cr, R *Ci, stride rs, stride csr, stride csi, INT v, INT ivs, INT ovs) { DK(KP968287244, +0.968287244361984016049539446938120421179794516); DK(KP875502302, +0.875502302409147941146295545768755143177842006); DK(KP1_150281458, +1.150281458948006242736771094910906776922003215); DK(KP1_040057143, +1.040057143777729238234261000998465604986476278); DK(KP1_200954543, +1.200954543865330565851538506669526018704025697); DK(KP769338817, +0.769338817572980603471413688209101117038278899); DK(KP600925212, +0.600925212577331548853203544578415991041882762); DK(KP1_033041561, +1.033041561246979445681802577138034271410067244); DK(KP1_007074065, +1.007074065727533254493747707736933954186697125); DK(KP503537032, +0.503537032863766627246873853868466977093348562); DK(KP581704778, +0.581704778510515730456870384989698884939833902); DK(KP859542535, +0.859542535098774820163672132761689612766401925); DK(KP166666666, +0.166666666666666666666666666666666666666666667); DK(KP2_000000000, +2.000000000000000000000000000000000000000000000); DK(KP301479260, +0.301479260047709873958013540496673347309208464); DK(KP226109445, +0.226109445035782405468510155372505010481906348); DK(KP686558370, +0.686558370781754340655719594850823015421401653); DK(KP514918778, +0.514918778086315755491789696138117261566051239); DK(KP957805992, +0.957805992594665126462521754605754580515587217); DK(KP522026385, +0.522026385161275033714027226654165028300441940); DK(KP853480001, +0.853480001859823990758994934970528322872359049); DK(KP038632954, +0.038632954644348171955506895830342264440241080); DK(KP612264650, +0.612264650376756543746494474777125408779395514); DK(KP302775637, +0.302775637731994646559610633735247973125648287); DK(KP866025403, +0.866025403784438646763723170752936183471402627); DK(KP500000000, +0.500000000000000000000000000000000000000000000); { INT i; for (i = v; i > 0; i = i - 1, R0 = R0 + ovs, R1 = R1 + ovs, Cr = Cr + ivs, Ci = Ci + ivs, MAKE_VOLATILE_STRIDE(52, rs), MAKE_VOLATILE_STRIDE(52, csr), MAKE_VOLATILE_STRIDE(52, csi)) { E TW, T14, TS, TO, T18, T1e, TY, TX, TQ, Tq, TP, Tl, T1d, Tr; { E T1, TN, T16, TJ, TV, TG, TU, Tf, T2, T3, Tb, Ti, T4; { E Ts, TB, Tx, Ty, Tv, TE, Tt, Tu, Tz, TC; Ts = Ci[WS(csi, 5)]; Tt = Ci[WS(csi, 2)]; Tu = Ci[WS(csi, 6)]; TB = Ci[WS(csi, 1)]; Tx = Ci[WS(csi, 3)]; Ty = Ci[WS(csi, 4)]; Tv = Tt + Tu; TE = Tu - Tt; T1 = Cr[0]; Tz = Tx + Ty; TC = Tx - Ty; { E TL, Tw, T7, Ta; TL = Ts + Tv; Tw = FNMS(KP500000000, Tv, Ts); T7 = Cr[WS(csr, 5)]; { E TD, TM, TA, TH; TD = FNMS(KP500000000, TC, TB); TM = TB + TC; TA = FMA(KP866025403, Tz, Tw); TH = FNMS(KP866025403, Tz, Tw); TN = FMA(KP302775637, TM, TL); T16 = FNMS(KP302775637, TL, TM); { E TF, TI, T8, T9; TF = FMA(KP866025403, TE, TD); TI = FNMS(KP866025403, TE, TD); T8 = Cr[WS(csr, 2)]; T9 = Cr[WS(csr, 6)]; TJ = FNMS(KP612264650, TI, TH); TV = FMA(KP612264650, TH, TI); TG = FNMS(KP038632954, TF, TA); TU = FMA(KP038632954, TA, TF); Tf = T8 - T9; Ta = T8 + T9; } } T2 = Cr[WS(csr, 1)]; T3 = Cr[WS(csr, 3)]; Tb = T7 + Ta; Ti = FMS(KP500000000, Ta, T7); T4 = Cr[WS(csr, 4)]; } } { E T17, TK, T5, Te, Tk, Td; TW = FMA(KP853480001, TV, TU); T17 = FNMS(KP853480001, TV, TU); TK = FNMS(KP853480001, TJ, TG); T14 = FMA(KP853480001, TJ, TG); T5 = T3 + T4; Te = T3 - T4; { E Tn, Tg, Th, T6; TS = FNMS(KP522026385, TK, TN); TO = FMA(KP957805992, TN, TK); Tn = Te - Tf; Tg = Te + Tf; Th = FNMS(KP500000000, T5, T2); T6 = T2 + T5; T18 = FNMS(KP522026385, T17, T16); T1e = FMA(KP957805992, T16, T17); { E Tm, Tj, Tc, Tp, To; Tm = Th + Ti; Tj = Th - Ti; Tc = T6 + Tb; Tp = T6 - Tb; To = FNMS(KP514918778, Tn, Tm); TY = FMA(KP686558370, Tm, Tn); TX = FNMS(KP226109445, Tg, Tj); Tk = FMA(KP301479260, Tj, Tg); R0[0] = FMA(KP2_000000000, Tc, T1); Td = FNMS(KP166666666, Tc, T1); TQ = FNMS(KP859542535, To, Tp); Tq = FMA(KP581704778, Tp, To); } } TP = FNMS(KP503537032, Tk, Td); Tl = FMA(KP1_007074065, Tk, Td); } } T1d = FNMS(KP1_033041561, Tq, Tl); Tr = FMA(KP1_033041561, Tq, Tl); { E T13, TR, T19, TZ; T13 = FNMS(KP600925212, TQ, TP); TR = FMA(KP600925212, TQ, TP); T19 = FMA(KP769338817, TY, TX); TZ = FNMS(KP769338817, TY, TX); R0[WS(rs, 4)] = FMA(KP1_200954543, T1e, T1d); R1[WS(rs, 2)] = FNMS(KP1_200954543, T1e, T1d); R0[WS(rs, 6)] = FMA(KP1_200954543, TO, Tr); R1[0] = FNMS(KP1_200954543, TO, Tr); { E T1b, T15, T11, TT; T1b = FNMS(KP1_040057143, T14, T13); T15 = FMA(KP1_040057143, T14, T13); T11 = FMA(KP1_150281458, TS, TR); TT = FNMS(KP1_150281458, TS, TR); { E T1c, T1a, T12, T10; T1c = FMA(KP875502302, T19, T18); T1a = FNMS(KP875502302, T19, T18); T12 = FMA(KP968287244, TZ, TW); T10 = FNMS(KP968287244, TZ, TW); R1[WS(rs, 5)] = FMA(KP1_150281458, T1c, T1b); R0[WS(rs, 3)] = FNMS(KP1_150281458, T1c, T1b); R1[WS(rs, 3)] = FMA(KP1_150281458, T1a, T15); R0[WS(rs, 1)] = FNMS(KP1_150281458, T1a, T15); R0[WS(rs, 5)] = FMA(KP1_040057143, T12, T11); R0[WS(rs, 2)] = FNMS(KP1_040057143, T12, T11); R1[WS(rs, 4)] = FMA(KP1_040057143, T10, TT); R1[WS(rs, 1)] = FNMS(KP1_040057143, T10, TT); } } } } } } static const kr2c_desc desc = { 13, "r2cb_13", {18, 0, 58, 0}, &GENUS }; void X(codelet_r2cb_13) (planner *p) { X(kr2c_register) (p, r2cb_13, &desc); } #else /* HAVE_FMA */ /* Generated by: ../../../genfft/gen_r2cb.native -compact -variables 4 -pipeline-latency 4 -sign 1 -n 13 -name r2cb_13 -include r2cb.h */ /* * This function contains 76 FP additions, 35 FP multiplications, * (or, 56 additions, 15 multiplications, 20 fused multiply/add), * 56 stack variables, 19 constants, and 26 memory accesses */ #include "r2cb.h" static void r2cb_13(R *R0, R *R1, R *Cr, R *Ci, stride rs, stride csr, stride csi, INT v, INT ivs, INT ovs) { DK(KP1_007074065, +1.007074065727533254493747707736933954186697125); DK(KP227708958, +0.227708958111581597949308691735310621069285120); DK(KP531932498, +0.531932498429674575175042127684371897596660533); DK(KP774781170, +0.774781170935234584261351932853525703557550433); DK(KP265966249, +0.265966249214837287587521063842185948798330267); DK(KP516520780, +0.516520780623489722840901288569017135705033622); DK(KP151805972, +0.151805972074387731966205794490207080712856746); DK(KP503537032, +0.503537032863766627246873853868466977093348562); DK(KP166666666, +0.166666666666666666666666666666666666666666667); DK(KP600925212, +0.600925212577331548853203544578415991041882762); DK(KP500000000, +0.500000000000000000000000000000000000000000000); DK(KP256247671, +0.256247671582936600958684654061725059144125175); DK(KP156891391, +0.156891391051584611046832726756003269660212636); DK(KP348277202, +0.348277202304271810011321589858529485233929352); DK(KP1_150281458, +1.150281458948006242736771094910906776922003215); DK(KP300238635, +0.300238635966332641462884626667381504676006424); DK(KP011599105, +0.011599105605768290721655456654083252189827041); DK(KP1_732050807, +1.732050807568877293527446341505872366942805254); DK(KP2_000000000, +2.000000000000000000000000000000000000000000000); { INT i; for (i = v; i > 0; i = i - 1, R0 = R0 + ovs, R1 = R1 + ovs, Cr = Cr + ivs, Ci = Ci + ivs, MAKE_VOLATILE_STRIDE(52, rs), MAKE_VOLATILE_STRIDE(52, csr), MAKE_VOLATILE_STRIDE(52, csi)) { E TG, TS, TR, T15, TJ, TT, T1, Tm, Tc, Td, Tg, Tj, Tk, Tn, To; E Tp; { E Ts, Tv, Tw, TE, TC, TB, Tz, TD, TA, TF; { E Tt, Tu, Tx, Ty; Ts = Ci[WS(csi, 1)]; Tt = Ci[WS(csi, 3)]; Tu = Ci[WS(csi, 4)]; Tv = Tt - Tu; Tw = FMS(KP2_000000000, Ts, Tv); TE = KP1_732050807 * (Tt + Tu); TC = Ci[WS(csi, 5)]; Tx = Ci[WS(csi, 6)]; Ty = Ci[WS(csi, 2)]; TB = Tx + Ty; Tz = KP1_732050807 * (Tx - Ty); TD = FNMS(KP2_000000000, TC, TB); } TA = Tw + Tz; TF = TD - TE; TG = FMA(KP011599105, TA, KP300238635 * TF); TS = FNMS(KP011599105, TF, KP300238635 * TA); { E TP, TQ, TH, TI; TP = Ts + Tv; TQ = TB + TC; TR = FNMS(KP348277202, TQ, KP1_150281458 * TP); T15 = FMA(KP348277202, TP, KP1_150281458 * TQ); TH = Tw - Tz; TI = TE + TD; TJ = FMA(KP156891391, TH, KP256247671 * TI); TT = FNMS(KP256247671, TH, KP156891391 * TI); } } { E Tb, Ti, Tf, T6, Th, Te; T1 = Cr[0]; { E T7, T8, T9, Ta; T7 = Cr[WS(csr, 5)]; T8 = Cr[WS(csr, 2)]; T9 = Cr[WS(csr, 6)]; Ta = T8 + T9; Tb = T7 + Ta; Ti = FNMS(KP500000000, Ta, T7); Tf = T8 - T9; } { E T2, T3, T4, T5; T2 = Cr[WS(csr, 1)]; T3 = Cr[WS(csr, 3)]; T4 = Cr[WS(csr, 4)]; T5 = T3 + T4; T6 = T2 + T5; Th = FNMS(KP500000000, T5, T2); Te = T3 - T4; } Tm = KP600925212 * (T6 - Tb); Tc = T6 + Tb; Td = FNMS(KP166666666, Tc, T1); Tg = Te + Tf; Tj = Th + Ti; Tk = FMA(KP503537032, Tg, KP151805972 * Tj); Tn = Th - Ti; To = Te - Tf; Tp = FNMS(KP265966249, To, KP516520780 * Tn); } R0[0] = FMA(KP2_000000000, Tc, T1); { E TK, T1b, TV, T12, T16, T18, TO, T1a, Tr, T17, T11, T13; { E TU, T14, TM, TN; TK = KP1_732050807 * (TG + TJ); T1b = KP1_732050807 * (TS - TT); TU = TS + TT; TV = TR - TU; T12 = FMA(KP2_000000000, TU, TR); T14 = TG - TJ; T16 = FMS(KP2_000000000, T14, T15); T18 = T14 + T15; TM = FMA(KP774781170, To, KP531932498 * Tn); TN = FNMS(KP1_007074065, Tj, KP227708958 * Tg); TO = TM - TN; T1a = TM + TN; { E Tl, Tq, TZ, T10; Tl = Td - Tk; Tq = Tm - Tp; Tr = Tl - Tq; T17 = Tq + Tl; TZ = FMA(KP2_000000000, Tk, Td); T10 = FMA(KP2_000000000, Tp, Tm); T11 = TZ - T10; T13 = T10 + TZ; } } R1[WS(rs, 2)] = T11 - T12; R0[WS(rs, 6)] = T13 - T16; R1[0] = T13 + T16; R0[WS(rs, 4)] = T11 + T12; { E TL, TW, T19, T1c; TL = Tr - TK; TW = TO - TV; R1[WS(rs, 3)] = TL - TW; R0[WS(rs, 1)] = TL + TW; T19 = T17 - T18; T1c = T1a + T1b; R1[WS(rs, 1)] = T19 - T1c; R1[WS(rs, 4)] = T1c + T19; } { E T1d, T1e, TX, TY; T1d = T1a - T1b; T1e = T17 + T18; R0[WS(rs, 2)] = T1d + T1e; R0[WS(rs, 5)] = T1e - T1d; TX = Tr + TK; TY = TO + TV; R0[WS(rs, 3)] = TX - TY; R1[WS(rs, 5)] = TX + TY; } } } } } static const kr2c_desc desc = { 13, "r2cb_13", {56, 15, 20, 0}, &GENUS }; void X(codelet_r2cb_13) (planner *p) { X(kr2c_register) (p, r2cb_13, &desc); } #endif /* HAVE_FMA */ fftw-3.3.3/rdft/scalar/r2cb/r2cb_14.c0000644000175000017500000002206112054410745013746 00000000000000/* * Copyright (c) 2003, 2007-11 Matteo Frigo * Copyright (c) 2003, 2007-11 Massachusetts Institute of Technology * * This program is free software; you can redistribute it and/or modify * it under the terms of the GNU General Public License as published by * the Free Software Foundation; either version 2 of the License, or * (at your option) any later version. * * This program is distributed in the hope that it will be useful, * but WITHOUT ANY WARRANTY; without even the implied warranty of * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the * GNU General Public License for more details. * * You should have received a copy of the GNU General Public License * along with this program; if not, write to the Free Software * Foundation, Inc., 51 Franklin Street, Fifth Floor, Boston, MA 02110-1301 USA * */ /* This file was automatically generated --- DO NOT EDIT */ /* Generated on Sun Nov 25 07:41:08 EST 2012 */ #include "codelet-rdft.h" #ifdef HAVE_FMA /* Generated by: ../../../genfft/gen_r2cb.native -fma -reorder-insns -schedule-for-pipeline -compact -variables 4 -pipeline-latency 4 -sign 1 -n 14 -name r2cb_14 -include r2cb.h */ /* * This function contains 62 FP additions, 44 FP multiplications, * (or, 18 additions, 0 multiplications, 44 fused multiply/add), * 58 stack variables, 7 constants, and 28 memory accesses */ #include "r2cb.h" static void r2cb_14(R *R0, R *R1, R *Cr, R *Ci, stride rs, stride csr, stride csi, INT v, INT ivs, INT ovs) { DK(KP1_949855824, +1.949855824363647214036263365987862434465571601); DK(KP1_801937735, +1.801937735804838252472204639014890102331838324); DK(KP692021471, +0.692021471630095869627814897002069140197260599); DK(KP2_000000000, +2.000000000000000000000000000000000000000000000); DK(KP356895867, +0.356895867892209443894399510021300583399127187); DK(KP801937735, +0.801937735804838252472204639014890102331838324); DK(KP554958132, +0.554958132087371191422194871006410481067288862); { INT i; for (i = v; i > 0; i = i - 1, R0 = R0 + ovs, R1 = R1 + ovs, Cr = Cr + ivs, Ci = Ci + ivs, MAKE_VOLATILE_STRIDE(56, rs), MAKE_VOLATILE_STRIDE(56, csr), MAKE_VOLATILE_STRIDE(56, csi)) { E Te, TO, TT, TG, TJ, TD, TR, TE; { E T3, TK, To, TM, Tu, TL, Tr, TS, TA, TN, TX, TF, Tv, T7, Tf; E T6, Th, Tc, T8, T1, T2; T1 = Cr[0]; T2 = Cr[WS(csr, 7)]; { E Ts, Tt, Tp, Tq, Tm, Tn; Tm = Ci[WS(csi, 4)]; Tn = Ci[WS(csi, 3)]; Ts = Ci[WS(csi, 6)]; Te = T1 + T2; T3 = T1 - T2; TK = Tm + Tn; To = Tm - Tn; Tt = Ci[WS(csi, 1)]; Tp = Ci[WS(csi, 2)]; Tq = Ci[WS(csi, 5)]; { E T4, T5, Ta, Tb; T4 = Cr[WS(csr, 2)]; TM = Ts + Tt; Tu = Ts - Tt; TL = Tp + Tq; Tr = Tp - Tq; TS = FMA(KP554958132, TK, TM); TA = FMA(KP554958132, To, Tu); TN = FMA(KP554958132, TM, TL); TX = FNMS(KP554958132, TL, TK); TF = FNMS(KP554958132, Tr, To); Tv = FMA(KP554958132, Tu, Tr); T5 = Cr[WS(csr, 5)]; Ta = Cr[WS(csr, 6)]; Tb = Cr[WS(csr, 1)]; T7 = Cr[WS(csr, 4)]; Tf = T4 + T5; T6 = T4 - T5; Th = Ta + Tb; Tc = Ta - Tb; T8 = Cr[WS(csr, 3)]; } } { E Tw, Tx, TP, Tg, T9, TY, TC, TI, TQ; Tw = FMA(KP801937735, Tv, To); Tx = FNMS(KP356895867, Tf, Th); TP = FNMS(KP356895867, T6, Tc); Tg = T7 + T8; T9 = T7 - T8; TY = FNMS(KP801937735, TX, TM); { E TB, TH, TV, Ty, Tl, Ti, TW, Tz; TB = FNMS(KP801937735, TA, Tr); Ti = Tf + Tg + Th; TC = FNMS(KP356895867, Th, Tg); { E Tj, Td, TU, Tk; Tj = FNMS(KP356895867, Tg, Tf); Td = T6 + T9 + Tc; TH = FNMS(KP356895867, T9, T6); TU = FNMS(KP356895867, Tc, T9); R0[0] = FMA(KP2_000000000, Ti, Te); Tk = FNMS(KP692021471, Tj, Th); R1[WS(rs, 3)] = FMA(KP2_000000000, Td, T3); TV = FNMS(KP692021471, TU, T6); Ty = FNMS(KP692021471, Tx, Tg); Tl = FNMS(KP1_801937735, Tk, Te); } TO = FMA(KP801937735, TN, TK); TW = FNMS(KP1_801937735, TV, T3); Tz = FNMS(KP1_801937735, Ty, Te); R0[WS(rs, 3)] = FMA(KP1_949855824, Tw, Tl); R0[WS(rs, 4)] = FNMS(KP1_949855824, Tw, Tl); R1[WS(rs, 5)] = FMA(KP1_949855824, TY, TW); R1[WS(rs, 1)] = FNMS(KP1_949855824, TY, TW); R0[WS(rs, 6)] = FMA(KP1_949855824, TB, Tz); R0[WS(rs, 1)] = FNMS(KP1_949855824, TB, Tz); TI = FNMS(KP692021471, TH, Tc); } TT = FNMS(KP801937735, TS, TL); TQ = FNMS(KP692021471, TP, T9); TG = FNMS(KP801937735, TF, Tu); TJ = FNMS(KP1_801937735, TI, T3); TD = FNMS(KP692021471, TC, Tf); TR = FNMS(KP1_801937735, TQ, T3); } } R1[WS(rs, 6)] = FMA(KP1_949855824, TO, TJ); R1[0] = FNMS(KP1_949855824, TO, TJ); TE = FNMS(KP1_801937735, TD, Te); R1[WS(rs, 2)] = FMA(KP1_949855824, TT, TR); R1[WS(rs, 4)] = FNMS(KP1_949855824, TT, TR); R0[WS(rs, 2)] = FMA(KP1_949855824, TG, TE); R0[WS(rs, 5)] = FNMS(KP1_949855824, TG, TE); } } } static const kr2c_desc desc = { 14, "r2cb_14", {18, 0, 44, 0}, &GENUS }; void X(codelet_r2cb_14) (planner *p) { X(kr2c_register) (p, r2cb_14, &desc); } #else /* HAVE_FMA */ /* Generated by: ../../../genfft/gen_r2cb.native -compact -variables 4 -pipeline-latency 4 -sign 1 -n 14 -name r2cb_14 -include r2cb.h */ /* * This function contains 62 FP additions, 38 FP multiplications, * (or, 36 additions, 12 multiplications, 26 fused multiply/add), * 28 stack variables, 7 constants, and 28 memory accesses */ #include "r2cb.h" static void r2cb_14(R *R0, R *R1, R *Cr, R *Ci, stride rs, stride csr, stride csi, INT v, INT ivs, INT ovs) { DK(KP1_801937735, +1.801937735804838252472204639014890102331838324); DK(KP445041867, +0.445041867912628808577805128993589518932711138); DK(KP1_246979603, +1.246979603717467061050009768008479621264549462); DK(KP867767478, +0.867767478235116240951536665696717509219981456); DK(KP1_949855824, +1.949855824363647214036263365987862434465571601); DK(KP1_563662964, +1.563662964936059617416889053348115500464669037); DK(KP2_000000000, +2.000000000000000000000000000000000000000000000); { INT i; for (i = v; i > 0; i = i - 1, R0 = R0 + ovs, R1 = R1 + ovs, Cr = Cr + ivs, Ci = Ci + ivs, MAKE_VOLATILE_STRIDE(56, rs), MAKE_VOLATILE_STRIDE(56, csr), MAKE_VOLATILE_STRIDE(56, csi)) { E T3, Td, T6, Te, Tq, Tz, Tn, Ty, Tc, Tg, Tk, Tx, T9, Tf, T1; E T2; T1 = Cr[0]; T2 = Cr[WS(csr, 7)]; T3 = T1 - T2; Td = T1 + T2; { E T4, T5, To, Tp; T4 = Cr[WS(csr, 2)]; T5 = Cr[WS(csr, 5)]; T6 = T4 - T5; Te = T4 + T5; To = Ci[WS(csi, 2)]; Tp = Ci[WS(csi, 5)]; Tq = To - Tp; Tz = To + Tp; } { E Tl, Tm, Ta, Tb; Tl = Ci[WS(csi, 6)]; Tm = Ci[WS(csi, 1)]; Tn = Tl - Tm; Ty = Tl + Tm; Ta = Cr[WS(csr, 6)]; Tb = Cr[WS(csr, 1)]; Tc = Ta - Tb; Tg = Ta + Tb; } { E Ti, Tj, T7, T8; Ti = Ci[WS(csi, 4)]; Tj = Ci[WS(csi, 3)]; Tk = Ti - Tj; Tx = Ti + Tj; T7 = Cr[WS(csr, 4)]; T8 = Cr[WS(csr, 3)]; T9 = T7 - T8; Tf = T7 + T8; } R1[WS(rs, 3)] = FMA(KP2_000000000, T6 + T9 + Tc, T3); R0[0] = FMA(KP2_000000000, Te + Tf + Tg, Td); { E Tr, Th, TE, TD; Tr = FNMS(KP1_949855824, Tn, KP1_563662964 * Tk) - (KP867767478 * Tq); Th = FMA(KP1_246979603, Tf, Td) + FNMA(KP445041867, Tg, KP1_801937735 * Te); R0[WS(rs, 2)] = Th - Tr; R0[WS(rs, 5)] = Th + Tr; TE = FMA(KP867767478, Tx, KP1_563662964 * Ty) - (KP1_949855824 * Tz); TD = FMA(KP1_246979603, Tc, T3) + FNMA(KP1_801937735, T9, KP445041867 * T6); R1[WS(rs, 2)] = TD - TE; R1[WS(rs, 4)] = TD + TE; } { E Tt, Ts, TA, Tw; Tt = FMA(KP867767478, Tk, KP1_563662964 * Tn) - (KP1_949855824 * Tq); Ts = FMA(KP1_246979603, Tg, Td) + FNMA(KP1_801937735, Tf, KP445041867 * Te); R0[WS(rs, 6)] = Ts - Tt; R0[WS(rs, 1)] = Ts + Tt; TA = FNMS(KP1_949855824, Ty, KP1_563662964 * Tx) - (KP867767478 * Tz); Tw = FMA(KP1_246979603, T9, T3) + FNMA(KP445041867, Tc, KP1_801937735 * T6); R1[WS(rs, 5)] = Tw - TA; R1[WS(rs, 1)] = Tw + TA; } { E TC, TB, Tv, Tu; TC = FMA(KP1_563662964, Tz, KP1_949855824 * Tx) + (KP867767478 * Ty); TB = FMA(KP1_246979603, T6, T3) + FNMA(KP1_801937735, Tc, KP445041867 * T9); R1[0] = TB - TC; R1[WS(rs, 6)] = TB + TC; Tv = FMA(KP1_563662964, Tq, KP1_949855824 * Tk) + (KP867767478 * Tn); Tu = FMA(KP1_246979603, Te, Td) + FNMA(KP1_801937735, Tg, KP445041867 * Tf); R0[WS(rs, 4)] = Tu - Tv; R0[WS(rs, 3)] = Tu + Tv; } } } } static const kr2c_desc desc = { 14, "r2cb_14", {36, 12, 26, 0}, &GENUS }; void X(codelet_r2cb_14) (planner *p) { X(kr2c_register) (p, r2cb_14, &desc); } #endif /* HAVE_FMA */ fftw-3.3.3/rdft/scalar/r2cb/r2cb_15.c0000644000175000017500000002331412054410744013750 00000000000000/* * Copyright (c) 2003, 2007-11 Matteo Frigo * Copyright (c) 2003, 2007-11 Massachusetts Institute of Technology * * This program is free software; you can redistribute it and/or modify * it under the terms of the GNU General Public License as published by * the Free Software Foundation; either version 2 of the License, or * (at your option) any later version. * * This program is distributed in the hope that it will be useful, * but WITHOUT ANY WARRANTY; without even the implied warranty of * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the * GNU General Public License for more details. * * You should have received a copy of the GNU General Public License * along with this program; if not, write to the Free Software * Foundation, Inc., 51 Franklin Street, Fifth Floor, Boston, MA 02110-1301 USA * */ /* This file was automatically generated --- DO NOT EDIT */ /* Generated on Sun Nov 25 07:41:08 EST 2012 */ #include "codelet-rdft.h" #ifdef HAVE_FMA /* Generated by: ../../../genfft/gen_r2cb.native -fma -reorder-insns -schedule-for-pipeline -compact -variables 4 -pipeline-latency 4 -sign 1 -n 15 -name r2cb_15 -include r2cb.h */ /* * This function contains 64 FP additions, 43 FP multiplications, * (or, 21 additions, 0 multiplications, 43 fused multiply/add), * 54 stack variables, 9 constants, and 30 memory accesses */ #include "r2cb.h" static void r2cb_15(R *R0, R *R1, R *Cr, R *Ci, stride rs, stride csr, stride csi, INT v, INT ivs, INT ovs) { DK(KP559016994, +0.559016994374947424102293417182819058860154590); DK(KP1_902113032, +1.902113032590307144232878666758764286811397268); DK(KP250000000, +0.250000000000000000000000000000000000000000000); DK(KP866025403, +0.866025403784438646763723170752936183471402627); DK(KP1_118033988, +1.118033988749894848204586834365638117720309180); DK(KP618033988, +0.618033988749894848204586834365638117720309180); DK(KP500000000, +0.500000000000000000000000000000000000000000000); DK(KP1_732050807, +1.732050807568877293527446341505872366942805254); DK(KP2_000000000, +2.000000000000000000000000000000000000000000000); { INT i; for (i = v; i > 0; i = i - 1, R0 = R0 + ovs, R1 = R1 + ovs, Cr = Cr + ivs, Ci = Ci + ivs, MAKE_VOLATILE_STRIDE(60, rs), MAKE_VOLATILE_STRIDE(60, csr), MAKE_VOLATILE_STRIDE(60, csi)) { E TL, Tz, TM, TK; { E T3, Th, Tt, TD, TI, TH, TY, TC, TZ, Tu, Tm, Tv, Tr, Te, TW; E Tg, T1, T2, T12, T10, TV; Tg = Ci[WS(csi, 5)]; T1 = Cr[0]; T2 = Cr[WS(csr, 5)]; { E T4, TA, T9, TF, T7, Tj, Tc, Tk, TG, Tq, Tf, Tl, TB; T4 = Cr[WS(csr, 3)]; TA = Ci[WS(csi, 3)]; T9 = Cr[WS(csr, 6)]; Tf = T1 - T2; T3 = FMA(KP2_000000000, T2, T1); TF = Ci[WS(csi, 6)]; { E Ta, Tb, T5, T6, To, Tp; T5 = Cr[WS(csr, 7)]; T6 = Cr[WS(csr, 2)]; Th = FMA(KP1_732050807, Tg, Tf); Tt = FNMS(KP1_732050807, Tg, Tf); Ta = Cr[WS(csr, 4)]; TD = T5 - T6; T7 = T5 + T6; Tb = Cr[WS(csr, 1)]; To = Ci[WS(csi, 4)]; Tp = Ci[WS(csi, 1)]; Tj = Ci[WS(csi, 7)]; Tc = Ta + Tb; TI = Ta - Tb; Tk = Ci[WS(csi, 2)]; TG = Tp - To; Tq = To + Tp; } Tl = Tj - Tk; TB = Tj + Tk; TH = FNMS(KP500000000, TG, TF); TY = TG + TF; TC = FMA(KP500000000, TB, TA); TZ = TA - TB; { E Ti, T8, Td, Tn; Ti = FNMS(KP2_000000000, T4, T7); T8 = T4 + T7; Td = T9 + Tc; Tn = FNMS(KP2_000000000, T9, Tc); Tu = FNMS(KP1_732050807, Tl, Ti); Tm = FMA(KP1_732050807, Tl, Ti); Tv = FNMS(KP1_732050807, Tq, Tn); Tr = FMA(KP1_732050807, Tq, Tn); Te = T8 + Td; TW = T8 - Td; } } T12 = FMA(KP618033988, TY, TZ); T10 = FNMS(KP618033988, TZ, TY); TV = FNMS(KP500000000, Te, T3); R0[0] = FMA(KP2_000000000, Te, T3); { E TJ, TE, TT, TP, TU, TS, Ty, Tw, Tx; { E TO, Ts, TQ, TN, TR, T11, TX; TO = Tr - Tm; Ts = Tm + Tr; T11 = FMA(KP1_118033988, TW, TV); TX = FNMS(KP1_118033988, TW, TV); TQ = FNMS(KP866025403, TI, TH); TJ = FMA(KP866025403, TI, TH); TN = FMA(KP250000000, Ts, Th); R0[WS(rs, 3)] = FNMS(KP1_902113032, T12, T11); R1[WS(rs, 4)] = FMA(KP1_902113032, T12, T11); R0[WS(rs, 6)] = FMA(KP1_902113032, T10, TX); R1[WS(rs, 1)] = FNMS(KP1_902113032, T10, TX); TR = FNMS(KP866025403, TD, TC); TE = FMA(KP866025403, TD, TC); R1[WS(rs, 2)] = Th - Ts; TT = FMA(KP559016994, TO, TN); TP = FNMS(KP559016994, TO, TN); TU = FMA(KP618033988, TQ, TR); TS = FNMS(KP618033988, TR, TQ); } Ty = Tv - Tu; Tw = Tu + Tv; R0[WS(rs, 7)] = FMA(KP1_902113032, TU, TT); R1[WS(rs, 5)] = FNMS(KP1_902113032, TU, TT); R0[WS(rs, 1)] = FMA(KP1_902113032, TS, TP); R0[WS(rs, 4)] = FNMS(KP1_902113032, TS, TP); Tx = FMA(KP250000000, Tw, Tt); R0[WS(rs, 5)] = Tt - Tw; TL = FNMS(KP559016994, Ty, Tx); Tz = FMA(KP559016994, Ty, Tx); TM = FNMS(KP618033988, TE, TJ); TK = FMA(KP618033988, TJ, TE); } } R1[WS(rs, 3)] = FMA(KP1_902113032, TM, TL); R1[WS(rs, 6)] = FNMS(KP1_902113032, TM, TL); R0[WS(rs, 2)] = FMA(KP1_902113032, TK, Tz); R1[0] = FNMS(KP1_902113032, TK, Tz); } } } static const kr2c_desc desc = { 15, "r2cb_15", {21, 0, 43, 0}, &GENUS }; void X(codelet_r2cb_15) (planner *p) { X(kr2c_register) (p, r2cb_15, &desc); } #else /* HAVE_FMA */ /* Generated by: ../../../genfft/gen_r2cb.native -compact -variables 4 -pipeline-latency 4 -sign 1 -n 15 -name r2cb_15 -include r2cb.h */ /* * This function contains 64 FP additions, 31 FP multiplications, * (or, 47 additions, 14 multiplications, 17 fused multiply/add), * 44 stack variables, 7 constants, and 30 memory accesses */ #include "r2cb.h" static void r2cb_15(R *R0, R *R1, R *Cr, R *Ci, stride rs, stride csr, stride csi, INT v, INT ivs, INT ovs) { DK(KP1_118033988, +1.118033988749894848204586834365638117720309180); DK(KP1_902113032, +1.902113032590307144232878666758764286811397268); DK(KP1_175570504, +1.175570504584946258337411909278145537195304875); DK(KP500000000, +0.500000000000000000000000000000000000000000000); DK(KP866025403, +0.866025403784438646763723170752936183471402627); DK(KP2_000000000, +2.000000000000000000000000000000000000000000000); DK(KP1_732050807, +1.732050807568877293527446341505872366942805254); { INT i; for (i = v; i > 0; i = i - 1, R0 = R0 + ovs, R1 = R1 + ovs, Cr = Cr + ivs, Ci = Ci + ivs, MAKE_VOLATILE_STRIDE(60, rs), MAKE_VOLATILE_STRIDE(60, csr), MAKE_VOLATILE_STRIDE(60, csi)) { E T3, Tu, Ti, TB, TZ, T10, TE, TG, TJ, Tn, Tv, Ts, Tw, T8, Td; E Te; { E Th, T1, T2, Tf, Tg; Tg = Ci[WS(csi, 5)]; Th = KP1_732050807 * Tg; T1 = Cr[0]; T2 = Cr[WS(csr, 5)]; Tf = T1 - T2; T3 = FMA(KP2_000000000, T2, T1); Tu = Tf - Th; Ti = Tf + Th; } { E T4, TD, T9, TI, T5, T6, T7, Ta, Tb, Tc, Tr, TH, Tm, TC, Tj; E To; T4 = Cr[WS(csr, 3)]; TD = Ci[WS(csi, 3)]; T9 = Cr[WS(csr, 6)]; TI = Ci[WS(csi, 6)]; T5 = Cr[WS(csr, 7)]; T6 = Cr[WS(csr, 2)]; T7 = T5 + T6; Ta = Cr[WS(csr, 4)]; Tb = Cr[WS(csr, 1)]; Tc = Ta + Tb; { E Tp, Tq, Tk, Tl; Tp = Ci[WS(csi, 4)]; Tq = Ci[WS(csi, 1)]; Tr = KP866025403 * (Tp + Tq); TH = Tp - Tq; Tk = Ci[WS(csi, 7)]; Tl = Ci[WS(csi, 2)]; Tm = KP866025403 * (Tk - Tl); TC = Tk + Tl; } TB = KP866025403 * (T5 - T6); TZ = TD - TC; T10 = TI - TH; TE = FMA(KP500000000, TC, TD); TG = KP866025403 * (Ta - Tb); TJ = FMA(KP500000000, TH, TI); Tj = FNMS(KP500000000, T7, T4); Tn = Tj - Tm; Tv = Tj + Tm; To = FNMS(KP500000000, Tc, T9); Ts = To - Tr; Tw = To + Tr; T8 = T4 + T7; Td = T9 + Tc; Te = T8 + Td; } R0[0] = FMA(KP2_000000000, Te, T3); { E T11, T13, TY, T12, TW, TX; T11 = FNMS(KP1_902113032, T10, KP1_175570504 * TZ); T13 = FMA(KP1_902113032, TZ, KP1_175570504 * T10); TW = FNMS(KP500000000, Te, T3); TX = KP1_118033988 * (T8 - Td); TY = TW - TX; T12 = TX + TW; R0[WS(rs, 6)] = TY - T11; R1[WS(rs, 4)] = T12 + T13; R1[WS(rs, 1)] = TY + T11; R0[WS(rs, 3)] = T12 - T13; } { E TP, Tt, TO, TT, TV, TR, TS, TU, TQ; TP = KP1_118033988 * (Tn - Ts); Tt = Tn + Ts; TO = FNMS(KP500000000, Tt, Ti); TR = TE - TB; TS = TJ - TG; TT = FNMS(KP1_902113032, TS, KP1_175570504 * TR); TV = FMA(KP1_902113032, TR, KP1_175570504 * TS); R1[WS(rs, 2)] = FMA(KP2_000000000, Tt, Ti); TU = TP + TO; R1[WS(rs, 5)] = TU - TV; R0[WS(rs, 7)] = TU + TV; TQ = TO - TP; R0[WS(rs, 1)] = TQ - TT; R0[WS(rs, 4)] = TQ + TT; } { E Tz, Tx, Ty, TL, TN, TF, TK, TM, TA; Tz = KP1_118033988 * (Tv - Tw); Tx = Tv + Tw; Ty = FNMS(KP500000000, Tx, Tu); TF = TB + TE; TK = TG + TJ; TL = FNMS(KP1_902113032, TK, KP1_175570504 * TF); TN = FMA(KP1_902113032, TF, KP1_175570504 * TK); R0[WS(rs, 5)] = FMA(KP2_000000000, Tx, Tu); TM = Tz + Ty; R1[0] = TM - TN; R0[WS(rs, 2)] = TM + TN; TA = Ty - Tz; R1[WS(rs, 3)] = TA - TL; R1[WS(rs, 6)] = TA + TL; } } } } static const kr2c_desc desc = { 15, "r2cb_15", {47, 14, 17, 0}, &GENUS }; void X(codelet_r2cb_15) (planner *p) { X(kr2c_register) (p, r2cb_15, &desc); } #endif /* HAVE_FMA */ fftw-3.3.3/rdft/scalar/r2cb/r2cb_16.c0000644000175000017500000002114412054410745013751 00000000000000/* * Copyright (c) 2003, 2007-11 Matteo Frigo * Copyright (c) 2003, 2007-11 Massachusetts Institute of Technology * * This program is free software; you can redistribute it and/or modify * it under the terms of the GNU General Public License as published by * the Free Software Foundation; either version 2 of the License, or * (at your option) any later version. * * This program is distributed in the hope that it will be useful, * but WITHOUT ANY WARRANTY; without even the implied warranty of * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the * GNU General Public License for more details. * * You should have received a copy of the GNU General Public License * along with this program; if not, write to the Free Software * Foundation, Inc., 51 Franklin Street, Fifth Floor, Boston, MA 02110-1301 USA * */ /* This file was automatically generated --- DO NOT EDIT */ /* Generated on Sun Nov 25 07:41:08 EST 2012 */ #include "codelet-rdft.h" #ifdef HAVE_FMA /* Generated by: ../../../genfft/gen_r2cb.native -fma -reorder-insns -schedule-for-pipeline -compact -variables 4 -pipeline-latency 4 -sign 1 -n 16 -name r2cb_16 -include r2cb.h */ /* * This function contains 58 FP additions, 32 FP multiplications, * (or, 26 additions, 0 multiplications, 32 fused multiply/add), * 47 stack variables, 4 constants, and 32 memory accesses */ #include "r2cb.h" static void r2cb_16(R *R0, R *R1, R *Cr, R *Ci, stride rs, stride csr, stride csi, INT v, INT ivs, INT ovs) { DK(KP1_847759065, +1.847759065022573512256366378793576573644833252); DK(KP414213562, +0.414213562373095048801688724209698078569671875); DK(KP1_414213562, +1.414213562373095048801688724209698078569671875); DK(KP2_000000000, +2.000000000000000000000000000000000000000000000); { INT i; for (i = v; i > 0; i = i - 1, R0 = R0 + ovs, R1 = R1 + ovs, Cr = Cr + ivs, Ci = Ci + ivs, MAKE_VOLATILE_STRIDE(64, rs), MAKE_VOLATILE_STRIDE(64, csr), MAKE_VOLATILE_STRIDE(64, csi)) { E TN, TS, TF, TI; { E T8, TD, Tj, TL, T5, TM, TE, To, Td, Tq, Tc, TP, Ty, Te, Tr; E Ts; { E T4, Ti, T1, T2; T4 = Cr[WS(csr, 4)]; Ti = Ci[WS(csi, 4)]; T1 = Cr[0]; T2 = Cr[WS(csr, 8)]; { E Tk, Tn, T6, T7; T6 = Cr[WS(csr, 2)]; T7 = Cr[WS(csr, 6)]; { E Tl, Th, T3, Tm; Tl = Ci[WS(csi, 2)]; Th = T1 - T2; T3 = T1 + T2; Tk = T6 - T7; T8 = T6 + T7; Tm = Ci[WS(csi, 6)]; TD = FMA(KP2_000000000, Ti, Th); Tj = FNMS(KP2_000000000, Ti, Th); TL = FNMS(KP2_000000000, T4, T3); T5 = FMA(KP2_000000000, T4, T3); Tn = Tl + Tm; TM = Tl - Tm; } { E Ta, Tb, Tw, Tx; Ta = Cr[WS(csr, 1)]; TE = Tk + Tn; To = Tk - Tn; Tb = Cr[WS(csr, 7)]; Tw = Ci[WS(csi, 1)]; Tx = Ci[WS(csi, 7)]; Td = Cr[WS(csr, 5)]; Tq = Ta - Tb; Tc = Ta + Tb; TP = Tw - Tx; Ty = Tw + Tx; Te = Cr[WS(csr, 3)]; Tr = Ci[WS(csi, 5)]; Ts = Ci[WS(csi, 3)]; } } } { E TV, TG, TW, TH, TB, Tp, TA, TC, TJ, TK; { E T9, Tz, Tg, Tu, TT, TU, TO, TR; TV = FNMS(KP2_000000000, T8, T5); T9 = FMA(KP2_000000000, T8, T5); { E Tv, Tf, TQ, Tt; Tv = Td - Te; Tf = Td + Te; TQ = Tr - Ts; Tt = Tr + Ts; TG = Ty - Tv; Tz = Tv + Ty; TO = Tc - Tf; Tg = Tc + Tf; TW = TQ + TP; TR = TP - TQ; TH = Tq + Tt; Tu = Tq - Tt; } TN = FNMS(KP2_000000000, TM, TL); TT = FMA(KP2_000000000, TM, TL); TU = TO + TR; TS = TO - TR; R0[0] = FMA(KP2_000000000, Tg, T9); R0[WS(rs, 4)] = FNMS(KP2_000000000, Tg, T9); R0[WS(rs, 7)] = FMA(KP1_414213562, TU, TT); R0[WS(rs, 3)] = FNMS(KP1_414213562, TU, TT); TB = FNMS(KP1_414213562, To, Tj); Tp = FMA(KP1_414213562, To, Tj); TA = FNMS(KP414213562, Tz, Tu); TC = FMA(KP414213562, Tu, Tz); } R0[WS(rs, 6)] = FMA(KP2_000000000, TW, TV); R0[WS(rs, 2)] = FNMS(KP2_000000000, TW, TV); R1[0] = FMA(KP1_847759065, TA, Tp); R1[WS(rs, 4)] = FNMS(KP1_847759065, TA, Tp); TF = FNMS(KP1_414213562, TE, TD); TJ = FMA(KP1_414213562, TE, TD); TK = FMA(KP414213562, TG, TH); TI = FNMS(KP414213562, TH, TG); R1[WS(rs, 6)] = FMA(KP1_847759065, TC, TB); R1[WS(rs, 2)] = FNMS(KP1_847759065, TC, TB); R1[WS(rs, 7)] = FMA(KP1_847759065, TK, TJ); R1[WS(rs, 3)] = FNMS(KP1_847759065, TK, TJ); } } R0[WS(rs, 1)] = FMA(KP1_414213562, TS, TN); R0[WS(rs, 5)] = FNMS(KP1_414213562, TS, TN); R1[WS(rs, 5)] = FMA(KP1_847759065, TI, TF); R1[WS(rs, 1)] = FNMS(KP1_847759065, TI, TF); } } } static const kr2c_desc desc = { 16, "r2cb_16", {26, 0, 32, 0}, &GENUS }; void X(codelet_r2cb_16) (planner *p) { X(kr2c_register) (p, r2cb_16, &desc); } #else /* HAVE_FMA */ /* Generated by: ../../../genfft/gen_r2cb.native -compact -variables 4 -pipeline-latency 4 -sign 1 -n 16 -name r2cb_16 -include r2cb.h */ /* * This function contains 58 FP additions, 18 FP multiplications, * (or, 54 additions, 14 multiplications, 4 fused multiply/add), * 31 stack variables, 4 constants, and 32 memory accesses */ #include "r2cb.h" static void r2cb_16(R *R0, R *R1, R *Cr, R *Ci, stride rs, stride csr, stride csi, INT v, INT ivs, INT ovs) { DK(KP1_847759065, +1.847759065022573512256366378793576573644833252); DK(KP765366864, +0.765366864730179543456919968060797733522689125); DK(KP1_414213562, +1.414213562373095048801688724209698078569671875); DK(KP2_000000000, +2.000000000000000000000000000000000000000000000); { INT i; for (i = v; i > 0; i = i - 1, R0 = R0 + ovs, R1 = R1 + ovs, Cr = Cr + ivs, Ci = Ci + ivs, MAKE_VOLATILE_STRIDE(64, rs), MAKE_VOLATILE_STRIDE(64, csr), MAKE_VOLATILE_STRIDE(64, csi)) { E T9, TS, Tl, TG, T6, TR, Ti, TD, Td, Tq, Tg, Tt, Tn, Tu, TV; E TU, TN, TK; { E T7, T8, TE, Tj, Tk, TF; T7 = Cr[WS(csr, 2)]; T8 = Cr[WS(csr, 6)]; TE = T7 - T8; Tj = Ci[WS(csi, 2)]; Tk = Ci[WS(csi, 6)]; TF = Tj + Tk; T9 = KP2_000000000 * (T7 + T8); TS = KP1_414213562 * (TE + TF); Tl = KP2_000000000 * (Tj - Tk); TG = KP1_414213562 * (TE - TF); } { E T5, TC, T3, TA; { E T4, TB, T1, T2; T4 = Cr[WS(csr, 4)]; T5 = KP2_000000000 * T4; TB = Ci[WS(csi, 4)]; TC = KP2_000000000 * TB; T1 = Cr[0]; T2 = Cr[WS(csr, 8)]; T3 = T1 + T2; TA = T1 - T2; } T6 = T3 + T5; TR = TA + TC; Ti = T3 - T5; TD = TA - TC; } { E TI, TM, TL, TJ; { E Tb, Tc, To, Tp; Tb = Cr[WS(csr, 1)]; Tc = Cr[WS(csr, 7)]; Td = Tb + Tc; TI = Tb - Tc; To = Ci[WS(csi, 1)]; Tp = Ci[WS(csi, 7)]; Tq = To - Tp; TM = To + Tp; } { E Te, Tf, Tr, Ts; Te = Cr[WS(csr, 5)]; Tf = Cr[WS(csr, 3)]; Tg = Te + Tf; TL = Te - Tf; Tr = Ci[WS(csi, 5)]; Ts = Ci[WS(csi, 3)]; Tt = Tr - Ts; TJ = Tr + Ts; } Tn = Td - Tg; Tu = Tq - Tt; TV = TM - TL; TU = TI + TJ; TN = TL + TM; TK = TI - TJ; } { E Ta, Th, TT, TW; Ta = T6 + T9; Th = KP2_000000000 * (Td + Tg); R0[WS(rs, 4)] = Ta - Th; R0[0] = Ta + Th; TT = TR - TS; TW = FNMS(KP1_847759065, TV, KP765366864 * TU); R1[WS(rs, 5)] = TT - TW; R1[WS(rs, 1)] = TT + TW; } { E TX, TY, Tm, Tv; TX = TR + TS; TY = FMA(KP1_847759065, TU, KP765366864 * TV); R1[WS(rs, 3)] = TX - TY; R1[WS(rs, 7)] = TX + TY; Tm = Ti - Tl; Tv = KP1_414213562 * (Tn - Tu); R0[WS(rs, 5)] = Tm - Tv; R0[WS(rs, 1)] = Tm + Tv; } { E Tw, Tx, TH, TO; Tw = Ti + Tl; Tx = KP1_414213562 * (Tn + Tu); R0[WS(rs, 3)] = Tw - Tx; R0[WS(rs, 7)] = Tw + Tx; TH = TD + TG; TO = FNMS(KP765366864, TN, KP1_847759065 * TK); R1[WS(rs, 4)] = TH - TO; R1[0] = TH + TO; } { E TP, TQ, Ty, Tz; TP = TD - TG; TQ = FMA(KP765366864, TK, KP1_847759065 * TN); R1[WS(rs, 2)] = TP - TQ; R1[WS(rs, 6)] = TP + TQ; Ty = T6 - T9; Tz = KP2_000000000 * (Tt + Tq); R0[WS(rs, 2)] = Ty - Tz; R0[WS(rs, 6)] = Ty + Tz; } } } } static const kr2c_desc desc = { 16, "r2cb_16", {54, 14, 4, 0}, &GENUS }; void X(codelet_r2cb_16) (planner *p) { X(kr2c_register) (p, r2cb_16, &desc); } #endif /* HAVE_FMA */ fftw-3.3.3/rdft/scalar/r2cb/r2cb_32.c0000644000175000017500000004403312054410750013745 00000000000000/* * Copyright (c) 2003, 2007-11 Matteo Frigo * Copyright (c) 2003, 2007-11 Massachusetts Institute of Technology * * This program is free software; you can redistribute it and/or modify * it under the terms of the GNU General Public License as published by * the Free Software Foundation; either version 2 of the License, or * (at your option) any later version. * * This program is distributed in the hope that it will be useful, * but WITHOUT ANY WARRANTY; without even the implied warranty of * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the * GNU General Public License for more details. * * You should have received a copy of the GNU General Public License * along with this program; if not, write to the Free Software * Foundation, Inc., 51 Franklin Street, Fifth Floor, Boston, MA 02110-1301 USA * */ /* This file was automatically generated --- DO NOT EDIT */ /* Generated on Sun Nov 25 07:41:09 EST 2012 */ #include "codelet-rdft.h" #ifdef HAVE_FMA /* Generated by: ../../../genfft/gen_r2cb.native -fma -reorder-insns -schedule-for-pipeline -compact -variables 4 -pipeline-latency 4 -sign 1 -n 32 -name r2cb_32 -include r2cb.h */ /* * This function contains 156 FP additions, 84 FP multiplications, * (or, 72 additions, 0 multiplications, 84 fused multiply/add), * 82 stack variables, 9 constants, and 64 memory accesses */ #include "r2cb.h" static void r2cb_32(R *R0, R *R1, R *Cr, R *Ci, stride rs, stride csr, stride csi, INT v, INT ivs, INT ovs) { DK(KP1_662939224, +1.662939224605090474157576755235811513477121624); DK(KP668178637, +0.668178637919298919997757686523080761552472251); DK(KP1_961570560, +1.961570560806460898252364472268478073947867462); DK(KP198912367, +0.198912367379658006911597622644676228597850501); DK(KP707106781, +0.707106781186547524400844362104849039284835938); DK(KP1_847759065, +1.847759065022573512256366378793576573644833252); DK(KP414213562, +0.414213562373095048801688724209698078569671875); DK(KP1_414213562, +1.414213562373095048801688724209698078569671875); DK(KP2_000000000, +2.000000000000000000000000000000000000000000000); { INT i; for (i = v; i > 0; i = i - 1, R0 = R0 + ovs, R1 = R1 + ovs, Cr = Cr + ivs, Ci = Ci + ivs, MAKE_VOLATILE_STRIDE(128, rs), MAKE_VOLATILE_STRIDE(128, csr), MAKE_VOLATILE_STRIDE(128, csi)) { E T1F, T1C, T1H, T1z, T1G, T1I; { E T8, T1t, Tz, T1R, T5, T1S, T1u, TE, T1w, TP, T1U, Tg, T2m, T1X, T1x; E TK, T1D, T1d, T20, To, T2p, T28, T1A, TW, T11, T1e, Tv, T25, T23, T2q; E T16, T1f, TA, TD; { E T4, Ty, T1, T2, T6, T7; T4 = Cr[WS(csr, 8)]; Ty = Ci[WS(csi, 8)]; T1 = Cr[0]; T2 = Cr[WS(csr, 16)]; T6 = Cr[WS(csr, 4)]; T7 = Cr[WS(csr, 12)]; { E TB, Tx, T3, TC; TB = Ci[WS(csi, 4)]; Tx = T1 - T2; T3 = T1 + T2; TA = T6 - T7; T8 = T6 + T7; TC = Ci[WS(csi, 12)]; T1t = FMA(KP2_000000000, Ty, Tx); Tz = FNMS(KP2_000000000, Ty, Tx); T1R = FNMS(KP2_000000000, T4, T3); T5 = FMA(KP2_000000000, T4, T3); TD = TB + TC; T1S = TB - TC; } } { E Td, TG, Tc, T1V, TO, Te, TH, TI; { E Ta, Tb, TM, TN; Ta = Cr[WS(csr, 2)]; T1u = TA + TD; TE = TA - TD; Tb = Cr[WS(csr, 14)]; TM = Ci[WS(csi, 2)]; TN = Ci[WS(csi, 14)]; Td = Cr[WS(csr, 10)]; TG = Ta - Tb; Tc = Ta + Tb; T1V = TM - TN; TO = TM + TN; Te = Cr[WS(csr, 6)]; TH = Ci[WS(csi, 10)]; TI = Ci[WS(csi, 6)]; } { E Tl, TS, Tk, T26, T1c, Tm, TT, TU; { E Ti, Tj, T1a, T1b; Ti = Cr[WS(csr, 1)]; { E TL, Tf, T1W, TJ; TL = Td - Te; Tf = Td + Te; T1W = TH - TI; TJ = TH + TI; T1w = TO - TL; TP = TL + TO; T1U = Tc - Tf; Tg = Tc + Tf; T2m = T1W + T1V; T1X = T1V - T1W; T1x = TG + TJ; TK = TG - TJ; Tj = Cr[WS(csr, 15)]; } T1a = Ci[WS(csi, 1)]; T1b = Ci[WS(csi, 15)]; Tl = Cr[WS(csr, 9)]; TS = Ti - Tj; Tk = Ti + Tj; T26 = T1a - T1b; T1c = T1a + T1b; Tm = Cr[WS(csr, 7)]; TT = Ci[WS(csi, 9)]; TU = Ci[WS(csi, 7)]; } { E Ts, TX, Tr, T22, T10, Tt, T13, T14; { E Tp, Tq, TY, TZ; Tp = Cr[WS(csr, 5)]; { E T19, Tn, T27, TV; T19 = Tl - Tm; Tn = Tl + Tm; T27 = TT - TU; TV = TT + TU; T1D = T1c - T19; T1d = T19 + T1c; T20 = Tk - Tn; To = Tk + Tn; T2p = T27 + T26; T28 = T26 - T27; T1A = TS + TV; TW = TS - TV; Tq = Cr[WS(csr, 11)]; } TY = Ci[WS(csi, 5)]; TZ = Ci[WS(csi, 11)]; Ts = Cr[WS(csr, 3)]; TX = Tp - Tq; Tr = Tp + Tq; T22 = TY - TZ; T10 = TY + TZ; Tt = Cr[WS(csr, 13)]; T13 = Ci[WS(csi, 3)]; T14 = Ci[WS(csi, 13)]; } { E T12, Tu, T21, T15; T11 = TX - T10; T1e = TX + T10; T12 = Ts - Tt; Tu = Ts + Tt; T21 = T14 - T13; T15 = T13 + T14; Tv = Tr + Tu; T25 = Tr - Tu; T23 = T21 - T22; T2q = T22 + T21; T16 = T12 - T15; T1f = T12 + T15; } } } } { E T1B, T1E, T1l, T1m, T1p, T1o, T1T, T1Y, T29, T2g, T2j, T2f, T2h, T24; { E T1g, T17, T2n, T2t, T2u, T2s; { E T2o, Tw, T2w, T2r, T2l, T9, Th, T2v; T2o = To - Tv; Tw = To + Tv; T2w = T2q + T2p; T2r = T2p - T2q; T1g = T1e - T1f; T1B = T1e + T1f; T17 = T11 + T16; T1E = T16 - T11; T2l = FNMS(KP2_000000000, T8, T5); T9 = FMA(KP2_000000000, T8, T5); Th = FMA(KP2_000000000, Tg, T9); T2v = FNMS(KP2_000000000, Tg, T9); T2n = FNMS(KP2_000000000, T2m, T2l); T2t = FMA(KP2_000000000, T2m, T2l); R0[WS(rs, 4)] = FNMS(KP2_000000000, T2w, T2v); R0[WS(rs, 12)] = FMA(KP2_000000000, T2w, T2v); R0[0] = FMA(KP2_000000000, Tw, Th); R0[WS(rs, 8)] = FNMS(KP2_000000000, Tw, Th); T2u = T2o + T2r; T2s = T2o - T2r; } { E T1j, TR, T18, T1h, TF, TQ; T1l = FNMS(KP1_414213562, TE, Tz); TF = FMA(KP1_414213562, TE, Tz); TQ = FNMS(KP414213562, TP, TK); T1m = FMA(KP414213562, TK, TP); R0[WS(rs, 2)] = FMA(KP1_414213562, T2s, T2n); R0[WS(rs, 10)] = FNMS(KP1_414213562, T2s, T2n); R0[WS(rs, 6)] = FNMS(KP1_414213562, T2u, T2t); R0[WS(rs, 14)] = FMA(KP1_414213562, T2u, T2t); T1j = FNMS(KP1_847759065, TQ, TF); TR = FMA(KP1_847759065, TQ, TF); T1p = FNMS(KP707106781, T17, TW); T18 = FMA(KP707106781, T17, TW); T1h = FMA(KP707106781, T1g, T1d); T1o = FNMS(KP707106781, T1g, T1d); { E T2d, T2e, T1k, T1i; T1T = FNMS(KP2_000000000, T1S, T1R); T2d = FMA(KP2_000000000, T1S, T1R); T2e = T1U + T1X; T1Y = T1U - T1X; T29 = T25 + T28; T2g = T28 - T25; T1k = FMA(KP198912367, T18, T1h); T1i = FNMS(KP198912367, T1h, T18); T2j = FMA(KP1_414213562, T2e, T2d); T2f = FNMS(KP1_414213562, T2e, T2d); R1[WS(rs, 4)] = FNMS(KP1_961570560, T1k, T1j); R1[WS(rs, 12)] = FMA(KP1_961570560, T1k, T1j); R1[0] = FMA(KP1_961570560, T1i, TR); R1[WS(rs, 8)] = FNMS(KP1_961570560, T1i, TR); T2h = T20 - T23; T24 = T20 + T23; } } } { E T1v, T1y, T1M, T1P, T1L, T1N; { E T1r, T1n, T2k, T2i; T2k = FMA(KP414213562, T2g, T2h); T2i = FNMS(KP414213562, T2h, T2g); T1r = FMA(KP1_847759065, T1m, T1l); T1n = FNMS(KP1_847759065, T1m, T1l); R0[WS(rs, 7)] = FNMS(KP1_847759065, T2k, T2j); R0[WS(rs, 15)] = FMA(KP1_847759065, T2k, T2j); R0[WS(rs, 11)] = FMA(KP1_847759065, T2i, T2f); R0[WS(rs, 3)] = FNMS(KP1_847759065, T2i, T2f); { E T1J, T1K, T1s, T1q; T1v = FNMS(KP1_414213562, T1u, T1t); T1J = FMA(KP1_414213562, T1u, T1t); T1K = FMA(KP414213562, T1w, T1x); T1y = FNMS(KP414213562, T1x, T1w); T1F = FNMS(KP707106781, T1E, T1D); T1M = FMA(KP707106781, T1E, T1D); T1s = FMA(KP668178637, T1o, T1p); T1q = FNMS(KP668178637, T1p, T1o); T1P = FMA(KP1_847759065, T1K, T1J); T1L = FNMS(KP1_847759065, T1K, T1J); R1[WS(rs, 6)] = FNMS(KP1_662939224, T1s, T1r); R1[WS(rs, 14)] = FMA(KP1_662939224, T1s, T1r); R1[WS(rs, 10)] = FMA(KP1_662939224, T1q, T1n); R1[WS(rs, 2)] = FNMS(KP1_662939224, T1q, T1n); T1N = FMA(KP707106781, T1B, T1A); T1C = FNMS(KP707106781, T1B, T1A); } } { E T2b, T1Z, T1Q, T1O, T2c, T2a; T1Q = FMA(KP198912367, T1M, T1N); T1O = FNMS(KP198912367, T1N, T1M); T2b = FNMS(KP1_414213562, T1Y, T1T); T1Z = FMA(KP1_414213562, T1Y, T1T); R1[WS(rs, 7)] = FNMS(KP1_961570560, T1Q, T1P); R1[WS(rs, 15)] = FMA(KP1_961570560, T1Q, T1P); R1[WS(rs, 11)] = FMA(KP1_961570560, T1O, T1L); R1[WS(rs, 3)] = FNMS(KP1_961570560, T1O, T1L); T2c = FMA(KP414213562, T24, T29); T2a = FNMS(KP414213562, T29, T24); T1H = FMA(KP1_847759065, T1y, T1v); T1z = FNMS(KP1_847759065, T1y, T1v); R0[WS(rs, 5)] = FNMS(KP1_847759065, T2c, T2b); R0[WS(rs, 13)] = FMA(KP1_847759065, T2c, T2b); R0[WS(rs, 1)] = FMA(KP1_847759065, T2a, T1Z); R0[WS(rs, 9)] = FNMS(KP1_847759065, T2a, T1Z); } } } } T1G = FNMS(KP668178637, T1F, T1C); T1I = FMA(KP668178637, T1C, T1F); R1[WS(rs, 5)] = FNMS(KP1_662939224, T1I, T1H); R1[WS(rs, 13)] = FMA(KP1_662939224, T1I, T1H); R1[WS(rs, 1)] = FMA(KP1_662939224, T1G, T1z); R1[WS(rs, 9)] = FNMS(KP1_662939224, T1G, T1z); } } } static const kr2c_desc desc = { 32, "r2cb_32", {72, 0, 84, 0}, &GENUS }; void X(codelet_r2cb_32) (planner *p) { X(kr2c_register) (p, r2cb_32, &desc); } #else /* HAVE_FMA */ /* Generated by: ../../../genfft/gen_r2cb.native -compact -variables 4 -pipeline-latency 4 -sign 1 -n 32 -name r2cb_32 -include r2cb.h */ /* * This function contains 156 FP additions, 50 FP multiplications, * (or, 140 additions, 34 multiplications, 16 fused multiply/add), * 54 stack variables, 9 constants, and 64 memory accesses */ #include "r2cb.h" static void r2cb_32(R *R0, R *R1, R *Cr, R *Ci, stride rs, stride csr, stride csi, INT v, INT ivs, INT ovs) { DK(KP1_662939224, +1.662939224605090474157576755235811513477121624); DK(KP1_111140466, +1.111140466039204449485661627897065748749874382); DK(KP1_961570560, +1.961570560806460898252364472268478073947867462); DK(KP390180644, +0.390180644032256535696569736954044481855383236); DK(KP765366864, +0.765366864730179543456919968060797733522689125); DK(KP1_847759065, +1.847759065022573512256366378793576573644833252); DK(KP707106781, +0.707106781186547524400844362104849039284835938); DK(KP1_414213562, +1.414213562373095048801688724209698078569671875); DK(KP2_000000000, +2.000000000000000000000000000000000000000000000); { INT i; for (i = v; i > 0; i = i - 1, R0 = R0 + ovs, R1 = R1 + ovs, Cr = Cr + ivs, Ci = Ci + ivs, MAKE_VOLATILE_STRIDE(128, rs), MAKE_VOLATILE_STRIDE(128, csr), MAKE_VOLATILE_STRIDE(128, csi)) { E T9, T2c, TB, T1y, T6, T2b, Ty, T1v, Th, T2e, T2f, TD, TK, T1C, T1F; E T1h, Tp, T2i, T2m, TN, T13, T1K, T1Y, T1k, Tw, TU, T1l, TW, T1V, T2j; E T1R, T2l; { E T7, T8, T1w, Tz, TA, T1x; T7 = Cr[WS(csr, 4)]; T8 = Cr[WS(csr, 12)]; T1w = T7 - T8; Tz = Ci[WS(csi, 4)]; TA = Ci[WS(csi, 12)]; T1x = Tz + TA; T9 = KP2_000000000 * (T7 + T8); T2c = KP1_414213562 * (T1w + T1x); TB = KP2_000000000 * (Tz - TA); T1y = KP1_414213562 * (T1w - T1x); } { E T5, T1u, T3, T1s; { E T4, T1t, T1, T2; T4 = Cr[WS(csr, 8)]; T5 = KP2_000000000 * T4; T1t = Ci[WS(csi, 8)]; T1u = KP2_000000000 * T1t; T1 = Cr[0]; T2 = Cr[WS(csr, 16)]; T3 = T1 + T2; T1s = T1 - T2; } T6 = T3 + T5; T2b = T1s + T1u; Ty = T3 - T5; T1v = T1s - T1u; } { E Td, T1A, TG, T1E, Tg, T1D, TJ, T1B; { E Tb, Tc, TE, TF; Tb = Cr[WS(csr, 2)]; Tc = Cr[WS(csr, 14)]; Td = Tb + Tc; T1A = Tb - Tc; TE = Ci[WS(csi, 2)]; TF = Ci[WS(csi, 14)]; TG = TE - TF; T1E = TE + TF; } { E Te, Tf, TH, TI; Te = Cr[WS(csr, 10)]; Tf = Cr[WS(csr, 6)]; Tg = Te + Tf; T1D = Te - Tf; TH = Ci[WS(csi, 10)]; TI = Ci[WS(csi, 6)]; TJ = TH - TI; T1B = TH + TI; } Th = KP2_000000000 * (Td + Tg); T2e = T1A + T1B; T2f = T1E - T1D; TD = Td - Tg; TK = TG - TJ; T1C = T1A - T1B; T1F = T1D + T1E; T1h = KP2_000000000 * (TJ + TG); } { E Tl, T1I, TZ, T1X, To, T1W, T12, T1J; { E Tj, Tk, TX, TY; Tj = Cr[WS(csr, 1)]; Tk = Cr[WS(csr, 15)]; Tl = Tj + Tk; T1I = Tj - Tk; TX = Ci[WS(csi, 1)]; TY = Ci[WS(csi, 15)]; TZ = TX - TY; T1X = TX + TY; } { E Tm, Tn, T10, T11; Tm = Cr[WS(csr, 9)]; Tn = Cr[WS(csr, 7)]; To = Tm + Tn; T1W = Tm - Tn; T10 = Ci[WS(csi, 9)]; T11 = Ci[WS(csi, 7)]; T12 = T10 - T11; T1J = T10 + T11; } Tp = Tl + To; T2i = T1I + T1J; T2m = T1X - T1W; TN = Tl - To; T13 = TZ - T12; T1K = T1I - T1J; T1Y = T1W + T1X; T1k = T12 + TZ; } { E Ts, T1L, TT, T1M, Tv, T1O, TQ, T1P; { E Tq, Tr, TR, TS; Tq = Cr[WS(csr, 5)]; Tr = Cr[WS(csr, 11)]; Ts = Tq + Tr; T1L = Tq - Tr; TR = Ci[WS(csi, 5)]; TS = Ci[WS(csi, 11)]; TT = TR - TS; T1M = TR + TS; } { E Tt, Tu, TO, TP; Tt = Cr[WS(csr, 3)]; Tu = Cr[WS(csr, 13)]; Tv = Tt + Tu; T1O = Tt - Tu; TO = Ci[WS(csi, 13)]; TP = Ci[WS(csi, 3)]; TQ = TO - TP; T1P = TP + TO; } Tw = Ts + Tv; TU = TQ - TT; T1l = TT + TQ; TW = Ts - Tv; { E T1T, T1U, T1N, T1Q; T1T = T1L + T1M; T1U = T1O + T1P; T1V = KP707106781 * (T1T - T1U); T2j = KP707106781 * (T1T + T1U); T1N = T1L - T1M; T1Q = T1O - T1P; T1R = KP707106781 * (T1N + T1Q); T2l = KP707106781 * (T1N - T1Q); } } { E Tx, T1r, Ti, T1q, Ta; Tx = KP2_000000000 * (Tp + Tw); T1r = KP2_000000000 * (T1l + T1k); Ta = T6 + T9; Ti = Ta + Th; T1q = Ta - Th; R0[WS(rs, 8)] = Ti - Tx; R0[WS(rs, 12)] = T1q + T1r; R0[0] = Ti + Tx; R0[WS(rs, 4)] = T1q - T1r; } { E T1i, T1o, T1n, T1p, T1g, T1j, T1m; T1g = T6 - T9; T1i = T1g - T1h; T1o = T1g + T1h; T1j = Tp - Tw; T1m = T1k - T1l; T1n = KP1_414213562 * (T1j - T1m); T1p = KP1_414213562 * (T1j + T1m); R0[WS(rs, 10)] = T1i - T1n; R0[WS(rs, 14)] = T1o + T1p; R0[WS(rs, 2)] = T1i + T1n; R0[WS(rs, 6)] = T1o - T1p; } { E TM, T16, T15, T17; { E TC, TL, TV, T14; TC = Ty - TB; TL = KP1_414213562 * (TD - TK); TM = TC + TL; T16 = TC - TL; TV = TN + TU; T14 = TW + T13; T15 = FNMS(KP765366864, T14, KP1_847759065 * TV); T17 = FMA(KP765366864, TV, KP1_847759065 * T14); } R0[WS(rs, 9)] = TM - T15; R0[WS(rs, 13)] = T16 + T17; R0[WS(rs, 1)] = TM + T15; R0[WS(rs, 5)] = T16 - T17; } { E T2t, T2x, T2w, T2y; { E T2r, T2s, T2u, T2v; T2r = T2b + T2c; T2s = FMA(KP1_847759065, T2e, KP765366864 * T2f); T2t = T2r - T2s; T2x = T2r + T2s; T2u = T2i + T2j; T2v = T2m - T2l; T2w = FNMS(KP1_961570560, T2v, KP390180644 * T2u); T2y = FMA(KP1_961570560, T2u, KP390180644 * T2v); } R1[WS(rs, 11)] = T2t - T2w; R1[WS(rs, 15)] = T2x + T2y; R1[WS(rs, 3)] = T2t + T2w; R1[WS(rs, 7)] = T2x - T2y; } { E T1a, T1e, T1d, T1f; { E T18, T19, T1b, T1c; T18 = Ty + TB; T19 = KP1_414213562 * (TD + TK); T1a = T18 - T19; T1e = T18 + T19; T1b = TN - TU; T1c = T13 - TW; T1d = FNMS(KP1_847759065, T1c, KP765366864 * T1b); T1f = FMA(KP1_847759065, T1b, KP765366864 * T1c); } R0[WS(rs, 11)] = T1a - T1d; R0[WS(rs, 15)] = T1e + T1f; R0[WS(rs, 3)] = T1a + T1d; R0[WS(rs, 7)] = T1e - T1f; } { E T25, T29, T28, T2a; { E T23, T24, T26, T27; T23 = T1v - T1y; T24 = FMA(KP765366864, T1C, KP1_847759065 * T1F); T25 = T23 - T24; T29 = T23 + T24; T26 = T1K - T1R; T27 = T1Y - T1V; T28 = FNMS(KP1_662939224, T27, KP1_111140466 * T26); T2a = FMA(KP1_662939224, T26, KP1_111140466 * T27); } R1[WS(rs, 10)] = T25 - T28; R1[WS(rs, 14)] = T29 + T2a; R1[WS(rs, 2)] = T25 + T28; R1[WS(rs, 6)] = T29 - T2a; } { E T2h, T2p, T2o, T2q; { E T2d, T2g, T2k, T2n; T2d = T2b - T2c; T2g = FNMS(KP1_847759065, T2f, KP765366864 * T2e); T2h = T2d + T2g; T2p = T2d - T2g; T2k = T2i - T2j; T2n = T2l + T2m; T2o = FNMS(KP1_111140466, T2n, KP1_662939224 * T2k); T2q = FMA(KP1_111140466, T2k, KP1_662939224 * T2n); } R1[WS(rs, 9)] = T2h - T2o; R1[WS(rs, 13)] = T2p + T2q; R1[WS(rs, 1)] = T2h + T2o; R1[WS(rs, 5)] = T2p - T2q; } { E T1H, T21, T20, T22; { E T1z, T1G, T1S, T1Z; T1z = T1v + T1y; T1G = FNMS(KP765366864, T1F, KP1_847759065 * T1C); T1H = T1z + T1G; T21 = T1z - T1G; T1S = T1K + T1R; T1Z = T1V + T1Y; T20 = FNMS(KP390180644, T1Z, KP1_961570560 * T1S); T22 = FMA(KP390180644, T1S, KP1_961570560 * T1Z); } R1[WS(rs, 8)] = T1H - T20; R1[WS(rs, 12)] = T21 + T22; R1[0] = T1H + T20; R1[WS(rs, 4)] = T21 - T22; } } } } static const kr2c_desc desc = { 32, "r2cb_32", {140, 34, 16, 0}, &GENUS }; void X(codelet_r2cb_32) (planner *p) { X(kr2c_register) (p, r2cb_32, &desc); } #endif /* HAVE_FMA */ fftw-3.3.3/rdft/scalar/r2cb/r2cb_64.c0000644000175000017500000012146612054410763013764 00000000000000/* * Copyright (c) 2003, 2007-11 Matteo Frigo * Copyright (c) 2003, 2007-11 Massachusetts Institute of Technology * * This program is free software; you can redistribute it and/or modify * it under the terms of the GNU General Public License as published by * the Free Software Foundation; either version 2 of the License, or * (at your option) any later version. * * This program is distributed in the hope that it will be useful, * but WITHOUT ANY WARRANTY; without even the implied warranty of * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the * GNU General Public License for more details. * * You should have received a copy of the GNU General Public License * along with this program; if not, write to the Free Software * Foundation, Inc., 51 Franklin Street, Fifth Floor, Boston, MA 02110-1301 USA * */ /* This file was automatically generated --- DO NOT EDIT */ /* Generated on Sun Nov 25 07:41:09 EST 2012 */ #include "codelet-rdft.h" #ifdef HAVE_FMA /* Generated by: ../../../genfft/gen_r2cb.native -fma -reorder-insns -schedule-for-pipeline -compact -variables 4 -pipeline-latency 4 -sign 1 -n 64 -name r2cb_64 -include r2cb.h */ /* * This function contains 394 FP additions, 216 FP multiplications, * (or, 178 additions, 0 multiplications, 216 fused multiply/add), * 143 stack variables, 18 constants, and 128 memory accesses */ #include "r2cb.h" static void r2cb_64(R *R0, R *R1, R *Cr, R *Ci, stride rs, stride csr, stride csi, INT v, INT ivs, INT ovs) { DK(KP1_990369453, +1.990369453344393772489673906218959843150949737); DK(KP1_546020906, +1.546020906725473921621813219516939601942082586); DK(KP098491403, +0.098491403357164253077197521291327432293052451); DK(KP820678790, +0.820678790828660330972281985331011598767386482); DK(KP1_913880671, +1.913880671464417729871595773960539938965698411); DK(KP1_763842528, +1.763842528696710059425513727320776699016885241); DK(KP303346683, +0.303346683607342391675883946941299872384187453); DK(KP534511135, +0.534511135950791641089685961295362908582039528); DK(KP923879532, +0.923879532511286756128183189396788286822416626); DK(KP1_662939224, +1.662939224605090474157576755235811513477121624); DK(KP1_961570560, +1.961570560806460898252364472268478073947867462); DK(KP668178637, +0.668178637919298919997757686523080761552472251); DK(KP198912367, +0.198912367379658006911597622644676228597850501); DK(KP707106781, +0.707106781186547524400844362104849039284835938); DK(KP1_847759065, +1.847759065022573512256366378793576573644833252); DK(KP414213562, +0.414213562373095048801688724209698078569671875); DK(KP1_414213562, +1.414213562373095048801688724209698078569671875); DK(KP2_000000000, +2.000000000000000000000000000000000000000000000); { INT i; for (i = v; i > 0; i = i - 1, R0 = R0 + ovs, R1 = R1 + ovs, Cr = Cr + ivs, Ci = Ci + ivs, MAKE_VOLATILE_STRIDE(256, rs), MAKE_VOLATILE_STRIDE(256, csr), MAKE_VOLATILE_STRIDE(256, csi)) { E T3d, T32, T37, T2Z, T3f, T3b, T3c, T35; { E T5H, T9, T5j, T4p, T2T, T1b, T3Z, T3j, Tg, T5I, T5k, T4u, T40, T3m, T2U; E T1m, T3o, T1s, T1J, T3r, T5K, Tw, T5N, T6c, T4A, T5n, T3s, T1D, T5m, T4F; E T3p, T1M, T3w, T1U, T2z, T3H, T5Q, TM, T6f, T5Y, T5q, T4M, T3I, T25, T5t; E T53, T3x, T2C, T3A, T5V, T11, T6g, T5T, T55, T4W, T3z, T2E, T2h, T2F, T2s; E T3L, T3E, T54, T4R; { E Td, T1c, Tc, T4r, T1k, Te, T1d, T1e; { E T3h, T15, T1a, T3i; { E T4, T14, T17, T13, T3, T16, T8, T18; T4 = Cr[WS(csr, 16)]; T14 = Ci[WS(csi, 16)]; { E T1, T2, T6, T7; T1 = Cr[0]; T2 = Cr[WS(csr, 32)]; T6 = Cr[WS(csr, 8)]; T7 = Cr[WS(csr, 24)]; T17 = Ci[WS(csi, 8)]; T13 = T1 - T2; T3 = T1 + T2; T16 = T6 - T7; T8 = T6 + T7; T18 = Ci[WS(csi, 24)]; } { E T4n, T5, T4o, T19; T4n = FNMS(KP2_000000000, T4, T3); T5 = FMA(KP2_000000000, T4, T3); T3h = FMA(KP2_000000000, T14, T13); T15 = FNMS(KP2_000000000, T14, T13); T4o = T17 - T18; T19 = T17 + T18; T5H = FNMS(KP2_000000000, T8, T5); T9 = FMA(KP2_000000000, T8, T5); T5j = FMA(KP2_000000000, T4o, T4n); T4p = FNMS(KP2_000000000, T4o, T4n); T1a = T16 - T19; T3i = T16 + T19; } } { E Ta, Tb, T1i, T1j; Ta = Cr[WS(csr, 4)]; T2T = FNMS(KP1_414213562, T1a, T15); T1b = FMA(KP1_414213562, T1a, T15); T3Z = FMA(KP1_414213562, T3i, T3h); T3j = FNMS(KP1_414213562, T3i, T3h); Tb = Cr[WS(csr, 28)]; T1i = Ci[WS(csi, 4)]; T1j = Ci[WS(csi, 28)]; Td = Cr[WS(csr, 20)]; T1c = Ta - Tb; Tc = Ta + Tb; T4r = T1i - T1j; T1k = T1i + T1j; Te = Cr[WS(csr, 12)]; T1d = Ci[WS(csi, 20)]; T1e = Ci[WS(csi, 12)]; } } { E T4B, T4E, T1K, T1L; { E T1o, Tk, T4C, T1I, T1F, Tn, T4D, T1r, Ts, T1t, Tr, T4y, T1w, Tt, T1z; E T1A; { E Tl, Tm, T1p, T1q; { E Ti, Tj, T1G, T1H, T1h, Tf; Ti = Cr[WS(csr, 2)]; T1h = Td - Te; Tf = Td + Te; { E T4s, T1f, T3k, T1l; T4s = T1d - T1e; T1f = T1d + T1e; T3k = T1k - T1h; T1l = T1h + T1k; { E T4q, T4t, T3l, T1g; T4q = Tc - Tf; Tg = Tc + Tf; T4t = T4r - T4s; T5I = T4s + T4r; T3l = T1c + T1f; T1g = T1c - T1f; T5k = T4q + T4t; T4u = T4q - T4t; T40 = FMA(KP414213562, T3k, T3l); T3m = FNMS(KP414213562, T3l, T3k); T2U = FMA(KP414213562, T1g, T1l); T1m = FNMS(KP414213562, T1l, T1g); Tj = Cr[WS(csr, 30)]; } } T1G = Ci[WS(csi, 2)]; T1H = Ci[WS(csi, 30)]; Tl = Cr[WS(csr, 18)]; T1o = Ti - Tj; Tk = Ti + Tj; T4C = T1G - T1H; T1I = T1G + T1H; Tm = Cr[WS(csr, 14)]; T1p = Ci[WS(csi, 18)]; T1q = Ci[WS(csi, 14)]; } { E Tp, Tq, T1u, T1v; Tp = Cr[WS(csr, 10)]; T1F = Tl - Tm; Tn = Tl + Tm; T4D = T1p - T1q; T1r = T1p + T1q; Tq = Cr[WS(csr, 22)]; T1u = Ci[WS(csi, 10)]; T1v = Ci[WS(csi, 22)]; Ts = Cr[WS(csr, 6)]; T1t = Tp - Tq; Tr = Tp + Tq; T4y = T1u - T1v; T1w = T1u + T1v; Tt = Cr[WS(csr, 26)]; T1z = Ci[WS(csi, 6)]; T1A = Ci[WS(csi, 26)]; } } { E T1y, T4x, T1B, T4w, To, Tv, Tu; T3o = T1o + T1r; T1s = T1o - T1r; T1y = Ts - Tt; Tu = Ts + Tt; T4x = T1A - T1z; T1B = T1z + T1A; T1J = T1F + T1I; T3r = T1I - T1F; T4w = Tk - Tn; To = Tk + Tn; Tv = Tr + Tu; T4B = Tr - Tu; { E T4z, T5L, T5M, T1x, T1C; T4E = T4C - T4D; T5L = T4D + T4C; T5M = T4y + T4x; T4z = T4x - T4y; T5K = To - Tv; Tw = To + Tv; T5N = T5L - T5M; T6c = T5M + T5L; T1K = T1t + T1w; T1x = T1t - T1w; T1C = T1y - T1B; T1L = T1y + T1B; T4A = T4w + T4z; T5n = T4w - T4z; T3s = T1C - T1x; T1D = T1x + T1C; } } } { E T4Z, T52, T2A, T2B; { E T1Q, TA, T50, T2y, T2v, TD, T51, T1T, TI, T1V, TH, T4K, T1Y, TJ, T21; E T22; { E TB, TC, T1R, T1S; { E Ty, Tz, T2w, T2x; Ty = Cr[WS(csr, 1)]; T5m = T4E - T4B; T4F = T4B + T4E; T3p = T1K + T1L; T1M = T1K - T1L; Tz = Cr[WS(csr, 31)]; T2w = Ci[WS(csi, 1)]; T2x = Ci[WS(csi, 31)]; TB = Cr[WS(csr, 17)]; T1Q = Ty - Tz; TA = Ty + Tz; T50 = T2w - T2x; T2y = T2w + T2x; TC = Cr[WS(csr, 15)]; T1R = Ci[WS(csi, 17)]; T1S = Ci[WS(csi, 15)]; } { E TF, TG, T1W, T1X; TF = Cr[WS(csr, 9)]; T2v = TB - TC; TD = TB + TC; T51 = T1R - T1S; T1T = T1R + T1S; TG = Cr[WS(csr, 23)]; T1W = Ci[WS(csi, 9)]; T1X = Ci[WS(csi, 23)]; TI = Cr[WS(csr, 7)]; T1V = TF - TG; TH = TF + TG; T4K = T1W - T1X; T1Y = T1W + T1X; TJ = Cr[WS(csr, 25)]; T21 = Ci[WS(csi, 7)]; T22 = Ci[WS(csi, 25)]; } } { E T20, T4J, T23, T4I, TE, TL, TK; T3w = T1Q + T1T; T1U = T1Q - T1T; T20 = TI - TJ; TK = TI + TJ; T4J = T22 - T21; T23 = T21 + T22; T2z = T2v + T2y; T3H = T2y - T2v; T4I = TA - TD; TE = TA + TD; TL = TH + TK; T4Z = TH - TK; { E T4L, T5W, T5X, T1Z, T24; T52 = T50 - T51; T5W = T51 + T50; T5X = T4K + T4J; T4L = T4J - T4K; T5Q = TE - TL; TM = TE + TL; T6f = T5X + T5W; T5Y = T5W - T5X; T2A = T1V + T1Y; T1Z = T1V - T1Y; T24 = T20 - T23; T2B = T20 + T23; T5q = T4I - T4L; T4M = T4I + T4L; T3I = T24 - T1Z; T25 = T1Z + T24; } } } { E T27, TP, T4O, T2f, T2c, TS, T4P, T2a, TX, T2i, TW, T4T, T2q, TY, T2j; E T2k; { E TQ, TR, T28, T29; { E TN, TO, T2d, T2e; TN = Cr[WS(csr, 5)]; T5t = T52 - T4Z; T53 = T4Z + T52; T3x = T2A + T2B; T2C = T2A - T2B; TO = Cr[WS(csr, 27)]; T2d = Ci[WS(csi, 5)]; T2e = Ci[WS(csi, 27)]; TQ = Cr[WS(csr, 21)]; T27 = TN - TO; TP = TN + TO; T4O = T2d - T2e; T2f = T2d + T2e; TR = Cr[WS(csr, 11)]; T28 = Ci[WS(csi, 21)]; T29 = Ci[WS(csi, 11)]; } { E TU, TV, T2o, T2p; TU = Cr[WS(csr, 3)]; T2c = TQ - TR; TS = TQ + TR; T4P = T28 - T29; T2a = T28 + T29; TV = Cr[WS(csr, 29)]; T2o = Ci[WS(csi, 3)]; T2p = Ci[WS(csi, 29)]; TX = Cr[WS(csr, 13)]; T2i = TU - TV; TW = TU + TV; T4T = T2p - T2o; T2q = T2o + T2p; TY = Cr[WS(csr, 19)]; T2j = Ci[WS(csi, 13)]; T2k = Ci[WS(csi, 19)]; } } { E T4N, T2n, T2l, T4Q, T2b, T2g, TT, TZ, T4U; T4N = TP - TS; TT = TP + TS; T2n = TX - TY; TZ = TX + TY; T4U = T2j - T2k; T2l = T2j + T2k; { E T5S, T10, T4S, T4V, T5R; T5S = T4P + T4O; T4Q = T4O - T4P; T10 = TW + TZ; T4S = TW - TZ; T4V = T4T - T4U; T5R = T4U + T4T; T3A = T27 + T2a; T2b = T27 - T2a; T5V = TT - T10; T11 = TT + T10; T6g = T5S + T5R; T5T = T5R - T5S; T55 = T4V - T4S; T4W = T4S + T4V; T2g = T2c + T2f; T3z = T2f - T2c; } { E T3D, T3C, T2m, T2r; T3D = T2i + T2l; T2m = T2i - T2l; T2r = T2n - T2q; T3C = T2n + T2q; T2E = FMA(KP414213562, T2b, T2g); T2h = FNMS(KP414213562, T2g, T2b); T2F = FNMS(KP414213562, T2m, T2r); T2s = FMA(KP414213562, T2r, T2m); T3L = FMA(KP414213562, T3C, T3D); T3E = FNMS(KP414213562, T3D, T3C); T54 = T4N + T4Q; T4R = T4N - T4Q; } } } } } } { E T3K, T3B, T5u, T5r, T5d, T5g; { E T6e, T6h, T6b, T5J, T5O, T5Z, T66, T69, T65, T67, T5U, T12, T6m, Th; T6e = TM - T11; T12 = TM + T11; T6m = T6g + T6f; T6h = T6f - T6g; T6b = FNMS(KP2_000000000, Tg, T9); Th = FMA(KP2_000000000, Tg, T9); T3K = FMA(KP414213562, T3z, T3A); T3B = FNMS(KP414213562, T3A, T3z); { E T63, T64, T6l, Tx; T5J = FNMS(KP2_000000000, T5I, T5H); T63 = FMA(KP2_000000000, T5I, T5H); T64 = T5K + T5N; T5O = T5K - T5N; T5Z = T5V + T5Y; T66 = T5Y - T5V; T6l = FNMS(KP2_000000000, Tw, Th); Tx = FMA(KP2_000000000, Tw, Th); T69 = FMA(KP1_414213562, T64, T63); T65 = FNMS(KP1_414213562, T64, T63); R0[WS(rs, 8)] = FNMS(KP2_000000000, T6m, T6l); R0[WS(rs, 24)] = FMA(KP2_000000000, T6m, T6l); R0[0] = FMA(KP2_000000000, T12, Tx); R0[WS(rs, 16)] = FNMS(KP2_000000000, T12, Tx); T67 = T5Q - T5T; T5U = T5Q + T5T; } { E T6j, T6d, T6a, T68; T6a = FMA(KP414213562, T66, T67); T68 = FNMS(KP414213562, T67, T66); T6j = FMA(KP2_000000000, T6c, T6b); T6d = FNMS(KP2_000000000, T6c, T6b); R0[WS(rs, 14)] = FNMS(KP1_847759065, T6a, T69); R0[WS(rs, 30)] = FMA(KP1_847759065, T6a, T69); R0[WS(rs, 22)] = FMA(KP1_847759065, T68, T65); R0[WS(rs, 6)] = FNMS(KP1_847759065, T68, T65); { E T61, T5P, T6k, T6i; T6k = T6e + T6h; T6i = T6e - T6h; T61 = FNMS(KP1_414213562, T5O, T5J); T5P = FMA(KP1_414213562, T5O, T5J); R0[WS(rs, 12)] = FNMS(KP1_414213562, T6k, T6j); R0[WS(rs, 28)] = FMA(KP1_414213562, T6k, T6j); R0[WS(rs, 4)] = FMA(KP1_414213562, T6i, T6d); R0[WS(rs, 20)] = FNMS(KP1_414213562, T6i, T6d); { E T5b, T4v, T5f, T4Y, T5e, T57, T4G, T5c; { E T4X, T56, T62, T60; T5u = T4W - T4R; T4X = T4R + T4W; T56 = T54 + T55; T5r = T54 - T55; T5b = FNMS(KP1_414213562, T4u, T4p); T4v = FMA(KP1_414213562, T4u, T4p); T62 = FMA(KP414213562, T5U, T5Z); T60 = FNMS(KP414213562, T5Z, T5U); T5f = FNMS(KP707106781, T4X, T4M); T4Y = FMA(KP707106781, T4X, T4M); T5e = FNMS(KP707106781, T56, T53); T57 = FMA(KP707106781, T56, T53); R0[WS(rs, 10)] = FNMS(KP1_847759065, T62, T61); R0[WS(rs, 26)] = FMA(KP1_847759065, T62, T61); R0[WS(rs, 2)] = FMA(KP1_847759065, T60, T5P); R0[WS(rs, 18)] = FNMS(KP1_847759065, T60, T5P); T4G = FNMS(KP414213562, T4F, T4A); T5c = FMA(KP414213562, T4A, T4F); } { E T5a, T59, T5h, T5i, T58, T4H; T5a = FMA(KP198912367, T4Y, T57); T58 = FNMS(KP198912367, T57, T4Y); T59 = FNMS(KP1_847759065, T4G, T4v); T4H = FMA(KP1_847759065, T4G, T4v); T5h = FMA(KP1_847759065, T5c, T5b); T5d = FNMS(KP1_847759065, T5c, T5b); T5i = FMA(KP668178637, T5e, T5f); T5g = FNMS(KP668178637, T5f, T5e); R0[WS(rs, 1)] = FMA(KP1_961570560, T58, T4H); R0[WS(rs, 17)] = FNMS(KP1_961570560, T58, T4H); R0[WS(rs, 29)] = FMA(KP1_662939224, T5i, T5h); R0[WS(rs, 13)] = FNMS(KP1_662939224, T5i, T5h); R0[WS(rs, 25)] = FMA(KP1_961570560, T5a, T59); R0[WS(rs, 9)] = FNMS(KP1_961570560, T5a, T59); } } } } } { E T43, T42, T46, T4a, T49, T3V, T3G, T47, T3P, T3v, T3X, T3T, T3U, T3N, T5B; E T5E; { E T5s, T5D, T5z, T5l, T5C, T5v, T5o, T5A; R0[WS(rs, 21)] = FMA(KP1_662939224, T5g, T5d); R0[WS(rs, 5)] = FNMS(KP1_662939224, T5g, T5d); T5s = FNMS(KP707106781, T5r, T5q); T5D = FMA(KP707106781, T5r, T5q); T5z = FMA(KP1_414213562, T5k, T5j); T5l = FNMS(KP1_414213562, T5k, T5j); T5C = FMA(KP707106781, T5u, T5t); T5v = FNMS(KP707106781, T5u, T5t); T5o = FNMS(KP414213562, T5n, T5m); T5A = FMA(KP414213562, T5m, T5n); { E T5y, T5x, T5F, T5G, T5w, T5p; T5y = FMA(KP668178637, T5s, T5v); T5w = FNMS(KP668178637, T5v, T5s); T5x = FMA(KP1_847759065, T5o, T5l); T5p = FNMS(KP1_847759065, T5o, T5l); T5F = FMA(KP1_847759065, T5A, T5z); T5B = FNMS(KP1_847759065, T5A, T5z); T5G = FMA(KP198912367, T5C, T5D); T5E = FNMS(KP198912367, T5D, T5C); R0[WS(rs, 3)] = FMA(KP1_662939224, T5w, T5p); R0[WS(rs, 19)] = FNMS(KP1_662939224, T5w, T5p); R0[WS(rs, 31)] = FMA(KP1_961570560, T5G, T5F); R0[WS(rs, 15)] = FNMS(KP1_961570560, T5G, T5F); R0[WS(rs, 27)] = FMA(KP1_662939224, T5y, T5x); R0[WS(rs, 11)] = FNMS(KP1_662939224, T5y, T5x); } } { E T3R, T3n, T3J, T3S, T3u, T3M; T3R = FMA(KP1_847759065, T3m, T3j); T3n = FNMS(KP1_847759065, T3m, T3j); R0[WS(rs, 23)] = FMA(KP1_961570560, T5E, T5B); R0[WS(rs, 7)] = FNMS(KP1_961570560, T5E, T5B); { E T3q, T3t, T3y, T3F; T43 = FMA(KP707106781, T3p, T3o); T3q = FNMS(KP707106781, T3p, T3o); T3t = FNMS(KP707106781, T3s, T3r); T42 = FMA(KP707106781, T3s, T3r); T46 = FMA(KP707106781, T3x, T3w); T3y = FNMS(KP707106781, T3x, T3w); T3F = T3B + T3E; T4a = T3B - T3E; T49 = FMA(KP707106781, T3I, T3H); T3J = FNMS(KP707106781, T3I, T3H); T3S = FMA(KP668178637, T3q, T3t); T3u = FNMS(KP668178637, T3t, T3q); T3V = FMA(KP923879532, T3F, T3y); T3G = FNMS(KP923879532, T3F, T3y); T3M = T3K - T3L; T47 = T3K + T3L; } T3P = FNMS(KP1_662939224, T3u, T3n); T3v = FMA(KP1_662939224, T3u, T3n); T3X = FMA(KP1_662939224, T3S, T3R); T3T = FNMS(KP1_662939224, T3S, T3R); T3U = FNMS(KP923879532, T3M, T3J); T3N = FMA(KP923879532, T3M, T3J); } { E T2X, T2W, T30, T34, T33, T2P, T2u, T31, T2J, T1P, T2R, T2N, T2O, T2H; { E T2L, T1n, T2D, T2M, T1O, T2G; T2L = FNMS(KP1_847759065, T1m, T1b); T1n = FMA(KP1_847759065, T1m, T1b); { E T3W, T3Y, T3Q, T3O; T3W = FNMS(KP534511135, T3V, T3U); T3Y = FMA(KP534511135, T3U, T3V); T3Q = FMA(KP303346683, T3G, T3N); T3O = FNMS(KP303346683, T3N, T3G); R1[WS(rs, 21)] = FMA(KP1_763842528, T3W, T3T); R1[WS(rs, 5)] = FNMS(KP1_763842528, T3W, T3T); R1[WS(rs, 29)] = FMA(KP1_763842528, T3Y, T3X); R1[WS(rs, 13)] = FNMS(KP1_763842528, T3Y, T3X); R1[WS(rs, 25)] = FMA(KP1_913880671, T3Q, T3P); R1[WS(rs, 9)] = FNMS(KP1_913880671, T3Q, T3P); R1[WS(rs, 1)] = FMA(KP1_913880671, T3O, T3v); R1[WS(rs, 17)] = FNMS(KP1_913880671, T3O, T3v); } { E T1E, T1N, T26, T2t; T2X = FNMS(KP707106781, T1D, T1s); T1E = FMA(KP707106781, T1D, T1s); T1N = FMA(KP707106781, T1M, T1J); T2W = FNMS(KP707106781, T1M, T1J); T30 = FNMS(KP707106781, T25, T1U); T26 = FMA(KP707106781, T25, T1U); T2t = T2h + T2s; T34 = T2s - T2h; T33 = FNMS(KP707106781, T2C, T2z); T2D = FMA(KP707106781, T2C, T2z); T2M = FMA(KP198912367, T1E, T1N); T1O = FNMS(KP198912367, T1N, T1E); T2P = FNMS(KP923879532, T2t, T26); T2u = FMA(KP923879532, T2t, T26); T2G = T2E + T2F; T31 = T2E - T2F; } T2J = FNMS(KP1_961570560, T1O, T1n); T1P = FMA(KP1_961570560, T1O, T1n); T2R = FMA(KP1_961570560, T2M, T2L); T2N = FNMS(KP1_961570560, T2M, T2L); T2O = FNMS(KP923879532, T2G, T2D); T2H = FMA(KP923879532, T2G, T2D); } { E T4j, T48, T4d, T45, T4l, T4h, T4i, T4b; { E T4f, T41, T4g, T44; T4f = FMA(KP1_847759065, T40, T3Z); T41 = FNMS(KP1_847759065, T40, T3Z); { E T2Q, T2S, T2K, T2I; T2Q = FNMS(KP820678790, T2P, T2O); T2S = FMA(KP820678790, T2O, T2P); T2K = FMA(KP098491403, T2u, T2H); T2I = FNMS(KP098491403, T2H, T2u); R1[WS(rs, 20)] = FMA(KP1_546020906, T2Q, T2N); R1[WS(rs, 4)] = FNMS(KP1_546020906, T2Q, T2N); R1[WS(rs, 28)] = FMA(KP1_546020906, T2S, T2R); R1[WS(rs, 12)] = FNMS(KP1_546020906, T2S, T2R); R1[WS(rs, 24)] = FMA(KP1_990369453, T2K, T2J); R1[WS(rs, 8)] = FNMS(KP1_990369453, T2K, T2J); R1[0] = FMA(KP1_990369453, T2I, T1P); R1[WS(rs, 16)] = FNMS(KP1_990369453, T2I, T1P); } T4g = FMA(KP198912367, T42, T43); T44 = FNMS(KP198912367, T43, T42); T4j = FMA(KP923879532, T47, T46); T48 = FNMS(KP923879532, T47, T46); T4d = FMA(KP1_961570560, T44, T41); T45 = FNMS(KP1_961570560, T44, T41); T4l = FMA(KP1_961570560, T4g, T4f); T4h = FNMS(KP1_961570560, T4g, T4f); T4i = FMA(KP923879532, T4a, T49); T4b = FNMS(KP923879532, T4a, T49); } { E T39, T2V, T3a, T2Y; T39 = FMA(KP1_847759065, T2U, T2T); T2V = FNMS(KP1_847759065, T2U, T2T); { E T4k, T4m, T4e, T4c; T4k = FNMS(KP098491403, T4j, T4i); T4m = FMA(KP098491403, T4i, T4j); T4e = FMA(KP820678790, T48, T4b); T4c = FNMS(KP820678790, T4b, T48); R1[WS(rs, 23)] = FMA(KP1_990369453, T4k, T4h); R1[WS(rs, 7)] = FNMS(KP1_990369453, T4k, T4h); R1[WS(rs, 31)] = FMA(KP1_990369453, T4m, T4l); R1[WS(rs, 15)] = FNMS(KP1_990369453, T4m, T4l); R1[WS(rs, 27)] = FMA(KP1_546020906, T4e, T4d); R1[WS(rs, 11)] = FNMS(KP1_546020906, T4e, T4d); R1[WS(rs, 3)] = FMA(KP1_546020906, T4c, T45); R1[WS(rs, 19)] = FNMS(KP1_546020906, T4c, T45); } T3a = FMA(KP668178637, T2W, T2X); T2Y = FNMS(KP668178637, T2X, T2W); T3d = FMA(KP923879532, T31, T30); T32 = FNMS(KP923879532, T31, T30); T37 = FMA(KP1_662939224, T2Y, T2V); T2Z = FNMS(KP1_662939224, T2Y, T2V); T3f = FMA(KP1_662939224, T3a, T39); T3b = FNMS(KP1_662939224, T3a, T39); T3c = FMA(KP923879532, T34, T33); T35 = FNMS(KP923879532, T34, T33); } } } } } } { E T3g, T3e, T36, T38; T3g = FMA(KP303346683, T3c, T3d); T3e = FNMS(KP303346683, T3d, T3c); T36 = FNMS(KP534511135, T35, T32); T38 = FMA(KP534511135, T32, T35); R1[WS(rs, 22)] = FMA(KP1_913880671, T3e, T3b); R1[WS(rs, 6)] = FNMS(KP1_913880671, T3e, T3b); R1[WS(rs, 30)] = FMA(KP1_913880671, T3g, T3f); R1[WS(rs, 14)] = FNMS(KP1_913880671, T3g, T3f); R1[WS(rs, 26)] = FMA(KP1_763842528, T38, T37); R1[WS(rs, 10)] = FNMS(KP1_763842528, T38, T37); R1[WS(rs, 2)] = FMA(KP1_763842528, T36, T2Z); R1[WS(rs, 18)] = FNMS(KP1_763842528, T36, T2Z); } } } } static const kr2c_desc desc = { 64, "r2cb_64", {178, 0, 216, 0}, &GENUS }; void X(codelet_r2cb_64) (planner *p) { X(kr2c_register) (p, r2cb_64, &desc); } #else /* HAVE_FMA */ /* Generated by: ../../../genfft/gen_r2cb.native -compact -variables 4 -pipeline-latency 4 -sign 1 -n 64 -name r2cb_64 -include r2cb.h */ /* * This function contains 394 FP additions, 134 FP multiplications, * (or, 342 additions, 82 multiplications, 52 fused multiply/add), * 110 stack variables, 19 constants, and 128 memory accesses */ #include "r2cb.h" static void r2cb_64(R *R0, R *R1, R *Cr, R *Ci, stride rs, stride csr, stride csi, INT v, INT ivs, INT ovs) { DK(KP1_268786568, +1.268786568327290996430343226450986741351374190); DK(KP1_546020906, +1.546020906725473921621813219516939601942082586); DK(KP196034280, +0.196034280659121203988391127777283691722273346); DK(KP1_990369453, +1.990369453344393772489673906218959843150949737); DK(KP942793473, +0.942793473651995297112775251810508755314920638); DK(KP1_763842528, +1.763842528696710059425513727320776699016885241); DK(KP580569354, +0.580569354508924735272384751634790549382952557); DK(KP1_913880671, +1.913880671464417729871595773960539938965698411); DK(KP1_111140466, +1.111140466039204449485661627897065748749874382); DK(KP1_662939224, +1.662939224605090474157576755235811513477121624); DK(KP390180644, +0.390180644032256535696569736954044481855383236); DK(KP1_961570560, +1.961570560806460898252364472268478073947867462); DK(KP923879532, +0.923879532511286756128183189396788286822416626); DK(KP382683432, +0.382683432365089771728459984030398866761344562); DK(KP707106781, +0.707106781186547524400844362104849039284835938); DK(KP765366864, +0.765366864730179543456919968060797733522689125); DK(KP1_847759065, +1.847759065022573512256366378793576573644833252); DK(KP1_414213562, +1.414213562373095048801688724209698078569671875); DK(KP2_000000000, +2.000000000000000000000000000000000000000000000); { INT i; for (i = v; i > 0; i = i - 1, R0 = R0 + ovs, R1 = R1 + ovs, Cr = Cr + ivs, Ci = Ci + ivs, MAKE_VOLATILE_STRIDE(256, rs), MAKE_VOLATILE_STRIDE(256, csr), MAKE_VOLATILE_STRIDE(256, csi)) { E Ta, T2S, T18, T2u, T3F, T4V, T5l, T61, Th, T2T, T1h, T2v, T3M, T4W, T5o; E T62, T3Q, T5q, T5u, T44, Tp, Tw, T2V, T2W, T2X, T2Y, T3X, T5t, T1r, T2x; E T41, T5r, T1A, T2y, T4a, T5y, T5N, T4H, TN, T31, T4E, T5z, T39, T3q, T1L; E T2B, T4h, T5M, T2h, T2F, T12, T36, T5D, T5J, T5G, T5K, T1U, T26, T23, T27; E T4p, T4z, T4w, T4A, T34, T3r; { E T5, T3A, T3, T3y, T9, T3C, T17, T3D, T6, T14; { E T4, T3z, T1, T2; T4 = Cr[WS(csr, 16)]; T5 = KP2_000000000 * T4; T3z = Ci[WS(csi, 16)]; T3A = KP2_000000000 * T3z; T1 = Cr[0]; T2 = Cr[WS(csr, 32)]; T3 = T1 + T2; T3y = T1 - T2; { E T7, T8, T15, T16; T7 = Cr[WS(csr, 8)]; T8 = Cr[WS(csr, 24)]; T9 = KP2_000000000 * (T7 + T8); T3C = T7 - T8; T15 = Ci[WS(csi, 8)]; T16 = Ci[WS(csi, 24)]; T17 = KP2_000000000 * (T15 - T16); T3D = T15 + T16; } } T6 = T3 + T5; Ta = T6 + T9; T2S = T6 - T9; T14 = T3 - T5; T18 = T14 - T17; T2u = T14 + T17; { E T3B, T3E, T5j, T5k; T3B = T3y - T3A; T3E = KP1_414213562 * (T3C - T3D); T3F = T3B + T3E; T4V = T3B - T3E; T5j = T3y + T3A; T5k = KP1_414213562 * (T3C + T3D); T5l = T5j - T5k; T61 = T5j + T5k; } } { E Td, T3G, T1c, T3K, Tg, T3J, T1f, T3H, T19, T1g; { E Tb, Tc, T1a, T1b; Tb = Cr[WS(csr, 4)]; Tc = Cr[WS(csr, 28)]; Td = Tb + Tc; T3G = Tb - Tc; T1a = Ci[WS(csi, 4)]; T1b = Ci[WS(csi, 28)]; T1c = T1a - T1b; T3K = T1a + T1b; } { E Te, Tf, T1d, T1e; Te = Cr[WS(csr, 20)]; Tf = Cr[WS(csr, 12)]; Tg = Te + Tf; T3J = Te - Tf; T1d = Ci[WS(csi, 20)]; T1e = Ci[WS(csi, 12)]; T1f = T1d - T1e; T3H = T1d + T1e; } Th = KP2_000000000 * (Td + Tg); T2T = KP2_000000000 * (T1f + T1c); T19 = Td - Tg; T1g = T1c - T1f; T1h = KP1_414213562 * (T19 - T1g); T2v = KP1_414213562 * (T19 + T1g); { E T3I, T3L, T5m, T5n; T3I = T3G - T3H; T3L = T3J + T3K; T3M = FNMS(KP765366864, T3L, KP1_847759065 * T3I); T4W = FMA(KP765366864, T3I, KP1_847759065 * T3L); T5m = T3G + T3H; T5n = T3K - T3J; T5o = FNMS(KP1_847759065, T5n, KP765366864 * T5m); T62 = FMA(KP1_847759065, T5m, KP765366864 * T5n); } } { E Tl, T3O, T1v, T43, To, T42, T1y, T3P, Ts, T3R, T1p, T3S, Tv, T3U, T1m; E T3V; { E Tj, Tk, T1t, T1u; Tj = Cr[WS(csr, 2)]; Tk = Cr[WS(csr, 30)]; Tl = Tj + Tk; T3O = Tj - Tk; T1t = Ci[WS(csi, 2)]; T1u = Ci[WS(csi, 30)]; T1v = T1t - T1u; T43 = T1t + T1u; } { E Tm, Tn, T1w, T1x; Tm = Cr[WS(csr, 18)]; Tn = Cr[WS(csr, 14)]; To = Tm + Tn; T42 = Tm - Tn; T1w = Ci[WS(csi, 18)]; T1x = Ci[WS(csi, 14)]; T1y = T1w - T1x; T3P = T1w + T1x; } { E Tq, Tr, T1n, T1o; Tq = Cr[WS(csr, 10)]; Tr = Cr[WS(csr, 22)]; Ts = Tq + Tr; T3R = Tq - Tr; T1n = Ci[WS(csi, 10)]; T1o = Ci[WS(csi, 22)]; T1p = T1n - T1o; T3S = T1n + T1o; } { E Tt, Tu, T1k, T1l; Tt = Cr[WS(csr, 6)]; Tu = Cr[WS(csr, 26)]; Tv = Tt + Tu; T3U = Tt - Tu; T1k = Ci[WS(csi, 26)]; T1l = Ci[WS(csi, 6)]; T1m = T1k - T1l; T3V = T1l + T1k; } T3Q = T3O - T3P; T5q = T3O + T3P; T5u = T43 - T42; T44 = T42 + T43; Tp = Tl + To; Tw = Ts + Tv; T2V = Tp - Tw; { E T3T, T3W, T1j, T1q; T2W = T1y + T1v; T2X = T1p + T1m; T2Y = T2W - T2X; T3T = T3R - T3S; T3W = T3U - T3V; T3X = KP707106781 * (T3T + T3W); T5t = KP707106781 * (T3T - T3W); T1j = Tl - To; T1q = T1m - T1p; T1r = T1j + T1q; T2x = T1j - T1q; { E T3Z, T40, T1s, T1z; T3Z = T3R + T3S; T40 = T3U + T3V; T41 = KP707106781 * (T3Z - T40); T5r = KP707106781 * (T3Z + T40); T1s = Ts - Tv; T1z = T1v - T1y; T1A = T1s + T1z; T2y = T1z - T1s; } } } { E TB, T48, T2c, T4G, TE, T4F, T2f, T49, TI, T4b, T1J, T4c, TL, T4e, T1G; E T4f; { E Tz, TA, T2a, T2b; Tz = Cr[WS(csr, 1)]; TA = Cr[WS(csr, 31)]; TB = Tz + TA; T48 = Tz - TA; T2a = Ci[WS(csi, 1)]; T2b = Ci[WS(csi, 31)]; T2c = T2a - T2b; T4G = T2a + T2b; } { E TC, TD, T2d, T2e; TC = Cr[WS(csr, 17)]; TD = Cr[WS(csr, 15)]; TE = TC + TD; T4F = TC - TD; T2d = Ci[WS(csi, 17)]; T2e = Ci[WS(csi, 15)]; T2f = T2d - T2e; T49 = T2d + T2e; } { E TG, TH, T1H, T1I; TG = Cr[WS(csr, 9)]; TH = Cr[WS(csr, 23)]; TI = TG + TH; T4b = TG - TH; T1H = Ci[WS(csi, 9)]; T1I = Ci[WS(csi, 23)]; T1J = T1H - T1I; T4c = T1H + T1I; } { E TJ, TK, T1E, T1F; TJ = Cr[WS(csr, 7)]; TK = Cr[WS(csr, 25)]; TL = TJ + TK; T4e = TJ - TK; T1E = Ci[WS(csi, 25)]; T1F = Ci[WS(csi, 7)]; T1G = T1E - T1F; T4f = T1F + T1E; } { E TF, TM, T1D, T1K; T4a = T48 - T49; T5y = T48 + T49; T5N = T4G - T4F; T4H = T4F + T4G; TF = TB + TE; TM = TI + TL; TN = TF + TM; T31 = TF - TM; { E T4C, T4D, T37, T38; T4C = T4b + T4c; T4D = T4e + T4f; T4E = KP707106781 * (T4C - T4D); T5z = KP707106781 * (T4C + T4D); T37 = T2f + T2c; T38 = T1J + T1G; T39 = T37 - T38; T3q = T38 + T37; } T1D = TB - TE; T1K = T1G - T1J; T1L = T1D + T1K; T2B = T1D - T1K; { E T4d, T4g, T29, T2g; T4d = T4b - T4c; T4g = T4e - T4f; T4h = KP707106781 * (T4d + T4g); T5M = KP707106781 * (T4d - T4g); T29 = TI - TL; T2g = T2c - T2f; T2h = T29 + T2g; T2F = T2g - T29; } } } { E TQ, T4j, T1P, T4n, TT, T4m, T1S, T4k, TX, T4q, T1Y, T4u, T10, T4t, T21; E T4r; { E TO, TP, T1N, T1O; TO = Cr[WS(csr, 5)]; TP = Cr[WS(csr, 27)]; TQ = TO + TP; T4j = TO - TP; T1N = Ci[WS(csi, 5)]; T1O = Ci[WS(csi, 27)]; T1P = T1N - T1O; T4n = T1N + T1O; } { E TR, TS, T1Q, T1R; TR = Cr[WS(csr, 21)]; TS = Cr[WS(csr, 11)]; TT = TR + TS; T4m = TR - TS; T1Q = Ci[WS(csi, 21)]; T1R = Ci[WS(csi, 11)]; T1S = T1Q - T1R; T4k = T1Q + T1R; } { E TV, TW, T1W, T1X; TV = Cr[WS(csr, 3)]; TW = Cr[WS(csr, 29)]; TX = TV + TW; T4q = TV - TW; T1W = Ci[WS(csi, 29)]; T1X = Ci[WS(csi, 3)]; T1Y = T1W - T1X; T4u = T1X + T1W; } { E TY, TZ, T1Z, T20; TY = Cr[WS(csr, 13)]; TZ = Cr[WS(csr, 19)]; T10 = TY + TZ; T4t = TY - TZ; T1Z = Ci[WS(csi, 13)]; T20 = Ci[WS(csi, 19)]; T21 = T1Z - T20; T4r = T1Z + T20; } { E TU, T11, T5B, T5C; TU = TQ + TT; T11 = TX + T10; T12 = TU + T11; T36 = TU - T11; T5B = T4j + T4k; T5C = T4n - T4m; T5D = FNMS(KP923879532, T5C, KP382683432 * T5B); T5J = FMA(KP923879532, T5B, KP382683432 * T5C); } { E T5E, T5F, T1M, T1T; T5E = T4q + T4r; T5F = T4t + T4u; T5G = FNMS(KP923879532, T5F, KP382683432 * T5E); T5K = FMA(KP923879532, T5E, KP382683432 * T5F); T1M = TQ - TT; T1T = T1P - T1S; T1U = T1M - T1T; T26 = T1M + T1T; } { E T1V, T22, T4l, T4o; T1V = TX - T10; T22 = T1Y - T21; T23 = T1V + T22; T27 = T22 - T1V; T4l = T4j - T4k; T4o = T4m + T4n; T4p = FNMS(KP382683432, T4o, KP923879532 * T4l); T4z = FMA(KP382683432, T4l, KP923879532 * T4o); } { E T4s, T4v, T32, T33; T4s = T4q - T4r; T4v = T4t - T4u; T4w = FMA(KP923879532, T4s, KP382683432 * T4v); T4A = FNMS(KP382683432, T4s, KP923879532 * T4v); T32 = T21 + T1Y; T33 = T1S + T1P; T34 = T32 - T33; T3r = T33 + T32; } } { E T13, T3x, Ty, T3w, Ti, Tx; T13 = KP2_000000000 * (TN + T12); T3x = KP2_000000000 * (T3r + T3q); Ti = Ta + Th; Tx = KP2_000000000 * (Tp + Tw); Ty = Ti + Tx; T3w = Ti - Tx; R0[WS(rs, 16)] = Ty - T13; R0[WS(rs, 24)] = T3w + T3x; R0[0] = Ty + T13; R0[WS(rs, 8)] = T3w - T3x; } { E T3g, T3k, T3j, T3l; { E T3e, T3f, T3h, T3i; T3e = T2S + T2T; T3f = KP1_414213562 * (T2V + T2Y); T3g = T3e - T3f; T3k = T3e + T3f; T3h = T31 - T34; T3i = T39 - T36; T3j = FNMS(KP1_847759065, T3i, KP765366864 * T3h); T3l = FMA(KP1_847759065, T3h, KP765366864 * T3i); } R0[WS(rs, 22)] = T3g - T3j; R0[WS(rs, 30)] = T3k + T3l; R0[WS(rs, 6)] = T3g + T3j; R0[WS(rs, 14)] = T3k - T3l; } { E T3o, T3u, T3t, T3v; { E T3m, T3n, T3p, T3s; T3m = Ta - Th; T3n = KP2_000000000 * (T2X + T2W); T3o = T3m - T3n; T3u = T3m + T3n; T3p = TN - T12; T3s = T3q - T3r; T3t = KP1_414213562 * (T3p - T3s); T3v = KP1_414213562 * (T3p + T3s); } R0[WS(rs, 20)] = T3o - T3t; R0[WS(rs, 28)] = T3u + T3v; R0[WS(rs, 4)] = T3o + T3t; R0[WS(rs, 12)] = T3u - T3v; } { E T30, T3c, T3b, T3d; { E T2U, T2Z, T35, T3a; T2U = T2S - T2T; T2Z = KP1_414213562 * (T2V - T2Y); T30 = T2U + T2Z; T3c = T2U - T2Z; T35 = T31 + T34; T3a = T36 + T39; T3b = FNMS(KP765366864, T3a, KP1_847759065 * T35); T3d = FMA(KP765366864, T35, KP1_847759065 * T3a); } R0[WS(rs, 18)] = T30 - T3b; R0[WS(rs, 26)] = T3c + T3d; R0[WS(rs, 2)] = T30 + T3b; R0[WS(rs, 10)] = T3c - T3d; } { E T25, T2p, T2i, T2q, T1C, T2k, T2o, T2s, T24, T28; T24 = KP707106781 * (T1U + T23); T25 = T1L + T24; T2p = T1L - T24; T28 = KP707106781 * (T26 + T27); T2i = T28 + T2h; T2q = T2h - T28; { E T1i, T1B, T2m, T2n; T1i = T18 + T1h; T1B = FNMS(KP765366864, T1A, KP1_847759065 * T1r); T1C = T1i + T1B; T2k = T1i - T1B; T2m = T18 - T1h; T2n = FMA(KP765366864, T1r, KP1_847759065 * T1A); T2o = T2m - T2n; T2s = T2m + T2n; } { E T2j, T2t, T2l, T2r; T2j = FNMS(KP390180644, T2i, KP1_961570560 * T25); R0[WS(rs, 17)] = T1C - T2j; R0[WS(rs, 1)] = T1C + T2j; T2t = FMA(KP1_662939224, T2p, KP1_111140466 * T2q); R0[WS(rs, 13)] = T2s - T2t; R0[WS(rs, 29)] = T2s + T2t; T2l = FMA(KP390180644, T25, KP1_961570560 * T2i); R0[WS(rs, 9)] = T2k - T2l; R0[WS(rs, 25)] = T2k + T2l; T2r = FNMS(KP1_662939224, T2q, KP1_111140466 * T2p); R0[WS(rs, 21)] = T2o - T2r; R0[WS(rs, 5)] = T2o + T2r; } } { E T2D, T2N, T2G, T2O, T2A, T2I, T2M, T2Q, T2C, T2E; T2C = KP707106781 * (T27 - T26); T2D = T2B + T2C; T2N = T2B - T2C; T2E = KP707106781 * (T1U - T23); T2G = T2E + T2F; T2O = T2F - T2E; { E T2w, T2z, T2K, T2L; T2w = T2u - T2v; T2z = FNMS(KP1_847759065, T2y, KP765366864 * T2x); T2A = T2w + T2z; T2I = T2w - T2z; T2K = T2u + T2v; T2L = FMA(KP1_847759065, T2x, KP765366864 * T2y); T2M = T2K - T2L; T2Q = T2K + T2L; } { E T2H, T2R, T2J, T2P; T2H = FNMS(KP1_111140466, T2G, KP1_662939224 * T2D); R0[WS(rs, 19)] = T2A - T2H; R0[WS(rs, 3)] = T2A + T2H; T2R = FMA(KP1_961570560, T2N, KP390180644 * T2O); R0[WS(rs, 15)] = T2Q - T2R; R0[WS(rs, 31)] = T2Q + T2R; T2J = FMA(KP1_111140466, T2D, KP1_662939224 * T2G); R0[WS(rs, 11)] = T2I - T2J; R0[WS(rs, 27)] = T2I + T2J; T2P = FNMS(KP1_961570560, T2O, KP390180644 * T2N); R0[WS(rs, 23)] = T2M - T2P; R0[WS(rs, 7)] = T2M + T2P; } } { E T5p, T5T, T5w, T5U, T5I, T5W, T5P, T5X, T5s, T5v; T5p = T5l + T5o; T5T = T5l - T5o; T5s = T5q - T5r; T5v = T5t + T5u; T5w = FNMS(KP1_111140466, T5v, KP1_662939224 * T5s); T5U = FMA(KP1_111140466, T5s, KP1_662939224 * T5v); { E T5A, T5H, T5L, T5O; T5A = T5y - T5z; T5H = T5D + T5G; T5I = T5A + T5H; T5W = T5A - T5H; T5L = T5J - T5K; T5O = T5M + T5N; T5P = T5L + T5O; T5X = T5O - T5L; } { E T5x, T5Q, T5Z, T60; T5x = T5p + T5w; T5Q = FNMS(KP580569354, T5P, KP1_913880671 * T5I); R1[WS(rs, 17)] = T5x - T5Q; R1[WS(rs, 1)] = T5x + T5Q; T5Z = T5T + T5U; T60 = FMA(KP1_763842528, T5W, KP942793473 * T5X); R1[WS(rs, 13)] = T5Z - T60; R1[WS(rs, 29)] = T5Z + T60; } { E T5R, T5S, T5V, T5Y; T5R = T5p - T5w; T5S = FMA(KP580569354, T5I, KP1_913880671 * T5P); R1[WS(rs, 9)] = T5R - T5S; R1[WS(rs, 25)] = T5R + T5S; T5V = T5T - T5U; T5Y = FNMS(KP1_763842528, T5X, KP942793473 * T5W); R1[WS(rs, 21)] = T5V - T5Y; R1[WS(rs, 5)] = T5V + T5Y; } } { E T3N, T4N, T46, T4O, T4y, T4Q, T4J, T4R, T3Y, T45; T3N = T3F + T3M; T4N = T3F - T3M; T3Y = T3Q + T3X; T45 = T41 + T44; T46 = FNMS(KP390180644, T45, KP1_961570560 * T3Y); T4O = FMA(KP390180644, T3Y, KP1_961570560 * T45); { E T4i, T4x, T4B, T4I; T4i = T4a + T4h; T4x = T4p + T4w; T4y = T4i + T4x; T4Q = T4i - T4x; T4B = T4z + T4A; T4I = T4E + T4H; T4J = T4B + T4I; T4R = T4I - T4B; } { E T47, T4K, T4T, T4U; T47 = T3N + T46; T4K = FNMS(KP196034280, T4J, KP1_990369453 * T4y); R1[WS(rs, 16)] = T47 - T4K; R1[0] = T47 + T4K; T4T = T4N + T4O; T4U = FMA(KP1_546020906, T4Q, KP1_268786568 * T4R); R1[WS(rs, 12)] = T4T - T4U; R1[WS(rs, 28)] = T4T + T4U; } { E T4L, T4M, T4P, T4S; T4L = T3N - T46; T4M = FMA(KP196034280, T4y, KP1_990369453 * T4J); R1[WS(rs, 8)] = T4L - T4M; R1[WS(rs, 24)] = T4L + T4M; T4P = T4N - T4O; T4S = FNMS(KP1_546020906, T4R, KP1_268786568 * T4Q); R1[WS(rs, 20)] = T4P - T4S; R1[WS(rs, 4)] = T4P + T4S; } } { E T63, T6h, T66, T6i, T6a, T6k, T6d, T6l, T64, T65; T63 = T61 - T62; T6h = T61 + T62; T64 = T5q + T5r; T65 = T5u - T5t; T66 = FNMS(KP1_961570560, T65, KP390180644 * T64); T6i = FMA(KP1_961570560, T64, KP390180644 * T65); { E T68, T69, T6b, T6c; T68 = T5y + T5z; T69 = T5J + T5K; T6a = T68 - T69; T6k = T68 + T69; T6b = T5D - T5G; T6c = T5N - T5M; T6d = T6b + T6c; T6l = T6c - T6b; } { E T67, T6e, T6n, T6o; T67 = T63 + T66; T6e = FNMS(KP1_268786568, T6d, KP1_546020906 * T6a); R1[WS(rs, 19)] = T67 - T6e; R1[WS(rs, 3)] = T67 + T6e; T6n = T6h + T6i; T6o = FMA(KP1_990369453, T6k, KP196034280 * T6l); R1[WS(rs, 15)] = T6n - T6o; R1[WS(rs, 31)] = T6n + T6o; } { E T6f, T6g, T6j, T6m; T6f = T63 - T66; T6g = FMA(KP1_268786568, T6a, KP1_546020906 * T6d); R1[WS(rs, 11)] = T6f - T6g; R1[WS(rs, 27)] = T6f + T6g; T6j = T6h - T6i; T6m = FNMS(KP1_990369453, T6l, KP196034280 * T6k); R1[WS(rs, 23)] = T6j - T6m; R1[WS(rs, 7)] = T6j + T6m; } } { E T4X, T5b, T50, T5c, T54, T5e, T57, T5f, T4Y, T4Z; T4X = T4V - T4W; T5b = T4V + T4W; T4Y = T3Q - T3X; T4Z = T44 - T41; T50 = FNMS(KP1_662939224, T4Z, KP1_111140466 * T4Y); T5c = FMA(KP1_662939224, T4Y, KP1_111140466 * T4Z); { E T52, T53, T55, T56; T52 = T4a - T4h; T53 = T4A - T4z; T54 = T52 + T53; T5e = T52 - T53; T55 = T4p - T4w; T56 = T4H - T4E; T57 = T55 + T56; T5f = T56 - T55; } { E T51, T58, T5h, T5i; T51 = T4X + T50; T58 = FNMS(KP942793473, T57, KP1_763842528 * T54); R1[WS(rs, 18)] = T51 - T58; R1[WS(rs, 2)] = T51 + T58; T5h = T5b + T5c; T5i = FMA(KP1_913880671, T5e, KP580569354 * T5f); R1[WS(rs, 14)] = T5h - T5i; R1[WS(rs, 30)] = T5h + T5i; } { E T59, T5a, T5d, T5g; T59 = T4X - T50; T5a = FMA(KP942793473, T54, KP1_763842528 * T57); R1[WS(rs, 10)] = T59 - T5a; R1[WS(rs, 26)] = T59 + T5a; T5d = T5b - T5c; T5g = FNMS(KP1_913880671, T5f, KP580569354 * T5e); R1[WS(rs, 22)] = T5d - T5g; R1[WS(rs, 6)] = T5d + T5g; } } } } } static const kr2c_desc desc = { 64, "r2cb_64", {342, 82, 52, 0}, &GENUS }; void X(codelet_r2cb_64) (planner *p) { X(kr2c_register) (p, r2cb_64, &desc); } #endif /* HAVE_FMA */ fftw-3.3.3/rdft/scalar/r2cb/r2cb_128.c0000644000175000017500000027670312054411060014041 00000000000000/* * Copyright (c) 2003, 2007-11 Matteo Frigo * Copyright (c) 2003, 2007-11 Massachusetts Institute of Technology * * This program is free software; you can redistribute it and/or modify * it under the terms of the GNU General Public License as published by * the Free Software Foundation; either version 2 of the License, or * (at your option) any later version. * * This program is distributed in the hope that it will be useful, * but WITHOUT ANY WARRANTY; without even the implied warranty of * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the * GNU General Public License for more details. * * You should have received a copy of the GNU General Public License * along with this program; if not, write to the Free Software * Foundation, Inc., 51 Franklin Street, Fifth Floor, Boston, MA 02110-1301 USA * */ /* This file was automatically generated --- DO NOT EDIT */ /* Generated on Sun Nov 25 07:41:09 EST 2012 */ #include "codelet-rdft.h" #ifdef HAVE_FMA /* Generated by: ../../../genfft/gen_r2cb.native -fma -reorder-insns -schedule-for-pipeline -compact -variables 4 -pipeline-latency 4 -sign 1 -n 128 -name r2cb_128 -include r2cb.h */ /* * This function contains 956 FP additions, 540 FP multiplications, * (or, 416 additions, 0 multiplications, 540 fused multiply/add), * 242 stack variables, 36 constants, and 256 memory accesses */ #include "r2cb.h" static void r2cb_128(R *R0, R *R1, R *Cr, R *Ci, stride rs, stride csr, stride csi, INT v, INT ivs, INT ovs) { DK(KP1_715457220, +1.715457220000544139804539968569540274084981599); DK(KP1_606415062, +1.606415062961289819613353025926283847759138854); DK(KP599376933, +0.599376933681923766271389869014404232837890546); DK(KP741650546, +0.741650546272035369581266691172079863842265220); DK(KP1_978353019, +1.978353019929561946903347476032486127967379067); DK(KP1_940062506, +1.940062506389087985207968414572200502913731924); DK(KP148335987, +0.148335987538347428753676511486911367000625355); DK(KP250486960, +0.250486960191305461595702160124721208578685568); DK(KP831469612, +0.831469612302545237078788377617905756738560812); DK(KP1_807978586, +1.807978586246886663172400594461074097420264050); DK(KP1_481902250, +1.481902250709918182351233794990325459457910619); DK(KP472964775, +0.472964775891319928124438237972992463904131113); DK(KP906347169, +0.906347169019147157946142717268914412664134293); DK(KP1_997590912, +1.997590912410344785429543209518201388886407229); DK(KP1_883088130, +1.883088130366041556825018805199004714371179592); DK(KP049126849, +0.049126849769467254105343321271313617079695752); DK(KP357805721, +0.357805721314524104672487743774474392487532769); DK(KP980785280, +0.980785280403230449126182236134239036973933731); DK(KP1_763842528, +1.763842528696710059425513727320776699016885241); DK(KP1_913880671, +1.913880671464417729871595773960539938965698411); DK(KP534511135, +0.534511135950791641089685961295362908582039528); DK(KP303346683, +0.303346683607342391675883946941299872384187453); DK(KP1_662939224, +1.662939224605090474157576755235811513477121624); DK(KP668178637, +0.668178637919298919997757686523080761552472251); DK(KP1_990369453, +1.990369453344393772489673906218959843150949737); DK(KP1_546020906, +1.546020906725473921621813219516939601942082586); DK(KP098491403, +0.098491403357164253077197521291327432293052451); DK(KP820678790, +0.820678790828660330972281985331011598767386482); DK(KP1_961570560, +1.961570560806460898252364472268478073947867462); DK(KP198912367, +0.198912367379658006911597622644676228597850501); DK(KP923879532, +0.923879532511286756128183189396788286822416626); DK(KP1_847759065, +1.847759065022573512256366378793576573644833252); DK(KP707106781, +0.707106781186547524400844362104849039284835938); DK(KP414213562, +0.414213562373095048801688724209698078569671875); DK(KP1_414213562, +1.414213562373095048801688724209698078569671875); DK(KP2_000000000, +2.000000000000000000000000000000000000000000000); { INT i; for (i = v; i > 0; i = i - 1, R0 = R0 + ovs, R1 = R1 + ovs, Cr = Cr + ivs, Ci = Ci + ivs, MAKE_VOLATILE_STRIDE(512, rs), MAKE_VOLATILE_STRIDE(512, csr), MAKE_VOLATILE_STRIDE(512, csi)) { E T9H, T9I, T9X, T9Y; { E Tdr, T9, Tcl, Ta9, T6b, T2d, T91, T7j, Tg, Tds, Tcm, Tae, T92, T7m, T6c; E T2o, Tdu, Tw, Tco, Tap, TeM, Tdx, T6f, T2G, T6e, T2P, T94, T7t, Tcp, Tak; E T95, T7q, TdM, T1i, TcL, TbD, Tf0, Te6, T6q, T42, T6B, T5t, T9r, T8j, TcA; E TaY, T9g, T7S, TdA, TM, Tcv, TaN, TeP, TdI, T6i, T38, T6l, T3F, T9b, T7J; E Tcs, Taw, T98, T7y, T1N, TeW, T6x, T4H, Te8, TdV, T6w, T4Q, T9j, T86, TcO; E TcI, T9k, T83, TbI, Tbl, T22, TeV, Te0, Te9, T58, T6u, T6t, T5h, T9m, T8d; E TcP, TcF, T9n, T8a, TbJ, Tbw, Te3, T1x, TcB, TbG, Tf1, TdP, T6C, T4p, T6r; E T5w, T9h, T8m, TcM, Tb9, T9s, T7Z, TaB, TaG, TdF, T11, Tct, TaQ, TeQ, TdD; E T6m, T3v, T7B, T7E, T6j, T3I, T99, T7M; { E TaU, TaX, T7Q, T7R, Tbk, Tbf; { E Td, T2e, Tc, Tab, T2m, Te, T2f, T2g; { E T7h, T27, T2c, T7i; { E T4, T26, T29, T25, T3, T28, T8, T2a; T4 = Cr[WS(csr, 32)]; T26 = Ci[WS(csi, 32)]; { E T1, T2, T6, T7; T1 = Cr[0]; T2 = Cr[WS(csr, 64)]; T6 = Cr[WS(csr, 16)]; T7 = Cr[WS(csr, 48)]; T29 = Ci[WS(csi, 16)]; T25 = T1 - T2; T3 = T1 + T2; T28 = T6 - T7; T8 = T6 + T7; T2a = Ci[WS(csi, 48)]; } { E Ta7, T5, Ta8, T2b; Ta7 = FNMS(KP2_000000000, T4, T3); T5 = FMA(KP2_000000000, T4, T3); T7h = FMA(KP2_000000000, T26, T25); T27 = FNMS(KP2_000000000, T26, T25); Ta8 = T29 - T2a; T2b = T29 + T2a; Tdr = FNMS(KP2_000000000, T8, T5); T9 = FMA(KP2_000000000, T8, T5); Tcl = FMA(KP2_000000000, Ta8, Ta7); Ta9 = FNMS(KP2_000000000, Ta8, Ta7); T2c = T28 - T2b; T7i = T28 + T2b; } } { E Ta, Tb, T2k, T2l; Ta = Cr[WS(csr, 8)]; T6b = FNMS(KP1_414213562, T2c, T27); T2d = FMA(KP1_414213562, T2c, T27); T91 = FMA(KP1_414213562, T7i, T7h); T7j = FNMS(KP1_414213562, T7i, T7h); Tb = Cr[WS(csr, 56)]; T2k = Ci[WS(csi, 8)]; T2l = Ci[WS(csi, 56)]; Td = Cr[WS(csr, 40)]; T2e = Ta - Tb; Tc = Ta + Tb; Tab = T2k - T2l; T2m = T2k + T2l; Te = Cr[WS(csr, 24)]; T2f = Ci[WS(csi, 40)]; T2g = Ci[WS(csi, 24)]; } } { E Tag, Taj, T7o, T7p; { E T2q, Tk, Tam, T2K, T2H, Tn, Tan, T2t, Tu, Tah, T2E, T2N, Tr, T2v, T2y; E Tai; { E Tl, Tm, T2r, T2s; { E Ti, Tj, T2j, Tf, T2I, T2J; Ti = Cr[WS(csr, 4)]; T2j = Td - Te; Tf = Td + Te; { E Tac, T2h, T7k, T2n; Tac = T2f - T2g; T2h = T2f + T2g; T7k = T2m - T2j; T2n = T2j + T2m; { E Taa, Tad, T7l, T2i; Taa = Tc - Tf; Tg = Tc + Tf; Tad = Tab - Tac; Tds = Tac + Tab; T7l = T2e + T2h; T2i = T2e - T2h; Tcm = Taa + Tad; Tae = Taa - Tad; T92 = FMA(KP414213562, T7k, T7l); T7m = FNMS(KP414213562, T7l, T7k); T6c = FMA(KP414213562, T2i, T2n); T2o = FNMS(KP414213562, T2n, T2i); Tj = Cr[WS(csr, 60)]; } } T2I = Ci[WS(csi, 4)]; T2J = Ci[WS(csi, 60)]; Tl = Cr[WS(csr, 36)]; T2q = Ti - Tj; Tk = Ti + Tj; Tam = T2I - T2J; T2K = T2I + T2J; Tm = Cr[WS(csr, 28)]; } T2r = Ci[WS(csi, 36)]; T2s = Ci[WS(csi, 28)]; { E Ts, Tt, T2B, T2C; Ts = Cr[WS(csr, 12)]; T2H = Tl - Tm; Tn = Tl + Tm; Tan = T2r - T2s; T2t = T2r + T2s; Tt = Cr[WS(csr, 52)]; T2B = Ci[WS(csi, 12)]; T2C = Ci[WS(csi, 52)]; { E Tp, T2A, T2D, Tq, T2w, T2x; Tp = Cr[WS(csr, 20)]; Tu = Ts + Tt; T2A = Ts - Tt; Tah = T2C - T2B; T2D = T2B + T2C; Tq = Cr[WS(csr, 44)]; T2w = Ci[WS(csi, 20)]; T2x = Ci[WS(csi, 44)]; T2E = T2A - T2D; T2N = T2A + T2D; Tr = Tp + Tq; T2v = Tp - Tq; T2y = T2w + T2x; Tai = T2w - T2x; } } } { E T2M, Tdv, Tdw, T2u, T2F, T7s, T7r, T2L, T2O; { E To, T2z, Tv, Tal, Tao; Tag = Tk - Tn; To = Tk + Tn; T2M = T2v + T2y; T2z = T2v - T2y; Tv = Tr + Tu; Tal = Tr - Tu; Tao = Tam - Tan; Tdv = Tan + Tam; Tdu = To - Tv; Tw = To + Tv; Tco = Tao - Tal; Tap = Tal + Tao; Tdw = Tai + Tah; Taj = Tah - Tai; T7o = T2q + T2t; T2u = T2q - T2t; T2F = T2z + T2E; T7s = T2E - T2z; } T7r = T2K - T2H; T2L = T2H + T2K; TeM = Tdw + Tdv; Tdx = Tdv - Tdw; T6f = FNMS(KP707106781, T2F, T2u); T2G = FMA(KP707106781, T2F, T2u); T2O = T2M - T2N; T7p = T2M + T2N; T6e = FNMS(KP707106781, T2O, T2L); T2P = FMA(KP707106781, T2O, T2L); T94 = FMA(KP707106781, T7s, T7r); T7t = FNMS(KP707106781, T7s, T7r); } } { E T3M, T16, TbA, T5o, T5l, T19, TbB, T3P, T1g, TaV, T40, T5r, T1d, T3R, T3U; E TaW; { E T17, T18, T3N, T3O; { E T14, T15, T5m, T5n; T14 = Cr[WS(csr, 1)]; Tcp = Tag - Taj; Tak = Tag + Taj; T95 = FMA(KP707106781, T7p, T7o); T7q = FNMS(KP707106781, T7p, T7o); T15 = Cr[WS(csr, 63)]; T5m = Ci[WS(csi, 1)]; T5n = Ci[WS(csi, 63)]; T17 = Cr[WS(csr, 33)]; T3M = T14 - T15; T16 = T14 + T15; TbA = T5m - T5n; T5o = T5m + T5n; T18 = Cr[WS(csr, 31)]; } T3N = Ci[WS(csi, 33)]; T3O = Ci[WS(csi, 31)]; { E T1e, T1f, T3X, T3Y; T1e = Cr[WS(csr, 15)]; T5l = T17 - T18; T19 = T17 + T18; TbB = T3N - T3O; T3P = T3N + T3O; T1f = Cr[WS(csr, 49)]; T3X = Ci[WS(csi, 15)]; T3Y = Ci[WS(csi, 49)]; { E T1b, T3W, T3Z, T1c, T3S, T3T; T1b = Cr[WS(csr, 17)]; T1g = T1e + T1f; T3W = T1e - T1f; TaV = T3Y - T3X; T3Z = T3X + T3Y; T1c = Cr[WS(csr, 47)]; T3S = Ci[WS(csi, 17)]; T3T = Ci[WS(csi, 47)]; T40 = T3W - T3Z; T5r = T3W + T3Z; T1d = T1b + T1c; T3R = T1b - T1c; T3U = T3S + T3T; TaW = T3S - T3T; } } } { E T5q, Te4, Te5, T3Q, T41, T8i, T8h, T5p, T5s; { E T1a, T3V, T1h, Tbz, TbC; TaU = T16 - T19; T1a = T16 + T19; T5q = T3R + T3U; T3V = T3R - T3U; T1h = T1d + T1g; Tbz = T1d - T1g; TbC = TbA - TbB; Te4 = TbB + TbA; TdM = T1a - T1h; T1i = T1a + T1h; TcL = TbC - Tbz; TbD = Tbz + TbC; Te5 = TaW + TaV; TaX = TaV - TaW; T7Q = T3M + T3P; T3Q = T3M - T3P; T41 = T3V + T40; T8i = T40 - T3V; } T8h = T5o - T5l; T5p = T5l + T5o; Tf0 = Te5 + Te4; Te6 = Te4 - Te5; T6q = FNMS(KP707106781, T41, T3Q); T42 = FMA(KP707106781, T41, T3Q); T5s = T5q - T5r; T7R = T5q + T5r; T6B = FNMS(KP707106781, T5s, T5p); T5t = FMA(KP707106781, T5s, T5p); T9r = FMA(KP707106781, T8i, T8h); T8j = FNMS(KP707106781, T8i, T8h); } } } } { E Tas, Tav, T7w, T7x; { E T2S, TA, TaK, T3A, T3x, TD, TaL, T2V, TK, Tat, T36, T3D, TH, T2X, T30; E Tau; { E TB, TC, T2T, T2U; { E Ty, Tz, T3y, T3z; Ty = Cr[WS(csr, 2)]; TcA = TaU - TaX; TaY = TaU + TaX; T9g = FMA(KP707106781, T7R, T7Q); T7S = FNMS(KP707106781, T7R, T7Q); Tz = Cr[WS(csr, 62)]; T3y = Ci[WS(csi, 2)]; T3z = Ci[WS(csi, 62)]; TB = Cr[WS(csr, 34)]; T2S = Ty - Tz; TA = Ty + Tz; TaK = T3y - T3z; T3A = T3y + T3z; TC = Cr[WS(csr, 30)]; } T2T = Ci[WS(csi, 34)]; T2U = Ci[WS(csi, 30)]; { E TI, TJ, T33, T34; TI = Cr[WS(csr, 14)]; T3x = TB - TC; TD = TB + TC; TaL = T2T - T2U; T2V = T2T + T2U; TJ = Cr[WS(csr, 50)]; T33 = Ci[WS(csi, 14)]; T34 = Ci[WS(csi, 50)]; { E TF, T32, T35, TG, T2Y, T2Z; TF = Cr[WS(csr, 18)]; TK = TI + TJ; T32 = TI - TJ; Tat = T34 - T33; T35 = T33 + T34; TG = Cr[WS(csr, 46)]; T2Y = Ci[WS(csi, 18)]; T2Z = Ci[WS(csi, 46)]; T36 = T32 - T35; T3D = T32 + T35; TH = TF + TG; T2X = TF - TG; T30 = T2Y + T2Z; Tau = T2Y - T2Z; } } } { E T3C, TdG, TdH, T2W, T37, T7I, T7H, T3B, T3E; { E TE, T31, TL, TaJ, TaM; Tas = TA - TD; TE = TA + TD; T3C = T2X + T30; T31 = T2X - T30; TL = TH + TK; TaJ = TH - TK; TaM = TaK - TaL; TdG = TaL + TaK; TdA = TE - TL; TM = TE + TL; Tcv = TaM - TaJ; TaN = TaJ + TaM; TdH = Tau + Tat; Tav = Tat - Tau; T7w = T2S + T2V; T2W = T2S - T2V; T37 = T31 + T36; T7I = T36 - T31; } T7H = T3A - T3x; T3B = T3x + T3A; TeP = TdH + TdG; TdI = TdG - TdH; T6i = FNMS(KP707106781, T37, T2W); T38 = FMA(KP707106781, T37, T2W); T3E = T3C - T3D; T7x = T3C + T3D; T6l = FNMS(KP707106781, T3E, T3B); T3F = FMA(KP707106781, T3E, T3B); T9b = FMA(KP707106781, T7I, T7H); T7J = FNMS(KP707106781, T7I, T7H); } } { E T4r, T4I, T1F, Tbb, T4u, T4L, Tbj, TdS, T1I, Tbd, T4N, T4A, T4B, T1L, Tbc; E T4E, T1M, Tbg; { E T1z, T1A, T1C, T1D, Tbi, Tbh; T1z = Cr[WS(csr, 5)]; Tcs = Tas - Tav; Taw = Tas + Tav; T98 = FMA(KP707106781, T7x, T7w); T7y = FNMS(KP707106781, T7x, T7w); T1A = Cr[WS(csr, 59)]; T1C = Cr[WS(csr, 37)]; T1D = Cr[WS(csr, 27)]; { E T4s, T1B, T1E, T4t, T4J, T4K; T4s = Ci[WS(csi, 37)]; T4r = T1z - T1A; T1B = T1z + T1A; T4I = T1C - T1D; T1E = T1C + T1D; T4t = Ci[WS(csi, 27)]; T4J = Ci[WS(csi, 5)]; T4K = Ci[WS(csi, 59)]; T1F = T1B + T1E; Tbb = T1B - T1E; T4u = T4s + T4t; Tbi = T4s - T4t; Tbh = T4J - T4K; T4L = T4J + T4K; } { E T1J, T4w, T4z, T1K, T4C, T4D; { E T1G, T1H, T4x, T4y; T1G = Cr[WS(csr, 21)]; Tbj = Tbh - Tbi; TdS = Tbi + Tbh; T1H = Cr[WS(csr, 43)]; T4x = Ci[WS(csi, 21)]; T4y = Ci[WS(csi, 43)]; T1J = Cr[WS(csr, 11)]; T4w = T1G - T1H; T1I = T1G + T1H; Tbd = T4x - T4y; T4z = T4x + T4y; T1K = Cr[WS(csr, 53)]; T4C = Ci[WS(csi, 11)]; T4D = Ci[WS(csi, 53)]; } T4N = T4w + T4z; T4A = T4w - T4z; T4B = T1J - T1K; T1L = T1J + T1K; Tbc = T4D - T4C; T4E = T4C + T4D; } } T1M = T1I + T1L; Tbg = T1I - T1L; { E TdT, Tbe, T4F, T4O; TdT = Tbd + Tbc; Tbe = Tbc - Tbd; T4F = T4B - T4E; T4O = T4B + T4E; { E TdR, TdU, T81, T4v, T4G, T85; TdR = T1F - T1M; T1N = T1F + T1M; TeW = TdT + TdS; TdU = TdS - TdT; T81 = T4r + T4u; T4v = T4r - T4u; T4G = T4A + T4F; T85 = T4F - T4A; { E T84, T4M, T4P, T82, TcG, TcH; T84 = T4L - T4I; T4M = T4I + T4L; T6x = FNMS(KP707106781, T4G, T4v); T4H = FMA(KP707106781, T4G, T4v); Te8 = TdR + TdU; TdV = TdR - TdU; T4P = T4N - T4O; T82 = T4N + T4O; Tbk = Tbg + Tbj; TcG = Tbj - Tbg; T6w = FNMS(KP707106781, T4P, T4M); T4Q = FMA(KP707106781, T4P, T4M); T9j = FMA(KP707106781, T85, T84); T86 = FNMS(KP707106781, T85, T84); TcH = Tbb - Tbe; Tbf = Tbb + Tbe; TcO = FMA(KP414213562, TcG, TcH); TcI = FNMS(KP414213562, TcH, TcG); T9k = FMA(KP707106781, T82, T81); T83 = FNMS(KP707106781, T82, T81); } } } } } { E T88, T89, Tbv, Tbq; { E T4S, T59, T4V, Tbm, T1U, T5c, TdX, Tbu, T1X, T53, Tbo, T52, T20, T54, T5e; E T51; { E T1R, T1Q, T1S, T1O, T1P; T1O = Cr[WS(csr, 3)]; T1P = Cr[WS(csr, 61)]; T1R = Cr[WS(csr, 29)]; TbI = FMA(KP414213562, Tbf, Tbk); Tbl = FNMS(KP414213562, Tbk, Tbf); T1Q = T1O + T1P; T4S = T1O - T1P; T1S = Cr[WS(csr, 35)]; { E Tbt, Tbs, T4X, T50; { E T5a, T5b, T4T, T4U, T1T; T4T = Ci[WS(csi, 29)]; T4U = Ci[WS(csi, 35)]; T1T = T1R + T1S; T59 = T1R - T1S; T5a = Ci[WS(csi, 3)]; Tbt = T4T - T4U; T4V = T4T + T4U; T5b = Ci[WS(csi, 61)]; Tbm = T1Q - T1T; T1U = T1Q + T1T; T5c = T5a + T5b; Tbs = T5b - T5a; } { E T4Y, T4Z, T1V, T1W, T1Y, T1Z; T1V = Cr[WS(csr, 13)]; T1W = Cr[WS(csr, 51)]; TdX = Tbt + Tbs; Tbu = Tbs - Tbt; T4Y = Ci[WS(csi, 13)]; T4X = T1V - T1W; T1X = T1V + T1W; T4Z = Ci[WS(csi, 51)]; T1Y = Cr[WS(csr, 19)]; T1Z = Cr[WS(csr, 45)]; T53 = Ci[WS(csi, 19)]; Tbo = T4Y - T4Z; T50 = T4Y + T4Z; T52 = T1Y - T1Z; T20 = T1Y + T1Z; T54 = Ci[WS(csi, 45)]; } T5e = T4X + T50; T51 = T4X - T50; } } { E T21, Tbr, T55, Tbn; T21 = T1X + T20; Tbr = T1X - T20; T55 = T53 + T54; Tbn = T54 - T53; { E T4W, TdW, Tbp, T5f, TdZ, T57, T8c, TdY, T56; T88 = T4S + T4V; T4W = T4S - T4V; T22 = T1U + T21; TdW = T1U - T21; TdY = Tbo + Tbn; Tbp = Tbn - Tbo; T56 = T52 - T55; T5f = T52 + T55; TeV = TdY + TdX; TdZ = TdX - TdY; T57 = T51 + T56; T8c = T56 - T51; { E T8b, T5d, T5g, TcD, TcE; T8b = T59 + T5c; T5d = T59 - T5c; T5g = T5e - T5f; T89 = T5e + T5f; Te0 = TdW + TdZ; Te9 = TdZ - TdW; T58 = FMA(KP707106781, T57, T4W); T6u = FNMS(KP707106781, T57, T4W); T6t = FNMS(KP707106781, T5g, T5d); T5h = FMA(KP707106781, T5g, T5d); Tbv = Tbr + Tbu; TcD = Tbu - Tbr; TcE = Tbm - Tbp; Tbq = Tbm + Tbp; T9m = FNMS(KP707106781, T8c, T8b); T8d = FMA(KP707106781, T8c, T8b); TcP = FNMS(KP414213562, TcD, TcE); TcF = FMA(KP414213562, TcE, TcD); } } } } { E Tb3, Tb8, T7V, T7Y; { E T7T, T4c, TaZ, T1p, TdO, Tb2, T7U, T47, T1t, T4e, T1s, Tb5, T4m, T1u, T4f; E T4g; { E T1m, T43, T1l, Tb0, T4b, T1n, T44, T45; { E T1j, T1k, T49, T4a; T1j = Cr[WS(csr, 9)]; T9n = FMA(KP707106781, T89, T88); T8a = FNMS(KP707106781, T89, T88); TbJ = FNMS(KP414213562, Tbq, Tbv); Tbw = FMA(KP414213562, Tbv, Tbq); T1k = Cr[WS(csr, 55)]; T49 = Ci[WS(csi, 9)]; T4a = Ci[WS(csi, 55)]; T1m = Cr[WS(csr, 41)]; T43 = T1j - T1k; T1l = T1j + T1k; Tb0 = T49 - T4a; T4b = T49 + T4a; T1n = Cr[WS(csr, 23)]; T44 = Ci[WS(csi, 41)]; T45 = Ci[WS(csi, 23)]; } { E T1q, T1r, T4k, T4l; T1q = Cr[WS(csr, 7)]; { E T48, T1o, Tb1, T46; T48 = T1m - T1n; T1o = T1m + T1n; Tb1 = T44 - T45; T46 = T44 + T45; T7T = T4b - T48; T4c = T48 + T4b; TaZ = T1l - T1o; T1p = T1l + T1o; TdO = Tb1 + Tb0; Tb2 = Tb0 - Tb1; T7U = T43 + T46; T47 = T43 - T46; T1r = Cr[WS(csr, 57)]; } T4k = Ci[WS(csi, 7)]; T4l = Ci[WS(csi, 57)]; T1t = Cr[WS(csr, 25)]; T4e = T1q - T1r; T1s = T1q + T1r; Tb5 = T4l - T4k; T4m = T4k + T4l; T1u = Cr[WS(csr, 39)]; T4f = Ci[WS(csi, 25)]; T4g = Ci[WS(csi, 39)]; } } { E T7W, TdN, T7X, T5u, T4d, T4o, T5v, T8k, T8l; { E T4n, T1w, T4i, TbE, TbF, Tb4, Tb7; { E T4j, T1v, Tb6, T4h; T4j = T1t - T1u; T1v = T1t + T1u; Tb6 = T4f - T4g; T4h = T4f + T4g; T7W = T4j + T4m; T4n = T4j - T4m; Tb4 = T1s - T1v; T1w = T1s + T1v; TdN = Tb6 + Tb5; Tb7 = Tb5 - Tb6; T7X = T4e + T4h; T4i = T4e - T4h; } Tb3 = TaZ - Tb2; TbE = TaZ + Tb2; TbF = Tb7 - Tb4; Tb8 = Tb4 + Tb7; Te3 = T1p - T1w; T1x = T1p + T1w; TcB = TbE - TbF; TbG = TbE + TbF; T5u = FMA(KP414213562, T47, T4c); T4d = FNMS(KP414213562, T4c, T47); T4o = FMA(KP414213562, T4n, T4i); T5v = FNMS(KP414213562, T4i, T4n); } Tf1 = TdO + TdN; TdP = TdN - TdO; T6C = T4o - T4d; T4p = T4d + T4o; T7V = FNMS(KP414213562, T7U, T7T); T8k = FMA(KP414213562, T7T, T7U); T8l = FMA(KP414213562, T7W, T7X); T7Y = FNMS(KP414213562, T7X, T7W); T6r = T5u - T5v; T5w = T5u + T5v; T9h = T8k + T8l; T8m = T8k - T8l; } } { E T7z, T3i, Tax, TT, TdC, TaA, T7A, T3d, TX, T3k, TW, TaD, T3s, TY, T3l; E T3m; { E TQ, T39, TP, Tay, T3h, TR, T3a, T3b; { E TN, TO, T3f, T3g; TN = Cr[WS(csr, 10)]; TcM = Tb8 - Tb3; Tb9 = Tb3 + Tb8; T9s = T7V - T7Y; T7Z = T7V + T7Y; TO = Cr[WS(csr, 54)]; T3f = Ci[WS(csi, 10)]; T3g = Ci[WS(csi, 54)]; TQ = Cr[WS(csr, 42)]; T39 = TN - TO; TP = TN + TO; Tay = T3f - T3g; T3h = T3f + T3g; TR = Cr[WS(csr, 22)]; T3a = Ci[WS(csi, 42)]; T3b = Ci[WS(csi, 22)]; } { E TU, TV, T3q, T3r; TU = Cr[WS(csr, 6)]; { E T3e, TS, Taz, T3c; T3e = TQ - TR; TS = TQ + TR; Taz = T3a - T3b; T3c = T3a + T3b; T7z = T3h - T3e; T3i = T3e + T3h; Tax = TP - TS; TT = TP + TS; TdC = Taz + Tay; TaA = Tay - Taz; T7A = T39 + T3c; T3d = T39 - T3c; TV = Cr[WS(csr, 58)]; } T3q = Ci[WS(csi, 6)]; T3r = Ci[WS(csi, 58)]; TX = Cr[WS(csr, 26)]; T3k = TU - TV; TW = TU + TV; TaD = T3r - T3q; T3s = T3q + T3r; TY = Cr[WS(csr, 38)]; T3l = Ci[WS(csi, 26)]; T3m = Ci[WS(csi, 38)]; } } { E T7C, TdB, T7D, T3G, T3j, T3u, T3H, T7K, T7L; { E T3t, T10, T3o, TaO, TaP, TaC, TaF; { E T3p, TZ, TaE, T3n; T3p = TX - TY; TZ = TX + TY; TaE = T3l - T3m; T3n = T3l + T3m; T7C = T3p + T3s; T3t = T3p - T3s; TaC = TW - TZ; T10 = TW + TZ; TdB = TaE + TaD; TaF = TaD - TaE; T7D = T3k + T3n; T3o = T3k - T3n; } TaB = Tax - TaA; TaO = Tax + TaA; TaP = TaF - TaC; TaG = TaC + TaF; TdF = TT - T10; T11 = TT + T10; Tct = TaO - TaP; TaQ = TaO + TaP; T3G = FMA(KP414213562, T3d, T3i); T3j = FNMS(KP414213562, T3i, T3d); T3u = FMA(KP414213562, T3t, T3o); T3H = FNMS(KP414213562, T3o, T3t); } TeQ = TdC + TdB; TdD = TdB - TdC; T6m = T3u - T3j; T3v = T3j + T3u; T7B = FNMS(KP414213562, T7A, T7z); T7K = FMA(KP414213562, T7z, T7A); T7L = FMA(KP414213562, T7C, T7D); T7E = FNMS(KP414213562, T7D, T7C); T6j = T3G - T3H; T3I = T3G + T3H; T99 = T7K + T7L; T7M = T7K - T7L; } } } } } { E Tcw, T9c, T7F, Tev, Teu, TeD, Tep, TeG, Tez, TeE, Tes; { E TbX, TbY, Tc7, TbP, Tar, Tc5, Tc1, Tc0, Tc4, Tba, TbS, TbL, TbQ, TaS, Tbx; E Tc8; { E TeO, TaH, TeR, TeL, TeU, TeZ, Tf2, TeX, Tfh, Tfn, Tfo, Tfm; { E T12, Tfg, Tfj, Tx, Tff, T24, Tfi, Tfk, Th, T1y, T23; TeO = TM - T11; T12 = TM + T11; Tcw = TaG - TaB; TaH = TaB + TaG; T9c = T7B - T7E; T7F = T7B + T7E; Tfg = TeQ + TeP; TeR = TeP - TeQ; TeL = FNMS(KP2_000000000, Tg, T9); Th = FMA(KP2_000000000, Tg, T9); T1y = T1i + T1x; TeU = T1i - T1x; TeZ = T1N - T22; T23 = T1N + T22; Tfj = Tf1 + Tf0; Tf2 = Tf0 - Tf1; Tx = FMA(KP2_000000000, Tw, Th); Tff = FNMS(KP2_000000000, Tw, Th); T24 = T1y + T23; Tfi = T1y - T23; TeX = TeV - TeW; Tfk = TeW + TeV; { E T13, Tfp, Tfl, Tfq; T13 = FMA(KP2_000000000, T12, Tx); Tfp = FNMS(KP2_000000000, T12, Tx); Tfh = FNMS(KP2_000000000, Tfg, Tff); Tfn = FMA(KP2_000000000, Tfg, Tff); Tfl = Tfj - Tfk; Tfq = Tfk + Tfj; R0[0] = FMA(KP2_000000000, T24, T13); R0[WS(rs, 32)] = FNMS(KP2_000000000, T24, T13); R0[WS(rs, 48)] = FMA(KP2_000000000, Tfq, Tfp); R0[WS(rs, 16)] = FNMS(KP2_000000000, Tfq, Tfp); Tfo = Tfi + Tfl; Tfm = Tfi - Tfl; } } { E Tf7, TeN, Tfa, Tf3, Tf8, TeS; R0[WS(rs, 8)] = FMA(KP1_414213562, Tfm, Tfh); R0[WS(rs, 40)] = FNMS(KP1_414213562, Tfm, Tfh); R0[WS(rs, 56)] = FMA(KP1_414213562, Tfo, Tfn); R0[WS(rs, 24)] = FNMS(KP1_414213562, Tfo, Tfn); Tf7 = FMA(KP2_000000000, TeM, TeL); TeN = FNMS(KP2_000000000, TeM, TeL); Tfa = Tf2 - TeZ; Tf3 = TeZ + Tf2; Tf8 = TeO + TeR; TeS = TeO - TeR; { E TbH, TbK, TaI, TaR; { E Taf, Tf9, Tfd, Tf5, TeT, Tfb, TeY, Taq; TbX = FNMS(KP1_414213562, Tae, Ta9); Taf = FMA(KP1_414213562, Tae, Ta9); Tf9 = FNMS(KP1_414213562, Tf8, Tf7); Tfd = FMA(KP1_414213562, Tf8, Tf7); Tf5 = FNMS(KP1_414213562, TeS, TeN); TeT = FMA(KP1_414213562, TeS, TeN); Tfb = TeU - TeX; TeY = TeU + TeX; Taq = FNMS(KP414213562, Tap, Tak); TbY = FMA(KP414213562, Tak, Tap); Tc7 = FNMS(KP707106781, TbG, TbD); TbH = FMA(KP707106781, TbG, TbD); { E Tfc, Tfe, Tf6, Tf4; Tfc = FNMS(KP414213562, Tfb, Tfa); Tfe = FMA(KP414213562, Tfa, Tfb); Tf6 = FMA(KP414213562, TeY, Tf3); Tf4 = FNMS(KP414213562, Tf3, TeY); TbP = FNMS(KP1_847759065, Taq, Taf); Tar = FMA(KP1_847759065, Taq, Taf); R0[WS(rs, 44)] = FMA(KP1_847759065, Tfc, Tf9); R0[WS(rs, 12)] = FNMS(KP1_847759065, Tfc, Tf9); R0[WS(rs, 60)] = FMA(KP1_847759065, Tfe, Tfd); R0[WS(rs, 28)] = FNMS(KP1_847759065, Tfe, Tfd); R0[WS(rs, 52)] = FMA(KP1_847759065, Tf6, Tf5); R0[WS(rs, 20)] = FNMS(KP1_847759065, Tf6, Tf5); R0[WS(rs, 4)] = FMA(KP1_847759065, Tf4, TeT); R0[WS(rs, 36)] = FNMS(KP1_847759065, Tf4, TeT); TbK = TbI + TbJ; Tc5 = TbI - TbJ; } } Tc1 = FNMS(KP707106781, TaH, Taw); TaI = FMA(KP707106781, TaH, Taw); TaR = FMA(KP707106781, TaQ, TaN); Tc0 = FNMS(KP707106781, TaQ, TaN); Tc4 = FNMS(KP707106781, Tb9, TaY); Tba = FMA(KP707106781, Tb9, TaY); TbS = FNMS(KP923879532, TbK, TbH); TbL = FMA(KP923879532, TbK, TbH); TbQ = FMA(KP198912367, TaI, TaR); TaS = FNMS(KP198912367, TaR, TaI); Tbx = Tbl + Tbw; Tc8 = Tbw - Tbl; } } } { E Ten, Teo, Tex, Tef, Tdz, Ter, Teq, TdQ, Tei, Teb, Teg, TdK, Te1, Tey; { E Te7, Tea, TdE, TdJ; { E Tdt, TbR, TbV, TbN, TaT, TbT, Tby, Tdy; Ten = FMA(KP2_000000000, Tds, Tdr); Tdt = FNMS(KP2_000000000, Tds, Tdr); TbR = FNMS(KP1_961570560, TbQ, TbP); TbV = FMA(KP1_961570560, TbQ, TbP); TbN = FNMS(KP1_961570560, TaS, Tar); TaT = FMA(KP1_961570560, TaS, Tar); TbT = FNMS(KP923879532, Tbx, Tba); Tby = FMA(KP923879532, Tbx, Tba); Tdy = Tdu - Tdx; Teo = Tdu + Tdx; Tex = Te6 - Te3; Te7 = Te3 + Te6; { E TbU, TbW, TbO, TbM; TbU = FNMS(KP820678790, TbT, TbS); TbW = FMA(KP820678790, TbS, TbT); TbO = FMA(KP098491403, Tby, TbL); TbM = FNMS(KP098491403, TbL, Tby); Tef = FNMS(KP1_414213562, Tdy, Tdt); Tdz = FMA(KP1_414213562, Tdy, Tdt); R0[WS(rs, 41)] = FMA(KP1_546020906, TbU, TbR); R0[WS(rs, 9)] = FNMS(KP1_546020906, TbU, TbR); R0[WS(rs, 57)] = FMA(KP1_546020906, TbW, TbV); R0[WS(rs, 25)] = FNMS(KP1_546020906, TbW, TbV); R0[WS(rs, 49)] = FMA(KP1_990369453, TbO, TbN); R0[WS(rs, 17)] = FNMS(KP1_990369453, TbO, TbN); R0[WS(rs, 1)] = FMA(KP1_990369453, TbM, TaT); R0[WS(rs, 33)] = FNMS(KP1_990369453, TbM, TaT); Tea = Te8 + Te9; Tev = Te8 - Te9; } } Ter = TdA - TdD; TdE = TdA + TdD; TdJ = TdF + TdI; Teq = TdI - TdF; Teu = TdM - TdP; TdQ = TdM + TdP; Tei = FNMS(KP707106781, Tea, Te7); Teb = FMA(KP707106781, Tea, Te7); Teg = FMA(KP414213562, TdE, TdJ); TdK = FNMS(KP414213562, TdJ, TdE); Te1 = TdV + Te0; Tey = Te0 - TdV; } { E Tcd, TbZ, Tcg, Tc9, Tce, Tc2; { E Teh, Tel, Ted, TdL, Tej, Te2; Teh = FNMS(KP1_847759065, Teg, Tef); Tel = FMA(KP1_847759065, Teg, Tef); Ted = FNMS(KP1_847759065, TdK, Tdz); TdL = FMA(KP1_847759065, TdK, Tdz); Tej = FNMS(KP707106781, Te1, TdQ); Te2 = FMA(KP707106781, Te1, TdQ); { E Tek, Tem, Tee, Tec; Tek = FNMS(KP668178637, Tej, Tei); Tem = FMA(KP668178637, Tei, Tej); Tee = FMA(KP198912367, Te2, Teb); Tec = FNMS(KP198912367, Teb, Te2); Tcd = FMA(KP1_847759065, TbY, TbX); TbZ = FNMS(KP1_847759065, TbY, TbX); R0[WS(rs, 42)] = FMA(KP1_662939224, Tek, Teh); R0[WS(rs, 10)] = FNMS(KP1_662939224, Tek, Teh); R0[WS(rs, 58)] = FMA(KP1_662939224, Tem, Tel); R0[WS(rs, 26)] = FNMS(KP1_662939224, Tem, Tel); R0[WS(rs, 50)] = FMA(KP1_961570560, Tee, Ted); R0[WS(rs, 18)] = FNMS(KP1_961570560, Tee, Ted); R0[WS(rs, 2)] = FMA(KP1_961570560, Tec, TdL); R0[WS(rs, 34)] = FNMS(KP1_961570560, Tec, TdL); } } Tcg = FMA(KP923879532, Tc8, Tc7); Tc9 = FNMS(KP923879532, Tc8, Tc7); Tce = FMA(KP668178637, Tc0, Tc1); Tc2 = FNMS(KP668178637, Tc1, Tc0); { E Tcf, Tcj, Tcb, Tc3, Tch, Tc6; Tcf = FNMS(KP1_662939224, Tce, Tcd); Tcj = FMA(KP1_662939224, Tce, Tcd); Tcb = FMA(KP1_662939224, Tc2, TbZ); Tc3 = FNMS(KP1_662939224, Tc2, TbZ); Tch = FMA(KP923879532, Tc5, Tc4); Tc6 = FNMS(KP923879532, Tc5, Tc4); { E Tci, Tck, Tcc, Tca; Tci = FNMS(KP303346683, Tch, Tcg); Tck = FMA(KP303346683, Tcg, Tch); Tcc = FMA(KP534511135, Tc6, Tc9); Tca = FNMS(KP534511135, Tc9, Tc6); TeD = FMA(KP1_414213562, Teo, Ten); Tep = FNMS(KP1_414213562, Teo, Ten); R0[WS(rs, 45)] = FMA(KP1_913880671, Tci, Tcf); R0[WS(rs, 13)] = FNMS(KP1_913880671, Tci, Tcf); R0[WS(rs, 61)] = FMA(KP1_913880671, Tck, Tcj); R0[WS(rs, 29)] = FNMS(KP1_913880671, Tck, Tcj); R0[WS(rs, 53)] = FMA(KP1_763842528, Tcc, Tcb); R0[WS(rs, 21)] = FNMS(KP1_763842528, Tcc, Tcb); R0[WS(rs, 5)] = FMA(KP1_763842528, Tca, Tc3); R0[WS(rs, 37)] = FNMS(KP1_763842528, Tca, Tc3); } } TeG = FMA(KP707106781, Tey, Tex); Tez = FNMS(KP707106781, Tey, Tex); TeE = FMA(KP414213562, Teq, Ter); Tes = FNMS(KP414213562, Ter, Teq); } } } { E T5L, T5M, T61, T62; { E Td3, Td4, Tdd, TcV, Tcr, Tdb, Td7, Td6, Tda, TcC, TcY, TcR, TcW, Tcy, TcJ; E Tde; { E TcN, TcQ, Tcu, Tcx; { E Tcn, TeF, TeJ, TeB, Tet, TeH, Tew, Tcq; Td3 = FMA(KP1_414213562, Tcm, Tcl); Tcn = FNMS(KP1_414213562, Tcm, Tcl); TeF = FNMS(KP1_847759065, TeE, TeD); TeJ = FMA(KP1_847759065, TeE, TeD); TeB = FMA(KP1_847759065, Tes, Tep); Tet = FNMS(KP1_847759065, Tes, Tep); TeH = FMA(KP707106781, Tev, Teu); Tew = FNMS(KP707106781, Tev, Teu); Tcq = FNMS(KP414213562, Tcp, Tco); Td4 = FMA(KP414213562, Tco, Tcp); Tdd = FMA(KP707106781, TcM, TcL); TcN = FNMS(KP707106781, TcM, TcL); { E TeI, TeK, TeC, TeA; TeI = FNMS(KP198912367, TeH, TeG); TeK = FMA(KP198912367, TeG, TeH); TeC = FMA(KP668178637, Tew, Tez); TeA = FNMS(KP668178637, Tez, Tew); TcV = FMA(KP1_847759065, Tcq, Tcn); Tcr = FNMS(KP1_847759065, Tcq, Tcn); R0[WS(rs, 46)] = FMA(KP1_961570560, TeI, TeF); R0[WS(rs, 14)] = FNMS(KP1_961570560, TeI, TeF); R0[WS(rs, 62)] = FMA(KP1_961570560, TeK, TeJ); R0[WS(rs, 30)] = FNMS(KP1_961570560, TeK, TeJ); R0[WS(rs, 54)] = FMA(KP1_662939224, TeC, TeB); R0[WS(rs, 22)] = FNMS(KP1_662939224, TeC, TeB); R0[WS(rs, 6)] = FMA(KP1_662939224, TeA, Tet); R0[WS(rs, 38)] = FNMS(KP1_662939224, TeA, Tet); TcQ = TcO - TcP; Tdb = TcO + TcP; } } Td7 = FMA(KP707106781, Tct, Tcs); Tcu = FNMS(KP707106781, Tct, Tcs); Tcx = FNMS(KP707106781, Tcw, Tcv); Td6 = FMA(KP707106781, Tcw, Tcv); Tda = FMA(KP707106781, TcB, TcA); TcC = FNMS(KP707106781, TcB, TcA); TcY = FNMS(KP923879532, TcQ, TcN); TcR = FMA(KP923879532, TcQ, TcN); TcW = FMA(KP668178637, Tcu, Tcx); Tcy = FNMS(KP668178637, Tcx, Tcu); TcJ = TcF - TcI; Tde = TcI + TcF; } { E Tdj, Td5, Tdm, Tdf, Tdk, Td8; { E TcX, Td1, TcT, Tcz, TcZ, TcK; TcX = FNMS(KP1_662939224, TcW, TcV); Td1 = FMA(KP1_662939224, TcW, TcV); TcT = FNMS(KP1_662939224, Tcy, Tcr); Tcz = FMA(KP1_662939224, Tcy, Tcr); TcZ = FNMS(KP923879532, TcJ, TcC); TcK = FMA(KP923879532, TcJ, TcC); { E Td0, Td2, TcU, TcS; Td0 = FNMS(KP534511135, TcZ, TcY); Td2 = FMA(KP534511135, TcY, TcZ); TcU = FMA(KP303346683, TcK, TcR); TcS = FNMS(KP303346683, TcR, TcK); Tdj = FMA(KP1_847759065, Td4, Td3); Td5 = FNMS(KP1_847759065, Td4, Td3); R0[WS(rs, 43)] = FMA(KP1_763842528, Td0, TcX); R0[WS(rs, 11)] = FNMS(KP1_763842528, Td0, TcX); R0[WS(rs, 59)] = FMA(KP1_763842528, Td2, Td1); R0[WS(rs, 27)] = FNMS(KP1_763842528, Td2, Td1); R0[WS(rs, 51)] = FMA(KP1_913880671, TcU, TcT); R0[WS(rs, 19)] = FNMS(KP1_913880671, TcU, TcT); R0[WS(rs, 3)] = FMA(KP1_913880671, TcS, Tcz); R0[WS(rs, 35)] = FNMS(KP1_913880671, TcS, Tcz); } } Tdm = FMA(KP923879532, Tde, Tdd); Tdf = FNMS(KP923879532, Tde, Tdd); Tdk = FMA(KP198912367, Td6, Td7); Td8 = FNMS(KP198912367, Td7, Td6); { E T5F, T2R, T5G, T3K, T64, T5S, T5X, T5x, T5U, T4q, T4R, T63, T5P, T5i, T5V; E T5A; { E T5N, T5O, T5R, T3w, T3J, T5Q, T5y, T5z; { E T2p, Tdl, Tdp, Tdh, Td9, Tdn, Tdc, T2Q; T5N = FNMS(KP1_847759065, T2o, T2d); T2p = FMA(KP1_847759065, T2o, T2d); Tdl = FNMS(KP1_961570560, Tdk, Tdj); Tdp = FMA(KP1_961570560, Tdk, Tdj); Tdh = FMA(KP1_961570560, Td8, Td5); Td9 = FNMS(KP1_961570560, Td8, Td5); Tdn = FMA(KP923879532, Tdb, Tda); Tdc = FNMS(KP923879532, Tdb, Tda); T2Q = FNMS(KP198912367, T2P, T2G); T5O = FMA(KP198912367, T2G, T2P); T5R = FNMS(KP923879532, T3v, T38); T3w = FMA(KP923879532, T3v, T38); { E Tdo, Tdq, Tdi, Tdg; Tdo = FNMS(KP098491403, Tdn, Tdm); Tdq = FMA(KP098491403, Tdm, Tdn); Tdi = FMA(KP820678790, Tdc, Tdf); Tdg = FNMS(KP820678790, Tdf, Tdc); T5F = FNMS(KP1_961570560, T2Q, T2p); T2R = FMA(KP1_961570560, T2Q, T2p); R0[WS(rs, 47)] = FMA(KP1_990369453, Tdo, Tdl); R0[WS(rs, 15)] = FNMS(KP1_990369453, Tdo, Tdl); R0[WS(rs, 63)] = FMA(KP1_990369453, Tdq, Tdp); R0[WS(rs, 31)] = FNMS(KP1_990369453, Tdq, Tdp); R0[WS(rs, 55)] = FMA(KP1_546020906, Tdi, Tdh); R0[WS(rs, 23)] = FNMS(KP1_546020906, Tdi, Tdh); R0[WS(rs, 7)] = FMA(KP1_546020906, Tdg, Td9); R0[WS(rs, 39)] = FNMS(KP1_546020906, Tdg, Td9); T3J = FMA(KP923879532, T3I, T3F); T5Q = FNMS(KP923879532, T3I, T3F); } } T5G = FMA(KP098491403, T3w, T3J); T3K = FNMS(KP098491403, T3J, T3w); T64 = FMA(KP820678790, T5Q, T5R); T5S = FNMS(KP820678790, T5R, T5Q); T5X = FNMS(KP923879532, T5w, T5t); T5x = FMA(KP923879532, T5w, T5t); T5U = FNMS(KP923879532, T4p, T42); T4q = FMA(KP923879532, T4p, T42); T4R = FNMS(KP198912367, T4Q, T4H); T5y = FMA(KP198912367, T4H, T4Q); T63 = FMA(KP1_961570560, T5O, T5N); T5P = FNMS(KP1_961570560, T5O, T5N); T5z = FNMS(KP198912367, T58, T5h); T5i = FMA(KP198912367, T5h, T58); T5V = T5y - T5z; T5A = T5y + T5z; } { E T5W, T5I, T5Z, T5J; { E T5D, T3L, T67, T5B, T5Y, T5j, T65, T69, T66, T5k; T5D = FNMS(KP1_990369453, T3K, T2R); T3L = FMA(KP1_990369453, T3K, T2R); T5W = FNMS(KP980785280, T5V, T5U); T67 = FMA(KP980785280, T5V, T5U); T5I = FNMS(KP980785280, T5A, T5x); T5B = FMA(KP980785280, T5A, T5x); T5Y = T5i - T4R; T5j = T4R + T5i; T65 = FNMS(KP1_546020906, T64, T63); T69 = FMA(KP1_546020906, T64, T63); T5Z = FNMS(KP980785280, T5Y, T5X); T66 = FMA(KP980785280, T5Y, T5X); T5J = FNMS(KP980785280, T5j, T4q); T5k = FMA(KP980785280, T5j, T4q); { E T68, T6a, T5E, T5C; T68 = FNMS(KP357805721, T67, T66); T6a = FMA(KP357805721, T66, T67); T5E = FMA(KP049126849, T5k, T5B); T5C = FNMS(KP049126849, T5B, T5k); R1[WS(rs, 60)] = FMA(KP1_883088130, T6a, T69); R1[WS(rs, 28)] = FNMS(KP1_883088130, T6a, T69); R1[WS(rs, 44)] = FMA(KP1_883088130, T68, T65); R1[WS(rs, 12)] = FNMS(KP1_883088130, T68, T65); R1[0] = FMA(KP1_997590912, T5C, T3L); R1[WS(rs, 32)] = FNMS(KP1_997590912, T5C, T3L); R1[WS(rs, 16)] = FNMS(KP1_997590912, T5E, T5D); R1[WS(rs, 48)] = FMA(KP1_997590912, T5E, T5D); } } { E T5H, T5K, T5T, T60; T5L = FMA(KP1_990369453, T5G, T5F); T5H = FNMS(KP1_990369453, T5G, T5F); T5K = FNMS(KP906347169, T5J, T5I); T5M = FMA(KP906347169, T5I, T5J); T61 = FMA(KP1_546020906, T5S, T5P); T5T = FNMS(KP1_546020906, T5S, T5P); T60 = FNMS(KP472964775, T5Z, T5W); T62 = FMA(KP472964775, T5W, T5Z); R1[WS(rs, 40)] = FMA(KP1_481902250, T5K, T5H); R1[WS(rs, 8)] = FNMS(KP1_481902250, T5K, T5H); R1[WS(rs, 4)] = FMA(KP1_807978586, T60, T5T); R1[WS(rs, 36)] = FNMS(KP1_807978586, T60, T5T); } } } } } { E T8B, T8C, T8R, T8S; { E T8v, T7v, T8w, T7O, T8N, T8n, T8U, T8I, T8T, T8F, T8K, T80, T87, T8e, T8L; E T8q; { E T8D, T8E, T8H, T8G, T8o, T8p; { E T7n, T7u, T7G, T7N; T8D = FMA(KP1_847759065, T7m, T7j); T7n = FNMS(KP1_847759065, T7m, T7j); R1[WS(rs, 52)] = FMA(KP1_807978586, T62, T61); R1[WS(rs, 20)] = FNMS(KP1_807978586, T62, T61); R1[WS(rs, 56)] = FMA(KP1_481902250, T5M, T5L); R1[WS(rs, 24)] = FNMS(KP1_481902250, T5M, T5L); T7u = FNMS(KP668178637, T7t, T7q); T8E = FMA(KP668178637, T7q, T7t); T8H = FMA(KP923879532, T7F, T7y); T7G = FNMS(KP923879532, T7F, T7y); T7N = FMA(KP923879532, T7M, T7J); T8G = FNMS(KP923879532, T7M, T7J); T8v = FNMS(KP1_662939224, T7u, T7n); T7v = FMA(KP1_662939224, T7u, T7n); T8w = FMA(KP303346683, T7G, T7N); T7O = FNMS(KP303346683, T7N, T7G); } T8N = FNMS(KP923879532, T8m, T8j); T8n = FMA(KP923879532, T8m, T8j); T8U = FMA(KP534511135, T8G, T8H); T8I = FNMS(KP534511135, T8H, T8G); T8T = FMA(KP1_662939224, T8E, T8D); T8F = FNMS(KP1_662939224, T8E, T8D); T8K = FMA(KP923879532, T7Z, T7S); T80 = FNMS(KP923879532, T7Z, T7S); T87 = FNMS(KP668178637, T86, T83); T8o = FMA(KP668178637, T83, T86); T8p = FMA(KP668178637, T8a, T8d); T8e = FNMS(KP668178637, T8d, T8a); T8L = T8o + T8p; T8q = T8o - T8p; } { E T8M, T8y, T8P, T8z; { E T8t, T7P, T8X, T8r, T8O, T8f, T8V, T8Z, T8W, T8g; T8t = FNMS(KP1_913880671, T7O, T7v); T7P = FMA(KP1_913880671, T7O, T7v); T8M = FNMS(KP831469612, T8L, T8K); T8X = FMA(KP831469612, T8L, T8K); T8y = FNMS(KP831469612, T8q, T8n); T8r = FMA(KP831469612, T8q, T8n); T8O = T8e - T87; T8f = T87 + T8e; T8V = FNMS(KP1_763842528, T8U, T8T); T8Z = FMA(KP1_763842528, T8U, T8T); T8P = FNMS(KP831469612, T8O, T8N); T8W = FMA(KP831469612, T8O, T8N); T8z = FNMS(KP831469612, T8f, T80); T8g = FMA(KP831469612, T8f, T80); { E T8Y, T90, T8u, T8s; T8Y = FNMS(KP250486960, T8X, T8W); T90 = FMA(KP250486960, T8W, T8X); T8u = FMA(KP148335987, T8g, T8r); T8s = FNMS(KP148335987, T8r, T8g); R1[WS(rs, 61)] = FMA(KP1_940062506, T90, T8Z); R1[WS(rs, 29)] = FNMS(KP1_940062506, T90, T8Z); R1[WS(rs, 45)] = FMA(KP1_940062506, T8Y, T8V); R1[WS(rs, 13)] = FNMS(KP1_940062506, T8Y, T8V); R1[WS(rs, 1)] = FMA(KP1_978353019, T8s, T7P); R1[WS(rs, 33)] = FNMS(KP1_978353019, T8s, T7P); R1[WS(rs, 17)] = FNMS(KP1_978353019, T8u, T8t); R1[WS(rs, 49)] = FMA(KP1_978353019, T8u, T8t); } } { E T8x, T8A, T8J, T8Q; T8B = FMA(KP1_913880671, T8w, T8v); T8x = FNMS(KP1_913880671, T8w, T8v); T8A = FNMS(KP741650546, T8z, T8y); T8C = FMA(KP741650546, T8y, T8z); T8R = FMA(KP1_763842528, T8I, T8F); T8J = FNMS(KP1_763842528, T8I, T8F); T8Q = FNMS(KP599376933, T8P, T8M); T8S = FMA(KP599376933, T8M, T8P); R1[WS(rs, 41)] = FMA(KP1_606415062, T8A, T8x); R1[WS(rs, 9)] = FNMS(KP1_606415062, T8A, T8x); R1[WS(rs, 5)] = FMA(KP1_715457220, T8Q, T8J); R1[WS(rs, 37)] = FNMS(KP1_715457220, T8Q, T8J); } } } { E T6R, T6S, T77, T78; { E T6L, T6h, T6M, T6o, T73, T6D, T7a, T6Y, T79, T6V, T70, T6s, T6y, T6v, T71; E T6G; { E T6T, T6U, T6X, T6W, T6E, T6F; { E T6d, T6g, T6k, T6n; T6T = FMA(KP1_847759065, T6c, T6b); T6d = FNMS(KP1_847759065, T6c, T6b); R1[WS(rs, 53)] = FMA(KP1_715457220, T8S, T8R); R1[WS(rs, 21)] = FNMS(KP1_715457220, T8S, T8R); R1[WS(rs, 57)] = FMA(KP1_606415062, T8C, T8B); R1[WS(rs, 25)] = FNMS(KP1_606415062, T8C, T8B); T6g = FNMS(KP668178637, T6f, T6e); T6U = FMA(KP668178637, T6e, T6f); T6X = FMA(KP923879532, T6j, T6i); T6k = FNMS(KP923879532, T6j, T6i); T6n = FNMS(KP923879532, T6m, T6l); T6W = FMA(KP923879532, T6m, T6l); T6L = FMA(KP1_662939224, T6g, T6d); T6h = FNMS(KP1_662939224, T6g, T6d); T6M = FMA(KP534511135, T6k, T6n); T6o = FNMS(KP534511135, T6n, T6k); } T73 = FMA(KP923879532, T6C, T6B); T6D = FNMS(KP923879532, T6C, T6B); T7a = FMA(KP303346683, T6W, T6X); T6Y = FNMS(KP303346683, T6X, T6W); T79 = FMA(KP1_662939224, T6U, T6T); T6V = FNMS(KP1_662939224, T6U, T6T); T70 = FMA(KP923879532, T6r, T6q); T6s = FNMS(KP923879532, T6r, T6q); T6y = FNMS(KP668178637, T6x, T6w); T6E = FMA(KP668178637, T6w, T6x); T6F = FNMS(KP668178637, T6t, T6u); T6v = FMA(KP668178637, T6u, T6t); T71 = T6E + T6F; T6G = T6E - T6F; } { E T72, T6O, T75, T6P; { E T6J, T6p, T7d, T6H, T74, T6z, T7b, T7f, T7c, T6A; T6J = FNMS(KP1_763842528, T6o, T6h); T6p = FMA(KP1_763842528, T6o, T6h); T72 = FNMS(KP831469612, T71, T70); T7d = FMA(KP831469612, T71, T70); T6O = FNMS(KP831469612, T6G, T6D); T6H = FMA(KP831469612, T6G, T6D); T74 = T6y + T6v; T6z = T6v - T6y; T7b = FNMS(KP1_913880671, T7a, T79); T7f = FMA(KP1_913880671, T7a, T79); T75 = FNMS(KP831469612, T74, T73); T7c = FMA(KP831469612, T74, T73); T6P = FNMS(KP831469612, T6z, T6s); T6A = FMA(KP831469612, T6z, T6s); { E T7e, T7g, T6K, T6I; T7e = FNMS(KP148335987, T7d, T7c); T7g = FMA(KP148335987, T7c, T7d); T6K = FMA(KP250486960, T6A, T6H); T6I = FNMS(KP250486960, T6H, T6A); R1[WS(rs, 62)] = FMA(KP1_978353019, T7g, T7f); R1[WS(rs, 30)] = FNMS(KP1_978353019, T7g, T7f); R1[WS(rs, 46)] = FMA(KP1_978353019, T7e, T7b); R1[WS(rs, 14)] = FNMS(KP1_978353019, T7e, T7b); R1[WS(rs, 2)] = FMA(KP1_940062506, T6I, T6p); R1[WS(rs, 34)] = FNMS(KP1_940062506, T6I, T6p); R1[WS(rs, 18)] = FNMS(KP1_940062506, T6K, T6J); R1[WS(rs, 50)] = FMA(KP1_940062506, T6K, T6J); } } { E T6N, T6Q, T6Z, T76; T6R = FMA(KP1_763842528, T6M, T6L); T6N = FNMS(KP1_763842528, T6M, T6L); T6Q = FNMS(KP599376933, T6P, T6O); T6S = FMA(KP599376933, T6O, T6P); T77 = FMA(KP1_913880671, T6Y, T6V); T6Z = FNMS(KP1_913880671, T6Y, T6V); T76 = FNMS(KP741650546, T75, T72); T78 = FMA(KP741650546, T72, T75); R1[WS(rs, 42)] = FMA(KP1_715457220, T6Q, T6N); R1[WS(rs, 10)] = FNMS(KP1_715457220, T6Q, T6N); R1[WS(rs, 6)] = FMA(KP1_606415062, T76, T6Z); R1[WS(rs, 38)] = FNMS(KP1_606415062, T76, T6Z); } } } { E T9B, T97, T9C, T9e, T9T, T9t, Ta0, T9O, T9Z, T9L, T9Q, T9i, T9l, T9o, T9R; E T9w; { E T9J, T9K, T9N, T9M, T9u, T9v; { E T93, T96, T9a, T9d; T9J = FMA(KP1_847759065, T92, T91); T93 = FNMS(KP1_847759065, T92, T91); R1[WS(rs, 54)] = FMA(KP1_606415062, T78, T77); R1[WS(rs, 22)] = FNMS(KP1_606415062, T78, T77); R1[WS(rs, 58)] = FMA(KP1_715457220, T6S, T6R); R1[WS(rs, 26)] = FNMS(KP1_715457220, T6S, T6R); T96 = FNMS(KP198912367, T95, T94); T9K = FMA(KP198912367, T94, T95); T9N = FMA(KP923879532, T99, T98); T9a = FNMS(KP923879532, T99, T98); T9d = FNMS(KP923879532, T9c, T9b); T9M = FMA(KP923879532, T9c, T9b); T9B = FMA(KP1_961570560, T96, T93); T97 = FNMS(KP1_961570560, T96, T93); T9C = FMA(KP820678790, T9a, T9d); T9e = FNMS(KP820678790, T9d, T9a); } T9T = FMA(KP923879532, T9s, T9r); T9t = FNMS(KP923879532, T9s, T9r); Ta0 = FMA(KP098491403, T9M, T9N); T9O = FNMS(KP098491403, T9N, T9M); T9Z = FMA(KP1_961570560, T9K, T9J); T9L = FNMS(KP1_961570560, T9K, T9J); T9Q = FMA(KP923879532, T9h, T9g); T9i = FNMS(KP923879532, T9h, T9g); T9l = FNMS(KP198912367, T9k, T9j); T9u = FMA(KP198912367, T9j, T9k); T9v = FMA(KP198912367, T9m, T9n); T9o = FNMS(KP198912367, T9n, T9m); T9R = T9u + T9v; T9w = T9u - T9v; } { E T9S, T9E, T9V, T9F; { E T9z, T9f, Ta3, T9x, T9U, T9p, Ta1, Ta5, Ta2, T9q; T9z = FNMS(KP1_546020906, T9e, T97); T9f = FMA(KP1_546020906, T9e, T97); T9S = FNMS(KP980785280, T9R, T9Q); Ta3 = FMA(KP980785280, T9R, T9Q); T9E = FNMS(KP980785280, T9w, T9t); T9x = FMA(KP980785280, T9w, T9t); T9U = T9l - T9o; T9p = T9l + T9o; Ta1 = FNMS(KP1_990369453, Ta0, T9Z); Ta5 = FMA(KP1_990369453, Ta0, T9Z); T9V = FNMS(KP980785280, T9U, T9T); Ta2 = FMA(KP980785280, T9U, T9T); T9F = FMA(KP980785280, T9p, T9i); T9q = FNMS(KP980785280, T9p, T9i); { E Ta4, Ta6, T9A, T9y; Ta4 = FNMS(KP049126849, Ta3, Ta2); Ta6 = FMA(KP049126849, Ta2, Ta3); T9A = FMA(KP357805721, T9q, T9x); T9y = FNMS(KP357805721, T9x, T9q); R1[WS(rs, 63)] = FMA(KP1_997590912, Ta6, Ta5); R1[WS(rs, 31)] = FNMS(KP1_997590912, Ta6, Ta5); R1[WS(rs, 47)] = FMA(KP1_997590912, Ta4, Ta1); R1[WS(rs, 15)] = FNMS(KP1_997590912, Ta4, Ta1); R1[WS(rs, 3)] = FMA(KP1_883088130, T9y, T9f); R1[WS(rs, 35)] = FNMS(KP1_883088130, T9y, T9f); R1[WS(rs, 19)] = FNMS(KP1_883088130, T9A, T9z); R1[WS(rs, 51)] = FMA(KP1_883088130, T9A, T9z); } } { E T9D, T9G, T9P, T9W; T9H = FMA(KP1_546020906, T9C, T9B); T9D = FNMS(KP1_546020906, T9C, T9B); T9G = FNMS(KP472964775, T9F, T9E); T9I = FMA(KP472964775, T9E, T9F); T9X = FMA(KP1_990369453, T9O, T9L); T9P = FNMS(KP1_990369453, T9O, T9L); T9W = FNMS(KP906347169, T9V, T9S); T9Y = FMA(KP906347169, T9S, T9V); R1[WS(rs, 43)] = FMA(KP1_807978586, T9G, T9D); R1[WS(rs, 11)] = FNMS(KP1_807978586, T9G, T9D); R1[WS(rs, 7)] = FMA(KP1_481902250, T9W, T9P); R1[WS(rs, 39)] = FNMS(KP1_481902250, T9W, T9P); } } } } } } } } R1[WS(rs, 55)] = FMA(KP1_481902250, T9Y, T9X); R1[WS(rs, 23)] = FNMS(KP1_481902250, T9Y, T9X); R1[WS(rs, 59)] = FMA(KP1_807978586, T9I, T9H); R1[WS(rs, 27)] = FNMS(KP1_807978586, T9I, T9H); } } } static const kr2c_desc desc = { 128, "r2cb_128", {416, 0, 540, 0}, &GENUS }; void X(codelet_r2cb_128) (planner *p) { X(kr2c_register) (p, r2cb_128, &desc); } #else /* HAVE_FMA */ /* Generated by: ../../../genfft/gen_r2cb.native -compact -variables 4 -pipeline-latency 4 -sign 1 -n 128 -name r2cb_128 -include r2cb.h */ /* * This function contains 956 FP additions, 342 FP multiplications, * (or, 812 additions, 198 multiplications, 144 fused multiply/add), * 198 stack variables, 39 constants, and 256 memory accesses */ #include "r2cb.h" static void r2cb_128(R *R0, R *R1, R *Cr, R *Ci, stride rs, stride csr, stride csi, INT v, INT ivs, INT ovs) { DK(KP1_028205488, +1.028205488386443453187387677937631545216098241); DK(KP1_715457220, +1.715457220000544139804539968569540274084981599); DK(KP1_606415062, +1.606415062961289819613353025926283847759138854); DK(KP1_191398608, +1.191398608984866686934073057659939779023852677); DK(KP1_940062506, +1.940062506389087985207968414572200502913731924); DK(KP485960359, +0.485960359806527779896548324154942236641981567); DK(KP293460948, +0.293460948910723503317700259293435639412430633); DK(KP1_978353019, +1.978353019929561946903347476032486127967379067); DK(KP831469612, +0.831469612302545237078788377617905756738560812); DK(KP555570233, +0.555570233019602224742830813948532874374937191); DK(KP855110186, +0.855110186860564188641933713777597068609157259); DK(KP1_807978586, +1.807978586246886663172400594461074097420264050); DK(KP1_481902250, +1.481902250709918182351233794990325459457910619); DK(KP1_343117909, +1.343117909694036801250753700854843606457501264); DK(KP1_883088130, +1.883088130366041556825018805199004714371179592); DK(KP673779706, +0.673779706784440101378506425238295140955533559); DK(KP098135348, +0.098135348654836028509909953885365316629490726); DK(KP1_997590912, +1.997590912410344785429543209518201388886407229); DK(KP980785280, +0.980785280403230449126182236134239036973933731); DK(KP195090322, +0.195090322016128267848284868477022240927691618); DK(KP580569354, +0.580569354508924735272384751634790549382952557); DK(KP1_913880671, +1.913880671464417729871595773960539938965698411); DK(KP942793473, +0.942793473651995297112775251810508755314920638); DK(KP1_763842528, +1.763842528696710059425513727320776699016885241); DK(KP1_111140466, +1.111140466039204449485661627897065748749874382); DK(KP1_662939224, +1.662939224605090474157576755235811513477121624); DK(KP1_268786568, +1.268786568327290996430343226450986741351374190); DK(KP1_546020906, +1.546020906725473921621813219516939601942082586); DK(KP196034280, +0.196034280659121203988391127777283691722273346); DK(KP1_990369453, +1.990369453344393772489673906218959843150949737); DK(KP390180644, +0.390180644032256535696569736954044481855383236); DK(KP1_961570560, +1.961570560806460898252364472268478073947867462); DK(KP382683432, +0.382683432365089771728459984030398866761344562); DK(KP923879532, +0.923879532511286756128183189396788286822416626); DK(KP707106781, +0.707106781186547524400844362104849039284835938); DK(KP765366864, +0.765366864730179543456919968060797733522689125); DK(KP1_847759065, +1.847759065022573512256366378793576573644833252); DK(KP1_414213562, +1.414213562373095048801688724209698078569671875); DK(KP2_000000000, +2.000000000000000000000000000000000000000000000); { INT i; for (i = v; i > 0; i = i - 1, R0 = R0 + ovs, R1 = R1 + ovs, Cr = Cr + ivs, Ci = Ci + ivs, MAKE_VOLATILE_STRIDE(512, rs), MAKE_VOLATILE_STRIDE(512, csr), MAKE_VOLATILE_STRIDE(512, csi)) { E Ta, T6q, T2a, T5k, T8x, Tbx, TcF, Ten, Th, T6r, T2j, T5l, T8E, Tby, TcI; E Teo, Tx, T6t, TcM, Teq, TcP, Ter, T2t, T5n, T2C, T5o, T8Q, TbA, T8X, TbB; E T6w, T7L, T1j, T6L, Tde, TeC, TdL, TeR, T3v, T5z, T4I, T5O, T9O, TbM, TaV; E Tc1, T78, T7Z, TN, T6z, TcU, Teu, Td8, Tey, T2N, T5r, T3j, T5v, T9a, TbE; E T9A, TbI, T6H, T7O, T1O, T7V, T48, T4u, Tds, TeG, T5E, T5K, Taf, TbP, Tdp; E TeF, T6U, T72, Tam, TbQ, T23, T7U, T4r, T4v, Tdz, TeJ, T5H, T5L, Tay, TbS; E Tdw, TeI, T6Z, T73, TaF, TbT, T1y, T75, Tdl, TeQ, TdI, TeD, T3O, T5N, T4z; E T5A, Ta3, Tc0, TaO, TbN, T6O, T80, T12, T6E, Td1, Tex, Td5, Tev, T36, T5u; E T3a, T5s, T9p, TbH, T9t, TbF, T6C, T7P; { E T5, T8s, T3, T8q, T9, T8u, T29, T8v, T6, T26; { E T4, T8r, T1, T2; T4 = Cr[WS(csr, 32)]; T5 = KP2_000000000 * T4; T8r = Ci[WS(csi, 32)]; T8s = KP2_000000000 * T8r; T1 = Cr[0]; T2 = Cr[WS(csr, 64)]; T3 = T1 + T2; T8q = T1 - T2; { E T7, T8, T27, T28; T7 = Cr[WS(csr, 16)]; T8 = Cr[WS(csr, 48)]; T9 = KP2_000000000 * (T7 + T8); T8u = T7 - T8; T27 = Ci[WS(csi, 16)]; T28 = Ci[WS(csi, 48)]; T29 = KP2_000000000 * (T27 - T28); T8v = T27 + T28; } } T6 = T3 + T5; Ta = T6 + T9; T6q = T6 - T9; T26 = T3 - T5; T2a = T26 - T29; T5k = T26 + T29; { E T8t, T8w, TcD, TcE; T8t = T8q - T8s; T8w = KP1_414213562 * (T8u - T8v); T8x = T8t + T8w; Tbx = T8t - T8w; TcD = T8q + T8s; TcE = KP1_414213562 * (T8u + T8v); TcF = TcD - TcE; Ten = TcD + TcE; } } { E Td, T8y, T2e, T8C, Tg, T8B, T2h, T8z, T2b, T2i; { E Tb, Tc, T2c, T2d; Tb = Cr[WS(csr, 8)]; Tc = Cr[WS(csr, 56)]; Td = Tb + Tc; T8y = Tb - Tc; T2c = Ci[WS(csi, 8)]; T2d = Ci[WS(csi, 56)]; T2e = T2c - T2d; T8C = T2c + T2d; } { E Te, Tf, T2f, T2g; Te = Cr[WS(csr, 40)]; Tf = Cr[WS(csr, 24)]; Tg = Te + Tf; T8B = Te - Tf; T2f = Ci[WS(csi, 40)]; T2g = Ci[WS(csi, 24)]; T2h = T2f - T2g; T8z = T2f + T2g; } Th = KP2_000000000 * (Td + Tg); T6r = KP2_000000000 * (T2h + T2e); T2b = Td - Tg; T2i = T2e - T2h; T2j = KP1_414213562 * (T2b - T2i); T5l = KP1_414213562 * (T2b + T2i); { E T8A, T8D, TcG, TcH; T8A = T8y - T8z; T8D = T8B + T8C; T8E = FNMS(KP765366864, T8D, KP1_847759065 * T8A); Tby = FMA(KP765366864, T8A, KP1_847759065 * T8D); TcG = T8y + T8z; TcH = T8C - T8B; TcI = FNMS(KP1_847759065, TcH, KP765366864 * TcG); Teo = FMA(KP1_847759065, TcG, KP765366864 * TcH); } } { E Tl, T8G, T2x, T8V, To, T8U, T2A, T8H, Tv, T8S, T2o, T8O, Ts, T8R, T2r; E T8L; { E Tj, Tk, T2y, T2z; Tj = Cr[WS(csr, 4)]; Tk = Cr[WS(csr, 60)]; Tl = Tj + Tk; T8G = Tj - Tk; { E T2v, T2w, Tm, Tn; T2v = Ci[WS(csi, 4)]; T2w = Ci[WS(csi, 60)]; T2x = T2v - T2w; T8V = T2v + T2w; Tm = Cr[WS(csr, 36)]; Tn = Cr[WS(csr, 28)]; To = Tm + Tn; T8U = Tm - Tn; } T2y = Ci[WS(csi, 36)]; T2z = Ci[WS(csi, 28)]; T2A = T2y - T2z; T8H = T2y + T2z; { E Tt, Tu, T8M, T2m, T2n, T8N; Tt = Cr[WS(csr, 12)]; Tu = Cr[WS(csr, 52)]; T8M = Tt - Tu; T2m = Ci[WS(csi, 52)]; T2n = Ci[WS(csi, 12)]; T8N = T2n + T2m; Tv = Tt + Tu; T8S = T8M + T8N; T2o = T2m - T2n; T8O = T8M - T8N; } { E Tq, Tr, T8J, T2p, T2q, T8K; Tq = Cr[WS(csr, 20)]; Tr = Cr[WS(csr, 44)]; T8J = Tq - Tr; T2p = Ci[WS(csi, 20)]; T2q = Ci[WS(csi, 44)]; T8K = T2p + T2q; Ts = Tq + Tr; T8R = T8J + T8K; T2r = T2p - T2q; T8L = T8J - T8K; } } { E Tp, Tw, TcK, TcL; Tp = Tl + To; Tw = Ts + Tv; Tx = KP2_000000000 * (Tp + Tw); T6t = Tp - Tw; TcK = T8G + T8H; TcL = KP707106781 * (T8R + T8S); TcM = TcK - TcL; Teq = TcK + TcL; } { E TcN, TcO, T2l, T2s; TcN = KP707106781 * (T8L - T8O); TcO = T8V - T8U; TcP = TcN + TcO; Ter = TcO - TcN; T2l = Tl - To; T2s = T2o - T2r; T2t = T2l + T2s; T5n = T2l - T2s; } { E T2u, T2B, T8I, T8P; T2u = Ts - Tv; T2B = T2x - T2A; T2C = T2u + T2B; T5o = T2B - T2u; T8I = T8G - T8H; T8P = KP707106781 * (T8L + T8O); T8Q = T8I + T8P; TbA = T8I - T8P; } { E T8T, T8W, T6u, T6v; T8T = KP707106781 * (T8R - T8S); T8W = T8U + T8V; T8X = T8T + T8W; TbB = T8W - T8T; T6u = T2A + T2x; T6v = T2r + T2o; T6w = T6u - T6v; T7L = KP2_000000000 * (T6v + T6u); } } { E T17, T9E, T4D, TaT, T1a, TaS, T4G, T9F, T1h, TaQ, T3q, T9M, T1e, TaP, T3t; E T9J; { E T15, T16, T4E, T4F; T15 = Cr[WS(csr, 1)]; T16 = Cr[WS(csr, 63)]; T17 = T15 + T16; T9E = T15 - T16; { E T4B, T4C, T18, T19; T4B = Ci[WS(csi, 1)]; T4C = Ci[WS(csi, 63)]; T4D = T4B - T4C; TaT = T4B + T4C; T18 = Cr[WS(csr, 33)]; T19 = Cr[WS(csr, 31)]; T1a = T18 + T19; TaS = T18 - T19; } T4E = Ci[WS(csi, 33)]; T4F = Ci[WS(csi, 31)]; T4G = T4E - T4F; T9F = T4E + T4F; { E T1f, T1g, T9K, T3o, T3p, T9L; T1f = Cr[WS(csr, 15)]; T1g = Cr[WS(csr, 49)]; T9K = T1f - T1g; T3o = Ci[WS(csi, 49)]; T3p = Ci[WS(csi, 15)]; T9L = T3p + T3o; T1h = T1f + T1g; TaQ = T9K + T9L; T3q = T3o - T3p; T9M = T9K - T9L; } { E T1c, T1d, T9H, T3r, T3s, T9I; T1c = Cr[WS(csr, 17)]; T1d = Cr[WS(csr, 47)]; T9H = T1c - T1d; T3r = Ci[WS(csi, 17)]; T3s = Ci[WS(csi, 47)]; T9I = T3r + T3s; T1e = T1c + T1d; TaP = T9H + T9I; T3t = T3r - T3s; T9J = T9H - T9I; } } { E T1b, T1i, Tdc, Tdd; T1b = T17 + T1a; T1i = T1e + T1h; T1j = T1b + T1i; T6L = T1b - T1i; Tdc = T9E + T9F; Tdd = KP707106781 * (TaP + TaQ); Tde = Tdc - Tdd; TeC = Tdc + Tdd; } { E TdJ, TdK, T3n, T3u; TdJ = KP707106781 * (T9J - T9M); TdK = TaT - TaS; TdL = TdJ + TdK; TeR = TdK - TdJ; T3n = T17 - T1a; T3u = T3q - T3t; T3v = T3n + T3u; T5z = T3n - T3u; } { E T4A, T4H, T9G, T9N; T4A = T1e - T1h; T4H = T4D - T4G; T4I = T4A + T4H; T5O = T4H - T4A; T9G = T9E - T9F; T9N = KP707106781 * (T9J + T9M); T9O = T9G + T9N; TbM = T9G - T9N; } { E TaR, TaU, T76, T77; TaR = KP707106781 * (TaP - TaQ); TaU = TaS + TaT; TaV = TaR + TaU; Tc1 = TaU - TaR; T76 = T4G + T4D; T77 = T3t + T3q; T78 = T76 - T77; T7Z = T77 + T76; } } { E TB, T90, T3e, T9y, TE, T9x, T3h, T91, TL, T9v, T2I, T98, TI, T9u, T2L; E T95; { E Tz, TA, T3f, T3g; Tz = Cr[WS(csr, 2)]; TA = Cr[WS(csr, 62)]; TB = Tz + TA; T90 = Tz - TA; { E T3c, T3d, TC, TD; T3c = Ci[WS(csi, 2)]; T3d = Ci[WS(csi, 62)]; T3e = T3c - T3d; T9y = T3c + T3d; TC = Cr[WS(csr, 34)]; TD = Cr[WS(csr, 30)]; TE = TC + TD; T9x = TC - TD; } T3f = Ci[WS(csi, 34)]; T3g = Ci[WS(csi, 30)]; T3h = T3f - T3g; T91 = T3f + T3g; { E TJ, TK, T96, T2G, T2H, T97; TJ = Cr[WS(csr, 14)]; TK = Cr[WS(csr, 50)]; T96 = TJ - TK; T2G = Ci[WS(csi, 50)]; T2H = Ci[WS(csi, 14)]; T97 = T2H + T2G; TL = TJ + TK; T9v = T96 + T97; T2I = T2G - T2H; T98 = T96 - T97; } { E TG, TH, T93, T2J, T2K, T94; TG = Cr[WS(csr, 18)]; TH = Cr[WS(csr, 46)]; T93 = TG - TH; T2J = Ci[WS(csi, 18)]; T2K = Ci[WS(csi, 46)]; T94 = T2J + T2K; TI = TG + TH; T9u = T93 + T94; T2L = T2J - T2K; T95 = T93 - T94; } } { E TF, TM, TcS, TcT; TF = TB + TE; TM = TI + TL; TN = TF + TM; T6z = TF - TM; TcS = T90 + T91; TcT = KP707106781 * (T9u + T9v); TcU = TcS - TcT; Teu = TcS + TcT; } { E Td6, Td7, T2F, T2M; Td6 = KP707106781 * (T95 - T98); Td7 = T9y - T9x; Td8 = Td6 + Td7; Tey = Td7 - Td6; T2F = TB - TE; T2M = T2I - T2L; T2N = T2F + T2M; T5r = T2F - T2M; } { E T3b, T3i, T92, T99; T3b = TI - TL; T3i = T3e - T3h; T3j = T3b + T3i; T5v = T3i - T3b; T92 = T90 - T91; T99 = KP707106781 * (T95 + T98); T9a = T92 + T99; TbE = T92 - T99; } { E T9w, T9z, T6F, T6G; T9w = KP707106781 * (T9u - T9v); T9z = T9x + T9y; T9A = T9w + T9z; TbI = T9z - T9w; T6F = T3h + T3e; T6G = T2L + T2I; T6H = T6F - T6G; T7O = T6G + T6F; } } { E T1G, Taj, T3Q, Ta5, T46, Tak, T6R, Ta6, T1N, Tag, Tah, T3X, T3Z, Taa, Tad; E T6S, Tdn, Tdo; { E T1A, T1B, T1C, T1D, T1E, T1F; T1A = Cr[WS(csr, 5)]; T1B = Cr[WS(csr, 59)]; T1C = T1A + T1B; T1D = Cr[WS(csr, 37)]; T1E = Cr[WS(csr, 27)]; T1F = T1D + T1E; T1G = T1C + T1F; Taj = T1D - T1E; T3Q = T1C - T1F; Ta5 = T1A - T1B; } { E T40, T41, T42, T43, T44, T45; T40 = Ci[WS(csi, 5)]; T41 = Ci[WS(csi, 59)]; T42 = T40 - T41; T43 = Ci[WS(csi, 37)]; T44 = Ci[WS(csi, 27)]; T45 = T43 - T44; T46 = T42 - T45; Tak = T40 + T41; T6R = T45 + T42; Ta6 = T43 + T44; } { E T1J, Ta8, T3W, Ta9, T1M, Tab, T3T, Tac; { E T1H, T1I, T3U, T3V; T1H = Cr[WS(csr, 21)]; T1I = Cr[WS(csr, 43)]; T1J = T1H + T1I; Ta8 = T1H - T1I; T3U = Ci[WS(csi, 21)]; T3V = Ci[WS(csi, 43)]; T3W = T3U - T3V; Ta9 = T3U + T3V; } { E T1K, T1L, T3R, T3S; T1K = Cr[WS(csr, 11)]; T1L = Cr[WS(csr, 53)]; T1M = T1K + T1L; Tab = T1K - T1L; T3R = Ci[WS(csi, 53)]; T3S = Ci[WS(csi, 11)]; T3T = T3R - T3S; Tac = T3S + T3R; } T1N = T1J + T1M; Tag = Ta8 + Ta9; Tah = Tab + Tac; T3X = T3T - T3W; T3Z = T1J - T1M; Taa = Ta8 - Ta9; Tad = Tab - Tac; T6S = T3W + T3T; } T1O = T1G + T1N; T7V = T6S + T6R; { E T3Y, T47, Tdq, Tdr; T3Y = T3Q + T3X; T47 = T3Z + T46; T48 = FNMS(KP382683432, T47, KP923879532 * T3Y); T4u = FMA(KP382683432, T3Y, KP923879532 * T47); Tdq = KP707106781 * (Taa - Tad); Tdr = Tak - Taj; Tds = Tdq + Tdr; TeG = Tdr - Tdq; } { E T5C, T5D, Ta7, Tae; T5C = T3Q - T3X; T5D = T46 - T3Z; T5E = FNMS(KP923879532, T5D, KP382683432 * T5C); T5K = FMA(KP923879532, T5C, KP382683432 * T5D); Ta7 = Ta5 - Ta6; Tae = KP707106781 * (Taa + Tad); Taf = Ta7 + Tae; TbP = Ta7 - Tae; } Tdn = Ta5 + Ta6; Tdo = KP707106781 * (Tag + Tah); Tdp = Tdn - Tdo; TeF = Tdn + Tdo; { E T6Q, T6T, Tai, Tal; T6Q = T1G - T1N; T6T = T6R - T6S; T6U = T6Q - T6T; T72 = T6Q + T6T; Tai = KP707106781 * (Tag - Tah); Tal = Taj + Tak; Tam = Tai + Tal; TbQ = Tal - Tai; } } { E T1V, TaC, T49, Tao, T4p, TaD, T6W, Tap, T22, Taz, TaA, T4g, T4i, Tat, Taw; E T6X, Tdu, Tdv; { E T1P, T1Q, T1R, T1S, T1T, T1U; T1P = Cr[WS(csr, 3)]; T1Q = Cr[WS(csr, 61)]; T1R = T1P + T1Q; T1S = Cr[WS(csr, 29)]; T1T = Cr[WS(csr, 35)]; T1U = T1S + T1T; T1V = T1R + T1U; TaC = T1S - T1T; T49 = T1R - T1U; Tao = T1P - T1Q; } { E T4j, T4k, T4l, T4m, T4n, T4o; T4j = Ci[WS(csi, 61)]; T4k = Ci[WS(csi, 3)]; T4l = T4j - T4k; T4m = Ci[WS(csi, 29)]; T4n = Ci[WS(csi, 35)]; T4o = T4m - T4n; T4p = T4l - T4o; TaD = T4k + T4j; T6W = T4o + T4l; Tap = T4m + T4n; } { E T1Y, Tar, T4f, Tas, T21, Tau, T4c, Tav; { E T1W, T1X, T4d, T4e; T1W = Cr[WS(csr, 13)]; T1X = Cr[WS(csr, 51)]; T1Y = T1W + T1X; Tar = T1W - T1X; T4d = Ci[WS(csi, 13)]; T4e = Ci[WS(csi, 51)]; T4f = T4d - T4e; Tas = T4d + T4e; } { E T1Z, T20, T4a, T4b; T1Z = Cr[WS(csr, 19)]; T20 = Cr[WS(csr, 45)]; T21 = T1Z + T20; Tau = T1Z - T20; T4a = Ci[WS(csi, 45)]; T4b = Ci[WS(csi, 19)]; T4c = T4a - T4b; Tav = T4b + T4a; } T22 = T1Y + T21; Taz = Tar + Tas; TaA = Tau + Tav; T4g = T4c - T4f; T4i = T1Y - T21; Tat = Tar - Tas; Taw = Tau - Tav; T6X = T4f + T4c; } T23 = T1V + T22; T7U = T6X + T6W; { E T4h, T4q, Tdx, Tdy; T4h = T49 + T4g; T4q = T4i + T4p; T4r = FMA(KP923879532, T4h, KP382683432 * T4q); T4v = FNMS(KP382683432, T4h, KP923879532 * T4q); Tdx = KP707106781 * (Tat - Taw); Tdy = TaC + TaD; Tdz = Tdx - Tdy; TeJ = Tdx + Tdy; } { E T5F, T5G, Taq, Tax; T5F = T49 - T4g; T5G = T4p - T4i; T5H = FMA(KP382683432, T5F, KP923879532 * T5G); T5L = FNMS(KP923879532, T5F, KP382683432 * T5G); Taq = Tao - Tap; Tax = KP707106781 * (Tat + Taw); Tay = Taq + Tax; TbS = Taq - Tax; } Tdu = Tao + Tap; Tdv = KP707106781 * (Taz + TaA); Tdw = Tdu - Tdv; TeI = Tdu + Tdv; { E T6V, T6Y, TaB, TaE; T6V = T1V - T22; T6Y = T6W - T6X; T6Z = T6V + T6Y; T73 = T6Y - T6V; TaB = KP707106781 * (Taz - TaA); TaE = TaC - TaD; TaF = TaB + TaE; TbT = TaE - TaB; } } { E T1m, T3z, T1p, T3C, T3w, T3D, Tdg, Tdf, T9U, T9R, T1t, T3I, T1w, T3L, T3F; E T3M, Tdj, Tdi, Ta1, T9Y; { E T9P, T9T, T9S, T9Q; { E T1k, T1l, T3x, T3y; T1k = Cr[WS(csr, 9)]; T1l = Cr[WS(csr, 55)]; T1m = T1k + T1l; T9P = T1k - T1l; T3x = Ci[WS(csi, 9)]; T3y = Ci[WS(csi, 55)]; T3z = T3x - T3y; T9T = T3x + T3y; } { E T1n, T1o, T3A, T3B; T1n = Cr[WS(csr, 41)]; T1o = Cr[WS(csr, 23)]; T1p = T1n + T1o; T9S = T1n - T1o; T3A = Ci[WS(csi, 41)]; T3B = Ci[WS(csi, 23)]; T3C = T3A - T3B; T9Q = T3A + T3B; } T3w = T1m - T1p; T3D = T3z - T3C; Tdg = T9T - T9S; Tdf = T9P + T9Q; T9U = T9S + T9T; T9R = T9P - T9Q; } { E T9W, Ta0, T9Z, T9X; { E T1r, T1s, T3G, T3H; T1r = Cr[WS(csr, 7)]; T1s = Cr[WS(csr, 57)]; T1t = T1r + T1s; T9W = T1r - T1s; T3G = Ci[WS(csi, 57)]; T3H = Ci[WS(csi, 7)]; T3I = T3G - T3H; Ta0 = T3H + T3G; } { E T1u, T1v, T3J, T3K; T1u = Cr[WS(csr, 25)]; T1v = Cr[WS(csr, 39)]; T1w = T1u + T1v; T9Z = T1u - T1v; T3J = Ci[WS(csi, 25)]; T3K = Ci[WS(csi, 39)]; T3L = T3J - T3K; T9X = T3J + T3K; } T3F = T1t - T1w; T3M = T3I - T3L; Tdj = T9Z + Ta0; Tdi = T9W + T9X; Ta1 = T9Z - Ta0; T9Y = T9W - T9X; } { E T1q, T1x, Tdh, Tdk; T1q = T1m + T1p; T1x = T1t + T1w; T1y = T1q + T1x; T75 = T1q - T1x; Tdh = FNMS(KP923879532, Tdg, KP382683432 * Tdf); Tdk = FNMS(KP923879532, Tdj, KP382683432 * Tdi); Tdl = Tdh + Tdk; TeQ = Tdh - Tdk; } { E TdG, TdH, T3E, T3N; TdG = FMA(KP923879532, Tdf, KP382683432 * Tdg); TdH = FMA(KP923879532, Tdi, KP382683432 * Tdj); TdI = TdG - TdH; TeD = TdG + TdH; T3E = T3w - T3D; T3N = T3F + T3M; T3O = KP707106781 * (T3E + T3N); T5N = KP707106781 * (T3E - T3N); } { E T4x, T4y, T9V, Ta2; T4x = T3w + T3D; T4y = T3M - T3F; T4z = KP707106781 * (T4x + T4y); T5A = KP707106781 * (T4y - T4x); T9V = FNMS(KP382683432, T9U, KP923879532 * T9R); Ta2 = FMA(KP923879532, T9Y, KP382683432 * Ta1); Ta3 = T9V + Ta2; Tc0 = T9V - Ta2; } { E TaM, TaN, T6M, T6N; TaM = FMA(KP382683432, T9R, KP923879532 * T9U); TaN = FNMS(KP382683432, T9Y, KP923879532 * Ta1); TaO = TaM + TaN; TbN = TaN - TaM; T6M = T3L + T3I; T6N = T3C + T3z; T6O = T6M - T6N; T80 = T6N + T6M; } } { E TQ, T2R, TT, T2U, T2O, T2V, TcW, TcV, T9g, T9d, TX, T30, T10, T33, T2X; E T34, TcZ, TcY, T9n, T9k; { E T9b, T9f, T9e, T9c; { E TO, TP, T2P, T2Q; TO = Cr[WS(csr, 10)]; TP = Cr[WS(csr, 54)]; TQ = TO + TP; T9b = TO - TP; T2P = Ci[WS(csi, 10)]; T2Q = Ci[WS(csi, 54)]; T2R = T2P - T2Q; T9f = T2P + T2Q; } { E TR, TS, T2S, T2T; TR = Cr[WS(csr, 42)]; TS = Cr[WS(csr, 22)]; TT = TR + TS; T9e = TR - TS; T2S = Ci[WS(csi, 42)]; T2T = Ci[WS(csi, 22)]; T2U = T2S - T2T; T9c = T2S + T2T; } T2O = TQ - TT; T2V = T2R - T2U; TcW = T9f - T9e; TcV = T9b + T9c; T9g = T9e + T9f; T9d = T9b - T9c; } { E T9i, T9m, T9l, T9j; { E TV, TW, T2Y, T2Z; TV = Cr[WS(csr, 6)]; TW = Cr[WS(csr, 58)]; TX = TV + TW; T9i = TV - TW; T2Y = Ci[WS(csi, 58)]; T2Z = Ci[WS(csi, 6)]; T30 = T2Y - T2Z; T9m = T2Z + T2Y; } { E TY, TZ, T31, T32; TY = Cr[WS(csr, 26)]; TZ = Cr[WS(csr, 38)]; T10 = TY + TZ; T9l = TY - TZ; T31 = Ci[WS(csi, 26)]; T32 = Ci[WS(csi, 38)]; T33 = T31 - T32; T9j = T31 + T32; } T2X = TX - T10; T34 = T30 - T33; TcZ = T9l + T9m; TcY = T9i + T9j; T9n = T9l - T9m; T9k = T9i - T9j; } { E TU, T11, TcX, Td0; TU = TQ + TT; T11 = TX + T10; T12 = TU + T11; T6E = TU - T11; TcX = FNMS(KP923879532, TcW, KP382683432 * TcV); Td0 = FNMS(KP923879532, TcZ, KP382683432 * TcY); Td1 = TcX + Td0; Tex = TcX - Td0; } { E Td3, Td4, T2W, T35; Td3 = FMA(KP923879532, TcV, KP382683432 * TcW); Td4 = FMA(KP923879532, TcY, KP382683432 * TcZ); Td5 = Td3 - Td4; Tev = Td3 + Td4; T2W = T2O - T2V; T35 = T2X + T34; T36 = KP707106781 * (T2W + T35); T5u = KP707106781 * (T2W - T35); } { E T38, T39, T9h, T9o; T38 = T2O + T2V; T39 = T34 - T2X; T3a = KP707106781 * (T38 + T39); T5s = KP707106781 * (T39 - T38); T9h = FNMS(KP382683432, T9g, KP923879532 * T9d); T9o = FMA(KP923879532, T9k, KP382683432 * T9n); T9p = T9h + T9o; TbH = T9h - T9o; } { E T9r, T9s, T6A, T6B; T9r = FMA(KP382683432, T9d, KP923879532 * T9g); T9s = FNMS(KP382683432, T9k, KP923879532 * T9n); T9t = T9r + T9s; TbF = T9s - T9r; T6A = T33 + T30; T6B = T2U + T2R; T6C = T6A - T6B; T7P = T6B + T6A; } } { E T13, T8f, Ty, T8e, T25, T8h, T8k, T8p, Ti, T14, T8o; T13 = KP2_000000000 * (TN + T12); T8f = KP2_000000000 * (T7P + T7O); Ti = Ta + Th; Ty = Ti + Tx; T8e = Ti - Tx; { E T1z, T24, T8i, T8j; T1z = T1j + T1y; T24 = T1O + T23; T25 = KP2_000000000 * (T1z + T24); T8h = T1z - T24; T8i = T80 + T7Z; T8j = T7V + T7U; T8k = T8i - T8j; T8p = KP2_000000000 * (T8j + T8i); } T14 = Ty + T13; R0[WS(rs, 32)] = T14 - T25; R0[0] = T14 + T25; T8o = Ty - T13; R0[WS(rs, 16)] = T8o - T8p; R0[WS(rs, 48)] = T8o + T8p; { E T8g, T8l, T8m, T8n; T8g = T8e - T8f; T8l = KP1_414213562 * (T8h - T8k); R0[WS(rs, 40)] = T8g - T8l; R0[WS(rs, 8)] = T8g + T8l; T8m = T8e + T8f; T8n = KP1_414213562 * (T8h + T8k); R0[WS(rs, 24)] = T8m - T8n; R0[WS(rs, 56)] = T8m + T8n; } } { E T7M, T86, T82, T8a, T7R, T87, T7X, T89, T7K, T7Y, T81; T7K = Ta - Th; T7M = T7K - T7L; T86 = T7K + T7L; T7Y = T1O - T23; T81 = T7Z - T80; T82 = T7Y + T81; T8a = T81 - T7Y; { E T7N, T7Q, T7T, T7W; T7N = TN - T12; T7Q = T7O - T7P; T7R = KP1_414213562 * (T7N - T7Q); T87 = KP1_414213562 * (T7N + T7Q); T7T = T1j - T1y; T7W = T7U - T7V; T7X = T7T + T7W; T89 = T7T - T7W; } { E T7S, T83, T8c, T8d; T7S = T7M + T7R; T83 = FNMS(KP765366864, T82, KP1_847759065 * T7X); R0[WS(rs, 36)] = T7S - T83; R0[WS(rs, 4)] = T7S + T83; T8c = T86 + T87; T8d = FMA(KP1_847759065, T89, KP765366864 * T8a); R0[WS(rs, 28)] = T8c - T8d; R0[WS(rs, 60)] = T8c + T8d; } { E T84, T85, T88, T8b; T84 = T7M - T7R; T85 = FMA(KP765366864, T7X, KP1_847759065 * T82); R0[WS(rs, 20)] = T84 - T85; R0[WS(rs, 52)] = T84 + T85; T88 = T86 - T87; T8b = FNMS(KP1_847759065, T8a, KP765366864 * T89); R0[WS(rs, 44)] = T88 - T8b; R0[WS(rs, 12)] = T88 + T8b; } } { E T2E, T4O, T4K, T4S, T3l, T4P, T4t, T4R; { E T2k, T2D, T4w, T4J; T2k = T2a + T2j; T2D = FNMS(KP765366864, T2C, KP1_847759065 * T2t); T2E = T2k + T2D; T4O = T2k - T2D; T4w = T4u + T4v; T4J = T4z + T4I; T4K = T4w + T4J; T4S = T4J - T4w; } { E T37, T3k, T3P, T4s; T37 = T2N + T36; T3k = T3a + T3j; T3l = FNMS(KP390180644, T3k, KP1_961570560 * T37); T4P = FMA(KP390180644, T37, KP1_961570560 * T3k); T3P = T3v + T3O; T4s = T48 + T4r; T4t = T3P + T4s; T4R = T3P - T4s; } { E T3m, T4L, T4U, T4V; T3m = T2E + T3l; T4L = FNMS(KP196034280, T4K, KP1_990369453 * T4t); R0[WS(rs, 33)] = T3m - T4L; R0[WS(rs, 1)] = T3m + T4L; T4U = T4O + T4P; T4V = FMA(KP1_546020906, T4R, KP1_268786568 * T4S); R0[WS(rs, 25)] = T4U - T4V; R0[WS(rs, 57)] = T4U + T4V; } { E T4M, T4N, T4Q, T4T; T4M = T2E - T3l; T4N = FMA(KP196034280, T4t, KP1_990369453 * T4K); R0[WS(rs, 17)] = T4M - T4N; R0[WS(rs, 49)] = T4M + T4N; T4Q = T4O - T4P; T4T = FNMS(KP1_546020906, T4S, KP1_268786568 * T4R); R0[WS(rs, 41)] = T4Q - T4T; R0[WS(rs, 9)] = T4Q + T4T; } } { E T6y, T7e, T7a, T7i, T6J, T7f, T71, T7h; { E T6s, T6x, T74, T79; T6s = T6q - T6r; T6x = KP1_414213562 * (T6t - T6w); T6y = T6s + T6x; T7e = T6s - T6x; T74 = KP707106781 * (T72 + T73); T79 = T75 + T78; T7a = T74 + T79; T7i = T79 - T74; } { E T6D, T6I, T6P, T70; T6D = T6z + T6C; T6I = T6E + T6H; T6J = FNMS(KP765366864, T6I, KP1_847759065 * T6D); T7f = FMA(KP765366864, T6D, KP1_847759065 * T6I); T6P = T6L + T6O; T70 = KP707106781 * (T6U + T6Z); T71 = T6P + T70; T7h = T6P - T70; } { E T6K, T7b, T7k, T7l; T6K = T6y + T6J; T7b = FNMS(KP390180644, T7a, KP1_961570560 * T71); R0[WS(rs, 34)] = T6K - T7b; R0[WS(rs, 2)] = T6K + T7b; T7k = T7e + T7f; T7l = FMA(KP1_662939224, T7h, KP1_111140466 * T7i); R0[WS(rs, 26)] = T7k - T7l; R0[WS(rs, 58)] = T7k + T7l; } { E T7c, T7d, T7g, T7j; T7c = T6y - T6J; T7d = FMA(KP390180644, T71, KP1_961570560 * T7a); R0[WS(rs, 18)] = T7c - T7d; R0[WS(rs, 50)] = T7c + T7d; T7g = T7e - T7f; T7j = FNMS(KP1_662939224, T7i, KP1_111140466 * T7h); R0[WS(rs, 42)] = T7g - T7j; R0[WS(rs, 10)] = T7g + T7j; } } { E T4Y, T5c, T58, T5g, T51, T5d, T55, T5f; { E T4W, T4X, T56, T57; T4W = T2a - T2j; T4X = FMA(KP765366864, T2t, KP1_847759065 * T2C); T4Y = T4W - T4X; T5c = T4W + T4X; T56 = T48 - T4r; T57 = T4I - T4z; T58 = T56 + T57; T5g = T57 - T56; } { E T4Z, T50, T53, T54; T4Z = T2N - T36; T50 = T3j - T3a; T51 = FNMS(KP1_662939224, T50, KP1_111140466 * T4Z); T5d = FMA(KP1_662939224, T4Z, KP1_111140466 * T50); T53 = T3v - T3O; T54 = T4v - T4u; T55 = T53 + T54; T5f = T53 - T54; } { E T52, T59, T5i, T5j; T52 = T4Y + T51; T59 = FNMS(KP942793473, T58, KP1_763842528 * T55); R0[WS(rs, 37)] = T52 - T59; R0[WS(rs, 5)] = T52 + T59; T5i = T5c + T5d; T5j = FMA(KP1_913880671, T5f, KP580569354 * T5g); R0[WS(rs, 29)] = T5i - T5j; R0[WS(rs, 61)] = T5i + T5j; } { E T5a, T5b, T5e, T5h; T5a = T4Y - T51; T5b = FMA(KP942793473, T55, KP1_763842528 * T58); R0[WS(rs, 21)] = T5a - T5b; R0[WS(rs, 53)] = T5a + T5b; T5e = T5c - T5d; T5h = FNMS(KP1_913880671, T5g, KP580569354 * T5f); R0[WS(rs, 45)] = T5e - T5h; R0[WS(rs, 13)] = T5e + T5h; } } { E T7o, T7C, T7y, T7G, T7r, T7D, T7v, T7F; { E T7m, T7n, T7w, T7x; T7m = T6q + T6r; T7n = KP1_414213562 * (T6t + T6w); T7o = T7m - T7n; T7C = T7m + T7n; T7w = KP707106781 * (T6U - T6Z); T7x = T78 - T75; T7y = T7w + T7x; T7G = T7x - T7w; } { E T7p, T7q, T7t, T7u; T7p = T6z - T6C; T7q = T6H - T6E; T7r = FNMS(KP1_847759065, T7q, KP765366864 * T7p); T7D = FMA(KP1_847759065, T7p, KP765366864 * T7q); T7t = T6L - T6O; T7u = KP707106781 * (T73 - T72); T7v = T7t + T7u; T7F = T7t - T7u; } { E T7s, T7z, T7I, T7J; T7s = T7o + T7r; T7z = FNMS(KP1_111140466, T7y, KP1_662939224 * T7v); R0[WS(rs, 38)] = T7s - T7z; R0[WS(rs, 6)] = T7s + T7z; T7I = T7C + T7D; T7J = FMA(KP1_961570560, T7F, KP390180644 * T7G); R0[WS(rs, 30)] = T7I - T7J; R0[WS(rs, 62)] = T7I + T7J; } { E T7A, T7B, T7E, T7H; T7A = T7o - T7r; T7B = FMA(KP1_111140466, T7v, KP1_662939224 * T7y); R0[WS(rs, 22)] = T7A - T7B; R0[WS(rs, 54)] = T7A + T7B; T7E = T7C - T7D; T7H = FNMS(KP1_961570560, T7G, KP390180644 * T7F); R0[WS(rs, 46)] = T7E - T7H; R0[WS(rs, 14)] = T7E + T7H; } } { E T5q, T5U, T5Q, T5Y, T5x, T5V, T5J, T5X; { E T5m, T5p, T5M, T5P; T5m = T5k - T5l; T5p = FNMS(KP1_847759065, T5o, KP765366864 * T5n); T5q = T5m + T5p; T5U = T5m - T5p; T5M = T5K + T5L; T5P = T5N + T5O; T5Q = T5M + T5P; T5Y = T5P - T5M; } { E T5t, T5w, T5B, T5I; T5t = T5r + T5s; T5w = T5u + T5v; T5x = FNMS(KP1_111140466, T5w, KP1_662939224 * T5t); T5V = FMA(KP1_111140466, T5t, KP1_662939224 * T5w); T5B = T5z + T5A; T5I = T5E + T5H; T5J = T5B + T5I; T5X = T5B - T5I; } { E T5y, T5R, T60, T61; T5y = T5q + T5x; T5R = FNMS(KP580569354, T5Q, KP1_913880671 * T5J); R0[WS(rs, 35)] = T5y - T5R; R0[WS(rs, 3)] = T5y + T5R; T60 = T5U + T5V; T61 = FMA(KP1_763842528, T5X, KP942793473 * T5Y); R0[WS(rs, 27)] = T60 - T61; R0[WS(rs, 59)] = T60 + T61; } { E T5S, T5T, T5W, T5Z; T5S = T5q - T5x; T5T = FMA(KP580569354, T5J, KP1_913880671 * T5Q); R0[WS(rs, 19)] = T5S - T5T; R0[WS(rs, 51)] = T5S + T5T; T5W = T5U - T5V; T5Z = FNMS(KP1_763842528, T5Y, KP942793473 * T5X); R0[WS(rs, 43)] = T5W - T5Z; R0[WS(rs, 11)] = T5W + T5Z; } } { E T64, T6i, T6e, T6m, T67, T6j, T6b, T6l; { E T62, T63, T6c, T6d; T62 = T5k + T5l; T63 = FMA(KP1_847759065, T5n, KP765366864 * T5o); T64 = T62 - T63; T6i = T62 + T63; T6c = T5E - T5H; T6d = T5O - T5N; T6e = T6c + T6d; T6m = T6d - T6c; } { E T65, T66, T69, T6a; T65 = T5r - T5s; T66 = T5v - T5u; T67 = FNMS(KP1_961570560, T66, KP390180644 * T65); T6j = FMA(KP1_961570560, T65, KP390180644 * T66); T69 = T5z - T5A; T6a = T5L - T5K; T6b = T69 + T6a; T6l = T69 - T6a; } { E T68, T6f, T6o, T6p; T68 = T64 + T67; T6f = FNMS(KP1_268786568, T6e, KP1_546020906 * T6b); R0[WS(rs, 39)] = T68 - T6f; R0[WS(rs, 7)] = T68 + T6f; T6o = T6i + T6j; T6p = FMA(KP1_990369453, T6l, KP196034280 * T6m); R0[WS(rs, 31)] = T6o - T6p; R0[WS(rs, 63)] = T6o + T6p; } { E T6g, T6h, T6k, T6n; T6g = T64 - T67; T6h = FMA(KP1_268786568, T6b, KP1_546020906 * T6e); R0[WS(rs, 23)] = T6g - T6h; R0[WS(rs, 55)] = T6g + T6h; T6k = T6i - T6j; T6n = FNMS(KP1_990369453, T6m, KP196034280 * T6l); R0[WS(rs, 47)] = T6k - T6n; R0[WS(rs, 15)] = T6k + T6n; } } { E T8Z, Tb1, T9C, Tb2, Tbe, Tbq, Tbb, Tbp, TaX, Tbs, Tb5, Tbi, TaI, Tbt, Tb4; E Tbl; { E T8F, T8Y, Tb9, Tba; T8F = T8x + T8E; T8Y = FNMS(KP390180644, T8X, KP1_961570560 * T8Q); T8Z = T8F + T8Y; Tb1 = T8F - T8Y; { E T9q, T9B, Tbc, Tbd; T9q = T9a + T9p; T9B = T9t + T9A; T9C = FNMS(KP196034280, T9B, KP1_990369453 * T9q); Tb2 = FMA(KP196034280, T9q, KP1_990369453 * T9B); Tbc = T9a - T9p; Tbd = T9A - T9t; Tbe = FNMS(KP1_546020906, Tbd, KP1_268786568 * Tbc); Tbq = FMA(KP1_546020906, Tbc, KP1_268786568 * Tbd); } Tb9 = T8x - T8E; Tba = FMA(KP390180644, T8Q, KP1_961570560 * T8X); Tbb = Tb9 - Tba; Tbp = Tb9 + Tba; { E TaW, Tbg, TaL, Tbh, TaJ, TaK; TaW = TaO + TaV; Tbg = T9O - Ta3; TaJ = FMA(KP195090322, Taf, KP980785280 * Tam); TaK = FNMS(KP195090322, Tay, KP980785280 * TaF); TaL = TaJ + TaK; Tbh = TaK - TaJ; TaX = TaL + TaW; Tbs = Tbg - Tbh; Tb5 = TaW - TaL; Tbi = Tbg + Tbh; } { E Ta4, Tbk, TaH, Tbj, Tan, TaG; Ta4 = T9O + Ta3; Tbk = TaV - TaO; Tan = FNMS(KP195090322, Tam, KP980785280 * Taf); TaG = FMA(KP980785280, Tay, KP195090322 * TaF); TaH = Tan + TaG; Tbj = Tan - TaG; TaI = Ta4 + TaH; Tbt = Tbk - Tbj; Tb4 = Ta4 - TaH; Tbl = Tbj + Tbk; } } { E T9D, TaY, Tbr, Tbu; T9D = T8Z + T9C; TaY = FNMS(KP098135348, TaX, KP1_997590912 * TaI); R1[WS(rs, 32)] = T9D - TaY; R1[0] = T9D + TaY; Tbr = Tbp - Tbq; Tbu = FNMS(KP1_883088130, Tbt, KP673779706 * Tbs); R1[WS(rs, 44)] = Tbr - Tbu; R1[WS(rs, 12)] = Tbr + Tbu; } { E Tbv, Tbw, TaZ, Tb0; Tbv = Tbp + Tbq; Tbw = FMA(KP1_883088130, Tbs, KP673779706 * Tbt); R1[WS(rs, 28)] = Tbv - Tbw; R1[WS(rs, 60)] = Tbv + Tbw; TaZ = T8Z - T9C; Tb0 = FMA(KP098135348, TaI, KP1_997590912 * TaX); R1[WS(rs, 16)] = TaZ - Tb0; R1[WS(rs, 48)] = TaZ + Tb0; } { E Tb3, Tb6, Tbf, Tbm; Tb3 = Tb1 - Tb2; Tb6 = FNMS(KP1_481902250, Tb5, KP1_343117909 * Tb4); R1[WS(rs, 40)] = Tb3 - Tb6; R1[WS(rs, 8)] = Tb3 + Tb6; Tbf = Tbb + Tbe; Tbm = FNMS(KP855110186, Tbl, KP1_807978586 * Tbi); R1[WS(rs, 36)] = Tbf - Tbm; R1[WS(rs, 4)] = Tbf + Tbm; } { E Tbn, Tbo, Tb7, Tb8; Tbn = Tbb - Tbe; Tbo = FMA(KP855110186, Tbi, KP1_807978586 * Tbl); R1[WS(rs, 20)] = Tbn - Tbo; R1[WS(rs, 52)] = Tbn + Tbo; Tb7 = Tb1 + Tb2; Tb8 = FMA(KP1_481902250, Tb4, KP1_343117909 * Tb5); R1[WS(rs, 24)] = Tb7 - Tb8; R1[WS(rs, 56)] = Tb7 + Tb8; } } { E TcR, TdR, Tda, TdS, Te4, Teg, Te1, Tef, TdN, Tei, TdV, Te8, TdC, Tej, TdU; E Teb; { E TcJ, TcQ, TdZ, Te0; TcJ = TcF + TcI; TcQ = FNMS(KP1_111140466, TcP, KP1_662939224 * TcM); TcR = TcJ + TcQ; TdR = TcJ - TcQ; { E Td2, Td9, Te2, Te3; Td2 = TcU + Td1; Td9 = Td5 + Td8; Tda = FNMS(KP580569354, Td9, KP1_913880671 * Td2); TdS = FMA(KP580569354, Td2, KP1_913880671 * Td9); Te2 = TcU - Td1; Te3 = Td8 - Td5; Te4 = FNMS(KP1_763842528, Te3, KP942793473 * Te2); Teg = FMA(KP1_763842528, Te2, KP942793473 * Te3); } TdZ = TcF - TcI; Te0 = FMA(KP1_111140466, TcM, KP1_662939224 * TcP); Te1 = TdZ - Te0; Tef = TdZ + Te0; { E TdM, Te6, TdF, Te7, TdD, TdE; TdM = TdI + TdL; Te6 = Tde - Tdl; TdD = FMA(KP555570233, Tdp, KP831469612 * Tds); TdE = FNMS(KP555570233, Tdw, KP831469612 * Tdz); TdF = TdD + TdE; Te7 = TdE - TdD; TdN = TdF + TdM; Tei = Te6 - Te7; TdV = TdM - TdF; Te8 = Te6 + Te7; } { E Tdm, Tea, TdB, Te9, Tdt, TdA; Tdm = Tde + Tdl; Tea = TdL - TdI; Tdt = FNMS(KP555570233, Tds, KP831469612 * Tdp); TdA = FMA(KP831469612, Tdw, KP555570233 * Tdz); TdB = Tdt + TdA; Te9 = Tdt - TdA; TdC = Tdm + TdB; Tej = Tea - Te9; TdU = Tdm - TdB; Teb = Te9 + Tea; } } { E Tdb, TdO, Teh, Tek; Tdb = TcR + Tda; TdO = FNMS(KP293460948, TdN, KP1_978353019 * TdC); R1[WS(rs, 33)] = Tdb - TdO; R1[WS(rs, 1)] = Tdb + TdO; Teh = Tef - Teg; Tek = FNMS(KP1_940062506, Tej, KP485960359 * Tei); R1[WS(rs, 45)] = Teh - Tek; R1[WS(rs, 13)] = Teh + Tek; } { E Tel, Tem, TdP, TdQ; Tel = Tef + Teg; Tem = FMA(KP1_940062506, Tei, KP485960359 * Tej); R1[WS(rs, 29)] = Tel - Tem; R1[WS(rs, 61)] = Tel + Tem; TdP = TcR - Tda; TdQ = FMA(KP293460948, TdC, KP1_978353019 * TdN); R1[WS(rs, 17)] = TdP - TdQ; R1[WS(rs, 49)] = TdP + TdQ; } { E TdT, TdW, Te5, Tec; TdT = TdR - TdS; TdW = FNMS(KP1_606415062, TdV, KP1_191398608 * TdU); R1[WS(rs, 41)] = TdT - TdW; R1[WS(rs, 9)] = TdT + TdW; Te5 = Te1 + Te4; Tec = FNMS(KP1_028205488, Teb, KP1_715457220 * Te8); R1[WS(rs, 37)] = Te5 - Tec; R1[WS(rs, 5)] = Te5 + Tec; } { E Ted, Tee, TdX, TdY; Ted = Te1 - Te4; Tee = FMA(KP1_028205488, Te8, KP1_715457220 * Teb); R1[WS(rs, 21)] = Ted - Tee; R1[WS(rs, 53)] = Ted + Tee; TdX = TdR + TdS; TdY = FMA(KP1_606415062, TdU, KP1_191398608 * TdV); R1[WS(rs, 25)] = TdX - TdY; R1[WS(rs, 57)] = TdX + TdY; } } { E TbD, Tc7, TbK, Tc8, Tck, Tcw, Tch, Tcv, Tc3, Tcy, Tcb, Tco, TbW, Tcz, Tca; E Tcr; { E Tbz, TbC, Tcf, Tcg; Tbz = Tbx - Tby; TbC = FNMS(KP1_662939224, TbB, KP1_111140466 * TbA); TbD = Tbz + TbC; Tc7 = Tbz - TbC; { E TbG, TbJ, Tci, Tcj; TbG = TbE + TbF; TbJ = TbH + TbI; TbK = FNMS(KP942793473, TbJ, KP1_763842528 * TbG); Tc8 = FMA(KP942793473, TbG, KP1_763842528 * TbJ); Tci = TbE - TbF; Tcj = TbI - TbH; Tck = FNMS(KP1_913880671, Tcj, KP580569354 * Tci); Tcw = FMA(KP1_913880671, Tci, KP580569354 * Tcj); } Tcf = Tbx + Tby; Tcg = FMA(KP1_662939224, TbA, KP1_111140466 * TbB); Tch = Tcf - Tcg; Tcv = Tcf + Tcg; { E Tc2, Tcm, TbZ, Tcn, TbX, TbY; Tc2 = Tc0 + Tc1; Tcm = TbM - TbN; TbX = FMA(KP831469612, TbP, KP555570233 * TbQ); TbY = FNMS(KP831469612, TbS, KP555570233 * TbT); TbZ = TbX + TbY; Tcn = TbY - TbX; Tc3 = TbZ + Tc2; Tcy = Tcm - Tcn; Tcb = Tc2 - TbZ; Tco = Tcm + Tcn; } { E TbO, Tcq, TbV, Tcp, TbR, TbU; TbO = TbM + TbN; Tcq = Tc1 - Tc0; TbR = FNMS(KP831469612, TbQ, KP555570233 * TbP); TbU = FMA(KP555570233, TbS, KP831469612 * TbT); TbV = TbR + TbU; Tcp = TbR - TbU; TbW = TbO + TbV; Tcz = Tcq - Tcp; Tca = TbO - TbV; Tcr = Tcp + Tcq; } } { E TbL, Tc4, Tcx, TcA; TbL = TbD + TbK; Tc4 = FNMS(KP485960359, Tc3, KP1_940062506 * TbW); R1[WS(rs, 34)] = TbL - Tc4; R1[WS(rs, 2)] = TbL + Tc4; Tcx = Tcv - Tcw; TcA = FNMS(KP1_978353019, Tcz, KP293460948 * Tcy); R1[WS(rs, 46)] = Tcx - TcA; R1[WS(rs, 14)] = Tcx + TcA; } { E TcB, TcC, Tc5, Tc6; TcB = Tcv + Tcw; TcC = FMA(KP1_978353019, Tcy, KP293460948 * Tcz); R1[WS(rs, 30)] = TcB - TcC; R1[WS(rs, 62)] = TcB + TcC; Tc5 = TbD - TbK; Tc6 = FMA(KP485960359, TbW, KP1_940062506 * Tc3); R1[WS(rs, 18)] = Tc5 - Tc6; R1[WS(rs, 50)] = Tc5 + Tc6; } { E Tc9, Tcc, Tcl, Tcs; Tc9 = Tc7 - Tc8; Tcc = FNMS(KP1_715457220, Tcb, KP1_028205488 * Tca); R1[WS(rs, 42)] = Tc9 - Tcc; R1[WS(rs, 10)] = Tc9 + Tcc; Tcl = Tch + Tck; Tcs = FNMS(KP1_191398608, Tcr, KP1_606415062 * Tco); R1[WS(rs, 38)] = Tcl - Tcs; R1[WS(rs, 6)] = Tcl + Tcs; } { E Tct, Tcu, Tcd, Tce; Tct = Tch - Tck; Tcu = FMA(KP1_191398608, Tco, KP1_606415062 * Tcr); R1[WS(rs, 22)] = Tct - Tcu; R1[WS(rs, 54)] = Tct + Tcu; Tcd = Tc7 + Tc8; Tce = FMA(KP1_715457220, Tca, KP1_028205488 * Tcb); R1[WS(rs, 26)] = Tcd - Tce; R1[WS(rs, 58)] = Tcd + Tce; } } { E Tet, TeX, TeA, TeY, Tfa, Tfm, Tf7, Tfl, TeT, Tfo, Tf1, Tfe, TeM, Tfp, Tf0; E Tfh; { E Tep, Tes, Tf5, Tf6; Tep = Ten - Teo; Tes = FNMS(KP1_961570560, Ter, KP390180644 * Teq); Tet = Tep + Tes; TeX = Tep - Tes; { E Tew, Tez, Tf8, Tf9; Tew = Teu - Tev; Tez = Tex + Tey; TeA = FNMS(KP1_268786568, Tez, KP1_546020906 * Tew); TeY = FMA(KP1_268786568, Tew, KP1_546020906 * Tez); Tf8 = Teu + Tev; Tf9 = Tey - Tex; Tfa = FNMS(KP1_990369453, Tf9, KP196034280 * Tf8); Tfm = FMA(KP1_990369453, Tf8, KP196034280 * Tf9); } Tf5 = Ten + Teo; Tf6 = FMA(KP1_961570560, Teq, KP390180644 * Ter); Tf7 = Tf5 - Tf6; Tfl = Tf5 + Tf6; { E TeS, Tfc, TeP, Tfd, TeN, TeO; TeS = TeQ + TeR; Tfc = TeC + TeD; TeN = FMA(KP980785280, TeF, KP195090322 * TeG); TeO = FMA(KP980785280, TeI, KP195090322 * TeJ); TeP = TeN - TeO; Tfd = TeN + TeO; TeT = TeP + TeS; Tfo = Tfc + Tfd; Tf1 = TeS - TeP; Tfe = Tfc - Tfd; } { E TeE, Tfg, TeL, Tff, TeH, TeK; TeE = TeC - TeD; Tfg = TeR - TeQ; TeH = FNMS(KP980785280, TeG, KP195090322 * TeF); TeK = FNMS(KP980785280, TeJ, KP195090322 * TeI); TeL = TeH + TeK; Tff = TeH - TeK; TeM = TeE + TeL; Tfp = Tfg - Tff; Tf0 = TeE - TeL; Tfh = Tff + Tfg; } } { E TeB, TeU, Tfn, Tfq; TeB = Tet + TeA; TeU = FNMS(KP673779706, TeT, KP1_883088130 * TeM); R1[WS(rs, 35)] = TeB - TeU; R1[WS(rs, 3)] = TeB + TeU; Tfn = Tfl - Tfm; Tfq = FNMS(KP1_997590912, Tfp, KP098135348 * Tfo); R1[WS(rs, 47)] = Tfn - Tfq; R1[WS(rs, 15)] = Tfn + Tfq; } { E Tfr, Tfs, TeV, TeW; Tfr = Tfl + Tfm; Tfs = FMA(KP1_997590912, Tfo, KP098135348 * Tfp); R1[WS(rs, 31)] = Tfr - Tfs; R1[WS(rs, 63)] = Tfr + Tfs; TeV = Tet - TeA; TeW = FMA(KP673779706, TeM, KP1_883088130 * TeT); R1[WS(rs, 19)] = TeV - TeW; R1[WS(rs, 51)] = TeV + TeW; } { E TeZ, Tf2, Tfb, Tfi; TeZ = TeX - TeY; Tf2 = FNMS(KP1_807978586, Tf1, KP855110186 * Tf0); R1[WS(rs, 43)] = TeZ - Tf2; R1[WS(rs, 11)] = TeZ + Tf2; Tfb = Tf7 + Tfa; Tfi = FNMS(KP1_343117909, Tfh, KP1_481902250 * Tfe); R1[WS(rs, 39)] = Tfb - Tfi; R1[WS(rs, 7)] = Tfb + Tfi; } { E Tfj, Tfk, Tf3, Tf4; Tfj = Tf7 - Tfa; Tfk = FMA(KP1_343117909, Tfe, KP1_481902250 * Tfh); R1[WS(rs, 23)] = Tfj - Tfk; R1[WS(rs, 55)] = Tfj + Tfk; Tf3 = TeX + TeY; Tf4 = FMA(KP1_807978586, Tf0, KP855110186 * Tf1); R1[WS(rs, 27)] = Tf3 - Tf4; R1[WS(rs, 59)] = Tf3 + Tf4; } } } } } static const kr2c_desc desc = { 128, "r2cb_128", {812, 198, 144, 0}, &GENUS }; void X(codelet_r2cb_128) (planner *p) { X(kr2c_register) (p, r2cb_128, &desc); } #endif /* HAVE_FMA */ fftw-3.3.3/rdft/scalar/r2cb/r2cb_20.c0000644000175000017500000002572312054410746013754 00000000000000/* * Copyright (c) 2003, 2007-11 Matteo Frigo * Copyright (c) 2003, 2007-11 Massachusetts Institute of Technology * * This program is free software; you can redistribute it and/or modify * it under the terms of the GNU General Public License as published by * the Free Software Foundation; either version 2 of the License, or * (at your option) any later version. * * This program is distributed in the hope that it will be useful, * but WITHOUT ANY WARRANTY; without even the implied warranty of * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the * GNU General Public License for more details. * * You should have received a copy of the GNU General Public License * along with this program; if not, write to the Free Software * Foundation, Inc., 51 Franklin Street, Fifth Floor, Boston, MA 02110-1301 USA * */ /* This file was automatically generated --- DO NOT EDIT */ /* Generated on Sun Nov 25 07:41:09 EST 2012 */ #include "codelet-rdft.h" #ifdef HAVE_FMA /* Generated by: ../../../genfft/gen_r2cb.native -fma -reorder-insns -schedule-for-pipeline -compact -variables 4 -pipeline-latency 4 -sign 1 -n 20 -name r2cb_20 -include r2cb.h */ /* * This function contains 86 FP additions, 44 FP multiplications, * (or, 42 additions, 0 multiplications, 44 fused multiply/add), * 69 stack variables, 5 constants, and 40 memory accesses */ #include "r2cb.h" static void r2cb_20(R *R0, R *R1, R *Cr, R *Ci, stride rs, stride csr, stride csi, INT v, INT ivs, INT ovs) { DK(KP1_902113032, +1.902113032590307144232878666758764286811397268); DK(KP1_118033988, +1.118033988749894848204586834365638117720309180); DK(KP500000000, +0.500000000000000000000000000000000000000000000); DK(KP618033988, +0.618033988749894848204586834365638117720309180); DK(KP2_000000000, +2.000000000000000000000000000000000000000000000); { INT i; for (i = v; i > 0; i = i - 1, R0 = R0 + ovs, R1 = R1 + ovs, Cr = Cr + ivs, Ci = Ci + ivs, MAKE_VOLATILE_STRIDE(80, rs), MAKE_VOLATILE_STRIDE(80, csr), MAKE_VOLATILE_STRIDE(80, csi)) { E TY, T1o, T1m, T14, T12, TX, T1n, T1j, TZ, T13; { E Tr, TD, Tl, T5, T1a, T1l, T1d, T1k, TT, T10, TO, T11, TE, TF, Tk; E TI, TC, T1i, To, TG, T16; { E T4, Tq, T1, T2; T4 = Cr[WS(csr, 5)]; Tq = Ci[WS(csi, 5)]; T1 = Cr[0]; T2 = Cr[WS(csr, 10)]; { E Ts, T8, T19, TR, T18, Tb, TS, Tv, Tx, Tf, Ty, T1c, TM, T1b, Ti; E Tz, Tt, Tu, TN, TA; { E TP, TQ, T9, Ta; { E T6, T7, Tp, T3; T6 = Cr[WS(csr, 4)]; T7 = Cr[WS(csr, 6)]; TP = Ci[WS(csi, 4)]; Tp = T1 - T2; T3 = T1 + T2; Ts = T6 - T7; T8 = T6 + T7; Tr = FMA(KP2_000000000, Tq, Tp); TD = FNMS(KP2_000000000, Tq, Tp); Tl = FMA(KP2_000000000, T4, T3); T5 = FNMS(KP2_000000000, T4, T3); TQ = Ci[WS(csi, 6)]; } T9 = Cr[WS(csr, 9)]; Ta = Cr[WS(csr, 1)]; Tt = Ci[WS(csi, 9)]; T19 = TP + TQ; TR = TP - TQ; T18 = T9 - Ta; Tb = T9 + Ta; Tu = Ci[WS(csi, 1)]; } { E TK, TL, Td, Te, Tg, Th; Td = Cr[WS(csr, 8)]; Te = Cr[WS(csr, 2)]; TK = Ci[WS(csi, 8)]; TS = Tt - Tu; Tv = Tt + Tu; Tx = Td - Te; Tf = Td + Te; TL = Ci[WS(csi, 2)]; Tg = Cr[WS(csr, 7)]; Th = Cr[WS(csr, 3)]; Ty = Ci[WS(csi, 7)]; T1c = TK + TL; TM = TK - TL; T1b = Tg - Th; Ti = Tg + Th; Tz = Ci[WS(csi, 3)]; } T1a = T18 + T19; T1l = T19 - T18; T1d = T1b + T1c; T1k = T1c - T1b; TT = TR - TS; T10 = TS + TR; TN = Tz - Ty; TA = Ty + Tz; TO = TM - TN; T11 = TN + TM; { E Tm, Tc, Tj, Tn, Tw, TB; Tm = T8 + Tb; Tc = T8 - Tb; Tj = Tf - Ti; Tn = Tf + Ti; TE = Ts - Tv; Tw = Ts + Tv; TB = Tx - TA; TF = Tx + TA; Tk = Tc + Tj; TI = Tc - Tj; TC = Tw + TB; T1i = Tw - TB; TY = Tm - Tn; To = Tm + Tn; } } } R0[WS(rs, 5)] = FMA(KP2_000000000, Tk, T5); R1[WS(rs, 7)] = FMA(KP2_000000000, TC, Tr); TG = TE + TF; T16 = TE - TF; R0[0] = FMA(KP2_000000000, To, Tl); { E TU, TW, T1g, T1e, T15, TV, TJ, TH, T1h, T1f, T17; TU = FNMS(KP618033988, TT, TO); TW = FMA(KP618033988, TO, TT); R1[WS(rs, 2)] = FMA(KP2_000000000, TG, TD); TH = FNMS(KP500000000, Tk, T5); T1g = FNMS(KP618033988, T1a, T1d); T1e = FMA(KP618033988, T1d, T1a); T15 = FNMS(KP500000000, TG, TD); TV = FMA(KP1_118033988, TI, TH); TJ = FNMS(KP1_118033988, TI, TH); T1o = FMA(KP618033988, T1k, T1l); T1m = FNMS(KP618033988, T1l, T1k); R0[WS(rs, 3)] = FNMS(KP1_902113032, TW, TV); R0[WS(rs, 7)] = FMA(KP1_902113032, TW, TV); R0[WS(rs, 1)] = FMA(KP1_902113032, TU, TJ); R0[WS(rs, 9)] = FNMS(KP1_902113032, TU, TJ); T1f = FNMS(KP1_118033988, T16, T15); T17 = FMA(KP1_118033988, T16, T15); T1h = FNMS(KP500000000, TC, Tr); R1[WS(rs, 6)] = FNMS(KP1_902113032, T1g, T1f); R1[WS(rs, 8)] = FMA(KP1_902113032, T1g, T1f); R1[WS(rs, 4)] = FMA(KP1_902113032, T1e, T17); R1[0] = FNMS(KP1_902113032, T1e, T17); T14 = FNMS(KP618033988, T10, T11); T12 = FMA(KP618033988, T11, T10); TX = FNMS(KP500000000, To, Tl); T1n = FMA(KP1_118033988, T1i, T1h); T1j = FNMS(KP1_118033988, T1i, T1h); } } R1[WS(rs, 5)] = FNMS(KP1_902113032, T1o, T1n); R1[WS(rs, 9)] = FMA(KP1_902113032, T1o, T1n); R1[WS(rs, 3)] = FMA(KP1_902113032, T1m, T1j); R1[WS(rs, 1)] = FNMS(KP1_902113032, T1m, T1j); TZ = FMA(KP1_118033988, TY, TX); T13 = FNMS(KP1_118033988, TY, TX); R0[WS(rs, 4)] = FNMS(KP1_902113032, T14, T13); R0[WS(rs, 6)] = FMA(KP1_902113032, T14, T13); R0[WS(rs, 2)] = FMA(KP1_902113032, T12, TZ); R0[WS(rs, 8)] = FNMS(KP1_902113032, T12, TZ); } } } static const kr2c_desc desc = { 20, "r2cb_20", {42, 0, 44, 0}, &GENUS }; void X(codelet_r2cb_20) (planner *p) { X(kr2c_register) (p, r2cb_20, &desc); } #else /* HAVE_FMA */ /* Generated by: ../../../genfft/gen_r2cb.native -compact -variables 4 -pipeline-latency 4 -sign 1 -n 20 -name r2cb_20 -include r2cb.h */ /* * This function contains 86 FP additions, 30 FP multiplications, * (or, 70 additions, 14 multiplications, 16 fused multiply/add), * 50 stack variables, 5 constants, and 40 memory accesses */ #include "r2cb.h" static void r2cb_20(R *R0, R *R1, R *Cr, R *Ci, stride rs, stride csr, stride csi, INT v, INT ivs, INT ovs) { DK(KP1_118033988, +1.118033988749894848204586834365638117720309180); DK(KP500000000, +0.500000000000000000000000000000000000000000000); DK(KP1_902113032, +1.902113032590307144232878666758764286811397268); DK(KP1_175570504, +1.175570504584946258337411909278145537195304875); DK(KP2_000000000, +2.000000000000000000000000000000000000000000000); { INT i; for (i = v; i > 0; i = i - 1, R0 = R0 + ovs, R1 = R1 + ovs, Cr = Cr + ivs, Ci = Ci + ivs, MAKE_VOLATILE_STRIDE(80, rs), MAKE_VOLATILE_STRIDE(80, csr), MAKE_VOLATILE_STRIDE(80, csi)) { E T6, TF, Tm, Tt, TQ, T1n, T1f, T12, T1m, TV, T13, T1c, Td, Tk, Tl; E Ty, TD, TE, Tn, To, Tp, TG, TH, TI; { E T5, Ts, T3, Tq; { E T4, Tr, T1, T2; T4 = Cr[WS(csr, 5)]; T5 = KP2_000000000 * T4; Tr = Ci[WS(csi, 5)]; Ts = KP2_000000000 * Tr; T1 = Cr[0]; T2 = Cr[WS(csr, 10)]; T3 = T1 + T2; Tq = T1 - T2; } T6 = T3 - T5; TF = Tq - Ts; Tm = T3 + T5; Tt = Tq + Ts; } { E T9, Tu, TO, T1b, Tc, T1a, Tx, TP, Tg, Tz, TT, T1e, Tj, T1d, TC; E TU; { E T7, T8, TM, TN; T7 = Cr[WS(csr, 4)]; T8 = Cr[WS(csr, 6)]; T9 = T7 + T8; Tu = T7 - T8; TM = Ci[WS(csi, 4)]; TN = Ci[WS(csi, 6)]; TO = TM - TN; T1b = TM + TN; } { E Ta, Tb, Tv, Tw; Ta = Cr[WS(csr, 9)]; Tb = Cr[WS(csr, 1)]; Tc = Ta + Tb; T1a = Ta - Tb; Tv = Ci[WS(csi, 9)]; Tw = Ci[WS(csi, 1)]; Tx = Tv + Tw; TP = Tv - Tw; } { E Te, Tf, TR, TS; Te = Cr[WS(csr, 8)]; Tf = Cr[WS(csr, 2)]; Tg = Te + Tf; Tz = Te - Tf; TR = Ci[WS(csi, 8)]; TS = Ci[WS(csi, 2)]; TT = TR - TS; T1e = TR + TS; } { E Th, Ti, TA, TB; Th = Cr[WS(csr, 7)]; Ti = Cr[WS(csr, 3)]; Tj = Th + Ti; T1d = Th - Ti; TA = Ci[WS(csi, 7)]; TB = Ci[WS(csi, 3)]; TC = TA + TB; TU = TB - TA; } TQ = TO - TP; T1n = T1e - T1d; T1f = T1d + T1e; T12 = TP + TO; T1m = T1b - T1a; TV = TT - TU; T13 = TU + TT; T1c = T1a + T1b; Td = T9 - Tc; Tk = Tg - Tj; Tl = Td + Tk; Ty = Tu + Tx; TD = Tz - TC; TE = Ty + TD; Tn = T9 + Tc; To = Tg + Tj; Tp = Tn + To; TG = Tu - Tx; TH = Tz + TC; TI = TG + TH; } R0[WS(rs, 5)] = FMA(KP2_000000000, Tl, T6); R1[WS(rs, 7)] = FMA(KP2_000000000, TE, Tt); R1[WS(rs, 2)] = FMA(KP2_000000000, TI, TF); R0[0] = FMA(KP2_000000000, Tp, Tm); { E TW, TY, TL, TX, TJ, TK; TW = FNMS(KP1_902113032, TV, KP1_175570504 * TQ); TY = FMA(KP1_902113032, TQ, KP1_175570504 * TV); TJ = FNMS(KP500000000, Tl, T6); TK = KP1_118033988 * (Td - Tk); TL = TJ - TK; TX = TK + TJ; R0[WS(rs, 1)] = TL - TW; R0[WS(rs, 7)] = TX + TY; R0[WS(rs, 9)] = TL + TW; R0[WS(rs, 3)] = TX - TY; } { E T1g, T1i, T19, T1h, T17, T18; T1g = FNMS(KP1_902113032, T1f, KP1_175570504 * T1c); T1i = FMA(KP1_902113032, T1c, KP1_175570504 * T1f); T17 = FNMS(KP500000000, TI, TF); T18 = KP1_118033988 * (TG - TH); T19 = T17 - T18; T1h = T18 + T17; R1[WS(rs, 8)] = T19 - T1g; R1[WS(rs, 4)] = T1h + T1i; R1[WS(rs, 6)] = T19 + T1g; R1[0] = T1h - T1i; } { E T1o, T1q, T1l, T1p, T1j, T1k; T1o = FNMS(KP1_902113032, T1n, KP1_175570504 * T1m); T1q = FMA(KP1_902113032, T1m, KP1_175570504 * T1n); T1j = FNMS(KP500000000, TE, Tt); T1k = KP1_118033988 * (Ty - TD); T1l = T1j - T1k; T1p = T1k + T1j; R1[WS(rs, 3)] = T1l - T1o; R1[WS(rs, 9)] = T1p + T1q; R1[WS(rs, 1)] = T1l + T1o; R1[WS(rs, 5)] = T1p - T1q; } { E T14, T16, T11, T15, TZ, T10; T14 = FNMS(KP1_902113032, T13, KP1_175570504 * T12); T16 = FMA(KP1_902113032, T12, KP1_175570504 * T13); TZ = FNMS(KP500000000, Tp, Tm); T10 = KP1_118033988 * (Tn - To); T11 = TZ - T10; T15 = T10 + TZ; R0[WS(rs, 6)] = T11 - T14; R0[WS(rs, 2)] = T15 + T16; R0[WS(rs, 4)] = T11 + T14; R0[WS(rs, 8)] = T15 - T16; } } } } static const kr2c_desc desc = { 20, "r2cb_20", {70, 14, 16, 0}, &GENUS }; void X(codelet_r2cb_20) (planner *p) { X(kr2c_register) (p, r2cb_20, &desc); } #endif /* HAVE_FMA */ fftw-3.3.3/rdft/scalar/r2cb/r2cb_25.c0000644000175000017500000005334612054410751013757 00000000000000/* * Copyright (c) 2003, 2007-11 Matteo Frigo * Copyright (c) 2003, 2007-11 Massachusetts Institute of Technology * * This program is free software; you can redistribute it and/or modify * it under the terms of the GNU General Public License as published by * the Free Software Foundation; either version 2 of the License, or * (at your option) any later version. * * This program is distributed in the hope that it will be useful, * but WITHOUT ANY WARRANTY; without even the implied warranty of * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the * GNU General Public License for more details. * * You should have received a copy of the GNU General Public License * along with this program; if not, write to the Free Software * Foundation, Inc., 51 Franklin Street, Fifth Floor, Boston, MA 02110-1301 USA * */ /* This file was automatically generated --- DO NOT EDIT */ /* Generated on Sun Nov 25 07:41:10 EST 2012 */ #include "codelet-rdft.h" #ifdef HAVE_FMA /* Generated by: ../../../genfft/gen_r2cb.native -fma -reorder-insns -schedule-for-pipeline -compact -variables 4 -pipeline-latency 4 -sign 1 -n 25 -name r2cb_25 -include r2cb.h */ /* * This function contains 152 FP additions, 120 FP multiplications, * (or, 32 additions, 0 multiplications, 120 fused multiply/add), * 115 stack variables, 44 constants, and 50 memory accesses */ #include "r2cb.h" static void r2cb_25(R *R0, R *R1, R *Cr, R *Ci, stride rs, stride csr, stride csi, INT v, INT ivs, INT ovs) { DK(KP979740652, +0.979740652857618686258237536568998933733477632); DK(KP438153340, +0.438153340021931793654057951961031291699532119); DK(KP1_752613360, +1.752613360087727174616231807844125166798128477); DK(KP963507348, +0.963507348203430549974383005744259307057084020); DK(KP1_606007150, +1.606007150877320829666881187140752009270929701); DK(KP1_721083328, +1.721083328735889354196523361841037632825608373); DK(KP1_011627398, +1.011627398597394192215998921771049272931807941); DK(KP595480289, +0.595480289600000014706716770488118292997907308); DK(KP641441904, +0.641441904830606407298806329068862424939687989); DK(KP452413526, +0.452413526233009763856834323966348796985206956); DK(KP1_809654104, +1.809654104932039055427337295865395187940827822); DK(KP933137358, +0.933137358350283770603023973254446451924190884); DK(KP1_666834356, +1.666834356657377354817925100486477686277992119); DK(KP1_842354653, +1.842354653930286640500894870830132058718564461); DK(KP1_082908895, +1.082908895072625554092571180165639018104066379); DK(KP662318342, +0.662318342759882818626911127577439236802190210); DK(KP576710603, +0.576710603632765877371579268136471017090111488); DK(KP484291580, +0.484291580564315559745084187732367906918006201); DK(KP1_937166322, +1.937166322257262238980336750929471627672024806); DK(KP1_898359647, +1.898359647016882523151110931686726543423167685); DK(KP1_386580726, +1.386580726567734802700860150804827247498955921); DK(KP904730450, +0.904730450839922351881287709692877908104763647); DK(KP1_115827804, +1.115827804063668528375399296931134075984874304); DK(KP634619297, +0.634619297544148100711287640319130485732531031); DK(KP470564281, +0.470564281212251493087595091036643380879947982); DK(KP499013364, +0.499013364214135780976168403431725276668452610); DK(KP1_996053456, +1.996053456856543123904673613726901106673810439); DK(KP559154169, +0.559154169276087864842202529084232643714075927); DK(KP683113946, +0.683113946453479238701949862233725244439656928); DK(KP730409924, +0.730409924561256563751459444999838399157094302); DK(KP549754652, +0.549754652192770074288023275540779861653779767); DK(KP256756360, +0.256756360367726783319498520922669048172391148); DK(KP451418159, +0.451418159099103183892477933432151804893354132); DK(KP846146756, +0.846146756728608505452954290121135880883743802); DK(KP1_902113032, +1.902113032590307144232878666758764286811397268); DK(KP062914667, +0.062914667253649757225485955897349402364686947); DK(KP939062505, +0.939062505817492352556001843133229685779824606); DK(KP951056516, +0.951056516295153572116439333379382143405698634); DK(KP559016994, +0.559016994374947424102293417182819058860154590); DK(KP250000000, +0.250000000000000000000000000000000000000000000); DK(KP1_118033988, +1.118033988749894848204586834365638117720309180); DK(KP2_000000000, +2.000000000000000000000000000000000000000000000); DK(KP500000000, +0.500000000000000000000000000000000000000000000); DK(KP618033988, +0.618033988749894848204586834365638117720309180); { INT i; for (i = v; i > 0; i = i - 1, R0 = R0 + ovs, R1 = R1 + ovs, Cr = Cr + ivs, Ci = Ci + ivs, MAKE_VOLATILE_STRIDE(100, rs), MAKE_VOLATILE_STRIDE(100, csr), MAKE_VOLATILE_STRIDE(100, csi)) { E T1H, T24, T22, T1W, T1Y, T1X, T1Z, T23; { E T1G, Tu, T5, T1F, Tr, Te, T2o, T1N, T2a, T1t, TR, T1K, T29, T1u, TG; E TU, TT, Tn, T1d, T1Q, T2p, T1T, T12, T1P, T1a; { E T1, T2, T3, Ts, Tt; Ts = Ci[WS(csi, 5)]; Tt = Ci[WS(csi, 10)]; T1 = Cr[0]; T2 = Cr[WS(csr, 5)]; T3 = Cr[WS(csr, 10)]; T1G = FMS(KP618033988, Ts, Tt); Tu = FMA(KP618033988, Tt, Ts); { E Tx, Tw, T1M, TQ, TM, T1J, TF, TL; { E T6, TH, TO, TP, TB, TI, Td, TJ, TE, T4, Tq, TK; T6 = Cr[WS(csr, 1)]; T4 = T2 + T3; Tq = T2 - T3; TH = Ci[WS(csi, 1)]; { E Ta, T9, Tb, T7, T8, Tp; T7 = Cr[WS(csr, 6)]; T8 = Cr[WS(csr, 4)]; Tp = FNMS(KP500000000, T4, T1); T5 = FMA(KP2_000000000, T4, T1); Ta = Cr[WS(csr, 11)]; TO = T7 - T8; T9 = T7 + T8; T1F = FNMS(KP1_118033988, Tq, Tp); Tr = FMA(KP1_118033988, Tq, Tp); Tb = Cr[WS(csr, 9)]; { E TC, TD, Tz, TA, Tc; Tz = Ci[WS(csi, 6)]; TA = Ci[WS(csi, 4)]; TP = Tb - Ta; Tc = Ta + Tb; TC = Ci[WS(csi, 11)]; TB = Tz + TA; TI = Tz - TA; TD = Ci[WS(csi, 9)]; Td = T9 + Tc; Tx = T9 - Tc; TJ = TC - TD; TE = TC + TD; } } Te = T6 + Td; Tw = FNMS(KP250000000, Td, T6); T1M = FMA(KP618033988, TO, TP); TQ = FNMS(KP618033988, TP, TO); TK = TI + TJ; TM = TI - TJ; T1J = FNMS(KP618033988, TB, TE); TF = FMA(KP618033988, TE, TB); TL = FNMS(KP250000000, TK, TH); T2o = TK + TH; } { E Tf, T14, T1b, T1c, Tm, TY, T15, T16, T11, T17, T19, T18; Tf = Cr[WS(csr, 2)]; { E T1L, TN, T1I, Ty; T1L = FNMS(KP559016994, TM, TL); TN = FMA(KP559016994, TM, TL); T1I = FNMS(KP559016994, Tx, Tw); Ty = FMA(KP559016994, Tx, Tw); T1N = FMA(KP951056516, T1M, T1L); T2a = FNMS(KP951056516, T1M, T1L); T1t = FNMS(KP951056516, TQ, TN); TR = FMA(KP951056516, TQ, TN); T1K = FMA(KP951056516, T1J, T1I); T29 = FNMS(KP951056516, T1J, T1I); T1u = FMA(KP951056516, TF, Ty); TG = FNMS(KP951056516, TF, Ty); T14 = Ci[WS(csi, 2)]; } { E Tg, Th, Tj, Tk; Tg = Cr[WS(csr, 7)]; Th = Cr[WS(csr, 3)]; Tj = Cr[WS(csr, 12)]; Tk = Cr[WS(csr, 8)]; { E TW, Ti, Tl, TX, TZ, T10; TW = Ci[WS(csi, 7)]; T1b = Th - Tg; Ti = Tg + Th; T1c = Tj - Tk; Tl = Tj + Tk; TX = Ci[WS(csi, 3)]; TZ = Ci[WS(csi, 12)]; T10 = Ci[WS(csi, 8)]; Tm = Ti + Tl; TU = Tl - Ti; TY = TW + TX; T15 = TW - TX; T16 = TZ - T10; T11 = TZ + T10; } } TT = FNMS(KP250000000, Tm, Tf); Tn = Tf + Tm; T17 = T15 + T16; T19 = T16 - T15; T1d = FNMS(KP618033988, T1c, T1b); T1Q = FMA(KP618033988, T1b, T1c); T18 = FNMS(KP250000000, T17, T14); T2p = T17 + T14; T1T = FNMS(KP618033988, TY, T11); T12 = FMA(KP618033988, T11, TY); T1P = FMA(KP559016994, T19, T18); T1a = FNMS(KP559016994, T19, T18); } } } { E T1R, T1e, T1q, T1U, T13, T1r, T2b, T28, T25, T2i, T2k; { E T2m, To, T26, T27, TV, T1S; T2m = Te - Tn; To = Te + Tn; TV = FNMS(KP559016994, TU, TT); T1S = FMA(KP559016994, TU, TT); T26 = FMA(KP951056516, T1Q, T1P); T1R = FNMS(KP951056516, T1Q, T1P); T1e = FNMS(KP951056516, T1d, T1a); T1q = FMA(KP951056516, T1d, T1a); T27 = FNMS(KP951056516, T1T, T1S); T1U = FMA(KP951056516, T1T, T1S); T13 = FNMS(KP951056516, T12, TV); T1r = FMA(KP951056516, T12, TV); { E T2g, T2q, T2s, T2h, T2n, T2r, T2l; T2g = FMA(KP939062505, T29, T2a); T2b = FNMS(KP939062505, T2a, T29); R0[0] = FMA(KP2_000000000, To, T5); T2l = FNMS(KP500000000, To, T5); T2q = FMA(KP618033988, T2p, T2o); T2s = FNMS(KP618033988, T2o, T2p); T28 = FNMS(KP062914667, T27, T26); T2h = FMA(KP062914667, T26, T27); T2n = FMA(KP1_118033988, T2m, T2l); T2r = FNMS(KP1_118033988, T2m, T2l); T25 = FMA(KP1_902113032, T1G, T1F); T1H = FNMS(KP1_902113032, T1G, T1F); T2i = FMA(KP846146756, T2h, T2g); T2k = FNMS(KP451418159, T2g, T2h); R0[WS(rs, 10)] = FMA(KP1_902113032, T2q, T2n); R1[WS(rs, 2)] = FNMS(KP1_902113032, T2q, T2n); R0[WS(rs, 5)] = FMA(KP1_902113032, T2s, T2r); R1[WS(rs, 7)] = FNMS(KP1_902113032, T2s, T2r); } } { E TS, T1f, T1p, Tv, T2e, T1o, T1m, T2d, T1k, T1l, T2c; TS = FNMS(KP256756360, TR, TG); T1k = FMA(KP256756360, TG, TR); T1l = FMA(KP549754652, T13, T1e); T1f = FNMS(KP549754652, T1e, T13); T1p = FMA(KP1_902113032, Tu, Tr); Tv = FNMS(KP1_902113032, Tu, Tr); T2e = FMA(KP730409924, T2b, T28); T2c = FNMS(KP730409924, T2b, T28); T1o = FNMS(KP683113946, T1k, T1l); T1m = FMA(KP559154169, T1l, T1k); R1[WS(rs, 1)] = FNMS(KP1_996053456, T2c, T25); T2d = FMA(KP499013364, T2c, T25); { E T1C, T1E, T1y, T1w; { E T1s, T1v, T1i, T1h, T1n, T1j; { E T1A, T1B, T2f, T2j, T1g; T1A = FNMS(KP470564281, T1q, T1r); T1s = FMA(KP470564281, T1r, T1q); T1v = FNMS(KP634619297, T1u, T1t); T1B = FMA(KP634619297, T1t, T1u); T2f = FMA(KP1_115827804, T2e, T2d); T2j = FNMS(KP1_115827804, T2e, T2d); T1i = FNMS(KP904730450, T1f, TS); T1g = FMA(KP904730450, T1f, TS); R1[WS(rs, 11)] = FMA(KP1_386580726, T2i, T2f); R0[WS(rs, 4)] = FNMS(KP1_386580726, T2i, T2f); R1[WS(rs, 6)] = FMA(KP1_898359647, T2k, T2j); R0[WS(rs, 9)] = FNMS(KP1_898359647, T2k, T2j); R1[0] = FMA(KP1_937166322, T1g, Tv); T1h = FNMS(KP484291580, T1g, Tv); T1C = FNMS(KP576710603, T1B, T1A); T1E = FMA(KP662318342, T1A, T1B); } T1n = FNMS(KP1_082908895, T1i, T1h); T1j = FMA(KP1_082908895, T1i, T1h); R1[WS(rs, 10)] = FMA(KP1_842354653, T1m, T1j); R0[WS(rs, 3)] = FNMS(KP1_842354653, T1m, T1j); R1[WS(rs, 5)] = FMA(KP1_666834356, T1o, T1n); R0[WS(rs, 8)] = FNMS(KP1_666834356, T1o, T1n); T1y = FNMS(KP933137358, T1v, T1s); T1w = FMA(KP933137358, T1v, T1s); } { E T1O, T20, T21, T1V, T1x, T1z, T1D; T1O = FNMS(KP549754652, T1N, T1K); T20 = FMA(KP549754652, T1K, T1N); T21 = FMA(KP634619297, T1R, T1U); T1V = FNMS(KP634619297, T1U, T1R); R0[WS(rs, 2)] = FNMS(KP1_809654104, T1w, T1p); T1x = FMA(KP452413526, T1w, T1p); T24 = FNMS(KP641441904, T20, T21); T22 = FMA(KP595480289, T21, T20); T1z = FNMS(KP1_011627398, T1y, T1x); T1D = FMA(KP1_011627398, T1y, T1x); R1[WS(rs, 9)] = FNMS(KP1_721083328, T1C, T1z); R0[WS(rs, 7)] = FMA(KP1_721083328, T1C, T1z); R0[WS(rs, 12)] = FMA(KP1_606007150, T1E, T1D); R1[WS(rs, 4)] = FNMS(KP1_606007150, T1E, T1D); T1W = FNMS(KP963507348, T1V, T1O); T1Y = FMA(KP963507348, T1V, T1O); } } } } } R0[WS(rs, 1)] = FMA(KP1_752613360, T1W, T1H); T1X = FNMS(KP438153340, T1W, T1H); T1Z = FMA(KP979740652, T1Y, T1X); T23 = FNMS(KP979740652, T1Y, T1X); R0[WS(rs, 11)] = FMA(KP1_666834356, T22, T1Z); R1[WS(rs, 3)] = FNMS(KP1_666834356, T22, T1Z); R1[WS(rs, 8)] = FNMS(KP1_606007150, T24, T23); R0[WS(rs, 6)] = FMA(KP1_606007150, T24, T23); } } } static const kr2c_desc desc = { 25, "r2cb_25", {32, 0, 120, 0}, &GENUS }; void X(codelet_r2cb_25) (planner *p) { X(kr2c_register) (p, r2cb_25, &desc); } #else /* HAVE_FMA */ /* Generated by: ../../../genfft/gen_r2cb.native -compact -variables 4 -pipeline-latency 4 -sign 1 -n 25 -name r2cb_25 -include r2cb.h */ /* * This function contains 152 FP additions, 98 FP multiplications, * (or, 100 additions, 46 multiplications, 52 fused multiply/add), * 65 stack variables, 21 constants, and 50 memory accesses */ #include "r2cb.h" static void r2cb_25(R *R0, R *R1, R *Cr, R *Ci, stride rs, stride csr, stride csi, INT v, INT ivs, INT ovs) { DK(KP425779291, +0.425779291565072648862502445744251703979973042); DK(KP904827052, +0.904827052466019527713668647932697593970413911); DK(KP535826794, +0.535826794978996618271308767867639978063575346); DK(KP844327925, +0.844327925502015078548558063966681505381659241); DK(KP876306680, +0.876306680043863587308115903922062583399064238); DK(KP481753674, +0.481753674101715274987191502872129653528542010); DK(KP968583161, +0.968583161128631119490168375464735813836012403); DK(KP248689887, +0.248689887164854788242283746006447968417567406); DK(KP062790519, +0.062790519529313376076178224565631133122484832); DK(KP998026728, +0.998026728428271561952336806863450553336905220); DK(KP728968627, +0.728968627421411523146730319055259111372571664); DK(KP684547105, +0.684547105928688673732283357621209269889519233); DK(KP250000000, +0.250000000000000000000000000000000000000000000); DK(KP587785252, +0.587785252292473129168705954639072768597652438); DK(KP951056516, +0.951056516295153572116439333379382143405698634); DK(KP559016994, +0.559016994374947424102293417182819058860154590); DK(KP500000000, +0.500000000000000000000000000000000000000000000); DK(KP2_000000000, +2.000000000000000000000000000000000000000000000); DK(KP1_118033988, +1.118033988749894848204586834365638117720309180); DK(KP1_175570504, +1.175570504584946258337411909278145537195304875); DK(KP1_902113032, +1.902113032590307144232878666758764286811397268); { INT i; for (i = v; i > 0; i = i - 1, R0 = R0 + ovs, R1 = R1 + ovs, Cr = Cr + ivs, Ci = Ci + ivs, MAKE_VOLATILE_STRIDE(100, rs), MAKE_VOLATILE_STRIDE(100, csr), MAKE_VOLATILE_STRIDE(100, csi)) { E Tu, T1G, T5, Tr, T1F, TN, TO, Te, TR, T27, T1r, T1N, TG, T26, T1q; E T1K, T1a, T1b, Tn, T1e, T2a, T1u, T1U, T13, T29, T1t, T1R, Ts, Tt; Ts = Ci[WS(csi, 5)]; Tt = Ci[WS(csi, 10)]; Tu = FMA(KP1_902113032, Ts, KP1_175570504 * Tt); T1G = FNMS(KP1_902113032, Tt, KP1_175570504 * Ts); { E T1, T4, Tp, T2, T3, Tq; T1 = Cr[0]; T2 = Cr[WS(csr, 5)]; T3 = Cr[WS(csr, 10)]; T4 = T2 + T3; Tp = KP1_118033988 * (T2 - T3); T5 = FMA(KP2_000000000, T4, T1); Tq = FNMS(KP500000000, T4, T1); Tr = Tp + Tq; T1F = Tq - Tp; } { E T6, Td, TI, Tw, TH, TB, TE, TM; T6 = Cr[WS(csr, 1)]; TN = Ci[WS(csi, 1)]; { E T7, T8, T9, Ta, Tb, Tc; T7 = Cr[WS(csr, 6)]; T8 = Cr[WS(csr, 4)]; T9 = T7 + T8; Ta = Cr[WS(csr, 11)]; Tb = Cr[WS(csr, 9)]; Tc = Ta + Tb; Td = T9 + Tc; TI = Ta - Tb; Tw = KP559016994 * (T9 - Tc); TH = T7 - T8; } { E Tz, TA, TK, TC, TD, TL; Tz = Ci[WS(csi, 6)]; TA = Ci[WS(csi, 4)]; TK = Tz - TA; TC = Ci[WS(csi, 11)]; TD = Ci[WS(csi, 9)]; TL = TC - TD; TB = Tz + TA; TO = TK + TL; TE = TC + TD; TM = KP559016994 * (TK - TL); } Te = T6 + Td; { E TJ, T1L, TQ, T1M, TP; TJ = FMA(KP951056516, TH, KP587785252 * TI); T1L = FNMS(KP951056516, TI, KP587785252 * TH); TP = FNMS(KP250000000, TO, TN); TQ = TM + TP; T1M = TP - TM; TR = TJ + TQ; T27 = T1M - T1L; T1r = TQ - TJ; T1N = T1L + T1M; } { E TF, T1J, Ty, T1I, Tx; TF = FMA(KP951056516, TB, KP587785252 * TE); T1J = FNMS(KP951056516, TE, KP587785252 * TB); Tx = FNMS(KP250000000, Td, T6); Ty = Tw + Tx; T1I = Tx - Tw; TG = Ty - TF; T26 = T1I + T1J; T1q = Ty + TF; T1K = T1I - T1J; } } { E Tf, Tm, T15, TT, T14, TY, T11, T19; Tf = Cr[WS(csr, 2)]; T1a = Ci[WS(csi, 2)]; { E Tg, Th, Ti, Tj, Tk, Tl; Tg = Cr[WS(csr, 7)]; Th = Cr[WS(csr, 3)]; Ti = Tg + Th; Tj = Cr[WS(csr, 12)]; Tk = Cr[WS(csr, 8)]; Tl = Tj + Tk; Tm = Ti + Tl; T15 = Tj - Tk; TT = KP559016994 * (Ti - Tl); T14 = Tg - Th; } { E TW, TX, T17, TZ, T10, T18; TW = Ci[WS(csi, 7)]; TX = Ci[WS(csi, 3)]; T17 = TW - TX; TZ = Ci[WS(csi, 12)]; T10 = Ci[WS(csi, 8)]; T18 = TZ - T10; TY = TW + TX; T1b = T17 + T18; T11 = TZ + T10; T19 = KP559016994 * (T17 - T18); } Tn = Tf + Tm; { E T16, T1S, T1d, T1T, T1c; T16 = FMA(KP951056516, T14, KP587785252 * T15); T1S = FNMS(KP951056516, T15, KP587785252 * T14); T1c = FNMS(KP250000000, T1b, T1a); T1d = T19 + T1c; T1T = T1c - T19; T1e = T16 + T1d; T2a = T1T - T1S; T1u = T1d - T16; T1U = T1S + T1T; } { E T12, T1Q, TV, T1P, TU; T12 = FMA(KP951056516, TY, KP587785252 * T11); T1Q = FNMS(KP951056516, T11, KP587785252 * TY); TU = FNMS(KP250000000, Tm, Tf); TV = TT + TU; T1P = TU - TT; T13 = TV - T12; T29 = T1P + T1Q; T1t = TV + T12; T1R = T1P - T1Q; } } { E T2m, To, T2l, T2q, T2s, T2o, T2p, T2r, T2n; T2m = KP1_118033988 * (Te - Tn); To = Te + Tn; T2l = FNMS(KP500000000, To, T5); T2o = TO + TN; T2p = T1b + T1a; T2q = FNMS(KP1_902113032, T2p, KP1_175570504 * T2o); T2s = FMA(KP1_902113032, T2o, KP1_175570504 * T2p); R0[0] = FMA(KP2_000000000, To, T5); T2r = T2m + T2l; R1[WS(rs, 2)] = T2r - T2s; R0[WS(rs, 10)] = T2r + T2s; T2n = T2l - T2m; R0[WS(rs, 5)] = T2n - T2q; R1[WS(rs, 7)] = T2n + T2q; } { E T2i, T2k, T25, T2c, T2d, T2e, T2j, T2f; { E T2g, T2h, T28, T2b; T2g = FMA(KP684547105, T26, KP728968627 * T27); T2h = FMA(KP998026728, T29, KP062790519 * T2a); T2i = FNMS(KP1_902113032, T2h, KP1_175570504 * T2g); T2k = FMA(KP1_902113032, T2g, KP1_175570504 * T2h); T25 = T1F + T1G; T28 = FNMS(KP684547105, T27, KP728968627 * T26); T2b = FNMS(KP998026728, T2a, KP062790519 * T29); T2c = T28 + T2b; T2d = FNMS(KP500000000, T2c, T25); T2e = KP1_118033988 * (T28 - T2b); } R1[WS(rs, 1)] = FMA(KP2_000000000, T2c, T25); T2j = T2e + T2d; R0[WS(rs, 4)] = T2j - T2k; R1[WS(rs, 11)] = T2j + T2k; T2f = T2d - T2e; R1[WS(rs, 6)] = T2f - T2i; R0[WS(rs, 9)] = T2f + T2i; } { E T1m, T1o, Tv, T1g, T1h, T1i, T1n, T1j; { E T1k, T1l, TS, T1f; T1k = FMA(KP248689887, TG, KP968583161 * TR); T1l = FMA(KP481753674, T13, KP876306680 * T1e); T1m = FNMS(KP1_902113032, T1l, KP1_175570504 * T1k); T1o = FMA(KP1_902113032, T1k, KP1_175570504 * T1l); Tv = Tr - Tu; TS = FNMS(KP248689887, TR, KP968583161 * TG); T1f = FNMS(KP481753674, T1e, KP876306680 * T13); T1g = TS + T1f; T1h = FNMS(KP500000000, T1g, Tv); T1i = KP1_118033988 * (TS - T1f); } R1[0] = FMA(KP2_000000000, T1g, Tv); T1n = T1i + T1h; R0[WS(rs, 3)] = T1n - T1o; R1[WS(rs, 10)] = T1n + T1o; T1j = T1h - T1i; R1[WS(rs, 5)] = T1j - T1m; R0[WS(rs, 8)] = T1j + T1m; } { E T1C, T1E, T1p, T1w, T1x, T1y, T1D, T1z; { E T1A, T1B, T1s, T1v; T1A = FMA(KP844327925, T1q, KP535826794 * T1r); T1B = FNMS(KP425779291, T1u, KP904827052 * T1t); T1C = FNMS(KP1_902113032, T1B, KP1_175570504 * T1A); T1E = FMA(KP1_902113032, T1A, KP1_175570504 * T1B); T1p = Tr + Tu; T1s = FNMS(KP844327925, T1r, KP535826794 * T1q); T1v = FMA(KP425779291, T1t, KP904827052 * T1u); T1w = T1s - T1v; T1x = FNMS(KP500000000, T1w, T1p); T1y = KP1_118033988 * (T1s + T1v); } R0[WS(rs, 2)] = FMA(KP2_000000000, T1w, T1p); T1D = T1x + T1y; R1[WS(rs, 4)] = T1D - T1E; R0[WS(rs, 12)] = T1E + T1D; T1z = T1x - T1y; R0[WS(rs, 7)] = T1z - T1C; R1[WS(rs, 9)] = T1C + T1z; } { E T22, T24, T1H, T1W, T1X, T1Y, T23, T1Z; { E T20, T21, T1O, T1V; T20 = FMA(KP481753674, T1K, KP876306680 * T1N); T21 = FMA(KP844327925, T1R, KP535826794 * T1U); T22 = FNMS(KP1_902113032, T21, KP1_175570504 * T20); T24 = FMA(KP1_902113032, T20, KP1_175570504 * T21); T1H = T1F - T1G; T1O = FNMS(KP481753674, T1N, KP876306680 * T1K); T1V = FNMS(KP844327925, T1U, KP535826794 * T1R); T1W = T1O + T1V; T1X = FNMS(KP500000000, T1W, T1H); T1Y = KP1_118033988 * (T1O - T1V); } R0[WS(rs, 1)] = FMA(KP2_000000000, T1W, T1H); T23 = T1Y + T1X; R1[WS(rs, 3)] = T23 - T24; R0[WS(rs, 11)] = T23 + T24; T1Z = T1X - T1Y; R0[WS(rs, 6)] = T1Z - T22; R1[WS(rs, 8)] = T1Z + T22; } } } } static const kr2c_desc desc = { 25, "r2cb_25", {100, 46, 52, 0}, &GENUS }; void X(codelet_r2cb_25) (planner *p) { X(kr2c_register) (p, r2cb_25, &desc); } #endif /* HAVE_FMA */ fftw-3.3.3/rdft/scalar/r2cb/hb_2.c0000644000175000017500000000670412054410750013426 00000000000000/* * Copyright (c) 2003, 2007-11 Matteo Frigo * Copyright (c) 2003, 2007-11 Massachusetts Institute of Technology * * This program is free software; you can redistribute it and/or modify * it under the terms of the GNU General Public License as published by * the Free Software Foundation; either version 2 of the License, or * (at your option) any later version. * * This program is distributed in the hope that it will be useful, * but WITHOUT ANY WARRANTY; without even the implied warranty of * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the * GNU General Public License for more details. * * You should have received a copy of the GNU General Public License * along with this program; if not, write to the Free Software * Foundation, Inc., 51 Franklin Street, Fifth Floor, Boston, MA 02110-1301 USA * */ /* This file was automatically generated --- DO NOT EDIT */ /* Generated on Sun Nov 25 07:41:12 EST 2012 */ #include "codelet-rdft.h" #ifdef HAVE_FMA /* Generated by: ../../../genfft/gen_hc2hc.native -fma -reorder-insns -schedule-for-pipeline -compact -variables 4 -pipeline-latency 4 -sign 1 -n 2 -dif -name hb_2 -include hb.h */ /* * This function contains 6 FP additions, 4 FP multiplications, * (or, 4 additions, 2 multiplications, 2 fused multiply/add), * 11 stack variables, 0 constants, and 8 memory accesses */ #include "hb.h" static void hb_2(R *cr, R *ci, const R *W, stride rs, INT mb, INT me, INT ms) { { INT m; for (m = mb, W = W + ((mb - 1) * 2); m < me; m = m + 1, cr = cr + ms, ci = ci - ms, W = W + 2, MAKE_VOLATILE_STRIDE(4, rs)) { E T5, T6, T9, T8, T7, Ta; { E T1, T2, T3, T4; T1 = cr[0]; T2 = ci[0]; T3 = ci[WS(rs, 1)]; T4 = cr[WS(rs, 1)]; T5 = W[0]; cr[0] = T1 + T2; T6 = T1 - T2; ci[0] = T3 - T4; T9 = T3 + T4; T8 = W[1]; T7 = T5 * T6; } Ta = T8 * T6; cr[WS(rs, 1)] = FNMS(T8, T9, T7); ci[WS(rs, 1)] = FMA(T5, T9, Ta); } } } static const tw_instr twinstr[] = { {TW_FULL, 1, 2}, {TW_NEXT, 1, 0} }; static const hc2hc_desc desc = { 2, "hb_2", twinstr, &GENUS, {4, 2, 2, 0} }; void X(codelet_hb_2) (planner *p) { X(khc2hc_register) (p, hb_2, &desc); } #else /* HAVE_FMA */ /* Generated by: ../../../genfft/gen_hc2hc.native -compact -variables 4 -pipeline-latency 4 -sign 1 -n 2 -dif -name hb_2 -include hb.h */ /* * This function contains 6 FP additions, 4 FP multiplications, * (or, 4 additions, 2 multiplications, 2 fused multiply/add), * 9 stack variables, 0 constants, and 8 memory accesses */ #include "hb.h" static void hb_2(R *cr, R *ci, const R *W, stride rs, INT mb, INT me, INT ms) { { INT m; for (m = mb, W = W + ((mb - 1) * 2); m < me; m = m + 1, cr = cr + ms, ci = ci - ms, W = W + 2, MAKE_VOLATILE_STRIDE(4, rs)) { E T1, T2, T6, T3, T4, T8, T5, T7; T1 = cr[0]; T2 = ci[0]; T6 = T1 - T2; T3 = ci[WS(rs, 1)]; T4 = cr[WS(rs, 1)]; T8 = T3 + T4; cr[0] = T1 + T2; ci[0] = T3 - T4; T5 = W[0]; T7 = W[1]; cr[WS(rs, 1)] = FNMS(T7, T8, T5 * T6); ci[WS(rs, 1)] = FMA(T7, T6, T5 * T8); } } } static const tw_instr twinstr[] = { {TW_FULL, 1, 2}, {TW_NEXT, 1, 0} }; static const hc2hc_desc desc = { 2, "hb_2", twinstr, &GENUS, {4, 2, 2, 0} }; void X(codelet_hb_2) (planner *p) { X(khc2hc_register) (p, hb_2, &desc); } #endif /* HAVE_FMA */ fftw-3.3.3/rdft/scalar/r2cb/hb_3.c0000644000175000017500000001136112054410750013422 00000000000000/* * Copyright (c) 2003, 2007-11 Matteo Frigo * Copyright (c) 2003, 2007-11 Massachusetts Institute of Technology * * This program is free software; you can redistribute it and/or modify * it under the terms of the GNU General Public License as published by * the Free Software Foundation; either version 2 of the License, or * (at your option) any later version. * * This program is distributed in the hope that it will be useful, * but WITHOUT ANY WARRANTY; without even the implied warranty of * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the * GNU General Public License for more details. * * You should have received a copy of the GNU General Public License * along with this program; if not, write to the Free Software * Foundation, Inc., 51 Franklin Street, Fifth Floor, Boston, MA 02110-1301 USA * */ /* This file was automatically generated --- DO NOT EDIT */ /* Generated on Sun Nov 25 07:41:12 EST 2012 */ #include "codelet-rdft.h" #ifdef HAVE_FMA /* Generated by: ../../../genfft/gen_hc2hc.native -fma -reorder-insns -schedule-for-pipeline -compact -variables 4 -pipeline-latency 4 -sign 1 -n 3 -dif -name hb_3 -include hb.h */ /* * This function contains 16 FP additions, 14 FP multiplications, * (or, 6 additions, 4 multiplications, 10 fused multiply/add), * 27 stack variables, 2 constants, and 12 memory accesses */ #include "hb.h" static void hb_3(R *cr, R *ci, const R *W, stride rs, INT mb, INT me, INT ms) { DK(KP866025403, +0.866025403784438646763723170752936183471402627); DK(KP500000000, +0.500000000000000000000000000000000000000000000); { INT m; for (m = mb, W = W + ((mb - 1) * 4); m < me; m = m + 1, cr = cr + ms, ci = ci - ms, W = W + 4, MAKE_VOLATILE_STRIDE(6, rs)) { E Tk, Tj, Tn, Tl, Tm, To; { E T1, Td, T7, T8, T4, Tg, T2, T3; T1 = cr[0]; T2 = cr[WS(rs, 1)]; T3 = ci[0]; Td = ci[WS(rs, 2)]; T7 = ci[WS(rs, 1)]; T8 = cr[WS(rs, 2)]; T4 = T2 + T3; Tg = T2 - T3; { E T5, Tc, Tf, Ta, T9, Te, T6, Th, Ti, Tb; T5 = W[0]; T9 = T7 + T8; Te = T7 - T8; cr[0] = T1 + T4; T6 = FNMS(KP500000000, T4, T1); Tc = W[1]; ci[0] = Td + Te; Tf = FNMS(KP500000000, Te, Td); Tk = FMA(KP866025403, T9, T6); Ta = FNMS(KP866025403, T9, T6); Tj = W[2]; Tn = FNMS(KP866025403, Tg, Tf); Th = FMA(KP866025403, Tg, Tf); Ti = Tc * Ta; Tb = T5 * Ta; Tl = Tj * Tk; Tm = W[3]; ci[WS(rs, 1)] = FMA(T5, Th, Ti); cr[WS(rs, 1)] = FNMS(Tc, Th, Tb); } } cr[WS(rs, 2)] = FNMS(Tm, Tn, Tl); To = Tm * Tk; ci[WS(rs, 2)] = FMA(Tj, Tn, To); } } } static const tw_instr twinstr[] = { {TW_FULL, 1, 3}, {TW_NEXT, 1, 0} }; static const hc2hc_desc desc = { 3, "hb_3", twinstr, &GENUS, {6, 4, 10, 0} }; void X(codelet_hb_3) (planner *p) { X(khc2hc_register) (p, hb_3, &desc); } #else /* HAVE_FMA */ /* Generated by: ../../../genfft/gen_hc2hc.native -compact -variables 4 -pipeline-latency 4 -sign 1 -n 3 -dif -name hb_3 -include hb.h */ /* * This function contains 16 FP additions, 12 FP multiplications, * (or, 10 additions, 6 multiplications, 6 fused multiply/add), * 15 stack variables, 2 constants, and 12 memory accesses */ #include "hb.h" static void hb_3(R *cr, R *ci, const R *W, stride rs, INT mb, INT me, INT ms) { DK(KP866025403, +0.866025403784438646763723170752936183471402627); DK(KP500000000, +0.500000000000000000000000000000000000000000000); { INT m; for (m = mb, W = W + ((mb - 1) * 4); m < me; m = m + 1, cr = cr + ms, ci = ci - ms, W = W + 4, MAKE_VOLATILE_STRIDE(6, rs)) { E T1, T4, Ta, Te, T5, T8, Tb, Tf; { E T2, T3, T6, T7; T1 = cr[0]; T2 = cr[WS(rs, 1)]; T3 = ci[0]; T4 = T2 + T3; Ta = FNMS(KP500000000, T4, T1); Te = KP866025403 * (T2 - T3); T5 = ci[WS(rs, 2)]; T6 = ci[WS(rs, 1)]; T7 = cr[WS(rs, 2)]; T8 = T6 - T7; Tb = KP866025403 * (T6 + T7); Tf = FNMS(KP500000000, T8, T5); } cr[0] = T1 + T4; ci[0] = T5 + T8; { E Tc, Tg, T9, Td; Tc = Ta - Tb; Tg = Te + Tf; T9 = W[0]; Td = W[1]; cr[WS(rs, 1)] = FNMS(Td, Tg, T9 * Tc); ci[WS(rs, 1)] = FMA(T9, Tg, Td * Tc); } { E Ti, Tk, Th, Tj; Ti = Ta + Tb; Tk = Tf - Te; Th = W[2]; Tj = W[3]; cr[WS(rs, 2)] = FNMS(Tj, Tk, Th * Ti); ci[WS(rs, 2)] = FMA(Th, Tk, Tj * Ti); } } } } static const tw_instr twinstr[] = { {TW_FULL, 1, 3}, {TW_NEXT, 1, 0} }; static const hc2hc_desc desc = { 3, "hb_3", twinstr, &GENUS, {10, 6, 6, 0} }; void X(codelet_hb_3) (planner *p) { X(khc2hc_register) (p, hb_3, &desc); } #endif /* HAVE_FMA */ fftw-3.3.3/rdft/scalar/r2cb/hb_4.c0000644000175000017500000001212412054410750013421 00000000000000/* * Copyright (c) 2003, 2007-11 Matteo Frigo * Copyright (c) 2003, 2007-11 Massachusetts Institute of Technology * * This program is free software; you can redistribute it and/or modify * it under the terms of the GNU General Public License as published by * the Free Software Foundation; either version 2 of the License, or * (at your option) any later version. * * This program is distributed in the hope that it will be useful, * but WITHOUT ANY WARRANTY; without even the implied warranty of * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the * GNU General Public License for more details. * * You should have received a copy of the GNU General Public License * along with this program; if not, write to the Free Software * Foundation, Inc., 51 Franklin Street, Fifth Floor, Boston, MA 02110-1301 USA * */ /* This file was automatically generated --- DO NOT EDIT */ /* Generated on Sun Nov 25 07:41:12 EST 2012 */ #include "codelet-rdft.h" #ifdef HAVE_FMA /* Generated by: ../../../genfft/gen_hc2hc.native -fma -reorder-insns -schedule-for-pipeline -compact -variables 4 -pipeline-latency 4 -sign 1 -n 4 -dif -name hb_4 -include hb.h */ /* * This function contains 22 FP additions, 12 FP multiplications, * (or, 16 additions, 6 multiplications, 6 fused multiply/add), * 27 stack variables, 0 constants, and 16 memory accesses */ #include "hb.h" static void hb_4(R *cr, R *ci, const R *W, stride rs, INT mb, INT me, INT ms) { { INT m; for (m = mb, W = W + ((mb - 1) * 6); m < me; m = m + 1, cr = cr + ms, ci = ci - ms, W = W + 6, MAKE_VOLATILE_STRIDE(8, rs)) { E T8, Th, Ta, T7, Ti, T9; { E Td, Tg, T3, T6, Tu, Tm, Tx, Tr; { E Tq, Tl, T4, T5, Tk, Tp; { E Tb, Tc, Te, Tf, T1, T2; Tb = ci[WS(rs, 3)]; Tc = cr[WS(rs, 2)]; Te = ci[WS(rs, 2)]; Tf = cr[WS(rs, 3)]; T1 = cr[0]; Tq = Tb + Tc; Td = Tb - Tc; T2 = ci[WS(rs, 1)]; Tl = Te + Tf; Tg = Te - Tf; T4 = cr[WS(rs, 1)]; T5 = ci[0]; T3 = T1 + T2; Tk = T1 - T2; } Tp = T4 - T5; T6 = T4 + T5; Tu = Tk + Tl; Tm = Tk - Tl; Tx = Tq - Tp; Tr = Tp + Tq; T8 = T3 - T6; } cr[0] = T3 + T6; { E Tj, To, Tw, Tv; Tj = W[0]; ci[0] = Td + Tg; To = W[1]; { E Tt, Ts, Tn, Ty; Tt = W[4]; Ts = Tj * Tr; Tn = Tj * Tm; Tw = W[5]; Ty = Tt * Tx; Tv = Tt * Tu; ci[WS(rs, 1)] = FMA(To, Tm, Ts); cr[WS(rs, 1)] = FNMS(To, Tr, Tn); ci[WS(rs, 3)] = FMA(Tw, Tu, Ty); } cr[WS(rs, 3)] = FNMS(Tw, Tx, Tv); Th = Td - Tg; Ta = W[3]; T7 = W[2]; } } Ti = Ta * T8; T9 = T7 * T8; ci[WS(rs, 2)] = FMA(T7, Th, Ti); cr[WS(rs, 2)] = FNMS(Ta, Th, T9); } } } static const tw_instr twinstr[] = { {TW_FULL, 1, 4}, {TW_NEXT, 1, 0} }; static const hc2hc_desc desc = { 4, "hb_4", twinstr, &GENUS, {16, 6, 6, 0} }; void X(codelet_hb_4) (planner *p) { X(khc2hc_register) (p, hb_4, &desc); } #else /* HAVE_FMA */ /* Generated by: ../../../genfft/gen_hc2hc.native -compact -variables 4 -pipeline-latency 4 -sign 1 -n 4 -dif -name hb_4 -include hb.h */ /* * This function contains 22 FP additions, 12 FP multiplications, * (or, 16 additions, 6 multiplications, 6 fused multiply/add), * 13 stack variables, 0 constants, and 16 memory accesses */ #include "hb.h" static void hb_4(R *cr, R *ci, const R *W, stride rs, INT mb, INT me, INT ms) { { INT m; for (m = mb, W = W + ((mb - 1) * 6); m < me; m = m + 1, cr = cr + ms, ci = ci - ms, W = W + 6, MAKE_VOLATILE_STRIDE(8, rs)) { E T3, Ti, T6, Tm, Tc, Tn, Tf, Tj; { E T1, T2, T4, T5; T1 = cr[0]; T2 = ci[WS(rs, 1)]; T3 = T1 + T2; Ti = T1 - T2; T4 = cr[WS(rs, 1)]; T5 = ci[0]; T6 = T4 + T5; Tm = T4 - T5; } { E Ta, Tb, Td, Te; Ta = ci[WS(rs, 3)]; Tb = cr[WS(rs, 2)]; Tc = Ta - Tb; Tn = Ta + Tb; Td = ci[WS(rs, 2)]; Te = cr[WS(rs, 3)]; Tf = Td - Te; Tj = Td + Te; } cr[0] = T3 + T6; ci[0] = Tc + Tf; { E T8, Tg, T7, T9; T8 = T3 - T6; Tg = Tc - Tf; T7 = W[2]; T9 = W[3]; cr[WS(rs, 2)] = FNMS(T9, Tg, T7 * T8); ci[WS(rs, 2)] = FMA(T9, T8, T7 * Tg); } { E Tk, To, Th, Tl; Tk = Ti - Tj; To = Tm + Tn; Th = W[0]; Tl = W[1]; cr[WS(rs, 1)] = FNMS(Tl, To, Th * Tk); ci[WS(rs, 1)] = FMA(Th, To, Tl * Tk); } { E Tq, Ts, Tp, Tr; Tq = Ti + Tj; Ts = Tn - Tm; Tp = W[4]; Tr = W[5]; cr[WS(rs, 3)] = FNMS(Tr, Ts, Tp * Tq); ci[WS(rs, 3)] = FMA(Tp, Ts, Tr * Tq); } } } } static const tw_instr twinstr[] = { {TW_FULL, 1, 4}, {TW_NEXT, 1, 0} }; static const hc2hc_desc desc = { 4, "hb_4", twinstr, &GENUS, {16, 6, 6, 0} }; void X(codelet_hb_4) (planner *p) { X(khc2hc_register) (p, hb_4, &desc); } #endif /* HAVE_FMA */ fftw-3.3.3/rdft/scalar/r2cb/hb_5.c0000644000175000017500000001666112054410750013434 00000000000000/* * Copyright (c) 2003, 2007-11 Matteo Frigo * Copyright (c) 2003, 2007-11 Massachusetts Institute of Technology * * This program is free software; you can redistribute it and/or modify * it under the terms of the GNU General Public License as published by * the Free Software Foundation; either version 2 of the License, or * (at your option) any later version. * * This program is distributed in the hope that it will be useful, * but WITHOUT ANY WARRANTY; without even the implied warranty of * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the * GNU General Public License for more details. * * You should have received a copy of the GNU General Public License * along with this program; if not, write to the Free Software * Foundation, Inc., 51 Franklin Street, Fifth Floor, Boston, MA 02110-1301 USA * */ /* This file was automatically generated --- DO NOT EDIT */ /* Generated on Sun Nov 25 07:41:12 EST 2012 */ #include "codelet-rdft.h" #ifdef HAVE_FMA /* Generated by: ../../../genfft/gen_hc2hc.native -fma -reorder-insns -schedule-for-pipeline -compact -variables 4 -pipeline-latency 4 -sign 1 -n 5 -dif -name hb_5 -include hb.h */ /* * This function contains 40 FP additions, 34 FP multiplications, * (or, 14 additions, 8 multiplications, 26 fused multiply/add), * 42 stack variables, 4 constants, and 20 memory accesses */ #include "hb.h" static void hb_5(R *cr, R *ci, const R *W, stride rs, INT mb, INT me, INT ms) { DK(KP951056516, +0.951056516295153572116439333379382143405698634); DK(KP559016994, +0.559016994374947424102293417182819058860154590); DK(KP250000000, +0.250000000000000000000000000000000000000000000); DK(KP618033988, +0.618033988749894848204586834365638117720309180); { INT m; for (m = mb, W = W + ((mb - 1) * 8); m < me; m = m + 1, cr = cr + ms, ci = ci - ms, W = W + 8, MAKE_VOLATILE_STRIDE(10, rs)) { E TQ, TP, TT, TR, TS, TU; { E T1, Tn, TM, Tw, Tb, T8, To, Tf, Ta, Tg, Th; { E T2, T3, T5, T6, T4, Tu; T1 = cr[0]; T2 = cr[WS(rs, 1)]; T3 = ci[0]; T5 = cr[WS(rs, 2)]; T6 = ci[WS(rs, 1)]; Tn = ci[WS(rs, 4)]; T4 = T2 + T3; Tu = T2 - T3; { E T7, Tv, Td, Te; T7 = T5 + T6; Tv = T5 - T6; Td = ci[WS(rs, 3)]; Te = cr[WS(rs, 4)]; TM = FNMS(KP618033988, Tu, Tv); Tw = FMA(KP618033988, Tv, Tu); Tb = T4 - T7; T8 = T4 + T7; To = Td - Te; Tf = Td + Te; Ta = FNMS(KP250000000, T8, T1); Tg = ci[WS(rs, 2)]; Th = cr[WS(rs, 3)]; } } cr[0] = T1 + T8; { E TG, T9, Tm, Tz, TH, TC, TA, Tk, Tt, TL, Tc, Ti, Tp, TI; TG = FNMS(KP559016994, Tb, Ta); Tc = FMA(KP559016994, Tb, Ta); T9 = W[0]; Ti = Tg + Th; Tp = Tg - Th; Tm = W[1]; { E Ts, Tj, Tr, Tq; Tz = W[6]; Ts = To - Tp; Tq = To + Tp; Tj = FMA(KP618033988, Ti, Tf); TH = FNMS(KP618033988, Tf, Ti); ci[0] = Tn + Tq; Tr = FNMS(KP250000000, Tq, Tn); TC = W[7]; TA = FMA(KP951056516, Tj, Tc); Tk = FNMS(KP951056516, Tj, Tc); Tt = FMA(KP559016994, Ts, Tr); TL = FNMS(KP559016994, Ts, Tr); } { E TE, TB, Ty, Tl, TD, Tx; TE = TC * TA; TB = Tz * TA; Ty = Tm * Tk; Tl = T9 * Tk; TD = FNMS(KP951056516, Tw, Tt); Tx = FMA(KP951056516, Tw, Tt); TI = FMA(KP951056516, TH, TG); TQ = FNMS(KP951056516, TH, TG); ci[WS(rs, 4)] = FMA(Tz, TD, TE); cr[WS(rs, 4)] = FNMS(TC, TD, TB); ci[WS(rs, 1)] = FMA(T9, Tx, Ty); cr[WS(rs, 1)] = FNMS(Tm, Tx, Tl); } { E TF, TK, TN, TJ, TO; TF = W[2]; TK = W[3]; TP = W[4]; TT = FMA(KP951056516, TM, TL); TN = FNMS(KP951056516, TM, TL); TJ = TF * TI; TO = TK * TI; TR = TP * TQ; TS = W[5]; cr[WS(rs, 2)] = FNMS(TK, TN, TJ); ci[WS(rs, 2)] = FMA(TF, TN, TO); } } } cr[WS(rs, 3)] = FNMS(TS, TT, TR); TU = TS * TQ; ci[WS(rs, 3)] = FMA(TP, TT, TU); } } } static const tw_instr twinstr[] = { {TW_FULL, 1, 5}, {TW_NEXT, 1, 0} }; static const hc2hc_desc desc = { 5, "hb_5", twinstr, &GENUS, {14, 8, 26, 0} }; void X(codelet_hb_5) (planner *p) { X(khc2hc_register) (p, hb_5, &desc); } #else /* HAVE_FMA */ /* Generated by: ../../../genfft/gen_hc2hc.native -compact -variables 4 -pipeline-latency 4 -sign 1 -n 5 -dif -name hb_5 -include hb.h */ /* * This function contains 40 FP additions, 28 FP multiplications, * (or, 26 additions, 14 multiplications, 14 fused multiply/add), * 27 stack variables, 4 constants, and 20 memory accesses */ #include "hb.h" static void hb_5(R *cr, R *ci, const R *W, stride rs, INT mb, INT me, INT ms) { DK(KP250000000, +0.250000000000000000000000000000000000000000000); DK(KP587785252, +0.587785252292473129168705954639072768597652438); DK(KP951056516, +0.951056516295153572116439333379382143405698634); DK(KP559016994, +0.559016994374947424102293417182819058860154590); { INT m; for (m = mb, W = W + ((mb - 1) * 8); m < me; m = m + 1, cr = cr + ms, ci = ci - ms, W = W + 8, MAKE_VOLATILE_STRIDE(10, rs)) { E T1, Tj, TG, Ts, T8, Ti, T9, Tn, TD, Tu, Tg, Tt; { E T4, Tq, T7, Tr; T1 = cr[0]; { E T2, T3, T5, T6; T2 = cr[WS(rs, 1)]; T3 = ci[0]; T4 = T2 + T3; Tq = T2 - T3; T5 = cr[WS(rs, 2)]; T6 = ci[WS(rs, 1)]; T7 = T5 + T6; Tr = T5 - T6; } Tj = KP559016994 * (T4 - T7); TG = FMA(KP951056516, Tq, KP587785252 * Tr); Ts = FNMS(KP951056516, Tr, KP587785252 * Tq); T8 = T4 + T7; Ti = FNMS(KP250000000, T8, T1); } { E Tc, Tl, Tf, Tm; T9 = ci[WS(rs, 4)]; { E Ta, Tb, Td, Te; Ta = ci[WS(rs, 3)]; Tb = cr[WS(rs, 4)]; Tc = Ta - Tb; Tl = Ta + Tb; Td = ci[WS(rs, 2)]; Te = cr[WS(rs, 3)]; Tf = Td - Te; Tm = Td + Te; } Tn = FNMS(KP951056516, Tm, KP587785252 * Tl); TD = FMA(KP951056516, Tl, KP587785252 * Tm); Tu = KP559016994 * (Tc - Tf); Tg = Tc + Tf; Tt = FNMS(KP250000000, Tg, T9); } cr[0] = T1 + T8; ci[0] = T9 + Tg; { E To, Ty, Tw, TA, Tk, Tv; Tk = Ti - Tj; To = Tk - Tn; Ty = Tk + Tn; Tv = Tt - Tu; Tw = Ts + Tv; TA = Tv - Ts; { E Th, Tp, Tx, Tz; Th = W[2]; Tp = W[3]; cr[WS(rs, 2)] = FNMS(Tp, Tw, Th * To); ci[WS(rs, 2)] = FMA(Th, Tw, Tp * To); Tx = W[4]; Tz = W[5]; cr[WS(rs, 3)] = FNMS(Tz, TA, Tx * Ty); ci[WS(rs, 3)] = FMA(Tx, TA, Tz * Ty); } } { E TE, TK, TI, TM, TC, TH; TC = Tj + Ti; TE = TC - TD; TK = TC + TD; TH = Tu + Tt; TI = TG + TH; TM = TH - TG; { E TB, TF, TJ, TL; TB = W[0]; TF = W[1]; cr[WS(rs, 1)] = FNMS(TF, TI, TB * TE); ci[WS(rs, 1)] = FMA(TB, TI, TF * TE); TJ = W[6]; TL = W[7]; cr[WS(rs, 4)] = FNMS(TL, TM, TJ * TK); ci[WS(rs, 4)] = FMA(TJ, TM, TL * TK); } } } } } static const tw_instr twinstr[] = { {TW_FULL, 1, 5}, {TW_NEXT, 1, 0} }; static const hc2hc_desc desc = { 5, "hb_5", twinstr, &GENUS, {26, 14, 14, 0} }; void X(codelet_hb_5) (planner *p) { X(khc2hc_register) (p, hb_5, &desc); } #endif /* HAVE_FMA */ fftw-3.3.3/rdft/scalar/r2cb/hb_6.c0000644000175000017500000001733012054410750013427 00000000000000/* * Copyright (c) 2003, 2007-11 Matteo Frigo * Copyright (c) 2003, 2007-11 Massachusetts Institute of Technology * * This program is free software; you can redistribute it and/or modify * it under the terms of the GNU General Public License as published by * the Free Software Foundation; either version 2 of the License, or * (at your option) any later version. * * This program is distributed in the hope that it will be useful, * but WITHOUT ANY WARRANTY; without even the implied warranty of * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the * GNU General Public License for more details. * * You should have received a copy of the GNU General Public License * along with this program; if not, write to the Free Software * Foundation, Inc., 51 Franklin Street, Fifth Floor, Boston, MA 02110-1301 USA * */ /* This file was automatically generated --- DO NOT EDIT */ /* Generated on Sun Nov 25 07:41:12 EST 2012 */ #include "codelet-rdft.h" #ifdef HAVE_FMA /* Generated by: ../../../genfft/gen_hc2hc.native -fma -reorder-insns -schedule-for-pipeline -compact -variables 4 -pipeline-latency 4 -sign 1 -n 6 -dif -name hb_6 -include hb.h */ /* * This function contains 46 FP additions, 32 FP multiplications, * (or, 24 additions, 10 multiplications, 22 fused multiply/add), * 45 stack variables, 2 constants, and 24 memory accesses */ #include "hb.h" static void hb_6(R *cr, R *ci, const R *W, stride rs, INT mb, INT me, INT ms) { DK(KP866025403, +0.866025403784438646763723170752936183471402627); DK(KP500000000, +0.500000000000000000000000000000000000000000000); { INT m; for (m = mb, W = W + ((mb - 1) * 10); m < me; m = m + 1, cr = cr + ms, ci = ci - ms, W = W + 10, MAKE_VOLATILE_STRIDE(12, rs)) { E TK, TR, TB, TM, TL, TS; { E Td, TN, TO, TJ, Tn, Tk, TC, T3, Tr, T4, T5, T7, T8; { E TH, Tg, Tj, TI, Th, Ti, T1, T2; { E Tb, Tc, Te, Tf; Tb = ci[WS(rs, 5)]; Tc = cr[WS(rs, 3)]; Te = ci[WS(rs, 3)]; Tf = cr[WS(rs, 5)]; Th = ci[WS(rs, 4)]; Td = Tb - Tc; TN = Tb + Tc; Ti = cr[WS(rs, 4)]; TH = Te + Tf; Tg = Te - Tf; } Tj = Th - Ti; TI = Th + Ti; T1 = cr[0]; T2 = ci[WS(rs, 2)]; TO = TH - TI; TJ = TH + TI; Tn = Tj - Tg; Tk = Tg + Tj; TC = T1 - T2; T3 = T1 + T2; Tr = FNMS(KP500000000, Tk, Td); T4 = cr[WS(rs, 2)]; T5 = ci[0]; T7 = ci[WS(rs, 1)]; T8 = cr[WS(rs, 1)]; } { E Tl, Tq, TQ, Ts, Ta, T10, TG; ci[0] = Td + Tk; { E T6, TD, T9, TE, TF; T6 = T4 + T5; TD = T4 - T5; T9 = T7 + T8; TE = T7 - T8; Tl = W[2]; Tq = W[3]; TQ = TD - TE; TF = TD + TE; Ts = T6 - T9; Ta = T6 + T9; T10 = TC + TF; TG = FNMS(KP500000000, TF, TC); } { E T13, TP, Tz, TZ, Tw, T14, Tv, Ty; { E Tt, T12, T11, Tp, Tm, To, Tu; T13 = TN + TO; TP = FNMS(KP500000000, TO, TN); cr[0] = T3 + Ta; Tm = FNMS(KP500000000, Ta, T3); Tz = FMA(KP866025403, Ts, Tr); Tt = FNMS(KP866025403, Ts, Tr); TZ = W[4]; To = FNMS(KP866025403, Tn, Tm); Tw = FMA(KP866025403, Tn, Tm); Tu = Tl * Tt; T12 = W[5]; T11 = TZ * T10; Tp = Tl * To; ci[WS(rs, 2)] = FMA(Tq, To, Tu); T14 = T12 * T10; cr[WS(rs, 3)] = FNMS(T12, T13, T11); cr[WS(rs, 2)] = FNMS(Tq, Tt, Tp); } ci[WS(rs, 3)] = FMA(TZ, T13, T14); Tv = W[6]; Ty = W[7]; { E TX, TT, TW, TV, TY, TU, TA, Tx; TK = FNMS(KP866025403, TJ, TG); TU = FMA(KP866025403, TJ, TG); TA = Tv * Tz; Tx = Tv * Tw; TX = FNMS(KP866025403, TQ, TP); TR = FMA(KP866025403, TQ, TP); ci[WS(rs, 4)] = FMA(Ty, Tw, TA); cr[WS(rs, 4)] = FNMS(Ty, Tz, Tx); TT = W[8]; TW = W[9]; TB = W[0]; TV = TT * TU; TY = TW * TU; TM = W[1]; TL = TB * TK; cr[WS(rs, 5)] = FNMS(TW, TX, TV); ci[WS(rs, 5)] = FMA(TT, TX, TY); } } } } cr[WS(rs, 1)] = FNMS(TM, TR, TL); TS = TM * TK; ci[WS(rs, 1)] = FMA(TB, TR, TS); } } } static const tw_instr twinstr[] = { {TW_FULL, 1, 6}, {TW_NEXT, 1, 0} }; static const hc2hc_desc desc = { 6, "hb_6", twinstr, &GENUS, {24, 10, 22, 0} }; void X(codelet_hb_6) (planner *p) { X(khc2hc_register) (p, hb_6, &desc); } #else /* HAVE_FMA */ /* Generated by: ../../../genfft/gen_hc2hc.native -compact -variables 4 -pipeline-latency 4 -sign 1 -n 6 -dif -name hb_6 -include hb.h */ /* * This function contains 46 FP additions, 28 FP multiplications, * (or, 32 additions, 14 multiplications, 14 fused multiply/add), * 27 stack variables, 2 constants, and 24 memory accesses */ #include "hb.h" static void hb_6(R *cr, R *ci, const R *W, stride rs, INT mb, INT me, INT ms) { DK(KP500000000, +0.500000000000000000000000000000000000000000000); DK(KP866025403, +0.866025403784438646763723170752936183471402627); { INT m; for (m = mb, W = W + ((mb - 1) * 10); m < me; m = m + 1, cr = cr + ms, ci = ci - ms, W = W + 10, MAKE_VOLATILE_STRIDE(12, rs)) { E T3, Ty, Ta, TO, Tr, TB, Td, TE, Tk, TL, Tn, TH; { E T1, T2, Tb, Tc; T1 = cr[0]; T2 = ci[WS(rs, 2)]; T3 = T1 + T2; Ty = T1 - T2; { E T6, Tz, T9, TA; { E T4, T5, T7, T8; T4 = cr[WS(rs, 2)]; T5 = ci[0]; T6 = T4 + T5; Tz = T4 - T5; T7 = ci[WS(rs, 1)]; T8 = cr[WS(rs, 1)]; T9 = T7 + T8; TA = T7 - T8; } Ta = T6 + T9; TO = KP866025403 * (Tz - TA); Tr = KP866025403 * (T6 - T9); TB = Tz + TA; } Tb = ci[WS(rs, 5)]; Tc = cr[WS(rs, 3)]; Td = Tb - Tc; TE = Tb + Tc; { E Tg, TG, Tj, TF; { E Te, Tf, Th, Ti; Te = ci[WS(rs, 3)]; Tf = cr[WS(rs, 5)]; Tg = Te - Tf; TG = Te + Tf; Th = ci[WS(rs, 4)]; Ti = cr[WS(rs, 4)]; Tj = Th - Ti; TF = Th + Ti; } Tk = Tg + Tj; TL = KP866025403 * (TG + TF); Tn = KP866025403 * (Tj - Tg); TH = TF - TG; } } cr[0] = T3 + Ta; ci[0] = Td + Tk; { E TC, TI, Tx, TD; TC = Ty + TB; TI = TE - TH; Tx = W[4]; TD = W[5]; cr[WS(rs, 3)] = FNMS(TD, TI, Tx * TC); ci[WS(rs, 3)] = FMA(TD, TC, Tx * TI); } { E To, Tu, Ts, Tw, Tm, Tq; Tm = FNMS(KP500000000, Ta, T3); To = Tm - Tn; Tu = Tm + Tn; Tq = FNMS(KP500000000, Tk, Td); Ts = Tq - Tr; Tw = Tr + Tq; { E Tl, Tp, Tt, Tv; Tl = W[2]; Tp = W[3]; cr[WS(rs, 2)] = FNMS(Tp, Ts, Tl * To); ci[WS(rs, 2)] = FMA(Tl, Ts, Tp * To); Tt = W[6]; Tv = W[7]; cr[WS(rs, 4)] = FNMS(Tv, Tw, Tt * Tu); ci[WS(rs, 4)] = FMA(Tt, Tw, Tv * Tu); } } { E TM, TS, TQ, TU, TK, TP; TK = FNMS(KP500000000, TB, Ty); TM = TK - TL; TS = TK + TL; TP = FMA(KP500000000, TH, TE); TQ = TO + TP; TU = TP - TO; { E TJ, TN, TR, TT; TJ = W[0]; TN = W[1]; cr[WS(rs, 1)] = FNMS(TN, TQ, TJ * TM); ci[WS(rs, 1)] = FMA(TN, TM, TJ * TQ); TR = W[8]; TT = W[9]; cr[WS(rs, 5)] = FNMS(TT, TU, TR * TS); ci[WS(rs, 5)] = FMA(TT, TS, TR * TU); } } } } } static const tw_instr twinstr[] = { {TW_FULL, 1, 6}, {TW_NEXT, 1, 0} }; static const hc2hc_desc desc = { 6, "hb_6", twinstr, &GENUS, {32, 14, 14, 0} }; void X(codelet_hb_6) (planner *p) { X(khc2hc_register) (p, hb_6, &desc); } #endif /* HAVE_FMA */ fftw-3.3.3/rdft/scalar/r2cb/hb_7.c0000644000175000017500000002501012054410751013423 00000000000000/* * Copyright (c) 2003, 2007-11 Matteo Frigo * Copyright (c) 2003, 2007-11 Massachusetts Institute of Technology * * This program is free software; you can redistribute it and/or modify * it under the terms of the GNU General Public License as published by * the Free Software Foundation; either version 2 of the License, or * (at your option) any later version. * * This program is distributed in the hope that it will be useful, * but WITHOUT ANY WARRANTY; without even the implied warranty of * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the * GNU General Public License for more details. * * You should have received a copy of the GNU General Public License * along with this program; if not, write to the Free Software * Foundation, Inc., 51 Franklin Street, Fifth Floor, Boston, MA 02110-1301 USA * */ /* This file was automatically generated --- DO NOT EDIT */ /* Generated on Sun Nov 25 07:41:12 EST 2012 */ #include "codelet-rdft.h" #ifdef HAVE_FMA /* Generated by: ../../../genfft/gen_hc2hc.native -fma -reorder-insns -schedule-for-pipeline -compact -variables 4 -pipeline-latency 4 -sign 1 -n 7 -dif -name hb_7 -include hb.h */ /* * This function contains 72 FP additions, 66 FP multiplications, * (or, 18 additions, 12 multiplications, 54 fused multiply/add), * 67 stack variables, 6 constants, and 28 memory accesses */ #include "hb.h" static void hb_7(R *cr, R *ci, const R *W, stride rs, INT mb, INT me, INT ms) { DK(KP974927912, +0.974927912181823607018131682993931217232785801); DK(KP900968867, +0.900968867902419126236102319507445051165919162); DK(KP801937735, +0.801937735804838252472204639014890102331838324); DK(KP692021471, +0.692021471630095869627814897002069140197260599); DK(KP356895867, +0.356895867892209443894399510021300583399127187); DK(KP554958132, +0.554958132087371191422194871006410481067288862); { INT m; for (m = mb, W = W + ((mb - 1) * 12); m < me; m = m + 1, cr = cr + ms, ci = ci - ms, W = W + 12, MAKE_VOLATILE_STRIDE(14, rs)) { E T1q, T1p, T1t, T1r, T1s, T1u; { E T1, T4, TC, T7, TB, Tt, TD, Ta, TA, T1l, TZ, T1b, Th, Tw, Td; E TP, Ti, Tj, Tl, Tm, T8, T9, T1a; T1 = cr[0]; { E T2, T3, T5, T6; T2 = cr[WS(rs, 1)]; T3 = ci[0]; T5 = cr[WS(rs, 2)]; T6 = ci[WS(rs, 1)]; T8 = cr[WS(rs, 3)]; T4 = T2 + T3; TC = T2 - T3; T7 = T5 + T6; TB = T5 - T6; T9 = ci[WS(rs, 2)]; } Tt = ci[WS(rs, 6)]; TD = FNMS(KP554958132, TC, TB); T1a = FNMS(KP356895867, T7, T4); Ta = T8 + T9; TA = T8 - T9; { E Tf, Tg, Tc, TO; Tf = ci[WS(rs, 3)]; Tg = cr[WS(rs, 4)]; T1l = FMA(KP554958132, TA, TC); TZ = FMA(KP554958132, TB, TA); Tc = FNMS(KP356895867, Ta, T7); TO = FNMS(KP356895867, T4, Ta); T1b = FNMS(KP692021471, T1a, Ta); Th = Tf + Tg; Tw = Tf - Tg; Td = FNMS(KP692021471, Tc, T4); TP = FNMS(KP692021471, TO, T7); } Ti = ci[WS(rs, 4)]; Tj = cr[WS(rs, 5)]; Tl = ci[WS(rs, 5)]; Tm = cr[WS(rs, 6)]; { E Ty, TS, TX, T1j, T1e, Tp, Tk, Tv; cr[0] = T1 + T4 + T7 + Ta; Tk = Ti + Tj; Tv = Ti - Tj; { E Tn, Tu, Tx, TR; Tn = Tl + Tm; Tu = Tl - Tm; Tx = FNMS(KP356895867, Tw, Tv); TR = FMA(KP554958132, Tk, Th); { E TW, T1i, T1d, To; TW = FNMS(KP356895867, Tu, Tw); T1i = FNMS(KP356895867, Tv, Tu); T1d = FMA(KP554958132, Th, Tn); To = FNMS(KP554958132, Tn, Tk); Ty = FNMS(KP692021471, Tx, Tu); TS = FNMS(KP801937735, TR, Tn); TX = FNMS(KP692021471, TW, Tv); T1j = FNMS(KP692021471, T1i, Tw); T1e = FMA(KP801937735, T1d, Tk); Tp = FNMS(KP801937735, To, Th); ci[0] = Tt + Tu + Tv + Tw; } } { E TL, TH, TK, TJ, TM, Te, Tz, TE; Te = FNMS(KP900968867, Td, T1); Tz = FNMS(KP900968867, Ty, Tt); TE = FNMS(KP801937735, TD, TA); { E Tb, TI, Tq, TF, Ts, Tr, TG; Tb = W[4]; TI = FMA(KP974927912, Tp, Te); Tq = FNMS(KP974927912, Tp, Te); TL = FNMS(KP974927912, TE, Tz); TF = FMA(KP974927912, TE, Tz); Ts = W[5]; Tr = Tb * Tq; TH = W[6]; TK = W[7]; TG = Ts * Tq; cr[WS(rs, 3)] = FNMS(Ts, TF, Tr); TJ = TH * TI; TM = TK * TI; ci[WS(rs, 3)] = FMA(Tb, TF, TG); } { E T14, T13, T17, T15, T16; { E TY, TT, T10, TQ; TQ = FNMS(KP900968867, TP, T1); cr[WS(rs, 4)] = FNMS(TK, TL, TJ); ci[WS(rs, 4)] = FMA(TH, TL, TM); TY = FNMS(KP900968867, TX, Tt); TT = FNMS(KP974927912, TS, TQ); T14 = FMA(KP974927912, TS, TQ); T10 = FNMS(KP801937735, TZ, TC); { E TN, TV, T11, TU, T12; TN = W[2]; TV = W[3]; T13 = W[8]; T11 = FMA(KP974927912, T10, TY); T17 = FNMS(KP974927912, T10, TY); TU = TN * TT; T12 = TV * TT; T15 = T13 * T14; T16 = W[9]; cr[WS(rs, 2)] = FNMS(TV, T11, TU); ci[WS(rs, 2)] = FMA(TN, T11, T12); } } { E T1k, T1f, T1m, T1c, T18; T1c = FNMS(KP900968867, T1b, T1); cr[WS(rs, 5)] = FNMS(T16, T17, T15); T18 = T16 * T14; T1k = FNMS(KP900968867, T1j, Tt); T1f = FNMS(KP974927912, T1e, T1c); T1q = FMA(KP974927912, T1e, T1c); ci[WS(rs, 5)] = FMA(T13, T17, T18); T1m = FMA(KP801937735, T1l, TB); { E T19, T1h, T1n, T1g, T1o; T19 = W[0]; T1h = W[1]; T1p = W[10]; T1t = FNMS(KP974927912, T1m, T1k); T1n = FMA(KP974927912, T1m, T1k); T1g = T19 * T1f; T1o = T1h * T1f; T1r = T1p * T1q; T1s = W[11]; cr[WS(rs, 1)] = FNMS(T1h, T1n, T1g); ci[WS(rs, 1)] = FMA(T19, T1n, T1o); } } } } } } cr[WS(rs, 6)] = FNMS(T1s, T1t, T1r); T1u = T1s * T1q; ci[WS(rs, 6)] = FMA(T1p, T1t, T1u); } } } static const tw_instr twinstr[] = { {TW_FULL, 1, 7}, {TW_NEXT, 1, 0} }; static const hc2hc_desc desc = { 7, "hb_7", twinstr, &GENUS, {18, 12, 54, 0} }; void X(codelet_hb_7) (planner *p) { X(khc2hc_register) (p, hb_7, &desc); } #else /* HAVE_FMA */ /* Generated by: ../../../genfft/gen_hc2hc.native -compact -variables 4 -pipeline-latency 4 -sign 1 -n 7 -dif -name hb_7 -include hb.h */ /* * This function contains 72 FP additions, 60 FP multiplications, * (or, 36 additions, 24 multiplications, 36 fused multiply/add), * 36 stack variables, 6 constants, and 28 memory accesses */ #include "hb.h" static void hb_7(R *cr, R *ci, const R *W, stride rs, INT mb, INT me, INT ms) { DK(KP222520933, +0.222520933956314404288902564496794759466355569); DK(KP900968867, +0.900968867902419126236102319507445051165919162); DK(KP623489801, +0.623489801858733530525004884004239810632274731); DK(KP781831482, +0.781831482468029808708444526674057750232334519); DK(KP974927912, +0.974927912181823607018131682993931217232785801); DK(KP433883739, +0.433883739117558120475768332848358754609990728); { INT m; for (m = mb, W = W + ((mb - 1) * 12); m < me; m = m + 1, cr = cr + ms, ci = ci - ms, W = W + 12, MAKE_VOLATILE_STRIDE(14, rs)) { E T1, T4, T7, Ta, Tx, TI, TV, TQ, TE, Tm, Tb, Te, Th, Tk, Tq; E TF, TR, TU, TJ, Tt; { E Tu, Tw, Tv, T2, T3; T1 = cr[0]; T2 = cr[WS(rs, 1)]; T3 = ci[0]; T4 = T2 + T3; Tu = T2 - T3; { E T5, T6, T8, T9; T5 = cr[WS(rs, 2)]; T6 = ci[WS(rs, 1)]; T7 = T5 + T6; Tw = T5 - T6; T8 = cr[WS(rs, 3)]; T9 = ci[WS(rs, 2)]; Ta = T8 + T9; Tv = T8 - T9; } Tx = FMA(KP433883739, Tu, KP974927912 * Tv) - (KP781831482 * Tw); TI = FMA(KP781831482, Tu, KP974927912 * Tw) + (KP433883739 * Tv); TV = FNMS(KP781831482, Tv, KP974927912 * Tu) - (KP433883739 * Tw); TQ = FMA(KP623489801, Ta, T1) + FNMA(KP900968867, T7, KP222520933 * T4); TE = FMA(KP623489801, T4, T1) + FNMA(KP900968867, Ta, KP222520933 * T7); Tm = FMA(KP623489801, T7, T1) + FNMA(KP222520933, Ta, KP900968867 * T4); } { E Tp, Tn, To, Tc, Td; Tb = ci[WS(rs, 6)]; Tc = ci[WS(rs, 5)]; Td = cr[WS(rs, 6)]; Te = Tc - Td; Tp = Tc + Td; { E Tf, Tg, Ti, Tj; Tf = ci[WS(rs, 4)]; Tg = cr[WS(rs, 5)]; Th = Tf - Tg; Tn = Tf + Tg; Ti = ci[WS(rs, 3)]; Tj = cr[WS(rs, 4)]; Tk = Ti - Tj; To = Ti + Tj; } Tq = FNMS(KP974927912, To, KP781831482 * Tn) - (KP433883739 * Tp); TF = FMA(KP781831482, Tp, KP974927912 * Tn) + (KP433883739 * To); TR = FMA(KP433883739, Tn, KP781831482 * To) - (KP974927912 * Tp); TU = FMA(KP623489801, Tk, Tb) + FNMA(KP900968867, Th, KP222520933 * Te); TJ = FMA(KP623489801, Te, Tb) + FNMA(KP900968867, Tk, KP222520933 * Th); Tt = FMA(KP623489801, Th, Tb) + FNMA(KP222520933, Tk, KP900968867 * Te); } cr[0] = T1 + T4 + T7 + Ta; ci[0] = Tb + Te + Th + Tk; { E Tr, Ty, Tl, Ts; Tr = Tm - Tq; Ty = Tt - Tx; Tl = W[6]; Ts = W[7]; cr[WS(rs, 4)] = FNMS(Ts, Ty, Tl * Tr); ci[WS(rs, 4)] = FMA(Tl, Ty, Ts * Tr); } { E TY, T10, TX, TZ; TY = TQ + TR; T10 = TV + TU; TX = W[2]; TZ = W[3]; cr[WS(rs, 2)] = FNMS(TZ, T10, TX * TY); ci[WS(rs, 2)] = FMA(TX, T10, TZ * TY); } { E TA, TC, Tz, TB; TA = Tm + Tq; TC = Tx + Tt; Tz = W[4]; TB = W[5]; cr[WS(rs, 3)] = FNMS(TB, TC, Tz * TA); ci[WS(rs, 3)] = FMA(Tz, TC, TB * TA); } { E TM, TO, TL, TN; TM = TE + TF; TO = TJ - TI; TL = W[10]; TN = W[11]; cr[WS(rs, 6)] = FNMS(TN, TO, TL * TM); ci[WS(rs, 6)] = FMA(TL, TO, TN * TM); } { E TS, TW, TP, TT; TS = TQ - TR; TW = TU - TV; TP = W[8]; TT = W[9]; cr[WS(rs, 5)] = FNMS(TT, TW, TP * TS); ci[WS(rs, 5)] = FMA(TP, TW, TT * TS); } { E TG, TK, TD, TH; TG = TE - TF; TK = TI + TJ; TD = W[0]; TH = W[1]; cr[WS(rs, 1)] = FNMS(TH, TK, TD * TG); ci[WS(rs, 1)] = FMA(TD, TK, TH * TG); } } } } static const tw_instr twinstr[] = { {TW_FULL, 1, 7}, {TW_NEXT, 1, 0} }; static const hc2hc_desc desc = { 7, "hb_7", twinstr, &GENUS, {36, 24, 36, 0} }; void X(codelet_hb_7) (planner *p) { X(khc2hc_register) (p, hb_7, &desc); } #endif /* HAVE_FMA */ fftw-3.3.3/rdft/scalar/r2cb/hb_8.c0000644000175000017500000002262012054410751013430 00000000000000/* * Copyright (c) 2003, 2007-11 Matteo Frigo * Copyright (c) 2003, 2007-11 Massachusetts Institute of Technology * * This program is free software; you can redistribute it and/or modify * it under the terms of the GNU General Public License as published by * the Free Software Foundation; either version 2 of the License, or * (at your option) any later version. * * This program is distributed in the hope that it will be useful, * but WITHOUT ANY WARRANTY; without even the implied warranty of * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the * GNU General Public License for more details. * * You should have received a copy of the GNU General Public License * along with this program; if not, write to the Free Software * Foundation, Inc., 51 Franklin Street, Fifth Floor, Boston, MA 02110-1301 USA * */ /* This file was automatically generated --- DO NOT EDIT */ /* Generated on Sun Nov 25 07:41:13 EST 2012 */ #include "codelet-rdft.h" #ifdef HAVE_FMA /* Generated by: ../../../genfft/gen_hc2hc.native -fma -reorder-insns -schedule-for-pipeline -compact -variables 4 -pipeline-latency 4 -sign 1 -n 8 -dif -name hb_8 -include hb.h */ /* * This function contains 66 FP additions, 36 FP multiplications, * (or, 44 additions, 14 multiplications, 22 fused multiply/add), * 52 stack variables, 1 constants, and 32 memory accesses */ #include "hb.h" static void hb_8(R *cr, R *ci, const R *W, stride rs, INT mb, INT me, INT ms) { DK(KP707106781, +0.707106781186547524400844362104849039284835938); { INT m; for (m = mb, W = W + ((mb - 1) * 14); m < me; m = m + 1, cr = cr + ms, ci = ci - ms, W = W + 14, MAKE_VOLATILE_STRIDE(16, rs)) { E Tw, TH, Tf, Ty, Tx, TI; { E TV, TD, T1i, T7, T1b, T1n, TQ, Tk, Tb, Tl, Ta, T1d, Tt, Tc, Tm; E Tn; { E T4, Tg, T3, T19, TC, T5, Th, Ti; { E T1, T2, TA, TB; T1 = cr[0]; T2 = ci[WS(rs, 3)]; TA = ci[WS(rs, 7)]; TB = cr[WS(rs, 4)]; T4 = cr[WS(rs, 2)]; Tg = T1 - T2; T3 = T1 + T2; T19 = TA - TB; TC = TA + TB; T5 = ci[WS(rs, 1)]; Th = ci[WS(rs, 5)]; Ti = cr[WS(rs, 6)]; } { E T8, T9, Tr, Ts; T8 = cr[WS(rs, 1)]; { E Tz, T6, T1a, Tj; Tz = T4 - T5; T6 = T4 + T5; T1a = Th - Ti; Tj = Th + Ti; TV = TC - Tz; TD = Tz + TC; T1i = T3 - T6; T7 = T3 + T6; T1b = T19 + T1a; T1n = T19 - T1a; TQ = Tg + Tj; Tk = Tg - Tj; T9 = ci[WS(rs, 2)]; } Tr = ci[WS(rs, 4)]; Ts = cr[WS(rs, 7)]; Tb = ci[0]; Tl = T8 - T9; Ta = T8 + T9; T1d = Tr - Ts; Tt = Tr + Ts; Tc = cr[WS(rs, 3)]; Tm = ci[WS(rs, 6)]; Tn = cr[WS(rs, 5)]; } } { E Te, T1e, Tv, TG, T13, T1k, T1s, T10, T1p, T1v, T1u, T1w, T1t; { E TP, T1o, T1j, TR, TU, TX, TW; TP = W[4]; { E Tq, Td, T1c, To; Tq = Tb - Tc; Td = Tb + Tc; T1c = Tm - Tn; To = Tm + Tn; { E Tu, TF, Tp, TE; Tu = Tq - Tt; TF = Tq + Tt; T1o = Ta - Td; Te = Ta + Td; T1j = T1d - T1c; T1e = T1c + T1d; Tp = Tl - To; TE = Tl + To; cr[0] = T7 + Te; ci[0] = T1b + T1e; TW = Tp - Tu; Tv = Tp + Tu; TR = TE + TF; TG = TE - TF; } } TU = W[5]; TX = FMA(KP707106781, TW, TV); T13 = FNMS(KP707106781, TW, TV); { E TS, TY, T1r, TT; T1k = T1i - T1j; T1s = T1i + T1j; TS = FNMS(KP707106781, TR, TQ); T10 = FMA(KP707106781, TR, TQ); T1p = T1n - T1o; T1v = T1o + T1n; TY = TP * TX; T1r = W[2]; TT = TP * TS; T1u = W[3]; ci[WS(rs, 3)] = FMA(TU, TS, TY); T1w = T1r * T1v; T1t = T1r * T1s; cr[WS(rs, 3)] = FNMS(TU, TX, TT); } } { E T1f, T15, T18, T17, T1g, T1h, T1m; { E TZ, T12, T16, T14, T11; ci[WS(rs, 2)] = FMA(T1u, T1s, T1w); cr[WS(rs, 2)] = FNMS(T1u, T1v, T1t); TZ = W[12]; T12 = W[13]; T1f = T1b - T1e; T16 = T7 - Te; T14 = TZ * T13; T11 = TZ * T10; T15 = W[6]; T18 = W[7]; ci[WS(rs, 7)] = FMA(T12, T10, T14); cr[WS(rs, 7)] = FNMS(T12, T13, T11); T17 = T15 * T16; T1g = T18 * T16; } cr[WS(rs, 4)] = FNMS(T18, T1f, T17); ci[WS(rs, 4)] = FMA(T15, T1f, T1g); T1h = W[10]; T1m = W[11]; { E TN, TJ, TM, TL, TO, TK, T1q, T1l; Tw = FNMS(KP707106781, Tv, Tk); TK = FMA(KP707106781, Tv, Tk); T1q = T1h * T1p; T1l = T1h * T1k; TN = FMA(KP707106781, TG, TD); TH = FNMS(KP707106781, TG, TD); ci[WS(rs, 6)] = FMA(T1m, T1k, T1q); cr[WS(rs, 6)] = FNMS(T1m, T1p, T1l); TJ = W[0]; TM = W[1]; Tf = W[8]; TL = TJ * TK; TO = TM * TK; Ty = W[9]; Tx = Tf * Tw; cr[WS(rs, 1)] = FNMS(TM, TN, TL); ci[WS(rs, 1)] = FMA(TJ, TN, TO); } } } } cr[WS(rs, 5)] = FNMS(Ty, TH, Tx); TI = Ty * Tw; ci[WS(rs, 5)] = FMA(Tf, TH, TI); } } } static const tw_instr twinstr[] = { {TW_FULL, 1, 8}, {TW_NEXT, 1, 0} }; static const hc2hc_desc desc = { 8, "hb_8", twinstr, &GENUS, {44, 14, 22, 0} }; void X(codelet_hb_8) (planner *p) { X(khc2hc_register) (p, hb_8, &desc); } #else /* HAVE_FMA */ /* Generated by: ../../../genfft/gen_hc2hc.native -compact -variables 4 -pipeline-latency 4 -sign 1 -n 8 -dif -name hb_8 -include hb.h */ /* * This function contains 66 FP additions, 32 FP multiplications, * (or, 52 additions, 18 multiplications, 14 fused multiply/add), * 30 stack variables, 1 constants, and 32 memory accesses */ #include "hb.h" static void hb_8(R *cr, R *ci, const R *W, stride rs, INT mb, INT me, INT ms) { DK(KP707106781, +0.707106781186547524400844362104849039284835938); { INT m; for (m = mb, W = W + ((mb - 1) * 14); m < me; m = m + 1, cr = cr + ms, ci = ci - ms, W = W + 14, MAKE_VOLATILE_STRIDE(16, rs)) { E T7, T18, T1c, To, Ty, TM, TY, TC, Te, TZ, T10, Tv, Tz, TP, TS; E TD; { E T3, TK, Tn, TL, T6, TW, Tk, TX; { E T1, T2, Tl, Tm; T1 = cr[0]; T2 = ci[WS(rs, 3)]; T3 = T1 + T2; TK = T1 - T2; Tl = ci[WS(rs, 5)]; Tm = cr[WS(rs, 6)]; Tn = Tl - Tm; TL = Tl + Tm; } { E T4, T5, Ti, Tj; T4 = cr[WS(rs, 2)]; T5 = ci[WS(rs, 1)]; T6 = T4 + T5; TW = T4 - T5; Ti = ci[WS(rs, 7)]; Tj = cr[WS(rs, 4)]; Tk = Ti - Tj; TX = Ti + Tj; } T7 = T3 + T6; T18 = TK + TL; T1c = TX - TW; To = Tk + Tn; Ty = T3 - T6; TM = TK - TL; TY = TW + TX; TC = Tk - Tn; } { E Ta, TN, Tu, TR, Td, TQ, Tr, TO; { E T8, T9, Ts, Tt; T8 = cr[WS(rs, 1)]; T9 = ci[WS(rs, 2)]; Ta = T8 + T9; TN = T8 - T9; Ts = ci[WS(rs, 4)]; Tt = cr[WS(rs, 7)]; Tu = Ts - Tt; TR = Ts + Tt; } { E Tb, Tc, Tp, Tq; Tb = ci[0]; Tc = cr[WS(rs, 3)]; Td = Tb + Tc; TQ = Tb - Tc; Tp = ci[WS(rs, 6)]; Tq = cr[WS(rs, 5)]; Tr = Tp - Tq; TO = Tp + Tq; } Te = Ta + Td; TZ = TN + TO; T10 = TQ + TR; Tv = Tr + Tu; Tz = Tu - Tr; TP = TN - TO; TS = TQ - TR; TD = Ta - Td; } cr[0] = T7 + Te; ci[0] = To + Tv; { E Tg, Tw, Tf, Th; Tg = T7 - Te; Tw = To - Tv; Tf = W[6]; Th = W[7]; cr[WS(rs, 4)] = FNMS(Th, Tw, Tf * Tg); ci[WS(rs, 4)] = FMA(Th, Tg, Tf * Tw); } { E TG, TI, TF, TH; TG = Ty + Tz; TI = TD + TC; TF = W[2]; TH = W[3]; cr[WS(rs, 2)] = FNMS(TH, TI, TF * TG); ci[WS(rs, 2)] = FMA(TF, TI, TH * TG); } { E TA, TE, Tx, TB; TA = Ty - Tz; TE = TC - TD; Tx = W[10]; TB = W[11]; cr[WS(rs, 6)] = FNMS(TB, TE, Tx * TA); ci[WS(rs, 6)] = FMA(Tx, TE, TB * TA); } { E T1a, T1g, T1e, T1i, T19, T1d; T19 = KP707106781 * (TZ + T10); T1a = T18 - T19; T1g = T18 + T19; T1d = KP707106781 * (TP - TS); T1e = T1c + T1d; T1i = T1c - T1d; { E T17, T1b, T1f, T1h; T17 = W[4]; T1b = W[5]; cr[WS(rs, 3)] = FNMS(T1b, T1e, T17 * T1a); ci[WS(rs, 3)] = FMA(T17, T1e, T1b * T1a); T1f = W[12]; T1h = W[13]; cr[WS(rs, 7)] = FNMS(T1h, T1i, T1f * T1g); ci[WS(rs, 7)] = FMA(T1f, T1i, T1h * T1g); } } { E TU, T14, T12, T16, TT, T11; TT = KP707106781 * (TP + TS); TU = TM - TT; T14 = TM + TT; T11 = KP707106781 * (TZ - T10); T12 = TY - T11; T16 = TY + T11; { E TJ, TV, T13, T15; TJ = W[8]; TV = W[9]; cr[WS(rs, 5)] = FNMS(TV, T12, TJ * TU); ci[WS(rs, 5)] = FMA(TV, TU, TJ * T12); T13 = W[0]; T15 = W[1]; cr[WS(rs, 1)] = FNMS(T15, T16, T13 * T14); ci[WS(rs, 1)] = FMA(T15, T14, T13 * T16); } } } } } static const tw_instr twinstr[] = { {TW_FULL, 1, 8}, {TW_NEXT, 1, 0} }; static const hc2hc_desc desc = { 8, "hb_8", twinstr, &GENUS, {52, 18, 14, 0} }; void X(codelet_hb_8) (planner *p) { X(khc2hc_register) (p, hb_8, &desc); } #endif /* HAVE_FMA */ fftw-3.3.3/rdft/scalar/r2cb/hb_9.c0000644000175000017500000003413312054410752013434 00000000000000/* * Copyright (c) 2003, 2007-11 Matteo Frigo * Copyright (c) 2003, 2007-11 Massachusetts Institute of Technology * * This program is free software; you can redistribute it and/or modify * it under the terms of the GNU General Public License as published by * the Free Software Foundation; either version 2 of the License, or * (at your option) any later version. * * This program is distributed in the hope that it will be useful, * but WITHOUT ANY WARRANTY; without even the implied warranty of * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the * GNU General Public License for more details. * * You should have received a copy of the GNU General Public License * along with this program; if not, write to the Free Software * Foundation, Inc., 51 Franklin Street, Fifth Floor, Boston, MA 02110-1301 USA * */ /* This file was automatically generated --- DO NOT EDIT */ /* Generated on Sun Nov 25 07:41:13 EST 2012 */ #include "codelet-rdft.h" #ifdef HAVE_FMA /* Generated by: ../../../genfft/gen_hc2hc.native -fma -reorder-insns -schedule-for-pipeline -compact -variables 4 -pipeline-latency 4 -sign 1 -n 9 -dif -name hb_9 -include hb.h */ /* * This function contains 96 FP additions, 88 FP multiplications, * (or, 24 additions, 16 multiplications, 72 fused multiply/add), * 69 stack variables, 10 constants, and 36 memory accesses */ #include "hb.h" static void hb_9(R *cr, R *ci, const R *W, stride rs, INT mb, INT me, INT ms) { DK(KP954188894, +0.954188894138671133499268364187245676532219158); DK(KP852868531, +0.852868531952443209628250963940074071936020296); DK(KP492403876, +0.492403876506104029683371512294761506835321626); DK(KP984807753, +0.984807753012208059366743024589523013670643252); DK(KP777861913, +0.777861913430206160028177977318626690410586096); DK(KP839099631, +0.839099631177280011763127298123181364687434283); DK(KP363970234, +0.363970234266202361351047882776834043890471784); DK(KP176326980, +0.176326980708464973471090386868618986121633062); DK(KP866025403, +0.866025403784438646763723170752936183471402627); DK(KP500000000, +0.500000000000000000000000000000000000000000000); { INT m; for (m = mb, W = W + ((mb - 1) * 16); m < me; m = m + 1, cr = cr + ms, ci = ci - ms, W = W + 16, MAKE_VOLATILE_STRIDE(18, rs)) { E T1X, T1S, T1U, T1P, T1Y, T1T; { E T5, Tl, TQ, T1y, T1b, T1J, Tg, TE, TW, T13, T10, Tz, Tw, TT, T1K; E T1B, T1L, T1E; { E T1, Th, T2, T3, Ti, Tj; T1 = cr[0]; Th = ci[WS(rs, 8)]; T2 = cr[WS(rs, 3)]; T3 = ci[WS(rs, 2)]; Ti = ci[WS(rs, 5)]; Tj = cr[WS(rs, 6)]; { E T12, Tb, TZ, TY, Ta, Tq, T11, Tr, Ts, TS, Te, Tt; { E T6, Tm, Tn, To, T9, Tc, Td, Tp; { E T7, T8, T1a, T4; T6 = cr[WS(rs, 1)]; T1a = T2 - T3; T4 = T2 + T3; { E TP, Tk, TO, T19; TP = Ti + Tj; Tk = Ti - Tj; T7 = cr[WS(rs, 4)]; T5 = T1 + T4; TO = FNMS(KP500000000, T4, T1); Tl = Th + Tk; T19 = FNMS(KP500000000, Tk, Th); TQ = FNMS(KP866025403, TP, TO); T1y = FMA(KP866025403, TP, TO); T1b = FMA(KP866025403, T1a, T19); T1J = FNMS(KP866025403, T1a, T19); T8 = ci[WS(rs, 1)]; } Tm = ci[WS(rs, 7)]; Tn = ci[WS(rs, 4)]; To = cr[WS(rs, 7)]; T9 = T7 + T8; T12 = T7 - T8; } Tb = cr[WS(rs, 2)]; TZ = Tn + To; Tp = Tn - To; TY = FNMS(KP500000000, T9, T6); Ta = T6 + T9; Tc = ci[WS(rs, 3)]; Td = ci[0]; Tq = Tm + Tp; T11 = FMS(KP500000000, Tp, Tm); Tr = ci[WS(rs, 6)]; Ts = cr[WS(rs, 5)]; TS = Td - Tc; Te = Tc + Td; Tt = cr[WS(rs, 8)]; } { E T1C, Tv, TR, T1D, T1z, T1A; { E TU, Tu, TV, Tf; TU = FNMS(KP500000000, Te, Tb); Tf = Tb + Te; Tu = Ts + Tt; TV = Ts - Tt; Tg = Ta + Tf; TE = Ta - Tf; TW = FMA(KP866025403, TV, TU); T1C = FNMS(KP866025403, TV, TU); Tv = Tr - Tu; TR = FMA(KP500000000, Tu, Tr); } T1z = FMA(KP866025403, T12, T11); T13 = FNMS(KP866025403, T12, T11); T10 = FNMS(KP866025403, TZ, TY); T1A = FMA(KP866025403, TZ, TY); Tz = Tv - Tq; Tw = Tq + Tv; T1D = FMA(KP866025403, TS, TR); TT = FNMS(KP866025403, TS, TR); T1K = FNMS(KP176326980, T1z, T1A); T1B = FMA(KP176326980, T1A, T1z); T1L = FNMS(KP363970234, T1C, T1D); T1E = FMA(KP363970234, T1D, T1C); } } } { E T1d, T14, T1c, TX; cr[0] = T5 + Tg; T1d = FNMS(KP839099631, T10, T13); T14 = FMA(KP839099631, T13, T10); T1c = FMA(KP176326980, TT, TW); TX = FNMS(KP176326980, TW, TT); ci[0] = Tl + Tw; { E TL, TK, TJ, Ty, TD; Ty = FNMS(KP500000000, Tg, T5); TD = FNMS(KP500000000, Tw, Tl); { E Tx, TC, TA, TI, TF; Tx = W[10]; TC = W[11]; TA = FNMS(KP866025403, Tz, Ty); TI = FMA(KP866025403, Tz, Ty); TF = FNMS(KP866025403, TE, TD); TL = FMA(KP866025403, TE, TD); { E TH, TB, TG, TM; TH = W[4]; TB = Tx * TA; TK = W[5]; TG = Tx * TF; TM = TH * TL; TJ = TH * TI; cr[WS(rs, 6)] = FNMS(TC, TF, TB); ci[WS(rs, 6)] = FMA(TC, TA, TG); ci[WS(rs, 3)] = FMA(TK, TI, TM); } } cr[WS(rs, 3)] = FNMS(TK, TL, TJ); { E T1k, T1p, T1l, T1q, T1m; { E T1e, T1j, T15, T1o; T1e = FNMS(KP777861913, T1d, T1c); T1j = FMA(KP777861913, T1d, T1c); T15 = FNMS(KP777861913, T14, TX); T1o = FMA(KP777861913, T14, TX); { E TN, T16, T1f, T17, T1s, T1v, T18, T1i, T1n, T1r, T1u; TN = W[0]; T16 = FNMS(KP984807753, T15, TQ); T1i = FMA(KP492403876, T15, TQ); T1f = FMA(KP984807753, T1e, T1b); T1n = FNMS(KP492403876, T1e, T1b); T17 = TN * T16; T1s = FMA(KP852868531, T1j, T1i); T1k = FNMS(KP852868531, T1j, T1i); T1v = FMA(KP852868531, T1o, T1n); T1p = FNMS(KP852868531, T1o, T1n); T18 = W[1]; T1r = W[6]; T1u = W[7]; { E T1h, T1g, T1w, T1t; T1h = W[12]; cr[WS(rs, 1)] = FNMS(T18, T1f, T17); T1g = T18 * T16; T1w = T1r * T1v; T1t = T1r * T1s; T1l = T1h * T1k; ci[WS(rs, 1)] = FMA(TN, T1f, T1g); ci[WS(rs, 4)] = FMA(T1u, T1s, T1w); cr[WS(rs, 4)] = FNMS(T1u, T1v, T1t); T1q = T1h * T1p; } T1m = W[13]; } } { E T1F, T1W, T1R, T1V, T1N, T1M, T1x, T1I; T1F = FNMS(KP954188894, T1E, T1B); T1W = FMA(KP954188894, T1E, T1B); T1M = FNMS(KP954188894, T1L, T1K); T1R = FMA(KP954188894, T1L, T1K); ci[WS(rs, 7)] = FMA(T1m, T1k, T1q); cr[WS(rs, 7)] = FNMS(T1m, T1p, T1l); T1V = FNMS(KP492403876, T1M, T1J); T1N = FMA(KP984807753, T1M, T1J); T1x = W[2]; T1I = W[3]; { E T23, T22, T20, T1Z, T24, T21; T1X = FMA(KP852868531, T1W, T1V); T23 = FNMS(KP852868531, T1W, T1V); { E T1G, T1Q, T1O, T1H; T1G = FMA(KP984807753, T1F, T1y); T1Q = FNMS(KP492403876, T1F, T1y); T1O = T1x * T1N; T22 = W[15]; T1H = T1x * T1G; T20 = FMA(KP852868531, T1R, T1Q); T1S = FNMS(KP852868531, T1R, T1Q); ci[WS(rs, 2)] = FMA(T1I, T1G, T1O); cr[WS(rs, 2)] = FNMS(T1I, T1N, T1H); T1Z = W[14]; T24 = T22 * T20; } T1U = W[9]; T21 = T1Z * T20; ci[WS(rs, 8)] = FMA(T1Z, T23, T24); T1P = W[8]; T1Y = T1U * T1S; cr[WS(rs, 8)] = FNMS(T22, T23, T21); } } } } } } T1T = T1P * T1S; ci[WS(rs, 5)] = FMA(T1P, T1X, T1Y); cr[WS(rs, 5)] = FNMS(T1U, T1X, T1T); } } } static const tw_instr twinstr[] = { {TW_FULL, 1, 9}, {TW_NEXT, 1, 0} }; static const hc2hc_desc desc = { 9, "hb_9", twinstr, &GENUS, {24, 16, 72, 0} }; void X(codelet_hb_9) (planner *p) { X(khc2hc_register) (p, hb_9, &desc); } #else /* HAVE_FMA */ /* Generated by: ../../../genfft/gen_hc2hc.native -compact -variables 4 -pipeline-latency 4 -sign 1 -n 9 -dif -name hb_9 -include hb.h */ /* * This function contains 96 FP additions, 72 FP multiplications, * (or, 60 additions, 36 multiplications, 36 fused multiply/add), * 53 stack variables, 8 constants, and 36 memory accesses */ #include "hb.h" static void hb_9(R *cr, R *ci, const R *W, stride rs, INT mb, INT me, INT ms) { DK(KP984807753, +0.984807753012208059366743024589523013670643252); DK(KP173648177, +0.173648177666930348851716626769314796000375677); DK(KP342020143, +0.342020143325668733044099614682259580763083368); DK(KP939692620, +0.939692620785908384054109277324731469936208134); DK(KP642787609, +0.642787609686539326322643409907263432907559884); DK(KP766044443, +0.766044443118978035202392650555416673935832457); DK(KP500000000, +0.500000000000000000000000000000000000000000000); DK(KP866025403, +0.866025403784438646763723170752936183471402627); { INT m; for (m = mb, W = W + ((mb - 1) * 16); m < me; m = m + 1, cr = cr + ms, ci = ci - ms, W = W + 16, MAKE_VOLATILE_STRIDE(18, rs)) { E T5, Tl, TM, T1o, T16, T1y, Ta, Tf, Tg, Tq, Tv, Tw, TT, T17, T1u; E T1A, T1r, T1z, T10, T18; { E T1, Th, T4, T14, Tk, TL, TK, T15; T1 = cr[0]; Th = ci[WS(rs, 8)]; { E T2, T3, Ti, Tj; T2 = cr[WS(rs, 3)]; T3 = ci[WS(rs, 2)]; T4 = T2 + T3; T14 = KP866025403 * (T2 - T3); Ti = ci[WS(rs, 5)]; Tj = cr[WS(rs, 6)]; Tk = Ti - Tj; TL = KP866025403 * (Ti + Tj); } T5 = T1 + T4; Tl = Th + Tk; TK = FNMS(KP500000000, T4, T1); TM = TK - TL; T1o = TK + TL; T15 = FNMS(KP500000000, Tk, Th); T16 = T14 + T15; T1y = T15 - T14; } { E T6, T9, TN, TQ, Tm, Tp, TO, TR, Tb, Te, TU, TX, Tr, Tu, TV; E TY; { E T7, T8, Tn, To; T6 = cr[WS(rs, 1)]; T7 = cr[WS(rs, 4)]; T8 = ci[WS(rs, 1)]; T9 = T7 + T8; TN = FNMS(KP500000000, T9, T6); TQ = KP866025403 * (T7 - T8); Tm = ci[WS(rs, 7)]; Tn = ci[WS(rs, 4)]; To = cr[WS(rs, 7)]; Tp = Tn - To; TO = KP866025403 * (Tn + To); TR = FNMS(KP500000000, Tp, Tm); } { E Tc, Td, Ts, Tt; Tb = cr[WS(rs, 2)]; Tc = ci[WS(rs, 3)]; Td = ci[0]; Te = Tc + Td; TU = FNMS(KP500000000, Te, Tb); TX = KP866025403 * (Tc - Td); Tr = ci[WS(rs, 6)]; Ts = cr[WS(rs, 5)]; Tt = cr[WS(rs, 8)]; Tu = Ts + Tt; TV = KP866025403 * (Ts - Tt); TY = FMA(KP500000000, Tu, Tr); } { E TP, TS, T1s, T1t; Ta = T6 + T9; Tf = Tb + Te; Tg = Ta + Tf; Tq = Tm + Tp; Tv = Tr - Tu; Tw = Tq + Tv; TP = TN - TO; TS = TQ + TR; TT = FNMS(KP642787609, TS, KP766044443 * TP); T17 = FMA(KP766044443, TS, KP642787609 * TP); T1s = TU - TV; T1t = TY - TX; T1u = FMA(KP939692620, T1s, KP342020143 * T1t); T1A = FNMS(KP939692620, T1t, KP342020143 * T1s); { E T1p, T1q, TW, TZ; T1p = TN + TO; T1q = TR - TQ; T1r = FNMS(KP984807753, T1q, KP173648177 * T1p); T1z = FMA(KP173648177, T1q, KP984807753 * T1p); TW = TU + TV; TZ = TX + TY; T10 = FNMS(KP984807753, TZ, KP173648177 * TW); T18 = FMA(KP984807753, TW, KP173648177 * TZ); } } } cr[0] = T5 + Tg; ci[0] = Tl + Tw; { E TA, TG, TE, TI; { E Ty, Tz, TC, TD; Ty = FNMS(KP500000000, Tg, T5); Tz = KP866025403 * (Tv - Tq); TA = Ty - Tz; TG = Ty + Tz; TC = FNMS(KP500000000, Tw, Tl); TD = KP866025403 * (Ta - Tf); TE = TC - TD; TI = TD + TC; } { E Tx, TB, TF, TH; Tx = W[10]; TB = W[11]; cr[WS(rs, 6)] = FNMS(TB, TE, Tx * TA); ci[WS(rs, 6)] = FMA(Tx, TE, TB * TA); TF = W[4]; TH = W[5]; cr[WS(rs, 3)] = FNMS(TH, TI, TF * TG); ci[WS(rs, 3)] = FMA(TF, TI, TH * TG); } } { E T1d, T1h, T12, T1c, T1a, T1g, T11, T19, TJ, T13; T1d = KP866025403 * (T18 - T17); T1h = KP866025403 * (TT - T10); T11 = TT + T10; T12 = TM + T11; T1c = FNMS(KP500000000, T11, TM); T19 = T17 + T18; T1a = T16 + T19; T1g = FNMS(KP500000000, T19, T16); TJ = W[0]; T13 = W[1]; cr[WS(rs, 1)] = FNMS(T13, T1a, TJ * T12); ci[WS(rs, 1)] = FMA(T13, T12, TJ * T1a); { E T1k, T1m, T1j, T1l; T1k = T1c + T1d; T1m = T1h + T1g; T1j = W[6]; T1l = W[7]; cr[WS(rs, 4)] = FNMS(T1l, T1m, T1j * T1k); ci[WS(rs, 4)] = FMA(T1j, T1m, T1l * T1k); } { E T1e, T1i, T1b, T1f; T1e = T1c - T1d; T1i = T1g - T1h; T1b = W[12]; T1f = W[13]; cr[WS(rs, 7)] = FNMS(T1f, T1i, T1b * T1e); ci[WS(rs, 7)] = FMA(T1b, T1i, T1f * T1e); } } { E T1F, T1J, T1w, T1E, T1C, T1I, T1v, T1B, T1n, T1x; T1F = KP866025403 * (T1A - T1z); T1J = KP866025403 * (T1r + T1u); T1v = T1r - T1u; T1w = T1o + T1v; T1E = FNMS(KP500000000, T1v, T1o); T1B = T1z + T1A; T1C = T1y + T1B; T1I = FNMS(KP500000000, T1B, T1y); T1n = W[2]; T1x = W[3]; cr[WS(rs, 2)] = FNMS(T1x, T1C, T1n * T1w); ci[WS(rs, 2)] = FMA(T1n, T1C, T1x * T1w); { E T1M, T1O, T1L, T1N; T1M = T1F + T1E; T1O = T1I + T1J; T1L = W[8]; T1N = W[9]; cr[WS(rs, 5)] = FNMS(T1N, T1O, T1L * T1M); ci[WS(rs, 5)] = FMA(T1N, T1M, T1L * T1O); } { E T1G, T1K, T1D, T1H; T1G = T1E - T1F; T1K = T1I - T1J; T1D = W[14]; T1H = W[15]; cr[WS(rs, 8)] = FNMS(T1H, T1K, T1D * T1G); ci[WS(rs, 8)] = FMA(T1H, T1G, T1D * T1K); } } } } } static const tw_instr twinstr[] = { {TW_FULL, 1, 9}, {TW_NEXT, 1, 0} }; static const hc2hc_desc desc = { 9, "hb_9", twinstr, &GENUS, {60, 36, 36, 0} }; void X(codelet_hb_9) (planner *p) { X(khc2hc_register) (p, hb_9, &desc); } #endif /* HAVE_FMA */ fftw-3.3.3/rdft/scalar/r2cb/hb_10.c0000644000175000017500000003307712054410752013512 00000000000000/* * Copyright (c) 2003, 2007-11 Matteo Frigo * Copyright (c) 2003, 2007-11 Massachusetts Institute of Technology * * This program is free software; you can redistribute it and/or modify * it under the terms of the GNU General Public License as published by * the Free Software Foundation; either version 2 of the License, or * (at your option) any later version. * * This program is distributed in the hope that it will be useful, * but WITHOUT ANY WARRANTY; without even the implied warranty of * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the * GNU General Public License for more details. * * You should have received a copy of the GNU General Public License * along with this program; if not, write to the Free Software * Foundation, Inc., 51 Franklin Street, Fifth Floor, Boston, MA 02110-1301 USA * */ /* This file was automatically generated --- DO NOT EDIT */ /* Generated on Sun Nov 25 07:41:13 EST 2012 */ #include "codelet-rdft.h" #ifdef HAVE_FMA /* Generated by: ../../../genfft/gen_hc2hc.native -fma -reorder-insns -schedule-for-pipeline -compact -variables 4 -pipeline-latency 4 -sign 1 -n 10 -dif -name hb_10 -include hb.h */ /* * This function contains 102 FP additions, 72 FP multiplications, * (or, 48 additions, 18 multiplications, 54 fused multiply/add), * 71 stack variables, 4 constants, and 40 memory accesses */ #include "hb.h" static void hb_10(R *cr, R *ci, const R *W, stride rs, INT mb, INT me, INT ms) { DK(KP951056516, +0.951056516295153572116439333379382143405698634); DK(KP559016994, +0.559016994374947424102293417182819058860154590); DK(KP250000000, +0.250000000000000000000000000000000000000000000); DK(KP618033988, +0.618033988749894848204586834365638117720309180); { INT m; for (m = mb, W = W + ((mb - 1) * 18); m < me; m = m + 1, cr = cr + ms, ci = ci - ms, W = W + 18, MAKE_VOLATILE_STRIDE(20, rs)) { E T21, T1Y, T1X; { E T1B, TH, T1g, T3, T1V, T1x, T1G, T1E, TM, TK, T11, TB, T7, T1m, T1J; E TO, Th, T1h, T6, T8, TF, TG, T1i, T9; TF = ci[WS(rs, 9)]; TG = cr[WS(rs, 5)]; { E T1u, Tp, Tu, T1s, Tz, T1v, Ts, Tv; { E Tx, Ty, Tn, To, Tq, Tr; Tn = ci[WS(rs, 5)]; To = cr[WS(rs, 9)]; Tx = ci[WS(rs, 6)]; T1B = TF + TG; TH = TF - TG; T1u = Tn + To; Tp = Tn - To; Ty = cr[WS(rs, 8)]; Tq = ci[WS(rs, 8)]; Tr = cr[WS(rs, 6)]; Tu = ci[WS(rs, 7)]; T1s = Tx + Ty; Tz = Tx - Ty; T1v = Tq + Tr; Ts = Tq - Tr; Tv = cr[WS(rs, 7)]; } { E T1, T1w, T1D, TJ, Tt, T1r, Tw, T2; T1 = cr[0]; T1w = T1u + T1v; T1D = T1u - T1v; TJ = Tp + Ts; Tt = Tp - Ts; T1r = Tu + Tv; Tw = Tu - Tv; T2 = ci[WS(rs, 4)]; { E Tb, Tc, Te, Tf; Tb = cr[WS(rs, 4)]; { E T1t, T1C, TI, TA; T1t = T1r + T1s; T1C = T1r - T1s; TI = Tw + Tz; TA = Tw - Tz; T1g = T1 - T2; T3 = T1 + T2; T1V = FNMS(KP618033988, T1t, T1w); T1x = FMA(KP618033988, T1w, T1t); T1G = T1C - T1D; T1E = T1C + T1D; TM = TI - TJ; TK = TI + TJ; T11 = FMA(KP618033988, Tt, TA); TB = FNMS(KP618033988, TA, Tt); Tc = ci[0]; } Te = ci[WS(rs, 3)]; Tf = cr[WS(rs, 1)]; { E T4, T1k, Td, T1l, Tg, T5; T4 = cr[WS(rs, 2)]; T1k = Tb - Tc; Td = Tb + Tc; T1l = Te - Tf; Tg = Te + Tf; T5 = ci[WS(rs, 2)]; T7 = ci[WS(rs, 1)]; T1m = T1k + T1l; T1J = T1k - T1l; TO = Td - Tg; Th = Td + Tg; T1h = T4 - T5; T6 = T4 + T5; T8 = cr[WS(rs, 3)]; } } } } ci[0] = TH + TK; T1i = T7 - T8; T9 = T7 + T8; { E T2d, T1F, T29, T1I, TP, T2c, T1p, Tl, T1o, Tk, T2b, T2e, T17, T14, T13; T2d = T1B + T1E; T1F = FNMS(KP250000000, T1E, T1B); { E T1j, Ta, T1n, Ti, T2a; T29 = W[8]; T1I = T1h - T1i; T1j = T1h + T1i; TP = T6 - T9; Ta = T6 + T9; T2c = W[9]; T1p = T1j - T1m; T1n = T1j + T1m; Tl = Ta - Th; Ti = Ta + Th; T1o = FNMS(KP250000000, T1n, T1g); T2a = T1g + T1n; cr[0] = T3 + Ti; Tk = FNMS(KP250000000, Ti, T3); T2b = T29 * T2a; T2e = T2c * T2a; } { E T16, TQ, T10, Tm, TL; T16 = FMA(KP618033988, TO, TP); TQ = FNMS(KP618033988, TP, TO); cr[WS(rs, 5)] = FNMS(T2c, T2d, T2b); ci[WS(rs, 5)] = FMA(T29, T2d, T2e); T10 = FMA(KP559016994, Tl, Tk); Tm = FNMS(KP559016994, Tl, Tk); TL = FNMS(KP250000000, TK, TH); { E TE, TU, T12, TR, TX, T1d, T1c, T19, TD, T1e, T1b, TW, TT; { E TC, T15, T1a, TS, Tj, TN; TE = W[3]; TC = FMA(KP951056516, TB, Tm); TU = FNMS(KP951056516, TB, Tm); TN = FNMS(KP559016994, TM, TL); T15 = FMA(KP559016994, TM, TL); T12 = FMA(KP951056516, T11, T10); T1a = FNMS(KP951056516, T11, T10); TS = TE * TC; TR = FNMS(KP951056516, TQ, TN); TX = FMA(KP951056516, TQ, TN); Tj = W[2]; T1d = FMA(KP951056516, T16, T15); T17 = FNMS(KP951056516, T16, T15); T1c = W[11]; T19 = W[10]; ci[WS(rs, 2)] = FMA(Tj, TR, TS); TD = Tj * TC; T1e = T1c * T1a; T1b = T19 * T1a; } cr[WS(rs, 2)] = FNMS(TE, TR, TD); ci[WS(rs, 6)] = FMA(T19, T1d, T1e); cr[WS(rs, 6)] = FNMS(T1c, T1d, T1b); TW = W[15]; TT = W[14]; { E TZ, T18, TY, TV; T14 = W[7]; TY = TW * TU; TV = TT * TU; TZ = W[6]; T18 = T14 * T12; ci[WS(rs, 8)] = FMA(TT, TX, TY); cr[WS(rs, 8)] = FNMS(TW, TX, TV); T13 = TZ * T12; ci[WS(rs, 4)] = FMA(TZ, T17, T18); } } } { E T20, T1K, T1q, T1U; T20 = FNMS(KP618033988, T1I, T1J); T1K = FMA(KP618033988, T1J, T1I); cr[WS(rs, 4)] = FNMS(T14, T17, T13); T1q = FMA(KP559016994, T1p, T1o); T1U = FNMS(KP559016994, T1p, T1o); { E T1A, T1O, T1W, T1R, T1L, T27, T26, T23, T1z, T28, T25, T1Q, T1N; { E T1y, T1Z, T24, T1M, T1f, T1H; T1A = W[1]; T1O = FMA(KP951056516, T1x, T1q); T1y = FNMS(KP951056516, T1x, T1q); T1Z = FNMS(KP559016994, T1G, T1F); T1H = FMA(KP559016994, T1G, T1F); T24 = FMA(KP951056516, T1V, T1U); T1W = FNMS(KP951056516, T1V, T1U); T1M = T1A * T1y; T1R = FNMS(KP951056516, T1K, T1H); T1L = FMA(KP951056516, T1K, T1H); T1f = W[0]; T21 = FMA(KP951056516, T20, T1Z); T27 = FNMS(KP951056516, T20, T1Z); T26 = W[13]; T23 = W[12]; ci[WS(rs, 1)] = FMA(T1f, T1L, T1M); T1z = T1f * T1y; T28 = T26 * T24; T25 = T23 * T24; } cr[WS(rs, 1)] = FNMS(T1A, T1L, T1z); ci[WS(rs, 7)] = FMA(T23, T27, T28); cr[WS(rs, 7)] = FNMS(T26, T27, T25); T1Q = W[17]; T1N = W[16]; { E T1T, T22, T1S, T1P; T1Y = W[5]; T1S = T1Q * T1O; T1P = T1N * T1O; T1T = W[4]; T22 = T1Y * T1W; ci[WS(rs, 9)] = FMA(T1N, T1R, T1S); cr[WS(rs, 9)] = FNMS(T1Q, T1R, T1P); T1X = T1T * T1W; ci[WS(rs, 3)] = FMA(T1T, T21, T22); } } } } } cr[WS(rs, 3)] = FNMS(T1Y, T21, T1X); } } } static const tw_instr twinstr[] = { {TW_FULL, 1, 10}, {TW_NEXT, 1, 0} }; static const hc2hc_desc desc = { 10, "hb_10", twinstr, &GENUS, {48, 18, 54, 0} }; void X(codelet_hb_10) (planner *p) { X(khc2hc_register) (p, hb_10, &desc); } #else /* HAVE_FMA */ /* Generated by: ../../../genfft/gen_hc2hc.native -compact -variables 4 -pipeline-latency 4 -sign 1 -n 10 -dif -name hb_10 -include hb.h */ /* * This function contains 102 FP additions, 60 FP multiplications, * (or, 72 additions, 30 multiplications, 30 fused multiply/add), * 41 stack variables, 4 constants, and 40 memory accesses */ #include "hb.h" static void hb_10(R *cr, R *ci, const R *W, stride rs, INT mb, INT me, INT ms) { DK(KP250000000, +0.250000000000000000000000000000000000000000000); DK(KP951056516, +0.951056516295153572116439333379382143405698634); DK(KP587785252, +0.587785252292473129168705954639072768597652438); DK(KP559016994, +0.559016994374947424102293417182819058860154590); { INT m; for (m = mb, W = W + ((mb - 1) * 18); m < me; m = m + 1, cr = cr + ms, ci = ci - ms, W = W + 18, MAKE_VOLATILE_STRIDE(20, rs)) { E T3, T18, TE, TF, T1B, T1A, T1f, T1t, Ti, Tl, TJ, T1i, Tt, TA, T1w; E T1v, T1p, T1E, TM, TO; { E T1, T2, TH, TI; T1 = cr[0]; T2 = ci[WS(rs, 4)]; T3 = T1 + T2; T18 = T1 - T2; { E T6, T19, Tg, T1d, T9, T1a, Td, T1c; { E T4, T5, Te, Tf; T4 = cr[WS(rs, 2)]; T5 = ci[WS(rs, 2)]; T6 = T4 + T5; T19 = T4 - T5; Te = ci[WS(rs, 3)]; Tf = cr[WS(rs, 1)]; Tg = Te + Tf; T1d = Te - Tf; } { E T7, T8, Tb, Tc; T7 = ci[WS(rs, 1)]; T8 = cr[WS(rs, 3)]; T9 = T7 + T8; T1a = T7 - T8; Tb = cr[WS(rs, 4)]; Tc = ci[0]; Td = Tb + Tc; T1c = Tb - Tc; } TE = T6 - T9; TF = Td - Tg; T1B = T1c - T1d; T1A = T19 - T1a; { E T1b, T1e, Ta, Th; T1b = T19 + T1a; T1e = T1c + T1d; T1f = T1b + T1e; T1t = KP559016994 * (T1b - T1e); Ta = T6 + T9; Th = Td + Tg; Ti = Ta + Th; Tl = KP559016994 * (Ta - Th); } } TH = ci[WS(rs, 9)]; TI = cr[WS(rs, 5)]; TJ = TH - TI; T1i = TH + TI; { E Tp, T1j, Tz, T1n, Ts, T1k, Tw, T1m; { E Tn, To, Tx, Ty; Tn = ci[WS(rs, 7)]; To = cr[WS(rs, 7)]; Tp = Tn - To; T1j = Tn + To; Tx = ci[WS(rs, 8)]; Ty = cr[WS(rs, 6)]; Tz = Tx - Ty; T1n = Tx + Ty; } { E Tq, Tr, Tu, Tv; Tq = ci[WS(rs, 6)]; Tr = cr[WS(rs, 8)]; Ts = Tq - Tr; T1k = Tq + Tr; Tu = ci[WS(rs, 5)]; Tv = cr[WS(rs, 9)]; Tw = Tu - Tv; T1m = Tu + Tv; } Tt = Tp - Ts; TA = Tw - Tz; T1w = T1m + T1n; T1v = T1j + T1k; { E T1l, T1o, TK, TL; T1l = T1j - T1k; T1o = T1m - T1n; T1p = T1l + T1o; T1E = KP559016994 * (T1l - T1o); TK = Tp + Ts; TL = Tw + Tz; TM = TK + TL; TO = KP559016994 * (TK - TL); } } } cr[0] = T3 + Ti; ci[0] = TJ + TM; { E T1g, T1q, T17, T1h; T1g = T18 + T1f; T1q = T1i + T1p; T17 = W[8]; T1h = W[9]; cr[WS(rs, 5)] = FNMS(T1h, T1q, T17 * T1g); ci[WS(rs, 5)] = FMA(T1h, T1g, T17 * T1q); } { E TB, TG, T11, TX, TP, T10, Tm, TW, TN, Tk; TB = FNMS(KP951056516, TA, KP587785252 * Tt); TG = FNMS(KP951056516, TF, KP587785252 * TE); T11 = FMA(KP951056516, TE, KP587785252 * TF); TX = FMA(KP951056516, Tt, KP587785252 * TA); TN = FNMS(KP250000000, TM, TJ); TP = TN - TO; T10 = TO + TN; Tk = FNMS(KP250000000, Ti, T3); Tm = Tk - Tl; TW = Tl + Tk; { E TC, TQ, Tj, TD; TC = Tm - TB; TQ = TG + TP; Tj = W[2]; TD = W[3]; cr[WS(rs, 2)] = FNMS(TD, TQ, Tj * TC); ci[WS(rs, 2)] = FMA(TD, TC, Tj * TQ); } { E T14, T16, T13, T15; T14 = TW - TX; T16 = T11 + T10; T13 = W[10]; T15 = W[11]; cr[WS(rs, 6)] = FNMS(T15, T16, T13 * T14); ci[WS(rs, 6)] = FMA(T15, T14, T13 * T16); } { E TS, TU, TR, TT; TS = Tm + TB; TU = TP - TG; TR = W[14]; TT = W[15]; cr[WS(rs, 8)] = FNMS(TT, TU, TR * TS); ci[WS(rs, 8)] = FMA(TT, TS, TR * TU); } { E TY, T12, TV, TZ; TY = TW + TX; T12 = T10 - T11; TV = W[6]; TZ = W[7]; cr[WS(rs, 4)] = FNMS(TZ, T12, TV * TY); ci[WS(rs, 4)] = FMA(TZ, TY, TV * T12); } } { E T1x, T1C, T1Q, T1N, T1F, T1R, T1u, T1M, T1D, T1s; T1x = FNMS(KP951056516, T1w, KP587785252 * T1v); T1C = FNMS(KP951056516, T1B, KP587785252 * T1A); T1Q = FMA(KP951056516, T1A, KP587785252 * T1B); T1N = FMA(KP951056516, T1v, KP587785252 * T1w); T1D = FNMS(KP250000000, T1p, T1i); T1F = T1D - T1E; T1R = T1E + T1D; T1s = FNMS(KP250000000, T1f, T18); T1u = T1s - T1t; T1M = T1t + T1s; { E T1y, T1G, T1r, T1z; T1y = T1u - T1x; T1G = T1C + T1F; T1r = W[12]; T1z = W[13]; cr[WS(rs, 7)] = FNMS(T1z, T1G, T1r * T1y); ci[WS(rs, 7)] = FMA(T1r, T1G, T1z * T1y); } { E T1U, T1W, T1T, T1V; T1U = T1M + T1N; T1W = T1R - T1Q; T1T = W[16]; T1V = W[17]; cr[WS(rs, 9)] = FNMS(T1V, T1W, T1T * T1U); ci[WS(rs, 9)] = FMA(T1T, T1W, T1V * T1U); } { E T1I, T1K, T1H, T1J; T1I = T1u + T1x; T1K = T1F - T1C; T1H = W[4]; T1J = W[5]; cr[WS(rs, 3)] = FNMS(T1J, T1K, T1H * T1I); ci[WS(rs, 3)] = FMA(T1H, T1K, T1J * T1I); } { E T1O, T1S, T1L, T1P; T1O = T1M - T1N; T1S = T1Q + T1R; T1L = W[0]; T1P = W[1]; cr[WS(rs, 1)] = FNMS(T1P, T1S, T1L * T1O); ci[WS(rs, 1)] = FMA(T1L, T1S, T1P * T1O); } } } } } static const tw_instr twinstr[] = { {TW_FULL, 1, 10}, {TW_NEXT, 1, 0} }; static const hc2hc_desc desc = { 10, "hb_10", twinstr, &GENUS, {72, 30, 30, 0} }; void X(codelet_hb_10) (planner *p) { X(khc2hc_register) (p, hb_10, &desc); } #endif /* HAVE_FMA */ fftw-3.3.3/rdft/scalar/r2cb/hb_12.c0000644000175000017500000003520612054410752013510 00000000000000/* * Copyright (c) 2003, 2007-11 Matteo Frigo * Copyright (c) 2003, 2007-11 Massachusetts Institute of Technology * * This program is free software; you can redistribute it and/or modify * it under the terms of the GNU General Public License as published by * the Free Software Foundation; either version 2 of the License, or * (at your option) any later version. * * This program is distributed in the hope that it will be useful, * but WITHOUT ANY WARRANTY; without even the implied warranty of * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the * GNU General Public License for more details. * * You should have received a copy of the GNU General Public License * along with this program; if not, write to the Free Software * Foundation, Inc., 51 Franklin Street, Fifth Floor, Boston, MA 02110-1301 USA * */ /* This file was automatically generated --- DO NOT EDIT */ /* Generated on Sun Nov 25 07:41:14 EST 2012 */ #include "codelet-rdft.h" #ifdef HAVE_FMA /* Generated by: ../../../genfft/gen_hc2hc.native -fma -reorder-insns -schedule-for-pipeline -compact -variables 4 -pipeline-latency 4 -sign 1 -n 12 -dif -name hb_12 -include hb.h */ /* * This function contains 118 FP additions, 68 FP multiplications, * (or, 72 additions, 22 multiplications, 46 fused multiply/add), * 64 stack variables, 2 constants, and 48 memory accesses */ #include "hb.h" static void hb_12(R *cr, R *ci, const R *W, stride rs, INT mb, INT me, INT ms) { DK(KP866025403, +0.866025403784438646763723170752936183471402627); DK(KP500000000, +0.500000000000000000000000000000000000000000000); { INT m; for (m = mb, W = W + ((mb - 1) * 22); m < me; m = m + 1, cr = cr + ms, ci = ci - ms, W = W + 22, MAKE_VOLATILE_STRIDE(24, rs)) { E T1U, T1X, T1W, T1Y, T1V; { E T18, T20, T2a, T1s, T21, T1b, T29, T1p, TO, T11, To, Tb, Tg, T23, T1f; E Ty, Tl, Tt, T1z, T2d, T1i, T24, T1w, T2c; { E T5, TN, Ta, TI; { E T1, TE, TM, T6, TJ, T1o, T4, T17, TH, TK, T7, T8; T1 = cr[0]; TE = ci[WS(rs, 11)]; TM = cr[WS(rs, 6)]; T6 = ci[WS(rs, 5)]; { E T2, T3, TF, TG; T2 = cr[WS(rs, 4)]; T3 = ci[WS(rs, 3)]; TF = ci[WS(rs, 7)]; TG = cr[WS(rs, 8)]; TJ = ci[WS(rs, 9)]; T1o = T2 - T3; T4 = T2 + T3; T17 = TF + TG; TH = TF - TG; TK = cr[WS(rs, 10)]; T7 = ci[WS(rs, 1)]; T8 = cr[WS(rs, 2)]; } { E T1a, T1r, T1q, T19, TL, T9, T16, T1n; T5 = T1 + T4; T16 = FNMS(KP500000000, T4, T1); T1a = TJ + TK; TL = TJ - TK; T1r = T7 - T8; T9 = T7 + T8; T18 = FNMS(KP866025403, T17, T16); T20 = FMA(KP866025403, T17, T16); T1q = FMA(KP500000000, TL, TM); TN = TL - TM; Ta = T6 + T9; T19 = FNMS(KP500000000, T9, T6); T1n = FNMS(KP500000000, TH, TE); TI = TE + TH; T2a = FMA(KP866025403, T1r, T1q); T1s = FNMS(KP866025403, T1r, T1q); T21 = FNMS(KP866025403, T1a, T19); T1b = FMA(KP866025403, T1a, T19); T29 = FNMS(KP866025403, T1o, T1n); T1p = FMA(KP866025403, T1o, T1n); } } { E Tc, Tp, Tx, Th, Tu, Tf, T1v, Ts, T1e, Tv, Ti, Tj; Tc = cr[WS(rs, 3)]; TO = TI - TN; T11 = TI + TN; Tp = ci[WS(rs, 8)]; To = T5 - Ta; Tb = T5 + Ta; Tx = cr[WS(rs, 9)]; Th = ci[WS(rs, 2)]; { E Td, Te, Tq, Tr; Td = ci[WS(rs, 4)]; Te = ci[0]; Tq = cr[WS(rs, 7)]; Tr = cr[WS(rs, 11)]; Tu = ci[WS(rs, 10)]; Tf = Td + Te; T1v = Td - Te; Ts = Tq + Tr; T1e = Tq - Tr; Tv = ci[WS(rs, 6)]; Ti = cr[WS(rs, 1)]; Tj = cr[WS(rs, 5)]; } { E T1h, T1y, T1x, T1g, Tw, Tk, T1d, T1u; T1d = FNMS(KP500000000, Tf, Tc); Tg = Tc + Tf; Tw = Tu + Tv; T1h = Tv - Tu; Tk = Ti + Tj; T1y = Ti - Tj; T23 = FNMS(KP866025403, T1e, T1d); T1f = FMA(KP866025403, T1e, T1d); Ty = Tw - Tx; T1x = FMA(KP500000000, Tw, Tx); T1g = FNMS(KP500000000, Tk, Th); Tl = Th + Tk; Tt = Tp - Ts; T1u = FMA(KP500000000, Ts, Tp); T1z = FNMS(KP866025403, T1y, T1x); T2d = FMA(KP866025403, T1y, T1x); T1i = FMA(KP866025403, T1h, T1g); T24 = FNMS(KP866025403, T1h, T1g); T1w = FMA(KP866025403, T1v, T1u); T2c = FNMS(KP866025403, T1v, T1u); } } } { E TY, T13, TX, T10; { E Tn, T12, TC, Tm, TD, TS, TA, Tz; Tn = W[16]; T12 = Tt + Ty; Tz = Tt - Ty; TC = W[17]; Tm = Tg + Tl; TD = Tg - Tl; TS = To + Tz; TA = To - Tz; { E TV, TU, TW, TT; { E TQ, TR, TP, TB; TV = TO - TD; TP = TD + TO; cr[0] = Tb + Tm; TB = Tn * TA; TQ = Tn * TP; TR = W[4]; cr[WS(rs, 9)] = FNMS(TC, TP, TB); TU = W[5]; ci[WS(rs, 9)] = FMA(TC, TA, TQ); TW = TR * TV; TT = TR * TS; } ci[WS(rs, 3)] = FMA(TU, TS, TW); cr[WS(rs, 3)] = FNMS(TU, TV, TT); TY = Tb - Tm; T13 = T11 - T12; TX = W[10]; T10 = W[11]; ci[0] = T11 + T12; } } { E T1K, T1Q, T1P, T1L, T2o, T2u, T2t, T2p; { E T1E, T1D, T1H, T1F, T1G, T1t, T1k, T1A; { E T1c, TZ, T14, T1j; T1K = T18 - T1b; T1c = T18 + T1b; TZ = TX * TY; T14 = T10 * TY; T1j = T1f + T1i; T1Q = T1f - T1i; T1P = T1p + T1s; T1t = T1p - T1s; cr[WS(rs, 6)] = FNMS(T10, T13, TZ); ci[WS(rs, 6)] = FMA(TX, T13, T14); T1E = T1c + T1j; T1k = T1c - T1j; T1A = T1w - T1z; T1L = T1w + T1z; } { E T15, T1m, T1B, T1l, T1C; T15 = W[18]; T1m = W[19]; T1D = W[6]; T1H = T1t + T1A; T1B = T1t - T1A; T1l = T15 * T1k; T1C = T1m * T1k; T1F = T1D * T1E; T1G = W[7]; cr[WS(rs, 10)] = FNMS(T1m, T1B, T1l); ci[WS(rs, 10)] = FMA(T15, T1B, T1C); } { E T26, T2i, T2l, T2f, T1Z, T28; { E T22, T1I, T25, T2b, T2e; T22 = T20 + T21; T2o = T20 - T21; cr[WS(rs, 4)] = FNMS(T1G, T1H, T1F); T1I = T1G * T1E; T2u = T23 - T24; T25 = T23 + T24; T2b = T29 - T2a; T2t = T29 + T2a; T2p = T2c + T2d; T2e = T2c - T2d; ci[WS(rs, 4)] = FMA(T1D, T1H, T1I); T26 = T22 - T25; T2i = T22 + T25; T2l = T2b + T2e; T2f = T2b - T2e; } T1Z = W[2]; T28 = W[3]; { E T2h, T2k, T27, T2g, T2j, T2m; T2h = W[14]; T2k = W[15]; T27 = T1Z * T26; T2g = T28 * T26; T2j = T2h * T2i; T2m = T2k * T2i; cr[WS(rs, 2)] = FNMS(T28, T2f, T27); ci[WS(rs, 2)] = FMA(T1Z, T2f, T2g); cr[WS(rs, 8)] = FNMS(T2k, T2l, T2j); ci[WS(rs, 8)] = FMA(T2h, T2l, T2m); } } } { E T2y, T2B, T2A, T2C, T2z; { E T2n, T2q, T2v, T2s, T2r, T2x, T2w; T2n = W[8]; T2y = T2o + T2p; T2q = T2o - T2p; T2B = T2t - T2u; T2v = T2t + T2u; T2s = W[9]; T2r = T2n * T2q; T2x = W[20]; T2w = T2n * T2v; T2A = W[21]; cr[WS(rs, 5)] = FNMS(T2s, T2v, T2r); T2C = T2x * T2B; T2z = T2x * T2y; ci[WS(rs, 5)] = FMA(T2s, T2q, T2w); } ci[WS(rs, 11)] = FMA(T2A, T2y, T2C); cr[WS(rs, 11)] = FNMS(T2A, T2B, T2z); { E T1J, T1M, T1R, T1O, T1N, T1T, T1S; T1J = W[0]; T1U = T1K + T1L; T1M = T1K - T1L; T1X = T1P - T1Q; T1R = T1P + T1Q; T1O = W[1]; T1N = T1J * T1M; T1T = W[12]; T1S = T1J * T1R; T1W = W[13]; cr[WS(rs, 1)] = FNMS(T1O, T1R, T1N); T1Y = T1T * T1X; T1V = T1T * T1U; ci[WS(rs, 1)] = FMA(T1O, T1M, T1S); } } } } } ci[WS(rs, 7)] = FMA(T1W, T1U, T1Y); cr[WS(rs, 7)] = FNMS(T1W, T1X, T1V); } } } static const tw_instr twinstr[] = { {TW_FULL, 1, 12}, {TW_NEXT, 1, 0} }; static const hc2hc_desc desc = { 12, "hb_12", twinstr, &GENUS, {72, 22, 46, 0} }; void X(codelet_hb_12) (planner *p) { X(khc2hc_register) (p, hb_12, &desc); } #else /* HAVE_FMA */ /* Generated by: ../../../genfft/gen_hc2hc.native -compact -variables 4 -pipeline-latency 4 -sign 1 -n 12 -dif -name hb_12 -include hb.h */ /* * This function contains 118 FP additions, 60 FP multiplications, * (or, 88 additions, 30 multiplications, 30 fused multiply/add), * 39 stack variables, 2 constants, and 48 memory accesses */ #include "hb.h" static void hb_12(R *cr, R *ci, const R *W, stride rs, INT mb, INT me, INT ms) { DK(KP500000000, +0.500000000000000000000000000000000000000000000); DK(KP866025403, +0.866025403784438646763723170752936183471402627); { INT m; for (m = mb, W = W + ((mb - 1) * 22); m < me; m = m + 1, cr = cr + ms, ci = ci - ms, W = W + 22, MAKE_VOLATILE_STRIDE(24, rs)) { E T5, TH, T12, T1M, T1i, T1U, Tg, Tt, T19, T1X, T1p, T1P, Ta, TM, T15; E T1N, T1l, T1V, Tl, Ty, T1c, T1Y, T1s, T1Q; { E T1, TD, T4, T1g, TG, T11, T10, T1h; T1 = cr[0]; TD = ci[WS(rs, 11)]; { E T2, T3, TE, TF; T2 = cr[WS(rs, 4)]; T3 = ci[WS(rs, 3)]; T4 = T2 + T3; T1g = KP866025403 * (T2 - T3); TE = ci[WS(rs, 7)]; TF = cr[WS(rs, 8)]; TG = TE - TF; T11 = KP866025403 * (TE + TF); } T5 = T1 + T4; TH = TD + TG; T10 = FNMS(KP500000000, T4, T1); T12 = T10 - T11; T1M = T10 + T11; T1h = FNMS(KP500000000, TG, TD); T1i = T1g + T1h; T1U = T1h - T1g; } { E Tc, Tp, Tf, T17, Ts, T1o, T18, T1n; Tc = cr[WS(rs, 3)]; Tp = ci[WS(rs, 8)]; { E Td, Te, Tq, Tr; Td = ci[WS(rs, 4)]; Te = ci[0]; Tf = Td + Te; T17 = KP866025403 * (Td - Te); Tq = cr[WS(rs, 7)]; Tr = cr[WS(rs, 11)]; Ts = Tq + Tr; T1o = KP866025403 * (Tq - Tr); } Tg = Tc + Tf; Tt = Tp - Ts; T18 = FMA(KP500000000, Ts, Tp); T19 = T17 + T18; T1X = T18 - T17; T1n = FNMS(KP500000000, Tf, Tc); T1p = T1n + T1o; T1P = T1n - T1o; } { E T6, TL, T9, T1j, TK, T14, T13, T1k; T6 = ci[WS(rs, 5)]; TL = cr[WS(rs, 6)]; { E T7, T8, TI, TJ; T7 = ci[WS(rs, 1)]; T8 = cr[WS(rs, 2)]; T9 = T7 + T8; T1j = KP866025403 * (T7 - T8); TI = ci[WS(rs, 9)]; TJ = cr[WS(rs, 10)]; TK = TI - TJ; T14 = KP866025403 * (TI + TJ); } Ta = T6 + T9; TM = TK - TL; T13 = FNMS(KP500000000, T9, T6); T15 = T13 + T14; T1N = T13 - T14; T1k = FMA(KP500000000, TK, TL); T1l = T1j - T1k; T1V = T1j + T1k; } { E Th, Tx, Tk, T1a, Tw, T1r, T1b, T1q; Th = ci[WS(rs, 2)]; Tx = cr[WS(rs, 9)]; { E Ti, Tj, Tu, Tv; Ti = cr[WS(rs, 1)]; Tj = cr[WS(rs, 5)]; Tk = Ti + Tj; T1a = KP866025403 * (Ti - Tj); Tu = ci[WS(rs, 10)]; Tv = ci[WS(rs, 6)]; Tw = Tu + Tv; T1r = KP866025403 * (Tv - Tu); } Tl = Th + Tk; Ty = Tw - Tx; T1b = FMA(KP500000000, Tw, Tx); T1c = T1a - T1b; T1Y = T1a + T1b; T1q = FNMS(KP500000000, Tk, Th); T1s = T1q + T1r; T1Q = T1q - T1r; } { E Tb, Tm, TU, TW, TX, TY, TT, TV; Tb = T5 + Ta; Tm = Tg + Tl; TU = Tb - Tm; TW = TH + TM; TX = Tt + Ty; TY = TW - TX; cr[0] = Tb + Tm; ci[0] = TW + TX; TT = W[10]; TV = W[11]; cr[WS(rs, 6)] = FNMS(TV, TY, TT * TU); ci[WS(rs, 6)] = FMA(TV, TU, TT * TY); } { E TA, TQ, TO, TS; { E To, Tz, TC, TN; To = T5 - Ta; Tz = Tt - Ty; TA = To - Tz; TQ = To + Tz; TC = Tg - Tl; TN = TH - TM; TO = TC + TN; TS = TN - TC; } { E Tn, TB, TP, TR; Tn = W[16]; TB = W[17]; cr[WS(rs, 9)] = FNMS(TB, TO, Tn * TA); ci[WS(rs, 9)] = FMA(Tn, TO, TB * TA); TP = W[4]; TR = W[5]; cr[WS(rs, 3)] = FNMS(TR, TS, TP * TQ); ci[WS(rs, 3)] = FMA(TP, TS, TR * TQ); } } { E T28, T2e, T2c, T2g; { E T26, T27, T2a, T2b; T26 = T1M - T1N; T27 = T1X + T1Y; T28 = T26 - T27; T2e = T26 + T27; T2a = T1U + T1V; T2b = T1P - T1Q; T2c = T2a + T2b; T2g = T2a - T2b; } { E T25, T29, T2d, T2f; T25 = W[8]; T29 = W[9]; cr[WS(rs, 5)] = FNMS(T29, T2c, T25 * T28); ci[WS(rs, 5)] = FMA(T25, T2c, T29 * T28); T2d = W[20]; T2f = W[21]; cr[WS(rs, 11)] = FNMS(T2f, T2g, T2d * T2e); ci[WS(rs, 11)] = FMA(T2d, T2g, T2f * T2e); } } { E T1S, T22, T20, T24; { E T1O, T1R, T1W, T1Z; T1O = T1M + T1N; T1R = T1P + T1Q; T1S = T1O - T1R; T22 = T1O + T1R; T1W = T1U - T1V; T1Z = T1X - T1Y; T20 = T1W - T1Z; T24 = T1W + T1Z; } { E T1L, T1T, T21, T23; T1L = W[2]; T1T = W[3]; cr[WS(rs, 2)] = FNMS(T1T, T20, T1L * T1S); ci[WS(rs, 2)] = FMA(T1T, T1S, T1L * T20); T21 = W[14]; T23 = W[15]; cr[WS(rs, 8)] = FNMS(T23, T24, T21 * T22); ci[WS(rs, 8)] = FMA(T23, T22, T21 * T24); } } { E T1C, T1I, T1G, T1K; { E T1A, T1B, T1E, T1F; T1A = T12 + T15; T1B = T1p + T1s; T1C = T1A - T1B; T1I = T1A + T1B; T1E = T1i + T1l; T1F = T19 + T1c; T1G = T1E - T1F; T1K = T1E + T1F; } { E T1z, T1D, T1H, T1J; T1z = W[18]; T1D = W[19]; cr[WS(rs, 10)] = FNMS(T1D, T1G, T1z * T1C); ci[WS(rs, 10)] = FMA(T1D, T1C, T1z * T1G); T1H = W[6]; T1J = W[7]; cr[WS(rs, 4)] = FNMS(T1J, T1K, T1H * T1I); ci[WS(rs, 4)] = FMA(T1J, T1I, T1H * T1K); } } { E T1e, T1w, T1u, T1y; { E T16, T1d, T1m, T1t; T16 = T12 - T15; T1d = T19 - T1c; T1e = T16 - T1d; T1w = T16 + T1d; T1m = T1i - T1l; T1t = T1p - T1s; T1u = T1m + T1t; T1y = T1m - T1t; } { E TZ, T1f, T1v, T1x; TZ = W[0]; T1f = W[1]; cr[WS(rs, 1)] = FNMS(T1f, T1u, TZ * T1e); ci[WS(rs, 1)] = FMA(TZ, T1u, T1f * T1e); T1v = W[12]; T1x = W[13]; cr[WS(rs, 7)] = FNMS(T1x, T1y, T1v * T1w); ci[WS(rs, 7)] = FMA(T1v, T1y, T1x * T1w); } } } } } static const tw_instr twinstr[] = { {TW_FULL, 1, 12}, {TW_NEXT, 1, 0} }; static const hc2hc_desc desc = { 12, "hb_12", twinstr, &GENUS, {88, 30, 30, 0} }; void X(codelet_hb_12) (planner *p) { X(khc2hc_register) (p, hb_12, &desc); } #endif /* HAVE_FMA */ fftw-3.3.3/rdft/scalar/r2cb/hb_15.c0000644000175000017500000005420612054410753013515 00000000000000/* * Copyright (c) 2003, 2007-11 Matteo Frigo * Copyright (c) 2003, 2007-11 Massachusetts Institute of Technology * * This program is free software; you can redistribute it and/or modify * it under the terms of the GNU General Public License as published by * the Free Software Foundation; either version 2 of the License, or * (at your option) any later version. * * This program is distributed in the hope that it will be useful, * but WITHOUT ANY WARRANTY; without even the implied warranty of * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the * GNU General Public License for more details. * * You should have received a copy of the GNU General Public License * along with this program; if not, write to the Free Software * Foundation, Inc., 51 Franklin Street, Fifth Floor, Boston, MA 02110-1301 USA * */ /* This file was automatically generated --- DO NOT EDIT */ /* Generated on Sun Nov 25 07:41:14 EST 2012 */ #include "codelet-rdft.h" #ifdef HAVE_FMA /* Generated by: ../../../genfft/gen_hc2hc.native -fma -reorder-insns -schedule-for-pipeline -compact -variables 4 -pipeline-latency 4 -sign 1 -n 15 -dif -name hb_15 -include hb.h */ /* * This function contains 184 FP additions, 140 FP multiplications, * (or, 72 additions, 28 multiplications, 112 fused multiply/add), * 93 stack variables, 6 constants, and 60 memory accesses */ #include "hb.h" static void hb_15(R *cr, R *ci, const R *W, stride rs, INT mb, INT me, INT ms) { DK(KP951056516, +0.951056516295153572116439333379382143405698634); DK(KP559016994, +0.559016994374947424102293417182819058860154590); DK(KP250000000, +0.250000000000000000000000000000000000000000000); DK(KP618033988, +0.618033988749894848204586834365638117720309180); DK(KP866025403, +0.866025403784438646763723170752936183471402627); DK(KP500000000, +0.500000000000000000000000000000000000000000000); { INT m; for (m = mb, W = W + ((mb - 1) * 28); m < me; m = m + 1, cr = cr + ms, ci = ci - ms, W = W + 28, MAKE_VOLATILE_STRIDE(30, rs)) { E T3v, T3u, T3r, T3w, T3t; { E T5, T11, T1C, T2U, T2f, T3f, T19, T18, TS, TH, T14, T16, T3g, T3a, Ts; E Tv, T37, T3h, T28, T2h, T1M, T21, T2g, T3n, T2X, T1P, T30, T3m, T1J, T2m; { E T1, TX, T2, T3, TY, TZ; T1 = cr[0]; TX = ci[WS(rs, 14)]; T2 = cr[WS(rs, 5)]; T3 = ci[WS(rs, 4)]; TY = ci[WS(rs, 9)]; TZ = cr[WS(rs, 10)]; { E T1W, T23, T1D, Ta, Tl, T1K, T1Z, T1H, T1G, Tf, TR, T1Y, T26, TI, T1O; E T1N, Tq, TG, T25, Tx, Ty, Tz, TL, T1E; { E Tb, TQ, TN, TO, Te; { E T6, Th, Ti, Tj, T9, Tc, Td, Tk; { E T7, T8, T2e, T4; T6 = cr[WS(rs, 3)]; T2e = T2 - T3; T4 = T2 + T3; { E T1B, T10, T1A, T2d; T1B = TY + TZ; T10 = TY - TZ; T7 = ci[WS(rs, 6)]; T5 = T1 + T4; T1A = FNMS(KP500000000, T4, T1); T11 = TX + T10; T2d = FNMS(KP500000000, T10, TX); T1C = FNMS(KP866025403, T1B, T1A); T2U = FMA(KP866025403, T1B, T1A); T2f = FMA(KP866025403, T2e, T2d); T3f = FNMS(KP866025403, T2e, T2d); T8 = ci[WS(rs, 1)]; } Th = cr[WS(rs, 6)]; Ti = ci[WS(rs, 3)]; Tj = cr[WS(rs, 1)]; T9 = T7 + T8; T1W = T7 - T8; } Tb = ci[WS(rs, 2)]; T23 = Ti - Tj; Tk = Ti + Tj; T1D = FNMS(KP500000000, T9, T6); Ta = T6 + T9; Tc = cr[WS(rs, 2)]; Tl = Th + Tk; T1K = FNMS(KP500000000, Tk, Th); Td = cr[WS(rs, 7)]; TQ = cr[WS(rs, 12)]; TN = ci[WS(rs, 12)]; TO = ci[WS(rs, 7)]; Te = Tc + Td; T1Z = Tc - Td; } { E Tm, TF, TC, TD, Tp, Tn, To, TP, TJ, TK, TE; Tm = ci[WS(rs, 5)]; T1H = TO - TN; TP = TN + TO; T1G = FNMS(KP500000000, Te, Tb); Tf = Tb + Te; Tn = ci[0]; TR = TP - TQ; T1Y = FMA(KP500000000, TP, TQ); To = cr[WS(rs, 4)]; TF = cr[WS(rs, 9)]; TC = ci[WS(rs, 10)]; TD = cr[WS(rs, 14)]; Tp = Tn + To; T26 = Tn - To; TI = ci[WS(rs, 11)]; T1O = TC + TD; TE = TC - TD; T1N = FNMS(KP500000000, Tp, Tm); Tq = Tm + Tp; TJ = cr[WS(rs, 8)]; TG = TE - TF; T25 = FMA(KP500000000, TE, TF); TK = cr[WS(rs, 13)]; Tx = ci[WS(rs, 8)]; Ty = ci[WS(rs, 13)]; Tz = cr[WS(rs, 11)]; TL = TJ + TK; T1E = TJ - TK; } } { E Tg, T1L, Tr, T22, T12, T1X, T38, T13, T39, T20; { E TA, T1V, TM, TB; Tg = Ta + Tf; T19 = Ta - Tf; T1L = Ty + Tz; TA = Ty - Tz; T1V = FMA(KP500000000, TL, TI); TM = TI - TL; T18 = Tl - Tq; Tr = Tl + Tq; TB = Tx + TA; T22 = FNMS(KP500000000, TA, Tx); T12 = TM + TR; TS = TM - TR; T1X = FMA(KP866025403, T1W, T1V); T38 = FNMS(KP866025403, T1W, T1V); T13 = TB + TG; TH = TB - TG; T39 = FMA(KP866025403, T1Z, T1Y); T20 = FNMS(KP866025403, T1Z, T1Y); } { E T35, T24, T27, T36; T14 = T12 + T13; T16 = T12 - T13; T3g = T38 - T39; T3a = T38 + T39; T35 = FNMS(KP866025403, T23, T22); T24 = FMA(KP866025403, T23, T22); Ts = Tg + Tr; Tv = Tg - Tr; T27 = FNMS(KP866025403, T26, T25); T36 = FMA(KP866025403, T26, T25); T37 = T35 + T36; T3h = T35 - T36; T28 = T24 + T27; T2h = T24 - T27; { E T1F, T1I, T2Y, T2Z, T2V, T2W; T2V = FNMS(KP866025403, T1E, T1D); T1F = FMA(KP866025403, T1E, T1D); T1I = FMA(KP866025403, T1H, T1G); T2W = FNMS(KP866025403, T1H, T1G); T2Y = FNMS(KP866025403, T1L, T1K); T1M = FMA(KP866025403, T1L, T1K); T21 = T1X + T20; T2g = T1X - T20; T3n = T2V - T2W; T2X = T2V + T2W; T2Z = FNMS(KP866025403, T1O, T1N); T1P = FMA(KP866025403, T1O, T1N); T30 = T2Y + T2Z; T3m = T2Y - T2Z; T1J = T1F + T1I; T2m = T1F - T1I; } } } } } { E T31, T33, T2n, T1Q; cr[0] = T5 + Ts; T31 = T2X + T30; T33 = T2X - T30; T2n = T1M - T1P; T1Q = T1M + T1P; ci[0] = T11 + T14; { E T1T, T1R, T1r, T1o, T1n; { E T1q, T1a, TT, T1l, Tu, T17, T1p, T15; T1q = FMA(KP618033988, T18, T19); T1a = FNMS(KP618033988, T19, T18); T1T = T1J - T1Q; T1R = T1J + T1Q; T15 = FNMS(KP250000000, T14, T11); TT = FNMS(KP618033988, TS, TH); T1l = FMA(KP618033988, TH, TS); Tu = FNMS(KP250000000, Ts, T5); T17 = FNMS(KP559016994, T16, T15); T1p = FMA(KP559016994, T16, T15); { E T1h, T1m, T1e, T1x, T1w, T1v, T1g, T1d; { E TW, T1b, Tt, T1u, TU, T1k, Tw; TW = W[5]; T1k = FMA(KP559016994, Tv, Tu); Tw = FNMS(KP559016994, Tv, Tu); T1b = FMA(KP951056516, T1a, T17); T1h = FNMS(KP951056516, T1a, T17); Tt = W[4]; T1m = FNMS(KP951056516, T1l, T1k); T1u = FMA(KP951056516, T1l, T1k); T1e = FMA(KP951056516, TT, Tw); TU = FNMS(KP951056516, TT, Tw); { E T1t, TV, T1c, T1y; T1x = FNMS(KP951056516, T1q, T1p); T1r = FMA(KP951056516, T1q, T1p); T1w = W[17]; T1t = W[16]; TV = Tt * TU; T1c = TW * TU; T1y = T1w * T1u; T1v = T1t * T1u; cr[WS(rs, 3)] = FNMS(TW, T1b, TV); ci[WS(rs, 3)] = FMA(Tt, T1b, T1c); ci[WS(rs, 9)] = FMA(T1t, T1x, T1y); } } cr[WS(rs, 9)] = FNMS(T1w, T1x, T1v); T1g = W[23]; T1d = W[22]; { E T1j, T1s, T1i, T1f; T1o = W[11]; T1i = T1g * T1e; T1f = T1d * T1e; T1j = W[10]; T1s = T1o * T1m; ci[WS(rs, 12)] = FMA(T1d, T1h, T1i); cr[WS(rs, 12)] = FNMS(T1g, T1h, T1f); T1n = T1j * T1m; ci[WS(rs, 6)] = FMA(T1j, T1r, T1s); } } } { E T2v, T2u, T2r, T2w, T2t; { E T1S, T2N, T2o, T2E, T2Q, T2P, T2k, T2S, T29, T2z, T2R, T2j, T2O, T2i; cr[WS(rs, 6)] = FNMS(T1o, T1r, T1n); T1S = FNMS(KP250000000, T1R, T1C); T2O = T1C + T1R; T2N = W[18]; T2o = FMA(KP618033988, T2n, T2m); T2E = FNMS(KP618033988, T2m, T2n); T2Q = W[19]; T2P = T2N * T2O; T2i = T2g + T2h; T2k = T2g - T2h; T2S = T2Q * T2O; T29 = FMA(KP618033988, T28, T21); T2z = FNMS(KP618033988, T21, T28); T2R = T2f + T2i; T2j = FNMS(KP250000000, T2i, T2f); { E T2D, T2p, T2I, T2A, T2a, T2s, T2c, T1z, T2l, T1U, T2y; cr[WS(rs, 10)] = FNMS(T2Q, T2R, T2P); T2l = FMA(KP559016994, T2k, T2j); T2D = FNMS(KP559016994, T2k, T2j); T1U = FMA(KP559016994, T1T, T1S); T2y = FNMS(KP559016994, T1T, T1S); ci[WS(rs, 10)] = FMA(T2N, T2R, T2S); T2p = FMA(KP951056516, T2o, T2l); T2v = FNMS(KP951056516, T2o, T2l); T2I = FNMS(KP951056516, T2z, T2y); T2A = FMA(KP951056516, T2z, T2y); T2a = FNMS(KP951056516, T29, T1U); T2s = FMA(KP951056516, T29, T1U); T2c = W[1]; T1z = W[0]; { E T2F, T2L, T2K, T2J; { E T2H, T2M, T2q, T2b; T2F = FNMS(KP951056516, T2E, T2D); T2L = FMA(KP951056516, T2E, T2D); T2K = W[25]; T2q = T2c * T2a; T2b = T1z * T2a; T2H = W[24]; T2M = T2K * T2I; ci[WS(rs, 1)] = FMA(T1z, T2p, T2q); cr[WS(rs, 1)] = FNMS(T2c, T2p, T2b); T2J = T2H * T2I; ci[WS(rs, 13)] = FMA(T2H, T2L, T2M); } { E T2C, T2x, T2G, T2B; T2C = W[13]; cr[WS(rs, 13)] = FNMS(T2K, T2L, T2J); T2x = W[12]; T2G = T2C * T2A; T2u = W[7]; T2B = T2x * T2A; T2r = W[6]; ci[WS(rs, 7)] = FMA(T2x, T2F, T2G); T2w = T2u * T2s; cr[WS(rs, 7)] = FNMS(T2C, T2F, T2B); T2t = T2r * T2s; } } } } { E T32, T3N, T3E, T3o, T3Q, T3P, T3k, T3S, T3z, T3b, T3j, T3R, T3O, T3i; ci[WS(rs, 4)] = FMA(T2r, T2v, T2w); cr[WS(rs, 4)] = FNMS(T2u, T2v, T2t); T3O = T2U + T31; T32 = FNMS(KP250000000, T31, T2U); T3N = W[8]; T3E = FMA(KP618033988, T3m, T3n); T3o = FNMS(KP618033988, T3n, T3m); T3Q = W[9]; T3P = T3N * T3O; T3k = T3g - T3h; T3i = T3g + T3h; T3S = T3Q * T3O; T3z = FMA(KP618033988, T37, T3a); T3b = FNMS(KP618033988, T3a, T37); T3j = FNMS(KP250000000, T3i, T3f); T3R = T3f + T3i; { E T3D, T3p, T3A, T3I, T3s, T3c, T3e, T2T, T3l, T3y, T34; cr[WS(rs, 5)] = FNMS(T3Q, T3R, T3P); T3D = FMA(KP559016994, T3k, T3j); T3l = FNMS(KP559016994, T3k, T3j); T3y = FMA(KP559016994, T33, T32); T34 = FNMS(KP559016994, T33, T32); ci[WS(rs, 5)] = FMA(T3N, T3R, T3S); T3v = FMA(KP951056516, T3o, T3l); T3p = FNMS(KP951056516, T3o, T3l); T3A = FNMS(KP951056516, T3z, T3y); T3I = FMA(KP951056516, T3z, T3y); T3s = FNMS(KP951056516, T3b, T34); T3c = FMA(KP951056516, T3b, T34); T3e = W[3]; T2T = W[2]; { E T3L, T3F, T3K, T3J; { E T3H, T3M, T3q, T3d; T3L = FNMS(KP951056516, T3E, T3D); T3F = FMA(KP951056516, T3E, T3D); T3K = W[27]; T3q = T3e * T3c; T3d = T2T * T3c; T3H = W[26]; T3M = T3K * T3I; ci[WS(rs, 2)] = FMA(T2T, T3p, T3q); cr[WS(rs, 2)] = FNMS(T3e, T3p, T3d); T3J = T3H * T3I; ci[WS(rs, 14)] = FMA(T3H, T3L, T3M); } { E T3C, T3x, T3G, T3B; T3C = W[21]; cr[WS(rs, 14)] = FNMS(T3K, T3L, T3J); T3x = W[20]; T3G = T3C * T3A; T3u = W[15]; T3B = T3x * T3A; T3r = W[14]; ci[WS(rs, 11)] = FMA(T3x, T3F, T3G); T3w = T3u * T3s; cr[WS(rs, 11)] = FNMS(T3C, T3F, T3B); T3t = T3r * T3s; } } } } } } } } ci[WS(rs, 8)] = FMA(T3r, T3v, T3w); cr[WS(rs, 8)] = FNMS(T3u, T3v, T3t); } } } static const tw_instr twinstr[] = { {TW_FULL, 1, 15}, {TW_NEXT, 1, 0} }; static const hc2hc_desc desc = { 15, "hb_15", twinstr, &GENUS, {72, 28, 112, 0} }; void X(codelet_hb_15) (planner *p) { X(khc2hc_register) (p, hb_15, &desc); } #else /* HAVE_FMA */ /* Generated by: ../../../genfft/gen_hc2hc.native -compact -variables 4 -pipeline-latency 4 -sign 1 -n 15 -dif -name hb_15 -include hb.h */ /* * This function contains 184 FP additions, 112 FP multiplications, * (or, 128 additions, 56 multiplications, 56 fused multiply/add), * 75 stack variables, 6 constants, and 60 memory accesses */ #include "hb.h" static void hb_15(R *cr, R *ci, const R *W, stride rs, INT mb, INT me, INT ms) { DK(KP559016994, +0.559016994374947424102293417182819058860154590); DK(KP250000000, +0.250000000000000000000000000000000000000000000); DK(KP951056516, +0.951056516295153572116439333379382143405698634); DK(KP587785252, +0.587785252292473129168705954639072768597652438); DK(KP500000000, +0.500000000000000000000000000000000000000000000); DK(KP866025403, +0.866025403784438646763723170752936183471402627); { INT m; for (m = mb, W = W + ((mb - 1) * 28); m < me; m = m + 1, cr = cr + ms, ci = ci - ms, W = W + 28, MAKE_VOLATILE_STRIDE(30, rs)) { E T5, T10, T1J, T2C, T2c, T2M, TH, T18, T17, TS, T2Q, T2R, T2S, Tg, Tr; E Ts, T11, T12, T13, T2N, T2O, T2P, T1u, T1x, T1y, T1W, T1Z, T28, T1P, T1S; E T27, T1B, T1E, T1F, T2G, T2H, T2I, T2D, T2E, T2F; { E T1, TW, T4, T2a, TZ, T1I, T1H, T2b; T1 = cr[0]; TW = ci[WS(rs, 14)]; { E T2, T3, TX, TY; T2 = cr[WS(rs, 5)]; T3 = ci[WS(rs, 4)]; T4 = T2 + T3; T2a = KP866025403 * (T2 - T3); TX = ci[WS(rs, 9)]; TY = cr[WS(rs, 10)]; TZ = TX - TY; T1I = KP866025403 * (TX + TY); } T5 = T1 + T4; T10 = TW + TZ; T1H = FNMS(KP500000000, T4, T1); T1J = T1H - T1I; T2C = T1H + T1I; T2b = FNMS(KP500000000, TZ, TW); T2c = T2a + T2b; T2M = T2b - T2a; } { E Ta, T1N, T1s, Tl, T1U, T1z, Tf, T1Q, T1v, TG, T1R, T1w, Tq, T1X, T1C; E TM, T1V, T1A, TB, T1O, T1t, TR, T1Y, T1D; { E T6, T7, T8, T9; T6 = cr[WS(rs, 3)]; T7 = ci[WS(rs, 6)]; T8 = ci[WS(rs, 1)]; T9 = T7 + T8; Ta = T6 + T9; T1N = KP866025403 * (T7 - T8); T1s = FNMS(KP500000000, T9, T6); } { E Th, Ti, Tj, Tk; Th = cr[WS(rs, 6)]; Ti = ci[WS(rs, 3)]; Tj = cr[WS(rs, 1)]; Tk = Ti + Tj; Tl = Th + Tk; T1U = KP866025403 * (Ti - Tj); T1z = FNMS(KP500000000, Tk, Th); } { E Tb, Tc, Td, Te; Tb = ci[WS(rs, 2)]; Tc = cr[WS(rs, 2)]; Td = cr[WS(rs, 7)]; Te = Tc + Td; Tf = Tb + Te; T1Q = KP866025403 * (Tc - Td); T1v = FNMS(KP500000000, Te, Tb); } { E TF, TC, TD, TE; TF = cr[WS(rs, 12)]; TC = ci[WS(rs, 12)]; TD = ci[WS(rs, 7)]; TE = TC + TD; TG = TE - TF; T1R = FMA(KP500000000, TE, TF); T1w = KP866025403 * (TD - TC); } { E Tm, Tn, To, Tp; Tm = ci[WS(rs, 5)]; Tn = ci[0]; To = cr[WS(rs, 4)]; Tp = Tn + To; Tq = Tm + Tp; T1X = KP866025403 * (Tn - To); T1C = FNMS(KP500000000, Tp, Tm); } { E TI, TJ, TK, TL; TI = ci[WS(rs, 8)]; TJ = ci[WS(rs, 13)]; TK = cr[WS(rs, 11)]; TL = TJ - TK; TM = TI + TL; T1V = FNMS(KP500000000, TL, TI); T1A = KP866025403 * (TJ + TK); } { E Tx, Ty, Tz, TA; Tx = ci[WS(rs, 11)]; Ty = cr[WS(rs, 8)]; Tz = cr[WS(rs, 13)]; TA = Ty + Tz; TB = Tx - TA; T1O = FMA(KP500000000, TA, Tx); T1t = KP866025403 * (Ty - Tz); } { E TQ, TN, TO, TP; TQ = cr[WS(rs, 9)]; TN = ci[WS(rs, 10)]; TO = cr[WS(rs, 14)]; TP = TN - TO; TR = TP - TQ; T1Y = FMA(KP500000000, TP, TQ); T1D = KP866025403 * (TN + TO); } TH = TB - TG; T18 = Tl - Tq; T17 = Ta - Tf; TS = TM - TR; T2Q = T1V - T1U; T2R = T1X + T1Y; T2S = T2Q - T2R; Tg = Ta + Tf; Tr = Tl + Tq; Ts = Tg + Tr; T11 = TB + TG; T12 = TM + TR; T13 = T11 + T12; T2N = T1O - T1N; T2O = T1Q + T1R; T2P = T2N - T2O; T1u = T1s + T1t; T1x = T1v + T1w; T1y = T1u + T1x; T1W = T1U + T1V; T1Z = T1X - T1Y; T28 = T1W + T1Z; T1P = T1N + T1O; T1S = T1Q - T1R; T27 = T1P + T1S; T1B = T1z + T1A; T1E = T1C + T1D; T1F = T1B + T1E; T2G = T1z - T1A; T2H = T1C - T1D; T2I = T2G + T2H; T2D = T1s - T1t; T2E = T1v - T1w; T2F = T2D + T2E; } cr[0] = T5 + Ts; ci[0] = T10 + T13; { E TT, T19, T1k, T1h, T16, T1l, Tw, T1g; TT = FNMS(KP951056516, TS, KP587785252 * TH); T19 = FNMS(KP951056516, T18, KP587785252 * T17); T1k = FMA(KP951056516, T17, KP587785252 * T18); T1h = FMA(KP951056516, TH, KP587785252 * TS); { E T14, T15, Tu, Tv; T14 = FNMS(KP250000000, T13, T10); T15 = KP559016994 * (T11 - T12); T16 = T14 - T15; T1l = T15 + T14; Tu = FNMS(KP250000000, Ts, T5); Tv = KP559016994 * (Tg - Tr); Tw = Tu - Tv; T1g = Tv + Tu; } { E TU, T1a, Tt, TV; TU = Tw + TT; T1a = T16 - T19; Tt = W[4]; TV = W[5]; cr[WS(rs, 3)] = FNMS(TV, T1a, Tt * TU); ci[WS(rs, 3)] = FMA(TV, TU, Tt * T1a); } { E T1o, T1q, T1n, T1p; T1o = T1g + T1h; T1q = T1l - T1k; T1n = W[16]; T1p = W[17]; cr[WS(rs, 9)] = FNMS(T1p, T1q, T1n * T1o); ci[WS(rs, 9)] = FMA(T1p, T1o, T1n * T1q); } { E T1c, T1e, T1b, T1d; T1c = Tw - TT; T1e = T19 + T16; T1b = W[22]; T1d = W[23]; cr[WS(rs, 12)] = FNMS(T1d, T1e, T1b * T1c); ci[WS(rs, 12)] = FMA(T1d, T1c, T1b * T1e); } { E T1i, T1m, T1f, T1j; T1i = T1g - T1h; T1m = T1k + T1l; T1f = W[10]; T1j = W[11]; cr[WS(rs, 6)] = FNMS(T1j, T1m, T1f * T1i); ci[WS(rs, 6)] = FMA(T1j, T1i, T1f * T1m); } } { E T21, T2n, T26, T2q, T1M, T2y, T2m, T2f, T2A, T2r, T2x, T2z; { E T1T, T20, T24, T25; T1T = T1P - T1S; T20 = T1W - T1Z; T21 = FMA(KP951056516, T1T, KP587785252 * T20); T2n = FNMS(KP951056516, T20, KP587785252 * T1T); T24 = T1u - T1x; T25 = T1B - T1E; T26 = FMA(KP951056516, T24, KP587785252 * T25); T2q = FNMS(KP951056516, T25, KP587785252 * T24); } { E T1G, T1K, T1L, T29, T2d, T2e; T1G = KP559016994 * (T1y - T1F); T1K = T1y + T1F; T1L = FNMS(KP250000000, T1K, T1J); T1M = T1G + T1L; T2y = T1J + T1K; T2m = T1L - T1G; T29 = KP559016994 * (T27 - T28); T2d = T27 + T28; T2e = FNMS(KP250000000, T2d, T2c); T2f = T29 + T2e; T2A = T2c + T2d; T2r = T2e - T29; } T2x = W[18]; T2z = W[19]; cr[WS(rs, 10)] = FNMS(T2z, T2A, T2x * T2y); ci[WS(rs, 10)] = FMA(T2z, T2y, T2x * T2A); { E T2u, T2w, T2t, T2v; T2u = T2m + T2n; T2w = T2r - T2q; T2t = W[24]; T2v = W[25]; cr[WS(rs, 13)] = FNMS(T2v, T2w, T2t * T2u); ci[WS(rs, 13)] = FMA(T2v, T2u, T2t * T2w); } { E T22, T2g, T1r, T23; T22 = T1M - T21; T2g = T26 + T2f; T1r = W[0]; T23 = W[1]; cr[WS(rs, 1)] = FNMS(T23, T2g, T1r * T22); ci[WS(rs, 1)] = FMA(T23, T22, T1r * T2g); } { E T2i, T2k, T2h, T2j; T2i = T1M + T21; T2k = T2f - T26; T2h = W[6]; T2j = W[7]; cr[WS(rs, 4)] = FNMS(T2j, T2k, T2h * T2i); ci[WS(rs, 4)] = FMA(T2j, T2i, T2h * T2k); } { E T2o, T2s, T2l, T2p; T2o = T2m - T2n; T2s = T2q + T2r; T2l = W[12]; T2p = W[13]; cr[WS(rs, 7)] = FNMS(T2p, T2s, T2l * T2o); ci[WS(rs, 7)] = FMA(T2p, T2o, T2l * T2s); } } { E T31, T3h, T36, T3k, T2K, T3g, T2Y, T2U, T3l, T39, T2B, T2L; { E T2Z, T30, T34, T35; T2Z = T2N + T2O; T30 = T2Q + T2R; T31 = FNMS(KP951056516, T30, KP587785252 * T2Z); T3h = FMA(KP951056516, T2Z, KP587785252 * T30); T34 = T2D - T2E; T35 = T2G - T2H; T36 = FNMS(KP951056516, T35, KP587785252 * T34); T3k = FMA(KP951056516, T34, KP587785252 * T35); } { E T2X, T2J, T2W, T38, T2T, T37; T2X = KP559016994 * (T2F - T2I); T2J = T2F + T2I; T2W = FNMS(KP250000000, T2J, T2C); T2K = T2C + T2J; T3g = T2X + T2W; T2Y = T2W - T2X; T38 = KP559016994 * (T2P - T2S); T2T = T2P + T2S; T37 = FNMS(KP250000000, T2T, T2M); T2U = T2M + T2T; T3l = T38 + T37; T39 = T37 - T38; } T2B = W[8]; T2L = W[9]; cr[WS(rs, 5)] = FNMS(T2L, T2U, T2B * T2K); ci[WS(rs, 5)] = FMA(T2L, T2K, T2B * T2U); { E T3o, T3q, T3n, T3p; T3o = T3g + T3h; T3q = T3l - T3k; T3n = W[26]; T3p = W[27]; cr[WS(rs, 14)] = FNMS(T3p, T3q, T3n * T3o); ci[WS(rs, 14)] = FMA(T3n, T3q, T3p * T3o); } { E T32, T3a, T2V, T33; T32 = T2Y - T31; T3a = T36 + T39; T2V = W[2]; T33 = W[3]; cr[WS(rs, 2)] = FNMS(T33, T3a, T2V * T32); ci[WS(rs, 2)] = FMA(T2V, T3a, T33 * T32); } { E T3c, T3e, T3b, T3d; T3c = T2Y + T31; T3e = T39 - T36; T3b = W[14]; T3d = W[15]; cr[WS(rs, 8)] = FNMS(T3d, T3e, T3b * T3c); ci[WS(rs, 8)] = FMA(T3b, T3e, T3d * T3c); } { E T3i, T3m, T3f, T3j; T3i = T3g - T3h; T3m = T3k + T3l; T3f = W[20]; T3j = W[21]; cr[WS(rs, 11)] = FNMS(T3j, T3m, T3f * T3i); ci[WS(rs, 11)] = FMA(T3f, T3m, T3j * T3i); } } } } } static const tw_instr twinstr[] = { {TW_FULL, 1, 15}, {TW_NEXT, 1, 0} }; static const hc2hc_desc desc = { 15, "hb_15", twinstr, &GENUS, {128, 56, 56, 0} }; void X(codelet_hb_15) (planner *p) { X(khc2hc_register) (p, hb_15, &desc); } #endif /* HAVE_FMA */ fftw-3.3.3/rdft/scalar/r2cb/hb_16.c0000644000175000017500000005007512054410753013516 00000000000000/* * Copyright (c) 2003, 2007-11 Matteo Frigo * Copyright (c) 2003, 2007-11 Massachusetts Institute of Technology * * This program is free software; you can redistribute it and/or modify * it under the terms of the GNU General Public License as published by * the Free Software Foundation; either version 2 of the License, or * (at your option) any later version. * * This program is distributed in the hope that it will be useful, * but WITHOUT ANY WARRANTY; without even the implied warranty of * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the * GNU General Public License for more details. * * You should have received a copy of the GNU General Public License * along with this program; if not, write to the Free Software * Foundation, Inc., 51 Franklin Street, Fifth Floor, Boston, MA 02110-1301 USA * */ /* This file was automatically generated --- DO NOT EDIT */ /* Generated on Sun Nov 25 07:41:14 EST 2012 */ #include "codelet-rdft.h" #ifdef HAVE_FMA /* Generated by: ../../../genfft/gen_hc2hc.native -fma -reorder-insns -schedule-for-pipeline -compact -variables 4 -pipeline-latency 4 -sign 1 -n 16 -dif -name hb_16 -include hb.h */ /* * This function contains 174 FP additions, 100 FP multiplications, * (or, 104 additions, 30 multiplications, 70 fused multiply/add), * 78 stack variables, 3 constants, and 64 memory accesses */ #include "hb.h" static void hb_16(R *cr, R *ci, const R *W, stride rs, INT mb, INT me, INT ms) { DK(KP923879532, +0.923879532511286756128183189396788286822416626); DK(KP707106781, +0.707106781186547524400844362104849039284835938); DK(KP414213562, +0.414213562373095048801688724209698078569671875); { INT m; for (m = mb, W = W + ((mb - 1) * 30); m < me; m = m + 1, cr = cr + ms, ci = ci - ms, W = W + 30, MAKE_VOLATILE_STRIDE(32, rs)) { E T1I, T1L, T1K, T1M, T1J; { E T1O, TA, T1h, T21, T3b, T2T, T3D, T3r, T1k, T1P, T3y, Tf, T36, T2A, T22; E TL, T3z, T3u, T2U, T2F, T2K, T2V, T12, Tu, T3E, TX, T1n, T17, T1T, T24; E T1W, T25; { E T2z, TF, TK, T2w; { E Tw, T3, T2x, TJ, T2Q, T1g, T1d, T6, TC, TB, Ta, T2R, Tz, TD, Tb; E Tc; { E T1e, T1f, T4, T5; { E T1, T2, TH, TI; T1 = cr[0]; T2 = ci[WS(rs, 7)]; TH = ci[WS(rs, 9)]; TI = cr[WS(rs, 14)]; T1e = ci[WS(rs, 15)]; Tw = T1 - T2; T3 = T1 + T2; T2x = TH - TI; TJ = TH + TI; T1f = cr[WS(rs, 8)]; T4 = cr[WS(rs, 4)]; T5 = ci[WS(rs, 3)]; } { E T8, T9, Tx, Ty; T8 = cr[WS(rs, 2)]; T2Q = T1e - T1f; T1g = T1e + T1f; T1d = T4 - T5; T6 = T4 + T5; T9 = ci[WS(rs, 5)]; Tx = ci[WS(rs, 11)]; Ty = cr[WS(rs, 12)]; TC = ci[WS(rs, 13)]; TB = T8 - T9; Ta = T8 + T9; T2R = Tx - Ty; Tz = Tx + Ty; TD = cr[WS(rs, 10)]; Tb = ci[WS(rs, 1)]; Tc = cr[WS(rs, 6)]; } } { E T2y, TE, TG, Te, T2P, T2S, T3p, Td; T1O = Tw + Tz; TA = Tw - Tz; T2y = TC - TD; TE = TC + TD; TG = Tb - Tc; Td = Tb + Tc; T1h = T1d + T1g; T21 = T1g - T1d; Te = Ta + Td; T2P = Ta - Td; T2S = T2Q - T2R; T3p = T2Q + T2R; { E T1i, T1j, T3q, T7; T3q = T2y + T2x; T2z = T2x - T2y; TF = TB - TE; T1i = TB + TE; T3b = T2S - T2P; T2T = T2P + T2S; TK = TG - TJ; T1j = TG + TJ; T3D = T3p - T3q; T3r = T3p + T3q; T2w = T3 - T6; T7 = T3 + T6; T1k = T1i - T1j; T1P = T1i + T1j; T3y = T7 - Te; Tf = T7 + Te; } } } { E T13, Ti, T2C, T11, T2D, T16, TY, Tl, TT, TS, Tp, T2H, TQ, TU, Tq; E Tr; { E T14, T15, Tj, Tk; { E Tg, Th, TZ, T10; Tg = cr[WS(rs, 1)]; T36 = T2w - T2z; T2A = T2w + T2z; T22 = TF - TK; TL = TF + TK; Th = ci[WS(rs, 6)]; TZ = ci[WS(rs, 14)]; T10 = cr[WS(rs, 9)]; T14 = ci[WS(rs, 10)]; T13 = Tg - Th; Ti = Tg + Th; T2C = TZ - T10; T11 = TZ + T10; T15 = cr[WS(rs, 13)]; Tj = cr[WS(rs, 5)]; Tk = ci[WS(rs, 2)]; } { E Tn, To, TO, TP; Tn = ci[0]; T2D = T14 - T15; T16 = T14 + T15; TY = Tj - Tk; Tl = Tj + Tk; To = cr[WS(rs, 7)]; TO = ci[WS(rs, 8)]; TP = cr[WS(rs, 15)]; TT = ci[WS(rs, 12)]; TS = Tn - To; Tp = Tn + To; T2H = TO - TP; TQ = TO + TP; TU = cr[WS(rs, 11)]; Tq = cr[WS(rs, 3)]; Tr = ci[WS(rs, 4)]; } } { E TV, TN, Tm, Tt; { E T2E, T3s, Ts, T3t, T2J, T2B, T2I, T2G; T2E = T2C - T2D; T3s = T2C + T2D; T2I = TT - TU; TV = TT + TU; TN = Tq - Tr; Ts = Tq + Tr; T3t = T2H + T2I; T2J = T2H - T2I; Tm = Ti + Tl; T2B = Ti - Tl; Tt = Tp + Ts; T2G = Tp - Ts; T3z = T3t - T3s; T3u = T3s + T3t; T2U = T2B + T2E; T2F = T2B - T2E; T2K = T2G + T2J; T2V = T2J - T2G; } { E T1U, T1V, T1R, T1S, TR, TW; TR = TN - TQ; T1U = TN + TQ; T1V = TS + TV; TW = TS - TV; T1R = T11 - TY; T12 = TY + T11; Tu = Tm + Tt; T3E = Tm - Tt; TX = FNMS(KP414213562, TW, TR); T1n = FMA(KP414213562, TR, TW); T17 = T13 - T16; T1S = T13 + T16; T1T = FNMS(KP414213562, T1S, T1R); T24 = FMA(KP414213562, T1R, T1S); T1W = FNMS(KP414213562, T1V, T1U); T25 = FMA(KP414213562, T1U, T1V); } } } } { E T18, T1m, T2W, T2L, T3j, T3i, T3h; { E T3m, T3v, T3l, T3o; cr[0] = Tf + Tu; T18 = FMA(KP414213562, T17, T12); T1m = FNMS(KP414213562, T12, T17); T3m = Tf - Tu; T3v = T3r - T3u; T3l = W[14]; T3o = W[15]; ci[0] = T3r + T3u; { E T3A, T3I, T3L, T3F, T3C, T3G, T3B, T3x, T3n, T3w, T3H, T3K; T3A = T3y - T3z; T3I = T3y + T3z; T3n = T3l * T3m; T3w = T3o * T3m; T3L = T3E + T3D; T3F = T3D - T3E; T3x = W[22]; cr[WS(rs, 8)] = FNMS(T3o, T3v, T3n); ci[WS(rs, 8)] = FMA(T3l, T3v, T3w); T3C = W[23]; T3G = T3x * T3F; T3B = T3x * T3A; ci[WS(rs, 12)] = FMA(T3C, T3A, T3G); cr[WS(rs, 12)] = FNMS(T3C, T3F, T3B); T3H = W[6]; T3K = W[7]; { E T3g, T38, T3d, T35, T3a; { E T37, T3c, T3M, T3J; T37 = T2V - T2U; T2W = T2U + T2V; T2L = T2F + T2K; T3c = T2F - T2K; T3M = T3H * T3L; T3J = T3H * T3I; T3g = FMA(KP707106781, T37, T36); T38 = FNMS(KP707106781, T37, T36); ci[WS(rs, 4)] = FMA(T3K, T3I, T3M); cr[WS(rs, 4)] = FNMS(T3K, T3L, T3J); T3d = FNMS(KP707106781, T3c, T3b); T3j = FMA(KP707106781, T3c, T3b); } T35 = W[26]; T3a = W[27]; { E T3f, T3e, T39, T3k; T3f = W[10]; T3i = W[11]; T3e = T35 * T3d; T39 = T35 * T38; T3k = T3f * T3j; T3h = T3f * T3g; ci[WS(rs, 14)] = FMA(T3a, T38, T3e); cr[WS(rs, 14)] = FNMS(T3a, T3d, T39); ci[WS(rs, 6)] = FMA(T3i, T3g, T3k); } } } } cr[WS(rs, 6)] = FNMS(T3i, T3j, T3h); { E T2g, T2m, T2l, T2h, T2d, T29, T2c, T2b, T2e; { E T33, T2Z, T32, T31, T34; { E T2v, T30, T2M, T2X, T2O, T2N, T2Y; T2v = W[18]; T30 = FMA(KP707106781, T2L, T2A); T2M = FNMS(KP707106781, T2L, T2A); T33 = FMA(KP707106781, T2W, T2T); T2X = FNMS(KP707106781, T2W, T2T); T2O = W[19]; T2N = T2v * T2M; T2Z = W[2]; T32 = W[3]; T2Y = T2O * T2M; cr[WS(rs, 10)] = FNMS(T2O, T2X, T2N); T31 = T2Z * T30; T34 = T32 * T30; ci[WS(rs, 10)] = FMA(T2v, T2X, T2Y); } { E T1Q, T1X, T23, T26; T2g = FMA(KP707106781, T1P, T1O); T1Q = FNMS(KP707106781, T1P, T1O); cr[WS(rs, 2)] = FNMS(T32, T33, T31); ci[WS(rs, 2)] = FMA(T2Z, T33, T34); T1X = T1T + T1W; T2m = T1W - T1T; T2l = FNMS(KP707106781, T22, T21); T23 = FMA(KP707106781, T22, T21); T26 = T24 - T25; T2h = T24 + T25; { E T1N, T2a, T1Y, T27, T20, T1Z, T28; T1N = W[20]; T2a = FNMS(KP923879532, T1X, T1Q); T1Y = FMA(KP923879532, T1X, T1Q); T2d = FMA(KP923879532, T26, T23); T27 = FNMS(KP923879532, T26, T23); T20 = W[21]; T1Z = T1N * T1Y; T29 = W[4]; T2c = W[5]; T28 = T20 * T1Y; cr[WS(rs, 11)] = FNMS(T20, T27, T1Z); T2b = T29 * T2a; T2e = T2c * T2a; ci[WS(rs, 11)] = FMA(T1N, T27, T28); } } } { E T1y, T1E, T1D, T1z, T1v, T1r, T1u, T1t, T1w; { E TM, T19, T1l, T1o; T1y = FMA(KP707106781, TL, TA); TM = FNMS(KP707106781, TL, TA); cr[WS(rs, 3)] = FNMS(T2c, T2d, T2b); ci[WS(rs, 3)] = FMA(T29, T2d, T2e); T19 = TX - T18; T1E = T18 + TX; T1D = FMA(KP707106781, T1k, T1h); T1l = FNMS(KP707106781, T1k, T1h); T1o = T1m - T1n; T1z = T1m + T1n; { E Tv, T1s, T1a, T1p, T1c, T1b, T1q; Tv = W[24]; T1s = FMA(KP923879532, T19, TM); T1a = FNMS(KP923879532, T19, TM); T1v = FMA(KP923879532, T1o, T1l); T1p = FNMS(KP923879532, T1o, T1l); T1c = W[25]; T1b = Tv * T1a; T1r = W[8]; T1u = W[9]; T1q = T1c * T1a; cr[WS(rs, 13)] = FNMS(T1c, T1p, T1b); T1t = T1r * T1s; T1w = T1u * T1s; ci[WS(rs, 13)] = FMA(Tv, T1p, T1q); } } { E T2q, T2t, T2s, T2u, T2r; cr[WS(rs, 5)] = FNMS(T1u, T1v, T1t); ci[WS(rs, 5)] = FMA(T1r, T1v, T1w); { E T2f, T2i, T2n, T2k, T2j, T2p, T2o; T2f = W[12]; T2q = FMA(KP923879532, T2h, T2g); T2i = FNMS(KP923879532, T2h, T2g); T2t = FNMS(KP923879532, T2m, T2l); T2n = FMA(KP923879532, T2m, T2l); T2k = W[13]; T2j = T2f * T2i; T2p = W[28]; T2o = T2f * T2n; T2s = W[29]; cr[WS(rs, 7)] = FNMS(T2k, T2n, T2j); T2u = T2p * T2t; T2r = T2p * T2q; ci[WS(rs, 7)] = FMA(T2k, T2i, T2o); } ci[WS(rs, 15)] = FMA(T2s, T2q, T2u); cr[WS(rs, 15)] = FNMS(T2s, T2t, T2r); { E T1x, T1A, T1F, T1C, T1B, T1H, T1G; T1x = W[16]; T1I = FMA(KP923879532, T1z, T1y); T1A = FNMS(KP923879532, T1z, T1y); T1L = FMA(KP923879532, T1E, T1D); T1F = FNMS(KP923879532, T1E, T1D); T1C = W[17]; T1B = T1x * T1A; T1H = W[0]; T1G = T1x * T1F; T1K = W[1]; cr[WS(rs, 9)] = FNMS(T1C, T1F, T1B); T1M = T1H * T1L; T1J = T1H * T1I; ci[WS(rs, 9)] = FMA(T1C, T1A, T1G); } } } } } } ci[WS(rs, 1)] = FMA(T1K, T1I, T1M); cr[WS(rs, 1)] = FNMS(T1K, T1L, T1J); } } } static const tw_instr twinstr[] = { {TW_FULL, 1, 16}, {TW_NEXT, 1, 0} }; static const hc2hc_desc desc = { 16, "hb_16", twinstr, &GENUS, {104, 30, 70, 0} }; void X(codelet_hb_16) (planner *p) { X(khc2hc_register) (p, hb_16, &desc); } #else /* HAVE_FMA */ /* Generated by: ../../../genfft/gen_hc2hc.native -compact -variables 4 -pipeline-latency 4 -sign 1 -n 16 -dif -name hb_16 -include hb.h */ /* * This function contains 174 FP additions, 84 FP multiplications, * (or, 136 additions, 46 multiplications, 38 fused multiply/add), * 50 stack variables, 3 constants, and 64 memory accesses */ #include "hb.h" static void hb_16(R *cr, R *ci, const R *W, stride rs, INT mb, INT me, INT ms) { DK(KP382683432, +0.382683432365089771728459984030398866761344562); DK(KP923879532, +0.923879532511286756128183189396788286822416626); DK(KP707106781, +0.707106781186547524400844362104849039284835938); { INT m; for (m = mb, W = W + ((mb - 1) * 30); m < me; m = m + 1, cr = cr + ms, ci = ci - ms, W = W + 30, MAKE_VOLATILE_STRIDE(32, rs)) { E T7, T2K, T2W, Tw, T17, T1S, T2k, T1w, Te, TD, T1x, T10, T2n, T2L, T1Z; E T2X, Tm, T1z, TN, T19, T2e, T2p, T2P, T2Z, Tt, T1A, TW, T1a, T27, T2q; E T2S, T30; { E T3, T1Q, T16, T1R, T6, T2i, T13, T2j; { E T1, T2, T14, T15; T1 = cr[0]; T2 = ci[WS(rs, 7)]; T3 = T1 + T2; T1Q = T1 - T2; T14 = ci[WS(rs, 11)]; T15 = cr[WS(rs, 12)]; T16 = T14 - T15; T1R = T14 + T15; } { E T4, T5, T11, T12; T4 = cr[WS(rs, 4)]; T5 = ci[WS(rs, 3)]; T6 = T4 + T5; T2i = T4 - T5; T11 = ci[WS(rs, 15)]; T12 = cr[WS(rs, 8)]; T13 = T11 - T12; T2j = T11 + T12; } T7 = T3 + T6; T2K = T1Q + T1R; T2W = T2j - T2i; Tw = T3 - T6; T17 = T13 - T16; T1S = T1Q - T1R; T2k = T2i + T2j; T1w = T13 + T16; } { E Ta, T1T, TC, T1U, Td, T1W, Tz, T1X; { E T8, T9, TA, TB; T8 = cr[WS(rs, 2)]; T9 = ci[WS(rs, 5)]; Ta = T8 + T9; T1T = T8 - T9; TA = ci[WS(rs, 13)]; TB = cr[WS(rs, 10)]; TC = TA - TB; T1U = TA + TB; } { E Tb, Tc, Tx, Ty; Tb = ci[WS(rs, 1)]; Tc = cr[WS(rs, 6)]; Td = Tb + Tc; T1W = Tb - Tc; Tx = ci[WS(rs, 9)]; Ty = cr[WS(rs, 14)]; Tz = Tx - Ty; T1X = Tx + Ty; } Te = Ta + Td; TD = Tz - TC; T1x = TC + Tz; T10 = Ta - Td; { E T2l, T2m, T1V, T1Y; T2l = T1T + T1U; T2m = T1W + T1X; T2n = KP707106781 * (T2l - T2m); T2L = KP707106781 * (T2l + T2m); T1V = T1T - T1U; T1Y = T1W - T1X; T1Z = KP707106781 * (T1V + T1Y); T2X = KP707106781 * (T1V - T1Y); } } { E Ti, T2b, TL, T2c, Tl, T28, TI, T29, TF, TM; { E Tg, Th, TJ, TK; Tg = cr[WS(rs, 1)]; Th = ci[WS(rs, 6)]; Ti = Tg + Th; T2b = Tg - Th; TJ = ci[WS(rs, 10)]; TK = cr[WS(rs, 13)]; TL = TJ - TK; T2c = TJ + TK; } { E Tj, Tk, TG, TH; Tj = cr[WS(rs, 5)]; Tk = ci[WS(rs, 2)]; Tl = Tj + Tk; T28 = Tj - Tk; TG = ci[WS(rs, 14)]; TH = cr[WS(rs, 9)]; TI = TG - TH; T29 = TG + TH; } Tm = Ti + Tl; T1z = TI + TL; TF = Ti - Tl; TM = TI - TL; TN = TF - TM; T19 = TF + TM; { E T2a, T2d, T2N, T2O; T2a = T28 + T29; T2d = T2b - T2c; T2e = FMA(KP923879532, T2a, KP382683432 * T2d); T2p = FNMS(KP382683432, T2a, KP923879532 * T2d); T2N = T2b + T2c; T2O = T29 - T28; T2P = FNMS(KP923879532, T2O, KP382683432 * T2N); T2Z = FMA(KP382683432, T2O, KP923879532 * T2N); } } { E Tp, T24, TU, T25, Ts, T21, TR, T22, TO, TV; { E Tn, To, TS, TT; Tn = ci[0]; To = cr[WS(rs, 7)]; Tp = Tn + To; T24 = Tn - To; TS = ci[WS(rs, 12)]; TT = cr[WS(rs, 11)]; TU = TS - TT; T25 = TS + TT; } { E Tq, Tr, TP, TQ; Tq = cr[WS(rs, 3)]; Tr = ci[WS(rs, 4)]; Ts = Tq + Tr; T21 = Tq - Tr; TP = ci[WS(rs, 8)]; TQ = cr[WS(rs, 15)]; TR = TP - TQ; T22 = TP + TQ; } Tt = Tp + Ts; T1A = TR + TU; TO = Tp - Ts; TV = TR - TU; TW = TO + TV; T1a = TV - TO; { E T23, T26, T2Q, T2R; T23 = T21 - T22; T26 = T24 - T25; T27 = FNMS(KP382683432, T26, KP923879532 * T23); T2q = FMA(KP382683432, T23, KP923879532 * T26); T2Q = T24 + T25; T2R = T21 + T22; T2S = FNMS(KP923879532, T2R, KP382683432 * T2Q); T30 = FMA(KP382683432, T2R, KP923879532 * T2Q); } } { E Tf, Tu, T1u, T1y, T1B, T1C, T1t, T1v; Tf = T7 + Te; Tu = Tm + Tt; T1u = Tf - Tu; T1y = T1w + T1x; T1B = T1z + T1A; T1C = T1y - T1B; cr[0] = Tf + Tu; ci[0] = T1y + T1B; T1t = W[14]; T1v = W[15]; cr[WS(rs, 8)] = FNMS(T1v, T1C, T1t * T1u); ci[WS(rs, 8)] = FMA(T1v, T1u, T1t * T1C); } { E T2U, T34, T32, T36; { E T2M, T2T, T2Y, T31; T2M = T2K - T2L; T2T = T2P + T2S; T2U = T2M - T2T; T34 = T2M + T2T; T2Y = T2W + T2X; T31 = T2Z - T30; T32 = T2Y - T31; T36 = T2Y + T31; } { E T2J, T2V, T33, T35; T2J = W[20]; T2V = W[21]; cr[WS(rs, 11)] = FNMS(T2V, T32, T2J * T2U); ci[WS(rs, 11)] = FMA(T2V, T2U, T2J * T32); T33 = W[4]; T35 = W[5]; cr[WS(rs, 3)] = FNMS(T35, T36, T33 * T34); ci[WS(rs, 3)] = FMA(T35, T34, T33 * T36); } } { E T3a, T3g, T3e, T3i; { E T38, T39, T3c, T3d; T38 = T2K + T2L; T39 = T2Z + T30; T3a = T38 - T39; T3g = T38 + T39; T3c = T2W - T2X; T3d = T2P - T2S; T3e = T3c + T3d; T3i = T3c - T3d; } { E T37, T3b, T3f, T3h; T37 = W[12]; T3b = W[13]; cr[WS(rs, 7)] = FNMS(T3b, T3e, T37 * T3a); ci[WS(rs, 7)] = FMA(T37, T3e, T3b * T3a); T3f = W[28]; T3h = W[29]; cr[WS(rs, 15)] = FNMS(T3h, T3i, T3f * T3g); ci[WS(rs, 15)] = FMA(T3f, T3i, T3h * T3g); } } { E TY, T1e, T1c, T1g; { E TE, TX, T18, T1b; TE = Tw + TD; TX = KP707106781 * (TN + TW); TY = TE - TX; T1e = TE + TX; T18 = T10 + T17; T1b = KP707106781 * (T19 + T1a); T1c = T18 - T1b; T1g = T18 + T1b; } { E Tv, TZ, T1d, T1f; Tv = W[18]; TZ = W[19]; cr[WS(rs, 10)] = FNMS(TZ, T1c, Tv * TY); ci[WS(rs, 10)] = FMA(TZ, TY, Tv * T1c); T1d = W[2]; T1f = W[3]; cr[WS(rs, 2)] = FNMS(T1f, T1g, T1d * T1e); ci[WS(rs, 2)] = FMA(T1f, T1e, T1d * T1g); } } { E T1k, T1q, T1o, T1s; { E T1i, T1j, T1m, T1n; T1i = Tw - TD; T1j = KP707106781 * (T1a - T19); T1k = T1i - T1j; T1q = T1i + T1j; T1m = T17 - T10; T1n = KP707106781 * (TN - TW); T1o = T1m - T1n; T1s = T1m + T1n; } { E T1h, T1l, T1p, T1r; T1h = W[26]; T1l = W[27]; cr[WS(rs, 14)] = FNMS(T1l, T1o, T1h * T1k); ci[WS(rs, 14)] = FMA(T1h, T1o, T1l * T1k); T1p = W[10]; T1r = W[11]; cr[WS(rs, 6)] = FNMS(T1r, T1s, T1p * T1q); ci[WS(rs, 6)] = FMA(T1p, T1s, T1r * T1q); } } { E T2g, T2u, T2s, T2w; { E T20, T2f, T2o, T2r; T20 = T1S - T1Z; T2f = T27 - T2e; T2g = T20 - T2f; T2u = T20 + T2f; T2o = T2k - T2n; T2r = T2p - T2q; T2s = T2o - T2r; T2w = T2o + T2r; } { E T1P, T2h, T2t, T2v; T1P = W[24]; T2h = W[25]; cr[WS(rs, 13)] = FNMS(T2h, T2s, T1P * T2g); ci[WS(rs, 13)] = FMA(T2h, T2g, T1P * T2s); T2t = W[8]; T2v = W[9]; cr[WS(rs, 5)] = FNMS(T2v, T2w, T2t * T2u); ci[WS(rs, 5)] = FMA(T2v, T2u, T2t * T2w); } } { E T2A, T2G, T2E, T2I; { E T2y, T2z, T2C, T2D; T2y = T1S + T1Z; T2z = T2p + T2q; T2A = T2y - T2z; T2G = T2y + T2z; T2C = T2k + T2n; T2D = T2e + T27; T2E = T2C - T2D; T2I = T2C + T2D; } { E T2x, T2B, T2F, T2H; T2x = W[16]; T2B = W[17]; cr[WS(rs, 9)] = FNMS(T2B, T2E, T2x * T2A); ci[WS(rs, 9)] = FMA(T2x, T2E, T2B * T2A); T2F = W[0]; T2H = W[1]; cr[WS(rs, 1)] = FNMS(T2H, T2I, T2F * T2G); ci[WS(rs, 1)] = FMA(T2F, T2I, T2H * T2G); } } { E T1G, T1M, T1K, T1O; { E T1E, T1F, T1I, T1J; T1E = T7 - Te; T1F = T1A - T1z; T1G = T1E - T1F; T1M = T1E + T1F; T1I = T1w - T1x; T1J = Tm - Tt; T1K = T1I - T1J; T1O = T1J + T1I; } { E T1D, T1H, T1L, T1N; T1D = W[22]; T1H = W[23]; cr[WS(rs, 12)] = FNMS(T1H, T1K, T1D * T1G); ci[WS(rs, 12)] = FMA(T1D, T1K, T1H * T1G); T1L = W[6]; T1N = W[7]; cr[WS(rs, 4)] = FNMS(T1N, T1O, T1L * T1M); ci[WS(rs, 4)] = FMA(T1L, T1O, T1N * T1M); } } } } } static const tw_instr twinstr[] = { {TW_FULL, 1, 16}, {TW_NEXT, 1, 0} }; static const hc2hc_desc desc = { 16, "hb_16", twinstr, &GENUS, {136, 46, 38, 0} }; void X(codelet_hb_16) (planner *p) { X(khc2hc_register) (p, hb_16, &desc); } #endif /* HAVE_FMA */ fftw-3.3.3/rdft/scalar/r2cb/hb_32.c0000644000175000017500000013377612054410762013526 00000000000000/* * Copyright (c) 2003, 2007-11 Matteo Frigo * Copyright (c) 2003, 2007-11 Massachusetts Institute of Technology * * This program is free software; you can redistribute it and/or modify * it under the terms of the GNU General Public License as published by * the Free Software Foundation; either version 2 of the License, or * (at your option) any later version. * * This program is distributed in the hope that it will be useful, * but WITHOUT ANY WARRANTY; without even the implied warranty of * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the * GNU General Public License for more details. * * You should have received a copy of the GNU General Public License * along with this program; if not, write to the Free Software * Foundation, Inc., 51 Franklin Street, Fifth Floor, Boston, MA 02110-1301 USA * */ /* This file was automatically generated --- DO NOT EDIT */ /* Generated on Sun Nov 25 07:41:15 EST 2012 */ #include "codelet-rdft.h" #ifdef HAVE_FMA /* Generated by: ../../../genfft/gen_hc2hc.native -fma -reorder-insns -schedule-for-pipeline -compact -variables 4 -pipeline-latency 4 -sign 1 -n 32 -dif -name hb_32 -include hb.h */ /* * This function contains 434 FP additions, 260 FP multiplications, * (or, 236 additions, 62 multiplications, 198 fused multiply/add), * 135 stack variables, 7 constants, and 128 memory accesses */ #include "hb.h" static void hb_32(R *cr, R *ci, const R *W, stride rs, INT mb, INT me, INT ms) { DK(KP980785280, +0.980785280403230449126182236134239036973933731); DK(KP198912367, +0.198912367379658006911597622644676228597850501); DK(KP831469612, +0.831469612302545237078788377617905756738560812); DK(KP668178637, +0.668178637919298919997757686523080761552472251); DK(KP923879532, +0.923879532511286756128183189396788286822416626); DK(KP414213562, +0.414213562373095048801688724209698078569671875); DK(KP707106781, +0.707106781186547524400844362104849039284835938); { INT m; for (m = mb, W = W + ((mb - 1) * 62); m < me; m = m + 1, cr = cr + ms, ci = ci - ms, W = W + 62, MAKE_VOLATILE_STRIDE(64, rs)) { E T5o, T5r, T5q, T5n, T5s, T5p; { E T5K, Tf, T8k, T7k, T8x, T7N, T3i, T1i, T3v, T2L, T5f, T4v, T6T, T6m, T52; E T42, TZ, T6X, T1X, T3p, T8p, T8B, T3o, T26, T58, T4n, T7T, T7z, T59, T4k; E T6p, T6a, TK, T6W, T2o, T3m, T8s, T8A, T3l, T2x, T55, T4g, T7S, T7G, T56; E T4d, T6o, T61, T5Q, T5N, T6f, Tu, T8y, T7r, T8l, T7Q, T3w, T1F, T45, T48; E T3j, T2O, T53, T4y; { E T62, T69, T4j, T4i; { E T6l, T6i, T40, T41; { E T12, T3, T2D, T6, T6g, T2G, T6h, T15, Td, T6k, T1g, T2J, Ta, T17, T1a; E T6j; { E T2E, T2F, T13, T14; { E T1, T2, T4, T5; T1 = cr[0]; T2 = ci[WS(rs, 15)]; T4 = cr[WS(rs, 8)]; T5 = ci[WS(rs, 7)]; T2E = ci[WS(rs, 31)]; T12 = T1 - T2; T3 = T1 + T2; T2D = T4 - T5; T6 = T4 + T5; T2F = cr[WS(rs, 16)]; } T13 = ci[WS(rs, 23)]; T14 = cr[WS(rs, 24)]; { E Tb, Tc, T1d, T1e; Tb = ci[WS(rs, 3)]; T6g = T2E - T2F; T2G = T2E + T2F; T6h = T13 - T14; T15 = T13 + T14; Tc = cr[WS(rs, 12)]; T1d = ci[WS(rs, 19)]; T1e = cr[WS(rs, 28)]; { E T8, T1c, T1f, T9, T18, T19; T8 = cr[WS(rs, 4)]; Td = Tb + Tc; T1c = Tb - Tc; T6k = T1d - T1e; T1f = T1d + T1e; T9 = ci[WS(rs, 11)]; T18 = ci[WS(rs, 27)]; T19 = cr[WS(rs, 20)]; T1g = T1c - T1f; T2J = T1c + T1f; Ta = T8 + T9; T17 = T8 - T9; T1a = T18 + T19; T6j = T18 - T19; } } } { E T2I, T7M, T7L, T16, T1h, T4u, T4t, T2H, T2K; { E T7i, T7, T1b, Te, T7j; T7i = T3 - T6; T7 = T3 + T6; T2I = T17 + T1a; T1b = T17 - T1a; Te = Ta + Td; T7M = Ta - Td; T7j = T6k - T6j; T6l = T6j + T6k; T6i = T6g + T6h; T7L = T6g - T6h; T5K = T7 - Te; Tf = T7 + Te; T8k = T7i + T7j; T7k = T7i - T7j; T40 = T12 + T15; T16 = T12 - T15; T1h = T1b + T1g; T4u = T1b - T1g; } T4t = T2G - T2D; T2H = T2D + T2G; T8x = T7M + T7L; T7N = T7L - T7M; T3i = FMA(KP707106781, T1h, T16); T1i = FNMS(KP707106781, T1h, T16); T2K = T2I - T2J; T41 = T2I + T2J; T3v = FMA(KP707106781, T2K, T2H); T2L = FNMS(KP707106781, T2K, T2H); T5f = FNMS(KP707106781, T4u, T4t); T4v = FMA(KP707106781, T4u, T4t); } } { E T1Y, T1H, TR, T7w, T1K, T21, T65, T7t, TV, T1M, TU, T67, T1U, TW, T1N; E T1O; { E TL, TM, TO, TP, T63, T64; TL = ci[0]; T6T = T6i + T6l; T6m = T6i - T6l; T52 = FMA(KP707106781, T41, T40); T42 = FNMS(KP707106781, T41, T40); TM = cr[WS(rs, 15)]; TO = cr[WS(rs, 7)]; TP = ci[WS(rs, 8)]; { E T1I, TN, TQ, T1J, T1Z, T20; T1I = ci[WS(rs, 16)]; T1Y = TL - TM; TN = TL + TM; T1H = TO - TP; TQ = TO + TP; T1J = cr[WS(rs, 31)]; T1Z = ci[WS(rs, 24)]; T20 = cr[WS(rs, 23)]; TR = TN + TQ; T7w = TN - TQ; T1K = T1I + T1J; T63 = T1I - T1J; T64 = T1Z - T20; T21 = T1Z + T20; } { E TS, TT, T1S, T1T; TS = cr[WS(rs, 3)]; T65 = T63 + T64; T7t = T63 - T64; TT = ci[WS(rs, 12)]; T1S = ci[WS(rs, 20)]; T1T = cr[WS(rs, 27)]; TV = ci[WS(rs, 4)]; T1M = TS - TT; TU = TS + TT; T67 = T1S - T1T; T1U = T1S + T1T; TW = cr[WS(rs, 11)]; T1N = ci[WS(rs, 28)]; T1O = cr[WS(rs, 19)]; } } { E T4l, T1L, T24, T23, T8n, T7v, T1W, T8o, T7y, T4m, T22, T25; { E T1V, T7u, T7x, T1Q, T1R, TX; T4l = T1H + T1K; T1L = T1H - T1K; T1R = TV - TW; TX = TV + TW; { E T66, T1P, TY, T68; T66 = T1N - T1O; T1P = T1N + T1O; T24 = T1R - T1U; T1V = T1R + T1U; T7u = TU - TX; TY = TU + TX; T68 = T66 + T67; T7x = T67 - T66; T23 = T1M - T1P; T1Q = T1M + T1P; TZ = TR + TY; T62 = TR - TY; T69 = T65 - T68; T6X = T65 + T68; } T8n = T7u + T7t; T7v = T7t - T7u; T4j = T1Q + T1V; T1W = T1Q - T1V; T8o = T7w + T7x; T7y = T7w - T7x; } T4i = T1Y + T21; T22 = T1Y - T21; T25 = T23 + T24; T4m = T23 - T24; T1X = FNMS(KP707106781, T1W, T1L); T3p = FMA(KP707106781, T1W, T1L); T8p = FNMS(KP414213562, T8o, T8n); T8B = FMA(KP414213562, T8n, T8o); T3o = FMA(KP707106781, T25, T22); T26 = FNMS(KP707106781, T25, T22); T58 = FMA(KP707106781, T4m, T4l); T4n = FNMS(KP707106781, T4m, T4l); T7T = FNMS(KP414213562, T7v, T7y); T7z = FMA(KP414213562, T7y, T7v); } } } { E T5T, T60, T4c, T4b; { E T2p, T28, TC, T7D, T2b, T2s, T5W, T7A, TG, T2d, TF, T5Y, T2l, TH, T2e; E T2f; { E Tw, Tx, Tz, TA, T5U, T5V; Tw = cr[WS(rs, 1)]; T59 = FMA(KP707106781, T4j, T4i); T4k = FNMS(KP707106781, T4j, T4i); T6p = T69 - T62; T6a = T62 + T69; Tx = ci[WS(rs, 14)]; Tz = cr[WS(rs, 9)]; TA = ci[WS(rs, 6)]; { E T29, Ty, TB, T2a, T2q, T2r; T29 = ci[WS(rs, 30)]; T2p = Tw - Tx; Ty = Tw + Tx; T28 = Tz - TA; TB = Tz + TA; T2a = cr[WS(rs, 17)]; T2q = ci[WS(rs, 22)]; T2r = cr[WS(rs, 25)]; TC = Ty + TB; T7D = Ty - TB; T2b = T29 + T2a; T5U = T29 - T2a; T5V = T2q - T2r; T2s = T2q + T2r; } { E TD, TE, T2j, T2k; TD = cr[WS(rs, 5)]; T5W = T5U + T5V; T7A = T5U - T5V; TE = ci[WS(rs, 10)]; T2j = ci[WS(rs, 18)]; T2k = cr[WS(rs, 29)]; TG = ci[WS(rs, 2)]; T2d = TD - TE; TF = TD + TE; T5Y = T2j - T2k; T2l = T2j + T2k; TH = cr[WS(rs, 13)]; T2e = ci[WS(rs, 26)]; T2f = cr[WS(rs, 21)]; } } { E T4e, T2c, T2v, T2u, T8q, T7C, T2n, T8r, T7F, T4f, T2t, T2w; { E T2m, T7B, T7E, T2h, T2i, TI; T4e = T2b - T28; T2c = T28 + T2b; T2i = TG - TH; TI = TG + TH; { E T5X, T2g, TJ, T5Z; T5X = T2e - T2f; T2g = T2e + T2f; T2v = T2i - T2l; T2m = T2i + T2l; T7B = TF - TI; TJ = TF + TI; T5Z = T5X + T5Y; T7E = T5Y - T5X; T2u = T2d - T2g; T2h = T2d + T2g; TK = TC + TJ; T5T = TC - TJ; T60 = T5W - T5Z; T6W = T5W + T5Z; } T8q = T7B + T7A; T7C = T7A - T7B; T4c = T2h + T2m; T2n = T2h - T2m; T8r = T7D + T7E; T7F = T7D - T7E; } T4b = T2p + T2s; T2t = T2p - T2s; T2w = T2u + T2v; T4f = T2v - T2u; T2o = FNMS(KP707106781, T2n, T2c); T3m = FMA(KP707106781, T2n, T2c); T8s = FMA(KP414213562, T8r, T8q); T8A = FNMS(KP414213562, T8q, T8r); T3l = FMA(KP707106781, T2w, T2t); T2x = FNMS(KP707106781, T2w, T2t); T55 = FMA(KP707106781, T4f, T4e); T4g = FNMS(KP707106781, T4f, T4e); T7S = FMA(KP414213562, T7C, T7F); T7G = FNMS(KP414213562, T7F, T7C); } } { E T44, T1D, Tm, T7o, T7p, T43, T1y, T47, T1s, Tt, T7m, T7l, T46, T1n; { E Tj, T1z, Ti, T5P, T1C, Tk, T1v, T1w; { E Tg, Th, T1A, T1B; Tg = cr[WS(rs, 2)]; T56 = FMA(KP707106781, T4c, T4b); T4d = FNMS(KP707106781, T4c, T4b); T6o = T5T + T60; T61 = T5T - T60; Th = ci[WS(rs, 13)]; T1A = ci[WS(rs, 21)]; T1B = cr[WS(rs, 26)]; Tj = cr[WS(rs, 10)]; T1z = Tg - Th; Ti = Tg + Th; T5P = T1A - T1B; T1C = T1A + T1B; Tk = ci[WS(rs, 5)]; T1v = ci[WS(rs, 29)]; T1w = cr[WS(rs, 18)]; } { E T1u, Tl, T5O, T1x; T44 = T1z + T1C; T1D = T1z - T1C; T1u = Tj - Tk; Tl = Tj + Tk; T5O = T1v - T1w; T1x = T1v + T1w; Tm = Ti + Tl; T7o = Ti - Tl; T7p = T5O - T5P; T5Q = T5O + T5P; T43 = T1x - T1u; T1y = T1u + T1x; } } { E Tq, T1o, Tp, T5M, T1r, Tr, T1k, T1l; { E Tn, To, T1p, T1q; Tn = ci[WS(rs, 1)]; To = cr[WS(rs, 14)]; T1p = ci[WS(rs, 25)]; T1q = cr[WS(rs, 22)]; Tq = cr[WS(rs, 6)]; T1o = Tn - To; Tp = Tn + To; T5M = T1p - T1q; T1r = T1p + T1q; Tr = ci[WS(rs, 9)]; T1k = ci[WS(rs, 17)]; T1l = cr[WS(rs, 30)]; } { E T1j, Ts, T5L, T1m; T47 = T1o + T1r; T1s = T1o - T1r; T1j = Tq - Tr; Ts = Tq + Tr; T5L = T1k - T1l; T1m = T1k + T1l; Tt = Tp + Ts; T7m = Tp - Ts; T7l = T5L - T5M; T5N = T5L + T5M; T46 = T1j + T1m; T1n = T1j - T1m; } } { E T7P, T7O, T2N, T1t, T1E, T2M, T7n, T7q, T4w, T4x; T7P = T7m + T7l; T7n = T7l - T7m; T7q = T7o + T7p; T7O = T7o - T7p; T6f = Tm - Tt; Tu = Tm + Tt; T8y = T7q + T7n; T7r = T7n - T7q; T2N = FMA(KP414213562, T1n, T1s); T1t = FNMS(KP414213562, T1s, T1n); T1E = FMA(KP414213562, T1D, T1y); T2M = FNMS(KP414213562, T1y, T1D); T8l = T7O + T7P; T7Q = T7O - T7P; T3w = T1E + T1t; T1F = T1t - T1E; T45 = FNMS(KP414213562, T44, T43); T4w = FMA(KP414213562, T43, T44); T4x = FMA(KP414213562, T46, T47); T48 = FNMS(KP414213562, T47, T46); T3j = T2M + T2N; T2O = T2M - T2N; T53 = T4w + T4x; T4y = T4w - T4x; } } } } { E T72, T5g, T49, T78, T77, T73, T7s, T7U, T7R, T7H, T3f, T3e, T3d; { E T5R, T8m, T8C, T8z, T8t, T8e, T86, T88, T8h, T8f, T8i, T8c, T8g; { E T6P, T6Q, T6Z, T6S, T6R; { E Tv, T10, T6V, T6Y, T6U; T72 = Tf - Tu; Tv = Tf + Tu; T6U = T5Q + T5N; T5R = T5N - T5Q; T5g = T48 - T45; T49 = T45 + T48; T10 = TK + TZ; T78 = TK - TZ; T77 = T6T - T6U; T6V = T6T + T6U; T6Y = T6W + T6X; T73 = T6X - T6W; T6P = W[30]; cr[0] = Tv + T10; T6Q = Tv - T10; ci[0] = T6V + T6Y; T6Z = T6V - T6Y; T6S = W[31]; T6R = T6P * T6Q; } { E T8O, T8W, T8Q, T8Z, T8X, T90, T8U, T8Y; { E T8R, T8S, T8M, T8N, T70; T8M = FMA(KP707106781, T8l, T8k); T8m = FNMS(KP707106781, T8l, T8k); T8C = T8A - T8B; T8N = T8A + T8B; T70 = T6S * T6Q; cr[WS(rs, 16)] = FNMS(T6S, T6Z, T6R); T8R = FMA(KP707106781, T8y, T8x); T8z = FNMS(KP707106781, T8y, T8x); T8O = FNMS(KP923879532, T8N, T8M); T8W = FMA(KP923879532, T8N, T8M); ci[WS(rs, 16)] = FMA(T6P, T6Z, T70); T8S = T8s + T8p; T8t = T8p - T8s; { E T8L, T8T, T8P, T8V; T8L = W[34]; T8Q = W[35]; T8V = W[2]; T8Z = FMA(KP923879532, T8S, T8R); T8T = FNMS(KP923879532, T8S, T8R); T8P = T8L * T8O; T8X = T8V * T8W; T90 = T8V * T8Z; T8U = T8L * T8T; cr[WS(rs, 18)] = FNMS(T8Q, T8T, T8P); T8Y = W[3]; } } { E T89, T8a, T84, T85; T84 = FNMS(KP707106781, T7r, T7k); T7s = FMA(KP707106781, T7r, T7k); ci[WS(rs, 18)] = FMA(T8Q, T8O, T8U); T85 = T7S + T7T; T7U = T7S - T7T; ci[WS(rs, 2)] = FMA(T8Y, T8W, T90); cr[WS(rs, 2)] = FNMS(T8Y, T8Z, T8X); T7R = FMA(KP707106781, T7Q, T7N); T89 = FNMS(KP707106781, T7Q, T7N); T8e = FMA(KP923879532, T85, T84); T86 = FNMS(KP923879532, T85, T84); T8a = T7G + T7z; T7H = T7z - T7G; { E T83, T8b, T87, T8d; T83 = W[26]; T88 = W[27]; T8d = W[58]; T8h = FMA(KP923879532, T8a, T89); T8b = FNMS(KP923879532, T8a, T89); T87 = T83 * T86; T8f = T8d * T8e; T8i = T8d * T8h; T8c = T83 * T8b; cr[WS(rs, 14)] = FNMS(T88, T8b, T87); T8g = W[59]; } } } } { E T5S, T6q, T6n, T6K, T6C, T6b, T6E, T6N, T6L, T6O, T6I, T6M; { E T6F, T6G, T6A, T6B; T6A = T5K - T5R; T5S = T5K + T5R; ci[WS(rs, 14)] = FMA(T88, T86, T8c); T6B = T6p - T6o; T6q = T6o + T6p; ci[WS(rs, 30)] = FMA(T8g, T8e, T8i); cr[WS(rs, 30)] = FNMS(T8g, T8h, T8f); T6n = T6f + T6m; T6F = T6m - T6f; T6K = FMA(KP707106781, T6B, T6A); T6C = FNMS(KP707106781, T6B, T6A); T6G = T61 - T6a; T6b = T61 + T6a; { E T6z, T6H, T6D, T6J; T6z = W[54]; T6E = W[55]; T6J = W[22]; T6N = FMA(KP707106781, T6G, T6F); T6H = FNMS(KP707106781, T6G, T6F); T6D = T6z * T6C; T6L = T6J * T6K; T6O = T6J * T6N; T6I = T6z * T6H; cr[WS(rs, 28)] = FNMS(T6E, T6H, T6D); T6M = W[23]; } } { E T8G, T8F, T8J, T8H, T8I, T8u; ci[WS(rs, 28)] = FMA(T6E, T6C, T6I); ci[WS(rs, 12)] = FMA(T6M, T6K, T6O); cr[WS(rs, 12)] = FNMS(T6M, T6N, T6L); T8G = FMA(KP923879532, T8t, T8m); T8u = FNMS(KP923879532, T8t, T8m); { E T8j, T8w, T8D, T8v, T8E; T8j = W[50]; T8w = W[51]; T8F = W[18]; T8J = FMA(KP923879532, T8C, T8z); T8D = FNMS(KP923879532, T8C, T8z); T8v = T8j * T8u; T8E = T8w * T8u; T8H = T8F * T8G; T8I = W[19]; cr[WS(rs, 26)] = FNMS(T8w, T8D, T8v); ci[WS(rs, 26)] = FMA(T8j, T8D, T8E); } { E T6c, T6u, T6x, T6r, T8K, T5J, T6e; cr[WS(rs, 10)] = FNMS(T8I, T8J, T8H); T8K = T8I * T8G; ci[WS(rs, 10)] = FMA(T8F, T8J, T8K); T6c = FNMS(KP707106781, T6b, T5S); T6u = FMA(KP707106781, T6b, T5S); T6x = FMA(KP707106781, T6q, T6n); T6r = FNMS(KP707106781, T6q, T6n); T5J = W[38]; T6e = W[39]; { E T6t, T6w, T6d, T6s, T6v, T6y; T6t = W[6]; T6w = W[7]; T6d = T5J * T6c; T6s = T6e * T6c; T6v = T6t * T6u; T6y = T6w * T6u; cr[WS(rs, 20)] = FNMS(T6e, T6r, T6d); ci[WS(rs, 20)] = FMA(T5J, T6r, T6s); cr[WS(rs, 4)] = FNMS(T6w, T6x, T6v); ci[WS(rs, 4)] = FMA(T6t, T6x, T6y); } } } } } { E T7c, T7f, T7e, T7g, T7d; { E T71, T74, T79, T76, T75, T7b, T7a; T71 = W[46]; T7c = T72 + T73; T74 = T72 - T73; T7f = T78 + T77; T79 = T77 - T78; T76 = W[47]; T75 = T71 * T74; T7b = W[14]; T7a = T71 * T79; T7e = W[15]; cr[WS(rs, 24)] = FNMS(T76, T79, T75); T7g = T7b * T7f; T7d = T7b * T7c; ci[WS(rs, 24)] = FMA(T76, T74, T7a); } { E T81, T7X, T80, T7Z, T82; ci[WS(rs, 8)] = FMA(T7e, T7c, T7g); cr[WS(rs, 8)] = FNMS(T7e, T7f, T7d); { E T7h, T7Y, T7I, T7V, T7K, T7J, T7W; T7h = W[42]; T7Y = FMA(KP923879532, T7H, T7s); T7I = FNMS(KP923879532, T7H, T7s); T81 = FMA(KP923879532, T7U, T7R); T7V = FNMS(KP923879532, T7U, T7R); T7K = W[43]; T7J = T7h * T7I; T7X = W[10]; T80 = W[11]; T7W = T7K * T7I; cr[WS(rs, 22)] = FNMS(T7K, T7V, T7J); T7Z = T7X * T7Y; T82 = T80 * T7Y; ci[WS(rs, 22)] = FMA(T7h, T7V, T7W); } { E T2P, T37, T1G, T32, T2R, T2Q, T38, T2z, T27, T2y; T2P = FMA(KP923879532, T2O, T2L); T37 = FNMS(KP923879532, T2O, T2L); cr[WS(rs, 6)] = FNMS(T80, T81, T7Z); ci[WS(rs, 6)] = FMA(T7X, T81, T82); T1G = FMA(KP923879532, T1F, T1i); T32 = FNMS(KP923879532, T1F, T1i); T2R = FNMS(KP668178637, T1X, T26); T27 = FMA(KP668178637, T26, T1X); T2y = FNMS(KP668178637, T2x, T2o); T2Q = FMA(KP668178637, T2o, T2x); T38 = T2y + T27; T2z = T27 - T2y; { E T2C, T2A, T3c, T34, T2U, T39, T36, T31; { E T11, T2W, T2S, T33; T11 = W[40]; T2C = W[41]; T2A = FNMS(KP831469612, T2z, T1G); T2W = FMA(KP831469612, T2z, T1G); T2S = T2Q - T2R; T33 = T2Q + T2R; { E T2V, T2B, T2T, T2Z, T2X, T2Y, T30; T2V = W[8]; T2B = T11 * T2A; T3c = FMA(KP831469612, T33, T32); T34 = FNMS(KP831469612, T33, T32); T2T = FNMS(KP831469612, T2S, T2P); T2Z = FMA(KP831469612, T2S, T2P); T2X = T2V * T2W; T2Y = W[9]; T30 = T2V * T2Z; cr[WS(rs, 21)] = FNMS(T2C, T2T, T2B); T2U = T11 * T2T; cr[WS(rs, 5)] = FNMS(T2Y, T2Z, T2X); ci[WS(rs, 5)] = FMA(T2Y, T2W, T30); } } T39 = FNMS(KP831469612, T38, T37); T3f = FMA(KP831469612, T38, T37); ci[WS(rs, 21)] = FMA(T2C, T2A, T2U); T36 = W[25]; T31 = W[24]; { E T3b, T3g, T3a, T35; T3e = W[57]; T3a = T36 * T34; T35 = T31 * T34; T3b = W[56]; T3g = T3e * T3c; ci[WS(rs, 13)] = FMA(T31, T39, T3a); cr[WS(rs, 13)] = FNMS(T36, T39, T35); T3d = T3b * T3c; ci[WS(rs, 29)] = FMA(T3b, T3f, T3g); } } } } } { E T4G, T4J, T4I, T4F, T4K; { E T4z, T4R, T4a, T4M, T4h, T4o, T4C, T4N, T4A, T4B; T4z = FMA(KP923879532, T4y, T4v); T4R = FNMS(KP923879532, T4y, T4v); T4a = FNMS(KP923879532, T49, T42); T4M = FMA(KP923879532, T49, T42); cr[WS(rs, 29)] = FNMS(T3e, T3f, T3d); T4h = FNMS(KP668178637, T4g, T4d); T4A = FMA(KP668178637, T4d, T4g); T4B = FMA(KP668178637, T4k, T4n); T4o = FNMS(KP668178637, T4n, T4k); T4C = T4A - T4B; T4N = T4A + T4B; { E T4W, T4Z, T4q, T4X, T50, T4Y; { E T4L, T4Q, T4O, T4p, T4S, T4P, T4U, T4V, T4T; T4L = W[20]; T4Q = W[21]; T4W = FMA(KP831469612, T4N, T4M); T4O = FNMS(KP831469612, T4N, T4M); T4p = T4h + T4o; T4S = T4h - T4o; T4P = T4L * T4O; T4V = W[52]; T4Z = FNMS(KP831469612, T4S, T4R); T4T = FMA(KP831469612, T4S, T4R); T4q = FNMS(KP831469612, T4p, T4a); T4G = FMA(KP831469612, T4p, T4a); cr[WS(rs, 11)] = FNMS(T4Q, T4T, T4P); T4U = T4L * T4T; T4X = T4V * T4W; T50 = T4V * T4Z; T4Y = W[53]; ci[WS(rs, 11)] = FMA(T4Q, T4O, T4U); } { E T4D, T4s, T3Z, T4E, T4r; T4J = FMA(KP831469612, T4C, T4z); T4D = FNMS(KP831469612, T4C, T4z); T4s = W[37]; ci[WS(rs, 27)] = FMA(T4Y, T4W, T50); cr[WS(rs, 27)] = FNMS(T4Y, T4Z, T4X); T3Z = W[36]; T4E = T4s * T4q; T4I = W[5]; T4r = T3Z * T4q; ci[WS(rs, 19)] = FMA(T3Z, T4D, T4E); T4F = W[4]; T4K = T4I * T4G; cr[WS(rs, 19)] = FNMS(T4s, T4D, T4r); } } } { E T3E, T3H, T3G, T3D, T3I; { E T3x, T3P, T3k, T3K, T3n, T3q, T3A, T3L, T4H, T3y, T3z; T3x = FMA(KP923879532, T3w, T3v); T3P = FNMS(KP923879532, T3w, T3v); T4H = T4F * T4G; ci[WS(rs, 3)] = FMA(T4F, T4J, T4K); T3k = FMA(KP923879532, T3j, T3i); T3K = FNMS(KP923879532, T3j, T3i); T3y = FMA(KP198912367, T3l, T3m); T3n = FNMS(KP198912367, T3m, T3l); cr[WS(rs, 3)] = FNMS(T4I, T4J, T4H); T3z = FNMS(KP198912367, T3o, T3p); T3q = FMA(KP198912367, T3p, T3o); T3A = T3y + T3z; T3L = T3z - T3y; { E T3U, T3X, T3s, T3V, T3Y, T3W; { E T3J, T3O, T3M, T3r, T3Q, T3N, T3S, T3T, T3R; T3J = W[48]; T3O = W[49]; T3U = FMA(KP980785280, T3L, T3K); T3M = FNMS(KP980785280, T3L, T3K); T3r = T3n + T3q; T3Q = T3n - T3q; T3N = T3J * T3M; T3T = W[16]; T3X = FMA(KP980785280, T3Q, T3P); T3R = FNMS(KP980785280, T3Q, T3P); T3s = FNMS(KP980785280, T3r, T3k); T3E = FMA(KP980785280, T3r, T3k); cr[WS(rs, 25)] = FNMS(T3O, T3R, T3N); T3S = T3J * T3R; T3V = T3T * T3U; T3Y = T3T * T3X; T3W = W[17]; ci[WS(rs, 25)] = FMA(T3O, T3M, T3S); } { E T3B, T3u, T3h, T3C, T3t; T3H = FMA(KP980785280, T3A, T3x); T3B = FNMS(KP980785280, T3A, T3x); T3u = W[33]; ci[WS(rs, 9)] = FMA(T3W, T3U, T3Y); cr[WS(rs, 9)] = FNMS(T3W, T3X, T3V); T3h = W[32]; T3C = T3u * T3s; T3G = W[1]; T3t = T3h * T3s; ci[WS(rs, 17)] = FMA(T3h, T3B, T3C); T3D = W[0]; T3I = T3G * T3E; cr[WS(rs, 17)] = FNMS(T3u, T3B, T3t); } } } { E T5h, T5z, T54, T5u, T57, T5a, T5k, T5v, T3F, T5i, T5j; T5h = FMA(KP923879532, T5g, T5f); T5z = FNMS(KP923879532, T5g, T5f); T3F = T3D * T3E; ci[WS(rs, 1)] = FMA(T3D, T3H, T3I); T54 = FNMS(KP923879532, T53, T52); T5u = FMA(KP923879532, T53, T52); T5i = FMA(KP198912367, T55, T56); T57 = FNMS(KP198912367, T56, T55); cr[WS(rs, 1)] = FNMS(T3G, T3H, T3F); T5j = FMA(KP198912367, T58, T59); T5a = FNMS(KP198912367, T59, T58); T5k = T5i - T5j; T5v = T5i + T5j; { E T5E, T5H, T5c, T5F, T5I, T5G; { E T5t, T5y, T5w, T5b, T5A, T5x, T5C, T5D, T5B; T5t = W[28]; T5y = W[29]; T5E = FMA(KP980785280, T5v, T5u); T5w = FNMS(KP980785280, T5v, T5u); T5b = T57 + T5a; T5A = T5a - T57; T5x = T5t * T5w; T5D = W[60]; T5H = FNMS(KP980785280, T5A, T5z); T5B = FMA(KP980785280, T5A, T5z); T5c = FMA(KP980785280, T5b, T54); T5o = FNMS(KP980785280, T5b, T54); cr[WS(rs, 15)] = FNMS(T5y, T5B, T5x); T5C = T5t * T5B; T5F = T5D * T5E; T5I = T5D * T5H; T5G = W[61]; ci[WS(rs, 15)] = FMA(T5y, T5w, T5C); } { E T5l, T5e, T51, T5m, T5d; T5r = FMA(KP980785280, T5k, T5h); T5l = FNMS(KP980785280, T5k, T5h); T5e = W[45]; ci[WS(rs, 31)] = FMA(T5G, T5E, T5I); cr[WS(rs, 31)] = FNMS(T5G, T5H, T5F); T51 = W[44]; T5m = T5e * T5c; T5q = W[13]; T5d = T51 * T5c; ci[WS(rs, 23)] = FMA(T51, T5l, T5m); T5n = W[12]; T5s = T5q * T5o; cr[WS(rs, 23)] = FNMS(T5e, T5l, T5d); } } } } } } } T5p = T5n * T5o; ci[WS(rs, 7)] = FMA(T5n, T5r, T5s); cr[WS(rs, 7)] = FNMS(T5q, T5r, T5p); } } } static const tw_instr twinstr[] = { {TW_FULL, 1, 32}, {TW_NEXT, 1, 0} }; static const hc2hc_desc desc = { 32, "hb_32", twinstr, &GENUS, {236, 62, 198, 0} }; void X(codelet_hb_32) (planner *p) { X(khc2hc_register) (p, hb_32, &desc); } #else /* HAVE_FMA */ /* Generated by: ../../../genfft/gen_hc2hc.native -compact -variables 4 -pipeline-latency 4 -sign 1 -n 32 -dif -name hb_32 -include hb.h */ /* * This function contains 434 FP additions, 208 FP multiplications, * (or, 340 additions, 114 multiplications, 94 fused multiply/add), * 98 stack variables, 7 constants, and 128 memory accesses */ #include "hb.h" static void hb_32(R *cr, R *ci, const R *W, stride rs, INT mb, INT me, INT ms) { DK(KP555570233, +0.555570233019602224742830813948532874374937191); DK(KP831469612, +0.831469612302545237078788377617905756738560812); DK(KP980785280, +0.980785280403230449126182236134239036973933731); DK(KP195090322, +0.195090322016128267848284868477022240927691618); DK(KP923879532, +0.923879532511286756128183189396788286822416626); DK(KP382683432, +0.382683432365089771728459984030398866761344562); DK(KP707106781, +0.707106781186547524400844362104849039284835938); { INT m; for (m = mb, W = W + ((mb - 1) * 62); m < me; m = m + 1, cr = cr + ms, ci = ci - ms, W = W + 62, MAKE_VOLATILE_STRIDE(64, rs)) { E T4o, T6y, T70, T5u, Tf, T12, T5x, T6z, T3m, T3Y, T29, T2y, T4v, T71, T2U; E T3M, Tu, T1U, T6D, T73, T6G, T74, T1h, T2z, T2X, T3o, T4D, T5A, T4K, T5z; E T30, T3n, TK, T1j, T6S, T7w, T6V, T7v, T1y, T2B, T3c, T3S, T4X, T61, T54; E T62, T3f, T3T, TZ, T1A, T6L, T7z, T6O, T7y, T1P, T2C, T35, T3P, T5g, T64; E T5n, T65, T38, T3Q; { E T3, T4m, T24, T4q, T27, T4t, T6, T5s, Ta, T4p, T1X, T5t, T20, T4n, Td; E T4s; { E T1, T2, T22, T23; T1 = cr[0]; T2 = ci[WS(rs, 15)]; T3 = T1 + T2; T4m = T1 - T2; T22 = ci[WS(rs, 27)]; T23 = cr[WS(rs, 20)]; T24 = T22 - T23; T4q = T22 + T23; } { E T25, T26, T4, T5; T25 = ci[WS(rs, 19)]; T26 = cr[WS(rs, 28)]; T27 = T25 - T26; T4t = T25 + T26; T4 = cr[WS(rs, 8)]; T5 = ci[WS(rs, 7)]; T6 = T4 + T5; T5s = T4 - T5; } { E T8, T9, T1V, T1W; T8 = cr[WS(rs, 4)]; T9 = ci[WS(rs, 11)]; Ta = T8 + T9; T4p = T8 - T9; T1V = ci[WS(rs, 31)]; T1W = cr[WS(rs, 16)]; T1X = T1V - T1W; T5t = T1V + T1W; } { E T1Y, T1Z, Tb, Tc; T1Y = ci[WS(rs, 23)]; T1Z = cr[WS(rs, 24)]; T20 = T1Y - T1Z; T4n = T1Y + T1Z; Tb = ci[WS(rs, 3)]; Tc = cr[WS(rs, 12)]; Td = Tb + Tc; T4s = Tb - Tc; } { E T7, Te, T21, T28; T4o = T4m - T4n; T6y = T4m + T4n; T70 = T5t - T5s; T5u = T5s + T5t; T7 = T3 + T6; Te = Ta + Td; Tf = T7 + Te; T12 = T7 - Te; { E T5v, T5w, T3k, T3l; T5v = T4p + T4q; T5w = T4s + T4t; T5x = KP707106781 * (T5v - T5w); T6z = KP707106781 * (T5v + T5w); T3k = T1X - T20; T3l = Ta - Td; T3m = T3k - T3l; T3Y = T3l + T3k; } T21 = T1X + T20; T28 = T24 + T27; T29 = T21 - T28; T2y = T21 + T28; { E T4r, T4u, T2S, T2T; T4r = T4p - T4q; T4u = T4s - T4t; T4v = KP707106781 * (T4r + T4u); T71 = KP707106781 * (T4r - T4u); T2S = T3 - T6; T2T = T27 - T24; T2U = T2S - T2T; T3M = T2S + T2T; } } } { E Ti, T4H, T1c, T4F, T1f, T4I, Tl, T4E, Tp, T4A, T15, T4y, T18, T4B, Ts; E T4x; { E Tg, Th, T1a, T1b; Tg = cr[WS(rs, 2)]; Th = ci[WS(rs, 13)]; Ti = Tg + Th; T4H = Tg - Th; T1a = ci[WS(rs, 29)]; T1b = cr[WS(rs, 18)]; T1c = T1a - T1b; T4F = T1a + T1b; } { E T1d, T1e, Tj, Tk; T1d = ci[WS(rs, 21)]; T1e = cr[WS(rs, 26)]; T1f = T1d - T1e; T4I = T1d + T1e; Tj = cr[WS(rs, 10)]; Tk = ci[WS(rs, 5)]; Tl = Tj + Tk; T4E = Tj - Tk; } { E Tn, To, T13, T14; Tn = ci[WS(rs, 1)]; To = cr[WS(rs, 14)]; Tp = Tn + To; T4A = Tn - To; T13 = ci[WS(rs, 17)]; T14 = cr[WS(rs, 30)]; T15 = T13 - T14; T4y = T13 + T14; } { E T16, T17, Tq, Tr; T16 = ci[WS(rs, 25)]; T17 = cr[WS(rs, 22)]; T18 = T16 - T17; T4B = T16 + T17; Tq = cr[WS(rs, 6)]; Tr = ci[WS(rs, 9)]; Ts = Tq + Tr; T4x = Tq - Tr; } { E Tm, Tt, T6B, T6C; Tm = Ti + Tl; Tt = Tp + Ts; Tu = Tm + Tt; T1U = Tm - Tt; T6B = T4H + T4I; T6C = T4F - T4E; T6D = FNMS(KP923879532, T6C, KP382683432 * T6B); T73 = FMA(KP382683432, T6C, KP923879532 * T6B); } { E T6E, T6F, T19, T1g; T6E = T4A + T4B; T6F = T4x + T4y; T6G = FNMS(KP923879532, T6F, KP382683432 * T6E); T74 = FMA(KP382683432, T6F, KP923879532 * T6E); T19 = T15 + T18; T1g = T1c + T1f; T1h = T19 - T1g; T2z = T1g + T19; } { E T2V, T2W, T4z, T4C; T2V = T15 - T18; T2W = Tp - Ts; T2X = T2V - T2W; T3o = T2W + T2V; T4z = T4x - T4y; T4C = T4A - T4B; T4D = FNMS(KP382683432, T4C, KP923879532 * T4z); T5A = FMA(KP382683432, T4z, KP923879532 * T4C); } { E T4G, T4J, T2Y, T2Z; T4G = T4E + T4F; T4J = T4H - T4I; T4K = FMA(KP923879532, T4G, KP382683432 * T4J); T5z = FNMS(KP382683432, T4G, KP923879532 * T4J); T2Y = Ti - Tl; T2Z = T1c - T1f; T30 = T2Y + T2Z; T3n = T2Y - T2Z; } } { E Ty, T4N, TB, T4Y, T1p, T4O, T1m, T4Z, TI, T52, T1w, T4V, TF, T51, T1t; E T4S; { E Tw, Tx, T1k, T1l; Tw = cr[WS(rs, 1)]; Tx = ci[WS(rs, 14)]; Ty = Tw + Tx; T4N = Tw - Tx; { E Tz, TA, T1n, T1o; Tz = cr[WS(rs, 9)]; TA = ci[WS(rs, 6)]; TB = Tz + TA; T4Y = Tz - TA; T1n = ci[WS(rs, 22)]; T1o = cr[WS(rs, 25)]; T1p = T1n - T1o; T4O = T1n + T1o; } T1k = ci[WS(rs, 30)]; T1l = cr[WS(rs, 17)]; T1m = T1k - T1l; T4Z = T1k + T1l; { E TG, TH, T4T, T1u, T1v, T4U; TG = ci[WS(rs, 2)]; TH = cr[WS(rs, 13)]; T4T = TG - TH; T1u = ci[WS(rs, 18)]; T1v = cr[WS(rs, 29)]; T4U = T1u + T1v; TI = TG + TH; T52 = T4T + T4U; T1w = T1u - T1v; T4V = T4T - T4U; } { E TD, TE, T4Q, T1r, T1s, T4R; TD = cr[WS(rs, 5)]; TE = ci[WS(rs, 10)]; T4Q = TD - TE; T1r = ci[WS(rs, 26)]; T1s = cr[WS(rs, 21)]; T4R = T1r + T1s; TF = TD + TE; T51 = T4Q + T4R; T1t = T1r - T1s; T4S = T4Q - T4R; } } { E TC, TJ, T6Q, T6R; TC = Ty + TB; TJ = TF + TI; TK = TC + TJ; T1j = TC - TJ; T6Q = T4Z - T4Y; T6R = KP707106781 * (T4S - T4V); T6S = T6Q + T6R; T7w = T6Q - T6R; } { E T6T, T6U, T1q, T1x; T6T = T4N + T4O; T6U = KP707106781 * (T51 + T52); T6V = T6T - T6U; T7v = T6T + T6U; T1q = T1m + T1p; T1x = T1t + T1w; T1y = T1q - T1x; T2B = T1q + T1x; } { E T3a, T3b, T4P, T4W; T3a = T1m - T1p; T3b = TF - TI; T3c = T3a - T3b; T3S = T3b + T3a; T4P = T4N - T4O; T4W = KP707106781 * (T4S + T4V); T4X = T4P - T4W; T61 = T4P + T4W; } { E T50, T53, T3d, T3e; T50 = T4Y + T4Z; T53 = KP707106781 * (T51 - T52); T54 = T50 - T53; T62 = T50 + T53; T3d = Ty - TB; T3e = T1w - T1t; T3f = T3d - T3e; T3T = T3d + T3e; } } { E TN, T56, TQ, T5h, T1G, T57, T1D, T5i, TX, T5l, T1N, T5e, TU, T5k, T1K; E T5b; { E TL, TM, T1B, T1C; TL = ci[0]; TM = cr[WS(rs, 15)]; TN = TL + TM; T56 = TL - TM; { E TO, TP, T1E, T1F; TO = cr[WS(rs, 7)]; TP = ci[WS(rs, 8)]; TQ = TO + TP; T5h = TO - TP; T1E = ci[WS(rs, 24)]; T1F = cr[WS(rs, 23)]; T1G = T1E - T1F; T57 = T1E + T1F; } T1B = ci[WS(rs, 16)]; T1C = cr[WS(rs, 31)]; T1D = T1B - T1C; T5i = T1B + T1C; { E TV, TW, T5c, T1L, T1M, T5d; TV = ci[WS(rs, 4)]; TW = cr[WS(rs, 11)]; T5c = TV - TW; T1L = ci[WS(rs, 20)]; T1M = cr[WS(rs, 27)]; T5d = T1L + T1M; TX = TV + TW; T5l = T5c + T5d; T1N = T1L - T1M; T5e = T5c - T5d; } { E TS, TT, T59, T1I, T1J, T5a; TS = cr[WS(rs, 3)]; TT = ci[WS(rs, 12)]; T59 = TS - TT; T1I = ci[WS(rs, 28)]; T1J = cr[WS(rs, 19)]; T5a = T1I + T1J; TU = TS + TT; T5k = T59 + T5a; T1K = T1I - T1J; T5b = T59 - T5a; } } { E TR, TY, T6J, T6K; TR = TN + TQ; TY = TU + TX; TZ = TR + TY; T1A = TR - TY; T6J = KP707106781 * (T5b - T5e); T6K = T5h + T5i; T6L = T6J - T6K; T7z = T6K + T6J; } { E T6M, T6N, T1H, T1O; T6M = T56 + T57; T6N = KP707106781 * (T5k + T5l); T6O = T6M - T6N; T7y = T6M + T6N; T1H = T1D + T1G; T1O = T1K + T1N; T1P = T1H - T1O; T2C = T1H + T1O; } { E T33, T34, T58, T5f; T33 = T1D - T1G; T34 = TU - TX; T35 = T33 - T34; T3P = T34 + T33; T58 = T56 - T57; T5f = KP707106781 * (T5b + T5e); T5g = T58 - T5f; T64 = T58 + T5f; } { E T5j, T5m, T36, T37; T5j = T5h - T5i; T5m = KP707106781 * (T5k - T5l); T5n = T5j - T5m; T65 = T5j + T5m; T36 = TN - TQ; T37 = T1N - T1K; T38 = T36 - T37; T3Q = T36 + T37; } } { E Tv, T10, T2w, T2A, T2D, T2E, T2v, T2x; Tv = Tf + Tu; T10 = TK + TZ; T2w = Tv - T10; T2A = T2y + T2z; T2D = T2B + T2C; T2E = T2A - T2D; cr[0] = Tv + T10; ci[0] = T2A + T2D; T2v = W[30]; T2x = W[31]; cr[WS(rs, 16)] = FNMS(T2x, T2E, T2v * T2w); ci[WS(rs, 16)] = FMA(T2x, T2w, T2v * T2E); } { E T2I, T2O, T2M, T2Q; { E T2G, T2H, T2K, T2L; T2G = Tf - Tu; T2H = T2C - T2B; T2I = T2G - T2H; T2O = T2G + T2H; T2K = T2y - T2z; T2L = TK - TZ; T2M = T2K - T2L; T2Q = T2L + T2K; } { E T2F, T2J, T2N, T2P; T2F = W[46]; T2J = W[47]; cr[WS(rs, 24)] = FNMS(T2J, T2M, T2F * T2I); ci[WS(rs, 24)] = FMA(T2F, T2M, T2J * T2I); T2N = W[14]; T2P = W[15]; cr[WS(rs, 8)] = FNMS(T2P, T2Q, T2N * T2O); ci[WS(rs, 8)] = FMA(T2N, T2Q, T2P * T2O); } } { E T1i, T2a, T2o, T2k, T2d, T2l, T1R, T2p; T1i = T12 + T1h; T2a = T1U + T29; T2o = T29 - T1U; T2k = T12 - T1h; { E T2b, T2c, T1z, T1Q; T2b = T1j + T1y; T2c = T1P - T1A; T2d = KP707106781 * (T2b + T2c); T2l = KP707106781 * (T2c - T2b); T1z = T1j - T1y; T1Q = T1A + T1P; T1R = KP707106781 * (T1z + T1Q); T2p = KP707106781 * (T1z - T1Q); } { E T1S, T2e, T11, T1T; T1S = T1i - T1R; T2e = T2a - T2d; T11 = W[38]; T1T = W[39]; cr[WS(rs, 20)] = FNMS(T1T, T2e, T11 * T1S); ci[WS(rs, 20)] = FMA(T1T, T1S, T11 * T2e); } { E T2s, T2u, T2r, T2t; T2s = T2k + T2l; T2u = T2o + T2p; T2r = W[22]; T2t = W[23]; cr[WS(rs, 12)] = FNMS(T2t, T2u, T2r * T2s); ci[WS(rs, 12)] = FMA(T2r, T2u, T2t * T2s); } { E T2g, T2i, T2f, T2h; T2g = T1i + T1R; T2i = T2a + T2d; T2f = W[6]; T2h = W[7]; cr[WS(rs, 4)] = FNMS(T2h, T2i, T2f * T2g); ci[WS(rs, 4)] = FMA(T2h, T2g, T2f * T2i); } { E T2m, T2q, T2j, T2n; T2m = T2k - T2l; T2q = T2o - T2p; T2j = W[54]; T2n = W[55]; cr[WS(rs, 28)] = FNMS(T2n, T2q, T2j * T2m); ci[WS(rs, 28)] = FMA(T2j, T2q, T2n * T2m); } } { E T3O, T4a, T40, T4e, T3V, T4f, T43, T4b, T3N, T3Z; T3N = KP707106781 * (T3n + T3o); T3O = T3M - T3N; T4a = T3M + T3N; T3Z = KP707106781 * (T30 + T2X); T40 = T3Y - T3Z; T4e = T3Y + T3Z; { E T3R, T3U, T41, T42; T3R = FNMS(KP382683432, T3Q, KP923879532 * T3P); T3U = FMA(KP923879532, T3S, KP382683432 * T3T); T3V = T3R - T3U; T4f = T3U + T3R; T41 = FNMS(KP382683432, T3S, KP923879532 * T3T); T42 = FMA(KP382683432, T3P, KP923879532 * T3Q); T43 = T41 - T42; T4b = T41 + T42; } { E T3W, T44, T3L, T3X; T3W = T3O - T3V; T44 = T40 - T43; T3L = W[50]; T3X = W[51]; cr[WS(rs, 26)] = FNMS(T3X, T44, T3L * T3W); ci[WS(rs, 26)] = FMA(T3X, T3W, T3L * T44); } { E T4i, T4k, T4h, T4j; T4i = T4a + T4b; T4k = T4e + T4f; T4h = W[2]; T4j = W[3]; cr[WS(rs, 2)] = FNMS(T4j, T4k, T4h * T4i); ci[WS(rs, 2)] = FMA(T4h, T4k, T4j * T4i); } { E T46, T48, T45, T47; T46 = T3O + T3V; T48 = T40 + T43; T45 = W[18]; T47 = W[19]; cr[WS(rs, 10)] = FNMS(T47, T48, T45 * T46); ci[WS(rs, 10)] = FMA(T47, T46, T45 * T48); } { E T4c, T4g, T49, T4d; T4c = T4a - T4b; T4g = T4e - T4f; T49 = W[34]; T4d = W[35]; cr[WS(rs, 18)] = FNMS(T4d, T4g, T49 * T4c); ci[WS(rs, 18)] = FMA(T49, T4g, T4d * T4c); } } { E T32, T3A, T3q, T3E, T3h, T3F, T3t, T3B, T31, T3p; T31 = KP707106781 * (T2X - T30); T32 = T2U - T31; T3A = T2U + T31; T3p = KP707106781 * (T3n - T3o); T3q = T3m - T3p; T3E = T3m + T3p; { E T39, T3g, T3r, T3s; T39 = FNMS(KP923879532, T38, KP382683432 * T35); T3g = FMA(KP382683432, T3c, KP923879532 * T3f); T3h = T39 - T3g; T3F = T3g + T39; T3r = FNMS(KP923879532, T3c, KP382683432 * T3f); T3s = FMA(KP923879532, T35, KP382683432 * T38); T3t = T3r - T3s; T3B = T3r + T3s; } { E T3i, T3u, T2R, T3j; T3i = T32 - T3h; T3u = T3q - T3t; T2R = W[58]; T3j = W[59]; cr[WS(rs, 30)] = FNMS(T3j, T3u, T2R * T3i); ci[WS(rs, 30)] = FMA(T3j, T3i, T2R * T3u); } { E T3I, T3K, T3H, T3J; T3I = T3A + T3B; T3K = T3E + T3F; T3H = W[10]; T3J = W[11]; cr[WS(rs, 6)] = FNMS(T3J, T3K, T3H * T3I); ci[WS(rs, 6)] = FMA(T3H, T3K, T3J * T3I); } { E T3w, T3y, T3v, T3x; T3w = T32 + T3h; T3y = T3q + T3t; T3v = W[26]; T3x = W[27]; cr[WS(rs, 14)] = FNMS(T3x, T3y, T3v * T3w); ci[WS(rs, 14)] = FMA(T3x, T3w, T3v * T3y); } { E T3C, T3G, T3z, T3D; T3C = T3A - T3B; T3G = T3E - T3F; T3z = W[42]; T3D = W[43]; cr[WS(rs, 22)] = FNMS(T3D, T3G, T3z * T3C); ci[WS(rs, 22)] = FMA(T3z, T3G, T3D * T3C); } } { E T60, T6m, T6f, T6n, T67, T6r, T6c, T6q; { E T5Y, T5Z, T6d, T6e; T5Y = T4o + T4v; T5Z = T5z + T5A; T60 = T5Y + T5Z; T6m = T5Y - T5Z; T6d = FMA(KP195090322, T61, KP980785280 * T62); T6e = FNMS(KP195090322, T64, KP980785280 * T65); T6f = T6d + T6e; T6n = T6e - T6d; } { E T63, T66, T6a, T6b; T63 = FNMS(KP195090322, T62, KP980785280 * T61); T66 = FMA(KP980785280, T64, KP195090322 * T65); T67 = T63 + T66; T6r = T63 - T66; T6a = T5u + T5x; T6b = T4K + T4D; T6c = T6a + T6b; T6q = T6a - T6b; } { E T68, T6g, T5X, T69; T68 = T60 - T67; T6g = T6c - T6f; T5X = W[32]; T69 = W[33]; cr[WS(rs, 17)] = FNMS(T69, T6g, T5X * T68); ci[WS(rs, 17)] = FMA(T69, T68, T5X * T6g); } { E T6u, T6w, T6t, T6v; T6u = T6m + T6n; T6w = T6q + T6r; T6t = W[16]; T6v = W[17]; cr[WS(rs, 9)] = FNMS(T6v, T6w, T6t * T6u); ci[WS(rs, 9)] = FMA(T6t, T6w, T6v * T6u); } { E T6i, T6k, T6h, T6j; T6i = T60 + T67; T6k = T6c + T6f; T6h = W[0]; T6j = W[1]; cr[WS(rs, 1)] = FNMS(T6j, T6k, T6h * T6i); ci[WS(rs, 1)] = FMA(T6j, T6i, T6h * T6k); } { E T6o, T6s, T6l, T6p; T6o = T6m - T6n; T6s = T6q - T6r; T6l = W[48]; T6p = W[49]; cr[WS(rs, 25)] = FNMS(T6p, T6s, T6l * T6o); ci[WS(rs, 25)] = FMA(T6l, T6s, T6p * T6o); } } { E T7u, T7Q, T7J, T7R, T7B, T7V, T7G, T7U; { E T7s, T7t, T7H, T7I; T7s = T6y + T6z; T7t = T73 + T74; T7u = T7s - T7t; T7Q = T7s + T7t; T7H = FMA(KP195090322, T7w, KP980785280 * T7v); T7I = FMA(KP195090322, T7z, KP980785280 * T7y); T7J = T7H - T7I; T7R = T7H + T7I; } { E T7x, T7A, T7E, T7F; T7x = FNMS(KP980785280, T7w, KP195090322 * T7v); T7A = FNMS(KP980785280, T7z, KP195090322 * T7y); T7B = T7x + T7A; T7V = T7x - T7A; T7E = T70 - T71; T7F = T6D - T6G; T7G = T7E + T7F; T7U = T7E - T7F; } { E T7C, T7K, T7r, T7D; T7C = T7u - T7B; T7K = T7G - T7J; T7r = W[44]; T7D = W[45]; cr[WS(rs, 23)] = FNMS(T7D, T7K, T7r * T7C); ci[WS(rs, 23)] = FMA(T7D, T7C, T7r * T7K); } { E T7Y, T80, T7X, T7Z; T7Y = T7Q + T7R; T80 = T7U - T7V; T7X = W[60]; T7Z = W[61]; cr[WS(rs, 31)] = FNMS(T7Z, T80, T7X * T7Y); ci[WS(rs, 31)] = FMA(T7X, T80, T7Z * T7Y); } { E T7M, T7O, T7L, T7N; T7M = T7u + T7B; T7O = T7G + T7J; T7L = W[12]; T7N = W[13]; cr[WS(rs, 7)] = FNMS(T7N, T7O, T7L * T7M); ci[WS(rs, 7)] = FMA(T7N, T7M, T7L * T7O); } { E T7S, T7W, T7P, T7T; T7S = T7Q - T7R; T7W = T7U + T7V; T7P = W[28]; T7T = W[29]; cr[WS(rs, 15)] = FNMS(T7T, T7W, T7P * T7S); ci[WS(rs, 15)] = FMA(T7P, T7W, T7T * T7S); } } { E T4M, T5M, T5F, T5N, T5p, T5R, T5C, T5Q; { E T4w, T4L, T5D, T5E; T4w = T4o - T4v; T4L = T4D - T4K; T4M = T4w + T4L; T5M = T4w - T4L; T5D = FMA(KP831469612, T4X, KP555570233 * T54); T5E = FNMS(KP831469612, T5g, KP555570233 * T5n); T5F = T5D + T5E; T5N = T5E - T5D; } { E T55, T5o, T5y, T5B; T55 = FNMS(KP831469612, T54, KP555570233 * T4X); T5o = FMA(KP555570233, T5g, KP831469612 * T5n); T5p = T55 + T5o; T5R = T55 - T5o; T5y = T5u - T5x; T5B = T5z - T5A; T5C = T5y + T5B; T5Q = T5y - T5B; } { E T5q, T5G, T4l, T5r; T5q = T4M - T5p; T5G = T5C - T5F; T4l = W[40]; T5r = W[41]; cr[WS(rs, 21)] = FNMS(T5r, T5G, T4l * T5q); ci[WS(rs, 21)] = FMA(T5r, T5q, T4l * T5G); } { E T5U, T5W, T5T, T5V; T5U = T5M + T5N; T5W = T5Q + T5R; T5T = W[24]; T5V = W[25]; cr[WS(rs, 13)] = FNMS(T5V, T5W, T5T * T5U); ci[WS(rs, 13)] = FMA(T5T, T5W, T5V * T5U); } { E T5I, T5K, T5H, T5J; T5I = T4M + T5p; T5K = T5C + T5F; T5H = W[8]; T5J = W[9]; cr[WS(rs, 5)] = FNMS(T5J, T5K, T5H * T5I); ci[WS(rs, 5)] = FMA(T5J, T5I, T5H * T5K); } { E T5O, T5S, T5L, T5P; T5O = T5M - T5N; T5S = T5Q - T5R; T5L = W[56]; T5P = W[57]; cr[WS(rs, 29)] = FNMS(T5P, T5S, T5L * T5O); ci[WS(rs, 29)] = FMA(T5L, T5S, T5P * T5O); } } { E T6I, T7g, T79, T7h, T6X, T7l, T76, T7k; { E T6A, T6H, T77, T78; T6A = T6y - T6z; T6H = T6D + T6G; T6I = T6A - T6H; T7g = T6A + T6H; T77 = FNMS(KP555570233, T6S, KP831469612 * T6V); T78 = FMA(KP555570233, T6L, KP831469612 * T6O); T79 = T77 - T78; T7h = T77 + T78; } { E T6P, T6W, T72, T75; T6P = FNMS(KP555570233, T6O, KP831469612 * T6L); T6W = FMA(KP831469612, T6S, KP555570233 * T6V); T6X = T6P - T6W; T7l = T6W + T6P; T72 = T70 + T71; T75 = T73 - T74; T76 = T72 - T75; T7k = T72 + T75; } { E T6Y, T7a, T6x, T6Z; T6Y = T6I - T6X; T7a = T76 - T79; T6x = W[52]; T6Z = W[53]; cr[WS(rs, 27)] = FNMS(T6Z, T7a, T6x * T6Y); ci[WS(rs, 27)] = FMA(T6Z, T6Y, T6x * T7a); } { E T7o, T7q, T7n, T7p; T7o = T7g + T7h; T7q = T7k + T7l; T7n = W[4]; T7p = W[5]; cr[WS(rs, 3)] = FNMS(T7p, T7q, T7n * T7o); ci[WS(rs, 3)] = FMA(T7n, T7q, T7p * T7o); } { E T7c, T7e, T7b, T7d; T7c = T6I + T6X; T7e = T76 + T79; T7b = W[20]; T7d = W[21]; cr[WS(rs, 11)] = FNMS(T7d, T7e, T7b * T7c); ci[WS(rs, 11)] = FMA(T7d, T7c, T7b * T7e); } { E T7i, T7m, T7f, T7j; T7i = T7g - T7h; T7m = T7k - T7l; T7f = W[36]; T7j = W[37]; cr[WS(rs, 19)] = FNMS(T7j, T7m, T7f * T7i); ci[WS(rs, 19)] = FMA(T7f, T7m, T7j * T7i); } } } } } static const tw_instr twinstr[] = { {TW_FULL, 1, 32}, {TW_NEXT, 1, 0} }; static const hc2hc_desc desc = { 32, "hb_32", twinstr, &GENUS, {340, 114, 94, 0} }; void X(codelet_hb_32) (planner *p) { X(khc2hc_register) (p, hb_32, &desc); } #endif /* HAVE_FMA */ fftw-3.3.3/rdft/scalar/r2cb/hb_64.c0000644000175000017500000032665712054411023013524 00000000000000/* * Copyright (c) 2003, 2007-11 Matteo Frigo * Copyright (c) 2003, 2007-11 Massachusetts Institute of Technology * * This program is free software; you can redistribute it and/or modify * it under the terms of the GNU General Public License as published by * the Free Software Foundation; either version 2 of the License, or * (at your option) any later version. * * This program is distributed in the hope that it will be useful, * but WITHOUT ANY WARRANTY; without even the implied warranty of * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the * GNU General Public License for more details. * * You should have received a copy of the GNU General Public License * along with this program; if not, write to the Free Software * Foundation, Inc., 51 Franklin Street, Fifth Floor, Boston, MA 02110-1301 USA * */ /* This file was automatically generated --- DO NOT EDIT */ /* Generated on Sun Nov 25 07:41:15 EST 2012 */ #include "codelet-rdft.h" #ifdef HAVE_FMA /* Generated by: ../../../genfft/gen_hc2hc.native -fma -reorder-insns -schedule-for-pipeline -compact -variables 4 -pipeline-latency 4 -sign 1 -n 64 -dif -name hb_64 -include hb.h */ /* * This function contains 1038 FP additions, 644 FP multiplications, * (or, 520 additions, 126 multiplications, 518 fused multiply/add), * 231 stack variables, 15 constants, and 256 memory accesses */ #include "hb.h" static void hb_64(R *cr, R *ci, const R *W, stride rs, INT mb, INT me, INT ms) { DK(KP881921264, +0.881921264348355029712756863660388349508442621); DK(KP534511135, +0.534511135950791641089685961295362908582039528); DK(KP956940335, +0.956940335732208864935797886980269969482849206); DK(KP303346683, +0.303346683607342391675883946941299872384187453); DK(KP995184726, +0.995184726672196886244836953109479921575474869); DK(KP098491403, +0.098491403357164253077197521291327432293052451); DK(KP831469612, +0.831469612302545237078788377617905756738560812); DK(KP773010453, +0.773010453362736960810906609758469800971041293); DK(KP980785280, +0.980785280403230449126182236134239036973933731); DK(KP820678790, +0.820678790828660330972281985331011598767386482); DK(KP923879532, +0.923879532511286756128183189396788286822416626); DK(KP198912367, +0.198912367379658006911597622644676228597850501); DK(KP668178637, +0.668178637919298919997757686523080761552472251); DK(KP414213562, +0.414213562373095048801688724209698078569671875); DK(KP707106781, +0.707106781186547524400844362104849039284835938); { INT m; for (m = mb, W = W + ((mb - 1) * 126); m < me; m = m + 1, cr = cr + ms, ci = ci - ms, W = W + 126, MAKE_VOLATILE_STRIDE(128, rs)) { E Tcx, Tcw, Tcv; { E Thy, Tv, T7n, T5B, TfP, Tey, Tkl, TjB, T6U, T2k, T7o, T2H, TiH, Tia, Tk8; E Tj8, T6V, T5E, Tbz, T9N, Tb7, T9Q, Tgh, Tev, Tb6, T8G, TbA, T8N, TfO, TcU; E Tgi, Td5, Ti3, T10, TjC, Tje, TiI, ThF, TeA, Tds, TjD, Tjb, TeB, Tdh, Tgl; E TfT, Tgk, TfW, T6Z, T7r, T5H, T39, Tbb, TbC, T9S, T8V, T72, T7q, T5G, T3A; E Tbe, TbD, T9T, T92, ThH, T1w, Tke, Tjq, Tkf, Tjt, TiK, ThO, Tgb, TgT, Tfc; E Tec, Tg8, TgU, Tfd, Tel, T77, T83, T6i, T5a, T7a, T82, T6j, T5n, Tbj, Tcc; E Tas, T9f, Tbm, Tcb, Tar, T9m, ThQ, T21, Tkb, Tjj, Tkc, Tjm, TiL, ThX, Tg4; E TgW, Tf9, TdL, Tg1, TgX, Tfa, TdU, T7e, T80, T6f, T4h, T9q, Tbr, T7h, T7Z; E T6g, T4u, T9D, T9C, Tbo, T9B, Tbp, T9x; { E T3v, T8Z, T8W, T90, T8X, T3y, T3q, T70; { E TcQ, TcT, Td4, TcZ; { E T24, T5t, T7, T27, T5w, Ti4, Tet, T2i, T5z, Te, Teu, Ti5, T5y, T2d, T8H; E T2u, Td0, Tm, Ti7, Td3, T8I, T2p, Tq, T2w, Tp, TcV, T2E, Tr, T2x, T2y; E Tes, Ter; { E T1, T2, T4, T5, T5u, T5v; T1 = cr[0]; T2 = ci[WS(rs, 31)]; T4 = cr[WS(rs, 16)]; T5 = ci[WS(rs, 15)]; { E T25, T3, T6, T26; T25 = ci[WS(rs, 47)]; T24 = T1 - T2; T3 = T1 + T2; T5t = T4 - T5; T6 = T4 + T5; T26 = cr[WS(rs, 48)]; T5u = ci[WS(rs, 63)]; T5v = cr[WS(rs, 32)]; TcQ = T3 - T6; T7 = T3 + T6; Tes = T25 - T26; T27 = T25 + T26; } Ter = T5u - T5v; T5w = T5u + T5v; } { E Ta, T29, Tb, TcR, T2h, Tc, T2a, T2b; { E T2f, T2g, T8, T9; T8 = cr[WS(rs, 8)]; T9 = ci[WS(rs, 23)]; Ti4 = Ter + Tes; Tet = Ter - Tes; T2f = ci[WS(rs, 39)]; T2g = cr[WS(rs, 56)]; Ta = T8 + T9; T29 = T8 - T9; Tb = ci[WS(rs, 7)]; TcR = T2f - T2g; T2h = T2f + T2g; Tc = cr[WS(rs, 24)]; T2a = ci[WS(rs, 55)]; T2b = cr[WS(rs, 40)]; } { E Tj, T2l, Ti, Td1, T2t, Tk, T2m, T2n; { E Tg, Th, T2r, T2s; Tg = cr[WS(rs, 4)]; { E T2e, Td, TcS, T2c; T2e = Tb - Tc; Td = Tb + Tc; TcS = T2a - T2b; T2c = T2a + T2b; T2i = T2e - T2h; T5z = T2e + T2h; Te = Ta + Td; Teu = Ta - Td; TcT = TcR - TcS; Ti5 = TcS + TcR; T5y = T29 + T2c; T2d = T29 - T2c; Th = ci[WS(rs, 27)]; } T2r = ci[WS(rs, 59)]; T2s = cr[WS(rs, 36)]; Tj = cr[WS(rs, 20)]; T2l = Tg - Th; Ti = Tg + Th; Td1 = T2r - T2s; T2t = T2r + T2s; Tk = ci[WS(rs, 11)]; T2m = ci[WS(rs, 43)]; T2n = cr[WS(rs, 52)]; } { E Tn, To, T2C, T2D; Tn = ci[WS(rs, 3)]; { E T2q, Tl, Td2, T2o; T2q = Tj - Tk; Tl = Tj + Tk; Td2 = T2m - T2n; T2o = T2m + T2n; T8H = T2t - T2q; T2u = T2q + T2t; Td0 = Ti - Tl; Tm = Ti + Tl; Ti7 = Td1 + Td2; Td3 = Td1 - Td2; T8I = T2l + T2o; T2p = T2l - T2o; To = cr[WS(rs, 28)]; } T2C = ci[WS(rs, 35)]; T2D = cr[WS(rs, 60)]; Tq = cr[WS(rs, 12)]; T2w = Tn - To; Tp = Tn + To; TcV = T2C - T2D; T2E = T2C + T2D; Tr = ci[WS(rs, 19)]; T2x = ci[WS(rs, 51)]; T2y = cr[WS(rs, 44)]; } } } { E Tj6, T8K, T8L, T9L, T8F, Ti6, T8E, T9M, T5C, T5D, Ti9, Tj7; { E T2F, Ti8, T2A, TjA, Tew, Tex, Tjz; { E Tf, TcY, TcX, Tu, T5x, T5A; Tj6 = T7 - Te; Tf = T7 + Te; { E T2B, Ts, TcW, T2z, Tt; T2B = Tq - Tr; Ts = Tq + Tr; TcW = T2x - T2y; T2z = T2x + T2y; T8K = T2B + T2E; T2F = T2B - T2E; TcY = Tp - Ts; Tt = Tp + Ts; TcX = TcV - TcW; Ti8 = TcV + TcW; T8L = T2w + T2z; T2A = T2w - T2z; Tu = Tm + Tt; TjA = Tm - Tt; } T9L = T5w - T5t; T5x = T5t + T5w; T5A = T5y - T5z; T8F = T5y + T5z; Td4 = Td0 + Td3; Tew = Td0 - Td3; Thy = Tf - Tu; Tv = Tf + Tu; T7n = FNMS(KP707106781, T5A, T5x); T5B = FMA(KP707106781, T5A, T5x); Tex = TcY + TcX; TcZ = TcX - TcY; Ti6 = Ti4 + Ti5; Tjz = Ti4 - Ti5; } { E T28, T2j, T2v, T2G; T8E = T24 + T27; T28 = T24 - T27; TfP = Tew + Tex; Tey = Tew - Tex; Tkl = TjA + Tjz; TjB = Tjz - TjA; T2j = T2d + T2i; T9M = T2d - T2i; T5C = FMA(KP414213562, T2p, T2u); T2v = FNMS(KP414213562, T2u, T2p); T2G = FMA(KP414213562, T2F, T2A); T5D = FNMS(KP414213562, T2A, T2F); T6U = FNMS(KP707106781, T2j, T28); T2k = FMA(KP707106781, T2j, T28); T7o = T2v - T2G; T2H = T2v + T2G; Ti9 = Ti7 + Ti8; Tj7 = Ti8 - Ti7; } } { E T8J, T9O, T9P, T8M; TiH = Ti6 + Ti9; Tia = Ti6 - Ti9; Tk8 = Tj6 + Tj7; Tj8 = Tj6 - Tj7; T8J = FNMS(KP414213562, T8I, T8H); T9O = FMA(KP414213562, T8H, T8I); T6V = T5D - T5C; T5E = T5C + T5D; Tbz = FNMS(KP707106781, T9M, T9L); T9N = FMA(KP707106781, T9M, T9L); T9P = FMA(KP414213562, T8K, T8L); T8M = FNMS(KP414213562, T8L, T8K); Tb7 = T9O + T9P; T9Q = T9O - T9P; Tgh = Teu + Tet; Tev = Tet - Teu; Tb6 = FMA(KP707106781, T8F, T8E); T8G = FNMS(KP707106781, T8F, T8E); TbA = T8M - T8J; T8N = T8J + T8M; } } } { E T8S, TC, Tdn, Tdk, ThC, T3e, T8P, T36, T2X, Tda, TY, ThA, Tdf, T35, T2S; E T3x, T3o, Tdl, TJ, ThD, Tdq, T3w, T3j, T34, TR, Tdc, Td9, Thz, T2N; { E TV, T2O, TU, Tdd, T2W, TW, T2P, T2Q; { E Tz, T3r, Ty, Tdj, T3u, TA, T3b, T3c; { E Tw, Tx, T3s, T3t; Tw = cr[WS(rs, 2)]; TfO = TcQ + TcT; TcU = TcQ - TcT; Tgi = Td4 + TcZ; Td5 = TcZ - Td4; Tx = ci[WS(rs, 29)]; T3s = ci[WS(rs, 45)]; T3t = cr[WS(rs, 50)]; Tz = cr[WS(rs, 18)]; T3r = Tw - Tx; Ty = Tw + Tx; Tdj = T3s - T3t; T3u = T3s + T3t; TA = ci[WS(rs, 13)]; T3b = ci[WS(rs, 61)]; T3c = cr[WS(rs, 34)]; } { E T3a, TB, Tdi, T3d; T8S = T3r + T3u; T3v = T3r - T3u; T3a = Tz - TA; TB = Tz + TA; Tdi = T3b - T3c; T3d = T3b + T3c; TC = Ty + TB; Tdn = Ty - TB; Tdk = Tdi - Tdj; ThC = Tdi + Tdj; T3e = T3a + T3d; T8P = T3d - T3a; } } { E TS, TT, T2U, T2V; TS = cr[WS(rs, 6)]; TT = ci[WS(rs, 25)]; T2U = ci[WS(rs, 41)]; T2V = cr[WS(rs, 54)]; TV = ci[WS(rs, 9)]; T2O = TS - TT; TU = TS + TT; Tdd = T2U - T2V; T2W = T2U + T2V; TW = cr[WS(rs, 22)]; T2P = ci[WS(rs, 57)]; T2Q = cr[WS(rs, 38)]; } { E TG, T3f, TF, Tdo, T3n, TH, T3g, T3h; { E TD, TE, T3l, T3m; TD = cr[WS(rs, 10)]; { E T2T, TX, Tde, T2R; T2T = TV - TW; TX = TV + TW; Tde = T2P - T2Q; T2R = T2P + T2Q; T36 = T2T - T2W; T2X = T2T + T2W; Tda = TU - TX; TY = TU + TX; ThA = Tde + Tdd; Tdf = Tdd - Tde; T35 = T2O - T2R; T2S = T2O + T2R; TE = ci[WS(rs, 21)]; } T3l = ci[WS(rs, 37)]; T3m = cr[WS(rs, 58)]; TG = ci[WS(rs, 5)]; T3f = TD - TE; TF = TD + TE; Tdo = T3l - T3m; T3n = T3l + T3m; TH = cr[WS(rs, 26)]; T3g = ci[WS(rs, 53)]; T3h = cr[WS(rs, 42)]; } { E TO, T30, TN, Td8, T33, TP, T2K, T2L; { E TL, TM, T31, T32; TL = ci[WS(rs, 1)]; { E T3k, TI, Tdp, T3i; T3k = TG - TH; TI = TG + TH; Tdp = T3g - T3h; T3i = T3g + T3h; T3x = T3k - T3n; T3o = T3k + T3n; Tdl = TF - TI; TJ = TF + TI; ThD = Tdp + Tdo; Tdq = Tdo - Tdp; T3w = T3f - T3i; T3j = T3f + T3i; TM = cr[WS(rs, 30)]; } T31 = ci[WS(rs, 49)]; T32 = cr[WS(rs, 46)]; TO = cr[WS(rs, 14)]; T30 = TL - TM; TN = TL + TM; Td8 = T31 - T32; T33 = T31 + T32; TP = ci[WS(rs, 17)]; T2K = ci[WS(rs, 33)]; T2L = cr[WS(rs, 62)]; } { E T2J, TQ, Td7, T2M; T8Z = T30 + T33; T34 = T30 - T33; T2J = TO - TP; TQ = TO + TP; Td7 = T2K - T2L; T2M = T2K + T2L; TR = TN + TQ; Tdc = TN - TQ; Td9 = Td7 - Td8; Thz = Td7 + Td8; T2N = T2J - T2M; T8W = T2J + T2M; } } } } { E Tja, Tj9, TfU, TfV, TfR, Tdb, Tdg, TfS; { E ThE, ThB, Tdm, Tdr; { E Tjc, TK, TZ, Tjd; Tjc = TC - TJ; TK = TC + TJ; TZ = TR + TY; Tja = TR - TY; Tjd = ThC - ThD; ThE = ThC + ThD; Tj9 = Thz - ThA; ThB = Thz + ThA; Ti3 = TK - TZ; T10 = TK + TZ; TjC = Tjc - Tjd; Tje = Tjc + Tjd; } TfU = Tdl + Tdk; Tdm = Tdk - Tdl; Tdr = Tdn - Tdq; TfV = Tdn + Tdq; TiI = ThE + ThB; ThF = ThB - ThE; TeA = FMA(KP414213562, Tdm, Tdr); Tds = FNMS(KP414213562, Tdr, Tdm); TfR = Tda + Td9; Tdb = Td9 - Tda; Tdg = Tdc - Tdf; TfS = Tdc + Tdf; } { E T2Z, T6X, T37, T2Y; TjD = Tja + Tj9; Tjb = Tj9 - Tja; TeB = FNMS(KP414213562, Tdb, Tdg); Tdh = FMA(KP414213562, Tdg, Tdb); T90 = T2S + T2X; T2Y = T2S - T2X; Tgl = FMA(KP414213562, TfR, TfS); TfT = FNMS(KP414213562, TfS, TfR); Tgk = FNMS(KP414213562, TfU, TfV); TfW = FMA(KP414213562, TfV, TfU); T2Z = FMA(KP707106781, T2Y, T2N); T6X = FNMS(KP707106781, T2Y, T2N); T37 = T35 + T36; T8X = T35 - T36; { E T8Q, T8T, T3p, T6Y, T38; T3y = T3w + T3x; T8Q = T3x - T3w; T8T = T3j + T3o; T3p = T3j - T3o; T6Y = FNMS(KP707106781, T37, T34); T38 = FMA(KP707106781, T37, T34); { E Tb9, T8R, Tba, T8U; Tb9 = FMA(KP707106781, T8Q, T8P); T8R = FNMS(KP707106781, T8Q, T8P); Tba = FMA(KP707106781, T8T, T8S); T8U = FNMS(KP707106781, T8T, T8S); T6Z = FMA(KP668178637, T6Y, T6X); T7r = FNMS(KP668178637, T6X, T6Y); T5H = FMA(KP198912367, T2Z, T38); T39 = FNMS(KP198912367, T38, T2Z); Tbb = FNMS(KP198912367, Tba, Tb9); TbC = FMA(KP198912367, Tb9, Tba); T9S = FNMS(KP668178637, T8R, T8U); T8V = FMA(KP668178637, T8U, T8R); T3q = FMA(KP707106781, T3p, T3e); T70 = FNMS(KP707106781, T3p, T3e); } } } } } } { E T97, Tbk, T9j, T9k, Tbh, T9i, Tbi, T9e; { E T9g, T5f, T18, Ted, TdY, ThI, T4A, T95, T9b, T57, T1u, Te1, Te4, ThM, T52; E T9c, T5h, T4K, TdZ, T1f, ThJ, Teg, T5g, T4F, T1j, Te8, T98, T4W, T4N, T1m; E Te7, T4Q, T1n, Te6; { E T1q, Te3, T4Y, T1t, Te2, T51; { E T15, T5b, T14, TdX, T5e, T16, T4x, T4y; { E T12, T13, T5c, T5d, T71, T3z; T12 = cr[WS(rs, 1)]; T71 = FNMS(KP707106781, T3y, T3v); T3z = FMA(KP707106781, T3y, T3v); { E Tbc, T8Y, Tbd, T91; Tbc = FMA(KP707106781, T8X, T8W); T8Y = FNMS(KP707106781, T8X, T8W); Tbd = FMA(KP707106781, T90, T8Z); T91 = FNMS(KP707106781, T90, T8Z); T72 = FNMS(KP668178637, T71, T70); T7q = FMA(KP668178637, T70, T71); T5G = FNMS(KP198912367, T3q, T3z); T3A = FMA(KP198912367, T3z, T3q); Tbe = FNMS(KP198912367, Tbd, Tbc); TbD = FMA(KP198912367, Tbc, Tbd); T9T = FNMS(KP668178637, T8Y, T91); T92 = FMA(KP668178637, T91, T8Y); T13 = ci[WS(rs, 30)]; } T5c = ci[WS(rs, 46)]; T5d = cr[WS(rs, 49)]; T15 = cr[WS(rs, 17)]; T5b = T12 - T13; T14 = T12 + T13; TdX = T5c - T5d; T5e = T5c + T5d; T16 = ci[WS(rs, 14)]; T4x = ci[WS(rs, 62)]; T4y = cr[WS(rs, 33)]; } { E T4w, T17, TdW, T4z; T9g = T5b + T5e; T5f = T5b - T5e; T4w = T15 - T16; T17 = T15 + T16; TdW = T4x - T4y; T4z = T4x + T4y; T18 = T14 + T17; Ted = T14 - T17; TdY = TdW - TdX; ThI = TdW + TdX; T4A = T4w + T4z; T95 = T4z - T4w; } } { E T1r, T53, T56, T1s, T4Z, T50; { E T1o, T1p, T54, T55; T1o = ci[WS(rs, 2)]; T1p = cr[WS(rs, 29)]; T54 = ci[WS(rs, 50)]; T55 = cr[WS(rs, 45)]; T1r = cr[WS(rs, 13)]; T53 = T1o - T1p; T1q = T1o + T1p; Te3 = T54 - T55; T56 = T54 + T55; T1s = ci[WS(rs, 18)]; T4Z = ci[WS(rs, 34)]; T50 = cr[WS(rs, 61)]; } T9b = T53 + T56; T57 = T53 - T56; T4Y = T1r - T1s; T1t = T1r + T1s; Te2 = T4Z - T50; T51 = T4Z + T50; } T1u = T1q + T1t; Te1 = T1q - T1t; Te4 = Te2 - Te3; ThM = Te2 + Te3; T52 = T4Y - T51; T9c = T4Y + T51; { E T1c, T4B, T1b, Tee, T4J, T1d, T4C, T4D; { E T19, T1a, T4H, T4I; T19 = cr[WS(rs, 9)]; T1a = ci[WS(rs, 22)]; T4H = ci[WS(rs, 38)]; T4I = cr[WS(rs, 57)]; T1c = ci[WS(rs, 6)]; T4B = T19 - T1a; T1b = T19 + T1a; Tee = T4H - T4I; T4J = T4H + T4I; T1d = cr[WS(rs, 25)]; T4C = ci[WS(rs, 54)]; T4D = cr[WS(rs, 41)]; } { E T1k, T4S, T4V, T1l, T4O, T4P; { E T1h, T1i, T4T, T4U; T1h = cr[WS(rs, 5)]; { E T4G, T1e, Tef, T4E; T4G = T1c - T1d; T1e = T1c + T1d; Tef = T4C - T4D; T4E = T4C + T4D; T5h = T4G - T4J; T4K = T4G + T4J; TdZ = T1b - T1e; T1f = T1b + T1e; ThJ = Tef + Tee; Teg = Tee - Tef; T5g = T4B - T4E; T4F = T4B + T4E; T1i = ci[WS(rs, 26)]; } T4T = ci[WS(rs, 42)]; T4U = cr[WS(rs, 53)]; T1k = cr[WS(rs, 21)]; T4S = T1h - T1i; T1j = T1h + T1i; Te8 = T4T - T4U; T4V = T4T + T4U; T1l = ci[WS(rs, 10)]; T4O = ci[WS(rs, 58)]; T4P = cr[WS(rs, 37)]; } T98 = T4S + T4V; T4W = T4S - T4V; T4N = T1k - T1l; T1m = T1k + T1l; Te7 = T4O - T4P; T4Q = T4O + T4P; } } } T1n = T1j + T1m; Te6 = T1j - T1m; { E Te9, ThL, T4R, T99; Te9 = Te7 - Te8; ThL = Te7 + Te8; T4R = T4N + T4Q; T99 = T4Q - T4N; { E Tjr, ThK, Tjs, ThN; { E T1g, T1v, Tjp, Tjo; Tjr = T18 - T1f; T1g = T18 + T1f; T1v = T1n + T1u; Tjp = T1n - T1u; ThK = ThI + ThJ; Tjo = ThI - ThJ; ThH = T1g - T1v; T1w = T1g + T1v; Tke = Tjp + Tjo; Tjq = Tjo - Tjp; Tjs = ThM - ThL; ThN = ThL + ThM; } { E Tg6, Te0, Tg9, Teh, Tej, Tei, Tga, Teb, Te5, Tea; Tg6 = TdZ + TdY; Te0 = TdY - TdZ; Tkf = Tjr + Tjs; Tjt = Tjr - Tjs; TiK = ThK + ThN; ThO = ThK - ThN; Tg9 = Ted + Teg; Teh = Ted - Teg; Tej = Te4 - Te1; Te5 = Te1 + Te4; Tea = Te6 - Te9; Tei = Te6 + Te9; Tga = Tea + Te5; Teb = Te5 - Tea; { E T9h, T4M, T78, T96, T5k, T5l, T75, T5j, T76, T59; { E T5i, Tg7, Tek, T4L, T4X, T58; T9h = T4F + T4K; T4L = T4F - T4K; Tgb = FNMS(KP707106781, Tga, Tg9); TgT = FMA(KP707106781, Tga, Tg9); Tfc = FMA(KP707106781, Teb, Te0); Tec = FNMS(KP707106781, Teb, Te0); Tg7 = Tei + Tej; Tek = Tei - Tej; T4M = FMA(KP707106781, T4L, T4A); T78 = FNMS(KP707106781, T4L, T4A); Tg8 = FNMS(KP707106781, Tg7, Tg6); TgU = FMA(KP707106781, Tg7, Tg6); Tfd = FMA(KP707106781, Tek, Teh); Tel = FNMS(KP707106781, Tek, Teh); T5i = T5g + T5h; T96 = T5h - T5g; T5k = FNMS(KP414213562, T4R, T4W); T4X = FMA(KP414213562, T4W, T4R); T58 = FNMS(KP414213562, T57, T52); T5l = FMA(KP414213562, T52, T57); T75 = FNMS(KP707106781, T5i, T5f); T5j = FMA(KP707106781, T5i, T5f); T76 = T4X - T58; T59 = T4X + T58; } { E T79, T5m, T9a, T9d; T77 = FNMS(KP923879532, T76, T75); T83 = FMA(KP923879532, T76, T75); T6i = FMA(KP923879532, T59, T4M); T5a = FNMS(KP923879532, T59, T4M); T79 = T5l - T5k; T5m = T5k + T5l; T97 = FNMS(KP707106781, T96, T95); Tbk = FMA(KP707106781, T96, T95); T7a = FNMS(KP923879532, T79, T78); T82 = FMA(KP923879532, T79, T78); T6j = FMA(KP923879532, T5m, T5j); T5n = FNMS(KP923879532, T5m, T5j); T9j = FNMS(KP414213562, T98, T99); T9a = FMA(KP414213562, T99, T98); T9d = FMA(KP414213562, T9c, T9b); T9k = FNMS(KP414213562, T9b, T9c); Tbh = FMA(KP707106781, T9h, T9g); T9i = FNMS(KP707106781, T9h, T9g); Tbi = T9a + T9d; T9e = T9a - T9d; } } } } } } { E T9z, T4m, T1D, TdM, ThR, Tdx, T3H, T9o, T9r, T4e, T1Z, TdA, TdD, ThV, T49; E T9s, T4o, T3R, Tdy, T1K, ThS, TdP, T4n, T3M, T1O, T3V, TdH, T3U, T1R, T3W; E T9u, T43; { E T1V, T46, TdC, T45, T1Y, T47, T48, TdB; { E Tdw, T3D, T3G, Tdv, T4a, T4d; { E T4i, T1z, T3E, T4l, T1C, T3F; { E T4j, T4k, T1A, T1B; { E T1x, Tbl, T9l, T1y; T1x = ci[0]; Tbj = FNMS(KP923879532, Tbi, Tbh); Tcc = FMA(KP923879532, Tbi, Tbh); Tas = FMA(KP923879532, T9e, T97); T9f = FNMS(KP923879532, T9e, T97); Tbl = T9j - T9k; T9l = T9j + T9k; T1y = cr[WS(rs, 31)]; T4j = ci[WS(rs, 48)]; Tbm = FNMS(KP923879532, Tbl, Tbk); Tcb = FMA(KP923879532, Tbl, Tbk); Tar = FNMS(KP923879532, T9l, T9i); T9m = FMA(KP923879532, T9l, T9i); T4i = T1x - T1y; T1z = T1x + T1y; T4k = cr[WS(rs, 47)]; } T1A = cr[WS(rs, 15)]; T1B = ci[WS(rs, 16)]; T3E = ci[WS(rs, 32)]; Tdw = T4j - T4k; T4l = T4j + T4k; T3D = T1A - T1B; T1C = T1A + T1B; T3F = cr[WS(rs, 63)]; } T9z = T4i + T4l; T4m = T4i - T4l; T1D = T1z + T1C; TdM = T1z - T1C; T3G = T3E + T3F; Tdv = T3E - T3F; } { E T4b, T4c, T1T, T1U, T1W, T1X; T1T = ci[WS(rs, 4)]; T1U = cr[WS(rs, 27)]; ThR = Tdv + Tdw; Tdx = Tdv - Tdw; T3H = T3D - T3G; T9o = T3D + T3G; T4a = T1T - T1U; T1V = T1T + T1U; T4b = ci[WS(rs, 52)]; T4c = cr[WS(rs, 43)]; T1W = cr[WS(rs, 11)]; T1X = ci[WS(rs, 20)]; T46 = ci[WS(rs, 36)]; TdC = T4b - T4c; T4d = T4b + T4c; T45 = T1W - T1X; T1Y = T1W + T1X; T47 = cr[WS(rs, 59)]; } T9r = T4a + T4d; T4e = T4a - T4d; } T1Z = T1V + T1Y; TdA = T1V - T1Y; T48 = T46 + T47; TdB = T46 - T47; { E T3I, T1G, T3J, TdN, T3Q, T3N, T1J, T3K, T3Z, T42; { E T3O, T3P, T1E, T1F, T1H, T1I; T1E = cr[WS(rs, 7)]; T1F = ci[WS(rs, 24)]; TdD = TdB - TdC; ThV = TdB + TdC; T49 = T45 - T48; T9s = T45 + T48; T3I = T1E - T1F; T1G = T1E + T1F; T3O = ci[WS(rs, 40)]; T3P = cr[WS(rs, 55)]; T1H = ci[WS(rs, 8)]; T1I = cr[WS(rs, 23)]; T3J = ci[WS(rs, 56)]; TdN = T3O - T3P; T3Q = T3O + T3P; T3N = T1H - T1I; T1J = T1H + T1I; T3K = cr[WS(rs, 39)]; } { E T40, T41, T1P, T1Q; { E T1M, TdO, T3L, T1N; T1M = cr[WS(rs, 3)]; T4o = T3N - T3Q; T3R = T3N + T3Q; Tdy = T1G - T1J; T1K = T1G + T1J; TdO = T3J - T3K; T3L = T3J + T3K; T1N = ci[WS(rs, 28)]; T40 = ci[WS(rs, 44)]; ThS = TdO + TdN; TdP = TdN - TdO; T4n = T3I - T3L; T3M = T3I + T3L; T3Z = T1M - T1N; T1O = T1M + T1N; T41 = cr[WS(rs, 51)]; } T1P = cr[WS(rs, 19)]; T1Q = ci[WS(rs, 12)]; T3V = ci[WS(rs, 60)]; TdH = T40 - T41; T42 = T40 + T41; T3U = T1P - T1Q; T1R = T1P + T1Q; T3W = cr[WS(rs, 35)]; } T9u = T3Z + T42; T43 = T3Z - T42; } } { E T1S, TdF, T3X, TdG; T1S = T1O + T1R; TdF = T1O - T1R; T3X = T3V + T3W; TdG = T3V - T3W; { E TdI, T3Y, T9v, ThT, ThW; { E Tjk, Tji, ThU, Tjh, T1L, T20, Tjl; Tjk = T1D - T1K; T1L = T1D + T1K; T20 = T1S + T1Z; Tji = T1S - T1Z; TdI = TdG - TdH; ThU = TdG + TdH; T3Y = T3U + T3X; T9v = T3U - T3X; ThQ = T1L - T20; T21 = T1L + T20; ThT = ThR + ThS; Tjh = ThR - ThS; Tjl = ThV - ThU; ThW = ThU + ThV; Tkb = Tji + Tjh; Tjj = Tjh - Tji; Tkc = Tjk + Tjl; Tjm = Tjk - Tjl; } { E TfZ, Tdz, Tg2, TdQ, TdS, TdR, Tg3, TdK, TdE, TdJ; TfZ = Tdy + Tdx; Tdz = Tdx - Tdy; Tg2 = TdM + TdP; TdQ = TdM - TdP; TdS = TdD - TdA; TdE = TdA + TdD; TiL = ThT + ThW; ThX = ThT - ThW; TdJ = TdF - TdI; TdR = TdF + TdI; Tg3 = TdJ + TdE; TdK = TdE - TdJ; { E T9A, T3T, T7f, T9p, T4r, T4s, T7c, T4q, T7d, T4g; { E T4p, Tg0, TdT, T3S, T44, T4f; T9A = T3M + T3R; T3S = T3M - T3R; Tg4 = FNMS(KP707106781, Tg3, Tg2); TgW = FMA(KP707106781, Tg3, Tg2); Tf9 = FMA(KP707106781, TdK, Tdz); TdL = FNMS(KP707106781, TdK, Tdz); Tg0 = TdR + TdS; TdT = TdR - TdS; T3T = FMA(KP707106781, T3S, T3H); T7f = FNMS(KP707106781, T3S, T3H); Tg1 = FNMS(KP707106781, Tg0, TfZ); TgX = FMA(KP707106781, Tg0, TfZ); Tfa = FMA(KP707106781, TdT, TdQ); TdU = FNMS(KP707106781, TdT, TdQ); T4p = T4n + T4o; T9p = T4n - T4o; T4r = FNMS(KP414213562, T3Y, T43); T44 = FMA(KP414213562, T43, T3Y); T4f = FNMS(KP414213562, T4e, T49); T4s = FMA(KP414213562, T49, T4e); T7c = FNMS(KP707106781, T4p, T4m); T4q = FMA(KP707106781, T4p, T4m); T7d = T44 - T4f; T4g = T44 + T4f; } { E T7g, T4t, T9t, T9w; T7e = FNMS(KP923879532, T7d, T7c); T80 = FMA(KP923879532, T7d, T7c); T6f = FMA(KP923879532, T4g, T3T); T4h = FNMS(KP923879532, T4g, T3T); T7g = T4s - T4r; T4t = T4r + T4s; T9q = FNMS(KP707106781, T9p, T9o); Tbr = FMA(KP707106781, T9p, T9o); T7h = FNMS(KP923879532, T7g, T7f); T7Z = FMA(KP923879532, T7g, T7f); T6g = FMA(KP923879532, T4t, T4q); T4u = FNMS(KP923879532, T4t, T4q); T9D = FNMS(KP414213562, T9r, T9s); T9t = FMA(KP414213562, T9s, T9r); T9w = FNMS(KP414213562, T9v, T9u); T9C = FMA(KP414213562, T9u, T9v); Tbo = FMA(KP707106781, T9A, T9z); T9B = FNMS(KP707106781, T9A, T9z); Tbp = T9w + T9t; T9x = T9t - T9w; } } } } } } } } { E Tbq, Tcf, Tav, T9y, Tbt, Tce, Tau, T9F, T6p, T6d, T6c, T6q, Thf, The, Thd; { E Tk9, Tkm, TjP, TjO, TjN; { E Tj0, TiS, TiU, Tj3, Tj1, Tj4, TiY, Tj2; { E TiQ, TiW, TiV, TiR, TiD, TiG, TiN, TiF, TiO; { E T11, T22, TiJ, TiE, TiM, Tbs, T9E; TiQ = Tv - T10; T11 = Tv + T10; Tbq = FNMS(KP923879532, Tbp, Tbo); Tcf = FMA(KP923879532, Tbp, Tbo); Tav = FMA(KP923879532, T9x, T9q); T9y = FNMS(KP923879532, T9x, T9q); Tbs = T9C + T9D; T9E = T9C - T9D; T22 = T1w + T21; TiW = T1w - T21; TiV = TiH - TiI; TiJ = TiH + TiI; Tbt = FNMS(KP923879532, Tbs, Tbr); Tce = FMA(KP923879532, Tbs, Tbr); Tau = FMA(KP923879532, T9E, T9B); T9F = FNMS(KP923879532, T9E, T9B); TiE = T11 - T22; TiR = TiL - TiK; TiM = TiK + TiL; cr[0] = T11 + T22; TiD = W[62]; TiG = W[63]; ci[0] = TiJ + TiM; TiN = TiJ - TiM; TiF = TiD * TiE; TiO = TiG * TiE; } cr[WS(rs, 32)] = FNMS(TiG, TiN, TiF); ci[WS(rs, 32)] = FMA(TiD, TiN, TiO); Tj0 = TiQ + TiR; TiS = TiQ - TiR; { E TiP, TiX, TiT, TiZ; TiP = W[94]; TiU = W[95]; TiZ = W[30]; Tj3 = TiW + TiV; TiX = TiV - TiW; TiT = TiP * TiS; Tj1 = TiZ * Tj0; Tj4 = TiZ * Tj3; TiY = TiP * TiX; cr[WS(rs, 48)] = FNMS(TiU, TiX, TiT); Tj2 = W[31]; } } { E Tii, Til, Tik, Tih, Tim; { E Tib, Tit, Tio, ThG, ThP, ThY, Tie, Tip, Tic, Tid; Tib = Ti3 + Tia; Tit = Tia - Ti3; ci[WS(rs, 48)] = FMA(TiU, TiS, TiY); Tio = Thy - ThF; ThG = Thy + ThF; ci[WS(rs, 16)] = FMA(Tj2, Tj0, Tj4); cr[WS(rs, 16)] = FNMS(Tj2, Tj3, Tj1); ThP = ThH - ThO; Tic = ThH + ThO; Tid = ThX - ThQ; ThY = ThQ + ThX; Tie = Tic + Tid; Tip = Tid - Tic; { E Tiy, TiB, Ti0, Tiz, TiC, TiA; { E Tin, Tis, Tiq, ThZ, Tiu, Tir, Tiw, Tix, Tiv; Tin = W[110]; Tis = W[111]; Tiy = FMA(KP707106781, Tip, Tio); Tiq = FNMS(KP707106781, Tip, Tio); ThZ = ThP + ThY; Tiu = ThP - ThY; Tir = Tin * Tiq; Tix = W[46]; TiB = FMA(KP707106781, Tiu, Tit); Tiv = FNMS(KP707106781, Tiu, Tit); Ti0 = FNMS(KP707106781, ThZ, ThG); Tii = FMA(KP707106781, ThZ, ThG); cr[WS(rs, 56)] = FNMS(Tis, Tiv, Tir); Tiw = Tin * Tiv; Tiz = Tix * Tiy; TiC = Tix * TiB; TiA = W[47]; ci[WS(rs, 56)] = FMA(Tis, Tiq, Tiw); } { E Tif, Ti2, Thx, Tig, Ti1; Til = FMA(KP707106781, Tie, Tib); Tif = FNMS(KP707106781, Tie, Tib); Ti2 = W[79]; ci[WS(rs, 24)] = FMA(TiA, Tiy, TiC); cr[WS(rs, 24)] = FNMS(TiA, TiB, Tiz); Thx = W[78]; Tig = Ti2 * Ti0; Tik = W[15]; Ti1 = Thx * Ti0; ci[WS(rs, 40)] = FMA(Thx, Tif, Tig); Tih = W[14]; Tim = Tik * Tii; cr[WS(rs, 40)] = FNMS(Ti2, Tif, Ti1); } } } { E TjF, TjI, TjU, Tk2, TjZ, Tk5, Tjw, TjM; { E TjX, TjG, Tju, Tjg, TjS, Tjn, TjH, Tjf, TjE, Tij, TjT, Tjv, TjY; TjE = TjC - TjD; Tk9 = TjC + TjD; Tij = Tih * Tii; ci[WS(rs, 8)] = FMA(Tih, Til, Tim); Tkm = Tje + Tjb; Tjf = Tjb - Tje; TjX = FNMS(KP707106781, TjE, TjB); TjF = FMA(KP707106781, TjE, TjB); cr[WS(rs, 8)] = FNMS(Tik, Til, Tij); TjG = FMA(KP414213562, Tjq, Tjt); Tju = FNMS(KP414213562, Tjt, Tjq); Tjg = FMA(KP707106781, Tjf, Tj8); TjS = FNMS(KP707106781, Tjf, Tj8); Tjn = FMA(KP414213562, Tjm, Tjj); TjH = FNMS(KP414213562, Tjj, Tjm); TjI = TjG - TjH; TjT = TjG + TjH; Tjv = Tjn - Tju; TjY = Tju + Tjn; TjU = FNMS(KP923879532, TjT, TjS); Tk2 = FMA(KP923879532, TjT, TjS); TjZ = FNMS(KP923879532, TjY, TjX); Tk5 = FMA(KP923879532, TjY, TjX); Tjw = FNMS(KP923879532, Tjv, Tjg); TjM = FMA(KP923879532, Tjv, Tjg); } { E Tk4, Tk3, TjR, TjW, TjJ, Tjy, Tj5; TjR = W[54]; TjW = W[55]; { E Tk1, Tk0, TjV, Tk6; Tk1 = W[118]; Tk4 = W[119]; Tk0 = TjR * TjZ; TjV = TjR * TjU; Tk6 = Tk1 * Tk5; Tk3 = Tk1 * Tk2; ci[WS(rs, 28)] = FMA(TjW, TjU, Tk0); cr[WS(rs, 28)] = FNMS(TjW, TjZ, TjV); ci[WS(rs, 60)] = FMA(Tk4, Tk2, Tk6); } cr[WS(rs, 60)] = FNMS(Tk4, Tk5, Tk3); TjP = FMA(KP923879532, TjI, TjF); TjJ = FNMS(KP923879532, TjI, TjF); Tjy = W[87]; Tj5 = W[86]; { E TjL, TjQ, TjK, Tjx; TjO = W[23]; TjK = Tjy * Tjw; Tjx = Tj5 * Tjw; TjL = W[22]; TjQ = TjO * TjM; ci[WS(rs, 44)] = FMA(Tj5, TjJ, TjK); cr[WS(rs, 44)] = FNMS(Tjy, TjJ, Tjx); TjN = TjL * TjM; ci[WS(rs, 12)] = FMA(TjL, TjP, TjQ); } } } } } { E T5T, T5S, T5R, Tkx, Tkw, Tkv; { E Tkn, Tkq, TkC, TkK, TkH, TkN, Tki, Tku; { E Tkg, Tko, TkF, Tka, TkA, Tkd, Tkp, TkB, Tkh, TkG; cr[WS(rs, 12)] = FNMS(TjO, TjP, TjN); Tkg = FMA(KP414213562, Tkf, Tke); Tko = FNMS(KP414213562, Tke, Tkf); TkF = FMA(KP707106781, Tkm, Tkl); Tkn = FNMS(KP707106781, Tkm, Tkl); Tka = FNMS(KP707106781, Tk9, Tk8); TkA = FMA(KP707106781, Tk9, Tk8); Tkd = FNMS(KP414213562, Tkc, Tkb); Tkp = FMA(KP414213562, Tkb, Tkc); Tkq = Tko - Tkp; TkB = Tko + Tkp; Tkh = Tkd - Tkg; TkG = Tkg + Tkd; TkC = FNMS(KP923879532, TkB, TkA); TkK = FMA(KP923879532, TkB, TkA); TkH = FNMS(KP923879532, TkG, TkF); TkN = FMA(KP923879532, TkG, TkF); Tki = FNMS(KP923879532, Tkh, Tka); Tku = FMA(KP923879532, Tkh, Tka); } { E TkM, TkL, Tkz, TkE, Tkr, Tkk, Tk7; Tkz = W[70]; TkE = W[71]; { E TkJ, TkI, TkD, TkO; TkJ = W[6]; TkM = W[7]; TkI = Tkz * TkH; TkD = Tkz * TkC; TkO = TkJ * TkN; TkL = TkJ * TkK; ci[WS(rs, 36)] = FMA(TkE, TkC, TkI); cr[WS(rs, 36)] = FNMS(TkE, TkH, TkD); ci[WS(rs, 4)] = FMA(TkM, TkK, TkO); } cr[WS(rs, 4)] = FNMS(TkM, TkN, TkL); Tkx = FMA(KP923879532, Tkq, Tkn); Tkr = FNMS(KP923879532, Tkq, Tkn); Tkk = W[103]; Tk7 = W[102]; { E Tkt, Tky, Tks, Tkj; Tkw = W[39]; Tks = Tkk * Tki; Tkj = Tk7 * Tki; Tkt = W[38]; Tky = Tkw * Tku; ci[WS(rs, 52)] = FMA(Tk7, Tkr, Tks); cr[WS(rs, 52)] = FNMS(Tkk, Tkr, Tkj); Tkv = Tkt * Tku; ci[WS(rs, 20)] = FMA(Tkt, Tkx, Tky); } } } { E T5J, T5M, T66, T5Y, T69, T63, T5Q, T5q; { E T5o, T4v, T61, T5X, T3C, T5W, T62, T5p; { E T5K, T5L, T5F, T5I, T2I, T3B; T5F = FNMS(KP923879532, T5E, T5B); T6p = FMA(KP923879532, T5E, T5B); T6d = T5G + T5H; T5I = T5G - T5H; cr[WS(rs, 20)] = FNMS(Tkw, Tkx, Tkv); T5o = FNMS(KP820678790, T5n, T5a); T5K = FMA(KP820678790, T5a, T5n); T5L = FNMS(KP820678790, T4h, T4u); T4v = FMA(KP820678790, T4u, T4h); T5J = FMA(KP980785280, T5I, T5F); T61 = FNMS(KP980785280, T5I, T5F); T2I = FNMS(KP923879532, T2H, T2k); T6c = FMA(KP923879532, T2H, T2k); T6q = T3A + T39; T3B = T39 - T3A; T5X = T5K + T5L; T5M = T5K - T5L; T3C = FMA(KP980785280, T3B, T2I); T5W = FNMS(KP980785280, T3B, T2I); } T62 = T5o + T4v; T5p = T4v - T5o; T66 = FMA(KP773010453, T5X, T5W); T5Y = FNMS(KP773010453, T5X, T5W); T69 = FMA(KP773010453, T62, T61); T63 = FNMS(KP773010453, T62, T61); T5Q = FMA(KP773010453, T5p, T3C); T5q = FNMS(KP773010453, T5p, T3C); } { E T68, T67, T5V, T60, T5N, T5s, T23; T5V = W[48]; T60 = W[49]; { E T65, T64, T5Z, T6a; T65 = W[112]; T68 = W[113]; T64 = T5V * T63; T5Z = T5V * T5Y; T6a = T65 * T69; T67 = T65 * T66; ci[WS(rs, 25)] = FMA(T60, T5Y, T64); cr[WS(rs, 25)] = FNMS(T60, T63, T5Z); ci[WS(rs, 57)] = FMA(T68, T66, T6a); } cr[WS(rs, 57)] = FNMS(T68, T69, T67); T5T = FMA(KP773010453, T5M, T5J); T5N = FNMS(KP773010453, T5M, T5J); T5s = W[81]; T23 = W[80]; { E T5P, T5U, T5O, T5r; T5S = W[17]; T5O = T5s * T5q; T5r = T23 * T5q; T5P = W[16]; T5U = T5S * T5Q; ci[WS(rs, 41)] = FMA(T23, T5N, T5O); cr[WS(rs, 41)] = FNMS(T5s, T5N, T5r); T5R = T5P * T5Q; ci[WS(rs, 9)] = FMA(T5P, T5T, T5U); } } } { E Th3, TgR, TgQ, Th4, TgN, TgM, TgL; { E TgG, TgF, Tge, Tgu, TgK, TgC, Tgx, Tgr; { E Tgp, Tgo, Tgd, Tgn, TfY, TgA, TgB, Tgq; { E Tgj, Tgm, Tg5, Tgc, TfQ, TfX; Tg5 = FMA(KP668178637, Tg4, Tg1); Tgp = FNMS(KP668178637, Tg1, Tg4); Tgo = FMA(KP668178637, Tg8, Tgb); Tgc = FNMS(KP668178637, Tgb, Tg8); cr[WS(rs, 9)] = FNMS(T5S, T5T, T5R); Th3 = FMA(KP707106781, Tgi, Tgh); Tgj = FNMS(KP707106781, Tgi, Tgh); Tgm = Tgk - Tgl; TgR = Tgk + Tgl; TgG = Tgc + Tg5; Tgd = Tg5 - Tgc; TfQ = FNMS(KP707106781, TfP, TfO); TgQ = FMA(KP707106781, TfP, TfO); Th4 = TfW + TfT; TfX = TfT - TfW; Tgn = FMA(KP923879532, Tgm, Tgj); TgF = FNMS(KP923879532, Tgm, Tgj); TfY = FMA(KP923879532, TfX, TfQ); TgA = FNMS(KP923879532, TfX, TfQ); } TgB = Tgo + Tgp; Tgq = Tgo - Tgp; Tge = FNMS(KP831469612, Tgd, TfY); Tgu = FMA(KP831469612, Tgd, TfY); TgK = FMA(KP831469612, TgB, TgA); TgC = FNMS(KP831469612, TgB, TgA); Tgx = FMA(KP831469612, Tgq, Tgn); Tgr = FNMS(KP831469612, Tgq, Tgn); } { E Tgw, Tgv, TfN, Tgg, TgH, TgE, Tgz; TfN = W[82]; Tgg = W[83]; { E Tgt, Tgs, Tgf, Tgy; Tgt = W[18]; Tgw = W[19]; Tgs = TfN * Tgr; Tgf = TfN * Tge; Tgy = Tgt * Tgx; Tgv = Tgt * Tgu; ci[WS(rs, 42)] = FMA(Tgg, Tge, Tgs); cr[WS(rs, 42)] = FNMS(Tgg, Tgr, Tgf); ci[WS(rs, 10)] = FMA(Tgw, Tgu, Tgy); } cr[WS(rs, 10)] = FNMS(Tgw, Tgx, Tgv); TgN = FMA(KP831469612, TgG, TgF); TgH = FNMS(KP831469612, TgG, TgF); TgE = W[51]; Tgz = W[50]; { E TgJ, TgO, TgI, TgD; TgM = W[115]; TgI = TgE * TgC; TgD = Tgz * TgC; TgJ = W[114]; TgO = TgM * TgK; ci[WS(rs, 26)] = FMA(Tgz, TgH, TgI); cr[WS(rs, 26)] = FNMS(TgE, TgH, TgD); TgL = TgJ * TgK; ci[WS(rs, 58)] = FMA(TgJ, TgN, TgO); } } } { E Th5, Th8, Ths, Thk, Thv, Thp, Thc, Th0; { E TgV, TgY, Thn, Thj, TgS, Thi, Th6, Th7, Tho, TgZ; cr[WS(rs, 58)] = FNMS(TgM, TgN, TgL); TgV = FNMS(KP198912367, TgU, TgT); Th6 = FMA(KP198912367, TgT, TgU); Th7 = FNMS(KP198912367, TgW, TgX); TgY = FMA(KP198912367, TgX, TgW); Th5 = FMA(KP923879532, Th4, Th3); Thn = FNMS(KP923879532, Th4, Th3); Thj = Th7 - Th6; Th8 = Th6 + Th7; TgS = FMA(KP923879532, TgR, TgQ); Thi = FNMS(KP923879532, TgR, TgQ); Tho = TgV - TgY; TgZ = TgV + TgY; Ths = FMA(KP980785280, Thj, Thi); Thk = FNMS(KP980785280, Thj, Thi); Thv = FMA(KP980785280, Tho, Thn); Thp = FNMS(KP980785280, Tho, Thn); Thc = FMA(KP980785280, TgZ, TgS); Th0 = FNMS(KP980785280, TgZ, TgS); } { E Thu, Tht, Thh, Thm, Th9, Th2, TgP; Thh = W[98]; Thm = W[99]; { E Thr, Thq, Thl, Thw; Thr = W[34]; Thu = W[35]; Thq = Thh * Thp; Thl = Thh * Thk; Thw = Thr * Thv; Tht = Thr * Ths; ci[WS(rs, 50)] = FMA(Thm, Thk, Thq); cr[WS(rs, 50)] = FNMS(Thm, Thp, Thl); ci[WS(rs, 18)] = FMA(Thu, Ths, Thw); } cr[WS(rs, 18)] = FNMS(Thu, Thv, Tht); Thf = FMA(KP980785280, Th8, Th5); Th9 = FNMS(KP980785280, Th8, Th5); Th2 = W[67]; TgP = W[66]; { E Thb, Thg, Tha, Th1; The = W[3]; Tha = Th2 * Th0; Th1 = TgP * Th0; Thb = W[2]; Thg = The * Thc; ci[WS(rs, 34)] = FMA(TgP, Th9, Tha); cr[WS(rs, 34)] = FNMS(Th2, Th9, Th1); Thd = Thb * Thc; ci[WS(rs, 2)] = FMA(Thb, Thf, Thg); } } } } } } { E Tcl, Tc9, Tc8, Tcm, T9R, T93, T8O, T9U, Tez, Tdt, Td6, TeC, Tfv, Tfu, Tft; E T8B, T8A, T8z; { E TbP, TbO, TbN, T6B, T6A, T6z, TaN, TaM, TaL; { E T6r, T6u, T6O, T6G, T6R, T6L, T6y, T6m; { E T6k, T6h, T6J, T6F, T6e, T6E, T6s, T6t, T6K, T6l; cr[WS(rs, 2)] = FNMS(The, Thf, Thd); T6k = FMA(KP098491403, T6j, T6i); T6s = FNMS(KP098491403, T6i, T6j); T6t = FMA(KP098491403, T6f, T6g); T6h = FNMS(KP098491403, T6g, T6f); T6r = FNMS(KP980785280, T6q, T6p); T6J = FMA(KP980785280, T6q, T6p); T6F = T6s + T6t; T6u = T6s - T6t; T6e = FNMS(KP980785280, T6d, T6c); T6E = FMA(KP980785280, T6d, T6c); T6K = T6k + T6h; T6l = T6h - T6k; T6O = FMA(KP995184726, T6F, T6E); T6G = FNMS(KP995184726, T6F, T6E); T6R = FMA(KP995184726, T6K, T6J); T6L = FNMS(KP995184726, T6K, T6J); T6y = FMA(KP995184726, T6l, T6e); T6m = FNMS(KP995184726, T6l, T6e); } { E T6Q, T6P, T6D, T6I, T6v, T6o, T6b; T6D = W[64]; T6I = W[65]; { E T6N, T6M, T6H, T6S; T6N = W[0]; T6Q = W[1]; T6M = T6D * T6L; T6H = T6D * T6G; T6S = T6N * T6R; T6P = T6N * T6O; ci[WS(rs, 33)] = FMA(T6I, T6G, T6M); cr[WS(rs, 33)] = FNMS(T6I, T6L, T6H); ci[WS(rs, 1)] = FMA(T6Q, T6O, T6S); } cr[WS(rs, 1)] = FNMS(T6Q, T6R, T6P); T6B = FMA(KP995184726, T6u, T6r); T6v = FNMS(KP995184726, T6u, T6r); T6o = W[97]; T6b = W[96]; { E T6x, T6C, T6w, T6n; T6A = W[33]; T6w = T6o * T6m; T6n = T6b * T6m; T6x = W[32]; T6C = T6A * T6y; ci[WS(rs, 49)] = FMA(T6b, T6v, T6w); cr[WS(rs, 49)] = FNMS(T6o, T6v, T6n); T6z = T6x * T6y; ci[WS(rs, 17)] = FMA(T6x, T6B, T6C); } } } { E TbF, TbI, Tc2, TbU, Tc5, TbZ, TbM, Tbw; { E Tbn, Tbu, TbX, TbT, Tbg, TbS, TbY, Tbv; { E TbG, TbH, TbB, TbE, Tb8, Tbf; TbB = FMA(KP923879532, TbA, Tbz); Tcl = FNMS(KP923879532, TbA, Tbz); Tc9 = TbC + TbD; TbE = TbC - TbD; cr[WS(rs, 17)] = FNMS(T6A, T6B, T6z); Tbn = FNMS(KP820678790, Tbm, Tbj); TbG = FMA(KP820678790, Tbj, Tbm); TbH = FMA(KP820678790, Tbq, Tbt); Tbu = FNMS(KP820678790, Tbt, Tbq); TbF = FMA(KP980785280, TbE, TbB); TbX = FNMS(KP980785280, TbE, TbB); Tb8 = FNMS(KP923879532, Tb7, Tb6); Tc8 = FMA(KP923879532, Tb7, Tb6); Tcm = Tbe - Tbb; Tbf = Tbb + Tbe; TbT = TbG + TbH; TbI = TbG - TbH; Tbg = FNMS(KP980785280, Tbf, Tb8); TbS = FMA(KP980785280, Tbf, Tb8); } TbY = Tbn - Tbu; Tbv = Tbn + Tbu; Tc2 = FMA(KP773010453, TbT, TbS); TbU = FNMS(KP773010453, TbT, TbS); Tc5 = FNMS(KP773010453, TbY, TbX); TbZ = FMA(KP773010453, TbY, TbX); TbM = FMA(KP773010453, Tbv, Tbg); Tbw = FNMS(KP773010453, Tbv, Tbg); } { E Tc4, Tc3, TbR, TbW, TbJ, Tby, Tb5; TbR = W[44]; TbW = W[45]; { E Tc1, Tc0, TbV, Tc6; Tc1 = W[108]; Tc4 = W[109]; Tc0 = TbR * TbZ; TbV = TbR * TbU; Tc6 = Tc1 * Tc5; Tc3 = Tc1 * Tc2; ci[WS(rs, 23)] = FMA(TbW, TbU, Tc0); cr[WS(rs, 23)] = FNMS(TbW, TbZ, TbV); ci[WS(rs, 55)] = FMA(Tc4, Tc2, Tc6); } cr[WS(rs, 55)] = FNMS(Tc4, Tc5, Tc3); TbP = FMA(KP773010453, TbI, TbF); TbJ = FNMS(KP773010453, TbI, TbF); Tby = W[77]; Tb5 = W[76]; { E TbL, TbQ, TbK, Tbx; TbO = W[13]; TbK = Tby * Tbw; Tbx = Tb5 * Tbw; TbL = W[12]; TbQ = TbO * TbM; ci[WS(rs, 39)] = FMA(Tb5, TbJ, TbK); cr[WS(rs, 39)] = FNMS(Tby, TbJ, Tbx); TbN = TbL * TbM; ci[WS(rs, 7)] = FMA(TbL, TbP, TbQ); } } } { E TaD, TaG, Tb0, TaS, Tb3, TaX, TaK, Tay; { E Tat, Taw, TaV, TaR, Taq, TaQ, TaW, Tax; { E TaE, TaF, TaB, TaC, Tao, Tap; TaB = FMA(KP923879532, T9Q, T9N); T9R = FNMS(KP923879532, T9Q, T9N); T93 = T8V + T92; TaC = T8V - T92; cr[WS(rs, 7)] = FNMS(TbO, TbP, TbN); Tat = FNMS(KP303346683, Tas, Tar); TaE = FMA(KP303346683, Tar, Tas); TaF = FMA(KP303346683, Tau, Tav); Taw = FNMS(KP303346683, Tav, Tau); TaD = FMA(KP831469612, TaC, TaB); TaV = FNMS(KP831469612, TaC, TaB); Tao = FNMS(KP923879532, T8N, T8G); T8O = FMA(KP923879532, T8N, T8G); T9U = T9S - T9T; Tap = T9S + T9T; TaR = TaE + TaF; TaG = TaE - TaF; Taq = FMA(KP831469612, Tap, Tao); TaQ = FNMS(KP831469612, Tap, Tao); } TaW = Tat - Taw; Tax = Tat + Taw; Tb0 = FMA(KP956940335, TaR, TaQ); TaS = FNMS(KP956940335, TaR, TaQ); Tb3 = FNMS(KP956940335, TaW, TaV); TaX = FMA(KP956940335, TaW, TaV); TaK = FMA(KP956940335, Tax, Taq); Tay = FNMS(KP956940335, Tax, Taq); } { E Tb2, Tb1, TaP, TaU, TaH, TaA, Tan; TaP = W[36]; TaU = W[37]; { E TaZ, TaY, TaT, Tb4; TaZ = W[100]; Tb2 = W[101]; TaY = TaP * TaX; TaT = TaP * TaS; Tb4 = TaZ * Tb3; Tb1 = TaZ * Tb0; ci[WS(rs, 19)] = FMA(TaU, TaS, TaY); cr[WS(rs, 19)] = FNMS(TaU, TaX, TaT); ci[WS(rs, 51)] = FMA(Tb2, Tb0, Tb4); } cr[WS(rs, 51)] = FNMS(Tb2, Tb3, Tb1); TaN = FMA(KP956940335, TaG, TaD); TaH = FNMS(KP956940335, TaG, TaD); TaA = W[69]; Tan = W[68]; { E TaJ, TaO, TaI, Taz; TaM = W[5]; TaI = TaA * Tay; Taz = Tan * Tay; TaJ = W[4]; TaO = TaM * TaK; ci[WS(rs, 35)] = FMA(Tan, TaH, TaI); cr[WS(rs, 35)] = FNMS(TaA, TaH, Taz); TaL = TaJ * TaK; ci[WS(rs, 3)] = FMA(TaJ, TaN, TaO); } } } { E Tfl, Tfo, TfI, TfA, TfL, TfF, Tfs, Tfg; { E Tfe, Tfb, TfD, Tfz, Tf8, Tfy, TfE, Tff; { E Tfm, Tfn, Tfj, Tfk, Tf6, Tf7; Tfj = FNMS(KP707106781, Tey, Tev); Tez = FMA(KP707106781, Tey, Tev); Tdt = Tdh - Tds; Tfk = Tds + Tdh; cr[WS(rs, 3)] = FNMS(TaM, TaN, TaL); Tfe = FNMS(KP198912367, Tfd, Tfc); Tfm = FMA(KP198912367, Tfc, Tfd); Tfn = FNMS(KP198912367, Tf9, Tfa); Tfb = FMA(KP198912367, Tfa, Tf9); Tfl = FNMS(KP923879532, Tfk, Tfj); TfD = FMA(KP923879532, Tfk, Tfj); Tf6 = FNMS(KP707106781, Td5, TcU); Td6 = FMA(KP707106781, Td5, TcU); TeC = TeA - TeB; Tf7 = TeA + TeB; Tfz = Tfm + Tfn; Tfo = Tfm - Tfn; Tf8 = FNMS(KP923879532, Tf7, Tf6); Tfy = FMA(KP923879532, Tf7, Tf6); } TfE = Tfe + Tfb; Tff = Tfb - Tfe; TfI = FMA(KP980785280, Tfz, Tfy); TfA = FNMS(KP980785280, Tfz, Tfy); TfL = FMA(KP980785280, TfE, TfD); TfF = FNMS(KP980785280, TfE, TfD); Tfs = FMA(KP980785280, Tff, Tf8); Tfg = FNMS(KP980785280, Tff, Tf8); } { E TfK, TfJ, Tfx, TfC, Tfp, Tfi, Tf5; Tfx = W[58]; TfC = W[59]; { E TfH, TfG, TfB, TfM; TfH = W[122]; TfK = W[123]; TfG = Tfx * TfF; TfB = Tfx * TfA; TfM = TfH * TfL; TfJ = TfH * TfI; ci[WS(rs, 30)] = FMA(TfC, TfA, TfG); cr[WS(rs, 30)] = FNMS(TfC, TfF, TfB); ci[WS(rs, 62)] = FMA(TfK, TfI, TfM); } cr[WS(rs, 62)] = FNMS(TfK, TfL, TfJ); Tfv = FMA(KP980785280, Tfo, Tfl); Tfp = FNMS(KP980785280, Tfo, Tfl); Tfi = W[91]; Tf5 = W[90]; { E Tfr, Tfw, Tfq, Tfh; Tfu = W[27]; Tfq = Tfi * Tfg; Tfh = Tf5 * Tfg; Tfr = W[26]; Tfw = Tfu * Tfs; ci[WS(rs, 46)] = FMA(Tf5, Tfp, Tfq); cr[WS(rs, 46)] = FNMS(Tfi, Tfp, Tfh); Tft = Tfr * Tfs; ci[WS(rs, 14)] = FMA(Tfr, Tfv, Tfw); } } } } { E T89, T7X, T7W, T8a, T7D, T7C, T7B; { E T7t, T7w, T7Q, T7I, T7T, T7N, T7A, T7k; { E T7b, T7i, T7L, T7H, T74, T7G, T7M, T7j; { E T7u, T7v, T7p, T7s, T6W, T73; T7p = FMA(KP923879532, T7o, T7n); T89 = FNMS(KP923879532, T7o, T7n); T7X = T7q + T7r; T7s = T7q - T7r; cr[WS(rs, 14)] = FNMS(Tfu, Tfv, Tft); T7b = FNMS(KP534511135, T7a, T77); T7u = FMA(KP534511135, T77, T7a); T7v = FNMS(KP534511135, T7e, T7h); T7i = FMA(KP534511135, T7h, T7e); T7t = FMA(KP831469612, T7s, T7p); T7L = FNMS(KP831469612, T7s, T7p); T6W = FMA(KP923879532, T6V, T6U); T7W = FNMS(KP923879532, T6V, T6U); T8a = T72 + T6Z; T73 = T6Z - T72; T7H = T7v - T7u; T7w = T7u + T7v; T74 = FMA(KP831469612, T73, T6W); T7G = FNMS(KP831469612, T73, T6W); } T7M = T7b - T7i; T7j = T7b + T7i; T7Q = FMA(KP881921264, T7H, T7G); T7I = FNMS(KP881921264, T7H, T7G); T7T = FMA(KP881921264, T7M, T7L); T7N = FNMS(KP881921264, T7M, T7L); T7A = FMA(KP881921264, T7j, T74); T7k = FNMS(KP881921264, T7j, T74); } { E T7S, T7R, T7F, T7K, T7x, T7m, T6T; T7F = W[104]; T7K = W[105]; { E T7P, T7O, T7J, T7U; T7P = W[40]; T7S = W[41]; T7O = T7F * T7N; T7J = T7F * T7I; T7U = T7P * T7T; T7R = T7P * T7Q; ci[WS(rs, 53)] = FMA(T7K, T7I, T7O); cr[WS(rs, 53)] = FNMS(T7K, T7N, T7J); ci[WS(rs, 21)] = FMA(T7S, T7Q, T7U); } cr[WS(rs, 21)] = FNMS(T7S, T7T, T7R); T7D = FMA(KP881921264, T7w, T7t); T7x = FNMS(KP881921264, T7w, T7t); T7m = W[73]; T6T = W[72]; { E T7z, T7E, T7y, T7l; T7C = W[9]; T7y = T7m * T7k; T7l = T6T * T7k; T7z = W[8]; T7E = T7C * T7A; ci[WS(rs, 37)] = FMA(T6T, T7x, T7y); cr[WS(rs, 37)] = FNMS(T7m, T7x, T7l); T7B = T7z * T7A; ci[WS(rs, 5)] = FMA(T7z, T7D, T7E); } } } { E T8u, T8t, T86, T8i, T8y, T8q, T8l, T8f; { E T8d, T8c, T85, T8b, T7Y, T8o, T81, T84, T8p, T8e; T81 = FMA(KP303346683, T80, T7Z); T8d = FNMS(KP303346683, T7Z, T80); T8c = FMA(KP303346683, T82, T83); T84 = FNMS(KP303346683, T83, T82); cr[WS(rs, 5)] = FNMS(T7C, T7D, T7B); T8u = T84 + T81; T85 = T81 - T84; T8b = FNMS(KP831469612, T8a, T89); T8t = FMA(KP831469612, T8a, T89); T7Y = FNMS(KP831469612, T7X, T7W); T8o = FMA(KP831469612, T7X, T7W); T8p = T8c + T8d; T8e = T8c - T8d; T86 = FNMS(KP956940335, T85, T7Y); T8i = FMA(KP956940335, T85, T7Y); T8y = FMA(KP956940335, T8p, T8o); T8q = FNMS(KP956940335, T8p, T8o); T8l = FMA(KP956940335, T8e, T8b); T8f = FNMS(KP956940335, T8e, T8b); } { E T8k, T8j, T7V, T88, T8v, T8s, T8n; T7V = W[88]; T88 = W[89]; { E T8h, T8g, T87, T8m; T8h = W[24]; T8k = W[25]; T8g = T7V * T8f; T87 = T7V * T86; T8m = T8h * T8l; T8j = T8h * T8i; ci[WS(rs, 45)] = FMA(T88, T86, T8g); cr[WS(rs, 45)] = FNMS(T88, T8f, T87); ci[WS(rs, 13)] = FMA(T8k, T8i, T8m); } cr[WS(rs, 13)] = FNMS(T8k, T8l, T8j); T8B = FMA(KP956940335, T8u, T8t); T8v = FNMS(KP956940335, T8u, T8t); T8s = W[57]; T8n = W[56]; { E T8x, T8C, T8w, T8r; T8A = W[121]; T8w = T8s * T8q; T8r = T8n * T8q; T8x = W[120]; T8C = T8A * T8y; ci[WS(rs, 29)] = FMA(T8n, T8v, T8w); cr[WS(rs, 29)] = FNMS(T8s, T8v, T8r); T8z = T8x * T8y; ci[WS(rs, 61)] = FMA(T8x, T8B, T8C); } } } } { E Ta5, Ta4, Ta3, TeN, TeM, TeL; { E T9V, T9Y, Tai, Taa, Tal, Taf, Ta2, T9I; { E T9n, T9G, Tad, Ta9, T94, Ta8, T9W, T9X, Tae, T9H; cr[WS(rs, 61)] = FNMS(T8A, T8B, T8z); T9n = FNMS(KP534511135, T9m, T9f); T9W = FMA(KP534511135, T9f, T9m); T9X = FMA(KP534511135, T9y, T9F); T9G = FNMS(KP534511135, T9F, T9y); T9V = FMA(KP831469612, T9U, T9R); Tad = FNMS(KP831469612, T9U, T9R); Ta9 = T9W + T9X; T9Y = T9W - T9X; T94 = FNMS(KP831469612, T93, T8O); Ta8 = FMA(KP831469612, T93, T8O); Tae = T9G - T9n; T9H = T9n + T9G; Tai = FMA(KP881921264, Ta9, Ta8); Taa = FNMS(KP881921264, Ta9, Ta8); Tal = FNMS(KP881921264, Tae, Tad); Taf = FMA(KP881921264, Tae, Tad); Ta2 = FNMS(KP881921264, T9H, T94); T9I = FMA(KP881921264, T9H, T94); } { E Tak, Taj, Ta7, Tac, T9Z, T9K, T8D; Ta7 = W[52]; Tac = W[53]; { E Tah, Tag, Tab, Tam; Tah = W[116]; Tak = W[117]; Tag = Ta7 * Taf; Tab = Ta7 * Taa; Tam = Tah * Tal; Taj = Tah * Tai; ci[WS(rs, 27)] = FMA(Tac, Taa, Tag); cr[WS(rs, 27)] = FNMS(Tac, Taf, Tab); ci[WS(rs, 59)] = FMA(Tak, Tai, Tam); } cr[WS(rs, 59)] = FNMS(Tak, Tal, Taj); Ta5 = FMA(KP881921264, T9Y, T9V); T9Z = FNMS(KP881921264, T9Y, T9V); T9K = W[85]; T8D = W[84]; { E Ta1, Ta6, Ta0, T9J; Ta4 = W[21]; Ta0 = T9K * T9I; T9J = T8D * T9I; Ta1 = W[20]; Ta6 = Ta4 * Ta2; ci[WS(rs, 43)] = FMA(T8D, T9Z, Ta0); cr[WS(rs, 43)] = FNMS(T9K, T9Z, T9J); Ta3 = Ta1 * Ta2; ci[WS(rs, 11)] = FMA(Ta1, Ta5, Ta6); } } } { E TeD, TeG, Tf0, TeS, Tf3, TeX, TeK, Teo; { E Tem, TdV, TeV, TeR, Tdu, TeQ, TeE, TeF, TeW, Ten; cr[WS(rs, 11)] = FNMS(Ta4, Ta5, Ta3); Tem = FMA(KP668178637, Tel, Tec); TeE = FNMS(KP668178637, Tec, Tel); TeF = FMA(KP668178637, TdL, TdU); TdV = FNMS(KP668178637, TdU, TdL); TeD = FNMS(KP923879532, TeC, Tez); TeV = FMA(KP923879532, TeC, Tez); TeR = TeE + TeF; TeG = TeE - TeF; Tdu = FNMS(KP923879532, Tdt, Td6); TeQ = FMA(KP923879532, Tdt, Td6); TeW = Tem + TdV; Ten = TdV - Tem; Tf0 = FMA(KP831469612, TeR, TeQ); TeS = FNMS(KP831469612, TeR, TeQ); Tf3 = FMA(KP831469612, TeW, TeV); TeX = FNMS(KP831469612, TeW, TeV); TeK = FMA(KP831469612, Ten, Tdu); Teo = FNMS(KP831469612, Ten, Tdu); } { E Tf2, Tf1, TeP, TeU, TeH, Teq, TcP; TeP = W[74]; TeU = W[75]; { E TeZ, TeY, TeT, Tf4; TeZ = W[10]; Tf2 = W[11]; TeY = TeP * TeX; TeT = TeP * TeS; Tf4 = TeZ * Tf3; Tf1 = TeZ * Tf0; ci[WS(rs, 38)] = FMA(TeU, TeS, TeY); cr[WS(rs, 38)] = FNMS(TeU, TeX, TeT); ci[WS(rs, 6)] = FMA(Tf2, Tf0, Tf4); } cr[WS(rs, 6)] = FNMS(Tf2, Tf3, Tf1); TeN = FMA(KP831469612, TeG, TeD); TeH = FNMS(KP831469612, TeG, TeD); Teq = W[107]; TcP = W[106]; { E TeJ, TeO, TeI, Tep; TeM = W[43]; TeI = Teq * Teo; Tep = TcP * Teo; TeJ = W[42]; TeO = TeM * TeK; ci[WS(rs, 54)] = FMA(TcP, TeH, TeI); cr[WS(rs, 54)] = FNMS(Teq, TeH, Tep); TeL = TeJ * TeK; ci[WS(rs, 22)] = FMA(TeJ, TeN, TeO); } } } { E Tcn, Tcq, TcK, TcC, TcN, TcH, Tcu, Tci; { E Tcd, Tcg, TcF, TcB, Tca, TcA, Tco, Tcp, TcG, Tch; cr[WS(rs, 22)] = FNMS(TeM, TeN, TeL); Tcd = FNMS(KP098491403, Tcc, Tcb); Tco = FMA(KP098491403, Tcb, Tcc); Tcp = FMA(KP098491403, Tce, Tcf); Tcg = FNMS(KP098491403, Tcf, Tce); Tcn = FMA(KP980785280, Tcm, Tcl); TcF = FNMS(KP980785280, Tcm, Tcl); TcB = Tco + Tcp; Tcq = Tco - Tcp; Tca = FNMS(KP980785280, Tc9, Tc8); TcA = FMA(KP980785280, Tc9, Tc8); TcG = Tcg - Tcd; Tch = Tcd + Tcg; TcK = FMA(KP995184726, TcB, TcA); TcC = FNMS(KP995184726, TcB, TcA); TcN = FNMS(KP995184726, TcG, TcF); TcH = FMA(KP995184726, TcG, TcF); Tcu = FNMS(KP995184726, Tch, Tca); Tci = FMA(KP995184726, Tch, Tca); } { E TcM, TcL, Tcz, TcE, Tcr, Tck, Tc7; Tcz = W[60]; TcE = W[61]; { E TcJ, TcI, TcD, TcO; TcJ = W[124]; TcM = W[125]; TcI = Tcz * TcH; TcD = Tcz * TcC; TcO = TcJ * TcN; TcL = TcJ * TcK; ci[WS(rs, 31)] = FMA(TcE, TcC, TcI); cr[WS(rs, 31)] = FNMS(TcE, TcH, TcD); ci[WS(rs, 63)] = FMA(TcM, TcK, TcO); } cr[WS(rs, 63)] = FNMS(TcM, TcN, TcL); Tcx = FMA(KP995184726, Tcq, Tcn); Tcr = FNMS(KP995184726, Tcq, Tcn); Tck = W[93]; Tc7 = W[92]; { E Tct, Tcy, Tcs, Tcj; Tcw = W[29]; Tcs = Tck * Tci; Tcj = Tc7 * Tci; Tct = W[28]; Tcy = Tcw * Tcu; ci[WS(rs, 47)] = FMA(Tc7, Tcr, Tcs); cr[WS(rs, 47)] = FNMS(Tck, Tcr, Tcj); Tcv = Tct * Tcu; ci[WS(rs, 15)] = FMA(Tct, Tcx, Tcy); } } } } } } } cr[WS(rs, 15)] = FNMS(Tcw, Tcx, Tcv); } } } static const tw_instr twinstr[] = { {TW_FULL, 1, 64}, {TW_NEXT, 1, 0} }; static const hc2hc_desc desc = { 64, "hb_64", twinstr, &GENUS, {520, 126, 518, 0} }; void X(codelet_hb_64) (planner *p) { X(khc2hc_register) (p, hb_64, &desc); } #else /* HAVE_FMA */ /* Generated by: ../../../genfft/gen_hc2hc.native -compact -variables 4 -pipeline-latency 4 -sign 1 -n 64 -dif -name hb_64 -include hb.h */ /* * This function contains 1038 FP additions, 500 FP multiplications, * (or, 808 additions, 270 multiplications, 230 fused multiply/add), * 196 stack variables, 15 constants, and 256 memory accesses */ #include "hb.h" static void hb_64(R *cr, R *ci, const R *W, stride rs, INT mb, INT me, INT ms) { DK(KP098017140, +0.098017140329560601994195563888641845861136673); DK(KP995184726, +0.995184726672196886244836953109479921575474869); DK(KP773010453, +0.773010453362736960810906609758469800971041293); DK(KP634393284, +0.634393284163645498215171613225493370675687095); DK(KP471396736, +0.471396736825997648556387625905254377657460319); DK(KP881921264, +0.881921264348355029712756863660388349508442621); DK(KP956940335, +0.956940335732208864935797886980269969482849206); DK(KP290284677, +0.290284677254462367636192375817395274691476278); DK(KP195090322, +0.195090322016128267848284868477022240927691618); DK(KP980785280, +0.980785280403230449126182236134239036973933731); DK(KP555570233, +0.555570233019602224742830813948532874374937191); DK(KP831469612, +0.831469612302545237078788377617905756738560812); DK(KP382683432, +0.382683432365089771728459984030398866761344562); DK(KP923879532, +0.923879532511286756128183189396788286822416626); DK(KP707106781, +0.707106781186547524400844362104849039284835938); { INT m; for (m = mb, W = W + ((mb - 1) * 126); m < me; m = m + 1, cr = cr + ms, ci = ci - ms, W = W + 126, MAKE_VOLATILE_STRIDE(128, rs)) { E Tf, T8C, Tfa, Thk, Tgg, ThM, T2c, T5O, T4K, T6g, Tag, TdE, TcA, Te6, T7P; E T94, TK, T7o, T38, T4P, Tfv, Thn, T5W, T6j, Tb0, TdK, Tfs, Tho, T8K, T97; E Tb7, TdL, TZ, T7l, T2P, T4Q, Tfo, Thq, T5T, T6k, TaH, TdH, Tfl, Thr, T8H; E T98, TaO, TdI, Tu, T95, Tfh, ThN, Tgj, Thl, T2v, T6h, T4N, T5P, Tav, Te7; E TcD, TdF, T7S, T8D, T1L, T20, T7A, T7D, T7G, T7H, T40, T62, Tg1, Thv, Tg8; E Thz, Tg5, Thw, T4t, T5Z, T4j, T60, T4w, T63, TbY, TdS, Tcd, TdQ, TfU, Thy; E T8P, T9z, T8S, T9A, Tcl, TdP, Tco, TdT, T1g, T1v, T7r, T7u, T7x, T7y, T3j; E T69, TfI, ThD, TfP, ThG, TfM, ThC, T3M, T66, T3C, T67, T3P, T6a, Tbl, TdZ; E TbA, TdX, TfB, ThF, T8W, T9C, T8Z, T9D, TbI, TdW, TbL, Te0; { E T3, Ta6, T6, Tcu, T4I, Ta7, T4F, Tcv, Td, Tcy, T27, Tae, Ta, Tcx, T2a; E Tab; { E T1, T2, T4D, T4E; T1 = cr[0]; T2 = ci[WS(rs, 31)]; T3 = T1 + T2; Ta6 = T1 - T2; { E T4, T5, T4G, T4H; T4 = cr[WS(rs, 16)]; T5 = ci[WS(rs, 15)]; T6 = T4 + T5; Tcu = T4 - T5; T4G = ci[WS(rs, 47)]; T4H = cr[WS(rs, 48)]; T4I = T4G - T4H; Ta7 = T4G + T4H; } T4D = ci[WS(rs, 63)]; T4E = cr[WS(rs, 32)]; T4F = T4D - T4E; Tcv = T4D + T4E; { E Tb, Tc, Tac, T25, T26, Tad; Tb = ci[WS(rs, 7)]; Tc = cr[WS(rs, 24)]; Tac = Tb - Tc; T25 = ci[WS(rs, 39)]; T26 = cr[WS(rs, 56)]; Tad = T25 + T26; Td = Tb + Tc; Tcy = Tac + Tad; T27 = T25 - T26; Tae = Tac - Tad; } { E T8, T9, Ta9, T28, T29, Taa; T8 = cr[WS(rs, 8)]; T9 = ci[WS(rs, 23)]; Ta9 = T8 - T9; T28 = ci[WS(rs, 55)]; T29 = cr[WS(rs, 40)]; Taa = T28 + T29; Ta = T8 + T9; Tcx = Ta9 + Taa; T2a = T28 - T29; Tab = Ta9 - Taa; } } { E T7, Te, Tf8, Tf9; T7 = T3 + T6; Te = Ta + Td; Tf = T7 + Te; T8C = T7 - Te; Tf8 = Ta6 + Ta7; Tf9 = KP707106781 * (Tcx + Tcy); Tfa = Tf8 - Tf9; Thk = Tf8 + Tf9; } { E Tge, Tgf, T24, T2b; Tge = Tcv - Tcu; Tgf = KP707106781 * (Tab - Tae); Tgg = Tge + Tgf; ThM = Tge - Tgf; T24 = T3 - T6; T2b = T27 - T2a; T2c = T24 + T2b; T5O = T24 - T2b; } { E T4C, T4J, Ta8, Taf; T4C = Ta - Td; T4J = T4F - T4I; T4K = T4C + T4J; T6g = T4J - T4C; Ta8 = Ta6 - Ta7; Taf = KP707106781 * (Tab + Tae); Tag = Ta8 - Taf; TdE = Ta8 + Taf; } { E Tcw, Tcz, T7N, T7O; Tcw = Tcu + Tcv; Tcz = KP707106781 * (Tcx - Tcy); TcA = Tcw - Tcz; Te6 = Tcw + Tcz; T7N = T4F + T4I; T7O = T2a + T27; T7P = T7N + T7O; T94 = T7N - T7O; } } { E TC, Tb1, T2Z, TaQ, T2X, Tb2, T7m, TaR, TJ, Tb4, Tb5, T2Q, T36, TaV, TaY; E T7n, Tfq, Tfr; { E Tw, Tx, Ty, Tz, TA, TB; Tw = cr[WS(rs, 2)]; Tx = ci[WS(rs, 29)]; Ty = Tw + Tx; Tz = cr[WS(rs, 18)]; TA = ci[WS(rs, 13)]; TB = Tz + TA; TC = Ty + TB; Tb1 = Tz - TA; T2Z = Ty - TB; TaQ = Tw - Tx; } { E T2R, T2S, T2T, T2U, T2V, T2W; T2R = ci[WS(rs, 61)]; T2S = cr[WS(rs, 34)]; T2T = T2R - T2S; T2U = ci[WS(rs, 45)]; T2V = cr[WS(rs, 50)]; T2W = T2U - T2V; T2X = T2T - T2W; Tb2 = T2R + T2S; T7m = T2T + T2W; TaR = T2U + T2V; } { E TF, TaT, T35, TaU, TI, TaW, T32, TaX; { E TD, TE, T33, T34; TD = cr[WS(rs, 10)]; TE = ci[WS(rs, 21)]; TF = TD + TE; TaT = TD - TE; T33 = ci[WS(rs, 53)]; T34 = cr[WS(rs, 42)]; T35 = T33 - T34; TaU = T33 + T34; } { E TG, TH, T30, T31; TG = ci[WS(rs, 5)]; TH = cr[WS(rs, 26)]; TI = TG + TH; TaW = TG - TH; T30 = ci[WS(rs, 37)]; T31 = cr[WS(rs, 58)]; T32 = T30 - T31; TaX = T30 + T31; } TJ = TF + TI; Tb4 = TaT + TaU; Tb5 = TaW + TaX; T2Q = TF - TI; T36 = T32 - T35; TaV = TaT - TaU; TaY = TaW - TaX; T7n = T35 + T32; } TK = TC + TJ; T7o = T7m + T7n; { E T2Y, T37, Tft, Tfu; T2Y = T2Q + T2X; T37 = T2Z + T36; T38 = FMA(KP923879532, T2Y, KP382683432 * T37); T4P = FNMS(KP382683432, T2Y, KP923879532 * T37); Tft = TaQ + TaR; Tfu = KP707106781 * (Tb4 + Tb5); Tfv = Tft - Tfu; Thn = Tft + Tfu; } { E T5U, T5V, TaS, TaZ; T5U = T2X - T2Q; T5V = T2Z - T36; T5W = FMA(KP382683432, T5U, KP923879532 * T5V); T6j = FNMS(KP923879532, T5U, KP382683432 * T5V); TaS = TaQ - TaR; TaZ = KP707106781 * (TaV + TaY); Tb0 = TaS - TaZ; TdK = TaS + TaZ; } Tfq = Tb2 - Tb1; Tfr = KP707106781 * (TaV - TaY); Tfs = Tfq + Tfr; Tho = Tfq - Tfr; { E T8I, T8J, Tb3, Tb6; T8I = TC - TJ; T8J = T7m - T7n; T8K = T8I + T8J; T97 = T8I - T8J; Tb3 = Tb1 + Tb2; Tb6 = KP707106781 * (Tb4 - Tb5); Tb7 = Tb3 - Tb6; TdL = Tb3 + Tb6; } } { E TR, TaI, T2G, Tax, T2E, TaJ, T7j, Tay, TY, TaL, TaM, T2x, T2N, TaC, TaF; E T7k, Tfj, Tfk; { E TL, TM, TN, TO, TP, TQ; TL = ci[WS(rs, 1)]; TM = cr[WS(rs, 30)]; TN = TL + TM; TO = cr[WS(rs, 14)]; TP = ci[WS(rs, 17)]; TQ = TO + TP; TR = TN + TQ; TaI = TL - TM; T2G = TN - TQ; Tax = TO - TP; } { E T2y, T2z, T2A, T2B, T2C, T2D; T2y = ci[WS(rs, 33)]; T2z = cr[WS(rs, 62)]; T2A = T2y - T2z; T2B = ci[WS(rs, 49)]; T2C = cr[WS(rs, 46)]; T2D = T2B - T2C; T2E = T2A - T2D; TaJ = T2B + T2C; T7j = T2A + T2D; Tay = T2y + T2z; } { E TU, TaA, T2M, TaB, TX, TaD, T2J, TaE; { E TS, TT, T2K, T2L; TS = cr[WS(rs, 6)]; TT = ci[WS(rs, 25)]; TU = TS + TT; TaA = TS - TT; T2K = ci[WS(rs, 57)]; T2L = cr[WS(rs, 38)]; T2M = T2K - T2L; TaB = T2K + T2L; } { E TV, TW, T2H, T2I; TV = ci[WS(rs, 9)]; TW = cr[WS(rs, 22)]; TX = TV + TW; TaD = TV - TW; T2H = ci[WS(rs, 41)]; T2I = cr[WS(rs, 54)]; T2J = T2H - T2I; TaE = T2H + T2I; } TY = TU + TX; TaL = TaA - TaB; TaM = TaD - TaE; T2x = TU - TX; T2N = T2J - T2M; TaC = TaA + TaB; TaF = TaD + TaE; T7k = T2M + T2J; } TZ = TR + TY; T7l = T7j + T7k; { E T2F, T2O, Tfm, Tfn; T2F = T2x + T2E; T2O = T2G + T2N; T2P = FNMS(KP382683432, T2O, KP923879532 * T2F); T4Q = FMA(KP382683432, T2F, KP923879532 * T2O); Tfm = TaI + TaJ; Tfn = KP707106781 * (TaC + TaF); Tfo = Tfm - Tfn; Thq = Tfm + Tfn; } { E T5R, T5S, Taz, TaG; T5R = T2E - T2x; T5S = T2G - T2N; T5T = FNMS(KP923879532, T5S, KP382683432 * T5R); T6k = FMA(KP923879532, T5R, KP382683432 * T5S); Taz = Tax - Tay; TaG = KP707106781 * (TaC - TaF); TaH = Taz - TaG; TdH = Taz + TaG; } Tfj = KP707106781 * (TaL - TaM); Tfk = Tax + Tay; Tfl = Tfj - Tfk; Thr = Tfk + Tfj; { E T8F, T8G, TaK, TaN; T8F = T7j - T7k; T8G = TR - TY; T8H = T8F - T8G; T98 = T8G + T8F; TaK = TaI - TaJ; TaN = KP707106781 * (TaL + TaM); TaO = TaK - TaN; TdI = TaK + TaN; } } { E Ti, T2j, Tl, T2g, T2d, T2k, Tfc, Tfb, Tat, Taq, Tp, T2s, Ts, T2p, T2m; E T2t, Tff, Tfe, Tam, Taj; { E Tar, Tas, Tao, Tap; { E Tg, Th, T2h, T2i; Tg = cr[WS(rs, 4)]; Th = ci[WS(rs, 27)]; Ti = Tg + Th; Tar = Tg - Th; T2h = ci[WS(rs, 43)]; T2i = cr[WS(rs, 52)]; T2j = T2h - T2i; Tas = T2h + T2i; } { E Tj, Tk, T2e, T2f; Tj = cr[WS(rs, 20)]; Tk = ci[WS(rs, 11)]; Tl = Tj + Tk; Tao = Tj - Tk; T2e = ci[WS(rs, 59)]; T2f = cr[WS(rs, 36)]; T2g = T2e - T2f; Tap = T2e + T2f; } T2d = Ti - Tl; T2k = T2g - T2j; Tfc = Tap - Tao; Tfb = Tar + Tas; Tat = Tar - Tas; Taq = Tao + Tap; } { E Tak, Tal, Tah, Tai; { E Tn, To, T2q, T2r; Tn = ci[WS(rs, 3)]; To = cr[WS(rs, 28)]; Tp = Tn + To; Tak = Tn - To; T2q = ci[WS(rs, 51)]; T2r = cr[WS(rs, 44)]; T2s = T2q - T2r; Tal = T2q + T2r; } { E Tq, Tr, T2n, T2o; Tq = cr[WS(rs, 12)]; Tr = ci[WS(rs, 19)]; Ts = Tq + Tr; Tah = Tq - Tr; T2n = ci[WS(rs, 35)]; T2o = cr[WS(rs, 60)]; T2p = T2n - T2o; Tai = T2n + T2o; } T2m = Tp - Ts; T2t = T2p - T2s; Tff = Tah + Tai; Tfe = Tak + Tal; Tam = Tak - Tal; Taj = Tah - Tai; } { E Tm, Tt, Tfd, Tfg; Tm = Ti + Tl; Tt = Tp + Ts; Tu = Tm + Tt; T95 = Tm - Tt; Tfd = FNMS(KP923879532, Tfc, KP382683432 * Tfb); Tfg = FNMS(KP923879532, Tff, KP382683432 * Tfe); Tfh = Tfd + Tfg; ThN = Tfd - Tfg; } { E Tgh, Tgi, T2l, T2u; Tgh = FMA(KP382683432, Tfc, KP923879532 * Tfb); Tgi = FMA(KP382683432, Tff, KP923879532 * Tfe); Tgj = Tgh - Tgi; Thl = Tgh + Tgi; T2l = T2d - T2k; T2u = T2m + T2t; T2v = KP707106781 * (T2l + T2u); T6h = KP707106781 * (T2l - T2u); } { E T4L, T4M, Tan, Tau; T4L = T2d + T2k; T4M = T2t - T2m; T4N = KP707106781 * (T4L + T4M); T5P = KP707106781 * (T4M - T4L); Tan = FNMS(KP382683432, Tam, KP923879532 * Taj); Tau = FMA(KP923879532, Taq, KP382683432 * Tat); Tav = Tan - Tau; Te7 = Tau + Tan; } { E TcB, TcC, T7Q, T7R; TcB = FNMS(KP382683432, Taq, KP923879532 * Tat); TcC = FMA(KP382683432, Taj, KP923879532 * Tam); TcD = TcB - TcC; TdF = TcB + TcC; T7Q = T2g + T2j; T7R = T2p + T2s; T7S = T7Q + T7R; T8D = T7R - T7Q; } } { E T1z, T1C, T1D, Tcf, TbO, T4o, T4r, T7B, Tcg, TbP, T1G, T3Y, T1J, T3V, T1K; E T7C, Tcj, Tci, TbW, TbT, T1S, TfV, TfW, T41, T48, Tc8, Tcb, T7E, T1Z, TfY; E TfZ, T4a, T4h, Tc1, Tc4, T7F; { E T1x, T1y, T1A, T1B; T1x = ci[0]; T1y = cr[WS(rs, 31)]; T1z = T1x + T1y; T1A = cr[WS(rs, 15)]; T1B = ci[WS(rs, 16)]; T1C = T1A + T1B; T1D = T1z + T1C; Tcf = T1A - T1B; TbO = T1x - T1y; } { E T4m, T4n, T4p, T4q; T4m = ci[WS(rs, 32)]; T4n = cr[WS(rs, 63)]; T4o = T4m - T4n; T4p = ci[WS(rs, 48)]; T4q = cr[WS(rs, 47)]; T4r = T4p - T4q; T7B = T4o + T4r; Tcg = T4m + T4n; TbP = T4p + T4q; } { E TbR, TbS, TbU, TbV; { E T1E, T1F, T3W, T3X; T1E = cr[WS(rs, 7)]; T1F = ci[WS(rs, 24)]; T1G = T1E + T1F; TbR = T1E - T1F; T3W = ci[WS(rs, 56)]; T3X = cr[WS(rs, 39)]; T3Y = T3W - T3X; TbS = T3W + T3X; } { E T1H, T1I, T3T, T3U; T1H = ci[WS(rs, 8)]; T1I = cr[WS(rs, 23)]; T1J = T1H + T1I; TbU = T1H - T1I; T3T = ci[WS(rs, 40)]; T3U = cr[WS(rs, 55)]; T3V = T3T - T3U; TbV = T3T + T3U; } T1K = T1G + T1J; T7C = T3Y + T3V; Tcj = TbU + TbV; Tci = TbR + TbS; TbW = TbU - TbV; TbT = TbR - TbS; } { E T1O, Tc9, T47, Tca, T1R, Tc6, T44, Tc7; { E T1M, T1N, T45, T46; T1M = cr[WS(rs, 3)]; T1N = ci[WS(rs, 28)]; T1O = T1M + T1N; Tc9 = T1M - T1N; T45 = ci[WS(rs, 44)]; T46 = cr[WS(rs, 51)]; T47 = T45 - T46; Tca = T45 + T46; } { E T1P, T1Q, T42, T43; T1P = cr[WS(rs, 19)]; T1Q = ci[WS(rs, 12)]; T1R = T1P + T1Q; Tc6 = T1P - T1Q; T42 = ci[WS(rs, 60)]; T43 = cr[WS(rs, 35)]; T44 = T42 - T43; Tc7 = T42 + T43; } T1S = T1O + T1R; TfV = Tc9 + Tca; TfW = Tc7 - Tc6; T41 = T1O - T1R; T48 = T44 - T47; Tc8 = Tc6 + Tc7; Tcb = Tc9 - Tca; T7E = T44 + T47; } { E T1V, Tc2, T4g, Tc3, T1Y, TbZ, T4d, Tc0; { E T1T, T1U, T4e, T4f; T1T = ci[WS(rs, 4)]; T1U = cr[WS(rs, 27)]; T1V = T1T + T1U; Tc2 = T1T - T1U; T4e = ci[WS(rs, 52)]; T4f = cr[WS(rs, 43)]; T4g = T4e - T4f; Tc3 = T4e + T4f; } { E T1W, T1X, T4b, T4c; T1W = cr[WS(rs, 11)]; T1X = ci[WS(rs, 20)]; T1Y = T1W + T1X; TbZ = T1W - T1X; T4b = ci[WS(rs, 36)]; T4c = cr[WS(rs, 59)]; T4d = T4b - T4c; Tc0 = T4b + T4c; } T1Z = T1V + T1Y; TfY = Tc2 + Tc3; TfZ = TbZ + Tc0; T4a = T1V - T1Y; T4h = T4d - T4g; Tc1 = TbZ - Tc0; Tc4 = Tc2 - Tc3; T7F = T4d + T4g; } T1L = T1D + T1K; T20 = T1S + T1Z; T7A = T1L - T20; T7D = T7B + T7C; T7G = T7E + T7F; T7H = T7D - T7G; { E T3S, T3Z, TfX, Tg0; T3S = T1z - T1C; T3Z = T3V - T3Y; T40 = T3S + T3Z; T62 = T3S - T3Z; TfX = FNMS(KP923879532, TfW, KP382683432 * TfV); Tg0 = FNMS(KP923879532, TfZ, KP382683432 * TfY); Tg1 = TfX + Tg0; Thv = TfX - Tg0; } { E Tg6, Tg7, Tg3, Tg4; Tg6 = FMA(KP382683432, TfW, KP923879532 * TfV); Tg7 = FMA(KP382683432, TfZ, KP923879532 * TfY); Tg8 = Tg6 - Tg7; Thz = Tg6 + Tg7; Tg3 = KP707106781 * (TbT - TbW); Tg4 = Tcf + Tcg; Tg5 = Tg3 - Tg4; Thw = Tg4 + Tg3; } { E T4l, T4s, T49, T4i; T4l = T1G - T1J; T4s = T4o - T4r; T4t = T4l + T4s; T5Z = T4s - T4l; T49 = T41 - T48; T4i = T4a + T4h; T4j = KP707106781 * (T49 + T4i); T60 = KP707106781 * (T49 - T4i); } { E T4u, T4v, TbQ, TbX; T4u = T41 + T48; T4v = T4h - T4a; T4w = KP707106781 * (T4u + T4v); T63 = KP707106781 * (T4v - T4u); TbQ = TbO - TbP; TbX = KP707106781 * (TbT + TbW); TbY = TbQ - TbX; TdS = TbQ + TbX; } { E Tc5, Tcc, TfS, TfT; Tc5 = FNMS(KP382683432, Tc4, KP923879532 * Tc1); Tcc = FMA(KP923879532, Tc8, KP382683432 * Tcb); Tcd = Tc5 - Tcc; TdQ = Tcc + Tc5; TfS = TbO + TbP; TfT = KP707106781 * (Tci + Tcj); TfU = TfS - TfT; Thy = TfS + TfT; } { E T8N, T8O, T8Q, T8R; T8N = T7B - T7C; T8O = T1S - T1Z; T8P = T8N - T8O; T9z = T8O + T8N; T8Q = T1D - T1K; T8R = T7F - T7E; T8S = T8Q - T8R; T9A = T8Q + T8R; } { E Tch, Tck, Tcm, Tcn; Tch = Tcf - Tcg; Tck = KP707106781 * (Tci - Tcj); Tcl = Tch - Tck; TdP = Tch + Tck; Tcm = FNMS(KP382683432, Tc8, KP923879532 * Tcb); Tcn = FMA(KP382683432, Tc1, KP923879532 * Tc4); Tco = Tcm - Tcn; TdT = Tcm + Tcn; } } { E T14, T17, T18, TbC, Tbb, T3H, T3K, T7s, TbD, Tbc, T1b, T3h, T1e, T3e, T1f; E T7t, TbG, TbF, Tbj, Tbg, T1n, TfC, TfD, T3k, T3r, Tbv, Tby, T7v, T1u, TfF; E TfG, T3t, T3A, Tbo, Tbr, T7w; { E T12, T13, T15, T16; T12 = cr[WS(rs, 1)]; T13 = ci[WS(rs, 30)]; T14 = T12 + T13; T15 = cr[WS(rs, 17)]; T16 = ci[WS(rs, 14)]; T17 = T15 + T16; T18 = T14 + T17; TbC = T15 - T16; Tbb = T12 - T13; } { E T3F, T3G, T3I, T3J; T3F = ci[WS(rs, 62)]; T3G = cr[WS(rs, 33)]; T3H = T3F - T3G; T3I = ci[WS(rs, 46)]; T3J = cr[WS(rs, 49)]; T3K = T3I - T3J; T7s = T3H + T3K; TbD = T3F + T3G; Tbc = T3I + T3J; } { E Tbe, Tbf, Tbh, Tbi; { E T19, T1a, T3f, T3g; T19 = cr[WS(rs, 9)]; T1a = ci[WS(rs, 22)]; T1b = T19 + T1a; Tbe = T19 - T1a; T3f = ci[WS(rs, 54)]; T3g = cr[WS(rs, 41)]; T3h = T3f - T3g; Tbf = T3f + T3g; } { E T1c, T1d, T3c, T3d; T1c = ci[WS(rs, 6)]; T1d = cr[WS(rs, 25)]; T1e = T1c + T1d; Tbh = T1c - T1d; T3c = ci[WS(rs, 38)]; T3d = cr[WS(rs, 57)]; T3e = T3c - T3d; Tbi = T3c + T3d; } T1f = T1b + T1e; T7t = T3h + T3e; TbG = Tbh + Tbi; TbF = Tbe + Tbf; Tbj = Tbh - Tbi; Tbg = Tbe - Tbf; } { E T1j, Tbw, T3q, Tbx, T1m, Tbt, T3n, Tbu; { E T1h, T1i, T3o, T3p; T1h = cr[WS(rs, 5)]; T1i = ci[WS(rs, 26)]; T1j = T1h + T1i; Tbw = T1h - T1i; T3o = ci[WS(rs, 42)]; T3p = cr[WS(rs, 53)]; T3q = T3o - T3p; Tbx = T3o + T3p; } { E T1k, T1l, T3l, T3m; T1k = cr[WS(rs, 21)]; T1l = ci[WS(rs, 10)]; T1m = T1k + T1l; Tbt = T1k - T1l; T3l = ci[WS(rs, 58)]; T3m = cr[WS(rs, 37)]; T3n = T3l - T3m; Tbu = T3l + T3m; } T1n = T1j + T1m; TfC = Tbw + Tbx; TfD = Tbu - Tbt; T3k = T1j - T1m; T3r = T3n - T3q; Tbv = Tbt + Tbu; Tby = Tbw - Tbx; T7v = T3n + T3q; } { E T1q, Tbp, T3z, Tbq, T1t, Tbm, T3w, Tbn; { E T1o, T1p, T3x, T3y; T1o = ci[WS(rs, 2)]; T1p = cr[WS(rs, 29)]; T1q = T1o + T1p; Tbp = T1o - T1p; T3x = ci[WS(rs, 50)]; T3y = cr[WS(rs, 45)]; T3z = T3x - T3y; Tbq = T3x + T3y; } { E T1r, T1s, T3u, T3v; T1r = cr[WS(rs, 13)]; T1s = ci[WS(rs, 18)]; T1t = T1r + T1s; Tbm = T1r - T1s; T3u = ci[WS(rs, 34)]; T3v = cr[WS(rs, 61)]; T3w = T3u - T3v; Tbn = T3u + T3v; } T1u = T1q + T1t; TfF = Tbp + Tbq; TfG = Tbm + Tbn; T3t = T1q - T1t; T3A = T3w - T3z; Tbo = Tbm - Tbn; Tbr = Tbp - Tbq; T7w = T3w + T3z; } T1g = T18 + T1f; T1v = T1n + T1u; T7r = T1g - T1v; T7u = T7s + T7t; T7x = T7v + T7w; T7y = T7u - T7x; { E T3b, T3i, TfE, TfH; T3b = T14 - T17; T3i = T3e - T3h; T3j = T3b + T3i; T69 = T3b - T3i; TfE = FNMS(KP923879532, TfD, KP382683432 * TfC); TfH = FNMS(KP923879532, TfG, KP382683432 * TfF); TfI = TfE + TfH; ThD = TfE - TfH; } { E TfN, TfO, TfK, TfL; TfN = FMA(KP382683432, TfD, KP923879532 * TfC); TfO = FMA(KP382683432, TfG, KP923879532 * TfF); TfP = TfN - TfO; ThG = TfN + TfO; TfK = TbD - TbC; TfL = KP707106781 * (Tbg - Tbj); TfM = TfK + TfL; ThC = TfK - TfL; } { E T3E, T3L, T3s, T3B; T3E = T1b - T1e; T3L = T3H - T3K; T3M = T3E + T3L; T66 = T3L - T3E; T3s = T3k - T3r; T3B = T3t + T3A; T3C = KP707106781 * (T3s + T3B); T67 = KP707106781 * (T3s - T3B); } { E T3N, T3O, Tbd, Tbk; T3N = T3k + T3r; T3O = T3A - T3t; T3P = KP707106781 * (T3N + T3O); T6a = KP707106781 * (T3O - T3N); Tbd = Tbb - Tbc; Tbk = KP707106781 * (Tbg + Tbj); Tbl = Tbd - Tbk; TdZ = Tbd + Tbk; } { E Tbs, Tbz, Tfz, TfA; Tbs = FNMS(KP382683432, Tbr, KP923879532 * Tbo); Tbz = FMA(KP923879532, Tbv, KP382683432 * Tby); TbA = Tbs - Tbz; TdX = Tbz + Tbs; Tfz = Tbb + Tbc; TfA = KP707106781 * (TbF + TbG); TfB = Tfz - TfA; ThF = Tfz + TfA; } { E T8U, T8V, T8X, T8Y; T8U = T7s - T7t; T8V = T1n - T1u; T8W = T8U - T8V; T9C = T8V + T8U; T8X = T18 - T1f; T8Y = T7w - T7v; T8Z = T8X - T8Y; T9D = T8X + T8Y; } { E TbE, TbH, TbJ, TbK; TbE = TbC + TbD; TbH = KP707106781 * (TbF - TbG); TbI = TbE - TbH; TdW = TbE + TbH; TbJ = FNMS(KP382683432, Tbv, KP923879532 * Tby); TbK = FMA(KP382683432, Tbo, KP923879532 * Tbr); TbL = TbJ - TbK; Te0 = TbJ + TbK; } } { E T11, T8q, T8n, T8r, T22, T8v, T8k, T8u; { E Tv, T10, T8l, T8m; Tv = Tf + Tu; T10 = TK + TZ; T11 = Tv + T10; T8q = Tv - T10; T8l = T7u + T7x; T8m = T7D + T7G; T8n = T8l + T8m; T8r = T8m - T8l; } { E T1w, T21, T8i, T8j; T1w = T1g + T1v; T21 = T1L + T20; T22 = T1w + T21; T8v = T1w - T21; T8i = T7P + T7S; T8j = T7o + T7l; T8k = T8i + T8j; T8u = T8i - T8j; } cr[0] = T11 + T22; ci[0] = T8k + T8n; { E T8g, T8o, T8f, T8h; T8g = T11 - T22; T8o = T8k - T8n; T8f = W[62]; T8h = W[63]; cr[WS(rs, 32)] = FNMS(T8h, T8o, T8f * T8g); ci[WS(rs, 32)] = FMA(T8h, T8g, T8f * T8o); } { E T8s, T8w, T8p, T8t; T8s = T8q - T8r; T8w = T8u - T8v; T8p = W[94]; T8t = W[95]; cr[WS(rs, 48)] = FNMS(T8t, T8w, T8p * T8s); ci[WS(rs, 48)] = FMA(T8p, T8w, T8t * T8s); } { E T8y, T8A, T8x, T8z; T8y = T8q + T8r; T8A = T8v + T8u; T8x = W[30]; T8z = W[31]; cr[WS(rs, 16)] = FNMS(T8z, T8A, T8x * T8y); ci[WS(rs, 16)] = FMA(T8x, T8A, T8z * T8y); } } { E T9y, T9U, T9N, T9V, T9F, T9Z, T9K, T9Y; { E T9w, T9x, T9L, T9M; T9w = T8C + T8D; T9x = KP707106781 * (T97 + T98); T9y = T9w - T9x; T9U = T9w + T9x; T9L = FNMS(KP382683432, T9C, KP923879532 * T9D); T9M = FMA(KP382683432, T9z, KP923879532 * T9A); T9N = T9L - T9M; T9V = T9L + T9M; } { E T9B, T9E, T9I, T9J; T9B = FNMS(KP382683432, T9A, KP923879532 * T9z); T9E = FMA(KP923879532, T9C, KP382683432 * T9D); T9F = T9B - T9E; T9Z = T9E + T9B; T9I = T95 + T94; T9J = KP707106781 * (T8K + T8H); T9K = T9I - T9J; T9Y = T9I + T9J; } { E T9G, T9O, T9v, T9H; T9G = T9y - T9F; T9O = T9K - T9N; T9v = W[102]; T9H = W[103]; cr[WS(rs, 52)] = FNMS(T9H, T9O, T9v * T9G); ci[WS(rs, 52)] = FMA(T9H, T9G, T9v * T9O); } { E Ta2, Ta4, Ta1, Ta3; Ta2 = T9U + T9V; Ta4 = T9Y + T9Z; Ta1 = W[6]; Ta3 = W[7]; cr[WS(rs, 4)] = FNMS(Ta3, Ta4, Ta1 * Ta2); ci[WS(rs, 4)] = FMA(Ta1, Ta4, Ta3 * Ta2); } { E T9Q, T9S, T9P, T9R; T9Q = T9y + T9F; T9S = T9K + T9N; T9P = W[38]; T9R = W[39]; cr[WS(rs, 20)] = FNMS(T9R, T9S, T9P * T9Q); ci[WS(rs, 20)] = FMA(T9R, T9Q, T9P * T9S); } { E T9W, Ta0, T9T, T9X; T9W = T9U - T9V; Ta0 = T9Y - T9Z; T9T = W[70]; T9X = W[71]; cr[WS(rs, 36)] = FNMS(T9X, Ta0, T9T * T9W); ci[WS(rs, 36)] = FMA(T9T, Ta0, T9X * T9W); } } { E T8M, T9k, T9d, T9l, T91, T9p, T9a, T9o; { E T8E, T8L, T9b, T9c; T8E = T8C - T8D; T8L = KP707106781 * (T8H - T8K); T8M = T8E - T8L; T9k = T8E + T8L; T9b = FNMS(KP923879532, T8W, KP382683432 * T8Z); T9c = FMA(KP923879532, T8P, KP382683432 * T8S); T9d = T9b - T9c; T9l = T9b + T9c; } { E T8T, T90, T96, T99; T8T = FNMS(KP923879532, T8S, KP382683432 * T8P); T90 = FMA(KP382683432, T8W, KP923879532 * T8Z); T91 = T8T - T90; T9p = T90 + T8T; T96 = T94 - T95; T99 = KP707106781 * (T97 - T98); T9a = T96 - T99; T9o = T96 + T99; } { E T92, T9e, T8B, T93; T92 = T8M - T91; T9e = T9a - T9d; T8B = W[118]; T93 = W[119]; cr[WS(rs, 60)] = FNMS(T93, T9e, T8B * T92); ci[WS(rs, 60)] = FMA(T93, T92, T8B * T9e); } { E T9s, T9u, T9r, T9t; T9s = T9k + T9l; T9u = T9o + T9p; T9r = W[22]; T9t = W[23]; cr[WS(rs, 12)] = FNMS(T9t, T9u, T9r * T9s); ci[WS(rs, 12)] = FMA(T9r, T9u, T9t * T9s); } { E T9g, T9i, T9f, T9h; T9g = T8M + T91; T9i = T9a + T9d; T9f = W[54]; T9h = W[55]; cr[WS(rs, 28)] = FNMS(T9h, T9i, T9f * T9g); ci[WS(rs, 28)] = FMA(T9h, T9g, T9f * T9i); } { E T9m, T9q, T9j, T9n; T9m = T9k - T9l; T9q = T9o - T9p; T9j = W[86]; T9n = W[87]; cr[WS(rs, 44)] = FNMS(T9n, T9q, T9j * T9m); ci[WS(rs, 44)] = FMA(T9j, T9q, T9n * T9m); } } { E T7q, T84, T7X, T85, T7J, T89, T7U, T88; { E T7i, T7p, T7V, T7W; T7i = Tf - Tu; T7p = T7l - T7o; T7q = T7i + T7p; T84 = T7i - T7p; T7V = T7r + T7y; T7W = T7H - T7A; T7X = KP707106781 * (T7V + T7W); T85 = KP707106781 * (T7W - T7V); } { E T7z, T7I, T7M, T7T; T7z = T7r - T7y; T7I = T7A + T7H; T7J = KP707106781 * (T7z + T7I); T89 = KP707106781 * (T7z - T7I); T7M = TK - TZ; T7T = T7P - T7S; T7U = T7M + T7T; T88 = T7T - T7M; } { E T7K, T7Y, T7h, T7L; T7K = T7q - T7J; T7Y = T7U - T7X; T7h = W[78]; T7L = W[79]; cr[WS(rs, 40)] = FNMS(T7L, T7Y, T7h * T7K); ci[WS(rs, 40)] = FMA(T7L, T7K, T7h * T7Y); } { E T8c, T8e, T8b, T8d; T8c = T84 + T85; T8e = T88 + T89; T8b = W[46]; T8d = W[47]; cr[WS(rs, 24)] = FNMS(T8d, T8e, T8b * T8c); ci[WS(rs, 24)] = FMA(T8b, T8e, T8d * T8c); } { E T80, T82, T7Z, T81; T80 = T7q + T7J; T82 = T7U + T7X; T7Z = W[14]; T81 = W[15]; cr[WS(rs, 8)] = FNMS(T81, T82, T7Z * T80); ci[WS(rs, 8)] = FMA(T81, T80, T7Z * T82); } { E T86, T8a, T83, T87; T86 = T84 - T85; T8a = T88 - T89; T83 = W[110]; T87 = W[111]; cr[WS(rs, 56)] = FNMS(T87, T8a, T83 * T86); ci[WS(rs, 56)] = FMA(T83, T8a, T87 * T86); } } { E T6K, T76, T6W, T7a, T6R, T7b, T6Z, T77; { E T6I, T6J, T6U, T6V; T6I = T5O + T5P; T6J = T6j + T6k; T6K = T6I - T6J; T76 = T6I + T6J; T6U = T6g + T6h; T6V = T5W + T5T; T6W = T6U - T6V; T7a = T6U + T6V; { E T6N, T6Y, T6Q, T6X; { E T6L, T6M, T6O, T6P; T6L = T5Z + T60; T6M = T62 + T63; T6N = FNMS(KP555570233, T6M, KP831469612 * T6L); T6Y = FMA(KP555570233, T6L, KP831469612 * T6M); T6O = T66 + T67; T6P = T69 + T6a; T6Q = FMA(KP831469612, T6O, KP555570233 * T6P); T6X = FNMS(KP555570233, T6O, KP831469612 * T6P); } T6R = T6N - T6Q; T7b = T6Q + T6N; T6Z = T6X - T6Y; T77 = T6X + T6Y; } } { E T6S, T70, T6H, T6T; T6S = T6K - T6R; T70 = T6W - T6Z; T6H = W[106]; T6T = W[107]; cr[WS(rs, 54)] = FNMS(T6T, T70, T6H * T6S); ci[WS(rs, 54)] = FMA(T6T, T6S, T6H * T70); } { E T7e, T7g, T7d, T7f; T7e = T76 + T77; T7g = T7a + T7b; T7d = W[10]; T7f = W[11]; cr[WS(rs, 6)] = FNMS(T7f, T7g, T7d * T7e); ci[WS(rs, 6)] = FMA(T7d, T7g, T7f * T7e); } { E T72, T74, T71, T73; T72 = T6K + T6R; T74 = T6W + T6Z; T71 = W[42]; T73 = W[43]; cr[WS(rs, 22)] = FNMS(T73, T74, T71 * T72); ci[WS(rs, 22)] = FMA(T73, T72, T71 * T74); } { E T78, T7c, T75, T79; T78 = T76 - T77; T7c = T7a - T7b; T75 = W[74]; T79 = W[75]; cr[WS(rs, 38)] = FNMS(T79, T7c, T75 * T78); ci[WS(rs, 38)] = FMA(T75, T7c, T79 * T78); } } { E T3a, T52, T4S, T56, T4z, T57, T4V, T53; { E T2w, T39, T4O, T4R; T2w = T2c - T2v; T39 = T2P - T38; T3a = T2w + T39; T52 = T2w - T39; T4O = T4K - T4N; T4R = T4P - T4Q; T4S = T4O + T4R; T56 = T4O - T4R; { E T3R, T4T, T4y, T4U; { E T3D, T3Q, T4k, T4x; T3D = T3j - T3C; T3Q = T3M - T3P; T3R = FNMS(KP831469612, T3Q, KP555570233 * T3D); T4T = FMA(KP831469612, T3D, KP555570233 * T3Q); T4k = T40 - T4j; T4x = T4t - T4w; T4y = FMA(KP555570233, T4k, KP831469612 * T4x); T4U = FNMS(KP831469612, T4k, KP555570233 * T4x); } T4z = T3R + T4y; T57 = T3R - T4y; T4V = T4T + T4U; T53 = T4U - T4T; } } { E T4A, T4W, T23, T4B; T4A = T3a - T4z; T4W = T4S - T4V; T23 = W[82]; T4B = W[83]; cr[WS(rs, 42)] = FNMS(T4B, T4W, T23 * T4A); ci[WS(rs, 42)] = FMA(T4B, T4A, T23 * T4W); } { E T5a, T5c, T59, T5b; T5a = T52 + T53; T5c = T56 + T57; T59 = W[50]; T5b = W[51]; cr[WS(rs, 26)] = FNMS(T5b, T5c, T59 * T5a); ci[WS(rs, 26)] = FMA(T59, T5c, T5b * T5a); } { E T4Y, T50, T4X, T4Z; T4Y = T3a + T4z; T50 = T4S + T4V; T4X = W[18]; T4Z = W[19]; cr[WS(rs, 10)] = FNMS(T4Z, T50, T4X * T4Y); ci[WS(rs, 10)] = FMA(T4Z, T4Y, T4X * T50); } { E T54, T58, T51, T55; T54 = T52 - T53; T58 = T56 - T57; T51 = W[114]; T55 = W[115]; cr[WS(rs, 58)] = FNMS(T55, T58, T51 * T54); ci[WS(rs, 58)] = FMA(T51, T58, T55 * T54); } } { E T5g, T5C, T5s, T5G, T5n, T5H, T5v, T5D; { E T5e, T5f, T5q, T5r; T5e = T2c + T2v; T5f = T4P + T4Q; T5g = T5e + T5f; T5C = T5e - T5f; T5q = T4K + T4N; T5r = T38 + T2P; T5s = T5q + T5r; T5G = T5q - T5r; { E T5j, T5t, T5m, T5u; { E T5h, T5i, T5k, T5l; T5h = T3j + T3C; T5i = T3M + T3P; T5j = FNMS(KP195090322, T5i, KP980785280 * T5h); T5t = FMA(KP195090322, T5h, KP980785280 * T5i); T5k = T40 + T4j; T5l = T4t + T4w; T5m = FMA(KP980785280, T5k, KP195090322 * T5l); T5u = FNMS(KP195090322, T5k, KP980785280 * T5l); } T5n = T5j + T5m; T5H = T5j - T5m; T5v = T5t + T5u; T5D = T5u - T5t; } } { E T5o, T5w, T5d, T5p; T5o = T5g - T5n; T5w = T5s - T5v; T5d = W[66]; T5p = W[67]; cr[WS(rs, 34)] = FNMS(T5p, T5w, T5d * T5o); ci[WS(rs, 34)] = FMA(T5p, T5o, T5d * T5w); } { E T5K, T5M, T5J, T5L; T5K = T5C + T5D; T5M = T5G + T5H; T5J = W[34]; T5L = W[35]; cr[WS(rs, 18)] = FNMS(T5L, T5M, T5J * T5K); ci[WS(rs, 18)] = FMA(T5J, T5M, T5L * T5K); } { E T5y, T5A, T5x, T5z; T5y = T5g + T5n; T5A = T5s + T5v; T5x = W[2]; T5z = W[3]; cr[WS(rs, 2)] = FNMS(T5z, T5A, T5x * T5y); ci[WS(rs, 2)] = FMA(T5z, T5y, T5x * T5A); } { E T5E, T5I, T5B, T5F; T5E = T5C - T5D; T5I = T5G - T5H; T5B = W[98]; T5F = W[99]; cr[WS(rs, 50)] = FNMS(T5F, T5I, T5B * T5E); ci[WS(rs, 50)] = FMA(T5B, T5I, T5F * T5E); } } { E T5Y, T6w, T6m, T6A, T6d, T6B, T6p, T6x; { E T5Q, T5X, T6i, T6l; T5Q = T5O - T5P; T5X = T5T - T5W; T5Y = T5Q - T5X; T6w = T5Q + T5X; T6i = T6g - T6h; T6l = T6j - T6k; T6m = T6i - T6l; T6A = T6i + T6l; { E T65, T6o, T6c, T6n; { E T61, T64, T68, T6b; T61 = T5Z - T60; T64 = T62 - T63; T65 = FNMS(KP980785280, T64, KP195090322 * T61); T6o = FMA(KP980785280, T61, KP195090322 * T64); T68 = T66 - T67; T6b = T69 - T6a; T6c = FMA(KP195090322, T68, KP980785280 * T6b); T6n = FNMS(KP980785280, T68, KP195090322 * T6b); } T6d = T65 - T6c; T6B = T6c + T65; T6p = T6n - T6o; T6x = T6n + T6o; } } { E T6e, T6q, T5N, T6f; T6e = T5Y - T6d; T6q = T6m - T6p; T5N = W[122]; T6f = W[123]; cr[WS(rs, 62)] = FNMS(T6f, T6q, T5N * T6e); ci[WS(rs, 62)] = FMA(T6f, T6e, T5N * T6q); } { E T6E, T6G, T6D, T6F; T6E = T6w + T6x; T6G = T6A + T6B; T6D = W[26]; T6F = W[27]; cr[WS(rs, 14)] = FNMS(T6F, T6G, T6D * T6E); ci[WS(rs, 14)] = FMA(T6D, T6G, T6F * T6E); } { E T6s, T6u, T6r, T6t; T6s = T5Y + T6d; T6u = T6m + T6p; T6r = W[58]; T6t = W[59]; cr[WS(rs, 30)] = FNMS(T6t, T6u, T6r * T6s); ci[WS(rs, 30)] = FMA(T6t, T6s, T6r * T6u); } { E T6y, T6C, T6v, T6z; T6y = T6w - T6x; T6C = T6A - T6B; T6v = W[90]; T6z = W[91]; cr[WS(rs, 46)] = FNMS(T6z, T6C, T6v * T6y); ci[WS(rs, 46)] = FMA(T6v, T6C, T6z * T6y); } } { E Tba, Tdw, TcS, Tdi, TcI, Tds, TcW, Td6, Tcr, TcX, TcL, TcT, Tdd, Tdx, Tdl; E Tdt; { E Taw, Tdg, Tb9, Tdh, TaP, Tb8; Taw = Tag - Tav; Tdg = TcA + TcD; TaP = FNMS(KP831469612, TaO, KP555570233 * TaH); Tb8 = FMA(KP831469612, Tb0, KP555570233 * Tb7); Tb9 = TaP - Tb8; Tdh = Tb8 + TaP; Tba = Taw + Tb9; Tdw = Tdg - Tdh; TcS = Taw - Tb9; Tdi = Tdg + Tdh; } { E TcE, Td4, TcH, Td5, TcF, TcG; TcE = TcA - TcD; Td4 = Tag + Tav; TcF = FNMS(KP831469612, Tb7, KP555570233 * Tb0); TcG = FMA(KP555570233, TaO, KP831469612 * TaH); TcH = TcF - TcG; Td5 = TcF + TcG; TcI = TcE + TcH; Tds = Td4 - Td5; TcW = TcE - TcH; Td6 = Td4 + Td5; } { E TbN, TcJ, Tcq, TcK; { E TbB, TbM, Tce, Tcp; TbB = Tbl - TbA; TbM = TbI - TbL; TbN = FNMS(KP956940335, TbM, KP290284677 * TbB); TcJ = FMA(KP956940335, TbB, KP290284677 * TbM); Tce = TbY - Tcd; Tcp = Tcl - Tco; Tcq = FMA(KP290284677, Tce, KP956940335 * Tcp); TcK = FNMS(KP956940335, Tce, KP290284677 * Tcp); } Tcr = TbN + Tcq; TcX = TbN - Tcq; TcL = TcJ + TcK; TcT = TcK - TcJ; } { E Td9, Tdj, Tdc, Tdk; { E Td7, Td8, Tda, Tdb; Td7 = Tbl + TbA; Td8 = TbI + TbL; Td9 = FNMS(KP471396736, Td8, KP881921264 * Td7); Tdj = FMA(KP471396736, Td7, KP881921264 * Td8); Tda = TbY + Tcd; Tdb = Tcl + Tco; Tdc = FMA(KP881921264, Tda, KP471396736 * Tdb); Tdk = FNMS(KP471396736, Tda, KP881921264 * Tdb); } Tdd = Td9 + Tdc; Tdx = Td9 - Tdc; Tdl = Tdj + Tdk; Tdt = Tdk - Tdj; } { E Tcs, TcM, Ta5, Tct; Tcs = Tba - Tcr; TcM = TcI - TcL; Ta5 = W[88]; Tct = W[89]; cr[WS(rs, 45)] = FNMS(Tct, TcM, Ta5 * Tcs); ci[WS(rs, 45)] = FMA(Tct, Tcs, Ta5 * TcM); } { E Tdu, Tdy, Tdr, Tdv; Tdu = Tds - Tdt; Tdy = Tdw - Tdx; Tdr = W[104]; Tdv = W[105]; cr[WS(rs, 53)] = FNMS(Tdv, Tdy, Tdr * Tdu); ci[WS(rs, 53)] = FMA(Tdr, Tdy, Tdv * Tdu); } { E TdA, TdC, Tdz, TdB; TdA = Tds + Tdt; TdC = Tdw + Tdx; Tdz = W[40]; TdB = W[41]; cr[WS(rs, 21)] = FNMS(TdB, TdC, Tdz * TdA); ci[WS(rs, 21)] = FMA(Tdz, TdC, TdB * TdA); } { E TcO, TcQ, TcN, TcP; TcO = Tba + Tcr; TcQ = TcI + TcL; TcN = W[24]; TcP = W[25]; cr[WS(rs, 13)] = FNMS(TcP, TcQ, TcN * TcO); ci[WS(rs, 13)] = FMA(TcP, TcO, TcN * TcQ); } { E TcU, TcY, TcR, TcV; TcU = TcS - TcT; TcY = TcW - TcX; TcR = W[120]; TcV = W[121]; cr[WS(rs, 61)] = FNMS(TcV, TcY, TcR * TcU); ci[WS(rs, 61)] = FMA(TcR, TcY, TcV * TcU); } { E Tde, Tdm, Td3, Tdf; Tde = Td6 - Tdd; Tdm = Tdi - Tdl; Td3 = W[72]; Tdf = W[73]; cr[WS(rs, 37)] = FNMS(Tdf, Tdm, Td3 * Tde); ci[WS(rs, 37)] = FMA(Tdf, Tde, Td3 * Tdm); } { E Tdo, Tdq, Tdn, Tdp; Tdo = Td6 + Tdd; Tdq = Tdi + Tdl; Tdn = W[8]; Tdp = W[9]; cr[WS(rs, 5)] = FNMS(Tdp, Tdq, Tdn * Tdo); ci[WS(rs, 5)] = FMA(Tdp, Tdo, Tdn * Tdq); } { E Td0, Td2, TcZ, Td1; Td0 = TcS + TcT; Td2 = TcW + TcX; TcZ = W[56]; Td1 = W[57]; cr[WS(rs, 29)] = FNMS(Td1, Td2, TcZ * Td0); ci[WS(rs, 29)] = FMA(TcZ, Td2, Td1 * Td0); } } { E Tfy, Thc, Tgy, TgY, Tgo, Th8, TgC, TgM, Tgb, TgD, Tgr, Tgz, TgT, Thd, Th1; E Th9; { E Tfi, TgW, Tfx, TgX, Tfp, Tfw; Tfi = Tfa - Tfh; TgW = Tgg + Tgj; Tfp = FNMS(KP555570233, Tfo, KP831469612 * Tfl); Tfw = FMA(KP831469612, Tfs, KP555570233 * Tfv); Tfx = Tfp - Tfw; TgX = Tfw + Tfp; Tfy = Tfi + Tfx; Thc = TgW - TgX; Tgy = Tfi - Tfx; TgY = TgW + TgX; } { E Tgk, TgK, Tgn, TgL, Tgl, Tgm; Tgk = Tgg - Tgj; TgK = Tfa + Tfh; Tgl = FNMS(KP555570233, Tfs, KP831469612 * Tfv); Tgm = FMA(KP555570233, Tfl, KP831469612 * Tfo); Tgn = Tgl - Tgm; TgL = Tgl + Tgm; Tgo = Tgk + Tgn; Th8 = TgK - TgL; TgC = Tgk - Tgn; TgM = TgK + TgL; } { E TfR, Tgp, Tga, Tgq; { E TfJ, TfQ, Tg2, Tg9; TfJ = TfB - TfI; TfQ = TfM - TfP; TfR = FNMS(KP881921264, TfQ, KP471396736 * TfJ); Tgp = FMA(KP881921264, TfJ, KP471396736 * TfQ); Tg2 = TfU - Tg1; Tg9 = Tg5 - Tg8; Tga = FMA(KP471396736, Tg2, KP881921264 * Tg9); Tgq = FNMS(KP881921264, Tg2, KP471396736 * Tg9); } Tgb = TfR + Tga; TgD = TfR - Tga; Tgr = Tgp + Tgq; Tgz = Tgq - Tgp; } { E TgP, TgZ, TgS, Th0; { E TgN, TgO, TgQ, TgR; TgN = TfB + TfI; TgO = TfM + TfP; TgP = FNMS(KP290284677, TgO, KP956940335 * TgN); TgZ = FMA(KP290284677, TgN, KP956940335 * TgO); TgQ = TfU + Tg1; TgR = Tg5 + Tg8; TgS = FMA(KP956940335, TgQ, KP290284677 * TgR); Th0 = FNMS(KP290284677, TgQ, KP956940335 * TgR); } TgT = TgP + TgS; Thd = TgP - TgS; Th1 = TgZ + Th0; Th9 = Th0 - TgZ; } { E Tgc, Tgs, Tf7, Tgd; Tgc = Tfy - Tgb; Tgs = Tgo - Tgr; Tf7 = W[84]; Tgd = W[85]; cr[WS(rs, 43)] = FNMS(Tgd, Tgs, Tf7 * Tgc); ci[WS(rs, 43)] = FMA(Tgd, Tgc, Tf7 * Tgs); } { E Tha, The, Th7, Thb; Tha = Th8 - Th9; The = Thc - Thd; Th7 = W[100]; Thb = W[101]; cr[WS(rs, 51)] = FNMS(Thb, The, Th7 * Tha); ci[WS(rs, 51)] = FMA(Th7, The, Thb * Tha); } { E Thg, Thi, Thf, Thh; Thg = Th8 + Th9; Thi = Thc + Thd; Thf = W[36]; Thh = W[37]; cr[WS(rs, 19)] = FNMS(Thh, Thi, Thf * Thg); ci[WS(rs, 19)] = FMA(Thf, Thi, Thh * Thg); } { E Tgu, Tgw, Tgt, Tgv; Tgu = Tfy + Tgb; Tgw = Tgo + Tgr; Tgt = W[20]; Tgv = W[21]; cr[WS(rs, 11)] = FNMS(Tgv, Tgw, Tgt * Tgu); ci[WS(rs, 11)] = FMA(Tgv, Tgu, Tgt * Tgw); } { E TgA, TgE, Tgx, TgB; TgA = Tgy - Tgz; TgE = TgC - TgD; Tgx = W[116]; TgB = W[117]; cr[WS(rs, 59)] = FNMS(TgB, TgE, Tgx * TgA); ci[WS(rs, 59)] = FMA(Tgx, TgE, TgB * TgA); } { E TgU, Th2, TgJ, TgV; TgU = TgM - TgT; Th2 = TgY - Th1; TgJ = W[68]; TgV = W[69]; cr[WS(rs, 35)] = FNMS(TgV, Th2, TgJ * TgU); ci[WS(rs, 35)] = FMA(TgV, TgU, TgJ * Th2); } { E Th4, Th6, Th3, Th5; Th4 = TgM + TgT; Th6 = TgY + Th1; Th3 = W[4]; Th5 = W[5]; cr[WS(rs, 3)] = FNMS(Th5, Th6, Th3 * Th4); ci[WS(rs, 3)] = FMA(Th5, Th4, Th3 * Th6); } { E TgG, TgI, TgF, TgH; TgG = Tgy + Tgz; TgI = TgC + TgD; TgF = W[52]; TgH = W[53]; cr[WS(rs, 27)] = FNMS(TgH, TgI, TgF * TgG); ci[WS(rs, 27)] = FMA(TgF, TgI, TgH * TgG); } } { E TdO, Tf0, Tem, TeM, Tec, TeW, Teq, TeA, Te3, Ter, Tef, Ten, TeH, Tf1, TeP; E TeX; { E TdG, TeK, TdN, TeL, TdJ, TdM; TdG = TdE - TdF; TeK = Te6 + Te7; TdJ = FNMS(KP195090322, TdI, KP980785280 * TdH); TdM = FMA(KP195090322, TdK, KP980785280 * TdL); TdN = TdJ - TdM; TeL = TdM + TdJ; TdO = TdG - TdN; Tf0 = TeK + TeL; Tem = TdG + TdN; TeM = TeK - TeL; } { E Te8, Tey, Teb, Tez, Te9, Tea; Te8 = Te6 - Te7; Tey = TdE + TdF; Te9 = FNMS(KP195090322, TdL, KP980785280 * TdK); Tea = FMA(KP980785280, TdI, KP195090322 * TdH); Teb = Te9 - Tea; Tez = Te9 + Tea; Tec = Te8 - Teb; TeW = Tey + Tez; Teq = Te8 + Teb; TeA = Tey - Tez; } { E TdV, Tee, Te2, Ted; { E TdR, TdU, TdY, Te1; TdR = TdP - TdQ; TdU = TdS - TdT; TdV = FNMS(KP773010453, TdU, KP634393284 * TdR); Tee = FMA(KP773010453, TdR, KP634393284 * TdU); TdY = TdW - TdX; Te1 = TdZ - Te0; Te2 = FMA(KP634393284, TdY, KP773010453 * Te1); Ted = FNMS(KP773010453, TdY, KP634393284 * Te1); } Te3 = TdV - Te2; Ter = Te2 + TdV; Tef = Ted - Tee; Ten = Ted + Tee; } { E TeD, TeO, TeG, TeN; { E TeB, TeC, TeE, TeF; TeB = TdP + TdQ; TeC = TdS + TdT; TeD = FNMS(KP098017140, TeC, KP995184726 * TeB); TeO = FMA(KP098017140, TeB, KP995184726 * TeC); TeE = TdW + TdX; TeF = TdZ + Te0; TeG = FMA(KP995184726, TeE, KP098017140 * TeF); TeN = FNMS(KP098017140, TeE, KP995184726 * TeF); } TeH = TeD - TeG; Tf1 = TeG + TeD; TeP = TeN - TeO; TeX = TeN + TeO; } { E Te4, Teg, TdD, Te5; Te4 = TdO - Te3; Teg = Tec - Tef; TdD = W[112]; Te5 = W[113]; cr[WS(rs, 57)] = FNMS(Te5, Teg, TdD * Te4); ci[WS(rs, 57)] = FMA(Te5, Te4, TdD * Teg); } { E TeY, Tf2, TeV, TeZ; TeY = TeW - TeX; Tf2 = Tf0 - Tf1; TeV = W[64]; TeZ = W[65]; cr[WS(rs, 33)] = FNMS(TeZ, Tf2, TeV * TeY); ci[WS(rs, 33)] = FMA(TeV, Tf2, TeZ * TeY); } { E Tf4, Tf6, Tf3, Tf5; Tf4 = TeW + TeX; Tf6 = Tf0 + Tf1; Tf3 = W[0]; Tf5 = W[1]; cr[WS(rs, 1)] = FNMS(Tf5, Tf6, Tf3 * Tf4); ci[WS(rs, 1)] = FMA(Tf3, Tf6, Tf5 * Tf4); } { E Tei, Tek, Teh, Tej; Tei = TdO + Te3; Tek = Tec + Tef; Teh = W[48]; Tej = W[49]; cr[WS(rs, 25)] = FNMS(Tej, Tek, Teh * Tei); ci[WS(rs, 25)] = FMA(Tej, Tei, Teh * Tek); } { E Teo, Tes, Tel, Tep; Teo = Tem - Ten; Tes = Teq - Ter; Tel = W[80]; Tep = W[81]; cr[WS(rs, 41)] = FNMS(Tep, Tes, Tel * Teo); ci[WS(rs, 41)] = FMA(Tel, Tes, Tep * Teo); } { E TeI, TeQ, Tex, TeJ; TeI = TeA - TeH; TeQ = TeM - TeP; Tex = W[96]; TeJ = W[97]; cr[WS(rs, 49)] = FNMS(TeJ, TeQ, Tex * TeI); ci[WS(rs, 49)] = FMA(TeJ, TeI, Tex * TeQ); } { E TeS, TeU, TeR, TeT; TeS = TeA + TeH; TeU = TeM + TeP; TeR = W[32]; TeT = W[33]; cr[WS(rs, 17)] = FNMS(TeT, TeU, TeR * TeS); ci[WS(rs, 17)] = FMA(TeT, TeS, TeR * TeU); } { E Teu, Tew, Tet, Tev; Teu = Tem + Ten; Tew = Teq + Ter; Tet = W[16]; Tev = W[17]; cr[WS(rs, 9)] = FNMS(Tev, Tew, Tet * Teu); ci[WS(rs, 9)] = FMA(Tet, Tew, Tev * Teu); } } { E Thu, TiG, Ti2, Tis, ThS, TiC, Ti6, Tig, ThJ, Ti7, ThV, Ti3, Tin, TiH, Tiv; E TiD; { E Thm, Tiq, Tht, Tir, Thp, Ths; Thm = Thk - Thl; Tiq = ThM - ThN; Thp = FNMS(KP980785280, Tho, KP195090322 * Thn); Ths = FNMS(KP980785280, Thr, KP195090322 * Thq); Tht = Thp + Ths; Tir = Thp - Ths; Thu = Thm - Tht; TiG = Tiq - Tir; Ti2 = Thm + Tht; Tis = Tiq + Tir; } { E ThO, Tie, ThR, Tif, ThP, ThQ; ThO = ThM + ThN; Tie = Thk + Thl; ThP = FMA(KP195090322, Tho, KP980785280 * Thn); ThQ = FMA(KP195090322, Thr, KP980785280 * Thq); ThR = ThP - ThQ; Tif = ThP + ThQ; ThS = ThO - ThR; TiC = Tie + Tif; Ti6 = ThO + ThR; Tig = Tie - Tif; } { E ThB, ThU, ThI, ThT; { E Thx, ThA, ThE, ThH; Thx = Thv - Thw; ThA = Thy - Thz; ThB = FNMS(KP634393284, ThA, KP773010453 * Thx); ThU = FMA(KP634393284, Thx, KP773010453 * ThA); ThE = ThC + ThD; ThH = ThF - ThG; ThI = FMA(KP773010453, ThE, KP634393284 * ThH); ThT = FNMS(KP634393284, ThE, KP773010453 * ThH); } ThJ = ThB - ThI; Ti7 = ThI + ThB; ThV = ThT - ThU; Ti3 = ThT + ThU; } { E Tij, Tit, Tim, Tiu; { E Tih, Tii, Tik, Til; Tih = ThF + ThG; Tii = ThC - ThD; Tij = FNMS(KP995184726, Tii, KP098017140 * Tih); Tit = FMA(KP098017140, Tii, KP995184726 * Tih); Tik = Thy + Thz; Til = Thw + Thv; Tim = FNMS(KP995184726, Til, KP098017140 * Tik); Tiu = FMA(KP098017140, Til, KP995184726 * Tik); } Tin = Tij + Tim; TiH = Tij - Tim; Tiv = Tit - Tiu; TiD = Tit + Tiu; } { E ThK, ThW, Thj, ThL; ThK = Thu - ThJ; ThW = ThS - ThV; Thj = W[108]; ThL = W[109]; cr[WS(rs, 55)] = FNMS(ThL, ThW, Thj * ThK); ci[WS(rs, 55)] = FMA(ThL, ThK, Thj * ThW); } { E TiE, TiI, TiB, TiF; TiE = TiC - TiD; TiI = TiG + TiH; TiB = W[60]; TiF = W[61]; cr[WS(rs, 31)] = FNMS(TiF, TiI, TiB * TiE); ci[WS(rs, 31)] = FMA(TiB, TiI, TiF * TiE); } { E TiK, TiM, TiJ, TiL; TiK = TiC + TiD; TiM = TiG - TiH; TiJ = W[124]; TiL = W[125]; cr[WS(rs, 63)] = FNMS(TiL, TiM, TiJ * TiK); ci[WS(rs, 63)] = FMA(TiJ, TiM, TiL * TiK); } { E ThY, Ti0, ThX, ThZ; ThY = Thu + ThJ; Ti0 = ThS + ThV; ThX = W[44]; ThZ = W[45]; cr[WS(rs, 23)] = FNMS(ThZ, Ti0, ThX * ThY); ci[WS(rs, 23)] = FMA(ThZ, ThY, ThX * Ti0); } { E Ti4, Ti8, Ti1, Ti5; Ti4 = Ti2 - Ti3; Ti8 = Ti6 - Ti7; Ti1 = W[76]; Ti5 = W[77]; cr[WS(rs, 39)] = FNMS(Ti5, Ti8, Ti1 * Ti4); ci[WS(rs, 39)] = FMA(Ti1, Ti8, Ti5 * Ti4); } { E Tio, Tiw, Tid, Tip; Tio = Tig - Tin; Tiw = Tis - Tiv; Tid = W[92]; Tip = W[93]; cr[WS(rs, 47)] = FNMS(Tip, Tiw, Tid * Tio); ci[WS(rs, 47)] = FMA(Tip, Tio, Tid * Tiw); } { E Tiy, TiA, Tix, Tiz; Tiy = Tig + Tin; TiA = Tis + Tiv; Tix = W[28]; Tiz = W[29]; cr[WS(rs, 15)] = FNMS(Tiz, TiA, Tix * Tiy); ci[WS(rs, 15)] = FMA(Tiz, Tiy, Tix * TiA); } { E Tia, Tic, Ti9, Tib; Tia = Ti2 + Ti3; Tic = Ti6 + Ti7; Ti9 = W[12]; Tib = W[13]; cr[WS(rs, 7)] = FNMS(Tib, Tic, Ti9 * Tia); ci[WS(rs, 7)] = FMA(Ti9, Tic, Tib * Tia); } } } } } static const tw_instr twinstr[] = { {TW_FULL, 1, 64}, {TW_NEXT, 1, 0} }; static const hc2hc_desc desc = { 64, "hb_64", twinstr, &GENUS, {808, 270, 230, 0} }; void X(codelet_hb_64) (planner *p) { X(khc2hc_register) (p, hb_64, &desc); } #endif /* HAVE_FMA */ fftw-3.3.3/rdft/scalar/r2cb/hb_20.c0000644000175000017500000006733512054410763013521 00000000000000/* * Copyright (c) 2003, 2007-11 Matteo Frigo * Copyright (c) 2003, 2007-11 Massachusetts Institute of Technology * * This program is free software; you can redistribute it and/or modify * it under the terms of the GNU General Public License as published by * the Free Software Foundation; either version 2 of the License, or * (at your option) any later version. * * This program is distributed in the hope that it will be useful, * but WITHOUT ANY WARRANTY; without even the implied warranty of * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the * GNU General Public License for more details. * * You should have received a copy of the GNU General Public License * along with this program; if not, write to the Free Software * Foundation, Inc., 51 Franklin Street, Fifth Floor, Boston, MA 02110-1301 USA * */ /* This file was automatically generated --- DO NOT EDIT */ /* Generated on Sun Nov 25 07:41:22 EST 2012 */ #include "codelet-rdft.h" #ifdef HAVE_FMA /* Generated by: ../../../genfft/gen_hc2hc.native -fma -reorder-insns -schedule-for-pipeline -compact -variables 4 -pipeline-latency 4 -sign 1 -n 20 -dif -name hb_20 -include hb.h */ /* * This function contains 246 FP additions, 148 FP multiplications, * (or, 136 additions, 38 multiplications, 110 fused multiply/add), * 101 stack variables, 4 constants, and 80 memory accesses */ #include "hb.h" static void hb_20(R *cr, R *ci, const R *W, stride rs, INT mb, INT me, INT ms) { DK(KP951056516, +0.951056516295153572116439333379382143405698634); DK(KP559016994, +0.559016994374947424102293417182819058860154590); DK(KP618033988, +0.618033988749894848204586834365638117720309180); DK(KP250000000, +0.250000000000000000000000000000000000000000000); { INT m; for (m = mb, W = W + ((mb - 1) * 38); m < me; m = m + 1, cr = cr + ms, ci = ci - ms, W = W + 38, MAKE_VOLATILE_STRIDE(40, rs)) { E T1T, T1Q, T1P; { E T2W, T4e, T7, TE, T3z, T4z, T1t, T2l, T3a, T3G, T13, T33, T3H, T1i, T2g; E T4H, T4G, T2d, T1B, T4u, T4B, T4A, T4r, T1A, T2s, T3l, T2t, T3s, T2o, T2q; E T1w, T1y, TC, T29, T3E, T3C, T4n, T4l, TN, TL; { E T4, T2U, T3, T2V, T1s, T5, T1n, T1o; { E T1, T2, T1q, T1r; T1 = cr[0]; T2 = ci[WS(rs, 9)]; T1q = ci[WS(rs, 14)]; T1r = cr[WS(rs, 15)]; T4 = cr[WS(rs, 5)]; T2U = T1 - T2; T3 = T1 + T2; T2V = T1q + T1r; T1s = T1q - T1r; T5 = ci[WS(rs, 4)]; T1n = ci[WS(rs, 19)]; T1o = cr[WS(rs, 10)]; } { E T3y, T6, T3x, T1p; T2W = T2U + T2V; T4e = T2U - T2V; T3y = T4 - T5; T6 = T4 + T5; T3x = T1n + T1o; T1p = T1n - T1o; T7 = T3 + T6; TE = T3 - T6; T3z = T3x - T3y; T4z = T3y + T3x; T1t = T1p - T1s; T2l = T1p + T1s; } } { E T2Z, T4f, Te, TF, T3o, T4p, T1a, T2b, TJ, TA, T4t, T3k, T4j, T39, T2f; E T12, T32, T4g, Tl, TG, T3r, T4q, T1h, T2c, T36, T4i, Tt, TI, T3h, T4s; E TV, T2e; { E Tb, T2X, Ta, T2Y, T19, Tc, T14, T15; { E T8, T9, T17, T18; T8 = cr[WS(rs, 4)]; T9 = ci[WS(rs, 5)]; T17 = ci[WS(rs, 10)]; T18 = cr[WS(rs, 19)]; Tb = cr[WS(rs, 9)]; T2X = T8 - T9; Ta = T8 + T9; T2Y = T17 + T18; T19 = T17 - T18; Tc = ci[0]; T14 = ci[WS(rs, 15)]; T15 = cr[WS(rs, 14)]; } { E T3n, Td, T3m, T16; T2Z = T2X + T2Y; T4f = T2X - T2Y; T3n = Tb - Tc; Td = Tb + Tc; T3m = T14 + T15; T16 = T14 - T15; Te = Ta + Td; TF = Ta - Td; T3o = T3m - T3n; T4p = T3n + T3m; T1a = T16 - T19; T2b = T16 + T19; } } { E TW, T37, Tw, T3i, Tz, TX, TZ, T10; { E Tu, Tv, Tx, Ty; Tu = ci[WS(rs, 7)]; Tv = cr[WS(rs, 2)]; Tx = ci[WS(rs, 2)]; Ty = cr[WS(rs, 7)]; TW = ci[WS(rs, 17)]; T37 = Tu - Tv; Tw = Tu + Tv; T3i = Tx - Ty; Tz = Tx + Ty; TX = cr[WS(rs, 12)]; TZ = ci[WS(rs, 12)]; T10 = cr[WS(rs, 17)]; } { E TY, T38, T11, T3j; TJ = Tw - Tz; TA = Tw + Tz; T3j = TW + TX; TY = TW - TX; T38 = TZ + T10; T11 = TZ - T10; T4t = T3i - T3j; T3k = T3i + T3j; T4j = T37 + T38; T39 = T37 - T38; T2f = TY + T11; T12 = TY - T11; } } { E Ti, T30, Th, T31, T1g, Tj, T1b, T1c; { E Tf, Tg, T1e, T1f; Tf = ci[WS(rs, 3)]; Tg = cr[WS(rs, 6)]; T1e = ci[WS(rs, 18)]; T1f = cr[WS(rs, 11)]; Ti = cr[WS(rs, 1)]; T30 = Tf - Tg; Th = Tf + Tg; T31 = T1e + T1f; T1g = T1e - T1f; Tj = ci[WS(rs, 8)]; T1b = ci[WS(rs, 13)]; T1c = cr[WS(rs, 16)]; } { E T3p, Tk, T3q, T1d; T32 = T30 + T31; T4g = T30 - T31; T3p = Ti - Tj; Tk = Ti + Tj; T3q = T1b + T1c; T1d = T1b - T1c; Tl = Th + Tk; TG = Th - Tk; T3r = T3p + T3q; T4q = T3p - T3q; T1h = T1d - T1g; T2c = T1d + T1g; } } { E Tq, T34, Tp, T35, TU, Tr, TP, TQ; { E Tn, To, TS, TT; Tn = cr[WS(rs, 8)]; To = ci[WS(rs, 1)]; TS = ci[WS(rs, 16)]; TT = cr[WS(rs, 13)]; Tq = ci[WS(rs, 6)]; T34 = Tn - To; Tp = Tn + To; T35 = TS + TT; TU = TS - TT; Tr = cr[WS(rs, 3)]; TP = ci[WS(rs, 11)]; TQ = cr[WS(rs, 18)]; } { E T3g, Ts, T3f, TR; T36 = T34 - T35; T4i = T34 + T35; T3g = Tq - Tr; Ts = Tq + Tr; T3f = TP + TQ; TR = TP - TQ; Tt = Tp + Ts; TI = Tp - Ts; T3h = T3f - T3g; T4s = T3g + T3f; TV = TR - TU; T2e = TR + TU; } } { E T1v, T1u, T2n, T4k, T4h, T2m, TH, TK; T3a = T36 + T39; T3G = T36 - T39; T13 = TV - T12; T1v = TV + T12; T33 = T2Z + T32; T3H = T2Z - T32; T1i = T1a - T1h; T1u = T1a + T1h; T2n = T2e + T2f; T2g = T2e - T2f; T4H = T4i - T4j; T4k = T4i + T4j; T4h = T4f + T4g; T4G = T4f - T4g; T2d = T2b - T2c; T2m = T2b + T2c; TH = TF + TG; T1B = TF - TG; T4u = T4s - T4t; T4B = T4s + T4t; T4A = T4p + T4q; T4r = T4p - T4q; T1A = TI - TJ; TK = TI + TJ; { E Tm, T3B, TB, T3A; Tm = Te + Tl; T2s = Te - Tl; T3l = T3h + T3k; T3B = T3h - T3k; TB = Tt + TA; T2t = Tt - TA; T3s = T3o + T3r; T3A = T3o - T3r; T2o = T2m + T2n; T2q = T2m - T2n; T1w = T1u + T1v; T1y = T1u - T1v; TC = Tm + TB; T29 = Tm - TB; T3E = T3A - T3B; T3C = T3A + T3B; T4n = T4h - T4k; T4l = T4h + T4k; TN = TH - TK; TL = TH + TK; } } } { E T3d, T3b, T4E, T1x, TM, T4m, T58, T5b, T4D, T5a, T5c, T59, T4C; cr[0] = T7 + TC; T3d = T33 - T3a; T3b = T33 + T3a; T4E = T4A - T4B; T4C = T4A + T4B; ci[0] = T2l + T2o; { E T25, T22, T21, T24, T23, T26, T57; T1x = FNMS(KP250000000, T1w, T1t); T25 = T1t + T1w; T22 = TE + TL; TM = FNMS(KP250000000, TL, TE); T21 = W[18]; T24 = W[19]; T4m = FNMS(KP250000000, T4l, T4e); T58 = T4e + T4l; T5b = T4z + T4C; T4D = FNMS(KP250000000, T4C, T4z); T23 = T21 * T22; T26 = T24 * T22; T57 = W[8]; T5a = W[9]; cr[WS(rs, 10)] = FNMS(T24, T25, T23); ci[WS(rs, 10)] = FMA(T21, T25, T26); T5c = T57 * T5b; T59 = T57 * T58; } { E T3U, T3Z, T3W, T40, T3V; { E T3c, T48, T4b, T3D, T47, T4a; T3c = FNMS(KP250000000, T3b, T2W); T48 = T2W + T3b; T4b = T3z + T3C; T3D = FNMS(KP250000000, T3C, T3z); ci[WS(rs, 5)] = FMA(T5a, T58, T5c); cr[WS(rs, 5)] = FNMS(T5a, T5b, T59); T47 = W[28]; T4a = W[29]; { E T3I, T3Y, T42, T3u, T3M, T3X, T3F; { E T3T, T3t, T4c, T49, T3e, T3S; T3T = FMA(KP618033988, T3l, T3s); T3t = FNMS(KP618033988, T3s, T3l); T4c = T47 * T4b; T49 = T47 * T48; T3I = FNMS(KP618033988, T3H, T3G); T3Y = FMA(KP618033988, T3G, T3H); ci[WS(rs, 15)] = FMA(T4a, T48, T4c); cr[WS(rs, 15)] = FNMS(T4a, T4b, T49); T3e = FNMS(KP559016994, T3d, T3c); T3S = FMA(KP559016994, T3d, T3c); T42 = FMA(KP951056516, T3T, T3S); T3U = FNMS(KP951056516, T3T, T3S); T3u = FNMS(KP951056516, T3t, T3e); T3M = FMA(KP951056516, T3t, T3e); T3X = FMA(KP559016994, T3E, T3D); T3F = FNMS(KP559016994, T3E, T3D); } { E T3P, T45, T44, T46, T43; { E T3w, T3J, T3v, T3K, T2T, T41; T2T = W[4]; T3w = W[5]; T3J = FMA(KP951056516, T3I, T3F); T3P = FNMS(KP951056516, T3I, T3F); T45 = FNMS(KP951056516, T3Y, T3X); T3Z = FMA(KP951056516, T3Y, T3X); T3v = T2T * T3u; T3K = T2T * T3J; T41 = W[36]; T44 = W[37]; cr[WS(rs, 3)] = FNMS(T3w, T3J, T3v); ci[WS(rs, 3)] = FMA(T3w, T3u, T3K); T46 = T41 * T45; T43 = T41 * T42; } { E T3O, T3Q, T3N, T3L, T3R; T3L = W[12]; T3O = W[13]; ci[WS(rs, 19)] = FMA(T44, T42, T46); cr[WS(rs, 19)] = FNMS(T44, T45, T43); T3Q = T3L * T3P; T3N = T3L * T3M; T3R = W[20]; T3W = W[21]; ci[WS(rs, 7)] = FMA(T3O, T3M, T3Q); cr[WS(rs, 7)] = FNMS(T3O, T3P, T3N); T40 = T3R * T3Z; T3V = T3R * T3U; } } } } { E T4U, T4Z, T4W, T50, T4V, T2L, T2I, T2H; { E T4T, T4v, T4I, T4Y, T4o, T4S; T4T = FNMS(KP618033988, T4r, T4u); T4v = FMA(KP618033988, T4u, T4r); ci[WS(rs, 11)] = FMA(T3W, T3U, T40); cr[WS(rs, 11)] = FNMS(T3W, T3Z, T3V); T4I = FMA(KP618033988, T4H, T4G); T4Y = FNMS(KP618033988, T4G, T4H); T4o = FMA(KP559016994, T4n, T4m); T4S = FNMS(KP559016994, T4n, T4m); { E T52, T4M, T55, T4P, T54, T56, T53; { E T4d, T4w, T4J, T4x, T4y, T4X, T4F, T51, T4K; T4d = W[0]; T4X = FNMS(KP559016994, T4E, T4D); T4F = FMA(KP559016994, T4E, T4D); T4U = FNMS(KP951056516, T4T, T4S); T52 = FMA(KP951056516, T4T, T4S); T4M = FMA(KP951056516, T4v, T4o); T4w = FNMS(KP951056516, T4v, T4o); T4Z = FMA(KP951056516, T4Y, T4X); T55 = FNMS(KP951056516, T4Y, T4X); T4P = FNMS(KP951056516, T4I, T4F); T4J = FMA(KP951056516, T4I, T4F); T4x = T4d * T4w; T4y = W[1]; T51 = W[32]; T4K = T4d * T4J; T54 = W[33]; cr[WS(rs, 1)] = FNMS(T4y, T4J, T4x); T56 = T51 * T55; T53 = T51 * T52; ci[WS(rs, 1)] = FMA(T4y, T4w, T4K); } { E T4O, T4Q, T4N, T4L, T4R; T4L = W[16]; ci[WS(rs, 17)] = FMA(T54, T52, T56); cr[WS(rs, 17)] = FNMS(T54, T55, T53); T4O = W[17]; T4Q = T4L * T4P; T4N = T4L * T4M; T4R = W[24]; T4W = W[25]; ci[WS(rs, 9)] = FMA(T4O, T4M, T4Q); cr[WS(rs, 9)] = FNMS(T4O, T4P, T4N); T50 = T4R * T4Z; T4V = T4R * T4U; } } } { E T2K, T2u, T2F, T2h, T28, T2J, T2r, T2p; T2K = FNMS(KP618033988, T2s, T2t); T2u = FMA(KP618033988, T2t, T2s); ci[WS(rs, 13)] = FMA(T4W, T4U, T50); cr[WS(rs, 13)] = FNMS(T4W, T4Z, T4V); T2p = FNMS(KP250000000, T2o, T2l); T2F = FNMS(KP618033988, T2d, T2g); T2h = FMA(KP618033988, T2g, T2d); T28 = FNMS(KP250000000, TC, T7); T2J = FNMS(KP559016994, T2q, T2p); T2r = FMA(KP559016994, T2q, T2p); { E T2B, T2G, T2y, T2R, T2Q, T2P, T2A, T2x; { E T2k, T2v, T27, T2O, T2i, T2a, T2E; T2k = W[7]; T2a = FMA(KP559016994, T29, T28); T2E = FNMS(KP559016994, T29, T28); T2B = FMA(KP951056516, T2u, T2r); T2v = FNMS(KP951056516, T2u, T2r); T27 = W[6]; T2O = FMA(KP951056516, T2F, T2E); T2G = FNMS(KP951056516, T2F, T2E); T2i = FMA(KP951056516, T2h, T2a); T2y = FNMS(KP951056516, T2h, T2a); { E T2N, T2j, T2w, T2S; T2L = FMA(KP951056516, T2K, T2J); T2R = FNMS(KP951056516, T2K, T2J); T2Q = W[23]; T2N = W[22]; T2j = T27 * T2i; T2w = T2k * T2i; T2S = T2Q * T2O; T2P = T2N * T2O; cr[WS(rs, 4)] = FNMS(T2k, T2v, T2j); ci[WS(rs, 4)] = FMA(T27, T2v, T2w); ci[WS(rs, 12)] = FMA(T2N, T2R, T2S); } } cr[WS(rs, 12)] = FNMS(T2Q, T2R, T2P); T2A = W[31]; T2x = W[30]; { E T2D, T2M, T2C, T2z; T2I = W[15]; T2C = T2A * T2y; T2z = T2x * T2y; T2D = W[14]; T2M = T2I * T2G; ci[WS(rs, 16)] = FMA(T2x, T2B, T2C); cr[WS(rs, 16)] = FNMS(T2A, T2B, T2z); T2H = T2D * T2G; ci[WS(rs, 8)] = FMA(T2D, T2L, T2M); } } } { E T1S, T1C, T1j, T1N, T1z, T1R; T1S = FMA(KP618033988, T1A, T1B); T1C = FNMS(KP618033988, T1B, T1A); cr[WS(rs, 8)] = FNMS(T2I, T2L, T2H); T1j = FNMS(KP618033988, T1i, T13); T1N = FMA(KP618033988, T13, T1i); T1z = FNMS(KP559016994, T1y, T1x); T1R = FMA(KP559016994, T1y, T1x); { E T1J, T1O, T1G, T1Z, T1Y, T1X, T1I, T1F; { E T1m, T1D, TD, T1W, T1k, T1M, TO; T1m = W[3]; T1M = FMA(KP559016994, TN, TM); TO = FNMS(KP559016994, TN, TM); T1D = FNMS(KP951056516, T1C, T1z); T1J = FMA(KP951056516, T1C, T1z); TD = W[2]; T1O = FNMS(KP951056516, T1N, T1M); T1W = FMA(KP951056516, T1N, T1M); T1G = FNMS(KP951056516, T1j, TO); T1k = FMA(KP951056516, T1j, TO); { E T1V, T1l, T1E, T20; T1Z = FNMS(KP951056516, T1S, T1R); T1T = FMA(KP951056516, T1S, T1R); T1Y = W[27]; T1V = W[26]; T1l = TD * T1k; T1E = T1m * T1k; T20 = T1Y * T1W; T1X = T1V * T1W; cr[WS(rs, 2)] = FNMS(T1m, T1D, T1l); ci[WS(rs, 2)] = FMA(TD, T1D, T1E); ci[WS(rs, 14)] = FMA(T1V, T1Z, T20); } } cr[WS(rs, 14)] = FNMS(T1Y, T1Z, T1X); T1I = W[35]; T1F = W[34]; { E T1L, T1U, T1K, T1H; T1Q = W[11]; T1K = T1I * T1G; T1H = T1F * T1G; T1L = W[10]; T1U = T1Q * T1O; ci[WS(rs, 18)] = FMA(T1F, T1J, T1K); cr[WS(rs, 18)] = FNMS(T1I, T1J, T1H); T1P = T1L * T1O; ci[WS(rs, 6)] = FMA(T1L, T1T, T1U); } } } } } } } cr[WS(rs, 6)] = FNMS(T1Q, T1T, T1P); } } } static const tw_instr twinstr[] = { {TW_FULL, 1, 20}, {TW_NEXT, 1, 0} }; static const hc2hc_desc desc = { 20, "hb_20", twinstr, &GENUS, {136, 38, 110, 0} }; void X(codelet_hb_20) (planner *p) { X(khc2hc_register) (p, hb_20, &desc); } #else /* HAVE_FMA */ /* Generated by: ../../../genfft/gen_hc2hc.native -compact -variables 4 -pipeline-latency 4 -sign 1 -n 20 -dif -name hb_20 -include hb.h */ /* * This function contains 246 FP additions, 124 FP multiplications, * (or, 184 additions, 62 multiplications, 62 fused multiply/add), * 97 stack variables, 4 constants, and 80 memory accesses */ #include "hb.h" static void hb_20(R *cr, R *ci, const R *W, stride rs, INT mb, INT me, INT ms) { DK(KP250000000, +0.250000000000000000000000000000000000000000000); DK(KP559016994, +0.559016994374947424102293417182819058860154590); DK(KP587785252, +0.587785252292473129168705954639072768597652438); DK(KP951056516, +0.951056516295153572116439333379382143405698634); { INT m; for (m = mb, W = W + ((mb - 1) * 38); m < me; m = m + 1, cr = cr + ms, ci = ci - ms, W = W + 38, MAKE_VOLATILE_STRIDE(40, rs)) { E T7, T3T, T49, TE, T1v, T2T, T3g, T2d, T13, T3n, T3o, T1i, T26, T4e, T4d; E T23, T1n, T42, T3Z, T1m, T2h, T2I, T2i, T2P, T30, T37, T38, Tm, TB, TC; E T46, T47, T4a, T2a, T2b, T2e, T1w, T1x, T1y, T3O, T3R, T3U, T3h, T3i, T3j; E TH, TK, TL; { E T3, T2R, T1u, T2S, T6, T3f, T1r, T3e; { E T1, T2, T1s, T1t; T1 = cr[0]; T2 = ci[WS(rs, 9)]; T3 = T1 + T2; T2R = T1 - T2; T1s = ci[WS(rs, 14)]; T1t = cr[WS(rs, 15)]; T1u = T1s - T1t; T2S = T1s + T1t; } { E T4, T5, T1p, T1q; T4 = cr[WS(rs, 5)]; T5 = ci[WS(rs, 4)]; T6 = T4 + T5; T3f = T4 - T5; T1p = ci[WS(rs, 19)]; T1q = cr[WS(rs, 10)]; T1r = T1p - T1q; T3e = T1p + T1q; } T7 = T3 + T6; T3T = T2R - T2S; T49 = T3f + T3e; TE = T3 - T6; T1v = T1r - T1u; T2T = T2R + T2S; T3g = T3e - T3f; T2d = T1r + T1u; } { E Te, T3M, T3X, TF, TV, T2E, T2W, T21, TA, T3Q, T41, TJ, T1h, T2O, T36; E T25, Tl, T3N, T3Y, TG, T12, T2H, T2Z, T22, Tt, T3P, T40, TI, T1a, T2L; E T33, T24; { E Ta, T2U, TU, T2V, Td, T2D, TR, T2C; { E T8, T9, TS, TT; T8 = cr[WS(rs, 4)]; T9 = ci[WS(rs, 5)]; Ta = T8 + T9; T2U = T8 - T9; TS = ci[WS(rs, 10)]; TT = cr[WS(rs, 19)]; TU = TS - TT; T2V = TS + TT; } { E Tb, Tc, TP, TQ; Tb = cr[WS(rs, 9)]; Tc = ci[0]; Td = Tb + Tc; T2D = Tb - Tc; TP = ci[WS(rs, 15)]; TQ = cr[WS(rs, 14)]; TR = TP - TQ; T2C = TP + TQ; } Te = Ta + Td; T3M = T2U - T2V; T3X = T2D + T2C; TF = Ta - Td; TV = TR - TU; T2E = T2C - T2D; T2W = T2U + T2V; T21 = TR + TU; } { E Tw, T34, Tz, T2M, T1d, T2N, T1g, T35; { E Tu, Tv, Tx, Ty; Tu = ci[WS(rs, 7)]; Tv = cr[WS(rs, 2)]; Tw = Tu + Tv; T34 = Tu - Tv; Tx = ci[WS(rs, 2)]; Ty = cr[WS(rs, 7)]; Tz = Tx + Ty; T2M = Tx - Ty; } { E T1b, T1c, T1e, T1f; T1b = ci[WS(rs, 17)]; T1c = cr[WS(rs, 12)]; T1d = T1b - T1c; T2N = T1b + T1c; T1e = ci[WS(rs, 12)]; T1f = cr[WS(rs, 17)]; T1g = T1e - T1f; T35 = T1e + T1f; } TA = Tw + Tz; T3Q = T34 + T35; T41 = T2M - T2N; TJ = Tw - Tz; T1h = T1d - T1g; T2O = T2M + T2N; T36 = T34 - T35; T25 = T1d + T1g; } { E Th, T2X, T11, T2Y, Tk, T2F, TY, T2G; { E Tf, Tg, TZ, T10; Tf = ci[WS(rs, 3)]; Tg = cr[WS(rs, 6)]; Th = Tf + Tg; T2X = Tf - Tg; TZ = ci[WS(rs, 18)]; T10 = cr[WS(rs, 11)]; T11 = TZ - T10; T2Y = TZ + T10; } { E Ti, Tj, TW, TX; Ti = cr[WS(rs, 1)]; Tj = ci[WS(rs, 8)]; Tk = Ti + Tj; T2F = Ti - Tj; TW = ci[WS(rs, 13)]; TX = cr[WS(rs, 16)]; TY = TW - TX; T2G = TW + TX; } Tl = Th + Tk; T3N = T2X - T2Y; T3Y = T2F - T2G; TG = Th - Tk; T12 = TY - T11; T2H = T2F + T2G; T2Z = T2X + T2Y; T22 = TY + T11; } { E Tp, T31, T19, T32, Ts, T2K, T16, T2J; { E Tn, To, T17, T18; Tn = cr[WS(rs, 8)]; To = ci[WS(rs, 1)]; Tp = Tn + To; T31 = Tn - To; T17 = ci[WS(rs, 16)]; T18 = cr[WS(rs, 13)]; T19 = T17 - T18; T32 = T17 + T18; } { E Tq, Tr, T14, T15; Tq = ci[WS(rs, 6)]; Tr = cr[WS(rs, 3)]; Ts = Tq + Tr; T2K = Tq - Tr; T14 = ci[WS(rs, 11)]; T15 = cr[WS(rs, 18)]; T16 = T14 - T15; T2J = T14 + T15; } Tt = Tp + Ts; T3P = T31 + T32; T40 = T2K + T2J; TI = Tp - Ts; T1a = T16 - T19; T2L = T2J - T2K; T33 = T31 - T32; T24 = T16 + T19; } T13 = TV - T12; T3n = T2W - T2Z; T3o = T33 - T36; T1i = T1a - T1h; T26 = T24 - T25; T4e = T3P - T3Q; T4d = T3M - T3N; T23 = T21 - T22; T1n = TI - TJ; T42 = T40 - T41; T3Z = T3X - T3Y; T1m = TF - TG; T2h = Te - Tl; T2I = T2E + T2H; T2i = Tt - TA; T2P = T2L + T2O; T30 = T2W + T2Z; T37 = T33 + T36; T38 = T30 + T37; Tm = Te + Tl; TB = Tt + TA; TC = Tm + TB; T46 = T3X + T3Y; T47 = T40 + T41; T4a = T46 + T47; T2a = T21 + T22; T2b = T24 + T25; T2e = T2a + T2b; T1w = TV + T12; T1x = T1a + T1h; T1y = T1w + T1x; T3O = T3M + T3N; T3R = T3P + T3Q; T3U = T3O + T3R; T3h = T2E - T2H; T3i = T2L - T2O; T3j = T3h + T3i; TH = TF + TG; TK = TI + TJ; TL = TH + TK; } cr[0] = T7 + TC; ci[0] = T2d + T2e; { E T1U, T1W, T1T, T1V; T1U = TE + TL; T1W = T1v + T1y; T1T = W[18]; T1V = W[19]; cr[WS(rs, 10)] = FNMS(T1V, T1W, T1T * T1U); ci[WS(rs, 10)] = FMA(T1V, T1U, T1T * T1W); } { E T4y, T4A, T4x, T4z; T4y = T3T + T3U; T4A = T49 + T4a; T4x = W[8]; T4z = W[9]; cr[WS(rs, 5)] = FNMS(T4z, T4A, T4x * T4y); ci[WS(rs, 5)] = FMA(T4x, T4A, T4z * T4y); } { E T3I, T3K, T3H, T3J; T3I = T2T + T38; T3K = T3g + T3j; T3H = W[28]; T3J = W[29]; cr[WS(rs, 15)] = FNMS(T3J, T3K, T3H * T3I); ci[WS(rs, 15)] = FMA(T3H, T3K, T3J * T3I); } { E T27, T2j, T2v, T2r, T2g, T2u, T20, T2q; T27 = FMA(KP951056516, T23, KP587785252 * T26); T2j = FMA(KP951056516, T2h, KP587785252 * T2i); T2v = FNMS(KP951056516, T2i, KP587785252 * T2h); T2r = FNMS(KP951056516, T26, KP587785252 * T23); { E T2c, T2f, T1Y, T1Z; T2c = KP559016994 * (T2a - T2b); T2f = FNMS(KP250000000, T2e, T2d); T2g = T2c + T2f; T2u = T2f - T2c; T1Y = KP559016994 * (Tm - TB); T1Z = FNMS(KP250000000, TC, T7); T20 = T1Y + T1Z; T2q = T1Z - T1Y; } { E T28, T2k, T1X, T29; T28 = T20 + T27; T2k = T2g - T2j; T1X = W[6]; T29 = W[7]; cr[WS(rs, 4)] = FNMS(T29, T2k, T1X * T28); ci[WS(rs, 4)] = FMA(T29, T28, T1X * T2k); } { E T2y, T2A, T2x, T2z; T2y = T2q - T2r; T2A = T2v + T2u; T2x = W[22]; T2z = W[23]; cr[WS(rs, 12)] = FNMS(T2z, T2A, T2x * T2y); ci[WS(rs, 12)] = FMA(T2z, T2y, T2x * T2A); } { E T2m, T2o, T2l, T2n; T2m = T20 - T27; T2o = T2j + T2g; T2l = W[30]; T2n = W[31]; cr[WS(rs, 16)] = FNMS(T2n, T2o, T2l * T2m); ci[WS(rs, 16)] = FMA(T2n, T2m, T2l * T2o); } { E T2s, T2w, T2p, T2t; T2s = T2q + T2r; T2w = T2u - T2v; T2p = W[14]; T2t = W[15]; cr[WS(rs, 8)] = FNMS(T2t, T2w, T2p * T2s); ci[WS(rs, 8)] = FMA(T2t, T2s, T2p * T2w); } } { E T43, T4f, T4r, T4m, T4c, T4q, T3W, T4n; T43 = FMA(KP951056516, T3Z, KP587785252 * T42); T4f = FMA(KP951056516, T4d, KP587785252 * T4e); T4r = FNMS(KP951056516, T4e, KP587785252 * T4d); T4m = FNMS(KP951056516, T42, KP587785252 * T3Z); { E T48, T4b, T3S, T3V; T48 = KP559016994 * (T46 - T47); T4b = FNMS(KP250000000, T4a, T49); T4c = T48 + T4b; T4q = T4b - T48; T3S = KP559016994 * (T3O - T3R); T3V = FNMS(KP250000000, T3U, T3T); T3W = T3S + T3V; T4n = T3V - T3S; } { E T44, T4g, T3L, T45; T44 = T3W - T43; T4g = T4c + T4f; T3L = W[0]; T45 = W[1]; cr[WS(rs, 1)] = FNMS(T45, T4g, T3L * T44); ci[WS(rs, 1)] = FMA(T3L, T4g, T45 * T44); } { E T4u, T4w, T4t, T4v; T4u = T4n - T4m; T4w = T4q + T4r; T4t = W[32]; T4v = W[33]; cr[WS(rs, 17)] = FNMS(T4v, T4w, T4t * T4u); ci[WS(rs, 17)] = FMA(T4t, T4w, T4v * T4u); } { E T4i, T4k, T4h, T4j; T4i = T43 + T3W; T4k = T4c - T4f; T4h = W[16]; T4j = W[17]; cr[WS(rs, 9)] = FNMS(T4j, T4k, T4h * T4i); ci[WS(rs, 9)] = FMA(T4h, T4k, T4j * T4i); } { E T4o, T4s, T4l, T4p; T4o = T4m + T4n; T4s = T4q - T4r; T4l = W[24]; T4p = W[25]; cr[WS(rs, 13)] = FNMS(T4p, T4s, T4l * T4o); ci[WS(rs, 13)] = FMA(T4l, T4s, T4p * T4o); } } { E T1j, T1o, T1M, T1J, T1B, T1N, TO, T1I; T1j = FNMS(KP951056516, T1i, KP587785252 * T13); T1o = FNMS(KP951056516, T1n, KP587785252 * T1m); T1M = FMA(KP951056516, T1m, KP587785252 * T1n); T1J = FMA(KP951056516, T13, KP587785252 * T1i); { E T1z, T1A, TM, TN; T1z = FNMS(KP250000000, T1y, T1v); T1A = KP559016994 * (T1w - T1x); T1B = T1z - T1A; T1N = T1A + T1z; TM = FNMS(KP250000000, TL, TE); TN = KP559016994 * (TH - TK); TO = TM - TN; T1I = TN + TM; } { E T1k, T1C, TD, T1l; T1k = TO - T1j; T1C = T1o + T1B; TD = W[2]; T1l = W[3]; cr[WS(rs, 2)] = FNMS(T1l, T1C, TD * T1k); ci[WS(rs, 2)] = FMA(T1l, T1k, TD * T1C); } { E T1Q, T1S, T1P, T1R; T1Q = T1I + T1J; T1S = T1N - T1M; T1P = W[26]; T1R = W[27]; cr[WS(rs, 14)] = FNMS(T1R, T1S, T1P * T1Q); ci[WS(rs, 14)] = FMA(T1R, T1Q, T1P * T1S); } { E T1E, T1G, T1D, T1F; T1E = TO + T1j; T1G = T1B - T1o; T1D = W[34]; T1F = W[35]; cr[WS(rs, 18)] = FNMS(T1F, T1G, T1D * T1E); ci[WS(rs, 18)] = FMA(T1F, T1E, T1D * T1G); } { E T1K, T1O, T1H, T1L; T1K = T1I - T1J; T1O = T1M + T1N; T1H = W[10]; T1L = W[11]; cr[WS(rs, 6)] = FNMS(T1L, T1O, T1H * T1K); ci[WS(rs, 6)] = FMA(T1L, T1K, T1H * T1O); } } { E T2Q, T3p, T3B, T3x, T3m, T3A, T3b, T3w; T2Q = FNMS(KP951056516, T2P, KP587785252 * T2I); T3p = FNMS(KP951056516, T3o, KP587785252 * T3n); T3B = FMA(KP951056516, T3n, KP587785252 * T3o); T3x = FMA(KP951056516, T2I, KP587785252 * T2P); { E T3k, T3l, T39, T3a; T3k = FNMS(KP250000000, T3j, T3g); T3l = KP559016994 * (T3h - T3i); T3m = T3k - T3l; T3A = T3l + T3k; T39 = FNMS(KP250000000, T38, T2T); T3a = KP559016994 * (T30 - T37); T3b = T39 - T3a; T3w = T3a + T39; } { E T3c, T3q, T2B, T3d; T3c = T2Q + T3b; T3q = T3m - T3p; T2B = W[4]; T3d = W[5]; cr[WS(rs, 3)] = FNMS(T3d, T3q, T2B * T3c); ci[WS(rs, 3)] = FMA(T2B, T3q, T3d * T3c); } { E T3E, T3G, T3D, T3F; T3E = T3x + T3w; T3G = T3A - T3B; T3D = W[36]; T3F = W[37]; cr[WS(rs, 19)] = FNMS(T3F, T3G, T3D * T3E); ci[WS(rs, 19)] = FMA(T3D, T3G, T3F * T3E); } { E T3s, T3u, T3r, T3t; T3s = T3b - T2Q; T3u = T3m + T3p; T3r = W[12]; T3t = W[13]; cr[WS(rs, 7)] = FNMS(T3t, T3u, T3r * T3s); ci[WS(rs, 7)] = FMA(T3r, T3u, T3t * T3s); } { E T3y, T3C, T3v, T3z; T3y = T3w - T3x; T3C = T3A + T3B; T3v = W[20]; T3z = W[21]; cr[WS(rs, 11)] = FNMS(T3z, T3C, T3v * T3y); ci[WS(rs, 11)] = FMA(T3v, T3C, T3z * T3y); } } } } } static const tw_instr twinstr[] = { {TW_FULL, 1, 20}, {TW_NEXT, 1, 0} }; static const hc2hc_desc desc = { 20, "hb_20", twinstr, &GENUS, {184, 62, 62, 0} }; void X(codelet_hb_20) (planner *p) { X(khc2hc_register) (p, hb_20, &desc); } #endif /* HAVE_FMA */ fftw-3.3.3/rdft/scalar/r2cb/hb_25.c0000644000175000017500000014321412054410771013514 00000000000000/* * Copyright (c) 2003, 2007-11 Matteo Frigo * Copyright (c) 2003, 2007-11 Massachusetts Institute of Technology * * This program is free software; you can redistribute it and/or modify * it under the terms of the GNU General Public License as published by * the Free Software Foundation; either version 2 of the License, or * (at your option) any later version. * * This program is distributed in the hope that it will be useful, * but WITHOUT ANY WARRANTY; without even the implied warranty of * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the * GNU General Public License for more details. * * You should have received a copy of the GNU General Public License * along with this program; if not, write to the Free Software * Foundation, Inc., 51 Franklin Street, Fifth Floor, Boston, MA 02110-1301 USA * */ /* This file was automatically generated --- DO NOT EDIT */ /* Generated on Sun Nov 25 07:41:23 EST 2012 */ #include "codelet-rdft.h" #ifdef HAVE_FMA /* Generated by: ../../../genfft/gen_hc2hc.native -fma -reorder-insns -schedule-for-pipeline -compact -variables 4 -pipeline-latency 4 -sign 1 -n 25 -dif -name hb_25 -include hb.h */ /* * This function contains 400 FP additions, 364 FP multiplications, * (or, 84 additions, 48 multiplications, 316 fused multiply/add), * 176 stack variables, 47 constants, and 100 memory accesses */ #include "hb.h" static void hb_25(R *cr, R *ci, const R *W, stride rs, INT mb, INT me, INT ms) { DK(KP921177326, +0.921177326965143320250447435415066029359282231); DK(KP833417178, +0.833417178328688677408962550243238843138996060); DK(KP541454447, +0.541454447536312777046285590082819509052033189); DK(KP968583161, +0.968583161128631119490168375464735813836012403); DK(KP242145790, +0.242145790282157779872542093866183953459003101); DK(KP904730450, +0.904730450839922351881287709692877908104763647); DK(KP559154169, +0.559154169276087864842202529084232643714075927); DK(KP683113946, +0.683113946453479238701949862233725244439656928); DK(KP831864738, +0.831864738706457140726048799369896829771167132); DK(KP871714437, +0.871714437527667770979999223229522602943903653); DK(KP803003575, +0.803003575438660414833440593570376004635464850); DK(KP554608978, +0.554608978404018097464974850792216217022558774); DK(KP248028675, +0.248028675328619457762448260696444630363259177); DK(KP992114701, +0.992114701314477831049793042785778521453036709); DK(KP851038619, +0.851038619207379630836264138867114231259902550); DK(KP525970792, +0.525970792408939708442463226536226366643874659); DK(KP726211448, +0.726211448929902658173535992263577167607493062); DK(KP912018591, +0.912018591466481957908415381764119056233607330); DK(KP912575812, +0.912575812670962425556968549836277086778922727); DK(KP943557151, +0.943557151597354104399655195398983005179443399); DK(KP994076283, +0.994076283785401014123185814696322018529298887); DK(KP614372930, +0.614372930789563808870829930444362096004872855); DK(KP621716863, +0.621716863012209892444754556304102309693593202); DK(KP772036680, +0.772036680810363904029489473607579825330539880); DK(KP734762448, +0.734762448793050413546343770063151342619912334); DK(KP860541664, +0.860541664367944677098261680920518816412804187); DK(KP949179823, +0.949179823508441261575555465843363271711583843); DK(KP557913902, +0.557913902031834264187699648465567037992437152); DK(KP998026728, +0.998026728428271561952336806863450553336905220); DK(KP249506682, +0.249506682107067890488084201715862638334226305); DK(KP906616052, +0.906616052148196230441134447086066874408359177); DK(KP681693190, +0.681693190061530575150324149145440022633095390); DK(KP560319534, +0.560319534973832390111614715371676131169633784); DK(KP845997307, +0.845997307939530944175097360758058292389769300); DK(KP968479752, +0.968479752739016373193524836781420152702090879); DK(KP062914667, +0.062914667253649757225485955897349402364686947); DK(KP827271945, +0.827271945972475634034355757144307982555673741); DK(KP470564281, +0.470564281212251493087595091036643380879947982); DK(KP126329378, +0.126329378446108174786050455341811215027378105); DK(KP256756360, +0.256756360367726783319498520922669048172391148); DK(KP634619297, +0.634619297544148100711287640319130485732531031); DK(KP549754652, +0.549754652192770074288023275540779861653779767); DK(KP939062505, +0.939062505817492352556001843133229685779824606); DK(KP951056516, +0.951056516295153572116439333379382143405698634); DK(KP559016994, +0.559016994374947424102293417182819058860154590); DK(KP618033988, +0.618033988749894848204586834365638117720309180); DK(KP250000000, +0.250000000000000000000000000000000000000000000); { INT m; for (m = mb, W = W + ((mb - 1) * 48); m < me; m = m + 1, cr = cr + ms, ci = ci - ms, W = W + 48, MAKE_VOLATILE_STRIDE(50, rs)) { E T3w, T3P, T2d, T3y, T3x, T3Q; { E T9, T3E, T1F, T3B, T6f, T7d, T5u, T6U, T4k, T2k, T5G, T1G, T19, T1H, T1s; E T1M, T1N, TP, TM, T7i, T77, T5X, T64, T4u, T4D, T3p, T2z, T74, T7h, T63; E T5Q, T4x, T4E, T3q, T2O, T4n, T4G, T3t, T3j, T5F, T70, T7f, T66, T5B, T4q; E T4H, T3s, T34, T5E, T6V; { E T2f, T2e, T6e, T2j, T5t, T6d; { E T1, T1x, T3C, T3D, T8, T2h, T1A, T1D, T2i, T3A, T1E, T3z; T1 = cr[0]; T1x = ci[WS(rs, 24)]; { E T2, T3, T5, T6; T2 = cr[WS(rs, 5)]; T3 = ci[WS(rs, 4)]; T5 = cr[WS(rs, 10)]; T6 = ci[WS(rs, 9)]; { E T1y, T4, T7, T1z, T1B, T1C; T1y = ci[WS(rs, 19)]; T3C = T2 - T3; T4 = T2 + T3; T3D = T5 - T6; T7 = T5 + T6; T1z = cr[WS(rs, 20)]; T1B = ci[WS(rs, 14)]; T1C = cr[WS(rs, 15)]; T8 = T4 + T7; T2f = T4 - T7; T2h = T1y + T1z; T1A = T1y - T1z; T1D = T1B - T1C; T2i = T1B + T1C; } } T2e = FNMS(KP250000000, T8, T1); T9 = T1 + T8; T3A = T1A - T1D; T1E = T1A + T1D; T3E = FMA(KP618033988, T3D, T3C); T6e = FNMS(KP618033988, T3C, T3D); T2j = FMA(KP618033988, T2i, T2h); T5t = FNMS(KP618033988, T2h, T2i); T1F = T1x + T1E; T3z = FNMS(KP250000000, T1E, T1x); T6d = FNMS(KP559016994, T3A, T3z); T3B = FMA(KP559016994, T3A, T3z); } { E T2x, T5V, T2m, T2l, Ti, T5w, T3h, T36, TK, T35, T2F, T5L, T2I, Tr, T2H; E T3a, T5z, T3d, T1r, T3c, T2q, T5S, T2t, TZ, T2s, T5O, T2M, T2B, Tt, T18; E T2A, Tx, T2V, T2Y, Tw, T30, T1i, T2X, Ty; { E T1j, T1k, T1p, T39, T1l; { E TC, TI, T3g, TD, TE; { E Ta, Te, Tf, Tb, Tc, T5s, T2g, T2w, Tg; Ta = cr[WS(rs, 1)]; T5s = FNMS(KP559016994, T2f, T2e); T2g = FMA(KP559016994, T2f, T2e); T6f = FNMS(KP951056516, T6e, T6d); T7d = FMA(KP951056516, T6e, T6d); Te = cr[WS(rs, 11)]; T5u = FMA(KP951056516, T5t, T5s); T6U = FNMS(KP951056516, T5t, T5s); T4k = FMA(KP951056516, T2j, T2g); T2k = FNMS(KP951056516, T2j, T2g); Tf = ci[WS(rs, 8)]; Tb = cr[WS(rs, 6)]; Tc = ci[WS(rs, 3)]; TC = cr[WS(rs, 3)]; T2w = Tf - Te; Tg = Te + Tf; { E T2v, Td, Th, TG, TH; T2v = Tb - Tc; Td = Tb + Tc; TG = ci[WS(rs, 11)]; TH = ci[WS(rs, 6)]; T2x = FNMS(KP618033988, T2w, T2v); T5V = FMA(KP618033988, T2v, T2w); Th = Td + Tg; T2m = Td - Tg; TI = TG + TH; T3g = TG - TH; T2l = FNMS(KP250000000, Th, Ta); Ti = Ta + Th; TD = cr[WS(rs, 8)]; TE = ci[WS(rs, 1)]; } } { E Tj, Tk, Tp, T2E, TJ, Tl; Tj = cr[WS(rs, 4)]; { E Tn, To, T3f, TF; Tn = ci[WS(rs, 10)]; To = ci[WS(rs, 5)]; T3f = TD - TE; TF = TD + TE; Tk = cr[WS(rs, 9)]; Tp = Tn + To; T2E = To - Tn; T5w = FNMS(KP618033988, T3f, T3g); T3h = FMA(KP618033988, T3g, T3f); T36 = TI - TF; TJ = TF + TI; Tl = ci[0]; } T1j = ci[WS(rs, 21)]; TK = TC + TJ; T35 = FNMS(KP250000000, TJ, TC); { E T1n, Tm, T2D, T1o, Tq; T1n = cr[WS(rs, 13)]; Tm = Tk + Tl; T2D = Tl - Tk; T1o = cr[WS(rs, 18)]; T1k = ci[WS(rs, 16)]; T2F = FMA(KP618033988, T2E, T2D); T5L = FNMS(KP618033988, T2D, T2E); T2I = Tm - Tp; Tq = Tm + Tp; T1p = T1n + T1o; T39 = T1o - T1n; Tr = Tj + Tq; T2H = FMS(KP250000000, Tq, Tj); T1l = cr[WS(rs, 23)]; } } } { E T10, T11, T16, T2L, T12; { E TR, TS, TX, T2p, T1q, TT; TR = ci[WS(rs, 23)]; { E TV, TW, T38, T1m; TV = ci[WS(rs, 13)]; TW = cr[WS(rs, 16)]; T38 = T1k + T1l; T1m = T1k - T1l; TS = ci[WS(rs, 18)]; TX = TV - TW; T2p = TV + TW; T3a = FMA(KP618033988, T39, T38); T5z = FNMS(KP618033988, T38, T39); T3d = T1m + T1p; T1q = T1m - T1p; TT = cr[WS(rs, 21)]; } T10 = ci[WS(rs, 20)]; T1r = T1j + T1q; T3c = FMS(KP250000000, T1q, T1j); { E T14, TU, T2o, T15, TY; T14 = cr[WS(rs, 14)]; TU = TS - TT; T2o = TS + TT; T15 = cr[WS(rs, 19)]; T11 = ci[WS(rs, 15)]; T2q = FMA(KP618033988, T2p, T2o); T5S = FNMS(KP618033988, T2o, T2p); T2t = TU - TX; TY = TU + TX; T16 = T14 + T15; T2L = T15 - T14; TZ = TR + TY; T2s = FNMS(KP250000000, TY, TR); T12 = cr[WS(rs, 24)]; } } { E T1a, T1e, T1d, T2T, T17, T1f; T1a = ci[WS(rs, 22)]; { E T1b, T1c, T2K, T13; T1b = ci[WS(rs, 17)]; T1c = cr[WS(rs, 22)]; T2K = T11 + T12; T13 = T11 - T12; T1e = ci[WS(rs, 12)]; T1d = T1b - T1c; T2T = T1b + T1c; T5O = FNMS(KP618033988, T2K, T2L); T2M = FMA(KP618033988, T2L, T2K); T2B = T13 + T16; T17 = T13 - T16; T1f = cr[WS(rs, 17)]; } Tt = cr[WS(rs, 2)]; T18 = T10 + T17; T2A = FMS(KP250000000, T17, T10); { E Tu, T1g, T2U, Tv, T1h; Tu = cr[WS(rs, 7)]; T1g = T1e - T1f; T2U = T1e + T1f; Tv = ci[WS(rs, 2)]; Tx = cr[WS(rs, 12)]; T2V = FMA(KP618033988, T2U, T2T); T5G = FNMS(KP618033988, T2T, T2U); T2Y = T1d - T1g; T1h = T1d + T1g; Tw = Tu + Tv; T30 = Tu - Tv; T1i = T1a + T1h; T2X = FMS(KP250000000, T1h, T1a); Ty = ci[WS(rs, 7)]; } } } } { E T32, T5D, T2R, T2Q, T2u, T2r, T4t; { E TA, T31, Tz, TB, Ts; T31 = Ty - Tx; Tz = Tx + Ty; T1G = TZ + T18; T19 = TZ - T18; T32 = FNMS(KP618033988, T31, T30); T5D = FMA(KP618033988, T30, T31); TA = Tw + Tz; T2R = Tz - Tw; T2Q = FNMS(KP250000000, TA, Tt); TB = Tt + TA; T1H = T1i + T1r; T1s = T1i - T1r; T1M = Ti - Tr; Ts = Ti + Tr; { E T2n, T5R, T5U, TL; T2n = FMA(KP559016994, T2m, T2l); T5R = FNMS(KP559016994, T2m, T2l); T5U = FNMS(KP559016994, T2t, T2s); T2u = FMA(KP559016994, T2t, T2s); TL = TB + TK; T1N = TB - TK; { E T5T, T75, T5W, T76; T5T = FMA(KP951056516, T5S, T5R); T75 = FNMS(KP951056516, T5S, T5R); T5W = FMA(KP951056516, T5V, T5U); T76 = FNMS(KP951056516, T5V, T5U); TP = Ts - TL; TM = Ts + TL; T2r = FNMS(KP951056516, T2q, T2n); T4t = FMA(KP951056516, T2q, T2n); T7i = FMA(KP939062505, T75, T76); T77 = FNMS(KP939062505, T76, T75); T5X = FNMS(KP549754652, T5W, T5T); T64 = FMA(KP549754652, T5T, T5W); } } } { E T2J, T2G, T4v, T5y, T37, T3e, T5v; { E T2C, T5K, T5N, T4s, T2y; T2C = FNMS(KP559016994, T2B, T2A); T5K = FMA(KP559016994, T2B, T2A); T5N = FMA(KP559016994, T2I, T2H); T2J = FNMS(KP559016994, T2I, T2H); T4s = FNMS(KP951056516, T2x, T2u); T2y = FMA(KP951056516, T2x, T2u); { E T73, T5M, T72, T5P; T73 = FMA(KP951056516, T5L, T5K); T5M = FNMS(KP951056516, T5L, T5K); T72 = FMA(KP951056516, T5O, T5N); T5P = FNMS(KP951056516, T5O, T5N); T4u = FNMS(KP634619297, T4t, T4s); T4D = FMA(KP634619297, T4s, T4t); T3p = FMA(KP256756360, T2r, T2y); T2z = FNMS(KP256756360, T2y, T2r); T74 = FMA(KP126329378, T73, T72); T7h = FNMS(KP126329378, T72, T73); T63 = FNMS(KP470564281, T5M, T5P); T5Q = FMA(KP470564281, T5P, T5M); T2G = FMA(KP951056516, T2F, T2C); T4v = FNMS(KP951056516, T2F, T2C); } T5y = FMA(KP559016994, T36, T35); T37 = FNMS(KP559016994, T36, T35); T3e = FNMS(KP559016994, T3d, T3c); T5v = FMA(KP559016994, T3d, T3c); } { E T5x, T6Y, T4w, T2N; T4w = FNMS(KP951056516, T2M, T2J); T2N = FMA(KP951056516, T2M, T2J); { E T4l, T3b, T4m, T3i; T4l = FMA(KP951056516, T3a, T37); T3b = FNMS(KP951056516, T3a, T37); T4m = FMA(KP951056516, T3h, T3e); T3i = FNMS(KP951056516, T3h, T3e); T4x = FNMS(KP827271945, T4w, T4v); T4E = FMA(KP827271945, T4v, T4w); T3q = FMA(KP634619297, T2G, T2N); T2O = FNMS(KP634619297, T2N, T2G); T4n = FNMS(KP126329378, T4m, T4l); T4G = FMA(KP126329378, T4l, T4m); T3t = FNMS(KP939062505, T3b, T3i); T3j = FMA(KP939062505, T3i, T3b); T5x = FMA(KP951056516, T5w, T5v); T6Y = FNMS(KP951056516, T5w, T5v); } { E T2S, T2Z, T5C, T6Z, T5A; T5F = FMA(KP559016994, T2R, T2Q); T2S = FNMS(KP559016994, T2R, T2Q); T2Z = FNMS(KP559016994, T2Y, T2X); T5C = FMA(KP559016994, T2Y, T2X); T6Z = FNMS(KP951056516, T5z, T5y); T5A = FMA(KP951056516, T5z, T5y); { E T4p, T2W, T4o, T33; T4p = FMA(KP951056516, T2V, T2S); T2W = FNMS(KP951056516, T2V, T2S); T4o = FMA(KP951056516, T32, T2Z); T33 = FNMS(KP951056516, T32, T2Z); T70 = FNMS(KP827271945, T6Z, T6Y); T7f = FMA(KP827271945, T6Y, T6Z); T66 = FNMS(KP062914667, T5x, T5A); T5B = FMA(KP062914667, T5A, T5x); T4q = FNMS(KP470564281, T4p, T4o); T4H = FMA(KP470564281, T4o, T4p); T3s = FNMS(KP549754652, T2W, T33); T34 = FMA(KP549754652, T33, T2W); T5E = FNMS(KP951056516, T5D, T5C); T6V = FMA(KP951056516, T5D, T5C); } } } } } } } { E T6X, T7e, T6A, T6F, T6C, T6G, T6B; cr[0] = T9 + TM; { E T67, T5I, T25, T22, T1X, T26, T21; { E T1I, T23, T1L, T1Z, T1t, TO, T24, T1O; { E T1K, T6W, T5H, T1J; T1K = T1G - T1H; T1I = T1G + T1H; T6W = FNMS(KP951056516, T5G, T5F); T5H = FMA(KP951056516, T5G, T5F); T1J = FNMS(KP250000000, T1I, T1F); T6X = FMA(KP062914667, T6W, T6V); T7e = FNMS(KP062914667, T6V, T6W); T67 = FNMS(KP634619297, T5E, T5H); T5I = FMA(KP634619297, T5H, T5E); T23 = FNMS(KP559016994, T1K, T1J); T1L = FMA(KP559016994, T1K, T1J); T1Z = FNMS(KP618033988, T19, T1s); T1t = FMA(KP618033988, T1s, T19); TO = FNMS(KP250000000, TM, T9); T24 = FNMS(KP618033988, T1M, T1N); T1O = FMA(KP618033988, T1N, T1M); } { E T2b, T2a, T1Y, TQ, T27; ci[0] = T1F + T1I; T2b = FMA(KP951056516, T24, T23); T25 = FNMS(KP951056516, T24, T23); T2a = W[29]; T1Y = FNMS(KP559016994, TP, TO); TQ = FMA(KP559016994, TP, TO); T27 = W[28]; { E T1V, T1P, T20, T1S, T1w, T1v, TN, T1Q; T1V = FNMS(KP951056516, T1O, T1L); T1P = FMA(KP951056516, T1O, T1L); { E T28, T1u, T29, T2c; T20 = FMA(KP951056516, T1Z, T1Y); T28 = FNMS(KP951056516, T1Z, T1Y); T1S = FMA(KP951056516, T1t, TQ); T1u = FNMS(KP951056516, T1t, TQ); T1w = W[9]; T29 = T27 * T28; T2c = T2a * T28; TN = W[8]; T1Q = T1w * T1u; cr[WS(rs, 15)] = FNMS(T2a, T2b, T29); ci[WS(rs, 15)] = FMA(T27, T2b, T2c); T1v = TN * T1u; } ci[WS(rs, 5)] = FMA(TN, T1P, T1Q); { E T1U, T1R, T1W, T1T; T1U = W[39]; cr[WS(rs, 5)] = FNMS(T1w, T1P, T1v); T1R = W[38]; T1W = T1U * T1S; T22 = W[19]; T1T = T1R * T1S; T1X = W[18]; ci[WS(rs, 20)] = FMA(T1R, T1V, T1W); T26 = T22 * T20; cr[WS(rs, 20)] = FNMS(T1U, T1V, T1T); T21 = T1X * T20; } } } } { E T6h, T6g, T5Y, T5J, T6z, T69, T6o, T6E; { E T6m, T6n, T65, T68; T65 = FMA(KP968479752, T64, T63); T6h = FNMS(KP968479752, T64, T63); ci[WS(rs, 10)] = FMA(T1X, T25, T26); T68 = FNMS(KP845997307, T67, T66); T6g = FMA(KP845997307, T67, T66); cr[WS(rs, 10)] = FNMS(T22, T25, T21); T6m = FNMS(KP968479752, T5X, T5Q); T5Y = FMA(KP968479752, T5X, T5Q); T5J = FMA(KP845997307, T5I, T5B); T6n = FNMS(KP845997307, T5I, T5B); T6z = FMA(KP560319534, T65, T68); T69 = FNMS(KP681693190, T68, T65); T6o = FMA(KP681693190, T6n, T6m); T6E = FNMS(KP560319534, T6m, T6n); } { E T62, T6l, T6I, T6L, T6H, T6K; { E T6Q, T6O, T6y, T6D, T6S; { E T6N, T5Z, T61, T6i, T6k; T6N = W[2]; T5Z = FMA(KP906616052, T5Y, T5J); T61 = FNMS(KP906616052, T5Y, T5J); T6i = FNMS(KP906616052, T6h, T6g); T6k = FMA(KP906616052, T6h, T6g); T6Q = W[3]; { E T60, T6j, T6R, T6P; T60 = FNMS(KP249506682, T5Z, T5u); T6O = FMA(KP998026728, T5Z, T5u); T6j = FNMS(KP249506682, T6i, T6f); T6R = FMA(KP998026728, T6i, T6f); T6y = FMA(KP557913902, T61, T60); T62 = FNMS(KP557913902, T61, T60); T6P = T6N * T6O; T6l = FNMS(KP557913902, T6k, T6j); T6D = FMA(KP557913902, T6k, T6j); T6S = T6N * T6R; cr[WS(rs, 2)] = FNMS(T6Q, T6R, T6P); } } T6A = FNMS(KP949179823, T6z, T6y); T6I = FMA(KP949179823, T6z, T6y); T6L = FNMS(KP949179823, T6E, T6D); T6F = FMA(KP949179823, T6E, T6D); ci[WS(rs, 2)] = FMA(T6Q, T6O, T6S); T6H = W[32]; T6K = W[33]; } { E T6a, T6s, T6v, T6p, T6c, T6q, T6b, T6M, T6J, T5r; T6a = FNMS(KP860541664, T69, T62); T6s = FMA(KP860541664, T69, T62); T6v = FMA(KP860541664, T6o, T6l); T6p = FNMS(KP860541664, T6o, T6l); T6M = T6H * T6L; T6J = T6H * T6I; T5r = W[12]; T6c = W[13]; ci[WS(rs, 17)] = FMA(T6K, T6I, T6M); cr[WS(rs, 17)] = FNMS(T6K, T6L, T6J); T6q = T5r * T6p; T6b = T5r * T6a; { E T6r, T6u, T6w, T6t, T6x; ci[WS(rs, 7)] = FMA(T6c, T6a, T6q); cr[WS(rs, 7)] = FNMS(T6c, T6p, T6b); T6r = W[42]; T6u = W[43]; T6w = T6r * T6v; T6t = T6r * T6s; T6x = W[22]; T6C = W[23]; ci[WS(rs, 22)] = FMA(T6u, T6s, T6w); cr[WS(rs, 22)] = FNMS(T6u, T6v, T6t); T6G = T6x * T6F; T6B = T6x * T6A; } } } } } { E T7u, T7D, T7n, T7w, T7v, T7E; { E T78, T7t, T7N, T71, T7C, T7S, T7y, T7k; { E T7j, T7g, T7A, T7B, T7r, T7s; T7r = FNMS(KP734762448, T7i, T7h); T7j = FMA(KP734762448, T7i, T7h); T7g = FMA(KP772036680, T7f, T7e); T7s = FNMS(KP772036680, T7f, T7e); ci[WS(rs, 12)] = FMA(T6C, T6A, T6G); cr[WS(rs, 12)] = FNMS(T6C, T6F, T6B); T7A = FNMS(KP734762448, T77, T74); T78 = FMA(KP734762448, T77, T74); T7t = FNMS(KP621716863, T7s, T7r); T7N = FMA(KP614372930, T7r, T7s); T71 = FMA(KP772036680, T70, T6X); T7B = FNMS(KP772036680, T70, T6X); T7C = FNMS(KP621716863, T7B, T7A); T7S = FMA(KP614372930, T7A, T7B); T7y = FNMS(KP994076283, T7j, T7g); T7k = FMA(KP994076283, T7j, T7g); } { E T7c, T6T, T7x, T7l, T79, T7p; T7c = W[5]; T6T = W[4]; T7x = FNMS(KP249506682, T7k, T7d); T7l = FMA(KP998026728, T7k, T7d); T79 = FMA(KP994076283, T78, T71); T7p = FNMS(KP994076283, T78, T71); { E T7z, T7Y, T7Z, T7T, T7q, T7O, T7X, T7L, T7Q, T7P, T7U; { E T7V, T80, T7b, T7m, T7W; { E T7R, T7o, T7a, T7M; T7V = W[34]; T7R = FMA(KP557913902, T7y, T7x); T7z = FNMS(KP557913902, T7y, T7x); T7Y = W[35]; T7o = FNMS(KP249506682, T79, T6U); T7a = FMA(KP998026728, T79, T6U); T7Z = FMA(KP949179823, T7S, T7R); T7T = FNMS(KP949179823, T7S, T7R); T7M = FMA(KP557913902, T7p, T7o); T7q = FNMS(KP557913902, T7p, T7o); T7b = T6T * T7a; T7m = T7c * T7a; T7W = FNMS(KP949179823, T7N, T7M); T7O = FMA(KP949179823, T7N, T7M); } cr[WS(rs, 3)] = FNMS(T7c, T7l, T7b); ci[WS(rs, 3)] = FMA(T6T, T7l, T7m); T80 = T7Y * T7W; T7X = T7V * T7W; T7L = W[24]; T7Q = W[25]; ci[WS(rs, 18)] = FMA(T7V, T7Z, T80); } cr[WS(rs, 18)] = FNMS(T7Y, T7Z, T7X); T7P = T7L * T7O; T7U = T7Q * T7O; { E T7J, T7F, T7I, T7H, T7K, T7G; T7u = FMA(KP943557151, T7t, T7q); T7G = FNMS(KP943557151, T7t, T7q); cr[WS(rs, 13)] = FNMS(T7Q, T7T, T7P); ci[WS(rs, 13)] = FMA(T7L, T7T, T7U); T7J = FMA(KP943557151, T7C, T7z); T7D = FNMS(KP943557151, T7C, T7z); T7F = W[44]; T7I = W[45]; T7n = W[14]; T7H = T7F * T7G; T7K = T7I * T7G; T7w = W[15]; T7v = T7n * T7u; cr[WS(rs, 23)] = FNMS(T7I, T7J, T7H); ci[WS(rs, 23)] = FMA(T7F, T7J, T7K); } } } } T7E = T7w * T7u; cr[WS(rs, 8)] = FNMS(T7w, T7D, T7v); { E T3F, T4K, T4X, T4j, T4M, T4L, T4Y; { E T4P, T4O, T4y, T4r, T4J, T57, T4N, T5c, T4W; { E T4U, T4V, T4F, T4I; T4F = FNMS(KP912575812, T4E, T4D); T4P = FMA(KP912575812, T4E, T4D); T4O = FMA(KP912018591, T4H, T4G); T4I = FNMS(KP912018591, T4H, T4G); ci[WS(rs, 8)] = FMA(T7n, T7D, T7E); T4y = FMA(KP912575812, T4x, T4u); T4U = FNMS(KP912575812, T4x, T4u); T4V = FMA(KP912018591, T4q, T4n); T4r = FNMS(KP912018591, T4q, T4n); T4J = FNMS(KP726211448, T4I, T4F); T57 = FMA(KP525970792, T4F, T4I); T3F = FMA(KP951056516, T3E, T3B); T4N = FNMS(KP951056516, T3E, T3B); T5c = FMA(KP525970792, T4U, T4V); T4W = FNMS(KP726211448, T4V, T4U); } { E T5o, T4S, T4B, T5l, T5p, T4R, T4A, T5m, T4Q, T4z; T5o = W[7]; T4Q = FMA(KP851038619, T4P, T4O); T4S = FNMS(KP851038619, T4P, T4O); T4z = FMA(KP851038619, T4y, T4r); T4B = FNMS(KP851038619, T4y, T4r); T5l = W[6]; T5p = FMA(KP992114701, T4Q, T4N); T4R = FNMS(KP248028675, T4Q, T4N); T4A = FMA(KP248028675, T4z, T4k); T5m = FNMS(KP992114701, T4z, T4k); { E T4T, T4C, T5d, T58, T55, T5a, T59, T5e; { E T5f, T5j, T5i, T5h, T5k, T5g; T5f = W[36]; { E T5b, T56, T5n, T5q; T4T = FNMS(KP554608978, T4S, T4R); T5b = FMA(KP554608978, T4S, T4R); T56 = FNMS(KP554608978, T4B, T4A); T4C = FMA(KP554608978, T4B, T4A); T5n = T5l * T5m; T5q = T5o * T5m; T5j = FMA(KP943557151, T5c, T5b); T5d = FNMS(KP943557151, T5c, T5b); T5g = FMA(KP943557151, T57, T56); T58 = FNMS(KP943557151, T57, T56); cr[WS(rs, 4)] = FNMS(T5o, T5p, T5n); ci[WS(rs, 4)] = FMA(T5l, T5p, T5q); } T5i = W[37]; T5h = T5f * T5g; T55 = W[26]; T5k = T5i * T5g; T5a = W[27]; cr[WS(rs, 19)] = FNMS(T5i, T5j, T5h); T59 = T55 * T58; ci[WS(rs, 19)] = FMA(T5f, T5j, T5k); } T5e = T5a * T58; { E T53, T4Z, T52, T51, T54, T50; cr[WS(rs, 14)] = FNMS(T5a, T5d, T59); T4K = FNMS(KP803003575, T4J, T4C); T50 = FMA(KP803003575, T4J, T4C); ci[WS(rs, 14)] = FMA(T55, T5d, T5e); T4X = FNMS(KP803003575, T4W, T4T); T53 = FMA(KP803003575, T4W, T4T); T4Z = W[46]; T52 = W[47]; T4j = W[16]; T51 = T4Z * T50; T54 = T52 * T50; T4M = W[17]; T4L = T4j * T4K; cr[WS(rs, 24)] = FNMS(T52, T53, T51); ci[WS(rs, 24)] = FMA(T4Z, T53, T54); } } } } T4Y = T4M * T4K; cr[WS(rs, 9)] = FNMS(T4M, T4X, T4L); { E T3G, T3H, T2P, T3k, T3Z, T3v, T3O, T44; { E T3M, T3N, T3r, T3u; T3G = FNMS(KP871714437, T3q, T3p); T3r = FMA(KP871714437, T3q, T3p); T3u = FNMS(KP831864738, T3t, T3s); T3H = FMA(KP831864738, T3t, T3s); ci[WS(rs, 9)] = FMA(T4j, T4X, T4Y); T3M = FNMS(KP871714437, T2O, T2z); T2P = FMA(KP871714437, T2O, T2z); T3k = FMA(KP831864738, T3j, T34); T3N = FNMS(KP831864738, T3j, T34); T3Z = FMA(KP683113946, T3r, T3u); T3v = FNMS(KP559154169, T3u, T3r); T3O = FMA(KP559154169, T3N, T3M); T44 = FNMS(KP683113946, T3M, T3N); } { E T4g, T3K, T3n, T4d, T3J, T4h, T4e, T3m, T3I, T3l; T4g = W[1]; T3K = FMA(KP904730450, T3H, T3G); T3I = FNMS(KP904730450, T3H, T3G); T3n = FNMS(KP904730450, T3k, T2P); T3l = FMA(KP904730450, T3k, T2P); T4d = W[0]; T3J = FNMS(KP242145790, T3I, T3F); T4h = FMA(KP968583161, T3I, T3F); T4e = FMA(KP968583161, T3l, T2k); T3m = FNMS(KP242145790, T3l, T2k); { E T3L, T3o, T45, T40, T3X, T42, T41, T46; { E T47, T4b, T4a, T49, T4c, T48; T47 = W[30]; { E T43, T3Y, T4f, T4i; T43 = FNMS(KP541454447, T3K, T3J); T3L = FMA(KP541454447, T3K, T3J); T3o = FMA(KP541454447, T3n, T3m); T3Y = FNMS(KP541454447, T3n, T3m); T4f = T4d * T4e; T4i = T4g * T4e; T45 = FNMS(KP833417178, T44, T43); T4b = FMA(KP833417178, T44, T43); T40 = FNMS(KP833417178, T3Z, T3Y); T48 = FMA(KP833417178, T3Z, T3Y); cr[WS(rs, 1)] = FNMS(T4g, T4h, T4f); ci[WS(rs, 1)] = FMA(T4d, T4h, T4i); } T4a = W[31]; T49 = T47 * T48; T3X = W[20]; T4c = T4a * T48; T42 = W[21]; cr[WS(rs, 16)] = FNMS(T4a, T4b, T49); T41 = T3X * T40; ci[WS(rs, 16)] = FMA(T47, T4b, T4c); } T46 = T42 * T40; { E T3V, T3R, T3U, T3T, T3W, T3S; cr[WS(rs, 11)] = FNMS(T42, T45, T41); T3S = FMA(KP921177326, T3v, T3o); T3w = FNMS(KP921177326, T3v, T3o); ci[WS(rs, 11)] = FMA(T3X, T45, T46); T3V = FNMS(KP921177326, T3O, T3L); T3P = FMA(KP921177326, T3O, T3L); T3R = W[40]; T3U = W[41]; T2d = W[10]; T3T = T3R * T3S; T3W = T3U * T3S; T3y = W[11]; T3x = T2d * T3w; cr[WS(rs, 21)] = FNMS(T3U, T3V, T3T); ci[WS(rs, 21)] = FMA(T3R, T3V, T3W); } } } } } } } } cr[WS(rs, 6)] = FNMS(T3y, T3P, T3x); T3Q = T3y * T3w; ci[WS(rs, 6)] = FMA(T2d, T3P, T3Q); } } } static const tw_instr twinstr[] = { {TW_FULL, 1, 25}, {TW_NEXT, 1, 0} }; static const hc2hc_desc desc = { 25, "hb_25", twinstr, &GENUS, {84, 48, 316, 0} }; void X(codelet_hb_25) (planner *p) { X(khc2hc_register) (p, hb_25, &desc); } #else /* HAVE_FMA */ /* Generated by: ../../../genfft/gen_hc2hc.native -compact -variables 4 -pipeline-latency 4 -sign 1 -n 25 -dif -name hb_25 -include hb.h */ /* * This function contains 400 FP additions, 280 FP multiplications, * (or, 260 additions, 140 multiplications, 140 fused multiply/add), * 107 stack variables, 20 constants, and 100 memory accesses */ #include "hb.h" static void hb_25(R *cr, R *ci, const R *W, stride rs, INT mb, INT me, INT ms) { DK(KP998026728, +0.998026728428271561952336806863450553336905220); DK(KP062790519, +0.062790519529313376076178224565631133122484832); DK(KP992114701, +0.992114701314477831049793042785778521453036709); DK(KP125333233, +0.125333233564304245373118759816508793942918247); DK(KP425779291, +0.425779291565072648862502445744251703979973042); DK(KP904827052, +0.904827052466019527713668647932697593970413911); DK(KP248689887, +0.248689887164854788242283746006447968417567406); DK(KP968583161, +0.968583161128631119490168375464735813836012403); DK(KP770513242, +0.770513242775789230803009636396177847271667672); DK(KP637423989, +0.637423989748689710176712811676016195434917298); DK(KP844327925, +0.844327925502015078548558063966681505381659241); DK(KP535826794, +0.535826794978996618271308767867639978063575346); DK(KP684547105, +0.684547105928688673732283357621209269889519233); DK(KP728968627, +0.728968627421411523146730319055259111372571664); DK(KP481753674, +0.481753674101715274987191502872129653528542010); DK(KP876306680, +0.876306680043863587308115903922062583399064238); DK(KP559016994, +0.559016994374947424102293417182819058860154590); DK(KP250000000, +0.250000000000000000000000000000000000000000000); DK(KP587785252, +0.587785252292473129168705954639072768597652438); DK(KP951056516, +0.951056516295153572116439333379382143405698634); { INT m; for (m = mb, W = W + ((mb - 1) * 48); m < me; m = m + 1, cr = cr + ms, ci = ci - ms, W = W + 48, MAKE_VOLATILE_STRIDE(50, rs)) { E T9, T5Q, T3y, T39, T5v, Ti, Tr, Ts, TZ, T18, T1z, T2k, T4l, T3h, T44; E T5d, T6C, T5C, T6o, T56, T6B, T5B, T6l, T2z, T4m, T3i, T47, T1K, T5w, T3c; E T3B, T5R, TB, TK, TL, T1i, T1r, T1A, T2P, T4o, T3k, T4b, T5s, T6F, T5F; E T6v, T5l, T6E, T5E, T6s, T34, T4p, T3l, T4e; { E T1, T4, T7, T8, T3x, T3w, T37, T38; T1 = cr[0]; { E T2, T3, T5, T6; T2 = cr[WS(rs, 5)]; T3 = ci[WS(rs, 4)]; T4 = T2 + T3; T5 = cr[WS(rs, 10)]; T6 = ci[WS(rs, 9)]; T7 = T5 + T6; T8 = T4 + T7; T3x = T5 - T6; T3w = T2 - T3; } T9 = T1 + T8; T5Q = FMA(KP951056516, T3w, KP587785252 * T3x); T3y = FNMS(KP951056516, T3x, KP587785252 * T3w); T37 = FNMS(KP250000000, T8, T1); T38 = KP559016994 * (T4 - T7); T39 = T37 - T38; T5v = T38 + T37; } { E Ta, T27, T53, T2f, Th, T26, T10, T2p, T58, T2x, T17, T2o, Tj, T2n, T5a; E T2t, Tq, T2s, TR, T2b, T51, T2h, TY, T2g; { E Tg, T2e, Td, T2d; Ta = cr[WS(rs, 1)]; { E Te, Tf, Tb, Tc; Te = cr[WS(rs, 11)]; Tf = ci[WS(rs, 8)]; Tg = Te + Tf; T2e = Te - Tf; Tb = cr[WS(rs, 6)]; Tc = ci[WS(rs, 3)]; Td = Tb + Tc; T2d = Tb - Tc; } T27 = KP559016994 * (Td - Tg); T53 = FMA(KP951056516, T2d, KP587785252 * T2e); T2f = FNMS(KP951056516, T2e, KP587785252 * T2d); Th = Td + Tg; T26 = FNMS(KP250000000, Th, Ta); } { E T16, T2w, T13, T2v; T10 = ci[WS(rs, 20)]; { E T14, T15, T11, T12; T14 = cr[WS(rs, 14)]; T15 = cr[WS(rs, 19)]; T16 = T14 + T15; T2w = T15 - T14; T11 = ci[WS(rs, 15)]; T12 = cr[WS(rs, 24)]; T13 = T11 - T12; T2v = T11 + T12; } T2p = KP559016994 * (T13 + T16); T58 = FMA(KP951056516, T2v, KP587785252 * T2w); T2x = FNMS(KP951056516, T2w, KP587785252 * T2v); T17 = T13 - T16; T2o = FNMS(KP250000000, T17, T10); } { E Tp, T2m, Tm, T2l; Tj = cr[WS(rs, 4)]; { E Tn, To, Tk, Tl; Tn = ci[WS(rs, 10)]; To = ci[WS(rs, 5)]; Tp = Tn + To; T2m = Tn - To; Tk = cr[WS(rs, 9)]; Tl = ci[0]; Tm = Tk + Tl; T2l = Tk - Tl; } T2n = FNMS(KP951056516, T2m, KP587785252 * T2l); T5a = FMA(KP951056516, T2l, KP587785252 * T2m); T2t = KP559016994 * (Tm - Tp); Tq = Tm + Tp; T2s = FNMS(KP250000000, Tq, Tj); } { E TX, T2a, TU, T29; TR = ci[WS(rs, 23)]; { E TV, TW, TS, TT; TV = ci[WS(rs, 13)]; TW = cr[WS(rs, 16)]; TX = TV - TW; T2a = TV + TW; TS = ci[WS(rs, 18)]; TT = cr[WS(rs, 21)]; TU = TS - TT; T29 = TS + TT; } T2b = FNMS(KP951056516, T2a, KP587785252 * T29); T51 = FMA(KP951056516, T29, KP587785252 * T2a); T2h = KP559016994 * (TU - TX); TY = TU + TX; T2g = FNMS(KP250000000, TY, TR); } Ti = Ta + Th; Tr = Tj + Tq; Ts = Ti + Tr; TZ = TR + TY; T18 = T10 + T17; T1z = TZ + T18; { E T2c, T42, T2j, T43, T28, T2i; T28 = T26 - T27; T2c = T28 - T2b; T42 = T28 + T2b; T2i = T2g - T2h; T2j = T2f + T2i; T43 = T2i - T2f; T2k = FNMS(KP481753674, T2j, KP876306680 * T2c); T4l = FMA(KP728968627, T43, KP684547105 * T42); T3h = FMA(KP876306680, T2j, KP481753674 * T2c); T44 = FNMS(KP684547105, T43, KP728968627 * T42); } { E T59, T6n, T5c, T6m, T57, T5b; T57 = T2t + T2s; T59 = T57 - T58; T6n = T57 + T58; T5b = T2o + T2p; T5c = T5a + T5b; T6m = T5b - T5a; T5d = FNMS(KP844327925, T5c, KP535826794 * T59); T6C = FMA(KP637423989, T6m, KP770513242 * T6n); T5C = FMA(KP535826794, T5c, KP844327925 * T59); T6o = FNMS(KP637423989, T6n, KP770513242 * T6m); } { E T52, T6j, T55, T6k, T50, T54; T50 = T27 + T26; T52 = T50 - T51; T6j = T50 + T51; T54 = T2h + T2g; T55 = T53 + T54; T6k = T54 - T53; T56 = FNMS(KP248689887, T55, KP968583161 * T52); T6B = FMA(KP535826794, T6k, KP844327925 * T6j); T5B = FMA(KP968583161, T55, KP248689887 * T52); T6l = FNMS(KP844327925, T6k, KP535826794 * T6j); } { E T2r, T45, T2y, T46, T2q, T2u; T2q = T2o - T2p; T2r = T2n + T2q; T45 = T2q - T2n; T2u = T2s - T2t; T2y = T2u - T2x; T46 = T2u + T2x; T2z = FMA(KP904827052, T2r, KP425779291 * T2y); T4m = FNMS(KP992114701, T45, KP125333233 * T46); T3i = FNMS(KP425779291, T2r, KP904827052 * T2y); T47 = FMA(KP125333233, T45, KP992114701 * T46); } } { E T1C, T1F, T1I, T1J, T3b, T3a, T3z, T3A; T1C = ci[WS(rs, 24)]; { E T1D, T1E, T1G, T1H; T1D = ci[WS(rs, 19)]; T1E = cr[WS(rs, 20)]; T1F = T1D - T1E; T1G = ci[WS(rs, 14)]; T1H = cr[WS(rs, 15)]; T1I = T1G - T1H; T1J = T1F + T1I; T3b = T1G + T1H; T3a = T1D + T1E; } T1K = T1C + T1J; T5w = FMA(KP951056516, T3a, KP587785252 * T3b); T3c = FNMS(KP951056516, T3b, KP587785252 * T3a); T3z = FNMS(KP250000000, T1J, T1C); T3A = KP559016994 * (T1F - T1I); T3B = T3z - T3A; T5R = T3A + T3z; } { E Tt, T2C, T5i, T2K, TA, T2B, T1a, T2G, T5g, T2M, T1h, T2L, TC, T2R, T5p; E T2Z, TJ, T2Q, T1j, T2V, T5n, T31, T1q, T30; { E Tw, T2I, Tz, T2J; Tt = cr[WS(rs, 2)]; { E Tu, Tv, Tx, Ty; Tu = cr[WS(rs, 7)]; Tv = ci[WS(rs, 2)]; Tw = Tu + Tv; T2I = Tu - Tv; Tx = cr[WS(rs, 12)]; Ty = ci[WS(rs, 7)]; Tz = Tx + Ty; T2J = Tx - Ty; } T2C = KP559016994 * (Tw - Tz); T5i = FMA(KP951056516, T2I, KP587785252 * T2J); T2K = FNMS(KP951056516, T2J, KP587785252 * T2I); TA = Tw + Tz; T2B = FNMS(KP250000000, TA, Tt); } { E T1d, T2E, T1g, T2F; T1a = ci[WS(rs, 22)]; { E T1b, T1c, T1e, T1f; T1b = ci[WS(rs, 17)]; T1c = cr[WS(rs, 22)]; T1d = T1b - T1c; T2E = T1b + T1c; T1e = ci[WS(rs, 12)]; T1f = cr[WS(rs, 17)]; T1g = T1e - T1f; T2F = T1e + T1f; } T2G = FNMS(KP951056516, T2F, KP587785252 * T2E); T5g = FMA(KP951056516, T2E, KP587785252 * T2F); T2M = KP559016994 * (T1d - T1g); T1h = T1d + T1g; T2L = FNMS(KP250000000, T1h, T1a); } { E TI, T2Y, TF, T2X; TC = cr[WS(rs, 3)]; { E TG, TH, TD, TE; TG = ci[WS(rs, 11)]; TH = ci[WS(rs, 6)]; TI = TG + TH; T2Y = TG - TH; TD = cr[WS(rs, 8)]; TE = ci[WS(rs, 1)]; TF = TD + TE; T2X = TD - TE; } T2R = KP559016994 * (TF - TI); T5p = FMA(KP951056516, T2X, KP587785252 * T2Y); T2Z = FNMS(KP951056516, T2Y, KP587785252 * T2X); TJ = TF + TI; T2Q = FNMS(KP250000000, TJ, TC); } { E T1p, T2U, T1m, T2T; T1j = ci[WS(rs, 21)]; { E T1n, T1o, T1k, T1l; T1n = cr[WS(rs, 13)]; T1o = cr[WS(rs, 18)]; T1p = T1n + T1o; T2U = T1o - T1n; T1k = ci[WS(rs, 16)]; T1l = cr[WS(rs, 23)]; T1m = T1k - T1l; T2T = T1k + T1l; } T2V = FNMS(KP951056516, T2U, KP587785252 * T2T); T5n = FMA(KP951056516, T2T, KP587785252 * T2U); T31 = KP559016994 * (T1m + T1p); T1q = T1m - T1p; T30 = FNMS(KP250000000, T1q, T1j); } TB = Tt + TA; TK = TC + TJ; TL = TB + TK; T1i = T1a + T1h; T1r = T1j + T1q; T1A = T1i + T1r; { E T2H, T49, T2O, T4a, T2D, T2N; T2D = T2B - T2C; T2H = T2D - T2G; T49 = T2D + T2G; T2N = T2L - T2M; T2O = T2K + T2N; T4a = T2N - T2K; T2P = FNMS(KP844327925, T2O, KP535826794 * T2H); T4o = FMA(KP062790519, T4a, KP998026728 * T49); T3k = FMA(KP535826794, T2O, KP844327925 * T2H); T4b = FNMS(KP998026728, T4a, KP062790519 * T49); } { E T5o, T6u, T5r, T6t, T5m, T5q; T5m = T2R + T2Q; T5o = T5m - T5n; T6u = T5m + T5n; T5q = T30 + T31; T5r = T5p + T5q; T6t = T5q - T5p; T5s = FNMS(KP684547105, T5r, KP728968627 * T5o); T6F = FNMS(KP992114701, T6t, KP125333233 * T6u); T5F = FMA(KP728968627, T5r, KP684547105 * T5o); T6v = FMA(KP125333233, T6t, KP992114701 * T6u); } { E T5h, T6r, T5k, T6q, T5f, T5j; T5f = T2C + T2B; T5h = T5f - T5g; T6r = T5f + T5g; T5j = T2M + T2L; T5k = T5i + T5j; T6q = T5j - T5i; T5l = FNMS(KP481753674, T5k, KP876306680 * T5h); T6E = FNMS(KP425779291, T6q, KP904827052 * T6r); T5E = FMA(KP876306680, T5k, KP481753674 * T5h); T6s = FMA(KP904827052, T6q, KP425779291 * T6r); } { E T2W, T4d, T33, T4c, T2S, T32; T2S = T2Q - T2R; T2W = T2S - T2V; T4d = T2S + T2V; T32 = T30 - T31; T33 = T2Z + T32; T4c = T32 - T2Z; T34 = FNMS(KP998026728, T33, KP062790519 * T2W); T4p = FNMS(KP637423989, T4c, KP770513242 * T4d); T3l = FMA(KP062790519, T33, KP998026728 * T2W); T4e = FMA(KP770513242, T4c, KP637423989 * T4d); } } { E TM, TQ, T1U, T1L, T1N, T1Z, T1t, T1V, T1y, T1Y; { E TO, TP, T1B, T1M; TO = KP559016994 * (Ts - TL); TM = Ts + TL; TP = FNMS(KP250000000, TM, T9); TQ = TO + TP; T1U = TP - TO; T1B = KP559016994 * (T1z - T1A); T1L = T1z + T1A; T1M = FNMS(KP250000000, T1L, T1K); T1N = T1B + T1M; T1Z = T1M - T1B; } { E T19, T1s, T1w, T1x; T19 = TZ - T18; T1s = T1i - T1r; T1t = FMA(KP951056516, T19, KP587785252 * T1s); T1V = FNMS(KP951056516, T1s, KP587785252 * T19); T1w = Ti - Tr; T1x = TB - TK; T1y = FMA(KP951056516, T1w, KP587785252 * T1x); T1Y = FNMS(KP951056516, T1x, KP587785252 * T1w); } cr[0] = T9 + TM; ci[0] = T1K + T1L; { E T1u, T1O, TN, T1v; T1u = TQ - T1t; T1O = T1y + T1N; TN = W[8]; T1v = W[9]; cr[WS(rs, 5)] = FNMS(T1v, T1O, TN * T1u); ci[WS(rs, 5)] = FMA(T1v, T1u, TN * T1O); } { E T22, T24, T21, T23; T22 = T1U + T1V; T24 = T1Z - T1Y; T21 = W[28]; T23 = W[29]; cr[WS(rs, 15)] = FNMS(T23, T24, T21 * T22); ci[WS(rs, 15)] = FMA(T23, T22, T21 * T24); } { E T1W, T20, T1T, T1X; T1W = T1U - T1V; T20 = T1Y + T1Z; T1T = W[18]; T1X = W[19]; cr[WS(rs, 10)] = FNMS(T1X, T20, T1T * T1W); ci[WS(rs, 10)] = FMA(T1X, T1W, T1T * T20); } { E T1Q, T1S, T1P, T1R; T1Q = TQ + T1t; T1S = T1N - T1y; T1P = W[38]; T1R = W[39]; cr[WS(rs, 20)] = FNMS(T1R, T1S, T1P * T1Q); ci[WS(rs, 20)] = FMA(T1R, T1Q, T1P * T1S); } } { E T6H, T71, T6M, T74, T6i, T6x, T6y, T6z, T6Q, T6R, T6P, T6S; { E T6D, T6G, T6K, T6L; T6D = T6B + T6C; T6G = T6E - T6F; T6H = FMA(KP951056516, T6D, KP587785252 * T6G); T71 = FNMS(KP951056516, T6G, KP587785252 * T6D); T6K = T6l - T6o; T6L = T6v - T6s; T6M = FMA(KP951056516, T6K, KP587785252 * T6L); T74 = FNMS(KP951056516, T6L, KP587785252 * T6K); } { E T6p, T6w, T6N, T6O; T6i = T5v + T5w; T6p = T6l + T6o; T6w = T6s + T6v; T6x = T6p - T6w; T6y = FNMS(KP250000000, T6x, T6i); T6z = KP559016994 * (T6p + T6w); T6Q = T5R - T5Q; T6N = T6B - T6C; T6O = T6E + T6F; T6R = T6N + T6O; T6P = KP559016994 * (T6N - T6O); T6S = FNMS(KP250000000, T6R, T6Q); } { E T7c, T7e, T7b, T7d; T7c = T6i + T6x; T7e = T6Q + T6R; T7b = W[6]; T7d = W[7]; cr[WS(rs, 4)] = FNMS(T7d, T7e, T7b * T7c); ci[WS(rs, 4)] = FMA(T7d, T7c, T7b * T7e); } { E T72, T78, T76, T7a, T70, T75; T70 = T6y - T6z; T72 = T70 - T71; T78 = T70 + T71; T75 = T6S - T6P; T76 = T74 + T75; T7a = T75 - T74; { E T6Z, T73, T77, T79; T6Z = W[26]; T73 = W[27]; cr[WS(rs, 14)] = FNMS(T73, T76, T6Z * T72); ci[WS(rs, 14)] = FMA(T73, T72, T6Z * T76); T77 = W[36]; T79 = W[37]; cr[WS(rs, 19)] = FNMS(T79, T7a, T77 * T78); ci[WS(rs, 19)] = FMA(T79, T78, T77 * T7a); } } { E T6I, T6W, T6U, T6Y, T6A, T6T; T6A = T6y + T6z; T6I = T6A - T6H; T6W = T6A + T6H; T6T = T6P + T6S; T6U = T6M + T6T; T6Y = T6T - T6M; { E T6h, T6J, T6V, T6X; T6h = W[16]; T6J = W[17]; cr[WS(rs, 9)] = FNMS(T6J, T6U, T6h * T6I); ci[WS(rs, 9)] = FMA(T6J, T6I, T6h * T6U); T6V = W[46]; T6X = W[47]; cr[WS(rs, 24)] = FNMS(T6X, T6Y, T6V * T6W); ci[WS(rs, 24)] = FMA(T6X, T6W, T6V * T6Y); } } } { E T3n, T3N, T3s, T3Q, T3d, T3e, T36, T3f, T3C, T3D, T3v, T3E; { E T3j, T3m, T3q, T3r; T3j = T3h - T3i; T3m = T3k - T3l; T3n = FMA(KP951056516, T3j, KP587785252 * T3m); T3N = FNMS(KP951056516, T3m, KP587785252 * T3j); T3q = T2k + T2z; T3r = T2P - T34; T3s = FMA(KP951056516, T3q, KP587785252 * T3r); T3Q = FNMS(KP951056516, T3r, KP587785252 * T3q); } { E T2A, T35, T3t, T3u; T3d = T39 - T3c; T2A = T2k - T2z; T35 = T2P + T34; T3e = T2A + T35; T36 = KP559016994 * (T2A - T35); T3f = FNMS(KP250000000, T3e, T3d); T3C = T3y + T3B; T3t = T3h + T3i; T3u = T3k + T3l; T3D = T3t + T3u; T3v = KP559016994 * (T3t - T3u); T3E = FNMS(KP250000000, T3D, T3C); } { E T3Y, T40, T3X, T3Z; T3Y = T3d + T3e; T40 = T3C + T3D; T3X = W[2]; T3Z = W[3]; cr[WS(rs, 2)] = FNMS(T3Z, T40, T3X * T3Y); ci[WS(rs, 2)] = FMA(T3Z, T3Y, T3X * T40); } { E T3O, T3U, T3S, T3W, T3M, T3R; T3M = T3f - T36; T3O = T3M - T3N; T3U = T3M + T3N; T3R = T3E - T3v; T3S = T3Q + T3R; T3W = T3R - T3Q; { E T3L, T3P, T3T, T3V; T3L = W[22]; T3P = W[23]; cr[WS(rs, 12)] = FNMS(T3P, T3S, T3L * T3O); ci[WS(rs, 12)] = FMA(T3P, T3O, T3L * T3S); T3T = W[32]; T3V = W[33]; cr[WS(rs, 17)] = FNMS(T3V, T3W, T3T * T3U); ci[WS(rs, 17)] = FMA(T3V, T3U, T3T * T3W); } } { E T3o, T3I, T3G, T3K, T3g, T3F; T3g = T36 + T3f; T3o = T3g - T3n; T3I = T3g + T3n; T3F = T3v + T3E; T3G = T3s + T3F; T3K = T3F - T3s; { E T25, T3p, T3H, T3J; T25 = W[12]; T3p = W[13]; cr[WS(rs, 7)] = FNMS(T3p, T3G, T25 * T3o); ci[WS(rs, 7)] = FMA(T3p, T3o, T25 * T3G); T3H = W[42]; T3J = W[43]; cr[WS(rs, 22)] = FNMS(T3J, T3K, T3H * T3I); ci[WS(rs, 22)] = FMA(T3J, T3I, T3H * T3K); } } } { E T4r, T4L, T4w, T4O, T4h, T4i, T4g, T4j, T4A, T4B, T4z, T4C; { E T4n, T4q, T4u, T4v; T4n = T4l - T4m; T4q = T4o - T4p; T4r = FMA(KP951056516, T4n, KP587785252 * T4q); T4L = FNMS(KP951056516, T4q, KP587785252 * T4n); T4u = T44 + T47; T4v = T4b + T4e; T4w = FMA(KP951056516, T4u, KP587785252 * T4v); T4O = FNMS(KP951056516, T4v, KP587785252 * T4u); } { E T48, T4f, T4x, T4y; T4h = T39 + T3c; T48 = T44 - T47; T4f = T4b - T4e; T4i = T48 + T4f; T4g = KP559016994 * (T48 - T4f); T4j = FNMS(KP250000000, T4i, T4h); T4A = T3B - T3y; T4x = T4l + T4m; T4y = T4o + T4p; T4B = T4x + T4y; T4z = KP559016994 * (T4x - T4y); T4C = FNMS(KP250000000, T4B, T4A); } { E T4W, T4Y, T4V, T4X; T4W = T4h + T4i; T4Y = T4A + T4B; T4V = W[4]; T4X = W[5]; cr[WS(rs, 3)] = FNMS(T4X, T4Y, T4V * T4W); ci[WS(rs, 3)] = FMA(T4X, T4W, T4V * T4Y); } { E T4M, T4S, T4Q, T4U, T4K, T4P; T4K = T4j - T4g; T4M = T4K - T4L; T4S = T4K + T4L; T4P = T4C - T4z; T4Q = T4O + T4P; T4U = T4P - T4O; { E T4J, T4N, T4R, T4T; T4J = W[24]; T4N = W[25]; cr[WS(rs, 13)] = FNMS(T4N, T4Q, T4J * T4M); ci[WS(rs, 13)] = FMA(T4N, T4M, T4J * T4Q); T4R = W[34]; T4T = W[35]; cr[WS(rs, 18)] = FNMS(T4T, T4U, T4R * T4S); ci[WS(rs, 18)] = FMA(T4T, T4S, T4R * T4U); } } { E T4s, T4G, T4E, T4I, T4k, T4D; T4k = T4g + T4j; T4s = T4k - T4r; T4G = T4k + T4r; T4D = T4z + T4C; T4E = T4w + T4D; T4I = T4D - T4w; { E T41, T4t, T4F, T4H; T41 = W[14]; T4t = W[15]; cr[WS(rs, 8)] = FNMS(T4t, T4E, T41 * T4s); ci[WS(rs, 8)] = FMA(T4t, T4s, T41 * T4E); T4F = W[44]; T4H = W[45]; cr[WS(rs, 23)] = FNMS(T4H, T4I, T4F * T4G); ci[WS(rs, 23)] = FMA(T4H, T4G, T4F * T4I); } } } { E T5H, T63, T5M, T66, T5x, T5y, T5u, T5z, T5S, T5T, T5P, T5U; { E T5D, T5G, T5K, T5L; T5D = T5B - T5C; T5G = T5E - T5F; T5H = FMA(KP951056516, T5D, KP587785252 * T5G); T63 = FNMS(KP951056516, T5G, KP587785252 * T5D); T5K = T56 - T5d; T5L = T5l - T5s; T5M = FMA(KP951056516, T5K, KP587785252 * T5L); T66 = FNMS(KP951056516, T5L, KP587785252 * T5K); } { E T5e, T5t, T5N, T5O; T5x = T5v - T5w; T5e = T56 + T5d; T5t = T5l + T5s; T5y = T5e + T5t; T5u = KP559016994 * (T5e - T5t); T5z = FNMS(KP250000000, T5y, T5x); T5S = T5Q + T5R; T5N = T5B + T5C; T5O = T5E + T5F; T5T = T5N + T5O; T5P = KP559016994 * (T5N - T5O); T5U = FNMS(KP250000000, T5T, T5S); } { E T6e, T6g, T6d, T6f; T6e = T5x + T5y; T6g = T5S + T5T; T6d = W[0]; T6f = W[1]; cr[WS(rs, 1)] = FNMS(T6f, T6g, T6d * T6e); ci[WS(rs, 1)] = FMA(T6f, T6e, T6d * T6g); } { E T64, T6a, T68, T6c, T62, T67; T62 = T5z - T5u; T64 = T62 - T63; T6a = T62 + T63; T67 = T5U - T5P; T68 = T66 + T67; T6c = T67 - T66; { E T61, T65, T69, T6b; T61 = W[20]; T65 = W[21]; cr[WS(rs, 11)] = FNMS(T65, T68, T61 * T64); ci[WS(rs, 11)] = FMA(T65, T64, T61 * T68); T69 = W[30]; T6b = W[31]; cr[WS(rs, 16)] = FNMS(T6b, T6c, T69 * T6a); ci[WS(rs, 16)] = FMA(T6b, T6a, T69 * T6c); } } { E T5I, T5Y, T5W, T60, T5A, T5V; T5A = T5u + T5z; T5I = T5A - T5H; T5Y = T5A + T5H; T5V = T5P + T5U; T5W = T5M + T5V; T60 = T5V - T5M; { E T4Z, T5J, T5X, T5Z; T4Z = W[10]; T5J = W[11]; cr[WS(rs, 6)] = FNMS(T5J, T5W, T4Z * T5I); ci[WS(rs, 6)] = FMA(T5J, T5I, T4Z * T5W); T5X = W[40]; T5Z = W[41]; cr[WS(rs, 21)] = FNMS(T5Z, T60, T5X * T5Y); ci[WS(rs, 21)] = FMA(T5Z, T5Y, T5X * T60); } } } } } } static const tw_instr twinstr[] = { {TW_FULL, 1, 25}, {TW_NEXT, 1, 0} }; static const hc2hc_desc desc = { 25, "hb_25", twinstr, &GENUS, {260, 140, 140, 0} }; void X(codelet_hb_25) (planner *p) { X(khc2hc_register) (p, hb_25, &desc); } #endif /* HAVE_FMA */ fftw-3.3.3/rdft/scalar/r2cb/hb2_4.c0000644000175000017500000001242112054410763013507 00000000000000/* * Copyright (c) 2003, 2007-11 Matteo Frigo * Copyright (c) 2003, 2007-11 Massachusetts Institute of Technology * * This program is free software; you can redistribute it and/or modify * it under the terms of the GNU General Public License as published by * the Free Software Foundation; either version 2 of the License, or * (at your option) any later version. * * This program is distributed in the hope that it will be useful, * but WITHOUT ANY WARRANTY; without even the implied warranty of * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the * GNU General Public License for more details. * * You should have received a copy of the GNU General Public License * along with this program; if not, write to the Free Software * Foundation, Inc., 51 Franklin Street, Fifth Floor, Boston, MA 02110-1301 USA * */ /* This file was automatically generated --- DO NOT EDIT */ /* Generated on Sun Nov 25 07:41:23 EST 2012 */ #include "codelet-rdft.h" #ifdef HAVE_FMA /* Generated by: ../../../genfft/gen_hc2hc.native -fma -reorder-insns -schedule-for-pipeline -compact -variables 4 -pipeline-latency 4 -sign 1 -twiddle-log3 -precompute-twiddles -n 4 -dif -name hb2_4 -include hb.h */ /* * This function contains 24 FP additions, 16 FP multiplications, * (or, 16 additions, 8 multiplications, 8 fused multiply/add), * 33 stack variables, 0 constants, and 16 memory accesses */ #include "hb.h" static void hb2_4(R *cr, R *ci, const R *W, stride rs, INT mb, INT me, INT ms) { { INT m; for (m = mb, W = W + ((mb - 1) * 4); m < me; m = m + 1, cr = cr + ms, ci = ci - ms, W = W + 4, MAKE_VOLATILE_STRIDE(8, rs)) { E Tg, Tc, Te, To, Tn; { E T7, Tb, T8, Ta; T7 = W[0]; Tb = W[3]; T8 = W[2]; Ta = W[1]; { E Tj, Tm, T3, T6, Tx, Tr, Tz, Tv, Td; { E Tu, T4, Tq, T5, Tp, Tt; { E Tk, Tl, T1, T2; { E Th, Tf, T9, Ti; Th = ci[WS(rs, 3)]; Tf = T7 * Tb; T9 = T7 * T8; Ti = cr[WS(rs, 2)]; Tk = ci[WS(rs, 2)]; Tg = FNMS(Ta, T8, Tf); Tc = FMA(Ta, Tb, T9); Tu = Th + Ti; Tj = Th - Ti; Tl = cr[WS(rs, 3)]; } T1 = cr[0]; T2 = ci[WS(rs, 1)]; T4 = cr[WS(rs, 1)]; Tm = Tk - Tl; Tq = Tk + Tl; T5 = ci[0]; T3 = T1 + T2; Tp = T1 - T2; } Tt = T4 - T5; T6 = T4 + T5; Tx = Tp + Tq; Tr = Tp - Tq; Tz = Tu - Tt; Tv = Tt + Tu; Td = T3 - T6; } { E Ts, Tw, TA, Ty; cr[0] = T3 + T6; Ts = T7 * Tr; ci[0] = Tj + Tm; Tw = T7 * Tv; TA = T8 * Tz; cr[WS(rs, 1)] = FNMS(Ta, Tv, Ts); Ty = T8 * Tx; ci[WS(rs, 1)] = FMA(Ta, Tr, Tw); ci[WS(rs, 3)] = FMA(Tb, Tx, TA); Te = Tc * Td; cr[WS(rs, 3)] = FNMS(Tb, Tz, Ty); To = Tg * Td; Tn = Tj - Tm; } } } ci[WS(rs, 2)] = FMA(Tc, Tn, To); cr[WS(rs, 2)] = FNMS(Tg, Tn, Te); } } } static const tw_instr twinstr[] = { {TW_CEXP, 1, 1}, {TW_CEXP, 1, 3}, {TW_NEXT, 1, 0} }; static const hc2hc_desc desc = { 4, "hb2_4", twinstr, &GENUS, {16, 8, 8, 0} }; void X(codelet_hb2_4) (planner *p) { X(khc2hc_register) (p, hb2_4, &desc); } #else /* HAVE_FMA */ /* Generated by: ../../../genfft/gen_hc2hc.native -compact -variables 4 -pipeline-latency 4 -sign 1 -twiddle-log3 -precompute-twiddles -n 4 -dif -name hb2_4 -include hb.h */ /* * This function contains 24 FP additions, 16 FP multiplications, * (or, 16 additions, 8 multiplications, 8 fused multiply/add), * 21 stack variables, 0 constants, and 16 memory accesses */ #include "hb.h" static void hb2_4(R *cr, R *ci, const R *W, stride rs, INT mb, INT me, INT ms) { { INT m; for (m = mb, W = W + ((mb - 1) * 4); m < me; m = m + 1, cr = cr + ms, ci = ci - ms, W = W + 4, MAKE_VOLATILE_STRIDE(8, rs)) { E T7, T9, T8, Ta, Tb, Td; T7 = W[0]; T9 = W[1]; T8 = W[2]; Ta = W[3]; Tb = FMA(T7, T8, T9 * Ta); Td = FNMS(T9, T8, T7 * Ta); { E T3, Tl, T6, To, Tg, Tp, Tj, Tm, Tc, Tk; { E T1, T2, T4, T5; T1 = cr[0]; T2 = ci[WS(rs, 1)]; T3 = T1 + T2; Tl = T1 - T2; T4 = cr[WS(rs, 1)]; T5 = ci[0]; T6 = T4 + T5; To = T4 - T5; } { E Te, Tf, Th, Ti; Te = ci[WS(rs, 3)]; Tf = cr[WS(rs, 2)]; Tg = Te - Tf; Tp = Te + Tf; Th = ci[WS(rs, 2)]; Ti = cr[WS(rs, 3)]; Tj = Th - Ti; Tm = Th + Ti; } cr[0] = T3 + T6; ci[0] = Tg + Tj; Tc = T3 - T6; Tk = Tg - Tj; cr[WS(rs, 2)] = FNMS(Td, Tk, Tb * Tc); ci[WS(rs, 2)] = FMA(Td, Tc, Tb * Tk); { E Tn, Tq, Tr, Ts; Tn = Tl - Tm; Tq = To + Tp; cr[WS(rs, 1)] = FNMS(T9, Tq, T7 * Tn); ci[WS(rs, 1)] = FMA(T7, Tq, T9 * Tn); Tr = Tl + Tm; Ts = Tp - To; cr[WS(rs, 3)] = FNMS(Ta, Ts, T8 * Tr); ci[WS(rs, 3)] = FMA(T8, Ts, Ta * Tr); } } } } } static const tw_instr twinstr[] = { {TW_CEXP, 1, 1}, {TW_CEXP, 1, 3}, {TW_NEXT, 1, 0} }; static const hc2hc_desc desc = { 4, "hb2_4", twinstr, &GENUS, {16, 8, 8, 0} }; void X(codelet_hb2_4) (planner *p) { X(khc2hc_register) (p, hb2_4, &desc); } #endif /* HAVE_FMA */ fftw-3.3.3/rdft/scalar/r2cb/hb2_8.c0000644000175000017500000002425212054410764013521 00000000000000/* * Copyright (c) 2003, 2007-11 Matteo Frigo * Copyright (c) 2003, 2007-11 Massachusetts Institute of Technology * * This program is free software; you can redistribute it and/or modify * it under the terms of the GNU General Public License as published by * the Free Software Foundation; either version 2 of the License, or * (at your option) any later version. * * This program is distributed in the hope that it will be useful, * but WITHOUT ANY WARRANTY; without even the implied warranty of * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the * GNU General Public License for more details. * * You should have received a copy of the GNU General Public License * along with this program; if not, write to the Free Software * Foundation, Inc., 51 Franklin Street, Fifth Floor, Boston, MA 02110-1301 USA * */ /* This file was automatically generated --- DO NOT EDIT */ /* Generated on Sun Nov 25 07:41:24 EST 2012 */ #include "codelet-rdft.h" #ifdef HAVE_FMA /* Generated by: ../../../genfft/gen_hc2hc.native -fma -reorder-insns -schedule-for-pipeline -compact -variables 4 -pipeline-latency 4 -sign 1 -twiddle-log3 -precompute-twiddles -n 8 -dif -name hb2_8 -include hb.h */ /* * This function contains 74 FP additions, 50 FP multiplications, * (or, 44 additions, 20 multiplications, 30 fused multiply/add), * 77 stack variables, 1 constants, and 32 memory accesses */ #include "hb.h" static void hb2_8(R *cr, R *ci, const R *W, stride rs, INT mb, INT me, INT ms) { DK(KP707106781, +0.707106781186547524400844362104849039284835938); { INT m; for (m = mb, W = W + ((mb - 1) * 6); m < me; m = m + 1, cr = cr + ms, ci = ci - ms, W = W + 6, MAKE_VOLATILE_STRIDE(16, rs)) { E Tf, Tg, Tl, Tp, Ti, Tj, T1o, T1u, Tk, T1b, To, T1e, TK, Tq, T13; E TP, T1p, T7, T1h, T1v, TZ, Tv, Tw, Ta, Tx, T1j, TE, TB, Td, Ty; E Th, T1n, T1t; Tf = W[0]; Tg = W[2]; Tl = W[4]; Tp = W[5]; Ti = W[1]; Th = Tf * Tg; T1n = Tf * Tl; T1t = Tf * Tp; Tj = W[3]; { E Tr, T3, Ts, T1f, TO, TL, T6, Tt; { E TM, TN, T4, T5; { E T1, Tn, T2, TJ, Tm; T1 = cr[0]; T1o = FMA(Ti, Tp, T1n); T1u = FNMS(Ti, Tl, T1t); Tk = FMA(Ti, Tj, Th); T1b = FNMS(Ti, Tj, Th); Tn = Tf * Tj; T2 = ci[WS(rs, 3)]; TM = ci[WS(rs, 7)]; TJ = Tk * Tp; Tm = Tk * Tl; To = FNMS(Ti, Tg, Tn); T1e = FMA(Ti, Tg, Tn); Tr = T1 - T2; T3 = T1 + T2; TK = FNMS(To, Tl, TJ); Tq = FMA(To, Tp, Tm); TN = cr[WS(rs, 4)]; } T4 = cr[WS(rs, 2)]; T5 = ci[WS(rs, 1)]; Ts = ci[WS(rs, 5)]; T1f = TM - TN; TO = TM + TN; TL = T4 - T5; T6 = T4 + T5; Tt = cr[WS(rs, 6)]; } { E TC, TD, Tb, Tc; { E T8, T1g, Tu, T9; T8 = cr[WS(rs, 1)]; T13 = TO - TL; TP = TL + TO; T1p = T3 - T6; T7 = T3 + T6; T1g = Ts - Tt; Tu = Ts + Tt; T9 = ci[WS(rs, 2)]; TC = ci[WS(rs, 4)]; T1h = T1f + T1g; T1v = T1f - T1g; TZ = Tr + Tu; Tv = Tr - Tu; Tw = T8 - T9; Ta = T8 + T9; TD = cr[WS(rs, 7)]; } Tb = ci[0]; Tc = cr[WS(rs, 3)]; Tx = ci[WS(rs, 6)]; T1j = TC - TD; TE = TC + TD; TB = Tb - Tc; Td = Tb + Tc; Ty = cr[WS(rs, 5)]; } } { E TR, TF, Te, T1w; TR = TB + TE; TF = TB - TE; Te = Ta + Td; T1w = Ta - Td; { E Tz, T1i, T1B, T1x, T1c; Tz = Tx + Ty; T1i = Tx - Ty; T1B = T1w + T1v; T1x = T1v - T1w; T1c = T7 - Te; cr[0] = T7 + Te; { E T1k, T1q, TQ, TA; T1k = T1i + T1j; T1q = T1j - T1i; TQ = Tw + Tz; TA = Tw - Tz; { E T1y, T1C, T1m, T1d; T1y = T1o * T1x; T1C = Tk * T1B; T1m = T1e * T1c; T1d = T1b * T1c; { E T1z, T1r, T1l, TG, T14; T1z = T1p + T1q; T1r = T1p - T1q; T1l = T1h - T1k; ci[0] = T1h + T1k; TG = TA + TF; T14 = TA - TF; { E T10, TS, T1s, T1A; T10 = TQ + TR; TS = TQ - TR; ci[WS(rs, 6)] = FMA(T1u, T1r, T1y); T1s = T1o * T1r; ci[WS(rs, 2)] = FMA(To, T1z, T1C); T1A = Tk * T1z; ci[WS(rs, 4)] = FMA(T1b, T1l, T1m); cr[WS(rs, 4)] = FNMS(T1e, T1l, T1d); { E T15, T19, TV, TH; T15 = FMA(KP707106781, T14, T13); T19 = FNMS(KP707106781, T14, T13); TV = FMA(KP707106781, TG, Tv); TH = FNMS(KP707106781, TG, Tv); { E TT, TX, T11, T17; TT = FNMS(KP707106781, TS, TP); TX = FMA(KP707106781, TS, TP); T11 = FNMS(KP707106781, T10, TZ); T17 = FMA(KP707106781, T10, TZ); cr[WS(rs, 6)] = FNMS(T1u, T1x, T1s); cr[WS(rs, 2)] = FNMS(To, T1B, T1A); { E T1a, T16, TU, TI; T1a = Tl * T19; T16 = Tg * T15; TU = TK * TH; TI = Tq * TH; { E TY, TW, T18, T12; TY = Ti * TV; TW = Tf * TV; T18 = Tl * T17; T12 = Tg * T11; ci[WS(rs, 7)] = FMA(Tp, T17, T1a); ci[WS(rs, 3)] = FMA(Tj, T11, T16); ci[WS(rs, 5)] = FMA(Tq, TT, TU); cr[WS(rs, 5)] = FNMS(TK, TT, TI); ci[WS(rs, 1)] = FMA(Tf, TX, TY); cr[WS(rs, 1)] = FNMS(Ti, TX, TW); cr[WS(rs, 7)] = FNMS(Tp, T19, T18); cr[WS(rs, 3)] = FNMS(Tj, T15, T12); } } } } } } } } } } } } } static const tw_instr twinstr[] = { {TW_CEXP, 1, 1}, {TW_CEXP, 1, 3}, {TW_CEXP, 1, 7}, {TW_NEXT, 1, 0} }; static const hc2hc_desc desc = { 8, "hb2_8", twinstr, &GENUS, {44, 20, 30, 0} }; void X(codelet_hb2_8) (planner *p) { X(khc2hc_register) (p, hb2_8, &desc); } #else /* HAVE_FMA */ /* Generated by: ../../../genfft/gen_hc2hc.native -compact -variables 4 -pipeline-latency 4 -sign 1 -twiddle-log3 -precompute-twiddles -n 8 -dif -name hb2_8 -include hb.h */ /* * This function contains 74 FP additions, 44 FP multiplications, * (or, 56 additions, 26 multiplications, 18 fused multiply/add), * 46 stack variables, 1 constants, and 32 memory accesses */ #include "hb.h" static void hb2_8(R *cr, R *ci, const R *W, stride rs, INT mb, INT me, INT ms) { DK(KP707106781, +0.707106781186547524400844362104849039284835938); { INT m; for (m = mb, W = W + ((mb - 1) * 6); m < me; m = m + 1, cr = cr + ms, ci = ci - ms, W = W + 6, MAKE_VOLATILE_STRIDE(16, rs)) { E Tf, Ti, Tg, Tj, Tl, Tp, TP, TR, TF, TG, TH, T15, TL, TT; { E Th, To, Tk, Tn; Tf = W[0]; Ti = W[1]; Tg = W[2]; Tj = W[3]; Th = Tf * Tg; To = Ti * Tg; Tk = Ti * Tj; Tn = Tf * Tj; Tl = Th - Tk; Tp = Tn + To; TP = Th + Tk; TR = Tn - To; TF = W[4]; TG = W[5]; TH = FMA(Tf, TF, Ti * TG); T15 = FNMS(TR, TF, TP * TG); TL = FNMS(Ti, TF, Tf * TG); TT = FMA(TP, TF, TR * TG); } { E T7, T1f, T1i, Tw, TI, TW, T18, TM, Te, T19, T1a, TD, TJ, TZ, T12; E TN, Tm, TE; { E T3, TU, Tv, TV, T6, T16, Ts, T17; { E T1, T2, Tt, Tu; T1 = cr[0]; T2 = ci[WS(rs, 3)]; T3 = T1 + T2; TU = T1 - T2; Tt = ci[WS(rs, 5)]; Tu = cr[WS(rs, 6)]; Tv = Tt - Tu; TV = Tt + Tu; } { E T4, T5, Tq, Tr; T4 = cr[WS(rs, 2)]; T5 = ci[WS(rs, 1)]; T6 = T4 + T5; T16 = T4 - T5; Tq = ci[WS(rs, 7)]; Tr = cr[WS(rs, 4)]; Ts = Tq - Tr; T17 = Tq + Tr; } T7 = T3 + T6; T1f = TU + TV; T1i = T17 - T16; Tw = Ts + Tv; TI = T3 - T6; TW = TU - TV; T18 = T16 + T17; TM = Ts - Tv; } { E Ta, TX, TC, T11, Td, T10, Tz, TY; { E T8, T9, TA, TB; T8 = cr[WS(rs, 1)]; T9 = ci[WS(rs, 2)]; Ta = T8 + T9; TX = T8 - T9; TA = ci[WS(rs, 4)]; TB = cr[WS(rs, 7)]; TC = TA - TB; T11 = TA + TB; } { E Tb, Tc, Tx, Ty; Tb = ci[0]; Tc = cr[WS(rs, 3)]; Td = Tb + Tc; T10 = Tb - Tc; Tx = ci[WS(rs, 6)]; Ty = cr[WS(rs, 5)]; Tz = Tx - Ty; TY = Tx + Ty; } Te = Ta + Td; T19 = TX + TY; T1a = T10 + T11; TD = Tz + TC; TJ = TC - Tz; TZ = TX - TY; T12 = T10 - T11; TN = Ta - Td; } cr[0] = T7 + Te; ci[0] = Tw + TD; Tm = T7 - Te; TE = Tw - TD; cr[WS(rs, 4)] = FNMS(Tp, TE, Tl * Tm); ci[WS(rs, 4)] = FMA(Tp, Tm, Tl * TE); { E TQ, TS, TK, TO; TQ = TI + TJ; TS = TN + TM; cr[WS(rs, 2)] = FNMS(TR, TS, TP * TQ); ci[WS(rs, 2)] = FMA(TP, TS, TR * TQ); TK = TI - TJ; TO = TM - TN; cr[WS(rs, 6)] = FNMS(TL, TO, TH * TK); ci[WS(rs, 6)] = FMA(TH, TO, TL * TK); } { E T1h, T1l, T1k, T1m, T1g, T1j; T1g = KP707106781 * (T19 + T1a); T1h = T1f - T1g; T1l = T1f + T1g; T1j = KP707106781 * (TZ - T12); T1k = T1i + T1j; T1m = T1i - T1j; cr[WS(rs, 3)] = FNMS(Tj, T1k, Tg * T1h); ci[WS(rs, 3)] = FMA(Tg, T1k, Tj * T1h); cr[WS(rs, 7)] = FNMS(TG, T1m, TF * T1l); ci[WS(rs, 7)] = FMA(TF, T1m, TG * T1l); } { E T14, T1d, T1c, T1e, T13, T1b; T13 = KP707106781 * (TZ + T12); T14 = TW - T13; T1d = TW + T13; T1b = KP707106781 * (T19 - T1a); T1c = T18 - T1b; T1e = T18 + T1b; cr[WS(rs, 5)] = FNMS(T15, T1c, TT * T14); ci[WS(rs, 5)] = FMA(T15, T14, TT * T1c); cr[WS(rs, 1)] = FNMS(Ti, T1e, Tf * T1d); ci[WS(rs, 1)] = FMA(Ti, T1d, Tf * T1e); } } } } } static const tw_instr twinstr[] = { {TW_CEXP, 1, 1}, {TW_CEXP, 1, 3}, {TW_CEXP, 1, 7}, {TW_NEXT, 1, 0} }; static const hc2hc_desc desc = { 8, "hb2_8", twinstr, &GENUS, {56, 26, 18, 0} }; void X(codelet_hb2_8) (planner *p) { X(khc2hc_register) (p, hb2_8, &desc); } #endif /* HAVE_FMA */ fftw-3.3.3/rdft/scalar/r2cb/hb2_16.c0000644000175000017500000005366012054410765013606 00000000000000/* * Copyright (c) 2003, 2007-11 Matteo Frigo * Copyright (c) 2003, 2007-11 Massachusetts Institute of Technology * * This program is free software; you can redistribute it and/or modify * it under the terms of the GNU General Public License as published by * the Free Software Foundation; either version 2 of the License, or * (at your option) any later version. * * This program is distributed in the hope that it will be useful, * but WITHOUT ANY WARRANTY; without even the implied warranty of * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the * GNU General Public License for more details. * * You should have received a copy of the GNU General Public License * along with this program; if not, write to the Free Software * Foundation, Inc., 51 Franklin Street, Fifth Floor, Boston, MA 02110-1301 USA * */ /* This file was automatically generated --- DO NOT EDIT */ /* Generated on Sun Nov 25 07:41:24 EST 2012 */ #include "codelet-rdft.h" #ifdef HAVE_FMA /* Generated by: ../../../genfft/gen_hc2hc.native -fma -reorder-insns -schedule-for-pipeline -compact -variables 4 -pipeline-latency 4 -sign 1 -twiddle-log3 -precompute-twiddles -n 16 -dif -name hb2_16 -include hb.h */ /* * This function contains 196 FP additions, 134 FP multiplications, * (or, 104 additions, 42 multiplications, 92 fused multiply/add), * 114 stack variables, 3 constants, and 64 memory accesses */ #include "hb.h" static void hb2_16(R *cr, R *ci, const R *W, stride rs, INT mb, INT me, INT ms) { DK(KP923879532, +0.923879532511286756128183189396788286822416626); DK(KP707106781, +0.707106781186547524400844362104849039284835938); DK(KP414213562, +0.414213562373095048801688724209698078569671875); { INT m; for (m = mb, W = W + ((mb - 1) * 8); m < me; m = m + 1, cr = cr + ms, ci = ci - ms, W = W + 8, MAKE_VOLATILE_STRIDE(32, rs)) { E Tv, TB, TF, Ty, T1J, T1O, T1N, T1K; { E Tw, T2z, T2C, Tx, T3f, T3l, T2F, T3r, Tz; Tv = W[0]; Tw = W[2]; T2z = W[6]; T2C = W[7]; TB = W[4]; Tx = Tv * Tw; T3f = Tv * T2z; T3l = Tv * T2C; T2F = Tv * TB; T3r = Tw * TB; TF = W[5]; Ty = W[1]; Tz = W[3]; { E T2G, T3z, T3m, T3g, T3L, T3s, T1V, TA, T3w, T3Q, T30, T3C, TE, T1X, T1D; E TG, T1G, T1o, T2p, T1Y, T2u, T2c, T1Z, TL, T1t, T2d, T35, T3n, T3R, T3F; E T20, T1w, T3M, Tf, T3h, T2L, T2e, TW, T3N, T3I, T2Q, T36, T2V, T37, T1d; E Tu, T3S, T18, T1z, T1i, T24, T2g, T27, T2h, TQ, TV; { E TH, T3, T2I, TU, T32, T1s, T1p, T6, TM, Ta, Tb, T33, TK, T2J, TP; E Tc, T4, T5; { E TS, TT, T1q, T1r; { E T1, T1n, TC, T2b, T1W, T2, T3v, T2Z, TD; T1 = cr[0]; T3v = Tw * TF; T2Z = Tv * TF; T2G = FNMS(Ty, TF, T2F); T3z = FMA(Ty, TF, T2F); T3m = FNMS(Ty, T2z, T3l); T3g = FMA(Ty, T2C, T3f); T3L = FNMS(Tz, TF, T3r); T3s = FMA(Tz, TF, T3r); T1V = FMA(Ty, Tz, Tx); TA = FNMS(Ty, Tz, Tx); TD = Tv * Tz; T3w = FNMS(Tz, TB, T3v); T3Q = FMA(Tz, TB, T3v); T30 = FMA(Ty, TB, T2Z); T3C = FNMS(Ty, TB, T2Z); T1n = TA * TF; TC = TA * TB; T2b = T1V * TF; T1W = T1V * TB; TE = FMA(Ty, Tw, TD); T1X = FNMS(Ty, Tw, TD); T2 = ci[WS(rs, 7)]; TS = ci[WS(rs, 9)]; T1D = FMA(TE, TF, TC); TG = FNMS(TE, TF, TC); T1G = FNMS(TE, TB, T1n); T1o = FMA(TE, TB, T1n); T2p = FMA(T1X, TF, T1W); T1Y = FNMS(T1X, TF, T1W); T2u = FNMS(T1X, TB, T2b); T2c = FMA(T1X, TB, T2b); TH = T1 - T2; T3 = T1 + T2; TT = cr[WS(rs, 14)]; } T1q = ci[WS(rs, 15)]; T1r = cr[WS(rs, 8)]; T4 = cr[WS(rs, 4)]; T2I = TS - TT; TU = TS + TT; T32 = T1q - T1r; T1s = T1q + T1r; T5 = ci[WS(rs, 3)]; } { E TI, TJ, T8, T9, TN, TO; T8 = cr[WS(rs, 2)]; T9 = ci[WS(rs, 5)]; TI = ci[WS(rs, 11)]; T1p = T4 - T5; T6 = T4 + T5; TM = T8 - T9; Ta = T8 + T9; TJ = cr[WS(rs, 12)]; TN = ci[WS(rs, 13)]; TO = cr[WS(rs, 10)]; Tb = ci[WS(rs, 1)]; T33 = TI - TJ; TK = TI + TJ; T2J = TN - TO; TP = TN + TO; Tc = cr[WS(rs, 6)]; } { E TR, Td, T3D, T34; T1Z = TH + TK; TL = TH - TK; T1t = T1p + T1s; T2d = T1s - T1p; TR = Tb - Tc; Td = Tb + Tc; T3D = T32 + T33; T34 = T32 - T33; { E Te, T2K, T1u, T1v, T31, T3E, T2H, T7; Te = Ta + Td; T31 = Ta - Td; T3E = T2J + T2I; T2K = T2I - T2J; TQ = TM - TP; T1u = TM + TP; T1v = TR + TU; TV = TR - TU; T35 = T31 + T34; T3n = T34 - T31; T3R = T3D - T3E; T3F = T3D + T3E; T2H = T3 - T6; T7 = T3 + T6; T20 = T1u + T1v; T1w = T1u - T1v; T3M = T7 - Te; Tf = T7 + Te; T3h = T2H - T2K; T2L = T2H + T2K; } } } { E T1e, Ti, T2N, T1c, T2O, T1h, T19, Tl, T13, Tp, Tq, T2S, T11, T2T, T16; E Tr, Tj, Tk, Tm, TY, Tt; { E T1a, T1b, Tg, Th, T1f, T1g; Tg = cr[WS(rs, 1)]; Th = ci[WS(rs, 6)]; T1a = ci[WS(rs, 14)]; T2e = TQ - TV; TW = TQ + TV; T1e = Tg - Th; Ti = Tg + Th; T1b = cr[WS(rs, 9)]; T1f = ci[WS(rs, 10)]; T1g = cr[WS(rs, 13)]; Tj = cr[WS(rs, 5)]; T2N = T1a - T1b; T1c = T1a + T1b; T2O = T1f - T1g; T1h = T1f + T1g; Tk = ci[WS(rs, 2)]; } { E TZ, T10, Tn, To, T14, T15; Tn = ci[0]; To = cr[WS(rs, 7)]; TZ = ci[WS(rs, 8)]; T19 = Tj - Tk; Tl = Tj + Tk; T13 = Tn - To; Tp = Tn + To; T10 = cr[WS(rs, 15)]; T14 = ci[WS(rs, 12)]; T15 = cr[WS(rs, 11)]; Tq = cr[WS(rs, 3)]; T2S = TZ - T10; T11 = TZ + T10; T2T = T14 - T15; T16 = T14 + T15; Tr = ci[WS(rs, 4)]; } { E T2P, T2U, T2M, Ts, T3G, T3H, T2R; T2P = T2N - T2O; T3G = T2N + T2O; T3H = T2S + T2T; T2U = T2S - T2T; Tm = Ti + Tl; T2M = Ti - Tl; TY = Tq - Tr; Ts = Tq + Tr; T3N = T3H - T3G; T3I = T3G + T3H; Tt = Tp + Ts; T2R = Tp - Ts; T2Q = T2M - T2P; T36 = T2M + T2P; T2V = T2R + T2U; T37 = T2U - T2R; } { E T25, T26, T22, T23, T12, T17; T12 = TY - T11; T25 = TY + T11; T26 = T13 + T16; T17 = T13 - T16; T22 = T1c - T19; T1d = T19 + T1c; Tu = Tm + Tt; T3S = Tm - Tt; T18 = FNMS(KP414213562, T17, T12); T1z = FMA(KP414213562, T12, T17); T1i = T1e - T1h; T23 = T1e + T1h; T24 = FNMS(KP414213562, T23, T22); T2g = FMA(KP414213562, T22, T23); T27 = FNMS(KP414213562, T26, T25); T2h = FMA(KP414213562, T25, T26); } } { E T1j, T1y, T3V, T3X, T3W, T38, T3i, T3o, T2W, T3K, T3B, T3A; cr[0] = Tf + Tu; T3A = Tf - Tu; T1j = FMA(KP414213562, T1i, T1d); T1y = FNMS(KP414213562, T1d, T1i); T3K = T3C * T3A; T3B = T3z * T3A; { E T3O, T3T, T3J, T3P, T3U; T3O = T3M - T3N; T3V = T3M + T3N; T3X = T3S + T3R; T3T = T3R - T3S; ci[0] = T3F + T3I; T3J = T3F - T3I; T3P = T3L * T3O; T3U = T3L * T3T; T3W = TA * T3V; cr[WS(rs, 8)] = FNMS(T3C, T3J, T3B); ci[WS(rs, 8)] = FMA(T3z, T3J, T3K); cr[WS(rs, 12)] = FNMS(T3Q, T3T, T3P); ci[WS(rs, 12)] = FMA(T3Q, T3O, T3U); T38 = T36 + T37; T3i = T37 - T36; T3o = T2Q - T2V; T2W = T2Q + T2V; } { E T2q, T21, T28, T2w, T2v, T2f, T2i, T2r; { E T2Y, T3a, T3c, T3d, T39, T3e, T3b, T2X, T3Y; cr[WS(rs, 4)] = FNMS(TE, T3X, T3W); T3Y = TA * T3X; { E T3t, T3j, T3x, T3p; T3t = FMA(KP707106781, T3i, T3h); T3j = FNMS(KP707106781, T3i, T3h); T3x = FMA(KP707106781, T3o, T3n); T3p = FNMS(KP707106781, T3o, T3n); ci[WS(rs, 4)] = FMA(TE, T3V, T3Y); { E T3u, T3k, T3y, T3q; T3u = T3s * T3t; T3k = T3g * T3j; T3y = T3s * T3x; T3q = T3g * T3p; cr[WS(rs, 6)] = FNMS(T3w, T3x, T3u); cr[WS(rs, 14)] = FNMS(T3m, T3p, T3k); ci[WS(rs, 6)] = FMA(T3w, T3t, T3y); ci[WS(rs, 14)] = FMA(T3m, T3j, T3q); T3b = FMA(KP707106781, T2W, T2L); T2X = FNMS(KP707106781, T2W, T2L); } } T2Y = T2G * T2X; T3a = T30 * T2X; T3c = T1V * T3b; T3d = FMA(KP707106781, T38, T35); T39 = FNMS(KP707106781, T38, T35); T3e = T1X * T3b; T2q = FMA(KP707106781, T20, T1Z); T21 = FNMS(KP707106781, T20, T1Z); cr[WS(rs, 2)] = FNMS(T1X, T3d, T3c); ci[WS(rs, 10)] = FMA(T2G, T39, T3a); cr[WS(rs, 10)] = FNMS(T30, T39, T2Y); ci[WS(rs, 2)] = FMA(T1V, T3d, T3e); T28 = T24 + T27; T2w = T27 - T24; T2v = FNMS(KP707106781, T2e, T2d); T2f = FMA(KP707106781, T2e, T2d); T2i = T2g - T2h; T2r = T2g + T2h; } { E TX, T1k, T1x, T1A; T1J = FMA(KP707106781, TW, TL); TX = FNMS(KP707106781, TW, TL); { E T2l, T29, T2n, T2j; T2l = FNMS(KP923879532, T28, T21); T29 = FMA(KP923879532, T28, T21); T2n = FMA(KP923879532, T2i, T2f); T2j = FNMS(KP923879532, T2i, T2f); { E T2o, T2m, T2k, T2a; T2o = Tz * T2l; T2m = Tw * T2l; T2k = T2c * T29; T2a = T1Y * T29; ci[WS(rs, 3)] = FMA(Tw, T2n, T2o); cr[WS(rs, 3)] = FNMS(Tz, T2n, T2m); ci[WS(rs, 11)] = FMA(T1Y, T2j, T2k); cr[WS(rs, 11)] = FNMS(T2c, T2j, T2a); T1k = T18 - T1j; T1O = T1j + T18; } } T1N = FMA(KP707106781, T1w, T1t); T1x = FNMS(KP707106781, T1w, T1t); T1A = T1y - T1z; T1K = T1y + T1z; { E T1E, T1l, T1H, T1B; T1E = FMA(KP923879532, T1k, TX); T1l = FNMS(KP923879532, T1k, TX); T1H = FMA(KP923879532, T1A, T1x); T1B = FNMS(KP923879532, T1A, T1x); { E T1I, T1F, T1C, T1m; T1I = T1G * T1E; T1F = T1D * T1E; T1C = T1o * T1l; T1m = TG * T1l; ci[WS(rs, 5)] = FMA(T1D, T1H, T1I); cr[WS(rs, 5)] = FNMS(T1G, T1H, T1F); ci[WS(rs, 13)] = FMA(TG, T1B, T1C); cr[WS(rs, 13)] = FNMS(T1o, T1B, T1m); } } { E T2A, T2s, T2D, T2x; T2A = FMA(KP923879532, T2r, T2q); T2s = FNMS(KP923879532, T2r, T2q); T2D = FNMS(KP923879532, T2w, T2v); T2x = FMA(KP923879532, T2w, T2v); { E T2B, T2t, T2E, T2y; T2B = T2z * T2A; T2t = T2p * T2s; T2E = T2z * T2D; T2y = T2p * T2x; cr[WS(rs, 15)] = FNMS(T2C, T2D, T2B); cr[WS(rs, 7)] = FNMS(T2u, T2x, T2t); ci[WS(rs, 15)] = FMA(T2C, T2A, T2E); ci[WS(rs, 7)] = FMA(T2u, T2s, T2y); } } } } } } } { E T1L, T1R, T1P, T1T; T1L = FNMS(KP923879532, T1K, T1J); T1R = FMA(KP923879532, T1K, T1J); T1P = FNMS(KP923879532, T1O, T1N); T1T = FMA(KP923879532, T1O, T1N); { E T1S, T1M, T1U, T1Q; T1S = Tv * T1R; T1M = TB * T1L; T1U = Tv * T1T; T1Q = TB * T1P; cr[WS(rs, 1)] = FNMS(Ty, T1T, T1S); cr[WS(rs, 9)] = FNMS(TF, T1P, T1M); ci[WS(rs, 1)] = FMA(Ty, T1R, T1U); ci[WS(rs, 9)] = FMA(TF, T1L, T1Q); } } } } } static const tw_instr twinstr[] = { {TW_CEXP, 1, 1}, {TW_CEXP, 1, 3}, {TW_CEXP, 1, 9}, {TW_CEXP, 1, 15}, {TW_NEXT, 1, 0} }; static const hc2hc_desc desc = { 16, "hb2_16", twinstr, &GENUS, {104, 42, 92, 0} }; void X(codelet_hb2_16) (planner *p) { X(khc2hc_register) (p, hb2_16, &desc); } #else /* HAVE_FMA */ /* Generated by: ../../../genfft/gen_hc2hc.native -compact -variables 4 -pipeline-latency 4 -sign 1 -twiddle-log3 -precompute-twiddles -n 16 -dif -name hb2_16 -include hb.h */ /* * This function contains 196 FP additions, 108 FP multiplications, * (or, 156 additions, 68 multiplications, 40 fused multiply/add), * 80 stack variables, 3 constants, and 64 memory accesses */ #include "hb.h" static void hb2_16(R *cr, R *ci, const R *W, stride rs, INT mb, INT me, INT ms) { DK(KP382683432, +0.382683432365089771728459984030398866761344562); DK(KP923879532, +0.923879532511286756128183189396788286822416626); DK(KP707106781, +0.707106781186547524400844362104849039284835938); { INT m; for (m = mb, W = W + ((mb - 1) * 8); m < me; m = m + 1, cr = cr + ms, ci = ci - ms, W = W + 8, MAKE_VOLATILE_STRIDE(32, rs)) { E Tv, Ty, T1l, T1n, T1p, T1t, T27, T25, Tz, Tw, TB, T21, T1P, T1H, T1X; E T17, T1L, T1N, T1v, T1w, T1x, T1B, T2F, T2T, T2b, T2R, T3j, T3x, T35, T3t; { E TA, T1J, T15, T1G, Tx, T1K, T16, T1F; { E T1m, T1s, T1o, T1r; Tv = W[0]; Ty = W[1]; T1l = W[2]; T1n = W[3]; T1m = Tv * T1l; T1s = Ty * T1l; T1o = Ty * T1n; T1r = Tv * T1n; T1p = T1m + T1o; T1t = T1r - T1s; T27 = T1r + T1s; T25 = T1m - T1o; Tz = W[5]; TA = Ty * Tz; T1J = T1l * Tz; T15 = Tv * Tz; T1G = T1n * Tz; Tw = W[4]; Tx = Tv * Tw; T1K = T1n * Tw; T16 = Ty * Tw; T1F = T1l * Tw; } TB = Tx - TA; T21 = T1J + T1K; T1P = T15 - T16; T1H = T1F + T1G; T1X = T1F - T1G; T17 = T15 + T16; T1L = T1J - T1K; T1N = Tx + TA; T1v = W[6]; T1w = W[7]; T1x = FMA(Tv, T1v, Ty * T1w); T1B = FNMS(Ty, T1v, Tv * T1w); { E T2D, T2E, T29, T2a; T2D = T25 * Tz; T2E = T27 * Tw; T2F = T2D + T2E; T2T = T2D - T2E; T29 = T25 * Tw; T2a = T27 * Tz; T2b = T29 - T2a; T2R = T29 + T2a; } { E T3h, T3i, T33, T34; T3h = T1p * Tz; T3i = T1t * Tw; T3j = T3h + T3i; T3x = T3h - T3i; T33 = T1p * Tw; T34 = T1t * Tz; T35 = T33 - T34; T3t = T33 + T34; } } { E T7, T36, T3k, TC, T1f, T2e, T2I, T1Q, Te, TJ, T1R, T18, T2L, T37, T2l; E T3l, Tm, T1T, TT, T1h, T2A, T2N, T3b, T3n, Tt, T1U, T12, T1i, T2t, T2O; E T3e, T3o; { E T3, T2c, T1e, T2d, T6, T2G, T1b, T2H; { E T1, T2, T1c, T1d; T1 = cr[0]; T2 = ci[WS(rs, 7)]; T3 = T1 + T2; T2c = T1 - T2; T1c = ci[WS(rs, 11)]; T1d = cr[WS(rs, 12)]; T1e = T1c - T1d; T2d = T1c + T1d; } { E T4, T5, T19, T1a; T4 = cr[WS(rs, 4)]; T5 = ci[WS(rs, 3)]; T6 = T4 + T5; T2G = T4 - T5; T19 = ci[WS(rs, 15)]; T1a = cr[WS(rs, 8)]; T1b = T19 - T1a; T2H = T19 + T1a; } T7 = T3 + T6; T36 = T2c + T2d; T3k = T2H - T2G; TC = T3 - T6; T1f = T1b - T1e; T2e = T2c - T2d; T2I = T2G + T2H; T1Q = T1b + T1e; } { E Ta, T2f, TI, T2g, Td, T2i, TF, T2j; { E T8, T9, TG, TH; T8 = cr[WS(rs, 2)]; T9 = ci[WS(rs, 5)]; Ta = T8 + T9; T2f = T8 - T9; TG = ci[WS(rs, 13)]; TH = cr[WS(rs, 10)]; TI = TG - TH; T2g = TG + TH; } { E Tb, Tc, TD, TE; Tb = ci[WS(rs, 1)]; Tc = cr[WS(rs, 6)]; Td = Tb + Tc; T2i = Tb - Tc; TD = ci[WS(rs, 9)]; TE = cr[WS(rs, 14)]; TF = TD - TE; T2j = TD + TE; } Te = Ta + Td; TJ = TF - TI; T1R = TI + TF; T18 = Ta - Td; { E T2J, T2K, T2h, T2k; T2J = T2f + T2g; T2K = T2i + T2j; T2L = KP707106781 * (T2J - T2K); T37 = KP707106781 * (T2J + T2K); T2h = T2f - T2g; T2k = T2i - T2j; T2l = KP707106781 * (T2h + T2k); T3l = KP707106781 * (T2h - T2k); } } { E Ti, T2x, TR, T2y, Tl, T2u, TO, T2v, TL, TS; { E Tg, Th, TP, TQ; Tg = cr[WS(rs, 1)]; Th = ci[WS(rs, 6)]; Ti = Tg + Th; T2x = Tg - Th; TP = ci[WS(rs, 10)]; TQ = cr[WS(rs, 13)]; TR = TP - TQ; T2y = TP + TQ; } { E Tj, Tk, TM, TN; Tj = cr[WS(rs, 5)]; Tk = ci[WS(rs, 2)]; Tl = Tj + Tk; T2u = Tj - Tk; TM = ci[WS(rs, 14)]; TN = cr[WS(rs, 9)]; TO = TM - TN; T2v = TM + TN; } Tm = Ti + Tl; T1T = TO + TR; TL = Ti - Tl; TS = TO - TR; TT = TL - TS; T1h = TL + TS; { E T2w, T2z, T39, T3a; T2w = T2u + T2v; T2z = T2x - T2y; T2A = FMA(KP923879532, T2w, KP382683432 * T2z); T2N = FNMS(KP382683432, T2w, KP923879532 * T2z); T39 = T2x + T2y; T3a = T2v - T2u; T3b = FNMS(KP923879532, T3a, KP382683432 * T39); T3n = FMA(KP382683432, T3a, KP923879532 * T39); } } { E Tp, T2q, T10, T2r, Ts, T2n, TX, T2o, TU, T11; { E Tn, To, TY, TZ; Tn = ci[0]; To = cr[WS(rs, 7)]; Tp = Tn + To; T2q = Tn - To; TY = ci[WS(rs, 12)]; TZ = cr[WS(rs, 11)]; T10 = TY - TZ; T2r = TY + TZ; } { E Tq, Tr, TV, TW; Tq = cr[WS(rs, 3)]; Tr = ci[WS(rs, 4)]; Ts = Tq + Tr; T2n = Tq - Tr; TV = ci[WS(rs, 8)]; TW = cr[WS(rs, 15)]; TX = TV - TW; T2o = TV + TW; } Tt = Tp + Ts; T1U = TX + T10; TU = Tp - Ts; T11 = TX - T10; T12 = TU + T11; T1i = T11 - TU; { E T2p, T2s, T3c, T3d; T2p = T2n - T2o; T2s = T2q - T2r; T2t = FNMS(KP382683432, T2s, KP923879532 * T2p); T2O = FMA(KP382683432, T2p, KP923879532 * T2s); T3c = T2q + T2r; T3d = T2n + T2o; T3e = FNMS(KP923879532, T3d, KP382683432 * T3c); T3o = FMA(KP382683432, T3d, KP923879532 * T3c); } } { E Tf, Tu, T1O, T1S, T1V, T1W; Tf = T7 + Te; Tu = Tm + Tt; T1O = Tf - Tu; T1S = T1Q + T1R; T1V = T1T + T1U; T1W = T1S - T1V; cr[0] = Tf + Tu; ci[0] = T1S + T1V; cr[WS(rs, 8)] = FNMS(T1P, T1W, T1N * T1O); ci[WS(rs, 8)] = FMA(T1P, T1O, T1N * T1W); } { E T3g, T3r, T3q, T3s; { E T38, T3f, T3m, T3p; T38 = T36 - T37; T3f = T3b + T3e; T3g = T38 - T3f; T3r = T38 + T3f; T3m = T3k + T3l; T3p = T3n - T3o; T3q = T3m - T3p; T3s = T3m + T3p; } cr[WS(rs, 11)] = FNMS(T3j, T3q, T35 * T3g); ci[WS(rs, 11)] = FMA(T3j, T3g, T35 * T3q); cr[WS(rs, 3)] = FNMS(T1n, T3s, T1l * T3r); ci[WS(rs, 3)] = FMA(T1n, T3r, T1l * T3s); } { E T3w, T3B, T3A, T3C; { E T3u, T3v, T3y, T3z; T3u = T36 + T37; T3v = T3n + T3o; T3w = T3u - T3v; T3B = T3u + T3v; T3y = T3k - T3l; T3z = T3b - T3e; T3A = T3y + T3z; T3C = T3y - T3z; } cr[WS(rs, 7)] = FNMS(T3x, T3A, T3t * T3w); ci[WS(rs, 7)] = FMA(T3t, T3A, T3x * T3w); cr[WS(rs, 15)] = FNMS(T1w, T3C, T1v * T3B); ci[WS(rs, 15)] = FMA(T1v, T3C, T1w * T3B); } { E T14, T1q, T1k, T1u; { E TK, T13, T1g, T1j; TK = TC + TJ; T13 = KP707106781 * (TT + T12); T14 = TK - T13; T1q = TK + T13; T1g = T18 + T1f; T1j = KP707106781 * (T1h + T1i); T1k = T1g - T1j; T1u = T1g + T1j; } cr[WS(rs, 10)] = FNMS(T17, T1k, TB * T14); ci[WS(rs, 10)] = FMA(T17, T14, TB * T1k); cr[WS(rs, 2)] = FNMS(T1t, T1u, T1p * T1q); ci[WS(rs, 2)] = FMA(T1t, T1q, T1p * T1u); } { E T1A, T1I, T1E, T1M; { E T1y, T1z, T1C, T1D; T1y = TC - TJ; T1z = KP707106781 * (T1i - T1h); T1A = T1y - T1z; T1I = T1y + T1z; T1C = T1f - T18; T1D = KP707106781 * (TT - T12); T1E = T1C - T1D; T1M = T1C + T1D; } cr[WS(rs, 14)] = FNMS(T1B, T1E, T1x * T1A); ci[WS(rs, 14)] = FMA(T1x, T1E, T1B * T1A); cr[WS(rs, 6)] = FNMS(T1L, T1M, T1H * T1I); ci[WS(rs, 6)] = FMA(T1H, T1M, T1L * T1I); } { E T2C, T2S, T2Q, T2U; { E T2m, T2B, T2M, T2P; T2m = T2e - T2l; T2B = T2t - T2A; T2C = T2m - T2B; T2S = T2m + T2B; T2M = T2I - T2L; T2P = T2N - T2O; T2Q = T2M - T2P; T2U = T2M + T2P; } cr[WS(rs, 13)] = FNMS(T2F, T2Q, T2b * T2C); ci[WS(rs, 13)] = FMA(T2F, T2C, T2b * T2Q); cr[WS(rs, 5)] = FNMS(T2T, T2U, T2R * T2S); ci[WS(rs, 5)] = FMA(T2T, T2S, T2R * T2U); } { E T2X, T31, T30, T32; { E T2V, T2W, T2Y, T2Z; T2V = T2e + T2l; T2W = T2N + T2O; T2X = T2V - T2W; T31 = T2V + T2W; T2Y = T2I + T2L; T2Z = T2A + T2t; T30 = T2Y - T2Z; T32 = T2Y + T2Z; } cr[WS(rs, 9)] = FNMS(Tz, T30, Tw * T2X); ci[WS(rs, 9)] = FMA(Tw, T30, Tz * T2X); cr[WS(rs, 1)] = FNMS(Ty, T32, Tv * T31); ci[WS(rs, 1)] = FMA(Tv, T32, Ty * T31); } { E T20, T26, T24, T28; { E T1Y, T1Z, T22, T23; T1Y = T7 - Te; T1Z = T1U - T1T; T20 = T1Y - T1Z; T26 = T1Y + T1Z; T22 = T1Q - T1R; T23 = Tm - Tt; T24 = T22 - T23; T28 = T23 + T22; } cr[WS(rs, 12)] = FNMS(T21, T24, T1X * T20); ci[WS(rs, 12)] = FMA(T1X, T24, T21 * T20); cr[WS(rs, 4)] = FNMS(T27, T28, T25 * T26); ci[WS(rs, 4)] = FMA(T25, T28, T27 * T26); } } } } } static const tw_instr twinstr[] = { {TW_CEXP, 1, 1}, {TW_CEXP, 1, 3}, {TW_CEXP, 1, 9}, {TW_CEXP, 1, 15}, {TW_NEXT, 1, 0} }; static const hc2hc_desc desc = { 16, "hb2_16", twinstr, &GENUS, {156, 68, 40, 0} }; void X(codelet_hb2_16) (planner *p) { X(khc2hc_register) (p, hb2_16, &desc); } #endif /* HAVE_FMA */ fftw-3.3.3/rdft/scalar/r2cb/hb2_32.c0000644000175000017500000014742012054410774013602 00000000000000/* * Copyright (c) 2003, 2007-11 Matteo Frigo * Copyright (c) 2003, 2007-11 Massachusetts Institute of Technology * * This program is free software; you can redistribute it and/or modify * it under the terms of the GNU General Public License as published by * the Free Software Foundation; either version 2 of the License, or * (at your option) any later version. * * This program is distributed in the hope that it will be useful, * but WITHOUT ANY WARRANTY; without even the implied warranty of * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the * GNU General Public License for more details. * * You should have received a copy of the GNU General Public License * along with this program; if not, write to the Free Software * Foundation, Inc., 51 Franklin Street, Fifth Floor, Boston, MA 02110-1301 USA * */ /* This file was automatically generated --- DO NOT EDIT */ /* Generated on Sun Nov 25 07:41:25 EST 2012 */ #include "codelet-rdft.h" #ifdef HAVE_FMA /* Generated by: ../../../genfft/gen_hc2hc.native -fma -reorder-insns -schedule-for-pipeline -compact -variables 4 -pipeline-latency 4 -sign 1 -twiddle-log3 -precompute-twiddles -n 32 -dif -name hb2_32 -include hb.h */ /* * This function contains 488 FP additions, 350 FP multiplications, * (or, 236 additions, 98 multiplications, 252 fused multiply/add), * 204 stack variables, 7 constants, and 128 memory accesses */ #include "hb.h" static void hb2_32(R *cr, R *ci, const R *W, stride rs, INT mb, INT me, INT ms) { DK(KP980785280, +0.980785280403230449126182236134239036973933731); DK(KP198912367, +0.198912367379658006911597622644676228597850501); DK(KP831469612, +0.831469612302545237078788377617905756738560812); DK(KP668178637, +0.668178637919298919997757686523080761552472251); DK(KP923879532, +0.923879532511286756128183189396788286822416626); DK(KP414213562, +0.414213562373095048801688724209698078569671875); DK(KP707106781, +0.707106781186547524400844362104849039284835938); { INT m; for (m = mb, W = W + ((mb - 1) * 8); m < me; m = m + 1, cr = cr + ms, ci = ci - ms, W = W + 8, MAKE_VOLATILE_STRIDE(64, rs)) { E T5u, T6b, T6e, T5I, T66, T60, T5U, T5R, T67, T5L, T61, T5x, T5A, T5D, T5O; E T62, T5V, T5P; { E T11, T14, T12, T37, T17, T1b, T39, T15, T7C, T8P, T8S, T7I, T98, T7e, T78; E T8V, T3d, T3x, T3a, T3v, T9s, T3G, T4p, T5X, T16, T9m, T3y, T4b, T3C, T4g; E T5Z, T1a, T4r, T3J, T2O, T1c, T4W, T4s, T3Y, T3K, T3l, T3e, T3i, T3q, T8K; E T8E, T8m, T7S, T5k, T5e; { E T13, T3c, T38, T3F, T7B, T9l, T77, T7d, T9r, T7H; T11 = W[2]; T14 = W[3]; T12 = W[4]; T37 = W[0]; T17 = W[6]; T1b = W[7]; T13 = T11 * T12; T3c = T37 * T14; T38 = T37 * T11; T3F = T37 * T12; T7B = T11 * T17; T9l = T12 * T17; T77 = T37 * T17; T7d = T37 * T1b; T9r = T12 * T1b; T7H = T11 * T1b; T39 = W[1]; T15 = W[5]; { E T3I, T19, T5d, T3b, T18, T2N; T7C = FMA(T14, T1b, T7B); T8P = FNMS(T14, T1b, T7B); T8S = FMA(T14, T17, T7H); T7I = FNMS(T14, T17, T7H); T98 = FNMS(T39, T17, T7d); T7e = FMA(T39, T17, T7d); T78 = FNMS(T39, T1b, T77); T8V = FMA(T39, T1b, T77); T3d = FMA(T39, T11, T3c); T3x = FNMS(T39, T11, T3c); T3a = FNMS(T39, T14, T38); T3v = FMA(T39, T14, T38); T9s = FNMS(T15, T17, T9r); T3G = FNMS(T39, T15, T3F); T4p = FMA(T39, T15, T3F); T5X = FNMS(T14, T15, T13); T16 = FMA(T14, T15, T13); T3I = T37 * T15; T19 = T11 * T15; T5d = T3v * T12; T3b = T3a * T12; T9m = FMA(T15, T1b, T9l); { E T3w, T3B, T5t, T5H; T3w = T3v * T17; T3B = T3v * T1b; T5t = T3a * T17; T5H = T3a * T1b; T3y = FNMS(T3x, T1b, T3w); T4b = FMA(T3x, T1b, T3w); T3C = FMA(T3x, T17, T3B); T4g = FNMS(T3x, T17, T3B); T5u = FMA(T3d, T1b, T5t); T6b = FNMS(T3d, T1b, T5t); T6e = FMA(T3d, T17, T5H); T5I = FNMS(T3d, T17, T5H); T18 = T16 * T17; T2N = T16 * T1b; T5Z = FMA(T14, T12, T19); T1a = FNMS(T14, T12, T19); } { E T3H, T3X, T4q, T4V, T5Y, T65; T4q = T4p * T17; T4V = T4p * T1b; T4r = FNMS(T39, T12, T3I); T3J = FMA(T39, T12, T3I); T2O = FNMS(T1a, T17, T2N); T1c = FMA(T1a, T1b, T18); T3H = T3G * T17; T4W = FNMS(T4r, T17, T4V); T4s = FMA(T4r, T1b, T4q); T3X = T3G * T1b; T5Y = T5X * T17; T65 = T5X * T1b; T3Y = FNMS(T3J, T17, T3X); T3K = FMA(T3J, T1b, T3H); { E T8J, T8D, T3h, T5j, T8l, T7R; T3h = T3a * T15; T66 = FNMS(T5Z, T17, T65); T60 = FMA(T5Z, T1b, T5Y); T3l = FNMS(T3d, T15, T3b); T3e = FMA(T3d, T15, T3b); T3i = FNMS(T3d, T12, T3h); T3q = FMA(T3d, T12, T3h); T8J = T3l * T1b; T8D = T3l * T17; T5j = T3v * T15; T8l = T3e * T1b; T7R = T3e * T17; T8K = FNMS(T3q, T17, T8J); T8E = FMA(T3q, T1b, T8D); T8m = FNMS(T3i, T17, T8l); T7S = FMA(T3i, T1b, T7R); T5U = FNMS(T3x, T12, T5j); T5k = FMA(T3x, T12, T5j); T5e = FNMS(T3x, T15, T5d); T5R = FMA(T3x, T15, T5d); } } } } { E T6O, T6i, T7s, T7o, T6j, Tf, T8W, T7V, T99, T8p, T3L, T1t, T3Z, T2X, T5J; E T4Z, T7t, T6W, T5v, T4v, TZ, T7x, T91, T9d, T28, T3S, T3R, T2h, T5B, T4Q; E T8v, T8a, T5C, T4N, T6Z, T6J, TK, T7w, T2z, T3P, T94, T9c, T3O, T2I, T5y; E T4J, T8u, T8h, T5z, T4G, T6Y, T6A, T6p, T6m, T6P, Tu, T9a, T82, T8X, T8s; E T40, T1Q, T4y, T4B, T3M, T30, T5w, T52; { E T6B, T6I, T4L, T4M, T4t, T4u; { E T1d, T3, T2P, T6, T6Q, T2S, T6R, T1g, Td, T6U, T1i, Ta, T2V, T1r, T6T; E T1l; { E T4, T5, T2Q, T2R, T1, T2, T1e, T1f; T1 = cr[0]; T2 = ci[WS(rs, 15)]; { E T6N, T6h, T7r, T7n; T6N = T5R * T1b; T6h = T5R * T17; T7r = T5e * T1b; T7n = T5e * T17; T6O = FNMS(T5U, T17, T6N); T6i = FMA(T5U, T1b, T6h); T7s = FNMS(T5k, T17, T7r); T7o = FMA(T5k, T1b, T7n); T1d = T1 - T2; T3 = T1 + T2; } T4 = cr[WS(rs, 8)]; T5 = ci[WS(rs, 7)]; T2Q = ci[WS(rs, 31)]; T2R = cr[WS(rs, 16)]; T1e = ci[WS(rs, 23)]; T2P = T4 - T5; T6 = T4 + T5; T6Q = T2Q - T2R; T2S = T2Q + T2R; T1f = cr[WS(rs, 24)]; { E T1o, T1n, T1p, Tb, Tc; Tb = ci[WS(rs, 3)]; Tc = cr[WS(rs, 12)]; T1o = ci[WS(rs, 19)]; T6R = T1e - T1f; T1g = T1e + T1f; T1n = Tb - Tc; Td = Tb + Tc; T1p = cr[WS(rs, 28)]; { E T1j, T1k, T8, T9, T1q; T8 = cr[WS(rs, 4)]; T9 = ci[WS(rs, 11)]; T1q = T1o + T1p; T6U = T1o - T1p; T1j = ci[WS(rs, 27)]; T1i = T8 - T9; Ta = T8 + T9; T1k = cr[WS(rs, 20)]; T2V = T1n + T1q; T1r = T1n - T1q; T6T = T1j - T1k; T1l = T1j + T1k; } } } { E T2U, T6V, T6S, T1h, T1s, T4Y, T4X, T2T, T2W; { E T7T, T8o, T1m, T7U, T7, Te, T8n; T7T = T3 - T6; T7 = T3 + T6; Te = Ta + Td; T8o = Ta - Td; T1m = T1i - T1l; T2U = T1i + T1l; T6j = T7 - Te; Tf = T7 + Te; T7U = T6U - T6T; T6V = T6T + T6U; T6S = T6Q + T6R; T8n = T6Q - T6R; T4t = T1d + T1g; T1h = T1d - T1g; T8W = T7T + T7U; T7V = T7T - T7U; T99 = T8o + T8n; T8p = T8n - T8o; T1s = T1m + T1r; T4Y = T1m - T1r; } T4X = T2S - T2P; T2T = T2P + T2S; T2W = T2U - T2V; T4u = T2U + T2V; T3L = FMA(KP707106781, T1s, T1h); T1t = FNMS(KP707106781, T1s, T1h); T3Z = FMA(KP707106781, T2W, T2T); T2X = FNMS(KP707106781, T2W, T2T); T5J = FNMS(KP707106781, T4Y, T4X); T4Z = FMA(KP707106781, T4Y, T4X); T7t = T6S + T6V; T6W = T6S - T6V; } } { E T29, T1S, T1V, T87, TR, T2c, T84, T6E, T1X, TU, T1Y, T6G, T25, T22, TX; E T1Z; { E TO, TN, TP, TL, TM, T6C, T6D; TL = ci[0]; TM = cr[WS(rs, 15)]; TO = cr[WS(rs, 7)]; T5v = FMA(KP707106781, T4u, T4t); T4v = FNMS(KP707106781, T4u, T4t); TN = TL + TM; T29 = TL - TM; TP = ci[WS(rs, 8)]; { E T2a, T2b, T1T, T1U, TQ; T1T = ci[WS(rs, 16)]; T1U = cr[WS(rs, 31)]; TQ = TO + TP; T1S = TO - TP; T2a = ci[WS(rs, 24)]; T6C = T1T - T1U; T1V = T1T + T1U; T2b = cr[WS(rs, 23)]; T87 = TN - TQ; TR = TN + TQ; T2c = T2a + T2b; T6D = T2a - T2b; } { E T23, T24, TS, TT, TV, TW; TS = cr[WS(rs, 3)]; TT = ci[WS(rs, 12)]; T84 = T6C - T6D; T6E = T6C + T6D; T23 = ci[WS(rs, 20)]; T1X = TS - TT; TU = TS + TT; T24 = cr[WS(rs, 27)]; TV = ci[WS(rs, 4)]; TW = cr[WS(rs, 11)]; T1Y = ci[WS(rs, 28)]; T6G = T23 - T24; T25 = T23 + T24; T22 = TV - TW; TX = TV + TW; T1Z = cr[WS(rs, 19)]; } } { E T4O, T1W, T2f, T26, T8Z, T86, T2e, T21, T89, T90; { E T85, TY, T6F, T20, T6H, T88; T4O = T1S + T1V; T1W = T1S - T1V; T2f = T22 - T25; T26 = T22 + T25; T85 = TU - TX; TY = TU + TX; T6F = T1Y - T1Z; T20 = T1Y + T1Z; T8Z = T85 + T84; T86 = T84 - T85; T6B = TR - TY; TZ = TR + TY; T6H = T6F + T6G; T88 = T6G - T6F; T2e = T1X - T20; T21 = T1X + T20; T7x = T6E + T6H; T6I = T6E - T6H; T89 = T87 - T88; T90 = T87 + T88; } { E T4P, T2d, T27, T2g; T2d = T29 - T2c; T4L = T29 + T2c; T4M = T21 + T26; T27 = T21 - T26; T2g = T2e + T2f; T4P = T2e - T2f; T91 = FNMS(KP414213562, T90, T8Z); T9d = FMA(KP414213562, T8Z, T90); T28 = FNMS(KP707106781, T27, T1W); T3S = FMA(KP707106781, T27, T1W); T3R = FMA(KP707106781, T2g, T2d); T2h = FNMS(KP707106781, T2g, T2d); T5B = FMA(KP707106781, T4P, T4O); T4Q = FNMS(KP707106781, T4P, T4O); T8v = FNMS(KP414213562, T86, T89); T8a = FMA(KP414213562, T89, T86); } } } { E T6s, T6z, T4F, T4E; { E T2A, T2j, TC, T8e, T2m, T2D, T6v, T8b, TG, T2o, TF, T6x, T2w, TH, T2p; E T2q; { E Tw, Tx, Tz, TA, T6t, T6u; Tw = cr[WS(rs, 1)]; T5C = FMA(KP707106781, T4M, T4L); T4N = FNMS(KP707106781, T4M, T4L); T6Z = T6I - T6B; T6J = T6B + T6I; Tx = ci[WS(rs, 14)]; Tz = cr[WS(rs, 9)]; TA = ci[WS(rs, 6)]; { E T2k, Ty, TB, T2l, T2B, T2C; T2k = ci[WS(rs, 30)]; T2A = Tw - Tx; Ty = Tw + Tx; T2j = Tz - TA; TB = Tz + TA; T2l = cr[WS(rs, 17)]; T2B = ci[WS(rs, 22)]; T2C = cr[WS(rs, 25)]; TC = Ty + TB; T8e = Ty - TB; T2m = T2k + T2l; T6t = T2k - T2l; T6u = T2B - T2C; T2D = T2B + T2C; } { E TD, TE, T2u, T2v; TD = cr[WS(rs, 5)]; T6v = T6t + T6u; T8b = T6t - T6u; TE = ci[WS(rs, 10)]; T2u = ci[WS(rs, 18)]; T2v = cr[WS(rs, 29)]; TG = ci[WS(rs, 2)]; T2o = TD - TE; TF = TD + TE; T6x = T2u - T2v; T2w = T2u + T2v; TH = cr[WS(rs, 13)]; T2p = ci[WS(rs, 26)]; T2q = cr[WS(rs, 21)]; } } { E T4H, T2n, T2G, T2F, T92, T8d, T2y, T93, T8g, T4I, T2E, T2H; { E T2x, T8c, T8f, T2s, T2t, TI; T4H = T2m - T2j; T2n = T2j + T2m; T2t = TG - TH; TI = TG + TH; { E T6w, T2r, TJ, T6y; T6w = T2p - T2q; T2r = T2p + T2q; T2G = T2t - T2w; T2x = T2t + T2w; T8c = TF - TI; TJ = TF + TI; T6y = T6w + T6x; T8f = T6x - T6w; T2F = T2o - T2r; T2s = T2o + T2r; TK = TC + TJ; T6s = TC - TJ; T6z = T6v - T6y; T7w = T6v + T6y; } T92 = T8c + T8b; T8d = T8b - T8c; T4F = T2s + T2x; T2y = T2s - T2x; T93 = T8e + T8f; T8g = T8e - T8f; } T4E = T2A + T2D; T2E = T2A - T2D; T2H = T2F + T2G; T4I = T2G - T2F; T2z = FNMS(KP707106781, T2y, T2n); T3P = FMA(KP707106781, T2y, T2n); T94 = FMA(KP414213562, T93, T92); T9c = FNMS(KP414213562, T92, T93); T3O = FMA(KP707106781, T2H, T2E); T2I = FNMS(KP707106781, T2H, T2E); T5y = FMA(KP707106781, T4I, T4H); T4J = FNMS(KP707106781, T4I, T4H); T8u = FMA(KP414213562, T8d, T8g); T8h = FNMS(KP414213562, T8g, T8d); } } { E T4x, T1O, Tm, T7Z, T80, T4w, T1J, T4A, T1D, Tt, T7X, T7W, T4z, T1y; { E Tj, T1K, Ti, T6o, T1N, Tk, T1G, T1H; { E Tg, Th, T1L, T1M; Tg = cr[WS(rs, 2)]; T5z = FMA(KP707106781, T4F, T4E); T4G = FNMS(KP707106781, T4F, T4E); T6Y = T6s + T6z; T6A = T6s - T6z; Th = ci[WS(rs, 13)]; T1L = ci[WS(rs, 21)]; T1M = cr[WS(rs, 26)]; Tj = cr[WS(rs, 10)]; T1K = Tg - Th; Ti = Tg + Th; T6o = T1L - T1M; T1N = T1L + T1M; Tk = ci[WS(rs, 5)]; T1G = ci[WS(rs, 29)]; T1H = cr[WS(rs, 18)]; } { E T1F, Tl, T6n, T1I; T4x = T1K + T1N; T1O = T1K - T1N; T1F = Tj - Tk; Tl = Tj + Tk; T6n = T1G - T1H; T1I = T1G + T1H; Tm = Ti + Tl; T7Z = Ti - Tl; T80 = T6n - T6o; T6p = T6n + T6o; T4w = T1I - T1F; T1J = T1F + T1I; } } { E Tq, T1z, Tp, T6l, T1C, Tr, T1v, T1w; { E Tn, To, T1A, T1B; Tn = ci[WS(rs, 1)]; To = cr[WS(rs, 14)]; T1A = ci[WS(rs, 25)]; T1B = cr[WS(rs, 22)]; Tq = cr[WS(rs, 6)]; T1z = Tn - To; Tp = Tn + To; T6l = T1A - T1B; T1C = T1A + T1B; Tr = ci[WS(rs, 9)]; T1v = ci[WS(rs, 17)]; T1w = cr[WS(rs, 30)]; } { E T1u, Ts, T6k, T1x; T4A = T1z + T1C; T1D = T1z - T1C; T1u = Tq - Tr; Ts = Tq + Tr; T6k = T1v - T1w; T1x = T1v + T1w; Tt = Tp + Ts; T7X = Tp - Ts; T7W = T6k - T6l; T6m = T6k + T6l; T4z = T1u + T1x; T1y = T1u - T1x; } } { E T8r, T8q, T2Z, T1E, T1P, T2Y, T7Y, T81, T50, T51; T8r = T7X + T7W; T7Y = T7W - T7X; T81 = T7Z + T80; T8q = T7Z - T80; T6P = Tm - Tt; Tu = Tm + Tt; T9a = T81 + T7Y; T82 = T7Y - T81; T2Z = FMA(KP414213562, T1y, T1D); T1E = FNMS(KP414213562, T1D, T1y); T1P = FMA(KP414213562, T1O, T1J); T2Y = FNMS(KP414213562, T1J, T1O); T8X = T8q + T8r; T8s = T8q - T8r; T40 = T1P + T1E; T1Q = T1E - T1P; T4y = FNMS(KP414213562, T4x, T4w); T50 = FMA(KP414213562, T4w, T4x); T51 = FMA(KP414213562, T4z, T4A); T4B = FNMS(KP414213562, T4A, T4z); T3M = T2Y + T2Z; T30 = T2Y - T2Z; T5w = T50 + T51; T52 = T50 - T51; } } } } { E T7D, T5K, T4C, T7K, T7J, T7E, T83, T8w, T8t, T8i, T6r, T70, T6X, T6K; { E T6q, T8Y, T9e, T9b, T95, T8L, T8Q, T8H, T8M, T8I, T8R; { E Tv, T10, T7v, T7y, T7u; T7D = Tf - Tu; Tv = Tf + Tu; T7u = T6p + T6m; T6q = T6m - T6p; T5K = T4B - T4y; T4C = T4y + T4B; T10 = TK + TZ; T7K = TK - TZ; T7J = T7t - T7u; T7v = T7t + T7u; T7y = T7w + T7x; T7E = T7x - T7w; { E T9t, T9x, T9p, T9u, T9q, T9y; { E T9n, T7z, T9o, T7A, T7q, T7p; T8Y = FNMS(KP707106781, T8X, T8W); T9n = FMA(KP707106781, T8X, T8W); cr[0] = Tv + T10; T7p = Tv - T10; ci[0] = T7v + T7y; T7z = T7v - T7y; T9o = T9c + T9d; T9e = T9c - T9d; T7A = T7s * T7p; T7q = T7o * T7p; T9b = FNMS(KP707106781, T9a, T99); T9t = FMA(KP707106781, T9a, T99); T9x = FMA(KP923879532, T9o, T9n); T9p = FNMS(KP923879532, T9o, T9n); ci[WS(rs, 16)] = FMA(T7o, T7z, T7A); cr[WS(rs, 16)] = FNMS(T7s, T7z, T7q); T9u = T94 + T91; T95 = T91 - T94; } T9q = T9m * T9p; T9y = T3v * T9x; { E T8F, T9z, T9v, T8G, T9A, T9w; T83 = FMA(KP707106781, T82, T7V); T8F = FNMS(KP707106781, T82, T7V); T9z = FMA(KP923879532, T9u, T9t); T9v = FNMS(KP923879532, T9u, T9t); T8G = T8u + T8v; T8w = T8u - T8v; T8t = FMA(KP707106781, T8s, T8p); T8L = FNMS(KP707106781, T8s, T8p); T9A = T3v * T9z; cr[WS(rs, 2)] = FNMS(T3x, T9z, T9y); T9w = T9m * T9v; cr[WS(rs, 18)] = FNMS(T9s, T9v, T9q); T8Q = FMA(KP923879532, T8G, T8F); T8H = FNMS(KP923879532, T8G, T8F); ci[WS(rs, 2)] = FMA(T3x, T9x, T9A); ci[WS(rs, 18)] = FMA(T9s, T9p, T9w); T8M = T8h + T8a; T8i = T8a - T8h; } T8I = T8E * T8H; T8R = T8P * T8Q; } } { E T7f, T7j, T7b, T7g, T7c, T7k; { E T79, T8T, T8N, T7a, T8U, T8O; T6r = T6j + T6q; T79 = T6j - T6q; T8T = FMA(KP923879532, T8M, T8L); T8N = FNMS(KP923879532, T8M, T8L); T7a = T6Z - T6Y; T70 = T6Y + T6Z; T6X = T6P + T6W; T7f = T6W - T6P; T8U = T8P * T8T; cr[WS(rs, 30)] = FNMS(T8S, T8T, T8R); T8O = T8E * T8N; cr[WS(rs, 14)] = FNMS(T8K, T8N, T8I); T7j = FMA(KP707106781, T7a, T79); T7b = FNMS(KP707106781, T7a, T79); ci[WS(rs, 30)] = FMA(T8S, T8Q, T8U); ci[WS(rs, 14)] = FMA(T8K, T8H, T8O); T7g = T6A - T6J; T6K = T6A + T6J; } T7c = T78 * T7b; T7k = T5X * T7j; { E T97, T9g, T9i, T9j, T9f, T9k, T9h, T96; { E T7l, T7h, T7m, T7i; T7l = FMA(KP707106781, T7g, T7f); T7h = FNMS(KP707106781, T7g, T7f); T7m = T5X * T7l; cr[WS(rs, 12)] = FNMS(T5Z, T7l, T7k); T7i = T78 * T7h; cr[WS(rs, 28)] = FNMS(T7e, T7h, T7c); T9h = FMA(KP923879532, T95, T8Y); T96 = FNMS(KP923879532, T95, T8Y); ci[WS(rs, 12)] = FMA(T5Z, T7j, T7m); ci[WS(rs, 28)] = FMA(T7e, T7b, T7i); } T97 = T8V * T96; T9g = T98 * T96; T9i = T3G * T9h; T9j = FMA(KP923879532, T9e, T9b); T9f = FNMS(KP923879532, T9e, T9b); T9k = T3J * T9h; cr[WS(rs, 10)] = FNMS(T3J, T9j, T9i); ci[WS(rs, 26)] = FMA(T8V, T9f, T9g); cr[WS(rs, 26)] = FNMS(T98, T9f, T97); ci[WS(rs, 10)] = FMA(T3G, T9j, T9k); } } } { E T31, T3r, T1R, T3m, T33, T32, T3s, T2K, T8z, T8j; { E T73, T6L, T75, T71; T73 = FMA(KP707106781, T6K, T6r); T6L = FNMS(KP707106781, T6K, T6r); T75 = FMA(KP707106781, T70, T6X); T71 = FNMS(KP707106781, T70, T6X); { E T76, T74, T72, T6M; T76 = T3d * T73; T74 = T3a * T73; T72 = T6O * T6L; T6M = T6i * T6L; ci[WS(rs, 4)] = FMA(T3a, T75, T76); cr[WS(rs, 4)] = FNMS(T3d, T75, T74); ci[WS(rs, 20)] = FMA(T6i, T71, T72); cr[WS(rs, 20)] = FNMS(T6O, T71, T6M); } } { E T7N, T7F, T7P, T7L; T7N = T7D + T7E; T7F = T7D - T7E; T7P = T7K + T7J; T7L = T7J - T7K; { E T7O, T7G, T7Q, T7M; T7O = T4p * T7N; T7G = T7C * T7F; T7Q = T4p * T7P; T7M = T7C * T7L; cr[WS(rs, 8)] = FNMS(T4r, T7P, T7O); cr[WS(rs, 24)] = FNMS(T7I, T7L, T7G); ci[WS(rs, 8)] = FMA(T4r, T7N, T7Q); ci[WS(rs, 24)] = FMA(T7I, T7F, T7M); } } T31 = FMA(KP923879532, T30, T2X); T3r = FNMS(KP923879532, T30, T2X); T8z = FMA(KP923879532, T8i, T83); T8j = FNMS(KP923879532, T8i, T83); { E T8B, T8x, T8C, T8A; T8B = FMA(KP923879532, T8w, T8t); T8x = FNMS(KP923879532, T8w, T8t); T8C = T1a * T8z; T8A = T16 * T8z; { E T8y, T8k, T2i, T2J; T8y = T8m * T8j; T8k = T7S * T8j; ci[WS(rs, 6)] = FMA(T16, T8B, T8C); cr[WS(rs, 6)] = FNMS(T1a, T8B, T8A); ci[WS(rs, 22)] = FMA(T7S, T8x, T8y); cr[WS(rs, 22)] = FNMS(T8m, T8x, T8k); T1R = FMA(KP923879532, T1Q, T1t); T3m = FNMS(KP923879532, T1Q, T1t); T33 = FNMS(KP668178637, T28, T2h); T2i = FMA(KP668178637, T2h, T28); T2J = FNMS(KP668178637, T2I, T2z); T32 = FMA(KP668178637, T2z, T2I); T3s = T2J + T2i; T2K = T2i - T2J; } } { E T5l, T53, T5f, T4D, T4K, T4R, T56, T5g; T5l = FNMS(KP923879532, T52, T4Z); T53 = FMA(KP923879532, T52, T4Z); { E T3t, T3D, T3f, T2L; T3t = FNMS(KP831469612, T3s, T3r); T3D = FMA(KP831469612, T3s, T3r); T3f = FMA(KP831469612, T2K, T1R); T2L = FNMS(KP831469612, T2K, T1R); { E T3n, T34, T3g, T2M; T3n = T32 + T33; T34 = T32 - T33; T3g = T3e * T3f; T2M = T1c * T2L; { E T3o, T3z, T3j, T35; T3o = FNMS(KP831469612, T3n, T3m); T3z = FMA(KP831469612, T3n, T3m); T3j = FMA(KP831469612, T34, T31); T35 = FNMS(KP831469612, T34, T31); { E T3u, T3p, T3E, T3A; T3u = T3q * T3o; T3p = T3l * T3o; T3E = T3C * T3z; T3A = T3y * T3z; { E T3k, T36, T54, T55; T3k = T3e * T3j; cr[WS(rs, 5)] = FNMS(T3i, T3j, T3g); T36 = T1c * T35; cr[WS(rs, 21)] = FNMS(T2O, T35, T2M); ci[WS(rs, 13)] = FMA(T3l, T3t, T3u); cr[WS(rs, 13)] = FNMS(T3q, T3t, T3p); ci[WS(rs, 29)] = FMA(T3y, T3D, T3E); cr[WS(rs, 29)] = FNMS(T3C, T3D, T3A); ci[WS(rs, 5)] = FMA(T3i, T3f, T3k); ci[WS(rs, 21)] = FMA(T2O, T2L, T36); T5f = FMA(KP923879532, T4C, T4v); T4D = FNMS(KP923879532, T4C, T4v); T4K = FNMS(KP668178637, T4J, T4G); T54 = FMA(KP668178637, T4G, T4J); T55 = FMA(KP668178637, T4N, T4Q); T4R = FNMS(KP668178637, T4Q, T4N); T56 = T54 - T55; T5g = T54 + T55; } } } } } { E T4h, T41, T4c, T3N, T3Q, T3T, T44, T4d; T4h = FNMS(KP923879532, T40, T3Z); T41 = FMA(KP923879532, T40, T3Z); { E T57, T5b, T5h, T5p; T57 = FNMS(KP831469612, T56, T53); T5b = FMA(KP831469612, T56, T53); T5h = FNMS(KP831469612, T5g, T5f); T5p = FMA(KP831469612, T5g, T5f); { E T5m, T4S, T5i, T5q; T5m = T4K - T4R; T4S = T4K + T4R; T5i = T5e * T5h; T5q = T17 * T5p; { E T5n, T5r, T59, T4T; T5n = FMA(KP831469612, T5m, T5l); T5r = FNMS(KP831469612, T5m, T5l); T59 = FMA(KP831469612, T4S, T4D); T4T = FNMS(KP831469612, T4S, T4D); { E T5o, T5s, T5c, T5a; T5o = T5e * T5n; cr[WS(rs, 11)] = FNMS(T5k, T5n, T5i); T5s = T17 * T5r; cr[WS(rs, 27)] = FNMS(T1b, T5r, T5q); T5c = T14 * T59; T5a = T11 * T59; { E T58, T4U, T42, T43; T58 = T4W * T4T; T4U = T4s * T4T; ci[WS(rs, 11)] = FMA(T5k, T5h, T5o); ci[WS(rs, 27)] = FMA(T1b, T5p, T5s); ci[WS(rs, 3)] = FMA(T11, T5b, T5c); cr[WS(rs, 3)] = FNMS(T14, T5b, T5a); ci[WS(rs, 19)] = FMA(T4s, T57, T58); cr[WS(rs, 19)] = FNMS(T4W, T57, T4U); T4c = FNMS(KP923879532, T3M, T3L); T3N = FMA(KP923879532, T3M, T3L); T3Q = FNMS(KP198912367, T3P, T3O); T42 = FMA(KP198912367, T3O, T3P); T43 = FNMS(KP198912367, T3R, T3S); T3T = FMA(KP198912367, T3S, T3R); T44 = T42 + T43; T4d = T43 - T42; } } } } } T67 = FNMS(KP923879532, T5K, T5J); T5L = FMA(KP923879532, T5K, T5J); { E T45, T49, T4e, T4l; T45 = FNMS(KP980785280, T44, T41); T49 = FMA(KP980785280, T44, T41); T4e = FNMS(KP980785280, T4d, T4c); T4l = FMA(KP980785280, T4d, T4c); { E T4i, T3U, T4f, T4m; T4i = T3Q - T3T; T3U = T3Q + T3T; T4f = T4b * T4e; T4m = T12 * T4l; { E T4j, T4n, T47, T3V; T4j = FNMS(KP980785280, T4i, T4h); T4n = FMA(KP980785280, T4i, T4h); T47 = FMA(KP980785280, T3U, T3N); T3V = FNMS(KP980785280, T3U, T3N); { E T4k, T4o, T4a, T48; T4k = T4b * T4j; cr[WS(rs, 25)] = FNMS(T4g, T4j, T4f); T4o = T12 * T4n; cr[WS(rs, 9)] = FNMS(T15, T4n, T4m); T4a = T39 * T47; T48 = T37 * T47; { E T46, T3W, T5M, T5N; T46 = T3Y * T3V; T3W = T3K * T3V; ci[WS(rs, 25)] = FMA(T4g, T4e, T4k); ci[WS(rs, 9)] = FMA(T15, T4l, T4o); ci[WS(rs, 1)] = FMA(T37, T49, T4a); cr[WS(rs, 1)] = FNMS(T39, T49, T48); ci[WS(rs, 17)] = FMA(T3K, T45, T46); cr[WS(rs, 17)] = FNMS(T3Y, T45, T3W); T61 = FMA(KP923879532, T5w, T5v); T5x = FNMS(KP923879532, T5w, T5v); T5A = FNMS(KP198912367, T5z, T5y); T5M = FMA(KP198912367, T5y, T5z); T5N = FMA(KP198912367, T5B, T5C); T5D = FNMS(KP198912367, T5C, T5B); T5O = T5M - T5N; T62 = T5M + T5N; } } } } } } } } } } } T5V = FMA(KP980785280, T5O, T5L); T5P = FNMS(KP980785280, T5O, T5L); { E T6c, T63, T5E, T68; T6c = FMA(KP980785280, T62, T61); T63 = FNMS(KP980785280, T62, T61); T5E = T5A + T5D; T68 = T5D - T5A; { E T64, T6d, T6f, T69; T64 = T60 * T63; T6d = T6b * T6c; T6f = FNMS(KP980785280, T68, T67); T69 = FMA(KP980785280, T68, T67); { E T5F, T5S, T6a, T6g; T5F = FMA(KP980785280, T5E, T5x); T5S = FNMS(KP980785280, T5E, T5x); T6a = T60 * T69; cr[WS(rs, 15)] = FNMS(T66, T69, T64); T6g = T6b * T6f; cr[WS(rs, 31)] = FNMS(T6e, T6f, T6d); { E T5W, T5T, T5Q, T5G; T5W = T5U * T5S; T5T = T5R * T5S; T5Q = T5I * T5F; T5G = T5u * T5F; ci[WS(rs, 15)] = FMA(T66, T63, T6a); ci[WS(rs, 31)] = FMA(T6e, T6c, T6g); ci[WS(rs, 7)] = FMA(T5R, T5V, T5W); cr[WS(rs, 7)] = FNMS(T5U, T5V, T5T); ci[WS(rs, 23)] = FMA(T5u, T5P, T5Q); cr[WS(rs, 23)] = FNMS(T5I, T5P, T5G); } } } } } } } static const tw_instr twinstr[] = { {TW_CEXP, 1, 1}, {TW_CEXP, 1, 3}, {TW_CEXP, 1, 9}, {TW_CEXP, 1, 27}, {TW_NEXT, 1, 0} }; static const hc2hc_desc desc = { 32, "hb2_32", twinstr, &GENUS, {236, 98, 252, 0} }; void X(codelet_hb2_32) (planner *p) { X(khc2hc_register) (p, hb2_32, &desc); } #else /* HAVE_FMA */ /* Generated by: ../../../genfft/gen_hc2hc.native -compact -variables 4 -pipeline-latency 4 -sign 1 -twiddle-log3 -precompute-twiddles -n 32 -dif -name hb2_32 -include hb.h */ /* * This function contains 488 FP additions, 280 FP multiplications, * (or, 376 additions, 168 multiplications, 112 fused multiply/add), * 160 stack variables, 7 constants, and 128 memory accesses */ #include "hb.h" static void hb2_32(R *cr, R *ci, const R *W, stride rs, INT mb, INT me, INT ms) { DK(KP555570233, +0.555570233019602224742830813948532874374937191); DK(KP831469612, +0.831469612302545237078788377617905756738560812); DK(KP980785280, +0.980785280403230449126182236134239036973933731); DK(KP195090322, +0.195090322016128267848284868477022240927691618); DK(KP923879532, +0.923879532511286756128183189396788286822416626); DK(KP382683432, +0.382683432365089771728459984030398866761344562); DK(KP707106781, +0.707106781186547524400844362104849039284835938); { INT m; for (m = mb, W = W + ((mb - 1) * 8); m < me; m = m + 1, cr = cr + ms, ci = ci - ms, W = W + 8, MAKE_VOLATILE_STRIDE(64, rs)) { E T11, T14, T12, T15, T17, T2z, T2B, T1c, T18, T1d, T1g, T1k, T2F, T2L, T3t; E T4H, T3h, T3V, T3b, T4v, T4T, T4X, T6t, T71, T6z, T75, T81, T8x, T8f, T8z; E T2R, T2V, T8p, T8t, T4r, T4t, T53, T69, T3n, T3r, T7P, T7T, T4P, T4R, T6F; E T6R, T1f, T2X, T1j, T2Y, T1l, T31, T2d, T2Z, T49, T4h, T4c, T4i, T4d, T4n; E T4f, T4j; { E T2P, T3q, T2U, T3l, T2Q, T3p, T2T, T3m, T2D, T3g, T2K, T39, T2E, T3f, T2J; E T3a; { E T13, T1b, T16, T1a; T11 = W[0]; T14 = W[1]; T12 = W[2]; T15 = W[3]; T13 = T11 * T12; T1b = T14 * T12; T16 = T14 * T15; T1a = T11 * T15; T17 = T13 + T16; T2z = T13 - T16; T2B = T1a + T1b; T1c = T1a - T1b; T18 = W[4]; T2P = T12 * T18; T3q = T14 * T18; T2U = T15 * T18; T3l = T11 * T18; T1d = W[5]; T2Q = T15 * T1d; T3p = T11 * T1d; T2T = T12 * T1d; T3m = T14 * T1d; T1g = W[6]; T2D = T11 * T1g; T3g = T15 * T1g; T2K = T14 * T1g; T39 = T12 * T1g; T1k = W[7]; T2E = T14 * T1k; T3f = T12 * T1k; T2J = T11 * T1k; T3a = T15 * T1k; } T2F = T2D - T2E; T2L = T2J + T2K; T3t = T39 - T3a; T4H = T2J - T2K; T3h = T3f - T3g; T3V = T3f + T3g; T3b = T39 + T3a; T4v = T2D + T2E; T4T = FMA(T18, T1g, T1d * T1k); T4X = FNMS(T1d, T1g, T18 * T1k); { E T6r, T6s, T6x, T6y; T6r = T17 * T1g; T6s = T1c * T1k; T6t = T6r - T6s; T71 = T6r + T6s; T6x = T17 * T1k; T6y = T1c * T1g; T6z = T6x + T6y; T75 = T6x - T6y; } { E T7Z, T80, T8d, T8e; T7Z = T2z * T1g; T80 = T2B * T1k; T81 = T7Z + T80; T8x = T7Z - T80; T8d = T2z * T1k; T8e = T2B * T1g; T8f = T8d - T8e; T8z = T8d + T8e; T2R = T2P - T2Q; T2V = T2T + T2U; T8p = FMA(T2R, T1g, T2V * T1k); T8t = FNMS(T2V, T1g, T2R * T1k); } T4r = T2P + T2Q; T4t = T2T - T2U; T53 = FMA(T4r, T1g, T4t * T1k); T69 = FNMS(T4t, T1g, T4r * T1k); T3n = T3l + T3m; T3r = T3p - T3q; T7P = FMA(T3n, T1g, T3r * T1k); T7T = FNMS(T3r, T1g, T3n * T1k); T4P = T3l - T3m; T4R = T3p + T3q; T6F = FMA(T4P, T1g, T4R * T1k); T6R = FNMS(T4R, T1g, T4P * T1k); { E T19, T1e, T1h, T1i; T19 = T17 * T18; T1e = T1c * T1d; T1f = T19 + T1e; T2X = T19 - T1e; T1h = T17 * T1d; T1i = T1c * T18; T1j = T1h - T1i; T2Y = T1h + T1i; } T1l = FMA(T1f, T1g, T1j * T1k); T31 = FNMS(T2Y, T1g, T2X * T1k); T2d = FNMS(T1j, T1g, T1f * T1k); T2Z = FMA(T2X, T1g, T2Y * T1k); { E T47, T48, T4a, T4b; T47 = T2z * T18; T48 = T2B * T1d; T49 = T47 - T48; T4h = T47 + T48; T4a = T2z * T1d; T4b = T2B * T18; T4c = T4a + T4b; T4i = T4a - T4b; } T4d = FMA(T49, T1g, T4c * T1k); T4n = FNMS(T4i, T1g, T4h * T1k); T4f = FNMS(T4c, T1g, T49 * T1k); T4j = FMA(T4h, T1g, T4i * T1k); } { E T56, T7b, T7C, T6c, Tf, T1m, T6f, T7c, T3Y, T4I, T2t, T32, T5d, T7D, T3w; E T4w, Tu, T2e, T7g, T7F, T7j, T7G, T1B, T33, T3z, T40, T5l, T6i, T5s, T6h; E T3C, T3Z, TK, T1D, T7v, T86, T7y, T85, T1S, T35, T3O, T4C, T5F, T6J, T5M; E T6K, T3R, T4D, TZ, T1U, T7o, T89, T7r, T88, T29, T36, T3H, T4z, T5Y, T6M; E T65, T6N, T3K, T4A; { E T3, T54, T2o, T58, T2r, T5b, T6, T6a, Ta, T57, T2h, T6b, T2k, T55, Td; E T5a; { E T1, T2, T2m, T2n; T1 = cr[0]; T2 = ci[WS(rs, 15)]; T3 = T1 + T2; T54 = T1 - T2; T2m = ci[WS(rs, 27)]; T2n = cr[WS(rs, 20)]; T2o = T2m - T2n; T58 = T2m + T2n; } { E T2p, T2q, T4, T5; T2p = ci[WS(rs, 19)]; T2q = cr[WS(rs, 28)]; T2r = T2p - T2q; T5b = T2p + T2q; T4 = cr[WS(rs, 8)]; T5 = ci[WS(rs, 7)]; T6 = T4 + T5; T6a = T4 - T5; } { E T8, T9, T2f, T2g; T8 = cr[WS(rs, 4)]; T9 = ci[WS(rs, 11)]; Ta = T8 + T9; T57 = T8 - T9; T2f = ci[WS(rs, 31)]; T2g = cr[WS(rs, 16)]; T2h = T2f - T2g; T6b = T2f + T2g; } { E T2i, T2j, Tb, Tc; T2i = ci[WS(rs, 23)]; T2j = cr[WS(rs, 24)]; T2k = T2i - T2j; T55 = T2i + T2j; Tb = ci[WS(rs, 3)]; Tc = cr[WS(rs, 12)]; Td = Tb + Tc; T5a = Tb - Tc; } { E T7, Te, T2l, T2s; T56 = T54 - T55; T7b = T54 + T55; T7C = T6b - T6a; T6c = T6a + T6b; T7 = T3 + T6; Te = Ta + Td; Tf = T7 + Te; T1m = T7 - Te; { E T6d, T6e, T3W, T3X; T6d = T57 + T58; T6e = T5a + T5b; T6f = KP707106781 * (T6d - T6e); T7c = KP707106781 * (T6d + T6e); T3W = T2h - T2k; T3X = Ta - Td; T3Y = T3W - T3X; T4I = T3X + T3W; } T2l = T2h + T2k; T2s = T2o + T2r; T2t = T2l - T2s; T32 = T2l + T2s; { E T59, T5c, T3u, T3v; T59 = T57 - T58; T5c = T5a - T5b; T5d = KP707106781 * (T59 + T5c); T7D = KP707106781 * (T59 - T5c); T3u = T3 - T6; T3v = T2r - T2o; T3w = T3u - T3v; T4w = T3u + T3v; } } } { E Ti, T5p, T1w, T5n, T1z, T5q, Tl, T5m, Tp, T5i, T1p, T5g, T1s, T5j, Ts; E T5f; { E Tg, Th, T1u, T1v; Tg = cr[WS(rs, 2)]; Th = ci[WS(rs, 13)]; Ti = Tg + Th; T5p = Tg - Th; T1u = ci[WS(rs, 29)]; T1v = cr[WS(rs, 18)]; T1w = T1u - T1v; T5n = T1u + T1v; } { E T1x, T1y, Tj, Tk; T1x = ci[WS(rs, 21)]; T1y = cr[WS(rs, 26)]; T1z = T1x - T1y; T5q = T1x + T1y; Tj = cr[WS(rs, 10)]; Tk = ci[WS(rs, 5)]; Tl = Tj + Tk; T5m = Tj - Tk; } { E Tn, To, T1n, T1o; Tn = ci[WS(rs, 1)]; To = cr[WS(rs, 14)]; Tp = Tn + To; T5i = Tn - To; T1n = ci[WS(rs, 17)]; T1o = cr[WS(rs, 30)]; T1p = T1n - T1o; T5g = T1n + T1o; } { E T1q, T1r, Tq, Tr; T1q = ci[WS(rs, 25)]; T1r = cr[WS(rs, 22)]; T1s = T1q - T1r; T5j = T1q + T1r; Tq = cr[WS(rs, 6)]; Tr = ci[WS(rs, 9)]; Ts = Tq + Tr; T5f = Tq - Tr; } { E Tm, Tt, T7e, T7f; Tm = Ti + Tl; Tt = Tp + Ts; Tu = Tm + Tt; T2e = Tm - Tt; T7e = T5p + T5q; T7f = T5n - T5m; T7g = FNMS(KP923879532, T7f, KP382683432 * T7e); T7F = FMA(KP382683432, T7f, KP923879532 * T7e); } { E T7h, T7i, T1t, T1A; T7h = T5i + T5j; T7i = T5f + T5g; T7j = FNMS(KP923879532, T7i, KP382683432 * T7h); T7G = FMA(KP382683432, T7i, KP923879532 * T7h); T1t = T1p + T1s; T1A = T1w + T1z; T1B = T1t - T1A; T33 = T1A + T1t; } { E T3x, T3y, T5h, T5k; T3x = T1p - T1s; T3y = Tp - Ts; T3z = T3x - T3y; T40 = T3y + T3x; T5h = T5f - T5g; T5k = T5i - T5j; T5l = FNMS(KP382683432, T5k, KP923879532 * T5h); T6i = FMA(KP382683432, T5h, KP923879532 * T5k); } { E T5o, T5r, T3A, T3B; T5o = T5m + T5n; T5r = T5p - T5q; T5s = FMA(KP923879532, T5o, KP382683432 * T5r); T6h = FNMS(KP382683432, T5o, KP923879532 * T5r); T3A = Ti - Tl; T3B = T1w - T1z; T3C = T3A + T3B; T3Z = T3A - T3B; } } { E Ty, T5v, TB, T5G, T1J, T5w, T1G, T5H, TI, T5K, T1Q, T5D, TF, T5J, T1N; E T5A; { E Tw, Tx, T1E, T1F; Tw = cr[WS(rs, 1)]; Tx = ci[WS(rs, 14)]; Ty = Tw + Tx; T5v = Tw - Tx; { E Tz, TA, T1H, T1I; Tz = cr[WS(rs, 9)]; TA = ci[WS(rs, 6)]; TB = Tz + TA; T5G = Tz - TA; T1H = ci[WS(rs, 22)]; T1I = cr[WS(rs, 25)]; T1J = T1H - T1I; T5w = T1H + T1I; } T1E = ci[WS(rs, 30)]; T1F = cr[WS(rs, 17)]; T1G = T1E - T1F; T5H = T1E + T1F; { E TG, TH, T5B, T1O, T1P, T5C; TG = ci[WS(rs, 2)]; TH = cr[WS(rs, 13)]; T5B = TG - TH; T1O = ci[WS(rs, 18)]; T1P = cr[WS(rs, 29)]; T5C = T1O + T1P; TI = TG + TH; T5K = T5B + T5C; T1Q = T1O - T1P; T5D = T5B - T5C; } { E TD, TE, T5y, T1L, T1M, T5z; TD = cr[WS(rs, 5)]; TE = ci[WS(rs, 10)]; T5y = TD - TE; T1L = ci[WS(rs, 26)]; T1M = cr[WS(rs, 21)]; T5z = T1L + T1M; TF = TD + TE; T5J = T5y + T5z; T1N = T1L - T1M; T5A = T5y - T5z; } } { E TC, TJ, T7t, T7u; TC = Ty + TB; TJ = TF + TI; TK = TC + TJ; T1D = TC - TJ; T7t = T5H - T5G; T7u = KP707106781 * (T5A - T5D); T7v = T7t + T7u; T86 = T7t - T7u; } { E T7w, T7x, T1K, T1R; T7w = T5v + T5w; T7x = KP707106781 * (T5J + T5K); T7y = T7w - T7x; T85 = T7w + T7x; T1K = T1G + T1J; T1R = T1N + T1Q; T1S = T1K - T1R; T35 = T1K + T1R; } { E T3M, T3N, T5x, T5E; T3M = T1G - T1J; T3N = TF - TI; T3O = T3M - T3N; T4C = T3N + T3M; T5x = T5v - T5w; T5E = KP707106781 * (T5A + T5D); T5F = T5x - T5E; T6J = T5x + T5E; } { E T5I, T5L, T3P, T3Q; T5I = T5G + T5H; T5L = KP707106781 * (T5J - T5K); T5M = T5I - T5L; T6K = T5I + T5L; T3P = Ty - TB; T3Q = T1Q - T1N; T3R = T3P - T3Q; T4D = T3P + T3Q; } } { E TN, T5O, TQ, T5Z, T20, T5P, T1X, T60, TX, T63, T27, T5W, TU, T62, T24; E T5T; { E TL, TM, T1V, T1W; TL = ci[0]; TM = cr[WS(rs, 15)]; TN = TL + TM; T5O = TL - TM; { E TO, TP, T1Y, T1Z; TO = cr[WS(rs, 7)]; TP = ci[WS(rs, 8)]; TQ = TO + TP; T5Z = TO - TP; T1Y = ci[WS(rs, 24)]; T1Z = cr[WS(rs, 23)]; T20 = T1Y - T1Z; T5P = T1Y + T1Z; } T1V = ci[WS(rs, 16)]; T1W = cr[WS(rs, 31)]; T1X = T1V - T1W; T60 = T1V + T1W; { E TV, TW, T5U, T25, T26, T5V; TV = ci[WS(rs, 4)]; TW = cr[WS(rs, 11)]; T5U = TV - TW; T25 = ci[WS(rs, 20)]; T26 = cr[WS(rs, 27)]; T5V = T25 + T26; TX = TV + TW; T63 = T5U + T5V; T27 = T25 - T26; T5W = T5U - T5V; } { E TS, TT, T5R, T22, T23, T5S; TS = cr[WS(rs, 3)]; TT = ci[WS(rs, 12)]; T5R = TS - TT; T22 = ci[WS(rs, 28)]; T23 = cr[WS(rs, 19)]; T5S = T22 + T23; TU = TS + TT; T62 = T5R + T5S; T24 = T22 - T23; T5T = T5R - T5S; } } { E TR, TY, T7m, T7n; TR = TN + TQ; TY = TU + TX; TZ = TR + TY; T1U = TR - TY; T7m = KP707106781 * (T5T - T5W); T7n = T5Z + T60; T7o = T7m - T7n; T89 = T7n + T7m; } { E T7p, T7q, T21, T28; T7p = T5O + T5P; T7q = KP707106781 * (T62 + T63); T7r = T7p - T7q; T88 = T7p + T7q; T21 = T1X + T20; T28 = T24 + T27; T29 = T21 - T28; T36 = T21 + T28; } { E T3F, T3G, T5Q, T5X; T3F = T1X - T20; T3G = TU - TX; T3H = T3F - T3G; T4z = T3G + T3F; T5Q = T5O - T5P; T5X = KP707106781 * (T5T + T5W); T5Y = T5Q - T5X; T6M = T5Q + T5X; } { E T61, T64, T3I, T3J; T61 = T5Z - T60; T64 = KP707106781 * (T62 - T63); T65 = T61 - T64; T6N = T61 + T64; T3I = TN - TQ; T3J = T27 - T24; T3K = T3I - T3J; T4A = T3I + T3J; } } { E Tv, T10, T30, T34, T37, T38; Tv = Tf + Tu; T10 = TK + TZ; T30 = Tv - T10; T34 = T32 + T33; T37 = T35 + T36; T38 = T34 - T37; cr[0] = Tv + T10; ci[0] = T34 + T37; cr[WS(rs, 16)] = FNMS(T31, T38, T2Z * T30); ci[WS(rs, 16)] = FMA(T31, T30, T2Z * T38); } { E T3e, T3o, T3k, T3s; { E T3c, T3d, T3i, T3j; T3c = Tf - Tu; T3d = T36 - T35; T3e = T3c - T3d; T3o = T3c + T3d; T3i = T32 - T33; T3j = TK - TZ; T3k = T3i - T3j; T3s = T3j + T3i; } cr[WS(rs, 24)] = FNMS(T3h, T3k, T3b * T3e); ci[WS(rs, 24)] = FMA(T3b, T3k, T3h * T3e); cr[WS(rs, 8)] = FNMS(T3r, T3s, T3n * T3o); ci[WS(rs, 8)] = FMA(T3n, T3s, T3r * T3o); } { E T1C, T2u, T2M, T2G, T2x, T2H, T2b, T2N; T1C = T1m + T1B; T2u = T2e + T2t; T2M = T2t - T2e; T2G = T1m - T1B; { E T2v, T2w, T1T, T2a; T2v = T1D + T1S; T2w = T29 - T1U; T2x = KP707106781 * (T2v + T2w); T2H = KP707106781 * (T2w - T2v); T1T = T1D - T1S; T2a = T1U + T29; T2b = KP707106781 * (T1T + T2a); T2N = KP707106781 * (T1T - T2a); } { E T2c, T2y, T2S, T2W; T2c = T1C - T2b; T2y = T2u - T2x; cr[WS(rs, 20)] = FNMS(T2d, T2y, T1l * T2c); ci[WS(rs, 20)] = FMA(T2d, T2c, T1l * T2y); T2S = T2G + T2H; T2W = T2M + T2N; cr[WS(rs, 12)] = FNMS(T2V, T2W, T2R * T2S); ci[WS(rs, 12)] = FMA(T2R, T2W, T2V * T2S); } { E T2A, T2C, T2I, T2O; T2A = T1C + T2b; T2C = T2u + T2x; cr[WS(rs, 4)] = FNMS(T2B, T2C, T2z * T2A); ci[WS(rs, 4)] = FMA(T2B, T2A, T2z * T2C); T2I = T2G - T2H; T2O = T2M - T2N; cr[WS(rs, 28)] = FNMS(T2L, T2O, T2F * T2I); ci[WS(rs, 28)] = FMA(T2F, T2O, T2L * T2I); } } { E T4y, T4U, T4K, T4Y, T4F, T4Z, T4N, T4V, T4x, T4J; T4x = KP707106781 * (T3Z + T40); T4y = T4w - T4x; T4U = T4w + T4x; T4J = KP707106781 * (T3C + T3z); T4K = T4I - T4J; T4Y = T4I + T4J; { E T4B, T4E, T4L, T4M; T4B = FNMS(KP382683432, T4A, KP923879532 * T4z); T4E = FMA(KP923879532, T4C, KP382683432 * T4D); T4F = T4B - T4E; T4Z = T4E + T4B; T4L = FNMS(KP382683432, T4C, KP923879532 * T4D); T4M = FMA(KP382683432, T4z, KP923879532 * T4A); T4N = T4L - T4M; T4V = T4L + T4M; } { E T4G, T4O, T51, T52; T4G = T4y - T4F; T4O = T4K - T4N; cr[WS(rs, 26)] = FNMS(T4H, T4O, T4v * T4G); ci[WS(rs, 26)] = FMA(T4H, T4G, T4v * T4O); T51 = T4U + T4V; T52 = T4Y + T4Z; cr[WS(rs, 2)] = FNMS(T1c, T52, T17 * T51); ci[WS(rs, 2)] = FMA(T17, T52, T1c * T51); } { E T4Q, T4S, T4W, T50; T4Q = T4y + T4F; T4S = T4K + T4N; cr[WS(rs, 10)] = FNMS(T4R, T4S, T4P * T4Q); ci[WS(rs, 10)] = FMA(T4R, T4Q, T4P * T4S); T4W = T4U - T4V; T50 = T4Y - T4Z; cr[WS(rs, 18)] = FNMS(T4X, T50, T4T * T4W); ci[WS(rs, 18)] = FMA(T4T, T50, T4X * T4W); } } { E T3E, T4k, T42, T4o, T3T, T4p, T45, T4l, T3D, T41; T3D = KP707106781 * (T3z - T3C); T3E = T3w - T3D; T4k = T3w + T3D; T41 = KP707106781 * (T3Z - T40); T42 = T3Y - T41; T4o = T3Y + T41; { E T3L, T3S, T43, T44; T3L = FNMS(KP923879532, T3K, KP382683432 * T3H); T3S = FMA(KP382683432, T3O, KP923879532 * T3R); T3T = T3L - T3S; T4p = T3S + T3L; T43 = FNMS(KP923879532, T3O, KP382683432 * T3R); T44 = FMA(KP923879532, T3H, KP382683432 * T3K); T45 = T43 - T44; T4l = T43 + T44; } { E T3U, T46, T4s, T4u; T3U = T3E - T3T; T46 = T42 - T45; cr[WS(rs, 30)] = FNMS(T3V, T46, T3t * T3U); ci[WS(rs, 30)] = FMA(T3V, T3U, T3t * T46); T4s = T4k + T4l; T4u = T4o + T4p; cr[WS(rs, 6)] = FNMS(T4t, T4u, T4r * T4s); ci[WS(rs, 6)] = FMA(T4r, T4u, T4t * T4s); } { E T4e, T4g, T4m, T4q; T4e = T3E + T3T; T4g = T42 + T45; cr[WS(rs, 14)] = FNMS(T4f, T4g, T4d * T4e); ci[WS(rs, 14)] = FMA(T4f, T4e, T4d * T4g); T4m = T4k - T4l; T4q = T4o - T4p; cr[WS(rs, 22)] = FNMS(T4n, T4q, T4j * T4m); ci[WS(rs, 22)] = FMA(T4j, T4q, T4n * T4m); } } { E T6I, T72, T6X, T73, T6P, T77, T6U, T76; { E T6G, T6H, T6V, T6W; T6G = T56 + T5d; T6H = T6h + T6i; T6I = T6G + T6H; T72 = T6G - T6H; T6V = FMA(KP195090322, T6J, KP980785280 * T6K); T6W = FNMS(KP195090322, T6M, KP980785280 * T6N); T6X = T6V + T6W; T73 = T6W - T6V; } { E T6L, T6O, T6S, T6T; T6L = FNMS(KP195090322, T6K, KP980785280 * T6J); T6O = FMA(KP980785280, T6M, KP195090322 * T6N); T6P = T6L + T6O; T77 = T6L - T6O; T6S = T6c + T6f; T6T = T5s + T5l; T6U = T6S + T6T; T76 = T6S - T6T; } { E T6Q, T6Y, T79, T7a; T6Q = T6I - T6P; T6Y = T6U - T6X; cr[WS(rs, 17)] = FNMS(T6R, T6Y, T6F * T6Q); ci[WS(rs, 17)] = FMA(T6R, T6Q, T6F * T6Y); T79 = T72 + T73; T7a = T76 + T77; cr[WS(rs, 9)] = FNMS(T1d, T7a, T18 * T79); ci[WS(rs, 9)] = FMA(T18, T7a, T1d * T79); } { E T6Z, T70, T74, T78; T6Z = T6I + T6P; T70 = T6U + T6X; cr[WS(rs, 1)] = FNMS(T14, T70, T11 * T6Z); ci[WS(rs, 1)] = FMA(T14, T6Z, T11 * T70); T74 = T72 - T73; T78 = T76 - T77; cr[WS(rs, 25)] = FNMS(T75, T78, T71 * T74); ci[WS(rs, 25)] = FMA(T71, T78, T75 * T74); } } { E T84, T8q, T8l, T8r, T8b, T8v, T8i, T8u; { E T82, T83, T8j, T8k; T82 = T7b + T7c; T83 = T7F + T7G; T84 = T82 - T83; T8q = T82 + T83; T8j = FMA(KP195090322, T86, KP980785280 * T85); T8k = FMA(KP195090322, T89, KP980785280 * T88); T8l = T8j - T8k; T8r = T8j + T8k; } { E T87, T8a, T8g, T8h; T87 = FNMS(KP980785280, T86, KP195090322 * T85); T8a = FNMS(KP980785280, T89, KP195090322 * T88); T8b = T87 + T8a; T8v = T87 - T8a; T8g = T7C - T7D; T8h = T7g - T7j; T8i = T8g + T8h; T8u = T8g - T8h; } { E T8c, T8m, T8y, T8A; T8c = T84 - T8b; T8m = T8i - T8l; cr[WS(rs, 23)] = FNMS(T8f, T8m, T81 * T8c); ci[WS(rs, 23)] = FMA(T8f, T8c, T81 * T8m); T8y = T8q + T8r; T8A = T8u - T8v; cr[WS(rs, 31)] = FNMS(T8z, T8A, T8x * T8y); ci[WS(rs, 31)] = FMA(T8x, T8A, T8z * T8y); } { E T8n, T8o, T8s, T8w; T8n = T84 + T8b; T8o = T8i + T8l; cr[WS(rs, 7)] = FNMS(T1j, T8o, T1f * T8n); ci[WS(rs, 7)] = FMA(T1j, T8n, T1f * T8o); T8s = T8q - T8r; T8w = T8u + T8v; cr[WS(rs, 15)] = FNMS(T8t, T8w, T8p * T8s); ci[WS(rs, 15)] = FMA(T8p, T8w, T8t * T8s); } } { E T5u, T6u, T6n, T6v, T67, T6B, T6k, T6A; { E T5e, T5t, T6l, T6m; T5e = T56 - T5d; T5t = T5l - T5s; T5u = T5e + T5t; T6u = T5e - T5t; T6l = FMA(KP831469612, T5F, KP555570233 * T5M); T6m = FNMS(KP831469612, T5Y, KP555570233 * T65); T6n = T6l + T6m; T6v = T6m - T6l; } { E T5N, T66, T6g, T6j; T5N = FNMS(KP831469612, T5M, KP555570233 * T5F); T66 = FMA(KP555570233, T5Y, KP831469612 * T65); T67 = T5N + T66; T6B = T5N - T66; T6g = T6c - T6f; T6j = T6h - T6i; T6k = T6g + T6j; T6A = T6g - T6j; } { E T68, T6o, T6D, T6E; T68 = T5u - T67; T6o = T6k - T6n; cr[WS(rs, 21)] = FNMS(T69, T6o, T53 * T68); ci[WS(rs, 21)] = FMA(T69, T68, T53 * T6o); T6D = T6u + T6v; T6E = T6A + T6B; cr[WS(rs, 13)] = FNMS(T4c, T6E, T49 * T6D); ci[WS(rs, 13)] = FMA(T49, T6E, T4c * T6D); } { E T6p, T6q, T6w, T6C; T6p = T5u + T67; T6q = T6k + T6n; cr[WS(rs, 5)] = FNMS(T4i, T6q, T4h * T6p); ci[WS(rs, 5)] = FMA(T4i, T6p, T4h * T6q); T6w = T6u - T6v; T6C = T6A - T6B; cr[WS(rs, 29)] = FNMS(T6z, T6C, T6t * T6w); ci[WS(rs, 29)] = FMA(T6t, T6C, T6z * T6w); } } { E T7l, T7Q, T7L, T7R, T7A, T7V, T7I, T7U; { E T7d, T7k, T7J, T7K; T7d = T7b - T7c; T7k = T7g + T7j; T7l = T7d - T7k; T7Q = T7d + T7k; T7J = FNMS(KP555570233, T7v, KP831469612 * T7y); T7K = FMA(KP555570233, T7o, KP831469612 * T7r); T7L = T7J - T7K; T7R = T7J + T7K; } { E T7s, T7z, T7E, T7H; T7s = FNMS(KP555570233, T7r, KP831469612 * T7o); T7z = FMA(KP831469612, T7v, KP555570233 * T7y); T7A = T7s - T7z; T7V = T7z + T7s; T7E = T7C + T7D; T7H = T7F - T7G; T7I = T7E - T7H; T7U = T7E + T7H; } { E T7B, T7M, T7X, T7Y; T7B = T7l - T7A; T7M = T7I - T7L; cr[WS(rs, 27)] = FNMS(T1k, T7M, T1g * T7B); ci[WS(rs, 27)] = FMA(T1k, T7B, T1g * T7M); T7X = T7Q + T7R; T7Y = T7U + T7V; cr[WS(rs, 3)] = FNMS(T15, T7Y, T12 * T7X); ci[WS(rs, 3)] = FMA(T12, T7Y, T15 * T7X); } { E T7N, T7O, T7S, T7W; T7N = T7l + T7A; T7O = T7I + T7L; cr[WS(rs, 11)] = FNMS(T2Y, T7O, T2X * T7N); ci[WS(rs, 11)] = FMA(T2Y, T7N, T2X * T7O); T7S = T7Q - T7R; T7W = T7U - T7V; cr[WS(rs, 19)] = FNMS(T7T, T7W, T7P * T7S); ci[WS(rs, 19)] = FMA(T7P, T7W, T7T * T7S); } } } } } } static const tw_instr twinstr[] = { {TW_CEXP, 1, 1}, {TW_CEXP, 1, 3}, {TW_CEXP, 1, 9}, {TW_CEXP, 1, 27}, {TW_NEXT, 1, 0} }; static const hc2hc_desc desc = { 32, "hb2_32", twinstr, &GENUS, {376, 168, 112, 0} }; void X(codelet_hb2_32) (planner *p) { X(khc2hc_register) (p, hb2_32, &desc); } #endif /* HAVE_FMA */ fftw-3.3.3/rdft/scalar/r2cb/hb2_5.c0000644000175000017500000001737212054410771013521 00000000000000/* * Copyright (c) 2003, 2007-11 Matteo Frigo * Copyright (c) 2003, 2007-11 Massachusetts Institute of Technology * * This program is free software; you can redistribute it and/or modify * it under the terms of the GNU General Public License as published by * the Free Software Foundation; either version 2 of the License, or * (at your option) any later version. * * This program is distributed in the hope that it will be useful, * but WITHOUT ANY WARRANTY; without even the implied warranty of * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the * GNU General Public License for more details. * * You should have received a copy of the GNU General Public License * along with this program; if not, write to the Free Software * Foundation, Inc., 51 Franklin Street, Fifth Floor, Boston, MA 02110-1301 USA * */ /* This file was automatically generated --- DO NOT EDIT */ /* Generated on Sun Nov 25 07:41:29 EST 2012 */ #include "codelet-rdft.h" #ifdef HAVE_FMA /* Generated by: ../../../genfft/gen_hc2hc.native -fma -reorder-insns -schedule-for-pipeline -compact -variables 4 -pipeline-latency 4 -sign 1 -twiddle-log3 -precompute-twiddles -n 5 -dif -name hb2_5 -include hb.h */ /* * This function contains 44 FP additions, 40 FP multiplications, * (or, 14 additions, 10 multiplications, 30 fused multiply/add), * 51 stack variables, 4 constants, and 20 memory accesses */ #include "hb.h" static void hb2_5(R *cr, R *ci, const R *W, stride rs, INT mb, INT me, INT ms) { DK(KP951056516, +0.951056516295153572116439333379382143405698634); DK(KP559016994, +0.559016994374947424102293417182819058860154590); DK(KP250000000, +0.250000000000000000000000000000000000000000000); DK(KP618033988, +0.618033988749894848204586834365638117720309180); { INT m; for (m = mb, W = W + ((mb - 1) * 4); m < me; m = m + 1, cr = cr + ms, ci = ci - ms, W = W + 4, MAKE_VOLATILE_STRIDE(10, rs)) { E T9, TB, Tz, Tm, T1, TG, TO, TJ, TC, Tn, Tg, To, Tf, Tw, TQ; E T8, Tb, Th, Ta, Ti, Tp; T9 = W[0]; TB = W[3]; Tz = W[2]; Tm = W[1]; { E T4, Tu, T5, T6; T1 = cr[0]; { E TF, TA, T2, T3; TF = T9 * TB; TA = T9 * Tz; T2 = cr[WS(rs, 1)]; T3 = ci[0]; TG = FMA(Tm, Tz, TF); TO = FNMS(Tm, Tz, TF); TJ = FMA(Tm, TB, TA); TC = FNMS(Tm, TB, TA); T4 = T2 + T3; Tu = T2 - T3; T5 = cr[WS(rs, 2)]; T6 = ci[WS(rs, 1)]; } Tn = ci[WS(rs, 4)]; { E Td, Te, T7, Tv; Td = ci[WS(rs, 3)]; Te = cr[WS(rs, 4)]; T7 = T5 + T6; Tv = T5 - T6; Tg = ci[WS(rs, 2)]; To = Td - Te; Tf = Td + Te; Tw = FMA(KP618033988, Tv, Tu); TQ = FNMS(KP618033988, Tu, Tv); T8 = T4 + T7; Tb = T4 - T7; Th = cr[WS(rs, 3)]; } } cr[0] = T1 + T8; Ta = FNMS(KP250000000, T8, T1); Ti = Tg + Th; Tp = Tg - Th; { E Tc, TK, Ts, Tq; Tc = FMA(KP559016994, Tb, Ta); TK = FNMS(KP559016994, Tb, Ta); Ts = To - Tp; Tq = To + Tp; { E Tj, TL, Tr, TM, TT; Tj = FMA(KP618033988, Ti, Tf); TL = FNMS(KP618033988, Tf, Ti); ci[0] = Tn + Tq; Tr = FNMS(KP250000000, Tq, Tn); TM = FMA(KP951056516, TL, TK); TT = FNMS(KP951056516, TL, TK); { E Tk, TD, Tt, TP; Tk = FNMS(KP951056516, Tj, Tc); TD = FMA(KP951056516, Tj, Tc); Tt = FMA(KP559016994, Ts, Tr); TP = FNMS(KP559016994, Ts, Tr); { E TW, TU, TS, TN; TW = TB * TT; TU = Tz * TT; TS = TO * TM; TN = TJ * TM; { E TI, TE, Ty, Tl; TI = TG * TD; TE = TC * TD; Ty = Tm * Tk; Tl = T9 * Tk; { E TR, TV, Tx, TH; TR = FNMS(KP951056516, TQ, TP); TV = FMA(KP951056516, TQ, TP); Tx = FMA(KP951056516, Tw, Tt); TH = FNMS(KP951056516, Tw, Tt); ci[WS(rs, 3)] = FMA(Tz, TV, TW); cr[WS(rs, 3)] = FNMS(TB, TV, TU); ci[WS(rs, 2)] = FMA(TJ, TR, TS); cr[WS(rs, 2)] = FNMS(TO, TR, TN); ci[WS(rs, 4)] = FMA(TC, TH, TI); cr[WS(rs, 4)] = FNMS(TG, TH, TE); ci[WS(rs, 1)] = FMA(T9, Tx, Ty); cr[WS(rs, 1)] = FNMS(Tm, Tx, Tl); } } } } } } } } } static const tw_instr twinstr[] = { {TW_CEXP, 1, 1}, {TW_CEXP, 1, 3}, {TW_NEXT, 1, 0} }; static const hc2hc_desc desc = { 5, "hb2_5", twinstr, &GENUS, {14, 10, 30, 0} }; void X(codelet_hb2_5) (planner *p) { X(khc2hc_register) (p, hb2_5, &desc); } #else /* HAVE_FMA */ /* Generated by: ../../../genfft/gen_hc2hc.native -compact -variables 4 -pipeline-latency 4 -sign 1 -twiddle-log3 -precompute-twiddles -n 5 -dif -name hb2_5 -include hb.h */ /* * This function contains 44 FP additions, 32 FP multiplications, * (or, 30 additions, 18 multiplications, 14 fused multiply/add), * 33 stack variables, 4 constants, and 20 memory accesses */ #include "hb.h" static void hb2_5(R *cr, R *ci, const R *W, stride rs, INT mb, INT me, INT ms) { DK(KP250000000, +0.250000000000000000000000000000000000000000000); DK(KP587785252, +0.587785252292473129168705954639072768597652438); DK(KP951056516, +0.951056516295153572116439333379382143405698634); DK(KP559016994, +0.559016994374947424102293417182819058860154590); { INT m; for (m = mb, W = W + ((mb - 1) * 4); m < me; m = m + 1, cr = cr + ms, ci = ci - ms, W = W + 4, MAKE_VOLATILE_STRIDE(10, rs)) { E Th, Tk, Ti, Tl, Tn, TP, Tx, TN; { E Tj, Tw, Tm, Tv; Th = W[0]; Tk = W[1]; Ti = W[2]; Tl = W[3]; Tj = Th * Ti; Tw = Tk * Ti; Tm = Tk * Tl; Tv = Th * Tl; Tn = Tj + Tm; TP = Tv + Tw; Tx = Tv - Tw; TN = Tj - Tm; } { E T1, Tp, TK, TA, T8, To, T9, Tt, TI, TC, Tg, TB; { E T4, Ty, T7, Tz; T1 = cr[0]; { E T2, T3, T5, T6; T2 = cr[WS(rs, 1)]; T3 = ci[0]; T4 = T2 + T3; Ty = T2 - T3; T5 = cr[WS(rs, 2)]; T6 = ci[WS(rs, 1)]; T7 = T5 + T6; Tz = T5 - T6; } Tp = KP559016994 * (T4 - T7); TK = FMA(KP951056516, Ty, KP587785252 * Tz); TA = FNMS(KP951056516, Tz, KP587785252 * Ty); T8 = T4 + T7; To = FNMS(KP250000000, T8, T1); } { E Tc, Tr, Tf, Ts; T9 = ci[WS(rs, 4)]; { E Ta, Tb, Td, Te; Ta = ci[WS(rs, 3)]; Tb = cr[WS(rs, 4)]; Tc = Ta - Tb; Tr = Ta + Tb; Td = ci[WS(rs, 2)]; Te = cr[WS(rs, 3)]; Tf = Td - Te; Ts = Td + Te; } Tt = FNMS(KP951056516, Ts, KP587785252 * Tr); TI = FMA(KP951056516, Tr, KP587785252 * Ts); TC = KP559016994 * (Tc - Tf); Tg = Tc + Tf; TB = FNMS(KP250000000, Tg, T9); } cr[0] = T1 + T8; ci[0] = T9 + Tg; { E Tu, TF, TE, TG, Tq, TD; Tq = To - Tp; Tu = Tq - Tt; TF = Tq + Tt; TD = TB - TC; TE = TA + TD; TG = TD - TA; cr[WS(rs, 2)] = FNMS(Tx, TE, Tn * Tu); ci[WS(rs, 2)] = FMA(Tn, TE, Tx * Tu); cr[WS(rs, 3)] = FNMS(Tl, TG, Ti * TF); ci[WS(rs, 3)] = FMA(Ti, TG, Tl * TF); } { E TJ, TO, TM, TQ, TH, TL; TH = Tp + To; TJ = TH - TI; TO = TH + TI; TL = TC + TB; TM = TK + TL; TQ = TL - TK; cr[WS(rs, 1)] = FNMS(Tk, TM, Th * TJ); ci[WS(rs, 1)] = FMA(Th, TM, Tk * TJ); cr[WS(rs, 4)] = FNMS(TP, TQ, TN * TO); ci[WS(rs, 4)] = FMA(TN, TQ, TP * TO); } } } } } static const tw_instr twinstr[] = { {TW_CEXP, 1, 1}, {TW_CEXP, 1, 3}, {TW_NEXT, 1, 0} }; static const hc2hc_desc desc = { 5, "hb2_5", twinstr, &GENUS, {30, 18, 14, 0} }; void X(codelet_hb2_5) (planner *p) { X(khc2hc_register) (p, hb2_5, &desc); } #endif /* HAVE_FMA */ fftw-3.3.3/rdft/scalar/r2cb/hb2_20.c0000644000175000017500000007455712054410772013607 00000000000000/* * Copyright (c) 2003, 2007-11 Matteo Frigo * Copyright (c) 2003, 2007-11 Massachusetts Institute of Technology * * This program is free software; you can redistribute it and/or modify * it under the terms of the GNU General Public License as published by * the Free Software Foundation; either version 2 of the License, or * (at your option) any later version. * * This program is distributed in the hope that it will be useful, * but WITHOUT ANY WARRANTY; without even the implied warranty of * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the * GNU General Public License for more details. * * You should have received a copy of the GNU General Public License * along with this program; if not, write to the Free Software * Foundation, Inc., 51 Franklin Street, Fifth Floor, Boston, MA 02110-1301 USA * */ /* This file was automatically generated --- DO NOT EDIT */ /* Generated on Sun Nov 25 07:41:29 EST 2012 */ #include "codelet-rdft.h" #ifdef HAVE_FMA /* Generated by: ../../../genfft/gen_hc2hc.native -fma -reorder-insns -schedule-for-pipeline -compact -variables 4 -pipeline-latency 4 -sign 1 -twiddle-log3 -precompute-twiddles -n 20 -dif -name hb2_20 -include hb.h */ /* * This function contains 276 FP additions, 198 FP multiplications, * (or, 136 additions, 58 multiplications, 140 fused multiply/add), * 153 stack variables, 4 constants, and 80 memory accesses */ #include "hb.h" static void hb2_20(R *cr, R *ci, const R *W, stride rs, INT mb, INT me, INT ms) { DK(KP951056516, +0.951056516295153572116439333379382143405698634); DK(KP559016994, +0.559016994374947424102293417182819058860154590); DK(KP618033988, +0.618033988749894848204586834365638117720309180); DK(KP250000000, +0.250000000000000000000000000000000000000000000); { INT m; for (m = mb, W = W + ((mb - 1) * 8); m < me; m = m + 1, cr = cr + ms, ci = ci - ms, W = W + 8, MAKE_VOLATILE_STRIDE(40, rs)) { E T1S, T1O, T1s, TI, T24, T1Y, T2g, T2k, TS, TR, T1I, T26, T1o, T20, T1F; E T25, TT, T1Z; { E TD, TH, TE, T1L, T1N, T1X, TG, T1V, T2Y, T2b, T29, T2s, T36, T3e, T31; E T2o, T3b, T5b, T2c, T2U, T4y, T4u, T2f, T5g, T47, T5p, T4b, T5l; { E T1r, TF, T2T, T1M, T1R, T2X, T2r, T4x; TD = W[0]; TH = W[3]; TE = W[2]; T1L = W[6]; T1N = W[7]; T1r = TD * TH; TF = TD * TE; T2T = TE * T1L; T1M = TD * T1L; T1R = TD * T1N; T2X = TE * T1N; T1X = W[5]; TG = W[1]; T1V = W[4]; T2Y = FNMS(TH, T1L, T2X); T2r = TD * T1X; { E T23, T2n, T1W, T2a; T23 = TE * T1X; T1S = FNMS(TG, T1L, T1R); T1O = FMA(TG, T1N, T1M); T2b = FMA(TG, TE, T1r); T1s = FNMS(TG, TE, T1r); T29 = FNMS(TG, TH, TF); TI = FMA(TG, TH, TF); T2n = TD * T1V; T1W = TE * T1V; T2s = FMA(TG, T1V, T2r); T36 = FNMS(TG, T1V, T2r); T3e = FMA(TH, T1V, T23); T24 = FNMS(TH, T1V, T23); T2a = T29 * T1V; T31 = FMA(TG, T1X, T2n); T2o = FNMS(TG, T1X, T2n); T3b = FNMS(TH, T1X, T1W); T1Y = FMA(TH, T1X, T1W); T5b = FNMS(T2b, T1X, T2a); T2c = FMA(T2b, T1X, T2a); T2U = FMA(TH, T1N, T2T); } T4x = T29 * T1N; { E T4t, T2d, T2j, T2e; T4t = T29 * T1L; T2e = T29 * T1X; T4y = FNMS(T2b, T1L, T4x); T4u = FMA(T2b, T1N, T4t); T2f = FNMS(T2b, T1V, T2e); T5g = FMA(T2b, T1V, T2e); T2d = T2c * T1L; T2j = T2c * T1N; T47 = TI * T1V; T2g = FMA(T2f, T1N, T2d); T2k = FNMS(T2f, T1L, T2j); T5p = TI * T1N; T4b = TI * T1X; T5l = TI * T1L; } } { E T4f, T48, T4c, T4k, T5m, T5q, T3j, T4B, T7, TJ, T4V, T3V, T1z, T2H, T3x; E T42, T18, T3q, T43, T1n, T2D, T53, T52, T2A, T1H, T4R, T4X, T4W, T4O, T1G; E T2O, T3I, T2P, T3P, T2K, T2M, T1C, T1E, TC, T2w, T40, T3Y, T4K, T4I, TQ; { E T1y, T3U, T1v, T3T; { E T3h, T3, T1t, T3i, T6, T1u; { E T1w, T1x, T1, T2, T4, T5; T1 = cr[0]; T2 = ci[WS(rs, 9)]; T1w = ci[WS(rs, 14)]; T4f = FNMS(T1s, T1X, T47); T48 = FMA(T1s, T1X, T47); T4c = FNMS(T1s, T1V, T4b); T4k = FMA(T1s, T1V, T4b); T5m = FMA(T1s, T1N, T5l); T5q = FNMS(T1s, T1L, T5p); T3h = T1 - T2; T3 = T1 + T2; T1x = cr[WS(rs, 15)]; T4 = cr[WS(rs, 5)]; T5 = ci[WS(rs, 4)]; T1t = ci[WS(rs, 19)]; T3i = T1w + T1x; T1y = T1w - T1x; T3U = T4 - T5; T6 = T4 + T5; T1u = cr[WS(rs, 10)]; } T3j = T3h + T3i; T4B = T3h - T3i; T7 = T3 + T6; TJ = T3 - T6; T1v = T1t - T1u; T3T = T1t + T1u; } { E T3m, T4C, Te, TK, T4M, T3L, T1f, T2y, TO, TA, T4Q, T3H, T3w, T4G, T2C; E T17, T3p, T4D, Tl, TL, T3O, T4N, T1m, T2z, T3t, T4F, Tt, TN, T3E, T4P; E T10, T2B; { E T3u, T13, T3v, T16; { E T1e, T3K, T1b, T3J; { E T3k, Ta, T19, T3l, Td, T1a; { E T1c, T1d, T8, T9, Tb, Tc; T8 = cr[WS(rs, 4)]; T9 = ci[WS(rs, 5)]; T4V = T3U + T3T; T3V = T3T - T3U; T1z = T1v - T1y; T2H = T1v + T1y; T3k = T8 - T9; Ta = T8 + T9; T1c = ci[WS(rs, 10)]; T1d = cr[WS(rs, 19)]; Tb = cr[WS(rs, 9)]; Tc = ci[0]; T19 = ci[WS(rs, 15)]; T3l = T1c + T1d; T1e = T1c - T1d; T3K = Tb - Tc; Td = Tb + Tc; T1a = cr[WS(rs, 14)]; } T3m = T3k + T3l; T4C = T3k - T3l; Te = Ta + Td; TK = Ta - Td; T1b = T19 - T1a; T3J = T19 + T1a; } { E Tw, T14, T3F, Tz, T3G, T15; { E Tx, Ty, Tu, Tv, T11, T12; Tu = ci[WS(rs, 7)]; Tv = cr[WS(rs, 2)]; T4M = T3K + T3J; T3L = T3J - T3K; T1f = T1b - T1e; T2y = T1b + T1e; T3u = Tu - Tv; Tw = Tu + Tv; Tx = ci[WS(rs, 2)]; Ty = cr[WS(rs, 7)]; T11 = ci[WS(rs, 17)]; T12 = cr[WS(rs, 12)]; T14 = ci[WS(rs, 12)]; T3F = Tx - Ty; Tz = Tx + Ty; T3G = T11 + T12; T13 = T11 - T12; T15 = cr[WS(rs, 17)]; } TO = Tw - Tz; TA = Tw + Tz; T4Q = T3F - T3G; T3H = T3F + T3G; T3v = T14 + T15; T16 = T14 - T15; } } { E Ti, T3n, Th, T3o, T1l, Tj, T1g, T1h; { E Tf, Tg, T1j, T1k; Tf = ci[WS(rs, 3)]; T3w = T3u - T3v; T4G = T3u + T3v; T2C = T13 + T16; T17 = T13 - T16; Tg = cr[WS(rs, 6)]; T1j = ci[WS(rs, 18)]; T1k = cr[WS(rs, 11)]; Ti = cr[WS(rs, 1)]; T3n = Tf - Tg; Th = Tf + Tg; T3o = T1j + T1k; T1l = T1j - T1k; Tj = ci[WS(rs, 8)]; T1g = ci[WS(rs, 13)]; T1h = cr[WS(rs, 16)]; } { E T3M, Tk, T3N, T1i; T3p = T3n + T3o; T4D = T3n - T3o; T3M = Ti - Tj; Tk = Ti + Tj; T3N = T1g + T1h; T1i = T1g - T1h; Tl = Th + Tk; TL = Th - Tk; T3O = T3M + T3N; T4N = T3M - T3N; T1m = T1i - T1l; T2z = T1i + T1l; } } { E Tq, T3r, Tp, T3s, TZ, Tr, TU, TV; { E Tn, To, TX, TY; Tn = cr[WS(rs, 8)]; To = ci[WS(rs, 1)]; TX = ci[WS(rs, 16)]; TY = cr[WS(rs, 13)]; Tq = ci[WS(rs, 6)]; T3r = Tn - To; Tp = Tn + To; T3s = TX + TY; TZ = TX - TY; Tr = cr[WS(rs, 3)]; TU = ci[WS(rs, 11)]; TV = cr[WS(rs, 18)]; } { E T3D, Ts, T3C, TW; T3t = T3r - T3s; T4F = T3r + T3s; T3D = Tq - Tr; Ts = Tq + Tr; T3C = TU + TV; TW = TU - TV; Tt = Tp + Ts; TN = Tp - Ts; T3E = T3C - T3D; T4P = T3D + T3C; T10 = TW - TZ; T2B = TW + TZ; } } } { E T1B, T1A, T2J, T4H, T4E, T2I, TM, TP; T3x = T3t + T3w; T42 = T3t - T3w; T18 = T10 - T17; T1B = T10 + T17; T3q = T3m + T3p; T43 = T3m - T3p; T1n = T1f - T1m; T1A = T1f + T1m; T2J = T2B + T2C; T2D = T2B - T2C; T53 = T4F - T4G; T4H = T4F + T4G; T4E = T4C + T4D; T52 = T4C - T4D; T2A = T2y - T2z; T2I = T2y + T2z; TM = TK + TL; T1H = TK - TL; T4R = T4P - T4Q; T4X = T4P + T4Q; T4W = T4M + T4N; T4O = T4M - T4N; T1G = TN - TO; TP = TN + TO; { E Tm, T3X, TB, T3W; Tm = Te + Tl; T2O = Te - Tl; T3I = T3E + T3H; T3X = T3E - T3H; TB = Tt + TA; T2P = Tt - TA; T3P = T3L + T3O; T3W = T3L - T3O; T2K = T2I + T2J; T2M = T2I - T2J; T1C = T1A + T1B; T1E = T1A - T1B; TC = Tm + TB; T2w = Tm - TB; T40 = T3W - T3X; T3Y = T3W + T3X; T4K = T4E - T4H; T4I = T4E + T4H; TS = TM - TP; TQ = TM + TP; } } } } { E T3A, T3y, T50, T1D, T2t, T2p, T4J, T5t, T5v, T4Z, T4Y; cr[0] = T7 + TC; T3A = T3q - T3x; T3y = T3q + T3x; T50 = T4W - T4X; T4Y = T4W + T4X; ci[0] = T2H + T2K; T1D = FNMS(KP250000000, T1C, T1z); T2t = T1z + T1C; T2p = TJ + TQ; TR = FNMS(KP250000000, TQ, TJ); T4J = FNMS(KP250000000, T4I, T4B); T5t = T4B + T4I; T5v = T4V + T4Y; T4Z = FNMS(KP250000000, T4Y, T4V); { E T4m, T44, T4i, T4p, T49, T3R, T4j, T4a, T3S, T4l, T41, T4q; { E T3z, T4v, T4w, T3Z, T4z; T3z = FNMS(KP250000000, T3y, T3j); T4v = T3j + T3y; { E T2u, T2q, T5u, T5w; T2u = T2s * T2p; T2q = T2o * T2p; T5u = T2c * T5t; T5w = T2c * T5v; ci[WS(rs, 10)] = FMA(T2o, T2t, T2u); cr[WS(rs, 10)] = FNMS(T2s, T2t, T2q); cr[WS(rs, 5)] = FNMS(T2f, T5v, T5u); ci[WS(rs, 5)] = FMA(T2f, T5t, T5w); T4w = T4u * T4v; } T3Z = FNMS(KP250000000, T3Y, T3V); T4z = T3V + T3Y; { E T3Q, T4h, T4A, T4g, T3B; T3Q = FNMS(KP618033988, T3P, T3I); T4h = FMA(KP618033988, T3I, T3P); cr[WS(rs, 15)] = FNMS(T4y, T4z, T4w); T4A = T4u * T4z; T4m = FMA(KP618033988, T42, T43); T44 = FNMS(KP618033988, T43, T42); T4g = FMA(KP559016994, T3A, T3z); T3B = FNMS(KP559016994, T3A, T3z); ci[WS(rs, 15)] = FMA(T4y, T4v, T4A); T4i = FNMS(KP951056516, T4h, T4g); T4p = FMA(KP951056516, T4h, T4g); T49 = FMA(KP951056516, T3Q, T3B); T3R = FNMS(KP951056516, T3Q, T3B); } T4j = T4f * T4i; T4a = T48 * T49; T3S = TE * T3R; T4l = FMA(KP559016994, T40, T3Z); T41 = FNMS(KP559016994, T40, T3Z); T4q = T1L * T4p; } { E T5d, T4S, T54, T5i, T4L, T5c; T5d = FNMS(KP618033988, T4O, T4R); T4S = FMA(KP618033988, T4R, T4O); { E T4n, T4r, T4d, T45; T4n = FMA(KP951056516, T4m, T4l); T4r = FNMS(KP951056516, T4m, T4l); T4d = FNMS(KP951056516, T44, T41); T45 = FMA(KP951056516, T44, T41); { E T4o, T4s, T4e, T46; T4o = T4f * T4n; cr[WS(rs, 11)] = FNMS(T4k, T4n, T4j); T4s = T1L * T4r; cr[WS(rs, 19)] = FNMS(T1N, T4r, T4q); T4e = T48 * T4d; cr[WS(rs, 7)] = FNMS(T4c, T4d, T4a); T46 = TE * T45; cr[WS(rs, 3)] = FNMS(TH, T45, T3S); ci[WS(rs, 11)] = FMA(T4k, T4i, T4o); ci[WS(rs, 19)] = FMA(T1N, T4p, T4s); ci[WS(rs, 7)] = FMA(T4c, T49, T4e); ci[WS(rs, 3)] = FMA(TH, T3R, T46); } } T54 = FMA(KP618033988, T53, T52); T5i = FNMS(KP618033988, T52, T53); T4L = FMA(KP559016994, T4K, T4J); T5c = FNMS(KP559016994, T4K, T4J); { E T38, T2Q, T33, T2E, T2v, T37, T2N, T5h, T51, T2L, T2x, T32; T38 = FNMS(KP618033988, T2O, T2P); T2Q = FMA(KP618033988, T2P, T2O); T5h = FNMS(KP559016994, T50, T4Z); T51 = FMA(KP559016994, T50, T4Z); { E T5e, T5n, T57, T4T; T5e = FNMS(KP951056516, T5d, T5c); T5n = FMA(KP951056516, T5d, T5c); T57 = FMA(KP951056516, T4S, T4L); T4T = FNMS(KP951056516, T4S, T4L); { E T5j, T5r, T59, T55; T5j = FMA(KP951056516, T5i, T5h); T5r = FNMS(KP951056516, T5i, T5h); T59 = FNMS(KP951056516, T54, T51); T55 = FMA(KP951056516, T54, T51); { E T5f, T5o, T58, T4U; T5f = T5b * T5e; T5o = T5m * T5n; T58 = T1V * T57; T4U = TD * T4T; { E T5k, T5s, T5a, T56; T5k = T5b * T5j; T5s = T5m * T5r; T5a = T1V * T59; T56 = TD * T55; cr[WS(rs, 13)] = FNMS(T5g, T5j, T5f); cr[WS(rs, 17)] = FNMS(T5q, T5r, T5o); cr[WS(rs, 9)] = FNMS(T1X, T59, T58); cr[WS(rs, 1)] = FNMS(TG, T55, T4U); ci[WS(rs, 13)] = FMA(T5g, T5e, T5k); ci[WS(rs, 17)] = FMA(T5q, T5n, T5s); ci[WS(rs, 9)] = FMA(T1X, T57, T5a); ci[WS(rs, 1)] = FMA(TG, T4T, T56); } } } } T2L = FNMS(KP250000000, T2K, T2H); T33 = FNMS(KP618033988, T2A, T2D); T2E = FMA(KP618033988, T2D, T2A); T2v = FNMS(KP250000000, TC, T7); T37 = FNMS(KP559016994, T2M, T2L); T2N = FMA(KP559016994, T2M, T2L); T1I = FNMS(KP618033988, T1H, T1G); T26 = FMA(KP618033988, T1G, T1H); T2x = FMA(KP559016994, T2w, T2v); T32 = FNMS(KP559016994, T2w, T2v); { E T3f, T39, T2R, T2Z; T3f = FNMS(KP951056516, T38, T37); T39 = FMA(KP951056516, T38, T37); T2R = FNMS(KP951056516, T2Q, T2N); T2Z = FMA(KP951056516, T2Q, T2N); { E T3c, T34, T2F, T2V; T3c = FMA(KP951056516, T33, T32); T34 = FNMS(KP951056516, T33, T32); T2F = FMA(KP951056516, T2E, T2x); T2V = FNMS(KP951056516, T2E, T2x); { E T3a, T35, T3g, T3d; T3a = T36 * T34; T35 = T31 * T34; T3g = T3e * T3c; T3d = T3b * T3c; { E T30, T2W, T2S, T2G; T30 = T2Y * T2V; T2W = T2U * T2V; T2S = T2b * T2F; T2G = T29 * T2F; ci[WS(rs, 8)] = FMA(T31, T39, T3a); cr[WS(rs, 8)] = FNMS(T36, T39, T35); ci[WS(rs, 12)] = FMA(T3b, T3f, T3g); cr[WS(rs, 12)] = FNMS(T3e, T3f, T3d); ci[WS(rs, 16)] = FMA(T2U, T2Z, T30); cr[WS(rs, 16)] = FNMS(T2Y, T2Z, T2W); ci[WS(rs, 4)] = FMA(T29, T2R, T2S); cr[WS(rs, 4)] = FNMS(T2b, T2R, T2G); } } } } T1o = FNMS(KP618033988, T1n, T18); T20 = FMA(KP618033988, T18, T1n); T1F = FNMS(KP559016994, T1E, T1D); T25 = FMA(KP559016994, T1E, T1D); } } } } } } TT = FNMS(KP559016994, TS, TR); T1Z = FMA(KP559016994, TS, TR); { E T2l, T27, T1J, T1T; T2l = FNMS(KP951056516, T26, T25); T27 = FMA(KP951056516, T26, T25); T1J = FNMS(KP951056516, T1I, T1F); T1T = FMA(KP951056516, T1I, T1F); { E T2h, T21, T1p, T1P; T2h = FMA(KP951056516, T20, T1Z); T21 = FNMS(KP951056516, T20, T1Z); T1p = FMA(KP951056516, T1o, TT); T1P = FNMS(KP951056516, T1o, TT); { E T28, T22, T2m, T2i; T28 = T24 * T21; T22 = T1Y * T21; T2m = T2k * T2h; T2i = T2g * T2h; { E T1U, T1Q, T1K, T1q; T1U = T1S * T1P; T1Q = T1O * T1P; T1K = T1s * T1p; T1q = TI * T1p; ci[WS(rs, 6)] = FMA(T1Y, T27, T28); cr[WS(rs, 6)] = FNMS(T24, T27, T22); ci[WS(rs, 14)] = FMA(T2g, T2l, T2m); cr[WS(rs, 14)] = FNMS(T2k, T2l, T2i); ci[WS(rs, 18)] = FMA(T1O, T1T, T1U); cr[WS(rs, 18)] = FNMS(T1S, T1T, T1Q); ci[WS(rs, 2)] = FMA(TI, T1J, T1K); cr[WS(rs, 2)] = FNMS(T1s, T1J, T1q); } } } } } } } static const tw_instr twinstr[] = { {TW_CEXP, 1, 1}, {TW_CEXP, 1, 3}, {TW_CEXP, 1, 9}, {TW_CEXP, 1, 19}, {TW_NEXT, 1, 0} }; static const hc2hc_desc desc = { 20, "hb2_20", twinstr, &GENUS, {136, 58, 140, 0} }; void X(codelet_hb2_20) (planner *p) { X(khc2hc_register) (p, hb2_20, &desc); } #else /* HAVE_FMA */ /* Generated by: ../../../genfft/gen_hc2hc.native -compact -variables 4 -pipeline-latency 4 -sign 1 -twiddle-log3 -precompute-twiddles -n 20 -dif -name hb2_20 -include hb.h */ /* * This function contains 276 FP additions, 164 FP multiplications, * (or, 204 additions, 92 multiplications, 72 fused multiply/add), * 137 stack variables, 4 constants, and 80 memory accesses */ #include "hb.h" static void hb2_20(R *cr, R *ci, const R *W, stride rs, INT mb, INT me, INT ms) { DK(KP250000000, +0.250000000000000000000000000000000000000000000); DK(KP559016994, +0.559016994374947424102293417182819058860154590); DK(KP587785252, +0.587785252292473129168705954639072768597652438); DK(KP951056516, +0.951056516295153572116439333379382143405698634); { INT m; for (m = mb, W = W + ((mb - 1) * 8); m < me; m = m + 1, cr = cr + ms, ci = ci - ms, W = W + 8, MAKE_VOLATILE_STRIDE(40, rs)) { E TD, TG, TE, TH, TJ, T1t, T27, T25, T1T, T1R, T1V, T2j, T2Z, T21, T2X; E T2T, T2n, T2P, T3V, T41, T3R, T3X, T29, T2c, T4H, T4L, T1L, T1M, T1N, T2d; E T4R, T1P, T4P, T49, T2N, T2f, T47, T2L; { E T1U, T2l, T1Z, T2i, T1S, T2m, T20, T2h; { E TF, T1s, TI, T1r; TD = W[0]; TG = W[1]; TE = W[2]; TH = W[3]; TF = TD * TE; T1s = TG * TE; TI = TG * TH; T1r = TD * TH; TJ = TF + TI; T1t = T1r - T1s; T27 = T1r + T1s; T25 = TF - TI; T1T = W[5]; T1U = TH * T1T; T2l = TD * T1T; T1Z = TE * T1T; T2i = TG * T1T; T1R = W[4]; T1S = TE * T1R; T2m = TG * T1R; T20 = TH * T1R; T2h = TD * T1R; } T1V = T1S + T1U; T2j = T2h - T2i; T2Z = T1Z + T20; T21 = T1Z - T20; T2X = T1S - T1U; T2T = T2l - T2m; T2n = T2l + T2m; T2P = T2h + T2i; { E T3T, T3U, T3P, T3Q; T3T = TJ * T1T; T3U = T1t * T1R; T3V = T3T - T3U; T41 = T3T + T3U; T3P = TJ * T1R; T3Q = T1t * T1T; T3R = T3P + T3Q; T3X = T3P - T3Q; { E T26, T28, T2a, T2b; T26 = T25 * T1R; T28 = T27 * T1T; T29 = T26 + T28; T2a = T25 * T1T; T2b = T27 * T1R; T2c = T2a - T2b; T4H = T26 - T28; T4L = T2a + T2b; T1L = W[6]; T1M = W[7]; T1N = FMA(TD, T1L, TG * T1M); T2d = FMA(T29, T1L, T2c * T1M); T4R = FNMS(T1t, T1L, TJ * T1M); T1P = FNMS(TG, T1L, TD * T1M); T4P = FMA(TJ, T1L, T1t * T1M); T49 = FNMS(T27, T1L, T25 * T1M); T2N = FNMS(TH, T1L, TE * T1M); T2f = FNMS(T2c, T1L, T29 * T1M); T47 = FMA(T25, T1L, T27 * T1M); T2L = FMA(TE, T1L, TH * T1M); } } } { E T7, T4i, T4x, TK, T1D, T3i, T3E, T2D, T19, T3L, T3M, T1o, T2x, T4C, T4B; E T2u, T1v, T4r, T4o, T1u, T2H, T37, T2I, T3e, T3p, T3w, T3x, Tm, TB, TC; E T4u, T4v, T4y, T2A, T2B, T2E, T1E, T1F, T1G, T4d, T4g, T4j, T3F, T3G, T3H; E TN, TQ, TR, T48, T4a; { E T3, T3g, T1C, T3h, T6, T3D, T1z, T3C; { E T1, T2, T1A, T1B; T1 = cr[0]; T2 = ci[WS(rs, 9)]; T3 = T1 + T2; T3g = T1 - T2; T1A = ci[WS(rs, 14)]; T1B = cr[WS(rs, 15)]; T1C = T1A - T1B; T3h = T1A + T1B; } { E T4, T5, T1x, T1y; T4 = cr[WS(rs, 5)]; T5 = ci[WS(rs, 4)]; T6 = T4 + T5; T3D = T4 - T5; T1x = ci[WS(rs, 19)]; T1y = cr[WS(rs, 10)]; T1z = T1x - T1y; T3C = T1x + T1y; } T7 = T3 + T6; T4i = T3g - T3h; T4x = T3D + T3C; TK = T3 - T6; T1D = T1z - T1C; T3i = T3g + T3h; T3E = T3C - T3D; T2D = T1z + T1C; } { E Te, T4b, T4m, TL, T11, T33, T3l, T2s, TA, T4f, T4q, TP, T1n, T3d, T3v; E T2w, Tl, T4c, T4n, TM, T18, T36, T3o, T2t, Tt, T4e, T4p, TO, T1g, T3a; E T3s, T2v; { E Ta, T3j, T10, T3k, Td, T32, TX, T31; { E T8, T9, TY, TZ; T8 = cr[WS(rs, 4)]; T9 = ci[WS(rs, 5)]; Ta = T8 + T9; T3j = T8 - T9; TY = ci[WS(rs, 10)]; TZ = cr[WS(rs, 19)]; T10 = TY - TZ; T3k = TY + TZ; } { E Tb, Tc, TV, TW; Tb = cr[WS(rs, 9)]; Tc = ci[0]; Td = Tb + Tc; T32 = Tb - Tc; TV = ci[WS(rs, 15)]; TW = cr[WS(rs, 14)]; TX = TV - TW; T31 = TV + TW; } Te = Ta + Td; T4b = T3j - T3k; T4m = T32 + T31; TL = Ta - Td; T11 = TX - T10; T33 = T31 - T32; T3l = T3j + T3k; T2s = TX + T10; } { E Tw, T3t, Tz, T3b, T1j, T3c, T1m, T3u; { E Tu, Tv, Tx, Ty; Tu = ci[WS(rs, 7)]; Tv = cr[WS(rs, 2)]; Tw = Tu + Tv; T3t = Tu - Tv; Tx = ci[WS(rs, 2)]; Ty = cr[WS(rs, 7)]; Tz = Tx + Ty; T3b = Tx - Ty; } { E T1h, T1i, T1k, T1l; T1h = ci[WS(rs, 17)]; T1i = cr[WS(rs, 12)]; T1j = T1h - T1i; T3c = T1h + T1i; T1k = ci[WS(rs, 12)]; T1l = cr[WS(rs, 17)]; T1m = T1k - T1l; T3u = T1k + T1l; } TA = Tw + Tz; T4f = T3t + T3u; T4q = T3b - T3c; TP = Tw - Tz; T1n = T1j - T1m; T3d = T3b + T3c; T3v = T3t - T3u; T2w = T1j + T1m; } { E Th, T3m, T17, T3n, Tk, T34, T14, T35; { E Tf, Tg, T15, T16; Tf = ci[WS(rs, 3)]; Tg = cr[WS(rs, 6)]; Th = Tf + Tg; T3m = Tf - Tg; T15 = ci[WS(rs, 18)]; T16 = cr[WS(rs, 11)]; T17 = T15 - T16; T3n = T15 + T16; } { E Ti, Tj, T12, T13; Ti = cr[WS(rs, 1)]; Tj = ci[WS(rs, 8)]; Tk = Ti + Tj; T34 = Ti - Tj; T12 = ci[WS(rs, 13)]; T13 = cr[WS(rs, 16)]; T14 = T12 - T13; T35 = T12 + T13; } Tl = Th + Tk; T4c = T3m - T3n; T4n = T34 - T35; TM = Th - Tk; T18 = T14 - T17; T36 = T34 + T35; T3o = T3m + T3n; T2t = T14 + T17; } { E Tp, T3q, T1f, T3r, Ts, T39, T1c, T38; { E Tn, To, T1d, T1e; Tn = cr[WS(rs, 8)]; To = ci[WS(rs, 1)]; Tp = Tn + To; T3q = Tn - To; T1d = ci[WS(rs, 16)]; T1e = cr[WS(rs, 13)]; T1f = T1d - T1e; T3r = T1d + T1e; } { E Tq, Tr, T1a, T1b; Tq = ci[WS(rs, 6)]; Tr = cr[WS(rs, 3)]; Ts = Tq + Tr; T39 = Tq - Tr; T1a = ci[WS(rs, 11)]; T1b = cr[WS(rs, 18)]; T1c = T1a - T1b; T38 = T1a + T1b; } Tt = Tp + Ts; T4e = T3q + T3r; T4p = T39 + T38; TO = Tp - Ts; T1g = T1c - T1f; T3a = T38 - T39; T3s = T3q - T3r; T2v = T1c + T1f; } T19 = T11 - T18; T3L = T3l - T3o; T3M = T3s - T3v; T1o = T1g - T1n; T2x = T2v - T2w; T4C = T4e - T4f; T4B = T4b - T4c; T2u = T2s - T2t; T1v = TO - TP; T4r = T4p - T4q; T4o = T4m - T4n; T1u = TL - TM; T2H = Te - Tl; T37 = T33 + T36; T2I = Tt - TA; T3e = T3a + T3d; T3p = T3l + T3o; T3w = T3s + T3v; T3x = T3p + T3w; Tm = Te + Tl; TB = Tt + TA; TC = Tm + TB; T4u = T4m + T4n; T4v = T4p + T4q; T4y = T4u + T4v; T2A = T2s + T2t; T2B = T2v + T2w; T2E = T2A + T2B; T1E = T11 + T18; T1F = T1g + T1n; T1G = T1E + T1F; T4d = T4b + T4c; T4g = T4e + T4f; T4j = T4d + T4g; T3F = T33 - T36; T3G = T3a - T3d; T3H = T3F + T3G; TN = TL + TM; TQ = TO + TP; TR = TN + TQ; } cr[0] = T7 + TC; ci[0] = T2D + T2E; { E T2k, T2o, T4T, T4U; T2k = TK + TR; T2o = T1D + T1G; cr[WS(rs, 10)] = FNMS(T2n, T2o, T2j * T2k); ci[WS(rs, 10)] = FMA(T2n, T2k, T2j * T2o); T4T = T4i + T4j; T4U = T4x + T4y; cr[WS(rs, 5)] = FNMS(T2c, T4U, T29 * T4T); ci[WS(rs, 5)] = FMA(T29, T4U, T2c * T4T); } T48 = T3i + T3x; T4a = T3E + T3H; cr[WS(rs, 15)] = FNMS(T49, T4a, T47 * T48); ci[WS(rs, 15)] = FMA(T47, T4a, T49 * T48); { E T2y, T2J, T2V, T2R, T2G, T2U, T2r, T2Q; T2y = FMA(KP951056516, T2u, KP587785252 * T2x); T2J = FMA(KP951056516, T2H, KP587785252 * T2I); T2V = FNMS(KP951056516, T2I, KP587785252 * T2H); T2R = FNMS(KP951056516, T2x, KP587785252 * T2u); { E T2C, T2F, T2p, T2q; T2C = KP559016994 * (T2A - T2B); T2F = FNMS(KP250000000, T2E, T2D); T2G = T2C + T2F; T2U = T2F - T2C; T2p = KP559016994 * (Tm - TB); T2q = FNMS(KP250000000, TC, T7); T2r = T2p + T2q; T2Q = T2q - T2p; } { E T2z, T2K, T2Y, T30; T2z = T2r + T2y; T2K = T2G - T2J; cr[WS(rs, 4)] = FNMS(T27, T2K, T25 * T2z); ci[WS(rs, 4)] = FMA(T27, T2z, T25 * T2K); T2Y = T2Q - T2R; T30 = T2V + T2U; cr[WS(rs, 12)] = FNMS(T2Z, T30, T2X * T2Y); ci[WS(rs, 12)] = FMA(T2Z, T2Y, T2X * T30); } { E T2M, T2O, T2S, T2W; T2M = T2r - T2y; T2O = T2J + T2G; cr[WS(rs, 16)] = FNMS(T2N, T2O, T2L * T2M); ci[WS(rs, 16)] = FMA(T2N, T2M, T2L * T2O); T2S = T2Q + T2R; T2W = T2U - T2V; cr[WS(rs, 8)] = FNMS(T2T, T2W, T2P * T2S); ci[WS(rs, 8)] = FMA(T2T, T2S, T2P * T2W); } } { E T4s, T4D, T4N, T4I, T4A, T4M, T4l, T4J; T4s = FMA(KP951056516, T4o, KP587785252 * T4r); T4D = FMA(KP951056516, T4B, KP587785252 * T4C); T4N = FNMS(KP951056516, T4C, KP587785252 * T4B); T4I = FNMS(KP951056516, T4r, KP587785252 * T4o); { E T4w, T4z, T4h, T4k; T4w = KP559016994 * (T4u - T4v); T4z = FNMS(KP250000000, T4y, T4x); T4A = T4w + T4z; T4M = T4z - T4w; T4h = KP559016994 * (T4d - T4g); T4k = FNMS(KP250000000, T4j, T4i); T4l = T4h + T4k; T4J = T4k - T4h; } { E T4t, T4E, T4Q, T4S; T4t = T4l - T4s; T4E = T4A + T4D; cr[WS(rs, 1)] = FNMS(TG, T4E, TD * T4t); ci[WS(rs, 1)] = FMA(TD, T4E, TG * T4t); T4Q = T4J - T4I; T4S = T4M + T4N; cr[WS(rs, 17)] = FNMS(T4R, T4S, T4P * T4Q); ci[WS(rs, 17)] = FMA(T4P, T4S, T4R * T4Q); } { E T4F, T4G, T4K, T4O; T4F = T4s + T4l; T4G = T4A - T4D; cr[WS(rs, 9)] = FNMS(T1T, T4G, T1R * T4F); ci[WS(rs, 9)] = FMA(T1R, T4G, T1T * T4F); T4K = T4I + T4J; T4O = T4M - T4N; cr[WS(rs, 13)] = FNMS(T4L, T4O, T4H * T4K); ci[WS(rs, 13)] = FMA(T4H, T4O, T4L * T4K); } } { E T1p, T1w, T22, T1X, T1J, T23, TU, T1W; T1p = FNMS(KP951056516, T1o, KP587785252 * T19); T1w = FNMS(KP951056516, T1v, KP587785252 * T1u); T22 = FMA(KP951056516, T1u, KP587785252 * T1v); T1X = FMA(KP951056516, T19, KP587785252 * T1o); { E T1H, T1I, TS, TT; T1H = FNMS(KP250000000, T1G, T1D); T1I = KP559016994 * (T1E - T1F); T1J = T1H - T1I; T23 = T1I + T1H; TS = FNMS(KP250000000, TR, TK); TT = KP559016994 * (TN - TQ); TU = TS - TT; T1W = TT + TS; } { E T1q, T1K, T2e, T2g; T1q = TU - T1p; T1K = T1w + T1J; cr[WS(rs, 2)] = FNMS(T1t, T1K, TJ * T1q); ci[WS(rs, 2)] = FMA(T1t, T1q, TJ * T1K); T2e = T1W + T1X; T2g = T23 - T22; cr[WS(rs, 14)] = FNMS(T2f, T2g, T2d * T2e); ci[WS(rs, 14)] = FMA(T2f, T2e, T2d * T2g); } { E T1O, T1Q, T1Y, T24; T1O = TU + T1p; T1Q = T1J - T1w; cr[WS(rs, 18)] = FNMS(T1P, T1Q, T1N * T1O); ci[WS(rs, 18)] = FMA(T1P, T1O, T1N * T1Q); T1Y = T1W - T1X; T24 = T22 + T23; cr[WS(rs, 6)] = FNMS(T21, T24, T1V * T1Y); ci[WS(rs, 6)] = FMA(T21, T1Y, T1V * T24); } } { E T3f, T3N, T43, T3Z, T3K, T42, T3A, T3Y; T3f = FNMS(KP951056516, T3e, KP587785252 * T37); T3N = FNMS(KP951056516, T3M, KP587785252 * T3L); T43 = FMA(KP951056516, T3L, KP587785252 * T3M); T3Z = FMA(KP951056516, T37, KP587785252 * T3e); { E T3I, T3J, T3y, T3z; T3I = FNMS(KP250000000, T3H, T3E); T3J = KP559016994 * (T3F - T3G); T3K = T3I - T3J; T42 = T3J + T3I; T3y = FNMS(KP250000000, T3x, T3i); T3z = KP559016994 * (T3p - T3w); T3A = T3y - T3z; T3Y = T3z + T3y; } { E T3B, T3O, T45, T46; T3B = T3f + T3A; T3O = T3K - T3N; cr[WS(rs, 3)] = FNMS(TH, T3O, TE * T3B); ci[WS(rs, 3)] = FMA(TE, T3O, TH * T3B); T45 = T3Z + T3Y; T46 = T42 - T43; cr[WS(rs, 19)] = FNMS(T1M, T46, T1L * T45); ci[WS(rs, 19)] = FMA(T1L, T46, T1M * T45); } { E T3S, T3W, T40, T44; T3S = T3A - T3f; T3W = T3K + T3N; cr[WS(rs, 7)] = FNMS(T3V, T3W, T3R * T3S); ci[WS(rs, 7)] = FMA(T3R, T3W, T3V * T3S); T40 = T3Y - T3Z; T44 = T42 + T43; cr[WS(rs, 11)] = FNMS(T41, T44, T3X * T40); ci[WS(rs, 11)] = FMA(T3X, T44, T41 * T40); } } } } } } static const tw_instr twinstr[] = { {TW_CEXP, 1, 1}, {TW_CEXP, 1, 3}, {TW_CEXP, 1, 9}, {TW_CEXP, 1, 19}, {TW_NEXT, 1, 0} }; static const hc2hc_desc desc = { 20, "hb2_20", twinstr, &GENUS, {204, 92, 72, 0} }; void X(codelet_hb2_20) (planner *p) { X(khc2hc_register) (p, hb2_20, &desc); } #endif /* HAVE_FMA */ fftw-3.3.3/rdft/scalar/r2cb/hb2_25.c0000644000175000017500000015444212054411000013565 00000000000000/* * Copyright (c) 2003, 2007-11 Matteo Frigo * Copyright (c) 2003, 2007-11 Massachusetts Institute of Technology * * This program is free software; you can redistribute it and/or modify * it under the terms of the GNU General Public License as published by * the Free Software Foundation; either version 2 of the License, or * (at your option) any later version. * * This program is distributed in the hope that it will be useful, * but WITHOUT ANY WARRANTY; without even the implied warranty of * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the * GNU General Public License for more details. * * You should have received a copy of the GNU General Public License * along with this program; if not, write to the Free Software * Foundation, Inc., 51 Franklin Street, Fifth Floor, Boston, MA 02110-1301 USA * */ /* This file was automatically generated --- DO NOT EDIT */ /* Generated on Sun Nov 25 07:41:30 EST 2012 */ #include "codelet-rdft.h" #ifdef HAVE_FMA /* Generated by: ../../../genfft/gen_hc2hc.native -fma -reorder-insns -schedule-for-pipeline -compact -variables 4 -pipeline-latency 4 -sign 1 -twiddle-log3 -precompute-twiddles -n 25 -dif -name hb2_25 -include hb.h */ /* * This function contains 440 FP additions, 434 FP multiplications, * (or, 84 additions, 78 multiplications, 356 fused multiply/add), * 234 stack variables, 47 constants, and 100 memory accesses */ #include "hb.h" static void hb2_25(R *cr, R *ci, const R *W, stride rs, INT mb, INT me, INT ms) { DK(KP833417178, +0.833417178328688677408962550243238843138996060); DK(KP921177326, +0.921177326965143320250447435415066029359282231); DK(KP541454447, +0.541454447536312777046285590082819509052033189); DK(KP968583161, +0.968583161128631119490168375464735813836012403); DK(KP242145790, +0.242145790282157779872542093866183953459003101); DK(KP904730450, +0.904730450839922351881287709692877908104763647); DK(KP803003575, +0.803003575438660414833440593570376004635464850); DK(KP554608978, +0.554608978404018097464974850792216217022558774); DK(KP683113946, +0.683113946453479238701949862233725244439656928); DK(KP559154169, +0.559154169276087864842202529084232643714075927); DK(KP248028675, +0.248028675328619457762448260696444630363259177); DK(KP992114701, +0.992114701314477831049793042785778521453036709); DK(KP831864738, +0.831864738706457140726048799369896829771167132); DK(KP871714437, +0.871714437527667770979999223229522602943903653); DK(KP851038619, +0.851038619207379630836264138867114231259902550); DK(KP943557151, +0.943557151597354104399655195398983005179443399); DK(KP726211448, +0.726211448929902658173535992263577167607493062); DK(KP525970792, +0.525970792408939708442463226536226366643874659); DK(KP912018591, +0.912018591466481957908415381764119056233607330); DK(KP912575812, +0.912575812670962425556968549836277086778922727); DK(KP994076283, +0.994076283785401014123185814696322018529298887); DK(KP614372930, +0.614372930789563808870829930444362096004872855); DK(KP621716863, +0.621716863012209892444754556304102309693593202); DK(KP860541664, +0.860541664367944677098261680920518816412804187); DK(KP949179823, +0.949179823508441261575555465843363271711583843); DK(KP557913902, +0.557913902031834264187699648465567037992437152); DK(KP998026728, +0.998026728428271561952336806863450553336905220); DK(KP249506682, +0.249506682107067890488084201715862638334226305); DK(KP772036680, +0.772036680810363904029489473607579825330539880); DK(KP906616052, +0.906616052148196230441134447086066874408359177); DK(KP734762448, +0.734762448793050413546343770063151342619912334); DK(KP560319534, +0.560319534973832390111614715371676131169633784); DK(KP681693190, +0.681693190061530575150324149145440022633095390); DK(KP845997307, +0.845997307939530944175097360758058292389769300); DK(KP968479752, +0.968479752739016373193524836781420152702090879); DK(KP062914667, +0.062914667253649757225485955897349402364686947); DK(KP827271945, +0.827271945972475634034355757144307982555673741); DK(KP470564281, +0.470564281212251493087595091036643380879947982); DK(KP126329378, +0.126329378446108174786050455341811215027378105); DK(KP256756360, +0.256756360367726783319498520922669048172391148); DK(KP634619297, +0.634619297544148100711287640319130485732531031); DK(KP549754652, +0.549754652192770074288023275540779861653779767); DK(KP939062505, +0.939062505817492352556001843133229685779824606); DK(KP951056516, +0.951056516295153572116439333379382143405698634); DK(KP559016994, +0.559016994374947424102293417182819058860154590); DK(KP618033988, +0.618033988749894848204586834365638117720309180); DK(KP250000000, +0.250000000000000000000000000000000000000000000); { INT m; for (m = mb, W = W + ((mb - 1) * 8); m < me; m = m + 1, cr = cr + ms, ci = ci - ms, W = W + 8, MAKE_VOLATILE_STRIDE(50, rs)) { E TN, TQ, T4e, T2y, T4i, T3U, T4u, T4o, T4G, T4C, T2F, T41, T3Q, T4q, T3a; E T3F, T4a, T4w, T46, T44; { E TT, TO, TR, T23, T2d, T2x, TP, TV, T2p, T85, T4d, T25, TX; TN = W[0]; TT = W[4]; TO = W[2]; TR = W[3]; T23 = W[6]; T2d = TN * TT; T2x = TO * TT; TP = TN * TO; TV = TN * TR; T2p = TT * T23; T85 = TN * T23; T4d = TO * T23; T25 = W[7]; TQ = W[1]; TX = W[5]; { E T86, T4n, TW, T4l, TS, T71, T2q, T4z, T2e, T8a, T2u, T76, T2k, T4B, T6E; E T6U, T6Y, T5T, T8i, T1I, T2a, T26, TY, T8d, T8s, T8o, T5C, T5w, T7g, T7c; E T5M, T5I, T9, T40, T1R, T3X, T6H, T7F, T5W, T7n, T4N, T68, T1S, T1k, T1T; E T1D, T1Y, T1Z, T10, TM, T7K, T7A, T6p, T6w, T4X, T56, T3K, T2U, T7x, T7J; E T6v, T6i, T50, T57, T3L, T39, T4Q, T59, T3O, T3E, T67, T7t, T7H, T6y, T63; E T4T, T5a, T3N, T3p, T66, T7o; { E T2A, T2z, T6G, T2E, T5V, T6F; { E T1, T1J, T3Y, T3Z, T8, T2C, T1M, T1P, T2D, T4h, T89, T2t, T3W, T1Q, T3V; T1 = cr[0]; T4e = FMA(TR, T25, T4d); T4h = TO * T25; T89 = TN * T25; T2t = TT * T25; T86 = FMA(TQ, T25, T85); T4n = FNMS(TQ, TO, TV); TW = FMA(TQ, TO, TV); T4l = FMA(TQ, TR, TP); TS = FNMS(TQ, TR, TP); T71 = FNMS(TR, TX, T2x); T2y = FMA(TR, TX, T2x); T2q = FMA(TX, T25, T2p); T4z = FMA(TQ, TX, T2d); T2e = FNMS(TQ, TX, T2d); { E T3T, T2j, T4t, T6T; T3T = TO * TX; T2j = TN * TX; T4i = FNMS(TR, T23, T4h); T8a = FNMS(TQ, T23, T89); T2u = FNMS(TX, T23, T2t); T4t = T4l * TX; T6T = T4l * T23; { E T6X, T4m, T1H, T29; T6X = T4l * T25; T4m = T4l * TT; T1H = TS * TX; T29 = TS * T25; { E T24, TU, T4F, T4A; T24 = TS * T23; TU = TS * TT; T4F = T4z * T25; T4A = T4z * T23; { E T8r, T8n, T5B, T5v; T8r = T2y * T25; T8n = T2y * T23; T5B = T2e * T25; T5v = T2e * T23; T3U = FNMS(TR, TT, T3T); T76 = FMA(TR, TT, T3T); T2k = FMA(TQ, TT, T2j); T4B = FNMS(TQ, TT, T2j); T4u = FMA(T4n, TT, T4t); T6E = FNMS(T4n, TT, T4t); T6U = FMA(T4n, T25, T6T); T6Y = FNMS(T4n, T23, T6X); T5T = FMA(T4n, TX, T4m); T4o = FNMS(T4n, TX, T4m); T8i = FMA(TW, TT, T1H); T1I = FNMS(TW, TT, T1H); T2a = FNMS(TW, T23, T29); T26 = FMA(TW, T25, T24); TY = FMA(TW, TX, TU); T8d = FNMS(TW, TX, TU); T8s = FNMS(T3U, T23, T8r); T8o = FMA(T3U, T25, T8n); T5C = FNMS(T2k, T23, T5B); T5w = FMA(T2k, T25, T5v); T4G = FNMS(T4B, T23, T4F); T4C = FMA(T4B, T25, T4A); { E T7f, T7b, T5L, T5H; T7f = T5T * T25; T7b = T5T * T23; T5L = TY * T25; T5H = TY * T23; T7g = FNMS(T6E, T23, T7f); T7c = FMA(T6E, T25, T7b); T5M = FNMS(T1I, T23, T5L); T5I = FMA(T1I, T25, T5H); T1J = ci[WS(rs, 24)]; } } } } } { E T2, T3, T5, T6; T2 = cr[WS(rs, 5)]; T3 = ci[WS(rs, 4)]; T5 = cr[WS(rs, 10)]; T6 = ci[WS(rs, 9)]; { E T1K, T4, T7, T1L, T1N, T1O; T1K = ci[WS(rs, 19)]; T3Y = T2 - T3; T4 = T2 + T3; T3Z = T5 - T6; T7 = T5 + T6; T1L = cr[WS(rs, 20)]; T1N = ci[WS(rs, 14)]; T1O = cr[WS(rs, 15)]; T8 = T4 + T7; T2A = T4 - T7; T2C = T1K + T1L; T1M = T1K - T1L; T1P = T1N - T1O; T2D = T1N + T1O; } } T2z = FNMS(KP250000000, T8, T1); T9 = T1 + T8; T3W = T1M - T1P; T1Q = T1M + T1P; T40 = FMA(KP618033988, T3Z, T3Y); T6G = FNMS(KP618033988, T3Y, T3Z); T2E = FMA(KP618033988, T2D, T2C); T5V = FNMS(KP618033988, T2C, T2D); T1R = T1J + T1Q; T3V = FNMS(KP250000000, T1Q, T1J); T6F = FNMS(KP559016994, T3W, T3V); T3X = FMA(KP559016994, T3W, T3V); } { E T2S, T6n, T2H, T2G, Ti, T5Y, T3C, T3r, TK, T3q, T30, T6d, T33, Tr, T32; E T3v, T61, T3y, T1C, T3x, T2L, T6k, T2O, T1a, T2N, T6g, T37, T2W, Tt, T1j; E T2V, Tx, T3g, T3j, Tw, T3l, T1t, T3i, Ty; { E T1u, T1v, T1A, T3u, T1w; { E TC, TI, T3B, TD, TE; { E Ta, Te, Tf, Tb, Tc, T5U, T2B, T2R, Tg; Ta = cr[WS(rs, 1)]; T5U = FNMS(KP559016994, T2A, T2z); T2B = FMA(KP559016994, T2A, T2z); T6H = FNMS(KP951056516, T6G, T6F); T7F = FMA(KP951056516, T6G, T6F); Te = cr[WS(rs, 11)]; T5W = FMA(KP951056516, T5V, T5U); T7n = FNMS(KP951056516, T5V, T5U); T4N = FMA(KP951056516, T2E, T2B); T2F = FNMS(KP951056516, T2E, T2B); Tf = ci[WS(rs, 8)]; Tb = cr[WS(rs, 6)]; Tc = ci[WS(rs, 3)]; TC = cr[WS(rs, 3)]; T2R = Tf - Te; Tg = Te + Tf; { E T2Q, Td, Th, TG, TH; T2Q = Tb - Tc; Td = Tb + Tc; TG = ci[WS(rs, 11)]; TH = ci[WS(rs, 6)]; T2S = FNMS(KP618033988, T2R, T2Q); T6n = FMA(KP618033988, T2Q, T2R); Th = Td + Tg; T2H = Td - Tg; TI = TG + TH; T3B = TG - TH; T2G = FNMS(KP250000000, Th, Ta); Ti = Ta + Th; TD = cr[WS(rs, 8)]; TE = ci[WS(rs, 1)]; } } { E Tj, Tk, Tp, T2Z, TJ, Tl; Tj = cr[WS(rs, 4)]; { E Tn, To, T3A, TF; Tn = ci[WS(rs, 10)]; To = ci[WS(rs, 5)]; T3A = TD - TE; TF = TD + TE; Tk = cr[WS(rs, 9)]; Tp = Tn + To; T2Z = To - Tn; T5Y = FNMS(KP618033988, T3A, T3B); T3C = FMA(KP618033988, T3B, T3A); T3r = TI - TF; TJ = TF + TI; Tl = ci[0]; } T1u = ci[WS(rs, 21)]; TK = TC + TJ; T3q = FNMS(KP250000000, TJ, TC); { E T1y, Tm, T2Y, T1z, Tq; T1y = cr[WS(rs, 13)]; Tm = Tk + Tl; T2Y = Tl - Tk; T1z = cr[WS(rs, 18)]; T1v = ci[WS(rs, 16)]; T30 = FMA(KP618033988, T2Z, T2Y); T6d = FNMS(KP618033988, T2Y, T2Z); T33 = Tm - Tp; Tq = Tm + Tp; T1A = T1y + T1z; T3u = T1z - T1y; Tr = Tj + Tq; T32 = FMS(KP250000000, Tq, Tj); T1w = cr[WS(rs, 23)]; } } } { E T1b, T1c, T1h, T36, T1d; { E T12, T13, T18, T2K, T1B, T14; T12 = ci[WS(rs, 23)]; { E T16, T17, T3t, T1x; T16 = ci[WS(rs, 13)]; T17 = cr[WS(rs, 16)]; T3t = T1v + T1w; T1x = T1v - T1w; T13 = ci[WS(rs, 18)]; T18 = T16 - T17; T2K = T16 + T17; T3v = FMA(KP618033988, T3u, T3t); T61 = FNMS(KP618033988, T3t, T3u); T3y = T1x + T1A; T1B = T1x - T1A; T14 = cr[WS(rs, 21)]; } T1b = ci[WS(rs, 20)]; T1C = T1u + T1B; T3x = FMS(KP250000000, T1B, T1u); { E T1f, T15, T2J, T1g, T19; T1f = cr[WS(rs, 14)]; T15 = T13 - T14; T2J = T13 + T14; T1g = cr[WS(rs, 19)]; T1c = ci[WS(rs, 15)]; T2L = FMA(KP618033988, T2K, T2J); T6k = FNMS(KP618033988, T2J, T2K); T2O = T15 - T18; T19 = T15 + T18; T1h = T1f + T1g; T36 = T1g - T1f; T1a = T12 + T19; T2N = FNMS(KP250000000, T19, T12); T1d = cr[WS(rs, 24)]; } } { E T1l, T1p, T1o, T3e, T1i, T1q; T1l = ci[WS(rs, 22)]; { E T1m, T1n, T35, T1e; T1m = ci[WS(rs, 17)]; T1n = cr[WS(rs, 22)]; T35 = T1c + T1d; T1e = T1c - T1d; T1p = ci[WS(rs, 12)]; T1o = T1m - T1n; T3e = T1m + T1n; T6g = FNMS(KP618033988, T35, T36); T37 = FMA(KP618033988, T36, T35); T2W = T1e + T1h; T1i = T1e - T1h; T1q = cr[WS(rs, 17)]; } Tt = cr[WS(rs, 2)]; T1j = T1b + T1i; T2V = FMS(KP250000000, T1i, T1b); { E Tu, T1r, T3f, Tv, T1s; Tu = cr[WS(rs, 7)]; T1r = T1p - T1q; T3f = T1p + T1q; Tv = ci[WS(rs, 2)]; Tx = cr[WS(rs, 12)]; T3g = FMA(KP618033988, T3f, T3e); T68 = FNMS(KP618033988, T3e, T3f); T3j = T1o - T1r; T1s = T1o + T1r; Tw = Tu + Tv; T3l = Tu - Tv; T1t = T1l + T1s; T3i = FMS(KP250000000, T1s, T1l); Ty = ci[WS(rs, 7)]; } } } } { E T3n, T65, T3c, T3b, T2P, T2M, T4W; { E TA, T3m, Tz, TB, Ts; T3m = Ty - Tx; Tz = Tx + Ty; T1S = T1a + T1j; T1k = T1a - T1j; T3n = FNMS(KP618033988, T3m, T3l); T65 = FMA(KP618033988, T3l, T3m); TA = Tw + Tz; T3c = Tz - Tw; T3b = FNMS(KP250000000, TA, Tt); TB = Tt + TA; T1T = T1t + T1C; T1D = T1t - T1C; T1Y = Ti - Tr; Ts = Ti + Tr; { E T2I, T6j, T6m, TL; T2I = FMA(KP559016994, T2H, T2G); T6j = FNMS(KP559016994, T2H, T2G); T6m = FNMS(KP559016994, T2O, T2N); T2P = FMA(KP559016994, T2O, T2N); TL = TB + TK; T1Z = TB - TK; { E T6l, T7y, T6o, T7z; T6l = FMA(KP951056516, T6k, T6j); T7y = FNMS(KP951056516, T6k, T6j); T6o = FMA(KP951056516, T6n, T6m); T7z = FNMS(KP951056516, T6n, T6m); T10 = Ts - TL; TM = Ts + TL; T2M = FNMS(KP951056516, T2L, T2I); T4W = FMA(KP951056516, T2L, T2I); T7K = FMA(KP939062505, T7y, T7z); T7A = FNMS(KP939062505, T7z, T7y); T6p = FNMS(KP549754652, T6o, T6l); T6w = FMA(KP549754652, T6l, T6o); } } } { E T34, T31, T4Y, T60, T3s, T3z, T5X; { E T2X, T6c, T6f, T4V, T2T; T2X = FNMS(KP559016994, T2W, T2V); T6c = FMA(KP559016994, T2W, T2V); T6f = FMA(KP559016994, T33, T32); T34 = FNMS(KP559016994, T33, T32); T4V = FNMS(KP951056516, T2S, T2P); T2T = FMA(KP951056516, T2S, T2P); { E T7w, T6e, T7v, T6h; T7w = FMA(KP951056516, T6d, T6c); T6e = FNMS(KP951056516, T6d, T6c); T7v = FMA(KP951056516, T6g, T6f); T6h = FNMS(KP951056516, T6g, T6f); T4X = FNMS(KP634619297, T4W, T4V); T56 = FMA(KP634619297, T4V, T4W); T3K = FMA(KP256756360, T2M, T2T); T2U = FNMS(KP256756360, T2T, T2M); T7x = FMA(KP126329378, T7w, T7v); T7J = FNMS(KP126329378, T7v, T7w); T6v = FNMS(KP470564281, T6e, T6h); T6i = FMA(KP470564281, T6h, T6e); T31 = FMA(KP951056516, T30, T2X); T4Y = FNMS(KP951056516, T30, T2X); } T60 = FMA(KP559016994, T3r, T3q); T3s = FNMS(KP559016994, T3r, T3q); T3z = FNMS(KP559016994, T3y, T3x); T5X = FMA(KP559016994, T3y, T3x); } { E T5Z, T7r, T4Z, T38; T4Z = FNMS(KP951056516, T37, T34); T38 = FMA(KP951056516, T37, T34); { E T4O, T3w, T4P, T3D; T4O = FMA(KP951056516, T3v, T3s); T3w = FNMS(KP951056516, T3v, T3s); T4P = FMA(KP951056516, T3C, T3z); T3D = FNMS(KP951056516, T3C, T3z); T50 = FNMS(KP827271945, T4Z, T4Y); T57 = FMA(KP827271945, T4Y, T4Z); T3L = FMA(KP634619297, T31, T38); T39 = FNMS(KP634619297, T38, T31); T4Q = FNMS(KP126329378, T4P, T4O); T59 = FMA(KP126329378, T4O, T4P); T3O = FNMS(KP939062505, T3w, T3D); T3E = FMA(KP939062505, T3D, T3w); T5Z = FMA(KP951056516, T5Y, T5X); T7r = FNMS(KP951056516, T5Y, T5X); } { E T3d, T3k, T64, T7s, T62; T67 = FMA(KP559016994, T3c, T3b); T3d = FNMS(KP559016994, T3c, T3b); T3k = FNMS(KP559016994, T3j, T3i); T64 = FMA(KP559016994, T3j, T3i); T7s = FNMS(KP951056516, T61, T60); T62 = FMA(KP951056516, T61, T60); { E T4S, T3h, T4R, T3o; T4S = FMA(KP951056516, T3g, T3d); T3h = FNMS(KP951056516, T3g, T3d); T4R = FMA(KP951056516, T3n, T3k); T3o = FNMS(KP951056516, T3n, T3k); T7t = FNMS(KP827271945, T7s, T7r); T7H = FMA(KP827271945, T7r, T7s); T6y = FNMS(KP062914667, T5Z, T62); T63 = FMA(KP062914667, T62, T5Z); T4T = FNMS(KP470564281, T4S, T4R); T5a = FMA(KP470564281, T4R, T4S); T3N = FNMS(KP549754652, T3h, T3o); T3p = FMA(KP549754652, T3o, T3h); T66 = FNMS(KP951056516, T65, T64); T7o = FMA(KP951056516, T65, T64); } } } } } } } { E T7q, T7G, T6J, T6I, T6q, T6b, T6B, T73, T6Q, T78, T6z, T6a; cr[0] = T9 + TM; { E T1U, T2l, T1X, T2g, T1E, TZ, T2m, T20, T2v, T2n; { E T1W, T7p, T69, T1V; T1W = T1S - T1T; T1U = T1S + T1T; T7p = FNMS(KP951056516, T68, T67); T69 = FMA(KP951056516, T68, T67); T1V = FNMS(KP250000000, T1U, T1R); T7q = FMA(KP062914667, T7p, T7o); T7G = FNMS(KP062914667, T7o, T7p); T6z = FNMS(KP634619297, T66, T69); T6a = FMA(KP634619297, T69, T66); T2l = FNMS(KP559016994, T1W, T1V); T1X = FMA(KP559016994, T1W, T1V); T2g = FNMS(KP618033988, T1k, T1D); T1E = FMA(KP618033988, T1D, T1k); TZ = FNMS(KP250000000, TM, T9); T2m = FNMS(KP618033988, T1Y, T1Z); T20 = FMA(KP618033988, T1Z, T1Y); } ci[0] = T1R + T1U; T2v = FMA(KP951056516, T2m, T2l); T2n = FNMS(KP951056516, T2m, T2l); { E T2b, T21, T2f, T11; T2b = FNMS(KP951056516, T20, T1X); T21 = FMA(KP951056516, T20, T1X); T2f = FNMS(KP559016994, T10, TZ); T11 = FMA(KP559016994, T10, TZ); { E T2h, T2r, T27, T1F; T2h = FMA(KP951056516, T2g, T2f); T2r = FNMS(KP951056516, T2g, T2f); T27 = FMA(KP951056516, T1E, T11); T1F = FNMS(KP951056516, T1E, T11); { E T2o, T2i, T2w, T2s; T2o = T2k * T2h; T2i = T2e * T2h; T2w = T2u * T2r; T2s = T2q * T2r; { E T2c, T28, T22, T1G; T2c = T2a * T27; T28 = T26 * T27; T22 = T1I * T1F; T1G = TY * T1F; ci[WS(rs, 15)] = FMA(T2q, T2v, T2w); cr[WS(rs, 15)] = FNMS(T2u, T2v, T2s); ci[WS(rs, 20)] = FMA(T26, T2b, T2c); cr[WS(rs, 20)] = FNMS(T2a, T2b, T28); ci[WS(rs, 5)] = FMA(TY, T21, T22); cr[WS(rs, 5)] = FNMS(T1I, T21, T1G); cr[WS(rs, 10)] = FNMS(T2k, T2n, T2i); ci[WS(rs, 10)] = FMA(T2e, T2n, T2o); } } } } } { E T6x, T6A, T6O, T6P; T6x = FMA(KP968479752, T6w, T6v); T6J = FNMS(KP968479752, T6w, T6v); T6I = FMA(KP845997307, T6z, T6y); T6A = FNMS(KP845997307, T6z, T6y); T6O = FNMS(KP968479752, T6p, T6i); T6q = FMA(KP968479752, T6p, T6i); T6b = FMA(KP845997307, T6a, T63); T6P = FNMS(KP845997307, T6a, T63); T6B = FNMS(KP681693190, T6A, T6x); T73 = FMA(KP560319534, T6x, T6A); T6Q = FMA(KP681693190, T6P, T6O); T78 = FNMS(KP560319534, T6O, T6P); } { E T7U, T8f, T7B, T7u, T82, T8k, T7Y, T7M; { E T7L, T7I, T80, T81; { E T7S, T6r, T6t, T6K, T6M, T7T, T6s, T7j; T7S = FNMS(KP734762448, T7K, T7J); T7L = FMA(KP734762448, T7K, T7J); T6r = FMA(KP906616052, T6q, T6b); T6t = FNMS(KP906616052, T6q, T6b); T6K = FNMS(KP906616052, T6J, T6I); T6M = FMA(KP906616052, T6J, T6I); T7I = FMA(KP772036680, T7H, T7G); T7T = FNMS(KP772036680, T7H, T7G); T6s = FNMS(KP249506682, T6r, T5W); T7j = FMA(KP998026728, T6r, T5W); { E T6L, T7l, T72, T6u; T6L = FNMS(KP249506682, T6K, T6H); T7l = FMA(KP998026728, T6K, T6H); T72 = FMA(KP557913902, T6t, T6s); T6u = FNMS(KP557913902, T6t, T6s); { E T7k, T6N, T77, T7m; T7k = T4l * T7j; T6N = FNMS(KP557913902, T6M, T6L); T77 = FMA(KP557913902, T6M, T6L); T7m = T4l * T7l; { E T74, T7d, T6V, T6C; T74 = FNMS(KP949179823, T73, T72); T7d = FMA(KP949179823, T73, T72); T6V = FMA(KP860541664, T6B, T6u); T6C = FNMS(KP860541664, T6B, T6u); cr[WS(rs, 2)] = FNMS(T4n, T7l, T7k); { E T7h, T79, T6R, T6Z; T7h = FNMS(KP949179823, T78, T77); T79 = FMA(KP949179823, T78, T77); T6R = FNMS(KP860541664, T6Q, T6N); T6Z = FMA(KP860541664, T6Q, T6N); ci[WS(rs, 2)] = FMA(T4n, T7j, T7m); { E T75, T7e, T6W, T6D; T75 = T71 * T74; T7e = T7c * T7d; T6W = T6U * T6V; T6D = T5T * T6C; { E T7a, T7i, T70, T6S; T7a = T71 * T79; T7i = T7c * T7h; T70 = T6U * T6Z; T6S = T5T * T6R; cr[WS(rs, 12)] = FNMS(T76, T79, T75); cr[WS(rs, 17)] = FNMS(T7g, T7h, T7e); cr[WS(rs, 22)] = FNMS(T6Y, T6Z, T6W); cr[WS(rs, 7)] = FNMS(T6E, T6R, T6D); ci[WS(rs, 12)] = FMA(T76, T74, T7a); ci[WS(rs, 17)] = FMA(T7g, T7d, T7i); ci[WS(rs, 22)] = FMA(T6Y, T6V, T70); ci[WS(rs, 7)] = FMA(T6E, T6C, T6S); T7U = FNMS(KP621716863, T7T, T7S); T8f = FMA(KP614372930, T7S, T7T); } } } } } } } T80 = FNMS(KP734762448, T7A, T7x); T7B = FMA(KP734762448, T7A, T7x); T7u = FMA(KP772036680, T7t, T7q); T81 = FNMS(KP772036680, T7t, T7q); T82 = FNMS(KP621716863, T81, T80); T8k = FMA(KP614372930, T80, T81); T7Y = FNMS(KP994076283, T7L, T7I); T7M = FMA(KP994076283, T7L, T7I); } { E T5y, T5c, T51, T4U, T5f, T5E, T5o, T5i, T5k; { E T5h, T5g, T5m, T5n, T58, T5b; T5h = FMA(KP912575812, T57, T56); T58 = FNMS(KP912575812, T57, T56); T5b = FNMS(KP912018591, T5a, T59); T5g = FMA(KP912018591, T5a, T59); { E T7X, T7N, T7C, T7Q; T7X = FNMS(KP249506682, T7M, T7F); T7N = FMA(KP998026728, T7M, T7F); T7C = FMA(KP994076283, T7B, T7u); T7Q = FNMS(KP994076283, T7B, T7u); T5y = FMA(KP525970792, T58, T5b); T5c = FNMS(KP726211448, T5b, T58); { E T7Z, T8j, T7P, T7D; T7Z = FNMS(KP557913902, T7Y, T7X); T8j = FMA(KP557913902, T7Y, T7X); T7P = FNMS(KP249506682, T7C, T7n); T7D = FMA(KP998026728, T7C, T7n); { E T8b, T83, T8t, T8l; T8b = FMA(KP943557151, T82, T7Z); T83 = FNMS(KP943557151, T82, T7Z); T8t = FMA(KP949179823, T8k, T8j); T8l = FNMS(KP949179823, T8k, T8j); { E T8e, T7R, T7O, T7E; T8e = FMA(KP557913902, T7Q, T7P); T7R = FNMS(KP557913902, T7Q, T7P); T7O = TR * T7D; T7E = TO * T7D; { E T8g, T8p, T7V, T87; T8g = FMA(KP949179823, T8f, T8e); T8p = FNMS(KP949179823, T8f, T8e); T7V = FMA(KP943557151, T7U, T7R); T87 = FNMS(KP943557151, T7U, T7R); ci[WS(rs, 3)] = FMA(TO, T7N, T7O); cr[WS(rs, 3)] = FNMS(TR, T7N, T7E); { E T8m, T8h, T8u, T8q; T8m = T8i * T8g; T8h = T8d * T8g; T8u = T8s * T8p; T8q = T8o * T8p; { E T84, T7W, T8c, T88; T84 = T4B * T7V; T7W = T4z * T7V; T8c = T8a * T87; T88 = T86 * T87; ci[WS(rs, 13)] = FMA(T8d, T8l, T8m); cr[WS(rs, 13)] = FNMS(T8i, T8l, T8h); ci[WS(rs, 18)] = FMA(T8o, T8t, T8u); cr[WS(rs, 18)] = FNMS(T8s, T8t, T8q); ci[WS(rs, 8)] = FMA(T4z, T83, T84); cr[WS(rs, 8)] = FNMS(T4B, T83, T7W); ci[WS(rs, 23)] = FMA(T86, T8b, T8c); cr[WS(rs, 23)] = FNMS(T8a, T8b, T88); } } } } } } } T51 = FMA(KP912575812, T50, T4X); T5m = FNMS(KP912575812, T50, T4X); T5n = FMA(KP912018591, T4T, T4Q); T4U = FNMS(KP912018591, T4T, T4Q); T41 = FMA(KP951056516, T40, T3X); T5f = FNMS(KP951056516, T40, T3X); T5E = FMA(KP525970792, T5m, T5n); T5o = FNMS(KP726211448, T5n, T5m); T5i = FMA(KP851038619, T5h, T5g); T5k = FNMS(KP851038619, T5h, T5g); } { E T42, T43, T48, T49, T3M, T3P; T3M = FMA(KP871714437, T3L, T3K); T42 = FNMS(KP871714437, T3L, T3K); T43 = FMA(KP831864738, T3O, T3N); T3P = FNMS(KP831864738, T3O, T3N); { E T5R, T5j, T54, T52; T5R = FMA(KP992114701, T5i, T5f); T5j = FNMS(KP248028675, T5i, T5f); T54 = FNMS(KP851038619, T51, T4U); T52 = FMA(KP851038619, T51, T4U); T3Q = FNMS(KP559154169, T3P, T3M); T4q = FMA(KP683113946, T3M, T3P); { E T5D, T5l, T5P, T53; T5D = FMA(KP554608978, T5k, T5j); T5l = FNMS(KP554608978, T5k, T5j); T5P = FNMS(KP992114701, T52, T4N); T53 = FMA(KP248028675, T52, T4N); { E T5p, T5t, T5F, T5N; T5p = FNMS(KP803003575, T5o, T5l); T5t = FMA(KP803003575, T5o, T5l); T5F = FNMS(KP943557151, T5E, T5D); T5N = FMA(KP943557151, T5E, T5D); { E T55, T5x, T5S, T5Q; T55 = FMA(KP554608978, T54, T53); T5x = FNMS(KP554608978, T54, T53); T5S = TW * T5P; T5Q = TS * T5P; { E T5J, T5z, T5r, T5d; T5J = FMA(KP943557151, T5y, T5x); T5z = FNMS(KP943557151, T5y, T5x); T5r = FMA(KP803003575, T5c, T55); T5d = FNMS(KP803003575, T5c, T55); ci[WS(rs, 4)] = FMA(TS, T5R, T5S); cr[WS(rs, 4)] = FNMS(TW, T5R, T5Q); { E T5G, T5A, T5O, T5K; T5G = T5C * T5z; T5A = T5w * T5z; T5O = T5M * T5J; T5K = T5I * T5J; { E T5q, T5e, T5u, T5s; T5q = TX * T5d; T5e = TT * T5d; T5u = T25 * T5r; T5s = T23 * T5r; ci[WS(rs, 14)] = FMA(T5w, T5F, T5G); cr[WS(rs, 14)] = FNMS(T5C, T5F, T5A); ci[WS(rs, 19)] = FMA(T5I, T5N, T5O); cr[WS(rs, 19)] = FNMS(T5M, T5N, T5K); ci[WS(rs, 9)] = FMA(TT, T5p, T5q); cr[WS(rs, 9)] = FNMS(TX, T5p, T5e); ci[WS(rs, 24)] = FMA(T23, T5t, T5u); cr[WS(rs, 24)] = FNMS(T25, T5t, T5s); } } } } } } } T48 = FNMS(KP871714437, T39, T2U); T3a = FMA(KP871714437, T39, T2U); T3F = FMA(KP831864738, T3E, T3p); T49 = FNMS(KP831864738, T3E, T3p); T4a = FMA(KP559154169, T49, T48); T4w = FNMS(KP683113946, T48, T49); T46 = FMA(KP904730450, T43, T42); T44 = FNMS(KP904730450, T43, T42); } } } } } } { E T45, T4L, T3G, T3I; T45 = FNMS(KP242145790, T44, T41); T4L = FMA(KP968583161, T44, T41); T3G = FMA(KP904730450, T3F, T3a); T3I = FNMS(KP904730450, T3F, T3a); { E T4v, T47, T4J, T3H; T4v = FNMS(KP541454447, T46, T45); T47 = FMA(KP541454447, T46, T45); T4J = FMA(KP968583161, T3G, T2F); T3H = FNMS(KP242145790, T3G, T2F); { E T4b, T4j, T4x, T4H; T4b = FMA(KP921177326, T4a, T47); T4j = FNMS(KP921177326, T4a, T47); T4x = FNMS(KP833417178, T4w, T4v); T4H = FMA(KP833417178, T4w, T4v); { E T3J, T4p, T4M, T4K; T3J = FMA(KP541454447, T3I, T3H); T4p = FNMS(KP541454447, T3I, T3H); T4M = TQ * T4J; T4K = TN * T4J; { E T4D, T4r, T4f, T3R; T4D = FMA(KP833417178, T4q, T4p); T4r = FNMS(KP833417178, T4q, T4p); T4f = FMA(KP921177326, T3Q, T3J); T3R = FNMS(KP921177326, T3Q, T3J); ci[WS(rs, 1)] = FMA(TN, T4L, T4M); cr[WS(rs, 1)] = FNMS(TQ, T4L, T4K); { E T4y, T4s, T4I, T4E; T4y = T4u * T4r; T4s = T4o * T4r; T4I = T4G * T4D; T4E = T4C * T4D; { E T4c, T3S, T4k, T4g; T4c = T3U * T3R; T3S = T2y * T3R; T4k = T4i * T4f; T4g = T4e * T4f; ci[WS(rs, 11)] = FMA(T4o, T4x, T4y); cr[WS(rs, 11)] = FNMS(T4u, T4x, T4s); ci[WS(rs, 16)] = FMA(T4C, T4H, T4I); cr[WS(rs, 16)] = FNMS(T4G, T4H, T4E); ci[WS(rs, 6)] = FMA(T2y, T4b, T4c); cr[WS(rs, 6)] = FNMS(T3U, T4b, T3S); ci[WS(rs, 21)] = FMA(T4e, T4j, T4k); cr[WS(rs, 21)] = FNMS(T4i, T4j, T4g); } } } } } } } } } } static const tw_instr twinstr[] = { {TW_CEXP, 1, 1}, {TW_CEXP, 1, 3}, {TW_CEXP, 1, 9}, {TW_CEXP, 1, 24}, {TW_NEXT, 1, 0} }; static const hc2hc_desc desc = { 25, "hb2_25", twinstr, &GENUS, {84, 78, 356, 0} }; void X(codelet_hb2_25) (planner *p) { X(khc2hc_register) (p, hb2_25, &desc); } #else /* HAVE_FMA */ /* Generated by: ../../../genfft/gen_hc2hc.native -compact -variables 4 -pipeline-latency 4 -sign 1 -twiddle-log3 -precompute-twiddles -n 25 -dif -name hb2_25 -include hb.h */ /* * This function contains 440 FP additions, 340 FP multiplications, * (or, 280 additions, 180 multiplications, 160 fused multiply/add), * 155 stack variables, 20 constants, and 100 memory accesses */ #include "hb.h" static void hb2_25(R *cr, R *ci, const R *W, stride rs, INT mb, INT me, INT ms) { DK(KP998026728, +0.998026728428271561952336806863450553336905220); DK(KP062790519, +0.062790519529313376076178224565631133122484832); DK(KP992114701, +0.992114701314477831049793042785778521453036709); DK(KP125333233, +0.125333233564304245373118759816508793942918247); DK(KP425779291, +0.425779291565072648862502445744251703979973042); DK(KP904827052, +0.904827052466019527713668647932697593970413911); DK(KP248689887, +0.248689887164854788242283746006447968417567406); DK(KP968583161, +0.968583161128631119490168375464735813836012403); DK(KP770513242, +0.770513242775789230803009636396177847271667672); DK(KP637423989, +0.637423989748689710176712811676016195434917298); DK(KP844327925, +0.844327925502015078548558063966681505381659241); DK(KP535826794, +0.535826794978996618271308767867639978063575346); DK(KP684547105, +0.684547105928688673732283357621209269889519233); DK(KP728968627, +0.728968627421411523146730319055259111372571664); DK(KP481753674, +0.481753674101715274987191502872129653528542010); DK(KP876306680, +0.876306680043863587308115903922062583399064238); DK(KP559016994, +0.559016994374947424102293417182819058860154590); DK(KP250000000, +0.250000000000000000000000000000000000000000000); DK(KP587785252, +0.587785252292473129168705954639072768597652438); DK(KP951056516, +0.951056516295153572116439333379382143405698634); { INT m; for (m = mb, W = W + ((mb - 1) * 8); m < me; m = m + 1, cr = cr + ms, ci = ci - ms, W = W + 8, MAKE_VOLATILE_STRIDE(50, rs)) { E TN, TQ, TO, TR, TT, TY, T2t, T2r, TZ, TU, T4f, T4l, T2d, T4v, T5m; E T2j, T5l, T4X, T2v, T11, T3R, T1L, T5d, T6x, T5h, T6t, T25, T26, T27, T29; E T6D, T7v, T49, T7l, T7p, T7t, T2p, T2n, T4b, T4p, T5n, T6B, T5b, T5p, T6p; E T6r, T59, T4r; { E T2c, T4j, T2h, T4e, T2b, T4k, T2i, T4d; { E TP, TX, TS, TW; TN = W[0]; TQ = W[1]; TO = W[2]; TR = W[3]; TP = TN * TO; TX = TQ * TO; TS = TQ * TR; TW = TN * TR; TT = TP - TS; TY = TW + TX; T2t = TW - TX; T2r = TP + TS; TZ = W[5]; T2c = TQ * TZ; T4j = TO * TZ; T2h = TN * TZ; T4e = TR * TZ; TU = W[4]; T2b = TN * TU; T4k = TR * TU; T2i = TQ * TU; T4d = TO * TU; } T4f = T4d - T4e; T4l = T4j + T4k; { E T2s, T2u, TV, T10, T3P, T3Q, T1J, T1K; T2d = T2b - T2c; T4v = T2b + T2c; T5m = T4j - T4k; T2j = T2h + T2i; T5l = T4d + T4e; T4X = T2h - T2i; T2s = T2r * TU; T2u = T2t * TZ; T2v = T2s + T2u; TV = TT * TU; T10 = TY * TZ; T11 = TV + T10; T3P = T2r * TZ; T3Q = T2t * TU; T3R = T3P - T3Q; T1J = TT * TZ; T1K = TY * TU; T1L = T1J - T1K; T5d = TV - T10; T6x = T3P + T3Q; T5h = T1J + T1K; T6t = T2s - T2u; T25 = W[6]; T26 = W[7]; T27 = FMA(TT, T25, TY * T26); T29 = FNMS(TY, T25, TT * T26); T6D = FNMS(T4X, T25, T4v * T26); T7v = FNMS(T1L, T25, T11 * T26); T49 = FMA(T2r, T25, T2t * T26); T7l = FMA(T2d, T25, T2j * T26); T7p = FNMS(T2j, T25, T2d * T26); T7t = FMA(T11, T25, T1L * T26); T2p = FNMS(TZ, T25, TU * T26); T2n = FMA(TU, T25, TZ * T26); T4b = FNMS(T2t, T25, T2r * T26); T4p = FMA(T2v, T25, T3R * T26); T5n = FMA(T5l, T25, T5m * T26); T6B = FMA(T4v, T25, T4X * T26); T5b = FNMS(TQ, T25, TN * T26); T5p = FNMS(T5m, T25, T5l * T26); T6p = FMA(TO, T25, TR * T26); T6r = FNMS(TR, T25, TO * T26); T59 = FMA(TN, T25, TQ * T26); T4r = FNMS(T3R, T25, T2v * T26); } } { E T9, T6i, T40, T3z, T5Y, Ti, Tr, Ts, T1d, T1m, T1P, T2K, T4P, T3H, T4y; E T5G, T71, T65, T6N, T5z, T70, T64, T6K, T2Z, T4Q, T3I, T4B, T20, T5Z, T3C; E T43, T6j, TB, TK, TL, T1w, T1F, T1Q, T3f, T4S, T3K, T4F, T5V, T74, T68; E T6U, T5O, T73, T67, T6R, T3u, T4T, T3L, T4I; { E T1, T4, T7, T8, T3Z, T3Y, T3x, T3y; T1 = cr[0]; { E T2, T3, T5, T6; T2 = cr[WS(rs, 5)]; T3 = ci[WS(rs, 4)]; T4 = T2 + T3; T5 = cr[WS(rs, 10)]; T6 = ci[WS(rs, 9)]; T7 = T5 + T6; T8 = T4 + T7; T3Z = T5 - T6; T3Y = T2 - T3; } T9 = T1 + T8; T6i = FMA(KP951056516, T3Y, KP587785252 * T3Z); T40 = FNMS(KP951056516, T3Z, KP587785252 * T3Y); T3x = FNMS(KP250000000, T8, T1); T3y = KP559016994 * (T4 - T7); T3z = T3x - T3y; T5Y = T3y + T3x; } { E Ta, T2x, T5w, T2F, Th, T2w, T1e, T2P, T5B, T2X, T1l, T2O, Tj, T2N, T5D; E T2T, Tq, T2S, T15, T2B, T5u, T2H, T1c, T2G; { E Tg, T2E, Td, T2D; Ta = cr[WS(rs, 1)]; { E Te, Tf, Tb, Tc; Te = cr[WS(rs, 11)]; Tf = ci[WS(rs, 8)]; Tg = Te + Tf; T2E = Te - Tf; Tb = cr[WS(rs, 6)]; Tc = ci[WS(rs, 3)]; Td = Tb + Tc; T2D = Tb - Tc; } T2x = KP559016994 * (Td - Tg); T5w = FMA(KP951056516, T2D, KP587785252 * T2E); T2F = FNMS(KP951056516, T2E, KP587785252 * T2D); Th = Td + Tg; T2w = FNMS(KP250000000, Th, Ta); } { E T1k, T2W, T1h, T2V; T1e = ci[WS(rs, 20)]; { E T1i, T1j, T1f, T1g; T1i = cr[WS(rs, 14)]; T1j = cr[WS(rs, 19)]; T1k = T1i + T1j; T2W = T1j - T1i; T1f = ci[WS(rs, 15)]; T1g = cr[WS(rs, 24)]; T1h = T1f - T1g; T2V = T1f + T1g; } T2P = KP559016994 * (T1h + T1k); T5B = FMA(KP951056516, T2V, KP587785252 * T2W); T2X = FNMS(KP951056516, T2W, KP587785252 * T2V); T1l = T1h - T1k; T2O = FNMS(KP250000000, T1l, T1e); } { E Tp, T2M, Tm, T2L; Tj = cr[WS(rs, 4)]; { E Tn, To, Tk, Tl; Tn = ci[WS(rs, 10)]; To = ci[WS(rs, 5)]; Tp = Tn + To; T2M = Tn - To; Tk = cr[WS(rs, 9)]; Tl = ci[0]; Tm = Tk + Tl; T2L = Tk - Tl; } T2N = FNMS(KP951056516, T2M, KP587785252 * T2L); T5D = FMA(KP951056516, T2L, KP587785252 * T2M); T2T = KP559016994 * (Tm - Tp); Tq = Tm + Tp; T2S = FNMS(KP250000000, Tq, Tj); } { E T1b, T2A, T18, T2z; T15 = ci[WS(rs, 23)]; { E T19, T1a, T16, T17; T19 = ci[WS(rs, 13)]; T1a = cr[WS(rs, 16)]; T1b = T19 - T1a; T2A = T19 + T1a; T16 = ci[WS(rs, 18)]; T17 = cr[WS(rs, 21)]; T18 = T16 - T17; T2z = T16 + T17; } T2B = FNMS(KP951056516, T2A, KP587785252 * T2z); T5u = FMA(KP951056516, T2z, KP587785252 * T2A); T2H = KP559016994 * (T18 - T1b); T1c = T18 + T1b; T2G = FNMS(KP250000000, T1c, T15); } Ti = Ta + Th; Tr = Tj + Tq; Ts = Ti + Tr; T1d = T15 + T1c; T1m = T1e + T1l; T1P = T1d + T1m; { E T2C, T4w, T2J, T4x, T2y, T2I; T2y = T2w - T2x; T2C = T2y - T2B; T4w = T2y + T2B; T2I = T2G - T2H; T2J = T2F + T2I; T4x = T2I - T2F; T2K = FNMS(KP481753674, T2J, KP876306680 * T2C); T4P = FMA(KP728968627, T4x, KP684547105 * T4w); T3H = FMA(KP876306680, T2J, KP481753674 * T2C); T4y = FNMS(KP684547105, T4x, KP728968627 * T4w); } { E T5C, T6M, T5F, T6L, T5A, T5E; T5A = T2T + T2S; T5C = T5A - T5B; T6M = T5A + T5B; T5E = T2O + T2P; T5F = T5D + T5E; T6L = T5E - T5D; T5G = FNMS(KP844327925, T5F, KP535826794 * T5C); T71 = FMA(KP637423989, T6L, KP770513242 * T6M); T65 = FMA(KP535826794, T5F, KP844327925 * T5C); T6N = FNMS(KP637423989, T6M, KP770513242 * T6L); } { E T5v, T6I, T5y, T6J, T5t, T5x; T5t = T2x + T2w; T5v = T5t - T5u; T6I = T5t + T5u; T5x = T2H + T2G; T5y = T5w + T5x; T6J = T5x - T5w; T5z = FNMS(KP248689887, T5y, KP968583161 * T5v); T70 = FMA(KP535826794, T6J, KP844327925 * T6I); T64 = FMA(KP968583161, T5y, KP248689887 * T5v); T6K = FNMS(KP844327925, T6J, KP535826794 * T6I); } { E T2R, T4z, T2Y, T4A, T2Q, T2U; T2Q = T2O - T2P; T2R = T2N + T2Q; T4z = T2Q - T2N; T2U = T2S - T2T; T2Y = T2U - T2X; T4A = T2U + T2X; T2Z = FMA(KP904827052, T2R, KP425779291 * T2Y); T4Q = FNMS(KP992114701, T4z, KP125333233 * T4A); T3I = FNMS(KP425779291, T2R, KP904827052 * T2Y); T4B = FMA(KP125333233, T4z, KP992114701 * T4A); } } { E T1S, T1V, T1Y, T1Z, T3B, T3A, T41, T42; T1S = ci[WS(rs, 24)]; { E T1T, T1U, T1W, T1X; T1T = ci[WS(rs, 19)]; T1U = cr[WS(rs, 20)]; T1V = T1T - T1U; T1W = ci[WS(rs, 14)]; T1X = cr[WS(rs, 15)]; T1Y = T1W - T1X; T1Z = T1V + T1Y; T3B = T1W + T1X; T3A = T1T + T1U; } T20 = T1S + T1Z; T5Z = FMA(KP951056516, T3A, KP587785252 * T3B); T3C = FNMS(KP951056516, T3B, KP587785252 * T3A); T41 = FNMS(KP250000000, T1Z, T1S); T42 = KP559016994 * (T1V - T1Y); T43 = T41 - T42; T6j = T42 + T41; } { E Tt, T32, T5L, T3a, TA, T31, T1o, T36, T5J, T3c, T1v, T3b, TC, T3h, T5S; E T3p, TJ, T3g, T1x, T3l, T5Q, T3r, T1E, T3q; { E Tw, T38, Tz, T39; Tt = cr[WS(rs, 2)]; { E Tu, Tv, Tx, Ty; Tu = cr[WS(rs, 7)]; Tv = ci[WS(rs, 2)]; Tw = Tu + Tv; T38 = Tu - Tv; Tx = cr[WS(rs, 12)]; Ty = ci[WS(rs, 7)]; Tz = Tx + Ty; T39 = Tx - Ty; } T32 = KP559016994 * (Tw - Tz); T5L = FMA(KP951056516, T38, KP587785252 * T39); T3a = FNMS(KP951056516, T39, KP587785252 * T38); TA = Tw + Tz; T31 = FNMS(KP250000000, TA, Tt); } { E T1r, T34, T1u, T35; T1o = ci[WS(rs, 22)]; { E T1p, T1q, T1s, T1t; T1p = ci[WS(rs, 17)]; T1q = cr[WS(rs, 22)]; T1r = T1p - T1q; T34 = T1p + T1q; T1s = ci[WS(rs, 12)]; T1t = cr[WS(rs, 17)]; T1u = T1s - T1t; T35 = T1s + T1t; } T36 = FNMS(KP951056516, T35, KP587785252 * T34); T5J = FMA(KP951056516, T34, KP587785252 * T35); T3c = KP559016994 * (T1r - T1u); T1v = T1r + T1u; T3b = FNMS(KP250000000, T1v, T1o); } { E TI, T3o, TF, T3n; TC = cr[WS(rs, 3)]; { E TG, TH, TD, TE; TG = ci[WS(rs, 11)]; TH = ci[WS(rs, 6)]; TI = TG + TH; T3o = TG - TH; TD = cr[WS(rs, 8)]; TE = ci[WS(rs, 1)]; TF = TD + TE; T3n = TD - TE; } T3h = KP559016994 * (TF - TI); T5S = FMA(KP951056516, T3n, KP587785252 * T3o); T3p = FNMS(KP951056516, T3o, KP587785252 * T3n); TJ = TF + TI; T3g = FNMS(KP250000000, TJ, TC); } { E T1D, T3k, T1A, T3j; T1x = ci[WS(rs, 21)]; { E T1B, T1C, T1y, T1z; T1B = cr[WS(rs, 13)]; T1C = cr[WS(rs, 18)]; T1D = T1B + T1C; T3k = T1C - T1B; T1y = ci[WS(rs, 16)]; T1z = cr[WS(rs, 23)]; T1A = T1y - T1z; T3j = T1y + T1z; } T3l = FNMS(KP951056516, T3k, KP587785252 * T3j); T5Q = FMA(KP951056516, T3j, KP587785252 * T3k); T3r = KP559016994 * (T1A + T1D); T1E = T1A - T1D; T3q = FNMS(KP250000000, T1E, T1x); } TB = Tt + TA; TK = TC + TJ; TL = TB + TK; T1w = T1o + T1v; T1F = T1x + T1E; T1Q = T1w + T1F; { E T37, T4D, T3e, T4E, T33, T3d; T33 = T31 - T32; T37 = T33 - T36; T4D = T33 + T36; T3d = T3b - T3c; T3e = T3a + T3d; T4E = T3d - T3a; T3f = FNMS(KP844327925, T3e, KP535826794 * T37); T4S = FMA(KP062790519, T4E, KP998026728 * T4D); T3K = FMA(KP535826794, T3e, KP844327925 * T37); T4F = FNMS(KP998026728, T4E, KP062790519 * T4D); } { E T5R, T6T, T5U, T6S, T5P, T5T; T5P = T3h + T3g; T5R = T5P - T5Q; T6T = T5P + T5Q; T5T = T3q + T3r; T5U = T5S + T5T; T6S = T5T - T5S; T5V = FNMS(KP684547105, T5U, KP728968627 * T5R); T74 = FNMS(KP992114701, T6S, KP125333233 * T6T); T68 = FMA(KP728968627, T5U, KP684547105 * T5R); T6U = FMA(KP125333233, T6S, KP992114701 * T6T); } { E T5K, T6Q, T5N, T6P, T5I, T5M; T5I = T32 + T31; T5K = T5I - T5J; T6Q = T5I + T5J; T5M = T3c + T3b; T5N = T5L + T5M; T6P = T5M - T5L; T5O = FNMS(KP481753674, T5N, KP876306680 * T5K); T73 = FNMS(KP425779291, T6P, KP904827052 * T6Q); T67 = FMA(KP876306680, T5N, KP481753674 * T5K); T6R = FMA(KP904827052, T6P, KP425779291 * T6Q); } { E T3m, T4H, T3t, T4G, T3i, T3s; T3i = T3g - T3h; T3m = T3i - T3l; T4H = T3i + T3l; T3s = T3q - T3r; T3t = T3p + T3s; T4G = T3s - T3p; T3u = FNMS(KP998026728, T3t, KP062790519 * T3m); T4T = FNMS(KP637423989, T4G, KP770513242 * T4H); T3L = FMA(KP062790519, T3t, KP998026728 * T3m); T4I = FMA(KP770513242, T4G, KP637423989 * T4H); } } { E TM, T14, T2e, T21, T23, T2l, T1H, T2f, T1O, T2k; { E T12, T13, T1R, T22; T12 = KP559016994 * (Ts - TL); TM = Ts + TL; T13 = FNMS(KP250000000, TM, T9); T14 = T12 + T13; T2e = T13 - T12; T1R = KP559016994 * (T1P - T1Q); T21 = T1P + T1Q; T22 = FNMS(KP250000000, T21, T20); T23 = T1R + T22; T2l = T22 - T1R; } { E T1n, T1G, T1M, T1N; T1n = T1d - T1m; T1G = T1w - T1F; T1H = FMA(KP951056516, T1n, KP587785252 * T1G); T2f = FNMS(KP951056516, T1G, KP587785252 * T1n); T1M = Ti - Tr; T1N = TB - TK; T1O = FMA(KP951056516, T1M, KP587785252 * T1N); T2k = FNMS(KP951056516, T1N, KP587785252 * T1M); } { E T1I, T24, T2o, T2q; cr[0] = T9 + TM; ci[0] = T20 + T21; T1I = T14 - T1H; T24 = T1O + T23; cr[WS(rs, 5)] = FNMS(T1L, T24, T11 * T1I); ci[WS(rs, 5)] = FMA(T1L, T1I, T11 * T24); T2o = T2e + T2f; T2q = T2l - T2k; cr[WS(rs, 15)] = FNMS(T2p, T2q, T2n * T2o); ci[WS(rs, 15)] = FMA(T2p, T2o, T2n * T2q); { E T2g, T2m, T28, T2a; T2g = T2e - T2f; T2m = T2k + T2l; cr[WS(rs, 10)] = FNMS(T2j, T2m, T2d * T2g); ci[WS(rs, 10)] = FMA(T2j, T2g, T2d * T2m); T28 = T14 + T1H; T2a = T23 - T1O; cr[WS(rs, 20)] = FNMS(T29, T2a, T27 * T28); ci[WS(rs, 20)] = FMA(T29, T28, T27 * T2a); } } } { E T76, T7n, T7a, T7q, T6H, T6W, T6X, T6Y, T7e, T7f, T7d, T7g, T7x, T7y; { E T72, T75, T78, T79; T72 = T70 + T71; T75 = T73 - T74; T76 = FMA(KP951056516, T72, KP587785252 * T75); T7n = FNMS(KP951056516, T75, KP587785252 * T72); T78 = T6K - T6N; T79 = T6U - T6R; T7a = FMA(KP951056516, T78, KP587785252 * T79); T7q = FNMS(KP951056516, T79, KP587785252 * T78); } { E T6O, T6V, T7b, T7c; T6H = T5Y + T5Z; T6O = T6K + T6N; T6V = T6R + T6U; T6W = T6O - T6V; T6X = FNMS(KP250000000, T6W, T6H); T6Y = KP559016994 * (T6O + T6V); T7e = T6j - T6i; T7b = T70 - T71; T7c = T73 + T74; T7f = T7b + T7c; T7d = KP559016994 * (T7b - T7c); T7g = FNMS(KP250000000, T7f, T7e); } T7x = T6H + T6W; T7y = T7e + T7f; cr[WS(rs, 4)] = FNMS(TY, T7y, TT * T7x); ci[WS(rs, 4)] = FMA(TY, T7x, TT * T7y); { E T7o, T7u, T7s, T7w, T7m, T7r; T7m = T6X - T6Y; T7o = T7m - T7n; T7u = T7m + T7n; T7r = T7g - T7d; T7s = T7q + T7r; T7w = T7r - T7q; cr[WS(rs, 14)] = FNMS(T7p, T7s, T7l * T7o); ci[WS(rs, 14)] = FMA(T7p, T7o, T7l * T7s); cr[WS(rs, 19)] = FNMS(T7v, T7w, T7t * T7u); ci[WS(rs, 19)] = FMA(T7v, T7u, T7t * T7w); } { E T77, T7j, T7i, T7k, T6Z, T7h; T6Z = T6X + T6Y; T77 = T6Z - T76; T7j = T6Z + T76; T7h = T7d + T7g; T7i = T7a + T7h; T7k = T7h - T7a; cr[WS(rs, 9)] = FNMS(TZ, T7i, TU * T77); ci[WS(rs, 9)] = FMA(TZ, T77, TU * T7i); cr[WS(rs, 24)] = FNMS(T26, T7k, T25 * T7j); ci[WS(rs, 24)] = FMA(T26, T7j, T25 * T7k); } } { E T3N, T4h, T3U, T4m, T3D, T3E, T3w, T3F, T44, T45, T3X, T46, T4t, T4u; { E T3J, T3M, T3S, T3T; T3J = T3H - T3I; T3M = T3K - T3L; T3N = FMA(KP951056516, T3J, KP587785252 * T3M); T4h = FNMS(KP951056516, T3M, KP587785252 * T3J); T3S = T2K + T2Z; T3T = T3f - T3u; T3U = FMA(KP951056516, T3S, KP587785252 * T3T); T4m = FNMS(KP951056516, T3T, KP587785252 * T3S); } { E T30, T3v, T3V, T3W; T3D = T3z - T3C; T30 = T2K - T2Z; T3v = T3f + T3u; T3E = T30 + T3v; T3w = KP559016994 * (T30 - T3v); T3F = FNMS(KP250000000, T3E, T3D); T44 = T40 + T43; T3V = T3H + T3I; T3W = T3K + T3L; T45 = T3V + T3W; T3X = KP559016994 * (T3V - T3W); T46 = FNMS(KP250000000, T45, T44); } T4t = T3D + T3E; T4u = T44 + T45; cr[WS(rs, 2)] = FNMS(T2t, T4u, T2r * T4t); ci[WS(rs, 2)] = FMA(T2t, T4t, T2r * T4u); { E T4i, T4q, T4o, T4s, T4g, T4n; T4g = T3F - T3w; T4i = T4g - T4h; T4q = T4g + T4h; T4n = T46 - T3X; T4o = T4m + T4n; T4s = T4n - T4m; cr[WS(rs, 12)] = FNMS(T4l, T4o, T4f * T4i); ci[WS(rs, 12)] = FMA(T4l, T4i, T4f * T4o); cr[WS(rs, 17)] = FNMS(T4r, T4s, T4p * T4q); ci[WS(rs, 17)] = FMA(T4r, T4q, T4p * T4s); } { E T3O, T4a, T48, T4c, T3G, T47; T3G = T3w + T3F; T3O = T3G - T3N; T4a = T3G + T3N; T47 = T3X + T46; T48 = T3U + T47; T4c = T47 - T3U; cr[WS(rs, 7)] = FNMS(T3R, T48, T2v * T3O); ci[WS(rs, 7)] = FMA(T3R, T3O, T2v * T48); cr[WS(rs, 22)] = FNMS(T4b, T4c, T49 * T4a); ci[WS(rs, 22)] = FMA(T4b, T4a, T49 * T4c); } } { E T4V, T5f, T50, T5i, T4L, T4M, T4K, T4N, T54, T55, T53, T56, T5r, T5s; { E T4R, T4U, T4Y, T4Z; T4R = T4P - T4Q; T4U = T4S - T4T; T4V = FMA(KP951056516, T4R, KP587785252 * T4U); T5f = FNMS(KP951056516, T4U, KP587785252 * T4R); T4Y = T4y + T4B; T4Z = T4F + T4I; T50 = FMA(KP951056516, T4Y, KP587785252 * T4Z); T5i = FNMS(KP951056516, T4Z, KP587785252 * T4Y); } { E T4C, T4J, T51, T52; T4L = T3z + T3C; T4C = T4y - T4B; T4J = T4F - T4I; T4M = T4C + T4J; T4K = KP559016994 * (T4C - T4J); T4N = FNMS(KP250000000, T4M, T4L); T54 = T43 - T40; T51 = T4P + T4Q; T52 = T4S + T4T; T55 = T51 + T52; T53 = KP559016994 * (T51 - T52); T56 = FNMS(KP250000000, T55, T54); } T5r = T4L + T4M; T5s = T54 + T55; cr[WS(rs, 3)] = FNMS(TR, T5s, TO * T5r); ci[WS(rs, 3)] = FMA(TR, T5r, TO * T5s); { E T5g, T5o, T5k, T5q, T5e, T5j; T5e = T4N - T4K; T5g = T5e - T5f; T5o = T5e + T5f; T5j = T56 - T53; T5k = T5i + T5j; T5q = T5j - T5i; cr[WS(rs, 13)] = FNMS(T5h, T5k, T5d * T5g); ci[WS(rs, 13)] = FMA(T5h, T5g, T5d * T5k); cr[WS(rs, 18)] = FNMS(T5p, T5q, T5n * T5o); ci[WS(rs, 18)] = FMA(T5p, T5o, T5n * T5q); } { E T4W, T5a, T58, T5c, T4O, T57; T4O = T4K + T4N; T4W = T4O - T4V; T5a = T4O + T4V; T57 = T53 + T56; T58 = T50 + T57; T5c = T57 - T50; cr[WS(rs, 8)] = FNMS(T4X, T58, T4v * T4W); ci[WS(rs, 8)] = FMA(T4X, T4W, T4v * T58); cr[WS(rs, 23)] = FNMS(T5b, T5c, T59 * T5a); ci[WS(rs, 23)] = FMA(T5b, T5a, T59 * T5c); } } { E T6a, T6v, T6e, T6y, T60, T61, T5X, T62, T6k, T6l, T6h, T6m, T6F, T6G; { E T66, T69, T6c, T6d; T66 = T64 - T65; T69 = T67 - T68; T6a = FMA(KP951056516, T66, KP587785252 * T69); T6v = FNMS(KP951056516, T69, KP587785252 * T66); T6c = T5z - T5G; T6d = T5O - T5V; T6e = FMA(KP951056516, T6c, KP587785252 * T6d); T6y = FNMS(KP951056516, T6d, KP587785252 * T6c); } { E T5H, T5W, T6f, T6g; T60 = T5Y - T5Z; T5H = T5z + T5G; T5W = T5O + T5V; T61 = T5H + T5W; T5X = KP559016994 * (T5H - T5W); T62 = FNMS(KP250000000, T61, T60); T6k = T6i + T6j; T6f = T64 + T65; T6g = T67 + T68; T6l = T6f + T6g; T6h = KP559016994 * (T6f - T6g); T6m = FNMS(KP250000000, T6l, T6k); } T6F = T60 + T61; T6G = T6k + T6l; cr[WS(rs, 1)] = FNMS(TQ, T6G, TN * T6F); ci[WS(rs, 1)] = FMA(TQ, T6F, TN * T6G); { E T6w, T6C, T6A, T6E, T6u, T6z; T6u = T62 - T5X; T6w = T6u - T6v; T6C = T6u + T6v; T6z = T6m - T6h; T6A = T6y + T6z; T6E = T6z - T6y; cr[WS(rs, 11)] = FNMS(T6x, T6A, T6t * T6w); ci[WS(rs, 11)] = FMA(T6x, T6w, T6t * T6A); cr[WS(rs, 16)] = FNMS(T6D, T6E, T6B * T6C); ci[WS(rs, 16)] = FMA(T6D, T6C, T6B * T6E); } { E T6b, T6q, T6o, T6s, T63, T6n; T63 = T5X + T62; T6b = T63 - T6a; T6q = T63 + T6a; T6n = T6h + T6m; T6o = T6e + T6n; T6s = T6n - T6e; cr[WS(rs, 6)] = FNMS(T5m, T6o, T5l * T6b); ci[WS(rs, 6)] = FMA(T5m, T6b, T5l * T6o); cr[WS(rs, 21)] = FNMS(T6r, T6s, T6p * T6q); ci[WS(rs, 21)] = FMA(T6r, T6q, T6p * T6s); } } } } } } static const tw_instr twinstr[] = { {TW_CEXP, 1, 1}, {TW_CEXP, 1, 3}, {TW_CEXP, 1, 9}, {TW_CEXP, 1, 24}, {TW_NEXT, 1, 0} }; static const hc2hc_desc desc = { 25, "hb2_25", twinstr, &GENUS, {280, 180, 160, 0} }; void X(codelet_hb2_25) (planner *p) { X(khc2hc_register) (p, hb2_25, &desc); } #endif /* HAVE_FMA */ fftw-3.3.3/rdft/scalar/r2cb/r2cbIII_2.c0000644000175000017500000000621712054410774014225 00000000000000/* * Copyright (c) 2003, 2007-11 Matteo Frigo * Copyright (c) 2003, 2007-11 Massachusetts Institute of Technology * * This program is free software; you can redistribute it and/or modify * it under the terms of the GNU General Public License as published by * the Free Software Foundation; either version 2 of the License, or * (at your option) any later version. * * This program is distributed in the hope that it will be useful, * but WITHOUT ANY WARRANTY; without even the implied warranty of * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the * GNU General Public License for more details. * * You should have received a copy of the GNU General Public License * along with this program; if not, write to the Free Software * Foundation, Inc., 51 Franklin Street, Fifth Floor, Boston, MA 02110-1301 USA * */ /* This file was automatically generated --- DO NOT EDIT */ /* Generated on Sun Nov 25 07:41:32 EST 2012 */ #include "codelet-rdft.h" #ifdef HAVE_FMA /* Generated by: ../../../genfft/gen_r2cb.native -fma -reorder-insns -schedule-for-pipeline -compact -variables 4 -pipeline-latency 4 -sign 1 -n 2 -name r2cbIII_2 -dft-III -include r2cbIII.h */ /* * This function contains 0 FP additions, 2 FP multiplications, * (or, 0 additions, 2 multiplications, 0 fused multiply/add), * 4 stack variables, 1 constants, and 4 memory accesses */ #include "r2cbIII.h" static void r2cbIII_2(R *R0, R *R1, R *Cr, R *Ci, stride rs, stride csr, stride csi, INT v, INT ivs, INT ovs) { DK(KP2_000000000, +2.000000000000000000000000000000000000000000000); { INT i; for (i = v; i > 0; i = i - 1, R0 = R0 + ovs, R1 = R1 + ovs, Cr = Cr + ivs, Ci = Ci + ivs, MAKE_VOLATILE_STRIDE(8, rs), MAKE_VOLATILE_STRIDE(8, csr), MAKE_VOLATILE_STRIDE(8, csi)) { E T1, T2; T1 = Cr[0]; T2 = Ci[0]; R0[0] = KP2_000000000 * T1; R1[0] = -(KP2_000000000 * T2); } } } static const kr2c_desc desc = { 2, "r2cbIII_2", {0, 2, 0, 0}, &GENUS }; void X(codelet_r2cbIII_2) (planner *p) { X(kr2c_register) (p, r2cbIII_2, &desc); } #else /* HAVE_FMA */ /* Generated by: ../../../genfft/gen_r2cb.native -compact -variables 4 -pipeline-latency 4 -sign 1 -n 2 -name r2cbIII_2 -dft-III -include r2cbIII.h */ /* * This function contains 0 FP additions, 2 FP multiplications, * (or, 0 additions, 2 multiplications, 0 fused multiply/add), * 4 stack variables, 1 constants, and 4 memory accesses */ #include "r2cbIII.h" static void r2cbIII_2(R *R0, R *R1, R *Cr, R *Ci, stride rs, stride csr, stride csi, INT v, INT ivs, INT ovs) { DK(KP2_000000000, +2.000000000000000000000000000000000000000000000); { INT i; for (i = v; i > 0; i = i - 1, R0 = R0 + ovs, R1 = R1 + ovs, Cr = Cr + ivs, Ci = Ci + ivs, MAKE_VOLATILE_STRIDE(8, rs), MAKE_VOLATILE_STRIDE(8, csr), MAKE_VOLATILE_STRIDE(8, csi)) { E T1, T2; T1 = Cr[0]; T2 = Ci[0]; R0[0] = KP2_000000000 * T1; R1[0] = -(KP2_000000000 * T2); } } } static const kr2c_desc desc = { 2, "r2cbIII_2", {0, 2, 0, 0}, &GENUS }; void X(codelet_r2cbIII_2) (planner *p) { X(kr2c_register) (p, r2cbIII_2, &desc); } #endif /* HAVE_FMA */ fftw-3.3.3/rdft/scalar/r2cb/r2cbIII_3.c0000644000175000017500000000704712054410775014231 00000000000000/* * Copyright (c) 2003, 2007-11 Matteo Frigo * Copyright (c) 2003, 2007-11 Massachusetts Institute of Technology * * This program is free software; you can redistribute it and/or modify * it under the terms of the GNU General Public License as published by * the Free Software Foundation; either version 2 of the License, or * (at your option) any later version. * * This program is distributed in the hope that it will be useful, * but WITHOUT ANY WARRANTY; without even the implied warranty of * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the * GNU General Public License for more details. * * You should have received a copy of the GNU General Public License * along with this program; if not, write to the Free Software * Foundation, Inc., 51 Franklin Street, Fifth Floor, Boston, MA 02110-1301 USA * */ /* This file was automatically generated --- DO NOT EDIT */ /* Generated on Sun Nov 25 07:41:32 EST 2012 */ #include "codelet-rdft.h" #ifdef HAVE_FMA /* Generated by: ../../../genfft/gen_r2cb.native -fma -reorder-insns -schedule-for-pipeline -compact -variables 4 -pipeline-latency 4 -sign 1 -n 3 -name r2cbIII_3 -dft-III -include r2cbIII.h */ /* * This function contains 4 FP additions, 3 FP multiplications, * (or, 1 additions, 0 multiplications, 3 fused multiply/add), * 7 stack variables, 2 constants, and 6 memory accesses */ #include "r2cbIII.h" static void r2cbIII_3(R *R0, R *R1, R *Cr, R *Ci, stride rs, stride csr, stride csi, INT v, INT ivs, INT ovs) { DK(KP1_732050807, +1.732050807568877293527446341505872366942805254); DK(KP2_000000000, +2.000000000000000000000000000000000000000000000); { INT i; for (i = v; i > 0; i = i - 1, R0 = R0 + ovs, R1 = R1 + ovs, Cr = Cr + ivs, Ci = Ci + ivs, MAKE_VOLATILE_STRIDE(12, rs), MAKE_VOLATILE_STRIDE(12, csr), MAKE_VOLATILE_STRIDE(12, csi)) { E T4, T1, T2, T3; T4 = Ci[0]; T1 = Cr[WS(csr, 1)]; T2 = Cr[0]; R0[0] = FMA(KP2_000000000, T2, T1); T3 = T2 - T1; R1[0] = FNMS(KP1_732050807, T4, T3); R0[WS(rs, 1)] = -(FMA(KP1_732050807, T4, T3)); } } } static const kr2c_desc desc = { 3, "r2cbIII_3", {1, 0, 3, 0}, &GENUS }; void X(codelet_r2cbIII_3) (planner *p) { X(kr2c_register) (p, r2cbIII_3, &desc); } #else /* HAVE_FMA */ /* Generated by: ../../../genfft/gen_r2cb.native -compact -variables 4 -pipeline-latency 4 -sign 1 -n 3 -name r2cbIII_3 -dft-III -include r2cbIII.h */ /* * This function contains 4 FP additions, 2 FP multiplications, * (or, 3 additions, 1 multiplications, 1 fused multiply/add), * 8 stack variables, 2 constants, and 6 memory accesses */ #include "r2cbIII.h" static void r2cbIII_3(R *R0, R *R1, R *Cr, R *Ci, stride rs, stride csr, stride csi, INT v, INT ivs, INT ovs) { DK(KP2_000000000, +2.000000000000000000000000000000000000000000000); DK(KP1_732050807, +1.732050807568877293527446341505872366942805254); { INT i; for (i = v; i > 0; i = i - 1, R0 = R0 + ovs, R1 = R1 + ovs, Cr = Cr + ivs, Ci = Ci + ivs, MAKE_VOLATILE_STRIDE(12, rs), MAKE_VOLATILE_STRIDE(12, csr), MAKE_VOLATILE_STRIDE(12, csi)) { E T5, T1, T2, T3, T4; T4 = Ci[0]; T5 = KP1_732050807 * T4; T1 = Cr[WS(csr, 1)]; T2 = Cr[0]; T3 = T2 - T1; R0[0] = FMA(KP2_000000000, T2, T1); R0[WS(rs, 1)] = -(T3 + T5); R1[0] = T3 - T5; } } } static const kr2c_desc desc = { 3, "r2cbIII_3", {3, 1, 1, 0}, &GENUS }; void X(codelet_r2cbIII_3) (planner *p) { X(kr2c_register) (p, r2cbIII_3, &desc); } #endif /* HAVE_FMA */ fftw-3.3.3/rdft/scalar/r2cb/r2cbIII_4.c0000644000175000017500000000737512054410775014236 00000000000000/* * Copyright (c) 2003, 2007-11 Matteo Frigo * Copyright (c) 2003, 2007-11 Massachusetts Institute of Technology * * This program is free software; you can redistribute it and/or modify * it under the terms of the GNU General Public License as published by * the Free Software Foundation; either version 2 of the License, or * (at your option) any later version. * * This program is distributed in the hope that it will be useful, * but WITHOUT ANY WARRANTY; without even the implied warranty of * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the * GNU General Public License for more details. * * You should have received a copy of the GNU General Public License * along with this program; if not, write to the Free Software * Foundation, Inc., 51 Franklin Street, Fifth Floor, Boston, MA 02110-1301 USA * */ /* This file was automatically generated --- DO NOT EDIT */ /* Generated on Sun Nov 25 07:41:33 EST 2012 */ #include "codelet-rdft.h" #ifdef HAVE_FMA /* Generated by: ../../../genfft/gen_r2cb.native -fma -reorder-insns -schedule-for-pipeline -compact -variables 4 -pipeline-latency 4 -sign 1 -n 4 -name r2cbIII_4 -dft-III -include r2cbIII.h */ /* * This function contains 6 FP additions, 4 FP multiplications, * (or, 6 additions, 4 multiplications, 0 fused multiply/add), * 9 stack variables, 2 constants, and 8 memory accesses */ #include "r2cbIII.h" static void r2cbIII_4(R *R0, R *R1, R *Cr, R *Ci, stride rs, stride csr, stride csi, INT v, INT ivs, INT ovs) { DK(KP1_414213562, +1.414213562373095048801688724209698078569671875); DK(KP2_000000000, +2.000000000000000000000000000000000000000000000); { INT i; for (i = v; i > 0; i = i - 1, R0 = R0 + ovs, R1 = R1 + ovs, Cr = Cr + ivs, Ci = Ci + ivs, MAKE_VOLATILE_STRIDE(16, rs), MAKE_VOLATILE_STRIDE(16, csr), MAKE_VOLATILE_STRIDE(16, csi)) { E T1, T2, T4, T5, T3, T6; T1 = Cr[0]; T2 = Cr[WS(csr, 1)]; T4 = Ci[0]; T5 = Ci[WS(csi, 1)]; R0[0] = KP2_000000000 * (T1 + T2); T3 = T1 - T2; R0[WS(rs, 1)] = KP2_000000000 * (T5 - T4); T6 = T4 + T5; R1[WS(rs, 1)] = -(KP1_414213562 * (T3 + T6)); R1[0] = KP1_414213562 * (T3 - T6); } } } static const kr2c_desc desc = { 4, "r2cbIII_4", {6, 4, 0, 0}, &GENUS }; void X(codelet_r2cbIII_4) (planner *p) { X(kr2c_register) (p, r2cbIII_4, &desc); } #else /* HAVE_FMA */ /* Generated by: ../../../genfft/gen_r2cb.native -compact -variables 4 -pipeline-latency 4 -sign 1 -n 4 -name r2cbIII_4 -dft-III -include r2cbIII.h */ /* * This function contains 6 FP additions, 4 FP multiplications, * (or, 6 additions, 4 multiplications, 0 fused multiply/add), * 9 stack variables, 2 constants, and 8 memory accesses */ #include "r2cbIII.h" static void r2cbIII_4(R *R0, R *R1, R *Cr, R *Ci, stride rs, stride csr, stride csi, INT v, INT ivs, INT ovs) { DK(KP1_414213562, +1.414213562373095048801688724209698078569671875); DK(KP2_000000000, +2.000000000000000000000000000000000000000000000); { INT i; for (i = v; i > 0; i = i - 1, R0 = R0 + ovs, R1 = R1 + ovs, Cr = Cr + ivs, Ci = Ci + ivs, MAKE_VOLATILE_STRIDE(16, rs), MAKE_VOLATILE_STRIDE(16, csr), MAKE_VOLATILE_STRIDE(16, csi)) { E T1, T2, T3, T4, T5, T6; T1 = Cr[0]; T2 = Cr[WS(csr, 1)]; T3 = T1 - T2; T4 = Ci[0]; T5 = Ci[WS(csi, 1)]; T6 = T4 + T5; R0[0] = KP2_000000000 * (T1 + T2); R0[WS(rs, 1)] = KP2_000000000 * (T5 - T4); R1[0] = KP1_414213562 * (T3 - T6); R1[WS(rs, 1)] = -(KP1_414213562 * (T3 + T6)); } } } static const kr2c_desc desc = { 4, "r2cbIII_4", {6, 4, 0, 0}, &GENUS }; void X(codelet_r2cbIII_4) (planner *p) { X(kr2c_register) (p, r2cbIII_4, &desc); } #endif /* HAVE_FMA */ fftw-3.3.3/rdft/scalar/r2cb/r2cbIII_5.c0000644000175000017500000001137412054410775014231 00000000000000/* * Copyright (c) 2003, 2007-11 Matteo Frigo * Copyright (c) 2003, 2007-11 Massachusetts Institute of Technology * * This program is free software; you can redistribute it and/or modify * it under the terms of the GNU General Public License as published by * the Free Software Foundation; either version 2 of the License, or * (at your option) any later version. * * This program is distributed in the hope that it will be useful, * but WITHOUT ANY WARRANTY; without even the implied warranty of * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the * GNU General Public License for more details. * * You should have received a copy of the GNU General Public License * along with this program; if not, write to the Free Software * Foundation, Inc., 51 Franklin Street, Fifth Floor, Boston, MA 02110-1301 USA * */ /* This file was automatically generated --- DO NOT EDIT */ /* Generated on Sun Nov 25 07:41:33 EST 2012 */ #include "codelet-rdft.h" #ifdef HAVE_FMA /* Generated by: ../../../genfft/gen_r2cb.native -fma -reorder-insns -schedule-for-pipeline -compact -variables 4 -pipeline-latency 4 -sign 1 -n 5 -name r2cbIII_5 -dft-III -include r2cbIII.h */ /* * This function contains 12 FP additions, 10 FP multiplications, * (or, 2 additions, 0 multiplications, 10 fused multiply/add), * 18 stack variables, 5 constants, and 10 memory accesses */ #include "r2cbIII.h" static void r2cbIII_5(R *R0, R *R1, R *Cr, R *Ci, stride rs, stride csr, stride csi, INT v, INT ivs, INT ovs) { DK(KP1_902113032, +1.902113032590307144232878666758764286811397268); DK(KP1_118033988, +1.118033988749894848204586834365638117720309180); DK(KP500000000, +0.500000000000000000000000000000000000000000000); DK(KP2_000000000, +2.000000000000000000000000000000000000000000000); DK(KP618033988, +0.618033988749894848204586834365638117720309180); { INT i; for (i = v; i > 0; i = i - 1, R0 = R0 + ovs, R1 = R1 + ovs, Cr = Cr + ivs, Ci = Ci + ivs, MAKE_VOLATILE_STRIDE(20, rs), MAKE_VOLATILE_STRIDE(20, csr), MAKE_VOLATILE_STRIDE(20, csi)) { E T1, T2, T3, Tc, Ta, T8, T9; T8 = Ci[WS(csi, 1)]; T9 = Ci[0]; T1 = Cr[WS(csr, 2)]; T2 = Cr[WS(csr, 1)]; T3 = Cr[0]; Tc = FMS(KP618033988, T8, T9); Ta = FMA(KP618033988, T9, T8); { E T6, T4, T5, T7, Tb; T6 = T3 - T2; T4 = T2 + T3; R0[0] = FMA(KP2_000000000, T4, T1); T5 = FNMS(KP500000000, T4, T1); T7 = FNMS(KP1_118033988, T6, T5); Tb = FMA(KP1_118033988, T6, T5); R0[WS(rs, 2)] = FNMS(KP1_902113032, Ta, T7); R1[0] = -(FMA(KP1_902113032, Ta, T7)); R1[WS(rs, 1)] = FMS(KP1_902113032, Tc, Tb); R0[WS(rs, 1)] = FMA(KP1_902113032, Tc, Tb); } } } } static const kr2c_desc desc = { 5, "r2cbIII_5", {2, 0, 10, 0}, &GENUS }; void X(codelet_r2cbIII_5) (planner *p) { X(kr2c_register) (p, r2cbIII_5, &desc); } #else /* HAVE_FMA */ /* Generated by: ../../../genfft/gen_r2cb.native -compact -variables 4 -pipeline-latency 4 -sign 1 -n 5 -name r2cbIII_5 -dft-III -include r2cbIII.h */ /* * This function contains 12 FP additions, 7 FP multiplications, * (or, 8 additions, 3 multiplications, 4 fused multiply/add), * 18 stack variables, 5 constants, and 10 memory accesses */ #include "r2cbIII.h" static void r2cbIII_5(R *R0, R *R1, R *Cr, R *Ci, stride rs, stride csr, stride csi, INT v, INT ivs, INT ovs) { DK(KP2_000000000, +2.000000000000000000000000000000000000000000000); DK(KP1_118033988, +1.118033988749894848204586834365638117720309180); DK(KP500000000, +0.500000000000000000000000000000000000000000000); DK(KP1_175570504, +1.175570504584946258337411909278145537195304875); DK(KP1_902113032, +1.902113032590307144232878666758764286811397268); { INT i; for (i = v; i > 0; i = i - 1, R0 = R0 + ovs, R1 = R1 + ovs, Cr = Cr + ivs, Ci = Ci + ivs, MAKE_VOLATILE_STRIDE(20, rs), MAKE_VOLATILE_STRIDE(20, csr), MAKE_VOLATILE_STRIDE(20, csi)) { E Ta, Tc, T1, T4, T5, T6, Tb, T7; { E T8, T9, T2, T3; T8 = Ci[WS(csi, 1)]; T9 = Ci[0]; Ta = FMA(KP1_902113032, T8, KP1_175570504 * T9); Tc = FNMS(KP1_902113032, T9, KP1_175570504 * T8); T1 = Cr[WS(csr, 2)]; T2 = Cr[WS(csr, 1)]; T3 = Cr[0]; T4 = T2 + T3; T5 = FMS(KP500000000, T4, T1); T6 = KP1_118033988 * (T3 - T2); } R0[0] = FMA(KP2_000000000, T4, T1); Tb = T6 - T5; R0[WS(rs, 1)] = Tb + Tc; R1[WS(rs, 1)] = Tc - Tb; T7 = T5 + T6; R1[0] = T7 - Ta; R0[WS(rs, 2)] = -(T7 + Ta); } } } static const kr2c_desc desc = { 5, "r2cbIII_5", {8, 3, 4, 0}, &GENUS }; void X(codelet_r2cbIII_5) (planner *p) { X(kr2c_register) (p, r2cbIII_5, &desc); } #endif /* HAVE_FMA */ fftw-3.3.3/rdft/scalar/r2cb/r2cbIII_6.c0000644000175000017500000001054212054410775014226 00000000000000/* * Copyright (c) 2003, 2007-11 Matteo Frigo * Copyright (c) 2003, 2007-11 Massachusetts Institute of Technology * * This program is free software; you can redistribute it and/or modify * it under the terms of the GNU General Public License as published by * the Free Software Foundation; either version 2 of the License, or * (at your option) any later version. * * This program is distributed in the hope that it will be useful, * but WITHOUT ANY WARRANTY; without even the implied warranty of * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the * GNU General Public License for more details. * * You should have received a copy of the GNU General Public License * along with this program; if not, write to the Free Software * Foundation, Inc., 51 Franklin Street, Fifth Floor, Boston, MA 02110-1301 USA * */ /* This file was automatically generated --- DO NOT EDIT */ /* Generated on Sun Nov 25 07:41:33 EST 2012 */ #include "codelet-rdft.h" #ifdef HAVE_FMA /* Generated by: ../../../genfft/gen_r2cb.native -fma -reorder-insns -schedule-for-pipeline -compact -variables 4 -pipeline-latency 4 -sign 1 -n 6 -name r2cbIII_6 -dft-III -include r2cbIII.h */ /* * This function contains 12 FP additions, 8 FP multiplications, * (or, 6 additions, 2 multiplications, 6 fused multiply/add), * 15 stack variables, 2 constants, and 12 memory accesses */ #include "r2cbIII.h" static void r2cbIII_6(R *R0, R *R1, R *Cr, R *Ci, stride rs, stride csr, stride csi, INT v, INT ivs, INT ovs) { DK(KP1_732050807, +1.732050807568877293527446341505872366942805254); DK(KP2_000000000, +2.000000000000000000000000000000000000000000000); { INT i; for (i = v; i > 0; i = i - 1, R0 = R0 + ovs, R1 = R1 + ovs, Cr = Cr + ivs, Ci = Ci + ivs, MAKE_VOLATILE_STRIDE(24, rs), MAKE_VOLATILE_STRIDE(24, csr), MAKE_VOLATILE_STRIDE(24, csi)) { E T1, T8, T2, T3, T5, T6; T1 = Cr[WS(csr, 1)]; T8 = Ci[WS(csi, 1)]; T2 = Cr[WS(csr, 2)]; T3 = Cr[0]; T5 = Ci[WS(csi, 2)]; T6 = Ci[0]; { E T4, Ta, T7, Tc, Tb, T9; T4 = T2 + T3; Ta = T2 - T3; T7 = T5 + T6; Tc = T5 - T6; Tb = FNMS(KP2_000000000, T1, T4); R0[0] = KP2_000000000 * (T1 + T4); T9 = FMA(KP2_000000000, T8, T7); R1[WS(rs, 1)] = KP2_000000000 * (T8 - T7); R0[WS(rs, 2)] = FMS(KP1_732050807, Tc, Tb); R0[WS(rs, 1)] = FMA(KP1_732050807, Tc, Tb); R1[WS(rs, 2)] = FMS(KP1_732050807, Ta, T9); R1[0] = -(FMA(KP1_732050807, Ta, T9)); } } } } static const kr2c_desc desc = { 6, "r2cbIII_6", {6, 2, 6, 0}, &GENUS }; void X(codelet_r2cbIII_6) (planner *p) { X(kr2c_register) (p, r2cbIII_6, &desc); } #else /* HAVE_FMA */ /* Generated by: ../../../genfft/gen_r2cb.native -compact -variables 4 -pipeline-latency 4 -sign 1 -n 6 -name r2cbIII_6 -dft-III -include r2cbIII.h */ /* * This function contains 12 FP additions, 6 FP multiplications, * (or, 10 additions, 4 multiplications, 2 fused multiply/add), * 15 stack variables, 2 constants, and 12 memory accesses */ #include "r2cbIII.h" static void r2cbIII_6(R *R0, R *R1, R *Cr, R *Ci, stride rs, stride csr, stride csi, INT v, INT ivs, INT ovs) { DK(KP2_000000000, +2.000000000000000000000000000000000000000000000); DK(KP1_732050807, +1.732050807568877293527446341505872366942805254); { INT i; for (i = v; i > 0; i = i - 1, R0 = R0 + ovs, R1 = R1 + ovs, Cr = Cr + ivs, Ci = Ci + ivs, MAKE_VOLATILE_STRIDE(24, rs), MAKE_VOLATILE_STRIDE(24, csr), MAKE_VOLATILE_STRIDE(24, csi)) { E T1, T6, T4, T5, T9, Tb, Ta, Tc; T1 = Cr[WS(csr, 1)]; T6 = Ci[WS(csi, 1)]; { E T2, T3, T7, T8; T2 = Cr[WS(csr, 2)]; T3 = Cr[0]; T4 = T2 + T3; T5 = KP1_732050807 * (T2 - T3); T7 = Ci[WS(csi, 2)]; T8 = Ci[0]; T9 = T7 + T8; Tb = KP1_732050807 * (T7 - T8); } R0[0] = KP2_000000000 * (T1 + T4); R1[WS(rs, 1)] = KP2_000000000 * (T6 - T9); Ta = FMA(KP2_000000000, T6, T9); R1[0] = -(T5 + Ta); R1[WS(rs, 2)] = T5 - Ta; Tc = FMS(KP2_000000000, T1, T4); R0[WS(rs, 1)] = Tb - Tc; R0[WS(rs, 2)] = Tc + Tb; } } } static const kr2c_desc desc = { 6, "r2cbIII_6", {10, 4, 2, 0}, &GENUS }; void X(codelet_r2cbIII_6) (planner *p) { X(kr2c_register) (p, r2cbIII_6, &desc); } #endif /* HAVE_FMA */ fftw-3.3.3/rdft/scalar/r2cb/r2cbIII_7.c0000644000175000017500000001367712054410776014244 00000000000000/* * Copyright (c) 2003, 2007-11 Matteo Frigo * Copyright (c) 2003, 2007-11 Massachusetts Institute of Technology * * This program is free software; you can redistribute it and/or modify * it under the terms of the GNU General Public License as published by * the Free Software Foundation; either version 2 of the License, or * (at your option) any later version. * * This program is distributed in the hope that it will be useful, * but WITHOUT ANY WARRANTY; without even the implied warranty of * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the * GNU General Public License for more details. * * You should have received a copy of the GNU General Public License * along with this program; if not, write to the Free Software * Foundation, Inc., 51 Franklin Street, Fifth Floor, Boston, MA 02110-1301 USA * */ /* This file was automatically generated --- DO NOT EDIT */ /* Generated on Sun Nov 25 07:41:33 EST 2012 */ #include "codelet-rdft.h" #ifdef HAVE_FMA /* Generated by: ../../../genfft/gen_r2cb.native -fma -reorder-insns -schedule-for-pipeline -compact -variables 4 -pipeline-latency 4 -sign 1 -n 7 -name r2cbIII_7 -dft-III -include r2cbIII.h */ /* * This function contains 24 FP additions, 22 FP multiplications, * (or, 2 additions, 0 multiplications, 22 fused multiply/add), * 31 stack variables, 7 constants, and 14 memory accesses */ #include "r2cbIII.h" static void r2cbIII_7(R *R0, R *R1, R *Cr, R *Ci, stride rs, stride csr, stride csi, INT v, INT ivs, INT ovs) { DK(KP1_949855824, +1.949855824363647214036263365987862434465571601); DK(KP1_801937735, +1.801937735804838252472204639014890102331838324); DK(KP2_000000000, +2.000000000000000000000000000000000000000000000); DK(KP692021471, +0.692021471630095869627814897002069140197260599); DK(KP801937735, +0.801937735804838252472204639014890102331838324); DK(KP356895867, +0.356895867892209443894399510021300583399127187); DK(KP554958132, +0.554958132087371191422194871006410481067288862); { INT i; for (i = v; i > 0; i = i - 1, R0 = R0 + ovs, R1 = R1 + ovs, Cr = Cr + ivs, Ci = Ci + ivs, MAKE_VOLATILE_STRIDE(28, rs), MAKE_VOLATILE_STRIDE(28, csr), MAKE_VOLATILE_STRIDE(28, csi)) { E Tn, Td, Tg, Ti, Tl, T8; { E T1, T9, Tb, Ta, T2, T4, Th, Tm, Tc, T3, Te; T1 = Cr[WS(csr, 3)]; T9 = Ci[WS(csi, 1)]; Tb = Ci[0]; Ta = Ci[WS(csi, 2)]; T2 = Cr[WS(csr, 2)]; T4 = Cr[0]; Th = FMA(KP554958132, T9, Tb); Tm = FNMS(KP554958132, Ta, T9); Tc = FMA(KP554958132, Tb, Ta); T3 = Cr[WS(csr, 1)]; Te = FNMS(KP356895867, T2, T4); Tn = FNMS(KP801937735, Tm, Tb); { E Tf, Tk, T7, T5, Tj, T6; Td = FMA(KP801937735, Tc, T9); T5 = T2 + T3 + T4; Tj = FNMS(KP356895867, T4, T3); T6 = FNMS(KP356895867, T3, T2); Tf = FNMS(KP692021471, Te, T3); R0[0] = FMA(KP2_000000000, T5, T1); Tk = FNMS(KP692021471, Tj, T2); T7 = FNMS(KP692021471, T6, T4); Tg = FNMS(KP1_801937735, Tf, T1); Ti = FNMS(KP801937735, Th, Ta); Tl = FNMS(KP1_801937735, Tk, T1); T8 = FNMS(KP1_801937735, T7, T1); } } R1[WS(rs, 2)] = FMS(KP1_949855824, Ti, Tg); R0[WS(rs, 1)] = FMA(KP1_949855824, Ti, Tg); R0[WS(rs, 2)] = FNMS(KP1_949855824, Tn, Tl); R1[WS(rs, 1)] = -(FMA(KP1_949855824, Tn, Tl)); R0[WS(rs, 3)] = FNMS(KP1_949855824, Td, T8); R1[0] = -(FMA(KP1_949855824, Td, T8)); } } } static const kr2c_desc desc = { 7, "r2cbIII_7", {2, 0, 22, 0}, &GENUS }; void X(codelet_r2cbIII_7) (planner *p) { X(kr2c_register) (p, r2cbIII_7, &desc); } #else /* HAVE_FMA */ /* Generated by: ../../../genfft/gen_r2cb.native -compact -variables 4 -pipeline-latency 4 -sign 1 -n 7 -name r2cbIII_7 -dft-III -include r2cbIII.h */ /* * This function contains 24 FP additions, 19 FP multiplications, * (or, 9 additions, 4 multiplications, 15 fused multiply/add), * 21 stack variables, 7 constants, and 14 memory accesses */ #include "r2cbIII.h" static void r2cbIII_7(R *R0, R *R1, R *Cr, R *Ci, stride rs, stride csr, stride csi, INT v, INT ivs, INT ovs) { DK(KP2_000000000, +2.000000000000000000000000000000000000000000000); DK(KP1_246979603, +1.246979603717467061050009768008479621264549462); DK(KP1_801937735, +1.801937735804838252472204639014890102331838324); DK(KP445041867, +0.445041867912628808577805128993589518932711138); DK(KP867767478, +0.867767478235116240951536665696717509219981456); DK(KP1_949855824, +1.949855824363647214036263365987862434465571601); DK(KP1_563662964, +1.563662964936059617416889053348115500464669037); { INT i; for (i = v; i > 0; i = i - 1, R0 = R0 + ovs, R1 = R1 + ovs, Cr = Cr + ivs, Ci = Ci + ivs, MAKE_VOLATILE_STRIDE(28, rs), MAKE_VOLATILE_STRIDE(28, csr), MAKE_VOLATILE_STRIDE(28, csi)) { E T9, Td, Tb, T1, T4, T2, T3, T5, Tc, Ta, T6, T8, T7; T6 = Ci[WS(csi, 2)]; T8 = Ci[0]; T7 = Ci[WS(csi, 1)]; T9 = FMA(KP1_563662964, T6, KP1_949855824 * T7) + (KP867767478 * T8); Td = FNMS(KP1_949855824, T8, KP1_563662964 * T7) - (KP867767478 * T6); Tb = FNMS(KP1_563662964, T8, KP1_949855824 * T6) - (KP867767478 * T7); T1 = Cr[WS(csr, 3)]; T4 = Cr[0]; T2 = Cr[WS(csr, 2)]; T3 = Cr[WS(csr, 1)]; T5 = FMA(KP445041867, T3, KP1_801937735 * T4) + FNMA(KP1_246979603, T2, T1); Tc = FMA(KP1_801937735, T2, KP445041867 * T4) + FNMA(KP1_246979603, T3, T1); Ta = FMA(KP1_246979603, T4, T1) + FNMA(KP1_801937735, T3, KP445041867 * T2); R1[0] = T5 - T9; R0[WS(rs, 3)] = -(T5 + T9); R0[WS(rs, 2)] = Td - Tc; R1[WS(rs, 1)] = Tc + Td; R1[WS(rs, 2)] = Tb - Ta; R0[WS(rs, 1)] = Ta + Tb; R0[0] = FMA(KP2_000000000, T2 + T3 + T4, T1); } } } static const kr2c_desc desc = { 7, "r2cbIII_7", {9, 4, 15, 0}, &GENUS }; void X(codelet_r2cbIII_7) (planner *p) { X(kr2c_register) (p, r2cbIII_7, &desc); } #endif /* HAVE_FMA */ fftw-3.3.3/rdft/scalar/r2cb/r2cbIII_8.c0000644000175000017500000001274612054410777014242 00000000000000/* * Copyright (c) 2003, 2007-11 Matteo Frigo * Copyright (c) 2003, 2007-11 Massachusetts Institute of Technology * * This program is free software; you can redistribute it and/or modify * it under the terms of the GNU General Public License as published by * the Free Software Foundation; either version 2 of the License, or * (at your option) any later version. * * This program is distributed in the hope that it will be useful, * but WITHOUT ANY WARRANTY; without even the implied warranty of * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the * GNU General Public License for more details. * * You should have received a copy of the GNU General Public License * along with this program; if not, write to the Free Software * Foundation, Inc., 51 Franklin Street, Fifth Floor, Boston, MA 02110-1301 USA * */ /* This file was automatically generated --- DO NOT EDIT */ /* Generated on Sun Nov 25 07:41:34 EST 2012 */ #include "codelet-rdft.h" #ifdef HAVE_FMA /* Generated by: ../../../genfft/gen_r2cb.native -fma -reorder-insns -schedule-for-pipeline -compact -variables 4 -pipeline-latency 4 -sign 1 -n 8 -name r2cbIII_8 -dft-III -include r2cbIII.h */ /* * This function contains 22 FP additions, 12 FP multiplications, * (or, 18 additions, 8 multiplications, 4 fused multiply/add), * 23 stack variables, 4 constants, and 16 memory accesses */ #include "r2cbIII.h" static void r2cbIII_8(R *R0, R *R1, R *Cr, R *Ci, stride rs, stride csr, stride csi, INT v, INT ivs, INT ovs) { DK(KP414213562, +0.414213562373095048801688724209698078569671875); DK(KP1_847759065, +1.847759065022573512256366378793576573644833252); DK(KP1_414213562, +1.414213562373095048801688724209698078569671875); DK(KP2_000000000, +2.000000000000000000000000000000000000000000000); { INT i; for (i = v; i > 0; i = i - 1, R0 = R0 + ovs, R1 = R1 + ovs, Cr = Cr + ivs, Ci = Ci + ivs, MAKE_VOLATILE_STRIDE(32, rs), MAKE_VOLATILE_STRIDE(32, csr), MAKE_VOLATILE_STRIDE(32, csi)) { E T4, T7, T3, Tl, Tf, T5, T8, T9, T6, Tc; { E T1, T2, Td, Te; T1 = Cr[0]; T2 = Cr[WS(csr, 3)]; Td = Ci[0]; Te = Ci[WS(csi, 3)]; T4 = Cr[WS(csr, 2)]; T7 = T1 - T2; T3 = T1 + T2; Tl = Te - Td; Tf = Td + Te; T5 = Cr[WS(csr, 1)]; T8 = Ci[WS(csi, 2)]; T9 = Ci[WS(csi, 1)]; } T6 = T4 + T5; Tc = T4 - T5; { E Ta, Tk, Tg, Th; Ta = T8 + T9; Tk = T8 - T9; Tg = Tc + Tf; Th = Tc - Tf; { E Tj, Tm, Tb, Ti; Tj = T3 - T6; R0[0] = KP2_000000000 * (T3 + T6); Tm = Tk + Tl; R0[WS(rs, 2)] = KP2_000000000 * (Tl - Tk); Tb = T7 - Ta; Ti = T7 + Ta; R0[WS(rs, 3)] = KP1_414213562 * (Tm - Tj); R0[WS(rs, 1)] = KP1_414213562 * (Tj + Tm); R1[WS(rs, 3)] = -(KP1_847759065 * (FNMS(KP414213562, Th, Ti))); R1[WS(rs, 1)] = KP1_847759065 * (FMA(KP414213562, Ti, Th)); R1[WS(rs, 2)] = -(KP1_847759065 * (FMA(KP414213562, Tb, Tg))); R1[0] = KP1_847759065 * (FNMS(KP414213562, Tg, Tb)); } } } } } static const kr2c_desc desc = { 8, "r2cbIII_8", {18, 8, 4, 0}, &GENUS }; void X(codelet_r2cbIII_8) (planner *p) { X(kr2c_register) (p, r2cbIII_8, &desc); } #else /* HAVE_FMA */ /* Generated by: ../../../genfft/gen_r2cb.native -compact -variables 4 -pipeline-latency 4 -sign 1 -n 8 -name r2cbIII_8 -dft-III -include r2cbIII.h */ /* * This function contains 22 FP additions, 12 FP multiplications, * (or, 18 additions, 8 multiplications, 4 fused multiply/add), * 19 stack variables, 4 constants, and 16 memory accesses */ #include "r2cbIII.h" static void r2cbIII_8(R *R0, R *R1, R *Cr, R *Ci, stride rs, stride csr, stride csi, INT v, INT ivs, INT ovs) { DK(KP1_414213562, +1.414213562373095048801688724209698078569671875); DK(KP765366864, +0.765366864730179543456919968060797733522689125); DK(KP1_847759065, +1.847759065022573512256366378793576573644833252); DK(KP2_000000000, +2.000000000000000000000000000000000000000000000); { INT i; for (i = v; i > 0; i = i - 1, R0 = R0 + ovs, R1 = R1 + ovs, Cr = Cr + ivs, Ci = Ci + ivs, MAKE_VOLATILE_STRIDE(32, rs), MAKE_VOLATILE_STRIDE(32, csr), MAKE_VOLATILE_STRIDE(32, csi)) { E T3, T7, Tf, Tl, T6, Tc, Ta, Tk, Tb, Tg; { E T1, T2, Td, Te; T1 = Cr[0]; T2 = Cr[WS(csr, 3)]; T3 = T1 + T2; T7 = T1 - T2; Td = Ci[0]; Te = Ci[WS(csi, 3)]; Tf = Td + Te; Tl = Te - Td; } { E T4, T5, T8, T9; T4 = Cr[WS(csr, 2)]; T5 = Cr[WS(csr, 1)]; T6 = T4 + T5; Tc = T4 - T5; T8 = Ci[WS(csi, 2)]; T9 = Ci[WS(csi, 1)]; Ta = T8 + T9; Tk = T8 - T9; } R0[0] = KP2_000000000 * (T3 + T6); R0[WS(rs, 2)] = KP2_000000000 * (Tl - Tk); Tb = T7 - Ta; Tg = Tc + Tf; R1[0] = FNMS(KP765366864, Tg, KP1_847759065 * Tb); R1[WS(rs, 2)] = -(FMA(KP765366864, Tb, KP1_847759065 * Tg)); { E Th, Ti, Tj, Tm; Th = T7 + Ta; Ti = Tc - Tf; R1[WS(rs, 1)] = FMA(KP765366864, Th, KP1_847759065 * Ti); R1[WS(rs, 3)] = FNMS(KP1_847759065, Th, KP765366864 * Ti); Tj = T3 - T6; Tm = Tk + Tl; R0[WS(rs, 1)] = KP1_414213562 * (Tj + Tm); R0[WS(rs, 3)] = KP1_414213562 * (Tm - Tj); } } } } static const kr2c_desc desc = { 8, "r2cbIII_8", {18, 8, 4, 0}, &GENUS }; void X(codelet_r2cbIII_8) (planner *p) { X(kr2c_register) (p, r2cbIII_8, &desc); } #endif /* HAVE_FMA */ fftw-3.3.3/rdft/scalar/r2cb/r2cbIII_9.c0000644000175000017500000001710512054410777014235 00000000000000/* * Copyright (c) 2003, 2007-11 Matteo Frigo * Copyright (c) 2003, 2007-11 Massachusetts Institute of Technology * * This program is free software; you can redistribute it and/or modify * it under the terms of the GNU General Public License as published by * the Free Software Foundation; either version 2 of the License, or * (at your option) any later version. * * This program is distributed in the hope that it will be useful, * but WITHOUT ANY WARRANTY; without even the implied warranty of * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the * GNU General Public License for more details. * * You should have received a copy of the GNU General Public License * along with this program; if not, write to the Free Software * Foundation, Inc., 51 Franklin Street, Fifth Floor, Boston, MA 02110-1301 USA * */ /* This file was automatically generated --- DO NOT EDIT */ /* Generated on Sun Nov 25 07:41:35 EST 2012 */ #include "codelet-rdft.h" #ifdef HAVE_FMA /* Generated by: ../../../genfft/gen_r2cb.native -fma -reorder-insns -schedule-for-pipeline -compact -variables 4 -pipeline-latency 4 -sign 1 -n 9 -name r2cbIII_9 -dft-III -include r2cbIII.h */ /* * This function contains 32 FP additions, 24 FP multiplications, * (or, 8 additions, 0 multiplications, 24 fused multiply/add), * 40 stack variables, 12 constants, and 18 memory accesses */ #include "r2cbIII.h" static void r2cbIII_9(R *R0, R *R1, R *Cr, R *Ci, stride rs, stride csr, stride csi, INT v, INT ivs, INT ovs) { DK(KP1_326827896, +1.326827896337876792410842639271782594433726619); DK(KP1_705737063, +1.705737063904886419256501927880148143872040591); DK(KP766044443, +0.766044443118978035202392650555416673935832457); DK(KP1_532088886, +1.532088886237956070404785301110833347871664914); DK(KP984807753, +0.984807753012208059366743024589523013670643252); DK(KP1_969615506, +1.969615506024416118733486049179046027341286503); DK(KP839099631, +0.839099631177280011763127298123181364687434283); DK(KP176326980, +0.176326980708464973471090386868618986121633062); DK(KP866025403, +0.866025403784438646763723170752936183471402627); DK(KP500000000, +0.500000000000000000000000000000000000000000000); DK(KP1_732050807, +1.732050807568877293527446341505872366942805254); DK(KP2_000000000, +2.000000000000000000000000000000000000000000000); { INT i; for (i = v; i > 0; i = i - 1, R0 = R0 + ovs, R1 = R1 + ovs, Cr = Cr + ivs, Ci = Ci + ivs, MAKE_VOLATILE_STRIDE(36, rs), MAKE_VOLATILE_STRIDE(36, csr), MAKE_VOLATILE_STRIDE(36, csi)) { E T4, Td, T3, Th, Tr, Tm, T7, Tc, Tj, Tg, T1, T2; Tg = Ci[WS(csi, 1)]; T1 = Cr[WS(csr, 4)]; T2 = Cr[WS(csr, 1)]; T4 = Cr[WS(csr, 3)]; Td = Ci[WS(csi, 3)]; { E T5, Tf, T6, Ta, Tb; T5 = Cr[0]; Tf = T2 - T1; T3 = FMA(KP2_000000000, T2, T1); T6 = Cr[WS(csr, 2)]; Ta = Ci[WS(csi, 2)]; Tb = Ci[0]; Th = FNMS(KP1_732050807, Tg, Tf); Tr = FMA(KP1_732050807, Tg, Tf); Tm = T5 - T6; T7 = T5 + T6; Tc = Ta - Tb; Tj = Tb + Ta; } { E Tw, Tq, Tv, Tp, Ti, T8; Ti = FNMS(KP500000000, T7, T4); T8 = T4 + T7; { E Te, Tl, Tt, Tk, T9; Te = Tc - Td; Tl = FMA(KP500000000, Tc, Td); Tt = FNMS(KP866025403, Tj, Ti); Tk = FMA(KP866025403, Tj, Ti); T9 = T8 - T3; R0[0] = FMA(KP2_000000000, T8, T3); { E Ts, Tn, Tu, To; Ts = FMA(KP866025403, Tm, Tl); Tn = FNMS(KP866025403, Tm, Tl); R0[WS(rs, 3)] = FMS(KP1_732050807, Te, T9); R1[WS(rs, 1)] = FMA(KP1_732050807, Te, T9); Tu = FMA(KP176326980, Tt, Ts); Tw = FNMS(KP176326980, Ts, Tt); To = FMA(KP839099631, Tn, Tk); Tq = FNMS(KP839099631, Tk, Tn); R0[WS(rs, 1)] = FMS(KP1_969615506, Tu, Tr); Tv = FMA(KP984807753, Tu, Tr); R1[0] = FNMS(KP1_532088886, To, Th); Tp = FMA(KP766044443, To, Th); } } R0[WS(rs, 4)] = FMS(KP1_705737063, Tw, Tv); R1[WS(rs, 2)] = FMA(KP1_705737063, Tw, Tv); R0[WS(rs, 2)] = FMS(KP1_326827896, Tq, Tp); R1[WS(rs, 3)] = FMA(KP1_326827896, Tq, Tp); } } } } static const kr2c_desc desc = { 9, "r2cbIII_9", {8, 0, 24, 0}, &GENUS }; void X(codelet_r2cbIII_9) (planner *p) { X(kr2c_register) (p, r2cbIII_9, &desc); } #else /* HAVE_FMA */ /* Generated by: ../../../genfft/gen_r2cb.native -compact -variables 4 -pipeline-latency 4 -sign 1 -n 9 -name r2cbIII_9 -dft-III -include r2cbIII.h */ /* * This function contains 32 FP additions, 18 FP multiplications, * (or, 22 additions, 8 multiplications, 10 fused multiply/add), * 35 stack variables, 12 constants, and 18 memory accesses */ #include "r2cbIII.h" static void r2cbIII_9(R *R0, R *R1, R *Cr, R *Ci, stride rs, stride csr, stride csi, INT v, INT ivs, INT ovs) { DK(KP642787609, +0.642787609686539326322643409907263432907559884); DK(KP766044443, +0.766044443118978035202392650555416673935832457); DK(KP1_326827896, +1.326827896337876792410842639271782594433726619); DK(KP1_113340798, +1.113340798452838732905825904094046265936583811); DK(KP984807753, +0.984807753012208059366743024589523013670643252); DK(KP173648177, +0.173648177666930348851716626769314796000375677); DK(KP1_705737063, +1.705737063904886419256501927880148143872040591); DK(KP300767466, +0.300767466360870593278543795225003852144476517); DK(KP500000000, +0.500000000000000000000000000000000000000000000); DK(KP866025403, +0.866025403784438646763723170752936183471402627); DK(KP2_000000000, +2.000000000000000000000000000000000000000000000); DK(KP1_732050807, +1.732050807568877293527446341505872366942805254); { INT i; for (i = v; i > 0; i = i - 1, R0 = R0 + ovs, R1 = R1 + ovs, Cr = Cr + ivs, Ci = Ci + ivs, MAKE_VOLATILE_STRIDE(36, rs), MAKE_VOLATILE_STRIDE(36, csr), MAKE_VOLATILE_STRIDE(36, csi)) { E T3, Ts, Ti, Td, Tc, T8, To, Tu, Tl, Tt, T9, Te; { E Th, T1, T2, Tf, Tg; Tg = Ci[WS(csi, 1)]; Th = KP1_732050807 * Tg; T1 = Cr[WS(csr, 4)]; T2 = Cr[WS(csr, 1)]; Tf = T2 - T1; T3 = FMA(KP2_000000000, T2, T1); Ts = Tf - Th; Ti = Tf + Th; } { E T4, T7, Tm, Tk, Tn, Tj; T4 = Cr[WS(csr, 3)]; Td = Ci[WS(csi, 3)]; { E T5, T6, Ta, Tb; T5 = Cr[0]; T6 = Cr[WS(csr, 2)]; T7 = T5 + T6; Tm = KP866025403 * (T6 - T5); Ta = Ci[WS(csi, 2)]; Tb = Ci[0]; Tc = Ta - Tb; Tk = KP866025403 * (Tb + Ta); } T8 = T4 + T7; Tn = FMA(KP500000000, Tc, Td); To = Tm - Tn; Tu = Tm + Tn; Tj = FMS(KP500000000, T7, T4); Tl = Tj + Tk; Tt = Tj - Tk; } R0[0] = FMA(KP2_000000000, T8, T3); T9 = T8 - T3; Te = KP1_732050807 * (Tc - Td); R1[WS(rs, 1)] = T9 + Te; R0[WS(rs, 3)] = Te - T9; { E Tr, Tp, Tq, Tx, Tv, Tw; Tr = FNMS(KP1_705737063, Tl, KP300767466 * To); Tp = FMA(KP173648177, Tl, KP984807753 * To); Tq = Ti - Tp; R0[WS(rs, 1)] = -(FMA(KP2_000000000, Tp, Ti)); R0[WS(rs, 4)] = Tr - Tq; R1[WS(rs, 2)] = Tq + Tr; Tx = FMA(KP1_113340798, Tt, KP1_326827896 * Tu); Tv = FNMS(KP642787609, Tu, KP766044443 * Tt); Tw = Tv - Ts; R1[0] = FMA(KP2_000000000, Tv, Ts); R1[WS(rs, 3)] = Tx - Tw; R0[WS(rs, 2)] = Tw + Tx; } } } } static const kr2c_desc desc = { 9, "r2cbIII_9", {22, 8, 10, 0}, &GENUS }; void X(codelet_r2cbIII_9) (planner *p) { X(kr2c_register) (p, r2cbIII_9, &desc); } #endif /* HAVE_FMA */ fftw-3.3.3/rdft/scalar/r2cb/r2cbIII_10.c0000644000175000017500000001530612054411000014262 00000000000000/* * Copyright (c) 2003, 2007-11 Matteo Frigo * Copyright (c) 2003, 2007-11 Massachusetts Institute of Technology * * This program is free software; you can redistribute it and/or modify * it under the terms of the GNU General Public License as published by * the Free Software Foundation; either version 2 of the License, or * (at your option) any later version. * * This program is distributed in the hope that it will be useful, * but WITHOUT ANY WARRANTY; without even the implied warranty of * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the * GNU General Public License for more details. * * You should have received a copy of the GNU General Public License * along with this program; if not, write to the Free Software * Foundation, Inc., 51 Franklin Street, Fifth Floor, Boston, MA 02110-1301 USA * */ /* This file was automatically generated --- DO NOT EDIT */ /* Generated on Sun Nov 25 07:41:35 EST 2012 */ #include "codelet-rdft.h" #ifdef HAVE_FMA /* Generated by: ../../../genfft/gen_r2cb.native -fma -reorder-insns -schedule-for-pipeline -compact -variables 4 -pipeline-latency 4 -sign 1 -n 10 -name r2cbIII_10 -dft-III -include r2cbIII.h */ /* * This function contains 32 FP additions, 28 FP multiplications, * (or, 14 additions, 10 multiplications, 18 fused multiply/add), * 38 stack variables, 5 constants, and 20 memory accesses */ #include "r2cbIII.h" static void r2cbIII_10(R *R0, R *R1, R *Cr, R *Ci, stride rs, stride csr, stride csi, INT v, INT ivs, INT ovs) { DK(KP951056516, +0.951056516295153572116439333379382143405698634); DK(KP559016994, +0.559016994374947424102293417182819058860154590); DK(KP250000000, +0.250000000000000000000000000000000000000000000); DK(KP618033988, +0.618033988749894848204586834365638117720309180); DK(KP2_000000000, +2.000000000000000000000000000000000000000000000); { INT i; for (i = v; i > 0; i = i - 1, R0 = R0 + ovs, R1 = R1 + ovs, Cr = Cr + ivs, Ci = Ci + ivs, MAKE_VOLATILE_STRIDE(40, rs), MAKE_VOLATILE_STRIDE(40, csr), MAKE_VOLATILE_STRIDE(40, csi)) { E Tq, Ti, Tk, Tu, Tw, Tp, Tb, Tj, Tr, Tv; { E T1, To, Ts, Tt, T8, Ta, Te, Tl, Tm, Th, Tn, T9; T1 = Cr[WS(csr, 2)]; To = Ci[WS(csi, 2)]; { E T2, T3, T5, T6; T2 = Cr[WS(csr, 4)]; T3 = Cr[0]; T5 = Cr[WS(csr, 3)]; T6 = Cr[WS(csr, 1)]; { E Tc, T4, T7, Td, Tf, Tg; Tc = Ci[WS(csi, 3)]; Ts = T2 - T3; T4 = T2 + T3; Tt = T5 - T6; T7 = T5 + T6; Td = Ci[WS(csi, 1)]; Tf = Ci[WS(csi, 4)]; Tg = Ci[0]; T8 = T4 + T7; Ta = T7 - T4; Te = Tc - Td; Tl = Tc + Td; Tm = Tf + Tg; Th = Tf - Tg; } } R0[0] = KP2_000000000 * (T1 + T8); Tn = Tl - Tm; Tq = Tl + Tm; Ti = FMA(KP618033988, Th, Te); Tk = FNMS(KP618033988, Te, Th); R1[WS(rs, 2)] = KP2_000000000 * (Tn - To); T9 = FMS(KP250000000, T8, T1); Tu = FMA(KP618033988, Tt, Ts); Tw = FNMS(KP618033988, Ts, Tt); Tp = FMA(KP250000000, Tn, To); Tb = FNMS(KP559016994, Ta, T9); Tj = FMA(KP559016994, Ta, T9); } Tr = FMA(KP559016994, Tq, Tp); Tv = FNMS(KP559016994, Tq, Tp); R0[WS(rs, 2)] = -(KP2_000000000 * (FNMS(KP951056516, Tk, Tj))); R0[WS(rs, 3)] = KP2_000000000 * (FMA(KP951056516, Tk, Tj)); R0[WS(rs, 4)] = -(KP2_000000000 * (FNMS(KP951056516, Ti, Tb))); R0[WS(rs, 1)] = KP2_000000000 * (FMA(KP951056516, Ti, Tb)); R1[WS(rs, 1)] = KP2_000000000 * (FMA(KP951056516, Tw, Tv)); R1[WS(rs, 3)] = KP2_000000000 * (FNMS(KP951056516, Tw, Tv)); R1[WS(rs, 4)] = -(KP2_000000000 * (FNMS(KP951056516, Tu, Tr))); R1[0] = -(KP2_000000000 * (FMA(KP951056516, Tu, Tr))); } } } static const kr2c_desc desc = { 10, "r2cbIII_10", {14, 10, 18, 0}, &GENUS }; void X(codelet_r2cbIII_10) (planner *p) { X(kr2c_register) (p, r2cbIII_10, &desc); } #else /* HAVE_FMA */ /* Generated by: ../../../genfft/gen_r2cb.native -compact -variables 4 -pipeline-latency 4 -sign 1 -n 10 -name r2cbIII_10 -dft-III -include r2cbIII.h */ /* * This function contains 32 FP additions, 16 FP multiplications, * (or, 26 additions, 10 multiplications, 6 fused multiply/add), * 22 stack variables, 5 constants, and 20 memory accesses */ #include "r2cbIII.h" static void r2cbIII_10(R *R0, R *R1, R *Cr, R *Ci, stride rs, stride csr, stride csi, INT v, INT ivs, INT ovs) { DK(KP500000000, +0.500000000000000000000000000000000000000000000); DK(KP1_902113032, +1.902113032590307144232878666758764286811397268); DK(KP1_175570504, +1.175570504584946258337411909278145537195304875); DK(KP2_000000000, +2.000000000000000000000000000000000000000000000); DK(KP1_118033988, +1.118033988749894848204586834365638117720309180); { INT i; for (i = v; i > 0; i = i - 1, R0 = R0 + ovs, R1 = R1 + ovs, Cr = Cr + ivs, Ci = Ci + ivs, MAKE_VOLATILE_STRIDE(40, rs), MAKE_VOLATILE_STRIDE(40, csr), MAKE_VOLATILE_STRIDE(40, csi)) { E T1, To, T8, Tq, Ta, Tp, Te, Ts, Th, Tn; T1 = Cr[WS(csr, 2)]; To = Ci[WS(csi, 2)]; { E T2, T3, T4, T5, T6, T7; T2 = Cr[WS(csr, 4)]; T3 = Cr[0]; T4 = T2 + T3; T5 = Cr[WS(csr, 3)]; T6 = Cr[WS(csr, 1)]; T7 = T5 + T6; T8 = T4 + T7; Tq = T5 - T6; Ta = KP1_118033988 * (T7 - T4); Tp = T2 - T3; } { E Tc, Td, Tm, Tf, Tg, Tl; Tc = Ci[WS(csi, 4)]; Td = Ci[0]; Tm = Tc + Td; Tf = Ci[WS(csi, 1)]; Tg = Ci[WS(csi, 3)]; Tl = Tg + Tf; Te = Tc - Td; Ts = KP1_118033988 * (Tl + Tm); Th = Tf - Tg; Tn = Tl - Tm; } R0[0] = KP2_000000000 * (T1 + T8); R1[WS(rs, 2)] = KP2_000000000 * (Tn - To); { E Ti, Tj, Tb, Tk, T9; Ti = FNMS(KP1_902113032, Th, KP1_175570504 * Te); Tj = FMA(KP1_175570504, Th, KP1_902113032 * Te); T9 = FNMS(KP2_000000000, T1, KP500000000 * T8); Tb = T9 - Ta; Tk = T9 + Ta; R0[WS(rs, 1)] = Tb + Ti; R0[WS(rs, 3)] = Tk + Tj; R0[WS(rs, 4)] = Ti - Tb; R0[WS(rs, 2)] = Tj - Tk; } { E Tr, Tv, Tu, Tw, Tt; Tr = FMA(KP1_902113032, Tp, KP1_175570504 * Tq); Tv = FNMS(KP1_175570504, Tp, KP1_902113032 * Tq); Tt = FMA(KP500000000, Tn, KP2_000000000 * To); Tu = Ts + Tt; Tw = Tt - Ts; R1[0] = -(Tr + Tu); R1[WS(rs, 3)] = Tw - Tv; R1[WS(rs, 4)] = Tr - Tu; R1[WS(rs, 1)] = Tv + Tw; } } } } static const kr2c_desc desc = { 10, "r2cbIII_10", {26, 10, 6, 0}, &GENUS }; void X(codelet_r2cbIII_10) (planner *p) { X(kr2c_register) (p, r2cbIII_10, &desc); } #endif /* HAVE_FMA */ fftw-3.3.3/rdft/scalar/r2cb/r2cbIII_12.c0000644000175000017500000001622412054411000014264 00000000000000/* * Copyright (c) 2003, 2007-11 Matteo Frigo * Copyright (c) 2003, 2007-11 Massachusetts Institute of Technology * * This program is free software; you can redistribute it and/or modify * it under the terms of the GNU General Public License as published by * the Free Software Foundation; either version 2 of the License, or * (at your option) any later version. * * This program is distributed in the hope that it will be useful, * but WITHOUT ANY WARRANTY; without even the implied warranty of * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the * GNU General Public License for more details. * * You should have received a copy of the GNU General Public License * along with this program; if not, write to the Free Software * Foundation, Inc., 51 Franklin Street, Fifth Floor, Boston, MA 02110-1301 USA * */ /* This file was automatically generated --- DO NOT EDIT */ /* Generated on Sun Nov 25 07:41:36 EST 2012 */ #include "codelet-rdft.h" #ifdef HAVE_FMA /* Generated by: ../../../genfft/gen_r2cb.native -fma -reorder-insns -schedule-for-pipeline -compact -variables 4 -pipeline-latency 4 -sign 1 -n 12 -name r2cbIII_12 -dft-III -include r2cbIII.h */ /* * This function contains 42 FP additions, 20 FP multiplications, * (or, 30 additions, 8 multiplications, 12 fused multiply/add), * 37 stack variables, 4 constants, and 24 memory accesses */ #include "r2cbIII.h" static void r2cbIII_12(R *R0, R *R1, R *Cr, R *Ci, stride rs, stride csr, stride csi, INT v, INT ivs, INT ovs) { DK(KP707106781, +0.707106781186547524400844362104849039284835938); DK(KP1_414213562, +1.414213562373095048801688724209698078569671875); DK(KP1_732050807, +1.732050807568877293527446341505872366942805254); DK(KP2_000000000, +2.000000000000000000000000000000000000000000000); { INT i; for (i = v; i > 0; i = i - 1, R0 = R0 + ovs, R1 = R1 + ovs, Cr = Cr + ivs, Ci = Ci + ivs, MAKE_VOLATILE_STRIDE(48, rs), MAKE_VOLATILE_STRIDE(48, csr), MAKE_VOLATILE_STRIDE(48, csi)) { E TE, TD, TF, TG; { E Tx, T6, Te, Tb, T5, Tw, Ts, To, Th, Ti, T9, TA; { E T1, Tq, Tc, Td, T4, T2, T3, T7, T8, Tr; T1 = Cr[WS(csr, 1)]; T2 = Cr[WS(csr, 5)]; T3 = Cr[WS(csr, 2)]; Tq = Ci[WS(csi, 1)]; Tc = Ci[WS(csi, 5)]; Td = Ci[WS(csi, 2)]; T4 = T2 + T3; Tx = T2 - T3; T6 = Cr[WS(csr, 4)]; Te = Tc + Td; Tr = Td - Tc; Tb = FNMS(KP2_000000000, T1, T4); T5 = T1 + T4; T7 = Cr[0]; Tw = FMA(KP2_000000000, Tq, Tr); Ts = Tq - Tr; T8 = Cr[WS(csr, 3)]; To = Ci[WS(csi, 4)]; Th = Ci[0]; Ti = Ci[WS(csi, 3)]; T9 = T7 + T8; TA = T7 - T8; } { E Tl, Tm, Tv, TC; { E Tf, Ty, Tk, TB; { E Tj, Tn, Tg, Ta; Tl = FNMS(KP1_732050807, Te, Tb); Tf = FMA(KP1_732050807, Te, Tb); Tj = Th + Ti; Tn = Ti - Th; Tg = FNMS(KP2_000000000, T6, T9); Ta = T6 + T9; { E Tu, Tt, Tz, Tp; Ty = FMA(KP1_732050807, Tx, Tw); TE = FNMS(KP1_732050807, Tx, Tw); Tz = FMA(KP2_000000000, To, Tn); Tp = Tn - To; Tm = FMA(KP1_732050807, Tj, Tg); Tk = FNMS(KP1_732050807, Tj, Tg); Tu = T5 - Ta; R0[0] = KP2_000000000 * (T5 + Ta); Tt = Tp - Ts; R0[WS(rs, 3)] = KP2_000000000 * (Ts + Tp); Tv = Tk - Tf; TD = FMA(KP1_732050807, TA, Tz); TB = FNMS(KP1_732050807, TA, Tz); R1[WS(rs, 4)] = KP1_414213562 * (Tu + Tt); R1[WS(rs, 1)] = KP1_414213562 * (Tt - Tu); } } R0[WS(rs, 2)] = Tf + Tk; TC = Ty + TB; R0[WS(rs, 5)] = TB - Ty; } R1[WS(rs, 3)] = KP707106781 * (Tv + TC); R1[0] = KP707106781 * (Tv - TC); TF = Tl - Tm; R0[WS(rs, 4)] = -(Tl + Tm); } } R0[WS(rs, 1)] = TD - TE; TG = TE + TD; R1[WS(rs, 5)] = KP707106781 * (TF - TG); R1[WS(rs, 2)] = KP707106781 * (TF + TG); } } } static const kr2c_desc desc = { 12, "r2cbIII_12", {30, 8, 12, 0}, &GENUS }; void X(codelet_r2cbIII_12) (planner *p) { X(kr2c_register) (p, r2cbIII_12, &desc); } #else /* HAVE_FMA */ /* Generated by: ../../../genfft/gen_r2cb.native -compact -variables 4 -pipeline-latency 4 -sign 1 -n 12 -name r2cbIII_12 -dft-III -include r2cbIII.h */ /* * This function contains 42 FP additions, 20 FP multiplications, * (or, 38 additions, 16 multiplications, 4 fused multiply/add), * 25 stack variables, 4 constants, and 24 memory accesses */ #include "r2cbIII.h" static void r2cbIII_12(R *R0, R *R1, R *Cr, R *Ci, stride rs, stride csr, stride csi, INT v, INT ivs, INT ovs) { DK(KP1_414213562, +1.414213562373095048801688724209698078569671875); DK(KP2_000000000, +2.000000000000000000000000000000000000000000000); DK(KP500000000, +0.500000000000000000000000000000000000000000000); DK(KP866025403, +0.866025403784438646763723170752936183471402627); { INT i; for (i = v; i > 0; i = i - 1, R0 = R0 + ovs, R1 = R1 + ovs, Cr = Cr + ivs, Ci = Ci + ivs, MAKE_VOLATILE_STRIDE(48, rs), MAKE_VOLATILE_STRIDE(48, csr), MAKE_VOLATILE_STRIDE(48, csi)) { E T5, Tw, Tb, Te, Tx, Ts, Ta, TA, Tg, Tj, Tz, Tp, Tt, Tu; { E T1, T2, T3, T4; T1 = Cr[WS(csr, 1)]; T2 = Cr[WS(csr, 5)]; T3 = Cr[WS(csr, 2)]; T4 = T2 + T3; T5 = T1 + T4; Tw = KP866025403 * (T2 - T3); Tb = FNMS(KP500000000, T4, T1); } { E Tq, Tc, Td, Tr; Tq = Ci[WS(csi, 1)]; Tc = Ci[WS(csi, 5)]; Td = Ci[WS(csi, 2)]; Tr = Td - Tc; Te = KP866025403 * (Tc + Td); Tx = FMA(KP500000000, Tr, Tq); Ts = Tq - Tr; } { E T6, T7, T8, T9; T6 = Cr[WS(csr, 4)]; T7 = Cr[0]; T8 = Cr[WS(csr, 3)]; T9 = T7 + T8; Ta = T6 + T9; TA = KP866025403 * (T7 - T8); Tg = FNMS(KP500000000, T9, T6); } { E To, Th, Ti, Tn; To = Ci[WS(csi, 4)]; Th = Ci[0]; Ti = Ci[WS(csi, 3)]; Tn = Ti - Th; Tj = KP866025403 * (Th + Ti); Tz = FMA(KP500000000, Tn, To); Tp = Tn - To; } R0[0] = KP2_000000000 * (T5 + Ta); R0[WS(rs, 3)] = KP2_000000000 * (Ts + Tp); Tt = Tp - Ts; Tu = T5 - Ta; R1[WS(rs, 1)] = KP1_414213562 * (Tt - Tu); R1[WS(rs, 4)] = KP1_414213562 * (Tu + Tt); { E Tf, Tk, Tv, Ty, TB, TC; Tf = Tb - Te; Tk = Tg + Tj; Tv = Tf - Tk; Ty = Tw + Tx; TB = Tz - TA; TC = Ty + TB; R0[WS(rs, 2)] = -(KP2_000000000 * (Tf + Tk)); R0[WS(rs, 5)] = KP2_000000000 * (TB - Ty); R1[0] = KP1_414213562 * (Tv - TC); R1[WS(rs, 3)] = KP1_414213562 * (Tv + TC); } { E Tl, Tm, TF, TD, TE, TG; Tl = Tb + Te; Tm = Tg - Tj; TF = Tm - Tl; TD = TA + Tz; TE = Tx - Tw; TG = TE + TD; R0[WS(rs, 4)] = KP2_000000000 * (Tl + Tm); R1[WS(rs, 2)] = KP1_414213562 * (TF + TG); R0[WS(rs, 1)] = KP2_000000000 * (TD - TE); R1[WS(rs, 5)] = KP1_414213562 * (TF - TG); } } } } static const kr2c_desc desc = { 12, "r2cbIII_12", {38, 16, 4, 0}, &GENUS }; void X(codelet_r2cbIII_12) (planner *p) { X(kr2c_register) (p, r2cbIII_12, &desc); } #endif /* HAVE_FMA */ fftw-3.3.3/rdft/scalar/r2cb/r2cbIII_15.c0000644000175000017500000002441212054411001014266 00000000000000/* * Copyright (c) 2003, 2007-11 Matteo Frigo * Copyright (c) 2003, 2007-11 Massachusetts Institute of Technology * * This program is free software; you can redistribute it and/or modify * it under the terms of the GNU General Public License as published by * the Free Software Foundation; either version 2 of the License, or * (at your option) any later version. * * This program is distributed in the hope that it will be useful, * but WITHOUT ANY WARRANTY; without even the implied warranty of * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the * GNU General Public License for more details. * * You should have received a copy of the GNU General Public License * along with this program; if not, write to the Free Software * Foundation, Inc., 51 Franklin Street, Fifth Floor, Boston, MA 02110-1301 USA * */ /* This file was automatically generated --- DO NOT EDIT */ /* Generated on Sun Nov 25 07:41:36 EST 2012 */ #include "codelet-rdft.h" #ifdef HAVE_FMA /* Generated by: ../../../genfft/gen_r2cb.native -fma -reorder-insns -schedule-for-pipeline -compact -variables 4 -pipeline-latency 4 -sign 1 -n 15 -name r2cbIII_15 -dft-III -include r2cbIII.h */ /* * This function contains 64 FP additions, 43 FP multiplications, * (or, 21 additions, 0 multiplications, 43 fused multiply/add), * 48 stack variables, 9 constants, and 30 memory accesses */ #include "r2cbIII.h" static void r2cbIII_15(R *R0, R *R1, R *Cr, R *Ci, stride rs, stride csr, stride csi, INT v, INT ivs, INT ovs) { DK(KP951056516, +0.951056516295153572116439333379382143405698634); DK(KP1_732050807, +1.732050807568877293527446341505872366942805254); DK(KP559016994, +0.559016994374947424102293417182819058860154590); DK(KP250000000, +0.250000000000000000000000000000000000000000000); DK(KP1_902113032, +1.902113032590307144232878666758764286811397268); DK(KP1_118033988, +1.118033988749894848204586834365638117720309180); DK(KP2_000000000, +2.000000000000000000000000000000000000000000000); DK(KP500000000, +0.500000000000000000000000000000000000000000000); DK(KP618033988, +0.618033988749894848204586834365638117720309180); { INT i; for (i = v; i > 0; i = i - 1, R0 = R0 + ovs, R1 = R1 + ovs, Cr = Cr + ivs, Ci = Ci + ivs, MAKE_VOLATILE_STRIDE(60, rs), MAKE_VOLATILE_STRIDE(60, csr), MAKE_VOLATILE_STRIDE(60, csi)) { E TX, Tv, To, TW, Tl, Tx, Ty, Tw; { E TA, Tk, T6, T5, Tz, Th, TI, Tp, Tu, TK, TR, Tn, Td, Tq; { E T1, T2, T3, Ti, Tj; Ti = Ci[WS(csi, 4)]; Tj = Ci[WS(csi, 1)]; T1 = Cr[WS(csr, 7)]; T2 = Cr[WS(csr, 4)]; T3 = Cr[WS(csr, 1)]; TA = FNMS(KP618033988, Ti, Tj); Tk = FMA(KP618033988, Tj, Ti); { E T7, TP, Tc, T8; T6 = Cr[WS(csr, 2)]; { E T4, Tg, Ta, Tb, Tf; T4 = T2 + T3; Tg = T2 - T3; Ta = Cr[WS(csr, 3)]; Tb = Cr[WS(csr, 6)]; T7 = Cr[0]; Tf = FNMS(KP500000000, T4, T1); T5 = FMA(KP2_000000000, T4, T1); TP = Ta - Tb; Tc = Ta + Tb; Tz = FNMS(KP1_118033988, Tg, Tf); Th = FMA(KP1_118033988, Tg, Tf); T8 = Cr[WS(csr, 5)]; } TI = Ci[WS(csi, 2)]; { E Ts, Tt, TQ, T9; Ts = Ci[WS(csi, 3)]; Tt = Ci[WS(csi, 6)]; TQ = T7 - T8; T9 = T7 + T8; Tp = Ci[0]; Tu = Ts - Tt; TK = Ts + Tt; TX = FMA(KP618033988, TP, TQ); TR = FNMS(KP618033988, TQ, TP); Tn = T9 - Tc; Td = T9 + Tc; Tq = Ci[WS(csi, 5)]; } } } { E TB, TF, TO, TG, TE; { E Tm, T11, TN, TD, TM, T12, TC; TB = FNMS(KP1_902113032, TA, Tz); TF = FMA(KP1_902113032, TA, Tz); { E Te, Tr, TJ, TL; Tm = FNMS(KP250000000, Td, T6); Te = T6 + Td; Tr = Tp + Tq; TJ = Tq - Tp; R0[0] = FMA(KP2_000000000, Te, T5); T11 = Te - T5; TN = TJ + TK; TL = TJ - TK; Tv = FMA(KP618033988, Tu, Tr); TD = FNMS(KP618033988, Tr, Tu); TM = FNMS(KP250000000, TL, TI); T12 = TL + TI; } TC = FNMS(KP559016994, Tn, Tm); To = FMA(KP559016994, Tn, Tm); R1[WS(rs, 2)] = FMA(KP1_732050807, T12, T11); R0[WS(rs, 5)] = FMS(KP1_732050807, T12, T11); TW = FMA(KP559016994, TN, TM); TO = FNMS(KP559016994, TN, TM); TG = FNMS(KP951056516, TD, TC); TE = FMA(KP951056516, TD, TC); } Tl = FNMS(KP1_902113032, Tk, Th); Tx = FMA(KP1_902113032, Tk, Th); { E TS, TU, TT, TH; TS = FMA(KP951056516, TR, TO); TU = FNMS(KP951056516, TR, TO); TT = TF - TG; R1[WS(rs, 1)] = -(FMA(KP2_000000000, TG, TF)); TH = TB - TE; R0[WS(rs, 6)] = FMA(KP2_000000000, TE, TB); R1[WS(rs, 6)] = -(FMA(KP1_732050807, TU, TT)); R0[WS(rs, 4)] = FNMS(KP1_732050807, TU, TT); R1[WS(rs, 3)] = -(FMA(KP1_732050807, TS, TH)); R0[WS(rs, 1)] = FNMS(KP1_732050807, TS, TH); } } } Ty = FNMS(KP951056516, Tv, To); Tw = FMA(KP951056516, Tv, To); { E T10, TY, TV, TZ; T10 = FMA(KP951056516, TX, TW); TY = FNMS(KP951056516, TX, TW); TV = Ty - Tx; R0[WS(rs, 3)] = FMA(KP2_000000000, Ty, Tx); TZ = Tl - Tw; R1[WS(rs, 4)] = -(FMA(KP2_000000000, Tw, Tl)); R1[WS(rs, 5)] = FMA(KP1_732050807, TY, TV); R1[0] = FNMS(KP1_732050807, TY, TV); R0[WS(rs, 2)] = FMA(KP1_732050807, T10, TZ); R0[WS(rs, 7)] = FNMS(KP1_732050807, T10, TZ); } } } } static const kr2c_desc desc = { 15, "r2cbIII_15", {21, 0, 43, 0}, &GENUS }; void X(codelet_r2cbIII_15) (planner *p) { X(kr2c_register) (p, r2cbIII_15, &desc); } #else /* HAVE_FMA */ /* Generated by: ../../../genfft/gen_r2cb.native -compact -variables 4 -pipeline-latency 4 -sign 1 -n 15 -name r2cbIII_15 -dft-III -include r2cbIII.h */ /* * This function contains 64 FP additions, 26 FP multiplications, * (or, 49 additions, 11 multiplications, 15 fused multiply/add), * 47 stack variables, 14 constants, and 30 memory accesses */ #include "r2cbIII.h" static void r2cbIII_15(R *R0, R *R1, R *Cr, R *Ci, stride rs, stride csr, stride csi, INT v, INT ivs, INT ovs) { DK(KP1_732050807, +1.732050807568877293527446341505872366942805254); DK(KP433012701, +0.433012701892219323381861585376468091735701313); DK(KP968245836, +0.968245836551854221294816349945599902708230426); DK(KP587785252, +0.587785252292473129168705954639072768597652438); DK(KP951056516, +0.951056516295153572116439333379382143405698634); DK(KP250000000, +0.250000000000000000000000000000000000000000000); DK(KP1_647278207, +1.647278207092663851754840078556380006059321028); DK(KP1_018073920, +1.018073920910254366901961726787815297021466329); DK(KP559016994, +0.559016994374947424102293417182819058860154590); DK(KP500000000, +0.500000000000000000000000000000000000000000000); DK(KP2_000000000, +2.000000000000000000000000000000000000000000000); DK(KP1_118033988, +1.118033988749894848204586834365638117720309180); DK(KP1_175570504, +1.175570504584946258337411909278145537195304875); DK(KP1_902113032, +1.902113032590307144232878666758764286811397268); { INT i; for (i = v; i > 0; i = i - 1, R0 = R0 + ovs, R1 = R1 + ovs, Cr = Cr + ivs, Ci = Ci + ivs, MAKE_VOLATILE_STRIDE(60, rs), MAKE_VOLATILE_STRIDE(60, csr), MAKE_VOLATILE_STRIDE(60, csi)) { E Tv, TD, T5, Ts, TC, T6, Tf, TW, TK, Td, Tg, TP, To, TN, TA; E TO, TQ, Tt, Tu, T12, Te, T11; Tt = Ci[WS(csi, 4)]; Tu = Ci[WS(csi, 1)]; Tv = FMA(KP1_902113032, Tt, KP1_175570504 * Tu); TD = FNMS(KP1_175570504, Tt, KP1_902113032 * Tu); { E T1, T4, Tq, T2, T3, Tr; T1 = Cr[WS(csr, 7)]; T2 = Cr[WS(csr, 4)]; T3 = Cr[WS(csr, 1)]; T4 = T2 + T3; Tq = KP1_118033988 * (T2 - T3); T5 = FMA(KP2_000000000, T4, T1); Tr = FNMS(KP500000000, T4, T1); Ts = Tq + Tr; TC = Tr - Tq; } { E Tc, TJ, T9, TI; T6 = Cr[WS(csr, 2)]; { E Ta, Tb, T7, T8; Ta = Cr[WS(csr, 3)]; Tb = Cr[WS(csr, 6)]; Tc = Ta + Tb; TJ = Ta - Tb; T7 = Cr[0]; T8 = Cr[WS(csr, 5)]; T9 = T7 + T8; TI = T7 - T8; } Tf = KP559016994 * (T9 - Tc); TW = FNMS(KP1_647278207, TJ, KP1_018073920 * TI); TK = FMA(KP1_647278207, TI, KP1_018073920 * TJ); Td = T9 + Tc; Tg = FNMS(KP250000000, Td, T6); } { E Tn, TM, Tk, TL; TP = Ci[WS(csi, 2)]; { E Tl, Tm, Ti, Tj; Tl = Ci[WS(csi, 3)]; Tm = Ci[WS(csi, 6)]; Tn = Tl - Tm; TM = Tl + Tm; Ti = Ci[0]; Tj = Ci[WS(csi, 5)]; Tk = Ti + Tj; TL = Ti - Tj; } To = FMA(KP951056516, Tk, KP587785252 * Tn); TN = KP968245836 * (TL - TM); TA = FNMS(KP587785252, Tk, KP951056516 * Tn); TO = TL + TM; TQ = FMA(KP433012701, TO, KP1_732050807 * TP); } T12 = KP1_732050807 * (TP - TO); Te = T6 + Td; T11 = Te - T5; R0[0] = FMA(KP2_000000000, Te, T5); R0[WS(rs, 5)] = T12 - T11; R1[WS(rs, 2)] = T11 + T12; { E TE, TG, TB, TF, TY, T10, Tz, TX, TV, TZ; TE = TC - TD; TG = TC + TD; Tz = Tg - Tf; TB = Tz + TA; TF = TA - Tz; TX = TN + TQ; TY = TW - TX; T10 = TW + TX; R0[WS(rs, 6)] = FMA(KP2_000000000, TB, TE); R1[WS(rs, 1)] = FMS(KP2_000000000, TF, TG); TV = TE - TB; R0[WS(rs, 1)] = TV + TY; R1[WS(rs, 3)] = TY - TV; TZ = TF + TG; R0[WS(rs, 4)] = TZ - T10; R1[WS(rs, 6)] = -(TZ + T10); } { E Tw, Ty, Tp, Tx, TS, TU, Th, TR, TH, TT; Tw = Ts - Tv; Ty = Ts + Tv; Th = Tf + Tg; Tp = Th + To; Tx = Th - To; TR = TN - TQ; TS = TK + TR; TU = TR - TK; R1[WS(rs, 4)] = -(FMA(KP2_000000000, Tp, Tw)); R0[WS(rs, 3)] = FMA(KP2_000000000, Tx, Ty); TH = Tx - Ty; R1[WS(rs, 5)] = TH - TS; R1[0] = TH + TS; TT = Tw - Tp; R0[WS(rs, 2)] = TT - TU; R0[WS(rs, 7)] = TT + TU; } } } } static const kr2c_desc desc = { 15, "r2cbIII_15", {49, 11, 15, 0}, &GENUS }; void X(codelet_r2cbIII_15) (planner *p) { X(kr2c_register) (p, r2cbIII_15, &desc); } #endif /* HAVE_FMA */ fftw-3.3.3/rdft/scalar/r2cb/r2cbIII_16.c0000644000175000017500000002404212054411002014267 00000000000000/* * Copyright (c) 2003, 2007-11 Matteo Frigo * Copyright (c) 2003, 2007-11 Massachusetts Institute of Technology * * This program is free software; you can redistribute it and/or modify * it under the terms of the GNU General Public License as published by * the Free Software Foundation; either version 2 of the License, or * (at your option) any later version. * * This program is distributed in the hope that it will be useful, * but WITHOUT ANY WARRANTY; without even the implied warranty of * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the * GNU General Public License for more details. * * You should have received a copy of the GNU General Public License * along with this program; if not, write to the Free Software * Foundation, Inc., 51 Franklin Street, Fifth Floor, Boston, MA 02110-1301 USA * */ /* This file was automatically generated --- DO NOT EDIT */ /* Generated on Sun Nov 25 07:41:36 EST 2012 */ #include "codelet-rdft.h" #ifdef HAVE_FMA /* Generated by: ../../../genfft/gen_r2cb.native -fma -reorder-insns -schedule-for-pipeline -compact -variables 4 -pipeline-latency 4 -sign 1 -n 16 -name r2cbIII_16 -dft-III -include r2cbIII.h */ /* * This function contains 66 FP additions, 36 FP multiplications, * (or, 46 additions, 16 multiplications, 20 fused multiply/add), * 55 stack variables, 9 constants, and 32 memory accesses */ #include "r2cbIII.h" static void r2cbIII_16(R *R0, R *R1, R *Cr, R *Ci, stride rs, stride csr, stride csi, INT v, INT ivs, INT ovs) { DK(KP668178637, +0.668178637919298919997757686523080761552472251); DK(KP1_662939224, +1.662939224605090474157576755235811513477121624); DK(KP198912367, +0.198912367379658006911597622644676228597850501); DK(KP1_961570560, +1.961570560806460898252364472268478073947867462); DK(KP707106781, +0.707106781186547524400844362104849039284835938); DK(KP1_414213562, +1.414213562373095048801688724209698078569671875); DK(KP414213562, +0.414213562373095048801688724209698078569671875); DK(KP1_847759065, +1.847759065022573512256366378793576573644833252); DK(KP2_000000000, +2.000000000000000000000000000000000000000000000); { INT i; for (i = v; i > 0; i = i - 1, R0 = R0 + ovs, R1 = R1 + ovs, Cr = Cr + ivs, Ci = Ci + ivs, MAKE_VOLATILE_STRIDE(64, rs), MAKE_VOLATILE_STRIDE(64, csr), MAKE_VOLATILE_STRIDE(64, csi)) { E TA, TD, Tv, TG, TE, TF; { E TK, TP, T7, T13, TW, TH, Tj, TC, To, Te, TX, TS, T12, Tt, TB; { E T4, Tf, T3, TU, Tz, T5, Tg, Th; { E T1, T2, Tx, Ty; T1 = Cr[0]; T2 = Cr[WS(csr, 7)]; Tx = Ci[0]; Ty = Ci[WS(csi, 7)]; T4 = Cr[WS(csr, 4)]; Tf = T1 - T2; T3 = T1 + T2; TU = Ty - Tx; Tz = Tx + Ty; T5 = Cr[WS(csr, 3)]; Tg = Ci[WS(csi, 4)]; Th = Ci[WS(csi, 3)]; } { E Tb, Tk, Ta, TR, Tn, Tc, Tq, Tr; { E T8, T9, Tl, Tm; T8 = Cr[WS(csr, 2)]; { E Tw, T6, TV, Ti; Tw = T4 - T5; T6 = T4 + T5; TV = Th - Tg; Ti = Tg + Th; TK = Tw - Tz; TA = Tw + Tz; TP = T3 - T6; T7 = T3 + T6; T13 = TV + TU; TW = TU - TV; TH = Tf + Ti; Tj = Tf - Ti; T9 = Cr[WS(csr, 5)]; } Tl = Ci[WS(csi, 2)]; Tm = Ci[WS(csi, 5)]; Tb = Cr[WS(csr, 1)]; Tk = T8 - T9; Ta = T8 + T9; TR = Tl - Tm; Tn = Tl + Tm; Tc = Cr[WS(csr, 6)]; Tq = Ci[WS(csi, 1)]; Tr = Ci[WS(csi, 6)]; } TC = Tk + Tn; To = Tk - Tn; { E Tp, Td, TQ, Ts; Tp = Tb - Tc; Td = Tb + Tc; TQ = Tr - Tq; Ts = Tq + Tr; Te = Ta + Td; TX = Ta - Td; TS = TQ - TR; T12 = TR + TQ; Tt = Tp - Ts; TB = Tp + Ts; } } } { E T10, TT, TY, TZ; R0[0] = KP2_000000000 * (T7 + Te); R0[WS(rs, 4)] = KP2_000000000 * (T13 - T12); T10 = TP - TS; TT = TP + TS; TY = TW - TX; TZ = TX + TW; { E T11, T14, TI, TL, Tu; T11 = T7 - Te; T14 = T12 + T13; R0[WS(rs, 5)] = KP1_847759065 * (FNMS(KP414213562, TT, TY)); R0[WS(rs, 1)] = KP1_847759065 * (FMA(KP414213562, TY, TT)); R0[WS(rs, 6)] = KP1_414213562 * (T14 - T11); R0[WS(rs, 2)] = KP1_414213562 * (T11 + T14); TD = TB - TC; TI = TC + TB; TL = To - Tt; Tu = To + Tt; { E TO, TJ, TN, TM; R0[WS(rs, 7)] = -(KP1_847759065 * (FNMS(KP414213562, TZ, T10))); R0[WS(rs, 3)] = KP1_847759065 * (FMA(KP414213562, T10, TZ)); TO = FMA(KP707106781, TI, TH); TJ = FNMS(KP707106781, TI, TH); TN = FMA(KP707106781, TL, TK); TM = FNMS(KP707106781, TL, TK); Tv = FMA(KP707106781, Tu, Tj); TG = FNMS(KP707106781, Tu, Tj); R1[WS(rs, 3)] = KP1_961570560 * (FMA(KP198912367, TO, TN)); R1[WS(rs, 7)] = -(KP1_961570560 * (FNMS(KP198912367, TN, TO))); R1[WS(rs, 5)] = KP1_662939224 * (FNMS(KP668178637, TJ, TM)); R1[WS(rs, 1)] = KP1_662939224 * (FMA(KP668178637, TM, TJ)); } } } } TE = FNMS(KP707106781, TD, TA); TF = FMA(KP707106781, TD, TA); R1[WS(rs, 2)] = -(KP1_662939224 * (FNMS(KP668178637, TG, TF))); R1[WS(rs, 6)] = -(KP1_662939224 * (FMA(KP668178637, TF, TG))); R1[WS(rs, 4)] = -(KP1_961570560 * (FMA(KP198912367, Tv, TE))); R1[0] = KP1_961570560 * (FNMS(KP198912367, TE, Tv)); } } } static const kr2c_desc desc = { 16, "r2cbIII_16", {46, 16, 20, 0}, &GENUS }; void X(codelet_r2cbIII_16) (planner *p) { X(kr2c_register) (p, r2cbIII_16, &desc); } #else /* HAVE_FMA */ /* Generated by: ../../../genfft/gen_r2cb.native -compact -variables 4 -pipeline-latency 4 -sign 1 -n 16 -name r2cbIII_16 -dft-III -include r2cbIII.h */ /* * This function contains 66 FP additions, 32 FP multiplications, * (or, 54 additions, 20 multiplications, 12 fused multiply/add), * 40 stack variables, 9 constants, and 32 memory accesses */ #include "r2cbIII.h" static void r2cbIII_16(R *R0, R *R1, R *Cr, R *Ci, stride rs, stride csr, stride csi, INT v, INT ivs, INT ovs) { DK(KP1_961570560, +1.961570560806460898252364472268478073947867462); DK(KP390180644, +0.390180644032256535696569736954044481855383236); DK(KP1_111140466, +1.111140466039204449485661627897065748749874382); DK(KP1_662939224, +1.662939224605090474157576755235811513477121624); DK(KP707106781, +0.707106781186547524400844362104849039284835938); DK(KP1_414213562, +1.414213562373095048801688724209698078569671875); DK(KP765366864, +0.765366864730179543456919968060797733522689125); DK(KP1_847759065, +1.847759065022573512256366378793576573644833252); DK(KP2_000000000, +2.000000000000000000000000000000000000000000000); { INT i; for (i = v; i > 0; i = i - 1, R0 = R0 + ovs, R1 = R1 + ovs, Cr = Cr + ivs, Ci = Ci + ivs, MAKE_VOLATILE_STRIDE(64, rs), MAKE_VOLATILE_STRIDE(64, csr), MAKE_VOLATILE_STRIDE(64, csi)) { E T7, TW, T13, Tj, TD, TK, TP, TH, Te, TX, T12, To, Tt, Tx, TS; E Tw, TT, TY; { E T3, Tf, TC, TV, T6, Tz, Ti, TU; { E T1, T2, TA, TB; T1 = Cr[0]; T2 = Cr[WS(csr, 7)]; T3 = T1 + T2; Tf = T1 - T2; TA = Ci[0]; TB = Ci[WS(csi, 7)]; TC = TA + TB; TV = TB - TA; } { E T4, T5, Tg, Th; T4 = Cr[WS(csr, 4)]; T5 = Cr[WS(csr, 3)]; T6 = T4 + T5; Tz = T4 - T5; Tg = Ci[WS(csi, 4)]; Th = Ci[WS(csi, 3)]; Ti = Tg + Th; TU = Tg - Th; } T7 = T3 + T6; TW = TU + TV; T13 = TV - TU; Tj = Tf - Ti; TD = Tz + TC; TK = Tz - TC; TP = T3 - T6; TH = Tf + Ti; } { E Ta, Tk, Tn, TR, Td, Tp, Ts, TQ; { E T8, T9, Tl, Tm; T8 = Cr[WS(csr, 2)]; T9 = Cr[WS(csr, 5)]; Ta = T8 + T9; Tk = T8 - T9; Tl = Ci[WS(csi, 2)]; Tm = Ci[WS(csi, 5)]; Tn = Tl + Tm; TR = Tl - Tm; } { E Tb, Tc, Tq, Tr; Tb = Cr[WS(csr, 1)]; Tc = Cr[WS(csr, 6)]; Td = Tb + Tc; Tp = Tb - Tc; Tq = Ci[WS(csi, 1)]; Tr = Ci[WS(csi, 6)]; Ts = Tq + Tr; TQ = Tr - Tq; } Te = Ta + Td; TX = Ta - Td; T12 = TR + TQ; To = Tk - Tn; Tt = Tp - Ts; Tx = Tp + Ts; TS = TQ - TR; Tw = Tk + Tn; } R0[0] = KP2_000000000 * (T7 + Te); R0[WS(rs, 4)] = KP2_000000000 * (T13 - T12); TT = TP + TS; TY = TW - TX; R0[WS(rs, 1)] = FMA(KP1_847759065, TT, KP765366864 * TY); R0[WS(rs, 5)] = FNMS(KP765366864, TT, KP1_847759065 * TY); { E T11, T14, TZ, T10; T11 = T7 - Te; T14 = T12 + T13; R0[WS(rs, 2)] = KP1_414213562 * (T11 + T14); R0[WS(rs, 6)] = KP1_414213562 * (T14 - T11); TZ = TP - TS; T10 = TX + TW; R0[WS(rs, 3)] = FMA(KP765366864, TZ, KP1_847759065 * T10); R0[WS(rs, 7)] = FNMS(KP1_847759065, TZ, KP765366864 * T10); } { E TJ, TN, TM, TO, TI, TL; TI = KP707106781 * (Tw + Tx); TJ = TH - TI; TN = TH + TI; TL = KP707106781 * (To - Tt); TM = TK - TL; TO = TL + TK; R1[WS(rs, 1)] = FMA(KP1_662939224, TJ, KP1_111140466 * TM); R1[WS(rs, 7)] = FNMS(KP1_961570560, TN, KP390180644 * TO); R1[WS(rs, 5)] = FNMS(KP1_111140466, TJ, KP1_662939224 * TM); R1[WS(rs, 3)] = FMA(KP390180644, TN, KP1_961570560 * TO); } { E Tv, TF, TE, TG, Tu, Ty; Tu = KP707106781 * (To + Tt); Tv = Tj + Tu; TF = Tj - Tu; Ty = KP707106781 * (Tw - Tx); TE = Ty + TD; TG = Ty - TD; R1[0] = FNMS(KP390180644, TE, KP1_961570560 * Tv); R1[WS(rs, 6)] = FNMS(KP1_662939224, TF, KP1_111140466 * TG); R1[WS(rs, 4)] = -(FMA(KP390180644, Tv, KP1_961570560 * TE)); R1[WS(rs, 2)] = FMA(KP1_111140466, TF, KP1_662939224 * TG); } } } } static const kr2c_desc desc = { 16, "r2cbIII_16", {54, 20, 12, 0}, &GENUS }; void X(codelet_r2cbIII_16) (planner *p) { X(kr2c_register) (p, r2cbIII_16, &desc); } #endif /* HAVE_FMA */ fftw-3.3.3/rdft/scalar/r2cb/r2cbIII_32.c0000644000175000017500000005240612054411010014271 00000000000000/* * Copyright (c) 2003, 2007-11 Matteo Frigo * Copyright (c) 2003, 2007-11 Massachusetts Institute of Technology * * This program is free software; you can redistribute it and/or modify * it under the terms of the GNU General Public License as published by * the Free Software Foundation; either version 2 of the License, or * (at your option) any later version. * * This program is distributed in the hope that it will be useful, * but WITHOUT ANY WARRANTY; without even the implied warranty of * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the * GNU General Public License for more details. * * You should have received a copy of the GNU General Public License * along with this program; if not, write to the Free Software * Foundation, Inc., 51 Franklin Street, Fifth Floor, Boston, MA 02110-1301 USA * */ /* This file was automatically generated --- DO NOT EDIT */ /* Generated on Sun Nov 25 07:41:37 EST 2012 */ #include "codelet-rdft.h" #ifdef HAVE_FMA /* Generated by: ../../../genfft/gen_r2cb.native -fma -reorder-insns -schedule-for-pipeline -compact -variables 4 -pipeline-latency 4 -sign 1 -n 32 -name r2cbIII_32 -dft-III -include r2cbIII.h */ /* * This function contains 174 FP additions, 100 FP multiplications, * (or, 106 additions, 32 multiplications, 68 fused multiply/add), * 101 stack variables, 18 constants, and 64 memory accesses */ #include "r2cbIII.h" static void r2cbIII_32(R *R0, R *R1, R *Cr, R *Ci, stride rs, stride csr, stride csi, INT v, INT ivs, INT ovs) { DK(KP534511135, +0.534511135950791641089685961295362908582039528); DK(KP1_763842528, +1.763842528696710059425513727320776699016885241); DK(KP303346683, +0.303346683607342391675883946941299872384187453); DK(KP1_913880671, +1.913880671464417729871595773960539938965698411); DK(KP098491403, +0.098491403357164253077197521291327432293052451); DK(KP1_990369453, +1.990369453344393772489673906218959843150949737); DK(KP820678790, +0.820678790828660330972281985331011598767386482); DK(KP1_546020906, +1.546020906725473921621813219516939601942082586); DK(KP1_847759065, +1.847759065022573512256366378793576573644833252); DK(KP923879532, +0.923879532511286756128183189396788286822416626); DK(KP668178637, +0.668178637919298919997757686523080761552472251); DK(KP1_662939224, +1.662939224605090474157576755235811513477121624); DK(KP198912367, +0.198912367379658006911597622644676228597850501); DK(KP1_961570560, +1.961570560806460898252364472268478073947867462); DK(KP707106781, +0.707106781186547524400844362104849039284835938); DK(KP1_414213562, +1.414213562373095048801688724209698078569671875); DK(KP2_000000000, +2.000000000000000000000000000000000000000000000); DK(KP414213562, +0.414213562373095048801688724209698078569671875); { INT i; for (i = v; i > 0; i = i - 1, R0 = R0 + ovs, R1 = R1 + ovs, Cr = Cr + ivs, Ci = Ci + ivs, MAKE_VOLATILE_STRIDE(128, rs), MAKE_VOLATILE_STRIDE(128, csr), MAKE_VOLATILE_STRIDE(128, csi)) { E T1N, T1K, T1Q, T1H, T1O, T1P; { E T1I, T1e, T1Z, T7, T2E, T2i, T1x, Tz, Te, T2j, T22, T2F, T1h, T1y, TK; E T1J, Tm, T2B, TX, Tp, T2m, T28, T1M, T1C, T1k, TW, TY, T2a, T14, T15; E Ts, TZ; { E TE, T1g, TJ, T1f; { E T4, Tv, T3, T2g, T1d, T5, Tw, Tx; { E T1, T2, T1b, T1c; T1 = Cr[0]; T2 = Cr[WS(csr, 15)]; T1b = Ci[0]; T1c = Ci[WS(csi, 15)]; T4 = Cr[WS(csr, 8)]; Tv = T1 - T2; T3 = T1 + T2; T2g = T1c - T1b; T1d = T1b + T1c; T5 = Cr[WS(csr, 7)]; Tw = Ci[WS(csi, 8)]; Tx = Ci[WS(csi, 7)]; } { E Tb, TA, Ta, T20, TD, Tc, TG, TH; { E T8, T9, TB, TC; T8 = Cr[WS(csr, 4)]; { E T1a, T6, T2h, Ty; T1a = T4 - T5; T6 = T4 + T5; T2h = Tx - Tw; Ty = Tw + Tx; T1I = T1a - T1d; T1e = T1a + T1d; T1Z = T3 - T6; T7 = T3 + T6; T2E = T2h + T2g; T2i = T2g - T2h; T1x = Tv + Ty; Tz = Tv - Ty; T9 = Cr[WS(csr, 11)]; } TB = Ci[WS(csi, 4)]; TC = Ci[WS(csi, 11)]; Tb = Cr[WS(csr, 3)]; TA = T8 - T9; Ta = T8 + T9; T20 = TC - TB; TD = TB + TC; Tc = Cr[WS(csr, 12)]; TG = Ci[WS(csi, 3)]; TH = Ci[WS(csi, 12)]; } { E TF, Td, T21, TI; TE = TA - TD; T1g = TA + TD; TF = Tb - Tc; Td = Tb + Tc; T21 = TG - TH; TI = TG + TH; Te = Ta + Td; T2j = Ta - Td; T22 = T20 - T21; T2F = T20 + T21; TJ = TF - TI; T1f = TF + TI; } } } { E TM, Ti, TN, T25, TU, TR, Tl, TO; { E TS, TT, Tg, Th, Tj, Tk; Tg = Cr[WS(csr, 2)]; Th = Cr[WS(csr, 13)]; T1h = T1f - T1g; T1y = T1g + T1f; TK = TE + TJ; T1J = TE - TJ; TM = Tg - Th; Ti = Tg + Th; TS = Ci[WS(csi, 2)]; TT = Ci[WS(csi, 13)]; Tj = Cr[WS(csr, 10)]; Tk = Cr[WS(csr, 5)]; TN = Ci[WS(csi, 10)]; T25 = TS - TT; TU = TS + TT; TR = Tj - Tk; Tl = Tj + Tk; TO = Ci[WS(csi, 5)]; } { E T12, T13, Tq, Tr; { E Tn, T1A, TV, T24, T26, TP, To, T27, T1B, TQ; Tn = Cr[WS(csr, 1)]; T1A = TR - TU; TV = TR + TU; T24 = Ti - Tl; Tm = Ti + Tl; T26 = TN - TO; TP = TN + TO; To = Cr[WS(csr, 14)]; T12 = Ci[WS(csi, 1)]; T27 = T25 - T26; T2B = T26 + T25; T1B = TM + TP; TQ = TM - TP; TX = Tn - To; Tp = Tn + To; T2m = T24 + T27; T28 = T24 - T27; T1M = FNMS(KP414213562, T1A, T1B); T1C = FMA(KP414213562, T1B, T1A); T1k = FMA(KP414213562, TQ, TV); TW = FNMS(KP414213562, TV, TQ); T13 = Ci[WS(csi, 14)]; } Tq = Cr[WS(csr, 6)]; Tr = Cr[WS(csr, 9)]; TY = Ci[WS(csi, 6)]; T2a = T13 - T12; T14 = T12 + T13; T15 = Tq - Tr; Ts = Tq + Tr; TZ = Ci[WS(csi, 9)]; } } } { E T1L, T1F, T23, T2n, T2k, T2e, T1p, T1t, T1s, T1i, T1o, T19, T1l, T1q; { E T2z, T2G, T2H, T2C, T1j, T17, T2r, T2s, T2u, T2v, T2K, T2D; { E T2L, T2d, T2l, T2O; { E Tf, T2N, Tu, T2M; { E T1D, T16, T29, Tt, T2b, T10; T2z = T7 - Te; Tf = T7 + Te; T1D = T15 + T14; T16 = T14 - T15; T29 = Tp - Ts; Tt = Tp + Ts; T2b = TY - TZ; T10 = TY + TZ; T2N = T2F + T2E; T2G = T2E - T2F; T2H = Tm - Tt; Tu = Tm + Tt; { E T2c, T2A, T1E, T11; T2c = T2a - T2b; T2A = T2b + T2a; T1E = TX + T10; T11 = TX - T10; T2L = Tf - Tu; T2d = T29 + T2c; T2l = T29 - T2c; T2C = T2A - T2B; T2M = T2B + T2A; T1L = FMA(KP414213562, T1D, T1E); T1F = FNMS(KP414213562, T1E, T1D); T1j = FMA(KP414213562, T11, T16); T17 = FNMS(KP414213562, T16, T11); T2O = T2M + T2N; } } R0[0] = KP2_000000000 * (Tf + Tu); R0[WS(rs, 8)] = KP2_000000000 * (T2N - T2M); } T23 = T1Z + T22; T2r = T1Z - T22; R0[WS(rs, 12)] = KP1_414213562 * (T2O - T2L); R0[WS(rs, 4)] = KP1_414213562 * (T2L + T2O); T2s = T2m + T2l; T2n = T2l - T2m; T2k = T2i - T2j; T2u = T2j + T2i; T2v = T28 - T2d; T2e = T28 + T2d; } { E T2y, T2t, T2x, T2w; T2y = FMA(KP707106781, T2s, T2r); T2t = FNMS(KP707106781, T2s, T2r); T2x = FMA(KP707106781, T2v, T2u); T2w = FNMS(KP707106781, T2v, T2u); R0[WS(rs, 7)] = KP1_961570560 * (FMA(KP198912367, T2y, T2x)); R0[WS(rs, 15)] = -(KP1_961570560 * (FNMS(KP198912367, T2x, T2y))); R0[WS(rs, 11)] = KP1_662939224 * (FNMS(KP668178637, T2t, T2w)); R0[WS(rs, 3)] = KP1_662939224 * (FMA(KP668178637, T2w, T2t)); T2K = T2z - T2C; T2D = T2z + T2C; } { E TL, T18, T2J, T2I; T1p = FNMS(KP707106781, TK, Tz); TL = FMA(KP707106781, TK, Tz); T18 = TW + T17; T1t = TW - T17; T1s = FMA(KP707106781, T1h, T1e); T1i = FNMS(KP707106781, T1h, T1e); T2J = T2H + T2G; T2I = T2G - T2H; T1o = FNMS(KP923879532, T18, TL); T19 = FMA(KP923879532, T18, TL); R0[WS(rs, 6)] = KP1_847759065 * (FMA(KP414213562, T2K, T2J)); R0[WS(rs, 14)] = -(KP1_847759065 * (FNMS(KP414213562, T2J, T2K))); R0[WS(rs, 10)] = KP1_847759065 * (FNMS(KP414213562, T2D, T2I)); R0[WS(rs, 2)] = KP1_847759065 * (FMA(KP414213562, T2I, T2D)); T1l = T1j - T1k; T1q = T1k + T1j; } } { E T1z, T1U, T1Y, T1T, T1V, T1G; { E T1w, T1r, T1n, T1m; T1n = FMA(KP923879532, T1l, T1i); T1m = FNMS(KP923879532, T1l, T1i); T1w = FMA(KP923879532, T1q, T1p); T1r = FNMS(KP923879532, T1q, T1p); R1[WS(rs, 4)] = -(KP1_546020906 * (FNMS(KP820678790, T1o, T1n))); R1[WS(rs, 12)] = -(KP1_546020906 * (FMA(KP820678790, T1n, T1o))); R1[WS(rs, 8)] = -(KP1_990369453 * (FMA(KP098491403, T19, T1m))); R1[0] = KP1_990369453 * (FNMS(KP098491403, T1m, T19)); { E T1R, T1S, T1v, T1u; T1z = FNMS(KP707106781, T1y, T1x); T1R = FMA(KP707106781, T1y, T1x); T1S = T1M + T1L; T1N = T1L - T1M; T1K = FNMS(KP707106781, T1J, T1I); T1U = FMA(KP707106781, T1J, T1I); T1v = FNMS(KP923879532, T1t, T1s); T1u = FMA(KP923879532, T1t, T1s); T1Y = FMA(KP923879532, T1S, T1R); T1T = FNMS(KP923879532, T1S, T1R); R1[WS(rs, 6)] = -(KP1_913880671 * (FNMS(KP303346683, T1w, T1v))); R1[WS(rs, 14)] = -(KP1_913880671 * (FMA(KP303346683, T1v, T1w))); R1[WS(rs, 10)] = -(KP1_763842528 * (FMA(KP534511135, T1r, T1u))); R1[WS(rs, 2)] = KP1_763842528 * (FNMS(KP534511135, T1u, T1r)); T1V = T1C + T1F; T1G = T1C - T1F; } } { E T2q, T2f, T1X, T1W, T2p, T2o; T1X = FMA(KP923879532, T1V, T1U); T1W = FNMS(KP923879532, T1V, T1U); T2q = FNMS(KP707106781, T2e, T23); T2f = FMA(KP707106781, T2e, T23); R1[WS(rs, 7)] = KP1_990369453 * (FMA(KP098491403, T1Y, T1X)); R1[WS(rs, 15)] = -(KP1_990369453 * (FNMS(KP098491403, T1X, T1Y))); R1[WS(rs, 11)] = KP1_546020906 * (FNMS(KP820678790, T1T, T1W)); R1[WS(rs, 3)] = KP1_546020906 * (FMA(KP820678790, T1W, T1T)); T2p = FNMS(KP707106781, T2n, T2k); T2o = FMA(KP707106781, T2n, T2k); T1Q = FNMS(KP923879532, T1G, T1z); T1H = FMA(KP923879532, T1G, T1z); R0[WS(rs, 5)] = KP1_662939224 * (FMA(KP668178637, T2q, T2p)); R0[WS(rs, 13)] = -(KP1_662939224 * (FNMS(KP668178637, T2p, T2q))); R0[WS(rs, 9)] = KP1_961570560 * (FNMS(KP198912367, T2f, T2o)); R0[WS(rs, 1)] = KP1_961570560 * (FMA(KP198912367, T2o, T2f)); } } } } T1O = FMA(KP923879532, T1N, T1K); T1P = FNMS(KP923879532, T1N, T1K); R1[WS(rs, 5)] = KP1_763842528 * (FMA(KP534511135, T1Q, T1P)); R1[WS(rs, 13)] = -(KP1_763842528 * (FNMS(KP534511135, T1P, T1Q))); R1[WS(rs, 9)] = KP1_913880671 * (FNMS(KP303346683, T1H, T1O)); R1[WS(rs, 1)] = KP1_913880671 * (FMA(KP303346683, T1O, T1H)); } } } static const kr2c_desc desc = { 32, "r2cbIII_32", {106, 32, 68, 0}, &GENUS }; void X(codelet_r2cbIII_32) (planner *p) { X(kr2c_register) (p, r2cbIII_32, &desc); } #else /* HAVE_FMA */ /* Generated by: ../../../genfft/gen_r2cb.native -compact -variables 4 -pipeline-latency 4 -sign 1 -n 32 -name r2cbIII_32 -dft-III -include r2cbIII.h */ /* * This function contains 174 FP additions, 84 FP multiplications, * (or, 138 additions, 48 multiplications, 36 fused multiply/add), * 66 stack variables, 19 constants, and 64 memory accesses */ #include "r2cbIII.h" static void r2cbIII_32(R *R0, R *R1, R *Cr, R *Ci, stride rs, stride csr, stride csi, INT v, INT ivs, INT ovs) { DK(KP1_913880671, +1.913880671464417729871595773960539938965698411); DK(KP580569354, +0.580569354508924735272384751634790549382952557); DK(KP942793473, +0.942793473651995297112775251810508755314920638); DK(KP1_763842528, +1.763842528696710059425513727320776699016885241); DK(KP1_546020906, +1.546020906725473921621813219516939601942082586); DK(KP1_268786568, +1.268786568327290996430343226450986741351374190); DK(KP196034280, +0.196034280659121203988391127777283691722273346); DK(KP1_990369453, +1.990369453344393772489673906218959843150949737); DK(KP765366864, +0.765366864730179543456919968060797733522689125); DK(KP1_847759065, +1.847759065022573512256366378793576573644833252); DK(KP1_961570560, +1.961570560806460898252364472268478073947867462); DK(KP390180644, +0.390180644032256535696569736954044481855383236); DK(KP1_111140466, +1.111140466039204449485661627897065748749874382); DK(KP1_662939224, +1.662939224605090474157576755235811513477121624); DK(KP1_414213562, +1.414213562373095048801688724209698078569671875); DK(KP2_000000000, +2.000000000000000000000000000000000000000000000); DK(KP382683432, +0.382683432365089771728459984030398866761344562); DK(KP923879532, +0.923879532511286756128183189396788286822416626); DK(KP707106781, +0.707106781186547524400844362104849039284835938); { INT i; for (i = v; i > 0; i = i - 1, R0 = R0 + ovs, R1 = R1 + ovs, Cr = Cr + ivs, Ci = Ci + ivs, MAKE_VOLATILE_STRIDE(128, rs), MAKE_VOLATILE_STRIDE(128, csr), MAKE_VOLATILE_STRIDE(128, csi)) { E T7, T2i, T2F, Tz, T1k, T1I, T1Z, T1x, Te, T22, T2E, T2j, T1f, T1y, TK; E T1J, Tm, T2B, TW, T1a, T1C, T1L, T28, T2l, Tt, T2A, T17, T1b, T1F, T1M; E T2d, T2m; { E T3, Tv, T1j, T2h, T6, T1g, Ty, T2g; { E T1, T2, T1h, T1i; T1 = Cr[0]; T2 = Cr[WS(csr, 15)]; T3 = T1 + T2; Tv = T1 - T2; T1h = Ci[0]; T1i = Ci[WS(csi, 15)]; T1j = T1h + T1i; T2h = T1i - T1h; } { E T4, T5, Tw, Tx; T4 = Cr[WS(csr, 8)]; T5 = Cr[WS(csr, 7)]; T6 = T4 + T5; T1g = T4 - T5; Tw = Ci[WS(csi, 8)]; Tx = Ci[WS(csi, 7)]; Ty = Tw + Tx; T2g = Tw - Tx; } T7 = T3 + T6; T2i = T2g + T2h; T2F = T2h - T2g; Tz = Tv - Ty; T1k = T1g + T1j; T1I = T1g - T1j; T1Z = T3 - T6; T1x = Tv + Ty; } { E Ta, TA, TD, T21, Td, TF, TI, T20; { E T8, T9, TB, TC; T8 = Cr[WS(csr, 4)]; T9 = Cr[WS(csr, 11)]; Ta = T8 + T9; TA = T8 - T9; TB = Ci[WS(csi, 4)]; TC = Ci[WS(csi, 11)]; TD = TB + TC; T21 = TB - TC; } { E Tb, Tc, TG, TH; Tb = Cr[WS(csr, 3)]; Tc = Cr[WS(csr, 12)]; Td = Tb + Tc; TF = Tb - Tc; TG = Ci[WS(csi, 3)]; TH = Ci[WS(csi, 12)]; TI = TG + TH; T20 = TH - TG; } Te = Ta + Td; T22 = T20 - T21; T2E = T21 + T20; T2j = Ta - Td; { E T1d, T1e, TE, TJ; T1d = TA + TD; T1e = TF + TI; T1f = KP707106781 * (T1d - T1e); T1y = KP707106781 * (T1d + T1e); TE = TA - TD; TJ = TF - TI; TK = KP707106781 * (TE + TJ); T1J = KP707106781 * (TE - TJ); } } { E Ti, TM, TU, T25, Tl, TR, TP, T26, TQ, TV; { E Tg, Th, TS, TT; Tg = Cr[WS(csr, 2)]; Th = Cr[WS(csr, 13)]; Ti = Tg + Th; TM = Tg - Th; TS = Ci[WS(csi, 2)]; TT = Ci[WS(csi, 13)]; TU = TS + TT; T25 = TS - TT; } { E Tj, Tk, TN, TO; Tj = Cr[WS(csr, 10)]; Tk = Cr[WS(csr, 5)]; Tl = Tj + Tk; TR = Tj - Tk; TN = Ci[WS(csi, 10)]; TO = Ci[WS(csi, 5)]; TP = TN + TO; T26 = TN - TO; } Tm = Ti + Tl; T2B = T26 + T25; TQ = TM - TP; TV = TR + TU; TW = FNMS(KP382683432, TV, KP923879532 * TQ); T1a = FMA(KP382683432, TQ, KP923879532 * TV); { E T1A, T1B, T24, T27; T1A = TM + TP; T1B = TU - TR; T1C = FNMS(KP923879532, T1B, KP382683432 * T1A); T1L = FMA(KP923879532, T1A, KP382683432 * T1B); T24 = Ti - Tl; T27 = T25 - T26; T28 = T24 - T27; T2l = T24 + T27; } } { E Tp, TX, T15, T2a, Ts, T12, T10, T2b, T11, T16; { E Tn, To, T13, T14; Tn = Cr[WS(csr, 1)]; To = Cr[WS(csr, 14)]; Tp = Tn + To; TX = Tn - To; T13 = Ci[WS(csi, 1)]; T14 = Ci[WS(csi, 14)]; T15 = T13 + T14; T2a = T14 - T13; } { E Tq, Tr, TY, TZ; Tq = Cr[WS(csr, 6)]; Tr = Cr[WS(csr, 9)]; Ts = Tq + Tr; T12 = Tq - Tr; TY = Ci[WS(csi, 6)]; TZ = Ci[WS(csi, 9)]; T10 = TY + TZ; T2b = TY - TZ; } Tt = Tp + Ts; T2A = T2b + T2a; T11 = TX - T10; T16 = T12 - T15; T17 = FMA(KP923879532, T11, KP382683432 * T16); T1b = FNMS(KP382683432, T11, KP923879532 * T16); { E T1D, T1E, T29, T2c; T1D = TX + T10; T1E = T12 + T15; T1F = FNMS(KP923879532, T1E, KP382683432 * T1D); T1M = FMA(KP923879532, T1D, KP382683432 * T1E); T29 = Tp - Ts; T2c = T2a - T2b; T2d = T29 + T2c; T2m = T2c - T29; } } { E Tf, Tu, T2L, T2M, T2N, T2O; Tf = T7 + Te; Tu = Tm + Tt; T2L = Tf - Tu; T2M = T2B + T2A; T2N = T2F - T2E; T2O = T2M + T2N; R0[0] = KP2_000000000 * (Tf + Tu); R0[WS(rs, 8)] = KP2_000000000 * (T2N - T2M); R0[WS(rs, 4)] = KP1_414213562 * (T2L + T2O); R0[WS(rs, 12)] = KP1_414213562 * (T2O - T2L); } { E T2t, T2x, T2w, T2y; { E T2r, T2s, T2u, T2v; T2r = T1Z - T22; T2s = KP707106781 * (T2m - T2l); T2t = T2r + T2s; T2x = T2r - T2s; T2u = T2j + T2i; T2v = KP707106781 * (T28 - T2d); T2w = T2u - T2v; T2y = T2v + T2u; } R0[WS(rs, 3)] = FMA(KP1_662939224, T2t, KP1_111140466 * T2w); R0[WS(rs, 15)] = FNMS(KP1_961570560, T2x, KP390180644 * T2y); R0[WS(rs, 11)] = FNMS(KP1_111140466, T2t, KP1_662939224 * T2w); R0[WS(rs, 7)] = FMA(KP390180644, T2x, KP1_961570560 * T2y); } { E T2D, T2J, T2I, T2K; { E T2z, T2C, T2G, T2H; T2z = T7 - Te; T2C = T2A - T2B; T2D = T2z + T2C; T2J = T2z - T2C; T2G = T2E + T2F; T2H = Tm - Tt; T2I = T2G - T2H; T2K = T2H + T2G; } R0[WS(rs, 2)] = FMA(KP1_847759065, T2D, KP765366864 * T2I); R0[WS(rs, 14)] = FNMS(KP1_847759065, T2J, KP765366864 * T2K); R0[WS(rs, 10)] = FNMS(KP765366864, T2D, KP1_847759065 * T2I); R0[WS(rs, 6)] = FMA(KP765366864, T2J, KP1_847759065 * T2K); } { E T19, T1n, T1m, T1o; { E TL, T18, T1c, T1l; TL = Tz + TK; T18 = TW + T17; T19 = TL + T18; T1n = TL - T18; T1c = T1a + T1b; T1l = T1f + T1k; T1m = T1c + T1l; T1o = T1c - T1l; } R1[0] = FNMS(KP196034280, T1m, KP1_990369453 * T19); R1[WS(rs, 12)] = FNMS(KP1_546020906, T1n, KP1_268786568 * T1o); R1[WS(rs, 8)] = -(FMA(KP196034280, T19, KP1_990369453 * T1m)); R1[WS(rs, 4)] = FMA(KP1_268786568, T1n, KP1_546020906 * T1o); } { E T1r, T1v, T1u, T1w; { E T1p, T1q, T1s, T1t; T1p = Tz - TK; T1q = T1b - T1a; T1r = T1p + T1q; T1v = T1p - T1q; T1s = T1f - T1k; T1t = TW - T17; T1u = T1s - T1t; T1w = T1t + T1s; } R1[WS(rs, 2)] = FMA(KP1_763842528, T1r, KP942793473 * T1u); R1[WS(rs, 14)] = FNMS(KP1_913880671, T1v, KP580569354 * T1w); R1[WS(rs, 10)] = FNMS(KP942793473, T1r, KP1_763842528 * T1u); R1[WS(rs, 6)] = FMA(KP580569354, T1v, KP1_913880671 * T1w); } { E T1T, T1X, T1W, T1Y; { E T1R, T1S, T1U, T1V; T1R = T1x + T1y; T1S = T1L + T1M; T1T = T1R - T1S; T1X = T1R + T1S; T1U = T1J + T1I; T1V = T1C - T1F; T1W = T1U - T1V; T1Y = T1V + T1U; } R1[WS(rs, 3)] = FMA(KP1_546020906, T1T, KP1_268786568 * T1W); R1[WS(rs, 15)] = FNMS(KP1_990369453, T1X, KP196034280 * T1Y); R1[WS(rs, 11)] = FNMS(KP1_268786568, T1T, KP1_546020906 * T1W); R1[WS(rs, 7)] = FMA(KP196034280, T1X, KP1_990369453 * T1Y); } { E T2f, T2p, T2o, T2q; { E T23, T2e, T2k, T2n; T23 = T1Z + T22; T2e = KP707106781 * (T28 + T2d); T2f = T23 + T2e; T2p = T23 - T2e; T2k = T2i - T2j; T2n = KP707106781 * (T2l + T2m); T2o = T2k - T2n; T2q = T2n + T2k; } R0[WS(rs, 1)] = FMA(KP1_961570560, T2f, KP390180644 * T2o); R0[WS(rs, 13)] = FNMS(KP1_662939224, T2p, KP1_111140466 * T2q); R0[WS(rs, 9)] = FNMS(KP390180644, T2f, KP1_961570560 * T2o); R0[WS(rs, 5)] = FMA(KP1_111140466, T2p, KP1_662939224 * T2q); } { E T1H, T1P, T1O, T1Q; { E T1z, T1G, T1K, T1N; T1z = T1x - T1y; T1G = T1C + T1F; T1H = T1z + T1G; T1P = T1z - T1G; T1K = T1I - T1J; T1N = T1L - T1M; T1O = T1K - T1N; T1Q = T1N + T1K; } R1[WS(rs, 1)] = FMA(KP1_913880671, T1H, KP580569354 * T1O); R1[WS(rs, 13)] = FNMS(KP1_763842528, T1P, KP942793473 * T1Q); R1[WS(rs, 9)] = FNMS(KP580569354, T1H, KP1_913880671 * T1O); R1[WS(rs, 5)] = FMA(KP942793473, T1P, KP1_763842528 * T1Q); } } } } static const kr2c_desc desc = { 32, "r2cbIII_32", {138, 48, 36, 0}, &GENUS }; void X(codelet_r2cbIII_32) (planner *p) { X(kr2c_register) (p, r2cbIII_32, &desc); } #endif /* HAVE_FMA */ fftw-3.3.3/rdft/scalar/r2cb/r2cbIII_64.c0000644000175000017500000014006712054411037014310 00000000000000/* * Copyright (c) 2003, 2007-11 Matteo Frigo * Copyright (c) 2003, 2007-11 Massachusetts Institute of Technology * * This program is free software; you can redistribute it and/or modify * it under the terms of the GNU General Public License as published by * the Free Software Foundation; either version 2 of the License, or * (at your option) any later version. * * This program is distributed in the hope that it will be useful, * but WITHOUT ANY WARRANTY; without even the implied warranty of * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the * GNU General Public License for more details. * * You should have received a copy of the GNU General Public License * along with this program; if not, write to the Free Software * Foundation, Inc., 51 Franklin Street, Fifth Floor, Boston, MA 02110-1301 USA * */ /* This file was automatically generated --- DO NOT EDIT */ /* Generated on Sun Nov 25 07:41:38 EST 2012 */ #include "codelet-rdft.h" #ifdef HAVE_FMA /* Generated by: ../../../genfft/gen_r2cb.native -fma -reorder-insns -schedule-for-pipeline -compact -variables 4 -pipeline-latency 4 -sign 1 -n 64 -name r2cbIII_64 -dft-III -include r2cbIII.h */ /* * This function contains 434 FP additions, 260 FP multiplications, * (or, 238 additions, 64 multiplications, 196 fused multiply/add), * 165 stack variables, 36 constants, and 128 memory accesses */ #include "r2cbIII.h" static void r2cbIII_64(R *R0, R *R1, R *Cr, R *Ci, stride rs, stride csr, stride csi, INT v, INT ivs, INT ovs) { DK(KP357805721, +0.357805721314524104672487743774474392487532769); DK(KP1_883088130, +1.883088130366041556825018805199004714371179592); DK(KP472964775, +0.472964775891319928124438237972992463904131113); DK(KP1_807978586, +1.807978586246886663172400594461074097420264050); DK(KP049126849, +0.049126849769467254105343321271313617079695752); DK(KP1_997590912, +1.997590912410344785429543209518201388886407229); DK(KP906347169, +0.906347169019147157946142717268914412664134293); DK(KP1_481902250, +1.481902250709918182351233794990325459457910619); DK(KP980785280, +0.980785280403230449126182236134239036973933731); DK(KP250486960, +0.250486960191305461595702160124721208578685568); DK(KP1_940062506, +1.940062506389087985207968414572200502913731924); DK(KP599376933, +0.599376933681923766271389869014404232837890546); DK(KP1_715457220, +1.715457220000544139804539968569540274084981599); DK(KP148335987, +0.148335987538347428753676511486911367000625355); DK(KP1_978353019, +1.978353019929561946903347476032486127967379067); DK(KP741650546, +0.741650546272035369581266691172079863842265220); DK(KP1_606415062, +1.606415062961289819613353025926283847759138854); DK(KP831469612, +0.831469612302545237078788377617905756738560812); DK(KP303346683, +0.303346683607342391675883946941299872384187453); DK(KP1_913880671, +1.913880671464417729871595773960539938965698411); DK(KP534511135, +0.534511135950791641089685961295362908582039528); DK(KP1_763842528, +1.763842528696710059425513727320776699016885241); DK(KP098491403, +0.098491403357164253077197521291327432293052451); DK(KP1_990369453, +1.990369453344393772489673906218959843150949737); DK(KP820678790, +0.820678790828660330972281985331011598767386482); DK(KP1_546020906, +1.546020906725473921621813219516939601942082586); DK(KP923879532, +0.923879532511286756128183189396788286822416626); DK(KP198912367, +0.198912367379658006911597622644676228597850501); DK(KP1_961570560, +1.961570560806460898252364472268478073947867462); DK(KP668178637, +0.668178637919298919997757686523080761552472251); DK(KP1_662939224, +1.662939224605090474157576755235811513477121624); DK(KP1_847759065, +1.847759065022573512256366378793576573644833252); DK(KP1_414213562, +1.414213562373095048801688724209698078569671875); DK(KP2_000000000, +2.000000000000000000000000000000000000000000000); DK(KP707106781, +0.707106781186547524400844362104849039284835938); DK(KP414213562, +0.414213562373095048801688724209698078569671875); { INT i; for (i = v; i > 0; i = i - 1, R0 = R0 + ovs, R1 = R1 + ovs, Cr = Cr + ivs, Ci = Ci + ivs, MAKE_VOLATILE_STRIDE(256, rs), MAKE_VOLATILE_STRIDE(256, csr), MAKE_VOLATILE_STRIDE(256, csi)) { E T43, T4b, T49, T4e, T3T, T46, T40, T4a; { E T3t, T15, T2E, T3U, T6b, Tf, T6Q, T6u, T5J, T4L, T3V, T1g, T5U, T5q, T3u; E T2H, T6v, Tu, T5r, T4V, T6R, T6e, T2K, T1s, T2J, T1D, T3X, T3B, T5s, T4Q; E T3Y, T3y, T6g, TK, T5M, T57, T6N, T6j, T35, T1W, T34, T25, T4i, T3J, T5N; E T52, T4j, T3G, T6l, TZ, T3L, T5P, T5i, T6M, T6o, T3M, T38, T2n, T37, T2w; E T4l, T3Q, T5Q, T5d; { E T3x, T3w, T3E, T3F; { E T5p, T5o, T2G, T2F; { E T11, T3, T5m, T2D, T2A, T6, T5n, T14, Tb, T16, Ta, T4I, T19, Tc, T1c; E T1d; { E T4, T5, T12, T13; { E T1, T2, T2B, T2C; T1 = Cr[0]; T2 = Cr[WS(csr, 31)]; T2B = Ci[0]; T2C = Ci[WS(csi, 31)]; T4 = Cr[WS(csr, 16)]; T11 = T1 - T2; T3 = T1 + T2; T5m = T2C - T2B; T2D = T2B + T2C; T5 = Cr[WS(csr, 15)]; T12 = Ci[WS(csi, 16)]; T13 = Ci[WS(csi, 15)]; } { E T8, T9, T17, T18; T8 = Cr[WS(csr, 8)]; T2A = T4 - T5; T6 = T4 + T5; T5n = T13 - T12; T14 = T12 + T13; T9 = Cr[WS(csr, 23)]; T17 = Ci[WS(csi, 8)]; T18 = Ci[WS(csi, 23)]; Tb = Cr[WS(csr, 7)]; T16 = T8 - T9; Ta = T8 + T9; T4I = T18 - T17; T19 = T17 + T18; Tc = Cr[WS(csr, 24)]; T1c = Ci[WS(csi, 7)]; T1d = Ci[WS(csi, 24)]; } } { E T1b, T4J, T1e, T4H, T7, Te, Td; T3t = T11 + T14; T15 = T11 - T14; T1b = Tb - Tc; Td = Tb + Tc; T4J = T1c - T1d; T1e = T1c + T1d; T2E = T2A + T2D; T3U = T2A - T2D; T4H = T3 - T6; T7 = T3 + T6; Te = Ta + Td; T5p = Ta - Td; { E T4K, T6s, T6t, T1a, T1f; T5o = T5m - T5n; T6s = T5n + T5m; T6t = T4I + T4J; T4K = T4I - T4J; T6b = T7 - Te; Tf = T7 + Te; T6Q = T6t + T6s; T6u = T6s - T6t; T2G = T16 + T19; T1a = T16 - T19; T1f = T1b - T1e; T2F = T1b + T1e; T5J = T4H - T4K; T4L = T4H + T4K; T3V = T1a - T1f; T1g = T1a + T1f; } } } { E T1i, Ti, T4O, T1q, T1n, Tl, T4N, T1l, Tq, T1t, Tp, T4T, T1A, Tr, T1u; E T1v; { E Tj, Tk, T1j, T1k; { E Tg, Th, T1o, T1p; Tg = Cr[WS(csr, 4)]; T5U = T5p + T5o; T5q = T5o - T5p; T3u = T2G + T2F; T2H = T2F - T2G; Th = Cr[WS(csr, 27)]; T1o = Ci[WS(csi, 4)]; T1p = Ci[WS(csi, 27)]; Tj = Cr[WS(csr, 20)]; T1i = Tg - Th; Ti = Tg + Th; T4O = T1p - T1o; T1q = T1o + T1p; Tk = Cr[WS(csr, 11)]; T1j = Ci[WS(csi, 20)]; T1k = Ci[WS(csi, 11)]; } { E Tn, To, T1y, T1z; Tn = Cr[WS(csr, 3)]; T1n = Tj - Tk; Tl = Tj + Tk; T4N = T1k - T1j; T1l = T1j + T1k; To = Cr[WS(csr, 28)]; T1y = Ci[WS(csi, 3)]; T1z = Ci[WS(csi, 28)]; Tq = Cr[WS(csr, 12)]; T1t = Tn - To; Tp = Tn + To; T4T = T1y - T1z; T1A = T1y + T1z; Tr = Cr[WS(csr, 19)]; T1u = Ci[WS(csi, 12)]; T1v = Ci[WS(csi, 19)]; } } { E T4M, T1B, T1w, T4P, T1m, T1r, Tm, Ts, T4S; T4M = Ti - Tl; Tm = Ti + Tl; T1B = Tq - Tr; Ts = Tq + Tr; T4S = T1v - T1u; T1w = T1u + T1v; { E T6c, Tt, T4R, T6d, T4U; T6c = T4N + T4O; T4P = T4N - T4O; Tt = Tp + Ts; T4R = Tp - Ts; T6d = T4S + T4T; T4U = T4S - T4T; T3x = T1i + T1l; T1m = T1i - T1l; T6v = Tm - Tt; Tu = Tm + Tt; T5r = T4R - T4U; T4V = T4R + T4U; T6R = T6c + T6d; T6e = T6c - T6d; T1r = T1n + T1q; T3w = T1n - T1q; } { E T3A, T3z, T1x, T1C; T3A = T1t + T1w; T1x = T1t - T1w; T1C = T1A - T1B; T3z = T1B + T1A; T2K = FMA(KP414213562, T1m, T1r); T1s = FNMS(KP414213562, T1r, T1m); T2J = FMA(KP414213562, T1x, T1C); T1D = FNMS(KP414213562, T1C, T1x); T3X = FMA(KP414213562, T3z, T3A); T3B = FNMS(KP414213562, T3A, T3z); T5s = T4M + T4P; T4Q = T4M - T4P; } } } } { E T1G, Ty, T54, T20, T1X, TB, T53, T1J, TI, T4Z, T1L, TF, T22, T1U, T50; E T1O; { E T1Y, T1Z, Tz, TA, Tw, Tx, T1H, T1I; Tw = Cr[WS(csr, 2)]; Tx = Cr[WS(csr, 29)]; T1Y = Ci[WS(csi, 2)]; T3Y = FNMS(KP414213562, T3w, T3x); T3y = FMA(KP414213562, T3x, T3w); T1G = Tw - Tx; Ty = Tw + Tx; T1Z = Ci[WS(csi, 29)]; Tz = Cr[WS(csr, 18)]; TA = Cr[WS(csr, 13)]; T1H = Ci[WS(csi, 18)]; T54 = T1Y - T1Z; T20 = T1Y + T1Z; T1X = Tz - TA; TB = Tz + TA; T1I = Ci[WS(csi, 13)]; { E T1R, T1Q, T1S, TG, TH; TG = Cr[WS(csr, 5)]; TH = Cr[WS(csr, 26)]; T1R = Ci[WS(csi, 5)]; T53 = T1H - T1I; T1J = T1H + T1I; T1Q = TG - TH; TI = TG + TH; T1S = Ci[WS(csi, 26)]; { E T1M, T1N, TD, TE, T1T; TD = Cr[WS(csr, 10)]; TE = Cr[WS(csr, 21)]; T1T = T1R + T1S; T4Z = T1S - T1R; T1M = Ci[WS(csi, 10)]; T1L = TD - TE; TF = TD + TE; T1N = Ci[WS(csi, 21)]; T22 = T1Q + T1T; T1U = T1Q - T1T; T50 = T1M - T1N; T1O = T1M + T1N; } } } { E T4Y, T23, T51, T1K, T1V, T3I, T3H, T21, T24; { E T56, T1P, T6h, T55, TC, TJ, T6i; T4Y = Ty - TB; TC = Ty + TB; TJ = TF + TI; T56 = TF - TI; T1P = T1L - T1O; T23 = T1L + T1O; T6h = T53 + T54; T55 = T53 - T54; T6g = TC - TJ; TK = TC + TJ; T6i = T50 + T4Z; T51 = T4Z - T50; T3E = T1G + T1J; T1K = T1G - T1J; T5M = T56 + T55; T57 = T55 - T56; T6N = T6i + T6h; T6j = T6h - T6i; T1V = T1P + T1U; T3I = T1P - T1U; } T3H = T1X - T20; T21 = T1X + T20; T24 = T22 - T23; T3F = T23 + T22; T35 = FNMS(KP707106781, T1V, T1K); T1W = FMA(KP707106781, T1V, T1K); T34 = FMA(KP707106781, T24, T21); T25 = FNMS(KP707106781, T24, T21); T4i = FMA(KP707106781, T3I, T3H); T3J = FNMS(KP707106781, T3I, T3H); T5N = T4Y - T51; T52 = T4Y + T51; } } { E T27, TN, T5f, T2q, T2r, TQ, T5e, T2a, TX, T5a, T2c, TU, T2t, T2l, T5b; E T2f; { E T2o, T2p, TO, TP, TL, TM, T28, T29; TL = Cr[WS(csr, 1)]; TM = Cr[WS(csr, 30)]; T2o = Ci[WS(csi, 1)]; T4j = FMA(KP707106781, T3F, T3E); T3G = FNMS(KP707106781, T3F, T3E); T27 = TL - TM; TN = TL + TM; T2p = Ci[WS(csi, 30)]; TO = Cr[WS(csr, 14)]; TP = Cr[WS(csr, 17)]; T28 = Ci[WS(csi, 14)]; T5f = T2p - T2o; T2q = T2o + T2p; T2r = TO - TP; TQ = TO + TP; T29 = Ci[WS(csi, 17)]; { E T2i, T2h, T2j, TV, TW; TV = Cr[WS(csr, 9)]; TW = Cr[WS(csr, 22)]; T2i = Ci[WS(csi, 9)]; T5e = T28 - T29; T2a = T28 + T29; T2h = TV - TW; TX = TV + TW; T2j = Ci[WS(csi, 22)]; { E T2d, T2e, TS, TT, T2k; TS = Cr[WS(csr, 6)]; TT = Cr[WS(csr, 25)]; T2k = T2i + T2j; T5a = T2j - T2i; T2d = Ci[WS(csi, 6)]; T2c = TS - TT; TU = TS + TT; T2e = Ci[WS(csi, 25)]; T2t = T2h + T2k; T2l = T2h - T2k; T5b = T2d - T2e; T2f = T2d + T2e; } } } { E T59, T2u, T5c, T2b, T2m, T3P, T3O, T2s, T2v; { E T5h, T2g, T6m, T5g, TR, TY, T6n; T59 = TN - TQ; TR = TN + TQ; TY = TU + TX; T5h = TU - TX; T2g = T2c - T2f; T2u = T2c + T2f; T6m = T5e + T5f; T5g = T5e - T5f; T6l = TR - TY; TZ = TR + TY; T6n = T5b + T5a; T5c = T5a - T5b; T3L = T27 + T2a; T2b = T27 - T2a; T5P = T5h + T5g; T5i = T5g - T5h; T6M = T6n + T6m; T6o = T6m - T6n; T2m = T2g + T2l; T3P = T2g - T2l; } T3O = T2r + T2q; T2s = T2q - T2r; T2v = T2t - T2u; T3M = T2u + T2t; T38 = FNMS(KP707106781, T2m, T2b); T2n = FMA(KP707106781, T2m, T2b); T37 = FNMS(KP707106781, T2v, T2s); T2w = FMA(KP707106781, T2v, T2s); T4l = FMA(KP707106781, T3P, T3O); T3Q = FNMS(KP707106781, T3P, T3O); T5Q = T59 - T5c; T5d = T59 + T5c; } } } { E T4m, T3N, T5t, T5L, T63, T4W, T5Y, T5X, T66, T5W, T67, T5S; { E T6T, T6S, T6W, T6P; { E T6L, T6O, T6Y, T6X, T6Z, Tv, T10, T70; T6L = Tf - Tu; Tv = Tf + Tu; T10 = TK + TZ; T6T = TK - TZ; T6O = T6M - T6N; T6Y = T6N + T6M; T4m = FMA(KP707106781, T3M, T3L); T3N = FNMS(KP707106781, T3M, T3L); T6X = Tv - T10; T6S = T6Q - T6R; T6Z = T6R + T6Q; R0[0] = KP2_000000000 * (Tv + T10); R0[WS(rs, 16)] = KP2_000000000 * (T6Z - T6Y); T70 = T6Y + T6Z; T6W = T6L - T6O; T6P = T6L + T6O; R0[WS(rs, 24)] = KP1_414213562 * (T70 - T6X); R0[WS(rs, 8)] = KP1_414213562 * (T6X + T70); } { E T6D, T6f, T6w, T6G, T6p, T6x, T6y, T6k, T6V, T6U; T6D = T6b - T6e; T6f = T6b + T6e; T6w = T6u - T6v; T6G = T6v + T6u; T6V = T6T + T6S; T6U = T6S - T6T; T6p = T6l + T6o; T6x = T6l - T6o; R0[WS(rs, 12)] = KP1_847759065 * (FMA(KP414213562, T6W, T6V)); R0[WS(rs, 28)] = -(KP1_847759065 * (FNMS(KP414213562, T6V, T6W))); R0[WS(rs, 20)] = KP1_847759065 * (FNMS(KP414213562, T6P, T6U)); R0[WS(rs, 4)] = KP1_847759065 * (FMA(KP414213562, T6U, T6P)); T6y = T6g + T6j; T6k = T6g - T6j; { E T5V, T5K, T5O, T5R; T5t = T5r - T5s; T5K = T5s + T5r; { E T6E, T6z, T6H, T6q; T6E = T6y + T6x; T6z = T6x - T6y; T6H = T6k - T6p; T6q = T6k + T6p; { E T6F, T6K, T6B, T6A; T6F = FNMS(KP707106781, T6E, T6D); T6K = FMA(KP707106781, T6E, T6D); T6B = FNMS(KP707106781, T6z, T6w); T6A = FMA(KP707106781, T6z, T6w); { E T6I, T6J, T6C, T6r; T6I = FNMS(KP707106781, T6H, T6G); T6J = FMA(KP707106781, T6H, T6G); T6C = FNMS(KP707106781, T6q, T6f); T6r = FMA(KP707106781, T6q, T6f); R0[WS(rs, 22)] = KP1_662939224 * (FNMS(KP668178637, T6F, T6I)); R0[WS(rs, 6)] = KP1_662939224 * (FMA(KP668178637, T6I, T6F)); R0[WS(rs, 30)] = -(KP1_961570560 * (FNMS(KP198912367, T6J, T6K))); R0[WS(rs, 14)] = KP1_961570560 * (FMA(KP198912367, T6K, T6J)); R0[WS(rs, 26)] = -(KP1_662939224 * (FNMS(KP668178637, T6B, T6C))); R0[WS(rs, 10)] = KP1_662939224 * (FMA(KP668178637, T6C, T6B)); R0[WS(rs, 18)] = KP1_961570560 * (FNMS(KP198912367, T6r, T6A)); R0[WS(rs, 2)] = KP1_961570560 * (FMA(KP198912367, T6A, T6r)); T5L = FNMS(KP707106781, T5K, T5J); T63 = FMA(KP707106781, T5K, T5J); } } } T5V = T4Q - T4V; T4W = T4Q + T4V; T5Y = FNMS(KP414213562, T5M, T5N); T5O = FMA(KP414213562, T5N, T5M); T5R = FNMS(KP414213562, T5Q, T5P); T5X = FMA(KP414213562, T5P, T5Q); T66 = FMA(KP707106781, T5V, T5U); T5W = FNMS(KP707106781, T5V, T5U); T67 = T5O + T5R; T5S = T5O - T5R; } } } { E T1h, T2L, T2I, T3h, T3p, T1E, T3n, T3s, T3b, T3k, T3e, T3o; { E T4X, T5B, T5v, T5w, T5E, T5u, T5F, T5k, T58, T5j; { E T68, T69, T62, T5T, T64, T5Z; T68 = FNMS(KP923879532, T67, T66); T69 = FMA(KP923879532, T67, T66); T62 = FNMS(KP923879532, T5S, T5L); T5T = FMA(KP923879532, T5S, T5L); T64 = T5Y + T5X; T5Z = T5X - T5Y; T4X = FMA(KP707106781, T4W, T4L); T5B = FNMS(KP707106781, T4W, T4L); { E T65, T6a, T61, T60; T65 = FNMS(KP923879532, T64, T63); T6a = FMA(KP923879532, T64, T63); T61 = FNMS(KP923879532, T5Z, T5W); T60 = FMA(KP923879532, T5Z, T5W); R0[WS(rs, 23)] = KP1_546020906 * (FNMS(KP820678790, T65, T68)); R0[WS(rs, 7)] = KP1_546020906 * (FMA(KP820678790, T68, T65)); R0[WS(rs, 31)] = -(KP1_990369453 * (FNMS(KP098491403, T69, T6a))); R0[WS(rs, 15)] = KP1_990369453 * (FMA(KP098491403, T6a, T69)); R0[WS(rs, 27)] = -(KP1_763842528 * (FNMS(KP534511135, T61, T62))); R0[WS(rs, 11)] = KP1_763842528 * (FMA(KP534511135, T62, T61)); R0[WS(rs, 19)] = KP1_913880671 * (FNMS(KP303346683, T5T, T60)); R0[WS(rs, 3)] = KP1_913880671 * (FMA(KP303346683, T60, T5T)); } } T5v = FNMS(KP414213562, T52, T57); T58 = FMA(KP414213562, T57, T52); T5j = FNMS(KP414213562, T5i, T5d); T5w = FMA(KP414213562, T5d, T5i); T5E = FNMS(KP707106781, T5t, T5q); T5u = FMA(KP707106781, T5t, T5q); T5F = T58 - T5j; T5k = T58 + T5j; { E T3l, T33, T3c, T3m, T3a, T3d; { E T39, T3f, T3g, T36; { E T31, T5G, T5H, T5A, T5l, T5C, T5x, T32; T1h = FMA(KP707106781, T1g, T15); T31 = FNMS(KP707106781, T1g, T15); T5G = FNMS(KP923879532, T5F, T5E); T5H = FMA(KP923879532, T5F, T5E); T5A = FNMS(KP923879532, T5k, T4X); T5l = FMA(KP923879532, T5k, T4X); T5C = T5w - T5v; T5x = T5v + T5w; T32 = T2K + T2J; T2L = T2J - T2K; T39 = FNMS(KP668178637, T38, T37); T3f = FMA(KP668178637, T37, T38); { E T5D, T5I, T5z, T5y; T5D = FNMS(KP923879532, T5C, T5B); T5I = FMA(KP923879532, T5C, T5B); T5z = FNMS(KP923879532, T5x, T5u); T5y = FMA(KP923879532, T5x, T5u); T3l = FMA(KP923879532, T32, T31); T33 = FNMS(KP923879532, T32, T31); R0[WS(rs, 21)] = KP1_763842528 * (FNMS(KP534511135, T5D, T5G)); R0[WS(rs, 5)] = KP1_763842528 * (FMA(KP534511135, T5G, T5D)); R0[WS(rs, 29)] = -(KP1_913880671 * (FNMS(KP303346683, T5H, T5I))); R0[WS(rs, 13)] = KP1_913880671 * (FMA(KP303346683, T5I, T5H)); R0[WS(rs, 25)] = -(KP1_546020906 * (FNMS(KP820678790, T5z, T5A))); R0[WS(rs, 9)] = KP1_546020906 * (FMA(KP820678790, T5A, T5z)); R0[WS(rs, 17)] = KP1_990369453 * (FNMS(KP098491403, T5l, T5y)); R0[WS(rs, 1)] = KP1_990369453 * (FMA(KP098491403, T5y, T5l)); T3g = FMA(KP668178637, T34, T35); T36 = FNMS(KP668178637, T35, T34); } } T2I = FNMS(KP707106781, T2H, T2E); T3c = FMA(KP707106781, T2H, T2E); T3m = T3g + T3f; T3h = T3f - T3g; T3p = T39 - T36; T3a = T36 + T39; T3d = T1s - T1D; T1E = T1s + T1D; } T3n = FNMS(KP831469612, T3m, T3l); T3s = FMA(KP831469612, T3m, T3l); T3b = FNMS(KP831469612, T3a, T33); T3k = FMA(KP831469612, T3a, T33); T3e = FMA(KP923879532, T3d, T3c); T3o = FNMS(KP923879532, T3d, T3c); } } { E T3v, T3Z, T3W, T4v, T4D, T3C, T4B, T4G, T4p, T4y, T4s, T4C; { E T4z, T4h, T4q, T4A, T4o, T4r; { E T4n, T4t, T4u, T4k, T4f, T4g; T3v = FNMS(KP707106781, T3u, T3t); T4f = FMA(KP707106781, T3u, T3t); T4g = T3Y + T3X; T3Z = T3X - T3Y; { E T3r, T3q, T3i, T3j; T3r = FNMS(KP831469612, T3p, T3o); T3q = FMA(KP831469612, T3p, T3o); T3i = FNMS(KP831469612, T3h, T3e); T3j = FMA(KP831469612, T3h, T3e); R1[WS(rs, 22)] = -(KP1_606415062 * (FMA(KP741650546, T3n, T3q))); R1[WS(rs, 6)] = KP1_606415062 * (FNMS(KP741650546, T3q, T3n)); R1[WS(rs, 30)] = -(KP1_978353019 * (FMA(KP148335987, T3r, T3s))); R1[WS(rs, 14)] = -(KP1_978353019 * (FNMS(KP148335987, T3s, T3r))); R1[WS(rs, 26)] = -(KP1_715457220 * (FMA(KP599376933, T3j, T3k))); R1[WS(rs, 10)] = -(KP1_715457220 * (FNMS(KP599376933, T3k, T3j))); R1[WS(rs, 18)] = -(KP1_940062506 * (FMA(KP250486960, T3b, T3i))); R1[WS(rs, 2)] = KP1_940062506 * (FNMS(KP250486960, T3i, T3b)); T4z = FMA(KP923879532, T4g, T4f); T4h = FNMS(KP923879532, T4g, T4f); } T4n = FNMS(KP198912367, T4m, T4l); T4t = FMA(KP198912367, T4l, T4m); T4u = FNMS(KP198912367, T4i, T4j); T4k = FMA(KP198912367, T4j, T4i); T3W = FNMS(KP707106781, T3V, T3U); T4q = FMA(KP707106781, T3V, T3U); T4A = T4u + T4t; T4v = T4t - T4u; T4D = T4k + T4n; T4o = T4k - T4n; T4r = T3y + T3B; T3C = T3y - T3B; } T4B = FNMS(KP980785280, T4A, T4z); T4G = FMA(KP980785280, T4A, T4z); T4p = FMA(KP980785280, T4o, T4h); T4y = FNMS(KP980785280, T4o, T4h); T4s = FNMS(KP923879532, T4r, T4q); T4C = FMA(KP923879532, T4r, T4q); } { E T2P, T2X, T2V, T30, T2z, T2S, T2M, T2W; { E T2T, T1F, T2U, T2y; { E T2x, T2N, T2O, T26; { E T4F, T4E, T4w, T4x; T4F = FMA(KP980785280, T4D, T4C); T4E = FNMS(KP980785280, T4D, T4C); T4w = FMA(KP980785280, T4v, T4s); T4x = FNMS(KP980785280, T4v, T4s); R1[WS(rs, 23)] = KP1_481902250 * (FNMS(KP906347169, T4B, T4E)); R1[WS(rs, 7)] = KP1_481902250 * (FMA(KP906347169, T4E, T4B)); R1[WS(rs, 31)] = -(KP1_997590912 * (FNMS(KP049126849, T4F, T4G))); R1[WS(rs, 15)] = KP1_997590912 * (FMA(KP049126849, T4G, T4F)); R1[WS(rs, 27)] = -(KP1_807978586 * (FNMS(KP472964775, T4x, T4y))); R1[WS(rs, 11)] = KP1_807978586 * (FMA(KP472964775, T4y, T4x)); R1[WS(rs, 19)] = KP1_883088130 * (FNMS(KP357805721, T4p, T4w)); R1[WS(rs, 3)] = KP1_883088130 * (FMA(KP357805721, T4w, T4p)); T2T = FNMS(KP923879532, T1E, T1h); T1F = FMA(KP923879532, T1E, T1h); } T2x = FNMS(KP198912367, T2w, T2n); T2N = FMA(KP198912367, T2n, T2w); T2O = FMA(KP198912367, T1W, T25); T26 = FNMS(KP198912367, T25, T1W); T2U = T2O + T2N; T2P = T2N - T2O; T2X = T26 - T2x; T2y = T26 + T2x; } T2V = FNMS(KP980785280, T2U, T2T); T30 = FMA(KP980785280, T2U, T2T); T2z = FMA(KP980785280, T2y, T1F); T2S = FNMS(KP980785280, T2y, T1F); T2M = FNMS(KP923879532, T2L, T2I); T2W = FMA(KP923879532, T2L, T2I); } { E T47, T3D, T48, T3S; { E T3K, T41, T42, T3R; { E T2Z, T2Y, T2Q, T2R; T2Z = FNMS(KP980785280, T2X, T2W); T2Y = FMA(KP980785280, T2X, T2W); T2Q = FNMS(KP980785280, T2P, T2M); T2R = FMA(KP980785280, T2P, T2M); R1[WS(rs, 20)] = -(KP1_807978586 * (FMA(KP472964775, T2V, T2Y))); R1[WS(rs, 4)] = KP1_807978586 * (FNMS(KP472964775, T2Y, T2V)); R1[WS(rs, 28)] = -(KP1_883088130 * (FMA(KP357805721, T2Z, T30))); R1[WS(rs, 12)] = -(KP1_883088130 * (FNMS(KP357805721, T30, T2Z))); R1[WS(rs, 24)] = -(KP1_481902250 * (FMA(KP906347169, T2R, T2S))); R1[WS(rs, 8)] = -(KP1_481902250 * (FNMS(KP906347169, T2S, T2R))); R1[WS(rs, 16)] = -(KP1_997590912 * (FMA(KP049126849, T2z, T2Q))); R1[0] = KP1_997590912 * (FNMS(KP049126849, T2Q, T2z)); T47 = FNMS(KP923879532, T3C, T3v); T3D = FMA(KP923879532, T3C, T3v); } T3K = FMA(KP668178637, T3J, T3G); T41 = FNMS(KP668178637, T3G, T3J); T42 = FMA(KP668178637, T3N, T3Q); T3R = FNMS(KP668178637, T3Q, T3N); T48 = T42 - T41; T43 = T41 + T42; T4b = T3K - T3R; T3S = T3K + T3R; } T49 = FNMS(KP831469612, T48, T47); T4e = FMA(KP831469612, T48, T47); T3T = FMA(KP831469612, T3S, T3D); T46 = FNMS(KP831469612, T3S, T3D); T40 = FMA(KP923879532, T3Z, T3W); T4a = FNMS(KP923879532, T3Z, T3W); } } } } } } { E T4d, T4c, T44, T45; T4d = FMA(KP831469612, T4b, T4a); T4c = FNMS(KP831469612, T4b, T4a); T44 = FMA(KP831469612, T43, T40); T45 = FNMS(KP831469612, T43, T40); R1[WS(rs, 21)] = KP1_715457220 * (FNMS(KP599376933, T49, T4c)); R1[WS(rs, 5)] = KP1_715457220 * (FMA(KP599376933, T4c, T49)); R1[WS(rs, 29)] = -(KP1_940062506 * (FNMS(KP250486960, T4d, T4e))); R1[WS(rs, 13)] = KP1_940062506 * (FMA(KP250486960, T4e, T4d)); R1[WS(rs, 25)] = -(KP1_606415062 * (FNMS(KP741650546, T45, T46))); R1[WS(rs, 9)] = KP1_606415062 * (FMA(KP741650546, T46, T45)); R1[WS(rs, 17)] = KP1_978353019 * (FNMS(KP148335987, T3T, T44)); R1[WS(rs, 1)] = KP1_978353019 * (FMA(KP148335987, T44, T3T)); } } } } static const kr2c_desc desc = { 64, "r2cbIII_64", {238, 64, 196, 0}, &GENUS }; void X(codelet_r2cbIII_64) (planner *p) { X(kr2c_register) (p, r2cbIII_64, &desc); } #else /* HAVE_FMA */ /* Generated by: ../../../genfft/gen_r2cb.native -compact -variables 4 -pipeline-latency 4 -sign 1 -n 64 -name r2cbIII_64 -dft-III -include r2cbIII.h */ /* * This function contains 434 FP additions, 208 FP multiplications, * (or, 342 additions, 116 multiplications, 92 fused multiply/add), * 130 stack variables, 39 constants, and 128 memory accesses */ #include "r2cbIII.h" static void r2cbIII_64(R *R0, R *R1, R *Cr, R *Ci, stride rs, stride csr, stride csi, INT v, INT ivs, INT ovs) { DK(KP1_343117909, +1.343117909694036801250753700854843606457501264); DK(KP1_481902250, +1.481902250709918182351233794990325459457910619); DK(KP1_807978586, +1.807978586246886663172400594461074097420264050); DK(KP855110186, +0.855110186860564188641933713777597068609157259); DK(KP1_997590912, +1.997590912410344785429543209518201388886407229); DK(KP098135348, +0.098135348654836028509909953885365316629490726); DK(KP673779706, +0.673779706784440101378506425238295140955533559); DK(KP1_883088130, +1.883088130366041556825018805199004714371179592); DK(KP195090322, +0.195090322016128267848284868477022240927691618); DK(KP980785280, +0.980785280403230449126182236134239036973933731); DK(KP1_191398608, +1.191398608984866686934073057659939779023852677); DK(KP1_606415062, +1.606415062961289819613353025926283847759138854); DK(KP1_715457220, +1.715457220000544139804539968569540274084981599); DK(KP1_028205488, +1.028205488386443453187387677937631545216098241); DK(KP1_978353019, +1.978353019929561946903347476032486127967379067); DK(KP293460948, +0.293460948910723503317700259293435639412430633); DK(KP485960359, +0.485960359806527779896548324154942236641981567); DK(KP1_940062506, +1.940062506389087985207968414572200502913731924); DK(KP555570233, +0.555570233019602224742830813948532874374937191); DK(KP831469612, +0.831469612302545237078788377617905756738560812); DK(KP1_268786568, +1.268786568327290996430343226450986741351374190); DK(KP1_546020906, +1.546020906725473921621813219516939601942082586); DK(KP1_763842528, +1.763842528696710059425513727320776699016885241); DK(KP942793473, +0.942793473651995297112775251810508755314920638); DK(KP1_990369453, +1.990369453344393772489673906218959843150949737); DK(KP196034280, +0.196034280659121203988391127777283691722273346); DK(KP580569354, +0.580569354508924735272384751634790549382952557); DK(KP1_913880671, +1.913880671464417729871595773960539938965698411); DK(KP1_662939224, +1.662939224605090474157576755235811513477121624); DK(KP1_111140466, +1.111140466039204449485661627897065748749874382); DK(KP390180644, +0.390180644032256535696569736954044481855383236); DK(KP1_961570560, +1.961570560806460898252364472268478073947867462); DK(KP765366864, +0.765366864730179543456919968060797733522689125); DK(KP1_847759065, +1.847759065022573512256366378793576573644833252); DK(KP1_414213562, +1.414213562373095048801688724209698078569671875); DK(KP2_000000000, +2.000000000000000000000000000000000000000000000); DK(KP382683432, +0.382683432365089771728459984030398866761344562); DK(KP923879532, +0.923879532511286756128183189396788286822416626); DK(KP707106781, +0.707106781186547524400844362104849039284835938); { INT i; for (i = v; i > 0; i = i - 1, R0 = R0 + ovs, R1 = R1 + ovs, Cr = Cr + ivs, Ci = Ci + ivs, MAKE_VOLATILE_STRIDE(256, rs), MAKE_VOLATILE_STRIDE(256, csr), MAKE_VOLATILE_STRIDE(256, csi)) { E T15, T3t, T3U, T2N, Tf, T6b, T6u, T6R, T4L, T5J, T1g, T3V, T5q, T5U, T2I; E T3u, Tu, T6v, T4V, T5s, T6e, T6Q, T1s, T2D, T1D, T2E, T3B, T3Y, T4Q, T5r; E T3y, T3X, TK, T6g, T57, T5N, T6j, T6N, T1W, T34, T25, T35, T3J, T4j, T52; E T5M, T3G, T4i, TZ, T6l, T5i, T5Q, T6o, T6M, T2n, T37, T2w, T38, T3Q, T4m; E T5d, T5P, T3N, T4l; { E T3, T11, T2M, T5n, T6, T2J, T14, T5m, Ta, T16, T19, T4J, Td, T1b, T1e; E T4I; { E T1, T2, T2K, T2L; T1 = Cr[0]; T2 = Cr[WS(csr, 31)]; T3 = T1 + T2; T11 = T1 - T2; T2K = Ci[0]; T2L = Ci[WS(csi, 31)]; T2M = T2K + T2L; T5n = T2L - T2K; } { E T4, T5, T12, T13; T4 = Cr[WS(csr, 16)]; T5 = Cr[WS(csr, 15)]; T6 = T4 + T5; T2J = T4 - T5; T12 = Ci[WS(csi, 16)]; T13 = Ci[WS(csi, 15)]; T14 = T12 + T13; T5m = T12 - T13; } { E T8, T9, T17, T18; T8 = Cr[WS(csr, 8)]; T9 = Cr[WS(csr, 23)]; Ta = T8 + T9; T16 = T8 - T9; T17 = Ci[WS(csi, 8)]; T18 = Ci[WS(csi, 23)]; T19 = T17 + T18; T4J = T17 - T18; } { E Tb, Tc, T1c, T1d; Tb = Cr[WS(csr, 7)]; Tc = Cr[WS(csr, 24)]; Td = Tb + Tc; T1b = Tb - Tc; T1c = Ci[WS(csi, 7)]; T1d = Ci[WS(csi, 24)]; T1e = T1c + T1d; T4I = T1d - T1c; } { E T7, Te, T1a, T1f; T15 = T11 - T14; T3t = T11 + T14; T3U = T2J - T2M; T2N = T2J + T2M; T7 = T3 + T6; Te = Ta + Td; Tf = T7 + Te; T6b = T7 - Te; { E T6s, T6t, T4H, T4K; T6s = T4J + T4I; T6t = T5n - T5m; T6u = T6s + T6t; T6R = T6t - T6s; T4H = T3 - T6; T4K = T4I - T4J; T4L = T4H + T4K; T5J = T4H - T4K; } T1a = T16 - T19; T1f = T1b - T1e; T1g = KP707106781 * (T1a + T1f); T3V = KP707106781 * (T1a - T1f); { E T5o, T5p, T2G, T2H; T5o = T5m + T5n; T5p = Ta - Td; T5q = T5o - T5p; T5U = T5p + T5o; T2G = T16 + T19; T2H = T1b + T1e; T2I = KP707106781 * (T2G - T2H); T3u = KP707106781 * (T2G + T2H); } } } { E Ti, T1i, T1q, T4N, Tl, T1n, T1l, T4O, Tp, T1t, T1B, T4S, Ts, T1y, T1w; E T4T; { E Tg, Th, T1o, T1p; Tg = Cr[WS(csr, 4)]; Th = Cr[WS(csr, 27)]; Ti = Tg + Th; T1i = Tg - Th; T1o = Ci[WS(csi, 4)]; T1p = Ci[WS(csi, 27)]; T1q = T1o + T1p; T4N = T1o - T1p; } { E Tj, Tk, T1j, T1k; Tj = Cr[WS(csr, 20)]; Tk = Cr[WS(csr, 11)]; Tl = Tj + Tk; T1n = Tj - Tk; T1j = Ci[WS(csi, 20)]; T1k = Ci[WS(csi, 11)]; T1l = T1j + T1k; T4O = T1j - T1k; } { E Tn, To, T1z, T1A; Tn = Cr[WS(csr, 3)]; To = Cr[WS(csr, 28)]; Tp = Tn + To; T1t = Tn - To; T1z = Ci[WS(csi, 3)]; T1A = Ci[WS(csi, 28)]; T1B = T1z + T1A; T4S = T1A - T1z; } { E Tq, Tr, T1u, T1v; Tq = Cr[WS(csr, 12)]; Tr = Cr[WS(csr, 19)]; Ts = Tq + Tr; T1y = Tq - Tr; T1u = Ci[WS(csi, 12)]; T1v = Ci[WS(csi, 19)]; T1w = T1u + T1v; T4T = T1u - T1v; } { E Tm, Tt, T4R, T4U; Tm = Ti + Tl; Tt = Tp + Ts; Tu = Tm + Tt; T6v = Tm - Tt; T4R = Tp - Ts; T4U = T4S - T4T; T4V = T4R + T4U; T5s = T4U - T4R; } { E T6c, T6d, T1m, T1r; T6c = T4T + T4S; T6d = T4O + T4N; T6e = T6c - T6d; T6Q = T6d + T6c; T1m = T1i - T1l; T1r = T1n + T1q; T1s = FNMS(KP382683432, T1r, KP923879532 * T1m); T2D = FMA(KP382683432, T1m, KP923879532 * T1r); } { E T1x, T1C, T3z, T3A; T1x = T1t - T1w; T1C = T1y - T1B; T1D = FMA(KP923879532, T1x, KP382683432 * T1C); T2E = FNMS(KP382683432, T1x, KP923879532 * T1C); T3z = T1t + T1w; T3A = T1y + T1B; T3B = FNMS(KP923879532, T3A, KP382683432 * T3z); T3Y = FMA(KP923879532, T3z, KP382683432 * T3A); } { E T4M, T4P, T3w, T3x; T4M = Ti - Tl; T4P = T4N - T4O; T4Q = T4M - T4P; T5r = T4M + T4P; T3w = T1i + T1l; T3x = T1q - T1n; T3y = FNMS(KP923879532, T3x, KP382683432 * T3w); T3X = FMA(KP923879532, T3w, KP382683432 * T3x); } } { E Ty, T1G, T23, T54, TB, T20, T1J, T55, TI, T4Z, T1U, T1Y, TF, T50, T1P; E T1X; { E Tw, Tx, T1H, T1I; Tw = Cr[WS(csr, 2)]; Tx = Cr[WS(csr, 29)]; Ty = Tw + Tx; T1G = Tw - Tx; { E T21, T22, Tz, TA; T21 = Ci[WS(csi, 2)]; T22 = Ci[WS(csi, 29)]; T23 = T21 + T22; T54 = T21 - T22; Tz = Cr[WS(csr, 18)]; TA = Cr[WS(csr, 13)]; TB = Tz + TA; T20 = Tz - TA; } T1H = Ci[WS(csi, 18)]; T1I = Ci[WS(csi, 13)]; T1J = T1H + T1I; T55 = T1H - T1I; { E TG, TH, T1Q, T1R, T1S, T1T; TG = Cr[WS(csr, 5)]; TH = Cr[WS(csr, 26)]; T1Q = TG - TH; T1R = Ci[WS(csi, 5)]; T1S = Ci[WS(csi, 26)]; T1T = T1R + T1S; TI = TG + TH; T4Z = T1S - T1R; T1U = T1Q - T1T; T1Y = T1Q + T1T; } { E TD, TE, T1L, T1M, T1N, T1O; TD = Cr[WS(csr, 10)]; TE = Cr[WS(csr, 21)]; T1L = TD - TE; T1M = Ci[WS(csi, 10)]; T1N = Ci[WS(csi, 21)]; T1O = T1M + T1N; TF = TD + TE; T50 = T1M - T1N; T1P = T1L - T1O; T1X = T1L + T1O; } } { E TC, TJ, T53, T56; TC = Ty + TB; TJ = TF + TI; TK = TC + TJ; T6g = TC - TJ; T53 = TF - TI; T56 = T54 - T55; T57 = T53 + T56; T5N = T56 - T53; } { E T6h, T6i, T1K, T1V; T6h = T55 + T54; T6i = T50 + T4Z; T6j = T6h - T6i; T6N = T6i + T6h; T1K = T1G - T1J; T1V = KP707106781 * (T1P + T1U); T1W = T1K + T1V; T34 = T1K - T1V; } { E T1Z, T24, T3H, T3I; T1Z = KP707106781 * (T1X - T1Y); T24 = T20 + T23; T25 = T1Z + T24; T35 = T24 - T1Z; T3H = KP707106781 * (T1P - T1U); T3I = T23 - T20; T3J = T3H + T3I; T4j = T3I - T3H; } { E T4Y, T51, T3E, T3F; T4Y = Ty - TB; T51 = T4Z - T50; T52 = T4Y + T51; T5M = T4Y - T51; T3E = T1G + T1J; T3F = KP707106781 * (T1X + T1Y); T3G = T3E - T3F; T4i = T3E + T3F; } } { E TN, T27, T2u, T5f, TQ, T2r, T2a, T5g, TX, T5a, T2l, T2p, TU, T5b, T2g; E T2o; { E TL, TM, T28, T29; TL = Cr[WS(csr, 1)]; TM = Cr[WS(csr, 30)]; TN = TL + TM; T27 = TL - TM; { E T2s, T2t, TO, TP; T2s = Ci[WS(csi, 1)]; T2t = Ci[WS(csi, 30)]; T2u = T2s + T2t; T5f = T2t - T2s; TO = Cr[WS(csr, 14)]; TP = Cr[WS(csr, 17)]; TQ = TO + TP; T2r = TO - TP; } T28 = Ci[WS(csi, 14)]; T29 = Ci[WS(csi, 17)]; T2a = T28 + T29; T5g = T28 - T29; { E TV, TW, T2h, T2i, T2j, T2k; TV = Cr[WS(csr, 9)]; TW = Cr[WS(csr, 22)]; T2h = TV - TW; T2i = Ci[WS(csi, 9)]; T2j = Ci[WS(csi, 22)]; T2k = T2i + T2j; TX = TV + TW; T5a = T2j - T2i; T2l = T2h - T2k; T2p = T2h + T2k; } { E TS, TT, T2c, T2d, T2e, T2f; TS = Cr[WS(csr, 6)]; TT = Cr[WS(csr, 25)]; T2c = TS - TT; T2d = Ci[WS(csi, 6)]; T2e = Ci[WS(csi, 25)]; T2f = T2d + T2e; TU = TS + TT; T5b = T2d - T2e; T2g = T2c - T2f; T2o = T2c + T2f; } } { E TR, TY, T5e, T5h; TR = TN + TQ; TY = TU + TX; TZ = TR + TY; T6l = TR - TY; T5e = TU - TX; T5h = T5f - T5g; T5i = T5e + T5h; T5Q = T5h - T5e; } { E T6m, T6n, T2b, T2m; T6m = T5g + T5f; T6n = T5b + T5a; T6o = T6m - T6n; T6M = T6n + T6m; T2b = T27 - T2a; T2m = KP707106781 * (T2g + T2l); T2n = T2b + T2m; T37 = T2b - T2m; } { E T2q, T2v, T3O, T3P; T2q = KP707106781 * (T2o - T2p); T2v = T2r - T2u; T2w = T2q + T2v; T38 = T2v - T2q; T3O = KP707106781 * (T2g - T2l); T3P = T2r + T2u; T3Q = T3O - T3P; T4m = T3O + T3P; } { E T59, T5c, T3L, T3M; T59 = TN - TQ; T5c = T5a - T5b; T5d = T59 + T5c; T5P = T59 - T5c; T3L = T27 + T2a; T3M = KP707106781 * (T2o + T2p); T3N = T3L - T3M; T4l = T3L + T3M; } } { E Tv, T10, T6X, T6Y, T6Z, T70; Tv = Tf + Tu; T10 = TK + TZ; T6X = Tv - T10; T6Y = T6N + T6M; T6Z = T6R - T6Q; T70 = T6Y + T6Z; R0[0] = KP2_000000000 * (Tv + T10); R0[WS(rs, 16)] = KP2_000000000 * (T6Z - T6Y); R0[WS(rs, 8)] = KP1_414213562 * (T6X + T70); R0[WS(rs, 24)] = KP1_414213562 * (T70 - T6X); } { E T6P, T6V, T6U, T6W; { E T6L, T6O, T6S, T6T; T6L = Tf - Tu; T6O = T6M - T6N; T6P = T6L + T6O; T6V = T6L - T6O; T6S = T6Q + T6R; T6T = TK - TZ; T6U = T6S - T6T; T6W = T6T + T6S; } R0[WS(rs, 4)] = FMA(KP1_847759065, T6P, KP765366864 * T6U); R0[WS(rs, 28)] = FNMS(KP1_847759065, T6V, KP765366864 * T6W); R0[WS(rs, 20)] = FNMS(KP765366864, T6P, KP1_847759065 * T6U); R0[WS(rs, 12)] = FMA(KP765366864, T6V, KP1_847759065 * T6W); } { E T6f, T6w, T6G, T6D, T6z, T6E, T6q, T6H; T6f = T6b + T6e; T6w = T6u - T6v; T6G = T6v + T6u; T6D = T6b - T6e; { E T6x, T6y, T6k, T6p; T6x = T6g + T6j; T6y = T6o - T6l; T6z = KP707106781 * (T6x + T6y); T6E = KP707106781 * (T6y - T6x); T6k = T6g - T6j; T6p = T6l + T6o; T6q = KP707106781 * (T6k + T6p); T6H = KP707106781 * (T6k - T6p); } { E T6r, T6A, T6J, T6K; T6r = T6f + T6q; T6A = T6w - T6z; R0[WS(rs, 2)] = FMA(KP1_961570560, T6r, KP390180644 * T6A); R0[WS(rs, 18)] = FNMS(KP390180644, T6r, KP1_961570560 * T6A); T6J = T6D - T6E; T6K = T6H + T6G; R0[WS(rs, 14)] = FMA(KP390180644, T6J, KP1_961570560 * T6K); R0[WS(rs, 30)] = FNMS(KP1_961570560, T6J, KP390180644 * T6K); } { E T6B, T6C, T6F, T6I; T6B = T6f - T6q; T6C = T6z + T6w; R0[WS(rs, 10)] = FMA(KP1_111140466, T6B, KP1_662939224 * T6C); R0[WS(rs, 26)] = FNMS(KP1_662939224, T6B, KP1_111140466 * T6C); T6F = T6D + T6E; T6I = T6G - T6H; R0[WS(rs, 6)] = FMA(KP1_662939224, T6F, KP1_111140466 * T6I); R0[WS(rs, 22)] = FNMS(KP1_111140466, T6F, KP1_662939224 * T6I); } } { E T5L, T63, T5W, T66, T5S, T67, T5Z, T64, T5K, T5V; T5K = KP707106781 * (T5s - T5r); T5L = T5J + T5K; T63 = T5J - T5K; T5V = KP707106781 * (T4Q - T4V); T5W = T5U - T5V; T66 = T5V + T5U; { E T5O, T5R, T5X, T5Y; T5O = FNMS(KP923879532, T5N, KP382683432 * T5M); T5R = FMA(KP382683432, T5P, KP923879532 * T5Q); T5S = T5O + T5R; T67 = T5O - T5R; T5X = FMA(KP923879532, T5M, KP382683432 * T5N); T5Y = FNMS(KP923879532, T5P, KP382683432 * T5Q); T5Z = T5X + T5Y; T64 = T5Y - T5X; } { E T5T, T60, T69, T6a; T5T = T5L + T5S; T60 = T5W - T5Z; R0[WS(rs, 3)] = FMA(KP1_913880671, T5T, KP580569354 * T60); R0[WS(rs, 19)] = FNMS(KP580569354, T5T, KP1_913880671 * T60); T69 = T63 - T64; T6a = T67 + T66; R0[WS(rs, 15)] = FMA(KP196034280, T69, KP1_990369453 * T6a); R0[WS(rs, 31)] = FNMS(KP1_990369453, T69, KP196034280 * T6a); } { E T61, T62, T65, T68; T61 = T5L - T5S; T62 = T5Z + T5W; R0[WS(rs, 11)] = FMA(KP942793473, T61, KP1_763842528 * T62); R0[WS(rs, 27)] = FNMS(KP1_763842528, T61, KP942793473 * T62); T65 = T63 + T64; T68 = T66 - T67; R0[WS(rs, 7)] = FMA(KP1_546020906, T65, KP1_268786568 * T68); R0[WS(rs, 23)] = FNMS(KP1_268786568, T65, KP1_546020906 * T68); } } { E T4X, T5B, T5u, T5E, T5k, T5F, T5x, T5C, T4W, T5t; T4W = KP707106781 * (T4Q + T4V); T4X = T4L + T4W; T5B = T4L - T4W; T5t = KP707106781 * (T5r + T5s); T5u = T5q - T5t; T5E = T5t + T5q; { E T58, T5j, T5v, T5w; T58 = FNMS(KP382683432, T57, KP923879532 * T52); T5j = FMA(KP923879532, T5d, KP382683432 * T5i); T5k = T58 + T5j; T5F = T58 - T5j; T5v = FMA(KP382683432, T52, KP923879532 * T57); T5w = FNMS(KP382683432, T5d, KP923879532 * T5i); T5x = T5v + T5w; T5C = T5w - T5v; } { E T5l, T5y, T5H, T5I; T5l = T4X + T5k; T5y = T5u - T5x; R0[WS(rs, 1)] = FMA(KP1_990369453, T5l, KP196034280 * T5y); R0[WS(rs, 17)] = FNMS(KP196034280, T5l, KP1_990369453 * T5y); T5H = T5B - T5C; T5I = T5F + T5E; R0[WS(rs, 13)] = FMA(KP580569354, T5H, KP1_913880671 * T5I); R0[WS(rs, 29)] = FNMS(KP1_913880671, T5H, KP580569354 * T5I); } { E T5z, T5A, T5D, T5G; T5z = T4X - T5k; T5A = T5x + T5u; R0[WS(rs, 9)] = FMA(KP1_268786568, T5z, KP1_546020906 * T5A); R0[WS(rs, 25)] = FNMS(KP1_546020906, T5z, KP1_268786568 * T5A); T5D = T5B + T5C; T5G = T5E - T5F; R0[WS(rs, 5)] = FMA(KP1_763842528, T5D, KP942793473 * T5G); R0[WS(rs, 21)] = FNMS(KP942793473, T5D, KP1_763842528 * T5G); } } { E T33, T3l, T3h, T3m, T3a, T3p, T3e, T3o; { E T31, T32, T3f, T3g; T31 = T15 - T1g; T32 = T2E - T2D; T33 = T31 + T32; T3l = T31 - T32; T3f = FMA(KP831469612, T34, KP555570233 * T35); T3g = FNMS(KP831469612, T37, KP555570233 * T38); T3h = T3f + T3g; T3m = T3g - T3f; } { E T36, T39, T3c, T3d; T36 = FNMS(KP831469612, T35, KP555570233 * T34); T39 = FMA(KP555570233, T37, KP831469612 * T38); T3a = T36 + T39; T3p = T36 - T39; T3c = T2I - T2N; T3d = T1s - T1D; T3e = T3c - T3d; T3o = T3d + T3c; } { E T3b, T3i, T3r, T3s; T3b = T33 + T3a; T3i = T3e - T3h; R1[WS(rs, 2)] = FMA(KP1_940062506, T3b, KP485960359 * T3i); R1[WS(rs, 18)] = FNMS(KP485960359, T3b, KP1_940062506 * T3i); T3r = T3l - T3m; T3s = T3p + T3o; R1[WS(rs, 14)] = FMA(KP293460948, T3r, KP1_978353019 * T3s); R1[WS(rs, 30)] = FNMS(KP1_978353019, T3r, KP293460948 * T3s); } { E T3j, T3k, T3n, T3q; T3j = T33 - T3a; T3k = T3h + T3e; R1[WS(rs, 10)] = FMA(KP1_028205488, T3j, KP1_715457220 * T3k); R1[WS(rs, 26)] = FNMS(KP1_715457220, T3j, KP1_028205488 * T3k); T3n = T3l + T3m; T3q = T3o - T3p; R1[WS(rs, 6)] = FMA(KP1_606415062, T3n, KP1_191398608 * T3q); R1[WS(rs, 22)] = FNMS(KP1_191398608, T3n, KP1_606415062 * T3q); } } { E T4h, T4z, T4v, T4A, T4o, T4D, T4s, T4C; { E T4f, T4g, T4t, T4u; T4f = T3t + T3u; T4g = T3X + T3Y; T4h = T4f - T4g; T4z = T4f + T4g; T4t = FMA(KP980785280, T4i, KP195090322 * T4j); T4u = FMA(KP980785280, T4l, KP195090322 * T4m); T4v = T4t - T4u; T4A = T4t + T4u; } { E T4k, T4n, T4q, T4r; T4k = FNMS(KP980785280, T4j, KP195090322 * T4i); T4n = FNMS(KP980785280, T4m, KP195090322 * T4l); T4o = T4k + T4n; T4D = T4k - T4n; T4q = T3V + T3U; T4r = T3y - T3B; T4s = T4q - T4r; T4C = T4r + T4q; } { E T4p, T4w, T4F, T4G; T4p = T4h + T4o; T4w = T4s - T4v; R1[WS(rs, 3)] = FMA(KP1_883088130, T4p, KP673779706 * T4w); R1[WS(rs, 19)] = FNMS(KP673779706, T4p, KP1_883088130 * T4w); T4F = T4z + T4A; T4G = T4D + T4C; R1[WS(rs, 15)] = FMA(KP098135348, T4F, KP1_997590912 * T4G); R1[WS(rs, 31)] = FNMS(KP1_997590912, T4F, KP098135348 * T4G); } { E T4x, T4y, T4B, T4E; T4x = T4h - T4o; T4y = T4v + T4s; R1[WS(rs, 11)] = FMA(KP855110186, T4x, KP1_807978586 * T4y); R1[WS(rs, 27)] = FNMS(KP1_807978586, T4x, KP855110186 * T4y); T4B = T4z - T4A; T4E = T4C - T4D; R1[WS(rs, 7)] = FMA(KP1_481902250, T4B, KP1_343117909 * T4E); R1[WS(rs, 23)] = FNMS(KP1_343117909, T4B, KP1_481902250 * T4E); } } { E T1F, T2T, T2P, T2W, T2y, T2X, T2C, T2U; { E T1h, T1E, T2F, T2O; T1h = T15 + T1g; T1E = T1s + T1D; T1F = T1h + T1E; T2T = T1h - T1E; T2F = T2D + T2E; T2O = T2I + T2N; T2P = T2F + T2O; T2W = T2F - T2O; } { E T26, T2x, T2A, T2B; T26 = FNMS(KP195090322, T25, KP980785280 * T1W); T2x = FMA(KP980785280, T2n, KP195090322 * T2w); T2y = T26 + T2x; T2X = T26 - T2x; T2A = FMA(KP195090322, T1W, KP980785280 * T25); T2B = FNMS(KP195090322, T2n, KP980785280 * T2w); T2C = T2A + T2B; T2U = T2B - T2A; } { E T2z, T2Q, T2Z, T30; T2z = T1F + T2y; T2Q = T2C + T2P; R1[0] = FNMS(KP098135348, T2Q, KP1_997590912 * T2z); R1[WS(rs, 16)] = -(FMA(KP098135348, T2z, KP1_997590912 * T2Q)); T2Z = T2T - T2U; T30 = T2X + T2W; R1[WS(rs, 12)] = FMA(KP673779706, T2Z, KP1_883088130 * T30); R1[WS(rs, 28)] = FNMS(KP1_883088130, T2Z, KP673779706 * T30); } { E T2R, T2S, T2V, T2Y; T2R = T1F - T2y; T2S = T2C - T2P; R1[WS(rs, 8)] = FMA(KP1_343117909, T2R, KP1_481902250 * T2S); R1[WS(rs, 24)] = FNMS(KP1_481902250, T2R, KP1_343117909 * T2S); T2V = T2T + T2U; T2Y = T2W - T2X; R1[WS(rs, 4)] = FMA(KP1_807978586, T2V, KP855110186 * T2Y); R1[WS(rs, 20)] = FNMS(KP855110186, T2V, KP1_807978586 * T2Y); } } { E T3D, T47, T43, T48, T3S, T4b, T40, T4a; { E T3v, T3C, T41, T42; T3v = T3t - T3u; T3C = T3y + T3B; T3D = T3v + T3C; T47 = T3v - T3C; T41 = FMA(KP555570233, T3G, KP831469612 * T3J); T42 = FNMS(KP555570233, T3N, KP831469612 * T3Q); T43 = T41 + T42; T48 = T42 - T41; } { E T3K, T3R, T3W, T3Z; T3K = FNMS(KP555570233, T3J, KP831469612 * T3G); T3R = FMA(KP831469612, T3N, KP555570233 * T3Q); T3S = T3K + T3R; T4b = T3K - T3R; T3W = T3U - T3V; T3Z = T3X - T3Y; T40 = T3W - T3Z; T4a = T3Z + T3W; } { E T3T, T44, T4d, T4e; T3T = T3D + T3S; T44 = T40 - T43; R1[WS(rs, 1)] = FMA(KP1_978353019, T3T, KP293460948 * T44); R1[WS(rs, 17)] = FNMS(KP293460948, T3T, KP1_978353019 * T44); T4d = T47 - T48; T4e = T4b + T4a; R1[WS(rs, 13)] = FMA(KP485960359, T4d, KP1_940062506 * T4e); R1[WS(rs, 29)] = FNMS(KP1_940062506, T4d, KP485960359 * T4e); } { E T45, T46, T49, T4c; T45 = T3D - T3S; T46 = T43 + T40; R1[WS(rs, 9)] = FMA(KP1_191398608, T45, KP1_606415062 * T46); R1[WS(rs, 25)] = FNMS(KP1_606415062, T45, KP1_191398608 * T46); T49 = T47 + T48; T4c = T4a - T4b; R1[WS(rs, 5)] = FMA(KP1_715457220, T49, KP1_028205488 * T4c); R1[WS(rs, 21)] = FNMS(KP1_028205488, T49, KP1_715457220 * T4c); } } } } } static const kr2c_desc desc = { 64, "r2cbIII_64", {342, 116, 92, 0}, &GENUS }; void X(codelet_r2cbIII_64) (planner *p) { X(kr2c_register) (p, r2cbIII_64, &desc); } #endif /* HAVE_FMA */ fftw-3.3.3/rdft/scalar/r2cb/r2cbIII_20.c0000644000175000017500000003023312054411011014261 00000000000000/* * Copyright (c) 2003, 2007-11 Matteo Frigo * Copyright (c) 2003, 2007-11 Massachusetts Institute of Technology * * This program is free software; you can redistribute it and/or modify * it under the terms of the GNU General Public License as published by * the Free Software Foundation; either version 2 of the License, or * (at your option) any later version. * * This program is distributed in the hope that it will be useful, * but WITHOUT ANY WARRANTY; without even the implied warranty of * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the * GNU General Public License for more details. * * You should have received a copy of the GNU General Public License * along with this program; if not, write to the Free Software * Foundation, Inc., 51 Franklin Street, Fifth Floor, Boston, MA 02110-1301 USA * */ /* This file was automatically generated --- DO NOT EDIT */ /* Generated on Sun Nov 25 07:41:44 EST 2012 */ #include "codelet-rdft.h" #ifdef HAVE_FMA /* Generated by: ../../../genfft/gen_r2cb.native -fma -reorder-insns -schedule-for-pipeline -compact -variables 4 -pipeline-latency 4 -sign 1 -n 20 -name r2cbIII_20 -dft-III -include r2cbIII.h */ /* * This function contains 94 FP additions, 56 FP multiplications, * (or, 58 additions, 20 multiplications, 36 fused multiply/add), * 59 stack variables, 6 constants, and 40 memory accesses */ #include "r2cbIII.h" static void r2cbIII_20(R *R0, R *R1, R *Cr, R *Ci, stride rs, stride csr, stride csi, INT v, INT ivs, INT ovs) { DK(KP1_414213562, +1.414213562373095048801688724209698078569671875); DK(KP951056516, +0.951056516295153572116439333379382143405698634); DK(KP559016994, +0.559016994374947424102293417182819058860154590); DK(KP2_000000000, +2.000000000000000000000000000000000000000000000); DK(KP250000000, +0.250000000000000000000000000000000000000000000); DK(KP618033988, +0.618033988749894848204586834365638117720309180); { INT i; for (i = v; i > 0; i = i - 1, R0 = R0 + ovs, R1 = R1 + ovs, Cr = Cr + ivs, Ci = Ci + ivs, MAKE_VOLATILE_STRIDE(80, rs), MAKE_VOLATILE_STRIDE(80, csr), MAKE_VOLATILE_STRIDE(80, csi)) { E TZ, TD, TW, Tw, Tt, TF, T1f, T1b; { E T1l, Tk, T9, Tj, Ta, TV, TI, Ts, TU, T1t, T11, Tx, T13, TC, T1a; E T1i, Th, Tv, Ty; { E TQ, TS, Tr, Tm, Tn; { E T1, T5, T6, T2, T3, T7, TY; T1 = Cr[WS(csr, 2)]; T5 = Cr[WS(csr, 9)]; T6 = Cr[WS(csr, 5)]; T2 = Cr[WS(csr, 6)]; T3 = Cr[WS(csr, 1)]; TQ = Ci[WS(csi, 2)]; T7 = T5 + T6; TY = T5 - T6; { E T4, TX, T8, Tp, Tq; T4 = T2 + T3; TX = T2 - T3; Tp = Ci[WS(csi, 5)]; Tq = Ci[WS(csi, 9)]; T1l = FNMS(KP618033988, TX, TY); TZ = FMA(KP618033988, TY, TX); Tk = T4 - T7; T8 = T4 + T7; TS = Tp + Tq; Tr = Tp - Tq; T9 = T1 + T8; Tj = FNMS(KP250000000, T8, T1); Tm = Ci[WS(csi, 6)]; Tn = Ci[WS(csi, 1)]; } } { E Tb, T19, Tg, Tc; Ta = Cr[WS(csr, 7)]; { E Te, Tf, To, TR, TT; Te = Cr[0]; Tf = Cr[WS(csr, 4)]; To = Tm + Tn; TR = Tm - Tn; Tb = Cr[WS(csr, 3)]; T19 = Te - Tf; Tg = Te + Tf; TT = TR - TS; TV = TR + TS; TI = FNMS(KP618033988, To, Tr); Ts = FMA(KP618033988, Tr, To); TU = FNMS(KP250000000, TT, TQ); T1t = TT + TQ; Tc = Cr[WS(csr, 8)]; } T11 = Ci[WS(csi, 7)]; { E TA, TB, Td, T18; TA = Ci[WS(csi, 4)]; TB = Ci[0]; Td = Tb + Tc; T18 = Tb - Tc; Tx = Ci[WS(csi, 3)]; T13 = TB + TA; TC = TA - TB; T1a = FMA(KP618033988, T19, T18); T1i = FNMS(KP618033988, T18, T19); Th = Td + Tg; Tv = Td - Tg; Ty = Ci[WS(csi, 8)]; } } } { E Tu, T1w, T16, TL, T15, T1u; { E Ti, T12, Tz, T14; Tu = FNMS(KP250000000, Th, Ta); Ti = Ta + Th; T12 = Tx - Ty; Tz = Tx + Ty; T1w = T9 - Ti; T14 = T12 - T13; T16 = T12 + T13; TL = FNMS(KP618033988, Tz, TC); TD = FMA(KP618033988, TC, Tz); T15 = FNMS(KP250000000, T14, T11); T1u = T14 + T11; R0[0] = KP2_000000000 * (T9 + Ti); } { E Tl, TJ, TN, T1q, T1m, TK, T1h, T17, TH, T1k, T1v; Tl = FMA(KP559016994, Tk, Tj); TH = FNMS(KP559016994, Tk, Tj); T1k = FNMS(KP559016994, TV, TU); TW = FMA(KP559016994, TV, TU); R0[WS(rs, 5)] = KP2_000000000 * (T1u - T1t); T1v = T1t + T1u; TJ = FNMS(KP951056516, TI, TH); TN = FMA(KP951056516, TI, TH); T1q = FMA(KP951056516, T1l, T1k); T1m = FNMS(KP951056516, T1l, T1k); R1[WS(rs, 7)] = KP1_414213562 * (T1w + T1v); R1[WS(rs, 2)] = KP1_414213562 * (T1v - T1w); Tw = FMA(KP559016994, Tv, Tu); TK = FNMS(KP559016994, Tv, Tu); T1h = FNMS(KP559016994, T16, T15); T17 = FMA(KP559016994, T16, T15); { E TM, TO, T1j, T1r; TM = FMA(KP951056516, TL, TK); TO = FNMS(KP951056516, TL, TK); T1j = FMA(KP951056516, T1i, T1h); T1r = FNMS(KP951056516, T1i, T1h); Tt = FNMS(KP951056516, Ts, Tl); TF = FMA(KP951056516, Ts, Tl); { E T1n, T1p, T1s, T1o; T1n = TN - TO; R0[WS(rs, 6)] = -(KP2_000000000 * (TN + TO)); T1p = TM - TJ; R0[WS(rs, 4)] = KP2_000000000 * (TJ + TM); T1s = T1q + T1r; R0[WS(rs, 9)] = KP2_000000000 * (T1r - T1q); T1o = T1m + T1j; R0[WS(rs, 1)] = KP2_000000000 * (T1j - T1m); R1[WS(rs, 6)] = KP1_414213562 * (T1p + T1s); R1[WS(rs, 1)] = KP1_414213562 * (T1p - T1s); R1[WS(rs, 3)] = KP1_414213562 * (T1n + T1o); R1[WS(rs, 8)] = KP1_414213562 * (T1n - T1o); T1f = FMA(KP951056516, T1a, T17); T1b = FNMS(KP951056516, T1a, T17); } } } } } { E TE, TG, T10, T1e; TE = FMA(KP951056516, TD, Tw); TG = FNMS(KP951056516, TD, Tw); T10 = FMA(KP951056516, TZ, TW); T1e = FNMS(KP951056516, TZ, TW); { E T1d, TP, T1g, T1c; T1d = TF - TG; R0[WS(rs, 2)] = -(KP2_000000000 * (TF + TG)); TP = Tt - TE; R0[WS(rs, 8)] = KP2_000000000 * (Tt + TE); T1g = T1e + T1f; R0[WS(rs, 7)] = KP2_000000000 * (T1e - T1f); T1c = T10 + T1b; R0[WS(rs, 3)] = KP2_000000000 * (T10 - T1b); R1[WS(rs, 9)] = -(KP1_414213562 * (T1d + T1g)); R1[WS(rs, 4)] = KP1_414213562 * (T1d - T1g); R1[WS(rs, 5)] = -(KP1_414213562 * (TP + T1c)); R1[0] = KP1_414213562 * (TP - T1c); } } } } } static const kr2c_desc desc = { 20, "r2cbIII_20", {58, 20, 36, 0}, &GENUS }; void X(codelet_r2cbIII_20) (planner *p) { X(kr2c_register) (p, r2cbIII_20, &desc); } #else /* HAVE_FMA */ /* Generated by: ../../../genfft/gen_r2cb.native -compact -variables 4 -pipeline-latency 4 -sign 1 -n 20 -name r2cbIII_20 -dft-III -include r2cbIII.h */ /* * This function contains 94 FP additions, 44 FP multiplications, * (or, 82 additions, 32 multiplications, 12 fused multiply/add), * 43 stack variables, 6 constants, and 40 memory accesses */ #include "r2cbIII.h" static void r2cbIII_20(R *R0, R *R1, R *Cr, R *Ci, stride rs, stride csr, stride csi, INT v, INT ivs, INT ovs) { DK(KP1_414213562, +1.414213562373095048801688724209698078569671875); DK(KP2_000000000, +2.000000000000000000000000000000000000000000000); DK(KP250000000, +0.250000000000000000000000000000000000000000000); DK(KP951056516, +0.951056516295153572116439333379382143405698634); DK(KP587785252, +0.587785252292473129168705954639072768597652438); DK(KP559016994, +0.559016994374947424102293417182819058860154590); { INT i; for (i = v; i > 0; i = i - 1, R0 = R0 + ovs, R1 = R1 + ovs, Cr = Cr + ivs, Ci = Ci + ivs, MAKE_VOLATILE_STRIDE(80, rs), MAKE_VOLATILE_STRIDE(80, csr), MAKE_VOLATILE_STRIDE(80, csi)) { E T1, Tj, T1k, T13, T8, Tk, T17, Ts, T16, TI, T18, T19, Ta, Tu, T1i; E TS, Th, Tv, TX, TD, TV, TL, TW, TY; { E T7, T12, T4, T11; T1 = Cr[WS(csr, 2)]; { E T5, T6, T2, T3; T5 = Cr[WS(csr, 9)]; T6 = Cr[WS(csr, 5)]; T7 = T5 + T6; T12 = T5 - T6; T2 = Cr[WS(csr, 6)]; T3 = Cr[WS(csr, 1)]; T4 = T2 + T3; T11 = T2 - T3; } Tj = KP559016994 * (T4 - T7); T1k = FNMS(KP951056516, T12, KP587785252 * T11); T13 = FMA(KP951056516, T11, KP587785252 * T12); T8 = T4 + T7; Tk = FNMS(KP250000000, T8, T1); } { E Tr, T15, To, T14; T17 = Ci[WS(csi, 2)]; { E Tp, Tq, Tm, Tn; Tp = Ci[WS(csi, 5)]; Tq = Ci[WS(csi, 9)]; Tr = Tp - Tq; T15 = Tp + Tq; Tm = Ci[WS(csi, 6)]; Tn = Ci[WS(csi, 1)]; To = Tm + Tn; T14 = Tm - Tn; } Ts = FMA(KP951056516, To, KP587785252 * Tr); T16 = KP559016994 * (T14 + T15); TI = FNMS(KP951056516, Tr, KP587785252 * To); T18 = T14 - T15; T19 = FNMS(KP250000000, T18, T17); } { E Tg, TR, Td, TQ; Ta = Cr[WS(csr, 7)]; { E Te, Tf, Tb, Tc; Te = Cr[0]; Tf = Cr[WS(csr, 4)]; Tg = Te + Tf; TR = Te - Tf; Tb = Cr[WS(csr, 3)]; Tc = Cr[WS(csr, 8)]; Td = Tb + Tc; TQ = Tb - Tc; } Tu = KP559016994 * (Td - Tg); T1i = FNMS(KP951056516, TR, KP587785252 * TQ); TS = FMA(KP951056516, TQ, KP587785252 * TR); Th = Td + Tg; Tv = FNMS(KP250000000, Th, Ta); } { E TC, TU, Tz, TT; TX = Ci[WS(csi, 7)]; { E TA, TB, Tx, Ty; TA = Ci[WS(csi, 4)]; TB = Ci[0]; TC = TA - TB; TU = TB + TA; Tx = Ci[WS(csi, 3)]; Ty = Ci[WS(csi, 8)]; Tz = Tx + Ty; TT = Ty - Tx; } TD = FMA(KP951056516, Tz, KP587785252 * TC); TV = KP559016994 * (TT - TU); TL = FNMS(KP587785252, Tz, KP951056516 * TC); TW = TT + TU; TY = FMA(KP250000000, TW, TX); } { E T9, Ti, T1w, T1t, T1u, T1v; T9 = T1 + T8; Ti = Ta + Th; T1w = T9 - Ti; T1t = T18 + T17; T1u = TX - TW; T1v = T1t + T1u; R0[0] = KP2_000000000 * (T9 + Ti); R0[WS(rs, 5)] = KP2_000000000 * (T1u - T1t); R1[WS(rs, 2)] = KP1_414213562 * (T1v - T1w); R1[WS(rs, 7)] = KP1_414213562 * (T1w + T1v); } { E TJ, TO, T1m, T1q, TM, TN, T1j, T1r; { E TH, T1l, TK, T1h; TH = Tk - Tj; TJ = TH + TI; TO = TH - TI; T1l = T19 - T16; T1m = T1k + T1l; T1q = T1l - T1k; TK = Tv - Tu; TM = TK + TL; TN = TL - TK; T1h = TV + TY; T1j = T1h - T1i; T1r = T1i + T1h; } R0[WS(rs, 4)] = KP2_000000000 * (TJ + TM); R0[WS(rs, 6)] = KP2_000000000 * (TN - TO); R0[WS(rs, 9)] = KP2_000000000 * (T1r - T1q); R0[WS(rs, 1)] = KP2_000000000 * (T1j - T1m); { E T1p, T1s, T1n, T1o; T1p = TM - TJ; T1s = T1q + T1r; R1[WS(rs, 1)] = KP1_414213562 * (T1p - T1s); R1[WS(rs, 6)] = KP1_414213562 * (T1p + T1s); T1n = TO + TN; T1o = T1m + T1j; R1[WS(rs, 8)] = KP1_414213562 * (T1n - T1o); R1[WS(rs, 3)] = KP1_414213562 * (T1n + T1o); } } { E Tt, TG, T1b, T1f, TE, TF, T10, T1e; { E Tl, T1a, Tw, TZ; Tl = Tj + Tk; Tt = Tl - Ts; TG = Tl + Ts; T1a = T16 + T19; T1b = T13 + T1a; T1f = T1a - T13; Tw = Tu + Tv; TE = Tw + TD; TF = TD - Tw; TZ = TV - TY; T10 = TS + TZ; T1e = TZ - TS; } R0[WS(rs, 8)] = KP2_000000000 * (Tt + TE); R0[WS(rs, 2)] = KP2_000000000 * (TF - TG); R0[WS(rs, 7)] = KP2_000000000 * (T1f + T1e); R0[WS(rs, 3)] = KP2_000000000 * (T1b + T10); { E T1d, T1g, TP, T1c; T1d = TG + TF; T1g = T1e - T1f; R1[WS(rs, 4)] = KP1_414213562 * (T1d + T1g); R1[WS(rs, 9)] = KP1_414213562 * (T1g - T1d); TP = Tt - TE; T1c = T10 - T1b; R1[0] = KP1_414213562 * (TP + T1c); R1[WS(rs, 5)] = KP1_414213562 * (T1c - TP); } } } } } static const kr2c_desc desc = { 20, "r2cbIII_20", {82, 32, 12, 0}, &GENUS }; void X(codelet_r2cbIII_20) (planner *p) { X(kr2c_register) (p, r2cbIII_20, &desc); } #endif /* HAVE_FMA */ fftw-3.3.3/rdft/scalar/r2cb/r2cbIII_25.c0000644000175000017500000005343012054411020014272 00000000000000/* * Copyright (c) 2003, 2007-11 Matteo Frigo * Copyright (c) 2003, 2007-11 Massachusetts Institute of Technology * * This program is free software; you can redistribute it and/or modify * it under the terms of the GNU General Public License as published by * the Free Software Foundation; either version 2 of the License, or * (at your option) any later version. * * This program is distributed in the hope that it will be useful, * but WITHOUT ANY WARRANTY; without even the implied warranty of * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the * GNU General Public License for more details. * * You should have received a copy of the GNU General Public License * along with this program; if not, write to the Free Software * Foundation, Inc., 51 Franklin Street, Fifth Floor, Boston, MA 02110-1301 USA * */ /* This file was automatically generated --- DO NOT EDIT */ /* Generated on Sun Nov 25 07:41:45 EST 2012 */ #include "codelet-rdft.h" #ifdef HAVE_FMA /* Generated by: ../../../genfft/gen_r2cb.native -fma -reorder-insns -schedule-for-pipeline -compact -variables 4 -pipeline-latency 4 -sign 1 -n 25 -name r2cbIII_25 -dft-III -include r2cbIII.h */ /* * This function contains 152 FP additions, 120 FP multiplications, * (or, 32 additions, 0 multiplications, 120 fused multiply/add), * 115 stack variables, 44 constants, and 50 memory accesses */ #include "r2cbIII.h" static void r2cbIII_25(R *R0, R *R1, R *Cr, R *Ci, stride rs, stride csr, stride csi, INT v, INT ivs, INT ovs) { DK(KP979740652, +0.979740652857618686258237536568998933733477632); DK(KP438153340, +0.438153340021931793654057951961031291699532119); DK(KP1_752613360, +1.752613360087727174616231807844125166798128477); DK(KP963507348, +0.963507348203430549974383005744259307057084020); DK(KP1_721083328, +1.721083328735889354196523361841037632825608373); DK(KP1_606007150, +1.606007150877320829666881187140752009270929701); DK(KP1_011627398, +1.011627398597394192215998921771049272931807941); DK(KP641441904, +0.641441904830606407298806329068862424939687989); DK(KP595480289, +0.595480289600000014706716770488118292997907308); DK(KP452413526, +0.452413526233009763856834323966348796985206956); DK(KP1_809654104, +1.809654104932039055427337295865395187940827822); DK(KP933137358, +0.933137358350283770603023973254446451924190884); DK(KP1_666834356, +1.666834356657377354817925100486477686277992119); DK(KP1_842354653, +1.842354653930286640500894870830132058718564461); DK(KP1_082908895, +1.082908895072625554092571180165639018104066379); DK(KP576710603, +0.576710603632765877371579268136471017090111488); DK(KP662318342, +0.662318342759882818626911127577439236802190210); DK(KP484291580, +0.484291580564315559745084187732367906918006201); DK(KP1_937166322, +1.937166322257262238980336750929471627672024806); DK(KP1_898359647, +1.898359647016882523151110931686726543423167685); DK(KP1_386580726, +1.386580726567734802700860150804827247498955921); DK(KP904730450, +0.904730450839922351881287709692877908104763647); DK(KP1_115827804, +1.115827804063668528375399296931134075984874304); DK(KP470564281, +0.470564281212251493087595091036643380879947982); DK(KP634619297, +0.634619297544148100711287640319130485732531031); DK(KP499013364, +0.499013364214135780976168403431725276668452610); DK(KP1_996053456, +1.996053456856543123904673613726901106673810439); DK(KP559154169, +0.559154169276087864842202529084232643714075927); DK(KP683113946, +0.683113946453479238701949862233725244439656928); DK(KP730409924, +0.730409924561256563751459444999838399157094302); DK(KP549754652, +0.549754652192770074288023275540779861653779767); DK(KP256756360, +0.256756360367726783319498520922669048172391148); DK(KP451418159, +0.451418159099103183892477933432151804893354132); DK(KP846146756, +0.846146756728608505452954290121135880883743802); DK(KP1_902113032, +1.902113032590307144232878666758764286811397268); DK(KP062914667, +0.062914667253649757225485955897349402364686947); DK(KP939062505, +0.939062505817492352556001843133229685779824606); DK(KP951056516, +0.951056516295153572116439333379382143405698634); DK(KP559016994, +0.559016994374947424102293417182819058860154590); DK(KP250000000, +0.250000000000000000000000000000000000000000000); DK(KP1_118033988, +1.118033988749894848204586834365638117720309180); DK(KP2_000000000, +2.000000000000000000000000000000000000000000000); DK(KP500000000, +0.500000000000000000000000000000000000000000000); DK(KP618033988, +0.618033988749894848204586834365638117720309180); { INT i; for (i = v; i > 0; i = i - 1, R0 = R0 + ovs, R1 = R1 + ovs, Cr = Cr + ivs, Ci = Ci + ivs, MAKE_VOLATILE_STRIDE(100, rs), MAKE_VOLATILE_STRIDE(100, csr), MAKE_VOLATILE_STRIDE(100, csi)) { E T1P, T2c, T2a, T24, T26, T25, T27, T2b; { E T1O, TS, T5, T1N, TP, Te, TA, T2i, T1V, T17, T1B, T2h, T1S, T10, T1C; E T1a, T19, Tn, T1h, T1l, T1Y, T1e, T21, TJ, T1g; { E T1, T2, T3, TQ, TR; TQ = Ci[WS(csi, 7)]; TR = Ci[WS(csi, 2)]; T1 = Cr[WS(csr, 12)]; T2 = Cr[WS(csr, 7)]; T3 = Cr[WS(csr, 2)]; T1O = FNMS(KP618033988, TQ, TR); TS = FMA(KP618033988, TR, TQ); { E TV, TU, T1U, T16, T12, T1R, TZ, T11; { E T6, Tz, T14, T15, TX, Tu, Td, Tx, TY, T4, TO, Ty; T6 = Cr[WS(csr, 11)]; T4 = T2 + T3; TO = T3 - T2; Tz = Ci[WS(csi, 11)]; { E Ta, T9, Tb, T7, T8, TN; T7 = Cr[WS(csr, 6)]; T8 = Cr[WS(csr, 8)]; TN = FNMS(KP500000000, T4, T1); T5 = FMA(KP2_000000000, T4, T1); Ta = Cr[WS(csr, 1)]; T14 = T8 - T7; T9 = T7 + T8; T1N = FMA(KP1_118033988, TO, TN); TP = FNMS(KP1_118033988, TO, TN); Tb = Cr[WS(csr, 3)]; { E Tv, Tw, Ts, Tt, Tc; Ts = Ci[WS(csi, 8)]; Tt = Ci[WS(csi, 6)]; T15 = Tb - Ta; Tc = Ta + Tb; Tv = Ci[WS(csi, 3)]; TX = Tt + Ts; Tu = Ts - Tt; Tw = Ci[WS(csi, 1)]; Td = T9 + Tc; TV = Tc - T9; Tx = Tv - Tw; TY = Tw + Tv; } } Te = T6 + Td; TU = FMS(KP250000000, Td, T6); T1U = FNMS(KP618033988, T14, T15); T16 = FMA(KP618033988, T15, T14); T12 = Tx - Tu; Ty = Tu + Tx; T1R = FNMS(KP618033988, TX, TY); TZ = FMA(KP618033988, TY, TX); TA = Ty - Tz; T11 = FMA(KP250000000, Ty, Tz); } { E Tf, TI, T1j, T1k, Tm, T1c, TD, TG, T1d, TH; Tf = Cr[WS(csr, 10)]; TI = Ci[WS(csi, 10)]; { E T13, T1T, TW, T1Q; T13 = FMA(KP559016994, T12, T11); T1T = FNMS(KP559016994, T12, T11); TW = FMA(KP559016994, TV, TU); T1Q = FNMS(KP559016994, TV, TU); T2i = FMA(KP951056516, T1U, T1T); T1V = FNMS(KP951056516, T1U, T1T); T17 = FMA(KP951056516, T16, T13); T1B = FNMS(KP951056516, T16, T13); T2h = FNMS(KP951056516, T1R, T1Q); T1S = FMA(KP951056516, T1R, T1Q); T10 = FNMS(KP951056516, TZ, TW); T1C = FMA(KP951056516, TZ, TW); { E Tg, Th, Tj, Tk; Tg = Cr[WS(csr, 5)]; Th = Cr[WS(csr, 9)]; Tj = Cr[0]; Tk = Cr[WS(csr, 4)]; { E TB, Ti, Tl, TC, TE, TF; TB = Ci[WS(csi, 9)]; T1j = Tg - Th; Ti = Tg + Th; T1k = Tk - Tj; Tl = Tj + Tk; TC = Ci[WS(csi, 5)]; TE = Ci[WS(csi, 4)]; TF = Ci[0]; Tm = Ti + Tl; T1a = Ti - Tl; T1c = TC + TB; TD = TB - TC; TG = TE - TF; T1d = TF + TE; } } } T19 = FMS(KP250000000, Tm, Tf); Tn = Tf + Tm; T1h = TD - TG; TH = TD + TG; T1l = FNMS(KP618033988, T1k, T1j); T1Y = FMA(KP618033988, T1j, T1k); T1e = FMA(KP618033988, T1d, T1c); T21 = FNMS(KP618033988, T1c, T1d); TJ = TH - TI; T1g = FMA(KP250000000, TH, TI); } } } { E T1Z, T1m, T1y, T22, T1f, T1z, T2j, T2g, T2d, T2q, T2s; { E Tq, To, T2e, T2f; Tq = Tn - Te; To = Te + Tn; { E T1i, T1X, T1b, T20; T1i = FNMS(KP559016994, T1h, T1g); T1X = FMA(KP559016994, T1h, T1g); T1b = FNMS(KP559016994, T1a, T19); T20 = FMA(KP559016994, T1a, T19); T2e = FMA(KP951056516, T1Y, T1X); T1Z = FNMS(KP951056516, T1Y, T1X); T1m = FNMS(KP951056516, T1l, T1i); T1y = FMA(KP951056516, T1l, T1i); T2f = FNMS(KP951056516, T21, T20); T22 = FMA(KP951056516, T21, T20); T1f = FNMS(KP951056516, T1e, T1b); T1z = FMA(KP951056516, T1e, T1b); } { E T2o, TK, TM, T2p, Tr, TL, Tp; T2o = FMA(KP939062505, T2h, T2i); T2j = FNMS(KP939062505, T2i, T2h); R0[0] = FMA(KP2_000000000, To, T5); Tp = FNMS(KP500000000, To, T5); TK = FMA(KP618033988, TJ, TA); TM = FNMS(KP618033988, TA, TJ); T2g = FNMS(KP062914667, T2f, T2e); T2p = FMA(KP062914667, T2e, T2f); Tr = FNMS(KP1_118033988, Tq, Tp); TL = FMA(KP1_118033988, Tq, Tp); T2d = FMA(KP1_902113032, T1O, T1N); T1P = FNMS(KP1_902113032, T1O, T1N); T2q = FMA(KP846146756, T2p, T2o); T2s = FNMS(KP451418159, T2o, T2p); R0[WS(rs, 10)] = FMA(KP1_902113032, TK, Tr); R1[WS(rs, 2)] = FMS(KP1_902113032, TK, Tr); R1[WS(rs, 7)] = FMS(KP1_902113032, TM, TL); R0[WS(rs, 5)] = FMA(KP1_902113032, TM, TL); } } { E T18, T1n, T1x, TT, T2m, T1w, T1u, T2l, T1s, T1t, T2k; T18 = FNMS(KP256756360, T17, T10); T1s = FMA(KP256756360, T10, T17); T1t = FMA(KP549754652, T1f, T1m); T1n = FNMS(KP549754652, T1m, T1f); T1x = FNMS(KP1_902113032, TS, TP); TT = FMA(KP1_902113032, TS, TP); T2m = FMA(KP730409924, T2j, T2g); T2k = FNMS(KP730409924, T2j, T2g); T1w = FNMS(KP683113946, T1s, T1t); T1u = FMA(KP559154169, T1t, T1s); R1[WS(rs, 1)] = -(FMA(KP1_996053456, T2k, T2d)); T2l = FNMS(KP499013364, T2k, T2d); { E T1K, T1M, T1G, T1E; { E T1D, T1A, T1q, T1p, T1v, T1r; { E T1I, T1J, T2n, T2r, T1o; T1I = FMA(KP634619297, T1B, T1C); T1D = FNMS(KP634619297, T1C, T1B); T1A = FMA(KP470564281, T1z, T1y); T1J = FNMS(KP470564281, T1y, T1z); T2n = FNMS(KP1_115827804, T2m, T2l); T2r = FMA(KP1_115827804, T2m, T2l); T1q = FNMS(KP904730450, T1n, T18); T1o = FMA(KP904730450, T1n, T18); R1[WS(rs, 11)] = FMS(KP1_386580726, T2q, T2n); R0[WS(rs, 4)] = FMA(KP1_386580726, T2q, T2n); R0[WS(rs, 9)] = FMA(KP1_898359647, T2s, T2r); R1[WS(rs, 6)] = FMS(KP1_898359647, T2s, T2r); R1[0] = FMS(KP1_937166322, T1o, TT); T1p = FMA(KP484291580, T1o, TT); T1K = FMA(KP662318342, T1J, T1I); T1M = FNMS(KP576710603, T1I, T1J); } T1v = FMA(KP1_082908895, T1q, T1p); T1r = FNMS(KP1_082908895, T1q, T1p); R1[WS(rs, 10)] = FMS(KP1_842354653, T1u, T1r); R0[WS(rs, 3)] = FMA(KP1_842354653, T1u, T1r); R0[WS(rs, 8)] = FMA(KP1_666834356, T1w, T1v); R1[WS(rs, 5)] = FMS(KP1_666834356, T1w, T1v); T1G = FNMS(KP933137358, T1D, T1A); T1E = FMA(KP933137358, T1D, T1A); } { E T23, T28, T29, T1W, T1F, T1H, T1L; T23 = FNMS(KP634619297, T22, T1Z); T28 = FMA(KP634619297, T1Z, T22); T29 = FMA(KP549754652, T1S, T1V); T1W = FNMS(KP549754652, T1V, T1S); R0[WS(rs, 2)] = FMA(KP1_809654104, T1E, T1x); T1F = FNMS(KP452413526, T1E, T1x); T2c = FMA(KP595480289, T28, T29); T2a = FNMS(KP641441904, T29, T28); T1H = FNMS(KP1_011627398, T1G, T1F); T1L = FMA(KP1_011627398, T1G, T1F); R0[WS(rs, 12)] = FNMS(KP1_606007150, T1K, T1H); R1[WS(rs, 4)] = -(FMA(KP1_606007150, T1K, T1H)); R1[WS(rs, 9)] = -(FMA(KP1_721083328, T1M, T1L)); R0[WS(rs, 7)] = FNMS(KP1_721083328, T1M, T1L); T24 = FNMS(KP963507348, T23, T1W); T26 = FMA(KP963507348, T23, T1W); } } } } } R0[WS(rs, 1)] = FNMS(KP1_752613360, T24, T1P); T25 = FMA(KP438153340, T24, T1P); T27 = FMA(KP979740652, T26, T25); T2b = FNMS(KP979740652, T26, T25); R1[WS(rs, 8)] = -(FMA(KP1_606007150, T2a, T27)); R0[WS(rs, 6)] = FNMS(KP1_606007150, T2a, T27); R1[WS(rs, 3)] = -(FMA(KP1_666834356, T2c, T2b)); R0[WS(rs, 11)] = FNMS(KP1_666834356, T2c, T2b); } } } static const kr2c_desc desc = { 25, "r2cbIII_25", {32, 0, 120, 0}, &GENUS }; void X(codelet_r2cbIII_25) (planner *p) { X(kr2c_register) (p, r2cbIII_25, &desc); } #else /* HAVE_FMA */ /* Generated by: ../../../genfft/gen_r2cb.native -compact -variables 4 -pipeline-latency 4 -sign 1 -n 25 -name r2cbIII_25 -dft-III -include r2cbIII.h */ /* * This function contains 152 FP additions, 98 FP multiplications, * (or, 100 additions, 46 multiplications, 52 fused multiply/add), * 65 stack variables, 21 constants, and 50 memory accesses */ #include "r2cbIII.h" static void r2cbIII_25(R *R0, R *R1, R *Cr, R *Ci, stride rs, stride csr, stride csi, INT v, INT ivs, INT ovs) { DK(KP968583161, +0.968583161128631119490168375464735813836012403); DK(KP248689887, +0.248689887164854788242283746006447968417567406); DK(KP684547105, +0.684547105928688673732283357621209269889519233); DK(KP728968627, +0.728968627421411523146730319055259111372571664); DK(KP062790519, +0.062790519529313376076178224565631133122484832); DK(KP998026728, +0.998026728428271561952336806863450553336905220); DK(KP876306680, +0.876306680043863587308115903922062583399064238); DK(KP481753674, +0.481753674101715274987191502872129653528542010); DK(KP535826794, +0.535826794978996618271308767867639978063575346); DK(KP844327925, +0.844327925502015078548558063966681505381659241); DK(KP904827052, +0.904827052466019527713668647932697593970413911); DK(KP425779291, +0.425779291565072648862502445744251703979973042); DK(KP250000000, +0.250000000000000000000000000000000000000000000); DK(KP951056516, +0.951056516295153572116439333379382143405698634); DK(KP587785252, +0.587785252292473129168705954639072768597652438); DK(KP559016994, +0.559016994374947424102293417182819058860154590); DK(KP500000000, +0.500000000000000000000000000000000000000000000); DK(KP2_000000000, +2.000000000000000000000000000000000000000000000); DK(KP1_118033988, +1.118033988749894848204586834365638117720309180); DK(KP1_175570504, +1.175570504584946258337411909278145537195304875); DK(KP1_902113032, +1.902113032590307144232878666758764286811397268); { INT i; for (i = v; i > 0; i = i - 1, R0 = R0 + ovs, R1 = R1 + ovs, Cr = Cr + ivs, Ci = Ci + ivs, MAKE_VOLATILE_STRIDE(100, rs), MAKE_VOLATILE_STRIDE(100, csr), MAKE_VOLATILE_STRIDE(100, csi)) { E TS, T1O, T5, TP, T1N, TI, TH, Te, T17, T2h, T1y, T1V, T10, T2g, T1x; E T1S, Tz, Ty, Tn, T1m, T2e, T1B, T22, T1f, T2d, T1A, T1Z, TQ, TR; TQ = Ci[WS(csi, 2)]; TR = Ci[WS(csi, 7)]; TS = FNMS(KP1_175570504, TR, KP1_902113032 * TQ); T1O = FMA(KP1_902113032, TR, KP1_175570504 * TQ); { E T1, T4, TN, T2, T3, TO; T1 = Cr[WS(csr, 12)]; T2 = Cr[WS(csr, 7)]; T3 = Cr[WS(csr, 2)]; T4 = T2 + T3; TN = KP1_118033988 * (T3 - T2); T5 = FMA(KP2_000000000, T4, T1); TO = FMS(KP500000000, T4, T1); TP = TN - TO; T1N = TO + TN; } { E T6, Td, T15, TU, T14, T11, TX, TY; T6 = Cr[WS(csr, 11)]; TI = Ci[WS(csi, 11)]; { E T7, T8, T9, Ta, Tb, Tc; T7 = Cr[WS(csr, 6)]; T8 = Cr[WS(csr, 8)]; T9 = T7 + T8; Ta = Cr[WS(csr, 1)]; Tb = Cr[WS(csr, 3)]; Tc = Ta + Tb; Td = T9 + Tc; T15 = Ta - Tb; TU = KP559016994 * (Tc - T9); T14 = T8 - T7; } { E TB, TC, TD, TE, TF, TG; TB = Ci[WS(csi, 6)]; TC = Ci[WS(csi, 8)]; TD = TB - TC; TE = Ci[WS(csi, 1)]; TF = Ci[WS(csi, 3)]; TG = TE - TF; TH = TD + TG; T11 = KP559016994 * (TD - TG); TX = TB + TC; TY = TE + TF; } Te = T6 + Td; { E T16, T1T, T13, T1U, T12; T16 = FMA(KP587785252, T14, KP951056516 * T15); T1T = FNMS(KP587785252, T15, KP951056516 * T14); T12 = FNMS(KP250000000, TH, TI); T13 = T11 - T12; T1U = T11 + T12; T17 = T13 - T16; T2h = T1T - T1U; T1y = T16 + T13; T1V = T1T + T1U; } { E TZ, T1R, TW, T1Q, TV; TZ = FNMS(KP951056516, TY, KP587785252 * TX); T1R = FMA(KP951056516, TX, KP587785252 * TY); TV = FMS(KP250000000, Td, T6); TW = TU - TV; T1Q = TV + TU; T10 = TW + TZ; T2g = T1Q + T1R; T1x = TZ - TW; T1S = T1Q - T1R; } } { E Tf, Tm, T1k, T19, T1j, T1g, T1c, T1d; Tf = Cr[WS(csr, 10)]; Tz = Ci[WS(csi, 10)]; { E Tg, Th, Ti, Tj, Tk, Tl; Tg = Cr[WS(csr, 5)]; Th = Cr[WS(csr, 9)]; Ti = Tg + Th; Tj = Cr[0]; Tk = Cr[WS(csr, 4)]; Tl = Tj + Tk; Tm = Ti + Tl; T1k = Tj - Tk; T19 = KP559016994 * (Tl - Ti); T1j = Th - Tg; } { E Ts, Tt, Tu, Tv, Tw, Tx; Ts = Ci[WS(csi, 4)]; Tt = Ci[0]; Tu = Ts - Tt; Tv = Ci[WS(csi, 5)]; Tw = Ci[WS(csi, 9)]; Tx = Tv - Tw; Ty = Tu - Tx; T1g = KP559016994 * (Tx + Tu); T1c = Tv + Tw; T1d = Tt + Ts; } Tn = Tf + Tm; { E T1l, T20, T1i, T21, T1h; T1l = FMA(KP587785252, T1j, KP951056516 * T1k); T20 = FNMS(KP587785252, T1k, KP951056516 * T1j); T1h = FMA(KP250000000, Ty, Tz); T1i = T1g - T1h; T21 = T1g + T1h; T1m = T1i - T1l; T2e = T21 - T20; T1B = T1l + T1i; T22 = T20 + T21; } { E T1e, T1Y, T1b, T1X, T1a; T1e = FNMS(KP951056516, T1d, KP587785252 * T1c); T1Y = FMA(KP951056516, T1c, KP587785252 * T1d); T1a = FMS(KP250000000, Tm, Tf); T1b = T19 - T1a; T1X = T1a + T19; T1f = T1b + T1e; T2d = T1X + T1Y; T1A = T1e - T1b; T1Z = T1X - T1Y; } } { E Tq, To, Tp, TK, TM, TA, TJ, TL, Tr; Tq = KP1_118033988 * (Tn - Te); To = Te + Tn; Tp = FMS(KP500000000, To, T5); TA = Ty - Tz; TJ = TH + TI; TK = FNMS(KP1_902113032, TJ, KP1_175570504 * TA); TM = FMA(KP1_175570504, TJ, KP1_902113032 * TA); R0[0] = FMA(KP2_000000000, To, T5); TL = Tq - Tp; R0[WS(rs, 5)] = TL + TM; R1[WS(rs, 7)] = TM - TL; Tr = Tp + Tq; R1[WS(rs, 2)] = Tr + TK; R0[WS(rs, 10)] = TK - Tr; } { E T2q, T2s, T2k, T2j, T2l, T2m, T2r, T2n; { E T2o, T2p, T2f, T2i; T2o = FNMS(KP904827052, T2d, KP425779291 * T2e); T2p = FNMS(KP535826794, T2h, KP844327925 * T2g); T2q = FNMS(KP1_902113032, T2p, KP1_175570504 * T2o); T2s = FMA(KP1_175570504, T2p, KP1_902113032 * T2o); T2k = T1N + T1O; T2f = FMA(KP425779291, T2d, KP904827052 * T2e); T2i = FMA(KP535826794, T2g, KP844327925 * T2h); T2j = T2f - T2i; T2l = FMA(KP500000000, T2j, T2k); T2m = KP1_118033988 * (T2i + T2f); } R0[WS(rs, 2)] = FMS(KP2_000000000, T2j, T2k); T2r = T2m - T2l; R0[WS(rs, 7)] = T2r + T2s; R1[WS(rs, 9)] = T2s - T2r; T2n = T2l + T2m; R1[WS(rs, 4)] = T2n + T2q; R0[WS(rs, 12)] = T2q - T2n; } { E T1u, T1w, TT, T1o, T1p, T1q, T1v, T1r; { E T1s, T1t, T18, T1n; T1s = FMA(KP481753674, T10, KP876306680 * T17); T1t = FMA(KP844327925, T1f, KP535826794 * T1m); T1u = FMA(KP1_902113032, T1s, KP1_175570504 * T1t); T1w = FNMS(KP1_175570504, T1s, KP1_902113032 * T1t); TT = TP - TS; T18 = FNMS(KP481753674, T17, KP876306680 * T10); T1n = FNMS(KP844327925, T1m, KP535826794 * T1f); T1o = T18 + T1n; T1p = FMS(KP500000000, T1o, TT); T1q = KP1_118033988 * (T1n - T18); } R0[WS(rs, 1)] = FMA(KP2_000000000, T1o, TT); T1v = T1q - T1p; R0[WS(rs, 6)] = T1v + T1w; R1[WS(rs, 8)] = T1w - T1v; T1r = T1p + T1q; R1[WS(rs, 3)] = T1r + T1u; R0[WS(rs, 11)] = T1u - T1r; } { E T1H, T1L, T1E, T1D, T1I, T1J, T1M, T1K; { E T1F, T1G, T1z, T1C; T1F = FNMS(KP062790519, T1B, KP998026728 * T1A); T1G = FNMS(KP684547105, T1x, KP728968627 * T1y); T1H = FNMS(KP1_902113032, T1G, KP1_175570504 * T1F); T1L = FMA(KP1_175570504, T1G, KP1_902113032 * T1F); T1E = TP + TS; T1z = FMA(KP728968627, T1x, KP684547105 * T1y); T1C = FMA(KP062790519, T1A, KP998026728 * T1B); T1D = T1z + T1C; T1I = FMA(KP500000000, T1D, T1E); T1J = KP1_118033988 * (T1C - T1z); } R1[WS(rs, 1)] = FMS(KP2_000000000, T1D, T1E); T1M = T1J - T1I; R0[WS(rs, 9)] = T1L - T1M; R1[WS(rs, 6)] = T1L + T1M; T1K = T1I + T1J; R1[WS(rs, 11)] = T1H - T1K; R0[WS(rs, 4)] = T1H + T1K; } { E T2a, T2c, T1P, T24, T25, T26, T2b, T27; { E T28, T29, T1W, T23; T28 = FMA(KP248689887, T1S, KP968583161 * T1V); T29 = FMA(KP481753674, T1Z, KP876306680 * T22); T2a = FMA(KP1_902113032, T28, KP1_175570504 * T29); T2c = FNMS(KP1_175570504, T28, KP1_902113032 * T29); T1P = T1N - T1O; T1W = FNMS(KP248689887, T1V, KP968583161 * T1S); T23 = FNMS(KP481753674, T22, KP876306680 * T1Z); T24 = T1W + T23; T25 = FMS(KP500000000, T24, T1P); T26 = KP1_118033988 * (T23 - T1W); } R1[0] = FMA(KP2_000000000, T24, T1P); T2b = T26 - T25; R1[WS(rs, 5)] = T2b + T2c; R0[WS(rs, 8)] = T2c - T2b; T27 = T25 + T26; R0[WS(rs, 3)] = T27 + T2a; R1[WS(rs, 10)] = T2a - T27; } } } } static const kr2c_desc desc = { 25, "r2cbIII_25", {100, 46, 52, 0}, &GENUS }; void X(codelet_r2cbIII_25) (planner *p) { X(kr2c_register) (p, r2cbIII_25, &desc); } #endif /* HAVE_FMA */ fftw-3.3.3/rdft/scalar/r2cb/hc2cb_2.c0000644000175000017500000000703212054411020014000 00000000000000/* * Copyright (c) 2003, 2007-11 Matteo Frigo * Copyright (c) 2003, 2007-11 Massachusetts Institute of Technology * * This program is free software; you can redistribute it and/or modify * it under the terms of the GNU General Public License as published by * the Free Software Foundation; either version 2 of the License, or * (at your option) any later version. * * This program is distributed in the hope that it will be useful, * but WITHOUT ANY WARRANTY; without even the implied warranty of * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the * GNU General Public License for more details. * * You should have received a copy of the GNU General Public License * along with this program; if not, write to the Free Software * Foundation, Inc., 51 Franklin Street, Fifth Floor, Boston, MA 02110-1301 USA * */ /* This file was automatically generated --- DO NOT EDIT */ /* Generated on Sun Nov 25 07:41:52 EST 2012 */ #include "codelet-rdft.h" #ifdef HAVE_FMA /* Generated by: ../../../genfft/gen_hc2c.native -fma -reorder-insns -schedule-for-pipeline -compact -variables 4 -pipeline-latency 4 -sign 1 -n 2 -dif -name hc2cb_2 -include hc2cb.h */ /* * This function contains 6 FP additions, 4 FP multiplications, * (or, 4 additions, 2 multiplications, 2 fused multiply/add), * 11 stack variables, 0 constants, and 8 memory accesses */ #include "hc2cb.h" static void hc2cb_2(R *Rp, R *Ip, R *Rm, R *Im, const R *W, stride rs, INT mb, INT me, INT ms) { { INT m; for (m = mb, W = W + ((mb - 1) * 2); m < me; m = m + 1, Rp = Rp + ms, Ip = Ip + ms, Rm = Rm - ms, Im = Im - ms, W = W + 2, MAKE_VOLATILE_STRIDE(8, rs)) { E T5, T6, T9, T8, T7, Ta; { E T1, T2, T3, T4; T1 = Rp[0]; T2 = Rm[0]; T3 = Ip[0]; T4 = Im[0]; T5 = W[0]; Rp[0] = T1 + T2; T6 = T1 - T2; Rm[0] = T3 - T4; T9 = T3 + T4; T8 = W[1]; T7 = T5 * T6; } Ta = T8 * T6; Ip[0] = FNMS(T8, T9, T7); Im[0] = FMA(T5, T9, Ta); } } } static const tw_instr twinstr[] = { {TW_FULL, 1, 2}, {TW_NEXT, 1, 0} }; static const hc2c_desc desc = { 2, "hc2cb_2", twinstr, &GENUS, {4, 2, 2, 0} }; void X(codelet_hc2cb_2) (planner *p) { X(khc2c_register) (p, hc2cb_2, &desc, HC2C_VIA_RDFT); } #else /* HAVE_FMA */ /* Generated by: ../../../genfft/gen_hc2c.native -compact -variables 4 -pipeline-latency 4 -sign 1 -n 2 -dif -name hc2cb_2 -include hc2cb.h */ /* * This function contains 6 FP additions, 4 FP multiplications, * (or, 4 additions, 2 multiplications, 2 fused multiply/add), * 9 stack variables, 0 constants, and 8 memory accesses */ #include "hc2cb.h" static void hc2cb_2(R *Rp, R *Ip, R *Rm, R *Im, const R *W, stride rs, INT mb, INT me, INT ms) { { INT m; for (m = mb, W = W + ((mb - 1) * 2); m < me; m = m + 1, Rp = Rp + ms, Ip = Ip + ms, Rm = Rm - ms, Im = Im - ms, W = W + 2, MAKE_VOLATILE_STRIDE(8, rs)) { E T1, T2, T6, T3, T4, T8, T5, T7; T1 = Rp[0]; T2 = Rm[0]; T6 = T1 - T2; T3 = Ip[0]; T4 = Im[0]; T8 = T3 + T4; Rp[0] = T1 + T2; Rm[0] = T3 - T4; T5 = W[0]; T7 = W[1]; Ip[0] = FNMS(T7, T8, T5 * T6); Im[0] = FMA(T7, T6, T5 * T8); } } } static const tw_instr twinstr[] = { {TW_FULL, 1, 2}, {TW_NEXT, 1, 0} }; static const hc2c_desc desc = { 2, "hc2cb_2", twinstr, &GENUS, {4, 2, 2, 0} }; void X(codelet_hc2cb_2) (planner *p) { X(khc2c_register) (p, hc2cb_2, &desc, HC2C_VIA_RDFT); } #endif /* HAVE_FMA */ fftw-3.3.3/rdft/scalar/r2cb/hc2cb_4.c0000644000175000017500000001227012054411020014002 00000000000000/* * Copyright (c) 2003, 2007-11 Matteo Frigo * Copyright (c) 2003, 2007-11 Massachusetts Institute of Technology * * This program is free software; you can redistribute it and/or modify * it under the terms of the GNU General Public License as published by * the Free Software Foundation; either version 2 of the License, or * (at your option) any later version. * * This program is distributed in the hope that it will be useful, * but WITHOUT ANY WARRANTY; without even the implied warranty of * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the * GNU General Public License for more details. * * You should have received a copy of the GNU General Public License * along with this program; if not, write to the Free Software * Foundation, Inc., 51 Franklin Street, Fifth Floor, Boston, MA 02110-1301 USA * */ /* This file was automatically generated --- DO NOT EDIT */ /* Generated on Sun Nov 25 07:41:52 EST 2012 */ #include "codelet-rdft.h" #ifdef HAVE_FMA /* Generated by: ../../../genfft/gen_hc2c.native -fma -reorder-insns -schedule-for-pipeline -compact -variables 4 -pipeline-latency 4 -sign 1 -n 4 -dif -name hc2cb_4 -include hc2cb.h */ /* * This function contains 22 FP additions, 12 FP multiplications, * (or, 16 additions, 6 multiplications, 6 fused multiply/add), * 25 stack variables, 0 constants, and 16 memory accesses */ #include "hc2cb.h" static void hc2cb_4(R *Rp, R *Ip, R *Rm, R *Im, const R *W, stride rs, INT mb, INT me, INT ms) { { INT m; for (m = mb, W = W + ((mb - 1) * 6); m < me; m = m + 1, Rp = Rp + ms, Ip = Ip + ms, Rm = Rm - ms, Im = Im - ms, W = W + 6, MAKE_VOLATILE_STRIDE(16, rs)) { E Th, Ta, T7, Ti, T9; { E Tq, Td, T3, Tg, Tu, Tm, T6, Tp; { E Tk, T4, Tl, T5; { E Tb, Tc, T1, T2, Te, Tf; Tb = Ip[0]; Tc = Im[WS(rs, 1)]; T1 = Rp[0]; T2 = Rm[WS(rs, 1)]; Te = Ip[WS(rs, 1)]; Tq = Tb + Tc; Td = Tb - Tc; Tf = Im[0]; Tk = T1 - T2; T3 = T1 + T2; T4 = Rp[WS(rs, 1)]; Tg = Te - Tf; Tl = Te + Tf; T5 = Rm[0]; } Tu = Tk + Tl; Tm = Tk - Tl; T6 = T4 + T5; Tp = T4 - T5; } Rm[0] = Td + Tg; { E Tx, Tr, T8, Tn, Ts, To, Tj; Tj = W[0]; Tx = Tq - Tp; Tr = Tp + Tq; Rp[0] = T3 + T6; T8 = T3 - T6; Tn = Tj * Tm; Ts = Tj * Tr; To = W[1]; { E Tt, Tw, Ty, Tv; Tt = W[4]; Tw = W[5]; Th = Td - Tg; Im[0] = FMA(To, Tm, Ts); Ip[0] = FNMS(To, Tr, Tn); Ty = Tt * Tx; Tv = Tt * Tu; Ta = W[3]; T7 = W[2]; Im[WS(rs, 1)] = FMA(Tw, Tu, Ty); Ip[WS(rs, 1)] = FNMS(Tw, Tx, Tv); Ti = Ta * T8; T9 = T7 * T8; } } } Rm[WS(rs, 1)] = FMA(T7, Th, Ti); Rp[WS(rs, 1)] = FNMS(Ta, Th, T9); } } } static const tw_instr twinstr[] = { {TW_FULL, 1, 4}, {TW_NEXT, 1, 0} }; static const hc2c_desc desc = { 4, "hc2cb_4", twinstr, &GENUS, {16, 6, 6, 0} }; void X(codelet_hc2cb_4) (planner *p) { X(khc2c_register) (p, hc2cb_4, &desc, HC2C_VIA_RDFT); } #else /* HAVE_FMA */ /* Generated by: ../../../genfft/gen_hc2c.native -compact -variables 4 -pipeline-latency 4 -sign 1 -n 4 -dif -name hc2cb_4 -include hc2cb.h */ /* * This function contains 22 FP additions, 12 FP multiplications, * (or, 16 additions, 6 multiplications, 6 fused multiply/add), * 13 stack variables, 0 constants, and 16 memory accesses */ #include "hc2cb.h" static void hc2cb_4(R *Rp, R *Ip, R *Rm, R *Im, const R *W, stride rs, INT mb, INT me, INT ms) { { INT m; for (m = mb, W = W + ((mb - 1) * 6); m < me; m = m + 1, Rp = Rp + ms, Ip = Ip + ms, Rm = Rm - ms, Im = Im - ms, W = W + 6, MAKE_VOLATILE_STRIDE(16, rs)) { E T3, Ti, Tc, Tn, T6, Tm, Tf, Tj; { E T1, T2, Ta, Tb; T1 = Rp[0]; T2 = Rm[WS(rs, 1)]; T3 = T1 + T2; Ti = T1 - T2; Ta = Ip[0]; Tb = Im[WS(rs, 1)]; Tc = Ta - Tb; Tn = Ta + Tb; } { E T4, T5, Td, Te; T4 = Rp[WS(rs, 1)]; T5 = Rm[0]; T6 = T4 + T5; Tm = T4 - T5; Td = Ip[WS(rs, 1)]; Te = Im[0]; Tf = Td - Te; Tj = Td + Te; } Rp[0] = T3 + T6; Rm[0] = Tc + Tf; { E T8, Tg, T7, T9; T8 = T3 - T6; Tg = Tc - Tf; T7 = W[2]; T9 = W[3]; Rp[WS(rs, 1)] = FNMS(T9, Tg, T7 * T8); Rm[WS(rs, 1)] = FMA(T9, T8, T7 * Tg); } { E Tk, To, Th, Tl; Tk = Ti - Tj; To = Tm + Tn; Th = W[0]; Tl = W[1]; Ip[0] = FNMS(Tl, To, Th * Tk); Im[0] = FMA(Th, To, Tl * Tk); } { E Tq, Ts, Tp, Tr; Tq = Ti + Tj; Ts = Tn - Tm; Tp = W[4]; Tr = W[5]; Ip[WS(rs, 1)] = FNMS(Tr, Ts, Tp * Tq); Im[WS(rs, 1)] = FMA(Tp, Ts, Tr * Tq); } } } } static const tw_instr twinstr[] = { {TW_FULL, 1, 4}, {TW_NEXT, 1, 0} }; static const hc2c_desc desc = { 4, "hc2cb_4", twinstr, &GENUS, {16, 6, 6, 0} }; void X(codelet_hc2cb_4) (planner *p) { X(khc2c_register) (p, hc2cb_4, &desc, HC2C_VIA_RDFT); } #endif /* HAVE_FMA */ fftw-3.3.3/rdft/scalar/r2cb/hc2cb_6.c0000644000175000017500000001737012054411020014012 00000000000000/* * Copyright (c) 2003, 2007-11 Matteo Frigo * Copyright (c) 2003, 2007-11 Massachusetts Institute of Technology * * This program is free software; you can redistribute it and/or modify * it under the terms of the GNU General Public License as published by * the Free Software Foundation; either version 2 of the License, or * (at your option) any later version. * * This program is distributed in the hope that it will be useful, * but WITHOUT ANY WARRANTY; without even the implied warranty of * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the * GNU General Public License for more details. * * You should have received a copy of the GNU General Public License * along with this program; if not, write to the Free Software * Foundation, Inc., 51 Franklin Street, Fifth Floor, Boston, MA 02110-1301 USA * */ /* This file was automatically generated --- DO NOT EDIT */ /* Generated on Sun Nov 25 07:41:52 EST 2012 */ #include "codelet-rdft.h" #ifdef HAVE_FMA /* Generated by: ../../../genfft/gen_hc2c.native -fma -reorder-insns -schedule-for-pipeline -compact -variables 4 -pipeline-latency 4 -sign 1 -n 6 -dif -name hc2cb_6 -include hc2cb.h */ /* * This function contains 46 FP additions, 32 FP multiplications, * (or, 24 additions, 10 multiplications, 22 fused multiply/add), * 45 stack variables, 2 constants, and 24 memory accesses */ #include "hc2cb.h" static void hc2cb_6(R *Rp, R *Ip, R *Rm, R *Im, const R *W, stride rs, INT mb, INT me, INT ms) { DK(KP866025403, +0.866025403784438646763723170752936183471402627); DK(KP500000000, +0.500000000000000000000000000000000000000000000); { INT m; for (m = mb, W = W + ((mb - 1) * 10); m < me; m = m + 1, Rp = Rp + ms, Ip = Ip + ms, Rm = Rm - ms, Im = Im - ms, W = W + 10, MAKE_VOLATILE_STRIDE(24, rs)) { E TK, TR, TB, TM, TL, TS; { E Td, TN, TO, TJ, Tn, Tk, TC, T3, Tr, T7, T8, T4, T5; { E TI, Tj, Tg, TH, Te, Tf, T1, T2; { E Tb, Tc, Th, Ti; Tb = Ip[0]; Tc = Im[WS(rs, 2)]; Th = Ip[WS(rs, 1)]; Ti = Im[WS(rs, 1)]; Te = Ip[WS(rs, 2)]; Td = Tb - Tc; TN = Tb + Tc; Tf = Im[0]; TI = Th + Ti; Tj = Th - Ti; } Tg = Te - Tf; TH = Te + Tf; T1 = Rp[0]; T2 = Rm[WS(rs, 2)]; TO = TH - TI; TJ = TH + TI; Tn = Tj - Tg; Tk = Tg + Tj; TC = T1 - T2; T3 = T1 + T2; Tr = FNMS(KP500000000, Tk, Td); T7 = Rm[WS(rs, 1)]; T8 = Rp[WS(rs, 1)]; T4 = Rp[WS(rs, 2)]; T5 = Rm[0]; } { E Tl, Tq, TQ, Ts, Ta, T10, TG; Rm[0] = Td + Tk; { E T9, TE, T6, TD, TF; T9 = T7 + T8; TE = T7 - T8; T6 = T4 + T5; TD = T4 - T5; Tl = W[2]; Tq = W[3]; TQ = TD - TE; TF = TD + TE; Ts = T6 - T9; Ta = T6 + T9; T10 = TC + TF; TG = FNMS(KP500000000, TF, TC); } { E T13, TP, Tz, TZ, Tw, T14, Tv, Ty; { E Tt, T12, T11, Tp, Tm, To, Tu; T13 = TN + TO; TP = FNMS(KP500000000, TO, TN); Rp[0] = T3 + Ta; Tm = FNMS(KP500000000, Ta, T3); Tz = FMA(KP866025403, Ts, Tr); Tt = FNMS(KP866025403, Ts, Tr); TZ = W[4]; To = FNMS(KP866025403, Tn, Tm); Tw = FMA(KP866025403, Tn, Tm); Tu = Tl * Tt; T12 = W[5]; T11 = TZ * T10; Tp = Tl * To; Rm[WS(rs, 1)] = FMA(Tq, To, Tu); T14 = T12 * T10; Ip[WS(rs, 1)] = FNMS(T12, T13, T11); Rp[WS(rs, 1)] = FNMS(Tq, Tt, Tp); } Im[WS(rs, 1)] = FMA(TZ, T13, T14); Tv = W[6]; Ty = W[7]; { E TX, TT, TW, TV, TY, TU, TA, Tx; TK = FNMS(KP866025403, TJ, TG); TU = FMA(KP866025403, TJ, TG); TA = Tv * Tz; Tx = Tv * Tw; TX = FNMS(KP866025403, TQ, TP); TR = FMA(KP866025403, TQ, TP); Rm[WS(rs, 2)] = FMA(Ty, Tw, TA); Rp[WS(rs, 2)] = FNMS(Ty, Tz, Tx); TT = W[8]; TW = W[9]; TB = W[0]; TV = TT * TU; TY = TW * TU; TM = W[1]; TL = TB * TK; Ip[WS(rs, 2)] = FNMS(TW, TX, TV); Im[WS(rs, 2)] = FMA(TT, TX, TY); } } } } Ip[0] = FNMS(TM, TR, TL); TS = TM * TK; Im[0] = FMA(TB, TR, TS); } } } static const tw_instr twinstr[] = { {TW_FULL, 1, 6}, {TW_NEXT, 1, 0} }; static const hc2c_desc desc = { 6, "hc2cb_6", twinstr, &GENUS, {24, 10, 22, 0} }; void X(codelet_hc2cb_6) (planner *p) { X(khc2c_register) (p, hc2cb_6, &desc, HC2C_VIA_RDFT); } #else /* HAVE_FMA */ /* Generated by: ../../../genfft/gen_hc2c.native -compact -variables 4 -pipeline-latency 4 -sign 1 -n 6 -dif -name hc2cb_6 -include hc2cb.h */ /* * This function contains 46 FP additions, 28 FP multiplications, * (or, 32 additions, 14 multiplications, 14 fused multiply/add), * 25 stack variables, 2 constants, and 24 memory accesses */ #include "hc2cb.h" static void hc2cb_6(R *Rp, R *Ip, R *Rm, R *Im, const R *W, stride rs, INT mb, INT me, INT ms) { DK(KP500000000, +0.500000000000000000000000000000000000000000000); DK(KP866025403, +0.866025403784438646763723170752936183471402627); { INT m; for (m = mb, W = W + ((mb - 1) * 10); m < me; m = m + 1, Rp = Rp + ms, Ip = Ip + ms, Rm = Rm - ms, Im = Im - ms, W = W + 10, MAKE_VOLATILE_STRIDE(24, rs)) { E T3, Ty, Td, TE, Ta, TO, Tr, TB, Tk, TL, Tn, TH; { E T1, T2, Tb, Tc; T1 = Rp[0]; T2 = Rm[WS(rs, 2)]; T3 = T1 + T2; Ty = T1 - T2; Tb = Ip[0]; Tc = Im[WS(rs, 2)]; Td = Tb - Tc; TE = Tb + Tc; } { E T6, Tz, T9, TA; { E T4, T5, T7, T8; T4 = Rp[WS(rs, 2)]; T5 = Rm[0]; T6 = T4 + T5; Tz = T4 - T5; T7 = Rm[WS(rs, 1)]; T8 = Rp[WS(rs, 1)]; T9 = T7 + T8; TA = T7 - T8; } Ta = T6 + T9; TO = KP866025403 * (Tz - TA); Tr = KP866025403 * (T6 - T9); TB = Tz + TA; } { E Tg, TG, Tj, TF; { E Te, Tf, Th, Ti; Te = Ip[WS(rs, 2)]; Tf = Im[0]; Tg = Te - Tf; TG = Te + Tf; Th = Ip[WS(rs, 1)]; Ti = Im[WS(rs, 1)]; Tj = Th - Ti; TF = Th + Ti; } Tk = Tg + Tj; TL = KP866025403 * (TG + TF); Tn = KP866025403 * (Tj - Tg); TH = TF - TG; } Rp[0] = T3 + Ta; Rm[0] = Td + Tk; { E TC, TI, Tx, TD; TC = Ty + TB; TI = TE - TH; Tx = W[4]; TD = W[5]; Ip[WS(rs, 1)] = FNMS(TD, TI, Tx * TC); Im[WS(rs, 1)] = FMA(TD, TC, Tx * TI); } { E To, Tu, Ts, Tw, Tm, Tq; Tm = FNMS(KP500000000, Ta, T3); To = Tm - Tn; Tu = Tm + Tn; Tq = FNMS(KP500000000, Tk, Td); Ts = Tq - Tr; Tw = Tr + Tq; { E Tl, Tp, Tt, Tv; Tl = W[2]; Tp = W[3]; Rp[WS(rs, 1)] = FNMS(Tp, Ts, Tl * To); Rm[WS(rs, 1)] = FMA(Tl, Ts, Tp * To); Tt = W[6]; Tv = W[7]; Rp[WS(rs, 2)] = FNMS(Tv, Tw, Tt * Tu); Rm[WS(rs, 2)] = FMA(Tt, Tw, Tv * Tu); } } { E TM, TS, TQ, TU, TK, TP; TK = FNMS(KP500000000, TB, Ty); TM = TK - TL; TS = TK + TL; TP = FMA(KP500000000, TH, TE); TQ = TO + TP; TU = TP - TO; { E TJ, TN, TR, TT; TJ = W[0]; TN = W[1]; Ip[0] = FNMS(TN, TQ, TJ * TM); Im[0] = FMA(TN, TM, TJ * TQ); TR = W[8]; TT = W[9]; Ip[WS(rs, 2)] = FNMS(TT, TU, TR * TS); Im[WS(rs, 2)] = FMA(TT, TS, TR * TU); } } } } } static const tw_instr twinstr[] = { {TW_FULL, 1, 6}, {TW_NEXT, 1, 0} }; static const hc2c_desc desc = { 6, "hc2cb_6", twinstr, &GENUS, {32, 14, 14, 0} }; void X(codelet_hc2cb_6) (planner *p) { X(khc2c_register) (p, hc2cb_6, &desc, HC2C_VIA_RDFT); } #endif /* HAVE_FMA */ fftw-3.3.3/rdft/scalar/r2cb/hc2cb_8.c0000644000175000017500000002267512054411021014021 00000000000000/* * Copyright (c) 2003, 2007-11 Matteo Frigo * Copyright (c) 2003, 2007-11 Massachusetts Institute of Technology * * This program is free software; you can redistribute it and/or modify * it under the terms of the GNU General Public License as published by * the Free Software Foundation; either version 2 of the License, or * (at your option) any later version. * * This program is distributed in the hope that it will be useful, * but WITHOUT ANY WARRANTY; without even the implied warranty of * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the * GNU General Public License for more details. * * You should have received a copy of the GNU General Public License * along with this program; if not, write to the Free Software * Foundation, Inc., 51 Franklin Street, Fifth Floor, Boston, MA 02110-1301 USA * */ /* This file was automatically generated --- DO NOT EDIT */ /* Generated on Sun Nov 25 07:41:52 EST 2012 */ #include "codelet-rdft.h" #ifdef HAVE_FMA /* Generated by: ../../../genfft/gen_hc2c.native -fma -reorder-insns -schedule-for-pipeline -compact -variables 4 -pipeline-latency 4 -sign 1 -n 8 -dif -name hc2cb_8 -include hc2cb.h */ /* * This function contains 66 FP additions, 36 FP multiplications, * (or, 44 additions, 14 multiplications, 22 fused multiply/add), * 52 stack variables, 1 constants, and 32 memory accesses */ #include "hc2cb.h" static void hc2cb_8(R *Rp, R *Ip, R *Rm, R *Im, const R *W, stride rs, INT mb, INT me, INT ms) { DK(KP707106781, +0.707106781186547524400844362104849039284835938); { INT m; for (m = mb, W = W + ((mb - 1) * 14); m < me; m = m + 1, Rp = Rp + ms, Ip = Ip + ms, Rm = Rm - ms, Im = Im - ms, W = W + 14, MAKE_VOLATILE_STRIDE(32, rs)) { E Tw, TH, Tf, Ty, Tx, TI; { E TV, TD, T1i, T7, T1b, T1n, TQ, Tk, Tp, TE, Te, T1o, T1e, T1j, Tu; E TF; { E T4, Tg, T3, T19, TC, T5, Th, Ti; { E T1, T2, TA, TB; T1 = Rp[0]; T2 = Rm[WS(rs, 3)]; TA = Ip[0]; TB = Im[WS(rs, 3)]; T4 = Rp[WS(rs, 2)]; Tg = T1 - T2; T3 = T1 + T2; T19 = TA - TB; TC = TA + TB; T5 = Rm[WS(rs, 1)]; Th = Ip[WS(rs, 2)]; Ti = Im[WS(rs, 1)]; } { E Tb, Tl, Ta, T1c, To, Tc, Tr, Ts; { E T8, T9, Tm, Tn; T8 = Rp[WS(rs, 1)]; { E Tz, T6, T1a, Tj; Tz = T4 - T5; T6 = T4 + T5; T1a = Th - Ti; Tj = Th + Ti; TV = TC - Tz; TD = Tz + TC; T1i = T3 - T6; T7 = T3 + T6; T1b = T19 + T1a; T1n = T19 - T1a; TQ = Tg + Tj; Tk = Tg - Tj; T9 = Rm[WS(rs, 2)]; } Tm = Ip[WS(rs, 1)]; Tn = Im[WS(rs, 2)]; Tb = Rm[0]; Tl = T8 - T9; Ta = T8 + T9; T1c = Tm - Tn; To = Tm + Tn; Tc = Rp[WS(rs, 3)]; Tr = Ip[WS(rs, 3)]; Ts = Im[0]; } { E Tq, Td, T1d, Tt; Tp = Tl - To; TE = Tl + To; Tq = Tb - Tc; Td = Tb + Tc; T1d = Tr - Ts; Tt = Tr + Ts; Te = Ta + Td; T1o = Ta - Td; T1e = T1c + T1d; T1j = T1d - T1c; Tu = Tq - Tt; TF = Tq + Tt; } } } { E TG, Tv, T10, T13, T1s, T1k, T1p, T1v, T1u, T1w, T1t, TR, TW; Rp[0] = T7 + Te; Rm[0] = T1b + T1e; TG = TE - TF; TR = TE + TF; TW = Tp - Tu; Tv = Tp + Tu; { E TP, TS, TX, TU, T1r, TT, TY; TP = W[4]; T10 = FMA(KP707106781, TR, TQ); TS = FNMS(KP707106781, TR, TQ); TX = FMA(KP707106781, TW, TV); T13 = FNMS(KP707106781, TW, TV); TU = W[5]; T1s = T1i + T1j; T1k = T1i - T1j; TT = TP * TS; TY = TP * TX; T1p = T1n - T1o; T1v = T1o + T1n; T1r = W[2]; Ip[WS(rs, 1)] = FNMS(TU, TX, TT); Im[WS(rs, 1)] = FMA(TU, TS, TY); T1u = W[3]; T1w = T1r * T1v; T1t = T1r * T1s; } { E T1f, T15, T18, T17, T1g, T1h, T1m; { E TZ, T12, T16, T14, T11; Rm[WS(rs, 1)] = FMA(T1u, T1s, T1w); Rp[WS(rs, 1)] = FNMS(T1u, T1v, T1t); TZ = W[12]; T12 = W[13]; T1f = T1b - T1e; T16 = T7 - Te; T14 = TZ * T13; T11 = TZ * T10; T15 = W[6]; T18 = W[7]; Im[WS(rs, 3)] = FMA(T12, T10, T14); Ip[WS(rs, 3)] = FNMS(T12, T13, T11); T17 = T15 * T16; T1g = T18 * T16; } Rp[WS(rs, 2)] = FNMS(T18, T1f, T17); Rm[WS(rs, 2)] = FMA(T15, T1f, T1g); T1h = W[10]; T1m = W[11]; { E TN, TJ, TM, TL, TO, TK, T1q, T1l; Tw = FNMS(KP707106781, Tv, Tk); TK = FMA(KP707106781, Tv, Tk); T1q = T1h * T1p; T1l = T1h * T1k; TN = FMA(KP707106781, TG, TD); TH = FNMS(KP707106781, TG, TD); Rm[WS(rs, 3)] = FMA(T1m, T1k, T1q); Rp[WS(rs, 3)] = FNMS(T1m, T1p, T1l); TJ = W[0]; TM = W[1]; Tf = W[8]; TL = TJ * TK; TO = TM * TK; Ty = W[9]; Tx = Tf * Tw; Ip[0] = FNMS(TM, TN, TL); Im[0] = FMA(TJ, TN, TO); } } } } Ip[WS(rs, 2)] = FNMS(Ty, TH, Tx); TI = Ty * Tw; Im[WS(rs, 2)] = FMA(Tf, TH, TI); } } } static const tw_instr twinstr[] = { {TW_FULL, 1, 8}, {TW_NEXT, 1, 0} }; static const hc2c_desc desc = { 8, "hc2cb_8", twinstr, &GENUS, {44, 14, 22, 0} }; void X(codelet_hc2cb_8) (planner *p) { X(khc2c_register) (p, hc2cb_8, &desc, HC2C_VIA_RDFT); } #else /* HAVE_FMA */ /* Generated by: ../../../genfft/gen_hc2c.native -compact -variables 4 -pipeline-latency 4 -sign 1 -n 8 -dif -name hc2cb_8 -include hc2cb.h */ /* * This function contains 66 FP additions, 32 FP multiplications, * (or, 52 additions, 18 multiplications, 14 fused multiply/add), * 30 stack variables, 1 constants, and 32 memory accesses */ #include "hc2cb.h" static void hc2cb_8(R *Rp, R *Ip, R *Rm, R *Im, const R *W, stride rs, INT mb, INT me, INT ms) { DK(KP707106781, +0.707106781186547524400844362104849039284835938); { INT m; for (m = mb, W = W + ((mb - 1) * 14); m < me; m = m + 1, Rp = Rp + ms, Ip = Ip + ms, Rm = Rm - ms, Im = Im - ms, W = W + 14, MAKE_VOLATILE_STRIDE(32, rs)) { E T7, T18, T1c, To, Ty, TM, TY, TC, Te, TZ, T10, Tv, Tz, TP, TS; E TD; { E T3, TK, Tk, TX, T6, TW, Tn, TL; { E T1, T2, Ti, Tj; T1 = Rp[0]; T2 = Rm[WS(rs, 3)]; T3 = T1 + T2; TK = T1 - T2; Ti = Ip[0]; Tj = Im[WS(rs, 3)]; Tk = Ti - Tj; TX = Ti + Tj; } { E T4, T5, Tl, Tm; T4 = Rp[WS(rs, 2)]; T5 = Rm[WS(rs, 1)]; T6 = T4 + T5; TW = T4 - T5; Tl = Ip[WS(rs, 2)]; Tm = Im[WS(rs, 1)]; Tn = Tl - Tm; TL = Tl + Tm; } T7 = T3 + T6; T18 = TK + TL; T1c = TX - TW; To = Tk + Tn; Ty = T3 - T6; TM = TK - TL; TY = TW + TX; TC = Tk - Tn; } { E Ta, TN, Tr, TO, Td, TQ, Tu, TR; { E T8, T9, Tp, Tq; T8 = Rp[WS(rs, 1)]; T9 = Rm[WS(rs, 2)]; Ta = T8 + T9; TN = T8 - T9; Tp = Ip[WS(rs, 1)]; Tq = Im[WS(rs, 2)]; Tr = Tp - Tq; TO = Tp + Tq; } { E Tb, Tc, Ts, Tt; Tb = Rm[0]; Tc = Rp[WS(rs, 3)]; Td = Tb + Tc; TQ = Tb - Tc; Ts = Ip[WS(rs, 3)]; Tt = Im[0]; Tu = Ts - Tt; TR = Ts + Tt; } Te = Ta + Td; TZ = TN + TO; T10 = TQ + TR; Tv = Tr + Tu; Tz = Tu - Tr; TP = TN - TO; TS = TQ - TR; TD = Ta - Td; } Rp[0] = T7 + Te; Rm[0] = To + Tv; { E Tg, Tw, Tf, Th; Tg = T7 - Te; Tw = To - Tv; Tf = W[6]; Th = W[7]; Rp[WS(rs, 2)] = FNMS(Th, Tw, Tf * Tg); Rm[WS(rs, 2)] = FMA(Th, Tg, Tf * Tw); } { E TG, TI, TF, TH; TG = Ty + Tz; TI = TD + TC; TF = W[2]; TH = W[3]; Rp[WS(rs, 1)] = FNMS(TH, TI, TF * TG); Rm[WS(rs, 1)] = FMA(TF, TI, TH * TG); } { E TA, TE, Tx, TB; TA = Ty - Tz; TE = TC - TD; Tx = W[10]; TB = W[11]; Rp[WS(rs, 3)] = FNMS(TB, TE, Tx * TA); Rm[WS(rs, 3)] = FMA(Tx, TE, TB * TA); } { E T1a, T1g, T1e, T1i, T19, T1d; T19 = KP707106781 * (TZ + T10); T1a = T18 - T19; T1g = T18 + T19; T1d = KP707106781 * (TP - TS); T1e = T1c + T1d; T1i = T1c - T1d; { E T17, T1b, T1f, T1h; T17 = W[4]; T1b = W[5]; Ip[WS(rs, 1)] = FNMS(T1b, T1e, T17 * T1a); Im[WS(rs, 1)] = FMA(T17, T1e, T1b * T1a); T1f = W[12]; T1h = W[13]; Ip[WS(rs, 3)] = FNMS(T1h, T1i, T1f * T1g); Im[WS(rs, 3)] = FMA(T1f, T1i, T1h * T1g); } } { E TU, T14, T12, T16, TT, T11; TT = KP707106781 * (TP + TS); TU = TM - TT; T14 = TM + TT; T11 = KP707106781 * (TZ - T10); T12 = TY - T11; T16 = TY + T11; { E TJ, TV, T13, T15; TJ = W[8]; TV = W[9]; Ip[WS(rs, 2)] = FNMS(TV, T12, TJ * TU); Im[WS(rs, 2)] = FMA(TV, TU, TJ * T12); T13 = W[0]; T15 = W[1]; Ip[0] = FNMS(T15, T16, T13 * T14); Im[0] = FMA(T15, T14, T13 * T16); } } } } } static const tw_instr twinstr[] = { {TW_FULL, 1, 8}, {TW_NEXT, 1, 0} }; static const hc2c_desc desc = { 8, "hc2cb_8", twinstr, &GENUS, {52, 18, 14, 0} }; void X(codelet_hc2cb_8) (planner *p) { X(khc2c_register) (p, hc2cb_8, &desc, HC2C_VIA_RDFT); } #endif /* HAVE_FMA */ fftw-3.3.3/rdft/scalar/r2cb/hc2cb_10.c0000644000175000017500000003267312054411021014071 00000000000000/* * Copyright (c) 2003, 2007-11 Matteo Frigo * Copyright (c) 2003, 2007-11 Massachusetts Institute of Technology * * This program is free software; you can redistribute it and/or modify * it under the terms of the GNU General Public License as published by * the Free Software Foundation; either version 2 of the License, or * (at your option) any later version. * * This program is distributed in the hope that it will be useful, * but WITHOUT ANY WARRANTY; without even the implied warranty of * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the * GNU General Public License for more details. * * You should have received a copy of the GNU General Public License * along with this program; if not, write to the Free Software * Foundation, Inc., 51 Franklin Street, Fifth Floor, Boston, MA 02110-1301 USA * */ /* This file was automatically generated --- DO NOT EDIT */ /* Generated on Sun Nov 25 07:41:53 EST 2012 */ #include "codelet-rdft.h" #ifdef HAVE_FMA /* Generated by: ../../../genfft/gen_hc2c.native -fma -reorder-insns -schedule-for-pipeline -compact -variables 4 -pipeline-latency 4 -sign 1 -n 10 -dif -name hc2cb_10 -include hc2cb.h */ /* * This function contains 102 FP additions, 72 FP multiplications, * (or, 48 additions, 18 multiplications, 54 fused multiply/add), * 71 stack variables, 4 constants, and 40 memory accesses */ #include "hc2cb.h" static void hc2cb_10(R *Rp, R *Ip, R *Rm, R *Im, const R *W, stride rs, INT mb, INT me, INT ms) { DK(KP951056516, +0.951056516295153572116439333379382143405698634); DK(KP559016994, +0.559016994374947424102293417182819058860154590); DK(KP250000000, +0.250000000000000000000000000000000000000000000); DK(KP618033988, +0.618033988749894848204586834365638117720309180); { INT m; for (m = mb, W = W + ((mb - 1) * 18); m < me; m = m + 1, Rp = Rp + ms, Ip = Ip + ms, Rm = Rm - ms, Im = Im - ms, W = W + 18, MAKE_VOLATILE_STRIDE(40, rs)) { E T21, T1Y, T1X; { E T1B, TH, T1g, T3, T1V, T1x, T1G, T1E, TM, TK, T11, TB, T7, T1m, T1J; E TO, Th, T1h, T6, T8, TF, TG, T1i, T9; TF = Ip[0]; TG = Im[WS(rs, 4)]; { E T1u, Tp, Tu, T1s, Tz, T1v, Ts, Tv; { E Tx, Ty, Tn, To, Tq, Tr; Tn = Ip[WS(rs, 4)]; To = Im[0]; Tx = Ip[WS(rs, 3)]; T1B = TF + TG; TH = TF - TG; T1u = Tn + To; Tp = Tn - To; Ty = Im[WS(rs, 1)]; Tq = Ip[WS(rs, 1)]; Tr = Im[WS(rs, 3)]; Tu = Ip[WS(rs, 2)]; T1s = Tx + Ty; Tz = Tx - Ty; T1v = Tq + Tr; Ts = Tq - Tr; Tv = Im[WS(rs, 2)]; } { E T1, T1w, T1D, TJ, Tt, T1r, Tw, T2; T1 = Rp[0]; T1w = T1u + T1v; T1D = T1u - T1v; TJ = Tp + Ts; Tt = Tp - Ts; T1r = Tu + Tv; Tw = Tu - Tv; T2 = Rm[WS(rs, 4)]; { E Tb, Tc, Te, Tf; Tb = Rp[WS(rs, 4)]; { E T1t, T1C, TI, TA; T1t = T1r + T1s; T1C = T1r - T1s; TI = Tw + Tz; TA = Tw - Tz; T1g = T1 - T2; T3 = T1 + T2; T1V = FNMS(KP618033988, T1t, T1w); T1x = FMA(KP618033988, T1w, T1t); T1G = T1C - T1D; T1E = T1C + T1D; TM = TI - TJ; TK = TI + TJ; T11 = FMA(KP618033988, Tt, TA); TB = FNMS(KP618033988, TA, Tt); Tc = Rm[0]; } Te = Rm[WS(rs, 3)]; Tf = Rp[WS(rs, 1)]; { E T4, T1k, Td, T1l, Tg, T5; T4 = Rp[WS(rs, 2)]; T1k = Tb - Tc; Td = Tb + Tc; T1l = Te - Tf; Tg = Te + Tf; T5 = Rm[WS(rs, 2)]; T7 = Rm[WS(rs, 1)]; T1m = T1k + T1l; T1J = T1k - T1l; TO = Td - Tg; Th = Td + Tg; T1h = T4 - T5; T6 = T4 + T5; T8 = Rp[WS(rs, 3)]; } } } } Rm[0] = TH + TK; T1i = T7 - T8; T9 = T7 + T8; { E T2d, T1F, T29, T1I, TP, T2c, T1p, Tl, T1o, Tk, T2b, T2e, T17, T14, T13; T2d = T1B + T1E; T1F = FNMS(KP250000000, T1E, T1B); { E T1j, Ta, T1n, Ti, T2a; T29 = W[8]; T1I = T1h - T1i; T1j = T1h + T1i; TP = T6 - T9; Ta = T6 + T9; T2c = W[9]; T1p = T1j - T1m; T1n = T1j + T1m; Tl = Ta - Th; Ti = Ta + Th; T1o = FNMS(KP250000000, T1n, T1g); T2a = T1g + T1n; Rp[0] = T3 + Ti; Tk = FNMS(KP250000000, Ti, T3); T2b = T29 * T2a; T2e = T2c * T2a; } { E T16, TQ, T10, Tm, TL; T16 = FMA(KP618033988, TO, TP); TQ = FNMS(KP618033988, TP, TO); Ip[WS(rs, 2)] = FNMS(T2c, T2d, T2b); Im[WS(rs, 2)] = FMA(T29, T2d, T2e); T10 = FMA(KP559016994, Tl, Tk); Tm = FNMS(KP559016994, Tl, Tk); TL = FNMS(KP250000000, TK, TH); { E TE, TU, T12, TR, TX, T1d, T1c, T19, TD, T1e, T1b, TW, TT; { E TC, T15, T1a, TS, Tj, TN; TE = W[3]; TC = FMA(KP951056516, TB, Tm); TU = FNMS(KP951056516, TB, Tm); TN = FNMS(KP559016994, TM, TL); T15 = FMA(KP559016994, TM, TL); T12 = FMA(KP951056516, T11, T10); T1a = FNMS(KP951056516, T11, T10); TS = TE * TC; TR = FNMS(KP951056516, TQ, TN); TX = FMA(KP951056516, TQ, TN); Tj = W[2]; T1d = FMA(KP951056516, T16, T15); T17 = FNMS(KP951056516, T16, T15); T1c = W[11]; T19 = W[10]; Rm[WS(rs, 1)] = FMA(Tj, TR, TS); TD = Tj * TC; T1e = T1c * T1a; T1b = T19 * T1a; } Rp[WS(rs, 1)] = FNMS(TE, TR, TD); Rm[WS(rs, 3)] = FMA(T19, T1d, T1e); Rp[WS(rs, 3)] = FNMS(T1c, T1d, T1b); TW = W[15]; TT = W[14]; { E TZ, T18, TY, TV; T14 = W[7]; TY = TW * TU; TV = TT * TU; TZ = W[6]; T18 = T14 * T12; Rm[WS(rs, 4)] = FMA(TT, TX, TY); Rp[WS(rs, 4)] = FNMS(TW, TX, TV); T13 = TZ * T12; Rm[WS(rs, 2)] = FMA(TZ, T17, T18); } } } { E T20, T1K, T1q, T1U; T20 = FNMS(KP618033988, T1I, T1J); T1K = FMA(KP618033988, T1J, T1I); Rp[WS(rs, 2)] = FNMS(T14, T17, T13); T1q = FMA(KP559016994, T1p, T1o); T1U = FNMS(KP559016994, T1p, T1o); { E T1A, T1O, T1W, T1R, T1L, T27, T26, T23, T1z, T28, T25, T1Q, T1N; { E T1y, T1Z, T24, T1M, T1f, T1H; T1A = W[1]; T1O = FMA(KP951056516, T1x, T1q); T1y = FNMS(KP951056516, T1x, T1q); T1Z = FNMS(KP559016994, T1G, T1F); T1H = FMA(KP559016994, T1G, T1F); T24 = FMA(KP951056516, T1V, T1U); T1W = FNMS(KP951056516, T1V, T1U); T1M = T1A * T1y; T1R = FNMS(KP951056516, T1K, T1H); T1L = FMA(KP951056516, T1K, T1H); T1f = W[0]; T21 = FMA(KP951056516, T20, T1Z); T27 = FNMS(KP951056516, T20, T1Z); T26 = W[13]; T23 = W[12]; Im[0] = FMA(T1f, T1L, T1M); T1z = T1f * T1y; T28 = T26 * T24; T25 = T23 * T24; } Ip[0] = FNMS(T1A, T1L, T1z); Im[WS(rs, 3)] = FMA(T23, T27, T28); Ip[WS(rs, 3)] = FNMS(T26, T27, T25); T1Q = W[17]; T1N = W[16]; { E T1T, T22, T1S, T1P; T1Y = W[5]; T1S = T1Q * T1O; T1P = T1N * T1O; T1T = W[4]; T22 = T1Y * T1W; Im[WS(rs, 4)] = FMA(T1N, T1R, T1S); Ip[WS(rs, 4)] = FNMS(T1Q, T1R, T1P); T1X = T1T * T1W; Im[WS(rs, 1)] = FMA(T1T, T21, T22); } } } } } Ip[WS(rs, 1)] = FNMS(T1Y, T21, T1X); } } } static const tw_instr twinstr[] = { {TW_FULL, 1, 10}, {TW_NEXT, 1, 0} }; static const hc2c_desc desc = { 10, "hc2cb_10", twinstr, &GENUS, {48, 18, 54, 0} }; void X(codelet_hc2cb_10) (planner *p) { X(khc2c_register) (p, hc2cb_10, &desc, HC2C_VIA_RDFT); } #else /* HAVE_FMA */ /* Generated by: ../../../genfft/gen_hc2c.native -compact -variables 4 -pipeline-latency 4 -sign 1 -n 10 -dif -name hc2cb_10 -include hc2cb.h */ /* * This function contains 102 FP additions, 60 FP multiplications, * (or, 72 additions, 30 multiplications, 30 fused multiply/add), * 39 stack variables, 4 constants, and 40 memory accesses */ #include "hc2cb.h" static void hc2cb_10(R *Rp, R *Ip, R *Rm, R *Im, const R *W, stride rs, INT mb, INT me, INT ms) { DK(KP250000000, +0.250000000000000000000000000000000000000000000); DK(KP951056516, +0.951056516295153572116439333379382143405698634); DK(KP587785252, +0.587785252292473129168705954639072768597652438); DK(KP559016994, +0.559016994374947424102293417182819058860154590); { INT m; for (m = mb, W = W + ((mb - 1) * 18); m < me; m = m + 1, Rp = Rp + ms, Ip = Ip + ms, Rm = Rm - ms, Im = Im - ms, W = W + 18, MAKE_VOLATILE_STRIDE(40, rs)) { E T3, T18, TJ, T1i, TE, TF, T1B, T1A, T1f, T1t, Ti, Tl, Tt, TA, T1w; E T1v, T1p, T1E, TM, TO; { E T1, T2, TH, TI; T1 = Rp[0]; T2 = Rm[WS(rs, 4)]; T3 = T1 + T2; T18 = T1 - T2; TH = Ip[0]; TI = Im[WS(rs, 4)]; TJ = TH - TI; T1i = TH + TI; } { E T6, T19, Tg, T1d, T9, T1a, Td, T1c; { E T4, T5, Te, Tf; T4 = Rp[WS(rs, 2)]; T5 = Rm[WS(rs, 2)]; T6 = T4 + T5; T19 = T4 - T5; Te = Rm[WS(rs, 3)]; Tf = Rp[WS(rs, 1)]; Tg = Te + Tf; T1d = Te - Tf; } { E T7, T8, Tb, Tc; T7 = Rm[WS(rs, 1)]; T8 = Rp[WS(rs, 3)]; T9 = T7 + T8; T1a = T7 - T8; Tb = Rp[WS(rs, 4)]; Tc = Rm[0]; Td = Tb + Tc; T1c = Tb - Tc; } TE = T6 - T9; TF = Td - Tg; T1B = T1c - T1d; T1A = T19 - T1a; { E T1b, T1e, Ta, Th; T1b = T19 + T1a; T1e = T1c + T1d; T1f = T1b + T1e; T1t = KP559016994 * (T1b - T1e); Ta = T6 + T9; Th = Td + Tg; Ti = Ta + Th; Tl = KP559016994 * (Ta - Th); } } { E Tp, T1j, Tz, T1n, Ts, T1k, Tw, T1m; { E Tn, To, Tx, Ty; Tn = Ip[WS(rs, 2)]; To = Im[WS(rs, 2)]; Tp = Tn - To; T1j = Tn + To; Tx = Ip[WS(rs, 1)]; Ty = Im[WS(rs, 3)]; Tz = Tx - Ty; T1n = Tx + Ty; } { E Tq, Tr, Tu, Tv; Tq = Ip[WS(rs, 3)]; Tr = Im[WS(rs, 1)]; Ts = Tq - Tr; T1k = Tq + Tr; Tu = Ip[WS(rs, 4)]; Tv = Im[0]; Tw = Tu - Tv; T1m = Tu + Tv; } Tt = Tp - Ts; TA = Tw - Tz; T1w = T1m + T1n; T1v = T1j + T1k; { E T1l, T1o, TK, TL; T1l = T1j - T1k; T1o = T1m - T1n; T1p = T1l + T1o; T1E = KP559016994 * (T1l - T1o); TK = Tp + Ts; TL = Tw + Tz; TM = TK + TL; TO = KP559016994 * (TK - TL); } } Rp[0] = T3 + Ti; Rm[0] = TJ + TM; { E T1g, T1q, T17, T1h; T1g = T18 + T1f; T1q = T1i + T1p; T17 = W[8]; T1h = W[9]; Ip[WS(rs, 2)] = FNMS(T1h, T1q, T17 * T1g); Im[WS(rs, 2)] = FMA(T1h, T1g, T17 * T1q); } { E TB, TG, T11, TX, TP, T10, Tm, TW, TN, Tk; TB = FNMS(KP951056516, TA, KP587785252 * Tt); TG = FNMS(KP951056516, TF, KP587785252 * TE); T11 = FMA(KP951056516, TE, KP587785252 * TF); TX = FMA(KP951056516, Tt, KP587785252 * TA); TN = FNMS(KP250000000, TM, TJ); TP = TN - TO; T10 = TO + TN; Tk = FNMS(KP250000000, Ti, T3); Tm = Tk - Tl; TW = Tl + Tk; { E TC, TQ, Tj, TD; TC = Tm - TB; TQ = TG + TP; Tj = W[2]; TD = W[3]; Rp[WS(rs, 1)] = FNMS(TD, TQ, Tj * TC); Rm[WS(rs, 1)] = FMA(TD, TC, Tj * TQ); } { E T14, T16, T13, T15; T14 = TW - TX; T16 = T11 + T10; T13 = W[10]; T15 = W[11]; Rp[WS(rs, 3)] = FNMS(T15, T16, T13 * T14); Rm[WS(rs, 3)] = FMA(T15, T14, T13 * T16); } { E TS, TU, TR, TT; TS = Tm + TB; TU = TP - TG; TR = W[14]; TT = W[15]; Rp[WS(rs, 4)] = FNMS(TT, TU, TR * TS); Rm[WS(rs, 4)] = FMA(TT, TS, TR * TU); } { E TY, T12, TV, TZ; TY = TW + TX; T12 = T10 - T11; TV = W[6]; TZ = W[7]; Rp[WS(rs, 2)] = FNMS(TZ, T12, TV * TY); Rm[WS(rs, 2)] = FMA(TZ, TY, TV * T12); } } { E T1x, T1C, T1Q, T1N, T1F, T1R, T1u, T1M, T1D, T1s; T1x = FNMS(KP951056516, T1w, KP587785252 * T1v); T1C = FNMS(KP951056516, T1B, KP587785252 * T1A); T1Q = FMA(KP951056516, T1A, KP587785252 * T1B); T1N = FMA(KP951056516, T1v, KP587785252 * T1w); T1D = FNMS(KP250000000, T1p, T1i); T1F = T1D - T1E; T1R = T1E + T1D; T1s = FNMS(KP250000000, T1f, T18); T1u = T1s - T1t; T1M = T1t + T1s; { E T1y, T1G, T1r, T1z; T1y = T1u - T1x; T1G = T1C + T1F; T1r = W[12]; T1z = W[13]; Ip[WS(rs, 3)] = FNMS(T1z, T1G, T1r * T1y); Im[WS(rs, 3)] = FMA(T1r, T1G, T1z * T1y); } { E T1U, T1W, T1T, T1V; T1U = T1M + T1N; T1W = T1R - T1Q; T1T = W[16]; T1V = W[17]; Ip[WS(rs, 4)] = FNMS(T1V, T1W, T1T * T1U); Im[WS(rs, 4)] = FMA(T1T, T1W, T1V * T1U); } { E T1I, T1K, T1H, T1J; T1I = T1u + T1x; T1K = T1F - T1C; T1H = W[4]; T1J = W[5]; Ip[WS(rs, 1)] = FNMS(T1J, T1K, T1H * T1I); Im[WS(rs, 1)] = FMA(T1H, T1K, T1J * T1I); } { E T1O, T1S, T1L, T1P; T1O = T1M - T1N; T1S = T1Q + T1R; T1L = W[0]; T1P = W[1]; Ip[0] = FNMS(T1P, T1S, T1L * T1O); Im[0] = FMA(T1L, T1S, T1P * T1O); } } } } } static const tw_instr twinstr[] = { {TW_FULL, 1, 10}, {TW_NEXT, 1, 0} }; static const hc2c_desc desc = { 10, "hc2cb_10", twinstr, &GENUS, {72, 30, 30, 0} }; void X(codelet_hc2cb_10) (planner *p) { X(khc2c_register) (p, hc2cb_10, &desc, HC2C_VIA_RDFT); } #endif /* HAVE_FMA */ fftw-3.3.3/rdft/scalar/r2cb/hc2cb_12.c0000644000175000017500000003531412054411021014066 00000000000000/* * Copyright (c) 2003, 2007-11 Matteo Frigo * Copyright (c) 2003, 2007-11 Massachusetts Institute of Technology * * This program is free software; you can redistribute it and/or modify * it under the terms of the GNU General Public License as published by * the Free Software Foundation; either version 2 of the License, or * (at your option) any later version. * * This program is distributed in the hope that it will be useful, * but WITHOUT ANY WARRANTY; without even the implied warranty of * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the * GNU General Public License for more details. * * You should have received a copy of the GNU General Public License * along with this program; if not, write to the Free Software * Foundation, Inc., 51 Franklin Street, Fifth Floor, Boston, MA 02110-1301 USA * */ /* This file was automatically generated --- DO NOT EDIT */ /* Generated on Sun Nov 25 07:41:53 EST 2012 */ #include "codelet-rdft.h" #ifdef HAVE_FMA /* Generated by: ../../../genfft/gen_hc2c.native -fma -reorder-insns -schedule-for-pipeline -compact -variables 4 -pipeline-latency 4 -sign 1 -n 12 -dif -name hc2cb_12 -include hc2cb.h */ /* * This function contains 118 FP additions, 68 FP multiplications, * (or, 72 additions, 22 multiplications, 46 fused multiply/add), * 64 stack variables, 2 constants, and 48 memory accesses */ #include "hc2cb.h" static void hc2cb_12(R *Rp, R *Ip, R *Rm, R *Im, const R *W, stride rs, INT mb, INT me, INT ms) { DK(KP866025403, +0.866025403784438646763723170752936183471402627); DK(KP500000000, +0.500000000000000000000000000000000000000000000); { INT m; for (m = mb, W = W + ((mb - 1) * 22); m < me; m = m + 1, Rp = Rp + ms, Ip = Ip + ms, Rm = Rm - ms, Im = Im - ms, W = W + 22, MAKE_VOLATILE_STRIDE(48, rs)) { E T1U, T1X, T1W, T1Y, T1V; { E T18, T20, T21, T1b, T2a, T1s, T29, T1p, TO, T11, To, Tb, Tg, T23, T1f; E Tl, Ty, Tt, T1i, T24, T1z, T2d, T1w, T2c; { E T5, Ta, TN, TI; { E T1, TE, T6, TM, T7, T1o, T4, T17, TH, T8, TJ, TK; T1 = Rp[0]; TE = Ip[0]; T6 = Rm[WS(rs, 5)]; TM = Im[WS(rs, 5)]; { E T2, T3, TF, TG; T2 = Rp[WS(rs, 4)]; T3 = Rm[WS(rs, 3)]; TF = Ip[WS(rs, 4)]; TG = Im[WS(rs, 3)]; T7 = Rm[WS(rs, 1)]; T1o = T2 - T3; T4 = T2 + T3; T17 = TF + TG; TH = TF - TG; T8 = Rp[WS(rs, 2)]; TJ = Ip[WS(rs, 2)]; TK = Im[WS(rs, 1)]; } { E T1r, T1a, T19, T1q, T9, TL, T16, T1n; T5 = T1 + T4; T16 = FNMS(KP500000000, T4, T1); T1r = T7 - T8; T9 = T7 + T8; T1a = TJ + TK; TL = TJ - TK; T18 = FNMS(KP866025403, T17, T16); T20 = FMA(KP866025403, T17, T16); T19 = FNMS(KP500000000, T9, T6); Ta = T6 + T9; TN = TL - TM; T1q = FMA(KP500000000, TL, TM); T1n = FNMS(KP500000000, TH, TE); TI = TE + TH; T21 = FNMS(KP866025403, T1a, T19); T1b = FMA(KP866025403, T1a, T19); T2a = FMA(KP866025403, T1r, T1q); T1s = FNMS(KP866025403, T1r, T1q); T29 = FNMS(KP866025403, T1o, T1n); T1p = FMA(KP866025403, T1o, T1n); } } { E Tc, Tp, Th, Tx, Ti, Tf, T1v, Ts, T1e, Tj, Tu, Tv; Tc = Rp[WS(rs, 3)]; TO = TI - TN; T11 = TI + TN; Tp = Ip[WS(rs, 3)]; To = T5 - Ta; Tb = T5 + Ta; Th = Rm[WS(rs, 2)]; Tx = Im[WS(rs, 2)]; { E Td, Te, Tq, Tr; Td = Rm[WS(rs, 4)]; Te = Rm[0]; Tq = Im[WS(rs, 4)]; Tr = Im[0]; Ti = Rp[WS(rs, 1)]; Tf = Td + Te; T1v = Td - Te; Ts = Tq + Tr; T1e = Tq - Tr; Tj = Rp[WS(rs, 5)]; Tu = Ip[WS(rs, 1)]; Tv = Ip[WS(rs, 5)]; } { E T1y, T1h, T1g, T1x, Tk, Tw, T1d, T1u; T1d = FNMS(KP500000000, Tf, Tc); Tg = Tc + Tf; Tk = Ti + Tj; T1y = Ti - Tj; Tw = Tu + Tv; T1h = Tv - Tu; T23 = FNMS(KP866025403, T1e, T1d); T1f = FMA(KP866025403, T1e, T1d); Tl = Th + Tk; T1g = FNMS(KP500000000, Tk, Th); T1x = FMA(KP500000000, Tw, Tx); Ty = Tw - Tx; Tt = Tp - Ts; T1u = FMA(KP500000000, Ts, Tp); T1i = FMA(KP866025403, T1h, T1g); T24 = FNMS(KP866025403, T1h, T1g); T1z = FNMS(KP866025403, T1y, T1x); T2d = FMA(KP866025403, T1y, T1x); T1w = FMA(KP866025403, T1v, T1u); T2c = FNMS(KP866025403, T1v, T1u); } } } { E TY, T13, TX, T10; { E Tn, T12, TC, Tm, TD, TS, TA, Tz; Tn = W[16]; T12 = Tt + Ty; Tz = Tt - Ty; TC = W[17]; Tm = Tg + Tl; TD = Tg - Tl; TS = To + Tz; TA = To - Tz; { E TV, TU, TW, TT; { E TQ, TR, TP, TB; TV = TO - TD; TP = TD + TO; Rp[0] = Tb + Tm; TB = Tn * TA; TQ = Tn * TP; TR = W[4]; Ip[WS(rs, 4)] = FNMS(TC, TP, TB); TU = W[5]; Im[WS(rs, 4)] = FMA(TC, TA, TQ); TW = TR * TV; TT = TR * TS; } Im[WS(rs, 1)] = FMA(TU, TS, TW); Ip[WS(rs, 1)] = FNMS(TU, TV, TT); TY = Tb - Tm; T13 = T11 - T12; TX = W[10]; T10 = W[11]; Rm[0] = T11 + T12; } } { E T1K, T1Q, T1P, T1L, T2o, T2u, T2t, T2p; { E T1E, T1D, T1H, T1F, T1G, T1t, T1k, T1A; { E T1c, TZ, T14, T1j; T1K = T18 - T1b; T1c = T18 + T1b; TZ = TX * TY; T14 = T10 * TY; T1j = T1f + T1i; T1Q = T1f - T1i; T1P = T1p + T1s; T1t = T1p - T1s; Rp[WS(rs, 3)] = FNMS(T10, T13, TZ); Rm[WS(rs, 3)] = FMA(TX, T13, T14); T1E = T1c + T1j; T1k = T1c - T1j; T1A = T1w - T1z; T1L = T1w + T1z; } { E T15, T1m, T1B, T1l, T1C; T15 = W[18]; T1m = W[19]; T1D = W[6]; T1H = T1t + T1A; T1B = T1t - T1A; T1l = T15 * T1k; T1C = T1m * T1k; T1F = T1D * T1E; T1G = W[7]; Rp[WS(rs, 5)] = FNMS(T1m, T1B, T1l); Rm[WS(rs, 5)] = FMA(T15, T1B, T1C); } { E T26, T2i, T2l, T2f, T1Z, T28; { E T22, T1I, T25, T2b, T2e; T22 = T20 + T21; T2o = T20 - T21; Rp[WS(rs, 2)] = FNMS(T1G, T1H, T1F); T1I = T1G * T1E; T2u = T23 - T24; T25 = T23 + T24; T2b = T29 - T2a; T2t = T29 + T2a; T2p = T2c + T2d; T2e = T2c - T2d; Rm[WS(rs, 2)] = FMA(T1D, T1H, T1I); T26 = T22 - T25; T2i = T22 + T25; T2l = T2b + T2e; T2f = T2b - T2e; } T1Z = W[2]; T28 = W[3]; { E T2h, T2k, T27, T2g, T2j, T2m; T2h = W[14]; T2k = W[15]; T27 = T1Z * T26; T2g = T28 * T26; T2j = T2h * T2i; T2m = T2k * T2i; Rp[WS(rs, 1)] = FNMS(T28, T2f, T27); Rm[WS(rs, 1)] = FMA(T1Z, T2f, T2g); Rp[WS(rs, 4)] = FNMS(T2k, T2l, T2j); Rm[WS(rs, 4)] = FMA(T2h, T2l, T2m); } } } { E T2y, T2B, T2A, T2C, T2z; { E T2n, T2q, T2v, T2s, T2r, T2x, T2w; T2n = W[8]; T2y = T2o + T2p; T2q = T2o - T2p; T2B = T2t - T2u; T2v = T2t + T2u; T2s = W[9]; T2r = T2n * T2q; T2x = W[20]; T2w = T2n * T2v; T2A = W[21]; Ip[WS(rs, 2)] = FNMS(T2s, T2v, T2r); T2C = T2x * T2B; T2z = T2x * T2y; Im[WS(rs, 2)] = FMA(T2s, T2q, T2w); } Im[WS(rs, 5)] = FMA(T2A, T2y, T2C); Ip[WS(rs, 5)] = FNMS(T2A, T2B, T2z); { E T1J, T1M, T1R, T1O, T1N, T1T, T1S; T1J = W[0]; T1U = T1K + T1L; T1M = T1K - T1L; T1X = T1P - T1Q; T1R = T1P + T1Q; T1O = W[1]; T1N = T1J * T1M; T1T = W[12]; T1S = T1J * T1R; T1W = W[13]; Ip[0] = FNMS(T1O, T1R, T1N); T1Y = T1T * T1X; T1V = T1T * T1U; Im[0] = FMA(T1O, T1M, T1S); } } } } } Im[WS(rs, 3)] = FMA(T1W, T1U, T1Y); Ip[WS(rs, 3)] = FNMS(T1W, T1X, T1V); } } } static const tw_instr twinstr[] = { {TW_FULL, 1, 12}, {TW_NEXT, 1, 0} }; static const hc2c_desc desc = { 12, "hc2cb_12", twinstr, &GENUS, {72, 22, 46, 0} }; void X(codelet_hc2cb_12) (planner *p) { X(khc2c_register) (p, hc2cb_12, &desc, HC2C_VIA_RDFT); } #else /* HAVE_FMA */ /* Generated by: ../../../genfft/gen_hc2c.native -compact -variables 4 -pipeline-latency 4 -sign 1 -n 12 -dif -name hc2cb_12 -include hc2cb.h */ /* * This function contains 118 FP additions, 60 FP multiplications, * (or, 88 additions, 30 multiplications, 30 fused multiply/add), * 39 stack variables, 2 constants, and 48 memory accesses */ #include "hc2cb.h" static void hc2cb_12(R *Rp, R *Ip, R *Rm, R *Im, const R *W, stride rs, INT mb, INT me, INT ms) { DK(KP500000000, +0.500000000000000000000000000000000000000000000); DK(KP866025403, +0.866025403784438646763723170752936183471402627); { INT m; for (m = mb, W = W + ((mb - 1) * 22); m < me; m = m + 1, Rp = Rp + ms, Ip = Ip + ms, Rm = Rm - ms, Im = Im - ms, W = W + 22, MAKE_VOLATILE_STRIDE(48, rs)) { E T5, TH, T12, T1M, T1i, T1U, Tl, Ty, T1c, T1Y, T1s, T1Q, Ta, TM, T15; E T1N, T1l, T1V, Tg, Tt, T19, T1X, T1p, T1P; { E T1, TD, T4, T1g, TG, T11, T10, T1h; T1 = Rp[0]; TD = Ip[0]; { E T2, T3, TE, TF; T2 = Rp[WS(rs, 4)]; T3 = Rm[WS(rs, 3)]; T4 = T2 + T3; T1g = KP866025403 * (T2 - T3); TE = Ip[WS(rs, 4)]; TF = Im[WS(rs, 3)]; TG = TE - TF; T11 = KP866025403 * (TE + TF); } T5 = T1 + T4; TH = TD + TG; T10 = FNMS(KP500000000, T4, T1); T12 = T10 - T11; T1M = T10 + T11; T1h = FNMS(KP500000000, TG, TD); T1i = T1g + T1h; T1U = T1h - T1g; } { E Th, Tx, Tk, T1a, Tw, T1r, T1b, T1q; Th = Rm[WS(rs, 2)]; Tx = Im[WS(rs, 2)]; { E Ti, Tj, Tu, Tv; Ti = Rp[WS(rs, 1)]; Tj = Rp[WS(rs, 5)]; Tk = Ti + Tj; T1a = KP866025403 * (Ti - Tj); Tu = Ip[WS(rs, 1)]; Tv = Ip[WS(rs, 5)]; Tw = Tu + Tv; T1r = KP866025403 * (Tv - Tu); } Tl = Th + Tk; Ty = Tw - Tx; T1b = FMA(KP500000000, Tw, Tx); T1c = T1a - T1b; T1Y = T1a + T1b; T1q = FNMS(KP500000000, Tk, Th); T1s = T1q + T1r; T1Q = T1q - T1r; } { E T6, TL, T9, T1j, TK, T14, T13, T1k; T6 = Rm[WS(rs, 5)]; TL = Im[WS(rs, 5)]; { E T7, T8, TI, TJ; T7 = Rm[WS(rs, 1)]; T8 = Rp[WS(rs, 2)]; T9 = T7 + T8; T1j = KP866025403 * (T7 - T8); TI = Ip[WS(rs, 2)]; TJ = Im[WS(rs, 1)]; TK = TI - TJ; T14 = KP866025403 * (TI + TJ); } Ta = T6 + T9; TM = TK - TL; T13 = FNMS(KP500000000, T9, T6); T15 = T13 + T14; T1N = T13 - T14; T1k = FMA(KP500000000, TK, TL); T1l = T1j - T1k; T1V = T1j + T1k; } { E Tc, Tp, Tf, T17, Ts, T1o, T18, T1n; Tc = Rp[WS(rs, 3)]; Tp = Ip[WS(rs, 3)]; { E Td, Te, Tq, Tr; Td = Rm[WS(rs, 4)]; Te = Rm[0]; Tf = Td + Te; T17 = KP866025403 * (Td - Te); Tq = Im[WS(rs, 4)]; Tr = Im[0]; Ts = Tq + Tr; T1o = KP866025403 * (Tq - Tr); } Tg = Tc + Tf; Tt = Tp - Ts; T18 = FMA(KP500000000, Ts, Tp); T19 = T17 + T18; T1X = T18 - T17; T1n = FNMS(KP500000000, Tf, Tc); T1p = T1n + T1o; T1P = T1n - T1o; } { E Tb, Tm, TU, TW, TX, TY, TT, TV; Tb = T5 + Ta; Tm = Tg + Tl; TU = Tb - Tm; TW = TH + TM; TX = Tt + Ty; TY = TW - TX; Rp[0] = Tb + Tm; Rm[0] = TW + TX; TT = W[10]; TV = W[11]; Rp[WS(rs, 3)] = FNMS(TV, TY, TT * TU); Rm[WS(rs, 3)] = FMA(TV, TU, TT * TY); } { E TA, TQ, TO, TS; { E To, Tz, TC, TN; To = T5 - Ta; Tz = Tt - Ty; TA = To - Tz; TQ = To + Tz; TC = Tg - Tl; TN = TH - TM; TO = TC + TN; TS = TN - TC; } { E Tn, TB, TP, TR; Tn = W[16]; TB = W[17]; Ip[WS(rs, 4)] = FNMS(TB, TO, Tn * TA); Im[WS(rs, 4)] = FMA(Tn, TO, TB * TA); TP = W[4]; TR = W[5]; Ip[WS(rs, 1)] = FNMS(TR, TS, TP * TQ); Im[WS(rs, 1)] = FMA(TP, TS, TR * TQ); } } { E T28, T2e, T2c, T2g; { E T26, T27, T2a, T2b; T26 = T1M - T1N; T27 = T1X + T1Y; T28 = T26 - T27; T2e = T26 + T27; T2a = T1U + T1V; T2b = T1P - T1Q; T2c = T2a + T2b; T2g = T2a - T2b; } { E T25, T29, T2d, T2f; T25 = W[8]; T29 = W[9]; Ip[WS(rs, 2)] = FNMS(T29, T2c, T25 * T28); Im[WS(rs, 2)] = FMA(T25, T2c, T29 * T28); T2d = W[20]; T2f = W[21]; Ip[WS(rs, 5)] = FNMS(T2f, T2g, T2d * T2e); Im[WS(rs, 5)] = FMA(T2d, T2g, T2f * T2e); } } { E T1S, T22, T20, T24; { E T1O, T1R, T1W, T1Z; T1O = T1M + T1N; T1R = T1P + T1Q; T1S = T1O - T1R; T22 = T1O + T1R; T1W = T1U - T1V; T1Z = T1X - T1Y; T20 = T1W - T1Z; T24 = T1W + T1Z; } { E T1L, T1T, T21, T23; T1L = W[2]; T1T = W[3]; Rp[WS(rs, 1)] = FNMS(T1T, T20, T1L * T1S); Rm[WS(rs, 1)] = FMA(T1T, T1S, T1L * T20); T21 = W[14]; T23 = W[15]; Rp[WS(rs, 4)] = FNMS(T23, T24, T21 * T22); Rm[WS(rs, 4)] = FMA(T23, T22, T21 * T24); } } { E T1C, T1I, T1G, T1K; { E T1A, T1B, T1E, T1F; T1A = T12 + T15; T1B = T1p + T1s; T1C = T1A - T1B; T1I = T1A + T1B; T1E = T1i + T1l; T1F = T19 + T1c; T1G = T1E - T1F; T1K = T1E + T1F; } { E T1z, T1D, T1H, T1J; T1z = W[18]; T1D = W[19]; Rp[WS(rs, 5)] = FNMS(T1D, T1G, T1z * T1C); Rm[WS(rs, 5)] = FMA(T1D, T1C, T1z * T1G); T1H = W[6]; T1J = W[7]; Rp[WS(rs, 2)] = FNMS(T1J, T1K, T1H * T1I); Rm[WS(rs, 2)] = FMA(T1J, T1I, T1H * T1K); } } { E T1e, T1w, T1u, T1y; { E T16, T1d, T1m, T1t; T16 = T12 - T15; T1d = T19 - T1c; T1e = T16 - T1d; T1w = T16 + T1d; T1m = T1i - T1l; T1t = T1p - T1s; T1u = T1m + T1t; T1y = T1m - T1t; } { E TZ, T1f, T1v, T1x; TZ = W[0]; T1f = W[1]; Ip[0] = FNMS(T1f, T1u, TZ * T1e); Im[0] = FMA(TZ, T1u, T1f * T1e); T1v = W[12]; T1x = W[13]; Ip[WS(rs, 3)] = FNMS(T1x, T1y, T1v * T1w); Im[WS(rs, 3)] = FMA(T1v, T1y, T1x * T1w); } } } } } static const tw_instr twinstr[] = { {TW_FULL, 1, 12}, {TW_NEXT, 1, 0} }; static const hc2c_desc desc = { 12, "hc2cb_12", twinstr, &GENUS, {88, 30, 30, 0} }; void X(codelet_hc2cb_12) (planner *p) { X(khc2c_register) (p, hc2cb_12, &desc, HC2C_VIA_RDFT); } #endif /* HAVE_FMA */ fftw-3.3.3/rdft/scalar/r2cb/hc2cb_16.c0000644000175000017500000005014312054411023014071 00000000000000/* * Copyright (c) 2003, 2007-11 Matteo Frigo * Copyright (c) 2003, 2007-11 Massachusetts Institute of Technology * * This program is free software; you can redistribute it and/or modify * it under the terms of the GNU General Public License as published by * the Free Software Foundation; either version 2 of the License, or * (at your option) any later version. * * This program is distributed in the hope that it will be useful, * but WITHOUT ANY WARRANTY; without even the implied warranty of * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the * GNU General Public License for more details. * * You should have received a copy of the GNU General Public License * along with this program; if not, write to the Free Software * Foundation, Inc., 51 Franklin Street, Fifth Floor, Boston, MA 02110-1301 USA * */ /* This file was automatically generated --- DO NOT EDIT */ /* Generated on Sun Nov 25 07:41:53 EST 2012 */ #include "codelet-rdft.h" #ifdef HAVE_FMA /* Generated by: ../../../genfft/gen_hc2c.native -fma -reorder-insns -schedule-for-pipeline -compact -variables 4 -pipeline-latency 4 -sign 1 -n 16 -dif -name hc2cb_16 -include hc2cb.h */ /* * This function contains 174 FP additions, 100 FP multiplications, * (or, 104 additions, 30 multiplications, 70 fused multiply/add), * 78 stack variables, 3 constants, and 64 memory accesses */ #include "hc2cb.h" static void hc2cb_16(R *Rp, R *Ip, R *Rm, R *Im, const R *W, stride rs, INT mb, INT me, INT ms) { DK(KP923879532, +0.923879532511286756128183189396788286822416626); DK(KP707106781, +0.707106781186547524400844362104849039284835938); DK(KP414213562, +0.414213562373095048801688724209698078569671875); { INT m; for (m = mb, W = W + ((mb - 1) * 30); m < me; m = m + 1, Rp = Rp + ms, Ip = Ip + ms, Rm = Rm - ms, Im = Im - ms, W = W + 30, MAKE_VOLATILE_STRIDE(64, rs)) { E T1I, T1L, T1K, T1M, T1J; { E T1O, TA, T1h, T21, T3b, T2T, T3D, T3r, T1k, T1P, T3y, Tf, T36, T2A, T22; E TL, T2F, T2U, T3u, T3z, T2K, T2V, T12, Tu, T3E, TX, T1n, T17, T1T, T24; E T1W, T25; { E T2z, TF, TK, T2w; { E Tw, T3, T2Q, T1g, T1d, T6, T2R, Tz, Tb, TB, Ta, T2y, TE, Tc, TH; E TI; { E T4, T5, Tx, Ty; { E T1, T2, T1e, T1f; T1 = Rp[0]; T2 = Rm[WS(rs, 7)]; T1e = Ip[0]; T1f = Im[WS(rs, 7)]; T4 = Rp[WS(rs, 4)]; Tw = T1 - T2; T3 = T1 + T2; T2Q = T1e - T1f; T1g = T1e + T1f; T5 = Rm[WS(rs, 3)]; Tx = Ip[WS(rs, 4)]; Ty = Im[WS(rs, 3)]; } { E T8, T9, TC, TD; T8 = Rp[WS(rs, 2)]; T1d = T4 - T5; T6 = T4 + T5; T2R = Tx - Ty; Tz = Tx + Ty; T9 = Rm[WS(rs, 5)]; TC = Ip[WS(rs, 2)]; TD = Im[WS(rs, 5)]; Tb = Rm[WS(rs, 1)]; TB = T8 - T9; Ta = T8 + T9; T2y = TC - TD; TE = TC + TD; Tc = Rp[WS(rs, 6)]; TH = Ip[WS(rs, 6)]; TI = Im[WS(rs, 1)]; } } { E TG, T2x, TJ, Te, T2P, T2S, T3p, Td; T1O = Tw + Tz; TA = Tw - Tz; TG = Tb - Tc; Td = Tb + Tc; T2x = TH - TI; TJ = TH + TI; T1h = T1d + T1g; T21 = T1g - T1d; Te = Ta + Td; T2P = Ta - Td; T2S = T2Q - T2R; T3p = T2Q + T2R; { E T1i, T1j, T3q, T7; T3q = T2y + T2x; T2z = T2x - T2y; TF = TB - TE; T1i = TB + TE; T3b = T2S - T2P; T2T = T2P + T2S; TK = TG - TJ; T1j = TG + TJ; T3D = T3p - T3q; T3r = T3p + T3q; T2w = T3 - T6; T7 = T3 + T6; T1k = T1i - T1j; T1P = T1i + T1j; T3y = T7 - Te; Tf = T7 + Te; } } } { E T13, Ti, T2C, T11, TY, Tl, T2D, T16, Tq, TS, Tp, T2H, TQ, Tr, TT; E TU; { E Tj, Tk, T14, T15; { E Tg, Th, TZ, T10; Tg = Rp[WS(rs, 1)]; T36 = T2w - T2z; T2A = T2w + T2z; T22 = TF - TK; TL = TF + TK; Th = Rm[WS(rs, 6)]; TZ = Ip[WS(rs, 1)]; T10 = Im[WS(rs, 6)]; Tj = Rp[WS(rs, 5)]; T13 = Tg - Th; Ti = Tg + Th; T2C = TZ - T10; T11 = TZ + T10; Tk = Rm[WS(rs, 2)]; T14 = Ip[WS(rs, 5)]; T15 = Im[WS(rs, 2)]; } { E Tn, To, TO, TP; Tn = Rm[0]; TY = Tj - Tk; Tl = Tj + Tk; T2D = T14 - T15; T16 = T14 + T15; To = Rp[WS(rs, 7)]; TO = Ip[WS(rs, 7)]; TP = Im[0]; Tq = Rp[WS(rs, 3)]; TS = Tn - To; Tp = Tn + To; T2H = TO - TP; TQ = TO + TP; Tr = Rm[WS(rs, 4)]; TT = Ip[WS(rs, 3)]; TU = Im[WS(rs, 4)]; } } { E TN, TV, Tm, Tt; { E T2E, T3s, Ts, T2B, T3t, T2J, T2I, T2G; T2E = T2C - T2D; T3s = T2C + T2D; TN = Tq - Tr; Ts = Tq + Tr; T2I = TT - TU; TV = TT + TU; T2B = Ti - Tl; Tm = Ti + Tl; T3t = T2H + T2I; T2J = T2H - T2I; Tt = Tp + Ts; T2G = Tp - Ts; T2F = T2B - T2E; T2U = T2B + T2E; T3u = T3s + T3t; T3z = T3t - T3s; T2K = T2G + T2J; T2V = T2J - T2G; } { E T1U, T1V, T1R, T1S, TR, TW; TR = TN - TQ; T1U = TN + TQ; T1V = TS + TV; TW = TS - TV; T1R = T11 - TY; T12 = TY + T11; Tu = Tm + Tt; T3E = Tm - Tt; TX = FNMS(KP414213562, TW, TR); T1n = FMA(KP414213562, TR, TW); T17 = T13 - T16; T1S = T13 + T16; T1T = FNMS(KP414213562, T1S, T1R); T24 = FMA(KP414213562, T1R, T1S); T1W = FNMS(KP414213562, T1V, T1U); T25 = FMA(KP414213562, T1U, T1V); } } } } { E T18, T1m, T2W, T2L, T3j, T3i, T3h; { E T3m, T3v, T3l, T3o; Rp[0] = Tf + Tu; T18 = FMA(KP414213562, T17, T12); T1m = FNMS(KP414213562, T12, T17); T3m = Tf - Tu; T3v = T3r - T3u; T3l = W[14]; T3o = W[15]; Rm[0] = T3r + T3u; { E T3A, T3I, T3L, T3F, T3C, T3G, T3B, T3x, T3n, T3w, T3H, T3K; T3A = T3y - T3z; T3I = T3y + T3z; T3n = T3l * T3m; T3w = T3o * T3m; T3L = T3E + T3D; T3F = T3D - T3E; T3x = W[22]; Rp[WS(rs, 4)] = FNMS(T3o, T3v, T3n); Rm[WS(rs, 4)] = FMA(T3l, T3v, T3w); T3C = W[23]; T3G = T3x * T3F; T3B = T3x * T3A; Rm[WS(rs, 6)] = FMA(T3C, T3A, T3G); Rp[WS(rs, 6)] = FNMS(T3C, T3F, T3B); T3H = W[6]; T3K = W[7]; { E T3g, T38, T3d, T35, T3a; { E T37, T3c, T3M, T3J; T37 = T2V - T2U; T2W = T2U + T2V; T2L = T2F + T2K; T3c = T2F - T2K; T3M = T3H * T3L; T3J = T3H * T3I; T3g = FMA(KP707106781, T37, T36); T38 = FNMS(KP707106781, T37, T36); Rm[WS(rs, 2)] = FMA(T3K, T3I, T3M); Rp[WS(rs, 2)] = FNMS(T3K, T3L, T3J); T3d = FNMS(KP707106781, T3c, T3b); T3j = FMA(KP707106781, T3c, T3b); } T35 = W[26]; T3a = W[27]; { E T3f, T3e, T39, T3k; T3f = W[10]; T3i = W[11]; T3e = T35 * T3d; T39 = T35 * T38; T3k = T3f * T3j; T3h = T3f * T3g; Rm[WS(rs, 7)] = FMA(T3a, T38, T3e); Rp[WS(rs, 7)] = FNMS(T3a, T3d, T39); Rm[WS(rs, 3)] = FMA(T3i, T3g, T3k); } } } } Rp[WS(rs, 3)] = FNMS(T3i, T3j, T3h); { E T2g, T2m, T2l, T2h, T2d, T29, T2c, T2b, T2e; { E T33, T2Z, T32, T31, T34; { E T2v, T30, T2M, T2X, T2O, T2N, T2Y; T2v = W[18]; T30 = FMA(KP707106781, T2L, T2A); T2M = FNMS(KP707106781, T2L, T2A); T33 = FMA(KP707106781, T2W, T2T); T2X = FNMS(KP707106781, T2W, T2T); T2O = W[19]; T2N = T2v * T2M; T2Z = W[2]; T32 = W[3]; T2Y = T2O * T2M; Rp[WS(rs, 5)] = FNMS(T2O, T2X, T2N); T31 = T2Z * T30; T34 = T32 * T30; Rm[WS(rs, 5)] = FMA(T2v, T2X, T2Y); } { E T1Q, T1X, T23, T26; T2g = FMA(KP707106781, T1P, T1O); T1Q = FNMS(KP707106781, T1P, T1O); Rp[WS(rs, 1)] = FNMS(T32, T33, T31); Rm[WS(rs, 1)] = FMA(T2Z, T33, T34); T1X = T1T + T1W; T2m = T1W - T1T; T2l = FNMS(KP707106781, T22, T21); T23 = FMA(KP707106781, T22, T21); T26 = T24 - T25; T2h = T24 + T25; { E T1N, T2a, T1Y, T27, T20, T1Z, T28; T1N = W[20]; T2a = FNMS(KP923879532, T1X, T1Q); T1Y = FMA(KP923879532, T1X, T1Q); T2d = FMA(KP923879532, T26, T23); T27 = FNMS(KP923879532, T26, T23); T20 = W[21]; T1Z = T1N * T1Y; T29 = W[4]; T2c = W[5]; T28 = T20 * T1Y; Ip[WS(rs, 5)] = FNMS(T20, T27, T1Z); T2b = T29 * T2a; T2e = T2c * T2a; Im[WS(rs, 5)] = FMA(T1N, T27, T28); } } } { E T1y, T1E, T1D, T1z, T1v, T1r, T1u, T1t, T1w; { E TM, T19, T1l, T1o; T1y = FMA(KP707106781, TL, TA); TM = FNMS(KP707106781, TL, TA); Ip[WS(rs, 1)] = FNMS(T2c, T2d, T2b); Im[WS(rs, 1)] = FMA(T29, T2d, T2e); T19 = TX - T18; T1E = T18 + TX; T1D = FMA(KP707106781, T1k, T1h); T1l = FNMS(KP707106781, T1k, T1h); T1o = T1m - T1n; T1z = T1m + T1n; { E Tv, T1s, T1a, T1p, T1c, T1b, T1q; Tv = W[24]; T1s = FMA(KP923879532, T19, TM); T1a = FNMS(KP923879532, T19, TM); T1v = FMA(KP923879532, T1o, T1l); T1p = FNMS(KP923879532, T1o, T1l); T1c = W[25]; T1b = Tv * T1a; T1r = W[8]; T1u = W[9]; T1q = T1c * T1a; Ip[WS(rs, 6)] = FNMS(T1c, T1p, T1b); T1t = T1r * T1s; T1w = T1u * T1s; Im[WS(rs, 6)] = FMA(Tv, T1p, T1q); } } { E T2q, T2t, T2s, T2u, T2r; Ip[WS(rs, 2)] = FNMS(T1u, T1v, T1t); Im[WS(rs, 2)] = FMA(T1r, T1v, T1w); { E T2f, T2i, T2n, T2k, T2j, T2p, T2o; T2f = W[12]; T2q = FMA(KP923879532, T2h, T2g); T2i = FNMS(KP923879532, T2h, T2g); T2t = FNMS(KP923879532, T2m, T2l); T2n = FMA(KP923879532, T2m, T2l); T2k = W[13]; T2j = T2f * T2i; T2p = W[28]; T2o = T2f * T2n; T2s = W[29]; Ip[WS(rs, 3)] = FNMS(T2k, T2n, T2j); T2u = T2p * T2t; T2r = T2p * T2q; Im[WS(rs, 3)] = FMA(T2k, T2i, T2o); } Im[WS(rs, 7)] = FMA(T2s, T2q, T2u); Ip[WS(rs, 7)] = FNMS(T2s, T2t, T2r); { E T1x, T1A, T1F, T1C, T1B, T1H, T1G; T1x = W[16]; T1I = FMA(KP923879532, T1z, T1y); T1A = FNMS(KP923879532, T1z, T1y); T1L = FMA(KP923879532, T1E, T1D); T1F = FNMS(KP923879532, T1E, T1D); T1C = W[17]; T1B = T1x * T1A; T1H = W[0]; T1G = T1x * T1F; T1K = W[1]; Ip[WS(rs, 4)] = FNMS(T1C, T1F, T1B); T1M = T1H * T1L; T1J = T1H * T1I; Im[WS(rs, 4)] = FMA(T1C, T1A, T1G); } } } } } } Im[0] = FMA(T1K, T1I, T1M); Ip[0] = FNMS(T1K, T1L, T1J); } } } static const tw_instr twinstr[] = { {TW_FULL, 1, 16}, {TW_NEXT, 1, 0} }; static const hc2c_desc desc = { 16, "hc2cb_16", twinstr, &GENUS, {104, 30, 70, 0} }; void X(codelet_hc2cb_16) (planner *p) { X(khc2c_register) (p, hc2cb_16, &desc, HC2C_VIA_RDFT); } #else /* HAVE_FMA */ /* Generated by: ../../../genfft/gen_hc2c.native -compact -variables 4 -pipeline-latency 4 -sign 1 -n 16 -dif -name hc2cb_16 -include hc2cb.h */ /* * This function contains 174 FP additions, 84 FP multiplications, * (or, 136 additions, 46 multiplications, 38 fused multiply/add), * 50 stack variables, 3 constants, and 64 memory accesses */ #include "hc2cb.h" static void hc2cb_16(R *Rp, R *Ip, R *Rm, R *Im, const R *W, stride rs, INT mb, INT me, INT ms) { DK(KP382683432, +0.382683432365089771728459984030398866761344562); DK(KP923879532, +0.923879532511286756128183189396788286822416626); DK(KP707106781, +0.707106781186547524400844362104849039284835938); { INT m; for (m = mb, W = W + ((mb - 1) * 30); m < me; m = m + 1, Rp = Rp + ms, Ip = Ip + ms, Rm = Rm - ms, Im = Im - ms, W = W + 30, MAKE_VOLATILE_STRIDE(64, rs)) { E T7, T2K, T2W, Tw, T17, T1S, T2k, T1w, Te, TD, T1x, T10, T2n, T2L, T1Z; E T2X, Tm, T1z, TN, T19, T2e, T2p, T2P, T2Z, Tt, T1A, TW, T1a, T27, T2q; E T2S, T30; { E T3, T1Q, T13, T2j, T6, T2i, T16, T1R; { E T1, T2, T11, T12; T1 = Rp[0]; T2 = Rm[WS(rs, 7)]; T3 = T1 + T2; T1Q = T1 - T2; T11 = Ip[0]; T12 = Im[WS(rs, 7)]; T13 = T11 - T12; T2j = T11 + T12; } { E T4, T5, T14, T15; T4 = Rp[WS(rs, 4)]; T5 = Rm[WS(rs, 3)]; T6 = T4 + T5; T2i = T4 - T5; T14 = Ip[WS(rs, 4)]; T15 = Im[WS(rs, 3)]; T16 = T14 - T15; T1R = T14 + T15; } T7 = T3 + T6; T2K = T1Q + T1R; T2W = T2j - T2i; Tw = T3 - T6; T17 = T13 - T16; T1S = T1Q - T1R; T2k = T2i + T2j; T1w = T13 + T16; } { E Ta, T1T, TC, T1U, Td, T1W, Tz, T1X; { E T8, T9, TA, TB; T8 = Rp[WS(rs, 2)]; T9 = Rm[WS(rs, 5)]; Ta = T8 + T9; T1T = T8 - T9; TA = Ip[WS(rs, 2)]; TB = Im[WS(rs, 5)]; TC = TA - TB; T1U = TA + TB; } { E Tb, Tc, Tx, Ty; Tb = Rm[WS(rs, 1)]; Tc = Rp[WS(rs, 6)]; Td = Tb + Tc; T1W = Tb - Tc; Tx = Ip[WS(rs, 6)]; Ty = Im[WS(rs, 1)]; Tz = Tx - Ty; T1X = Tx + Ty; } Te = Ta + Td; TD = Tz - TC; T1x = TC + Tz; T10 = Ta - Td; { E T2l, T2m, T1V, T1Y; T2l = T1T + T1U; T2m = T1W + T1X; T2n = KP707106781 * (T2l - T2m); T2L = KP707106781 * (T2l + T2m); T1V = T1T - T1U; T1Y = T1W - T1X; T1Z = KP707106781 * (T1V + T1Y); T2X = KP707106781 * (T1V - T1Y); } } { E Ti, T2b, TI, T29, Tl, T28, TL, T2c, TF, TM; { E Tg, Th, TG, TH; Tg = Rp[WS(rs, 1)]; Th = Rm[WS(rs, 6)]; Ti = Tg + Th; T2b = Tg - Th; TG = Ip[WS(rs, 1)]; TH = Im[WS(rs, 6)]; TI = TG - TH; T29 = TG + TH; } { E Tj, Tk, TJ, TK; Tj = Rp[WS(rs, 5)]; Tk = Rm[WS(rs, 2)]; Tl = Tj + Tk; T28 = Tj - Tk; TJ = Ip[WS(rs, 5)]; TK = Im[WS(rs, 2)]; TL = TJ - TK; T2c = TJ + TK; } Tm = Ti + Tl; T1z = TI + TL; TF = Ti - Tl; TM = TI - TL; TN = TF - TM; T19 = TF + TM; { E T2a, T2d, T2N, T2O; T2a = T28 + T29; T2d = T2b - T2c; T2e = FMA(KP923879532, T2a, KP382683432 * T2d); T2p = FNMS(KP382683432, T2a, KP923879532 * T2d); T2N = T2b + T2c; T2O = T29 - T28; T2P = FNMS(KP923879532, T2O, KP382683432 * T2N); T2Z = FMA(KP382683432, T2O, KP923879532 * T2N); } } { E Tp, T24, TR, T22, Ts, T21, TU, T25, TO, TV; { E Tn, To, TP, TQ; Tn = Rm[0]; To = Rp[WS(rs, 7)]; Tp = Tn + To; T24 = Tn - To; TP = Ip[WS(rs, 7)]; TQ = Im[0]; TR = TP - TQ; T22 = TP + TQ; } { E Tq, Tr, TS, TT; Tq = Rp[WS(rs, 3)]; Tr = Rm[WS(rs, 4)]; Ts = Tq + Tr; T21 = Tq - Tr; TS = Ip[WS(rs, 3)]; TT = Im[WS(rs, 4)]; TU = TS - TT; T25 = TS + TT; } Tt = Tp + Ts; T1A = TR + TU; TO = Tp - Ts; TV = TR - TU; TW = TO + TV; T1a = TV - TO; { E T23, T26, T2Q, T2R; T23 = T21 - T22; T26 = T24 - T25; T27 = FNMS(KP382683432, T26, KP923879532 * T23); T2q = FMA(KP382683432, T23, KP923879532 * T26); T2Q = T24 + T25; T2R = T21 + T22; T2S = FNMS(KP923879532, T2R, KP382683432 * T2Q); T30 = FMA(KP382683432, T2R, KP923879532 * T2Q); } } { E Tf, Tu, T1u, T1y, T1B, T1C, T1t, T1v; Tf = T7 + Te; Tu = Tm + Tt; T1u = Tf - Tu; T1y = T1w + T1x; T1B = T1z + T1A; T1C = T1y - T1B; Rp[0] = Tf + Tu; Rm[0] = T1y + T1B; T1t = W[14]; T1v = W[15]; Rp[WS(rs, 4)] = FNMS(T1v, T1C, T1t * T1u); Rm[WS(rs, 4)] = FMA(T1v, T1u, T1t * T1C); } { E T2U, T34, T32, T36; { E T2M, T2T, T2Y, T31; T2M = T2K - T2L; T2T = T2P + T2S; T2U = T2M - T2T; T34 = T2M + T2T; T2Y = T2W + T2X; T31 = T2Z - T30; T32 = T2Y - T31; T36 = T2Y + T31; } { E T2J, T2V, T33, T35; T2J = W[20]; T2V = W[21]; Ip[WS(rs, 5)] = FNMS(T2V, T32, T2J * T2U); Im[WS(rs, 5)] = FMA(T2V, T2U, T2J * T32); T33 = W[4]; T35 = W[5]; Ip[WS(rs, 1)] = FNMS(T35, T36, T33 * T34); Im[WS(rs, 1)] = FMA(T35, T34, T33 * T36); } } { E T3a, T3g, T3e, T3i; { E T38, T39, T3c, T3d; T38 = T2K + T2L; T39 = T2Z + T30; T3a = T38 - T39; T3g = T38 + T39; T3c = T2W - T2X; T3d = T2P - T2S; T3e = T3c + T3d; T3i = T3c - T3d; } { E T37, T3b, T3f, T3h; T37 = W[12]; T3b = W[13]; Ip[WS(rs, 3)] = FNMS(T3b, T3e, T37 * T3a); Im[WS(rs, 3)] = FMA(T37, T3e, T3b * T3a); T3f = W[28]; T3h = W[29]; Ip[WS(rs, 7)] = FNMS(T3h, T3i, T3f * T3g); Im[WS(rs, 7)] = FMA(T3f, T3i, T3h * T3g); } } { E TY, T1e, T1c, T1g; { E TE, TX, T18, T1b; TE = Tw + TD; TX = KP707106781 * (TN + TW); TY = TE - TX; T1e = TE + TX; T18 = T10 + T17; T1b = KP707106781 * (T19 + T1a); T1c = T18 - T1b; T1g = T18 + T1b; } { E Tv, TZ, T1d, T1f; Tv = W[18]; TZ = W[19]; Rp[WS(rs, 5)] = FNMS(TZ, T1c, Tv * TY); Rm[WS(rs, 5)] = FMA(TZ, TY, Tv * T1c); T1d = W[2]; T1f = W[3]; Rp[WS(rs, 1)] = FNMS(T1f, T1g, T1d * T1e); Rm[WS(rs, 1)] = FMA(T1f, T1e, T1d * T1g); } } { E T1k, T1q, T1o, T1s; { E T1i, T1j, T1m, T1n; T1i = Tw - TD; T1j = KP707106781 * (T1a - T19); T1k = T1i - T1j; T1q = T1i + T1j; T1m = T17 - T10; T1n = KP707106781 * (TN - TW); T1o = T1m - T1n; T1s = T1m + T1n; } { E T1h, T1l, T1p, T1r; T1h = W[26]; T1l = W[27]; Rp[WS(rs, 7)] = FNMS(T1l, T1o, T1h * T1k); Rm[WS(rs, 7)] = FMA(T1h, T1o, T1l * T1k); T1p = W[10]; T1r = W[11]; Rp[WS(rs, 3)] = FNMS(T1r, T1s, T1p * T1q); Rm[WS(rs, 3)] = FMA(T1p, T1s, T1r * T1q); } } { E T2g, T2u, T2s, T2w; { E T20, T2f, T2o, T2r; T20 = T1S - T1Z; T2f = T27 - T2e; T2g = T20 - T2f; T2u = T20 + T2f; T2o = T2k - T2n; T2r = T2p - T2q; T2s = T2o - T2r; T2w = T2o + T2r; } { E T1P, T2h, T2t, T2v; T1P = W[24]; T2h = W[25]; Ip[WS(rs, 6)] = FNMS(T2h, T2s, T1P * T2g); Im[WS(rs, 6)] = FMA(T2h, T2g, T1P * T2s); T2t = W[8]; T2v = W[9]; Ip[WS(rs, 2)] = FNMS(T2v, T2w, T2t * T2u); Im[WS(rs, 2)] = FMA(T2v, T2u, T2t * T2w); } } { E T2A, T2G, T2E, T2I; { E T2y, T2z, T2C, T2D; T2y = T1S + T1Z; T2z = T2p + T2q; T2A = T2y - T2z; T2G = T2y + T2z; T2C = T2k + T2n; T2D = T2e + T27; T2E = T2C - T2D; T2I = T2C + T2D; } { E T2x, T2B, T2F, T2H; T2x = W[16]; T2B = W[17]; Ip[WS(rs, 4)] = FNMS(T2B, T2E, T2x * T2A); Im[WS(rs, 4)] = FMA(T2x, T2E, T2B * T2A); T2F = W[0]; T2H = W[1]; Ip[0] = FNMS(T2H, T2I, T2F * T2G); Im[0] = FMA(T2F, T2I, T2H * T2G); } } { E T1G, T1M, T1K, T1O; { E T1E, T1F, T1I, T1J; T1E = T7 - Te; T1F = T1A - T1z; T1G = T1E - T1F; T1M = T1E + T1F; T1I = T1w - T1x; T1J = Tm - Tt; T1K = T1I - T1J; T1O = T1J + T1I; } { E T1D, T1H, T1L, T1N; T1D = W[22]; T1H = W[23]; Rp[WS(rs, 6)] = FNMS(T1H, T1K, T1D * T1G); Rm[WS(rs, 6)] = FMA(T1D, T1K, T1H * T1G); T1L = W[6]; T1N = W[7]; Rp[WS(rs, 2)] = FNMS(T1N, T1O, T1L * T1M); Rm[WS(rs, 2)] = FMA(T1L, T1O, T1N * T1M); } } } } } static const tw_instr twinstr[] = { {TW_FULL, 1, 16}, {TW_NEXT, 1, 0} }; static const hc2c_desc desc = { 16, "hc2cb_16", twinstr, &GENUS, {136, 46, 38, 0} }; void X(codelet_hc2cb_16) (planner *p) { X(khc2c_register) (p, hc2cb_16, &desc, HC2C_VIA_RDFT); } #endif /* HAVE_FMA */ fftw-3.3.3/rdft/scalar/r2cb/hc2cb_32.c0000644000175000017500000013323312054411032014071 00000000000000/* * Copyright (c) 2003, 2007-11 Matteo Frigo * Copyright (c) 2003, 2007-11 Massachusetts Institute of Technology * * This program is free software; you can redistribute it and/or modify * it under the terms of the GNU General Public License as published by * the Free Software Foundation; either version 2 of the License, or * (at your option) any later version. * * This program is distributed in the hope that it will be useful, * but WITHOUT ANY WARRANTY; without even the implied warranty of * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the * GNU General Public License for more details. * * You should have received a copy of the GNU General Public License * along with this program; if not, write to the Free Software * Foundation, Inc., 51 Franklin Street, Fifth Floor, Boston, MA 02110-1301 USA * */ /* This file was automatically generated --- DO NOT EDIT */ /* Generated on Sun Nov 25 07:41:55 EST 2012 */ #include "codelet-rdft.h" #ifdef HAVE_FMA /* Generated by: ../../../genfft/gen_hc2c.native -fma -reorder-insns -schedule-for-pipeline -compact -variables 4 -pipeline-latency 4 -sign 1 -n 32 -dif -name hc2cb_32 -include hc2cb.h */ /* * This function contains 434 FP additions, 260 FP multiplications, * (or, 236 additions, 62 multiplications, 198 fused multiply/add), * 137 stack variables, 7 constants, and 128 memory accesses */ #include "hc2cb.h" static void hc2cb_32(R *Rp, R *Ip, R *Rm, R *Im, const R *W, stride rs, INT mb, INT me, INT ms) { DK(KP980785280, +0.980785280403230449126182236134239036973933731); DK(KP198912367, +0.198912367379658006911597622644676228597850501); DK(KP831469612, +0.831469612302545237078788377617905756738560812); DK(KP668178637, +0.668178637919298919997757686523080761552472251); DK(KP923879532, +0.923879532511286756128183189396788286822416626); DK(KP414213562, +0.414213562373095048801688724209698078569671875); DK(KP707106781, +0.707106781186547524400844362104849039284835938); { INT m; for (m = mb, W = W + ((mb - 1) * 62); m < me; m = m + 1, Rp = Rp + ms, Ip = Ip + ms, Rm = Rm - ms, Im = Im - ms, W = W + 62, MAKE_VOLATILE_STRIDE(128, rs)) { E T5o, T5r, T5q, T5n, T5s, T5p; { E T5K, Tf, T8k, T7k, T8x, T7N, T3i, T1i, T3v, T2L, T5f, T4v, T6T, T6m, T52; E T42, TZ, T6X, T3p, T1X, T8B, T8p, T3o, T26, T58, T4n, T7T, T7z, T59, T4k; E T6p, T6a, TK, T6W, T8s, T8A, T2o, T3m, T3l, T2x, T55, T4g, T7S, T7G, T56; E T4d, T6o, T61, T5Q, T5N, T6f, Tu, T8y, T7r, T8l, T7Q, T3w, T1F, T45, T48; E T3j, T2O, T53, T4y, T62, T69; { E T6l, T6i, T40, T41; { E T12, T3, T6g, T2G, T2D, T6, T6h, T15, Td, T6k, T1g, T2J, Ta, T17, T1a; E T6j; { E T4, T5, T13, T14; { E T1, T2, T2E, T2F; T1 = Rp[0]; T2 = Rm[WS(rs, 15)]; T2E = Ip[0]; T2F = Im[WS(rs, 15)]; T4 = Rp[WS(rs, 8)]; T12 = T1 - T2; T3 = T1 + T2; T6g = T2E - T2F; T2G = T2E + T2F; T5 = Rm[WS(rs, 7)]; } T13 = Ip[WS(rs, 8)]; T14 = Im[WS(rs, 7)]; { E Tb, Tc, T1d, T1e; Tb = Rm[WS(rs, 3)]; T2D = T4 - T5; T6 = T4 + T5; T6h = T13 - T14; T15 = T13 + T14; Tc = Rp[WS(rs, 12)]; T1d = Ip[WS(rs, 12)]; T1e = Im[WS(rs, 3)]; { E T8, T1c, T1f, T9, T18, T19; T8 = Rp[WS(rs, 4)]; Td = Tb + Tc; T1c = Tb - Tc; T6k = T1d - T1e; T1f = T1d + T1e; T9 = Rm[WS(rs, 11)]; T18 = Ip[WS(rs, 4)]; T19 = Im[WS(rs, 11)]; T1g = T1c - T1f; T2J = T1c + T1f; Ta = T8 + T9; T17 = T8 - T9; T1a = T18 + T19; T6j = T18 - T19; } } } { E T2I, T7M, T7L, T16, T1h, T4u, T4t, T2H, T2K; { E T7i, T7, T1b, Te, T7j; T7i = T3 - T6; T7 = T3 + T6; T2I = T17 + T1a; T1b = T17 - T1a; Te = Ta + Td; T7M = Ta - Td; T7j = T6k - T6j; T6l = T6j + T6k; T6i = T6g + T6h; T7L = T6g - T6h; T5K = T7 - Te; Tf = T7 + Te; T8k = T7i + T7j; T7k = T7i - T7j; T40 = T12 + T15; T16 = T12 - T15; T1h = T1b + T1g; T4u = T1b - T1g; } T4t = T2G - T2D; T2H = T2D + T2G; T8x = T7M + T7L; T7N = T7L - T7M; T3i = FMA(KP707106781, T1h, T16); T1i = FNMS(KP707106781, T1h, T16); T2K = T2I - T2J; T41 = T2I + T2J; T3v = FMA(KP707106781, T2K, T2H); T2L = FNMS(KP707106781, T2K, T2H); T5f = FNMS(KP707106781, T4u, T4t); T4v = FMA(KP707106781, T4u, T4t); } } { E T1Y, T1H, TR, T7w, T1K, T21, T65, T7t, TU, T66, T23, T1Q, T1R, TX, T67; E T1U, TY, T7u; { E TL, TM, TO, TP, T63, T64; TL = Rm[0]; T6T = T6i + T6l; T6m = T6i - T6l; T52 = FMA(KP707106781, T41, T40); T42 = FNMS(KP707106781, T41, T40); TM = Rp[WS(rs, 15)]; TO = Rp[WS(rs, 7)]; TP = Rm[WS(rs, 8)]; { E T1I, TN, TQ, T1J, T1Z, T20; T1I = Ip[WS(rs, 15)]; T1Y = TL - TM; TN = TL + TM; T1H = TO - TP; TQ = TO + TP; T1J = Im[0]; T1Z = Ip[WS(rs, 7)]; T20 = Im[WS(rs, 8)]; TR = TN + TQ; T7w = TN - TQ; T1K = T1I + T1J; T63 = T1I - T1J; T64 = T1Z - T20; T21 = T1Z + T20; } { E TV, T1M, T1P, TW, T1S, T1T; { E TS, TT, T1N, T1O; TS = Rp[WS(rs, 3)]; T65 = T63 + T64; T7t = T63 - T64; TT = Rm[WS(rs, 12)]; T1N = Ip[WS(rs, 3)]; T1O = Im[WS(rs, 12)]; TV = Rm[WS(rs, 4)]; T1M = TS - TT; TU = TS + TT; T66 = T1N - T1O; T1P = T1N + T1O; TW = Rp[WS(rs, 11)]; T1S = Ip[WS(rs, 11)]; T1T = Im[WS(rs, 4)]; } T23 = T1M - T1P; T1Q = T1M + T1P; T1R = TV - TW; TX = TV + TW; T67 = T1S - T1T; T1U = T1S + T1T; } } TY = TU + TX; T7u = TU - TX; { E T7x, T68, T1V, T24; T7x = T67 - T66; T68 = T66 + T67; T1V = T1R + T1U; T24 = T1R - T1U; { E T4l, T1L, T1W, T4j, T7v, T8n, T8o, T7y; T62 = TR - TY; TZ = TR + TY; T6X = T65 + T68; T69 = T65 - T68; T4l = T1H + T1K; T1L = T1H - T1K; T1W = T1Q - T1V; T4j = T1Q + T1V; T7v = T7t - T7u; T8n = T7u + T7t; T8o = T7w + T7x; T7y = T7w - T7x; { E T4i, T22, T25, T4m; T4i = T1Y + T21; T22 = T1Y - T21; T3p = FMA(KP707106781, T1W, T1L); T1X = FNMS(KP707106781, T1W, T1L); T8B = FMA(KP414213562, T8n, T8o); T8p = FNMS(KP414213562, T8o, T8n); T25 = T23 + T24; T4m = T23 - T24; T3o = FMA(KP707106781, T25, T22); T26 = FNMS(KP707106781, T25, T22); T58 = FMA(KP707106781, T4m, T4l); T4n = FNMS(KP707106781, T4m, T4l); T7T = FNMS(KP414213562, T7v, T7y); T7z = FMA(KP414213562, T7y, T7v); T59 = FMA(KP707106781, T4j, T4i); T4k = FNMS(KP707106781, T4j, T4i); } } } } } { E T5T, T60, T4c, T4b; { E T2p, T28, T2b, T7D, TC, T2s, T7A, T5W, TF, T2j, T5X, T2i, TI, T2k, T2u; E T2h; { E Tz, Ty, TA, Tw, Tx; Tw = Rp[WS(rs, 1)]; Tx = Rm[WS(rs, 14)]; Tz = Rp[WS(rs, 9)]; T6p = T69 - T62; T6a = T62 + T69; Ty = Tw + Tx; T2p = Tw - Tx; TA = Rm[WS(rs, 6)]; { E T5U, T5V, T2d, T2g; { E T2q, T2r, T29, T2a, TB; T29 = Ip[WS(rs, 1)]; T2a = Im[WS(rs, 14)]; TB = Tz + TA; T28 = Tz - TA; T2q = Ip[WS(rs, 9)]; T5U = T29 - T2a; T2b = T29 + T2a; T2r = Im[WS(rs, 6)]; T7D = Ty - TB; TC = Ty + TB; T2s = T2q + T2r; T5V = T2q - T2r; } { E T2e, T2f, TD, TE, TG, TH; TD = Rp[WS(rs, 5)]; TE = Rm[WS(rs, 10)]; T7A = T5U - T5V; T5W = T5U + T5V; T2e = Ip[WS(rs, 5)]; T2d = TD - TE; TF = TD + TE; T2f = Im[WS(rs, 10)]; TG = Rm[WS(rs, 2)]; TH = Rp[WS(rs, 13)]; T2j = Ip[WS(rs, 13)]; T5X = T2e - T2f; T2g = T2e + T2f; T2i = TG - TH; TI = TG + TH; T2k = Im[WS(rs, 2)]; } T2u = T2d - T2g; T2h = T2d + T2g; } } { E TJ, T7B, T2l, T5Y; TJ = TF + TI; T7B = TF - TI; T2l = T2j + T2k; T5Y = T2j - T2k; { E T4e, T2c, T2v, T8q, T7C, T7F, T8r, T2n, T7E, T2m, T5Z, T4f, T2t, T2w; T4e = T2b - T28; T2c = T28 + T2b; TK = TC + TJ; T5T = TC - TJ; T7E = T5Y - T5X; T5Z = T5X + T5Y; T2m = T2i + T2l; T2v = T2i - T2l; T60 = T5W - T5Z; T6W = T5W + T5Z; T8q = T7B + T7A; T7C = T7A - T7B; T7F = T7D - T7E; T8r = T7D + T7E; T2n = T2h - T2m; T4c = T2h + T2m; T4b = T2p + T2s; T2t = T2p - T2s; T2w = T2u + T2v; T4f = T2v - T2u; T8s = FMA(KP414213562, T8r, T8q); T8A = FNMS(KP414213562, T8q, T8r); T2o = FNMS(KP707106781, T2n, T2c); T3m = FMA(KP707106781, T2n, T2c); T3l = FMA(KP707106781, T2w, T2t); T2x = FNMS(KP707106781, T2w, T2t); T55 = FMA(KP707106781, T4f, T4e); T4g = FNMS(KP707106781, T4f, T4e); T7S = FMA(KP414213562, T7C, T7F); T7G = FNMS(KP414213562, T7F, T7C); } } } { E T43, T1y, T7o, Tm, T7p, T44, T1D, Tq, T1o, Tp, T5L, T1m, Tr, T1p, T1q; { E Tj, T1z, Ti, T5O, T1x, Tk, T1A, T1B; { E Tg, Th, T1v, T1w; Tg = Rp[WS(rs, 2)]; T56 = FMA(KP707106781, T4c, T4b); T4d = FNMS(KP707106781, T4c, T4b); T6o = T5T + T60; T61 = T5T - T60; Th = Rm[WS(rs, 13)]; T1v = Ip[WS(rs, 2)]; T1w = Im[WS(rs, 13)]; Tj = Rp[WS(rs, 10)]; T1z = Tg - Th; Ti = Tg + Th; T5O = T1v - T1w; T1x = T1v + T1w; Tk = Rm[WS(rs, 5)]; T1A = Ip[WS(rs, 10)]; T1B = Im[WS(rs, 5)]; } { E Tn, To, T1k, T1l; Tn = Rm[WS(rs, 1)]; { E T1u, Tl, T5P, T1C; T1u = Tj - Tk; Tl = Tj + Tk; T5P = T1A - T1B; T1C = T1A + T1B; T43 = T1x - T1u; T1y = T1u + T1x; T7o = Ti - Tl; Tm = Ti + Tl; T5Q = T5O + T5P; T7p = T5O - T5P; T44 = T1z + T1C; T1D = T1z - T1C; To = Rp[WS(rs, 14)]; } T1k = Ip[WS(rs, 14)]; T1l = Im[WS(rs, 1)]; Tq = Rp[WS(rs, 6)]; T1o = Tn - To; Tp = Tn + To; T5L = T1k - T1l; T1m = T1k + T1l; Tr = Rm[WS(rs, 9)]; T1p = Ip[WS(rs, 6)]; T1q = Im[WS(rs, 9)]; } } { E T46, T47, T7P, T7O, T2N, T1t, T1E, T2M, T4w, T4x; { E T1n, Tt, T1s, T7n, T7q, T7m, T7l; { E T1j, Ts, T5M, T1r; T1j = Tq - Tr; Ts = Tq + Tr; T5M = T1p - T1q; T1r = T1p + T1q; T46 = T1j + T1m; T1n = T1j - T1m; T7m = Tp - Ts; Tt = Tp + Ts; T5N = T5L + T5M; T7l = T5L - T5M; T47 = T1o + T1r; T1s = T1o - T1r; } T7P = T7m + T7l; T7n = T7l - T7m; T7q = T7o + T7p; T7O = T7o - T7p; T6f = Tm - Tt; Tu = Tm + Tt; T8y = T7q + T7n; T7r = T7n - T7q; T2N = FMA(KP414213562, T1n, T1s); T1t = FNMS(KP414213562, T1s, T1n); T1E = FMA(KP414213562, T1D, T1y); T2M = FNMS(KP414213562, T1y, T1D); } T8l = T7O + T7P; T7Q = T7O - T7P; T3w = T1E + T1t; T1F = T1t - T1E; T45 = FNMS(KP414213562, T44, T43); T4w = FMA(KP414213562, T43, T44); T4x = FMA(KP414213562, T46, T47); T48 = FNMS(KP414213562, T47, T46); T3j = T2M + T2N; T2O = T2M - T2N; T53 = T4w + T4x; T4y = T4w - T4x; } } } { E T72, T5g, T49, T78, T77, T73, T7s, T7U, T7R, T7H, T3f, T3e, T3d; { E T5R, T8m, T8C, T8z, T8t, T8e, T86, T88, T8h, T8f, T8i, T8c, T8g; { E T6P, T6Q, T6Z, T6S, T6R; { E Tv, T10, T6V, T6Y, T6U; T72 = Tf - Tu; Tv = Tf + Tu; T6U = T5Q + T5N; T5R = T5N - T5Q; T5g = T48 - T45; T49 = T45 + T48; T10 = TK + TZ; T78 = TK - TZ; T77 = T6T - T6U; T6V = T6T + T6U; T6Y = T6W + T6X; T73 = T6X - T6W; T6P = W[30]; Rp[0] = Tv + T10; T6Q = Tv - T10; Rm[0] = T6V + T6Y; T6Z = T6V - T6Y; T6S = W[31]; T6R = T6P * T6Q; } { E T8O, T8W, T8Q, T8Z, T8X, T90, T8U, T8Y; { E T8R, T8S, T8M, T8N, T70; T8M = FMA(KP707106781, T8l, T8k); T8m = FNMS(KP707106781, T8l, T8k); T8C = T8A - T8B; T8N = T8A + T8B; T70 = T6S * T6Q; Rp[WS(rs, 8)] = FNMS(T6S, T6Z, T6R); T8R = FMA(KP707106781, T8y, T8x); T8z = FNMS(KP707106781, T8y, T8x); T8O = FNMS(KP923879532, T8N, T8M); T8W = FMA(KP923879532, T8N, T8M); Rm[WS(rs, 8)] = FMA(T6P, T6Z, T70); T8S = T8s + T8p; T8t = T8p - T8s; { E T8L, T8T, T8P, T8V; T8L = W[34]; T8Q = W[35]; T8V = W[2]; T8Z = FMA(KP923879532, T8S, T8R); T8T = FNMS(KP923879532, T8S, T8R); T8P = T8L * T8O; T8X = T8V * T8W; T90 = T8V * T8Z; T8U = T8L * T8T; Rp[WS(rs, 9)] = FNMS(T8Q, T8T, T8P); T8Y = W[3]; } } { E T89, T8a, T84, T85; T84 = FNMS(KP707106781, T7r, T7k); T7s = FMA(KP707106781, T7r, T7k); Rm[WS(rs, 9)] = FMA(T8Q, T8O, T8U); T85 = T7S + T7T; T7U = T7S - T7T; Rm[WS(rs, 1)] = FMA(T8Y, T8W, T90); Rp[WS(rs, 1)] = FNMS(T8Y, T8Z, T8X); T7R = FMA(KP707106781, T7Q, T7N); T89 = FNMS(KP707106781, T7Q, T7N); T8e = FMA(KP923879532, T85, T84); T86 = FNMS(KP923879532, T85, T84); T8a = T7G + T7z; T7H = T7z - T7G; { E T83, T8b, T87, T8d; T83 = W[26]; T88 = W[27]; T8d = W[58]; T8h = FMA(KP923879532, T8a, T89); T8b = FNMS(KP923879532, T8a, T89); T87 = T83 * T86; T8f = T8d * T8e; T8i = T8d * T8h; T8c = T83 * T8b; Rp[WS(rs, 7)] = FNMS(T88, T8b, T87); T8g = W[59]; } } } } { E T5S, T6q, T6n, T6K, T6C, T6b, T6E, T6N, T6L, T6O, T6I, T6M; { E T6F, T6G, T6A, T6B; T6A = T5K - T5R; T5S = T5K + T5R; Rm[WS(rs, 7)] = FMA(T88, T86, T8c); T6B = T6p - T6o; T6q = T6o + T6p; Rm[WS(rs, 15)] = FMA(T8g, T8e, T8i); Rp[WS(rs, 15)] = FNMS(T8g, T8h, T8f); T6n = T6f + T6m; T6F = T6m - T6f; T6K = FMA(KP707106781, T6B, T6A); T6C = FNMS(KP707106781, T6B, T6A); T6G = T61 - T6a; T6b = T61 + T6a; { E T6z, T6H, T6D, T6J; T6z = W[54]; T6E = W[55]; T6J = W[22]; T6N = FMA(KP707106781, T6G, T6F); T6H = FNMS(KP707106781, T6G, T6F); T6D = T6z * T6C; T6L = T6J * T6K; T6O = T6J * T6N; T6I = T6z * T6H; Rp[WS(rs, 14)] = FNMS(T6E, T6H, T6D); T6M = W[23]; } } { E T8G, T8F, T8J, T8H, T8I, T8u; Rm[WS(rs, 14)] = FMA(T6E, T6C, T6I); Rm[WS(rs, 6)] = FMA(T6M, T6K, T6O); Rp[WS(rs, 6)] = FNMS(T6M, T6N, T6L); T8G = FMA(KP923879532, T8t, T8m); T8u = FNMS(KP923879532, T8t, T8m); { E T8j, T8w, T8D, T8v, T8E; T8j = W[50]; T8w = W[51]; T8F = W[18]; T8J = FMA(KP923879532, T8C, T8z); T8D = FNMS(KP923879532, T8C, T8z); T8v = T8j * T8u; T8E = T8w * T8u; T8H = T8F * T8G; T8I = W[19]; Rp[WS(rs, 13)] = FNMS(T8w, T8D, T8v); Rm[WS(rs, 13)] = FMA(T8j, T8D, T8E); } { E T6c, T6u, T6x, T6r, T8K, T5J, T6e; Rp[WS(rs, 5)] = FNMS(T8I, T8J, T8H); T8K = T8I * T8G; Rm[WS(rs, 5)] = FMA(T8F, T8J, T8K); T6c = FNMS(KP707106781, T6b, T5S); T6u = FMA(KP707106781, T6b, T5S); T6x = FMA(KP707106781, T6q, T6n); T6r = FNMS(KP707106781, T6q, T6n); T5J = W[38]; T6e = W[39]; { E T6t, T6w, T6d, T6s, T6v, T6y; T6t = W[6]; T6w = W[7]; T6d = T5J * T6c; T6s = T6e * T6c; T6v = T6t * T6u; T6y = T6w * T6u; Rp[WS(rs, 10)] = FNMS(T6e, T6r, T6d); Rm[WS(rs, 10)] = FMA(T5J, T6r, T6s); Rp[WS(rs, 2)] = FNMS(T6w, T6x, T6v); Rm[WS(rs, 2)] = FMA(T6t, T6x, T6y); } } } } } { E T7c, T7f, T7e, T7g, T7d; { E T71, T74, T79, T76, T75, T7b, T7a; T71 = W[46]; T7c = T72 + T73; T74 = T72 - T73; T7f = T78 + T77; T79 = T77 - T78; T76 = W[47]; T75 = T71 * T74; T7b = W[14]; T7a = T71 * T79; T7e = W[15]; Rp[WS(rs, 12)] = FNMS(T76, T79, T75); T7g = T7b * T7f; T7d = T7b * T7c; Rm[WS(rs, 12)] = FMA(T76, T74, T7a); } { E T81, T7X, T80, T7Z, T82; Rm[WS(rs, 4)] = FMA(T7e, T7c, T7g); Rp[WS(rs, 4)] = FNMS(T7e, T7f, T7d); { E T7h, T7Y, T7I, T7V, T7K, T7J, T7W; T7h = W[42]; T7Y = FMA(KP923879532, T7H, T7s); T7I = FNMS(KP923879532, T7H, T7s); T81 = FMA(KP923879532, T7U, T7R); T7V = FNMS(KP923879532, T7U, T7R); T7K = W[43]; T7J = T7h * T7I; T7X = W[10]; T80 = W[11]; T7W = T7K * T7I; Rp[WS(rs, 11)] = FNMS(T7K, T7V, T7J); T7Z = T7X * T7Y; T82 = T80 * T7Y; Rm[WS(rs, 11)] = FMA(T7h, T7V, T7W); } { E T2P, T37, T1G, T32, T2R, T2Q, T38, T2z, T27, T2y; T2P = FMA(KP923879532, T2O, T2L); T37 = FNMS(KP923879532, T2O, T2L); Rp[WS(rs, 3)] = FNMS(T80, T81, T7Z); Rm[WS(rs, 3)] = FMA(T7X, T81, T82); T1G = FMA(KP923879532, T1F, T1i); T32 = FNMS(KP923879532, T1F, T1i); T2R = FNMS(KP668178637, T1X, T26); T27 = FMA(KP668178637, T26, T1X); T2y = FNMS(KP668178637, T2x, T2o); T2Q = FMA(KP668178637, T2o, T2x); T38 = T2y + T27; T2z = T27 - T2y; { E T2C, T2A, T3c, T34, T2U, T39, T36, T31; { E T11, T2W, T2S, T33; T11 = W[40]; T2C = W[41]; T2A = FNMS(KP831469612, T2z, T1G); T2W = FMA(KP831469612, T2z, T1G); T2S = T2Q - T2R; T33 = T2Q + T2R; { E T2V, T2B, T2T, T2Z, T2X, T2Y, T30; T2V = W[8]; T2B = T11 * T2A; T3c = FMA(KP831469612, T33, T32); T34 = FNMS(KP831469612, T33, T32); T2T = FNMS(KP831469612, T2S, T2P); T2Z = FMA(KP831469612, T2S, T2P); T2X = T2V * T2W; T2Y = W[9]; T30 = T2V * T2Z; Ip[WS(rs, 10)] = FNMS(T2C, T2T, T2B); T2U = T11 * T2T; Ip[WS(rs, 2)] = FNMS(T2Y, T2Z, T2X); Im[WS(rs, 2)] = FMA(T2Y, T2W, T30); } } T39 = FNMS(KP831469612, T38, T37); T3f = FMA(KP831469612, T38, T37); Im[WS(rs, 10)] = FMA(T2C, T2A, T2U); T36 = W[25]; T31 = W[24]; { E T3b, T3g, T3a, T35; T3e = W[57]; T3a = T36 * T34; T35 = T31 * T34; T3b = W[56]; T3g = T3e * T3c; Im[WS(rs, 6)] = FMA(T31, T39, T3a); Ip[WS(rs, 6)] = FNMS(T36, T39, T35); T3d = T3b * T3c; Im[WS(rs, 14)] = FMA(T3b, T3f, T3g); } } } } } { E T4G, T4J, T4I, T4F, T4K; { E T4z, T4R, T4a, T4M, T4h, T4o, T4C, T4N, T4A, T4B; T4z = FMA(KP923879532, T4y, T4v); T4R = FNMS(KP923879532, T4y, T4v); T4a = FNMS(KP923879532, T49, T42); T4M = FMA(KP923879532, T49, T42); Ip[WS(rs, 14)] = FNMS(T3e, T3f, T3d); T4h = FNMS(KP668178637, T4g, T4d); T4A = FMA(KP668178637, T4d, T4g); T4B = FMA(KP668178637, T4k, T4n); T4o = FNMS(KP668178637, T4n, T4k); T4C = T4A - T4B; T4N = T4A + T4B; { E T4W, T4Z, T4q, T4X, T50, T4Y; { E T4L, T4Q, T4O, T4p, T4S, T4P, T4U, T4V, T4T; T4L = W[20]; T4Q = W[21]; T4W = FMA(KP831469612, T4N, T4M); T4O = FNMS(KP831469612, T4N, T4M); T4p = T4h + T4o; T4S = T4h - T4o; T4P = T4L * T4O; T4V = W[52]; T4Z = FNMS(KP831469612, T4S, T4R); T4T = FMA(KP831469612, T4S, T4R); T4q = FNMS(KP831469612, T4p, T4a); T4G = FMA(KP831469612, T4p, T4a); Ip[WS(rs, 5)] = FNMS(T4Q, T4T, T4P); T4U = T4L * T4T; T4X = T4V * T4W; T50 = T4V * T4Z; T4Y = W[53]; Im[WS(rs, 5)] = FMA(T4Q, T4O, T4U); } { E T4D, T4s, T3Z, T4E, T4r; T4J = FMA(KP831469612, T4C, T4z); T4D = FNMS(KP831469612, T4C, T4z); T4s = W[37]; Im[WS(rs, 13)] = FMA(T4Y, T4W, T50); Ip[WS(rs, 13)] = FNMS(T4Y, T4Z, T4X); T3Z = W[36]; T4E = T4s * T4q; T4I = W[5]; T4r = T3Z * T4q; Im[WS(rs, 9)] = FMA(T3Z, T4D, T4E); T4F = W[4]; T4K = T4I * T4G; Ip[WS(rs, 9)] = FNMS(T4s, T4D, T4r); } } } { E T3E, T3H, T3G, T3D, T3I; { E T3x, T3P, T3k, T3K, T3n, T3q, T3A, T3L, T4H, T3y, T3z; T3x = FMA(KP923879532, T3w, T3v); T3P = FNMS(KP923879532, T3w, T3v); T4H = T4F * T4G; Im[WS(rs, 1)] = FMA(T4F, T4J, T4K); T3k = FMA(KP923879532, T3j, T3i); T3K = FNMS(KP923879532, T3j, T3i); T3y = FMA(KP198912367, T3l, T3m); T3n = FNMS(KP198912367, T3m, T3l); Ip[WS(rs, 1)] = FNMS(T4I, T4J, T4H); T3z = FNMS(KP198912367, T3o, T3p); T3q = FMA(KP198912367, T3p, T3o); T3A = T3y + T3z; T3L = T3z - T3y; { E T3U, T3X, T3s, T3V, T3Y, T3W; { E T3J, T3O, T3M, T3r, T3Q, T3N, T3S, T3T, T3R; T3J = W[48]; T3O = W[49]; T3U = FMA(KP980785280, T3L, T3K); T3M = FNMS(KP980785280, T3L, T3K); T3r = T3n + T3q; T3Q = T3n - T3q; T3N = T3J * T3M; T3T = W[16]; T3X = FMA(KP980785280, T3Q, T3P); T3R = FNMS(KP980785280, T3Q, T3P); T3s = FNMS(KP980785280, T3r, T3k); T3E = FMA(KP980785280, T3r, T3k); Ip[WS(rs, 12)] = FNMS(T3O, T3R, T3N); T3S = T3J * T3R; T3V = T3T * T3U; T3Y = T3T * T3X; T3W = W[17]; Im[WS(rs, 12)] = FMA(T3O, T3M, T3S); } { E T3B, T3u, T3h, T3C, T3t; T3H = FMA(KP980785280, T3A, T3x); T3B = FNMS(KP980785280, T3A, T3x); T3u = W[33]; Im[WS(rs, 4)] = FMA(T3W, T3U, T3Y); Ip[WS(rs, 4)] = FNMS(T3W, T3X, T3V); T3h = W[32]; T3C = T3u * T3s; T3G = W[1]; T3t = T3h * T3s; Im[WS(rs, 8)] = FMA(T3h, T3B, T3C); T3D = W[0]; T3I = T3G * T3E; Ip[WS(rs, 8)] = FNMS(T3u, T3B, T3t); } } } { E T5h, T5z, T54, T5u, T57, T5a, T5k, T5v, T3F, T5i, T5j; T5h = FMA(KP923879532, T5g, T5f); T5z = FNMS(KP923879532, T5g, T5f); T3F = T3D * T3E; Im[0] = FMA(T3D, T3H, T3I); T54 = FNMS(KP923879532, T53, T52); T5u = FMA(KP923879532, T53, T52); T5i = FMA(KP198912367, T55, T56); T57 = FNMS(KP198912367, T56, T55); Ip[0] = FNMS(T3G, T3H, T3F); T5j = FMA(KP198912367, T58, T59); T5a = FNMS(KP198912367, T59, T58); T5k = T5i - T5j; T5v = T5i + T5j; { E T5E, T5H, T5c, T5F, T5I, T5G; { E T5t, T5y, T5w, T5b, T5A, T5x, T5C, T5D, T5B; T5t = W[28]; T5y = W[29]; T5E = FMA(KP980785280, T5v, T5u); T5w = FNMS(KP980785280, T5v, T5u); T5b = T57 + T5a; T5A = T5a - T57; T5x = T5t * T5w; T5D = W[60]; T5H = FNMS(KP980785280, T5A, T5z); T5B = FMA(KP980785280, T5A, T5z); T5c = FMA(KP980785280, T5b, T54); T5o = FNMS(KP980785280, T5b, T54); Ip[WS(rs, 7)] = FNMS(T5y, T5B, T5x); T5C = T5t * T5B; T5F = T5D * T5E; T5I = T5D * T5H; T5G = W[61]; Im[WS(rs, 7)] = FMA(T5y, T5w, T5C); } { E T5l, T5e, T51, T5m, T5d; T5r = FMA(KP980785280, T5k, T5h); T5l = FNMS(KP980785280, T5k, T5h); T5e = W[45]; Im[WS(rs, 15)] = FMA(T5G, T5E, T5I); Ip[WS(rs, 15)] = FNMS(T5G, T5H, T5F); T51 = W[44]; T5m = T5e * T5c; T5q = W[13]; T5d = T51 * T5c; Im[WS(rs, 11)] = FMA(T51, T5l, T5m); T5n = W[12]; T5s = T5q * T5o; Ip[WS(rs, 11)] = FNMS(T5e, T5l, T5d); } } } } } } } T5p = T5n * T5o; Im[WS(rs, 3)] = FMA(T5n, T5r, T5s); Ip[WS(rs, 3)] = FNMS(T5q, T5r, T5p); } } } static const tw_instr twinstr[] = { {TW_FULL, 1, 32}, {TW_NEXT, 1, 0} }; static const hc2c_desc desc = { 32, "hc2cb_32", twinstr, &GENUS, {236, 62, 198, 0} }; void X(codelet_hc2cb_32) (planner *p) { X(khc2c_register) (p, hc2cb_32, &desc, HC2C_VIA_RDFT); } #else /* HAVE_FMA */ /* Generated by: ../../../genfft/gen_hc2c.native -compact -variables 4 -pipeline-latency 4 -sign 1 -n 32 -dif -name hc2cb_32 -include hc2cb.h */ /* * This function contains 434 FP additions, 208 FP multiplications, * (or, 340 additions, 114 multiplications, 94 fused multiply/add), * 98 stack variables, 7 constants, and 128 memory accesses */ #include "hc2cb.h" static void hc2cb_32(R *Rp, R *Ip, R *Rm, R *Im, const R *W, stride rs, INT mb, INT me, INT ms) { DK(KP555570233, +0.555570233019602224742830813948532874374937191); DK(KP831469612, +0.831469612302545237078788377617905756738560812); DK(KP980785280, +0.980785280403230449126182236134239036973933731); DK(KP195090322, +0.195090322016128267848284868477022240927691618); DK(KP923879532, +0.923879532511286756128183189396788286822416626); DK(KP382683432, +0.382683432365089771728459984030398866761344562); DK(KP707106781, +0.707106781186547524400844362104849039284835938); { INT m; for (m = mb, W = W + ((mb - 1) * 62); m < me; m = m + 1, Rp = Rp + ms, Ip = Ip + ms, Rm = Rm - ms, Im = Im - ms, W = W + 62, MAKE_VOLATILE_STRIDE(128, rs)) { E T4o, T6y, T70, T5u, Tf, T12, T5x, T6z, T3m, T3Y, T29, T2y, T4v, T71, T2U; E T3M, Tu, T1U, T6D, T73, T6G, T74, T1h, T2z, T2X, T3o, T4D, T5A, T4K, T5z; E T30, T3n, TK, T1j, T6S, T7w, T6V, T7v, T1y, T2B, T3c, T3S, T4X, T61, T54; E T62, T3f, T3T, TZ, T1A, T6L, T7z, T6O, T7y, T1P, T2C, T35, T3P, T5g, T64; E T5n, T65, T38, T3Q; { E T3, T4m, T1X, T5t, T6, T5s, T20, T4n, Ta, T4p, T24, T4q, Td, T4s, T27; E T4t; { E T1, T2, T1V, T1W; T1 = Rp[0]; T2 = Rm[WS(rs, 15)]; T3 = T1 + T2; T4m = T1 - T2; T1V = Ip[0]; T1W = Im[WS(rs, 15)]; T1X = T1V - T1W; T5t = T1V + T1W; } { E T4, T5, T1Y, T1Z; T4 = Rp[WS(rs, 8)]; T5 = Rm[WS(rs, 7)]; T6 = T4 + T5; T5s = T4 - T5; T1Y = Ip[WS(rs, 8)]; T1Z = Im[WS(rs, 7)]; T20 = T1Y - T1Z; T4n = T1Y + T1Z; } { E T8, T9, T22, T23; T8 = Rp[WS(rs, 4)]; T9 = Rm[WS(rs, 11)]; Ta = T8 + T9; T4p = T8 - T9; T22 = Ip[WS(rs, 4)]; T23 = Im[WS(rs, 11)]; T24 = T22 - T23; T4q = T22 + T23; } { E Tb, Tc, T25, T26; Tb = Rm[WS(rs, 3)]; Tc = Rp[WS(rs, 12)]; Td = Tb + Tc; T4s = Tb - Tc; T25 = Ip[WS(rs, 12)]; T26 = Im[WS(rs, 3)]; T27 = T25 - T26; T4t = T25 + T26; } { E T7, Te, T21, T28; T4o = T4m - T4n; T6y = T4m + T4n; T70 = T5t - T5s; T5u = T5s + T5t; T7 = T3 + T6; Te = Ta + Td; Tf = T7 + Te; T12 = T7 - Te; { E T5v, T5w, T3k, T3l; T5v = T4p + T4q; T5w = T4s + T4t; T5x = KP707106781 * (T5v - T5w); T6z = KP707106781 * (T5v + T5w); T3k = T1X - T20; T3l = Ta - Td; T3m = T3k - T3l; T3Y = T3l + T3k; } T21 = T1X + T20; T28 = T24 + T27; T29 = T21 - T28; T2y = T21 + T28; { E T4r, T4u, T2S, T2T; T4r = T4p - T4q; T4u = T4s - T4t; T4v = KP707106781 * (T4r + T4u); T71 = KP707106781 * (T4r - T4u); T2S = T3 - T6; T2T = T27 - T24; T2U = T2S - T2T; T3M = T2S + T2T; } } } { E Ti, T4H, T1c, T4F, Tl, T4E, T1f, T4I, Tp, T4A, T15, T4y, Ts, T4x, T18; E T4B; { E Tg, Th, T1a, T1b; Tg = Rp[WS(rs, 2)]; Th = Rm[WS(rs, 13)]; Ti = Tg + Th; T4H = Tg - Th; T1a = Ip[WS(rs, 2)]; T1b = Im[WS(rs, 13)]; T1c = T1a - T1b; T4F = T1a + T1b; } { E Tj, Tk, T1d, T1e; Tj = Rp[WS(rs, 10)]; Tk = Rm[WS(rs, 5)]; Tl = Tj + Tk; T4E = Tj - Tk; T1d = Ip[WS(rs, 10)]; T1e = Im[WS(rs, 5)]; T1f = T1d - T1e; T4I = T1d + T1e; } { E Tn, To, T13, T14; Tn = Rm[WS(rs, 1)]; To = Rp[WS(rs, 14)]; Tp = Tn + To; T4A = Tn - To; T13 = Ip[WS(rs, 14)]; T14 = Im[WS(rs, 1)]; T15 = T13 - T14; T4y = T13 + T14; } { E Tq, Tr, T16, T17; Tq = Rp[WS(rs, 6)]; Tr = Rm[WS(rs, 9)]; Ts = Tq + Tr; T4x = Tq - Tr; T16 = Ip[WS(rs, 6)]; T17 = Im[WS(rs, 9)]; T18 = T16 - T17; T4B = T16 + T17; } { E Tm, Tt, T6B, T6C; Tm = Ti + Tl; Tt = Tp + Ts; Tu = Tm + Tt; T1U = Tm - Tt; T6B = T4H + T4I; T6C = T4F - T4E; T6D = FNMS(KP923879532, T6C, KP382683432 * T6B); T73 = FMA(KP382683432, T6C, KP923879532 * T6B); } { E T6E, T6F, T19, T1g; T6E = T4A + T4B; T6F = T4x + T4y; T6G = FNMS(KP923879532, T6F, KP382683432 * T6E); T74 = FMA(KP382683432, T6F, KP923879532 * T6E); T19 = T15 + T18; T1g = T1c + T1f; T1h = T19 - T1g; T2z = T1g + T19; } { E T2V, T2W, T4z, T4C; T2V = T15 - T18; T2W = Tp - Ts; T2X = T2V - T2W; T3o = T2W + T2V; T4z = T4x - T4y; T4C = T4A - T4B; T4D = FNMS(KP382683432, T4C, KP923879532 * T4z); T5A = FMA(KP382683432, T4z, KP923879532 * T4C); } { E T4G, T4J, T2Y, T2Z; T4G = T4E + T4F; T4J = T4H - T4I; T4K = FMA(KP923879532, T4G, KP382683432 * T4J); T5z = FNMS(KP382683432, T4G, KP923879532 * T4J); T2Y = Ti - Tl; T2Z = T1c - T1f; T30 = T2Y + T2Z; T3n = T2Y - T2Z; } } { E Ty, T4N, T1m, T4Z, TB, T4Y, T1p, T4O, TI, T52, T1w, T4V, TF, T51, T1t; E T4S; { E Tw, Tx, T1n, T1o; Tw = Rp[WS(rs, 1)]; Tx = Rm[WS(rs, 14)]; Ty = Tw + Tx; T4N = Tw - Tx; { E T1k, T1l, Tz, TA; T1k = Ip[WS(rs, 1)]; T1l = Im[WS(rs, 14)]; T1m = T1k - T1l; T4Z = T1k + T1l; Tz = Rp[WS(rs, 9)]; TA = Rm[WS(rs, 6)]; TB = Tz + TA; T4Y = Tz - TA; } T1n = Ip[WS(rs, 9)]; T1o = Im[WS(rs, 6)]; T1p = T1n - T1o; T4O = T1n + T1o; { E TG, TH, T4T, T1u, T1v, T4U; TG = Rm[WS(rs, 2)]; TH = Rp[WS(rs, 13)]; T4T = TG - TH; T1u = Ip[WS(rs, 13)]; T1v = Im[WS(rs, 2)]; T4U = T1u + T1v; TI = TG + TH; T52 = T4T + T4U; T1w = T1u - T1v; T4V = T4T - T4U; } { E TD, TE, T4Q, T1r, T1s, T4R; TD = Rp[WS(rs, 5)]; TE = Rm[WS(rs, 10)]; T4Q = TD - TE; T1r = Ip[WS(rs, 5)]; T1s = Im[WS(rs, 10)]; T4R = T1r + T1s; TF = TD + TE; T51 = T4Q + T4R; T1t = T1r - T1s; T4S = T4Q - T4R; } } { E TC, TJ, T6Q, T6R; TC = Ty + TB; TJ = TF + TI; TK = TC + TJ; T1j = TC - TJ; T6Q = T4Z - T4Y; T6R = KP707106781 * (T4S - T4V); T6S = T6Q + T6R; T7w = T6Q - T6R; } { E T6T, T6U, T1q, T1x; T6T = T4N + T4O; T6U = KP707106781 * (T51 + T52); T6V = T6T - T6U; T7v = T6T + T6U; T1q = T1m + T1p; T1x = T1t + T1w; T1y = T1q - T1x; T2B = T1q + T1x; } { E T3a, T3b, T4P, T4W; T3a = T1m - T1p; T3b = TF - TI; T3c = T3a - T3b; T3S = T3b + T3a; T4P = T4N - T4O; T4W = KP707106781 * (T4S + T4V); T4X = T4P - T4W; T61 = T4P + T4W; } { E T50, T53, T3d, T3e; T50 = T4Y + T4Z; T53 = KP707106781 * (T51 - T52); T54 = T50 - T53; T62 = T50 + T53; T3d = Ty - TB; T3e = T1w - T1t; T3f = T3d - T3e; T3T = T3d + T3e; } } { E TN, T56, T1D, T5i, TQ, T5h, T1G, T57, TX, T5l, T1N, T5e, TU, T5k, T1K; E T5b; { E TL, TM, T1E, T1F; TL = Rm[0]; TM = Rp[WS(rs, 15)]; TN = TL + TM; T56 = TL - TM; { E T1B, T1C, TO, TP; T1B = Ip[WS(rs, 15)]; T1C = Im[0]; T1D = T1B - T1C; T5i = T1B + T1C; TO = Rp[WS(rs, 7)]; TP = Rm[WS(rs, 8)]; TQ = TO + TP; T5h = TO - TP; } T1E = Ip[WS(rs, 7)]; T1F = Im[WS(rs, 8)]; T1G = T1E - T1F; T57 = T1E + T1F; { E TV, TW, T5c, T1L, T1M, T5d; TV = Rm[WS(rs, 4)]; TW = Rp[WS(rs, 11)]; T5c = TV - TW; T1L = Ip[WS(rs, 11)]; T1M = Im[WS(rs, 4)]; T5d = T1L + T1M; TX = TV + TW; T5l = T5c + T5d; T1N = T1L - T1M; T5e = T5c - T5d; } { E TS, TT, T59, T1I, T1J, T5a; TS = Rp[WS(rs, 3)]; TT = Rm[WS(rs, 12)]; T59 = TS - TT; T1I = Ip[WS(rs, 3)]; T1J = Im[WS(rs, 12)]; T5a = T1I + T1J; TU = TS + TT; T5k = T59 + T5a; T1K = T1I - T1J; T5b = T59 - T5a; } } { E TR, TY, T6J, T6K; TR = TN + TQ; TY = TU + TX; TZ = TR + TY; T1A = TR - TY; T6J = KP707106781 * (T5b - T5e); T6K = T5h + T5i; T6L = T6J - T6K; T7z = T6K + T6J; } { E T6M, T6N, T1H, T1O; T6M = T56 + T57; T6N = KP707106781 * (T5k + T5l); T6O = T6M - T6N; T7y = T6M + T6N; T1H = T1D + T1G; T1O = T1K + T1N; T1P = T1H - T1O; T2C = T1H + T1O; } { E T33, T34, T58, T5f; T33 = T1D - T1G; T34 = TU - TX; T35 = T33 - T34; T3P = T34 + T33; T58 = T56 - T57; T5f = KP707106781 * (T5b + T5e); T5g = T58 - T5f; T64 = T58 + T5f; } { E T5j, T5m, T36, T37; T5j = T5h - T5i; T5m = KP707106781 * (T5k - T5l); T5n = T5j - T5m; T65 = T5j + T5m; T36 = TN - TQ; T37 = T1N - T1K; T38 = T36 - T37; T3Q = T36 + T37; } } { E Tv, T10, T2w, T2A, T2D, T2E, T2v, T2x; Tv = Tf + Tu; T10 = TK + TZ; T2w = Tv - T10; T2A = T2y + T2z; T2D = T2B + T2C; T2E = T2A - T2D; Rp[0] = Tv + T10; Rm[0] = T2A + T2D; T2v = W[30]; T2x = W[31]; Rp[WS(rs, 8)] = FNMS(T2x, T2E, T2v * T2w); Rm[WS(rs, 8)] = FMA(T2x, T2w, T2v * T2E); } { E T2I, T2O, T2M, T2Q; { E T2G, T2H, T2K, T2L; T2G = Tf - Tu; T2H = T2C - T2B; T2I = T2G - T2H; T2O = T2G + T2H; T2K = T2y - T2z; T2L = TK - TZ; T2M = T2K - T2L; T2Q = T2L + T2K; } { E T2F, T2J, T2N, T2P; T2F = W[46]; T2J = W[47]; Rp[WS(rs, 12)] = FNMS(T2J, T2M, T2F * T2I); Rm[WS(rs, 12)] = FMA(T2F, T2M, T2J * T2I); T2N = W[14]; T2P = W[15]; Rp[WS(rs, 4)] = FNMS(T2P, T2Q, T2N * T2O); Rm[WS(rs, 4)] = FMA(T2N, T2Q, T2P * T2O); } } { E T1i, T2a, T2o, T2k, T2d, T2l, T1R, T2p; T1i = T12 + T1h; T2a = T1U + T29; T2o = T29 - T1U; T2k = T12 - T1h; { E T2b, T2c, T1z, T1Q; T2b = T1j + T1y; T2c = T1P - T1A; T2d = KP707106781 * (T2b + T2c); T2l = KP707106781 * (T2c - T2b); T1z = T1j - T1y; T1Q = T1A + T1P; T1R = KP707106781 * (T1z + T1Q); T2p = KP707106781 * (T1z - T1Q); } { E T1S, T2e, T11, T1T; T1S = T1i - T1R; T2e = T2a - T2d; T11 = W[38]; T1T = W[39]; Rp[WS(rs, 10)] = FNMS(T1T, T2e, T11 * T1S); Rm[WS(rs, 10)] = FMA(T1T, T1S, T11 * T2e); } { E T2s, T2u, T2r, T2t; T2s = T2k + T2l; T2u = T2o + T2p; T2r = W[22]; T2t = W[23]; Rp[WS(rs, 6)] = FNMS(T2t, T2u, T2r * T2s); Rm[WS(rs, 6)] = FMA(T2r, T2u, T2t * T2s); } { E T2g, T2i, T2f, T2h; T2g = T1i + T1R; T2i = T2a + T2d; T2f = W[6]; T2h = W[7]; Rp[WS(rs, 2)] = FNMS(T2h, T2i, T2f * T2g); Rm[WS(rs, 2)] = FMA(T2h, T2g, T2f * T2i); } { E T2m, T2q, T2j, T2n; T2m = T2k - T2l; T2q = T2o - T2p; T2j = W[54]; T2n = W[55]; Rp[WS(rs, 14)] = FNMS(T2n, T2q, T2j * T2m); Rm[WS(rs, 14)] = FMA(T2j, T2q, T2n * T2m); } } { E T3O, T4a, T40, T4e, T3V, T4f, T43, T4b, T3N, T3Z; T3N = KP707106781 * (T3n + T3o); T3O = T3M - T3N; T4a = T3M + T3N; T3Z = KP707106781 * (T30 + T2X); T40 = T3Y - T3Z; T4e = T3Y + T3Z; { E T3R, T3U, T41, T42; T3R = FNMS(KP382683432, T3Q, KP923879532 * T3P); T3U = FMA(KP923879532, T3S, KP382683432 * T3T); T3V = T3R - T3U; T4f = T3U + T3R; T41 = FNMS(KP382683432, T3S, KP923879532 * T3T); T42 = FMA(KP382683432, T3P, KP923879532 * T3Q); T43 = T41 - T42; T4b = T41 + T42; } { E T3W, T44, T3L, T3X; T3W = T3O - T3V; T44 = T40 - T43; T3L = W[50]; T3X = W[51]; Rp[WS(rs, 13)] = FNMS(T3X, T44, T3L * T3W); Rm[WS(rs, 13)] = FMA(T3X, T3W, T3L * T44); } { E T4i, T4k, T4h, T4j; T4i = T4a + T4b; T4k = T4e + T4f; T4h = W[2]; T4j = W[3]; Rp[WS(rs, 1)] = FNMS(T4j, T4k, T4h * T4i); Rm[WS(rs, 1)] = FMA(T4h, T4k, T4j * T4i); } { E T46, T48, T45, T47; T46 = T3O + T3V; T48 = T40 + T43; T45 = W[18]; T47 = W[19]; Rp[WS(rs, 5)] = FNMS(T47, T48, T45 * T46); Rm[WS(rs, 5)] = FMA(T47, T46, T45 * T48); } { E T4c, T4g, T49, T4d; T4c = T4a - T4b; T4g = T4e - T4f; T49 = W[34]; T4d = W[35]; Rp[WS(rs, 9)] = FNMS(T4d, T4g, T49 * T4c); Rm[WS(rs, 9)] = FMA(T49, T4g, T4d * T4c); } } { E T32, T3A, T3q, T3E, T3h, T3F, T3t, T3B, T31, T3p; T31 = KP707106781 * (T2X - T30); T32 = T2U - T31; T3A = T2U + T31; T3p = KP707106781 * (T3n - T3o); T3q = T3m - T3p; T3E = T3m + T3p; { E T39, T3g, T3r, T3s; T39 = FNMS(KP923879532, T38, KP382683432 * T35); T3g = FMA(KP382683432, T3c, KP923879532 * T3f); T3h = T39 - T3g; T3F = T3g + T39; T3r = FNMS(KP923879532, T3c, KP382683432 * T3f); T3s = FMA(KP923879532, T35, KP382683432 * T38); T3t = T3r - T3s; T3B = T3r + T3s; } { E T3i, T3u, T2R, T3j; T3i = T32 - T3h; T3u = T3q - T3t; T2R = W[58]; T3j = W[59]; Rp[WS(rs, 15)] = FNMS(T3j, T3u, T2R * T3i); Rm[WS(rs, 15)] = FMA(T3j, T3i, T2R * T3u); } { E T3I, T3K, T3H, T3J; T3I = T3A + T3B; T3K = T3E + T3F; T3H = W[10]; T3J = W[11]; Rp[WS(rs, 3)] = FNMS(T3J, T3K, T3H * T3I); Rm[WS(rs, 3)] = FMA(T3H, T3K, T3J * T3I); } { E T3w, T3y, T3v, T3x; T3w = T32 + T3h; T3y = T3q + T3t; T3v = W[26]; T3x = W[27]; Rp[WS(rs, 7)] = FNMS(T3x, T3y, T3v * T3w); Rm[WS(rs, 7)] = FMA(T3x, T3w, T3v * T3y); } { E T3C, T3G, T3z, T3D; T3C = T3A - T3B; T3G = T3E - T3F; T3z = W[42]; T3D = W[43]; Rp[WS(rs, 11)] = FNMS(T3D, T3G, T3z * T3C); Rm[WS(rs, 11)] = FMA(T3z, T3G, T3D * T3C); } } { E T60, T6m, T6f, T6n, T67, T6r, T6c, T6q; { E T5Y, T5Z, T6d, T6e; T5Y = T4o + T4v; T5Z = T5z + T5A; T60 = T5Y + T5Z; T6m = T5Y - T5Z; T6d = FMA(KP195090322, T61, KP980785280 * T62); T6e = FNMS(KP195090322, T64, KP980785280 * T65); T6f = T6d + T6e; T6n = T6e - T6d; } { E T63, T66, T6a, T6b; T63 = FNMS(KP195090322, T62, KP980785280 * T61); T66 = FMA(KP980785280, T64, KP195090322 * T65); T67 = T63 + T66; T6r = T63 - T66; T6a = T5u + T5x; T6b = T4K + T4D; T6c = T6a + T6b; T6q = T6a - T6b; } { E T68, T6g, T5X, T69; T68 = T60 - T67; T6g = T6c - T6f; T5X = W[32]; T69 = W[33]; Ip[WS(rs, 8)] = FNMS(T69, T6g, T5X * T68); Im[WS(rs, 8)] = FMA(T69, T68, T5X * T6g); } { E T6u, T6w, T6t, T6v; T6u = T6m + T6n; T6w = T6q + T6r; T6t = W[16]; T6v = W[17]; Ip[WS(rs, 4)] = FNMS(T6v, T6w, T6t * T6u); Im[WS(rs, 4)] = FMA(T6t, T6w, T6v * T6u); } { E T6i, T6k, T6h, T6j; T6i = T60 + T67; T6k = T6c + T6f; T6h = W[0]; T6j = W[1]; Ip[0] = FNMS(T6j, T6k, T6h * T6i); Im[0] = FMA(T6j, T6i, T6h * T6k); } { E T6o, T6s, T6l, T6p; T6o = T6m - T6n; T6s = T6q - T6r; T6l = W[48]; T6p = W[49]; Ip[WS(rs, 12)] = FNMS(T6p, T6s, T6l * T6o); Im[WS(rs, 12)] = FMA(T6l, T6s, T6p * T6o); } } { E T7u, T7Q, T7J, T7R, T7B, T7V, T7G, T7U; { E T7s, T7t, T7H, T7I; T7s = T6y + T6z; T7t = T73 + T74; T7u = T7s - T7t; T7Q = T7s + T7t; T7H = FMA(KP195090322, T7w, KP980785280 * T7v); T7I = FMA(KP195090322, T7z, KP980785280 * T7y); T7J = T7H - T7I; T7R = T7H + T7I; } { E T7x, T7A, T7E, T7F; T7x = FNMS(KP980785280, T7w, KP195090322 * T7v); T7A = FNMS(KP980785280, T7z, KP195090322 * T7y); T7B = T7x + T7A; T7V = T7x - T7A; T7E = T70 - T71; T7F = T6D - T6G; T7G = T7E + T7F; T7U = T7E - T7F; } { E T7C, T7K, T7r, T7D; T7C = T7u - T7B; T7K = T7G - T7J; T7r = W[44]; T7D = W[45]; Ip[WS(rs, 11)] = FNMS(T7D, T7K, T7r * T7C); Im[WS(rs, 11)] = FMA(T7D, T7C, T7r * T7K); } { E T7Y, T80, T7X, T7Z; T7Y = T7Q + T7R; T80 = T7U - T7V; T7X = W[60]; T7Z = W[61]; Ip[WS(rs, 15)] = FNMS(T7Z, T80, T7X * T7Y); Im[WS(rs, 15)] = FMA(T7X, T80, T7Z * T7Y); } { E T7M, T7O, T7L, T7N; T7M = T7u + T7B; T7O = T7G + T7J; T7L = W[12]; T7N = W[13]; Ip[WS(rs, 3)] = FNMS(T7N, T7O, T7L * T7M); Im[WS(rs, 3)] = FMA(T7N, T7M, T7L * T7O); } { E T7S, T7W, T7P, T7T; T7S = T7Q - T7R; T7W = T7U + T7V; T7P = W[28]; T7T = W[29]; Ip[WS(rs, 7)] = FNMS(T7T, T7W, T7P * T7S); Im[WS(rs, 7)] = FMA(T7P, T7W, T7T * T7S); } } { E T4M, T5M, T5F, T5N, T5p, T5R, T5C, T5Q; { E T4w, T4L, T5D, T5E; T4w = T4o - T4v; T4L = T4D - T4K; T4M = T4w + T4L; T5M = T4w - T4L; T5D = FMA(KP831469612, T4X, KP555570233 * T54); T5E = FNMS(KP831469612, T5g, KP555570233 * T5n); T5F = T5D + T5E; T5N = T5E - T5D; } { E T55, T5o, T5y, T5B; T55 = FNMS(KP831469612, T54, KP555570233 * T4X); T5o = FMA(KP555570233, T5g, KP831469612 * T5n); T5p = T55 + T5o; T5R = T55 - T5o; T5y = T5u - T5x; T5B = T5z - T5A; T5C = T5y + T5B; T5Q = T5y - T5B; } { E T5q, T5G, T4l, T5r; T5q = T4M - T5p; T5G = T5C - T5F; T4l = W[40]; T5r = W[41]; Ip[WS(rs, 10)] = FNMS(T5r, T5G, T4l * T5q); Im[WS(rs, 10)] = FMA(T5r, T5q, T4l * T5G); } { E T5U, T5W, T5T, T5V; T5U = T5M + T5N; T5W = T5Q + T5R; T5T = W[24]; T5V = W[25]; Ip[WS(rs, 6)] = FNMS(T5V, T5W, T5T * T5U); Im[WS(rs, 6)] = FMA(T5T, T5W, T5V * T5U); } { E T5I, T5K, T5H, T5J; T5I = T4M + T5p; T5K = T5C + T5F; T5H = W[8]; T5J = W[9]; Ip[WS(rs, 2)] = FNMS(T5J, T5K, T5H * T5I); Im[WS(rs, 2)] = FMA(T5J, T5I, T5H * T5K); } { E T5O, T5S, T5L, T5P; T5O = T5M - T5N; T5S = T5Q - T5R; T5L = W[56]; T5P = W[57]; Ip[WS(rs, 14)] = FNMS(T5P, T5S, T5L * T5O); Im[WS(rs, 14)] = FMA(T5L, T5S, T5P * T5O); } } { E T6I, T7g, T79, T7h, T6X, T7l, T76, T7k; { E T6A, T6H, T77, T78; T6A = T6y - T6z; T6H = T6D + T6G; T6I = T6A - T6H; T7g = T6A + T6H; T77 = FNMS(KP555570233, T6S, KP831469612 * T6V); T78 = FMA(KP555570233, T6L, KP831469612 * T6O); T79 = T77 - T78; T7h = T77 + T78; } { E T6P, T6W, T72, T75; T6P = FNMS(KP555570233, T6O, KP831469612 * T6L); T6W = FMA(KP831469612, T6S, KP555570233 * T6V); T6X = T6P - T6W; T7l = T6W + T6P; T72 = T70 + T71; T75 = T73 - T74; T76 = T72 - T75; T7k = T72 + T75; } { E T6Y, T7a, T6x, T6Z; T6Y = T6I - T6X; T7a = T76 - T79; T6x = W[52]; T6Z = W[53]; Ip[WS(rs, 13)] = FNMS(T6Z, T7a, T6x * T6Y); Im[WS(rs, 13)] = FMA(T6Z, T6Y, T6x * T7a); } { E T7o, T7q, T7n, T7p; T7o = T7g + T7h; T7q = T7k + T7l; T7n = W[4]; T7p = W[5]; Ip[WS(rs, 1)] = FNMS(T7p, T7q, T7n * T7o); Im[WS(rs, 1)] = FMA(T7n, T7q, T7p * T7o); } { E T7c, T7e, T7b, T7d; T7c = T6I + T6X; T7e = T76 + T79; T7b = W[20]; T7d = W[21]; Ip[WS(rs, 5)] = FNMS(T7d, T7e, T7b * T7c); Im[WS(rs, 5)] = FMA(T7d, T7c, T7b * T7e); } { E T7i, T7m, T7f, T7j; T7i = T7g - T7h; T7m = T7k - T7l; T7f = W[36]; T7j = W[37]; Ip[WS(rs, 9)] = FNMS(T7j, T7m, T7f * T7i); Im[WS(rs, 9)] = FMA(T7f, T7m, T7j * T7i); } } } } } static const tw_instr twinstr[] = { {TW_FULL, 1, 32}, {TW_NEXT, 1, 0} }; static const hc2c_desc desc = { 32, "hc2cb_32", twinstr, &GENUS, {340, 114, 94, 0} }; void X(codelet_hc2cb_32) (planner *p) { X(khc2c_register) (p, hc2cb_32, &desc, HC2C_VIA_RDFT); } #endif /* HAVE_FMA */ fftw-3.3.3/rdft/scalar/r2cb/hc2cb_20.c0000644000175000017500000006753312054411025014101 00000000000000/* * Copyright (c) 2003, 2007-11 Matteo Frigo * Copyright (c) 2003, 2007-11 Massachusetts Institute of Technology * * This program is free software; you can redistribute it and/or modify * it under the terms of the GNU General Public License as published by * the Free Software Foundation; either version 2 of the License, or * (at your option) any later version. * * This program is distributed in the hope that it will be useful, * but WITHOUT ANY WARRANTY; without even the implied warranty of * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the * GNU General Public License for more details. * * You should have received a copy of the GNU General Public License * along with this program; if not, write to the Free Software * Foundation, Inc., 51 Franklin Street, Fifth Floor, Boston, MA 02110-1301 USA * */ /* This file was automatically generated --- DO NOT EDIT */ /* Generated on Sun Nov 25 07:41:55 EST 2012 */ #include "codelet-rdft.h" #ifdef HAVE_FMA /* Generated by: ../../../genfft/gen_hc2c.native -fma -reorder-insns -schedule-for-pipeline -compact -variables 4 -pipeline-latency 4 -sign 1 -n 20 -dif -name hc2cb_20 -include hc2cb.h */ /* * This function contains 246 FP additions, 148 FP multiplications, * (or, 136 additions, 38 multiplications, 110 fused multiply/add), * 112 stack variables, 4 constants, and 80 memory accesses */ #include "hc2cb.h" static void hc2cb_20(R *Rp, R *Ip, R *Rm, R *Im, const R *W, stride rs, INT mb, INT me, INT ms) { DK(KP951056516, +0.951056516295153572116439333379382143405698634); DK(KP559016994, +0.559016994374947424102293417182819058860154590); DK(KP618033988, +0.618033988749894848204586834365638117720309180); DK(KP250000000, +0.250000000000000000000000000000000000000000000); { INT m; for (m = mb, W = W + ((mb - 1) * 38); m < me; m = m + 1, Rp = Rp + ms, Ip = Ip + ms, Rm = Rm - ms, Im = Im - ms, W = W + 38, MAKE_VOLATILE_STRIDE(80, rs)) { E T1T, T1Q, T1P; { E T3z, T4z, TE, T7, T2W, T4e, T2l, T1t, T33, T3H, T3G, T3a, T1i, T2g, T13; E T4H, T4G, T2d, T1B, T4u, T4B, T4A, T4r, T1A, T2s, T3l, T2t, T3s, T2o, T2q; E T1w, T1y, TC, T29, T3E, T3C, T4n, T4l, TN, TL; { E T4, T2U, T3, T3x, T1p, T5, T1q, T1r; { E T1, T2, T1n, T1o; T1 = Rp[0]; T2 = Rm[WS(rs, 9)]; T1n = Ip[0]; T1o = Im[WS(rs, 9)]; T4 = Rp[WS(rs, 5)]; T2U = T1 - T2; T3 = T1 + T2; T3x = T1n + T1o; T1p = T1n - T1o; T5 = Rm[WS(rs, 4)]; T1q = Ip[WS(rs, 5)]; T1r = Im[WS(rs, 4)]; } { E T3o, T4p, TF, Te, T2Z, T4f, T2b, T1a, T3k, T4t, TJ, TA, T39, T4j, T2f; E T12, T3r, T4q, TG, Tl, T32, T4g, T2c, T1h, Tq, T34, Tp, T3f, TR, Tr; E TS, TT; { E Tx, T37, Tw, T3j, TY, Ty, TZ, T10; { E Tb, T2X, Ta, T3m, T16, Tc, T17, T18; { E T8, T9, T14, T15; T8 = Rp[WS(rs, 4)]; { E T3y, T6, T2V, T1s; T3y = T4 - T5; T6 = T4 + T5; T2V = T1q + T1r; T1s = T1q - T1r; T3z = T3x - T3y; T4z = T3y + T3x; TE = T3 - T6; T7 = T3 + T6; T2W = T2U + T2V; T4e = T2U - T2V; T2l = T1p + T1s; T1t = T1p - T1s; T9 = Rm[WS(rs, 5)]; } T14 = Ip[WS(rs, 4)]; T15 = Im[WS(rs, 5)]; Tb = Rp[WS(rs, 9)]; T2X = T8 - T9; Ta = T8 + T9; T3m = T14 + T15; T16 = T14 - T15; Tc = Rm[0]; T17 = Ip[WS(rs, 9)]; T18 = Im[0]; } { E Tu, Tv, TW, TX; Tu = Rm[WS(rs, 7)]; { E T3n, Td, T2Y, T19; T3n = Tb - Tc; Td = Tb + Tc; T2Y = T17 + T18; T19 = T17 - T18; T3o = T3m - T3n; T4p = T3n + T3m; TF = Ta - Td; Te = Ta + Td; T2Z = T2X + T2Y; T4f = T2X - T2Y; T2b = T16 + T19; T1a = T16 - T19; Tv = Rp[WS(rs, 2)]; } TW = Ip[WS(rs, 2)]; TX = Im[WS(rs, 7)]; Tx = Rm[WS(rs, 2)]; T37 = Tu - Tv; Tw = Tu + Tv; T3j = TW + TX; TY = TW - TX; Ty = Rp[WS(rs, 7)]; TZ = Ip[WS(rs, 7)]; T10 = Im[WS(rs, 2)]; } } { E Ti, T30, Th, T3q, T1d, Tj, T1e, T1f; { E Tf, Tg, T1b, T1c; Tf = Rm[WS(rs, 3)]; { E T3i, Tz, T38, T11; T3i = Tx - Ty; Tz = Tx + Ty; T38 = TZ + T10; T11 = TZ - T10; T3k = T3i + T3j; T4t = T3i - T3j; TJ = Tw - Tz; TA = Tw + Tz; T39 = T37 - T38; T4j = T37 + T38; T2f = TY + T11; T12 = TY - T11; Tg = Rp[WS(rs, 6)]; } T1b = Ip[WS(rs, 6)]; T1c = Im[WS(rs, 3)]; Ti = Rp[WS(rs, 1)]; T30 = Tf - Tg; Th = Tf + Tg; T3q = T1b + T1c; T1d = T1b - T1c; Tj = Rm[WS(rs, 8)]; T1e = Ip[WS(rs, 1)]; T1f = Im[WS(rs, 8)]; } { E Tn, To, TP, TQ; Tn = Rp[WS(rs, 8)]; { E T3p, Tk, T31, T1g; T3p = Ti - Tj; Tk = Ti + Tj; T31 = T1e + T1f; T1g = T1e - T1f; T3r = T3p + T3q; T4q = T3p - T3q; TG = Th - Tk; Tl = Th + Tk; T32 = T30 + T31; T4g = T30 - T31; T2c = T1d + T1g; T1h = T1d - T1g; To = Rm[WS(rs, 1)]; } TP = Ip[WS(rs, 8)]; TQ = Im[WS(rs, 1)]; Tq = Rm[WS(rs, 6)]; T34 = Tn - To; Tp = Tn + To; T3f = TP + TQ; TR = TP - TQ; Tr = Rp[WS(rs, 3)]; TS = Ip[WS(rs, 3)]; TT = Im[WS(rs, 6)]; } } } { E T3h, Tt, T1u, T2n, T1v, T4k, T4h, T2m, TH, TK, T4s, TI; T33 = T2Z + T32; T3H = T2Z - T32; { E T3g, Ts, T35, TU; T3g = Tq - Tr; Ts = Tq + Tr; T35 = TS + TT; TU = TS - TT; T3h = T3f - T3g; T4s = T3g + T3f; TI = Tp - Ts; Tt = Tp + Ts; { E T36, T4i, T2e, TV; T36 = T34 - T35; T4i = T34 + T35; T2e = TR + TU; TV = TR - TU; T3G = T36 - T39; T3a = T36 + T39; T1u = T1a + T1h; T1i = T1a - T1h; T2g = T2e - T2f; T2n = T2e + T2f; T1v = TV + T12; T13 = TV - T12; T4H = T4i - T4j; T4k = T4i + T4j; } } T4h = T4f + T4g; T4G = T4f - T4g; T2d = T2b - T2c; T2m = T2b + T2c; TH = TF + TG; T1B = TF - TG; T4u = T4s - T4t; T4B = T4s + T4t; T4A = T4p + T4q; T4r = T4p - T4q; T1A = TI - TJ; TK = TI + TJ; { E Tm, T3B, TB, T3A; Tm = Te + Tl; T2s = Te - Tl; T3l = T3h + T3k; T3B = T3h - T3k; TB = Tt + TA; T2t = Tt - TA; T3s = T3o + T3r; T3A = T3o - T3r; T2o = T2m + T2n; T2q = T2m - T2n; T1w = T1u + T1v; T1y = T1u - T1v; TC = Tm + TB; T29 = Tm - TB; T3E = T3A - T3B; T3C = T3A + T3B; T4n = T4h - T4k; T4l = T4h + T4k; TN = TH - TK; TL = TH + TK; } } } } { E T3d, T3b, T4E, T1x, TM, T4m, T58, T5b, T4D, T5a, T5c, T59, T4C; Rp[0] = T7 + TC; T3d = T33 - T3a; T3b = T33 + T3a; T4E = T4A - T4B; T4C = T4A + T4B; Rm[0] = T2l + T2o; { E T25, T22, T21, T24, T23, T26, T57; T1x = FNMS(KP250000000, T1w, T1t); T25 = T1t + T1w; T22 = TE + TL; TM = FNMS(KP250000000, TL, TE); T21 = W[18]; T24 = W[19]; T4m = FNMS(KP250000000, T4l, T4e); T58 = T4e + T4l; T5b = T4z + T4C; T4D = FNMS(KP250000000, T4C, T4z); T23 = T21 * T22; T26 = T24 * T22; T57 = W[8]; T5a = W[9]; Rp[WS(rs, 5)] = FNMS(T24, T25, T23); Rm[WS(rs, 5)] = FMA(T21, T25, T26); T5c = T57 * T5b; T59 = T57 * T58; } { E T3U, T3Z, T3W, T40, T3V; { E T3c, T48, T4b, T3D, T47, T4a; T3c = FNMS(KP250000000, T3b, T2W); T48 = T2W + T3b; T4b = T3z + T3C; T3D = FNMS(KP250000000, T3C, T3z); Im[WS(rs, 2)] = FMA(T5a, T58, T5c); Ip[WS(rs, 2)] = FNMS(T5a, T5b, T59); T47 = W[28]; T4a = W[29]; { E T3I, T3Y, T42, T3u, T3M, T3X, T3F; { E T3T, T3t, T4c, T49, T3e, T3S; T3T = FMA(KP618033988, T3l, T3s); T3t = FNMS(KP618033988, T3s, T3l); T4c = T47 * T4b; T49 = T47 * T48; T3I = FNMS(KP618033988, T3H, T3G); T3Y = FMA(KP618033988, T3G, T3H); Im[WS(rs, 7)] = FMA(T4a, T48, T4c); Ip[WS(rs, 7)] = FNMS(T4a, T4b, T49); T3e = FNMS(KP559016994, T3d, T3c); T3S = FMA(KP559016994, T3d, T3c); T42 = FMA(KP951056516, T3T, T3S); T3U = FNMS(KP951056516, T3T, T3S); T3u = FNMS(KP951056516, T3t, T3e); T3M = FMA(KP951056516, T3t, T3e); T3X = FMA(KP559016994, T3E, T3D); T3F = FNMS(KP559016994, T3E, T3D); } { E T3P, T45, T44, T46, T43; { E T3w, T3J, T3v, T3K, T2T, T41; T2T = W[4]; T3w = W[5]; T3J = FMA(KP951056516, T3I, T3F); T3P = FNMS(KP951056516, T3I, T3F); T45 = FNMS(KP951056516, T3Y, T3X); T3Z = FMA(KP951056516, T3Y, T3X); T3v = T2T * T3u; T3K = T2T * T3J; T41 = W[36]; T44 = W[37]; Ip[WS(rs, 1)] = FNMS(T3w, T3J, T3v); Im[WS(rs, 1)] = FMA(T3w, T3u, T3K); T46 = T41 * T45; T43 = T41 * T42; } { E T3O, T3Q, T3N, T3L, T3R; T3L = W[12]; T3O = W[13]; Im[WS(rs, 9)] = FMA(T44, T42, T46); Ip[WS(rs, 9)] = FNMS(T44, T45, T43); T3Q = T3L * T3P; T3N = T3L * T3M; T3R = W[20]; T3W = W[21]; Im[WS(rs, 3)] = FMA(T3O, T3M, T3Q); Ip[WS(rs, 3)] = FNMS(T3O, T3P, T3N); T40 = T3R * T3Z; T3V = T3R * T3U; } } } } { E T4U, T4Z, T4W, T50, T4V, T2L, T2I, T2H; { E T4T, T4v, T4I, T4Y, T4o, T4S; T4T = FNMS(KP618033988, T4r, T4u); T4v = FMA(KP618033988, T4u, T4r); Im[WS(rs, 5)] = FMA(T3W, T3U, T40); Ip[WS(rs, 5)] = FNMS(T3W, T3Z, T3V); T4I = FMA(KP618033988, T4H, T4G); T4Y = FNMS(KP618033988, T4G, T4H); T4o = FMA(KP559016994, T4n, T4m); T4S = FNMS(KP559016994, T4n, T4m); { E T52, T4M, T55, T4P, T54, T56, T53; { E T4d, T4w, T4J, T4x, T4y, T4X, T4F, T51, T4K; T4d = W[0]; T4X = FNMS(KP559016994, T4E, T4D); T4F = FMA(KP559016994, T4E, T4D); T4U = FNMS(KP951056516, T4T, T4S); T52 = FMA(KP951056516, T4T, T4S); T4M = FMA(KP951056516, T4v, T4o); T4w = FNMS(KP951056516, T4v, T4o); T4Z = FMA(KP951056516, T4Y, T4X); T55 = FNMS(KP951056516, T4Y, T4X); T4P = FNMS(KP951056516, T4I, T4F); T4J = FMA(KP951056516, T4I, T4F); T4x = T4d * T4w; T4y = W[1]; T51 = W[32]; T4K = T4d * T4J; T54 = W[33]; Ip[0] = FNMS(T4y, T4J, T4x); T56 = T51 * T55; T53 = T51 * T52; Im[0] = FMA(T4y, T4w, T4K); } { E T4O, T4Q, T4N, T4L, T4R; T4L = W[16]; Im[WS(rs, 8)] = FMA(T54, T52, T56); Ip[WS(rs, 8)] = FNMS(T54, T55, T53); T4O = W[17]; T4Q = T4L * T4P; T4N = T4L * T4M; T4R = W[24]; T4W = W[25]; Im[WS(rs, 4)] = FMA(T4O, T4M, T4Q); Ip[WS(rs, 4)] = FNMS(T4O, T4P, T4N); T50 = T4R * T4Z; T4V = T4R * T4U; } } } { E T2K, T2u, T2F, T2h, T28, T2J, T2r, T2p; T2K = FNMS(KP618033988, T2s, T2t); T2u = FMA(KP618033988, T2t, T2s); Im[WS(rs, 6)] = FMA(T4W, T4U, T50); Ip[WS(rs, 6)] = FNMS(T4W, T4Z, T4V); T2p = FNMS(KP250000000, T2o, T2l); T2F = FNMS(KP618033988, T2d, T2g); T2h = FMA(KP618033988, T2g, T2d); T28 = FNMS(KP250000000, TC, T7); T2J = FNMS(KP559016994, T2q, T2p); T2r = FMA(KP559016994, T2q, T2p); { E T2B, T2G, T2y, T2R, T2Q, T2P, T2A, T2x; { E T2k, T2v, T27, T2O, T2i, T2a, T2E; T2k = W[7]; T2a = FMA(KP559016994, T29, T28); T2E = FNMS(KP559016994, T29, T28); T2B = FMA(KP951056516, T2u, T2r); T2v = FNMS(KP951056516, T2u, T2r); T27 = W[6]; T2O = FMA(KP951056516, T2F, T2E); T2G = FNMS(KP951056516, T2F, T2E); T2i = FMA(KP951056516, T2h, T2a); T2y = FNMS(KP951056516, T2h, T2a); { E T2N, T2j, T2w, T2S; T2L = FMA(KP951056516, T2K, T2J); T2R = FNMS(KP951056516, T2K, T2J); T2Q = W[23]; T2N = W[22]; T2j = T27 * T2i; T2w = T2k * T2i; T2S = T2Q * T2O; T2P = T2N * T2O; Rp[WS(rs, 2)] = FNMS(T2k, T2v, T2j); Rm[WS(rs, 2)] = FMA(T27, T2v, T2w); Rm[WS(rs, 6)] = FMA(T2N, T2R, T2S); } } Rp[WS(rs, 6)] = FNMS(T2Q, T2R, T2P); T2A = W[31]; T2x = W[30]; { E T2D, T2M, T2C, T2z; T2I = W[15]; T2C = T2A * T2y; T2z = T2x * T2y; T2D = W[14]; T2M = T2I * T2G; Rm[WS(rs, 8)] = FMA(T2x, T2B, T2C); Rp[WS(rs, 8)] = FNMS(T2A, T2B, T2z); T2H = T2D * T2G; Rm[WS(rs, 4)] = FMA(T2D, T2L, T2M); } } } { E T1S, T1C, T1j, T1N, T1z, T1R; T1S = FMA(KP618033988, T1A, T1B); T1C = FNMS(KP618033988, T1B, T1A); Rp[WS(rs, 4)] = FNMS(T2I, T2L, T2H); T1j = FNMS(KP618033988, T1i, T13); T1N = FMA(KP618033988, T13, T1i); T1z = FNMS(KP559016994, T1y, T1x); T1R = FMA(KP559016994, T1y, T1x); { E T1J, T1O, T1G, T1Z, T1Y, T1X, T1I, T1F; { E T1m, T1D, TD, T1W, T1k, T1M, TO; T1m = W[3]; T1M = FMA(KP559016994, TN, TM); TO = FNMS(KP559016994, TN, TM); T1D = FNMS(KP951056516, T1C, T1z); T1J = FMA(KP951056516, T1C, T1z); TD = W[2]; T1O = FNMS(KP951056516, T1N, T1M); T1W = FMA(KP951056516, T1N, T1M); T1G = FNMS(KP951056516, T1j, TO); T1k = FMA(KP951056516, T1j, TO); { E T1V, T1l, T1E, T20; T1Z = FNMS(KP951056516, T1S, T1R); T1T = FMA(KP951056516, T1S, T1R); T1Y = W[27]; T1V = W[26]; T1l = TD * T1k; T1E = T1m * T1k; T20 = T1Y * T1W; T1X = T1V * T1W; Rp[WS(rs, 1)] = FNMS(T1m, T1D, T1l); Rm[WS(rs, 1)] = FMA(TD, T1D, T1E); Rm[WS(rs, 7)] = FMA(T1V, T1Z, T20); } } Rp[WS(rs, 7)] = FNMS(T1Y, T1Z, T1X); T1I = W[35]; T1F = W[34]; { E T1L, T1U, T1K, T1H; T1Q = W[11]; T1K = T1I * T1G; T1H = T1F * T1G; T1L = W[10]; T1U = T1Q * T1O; Rm[WS(rs, 9)] = FMA(T1F, T1J, T1K); Rp[WS(rs, 9)] = FNMS(T1I, T1J, T1H); T1P = T1L * T1O; Rm[WS(rs, 3)] = FMA(T1L, T1T, T1U); } } } } } } } Rp[WS(rs, 3)] = FNMS(T1Q, T1T, T1P); } } } static const tw_instr twinstr[] = { {TW_FULL, 1, 20}, {TW_NEXT, 1, 0} }; static const hc2c_desc desc = { 20, "hc2cb_20", twinstr, &GENUS, {136, 38, 110, 0} }; void X(codelet_hc2cb_20) (planner *p) { X(khc2c_register) (p, hc2cb_20, &desc, HC2C_VIA_RDFT); } #else /* HAVE_FMA */ /* Generated by: ../../../genfft/gen_hc2c.native -compact -variables 4 -pipeline-latency 4 -sign 1 -n 20 -dif -name hc2cb_20 -include hc2cb.h */ /* * This function contains 246 FP additions, 124 FP multiplications, * (or, 184 additions, 62 multiplications, 62 fused multiply/add), * 97 stack variables, 4 constants, and 80 memory accesses */ #include "hc2cb.h" static void hc2cb_20(R *Rp, R *Ip, R *Rm, R *Im, const R *W, stride rs, INT mb, INT me, INT ms) { DK(KP250000000, +0.250000000000000000000000000000000000000000000); DK(KP559016994, +0.559016994374947424102293417182819058860154590); DK(KP587785252, +0.587785252292473129168705954639072768597652438); DK(KP951056516, +0.951056516295153572116439333379382143405698634); { INT m; for (m = mb, W = W + ((mb - 1) * 38); m < me; m = m + 1, Rp = Rp + ms, Ip = Ip + ms, Rm = Rm - ms, Im = Im - ms, W = W + 38, MAKE_VOLATILE_STRIDE(80, rs)) { E T7, T3T, T49, TE, T1v, T2T, T3g, T2d, T13, T3n, T3o, T1i, T26, T4e, T4d; E T23, T1n, T42, T3Z, T1m, T2h, T2I, T2i, T2P, T30, T37, T38, Tm, TB, TC; E T46, T47, T4a, T2a, T2b, T2e, T1w, T1x, T1y, T3O, T3R, T3U, T3h, T3i, T3j; E TH, TK, TL; { E T3, T2R, T1r, T3e, T6, T3f, T1u, T2S; { E T1, T2, T1p, T1q; T1 = Rp[0]; T2 = Rm[WS(rs, 9)]; T3 = T1 + T2; T2R = T1 - T2; T1p = Ip[0]; T1q = Im[WS(rs, 9)]; T1r = T1p - T1q; T3e = T1p + T1q; } { E T4, T5, T1s, T1t; T4 = Rp[WS(rs, 5)]; T5 = Rm[WS(rs, 4)]; T6 = T4 + T5; T3f = T4 - T5; T1s = Ip[WS(rs, 5)]; T1t = Im[WS(rs, 4)]; T1u = T1s - T1t; T2S = T1s + T1t; } T7 = T3 + T6; T3T = T2R - T2S; T49 = T3f + T3e; TE = T3 - T6; T1v = T1r - T1u; T2T = T2R + T2S; T3g = T3e - T3f; T2d = T1r + T1u; } { E Te, T3M, T3X, TF, TV, T2E, T2W, T21, TA, T3Q, T41, TJ, T1h, T2O, T36; E T25, Tl, T3N, T3Y, TG, T12, T2H, T2Z, T22, Tt, T3P, T40, TI, T1a, T2L; E T33, T24; { E Ta, T2U, TR, T2C, Td, T2D, TU, T2V; { E T8, T9, TP, TQ; T8 = Rp[WS(rs, 4)]; T9 = Rm[WS(rs, 5)]; Ta = T8 + T9; T2U = T8 - T9; TP = Ip[WS(rs, 4)]; TQ = Im[WS(rs, 5)]; TR = TP - TQ; T2C = TP + TQ; } { E Tb, Tc, TS, TT; Tb = Rp[WS(rs, 9)]; Tc = Rm[0]; Td = Tb + Tc; T2D = Tb - Tc; TS = Ip[WS(rs, 9)]; TT = Im[0]; TU = TS - TT; T2V = TS + TT; } Te = Ta + Td; T3M = T2U - T2V; T3X = T2D + T2C; TF = Ta - Td; TV = TR - TU; T2E = T2C - T2D; T2W = T2U + T2V; T21 = TR + TU; } { E Tw, T34, T1d, T2N, Tz, T2M, T1g, T35; { E Tu, Tv, T1b, T1c; Tu = Rm[WS(rs, 7)]; Tv = Rp[WS(rs, 2)]; Tw = Tu + Tv; T34 = Tu - Tv; T1b = Ip[WS(rs, 2)]; T1c = Im[WS(rs, 7)]; T1d = T1b - T1c; T2N = T1b + T1c; } { E Tx, Ty, T1e, T1f; Tx = Rm[WS(rs, 2)]; Ty = Rp[WS(rs, 7)]; Tz = Tx + Ty; T2M = Tx - Ty; T1e = Ip[WS(rs, 7)]; T1f = Im[WS(rs, 2)]; T1g = T1e - T1f; T35 = T1e + T1f; } TA = Tw + Tz; T3Q = T34 + T35; T41 = T2M - T2N; TJ = Tw - Tz; T1h = T1d - T1g; T2O = T2M + T2N; T36 = T34 - T35; T25 = T1d + T1g; } { E Th, T2X, TY, T2G, Tk, T2F, T11, T2Y; { E Tf, Tg, TW, TX; Tf = Rm[WS(rs, 3)]; Tg = Rp[WS(rs, 6)]; Th = Tf + Tg; T2X = Tf - Tg; TW = Ip[WS(rs, 6)]; TX = Im[WS(rs, 3)]; TY = TW - TX; T2G = TW + TX; } { E Ti, Tj, TZ, T10; Ti = Rp[WS(rs, 1)]; Tj = Rm[WS(rs, 8)]; Tk = Ti + Tj; T2F = Ti - Tj; TZ = Ip[WS(rs, 1)]; T10 = Im[WS(rs, 8)]; T11 = TZ - T10; T2Y = TZ + T10; } Tl = Th + Tk; T3N = T2X - T2Y; T3Y = T2F - T2G; TG = Th - Tk; T12 = TY - T11; T2H = T2F + T2G; T2Z = T2X + T2Y; T22 = TY + T11; } { E Tp, T31, T16, T2J, Ts, T2K, T19, T32; { E Tn, To, T14, T15; Tn = Rp[WS(rs, 8)]; To = Rm[WS(rs, 1)]; Tp = Tn + To; T31 = Tn - To; T14 = Ip[WS(rs, 8)]; T15 = Im[WS(rs, 1)]; T16 = T14 - T15; T2J = T14 + T15; } { E Tq, Tr, T17, T18; Tq = Rm[WS(rs, 6)]; Tr = Rp[WS(rs, 3)]; Ts = Tq + Tr; T2K = Tq - Tr; T17 = Ip[WS(rs, 3)]; T18 = Im[WS(rs, 6)]; T19 = T17 - T18; T32 = T17 + T18; } Tt = Tp + Ts; T3P = T31 + T32; T40 = T2K + T2J; TI = Tp - Ts; T1a = T16 - T19; T2L = T2J - T2K; T33 = T31 - T32; T24 = T16 + T19; } T13 = TV - T12; T3n = T2W - T2Z; T3o = T33 - T36; T1i = T1a - T1h; T26 = T24 - T25; T4e = T3P - T3Q; T4d = T3M - T3N; T23 = T21 - T22; T1n = TI - TJ; T42 = T40 - T41; T3Z = T3X - T3Y; T1m = TF - TG; T2h = Te - Tl; T2I = T2E + T2H; T2i = Tt - TA; T2P = T2L + T2O; T30 = T2W + T2Z; T37 = T33 + T36; T38 = T30 + T37; Tm = Te + Tl; TB = Tt + TA; TC = Tm + TB; T46 = T3X + T3Y; T47 = T40 + T41; T4a = T46 + T47; T2a = T21 + T22; T2b = T24 + T25; T2e = T2a + T2b; T1w = TV + T12; T1x = T1a + T1h; T1y = T1w + T1x; T3O = T3M + T3N; T3R = T3P + T3Q; T3U = T3O + T3R; T3h = T2E - T2H; T3i = T2L - T2O; T3j = T3h + T3i; TH = TF + TG; TK = TI + TJ; TL = TH + TK; } Rp[0] = T7 + TC; Rm[0] = T2d + T2e; { E T1U, T1W, T1T, T1V; T1U = TE + TL; T1W = T1v + T1y; T1T = W[18]; T1V = W[19]; Rp[WS(rs, 5)] = FNMS(T1V, T1W, T1T * T1U); Rm[WS(rs, 5)] = FMA(T1V, T1U, T1T * T1W); } { E T4y, T4A, T4x, T4z; T4y = T3T + T3U; T4A = T49 + T4a; T4x = W[8]; T4z = W[9]; Ip[WS(rs, 2)] = FNMS(T4z, T4A, T4x * T4y); Im[WS(rs, 2)] = FMA(T4x, T4A, T4z * T4y); } { E T3I, T3K, T3H, T3J; T3I = T2T + T38; T3K = T3g + T3j; T3H = W[28]; T3J = W[29]; Ip[WS(rs, 7)] = FNMS(T3J, T3K, T3H * T3I); Im[WS(rs, 7)] = FMA(T3H, T3K, T3J * T3I); } { E T27, T2j, T2v, T2r, T2g, T2u, T20, T2q; T27 = FMA(KP951056516, T23, KP587785252 * T26); T2j = FMA(KP951056516, T2h, KP587785252 * T2i); T2v = FNMS(KP951056516, T2i, KP587785252 * T2h); T2r = FNMS(KP951056516, T26, KP587785252 * T23); { E T2c, T2f, T1Y, T1Z; T2c = KP559016994 * (T2a - T2b); T2f = FNMS(KP250000000, T2e, T2d); T2g = T2c + T2f; T2u = T2f - T2c; T1Y = KP559016994 * (Tm - TB); T1Z = FNMS(KP250000000, TC, T7); T20 = T1Y + T1Z; T2q = T1Z - T1Y; } { E T28, T2k, T1X, T29; T28 = T20 + T27; T2k = T2g - T2j; T1X = W[6]; T29 = W[7]; Rp[WS(rs, 2)] = FNMS(T29, T2k, T1X * T28); Rm[WS(rs, 2)] = FMA(T29, T28, T1X * T2k); } { E T2y, T2A, T2x, T2z; T2y = T2q - T2r; T2A = T2v + T2u; T2x = W[22]; T2z = W[23]; Rp[WS(rs, 6)] = FNMS(T2z, T2A, T2x * T2y); Rm[WS(rs, 6)] = FMA(T2z, T2y, T2x * T2A); } { E T2m, T2o, T2l, T2n; T2m = T20 - T27; T2o = T2j + T2g; T2l = W[30]; T2n = W[31]; Rp[WS(rs, 8)] = FNMS(T2n, T2o, T2l * T2m); Rm[WS(rs, 8)] = FMA(T2n, T2m, T2l * T2o); } { E T2s, T2w, T2p, T2t; T2s = T2q + T2r; T2w = T2u - T2v; T2p = W[14]; T2t = W[15]; Rp[WS(rs, 4)] = FNMS(T2t, T2w, T2p * T2s); Rm[WS(rs, 4)] = FMA(T2t, T2s, T2p * T2w); } } { E T43, T4f, T4r, T4m, T4c, T4q, T3W, T4n; T43 = FMA(KP951056516, T3Z, KP587785252 * T42); T4f = FMA(KP951056516, T4d, KP587785252 * T4e); T4r = FNMS(KP951056516, T4e, KP587785252 * T4d); T4m = FNMS(KP951056516, T42, KP587785252 * T3Z); { E T48, T4b, T3S, T3V; T48 = KP559016994 * (T46 - T47); T4b = FNMS(KP250000000, T4a, T49); T4c = T48 + T4b; T4q = T4b - T48; T3S = KP559016994 * (T3O - T3R); T3V = FNMS(KP250000000, T3U, T3T); T3W = T3S + T3V; T4n = T3V - T3S; } { E T44, T4g, T3L, T45; T44 = T3W - T43; T4g = T4c + T4f; T3L = W[0]; T45 = W[1]; Ip[0] = FNMS(T45, T4g, T3L * T44); Im[0] = FMA(T3L, T4g, T45 * T44); } { E T4u, T4w, T4t, T4v; T4u = T4n - T4m; T4w = T4q + T4r; T4t = W[32]; T4v = W[33]; Ip[WS(rs, 8)] = FNMS(T4v, T4w, T4t * T4u); Im[WS(rs, 8)] = FMA(T4t, T4w, T4v * T4u); } { E T4i, T4k, T4h, T4j; T4i = T43 + T3W; T4k = T4c - T4f; T4h = W[16]; T4j = W[17]; Ip[WS(rs, 4)] = FNMS(T4j, T4k, T4h * T4i); Im[WS(rs, 4)] = FMA(T4h, T4k, T4j * T4i); } { E T4o, T4s, T4l, T4p; T4o = T4m + T4n; T4s = T4q - T4r; T4l = W[24]; T4p = W[25]; Ip[WS(rs, 6)] = FNMS(T4p, T4s, T4l * T4o); Im[WS(rs, 6)] = FMA(T4l, T4s, T4p * T4o); } } { E T1j, T1o, T1M, T1J, T1B, T1N, TO, T1I; T1j = FNMS(KP951056516, T1i, KP587785252 * T13); T1o = FNMS(KP951056516, T1n, KP587785252 * T1m); T1M = FMA(KP951056516, T1m, KP587785252 * T1n); T1J = FMA(KP951056516, T13, KP587785252 * T1i); { E T1z, T1A, TM, TN; T1z = FNMS(KP250000000, T1y, T1v); T1A = KP559016994 * (T1w - T1x); T1B = T1z - T1A; T1N = T1A + T1z; TM = FNMS(KP250000000, TL, TE); TN = KP559016994 * (TH - TK); TO = TM - TN; T1I = TN + TM; } { E T1k, T1C, TD, T1l; T1k = TO - T1j; T1C = T1o + T1B; TD = W[2]; T1l = W[3]; Rp[WS(rs, 1)] = FNMS(T1l, T1C, TD * T1k); Rm[WS(rs, 1)] = FMA(T1l, T1k, TD * T1C); } { E T1Q, T1S, T1P, T1R; T1Q = T1I + T1J; T1S = T1N - T1M; T1P = W[26]; T1R = W[27]; Rp[WS(rs, 7)] = FNMS(T1R, T1S, T1P * T1Q); Rm[WS(rs, 7)] = FMA(T1R, T1Q, T1P * T1S); } { E T1E, T1G, T1D, T1F; T1E = TO + T1j; T1G = T1B - T1o; T1D = W[34]; T1F = W[35]; Rp[WS(rs, 9)] = FNMS(T1F, T1G, T1D * T1E); Rm[WS(rs, 9)] = FMA(T1F, T1E, T1D * T1G); } { E T1K, T1O, T1H, T1L; T1K = T1I - T1J; T1O = T1M + T1N; T1H = W[10]; T1L = W[11]; Rp[WS(rs, 3)] = FNMS(T1L, T1O, T1H * T1K); Rm[WS(rs, 3)] = FMA(T1L, T1K, T1H * T1O); } } { E T2Q, T3p, T3B, T3x, T3m, T3A, T3b, T3w; T2Q = FNMS(KP951056516, T2P, KP587785252 * T2I); T3p = FNMS(KP951056516, T3o, KP587785252 * T3n); T3B = FMA(KP951056516, T3n, KP587785252 * T3o); T3x = FMA(KP951056516, T2I, KP587785252 * T2P); { E T3k, T3l, T39, T3a; T3k = FNMS(KP250000000, T3j, T3g); T3l = KP559016994 * (T3h - T3i); T3m = T3k - T3l; T3A = T3l + T3k; T39 = FNMS(KP250000000, T38, T2T); T3a = KP559016994 * (T30 - T37); T3b = T39 - T3a; T3w = T3a + T39; } { E T3c, T3q, T2B, T3d; T3c = T2Q + T3b; T3q = T3m - T3p; T2B = W[4]; T3d = W[5]; Ip[WS(rs, 1)] = FNMS(T3d, T3q, T2B * T3c); Im[WS(rs, 1)] = FMA(T2B, T3q, T3d * T3c); } { E T3E, T3G, T3D, T3F; T3E = T3x + T3w; T3G = T3A - T3B; T3D = W[36]; T3F = W[37]; Ip[WS(rs, 9)] = FNMS(T3F, T3G, T3D * T3E); Im[WS(rs, 9)] = FMA(T3D, T3G, T3F * T3E); } { E T3s, T3u, T3r, T3t; T3s = T3b - T2Q; T3u = T3m + T3p; T3r = W[12]; T3t = W[13]; Ip[WS(rs, 3)] = FNMS(T3t, T3u, T3r * T3s); Im[WS(rs, 3)] = FMA(T3r, T3u, T3t * T3s); } { E T3y, T3C, T3v, T3z; T3y = T3w - T3x; T3C = T3A + T3B; T3v = W[20]; T3z = W[21]; Ip[WS(rs, 5)] = FNMS(T3z, T3C, T3v * T3y); Im[WS(rs, 5)] = FMA(T3v, T3C, T3z * T3y); } } } } } static const tw_instr twinstr[] = { {TW_FULL, 1, 20}, {TW_NEXT, 1, 0} }; static const hc2c_desc desc = { 20, "hc2cb_20", twinstr, &GENUS, {184, 62, 62, 0} }; void X(codelet_hc2cb_20) (planner *p) { X(khc2c_register) (p, hc2cb_20, &desc, HC2C_VIA_RDFT); } #endif /* HAVE_FMA */ fftw-3.3.3/rdft/scalar/r2cb/hc2cb2_4.c0000644000175000017500000001255112054411025014073 00000000000000/* * Copyright (c) 2003, 2007-11 Matteo Frigo * Copyright (c) 2003, 2007-11 Massachusetts Institute of Technology * * This program is free software; you can redistribute it and/or modify * it under the terms of the GNU General Public License as published by * the Free Software Foundation; either version 2 of the License, or * (at your option) any later version. * * This program is distributed in the hope that it will be useful, * but WITHOUT ANY WARRANTY; without even the implied warranty of * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the * GNU General Public License for more details. * * You should have received a copy of the GNU General Public License * along with this program; if not, write to the Free Software * Foundation, Inc., 51 Franklin Street, Fifth Floor, Boston, MA 02110-1301 USA * */ /* This file was automatically generated --- DO NOT EDIT */ /* Generated on Sun Nov 25 07:41:57 EST 2012 */ #include "codelet-rdft.h" #ifdef HAVE_FMA /* Generated by: ../../../genfft/gen_hc2c.native -fma -reorder-insns -schedule-for-pipeline -compact -variables 4 -pipeline-latency 4 -sign 1 -twiddle-log3 -precompute-twiddles -n 4 -dif -name hc2cb2_4 -include hc2cb.h */ /* * This function contains 24 FP additions, 16 FP multiplications, * (or, 16 additions, 8 multiplications, 8 fused multiply/add), * 30 stack variables, 0 constants, and 16 memory accesses */ #include "hc2cb.h" static void hc2cb2_4(R *Rp, R *Ip, R *Rm, R *Im, const R *W, stride rs, INT mb, INT me, INT ms) { { INT m; for (m = mb, W = W + ((mb - 1) * 4); m < me; m = m + 1, Rp = Rp + ms, Ip = Ip + ms, Rm = Rm - ms, Im = Im - ms, W = W + 4, MAKE_VOLATILE_STRIDE(16, rs)) { E Tg, Tc, Te, To, Tn; { E T7, Tb, T8, Ta; T7 = W[0]; Tb = W[3]; T8 = W[2]; Ta = W[1]; { E Tu, Tj, T3, Tm, Tx, Tr, T6, Tt; { E T4, Tp, Tq, T5; { E T1, T2, Tk, Tl; { E Th, Tf, T9, Ti; Th = Ip[0]; Tf = T7 * Tb; T9 = T7 * T8; Ti = Im[WS(rs, 1)]; T1 = Rp[0]; Tg = FNMS(Ta, T8, Tf); Tc = FMA(Ta, Tb, T9); Tu = Th + Ti; Tj = Th - Ti; T2 = Rm[WS(rs, 1)]; } Tk = Ip[WS(rs, 1)]; Tl = Im[0]; T4 = Rp[WS(rs, 1)]; T3 = T1 + T2; Tp = T1 - T2; Tm = Tk - Tl; Tq = Tk + Tl; T5 = Rm[0]; } Tx = Tp + Tq; Tr = Tp - Tq; T6 = T4 + T5; Tt = T4 - T5; } { E Tz, Tv, Td, Ts, Tw, TA, Ty; Rm[0] = Tj + Tm; Ts = T7 * Tr; Tz = Tu - Tt; Tv = Tt + Tu; Rp[0] = T3 + T6; Td = T3 - T6; Ip[0] = FNMS(Ta, Tv, Ts); Tw = T7 * Tv; TA = T8 * Tz; Ty = T8 * Tx; Te = Tc * Td; Im[0] = FMA(Ta, Tr, Tw); Im[WS(rs, 1)] = FMA(Tb, Tx, TA); Ip[WS(rs, 1)] = FNMS(Tb, Tz, Ty); To = Tg * Td; Tn = Tj - Tm; } } } Rm[WS(rs, 1)] = FMA(Tc, Tn, To); Rp[WS(rs, 1)] = FNMS(Tg, Tn, Te); } } } static const tw_instr twinstr[] = { {TW_CEXP, 1, 1}, {TW_CEXP, 1, 3}, {TW_NEXT, 1, 0} }; static const hc2c_desc desc = { 4, "hc2cb2_4", twinstr, &GENUS, {16, 8, 8, 0} }; void X(codelet_hc2cb2_4) (planner *p) { X(khc2c_register) (p, hc2cb2_4, &desc, HC2C_VIA_RDFT); } #else /* HAVE_FMA */ /* Generated by: ../../../genfft/gen_hc2c.native -compact -variables 4 -pipeline-latency 4 -sign 1 -twiddle-log3 -precompute-twiddles -n 4 -dif -name hc2cb2_4 -include hc2cb.h */ /* * This function contains 24 FP additions, 16 FP multiplications, * (or, 16 additions, 8 multiplications, 8 fused multiply/add), * 21 stack variables, 0 constants, and 16 memory accesses */ #include "hc2cb.h" static void hc2cb2_4(R *Rp, R *Ip, R *Rm, R *Im, const R *W, stride rs, INT mb, INT me, INT ms) { { INT m; for (m = mb, W = W + ((mb - 1) * 4); m < me; m = m + 1, Rp = Rp + ms, Ip = Ip + ms, Rm = Rm - ms, Im = Im - ms, W = W + 4, MAKE_VOLATILE_STRIDE(16, rs)) { E T7, T9, T8, Ta, Tb, Td; T7 = W[0]; T9 = W[1]; T8 = W[2]; Ta = W[3]; Tb = FMA(T7, T8, T9 * Ta); Td = FNMS(T9, T8, T7 * Ta); { E T3, Tl, Tg, Tp, T6, To, Tj, Tm, Tc, Tk; { E T1, T2, Te, Tf; T1 = Rp[0]; T2 = Rm[WS(rs, 1)]; T3 = T1 + T2; Tl = T1 - T2; Te = Ip[0]; Tf = Im[WS(rs, 1)]; Tg = Te - Tf; Tp = Te + Tf; } { E T4, T5, Th, Ti; T4 = Rp[WS(rs, 1)]; T5 = Rm[0]; T6 = T4 + T5; To = T4 - T5; Th = Ip[WS(rs, 1)]; Ti = Im[0]; Tj = Th - Ti; Tm = Th + Ti; } Rp[0] = T3 + T6; Rm[0] = Tg + Tj; Tc = T3 - T6; Tk = Tg - Tj; Rp[WS(rs, 1)] = FNMS(Td, Tk, Tb * Tc); Rm[WS(rs, 1)] = FMA(Td, Tc, Tb * Tk); { E Tn, Tq, Tr, Ts; Tn = Tl - Tm; Tq = To + Tp; Ip[0] = FNMS(T9, Tq, T7 * Tn); Im[0] = FMA(T7, Tq, T9 * Tn); Tr = Tl + Tm; Ts = Tp - To; Ip[WS(rs, 1)] = FNMS(Ta, Ts, T8 * Tr); Im[WS(rs, 1)] = FMA(T8, Ts, Ta * Tr); } } } } } static const tw_instr twinstr[] = { {TW_CEXP, 1, 1}, {TW_CEXP, 1, 3}, {TW_NEXT, 1, 0} }; static const hc2c_desc desc = { 4, "hc2cb2_4", twinstr, &GENUS, {16, 8, 8, 0} }; void X(codelet_hc2cb2_4) (planner *p) { X(khc2c_register) (p, hc2cb2_4, &desc, HC2C_VIA_RDFT); } #endif /* HAVE_FMA */ fftw-3.3.3/rdft/scalar/r2cb/hc2cb2_8.c0000644000175000017500000002377612054411025014112 00000000000000/* * Copyright (c) 2003, 2007-11 Matteo Frigo * Copyright (c) 2003, 2007-11 Massachusetts Institute of Technology * * This program is free software; you can redistribute it and/or modify * it under the terms of the GNU General Public License as published by * the Free Software Foundation; either version 2 of the License, or * (at your option) any later version. * * This program is distributed in the hope that it will be useful, * but WITHOUT ANY WARRANTY; without even the implied warranty of * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the * GNU General Public License for more details. * * You should have received a copy of the GNU General Public License * along with this program; if not, write to the Free Software * Foundation, Inc., 51 Franklin Street, Fifth Floor, Boston, MA 02110-1301 USA * */ /* This file was automatically generated --- DO NOT EDIT */ /* Generated on Sun Nov 25 07:41:57 EST 2012 */ #include "codelet-rdft.h" #ifdef HAVE_FMA /* Generated by: ../../../genfft/gen_hc2c.native -fma -reorder-insns -schedule-for-pipeline -compact -variables 4 -pipeline-latency 4 -sign 1 -twiddle-log3 -precompute-twiddles -n 8 -dif -name hc2cb2_8 -include hc2cb.h */ /* * This function contains 74 FP additions, 50 FP multiplications, * (or, 44 additions, 20 multiplications, 30 fused multiply/add), * 64 stack variables, 1 constants, and 32 memory accesses */ #include "hc2cb.h" static void hc2cb2_8(R *Rp, R *Ip, R *Rm, R *Im, const R *W, stride rs, INT mb, INT me, INT ms) { DK(KP707106781, +0.707106781186547524400844362104849039284835938); { INT m; for (m = mb, W = W + ((mb - 1) * 6); m < me; m = m + 1, Rp = Rp + ms, Ip = Ip + ms, Rm = Rm - ms, Im = Im - ms, W = W + 6, MAKE_VOLATILE_STRIDE(32, rs)) { E Tf, Ti, TK, Tq, TH, TT, TX, TW, TY, TU, TI; { E Tg, Tl, Tp, Th, T1n, T1t, Tj; Tf = W[0]; Tg = W[2]; Tl = W[4]; Tp = W[5]; Ti = W[1]; Th = Tf * Tg; T1n = Tf * Tl; T1t = Tf * Tp; Tj = W[3]; { E T1o, T1u, Tk, T1b, To, T1e, T13, TP, T1p, T7, T1h, T1v, TZ, Tv, T1i; E TB, TA, TQ, Te, T1w, TE, T1j; { E Tr, T3, Ts, T1f, TO, TL, T6, Tt; { E TM, TN, T4, T5; { E T1, Tn, T2, TJ, Tm; T1 = Rp[0]; T1o = FMA(Ti, Tp, T1n); T1u = FNMS(Ti, Tl, T1t); Tk = FMA(Ti, Tj, Th); T1b = FNMS(Ti, Tj, Th); Tn = Tf * Tj; T2 = Rm[WS(rs, 3)]; TM = Ip[0]; TJ = Tk * Tp; Tm = Tk * Tl; To = FNMS(Ti, Tg, Tn); T1e = FMA(Ti, Tg, Tn); Tr = T1 - T2; T3 = T1 + T2; TK = FNMS(To, Tl, TJ); Tq = FMA(To, Tp, Tm); TN = Im[WS(rs, 3)]; } T4 = Rp[WS(rs, 2)]; T5 = Rm[WS(rs, 1)]; Ts = Ip[WS(rs, 2)]; T1f = TM - TN; TO = TM + TN; TL = T4 - T5; T6 = T4 + T5; Tt = Im[WS(rs, 1)]; } { E Tw, Ta, TC, Tz, Td, TD; { E Tx, Ty, Tb, Tc; { E T8, T1g, Tu, T9; T8 = Rp[WS(rs, 1)]; T13 = TO - TL; TP = TL + TO; T1p = T3 - T6; T7 = T3 + T6; T1g = Ts - Tt; Tu = Ts + Tt; T9 = Rm[WS(rs, 2)]; Tx = Ip[WS(rs, 1)]; T1h = T1f + T1g; T1v = T1f - T1g; TZ = Tr + Tu; Tv = Tr - Tu; Tw = T8 - T9; Ta = T8 + T9; Ty = Im[WS(rs, 2)]; } Tb = Rm[0]; Tc = Rp[WS(rs, 3)]; TC = Ip[WS(rs, 3)]; T1i = Tx - Ty; Tz = Tx + Ty; TB = Tb - Tc; Td = Tb + Tc; TD = Im[0]; } TA = Tw - Tz; TQ = Tw + Tz; Te = Ta + Td; T1w = Ta - Td; TE = TC + TD; T1j = TC - TD; } } { E T1x, T1k, T1r, TG, TS, T19, T15, T17, T11, T16, T12; { E T1B, T1z, T10, T1A, T1C; T1x = T1v - T1w; T1B = T1w + T1v; Rp[0] = T7 + Te; { E T1q, TR, TF, T14; T1k = T1i + T1j; T1q = T1j - T1i; TR = TB + TE; TF = TB - TE; T1r = T1p - T1q; T1z = T1p + T1q; Rm[0] = T1h + T1k; TG = TA + TF; T14 = TA - TF; TS = TQ - TR; T10 = TQ + TR; T1A = Tk * T1z; T19 = FNMS(KP707106781, T14, T13); T15 = FMA(KP707106781, T14, T13); T1C = Tk * T1B; } T17 = FMA(KP707106781, T10, TZ); T11 = FNMS(KP707106781, T10, TZ); Rp[WS(rs, 1)] = FNMS(To, T1B, T1A); T16 = Tg * T15; Rm[WS(rs, 1)] = FMA(To, T1z, T1C); } T12 = Tg * T11; { E T1l, T1a, T1c, T18; Im[WS(rs, 1)] = FMA(Tj, T11, T16); Ip[WS(rs, 1)] = FNMS(Tj, T15, T12); T18 = Tl * T17; T1l = T1h - T1k; T1a = Tl * T19; T1c = T7 - Te; Ip[WS(rs, 3)] = FNMS(Tp, T19, T18); { E T1s, T1m, T1d, T1y, TV; Im[WS(rs, 3)] = FMA(Tp, T17, T1a); T1m = T1e * T1c; T1d = T1b * T1c; T1s = T1o * T1r; Rm[WS(rs, 2)] = FMA(T1b, T1l, T1m); Rp[WS(rs, 2)] = FNMS(T1e, T1l, T1d); Rp[WS(rs, 3)] = FNMS(T1u, T1x, T1s); T1y = T1o * T1x; TV = FMA(KP707106781, TG, Tv); TH = FNMS(KP707106781, TG, Tv); TT = FNMS(KP707106781, TS, TP); TX = FMA(KP707106781, TS, TP); Rm[WS(rs, 3)] = FMA(T1u, T1r, T1y); TW = Tf * TV; TY = Ti * TV; } } } } } Ip[0] = FNMS(Ti, TX, TW); Im[0] = FMA(Tf, TX, TY); TU = TK * TH; TI = Tq * TH; Im[WS(rs, 2)] = FMA(Tq, TT, TU); Ip[WS(rs, 2)] = FNMS(TK, TT, TI); } } } static const tw_instr twinstr[] = { {TW_CEXP, 1, 1}, {TW_CEXP, 1, 3}, {TW_CEXP, 1, 7}, {TW_NEXT, 1, 0} }; static const hc2c_desc desc = { 8, "hc2cb2_8", twinstr, &GENUS, {44, 20, 30, 0} }; void X(codelet_hc2cb2_8) (planner *p) { X(khc2c_register) (p, hc2cb2_8, &desc, HC2C_VIA_RDFT); } #else /* HAVE_FMA */ /* Generated by: ../../../genfft/gen_hc2c.native -compact -variables 4 -pipeline-latency 4 -sign 1 -twiddle-log3 -precompute-twiddles -n 8 -dif -name hc2cb2_8 -include hc2cb.h */ /* * This function contains 74 FP additions, 44 FP multiplications, * (or, 56 additions, 26 multiplications, 18 fused multiply/add), * 46 stack variables, 1 constants, and 32 memory accesses */ #include "hc2cb.h" static void hc2cb2_8(R *Rp, R *Ip, R *Rm, R *Im, const R *W, stride rs, INT mb, INT me, INT ms) { DK(KP707106781, +0.707106781186547524400844362104849039284835938); { INT m; for (m = mb, W = W + ((mb - 1) * 6); m < me; m = m + 1, Rp = Rp + ms, Ip = Ip + ms, Rm = Rm - ms, Im = Im - ms, W = W + 6, MAKE_VOLATILE_STRIDE(32, rs)) { E Tf, Ti, Tg, Tj, Tl, Tp, TP, TR, TF, TG, TH, T15, TL, TT; { E Th, To, Tk, Tn; Tf = W[0]; Ti = W[1]; Tg = W[2]; Tj = W[3]; Th = Tf * Tg; To = Ti * Tg; Tk = Ti * Tj; Tn = Tf * Tj; Tl = Th - Tk; Tp = Tn + To; TP = Th + Tk; TR = Tn - To; TF = W[4]; TG = W[5]; TH = FMA(Tf, TF, Ti * TG); T15 = FNMS(TR, TF, TP * TG); TL = FNMS(Ti, TF, Tf * TG); TT = FMA(TP, TF, TR * TG); } { E T7, T1f, T1i, Tw, TI, TW, T18, TM, Te, T19, T1a, TD, TJ, TZ, T12; E TN, Tm, TE; { E T3, TU, Ts, T17, T6, T16, Tv, TV; { E T1, T2, Tq, Tr; T1 = Rp[0]; T2 = Rm[WS(rs, 3)]; T3 = T1 + T2; TU = T1 - T2; Tq = Ip[0]; Tr = Im[WS(rs, 3)]; Ts = Tq - Tr; T17 = Tq + Tr; } { E T4, T5, Tt, Tu; T4 = Rp[WS(rs, 2)]; T5 = Rm[WS(rs, 1)]; T6 = T4 + T5; T16 = T4 - T5; Tt = Ip[WS(rs, 2)]; Tu = Im[WS(rs, 1)]; Tv = Tt - Tu; TV = Tt + Tu; } T7 = T3 + T6; T1f = TU + TV; T1i = T17 - T16; Tw = Ts + Tv; TI = T3 - T6; TW = TU - TV; T18 = T16 + T17; TM = Ts - Tv; } { E Ta, TX, Tz, TY, Td, T10, TC, T11; { E T8, T9, Tx, Ty; T8 = Rp[WS(rs, 1)]; T9 = Rm[WS(rs, 2)]; Ta = T8 + T9; TX = T8 - T9; Tx = Ip[WS(rs, 1)]; Ty = Im[WS(rs, 2)]; Tz = Tx - Ty; TY = Tx + Ty; } { E Tb, Tc, TA, TB; Tb = Rm[0]; Tc = Rp[WS(rs, 3)]; Td = Tb + Tc; T10 = Tb - Tc; TA = Ip[WS(rs, 3)]; TB = Im[0]; TC = TA - TB; T11 = TA + TB; } Te = Ta + Td; T19 = TX + TY; T1a = T10 + T11; TD = Tz + TC; TJ = TC - Tz; TZ = TX - TY; T12 = T10 - T11; TN = Ta - Td; } Rp[0] = T7 + Te; Rm[0] = Tw + TD; Tm = T7 - Te; TE = Tw - TD; Rp[WS(rs, 2)] = FNMS(Tp, TE, Tl * Tm); Rm[WS(rs, 2)] = FMA(Tp, Tm, Tl * TE); { E TQ, TS, TK, TO; TQ = TI + TJ; TS = TN + TM; Rp[WS(rs, 1)] = FNMS(TR, TS, TP * TQ); Rm[WS(rs, 1)] = FMA(TP, TS, TR * TQ); TK = TI - TJ; TO = TM - TN; Rp[WS(rs, 3)] = FNMS(TL, TO, TH * TK); Rm[WS(rs, 3)] = FMA(TH, TO, TL * TK); } { E T1h, T1l, T1k, T1m, T1g, T1j; T1g = KP707106781 * (T19 + T1a); T1h = T1f - T1g; T1l = T1f + T1g; T1j = KP707106781 * (TZ - T12); T1k = T1i + T1j; T1m = T1i - T1j; Ip[WS(rs, 1)] = FNMS(Tj, T1k, Tg * T1h); Im[WS(rs, 1)] = FMA(Tg, T1k, Tj * T1h); Ip[WS(rs, 3)] = FNMS(TG, T1m, TF * T1l); Im[WS(rs, 3)] = FMA(TF, T1m, TG * T1l); } { E T14, T1d, T1c, T1e, T13, T1b; T13 = KP707106781 * (TZ + T12); T14 = TW - T13; T1d = TW + T13; T1b = KP707106781 * (T19 - T1a); T1c = T18 - T1b; T1e = T18 + T1b; Ip[WS(rs, 2)] = FNMS(T15, T1c, TT * T14); Im[WS(rs, 2)] = FMA(T15, T14, TT * T1c); Ip[0] = FNMS(Ti, T1e, Tf * T1d); Im[0] = FMA(Ti, T1d, Tf * T1e); } } } } } static const tw_instr twinstr[] = { {TW_CEXP, 1, 1}, {TW_CEXP, 1, 3}, {TW_CEXP, 1, 7}, {TW_NEXT, 1, 0} }; static const hc2c_desc desc = { 8, "hc2cb2_8", twinstr, &GENUS, {56, 26, 18, 0} }; void X(codelet_hc2cb2_8) (planner *p) { X(khc2c_register) (p, hc2cb2_8, &desc, HC2C_VIA_RDFT); } #endif /* HAVE_FMA */ fftw-3.3.3/rdft/scalar/r2cb/hc2cb2_16.c0000644000175000017500000005462012054411026014162 00000000000000/* * Copyright (c) 2003, 2007-11 Matteo Frigo * Copyright (c) 2003, 2007-11 Massachusetts Institute of Technology * * This program is free software; you can redistribute it and/or modify * it under the terms of the GNU General Public License as published by * the Free Software Foundation; either version 2 of the License, or * (at your option) any later version. * * This program is distributed in the hope that it will be useful, * but WITHOUT ANY WARRANTY; without even the implied warranty of * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the * GNU General Public License for more details. * * You should have received a copy of the GNU General Public License * along with this program; if not, write to the Free Software * Foundation, Inc., 51 Franklin Street, Fifth Floor, Boston, MA 02110-1301 USA * */ /* This file was automatically generated --- DO NOT EDIT */ /* Generated on Sun Nov 25 07:41:57 EST 2012 */ #include "codelet-rdft.h" #ifdef HAVE_FMA /* Generated by: ../../../genfft/gen_hc2c.native -fma -reorder-insns -schedule-for-pipeline -compact -variables 4 -pipeline-latency 4 -sign 1 -twiddle-log3 -precompute-twiddles -n 16 -dif -name hc2cb2_16 -include hc2cb.h */ /* * This function contains 196 FP additions, 134 FP multiplications, * (or, 104 additions, 42 multiplications, 92 fused multiply/add), * 112 stack variables, 3 constants, and 64 memory accesses */ #include "hc2cb.h" static void hc2cb2_16(R *Rp, R *Ip, R *Rm, R *Im, const R *W, stride rs, INT mb, INT me, INT ms) { DK(KP923879532, +0.923879532511286756128183189396788286822416626); DK(KP707106781, +0.707106781186547524400844362104849039284835938); DK(KP414213562, +0.414213562373095048801688724209698078569671875); { INT m; for (m = mb, W = W + ((mb - 1) * 8); m < me; m = m + 1, Rp = Rp + ms, Ip = Ip + ms, Rm = Rm - ms, Im = Im - ms, W = W + 8, MAKE_VOLATILE_STRIDE(64, rs)) { E Tv, TB, TF, Ty, T1J, T1O, T1N, T1K; { E Tw, T2z, T2C, Tx, T3f, T3l, T2F, T3r, Tz; Tv = W[0]; Tw = W[2]; T2z = W[6]; T2C = W[7]; TB = W[4]; Tx = Tv * Tw; T3f = Tv * T2z; T3l = Tv * T2C; T2F = Tv * TB; T3r = Tw * TB; TF = W[5]; Ty = W[1]; Tz = W[3]; { E T2G, T3z, T3m, T3g, T3L, T3s, T1V, TA, T3w, T3Q, T30, T3C, TE, T1X, T1D; E TG, T1G, T1o, T2p, T1Y, T2u, T2c, T1Z, TL, T1t, T2d, T3n, T35, T3R, T3F; E T1w, T20, T3M, Tf, T3h, T2L, T2e, TW, T2Q, T36, T3I, T3N, T2V, T37, T1d; E Tu, T3S, T18, T1z, T1i, T24, T2g, T27, T2h; { E T2K, TQ, TV, T2H; { E TH, T3, T32, T1s, T1p, T6, T33, TK, TM, Ta, TS, T2J, TP, TR, Td; E TT, TI, TJ; { E T1q, T1r, T4, T5; { E T1, T1n, TC, T2b, T1W, T2, T3v, T2Z, TD; T1 = Rp[0]; T3v = Tw * TF; T2Z = Tv * TF; T2G = FNMS(Ty, TF, T2F); T3z = FMA(Ty, TF, T2F); T3m = FNMS(Ty, T2z, T3l); T3g = FMA(Ty, T2C, T3f); T3L = FNMS(Tz, TF, T3r); T3s = FMA(Tz, TF, T3r); T1V = FMA(Ty, Tz, Tx); TA = FNMS(Ty, Tz, Tx); TD = Tv * Tz; T3w = FNMS(Tz, TB, T3v); T3Q = FMA(Tz, TB, T3v); T30 = FMA(Ty, TB, T2Z); T3C = FNMS(Ty, TB, T2Z); T1n = TA * TF; TC = TA * TB; T2b = T1V * TF; T1W = T1V * TB; TE = FMA(Ty, Tw, TD); T1X = FNMS(Ty, Tw, TD); T2 = Rm[WS(rs, 7)]; T1q = Ip[0]; T1D = FMA(TE, TF, TC); TG = FNMS(TE, TF, TC); T1G = FNMS(TE, TB, T1n); T1o = FMA(TE, TB, T1n); T2p = FMA(T1X, TF, T1W); T1Y = FNMS(T1X, TF, T1W); T2u = FNMS(T1X, TB, T2b); T2c = FMA(T1X, TB, T2b); TH = T1 - T2; T3 = T1 + T2; T1r = Im[WS(rs, 7)]; } T4 = Rp[WS(rs, 4)]; T5 = Rm[WS(rs, 3)]; TI = Ip[WS(rs, 4)]; T32 = T1q - T1r; T1s = T1q + T1r; T1p = T4 - T5; T6 = T4 + T5; TJ = Im[WS(rs, 3)]; } { E TN, TO, T8, T9, Tb, Tc; T8 = Rp[WS(rs, 2)]; T9 = Rm[WS(rs, 5)]; TN = Ip[WS(rs, 2)]; T33 = TI - TJ; TK = TI + TJ; TM = T8 - T9; Ta = T8 + T9; TO = Im[WS(rs, 5)]; Tb = Rm[WS(rs, 1)]; Tc = Rp[WS(rs, 6)]; TS = Ip[WS(rs, 6)]; T2J = TN - TO; TP = TN + TO; TR = Tb - Tc; Td = Tb + Tc; TT = Im[WS(rs, 1)]; } { E T2I, TU, Te, T31, T34, T3D; T1Z = TH + TK; TL = TH - TK; T1t = T1p + T1s; T2d = T1s - T1p; T2I = TS - TT; TU = TS + TT; Te = Ta + Td; T31 = Ta - Td; T34 = T32 - T33; T3D = T32 + T33; { E T1u, T1v, T3E, T7; T3E = T2J + T2I; T2K = T2I - T2J; TQ = TM - TP; T1u = TM + TP; T3n = T34 - T31; T35 = T31 + T34; TV = TR - TU; T1v = TR + TU; T3R = T3D - T3E; T3F = T3D + T3E; T2H = T3 - T6; T7 = T3 + T6; T1w = T1u - T1v; T20 = T1u + T1v; T3M = T7 - Te; Tf = T7 + Te; } } } { E T1e, Ti, T2N, T1c, T19, Tl, T2O, T1h, Tq, T13, Tp, T2S, T11, Tr, T14; E T15; { E Tj, Tk, T1f, T1g; { E Tg, Th, T1a, T1b; Tg = Rp[WS(rs, 1)]; T3h = T2H - T2K; T2L = T2H + T2K; T2e = TQ - TV; TW = TQ + TV; Th = Rm[WS(rs, 6)]; T1a = Ip[WS(rs, 1)]; T1b = Im[WS(rs, 6)]; Tj = Rp[WS(rs, 5)]; T1e = Tg - Th; Ti = Tg + Th; T2N = T1a - T1b; T1c = T1a + T1b; Tk = Rm[WS(rs, 2)]; T1f = Ip[WS(rs, 5)]; T1g = Im[WS(rs, 2)]; } { E Tn, To, TZ, T10; Tn = Rm[0]; T19 = Tj - Tk; Tl = Tj + Tk; T2O = T1f - T1g; T1h = T1f + T1g; To = Rp[WS(rs, 7)]; TZ = Ip[WS(rs, 7)]; T10 = Im[0]; Tq = Rp[WS(rs, 3)]; T13 = Tn - To; Tp = Tn + To; T2S = TZ - T10; T11 = TZ + T10; Tr = Rm[WS(rs, 4)]; T14 = Ip[WS(rs, 3)]; T15 = Im[WS(rs, 4)]; } } { E TY, T16, Tm, Tt; { E T2P, T3G, Ts, T2M, T3H, T2U, T2T, T2R; T2P = T2N - T2O; T3G = T2N + T2O; TY = Tq - Tr; Ts = Tq + Tr; T2T = T14 - T15; T16 = T14 + T15; T2M = Ti - Tl; Tm = Ti + Tl; T3H = T2S + T2T; T2U = T2S - T2T; Tt = Tp + Ts; T2R = Tp - Ts; T2Q = T2M - T2P; T36 = T2M + T2P; T3I = T3G + T3H; T3N = T3H - T3G; T2V = T2R + T2U; T37 = T2U - T2R; } { E T25, T26, T22, T23, T12, T17; T12 = TY - T11; T25 = TY + T11; T26 = T13 + T16; T17 = T13 - T16; T22 = T1c - T19; T1d = T19 + T1c; Tu = Tm + Tt; T3S = Tm - Tt; T18 = FNMS(KP414213562, T17, T12); T1z = FMA(KP414213562, T12, T17); T1i = T1e - T1h; T23 = T1e + T1h; T24 = FNMS(KP414213562, T23, T22); T2g = FMA(KP414213562, T22, T23); T27 = FNMS(KP414213562, T26, T25); T2h = FMA(KP414213562, T25, T26); } } } } { E T1j, T1y, T3V, T3X, T3W, T38, T3i, T3o, T2W, T3K, T3B, T3A; Rp[0] = Tf + Tu; T3A = Tf - Tu; T1j = FMA(KP414213562, T1i, T1d); T1y = FNMS(KP414213562, T1d, T1i); T3K = T3C * T3A; T3B = T3z * T3A; { E T3O, T3T, T3J, T3P, T3U; T3O = T3M - T3N; T3V = T3M + T3N; T3X = T3S + T3R; T3T = T3R - T3S; Rm[0] = T3F + T3I; T3J = T3F - T3I; T3P = T3L * T3O; T3U = T3L * T3T; T3W = TA * T3V; Rp[WS(rs, 4)] = FNMS(T3C, T3J, T3B); Rm[WS(rs, 4)] = FMA(T3z, T3J, T3K); Rp[WS(rs, 6)] = FNMS(T3Q, T3T, T3P); Rm[WS(rs, 6)] = FMA(T3Q, T3O, T3U); T38 = T36 + T37; T3i = T37 - T36; T3o = T2Q - T2V; T2W = T2Q + T2V; } { E T2q, T21, T28, T2w, T2v, T2f, T2i, T2r; { E T2Y, T3a, T3c, T3d, T39, T3e, T3b, T2X, T3Y; Rp[WS(rs, 2)] = FNMS(TE, T3X, T3W); T3Y = TA * T3X; { E T3t, T3j, T3x, T3p; T3t = FMA(KP707106781, T3i, T3h); T3j = FNMS(KP707106781, T3i, T3h); T3x = FMA(KP707106781, T3o, T3n); T3p = FNMS(KP707106781, T3o, T3n); Rm[WS(rs, 2)] = FMA(TE, T3V, T3Y); { E T3u, T3k, T3y, T3q; T3u = T3s * T3t; T3k = T3g * T3j; T3y = T3s * T3x; T3q = T3g * T3p; Rp[WS(rs, 3)] = FNMS(T3w, T3x, T3u); Rp[WS(rs, 7)] = FNMS(T3m, T3p, T3k); Rm[WS(rs, 3)] = FMA(T3w, T3t, T3y); Rm[WS(rs, 7)] = FMA(T3m, T3j, T3q); T3b = FMA(KP707106781, T2W, T2L); T2X = FNMS(KP707106781, T2W, T2L); } } T2Y = T2G * T2X; T3a = T30 * T2X; T3c = T1V * T3b; T3d = FMA(KP707106781, T38, T35); T39 = FNMS(KP707106781, T38, T35); T3e = T1X * T3b; T2q = FMA(KP707106781, T20, T1Z); T21 = FNMS(KP707106781, T20, T1Z); Rp[WS(rs, 1)] = FNMS(T1X, T3d, T3c); Rm[WS(rs, 5)] = FMA(T2G, T39, T3a); Rp[WS(rs, 5)] = FNMS(T30, T39, T2Y); Rm[WS(rs, 1)] = FMA(T1V, T3d, T3e); T28 = T24 + T27; T2w = T27 - T24; T2v = FNMS(KP707106781, T2e, T2d); T2f = FMA(KP707106781, T2e, T2d); T2i = T2g - T2h; T2r = T2g + T2h; } { E TX, T1k, T1x, T1A; T1J = FMA(KP707106781, TW, TL); TX = FNMS(KP707106781, TW, TL); { E T2l, T29, T2n, T2j; T2l = FNMS(KP923879532, T28, T21); T29 = FMA(KP923879532, T28, T21); T2n = FMA(KP923879532, T2i, T2f); T2j = FNMS(KP923879532, T2i, T2f); { E T2o, T2m, T2k, T2a; T2o = Tz * T2l; T2m = Tw * T2l; T2k = T2c * T29; T2a = T1Y * T29; Im[WS(rs, 1)] = FMA(Tw, T2n, T2o); Ip[WS(rs, 1)] = FNMS(Tz, T2n, T2m); Im[WS(rs, 5)] = FMA(T1Y, T2j, T2k); Ip[WS(rs, 5)] = FNMS(T2c, T2j, T2a); T1k = T18 - T1j; T1O = T1j + T18; } } T1N = FMA(KP707106781, T1w, T1t); T1x = FNMS(KP707106781, T1w, T1t); T1A = T1y - T1z; T1K = T1y + T1z; { E T1E, T1l, T1H, T1B; T1E = FMA(KP923879532, T1k, TX); T1l = FNMS(KP923879532, T1k, TX); T1H = FMA(KP923879532, T1A, T1x); T1B = FNMS(KP923879532, T1A, T1x); { E T1I, T1F, T1C, T1m; T1I = T1G * T1E; T1F = T1D * T1E; T1C = T1o * T1l; T1m = TG * T1l; Im[WS(rs, 2)] = FMA(T1D, T1H, T1I); Ip[WS(rs, 2)] = FNMS(T1G, T1H, T1F); Im[WS(rs, 6)] = FMA(TG, T1B, T1C); Ip[WS(rs, 6)] = FNMS(T1o, T1B, T1m); } } { E T2A, T2s, T2D, T2x; T2A = FMA(KP923879532, T2r, T2q); T2s = FNMS(KP923879532, T2r, T2q); T2D = FNMS(KP923879532, T2w, T2v); T2x = FMA(KP923879532, T2w, T2v); { E T2B, T2t, T2E, T2y; T2B = T2z * T2A; T2t = T2p * T2s; T2E = T2z * T2D; T2y = T2p * T2x; Ip[WS(rs, 7)] = FNMS(T2C, T2D, T2B); Ip[WS(rs, 3)] = FNMS(T2u, T2x, T2t); Im[WS(rs, 7)] = FMA(T2C, T2A, T2E); Im[WS(rs, 3)] = FMA(T2u, T2s, T2y); } } } } } } } { E T1L, T1R, T1P, T1T; T1L = FNMS(KP923879532, T1K, T1J); T1R = FMA(KP923879532, T1K, T1J); T1P = FNMS(KP923879532, T1O, T1N); T1T = FMA(KP923879532, T1O, T1N); { E T1S, T1M, T1U, T1Q; T1S = Tv * T1R; T1M = TB * T1L; T1U = Tv * T1T; T1Q = TB * T1P; Ip[0] = FNMS(Ty, T1T, T1S); Ip[WS(rs, 4)] = FNMS(TF, T1P, T1M); Im[0] = FMA(Ty, T1R, T1U); Im[WS(rs, 4)] = FMA(TF, T1L, T1Q); } } } } } static const tw_instr twinstr[] = { {TW_CEXP, 1, 1}, {TW_CEXP, 1, 3}, {TW_CEXP, 1, 9}, {TW_CEXP, 1, 15}, {TW_NEXT, 1, 0} }; static const hc2c_desc desc = { 16, "hc2cb2_16", twinstr, &GENUS, {104, 42, 92, 0} }; void X(codelet_hc2cb2_16) (planner *p) { X(khc2c_register) (p, hc2cb2_16, &desc, HC2C_VIA_RDFT); } #else /* HAVE_FMA */ /* Generated by: ../../../genfft/gen_hc2c.native -compact -variables 4 -pipeline-latency 4 -sign 1 -twiddle-log3 -precompute-twiddles -n 16 -dif -name hc2cb2_16 -include hc2cb.h */ /* * This function contains 196 FP additions, 108 FP multiplications, * (or, 156 additions, 68 multiplications, 40 fused multiply/add), * 80 stack variables, 3 constants, and 64 memory accesses */ #include "hc2cb.h" static void hc2cb2_16(R *Rp, R *Ip, R *Rm, R *Im, const R *W, stride rs, INT mb, INT me, INT ms) { DK(KP382683432, +0.382683432365089771728459984030398866761344562); DK(KP923879532, +0.923879532511286756128183189396788286822416626); DK(KP707106781, +0.707106781186547524400844362104849039284835938); { INT m; for (m = mb, W = W + ((mb - 1) * 8); m < me; m = m + 1, Rp = Rp + ms, Ip = Ip + ms, Rm = Rm - ms, Im = Im - ms, W = W + 8, MAKE_VOLATILE_STRIDE(64, rs)) { E Tv, Ty, T1l, T1n, T1p, T1t, T27, T25, Tz, Tw, TB, T21, T1P, T1H, T1X; E T17, T1L, T1N, T1v, T1w, T1x, T1B, T2F, T2T, T2b, T2R, T3j, T3x, T35, T3t; { E TA, T1J, T15, T1G, Tx, T1K, T16, T1F; { E T1m, T1s, T1o, T1r; Tv = W[0]; Ty = W[1]; T1l = W[2]; T1n = W[3]; T1m = Tv * T1l; T1s = Ty * T1l; T1o = Ty * T1n; T1r = Tv * T1n; T1p = T1m + T1o; T1t = T1r - T1s; T27 = T1r + T1s; T25 = T1m - T1o; Tz = W[5]; TA = Ty * Tz; T1J = T1l * Tz; T15 = Tv * Tz; T1G = T1n * Tz; Tw = W[4]; Tx = Tv * Tw; T1K = T1n * Tw; T16 = Ty * Tw; T1F = T1l * Tw; } TB = Tx - TA; T21 = T1J + T1K; T1P = T15 - T16; T1H = T1F + T1G; T1X = T1F - T1G; T17 = T15 + T16; T1L = T1J - T1K; T1N = Tx + TA; T1v = W[6]; T1w = W[7]; T1x = FMA(Tv, T1v, Ty * T1w); T1B = FNMS(Ty, T1v, Tv * T1w); { E T2D, T2E, T29, T2a; T2D = T25 * Tz; T2E = T27 * Tw; T2F = T2D + T2E; T2T = T2D - T2E; T29 = T25 * Tw; T2a = T27 * Tz; T2b = T29 - T2a; T2R = T29 + T2a; } { E T3h, T3i, T33, T34; T3h = T1p * Tz; T3i = T1t * Tw; T3j = T3h + T3i; T3x = T3h - T3i; T33 = T1p * Tw; T34 = T1t * Tz; T35 = T33 - T34; T3t = T33 + T34; } } { E T7, T36, T3k, TC, T1f, T2e, T2I, T1Q, Te, TJ, T1R, T18, T2L, T37, T2l; E T3l, Tm, T1T, TT, T1h, T2A, T2N, T3b, T3n, Tt, T1U, T12, T1i, T2t, T2O; E T3e, T3o; { E T3, T2c, T1b, T2H, T6, T2G, T1e, T2d; { E T1, T2, T19, T1a; T1 = Rp[0]; T2 = Rm[WS(rs, 7)]; T3 = T1 + T2; T2c = T1 - T2; T19 = Ip[0]; T1a = Im[WS(rs, 7)]; T1b = T19 - T1a; T2H = T19 + T1a; } { E T4, T5, T1c, T1d; T4 = Rp[WS(rs, 4)]; T5 = Rm[WS(rs, 3)]; T6 = T4 + T5; T2G = T4 - T5; T1c = Ip[WS(rs, 4)]; T1d = Im[WS(rs, 3)]; T1e = T1c - T1d; T2d = T1c + T1d; } T7 = T3 + T6; T36 = T2c + T2d; T3k = T2H - T2G; TC = T3 - T6; T1f = T1b - T1e; T2e = T2c - T2d; T2I = T2G + T2H; T1Q = T1b + T1e; } { E Ta, T2f, TI, T2g, Td, T2i, TF, T2j; { E T8, T9, TG, TH; T8 = Rp[WS(rs, 2)]; T9 = Rm[WS(rs, 5)]; Ta = T8 + T9; T2f = T8 - T9; TG = Ip[WS(rs, 2)]; TH = Im[WS(rs, 5)]; TI = TG - TH; T2g = TG + TH; } { E Tb, Tc, TD, TE; Tb = Rm[WS(rs, 1)]; Tc = Rp[WS(rs, 6)]; Td = Tb + Tc; T2i = Tb - Tc; TD = Ip[WS(rs, 6)]; TE = Im[WS(rs, 1)]; TF = TD - TE; T2j = TD + TE; } Te = Ta + Td; TJ = TF - TI; T1R = TI + TF; T18 = Ta - Td; { E T2J, T2K, T2h, T2k; T2J = T2f + T2g; T2K = T2i + T2j; T2L = KP707106781 * (T2J - T2K); T37 = KP707106781 * (T2J + T2K); T2h = T2f - T2g; T2k = T2i - T2j; T2l = KP707106781 * (T2h + T2k); T3l = KP707106781 * (T2h - T2k); } } { E Ti, T2x, TO, T2v, Tl, T2u, TR, T2y, TL, TS; { E Tg, Th, TM, TN; Tg = Rp[WS(rs, 1)]; Th = Rm[WS(rs, 6)]; Ti = Tg + Th; T2x = Tg - Th; TM = Ip[WS(rs, 1)]; TN = Im[WS(rs, 6)]; TO = TM - TN; T2v = TM + TN; } { E Tj, Tk, TP, TQ; Tj = Rp[WS(rs, 5)]; Tk = Rm[WS(rs, 2)]; Tl = Tj + Tk; T2u = Tj - Tk; TP = Ip[WS(rs, 5)]; TQ = Im[WS(rs, 2)]; TR = TP - TQ; T2y = TP + TQ; } Tm = Ti + Tl; T1T = TO + TR; TL = Ti - Tl; TS = TO - TR; TT = TL - TS; T1h = TL + TS; { E T2w, T2z, T39, T3a; T2w = T2u + T2v; T2z = T2x - T2y; T2A = FMA(KP923879532, T2w, KP382683432 * T2z); T2N = FNMS(KP382683432, T2w, KP923879532 * T2z); T39 = T2x + T2y; T3a = T2v - T2u; T3b = FNMS(KP923879532, T3a, KP382683432 * T39); T3n = FMA(KP382683432, T3a, KP923879532 * T39); } } { E Tp, T2q, TX, T2o, Ts, T2n, T10, T2r, TU, T11; { E Tn, To, TV, TW; Tn = Rm[0]; To = Rp[WS(rs, 7)]; Tp = Tn + To; T2q = Tn - To; TV = Ip[WS(rs, 7)]; TW = Im[0]; TX = TV - TW; T2o = TV + TW; } { E Tq, Tr, TY, TZ; Tq = Rp[WS(rs, 3)]; Tr = Rm[WS(rs, 4)]; Ts = Tq + Tr; T2n = Tq - Tr; TY = Ip[WS(rs, 3)]; TZ = Im[WS(rs, 4)]; T10 = TY - TZ; T2r = TY + TZ; } Tt = Tp + Ts; T1U = TX + T10; TU = Tp - Ts; T11 = TX - T10; T12 = TU + T11; T1i = T11 - TU; { E T2p, T2s, T3c, T3d; T2p = T2n - T2o; T2s = T2q - T2r; T2t = FNMS(KP382683432, T2s, KP923879532 * T2p); T2O = FMA(KP382683432, T2p, KP923879532 * T2s); T3c = T2q + T2r; T3d = T2n + T2o; T3e = FNMS(KP923879532, T3d, KP382683432 * T3c); T3o = FMA(KP382683432, T3d, KP923879532 * T3c); } } { E Tf, Tu, T1O, T1S, T1V, T1W; Tf = T7 + Te; Tu = Tm + Tt; T1O = Tf - Tu; T1S = T1Q + T1R; T1V = T1T + T1U; T1W = T1S - T1V; Rp[0] = Tf + Tu; Rm[0] = T1S + T1V; Rp[WS(rs, 4)] = FNMS(T1P, T1W, T1N * T1O); Rm[WS(rs, 4)] = FMA(T1P, T1O, T1N * T1W); } { E T3g, T3r, T3q, T3s; { E T38, T3f, T3m, T3p; T38 = T36 - T37; T3f = T3b + T3e; T3g = T38 - T3f; T3r = T38 + T3f; T3m = T3k + T3l; T3p = T3n - T3o; T3q = T3m - T3p; T3s = T3m + T3p; } Ip[WS(rs, 5)] = FNMS(T3j, T3q, T35 * T3g); Im[WS(rs, 5)] = FMA(T3j, T3g, T35 * T3q); Ip[WS(rs, 1)] = FNMS(T1n, T3s, T1l * T3r); Im[WS(rs, 1)] = FMA(T1n, T3r, T1l * T3s); } { E T3w, T3B, T3A, T3C; { E T3u, T3v, T3y, T3z; T3u = T36 + T37; T3v = T3n + T3o; T3w = T3u - T3v; T3B = T3u + T3v; T3y = T3k - T3l; T3z = T3b - T3e; T3A = T3y + T3z; T3C = T3y - T3z; } Ip[WS(rs, 3)] = FNMS(T3x, T3A, T3t * T3w); Im[WS(rs, 3)] = FMA(T3t, T3A, T3x * T3w); Ip[WS(rs, 7)] = FNMS(T1w, T3C, T1v * T3B); Im[WS(rs, 7)] = FMA(T1v, T3C, T1w * T3B); } { E T14, T1q, T1k, T1u; { E TK, T13, T1g, T1j; TK = TC + TJ; T13 = KP707106781 * (TT + T12); T14 = TK - T13; T1q = TK + T13; T1g = T18 + T1f; T1j = KP707106781 * (T1h + T1i); T1k = T1g - T1j; T1u = T1g + T1j; } Rp[WS(rs, 5)] = FNMS(T17, T1k, TB * T14); Rm[WS(rs, 5)] = FMA(T17, T14, TB * T1k); Rp[WS(rs, 1)] = FNMS(T1t, T1u, T1p * T1q); Rm[WS(rs, 1)] = FMA(T1t, T1q, T1p * T1u); } { E T1A, T1I, T1E, T1M; { E T1y, T1z, T1C, T1D; T1y = TC - TJ; T1z = KP707106781 * (T1i - T1h); T1A = T1y - T1z; T1I = T1y + T1z; T1C = T1f - T18; T1D = KP707106781 * (TT - T12); T1E = T1C - T1D; T1M = T1C + T1D; } Rp[WS(rs, 7)] = FNMS(T1B, T1E, T1x * T1A); Rm[WS(rs, 7)] = FMA(T1x, T1E, T1B * T1A); Rp[WS(rs, 3)] = FNMS(T1L, T1M, T1H * T1I); Rm[WS(rs, 3)] = FMA(T1H, T1M, T1L * T1I); } { E T2C, T2S, T2Q, T2U; { E T2m, T2B, T2M, T2P; T2m = T2e - T2l; T2B = T2t - T2A; T2C = T2m - T2B; T2S = T2m + T2B; T2M = T2I - T2L; T2P = T2N - T2O; T2Q = T2M - T2P; T2U = T2M + T2P; } Ip[WS(rs, 6)] = FNMS(T2F, T2Q, T2b * T2C); Im[WS(rs, 6)] = FMA(T2F, T2C, T2b * T2Q); Ip[WS(rs, 2)] = FNMS(T2T, T2U, T2R * T2S); Im[WS(rs, 2)] = FMA(T2T, T2S, T2R * T2U); } { E T2X, T31, T30, T32; { E T2V, T2W, T2Y, T2Z; T2V = T2e + T2l; T2W = T2N + T2O; T2X = T2V - T2W; T31 = T2V + T2W; T2Y = T2I + T2L; T2Z = T2A + T2t; T30 = T2Y - T2Z; T32 = T2Y + T2Z; } Ip[WS(rs, 4)] = FNMS(Tz, T30, Tw * T2X); Im[WS(rs, 4)] = FMA(Tw, T30, Tz * T2X); Ip[0] = FNMS(Ty, T32, Tv * T31); Im[0] = FMA(Tv, T32, Ty * T31); } { E T20, T26, T24, T28; { E T1Y, T1Z, T22, T23; T1Y = T7 - Te; T1Z = T1U - T1T; T20 = T1Y - T1Z; T26 = T1Y + T1Z; T22 = T1Q - T1R; T23 = Tm - Tt; T24 = T22 - T23; T28 = T23 + T22; } Rp[WS(rs, 6)] = FNMS(T21, T24, T1X * T20); Rm[WS(rs, 6)] = FMA(T1X, T24, T21 * T20); Rp[WS(rs, 2)] = FNMS(T27, T28, T25 * T26); Rm[WS(rs, 2)] = FMA(T25, T28, T27 * T26); } } } } } static const tw_instr twinstr[] = { {TW_CEXP, 1, 1}, {TW_CEXP, 1, 3}, {TW_CEXP, 1, 9}, {TW_CEXP, 1, 15}, {TW_NEXT, 1, 0} }; static const hc2c_desc desc = { 16, "hc2cb2_16", twinstr, &GENUS, {156, 68, 40, 0} }; void X(codelet_hc2cb2_16) (planner *p) { X(khc2c_register) (p, hc2cb2_16, &desc, HC2C_VIA_RDFT); } #endif /* HAVE_FMA */ fftw-3.3.3/rdft/scalar/r2cb/hc2cb2_32.c0000644000175000017500000014715712054411036014171 00000000000000/* * Copyright (c) 2003, 2007-11 Matteo Frigo * Copyright (c) 2003, 2007-11 Massachusetts Institute of Technology * * This program is free software; you can redistribute it and/or modify * it under the terms of the GNU General Public License as published by * the Free Software Foundation; either version 2 of the License, or * (at your option) any later version. * * This program is distributed in the hope that it will be useful, * but WITHOUT ANY WARRANTY; without even the implied warranty of * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the * GNU General Public License for more details. * * You should have received a copy of the GNU General Public License * along with this program; if not, write to the Free Software * Foundation, Inc., 51 Franklin Street, Fifth Floor, Boston, MA 02110-1301 USA * */ /* This file was automatically generated --- DO NOT EDIT */ /* Generated on Sun Nov 25 07:41:58 EST 2012 */ #include "codelet-rdft.h" #ifdef HAVE_FMA /* Generated by: ../../../genfft/gen_hc2c.native -fma -reorder-insns -schedule-for-pipeline -compact -variables 4 -pipeline-latency 4 -sign 1 -twiddle-log3 -precompute-twiddles -n 32 -dif -name hc2cb2_32 -include hc2cb.h */ /* * This function contains 488 FP additions, 350 FP multiplications, * (or, 236 additions, 98 multiplications, 252 fused multiply/add), * 204 stack variables, 7 constants, and 128 memory accesses */ #include "hc2cb.h" static void hc2cb2_32(R *Rp, R *Ip, R *Rm, R *Im, const R *W, stride rs, INT mb, INT me, INT ms) { DK(KP980785280, +0.980785280403230449126182236134239036973933731); DK(KP198912367, +0.198912367379658006911597622644676228597850501); DK(KP831469612, +0.831469612302545237078788377617905756738560812); DK(KP668178637, +0.668178637919298919997757686523080761552472251); DK(KP923879532, +0.923879532511286756128183189396788286822416626); DK(KP414213562, +0.414213562373095048801688724209698078569671875); DK(KP707106781, +0.707106781186547524400844362104849039284835938); { INT m; for (m = mb, W = W + ((mb - 1) * 8); m < me; m = m + 1, Rp = Rp + ms, Ip = Ip + ms, Rm = Rm - ms, Im = Im - ms, W = W + 8, MAKE_VOLATILE_STRIDE(128, rs)) { E T5u, T6b, T6e, T5I, T66, T60, T5U, T5R, T67, T5L, T61, T5x, T5A, T5D, T5O; E T62, T5V, T5P; { E T11, T14, T12, T37, T17, T1b, T39, T15, T7C, T8P, T8S, T7I, T98, T7e, T78; E T8V, T3d, T3x, T3a, T3v, T9s, T3G, T4p, T5X, T16, T9m, T3y, T4b, T3C, T4g; E T5Z, T1a, T4r, T3J, T2O, T1c, T4W, T4s, T3Y, T3K, T3l, T3e, T3i, T3q, T8K; E T8E, T8m, T7S, T5k, T5e; { E T13, T3c, T38, T3F, T7B, T9l, T77, T7d, T9r, T7H; T11 = W[2]; T14 = W[3]; T12 = W[4]; T37 = W[0]; T17 = W[6]; T1b = W[7]; T13 = T11 * T12; T3c = T37 * T14; T38 = T37 * T11; T3F = T37 * T12; T7B = T11 * T17; T9l = T12 * T17; T77 = T37 * T17; T7d = T37 * T1b; T9r = T12 * T1b; T7H = T11 * T1b; T39 = W[1]; T15 = W[5]; { E T3I, T19, T5d, T3b, T18, T2N; T7C = FMA(T14, T1b, T7B); T8P = FNMS(T14, T1b, T7B); T8S = FMA(T14, T17, T7H); T7I = FNMS(T14, T17, T7H); T98 = FNMS(T39, T17, T7d); T7e = FMA(T39, T17, T7d); T78 = FNMS(T39, T1b, T77); T8V = FMA(T39, T1b, T77); T3d = FMA(T39, T11, T3c); T3x = FNMS(T39, T11, T3c); T3a = FNMS(T39, T14, T38); T3v = FMA(T39, T14, T38); T9s = FNMS(T15, T17, T9r); T3G = FNMS(T39, T15, T3F); T4p = FMA(T39, T15, T3F); T5X = FNMS(T14, T15, T13); T16 = FMA(T14, T15, T13); T3I = T37 * T15; T19 = T11 * T15; T5d = T3v * T12; T3b = T3a * T12; T9m = FMA(T15, T1b, T9l); { E T3w, T3B, T5t, T5H; T3w = T3v * T17; T3B = T3v * T1b; T5t = T3a * T17; T5H = T3a * T1b; T3y = FNMS(T3x, T1b, T3w); T4b = FMA(T3x, T1b, T3w); T3C = FMA(T3x, T17, T3B); T4g = FNMS(T3x, T17, T3B); T5u = FMA(T3d, T1b, T5t); T6b = FNMS(T3d, T1b, T5t); T6e = FMA(T3d, T17, T5H); T5I = FNMS(T3d, T17, T5H); T18 = T16 * T17; T2N = T16 * T1b; T5Z = FMA(T14, T12, T19); T1a = FNMS(T14, T12, T19); } { E T3H, T3X, T4q, T4V, T5Y, T65; T4q = T4p * T17; T4V = T4p * T1b; T4r = FNMS(T39, T12, T3I); T3J = FMA(T39, T12, T3I); T2O = FNMS(T1a, T17, T2N); T1c = FMA(T1a, T1b, T18); T3H = T3G * T17; T4W = FNMS(T4r, T17, T4V); T4s = FMA(T4r, T1b, T4q); T3X = T3G * T1b; T5Y = T5X * T17; T65 = T5X * T1b; T3Y = FNMS(T3J, T17, T3X); T3K = FMA(T3J, T1b, T3H); { E T8J, T8D, T3h, T5j, T8l, T7R; T3h = T3a * T15; T66 = FNMS(T5Z, T17, T65); T60 = FMA(T5Z, T1b, T5Y); T3l = FNMS(T3d, T15, T3b); T3e = FMA(T3d, T15, T3b); T3i = FNMS(T3d, T12, T3h); T3q = FMA(T3d, T12, T3h); T8J = T3l * T1b; T8D = T3l * T17; T5j = T3v * T15; T8l = T3e * T1b; T7R = T3e * T17; T8K = FNMS(T3q, T17, T8J); T8E = FMA(T3q, T1b, T8D); T8m = FNMS(T3i, T17, T8l); T7S = FMA(T3i, T1b, T7R); T5U = FNMS(T3x, T12, T5j); T5k = FMA(T3x, T12, T5j); T5e = FNMS(T3x, T15, T5d); T5R = FMA(T3x, T15, T5d); } } } } { E T6O, T6i, T7s, T7o, T6j, Tf, T8W, T7V, T99, T8p, T3L, T1t, T3Z, T2X, T5J; E T4Z, T7t, T6W, T5v, T4v, TZ, T7x, T91, T9d, T28, T3S, T3R, T2h, T5B, T4Q; E T8v, T8a, T5C, T4N, T6Z, T6J, TK, T7w, T3P, T2z, T9c, T94, T3O, T2I, T5y; E T4J, T8u, T8h, T5z, T4G, T6Y, T6A, T6P, Tu, T9a, T82, T8X, T8s, T4y, T40; E T1Q, T3M, T30, T4B, T5w, T52, T7u, T6q; { E T6B, T6I, T4M, T4L, T4t, T4u, T6s, T6z; { E T1d, T3, T6Q, T2S, T2P, T6, T6R, T1g, Td, T6U, T1i, Ta, T2V, T1r, T6T; E T1l; { E T2Q, T2R, T4, T5, T1, T2, T1e, T1f; T1 = Rp[0]; T2 = Rm[WS(rs, 15)]; { E T6N, T6h, T7r, T7n; T6N = T5R * T1b; T6h = T5R * T17; T7r = T5e * T1b; T7n = T5e * T17; T6O = FNMS(T5U, T17, T6N); T6i = FMA(T5U, T1b, T6h); T7s = FNMS(T5k, T17, T7r); T7o = FMA(T5k, T1b, T7n); T1d = T1 - T2; T3 = T1 + T2; } T2Q = Ip[0]; T2R = Im[WS(rs, 15)]; T4 = Rp[WS(rs, 8)]; T5 = Rm[WS(rs, 7)]; T1e = Ip[WS(rs, 8)]; T6Q = T2Q - T2R; T2S = T2Q + T2R; T2P = T4 - T5; T6 = T4 + T5; T1f = Im[WS(rs, 7)]; { E T1o, T1n, T1p, Tb, Tc; Tb = Rm[WS(rs, 3)]; Tc = Rp[WS(rs, 12)]; T1o = Ip[WS(rs, 12)]; T6R = T1e - T1f; T1g = T1e + T1f; T1n = Tb - Tc; Td = Tb + Tc; T1p = Im[WS(rs, 3)]; { E T1j, T1k, T8, T9, T1q; T8 = Rp[WS(rs, 4)]; T9 = Rm[WS(rs, 11)]; T1q = T1o + T1p; T6U = T1o - T1p; T1j = Ip[WS(rs, 4)]; T1i = T8 - T9; Ta = T8 + T9; T1k = Im[WS(rs, 11)]; T2V = T1n + T1q; T1r = T1n - T1q; T6T = T1j - T1k; T1l = T1j + T1k; } } } { E T2U, T6V, T6S, T1h, T1s, T4Y, T4X, T2T, T2W; { E T7T, T8o, T1m, T7U, T7, Te, T8n; T7T = T3 - T6; T7 = T3 + T6; Te = Ta + Td; T8o = Ta - Td; T1m = T1i - T1l; T2U = T1i + T1l; T6j = T7 - Te; Tf = T7 + Te; T7U = T6U - T6T; T6V = T6T + T6U; T6S = T6Q + T6R; T8n = T6Q - T6R; T4t = T1d + T1g; T1h = T1d - T1g; T8W = T7T + T7U; T7V = T7T - T7U; T99 = T8o + T8n; T8p = T8n - T8o; T1s = T1m + T1r; T4Y = T1m - T1r; } T4X = T2S - T2P; T2T = T2P + T2S; T2W = T2U - T2V; T4u = T2U + T2V; T3L = FMA(KP707106781, T1s, T1h); T1t = FNMS(KP707106781, T1s, T1h); T3Z = FMA(KP707106781, T2W, T2T); T2X = FNMS(KP707106781, T2W, T2T); T5J = FNMS(KP707106781, T4Y, T4X); T4Z = FMA(KP707106781, T4Y, T4X); T7t = T6S + T6V; T6W = T6S - T6V; } } { E T29, T1S, T1V, T87, TR, T2c, T84, T6E, TU, T23, T6F, T22, TX, T24, T2e; E T21; { E TO, TN, TP, TL, TM; TL = Rm[0]; TM = Rp[WS(rs, 15)]; TO = Rp[WS(rs, 7)]; T5v = FMA(KP707106781, T4u, T4t); T4v = FNMS(KP707106781, T4u, T4t); TN = TL + TM; T29 = TL - TM; TP = Rm[WS(rs, 8)]; { E T6C, T6D, T1X, T20; { E T2a, T2b, T1T, T1U, TQ; T1T = Ip[WS(rs, 15)]; T1U = Im[0]; TQ = TO + TP; T1S = TO - TP; T2a = Ip[WS(rs, 7)]; T6C = T1T - T1U; T1V = T1T + T1U; T2b = Im[WS(rs, 8)]; T87 = TN - TQ; TR = TN + TQ; T2c = T2a + T2b; T6D = T2a - T2b; } { E T1Y, T1Z, TS, TT, TV, TW; TS = Rp[WS(rs, 3)]; TT = Rm[WS(rs, 12)]; T84 = T6C - T6D; T6E = T6C + T6D; T1Y = Ip[WS(rs, 3)]; T1X = TS - TT; TU = TS + TT; T1Z = Im[WS(rs, 12)]; TV = Rm[WS(rs, 4)]; TW = Rp[WS(rs, 11)]; T23 = Ip[WS(rs, 11)]; T6F = T1Y - T1Z; T20 = T1Y + T1Z; T22 = TV - TW; TX = TV + TW; T24 = Im[WS(rs, 4)]; } T2e = T1X - T20; T21 = T1X + T20; } } { E TY, T85, T25, T6G; TY = TU + TX; T85 = TU - TX; T25 = T23 + T24; T6G = T23 - T24; { E T4O, T1W, T2f, T8Z, T86, T89, T90, T27, T88, T26, T6H, T4P, T2d, T2g; T4O = T1S + T1V; T1W = T1S - T1V; TZ = TR + TY; T6B = TR - TY; T88 = T6G - T6F; T6H = T6F + T6G; T26 = T22 + T25; T2f = T22 - T25; T6I = T6E - T6H; T7x = T6E + T6H; T8Z = T85 + T84; T86 = T84 - T85; T89 = T87 - T88; T90 = T87 + T88; T27 = T21 - T26; T4M = T21 + T26; T4L = T29 + T2c; T2d = T29 - T2c; T2g = T2e + T2f; T4P = T2e - T2f; T91 = FNMS(KP414213562, T90, T8Z); T9d = FMA(KP414213562, T8Z, T90); T28 = FNMS(KP707106781, T27, T1W); T3S = FMA(KP707106781, T27, T1W); T3R = FMA(KP707106781, T2g, T2d); T2h = FNMS(KP707106781, T2g, T2d); T5B = FMA(KP707106781, T4P, T4O); T4Q = FNMS(KP707106781, T4P, T4O); T8v = FNMS(KP414213562, T86, T89); T8a = FMA(KP414213562, T89, T86); } } } { E T2A, T2j, TC, T8e, T2m, T2D, T6v, T8b, TF, T6w, T2F, T2s, T2t, TI, T6x; E T2w, TJ, T8c; { E Tw, Tx, Tz, TA, T6t, T6u; Tw = Rp[WS(rs, 1)]; T5C = FMA(KP707106781, T4M, T4L); T4N = FNMS(KP707106781, T4M, T4L); T6Z = T6I - T6B; T6J = T6B + T6I; Tx = Rm[WS(rs, 14)]; Tz = Rp[WS(rs, 9)]; TA = Rm[WS(rs, 6)]; { E T2k, Ty, TB, T2l, T2B, T2C; T2k = Ip[WS(rs, 1)]; T2A = Tw - Tx; Ty = Tw + Tx; T2j = Tz - TA; TB = Tz + TA; T2l = Im[WS(rs, 14)]; T2B = Ip[WS(rs, 9)]; T2C = Im[WS(rs, 6)]; TC = Ty + TB; T8e = Ty - TB; T2m = T2k + T2l; T6t = T2k - T2l; T6u = T2B - T2C; T2D = T2B + T2C; } { E TG, T2o, T2r, TH, T2u, T2v; { E TD, TE, T2p, T2q; TD = Rp[WS(rs, 5)]; T6v = T6t + T6u; T8b = T6t - T6u; TE = Rm[WS(rs, 10)]; T2p = Ip[WS(rs, 5)]; T2q = Im[WS(rs, 10)]; TG = Rm[WS(rs, 2)]; T2o = TD - TE; TF = TD + TE; T6w = T2p - T2q; T2r = T2p + T2q; TH = Rp[WS(rs, 13)]; T2u = Ip[WS(rs, 13)]; T2v = Im[WS(rs, 2)]; } T2F = T2o - T2r; T2s = T2o + T2r; T2t = TG - TH; TI = TG + TH; T6x = T2u - T2v; T2w = T2u + T2v; } } TJ = TF + TI; T8c = TF - TI; { E T8f, T6y, T2x, T2G; T8f = T6x - T6w; T6y = T6w + T6x; T2x = T2t + T2w; T2G = T2t - T2w; { E T4H, T2n, T2y, T4F, T8d, T92, T93, T8g; T6s = TC - TJ; TK = TC + TJ; T7w = T6v + T6y; T6z = T6v - T6y; T4H = T2m - T2j; T2n = T2j + T2m; T2y = T2s - T2x; T4F = T2s + T2x; T8d = T8b - T8c; T92 = T8c + T8b; T93 = T8e + T8f; T8g = T8e - T8f; { E T4E, T2E, T2H, T4I; T4E = T2A + T2D; T2E = T2A - T2D; T3P = FMA(KP707106781, T2y, T2n); T2z = FNMS(KP707106781, T2y, T2n); T9c = FNMS(KP414213562, T92, T93); T94 = FMA(KP414213562, T93, T92); T2H = T2F + T2G; T4I = T2G - T2F; T3O = FMA(KP707106781, T2H, T2E); T2I = FNMS(KP707106781, T2H, T2E); T5y = FMA(KP707106781, T4I, T4H); T4J = FNMS(KP707106781, T4I, T4H); T8u = FMA(KP414213562, T8d, T8g); T8h = FNMS(KP414213562, T8g, T8d); T5z = FMA(KP707106781, T4F, T4E); T4G = FNMS(KP707106781, T4F, T4E); } } } } { E T4w, T1J, T7Z, Tm, T6p, T80, T4x, T1O, T1z, Tp, T1A, T6k, T1x, T1u, Ts; E T1B; { E T1K, Ti, T1L, T6n, T1I, T1F, Tl, T1M; { E T1G, T1H, Tg, Th, Tj, Tk; Tg = Rp[WS(rs, 2)]; Th = Rm[WS(rs, 13)]; T1G = Ip[WS(rs, 2)]; T6Y = T6s + T6z; T6A = T6s - T6z; T1K = Tg - Th; Ti = Tg + Th; T1H = Im[WS(rs, 13)]; Tj = Rp[WS(rs, 10)]; Tk = Rm[WS(rs, 5)]; T1L = Ip[WS(rs, 10)]; T6n = T1G - T1H; T1I = T1G + T1H; T1F = Tj - Tk; Tl = Tj + Tk; T1M = Im[WS(rs, 5)]; } { E T1v, T1w, Tq, Tr; { E Tn, T1N, T6o, To; Tn = Rm[WS(rs, 1)]; T4w = T1I - T1F; T1J = T1F + T1I; T7Z = Ti - Tl; Tm = Ti + Tl; T1N = T1L + T1M; T6o = T1L - T1M; To = Rp[WS(rs, 14)]; T1v = Ip[WS(rs, 14)]; T6p = T6n + T6o; T80 = T6n - T6o; T4x = T1K + T1N; T1O = T1K - T1N; T1z = Tn - To; Tp = Tn + To; T1w = Im[WS(rs, 1)]; } Tq = Rp[WS(rs, 6)]; Tr = Rm[WS(rs, 9)]; T1A = Ip[WS(rs, 6)]; T6k = T1v - T1w; T1x = T1v + T1w; T1u = Tq - Tr; Ts = Tq + Tr; T1B = Im[WS(rs, 9)]; } } { E T4z, T6m, T4A, T2Z, T1E, T1P, T2Y, T50, T51; { E T1y, T81, T8q, T1D, T7Y, T8r; { E T7X, Tt, T1C, T6l, T7W; T4z = T1u + T1x; T1y = T1u - T1x; T7X = Tp - Ts; Tt = Tp + Ts; T1C = T1A + T1B; T6l = T1A - T1B; T81 = T7Z + T80; T8q = T7Z - T80; T6m = T6k + T6l; T7W = T6k - T6l; T4A = T1z + T1C; T1D = T1z - T1C; T6P = Tm - Tt; Tu = Tm + Tt; T7Y = T7W - T7X; T8r = T7X + T7W; } T2Z = FMA(KP414213562, T1y, T1D); T1E = FNMS(KP414213562, T1D, T1y); T9a = T81 + T7Y; T82 = T7Y - T81; T8X = T8q + T8r; T8s = T8q - T8r; T1P = FMA(KP414213562, T1O, T1J); T2Y = FNMS(KP414213562, T1J, T1O); } T4y = FNMS(KP414213562, T4x, T4w); T50 = FMA(KP414213562, T4w, T4x); T40 = T1P + T1E; T1Q = T1E - T1P; T3M = T2Y + T2Z; T30 = T2Y - T2Z; T51 = FMA(KP414213562, T4z, T4A); T4B = FNMS(KP414213562, T4A, T4z); T5w = T50 + T51; T52 = T50 - T51; T7u = T6p + T6m; T6q = T6m - T6p; } } } { E T7D, T7K, T7J, T5K, T4C, T7E, T83, T8w, T8t, T8i, T6r, T70, T6X, T6K; { E T8Y, T9e, T9b, T95, T8F, T8G, T8L, T8M; { E T7v, T7p, T7y, Tv, T10; T7D = Tf - Tu; Tv = Tf + Tu; T10 = TK + TZ; T7K = TK - TZ; T7J = T7t - T7u; T7v = T7t + T7u; T5K = T4B - T4y; T4C = T4y + T4B; T7p = Tv - T10; T7E = T7x - T7w; T7y = T7w + T7x; Rp[0] = Tv + T10; { E T9p, T9x, T9z, T9v; { E T9n, T7A, T7q, T7z, T9o, T9t, T9u; T8Y = FNMS(KP707106781, T8X, T8W); T9n = FMA(KP707106781, T8X, T8W); T7A = T7s * T7p; T7q = T7o * T7p; Rm[0] = T7v + T7y; T7z = T7v - T7y; T9o = T9c + T9d; T9e = T9c - T9d; T9b = FNMS(KP707106781, T9a, T99); T9t = FMA(KP707106781, T9a, T99); T9u = T94 + T91; T95 = T91 - T94; Rm[WS(rs, 8)] = FMA(T7o, T7z, T7A); Rp[WS(rs, 8)] = FNMS(T7s, T7z, T7q); T9p = FNMS(KP923879532, T9o, T9n); T9x = FMA(KP923879532, T9o, T9n); T9z = FMA(KP923879532, T9u, T9t); T9v = FNMS(KP923879532, T9u, T9t); } { E T9y, T9q, T9w, T9A; T9y = T3v * T9x; T9q = T9m * T9p; T9w = T9m * T9v; T9A = T3v * T9z; Rp[WS(rs, 1)] = FNMS(T3x, T9z, T9y); Rp[WS(rs, 9)] = FNMS(T9s, T9v, T9q); Rm[WS(rs, 9)] = FMA(T9s, T9p, T9w); Rm[WS(rs, 1)] = FMA(T3x, T9x, T9A); } } T83 = FMA(KP707106781, T82, T7V); T8F = FNMS(KP707106781, T82, T7V); T8G = T8u + T8v; T8w = T8u - T8v; T8t = FMA(KP707106781, T8s, T8p); T8L = FNMS(KP707106781, T8s, T8p); T8M = T8h + T8a; T8i = T8a - T8h; } { E T79, T7a, T7f, T7g; T6r = T6j + T6q; T79 = T6j - T6q; { E T8Q, T8H, T8T, T8N; T8Q = FMA(KP923879532, T8G, T8F); T8H = FNMS(KP923879532, T8G, T8F); T8T = FMA(KP923879532, T8M, T8L); T8N = FNMS(KP923879532, T8M, T8L); { E T8R, T8I, T8U, T8O; T8R = T8P * T8Q; T8I = T8E * T8H; T8U = T8P * T8T; T8O = T8E * T8N; Rp[WS(rs, 15)] = FNMS(T8S, T8T, T8R); Rp[WS(rs, 7)] = FNMS(T8K, T8N, T8I); Rm[WS(rs, 15)] = FMA(T8S, T8Q, T8U); Rm[WS(rs, 7)] = FMA(T8K, T8H, T8O); T7a = T6Z - T6Y; T70 = T6Y + T6Z; } } T6X = T6P + T6W; T7f = T6W - T6P; T7g = T6A - T6J; T6K = T6A + T6J; { E T7j, T7b, T7l, T7h; T7j = FMA(KP707106781, T7a, T79); T7b = FNMS(KP707106781, T7a, T79); T7l = FMA(KP707106781, T7g, T7f); T7h = FNMS(KP707106781, T7g, T7f); { E T7k, T7c, T7m, T7i; T7k = T5X * T7j; T7c = T78 * T7b; T7m = T5X * T7l; T7i = T78 * T7h; Rp[WS(rs, 6)] = FNMS(T5Z, T7l, T7k); Rp[WS(rs, 14)] = FNMS(T7e, T7h, T7c); Rm[WS(rs, 6)] = FMA(T5Z, T7j, T7m); Rm[WS(rs, 14)] = FMA(T7e, T7b, T7i); } } { E T9h, T96, T9j, T9f; T9h = FMA(KP923879532, T95, T8Y); T96 = FNMS(KP923879532, T95, T8Y); T9j = FMA(KP923879532, T9e, T9b); T9f = FNMS(KP923879532, T9e, T9b); { E T9k, T9i, T9g, T97; T9k = T3J * T9h; T9i = T3G * T9h; T9g = T98 * T96; T97 = T8V * T96; Rm[WS(rs, 5)] = FMA(T3G, T9j, T9k); Rp[WS(rs, 5)] = FNMS(T3J, T9j, T9i); Rm[WS(rs, 13)] = FMA(T8V, T9f, T9g); Rp[WS(rs, 13)] = FNMS(T98, T9f, T97); } } } } { E T31, T3r, T1R, T3m, T33, T32, T3s, T2K, T8z, T8j; { E T73, T6L, T75, T71; T73 = FMA(KP707106781, T6K, T6r); T6L = FNMS(KP707106781, T6K, T6r); T75 = FMA(KP707106781, T70, T6X); T71 = FNMS(KP707106781, T70, T6X); { E T76, T74, T72, T6M; T76 = T3d * T73; T74 = T3a * T73; T72 = T6O * T6L; T6M = T6i * T6L; Rm[WS(rs, 2)] = FMA(T3a, T75, T76); Rp[WS(rs, 2)] = FNMS(T3d, T75, T74); Rm[WS(rs, 10)] = FMA(T6i, T71, T72); Rp[WS(rs, 10)] = FNMS(T6O, T71, T6M); } } { E T7N, T7F, T7P, T7L; T7N = T7D + T7E; T7F = T7D - T7E; T7P = T7K + T7J; T7L = T7J - T7K; { E T7O, T7G, T7Q, T7M; T7O = T4p * T7N; T7G = T7C * T7F; T7Q = T4p * T7P; T7M = T7C * T7L; Rp[WS(rs, 4)] = FNMS(T4r, T7P, T7O); Rp[WS(rs, 12)] = FNMS(T7I, T7L, T7G); Rm[WS(rs, 4)] = FMA(T4r, T7N, T7Q); Rm[WS(rs, 12)] = FMA(T7I, T7F, T7M); } } T31 = FMA(KP923879532, T30, T2X); T3r = FNMS(KP923879532, T30, T2X); T8z = FMA(KP923879532, T8i, T83); T8j = FNMS(KP923879532, T8i, T83); { E T8B, T8x, T8C, T8A; T8B = FMA(KP923879532, T8w, T8t); T8x = FNMS(KP923879532, T8w, T8t); T8C = T1a * T8z; T8A = T16 * T8z; { E T8y, T8k, T2i, T2J; T8y = T8m * T8j; T8k = T7S * T8j; Rm[WS(rs, 3)] = FMA(T16, T8B, T8C); Rp[WS(rs, 3)] = FNMS(T1a, T8B, T8A); Rm[WS(rs, 11)] = FMA(T7S, T8x, T8y); Rp[WS(rs, 11)] = FNMS(T8m, T8x, T8k); T1R = FMA(KP923879532, T1Q, T1t); T3m = FNMS(KP923879532, T1Q, T1t); T33 = FNMS(KP668178637, T28, T2h); T2i = FMA(KP668178637, T2h, T28); T2J = FNMS(KP668178637, T2I, T2z); T32 = FMA(KP668178637, T2z, T2I); T3s = T2J + T2i; T2K = T2i - T2J; } } { E T5l, T53, T5f, T4D, T4K, T4R, T56, T5g; T5l = FNMS(KP923879532, T52, T4Z); T53 = FMA(KP923879532, T52, T4Z); { E T3t, T3D, T3f, T2L; T3t = FNMS(KP831469612, T3s, T3r); T3D = FMA(KP831469612, T3s, T3r); T3f = FMA(KP831469612, T2K, T1R); T2L = FNMS(KP831469612, T2K, T1R); { E T3n, T34, T3g, T2M; T3n = T32 + T33; T34 = T32 - T33; T3g = T3e * T3f; T2M = T1c * T2L; { E T3o, T3z, T3j, T35; T3o = FNMS(KP831469612, T3n, T3m); T3z = FMA(KP831469612, T3n, T3m); T3j = FMA(KP831469612, T34, T31); T35 = FNMS(KP831469612, T34, T31); { E T3u, T3p, T3E, T3A; T3u = T3q * T3o; T3p = T3l * T3o; T3E = T3C * T3z; T3A = T3y * T3z; { E T3k, T36, T54, T55; T3k = T3e * T3j; Ip[WS(rs, 2)] = FNMS(T3i, T3j, T3g); T36 = T1c * T35; Ip[WS(rs, 10)] = FNMS(T2O, T35, T2M); Im[WS(rs, 6)] = FMA(T3l, T3t, T3u); Ip[WS(rs, 6)] = FNMS(T3q, T3t, T3p); Im[WS(rs, 14)] = FMA(T3y, T3D, T3E); Ip[WS(rs, 14)] = FNMS(T3C, T3D, T3A); Im[WS(rs, 2)] = FMA(T3i, T3f, T3k); Im[WS(rs, 10)] = FMA(T2O, T2L, T36); T5f = FMA(KP923879532, T4C, T4v); T4D = FNMS(KP923879532, T4C, T4v); T4K = FNMS(KP668178637, T4J, T4G); T54 = FMA(KP668178637, T4G, T4J); T55 = FMA(KP668178637, T4N, T4Q); T4R = FNMS(KP668178637, T4Q, T4N); T56 = T54 - T55; T5g = T54 + T55; } } } } } { E T4h, T41, T4c, T3N, T3Q, T3T, T44, T4d; T4h = FNMS(KP923879532, T40, T3Z); T41 = FMA(KP923879532, T40, T3Z); { E T57, T5b, T5h, T5p; T57 = FNMS(KP831469612, T56, T53); T5b = FMA(KP831469612, T56, T53); T5h = FNMS(KP831469612, T5g, T5f); T5p = FMA(KP831469612, T5g, T5f); { E T5m, T4S, T5i, T5q; T5m = T4K - T4R; T4S = T4K + T4R; T5i = T5e * T5h; T5q = T17 * T5p; { E T5n, T5r, T59, T4T; T5n = FMA(KP831469612, T5m, T5l); T5r = FNMS(KP831469612, T5m, T5l); T59 = FMA(KP831469612, T4S, T4D); T4T = FNMS(KP831469612, T4S, T4D); { E T5o, T5s, T5c, T5a; T5o = T5e * T5n; Ip[WS(rs, 5)] = FNMS(T5k, T5n, T5i); T5s = T17 * T5r; Ip[WS(rs, 13)] = FNMS(T1b, T5r, T5q); T5c = T14 * T59; T5a = T11 * T59; { E T58, T4U, T42, T43; T58 = T4W * T4T; T4U = T4s * T4T; Im[WS(rs, 5)] = FMA(T5k, T5h, T5o); Im[WS(rs, 13)] = FMA(T1b, T5p, T5s); Im[WS(rs, 1)] = FMA(T11, T5b, T5c); Ip[WS(rs, 1)] = FNMS(T14, T5b, T5a); Im[WS(rs, 9)] = FMA(T4s, T57, T58); Ip[WS(rs, 9)] = FNMS(T4W, T57, T4U); T4c = FNMS(KP923879532, T3M, T3L); T3N = FMA(KP923879532, T3M, T3L); T3Q = FNMS(KP198912367, T3P, T3O); T42 = FMA(KP198912367, T3O, T3P); T43 = FNMS(KP198912367, T3R, T3S); T3T = FMA(KP198912367, T3S, T3R); T44 = T42 + T43; T4d = T43 - T42; } } } } } T67 = FNMS(KP923879532, T5K, T5J); T5L = FMA(KP923879532, T5K, T5J); { E T45, T49, T4e, T4l; T45 = FNMS(KP980785280, T44, T41); T49 = FMA(KP980785280, T44, T41); T4e = FNMS(KP980785280, T4d, T4c); T4l = FMA(KP980785280, T4d, T4c); { E T4i, T3U, T4f, T4m; T4i = T3Q - T3T; T3U = T3Q + T3T; T4f = T4b * T4e; T4m = T12 * T4l; { E T4j, T4n, T47, T3V; T4j = FNMS(KP980785280, T4i, T4h); T4n = FMA(KP980785280, T4i, T4h); T47 = FMA(KP980785280, T3U, T3N); T3V = FNMS(KP980785280, T3U, T3N); { E T4k, T4o, T4a, T48; T4k = T4b * T4j; Ip[WS(rs, 12)] = FNMS(T4g, T4j, T4f); T4o = T12 * T4n; Ip[WS(rs, 4)] = FNMS(T15, T4n, T4m); T4a = T39 * T47; T48 = T37 * T47; { E T46, T3W, T5M, T5N; T46 = T3Y * T3V; T3W = T3K * T3V; Im[WS(rs, 12)] = FMA(T4g, T4e, T4k); Im[WS(rs, 4)] = FMA(T15, T4l, T4o); Im[0] = FMA(T37, T49, T4a); Ip[0] = FNMS(T39, T49, T48); Im[WS(rs, 8)] = FMA(T3K, T45, T46); Ip[WS(rs, 8)] = FNMS(T3Y, T45, T3W); T61 = FMA(KP923879532, T5w, T5v); T5x = FNMS(KP923879532, T5w, T5v); T5A = FNMS(KP198912367, T5z, T5y); T5M = FMA(KP198912367, T5y, T5z); T5N = FMA(KP198912367, T5B, T5C); T5D = FNMS(KP198912367, T5C, T5B); T5O = T5M - T5N; T62 = T5M + T5N; } } } } } } } } } } } T5V = FMA(KP980785280, T5O, T5L); T5P = FNMS(KP980785280, T5O, T5L); { E T6c, T63, T5E, T68; T6c = FMA(KP980785280, T62, T61); T63 = FNMS(KP980785280, T62, T61); T5E = T5A + T5D; T68 = T5D - T5A; { E T64, T6d, T6f, T69; T64 = T60 * T63; T6d = T6b * T6c; T6f = FNMS(KP980785280, T68, T67); T69 = FMA(KP980785280, T68, T67); { E T5F, T5S, T6a, T6g; T5F = FMA(KP980785280, T5E, T5x); T5S = FNMS(KP980785280, T5E, T5x); T6a = T60 * T69; Ip[WS(rs, 7)] = FNMS(T66, T69, T64); T6g = T6b * T6f; Ip[WS(rs, 15)] = FNMS(T6e, T6f, T6d); { E T5W, T5T, T5Q, T5G; T5W = T5U * T5S; T5T = T5R * T5S; T5Q = T5I * T5F; T5G = T5u * T5F; Im[WS(rs, 7)] = FMA(T66, T63, T6a); Im[WS(rs, 15)] = FMA(T6e, T6c, T6g); Im[WS(rs, 3)] = FMA(T5R, T5V, T5W); Ip[WS(rs, 3)] = FNMS(T5U, T5V, T5T); Im[WS(rs, 11)] = FMA(T5u, T5P, T5Q); Ip[WS(rs, 11)] = FNMS(T5I, T5P, T5G); } } } } } } } static const tw_instr twinstr[] = { {TW_CEXP, 1, 1}, {TW_CEXP, 1, 3}, {TW_CEXP, 1, 9}, {TW_CEXP, 1, 27}, {TW_NEXT, 1, 0} }; static const hc2c_desc desc = { 32, "hc2cb2_32", twinstr, &GENUS, {236, 98, 252, 0} }; void X(codelet_hc2cb2_32) (planner *p) { X(khc2c_register) (p, hc2cb2_32, &desc, HC2C_VIA_RDFT); } #else /* HAVE_FMA */ /* Generated by: ../../../genfft/gen_hc2c.native -compact -variables 4 -pipeline-latency 4 -sign 1 -twiddle-log3 -precompute-twiddles -n 32 -dif -name hc2cb2_32 -include hc2cb.h */ /* * This function contains 488 FP additions, 280 FP multiplications, * (or, 376 additions, 168 multiplications, 112 fused multiply/add), * 160 stack variables, 7 constants, and 128 memory accesses */ #include "hc2cb.h" static void hc2cb2_32(R *Rp, R *Ip, R *Rm, R *Im, const R *W, stride rs, INT mb, INT me, INT ms) { DK(KP555570233, +0.555570233019602224742830813948532874374937191); DK(KP831469612, +0.831469612302545237078788377617905756738560812); DK(KP980785280, +0.980785280403230449126182236134239036973933731); DK(KP195090322, +0.195090322016128267848284868477022240927691618); DK(KP923879532, +0.923879532511286756128183189396788286822416626); DK(KP382683432, +0.382683432365089771728459984030398866761344562); DK(KP707106781, +0.707106781186547524400844362104849039284835938); { INT m; for (m = mb, W = W + ((mb - 1) * 8); m < me; m = m + 1, Rp = Rp + ms, Ip = Ip + ms, Rm = Rm - ms, Im = Im - ms, W = W + 8, MAKE_VOLATILE_STRIDE(128, rs)) { E T11, T14, T12, T15, T17, T2z, T2B, T1c, T18, T1d, T1g, T1k, T2F, T2L, T3t; E T4H, T3h, T3V, T3b, T4v, T4T, T4X, T6t, T71, T6z, T75, T81, T8x, T8f, T8z; E T2R, T2V, T8p, T8t, T4r, T4t, T53, T69, T3n, T3r, T7P, T7T, T4P, T4R, T6F; E T6R, T1f, T2X, T1j, T2Y, T1l, T31, T2d, T2Z, T49, T4h, T4c, T4i, T4d, T4n; E T4f, T4j; { E T2P, T3q, T2U, T3l, T2Q, T3p, T2T, T3m, T2D, T3g, T2K, T39, T2E, T3f, T2J; E T3a; { E T13, T1b, T16, T1a; T11 = W[0]; T14 = W[1]; T12 = W[2]; T15 = W[3]; T13 = T11 * T12; T1b = T14 * T12; T16 = T14 * T15; T1a = T11 * T15; T17 = T13 + T16; T2z = T13 - T16; T2B = T1a + T1b; T1c = T1a - T1b; T18 = W[4]; T2P = T12 * T18; T3q = T14 * T18; T2U = T15 * T18; T3l = T11 * T18; T1d = W[5]; T2Q = T15 * T1d; T3p = T11 * T1d; T2T = T12 * T1d; T3m = T14 * T1d; T1g = W[6]; T2D = T11 * T1g; T3g = T15 * T1g; T2K = T14 * T1g; T39 = T12 * T1g; T1k = W[7]; T2E = T14 * T1k; T3f = T12 * T1k; T2J = T11 * T1k; T3a = T15 * T1k; } T2F = T2D - T2E; T2L = T2J + T2K; T3t = T39 - T3a; T4H = T2J - T2K; T3h = T3f - T3g; T3V = T3f + T3g; T3b = T39 + T3a; T4v = T2D + T2E; T4T = FMA(T18, T1g, T1d * T1k); T4X = FNMS(T1d, T1g, T18 * T1k); { E T6r, T6s, T6x, T6y; T6r = T17 * T1g; T6s = T1c * T1k; T6t = T6r - T6s; T71 = T6r + T6s; T6x = T17 * T1k; T6y = T1c * T1g; T6z = T6x + T6y; T75 = T6x - T6y; } { E T7Z, T80, T8d, T8e; T7Z = T2z * T1g; T80 = T2B * T1k; T81 = T7Z + T80; T8x = T7Z - T80; T8d = T2z * T1k; T8e = T2B * T1g; T8f = T8d - T8e; T8z = T8d + T8e; T2R = T2P - T2Q; T2V = T2T + T2U; T8p = FMA(T2R, T1g, T2V * T1k); T8t = FNMS(T2V, T1g, T2R * T1k); } T4r = T2P + T2Q; T4t = T2T - T2U; T53 = FMA(T4r, T1g, T4t * T1k); T69 = FNMS(T4t, T1g, T4r * T1k); T3n = T3l + T3m; T3r = T3p - T3q; T7P = FMA(T3n, T1g, T3r * T1k); T7T = FNMS(T3r, T1g, T3n * T1k); T4P = T3l - T3m; T4R = T3p + T3q; T6F = FMA(T4P, T1g, T4R * T1k); T6R = FNMS(T4R, T1g, T4P * T1k); { E T19, T1e, T1h, T1i; T19 = T17 * T18; T1e = T1c * T1d; T1f = T19 + T1e; T2X = T19 - T1e; T1h = T17 * T1d; T1i = T1c * T18; T1j = T1h - T1i; T2Y = T1h + T1i; } T1l = FMA(T1f, T1g, T1j * T1k); T31 = FNMS(T2Y, T1g, T2X * T1k); T2d = FNMS(T1j, T1g, T1f * T1k); T2Z = FMA(T2X, T1g, T2Y * T1k); { E T47, T48, T4a, T4b; T47 = T2z * T18; T48 = T2B * T1d; T49 = T47 - T48; T4h = T47 + T48; T4a = T2z * T1d; T4b = T2B * T18; T4c = T4a + T4b; T4i = T4a - T4b; } T4d = FMA(T49, T1g, T4c * T1k); T4n = FNMS(T4i, T1g, T4h * T1k); T4f = FNMS(T4c, T1g, T49 * T1k); T4j = FMA(T4h, T1g, T4i * T1k); } { E T56, T7b, T7C, T6c, Tf, T1m, T6f, T7c, T3Y, T4I, T2t, T32, T5d, T7D, T3w; E T4w, Tu, T2e, T7g, T7F, T7j, T7G, T1B, T33, T3z, T40, T5l, T6i, T5s, T6h; E T3C, T3Z, TK, T1D, T7v, T86, T7y, T85, T1S, T35, T3O, T4C, T5F, T6J, T5M; E T6K, T3R, T4D, TZ, T1U, T7o, T89, T7r, T88, T29, T36, T3H, T4z, T5Y, T6M; E T65, T6N, T3K, T4A; { E T3, T54, T2h, T6b, T6, T6a, T2k, T55, Ta, T57, T2o, T58, Td, T5a, T2r; E T5b; { E T1, T2, T2f, T2g; T1 = Rp[0]; T2 = Rm[WS(rs, 15)]; T3 = T1 + T2; T54 = T1 - T2; T2f = Ip[0]; T2g = Im[WS(rs, 15)]; T2h = T2f - T2g; T6b = T2f + T2g; } { E T4, T5, T2i, T2j; T4 = Rp[WS(rs, 8)]; T5 = Rm[WS(rs, 7)]; T6 = T4 + T5; T6a = T4 - T5; T2i = Ip[WS(rs, 8)]; T2j = Im[WS(rs, 7)]; T2k = T2i - T2j; T55 = T2i + T2j; } { E T8, T9, T2m, T2n; T8 = Rp[WS(rs, 4)]; T9 = Rm[WS(rs, 11)]; Ta = T8 + T9; T57 = T8 - T9; T2m = Ip[WS(rs, 4)]; T2n = Im[WS(rs, 11)]; T2o = T2m - T2n; T58 = T2m + T2n; } { E Tb, Tc, T2p, T2q; Tb = Rm[WS(rs, 3)]; Tc = Rp[WS(rs, 12)]; Td = Tb + Tc; T5a = Tb - Tc; T2p = Ip[WS(rs, 12)]; T2q = Im[WS(rs, 3)]; T2r = T2p - T2q; T5b = T2p + T2q; } { E T7, Te, T2l, T2s; T56 = T54 - T55; T7b = T54 + T55; T7C = T6b - T6a; T6c = T6a + T6b; T7 = T3 + T6; Te = Ta + Td; Tf = T7 + Te; T1m = T7 - Te; { E T6d, T6e, T3W, T3X; T6d = T57 + T58; T6e = T5a + T5b; T6f = KP707106781 * (T6d - T6e); T7c = KP707106781 * (T6d + T6e); T3W = T2h - T2k; T3X = Ta - Td; T3Y = T3W - T3X; T4I = T3X + T3W; } T2l = T2h + T2k; T2s = T2o + T2r; T2t = T2l - T2s; T32 = T2l + T2s; { E T59, T5c, T3u, T3v; T59 = T57 - T58; T5c = T5a - T5b; T5d = KP707106781 * (T59 + T5c); T7D = KP707106781 * (T59 - T5c); T3u = T3 - T6; T3v = T2r - T2o; T3w = T3u - T3v; T4w = T3u + T3v; } } } { E Ti, T5p, T1w, T5n, Tl, T5m, T1z, T5q, Tp, T5i, T1p, T5g, Ts, T5f, T1s; E T5j; { E Tg, Th, T1u, T1v; Tg = Rp[WS(rs, 2)]; Th = Rm[WS(rs, 13)]; Ti = Tg + Th; T5p = Tg - Th; T1u = Ip[WS(rs, 2)]; T1v = Im[WS(rs, 13)]; T1w = T1u - T1v; T5n = T1u + T1v; } { E Tj, Tk, T1x, T1y; Tj = Rp[WS(rs, 10)]; Tk = Rm[WS(rs, 5)]; Tl = Tj + Tk; T5m = Tj - Tk; T1x = Ip[WS(rs, 10)]; T1y = Im[WS(rs, 5)]; T1z = T1x - T1y; T5q = T1x + T1y; } { E Tn, To, T1n, T1o; Tn = Rm[WS(rs, 1)]; To = Rp[WS(rs, 14)]; Tp = Tn + To; T5i = Tn - To; T1n = Ip[WS(rs, 14)]; T1o = Im[WS(rs, 1)]; T1p = T1n - T1o; T5g = T1n + T1o; } { E Tq, Tr, T1q, T1r; Tq = Rp[WS(rs, 6)]; Tr = Rm[WS(rs, 9)]; Ts = Tq + Tr; T5f = Tq - Tr; T1q = Ip[WS(rs, 6)]; T1r = Im[WS(rs, 9)]; T1s = T1q - T1r; T5j = T1q + T1r; } { E Tm, Tt, T7e, T7f; Tm = Ti + Tl; Tt = Tp + Ts; Tu = Tm + Tt; T2e = Tm - Tt; T7e = T5p + T5q; T7f = T5n - T5m; T7g = FNMS(KP923879532, T7f, KP382683432 * T7e); T7F = FMA(KP382683432, T7f, KP923879532 * T7e); } { E T7h, T7i, T1t, T1A; T7h = T5i + T5j; T7i = T5f + T5g; T7j = FNMS(KP923879532, T7i, KP382683432 * T7h); T7G = FMA(KP382683432, T7i, KP923879532 * T7h); T1t = T1p + T1s; T1A = T1w + T1z; T1B = T1t - T1A; T33 = T1A + T1t; } { E T3x, T3y, T5h, T5k; T3x = T1p - T1s; T3y = Tp - Ts; T3z = T3x - T3y; T40 = T3y + T3x; T5h = T5f - T5g; T5k = T5i - T5j; T5l = FNMS(KP382683432, T5k, KP923879532 * T5h); T6i = FMA(KP382683432, T5h, KP923879532 * T5k); } { E T5o, T5r, T3A, T3B; T5o = T5m + T5n; T5r = T5p - T5q; T5s = FMA(KP923879532, T5o, KP382683432 * T5r); T6h = FNMS(KP382683432, T5o, KP923879532 * T5r); T3A = Ti - Tl; T3B = T1w - T1z; T3C = T3A + T3B; T3Z = T3A - T3B; } } { E Ty, T5v, T1G, T5H, TB, T5G, T1J, T5w, TI, T5K, T1Q, T5D, TF, T5J, T1N; E T5A; { E Tw, Tx, T1H, T1I; Tw = Rp[WS(rs, 1)]; Tx = Rm[WS(rs, 14)]; Ty = Tw + Tx; T5v = Tw - Tx; { E T1E, T1F, Tz, TA; T1E = Ip[WS(rs, 1)]; T1F = Im[WS(rs, 14)]; T1G = T1E - T1F; T5H = T1E + T1F; Tz = Rp[WS(rs, 9)]; TA = Rm[WS(rs, 6)]; TB = Tz + TA; T5G = Tz - TA; } T1H = Ip[WS(rs, 9)]; T1I = Im[WS(rs, 6)]; T1J = T1H - T1I; T5w = T1H + T1I; { E TG, TH, T5B, T1O, T1P, T5C; TG = Rm[WS(rs, 2)]; TH = Rp[WS(rs, 13)]; T5B = TG - TH; T1O = Ip[WS(rs, 13)]; T1P = Im[WS(rs, 2)]; T5C = T1O + T1P; TI = TG + TH; T5K = T5B + T5C; T1Q = T1O - T1P; T5D = T5B - T5C; } { E TD, TE, T5y, T1L, T1M, T5z; TD = Rp[WS(rs, 5)]; TE = Rm[WS(rs, 10)]; T5y = TD - TE; T1L = Ip[WS(rs, 5)]; T1M = Im[WS(rs, 10)]; T5z = T1L + T1M; TF = TD + TE; T5J = T5y + T5z; T1N = T1L - T1M; T5A = T5y - T5z; } } { E TC, TJ, T7t, T7u; TC = Ty + TB; TJ = TF + TI; TK = TC + TJ; T1D = TC - TJ; T7t = T5H - T5G; T7u = KP707106781 * (T5A - T5D); T7v = T7t + T7u; T86 = T7t - T7u; } { E T7w, T7x, T1K, T1R; T7w = T5v + T5w; T7x = KP707106781 * (T5J + T5K); T7y = T7w - T7x; T85 = T7w + T7x; T1K = T1G + T1J; T1R = T1N + T1Q; T1S = T1K - T1R; T35 = T1K + T1R; } { E T3M, T3N, T5x, T5E; T3M = T1G - T1J; T3N = TF - TI; T3O = T3M - T3N; T4C = T3N + T3M; T5x = T5v - T5w; T5E = KP707106781 * (T5A + T5D); T5F = T5x - T5E; T6J = T5x + T5E; } { E T5I, T5L, T3P, T3Q; T5I = T5G + T5H; T5L = KP707106781 * (T5J - T5K); T5M = T5I - T5L; T6K = T5I + T5L; T3P = Ty - TB; T3Q = T1Q - T1N; T3R = T3P - T3Q; T4D = T3P + T3Q; } } { E TN, T5O, T1X, T60, TQ, T5Z, T20, T5P, TX, T63, T27, T5W, TU, T62, T24; E T5T; { E TL, TM, T1Y, T1Z; TL = Rm[0]; TM = Rp[WS(rs, 15)]; TN = TL + TM; T5O = TL - TM; { E T1V, T1W, TO, TP; T1V = Ip[WS(rs, 15)]; T1W = Im[0]; T1X = T1V - T1W; T60 = T1V + T1W; TO = Rp[WS(rs, 7)]; TP = Rm[WS(rs, 8)]; TQ = TO + TP; T5Z = TO - TP; } T1Y = Ip[WS(rs, 7)]; T1Z = Im[WS(rs, 8)]; T20 = T1Y - T1Z; T5P = T1Y + T1Z; { E TV, TW, T5U, T25, T26, T5V; TV = Rm[WS(rs, 4)]; TW = Rp[WS(rs, 11)]; T5U = TV - TW; T25 = Ip[WS(rs, 11)]; T26 = Im[WS(rs, 4)]; T5V = T25 + T26; TX = TV + TW; T63 = T5U + T5V; T27 = T25 - T26; T5W = T5U - T5V; } { E TS, TT, T5R, T22, T23, T5S; TS = Rp[WS(rs, 3)]; TT = Rm[WS(rs, 12)]; T5R = TS - TT; T22 = Ip[WS(rs, 3)]; T23 = Im[WS(rs, 12)]; T5S = T22 + T23; TU = TS + TT; T62 = T5R + T5S; T24 = T22 - T23; T5T = T5R - T5S; } } { E TR, TY, T7m, T7n; TR = TN + TQ; TY = TU + TX; TZ = TR + TY; T1U = TR - TY; T7m = KP707106781 * (T5T - T5W); T7n = T5Z + T60; T7o = T7m - T7n; T89 = T7n + T7m; } { E T7p, T7q, T21, T28; T7p = T5O + T5P; T7q = KP707106781 * (T62 + T63); T7r = T7p - T7q; T88 = T7p + T7q; T21 = T1X + T20; T28 = T24 + T27; T29 = T21 - T28; T36 = T21 + T28; } { E T3F, T3G, T5Q, T5X; T3F = T1X - T20; T3G = TU - TX; T3H = T3F - T3G; T4z = T3G + T3F; T5Q = T5O - T5P; T5X = KP707106781 * (T5T + T5W); T5Y = T5Q - T5X; T6M = T5Q + T5X; } { E T61, T64, T3I, T3J; T61 = T5Z - T60; T64 = KP707106781 * (T62 - T63); T65 = T61 - T64; T6N = T61 + T64; T3I = TN - TQ; T3J = T27 - T24; T3K = T3I - T3J; T4A = T3I + T3J; } } { E Tv, T10, T30, T34, T37, T38; Tv = Tf + Tu; T10 = TK + TZ; T30 = Tv - T10; T34 = T32 + T33; T37 = T35 + T36; T38 = T34 - T37; Rp[0] = Tv + T10; Rm[0] = T34 + T37; Rp[WS(rs, 8)] = FNMS(T31, T38, T2Z * T30); Rm[WS(rs, 8)] = FMA(T31, T30, T2Z * T38); } { E T3e, T3o, T3k, T3s; { E T3c, T3d, T3i, T3j; T3c = Tf - Tu; T3d = T36 - T35; T3e = T3c - T3d; T3o = T3c + T3d; T3i = T32 - T33; T3j = TK - TZ; T3k = T3i - T3j; T3s = T3j + T3i; } Rp[WS(rs, 12)] = FNMS(T3h, T3k, T3b * T3e); Rm[WS(rs, 12)] = FMA(T3b, T3k, T3h * T3e); Rp[WS(rs, 4)] = FNMS(T3r, T3s, T3n * T3o); Rm[WS(rs, 4)] = FMA(T3n, T3s, T3r * T3o); } { E T1C, T2u, T2M, T2G, T2x, T2H, T2b, T2N; T1C = T1m + T1B; T2u = T2e + T2t; T2M = T2t - T2e; T2G = T1m - T1B; { E T2v, T2w, T1T, T2a; T2v = T1D + T1S; T2w = T29 - T1U; T2x = KP707106781 * (T2v + T2w); T2H = KP707106781 * (T2w - T2v); T1T = T1D - T1S; T2a = T1U + T29; T2b = KP707106781 * (T1T + T2a); T2N = KP707106781 * (T1T - T2a); } { E T2c, T2y, T2S, T2W; T2c = T1C - T2b; T2y = T2u - T2x; Rp[WS(rs, 10)] = FNMS(T2d, T2y, T1l * T2c); Rm[WS(rs, 10)] = FMA(T2d, T2c, T1l * T2y); T2S = T2G + T2H; T2W = T2M + T2N; Rp[WS(rs, 6)] = FNMS(T2V, T2W, T2R * T2S); Rm[WS(rs, 6)] = FMA(T2R, T2W, T2V * T2S); } { E T2A, T2C, T2I, T2O; T2A = T1C + T2b; T2C = T2u + T2x; Rp[WS(rs, 2)] = FNMS(T2B, T2C, T2z * T2A); Rm[WS(rs, 2)] = FMA(T2B, T2A, T2z * T2C); T2I = T2G - T2H; T2O = T2M - T2N; Rp[WS(rs, 14)] = FNMS(T2L, T2O, T2F * T2I); Rm[WS(rs, 14)] = FMA(T2F, T2O, T2L * T2I); } } { E T4y, T4U, T4K, T4Y, T4F, T4Z, T4N, T4V, T4x, T4J; T4x = KP707106781 * (T3Z + T40); T4y = T4w - T4x; T4U = T4w + T4x; T4J = KP707106781 * (T3C + T3z); T4K = T4I - T4J; T4Y = T4I + T4J; { E T4B, T4E, T4L, T4M; T4B = FNMS(KP382683432, T4A, KP923879532 * T4z); T4E = FMA(KP923879532, T4C, KP382683432 * T4D); T4F = T4B - T4E; T4Z = T4E + T4B; T4L = FNMS(KP382683432, T4C, KP923879532 * T4D); T4M = FMA(KP382683432, T4z, KP923879532 * T4A); T4N = T4L - T4M; T4V = T4L + T4M; } { E T4G, T4O, T51, T52; T4G = T4y - T4F; T4O = T4K - T4N; Rp[WS(rs, 13)] = FNMS(T4H, T4O, T4v * T4G); Rm[WS(rs, 13)] = FMA(T4H, T4G, T4v * T4O); T51 = T4U + T4V; T52 = T4Y + T4Z; Rp[WS(rs, 1)] = FNMS(T1c, T52, T17 * T51); Rm[WS(rs, 1)] = FMA(T17, T52, T1c * T51); } { E T4Q, T4S, T4W, T50; T4Q = T4y + T4F; T4S = T4K + T4N; Rp[WS(rs, 5)] = FNMS(T4R, T4S, T4P * T4Q); Rm[WS(rs, 5)] = FMA(T4R, T4Q, T4P * T4S); T4W = T4U - T4V; T50 = T4Y - T4Z; Rp[WS(rs, 9)] = FNMS(T4X, T50, T4T * T4W); Rm[WS(rs, 9)] = FMA(T4T, T50, T4X * T4W); } } { E T3E, T4k, T42, T4o, T3T, T4p, T45, T4l, T3D, T41; T3D = KP707106781 * (T3z - T3C); T3E = T3w - T3D; T4k = T3w + T3D; T41 = KP707106781 * (T3Z - T40); T42 = T3Y - T41; T4o = T3Y + T41; { E T3L, T3S, T43, T44; T3L = FNMS(KP923879532, T3K, KP382683432 * T3H); T3S = FMA(KP382683432, T3O, KP923879532 * T3R); T3T = T3L - T3S; T4p = T3S + T3L; T43 = FNMS(KP923879532, T3O, KP382683432 * T3R); T44 = FMA(KP923879532, T3H, KP382683432 * T3K); T45 = T43 - T44; T4l = T43 + T44; } { E T3U, T46, T4s, T4u; T3U = T3E - T3T; T46 = T42 - T45; Rp[WS(rs, 15)] = FNMS(T3V, T46, T3t * T3U); Rm[WS(rs, 15)] = FMA(T3V, T3U, T3t * T46); T4s = T4k + T4l; T4u = T4o + T4p; Rp[WS(rs, 3)] = FNMS(T4t, T4u, T4r * T4s); Rm[WS(rs, 3)] = FMA(T4r, T4u, T4t * T4s); } { E T4e, T4g, T4m, T4q; T4e = T3E + T3T; T4g = T42 + T45; Rp[WS(rs, 7)] = FNMS(T4f, T4g, T4d * T4e); Rm[WS(rs, 7)] = FMA(T4f, T4e, T4d * T4g); T4m = T4k - T4l; T4q = T4o - T4p; Rp[WS(rs, 11)] = FNMS(T4n, T4q, T4j * T4m); Rm[WS(rs, 11)] = FMA(T4j, T4q, T4n * T4m); } } { E T6I, T72, T6X, T73, T6P, T77, T6U, T76; { E T6G, T6H, T6V, T6W; T6G = T56 + T5d; T6H = T6h + T6i; T6I = T6G + T6H; T72 = T6G - T6H; T6V = FMA(KP195090322, T6J, KP980785280 * T6K); T6W = FNMS(KP195090322, T6M, KP980785280 * T6N); T6X = T6V + T6W; T73 = T6W - T6V; } { E T6L, T6O, T6S, T6T; T6L = FNMS(KP195090322, T6K, KP980785280 * T6J); T6O = FMA(KP980785280, T6M, KP195090322 * T6N); T6P = T6L + T6O; T77 = T6L - T6O; T6S = T6c + T6f; T6T = T5s + T5l; T6U = T6S + T6T; T76 = T6S - T6T; } { E T6Q, T6Y, T79, T7a; T6Q = T6I - T6P; T6Y = T6U - T6X; Ip[WS(rs, 8)] = FNMS(T6R, T6Y, T6F * T6Q); Im[WS(rs, 8)] = FMA(T6R, T6Q, T6F * T6Y); T79 = T72 + T73; T7a = T76 + T77; Ip[WS(rs, 4)] = FNMS(T1d, T7a, T18 * T79); Im[WS(rs, 4)] = FMA(T18, T7a, T1d * T79); } { E T6Z, T70, T74, T78; T6Z = T6I + T6P; T70 = T6U + T6X; Ip[0] = FNMS(T14, T70, T11 * T6Z); Im[0] = FMA(T14, T6Z, T11 * T70); T74 = T72 - T73; T78 = T76 - T77; Ip[WS(rs, 12)] = FNMS(T75, T78, T71 * T74); Im[WS(rs, 12)] = FMA(T71, T78, T75 * T74); } } { E T84, T8q, T8l, T8r, T8b, T8v, T8i, T8u; { E T82, T83, T8j, T8k; T82 = T7b + T7c; T83 = T7F + T7G; T84 = T82 - T83; T8q = T82 + T83; T8j = FMA(KP195090322, T86, KP980785280 * T85); T8k = FMA(KP195090322, T89, KP980785280 * T88); T8l = T8j - T8k; T8r = T8j + T8k; } { E T87, T8a, T8g, T8h; T87 = FNMS(KP980785280, T86, KP195090322 * T85); T8a = FNMS(KP980785280, T89, KP195090322 * T88); T8b = T87 + T8a; T8v = T87 - T8a; T8g = T7C - T7D; T8h = T7g - T7j; T8i = T8g + T8h; T8u = T8g - T8h; } { E T8c, T8m, T8y, T8A; T8c = T84 - T8b; T8m = T8i - T8l; Ip[WS(rs, 11)] = FNMS(T8f, T8m, T81 * T8c); Im[WS(rs, 11)] = FMA(T8f, T8c, T81 * T8m); T8y = T8q + T8r; T8A = T8u - T8v; Ip[WS(rs, 15)] = FNMS(T8z, T8A, T8x * T8y); Im[WS(rs, 15)] = FMA(T8x, T8A, T8z * T8y); } { E T8n, T8o, T8s, T8w; T8n = T84 + T8b; T8o = T8i + T8l; Ip[WS(rs, 3)] = FNMS(T1j, T8o, T1f * T8n); Im[WS(rs, 3)] = FMA(T1j, T8n, T1f * T8o); T8s = T8q - T8r; T8w = T8u + T8v; Ip[WS(rs, 7)] = FNMS(T8t, T8w, T8p * T8s); Im[WS(rs, 7)] = FMA(T8p, T8w, T8t * T8s); } } { E T5u, T6u, T6n, T6v, T67, T6B, T6k, T6A; { E T5e, T5t, T6l, T6m; T5e = T56 - T5d; T5t = T5l - T5s; T5u = T5e + T5t; T6u = T5e - T5t; T6l = FMA(KP831469612, T5F, KP555570233 * T5M); T6m = FNMS(KP831469612, T5Y, KP555570233 * T65); T6n = T6l + T6m; T6v = T6m - T6l; } { E T5N, T66, T6g, T6j; T5N = FNMS(KP831469612, T5M, KP555570233 * T5F); T66 = FMA(KP555570233, T5Y, KP831469612 * T65); T67 = T5N + T66; T6B = T5N - T66; T6g = T6c - T6f; T6j = T6h - T6i; T6k = T6g + T6j; T6A = T6g - T6j; } { E T68, T6o, T6D, T6E; T68 = T5u - T67; T6o = T6k - T6n; Ip[WS(rs, 10)] = FNMS(T69, T6o, T53 * T68); Im[WS(rs, 10)] = FMA(T69, T68, T53 * T6o); T6D = T6u + T6v; T6E = T6A + T6B; Ip[WS(rs, 6)] = FNMS(T4c, T6E, T49 * T6D); Im[WS(rs, 6)] = FMA(T49, T6E, T4c * T6D); } { E T6p, T6q, T6w, T6C; T6p = T5u + T67; T6q = T6k + T6n; Ip[WS(rs, 2)] = FNMS(T4i, T6q, T4h * T6p); Im[WS(rs, 2)] = FMA(T4i, T6p, T4h * T6q); T6w = T6u - T6v; T6C = T6A - T6B; Ip[WS(rs, 14)] = FNMS(T6z, T6C, T6t * T6w); Im[WS(rs, 14)] = FMA(T6t, T6C, T6z * T6w); } } { E T7l, T7Q, T7L, T7R, T7A, T7V, T7I, T7U; { E T7d, T7k, T7J, T7K; T7d = T7b - T7c; T7k = T7g + T7j; T7l = T7d - T7k; T7Q = T7d + T7k; T7J = FNMS(KP555570233, T7v, KP831469612 * T7y); T7K = FMA(KP555570233, T7o, KP831469612 * T7r); T7L = T7J - T7K; T7R = T7J + T7K; } { E T7s, T7z, T7E, T7H; T7s = FNMS(KP555570233, T7r, KP831469612 * T7o); T7z = FMA(KP831469612, T7v, KP555570233 * T7y); T7A = T7s - T7z; T7V = T7z + T7s; T7E = T7C + T7D; T7H = T7F - T7G; T7I = T7E - T7H; T7U = T7E + T7H; } { E T7B, T7M, T7X, T7Y; T7B = T7l - T7A; T7M = T7I - T7L; Ip[WS(rs, 13)] = FNMS(T1k, T7M, T1g * T7B); Im[WS(rs, 13)] = FMA(T1k, T7B, T1g * T7M); T7X = T7Q + T7R; T7Y = T7U + T7V; Ip[WS(rs, 1)] = FNMS(T15, T7Y, T12 * T7X); Im[WS(rs, 1)] = FMA(T12, T7Y, T15 * T7X); } { E T7N, T7O, T7S, T7W; T7N = T7l + T7A; T7O = T7I + T7L; Ip[WS(rs, 5)] = FNMS(T2Y, T7O, T2X * T7N); Im[WS(rs, 5)] = FMA(T2Y, T7N, T2X * T7O); T7S = T7Q - T7R; T7W = T7U - T7V; Ip[WS(rs, 9)] = FNMS(T7T, T7W, T7P * T7S); Im[WS(rs, 9)] = FMA(T7P, T7W, T7T * T7S); } } } } } } static const tw_instr twinstr[] = { {TW_CEXP, 1, 1}, {TW_CEXP, 1, 3}, {TW_CEXP, 1, 9}, {TW_CEXP, 1, 27}, {TW_NEXT, 1, 0} }; static const hc2c_desc desc = { 32, "hc2cb2_32", twinstr, &GENUS, {376, 168, 112, 0} }; void X(codelet_hc2cb2_32) (planner *p) { X(khc2c_register) (p, hc2cb2_32, &desc, HC2C_VIA_RDFT); } #endif /* HAVE_FMA */ fftw-3.3.3/rdft/scalar/r2cb/hc2cb2_20.c0000644000175000017500000007524612054411033014162 00000000000000/* * Copyright (c) 2003, 2007-11 Matteo Frigo * Copyright (c) 2003, 2007-11 Massachusetts Institute of Technology * * This program is free software; you can redistribute it and/or modify * it under the terms of the GNU General Public License as published by * the Free Software Foundation; either version 2 of the License, or * (at your option) any later version. * * This program is distributed in the hope that it will be useful, * but WITHOUT ANY WARRANTY; without even the implied warranty of * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the * GNU General Public License for more details. * * You should have received a copy of the GNU General Public License * along with this program; if not, write to the Free Software * Foundation, Inc., 51 Franklin Street, Fifth Floor, Boston, MA 02110-1301 USA * */ /* This file was automatically generated --- DO NOT EDIT */ /* Generated on Sun Nov 25 07:42:02 EST 2012 */ #include "codelet-rdft.h" #ifdef HAVE_FMA /* Generated by: ../../../genfft/gen_hc2c.native -fma -reorder-insns -schedule-for-pipeline -compact -variables 4 -pipeline-latency 4 -sign 1 -twiddle-log3 -precompute-twiddles -n 20 -dif -name hc2cb2_20 -include hc2cb.h */ /* * This function contains 276 FP additions, 198 FP multiplications, * (or, 136 additions, 58 multiplications, 140 fused multiply/add), * 160 stack variables, 4 constants, and 80 memory accesses */ #include "hc2cb.h" static void hc2cb2_20(R *Rp, R *Ip, R *Rm, R *Im, const R *W, stride rs, INT mb, INT me, INT ms) { DK(KP951056516, +0.951056516295153572116439333379382143405698634); DK(KP559016994, +0.559016994374947424102293417182819058860154590); DK(KP618033988, +0.618033988749894848204586834365638117720309180); DK(KP250000000, +0.250000000000000000000000000000000000000000000); { INT m; for (m = mb, W = W + ((mb - 1) * 8); m < me; m = m + 1, Rp = Rp + ms, Ip = Ip + ms, Rm = Rm - ms, Im = Im - ms, W = W + 8, MAKE_VOLATILE_STRIDE(80, rs)) { E T1S, T1O, T1s, TI, T24, T1Y, T2g, T2k, TS, TR, T1I, T26, T1o, T20, T1F; E T25, TT, T1Z; { E TD, TH, TE, T1L, T1N, T1X, TG, T1V, T2Y, T2b, T29, T2s, T36, T3e, T31; E T2o, T3b, T5b, T2c, T2U, T4y, T4u, T2f, T5g, T47, T5p, T4b, T5l; { E T1r, TF, T2T, T1M, T1R, T2X, T2r, T4x; TD = W[0]; TH = W[3]; TE = W[2]; T1L = W[6]; T1N = W[7]; T1r = TD * TH; TF = TD * TE; T2T = TE * T1L; T1M = TD * T1L; T1R = TD * T1N; T2X = TE * T1N; T1X = W[5]; TG = W[1]; T1V = W[4]; T2Y = FNMS(TH, T1L, T2X); T2r = TD * T1X; { E T23, T2n, T1W, T2a; T23 = TE * T1X; T1S = FNMS(TG, T1L, T1R); T1O = FMA(TG, T1N, T1M); T2b = FMA(TG, TE, T1r); T1s = FNMS(TG, TE, T1r); T29 = FNMS(TG, TH, TF); TI = FMA(TG, TH, TF); T2n = TD * T1V; T1W = TE * T1V; T2s = FMA(TG, T1V, T2r); T36 = FNMS(TG, T1V, T2r); T3e = FMA(TH, T1V, T23); T24 = FNMS(TH, T1V, T23); T2a = T29 * T1V; T31 = FMA(TG, T1X, T2n); T2o = FNMS(TG, T1X, T2n); T3b = FNMS(TH, T1X, T1W); T1Y = FMA(TH, T1X, T1W); T5b = FNMS(T2b, T1X, T2a); T2c = FMA(T2b, T1X, T2a); T2U = FMA(TH, T1N, T2T); } T4x = T29 * T1N; { E T4t, T2d, T2j, T2e; T4t = T29 * T1L; T2e = T29 * T1X; T4y = FNMS(T2b, T1L, T4x); T4u = FMA(T2b, T1N, T4t); T2f = FNMS(T2b, T1V, T2e); T5g = FMA(T2b, T1V, T2e); T2d = T2c * T1L; T2j = T2c * T1N; T47 = TI * T1V; T2g = FMA(T2f, T1N, T2d); T2k = FNMS(T2f, T1L, T2j); T5p = TI * T1N; T4b = TI * T1X; T5l = TI * T1L; } } { E T4f, T48, T4c, T4k, T5m, T5q, T3V, T4V, TJ, T7, T3j, T4B, T2H, T1z, T3q; E T43, T1n, T52, T42, T3x, T53, T2D, T18, T2A, T1H, T4R, T4X, T4W, T4O, T1G; E T2O, T3I, T2P, T3P, T2K, T2M, T1C, T1E, TC, T2w, T40, T3Y, T4K, T4I, TQ; { E T3h, T3, T1w, T3T, T1v, T3U, T6, T1x; { E T1t, T1u, T1, T2, T4, T5; T1 = Rp[0]; T2 = Rm[WS(rs, 9)]; T1t = Ip[0]; T4f = FNMS(T1s, T1X, T47); T48 = FMA(T1s, T1X, T47); T4c = FNMS(T1s, T1V, T4b); T4k = FMA(T1s, T1V, T4b); T5m = FMA(T1s, T1N, T5l); T5q = FNMS(T1s, T1L, T5p); T3h = T1 - T2; T3 = T1 + T2; T1u = Im[WS(rs, 9)]; T4 = Rp[WS(rs, 5)]; T5 = Rm[WS(rs, 4)]; T1w = Ip[WS(rs, 5)]; T3T = T1t + T1u; T1v = T1t - T1u; T3U = T4 - T5; T6 = T4 + T5; T1x = Im[WS(rs, 4)]; } { E T3L, T4M, TK, Te, T3m, T4C, T2y, T1f, T3H, T4Q, TO, TA, T3w, T4G, T2C; E T17, T3O, T4N, TL, Tl, T3p, T4D, T2z, T1m, T3r, Tp, TX, T3C, TW, T3D; E Ts, TY; { E T3u, Tw, T14, T3G, T13, T3F, Tz, T15; { E T3k, Ta, T1c, T3J, T1b, T3K, Td, T1d; { E T19, T1a, Tb, Tc; { E T8, T3i, T1y, T9; T8 = Rp[WS(rs, 4)]; T3V = T3T - T3U; T4V = T3U + T3T; TJ = T3 - T6; T7 = T3 + T6; T3i = T1w + T1x; T1y = T1w - T1x; T9 = Rm[WS(rs, 5)]; T19 = Ip[WS(rs, 4)]; T3j = T3h + T3i; T4B = T3h - T3i; T2H = T1v + T1y; T1z = T1v - T1y; T3k = T8 - T9; Ta = T8 + T9; T1a = Im[WS(rs, 5)]; } Tb = Rp[WS(rs, 9)]; Tc = Rm[0]; T1c = Ip[WS(rs, 9)]; T3J = T19 + T1a; T1b = T19 - T1a; T3K = Tb - Tc; Td = Tb + Tc; T1d = Im[0]; } { E T11, T12, Tx, Ty; { E Tu, T3l, T1e, Tv; Tu = Rm[WS(rs, 7)]; T3L = T3J - T3K; T4M = T3K + T3J; TK = Ta - Td; Te = Ta + Td; T3l = T1c + T1d; T1e = T1c - T1d; Tv = Rp[WS(rs, 2)]; T11 = Ip[WS(rs, 2)]; T3m = T3k + T3l; T4C = T3k - T3l; T2y = T1b + T1e; T1f = T1b - T1e; T3u = Tu - Tv; Tw = Tu + Tv; T12 = Im[WS(rs, 7)]; } Tx = Rm[WS(rs, 2)]; Ty = Rp[WS(rs, 7)]; T14 = Ip[WS(rs, 7)]; T3G = T11 + T12; T13 = T11 - T12; T3F = Tx - Ty; Tz = Tx + Ty; T15 = Im[WS(rs, 2)]; } } { E T3n, Th, T1j, T3N, T1i, T3M, Tk, T1k; { E T1g, T1h, Ti, Tj; { E Tf, T3v, T16, Tg; Tf = Rm[WS(rs, 3)]; T3H = T3F + T3G; T4Q = T3F - T3G; TO = Tw - Tz; TA = Tw + Tz; T3v = T14 + T15; T16 = T14 - T15; Tg = Rp[WS(rs, 6)]; T1g = Ip[WS(rs, 6)]; T3w = T3u - T3v; T4G = T3u + T3v; T2C = T13 + T16; T17 = T13 - T16; T3n = Tf - Tg; Th = Tf + Tg; T1h = Im[WS(rs, 3)]; } Ti = Rp[WS(rs, 1)]; Tj = Rm[WS(rs, 8)]; T1j = Ip[WS(rs, 1)]; T3N = T1g + T1h; T1i = T1g - T1h; T3M = Ti - Tj; Tk = Ti + Tj; T1k = Im[WS(rs, 8)]; } { E TU, TV, Tq, Tr; { E Tn, T3o, T1l, To; Tn = Rp[WS(rs, 8)]; T3O = T3M + T3N; T4N = T3M - T3N; TL = Th - Tk; Tl = Th + Tk; T3o = T1j + T1k; T1l = T1j - T1k; To = Rm[WS(rs, 1)]; TU = Ip[WS(rs, 8)]; T3p = T3n + T3o; T4D = T3n - T3o; T2z = T1i + T1l; T1m = T1i - T1l; T3r = Tn - To; Tp = Tn + To; TV = Im[WS(rs, 1)]; } Tq = Rm[WS(rs, 6)]; Tr = Rp[WS(rs, 3)]; TX = Ip[WS(rs, 3)]; T3C = TU + TV; TW = TU - TV; T3D = Tq - Tr; Ts = Tq + Tr; TY = Im[WS(rs, 6)]; } } } { E T3E, Tt, T1A, T4E, T4H, T2J, T1B, T2I, TM, TP; { E T4P, TN, T3s, TZ; T3q = T3m + T3p; T43 = T3m - T3p; T3E = T3C - T3D; T4P = T3D + T3C; TN = Tp - Ts; Tt = Tp + Ts; T3s = TX + TY; TZ = TX - TY; T1n = T1f - T1m; T1A = T1f + T1m; T4E = T4C + T4D; T52 = T4C - T4D; { E T3t, T4F, T2B, T10; T3t = T3r - T3s; T4F = T3r + T3s; T2B = TW + TZ; T10 = TW - TZ; T42 = T3t - T3w; T3x = T3t + T3w; T4H = T4F + T4G; T53 = T4F - T4G; T2D = T2B - T2C; T2J = T2B + T2C; T1B = T10 + T17; T18 = T10 - T17; T2A = T2y - T2z; T2I = T2y + T2z; TM = TK + TL; T1H = TK - TL; } T4R = T4P - T4Q; T4X = T4P + T4Q; T4W = T4M + T4N; T4O = T4M - T4N; T1G = TN - TO; TP = TN + TO; } { E Tm, T3X, TB, T3W; Tm = Te + Tl; T2O = Te - Tl; T3I = T3E + T3H; T3X = T3E - T3H; TB = Tt + TA; T2P = Tt - TA; T3P = T3L + T3O; T3W = T3L - T3O; T2K = T2I + T2J; T2M = T2I - T2J; T1C = T1A + T1B; T1E = T1A - T1B; TC = Tm + TB; T2w = Tm - TB; T40 = T3W - T3X; T3Y = T3W + T3X; T4K = T4E - T4H; T4I = T4E + T4H; TS = TM - TP; TQ = TM + TP; } } } } { E T3A, T3y, T50, T1D, T2t, T2p, T4J, T5t, T5v, T4Z, T4Y; Rp[0] = T7 + TC; T3A = T3q - T3x; T3y = T3q + T3x; T50 = T4W - T4X; T4Y = T4W + T4X; Rm[0] = T2H + T2K; T1D = FNMS(KP250000000, T1C, T1z); T2t = T1z + T1C; T2p = TJ + TQ; TR = FNMS(KP250000000, TQ, TJ); T4J = FNMS(KP250000000, T4I, T4B); T5t = T4B + T4I; T5v = T4V + T4Y; T4Z = FNMS(KP250000000, T4Y, T4V); { E T4m, T44, T4i, T4p, T49, T3R, T4j, T4a, T3S, T4l, T41, T4q; { E T3z, T4v, T4w, T3Z, T4z; T3z = FNMS(KP250000000, T3y, T3j); T4v = T3j + T3y; { E T2u, T2q, T5u, T5w; T2u = T2s * T2p; T2q = T2o * T2p; T5u = T2c * T5t; T5w = T2c * T5v; Rm[WS(rs, 5)] = FMA(T2o, T2t, T2u); Rp[WS(rs, 5)] = FNMS(T2s, T2t, T2q); Ip[WS(rs, 2)] = FNMS(T2f, T5v, T5u); Im[WS(rs, 2)] = FMA(T2f, T5t, T5w); T4w = T4u * T4v; } T3Z = FNMS(KP250000000, T3Y, T3V); T4z = T3V + T3Y; { E T3Q, T4h, T4A, T4g, T3B; T3Q = FNMS(KP618033988, T3P, T3I); T4h = FMA(KP618033988, T3I, T3P); Ip[WS(rs, 7)] = FNMS(T4y, T4z, T4w); T4A = T4u * T4z; T4m = FMA(KP618033988, T42, T43); T44 = FNMS(KP618033988, T43, T42); T4g = FMA(KP559016994, T3A, T3z); T3B = FNMS(KP559016994, T3A, T3z); Im[WS(rs, 7)] = FMA(T4y, T4v, T4A); T4i = FNMS(KP951056516, T4h, T4g); T4p = FMA(KP951056516, T4h, T4g); T49 = FMA(KP951056516, T3Q, T3B); T3R = FNMS(KP951056516, T3Q, T3B); } T4j = T4f * T4i; T4a = T48 * T49; T3S = TE * T3R; T4l = FMA(KP559016994, T40, T3Z); T41 = FNMS(KP559016994, T40, T3Z); T4q = T1L * T4p; } { E T5d, T4S, T54, T5i, T4L, T5c; T5d = FNMS(KP618033988, T4O, T4R); T4S = FMA(KP618033988, T4R, T4O); { E T4n, T4r, T4d, T45; T4n = FMA(KP951056516, T4m, T4l); T4r = FNMS(KP951056516, T4m, T4l); T4d = FNMS(KP951056516, T44, T41); T45 = FMA(KP951056516, T44, T41); { E T4o, T4s, T4e, T46; T4o = T4f * T4n; Ip[WS(rs, 5)] = FNMS(T4k, T4n, T4j); T4s = T1L * T4r; Ip[WS(rs, 9)] = FNMS(T1N, T4r, T4q); T4e = T48 * T4d; Ip[WS(rs, 3)] = FNMS(T4c, T4d, T4a); T46 = TE * T45; Ip[WS(rs, 1)] = FNMS(TH, T45, T3S); Im[WS(rs, 5)] = FMA(T4k, T4i, T4o); Im[WS(rs, 9)] = FMA(T1N, T4p, T4s); Im[WS(rs, 3)] = FMA(T4c, T49, T4e); Im[WS(rs, 1)] = FMA(TH, T3R, T46); } } T54 = FMA(KP618033988, T53, T52); T5i = FNMS(KP618033988, T52, T53); T4L = FMA(KP559016994, T4K, T4J); T5c = FNMS(KP559016994, T4K, T4J); { E T38, T2Q, T33, T2E, T2v, T37, T2N, T5h, T51, T2L, T2x, T32; T38 = FNMS(KP618033988, T2O, T2P); T2Q = FMA(KP618033988, T2P, T2O); T5h = FNMS(KP559016994, T50, T4Z); T51 = FMA(KP559016994, T50, T4Z); { E T5e, T5n, T57, T4T; T5e = FNMS(KP951056516, T5d, T5c); T5n = FMA(KP951056516, T5d, T5c); T57 = FMA(KP951056516, T4S, T4L); T4T = FNMS(KP951056516, T4S, T4L); { E T5j, T5r, T59, T55; T5j = FMA(KP951056516, T5i, T5h); T5r = FNMS(KP951056516, T5i, T5h); T59 = FNMS(KP951056516, T54, T51); T55 = FMA(KP951056516, T54, T51); { E T5f, T5o, T58, T4U; T5f = T5b * T5e; T5o = T5m * T5n; T58 = T1V * T57; T4U = TD * T4T; { E T5k, T5s, T5a, T56; T5k = T5b * T5j; T5s = T5m * T5r; T5a = T1V * T59; T56 = TD * T55; Ip[WS(rs, 6)] = FNMS(T5g, T5j, T5f); Ip[WS(rs, 8)] = FNMS(T5q, T5r, T5o); Ip[WS(rs, 4)] = FNMS(T1X, T59, T58); Ip[0] = FNMS(TG, T55, T4U); Im[WS(rs, 6)] = FMA(T5g, T5e, T5k); Im[WS(rs, 8)] = FMA(T5q, T5n, T5s); Im[WS(rs, 4)] = FMA(T1X, T57, T5a); Im[0] = FMA(TG, T4T, T56); } } } } T2L = FNMS(KP250000000, T2K, T2H); T33 = FNMS(KP618033988, T2A, T2D); T2E = FMA(KP618033988, T2D, T2A); T2v = FNMS(KP250000000, TC, T7); T37 = FNMS(KP559016994, T2M, T2L); T2N = FMA(KP559016994, T2M, T2L); T1I = FNMS(KP618033988, T1H, T1G); T26 = FMA(KP618033988, T1G, T1H); T2x = FMA(KP559016994, T2w, T2v); T32 = FNMS(KP559016994, T2w, T2v); { E T3f, T39, T2R, T2Z; T3f = FNMS(KP951056516, T38, T37); T39 = FMA(KP951056516, T38, T37); T2R = FNMS(KP951056516, T2Q, T2N); T2Z = FMA(KP951056516, T2Q, T2N); { E T3c, T34, T2F, T2V; T3c = FMA(KP951056516, T33, T32); T34 = FNMS(KP951056516, T33, T32); T2F = FMA(KP951056516, T2E, T2x); T2V = FNMS(KP951056516, T2E, T2x); { E T3a, T35, T3g, T3d; T3a = T36 * T34; T35 = T31 * T34; T3g = T3e * T3c; T3d = T3b * T3c; { E T30, T2W, T2S, T2G; T30 = T2Y * T2V; T2W = T2U * T2V; T2S = T2b * T2F; T2G = T29 * T2F; Rm[WS(rs, 4)] = FMA(T31, T39, T3a); Rp[WS(rs, 4)] = FNMS(T36, T39, T35); Rm[WS(rs, 6)] = FMA(T3b, T3f, T3g); Rp[WS(rs, 6)] = FNMS(T3e, T3f, T3d); Rm[WS(rs, 8)] = FMA(T2U, T2Z, T30); Rp[WS(rs, 8)] = FNMS(T2Y, T2Z, T2W); Rm[WS(rs, 2)] = FMA(T29, T2R, T2S); Rp[WS(rs, 2)] = FNMS(T2b, T2R, T2G); } } } } T1o = FNMS(KP618033988, T1n, T18); T20 = FMA(KP618033988, T18, T1n); T1F = FNMS(KP559016994, T1E, T1D); T25 = FMA(KP559016994, T1E, T1D); } } } } } } TT = FNMS(KP559016994, TS, TR); T1Z = FMA(KP559016994, TS, TR); { E T2l, T27, T1J, T1T; T2l = FNMS(KP951056516, T26, T25); T27 = FMA(KP951056516, T26, T25); T1J = FNMS(KP951056516, T1I, T1F); T1T = FMA(KP951056516, T1I, T1F); { E T2h, T21, T1p, T1P; T2h = FMA(KP951056516, T20, T1Z); T21 = FNMS(KP951056516, T20, T1Z); T1p = FMA(KP951056516, T1o, TT); T1P = FNMS(KP951056516, T1o, TT); { E T28, T22, T2m, T2i; T28 = T24 * T21; T22 = T1Y * T21; T2m = T2k * T2h; T2i = T2g * T2h; { E T1U, T1Q, T1K, T1q; T1U = T1S * T1P; T1Q = T1O * T1P; T1K = T1s * T1p; T1q = TI * T1p; Rm[WS(rs, 3)] = FMA(T1Y, T27, T28); Rp[WS(rs, 3)] = FNMS(T24, T27, T22); Rm[WS(rs, 7)] = FMA(T2g, T2l, T2m); Rp[WS(rs, 7)] = FNMS(T2k, T2l, T2i); Rm[WS(rs, 9)] = FMA(T1O, T1T, T1U); Rp[WS(rs, 9)] = FNMS(T1S, T1T, T1Q); Rm[WS(rs, 1)] = FMA(TI, T1J, T1K); Rp[WS(rs, 1)] = FNMS(T1s, T1J, T1q); } } } } } } } static const tw_instr twinstr[] = { {TW_CEXP, 1, 1}, {TW_CEXP, 1, 3}, {TW_CEXP, 1, 9}, {TW_CEXP, 1, 19}, {TW_NEXT, 1, 0} }; static const hc2c_desc desc = { 20, "hc2cb2_20", twinstr, &GENUS, {136, 58, 140, 0} }; void X(codelet_hc2cb2_20) (planner *p) { X(khc2c_register) (p, hc2cb2_20, &desc, HC2C_VIA_RDFT); } #else /* HAVE_FMA */ /* Generated by: ../../../genfft/gen_hc2c.native -compact -variables 4 -pipeline-latency 4 -sign 1 -twiddle-log3 -precompute-twiddles -n 20 -dif -name hc2cb2_20 -include hc2cb.h */ /* * This function contains 276 FP additions, 164 FP multiplications, * (or, 204 additions, 92 multiplications, 72 fused multiply/add), * 137 stack variables, 4 constants, and 80 memory accesses */ #include "hc2cb.h" static void hc2cb2_20(R *Rp, R *Ip, R *Rm, R *Im, const R *W, stride rs, INT mb, INT me, INT ms) { DK(KP250000000, +0.250000000000000000000000000000000000000000000); DK(KP559016994, +0.559016994374947424102293417182819058860154590); DK(KP587785252, +0.587785252292473129168705954639072768597652438); DK(KP951056516, +0.951056516295153572116439333379382143405698634); { INT m; for (m = mb, W = W + ((mb - 1) * 8); m < me; m = m + 1, Rp = Rp + ms, Ip = Ip + ms, Rm = Rm - ms, Im = Im - ms, W = W + 8, MAKE_VOLATILE_STRIDE(80, rs)) { E TD, TG, TE, TH, TJ, T1t, T27, T25, T1T, T1R, T1V, T2j, T2Z, T21, T2X; E T2T, T2n, T2P, T3V, T41, T3R, T3X, T29, T2c, T4H, T4L, T1L, T1M, T1N, T2d; E T4R, T1P, T4P, T49, T2N, T2f, T47, T2L; { E T1U, T2l, T1Z, T2i, T1S, T2m, T20, T2h; { E TF, T1s, TI, T1r; TD = W[0]; TG = W[1]; TE = W[2]; TH = W[3]; TF = TD * TE; T1s = TG * TE; TI = TG * TH; T1r = TD * TH; TJ = TF + TI; T1t = T1r - T1s; T27 = T1r + T1s; T25 = TF - TI; T1T = W[5]; T1U = TH * T1T; T2l = TD * T1T; T1Z = TE * T1T; T2i = TG * T1T; T1R = W[4]; T1S = TE * T1R; T2m = TG * T1R; T20 = TH * T1R; T2h = TD * T1R; } T1V = T1S + T1U; T2j = T2h - T2i; T2Z = T1Z + T20; T21 = T1Z - T20; T2X = T1S - T1U; T2T = T2l - T2m; T2n = T2l + T2m; T2P = T2h + T2i; { E T3T, T3U, T3P, T3Q; T3T = TJ * T1T; T3U = T1t * T1R; T3V = T3T - T3U; T41 = T3T + T3U; T3P = TJ * T1R; T3Q = T1t * T1T; T3R = T3P + T3Q; T3X = T3P - T3Q; { E T26, T28, T2a, T2b; T26 = T25 * T1R; T28 = T27 * T1T; T29 = T26 + T28; T2a = T25 * T1T; T2b = T27 * T1R; T2c = T2a - T2b; T4H = T26 - T28; T4L = T2a + T2b; T1L = W[6]; T1M = W[7]; T1N = FMA(TD, T1L, TG * T1M); T2d = FMA(T29, T1L, T2c * T1M); T4R = FNMS(T1t, T1L, TJ * T1M); T1P = FNMS(TG, T1L, TD * T1M); T4P = FMA(TJ, T1L, T1t * T1M); T49 = FNMS(T27, T1L, T25 * T1M); T2N = FNMS(TH, T1L, TE * T1M); T2f = FNMS(T2c, T1L, T29 * T1M); T47 = FMA(T25, T1L, T27 * T1M); T2L = FMA(TE, T1L, TH * T1M); } } } { E T7, T4i, T4x, TK, T1D, T3i, T3E, T2D, T19, T3L, T3M, T1o, T2x, T4C, T4B; E T2u, T1v, T4r, T4o, T1u, T2H, T37, T2I, T3e, T3p, T3w, T3x, Tm, TB, TC; E T4u, T4v, T4y, T2A, T2B, T2E, T1E, T1F, T1G, T4d, T4g, T4j, T3F, T3G, T3H; E TN, TQ, TR, T48, T4a; { E T3, T3g, T1z, T3C, T6, T3D, T1C, T3h; { E T1, T2, T1x, T1y; T1 = Rp[0]; T2 = Rm[WS(rs, 9)]; T3 = T1 + T2; T3g = T1 - T2; T1x = Ip[0]; T1y = Im[WS(rs, 9)]; T1z = T1x - T1y; T3C = T1x + T1y; } { E T4, T5, T1A, T1B; T4 = Rp[WS(rs, 5)]; T5 = Rm[WS(rs, 4)]; T6 = T4 + T5; T3D = T4 - T5; T1A = Ip[WS(rs, 5)]; T1B = Im[WS(rs, 4)]; T1C = T1A - T1B; T3h = T1A + T1B; } T7 = T3 + T6; T4i = T3g - T3h; T4x = T3D + T3C; TK = T3 - T6; T1D = T1z - T1C; T3i = T3g + T3h; T3E = T3C - T3D; T2D = T1z + T1C; } { E Te, T4b, T4m, TL, T11, T33, T3l, T2s, TA, T4f, T4q, TP, T1n, T3d, T3v; E T2w, Tl, T4c, T4n, TM, T18, T36, T3o, T2t, Tt, T4e, T4p, TO, T1g, T3a; E T3s, T2v; { E Ta, T3j, TX, T31, Td, T32, T10, T3k; { E T8, T9, TV, TW; T8 = Rp[WS(rs, 4)]; T9 = Rm[WS(rs, 5)]; Ta = T8 + T9; T3j = T8 - T9; TV = Ip[WS(rs, 4)]; TW = Im[WS(rs, 5)]; TX = TV - TW; T31 = TV + TW; } { E Tb, Tc, TY, TZ; Tb = Rp[WS(rs, 9)]; Tc = Rm[0]; Td = Tb + Tc; T32 = Tb - Tc; TY = Ip[WS(rs, 9)]; TZ = Im[0]; T10 = TY - TZ; T3k = TY + TZ; } Te = Ta + Td; T4b = T3j - T3k; T4m = T32 + T31; TL = Ta - Td; T11 = TX - T10; T33 = T31 - T32; T3l = T3j + T3k; T2s = TX + T10; } { E Tw, T3t, T1j, T3c, Tz, T3b, T1m, T3u; { E Tu, Tv, T1h, T1i; Tu = Rm[WS(rs, 7)]; Tv = Rp[WS(rs, 2)]; Tw = Tu + Tv; T3t = Tu - Tv; T1h = Ip[WS(rs, 2)]; T1i = Im[WS(rs, 7)]; T1j = T1h - T1i; T3c = T1h + T1i; } { E Tx, Ty, T1k, T1l; Tx = Rm[WS(rs, 2)]; Ty = Rp[WS(rs, 7)]; Tz = Tx + Ty; T3b = Tx - Ty; T1k = Ip[WS(rs, 7)]; T1l = Im[WS(rs, 2)]; T1m = T1k - T1l; T3u = T1k + T1l; } TA = Tw + Tz; T4f = T3t + T3u; T4q = T3b - T3c; TP = Tw - Tz; T1n = T1j - T1m; T3d = T3b + T3c; T3v = T3t - T3u; T2w = T1j + T1m; } { E Th, T3m, T14, T35, Tk, T34, T17, T3n; { E Tf, Tg, T12, T13; Tf = Rm[WS(rs, 3)]; Tg = Rp[WS(rs, 6)]; Th = Tf + Tg; T3m = Tf - Tg; T12 = Ip[WS(rs, 6)]; T13 = Im[WS(rs, 3)]; T14 = T12 - T13; T35 = T12 + T13; } { E Ti, Tj, T15, T16; Ti = Rp[WS(rs, 1)]; Tj = Rm[WS(rs, 8)]; Tk = Ti + Tj; T34 = Ti - Tj; T15 = Ip[WS(rs, 1)]; T16 = Im[WS(rs, 8)]; T17 = T15 - T16; T3n = T15 + T16; } Tl = Th + Tk; T4c = T3m - T3n; T4n = T34 - T35; TM = Th - Tk; T18 = T14 - T17; T36 = T34 + T35; T3o = T3m + T3n; T2t = T14 + T17; } { E Tp, T3q, T1c, T38, Ts, T39, T1f, T3r; { E Tn, To, T1a, T1b; Tn = Rp[WS(rs, 8)]; To = Rm[WS(rs, 1)]; Tp = Tn + To; T3q = Tn - To; T1a = Ip[WS(rs, 8)]; T1b = Im[WS(rs, 1)]; T1c = T1a - T1b; T38 = T1a + T1b; } { E Tq, Tr, T1d, T1e; Tq = Rm[WS(rs, 6)]; Tr = Rp[WS(rs, 3)]; Ts = Tq + Tr; T39 = Tq - Tr; T1d = Ip[WS(rs, 3)]; T1e = Im[WS(rs, 6)]; T1f = T1d - T1e; T3r = T1d + T1e; } Tt = Tp + Ts; T4e = T3q + T3r; T4p = T39 + T38; TO = Tp - Ts; T1g = T1c - T1f; T3a = T38 - T39; T3s = T3q - T3r; T2v = T1c + T1f; } T19 = T11 - T18; T3L = T3l - T3o; T3M = T3s - T3v; T1o = T1g - T1n; T2x = T2v - T2w; T4C = T4e - T4f; T4B = T4b - T4c; T2u = T2s - T2t; T1v = TO - TP; T4r = T4p - T4q; T4o = T4m - T4n; T1u = TL - TM; T2H = Te - Tl; T37 = T33 + T36; T2I = Tt - TA; T3e = T3a + T3d; T3p = T3l + T3o; T3w = T3s + T3v; T3x = T3p + T3w; Tm = Te + Tl; TB = Tt + TA; TC = Tm + TB; T4u = T4m + T4n; T4v = T4p + T4q; T4y = T4u + T4v; T2A = T2s + T2t; T2B = T2v + T2w; T2E = T2A + T2B; T1E = T11 + T18; T1F = T1g + T1n; T1G = T1E + T1F; T4d = T4b + T4c; T4g = T4e + T4f; T4j = T4d + T4g; T3F = T33 - T36; T3G = T3a - T3d; T3H = T3F + T3G; TN = TL + TM; TQ = TO + TP; TR = TN + TQ; } Rp[0] = T7 + TC; Rm[0] = T2D + T2E; { E T2k, T2o, T4T, T4U; T2k = TK + TR; T2o = T1D + T1G; Rp[WS(rs, 5)] = FNMS(T2n, T2o, T2j * T2k); Rm[WS(rs, 5)] = FMA(T2n, T2k, T2j * T2o); T4T = T4i + T4j; T4U = T4x + T4y; Ip[WS(rs, 2)] = FNMS(T2c, T4U, T29 * T4T); Im[WS(rs, 2)] = FMA(T29, T4U, T2c * T4T); } T48 = T3i + T3x; T4a = T3E + T3H; Ip[WS(rs, 7)] = FNMS(T49, T4a, T47 * T48); Im[WS(rs, 7)] = FMA(T47, T4a, T49 * T48); { E T2y, T2J, T2V, T2R, T2G, T2U, T2r, T2Q; T2y = FMA(KP951056516, T2u, KP587785252 * T2x); T2J = FMA(KP951056516, T2H, KP587785252 * T2I); T2V = FNMS(KP951056516, T2I, KP587785252 * T2H); T2R = FNMS(KP951056516, T2x, KP587785252 * T2u); { E T2C, T2F, T2p, T2q; T2C = KP559016994 * (T2A - T2B); T2F = FNMS(KP250000000, T2E, T2D); T2G = T2C + T2F; T2U = T2F - T2C; T2p = KP559016994 * (Tm - TB); T2q = FNMS(KP250000000, TC, T7); T2r = T2p + T2q; T2Q = T2q - T2p; } { E T2z, T2K, T2Y, T30; T2z = T2r + T2y; T2K = T2G - T2J; Rp[WS(rs, 2)] = FNMS(T27, T2K, T25 * T2z); Rm[WS(rs, 2)] = FMA(T27, T2z, T25 * T2K); T2Y = T2Q - T2R; T30 = T2V + T2U; Rp[WS(rs, 6)] = FNMS(T2Z, T30, T2X * T2Y); Rm[WS(rs, 6)] = FMA(T2Z, T2Y, T2X * T30); } { E T2M, T2O, T2S, T2W; T2M = T2r - T2y; T2O = T2J + T2G; Rp[WS(rs, 8)] = FNMS(T2N, T2O, T2L * T2M); Rm[WS(rs, 8)] = FMA(T2N, T2M, T2L * T2O); T2S = T2Q + T2R; T2W = T2U - T2V; Rp[WS(rs, 4)] = FNMS(T2T, T2W, T2P * T2S); Rm[WS(rs, 4)] = FMA(T2T, T2S, T2P * T2W); } } { E T4s, T4D, T4N, T4I, T4A, T4M, T4l, T4J; T4s = FMA(KP951056516, T4o, KP587785252 * T4r); T4D = FMA(KP951056516, T4B, KP587785252 * T4C); T4N = FNMS(KP951056516, T4C, KP587785252 * T4B); T4I = FNMS(KP951056516, T4r, KP587785252 * T4o); { E T4w, T4z, T4h, T4k; T4w = KP559016994 * (T4u - T4v); T4z = FNMS(KP250000000, T4y, T4x); T4A = T4w + T4z; T4M = T4z - T4w; T4h = KP559016994 * (T4d - T4g); T4k = FNMS(KP250000000, T4j, T4i); T4l = T4h + T4k; T4J = T4k - T4h; } { E T4t, T4E, T4Q, T4S; T4t = T4l - T4s; T4E = T4A + T4D; Ip[0] = FNMS(TG, T4E, TD * T4t); Im[0] = FMA(TD, T4E, TG * T4t); T4Q = T4J - T4I; T4S = T4M + T4N; Ip[WS(rs, 8)] = FNMS(T4R, T4S, T4P * T4Q); Im[WS(rs, 8)] = FMA(T4P, T4S, T4R * T4Q); } { E T4F, T4G, T4K, T4O; T4F = T4s + T4l; T4G = T4A - T4D; Ip[WS(rs, 4)] = FNMS(T1T, T4G, T1R * T4F); Im[WS(rs, 4)] = FMA(T1R, T4G, T1T * T4F); T4K = T4I + T4J; T4O = T4M - T4N; Ip[WS(rs, 6)] = FNMS(T4L, T4O, T4H * T4K); Im[WS(rs, 6)] = FMA(T4H, T4O, T4L * T4K); } } { E T1p, T1w, T22, T1X, T1J, T23, TU, T1W; T1p = FNMS(KP951056516, T1o, KP587785252 * T19); T1w = FNMS(KP951056516, T1v, KP587785252 * T1u); T22 = FMA(KP951056516, T1u, KP587785252 * T1v); T1X = FMA(KP951056516, T19, KP587785252 * T1o); { E T1H, T1I, TS, TT; T1H = FNMS(KP250000000, T1G, T1D); T1I = KP559016994 * (T1E - T1F); T1J = T1H - T1I; T23 = T1I + T1H; TS = FNMS(KP250000000, TR, TK); TT = KP559016994 * (TN - TQ); TU = TS - TT; T1W = TT + TS; } { E T1q, T1K, T2e, T2g; T1q = TU - T1p; T1K = T1w + T1J; Rp[WS(rs, 1)] = FNMS(T1t, T1K, TJ * T1q); Rm[WS(rs, 1)] = FMA(T1t, T1q, TJ * T1K); T2e = T1W + T1X; T2g = T23 - T22; Rp[WS(rs, 7)] = FNMS(T2f, T2g, T2d * T2e); Rm[WS(rs, 7)] = FMA(T2f, T2e, T2d * T2g); } { E T1O, T1Q, T1Y, T24; T1O = TU + T1p; T1Q = T1J - T1w; Rp[WS(rs, 9)] = FNMS(T1P, T1Q, T1N * T1O); Rm[WS(rs, 9)] = FMA(T1P, T1O, T1N * T1Q); T1Y = T1W - T1X; T24 = T22 + T23; Rp[WS(rs, 3)] = FNMS(T21, T24, T1V * T1Y); Rm[WS(rs, 3)] = FMA(T21, T1Y, T1V * T24); } } { E T3f, T3N, T43, T3Z, T3K, T42, T3A, T3Y; T3f = FNMS(KP951056516, T3e, KP587785252 * T37); T3N = FNMS(KP951056516, T3M, KP587785252 * T3L); T43 = FMA(KP951056516, T3L, KP587785252 * T3M); T3Z = FMA(KP951056516, T37, KP587785252 * T3e); { E T3I, T3J, T3y, T3z; T3I = FNMS(KP250000000, T3H, T3E); T3J = KP559016994 * (T3F - T3G); T3K = T3I - T3J; T42 = T3J + T3I; T3y = FNMS(KP250000000, T3x, T3i); T3z = KP559016994 * (T3p - T3w); T3A = T3y - T3z; T3Y = T3z + T3y; } { E T3B, T3O, T45, T46; T3B = T3f + T3A; T3O = T3K - T3N; Ip[WS(rs, 1)] = FNMS(TH, T3O, TE * T3B); Im[WS(rs, 1)] = FMA(TE, T3O, TH * T3B); T45 = T3Z + T3Y; T46 = T42 - T43; Ip[WS(rs, 9)] = FNMS(T1M, T46, T1L * T45); Im[WS(rs, 9)] = FMA(T1L, T46, T1M * T45); } { E T3S, T3W, T40, T44; T3S = T3A - T3f; T3W = T3K + T3N; Ip[WS(rs, 3)] = FNMS(T3V, T3W, T3R * T3S); Im[WS(rs, 3)] = FMA(T3R, T3W, T3V * T3S); T40 = T3Y - T3Z; T44 = T42 + T43; Ip[WS(rs, 5)] = FNMS(T41, T44, T3X * T40); Im[WS(rs, 5)] = FMA(T3X, T44, T41 * T40); } } } } } } static const tw_instr twinstr[] = { {TW_CEXP, 1, 1}, {TW_CEXP, 1, 3}, {TW_CEXP, 1, 9}, {TW_CEXP, 1, 19}, {TW_NEXT, 1, 0} }; static const hc2c_desc desc = { 20, "hc2cb2_20", twinstr, &GENUS, {204, 92, 72, 0} }; void X(codelet_hc2cb2_20) (planner *p) { X(khc2c_register) (p, hc2cb2_20, &desc, HC2C_VIA_RDFT); } #endif /* HAVE_FMA */ fftw-3.3.3/rdft/scalar/r2cb/hc2cbdft_2.c0000644000175000017500000000746412054411033014513 00000000000000/* * Copyright (c) 2003, 2007-11 Matteo Frigo * Copyright (c) 2003, 2007-11 Massachusetts Institute of Technology * * This program is free software; you can redistribute it and/or modify * it under the terms of the GNU General Public License as published by * the Free Software Foundation; either version 2 of the License, or * (at your option) any later version. * * This program is distributed in the hope that it will be useful, * but WITHOUT ANY WARRANTY; without even the implied warranty of * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the * GNU General Public License for more details. * * You should have received a copy of the GNU General Public License * along with this program; if not, write to the Free Software * Foundation, Inc., 51 Franklin Street, Fifth Floor, Boston, MA 02110-1301 USA * */ /* This file was automatically generated --- DO NOT EDIT */ /* Generated on Sun Nov 25 07:42:03 EST 2012 */ #include "codelet-rdft.h" #ifdef HAVE_FMA /* Generated by: ../../../genfft/gen_hc2cdft.native -fma -reorder-insns -schedule-for-pipeline -compact -variables 4 -pipeline-latency 4 -sign 1 -n 2 -dif -name hc2cbdft_2 -include hc2cb.h */ /* * This function contains 10 FP additions, 4 FP multiplications, * (or, 8 additions, 2 multiplications, 2 fused multiply/add), * 11 stack variables, 0 constants, and 8 memory accesses */ #include "hc2cb.h" static void hc2cbdft_2(R *Rp, R *Ip, R *Rm, R *Im, const R *W, stride rs, INT mb, INT me, INT ms) { { INT m; for (m = mb, W = W + ((mb - 1) * 2); m < me; m = m + 1, Rp = Rp + ms, Ip = Ip + ms, Rm = Rm - ms, Im = Im - ms, W = W + 2, MAKE_VOLATILE_STRIDE(8, rs)) { E T9, Ta, T3, Tc, T7, T4; { E T1, T2, T5, T6; T1 = Ip[0]; T2 = Im[0]; T5 = Rp[0]; T6 = Rm[0]; T9 = W[1]; Ta = T1 + T2; T3 = T1 - T2; Tc = T5 + T6; T7 = T5 - T6; T4 = W[0]; } { E Td, T8, Te, Tb; Td = T9 * T7; T8 = T4 * T7; Te = FMA(T4, Ta, Td); Tb = FNMS(T9, Ta, T8); Rm[0] = Tc + Te; Rp[0] = Tc - Te; Im[0] = Tb - T3; Ip[0] = T3 + Tb; } } } } static const tw_instr twinstr[] = { {TW_FULL, 1, 2}, {TW_NEXT, 1, 0} }; static const hc2c_desc desc = { 2, "hc2cbdft_2", twinstr, &GENUS, {8, 2, 2, 0} }; void X(codelet_hc2cbdft_2) (planner *p) { X(khc2c_register) (p, hc2cbdft_2, &desc, HC2C_VIA_DFT); } #else /* HAVE_FMA */ /* Generated by: ../../../genfft/gen_hc2cdft.native -compact -variables 4 -pipeline-latency 4 -sign 1 -n 2 -dif -name hc2cbdft_2 -include hc2cb.h */ /* * This function contains 10 FP additions, 4 FP multiplications, * (or, 8 additions, 2 multiplications, 2 fused multiply/add), * 9 stack variables, 0 constants, and 8 memory accesses */ #include "hc2cb.h" static void hc2cbdft_2(R *Rp, R *Ip, R *Rm, R *Im, const R *W, stride rs, INT mb, INT me, INT ms) { { INT m; for (m = mb, W = W + ((mb - 1) * 2); m < me; m = m + 1, Rp = Rp + ms, Ip = Ip + ms, Rm = Rm - ms, Im = Im - ms, W = W + 2, MAKE_VOLATILE_STRIDE(8, rs)) { E T3, T9, T7, Tb; { E T1, T2, T5, T6; T1 = Ip[0]; T2 = Im[0]; T3 = T1 - T2; T9 = T1 + T2; T5 = Rp[0]; T6 = Rm[0]; T7 = T5 - T6; Tb = T5 + T6; } { E Ta, Tc, T4, T8; T4 = W[0]; T8 = W[1]; Ta = FNMS(T8, T9, T4 * T7); Tc = FMA(T8, T7, T4 * T9); Ip[0] = T3 + Ta; Rp[0] = Tb - Tc; Im[0] = Ta - T3; Rm[0] = Tb + Tc; } } } } static const tw_instr twinstr[] = { {TW_FULL, 1, 2}, {TW_NEXT, 1, 0} }; static const hc2c_desc desc = { 2, "hc2cbdft_2", twinstr, &GENUS, {8, 2, 2, 0} }; void X(codelet_hc2cbdft_2) (planner *p) { X(khc2c_register) (p, hc2cbdft_2, &desc, HC2C_VIA_DFT); } #endif /* HAVE_FMA */ fftw-3.3.3/rdft/scalar/r2cb/hc2cbdft_4.c0000644000175000017500000001307512054411033014510 00000000000000/* * Copyright (c) 2003, 2007-11 Matteo Frigo * Copyright (c) 2003, 2007-11 Massachusetts Institute of Technology * * This program is free software; you can redistribute it and/or modify * it under the terms of the GNU General Public License as published by * the Free Software Foundation; either version 2 of the License, or * (at your option) any later version. * * This program is distributed in the hope that it will be useful, * but WITHOUT ANY WARRANTY; without even the implied warranty of * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the * GNU General Public License for more details. * * You should have received a copy of the GNU General Public License * along with this program; if not, write to the Free Software * Foundation, Inc., 51 Franklin Street, Fifth Floor, Boston, MA 02110-1301 USA * */ /* This file was automatically generated --- DO NOT EDIT */ /* Generated on Sun Nov 25 07:42:03 EST 2012 */ #include "codelet-rdft.h" #ifdef HAVE_FMA /* Generated by: ../../../genfft/gen_hc2cdft.native -fma -reorder-insns -schedule-for-pipeline -compact -variables 4 -pipeline-latency 4 -sign 1 -n 4 -dif -name hc2cbdft_4 -include hc2cb.h */ /* * This function contains 30 FP additions, 12 FP multiplications, * (or, 24 additions, 6 multiplications, 6 fused multiply/add), * 35 stack variables, 0 constants, and 16 memory accesses */ #include "hc2cb.h" static void hc2cbdft_4(R *Rp, R *Ip, R *Rm, R *Im, const R *W, stride rs, INT mb, INT me, INT ms) { { INT m; for (m = mb, W = W + ((mb - 1) * 6); m < me; m = m + 1, Rp = Rp + ms, Ip = Ip + ms, Rm = Rm - ms, Im = Im - ms, W = W + 6, MAKE_VOLATILE_STRIDE(16, rs)) { E Ty, TB, Tw, TE, TA, TF, Tz, TG, TC; { E T4, Tg, T3, Tm, Tc, T5, Th, Ti; { E T1, T2, Ta, Tb; T1 = Rp[0]; T2 = Rm[WS(rs, 1)]; Ta = Ip[0]; Tb = Im[WS(rs, 1)]; T4 = Rp[WS(rs, 1)]; Tg = T1 - T2; T3 = T1 + T2; Tm = Ta - Tb; Tc = Ta + Tb; T5 = Rm[0]; Th = Ip[WS(rs, 1)]; Ti = Im[0]; } { E T8, Td, T7, Ts, To, Tv, Tk, Te, Tf; T8 = W[0]; { E T9, T6, Tn, Tj; T9 = T4 - T5; T6 = T4 + T5; Tn = Th - Ti; Tj = Th + Ti; Ty = Tc - T9; Td = T9 + Tc; T7 = T3 + T6; Ts = T3 - T6; To = Tm + Tn; Tv = Tm - Tn; TB = Tg + Tj; Tk = Tg - Tj; Te = T8 * Td; } Tf = W[1]; { E Tr, Tu, Tt, TD, Tx, Tp, Tl, Tq; Tr = W[2]; Tp = T8 * Tk; Tu = W[3]; Tl = FMA(Tf, Tk, Te); Tt = Tr * Ts; Tq = FNMS(Tf, Td, Tp); TD = Tu * Ts; Rm[0] = T7 + Tl; Rp[0] = T7 - Tl; Im[0] = Tq - To; Ip[0] = To + Tq; Tx = W[4]; Tw = FNMS(Tu, Tv, Tt); TE = FMA(Tr, Tv, TD); TA = W[5]; TF = Tx * TB; Tz = Tx * Ty; } } } TG = FNMS(TA, Ty, TF); TC = FMA(TA, TB, Tz); Im[WS(rs, 1)] = TG - TE; Ip[WS(rs, 1)] = TE + TG; Rm[WS(rs, 1)] = Tw + TC; Rp[WS(rs, 1)] = Tw - TC; } } } static const tw_instr twinstr[] = { {TW_FULL, 1, 4}, {TW_NEXT, 1, 0} }; static const hc2c_desc desc = { 4, "hc2cbdft_4", twinstr, &GENUS, {24, 6, 6, 0} }; void X(codelet_hc2cbdft_4) (planner *p) { X(khc2c_register) (p, hc2cbdft_4, &desc, HC2C_VIA_DFT); } #else /* HAVE_FMA */ /* Generated by: ../../../genfft/gen_hc2cdft.native -compact -variables 4 -pipeline-latency 4 -sign 1 -n 4 -dif -name hc2cbdft_4 -include hc2cb.h */ /* * This function contains 30 FP additions, 12 FP multiplications, * (or, 24 additions, 6 multiplications, 6 fused multiply/add), * 19 stack variables, 0 constants, and 16 memory accesses */ #include "hc2cb.h" static void hc2cbdft_4(R *Rp, R *Ip, R *Rm, R *Im, const R *W, stride rs, INT mb, INT me, INT ms) { { INT m; for (m = mb, W = W + ((mb - 1) * 6); m < me; m = m + 1, Rp = Rp + ms, Ip = Ip + ms, Rm = Rm - ms, Im = Im - ms, W = W + 6, MAKE_VOLATILE_STRIDE(16, rs)) { E T3, Tl, T6, Tm, Td, Tj, Tx, Tv, Ts, Tq; { E Tf, Tc, T9, Ti; { E T1, T2, Ta, Tb; T1 = Rp[0]; T2 = Rm[WS(rs, 1)]; T3 = T1 + T2; Tf = T1 - T2; Ta = Ip[0]; Tb = Im[WS(rs, 1)]; Tc = Ta + Tb; Tl = Ta - Tb; } { E T4, T5, Tg, Th; T4 = Rp[WS(rs, 1)]; T5 = Rm[0]; T6 = T4 + T5; T9 = T4 - T5; Tg = Ip[WS(rs, 1)]; Th = Im[0]; Ti = Tg + Th; Tm = Tg - Th; } Td = T9 + Tc; Tj = Tf - Ti; Tx = Tf + Ti; Tv = Tc - T9; Ts = Tl - Tm; Tq = T3 - T6; } { E T7, Tn, Tk, To, T8, Te; T7 = T3 + T6; Tn = Tl + Tm; T8 = W[0]; Te = W[1]; Tk = FMA(T8, Td, Te * Tj); To = FNMS(Te, Td, T8 * Tj); Rp[0] = T7 - Tk; Ip[0] = Tn + To; Rm[0] = T7 + Tk; Im[0] = To - Tn; } { E Tt, Tz, Ty, TA; { E Tp, Tr, Tu, Tw; Tp = W[2]; Tr = W[3]; Tt = FNMS(Tr, Ts, Tp * Tq); Tz = FMA(Tr, Tq, Tp * Ts); Tu = W[4]; Tw = W[5]; Ty = FMA(Tu, Tv, Tw * Tx); TA = FNMS(Tw, Tv, Tu * Tx); } Rp[WS(rs, 1)] = Tt - Ty; Ip[WS(rs, 1)] = Tz + TA; Rm[WS(rs, 1)] = Tt + Ty; Im[WS(rs, 1)] = TA - Tz; } } } } static const tw_instr twinstr[] = { {TW_FULL, 1, 4}, {TW_NEXT, 1, 0} }; static const hc2c_desc desc = { 4, "hc2cbdft_4", twinstr, &GENUS, {24, 6, 6, 0} }; void X(codelet_hc2cbdft_4) (planner *p) { X(khc2c_register) (p, hc2cbdft_4, &desc, HC2C_VIA_DFT); } #endif /* HAVE_FMA */ fftw-3.3.3/rdft/scalar/r2cb/hc2cbdft_6.c0000644000175000017500000002104512054411034014507 00000000000000/* * Copyright (c) 2003, 2007-11 Matteo Frigo * Copyright (c) 2003, 2007-11 Massachusetts Institute of Technology * * This program is free software; you can redistribute it and/or modify * it under the terms of the GNU General Public License as published by * the Free Software Foundation; either version 2 of the License, or * (at your option) any later version. * * This program is distributed in the hope that it will be useful, * but WITHOUT ANY WARRANTY; without even the implied warranty of * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the * GNU General Public License for more details. * * You should have received a copy of the GNU General Public License * along with this program; if not, write to the Free Software * Foundation, Inc., 51 Franklin Street, Fifth Floor, Boston, MA 02110-1301 USA * */ /* This file was automatically generated --- DO NOT EDIT */ /* Generated on Sun Nov 25 07:42:04 EST 2012 */ #include "codelet-rdft.h" #ifdef HAVE_FMA /* Generated by: ../../../genfft/gen_hc2cdft.native -fma -reorder-insns -schedule-for-pipeline -compact -variables 4 -pipeline-latency 4 -sign 1 -n 6 -dif -name hc2cbdft_6 -include hc2cb.h */ /* * This function contains 58 FP additions, 32 FP multiplications, * (or, 36 additions, 10 multiplications, 22 fused multiply/add), * 52 stack variables, 2 constants, and 24 memory accesses */ #include "hc2cb.h" static void hc2cbdft_6(R *Rp, R *Ip, R *Rm, R *Im, const R *W, stride rs, INT mb, INT me, INT ms) { DK(KP866025403, +0.866025403784438646763723170752936183471402627); DK(KP500000000, +0.500000000000000000000000000000000000000000000); { INT m; for (m = mb, W = W + ((mb - 1) * 10); m < me; m = m + 1, Rp = Rp + ms, Ip = Ip + ms, Rm = Rm - ms, Im = Im - ms, W = W + 10, MAKE_VOLATILE_STRIDE(24, rs)) { E T18, T1b, T16, T1e, T1a, T1f, T19, T1g, T1c; { E Tw, T4, TV, Tj, TP, TH, Tr, TY, T5, T6, Ta, Ty; { E Tg, TF, Tf, TD, Tp, Th; { E Td, Te, Tn, To; Td = Ip[WS(rs, 1)]; Te = Im[WS(rs, 1)]; Tn = Ip[0]; To = Im[WS(rs, 2)]; Tg = Ip[WS(rs, 2)]; TF = Te + Td; Tf = Td - Te; TD = Tn + To; Tp = Tn - To; Th = Im[0]; } { E T2, T3, T8, T9; T2 = Rp[0]; T3 = Rm[WS(rs, 2)]; { E Tq, TE, Ti, TG; T8 = Rm[WS(rs, 1)]; TE = Tg + Th; Ti = Tg - Th; Tw = T2 - T3; T4 = T2 + T3; TG = TE - TF; TV = TF + TE; Tq = Tf + Ti; Tj = Tf - Ti; TP = FNMS(KP500000000, TG, TD); TH = TD + TG; T9 = Rp[WS(rs, 1)]; Tr = FNMS(KP500000000, Tq, Tp); TY = Tp + Tq; } T5 = Rp[WS(rs, 2)]; T6 = Rm[0]; Ta = T8 + T9; Ty = T8 - T9; } } { E TO, TT, Ts, TA, TR, Tc, TN, TW, TS, Tx, T7; Tx = T5 - T6; T7 = T5 + T6; TO = W[0]; TT = W[1]; { E Tz, TQ, Tb, TU; Tz = Tx + Ty; TQ = Tx - Ty; Tb = T7 + Ta; Ts = T7 - Ta; TU = FNMS(KP500000000, Tz, Tw); TA = Tw + Tz; TR = FMA(KP866025403, TQ, TP); T18 = FNMS(KP866025403, TQ, TP); Tc = FNMS(KP500000000, Tb, T4); TN = T4 + Tb; T1b = FMA(KP866025403, TV, TU); TW = FNMS(KP866025403, TV, TU); TS = TO * TR; } { E T15, Tt, T12, T1, Tm, TI, TM, Tl, TJ; { E Tv, TC, TB, TL, Tk, TZ, TX, T10; T15 = FMA(KP866025403, Ts, Tr); Tt = FNMS(KP866025403, Ts, Tr); TZ = TO * TW; TX = FMA(TT, TW, TS); Tv = W[4]; TC = W[5]; T10 = FNMS(TT, TR, TZ); Rm[0] = TN + TX; Rp[0] = TN - TX; TB = Tv * TA; Im[0] = T10 - TY; Ip[0] = TY + T10; TL = TC * TA; Tk = FNMS(KP866025403, Tj, Tc); T12 = FMA(KP866025403, Tj, Tc); T1 = W[3]; Tm = W[2]; TI = FNMS(TC, TH, TB); TM = FMA(Tv, TH, TL); Tl = T1 * Tk; TJ = Tm * Tk; } { E T11, T14, T13, T1d, T17, Tu, TK; Tu = FMA(Tm, Tt, Tl); TK = FNMS(T1, Tt, TJ); T11 = W[6]; T14 = W[7]; Im[WS(rs, 1)] = TI - Tu; Ip[WS(rs, 1)] = Tu + TI; Rm[WS(rs, 1)] = TK + TM; Rp[WS(rs, 1)] = TK - TM; T13 = T11 * T12; T1d = T14 * T12; T17 = W[8]; T16 = FNMS(T14, T15, T13); T1e = FMA(T11, T15, T1d); T1a = W[9]; T1f = T17 * T1b; T19 = T17 * T18; } } } } T1g = FNMS(T1a, T18, T1f); T1c = FMA(T1a, T1b, T19); Im[WS(rs, 2)] = T1g - T1e; Ip[WS(rs, 2)] = T1e + T1g; Rm[WS(rs, 2)] = T16 + T1c; Rp[WS(rs, 2)] = T16 - T1c; } } } static const tw_instr twinstr[] = { {TW_FULL, 1, 6}, {TW_NEXT, 1, 0} }; static const hc2c_desc desc = { 6, "hc2cbdft_6", twinstr, &GENUS, {36, 10, 22, 0} }; void X(codelet_hc2cbdft_6) (planner *p) { X(khc2c_register) (p, hc2cbdft_6, &desc, HC2C_VIA_DFT); } #else /* HAVE_FMA */ /* Generated by: ../../../genfft/gen_hc2cdft.native -compact -variables 4 -pipeline-latency 4 -sign 1 -n 6 -dif -name hc2cbdft_6 -include hc2cb.h */ /* * This function contains 58 FP additions, 28 FP multiplications, * (or, 44 additions, 14 multiplications, 14 fused multiply/add), * 29 stack variables, 2 constants, and 24 memory accesses */ #include "hc2cb.h" static void hc2cbdft_6(R *Rp, R *Ip, R *Rm, R *Im, const R *W, stride rs, INT mb, INT me, INT ms) { DK(KP500000000, +0.500000000000000000000000000000000000000000000); DK(KP866025403, +0.866025403784438646763723170752936183471402627); { INT m; for (m = mb, W = W + ((mb - 1) * 10); m < me; m = m + 1, Rp = Rp + ms, Ip = Ip + ms, Rm = Rm - ms, Im = Im - ms, W = W + 10, MAKE_VOLATILE_STRIDE(24, rs)) { E T4, Tv, Tr, TL, Tb, Tc, Ty, TP, To, TB, Tj, TQ, Tp, Tq, TE; E TM; { E Ta, Tx, T7, Tw, T2, T3; T2 = Rp[0]; T3 = Rm[WS(rs, 2)]; T4 = T2 + T3; Tv = T2 - T3; { E T8, T9, T5, T6; T8 = Rm[WS(rs, 1)]; T9 = Rp[WS(rs, 1)]; Ta = T8 + T9; Tx = T8 - T9; T5 = Rp[WS(rs, 2)]; T6 = Rm[0]; T7 = T5 + T6; Tw = T5 - T6; } Tr = KP866025403 * (T7 - Ta); TL = KP866025403 * (Tw - Tx); Tb = T7 + Ta; Tc = FNMS(KP500000000, Tb, T4); Ty = Tw + Tx; TP = FNMS(KP500000000, Ty, Tv); } { E Tf, TC, Ti, TD, Td, Te; Td = Ip[WS(rs, 1)]; Te = Im[WS(rs, 1)]; Tf = Td - Te; TC = Te + Td; { E Tm, Tn, Tg, Th; Tm = Ip[0]; Tn = Im[WS(rs, 2)]; To = Tm - Tn; TB = Tm + Tn; Tg = Ip[WS(rs, 2)]; Th = Im[0]; Ti = Tg - Th; TD = Tg + Th; } Tj = KP866025403 * (Tf - Ti); TQ = KP866025403 * (TC + TD); Tp = Tf + Ti; Tq = FNMS(KP500000000, Tp, To); TE = TC - TD; TM = FMA(KP500000000, TE, TB); } { E TJ, TT, TS, TU; TJ = T4 + Tb; TT = To + Tp; { E TN, TR, TK, TO; TN = TL + TM; TR = TP - TQ; TK = W[0]; TO = W[1]; TS = FMA(TK, TN, TO * TR); TU = FNMS(TO, TN, TK * TR); } Rp[0] = TJ - TS; Ip[0] = TT + TU; Rm[0] = TJ + TS; Im[0] = TU - TT; } { E TZ, T15, T14, T16; { E TW, TY, TV, TX; TW = Tc + Tj; TY = Tr + Tq; TV = W[6]; TX = W[7]; TZ = FNMS(TX, TY, TV * TW); T15 = FMA(TX, TW, TV * TY); } { E T11, T13, T10, T12; T11 = TM - TL; T13 = TP + TQ; T10 = W[8]; T12 = W[9]; T14 = FMA(T10, T11, T12 * T13); T16 = FNMS(T12, T11, T10 * T13); } Rp[WS(rs, 2)] = TZ - T14; Ip[WS(rs, 2)] = T15 + T16; Rm[WS(rs, 2)] = TZ + T14; Im[WS(rs, 2)] = T16 - T15; } { E Tt, TH, TG, TI; { E Tk, Ts, T1, Tl; Tk = Tc - Tj; Ts = Tq - Tr; T1 = W[3]; Tl = W[2]; Tt = FMA(T1, Tk, Tl * Ts); TH = FNMS(T1, Ts, Tl * Tk); } { E Tz, TF, Tu, TA; Tz = Tv + Ty; TF = TB - TE; Tu = W[4]; TA = W[5]; TG = FNMS(TA, TF, Tu * Tz); TI = FMA(TA, Tz, Tu * TF); } Ip[WS(rs, 1)] = Tt + TG; Rp[WS(rs, 1)] = TH - TI; Im[WS(rs, 1)] = TG - Tt; Rm[WS(rs, 1)] = TH + TI; } } } } static const tw_instr twinstr[] = { {TW_FULL, 1, 6}, {TW_NEXT, 1, 0} }; static const hc2c_desc desc = { 6, "hc2cbdft_6", twinstr, &GENUS, {44, 14, 14, 0} }; void X(codelet_hc2cbdft_6) (planner *p) { X(khc2c_register) (p, hc2cbdft_6, &desc, HC2C_VIA_DFT); } #endif /* HAVE_FMA */ fftw-3.3.3/rdft/scalar/r2cb/hc2cbdft_8.c0000644000175000017500000002471112054411034014514 00000000000000/* * Copyright (c) 2003, 2007-11 Matteo Frigo * Copyright (c) 2003, 2007-11 Massachusetts Institute of Technology * * This program is free software; you can redistribute it and/or modify * it under the terms of the GNU General Public License as published by * the Free Software Foundation; either version 2 of the License, or * (at your option) any later version. * * This program is distributed in the hope that it will be useful, * but WITHOUT ANY WARRANTY; without even the implied warranty of * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the * GNU General Public License for more details. * * You should have received a copy of the GNU General Public License * along with this program; if not, write to the Free Software * Foundation, Inc., 51 Franklin Street, Fifth Floor, Boston, MA 02110-1301 USA * */ /* This file was automatically generated --- DO NOT EDIT */ /* Generated on Sun Nov 25 07:42:04 EST 2012 */ #include "codelet-rdft.h" #ifdef HAVE_FMA /* Generated by: ../../../genfft/gen_hc2cdft.native -fma -reorder-insns -schedule-for-pipeline -compact -variables 4 -pipeline-latency 4 -sign 1 -n 8 -dif -name hc2cbdft_8 -include hc2cb.h */ /* * This function contains 82 FP additions, 36 FP multiplications, * (or, 60 additions, 14 multiplications, 22 fused multiply/add), * 55 stack variables, 1 constants, and 32 memory accesses */ #include "hc2cb.h" static void hc2cbdft_8(R *Rp, R *Ip, R *Rm, R *Im, const R *W, stride rs, INT mb, INT me, INT ms) { DK(KP707106781, +0.707106781186547524400844362104849039284835938); { INT m; for (m = mb, W = W + ((mb - 1) * 14); m < me; m = m + 1, Rp = Rp + ms, Ip = Ip + ms, Rm = Rm - ms, Im = Im - ms, W = W + 14, MAKE_VOLATILE_STRIDE(32, rs)) { E T1m, T1r, T1i, T1u, T1o, T1v, T1n, T1w, T1s; { E T1k, Tl, T1p, TE, TP, T1g, TM, T1b, T1f, T1a, TU, Tf, T1l, TH, Tw; E T1q; { E TA, T3, TN, Tk, Th, T6, TO, TD, Tb, Tm, Ta, TK, Tp, Tc, Ts; E Tt; { E T4, T5, TB, TC; { E T1, T2, Ti, Tj; T1 = Rp[0]; T2 = Rm[WS(rs, 3)]; Ti = Ip[0]; Tj = Im[WS(rs, 3)]; T4 = Rp[WS(rs, 2)]; TA = T1 - T2; T3 = T1 + T2; TN = Ti - Tj; Tk = Ti + Tj; T5 = Rm[WS(rs, 1)]; TB = Ip[WS(rs, 2)]; TC = Im[WS(rs, 1)]; } { E T8, T9, Tn, To; T8 = Rp[WS(rs, 1)]; Th = T4 - T5; T6 = T4 + T5; TO = TB - TC; TD = TB + TC; T9 = Rm[WS(rs, 2)]; Tn = Ip[WS(rs, 1)]; To = Im[WS(rs, 2)]; Tb = Rm[0]; Tm = T8 - T9; Ta = T8 + T9; TK = Tn - To; Tp = Tn + To; Tc = Rp[WS(rs, 3)]; Ts = Im[0]; Tt = Ip[WS(rs, 3)]; } } { E Tr, Td, Tu, TL, Te, T7; T1k = Tk - Th; Tl = Th + Tk; Tr = Tb - Tc; Td = Tb + Tc; TL = Tt - Ts; Tu = Ts + Tt; T1p = TA + TD; TE = TA - TD; TP = TN + TO; T1g = TN - TO; TM = TK + TL; T1b = TL - TK; T1f = Ta - Td; Te = Ta + Td; T1a = T3 - T6; T7 = T3 + T6; { E Tq, TF, TG, Tv; Tq = Tm + Tp; TF = Tm - Tp; TG = Tr - Tu; Tv = Tr + Tu; TU = T7 - Te; Tf = T7 + Te; T1l = TF - TG; TH = TF + TG; Tw = Tq - Tv; T1q = Tq + Tv; } } } { E TX, T10, T1c, T13, T1h, T1E, T1H, T1C, T1K, T1G, T1L, T1F; { E TQ, Tx, T1y, TI, Tg, Tz; TX = TP - TM; TQ = TM + TP; Tx = FMA(KP707106781, Tw, Tl); T10 = FNMS(KP707106781, Tw, Tl); T1c = T1a + T1b; T1y = T1a - T1b; T13 = FNMS(KP707106781, TH, TE); TI = FMA(KP707106781, TH, TE); Tg = W[0]; Tz = W[1]; { E T1B, T1A, T1x, T1J, T1z, T1D; { E TR, Ty, TS, TJ; T1B = T1g - T1f; T1h = T1f + T1g; T1A = W[11]; TR = Tg * TI; Ty = Tg * Tx; T1x = W[10]; T1J = T1A * T1y; TS = FNMS(Tz, Tx, TR); TJ = FMA(Tz, TI, Ty); T1z = T1x * T1y; T1m = FMA(KP707106781, T1l, T1k); T1E = FNMS(KP707106781, T1l, T1k); Im[0] = TS - TQ; Ip[0] = TQ + TS; Rm[0] = Tf + TJ; Rp[0] = Tf - TJ; T1H = FMA(KP707106781, T1q, T1p); T1r = FNMS(KP707106781, T1q, T1p); T1D = W[12]; } T1C = FNMS(T1A, T1B, T1z); T1K = FMA(T1x, T1B, T1J); T1G = W[13]; T1L = T1D * T1H; T1F = T1D * T1E; } } { E TY, T16, T12, T17, T11; { E TW, TT, T15, TV, TZ, T1M, T1I; TW = W[7]; T1M = FNMS(T1G, T1E, T1L); T1I = FMA(T1G, T1H, T1F); TT = W[6]; T15 = TW * TU; Im[WS(rs, 3)] = T1M - T1K; Ip[WS(rs, 3)] = T1K + T1M; Rm[WS(rs, 3)] = T1C + T1I; Rp[WS(rs, 3)] = T1C - T1I; TV = TT * TU; TZ = W[8]; TY = FNMS(TW, TX, TV); T16 = FMA(TT, TX, T15); T12 = W[9]; T17 = TZ * T13; T11 = TZ * T10; } { E T1e, T19, T1t, T1d, T1j, T18, T14; T1e = W[3]; T18 = FNMS(T12, T10, T17); T14 = FMA(T12, T13, T11); T19 = W[2]; T1t = T1e * T1c; Im[WS(rs, 2)] = T18 - T16; Ip[WS(rs, 2)] = T16 + T18; Rm[WS(rs, 2)] = TY + T14; Rp[WS(rs, 2)] = TY - T14; T1d = T19 * T1c; T1j = W[4]; T1i = FNMS(T1e, T1h, T1d); T1u = FMA(T19, T1h, T1t); T1o = W[5]; T1v = T1j * T1r; T1n = T1j * T1m; } } } } T1w = FNMS(T1o, T1m, T1v); T1s = FMA(T1o, T1r, T1n); Im[WS(rs, 1)] = T1w - T1u; Ip[WS(rs, 1)] = T1u + T1w; Rm[WS(rs, 1)] = T1i + T1s; Rp[WS(rs, 1)] = T1i - T1s; } } } static const tw_instr twinstr[] = { {TW_FULL, 1, 8}, {TW_NEXT, 1, 0} }; static const hc2c_desc desc = { 8, "hc2cbdft_8", twinstr, &GENUS, {60, 14, 22, 0} }; void X(codelet_hc2cbdft_8) (planner *p) { X(khc2c_register) (p, hc2cbdft_8, &desc, HC2C_VIA_DFT); } #else /* HAVE_FMA */ /* Generated by: ../../../genfft/gen_hc2cdft.native -compact -variables 4 -pipeline-latency 4 -sign 1 -n 8 -dif -name hc2cbdft_8 -include hc2cb.h */ /* * This function contains 82 FP additions, 32 FP multiplications, * (or, 68 additions, 18 multiplications, 14 fused multiply/add), * 30 stack variables, 1 constants, and 32 memory accesses */ #include "hc2cb.h" static void hc2cbdft_8(R *Rp, R *Ip, R *Rm, R *Im, const R *W, stride rs, INT mb, INT me, INT ms) { DK(KP707106781, +0.707106781186547524400844362104849039284835938); { INT m; for (m = mb, W = W + ((mb - 1) * 14); m < me; m = m + 1, Rp = Rp + ms, Ip = Ip + ms, Rm = Rm - ms, Im = Im - ms, W = W + 14, MAKE_VOLATILE_STRIDE(32, rs)) { E T7, T1d, T1h, Tl, TG, T14, T19, TO, Te, TL, T18, T15, TB, T1e, Tw; E T1i; { E T3, TC, Tk, TM, T6, Th, TF, TN; { E T1, T2, Ti, Tj; T1 = Rp[0]; T2 = Rm[WS(rs, 3)]; T3 = T1 + T2; TC = T1 - T2; Ti = Ip[0]; Tj = Im[WS(rs, 3)]; Tk = Ti + Tj; TM = Ti - Tj; } { E T4, T5, TD, TE; T4 = Rp[WS(rs, 2)]; T5 = Rm[WS(rs, 1)]; T6 = T4 + T5; Th = T4 - T5; TD = Ip[WS(rs, 2)]; TE = Im[WS(rs, 1)]; TF = TD + TE; TN = TD - TE; } T7 = T3 + T6; T1d = Tk - Th; T1h = TC + TF; Tl = Th + Tk; TG = TC - TF; T14 = T3 - T6; T19 = TM - TN; TO = TM + TN; } { E Ta, Tm, Tp, TJ, Td, Tr, Tu, TK; { E T8, T9, Tn, To; T8 = Rp[WS(rs, 1)]; T9 = Rm[WS(rs, 2)]; Ta = T8 + T9; Tm = T8 - T9; Tn = Ip[WS(rs, 1)]; To = Im[WS(rs, 2)]; Tp = Tn + To; TJ = Tn - To; } { E Tb, Tc, Ts, Tt; Tb = Rm[0]; Tc = Rp[WS(rs, 3)]; Td = Tb + Tc; Tr = Tb - Tc; Ts = Im[0]; Tt = Ip[WS(rs, 3)]; Tu = Ts + Tt; TK = Tt - Ts; } Te = Ta + Td; TL = TJ + TK; T18 = Ta - Td; T15 = TK - TJ; { E Tz, TA, Tq, Tv; Tz = Tm - Tp; TA = Tr - Tu; TB = KP707106781 * (Tz + TA); T1e = KP707106781 * (Tz - TA); Tq = Tm + Tp; Tv = Tr + Tu; Tw = KP707106781 * (Tq - Tv); T1i = KP707106781 * (Tq + Tv); } } { E Tf, TP, TI, TQ; Tf = T7 + Te; TP = TL + TO; { E Tx, TH, Tg, Ty; Tx = Tl + Tw; TH = TB + TG; Tg = W[0]; Ty = W[1]; TI = FMA(Tg, Tx, Ty * TH); TQ = FNMS(Ty, Tx, Tg * TH); } Rp[0] = Tf - TI; Ip[0] = TP + TQ; Rm[0] = Tf + TI; Im[0] = TQ - TP; } { E T1r, T1x, T1w, T1y; { E T1o, T1q, T1n, T1p; T1o = T14 - T15; T1q = T19 - T18; T1n = W[10]; T1p = W[11]; T1r = FNMS(T1p, T1q, T1n * T1o); T1x = FMA(T1p, T1o, T1n * T1q); } { E T1t, T1v, T1s, T1u; T1t = T1d - T1e; T1v = T1i + T1h; T1s = W[12]; T1u = W[13]; T1w = FMA(T1s, T1t, T1u * T1v); T1y = FNMS(T1u, T1t, T1s * T1v); } Rp[WS(rs, 3)] = T1r - T1w; Ip[WS(rs, 3)] = T1x + T1y; Rm[WS(rs, 3)] = T1r + T1w; Im[WS(rs, 3)] = T1y - T1x; } { E TV, T11, T10, T12; { E TS, TU, TR, TT; TS = T7 - Te; TU = TO - TL; TR = W[6]; TT = W[7]; TV = FNMS(TT, TU, TR * TS); T11 = FMA(TT, TS, TR * TU); } { E TX, TZ, TW, TY; TX = Tl - Tw; TZ = TG - TB; TW = W[8]; TY = W[9]; T10 = FMA(TW, TX, TY * TZ); T12 = FNMS(TY, TX, TW * TZ); } Rp[WS(rs, 2)] = TV - T10; Ip[WS(rs, 2)] = T11 + T12; Rm[WS(rs, 2)] = TV + T10; Im[WS(rs, 2)] = T12 - T11; } { E T1b, T1l, T1k, T1m; { E T16, T1a, T13, T17; T16 = T14 + T15; T1a = T18 + T19; T13 = W[2]; T17 = W[3]; T1b = FNMS(T17, T1a, T13 * T16); T1l = FMA(T17, T16, T13 * T1a); } { E T1f, T1j, T1c, T1g; T1f = T1d + T1e; T1j = T1h - T1i; T1c = W[4]; T1g = W[5]; T1k = FMA(T1c, T1f, T1g * T1j); T1m = FNMS(T1g, T1f, T1c * T1j); } Rp[WS(rs, 1)] = T1b - T1k; Ip[WS(rs, 1)] = T1l + T1m; Rm[WS(rs, 1)] = T1b + T1k; Im[WS(rs, 1)] = T1m - T1l; } } } } static const tw_instr twinstr[] = { {TW_FULL, 1, 8}, {TW_NEXT, 1, 0} }; static const hc2c_desc desc = { 8, "hc2cbdft_8", twinstr, &GENUS, {68, 18, 14, 0} }; void X(codelet_hc2cbdft_8) (planner *p) { X(khc2c_register) (p, hc2cbdft_8, &desc, HC2C_VIA_DFT); } #endif /* HAVE_FMA */ fftw-3.3.3/rdft/scalar/r2cb/hc2cbdft_10.c0000644000175000017500000003531312054411034014565 00000000000000/* * Copyright (c) 2003, 2007-11 Matteo Frigo * Copyright (c) 2003, 2007-11 Massachusetts Institute of Technology * * This program is free software; you can redistribute it and/or modify * it under the terms of the GNU General Public License as published by * the Free Software Foundation; either version 2 of the License, or * (at your option) any later version. * * This program is distributed in the hope that it will be useful, * but WITHOUT ANY WARRANTY; without even the implied warranty of * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the * GNU General Public License for more details. * * You should have received a copy of the GNU General Public License * along with this program; if not, write to the Free Software * Foundation, Inc., 51 Franklin Street, Fifth Floor, Boston, MA 02110-1301 USA * */ /* This file was automatically generated --- DO NOT EDIT */ /* Generated on Sun Nov 25 07:42:04 EST 2012 */ #include "codelet-rdft.h" #ifdef HAVE_FMA /* Generated by: ../../../genfft/gen_hc2cdft.native -fma -reorder-insns -schedule-for-pipeline -compact -variables 4 -pipeline-latency 4 -sign 1 -n 10 -dif -name hc2cbdft_10 -include hc2cb.h */ /* * This function contains 122 FP additions, 72 FP multiplications, * (or, 68 additions, 18 multiplications, 54 fused multiply/add), * 95 stack variables, 4 constants, and 40 memory accesses */ #include "hc2cb.h" static void hc2cbdft_10(R *Rp, R *Ip, R *Rm, R *Im, const R *W, stride rs, INT mb, INT me, INT ms) { DK(KP951056516, +0.951056516295153572116439333379382143405698634); DK(KP559016994, +0.559016994374947424102293417182819058860154590); DK(KP250000000, +0.250000000000000000000000000000000000000000000); DK(KP618033988, +0.618033988749894848204586834365638117720309180); { INT m; for (m = mb, W = W + ((mb - 1) * 18); m < me; m = m + 1, Rp = Rp + ms, Ip = Ip + ms, Rm = Rm - ms, Im = Im - ms, W = W + 18, MAKE_VOLATILE_STRIDE(40, rs)) { E T2d, T2f; { E T1g, TQ, T1z, TZ, Tu, T23, T1p, T14, Tt, T27, T13, Tj, Tz, T1i, T18; E TJ, TS, T19, Ty, TA; { E Tl, T3, T7, Tm, T6, Tr, TY, T1n, Th, T8, T1, T2; T1 = Rp[0]; T2 = Rm[WS(rs, 4)]; { E Te, Tp, Td, Tf, Tb, Tc; Tb = Rp[WS(rs, 4)]; Tc = Rm[0]; Te = Rm[WS(rs, 3)]; Tl = T1 - T2; T3 = T1 + T2; Tp = Tb - Tc; Td = Tb + Tc; Tf = Rp[WS(rs, 1)]; { E T4, T5, Tq, Tg; T4 = Rp[WS(rs, 2)]; T5 = Rm[WS(rs, 2)]; T7 = Rm[WS(rs, 1)]; Tq = Te - Tf; Tg = Te + Tf; Tm = T4 - T5; T6 = T4 + T5; Tr = Tp + Tq; TY = Tp - Tq; T1n = Td - Tg; Th = Td + Tg; T8 = Rp[WS(rs, 3)]; } } { E TO, Tn, T9, TP; TO = Ip[0]; Tn = T7 - T8; T9 = T7 + T8; TP = Im[WS(rs, 4)]; { E TG, TH, TF, T16, TD, TE, Ti; TD = Ip[WS(rs, 4)]; { E TX, To, T1o, Ta, Ts; TX = Tm - Tn; To = Tm + Tn; T1o = T6 - T9; Ta = T6 + T9; T1g = TO - TP; TQ = TO + TP; T1z = FNMS(KP618033988, TX, TY); TZ = FMA(KP618033988, TY, TX); Ts = To + Tr; Tu = To - Tr; T23 = FMA(KP618033988, T1n, T1o); T1p = FNMS(KP618033988, T1o, T1n); Ti = Ta + Th; T14 = Ta - Th; Tt = FNMS(KP250000000, Ts, Tl); T27 = Tl + Ts; TE = Im[0]; } T13 = FNMS(KP250000000, Ti, T3); Tj = T3 + Ti; TG = Im[WS(rs, 3)]; TH = Ip[WS(rs, 1)]; TF = TD + TE; T16 = TD - TE; { E Tw, T17, TI, Tx; Tw = Ip[WS(rs, 2)]; T17 = TH - TG; TI = TG + TH; Tx = Im[WS(rs, 2)]; Tz = Im[WS(rs, 1)]; T1i = T16 + T17; T18 = T16 - T17; TJ = TF + TI; TS = TF - TI; T19 = Tw - Tx; Ty = Tw + Tx; TA = Ip[WS(rs, 3)]; } } } } { E T26, T2y, T2a, T28, T1q, T1K, T24, T2k, T10, T1Q, T1A, T2q, T29, Tk, TN; E T2c, T1M, T1P, T2w, TM, T1O, T1S, T1s, T1x, T2m, T2p, T1w, T1C, T2o, T2s; E T12, T1f, T1G, T1J, T1I, T1E, T1e, T1U, T1W, T21, T2g, T2j, T20, T2e, T2i; E T2u, T1a, TB; T1a = TA - Tz; TB = Tz + TA; { E T1Y, T1c, T1u, T1t, T1N, TL, TK, Tv, T2n, T1v; { E T1l, TV, T1k, TU, T1b, T1h; T26 = W[9]; T1b = T19 - T1a; T1h = T19 + T1a; { E TC, TR, T1j, TT; TC = Ty + TB; TR = Ty - TB; T1Y = FMA(KP618033988, T18, T1b); T1c = FNMS(KP618033988, T1b, T18); T1j = T1h + T1i; T1l = T1h - T1i; T1u = FNMS(KP618033988, TC, TJ); TK = FMA(KP618033988, TJ, TC); TT = TR + TS; TV = TR - TS; T2y = T1g + T1j; T1k = FNMS(KP250000000, T1j, T1g); T2a = TQ + TT; TU = FNMS(KP250000000, TT, TQ); T28 = T26 * T27; } { E T22, T1m, T1y, TW; T22 = FMA(KP559016994, T1l, T1k); T1m = FNMS(KP559016994, T1l, T1k); T1y = FNMS(KP559016994, TV, TU); TW = FMA(KP559016994, TV, TU); T1q = FNMS(KP951056516, T1p, T1m); T1K = FMA(KP951056516, T1p, T1m); T24 = FNMS(KP951056516, T23, T22); T2k = FMA(KP951056516, T23, T22); T10 = FMA(KP951056516, TZ, TW); T1Q = FNMS(KP951056516, TZ, TW); T1A = FMA(KP951056516, T1z, T1y); T2q = FNMS(KP951056516, T1z, T1y); T29 = W[8]; } } Tv = FMA(KP559016994, Tu, Tt); T1t = FNMS(KP559016994, Tu, Tt); Tk = W[1]; TN = W[0]; T2c = T29 * T27; T1N = FMA(KP951056516, TK, Tv); TL = FNMS(KP951056516, TK, Tv); T1M = W[17]; T1P = W[16]; T2w = TN * TL; TM = Tk * TL; T1O = T1M * T1N; T1S = T1P * T1N; T2n = FMA(KP951056516, T1u, T1t); T1v = FNMS(KP951056516, T1u, T1t); T1s = W[5]; T1x = W[4]; T2m = W[13]; T2p = W[12]; T1w = T1s * T1v; T1C = T1x * T1v; T2o = T2m * T2n; T2s = T2p * T2n; { E T1X, T1d, T1H, T15, T2h, T1Z; T1X = FMA(KP559016994, T14, T13); T15 = FNMS(KP559016994, T14, T13); T12 = W[2]; T1f = W[3]; T1G = W[14]; T1d = FMA(KP951056516, T1c, T15); T1H = FNMS(KP951056516, T1c, T15); T1J = W[15]; T1I = T1G * T1H; T1E = T1f * T1d; T1e = T12 * T1d; T1U = T1J * T1H; T2h = FNMS(KP951056516, T1Y, T1X); T1Z = FMA(KP951056516, T1Y, T1X); T1W = W[6]; T21 = W[7]; T2g = W[10]; T2j = W[11]; T20 = T1W * T1Z; T2e = T21 * T1Z; T2i = T2g * T2h; T2u = T2j * T2h; } } { E T1D, T1F, T1L, T1R; { E T11, T2x, T1r, T1B; T11 = FMA(TN, T10, TM); T2x = FNMS(Tk, T10, T2w); T1r = FNMS(T1f, T1q, T1e); T1B = FMA(T1x, T1A, T1w); Rm[0] = Tj + T11; Rp[0] = Tj - T11; Ip[0] = T2x + T2y; Im[0] = T2x - T2y; Rp[WS(rs, 1)] = T1r - T1B; Rm[WS(rs, 1)] = T1B + T1r; T1D = FNMS(T1s, T1A, T1C); T1F = FMA(T12, T1q, T1E); T1L = FNMS(T1J, T1K, T1I); T1R = FMA(T1P, T1Q, T1O); } { E T1T, T1V, T2t, T2v; T1T = FNMS(T1M, T1Q, T1S); Ip[WS(rs, 1)] = T1D + T1F; Im[WS(rs, 1)] = T1D - T1F; Rm[WS(rs, 4)] = T1R + T1L; Rp[WS(rs, 4)] = T1L - T1R; T1V = FMA(T1G, T1K, T1U); T2t = FNMS(T2m, T2q, T2s); T2v = FMA(T2g, T2k, T2u); { E T2l, T2r, T25, T2b; T2l = FNMS(T2j, T2k, T2i); Ip[WS(rs, 4)] = T1T + T1V; Im[WS(rs, 4)] = T1T - T1V; Ip[WS(rs, 3)] = T2t + T2v; Im[WS(rs, 3)] = T2t - T2v; T2r = FMA(T2p, T2q, T2o); T25 = FNMS(T21, T24, T20); T2b = FMA(T29, T2a, T28); T2d = FNMS(T26, T2a, T2c); Rm[WS(rs, 3)] = T2r + T2l; Rp[WS(rs, 3)] = T2l - T2r; Rm[WS(rs, 2)] = T2b + T25; Rp[WS(rs, 2)] = T25 - T2b; T2f = FMA(T1W, T24, T2e); } } } } } Ip[WS(rs, 2)] = T2d + T2f; Im[WS(rs, 2)] = T2d - T2f; } } } static const tw_instr twinstr[] = { {TW_FULL, 1, 10}, {TW_NEXT, 1, 0} }; static const hc2c_desc desc = { 10, "hc2cbdft_10", twinstr, &GENUS, {68, 18, 54, 0} }; void X(codelet_hc2cbdft_10) (planner *p) { X(khc2c_register) (p, hc2cbdft_10, &desc, HC2C_VIA_DFT); } #else /* HAVE_FMA */ /* Generated by: ../../../genfft/gen_hc2cdft.native -compact -variables 4 -pipeline-latency 4 -sign 1 -n 10 -dif -name hc2cbdft_10 -include hc2cb.h */ /* * This function contains 122 FP additions, 60 FP multiplications, * (or, 92 additions, 30 multiplications, 30 fused multiply/add), * 61 stack variables, 4 constants, and 40 memory accesses */ #include "hc2cb.h" static void hc2cbdft_10(R *Rp, R *Ip, R *Rm, R *Im, const R *W, stride rs, INT mb, INT me, INT ms) { DK(KP951056516, +0.951056516295153572116439333379382143405698634); DK(KP587785252, +0.587785252292473129168705954639072768597652438); DK(KP250000000, +0.250000000000000000000000000000000000000000000); DK(KP559016994, +0.559016994374947424102293417182819058860154590); { INT m; for (m = mb, W = W + ((mb - 1) * 18); m < me; m = m + 1, Rp = Rp + ms, Ip = Ip + ms, Rm = Rm - ms, Im = Im - ms, W = W + 18, MAKE_VOLATILE_STRIDE(40, rs)) { E T3, TS, TR, T13, Ti, T12, TT, TU, T1g, T1T, Tr, T1s, TJ, T1h, TG; E T1m, TK, TL, T1k, T1l, T1b, T1P, TY, T1w; { E Td, To, Tg, Tp, Th, TQ, T6, Tl, T9, Tm, Ta, TP, T1, T2; T1 = Rp[0]; T2 = Rm[WS(rs, 4)]; T3 = T1 + T2; TS = T1 - T2; { E Tb, Tc, Te, Tf; Tb = Rp[WS(rs, 4)]; Tc = Rm[0]; Td = Tb + Tc; To = Tb - Tc; Te = Rm[WS(rs, 3)]; Tf = Rp[WS(rs, 1)]; Tg = Te + Tf; Tp = Te - Tf; } Th = Td + Tg; TQ = To + Tp; { E T4, T5, T7, T8; T4 = Rp[WS(rs, 2)]; T5 = Rm[WS(rs, 2)]; T6 = T4 + T5; Tl = T4 - T5; T7 = Rm[WS(rs, 1)]; T8 = Rp[WS(rs, 3)]; T9 = T7 + T8; Tm = T7 - T8; } Ta = T6 + T9; TP = Tl + Tm; TR = KP559016994 * (TP - TQ); T13 = KP559016994 * (Ta - Th); Ti = Ta + Th; T12 = FNMS(KP250000000, Ti, T3); TT = TP + TQ; TU = FNMS(KP250000000, TT, TS); { E T1e, T1f, Tn, Tq; T1e = T6 - T9; T1f = Td - Tg; T1g = FNMS(KP951056516, T1f, KP587785252 * T1e); T1T = FMA(KP951056516, T1e, KP587785252 * T1f); Tn = Tl - Tm; Tq = To - Tp; Tr = FMA(KP951056516, Tn, KP587785252 * Tq); T1s = FNMS(KP951056516, Tq, KP587785252 * Tn); } } { E TB, T18, TE, T19, TF, T1j, Tu, T15, Tx, T16, Ty, T1i, TH, TI; TH = Ip[0]; TI = Im[WS(rs, 4)]; TJ = TH + TI; T1h = TH - TI; { E Tz, TA, TC, TD; Tz = Ip[WS(rs, 4)]; TA = Im[0]; TB = Tz + TA; T18 = Tz - TA; TC = Im[WS(rs, 3)]; TD = Ip[WS(rs, 1)]; TE = TC + TD; T19 = TD - TC; } TF = TB - TE; T1j = T18 + T19; { E Ts, Tt, Tv, Tw; Ts = Ip[WS(rs, 2)]; Tt = Im[WS(rs, 2)]; Tu = Ts + Tt; T15 = Ts - Tt; Tv = Im[WS(rs, 1)]; Tw = Ip[WS(rs, 3)]; Tx = Tv + Tw; T16 = Tw - Tv; } Ty = Tu - Tx; T1i = T15 + T16; TG = KP559016994 * (Ty - TF); T1m = KP559016994 * (T1i - T1j); TK = Ty + TF; TL = FNMS(KP250000000, TK, TJ); T1k = T1i + T1j; T1l = FNMS(KP250000000, T1k, T1h); { E T17, T1a, TW, TX; T17 = T15 - T16; T1a = T18 - T19; T1b = FNMS(KP951056516, T1a, KP587785252 * T17); T1P = FMA(KP951056516, T17, KP587785252 * T1a); TW = Tu + Tx; TX = TB + TE; TY = FMA(KP951056516, TW, KP587785252 * TX); T1w = FNMS(KP951056516, TX, KP587785252 * TW); } } { E Tj, T2g, TN, T1H, T1U, T26, TZ, T1J, T1Q, T24, T1c, T1C, T1t, T29, T1o; E T1E, T1x, T2b, T20, T21, TM, T1S, TV; Tj = T3 + Ti; T2g = T1h + T1k; TM = TG + TL; TN = Tr + TM; T1H = TM - Tr; T1S = T1m + T1l; T1U = T1S - T1T; T26 = T1T + T1S; TV = TR + TU; TZ = TV - TY; T1J = TV + TY; { E T1O, T14, T1r, T1n, T1v; T1O = T13 + T12; T1Q = T1O + T1P; T24 = T1O - T1P; T14 = T12 - T13; T1c = T14 - T1b; T1C = T14 + T1b; T1r = TL - TG; T1t = T1r - T1s; T29 = T1s + T1r; T1n = T1l - T1m; T1o = T1g + T1n; T1E = T1n - T1g; T1v = TU - TR; T1x = T1v + T1w; T2b = T1v - T1w; { E T1X, T1Z, T1W, T1Y; T1X = TS + TT; T1Z = TJ + TK; T1W = W[9]; T1Y = W[8]; T20 = FMA(T1W, T1X, T1Y * T1Z); T21 = FNMS(T1W, T1Z, T1Y * T1X); } } { E T10, T2f, Tk, TO; Tk = W[0]; TO = W[1]; T10 = FMA(Tk, TN, TO * TZ); T2f = FNMS(TO, TN, Tk * TZ); Rp[0] = Tj - T10; Ip[0] = T2f + T2g; Rm[0] = Tj + T10; Im[0] = T2f - T2g; } { E T1V, T22, T1N, T1R; T1N = W[6]; T1R = W[7]; T1V = FNMS(T1R, T1U, T1N * T1Q); T22 = FMA(T1R, T1Q, T1N * T1U); Rp[WS(rs, 2)] = T1V - T20; Ip[WS(rs, 2)] = T21 + T22; Rm[WS(rs, 2)] = T20 + T1V; Im[WS(rs, 2)] = T21 - T22; } { E T1p, T1A, T1y, T1z; { E T11, T1d, T1q, T1u; T11 = W[2]; T1d = W[3]; T1p = FNMS(T1d, T1o, T11 * T1c); T1A = FMA(T1d, T1c, T11 * T1o); T1q = W[4]; T1u = W[5]; T1y = FMA(T1q, T1t, T1u * T1x); T1z = FNMS(T1u, T1t, T1q * T1x); } Rp[WS(rs, 1)] = T1p - T1y; Ip[WS(rs, 1)] = T1z + T1A; Rm[WS(rs, 1)] = T1y + T1p; Im[WS(rs, 1)] = T1z - T1A; } { E T1F, T1M, T1K, T1L; { E T1B, T1D, T1G, T1I; T1B = W[14]; T1D = W[15]; T1F = FNMS(T1D, T1E, T1B * T1C); T1M = FMA(T1D, T1C, T1B * T1E); T1G = W[16]; T1I = W[17]; T1K = FMA(T1G, T1H, T1I * T1J); T1L = FNMS(T1I, T1H, T1G * T1J); } Rp[WS(rs, 4)] = T1F - T1K; Ip[WS(rs, 4)] = T1L + T1M; Rm[WS(rs, 4)] = T1K + T1F; Im[WS(rs, 4)] = T1L - T1M; } { E T27, T2e, T2c, T2d; { E T23, T25, T28, T2a; T23 = W[10]; T25 = W[11]; T27 = FNMS(T25, T26, T23 * T24); T2e = FMA(T25, T24, T23 * T26); T28 = W[12]; T2a = W[13]; T2c = FMA(T28, T29, T2a * T2b); T2d = FNMS(T2a, T29, T28 * T2b); } Rp[WS(rs, 3)] = T27 - T2c; Ip[WS(rs, 3)] = T2d + T2e; Rm[WS(rs, 3)] = T2c + T27; Im[WS(rs, 3)] = T2d - T2e; } } } } } static const tw_instr twinstr[] = { {TW_FULL, 1, 10}, {TW_NEXT, 1, 0} }; static const hc2c_desc desc = { 10, "hc2cbdft_10", twinstr, &GENUS, {92, 30, 30, 0} }; void X(codelet_hc2cbdft_10) (planner *p) { X(khc2c_register) (p, hc2cbdft_10, &desc, HC2C_VIA_DFT); } #endif /* HAVE_FMA */ fftw-3.3.3/rdft/scalar/r2cb/hc2cbdft_12.c0000644000175000017500000003775512054411035014604 00000000000000/* * Copyright (c) 2003, 2007-11 Matteo Frigo * Copyright (c) 2003, 2007-11 Massachusetts Institute of Technology * * This program is free software; you can redistribute it and/or modify * it under the terms of the GNU General Public License as published by * the Free Software Foundation; either version 2 of the License, or * (at your option) any later version. * * This program is distributed in the hope that it will be useful, * but WITHOUT ANY WARRANTY; without even the implied warranty of * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the * GNU General Public License for more details. * * You should have received a copy of the GNU General Public License * along with this program; if not, write to the Free Software * Foundation, Inc., 51 Franklin Street, Fifth Floor, Boston, MA 02110-1301 USA * */ /* This file was automatically generated --- DO NOT EDIT */ /* Generated on Sun Nov 25 07:42:04 EST 2012 */ #include "codelet-rdft.h" #ifdef HAVE_FMA /* Generated by: ../../../genfft/gen_hc2cdft.native -fma -reorder-insns -schedule-for-pipeline -compact -variables 4 -pipeline-latency 4 -sign 1 -n 12 -dif -name hc2cbdft_12 -include hc2cb.h */ /* * This function contains 142 FP additions, 68 FP multiplications, * (or, 96 additions, 22 multiplications, 46 fused multiply/add), * 81 stack variables, 2 constants, and 48 memory accesses */ #include "hc2cb.h" static void hc2cbdft_12(R *Rp, R *Ip, R *Rm, R *Im, const R *W, stride rs, INT mb, INT me, INT ms) { DK(KP866025403, +0.866025403784438646763723170752936183471402627); DK(KP500000000, +0.500000000000000000000000000000000000000000000); { INT m; for (m = mb, W = W + ((mb - 1) * 22); m < me; m = m + 1, Rp = Rp + ms, Ip = Ip + ms, Rm = Rm - ms, Im = Im - ms, W = W + 22, MAKE_VOLATILE_STRIDE(48, rs)) { E T2S, T2V, T2w, T2Z, T2T, T2I, T2Q, T2Y, T2U, T2K, T2G, T30, T2W; { E Tb, T1Z, T2D, T1E, T1N, T2y, TD, T2t, T1U, T1e, T2o, TY, T1f, TI, T1g; E TN, Tm, T1V, T2z, T1H, T1Q, T2E, T19, T2u; { E T1c, TU, T1d, TX; { E Tu, T6, TT, TS, T5, Tt, Tw, Tx, TB, T9, Ty; { E T1, Tp, Tq, Tr, T4, T2, T3, T7, T8, Ts; T1 = Rp[0]; T2 = Rp[WS(rs, 4)]; T3 = Rm[WS(rs, 3)]; Tp = Ip[0]; Tq = Ip[WS(rs, 4)]; Tr = Im[WS(rs, 3)]; T4 = T2 + T3; Tu = T2 - T3; T6 = Rm[WS(rs, 5)]; TT = Tr + Tq; Ts = Tq - Tr; TS = FNMS(KP500000000, T4, T1); T5 = T1 + T4; T7 = Rm[WS(rs, 1)]; T8 = Rp[WS(rs, 2)]; T1c = Tp + Ts; Tt = FNMS(KP500000000, Ts, Tp); Tw = Im[WS(rs, 5)]; Tx = Im[WS(rs, 1)]; TB = T7 - T8; T9 = T7 + T8; Ty = Ip[WS(rs, 2)]; } { E T1L, Tv, Ta, TV, TW, Tz; T1L = FNMS(KP866025403, Tu, Tt); Tv = FMA(KP866025403, Tu, Tt); Ta = T6 + T9; TV = FNMS(KP500000000, T9, T6); TW = Tx + Ty; Tz = Tx - Ty; { E TC, T1M, T1C, TA, T1D; T1C = FMA(KP866025403, TT, TS); TU = FNMS(KP866025403, TT, TS); T1d = Tw + Tz; TA = FNMS(KP500000000, Tz, Tw); T1D = FNMS(KP866025403, TW, TV); TX = FMA(KP866025403, TW, TV); Tb = T5 + Ta; T1Z = T5 - Ta; TC = FNMS(KP866025403, TB, TA); T1M = FMA(KP866025403, TB, TA); T2D = T1C - T1D; T1E = T1C + T1D; T1N = T1L - T1M; T2y = T1L + T1M; TD = Tv + TC; T2t = Tv - TC; } } } { E T12, Th, TH, TE, Tg, T11, T14, TK, T17, Tk, TL; { E Tc, TZ, TF, TG, Tf, Td, Te, Ti, Tj, T10; Tc = Rp[WS(rs, 3)]; T1U = T1c + T1d; T1e = T1c - T1d; T2o = TU + TX; TY = TU - TX; Td = Rm[WS(rs, 4)]; Te = Rm[0]; TZ = Ip[WS(rs, 3)]; TF = Im[WS(rs, 4)]; TG = Im[0]; Tf = Td + Te; T12 = Td - Te; Th = Rm[WS(rs, 2)]; TH = TF - TG; T10 = TF + TG; TE = FNMS(KP500000000, Tf, Tc); Tg = Tc + Tf; Ti = Rp[WS(rs, 1)]; Tj = Rp[WS(rs, 5)]; T1f = TZ - T10; T11 = FMA(KP500000000, T10, TZ); T14 = Im[WS(rs, 2)]; TK = Ip[WS(rs, 5)]; T17 = Ti - Tj; Tk = Ti + Tj; TL = Ip[WS(rs, 1)]; } { E T1O, T13, Tl, TJ, TM, T15; T1O = FNMS(KP866025403, T12, T11); T13 = FMA(KP866025403, T12, T11); Tl = Th + Tk; TJ = FNMS(KP500000000, Tk, Th); TM = TK - TL; T15 = TK + TL; { E T18, T1P, T1F, T16, T1G; T1F = FNMS(KP866025403, TH, TE); TI = FMA(KP866025403, TH, TE); T1g = T15 - T14; T16 = FMA(KP500000000, T15, T14); T1G = FNMS(KP866025403, TM, TJ); TN = FMA(KP866025403, TM, TJ); Tm = Tg + Tl; T1V = Tg - Tl; T18 = FNMS(KP866025403, T17, T16); T1P = FMA(KP866025403, T17, T16); T2z = T1F - T1G; T1H = T1F + T1G; T1Q = T1O - T1P; T2E = T1O + T1P; T19 = T13 + T18; T2u = T13 - T18; } } } } { E T20, T2p, T1v, T1s, T1q, T1y, T1u, T1z, T1t; { E T1m, Tn, T1a, T1p, T1i, To, TP, TR, T1h, TO; T1m = Tb - Tm; Tn = Tb + Tm; T20 = T1f - T1g; T1h = T1f + T1g; T2p = TI + TN; TO = TI - TN; T1a = TY - T19; T1v = TY + T19; T1p = T1e - T1h; T1i = T1e + T1h; To = W[0]; T1s = TD - TO; TP = TD + TO; TR = W[1]; { E T1l, T1o, T1n, T1x, T1r; { E T1j, TQ, T1k, T1b; T1j = To * T1a; TQ = To * TP; T1l = W[10]; T1k = FNMS(TR, TP, T1j); T1b = FMA(TR, T1a, TQ); T1o = W[11]; T1n = T1l * T1m; Im[0] = T1k - T1i; Ip[0] = T1i + T1k; Rm[0] = Tn + T1b; Rp[0] = Tn - T1b; T1x = T1o * T1m; T1r = W[12]; } T1q = FNMS(T1o, T1p, T1n); T1y = FMA(T1l, T1p, T1x); T1u = W[13]; T1z = T1r * T1v; T1t = T1r * T1s; } } { E T2e, T2h, T1S, T2j, T2f, T26, T2c, T2m, T2g, T24, T22; { E T2b, T1R, T27, T2a, T1B, T29, T2l, T1K, T1J, T1W, T21, T25, T2d, T23, T1X; E T1Y; { E T1I, T28, T1A, T1w, T1T; T1A = FNMS(T1u, T1s, T1z); T1w = FMA(T1u, T1v, T1t); T1I = T1E - T1H; T28 = T1E + T1H; T2b = T1N + T1Q; T1R = T1N - T1Q; Im[WS(rs, 3)] = T1A - T1y; Ip[WS(rs, 3)] = T1y + T1A; Rm[WS(rs, 3)] = T1q + T1w; Rp[WS(rs, 3)] = T1q - T1w; T27 = W[14]; T2a = W[15]; T1B = W[2]; T29 = T27 * T28; T2l = T2a * T28; T1K = W[3]; T1J = T1B * T1I; T1W = T1U - T1V; T2e = T1V + T1U; T2h = T1Z - T20; T21 = T1Z + T20; T25 = T1K * T1I; T1T = W[4]; T2d = W[16]; T23 = T1T * T21; T1X = T1T * T1W; } T1S = FNMS(T1K, T1R, T1J); T2j = T2d * T2h; T2f = T2d * T2e; T26 = FMA(T1B, T1R, T25); T1Y = W[5]; T2c = FNMS(T2a, T2b, T29); T2m = FMA(T27, T2b, T2l); T2g = W[17]; T24 = FNMS(T1Y, T1W, T23); T22 = FMA(T1Y, T21, T1X); } { E T2L, T2O, T2P, T2v, T2N, T2X, T2n, T2s, T2A, T2F, T2r, T2H, T2R, T2J, T2B; E T2C; { E T2q, T2k, T2i, T2M, T2x; T2k = FNMS(T2g, T2e, T2j); T2i = FMA(T2g, T2h, T2f); Im[WS(rs, 1)] = T24 - T26; Ip[WS(rs, 1)] = T24 + T26; Rm[WS(rs, 1)] = T22 + T1S; Rp[WS(rs, 1)] = T1S - T22; Im[WS(rs, 4)] = T2k - T2m; Ip[WS(rs, 4)] = T2k + T2m; Rm[WS(rs, 4)] = T2i + T2c; Rp[WS(rs, 4)] = T2c - T2i; T2q = T2o + T2p; T2M = T2o - T2p; T2L = W[18]; T2O = W[19]; T2P = T2t - T2u; T2v = T2t + T2u; T2N = T2L * T2M; T2X = T2O * T2M; T2n = W[6]; T2s = W[7]; T2S = T2y - T2z; T2A = T2y + T2z; T2F = T2D - T2E; T2V = T2D + T2E; T2r = T2n * T2q; T2H = T2s * T2q; T2x = W[8]; T2R = W[20]; T2J = T2x * T2F; T2B = T2x * T2A; } T2w = FNMS(T2s, T2v, T2r); T2Z = T2R * T2V; T2T = T2R * T2S; T2I = FMA(T2n, T2v, T2H); T2C = W[9]; T2Q = FNMS(T2O, T2P, T2N); T2Y = FMA(T2L, T2P, T2X); T2U = W[21]; T2K = FNMS(T2C, T2A, T2J); T2G = FMA(T2C, T2F, T2B); } } } } T30 = FNMS(T2U, T2S, T2Z); T2W = FMA(T2U, T2V, T2T); Im[WS(rs, 2)] = T2K - T2I; Ip[WS(rs, 2)] = T2I + T2K; Rm[WS(rs, 2)] = T2w + T2G; Rp[WS(rs, 2)] = T2w - T2G; Im[WS(rs, 5)] = T30 - T2Y; Ip[WS(rs, 5)] = T2Y + T30; Rm[WS(rs, 5)] = T2Q + T2W; Rp[WS(rs, 5)] = T2Q - T2W; } } } static const tw_instr twinstr[] = { {TW_FULL, 1, 12}, {TW_NEXT, 1, 0} }; static const hc2c_desc desc = { 12, "hc2cbdft_12", twinstr, &GENUS, {96, 22, 46, 0} }; void X(codelet_hc2cbdft_12) (planner *p) { X(khc2c_register) (p, hc2cbdft_12, &desc, HC2C_VIA_DFT); } #else /* HAVE_FMA */ /* Generated by: ../../../genfft/gen_hc2cdft.native -compact -variables 4 -pipeline-latency 4 -sign 1 -n 12 -dif -name hc2cbdft_12 -include hc2cb.h */ /* * This function contains 142 FP additions, 60 FP multiplications, * (or, 112 additions, 30 multiplications, 30 fused multiply/add), * 47 stack variables, 2 constants, and 48 memory accesses */ #include "hc2cb.h" static void hc2cbdft_12(R *Rp, R *Ip, R *Rm, R *Im, const R *W, stride rs, INT mb, INT me, INT ms) { DK(KP500000000, +0.500000000000000000000000000000000000000000000); DK(KP866025403, +0.866025403784438646763723170752936183471402627); { INT m; for (m = mb, W = W + ((mb - 1) * 22); m < me; m = m + 1, Rp = Rp + ms, Ip = Ip + ms, Rm = Rm - ms, Im = Im - ms, W = W + 22, MAKE_VOLATILE_STRIDE(48, rs)) { E Tv, T1E, TC, T1F, TW, T1x, TT, T1w, T1d, T1N, Tb, T1R, TI, T1z, TN; E T1A, T17, T1I, T12, T1H, T1g, T1S, Tm, T1O; { E T1, Tq, T6, TA, T4, Tp, Tt, TS, T9, Tw, Tz, TV; T1 = Rp[0]; Tq = Ip[0]; T6 = Rm[WS(rs, 5)]; TA = Im[WS(rs, 5)]; { E T2, T3, Tr, Ts; T2 = Rp[WS(rs, 4)]; T3 = Rm[WS(rs, 3)]; T4 = T2 + T3; Tp = KP866025403 * (T2 - T3); Tr = Im[WS(rs, 3)]; Ts = Ip[WS(rs, 4)]; Tt = Tr - Ts; TS = KP866025403 * (Tr + Ts); } { E T7, T8, Tx, Ty; T7 = Rm[WS(rs, 1)]; T8 = Rp[WS(rs, 2)]; T9 = T7 + T8; Tw = KP866025403 * (T7 - T8); Tx = Im[WS(rs, 1)]; Ty = Ip[WS(rs, 2)]; Tz = Tx - Ty; TV = KP866025403 * (Tx + Ty); } { E Tu, TB, TU, TR; Tu = FMA(KP500000000, Tt, Tq); Tv = Tp + Tu; T1E = Tu - Tp; TB = FMS(KP500000000, Tz, TA); TC = Tw + TB; T1F = TB - Tw; TU = FNMS(KP500000000, T9, T6); TW = TU + TV; T1x = TU - TV; TR = FNMS(KP500000000, T4, T1); TT = TR - TS; T1w = TR + TS; { E T1b, T1c, T5, Ta; T1b = Tq - Tt; T1c = Tz + TA; T1d = T1b - T1c; T1N = T1b + T1c; T5 = T1 + T4; Ta = T6 + T9; Tb = T5 + Ta; T1R = T5 - Ta; } } } { E Tc, T10, Th, T15, Tf, TY, TH, TZ, Tk, T13, TM, T14; Tc = Rp[WS(rs, 3)]; T10 = Ip[WS(rs, 3)]; Th = Rm[WS(rs, 2)]; T15 = Im[WS(rs, 2)]; { E Td, Te, TF, TG; Td = Rm[WS(rs, 4)]; Te = Rm[0]; Tf = Td + Te; TY = KP866025403 * (Td - Te); TF = Im[WS(rs, 4)]; TG = Im[0]; TH = KP866025403 * (TF - TG); TZ = TF + TG; } { E Ti, Tj, TK, TL; Ti = Rp[WS(rs, 1)]; Tj = Rp[WS(rs, 5)]; Tk = Ti + Tj; T13 = KP866025403 * (Ti - Tj); TK = Ip[WS(rs, 5)]; TL = Ip[WS(rs, 1)]; TM = KP866025403 * (TK - TL); T14 = TK + TL; } { E TE, TJ, T16, T11; TE = FNMS(KP500000000, Tf, Tc); TI = TE + TH; T1z = TE - TH; TJ = FNMS(KP500000000, Tk, Th); TN = TJ + TM; T1A = TJ - TM; T16 = FMA(KP500000000, T14, T15); T17 = T13 - T16; T1I = T13 + T16; T11 = FMA(KP500000000, TZ, T10); T12 = TY + T11; T1H = T11 - TY; { E T1e, T1f, Tg, Tl; T1e = T10 - TZ; T1f = T14 - T15; T1g = T1e + T1f; T1S = T1e - T1f; Tg = Tc + Tf; Tl = Th + Tk; Tm = Tg + Tl; T1O = Tg - Tl; } } } { E Tn, T1h, TP, T1p, T19, T1r, T1n, T1t; Tn = Tb + Tm; T1h = T1d + T1g; { E TD, TO, TX, T18; TD = Tv - TC; TO = TI - TN; TP = TD + TO; T1p = TD - TO; TX = TT - TW; T18 = T12 - T17; T19 = TX - T18; T1r = TX + T18; { E T1k, T1m, T1j, T1l; T1k = Tb - Tm; T1m = T1d - T1g; T1j = W[10]; T1l = W[11]; T1n = FNMS(T1l, T1m, T1j * T1k); T1t = FMA(T1l, T1k, T1j * T1m); } } { E T1a, T1i, To, TQ; To = W[0]; TQ = W[1]; T1a = FMA(To, TP, TQ * T19); T1i = FNMS(TQ, TP, To * T19); Rp[0] = Tn - T1a; Ip[0] = T1h + T1i; Rm[0] = Tn + T1a; Im[0] = T1i - T1h; } { E T1s, T1u, T1o, T1q; T1o = W[12]; T1q = W[13]; T1s = FMA(T1o, T1p, T1q * T1r); T1u = FNMS(T1q, T1p, T1o * T1r); Rp[WS(rs, 3)] = T1n - T1s; Ip[WS(rs, 3)] = T1t + T1u; Rm[WS(rs, 3)] = T1n + T1s; Im[WS(rs, 3)] = T1u - T1t; } } { E T1C, T1Y, T1K, T20, T1U, T1V, T26, T27; { E T1y, T1B, T1G, T1J; T1y = T1w + T1x; T1B = T1z + T1A; T1C = T1y - T1B; T1Y = T1y + T1B; T1G = T1E + T1F; T1J = T1H - T1I; T1K = T1G - T1J; T20 = T1G + T1J; } { E T1P, T1T, T1M, T1Q; T1P = T1N - T1O; T1T = T1R + T1S; T1M = W[4]; T1Q = W[5]; T1U = FMA(T1M, T1P, T1Q * T1T); T1V = FNMS(T1Q, T1P, T1M * T1T); } { E T23, T25, T22, T24; T23 = T1O + T1N; T25 = T1R - T1S; T22 = W[16]; T24 = W[17]; T26 = FMA(T22, T23, T24 * T25); T27 = FNMS(T24, T23, T22 * T25); } { E T1L, T1W, T1v, T1D; T1v = W[2]; T1D = W[3]; T1L = FNMS(T1D, T1K, T1v * T1C); T1W = FMA(T1D, T1C, T1v * T1K); Rp[WS(rs, 1)] = T1L - T1U; Ip[WS(rs, 1)] = T1V + T1W; Rm[WS(rs, 1)] = T1U + T1L; Im[WS(rs, 1)] = T1V - T1W; } { E T21, T28, T1X, T1Z; T1X = W[14]; T1Z = W[15]; T21 = FNMS(T1Z, T20, T1X * T1Y); T28 = FMA(T1Z, T1Y, T1X * T20); Rp[WS(rs, 4)] = T21 - T26; Ip[WS(rs, 4)] = T27 + T28; Rm[WS(rs, 4)] = T26 + T21; Im[WS(rs, 4)] = T27 - T28; } } { E T2c, T2u, T2p, T2B, T2g, T2w, T2l, T2z; { E T2a, T2b, T2n, T2o; T2a = TT + TW; T2b = TI + TN; T2c = T2a + T2b; T2u = T2a - T2b; T2n = T1w - T1x; T2o = T1H + T1I; T2p = T2n - T2o; T2B = T2n + T2o; } { E T2e, T2f, T2j, T2k; T2e = Tv + TC; T2f = T12 + T17; T2g = T2e + T2f; T2w = T2e - T2f; T2j = T1E - T1F; T2k = T1z - T1A; T2l = T2j + T2k; T2z = T2j - T2k; } { E T2h, T2r, T2q, T2s; { E T29, T2d, T2i, T2m; T29 = W[6]; T2d = W[7]; T2h = FNMS(T2d, T2g, T29 * T2c); T2r = FMA(T2d, T2c, T29 * T2g); T2i = W[8]; T2m = W[9]; T2q = FMA(T2i, T2l, T2m * T2p); T2s = FNMS(T2m, T2l, T2i * T2p); } Rp[WS(rs, 2)] = T2h - T2q; Ip[WS(rs, 2)] = T2r + T2s; Rm[WS(rs, 2)] = T2h + T2q; Im[WS(rs, 2)] = T2s - T2r; } { E T2x, T2D, T2C, T2E; { E T2t, T2v, T2y, T2A; T2t = W[18]; T2v = W[19]; T2x = FNMS(T2v, T2w, T2t * T2u); T2D = FMA(T2v, T2u, T2t * T2w); T2y = W[20]; T2A = W[21]; T2C = FMA(T2y, T2z, T2A * T2B); T2E = FNMS(T2A, T2z, T2y * T2B); } Rp[WS(rs, 5)] = T2x - T2C; Ip[WS(rs, 5)] = T2D + T2E; Rm[WS(rs, 5)] = T2x + T2C; Im[WS(rs, 5)] = T2E - T2D; } } } } } static const tw_instr twinstr[] = { {TW_FULL, 1, 12}, {TW_NEXT, 1, 0} }; static const hc2c_desc desc = { 12, "hc2cbdft_12", twinstr, &GENUS, {112, 30, 30, 0} }; void X(codelet_hc2cbdft_12) (planner *p) { X(khc2c_register) (p, hc2cbdft_12, &desc, HC2C_VIA_DFT); } #endif /* HAVE_FMA */ fftw-3.3.3/rdft/scalar/r2cb/hc2cbdft_16.c0000644000175000017500000005436012054411036014600 00000000000000/* * Copyright (c) 2003, 2007-11 Matteo Frigo * Copyright (c) 2003, 2007-11 Massachusetts Institute of Technology * * This program is free software; you can redistribute it and/or modify * it under the terms of the GNU General Public License as published by * the Free Software Foundation; either version 2 of the License, or * (at your option) any later version. * * This program is distributed in the hope that it will be useful, * but WITHOUT ANY WARRANTY; without even the implied warranty of * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the * GNU General Public License for more details. * * You should have received a copy of the GNU General Public License * along with this program; if not, write to the Free Software * Foundation, Inc., 51 Franklin Street, Fifth Floor, Boston, MA 02110-1301 USA * */ /* This file was automatically generated --- DO NOT EDIT */ /* Generated on Sun Nov 25 07:42:05 EST 2012 */ #include "codelet-rdft.h" #ifdef HAVE_FMA /* Generated by: ../../../genfft/gen_hc2cdft.native -fma -reorder-insns -schedule-for-pipeline -compact -variables 4 -pipeline-latency 4 -sign 1 -n 16 -dif -name hc2cbdft_16 -include hc2cb.h */ /* * This function contains 206 FP additions, 100 FP multiplications, * (or, 136 additions, 30 multiplications, 70 fused multiply/add), * 97 stack variables, 3 constants, and 64 memory accesses */ #include "hc2cb.h" static void hc2cbdft_16(R *Rp, R *Ip, R *Rm, R *Im, const R *W, stride rs, INT mb, INT me, INT ms) { DK(KP923879532, +0.923879532511286756128183189396788286822416626); DK(KP414213562, +0.414213562373095048801688724209698078569671875); DK(KP707106781, +0.707106781186547524400844362104849039284835938); { INT m; for (m = mb, W = W + ((mb - 1) * 30); m < me; m = m + 1, Rp = Rp + ms, Ip = Ip + ms, Rm = Rm - ms, Im = Im - ms, W = W + 30, MAKE_VOLATILE_STRIDE(64, rs)) { E T3w, T3z, T2Y, T3D, T3x, T3m, T3u, T3C, T3y, T3o, T3k, T3E, T3A; { E T20, Tf, T3Q, T32, T3V, T3f, T2a, TN, T2f, T1m, T3G, T2G, T3L, T2T, T26; E T1F, T3M, T2N, T3H, T2W, T25, Tu, T1n, T1o, T3R, T3i, T2g, T1a, T21, T1y; E T3W, T39; { E T2R, T1B, T2S, T1E; { E T1e, T3, T1C, TA, Tx, T6, T1D, T1h, Td, T1A, TL, T1k, Ta, TC, TF; E T1z; { E T4, T5, T1f, T1g; { E T1, T2, Ty, Tz; T1 = Rp[0]; T2 = Rm[WS(rs, 7)]; Ty = Ip[0]; Tz = Im[WS(rs, 7)]; T4 = Rp[WS(rs, 4)]; T1e = T1 - T2; T3 = T1 + T2; T1C = Ty - Tz; TA = Ty + Tz; T5 = Rm[WS(rs, 3)]; } T1f = Ip[WS(rs, 4)]; T1g = Im[WS(rs, 3)]; { E Tb, Tc, TI, TJ; Tb = Rm[WS(rs, 1)]; Tx = T4 - T5; T6 = T4 + T5; T1D = T1f - T1g; T1h = T1f + T1g; Tc = Rp[WS(rs, 6)]; TI = Im[WS(rs, 1)]; TJ = Ip[WS(rs, 6)]; { E T8, TH, TK, T9, TD, TE; T8 = Rp[WS(rs, 2)]; Td = Tb + Tc; TH = Tb - Tc; T1A = TJ - TI; TK = TI + TJ; T9 = Rm[WS(rs, 5)]; TD = Ip[WS(rs, 2)]; TE = Im[WS(rs, 5)]; TL = TH + TK; T1k = TH - TK; Ta = T8 + T9; TC = T8 - T9; TF = TD + TE; T1z = TD - TE; } } } { E T2E, TB, T1l, T1i, T3d, T3e, TM, T2F; { E T7, TG, Te, T30, T31, T1j; T2E = T3 - T6; T7 = T3 + T6; T1j = TC - TF; TG = TC + TF; Te = Ta + Td; T2R = Ta - Td; TB = Tx + TA; T30 = TA - Tx; T31 = T1j - T1k; T1l = T1j + T1k; T1i = T1e - T1h; T3d = T1e + T1h; T20 = T7 - Te; Tf = T7 + Te; T3Q = FNMS(KP707106781, T31, T30); T32 = FMA(KP707106781, T31, T30); T3e = TG + TL; TM = TG - TL; } T3V = FMA(KP707106781, T3e, T3d); T3f = FNMS(KP707106781, T3e, T3d); T2a = FNMS(KP707106781, TM, TB); TN = FMA(KP707106781, TM, TB); T2F = T1A - T1z; T1B = T1z + T1A; T2f = FNMS(KP707106781, T1l, T1i); T1m = FMA(KP707106781, T1l, T1i); T3G = T2E - T2F; T2G = T2E + T2F; T2S = T1C - T1D; T1E = T1C + T1D; } } { E T34, TS, T2H, Tm, T1u, T2I, T33, TX, Tq, T14, Tp, T1v, T12, Tr, T15; E T16; { E Tj, TT, Ti, T1s, TR, Tk, TU, TV; { E Tg, Th, TP, TQ; Tg = Rp[WS(rs, 1)]; T3L = T2S - T2R; T2T = T2R + T2S; T26 = T1E - T1B; T1F = T1B + T1E; Th = Rm[WS(rs, 6)]; TP = Ip[WS(rs, 1)]; TQ = Im[WS(rs, 6)]; Tj = Rp[WS(rs, 5)]; TT = Tg - Th; Ti = Tg + Th; T1s = TP - TQ; TR = TP + TQ; Tk = Rm[WS(rs, 2)]; TU = Ip[WS(rs, 5)]; TV = Im[WS(rs, 2)]; } { E Tn, To, T10, T11; Tn = Rm[0]; { E TO, Tl, T1t, TW; TO = Tj - Tk; Tl = Tj + Tk; T1t = TU - TV; TW = TU + TV; T34 = TR - TO; TS = TO + TR; T2H = Ti - Tl; Tm = Ti + Tl; T1u = T1s + T1t; T2I = T1s - T1t; T33 = TT + TW; TX = TT - TW; To = Rp[WS(rs, 7)]; } T10 = Im[0]; T11 = Ip[WS(rs, 7)]; Tq = Rp[WS(rs, 3)]; T14 = Tn - To; Tp = Tn + To; T1v = T11 - T10; T12 = T10 + T11; Tr = Rm[WS(rs, 4)]; T15 = Ip[WS(rs, 3)]; T16 = Im[WS(rs, 4)]; } } { E T13, T1x, T18, T35, T3g, T3h, T38, TY, T19; { E T2U, T2J, T37, Tt, T36, T2V, T2M, T2K, T2L; T2U = T2H + T2I; T2J = T2H - T2I; { E TZ, Ts, T1w, T17; TZ = Tq - Tr; Ts = Tq + Tr; T1w = T15 - T16; T17 = T15 + T16; T37 = TZ + T12; T13 = TZ - T12; T2K = Tp - Ts; Tt = Tp + Ts; T1x = T1v + T1w; T2L = T1v - T1w; T36 = T14 + T17; T18 = T14 - T17; } T2V = T2L - T2K; T2M = T2K + T2L; T3M = T2J - T2M; T2N = T2J + T2M; T3H = T2V - T2U; T2W = T2U + T2V; T35 = FMA(KP414213562, T34, T33); T3g = FNMS(KP414213562, T33, T34); T25 = Tm - Tt; Tu = Tm + Tt; T3h = FNMS(KP414213562, T36, T37); T38 = FMA(KP414213562, T37, T36); } T1n = FNMS(KP414213562, TS, TX); TY = FMA(KP414213562, TX, TS); T19 = FNMS(KP414213562, T18, T13); T1o = FMA(KP414213562, T13, T18); T3R = T3h - T3g; T3i = T3g + T3h; T2g = T19 - TY; T1a = TY + T19; T21 = T1x - T1u; T1y = T1u + T1x; T3W = T35 + T38; T39 = T35 - T38; } } } { E T27, T22, T2c, T2u, T2x, T2h, T2s, T2A, T2w, T2B, T2v; { E T1K, Tv, T1G, T1N, T1Q, T1b, T2b, T1p, Tw, T1d; T1K = Tf - Tu; Tv = Tf + Tu; T1G = T1y + T1F; T1N = T1F - T1y; T1Q = FNMS(KP923879532, T1a, TN); T1b = FMA(KP923879532, T1a, TN); T2b = T1n - T1o; T1p = T1n + T1o; Tw = W[0]; T1d = W[1]; { E T1T, T1O, T1W, T1S, T1X, T1R; { E T1J, T1M, T1L, T1V, T1P, T1q; T1T = FNMS(KP923879532, T1p, T1m); T1q = FMA(KP923879532, T1p, T1m); { E T1c, T1I, T1H, T1r; T1c = Tw * T1b; T1J = W[14]; T1H = Tw * T1q; T1r = FMA(T1d, T1q, T1c); T1M = W[15]; T1L = T1J * T1K; T1I = FNMS(T1d, T1b, T1H); Rm[0] = Tv + T1r; Rp[0] = Tv - T1r; T1V = T1M * T1K; Im[0] = T1I - T1G; Ip[0] = T1G + T1I; T1P = W[16]; } T1O = FNMS(T1M, T1N, T1L); T1W = FMA(T1J, T1N, T1V); T1S = W[17]; T1X = T1P * T1T; T1R = T1P * T1Q; } { E T2r, T2n, T2q, T2p, T2z, T2t, T2o, T1Y, T1U; T27 = T25 + T26; T2r = T26 - T25; T2o = T20 - T21; T22 = T20 + T21; T1Y = FNMS(T1S, T1Q, T1X); T1U = FMA(T1S, T1T, T1R); T2n = W[22]; T2q = W[23]; Im[WS(rs, 4)] = T1Y - T1W; Ip[WS(rs, 4)] = T1W + T1Y; Rm[WS(rs, 4)] = T1O + T1U; Rp[WS(rs, 4)] = T1O - T1U; T2p = T2n * T2o; T2z = T2q * T2o; T2c = FMA(KP923879532, T2b, T2a); T2u = FNMS(KP923879532, T2b, T2a); T2x = FNMS(KP923879532, T2g, T2f); T2h = FMA(KP923879532, T2g, T2f); T2t = W[24]; T2s = FNMS(T2q, T2r, T2p); T2A = FMA(T2n, T2r, T2z); T2w = W[25]; T2B = T2t * T2x; T2v = T2t * T2u; } } } { E T28, T2k, T2e, T2l, T2d; { E T1Z, T24, T23, T2j, T29, T2C, T2y; T2C = FNMS(T2w, T2u, T2B); T2y = FMA(T2w, T2x, T2v); T1Z = W[6]; T24 = W[7]; Im[WS(rs, 6)] = T2C - T2A; Ip[WS(rs, 6)] = T2A + T2C; Rm[WS(rs, 6)] = T2s + T2y; Rp[WS(rs, 6)] = T2s - T2y; T23 = T1Z * T22; T2j = T24 * T22; T29 = W[8]; T28 = FNMS(T24, T27, T23); T2k = FMA(T1Z, T27, T2j); T2e = W[9]; T2l = T29 * T2h; T2d = T29 * T2c; } { E T4a, T4d, T3O, T4h, T4b, T40, T48, T4g, T4c, T42, T3Y; { E T3N, T47, T43, T46, T3F, T45, T4f, T3K, T3J, T3S, T3X, T3Z, T49, T41, T3T; E T3U; { E T44, T3I, T2m, T2i, T3P; T44 = FNMS(KP707106781, T3H, T3G); T3I = FMA(KP707106781, T3H, T3G); T2m = FNMS(T2e, T2c, T2l); T2i = FMA(T2e, T2h, T2d); T3N = FMA(KP707106781, T3M, T3L); T47 = FNMS(KP707106781, T3M, T3L); Im[WS(rs, 2)] = T2m - T2k; Ip[WS(rs, 2)] = T2k + T2m; Rm[WS(rs, 2)] = T28 + T2i; Rp[WS(rs, 2)] = T28 - T2i; T43 = W[26]; T46 = W[27]; T3F = W[10]; T45 = T43 * T44; T4f = T46 * T44; T3K = W[11]; T3J = T3F * T3I; T4a = FNMS(KP923879532, T3R, T3Q); T3S = FMA(KP923879532, T3R, T3Q); T3X = FNMS(KP923879532, T3W, T3V); T4d = FMA(KP923879532, T3W, T3V); T3Z = T3K * T3I; T3P = W[12]; T49 = W[28]; T41 = T3P * T3X; T3T = T3P * T3S; } T3O = FNMS(T3K, T3N, T3J); T4h = T49 * T4d; T4b = T49 * T4a; T40 = FMA(T3F, T3N, T3Z); T3U = W[13]; T48 = FNMS(T46, T47, T45); T4g = FMA(T43, T47, T4f); T4c = W[29]; T42 = FNMS(T3U, T3S, T41); T3Y = FMA(T3U, T3X, T3T); } { E T3t, T2X, T3p, T3s, T2D, T3r, T3B, T2Q, T2P, T3a, T3j, T3l, T3v, T3n, T3b; E T3c; { E T2O, T3q, T4i, T4e, T2Z; T4i = FNMS(T4c, T4a, T4h); T4e = FMA(T4c, T4d, T4b); Im[WS(rs, 3)] = T42 - T40; Ip[WS(rs, 3)] = T40 + T42; Rm[WS(rs, 3)] = T3O + T3Y; Rp[WS(rs, 3)] = T3O - T3Y; Im[WS(rs, 7)] = T4i - T4g; Ip[WS(rs, 7)] = T4g + T4i; Rm[WS(rs, 7)] = T48 + T4e; Rp[WS(rs, 7)] = T48 - T4e; T3t = FNMS(KP707106781, T2W, T2T); T2X = FMA(KP707106781, T2W, T2T); T2O = FMA(KP707106781, T2N, T2G); T3q = FNMS(KP707106781, T2N, T2G); T3p = W[18]; T3s = W[19]; T2D = W[2]; T3r = T3p * T3q; T3B = T3s * T3q; T2Q = W[3]; T2P = T2D * T2O; T3a = FMA(KP923879532, T39, T32); T3w = FNMS(KP923879532, T39, T32); T3z = FMA(KP923879532, T3i, T3f); T3j = FNMS(KP923879532, T3i, T3f); T3l = T2Q * T2O; T2Z = W[4]; T3v = W[20]; T3n = T2Z * T3j; T3b = T2Z * T3a; } T2Y = FNMS(T2Q, T2X, T2P); T3D = T3v * T3z; T3x = T3v * T3w; T3m = FMA(T2D, T2X, T3l); T3c = W[5]; T3u = FNMS(T3s, T3t, T3r); T3C = FMA(T3p, T3t, T3B); T3y = W[21]; T3o = FNMS(T3c, T3a, T3n); T3k = FMA(T3c, T3j, T3b); } } } } } T3E = FNMS(T3y, T3w, T3D); T3A = FMA(T3y, T3z, T3x); Im[WS(rs, 1)] = T3o - T3m; Ip[WS(rs, 1)] = T3m + T3o; Rm[WS(rs, 1)] = T2Y + T3k; Rp[WS(rs, 1)] = T2Y - T3k; Im[WS(rs, 5)] = T3E - T3C; Ip[WS(rs, 5)] = T3C + T3E; Rm[WS(rs, 5)] = T3u + T3A; Rp[WS(rs, 5)] = T3u - T3A; } } } static const tw_instr twinstr[] = { {TW_FULL, 1, 16}, {TW_NEXT, 1, 0} }; static const hc2c_desc desc = { 16, "hc2cbdft_16", twinstr, &GENUS, {136, 30, 70, 0} }; void X(codelet_hc2cbdft_16) (planner *p) { X(khc2c_register) (p, hc2cbdft_16, &desc, HC2C_VIA_DFT); } #else /* HAVE_FMA */ /* Generated by: ../../../genfft/gen_hc2cdft.native -compact -variables 4 -pipeline-latency 4 -sign 1 -n 16 -dif -name hc2cbdft_16 -include hc2cb.h */ /* * This function contains 206 FP additions, 84 FP multiplications, * (or, 168 additions, 46 multiplications, 38 fused multiply/add), * 60 stack variables, 3 constants, and 64 memory accesses */ #include "hc2cb.h" static void hc2cbdft_16(R *Rp, R *Ip, R *Rm, R *Im, const R *W, stride rs, INT mb, INT me, INT ms) { DK(KP923879532, +0.923879532511286756128183189396788286822416626); DK(KP382683432, +0.382683432365089771728459984030398866761344562); DK(KP707106781, +0.707106781186547524400844362104849039284835938); { INT m; for (m = mb, W = W + ((mb - 1) * 30); m < me; m = m + 1, Rp = Rp + ms, Ip = Ip + ms, Rm = Rm - ms, Im = Im - ms, W = W + 30, MAKE_VOLATILE_STRIDE(64, rs)) { E TB, T2L, T30, T1n, Tf, T1U, T2H, T3p, T1E, T1Z, TM, T31, T2s, T3k, T1i; E T2M, Tu, T1Y, T2Q, T2X, T2T, T2Y, TY, T1d, T19, T1e, T2v, T2C, T2y, T2D; E T1x, T1V; { E T3, T1j, TA, T1B, T6, Tx, T1m, T1C, Ta, TC, TF, T1y, Td, TH, TK; E T1z; { E T1, T2, Ty, Tz; T1 = Rp[0]; T2 = Rm[WS(rs, 7)]; T3 = T1 + T2; T1j = T1 - T2; Ty = Ip[0]; Tz = Im[WS(rs, 7)]; TA = Ty + Tz; T1B = Ty - Tz; } { E T4, T5, T1k, T1l; T4 = Rp[WS(rs, 4)]; T5 = Rm[WS(rs, 3)]; T6 = T4 + T5; Tx = T4 - T5; T1k = Ip[WS(rs, 4)]; T1l = Im[WS(rs, 3)]; T1m = T1k + T1l; T1C = T1k - T1l; } { E T8, T9, TD, TE; T8 = Rp[WS(rs, 2)]; T9 = Rm[WS(rs, 5)]; Ta = T8 + T9; TC = T8 - T9; TD = Ip[WS(rs, 2)]; TE = Im[WS(rs, 5)]; TF = TD + TE; T1y = TD - TE; } { E Tb, Tc, TI, TJ; Tb = Rm[WS(rs, 1)]; Tc = Rp[WS(rs, 6)]; Td = Tb + Tc; TH = Tb - Tc; TI = Im[WS(rs, 1)]; TJ = Ip[WS(rs, 6)]; TK = TI + TJ; T1z = TJ - TI; } { E T7, Te, TG, TL; TB = Tx + TA; T2L = TA - Tx; T30 = T1j + T1m; T1n = T1j - T1m; T7 = T3 + T6; Te = Ta + Td; Tf = T7 + Te; T1U = T7 - Te; { E T2F, T2G, T1A, T1D; T2F = Ta - Td; T2G = T1B - T1C; T2H = T2F + T2G; T3p = T2G - T2F; T1A = T1y + T1z; T1D = T1B + T1C; T1E = T1A + T1D; T1Z = T1D - T1A; } TG = TC + TF; TL = TH + TK; TM = KP707106781 * (TG - TL); T31 = KP707106781 * (TG + TL); { E T2q, T2r, T1g, T1h; T2q = T3 - T6; T2r = T1z - T1y; T2s = T2q + T2r; T3k = T2q - T2r; T1g = TC - TF; T1h = TH - TK; T1i = KP707106781 * (T1g + T1h); T2M = KP707106781 * (T1g - T1h); } } } { E Ti, TT, TR, T1r, Tl, TO, TW, T1s, Tp, T14, T12, T1u, Ts, TZ, T17; E T1v; { E Tg, Th, TP, TQ; Tg = Rp[WS(rs, 1)]; Th = Rm[WS(rs, 6)]; Ti = Tg + Th; TT = Tg - Th; TP = Ip[WS(rs, 1)]; TQ = Im[WS(rs, 6)]; TR = TP + TQ; T1r = TP - TQ; } { E Tj, Tk, TU, TV; Tj = Rp[WS(rs, 5)]; Tk = Rm[WS(rs, 2)]; Tl = Tj + Tk; TO = Tj - Tk; TU = Ip[WS(rs, 5)]; TV = Im[WS(rs, 2)]; TW = TU + TV; T1s = TU - TV; } { E Tn, To, T10, T11; Tn = Rm[0]; To = Rp[WS(rs, 7)]; Tp = Tn + To; T14 = Tn - To; T10 = Im[0]; T11 = Ip[WS(rs, 7)]; T12 = T10 + T11; T1u = T11 - T10; } { E Tq, Tr, T15, T16; Tq = Rp[WS(rs, 3)]; Tr = Rm[WS(rs, 4)]; Ts = Tq + Tr; TZ = Tq - Tr; T15 = Ip[WS(rs, 3)]; T16 = Im[WS(rs, 4)]; T17 = T15 + T16; T1v = T15 - T16; } { E Tm, Tt, T2O, T2P; Tm = Ti + Tl; Tt = Tp + Ts; Tu = Tm + Tt; T1Y = Tm - Tt; T2O = TR - TO; T2P = TT + TW; T2Q = FMA(KP382683432, T2O, KP923879532 * T2P); T2X = FNMS(KP923879532, T2O, KP382683432 * T2P); } { E T2R, T2S, TS, TX; T2R = TZ + T12; T2S = T14 + T17; T2T = FMA(KP382683432, T2R, KP923879532 * T2S); T2Y = FNMS(KP923879532, T2R, KP382683432 * T2S); TS = TO + TR; TX = TT - TW; TY = FMA(KP923879532, TS, KP382683432 * TX); T1d = FNMS(KP382683432, TS, KP923879532 * TX); } { E T13, T18, T2t, T2u; T13 = TZ - T12; T18 = T14 - T17; T19 = FNMS(KP382683432, T18, KP923879532 * T13); T1e = FMA(KP382683432, T13, KP923879532 * T18); T2t = Ti - Tl; T2u = T1r - T1s; T2v = T2t - T2u; T2C = T2t + T2u; } { E T2w, T2x, T1t, T1w; T2w = Tp - Ts; T2x = T1u - T1v; T2y = T2w + T2x; T2D = T2x - T2w; T1t = T1r + T1s; T1w = T1u + T1v; T1x = T1t + T1w; T1V = T1w - T1t; } } { E Tv, T1F, T1b, T1N, T1p, T1P, T1L, T1R; Tv = Tf + Tu; T1F = T1x + T1E; { E TN, T1a, T1f, T1o; TN = TB + TM; T1a = TY + T19; T1b = TN + T1a; T1N = TN - T1a; T1f = T1d + T1e; T1o = T1i + T1n; T1p = T1f + T1o; T1P = T1o - T1f; { E T1I, T1K, T1H, T1J; T1I = Tf - Tu; T1K = T1E - T1x; T1H = W[14]; T1J = W[15]; T1L = FNMS(T1J, T1K, T1H * T1I); T1R = FMA(T1J, T1I, T1H * T1K); } } { E T1q, T1G, Tw, T1c; Tw = W[0]; T1c = W[1]; T1q = FMA(Tw, T1b, T1c * T1p); T1G = FNMS(T1c, T1b, Tw * T1p); Rp[0] = Tv - T1q; Ip[0] = T1F + T1G; Rm[0] = Tv + T1q; Im[0] = T1G - T1F; } { E T1Q, T1S, T1M, T1O; T1M = W[16]; T1O = W[17]; T1Q = FMA(T1M, T1N, T1O * T1P); T1S = FNMS(T1O, T1N, T1M * T1P); Rp[WS(rs, 4)] = T1L - T1Q; Ip[WS(rs, 4)] = T1R + T1S; Rm[WS(rs, 4)] = T1L + T1Q; Im[WS(rs, 4)] = T1S - T1R; } } { E T25, T2j, T29, T2l, T21, T2b, T2h, T2n; { E T23, T24, T27, T28; T23 = TB - TM; T24 = T1d - T1e; T25 = T23 + T24; T2j = T23 - T24; T27 = T19 - TY; T28 = T1n - T1i; T29 = T27 + T28; T2l = T28 - T27; } { E T1W, T20, T1T, T1X; T1W = T1U + T1V; T20 = T1Y + T1Z; T1T = W[6]; T1X = W[7]; T21 = FNMS(T1X, T20, T1T * T1W); T2b = FMA(T1X, T1W, T1T * T20); } { E T2e, T2g, T2d, T2f; T2e = T1U - T1V; T2g = T1Z - T1Y; T2d = W[22]; T2f = W[23]; T2h = FNMS(T2f, T2g, T2d * T2e); T2n = FMA(T2f, T2e, T2d * T2g); } { E T2a, T2c, T22, T26; T22 = W[8]; T26 = W[9]; T2a = FMA(T22, T25, T26 * T29); T2c = FNMS(T26, T25, T22 * T29); Rp[WS(rs, 2)] = T21 - T2a; Ip[WS(rs, 2)] = T2b + T2c; Rm[WS(rs, 2)] = T21 + T2a; Im[WS(rs, 2)] = T2c - T2b; } { E T2m, T2o, T2i, T2k; T2i = W[24]; T2k = W[25]; T2m = FMA(T2i, T2j, T2k * T2l); T2o = FNMS(T2k, T2j, T2i * T2l); Rp[WS(rs, 6)] = T2h - T2m; Ip[WS(rs, 6)] = T2n + T2o; Rm[WS(rs, 6)] = T2h + T2m; Im[WS(rs, 6)] = T2o - T2n; } } { E T2A, T38, T2I, T3a, T2V, T3d, T33, T3f, T2z, T2E; T2z = KP707106781 * (T2v + T2y); T2A = T2s + T2z; T38 = T2s - T2z; T2E = KP707106781 * (T2C + T2D); T2I = T2E + T2H; T3a = T2H - T2E; { E T2N, T2U, T2Z, T32; T2N = T2L + T2M; T2U = T2Q - T2T; T2V = T2N + T2U; T3d = T2N - T2U; T2Z = T2X + T2Y; T32 = T30 - T31; T33 = T2Z + T32; T3f = T32 - T2Z; } { E T2J, T35, T34, T36; { E T2p, T2B, T2K, T2W; T2p = W[2]; T2B = W[3]; T2J = FNMS(T2B, T2I, T2p * T2A); T35 = FMA(T2B, T2A, T2p * T2I); T2K = W[4]; T2W = W[5]; T34 = FMA(T2K, T2V, T2W * T33); T36 = FNMS(T2W, T2V, T2K * T33); } Rp[WS(rs, 1)] = T2J - T34; Ip[WS(rs, 1)] = T35 + T36; Rm[WS(rs, 1)] = T2J + T34; Im[WS(rs, 1)] = T36 - T35; } { E T3b, T3h, T3g, T3i; { E T37, T39, T3c, T3e; T37 = W[18]; T39 = W[19]; T3b = FNMS(T39, T3a, T37 * T38); T3h = FMA(T39, T38, T37 * T3a); T3c = W[20]; T3e = W[21]; T3g = FMA(T3c, T3d, T3e * T3f); T3i = FNMS(T3e, T3d, T3c * T3f); } Rp[WS(rs, 5)] = T3b - T3g; Ip[WS(rs, 5)] = T3h + T3i; Rm[WS(rs, 5)] = T3b + T3g; Im[WS(rs, 5)] = T3i - T3h; } } { E T3m, T3E, T3q, T3G, T3v, T3J, T3z, T3L, T3l, T3o; T3l = KP707106781 * (T2D - T2C); T3m = T3k + T3l; T3E = T3k - T3l; T3o = KP707106781 * (T2v - T2y); T3q = T3o + T3p; T3G = T3p - T3o; { E T3t, T3u, T3x, T3y; T3t = T2L - T2M; T3u = T2X - T2Y; T3v = T3t + T3u; T3J = T3t - T3u; T3x = T31 + T30; T3y = T2Q + T2T; T3z = T3x - T3y; T3L = T3y + T3x; } { E T3r, T3B, T3A, T3C; { E T3j, T3n, T3s, T3w; T3j = W[10]; T3n = W[11]; T3r = FNMS(T3n, T3q, T3j * T3m); T3B = FMA(T3n, T3m, T3j * T3q); T3s = W[12]; T3w = W[13]; T3A = FMA(T3s, T3v, T3w * T3z); T3C = FNMS(T3w, T3v, T3s * T3z); } Rp[WS(rs, 3)] = T3r - T3A; Ip[WS(rs, 3)] = T3B + T3C; Rm[WS(rs, 3)] = T3r + T3A; Im[WS(rs, 3)] = T3C - T3B; } { E T3H, T3N, T3M, T3O; { E T3D, T3F, T3I, T3K; T3D = W[26]; T3F = W[27]; T3H = FNMS(T3F, T3G, T3D * T3E); T3N = FMA(T3F, T3E, T3D * T3G); T3I = W[28]; T3K = W[29]; T3M = FMA(T3I, T3J, T3K * T3L); T3O = FNMS(T3K, T3J, T3I * T3L); } Rp[WS(rs, 7)] = T3H - T3M; Ip[WS(rs, 7)] = T3N + T3O; Rm[WS(rs, 7)] = T3H + T3M; Im[WS(rs, 7)] = T3O - T3N; } } } } } static const tw_instr twinstr[] = { {TW_FULL, 1, 16}, {TW_NEXT, 1, 0} }; static const hc2c_desc desc = { 16, "hc2cbdft_16", twinstr, &GENUS, {168, 46, 38, 0} }; void X(codelet_hc2cbdft_16) (planner *p) { X(khc2c_register) (p, hc2cbdft_16, &desc, HC2C_VIA_DFT); } #endif /* HAVE_FMA */ fftw-3.3.3/rdft/scalar/r2cb/hc2cbdft_32.c0000644000175000017500000014437712054411046014607 00000000000000/* * Copyright (c) 2003, 2007-11 Matteo Frigo * Copyright (c) 2003, 2007-11 Massachusetts Institute of Technology * * This program is free software; you can redistribute it and/or modify * it under the terms of the GNU General Public License as published by * the Free Software Foundation; either version 2 of the License, or * (at your option) any later version. * * This program is distributed in the hope that it will be useful, * but WITHOUT ANY WARRANTY; without even the implied warranty of * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the * GNU General Public License for more details. * * You should have received a copy of the GNU General Public License * along with this program; if not, write to the Free Software * Foundation, Inc., 51 Franklin Street, Fifth Floor, Boston, MA 02110-1301 USA * */ /* This file was automatically generated --- DO NOT EDIT */ /* Generated on Sun Nov 25 07:42:06 EST 2012 */ #include "codelet-rdft.h" #ifdef HAVE_FMA /* Generated by: ../../../genfft/gen_hc2cdft.native -fma -reorder-insns -schedule-for-pipeline -compact -variables 4 -pipeline-latency 4 -sign 1 -n 32 -dif -name hc2cbdft_32 -include hc2cb.h */ /* * This function contains 498 FP additions, 260 FP multiplications, * (or, 300 additions, 62 multiplications, 198 fused multiply/add), * 165 stack variables, 7 constants, and 128 memory accesses */ #include "hc2cb.h" static void hc2cbdft_32(R *Rp, R *Ip, R *Rm, R *Im, const R *W, stride rs, INT mb, INT me, INT ms) { DK(KP831469612, +0.831469612302545237078788377617905756738560812); DK(KP980785280, +0.980785280403230449126182236134239036973933731); DK(KP923879532, +0.923879532511286756128183189396788286822416626); DK(KP198912367, +0.198912367379658006911597622644676228597850501); DK(KP668178637, +0.668178637919298919997757686523080761552472251); DK(KP707106781, +0.707106781186547524400844362104849039284835938); DK(KP414213562, +0.414213562373095048801688724209698078569671875); { INT m; for (m = mb, W = W + ((mb - 1) * 62); m < me; m = m + 1, Rp = Rp + ms, Ip = Ip + ms, Rm = Rm - ms, Im = Im - ms, W = W + 62, MAKE_VOLATILE_STRIDE(128, rs)) { E T8e, T8h, T7S, T8l, T8f, T84, T8c, T8k, T8g, T86, T82, T8m, T8i; { E T4B, T3h, T3K, Tv, T8Y, T6T, T8L, T7i, T8X, T7f, T4Y, T1G, T4K, T1j, T4X; E T2M, T8C, T6d, T8o, T66, T8K, T6M, T4L, T2P, T4C, T3o, T5q, T4q, T8p, T6C; E T8B, T6z, T72, T2u, T75, T10, T3P, T3a, T3L, T4t, T4E, T8F, T8t, T4F, T4w; E T8E, T8w, T6E, T6l, T6F, T6s, T76, T4P, T51, T2R, T28, T8P, T90, T7k, T71; E T2p, T4R, T2x, T73, T6x, T6y; { E T3l, T16, T3m, T2H, T2E, T13, T64, T7, T3i, T2J, T1c, T3j, T1h, T2K, Te; E T1z, T6R, T6a, Tt, T3g, T6b, T1E, T6Q, Tj, T1p, Ti, T3b, T1n, Tk, T1q; E T1r; { E T1, T2, T4, T5; { E T14, T15, T2F, T2G; T14 = Ip[0]; T15 = Im[WS(rs, 15)]; T2F = Ip[WS(rs, 8)]; T2G = Im[WS(rs, 7)]; T1 = Rp[0]; T3l = T14 - T15; T16 = T14 + T15; T3m = T2F - T2G; T2H = T2F + T2G; T2 = Rm[WS(rs, 15)]; T4 = Rp[WS(rs, 8)]; T5 = Rm[WS(rs, 7)]; } { E T1b, T1e, T18, Ta, T1f, Tb, Tc, T8, T9, T1g, T1d, Td; { E T19, T3, T6, T1a; T19 = Ip[WS(rs, 4)]; T2E = T1 - T2; T3 = T1 + T2; T13 = T4 - T5; T6 = T4 + T5; T1a = Im[WS(rs, 11)]; T8 = Rp[WS(rs, 4)]; T9 = Rm[WS(rs, 11)]; T64 = T3 - T6; T7 = T3 + T6; T1b = T19 + T1a; T3i = T19 - T1a; } T1e = Im[WS(rs, 3)]; T18 = T8 - T9; Ta = T8 + T9; T1f = Ip[WS(rs, 12)]; Tb = Rm[WS(rs, 3)]; Tc = Rp[WS(rs, 12)]; T2J = T18 - T1b; T1c = T18 + T1b; T1g = T1e + T1f; T3j = T1f - T1e; T1d = Tb - Tc; Td = Tb + Tc; T1h = T1d + T1g; T2K = T1d - T1g; T6x = Ta - Td; Te = Ta + Td; } { E Tq, T1A, Tp, T3e, T1y, Tr, T1B, T1C; { E Tn, To, T1w, T1x; Tn = Rm[WS(rs, 1)]; To = Rp[WS(rs, 14)]; T1w = Im[WS(rs, 1)]; T1x = Ip[WS(rs, 14)]; Tq = Rp[WS(rs, 6)]; T1A = Tn - To; Tp = Tn + To; T3e = T1x - T1w; T1y = T1w + T1x; Tr = Rm[WS(rs, 9)]; T1B = Ip[WS(rs, 6)]; T1C = Im[WS(rs, 9)]; } { E Tg, Th, T1l, T1m; Tg = Rp[WS(rs, 2)]; { E T1v, Ts, T3f, T1D; T1v = Tq - Tr; Ts = Tq + Tr; T3f = T1B - T1C; T1D = T1B + T1C; T1z = T1v - T1y; T6R = T1v + T1y; T6a = Tp - Ts; Tt = Tp + Ts; T3g = T3e + T3f; T6b = T3e - T3f; T1E = T1A - T1D; T6Q = T1A + T1D; Th = Rm[WS(rs, 13)]; } T1l = Ip[WS(rs, 2)]; T1m = Im[WS(rs, 13)]; Tj = Rp[WS(rs, 10)]; T1p = Tg - Th; Ti = Tg + Th; T3b = T1l - T1m; T1n = T1l + T1m; Tk = Rm[WS(rs, 5)]; T1q = Ip[WS(rs, 10)]; T1r = Im[WS(rs, 5)]; } } } { E T4o, T67, T68, T4p, T2I, T1i, T2N, T1u, T1F, T2O, T6K, T17; { E Tf, T1o, T1t, Tu, T7g, T6P, T6S, T7h, T7d, T7e; { E T6O, T6N, T1k, Tl; T4o = T7 - Te; Tf = T7 + Te; T1k = Tj - Tk; Tl = Tj + Tk; { E T3c, T1s, Tm, T3d; T3c = T1q - T1r; T1s = T1q + T1r; T1o = T1k + T1n; T6O = T1n - T1k; T67 = Ti - Tl; Tm = Ti + Tl; T3d = T3b + T3c; T68 = T3b - T3c; T1t = T1p - T1s; T6N = T1p + T1s; T4B = Tm - Tt; Tu = Tm + Tt; T4p = T3g - T3d; T3h = T3d + T3g; } T7g = FNMS(KP414213562, T6N, T6O); T6P = FMA(KP414213562, T6O, T6N); T6S = FMA(KP414213562, T6R, T6Q); T7h = FNMS(KP414213562, T6Q, T6R); } T3K = Tf - Tu; Tv = Tf + Tu; T8Y = T6P + T6S; T6T = T6P - T6S; T2I = T2E - T2H; T7d = T2E + T2H; T7e = T1c + T1h; T1i = T1c - T1h; T2N = FNMS(KP414213562, T1o, T1t); T1u = FMA(KP414213562, T1t, T1o); T8L = T7h - T7g; T7i = T7g + T7h; T8X = FMA(KP707106781, T7e, T7d); T7f = FNMS(KP707106781, T7e, T7d); T1F = FNMS(KP414213562, T1E, T1z); T2O = FMA(KP414213562, T1z, T1E); T6K = T16 - T13; T17 = T13 + T16; } { E T6L, T6A, T6B, T65, T3k, T2L, T69, T6c, T3n; T4Y = T1F - T1u; T1G = T1u + T1F; T4K = FNMS(KP707106781, T1i, T17); T1j = FMA(KP707106781, T1i, T17); T2L = T2J + T2K; T6L = T2J - T2K; T6A = T67 + T68; T69 = T67 - T68; T6c = T6a + T6b; T6B = T6b - T6a; T4X = FNMS(KP707106781, T2L, T2I); T2M = FMA(KP707106781, T2L, T2I); T8C = T69 - T6c; T6d = T69 + T6c; T65 = T3j - T3i; T3k = T3i + T3j; T8o = T64 - T65; T66 = T64 + T65; T8K = FNMS(KP707106781, T6L, T6K); T6M = FMA(KP707106781, T6L, T6K); T3n = T3l + T3m; T6y = T3l - T3m; T4L = T2N - T2O; T2P = T2N + T2O; T4C = T3n - T3k; T3o = T3k + T3n; T5q = T4o - T4p; T4q = T4o + T4p; T8p = T6B - T6A; T6C = T6A + T6B; } } } { E T1M, T6V, T6f, TC, T31, T6j, T23, T6Y, T2v, T2i, TY, T6p, T6n, T35, T2n; E T2w, T24, T1R, TJ, T6i, T6g, T2Y, T1W, T25, T2q, TN, T2r, T36, T2c, T29; E TQ, T2s; { E TU, T2k, T33, T2j, TX, T2l, T2m, T34; { E T1Z, Ty, T20, T2Z, T1L, T1I, TB, T21, T2e, T2h; { E T1J, T1K, Tw, Tx, Tz, TA; Tw = Rp[WS(rs, 1)]; Tx = Rm[WS(rs, 14)]; T1J = Ip[WS(rs, 1)]; T8B = T6y - T6x; T6z = T6x + T6y; T1Z = Tw - Tx; Ty = Tw + Tx; T1K = Im[WS(rs, 14)]; Tz = Rp[WS(rs, 9)]; TA = Rm[WS(rs, 6)]; T20 = Ip[WS(rs, 9)]; T2Z = T1J - T1K; T1L = T1J + T1K; T1I = Tz - TA; TB = Tz + TA; T21 = Im[WS(rs, 6)]; } { E T2f, T2g, TV, TW; { E TS, T30, T22, TT; TS = Rp[WS(rs, 3)]; T1M = T1I + T1L; T6V = T1L - T1I; T6f = Ty - TB; TC = Ty + TB; T30 = T20 - T21; T22 = T20 + T21; TT = Rm[WS(rs, 12)]; T2f = Ip[WS(rs, 3)]; T31 = T2Z + T30; T6j = T2Z - T30; T23 = T1Z - T22; T6Y = T1Z + T22; T2e = TS - TT; TU = TS + TT; T2g = Im[WS(rs, 12)]; } TV = Rm[WS(rs, 4)]; TW = Rp[WS(rs, 11)]; T2k = Im[WS(rs, 4)]; T33 = T2f - T2g; T2h = T2f + T2g; T2j = TV - TW; TX = TV + TW; T2l = Ip[WS(rs, 11)]; } T2v = T2e - T2h; T2i = T2e + T2h; } TY = TU + TX; T6p = TU - TX; T2m = T2k + T2l; T34 = T2l - T2k; { E TF, T1T, T2W, T1S, TI, T1U, T1N, T1Q, T1V, T2X; { E T1O, T1P, TD, TE, TG, TH; TD = Rp[WS(rs, 5)]; TE = Rm[WS(rs, 10)]; T6n = T34 - T33; T35 = T33 + T34; T2n = T2j + T2m; T2w = T2j - T2m; T1N = TD - TE; TF = TD + TE; T1O = Ip[WS(rs, 5)]; T1P = Im[WS(rs, 10)]; TG = Rm[WS(rs, 2)]; TH = Rp[WS(rs, 13)]; T1T = Im[WS(rs, 2)]; T2W = T1O - T1P; T1Q = T1O + T1P; T1S = TG - TH; TI = TG + TH; T1U = Ip[WS(rs, 13)]; } T24 = T1N - T1Q; T1R = T1N + T1Q; TJ = TF + TI; T6i = TF - TI; T1V = T1T + T1U; T2X = T1U - T1T; { E T2a, T2b, TL, TM, TO, TP; TL = Rm[0]; TM = Rp[WS(rs, 15)]; T6g = T2X - T2W; T2Y = T2W + T2X; T1W = T1S + T1V; T25 = T1S - T1V; T2q = TL - TM; TN = TL + TM; T2a = Im[0]; T2b = Ip[WS(rs, 15)]; TO = Rp[WS(rs, 7)]; TP = Rm[WS(rs, 8)]; T2r = Ip[WS(rs, 7)]; T36 = T2b - T2a; T2c = T2a + T2b; T29 = TO - TP; TQ = TO + TP; T2s = Im[WS(rs, 8)]; } } } { E T2d, T4u, T4v, T6r, T6o, T6k, T8u, T8v, T6h; { E T4r, T6m, T32, T4s, T6q, T39, T8r, T8s; { E TK, TR, T37, T2t, TZ, T38; T4r = TC - TJ; TK = TC + TJ; T2d = T29 - T2c; T72 = T29 + T2c; T6m = TN - TQ; TR = TN + TQ; T37 = T2r - T2s; T2t = T2r + T2s; T32 = T2Y + T31; T4s = T31 - T2Y; T4u = TR - TY; TZ = TR + TY; T38 = T36 + T37; T6q = T36 - T37; T2u = T2q - T2t; T75 = T2q + T2t; T10 = TK + TZ; T3P = TK - TZ; T4v = T38 - T35; T39 = T35 + T38; } T8r = T6q - T6p; T6r = T6p + T6q; T3a = T32 + T39; T3L = T39 - T32; T8s = T6m - T6n; T6o = T6m + T6n; T4t = T4r - T4s; T4E = T4r + T4s; T8F = FNMS(KP414213562, T8r, T8s); T8t = FMA(KP414213562, T8s, T8r); T6k = T6i + T6j; T8u = T6j - T6i; T8v = T6f - T6g; T6h = T6f + T6g; } { E T6Z, T1Y, T4O, T26, T6W, T1X, T2o, T4N, T27; T4F = T4v - T4u; T4w = T4u + T4v; T8E = FMA(KP414213562, T8u, T8v); T8w = FNMS(KP414213562, T8v, T8u); T6Z = T1R + T1W; T1X = T1R - T1W; T6E = FMA(KP414213562, T6h, T6k); T6l = FNMS(KP414213562, T6k, T6h); T6F = FNMS(KP414213562, T6o, T6r); T6s = FMA(KP414213562, T6r, T6o); T1Y = FMA(KP707106781, T1X, T1M); T4O = FNMS(KP707106781, T1X, T1M); T26 = T24 + T25; T6W = T25 - T24; T76 = T2i + T2n; T2o = T2i - T2n; T4N = FNMS(KP707106781, T26, T23); T27 = FMA(KP707106781, T26, T23); { E T8O, T6X, T8N, T70; T8O = FMA(KP707106781, T6W, T6V); T6X = FNMS(KP707106781, T6W, T6V); T8N = FMA(KP707106781, T6Z, T6Y); T70 = FNMS(KP707106781, T6Z, T6Y); T4P = FMA(KP668178637, T4O, T4N); T51 = FNMS(KP668178637, T4N, T4O); T2R = FNMS(KP198912367, T1Y, T27); T28 = FMA(KP198912367, T27, T1Y); T8P = FMA(KP198912367, T8O, T8N); T90 = FNMS(KP198912367, T8N, T8O); T7k = FNMS(KP668178637, T6X, T70); T71 = FMA(KP668178637, T70, T6X); T2p = FMA(KP707106781, T2o, T2d); T4R = FNMS(KP707106781, T2o, T2d); } T2x = T2v + T2w; T73 = T2v - T2w; } } } { E T8S, T91, T7l, T78, T5U, T5X, T5y, T61, T5V, T5K, T5S, T60, T5W, T5M, T5I; { E T4S, T50, T4e, T4h, T3S, T4l, T4f, T44, T4c, T4k, T4g, T46, T42; { E T3Q, T3U, T40, T3Z, T3V, T3A, T3D, T3H, T3B, T3y, T3G, T3C; { E T11, T3t, T3w, T3q, T3x, T3v, T3F, T12, T2B, T2U, T3z, T2C; { E T3u, T2S, T2z, T3p, T4Q, T2y; T3u = Tv - T10; T11 = Tv + T10; T4Q = FNMS(KP707106781, T2x, T2u); T2y = FMA(KP707106781, T2x, T2u); { E T8R, T74, T8Q, T77; T8R = FMA(KP707106781, T73, T72); T74 = FNMS(KP707106781, T73, T72); T8Q = FMA(KP707106781, T76, T75); T77 = FNMS(KP707106781, T76, T75); T4S = FNMS(KP668178637, T4R, T4Q); T50 = FMA(KP668178637, T4Q, T4R); T2S = FMA(KP198912367, T2p, T2y); T2z = FNMS(KP198912367, T2y, T2p); T8S = FMA(KP198912367, T8R, T8Q); T91 = FNMS(KP198912367, T8Q, T8R); T7l = FNMS(KP668178637, T74, T77); T78 = FMA(KP668178637, T77, T74); T3Q = T3o - T3h; T3p = T3h + T3o; } T3t = W[30]; T3w = W[31]; T3q = T3a + T3p; T3x = T3p - T3a; T3v = T3t * T3u; T3F = T3w * T3u; { E T1H, T2A, T2Q, T2T; T3U = FNMS(KP923879532, T1G, T1j); T1H = FMA(KP923879532, T1G, T1j); T2A = T28 + T2z; T40 = T2z - T28; T3Z = FNMS(KP923879532, T2P, T2M); T2Q = FMA(KP923879532, T2P, T2M); T2T = T2R + T2S; T3V = T2R - T2S; T12 = W[0]; T3A = FNMS(KP980785280, T2A, T1H); T2B = FMA(KP980785280, T2A, T1H); T3D = FNMS(KP980785280, T2T, T2Q); T2U = FMA(KP980785280, T2T, T2Q); T3z = W[32]; T2C = T12 * T2B; } } { E T2V, T3s, T2D, T3r; T2D = W[1]; T3r = T12 * T2U; T3H = T3z * T3D; T3B = T3z * T3A; T2V = FMA(T2D, T2U, T2C); T3s = FNMS(T2D, T2B, T3r); T3y = FNMS(T3w, T3x, T3v); T3G = FMA(T3t, T3x, T3F); Rm[0] = T11 + T2V; Rp[0] = T11 - T2V; Im[0] = T3s - T3q; Ip[0] = T3q + T3s; T3C = W[33]; } } { E T4b, T3R, T47, T4a, T3J, T49, T4j, T3O, T3N, T43, T3W, T3T, T41, T4d, T3X; E T45, T3Y; { E T3M, T48, T3I, T3E; T3M = T3K + T3L; T48 = T3K - T3L; T3I = FNMS(T3C, T3A, T3H); T3E = FMA(T3C, T3D, T3B); T4b = T3Q - T3P; T3R = T3P + T3Q; Im[WS(rs, 8)] = T3I - T3G; Ip[WS(rs, 8)] = T3G + T3I; Rm[WS(rs, 8)] = T3y + T3E; Rp[WS(rs, 8)] = T3y - T3E; T47 = W[46]; T4a = W[47]; T3J = W[14]; T49 = T47 * T48; T4j = T4a * T48; T3O = W[15]; T3N = T3J * T3M; T43 = T3O * T3M; T3W = FMA(KP980785280, T3V, T3U); T4e = FNMS(KP980785280, T3V, T3U); T3T = W[16]; T4h = FNMS(KP980785280, T40, T3Z); T41 = FMA(KP980785280, T40, T3Z); T4d = W[48]; T3X = T3T * T3W; } T3S = FNMS(T3O, T3R, T3N); T45 = T3T * T41; T4l = T4d * T4h; T4f = T4d * T4e; T44 = FMA(T3J, T3R, T43); T3Y = W[17]; T4c = FNMS(T4a, T4b, T49); T4k = FMA(T47, T4b, T4j); T4g = W[49]; T46 = FNMS(T3Y, T3W, T45); T42 = FMA(T3Y, T41, T3X); } } { E T5v, T5r, T5w, T5A, T5G, T5F, T5B, T5g, T5j, T4I, T5n, T5h, T56, T5e, T5m; E T5i, T58, T54; { E T4n, T4A, T5d, T4H, T59, T5c, T55, T4z, T5b, T5l, T4J, T4U, T53, T5f, T4V; E T57, T4W; { E T4D, T4G, T4m, T4i, T5a, T4y, T4x; T5v = T4C - T4B; T4D = T4B + T4C; T4m = FNMS(T4g, T4e, T4l); T4i = FMA(T4g, T4h, T4f); Im[WS(rs, 4)] = T46 - T44; Ip[WS(rs, 4)] = T44 + T46; Rm[WS(rs, 4)] = T3S + T42; Rp[WS(rs, 4)] = T3S - T42; Im[WS(rs, 12)] = T4m - T4k; Ip[WS(rs, 12)] = T4k + T4m; Rm[WS(rs, 12)] = T4c + T4i; Rp[WS(rs, 12)] = T4c - T4i; T4G = T4E + T4F; T5r = T4F - T4E; T5w = T4t - T4w; T4x = T4t + T4w; T4n = W[6]; T4A = W[7]; T5d = FNMS(KP707106781, T4G, T4D); T4H = FMA(KP707106781, T4G, T4D); T5a = FNMS(KP707106781, T4x, T4q); T4y = FMA(KP707106781, T4x, T4q); T59 = W[38]; T5c = W[39]; { E T4M, T4T, T4Z, T52; T4M = FMA(KP923879532, T4L, T4K); T5A = FNMS(KP923879532, T4L, T4K); T55 = T4A * T4y; T4z = T4n * T4y; T5b = T59 * T5a; T5l = T5c * T5a; T5G = T4P + T4S; T4T = T4P - T4S; T4Z = FMA(KP923879532, T4Y, T4X); T5F = FNMS(KP923879532, T4Y, T4X); T5B = T51 + T50; T52 = T50 - T51; T4J = W[8]; T4U = FMA(KP831469612, T4T, T4M); T5g = FNMS(KP831469612, T4T, T4M); T53 = FMA(KP831469612, T52, T4Z); T5j = FNMS(KP831469612, T52, T4Z); T5f = W[40]; T4V = T4J * T4U; } } T4I = FNMS(T4A, T4H, T4z); T57 = T4J * T53; T5n = T5f * T5j; T5h = T5f * T5g; T56 = FMA(T4n, T4H, T55); T4W = W[9]; T5e = FNMS(T5c, T5d, T5b); T5m = FMA(T59, T5d, T5l); T5i = W[41]; T58 = FNMS(T4W, T4U, T57); T54 = FMA(T4W, T53, T4V); } { E T5p, T5u, T5x, T5R, T5N, T5Q, T5J, T5t, T5P, T5Z, T5z, T5C, T5H, T5T, T5D; E T5L, T5E; { E T5o, T5k, T5s, T5O; T5o = FNMS(T5i, T5g, T5n); T5k = FMA(T5i, T5j, T5h); Im[WS(rs, 2)] = T58 - T56; Ip[WS(rs, 2)] = T56 + T58; Rm[WS(rs, 2)] = T4I + T54; Rp[WS(rs, 2)] = T4I - T54; Im[WS(rs, 10)] = T5o - T5m; Ip[WS(rs, 10)] = T5m + T5o; Rm[WS(rs, 10)] = T5e + T5k; Rp[WS(rs, 10)] = T5e - T5k; T5p = W[22]; T5u = W[23]; T5x = FMA(KP707106781, T5w, T5v); T5R = FNMS(KP707106781, T5w, T5v); T5s = FMA(KP707106781, T5r, T5q); T5O = FNMS(KP707106781, T5r, T5q); T5N = W[54]; T5Q = W[55]; T5J = T5u * T5s; T5t = T5p * T5s; T5P = T5N * T5O; T5Z = T5Q * T5O; T5z = W[24]; T5U = FMA(KP831469612, T5B, T5A); T5C = FNMS(KP831469612, T5B, T5A); T5X = FMA(KP831469612, T5G, T5F); T5H = FNMS(KP831469612, T5G, T5F); T5T = W[56]; T5D = T5z * T5C; } T5y = FNMS(T5u, T5x, T5t); T5L = T5z * T5H; T61 = T5T * T5X; T5V = T5T * T5U; T5K = FMA(T5p, T5x, T5J); T5E = W[25]; T5S = FNMS(T5Q, T5R, T5P); T60 = FMA(T5N, T5R, T5Z); T5W = W[57]; T5M = FNMS(T5E, T5C, T5L); T5I = FMA(T5E, T5H, T5D); } } } { E T7P, T7L, T7K, T7Q, T7U, T80, T7Z, T7V, T9v, T9r, T9q, T9w, T9A, T9G, T9F; E T9B, T9g, T9j, T8I, T9n, T9h, T96, T9e, T9m, T9i, T98, T94; { E T7A, T7D, T6I, T7H, T7B, T7q, T7y, T7G, T7C, T7s, T7o; { E T63, T7x, T6H, T6w, T7t, T7w, T6v, T7p, T7v, T7F, T6J, T7a, T7n, T7z, T7b; E T7r, T7c; { E T6D, T6G, T62, T5Y; T7P = FNMS(KP707106781, T6C, T6z); T6D = FMA(KP707106781, T6C, T6z); T62 = FNMS(T5W, T5U, T61); T5Y = FMA(T5W, T5X, T5V); Im[WS(rs, 6)] = T5M - T5K; Ip[WS(rs, 6)] = T5K + T5M; Rm[WS(rs, 6)] = T5y + T5I; Rp[WS(rs, 6)] = T5y - T5I; Im[WS(rs, 14)] = T62 - T60; Ip[WS(rs, 14)] = T60 + T62; Rm[WS(rs, 14)] = T5S + T5Y; Rp[WS(rs, 14)] = T5S - T5Y; T6G = T6E + T6F; T7L = T6F - T6E; { E T6e, T6t, T7u, T6u; T7K = FNMS(KP707106781, T6d, T66); T6e = FMA(KP707106781, T6d, T66); T6t = T6l + T6s; T7Q = T6l - T6s; T63 = W[2]; T7x = FNMS(KP923879532, T6G, T6D); T6H = FMA(KP923879532, T6G, T6D); T7u = FNMS(KP923879532, T6t, T6e); T6u = FMA(KP923879532, T6t, T6e); T6w = W[3]; T7t = W[34]; T7w = W[35]; T6v = T63 * T6u; T7p = T6w * T6u; T7v = T7t * T7u; T7F = T7w * T7u; } { E T6U, T79, T7j, T7m; T7U = FNMS(KP923879532, T6T, T6M); T6U = FMA(KP923879532, T6T, T6M); T79 = T71 - T78; T80 = T71 + T78; T7Z = FMA(KP923879532, T7i, T7f); T7j = FNMS(KP923879532, T7i, T7f); T7m = T7k + T7l; T7V = T7k - T7l; T6J = W[4]; T7A = FNMS(KP831469612, T79, T6U); T7a = FMA(KP831469612, T79, T6U); T7D = FNMS(KP831469612, T7m, T7j); T7n = FMA(KP831469612, T7m, T7j); T7z = W[36]; T7b = T6J * T7a; } } T6I = FNMS(T6w, T6H, T6v); T7r = T6J * T7n; T7H = T7z * T7D; T7B = T7z * T7A; T7q = FMA(T63, T6H, T7p); T7c = W[5]; T7y = FNMS(T7w, T7x, T7v); T7G = FMA(T7t, T7x, T7F); T7C = W[37]; T7s = FNMS(T7c, T7a, T7r); T7o = FMA(T7c, T7n, T7b); } { E T8n, T9d, T8H, T8A, T99, T9c, T8z, T95, T9b, T9l, T8J, T8U, T93, T9f, T8V; E T97, T8W; { E T8D, T8G, T7I, T7E; T9v = FNMS(KP707106781, T8C, T8B); T8D = FMA(KP707106781, T8C, T8B); T7I = FNMS(T7C, T7A, T7H); T7E = FMA(T7C, T7D, T7B); Im[WS(rs, 1)] = T7s - T7q; Ip[WS(rs, 1)] = T7q + T7s; Rm[WS(rs, 1)] = T6I + T7o; Rp[WS(rs, 1)] = T6I - T7o; Im[WS(rs, 9)] = T7I - T7G; Ip[WS(rs, 9)] = T7G + T7I; Rm[WS(rs, 9)] = T7y + T7E; Rp[WS(rs, 9)] = T7y - T7E; T8G = T8E - T8F; T9r = T8E + T8F; { E T8q, T8x, T9a, T8y; T9q = FNMS(KP707106781, T8p, T8o); T8q = FMA(KP707106781, T8p, T8o); T8x = T8t - T8w; T9w = T8w + T8t; T8n = W[10]; T9d = FNMS(KP923879532, T8G, T8D); T8H = FMA(KP923879532, T8G, T8D); T9a = FNMS(KP923879532, T8x, T8q); T8y = FMA(KP923879532, T8x, T8q); T8A = W[11]; T99 = W[42]; T9c = W[43]; T8z = T8n * T8y; T95 = T8A * T8y; T9b = T99 * T9a; T9l = T9c * T9a; } { E T8M, T8T, T8Z, T92; T9A = FNMS(KP923879532, T8L, T8K); T8M = FMA(KP923879532, T8L, T8K); T8T = T8P - T8S; T9G = T8P + T8S; T9F = FMA(KP923879532, T8Y, T8X); T8Z = FNMS(KP923879532, T8Y, T8X); T92 = T90 + T91; T9B = T91 - T90; T8J = W[12]; T9g = FNMS(KP980785280, T8T, T8M); T8U = FMA(KP980785280, T8T, T8M); T9j = FMA(KP980785280, T92, T8Z); T93 = FNMS(KP980785280, T92, T8Z); T9f = W[44]; T8V = T8J * T8U; } } T8I = FNMS(T8A, T8H, T8z); T97 = T8J * T93; T9n = T9f * T9j; T9h = T9f * T9g; T96 = FMA(T8n, T8H, T95); T8W = W[13]; T9e = FNMS(T9c, T9d, T9b); T9m = FMA(T99, T9d, T9l); T9i = W[45]; T98 = FNMS(T8W, T8U, T97); T94 = FMA(T8W, T93, T8V); } } { E T9U, T9X, T9y, Ta1, T9V, T9K, T9S, Ta0, T9W, T9M, T9I; { E T9p, T9R, T9x, T9u, T9N, T9Q, T9t, T9J, T9P, T9Z, T9z, T9C, T9H, T9T, T9D; E T9L, T9E; { E T9o, T9k, T9O, T9s; T9o = FNMS(T9i, T9g, T9n); T9k = FMA(T9i, T9j, T9h); Im[WS(rs, 3)] = T98 - T96; Ip[WS(rs, 3)] = T96 + T98; Rm[WS(rs, 3)] = T8I + T94; Rp[WS(rs, 3)] = T8I - T94; Im[WS(rs, 11)] = T9o - T9m; Ip[WS(rs, 11)] = T9m + T9o; Rm[WS(rs, 11)] = T9e + T9k; Rp[WS(rs, 11)] = T9e - T9k; T9p = W[26]; T9R = FMA(KP923879532, T9w, T9v); T9x = FNMS(KP923879532, T9w, T9v); T9O = FMA(KP923879532, T9r, T9q); T9s = FNMS(KP923879532, T9r, T9q); T9u = W[27]; T9N = W[58]; T9Q = W[59]; T9t = T9p * T9s; T9J = T9u * T9s; T9P = T9N * T9O; T9Z = T9Q * T9O; T9z = W[28]; T9U = FNMS(KP980785280, T9B, T9A); T9C = FMA(KP980785280, T9B, T9A); T9X = FMA(KP980785280, T9G, T9F); T9H = FNMS(KP980785280, T9G, T9F); T9T = W[60]; T9D = T9z * T9C; } T9y = FNMS(T9u, T9x, T9t); T9L = T9z * T9H; Ta1 = T9T * T9X; T9V = T9T * T9U; T9K = FMA(T9p, T9x, T9J); T9E = W[29]; T9S = FNMS(T9Q, T9R, T9P); Ta0 = FMA(T9N, T9R, T9Z); T9W = W[61]; T9M = FNMS(T9E, T9C, T9L); T9I = FMA(T9E, T9H, T9D); } { E T7J, T8b, T7R, T7O, T87, T8a, T7N, T83, T89, T8j, T7T, T7W, T81, T8d, T7X; E T85, T7Y; { E Ta2, T9Y, T88, T7M; Ta2 = FNMS(T9W, T9U, Ta1); T9Y = FMA(T9W, T9X, T9V); Im[WS(rs, 7)] = T9M - T9K; Ip[WS(rs, 7)] = T9K + T9M; Rm[WS(rs, 7)] = T9y + T9I; Rp[WS(rs, 7)] = T9y - T9I; Im[WS(rs, 15)] = Ta2 - Ta0; Ip[WS(rs, 15)] = Ta0 + Ta2; Rm[WS(rs, 15)] = T9S + T9Y; Rp[WS(rs, 15)] = T9S - T9Y; T7J = W[18]; T8b = FNMS(KP923879532, T7Q, T7P); T7R = FMA(KP923879532, T7Q, T7P); T88 = FNMS(KP923879532, T7L, T7K); T7M = FMA(KP923879532, T7L, T7K); T7O = W[19]; T87 = W[50]; T8a = W[51]; T7N = T7J * T7M; T83 = T7O * T7M; T89 = T87 * T88; T8j = T8a * T88; T7T = W[20]; T8e = FNMS(KP831469612, T7V, T7U); T7W = FMA(KP831469612, T7V, T7U); T8h = FMA(KP831469612, T80, T7Z); T81 = FNMS(KP831469612, T80, T7Z); T8d = W[52]; T7X = T7T * T7W; } T7S = FNMS(T7O, T7R, T7N); T85 = T7T * T81; T8l = T8d * T8h; T8f = T8d * T8e; T84 = FMA(T7J, T7R, T83); T7Y = W[21]; T8c = FNMS(T8a, T8b, T89); T8k = FMA(T87, T8b, T8j); T8g = W[53]; T86 = FNMS(T7Y, T7W, T85); T82 = FMA(T7Y, T81, T7X); } } } } } T8m = FNMS(T8g, T8e, T8l); T8i = FMA(T8g, T8h, T8f); Im[WS(rs, 5)] = T86 - T84; Ip[WS(rs, 5)] = T84 + T86; Rm[WS(rs, 5)] = T7S + T82; Rp[WS(rs, 5)] = T7S - T82; Im[WS(rs, 13)] = T8m - T8k; Ip[WS(rs, 13)] = T8k + T8m; Rm[WS(rs, 13)] = T8c + T8i; Rp[WS(rs, 13)] = T8c - T8i; } } } static const tw_instr twinstr[] = { {TW_FULL, 1, 32}, {TW_NEXT, 1, 0} }; static const hc2c_desc desc = { 32, "hc2cbdft_32", twinstr, &GENUS, {300, 62, 198, 0} }; void X(codelet_hc2cbdft_32) (planner *p) { X(khc2c_register) (p, hc2cbdft_32, &desc, HC2C_VIA_DFT); } #else /* HAVE_FMA */ /* Generated by: ../../../genfft/gen_hc2cdft.native -compact -variables 4 -pipeline-latency 4 -sign 1 -n 32 -dif -name hc2cbdft_32 -include hc2cb.h */ /* * This function contains 498 FP additions, 208 FP multiplications, * (or, 404 additions, 114 multiplications, 94 fused multiply/add), * 102 stack variables, 7 constants, and 128 memory accesses */ #include "hc2cb.h" static void hc2cbdft_32(R *Rp, R *Ip, R *Rm, R *Im, const R *W, stride rs, INT mb, INT me, INT ms) { DK(KP831469612, +0.831469612302545237078788377617905756738560812); DK(KP555570233, +0.555570233019602224742830813948532874374937191); DK(KP195090322, +0.195090322016128267848284868477022240927691618); DK(KP980785280, +0.980785280403230449126182236134239036973933731); DK(KP923879532, +0.923879532511286756128183189396788286822416626); DK(KP382683432, +0.382683432365089771728459984030398866761344562); DK(KP707106781, +0.707106781186547524400844362104849039284835938); { INT m; for (m = mb, W = W + ((mb - 1) * 62); m < me; m = m + 1, Rp = Rp + ms, Ip = Ip + ms, Rm = Rm - ms, Im = Im - ms, W = W + 62, MAKE_VOLATILE_STRIDE(128, rs)) { E Tf, T4a, T6h, T7Z, T6P, T8e, T1j, T4v, T2R, T4L, T5C, T7E, T6a, T7U, T3n; E T4q, TZ, T38, T2p, T4B, T7M, T7R, T2y, T4C, T5Y, T63, T6C, T86, T4i, T4n; E T6z, T85, TK, T31, T1Y, T4y, T7J, T7Q, T27, T4z, T5R, T62, T6v, T83, T4f; E T4m, T6s, T82, Tu, T4p, T6o, T8f, T6M, T80, T1G, T4K, T2I, T4w, T5J, T7T; E T67, T7F, T3g, T4b; { E T3, T2M, T16, T3k, T6, T13, T2P, T3l, Td, T3i, T1h, T2K, Ta, T3h, T1c; E T2J; { E T1, T2, T2N, T2O; T1 = Rp[0]; T2 = Rm[WS(rs, 15)]; T3 = T1 + T2; T2M = T1 - T2; { E T14, T15, T4, T5; T14 = Ip[0]; T15 = Im[WS(rs, 15)]; T16 = T14 + T15; T3k = T14 - T15; T4 = Rp[WS(rs, 8)]; T5 = Rm[WS(rs, 7)]; T6 = T4 + T5; T13 = T4 - T5; } T2N = Ip[WS(rs, 8)]; T2O = Im[WS(rs, 7)]; T2P = T2N + T2O; T3l = T2N - T2O; { E Tb, Tc, T1d, T1e, T1f, T1g; Tb = Rm[WS(rs, 3)]; Tc = Rp[WS(rs, 12)]; T1d = Tb - Tc; T1e = Im[WS(rs, 3)]; T1f = Ip[WS(rs, 12)]; T1g = T1e + T1f; Td = Tb + Tc; T3i = T1f - T1e; T1h = T1d + T1g; T2K = T1d - T1g; } { E T8, T9, T18, T19, T1a, T1b; T8 = Rp[WS(rs, 4)]; T9 = Rm[WS(rs, 11)]; T18 = T8 - T9; T19 = Ip[WS(rs, 4)]; T1a = Im[WS(rs, 11)]; T1b = T19 + T1a; Ta = T8 + T9; T3h = T19 - T1a; T1c = T18 + T1b; T2J = T18 - T1b; } } { E T7, Te, T6f, T6g; T7 = T3 + T6; Te = Ta + Td; Tf = T7 + Te; T4a = T7 - Te; T6f = T16 - T13; T6g = KP707106781 * (T2J - T2K); T6h = T6f + T6g; T7Z = T6f - T6g; } { E T6N, T6O, T17, T1i; T6N = T2M + T2P; T6O = KP707106781 * (T1c + T1h); T6P = T6N - T6O; T8e = T6O + T6N; T17 = T13 + T16; T1i = KP707106781 * (T1c - T1h); T1j = T17 + T1i; T4v = T17 - T1i; } { E T2L, T2Q, T5A, T5B; T2L = KP707106781 * (T2J + T2K); T2Q = T2M - T2P; T2R = T2L + T2Q; T4L = T2Q - T2L; T5A = T3 - T6; T5B = T3i - T3h; T5C = T5A + T5B; T7E = T5A - T5B; } { E T68, T69, T3j, T3m; T68 = Ta - Td; T69 = T3k - T3l; T6a = T68 + T69; T7U = T69 - T68; T3j = T3h + T3i; T3m = T3k + T3l; T3n = T3j + T3m; T4q = T3m - T3j; } } { E TR, T5S, T29, T2t, T2c, T5W, T2w, T37, TY, T5T, T5V, T2i, T2n, T2r, T34; E T2q, T6A, T6B; { E TL, TM, TN, TO, TP, TQ; TL = Rm[0]; TM = Rp[WS(rs, 15)]; TN = TL + TM; TO = Rp[WS(rs, 7)]; TP = Rm[WS(rs, 8)]; TQ = TO + TP; TR = TN + TQ; T5S = TN - TQ; T29 = TO - TP; T2t = TL - TM; } { E T2a, T2b, T35, T2u, T2v, T36; T2a = Im[0]; T2b = Ip[WS(rs, 15)]; T35 = T2b - T2a; T2u = Ip[WS(rs, 7)]; T2v = Im[WS(rs, 8)]; T36 = T2u - T2v; T2c = T2a + T2b; T5W = T35 - T36; T2w = T2u + T2v; T37 = T35 + T36; } { E TU, T2e, T2h, T32, TX, T2j, T2m, T33; { E TS, TT, T2f, T2g; TS = Rp[WS(rs, 3)]; TT = Rm[WS(rs, 12)]; TU = TS + TT; T2e = TS - TT; T2f = Ip[WS(rs, 3)]; T2g = Im[WS(rs, 12)]; T2h = T2f + T2g; T32 = T2f - T2g; } { E TV, TW, T2k, T2l; TV = Rm[WS(rs, 4)]; TW = Rp[WS(rs, 11)]; TX = TV + TW; T2j = TV - TW; T2k = Im[WS(rs, 4)]; T2l = Ip[WS(rs, 11)]; T2m = T2k + T2l; T33 = T2l - T2k; } TY = TU + TX; T5T = T33 - T32; T5V = TU - TX; T2i = T2e + T2h; T2n = T2j + T2m; T2r = T2j - T2m; T34 = T32 + T33; T2q = T2e - T2h; } TZ = TR + TY; T38 = T34 + T37; { E T2d, T2o, T7K, T7L; T2d = T29 - T2c; T2o = KP707106781 * (T2i - T2n); T2p = T2d + T2o; T4B = T2d - T2o; T7K = T5S - T5T; T7L = T5W - T5V; T7M = FMA(KP382683432, T7K, KP923879532 * T7L); T7R = FNMS(KP923879532, T7K, KP382683432 * T7L); } { E T2s, T2x, T5U, T5X; T2s = KP707106781 * (T2q + T2r); T2x = T2t - T2w; T2y = T2s + T2x; T4C = T2x - T2s; T5U = T5S + T5T; T5X = T5V + T5W; T5Y = FMA(KP923879532, T5U, KP382683432 * T5X); T63 = FNMS(KP382683432, T5U, KP923879532 * T5X); } T6A = T2t + T2w; T6B = KP707106781 * (T2i + T2n); T6C = T6A - T6B; T86 = T6B + T6A; { E T4g, T4h, T6x, T6y; T4g = TR - TY; T4h = T37 - T34; T4i = T4g + T4h; T4n = T4h - T4g; T6x = KP707106781 * (T2q - T2r); T6y = T29 + T2c; T6z = T6x - T6y; T85 = T6y + T6x; } } { E TC, T5L, T1I, T22, T1L, T5P, T25, T30, TJ, T5M, T5O, T1R, T1W, T20, T2X; E T1Z, T6t, T6u; { E Tw, Tx, Ty, Tz, TA, TB; Tw = Rp[WS(rs, 1)]; Tx = Rm[WS(rs, 14)]; Ty = Tw + Tx; Tz = Rp[WS(rs, 9)]; TA = Rm[WS(rs, 6)]; TB = Tz + TA; TC = Ty + TB; T5L = Ty - TB; T1I = Tz - TA; T22 = Tw - Tx; } { E T1J, T1K, T2Y, T23, T24, T2Z; T1J = Ip[WS(rs, 1)]; T1K = Im[WS(rs, 14)]; T2Y = T1J - T1K; T23 = Ip[WS(rs, 9)]; T24 = Im[WS(rs, 6)]; T2Z = T23 - T24; T1L = T1J + T1K; T5P = T2Y - T2Z; T25 = T23 + T24; T30 = T2Y + T2Z; } { E TF, T1N, T1Q, T2V, TI, T1S, T1V, T2W; { E TD, TE, T1O, T1P; TD = Rp[WS(rs, 5)]; TE = Rm[WS(rs, 10)]; TF = TD + TE; T1N = TD - TE; T1O = Ip[WS(rs, 5)]; T1P = Im[WS(rs, 10)]; T1Q = T1O + T1P; T2V = T1O - T1P; } { E TG, TH, T1T, T1U; TG = Rm[WS(rs, 2)]; TH = Rp[WS(rs, 13)]; TI = TG + TH; T1S = TG - TH; T1T = Im[WS(rs, 2)]; T1U = Ip[WS(rs, 13)]; T1V = T1T + T1U; T2W = T1U - T1T; } TJ = TF + TI; T5M = T2W - T2V; T5O = TF - TI; T1R = T1N + T1Q; T1W = T1S + T1V; T20 = T1S - T1V; T2X = T2V + T2W; T1Z = T1N - T1Q; } TK = TC + TJ; T31 = T2X + T30; { E T1M, T1X, T7H, T7I; T1M = T1I + T1L; T1X = KP707106781 * (T1R - T1W); T1Y = T1M + T1X; T4y = T1M - T1X; T7H = T5L - T5M; T7I = T5P - T5O; T7J = FNMS(KP923879532, T7I, KP382683432 * T7H); T7Q = FMA(KP923879532, T7H, KP382683432 * T7I); } { E T21, T26, T5N, T5Q; T21 = KP707106781 * (T1Z + T20); T26 = T22 - T25; T27 = T21 + T26; T4z = T26 - T21; T5N = T5L + T5M; T5Q = T5O + T5P; T5R = FNMS(KP382683432, T5Q, KP923879532 * T5N); T62 = FMA(KP382683432, T5N, KP923879532 * T5Q); } T6t = T22 + T25; T6u = KP707106781 * (T1R + T1W); T6v = T6t - T6u; T83 = T6u + T6t; { E T4d, T4e, T6q, T6r; T4d = TC - TJ; T4e = T30 - T2X; T4f = T4d - T4e; T4m = T4d + T4e; T6q = T1L - T1I; T6r = KP707106781 * (T1Z - T20); T6s = T6q + T6r; T82 = T6q - T6r; } } { E Ti, T3a, Tl, T3b, T1o, T1t, T6j, T6i, T5E, T5D, Tp, T3d, Ts, T3e, T1z; E T1E, T6m, T6l, T5H, T5G; { E T1p, T1n, T1k, T1s; { E Tg, Th, T1l, T1m; Tg = Rp[WS(rs, 2)]; Th = Rm[WS(rs, 13)]; Ti = Tg + Th; T1p = Tg - Th; T1l = Ip[WS(rs, 2)]; T1m = Im[WS(rs, 13)]; T1n = T1l + T1m; T3a = T1l - T1m; } { E Tj, Tk, T1q, T1r; Tj = Rp[WS(rs, 10)]; Tk = Rm[WS(rs, 5)]; Tl = Tj + Tk; T1k = Tj - Tk; T1q = Ip[WS(rs, 10)]; T1r = Im[WS(rs, 5)]; T1s = T1q + T1r; T3b = T1q - T1r; } T1o = T1k + T1n; T1t = T1p - T1s; T6j = T1p + T1s; T6i = T1n - T1k; T5E = T3a - T3b; T5D = Ti - Tl; } { E T1A, T1y, T1v, T1D; { E Tn, To, T1w, T1x; Tn = Rm[WS(rs, 1)]; To = Rp[WS(rs, 14)]; Tp = Tn + To; T1A = Tn - To; T1w = Im[WS(rs, 1)]; T1x = Ip[WS(rs, 14)]; T1y = T1w + T1x; T3d = T1x - T1w; } { E Tq, Tr, T1B, T1C; Tq = Rp[WS(rs, 6)]; Tr = Rm[WS(rs, 9)]; Ts = Tq + Tr; T1v = Tq - Tr; T1B = Ip[WS(rs, 6)]; T1C = Im[WS(rs, 9)]; T1D = T1B + T1C; T3e = T1B - T1C; } T1z = T1v - T1y; T1E = T1A - T1D; T6m = T1A + T1D; T6l = T1v + T1y; T5H = T3d - T3e; T5G = Tp - Ts; } { E Tm, Tt, T6k, T6n; Tm = Ti + Tl; Tt = Tp + Ts; Tu = Tm + Tt; T4p = Tm - Tt; T6k = FMA(KP382683432, T6i, KP923879532 * T6j); T6n = FMA(KP382683432, T6l, KP923879532 * T6m); T6o = T6k - T6n; T8f = T6k + T6n; } { E T6K, T6L, T1u, T1F; T6K = FNMS(KP923879532, T6i, KP382683432 * T6j); T6L = FNMS(KP923879532, T6l, KP382683432 * T6m); T6M = T6K + T6L; T80 = T6K - T6L; T1u = FMA(KP923879532, T1o, KP382683432 * T1t); T1F = FNMS(KP382683432, T1E, KP923879532 * T1z); T1G = T1u + T1F; T4K = T1F - T1u; } { E T2G, T2H, T5F, T5I; T2G = FNMS(KP382683432, T1o, KP923879532 * T1t); T2H = FMA(KP382683432, T1z, KP923879532 * T1E); T2I = T2G + T2H; T4w = T2G - T2H; T5F = T5D - T5E; T5I = T5G + T5H; T5J = KP707106781 * (T5F + T5I); T7T = KP707106781 * (T5F - T5I); } { E T65, T66, T3c, T3f; T65 = T5D + T5E; T66 = T5H - T5G; T67 = KP707106781 * (T65 + T66); T7F = KP707106781 * (T66 - T65); T3c = T3a + T3b; T3f = T3d + T3e; T3g = T3c + T3f; T4b = T3f - T3c; } } { E T11, T3s, T3p, T3u, T3K, T40, T3G, T3Y, T2T, T43, T3z, T3P, T2B, T45, T3x; E T3T; { E Tv, T10, T3E, T3F; Tv = Tf + Tu; T10 = TK + TZ; T11 = Tv + T10; T3s = Tv - T10; { E T39, T3o, T3I, T3J; T39 = T31 + T38; T3o = T3g + T3n; T3p = T39 + T3o; T3u = T3o - T39; T3I = TK - TZ; T3J = T3n - T3g; T3K = T3I + T3J; T40 = T3J - T3I; } T3E = Tf - Tu; T3F = T38 - T31; T3G = T3E + T3F; T3Y = T3E - T3F; { E T2S, T3N, T2F, T3O, T2D, T2E; T2S = T2I + T2R; T3N = T1j - T1G; T2D = FNMS(KP195090322, T1Y, KP980785280 * T27); T2E = FMA(KP195090322, T2p, KP980785280 * T2y); T2F = T2D + T2E; T3O = T2D - T2E; T2T = T2F + T2S; T43 = T3N - T3O; T3z = T2S - T2F; T3P = T3N + T3O; } { E T1H, T3S, T2A, T3R, T28, T2z; T1H = T1j + T1G; T3S = T2R - T2I; T28 = FMA(KP980785280, T1Y, KP195090322 * T27); T2z = FNMS(KP195090322, T2y, KP980785280 * T2p); T2A = T28 + T2z; T3R = T2z - T28; T2B = T1H + T2A; T45 = T3S - T3R; T3x = T1H - T2A; T3T = T3R + T3S; } } { E T2U, T3q, T12, T2C; T12 = W[0]; T2C = W[1]; T2U = FMA(T12, T2B, T2C * T2T); T3q = FNMS(T2C, T2B, T12 * T2T); Rp[0] = T11 - T2U; Ip[0] = T3p + T3q; Rm[0] = T11 + T2U; Im[0] = T3q - T3p; } { E T41, T47, T46, T48; { E T3X, T3Z, T42, T44; T3X = W[46]; T3Z = W[47]; T41 = FNMS(T3Z, T40, T3X * T3Y); T47 = FMA(T3Z, T3Y, T3X * T40); T42 = W[48]; T44 = W[49]; T46 = FMA(T42, T43, T44 * T45); T48 = FNMS(T44, T43, T42 * T45); } Rp[WS(rs, 12)] = T41 - T46; Ip[WS(rs, 12)] = T47 + T48; Rm[WS(rs, 12)] = T41 + T46; Im[WS(rs, 12)] = T48 - T47; } { E T3v, T3B, T3A, T3C; { E T3r, T3t, T3w, T3y; T3r = W[30]; T3t = W[31]; T3v = FNMS(T3t, T3u, T3r * T3s); T3B = FMA(T3t, T3s, T3r * T3u); T3w = W[32]; T3y = W[33]; T3A = FMA(T3w, T3x, T3y * T3z); T3C = FNMS(T3y, T3x, T3w * T3z); } Rp[WS(rs, 8)] = T3v - T3A; Ip[WS(rs, 8)] = T3B + T3C; Rm[WS(rs, 8)] = T3v + T3A; Im[WS(rs, 8)] = T3C - T3B; } { E T3L, T3V, T3U, T3W; { E T3D, T3H, T3M, T3Q; T3D = W[14]; T3H = W[15]; T3L = FNMS(T3H, T3K, T3D * T3G); T3V = FMA(T3H, T3G, T3D * T3K); T3M = W[16]; T3Q = W[17]; T3U = FMA(T3M, T3P, T3Q * T3T); T3W = FNMS(T3Q, T3P, T3M * T3T); } Rp[WS(rs, 4)] = T3L - T3U; Ip[WS(rs, 4)] = T3V + T3W; Rm[WS(rs, 4)] = T3L + T3U; Im[WS(rs, 4)] = T3W - T3V; } } { E T7O, T8m, T7W, T8o, T8E, T8U, T8A, T8S, T8h, T8X, T8t, T8J, T89, T8Z, T8r; E T8N; { E T7G, T7N, T8y, T8z; T7G = T7E + T7F; T7N = T7J + T7M; T7O = T7G + T7N; T8m = T7G - T7N; { E T7S, T7V, T8C, T8D; T7S = T7Q + T7R; T7V = T7T + T7U; T7W = T7S + T7V; T8o = T7V - T7S; T8C = T7J - T7M; T8D = T7U - T7T; T8E = T8C + T8D; T8U = T8D - T8C; } T8y = T7E - T7F; T8z = T7R - T7Q; T8A = T8y + T8z; T8S = T8y - T8z; { E T8g, T8H, T8d, T8I, T8b, T8c; T8g = T8e - T8f; T8H = T7Z - T80; T8b = FNMS(KP980785280, T82, KP195090322 * T83); T8c = FNMS(KP980785280, T85, KP195090322 * T86); T8d = T8b + T8c; T8I = T8b - T8c; T8h = T8d + T8g; T8X = T8H - T8I; T8t = T8g - T8d; T8J = T8H + T8I; } { E T81, T8L, T88, T8M, T84, T87; T81 = T7Z + T80; T8L = T8f + T8e; T84 = FMA(KP195090322, T82, KP980785280 * T83); T87 = FMA(KP195090322, T85, KP980785280 * T86); T88 = T84 - T87; T8M = T84 + T87; T89 = T81 + T88; T8Z = T8M + T8L; T8r = T81 - T88; T8N = T8L - T8M; } } { E T7X, T8j, T8i, T8k; { E T7D, T7P, T7Y, T8a; T7D = W[10]; T7P = W[11]; T7X = FNMS(T7P, T7W, T7D * T7O); T8j = FMA(T7P, T7O, T7D * T7W); T7Y = W[12]; T8a = W[13]; T8i = FMA(T7Y, T89, T8a * T8h); T8k = FNMS(T8a, T89, T7Y * T8h); } Rp[WS(rs, 3)] = T7X - T8i; Ip[WS(rs, 3)] = T8j + T8k; Rm[WS(rs, 3)] = T7X + T8i; Im[WS(rs, 3)] = T8k - T8j; } { E T8V, T91, T90, T92; { E T8R, T8T, T8W, T8Y; T8R = W[58]; T8T = W[59]; T8V = FNMS(T8T, T8U, T8R * T8S); T91 = FMA(T8T, T8S, T8R * T8U); T8W = W[60]; T8Y = W[61]; T90 = FMA(T8W, T8X, T8Y * T8Z); T92 = FNMS(T8Y, T8X, T8W * T8Z); } Rp[WS(rs, 15)] = T8V - T90; Ip[WS(rs, 15)] = T91 + T92; Rm[WS(rs, 15)] = T8V + T90; Im[WS(rs, 15)] = T92 - T91; } { E T8p, T8v, T8u, T8w; { E T8l, T8n, T8q, T8s; T8l = W[42]; T8n = W[43]; T8p = FNMS(T8n, T8o, T8l * T8m); T8v = FMA(T8n, T8m, T8l * T8o); T8q = W[44]; T8s = W[45]; T8u = FMA(T8q, T8r, T8s * T8t); T8w = FNMS(T8s, T8r, T8q * T8t); } Rp[WS(rs, 11)] = T8p - T8u; Ip[WS(rs, 11)] = T8v + T8w; Rm[WS(rs, 11)] = T8p + T8u; Im[WS(rs, 11)] = T8w - T8v; } { E T8F, T8P, T8O, T8Q; { E T8x, T8B, T8G, T8K; T8x = W[26]; T8B = W[27]; T8F = FNMS(T8B, T8E, T8x * T8A); T8P = FMA(T8B, T8A, T8x * T8E); T8G = W[28]; T8K = W[29]; T8O = FMA(T8G, T8J, T8K * T8N); T8Q = FNMS(T8K, T8J, T8G * T8N); } Rp[WS(rs, 7)] = T8F - T8O; Ip[WS(rs, 7)] = T8P + T8Q; Rm[WS(rs, 7)] = T8F + T8O; Im[WS(rs, 7)] = T8Q - T8P; } } { E T4k, T4S, T4s, T4U, T5a, T5q, T56, T5o, T4N, T5t, T4Z, T5f, T4F, T5v, T4X; E T5j; { E T4c, T4j, T54, T55; T4c = T4a + T4b; T4j = KP707106781 * (T4f + T4i); T4k = T4c + T4j; T4S = T4c - T4j; { E T4o, T4r, T58, T59; T4o = KP707106781 * (T4m + T4n); T4r = T4p + T4q; T4s = T4o + T4r; T4U = T4r - T4o; T58 = KP707106781 * (T4f - T4i); T59 = T4q - T4p; T5a = T58 + T59; T5q = T59 - T58; } T54 = T4a - T4b; T55 = KP707106781 * (T4n - T4m); T56 = T54 + T55; T5o = T54 - T55; { E T4M, T5d, T4J, T5e, T4H, T4I; T4M = T4K + T4L; T5d = T4v - T4w; T4H = FNMS(KP831469612, T4y, KP555570233 * T4z); T4I = FMA(KP831469612, T4B, KP555570233 * T4C); T4J = T4H + T4I; T5e = T4H - T4I; T4N = T4J + T4M; T5t = T5d - T5e; T4Z = T4M - T4J; T5f = T5d + T5e; } { E T4x, T5i, T4E, T5h, T4A, T4D; T4x = T4v + T4w; T5i = T4L - T4K; T4A = FMA(KP555570233, T4y, KP831469612 * T4z); T4D = FNMS(KP831469612, T4C, KP555570233 * T4B); T4E = T4A + T4D; T5h = T4D - T4A; T4F = T4x + T4E; T5v = T5i - T5h; T4X = T4x - T4E; T5j = T5h + T5i; } } { E T4t, T4P, T4O, T4Q; { E T49, T4l, T4u, T4G; T49 = W[6]; T4l = W[7]; T4t = FNMS(T4l, T4s, T49 * T4k); T4P = FMA(T4l, T4k, T49 * T4s); T4u = W[8]; T4G = W[9]; T4O = FMA(T4u, T4F, T4G * T4N); T4Q = FNMS(T4G, T4F, T4u * T4N); } Rp[WS(rs, 2)] = T4t - T4O; Ip[WS(rs, 2)] = T4P + T4Q; Rm[WS(rs, 2)] = T4t + T4O; Im[WS(rs, 2)] = T4Q - T4P; } { E T5r, T5x, T5w, T5y; { E T5n, T5p, T5s, T5u; T5n = W[54]; T5p = W[55]; T5r = FNMS(T5p, T5q, T5n * T5o); T5x = FMA(T5p, T5o, T5n * T5q); T5s = W[56]; T5u = W[57]; T5w = FMA(T5s, T5t, T5u * T5v); T5y = FNMS(T5u, T5t, T5s * T5v); } Rp[WS(rs, 14)] = T5r - T5w; Ip[WS(rs, 14)] = T5x + T5y; Rm[WS(rs, 14)] = T5r + T5w; Im[WS(rs, 14)] = T5y - T5x; } { E T4V, T51, T50, T52; { E T4R, T4T, T4W, T4Y; T4R = W[38]; T4T = W[39]; T4V = FNMS(T4T, T4U, T4R * T4S); T51 = FMA(T4T, T4S, T4R * T4U); T4W = W[40]; T4Y = W[41]; T50 = FMA(T4W, T4X, T4Y * T4Z); T52 = FNMS(T4Y, T4X, T4W * T4Z); } Rp[WS(rs, 10)] = T4V - T50; Ip[WS(rs, 10)] = T51 + T52; Rm[WS(rs, 10)] = T4V + T50; Im[WS(rs, 10)] = T52 - T51; } { E T5b, T5l, T5k, T5m; { E T53, T57, T5c, T5g; T53 = W[22]; T57 = W[23]; T5b = FNMS(T57, T5a, T53 * T56); T5l = FMA(T57, T56, T53 * T5a); T5c = W[24]; T5g = W[25]; T5k = FMA(T5c, T5f, T5g * T5j); T5m = FNMS(T5g, T5f, T5c * T5j); } Rp[WS(rs, 6)] = T5b - T5k; Ip[WS(rs, 6)] = T5l + T5m; Rm[WS(rs, 6)] = T5b + T5k; Im[WS(rs, 6)] = T5m - T5l; } } { E T60, T6W, T6c, T6Y, T7e, T7u, T7a, T7s, T6R, T7x, T73, T7j, T6F, T7z, T71; E T7n; { E T5K, T5Z, T78, T79; T5K = T5C + T5J; T5Z = T5R + T5Y; T60 = T5K + T5Z; T6W = T5K - T5Z; { E T64, T6b, T7c, T7d; T64 = T62 + T63; T6b = T67 + T6a; T6c = T64 + T6b; T6Y = T6b - T64; T7c = T5R - T5Y; T7d = T6a - T67; T7e = T7c + T7d; T7u = T7d - T7c; } T78 = T5C - T5J; T79 = T63 - T62; T7a = T78 + T79; T7s = T78 - T79; { E T6Q, T7h, T6J, T7i, T6H, T6I; T6Q = T6M + T6P; T7h = T6h - T6o; T6H = FNMS(KP555570233, T6s, KP831469612 * T6v); T6I = FMA(KP555570233, T6z, KP831469612 * T6C); T6J = T6H + T6I; T7i = T6H - T6I; T6R = T6J + T6Q; T7x = T7h - T7i; T73 = T6Q - T6J; T7j = T7h + T7i; } { E T6p, T7m, T6E, T7l, T6w, T6D; T6p = T6h + T6o; T7m = T6P - T6M; T6w = FMA(KP831469612, T6s, KP555570233 * T6v); T6D = FNMS(KP555570233, T6C, KP831469612 * T6z); T6E = T6w + T6D; T7l = T6D - T6w; T6F = T6p + T6E; T7z = T7m - T7l; T71 = T6p - T6E; T7n = T7l + T7m; } } { E T6d, T6T, T6S, T6U; { E T5z, T61, T6e, T6G; T5z = W[2]; T61 = W[3]; T6d = FNMS(T61, T6c, T5z * T60); T6T = FMA(T61, T60, T5z * T6c); T6e = W[4]; T6G = W[5]; T6S = FMA(T6e, T6F, T6G * T6R); T6U = FNMS(T6G, T6F, T6e * T6R); } Rp[WS(rs, 1)] = T6d - T6S; Ip[WS(rs, 1)] = T6T + T6U; Rm[WS(rs, 1)] = T6d + T6S; Im[WS(rs, 1)] = T6U - T6T; } { E T7v, T7B, T7A, T7C; { E T7r, T7t, T7w, T7y; T7r = W[50]; T7t = W[51]; T7v = FNMS(T7t, T7u, T7r * T7s); T7B = FMA(T7t, T7s, T7r * T7u); T7w = W[52]; T7y = W[53]; T7A = FMA(T7w, T7x, T7y * T7z); T7C = FNMS(T7y, T7x, T7w * T7z); } Rp[WS(rs, 13)] = T7v - T7A; Ip[WS(rs, 13)] = T7B + T7C; Rm[WS(rs, 13)] = T7v + T7A; Im[WS(rs, 13)] = T7C - T7B; } { E T6Z, T75, T74, T76; { E T6V, T6X, T70, T72; T6V = W[34]; T6X = W[35]; T6Z = FNMS(T6X, T6Y, T6V * T6W); T75 = FMA(T6X, T6W, T6V * T6Y); T70 = W[36]; T72 = W[37]; T74 = FMA(T70, T71, T72 * T73); T76 = FNMS(T72, T71, T70 * T73); } Rp[WS(rs, 9)] = T6Z - T74; Ip[WS(rs, 9)] = T75 + T76; Rm[WS(rs, 9)] = T6Z + T74; Im[WS(rs, 9)] = T76 - T75; } { E T7f, T7p, T7o, T7q; { E T77, T7b, T7g, T7k; T77 = W[18]; T7b = W[19]; T7f = FNMS(T7b, T7e, T77 * T7a); T7p = FMA(T7b, T7a, T77 * T7e); T7g = W[20]; T7k = W[21]; T7o = FMA(T7g, T7j, T7k * T7n); T7q = FNMS(T7k, T7j, T7g * T7n); } Rp[WS(rs, 5)] = T7f - T7o; Ip[WS(rs, 5)] = T7p + T7q; Rm[WS(rs, 5)] = T7f + T7o; Im[WS(rs, 5)] = T7q - T7p; } } } } } static const tw_instr twinstr[] = { {TW_FULL, 1, 32}, {TW_NEXT, 1, 0} }; static const hc2c_desc desc = { 32, "hc2cbdft_32", twinstr, &GENUS, {404, 114, 94, 0} }; void X(codelet_hc2cbdft_32) (planner *p) { X(khc2c_register) (p, hc2cbdft_32, &desc, HC2C_VIA_DFT); } #endif /* HAVE_FMA */ fftw-3.3.3/rdft/scalar/r2cb/hc2cbdft_20.c0000644000175000017500000007336112054411040014570 00000000000000/* * Copyright (c) 2003, 2007-11 Matteo Frigo * Copyright (c) 2003, 2007-11 Massachusetts Institute of Technology * * This program is free software; you can redistribute it and/or modify * it under the terms of the GNU General Public License as published by * the Free Software Foundation; either version 2 of the License, or * (at your option) any later version. * * This program is distributed in the hope that it will be useful, * but WITHOUT ANY WARRANTY; without even the implied warranty of * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the * GNU General Public License for more details. * * You should have received a copy of the GNU General Public License * along with this program; if not, write to the Free Software * Foundation, Inc., 51 Franklin Street, Fifth Floor, Boston, MA 02110-1301 USA * */ /* This file was automatically generated --- DO NOT EDIT */ /* Generated on Sun Nov 25 07:42:07 EST 2012 */ #include "codelet-rdft.h" #ifdef HAVE_FMA /* Generated by: ../../../genfft/gen_hc2cdft.native -fma -reorder-insns -schedule-for-pipeline -compact -variables 4 -pipeline-latency 4 -sign 1 -n 20 -dif -name hc2cbdft_20 -include hc2cb.h */ /* * This function contains 286 FP additions, 148 FP multiplications, * (or, 176 additions, 38 multiplications, 110 fused multiply/add), * 122 stack variables, 4 constants, and 80 memory accesses */ #include "hc2cb.h" static void hc2cbdft_20(R *Rp, R *Ip, R *Rm, R *Im, const R *W, stride rs, INT mb, INT me, INT ms) { DK(KP951056516, +0.951056516295153572116439333379382143405698634); DK(KP559016994, +0.559016994374947424102293417182819058860154590); DK(KP250000000, +0.250000000000000000000000000000000000000000000); DK(KP618033988, +0.618033988749894848204586834365638117720309180); { INT m; for (m = mb, W = W + ((mb - 1) * 38); m < me; m = m + 1, Rp = Rp + ms, Ip = Ip + ms, Rm = Rm - ms, Im = Im - ms, W = W + 38, MAKE_VOLATILE_STRIDE(80, rs)) { E T5s, T5v, T5t, T5z, T5q, T5y, T5u, T5A, T5w; { E T3T, T27, T2o, T41, T2p, T40, TU, T15, T2Q, T1N, T2L, T1w, T59, T4n, T5e; E T4A, T2m, T24, T2Z, T2h, T4J, T3P, T3Y, T3W, T2d, TJ, T3H, T2c, TD, T52; E T3G, T1E, T4f, T5I, T4e, T4w, T5L, T4v, T1J, T1H; { E T1A, T3, T25, TI, TF, T6, T26, T1D, TO, T47, T3z, Te, T1S, T3M, T1e; E T4k, TZ, T4a, T3C, Tt, T1Z, T3J, T1p, T4h, T14, T4b, T3D, TA, T22, T3K; E T1u, T4i, Ti, T1f, Th, T1T, TS, Tj, T1g, T1h; { E T4, T5, T1B, T1C; { E T1, T2, TG, TH; T1 = Rp[0]; T2 = Rm[WS(rs, 9)]; TG = Ip[0]; TH = Im[WS(rs, 9)]; T4 = Rp[WS(rs, 5)]; T1A = T1 - T2; T3 = T1 + T2; T25 = TG - TH; TI = TG + TH; T5 = Rm[WS(rs, 4)]; T1B = Ip[WS(rs, 5)]; T1C = Im[WS(rs, 4)]; } { E Tq, T1l, Tp, T1X, TY, Tr, T1m, T1n; { E Tb, T1a, Ta, T1Q, TN, Tc, T1b, T1c; { E T8, T9, TL, TM; T8 = Rp[WS(rs, 4)]; TF = T4 - T5; T6 = T4 + T5; T26 = T1B - T1C; T1D = T1B + T1C; T9 = Rm[WS(rs, 5)]; TL = Ip[WS(rs, 4)]; TM = Im[WS(rs, 5)]; Tb = Rp[WS(rs, 9)]; T1a = T8 - T9; Ta = T8 + T9; T1Q = TL - TM; TN = TL + TM; Tc = Rm[0]; T1b = Ip[WS(rs, 9)]; T1c = Im[0]; } { E Tn, To, TW, TX; Tn = Rp[WS(rs, 8)]; { E TK, Td, T1R, T1d; TK = Tb - Tc; Td = Tb + Tc; T1R = T1b - T1c; T1d = T1b + T1c; TO = TK + TN; T47 = TN - TK; T3z = Ta - Td; Te = Ta + Td; T1S = T1Q + T1R; T3M = T1Q - T1R; T1e = T1a - T1d; T4k = T1a + T1d; To = Rm[WS(rs, 1)]; } TW = Ip[WS(rs, 8)]; TX = Im[WS(rs, 1)]; Tq = Rm[WS(rs, 6)]; T1l = Tn - To; Tp = Tn + To; T1X = TW - TX; TY = TW + TX; Tr = Rp[WS(rs, 3)]; T1m = Im[WS(rs, 6)]; T1n = Ip[WS(rs, 3)]; } } { E Tx, T1q, Tw, T20, T13, Ty, T1r, T1s; { E Tu, Tv, T11, T12; Tu = Rm[WS(rs, 7)]; { E TV, Ts, T1Y, T1o; TV = Tq - Tr; Ts = Tq + Tr; T1Y = T1n - T1m; T1o = T1m + T1n; TZ = TV + TY; T4a = TY - TV; T3C = Tp - Ts; Tt = Tp + Ts; T1Z = T1X + T1Y; T3J = T1X - T1Y; T1p = T1l + T1o; T4h = T1l - T1o; Tv = Rp[WS(rs, 2)]; } T11 = Im[WS(rs, 7)]; T12 = Ip[WS(rs, 2)]; Tx = Rm[WS(rs, 2)]; T1q = Tu - Tv; Tw = Tu + Tv; T20 = T12 - T11; T13 = T11 + T12; Ty = Rp[WS(rs, 7)]; T1r = Im[WS(rs, 2)]; T1s = Ip[WS(rs, 7)]; } { E Tf, Tg, TQ, TR; Tf = Rm[WS(rs, 3)]; { E T10, Tz, T21, T1t; T10 = Tx - Ty; Tz = Tx + Ty; T21 = T1s - T1r; T1t = T1r + T1s; T14 = T10 - T13; T4b = T10 + T13; T3D = Tw - Tz; TA = Tw + Tz; T22 = T20 + T21; T3K = T20 - T21; T1u = T1q + T1t; T4i = T1q - T1t; Tg = Rp[WS(rs, 6)]; } TQ = Im[WS(rs, 3)]; TR = Ip[WS(rs, 6)]; Ti = Rp[WS(rs, 1)]; T1f = Tf - Tg; Th = Tf + Tg; T1T = TR - TQ; TS = TQ + TR; Tj = Rm[WS(rs, 8)]; T1g = Ip[WS(rs, 1)]; T1h = Im[WS(rs, 8)]; } } } } { E T1V, T3N, TB, T3B, Tm, T3E, T1F, T1G, T4t, T4j, T4m, T4s, T4c, T4y, T4z; E T49, T3y, T7; { E TT, T48, T1j, T4l, T3A, Tl; T3T = T25 - T26; T27 = T25 + T26; { E TP, Tk, T1U, T1i; TP = Ti - Tj; Tk = Ti + Tj; T1U = T1g - T1h; T1i = T1g + T1h; TT = TP - TS; T48 = TP + TS; T3A = Th - Tk; Tl = Th + Tk; T1V = T1T + T1U; T3N = T1T - T1U; T1j = T1f - T1i; T4l = T1f + T1i; T2o = Tt - TA; TB = Tt + TA; } T41 = T3z - T3A; T3B = T3z + T3A; Tm = Te + Tl; T2p = Te - Tl; { E T1L, T1M, T1k, T1v; T40 = T3C - T3D; T3E = T3C + T3D; TU = TO + TT; T1L = TO - TT; T1M = TZ - T14; T15 = TZ + T14; T1F = T1e + T1j; T1k = T1e - T1j; T1v = T1p - T1u; T1G = T1p + T1u; T4t = T4h + T4i; T4j = T4h - T4i; T2Q = FNMS(KP618033988, T1L, T1M); T1N = FMA(KP618033988, T1M, T1L); T2L = FNMS(KP618033988, T1k, T1v); T1w = FMA(KP618033988, T1v, T1k); T4m = T4k - T4l; T4s = T4k + T4l; T4c = T4a - T4b; T4y = T4a + T4b; T4z = T47 + T48; T49 = T47 - T48; } } { E T2g, T1W, T23, T2f; T2g = T1S - T1V; T1W = T1S + T1V; T59 = FMA(KP618033988, T4j, T4m); T4n = FNMS(KP618033988, T4m, T4j); T5e = FMA(KP618033988, T4y, T4z); T4A = FNMS(KP618033988, T4z, T4y); T23 = T1Z + T22; T2f = T1Z - T22; { E T3V, T3L, T3O, T3U; T3V = T3J + T3K; T3L = T3J - T3K; T2m = T1W - T23; T24 = T1W + T23; T2Z = FMA(KP618033988, T2f, T2g); T2h = FNMS(KP618033988, T2g, T2f); T3O = T3M - T3N; T3U = T3M + T3N; T3y = T3 - T6; T7 = T3 + T6; T4J = FMA(KP618033988, T3L, T3O); T3P = FNMS(KP618033988, T3O, T3L); T3Y = T3U - T3V; T3W = T3U + T3V; } } { E T46, TC, T3F, T4r, T4d, T4u; TC = Tm + TB; T2d = Tm - TB; TJ = TF + TI; T46 = TI - TF; T3H = T3B - T3E; T3F = T3B + T3E; T2c = FNMS(KP250000000, TC, T7); TD = T7 + TC; T52 = T3y + T3F; T3G = FNMS(KP250000000, T3F, T3y); T4r = T1A + T1D; T1E = T1A - T1D; T4f = T49 - T4c; T4d = T49 + T4c; T5I = T46 + T4d; T4e = FNMS(KP250000000, T4d, T46); T4w = T4s - T4t; T4u = T4s + T4t; T5L = T4u + T4r; T4v = FNMS(KP250000000, T4u, T4r); T1J = T1F - T1G; T1H = T1F + T1G; } } } { E T38, T3b, T39, T3f, T36, T3e, T3a; { E T28, T3r, T3o, T3v, T3p, T2b, T2k, T35, T3l, T2H, T2r, T2j, T2z, T2D, T2G; E T2X, T2F, T2T, T32, T3h, T3k, T31, T3d, T3j, T3t, T1x, T2u, T1O, T2x, T2v; E T1y, T2B, T29, T2J, T2M, T2R, T2N, T2V; { E T2l, T1I, T18, T2q, T34, T17, T16, T3n; T28 = T24 + T27; T2l = FNMS(KP250000000, T24, T27); T3r = T1H + T1E; T1I = FNMS(KP250000000, T1H, T1E); T18 = TU - T15; T16 = TU + T15; T3n = W[8]; T2q = FNMS(KP618033988, T2p, T2o); T34 = FMA(KP618033988, T2o, T2p); T17 = FNMS(KP250000000, T16, TJ); T3o = TJ + T16; T3v = T3n * T3r; T3p = T3n * T3o; { E T2Y, T2E, T3i, T30; { E T2e, T33, T2n, T2i; T2Y = FMA(KP559016994, T2d, T2c); T2e = FNMS(KP559016994, T2d, T2c); T2b = W[14]; T2k = W[15]; T33 = FMA(KP559016994, T2m, T2l); T2n = FNMS(KP559016994, T2m, T2l); T2E = FMA(KP951056516, T2h, T2e); T2i = FNMS(KP951056516, T2h, T2e); T35 = FMA(KP951056516, T34, T33); T3l = FNMS(KP951056516, T34, T33); T2H = FNMS(KP951056516, T2q, T2n); T2r = FMA(KP951056516, T2q, T2n); T2j = T2b * T2i; T2z = T2k * T2i; T2D = W[22]; T2G = W[23]; } T2X = W[30]; T2F = T2D * T2E; T2T = T2G * T2E; T3i = FMA(KP951056516, T2Z, T2Y); T30 = FNMS(KP951056516, T2Z, T2Y); T32 = W[31]; T3h = W[6]; T3k = W[7]; T31 = T2X * T30; T3d = T32 * T30; T3j = T3h * T3i; T3t = T3k * T3i; } { E T2K, T2P, TE, T19, T1K, T2t, T37; T2K = FNMS(KP559016994, T18, T17); T19 = FMA(KP559016994, T18, T17); T1K = FMA(KP559016994, T1J, T1I); T2P = FNMS(KP559016994, T1J, T1I); TE = W[0]; T2t = W[16]; T1x = FMA(KP951056516, T1w, T19); T2u = FNMS(KP951056516, T1w, T19); T1O = FNMS(KP951056516, T1N, T1K); T2x = FMA(KP951056516, T1N, T1K); T2v = T2t * T2u; T1y = TE * T1x; T2B = T2t * T2x; T29 = TE * T1O; T2J = W[24]; T37 = W[32]; T2M = FMA(KP951056516, T2L, T2K); T38 = FNMS(KP951056516, T2L, T2K); T2R = FNMS(KP951056516, T2Q, T2P); T3b = FMA(KP951056516, T2Q, T2P); T39 = T37 * T38; T2N = T2J * T2M; T3f = T37 * T3b; } } T2V = T2J * T2R; { E T3m, T3u, T3q, T2a, T1P, T1z; T1z = W[1]; T3m = FNMS(T3k, T3l, T3j); T3u = FMA(T3h, T3l, T3t); T3q = W[9]; T2a = FNMS(T1z, T1x, T29); T1P = FMA(T1z, T1O, T1y); { E T2s, T2A, T2w, T3w, T3s; T2s = FNMS(T2k, T2r, T2j); T3w = FNMS(T3q, T3o, T3v); T3s = FMA(T3q, T3r, T3p); Im[0] = T2a - T28; Ip[0] = T28 + T2a; Rm[0] = TD + T1P; Rp[0] = TD - T1P; Im[WS(rs, 2)] = T3w - T3u; Ip[WS(rs, 2)] = T3u + T3w; Rm[WS(rs, 2)] = T3m + T3s; Rp[WS(rs, 2)] = T3m - T3s; T2A = FMA(T2b, T2r, T2z); T2w = W[17]; { E T2I, T2U, T2O, T2C, T2y, T2W, T2S; T2I = FNMS(T2G, T2H, T2F); T2U = FMA(T2D, T2H, T2T); T2O = W[25]; T2C = FNMS(T2w, T2u, T2B); T2y = FMA(T2w, T2x, T2v); T36 = FNMS(T32, T35, T31); T2W = FNMS(T2O, T2M, T2V); T2S = FMA(T2O, T2R, T2N); Im[WS(rs, 4)] = T2C - T2A; Ip[WS(rs, 4)] = T2A + T2C; Rm[WS(rs, 4)] = T2s + T2y; Rp[WS(rs, 4)] = T2s - T2y; Im[WS(rs, 6)] = T2W - T2U; Ip[WS(rs, 6)] = T2U + T2W; Rm[WS(rs, 6)] = T2I + T2S; Rp[WS(rs, 6)] = T2I - T2S; T3e = FMA(T2X, T35, T3d); T3a = W[33]; } } } } { E T55, T51, T54, T53, T5h, T5P, T5J, T3x, T4P, T5F, T5p, T43, T3R, T3S, T5l; E T5o, T4D, T5n, T5x, T4H, T4M, T5B, T5E, T4L, T4X, T5D, T5N, T4S, T4o, T4V; E T4B, T4T, T4p, T4Z, T4F, T57, T5a, T5f, T5b, T5j; { E T3X, T4O, T42, T3g, T3c, T5H; T55 = T3W + T3T; T3X = FNMS(KP250000000, T3W, T3T); T51 = W[18]; T3g = FNMS(T3a, T38, T3f); T3c = FMA(T3a, T3b, T39); T54 = W[19]; T53 = T51 * T52; Im[WS(rs, 8)] = T3g - T3e; Ip[WS(rs, 8)] = T3e + T3g; Rm[WS(rs, 8)] = T36 + T3c; Rp[WS(rs, 8)] = T36 - T3c; T5h = T54 * T52; T5H = W[28]; T4O = FMA(KP618033988, T40, T41); T42 = FNMS(KP618033988, T41, T40); T5P = T5H * T5L; T5J = T5H * T5I; { E T4I, T5m, T3Q, T3I, T3Z, T4N, T4K, T5C; T3I = FNMS(KP559016994, T3H, T3G); T4I = FMA(KP559016994, T3H, T3G); T3Z = FNMS(KP559016994, T3Y, T3X); T4N = FMA(KP559016994, T3Y, T3X); T3x = W[2]; T5m = FNMS(KP951056516, T3P, T3I); T3Q = FMA(KP951056516, T3P, T3I); T4P = FMA(KP951056516, T4O, T4N); T5F = FNMS(KP951056516, T4O, T4N); T5p = FMA(KP951056516, T42, T3Z); T43 = FNMS(KP951056516, T42, T3Z); T3R = T3x * T3Q; T3S = W[3]; T5l = W[34]; T5o = W[35]; T4D = T3S * T3Q; T5n = T5l * T5m; T5x = T5o * T5m; T4K = FNMS(KP951056516, T4J, T4I); T5C = FMA(KP951056516, T4J, T4I); T4H = W[10]; T4M = W[11]; T5B = W[26]; T5E = W[27]; T4L = T4H * T4K; T4X = T4M * T4K; T5D = T5B * T5C; T5N = T5E * T5C; } { E T58, T5d, T45, T4g, T4x, T4R, T5r; T4g = FNMS(KP559016994, T4f, T4e); T58 = FMA(KP559016994, T4f, T4e); T5d = FMA(KP559016994, T4w, T4v); T4x = FNMS(KP559016994, T4w, T4v); T45 = W[4]; T4R = W[12]; T4S = FNMS(KP951056516, T4n, T4g); T4o = FMA(KP951056516, T4n, T4g); T4V = FMA(KP951056516, T4A, T4x); T4B = FNMS(KP951056516, T4A, T4x); T4T = T4R * T4S; T4p = T45 * T4o; T4Z = T4R * T4V; T4F = T45 * T4B; T57 = W[20]; T5r = W[36]; T5s = FNMS(KP951056516, T59, T58); T5a = FMA(KP951056516, T59, T58); T5v = FMA(KP951056516, T5e, T5d); T5f = FNMS(KP951056516, T5e, T5d); T5t = T5r * T5s; T5b = T57 * T5a; T5z = T5r * T5v; } } T5j = T57 * T5f; { E T44, T4E, T5G, T5O, T5K, T4G, T4C, T4q; T44 = FNMS(T3S, T43, T3R); T4E = FMA(T3x, T43, T4D); T4q = W[5]; T5G = FNMS(T5E, T5F, T5D); T5O = FMA(T5B, T5F, T5N); T5K = W[29]; T4G = FNMS(T4q, T4o, T4F); T4C = FMA(T4q, T4B, T4p); { E T4Q, T4Y, T4U, T5Q, T5M; T4Q = FNMS(T4M, T4P, T4L); T5Q = FNMS(T5K, T5I, T5P); T5M = FMA(T5K, T5L, T5J); Im[WS(rs, 1)] = T4G - T4E; Ip[WS(rs, 1)] = T4E + T4G; Rm[WS(rs, 1)] = T44 + T4C; Rp[WS(rs, 1)] = T44 - T4C; Im[WS(rs, 7)] = T5Q - T5O; Ip[WS(rs, 7)] = T5O + T5Q; Rm[WS(rs, 7)] = T5G + T5M; Rp[WS(rs, 7)] = T5G - T5M; T4Y = FMA(T4H, T4P, T4X); T4U = W[13]; { E T56, T5i, T5c, T50, T4W, T5k, T5g; T56 = FNMS(T54, T55, T53); T5i = FMA(T51, T55, T5h); T5c = W[21]; T50 = FNMS(T4U, T4S, T4Z); T4W = FMA(T4U, T4V, T4T); T5q = FNMS(T5o, T5p, T5n); T5k = FNMS(T5c, T5a, T5j); T5g = FMA(T5c, T5f, T5b); Im[WS(rs, 3)] = T50 - T4Y; Ip[WS(rs, 3)] = T4Y + T50; Rm[WS(rs, 3)] = T4Q + T4W; Rp[WS(rs, 3)] = T4Q - T4W; Im[WS(rs, 5)] = T5k - T5i; Ip[WS(rs, 5)] = T5i + T5k; Rm[WS(rs, 5)] = T56 + T5g; Rp[WS(rs, 5)] = T56 - T5g; T5y = FMA(T5l, T5p, T5x); T5u = W[37]; } } } } } } T5A = FNMS(T5u, T5s, T5z); T5w = FMA(T5u, T5v, T5t); Im[WS(rs, 9)] = T5A - T5y; Ip[WS(rs, 9)] = T5y + T5A; Rm[WS(rs, 9)] = T5q + T5w; Rp[WS(rs, 9)] = T5q - T5w; } } } static const tw_instr twinstr[] = { {TW_FULL, 1, 20}, {TW_NEXT, 1, 0} }; static const hc2c_desc desc = { 20, "hc2cbdft_20", twinstr, &GENUS, {176, 38, 110, 0} }; void X(codelet_hc2cbdft_20) (planner *p) { X(khc2c_register) (p, hc2cbdft_20, &desc, HC2C_VIA_DFT); } #else /* HAVE_FMA */ /* Generated by: ../../../genfft/gen_hc2cdft.native -compact -variables 4 -pipeline-latency 4 -sign 1 -n 20 -dif -name hc2cbdft_20 -include hc2cb.h */ /* * This function contains 286 FP additions, 124 FP multiplications, * (or, 224 additions, 62 multiplications, 62 fused multiply/add), * 89 stack variables, 4 constants, and 80 memory accesses */ #include "hc2cb.h" static void hc2cbdft_20(R *Rp, R *Ip, R *Rm, R *Im, const R *W, stride rs, INT mb, INT me, INT ms) { DK(KP250000000, +0.250000000000000000000000000000000000000000000); DK(KP951056516, +0.951056516295153572116439333379382143405698634); DK(KP587785252, +0.587785252292473129168705954639072768597652438); DK(KP559016994, +0.559016994374947424102293417182819058860154590); { INT m; for (m = mb, W = W + ((mb - 1) * 38); m < me; m = m + 1, Rp = Rp + ms, Ip = Ip + ms, Rm = Rm - ms, Im = Im - ms, W = W + 38, MAKE_VOLATILE_STRIDE(80, rs)) { E T7, T3N, T4a, T16, T1G, T3g, T3D, T26, T1k, T3A, T3B, T1v, T2e, T48, T47; E T2d, T1L, T43, T40, T1K, T2l, T3t, T2m, T3w, T3n, T3p, TC, T2b, T4d, T4f; E T23, T2j, T1B, T1H, T3U, T3W, T3G, T3I, T11, T17; { E T3, T1C, T15, T24, T6, T12, T1F, T25; { E T1, T2, T13, T14; T1 = Rp[0]; T2 = Rm[WS(rs, 9)]; T3 = T1 + T2; T1C = T1 - T2; T13 = Ip[0]; T14 = Im[WS(rs, 9)]; T15 = T13 + T14; T24 = T13 - T14; } { E T4, T5, T1D, T1E; T4 = Rp[WS(rs, 5)]; T5 = Rm[WS(rs, 4)]; T6 = T4 + T5; T12 = T4 - T5; T1D = Ip[WS(rs, 5)]; T1E = Im[WS(rs, 4)]; T1F = T1D + T1E; T25 = T1D - T1E; } T7 = T3 + T6; T3N = T15 - T12; T4a = T1C + T1F; T16 = T12 + T15; T1G = T1C - T1F; T3g = T3 - T6; T3D = T24 - T25; T26 = T24 + T25; } { E Te, T3O, T3Y, TJ, T1e, T3h, T3r, T1R, TA, T3S, T42, TZ, T1u, T3l, T3v; E T21, Tl, T3P, T3Z, TO, T1j, T3i, T3s, T1U, Tt, T3R, T41, TU, T1p, T3k; E T3u, T1Y; { E Ta, T1a, TI, T1P, Td, TF, T1d, T1Q; { E T8, T9, TG, TH; T8 = Rp[WS(rs, 4)]; T9 = Rm[WS(rs, 5)]; Ta = T8 + T9; T1a = T8 - T9; TG = Ip[WS(rs, 4)]; TH = Im[WS(rs, 5)]; TI = TG + TH; T1P = TG - TH; } { E Tb, Tc, T1b, T1c; Tb = Rp[WS(rs, 9)]; Tc = Rm[0]; Td = Tb + Tc; TF = Tb - Tc; T1b = Ip[WS(rs, 9)]; T1c = Im[0]; T1d = T1b + T1c; T1Q = T1b - T1c; } Te = Ta + Td; T3O = TI - TF; T3Y = T1a + T1d; TJ = TF + TI; T1e = T1a - T1d; T3h = Ta - Td; T3r = T1P - T1Q; T1R = T1P + T1Q; } { E Tw, T1q, TY, T1Z, Tz, TV, T1t, T20; { E Tu, Tv, TW, TX; Tu = Rm[WS(rs, 7)]; Tv = Rp[WS(rs, 2)]; Tw = Tu + Tv; T1q = Tu - Tv; TW = Im[WS(rs, 7)]; TX = Ip[WS(rs, 2)]; TY = TW + TX; T1Z = TX - TW; } { E Tx, Ty, T1r, T1s; Tx = Rm[WS(rs, 2)]; Ty = Rp[WS(rs, 7)]; Tz = Tx + Ty; TV = Tx - Ty; T1r = Im[WS(rs, 2)]; T1s = Ip[WS(rs, 7)]; T1t = T1r + T1s; T20 = T1s - T1r; } TA = Tw + Tz; T3S = TV + TY; T42 = T1q - T1t; TZ = TV - TY; T1u = T1q + T1t; T3l = Tw - Tz; T3v = T1Z - T20; T21 = T1Z + T20; } { E Th, T1f, TN, T1S, Tk, TK, T1i, T1T; { E Tf, Tg, TL, TM; Tf = Rm[WS(rs, 3)]; Tg = Rp[WS(rs, 6)]; Th = Tf + Tg; T1f = Tf - Tg; TL = Im[WS(rs, 3)]; TM = Ip[WS(rs, 6)]; TN = TL + TM; T1S = TM - TL; } { E Ti, Tj, T1g, T1h; Ti = Rp[WS(rs, 1)]; Tj = Rm[WS(rs, 8)]; Tk = Ti + Tj; TK = Ti - Tj; T1g = Ip[WS(rs, 1)]; T1h = Im[WS(rs, 8)]; T1i = T1g + T1h; T1T = T1g - T1h; } Tl = Th + Tk; T3P = TK + TN; T3Z = T1f + T1i; TO = TK - TN; T1j = T1f - T1i; T3i = Th - Tk; T3s = T1S - T1T; T1U = T1S + T1T; } { E Tp, T1l, TT, T1W, Ts, TQ, T1o, T1X; { E Tn, To, TR, TS; Tn = Rp[WS(rs, 8)]; To = Rm[WS(rs, 1)]; Tp = Tn + To; T1l = Tn - To; TR = Ip[WS(rs, 8)]; TS = Im[WS(rs, 1)]; TT = TR + TS; T1W = TR - TS; } { E Tq, Tr, T1m, T1n; Tq = Rm[WS(rs, 6)]; Tr = Rp[WS(rs, 3)]; Ts = Tq + Tr; TQ = Tq - Tr; T1m = Im[WS(rs, 6)]; T1n = Ip[WS(rs, 3)]; T1o = T1m + T1n; T1X = T1n - T1m; } Tt = Tp + Ts; T3R = TT - TQ; T41 = T1l - T1o; TU = TQ + TT; T1p = T1l + T1o; T3k = Tp - Ts; T3u = T1W - T1X; T1Y = T1W + T1X; } T1k = T1e - T1j; T3A = T3h - T3i; T3B = T3k - T3l; T1v = T1p - T1u; T2e = T1Y - T21; T48 = T3R + T3S; T47 = T3O + T3P; T2d = T1R - T1U; T1L = TU - TZ; T43 = T41 - T42; T40 = T3Y - T3Z; T1K = TJ - TO; T2l = Te - Tl; T3t = T3r - T3s; T2m = Tt - TA; T3w = T3u - T3v; { E T3j, T3m, Tm, TB; T3j = T3h + T3i; T3m = T3k + T3l; T3n = T3j + T3m; T3p = KP559016994 * (T3j - T3m); Tm = Te + Tl; TB = Tt + TA; TC = Tm + TB; T2b = KP559016994 * (Tm - TB); } { E T4b, T4c, T3Q, T3T; T4b = T3Y + T3Z; T4c = T41 + T42; T4d = T4b + T4c; T4f = KP559016994 * (T4b - T4c); { E T1V, T22, T1z, T1A; T1V = T1R + T1U; T22 = T1Y + T21; T23 = T1V + T22; T2j = KP559016994 * (T1V - T22); T1z = T1e + T1j; T1A = T1p + T1u; T1B = KP559016994 * (T1z - T1A); T1H = T1z + T1A; } T3Q = T3O - T3P; T3T = T3R - T3S; T3U = T3Q + T3T; T3W = KP559016994 * (T3Q - T3T); { E T3E, T3F, TP, T10; T3E = T3r + T3s; T3F = T3u + T3v; T3G = T3E + T3F; T3I = KP559016994 * (T3E - T3F); TP = TJ + TO; T10 = TU + TZ; T11 = KP559016994 * (TP - T10); T17 = TP + T10; } } } { E TD, T27, T3c, T3e, T2o, T36, T2A, T2U, T1N, T2Z, T2t, T2J, T1x, T2X, T2r; E T2F, T2g, T34, T2y, T2Q; TD = T7 + TC; T27 = T23 + T26; { E T39, T3b, T38, T3a; T39 = T16 + T17; T3b = T1H + T1G; T38 = W[8]; T3a = W[9]; T3c = FMA(T38, T39, T3a * T3b); T3e = FNMS(T3a, T39, T38 * T3b); } { E T2n, T2S, T2k, T2T, T2i; T2n = FNMS(KP951056516, T2m, KP587785252 * T2l); T2S = FMA(KP951056516, T2l, KP587785252 * T2m); T2i = FNMS(KP250000000, T23, T26); T2k = T2i - T2j; T2T = T2j + T2i; T2o = T2k - T2n; T36 = T2T - T2S; T2A = T2n + T2k; T2U = T2S + T2T; } { E T1M, T2H, T1J, T2I, T1I; T1M = FMA(KP951056516, T1K, KP587785252 * T1L); T2H = FNMS(KP951056516, T1L, KP587785252 * T1K); T1I = FNMS(KP250000000, T1H, T1G); T1J = T1B + T1I; T2I = T1I - T1B; T1N = T1J - T1M; T2Z = T2I - T2H; T2t = T1M + T1J; T2J = T2H + T2I; } { E T1w, T2E, T19, T2D, T18; T1w = FMA(KP951056516, T1k, KP587785252 * T1v); T2E = FNMS(KP951056516, T1v, KP587785252 * T1k); T18 = FNMS(KP250000000, T17, T16); T19 = T11 + T18; T2D = T18 - T11; T1x = T19 + T1w; T2X = T2D + T2E; T2r = T19 - T1w; T2F = T2D - T2E; } { E T2f, T2P, T2c, T2O, T2a; T2f = FNMS(KP951056516, T2e, KP587785252 * T2d); T2P = FMA(KP951056516, T2d, KP587785252 * T2e); T2a = FNMS(KP250000000, TC, T7); T2c = T2a - T2b; T2O = T2b + T2a; T2g = T2c + T2f; T34 = T2O + T2P; T2y = T2c - T2f; T2Q = T2O - T2P; } { E T1O, T28, TE, T1y; TE = W[0]; T1y = W[1]; T1O = FMA(TE, T1x, T1y * T1N); T28 = FNMS(T1y, T1x, TE * T1N); Rp[0] = TD - T1O; Ip[0] = T27 + T28; Rm[0] = TD + T1O; Im[0] = T28 - T27; } { E T37, T3d, T33, T35; T33 = W[6]; T35 = W[7]; T37 = FNMS(T35, T36, T33 * T34); T3d = FMA(T35, T34, T33 * T36); Rp[WS(rs, 2)] = T37 - T3c; Ip[WS(rs, 2)] = T3d + T3e; Rm[WS(rs, 2)] = T37 + T3c; Im[WS(rs, 2)] = T3e - T3d; } { E T2p, T2v, T2u, T2w; { E T29, T2h, T2q, T2s; T29 = W[14]; T2h = W[15]; T2p = FNMS(T2h, T2o, T29 * T2g); T2v = FMA(T2h, T2g, T29 * T2o); T2q = W[16]; T2s = W[17]; T2u = FMA(T2q, T2r, T2s * T2t); T2w = FNMS(T2s, T2r, T2q * T2t); } Rp[WS(rs, 4)] = T2p - T2u; Ip[WS(rs, 4)] = T2v + T2w; Rm[WS(rs, 4)] = T2p + T2u; Im[WS(rs, 4)] = T2w - T2v; } { E T2B, T2L, T2K, T2M; { E T2x, T2z, T2C, T2G; T2x = W[22]; T2z = W[23]; T2B = FNMS(T2z, T2A, T2x * T2y); T2L = FMA(T2z, T2y, T2x * T2A); T2C = W[24]; T2G = W[25]; T2K = FMA(T2C, T2F, T2G * T2J); T2M = FNMS(T2G, T2F, T2C * T2J); } Rp[WS(rs, 6)] = T2B - T2K; Ip[WS(rs, 6)] = T2L + T2M; Rm[WS(rs, 6)] = T2B + T2K; Im[WS(rs, 6)] = T2M - T2L; } { E T2V, T31, T30, T32; { E T2N, T2R, T2W, T2Y; T2N = W[30]; T2R = W[31]; T2V = FNMS(T2R, T2U, T2N * T2Q); T31 = FMA(T2R, T2Q, T2N * T2U); T2W = W[32]; T2Y = W[33]; T30 = FMA(T2W, T2X, T2Y * T2Z); T32 = FNMS(T2Y, T2X, T2W * T2Z); } Rp[WS(rs, 8)] = T2V - T30; Ip[WS(rs, 8)] = T31 + T32; Rm[WS(rs, 8)] = T2V + T30; Im[WS(rs, 8)] = T32 - T31; } } { E T4F, T4P, T5c, T5e, T3y, T54, T4o, T4S, T4h, T4Z, T4x, T4N, T45, T4X, T4v; E T4J, T3K, T56, T4s, T4U; { E T4C, T4E, T4B, T4D; T4C = T3g + T3n; T4E = T3G + T3D; T4B = W[18]; T4D = W[19]; T4F = FNMS(T4D, T4E, T4B * T4C); T4P = FMA(T4D, T4C, T4B * T4E); } { E T59, T5b, T58, T5a; T59 = T3N + T3U; T5b = T4d + T4a; T58 = W[28]; T5a = W[29]; T5c = FMA(T58, T59, T5a * T5b); T5e = FNMS(T5a, T59, T58 * T5b); } { E T3x, T4n, T3q, T4m, T3o; T3x = FNMS(KP951056516, T3w, KP587785252 * T3t); T4n = FMA(KP951056516, T3t, KP587785252 * T3w); T3o = FNMS(KP250000000, T3n, T3g); T3q = T3o - T3p; T4m = T3p + T3o; T3y = T3q - T3x; T54 = T4m + T4n; T4o = T4m - T4n; T4S = T3q + T3x; } { E T49, T4M, T4g, T4L, T4e; T49 = FNMS(KP951056516, T48, KP587785252 * T47); T4M = FMA(KP951056516, T47, KP587785252 * T48); T4e = FNMS(KP250000000, T4d, T4a); T4g = T4e - T4f; T4L = T4f + T4e; T4h = T49 + T4g; T4Z = T4M + T4L; T4x = T4g - T49; T4N = T4L - T4M; } { E T44, T4I, T3X, T4H, T3V; T44 = FNMS(KP951056516, T43, KP587785252 * T40); T4I = FMA(KP951056516, T40, KP587785252 * T43); T3V = FNMS(KP250000000, T3U, T3N); T3X = T3V - T3W; T4H = T3W + T3V; T45 = T3X - T44; T4X = T4H - T4I; T4v = T3X + T44; T4J = T4H + T4I; } { E T3C, T4q, T3J, T4r, T3H; T3C = FNMS(KP951056516, T3B, KP587785252 * T3A); T4q = FMA(KP951056516, T3A, KP587785252 * T3B); T3H = FNMS(KP250000000, T3G, T3D); T3J = T3H - T3I; T4r = T3I + T3H; T3K = T3C + T3J; T56 = T4r - T4q; T4s = T4q + T4r; T4U = T3J - T3C; } { E T4O, T4Q, T4G, T4K; T4G = W[20]; T4K = W[21]; T4O = FMA(T4G, T4J, T4K * T4N); T4Q = FNMS(T4K, T4J, T4G * T4N); Rp[WS(rs, 5)] = T4F - T4O; Ip[WS(rs, 5)] = T4P + T4Q; Rm[WS(rs, 5)] = T4F + T4O; Im[WS(rs, 5)] = T4Q - T4P; } { E T57, T5d, T53, T55; T53 = W[26]; T55 = W[27]; T57 = FNMS(T55, T56, T53 * T54); T5d = FMA(T55, T54, T53 * T56); Rp[WS(rs, 7)] = T57 - T5c; Ip[WS(rs, 7)] = T5d + T5e; Rm[WS(rs, 7)] = T57 + T5c; Im[WS(rs, 7)] = T5e - T5d; } { E T3L, T4j, T4i, T4k; { E T3f, T3z, T3M, T46; T3f = W[2]; T3z = W[3]; T3L = FNMS(T3z, T3K, T3f * T3y); T4j = FMA(T3z, T3y, T3f * T3K); T3M = W[4]; T46 = W[5]; T4i = FMA(T3M, T45, T46 * T4h); T4k = FNMS(T46, T45, T3M * T4h); } Rp[WS(rs, 1)] = T3L - T4i; Ip[WS(rs, 1)] = T4j + T4k; Rm[WS(rs, 1)] = T3L + T4i; Im[WS(rs, 1)] = T4k - T4j; } { E T4t, T4z, T4y, T4A; { E T4l, T4p, T4u, T4w; T4l = W[10]; T4p = W[11]; T4t = FNMS(T4p, T4s, T4l * T4o); T4z = FMA(T4p, T4o, T4l * T4s); T4u = W[12]; T4w = W[13]; T4y = FMA(T4u, T4v, T4w * T4x); T4A = FNMS(T4w, T4v, T4u * T4x); } Rp[WS(rs, 3)] = T4t - T4y; Ip[WS(rs, 3)] = T4z + T4A; Rm[WS(rs, 3)] = T4t + T4y; Im[WS(rs, 3)] = T4A - T4z; } { E T4V, T51, T50, T52; { E T4R, T4T, T4W, T4Y; T4R = W[34]; T4T = W[35]; T4V = FNMS(T4T, T4U, T4R * T4S); T51 = FMA(T4T, T4S, T4R * T4U); T4W = W[36]; T4Y = W[37]; T50 = FMA(T4W, T4X, T4Y * T4Z); T52 = FNMS(T4Y, T4X, T4W * T4Z); } Rp[WS(rs, 9)] = T4V - T50; Ip[WS(rs, 9)] = T51 + T52; Rm[WS(rs, 9)] = T4V + T50; Im[WS(rs, 9)] = T52 - T51; } } } } } static const tw_instr twinstr[] = { {TW_FULL, 1, 20}, {TW_NEXT, 1, 0} }; static const hc2c_desc desc = { 20, "hc2cbdft_20", twinstr, &GENUS, {224, 62, 62, 0} }; void X(codelet_hc2cbdft_20) (planner *p) { X(khc2c_register) (p, hc2cbdft_20, &desc, HC2C_VIA_DFT); } #endif /* HAVE_FMA */ fftw-3.3.3/rdft/scalar/r2cb/hc2cbdft2_4.c0000644000175000017500000001310712054411037014572 00000000000000/* * Copyright (c) 2003, 2007-11 Matteo Frigo * Copyright (c) 2003, 2007-11 Massachusetts Institute of Technology * * This program is free software; you can redistribute it and/or modify * it under the terms of the GNU General Public License as published by * the Free Software Foundation; either version 2 of the License, or * (at your option) any later version. * * This program is distributed in the hope that it will be useful, * but WITHOUT ANY WARRANTY; without even the implied warranty of * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the * GNU General Public License for more details. * * You should have received a copy of the GNU General Public License * along with this program; if not, write to the Free Software * Foundation, Inc., 51 Franklin Street, Fifth Floor, Boston, MA 02110-1301 USA * */ /* This file was automatically generated --- DO NOT EDIT */ /* Generated on Sun Nov 25 07:42:07 EST 2012 */ #include "codelet-rdft.h" #ifdef HAVE_FMA /* Generated by: ../../../genfft/gen_hc2cdft.native -fma -reorder-insns -schedule-for-pipeline -compact -variables 4 -pipeline-latency 4 -sign 1 -n 4 -dif -name hc2cbdft2_4 -include hc2cb.h */ /* * This function contains 30 FP additions, 12 FP multiplications, * (or, 24 additions, 6 multiplications, 6 fused multiply/add), * 35 stack variables, 0 constants, and 16 memory accesses */ #include "hc2cb.h" static void hc2cbdft2_4(R *Rp, R *Ip, R *Rm, R *Im, const R *W, stride rs, INT mb, INT me, INT ms) { { INT m; for (m = mb, W = W + ((mb - 1) * 6); m < me; m = m + 1, Rp = Rp + ms, Ip = Ip + ms, Rm = Rm - ms, Im = Im - ms, W = W + 6, MAKE_VOLATILE_STRIDE(16, rs)) { E Ty, TB, Tw, TE, TA, TF, Tz, TG, TC; { E T4, Tg, T3, Tm, Tc, T5, Th, Ti; { E T1, T2, Ta, Tb; T1 = Rp[0]; T2 = Rm[WS(rs, 1)]; Ta = Ip[0]; Tb = Im[WS(rs, 1)]; T4 = Rp[WS(rs, 1)]; Tg = T1 - T2; T3 = T1 + T2; Tm = Ta - Tb; Tc = Ta + Tb; T5 = Rm[0]; Th = Ip[WS(rs, 1)]; Ti = Im[0]; } { E T8, Td, T7, Ts, To, Tv, Tk, Te, Tf; T8 = W[0]; { E T9, T6, Tn, Tj; T9 = T4 - T5; T6 = T4 + T5; Tn = Th - Ti; Tj = Th + Ti; Ty = Tc - T9; Td = T9 + Tc; T7 = T3 + T6; Ts = T3 - T6; To = Tm + Tn; Tv = Tm - Tn; TB = Tg + Tj; Tk = Tg - Tj; Te = T8 * Td; } Tf = W[1]; { E Tr, Tu, Tt, TD, Tx, Tp, Tl, Tq; Tr = W[2]; Tp = T8 * Tk; Tu = W[3]; Tl = FMA(Tf, Tk, Te); Tt = Tr * Ts; Tq = FNMS(Tf, Td, Tp); TD = Tu * Ts; Rm[0] = T7 + Tl; Rp[0] = T7 - Tl; Im[0] = Tq - To; Ip[0] = To + Tq; Tx = W[4]; Tw = FNMS(Tu, Tv, Tt); TE = FMA(Tr, Tv, TD); TA = W[5]; TF = Tx * TB; Tz = Tx * Ty; } } } TG = FNMS(TA, Ty, TF); TC = FMA(TA, TB, Tz); Im[WS(rs, 1)] = TG - TE; Ip[WS(rs, 1)] = TE + TG; Rm[WS(rs, 1)] = Tw + TC; Rp[WS(rs, 1)] = Tw - TC; } } } static const tw_instr twinstr[] = { {TW_FULL, 1, 4}, {TW_NEXT, 1, 0} }; static const hc2c_desc desc = { 4, "hc2cbdft2_4", twinstr, &GENUS, {24, 6, 6, 0} }; void X(codelet_hc2cbdft2_4) (planner *p) { X(khc2c_register) (p, hc2cbdft2_4, &desc, HC2C_VIA_DFT); } #else /* HAVE_FMA */ /* Generated by: ../../../genfft/gen_hc2cdft.native -compact -variables 4 -pipeline-latency 4 -sign 1 -n 4 -dif -name hc2cbdft2_4 -include hc2cb.h */ /* * This function contains 30 FP additions, 12 FP multiplications, * (or, 24 additions, 6 multiplications, 6 fused multiply/add), * 19 stack variables, 0 constants, and 16 memory accesses */ #include "hc2cb.h" static void hc2cbdft2_4(R *Rp, R *Ip, R *Rm, R *Im, const R *W, stride rs, INT mb, INT me, INT ms) { { INT m; for (m = mb, W = W + ((mb - 1) * 6); m < me; m = m + 1, Rp = Rp + ms, Ip = Ip + ms, Rm = Rm - ms, Im = Im - ms, W = W + 6, MAKE_VOLATILE_STRIDE(16, rs)) { E T3, Tl, T6, Tm, Td, Tj, Tx, Tv, Ts, Tq; { E Tf, Tc, T9, Ti; { E T1, T2, Ta, Tb; T1 = Rp[0]; T2 = Rm[WS(rs, 1)]; T3 = T1 + T2; Tf = T1 - T2; Ta = Ip[0]; Tb = Im[WS(rs, 1)]; Tc = Ta + Tb; Tl = Ta - Tb; } { E T4, T5, Tg, Th; T4 = Rp[WS(rs, 1)]; T5 = Rm[0]; T6 = T4 + T5; T9 = T4 - T5; Tg = Ip[WS(rs, 1)]; Th = Im[0]; Ti = Tg + Th; Tm = Tg - Th; } Td = T9 + Tc; Tj = Tf - Ti; Tx = Tf + Ti; Tv = Tc - T9; Ts = Tl - Tm; Tq = T3 - T6; } { E T7, Tn, Tk, To, T8, Te; T7 = T3 + T6; Tn = Tl + Tm; T8 = W[0]; Te = W[1]; Tk = FMA(T8, Td, Te * Tj); To = FNMS(Te, Td, T8 * Tj); Rp[0] = T7 - Tk; Ip[0] = Tn + To; Rm[0] = T7 + Tk; Im[0] = To - Tn; } { E Tt, Tz, Ty, TA; { E Tp, Tr, Tu, Tw; Tp = W[2]; Tr = W[3]; Tt = FNMS(Tr, Ts, Tp * Tq); Tz = FMA(Tr, Tq, Tp * Ts); Tu = W[4]; Tw = W[5]; Ty = FMA(Tu, Tv, Tw * Tx); TA = FNMS(Tw, Tv, Tu * Tx); } Rp[WS(rs, 1)] = Tt - Ty; Ip[WS(rs, 1)] = Tz + TA; Rm[WS(rs, 1)] = Tt + Ty; Im[WS(rs, 1)] = TA - Tz; } } } } static const tw_instr twinstr[] = { {TW_FULL, 1, 4}, {TW_NEXT, 1, 0} }; static const hc2c_desc desc = { 4, "hc2cbdft2_4", twinstr, &GENUS, {24, 6, 6, 0} }; void X(codelet_hc2cbdft2_4) (planner *p) { X(khc2c_register) (p, hc2cbdft2_4, &desc, HC2C_VIA_DFT); } #endif /* HAVE_FMA */ fftw-3.3.3/rdft/scalar/r2cb/hc2cbdft2_8.c0000644000175000017500000002472312054411037014604 00000000000000/* * Copyright (c) 2003, 2007-11 Matteo Frigo * Copyright (c) 2003, 2007-11 Massachusetts Institute of Technology * * This program is free software; you can redistribute it and/or modify * it under the terms of the GNU General Public License as published by * the Free Software Foundation; either version 2 of the License, or * (at your option) any later version. * * This program is distributed in the hope that it will be useful, * but WITHOUT ANY WARRANTY; without even the implied warranty of * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the * GNU General Public License for more details. * * You should have received a copy of the GNU General Public License * along with this program; if not, write to the Free Software * Foundation, Inc., 51 Franklin Street, Fifth Floor, Boston, MA 02110-1301 USA * */ /* This file was automatically generated --- DO NOT EDIT */ /* Generated on Sun Nov 25 07:42:07 EST 2012 */ #include "codelet-rdft.h" #ifdef HAVE_FMA /* Generated by: ../../../genfft/gen_hc2cdft.native -fma -reorder-insns -schedule-for-pipeline -compact -variables 4 -pipeline-latency 4 -sign 1 -n 8 -dif -name hc2cbdft2_8 -include hc2cb.h */ /* * This function contains 82 FP additions, 36 FP multiplications, * (or, 60 additions, 14 multiplications, 22 fused multiply/add), * 55 stack variables, 1 constants, and 32 memory accesses */ #include "hc2cb.h" static void hc2cbdft2_8(R *Rp, R *Ip, R *Rm, R *Im, const R *W, stride rs, INT mb, INT me, INT ms) { DK(KP707106781, +0.707106781186547524400844362104849039284835938); { INT m; for (m = mb, W = W + ((mb - 1) * 14); m < me; m = m + 1, Rp = Rp + ms, Ip = Ip + ms, Rm = Rm - ms, Im = Im - ms, W = W + 14, MAKE_VOLATILE_STRIDE(32, rs)) { E T1m, T1r, T1i, T1u, T1o, T1v, T1n, T1w, T1s; { E T1k, Tl, T1p, TE, TP, T1g, TM, T1b, T1f, T1a, TU, Tf, T1l, TH, Tw; E T1q; { E TA, T3, TN, Tk, Th, T6, TO, TD, Tb, Tm, Ta, TK, Tp, Tc, Ts; E Tt; { E T4, T5, TB, TC; { E T1, T2, Ti, Tj; T1 = Rp[0]; T2 = Rm[WS(rs, 3)]; Ti = Ip[0]; Tj = Im[WS(rs, 3)]; T4 = Rp[WS(rs, 2)]; TA = T1 - T2; T3 = T1 + T2; TN = Ti - Tj; Tk = Ti + Tj; T5 = Rm[WS(rs, 1)]; TB = Ip[WS(rs, 2)]; TC = Im[WS(rs, 1)]; } { E T8, T9, Tn, To; T8 = Rp[WS(rs, 1)]; Th = T4 - T5; T6 = T4 + T5; TO = TB - TC; TD = TB + TC; T9 = Rm[WS(rs, 2)]; Tn = Ip[WS(rs, 1)]; To = Im[WS(rs, 2)]; Tb = Rm[0]; Tm = T8 - T9; Ta = T8 + T9; TK = Tn - To; Tp = Tn + To; Tc = Rp[WS(rs, 3)]; Ts = Im[0]; Tt = Ip[WS(rs, 3)]; } } { E Tr, Td, Tu, TL, Te, T7; T1k = Tk - Th; Tl = Th + Tk; Tr = Tb - Tc; Td = Tb + Tc; TL = Tt - Ts; Tu = Ts + Tt; T1p = TA + TD; TE = TA - TD; TP = TN + TO; T1g = TN - TO; TM = TK + TL; T1b = TL - TK; T1f = Ta - Td; Te = Ta + Td; T1a = T3 - T6; T7 = T3 + T6; { E Tq, TF, TG, Tv; Tq = Tm + Tp; TF = Tm - Tp; TG = Tr - Tu; Tv = Tr + Tu; TU = T7 - Te; Tf = T7 + Te; T1l = TF - TG; TH = TF + TG; Tw = Tq - Tv; T1q = Tq + Tv; } } } { E TX, T10, T1c, T13, T1h, T1E, T1H, T1C, T1K, T1G, T1L, T1F; { E TQ, Tx, T1y, TI, Tg, Tz; TX = TP - TM; TQ = TM + TP; Tx = FMA(KP707106781, Tw, Tl); T10 = FNMS(KP707106781, Tw, Tl); T1c = T1a + T1b; T1y = T1a - T1b; T13 = FNMS(KP707106781, TH, TE); TI = FMA(KP707106781, TH, TE); Tg = W[0]; Tz = W[1]; { E T1B, T1A, T1x, T1J, T1z, T1D; { E TR, Ty, TS, TJ; T1B = T1g - T1f; T1h = T1f + T1g; T1A = W[11]; TR = Tg * TI; Ty = Tg * Tx; T1x = W[10]; T1J = T1A * T1y; TS = FNMS(Tz, Tx, TR); TJ = FMA(Tz, TI, Ty); T1z = T1x * T1y; T1m = FMA(KP707106781, T1l, T1k); T1E = FNMS(KP707106781, T1l, T1k); Im[0] = TS - TQ; Ip[0] = TQ + TS; Rm[0] = Tf + TJ; Rp[0] = Tf - TJ; T1H = FMA(KP707106781, T1q, T1p); T1r = FNMS(KP707106781, T1q, T1p); T1D = W[12]; } T1C = FNMS(T1A, T1B, T1z); T1K = FMA(T1x, T1B, T1J); T1G = W[13]; T1L = T1D * T1H; T1F = T1D * T1E; } } { E TY, T16, T12, T17, T11; { E TW, TT, T15, TV, TZ, T1M, T1I; TW = W[7]; T1M = FNMS(T1G, T1E, T1L); T1I = FMA(T1G, T1H, T1F); TT = W[6]; T15 = TW * TU; Im[WS(rs, 3)] = T1M - T1K; Ip[WS(rs, 3)] = T1K + T1M; Rm[WS(rs, 3)] = T1C + T1I; Rp[WS(rs, 3)] = T1C - T1I; TV = TT * TU; TZ = W[8]; TY = FNMS(TW, TX, TV); T16 = FMA(TT, TX, T15); T12 = W[9]; T17 = TZ * T13; T11 = TZ * T10; } { E T1e, T19, T1t, T1d, T1j, T18, T14; T1e = W[3]; T18 = FNMS(T12, T10, T17); T14 = FMA(T12, T13, T11); T19 = W[2]; T1t = T1e * T1c; Im[WS(rs, 2)] = T18 - T16; Ip[WS(rs, 2)] = T16 + T18; Rm[WS(rs, 2)] = TY + T14; Rp[WS(rs, 2)] = TY - T14; T1d = T19 * T1c; T1j = W[4]; T1i = FNMS(T1e, T1h, T1d); T1u = FMA(T19, T1h, T1t); T1o = W[5]; T1v = T1j * T1r; T1n = T1j * T1m; } } } } T1w = FNMS(T1o, T1m, T1v); T1s = FMA(T1o, T1r, T1n); Im[WS(rs, 1)] = T1w - T1u; Ip[WS(rs, 1)] = T1u + T1w; Rm[WS(rs, 1)] = T1i + T1s; Rp[WS(rs, 1)] = T1i - T1s; } } } static const tw_instr twinstr[] = { {TW_FULL, 1, 8}, {TW_NEXT, 1, 0} }; static const hc2c_desc desc = { 8, "hc2cbdft2_8", twinstr, &GENUS, {60, 14, 22, 0} }; void X(codelet_hc2cbdft2_8) (planner *p) { X(khc2c_register) (p, hc2cbdft2_8, &desc, HC2C_VIA_DFT); } #else /* HAVE_FMA */ /* Generated by: ../../../genfft/gen_hc2cdft.native -compact -variables 4 -pipeline-latency 4 -sign 1 -n 8 -dif -name hc2cbdft2_8 -include hc2cb.h */ /* * This function contains 82 FP additions, 32 FP multiplications, * (or, 68 additions, 18 multiplications, 14 fused multiply/add), * 30 stack variables, 1 constants, and 32 memory accesses */ #include "hc2cb.h" static void hc2cbdft2_8(R *Rp, R *Ip, R *Rm, R *Im, const R *W, stride rs, INT mb, INT me, INT ms) { DK(KP707106781, +0.707106781186547524400844362104849039284835938); { INT m; for (m = mb, W = W + ((mb - 1) * 14); m < me; m = m + 1, Rp = Rp + ms, Ip = Ip + ms, Rm = Rm - ms, Im = Im - ms, W = W + 14, MAKE_VOLATILE_STRIDE(32, rs)) { E T7, T1d, T1h, Tl, TG, T14, T19, TO, Te, TL, T18, T15, TB, T1e, Tw; E T1i; { E T3, TC, Tk, TM, T6, Th, TF, TN; { E T1, T2, Ti, Tj; T1 = Rp[0]; T2 = Rm[WS(rs, 3)]; T3 = T1 + T2; TC = T1 - T2; Ti = Ip[0]; Tj = Im[WS(rs, 3)]; Tk = Ti + Tj; TM = Ti - Tj; } { E T4, T5, TD, TE; T4 = Rp[WS(rs, 2)]; T5 = Rm[WS(rs, 1)]; T6 = T4 + T5; Th = T4 - T5; TD = Ip[WS(rs, 2)]; TE = Im[WS(rs, 1)]; TF = TD + TE; TN = TD - TE; } T7 = T3 + T6; T1d = Tk - Th; T1h = TC + TF; Tl = Th + Tk; TG = TC - TF; T14 = T3 - T6; T19 = TM - TN; TO = TM + TN; } { E Ta, Tm, Tp, TJ, Td, Tr, Tu, TK; { E T8, T9, Tn, To; T8 = Rp[WS(rs, 1)]; T9 = Rm[WS(rs, 2)]; Ta = T8 + T9; Tm = T8 - T9; Tn = Ip[WS(rs, 1)]; To = Im[WS(rs, 2)]; Tp = Tn + To; TJ = Tn - To; } { E Tb, Tc, Ts, Tt; Tb = Rm[0]; Tc = Rp[WS(rs, 3)]; Td = Tb + Tc; Tr = Tb - Tc; Ts = Im[0]; Tt = Ip[WS(rs, 3)]; Tu = Ts + Tt; TK = Tt - Ts; } Te = Ta + Td; TL = TJ + TK; T18 = Ta - Td; T15 = TK - TJ; { E Tz, TA, Tq, Tv; Tz = Tm - Tp; TA = Tr - Tu; TB = KP707106781 * (Tz + TA); T1e = KP707106781 * (Tz - TA); Tq = Tm + Tp; Tv = Tr + Tu; Tw = KP707106781 * (Tq - Tv); T1i = KP707106781 * (Tq + Tv); } } { E Tf, TP, TI, TQ; Tf = T7 + Te; TP = TL + TO; { E Tx, TH, Tg, Ty; Tx = Tl + Tw; TH = TB + TG; Tg = W[0]; Ty = W[1]; TI = FMA(Tg, Tx, Ty * TH); TQ = FNMS(Ty, Tx, Tg * TH); } Rp[0] = Tf - TI; Ip[0] = TP + TQ; Rm[0] = Tf + TI; Im[0] = TQ - TP; } { E T1r, T1x, T1w, T1y; { E T1o, T1q, T1n, T1p; T1o = T14 - T15; T1q = T19 - T18; T1n = W[10]; T1p = W[11]; T1r = FNMS(T1p, T1q, T1n * T1o); T1x = FMA(T1p, T1o, T1n * T1q); } { E T1t, T1v, T1s, T1u; T1t = T1d - T1e; T1v = T1i + T1h; T1s = W[12]; T1u = W[13]; T1w = FMA(T1s, T1t, T1u * T1v); T1y = FNMS(T1u, T1t, T1s * T1v); } Rp[WS(rs, 3)] = T1r - T1w; Ip[WS(rs, 3)] = T1x + T1y; Rm[WS(rs, 3)] = T1r + T1w; Im[WS(rs, 3)] = T1y - T1x; } { E TV, T11, T10, T12; { E TS, TU, TR, TT; TS = T7 - Te; TU = TO - TL; TR = W[6]; TT = W[7]; TV = FNMS(TT, TU, TR * TS); T11 = FMA(TT, TS, TR * TU); } { E TX, TZ, TW, TY; TX = Tl - Tw; TZ = TG - TB; TW = W[8]; TY = W[9]; T10 = FMA(TW, TX, TY * TZ); T12 = FNMS(TY, TX, TW * TZ); } Rp[WS(rs, 2)] = TV - T10; Ip[WS(rs, 2)] = T11 + T12; Rm[WS(rs, 2)] = TV + T10; Im[WS(rs, 2)] = T12 - T11; } { E T1b, T1l, T1k, T1m; { E T16, T1a, T13, T17; T16 = T14 + T15; T1a = T18 + T19; T13 = W[2]; T17 = W[3]; T1b = FNMS(T17, T1a, T13 * T16); T1l = FMA(T17, T16, T13 * T1a); } { E T1f, T1j, T1c, T1g; T1f = T1d + T1e; T1j = T1h - T1i; T1c = W[4]; T1g = W[5]; T1k = FMA(T1c, T1f, T1g * T1j); T1m = FNMS(T1g, T1f, T1c * T1j); } Rp[WS(rs, 1)] = T1b - T1k; Ip[WS(rs, 1)] = T1l + T1m; Rm[WS(rs, 1)] = T1b + T1k; Im[WS(rs, 1)] = T1m - T1l; } } } } static const tw_instr twinstr[] = { {TW_FULL, 1, 8}, {TW_NEXT, 1, 0} }; static const hc2c_desc desc = { 8, "hc2cbdft2_8", twinstr, &GENUS, {68, 18, 14, 0} }; void X(codelet_hc2cbdft2_8) (planner *p) { X(khc2c_register) (p, hc2cbdft2_8, &desc, HC2C_VIA_DFT); } #endif /* HAVE_FMA */ fftw-3.3.3/rdft/scalar/r2cb/hc2cbdft2_16.c0000644000175000017500000005437212054411041014661 00000000000000/* * Copyright (c) 2003, 2007-11 Matteo Frigo * Copyright (c) 2003, 2007-11 Massachusetts Institute of Technology * * This program is free software; you can redistribute it and/or modify * it under the terms of the GNU General Public License as published by * the Free Software Foundation; either version 2 of the License, or * (at your option) any later version. * * This program is distributed in the hope that it will be useful, * but WITHOUT ANY WARRANTY; without even the implied warranty of * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the * GNU General Public License for more details. * * You should have received a copy of the GNU General Public License * along with this program; if not, write to the Free Software * Foundation, Inc., 51 Franklin Street, Fifth Floor, Boston, MA 02110-1301 USA * */ /* This file was automatically generated --- DO NOT EDIT */ /* Generated on Sun Nov 25 07:42:07 EST 2012 */ #include "codelet-rdft.h" #ifdef HAVE_FMA /* Generated by: ../../../genfft/gen_hc2cdft.native -fma -reorder-insns -schedule-for-pipeline -compact -variables 4 -pipeline-latency 4 -sign 1 -n 16 -dif -name hc2cbdft2_16 -include hc2cb.h */ /* * This function contains 206 FP additions, 100 FP multiplications, * (or, 136 additions, 30 multiplications, 70 fused multiply/add), * 97 stack variables, 3 constants, and 64 memory accesses */ #include "hc2cb.h" static void hc2cbdft2_16(R *Rp, R *Ip, R *Rm, R *Im, const R *W, stride rs, INT mb, INT me, INT ms) { DK(KP923879532, +0.923879532511286756128183189396788286822416626); DK(KP414213562, +0.414213562373095048801688724209698078569671875); DK(KP707106781, +0.707106781186547524400844362104849039284835938); { INT m; for (m = mb, W = W + ((mb - 1) * 30); m < me; m = m + 1, Rp = Rp + ms, Ip = Ip + ms, Rm = Rm - ms, Im = Im - ms, W = W + 30, MAKE_VOLATILE_STRIDE(64, rs)) { E T3w, T3z, T2Y, T3D, T3x, T3m, T3u, T3C, T3y, T3o, T3k, T3E, T3A; { E T20, Tf, T3Q, T32, T3V, T3f, T2a, TN, T2f, T1m, T3G, T2G, T3L, T2T, T26; E T1F, T3M, T2N, T3H, T2W, T25, Tu, T1n, T1o, T3R, T3i, T2g, T1a, T21, T1y; E T3W, T39; { E T2R, T1B, T2S, T1E; { E T1e, T3, T1C, TA, Tx, T6, T1D, T1h, Td, T1A, TL, T1k, Ta, TC, TF; E T1z; { E T4, T5, T1f, T1g; { E T1, T2, Ty, Tz; T1 = Rp[0]; T2 = Rm[WS(rs, 7)]; Ty = Ip[0]; Tz = Im[WS(rs, 7)]; T4 = Rp[WS(rs, 4)]; T1e = T1 - T2; T3 = T1 + T2; T1C = Ty - Tz; TA = Ty + Tz; T5 = Rm[WS(rs, 3)]; } T1f = Ip[WS(rs, 4)]; T1g = Im[WS(rs, 3)]; { E Tb, Tc, TI, TJ; Tb = Rm[WS(rs, 1)]; Tx = T4 - T5; T6 = T4 + T5; T1D = T1f - T1g; T1h = T1f + T1g; Tc = Rp[WS(rs, 6)]; TI = Im[WS(rs, 1)]; TJ = Ip[WS(rs, 6)]; { E T8, TH, TK, T9, TD, TE; T8 = Rp[WS(rs, 2)]; Td = Tb + Tc; TH = Tb - Tc; T1A = TJ - TI; TK = TI + TJ; T9 = Rm[WS(rs, 5)]; TD = Ip[WS(rs, 2)]; TE = Im[WS(rs, 5)]; TL = TH + TK; T1k = TH - TK; Ta = T8 + T9; TC = T8 - T9; TF = TD + TE; T1z = TD - TE; } } } { E T2E, TB, T1l, T1i, T3d, T3e, TM, T2F; { E T7, TG, Te, T30, T31, T1j; T2E = T3 - T6; T7 = T3 + T6; T1j = TC - TF; TG = TC + TF; Te = Ta + Td; T2R = Ta - Td; TB = Tx + TA; T30 = TA - Tx; T31 = T1j - T1k; T1l = T1j + T1k; T1i = T1e - T1h; T3d = T1e + T1h; T20 = T7 - Te; Tf = T7 + Te; T3Q = FNMS(KP707106781, T31, T30); T32 = FMA(KP707106781, T31, T30); T3e = TG + TL; TM = TG - TL; } T3V = FMA(KP707106781, T3e, T3d); T3f = FNMS(KP707106781, T3e, T3d); T2a = FNMS(KP707106781, TM, TB); TN = FMA(KP707106781, TM, TB); T2F = T1A - T1z; T1B = T1z + T1A; T2f = FNMS(KP707106781, T1l, T1i); T1m = FMA(KP707106781, T1l, T1i); T3G = T2E - T2F; T2G = T2E + T2F; T2S = T1C - T1D; T1E = T1C + T1D; } } { E T34, TS, T2H, Tm, T1u, T2I, T33, TX, Tq, T14, Tp, T1v, T12, Tr, T15; E T16; { E Tj, TT, Ti, T1s, TR, Tk, TU, TV; { E Tg, Th, TP, TQ; Tg = Rp[WS(rs, 1)]; T3L = T2S - T2R; T2T = T2R + T2S; T26 = T1E - T1B; T1F = T1B + T1E; Th = Rm[WS(rs, 6)]; TP = Ip[WS(rs, 1)]; TQ = Im[WS(rs, 6)]; Tj = Rp[WS(rs, 5)]; TT = Tg - Th; Ti = Tg + Th; T1s = TP - TQ; TR = TP + TQ; Tk = Rm[WS(rs, 2)]; TU = Ip[WS(rs, 5)]; TV = Im[WS(rs, 2)]; } { E Tn, To, T10, T11; Tn = Rm[0]; { E TO, Tl, T1t, TW; TO = Tj - Tk; Tl = Tj + Tk; T1t = TU - TV; TW = TU + TV; T34 = TR - TO; TS = TO + TR; T2H = Ti - Tl; Tm = Ti + Tl; T1u = T1s + T1t; T2I = T1s - T1t; T33 = TT + TW; TX = TT - TW; To = Rp[WS(rs, 7)]; } T10 = Im[0]; T11 = Ip[WS(rs, 7)]; Tq = Rp[WS(rs, 3)]; T14 = Tn - To; Tp = Tn + To; T1v = T11 - T10; T12 = T10 + T11; Tr = Rm[WS(rs, 4)]; T15 = Ip[WS(rs, 3)]; T16 = Im[WS(rs, 4)]; } } { E T13, T1x, T18, T35, T3g, T3h, T38, TY, T19; { E T2U, T2J, T37, Tt, T36, T2V, T2M, T2K, T2L; T2U = T2H + T2I; T2J = T2H - T2I; { E TZ, Ts, T1w, T17; TZ = Tq - Tr; Ts = Tq + Tr; T1w = T15 - T16; T17 = T15 + T16; T37 = TZ + T12; T13 = TZ - T12; T2K = Tp - Ts; Tt = Tp + Ts; T1x = T1v + T1w; T2L = T1v - T1w; T36 = T14 + T17; T18 = T14 - T17; } T2V = T2L - T2K; T2M = T2K + T2L; T3M = T2J - T2M; T2N = T2J + T2M; T3H = T2V - T2U; T2W = T2U + T2V; T35 = FMA(KP414213562, T34, T33); T3g = FNMS(KP414213562, T33, T34); T25 = Tm - Tt; Tu = Tm + Tt; T3h = FNMS(KP414213562, T36, T37); T38 = FMA(KP414213562, T37, T36); } T1n = FNMS(KP414213562, TS, TX); TY = FMA(KP414213562, TX, TS); T19 = FNMS(KP414213562, T18, T13); T1o = FMA(KP414213562, T13, T18); T3R = T3h - T3g; T3i = T3g + T3h; T2g = T19 - TY; T1a = TY + T19; T21 = T1x - T1u; T1y = T1u + T1x; T3W = T35 + T38; T39 = T35 - T38; } } } { E T27, T22, T2c, T2u, T2x, T2h, T2s, T2A, T2w, T2B, T2v; { E T1K, Tv, T1G, T1N, T1Q, T1b, T2b, T1p, Tw, T1d; T1K = Tf - Tu; Tv = Tf + Tu; T1G = T1y + T1F; T1N = T1F - T1y; T1Q = FNMS(KP923879532, T1a, TN); T1b = FMA(KP923879532, T1a, TN); T2b = T1n - T1o; T1p = T1n + T1o; Tw = W[0]; T1d = W[1]; { E T1T, T1O, T1W, T1S, T1X, T1R; { E T1J, T1M, T1L, T1V, T1P, T1q; T1T = FNMS(KP923879532, T1p, T1m); T1q = FMA(KP923879532, T1p, T1m); { E T1c, T1I, T1H, T1r; T1c = Tw * T1b; T1J = W[14]; T1H = Tw * T1q; T1r = FMA(T1d, T1q, T1c); T1M = W[15]; T1L = T1J * T1K; T1I = FNMS(T1d, T1b, T1H); Rm[0] = Tv + T1r; Rp[0] = Tv - T1r; T1V = T1M * T1K; Im[0] = T1I - T1G; Ip[0] = T1G + T1I; T1P = W[16]; } T1O = FNMS(T1M, T1N, T1L); T1W = FMA(T1J, T1N, T1V); T1S = W[17]; T1X = T1P * T1T; T1R = T1P * T1Q; } { E T2r, T2n, T2q, T2p, T2z, T2t, T2o, T1Y, T1U; T27 = T25 + T26; T2r = T26 - T25; T2o = T20 - T21; T22 = T20 + T21; T1Y = FNMS(T1S, T1Q, T1X); T1U = FMA(T1S, T1T, T1R); T2n = W[22]; T2q = W[23]; Im[WS(rs, 4)] = T1Y - T1W; Ip[WS(rs, 4)] = T1W + T1Y; Rm[WS(rs, 4)] = T1O + T1U; Rp[WS(rs, 4)] = T1O - T1U; T2p = T2n * T2o; T2z = T2q * T2o; T2c = FMA(KP923879532, T2b, T2a); T2u = FNMS(KP923879532, T2b, T2a); T2x = FNMS(KP923879532, T2g, T2f); T2h = FMA(KP923879532, T2g, T2f); T2t = W[24]; T2s = FNMS(T2q, T2r, T2p); T2A = FMA(T2n, T2r, T2z); T2w = W[25]; T2B = T2t * T2x; T2v = T2t * T2u; } } } { E T28, T2k, T2e, T2l, T2d; { E T1Z, T24, T23, T2j, T29, T2C, T2y; T2C = FNMS(T2w, T2u, T2B); T2y = FMA(T2w, T2x, T2v); T1Z = W[6]; T24 = W[7]; Im[WS(rs, 6)] = T2C - T2A; Ip[WS(rs, 6)] = T2A + T2C; Rm[WS(rs, 6)] = T2s + T2y; Rp[WS(rs, 6)] = T2s - T2y; T23 = T1Z * T22; T2j = T24 * T22; T29 = W[8]; T28 = FNMS(T24, T27, T23); T2k = FMA(T1Z, T27, T2j); T2e = W[9]; T2l = T29 * T2h; T2d = T29 * T2c; } { E T4a, T4d, T3O, T4h, T4b, T40, T48, T4g, T4c, T42, T3Y; { E T3N, T47, T43, T46, T3F, T45, T4f, T3K, T3J, T3S, T3X, T3Z, T49, T41, T3T; E T3U; { E T44, T3I, T2m, T2i, T3P; T44 = FNMS(KP707106781, T3H, T3G); T3I = FMA(KP707106781, T3H, T3G); T2m = FNMS(T2e, T2c, T2l); T2i = FMA(T2e, T2h, T2d); T3N = FMA(KP707106781, T3M, T3L); T47 = FNMS(KP707106781, T3M, T3L); Im[WS(rs, 2)] = T2m - T2k; Ip[WS(rs, 2)] = T2k + T2m; Rm[WS(rs, 2)] = T28 + T2i; Rp[WS(rs, 2)] = T28 - T2i; T43 = W[26]; T46 = W[27]; T3F = W[10]; T45 = T43 * T44; T4f = T46 * T44; T3K = W[11]; T3J = T3F * T3I; T4a = FNMS(KP923879532, T3R, T3Q); T3S = FMA(KP923879532, T3R, T3Q); T3X = FNMS(KP923879532, T3W, T3V); T4d = FMA(KP923879532, T3W, T3V); T3Z = T3K * T3I; T3P = W[12]; T49 = W[28]; T41 = T3P * T3X; T3T = T3P * T3S; } T3O = FNMS(T3K, T3N, T3J); T4h = T49 * T4d; T4b = T49 * T4a; T40 = FMA(T3F, T3N, T3Z); T3U = W[13]; T48 = FNMS(T46, T47, T45); T4g = FMA(T43, T47, T4f); T4c = W[29]; T42 = FNMS(T3U, T3S, T41); T3Y = FMA(T3U, T3X, T3T); } { E T3t, T2X, T3p, T3s, T2D, T3r, T3B, T2Q, T2P, T3a, T3j, T3l, T3v, T3n, T3b; E T3c; { E T2O, T3q, T4i, T4e, T2Z; T4i = FNMS(T4c, T4a, T4h); T4e = FMA(T4c, T4d, T4b); Im[WS(rs, 3)] = T42 - T40; Ip[WS(rs, 3)] = T40 + T42; Rm[WS(rs, 3)] = T3O + T3Y; Rp[WS(rs, 3)] = T3O - T3Y; Im[WS(rs, 7)] = T4i - T4g; Ip[WS(rs, 7)] = T4g + T4i; Rm[WS(rs, 7)] = T48 + T4e; Rp[WS(rs, 7)] = T48 - T4e; T3t = FNMS(KP707106781, T2W, T2T); T2X = FMA(KP707106781, T2W, T2T); T2O = FMA(KP707106781, T2N, T2G); T3q = FNMS(KP707106781, T2N, T2G); T3p = W[18]; T3s = W[19]; T2D = W[2]; T3r = T3p * T3q; T3B = T3s * T3q; T2Q = W[3]; T2P = T2D * T2O; T3a = FMA(KP923879532, T39, T32); T3w = FNMS(KP923879532, T39, T32); T3z = FMA(KP923879532, T3i, T3f); T3j = FNMS(KP923879532, T3i, T3f); T3l = T2Q * T2O; T2Z = W[4]; T3v = W[20]; T3n = T2Z * T3j; T3b = T2Z * T3a; } T2Y = FNMS(T2Q, T2X, T2P); T3D = T3v * T3z; T3x = T3v * T3w; T3m = FMA(T2D, T2X, T3l); T3c = W[5]; T3u = FNMS(T3s, T3t, T3r); T3C = FMA(T3p, T3t, T3B); T3y = W[21]; T3o = FNMS(T3c, T3a, T3n); T3k = FMA(T3c, T3j, T3b); } } } } } T3E = FNMS(T3y, T3w, T3D); T3A = FMA(T3y, T3z, T3x); Im[WS(rs, 1)] = T3o - T3m; Ip[WS(rs, 1)] = T3m + T3o; Rm[WS(rs, 1)] = T2Y + T3k; Rp[WS(rs, 1)] = T2Y - T3k; Im[WS(rs, 5)] = T3E - T3C; Ip[WS(rs, 5)] = T3C + T3E; Rm[WS(rs, 5)] = T3u + T3A; Rp[WS(rs, 5)] = T3u - T3A; } } } static const tw_instr twinstr[] = { {TW_FULL, 1, 16}, {TW_NEXT, 1, 0} }; static const hc2c_desc desc = { 16, "hc2cbdft2_16", twinstr, &GENUS, {136, 30, 70, 0} }; void X(codelet_hc2cbdft2_16) (planner *p) { X(khc2c_register) (p, hc2cbdft2_16, &desc, HC2C_VIA_DFT); } #else /* HAVE_FMA */ /* Generated by: ../../../genfft/gen_hc2cdft.native -compact -variables 4 -pipeline-latency 4 -sign 1 -n 16 -dif -name hc2cbdft2_16 -include hc2cb.h */ /* * This function contains 206 FP additions, 84 FP multiplications, * (or, 168 additions, 46 multiplications, 38 fused multiply/add), * 60 stack variables, 3 constants, and 64 memory accesses */ #include "hc2cb.h" static void hc2cbdft2_16(R *Rp, R *Ip, R *Rm, R *Im, const R *W, stride rs, INT mb, INT me, INT ms) { DK(KP923879532, +0.923879532511286756128183189396788286822416626); DK(KP382683432, +0.382683432365089771728459984030398866761344562); DK(KP707106781, +0.707106781186547524400844362104849039284835938); { INT m; for (m = mb, W = W + ((mb - 1) * 30); m < me; m = m + 1, Rp = Rp + ms, Ip = Ip + ms, Rm = Rm - ms, Im = Im - ms, W = W + 30, MAKE_VOLATILE_STRIDE(64, rs)) { E TB, T2L, T30, T1n, Tf, T1U, T2H, T3p, T1E, T1Z, TM, T31, T2s, T3k, T1i; E T2M, Tu, T1Y, T2Q, T2X, T2T, T2Y, TY, T1d, T19, T1e, T2v, T2C, T2y, T2D; E T1x, T1V; { E T3, T1j, TA, T1B, T6, Tx, T1m, T1C, Ta, TC, TF, T1y, Td, TH, TK; E T1z; { E T1, T2, Ty, Tz; T1 = Rp[0]; T2 = Rm[WS(rs, 7)]; T3 = T1 + T2; T1j = T1 - T2; Ty = Ip[0]; Tz = Im[WS(rs, 7)]; TA = Ty + Tz; T1B = Ty - Tz; } { E T4, T5, T1k, T1l; T4 = Rp[WS(rs, 4)]; T5 = Rm[WS(rs, 3)]; T6 = T4 + T5; Tx = T4 - T5; T1k = Ip[WS(rs, 4)]; T1l = Im[WS(rs, 3)]; T1m = T1k + T1l; T1C = T1k - T1l; } { E T8, T9, TD, TE; T8 = Rp[WS(rs, 2)]; T9 = Rm[WS(rs, 5)]; Ta = T8 + T9; TC = T8 - T9; TD = Ip[WS(rs, 2)]; TE = Im[WS(rs, 5)]; TF = TD + TE; T1y = TD - TE; } { E Tb, Tc, TI, TJ; Tb = Rm[WS(rs, 1)]; Tc = Rp[WS(rs, 6)]; Td = Tb + Tc; TH = Tb - Tc; TI = Im[WS(rs, 1)]; TJ = Ip[WS(rs, 6)]; TK = TI + TJ; T1z = TJ - TI; } { E T7, Te, TG, TL; TB = Tx + TA; T2L = TA - Tx; T30 = T1j + T1m; T1n = T1j - T1m; T7 = T3 + T6; Te = Ta + Td; Tf = T7 + Te; T1U = T7 - Te; { E T2F, T2G, T1A, T1D; T2F = Ta - Td; T2G = T1B - T1C; T2H = T2F + T2G; T3p = T2G - T2F; T1A = T1y + T1z; T1D = T1B + T1C; T1E = T1A + T1D; T1Z = T1D - T1A; } TG = TC + TF; TL = TH + TK; TM = KP707106781 * (TG - TL); T31 = KP707106781 * (TG + TL); { E T2q, T2r, T1g, T1h; T2q = T3 - T6; T2r = T1z - T1y; T2s = T2q + T2r; T3k = T2q - T2r; T1g = TC - TF; T1h = TH - TK; T1i = KP707106781 * (T1g + T1h); T2M = KP707106781 * (T1g - T1h); } } } { E Ti, TT, TR, T1r, Tl, TO, TW, T1s, Tp, T14, T12, T1u, Ts, TZ, T17; E T1v; { E Tg, Th, TP, TQ; Tg = Rp[WS(rs, 1)]; Th = Rm[WS(rs, 6)]; Ti = Tg + Th; TT = Tg - Th; TP = Ip[WS(rs, 1)]; TQ = Im[WS(rs, 6)]; TR = TP + TQ; T1r = TP - TQ; } { E Tj, Tk, TU, TV; Tj = Rp[WS(rs, 5)]; Tk = Rm[WS(rs, 2)]; Tl = Tj + Tk; TO = Tj - Tk; TU = Ip[WS(rs, 5)]; TV = Im[WS(rs, 2)]; TW = TU + TV; T1s = TU - TV; } { E Tn, To, T10, T11; Tn = Rm[0]; To = Rp[WS(rs, 7)]; Tp = Tn + To; T14 = Tn - To; T10 = Im[0]; T11 = Ip[WS(rs, 7)]; T12 = T10 + T11; T1u = T11 - T10; } { E Tq, Tr, T15, T16; Tq = Rp[WS(rs, 3)]; Tr = Rm[WS(rs, 4)]; Ts = Tq + Tr; TZ = Tq - Tr; T15 = Ip[WS(rs, 3)]; T16 = Im[WS(rs, 4)]; T17 = T15 + T16; T1v = T15 - T16; } { E Tm, Tt, T2O, T2P; Tm = Ti + Tl; Tt = Tp + Ts; Tu = Tm + Tt; T1Y = Tm - Tt; T2O = TR - TO; T2P = TT + TW; T2Q = FMA(KP382683432, T2O, KP923879532 * T2P); T2X = FNMS(KP923879532, T2O, KP382683432 * T2P); } { E T2R, T2S, TS, TX; T2R = TZ + T12; T2S = T14 + T17; T2T = FMA(KP382683432, T2R, KP923879532 * T2S); T2Y = FNMS(KP923879532, T2R, KP382683432 * T2S); TS = TO + TR; TX = TT - TW; TY = FMA(KP923879532, TS, KP382683432 * TX); T1d = FNMS(KP382683432, TS, KP923879532 * TX); } { E T13, T18, T2t, T2u; T13 = TZ - T12; T18 = T14 - T17; T19 = FNMS(KP382683432, T18, KP923879532 * T13); T1e = FMA(KP382683432, T13, KP923879532 * T18); T2t = Ti - Tl; T2u = T1r - T1s; T2v = T2t - T2u; T2C = T2t + T2u; } { E T2w, T2x, T1t, T1w; T2w = Tp - Ts; T2x = T1u - T1v; T2y = T2w + T2x; T2D = T2x - T2w; T1t = T1r + T1s; T1w = T1u + T1v; T1x = T1t + T1w; T1V = T1w - T1t; } } { E Tv, T1F, T1b, T1N, T1p, T1P, T1L, T1R; Tv = Tf + Tu; T1F = T1x + T1E; { E TN, T1a, T1f, T1o; TN = TB + TM; T1a = TY + T19; T1b = TN + T1a; T1N = TN - T1a; T1f = T1d + T1e; T1o = T1i + T1n; T1p = T1f + T1o; T1P = T1o - T1f; { E T1I, T1K, T1H, T1J; T1I = Tf - Tu; T1K = T1E - T1x; T1H = W[14]; T1J = W[15]; T1L = FNMS(T1J, T1K, T1H * T1I); T1R = FMA(T1J, T1I, T1H * T1K); } } { E T1q, T1G, Tw, T1c; Tw = W[0]; T1c = W[1]; T1q = FMA(Tw, T1b, T1c * T1p); T1G = FNMS(T1c, T1b, Tw * T1p); Rp[0] = Tv - T1q; Ip[0] = T1F + T1G; Rm[0] = Tv + T1q; Im[0] = T1G - T1F; } { E T1Q, T1S, T1M, T1O; T1M = W[16]; T1O = W[17]; T1Q = FMA(T1M, T1N, T1O * T1P); T1S = FNMS(T1O, T1N, T1M * T1P); Rp[WS(rs, 4)] = T1L - T1Q; Ip[WS(rs, 4)] = T1R + T1S; Rm[WS(rs, 4)] = T1L + T1Q; Im[WS(rs, 4)] = T1S - T1R; } } { E T25, T2j, T29, T2l, T21, T2b, T2h, T2n; { E T23, T24, T27, T28; T23 = TB - TM; T24 = T1d - T1e; T25 = T23 + T24; T2j = T23 - T24; T27 = T19 - TY; T28 = T1n - T1i; T29 = T27 + T28; T2l = T28 - T27; } { E T1W, T20, T1T, T1X; T1W = T1U + T1V; T20 = T1Y + T1Z; T1T = W[6]; T1X = W[7]; T21 = FNMS(T1X, T20, T1T * T1W); T2b = FMA(T1X, T1W, T1T * T20); } { E T2e, T2g, T2d, T2f; T2e = T1U - T1V; T2g = T1Z - T1Y; T2d = W[22]; T2f = W[23]; T2h = FNMS(T2f, T2g, T2d * T2e); T2n = FMA(T2f, T2e, T2d * T2g); } { E T2a, T2c, T22, T26; T22 = W[8]; T26 = W[9]; T2a = FMA(T22, T25, T26 * T29); T2c = FNMS(T26, T25, T22 * T29); Rp[WS(rs, 2)] = T21 - T2a; Ip[WS(rs, 2)] = T2b + T2c; Rm[WS(rs, 2)] = T21 + T2a; Im[WS(rs, 2)] = T2c - T2b; } { E T2m, T2o, T2i, T2k; T2i = W[24]; T2k = W[25]; T2m = FMA(T2i, T2j, T2k * T2l); T2o = FNMS(T2k, T2j, T2i * T2l); Rp[WS(rs, 6)] = T2h - T2m; Ip[WS(rs, 6)] = T2n + T2o; Rm[WS(rs, 6)] = T2h + T2m; Im[WS(rs, 6)] = T2o - T2n; } } { E T2A, T38, T2I, T3a, T2V, T3d, T33, T3f, T2z, T2E; T2z = KP707106781 * (T2v + T2y); T2A = T2s + T2z; T38 = T2s - T2z; T2E = KP707106781 * (T2C + T2D); T2I = T2E + T2H; T3a = T2H - T2E; { E T2N, T2U, T2Z, T32; T2N = T2L + T2M; T2U = T2Q - T2T; T2V = T2N + T2U; T3d = T2N - T2U; T2Z = T2X + T2Y; T32 = T30 - T31; T33 = T2Z + T32; T3f = T32 - T2Z; } { E T2J, T35, T34, T36; { E T2p, T2B, T2K, T2W; T2p = W[2]; T2B = W[3]; T2J = FNMS(T2B, T2I, T2p * T2A); T35 = FMA(T2B, T2A, T2p * T2I); T2K = W[4]; T2W = W[5]; T34 = FMA(T2K, T2V, T2W * T33); T36 = FNMS(T2W, T2V, T2K * T33); } Rp[WS(rs, 1)] = T2J - T34; Ip[WS(rs, 1)] = T35 + T36; Rm[WS(rs, 1)] = T2J + T34; Im[WS(rs, 1)] = T36 - T35; } { E T3b, T3h, T3g, T3i; { E T37, T39, T3c, T3e; T37 = W[18]; T39 = W[19]; T3b = FNMS(T39, T3a, T37 * T38); T3h = FMA(T39, T38, T37 * T3a); T3c = W[20]; T3e = W[21]; T3g = FMA(T3c, T3d, T3e * T3f); T3i = FNMS(T3e, T3d, T3c * T3f); } Rp[WS(rs, 5)] = T3b - T3g; Ip[WS(rs, 5)] = T3h + T3i; Rm[WS(rs, 5)] = T3b + T3g; Im[WS(rs, 5)] = T3i - T3h; } } { E T3m, T3E, T3q, T3G, T3v, T3J, T3z, T3L, T3l, T3o; T3l = KP707106781 * (T2D - T2C); T3m = T3k + T3l; T3E = T3k - T3l; T3o = KP707106781 * (T2v - T2y); T3q = T3o + T3p; T3G = T3p - T3o; { E T3t, T3u, T3x, T3y; T3t = T2L - T2M; T3u = T2X - T2Y; T3v = T3t + T3u; T3J = T3t - T3u; T3x = T31 + T30; T3y = T2Q + T2T; T3z = T3x - T3y; T3L = T3y + T3x; } { E T3r, T3B, T3A, T3C; { E T3j, T3n, T3s, T3w; T3j = W[10]; T3n = W[11]; T3r = FNMS(T3n, T3q, T3j * T3m); T3B = FMA(T3n, T3m, T3j * T3q); T3s = W[12]; T3w = W[13]; T3A = FMA(T3s, T3v, T3w * T3z); T3C = FNMS(T3w, T3v, T3s * T3z); } Rp[WS(rs, 3)] = T3r - T3A; Ip[WS(rs, 3)] = T3B + T3C; Rm[WS(rs, 3)] = T3r + T3A; Im[WS(rs, 3)] = T3C - T3B; } { E T3H, T3N, T3M, T3O; { E T3D, T3F, T3I, T3K; T3D = W[26]; T3F = W[27]; T3H = FNMS(T3F, T3G, T3D * T3E); T3N = FMA(T3F, T3E, T3D * T3G); T3I = W[28]; T3K = W[29]; T3M = FMA(T3I, T3J, T3K * T3L); T3O = FNMS(T3K, T3J, T3I * T3L); } Rp[WS(rs, 7)] = T3H - T3M; Ip[WS(rs, 7)] = T3N + T3O; Rm[WS(rs, 7)] = T3H + T3M; Im[WS(rs, 7)] = T3O - T3N; } } } } } static const tw_instr twinstr[] = { {TW_FULL, 1, 16}, {TW_NEXT, 1, 0} }; static const hc2c_desc desc = { 16, "hc2cbdft2_16", twinstr, &GENUS, {168, 46, 38, 0} }; void X(codelet_hc2cbdft2_16) (planner *p) { X(khc2c_register) (p, hc2cbdft2_16, &desc, HC2C_VIA_DFT); } #endif /* HAVE_FMA */ fftw-3.3.3/rdft/scalar/r2cb/hc2cbdft2_32.c0000644000175000017500000014441112054411051014652 00000000000000/* * Copyright (c) 2003, 2007-11 Matteo Frigo * Copyright (c) 2003, 2007-11 Massachusetts Institute of Technology * * This program is free software; you can redistribute it and/or modify * it under the terms of the GNU General Public License as published by * the Free Software Foundation; either version 2 of the License, or * (at your option) any later version. * * This program is distributed in the hope that it will be useful, * but WITHOUT ANY WARRANTY; without even the implied warranty of * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the * GNU General Public License for more details. * * You should have received a copy of the GNU General Public License * along with this program; if not, write to the Free Software * Foundation, Inc., 51 Franklin Street, Fifth Floor, Boston, MA 02110-1301 USA * */ /* This file was automatically generated --- DO NOT EDIT */ /* Generated on Sun Nov 25 07:42:08 EST 2012 */ #include "codelet-rdft.h" #ifdef HAVE_FMA /* Generated by: ../../../genfft/gen_hc2cdft.native -fma -reorder-insns -schedule-for-pipeline -compact -variables 4 -pipeline-latency 4 -sign 1 -n 32 -dif -name hc2cbdft2_32 -include hc2cb.h */ /* * This function contains 498 FP additions, 260 FP multiplications, * (or, 300 additions, 62 multiplications, 198 fused multiply/add), * 165 stack variables, 7 constants, and 128 memory accesses */ #include "hc2cb.h" static void hc2cbdft2_32(R *Rp, R *Ip, R *Rm, R *Im, const R *W, stride rs, INT mb, INT me, INT ms) { DK(KP831469612, +0.831469612302545237078788377617905756738560812); DK(KP980785280, +0.980785280403230449126182236134239036973933731); DK(KP923879532, +0.923879532511286756128183189396788286822416626); DK(KP198912367, +0.198912367379658006911597622644676228597850501); DK(KP668178637, +0.668178637919298919997757686523080761552472251); DK(KP707106781, +0.707106781186547524400844362104849039284835938); DK(KP414213562, +0.414213562373095048801688724209698078569671875); { INT m; for (m = mb, W = W + ((mb - 1) * 62); m < me; m = m + 1, Rp = Rp + ms, Ip = Ip + ms, Rm = Rm - ms, Im = Im - ms, W = W + 62, MAKE_VOLATILE_STRIDE(128, rs)) { E T8e, T8h, T7S, T8l, T8f, T84, T8c, T8k, T8g, T86, T82, T8m, T8i; { E T4B, T3h, T3K, Tv, T8Y, T6T, T8L, T7i, T8X, T7f, T4Y, T1G, T4K, T1j, T4X; E T2M, T8C, T6d, T8o, T66, T8K, T6M, T4L, T2P, T4C, T3o, T5q, T4q, T8p, T6C; E T8B, T6z, T72, T2u, T75, T10, T3P, T3a, T3L, T4t, T4E, T8F, T8t, T4F, T4w; E T8E, T8w, T6E, T6l, T6F, T6s, T76, T4P, T51, T2R, T28, T8P, T90, T7k, T71; E T2p, T4R, T2x, T73, T6x, T6y; { E T3l, T16, T3m, T2H, T2E, T13, T64, T7, T3i, T2J, T1c, T3j, T1h, T2K, Te; E T1z, T6R, T6a, Tt, T3g, T6b, T1E, T6Q, Tj, T1p, Ti, T3b, T1n, Tk, T1q; E T1r; { E T1, T2, T4, T5; { E T14, T15, T2F, T2G; T14 = Ip[0]; T15 = Im[WS(rs, 15)]; T2F = Ip[WS(rs, 8)]; T2G = Im[WS(rs, 7)]; T1 = Rp[0]; T3l = T14 - T15; T16 = T14 + T15; T3m = T2F - T2G; T2H = T2F + T2G; T2 = Rm[WS(rs, 15)]; T4 = Rp[WS(rs, 8)]; T5 = Rm[WS(rs, 7)]; } { E T1b, T1e, T18, Ta, T1f, Tb, Tc, T8, T9, T1g, T1d, Td; { E T19, T3, T6, T1a; T19 = Ip[WS(rs, 4)]; T2E = T1 - T2; T3 = T1 + T2; T13 = T4 - T5; T6 = T4 + T5; T1a = Im[WS(rs, 11)]; T8 = Rp[WS(rs, 4)]; T9 = Rm[WS(rs, 11)]; T64 = T3 - T6; T7 = T3 + T6; T1b = T19 + T1a; T3i = T19 - T1a; } T1e = Im[WS(rs, 3)]; T18 = T8 - T9; Ta = T8 + T9; T1f = Ip[WS(rs, 12)]; Tb = Rm[WS(rs, 3)]; Tc = Rp[WS(rs, 12)]; T2J = T18 - T1b; T1c = T18 + T1b; T1g = T1e + T1f; T3j = T1f - T1e; T1d = Tb - Tc; Td = Tb + Tc; T1h = T1d + T1g; T2K = T1d - T1g; T6x = Ta - Td; Te = Ta + Td; } { E Tq, T1A, Tp, T3e, T1y, Tr, T1B, T1C; { E Tn, To, T1w, T1x; Tn = Rm[WS(rs, 1)]; To = Rp[WS(rs, 14)]; T1w = Im[WS(rs, 1)]; T1x = Ip[WS(rs, 14)]; Tq = Rp[WS(rs, 6)]; T1A = Tn - To; Tp = Tn + To; T3e = T1x - T1w; T1y = T1w + T1x; Tr = Rm[WS(rs, 9)]; T1B = Ip[WS(rs, 6)]; T1C = Im[WS(rs, 9)]; } { E Tg, Th, T1l, T1m; Tg = Rp[WS(rs, 2)]; { E T1v, Ts, T3f, T1D; T1v = Tq - Tr; Ts = Tq + Tr; T3f = T1B - T1C; T1D = T1B + T1C; T1z = T1v - T1y; T6R = T1v + T1y; T6a = Tp - Ts; Tt = Tp + Ts; T3g = T3e + T3f; T6b = T3e - T3f; T1E = T1A - T1D; T6Q = T1A + T1D; Th = Rm[WS(rs, 13)]; } T1l = Ip[WS(rs, 2)]; T1m = Im[WS(rs, 13)]; Tj = Rp[WS(rs, 10)]; T1p = Tg - Th; Ti = Tg + Th; T3b = T1l - T1m; T1n = T1l + T1m; Tk = Rm[WS(rs, 5)]; T1q = Ip[WS(rs, 10)]; T1r = Im[WS(rs, 5)]; } } } { E T4o, T67, T68, T4p, T2I, T1i, T2N, T1u, T1F, T2O, T6K, T17; { E Tf, T1o, T1t, Tu, T7g, T6P, T6S, T7h, T7d, T7e; { E T6O, T6N, T1k, Tl; T4o = T7 - Te; Tf = T7 + Te; T1k = Tj - Tk; Tl = Tj + Tk; { E T3c, T1s, Tm, T3d; T3c = T1q - T1r; T1s = T1q + T1r; T1o = T1k + T1n; T6O = T1n - T1k; T67 = Ti - Tl; Tm = Ti + Tl; T3d = T3b + T3c; T68 = T3b - T3c; T1t = T1p - T1s; T6N = T1p + T1s; T4B = Tm - Tt; Tu = Tm + Tt; T4p = T3g - T3d; T3h = T3d + T3g; } T7g = FNMS(KP414213562, T6N, T6O); T6P = FMA(KP414213562, T6O, T6N); T6S = FMA(KP414213562, T6R, T6Q); T7h = FNMS(KP414213562, T6Q, T6R); } T3K = Tf - Tu; Tv = Tf + Tu; T8Y = T6P + T6S; T6T = T6P - T6S; T2I = T2E - T2H; T7d = T2E + T2H; T7e = T1c + T1h; T1i = T1c - T1h; T2N = FNMS(KP414213562, T1o, T1t); T1u = FMA(KP414213562, T1t, T1o); T8L = T7h - T7g; T7i = T7g + T7h; T8X = FMA(KP707106781, T7e, T7d); T7f = FNMS(KP707106781, T7e, T7d); T1F = FNMS(KP414213562, T1E, T1z); T2O = FMA(KP414213562, T1z, T1E); T6K = T16 - T13; T17 = T13 + T16; } { E T6L, T6A, T6B, T65, T3k, T2L, T69, T6c, T3n; T4Y = T1F - T1u; T1G = T1u + T1F; T4K = FNMS(KP707106781, T1i, T17); T1j = FMA(KP707106781, T1i, T17); T2L = T2J + T2K; T6L = T2J - T2K; T6A = T67 + T68; T69 = T67 - T68; T6c = T6a + T6b; T6B = T6b - T6a; T4X = FNMS(KP707106781, T2L, T2I); T2M = FMA(KP707106781, T2L, T2I); T8C = T69 - T6c; T6d = T69 + T6c; T65 = T3j - T3i; T3k = T3i + T3j; T8o = T64 - T65; T66 = T64 + T65; T8K = FNMS(KP707106781, T6L, T6K); T6M = FMA(KP707106781, T6L, T6K); T3n = T3l + T3m; T6y = T3l - T3m; T4L = T2N - T2O; T2P = T2N + T2O; T4C = T3n - T3k; T3o = T3k + T3n; T5q = T4o - T4p; T4q = T4o + T4p; T8p = T6B - T6A; T6C = T6A + T6B; } } } { E T1M, T6V, T6f, TC, T31, T6j, T23, T6Y, T2v, T2i, TY, T6p, T6n, T35, T2n; E T2w, T24, T1R, TJ, T6i, T6g, T2Y, T1W, T25, T2q, TN, T2r, T36, T2c, T29; E TQ, T2s; { E TU, T2k, T33, T2j, TX, T2l, T2m, T34; { E T1Z, Ty, T20, T2Z, T1L, T1I, TB, T21, T2e, T2h; { E T1J, T1K, Tw, Tx, Tz, TA; Tw = Rp[WS(rs, 1)]; Tx = Rm[WS(rs, 14)]; T1J = Ip[WS(rs, 1)]; T8B = T6y - T6x; T6z = T6x + T6y; T1Z = Tw - Tx; Ty = Tw + Tx; T1K = Im[WS(rs, 14)]; Tz = Rp[WS(rs, 9)]; TA = Rm[WS(rs, 6)]; T20 = Ip[WS(rs, 9)]; T2Z = T1J - T1K; T1L = T1J + T1K; T1I = Tz - TA; TB = Tz + TA; T21 = Im[WS(rs, 6)]; } { E T2f, T2g, TV, TW; { E TS, T30, T22, TT; TS = Rp[WS(rs, 3)]; T1M = T1I + T1L; T6V = T1L - T1I; T6f = Ty - TB; TC = Ty + TB; T30 = T20 - T21; T22 = T20 + T21; TT = Rm[WS(rs, 12)]; T2f = Ip[WS(rs, 3)]; T31 = T2Z + T30; T6j = T2Z - T30; T23 = T1Z - T22; T6Y = T1Z + T22; T2e = TS - TT; TU = TS + TT; T2g = Im[WS(rs, 12)]; } TV = Rm[WS(rs, 4)]; TW = Rp[WS(rs, 11)]; T2k = Im[WS(rs, 4)]; T33 = T2f - T2g; T2h = T2f + T2g; T2j = TV - TW; TX = TV + TW; T2l = Ip[WS(rs, 11)]; } T2v = T2e - T2h; T2i = T2e + T2h; } TY = TU + TX; T6p = TU - TX; T2m = T2k + T2l; T34 = T2l - T2k; { E TF, T1T, T2W, T1S, TI, T1U, T1N, T1Q, T1V, T2X; { E T1O, T1P, TD, TE, TG, TH; TD = Rp[WS(rs, 5)]; TE = Rm[WS(rs, 10)]; T6n = T34 - T33; T35 = T33 + T34; T2n = T2j + T2m; T2w = T2j - T2m; T1N = TD - TE; TF = TD + TE; T1O = Ip[WS(rs, 5)]; T1P = Im[WS(rs, 10)]; TG = Rm[WS(rs, 2)]; TH = Rp[WS(rs, 13)]; T1T = Im[WS(rs, 2)]; T2W = T1O - T1P; T1Q = T1O + T1P; T1S = TG - TH; TI = TG + TH; T1U = Ip[WS(rs, 13)]; } T24 = T1N - T1Q; T1R = T1N + T1Q; TJ = TF + TI; T6i = TF - TI; T1V = T1T + T1U; T2X = T1U - T1T; { E T2a, T2b, TL, TM, TO, TP; TL = Rm[0]; TM = Rp[WS(rs, 15)]; T6g = T2X - T2W; T2Y = T2W + T2X; T1W = T1S + T1V; T25 = T1S - T1V; T2q = TL - TM; TN = TL + TM; T2a = Im[0]; T2b = Ip[WS(rs, 15)]; TO = Rp[WS(rs, 7)]; TP = Rm[WS(rs, 8)]; T2r = Ip[WS(rs, 7)]; T36 = T2b - T2a; T2c = T2a + T2b; T29 = TO - TP; TQ = TO + TP; T2s = Im[WS(rs, 8)]; } } } { E T2d, T4u, T4v, T6r, T6o, T6k, T8u, T8v, T6h; { E T4r, T6m, T32, T4s, T6q, T39, T8r, T8s; { E TK, TR, T37, T2t, TZ, T38; T4r = TC - TJ; TK = TC + TJ; T2d = T29 - T2c; T72 = T29 + T2c; T6m = TN - TQ; TR = TN + TQ; T37 = T2r - T2s; T2t = T2r + T2s; T32 = T2Y + T31; T4s = T31 - T2Y; T4u = TR - TY; TZ = TR + TY; T38 = T36 + T37; T6q = T36 - T37; T2u = T2q - T2t; T75 = T2q + T2t; T10 = TK + TZ; T3P = TK - TZ; T4v = T38 - T35; T39 = T35 + T38; } T8r = T6q - T6p; T6r = T6p + T6q; T3a = T32 + T39; T3L = T39 - T32; T8s = T6m - T6n; T6o = T6m + T6n; T4t = T4r - T4s; T4E = T4r + T4s; T8F = FNMS(KP414213562, T8r, T8s); T8t = FMA(KP414213562, T8s, T8r); T6k = T6i + T6j; T8u = T6j - T6i; T8v = T6f - T6g; T6h = T6f + T6g; } { E T6Z, T1Y, T4O, T26, T6W, T1X, T2o, T4N, T27; T4F = T4v - T4u; T4w = T4u + T4v; T8E = FMA(KP414213562, T8u, T8v); T8w = FNMS(KP414213562, T8v, T8u); T6Z = T1R + T1W; T1X = T1R - T1W; T6E = FMA(KP414213562, T6h, T6k); T6l = FNMS(KP414213562, T6k, T6h); T6F = FNMS(KP414213562, T6o, T6r); T6s = FMA(KP414213562, T6r, T6o); T1Y = FMA(KP707106781, T1X, T1M); T4O = FNMS(KP707106781, T1X, T1M); T26 = T24 + T25; T6W = T25 - T24; T76 = T2i + T2n; T2o = T2i - T2n; T4N = FNMS(KP707106781, T26, T23); T27 = FMA(KP707106781, T26, T23); { E T8O, T6X, T8N, T70; T8O = FMA(KP707106781, T6W, T6V); T6X = FNMS(KP707106781, T6W, T6V); T8N = FMA(KP707106781, T6Z, T6Y); T70 = FNMS(KP707106781, T6Z, T6Y); T4P = FMA(KP668178637, T4O, T4N); T51 = FNMS(KP668178637, T4N, T4O); T2R = FNMS(KP198912367, T1Y, T27); T28 = FMA(KP198912367, T27, T1Y); T8P = FMA(KP198912367, T8O, T8N); T90 = FNMS(KP198912367, T8N, T8O); T7k = FNMS(KP668178637, T6X, T70); T71 = FMA(KP668178637, T70, T6X); T2p = FMA(KP707106781, T2o, T2d); T4R = FNMS(KP707106781, T2o, T2d); } T2x = T2v + T2w; T73 = T2v - T2w; } } } { E T8S, T91, T7l, T78, T5U, T5X, T5y, T61, T5V, T5K, T5S, T60, T5W, T5M, T5I; { E T4S, T50, T4e, T4h, T3S, T4l, T4f, T44, T4c, T4k, T4g, T46, T42; { E T3Q, T3U, T40, T3Z, T3V, T3A, T3D, T3H, T3B, T3y, T3G, T3C; { E T11, T3t, T3w, T3q, T3x, T3v, T3F, T12, T2B, T2U, T3z, T2C; { E T3u, T2S, T2z, T3p, T4Q, T2y; T3u = Tv - T10; T11 = Tv + T10; T4Q = FNMS(KP707106781, T2x, T2u); T2y = FMA(KP707106781, T2x, T2u); { E T8R, T74, T8Q, T77; T8R = FMA(KP707106781, T73, T72); T74 = FNMS(KP707106781, T73, T72); T8Q = FMA(KP707106781, T76, T75); T77 = FNMS(KP707106781, T76, T75); T4S = FNMS(KP668178637, T4R, T4Q); T50 = FMA(KP668178637, T4Q, T4R); T2S = FMA(KP198912367, T2p, T2y); T2z = FNMS(KP198912367, T2y, T2p); T8S = FMA(KP198912367, T8R, T8Q); T91 = FNMS(KP198912367, T8Q, T8R); T7l = FNMS(KP668178637, T74, T77); T78 = FMA(KP668178637, T77, T74); T3Q = T3o - T3h; T3p = T3h + T3o; } T3t = W[30]; T3w = W[31]; T3q = T3a + T3p; T3x = T3p - T3a; T3v = T3t * T3u; T3F = T3w * T3u; { E T1H, T2A, T2Q, T2T; T3U = FNMS(KP923879532, T1G, T1j); T1H = FMA(KP923879532, T1G, T1j); T2A = T28 + T2z; T40 = T2z - T28; T3Z = FNMS(KP923879532, T2P, T2M); T2Q = FMA(KP923879532, T2P, T2M); T2T = T2R + T2S; T3V = T2R - T2S; T12 = W[0]; T3A = FNMS(KP980785280, T2A, T1H); T2B = FMA(KP980785280, T2A, T1H); T3D = FNMS(KP980785280, T2T, T2Q); T2U = FMA(KP980785280, T2T, T2Q); T3z = W[32]; T2C = T12 * T2B; } } { E T2V, T3s, T2D, T3r; T2D = W[1]; T3r = T12 * T2U; T3H = T3z * T3D; T3B = T3z * T3A; T2V = FMA(T2D, T2U, T2C); T3s = FNMS(T2D, T2B, T3r); T3y = FNMS(T3w, T3x, T3v); T3G = FMA(T3t, T3x, T3F); Rm[0] = T11 + T2V; Rp[0] = T11 - T2V; Im[0] = T3s - T3q; Ip[0] = T3q + T3s; T3C = W[33]; } } { E T4b, T3R, T47, T4a, T3J, T49, T4j, T3O, T3N, T43, T3W, T3T, T41, T4d, T3X; E T45, T3Y; { E T3M, T48, T3I, T3E; T3M = T3K + T3L; T48 = T3K - T3L; T3I = FNMS(T3C, T3A, T3H); T3E = FMA(T3C, T3D, T3B); T4b = T3Q - T3P; T3R = T3P + T3Q; Im[WS(rs, 8)] = T3I - T3G; Ip[WS(rs, 8)] = T3G + T3I; Rm[WS(rs, 8)] = T3y + T3E; Rp[WS(rs, 8)] = T3y - T3E; T47 = W[46]; T4a = W[47]; T3J = W[14]; T49 = T47 * T48; T4j = T4a * T48; T3O = W[15]; T3N = T3J * T3M; T43 = T3O * T3M; T3W = FMA(KP980785280, T3V, T3U); T4e = FNMS(KP980785280, T3V, T3U); T3T = W[16]; T4h = FNMS(KP980785280, T40, T3Z); T41 = FMA(KP980785280, T40, T3Z); T4d = W[48]; T3X = T3T * T3W; } T3S = FNMS(T3O, T3R, T3N); T45 = T3T * T41; T4l = T4d * T4h; T4f = T4d * T4e; T44 = FMA(T3J, T3R, T43); T3Y = W[17]; T4c = FNMS(T4a, T4b, T49); T4k = FMA(T47, T4b, T4j); T4g = W[49]; T46 = FNMS(T3Y, T3W, T45); T42 = FMA(T3Y, T41, T3X); } } { E T5v, T5r, T5w, T5A, T5G, T5F, T5B, T5g, T5j, T4I, T5n, T5h, T56, T5e, T5m; E T5i, T58, T54; { E T4n, T4A, T5d, T4H, T59, T5c, T55, T4z, T5b, T5l, T4J, T4U, T53, T5f, T4V; E T57, T4W; { E T4D, T4G, T4m, T4i, T5a, T4y, T4x; T5v = T4C - T4B; T4D = T4B + T4C; T4m = FNMS(T4g, T4e, T4l); T4i = FMA(T4g, T4h, T4f); Im[WS(rs, 4)] = T46 - T44; Ip[WS(rs, 4)] = T44 + T46; Rm[WS(rs, 4)] = T3S + T42; Rp[WS(rs, 4)] = T3S - T42; Im[WS(rs, 12)] = T4m - T4k; Ip[WS(rs, 12)] = T4k + T4m; Rm[WS(rs, 12)] = T4c + T4i; Rp[WS(rs, 12)] = T4c - T4i; T4G = T4E + T4F; T5r = T4F - T4E; T5w = T4t - T4w; T4x = T4t + T4w; T4n = W[6]; T4A = W[7]; T5d = FNMS(KP707106781, T4G, T4D); T4H = FMA(KP707106781, T4G, T4D); T5a = FNMS(KP707106781, T4x, T4q); T4y = FMA(KP707106781, T4x, T4q); T59 = W[38]; T5c = W[39]; { E T4M, T4T, T4Z, T52; T4M = FMA(KP923879532, T4L, T4K); T5A = FNMS(KP923879532, T4L, T4K); T55 = T4A * T4y; T4z = T4n * T4y; T5b = T59 * T5a; T5l = T5c * T5a; T5G = T4P + T4S; T4T = T4P - T4S; T4Z = FMA(KP923879532, T4Y, T4X); T5F = FNMS(KP923879532, T4Y, T4X); T5B = T51 + T50; T52 = T50 - T51; T4J = W[8]; T4U = FMA(KP831469612, T4T, T4M); T5g = FNMS(KP831469612, T4T, T4M); T53 = FMA(KP831469612, T52, T4Z); T5j = FNMS(KP831469612, T52, T4Z); T5f = W[40]; T4V = T4J * T4U; } } T4I = FNMS(T4A, T4H, T4z); T57 = T4J * T53; T5n = T5f * T5j; T5h = T5f * T5g; T56 = FMA(T4n, T4H, T55); T4W = W[9]; T5e = FNMS(T5c, T5d, T5b); T5m = FMA(T59, T5d, T5l); T5i = W[41]; T58 = FNMS(T4W, T4U, T57); T54 = FMA(T4W, T53, T4V); } { E T5p, T5u, T5x, T5R, T5N, T5Q, T5J, T5t, T5P, T5Z, T5z, T5C, T5H, T5T, T5D; E T5L, T5E; { E T5o, T5k, T5s, T5O; T5o = FNMS(T5i, T5g, T5n); T5k = FMA(T5i, T5j, T5h); Im[WS(rs, 2)] = T58 - T56; Ip[WS(rs, 2)] = T56 + T58; Rm[WS(rs, 2)] = T4I + T54; Rp[WS(rs, 2)] = T4I - T54; Im[WS(rs, 10)] = T5o - T5m; Ip[WS(rs, 10)] = T5m + T5o; Rm[WS(rs, 10)] = T5e + T5k; Rp[WS(rs, 10)] = T5e - T5k; T5p = W[22]; T5u = W[23]; T5x = FMA(KP707106781, T5w, T5v); T5R = FNMS(KP707106781, T5w, T5v); T5s = FMA(KP707106781, T5r, T5q); T5O = FNMS(KP707106781, T5r, T5q); T5N = W[54]; T5Q = W[55]; T5J = T5u * T5s; T5t = T5p * T5s; T5P = T5N * T5O; T5Z = T5Q * T5O; T5z = W[24]; T5U = FMA(KP831469612, T5B, T5A); T5C = FNMS(KP831469612, T5B, T5A); T5X = FMA(KP831469612, T5G, T5F); T5H = FNMS(KP831469612, T5G, T5F); T5T = W[56]; T5D = T5z * T5C; } T5y = FNMS(T5u, T5x, T5t); T5L = T5z * T5H; T61 = T5T * T5X; T5V = T5T * T5U; T5K = FMA(T5p, T5x, T5J); T5E = W[25]; T5S = FNMS(T5Q, T5R, T5P); T60 = FMA(T5N, T5R, T5Z); T5W = W[57]; T5M = FNMS(T5E, T5C, T5L); T5I = FMA(T5E, T5H, T5D); } } } { E T7P, T7L, T7K, T7Q, T7U, T80, T7Z, T7V, T9v, T9r, T9q, T9w, T9A, T9G, T9F; E T9B, T9g, T9j, T8I, T9n, T9h, T96, T9e, T9m, T9i, T98, T94; { E T7A, T7D, T6I, T7H, T7B, T7q, T7y, T7G, T7C, T7s, T7o; { E T63, T7x, T6H, T6w, T7t, T7w, T6v, T7p, T7v, T7F, T6J, T7a, T7n, T7z, T7b; E T7r, T7c; { E T6D, T6G, T62, T5Y; T7P = FNMS(KP707106781, T6C, T6z); T6D = FMA(KP707106781, T6C, T6z); T62 = FNMS(T5W, T5U, T61); T5Y = FMA(T5W, T5X, T5V); Im[WS(rs, 6)] = T5M - T5K; Ip[WS(rs, 6)] = T5K + T5M; Rm[WS(rs, 6)] = T5y + T5I; Rp[WS(rs, 6)] = T5y - T5I; Im[WS(rs, 14)] = T62 - T60; Ip[WS(rs, 14)] = T60 + T62; Rm[WS(rs, 14)] = T5S + T5Y; Rp[WS(rs, 14)] = T5S - T5Y; T6G = T6E + T6F; T7L = T6F - T6E; { E T6e, T6t, T7u, T6u; T7K = FNMS(KP707106781, T6d, T66); T6e = FMA(KP707106781, T6d, T66); T6t = T6l + T6s; T7Q = T6l - T6s; T63 = W[2]; T7x = FNMS(KP923879532, T6G, T6D); T6H = FMA(KP923879532, T6G, T6D); T7u = FNMS(KP923879532, T6t, T6e); T6u = FMA(KP923879532, T6t, T6e); T6w = W[3]; T7t = W[34]; T7w = W[35]; T6v = T63 * T6u; T7p = T6w * T6u; T7v = T7t * T7u; T7F = T7w * T7u; } { E T6U, T79, T7j, T7m; T7U = FNMS(KP923879532, T6T, T6M); T6U = FMA(KP923879532, T6T, T6M); T79 = T71 - T78; T80 = T71 + T78; T7Z = FMA(KP923879532, T7i, T7f); T7j = FNMS(KP923879532, T7i, T7f); T7m = T7k + T7l; T7V = T7k - T7l; T6J = W[4]; T7A = FNMS(KP831469612, T79, T6U); T7a = FMA(KP831469612, T79, T6U); T7D = FNMS(KP831469612, T7m, T7j); T7n = FMA(KP831469612, T7m, T7j); T7z = W[36]; T7b = T6J * T7a; } } T6I = FNMS(T6w, T6H, T6v); T7r = T6J * T7n; T7H = T7z * T7D; T7B = T7z * T7A; T7q = FMA(T63, T6H, T7p); T7c = W[5]; T7y = FNMS(T7w, T7x, T7v); T7G = FMA(T7t, T7x, T7F); T7C = W[37]; T7s = FNMS(T7c, T7a, T7r); T7o = FMA(T7c, T7n, T7b); } { E T8n, T9d, T8H, T8A, T99, T9c, T8z, T95, T9b, T9l, T8J, T8U, T93, T9f, T8V; E T97, T8W; { E T8D, T8G, T7I, T7E; T9v = FNMS(KP707106781, T8C, T8B); T8D = FMA(KP707106781, T8C, T8B); T7I = FNMS(T7C, T7A, T7H); T7E = FMA(T7C, T7D, T7B); Im[WS(rs, 1)] = T7s - T7q; Ip[WS(rs, 1)] = T7q + T7s; Rm[WS(rs, 1)] = T6I + T7o; Rp[WS(rs, 1)] = T6I - T7o; Im[WS(rs, 9)] = T7I - T7G; Ip[WS(rs, 9)] = T7G + T7I; Rm[WS(rs, 9)] = T7y + T7E; Rp[WS(rs, 9)] = T7y - T7E; T8G = T8E - T8F; T9r = T8E + T8F; { E T8q, T8x, T9a, T8y; T9q = FNMS(KP707106781, T8p, T8o); T8q = FMA(KP707106781, T8p, T8o); T8x = T8t - T8w; T9w = T8w + T8t; T8n = W[10]; T9d = FNMS(KP923879532, T8G, T8D); T8H = FMA(KP923879532, T8G, T8D); T9a = FNMS(KP923879532, T8x, T8q); T8y = FMA(KP923879532, T8x, T8q); T8A = W[11]; T99 = W[42]; T9c = W[43]; T8z = T8n * T8y; T95 = T8A * T8y; T9b = T99 * T9a; T9l = T9c * T9a; } { E T8M, T8T, T8Z, T92; T9A = FNMS(KP923879532, T8L, T8K); T8M = FMA(KP923879532, T8L, T8K); T8T = T8P - T8S; T9G = T8P + T8S; T9F = FMA(KP923879532, T8Y, T8X); T8Z = FNMS(KP923879532, T8Y, T8X); T92 = T90 + T91; T9B = T91 - T90; T8J = W[12]; T9g = FNMS(KP980785280, T8T, T8M); T8U = FMA(KP980785280, T8T, T8M); T9j = FMA(KP980785280, T92, T8Z); T93 = FNMS(KP980785280, T92, T8Z); T9f = W[44]; T8V = T8J * T8U; } } T8I = FNMS(T8A, T8H, T8z); T97 = T8J * T93; T9n = T9f * T9j; T9h = T9f * T9g; T96 = FMA(T8n, T8H, T95); T8W = W[13]; T9e = FNMS(T9c, T9d, T9b); T9m = FMA(T99, T9d, T9l); T9i = W[45]; T98 = FNMS(T8W, T8U, T97); T94 = FMA(T8W, T93, T8V); } } { E T9U, T9X, T9y, Ta1, T9V, T9K, T9S, Ta0, T9W, T9M, T9I; { E T9p, T9R, T9x, T9u, T9N, T9Q, T9t, T9J, T9P, T9Z, T9z, T9C, T9H, T9T, T9D; E T9L, T9E; { E T9o, T9k, T9O, T9s; T9o = FNMS(T9i, T9g, T9n); T9k = FMA(T9i, T9j, T9h); Im[WS(rs, 3)] = T98 - T96; Ip[WS(rs, 3)] = T96 + T98; Rm[WS(rs, 3)] = T8I + T94; Rp[WS(rs, 3)] = T8I - T94; Im[WS(rs, 11)] = T9o - T9m; Ip[WS(rs, 11)] = T9m + T9o; Rm[WS(rs, 11)] = T9e + T9k; Rp[WS(rs, 11)] = T9e - T9k; T9p = W[26]; T9R = FMA(KP923879532, T9w, T9v); T9x = FNMS(KP923879532, T9w, T9v); T9O = FMA(KP923879532, T9r, T9q); T9s = FNMS(KP923879532, T9r, T9q); T9u = W[27]; T9N = W[58]; T9Q = W[59]; T9t = T9p * T9s; T9J = T9u * T9s; T9P = T9N * T9O; T9Z = T9Q * T9O; T9z = W[28]; T9U = FNMS(KP980785280, T9B, T9A); T9C = FMA(KP980785280, T9B, T9A); T9X = FMA(KP980785280, T9G, T9F); T9H = FNMS(KP980785280, T9G, T9F); T9T = W[60]; T9D = T9z * T9C; } T9y = FNMS(T9u, T9x, T9t); T9L = T9z * T9H; Ta1 = T9T * T9X; T9V = T9T * T9U; T9K = FMA(T9p, T9x, T9J); T9E = W[29]; T9S = FNMS(T9Q, T9R, T9P); Ta0 = FMA(T9N, T9R, T9Z); T9W = W[61]; T9M = FNMS(T9E, T9C, T9L); T9I = FMA(T9E, T9H, T9D); } { E T7J, T8b, T7R, T7O, T87, T8a, T7N, T83, T89, T8j, T7T, T7W, T81, T8d, T7X; E T85, T7Y; { E Ta2, T9Y, T88, T7M; Ta2 = FNMS(T9W, T9U, Ta1); T9Y = FMA(T9W, T9X, T9V); Im[WS(rs, 7)] = T9M - T9K; Ip[WS(rs, 7)] = T9K + T9M; Rm[WS(rs, 7)] = T9y + T9I; Rp[WS(rs, 7)] = T9y - T9I; Im[WS(rs, 15)] = Ta2 - Ta0; Ip[WS(rs, 15)] = Ta0 + Ta2; Rm[WS(rs, 15)] = T9S + T9Y; Rp[WS(rs, 15)] = T9S - T9Y; T7J = W[18]; T8b = FNMS(KP923879532, T7Q, T7P); T7R = FMA(KP923879532, T7Q, T7P); T88 = FNMS(KP923879532, T7L, T7K); T7M = FMA(KP923879532, T7L, T7K); T7O = W[19]; T87 = W[50]; T8a = W[51]; T7N = T7J * T7M; T83 = T7O * T7M; T89 = T87 * T88; T8j = T8a * T88; T7T = W[20]; T8e = FNMS(KP831469612, T7V, T7U); T7W = FMA(KP831469612, T7V, T7U); T8h = FMA(KP831469612, T80, T7Z); T81 = FNMS(KP831469612, T80, T7Z); T8d = W[52]; T7X = T7T * T7W; } T7S = FNMS(T7O, T7R, T7N); T85 = T7T * T81; T8l = T8d * T8h; T8f = T8d * T8e; T84 = FMA(T7J, T7R, T83); T7Y = W[21]; T8c = FNMS(T8a, T8b, T89); T8k = FMA(T87, T8b, T8j); T8g = W[53]; T86 = FNMS(T7Y, T7W, T85); T82 = FMA(T7Y, T81, T7X); } } } } } T8m = FNMS(T8g, T8e, T8l); T8i = FMA(T8g, T8h, T8f); Im[WS(rs, 5)] = T86 - T84; Ip[WS(rs, 5)] = T84 + T86; Rm[WS(rs, 5)] = T7S + T82; Rp[WS(rs, 5)] = T7S - T82; Im[WS(rs, 13)] = T8m - T8k; Ip[WS(rs, 13)] = T8k + T8m; Rm[WS(rs, 13)] = T8c + T8i; Rp[WS(rs, 13)] = T8c - T8i; } } } static const tw_instr twinstr[] = { {TW_FULL, 1, 32}, {TW_NEXT, 1, 0} }; static const hc2c_desc desc = { 32, "hc2cbdft2_32", twinstr, &GENUS, {300, 62, 198, 0} }; void X(codelet_hc2cbdft2_32) (planner *p) { X(khc2c_register) (p, hc2cbdft2_32, &desc, HC2C_VIA_DFT); } #else /* HAVE_FMA */ /* Generated by: ../../../genfft/gen_hc2cdft.native -compact -variables 4 -pipeline-latency 4 -sign 1 -n 32 -dif -name hc2cbdft2_32 -include hc2cb.h */ /* * This function contains 498 FP additions, 208 FP multiplications, * (or, 404 additions, 114 multiplications, 94 fused multiply/add), * 102 stack variables, 7 constants, and 128 memory accesses */ #include "hc2cb.h" static void hc2cbdft2_32(R *Rp, R *Ip, R *Rm, R *Im, const R *W, stride rs, INT mb, INT me, INT ms) { DK(KP831469612, +0.831469612302545237078788377617905756738560812); DK(KP555570233, +0.555570233019602224742830813948532874374937191); DK(KP195090322, +0.195090322016128267848284868477022240927691618); DK(KP980785280, +0.980785280403230449126182236134239036973933731); DK(KP923879532, +0.923879532511286756128183189396788286822416626); DK(KP382683432, +0.382683432365089771728459984030398866761344562); DK(KP707106781, +0.707106781186547524400844362104849039284835938); { INT m; for (m = mb, W = W + ((mb - 1) * 62); m < me; m = m + 1, Rp = Rp + ms, Ip = Ip + ms, Rm = Rm - ms, Im = Im - ms, W = W + 62, MAKE_VOLATILE_STRIDE(128, rs)) { E Tf, T4a, T6h, T7Z, T6P, T8e, T1j, T4v, T2R, T4L, T5C, T7E, T6a, T7U, T3n; E T4q, TZ, T38, T2p, T4B, T7M, T7R, T2y, T4C, T5Y, T63, T6C, T86, T4i, T4n; E T6z, T85, TK, T31, T1Y, T4y, T7J, T7Q, T27, T4z, T5R, T62, T6v, T83, T4f; E T4m, T6s, T82, Tu, T4p, T6o, T8f, T6M, T80, T1G, T4K, T2I, T4w, T5J, T7T; E T67, T7F, T3g, T4b; { E T3, T2M, T16, T3k, T6, T13, T2P, T3l, Td, T3i, T1h, T2K, Ta, T3h, T1c; E T2J; { E T1, T2, T2N, T2O; T1 = Rp[0]; T2 = Rm[WS(rs, 15)]; T3 = T1 + T2; T2M = T1 - T2; { E T14, T15, T4, T5; T14 = Ip[0]; T15 = Im[WS(rs, 15)]; T16 = T14 + T15; T3k = T14 - T15; T4 = Rp[WS(rs, 8)]; T5 = Rm[WS(rs, 7)]; T6 = T4 + T5; T13 = T4 - T5; } T2N = Ip[WS(rs, 8)]; T2O = Im[WS(rs, 7)]; T2P = T2N + T2O; T3l = T2N - T2O; { E Tb, Tc, T1d, T1e, T1f, T1g; Tb = Rm[WS(rs, 3)]; Tc = Rp[WS(rs, 12)]; T1d = Tb - Tc; T1e = Im[WS(rs, 3)]; T1f = Ip[WS(rs, 12)]; T1g = T1e + T1f; Td = Tb + Tc; T3i = T1f - T1e; T1h = T1d + T1g; T2K = T1d - T1g; } { E T8, T9, T18, T19, T1a, T1b; T8 = Rp[WS(rs, 4)]; T9 = Rm[WS(rs, 11)]; T18 = T8 - T9; T19 = Ip[WS(rs, 4)]; T1a = Im[WS(rs, 11)]; T1b = T19 + T1a; Ta = T8 + T9; T3h = T19 - T1a; T1c = T18 + T1b; T2J = T18 - T1b; } } { E T7, Te, T6f, T6g; T7 = T3 + T6; Te = Ta + Td; Tf = T7 + Te; T4a = T7 - Te; T6f = T16 - T13; T6g = KP707106781 * (T2J - T2K); T6h = T6f + T6g; T7Z = T6f - T6g; } { E T6N, T6O, T17, T1i; T6N = T2M + T2P; T6O = KP707106781 * (T1c + T1h); T6P = T6N - T6O; T8e = T6O + T6N; T17 = T13 + T16; T1i = KP707106781 * (T1c - T1h); T1j = T17 + T1i; T4v = T17 - T1i; } { E T2L, T2Q, T5A, T5B; T2L = KP707106781 * (T2J + T2K); T2Q = T2M - T2P; T2R = T2L + T2Q; T4L = T2Q - T2L; T5A = T3 - T6; T5B = T3i - T3h; T5C = T5A + T5B; T7E = T5A - T5B; } { E T68, T69, T3j, T3m; T68 = Ta - Td; T69 = T3k - T3l; T6a = T68 + T69; T7U = T69 - T68; T3j = T3h + T3i; T3m = T3k + T3l; T3n = T3j + T3m; T4q = T3m - T3j; } } { E TR, T5S, T29, T2t, T2c, T5W, T2w, T37, TY, T5T, T5V, T2i, T2n, T2r, T34; E T2q, T6A, T6B; { E TL, TM, TN, TO, TP, TQ; TL = Rm[0]; TM = Rp[WS(rs, 15)]; TN = TL + TM; TO = Rp[WS(rs, 7)]; TP = Rm[WS(rs, 8)]; TQ = TO + TP; TR = TN + TQ; T5S = TN - TQ; T29 = TO - TP; T2t = TL - TM; } { E T2a, T2b, T35, T2u, T2v, T36; T2a = Im[0]; T2b = Ip[WS(rs, 15)]; T35 = T2b - T2a; T2u = Ip[WS(rs, 7)]; T2v = Im[WS(rs, 8)]; T36 = T2u - T2v; T2c = T2a + T2b; T5W = T35 - T36; T2w = T2u + T2v; T37 = T35 + T36; } { E TU, T2e, T2h, T32, TX, T2j, T2m, T33; { E TS, TT, T2f, T2g; TS = Rp[WS(rs, 3)]; TT = Rm[WS(rs, 12)]; TU = TS + TT; T2e = TS - TT; T2f = Ip[WS(rs, 3)]; T2g = Im[WS(rs, 12)]; T2h = T2f + T2g; T32 = T2f - T2g; } { E TV, TW, T2k, T2l; TV = Rm[WS(rs, 4)]; TW = Rp[WS(rs, 11)]; TX = TV + TW; T2j = TV - TW; T2k = Im[WS(rs, 4)]; T2l = Ip[WS(rs, 11)]; T2m = T2k + T2l; T33 = T2l - T2k; } TY = TU + TX; T5T = T33 - T32; T5V = TU - TX; T2i = T2e + T2h; T2n = T2j + T2m; T2r = T2j - T2m; T34 = T32 + T33; T2q = T2e - T2h; } TZ = TR + TY; T38 = T34 + T37; { E T2d, T2o, T7K, T7L; T2d = T29 - T2c; T2o = KP707106781 * (T2i - T2n); T2p = T2d + T2o; T4B = T2d - T2o; T7K = T5S - T5T; T7L = T5W - T5V; T7M = FMA(KP382683432, T7K, KP923879532 * T7L); T7R = FNMS(KP923879532, T7K, KP382683432 * T7L); } { E T2s, T2x, T5U, T5X; T2s = KP707106781 * (T2q + T2r); T2x = T2t - T2w; T2y = T2s + T2x; T4C = T2x - T2s; T5U = T5S + T5T; T5X = T5V + T5W; T5Y = FMA(KP923879532, T5U, KP382683432 * T5X); T63 = FNMS(KP382683432, T5U, KP923879532 * T5X); } T6A = T2t + T2w; T6B = KP707106781 * (T2i + T2n); T6C = T6A - T6B; T86 = T6B + T6A; { E T4g, T4h, T6x, T6y; T4g = TR - TY; T4h = T37 - T34; T4i = T4g + T4h; T4n = T4h - T4g; T6x = KP707106781 * (T2q - T2r); T6y = T29 + T2c; T6z = T6x - T6y; T85 = T6y + T6x; } } { E TC, T5L, T1I, T22, T1L, T5P, T25, T30, TJ, T5M, T5O, T1R, T1W, T20, T2X; E T1Z, T6t, T6u; { E Tw, Tx, Ty, Tz, TA, TB; Tw = Rp[WS(rs, 1)]; Tx = Rm[WS(rs, 14)]; Ty = Tw + Tx; Tz = Rp[WS(rs, 9)]; TA = Rm[WS(rs, 6)]; TB = Tz + TA; TC = Ty + TB; T5L = Ty - TB; T1I = Tz - TA; T22 = Tw - Tx; } { E T1J, T1K, T2Y, T23, T24, T2Z; T1J = Ip[WS(rs, 1)]; T1K = Im[WS(rs, 14)]; T2Y = T1J - T1K; T23 = Ip[WS(rs, 9)]; T24 = Im[WS(rs, 6)]; T2Z = T23 - T24; T1L = T1J + T1K; T5P = T2Y - T2Z; T25 = T23 + T24; T30 = T2Y + T2Z; } { E TF, T1N, T1Q, T2V, TI, T1S, T1V, T2W; { E TD, TE, T1O, T1P; TD = Rp[WS(rs, 5)]; TE = Rm[WS(rs, 10)]; TF = TD + TE; T1N = TD - TE; T1O = Ip[WS(rs, 5)]; T1P = Im[WS(rs, 10)]; T1Q = T1O + T1P; T2V = T1O - T1P; } { E TG, TH, T1T, T1U; TG = Rm[WS(rs, 2)]; TH = Rp[WS(rs, 13)]; TI = TG + TH; T1S = TG - TH; T1T = Im[WS(rs, 2)]; T1U = Ip[WS(rs, 13)]; T1V = T1T + T1U; T2W = T1U - T1T; } TJ = TF + TI; T5M = T2W - T2V; T5O = TF - TI; T1R = T1N + T1Q; T1W = T1S + T1V; T20 = T1S - T1V; T2X = T2V + T2W; T1Z = T1N - T1Q; } TK = TC + TJ; T31 = T2X + T30; { E T1M, T1X, T7H, T7I; T1M = T1I + T1L; T1X = KP707106781 * (T1R - T1W); T1Y = T1M + T1X; T4y = T1M - T1X; T7H = T5L - T5M; T7I = T5P - T5O; T7J = FNMS(KP923879532, T7I, KP382683432 * T7H); T7Q = FMA(KP923879532, T7H, KP382683432 * T7I); } { E T21, T26, T5N, T5Q; T21 = KP707106781 * (T1Z + T20); T26 = T22 - T25; T27 = T21 + T26; T4z = T26 - T21; T5N = T5L + T5M; T5Q = T5O + T5P; T5R = FNMS(KP382683432, T5Q, KP923879532 * T5N); T62 = FMA(KP382683432, T5N, KP923879532 * T5Q); } T6t = T22 + T25; T6u = KP707106781 * (T1R + T1W); T6v = T6t - T6u; T83 = T6u + T6t; { E T4d, T4e, T6q, T6r; T4d = TC - TJ; T4e = T30 - T2X; T4f = T4d - T4e; T4m = T4d + T4e; T6q = T1L - T1I; T6r = KP707106781 * (T1Z - T20); T6s = T6q + T6r; T82 = T6q - T6r; } } { E Ti, T3a, Tl, T3b, T1o, T1t, T6j, T6i, T5E, T5D, Tp, T3d, Ts, T3e, T1z; E T1E, T6m, T6l, T5H, T5G; { E T1p, T1n, T1k, T1s; { E Tg, Th, T1l, T1m; Tg = Rp[WS(rs, 2)]; Th = Rm[WS(rs, 13)]; Ti = Tg + Th; T1p = Tg - Th; T1l = Ip[WS(rs, 2)]; T1m = Im[WS(rs, 13)]; T1n = T1l + T1m; T3a = T1l - T1m; } { E Tj, Tk, T1q, T1r; Tj = Rp[WS(rs, 10)]; Tk = Rm[WS(rs, 5)]; Tl = Tj + Tk; T1k = Tj - Tk; T1q = Ip[WS(rs, 10)]; T1r = Im[WS(rs, 5)]; T1s = T1q + T1r; T3b = T1q - T1r; } T1o = T1k + T1n; T1t = T1p - T1s; T6j = T1p + T1s; T6i = T1n - T1k; T5E = T3a - T3b; T5D = Ti - Tl; } { E T1A, T1y, T1v, T1D; { E Tn, To, T1w, T1x; Tn = Rm[WS(rs, 1)]; To = Rp[WS(rs, 14)]; Tp = Tn + To; T1A = Tn - To; T1w = Im[WS(rs, 1)]; T1x = Ip[WS(rs, 14)]; T1y = T1w + T1x; T3d = T1x - T1w; } { E Tq, Tr, T1B, T1C; Tq = Rp[WS(rs, 6)]; Tr = Rm[WS(rs, 9)]; Ts = Tq + Tr; T1v = Tq - Tr; T1B = Ip[WS(rs, 6)]; T1C = Im[WS(rs, 9)]; T1D = T1B + T1C; T3e = T1B - T1C; } T1z = T1v - T1y; T1E = T1A - T1D; T6m = T1A + T1D; T6l = T1v + T1y; T5H = T3d - T3e; T5G = Tp - Ts; } { E Tm, Tt, T6k, T6n; Tm = Ti + Tl; Tt = Tp + Ts; Tu = Tm + Tt; T4p = Tm - Tt; T6k = FMA(KP382683432, T6i, KP923879532 * T6j); T6n = FMA(KP382683432, T6l, KP923879532 * T6m); T6o = T6k - T6n; T8f = T6k + T6n; } { E T6K, T6L, T1u, T1F; T6K = FNMS(KP923879532, T6i, KP382683432 * T6j); T6L = FNMS(KP923879532, T6l, KP382683432 * T6m); T6M = T6K + T6L; T80 = T6K - T6L; T1u = FMA(KP923879532, T1o, KP382683432 * T1t); T1F = FNMS(KP382683432, T1E, KP923879532 * T1z); T1G = T1u + T1F; T4K = T1F - T1u; } { E T2G, T2H, T5F, T5I; T2G = FNMS(KP382683432, T1o, KP923879532 * T1t); T2H = FMA(KP382683432, T1z, KP923879532 * T1E); T2I = T2G + T2H; T4w = T2G - T2H; T5F = T5D - T5E; T5I = T5G + T5H; T5J = KP707106781 * (T5F + T5I); T7T = KP707106781 * (T5F - T5I); } { E T65, T66, T3c, T3f; T65 = T5D + T5E; T66 = T5H - T5G; T67 = KP707106781 * (T65 + T66); T7F = KP707106781 * (T66 - T65); T3c = T3a + T3b; T3f = T3d + T3e; T3g = T3c + T3f; T4b = T3f - T3c; } } { E T11, T3s, T3p, T3u, T3K, T40, T3G, T3Y, T2T, T43, T3z, T3P, T2B, T45, T3x; E T3T; { E Tv, T10, T3E, T3F; Tv = Tf + Tu; T10 = TK + TZ; T11 = Tv + T10; T3s = Tv - T10; { E T39, T3o, T3I, T3J; T39 = T31 + T38; T3o = T3g + T3n; T3p = T39 + T3o; T3u = T3o - T39; T3I = TK - TZ; T3J = T3n - T3g; T3K = T3I + T3J; T40 = T3J - T3I; } T3E = Tf - Tu; T3F = T38 - T31; T3G = T3E + T3F; T3Y = T3E - T3F; { E T2S, T3N, T2F, T3O, T2D, T2E; T2S = T2I + T2R; T3N = T1j - T1G; T2D = FNMS(KP195090322, T1Y, KP980785280 * T27); T2E = FMA(KP195090322, T2p, KP980785280 * T2y); T2F = T2D + T2E; T3O = T2D - T2E; T2T = T2F + T2S; T43 = T3N - T3O; T3z = T2S - T2F; T3P = T3N + T3O; } { E T1H, T3S, T2A, T3R, T28, T2z; T1H = T1j + T1G; T3S = T2R - T2I; T28 = FMA(KP980785280, T1Y, KP195090322 * T27); T2z = FNMS(KP195090322, T2y, KP980785280 * T2p); T2A = T28 + T2z; T3R = T2z - T28; T2B = T1H + T2A; T45 = T3S - T3R; T3x = T1H - T2A; T3T = T3R + T3S; } } { E T2U, T3q, T12, T2C; T12 = W[0]; T2C = W[1]; T2U = FMA(T12, T2B, T2C * T2T); T3q = FNMS(T2C, T2B, T12 * T2T); Rp[0] = T11 - T2U; Ip[0] = T3p + T3q; Rm[0] = T11 + T2U; Im[0] = T3q - T3p; } { E T41, T47, T46, T48; { E T3X, T3Z, T42, T44; T3X = W[46]; T3Z = W[47]; T41 = FNMS(T3Z, T40, T3X * T3Y); T47 = FMA(T3Z, T3Y, T3X * T40); T42 = W[48]; T44 = W[49]; T46 = FMA(T42, T43, T44 * T45); T48 = FNMS(T44, T43, T42 * T45); } Rp[WS(rs, 12)] = T41 - T46; Ip[WS(rs, 12)] = T47 + T48; Rm[WS(rs, 12)] = T41 + T46; Im[WS(rs, 12)] = T48 - T47; } { E T3v, T3B, T3A, T3C; { E T3r, T3t, T3w, T3y; T3r = W[30]; T3t = W[31]; T3v = FNMS(T3t, T3u, T3r * T3s); T3B = FMA(T3t, T3s, T3r * T3u); T3w = W[32]; T3y = W[33]; T3A = FMA(T3w, T3x, T3y * T3z); T3C = FNMS(T3y, T3x, T3w * T3z); } Rp[WS(rs, 8)] = T3v - T3A; Ip[WS(rs, 8)] = T3B + T3C; Rm[WS(rs, 8)] = T3v + T3A; Im[WS(rs, 8)] = T3C - T3B; } { E T3L, T3V, T3U, T3W; { E T3D, T3H, T3M, T3Q; T3D = W[14]; T3H = W[15]; T3L = FNMS(T3H, T3K, T3D * T3G); T3V = FMA(T3H, T3G, T3D * T3K); T3M = W[16]; T3Q = W[17]; T3U = FMA(T3M, T3P, T3Q * T3T); T3W = FNMS(T3Q, T3P, T3M * T3T); } Rp[WS(rs, 4)] = T3L - T3U; Ip[WS(rs, 4)] = T3V + T3W; Rm[WS(rs, 4)] = T3L + T3U; Im[WS(rs, 4)] = T3W - T3V; } } { E T7O, T8m, T7W, T8o, T8E, T8U, T8A, T8S, T8h, T8X, T8t, T8J, T89, T8Z, T8r; E T8N; { E T7G, T7N, T8y, T8z; T7G = T7E + T7F; T7N = T7J + T7M; T7O = T7G + T7N; T8m = T7G - T7N; { E T7S, T7V, T8C, T8D; T7S = T7Q + T7R; T7V = T7T + T7U; T7W = T7S + T7V; T8o = T7V - T7S; T8C = T7J - T7M; T8D = T7U - T7T; T8E = T8C + T8D; T8U = T8D - T8C; } T8y = T7E - T7F; T8z = T7R - T7Q; T8A = T8y + T8z; T8S = T8y - T8z; { E T8g, T8H, T8d, T8I, T8b, T8c; T8g = T8e - T8f; T8H = T7Z - T80; T8b = FNMS(KP980785280, T82, KP195090322 * T83); T8c = FNMS(KP980785280, T85, KP195090322 * T86); T8d = T8b + T8c; T8I = T8b - T8c; T8h = T8d + T8g; T8X = T8H - T8I; T8t = T8g - T8d; T8J = T8H + T8I; } { E T81, T8L, T88, T8M, T84, T87; T81 = T7Z + T80; T8L = T8f + T8e; T84 = FMA(KP195090322, T82, KP980785280 * T83); T87 = FMA(KP195090322, T85, KP980785280 * T86); T88 = T84 - T87; T8M = T84 + T87; T89 = T81 + T88; T8Z = T8M + T8L; T8r = T81 - T88; T8N = T8L - T8M; } } { E T7X, T8j, T8i, T8k; { E T7D, T7P, T7Y, T8a; T7D = W[10]; T7P = W[11]; T7X = FNMS(T7P, T7W, T7D * T7O); T8j = FMA(T7P, T7O, T7D * T7W); T7Y = W[12]; T8a = W[13]; T8i = FMA(T7Y, T89, T8a * T8h); T8k = FNMS(T8a, T89, T7Y * T8h); } Rp[WS(rs, 3)] = T7X - T8i; Ip[WS(rs, 3)] = T8j + T8k; Rm[WS(rs, 3)] = T7X + T8i; Im[WS(rs, 3)] = T8k - T8j; } { E T8V, T91, T90, T92; { E T8R, T8T, T8W, T8Y; T8R = W[58]; T8T = W[59]; T8V = FNMS(T8T, T8U, T8R * T8S); T91 = FMA(T8T, T8S, T8R * T8U); T8W = W[60]; T8Y = W[61]; T90 = FMA(T8W, T8X, T8Y * T8Z); T92 = FNMS(T8Y, T8X, T8W * T8Z); } Rp[WS(rs, 15)] = T8V - T90; Ip[WS(rs, 15)] = T91 + T92; Rm[WS(rs, 15)] = T8V + T90; Im[WS(rs, 15)] = T92 - T91; } { E T8p, T8v, T8u, T8w; { E T8l, T8n, T8q, T8s; T8l = W[42]; T8n = W[43]; T8p = FNMS(T8n, T8o, T8l * T8m); T8v = FMA(T8n, T8m, T8l * T8o); T8q = W[44]; T8s = W[45]; T8u = FMA(T8q, T8r, T8s * T8t); T8w = FNMS(T8s, T8r, T8q * T8t); } Rp[WS(rs, 11)] = T8p - T8u; Ip[WS(rs, 11)] = T8v + T8w; Rm[WS(rs, 11)] = T8p + T8u; Im[WS(rs, 11)] = T8w - T8v; } { E T8F, T8P, T8O, T8Q; { E T8x, T8B, T8G, T8K; T8x = W[26]; T8B = W[27]; T8F = FNMS(T8B, T8E, T8x * T8A); T8P = FMA(T8B, T8A, T8x * T8E); T8G = W[28]; T8K = W[29]; T8O = FMA(T8G, T8J, T8K * T8N); T8Q = FNMS(T8K, T8J, T8G * T8N); } Rp[WS(rs, 7)] = T8F - T8O; Ip[WS(rs, 7)] = T8P + T8Q; Rm[WS(rs, 7)] = T8F + T8O; Im[WS(rs, 7)] = T8Q - T8P; } } { E T4k, T4S, T4s, T4U, T5a, T5q, T56, T5o, T4N, T5t, T4Z, T5f, T4F, T5v, T4X; E T5j; { E T4c, T4j, T54, T55; T4c = T4a + T4b; T4j = KP707106781 * (T4f + T4i); T4k = T4c + T4j; T4S = T4c - T4j; { E T4o, T4r, T58, T59; T4o = KP707106781 * (T4m + T4n); T4r = T4p + T4q; T4s = T4o + T4r; T4U = T4r - T4o; T58 = KP707106781 * (T4f - T4i); T59 = T4q - T4p; T5a = T58 + T59; T5q = T59 - T58; } T54 = T4a - T4b; T55 = KP707106781 * (T4n - T4m); T56 = T54 + T55; T5o = T54 - T55; { E T4M, T5d, T4J, T5e, T4H, T4I; T4M = T4K + T4L; T5d = T4v - T4w; T4H = FNMS(KP831469612, T4y, KP555570233 * T4z); T4I = FMA(KP831469612, T4B, KP555570233 * T4C); T4J = T4H + T4I; T5e = T4H - T4I; T4N = T4J + T4M; T5t = T5d - T5e; T4Z = T4M - T4J; T5f = T5d + T5e; } { E T4x, T5i, T4E, T5h, T4A, T4D; T4x = T4v + T4w; T5i = T4L - T4K; T4A = FMA(KP555570233, T4y, KP831469612 * T4z); T4D = FNMS(KP831469612, T4C, KP555570233 * T4B); T4E = T4A + T4D; T5h = T4D - T4A; T4F = T4x + T4E; T5v = T5i - T5h; T4X = T4x - T4E; T5j = T5h + T5i; } } { E T4t, T4P, T4O, T4Q; { E T49, T4l, T4u, T4G; T49 = W[6]; T4l = W[7]; T4t = FNMS(T4l, T4s, T49 * T4k); T4P = FMA(T4l, T4k, T49 * T4s); T4u = W[8]; T4G = W[9]; T4O = FMA(T4u, T4F, T4G * T4N); T4Q = FNMS(T4G, T4F, T4u * T4N); } Rp[WS(rs, 2)] = T4t - T4O; Ip[WS(rs, 2)] = T4P + T4Q; Rm[WS(rs, 2)] = T4t + T4O; Im[WS(rs, 2)] = T4Q - T4P; } { E T5r, T5x, T5w, T5y; { E T5n, T5p, T5s, T5u; T5n = W[54]; T5p = W[55]; T5r = FNMS(T5p, T5q, T5n * T5o); T5x = FMA(T5p, T5o, T5n * T5q); T5s = W[56]; T5u = W[57]; T5w = FMA(T5s, T5t, T5u * T5v); T5y = FNMS(T5u, T5t, T5s * T5v); } Rp[WS(rs, 14)] = T5r - T5w; Ip[WS(rs, 14)] = T5x + T5y; Rm[WS(rs, 14)] = T5r + T5w; Im[WS(rs, 14)] = T5y - T5x; } { E T4V, T51, T50, T52; { E T4R, T4T, T4W, T4Y; T4R = W[38]; T4T = W[39]; T4V = FNMS(T4T, T4U, T4R * T4S); T51 = FMA(T4T, T4S, T4R * T4U); T4W = W[40]; T4Y = W[41]; T50 = FMA(T4W, T4X, T4Y * T4Z); T52 = FNMS(T4Y, T4X, T4W * T4Z); } Rp[WS(rs, 10)] = T4V - T50; Ip[WS(rs, 10)] = T51 + T52; Rm[WS(rs, 10)] = T4V + T50; Im[WS(rs, 10)] = T52 - T51; } { E T5b, T5l, T5k, T5m; { E T53, T57, T5c, T5g; T53 = W[22]; T57 = W[23]; T5b = FNMS(T57, T5a, T53 * T56); T5l = FMA(T57, T56, T53 * T5a); T5c = W[24]; T5g = W[25]; T5k = FMA(T5c, T5f, T5g * T5j); T5m = FNMS(T5g, T5f, T5c * T5j); } Rp[WS(rs, 6)] = T5b - T5k; Ip[WS(rs, 6)] = T5l + T5m; Rm[WS(rs, 6)] = T5b + T5k; Im[WS(rs, 6)] = T5m - T5l; } } { E T60, T6W, T6c, T6Y, T7e, T7u, T7a, T7s, T6R, T7x, T73, T7j, T6F, T7z, T71; E T7n; { E T5K, T5Z, T78, T79; T5K = T5C + T5J; T5Z = T5R + T5Y; T60 = T5K + T5Z; T6W = T5K - T5Z; { E T64, T6b, T7c, T7d; T64 = T62 + T63; T6b = T67 + T6a; T6c = T64 + T6b; T6Y = T6b - T64; T7c = T5R - T5Y; T7d = T6a - T67; T7e = T7c + T7d; T7u = T7d - T7c; } T78 = T5C - T5J; T79 = T63 - T62; T7a = T78 + T79; T7s = T78 - T79; { E T6Q, T7h, T6J, T7i, T6H, T6I; T6Q = T6M + T6P; T7h = T6h - T6o; T6H = FNMS(KP555570233, T6s, KP831469612 * T6v); T6I = FMA(KP555570233, T6z, KP831469612 * T6C); T6J = T6H + T6I; T7i = T6H - T6I; T6R = T6J + T6Q; T7x = T7h - T7i; T73 = T6Q - T6J; T7j = T7h + T7i; } { E T6p, T7m, T6E, T7l, T6w, T6D; T6p = T6h + T6o; T7m = T6P - T6M; T6w = FMA(KP831469612, T6s, KP555570233 * T6v); T6D = FNMS(KP555570233, T6C, KP831469612 * T6z); T6E = T6w + T6D; T7l = T6D - T6w; T6F = T6p + T6E; T7z = T7m - T7l; T71 = T6p - T6E; T7n = T7l + T7m; } } { E T6d, T6T, T6S, T6U; { E T5z, T61, T6e, T6G; T5z = W[2]; T61 = W[3]; T6d = FNMS(T61, T6c, T5z * T60); T6T = FMA(T61, T60, T5z * T6c); T6e = W[4]; T6G = W[5]; T6S = FMA(T6e, T6F, T6G * T6R); T6U = FNMS(T6G, T6F, T6e * T6R); } Rp[WS(rs, 1)] = T6d - T6S; Ip[WS(rs, 1)] = T6T + T6U; Rm[WS(rs, 1)] = T6d + T6S; Im[WS(rs, 1)] = T6U - T6T; } { E T7v, T7B, T7A, T7C; { E T7r, T7t, T7w, T7y; T7r = W[50]; T7t = W[51]; T7v = FNMS(T7t, T7u, T7r * T7s); T7B = FMA(T7t, T7s, T7r * T7u); T7w = W[52]; T7y = W[53]; T7A = FMA(T7w, T7x, T7y * T7z); T7C = FNMS(T7y, T7x, T7w * T7z); } Rp[WS(rs, 13)] = T7v - T7A; Ip[WS(rs, 13)] = T7B + T7C; Rm[WS(rs, 13)] = T7v + T7A; Im[WS(rs, 13)] = T7C - T7B; } { E T6Z, T75, T74, T76; { E T6V, T6X, T70, T72; T6V = W[34]; T6X = W[35]; T6Z = FNMS(T6X, T6Y, T6V * T6W); T75 = FMA(T6X, T6W, T6V * T6Y); T70 = W[36]; T72 = W[37]; T74 = FMA(T70, T71, T72 * T73); T76 = FNMS(T72, T71, T70 * T73); } Rp[WS(rs, 9)] = T6Z - T74; Ip[WS(rs, 9)] = T75 + T76; Rm[WS(rs, 9)] = T6Z + T74; Im[WS(rs, 9)] = T76 - T75; } { E T7f, T7p, T7o, T7q; { E T77, T7b, T7g, T7k; T77 = W[18]; T7b = W[19]; T7f = FNMS(T7b, T7e, T77 * T7a); T7p = FMA(T7b, T7a, T77 * T7e); T7g = W[20]; T7k = W[21]; T7o = FMA(T7g, T7j, T7k * T7n); T7q = FNMS(T7k, T7j, T7g * T7n); } Rp[WS(rs, 5)] = T7f - T7o; Ip[WS(rs, 5)] = T7p + T7q; Rm[WS(rs, 5)] = T7f + T7o; Im[WS(rs, 5)] = T7q - T7p; } } } } } static const tw_instr twinstr[] = { {TW_FULL, 1, 32}, {TW_NEXT, 1, 0} }; static const hc2c_desc desc = { 32, "hc2cbdft2_32", twinstr, &GENUS, {404, 114, 94, 0} }; void X(codelet_hc2cbdft2_32) (planner *p) { X(khc2c_register) (p, hc2cbdft2_32, &desc, HC2C_VIA_DFT); } #endif /* HAVE_FMA */ fftw-3.3.3/rdft/scalar/r2cb/hc2cbdft2_20.c0000644000175000017500000007337312054411042014657 00000000000000/* * Copyright (c) 2003, 2007-11 Matteo Frigo * Copyright (c) 2003, 2007-11 Massachusetts Institute of Technology * * This program is free software; you can redistribute it and/or modify * it under the terms of the GNU General Public License as published by * the Free Software Foundation; either version 2 of the License, or * (at your option) any later version. * * This program is distributed in the hope that it will be useful, * but WITHOUT ANY WARRANTY; without even the implied warranty of * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the * GNU General Public License for more details. * * You should have received a copy of the GNU General Public License * along with this program; if not, write to the Free Software * Foundation, Inc., 51 Franklin Street, Fifth Floor, Boston, MA 02110-1301 USA * */ /* This file was automatically generated --- DO NOT EDIT */ /* Generated on Sun Nov 25 07:42:09 EST 2012 */ #include "codelet-rdft.h" #ifdef HAVE_FMA /* Generated by: ../../../genfft/gen_hc2cdft.native -fma -reorder-insns -schedule-for-pipeline -compact -variables 4 -pipeline-latency 4 -sign 1 -n 20 -dif -name hc2cbdft2_20 -include hc2cb.h */ /* * This function contains 286 FP additions, 148 FP multiplications, * (or, 176 additions, 38 multiplications, 110 fused multiply/add), * 122 stack variables, 4 constants, and 80 memory accesses */ #include "hc2cb.h" static void hc2cbdft2_20(R *Rp, R *Ip, R *Rm, R *Im, const R *W, stride rs, INT mb, INT me, INT ms) { DK(KP951056516, +0.951056516295153572116439333379382143405698634); DK(KP559016994, +0.559016994374947424102293417182819058860154590); DK(KP250000000, +0.250000000000000000000000000000000000000000000); DK(KP618033988, +0.618033988749894848204586834365638117720309180); { INT m; for (m = mb, W = W + ((mb - 1) * 38); m < me; m = m + 1, Rp = Rp + ms, Ip = Ip + ms, Rm = Rm - ms, Im = Im - ms, W = W + 38, MAKE_VOLATILE_STRIDE(80, rs)) { E T5s, T5v, T5t, T5z, T5q, T5y, T5u, T5A, T5w; { E T3T, T27, T2o, T41, T2p, T40, TU, T15, T2Q, T1N, T2L, T1w, T59, T4n, T5e; E T4A, T2m, T24, T2Z, T2h, T4J, T3P, T3Y, T3W, T2d, TJ, T3H, T2c, TD, T52; E T3G, T1E, T4f, T5I, T4e, T4w, T5L, T4v, T1J, T1H; { E T1A, T3, T25, TI, TF, T6, T26, T1D, TO, T47, T3z, Te, T1S, T3M, T1e; E T4k, TZ, T4a, T3C, Tt, T1Z, T3J, T1p, T4h, T14, T4b, T3D, TA, T22, T3K; E T1u, T4i, Ti, T1f, Th, T1T, TS, Tj, T1g, T1h; { E T4, T5, T1B, T1C; { E T1, T2, TG, TH; T1 = Rp[0]; T2 = Rm[WS(rs, 9)]; TG = Ip[0]; TH = Im[WS(rs, 9)]; T4 = Rp[WS(rs, 5)]; T1A = T1 - T2; T3 = T1 + T2; T25 = TG - TH; TI = TG + TH; T5 = Rm[WS(rs, 4)]; T1B = Ip[WS(rs, 5)]; T1C = Im[WS(rs, 4)]; } { E Tq, T1l, Tp, T1X, TY, Tr, T1m, T1n; { E Tb, T1a, Ta, T1Q, TN, Tc, T1b, T1c; { E T8, T9, TL, TM; T8 = Rp[WS(rs, 4)]; TF = T4 - T5; T6 = T4 + T5; T26 = T1B - T1C; T1D = T1B + T1C; T9 = Rm[WS(rs, 5)]; TL = Ip[WS(rs, 4)]; TM = Im[WS(rs, 5)]; Tb = Rp[WS(rs, 9)]; T1a = T8 - T9; Ta = T8 + T9; T1Q = TL - TM; TN = TL + TM; Tc = Rm[0]; T1b = Ip[WS(rs, 9)]; T1c = Im[0]; } { E Tn, To, TW, TX; Tn = Rp[WS(rs, 8)]; { E TK, Td, T1R, T1d; TK = Tb - Tc; Td = Tb + Tc; T1R = T1b - T1c; T1d = T1b + T1c; TO = TK + TN; T47 = TN - TK; T3z = Ta - Td; Te = Ta + Td; T1S = T1Q + T1R; T3M = T1Q - T1R; T1e = T1a - T1d; T4k = T1a + T1d; To = Rm[WS(rs, 1)]; } TW = Ip[WS(rs, 8)]; TX = Im[WS(rs, 1)]; Tq = Rm[WS(rs, 6)]; T1l = Tn - To; Tp = Tn + To; T1X = TW - TX; TY = TW + TX; Tr = Rp[WS(rs, 3)]; T1m = Im[WS(rs, 6)]; T1n = Ip[WS(rs, 3)]; } } { E Tx, T1q, Tw, T20, T13, Ty, T1r, T1s; { E Tu, Tv, T11, T12; Tu = Rm[WS(rs, 7)]; { E TV, Ts, T1Y, T1o; TV = Tq - Tr; Ts = Tq + Tr; T1Y = T1n - T1m; T1o = T1m + T1n; TZ = TV + TY; T4a = TY - TV; T3C = Tp - Ts; Tt = Tp + Ts; T1Z = T1X + T1Y; T3J = T1X - T1Y; T1p = T1l + T1o; T4h = T1l - T1o; Tv = Rp[WS(rs, 2)]; } T11 = Im[WS(rs, 7)]; T12 = Ip[WS(rs, 2)]; Tx = Rm[WS(rs, 2)]; T1q = Tu - Tv; Tw = Tu + Tv; T20 = T12 - T11; T13 = T11 + T12; Ty = Rp[WS(rs, 7)]; T1r = Im[WS(rs, 2)]; T1s = Ip[WS(rs, 7)]; } { E Tf, Tg, TQ, TR; Tf = Rm[WS(rs, 3)]; { E T10, Tz, T21, T1t; T10 = Tx - Ty; Tz = Tx + Ty; T21 = T1s - T1r; T1t = T1r + T1s; T14 = T10 - T13; T4b = T10 + T13; T3D = Tw - Tz; TA = Tw + Tz; T22 = T20 + T21; T3K = T20 - T21; T1u = T1q + T1t; T4i = T1q - T1t; Tg = Rp[WS(rs, 6)]; } TQ = Im[WS(rs, 3)]; TR = Ip[WS(rs, 6)]; Ti = Rp[WS(rs, 1)]; T1f = Tf - Tg; Th = Tf + Tg; T1T = TR - TQ; TS = TQ + TR; Tj = Rm[WS(rs, 8)]; T1g = Ip[WS(rs, 1)]; T1h = Im[WS(rs, 8)]; } } } } { E T1V, T3N, TB, T3B, Tm, T3E, T1F, T1G, T4t, T4j, T4m, T4s, T4c, T4y, T4z; E T49, T3y, T7; { E TT, T48, T1j, T4l, T3A, Tl; T3T = T25 - T26; T27 = T25 + T26; { E TP, Tk, T1U, T1i; TP = Ti - Tj; Tk = Ti + Tj; T1U = T1g - T1h; T1i = T1g + T1h; TT = TP - TS; T48 = TP + TS; T3A = Th - Tk; Tl = Th + Tk; T1V = T1T + T1U; T3N = T1T - T1U; T1j = T1f - T1i; T4l = T1f + T1i; T2o = Tt - TA; TB = Tt + TA; } T41 = T3z - T3A; T3B = T3z + T3A; Tm = Te + Tl; T2p = Te - Tl; { E T1L, T1M, T1k, T1v; T40 = T3C - T3D; T3E = T3C + T3D; TU = TO + TT; T1L = TO - TT; T1M = TZ - T14; T15 = TZ + T14; T1F = T1e + T1j; T1k = T1e - T1j; T1v = T1p - T1u; T1G = T1p + T1u; T4t = T4h + T4i; T4j = T4h - T4i; T2Q = FNMS(KP618033988, T1L, T1M); T1N = FMA(KP618033988, T1M, T1L); T2L = FNMS(KP618033988, T1k, T1v); T1w = FMA(KP618033988, T1v, T1k); T4m = T4k - T4l; T4s = T4k + T4l; T4c = T4a - T4b; T4y = T4a + T4b; T4z = T47 + T48; T49 = T47 - T48; } } { E T2g, T1W, T23, T2f; T2g = T1S - T1V; T1W = T1S + T1V; T59 = FMA(KP618033988, T4j, T4m); T4n = FNMS(KP618033988, T4m, T4j); T5e = FMA(KP618033988, T4y, T4z); T4A = FNMS(KP618033988, T4z, T4y); T23 = T1Z + T22; T2f = T1Z - T22; { E T3V, T3L, T3O, T3U; T3V = T3J + T3K; T3L = T3J - T3K; T2m = T1W - T23; T24 = T1W + T23; T2Z = FMA(KP618033988, T2f, T2g); T2h = FNMS(KP618033988, T2g, T2f); T3O = T3M - T3N; T3U = T3M + T3N; T3y = T3 - T6; T7 = T3 + T6; T4J = FMA(KP618033988, T3L, T3O); T3P = FNMS(KP618033988, T3O, T3L); T3Y = T3U - T3V; T3W = T3U + T3V; } } { E T46, TC, T3F, T4r, T4d, T4u; TC = Tm + TB; T2d = Tm - TB; TJ = TF + TI; T46 = TI - TF; T3H = T3B - T3E; T3F = T3B + T3E; T2c = FNMS(KP250000000, TC, T7); TD = T7 + TC; T52 = T3y + T3F; T3G = FNMS(KP250000000, T3F, T3y); T4r = T1A + T1D; T1E = T1A - T1D; T4f = T49 - T4c; T4d = T49 + T4c; T5I = T46 + T4d; T4e = FNMS(KP250000000, T4d, T46); T4w = T4s - T4t; T4u = T4s + T4t; T5L = T4u + T4r; T4v = FNMS(KP250000000, T4u, T4r); T1J = T1F - T1G; T1H = T1F + T1G; } } } { E T38, T3b, T39, T3f, T36, T3e, T3a; { E T28, T3r, T3o, T3v, T3p, T2b, T2k, T35, T3l, T2H, T2r, T2j, T2z, T2D, T2G; E T2X, T2F, T2T, T32, T3h, T3k, T31, T3d, T3j, T3t, T1x, T2u, T1O, T2x, T2v; E T1y, T2B, T29, T2J, T2M, T2R, T2N, T2V; { E T2l, T1I, T18, T2q, T34, T17, T16, T3n; T28 = T24 + T27; T2l = FNMS(KP250000000, T24, T27); T3r = T1H + T1E; T1I = FNMS(KP250000000, T1H, T1E); T18 = TU - T15; T16 = TU + T15; T3n = W[8]; T2q = FNMS(KP618033988, T2p, T2o); T34 = FMA(KP618033988, T2o, T2p); T17 = FNMS(KP250000000, T16, TJ); T3o = TJ + T16; T3v = T3n * T3r; T3p = T3n * T3o; { E T2Y, T2E, T3i, T30; { E T2e, T33, T2n, T2i; T2Y = FMA(KP559016994, T2d, T2c); T2e = FNMS(KP559016994, T2d, T2c); T2b = W[14]; T2k = W[15]; T33 = FMA(KP559016994, T2m, T2l); T2n = FNMS(KP559016994, T2m, T2l); T2E = FMA(KP951056516, T2h, T2e); T2i = FNMS(KP951056516, T2h, T2e); T35 = FMA(KP951056516, T34, T33); T3l = FNMS(KP951056516, T34, T33); T2H = FNMS(KP951056516, T2q, T2n); T2r = FMA(KP951056516, T2q, T2n); T2j = T2b * T2i; T2z = T2k * T2i; T2D = W[22]; T2G = W[23]; } T2X = W[30]; T2F = T2D * T2E; T2T = T2G * T2E; T3i = FMA(KP951056516, T2Z, T2Y); T30 = FNMS(KP951056516, T2Z, T2Y); T32 = W[31]; T3h = W[6]; T3k = W[7]; T31 = T2X * T30; T3d = T32 * T30; T3j = T3h * T3i; T3t = T3k * T3i; } { E T2K, T2P, TE, T19, T1K, T2t, T37; T2K = FNMS(KP559016994, T18, T17); T19 = FMA(KP559016994, T18, T17); T1K = FMA(KP559016994, T1J, T1I); T2P = FNMS(KP559016994, T1J, T1I); TE = W[0]; T2t = W[16]; T1x = FMA(KP951056516, T1w, T19); T2u = FNMS(KP951056516, T1w, T19); T1O = FNMS(KP951056516, T1N, T1K); T2x = FMA(KP951056516, T1N, T1K); T2v = T2t * T2u; T1y = TE * T1x; T2B = T2t * T2x; T29 = TE * T1O; T2J = W[24]; T37 = W[32]; T2M = FMA(KP951056516, T2L, T2K); T38 = FNMS(KP951056516, T2L, T2K); T2R = FNMS(KP951056516, T2Q, T2P); T3b = FMA(KP951056516, T2Q, T2P); T39 = T37 * T38; T2N = T2J * T2M; T3f = T37 * T3b; } } T2V = T2J * T2R; { E T3m, T3u, T3q, T2a, T1P, T1z; T1z = W[1]; T3m = FNMS(T3k, T3l, T3j); T3u = FMA(T3h, T3l, T3t); T3q = W[9]; T2a = FNMS(T1z, T1x, T29); T1P = FMA(T1z, T1O, T1y); { E T2s, T2A, T2w, T3w, T3s; T2s = FNMS(T2k, T2r, T2j); T3w = FNMS(T3q, T3o, T3v); T3s = FMA(T3q, T3r, T3p); Im[0] = T2a - T28; Ip[0] = T28 + T2a; Rm[0] = TD + T1P; Rp[0] = TD - T1P; Im[WS(rs, 2)] = T3w - T3u; Ip[WS(rs, 2)] = T3u + T3w; Rm[WS(rs, 2)] = T3m + T3s; Rp[WS(rs, 2)] = T3m - T3s; T2A = FMA(T2b, T2r, T2z); T2w = W[17]; { E T2I, T2U, T2O, T2C, T2y, T2W, T2S; T2I = FNMS(T2G, T2H, T2F); T2U = FMA(T2D, T2H, T2T); T2O = W[25]; T2C = FNMS(T2w, T2u, T2B); T2y = FMA(T2w, T2x, T2v); T36 = FNMS(T32, T35, T31); T2W = FNMS(T2O, T2M, T2V); T2S = FMA(T2O, T2R, T2N); Im[WS(rs, 4)] = T2C - T2A; Ip[WS(rs, 4)] = T2A + T2C; Rm[WS(rs, 4)] = T2s + T2y; Rp[WS(rs, 4)] = T2s - T2y; Im[WS(rs, 6)] = T2W - T2U; Ip[WS(rs, 6)] = T2U + T2W; Rm[WS(rs, 6)] = T2I + T2S; Rp[WS(rs, 6)] = T2I - T2S; T3e = FMA(T2X, T35, T3d); T3a = W[33]; } } } } { E T55, T51, T54, T53, T5h, T5P, T5J, T3x, T4P, T5F, T5p, T43, T3R, T3S, T5l; E T5o, T4D, T5n, T5x, T4H, T4M, T5B, T5E, T4L, T4X, T5D, T5N, T4S, T4o, T4V; E T4B, T4T, T4p, T4Z, T4F, T57, T5a, T5f, T5b, T5j; { E T3X, T4O, T42, T3g, T3c, T5H; T55 = T3W + T3T; T3X = FNMS(KP250000000, T3W, T3T); T51 = W[18]; T3g = FNMS(T3a, T38, T3f); T3c = FMA(T3a, T3b, T39); T54 = W[19]; T53 = T51 * T52; Im[WS(rs, 8)] = T3g - T3e; Ip[WS(rs, 8)] = T3e + T3g; Rm[WS(rs, 8)] = T36 + T3c; Rp[WS(rs, 8)] = T36 - T3c; T5h = T54 * T52; T5H = W[28]; T4O = FMA(KP618033988, T40, T41); T42 = FNMS(KP618033988, T41, T40); T5P = T5H * T5L; T5J = T5H * T5I; { E T4I, T5m, T3Q, T3I, T3Z, T4N, T4K, T5C; T3I = FNMS(KP559016994, T3H, T3G); T4I = FMA(KP559016994, T3H, T3G); T3Z = FNMS(KP559016994, T3Y, T3X); T4N = FMA(KP559016994, T3Y, T3X); T3x = W[2]; T5m = FNMS(KP951056516, T3P, T3I); T3Q = FMA(KP951056516, T3P, T3I); T4P = FMA(KP951056516, T4O, T4N); T5F = FNMS(KP951056516, T4O, T4N); T5p = FMA(KP951056516, T42, T3Z); T43 = FNMS(KP951056516, T42, T3Z); T3R = T3x * T3Q; T3S = W[3]; T5l = W[34]; T5o = W[35]; T4D = T3S * T3Q; T5n = T5l * T5m; T5x = T5o * T5m; T4K = FNMS(KP951056516, T4J, T4I); T5C = FMA(KP951056516, T4J, T4I); T4H = W[10]; T4M = W[11]; T5B = W[26]; T5E = W[27]; T4L = T4H * T4K; T4X = T4M * T4K; T5D = T5B * T5C; T5N = T5E * T5C; } { E T58, T5d, T45, T4g, T4x, T4R, T5r; T4g = FNMS(KP559016994, T4f, T4e); T58 = FMA(KP559016994, T4f, T4e); T5d = FMA(KP559016994, T4w, T4v); T4x = FNMS(KP559016994, T4w, T4v); T45 = W[4]; T4R = W[12]; T4S = FNMS(KP951056516, T4n, T4g); T4o = FMA(KP951056516, T4n, T4g); T4V = FMA(KP951056516, T4A, T4x); T4B = FNMS(KP951056516, T4A, T4x); T4T = T4R * T4S; T4p = T45 * T4o; T4Z = T4R * T4V; T4F = T45 * T4B; T57 = W[20]; T5r = W[36]; T5s = FNMS(KP951056516, T59, T58); T5a = FMA(KP951056516, T59, T58); T5v = FMA(KP951056516, T5e, T5d); T5f = FNMS(KP951056516, T5e, T5d); T5t = T5r * T5s; T5b = T57 * T5a; T5z = T5r * T5v; } } T5j = T57 * T5f; { E T44, T4E, T5G, T5O, T5K, T4G, T4C, T4q; T44 = FNMS(T3S, T43, T3R); T4E = FMA(T3x, T43, T4D); T4q = W[5]; T5G = FNMS(T5E, T5F, T5D); T5O = FMA(T5B, T5F, T5N); T5K = W[29]; T4G = FNMS(T4q, T4o, T4F); T4C = FMA(T4q, T4B, T4p); { E T4Q, T4Y, T4U, T5Q, T5M; T4Q = FNMS(T4M, T4P, T4L); T5Q = FNMS(T5K, T5I, T5P); T5M = FMA(T5K, T5L, T5J); Im[WS(rs, 1)] = T4G - T4E; Ip[WS(rs, 1)] = T4E + T4G; Rm[WS(rs, 1)] = T44 + T4C; Rp[WS(rs, 1)] = T44 - T4C; Im[WS(rs, 7)] = T5Q - T5O; Ip[WS(rs, 7)] = T5O + T5Q; Rm[WS(rs, 7)] = T5G + T5M; Rp[WS(rs, 7)] = T5G - T5M; T4Y = FMA(T4H, T4P, T4X); T4U = W[13]; { E T56, T5i, T5c, T50, T4W, T5k, T5g; T56 = FNMS(T54, T55, T53); T5i = FMA(T51, T55, T5h); T5c = W[21]; T50 = FNMS(T4U, T4S, T4Z); T4W = FMA(T4U, T4V, T4T); T5q = FNMS(T5o, T5p, T5n); T5k = FNMS(T5c, T5a, T5j); T5g = FMA(T5c, T5f, T5b); Im[WS(rs, 3)] = T50 - T4Y; Ip[WS(rs, 3)] = T4Y + T50; Rm[WS(rs, 3)] = T4Q + T4W; Rp[WS(rs, 3)] = T4Q - T4W; Im[WS(rs, 5)] = T5k - T5i; Ip[WS(rs, 5)] = T5i + T5k; Rm[WS(rs, 5)] = T56 + T5g; Rp[WS(rs, 5)] = T56 - T5g; T5y = FMA(T5l, T5p, T5x); T5u = W[37]; } } } } } } T5A = FNMS(T5u, T5s, T5z); T5w = FMA(T5u, T5v, T5t); Im[WS(rs, 9)] = T5A - T5y; Ip[WS(rs, 9)] = T5y + T5A; Rm[WS(rs, 9)] = T5q + T5w; Rp[WS(rs, 9)] = T5q - T5w; } } } static const tw_instr twinstr[] = { {TW_FULL, 1, 20}, {TW_NEXT, 1, 0} }; static const hc2c_desc desc = { 20, "hc2cbdft2_20", twinstr, &GENUS, {176, 38, 110, 0} }; void X(codelet_hc2cbdft2_20) (planner *p) { X(khc2c_register) (p, hc2cbdft2_20, &desc, HC2C_VIA_DFT); } #else /* HAVE_FMA */ /* Generated by: ../../../genfft/gen_hc2cdft.native -compact -variables 4 -pipeline-latency 4 -sign 1 -n 20 -dif -name hc2cbdft2_20 -include hc2cb.h */ /* * This function contains 286 FP additions, 124 FP multiplications, * (or, 224 additions, 62 multiplications, 62 fused multiply/add), * 89 stack variables, 4 constants, and 80 memory accesses */ #include "hc2cb.h" static void hc2cbdft2_20(R *Rp, R *Ip, R *Rm, R *Im, const R *W, stride rs, INT mb, INT me, INT ms) { DK(KP250000000, +0.250000000000000000000000000000000000000000000); DK(KP951056516, +0.951056516295153572116439333379382143405698634); DK(KP587785252, +0.587785252292473129168705954639072768597652438); DK(KP559016994, +0.559016994374947424102293417182819058860154590); { INT m; for (m = mb, W = W + ((mb - 1) * 38); m < me; m = m + 1, Rp = Rp + ms, Ip = Ip + ms, Rm = Rm - ms, Im = Im - ms, W = W + 38, MAKE_VOLATILE_STRIDE(80, rs)) { E T7, T3N, T4a, T16, T1G, T3g, T3D, T26, T1k, T3A, T3B, T1v, T2e, T48, T47; E T2d, T1L, T43, T40, T1K, T2l, T3t, T2m, T3w, T3n, T3p, TC, T2b, T4d, T4f; E T23, T2j, T1B, T1H, T3U, T3W, T3G, T3I, T11, T17; { E T3, T1C, T15, T24, T6, T12, T1F, T25; { E T1, T2, T13, T14; T1 = Rp[0]; T2 = Rm[WS(rs, 9)]; T3 = T1 + T2; T1C = T1 - T2; T13 = Ip[0]; T14 = Im[WS(rs, 9)]; T15 = T13 + T14; T24 = T13 - T14; } { E T4, T5, T1D, T1E; T4 = Rp[WS(rs, 5)]; T5 = Rm[WS(rs, 4)]; T6 = T4 + T5; T12 = T4 - T5; T1D = Ip[WS(rs, 5)]; T1E = Im[WS(rs, 4)]; T1F = T1D + T1E; T25 = T1D - T1E; } T7 = T3 + T6; T3N = T15 - T12; T4a = T1C + T1F; T16 = T12 + T15; T1G = T1C - T1F; T3g = T3 - T6; T3D = T24 - T25; T26 = T24 + T25; } { E Te, T3O, T3Y, TJ, T1e, T3h, T3r, T1R, TA, T3S, T42, TZ, T1u, T3l, T3v; E T21, Tl, T3P, T3Z, TO, T1j, T3i, T3s, T1U, Tt, T3R, T41, TU, T1p, T3k; E T3u, T1Y; { E Ta, T1a, TI, T1P, Td, TF, T1d, T1Q; { E T8, T9, TG, TH; T8 = Rp[WS(rs, 4)]; T9 = Rm[WS(rs, 5)]; Ta = T8 + T9; T1a = T8 - T9; TG = Ip[WS(rs, 4)]; TH = Im[WS(rs, 5)]; TI = TG + TH; T1P = TG - TH; } { E Tb, Tc, T1b, T1c; Tb = Rp[WS(rs, 9)]; Tc = Rm[0]; Td = Tb + Tc; TF = Tb - Tc; T1b = Ip[WS(rs, 9)]; T1c = Im[0]; T1d = T1b + T1c; T1Q = T1b - T1c; } Te = Ta + Td; T3O = TI - TF; T3Y = T1a + T1d; TJ = TF + TI; T1e = T1a - T1d; T3h = Ta - Td; T3r = T1P - T1Q; T1R = T1P + T1Q; } { E Tw, T1q, TY, T1Z, Tz, TV, T1t, T20; { E Tu, Tv, TW, TX; Tu = Rm[WS(rs, 7)]; Tv = Rp[WS(rs, 2)]; Tw = Tu + Tv; T1q = Tu - Tv; TW = Im[WS(rs, 7)]; TX = Ip[WS(rs, 2)]; TY = TW + TX; T1Z = TX - TW; } { E Tx, Ty, T1r, T1s; Tx = Rm[WS(rs, 2)]; Ty = Rp[WS(rs, 7)]; Tz = Tx + Ty; TV = Tx - Ty; T1r = Im[WS(rs, 2)]; T1s = Ip[WS(rs, 7)]; T1t = T1r + T1s; T20 = T1s - T1r; } TA = Tw + Tz; T3S = TV + TY; T42 = T1q - T1t; TZ = TV - TY; T1u = T1q + T1t; T3l = Tw - Tz; T3v = T1Z - T20; T21 = T1Z + T20; } { E Th, T1f, TN, T1S, Tk, TK, T1i, T1T; { E Tf, Tg, TL, TM; Tf = Rm[WS(rs, 3)]; Tg = Rp[WS(rs, 6)]; Th = Tf + Tg; T1f = Tf - Tg; TL = Im[WS(rs, 3)]; TM = Ip[WS(rs, 6)]; TN = TL + TM; T1S = TM - TL; } { E Ti, Tj, T1g, T1h; Ti = Rp[WS(rs, 1)]; Tj = Rm[WS(rs, 8)]; Tk = Ti + Tj; TK = Ti - Tj; T1g = Ip[WS(rs, 1)]; T1h = Im[WS(rs, 8)]; T1i = T1g + T1h; T1T = T1g - T1h; } Tl = Th + Tk; T3P = TK + TN; T3Z = T1f + T1i; TO = TK - TN; T1j = T1f - T1i; T3i = Th - Tk; T3s = T1S - T1T; T1U = T1S + T1T; } { E Tp, T1l, TT, T1W, Ts, TQ, T1o, T1X; { E Tn, To, TR, TS; Tn = Rp[WS(rs, 8)]; To = Rm[WS(rs, 1)]; Tp = Tn + To; T1l = Tn - To; TR = Ip[WS(rs, 8)]; TS = Im[WS(rs, 1)]; TT = TR + TS; T1W = TR - TS; } { E Tq, Tr, T1m, T1n; Tq = Rm[WS(rs, 6)]; Tr = Rp[WS(rs, 3)]; Ts = Tq + Tr; TQ = Tq - Tr; T1m = Im[WS(rs, 6)]; T1n = Ip[WS(rs, 3)]; T1o = T1m + T1n; T1X = T1n - T1m; } Tt = Tp + Ts; T3R = TT - TQ; T41 = T1l - T1o; TU = TQ + TT; T1p = T1l + T1o; T3k = Tp - Ts; T3u = T1W - T1X; T1Y = T1W + T1X; } T1k = T1e - T1j; T3A = T3h - T3i; T3B = T3k - T3l; T1v = T1p - T1u; T2e = T1Y - T21; T48 = T3R + T3S; T47 = T3O + T3P; T2d = T1R - T1U; T1L = TU - TZ; T43 = T41 - T42; T40 = T3Y - T3Z; T1K = TJ - TO; T2l = Te - Tl; T3t = T3r - T3s; T2m = Tt - TA; T3w = T3u - T3v; { E T3j, T3m, Tm, TB; T3j = T3h + T3i; T3m = T3k + T3l; T3n = T3j + T3m; T3p = KP559016994 * (T3j - T3m); Tm = Te + Tl; TB = Tt + TA; TC = Tm + TB; T2b = KP559016994 * (Tm - TB); } { E T4b, T4c, T3Q, T3T; T4b = T3Y + T3Z; T4c = T41 + T42; T4d = T4b + T4c; T4f = KP559016994 * (T4b - T4c); { E T1V, T22, T1z, T1A; T1V = T1R + T1U; T22 = T1Y + T21; T23 = T1V + T22; T2j = KP559016994 * (T1V - T22); T1z = T1e + T1j; T1A = T1p + T1u; T1B = KP559016994 * (T1z - T1A); T1H = T1z + T1A; } T3Q = T3O - T3P; T3T = T3R - T3S; T3U = T3Q + T3T; T3W = KP559016994 * (T3Q - T3T); { E T3E, T3F, TP, T10; T3E = T3r + T3s; T3F = T3u + T3v; T3G = T3E + T3F; T3I = KP559016994 * (T3E - T3F); TP = TJ + TO; T10 = TU + TZ; T11 = KP559016994 * (TP - T10); T17 = TP + T10; } } } { E TD, T27, T3c, T3e, T2o, T36, T2A, T2U, T1N, T2Z, T2t, T2J, T1x, T2X, T2r; E T2F, T2g, T34, T2y, T2Q; TD = T7 + TC; T27 = T23 + T26; { E T39, T3b, T38, T3a; T39 = T16 + T17; T3b = T1H + T1G; T38 = W[8]; T3a = W[9]; T3c = FMA(T38, T39, T3a * T3b); T3e = FNMS(T3a, T39, T38 * T3b); } { E T2n, T2S, T2k, T2T, T2i; T2n = FNMS(KP951056516, T2m, KP587785252 * T2l); T2S = FMA(KP951056516, T2l, KP587785252 * T2m); T2i = FNMS(KP250000000, T23, T26); T2k = T2i - T2j; T2T = T2j + T2i; T2o = T2k - T2n; T36 = T2T - T2S; T2A = T2n + T2k; T2U = T2S + T2T; } { E T1M, T2H, T1J, T2I, T1I; T1M = FMA(KP951056516, T1K, KP587785252 * T1L); T2H = FNMS(KP951056516, T1L, KP587785252 * T1K); T1I = FNMS(KP250000000, T1H, T1G); T1J = T1B + T1I; T2I = T1I - T1B; T1N = T1J - T1M; T2Z = T2I - T2H; T2t = T1M + T1J; T2J = T2H + T2I; } { E T1w, T2E, T19, T2D, T18; T1w = FMA(KP951056516, T1k, KP587785252 * T1v); T2E = FNMS(KP951056516, T1v, KP587785252 * T1k); T18 = FNMS(KP250000000, T17, T16); T19 = T11 + T18; T2D = T18 - T11; T1x = T19 + T1w; T2X = T2D + T2E; T2r = T19 - T1w; T2F = T2D - T2E; } { E T2f, T2P, T2c, T2O, T2a; T2f = FNMS(KP951056516, T2e, KP587785252 * T2d); T2P = FMA(KP951056516, T2d, KP587785252 * T2e); T2a = FNMS(KP250000000, TC, T7); T2c = T2a - T2b; T2O = T2b + T2a; T2g = T2c + T2f; T34 = T2O + T2P; T2y = T2c - T2f; T2Q = T2O - T2P; } { E T1O, T28, TE, T1y; TE = W[0]; T1y = W[1]; T1O = FMA(TE, T1x, T1y * T1N); T28 = FNMS(T1y, T1x, TE * T1N); Rp[0] = TD - T1O; Ip[0] = T27 + T28; Rm[0] = TD + T1O; Im[0] = T28 - T27; } { E T37, T3d, T33, T35; T33 = W[6]; T35 = W[7]; T37 = FNMS(T35, T36, T33 * T34); T3d = FMA(T35, T34, T33 * T36); Rp[WS(rs, 2)] = T37 - T3c; Ip[WS(rs, 2)] = T3d + T3e; Rm[WS(rs, 2)] = T37 + T3c; Im[WS(rs, 2)] = T3e - T3d; } { E T2p, T2v, T2u, T2w; { E T29, T2h, T2q, T2s; T29 = W[14]; T2h = W[15]; T2p = FNMS(T2h, T2o, T29 * T2g); T2v = FMA(T2h, T2g, T29 * T2o); T2q = W[16]; T2s = W[17]; T2u = FMA(T2q, T2r, T2s * T2t); T2w = FNMS(T2s, T2r, T2q * T2t); } Rp[WS(rs, 4)] = T2p - T2u; Ip[WS(rs, 4)] = T2v + T2w; Rm[WS(rs, 4)] = T2p + T2u; Im[WS(rs, 4)] = T2w - T2v; } { E T2B, T2L, T2K, T2M; { E T2x, T2z, T2C, T2G; T2x = W[22]; T2z = W[23]; T2B = FNMS(T2z, T2A, T2x * T2y); T2L = FMA(T2z, T2y, T2x * T2A); T2C = W[24]; T2G = W[25]; T2K = FMA(T2C, T2F, T2G * T2J); T2M = FNMS(T2G, T2F, T2C * T2J); } Rp[WS(rs, 6)] = T2B - T2K; Ip[WS(rs, 6)] = T2L + T2M; Rm[WS(rs, 6)] = T2B + T2K; Im[WS(rs, 6)] = T2M - T2L; } { E T2V, T31, T30, T32; { E T2N, T2R, T2W, T2Y; T2N = W[30]; T2R = W[31]; T2V = FNMS(T2R, T2U, T2N * T2Q); T31 = FMA(T2R, T2Q, T2N * T2U); T2W = W[32]; T2Y = W[33]; T30 = FMA(T2W, T2X, T2Y * T2Z); T32 = FNMS(T2Y, T2X, T2W * T2Z); } Rp[WS(rs, 8)] = T2V - T30; Ip[WS(rs, 8)] = T31 + T32; Rm[WS(rs, 8)] = T2V + T30; Im[WS(rs, 8)] = T32 - T31; } } { E T4F, T4P, T5c, T5e, T3y, T54, T4o, T4S, T4h, T4Z, T4x, T4N, T45, T4X, T4v; E T4J, T3K, T56, T4s, T4U; { E T4C, T4E, T4B, T4D; T4C = T3g + T3n; T4E = T3G + T3D; T4B = W[18]; T4D = W[19]; T4F = FNMS(T4D, T4E, T4B * T4C); T4P = FMA(T4D, T4C, T4B * T4E); } { E T59, T5b, T58, T5a; T59 = T3N + T3U; T5b = T4d + T4a; T58 = W[28]; T5a = W[29]; T5c = FMA(T58, T59, T5a * T5b); T5e = FNMS(T5a, T59, T58 * T5b); } { E T3x, T4n, T3q, T4m, T3o; T3x = FNMS(KP951056516, T3w, KP587785252 * T3t); T4n = FMA(KP951056516, T3t, KP587785252 * T3w); T3o = FNMS(KP250000000, T3n, T3g); T3q = T3o - T3p; T4m = T3p + T3o; T3y = T3q - T3x; T54 = T4m + T4n; T4o = T4m - T4n; T4S = T3q + T3x; } { E T49, T4M, T4g, T4L, T4e; T49 = FNMS(KP951056516, T48, KP587785252 * T47); T4M = FMA(KP951056516, T47, KP587785252 * T48); T4e = FNMS(KP250000000, T4d, T4a); T4g = T4e - T4f; T4L = T4f + T4e; T4h = T49 + T4g; T4Z = T4M + T4L; T4x = T4g - T49; T4N = T4L - T4M; } { E T44, T4I, T3X, T4H, T3V; T44 = FNMS(KP951056516, T43, KP587785252 * T40); T4I = FMA(KP951056516, T40, KP587785252 * T43); T3V = FNMS(KP250000000, T3U, T3N); T3X = T3V - T3W; T4H = T3W + T3V; T45 = T3X - T44; T4X = T4H - T4I; T4v = T3X + T44; T4J = T4H + T4I; } { E T3C, T4q, T3J, T4r, T3H; T3C = FNMS(KP951056516, T3B, KP587785252 * T3A); T4q = FMA(KP951056516, T3A, KP587785252 * T3B); T3H = FNMS(KP250000000, T3G, T3D); T3J = T3H - T3I; T4r = T3I + T3H; T3K = T3C + T3J; T56 = T4r - T4q; T4s = T4q + T4r; T4U = T3J - T3C; } { E T4O, T4Q, T4G, T4K; T4G = W[20]; T4K = W[21]; T4O = FMA(T4G, T4J, T4K * T4N); T4Q = FNMS(T4K, T4J, T4G * T4N); Rp[WS(rs, 5)] = T4F - T4O; Ip[WS(rs, 5)] = T4P + T4Q; Rm[WS(rs, 5)] = T4F + T4O; Im[WS(rs, 5)] = T4Q - T4P; } { E T57, T5d, T53, T55; T53 = W[26]; T55 = W[27]; T57 = FNMS(T55, T56, T53 * T54); T5d = FMA(T55, T54, T53 * T56); Rp[WS(rs, 7)] = T57 - T5c; Ip[WS(rs, 7)] = T5d + T5e; Rm[WS(rs, 7)] = T57 + T5c; Im[WS(rs, 7)] = T5e - T5d; } { E T3L, T4j, T4i, T4k; { E T3f, T3z, T3M, T46; T3f = W[2]; T3z = W[3]; T3L = FNMS(T3z, T3K, T3f * T3y); T4j = FMA(T3z, T3y, T3f * T3K); T3M = W[4]; T46 = W[5]; T4i = FMA(T3M, T45, T46 * T4h); T4k = FNMS(T46, T45, T3M * T4h); } Rp[WS(rs, 1)] = T3L - T4i; Ip[WS(rs, 1)] = T4j + T4k; Rm[WS(rs, 1)] = T3L + T4i; Im[WS(rs, 1)] = T4k - T4j; } { E T4t, T4z, T4y, T4A; { E T4l, T4p, T4u, T4w; T4l = W[10]; T4p = W[11]; T4t = FNMS(T4p, T4s, T4l * T4o); T4z = FMA(T4p, T4o, T4l * T4s); T4u = W[12]; T4w = W[13]; T4y = FMA(T4u, T4v, T4w * T4x); T4A = FNMS(T4w, T4v, T4u * T4x); } Rp[WS(rs, 3)] = T4t - T4y; Ip[WS(rs, 3)] = T4z + T4A; Rm[WS(rs, 3)] = T4t + T4y; Im[WS(rs, 3)] = T4A - T4z; } { E T4V, T51, T50, T52; { E T4R, T4T, T4W, T4Y; T4R = W[34]; T4T = W[35]; T4V = FNMS(T4T, T4U, T4R * T4S); T51 = FMA(T4T, T4S, T4R * T4U); T4W = W[36]; T4Y = W[37]; T50 = FMA(T4W, T4X, T4Y * T4Z); T52 = FNMS(T4Y, T4X, T4W * T4Z); } Rp[WS(rs, 9)] = T4V - T50; Ip[WS(rs, 9)] = T51 + T52; Rm[WS(rs, 9)] = T4V + T50; Im[WS(rs, 9)] = T52 - T51; } } } } } static const tw_instr twinstr[] = { {TW_FULL, 1, 20}, {TW_NEXT, 1, 0} }; static const hc2c_desc desc = { 20, "hc2cbdft2_20", twinstr, &GENUS, {224, 62, 62, 0} }; void X(codelet_hc2cbdft2_20) (planner *p) { X(khc2c_register) (p, hc2cbdft2_20, &desc, HC2C_VIA_DFT); } #endif /* HAVE_FMA */ fftw-3.3.3/rdft/scalar/r2cb/codlist.c0000644000175000017500000001513612054411043014250 00000000000000#include "ifftw.h" extern void X(codelet_r2cb_2)(planner *); extern void X(codelet_r2cb_3)(planner *); extern void X(codelet_r2cb_4)(planner *); extern void X(codelet_r2cb_5)(planner *); extern void X(codelet_r2cb_6)(planner *); extern void X(codelet_r2cb_7)(planner *); extern void X(codelet_r2cb_8)(planner *); extern void X(codelet_r2cb_9)(planner *); extern void X(codelet_r2cb_10)(planner *); extern void X(codelet_r2cb_11)(planner *); extern void X(codelet_r2cb_12)(planner *); extern void X(codelet_r2cb_13)(planner *); extern void X(codelet_r2cb_14)(planner *); extern void X(codelet_r2cb_15)(planner *); extern void X(codelet_r2cb_16)(planner *); extern void X(codelet_r2cb_32)(planner *); extern void X(codelet_r2cb_64)(planner *); extern void X(codelet_r2cb_128)(planner *); extern void X(codelet_r2cb_20)(planner *); extern void X(codelet_r2cb_25)(planner *); extern void X(codelet_hb_2)(planner *); extern void X(codelet_hb_3)(planner *); extern void X(codelet_hb_4)(planner *); extern void X(codelet_hb_5)(planner *); extern void X(codelet_hb_6)(planner *); extern void X(codelet_hb_7)(planner *); extern void X(codelet_hb_8)(planner *); extern void X(codelet_hb_9)(planner *); extern void X(codelet_hb_10)(planner *); extern void X(codelet_hb_12)(planner *); extern void X(codelet_hb_15)(planner *); extern void X(codelet_hb_16)(planner *); extern void X(codelet_hb_32)(planner *); extern void X(codelet_hb_64)(planner *); extern void X(codelet_hb_20)(planner *); extern void X(codelet_hb_25)(planner *); extern void X(codelet_hb2_4)(planner *); extern void X(codelet_hb2_8)(planner *); extern void X(codelet_hb2_16)(planner *); extern void X(codelet_hb2_32)(planner *); extern void X(codelet_hb2_5)(planner *); extern void X(codelet_hb2_20)(planner *); extern void X(codelet_hb2_25)(planner *); extern void X(codelet_r2cbIII_2)(planner *); extern void X(codelet_r2cbIII_3)(planner *); extern void X(codelet_r2cbIII_4)(planner *); extern void X(codelet_r2cbIII_5)(planner *); extern void X(codelet_r2cbIII_6)(planner *); extern void X(codelet_r2cbIII_7)(planner *); extern void X(codelet_r2cbIII_8)(planner *); extern void X(codelet_r2cbIII_9)(planner *); extern void X(codelet_r2cbIII_10)(planner *); extern void X(codelet_r2cbIII_12)(planner *); extern void X(codelet_r2cbIII_15)(planner *); extern void X(codelet_r2cbIII_16)(planner *); extern void X(codelet_r2cbIII_32)(planner *); extern void X(codelet_r2cbIII_64)(planner *); extern void X(codelet_r2cbIII_20)(planner *); extern void X(codelet_r2cbIII_25)(planner *); extern void X(codelet_hc2cb_2)(planner *); extern void X(codelet_hc2cb_4)(planner *); extern void X(codelet_hc2cb_6)(planner *); extern void X(codelet_hc2cb_8)(planner *); extern void X(codelet_hc2cb_10)(planner *); extern void X(codelet_hc2cb_12)(planner *); extern void X(codelet_hc2cb_16)(planner *); extern void X(codelet_hc2cb_32)(planner *); extern void X(codelet_hc2cb_20)(planner *); extern void X(codelet_hc2cb2_4)(planner *); extern void X(codelet_hc2cb2_8)(planner *); extern void X(codelet_hc2cb2_16)(planner *); extern void X(codelet_hc2cb2_32)(planner *); extern void X(codelet_hc2cb2_20)(planner *); extern void X(codelet_hc2cbdft_2)(planner *); extern void X(codelet_hc2cbdft_4)(planner *); extern void X(codelet_hc2cbdft_6)(planner *); extern void X(codelet_hc2cbdft_8)(planner *); extern void X(codelet_hc2cbdft_10)(planner *); extern void X(codelet_hc2cbdft_12)(planner *); extern void X(codelet_hc2cbdft_16)(planner *); extern void X(codelet_hc2cbdft_32)(planner *); extern void X(codelet_hc2cbdft_20)(planner *); extern void X(codelet_hc2cbdft2_4)(planner *); extern void X(codelet_hc2cbdft2_8)(planner *); extern void X(codelet_hc2cbdft2_16)(planner *); extern void X(codelet_hc2cbdft2_32)(planner *); extern void X(codelet_hc2cbdft2_20)(planner *); extern const solvtab X(solvtab_rdft_r2cb); const solvtab X(solvtab_rdft_r2cb) = { SOLVTAB(X(codelet_r2cb_2)), SOLVTAB(X(codelet_r2cb_3)), SOLVTAB(X(codelet_r2cb_4)), SOLVTAB(X(codelet_r2cb_5)), SOLVTAB(X(codelet_r2cb_6)), SOLVTAB(X(codelet_r2cb_7)), SOLVTAB(X(codelet_r2cb_8)), SOLVTAB(X(codelet_r2cb_9)), SOLVTAB(X(codelet_r2cb_10)), SOLVTAB(X(codelet_r2cb_11)), SOLVTAB(X(codelet_r2cb_12)), SOLVTAB(X(codelet_r2cb_13)), SOLVTAB(X(codelet_r2cb_14)), SOLVTAB(X(codelet_r2cb_15)), SOLVTAB(X(codelet_r2cb_16)), SOLVTAB(X(codelet_r2cb_32)), SOLVTAB(X(codelet_r2cb_64)), SOLVTAB(X(codelet_r2cb_128)), SOLVTAB(X(codelet_r2cb_20)), SOLVTAB(X(codelet_r2cb_25)), SOLVTAB(X(codelet_hb_2)), SOLVTAB(X(codelet_hb_3)), SOLVTAB(X(codelet_hb_4)), SOLVTAB(X(codelet_hb_5)), SOLVTAB(X(codelet_hb_6)), SOLVTAB(X(codelet_hb_7)), SOLVTAB(X(codelet_hb_8)), SOLVTAB(X(codelet_hb_9)), SOLVTAB(X(codelet_hb_10)), SOLVTAB(X(codelet_hb_12)), SOLVTAB(X(codelet_hb_15)), SOLVTAB(X(codelet_hb_16)), SOLVTAB(X(codelet_hb_32)), SOLVTAB(X(codelet_hb_64)), SOLVTAB(X(codelet_hb_20)), SOLVTAB(X(codelet_hb_25)), SOLVTAB(X(codelet_hb2_4)), SOLVTAB(X(codelet_hb2_8)), SOLVTAB(X(codelet_hb2_16)), SOLVTAB(X(codelet_hb2_32)), SOLVTAB(X(codelet_hb2_5)), SOLVTAB(X(codelet_hb2_20)), SOLVTAB(X(codelet_hb2_25)), SOLVTAB(X(codelet_r2cbIII_2)), SOLVTAB(X(codelet_r2cbIII_3)), SOLVTAB(X(codelet_r2cbIII_4)), SOLVTAB(X(codelet_r2cbIII_5)), SOLVTAB(X(codelet_r2cbIII_6)), SOLVTAB(X(codelet_r2cbIII_7)), SOLVTAB(X(codelet_r2cbIII_8)), SOLVTAB(X(codelet_r2cbIII_9)), SOLVTAB(X(codelet_r2cbIII_10)), SOLVTAB(X(codelet_r2cbIII_12)), SOLVTAB(X(codelet_r2cbIII_15)), SOLVTAB(X(codelet_r2cbIII_16)), SOLVTAB(X(codelet_r2cbIII_32)), SOLVTAB(X(codelet_r2cbIII_64)), SOLVTAB(X(codelet_r2cbIII_20)), SOLVTAB(X(codelet_r2cbIII_25)), SOLVTAB(X(codelet_hc2cb_2)), SOLVTAB(X(codelet_hc2cb_4)), SOLVTAB(X(codelet_hc2cb_6)), SOLVTAB(X(codelet_hc2cb_8)), SOLVTAB(X(codelet_hc2cb_10)), SOLVTAB(X(codelet_hc2cb_12)), SOLVTAB(X(codelet_hc2cb_16)), SOLVTAB(X(codelet_hc2cb_32)), SOLVTAB(X(codelet_hc2cb_20)), SOLVTAB(X(codelet_hc2cb2_4)), SOLVTAB(X(codelet_hc2cb2_8)), SOLVTAB(X(codelet_hc2cb2_16)), SOLVTAB(X(codelet_hc2cb2_32)), SOLVTAB(X(codelet_hc2cb2_20)), SOLVTAB(X(codelet_hc2cbdft_2)), SOLVTAB(X(codelet_hc2cbdft_4)), SOLVTAB(X(codelet_hc2cbdft_6)), SOLVTAB(X(codelet_hc2cbdft_8)), SOLVTAB(X(codelet_hc2cbdft_10)), SOLVTAB(X(codelet_hc2cbdft_12)), SOLVTAB(X(codelet_hc2cbdft_16)), SOLVTAB(X(codelet_hc2cbdft_32)), SOLVTAB(X(codelet_hc2cbdft_20)), SOLVTAB(X(codelet_hc2cbdft2_4)), SOLVTAB(X(codelet_hc2cbdft2_8)), SOLVTAB(X(codelet_hc2cbdft2_16)), SOLVTAB(X(codelet_hc2cbdft2_32)), SOLVTAB(X(codelet_hc2cbdft2_20)), SOLVTAB_END }; fftw-3.3.3/rdft/scalar/r2r/0000755000175000017500000000000012054411110012365 500000000000000fftw-3.3.3/rdft/scalar/r2r/Makefile.am0000644000175000017500000000775412054410050014360 00000000000000# This Makefile.am specifies a set of codelets, efficient transforms # of small sizes, that are used as building blocks (kernels) by FFTW # to build up large transforms, as well as the options for generating # and compiling them. # You can customize FFTW for special needs, e.g. to handle certain # sizes more efficiently, by adding new codelets to the lists of those # included by default. If you change the list of codelets, any new # ones you added will be automatically generated when you run the # bootstrap script (see "Generating your own code" in the FFTW # manual). ########################################################################### AM_CPPFLAGS = -I$(top_srcdir)/kernel -I$(top_srcdir)/rdft \ -I$(top_srcdir)/rdft/scalar noinst_LTLIBRARIES = librdft_scalar_r2r.la ########################################################################### # The following lines specify the REDFT/RODFT/DHT sizes for which to generate # specialized codelets. Currently, only REDFT01/10 of size 8 (used in JPEG). # e_ is a hard-coded REDFT FFT (DCT) of size E00 = # e00_2.c e00_3.c e00_4.c e00_5.c e00_6.c e00_7.c e00_8.c E01 = e01_8.c # e01_2.c e01_3.c e01_4.c e01_5.c e01_6.c e01_7.c E10 = e10_8.c # e10_2.c e10_3.c e10_4.c e10_5.c e10_6.c e10_7.c E11 = # e11_2.c e11_3.c e11_4.c e11_5.c e11_6.c e11_7.c e11_8.c # o_ is a hard-coded RODFT FFT (DST) of size O00 = # o00_2.c o00_3.c o00_4.c o00_5.c o00_6.c o00_7.c o00_8.c O01 = # o01_2.c o01_3.c o01_4.c o01_5.c o01_6.c o01_7.c o01_8.c O10 = # o10_2.c o10_3.c o10_4.c o10_5.c o10_6.c o10_7.c o10_8.c O11 = # o11_2.c o11_3.c o11_4.c o11_5.c o11_6.c o11_7.c o11_8.c # dht_ is a hard-coded DHT of size DHT = # dht_2.c dht_3.c dht_4.c dht_5.c dht_6.c dht_7.c dht_8.c ########################################################################### ALL_CODELETS = $(E00) $(E01) $(E10) $(E11) $(O00) $(O01) $(O10) $(O11) $(DHT) BUILT_SOURCES= $(ALL_CODELETS) $(CODLIST) librdft_scalar_r2r_la_SOURCES = $(BUILT_SOURCES) SOLVTAB_NAME = X(solvtab_rdft_r2r) XRENAME=X # special rules for regenerating codelets. include $(top_srcdir)/support/Makefile.codelets if MAINTAINER_MODE FLAGS_E00=$(RDFT_FLAGS_COMMON) FLAGS_E01=$(RDFT_FLAGS_COMMON) FLAGS_E10=$(RDFT_FLAGS_COMMON) FLAGS_E11=$(RDFT_FLAGS_COMMON) FLAGS_O00=$(RDFT_FLAGS_COMMON) FLAGS_O01=$(RDFT_FLAGS_COMMON) FLAGS_O10=$(RDFT_FLAGS_COMMON) FLAGS_O11=$(RDFT_FLAGS_COMMON) FLAGS_DHT=$(RDFT_FLAGS_COMMON) e00_%.c: $(CODELET_DEPS) $(GEN_R2R) ($(PRELUDE_COMMANDS_RDFT); $(TWOVERS) $(GEN_R2R) $(FLAGS_E00) -redft00 -n $* -name e00_$* -include "r2r.h") | $(ADD_DATE) | $(INDENT) >$@ e01_%.c: $(CODELET_DEPS) $(GEN_R2R) ($(PRELUDE_COMMANDS_RDFT); $(TWOVERS) $(GEN_R2R) $(FLAGS_E01) -redft01 -n $* -name e01_$* -include "r2r.h") | $(ADD_DATE) | $(INDENT) >$@ e10_%.c: $(CODELET_DEPS) $(GEN_R2R) ($(PRELUDE_COMMANDS_RDFT); $(TWOVERS) $(GEN_R2R) $(FLAGS_E10) -redft10 -n $* -name e10_$* -include "r2r.h") | $(ADD_DATE) | $(INDENT) >$@ e11_%.c: $(CODELET_DEPS) $(GEN_R2R) ($(PRELUDE_COMMANDS_RDFT); $(TWOVERS) $(GEN_R2R) $(FLAGS_E11) -redft11 -n $* -name e11_$* -include "r2r.h") | $(ADD_DATE) | $(INDENT) >$@ o00_%.c: $(CODELET_DEPS) $(GEN_R2R) ($(PRELUDE_COMMANDS_RDFT); $(TWOVERS) $(GEN_R2R) $(FLAGS_O00) -rodft00 -n $* -name o00_$* -include "r2r.h") | $(ADD_DATE) | $(INDENT) >$@ o01_%.c: $(CODELET_DEPS) $(GEN_R2R) ($(PRELUDE_COMMANDS_RDFT); $(TWOVERS) $(GEN_R2R) $(FLAGS_O01) -rodft01 -n $* -name o01_$* -include "r2r.h") | $(ADD_DATE) | $(INDENT) >$@ o10_%.c: $(CODELET_DEPS) $(GEN_R2R) ($(PRELUDE_COMMANDS_RDFT); $(TWOVERS) $(GEN_R2R) $(FLAGS_O10) -rodft10 -n $* -name o10_$* -include "r2r.h") | $(ADD_DATE) | $(INDENT) >$@ o11_%.c: $(CODELET_DEPS) $(GEN_R2R) ($(PRELUDE_COMMANDS_RDFT); $(TWOVERS) $(GEN_R2R) $(FLAGS_O11) -rodft11 -n $* -name o11_$* -include "r2r.h") | $(ADD_DATE) | $(INDENT) >$@ dht_%.c: $(CODELET_DEPS) $(GEN_R2R) ($(PRELUDE_COMMANDS_RDFT); $(TWOVERS) $(GEN_R2R) $(FLAGS_DHT) -dht -sign 1 -n $* -name dht_$* -include "r2r.h") | $(ADD_DATE) | $(INDENT) >$@ endif # MAINTAINER_MODE fftw-3.3.3/rdft/scalar/r2r/Makefile.in0000644000175000017500000006016412054410200014360 00000000000000# Makefile.in generated by automake 1.11.6 from Makefile.am. # @configure_input@ # Copyright (C) 1994, 1995, 1996, 1997, 1998, 1999, 2000, 2001, 2002, # 2003, 2004, 2005, 2006, 2007, 2008, 2009, 2010, 2011 Free Software # Foundation, Inc. # This Makefile.in is free software; the Free Software Foundation # gives unlimited permission to copy and/or distribute it, # with or without modifications, as long as this notice is preserved. # This program is distributed in the hope that it will be useful, # but WITHOUT ANY WARRANTY, to the extent permitted by law; without # even the implied warranty of MERCHANTABILITY or FITNESS FOR A # PARTICULAR PURPOSE. @SET_MAKE@ # This Makefile.am specifies a set of codelets, efficient transforms # of small sizes, that are used as building blocks (kernels) by FFTW # to build up large transforms, as well as the options for generating # and compiling them. # You can customize FFTW for special needs, e.g. to handle certain # sizes more efficiently, by adding new codelets to the lists of those # included by default. If you change the list of codelets, any new # ones you added will be automatically generated when you run the # bootstrap script (see "Generating your own code" in the FFTW # manual). # -*- makefile -*- # This file contains special make rules to generate codelets. # Most of this file requires GNU make . VPATH = @srcdir@ am__make_dryrun = \ { \ am__dry=no; \ case $$MAKEFLAGS in \ *\\[\ \ ]*) \ echo 'am--echo: ; @echo "AM" OK' | $(MAKE) -f - 2>/dev/null \ | grep '^AM OK$$' >/dev/null || am__dry=yes;; \ *) \ for am__flg in $$MAKEFLAGS; do \ case $$am__flg in \ *=*|--*) ;; \ *n*) am__dry=yes; break;; \ esac; \ done;; \ esac; \ test $$am__dry = yes; \ } pkgdatadir = $(datadir)/@PACKAGE@ pkgincludedir = $(includedir)/@PACKAGE@ pkglibdir = $(libdir)/@PACKAGE@ pkglibexecdir = $(libexecdir)/@PACKAGE@ am__cd = CDPATH="$${ZSH_VERSION+.}$(PATH_SEPARATOR)" && cd install_sh_DATA = $(install_sh) -c -m 644 install_sh_PROGRAM = $(install_sh) -c install_sh_SCRIPT = $(install_sh) -c INSTALL_HEADER = $(INSTALL_DATA) transform = $(program_transform_name) NORMAL_INSTALL = : PRE_INSTALL = : POST_INSTALL = : NORMAL_UNINSTALL = : PRE_UNINSTALL = : POST_UNINSTALL = : build_triplet = @build@ host_triplet = @host@ DIST_COMMON = $(srcdir)/Makefile.am $(srcdir)/Makefile.in \ $(top_srcdir)/support/Makefile.codelets subdir = rdft/scalar/r2r ACLOCAL_M4 = $(top_srcdir)/aclocal.m4 am__aclocal_m4_deps = $(top_srcdir)/m4/acx_mpi.m4 \ $(top_srcdir)/m4/acx_pthread.m4 \ $(top_srcdir)/m4/ax_cc_maxopt.m4 \ $(top_srcdir)/m4/ax_check_compiler_flags.m4 \ $(top_srcdir)/m4/ax_compiler_vendor.m4 \ $(top_srcdir)/m4/ax_gcc_aligns_stack.m4 \ $(top_srcdir)/m4/ax_gcc_version.m4 \ $(top_srcdir)/m4/ax_openmp.m4 $(top_srcdir)/m4/libtool.m4 \ $(top_srcdir)/m4/ltoptions.m4 $(top_srcdir)/m4/ltsugar.m4 \ $(top_srcdir)/m4/ltversion.m4 $(top_srcdir)/m4/lt~obsolete.m4 \ $(top_srcdir)/configure.ac am__configure_deps = $(am__aclocal_m4_deps) $(CONFIGURE_DEPENDENCIES) \ $(ACLOCAL_M4) mkinstalldirs = $(install_sh) -d CONFIG_HEADER = $(top_builddir)/config.h CONFIG_CLEAN_FILES = CONFIG_CLEAN_VPATH_FILES = LTLIBRARIES = $(noinst_LTLIBRARIES) librdft_scalar_r2r_la_LIBADD = am__objects_1 = am__objects_2 = e01_8.lo am__objects_3 = e10_8.lo am__objects_4 = $(am__objects_1) $(am__objects_2) $(am__objects_3) \ $(am__objects_1) $(am__objects_1) $(am__objects_1) \ $(am__objects_1) $(am__objects_1) $(am__objects_1) am__objects_5 = codlist.lo am__objects_6 = $(am__objects_4) $(am__objects_5) am_librdft_scalar_r2r_la_OBJECTS = $(am__objects_6) librdft_scalar_r2r_la_OBJECTS = $(am_librdft_scalar_r2r_la_OBJECTS) DEFAULT_INCLUDES = -I.@am__isrc@ -I$(top_builddir) depcomp = $(SHELL) $(top_srcdir)/depcomp am__depfiles_maybe = depfiles am__mv = mv -f COMPILE = $(CC) $(DEFS) $(DEFAULT_INCLUDES) $(INCLUDES) $(AM_CPPFLAGS) \ $(CPPFLAGS) $(AM_CFLAGS) $(CFLAGS) LTCOMPILE = $(LIBTOOL) --tag=CC $(AM_LIBTOOLFLAGS) $(LIBTOOLFLAGS) \ --mode=compile $(CC) $(DEFS) $(DEFAULT_INCLUDES) $(INCLUDES) \ $(AM_CPPFLAGS) $(CPPFLAGS) $(AM_CFLAGS) $(CFLAGS) CCLD = $(CC) LINK = $(LIBTOOL) --tag=CC $(AM_LIBTOOLFLAGS) $(LIBTOOLFLAGS) \ --mode=link $(CCLD) $(AM_CFLAGS) $(CFLAGS) $(AM_LDFLAGS) \ $(LDFLAGS) -o $@ SOURCES = $(librdft_scalar_r2r_la_SOURCES) DIST_SOURCES = $(librdft_scalar_r2r_la_SOURCES) am__can_run_installinfo = \ case $$AM_UPDATE_INFO_DIR in \ n|no|NO) false;; \ *) (install-info --version) >/dev/null 2>&1;; \ esac ETAGS = etags CTAGS = ctags DISTFILES = $(DIST_COMMON) $(DIST_SOURCES) $(TEXINFOS) $(EXTRA_DIST) ACLOCAL = @ACLOCAL@ ALLOCA = @ALLOCA@ ALTIVEC_CFLAGS = @ALTIVEC_CFLAGS@ AMTAR = @AMTAR@ AR = @AR@ AS = @AS@ AUTOCONF = @AUTOCONF@ AUTOHEADER = @AUTOHEADER@ AUTOMAKE = @AUTOMAKE@ AVX_CFLAGS = @AVX_CFLAGS@ AWK = @AWK@ CC = @CC@ CCDEPMODE = @CCDEPMODE@ CFLAGS = @CFLAGS@ CHECK_PL_OPTS = @CHECK_PL_OPTS@ CPP = @CPP@ CPPFLAGS = @CPPFLAGS@ CYGPATH_W = @CYGPATH_W@ C_FFTW_R2R_KIND = @C_FFTW_R2R_KIND@ C_MPI_FINT = @C_MPI_FINT@ DEFS = @DEFS@ DEPDIR = @DEPDIR@ DLLTOOL = @DLLTOOL@ DSYMUTIL = @DSYMUTIL@ DUMPBIN = @DUMPBIN@ ECHO_C = @ECHO_C@ ECHO_N = @ECHO_N@ ECHO_T = @ECHO_T@ EGREP = @EGREP@ EXEEXT = @EXEEXT@ F77 = @F77@ FFLAGS = @FFLAGS@ FGREP = @FGREP@ FLIBS = @FLIBS@ GREP = @GREP@ INSTALL = @INSTALL@ INSTALL_DATA = @INSTALL_DATA@ INSTALL_PROGRAM = @INSTALL_PROGRAM@ INSTALL_SCRIPT = @INSTALL_SCRIPT@ INSTALL_STRIP_PROGRAM = @INSTALL_STRIP_PROGRAM@ LD = @LD@ LDFLAGS = @LDFLAGS@ LIBOBJS = @LIBOBJS@ LIBQUADMATH = @LIBQUADMATH@ LIBS = @LIBS@ LIBTOOL = @LIBTOOL@ LIPO = @LIPO@ LN_S = @LN_S@ LTLIBOBJS = @LTLIBOBJS@ MAINT = @MAINT@ MAKEINFO = @MAKEINFO@ MANIFEST_TOOL = @MANIFEST_TOOL@ MKDIR_P = @MKDIR_P@ MPICC = @MPICC@ MPILIBS = @MPILIBS@ MPIRUN = @MPIRUN@ NEON_CFLAGS = @NEON_CFLAGS@ NM = @NM@ NMEDIT = @NMEDIT@ OBJDUMP = @OBJDUMP@ OBJEXT = @OBJEXT@ OCAMLBUILD = @OCAMLBUILD@ OPENMP_CFLAGS = @OPENMP_CFLAGS@ OTOOL = @OTOOL@ OTOOL64 = @OTOOL64@ PACKAGE = @PACKAGE@ PACKAGE_BUGREPORT = @PACKAGE_BUGREPORT@ PACKAGE_NAME = @PACKAGE_NAME@ PACKAGE_STRING = @PACKAGE_STRING@ PACKAGE_TARNAME = @PACKAGE_TARNAME@ PACKAGE_URL = @PACKAGE_URL@ PACKAGE_VERSION = @PACKAGE_VERSION@ PATH_SEPARATOR = @PATH_SEPARATOR@ POW_LIB = @POW_LIB@ PRECISION = @PRECISION@ PREC_SUFFIX = @PREC_SUFFIX@ PTHREAD_CC = @PTHREAD_CC@ PTHREAD_CFLAGS = @PTHREAD_CFLAGS@ PTHREAD_LIBS = @PTHREAD_LIBS@ RANLIB = @RANLIB@ SED = @SED@ SET_MAKE = @SET_MAKE@ SHARED_VERSION_INFO = @SHARED_VERSION_INFO@ SHELL = @SHELL@ SSE2_CFLAGS = @SSE2_CFLAGS@ STACK_ALIGN_CFLAGS = @STACK_ALIGN_CFLAGS@ STRIP = @STRIP@ THREADLIBS = @THREADLIBS@ VERSION = @VERSION@ abs_builddir = @abs_builddir@ abs_srcdir = @abs_srcdir@ abs_top_builddir = @abs_top_builddir@ abs_top_srcdir = @abs_top_srcdir@ ac_ct_AR = @ac_ct_AR@ ac_ct_CC = @ac_ct_CC@ ac_ct_DUMPBIN = @ac_ct_DUMPBIN@ ac_ct_F77 = @ac_ct_F77@ acx_pthread_config = @acx_pthread_config@ am__include = @am__include@ am__leading_dot = @am__leading_dot@ am__quote = @am__quote@ am__tar = @am__tar@ am__untar = @am__untar@ bindir = @bindir@ build = @build@ build_alias = @build_alias@ build_cpu = @build_cpu@ build_os = @build_os@ build_vendor = @build_vendor@ builddir = @builddir@ datadir = @datadir@ datarootdir = @datarootdir@ docdir = @docdir@ dvidir = @dvidir@ exec_prefix = @exec_prefix@ host = @host@ host_alias = @host_alias@ host_cpu = @host_cpu@ host_os = @host_os@ host_vendor = @host_vendor@ htmldir = @htmldir@ includedir = @includedir@ infodir = @infodir@ install_sh = @install_sh@ libdir = @libdir@ libexecdir = @libexecdir@ localedir = @localedir@ localstatedir = @localstatedir@ mandir = @mandir@ mkdir_p = @mkdir_p@ oldincludedir = @oldincludedir@ pdfdir = @pdfdir@ prefix = @prefix@ program_transform_name = @program_transform_name@ psdir = @psdir@ sbindir = @sbindir@ sharedstatedir = @sharedstatedir@ srcdir = @srcdir@ sysconfdir = @sysconfdir@ target_alias = @target_alias@ top_build_prefix = @top_build_prefix@ top_builddir = @top_builddir@ top_srcdir = @top_srcdir@ ########################################################################### AM_CPPFLAGS = -I$(top_srcdir)/kernel -I$(top_srcdir)/rdft \ -I$(top_srcdir)/rdft/scalar noinst_LTLIBRARIES = librdft_scalar_r2r.la ########################################################################### # The following lines specify the REDFT/RODFT/DHT sizes for which to generate # specialized codelets. Currently, only REDFT01/10 of size 8 (used in JPEG). # e_ is a hard-coded REDFT FFT (DCT) of size E00 = # e00_2.c e00_3.c e00_4.c e00_5.c e00_6.c e00_7.c e00_8.c E01 = e01_8.c # e01_2.c e01_3.c e01_4.c e01_5.c e01_6.c e01_7.c E10 = e10_8.c # e10_2.c e10_3.c e10_4.c e10_5.c e10_6.c e10_7.c E11 = # e11_2.c e11_3.c e11_4.c e11_5.c e11_6.c e11_7.c e11_8.c # o_ is a hard-coded RODFT FFT (DST) of size O00 = # o00_2.c o00_3.c o00_4.c o00_5.c o00_6.c o00_7.c o00_8.c O01 = # o01_2.c o01_3.c o01_4.c o01_5.c o01_6.c o01_7.c o01_8.c O10 = # o10_2.c o10_3.c o10_4.c o10_5.c o10_6.c o10_7.c o10_8.c O11 = # o11_2.c o11_3.c o11_4.c o11_5.c o11_6.c o11_7.c o11_8.c # dht_ is a hard-coded DHT of size DHT = # dht_2.c dht_3.c dht_4.c dht_5.c dht_6.c dht_7.c dht_8.c ########################################################################### ALL_CODELETS = $(E00) $(E01) $(E10) $(E11) $(O00) $(O01) $(O10) $(O11) $(DHT) BUILT_SOURCES = $(ALL_CODELETS) $(CODLIST) librdft_scalar_r2r_la_SOURCES = $(BUILT_SOURCES) SOLVTAB_NAME = X(solvtab_rdft_r2r) XRENAME = X CODLIST = codlist.c CODELET_NAME = codelet_ @MAINTAINER_MODE_TRUE@INDENT = indent -kr -cs -i5 -l800 -fca -nfc1 -sc -sob -cli4 -TR -Tplanner -TV @MAINTAINER_MODE_TRUE@TWOVERS = sh ${top_srcdir}/support/twovers.sh @MAINTAINER_MODE_TRUE@GENFFTDIR = ${top_builddir}/genfft @MAINTAINER_MODE_TRUE@GEN_NOTW = ${GENFFTDIR}/gen_notw.native @MAINTAINER_MODE_TRUE@GEN_NOTW_C = ${GENFFTDIR}/gen_notw_c.native @MAINTAINER_MODE_TRUE@GEN_TWIDDLE = ${GENFFTDIR}/gen_twiddle.native @MAINTAINER_MODE_TRUE@GEN_TWIDDLE_C = ${GENFFTDIR}/gen_twiddle_c.native @MAINTAINER_MODE_TRUE@GEN_TWIDSQ = ${GENFFTDIR}/gen_twidsq.native @MAINTAINER_MODE_TRUE@GEN_TWIDSQ_C = ${GENFFTDIR}/gen_twidsq_c.native @MAINTAINER_MODE_TRUE@GEN_R2CF = ${GENFFTDIR}/gen_r2cf.native @MAINTAINER_MODE_TRUE@GEN_R2CB = ${GENFFTDIR}/gen_r2cb.native @MAINTAINER_MODE_TRUE@GEN_HC2HC = ${GENFFTDIR}/gen_hc2hc.native @MAINTAINER_MODE_TRUE@GEN_HC2C = ${GENFFTDIR}/gen_hc2c.native @MAINTAINER_MODE_TRUE@GEN_HC2CDFT = ${GENFFTDIR}/gen_hc2cdft.native @MAINTAINER_MODE_TRUE@GEN_HC2CDFT_C = ${GENFFTDIR}/gen_hc2cdft_c.native @MAINTAINER_MODE_TRUE@GEN_R2R = ${GENFFTDIR}/gen_r2r.native @MAINTAINER_MODE_TRUE@PRELUDE_DFT = ${top_srcdir}/support/codelet_prelude.dft @MAINTAINER_MODE_TRUE@PRELUDE_RDFT = ${top_srcdir}/support/codelet_prelude.rdft @MAINTAINER_MODE_TRUE@ADD_DATE = sed -e s/@DATE@/"`date`"/ @MAINTAINER_MODE_TRUE@COPYRIGHT = ${top_srcdir}/COPYRIGHT @MAINTAINER_MODE_TRUE@CODELET_DEPS = $(COPYRIGHT) $(PRELUDE) @MAINTAINER_MODE_TRUE@PRELUDE_COMMANDS_DFT = cat $(COPYRIGHT) $(PRELUDE_DFT) @MAINTAINER_MODE_TRUE@PRELUDE_COMMANDS_RDFT = cat $(COPYRIGHT) $(PRELUDE_RDFT) @MAINTAINER_MODE_TRUE@FLAGS_COMMON = -compact -variables 4 @MAINTAINER_MODE_TRUE@DFT_FLAGS_COMMON = $(FLAGS_COMMON) -pipeline-latency 4 @MAINTAINER_MODE_TRUE@RDFT_FLAGS_COMMON = $(FLAGS_COMMON) -pipeline-latency 4 # special rules for regenerating codelets. @MAINTAINER_MODE_TRUE@FLAGS_E00 = $(RDFT_FLAGS_COMMON) @MAINTAINER_MODE_TRUE@FLAGS_E01 = $(RDFT_FLAGS_COMMON) @MAINTAINER_MODE_TRUE@FLAGS_E10 = $(RDFT_FLAGS_COMMON) @MAINTAINER_MODE_TRUE@FLAGS_E11 = $(RDFT_FLAGS_COMMON) @MAINTAINER_MODE_TRUE@FLAGS_O00 = $(RDFT_FLAGS_COMMON) @MAINTAINER_MODE_TRUE@FLAGS_O01 = $(RDFT_FLAGS_COMMON) @MAINTAINER_MODE_TRUE@FLAGS_O10 = $(RDFT_FLAGS_COMMON) @MAINTAINER_MODE_TRUE@FLAGS_O11 = $(RDFT_FLAGS_COMMON) @MAINTAINER_MODE_TRUE@FLAGS_DHT = $(RDFT_FLAGS_COMMON) all: $(BUILT_SOURCES) $(MAKE) $(AM_MAKEFLAGS) all-am .SUFFIXES: .SUFFIXES: .c .lo .o .obj $(srcdir)/Makefile.in: @MAINTAINER_MODE_TRUE@ $(srcdir)/Makefile.am $(top_srcdir)/support/Makefile.codelets $(am__configure_deps) @for dep in $?; do \ case '$(am__configure_deps)' in \ *$$dep*) \ ( cd $(top_builddir) && $(MAKE) $(AM_MAKEFLAGS) am--refresh ) \ && { if test -f $@; then exit 0; else break; fi; }; \ exit 1;; \ esac; \ done; \ echo ' cd $(top_srcdir) && $(AUTOMAKE) --gnu rdft/scalar/r2r/Makefile'; \ $(am__cd) $(top_srcdir) && \ $(AUTOMAKE) --gnu rdft/scalar/r2r/Makefile .PRECIOUS: Makefile Makefile: $(srcdir)/Makefile.in $(top_builddir)/config.status @case '$?' in \ *config.status*) \ cd $(top_builddir) && $(MAKE) $(AM_MAKEFLAGS) am--refresh;; \ *) \ echo ' cd $(top_builddir) && $(SHELL) ./config.status $(subdir)/$@ $(am__depfiles_maybe)'; \ cd $(top_builddir) && $(SHELL) ./config.status $(subdir)/$@ $(am__depfiles_maybe);; \ esac; $(top_srcdir)/support/Makefile.codelets: $(top_builddir)/config.status: $(top_srcdir)/configure $(CONFIG_STATUS_DEPENDENCIES) cd $(top_builddir) && $(MAKE) $(AM_MAKEFLAGS) am--refresh $(top_srcdir)/configure: @MAINTAINER_MODE_TRUE@ $(am__configure_deps) cd $(top_builddir) && $(MAKE) $(AM_MAKEFLAGS) am--refresh $(ACLOCAL_M4): @MAINTAINER_MODE_TRUE@ $(am__aclocal_m4_deps) cd $(top_builddir) && $(MAKE) $(AM_MAKEFLAGS) am--refresh $(am__aclocal_m4_deps): clean-noinstLTLIBRARIES: -test -z "$(noinst_LTLIBRARIES)" || rm -f $(noinst_LTLIBRARIES) @list='$(noinst_LTLIBRARIES)'; for p in $$list; do \ dir="`echo $$p | sed -e 's|/[^/]*$$||'`"; \ test "$$dir" != "$$p" || dir=.; \ echo "rm -f \"$${dir}/so_locations\""; \ rm -f "$${dir}/so_locations"; \ done librdft_scalar_r2r.la: $(librdft_scalar_r2r_la_OBJECTS) $(librdft_scalar_r2r_la_DEPENDENCIES) $(EXTRA_librdft_scalar_r2r_la_DEPENDENCIES) $(LINK) $(librdft_scalar_r2r_la_OBJECTS) $(librdft_scalar_r2r_la_LIBADD) $(LIBS) mostlyclean-compile: -rm -f *.$(OBJEXT) distclean-compile: -rm -f *.tab.c @AMDEP_TRUE@@am__include@ @am__quote@./$(DEPDIR)/codlist.Plo@am__quote@ @AMDEP_TRUE@@am__include@ @am__quote@./$(DEPDIR)/e01_8.Plo@am__quote@ @AMDEP_TRUE@@am__include@ @am__quote@./$(DEPDIR)/e10_8.Plo@am__quote@ .c.o: @am__fastdepCC_TRUE@ $(COMPILE) -MT $@ -MD -MP -MF $(DEPDIR)/$*.Tpo -c -o $@ $< @am__fastdepCC_TRUE@ $(am__mv) $(DEPDIR)/$*.Tpo $(DEPDIR)/$*.Po @AMDEP_TRUE@@am__fastdepCC_FALSE@ source='$<' object='$@' libtool=no @AMDEPBACKSLASH@ @AMDEP_TRUE@@am__fastdepCC_FALSE@ DEPDIR=$(DEPDIR) $(CCDEPMODE) $(depcomp) @AMDEPBACKSLASH@ @am__fastdepCC_FALSE@ $(COMPILE) -c $< .c.obj: @am__fastdepCC_TRUE@ $(COMPILE) -MT $@ -MD -MP -MF $(DEPDIR)/$*.Tpo -c -o $@ `$(CYGPATH_W) '$<'` @am__fastdepCC_TRUE@ $(am__mv) $(DEPDIR)/$*.Tpo $(DEPDIR)/$*.Po @AMDEP_TRUE@@am__fastdepCC_FALSE@ source='$<' object='$@' libtool=no @AMDEPBACKSLASH@ @AMDEP_TRUE@@am__fastdepCC_FALSE@ DEPDIR=$(DEPDIR) $(CCDEPMODE) $(depcomp) @AMDEPBACKSLASH@ @am__fastdepCC_FALSE@ $(COMPILE) -c `$(CYGPATH_W) '$<'` .c.lo: @am__fastdepCC_TRUE@ $(LTCOMPILE) -MT $@ -MD -MP -MF $(DEPDIR)/$*.Tpo -c -o $@ $< @am__fastdepCC_TRUE@ $(am__mv) $(DEPDIR)/$*.Tpo $(DEPDIR)/$*.Plo @AMDEP_TRUE@@am__fastdepCC_FALSE@ source='$<' object='$@' libtool=yes @AMDEPBACKSLASH@ @AMDEP_TRUE@@am__fastdepCC_FALSE@ DEPDIR=$(DEPDIR) $(CCDEPMODE) $(depcomp) @AMDEPBACKSLASH@ @am__fastdepCC_FALSE@ $(LTCOMPILE) -c -o $@ $< mostlyclean-libtool: -rm -f *.lo clean-libtool: -rm -rf .libs _libs ID: $(HEADERS) $(SOURCES) $(LISP) $(TAGS_FILES) list='$(SOURCES) $(HEADERS) $(LISP) $(TAGS_FILES)'; \ unique=`for i in $$list; do \ if test -f "$$i"; then echo $$i; else echo $(srcdir)/$$i; fi; \ done | \ $(AWK) '{ files[$$0] = 1; nonempty = 1; } \ END { if (nonempty) { for (i in files) print i; }; }'`; \ mkid -fID $$unique tags: TAGS TAGS: $(HEADERS) $(SOURCES) $(TAGS_DEPENDENCIES) \ $(TAGS_FILES) $(LISP) set x; \ here=`pwd`; \ list='$(SOURCES) $(HEADERS) $(LISP) $(TAGS_FILES)'; \ unique=`for i in $$list; do \ if test -f "$$i"; then echo $$i; else echo $(srcdir)/$$i; fi; \ done | \ $(AWK) '{ files[$$0] = 1; nonempty = 1; } \ END { if (nonempty) { for (i in files) print i; }; }'`; \ shift; \ if test -z "$(ETAGS_ARGS)$$*$$unique"; then :; else \ test -n "$$unique" || unique=$$empty_fix; \ if test $$# -gt 0; then \ $(ETAGS) $(ETAGSFLAGS) $(AM_ETAGSFLAGS) $(ETAGS_ARGS) \ "$$@" $$unique; \ else \ $(ETAGS) $(ETAGSFLAGS) $(AM_ETAGSFLAGS) $(ETAGS_ARGS) \ $$unique; \ fi; \ fi ctags: CTAGS CTAGS: $(HEADERS) $(SOURCES) $(TAGS_DEPENDENCIES) \ $(TAGS_FILES) $(LISP) list='$(SOURCES) $(HEADERS) $(LISP) $(TAGS_FILES)'; \ unique=`for i in $$list; do \ if test -f "$$i"; then echo $$i; else echo $(srcdir)/$$i; fi; \ done | \ $(AWK) '{ files[$$0] = 1; nonempty = 1; } \ END { if (nonempty) { for (i in files) print i; }; }'`; \ test -z "$(CTAGS_ARGS)$$unique" \ || $(CTAGS) $(CTAGSFLAGS) $(AM_CTAGSFLAGS) $(CTAGS_ARGS) \ $$unique GTAGS: here=`$(am__cd) $(top_builddir) && pwd` \ && $(am__cd) $(top_srcdir) \ && gtags -i $(GTAGS_ARGS) "$$here" distclean-tags: -rm -f TAGS ID GTAGS GRTAGS GSYMS GPATH tags distdir: $(DISTFILES) @srcdirstrip=`echo "$(srcdir)" | sed 's/[].[^$$\\*]/\\\\&/g'`; \ topsrcdirstrip=`echo "$(top_srcdir)" | sed 's/[].[^$$\\*]/\\\\&/g'`; \ list='$(DISTFILES)'; \ dist_files=`for file in $$list; do echo $$file; done | \ sed -e "s|^$$srcdirstrip/||;t" \ -e "s|^$$topsrcdirstrip/|$(top_builddir)/|;t"`; \ case $$dist_files in \ */*) $(MKDIR_P) `echo "$$dist_files" | \ sed '/\//!d;s|^|$(distdir)/|;s,/[^/]*$$,,' | \ sort -u` ;; \ esac; \ for file in $$dist_files; do \ if test -f $$file || test -d $$file; then d=.; else d=$(srcdir); fi; \ if test -d $$d/$$file; then \ dir=`echo "/$$file" | sed -e 's,/[^/]*$$,,'`; \ if test -d "$(distdir)/$$file"; then \ find "$(distdir)/$$file" -type d ! -perm -700 -exec chmod u+rwx {} \;; \ fi; \ if test -d $(srcdir)/$$file && test $$d != $(srcdir); then \ cp -fpR $(srcdir)/$$file "$(distdir)$$dir" || exit 1; \ find "$(distdir)/$$file" -type d ! -perm -700 -exec chmod u+rwx {} \;; \ fi; \ cp -fpR $$d/$$file "$(distdir)$$dir" || exit 1; \ else \ test -f "$(distdir)/$$file" \ || cp -p $$d/$$file "$(distdir)/$$file" \ || exit 1; \ fi; \ done check-am: all-am check: $(BUILT_SOURCES) $(MAKE) $(AM_MAKEFLAGS) check-am all-am: Makefile $(LTLIBRARIES) installdirs: install: $(BUILT_SOURCES) $(MAKE) $(AM_MAKEFLAGS) install-am install-exec: install-exec-am install-data: install-data-am uninstall: uninstall-am install-am: all-am @$(MAKE) $(AM_MAKEFLAGS) install-exec-am install-data-am installcheck: installcheck-am install-strip: if test -z '$(STRIP)'; then \ $(MAKE) $(AM_MAKEFLAGS) INSTALL_PROGRAM="$(INSTALL_STRIP_PROGRAM)" \ install_sh_PROGRAM="$(INSTALL_STRIP_PROGRAM)" INSTALL_STRIP_FLAG=-s \ install; \ else \ $(MAKE) $(AM_MAKEFLAGS) INSTALL_PROGRAM="$(INSTALL_STRIP_PROGRAM)" \ install_sh_PROGRAM="$(INSTALL_STRIP_PROGRAM)" INSTALL_STRIP_FLAG=-s \ "INSTALL_PROGRAM_ENV=STRIPPROG='$(STRIP)'" install; \ fi mostlyclean-generic: clean-generic: distclean-generic: -test -z "$(CONFIG_CLEAN_FILES)" || rm -f $(CONFIG_CLEAN_FILES) -test . = "$(srcdir)" || test -z "$(CONFIG_CLEAN_VPATH_FILES)" || rm -f $(CONFIG_CLEAN_VPATH_FILES) maintainer-clean-generic: @echo "This command is intended for maintainers to use" @echo "it deletes files that may require special tools to rebuild." -test -z "$(BUILT_SOURCES)" || rm -f $(BUILT_SOURCES) clean: clean-am clean-am: clean-generic clean-libtool clean-noinstLTLIBRARIES \ mostlyclean-am distclean: distclean-am -rm -rf ./$(DEPDIR) -rm -f Makefile distclean-am: clean-am distclean-compile distclean-generic \ distclean-tags dvi: dvi-am dvi-am: html: html-am html-am: info: info-am info-am: install-data-am: install-dvi: install-dvi-am install-dvi-am: install-exec-am: install-html: install-html-am install-html-am: install-info: install-info-am install-info-am: install-man: install-pdf: install-pdf-am install-pdf-am: install-ps: install-ps-am install-ps-am: installcheck-am: maintainer-clean: maintainer-clean-am -rm -rf ./$(DEPDIR) -rm -f Makefile maintainer-clean-am: distclean-am maintainer-clean-generic \ maintainer-clean-local mostlyclean: mostlyclean-am mostlyclean-am: mostlyclean-compile mostlyclean-generic \ mostlyclean-libtool pdf: pdf-am pdf-am: ps: ps-am ps-am: uninstall-am: .MAKE: all check install install-am install-strip .PHONY: CTAGS GTAGS all all-am check check-am clean clean-generic \ clean-libtool clean-noinstLTLIBRARIES ctags distclean \ distclean-compile distclean-generic distclean-libtool \ distclean-tags distdir dvi dvi-am html html-am info info-am \ install install-am install-data install-data-am install-dvi \ install-dvi-am install-exec install-exec-am install-html \ install-html-am install-info install-info-am install-man \ install-pdf install-pdf-am install-ps install-ps-am \ install-strip installcheck installcheck-am installdirs \ maintainer-clean maintainer-clean-generic \ maintainer-clean-local mostlyclean mostlyclean-compile \ mostlyclean-generic mostlyclean-libtool pdf pdf-am ps ps-am \ tags uninstall uninstall-am # rule to build codlist $(CODLIST): Makefile ( \ echo "#include \"ifftw.h\""; \ echo $(INCLUDE_SIMD_HEADER); \ echo; \ for i in $(ALL_CODELETS) NIL; do \ if test "$$i" != NIL; then \ j=`basename $$i | sed -e 's/[.][cS]$$//g'`; \ echo "extern void $(XRENAME)($(CODELET_NAME)$$j)(planner *);"; \ fi \ done; \ echo; \ echo; \ echo "extern const solvtab $(SOLVTAB_NAME);"; \ echo "const solvtab $(SOLVTAB_NAME) = {"; \ for i in $(ALL_CODELETS) NIL; do \ if test "$$i" != NIL; then \ j=`basename $$i | sed -e 's/[.][cS]$$//g'`; \ echo " SOLVTAB($(XRENAME)($(CODELET_NAME)$$j)),"; \ fi \ done; \ echo " SOLVTAB_END"; \ echo "};"; \ ) >$@ # only delete codlist.c in maintainer-mode, since it is included in the dist # FIXME: is there a way to delete in 'make clean' only when builddir != srcdir? maintainer-clean-local: rm -f $(CODLIST) # cancel the hideous builtin rules that cause an infinite loop @MAINTAINER_MODE_TRUE@%: %.o @MAINTAINER_MODE_TRUE@%: %.s @MAINTAINER_MODE_TRUE@%: %.c @MAINTAINER_MODE_TRUE@%: %.S @MAINTAINER_MODE_TRUE@e00_%.c: $(CODELET_DEPS) $(GEN_R2R) @MAINTAINER_MODE_TRUE@ ($(PRELUDE_COMMANDS_RDFT); $(TWOVERS) $(GEN_R2R) $(FLAGS_E00) -redft00 -n $* -name e00_$* -include "r2r.h") | $(ADD_DATE) | $(INDENT) >$@ @MAINTAINER_MODE_TRUE@e01_%.c: $(CODELET_DEPS) $(GEN_R2R) @MAINTAINER_MODE_TRUE@ ($(PRELUDE_COMMANDS_RDFT); $(TWOVERS) $(GEN_R2R) $(FLAGS_E01) -redft01 -n $* -name e01_$* -include "r2r.h") | $(ADD_DATE) | $(INDENT) >$@ @MAINTAINER_MODE_TRUE@e10_%.c: $(CODELET_DEPS) $(GEN_R2R) @MAINTAINER_MODE_TRUE@ ($(PRELUDE_COMMANDS_RDFT); $(TWOVERS) $(GEN_R2R) $(FLAGS_E10) -redft10 -n $* -name e10_$* -include "r2r.h") | $(ADD_DATE) | $(INDENT) >$@ @MAINTAINER_MODE_TRUE@e11_%.c: $(CODELET_DEPS) $(GEN_R2R) @MAINTAINER_MODE_TRUE@ ($(PRELUDE_COMMANDS_RDFT); $(TWOVERS) $(GEN_R2R) $(FLAGS_E11) -redft11 -n $* -name e11_$* -include "r2r.h") | $(ADD_DATE) | $(INDENT) >$@ @MAINTAINER_MODE_TRUE@o00_%.c: $(CODELET_DEPS) $(GEN_R2R) @MAINTAINER_MODE_TRUE@ ($(PRELUDE_COMMANDS_RDFT); $(TWOVERS) $(GEN_R2R) $(FLAGS_O00) -rodft00 -n $* -name o00_$* -include "r2r.h") | $(ADD_DATE) | $(INDENT) >$@ @MAINTAINER_MODE_TRUE@o01_%.c: $(CODELET_DEPS) $(GEN_R2R) @MAINTAINER_MODE_TRUE@ ($(PRELUDE_COMMANDS_RDFT); $(TWOVERS) $(GEN_R2R) $(FLAGS_O01) -rodft01 -n $* -name o01_$* -include "r2r.h") | $(ADD_DATE) | $(INDENT) >$@ @MAINTAINER_MODE_TRUE@o10_%.c: $(CODELET_DEPS) $(GEN_R2R) @MAINTAINER_MODE_TRUE@ ($(PRELUDE_COMMANDS_RDFT); $(TWOVERS) $(GEN_R2R) $(FLAGS_O10) -rodft10 -n $* -name o10_$* -include "r2r.h") | $(ADD_DATE) | $(INDENT) >$@ @MAINTAINER_MODE_TRUE@o11_%.c: $(CODELET_DEPS) $(GEN_R2R) @MAINTAINER_MODE_TRUE@ ($(PRELUDE_COMMANDS_RDFT); $(TWOVERS) $(GEN_R2R) $(FLAGS_O11) -rodft11 -n $* -name o11_$* -include "r2r.h") | $(ADD_DATE) | $(INDENT) >$@ @MAINTAINER_MODE_TRUE@dht_%.c: $(CODELET_DEPS) $(GEN_R2R) @MAINTAINER_MODE_TRUE@ ($(PRELUDE_COMMANDS_RDFT); $(TWOVERS) $(GEN_R2R) $(FLAGS_DHT) -dht -sign 1 -n $* -name dht_$* -include "r2r.h") | $(ADD_DATE) | $(INDENT) >$@ # Tell versions [3.59,3.63) of GNU make to not export all variables. # Otherwise a system limit (for SysV at least) may be exceeded. .NOEXPORT: fftw-3.3.3/rdft/scalar/r2r/e01_8.c0000644000175000017500000001441012054411064013275 00000000000000/* * Copyright (c) 2003, 2007-11 Matteo Frigo * Copyright (c) 2003, 2007-11 Massachusetts Institute of Technology * * This program is free software; you can redistribute it and/or modify * it under the terms of the GNU General Public License as published by * the Free Software Foundation; either version 2 of the License, or * (at your option) any later version. * * This program is distributed in the hope that it will be useful, * but WITHOUT ANY WARRANTY; without even the implied warranty of * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the * GNU General Public License for more details. * * You should have received a copy of the GNU General Public License * along with this program; if not, write to the Free Software * Foundation, Inc., 51 Franklin Street, Fifth Floor, Boston, MA 02110-1301 USA * */ /* This file was automatically generated --- DO NOT EDIT */ /* Generated on Sun Nov 25 07:42:27 EST 2012 */ #include "codelet-rdft.h" #ifdef HAVE_FMA /* Generated by: ../../../genfft/gen_r2r.native -fma -reorder-insns -schedule-for-pipeline -compact -variables 4 -pipeline-latency 4 -redft01 -n 8 -name e01_8 -include r2r.h */ /* * This function contains 26 FP additions, 24 FP multiplications, * (or, 2 additions, 0 multiplications, 24 fused multiply/add), * 27 stack variables, 8 constants, and 16 memory accesses */ #include "r2r.h" static void e01_8(const R *I, R *O, stride is, stride os, INT v, INT ivs, INT ovs) { DK(KP1_961570560, +1.961570560806460898252364472268478073947867462); DK(KP1_662939224, +1.662939224605090474157576755235811513477121624); DK(KP198912367, +0.198912367379658006911597622644676228597850501); DK(KP668178637, +0.668178637919298919997757686523080761552472251); DK(KP707106781, +0.707106781186547524400844362104849039284835938); DK(KP1_847759065, +1.847759065022573512256366378793576573644833252); DK(KP414213562, +0.414213562373095048801688724209698078569671875); DK(KP1_414213562, +1.414213562373095048801688724209698078569671875); { INT i; for (i = v; i > 0; i = i - 1, I = I + ivs, O = O + ovs, MAKE_VOLATILE_STRIDE(16, is), MAKE_VOLATILE_STRIDE(16, os)) { E T8, Td, Th, T7, Tp, Tl, Te, Tb; { E Tj, T3, Tk, T6, T9, Ta; { E T1, T2, T4, T5; T1 = I[0]; T2 = I[WS(is, 4)]; T4 = I[WS(is, 2)]; T5 = I[WS(is, 6)]; T8 = I[WS(is, 1)]; Tj = FNMS(KP1_414213562, T2, T1); T3 = FMA(KP1_414213562, T2, T1); Tk = FMS(KP414213562, T4, T5); T6 = FMA(KP414213562, T5, T4); Td = I[WS(is, 7)]; T9 = I[WS(is, 5)]; Ta = I[WS(is, 3)]; } Th = FNMS(KP1_847759065, T6, T3); T7 = FMA(KP1_847759065, T6, T3); Tp = FNMS(KP1_847759065, Tk, Tj); Tl = FMA(KP1_847759065, Tk, Tj); Te = Ta - T9; Tb = T9 + Ta; } { E Tn, Tf, Tc, Tm; Tn = FNMS(KP707106781, Te, Td); Tf = FMA(KP707106781, Te, Td); Tc = FMA(KP707106781, Tb, T8); Tm = FNMS(KP707106781, Tb, T8); { E Tq, To, Tg, Ti; Tq = FMA(KP668178637, Tm, Tn); To = FNMS(KP668178637, Tn, Tm); Tg = FMA(KP198912367, Tf, Tc); Ti = FNMS(KP198912367, Tc, Tf); O[WS(os, 1)] = FMA(KP1_662939224, To, Tl); O[WS(os, 6)] = FNMS(KP1_662939224, To, Tl); O[WS(os, 2)] = FMA(KP1_662939224, Tq, Tp); O[WS(os, 5)] = FNMS(KP1_662939224, Tq, Tp); O[WS(os, 4)] = FMA(KP1_961570560, Ti, Th); O[WS(os, 3)] = FNMS(KP1_961570560, Ti, Th); O[0] = FMA(KP1_961570560, Tg, T7); O[WS(os, 7)] = FNMS(KP1_961570560, Tg, T7); } } } } } static const kr2r_desc desc = { 8, "e01_8", {2, 0, 24, 0}, &GENUS, REDFT01 }; void X(codelet_e01_8) (planner *p) { X(kr2r_register) (p, e01_8, &desc); } #else /* HAVE_FMA */ /* Generated by: ../../../genfft/gen_r2r.native -compact -variables 4 -pipeline-latency 4 -redft01 -n 8 -name e01_8 -include r2r.h */ /* * This function contains 26 FP additions, 15 FP multiplications, * (or, 20 additions, 9 multiplications, 6 fused multiply/add), * 28 stack variables, 8 constants, and 16 memory accesses */ #include "r2r.h" static void e01_8(const R *I, R *O, stride is, stride os, INT v, INT ivs, INT ovs) { DK(KP1_662939224, +1.662939224605090474157576755235811513477121624); DK(KP1_111140466, +1.111140466039204449485661627897065748749874382); DK(KP390180644, +0.390180644032256535696569736954044481855383236); DK(KP1_961570560, +1.961570560806460898252364472268478073947867462); DK(KP707106781, +0.707106781186547524400844362104849039284835938); DK(KP1_414213562, +1.414213562373095048801688724209698078569671875); DK(KP765366864, +0.765366864730179543456919968060797733522689125); DK(KP1_847759065, +1.847759065022573512256366378793576573644833252); { INT i; for (i = v; i > 0; i = i - 1, I = I + ivs, O = O + ovs, MAKE_VOLATILE_STRIDE(16, is), MAKE_VOLATILE_STRIDE(16, os)) { E T7, Tl, T4, Tk, Td, To, Tg, Tn; { E T5, T6, T1, T3, T2; T5 = I[WS(is, 2)]; T6 = I[WS(is, 6)]; T7 = FMA(KP1_847759065, T5, KP765366864 * T6); Tl = FNMS(KP1_847759065, T6, KP765366864 * T5); T1 = I[0]; T2 = I[WS(is, 4)]; T3 = KP1_414213562 * T2; T4 = T1 + T3; Tk = T1 - T3; { E T9, Tf, Tc, Te, Ta, Tb; T9 = I[WS(is, 1)]; Tf = I[WS(is, 7)]; Ta = I[WS(is, 5)]; Tb = I[WS(is, 3)]; Tc = KP707106781 * (Ta + Tb); Te = KP707106781 * (Ta - Tb); Td = T9 + Tc; To = Te + Tf; Tg = Te - Tf; Tn = T9 - Tc; } } { E T8, Th, Tq, Tr; T8 = T4 + T7; Th = FNMS(KP390180644, Tg, KP1_961570560 * Td); O[WS(os, 7)] = T8 - Th; O[0] = T8 + Th; Tq = Tk - Tl; Tr = FMA(KP1_111140466, Tn, KP1_662939224 * To); O[WS(os, 5)] = Tq - Tr; O[WS(os, 2)] = Tq + Tr; } { E Ti, Tj, Tm, Tp; Ti = T4 - T7; Tj = FMA(KP390180644, Td, KP1_961570560 * Tg); O[WS(os, 4)] = Ti - Tj; O[WS(os, 3)] = Ti + Tj; Tm = Tk + Tl; Tp = FNMS(KP1_111140466, To, KP1_662939224 * Tn); O[WS(os, 6)] = Tm - Tp; O[WS(os, 1)] = Tm + Tp; } } } } static const kr2r_desc desc = { 8, "e01_8", {20, 9, 6, 0}, &GENUS, REDFT01 }; void X(codelet_e01_8) (planner *p) { X(kr2r_register) (p, e01_8, &desc); } #endif /* HAVE_FMA */ fftw-3.3.3/rdft/scalar/r2r/e10_8.c0000644000175000017500000001444412054411064013304 00000000000000/* * Copyright (c) 2003, 2007-11 Matteo Frigo * Copyright (c) 2003, 2007-11 Massachusetts Institute of Technology * * This program is free software; you can redistribute it and/or modify * it under the terms of the GNU General Public License as published by * the Free Software Foundation; either version 2 of the License, or * (at your option) any later version. * * This program is distributed in the hope that it will be useful, * but WITHOUT ANY WARRANTY; without even the implied warranty of * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the * GNU General Public License for more details. * * You should have received a copy of the GNU General Public License * along with this program; if not, write to the Free Software * Foundation, Inc., 51 Franklin Street, Fifth Floor, Boston, MA 02110-1301 USA * */ /* This file was automatically generated --- DO NOT EDIT */ /* Generated on Sun Nov 25 07:42:27 EST 2012 */ #include "codelet-rdft.h" #ifdef HAVE_FMA /* Generated by: ../../../genfft/gen_r2r.native -fma -reorder-insns -schedule-for-pipeline -compact -variables 4 -pipeline-latency 4 -redft10 -n 8 -name e10_8 -include r2r.h */ /* * This function contains 26 FP additions, 18 FP multiplications, * (or, 16 additions, 8 multiplications, 10 fused multiply/add), * 28 stack variables, 9 constants, and 16 memory accesses */ #include "r2r.h" static void e10_8(const R *I, R *O, stride is, stride os, INT v, INT ivs, INT ovs) { DK(KP668178637, +0.668178637919298919997757686523080761552472251); DK(KP1_662939224, +1.662939224605090474157576755235811513477121624); DK(KP198912367, +0.198912367379658006911597622644676228597850501); DK(KP1_961570560, +1.961570560806460898252364472268478073947867462); DK(KP707106781, +0.707106781186547524400844362104849039284835938); DK(KP1_414213562, +1.414213562373095048801688724209698078569671875); DK(KP2_000000000, +2.000000000000000000000000000000000000000000000); DK(KP414213562, +0.414213562373095048801688724209698078569671875); DK(KP1_847759065, +1.847759065022573512256366378793576573644833252); { INT i; for (i = v; i > 0; i = i - 1, I = I + ivs, O = O + ovs, MAKE_VOLATILE_STRIDE(16, is), MAKE_VOLATILE_STRIDE(16, os)) { E T3, Te, Tl, Tp, Tm, T6, Tn, T9; { E T4, Tj, Tk, T5, T7, T8; { E T1, T2, Tc, Td; T1 = I[0]; T2 = I[WS(is, 7)]; Tc = I[WS(is, 4)]; Td = I[WS(is, 3)]; T4 = I[WS(is, 2)]; Tj = T1 + T2; T3 = T1 - T2; Tk = Tc + Td; Te = Tc - Td; T5 = I[WS(is, 5)]; T7 = I[WS(is, 1)]; T8 = I[WS(is, 6)]; } Tl = Tj - Tk; Tp = Tj + Tk; Tm = T4 + T5; T6 = T4 - T5; Tn = T7 + T8; T9 = T7 - T8; } { E Tg, Ti, Tb, Th; { E Tq, To, Ta, Tf; Tq = Tm + Tn; To = Tm - Tn; Ta = T6 + T9; Tf = T6 - T9; O[WS(os, 6)] = KP1_847759065 * (FMA(KP414213562, Tl, To)); O[WS(os, 2)] = KP1_847759065 * (FNMS(KP414213562, To, Tl)); O[0] = KP2_000000000 * (Tp + Tq); O[WS(os, 4)] = KP1_414213562 * (Tp - Tq); Tg = FNMS(KP707106781, Tf, Te); Ti = FMA(KP707106781, Tf, Te); Tb = FNMS(KP707106781, Ta, T3); Th = FMA(KP707106781, Ta, T3); } O[WS(os, 7)] = KP1_961570560 * (FMA(KP198912367, Th, Ti)); O[WS(os, 1)] = KP1_961570560 * (FNMS(KP198912367, Ti, Th)); O[WS(os, 5)] = -(KP1_662939224 * (FNMS(KP668178637, Tb, Tg))); O[WS(os, 3)] = KP1_662939224 * (FMA(KP668178637, Tg, Tb)); } } } } static const kr2r_desc desc = { 8, "e10_8", {16, 8, 10, 0}, &GENUS, REDFT10 }; void X(codelet_e10_8) (planner *p) { X(kr2r_register) (p, e10_8, &desc); } #else /* HAVE_FMA */ /* Generated by: ../../../genfft/gen_r2r.native -compact -variables 4 -pipeline-latency 4 -redft10 -n 8 -name e10_8 -include r2r.h */ /* * This function contains 26 FP additions, 16 FP multiplications, * (or, 20 additions, 10 multiplications, 6 fused multiply/add), * 28 stack variables, 9 constants, and 16 memory accesses */ #include "r2r.h" static void e10_8(const R *I, R *O, stride is, stride os, INT v, INT ivs, INT ovs) { DK(KP765366864, +0.765366864730179543456919968060797733522689125); DK(KP1_847759065, +1.847759065022573512256366378793576573644833252); DK(KP390180644, +0.390180644032256535696569736954044481855383236); DK(KP1_961570560, +1.961570560806460898252364472268478073947867462); DK(KP2_000000000, +2.000000000000000000000000000000000000000000000); DK(KP1_414213562, +1.414213562373095048801688724209698078569671875); DK(KP1_111140466, +1.111140466039204449485661627897065748749874382); DK(KP1_662939224, +1.662939224605090474157576755235811513477121624); DK(KP707106781, +0.707106781186547524400844362104849039284835938); { INT i; for (i = v; i > 0; i = i - 1, I = I + ivs, O = O + ovs, MAKE_VOLATILE_STRIDE(16, is), MAKE_VOLATILE_STRIDE(16, os)) { E T3, Tj, Tf, Tk, Ta, Tn, Tc, Tm; { E T1, T2, Td, Te; T1 = I[0]; T2 = I[WS(is, 7)]; T3 = T1 - T2; Tj = T1 + T2; Td = I[WS(is, 4)]; Te = I[WS(is, 3)]; Tf = Td - Te; Tk = Td + Te; { E T4, T5, T6, T7, T8, T9; T4 = I[WS(is, 2)]; T5 = I[WS(is, 5)]; T6 = T4 - T5; T7 = I[WS(is, 1)]; T8 = I[WS(is, 6)]; T9 = T7 - T8; Ta = KP707106781 * (T6 + T9); Tn = T7 + T8; Tc = KP707106781 * (T6 - T9); Tm = T4 + T5; } } { E Tb, Tg, Tp, Tq; Tb = T3 - Ta; Tg = Tc - Tf; O[WS(os, 3)] = FNMS(KP1_111140466, Tg, KP1_662939224 * Tb); O[WS(os, 5)] = FMA(KP1_662939224, Tg, KP1_111140466 * Tb); Tp = Tj + Tk; Tq = Tm + Tn; O[WS(os, 4)] = KP1_414213562 * (Tp - Tq); O[0] = KP2_000000000 * (Tp + Tq); } { E Th, Ti, Tl, To; Th = T3 + Ta; Ti = Tf + Tc; O[WS(os, 1)] = FNMS(KP390180644, Ti, KP1_961570560 * Th); O[WS(os, 7)] = FMA(KP1_961570560, Ti, KP390180644 * Th); Tl = Tj - Tk; To = Tm - Tn; O[WS(os, 2)] = FNMS(KP765366864, To, KP1_847759065 * Tl); O[WS(os, 6)] = FMA(KP765366864, Tl, KP1_847759065 * To); } } } } static const kr2r_desc desc = { 8, "e10_8", {20, 10, 6, 0}, &GENUS, REDFT10 }; void X(codelet_e10_8) (planner *p) { X(kr2r_register) (p, e10_8, &desc); } #endif /* HAVE_FMA */ fftw-3.3.3/rdft/scalar/r2r/codlist.c0000644000175000017500000000040712054411063014122 00000000000000#include "ifftw.h" extern void X(codelet_e01_8)(planner *); extern void X(codelet_e10_8)(planner *); extern const solvtab X(solvtab_rdft_r2r); const solvtab X(solvtab_rdft_r2r) = { SOLVTAB(X(codelet_e01_8)), SOLVTAB(X(codelet_e10_8)), SOLVTAB_END }; fftw-3.3.3/rdft/simd/0000755000175000017500000000000012054411110011347 500000000000000fftw-3.3.3/rdft/simd/Makefile.am0000644000175000017500000000013112054410050013320 00000000000000SUBDIRS = common sse2 avx altivec neon EXTRA_DIST = hc2cbv.h hc2cfv.h codlist.mk simd.mk fftw-3.3.3/rdft/simd/Makefile.in0000644000175000017500000004435712054410200013350 00000000000000# Makefile.in generated by automake 1.11.6 from Makefile.am. # @configure_input@ # Copyright (C) 1994, 1995, 1996, 1997, 1998, 1999, 2000, 2001, 2002, # 2003, 2004, 2005, 2006, 2007, 2008, 2009, 2010, 2011 Free Software # Foundation, Inc. # This Makefile.in is free software; the Free Software Foundation # gives unlimited permission to copy and/or distribute it, # with or without modifications, as long as this notice is preserved. # This program is distributed in the hope that it will be useful, # but WITHOUT ANY WARRANTY, to the extent permitted by law; without # even the implied warranty of MERCHANTABILITY or FITNESS FOR A # PARTICULAR PURPOSE. @SET_MAKE@ VPATH = @srcdir@ am__make_dryrun = \ { \ am__dry=no; \ case $$MAKEFLAGS in \ *\\[\ \ ]*) \ echo 'am--echo: ; @echo "AM" OK' | $(MAKE) -f - 2>/dev/null \ | grep '^AM OK$$' >/dev/null || am__dry=yes;; \ *) \ for am__flg in $$MAKEFLAGS; do \ case $$am__flg in \ *=*|--*) ;; \ *n*) am__dry=yes; break;; \ esac; \ done;; \ esac; \ test $$am__dry = yes; \ } pkgdatadir = $(datadir)/@PACKAGE@ pkgincludedir = $(includedir)/@PACKAGE@ pkglibdir = $(libdir)/@PACKAGE@ pkglibexecdir = $(libexecdir)/@PACKAGE@ am__cd = CDPATH="$${ZSH_VERSION+.}$(PATH_SEPARATOR)" && cd install_sh_DATA = $(install_sh) -c -m 644 install_sh_PROGRAM = $(install_sh) -c install_sh_SCRIPT = $(install_sh) -c INSTALL_HEADER = $(INSTALL_DATA) transform = $(program_transform_name) NORMAL_INSTALL = : PRE_INSTALL = : POST_INSTALL = : NORMAL_UNINSTALL = : PRE_UNINSTALL = : POST_UNINSTALL = : build_triplet = @build@ host_triplet = @host@ subdir = rdft/simd DIST_COMMON = $(srcdir)/Makefile.am $(srcdir)/Makefile.in ACLOCAL_M4 = $(top_srcdir)/aclocal.m4 am__aclocal_m4_deps = $(top_srcdir)/m4/acx_mpi.m4 \ $(top_srcdir)/m4/acx_pthread.m4 \ $(top_srcdir)/m4/ax_cc_maxopt.m4 \ $(top_srcdir)/m4/ax_check_compiler_flags.m4 \ $(top_srcdir)/m4/ax_compiler_vendor.m4 \ $(top_srcdir)/m4/ax_gcc_aligns_stack.m4 \ $(top_srcdir)/m4/ax_gcc_version.m4 \ $(top_srcdir)/m4/ax_openmp.m4 $(top_srcdir)/m4/libtool.m4 \ $(top_srcdir)/m4/ltoptions.m4 $(top_srcdir)/m4/ltsugar.m4 \ $(top_srcdir)/m4/ltversion.m4 $(top_srcdir)/m4/lt~obsolete.m4 \ $(top_srcdir)/configure.ac am__configure_deps = $(am__aclocal_m4_deps) $(CONFIGURE_DEPENDENCIES) \ $(ACLOCAL_M4) mkinstalldirs = $(install_sh) -d CONFIG_HEADER = $(top_builddir)/config.h CONFIG_CLEAN_FILES = CONFIG_CLEAN_VPATH_FILES = SOURCES = DIST_SOURCES = RECURSIVE_TARGETS = all-recursive check-recursive dvi-recursive \ html-recursive info-recursive install-data-recursive \ install-dvi-recursive install-exec-recursive \ install-html-recursive install-info-recursive \ install-pdf-recursive install-ps-recursive install-recursive \ installcheck-recursive installdirs-recursive pdf-recursive \ ps-recursive uninstall-recursive am__can_run_installinfo = \ case $$AM_UPDATE_INFO_DIR in \ n|no|NO) false;; \ *) (install-info --version) >/dev/null 2>&1;; \ esac RECURSIVE_CLEAN_TARGETS = mostlyclean-recursive clean-recursive \ distclean-recursive maintainer-clean-recursive AM_RECURSIVE_TARGETS = $(RECURSIVE_TARGETS:-recursive=) \ $(RECURSIVE_CLEAN_TARGETS:-recursive=) tags TAGS ctags CTAGS \ distdir ETAGS = etags CTAGS = ctags DIST_SUBDIRS = $(SUBDIRS) DISTFILES = $(DIST_COMMON) $(DIST_SOURCES) $(TEXINFOS) $(EXTRA_DIST) am__relativize = \ dir0=`pwd`; \ sed_first='s,^\([^/]*\)/.*$$,\1,'; \ sed_rest='s,^[^/]*/*,,'; \ sed_last='s,^.*/\([^/]*\)$$,\1,'; \ sed_butlast='s,/*[^/]*$$,,'; \ while test -n "$$dir1"; do \ first=`echo "$$dir1" | sed -e "$$sed_first"`; \ if test "$$first" != "."; then \ if test "$$first" = ".."; then \ dir2=`echo "$$dir0" | sed -e "$$sed_last"`/"$$dir2"; \ dir0=`echo "$$dir0" | sed -e "$$sed_butlast"`; \ else \ first2=`echo "$$dir2" | sed -e "$$sed_first"`; \ if test "$$first2" = "$$first"; then \ dir2=`echo "$$dir2" | sed -e "$$sed_rest"`; \ else \ dir2="../$$dir2"; \ fi; \ dir0="$$dir0"/"$$first"; \ fi; \ fi; \ dir1=`echo "$$dir1" | sed -e "$$sed_rest"`; \ done; \ reldir="$$dir2" ACLOCAL = @ACLOCAL@ ALLOCA = @ALLOCA@ ALTIVEC_CFLAGS = @ALTIVEC_CFLAGS@ AMTAR = @AMTAR@ AR = @AR@ AS = @AS@ AUTOCONF = @AUTOCONF@ AUTOHEADER = @AUTOHEADER@ AUTOMAKE = @AUTOMAKE@ AVX_CFLAGS = @AVX_CFLAGS@ AWK = @AWK@ CC = @CC@ CCDEPMODE = @CCDEPMODE@ CFLAGS = @CFLAGS@ CHECK_PL_OPTS = @CHECK_PL_OPTS@ CPP = @CPP@ CPPFLAGS = @CPPFLAGS@ CYGPATH_W = @CYGPATH_W@ C_FFTW_R2R_KIND = @C_FFTW_R2R_KIND@ C_MPI_FINT = @C_MPI_FINT@ DEFS = @DEFS@ DEPDIR = @DEPDIR@ DLLTOOL = @DLLTOOL@ DSYMUTIL = @DSYMUTIL@ DUMPBIN = @DUMPBIN@ ECHO_C = @ECHO_C@ ECHO_N = @ECHO_N@ ECHO_T = @ECHO_T@ EGREP = @EGREP@ EXEEXT = @EXEEXT@ F77 = @F77@ FFLAGS = @FFLAGS@ FGREP = @FGREP@ FLIBS = @FLIBS@ GREP = @GREP@ INSTALL = @INSTALL@ INSTALL_DATA = @INSTALL_DATA@ INSTALL_PROGRAM = @INSTALL_PROGRAM@ INSTALL_SCRIPT = @INSTALL_SCRIPT@ INSTALL_STRIP_PROGRAM = @INSTALL_STRIP_PROGRAM@ LD = @LD@ LDFLAGS = @LDFLAGS@ LIBOBJS = @LIBOBJS@ LIBQUADMATH = @LIBQUADMATH@ LIBS = @LIBS@ LIBTOOL = @LIBTOOL@ LIPO = @LIPO@ LN_S = @LN_S@ LTLIBOBJS = @LTLIBOBJS@ MAINT = @MAINT@ MAKEINFO = @MAKEINFO@ MANIFEST_TOOL = @MANIFEST_TOOL@ MKDIR_P = @MKDIR_P@ MPICC = @MPICC@ MPILIBS = @MPILIBS@ MPIRUN = @MPIRUN@ NEON_CFLAGS = @NEON_CFLAGS@ NM = @NM@ NMEDIT = @NMEDIT@ OBJDUMP = @OBJDUMP@ OBJEXT = @OBJEXT@ OCAMLBUILD = @OCAMLBUILD@ OPENMP_CFLAGS = @OPENMP_CFLAGS@ OTOOL = @OTOOL@ OTOOL64 = @OTOOL64@ PACKAGE = @PACKAGE@ PACKAGE_BUGREPORT = @PACKAGE_BUGREPORT@ PACKAGE_NAME = @PACKAGE_NAME@ PACKAGE_STRING = @PACKAGE_STRING@ PACKAGE_TARNAME = @PACKAGE_TARNAME@ PACKAGE_URL = @PACKAGE_URL@ PACKAGE_VERSION = @PACKAGE_VERSION@ PATH_SEPARATOR = @PATH_SEPARATOR@ POW_LIB = @POW_LIB@ PRECISION = @PRECISION@ PREC_SUFFIX = @PREC_SUFFIX@ PTHREAD_CC = @PTHREAD_CC@ PTHREAD_CFLAGS = @PTHREAD_CFLAGS@ PTHREAD_LIBS = @PTHREAD_LIBS@ RANLIB = @RANLIB@ SED = @SED@ SET_MAKE = @SET_MAKE@ SHARED_VERSION_INFO = @SHARED_VERSION_INFO@ SHELL = @SHELL@ SSE2_CFLAGS = @SSE2_CFLAGS@ STACK_ALIGN_CFLAGS = @STACK_ALIGN_CFLAGS@ STRIP = @STRIP@ THREADLIBS = @THREADLIBS@ VERSION = @VERSION@ abs_builddir = @abs_builddir@ abs_srcdir = @abs_srcdir@ abs_top_builddir = @abs_top_builddir@ abs_top_srcdir = @abs_top_srcdir@ ac_ct_AR = @ac_ct_AR@ ac_ct_CC = @ac_ct_CC@ ac_ct_DUMPBIN = @ac_ct_DUMPBIN@ ac_ct_F77 = @ac_ct_F77@ acx_pthread_config = @acx_pthread_config@ am__include = @am__include@ am__leading_dot = @am__leading_dot@ am__quote = @am__quote@ am__tar = @am__tar@ am__untar = @am__untar@ bindir = @bindir@ build = @build@ build_alias = @build_alias@ build_cpu = @build_cpu@ build_os = @build_os@ build_vendor = @build_vendor@ builddir = @builddir@ datadir = @datadir@ datarootdir = @datarootdir@ docdir = @docdir@ dvidir = @dvidir@ exec_prefix = @exec_prefix@ host = @host@ host_alias = @host_alias@ host_cpu = @host_cpu@ host_os = @host_os@ host_vendor = @host_vendor@ htmldir = @htmldir@ includedir = @includedir@ infodir = @infodir@ install_sh = @install_sh@ libdir = @libdir@ libexecdir = @libexecdir@ localedir = @localedir@ localstatedir = @localstatedir@ mandir = @mandir@ mkdir_p = @mkdir_p@ oldincludedir = @oldincludedir@ pdfdir = @pdfdir@ prefix = @prefix@ program_transform_name = @program_transform_name@ psdir = @psdir@ sbindir = @sbindir@ sharedstatedir = @sharedstatedir@ srcdir = @srcdir@ sysconfdir = @sysconfdir@ target_alias = @target_alias@ top_build_prefix = @top_build_prefix@ top_builddir = @top_builddir@ top_srcdir = @top_srcdir@ SUBDIRS = common sse2 avx altivec neon EXTRA_DIST = hc2cbv.h hc2cfv.h codlist.mk simd.mk all: all-recursive .SUFFIXES: $(srcdir)/Makefile.in: @MAINTAINER_MODE_TRUE@ $(srcdir)/Makefile.am $(am__configure_deps) @for dep in $?; do \ case '$(am__configure_deps)' in \ *$$dep*) \ ( cd $(top_builddir) && $(MAKE) $(AM_MAKEFLAGS) am--refresh ) \ && { if test -f $@; then exit 0; else break; fi; }; \ exit 1;; \ esac; \ done; \ echo ' cd $(top_srcdir) && $(AUTOMAKE) --gnu rdft/simd/Makefile'; \ $(am__cd) $(top_srcdir) && \ $(AUTOMAKE) --gnu rdft/simd/Makefile .PRECIOUS: Makefile Makefile: $(srcdir)/Makefile.in $(top_builddir)/config.status @case '$?' in \ *config.status*) \ cd $(top_builddir) && $(MAKE) $(AM_MAKEFLAGS) am--refresh;; \ *) \ echo ' cd $(top_builddir) && $(SHELL) ./config.status $(subdir)/$@ $(am__depfiles_maybe)'; \ cd $(top_builddir) && $(SHELL) ./config.status $(subdir)/$@ $(am__depfiles_maybe);; \ esac; $(top_builddir)/config.status: $(top_srcdir)/configure $(CONFIG_STATUS_DEPENDENCIES) cd $(top_builddir) && $(MAKE) $(AM_MAKEFLAGS) am--refresh $(top_srcdir)/configure: @MAINTAINER_MODE_TRUE@ $(am__configure_deps) cd $(top_builddir) && $(MAKE) $(AM_MAKEFLAGS) am--refresh $(ACLOCAL_M4): @MAINTAINER_MODE_TRUE@ $(am__aclocal_m4_deps) cd $(top_builddir) && $(MAKE) $(AM_MAKEFLAGS) am--refresh $(am__aclocal_m4_deps): mostlyclean-libtool: -rm -f *.lo clean-libtool: -rm -rf .libs _libs # This directory's subdirectories are mostly independent; you can cd # into them and run `make' without going through this Makefile. # To change the values of `make' variables: instead of editing Makefiles, # (1) if the variable is set in `config.status', edit `config.status' # (which will cause the Makefiles to be regenerated when you run `make'); # (2) otherwise, pass the desired values on the `make' command line. $(RECURSIVE_TARGETS): @fail= failcom='exit 1'; \ for f in x $$MAKEFLAGS; do \ case $$f in \ *=* | --[!k]*);; \ *k*) failcom='fail=yes';; \ esac; \ done; \ dot_seen=no; \ target=`echo $@ | sed s/-recursive//`; \ list='$(SUBDIRS)'; for subdir in $$list; do \ echo "Making $$target in $$subdir"; \ if test "$$subdir" = "."; then \ dot_seen=yes; \ local_target="$$target-am"; \ else \ local_target="$$target"; \ fi; \ ($(am__cd) $$subdir && $(MAKE) $(AM_MAKEFLAGS) $$local_target) \ || eval $$failcom; \ done; \ if test "$$dot_seen" = "no"; then \ $(MAKE) $(AM_MAKEFLAGS) "$$target-am" || exit 1; \ fi; test -z "$$fail" $(RECURSIVE_CLEAN_TARGETS): @fail= failcom='exit 1'; \ for f in x $$MAKEFLAGS; do \ case $$f in \ *=* | --[!k]*);; \ *k*) failcom='fail=yes';; \ esac; \ done; \ dot_seen=no; \ case "$@" in \ distclean-* | maintainer-clean-*) list='$(DIST_SUBDIRS)' ;; \ *) list='$(SUBDIRS)' ;; \ esac; \ rev=''; for subdir in $$list; do \ if test "$$subdir" = "."; then :; else \ rev="$$subdir $$rev"; \ fi; \ done; \ rev="$$rev ."; \ target=`echo $@ | sed s/-recursive//`; \ for subdir in $$rev; do \ echo "Making $$target in $$subdir"; \ if test "$$subdir" = "."; then \ local_target="$$target-am"; \ else \ local_target="$$target"; \ fi; \ ($(am__cd) $$subdir && $(MAKE) $(AM_MAKEFLAGS) $$local_target) \ || eval $$failcom; \ done && test -z "$$fail" tags-recursive: list='$(SUBDIRS)'; for subdir in $$list; do \ test "$$subdir" = . || ($(am__cd) $$subdir && $(MAKE) $(AM_MAKEFLAGS) tags); \ done ctags-recursive: list='$(SUBDIRS)'; for subdir in $$list; do \ test "$$subdir" = . || ($(am__cd) $$subdir && $(MAKE) $(AM_MAKEFLAGS) ctags); \ done ID: $(HEADERS) $(SOURCES) $(LISP) $(TAGS_FILES) list='$(SOURCES) $(HEADERS) $(LISP) $(TAGS_FILES)'; \ unique=`for i in $$list; do \ if test -f "$$i"; then echo $$i; else echo $(srcdir)/$$i; fi; \ done | \ $(AWK) '{ files[$$0] = 1; nonempty = 1; } \ END { if (nonempty) { for (i in files) print i; }; }'`; \ mkid -fID $$unique tags: TAGS TAGS: tags-recursive $(HEADERS) $(SOURCES) $(TAGS_DEPENDENCIES) \ $(TAGS_FILES) $(LISP) set x; \ here=`pwd`; \ if ($(ETAGS) --etags-include --version) >/dev/null 2>&1; then \ include_option=--etags-include; \ empty_fix=.; \ else \ include_option=--include; \ empty_fix=; \ fi; \ list='$(SUBDIRS)'; for subdir in $$list; do \ if test "$$subdir" = .; then :; else \ test ! -f $$subdir/TAGS || \ set "$$@" "$$include_option=$$here/$$subdir/TAGS"; \ fi; \ done; \ list='$(SOURCES) $(HEADERS) $(LISP) $(TAGS_FILES)'; \ unique=`for i in $$list; do \ if test -f "$$i"; then echo $$i; else echo $(srcdir)/$$i; fi; \ done | \ $(AWK) '{ files[$$0] = 1; nonempty = 1; } \ END { if (nonempty) { for (i in files) print i; }; }'`; \ shift; \ if test -z "$(ETAGS_ARGS)$$*$$unique"; then :; else \ test -n "$$unique" || unique=$$empty_fix; \ if test $$# -gt 0; then \ $(ETAGS) $(ETAGSFLAGS) $(AM_ETAGSFLAGS) $(ETAGS_ARGS) \ "$$@" $$unique; \ else \ $(ETAGS) $(ETAGSFLAGS) $(AM_ETAGSFLAGS) $(ETAGS_ARGS) \ $$unique; \ fi; \ fi ctags: CTAGS CTAGS: ctags-recursive $(HEADERS) $(SOURCES) $(TAGS_DEPENDENCIES) \ $(TAGS_FILES) $(LISP) list='$(SOURCES) $(HEADERS) $(LISP) $(TAGS_FILES)'; \ unique=`for i in $$list; do \ if test -f "$$i"; then echo $$i; else echo $(srcdir)/$$i; fi; \ done | \ $(AWK) '{ files[$$0] = 1; nonempty = 1; } \ END { if (nonempty) { for (i in files) print i; }; }'`; \ test -z "$(CTAGS_ARGS)$$unique" \ || $(CTAGS) $(CTAGSFLAGS) $(AM_CTAGSFLAGS) $(CTAGS_ARGS) \ $$unique GTAGS: here=`$(am__cd) $(top_builddir) && pwd` \ && $(am__cd) $(top_srcdir) \ && gtags -i $(GTAGS_ARGS) "$$here" distclean-tags: -rm -f TAGS ID GTAGS GRTAGS GSYMS GPATH tags distdir: $(DISTFILES) @srcdirstrip=`echo "$(srcdir)" | sed 's/[].[^$$\\*]/\\\\&/g'`; \ topsrcdirstrip=`echo "$(top_srcdir)" | sed 's/[].[^$$\\*]/\\\\&/g'`; \ list='$(DISTFILES)'; \ dist_files=`for file in $$list; do echo $$file; done | \ sed -e "s|^$$srcdirstrip/||;t" \ -e "s|^$$topsrcdirstrip/|$(top_builddir)/|;t"`; \ case $$dist_files in \ */*) $(MKDIR_P) `echo "$$dist_files" | \ sed '/\//!d;s|^|$(distdir)/|;s,/[^/]*$$,,' | \ sort -u` ;; \ esac; \ for file in $$dist_files; do \ if test -f $$file || test -d $$file; then d=.; else d=$(srcdir); fi; \ if test -d $$d/$$file; then \ dir=`echo "/$$file" | sed -e 's,/[^/]*$$,,'`; \ if test -d "$(distdir)/$$file"; then \ find "$(distdir)/$$file" -type d ! -perm -700 -exec chmod u+rwx {} \;; \ fi; \ if test -d $(srcdir)/$$file && test $$d != $(srcdir); then \ cp -fpR $(srcdir)/$$file "$(distdir)$$dir" || exit 1; \ find "$(distdir)/$$file" -type d ! -perm -700 -exec chmod u+rwx {} \;; \ fi; \ cp -fpR $$d/$$file "$(distdir)$$dir" || exit 1; \ else \ test -f "$(distdir)/$$file" \ || cp -p $$d/$$file "$(distdir)/$$file" \ || exit 1; \ fi; \ done @list='$(DIST_SUBDIRS)'; for subdir in $$list; do \ if test "$$subdir" = .; then :; else \ $(am__make_dryrun) \ || test -d "$(distdir)/$$subdir" \ || $(MKDIR_P) "$(distdir)/$$subdir" \ || exit 1; \ dir1=$$subdir; dir2="$(distdir)/$$subdir"; \ $(am__relativize); \ new_distdir=$$reldir; \ dir1=$$subdir; dir2="$(top_distdir)"; \ $(am__relativize); \ new_top_distdir=$$reldir; \ echo " (cd $$subdir && $(MAKE) $(AM_MAKEFLAGS) top_distdir="$$new_top_distdir" distdir="$$new_distdir" \\"; \ echo " am__remove_distdir=: am__skip_length_check=: am__skip_mode_fix=: distdir)"; \ ($(am__cd) $$subdir && \ $(MAKE) $(AM_MAKEFLAGS) \ top_distdir="$$new_top_distdir" \ distdir="$$new_distdir" \ am__remove_distdir=: \ am__skip_length_check=: \ am__skip_mode_fix=: \ distdir) \ || exit 1; \ fi; \ done check-am: all-am check: check-recursive all-am: Makefile installdirs: installdirs-recursive installdirs-am: install: install-recursive install-exec: install-exec-recursive install-data: install-data-recursive uninstall: uninstall-recursive install-am: all-am @$(MAKE) $(AM_MAKEFLAGS) install-exec-am install-data-am installcheck: installcheck-recursive install-strip: if test -z '$(STRIP)'; then \ $(MAKE) $(AM_MAKEFLAGS) INSTALL_PROGRAM="$(INSTALL_STRIP_PROGRAM)" \ install_sh_PROGRAM="$(INSTALL_STRIP_PROGRAM)" INSTALL_STRIP_FLAG=-s \ install; \ else \ $(MAKE) $(AM_MAKEFLAGS) INSTALL_PROGRAM="$(INSTALL_STRIP_PROGRAM)" \ install_sh_PROGRAM="$(INSTALL_STRIP_PROGRAM)" INSTALL_STRIP_FLAG=-s \ "INSTALL_PROGRAM_ENV=STRIPPROG='$(STRIP)'" install; \ fi mostlyclean-generic: clean-generic: distclean-generic: -test -z "$(CONFIG_CLEAN_FILES)" || rm -f $(CONFIG_CLEAN_FILES) -test . = "$(srcdir)" || test -z "$(CONFIG_CLEAN_VPATH_FILES)" || rm -f $(CONFIG_CLEAN_VPATH_FILES) maintainer-clean-generic: @echo "This command is intended for maintainers to use" @echo "it deletes files that may require special tools to rebuild." clean: clean-recursive clean-am: clean-generic clean-libtool mostlyclean-am distclean: distclean-recursive -rm -f Makefile distclean-am: clean-am distclean-generic distclean-tags dvi: dvi-recursive dvi-am: html: html-recursive html-am: info: info-recursive info-am: install-data-am: install-dvi: install-dvi-recursive install-dvi-am: install-exec-am: install-html: install-html-recursive install-html-am: install-info: install-info-recursive install-info-am: install-man: install-pdf: install-pdf-recursive install-pdf-am: install-ps: install-ps-recursive install-ps-am: installcheck-am: maintainer-clean: maintainer-clean-recursive -rm -f Makefile maintainer-clean-am: distclean-am maintainer-clean-generic mostlyclean: mostlyclean-recursive mostlyclean-am: mostlyclean-generic mostlyclean-libtool pdf: pdf-recursive pdf-am: ps: ps-recursive ps-am: uninstall-am: .MAKE: $(RECURSIVE_CLEAN_TARGETS) $(RECURSIVE_TARGETS) ctags-recursive \ install-am install-strip tags-recursive .PHONY: $(RECURSIVE_CLEAN_TARGETS) $(RECURSIVE_TARGETS) CTAGS GTAGS \ all all-am check check-am clean clean-generic clean-libtool \ ctags ctags-recursive distclean distclean-generic \ distclean-libtool distclean-tags distdir dvi dvi-am html \ html-am info info-am install install-am install-data \ install-data-am install-dvi install-dvi-am install-exec \ install-exec-am install-html install-html-am install-info \ install-info-am install-man install-pdf install-pdf-am \ install-ps install-ps-am install-strip installcheck \ installcheck-am installdirs installdirs-am maintainer-clean \ maintainer-clean-generic mostlyclean mostlyclean-generic \ mostlyclean-libtool pdf pdf-am ps ps-am tags tags-recursive \ uninstall uninstall-am # Tell versions [3.59,3.63) of GNU make to not export all variables. # Otherwise a system limit (for SysV at least) may be exceeded. .NOEXPORT: fftw-3.3.3/rdft/simd/hc2cbv.h0000644000175000017500000000175212054410050012616 00000000000000/* * Copyright (c) 2003, 2007-11 Matteo Frigo * Copyright (c) 2003, 2007-11 Massachusetts Institute of Technology * * This program is free software; you can redistribute it and/or modify * it under the terms of the GNU General Public License as published by * the Free Software Foundation; either version 2 of the License, or * (at your option) any later version. * * This program is distributed in the hope that it will be useful, * but WITHOUT ANY WARRANTY; without even the implied warranty of * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the * GNU General Public License for more details. * * You should have received a copy of the GNU General Public License * along with this program; if not, write to the Free Software * Foundation, Inc., 51 Franklin Street, Fifth Floor, Boston, MA 02110-1301 USA * */ #include SIMD_HEADER #define VTW VTW3 #define TWVL TWVL3 #define LDW(x) LDA(x, 0, 0) #define GENUS XSIMD(rdft_hc2cbv_genus) extern const hc2c_genus GENUS; fftw-3.3.3/rdft/simd/hc2cfv.h0000644000175000017500000000175212054410050012622 00000000000000/* * Copyright (c) 2003, 2007-11 Matteo Frigo * Copyright (c) 2003, 2007-11 Massachusetts Institute of Technology * * This program is free software; you can redistribute it and/or modify * it under the terms of the GNU General Public License as published by * the Free Software Foundation; either version 2 of the License, or * (at your option) any later version. * * This program is distributed in the hope that it will be useful, * but WITHOUT ANY WARRANTY; without even the implied warranty of * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the * GNU General Public License for more details. * * You should have received a copy of the GNU General Public License * along with this program; if not, write to the Free Software * Foundation, Inc., 51 Franklin Street, Fifth Floor, Boston, MA 02110-1301 USA * */ #include SIMD_HEADER #define VTW VTW3 #define TWVL TWVL3 #define LDW(x) LDA(x, 0, 0) #define GENUS XSIMD(rdft_hc2cfv_genus) extern const hc2c_genus GENUS; fftw-3.3.3/rdft/simd/codlist.mk0000644000175000017500000000200112054410050013254 00000000000000# This file contains a standard list of RDFT SIMD codelets. It is # included by common/Makefile to generate the C files with the actual # codelets in them. It is included by {sse,sse2,...}/Makefile to # generate and compile stub files that include common/*.c # You can customize FFTW for special needs, e.g. to handle certain # sizes more efficiently, by adding new codelets to the lists of those # included by default. If you change the list of codelets, any new # ones you added will be automatically generated when you run the # bootstrap script (see "Generating your own code" in the FFTW # manual). HC2CFDFTV = hc2cfdftv_2.c hc2cfdftv_4.c hc2cfdftv_6.c hc2cfdftv_8.c \ hc2cfdftv_10.c hc2cfdftv_12.c hc2cfdftv_16.c hc2cfdftv_32.c \ hc2cfdftv_20.c HC2CBDFTV = hc2cbdftv_2.c hc2cbdftv_4.c hc2cbdftv_6.c hc2cbdftv_8.c \ hc2cbdftv_10.c hc2cbdftv_12.c hc2cbdftv_16.c hc2cbdftv_32.c \ hc2cbdftv_20.c ########################################################################### SIMD_CODELETS = $(HC2CFDFTV) $(HC2CBDFTV) fftw-3.3.3/rdft/simd/simd.mk0000644000175000017500000000054612054410050012563 00000000000000AM_CPPFLAGS = -I$(top_srcdir)/kernel -I$(top_srcdir)/rdft \ -I$(top_srcdir)/rdft/simd -I$(top_srcdir)/simd-support EXTRA_DIST = $(SIMD_CODELETS) genus.c codlist.c $(EXTRA_DIST): Makefile ( \ echo "/* Generated automatically. DO NOT EDIT! */"; \ echo "#define SIMD_HEADER \"$(SIMD_HEADER)\""; \ echo "#include \"../common/"$*".c\""; \ ) >$@ fftw-3.3.3/rdft/simd/common/0000755000175000017500000000000012054411110012637 500000000000000fftw-3.3.3/rdft/simd/common/Makefile.am0000644000175000017500000000165212054410050014621 00000000000000# include the list of codelets include $(top_srcdir)/rdft/simd/codlist.mk ALL_CODELETS = $(SIMD_CODELETS) BUILT_SOURCES= $(SIMD_CODELETS) $(CODLIST) EXTRA_DIST = $(BUILT_SOURCES) genus.c INCLUDE_SIMD_HEADER="\#include SIMD_HEADER" XRENAME=XSIMD SOLVTAB_NAME = XSIMD(solvtab_rdft) # include special rules for regenerating codelets. include $(top_srcdir)/support/Makefile.codelets if MAINTAINER_MODE FLAGS_HC2C=-simd $(FLAGS_COMMON) -pipeline-latency 8 -trivial-stores -variables 32 -no-generate-bytw hc2cfdftv_%.c: $(CODELET_DEPS) $(GEN_HC2CDFT_C) ($(PRELUDE_COMMANDS_RDFT); $(TWOVERS) $(GEN_HC2CDFT_C) $(FLAGS_HC2C) -n $* -dit -name hc2cfdftv_$* -include "hc2cfv.h") | $(ADD_DATE) | $(INDENT) >$@ hc2cbdftv_%.c: $(CODELET_DEPS) $(GEN_HC2CDFT_C) ($(PRELUDE_COMMANDS_RDFT); $(TWOVERS) $(GEN_HC2CDFT_C) $(FLAGS_HC2C) -n $* -dif -sign 1 -name hc2cbdftv_$* -include "hc2cbv.h") | $(ADD_DATE) | $(INDENT) >$@ endif # MAINTAINER_MODE fftw-3.3.3/rdft/simd/common/Makefile.in0000644000175000017500000004116412054410200014631 00000000000000# Makefile.in generated by automake 1.11.6 from Makefile.am. # @configure_input@ # Copyright (C) 1994, 1995, 1996, 1997, 1998, 1999, 2000, 2001, 2002, # 2003, 2004, 2005, 2006, 2007, 2008, 2009, 2010, 2011 Free Software # Foundation, Inc. # This Makefile.in is free software; the Free Software Foundation # gives unlimited permission to copy and/or distribute it, # with or without modifications, as long as this notice is preserved. # This program is distributed in the hope that it will be useful, # but WITHOUT ANY WARRANTY, to the extent permitted by law; without # even the implied warranty of MERCHANTABILITY or FITNESS FOR A # PARTICULAR PURPOSE. @SET_MAKE@ # include the list of codelets # This file contains a standard list of RDFT SIMD codelets. It is # included by common/Makefile to generate the C files with the actual # codelets in them. It is included by {sse,sse2,...}/Makefile to # generate and compile stub files that include common/*.c # You can customize FFTW for special needs, e.g. to handle certain # sizes more efficiently, by adding new codelets to the lists of those # included by default. If you change the list of codelets, any new # ones you added will be automatically generated when you run the # bootstrap script (see "Generating your own code" in the FFTW # manual). # -*- makefile -*- # This file contains special make rules to generate codelets. # Most of this file requires GNU make . VPATH = @srcdir@ am__make_dryrun = \ { \ am__dry=no; \ case $$MAKEFLAGS in \ *\\[\ \ ]*) \ echo 'am--echo: ; @echo "AM" OK' | $(MAKE) -f - 2>/dev/null \ | grep '^AM OK$$' >/dev/null || am__dry=yes;; \ *) \ for am__flg in $$MAKEFLAGS; do \ case $$am__flg in \ *=*|--*) ;; \ *n*) am__dry=yes; break;; \ esac; \ done;; \ esac; \ test $$am__dry = yes; \ } pkgdatadir = $(datadir)/@PACKAGE@ pkgincludedir = $(includedir)/@PACKAGE@ pkglibdir = $(libdir)/@PACKAGE@ pkglibexecdir = $(libexecdir)/@PACKAGE@ am__cd = CDPATH="$${ZSH_VERSION+.}$(PATH_SEPARATOR)" && cd install_sh_DATA = $(install_sh) -c -m 644 install_sh_PROGRAM = $(install_sh) -c install_sh_SCRIPT = $(install_sh) -c INSTALL_HEADER = $(INSTALL_DATA) transform = $(program_transform_name) NORMAL_INSTALL = : PRE_INSTALL = : POST_INSTALL = : NORMAL_UNINSTALL = : PRE_UNINSTALL = : POST_UNINSTALL = : build_triplet = @build@ host_triplet = @host@ DIST_COMMON = $(srcdir)/Makefile.am $(srcdir)/Makefile.in \ $(top_srcdir)/rdft/simd/codlist.mk \ $(top_srcdir)/support/Makefile.codelets subdir = rdft/simd/common ACLOCAL_M4 = $(top_srcdir)/aclocal.m4 am__aclocal_m4_deps = $(top_srcdir)/m4/acx_mpi.m4 \ $(top_srcdir)/m4/acx_pthread.m4 \ $(top_srcdir)/m4/ax_cc_maxopt.m4 \ $(top_srcdir)/m4/ax_check_compiler_flags.m4 \ $(top_srcdir)/m4/ax_compiler_vendor.m4 \ $(top_srcdir)/m4/ax_gcc_aligns_stack.m4 \ $(top_srcdir)/m4/ax_gcc_version.m4 \ $(top_srcdir)/m4/ax_openmp.m4 $(top_srcdir)/m4/libtool.m4 \ $(top_srcdir)/m4/ltoptions.m4 $(top_srcdir)/m4/ltsugar.m4 \ $(top_srcdir)/m4/ltversion.m4 $(top_srcdir)/m4/lt~obsolete.m4 \ $(top_srcdir)/configure.ac am__configure_deps = $(am__aclocal_m4_deps) $(CONFIGURE_DEPENDENCIES) \ $(ACLOCAL_M4) mkinstalldirs = $(install_sh) -d CONFIG_HEADER = $(top_builddir)/config.h CONFIG_CLEAN_FILES = CONFIG_CLEAN_VPATH_FILES = SOURCES = DIST_SOURCES = am__can_run_installinfo = \ case $$AM_UPDATE_INFO_DIR in \ n|no|NO) false;; \ *) (install-info --version) >/dev/null 2>&1;; \ esac DISTFILES = $(DIST_COMMON) $(DIST_SOURCES) $(TEXINFOS) $(EXTRA_DIST) ACLOCAL = @ACLOCAL@ ALLOCA = @ALLOCA@ ALTIVEC_CFLAGS = @ALTIVEC_CFLAGS@ AMTAR = @AMTAR@ AR = @AR@ AS = @AS@ AUTOCONF = @AUTOCONF@ AUTOHEADER = @AUTOHEADER@ AUTOMAKE = @AUTOMAKE@ AVX_CFLAGS = @AVX_CFLAGS@ AWK = @AWK@ CC = @CC@ CCDEPMODE = @CCDEPMODE@ CFLAGS = @CFLAGS@ CHECK_PL_OPTS = @CHECK_PL_OPTS@ CPP = @CPP@ CPPFLAGS = @CPPFLAGS@ CYGPATH_W = @CYGPATH_W@ C_FFTW_R2R_KIND = @C_FFTW_R2R_KIND@ C_MPI_FINT = @C_MPI_FINT@ DEFS = @DEFS@ DEPDIR = @DEPDIR@ DLLTOOL = @DLLTOOL@ DSYMUTIL = @DSYMUTIL@ DUMPBIN = @DUMPBIN@ ECHO_C = @ECHO_C@ ECHO_N = @ECHO_N@ ECHO_T = @ECHO_T@ EGREP = @EGREP@ EXEEXT = @EXEEXT@ F77 = @F77@ FFLAGS = @FFLAGS@ FGREP = @FGREP@ FLIBS = @FLIBS@ GREP = @GREP@ INSTALL = @INSTALL@ INSTALL_DATA = @INSTALL_DATA@ INSTALL_PROGRAM = @INSTALL_PROGRAM@ INSTALL_SCRIPT = @INSTALL_SCRIPT@ INSTALL_STRIP_PROGRAM = @INSTALL_STRIP_PROGRAM@ LD = @LD@ LDFLAGS = @LDFLAGS@ LIBOBJS = @LIBOBJS@ LIBQUADMATH = @LIBQUADMATH@ LIBS = @LIBS@ LIBTOOL = @LIBTOOL@ LIPO = @LIPO@ LN_S = @LN_S@ LTLIBOBJS = @LTLIBOBJS@ MAINT = @MAINT@ MAKEINFO = @MAKEINFO@ MANIFEST_TOOL = @MANIFEST_TOOL@ MKDIR_P = @MKDIR_P@ MPICC = @MPICC@ MPILIBS = @MPILIBS@ MPIRUN = @MPIRUN@ NEON_CFLAGS = @NEON_CFLAGS@ NM = @NM@ NMEDIT = @NMEDIT@ OBJDUMP = @OBJDUMP@ OBJEXT = @OBJEXT@ OCAMLBUILD = @OCAMLBUILD@ OPENMP_CFLAGS = @OPENMP_CFLAGS@ OTOOL = @OTOOL@ OTOOL64 = @OTOOL64@ PACKAGE = @PACKAGE@ PACKAGE_BUGREPORT = @PACKAGE_BUGREPORT@ PACKAGE_NAME = @PACKAGE_NAME@ PACKAGE_STRING = @PACKAGE_STRING@ PACKAGE_TARNAME = @PACKAGE_TARNAME@ PACKAGE_URL = @PACKAGE_URL@ PACKAGE_VERSION = @PACKAGE_VERSION@ PATH_SEPARATOR = @PATH_SEPARATOR@ POW_LIB = @POW_LIB@ PRECISION = @PRECISION@ PREC_SUFFIX = @PREC_SUFFIX@ PTHREAD_CC = @PTHREAD_CC@ PTHREAD_CFLAGS = @PTHREAD_CFLAGS@ PTHREAD_LIBS = @PTHREAD_LIBS@ RANLIB = @RANLIB@ SED = @SED@ SET_MAKE = @SET_MAKE@ SHARED_VERSION_INFO = @SHARED_VERSION_INFO@ SHELL = @SHELL@ SSE2_CFLAGS = @SSE2_CFLAGS@ STACK_ALIGN_CFLAGS = @STACK_ALIGN_CFLAGS@ STRIP = @STRIP@ THREADLIBS = @THREADLIBS@ VERSION = @VERSION@ abs_builddir = @abs_builddir@ abs_srcdir = @abs_srcdir@ abs_top_builddir = @abs_top_builddir@ abs_top_srcdir = @abs_top_srcdir@ ac_ct_AR = @ac_ct_AR@ ac_ct_CC = @ac_ct_CC@ ac_ct_DUMPBIN = @ac_ct_DUMPBIN@ ac_ct_F77 = @ac_ct_F77@ acx_pthread_config = @acx_pthread_config@ am__include = @am__include@ am__leading_dot = @am__leading_dot@ am__quote = @am__quote@ am__tar = @am__tar@ am__untar = @am__untar@ bindir = @bindir@ build = @build@ build_alias = @build_alias@ build_cpu = @build_cpu@ build_os = @build_os@ build_vendor = @build_vendor@ builddir = @builddir@ datadir = @datadir@ datarootdir = @datarootdir@ docdir = @docdir@ dvidir = @dvidir@ exec_prefix = @exec_prefix@ host = @host@ host_alias = @host_alias@ host_cpu = @host_cpu@ host_os = @host_os@ host_vendor = @host_vendor@ htmldir = @htmldir@ includedir = @includedir@ infodir = @infodir@ install_sh = @install_sh@ libdir = @libdir@ libexecdir = @libexecdir@ localedir = @localedir@ localstatedir = @localstatedir@ mandir = @mandir@ mkdir_p = @mkdir_p@ oldincludedir = @oldincludedir@ pdfdir = @pdfdir@ prefix = @prefix@ program_transform_name = @program_transform_name@ psdir = @psdir@ sbindir = @sbindir@ sharedstatedir = @sharedstatedir@ srcdir = @srcdir@ sysconfdir = @sysconfdir@ target_alias = @target_alias@ top_build_prefix = @top_build_prefix@ top_builddir = @top_builddir@ top_srcdir = @top_srcdir@ HC2CFDFTV = hc2cfdftv_2.c hc2cfdftv_4.c hc2cfdftv_6.c hc2cfdftv_8.c \ hc2cfdftv_10.c hc2cfdftv_12.c hc2cfdftv_16.c hc2cfdftv_32.c \ hc2cfdftv_20.c HC2CBDFTV = hc2cbdftv_2.c hc2cbdftv_4.c hc2cbdftv_6.c hc2cbdftv_8.c \ hc2cbdftv_10.c hc2cbdftv_12.c hc2cbdftv_16.c hc2cbdftv_32.c \ hc2cbdftv_20.c ########################################################################### SIMD_CODELETS = $(HC2CFDFTV) $(HC2CBDFTV) ALL_CODELETS = $(SIMD_CODELETS) BUILT_SOURCES = $(SIMD_CODELETS) $(CODLIST) EXTRA_DIST = $(BUILT_SOURCES) genus.c INCLUDE_SIMD_HEADER = "\#include SIMD_HEADER" XRENAME = XSIMD SOLVTAB_NAME = XSIMD(solvtab_rdft) CODLIST = codlist.c CODELET_NAME = codelet_ @MAINTAINER_MODE_TRUE@INDENT = indent -kr -cs -i5 -l800 -fca -nfc1 -sc -sob -cli4 -TR -Tplanner -TV @MAINTAINER_MODE_TRUE@TWOVERS = sh ${top_srcdir}/support/twovers.sh @MAINTAINER_MODE_TRUE@GENFFTDIR = ${top_builddir}/genfft @MAINTAINER_MODE_TRUE@GEN_NOTW = ${GENFFTDIR}/gen_notw.native @MAINTAINER_MODE_TRUE@GEN_NOTW_C = ${GENFFTDIR}/gen_notw_c.native @MAINTAINER_MODE_TRUE@GEN_TWIDDLE = ${GENFFTDIR}/gen_twiddle.native @MAINTAINER_MODE_TRUE@GEN_TWIDDLE_C = ${GENFFTDIR}/gen_twiddle_c.native @MAINTAINER_MODE_TRUE@GEN_TWIDSQ = ${GENFFTDIR}/gen_twidsq.native @MAINTAINER_MODE_TRUE@GEN_TWIDSQ_C = ${GENFFTDIR}/gen_twidsq_c.native @MAINTAINER_MODE_TRUE@GEN_R2CF = ${GENFFTDIR}/gen_r2cf.native @MAINTAINER_MODE_TRUE@GEN_R2CB = ${GENFFTDIR}/gen_r2cb.native @MAINTAINER_MODE_TRUE@GEN_HC2HC = ${GENFFTDIR}/gen_hc2hc.native @MAINTAINER_MODE_TRUE@GEN_HC2C = ${GENFFTDIR}/gen_hc2c.native @MAINTAINER_MODE_TRUE@GEN_HC2CDFT = ${GENFFTDIR}/gen_hc2cdft.native @MAINTAINER_MODE_TRUE@GEN_HC2CDFT_C = ${GENFFTDIR}/gen_hc2cdft_c.native @MAINTAINER_MODE_TRUE@GEN_R2R = ${GENFFTDIR}/gen_r2r.native @MAINTAINER_MODE_TRUE@PRELUDE_DFT = ${top_srcdir}/support/codelet_prelude.dft @MAINTAINER_MODE_TRUE@PRELUDE_RDFT = ${top_srcdir}/support/codelet_prelude.rdft @MAINTAINER_MODE_TRUE@ADD_DATE = sed -e s/@DATE@/"`date`"/ @MAINTAINER_MODE_TRUE@COPYRIGHT = ${top_srcdir}/COPYRIGHT @MAINTAINER_MODE_TRUE@CODELET_DEPS = $(COPYRIGHT) $(PRELUDE) @MAINTAINER_MODE_TRUE@PRELUDE_COMMANDS_DFT = cat $(COPYRIGHT) $(PRELUDE_DFT) @MAINTAINER_MODE_TRUE@PRELUDE_COMMANDS_RDFT = cat $(COPYRIGHT) $(PRELUDE_RDFT) @MAINTAINER_MODE_TRUE@FLAGS_COMMON = -compact -variables 4 @MAINTAINER_MODE_TRUE@DFT_FLAGS_COMMON = $(FLAGS_COMMON) -pipeline-latency 4 @MAINTAINER_MODE_TRUE@RDFT_FLAGS_COMMON = $(FLAGS_COMMON) -pipeline-latency 4 # include special rules for regenerating codelets. @MAINTAINER_MODE_TRUE@FLAGS_HC2C = -simd $(FLAGS_COMMON) -pipeline-latency 8 -trivial-stores -variables 32 -no-generate-bytw all: $(BUILT_SOURCES) $(MAKE) $(AM_MAKEFLAGS) all-am .SUFFIXES: $(srcdir)/Makefile.in: @MAINTAINER_MODE_TRUE@ $(srcdir)/Makefile.am $(top_srcdir)/rdft/simd/codlist.mk $(top_srcdir)/support/Makefile.codelets $(am__configure_deps) @for dep in $?; do \ case '$(am__configure_deps)' in \ *$$dep*) \ ( cd $(top_builddir) && $(MAKE) $(AM_MAKEFLAGS) am--refresh ) \ && { if test -f $@; then exit 0; else break; fi; }; \ exit 1;; \ esac; \ done; \ echo ' cd $(top_srcdir) && $(AUTOMAKE) --gnu rdft/simd/common/Makefile'; \ $(am__cd) $(top_srcdir) && \ $(AUTOMAKE) --gnu rdft/simd/common/Makefile .PRECIOUS: Makefile Makefile: $(srcdir)/Makefile.in $(top_builddir)/config.status @case '$?' in \ *config.status*) \ cd $(top_builddir) && $(MAKE) $(AM_MAKEFLAGS) am--refresh;; \ *) \ echo ' cd $(top_builddir) && $(SHELL) ./config.status $(subdir)/$@ $(am__depfiles_maybe)'; \ cd $(top_builddir) && $(SHELL) ./config.status $(subdir)/$@ $(am__depfiles_maybe);; \ esac; $(top_srcdir)/rdft/simd/codlist.mk $(top_srcdir)/support/Makefile.codelets: $(top_builddir)/config.status: $(top_srcdir)/configure $(CONFIG_STATUS_DEPENDENCIES) cd $(top_builddir) && $(MAKE) $(AM_MAKEFLAGS) am--refresh $(top_srcdir)/configure: @MAINTAINER_MODE_TRUE@ $(am__configure_deps) cd $(top_builddir) && $(MAKE) $(AM_MAKEFLAGS) am--refresh $(ACLOCAL_M4): @MAINTAINER_MODE_TRUE@ $(am__aclocal_m4_deps) cd $(top_builddir) && $(MAKE) $(AM_MAKEFLAGS) am--refresh $(am__aclocal_m4_deps): mostlyclean-libtool: -rm -f *.lo clean-libtool: -rm -rf .libs _libs tags: TAGS TAGS: ctags: CTAGS CTAGS: distdir: $(DISTFILES) @srcdirstrip=`echo "$(srcdir)" | sed 's/[].[^$$\\*]/\\\\&/g'`; \ topsrcdirstrip=`echo "$(top_srcdir)" | sed 's/[].[^$$\\*]/\\\\&/g'`; \ list='$(DISTFILES)'; \ dist_files=`for file in $$list; do echo $$file; done | \ sed -e "s|^$$srcdirstrip/||;t" \ -e "s|^$$topsrcdirstrip/|$(top_builddir)/|;t"`; \ case $$dist_files in \ */*) $(MKDIR_P) `echo "$$dist_files" | \ sed '/\//!d;s|^|$(distdir)/|;s,/[^/]*$$,,' | \ sort -u` ;; \ esac; \ for file in $$dist_files; do \ if test -f $$file || test -d $$file; then d=.; else d=$(srcdir); fi; \ if test -d $$d/$$file; then \ dir=`echo "/$$file" | sed -e 's,/[^/]*$$,,'`; \ if test -d "$(distdir)/$$file"; then \ find "$(distdir)/$$file" -type d ! -perm -700 -exec chmod u+rwx {} \;; \ fi; \ if test -d $(srcdir)/$$file && test $$d != $(srcdir); then \ cp -fpR $(srcdir)/$$file "$(distdir)$$dir" || exit 1; \ find "$(distdir)/$$file" -type d ! -perm -700 -exec chmod u+rwx {} \;; \ fi; \ cp -fpR $$d/$$file "$(distdir)$$dir" || exit 1; \ else \ test -f "$(distdir)/$$file" \ || cp -p $$d/$$file "$(distdir)/$$file" \ || exit 1; \ fi; \ done check-am: all-am check: $(BUILT_SOURCES) $(MAKE) $(AM_MAKEFLAGS) check-am all-am: Makefile installdirs: install: $(BUILT_SOURCES) $(MAKE) $(AM_MAKEFLAGS) install-am install-exec: install-exec-am install-data: install-data-am uninstall: uninstall-am install-am: all-am @$(MAKE) $(AM_MAKEFLAGS) install-exec-am install-data-am installcheck: installcheck-am install-strip: if test -z '$(STRIP)'; then \ $(MAKE) $(AM_MAKEFLAGS) INSTALL_PROGRAM="$(INSTALL_STRIP_PROGRAM)" \ install_sh_PROGRAM="$(INSTALL_STRIP_PROGRAM)" INSTALL_STRIP_FLAG=-s \ install; \ else \ $(MAKE) $(AM_MAKEFLAGS) INSTALL_PROGRAM="$(INSTALL_STRIP_PROGRAM)" \ install_sh_PROGRAM="$(INSTALL_STRIP_PROGRAM)" INSTALL_STRIP_FLAG=-s \ "INSTALL_PROGRAM_ENV=STRIPPROG='$(STRIP)'" install; \ fi mostlyclean-generic: clean-generic: distclean-generic: -test -z "$(CONFIG_CLEAN_FILES)" || rm -f $(CONFIG_CLEAN_FILES) -test . = "$(srcdir)" || test -z "$(CONFIG_CLEAN_VPATH_FILES)" || rm -f $(CONFIG_CLEAN_VPATH_FILES) maintainer-clean-generic: @echo "This command is intended for maintainers to use" @echo "it deletes files that may require special tools to rebuild." -test -z "$(BUILT_SOURCES)" || rm -f $(BUILT_SOURCES) clean: clean-am clean-am: clean-generic clean-libtool mostlyclean-am distclean: distclean-am -rm -f Makefile distclean-am: clean-am distclean-generic dvi: dvi-am dvi-am: html: html-am html-am: info: info-am info-am: install-data-am: install-dvi: install-dvi-am install-dvi-am: install-exec-am: install-html: install-html-am install-html-am: install-info: install-info-am install-info-am: install-man: install-pdf: install-pdf-am install-pdf-am: install-ps: install-ps-am install-ps-am: installcheck-am: maintainer-clean: maintainer-clean-am -rm -f Makefile maintainer-clean-am: distclean-am maintainer-clean-generic \ maintainer-clean-local mostlyclean: mostlyclean-am mostlyclean-am: mostlyclean-generic mostlyclean-libtool pdf: pdf-am pdf-am: ps: ps-am ps-am: uninstall-am: .MAKE: all check install install-am install-strip .PHONY: all all-am check check-am clean clean-generic clean-libtool \ distclean distclean-generic distclean-libtool distdir dvi \ dvi-am html html-am info info-am install install-am \ install-data install-data-am install-dvi install-dvi-am \ install-exec install-exec-am install-html install-html-am \ install-info install-info-am install-man install-pdf \ install-pdf-am install-ps install-ps-am install-strip \ installcheck installcheck-am installdirs maintainer-clean \ maintainer-clean-generic maintainer-clean-local mostlyclean \ mostlyclean-generic mostlyclean-libtool pdf pdf-am ps ps-am \ uninstall uninstall-am # rule to build codlist $(CODLIST): Makefile ( \ echo "#include \"ifftw.h\""; \ echo $(INCLUDE_SIMD_HEADER); \ echo; \ for i in $(ALL_CODELETS) NIL; do \ if test "$$i" != NIL; then \ j=`basename $$i | sed -e 's/[.][cS]$$//g'`; \ echo "extern void $(XRENAME)($(CODELET_NAME)$$j)(planner *);"; \ fi \ done; \ echo; \ echo; \ echo "extern const solvtab $(SOLVTAB_NAME);"; \ echo "const solvtab $(SOLVTAB_NAME) = {"; \ for i in $(ALL_CODELETS) NIL; do \ if test "$$i" != NIL; then \ j=`basename $$i | sed -e 's/[.][cS]$$//g'`; \ echo " SOLVTAB($(XRENAME)($(CODELET_NAME)$$j)),"; \ fi \ done; \ echo " SOLVTAB_END"; \ echo "};"; \ ) >$@ # only delete codlist.c in maintainer-mode, since it is included in the dist # FIXME: is there a way to delete in 'make clean' only when builddir != srcdir? maintainer-clean-local: rm -f $(CODLIST) # cancel the hideous builtin rules that cause an infinite loop @MAINTAINER_MODE_TRUE@%: %.o @MAINTAINER_MODE_TRUE@%: %.s @MAINTAINER_MODE_TRUE@%: %.c @MAINTAINER_MODE_TRUE@%: %.S @MAINTAINER_MODE_TRUE@hc2cfdftv_%.c: $(CODELET_DEPS) $(GEN_HC2CDFT_C) @MAINTAINER_MODE_TRUE@ ($(PRELUDE_COMMANDS_RDFT); $(TWOVERS) $(GEN_HC2CDFT_C) $(FLAGS_HC2C) -n $* -dit -name hc2cfdftv_$* -include "hc2cfv.h") | $(ADD_DATE) | $(INDENT) >$@ @MAINTAINER_MODE_TRUE@hc2cbdftv_%.c: $(CODELET_DEPS) $(GEN_HC2CDFT_C) @MAINTAINER_MODE_TRUE@ ($(PRELUDE_COMMANDS_RDFT); $(TWOVERS) $(GEN_HC2CDFT_C) $(FLAGS_HC2C) -n $* -dif -sign 1 -name hc2cbdftv_$* -include "hc2cbv.h") | $(ADD_DATE) | $(INDENT) >$@ # Tell versions [3.59,3.63) of GNU make to not export all variables. # Otherwise a system limit (for SysV at least) may be exceeded. .NOEXPORT: fftw-3.3.3/rdft/simd/common/hc2cfdftv_2.c0000644000175000017500000001015112054411065015024 00000000000000/* * Copyright (c) 2003, 2007-11 Matteo Frigo * Copyright (c) 2003, 2007-11 Massachusetts Institute of Technology * * This program is free software; you can redistribute it and/or modify * it under the terms of the GNU General Public License as published by * the Free Software Foundation; either version 2 of the License, or * (at your option) any later version. * * This program is distributed in the hope that it will be useful, * but WITHOUT ANY WARRANTY; without even the implied warranty of * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the * GNU General Public License for more details. * * You should have received a copy of the GNU General Public License * along with this program; if not, write to the Free Software * Foundation, Inc., 51 Franklin Street, Fifth Floor, Boston, MA 02110-1301 USA * */ /* This file was automatically generated --- DO NOT EDIT */ /* Generated on Sun Nov 25 07:42:29 EST 2012 */ #include "codelet-rdft.h" #ifdef HAVE_FMA /* Generated by: ../../../genfft/gen_hc2cdft_c.native -fma -reorder-insns -schedule-for-pipeline -simd -compact -variables 4 -pipeline-latency 8 -trivial-stores -variables 32 -no-generate-bytw -n 2 -dit -name hc2cfdftv_2 -include hc2cfv.h */ /* * This function contains 5 FP additions, 6 FP multiplications, * (or, 3 additions, 4 multiplications, 2 fused multiply/add), * 9 stack variables, 1 constants, and 4 memory accesses */ #include "hc2cfv.h" static void hc2cfdftv_2(R *Rp, R *Ip, R *Rm, R *Im, const R *W, stride rs, INT mb, INT me, INT ms) { DVK(KP500000000, +0.500000000000000000000000000000000000000000000); { INT m; for (m = mb, W = W + ((mb - 1) * ((TWVL / VL) * 2)); m < me; m = m + VL, Rp = Rp + (VL * ms), Ip = Ip + (VL * ms), Rm = Rm - (VL * ms), Im = Im - (VL * ms), W = W + (TWVL * 2), MAKE_VOLATILE_STRIDE(8, rs)) { V T1, T2, T4, T3, T5, T7, T6; T1 = LD(&(Rp[0]), ms, &(Rp[0])); T2 = LD(&(Rm[0]), -ms, &(Rm[0])); T4 = LDW(&(W[0])); T3 = VFMACONJ(T2, T1); T5 = VZMULIJ(T4, VFNMSCONJ(T2, T1)); T7 = VCONJ(VMUL(LDK(KP500000000), VADD(T3, T5))); T6 = VMUL(LDK(KP500000000), VSUB(T3, T5)); ST(&(Rm[0]), T7, -ms, &(Rm[0])); ST(&(Rp[0]), T6, ms, &(Rp[0])); } } VLEAVE(); } static const tw_instr twinstr[] = { VTW(1, 1), {TW_NEXT, VL, 0} }; static const hc2c_desc desc = { 2, XSIMD_STRING("hc2cfdftv_2"), twinstr, &GENUS, {3, 4, 2, 0} }; void XSIMD(codelet_hc2cfdftv_2) (planner *p) { X(khc2c_register) (p, hc2cfdftv_2, &desc, HC2C_VIA_DFT); } #else /* HAVE_FMA */ /* Generated by: ../../../genfft/gen_hc2cdft_c.native -simd -compact -variables 4 -pipeline-latency 8 -trivial-stores -variables 32 -no-generate-bytw -n 2 -dit -name hc2cfdftv_2 -include hc2cfv.h */ /* * This function contains 5 FP additions, 4 FP multiplications, * (or, 5 additions, 4 multiplications, 0 fused multiply/add), * 10 stack variables, 1 constants, and 4 memory accesses */ #include "hc2cfv.h" static void hc2cfdftv_2(R *Rp, R *Ip, R *Rm, R *Im, const R *W, stride rs, INT mb, INT me, INT ms) { DVK(KP500000000, +0.500000000000000000000000000000000000000000000); { INT m; for (m = mb, W = W + ((mb - 1) * ((TWVL / VL) * 2)); m < me; m = m + VL, Rp = Rp + (VL * ms), Ip = Ip + (VL * ms), Rm = Rm - (VL * ms), Im = Im - (VL * ms), W = W + (TWVL * 2), MAKE_VOLATILE_STRIDE(8, rs)) { V T4, T6, T1, T3, T2, T5, T7, T8; T1 = LD(&(Rp[0]), ms, &(Rp[0])); T2 = LD(&(Rm[0]), -ms, &(Rm[0])); T3 = VCONJ(T2); T4 = VADD(T1, T3); T5 = LDW(&(W[0])); T6 = VZMULIJ(T5, VSUB(T3, T1)); T7 = VCONJ(VMUL(LDK(KP500000000), VSUB(T4, T6))); ST(&(Rm[0]), T7, -ms, &(Rm[0])); T8 = VMUL(LDK(KP500000000), VADD(T4, T6)); ST(&(Rp[0]), T8, ms, &(Rp[0])); } } VLEAVE(); } static const tw_instr twinstr[] = { VTW(1, 1), {TW_NEXT, VL, 0} }; static const hc2c_desc desc = { 2, XSIMD_STRING("hc2cfdftv_2"), twinstr, &GENUS, {5, 4, 0, 0} }; void XSIMD(codelet_hc2cfdftv_2) (planner *p) { X(khc2c_register) (p, hc2cfdftv_2, &desc, HC2C_VIA_DFT); } #endif /* HAVE_FMA */ fftw-3.3.3/rdft/simd/common/hc2cfdftv_4.c0000644000175000017500000001277012054411065015037 00000000000000/* * Copyright (c) 2003, 2007-11 Matteo Frigo * Copyright (c) 2003, 2007-11 Massachusetts Institute of Technology * * This program is free software; you can redistribute it and/or modify * it under the terms of the GNU General Public License as published by * the Free Software Foundation; either version 2 of the License, or * (at your option) any later version. * * This program is distributed in the hope that it will be useful, * but WITHOUT ANY WARRANTY; without even the implied warranty of * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the * GNU General Public License for more details. * * You should have received a copy of the GNU General Public License * along with this program; if not, write to the Free Software * Foundation, Inc., 51 Franklin Street, Fifth Floor, Boston, MA 02110-1301 USA * */ /* This file was automatically generated --- DO NOT EDIT */ /* Generated on Sun Nov 25 07:42:29 EST 2012 */ #include "codelet-rdft.h" #ifdef HAVE_FMA /* Generated by: ../../../genfft/gen_hc2cdft_c.native -fma -reorder-insns -schedule-for-pipeline -simd -compact -variables 4 -pipeline-latency 8 -trivial-stores -variables 32 -no-generate-bytw -n 4 -dit -name hc2cfdftv_4 -include hc2cfv.h */ /* * This function contains 15 FP additions, 16 FP multiplications, * (or, 9 additions, 10 multiplications, 6 fused multiply/add), * 21 stack variables, 1 constants, and 8 memory accesses */ #include "hc2cfv.h" static void hc2cfdftv_4(R *Rp, R *Ip, R *Rm, R *Im, const R *W, stride rs, INT mb, INT me, INT ms) { DVK(KP500000000, +0.500000000000000000000000000000000000000000000); { INT m; for (m = mb, W = W + ((mb - 1) * ((TWVL / VL) * 6)); m < me; m = m + VL, Rp = Rp + (VL * ms), Ip = Ip + (VL * ms), Rm = Rm - (VL * ms), Im = Im - (VL * ms), W = W + (TWVL * 6), MAKE_VOLATILE_STRIDE(16, rs)) { V T1, T2, Tb, T5, T6, T4, T9, T3, Tc, T7, Ta, Tg, T8, Td, Th; V Tf, Te, Ti, Tj; T1 = LD(&(Rp[0]), ms, &(Rp[0])); T2 = LD(&(Rm[0]), -ms, &(Rm[0])); Tb = LDW(&(W[0])); T5 = LD(&(Rp[WS(rs, 1)]), ms, &(Rp[WS(rs, 1)])); T6 = LD(&(Rm[WS(rs, 1)]), -ms, &(Rm[WS(rs, 1)])); T4 = LDW(&(W[TWVL * 2])); T9 = LDW(&(W[TWVL * 4])); T3 = VFMACONJ(T2, T1); Tc = VZMULIJ(Tb, VFNMSCONJ(T2, T1)); T7 = VZMULJ(T4, VFMACONJ(T6, T5)); Ta = VZMULIJ(T9, VFNMSCONJ(T6, T5)); Tg = VADD(T3, T7); T8 = VSUB(T3, T7); Td = VSUB(Ta, Tc); Th = VADD(Tc, Ta); Tf = VCONJ(VMUL(LDK(KP500000000), VFMAI(Td, T8))); Te = VMUL(LDK(KP500000000), VFNMSI(Td, T8)); Ti = VMUL(LDK(KP500000000), VSUB(Tg, Th)); Tj = VCONJ(VMUL(LDK(KP500000000), VADD(Th, Tg))); ST(&(Rm[0]), Tf, -ms, &(Rm[0])); ST(&(Rp[WS(rs, 1)]), Te, ms, &(Rp[WS(rs, 1)])); ST(&(Rp[0]), Ti, ms, &(Rp[0])); ST(&(Rm[WS(rs, 1)]), Tj, -ms, &(Rm[WS(rs, 1)])); } } VLEAVE(); } static const tw_instr twinstr[] = { VTW(1, 1), VTW(1, 2), VTW(1, 3), {TW_NEXT, VL, 0} }; static const hc2c_desc desc = { 4, XSIMD_STRING("hc2cfdftv_4"), twinstr, &GENUS, {9, 10, 6, 0} }; void XSIMD(codelet_hc2cfdftv_4) (planner *p) { X(khc2c_register) (p, hc2cfdftv_4, &desc, HC2C_VIA_DFT); } #else /* HAVE_FMA */ /* Generated by: ../../../genfft/gen_hc2cdft_c.native -simd -compact -variables 4 -pipeline-latency 8 -trivial-stores -variables 32 -no-generate-bytw -n 4 -dit -name hc2cfdftv_4 -include hc2cfv.h */ /* * This function contains 15 FP additions, 10 FP multiplications, * (or, 15 additions, 10 multiplications, 0 fused multiply/add), * 23 stack variables, 1 constants, and 8 memory accesses */ #include "hc2cfv.h" static void hc2cfdftv_4(R *Rp, R *Ip, R *Rm, R *Im, const R *W, stride rs, INT mb, INT me, INT ms) { DVK(KP500000000, +0.500000000000000000000000000000000000000000000); { INT m; for (m = mb, W = W + ((mb - 1) * ((TWVL / VL) * 6)); m < me; m = m + VL, Rp = Rp + (VL * ms), Ip = Ip + (VL * ms), Rm = Rm - (VL * ms), Im = Im - (VL * ms), W = W + (TWVL * 6), MAKE_VOLATILE_STRIDE(16, rs)) { V T4, Tc, T9, Te, T1, T3, T2, Tb, T6, T8, T7, T5, Td, Tg, Th; V Ta, Tf, Tk, Tl, Ti, Tj; T1 = LD(&(Rp[0]), ms, &(Rp[0])); T2 = LD(&(Rm[0]), -ms, &(Rm[0])); T3 = VCONJ(T2); T4 = VADD(T1, T3); Tb = LDW(&(W[0])); Tc = VZMULIJ(Tb, VSUB(T3, T1)); T6 = LD(&(Rp[WS(rs, 1)]), ms, &(Rp[WS(rs, 1)])); T7 = LD(&(Rm[WS(rs, 1)]), -ms, &(Rm[WS(rs, 1)])); T8 = VCONJ(T7); T5 = LDW(&(W[TWVL * 2])); T9 = VZMULJ(T5, VADD(T6, T8)); Td = LDW(&(W[TWVL * 4])); Te = VZMULIJ(Td, VSUB(T8, T6)); Ta = VSUB(T4, T9); Tf = VBYI(VSUB(Tc, Te)); Tg = VMUL(LDK(KP500000000), VSUB(Ta, Tf)); Th = VCONJ(VMUL(LDK(KP500000000), VADD(Ta, Tf))); ST(&(Rp[WS(rs, 1)]), Tg, ms, &(Rp[WS(rs, 1)])); ST(&(Rm[0]), Th, -ms, &(Rm[0])); Ti = VADD(T4, T9); Tj = VADD(Tc, Te); Tk = VCONJ(VMUL(LDK(KP500000000), VSUB(Ti, Tj))); Tl = VMUL(LDK(KP500000000), VADD(Ti, Tj)); ST(&(Rm[WS(rs, 1)]), Tk, -ms, &(Rm[WS(rs, 1)])); ST(&(Rp[0]), Tl, ms, &(Rp[0])); } } VLEAVE(); } static const tw_instr twinstr[] = { VTW(1, 1), VTW(1, 2), VTW(1, 3), {TW_NEXT, VL, 0} }; static const hc2c_desc desc = { 4, XSIMD_STRING("hc2cfdftv_4"), twinstr, &GENUS, {15, 10, 0, 0} }; void XSIMD(codelet_hc2cfdftv_4) (planner *p) { X(khc2c_register) (p, hc2cfdftv_4, &desc, HC2C_VIA_DFT); } #endif /* HAVE_FMA */ fftw-3.3.3/rdft/simd/common/hc2cfdftv_6.c0000644000175000017500000001670312054411065015041 00000000000000/* * Copyright (c) 2003, 2007-11 Matteo Frigo * Copyright (c) 2003, 2007-11 Massachusetts Institute of Technology * * This program is free software; you can redistribute it and/or modify * it under the terms of the GNU General Public License as published by * the Free Software Foundation; either version 2 of the License, or * (at your option) any later version. * * This program is distributed in the hope that it will be useful, * but WITHOUT ANY WARRANTY; without even the implied warranty of * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the * GNU General Public License for more details. * * You should have received a copy of the GNU General Public License * along with this program; if not, write to the Free Software * Foundation, Inc., 51 Franklin Street, Fifth Floor, Boston, MA 02110-1301 USA * */ /* This file was automatically generated --- DO NOT EDIT */ /* Generated on Sun Nov 25 07:42:29 EST 2012 */ #include "codelet-rdft.h" #ifdef HAVE_FMA /* Generated by: ../../../genfft/gen_hc2cdft_c.native -fma -reorder-insns -schedule-for-pipeline -simd -compact -variables 4 -pipeline-latency 8 -trivial-stores -variables 32 -no-generate-bytw -n 6 -dit -name hc2cfdftv_6 -include hc2cfv.h */ /* * This function contains 29 FP additions, 30 FP multiplications, * (or, 17 additions, 18 multiplications, 12 fused multiply/add), * 38 stack variables, 2 constants, and 12 memory accesses */ #include "hc2cfv.h" static void hc2cfdftv_6(R *Rp, R *Ip, R *Rm, R *Im, const R *W, stride rs, INT mb, INT me, INT ms) { DVK(KP500000000, +0.500000000000000000000000000000000000000000000); DVK(KP866025403, +0.866025403784438646763723170752936183471402627); { INT m; for (m = mb, W = W + ((mb - 1) * ((TWVL / VL) * 10)); m < me; m = m + VL, Rp = Rp + (VL * ms), Ip = Ip + (VL * ms), Rm = Rm - (VL * ms), Im = Im - (VL * ms), W = W + (TWVL * 10), MAKE_VOLATILE_STRIDE(24, rs)) { V T5, T6, T3, Tj, T4, T9, Te, Th, T1, T2, Ti, Tc, Td, Tb, Tg; V T7, Ta, Tt, Tk, Tr, T8, Ts, Tf, Tx, Tu, To, Tl, Tw, Tv, Tn; V Tm, Tz, Ty, Tp, Tq; T1 = LD(&(Rp[0]), ms, &(Rp[0])); T2 = LD(&(Rm[0]), -ms, &(Rm[0])); Ti = LDW(&(W[0])); Tc = LD(&(Rp[WS(rs, 2)]), ms, &(Rp[0])); Td = LD(&(Rm[WS(rs, 2)]), -ms, &(Rm[0])); Tb = LDW(&(W[TWVL * 8])); Tg = LDW(&(W[TWVL * 6])); T5 = LD(&(Rp[WS(rs, 1)]), ms, &(Rp[WS(rs, 1)])); T6 = LD(&(Rm[WS(rs, 1)]), -ms, &(Rm[WS(rs, 1)])); T3 = VFMACONJ(T2, T1); Tj = VZMULIJ(Ti, VFNMSCONJ(T2, T1)); T4 = LDW(&(W[TWVL * 4])); T9 = LDW(&(W[TWVL * 2])); Te = VZMULIJ(Tb, VFNMSCONJ(Td, Tc)); Th = VZMULJ(Tg, VFMACONJ(Td, Tc)); T7 = VZMULIJ(T4, VFNMSCONJ(T6, T5)); Ta = VZMULJ(T9, VFMACONJ(T6, T5)); Tt = VADD(Tj, Th); Tk = VSUB(Th, Tj); Tr = VADD(T3, T7); T8 = VSUB(T3, T7); Ts = VADD(Ta, Te); Tf = VSUB(Ta, Te); Tx = VMUL(LDK(KP866025403), VSUB(Tt, Ts)); Tu = VADD(Ts, Tt); To = VMUL(LDK(KP866025403), VSUB(Tk, Tf)); Tl = VADD(Tf, Tk); Tw = VFNMS(LDK(KP500000000), Tu, Tr); Tv = VCONJ(VMUL(LDK(KP500000000), VADD(Tr, Tu))); Tn = VFNMS(LDK(KP500000000), Tl, T8); Tm = VMUL(LDK(KP500000000), VADD(T8, Tl)); Tz = VMUL(LDK(KP500000000), VFMAI(Tx, Tw)); Ty = VCONJ(VMUL(LDK(KP500000000), VFNMSI(Tx, Tw))); ST(&(Rm[WS(rs, 2)]), Tv, -ms, &(Rm[0])); Tp = VMUL(LDK(KP500000000), VFNMSI(To, Tn)); Tq = VCONJ(VMUL(LDK(KP500000000), VFMAI(To, Tn))); ST(&(Rp[0]), Tm, ms, &(Rp[0])); ST(&(Rp[WS(rs, 1)]), Tz, ms, &(Rp[WS(rs, 1)])); ST(&(Rm[0]), Ty, -ms, &(Rm[0])); ST(&(Rm[WS(rs, 1)]), Tq, -ms, &(Rm[WS(rs, 1)])); ST(&(Rp[WS(rs, 2)]), Tp, ms, &(Rp[0])); } } VLEAVE(); } static const tw_instr twinstr[] = { VTW(1, 1), VTW(1, 2), VTW(1, 3), VTW(1, 4), VTW(1, 5), {TW_NEXT, VL, 0} }; static const hc2c_desc desc = { 6, XSIMD_STRING("hc2cfdftv_6"), twinstr, &GENUS, {17, 18, 12, 0} }; void XSIMD(codelet_hc2cfdftv_6) (planner *p) { X(khc2c_register) (p, hc2cfdftv_6, &desc, HC2C_VIA_DFT); } #else /* HAVE_FMA */ /* Generated by: ../../../genfft/gen_hc2cdft_c.native -simd -compact -variables 4 -pipeline-latency 8 -trivial-stores -variables 32 -no-generate-bytw -n 6 -dit -name hc2cfdftv_6 -include hc2cfv.h */ /* * This function contains 29 FP additions, 20 FP multiplications, * (or, 27 additions, 18 multiplications, 2 fused multiply/add), * 42 stack variables, 3 constants, and 12 memory accesses */ #include "hc2cfv.h" static void hc2cfdftv_6(R *Rp, R *Ip, R *Rm, R *Im, const R *W, stride rs, INT mb, INT me, INT ms) { DVK(KP250000000, +0.250000000000000000000000000000000000000000000); DVK(KP866025403, +0.866025403784438646763723170752936183471402627); DVK(KP500000000, +0.500000000000000000000000000000000000000000000); { INT m; for (m = mb, W = W + ((mb - 1) * ((TWVL / VL) * 10)); m < me; m = m + VL, Rp = Rp + (VL * ms), Ip = Ip + (VL * ms), Rm = Rm - (VL * ms), Im = Im - (VL * ms), W = W + (TWVL * 10), MAKE_VOLATILE_STRIDE(24, rs)) { V Ta, Tu, Tn, Tw, Ti, Tv, T1, T8, Tg, Tf, T7, T3, Te, T6, T2; V T4, T9, T5, Tk, Tm, Tj, Tl, Tc, Th, Tb, Td, Tr, Tp, Tq, To; V Tt, Ts, TA, Ty, Tz, Tx, TC, TB; T1 = LD(&(Rp[0]), ms, &(Rp[0])); T8 = LD(&(Rp[WS(rs, 1)]), ms, &(Rp[WS(rs, 1)])); Tg = LD(&(Rp[WS(rs, 2)]), ms, &(Rp[0])); Te = LD(&(Rm[WS(rs, 2)]), -ms, &(Rm[0])); Tf = VCONJ(Te); T6 = LD(&(Rm[WS(rs, 1)]), -ms, &(Rm[WS(rs, 1)])); T7 = VCONJ(T6); T2 = LD(&(Rm[0]), -ms, &(Rm[0])); T3 = VCONJ(T2); T4 = VADD(T1, T3); T5 = LDW(&(W[TWVL * 4])); T9 = VZMULIJ(T5, VSUB(T7, T8)); Ta = VADD(T4, T9); Tu = VSUB(T4, T9); Tj = LDW(&(W[0])); Tk = VZMULIJ(Tj, VSUB(T3, T1)); Tl = LDW(&(W[TWVL * 6])); Tm = VZMULJ(Tl, VADD(Tf, Tg)); Tn = VADD(Tk, Tm); Tw = VSUB(Tm, Tk); Tb = LDW(&(W[TWVL * 2])); Tc = VZMULJ(Tb, VADD(T7, T8)); Td = LDW(&(W[TWVL * 8])); Th = VZMULIJ(Td, VSUB(Tf, Tg)); Ti = VADD(Tc, Th); Tv = VSUB(Tc, Th); Tr = VMUL(LDK(KP500000000), VBYI(VMUL(LDK(KP866025403), VSUB(Tn, Ti)))); To = VADD(Ti, Tn); Tp = VMUL(LDK(KP500000000), VADD(Ta, To)); Tq = VFNMS(LDK(KP250000000), To, VMUL(LDK(KP500000000), Ta)); ST(&(Rp[0]), Tp, ms, &(Rp[0])); Tt = VCONJ(VADD(Tq, Tr)); ST(&(Rm[WS(rs, 1)]), Tt, -ms, &(Rm[WS(rs, 1)])); Ts = VSUB(Tq, Tr); ST(&(Rp[WS(rs, 2)]), Ts, ms, &(Rp[0])); TA = VMUL(LDK(KP500000000), VBYI(VMUL(LDK(KP866025403), VSUB(Tw, Tv)))); Tx = VADD(Tv, Tw); Ty = VCONJ(VMUL(LDK(KP500000000), VADD(Tu, Tx))); Tz = VFNMS(LDK(KP250000000), Tx, VMUL(LDK(KP500000000), Tu)); ST(&(Rm[WS(rs, 2)]), Ty, -ms, &(Rm[0])); TC = VADD(Tz, TA); ST(&(Rp[WS(rs, 1)]), TC, ms, &(Rp[WS(rs, 1)])); TB = VCONJ(VSUB(Tz, TA)); ST(&(Rm[0]), TB, -ms, &(Rm[0])); } } VLEAVE(); } static const tw_instr twinstr[] = { VTW(1, 1), VTW(1, 2), VTW(1, 3), VTW(1, 4), VTW(1, 5), {TW_NEXT, VL, 0} }; static const hc2c_desc desc = { 6, XSIMD_STRING("hc2cfdftv_6"), twinstr, &GENUS, {27, 18, 2, 0} }; void XSIMD(codelet_hc2cfdftv_6) (planner *p) { X(khc2c_register) (p, hc2cfdftv_6, &desc, HC2C_VIA_DFT); } #endif /* HAVE_FMA */ fftw-3.3.3/rdft/simd/common/hc2cfdftv_8.c0000644000175000017500000002157312054411065015044 00000000000000/* * Copyright (c) 2003, 2007-11 Matteo Frigo * Copyright (c) 2003, 2007-11 Massachusetts Institute of Technology * * This program is free software; you can redistribute it and/or modify * it under the terms of the GNU General Public License as published by * the Free Software Foundation; either version 2 of the License, or * (at your option) any later version. * * This program is distributed in the hope that it will be useful, * but WITHOUT ANY WARRANTY; without even the implied warranty of * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the * GNU General Public License for more details. * * You should have received a copy of the GNU General Public License * along with this program; if not, write to the Free Software * Foundation, Inc., 51 Franklin Street, Fifth Floor, Boston, MA 02110-1301 USA * */ /* This file was automatically generated --- DO NOT EDIT */ /* Generated on Sun Nov 25 07:42:29 EST 2012 */ #include "codelet-rdft.h" #ifdef HAVE_FMA /* Generated by: ../../../genfft/gen_hc2cdft_c.native -fma -reorder-insns -schedule-for-pipeline -simd -compact -variables 4 -pipeline-latency 8 -trivial-stores -variables 32 -no-generate-bytw -n 8 -dit -name hc2cfdftv_8 -include hc2cfv.h */ /* * This function contains 41 FP additions, 40 FP multiplications, * (or, 23 additions, 22 multiplications, 18 fused multiply/add), * 52 stack variables, 2 constants, and 16 memory accesses */ #include "hc2cfv.h" static void hc2cfdftv_8(R *Rp, R *Ip, R *Rm, R *Im, const R *W, stride rs, INT mb, INT me, INT ms) { DVK(KP500000000, +0.500000000000000000000000000000000000000000000); DVK(KP707106781, +0.707106781186547524400844362104849039284835938); { INT m; for (m = mb, W = W + ((mb - 1) * ((TWVL / VL) * 14)); m < me; m = m + VL, Rp = Rp + (VL * ms), Ip = Ip + (VL * ms), Rm = Rm - (VL * ms), Im = Im - (VL * ms), W = W + (TWVL * 14), MAKE_VOLATILE_STRIDE(32, rs)) { V T3, Tc, Tl, Ts, Tf, Tg, Te, Tp, T7, Ta, T1, T2, Tb, Tj, Tk; V Ti, Tr, T5, T6, T4, T9, Th, Tq, TC, T8, Td, TF, Tm, TG, TD; V Tt, Tu, Tn, TH, TL, TE, TK, Tz, Tv, Ty, To, TJ, TI, TN, TM; V TB, TA, Tx, Tw; T1 = LD(&(Rp[0]), ms, &(Rp[0])); T2 = LD(&(Rm[0]), -ms, &(Rm[0])); Tb = LDW(&(W[0])); Tj = LD(&(Rp[WS(rs, 3)]), ms, &(Rp[WS(rs, 1)])); Tk = LD(&(Rm[WS(rs, 3)]), -ms, &(Rm[WS(rs, 1)])); Ti = LDW(&(W[TWVL * 12])); Tr = LDW(&(W[TWVL * 10])); T5 = LD(&(Rp[WS(rs, 2)]), ms, &(Rp[0])); T6 = LD(&(Rm[WS(rs, 2)]), -ms, &(Rm[0])); T3 = VFMACONJ(T2, T1); Tc = VZMULIJ(Tb, VFNMSCONJ(T2, T1)); T4 = LDW(&(W[TWVL * 6])); T9 = LDW(&(W[TWVL * 8])); Tl = VZMULIJ(Ti, VFNMSCONJ(Tk, Tj)); Ts = VZMULJ(Tr, VFMACONJ(Tk, Tj)); Tf = LD(&(Rp[WS(rs, 1)]), ms, &(Rp[WS(rs, 1)])); Tg = LD(&(Rm[WS(rs, 1)]), -ms, &(Rm[WS(rs, 1)])); Te = LDW(&(W[TWVL * 4])); Tp = LDW(&(W[TWVL * 2])); T7 = VZMULJ(T4, VFMACONJ(T6, T5)); Ta = VZMULIJ(T9, VFNMSCONJ(T6, T5)); Th = VZMULIJ(Te, VFNMSCONJ(Tg, Tf)); Tq = VZMULJ(Tp, VFMACONJ(Tg, Tf)); TC = VADD(T3, T7); T8 = VSUB(T3, T7); Td = VSUB(Ta, Tc); TF = VADD(Tc, Ta); Tm = VSUB(Th, Tl); TG = VADD(Th, Tl); TD = VADD(Tq, Ts); Tt = VSUB(Tq, Ts); Tu = VSUB(Tm, Td); Tn = VADD(Td, Tm); TH = VSUB(TF, TG); TL = VADD(TF, TG); TE = VSUB(TC, TD); TK = VADD(TC, TD); Tz = VFMA(LDK(KP707106781), Tu, Tt); Tv = VFNMS(LDK(KP707106781), Tu, Tt); Ty = VFNMS(LDK(KP707106781), Tn, T8); To = VFMA(LDK(KP707106781), Tn, T8); TJ = VCONJ(VMUL(LDK(KP500000000), VFNMSI(TH, TE))); TI = VMUL(LDK(KP500000000), VFMAI(TH, TE)); TN = VCONJ(VMUL(LDK(KP500000000), VADD(TL, TK))); TM = VMUL(LDK(KP500000000), VSUB(TK, TL)); TB = VMUL(LDK(KP500000000), VFMAI(Tz, Ty)); TA = VCONJ(VMUL(LDK(KP500000000), VFNMSI(Tz, Ty))); Tx = VCONJ(VMUL(LDK(KP500000000), VFMAI(Tv, To))); Tw = VMUL(LDK(KP500000000), VFNMSI(Tv, To)); ST(&(Rm[WS(rs, 1)]), TJ, -ms, &(Rm[WS(rs, 1)])); ST(&(Rp[WS(rs, 2)]), TI, ms, &(Rp[0])); ST(&(Rm[WS(rs, 3)]), TN, -ms, &(Rm[WS(rs, 1)])); ST(&(Rp[0]), TM, ms, &(Rp[0])); ST(&(Rp[WS(rs, 3)]), TB, ms, &(Rp[WS(rs, 1)])); ST(&(Rm[WS(rs, 2)]), TA, -ms, &(Rm[0])); ST(&(Rm[0]), Tx, -ms, &(Rm[0])); ST(&(Rp[WS(rs, 1)]), Tw, ms, &(Rp[WS(rs, 1)])); } } VLEAVE(); } static const tw_instr twinstr[] = { VTW(1, 1), VTW(1, 2), VTW(1, 3), VTW(1, 4), VTW(1, 5), VTW(1, 6), VTW(1, 7), {TW_NEXT, VL, 0} }; static const hc2c_desc desc = { 8, XSIMD_STRING("hc2cfdftv_8"), twinstr, &GENUS, {23, 22, 18, 0} }; void XSIMD(codelet_hc2cfdftv_8) (planner *p) { X(khc2c_register) (p, hc2cfdftv_8, &desc, HC2C_VIA_DFT); } #else /* HAVE_FMA */ /* Generated by: ../../../genfft/gen_hc2cdft_c.native -simd -compact -variables 4 -pipeline-latency 8 -trivial-stores -variables 32 -no-generate-bytw -n 8 -dit -name hc2cfdftv_8 -include hc2cfv.h */ /* * This function contains 41 FP additions, 23 FP multiplications, * (or, 41 additions, 23 multiplications, 0 fused multiply/add), * 57 stack variables, 3 constants, and 16 memory accesses */ #include "hc2cfv.h" static void hc2cfdftv_8(R *Rp, R *Ip, R *Rm, R *Im, const R *W, stride rs, INT mb, INT me, INT ms) { DVK(KP707106781, +0.707106781186547524400844362104849039284835938); DVK(KP353553390, +0.353553390593273762200422181052424519642417969); DVK(KP500000000, +0.500000000000000000000000000000000000000000000); { INT m; for (m = mb, W = W + ((mb - 1) * ((TWVL / VL) * 14)); m < me; m = m + VL, Rp = Rp + (VL * ms), Ip = Ip + (VL * ms), Rm = Rm - (VL * ms), Im = Im - (VL * ms), W = W + (TWVL * 14), MAKE_VOLATILE_STRIDE(32, rs)) { V Ta, TE, Tr, TF, Tl, TK, Tw, TG, T1, T6, T3, T8, T2, T7, T4; V T9, T5, To, Tq, Tn, Tp, Tc, Th, Te, Tj, Td, Ti, Tf, Tk, Tb; V Tg, Tt, Tv, Ts, Tu, Ty, Tz, Tm, Tx, TC, TD, TA, TB, TI, TO; V TL, TP, TH, TJ, TM, TR, TN, TQ; T1 = LD(&(Rp[0]), ms, &(Rp[0])); T6 = LD(&(Rp[WS(rs, 2)]), ms, &(Rp[0])); T2 = LD(&(Rm[0]), -ms, &(Rm[0])); T3 = VCONJ(T2); T7 = LD(&(Rm[WS(rs, 2)]), -ms, &(Rm[0])); T8 = VCONJ(T7); T4 = VADD(T1, T3); T5 = LDW(&(W[TWVL * 6])); T9 = VZMULJ(T5, VADD(T6, T8)); Ta = VADD(T4, T9); TE = VMUL(LDK(KP500000000), VSUB(T4, T9)); Tn = LDW(&(W[0])); To = VZMULIJ(Tn, VSUB(T3, T1)); Tp = LDW(&(W[TWVL * 8])); Tq = VZMULIJ(Tp, VSUB(T8, T6)); Tr = VADD(To, Tq); TF = VSUB(To, Tq); Tc = LD(&(Rp[WS(rs, 1)]), ms, &(Rp[WS(rs, 1)])); Th = LD(&(Rp[WS(rs, 3)]), ms, &(Rp[WS(rs, 1)])); Td = LD(&(Rm[WS(rs, 1)]), -ms, &(Rm[WS(rs, 1)])); Te = VCONJ(Td); Ti = LD(&(Rm[WS(rs, 3)]), -ms, &(Rm[WS(rs, 1)])); Tj = VCONJ(Ti); Tb = LDW(&(W[TWVL * 2])); Tf = VZMULJ(Tb, VADD(Tc, Te)); Tg = LDW(&(W[TWVL * 10])); Tk = VZMULJ(Tg, VADD(Th, Tj)); Tl = VADD(Tf, Tk); TK = VSUB(Tf, Tk); Ts = LDW(&(W[TWVL * 4])); Tt = VZMULIJ(Ts, VSUB(Te, Tc)); Tu = LDW(&(W[TWVL * 12])); Tv = VZMULIJ(Tu, VSUB(Tj, Th)); Tw = VADD(Tt, Tv); TG = VSUB(Tv, Tt); Tm = VADD(Ta, Tl); Tx = VADD(Tr, Tw); Ty = VCONJ(VMUL(LDK(KP500000000), VSUB(Tm, Tx))); Tz = VMUL(LDK(KP500000000), VADD(Tm, Tx)); ST(&(Rm[WS(rs, 3)]), Ty, -ms, &(Rm[WS(rs, 1)])); ST(&(Rp[0]), Tz, ms, &(Rp[0])); TA = VSUB(Ta, Tl); TB = VBYI(VSUB(Tw, Tr)); TC = VCONJ(VMUL(LDK(KP500000000), VSUB(TA, TB))); TD = VMUL(LDK(KP500000000), VADD(TA, TB)); ST(&(Rm[WS(rs, 1)]), TC, -ms, &(Rm[WS(rs, 1)])); ST(&(Rp[WS(rs, 2)]), TD, ms, &(Rp[0])); TH = VMUL(LDK(KP353553390), VADD(TF, TG)); TI = VADD(TE, TH); TO = VSUB(TE, TH); TJ = VMUL(LDK(KP707106781), VSUB(TG, TF)); TL = VMUL(LDK(KP500000000), VBYI(VSUB(TJ, TK))); TP = VMUL(LDK(KP500000000), VBYI(VADD(TK, TJ))); TM = VCONJ(VSUB(TI, TL)); ST(&(Rm[0]), TM, -ms, &(Rm[0])); TR = VADD(TO, TP); ST(&(Rp[WS(rs, 3)]), TR, ms, &(Rp[WS(rs, 1)])); TN = VADD(TI, TL); ST(&(Rp[WS(rs, 1)]), TN, ms, &(Rp[WS(rs, 1)])); TQ = VCONJ(VSUB(TO, TP)); ST(&(Rm[WS(rs, 2)]), TQ, -ms, &(Rm[0])); } } VLEAVE(); } static const tw_instr twinstr[] = { VTW(1, 1), VTW(1, 2), VTW(1, 3), VTW(1, 4), VTW(1, 5), VTW(1, 6), VTW(1, 7), {TW_NEXT, VL, 0} }; static const hc2c_desc desc = { 8, XSIMD_STRING("hc2cfdftv_8"), twinstr, &GENUS, {41, 23, 0, 0} }; void XSIMD(codelet_hc2cfdftv_8) (planner *p) { X(khc2c_register) (p, hc2cfdftv_8, &desc, HC2C_VIA_DFT); } #endif /* HAVE_FMA */ fftw-3.3.3/rdft/simd/common/hc2cfdftv_10.c0000644000175000017500000002700312054411065015107 00000000000000/* * Copyright (c) 2003, 2007-11 Matteo Frigo * Copyright (c) 2003, 2007-11 Massachusetts Institute of Technology * * This program is free software; you can redistribute it and/or modify * it under the terms of the GNU General Public License as published by * the Free Software Foundation; either version 2 of the License, or * (at your option) any later version. * * This program is distributed in the hope that it will be useful, * but WITHOUT ANY WARRANTY; without even the implied warranty of * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the * GNU General Public License for more details. * * You should have received a copy of the GNU General Public License * along with this program; if not, write to the Free Software * Foundation, Inc., 51 Franklin Street, Fifth Floor, Boston, MA 02110-1301 USA * */ /* This file was automatically generated --- DO NOT EDIT */ /* Generated on Sun Nov 25 07:42:29 EST 2012 */ #include "codelet-rdft.h" #ifdef HAVE_FMA /* Generated by: ../../../genfft/gen_hc2cdft_c.native -fma -reorder-insns -schedule-for-pipeline -simd -compact -variables 4 -pipeline-latency 8 -trivial-stores -variables 32 -no-generate-bytw -n 10 -dit -name hc2cfdftv_10 -include hc2cfv.h */ /* * This function contains 61 FP additions, 60 FP multiplications, * (or, 33 additions, 32 multiplications, 28 fused multiply/add), * 77 stack variables, 5 constants, and 20 memory accesses */ #include "hc2cfv.h" static void hc2cfdftv_10(R *Rp, R *Ip, R *Rm, R *Im, const R *W, stride rs, INT mb, INT me, INT ms) { DVK(KP559016994, +0.559016994374947424102293417182819058860154590); DVK(KP500000000, +0.500000000000000000000000000000000000000000000); DVK(KP250000000, +0.250000000000000000000000000000000000000000000); DVK(KP618033988, +0.618033988749894848204586834365638117720309180); DVK(KP951056516, +0.951056516295153572116439333379382143405698634); { INT m; for (m = mb, W = W + ((mb - 1) * ((TWVL / VL) * 18)); m < me; m = m + VL, Rp = Rp + (VL * ms), Ip = Ip + (VL * ms), Rm = Rm - (VL * ms), Im = Im - (VL * ms), W = W + (TWVL * 18), MAKE_VOLATILE_STRIDE(40, rs)) { V T5, T6, Tw, Tr, Tc, Tj, Tl, Tm, Tk, Ts, Tg, Ty, T3, T4, T1; V T2, Tv, Tq, Ta, Tb, T9, Ti, Te, Tf, Td, Tx, Tn, Tt, Th, TQ; V TT, Tz, T7, TR, To, Tu, TU; T1 = LD(&(Rp[0]), ms, &(Rp[0])); T2 = LD(&(Rm[0]), -ms, &(Rm[0])); Tv = LDW(&(W[0])); T5 = LD(&(Rp[WS(rs, 2)]), ms, &(Rp[0])); T6 = LD(&(Rm[WS(rs, 2)]), -ms, &(Rm[0])); Tq = LDW(&(W[TWVL * 6])); Ta = LD(&(Rp[WS(rs, 1)]), ms, &(Rp[WS(rs, 1)])); Tb = LD(&(Rm[WS(rs, 1)]), -ms, &(Rm[WS(rs, 1)])); T9 = LDW(&(W[TWVL * 2])); Ti = LDW(&(W[TWVL * 4])); Tw = VZMULIJ(Tv, VFNMSCONJ(T2, T1)); Te = LD(&(Rp[WS(rs, 3)]), ms, &(Rp[WS(rs, 1)])); Tf = LD(&(Rm[WS(rs, 3)]), -ms, &(Rm[WS(rs, 1)])); Tr = VZMULJ(Tq, VFMACONJ(T6, T5)); Td = LDW(&(W[TWVL * 12])); Tx = LDW(&(W[TWVL * 10])); Tc = VZMULJ(T9, VFMACONJ(Tb, Ta)); Tj = VZMULIJ(Ti, VFNMSCONJ(Tb, Ta)); Tl = LD(&(Rp[WS(rs, 4)]), ms, &(Rp[0])); Tm = LD(&(Rm[WS(rs, 4)]), -ms, &(Rm[0])); Tk = LDW(&(W[TWVL * 14])); Ts = LDW(&(W[TWVL * 16])); Tg = VZMULIJ(Td, VFNMSCONJ(Tf, Te)); Ty = VZMULJ(Tx, VFMACONJ(Tf, Te)); T3 = VFMACONJ(T2, T1); T4 = LDW(&(W[TWVL * 8])); Tn = VZMULJ(Tk, VFMACONJ(Tm, Tl)); Tt = VZMULIJ(Ts, VFNMSCONJ(Tm, Tl)); Th = VSUB(Tc, Tg); TQ = VADD(Tc, Tg); TT = VADD(Tw, Ty); Tz = VSUB(Tw, Ty); T7 = VZMULIJ(T4, VFNMSCONJ(T6, T5)); TR = VADD(Tj, Tn); To = VSUB(Tj, Tn); Tu = VSUB(Tr, Tt); TU = VADD(Tr, Tt); { V TP, T8, TS, T11, Tp, TH, TA, TG, TV, T12, TE, TB, TM, TI, TZ; V TW, T17, T13, TD, TC, TY, TX, TL, TF, T10, T16, TN, TO, TK, TJ; V T18, T19, T15, T14; TP = VADD(T3, T7); T8 = VSUB(T3, T7); TS = VADD(TQ, TR); T11 = VSUB(TQ, TR); Tp = VSUB(Th, To); TH = VADD(Th, To); TA = VSUB(Tu, Tz); TG = VADD(Tz, Tu); TV = VADD(TT, TU); T12 = VSUB(TU, TT); TE = VSUB(Tp, TA); TB = VADD(Tp, TA); TM = VMUL(LDK(KP951056516), VFMA(LDK(KP618033988), TG, TH)); TI = VMUL(LDK(KP951056516), VFNMS(LDK(KP618033988), TH, TG)); TZ = VSUB(TS, TV); TW = VADD(TS, TV); T17 = VMUL(LDK(KP951056516), VFNMS(LDK(KP618033988), T11, T12)); T13 = VMUL(LDK(KP951056516), VFMA(LDK(KP618033988), T12, T11)); TD = VFNMS(LDK(KP250000000), TB, T8); TC = VMUL(LDK(KP500000000), VADD(T8, TB)); TY = VFNMS(LDK(KP250000000), TW, TP); TX = VCONJ(VMUL(LDK(KP500000000), VADD(TP, TW))); TL = VFMA(LDK(KP559016994), TE, TD); TF = VFNMS(LDK(KP559016994), TE, TD); ST(&(Rp[0]), TC, ms, &(Rp[0])); T10 = VFMA(LDK(KP559016994), TZ, TY); T16 = VFNMS(LDK(KP559016994), TZ, TY); ST(&(Rm[WS(rs, 4)]), TX, -ms, &(Rm[0])); TN = VCONJ(VMUL(LDK(KP500000000), VFNMSI(TM, TL))); TO = VMUL(LDK(KP500000000), VFMAI(TM, TL)); TK = VMUL(LDK(KP500000000), VFMAI(TI, TF)); TJ = VCONJ(VMUL(LDK(KP500000000), VFNMSI(TI, TF))); T18 = VMUL(LDK(KP500000000), VFNMSI(T17, T16)); T19 = VCONJ(VMUL(LDK(KP500000000), VFMAI(T17, T16))); T15 = VCONJ(VMUL(LDK(KP500000000), VFMAI(T13, T10))); T14 = VMUL(LDK(KP500000000), VFNMSI(T13, T10)); ST(&(Rm[WS(rs, 3)]), TN, -ms, &(Rm[WS(rs, 1)])); ST(&(Rp[WS(rs, 4)]), TO, ms, &(Rp[0])); ST(&(Rp[WS(rs, 2)]), TK, ms, &(Rp[0])); ST(&(Rm[WS(rs, 1)]), TJ, -ms, &(Rm[WS(rs, 1)])); ST(&(Rp[WS(rs, 3)]), T18, ms, &(Rp[WS(rs, 1)])); ST(&(Rm[WS(rs, 2)]), T19, -ms, &(Rm[0])); ST(&(Rm[0]), T15, -ms, &(Rm[0])); ST(&(Rp[WS(rs, 1)]), T14, ms, &(Rp[WS(rs, 1)])); } } } VLEAVE(); } static const tw_instr twinstr[] = { VTW(1, 1), VTW(1, 2), VTW(1, 3), VTW(1, 4), VTW(1, 5), VTW(1, 6), VTW(1, 7), VTW(1, 8), VTW(1, 9), {TW_NEXT, VL, 0} }; static const hc2c_desc desc = { 10, XSIMD_STRING("hc2cfdftv_10"), twinstr, &GENUS, {33, 32, 28, 0} }; void XSIMD(codelet_hc2cfdftv_10) (planner *p) { X(khc2c_register) (p, hc2cfdftv_10, &desc, HC2C_VIA_DFT); } #else /* HAVE_FMA */ /* Generated by: ../../../genfft/gen_hc2cdft_c.native -simd -compact -variables 4 -pipeline-latency 8 -trivial-stores -variables 32 -no-generate-bytw -n 10 -dit -name hc2cfdftv_10 -include hc2cfv.h */ /* * This function contains 61 FP additions, 38 FP multiplications, * (or, 55 additions, 32 multiplications, 6 fused multiply/add), * 82 stack variables, 5 constants, and 20 memory accesses */ #include "hc2cfv.h" static void hc2cfdftv_10(R *Rp, R *Ip, R *Rm, R *Im, const R *W, stride rs, INT mb, INT me, INT ms) { DVK(KP125000000, +0.125000000000000000000000000000000000000000000); DVK(KP279508497, +0.279508497187473712051146708591409529430077295); DVK(KP587785252, +0.587785252292473129168705954639072768597652438); DVK(KP951056516, +0.951056516295153572116439333379382143405698634); DVK(KP500000000, +0.500000000000000000000000000000000000000000000); { INT m; for (m = mb, W = W + ((mb - 1) * ((TWVL / VL) * 18)); m < me; m = m + VL, Rp = Rp + (VL * ms), Ip = Ip + (VL * ms), Rm = Rm - (VL * ms), Im = Im - (VL * ms), W = W + (TWVL * 18), MAKE_VOLATILE_STRIDE(40, rs)) { V Tl, Tt, Tu, TY, TZ, T10, Tz, TE, TF, TV, TW, TX, Ta, TU, TN; V TR, TH, TQ, TK, TL, TM, TI, TG, TJ, TT, TO, TP, TS, T18, T1c; V T12, T1b, T15, T16, T17, T14, T11, T13, T1e, T19, T1a, T1d; { V T1, T3, Ty, T8, T7, TB, Tf, Ts, Tk, Tw, Tq, TD, T2, Tx, T6; V TA, Tc, Te, Td, Tb, Tr, Tj, Ti, Th, Tg, Tv, Tn, Tp, To, Tm; V TC, T4, T9, T5; T1 = LD(&(Rp[0]), ms, &(Rp[0])); T2 = LD(&(Rm[0]), -ms, &(Rm[0])); T3 = VCONJ(T2); Tx = LDW(&(W[0])); Ty = VZMULIJ(Tx, VSUB(T3, T1)); T8 = LD(&(Rp[WS(rs, 2)]), ms, &(Rp[0])); T6 = LD(&(Rm[WS(rs, 2)]), -ms, &(Rm[0])); T7 = VCONJ(T6); TA = LDW(&(W[TWVL * 6])); TB = VZMULJ(TA, VADD(T7, T8)); Tc = LD(&(Rp[WS(rs, 1)]), ms, &(Rp[WS(rs, 1)])); Td = LD(&(Rm[WS(rs, 1)]), -ms, &(Rm[WS(rs, 1)])); Te = VCONJ(Td); Tb = LDW(&(W[TWVL * 2])); Tf = VZMULJ(Tb, VADD(Tc, Te)); Tr = LDW(&(W[TWVL * 4])); Ts = VZMULIJ(Tr, VSUB(Te, Tc)); Tj = LD(&(Rp[WS(rs, 3)]), ms, &(Rp[WS(rs, 1)])); Th = LD(&(Rm[WS(rs, 3)]), -ms, &(Rm[WS(rs, 1)])); Ti = VCONJ(Th); Tg = LDW(&(W[TWVL * 12])); Tk = VZMULIJ(Tg, VSUB(Ti, Tj)); Tv = LDW(&(W[TWVL * 10])); Tw = VZMULJ(Tv, VADD(Ti, Tj)); Tn = LD(&(Rp[WS(rs, 4)]), ms, &(Rp[0])); To = LD(&(Rm[WS(rs, 4)]), -ms, &(Rm[0])); Tp = VCONJ(To); Tm = LDW(&(W[TWVL * 14])); Tq = VZMULJ(Tm, VADD(Tn, Tp)); TC = LDW(&(W[TWVL * 16])); TD = VZMULIJ(TC, VSUB(Tp, Tn)); Tl = VSUB(Tf, Tk); Tt = VSUB(Tq, Ts); Tu = VADD(Tl, Tt); TY = VADD(Ty, Tw); TZ = VADD(TB, TD); T10 = VADD(TY, TZ); Tz = VSUB(Tw, Ty); TE = VSUB(TB, TD); TF = VADD(Tz, TE); TV = VADD(Tf, Tk); TW = VADD(Ts, Tq); TX = VADD(TV, TW); T4 = VADD(T1, T3); T5 = LDW(&(W[TWVL * 8])); T9 = VZMULIJ(T5, VSUB(T7, T8)); Ta = VSUB(T4, T9); TU = VADD(T4, T9); } TL = VSUB(Tl, Tt); TM = VSUB(TE, Tz); TN = VMUL(LDK(KP500000000), VBYI(VFMA(LDK(KP951056516), TL, VMUL(LDK(KP587785252), TM)))); TR = VMUL(LDK(KP500000000), VBYI(VFNMS(LDK(KP587785252), TL, VMUL(LDK(KP951056516), TM)))); TI = VMUL(LDK(KP279508497), VSUB(Tu, TF)); TG = VADD(Tu, TF); TJ = VFNMS(LDK(KP125000000), TG, VMUL(LDK(KP500000000), Ta)); TH = VCONJ(VMUL(LDK(KP500000000), VADD(Ta, TG))); TQ = VSUB(TJ, TI); TK = VADD(TI, TJ); ST(&(Rm[WS(rs, 4)]), TH, -ms, &(Rm[0])); TT = VCONJ(VADD(TQ, TR)); ST(&(Rm[WS(rs, 2)]), TT, -ms, &(Rm[0])); TO = VSUB(TK, TN); ST(&(Rp[WS(rs, 1)]), TO, ms, &(Rp[WS(rs, 1)])); TP = VCONJ(VADD(TK, TN)); ST(&(Rm[0]), TP, -ms, &(Rm[0])); TS = VSUB(TQ, TR); ST(&(Rp[WS(rs, 3)]), TS, ms, &(Rp[WS(rs, 1)])); T16 = VSUB(TZ, TY); T17 = VSUB(TV, TW); T18 = VMUL(LDK(KP500000000), VBYI(VFNMS(LDK(KP587785252), T17, VMUL(LDK(KP951056516), T16)))); T1c = VMUL(LDK(KP500000000), VBYI(VFMA(LDK(KP951056516), T17, VMUL(LDK(KP587785252), T16)))); T14 = VMUL(LDK(KP279508497), VSUB(TX, T10)); T11 = VADD(TX, T10); T13 = VFNMS(LDK(KP125000000), T11, VMUL(LDK(KP500000000), TU)); T12 = VMUL(LDK(KP500000000), VADD(TU, T11)); T1b = VADD(T14, T13); T15 = VSUB(T13, T14); ST(&(Rp[0]), T12, ms, &(Rp[0])); T1e = VADD(T1b, T1c); ST(&(Rp[WS(rs, 4)]), T1e, ms, &(Rp[0])); T19 = VCONJ(VSUB(T15, T18)); ST(&(Rm[WS(rs, 1)]), T19, -ms, &(Rm[WS(rs, 1)])); T1a = VADD(T15, T18); ST(&(Rp[WS(rs, 2)]), T1a, ms, &(Rp[0])); T1d = VCONJ(VSUB(T1b, T1c)); ST(&(Rm[WS(rs, 3)]), T1d, -ms, &(Rm[WS(rs, 1)])); } } VLEAVE(); } static const tw_instr twinstr[] = { VTW(1, 1), VTW(1, 2), VTW(1, 3), VTW(1, 4), VTW(1, 5), VTW(1, 6), VTW(1, 7), VTW(1, 8), VTW(1, 9), {TW_NEXT, VL, 0} }; static const hc2c_desc desc = { 10, XSIMD_STRING("hc2cfdftv_10"), twinstr, &GENUS, {55, 32, 6, 0} }; void XSIMD(codelet_hc2cfdftv_10) (planner *p) { X(khc2c_register) (p, hc2cfdftv_10, &desc, HC2C_VIA_DFT); } #endif /* HAVE_FMA */ fftw-3.3.3/rdft/simd/common/hc2cfdftv_12.c0000644000175000017500000003071212054411065015112 00000000000000/* * Copyright (c) 2003, 2007-11 Matteo Frigo * Copyright (c) 2003, 2007-11 Massachusetts Institute of Technology * * This program is free software; you can redistribute it and/or modify * it under the terms of the GNU General Public License as published by * the Free Software Foundation; either version 2 of the License, or * (at your option) any later version. * * This program is distributed in the hope that it will be useful, * but WITHOUT ANY WARRANTY; without even the implied warranty of * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the * GNU General Public License for more details. * * You should have received a copy of the GNU General Public License * along with this program; if not, write to the Free Software * Foundation, Inc., 51 Franklin Street, Fifth Floor, Boston, MA 02110-1301 USA * */ /* This file was automatically generated --- DO NOT EDIT */ /* Generated on Sun Nov 25 07:42:29 EST 2012 */ #include "codelet-rdft.h" #ifdef HAVE_FMA /* Generated by: ../../../genfft/gen_hc2cdft_c.native -fma -reorder-insns -schedule-for-pipeline -simd -compact -variables 4 -pipeline-latency 8 -trivial-stores -variables 32 -no-generate-bytw -n 12 -dit -name hc2cfdftv_12 -include hc2cfv.h */ /* * This function contains 71 FP additions, 66 FP multiplications, * (or, 41 additions, 36 multiplications, 30 fused multiply/add), * 86 stack variables, 2 constants, and 24 memory accesses */ #include "hc2cfv.h" static void hc2cfdftv_12(R *Rp, R *Ip, R *Rm, R *Im, const R *W, stride rs, INT mb, INT me, INT ms) { DVK(KP866025403, +0.866025403784438646763723170752936183471402627); DVK(KP500000000, +0.500000000000000000000000000000000000000000000); { INT m; for (m = mb, W = W + ((mb - 1) * ((TWVL / VL) * 22)); m < me; m = m + VL, Rp = Rp + (VL * ms), Ip = Ip + (VL * ms), Rm = Rm - (VL * ms), Im = Im - (VL * ms), W = W + (TWVL * 22), MAKE_VOLATILE_STRIDE(48, rs)) { V T3, T7, TH, TE, Th, TC, Tq, T11, TU, Tx, Tb, Tz, Tu, Tw, Tp; V Tl, T9, Ta, T8, Ty, Tn, To, Tm, TG, T1, T2, Tt, T5, T6, T4; V Tv, Tj, Tk, Ti, TD, Tf, Tg, Te, TB, TT, TF, TR, Tr; T1 = LD(&(Rp[0]), ms, &(Rp[0])); T2 = LD(&(Rm[0]), -ms, &(Rm[0])); Tt = LDW(&(W[0])); T5 = LD(&(Rp[WS(rs, 2)]), ms, &(Rp[0])); T6 = LD(&(Rm[WS(rs, 2)]), -ms, &(Rm[0])); T4 = LDW(&(W[TWVL * 6])); Tv = LDW(&(W[TWVL * 8])); Tn = LD(&(Rp[WS(rs, 1)]), ms, &(Rp[WS(rs, 1)])); To = LD(&(Rm[WS(rs, 1)]), -ms, &(Rm[WS(rs, 1)])); T3 = VFMACONJ(T2, T1); Tu = VZMULIJ(Tt, VFNMSCONJ(T2, T1)); Tm = LDW(&(W[TWVL * 2])); TG = LDW(&(W[TWVL * 4])); T7 = VZMULJ(T4, VFMACONJ(T6, T5)); Tw = VZMULIJ(Tv, VFNMSCONJ(T6, T5)); Tj = LD(&(Rp[WS(rs, 5)]), ms, &(Rp[WS(rs, 1)])); Tk = LD(&(Rm[WS(rs, 5)]), -ms, &(Rm[WS(rs, 1)])); Ti = LDW(&(W[TWVL * 18])); TD = LDW(&(W[TWVL * 20])); Tp = VZMULJ(Tm, VFMACONJ(To, Tn)); TH = VZMULIJ(TG, VFNMSCONJ(To, Tn)); Tf = LD(&(Rp[WS(rs, 3)]), ms, &(Rp[WS(rs, 1)])); Tg = LD(&(Rm[WS(rs, 3)]), -ms, &(Rm[WS(rs, 1)])); Te = LDW(&(W[TWVL * 10])); TB = LDW(&(W[TWVL * 12])); Tl = VZMULJ(Ti, VFMACONJ(Tk, Tj)); TE = VZMULIJ(TD, VFNMSCONJ(Tk, Tj)); T9 = LD(&(Rp[WS(rs, 4)]), ms, &(Rp[0])); Ta = LD(&(Rm[WS(rs, 4)]), -ms, &(Rm[0])); T8 = LDW(&(W[TWVL * 14])); Ty = LDW(&(W[TWVL * 16])); Th = VZMULJ(Te, VFMACONJ(Tg, Tf)); TC = VZMULIJ(TB, VFNMSCONJ(Tg, Tf)); Tq = VADD(Tl, Tp); T11 = VSUB(Tp, Tl); TU = VSUB(Tu, Tw); Tx = VADD(Tu, Tw); Tb = VZMULJ(T8, VFMACONJ(Ta, T9)); Tz = VZMULIJ(Ty, VFNMSCONJ(Ta, T9)); TT = VSUB(TC, TE); TF = VADD(TC, TE); TR = VFNMS(LDK(KP500000000), Tq, Th); Tr = VADD(Th, Tq); { V TX, TA, T1d, TV, TY, TI, T1e, T12, TQ, Td, T10, Tc, T1a, TN, TJ; V T1j, T1f, T1b, TS, TM, Ts, T17, T13, TZ, T1i, T1c, T16, TW, TP, TO; V TL, TK, T1k, T1l, T1h, T1g, T18, T19, T15, T14; T10 = VSUB(Tb, T7); Tc = VADD(T7, Tb); TX = VFNMS(LDK(KP500000000), Tx, Tz); TA = VADD(Tx, Tz); T1d = VADD(TU, TT); TV = VSUB(TT, TU); TY = VFNMS(LDK(KP500000000), TF, TH); TI = VADD(TF, TH); T1e = VADD(T10, T11); T12 = VSUB(T10, T11); TQ = VFNMS(LDK(KP500000000), Tc, T3); Td = VADD(T3, Tc); T1a = VADD(TX, TY); TZ = VSUB(TX, TY); TN = VADD(TA, TI); TJ = VSUB(TA, TI); T1j = VMUL(LDK(KP866025403), VADD(T1d, T1e)); T1f = VMUL(LDK(KP866025403), VSUB(T1d, T1e)); T1b = VADD(TQ, TR); TS = VSUB(TQ, TR); TM = VADD(Td, Tr); Ts = VSUB(Td, Tr); T17 = VFMA(LDK(KP866025403), T12, TZ); T13 = VFNMS(LDK(KP866025403), T12, TZ); T1i = VSUB(T1b, T1a); T1c = VADD(T1a, T1b); T16 = VFNMS(LDK(KP866025403), TV, TS); TW = VFMA(LDK(KP866025403), TV, TS); TP = VCONJ(VMUL(LDK(KP500000000), VADD(TN, TM))); TO = VMUL(LDK(KP500000000), VSUB(TM, TN)); TL = VCONJ(VMUL(LDK(KP500000000), VFNMSI(TJ, Ts))); TK = VMUL(LDK(KP500000000), VFMAI(TJ, Ts)); T1k = VCONJ(VMUL(LDK(KP500000000), VFNMSI(T1j, T1i))); T1l = VMUL(LDK(KP500000000), VFMAI(T1j, T1i)); T1h = VMUL(LDK(KP500000000), VFMAI(T1f, T1c)); T1g = VCONJ(VMUL(LDK(KP500000000), VFNMSI(T1f, T1c))); T18 = VMUL(LDK(KP500000000), VFNMSI(T17, T16)); T19 = VCONJ(VMUL(LDK(KP500000000), VFMAI(T17, T16))); T15 = VCONJ(VMUL(LDK(KP500000000), VFMAI(T13, TW))); T14 = VMUL(LDK(KP500000000), VFNMSI(T13, TW)); ST(&(Rm[WS(rs, 5)]), TP, -ms, &(Rm[WS(rs, 1)])); ST(&(Rp[0]), TO, ms, &(Rp[0])); ST(&(Rm[WS(rs, 2)]), TL, -ms, &(Rm[0])); ST(&(Rp[WS(rs, 3)]), TK, ms, &(Rp[WS(rs, 1)])); ST(&(Rm[WS(rs, 3)]), T1k, -ms, &(Rm[WS(rs, 1)])); ST(&(Rp[WS(rs, 4)]), T1l, ms, &(Rp[0])); ST(&(Rp[WS(rs, 2)]), T1h, ms, &(Rp[0])); ST(&(Rm[WS(rs, 1)]), T1g, -ms, &(Rm[WS(rs, 1)])); ST(&(Rp[WS(rs, 5)]), T18, ms, &(Rp[WS(rs, 1)])); ST(&(Rm[WS(rs, 4)]), T19, -ms, &(Rm[0])); ST(&(Rm[0]), T15, -ms, &(Rm[0])); ST(&(Rp[WS(rs, 1)]), T14, ms, &(Rp[WS(rs, 1)])); } } } VLEAVE(); } static const tw_instr twinstr[] = { VTW(1, 1), VTW(1, 2), VTW(1, 3), VTW(1, 4), VTW(1, 5), VTW(1, 6), VTW(1, 7), VTW(1, 8), VTW(1, 9), VTW(1, 10), VTW(1, 11), {TW_NEXT, VL, 0} }; static const hc2c_desc desc = { 12, XSIMD_STRING("hc2cfdftv_12"), twinstr, &GENUS, {41, 36, 30, 0} }; void XSIMD(codelet_hc2cfdftv_12) (planner *p) { X(khc2c_register) (p, hc2cfdftv_12, &desc, HC2C_VIA_DFT); } #else /* HAVE_FMA */ /* Generated by: ../../../genfft/gen_hc2cdft_c.native -simd -compact -variables 4 -pipeline-latency 8 -trivial-stores -variables 32 -no-generate-bytw -n 12 -dit -name hc2cfdftv_12 -include hc2cfv.h */ /* * This function contains 71 FP additions, 41 FP multiplications, * (or, 67 additions, 37 multiplications, 4 fused multiply/add), * 58 stack variables, 4 constants, and 24 memory accesses */ #include "hc2cfv.h" static void hc2cfdftv_12(R *Rp, R *Ip, R *Rm, R *Im, const R *W, stride rs, INT mb, INT me, INT ms) { DVK(KP433012701, +0.433012701892219323381861585376468091735701313); DVK(KP866025403, +0.866025403784438646763723170752936183471402627); DVK(KP250000000, +0.250000000000000000000000000000000000000000000); DVK(KP500000000, +0.500000000000000000000000000000000000000000000); { INT m; for (m = mb, W = W + ((mb - 1) * ((TWVL / VL) * 22)); m < me; m = m + VL, Rp = Rp + (VL * ms), Ip = Ip + (VL * ms), Rm = Rm - (VL * ms), Im = Im - (VL * ms), W = W + (TWVL * 22), MAKE_VOLATILE_STRIDE(48, rs)) { V TX, T13, T4, Tf, TZ, TD, TF, T17, TW, T14, Tw, Tl, T10, TL, TN; V T16; { V T1, T3, TA, Tb, Td, Te, T9, TC, T2, Tz, Tc, Ta, T6, T8, T7; V T5, TB, TE, Ti, Tk, TI, Ts, Tu, Tv, Tq, TK, Tj, TH, Tt, Tr; V Tn, Tp, To, Tm, TJ, Th, TM; T1 = LD(&(Rp[0]), ms, &(Rp[0])); T2 = LD(&(Rm[0]), -ms, &(Rm[0])); T3 = VCONJ(T2); Tz = LDW(&(W[0])); TA = VZMULIJ(Tz, VSUB(T3, T1)); Tb = LD(&(Rp[WS(rs, 4)]), ms, &(Rp[0])); Tc = LD(&(Rm[WS(rs, 4)]), -ms, &(Rm[0])); Td = VCONJ(Tc); Ta = LDW(&(W[TWVL * 14])); Te = VZMULJ(Ta, VADD(Tb, Td)); T6 = LD(&(Rp[WS(rs, 2)]), ms, &(Rp[0])); T7 = LD(&(Rm[WS(rs, 2)]), -ms, &(Rm[0])); T8 = VCONJ(T7); T5 = LDW(&(W[TWVL * 6])); T9 = VZMULJ(T5, VADD(T6, T8)); TB = LDW(&(W[TWVL * 8])); TC = VZMULIJ(TB, VSUB(T8, T6)); TX = VSUB(TC, TA); T13 = VSUB(Te, T9); T4 = VADD(T1, T3); Tf = VADD(T9, Te); TZ = VFNMS(LDK(KP250000000), Tf, VMUL(LDK(KP500000000), T4)); TD = VADD(TA, TC); TE = LDW(&(W[TWVL * 16])); TF = VZMULIJ(TE, VSUB(Td, Tb)); T17 = VFNMS(LDK(KP500000000), TD, TF); Ti = LD(&(Rp[WS(rs, 3)]), ms, &(Rp[WS(rs, 1)])); Tj = LD(&(Rm[WS(rs, 3)]), -ms, &(Rm[WS(rs, 1)])); Tk = VCONJ(Tj); TH = LDW(&(W[TWVL * 12])); TI = VZMULIJ(TH, VSUB(Tk, Ti)); Ts = LD(&(Rp[WS(rs, 1)]), ms, &(Rp[WS(rs, 1)])); Tt = LD(&(Rm[WS(rs, 1)]), -ms, &(Rm[WS(rs, 1)])); Tu = VCONJ(Tt); Tr = LDW(&(W[TWVL * 2])); Tv = VZMULJ(Tr, VADD(Ts, Tu)); Tn = LD(&(Rp[WS(rs, 5)]), ms, &(Rp[WS(rs, 1)])); To = LD(&(Rm[WS(rs, 5)]), -ms, &(Rm[WS(rs, 1)])); Tp = VCONJ(To); Tm = LDW(&(W[TWVL * 18])); Tq = VZMULJ(Tm, VADD(Tn, Tp)); TJ = LDW(&(W[TWVL * 20])); TK = VZMULIJ(TJ, VSUB(Tp, Tn)); TW = VSUB(TK, TI); T14 = VSUB(Tv, Tq); Tw = VADD(Tq, Tv); Th = LDW(&(W[TWVL * 10])); Tl = VZMULJ(Th, VADD(Ti, Tk)); T10 = VFNMS(LDK(KP250000000), Tw, VMUL(LDK(KP500000000), Tl)); TL = VADD(TI, TK); TM = LDW(&(W[TWVL * 4])); TN = VZMULIJ(TM, VSUB(Tu, Ts)); T16 = VFNMS(LDK(KP500000000), TL, TN); } { V Ty, TS, TP, TT, Tg, Tx, TG, TO, TQ, TV, TR, TU, T1i, T1o, T1l; V T1p, T1g, T1h, T1j, T1k, T1m, T1r, T1n, T1q, T12, T1c, T19, T1d, TY, T11; V T15, T18, T1a, T1f, T1b, T1e; Tg = VADD(T4, Tf); Tx = VADD(Tl, Tw); Ty = VADD(Tg, Tx); TS = VSUB(Tg, Tx); TG = VADD(TD, TF); TO = VADD(TL, TN); TP = VADD(TG, TO); TT = VBYI(VSUB(TO, TG)); TQ = VCONJ(VMUL(LDK(KP500000000), VSUB(Ty, TP))); ST(&(Rm[WS(rs, 5)]), TQ, -ms, &(Rm[WS(rs, 1)])); TV = VMUL(LDK(KP500000000), VADD(TS, TT)); ST(&(Rp[WS(rs, 3)]), TV, ms, &(Rp[WS(rs, 1)])); TR = VMUL(LDK(KP500000000), VADD(Ty, TP)); ST(&(Rp[0]), TR, ms, &(Rp[0])); TU = VCONJ(VMUL(LDK(KP500000000), VSUB(TS, TT))); ST(&(Rm[WS(rs, 2)]), TU, -ms, &(Rm[0])); T1g = VADD(TX, TW); T1h = VADD(T13, T14); T1i = VMUL(LDK(KP500000000), VBYI(VMUL(LDK(KP866025403), VSUB(T1g, T1h)))); T1o = VMUL(LDK(KP500000000), VBYI(VMUL(LDK(KP866025403), VADD(T1g, T1h)))); T1j = VADD(TZ, T10); T1k = VMUL(LDK(KP500000000), VADD(T17, T16)); T1l = VSUB(T1j, T1k); T1p = VADD(T1j, T1k); T1m = VADD(T1i, T1l); ST(&(Rp[WS(rs, 2)]), T1m, ms, &(Rp[0])); T1r = VCONJ(VSUB(T1p, T1o)); ST(&(Rm[WS(rs, 3)]), T1r, -ms, &(Rm[WS(rs, 1)])); T1n = VCONJ(VSUB(T1l, T1i)); ST(&(Rm[WS(rs, 1)]), T1n, -ms, &(Rm[WS(rs, 1)])); T1q = VADD(T1o, T1p); ST(&(Rp[WS(rs, 4)]), T1q, ms, &(Rp[0])); TY = VMUL(LDK(KP433012701), VSUB(TW, TX)); T11 = VSUB(TZ, T10); T12 = VADD(TY, T11); T1c = VSUB(T11, TY); T15 = VMUL(LDK(KP866025403), VSUB(T13, T14)); T18 = VSUB(T16, T17); T19 = VMUL(LDK(KP500000000), VBYI(VSUB(T15, T18))); T1d = VMUL(LDK(KP500000000), VBYI(VADD(T15, T18))); T1a = VCONJ(VSUB(T12, T19)); ST(&(Rm[0]), T1a, -ms, &(Rm[0])); T1f = VCONJ(VADD(T1c, T1d)); ST(&(Rm[WS(rs, 4)]), T1f, -ms, &(Rm[0])); T1b = VADD(T12, T19); ST(&(Rp[WS(rs, 1)]), T1b, ms, &(Rp[WS(rs, 1)])); T1e = VSUB(T1c, T1d); ST(&(Rp[WS(rs, 5)]), T1e, ms, &(Rp[WS(rs, 1)])); } } } VLEAVE(); } static const tw_instr twinstr[] = { VTW(1, 1), VTW(1, 2), VTW(1, 3), VTW(1, 4), VTW(1, 5), VTW(1, 6), VTW(1, 7), VTW(1, 8), VTW(1, 9), VTW(1, 10), VTW(1, 11), {TW_NEXT, VL, 0} }; static const hc2c_desc desc = { 12, XSIMD_STRING("hc2cfdftv_12"), twinstr, &GENUS, {67, 37, 4, 0} }; void XSIMD(codelet_hc2cfdftv_12) (planner *p) { X(khc2c_register) (p, hc2cfdftv_12, &desc, HC2C_VIA_DFT); } #endif /* HAVE_FMA */ fftw-3.3.3/rdft/simd/common/hc2cfdftv_16.c0000644000175000017500000004104412054411066015117 00000000000000/* * Copyright (c) 2003, 2007-11 Matteo Frigo * Copyright (c) 2003, 2007-11 Massachusetts Institute of Technology * * This program is free software; you can redistribute it and/or modify * it under the terms of the GNU General Public License as published by * the Free Software Foundation; either version 2 of the License, or * (at your option) any later version. * * This program is distributed in the hope that it will be useful, * but WITHOUT ANY WARRANTY; without even the implied warranty of * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the * GNU General Public License for more details. * * You should have received a copy of the GNU General Public License * along with this program; if not, write to the Free Software * Foundation, Inc., 51 Franklin Street, Fifth Floor, Boston, MA 02110-1301 USA * */ /* This file was automatically generated --- DO NOT EDIT */ /* Generated on Sun Nov 25 07:42:29 EST 2012 */ #include "codelet-rdft.h" #ifdef HAVE_FMA /* Generated by: ../../../genfft/gen_hc2cdft_c.native -fma -reorder-insns -schedule-for-pipeline -simd -compact -variables 4 -pipeline-latency 8 -trivial-stores -variables 32 -no-generate-bytw -n 16 -dit -name hc2cfdftv_16 -include hc2cfv.h */ /* * This function contains 103 FP additions, 96 FP multiplications, * (or, 53 additions, 46 multiplications, 50 fused multiply/add), * 92 stack variables, 4 constants, and 32 memory accesses */ #include "hc2cfv.h" static void hc2cfdftv_16(R *Rp, R *Ip, R *Rm, R *Im, const R *W, stride rs, INT mb, INT me, INT ms) { DVK(KP500000000, +0.500000000000000000000000000000000000000000000); DVK(KP923879532, +0.923879532511286756128183189396788286822416626); DVK(KP707106781, +0.707106781186547524400844362104849039284835938); DVK(KP414213562, +0.414213562373095048801688724209698078569671875); { INT m; for (m = mb, W = W + ((mb - 1) * ((TWVL / VL) * 30)); m < me; m = m + VL, Rp = Rp + (VL * ms), Ip = Ip + (VL * ms), Rm = Rm - (VL * ms), Im = Im - (VL * ms), W = W + (TWVL * 30), MAKE_VOLATILE_STRIDE(64, rs)) { V T8, Tc, TQ, TZ, T1J, T1x, T12, TH, T1I, T1q, Tp, TJ, Te, Tf, Td; V TN, Tj, Tk, Ti, TK, Tg, TO, Tl, TL, T1r, Th, TR, T1y, T1s, Tq; V TM, T1z, T1N, T1t, T10, Tr, T13, TS, T1K, T1A, T1E, T1u, T1f, T11, T1c; V Ts, T1d, T14, T1g, TT; { V T3, Tw, TF, TW, Tz, TA, Ty, TX, T7, Tu, T1, T2, Tv, TD, TE; V TC, TV, T5, T6, T4, Tt, TB, TY, T1o, T1v, Tx, Ta, Tb, T9, TP; V T1w, TG, T1p, Tn, To, Tm, TI; T1 = LD(&(Rp[0]), ms, &(Rp[0])); T2 = LD(&(Rm[0]), -ms, &(Rm[0])); Tv = LDW(&(W[0])); TD = LD(&(Rp[WS(rs, 2)]), ms, &(Rp[0])); TE = LD(&(Rm[WS(rs, 2)]), -ms, &(Rm[0])); TC = LDW(&(W[TWVL * 8])); TV = LDW(&(W[TWVL * 6])); T5 = LD(&(Rp[WS(rs, 4)]), ms, &(Rp[0])); T6 = LD(&(Rm[WS(rs, 4)]), -ms, &(Rm[0])); T3 = VFMACONJ(T2, T1); Tw = VZMULIJ(Tv, VFNMSCONJ(T2, T1)); T4 = LDW(&(W[TWVL * 14])); Tt = LDW(&(W[TWVL * 16])); TF = VZMULIJ(TC, VFNMSCONJ(TE, TD)); TW = VZMULJ(TV, VFMACONJ(TE, TD)); Tz = LD(&(Rp[WS(rs, 6)]), ms, &(Rp[0])); TA = LD(&(Rm[WS(rs, 6)]), -ms, &(Rm[0])); Ty = LDW(&(W[TWVL * 24])); TX = LDW(&(W[TWVL * 22])); T7 = VZMULJ(T4, VFMACONJ(T6, T5)); Tu = VZMULIJ(Tt, VFNMSCONJ(T6, T5)); Ta = LD(&(Rp[WS(rs, 1)]), ms, &(Rp[WS(rs, 1)])); Tb = LD(&(Rm[WS(rs, 1)]), -ms, &(Rm[WS(rs, 1)])); T9 = LDW(&(W[TWVL * 2])); TP = LDW(&(W[TWVL * 4])); TB = VZMULIJ(Ty, VFNMSCONJ(TA, Tz)); TY = VZMULJ(TX, VFMACONJ(TA, Tz)); T1o = VADD(T3, T7); T8 = VSUB(T3, T7); T1v = VADD(Tw, Tu); Tx = VSUB(Tu, Tw); Tc = VZMULJ(T9, VFMACONJ(Tb, Ta)); TQ = VZMULIJ(TP, VFNMSCONJ(Tb, Ta)); T1w = VADD(TF, TB); TG = VSUB(TB, TF); T1p = VADD(TW, TY); TZ = VSUB(TW, TY); Tn = LD(&(Rp[WS(rs, 3)]), ms, &(Rp[WS(rs, 1)])); To = LD(&(Rm[WS(rs, 3)]), -ms, &(Rm[WS(rs, 1)])); Tm = LDW(&(W[TWVL * 10])); TI = LDW(&(W[TWVL * 12])); T1J = VSUB(T1w, T1v); T1x = VADD(T1v, T1w); T12 = VFMA(LDK(KP414213562), Tx, TG); TH = VFNMS(LDK(KP414213562), TG, Tx); T1I = VSUB(T1o, T1p); T1q = VADD(T1o, T1p); Tp = VZMULJ(Tm, VFMACONJ(To, Tn)); TJ = VZMULIJ(TI, VFNMSCONJ(To, Tn)); Te = LD(&(Rp[WS(rs, 5)]), ms, &(Rp[WS(rs, 1)])); Tf = LD(&(Rm[WS(rs, 5)]), -ms, &(Rm[WS(rs, 1)])); Td = LDW(&(W[TWVL * 18])); TN = LDW(&(W[TWVL * 20])); Tj = LD(&(Rp[WS(rs, 7)]), ms, &(Rp[WS(rs, 1)])); Tk = LD(&(Rm[WS(rs, 7)]), -ms, &(Rm[WS(rs, 1)])); Ti = LDW(&(W[TWVL * 26])); TK = LDW(&(W[TWVL * 28])); } Tg = VZMULJ(Td, VFMACONJ(Tf, Te)); TO = VZMULIJ(TN, VFNMSCONJ(Tf, Te)); Tl = VZMULJ(Ti, VFMACONJ(Tk, Tj)); TL = VZMULIJ(TK, VFNMSCONJ(Tk, Tj)); T1r = VADD(Tc, Tg); Th = VSUB(Tc, Tg); TR = VSUB(TO, TQ); T1y = VADD(TQ, TO); T1s = VADD(Tl, Tp); Tq = VSUB(Tl, Tp); TM = VSUB(TJ, TL); T1z = VADD(TL, TJ); T1N = VSUB(T1s, T1r); T1t = VADD(T1r, T1s); T10 = VSUB(Tq, Th); Tr = VADD(Th, Tq); T13 = VFNMS(LDK(KP414213562), TM, TR); TS = VFMA(LDK(KP414213562), TR, TM); T1K = VSUB(T1y, T1z); T1A = VADD(T1y, T1z); T1E = VADD(T1q, T1t); T1u = VSUB(T1q, T1t); T1f = VFMA(LDK(KP707106781), T10, TZ); T11 = VFNMS(LDK(KP707106781), T10, TZ); T1c = VFNMS(LDK(KP707106781), Tr, T8); Ts = VFMA(LDK(KP707106781), Tr, T8); T1d = VSUB(T12, T13); T14 = VADD(T12, T13); T1g = VSUB(TS, TH); TT = VADD(TH, TS); { V T1O, T1L, T1F, T1B, T1k, T1e, T19, T15, T1l, T1h, T18, TU, T1T, T1P, T1S; V T1M, T1H, T1G, T1D, T1C, T1m, T1n, T1j, T1i, T1a, T1b, T17, T16, T1U, T1V; V T1R, T1Q; T1O = VSUB(T1K, T1J); T1L = VADD(T1J, T1K); T1F = VADD(T1x, T1A); T1B = VSUB(T1x, T1A); T1k = VFNMS(LDK(KP923879532), T1d, T1c); T1e = VFMA(LDK(KP923879532), T1d, T1c); T19 = VFNMS(LDK(KP923879532), T14, T11); T15 = VFMA(LDK(KP923879532), T14, T11); T1l = VFNMS(LDK(KP923879532), T1g, T1f); T1h = VFMA(LDK(KP923879532), T1g, T1f); T18 = VFNMS(LDK(KP923879532), TT, Ts); TU = VFMA(LDK(KP923879532), TT, Ts); T1T = VFNMS(LDK(KP707106781), T1O, T1N); T1P = VFMA(LDK(KP707106781), T1O, T1N); T1S = VFNMS(LDK(KP707106781), T1L, T1I); T1M = VFMA(LDK(KP707106781), T1L, T1I); T1H = VCONJ(VMUL(LDK(KP500000000), VADD(T1F, T1E))); T1G = VMUL(LDK(KP500000000), VSUB(T1E, T1F)); T1D = VCONJ(VMUL(LDK(KP500000000), VFNMSI(T1B, T1u))); T1C = VMUL(LDK(KP500000000), VFMAI(T1B, T1u)); T1m = VMUL(LDK(KP500000000), VFNMSI(T1l, T1k)); T1n = VCONJ(VMUL(LDK(KP500000000), VFMAI(T1l, T1k))); T1j = VMUL(LDK(KP500000000), VFMAI(T1h, T1e)); T1i = VCONJ(VMUL(LDK(KP500000000), VFNMSI(T1h, T1e))); T1a = VCONJ(VMUL(LDK(KP500000000), VFNMSI(T19, T18))); T1b = VMUL(LDK(KP500000000), VFMAI(T19, T18)); T17 = VCONJ(VMUL(LDK(KP500000000), VFMAI(T15, TU))); T16 = VMUL(LDK(KP500000000), VFNMSI(T15, TU)); T1U = VMUL(LDK(KP500000000), VFNMSI(T1T, T1S)); T1V = VCONJ(VMUL(LDK(KP500000000), VFMAI(T1T, T1S))); T1R = VMUL(LDK(KP500000000), VFMAI(T1P, T1M)); T1Q = VCONJ(VMUL(LDK(KP500000000), VFNMSI(T1P, T1M))); ST(&(Rm[WS(rs, 7)]), T1H, -ms, &(Rm[WS(rs, 1)])); ST(&(Rp[0]), T1G, ms, &(Rp[0])); ST(&(Rm[WS(rs, 3)]), T1D, -ms, &(Rm[WS(rs, 1)])); ST(&(Rp[WS(rs, 4)]), T1C, ms, &(Rp[0])); ST(&(Rp[WS(rs, 5)]), T1m, ms, &(Rp[WS(rs, 1)])); ST(&(Rm[WS(rs, 4)]), T1n, -ms, &(Rm[0])); ST(&(Rp[WS(rs, 3)]), T1j, ms, &(Rp[WS(rs, 1)])); ST(&(Rm[WS(rs, 2)]), T1i, -ms, &(Rm[0])); ST(&(Rm[WS(rs, 6)]), T1a, -ms, &(Rm[0])); ST(&(Rp[WS(rs, 7)]), T1b, ms, &(Rp[WS(rs, 1)])); ST(&(Rm[0]), T17, -ms, &(Rm[0])); ST(&(Rp[WS(rs, 1)]), T16, ms, &(Rp[WS(rs, 1)])); ST(&(Rp[WS(rs, 6)]), T1U, ms, &(Rp[0])); ST(&(Rm[WS(rs, 5)]), T1V, -ms, &(Rm[WS(rs, 1)])); ST(&(Rp[WS(rs, 2)]), T1R, ms, &(Rp[0])); ST(&(Rm[WS(rs, 1)]), T1Q, -ms, &(Rm[WS(rs, 1)])); } } } VLEAVE(); } static const tw_instr twinstr[] = { VTW(1, 1), VTW(1, 2), VTW(1, 3), VTW(1, 4), VTW(1, 5), VTW(1, 6), VTW(1, 7), VTW(1, 8), VTW(1, 9), VTW(1, 10), VTW(1, 11), VTW(1, 12), VTW(1, 13), VTW(1, 14), VTW(1, 15), {TW_NEXT, VL, 0} }; static const hc2c_desc desc = { 16, XSIMD_STRING("hc2cfdftv_16"), twinstr, &GENUS, {53, 46, 50, 0} }; void XSIMD(codelet_hc2cfdftv_16) (planner *p) { X(khc2c_register) (p, hc2cfdftv_16, &desc, HC2C_VIA_DFT); } #else /* HAVE_FMA */ /* Generated by: ../../../genfft/gen_hc2cdft_c.native -simd -compact -variables 4 -pipeline-latency 8 -trivial-stores -variables 32 -no-generate-bytw -n 16 -dit -name hc2cfdftv_16 -include hc2cfv.h */ /* * This function contains 103 FP additions, 56 FP multiplications, * (or, 99 additions, 52 multiplications, 4 fused multiply/add), * 101 stack variables, 5 constants, and 32 memory accesses */ #include "hc2cfv.h" static void hc2cfdftv_16(R *Rp, R *Ip, R *Rm, R *Im, const R *W, stride rs, INT mb, INT me, INT ms) { DVK(KP707106781, +0.707106781186547524400844362104849039284835938); DVK(KP353553390, +0.353553390593273762200422181052424519642417969); DVK(KP500000000, +0.500000000000000000000000000000000000000000000); DVK(KP382683432, +0.382683432365089771728459984030398866761344562); DVK(KP923879532, +0.923879532511286756128183189396788286822416626); { INT m; for (m = mb, W = W + ((mb - 1) * ((TWVL / VL) * 30)); m < me; m = m + VL, Rp = Rp + (VL * ms), Ip = Ip + (VL * ms), Rm = Rm - (VL * ms), Im = Im - (VL * ms), W = W + (TWVL * 30), MAKE_VOLATILE_STRIDE(64, rs)) { V T1D, T1E, T1R, TP, T1b, Ta, T1w, T18, T1x, T1z, T1A, T1G, T1H, T1S, Tx; V T13, T10, T1a, T1, T3, TA, TM, TL, TN, T6, T8, TC, TH, TG, TI; V T2, Tz, TK, TJ, T7, TB, TF, TE, TD, TO, T4, T9, T5, T15, T17; V T14, T16; T1 = LD(&(Rp[0]), ms, &(Rp[0])); T2 = LD(&(Rm[0]), -ms, &(Rm[0])); T3 = VCONJ(T2); Tz = LDW(&(W[0])); TA = VZMULIJ(Tz, VSUB(T3, T1)); TM = LD(&(Rp[WS(rs, 6)]), ms, &(Rp[0])); TK = LD(&(Rm[WS(rs, 6)]), -ms, &(Rm[0])); TL = VCONJ(TK); TJ = LDW(&(W[TWVL * 24])); TN = VZMULIJ(TJ, VSUB(TL, TM)); T6 = LD(&(Rp[WS(rs, 4)]), ms, &(Rp[0])); T7 = LD(&(Rm[WS(rs, 4)]), -ms, &(Rm[0])); T8 = VCONJ(T7); TB = LDW(&(W[TWVL * 16])); TC = VZMULIJ(TB, VSUB(T8, T6)); TH = LD(&(Rp[WS(rs, 2)]), ms, &(Rp[0])); TF = LD(&(Rm[WS(rs, 2)]), -ms, &(Rm[0])); TG = VCONJ(TF); TE = LDW(&(W[TWVL * 8])); TI = VZMULIJ(TE, VSUB(TG, TH)); T1D = VADD(TA, TC); T1E = VADD(TI, TN); T1R = VSUB(T1D, T1E); TD = VSUB(TA, TC); TO = VSUB(TI, TN); TP = VFNMS(LDK(KP382683432), TO, VMUL(LDK(KP923879532), TD)); T1b = VFMA(LDK(KP382683432), TD, VMUL(LDK(KP923879532), TO)); T4 = VADD(T1, T3); T5 = LDW(&(W[TWVL * 14])); T9 = VZMULJ(T5, VADD(T6, T8)); Ta = VMUL(LDK(KP500000000), VSUB(T4, T9)); T1w = VADD(T4, T9); T14 = LDW(&(W[TWVL * 6])); T15 = VZMULJ(T14, VADD(TH, TG)); T16 = LDW(&(W[TWVL * 22])); T17 = VZMULJ(T16, VADD(TM, TL)); T18 = VSUB(T15, T17); T1x = VADD(T15, T17); { V Tf, TR, Tv, TY, Tk, TT, Tq, TW, Tc, Te, Td, Tb, TQ, Ts, Tu; V Tt, Tr, TX, Th, Tj, Ti, Tg, TS, Tn, Tp, To, Tm, TV, Tl, Tw; V TU, TZ; Tc = LD(&(Rp[WS(rs, 1)]), ms, &(Rp[WS(rs, 1)])); Td = LD(&(Rm[WS(rs, 1)]), -ms, &(Rm[WS(rs, 1)])); Te = VCONJ(Td); Tb = LDW(&(W[TWVL * 2])); Tf = VZMULJ(Tb, VADD(Tc, Te)); TQ = LDW(&(W[TWVL * 4])); TR = VZMULIJ(TQ, VSUB(Te, Tc)); Ts = LD(&(Rp[WS(rs, 3)]), ms, &(Rp[WS(rs, 1)])); Tt = LD(&(Rm[WS(rs, 3)]), -ms, &(Rm[WS(rs, 1)])); Tu = VCONJ(Tt); Tr = LDW(&(W[TWVL * 10])); Tv = VZMULJ(Tr, VADD(Ts, Tu)); TX = LDW(&(W[TWVL * 12])); TY = VZMULIJ(TX, VSUB(Tu, Ts)); Th = LD(&(Rp[WS(rs, 5)]), ms, &(Rp[WS(rs, 1)])); Ti = LD(&(Rm[WS(rs, 5)]), -ms, &(Rm[WS(rs, 1)])); Tj = VCONJ(Ti); Tg = LDW(&(W[TWVL * 18])); Tk = VZMULJ(Tg, VADD(Th, Tj)); TS = LDW(&(W[TWVL * 20])); TT = VZMULIJ(TS, VSUB(Tj, Th)); Tn = LD(&(Rp[WS(rs, 7)]), ms, &(Rp[WS(rs, 1)])); To = LD(&(Rm[WS(rs, 7)]), -ms, &(Rm[WS(rs, 1)])); Tp = VCONJ(To); Tm = LDW(&(W[TWVL * 26])); Tq = VZMULJ(Tm, VADD(Tn, Tp)); TV = LDW(&(W[TWVL * 28])); TW = VZMULIJ(TV, VSUB(Tp, Tn)); T1z = VADD(Tf, Tk); T1A = VADD(Tq, Tv); T1G = VADD(TR, TT); T1H = VADD(TW, TY); T1S = VSUB(T1H, T1G); Tl = VSUB(Tf, Tk); Tw = VSUB(Tq, Tv); Tx = VMUL(LDK(KP353553390), VADD(Tl, Tw)); T13 = VMUL(LDK(KP707106781), VSUB(Tw, Tl)); TU = VSUB(TR, TT); TZ = VSUB(TW, TY); T10 = VFMA(LDK(KP382683432), TU, VMUL(LDK(KP923879532), TZ)); T1a = VFNMS(LDK(KP923879532), TU, VMUL(LDK(KP382683432), TZ)); } { V T1U, T20, T1X, T21, T1Q, T1T, T1V, T1W, T1Y, T23, T1Z, T22, T1C, T1M, T1J; V T1N, T1y, T1B, T1F, T1I, T1K, T1P, T1L, T1O, T12, T1g, T1d, T1h, Ty, T11; V T19, T1c, T1e, T1j, T1f, T1i, T1m, T1s, T1p, T1t, T1k, T1l, T1n, T1o, T1q; V T1v, T1r, T1u; T1Q = VMUL(LDK(KP500000000), VSUB(T1w, T1x)); T1T = VMUL(LDK(KP353553390), VADD(T1R, T1S)); T1U = VADD(T1Q, T1T); T20 = VSUB(T1Q, T1T); T1V = VSUB(T1A, T1z); T1W = VMUL(LDK(KP707106781), VSUB(T1S, T1R)); T1X = VMUL(LDK(KP500000000), VBYI(VADD(T1V, T1W))); T21 = VMUL(LDK(KP500000000), VBYI(VSUB(T1W, T1V))); T1Y = VCONJ(VSUB(T1U, T1X)); ST(&(Rm[WS(rs, 1)]), T1Y, -ms, &(Rm[WS(rs, 1)])); T23 = VADD(T20, T21); ST(&(Rp[WS(rs, 6)]), T23, ms, &(Rp[0])); T1Z = VADD(T1U, T1X); ST(&(Rp[WS(rs, 2)]), T1Z, ms, &(Rp[0])); T22 = VCONJ(VSUB(T20, T21)); ST(&(Rm[WS(rs, 5)]), T22, -ms, &(Rm[WS(rs, 1)])); T1y = VADD(T1w, T1x); T1B = VADD(T1z, T1A); T1C = VADD(T1y, T1B); T1M = VSUB(T1y, T1B); T1F = VADD(T1D, T1E); T1I = VADD(T1G, T1H); T1J = VADD(T1F, T1I); T1N = VBYI(VSUB(T1I, T1F)); T1K = VCONJ(VMUL(LDK(KP500000000), VSUB(T1C, T1J))); ST(&(Rm[WS(rs, 7)]), T1K, -ms, &(Rm[WS(rs, 1)])); T1P = VMUL(LDK(KP500000000), VADD(T1M, T1N)); ST(&(Rp[WS(rs, 4)]), T1P, ms, &(Rp[0])); T1L = VMUL(LDK(KP500000000), VADD(T1C, T1J)); ST(&(Rp[0]), T1L, ms, &(Rp[0])); T1O = VCONJ(VMUL(LDK(KP500000000), VSUB(T1M, T1N))); ST(&(Rm[WS(rs, 3)]), T1O, -ms, &(Rm[WS(rs, 1)])); Ty = VADD(Ta, Tx); T11 = VMUL(LDK(KP500000000), VADD(TP, T10)); T12 = VADD(Ty, T11); T1g = VSUB(Ty, T11); T19 = VSUB(T13, T18); T1c = VSUB(T1a, T1b); T1d = VMUL(LDK(KP500000000), VBYI(VADD(T19, T1c))); T1h = VMUL(LDK(KP500000000), VBYI(VSUB(T1c, T19))); T1e = VCONJ(VSUB(T12, T1d)); ST(&(Rm[0]), T1e, -ms, &(Rm[0])); T1j = VADD(T1g, T1h); ST(&(Rp[WS(rs, 7)]), T1j, ms, &(Rp[WS(rs, 1)])); T1f = VADD(T12, T1d); ST(&(Rp[WS(rs, 1)]), T1f, ms, &(Rp[WS(rs, 1)])); T1i = VCONJ(VSUB(T1g, T1h)); ST(&(Rm[WS(rs, 6)]), T1i, -ms, &(Rm[0])); T1k = VSUB(T10, TP); T1l = VADD(T18, T13); T1m = VMUL(LDK(KP500000000), VBYI(VSUB(T1k, T1l))); T1s = VMUL(LDK(KP500000000), VBYI(VADD(T1l, T1k))); T1n = VSUB(Ta, Tx); T1o = VMUL(LDK(KP500000000), VADD(T1b, T1a)); T1p = VSUB(T1n, T1o); T1t = VADD(T1n, T1o); T1q = VADD(T1m, T1p); ST(&(Rp[WS(rs, 5)]), T1q, ms, &(Rp[WS(rs, 1)])); T1v = VCONJ(VSUB(T1t, T1s)); ST(&(Rm[WS(rs, 2)]), T1v, -ms, &(Rm[0])); T1r = VCONJ(VSUB(T1p, T1m)); ST(&(Rm[WS(rs, 4)]), T1r, -ms, &(Rm[0])); T1u = VADD(T1s, T1t); ST(&(Rp[WS(rs, 3)]), T1u, ms, &(Rp[WS(rs, 1)])); } } } VLEAVE(); } static const tw_instr twinstr[] = { VTW(1, 1), VTW(1, 2), VTW(1, 3), VTW(1, 4), VTW(1, 5), VTW(1, 6), VTW(1, 7), VTW(1, 8), VTW(1, 9), VTW(1, 10), VTW(1, 11), VTW(1, 12), VTW(1, 13), VTW(1, 14), VTW(1, 15), {TW_NEXT, VL, 0} }; static const hc2c_desc desc = { 16, XSIMD_STRING("hc2cfdftv_16"), twinstr, &GENUS, {99, 52, 4, 0} }; void XSIMD(codelet_hc2cfdftv_16) (planner *p) { X(khc2c_register) (p, hc2cfdftv_16, &desc, HC2C_VIA_DFT); } #endif /* HAVE_FMA */ fftw-3.3.3/rdft/simd/common/hc2cfdftv_32.c0000644000175000017500000010405312054411072015112 00000000000000/* * Copyright (c) 2003, 2007-11 Matteo Frigo * Copyright (c) 2003, 2007-11 Massachusetts Institute of Technology * * This program is free software; you can redistribute it and/or modify * it under the terms of the GNU General Public License as published by * the Free Software Foundation; either version 2 of the License, or * (at your option) any later version. * * This program is distributed in the hope that it will be useful, * but WITHOUT ANY WARRANTY; without even the implied warranty of * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the * GNU General Public License for more details. * * You should have received a copy of the GNU General Public License * along with this program; if not, write to the Free Software * Foundation, Inc., 51 Franklin Street, Fifth Floor, Boston, MA 02110-1301 USA * */ /* This file was automatically generated --- DO NOT EDIT */ /* Generated on Sun Nov 25 07:42:29 EST 2012 */ #include "codelet-rdft.h" #ifdef HAVE_FMA /* Generated by: ../../../genfft/gen_hc2cdft_c.native -fma -reorder-insns -schedule-for-pipeline -simd -compact -variables 4 -pipeline-latency 8 -trivial-stores -variables 32 -no-generate-bytw -n 32 -dit -name hc2cfdftv_32 -include hc2cfv.h */ /* * This function contains 249 FP additions, 224 FP multiplications, * (or, 119 additions, 94 multiplications, 130 fused multiply/add), * 167 stack variables, 8 constants, and 64 memory accesses */ #include "hc2cfv.h" static void hc2cfdftv_32(R *Rp, R *Ip, R *Rm, R *Im, const R *W, stride rs, INT mb, INT me, INT ms) { DVK(KP831469612, +0.831469612302545237078788377617905756738560812); DVK(KP980785280, +0.980785280403230449126182236134239036973933731); DVK(KP668178637, +0.668178637919298919997757686523080761552472251); DVK(KP198912367, +0.198912367379658006911597622644676228597850501); DVK(KP500000000, +0.500000000000000000000000000000000000000000000); DVK(KP923879532, +0.923879532511286756128183189396788286822416626); DVK(KP414213562, +0.414213562373095048801688724209698078569671875); DVK(KP707106781, +0.707106781186547524400844362104849039284835938); { INT m; for (m = mb, W = W + ((mb - 1) * ((TWVL / VL) * 62)); m < me; m = m + VL, Rp = Rp + (VL * ms), Ip = Ip + (VL * ms), Rm = Rm - (VL * ms), Im = Im - (VL * ms), W = W + (TWVL * 62), MAKE_VOLATILE_STRIDE(128, rs)) { V T2m, T2b, T2c, T2d, T2v, T2r, T20, T2i, T2n, T2e, T2o, T2u, T2j, T2f, T2t; V T2s, T2x, T2w, T2l, T2k, T2h, T2g; { V T41, T3B, T40, T3a, T2J, T27, T2y, Ts, T2C, T1X, T2B, T1Q, T3F, T3w, T4l; V T49, T1b, T1s, T3c, TB, T1f, T3g, T44, T1l, T3k, T3o, T4b, T28, T14, T1d; V T3b, TK; { V T1V, T1E, T3A, Th, T3v, T47, T1J, T3q, T8, T38, T25, T39, T3z, Tq, T1O; V T3r, T3, T7, T3u, T24, T22, T3t, T1I, Tn, T1G, To, Tm, T1K, Tl, T1N; V Tp, T1L, TU, T3f, T3m, T13, T3e, T3n, T1i, TH, TI, T1k, TG, TF, T1c; V TJ; { V T1x, T1y, T1U, T1B, T1S, T1C, T1A, T23, T21, T1z, T1, T2, T1T, T5, T6; V T1R, T4, T1w, Ta, Tb, T1H, Te, Tf, Td, Tc, T1F, T9, T1D, Tj, Tk; V Ti, Tg, T1M; T1 = LD(&(Rp[0]), ms, &(Rp[0])); T2 = LD(&(Rm[0]), -ms, &(Rm[0])); T1T = LDW(&(W[0])); T5 = LD(&(Rp[WS(rs, 8)]), ms, &(Rp[0])); T6 = LD(&(Rm[WS(rs, 8)]), -ms, &(Rm[0])); T1R = LDW(&(W[TWVL * 32])); T4 = LDW(&(W[TWVL * 30])); T1x = LD(&(Rp[WS(rs, 12)]), ms, &(Rp[0])); T1y = LD(&(Rm[WS(rs, 12)]), -ms, &(Rm[0])); T3 = VFMACONJ(T2, T1); T1U = VZMULIJ(T1T, VFNMSCONJ(T2, T1)); T1w = LDW(&(W[TWVL * 48])); T1B = LD(&(Rp[WS(rs, 4)]), ms, &(Rp[0])); T1S = VZMULIJ(T1R, VFNMSCONJ(T6, T5)); T7 = VZMULJ(T4, VFMACONJ(T6, T5)); T1C = LD(&(Rm[WS(rs, 4)]), -ms, &(Rm[0])); T1A = LDW(&(W[TWVL * 16])); T23 = LDW(&(W[TWVL * 46])); T21 = LDW(&(W[TWVL * 14])); T1z = VZMULIJ(T1w, VFNMSCONJ(T1y, T1x)); Ta = LD(&(Rp[WS(rs, 2)]), ms, &(Rp[0])); T3u = VADD(T1U, T1S); T1V = VSUB(T1S, T1U); Tb = LD(&(Rm[WS(rs, 2)]), -ms, &(Rm[0])); T9 = LDW(&(W[TWVL * 6])); T1D = VZMULIJ(T1A, VFNMSCONJ(T1C, T1B)); T24 = VZMULJ(T23, VFMACONJ(T1y, T1x)); T22 = VZMULJ(T21, VFMACONJ(T1C, T1B)); T1H = LDW(&(W[TWVL * 8])); Te = LD(&(Rp[WS(rs, 10)]), ms, &(Rp[0])); Tf = LD(&(Rm[WS(rs, 10)]), -ms, &(Rm[0])); Td = LDW(&(W[TWVL * 38])); Tc = VZMULJ(T9, VFMACONJ(Tb, Ta)); T1E = VSUB(T1z, T1D); T3t = VADD(T1D, T1z); T1F = LDW(&(W[TWVL * 40])); Tj = LD(&(Rp[WS(rs, 14)]), ms, &(Rp[0])); T1I = VZMULIJ(T1H, VFNMSCONJ(Tb, Ta)); Tk = LD(&(Rm[WS(rs, 14)]), -ms, &(Rm[0])); Ti = LDW(&(W[TWVL * 54])); Tg = VZMULJ(Td, VFMACONJ(Tf, Te)); T1M = LDW(&(W[TWVL * 56])); Tn = LD(&(Rp[WS(rs, 6)]), ms, &(Rp[0])); T1G = VZMULIJ(T1F, VFNMSCONJ(Tf, Te)); To = LD(&(Rm[WS(rs, 6)]), -ms, &(Rm[0])); Tm = LDW(&(W[TWVL * 22])); T1K = LDW(&(W[TWVL * 24])); Tl = VZMULJ(Ti, VFMACONJ(Tk, Tj)); T3A = VADD(Tc, Tg); Th = VSUB(Tc, Tg); T1N = VZMULIJ(T1M, VFNMSCONJ(Tk, Tj)); } T3v = VSUB(T3t, T3u); T47 = VADD(T3u, T3t); T1J = VSUB(T1G, T1I); T3q = VADD(T1I, T1G); Tp = VZMULJ(Tm, VFMACONJ(To, Tn)); T1L = VZMULIJ(T1K, VFNMSCONJ(To, Tn)); T8 = VSUB(T3, T7); T38 = VADD(T3, T7); T25 = VSUB(T22, T24); T39 = VADD(T22, T24); T3z = VADD(Tl, Tp); Tq = VSUB(Tl, Tp); T1O = VSUB(T1L, T1N); T3r = VADD(T1N, T1L); { V T10, T11, TZ, T1o, TY, T1r, TN, TO, TM, T19, TR, TS, TQ, T17, T26; V Tr, T1W, T1P, T3s, T48, TW, TX, TP, T1a, TV, T1q, TT, T18, Ty, Tz; V Tx, Tw, T1j, Tu, T12, T1p, Tv, Tt, T1h, TD, TA, TE, TC, T1e; TN = LD(&(Rp[WS(rs, 3)]), ms, &(Rp[WS(rs, 1)])); TO = LD(&(Rm[WS(rs, 3)]), -ms, &(Rm[WS(rs, 1)])); T41 = VADD(T3A, T3z); T3B = VSUB(T3z, T3A); T26 = VSUB(Tq, Th); Tr = VADD(Th, Tq); T1W = VADD(T1J, T1O); T1P = VSUB(T1J, T1O); T3s = VSUB(T3q, T3r); T48 = VADD(T3q, T3r); T40 = VADD(T38, T39); T3a = VSUB(T38, T39); T2J = VFNMS(LDK(KP707106781), T26, T25); T27 = VFMA(LDK(KP707106781), T26, T25); T2y = VFMA(LDK(KP707106781), Tr, T8); Ts = VFNMS(LDK(KP707106781), Tr, T8); T2C = VFMA(LDK(KP707106781), T1W, T1V); T1X = VFNMS(LDK(KP707106781), T1W, T1V); T2B = VFMA(LDK(KP707106781), T1P, T1E); T1Q = VFNMS(LDK(KP707106781), T1P, T1E); T3F = VFMA(LDK(KP414213562), T3s, T3v); T3w = VFNMS(LDK(KP414213562), T3v, T3s); T4l = VSUB(T48, T47); T49 = VADD(T47, T48); TM = LDW(&(W[TWVL * 10])); T19 = LDW(&(W[TWVL * 12])); TR = LD(&(Rp[WS(rs, 11)]), ms, &(Rp[WS(rs, 1)])); TS = LD(&(Rm[WS(rs, 11)]), -ms, &(Rm[WS(rs, 1)])); TQ = LDW(&(W[TWVL * 42])); T17 = LDW(&(W[TWVL * 44])); TW = LD(&(Rp[WS(rs, 15)]), ms, &(Rp[WS(rs, 1)])); TX = LD(&(Rm[WS(rs, 15)]), -ms, &(Rm[WS(rs, 1)])); TP = VZMULJ(TM, VFMACONJ(TO, TN)); T1a = VZMULIJ(T19, VFNMSCONJ(TO, TN)); TV = LDW(&(W[TWVL * 58])); T1q = LDW(&(W[TWVL * 60])); TT = VZMULJ(TQ, VFMACONJ(TS, TR)); T18 = VZMULIJ(T17, VFNMSCONJ(TS, TR)); T10 = LD(&(Rp[WS(rs, 7)]), ms, &(Rp[WS(rs, 1)])); T11 = LD(&(Rm[WS(rs, 7)]), -ms, &(Rm[WS(rs, 1)])); TZ = LDW(&(W[TWVL * 26])); T1o = LDW(&(W[TWVL * 28])); TY = VZMULJ(TV, VFMACONJ(TX, TW)); T1r = VZMULIJ(T1q, VFNMSCONJ(TX, TW)); TU = VSUB(TP, TT); T3f = VADD(TP, TT); T1b = VSUB(T18, T1a); T3m = VADD(T1a, T18); Tu = LD(&(Rp[WS(rs, 5)]), ms, &(Rp[WS(rs, 1)])); T12 = VZMULJ(TZ, VFMACONJ(T11, T10)); T1p = VZMULIJ(T1o, VFNMSCONJ(T11, T10)); Tv = LD(&(Rm[WS(rs, 5)]), -ms, &(Rm[WS(rs, 1)])); Tt = LDW(&(W[TWVL * 18])); T1h = LDW(&(W[TWVL * 20])); Ty = LD(&(Rp[WS(rs, 13)]), ms, &(Rp[WS(rs, 1)])); Tz = LD(&(Rm[WS(rs, 13)]), -ms, &(Rm[WS(rs, 1)])); Tx = LDW(&(W[TWVL * 50])); T13 = VSUB(TY, T12); T3e = VADD(TY, T12); T1s = VSUB(T1p, T1r); T3n = VADD(T1r, T1p); Tw = VZMULJ(Tt, VFMACONJ(Tv, Tu)); T1i = VZMULIJ(T1h, VFNMSCONJ(Tv, Tu)); T1j = LDW(&(W[TWVL * 52])); TD = LD(&(Rp[WS(rs, 1)]), ms, &(Rp[WS(rs, 1)])); TA = VZMULJ(Tx, VFMACONJ(Tz, Ty)); TE = LD(&(Rm[WS(rs, 1)]), -ms, &(Rm[WS(rs, 1)])); TC = LDW(&(W[TWVL * 2])); T1e = LDW(&(W[TWVL * 4])); TH = LD(&(Rp[WS(rs, 9)]), ms, &(Rp[WS(rs, 1)])); TI = LD(&(Rm[WS(rs, 9)]), -ms, &(Rm[WS(rs, 1)])); T1k = VZMULIJ(T1j, VFNMSCONJ(Tz, Ty)); TG = LDW(&(W[TWVL * 34])); T3c = VADD(Tw, TA); TB = VSUB(Tw, TA); TF = VZMULJ(TC, VFMACONJ(TE, TD)); T1f = VZMULIJ(T1e, VFNMSCONJ(TE, TD)); T1c = LDW(&(W[TWVL * 36])); } T3g = VSUB(T3e, T3f); T44 = VADD(T3e, T3f); T1l = VSUB(T1i, T1k); T3k = VADD(T1i, T1k); TJ = VZMULJ(TG, VFMACONJ(TI, TH)); T3o = VSUB(T3m, T3n); T4b = VADD(T3n, T3m); T28 = VFMA(LDK(KP414213562), TU, T13); T14 = VFNMS(LDK(KP414213562), T13, TU); T1d = VZMULIJ(T1c, VFNMSCONJ(TI, TH)); T3b = VADD(TF, TJ); TK = VSUB(TF, TJ); } { V T4k, T4p, T2z, T2a, T2K, T15, T2E, T1n, T2F, T1u, T4c, T3R, T3D, T3i, T3O; V T46, T4g, T3G, T3P, T3S, T3x, T4q, T4n, T42, T1g, T3j, T3E, T3p, T4m, T3d; V T43, T29, TL, T1m, T1t, T3l, T4a, T3C, T3h, T45, T3Q, T3W, T4d, T4h, T3H; V T3L, T3y, T3K, T4r, T4v, T4o, T4u, T4j, T4i, T4e, T4f, T3N, T3M, T3I, T3J; V T4x, T4w, T4s, T4t; T42 = VADD(T40, T41); T4k = VSUB(T40, T41); T1g = VSUB(T1d, T1f); T3j = VADD(T1f, T1d); T3d = VSUB(T3b, T3c); T43 = VADD(T3b, T3c); T29 = VFNMS(LDK(KP414213562), TB, TK); TL = VFMA(LDK(KP414213562), TK, TB); T1m = VSUB(T1g, T1l); T1t = VADD(T1g, T1l); T3l = VSUB(T3j, T3k); T4a = VADD(T3j, T3k); T3C = VSUB(T3g, T3d); T3h = VADD(T3d, T3g); T45 = VADD(T43, T44); T4p = VSUB(T44, T43); T2z = VADD(T29, T28); T2a = VSUB(T28, T29); T2K = VADD(TL, T14); T15 = VSUB(TL, T14); T2E = VFMA(LDK(KP707106781), T1m, T1b); T1n = VFNMS(LDK(KP707106781), T1m, T1b); T2F = VFMA(LDK(KP707106781), T1t, T1s); T1u = VFNMS(LDK(KP707106781), T1t, T1s); T3E = VFNMS(LDK(KP414213562), T3l, T3o); T3p = VFMA(LDK(KP414213562), T3o, T3l); T4m = VSUB(T4a, T4b); T4c = VADD(T4a, T4b); T3R = VFMA(LDK(KP707106781), T3C, T3B); T3D = VFNMS(LDK(KP707106781), T3C, T3B); T3i = VFNMS(LDK(KP707106781), T3h, T3a); T3O = VFMA(LDK(KP707106781), T3h, T3a); T46 = VSUB(T42, T45); T4g = VADD(T42, T45); T3G = VSUB(T3E, T3F); T3P = VADD(T3F, T3E); T3S = VADD(T3w, T3p); T3x = VSUB(T3p, T3w); T4q = VSUB(T4m, T4l); T4n = VADD(T4l, T4m); T4d = VSUB(T49, T4c); T4h = VADD(T49, T4c); T3H = VFNMS(LDK(KP923879532), T3G, T3D); T3L = VFMA(LDK(KP923879532), T3G, T3D); T3y = VFMA(LDK(KP923879532), T3x, T3i); T3K = VFNMS(LDK(KP923879532), T3x, T3i); T4r = VFMA(LDK(KP707106781), T4q, T4p); T4v = VFNMS(LDK(KP707106781), T4q, T4p); T4o = VFMA(LDK(KP707106781), T4n, T4k); T4u = VFNMS(LDK(KP707106781), T4n, T4k); T3Q = VFMA(LDK(KP923879532), T3P, T3O); T3W = VFNMS(LDK(KP923879532), T3P, T3O); T4j = VCONJ(VMUL(LDK(KP500000000), VADD(T4h, T4g))); T4i = VMUL(LDK(KP500000000), VSUB(T4g, T4h)); T4e = VMUL(LDK(KP500000000), VFMAI(T4d, T46)); T4f = VCONJ(VMUL(LDK(KP500000000), VFNMSI(T4d, T46))); T3N = VMUL(LDK(KP500000000), VFMAI(T3L, T3K)); T3M = VCONJ(VMUL(LDK(KP500000000), VFNMSI(T3L, T3K))); T3I = VMUL(LDK(KP500000000), VFNMSI(T3H, T3y)); T3J = VCONJ(VMUL(LDK(KP500000000), VFMAI(T3H, T3y))); T4x = VCONJ(VMUL(LDK(KP500000000), VFMAI(T4v, T4u))); T4w = VMUL(LDK(KP500000000), VFNMSI(T4v, T4u)); T4s = VCONJ(VMUL(LDK(KP500000000), VFNMSI(T4r, T4o))); T4t = VMUL(LDK(KP500000000), VFMAI(T4r, T4o)); ST(&(Rp[0]), T4i, ms, &(Rp[0])); ST(&(Rm[WS(rs, 15)]), T4j, -ms, &(Rm[WS(rs, 1)])); ST(&(Rm[WS(rs, 7)]), T4f, -ms, &(Rm[WS(rs, 1)])); ST(&(Rp[WS(rs, 8)]), T4e, ms, &(Rp[0])); ST(&(Rm[WS(rs, 9)]), T3M, -ms, &(Rm[WS(rs, 1)])); ST(&(Rp[WS(rs, 10)]), T3N, ms, &(Rp[0])); ST(&(Rm[WS(rs, 5)]), T3J, -ms, &(Rm[WS(rs, 1)])); ST(&(Rp[WS(rs, 6)]), T3I, ms, &(Rp[0])); ST(&(Rp[WS(rs, 12)]), T4w, ms, &(Rp[0])); ST(&(Rm[WS(rs, 11)]), T4x, -ms, &(Rm[WS(rs, 1)])); ST(&(Rp[WS(rs, 4)]), T4t, ms, &(Rp[0])); ST(&(Rm[WS(rs, 3)]), T4s, -ms, &(Rm[WS(rs, 1)])); { V T2A, T2W, T2L, T2Z, T2D, T2N, T2M, T2G, T3T, T3X, T16, T2p, T1v, T35, T31; V T2I, T2S, T34, T2Y, T2P, T2T, T1Y, T2H, T30, T3Z, T3Y, T3U, T3V, T2O, T2X; V T32, T33, T36, T37, T2U, T2V, T2Q, T2R, T1Z, T2q; T2A = VFNMS(LDK(KP923879532), T2z, T2y); T2W = VFMA(LDK(KP923879532), T2z, T2y); T2L = VFNMS(LDK(KP923879532), T2K, T2J); T2Z = VFMA(LDK(KP923879532), T2K, T2J); T2D = VFMA(LDK(KP198912367), T2C, T2B); T2N = VFNMS(LDK(KP198912367), T2B, T2C); T2M = VFMA(LDK(KP198912367), T2E, T2F); T2G = VFNMS(LDK(KP198912367), T2F, T2E); T3T = VFMA(LDK(KP923879532), T3S, T3R); T3X = VFNMS(LDK(KP923879532), T3S, T3R); T16 = VFNMS(LDK(KP923879532), T15, Ts); T2m = VFMA(LDK(KP923879532), T15, Ts); T2H = VSUB(T2D, T2G); T30 = VADD(T2D, T2G); T2b = VFNMS(LDK(KP923879532), T2a, T27); T2p = VFMA(LDK(KP923879532), T2a, T27); T1v = VFMA(LDK(KP668178637), T1u, T1n); T2c = VFNMS(LDK(KP668178637), T1n, T1u); T3Z = VCONJ(VMUL(LDK(KP500000000), VFMAI(T3X, T3W))); T3Y = VMUL(LDK(KP500000000), VFNMSI(T3X, T3W)); T3U = VCONJ(VMUL(LDK(KP500000000), VFNMSI(T3T, T3Q))); T3V = VMUL(LDK(KP500000000), VFMAI(T3T, T3Q)); T2O = VSUB(T2M, T2N); T2X = VADD(T2N, T2M); T35 = VFNMS(LDK(KP980785280), T30, T2Z); T31 = VFMA(LDK(KP980785280), T30, T2Z); T2I = VFMA(LDK(KP980785280), T2H, T2A); T2S = VFNMS(LDK(KP980785280), T2H, T2A); ST(&(Rp[WS(rs, 14)]), T3Y, ms, &(Rp[0])); ST(&(Rm[WS(rs, 13)]), T3Z, -ms, &(Rm[WS(rs, 1)])); ST(&(Rp[WS(rs, 2)]), T3V, ms, &(Rp[0])); ST(&(Rm[WS(rs, 1)]), T3U, -ms, &(Rm[WS(rs, 1)])); T34 = VFNMS(LDK(KP980785280), T2X, T2W); T2Y = VFMA(LDK(KP980785280), T2X, T2W); T2P = VFMA(LDK(KP980785280), T2O, T2L); T2T = VFNMS(LDK(KP980785280), T2O, T2L); T2d = VFMA(LDK(KP668178637), T1Q, T1X); T1Y = VFNMS(LDK(KP668178637), T1X, T1Q); T32 = VMUL(LDK(KP500000000), VFNMSI(T31, T2Y)); T33 = VCONJ(VMUL(LDK(KP500000000), VFMAI(T31, T2Y))); T36 = VCONJ(VMUL(LDK(KP500000000), VFNMSI(T35, T34))); T37 = VMUL(LDK(KP500000000), VFMAI(T35, T34)); T2U = VMUL(LDK(KP500000000), VFNMSI(T2T, T2S)); T2V = VCONJ(VMUL(LDK(KP500000000), VFMAI(T2T, T2S))); T2Q = VCONJ(VMUL(LDK(KP500000000), VFNMSI(T2P, T2I))); T2R = VMUL(LDK(KP500000000), VFMAI(T2P, T2I)); T1Z = VSUB(T1v, T1Y); T2q = VADD(T1Y, T1v); ST(&(Rm[0]), T33, -ms, &(Rm[0])); ST(&(Rp[WS(rs, 1)]), T32, ms, &(Rp[WS(rs, 1)])); ST(&(Rp[WS(rs, 15)]), T37, ms, &(Rp[WS(rs, 1)])); ST(&(Rm[WS(rs, 14)]), T36, -ms, &(Rm[0])); ST(&(Rm[WS(rs, 8)]), T2V, -ms, &(Rm[0])); ST(&(Rp[WS(rs, 9)]), T2U, ms, &(Rp[WS(rs, 1)])); ST(&(Rp[WS(rs, 7)]), T2R, ms, &(Rp[WS(rs, 1)])); ST(&(Rm[WS(rs, 6)]), T2Q, -ms, &(Rm[0])); T2v = VFNMS(LDK(KP831469612), T2q, T2p); T2r = VFMA(LDK(KP831469612), T2q, T2p); T20 = VFMA(LDK(KP831469612), T1Z, T16); T2i = VFNMS(LDK(KP831469612), T1Z, T16); } } } T2n = VADD(T2d, T2c); T2e = VSUB(T2c, T2d); T2o = VFMA(LDK(KP831469612), T2n, T2m); T2u = VFNMS(LDK(KP831469612), T2n, T2m); T2j = VFMA(LDK(KP831469612), T2e, T2b); T2f = VFNMS(LDK(KP831469612), T2e, T2b); T2t = VCONJ(VMUL(LDK(KP500000000), VFNMSI(T2r, T2o))); T2s = VMUL(LDK(KP500000000), VFMAI(T2r, T2o)); T2x = VCONJ(VMUL(LDK(KP500000000), VFMAI(T2v, T2u))); T2w = VMUL(LDK(KP500000000), VFNMSI(T2v, T2u)); T2l = VCONJ(VMUL(LDK(KP500000000), VFNMSI(T2j, T2i))); T2k = VMUL(LDK(KP500000000), VFMAI(T2j, T2i)); T2h = VCONJ(VMUL(LDK(KP500000000), VFMAI(T2f, T20))); T2g = VMUL(LDK(KP500000000), VFNMSI(T2f, T20)); ST(&(Rm[WS(rs, 2)]), T2t, -ms, &(Rm[0])); ST(&(Rp[WS(rs, 3)]), T2s, ms, &(Rp[WS(rs, 1)])); ST(&(Rm[WS(rs, 12)]), T2x, -ms, &(Rm[0])); ST(&(Rp[WS(rs, 13)]), T2w, ms, &(Rp[WS(rs, 1)])); ST(&(Rm[WS(rs, 10)]), T2l, -ms, &(Rm[0])); ST(&(Rp[WS(rs, 11)]), T2k, ms, &(Rp[WS(rs, 1)])); ST(&(Rm[WS(rs, 4)]), T2h, -ms, &(Rm[0])); ST(&(Rp[WS(rs, 5)]), T2g, ms, &(Rp[WS(rs, 1)])); } } VLEAVE(); } static const tw_instr twinstr[] = { VTW(1, 1), VTW(1, 2), VTW(1, 3), VTW(1, 4), VTW(1, 5), VTW(1, 6), VTW(1, 7), VTW(1, 8), VTW(1, 9), VTW(1, 10), VTW(1, 11), VTW(1, 12), VTW(1, 13), VTW(1, 14), VTW(1, 15), VTW(1, 16), VTW(1, 17), VTW(1, 18), VTW(1, 19), VTW(1, 20), VTW(1, 21), VTW(1, 22), VTW(1, 23), VTW(1, 24), VTW(1, 25), VTW(1, 26), VTW(1, 27), VTW(1, 28), VTW(1, 29), VTW(1, 30), VTW(1, 31), {TW_NEXT, VL, 0} }; static const hc2c_desc desc = { 32, XSIMD_STRING("hc2cfdftv_32"), twinstr, &GENUS, {119, 94, 130, 0} }; void XSIMD(codelet_hc2cfdftv_32) (planner *p) { X(khc2c_register) (p, hc2cfdftv_32, &desc, HC2C_VIA_DFT); } #else /* HAVE_FMA */ /* Generated by: ../../../genfft/gen_hc2cdft_c.native -simd -compact -variables 4 -pipeline-latency 8 -trivial-stores -variables 32 -no-generate-bytw -n 32 -dit -name hc2cfdftv_32 -include hc2cfv.h */ /* * This function contains 249 FP additions, 133 FP multiplications, * (or, 233 additions, 117 multiplications, 16 fused multiply/add), * 130 stack variables, 9 constants, and 64 memory accesses */ #include "hc2cfv.h" static void hc2cfdftv_32(R *Rp, R *Ip, R *Rm, R *Im, const R *W, stride rs, INT mb, INT me, INT ms) { DVK(KP555570233, +0.555570233019602224742830813948532874374937191); DVK(KP831469612, +0.831469612302545237078788377617905756738560812); DVK(KP195090322, +0.195090322016128267848284868477022240927691618); DVK(KP980785280, +0.980785280403230449126182236134239036973933731); DVK(KP382683432, +0.382683432365089771728459984030398866761344562); DVK(KP923879532, +0.923879532511286756128183189396788286822416626); DVK(KP707106781, +0.707106781186547524400844362104849039284835938); DVK(KP353553390, +0.353553390593273762200422181052424519642417969); DVK(KP500000000, +0.500000000000000000000000000000000000000000000); { INT m; for (m = mb, W = W + ((mb - 1) * ((TWVL / VL) * 62)); m < me; m = m + VL, Rp = Rp + (VL * ms), Ip = Ip + (VL * ms), Rm = Rm - (VL * ms), Im = Im - (VL * ms), W = W + (TWVL * 62), MAKE_VOLATILE_STRIDE(128, rs)) { V Ta, T2m, Tx, T2h, T3R, T4h, T3q, T4g, T3B, T4n, T3E, T4o, T1B, T2S, T1O; V T2R, TV, T2p, T1i, T2o, T3L, T4q, T3I, T4r, T3w, T4k, T3t, T4j, T26, T2V; V T2d, T2U; { V T4, T1m, T1H, T2j, T1M, T2l, T9, T1o, Tf, T1r, Tq, T1w, Tv, T1y, Tk; V T1t, Tl, Tw, T3P, T3Q, T3o, T3p, T3z, T3A, T3C, T3D, T1p, T1N, T1A, T1C; V T1u, T1z; { V T1, T3, T2, T1l, T1G, T1F, T1E, T1D, T2i, T1L, T1K, T1J, T1I, T2k, T6; V T8, T7, T5, T1n, Tc, Te, Td, Tb, T1q, Tn, Tp, To, Tm, T1v, Ts; V Tu, Tt, Tr, T1x, Th, Tj, Ti, Tg, T1s; T1 = LD(&(Rp[0]), ms, &(Rp[0])); T2 = LD(&(Rm[0]), -ms, &(Rm[0])); T3 = VCONJ(T2); T4 = VADD(T1, T3); T1l = LDW(&(W[0])); T1m = VZMULIJ(T1l, VSUB(T3, T1)); T1G = LD(&(Rp[WS(rs, 4)]), ms, &(Rp[0])); T1E = LD(&(Rm[WS(rs, 4)]), -ms, &(Rm[0])); T1F = VCONJ(T1E); T1D = LDW(&(W[TWVL * 16])); T1H = VZMULIJ(T1D, VSUB(T1F, T1G)); T2i = LDW(&(W[TWVL * 14])); T2j = VZMULJ(T2i, VADD(T1G, T1F)); T1L = LD(&(Rp[WS(rs, 12)]), ms, &(Rp[0])); T1J = LD(&(Rm[WS(rs, 12)]), -ms, &(Rm[0])); T1K = VCONJ(T1J); T1I = LDW(&(W[TWVL * 48])); T1M = VZMULIJ(T1I, VSUB(T1K, T1L)); T2k = LDW(&(W[TWVL * 46])); T2l = VZMULJ(T2k, VADD(T1L, T1K)); T6 = LD(&(Rp[WS(rs, 8)]), ms, &(Rp[0])); T7 = LD(&(Rm[WS(rs, 8)]), -ms, &(Rm[0])); T8 = VCONJ(T7); T5 = LDW(&(W[TWVL * 30])); T9 = VZMULJ(T5, VADD(T6, T8)); T1n = LDW(&(W[TWVL * 32])); T1o = VZMULIJ(T1n, VSUB(T8, T6)); Tc = LD(&(Rp[WS(rs, 2)]), ms, &(Rp[0])); Td = LD(&(Rm[WS(rs, 2)]), -ms, &(Rm[0])); Te = VCONJ(Td); Tb = LDW(&(W[TWVL * 6])); Tf = VZMULJ(Tb, VADD(Tc, Te)); T1q = LDW(&(W[TWVL * 8])); T1r = VZMULIJ(T1q, VSUB(Te, Tc)); Tn = LD(&(Rp[WS(rs, 14)]), ms, &(Rp[0])); To = LD(&(Rm[WS(rs, 14)]), -ms, &(Rm[0])); Tp = VCONJ(To); Tm = LDW(&(W[TWVL * 54])); Tq = VZMULJ(Tm, VADD(Tn, Tp)); T1v = LDW(&(W[TWVL * 56])); T1w = VZMULIJ(T1v, VSUB(Tp, Tn)); Ts = LD(&(Rp[WS(rs, 6)]), ms, &(Rp[0])); Tt = LD(&(Rm[WS(rs, 6)]), -ms, &(Rm[0])); Tu = VCONJ(Tt); Tr = LDW(&(W[TWVL * 22])); Tv = VZMULJ(Tr, VADD(Ts, Tu)); T1x = LDW(&(W[TWVL * 24])); T1y = VZMULIJ(T1x, VSUB(Tu, Ts)); Th = LD(&(Rp[WS(rs, 10)]), ms, &(Rp[0])); Ti = LD(&(Rm[WS(rs, 10)]), -ms, &(Rm[0])); Tj = VCONJ(Ti); Tg = LDW(&(W[TWVL * 38])); Tk = VZMULJ(Tg, VADD(Th, Tj)); T1s = LDW(&(W[TWVL * 40])); T1t = VZMULIJ(T1s, VSUB(Tj, Th)); } Ta = VMUL(LDK(KP500000000), VSUB(T4, T9)); T2m = VSUB(T2j, T2l); Tl = VSUB(Tf, Tk); Tw = VSUB(Tq, Tv); Tx = VMUL(LDK(KP353553390), VADD(Tl, Tw)); T2h = VMUL(LDK(KP707106781), VSUB(Tw, Tl)); T3P = VADD(Tq, Tv); T3Q = VADD(Tf, Tk); T3R = VSUB(T3P, T3Q); T4h = VADD(T3Q, T3P); T3o = VADD(T4, T9); T3p = VADD(T2j, T2l); T3q = VMUL(LDK(KP500000000), VSUB(T3o, T3p)); T4g = VADD(T3o, T3p); T3z = VADD(T1m, T1o); T3A = VADD(T1H, T1M); T3B = VSUB(T3z, T3A); T4n = VADD(T3z, T3A); T3C = VADD(T1w, T1y); T3D = VADD(T1r, T1t); T3E = VSUB(T3C, T3D); T4o = VADD(T3D, T3C); T1p = VSUB(T1m, T1o); T1N = VSUB(T1H, T1M); T1u = VSUB(T1r, T1t); T1z = VSUB(T1w, T1y); T1A = VMUL(LDK(KP707106781), VADD(T1u, T1z)); T1C = VMUL(LDK(KP707106781), VSUB(T1z, T1u)); T1B = VADD(T1p, T1A); T2S = VADD(T1N, T1C); T1O = VSUB(T1C, T1N); T2R = VSUB(T1p, T1A); } { V TD, T1R, T1b, T29, T1g, T2b, TI, T1T, TO, T1Y, T10, T22, T15, T24, TT; V T1W, TJ, TU, T16, T1h, T3J, T3K, T3G, T3H, T3u, T3v, T3r, T3s, T25, T2c; V T20, T27, T1U, T1Z; { V TA, TC, TB, Tz, T1Q, T18, T1a, T19, T17, T28, T1d, T1f, T1e, T1c, T2a; V TF, TH, TG, TE, T1S, TL, TN, TM, TK, T1X, TX, TZ, TY, TW, T21; V T12, T14, T13, T11, T23, TQ, TS, TR, TP, T1V; TA = LD(&(Rp[WS(rs, 1)]), ms, &(Rp[WS(rs, 1)])); TB = LD(&(Rm[WS(rs, 1)]), -ms, &(Rm[WS(rs, 1)])); TC = VCONJ(TB); Tz = LDW(&(W[TWVL * 2])); TD = VZMULJ(Tz, VADD(TA, TC)); T1Q = LDW(&(W[TWVL * 4])); T1R = VZMULIJ(T1Q, VSUB(TC, TA)); T18 = LD(&(Rp[WS(rs, 3)]), ms, &(Rp[WS(rs, 1)])); T19 = LD(&(Rm[WS(rs, 3)]), -ms, &(Rm[WS(rs, 1)])); T1a = VCONJ(T19); T17 = LDW(&(W[TWVL * 10])); T1b = VZMULJ(T17, VADD(T18, T1a)); T28 = LDW(&(W[TWVL * 12])); T29 = VZMULIJ(T28, VSUB(T1a, T18)); T1d = LD(&(Rp[WS(rs, 11)]), ms, &(Rp[WS(rs, 1)])); T1e = LD(&(Rm[WS(rs, 11)]), -ms, &(Rm[WS(rs, 1)])); T1f = VCONJ(T1e); T1c = LDW(&(W[TWVL * 42])); T1g = VZMULJ(T1c, VADD(T1d, T1f)); T2a = LDW(&(W[TWVL * 44])); T2b = VZMULIJ(T2a, VSUB(T1f, T1d)); TF = LD(&(Rp[WS(rs, 9)]), ms, &(Rp[WS(rs, 1)])); TG = LD(&(Rm[WS(rs, 9)]), -ms, &(Rm[WS(rs, 1)])); TH = VCONJ(TG); TE = LDW(&(W[TWVL * 34])); TI = VZMULJ(TE, VADD(TF, TH)); T1S = LDW(&(W[TWVL * 36])); T1T = VZMULIJ(T1S, VSUB(TH, TF)); TL = LD(&(Rp[WS(rs, 5)]), ms, &(Rp[WS(rs, 1)])); TM = LD(&(Rm[WS(rs, 5)]), -ms, &(Rm[WS(rs, 1)])); TN = VCONJ(TM); TK = LDW(&(W[TWVL * 18])); TO = VZMULJ(TK, VADD(TL, TN)); T1X = LDW(&(W[TWVL * 20])); T1Y = VZMULIJ(T1X, VSUB(TN, TL)); TX = LD(&(Rp[WS(rs, 15)]), ms, &(Rp[WS(rs, 1)])); TY = LD(&(Rm[WS(rs, 15)]), -ms, &(Rm[WS(rs, 1)])); TZ = VCONJ(TY); TW = LDW(&(W[TWVL * 58])); T10 = VZMULJ(TW, VADD(TX, TZ)); T21 = LDW(&(W[TWVL * 60])); T22 = VZMULIJ(T21, VSUB(TZ, TX)); T12 = LD(&(Rp[WS(rs, 7)]), ms, &(Rp[WS(rs, 1)])); T13 = LD(&(Rm[WS(rs, 7)]), -ms, &(Rm[WS(rs, 1)])); T14 = VCONJ(T13); T11 = LDW(&(W[TWVL * 26])); T15 = VZMULJ(T11, VADD(T12, T14)); T23 = LDW(&(W[TWVL * 28])); T24 = VZMULIJ(T23, VSUB(T14, T12)); TQ = LD(&(Rp[WS(rs, 13)]), ms, &(Rp[WS(rs, 1)])); TR = LD(&(Rm[WS(rs, 13)]), -ms, &(Rm[WS(rs, 1)])); TS = VCONJ(TR); TP = LDW(&(W[TWVL * 50])); TT = VZMULJ(TP, VADD(TQ, TS)); T1V = LDW(&(W[TWVL * 52])); T1W = VZMULIJ(T1V, VSUB(TS, TQ)); } TJ = VSUB(TD, TI); TU = VSUB(TO, TT); TV = VFNMS(LDK(KP382683432), TU, VMUL(LDK(KP923879532), TJ)); T2p = VFMA(LDK(KP382683432), TJ, VMUL(LDK(KP923879532), TU)); T16 = VSUB(T10, T15); T1h = VSUB(T1b, T1g); T1i = VFMA(LDK(KP923879532), T16, VMUL(LDK(KP382683432), T1h)); T2o = VFNMS(LDK(KP923879532), T1h, VMUL(LDK(KP382683432), T16)); T3J = VADD(T1Y, T1W); T3K = VADD(T1R, T1T); T3L = VSUB(T3J, T3K); T4q = VADD(T3K, T3J); T3G = VADD(T22, T24); T3H = VADD(T29, T2b); T3I = VSUB(T3G, T3H); T4r = VADD(T3G, T3H); T3u = VADD(T10, T15); T3v = VADD(T1b, T1g); T3w = VSUB(T3u, T3v); T4k = VADD(T3u, T3v); T3r = VADD(TD, TI); T3s = VADD(TO, TT); T3t = VSUB(T3r, T3s); T4j = VADD(T3r, T3s); T25 = VSUB(T22, T24); T2c = VSUB(T29, T2b); T1U = VSUB(T1R, T1T); T1Z = VSUB(T1W, T1Y); T20 = VMUL(LDK(KP707106781), VADD(T1U, T1Z)); T27 = VMUL(LDK(KP707106781), VSUB(T1Z, T1U)); T26 = VADD(T20, T25); T2V = VADD(T27, T2c); T2d = VSUB(T27, T2c); T2U = VSUB(T25, T20); } { V T4m, T4w, T4t, T4x, T4i, T4l, T4p, T4s, T4u, T4z, T4v, T4y, T4E, T4L, T4H; V T4K, T4A, T4F, T4D, T4G, T4B, T4C, T4I, T4N, T4J, T4M, T3O, T4c, T4d, T3X; V T40, T46, T49, T41, T3y, T47, T3T, T45, T3N, T44, T3W, T48, T3x, T3S, T3F; V T3M, T3U, T3V, T3Y, T4e, T4f, T3Z, T42, T4a, T4b, T43; T4i = VADD(T4g, T4h); T4l = VADD(T4j, T4k); T4m = VADD(T4i, T4l); T4w = VSUB(T4i, T4l); T4p = VADD(T4n, T4o); T4s = VADD(T4q, T4r); T4t = VADD(T4p, T4s); T4x = VBYI(VSUB(T4s, T4p)); T4u = VCONJ(VMUL(LDK(KP500000000), VSUB(T4m, T4t))); ST(&(Rm[WS(rs, 15)]), T4u, -ms, &(Rm[WS(rs, 1)])); T4z = VMUL(LDK(KP500000000), VADD(T4w, T4x)); ST(&(Rp[WS(rs, 8)]), T4z, ms, &(Rp[0])); T4v = VMUL(LDK(KP500000000), VADD(T4m, T4t)); ST(&(Rp[0]), T4v, ms, &(Rp[0])); T4y = VCONJ(VMUL(LDK(KP500000000), VSUB(T4w, T4x))); ST(&(Rm[WS(rs, 7)]), T4y, -ms, &(Rm[WS(rs, 1)])); T4A = VMUL(LDK(KP500000000), VSUB(T4g, T4h)); T4F = VSUB(T4k, T4j); T4B = VSUB(T4n, T4o); T4C = VSUB(T4r, T4q); T4D = VMUL(LDK(KP353553390), VADD(T4B, T4C)); T4G = VMUL(LDK(KP707106781), VSUB(T4C, T4B)); T4E = VADD(T4A, T4D); T4L = VMUL(LDK(KP500000000), VBYI(VSUB(T4G, T4F))); T4H = VMUL(LDK(KP500000000), VBYI(VADD(T4F, T4G))); T4K = VSUB(T4A, T4D); T4I = VCONJ(VSUB(T4E, T4H)); ST(&(Rm[WS(rs, 3)]), T4I, -ms, &(Rm[WS(rs, 1)])); T4N = VADD(T4K, T4L); ST(&(Rp[WS(rs, 12)]), T4N, ms, &(Rp[0])); T4J = VADD(T4E, T4H); ST(&(Rp[WS(rs, 4)]), T4J, ms, &(Rp[0])); T4M = VCONJ(VSUB(T4K, T4L)); ST(&(Rm[WS(rs, 11)]), T4M, -ms, &(Rm[WS(rs, 1)])); T3x = VMUL(LDK(KP353553390), VADD(T3t, T3w)); T3y = VADD(T3q, T3x); T47 = VSUB(T3q, T3x); T3S = VMUL(LDK(KP707106781), VSUB(T3w, T3t)); T3T = VADD(T3R, T3S); T45 = VSUB(T3S, T3R); T3F = VFMA(LDK(KP923879532), T3B, VMUL(LDK(KP382683432), T3E)); T3M = VFNMS(LDK(KP382683432), T3L, VMUL(LDK(KP923879532), T3I)); T3N = VMUL(LDK(KP500000000), VADD(T3F, T3M)); T44 = VSUB(T3M, T3F); T3U = VFNMS(LDK(KP382683432), T3B, VMUL(LDK(KP923879532), T3E)); T3V = VFMA(LDK(KP923879532), T3L, VMUL(LDK(KP382683432), T3I)); T3W = VADD(T3U, T3V); T48 = VMUL(LDK(KP500000000), VSUB(T3V, T3U)); T3O = VADD(T3y, T3N); T4c = VMUL(LDK(KP500000000), VBYI(VADD(T45, T44))); T4d = VADD(T47, T48); T3X = VMUL(LDK(KP500000000), VBYI(VADD(T3T, T3W))); T40 = VSUB(T3y, T3N); T46 = VMUL(LDK(KP500000000), VBYI(VSUB(T44, T45))); T49 = VSUB(T47, T48); T41 = VMUL(LDK(KP500000000), VBYI(VSUB(T3W, T3T))); T3Y = VCONJ(VSUB(T3O, T3X)); ST(&(Rm[WS(rs, 1)]), T3Y, -ms, &(Rm[WS(rs, 1)])); T4e = VADD(T4c, T4d); ST(&(Rp[WS(rs, 6)]), T4e, ms, &(Rp[0])); T4f = VCONJ(VSUB(T4d, T4c)); ST(&(Rm[WS(rs, 5)]), T4f, -ms, &(Rm[WS(rs, 1)])); T3Z = VADD(T3O, T3X); ST(&(Rp[WS(rs, 2)]), T3Z, ms, &(Rp[0])); T42 = VCONJ(VSUB(T40, T41)); ST(&(Rm[WS(rs, 13)]), T42, -ms, &(Rm[WS(rs, 1)])); T4a = VADD(T46, T49); ST(&(Rp[WS(rs, 10)]), T4a, ms, &(Rp[0])); T4b = VCONJ(VSUB(T49, T46)); ST(&(Rm[WS(rs, 9)]), T4b, -ms, &(Rm[WS(rs, 1)])); T43 = VADD(T40, T41); ST(&(Rp[WS(rs, 14)]), T43, ms, &(Rp[0])); { V T2g, T2K, T2L, T2v, T2y, T2E, T2H, T2z, T1k, T2F, T2u, T2G, T2f, T2C, T2r; V T2D, Ty, T1j, T2s, T2t, T1P, T2e, T2n, T2q, T2w, T2M, T2N, T2x, T2A, T2I; V T2J, T2B; Ty = VADD(Ta, Tx); T1j = VMUL(LDK(KP500000000), VADD(TV, T1i)); T1k = VADD(Ty, T1j); T2F = VSUB(Ty, T1j); T2s = VFNMS(LDK(KP195090322), T1B, VMUL(LDK(KP980785280), T1O)); T2t = VFMA(LDK(KP195090322), T26, VMUL(LDK(KP980785280), T2d)); T2u = VADD(T2s, T2t); T2G = VMUL(LDK(KP500000000), VSUB(T2t, T2s)); T1P = VFMA(LDK(KP980785280), T1B, VMUL(LDK(KP195090322), T1O)); T2e = VFNMS(LDK(KP195090322), T2d, VMUL(LDK(KP980785280), T26)); T2f = VMUL(LDK(KP500000000), VADD(T1P, T2e)); T2C = VSUB(T2e, T1P); T2n = VSUB(T2h, T2m); T2q = VSUB(T2o, T2p); T2r = VADD(T2n, T2q); T2D = VSUB(T2q, T2n); T2g = VADD(T1k, T2f); T2K = VMUL(LDK(KP500000000), VBYI(VADD(T2D, T2C))); T2L = VADD(T2F, T2G); T2v = VMUL(LDK(KP500000000), VBYI(VADD(T2r, T2u))); T2y = VSUB(T1k, T2f); T2E = VMUL(LDK(KP500000000), VBYI(VSUB(T2C, T2D))); T2H = VSUB(T2F, T2G); T2z = VMUL(LDK(KP500000000), VBYI(VSUB(T2u, T2r))); T2w = VCONJ(VSUB(T2g, T2v)); ST(&(Rm[0]), T2w, -ms, &(Rm[0])); T2M = VADD(T2K, T2L); ST(&(Rp[WS(rs, 7)]), T2M, ms, &(Rp[WS(rs, 1)])); T2N = VCONJ(VSUB(T2L, T2K)); ST(&(Rm[WS(rs, 6)]), T2N, -ms, &(Rm[0])); T2x = VADD(T2g, T2v); ST(&(Rp[WS(rs, 1)]), T2x, ms, &(Rp[WS(rs, 1)])); T2A = VCONJ(VSUB(T2y, T2z)); ST(&(Rm[WS(rs, 14)]), T2A, -ms, &(Rm[0])); T2I = VADD(T2E, T2H); ST(&(Rp[WS(rs, 9)]), T2I, ms, &(Rp[WS(rs, 1)])); T2J = VCONJ(VSUB(T2H, T2E)); ST(&(Rm[WS(rs, 8)]), T2J, -ms, &(Rm[0])); T2B = VADD(T2y, T2z); ST(&(Rp[WS(rs, 15)]), T2B, ms, &(Rp[WS(rs, 1)])); } { V T2Y, T3k, T3l, T35, T38, T3e, T3h, T39, T2Q, T3f, T34, T3g, T2X, T3c, T31; V T3d, T2O, T2P, T32, T33, T2T, T2W, T2Z, T30, T36, T3m, T3n, T37, T3a, T3i; V T3j, T3b; T2O = VSUB(Ta, Tx); T2P = VMUL(LDK(KP500000000), VADD(T2p, T2o)); T2Q = VADD(T2O, T2P); T3f = VSUB(T2O, T2P); T32 = VFNMS(LDK(KP555570233), T2R, VMUL(LDK(KP831469612), T2S)); T33 = VFMA(LDK(KP555570233), T2U, VMUL(LDK(KP831469612), T2V)); T34 = VADD(T32, T33); T3g = VMUL(LDK(KP500000000), VSUB(T33, T32)); T2T = VFMA(LDK(KP831469612), T2R, VMUL(LDK(KP555570233), T2S)); T2W = VFNMS(LDK(KP555570233), T2V, VMUL(LDK(KP831469612), T2U)); T2X = VMUL(LDK(KP500000000), VADD(T2T, T2W)); T3c = VSUB(T2W, T2T); T2Z = VADD(T2m, T2h); T30 = VSUB(T1i, TV); T31 = VADD(T2Z, T30); T3d = VSUB(T30, T2Z); T2Y = VADD(T2Q, T2X); T3k = VMUL(LDK(KP500000000), VBYI(VADD(T3d, T3c))); T3l = VADD(T3f, T3g); T35 = VMUL(LDK(KP500000000), VBYI(VADD(T31, T34))); T38 = VSUB(T2Q, T2X); T3e = VMUL(LDK(KP500000000), VBYI(VSUB(T3c, T3d))); T3h = VSUB(T3f, T3g); T39 = VMUL(LDK(KP500000000), VBYI(VSUB(T34, T31))); T36 = VCONJ(VSUB(T2Y, T35)); ST(&(Rm[WS(rs, 2)]), T36, -ms, &(Rm[0])); T3m = VADD(T3k, T3l); ST(&(Rp[WS(rs, 5)]), T3m, ms, &(Rp[WS(rs, 1)])); T3n = VCONJ(VSUB(T3l, T3k)); ST(&(Rm[WS(rs, 4)]), T3n, -ms, &(Rm[0])); T37 = VADD(T2Y, T35); ST(&(Rp[WS(rs, 3)]), T37, ms, &(Rp[WS(rs, 1)])); T3a = VCONJ(VSUB(T38, T39)); ST(&(Rm[WS(rs, 12)]), T3a, -ms, &(Rm[0])); T3i = VADD(T3e, T3h); ST(&(Rp[WS(rs, 11)]), T3i, ms, &(Rp[WS(rs, 1)])); T3j = VCONJ(VSUB(T3h, T3e)); ST(&(Rm[WS(rs, 10)]), T3j, -ms, &(Rm[0])); T3b = VADD(T38, T39); ST(&(Rp[WS(rs, 13)]), T3b, ms, &(Rp[WS(rs, 1)])); } } } } VLEAVE(); } static const tw_instr twinstr[] = { VTW(1, 1), VTW(1, 2), VTW(1, 3), VTW(1, 4), VTW(1, 5), VTW(1, 6), VTW(1, 7), VTW(1, 8), VTW(1, 9), VTW(1, 10), VTW(1, 11), VTW(1, 12), VTW(1, 13), VTW(1, 14), VTW(1, 15), VTW(1, 16), VTW(1, 17), VTW(1, 18), VTW(1, 19), VTW(1, 20), VTW(1, 21), VTW(1, 22), VTW(1, 23), VTW(1, 24), VTW(1, 25), VTW(1, 26), VTW(1, 27), VTW(1, 28), VTW(1, 29), VTW(1, 30), VTW(1, 31), {TW_NEXT, VL, 0} }; static const hc2c_desc desc = { 32, XSIMD_STRING("hc2cfdftv_32"), twinstr, &GENUS, {233, 117, 16, 0} }; void XSIMD(codelet_hc2cfdftv_32) (planner *p) { X(khc2c_register) (p, hc2cfdftv_32, &desc, HC2C_VIA_DFT); } #endif /* HAVE_FMA */ fftw-3.3.3/rdft/simd/common/hc2cfdftv_20.c0000644000175000017500000005210512054411066015112 00000000000000/* * Copyright (c) 2003, 2007-11 Matteo Frigo * Copyright (c) 2003, 2007-11 Massachusetts Institute of Technology * * This program is free software; you can redistribute it and/or modify * it under the terms of the GNU General Public License as published by * the Free Software Foundation; either version 2 of the License, or * (at your option) any later version. * * This program is distributed in the hope that it will be useful, * but WITHOUT ANY WARRANTY; without even the implied warranty of * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the * GNU General Public License for more details. * * You should have received a copy of the GNU General Public License * along with this program; if not, write to the Free Software * Foundation, Inc., 51 Franklin Street, Fifth Floor, Boston, MA 02110-1301 USA * */ /* This file was automatically generated --- DO NOT EDIT */ /* Generated on Sun Nov 25 07:42:29 EST 2012 */ #include "codelet-rdft.h" #ifdef HAVE_FMA /* Generated by: ../../../genfft/gen_hc2cdft_c.native -fma -reorder-insns -schedule-for-pipeline -simd -compact -variables 4 -pipeline-latency 8 -trivial-stores -variables 32 -no-generate-bytw -n 20 -dit -name hc2cfdftv_20 -include hc2cfv.h */ /* * This function contains 143 FP additions, 128 FP multiplications, * (or, 77 additions, 62 multiplications, 66 fused multiply/add), * 130 stack variables, 5 constants, and 40 memory accesses */ #include "hc2cfv.h" static void hc2cfdftv_20(R *Rp, R *Ip, R *Rm, R *Im, const R *W, stride rs, INT mb, INT me, INT ms) { DVK(KP559016994, +0.559016994374947424102293417182819058860154590); DVK(KP500000000, +0.500000000000000000000000000000000000000000000); DVK(KP951056516, +0.951056516295153572116439333379382143405698634); DVK(KP250000000, +0.250000000000000000000000000000000000000000000); DVK(KP618033988, +0.618033988749894848204586834365638117720309180); { INT m; for (m = mb, W = W + ((mb - 1) * ((TWVL / VL) * 38)); m < me; m = m + VL, Rp = Rp + (VL * ms), Ip = Ip + (VL * ms), Rm = Rm - (VL * ms), Im = Im - (VL * ms), W = W + (TWVL * 38), MAKE_VOLATILE_STRIDE(80, rs)) { V T2g, T2f, T2w, T2k, T2A, T2u, T2e, T2o, T1O, T2b, T2i, T1R, T1X, T1k, TN; V T1w, T1G, T1t, Ti, T2c, T12, T1x, T2j, T1U, T1y, T1d, T24, T2v, T2h, T2x; V T2B, T2p, T2l, T2z, T2y, T2D, T2C, T2r, T2q, T2n, T2m; { V T3, T7, TC, T1Y, Tc, Tg, Tn, T1P, T1Z, Tw, T1S, TS, TY, TZ, T1Q; V TL, T17, T21, TW, T19, TX, T1a, T8, T20, Th, Tx, T1u, T1v, TM, T10; V T1b, T22, T11, T1T, T1c, T23; { V Ta, Tb, Tz, Te, TB, Tf, Tl, T9, Td, Tk, T1, T2, Ty, T5, T6; V TA, T4, Tj, Tt, Tu, Ts, TQ, Tr, TP, Tp, Tq, Tm, To, TO, TG; V T14, TK, T16, TE, TF, Tv, TD, T13, TR, TI, TJ, TH, T15, TU, TV; V TT, T18; T1 = LD(&(Rp[0]), ms, &(Rp[0])); T2 = LD(&(Rm[0]), -ms, &(Rm[0])); Ty = LDW(&(W[0])); T5 = LD(&(Rp[WS(rs, 5)]), ms, &(Rp[WS(rs, 1)])); T6 = LD(&(Rm[WS(rs, 5)]), -ms, &(Rm[WS(rs, 1)])); TA = LDW(&(W[TWVL * 20])); T4 = LDW(&(W[TWVL * 18])); Ta = LD(&(Rp[WS(rs, 2)]), ms, &(Rp[0])); Tb = LD(&(Rm[WS(rs, 2)]), -ms, &(Rm[0])); T3 = VFMACONJ(T2, T1); Tz = VZMULIJ(Ty, VFNMSCONJ(T2, T1)); Tj = LDW(&(W[TWVL * 6])); Te = LD(&(Rp[WS(rs, 7)]), ms, &(Rp[WS(rs, 1)])); TB = VZMULIJ(TA, VFNMSCONJ(T6, T5)); T7 = VZMULJ(T4, VFMACONJ(T6, T5)); Tf = LD(&(Rm[WS(rs, 7)]), -ms, &(Rm[WS(rs, 1)])); Tl = LDW(&(W[TWVL * 26])); T9 = LDW(&(W[TWVL * 8])); Td = LDW(&(W[TWVL * 28])); Tk = VZMULJ(Tj, VFMACONJ(Tb, Ta)); Tp = LD(&(Rp[WS(rs, 4)]), ms, &(Rp[0])); TC = VADD(Tz, TB); T1Y = VSUB(TB, Tz); Tq = LD(&(Rm[WS(rs, 4)]), -ms, &(Rm[0])); Tm = VZMULJ(Tl, VFMACONJ(Tf, Te)); Tc = VZMULIJ(T9, VFNMSCONJ(Tb, Ta)); Tg = VZMULIJ(Td, VFNMSCONJ(Tf, Te)); To = LDW(&(W[TWVL * 16])); TO = LDW(&(W[TWVL * 14])); Tt = LD(&(Rp[WS(rs, 9)]), ms, &(Rp[WS(rs, 1)])); Tu = LD(&(Rm[WS(rs, 9)]), -ms, &(Rm[WS(rs, 1)])); Ts = LDW(&(W[TWVL * 36])); Tn = VADD(Tk, Tm); T1P = VSUB(Tk, Tm); TQ = LDW(&(W[TWVL * 34])); Tr = VZMULIJ(To, VFNMSCONJ(Tq, Tp)); TP = VZMULJ(TO, VFMACONJ(Tq, Tp)); TE = LD(&(Rp[WS(rs, 8)]), ms, &(Rp[0])); TF = LD(&(Rm[WS(rs, 8)]), -ms, &(Rm[0])); Tv = VZMULIJ(Ts, VFNMSCONJ(Tu, Tt)); TD = LDW(&(W[TWVL * 30])); T13 = LDW(&(W[TWVL * 32])); TR = VZMULJ(TQ, VFMACONJ(Tu, Tt)); TI = LD(&(Rp[WS(rs, 3)]), ms, &(Rp[WS(rs, 1)])); TJ = LD(&(Rm[WS(rs, 3)]), -ms, &(Rm[WS(rs, 1)])); TH = LDW(&(W[TWVL * 10])); T15 = LDW(&(W[TWVL * 12])); T1Z = VSUB(Tv, Tr); Tw = VADD(Tr, Tv); TG = VZMULJ(TD, VFMACONJ(TF, TE)); T14 = VZMULIJ(T13, VFNMSCONJ(TF, TE)); T1S = VSUB(TP, TR); TS = VADD(TP, TR); TK = VZMULJ(TH, VFMACONJ(TJ, TI)); T16 = VZMULIJ(T15, VFNMSCONJ(TJ, TI)); TU = LD(&(Rp[WS(rs, 6)]), ms, &(Rp[0])); TV = LD(&(Rm[WS(rs, 6)]), -ms, &(Rm[0])); TT = LDW(&(W[TWVL * 24])); T18 = LDW(&(W[TWVL * 22])); TY = LD(&(Rp[WS(rs, 1)]), ms, &(Rp[WS(rs, 1)])); TZ = LD(&(Rm[WS(rs, 1)]), -ms, &(Rm[WS(rs, 1)])); T1Q = VSUB(TK, TG); TL = VADD(TG, TK); T17 = VADD(T14, T16); T21 = VSUB(T16, T14); TW = VZMULIJ(TT, VFNMSCONJ(TV, TU)); T19 = VZMULJ(T18, VFMACONJ(TV, TU)); TX = LDW(&(W[TWVL * 4])); T1a = LDW(&(W[TWVL * 2])); } T1O = VSUB(T3, T7); T8 = VADD(T3, T7); T20 = VADD(T1Y, T1Z); T2b = VSUB(T1Y, T1Z); T2i = VADD(T1P, T1Q); T1R = VSUB(T1P, T1Q); Th = VADD(Tc, Tg); T1X = VSUB(Tg, Tc); Tx = VSUB(Tn, Tw); T1u = VADD(Tn, Tw); T1v = VADD(TC, TL); TM = VSUB(TC, TL); T10 = VZMULIJ(TX, VFNMSCONJ(TZ, TY)); T1b = VZMULJ(T1a, VFMACONJ(TZ, TY)); T1k = VADD(Tx, TM); TN = VSUB(Tx, TM); T22 = VSUB(T10, TW); T11 = VADD(TW, T10); T1T = VSUB(T1b, T19); T1c = VADD(T19, T1b); T1w = VADD(T1u, T1v); T1G = VSUB(T1u, T1v); T1t = VADD(T8, Th); Ti = VSUB(T8, Th); T23 = VADD(T21, T22); T2c = VSUB(T21, T22); T12 = VSUB(TS, T11); T1x = VADD(TS, T11); T2j = VADD(T1S, T1T); T1U = VSUB(T1S, T1T); T1y = VADD(T17, T1c); T1d = VSUB(T17, T1c); T2g = VSUB(T23, T20); T24 = VADD(T20, T23); } { V T2d, T2t, T29, T25, T1m, T1q, T1i, T1H, T1L, T1D, T1A, T28, T1W, T1h, T1g; V T1e, T1l, T1z, T1F, T1V, T1f, T1C, T1B, T26, T27, T2a, T2s, T1j, T1p, T1K; V T1E, T1n, T1o, T1s, T1r, T1I, T1J, T1N, T1M; T2d = VFMA(LDK(KP618033988), T2c, T2b); T2t = VFNMS(LDK(KP618033988), T2b, T2c); T1e = VSUB(T12, T1d); T1l = VADD(T12, T1d); T1z = VADD(T1x, T1y); T1F = VSUB(T1x, T1y); T1V = VADD(T1R, T1U); T29 = VSUB(T1R, T1U); T2f = VFNMS(LDK(KP250000000), T24, T1X); T25 = VADD(T1X, T24); T1m = VMUL(LDK(KP951056516), VFMA(LDK(KP618033988), T1l, T1k)); T1q = VMUL(LDK(KP951056516), VFNMS(LDK(KP618033988), T1k, T1l)); T1i = VSUB(TN, T1e); T1f = VADD(TN, T1e); T1H = VMUL(LDK(KP951056516), VFNMS(LDK(KP618033988), T1G, T1F)); T1L = VMUL(LDK(KP951056516), VFMA(LDK(KP618033988), T1F, T1G)); T1D = VSUB(T1w, T1z); T1A = VADD(T1w, T1z); T28 = VFNMS(LDK(KP250000000), T1V, T1O); T1W = VADD(T1O, T1V); T1h = VFNMS(LDK(KP250000000), T1f, Ti); T1g = VMUL(LDK(KP500000000), VADD(Ti, T1f)); T2w = VFNMS(LDK(KP618033988), T2i, T2j); T2k = VFMA(LDK(KP618033988), T2j, T2i); T1C = VFNMS(LDK(KP250000000), T1A, T1t); T1B = VCONJ(VMUL(LDK(KP500000000), VADD(T1t, T1A))); T26 = VMUL(LDK(KP500000000), VFNMSI(T25, T1W)); T27 = VCONJ(VMUL(LDK(KP500000000), VFMAI(T25, T1W))); T2a = VFMA(LDK(KP559016994), T29, T28); T2s = VFNMS(LDK(KP559016994), T29, T28); ST(&(Rp[0]), T1g, ms, &(Rp[0])); T1j = VFMA(LDK(KP559016994), T1i, T1h); T1p = VFNMS(LDK(KP559016994), T1i, T1h); ST(&(Rm[WS(rs, 9)]), T1B, -ms, &(Rm[WS(rs, 1)])); T1K = VFMA(LDK(KP559016994), T1D, T1C); T1E = VFNMS(LDK(KP559016994), T1D, T1C); ST(&(Rm[WS(rs, 4)]), T27, -ms, &(Rm[0])); ST(&(Rp[WS(rs, 5)]), T26, ms, &(Rp[WS(rs, 1)])); T2A = VFMA(LDK(KP951056516), T2t, T2s); T2u = VFNMS(LDK(KP951056516), T2t, T2s); T2e = VFNMS(LDK(KP951056516), T2d, T2a); T2o = VFMA(LDK(KP951056516), T2d, T2a); T1n = VCONJ(VMUL(LDK(KP500000000), VFNMSI(T1m, T1j))); T1o = VMUL(LDK(KP500000000), VFMAI(T1m, T1j)); T1s = VCONJ(VMUL(LDK(KP500000000), VFMAI(T1q, T1p))); T1r = VMUL(LDK(KP500000000), VFNMSI(T1q, T1p)); T1I = VCONJ(VMUL(LDK(KP500000000), VFNMSI(T1H, T1E))); T1J = VMUL(LDK(KP500000000), VFMAI(T1H, T1E)); T1N = VCONJ(VMUL(LDK(KP500000000), VFMAI(T1L, T1K))); T1M = VMUL(LDK(KP500000000), VFNMSI(T1L, T1K)); ST(&(Rp[WS(rs, 4)]), T1o, ms, &(Rp[0])); ST(&(Rm[WS(rs, 3)]), T1n, -ms, &(Rm[WS(rs, 1)])); ST(&(Rp[WS(rs, 8)]), T1r, ms, &(Rp[0])); ST(&(Rm[WS(rs, 7)]), T1s, -ms, &(Rm[WS(rs, 1)])); ST(&(Rp[WS(rs, 2)]), T1J, ms, &(Rp[0])); ST(&(Rm[WS(rs, 1)]), T1I, -ms, &(Rm[WS(rs, 1)])); ST(&(Rp[WS(rs, 6)]), T1M, ms, &(Rp[0])); ST(&(Rm[WS(rs, 5)]), T1N, -ms, &(Rm[WS(rs, 1)])); } T2v = VFMA(LDK(KP559016994), T2g, T2f); T2h = VFNMS(LDK(KP559016994), T2g, T2f); T2x = VFNMS(LDK(KP951056516), T2w, T2v); T2B = VFMA(LDK(KP951056516), T2w, T2v); T2p = VFMA(LDK(KP951056516), T2k, T2h); T2l = VFNMS(LDK(KP951056516), T2k, T2h); T2z = VMUL(LDK(KP500000000), VFMAI(T2x, T2u)); T2y = VCONJ(VMUL(LDK(KP500000000), VFNMSI(T2x, T2u))); T2D = VMUL(LDK(KP500000000), VFMAI(T2B, T2A)); T2C = VCONJ(VMUL(LDK(KP500000000), VFNMSI(T2B, T2A))); T2r = VCONJ(VMUL(LDK(KP500000000), VFMAI(T2p, T2o))); T2q = VMUL(LDK(KP500000000), VFNMSI(T2p, T2o)); T2n = VCONJ(VMUL(LDK(KP500000000), VFMAI(T2l, T2e))); T2m = VMUL(LDK(KP500000000), VFNMSI(T2l, T2e)); ST(&(Rp[WS(rs, 3)]), T2z, ms, &(Rp[WS(rs, 1)])); ST(&(Rm[WS(rs, 2)]), T2y, -ms, &(Rm[0])); ST(&(Rp[WS(rs, 7)]), T2D, ms, &(Rp[WS(rs, 1)])); ST(&(Rm[WS(rs, 6)]), T2C, -ms, &(Rm[0])); ST(&(Rm[0]), T2r, -ms, &(Rm[0])); ST(&(Rp[WS(rs, 1)]), T2q, ms, &(Rp[WS(rs, 1)])); ST(&(Rm[WS(rs, 8)]), T2n, -ms, &(Rm[0])); ST(&(Rp[WS(rs, 9)]), T2m, ms, &(Rp[WS(rs, 1)])); } } VLEAVE(); } static const tw_instr twinstr[] = { VTW(1, 1), VTW(1, 2), VTW(1, 3), VTW(1, 4), VTW(1, 5), VTW(1, 6), VTW(1, 7), VTW(1, 8), VTW(1, 9), VTW(1, 10), VTW(1, 11), VTW(1, 12), VTW(1, 13), VTW(1, 14), VTW(1, 15), VTW(1, 16), VTW(1, 17), VTW(1, 18), VTW(1, 19), {TW_NEXT, VL, 0} }; static const hc2c_desc desc = { 20, XSIMD_STRING("hc2cfdftv_20"), twinstr, &GENUS, {77, 62, 66, 0} }; void XSIMD(codelet_hc2cfdftv_20) (planner *p) { X(khc2c_register) (p, hc2cfdftv_20, &desc, HC2C_VIA_DFT); } #else /* HAVE_FMA */ /* Generated by: ../../../genfft/gen_hc2cdft_c.native -simd -compact -variables 4 -pipeline-latency 8 -trivial-stores -variables 32 -no-generate-bytw -n 20 -dit -name hc2cfdftv_20 -include hc2cfv.h */ /* * This function contains 143 FP additions, 77 FP multiplications, * (or, 131 additions, 65 multiplications, 12 fused multiply/add), * 141 stack variables, 9 constants, and 40 memory accesses */ #include "hc2cfv.h" static void hc2cfdftv_20(R *Rp, R *Ip, R *Rm, R *Im, const R *W, stride rs, INT mb, INT me, INT ms) { DVK(KP293892626, +0.293892626146236564584352977319536384298826219); DVK(KP475528258, +0.475528258147576786058219666689691071702849317); DVK(KP559016994, +0.559016994374947424102293417182819058860154590); DVK(KP250000000, +0.250000000000000000000000000000000000000000000); DVK(KP125000000, +0.125000000000000000000000000000000000000000000); DVK(KP279508497, +0.279508497187473712051146708591409529430077295); DVK(KP587785252, +0.587785252292473129168705954639072768597652438); DVK(KP951056516, +0.951056516295153572116439333379382143405698634); DVK(KP500000000, +0.500000000000000000000000000000000000000000000); { INT m; for (m = mb, W = W + ((mb - 1) * ((TWVL / VL) * 38)); m < me; m = m + VL, Rp = Rp + (VL * ms), Ip = Ip + (VL * ms), Rm = Rm - (VL * ms), Im = Im - (VL * ms), W = W + (TWVL * 38), MAKE_VOLATILE_STRIDE(80, rs)) { V TW, T1x, T2i, T2A, T1r, T1s, T1a, T1y, T1l, Tn, TK, TL, T1p, T1o, T27; V T2t, T2a, T2u, T2e, T2C, T20, T2w, T23, T2x, T2d, T2B, T1W, T1X, T1U, T1V; V T2z, T2K, T2G, T2N, T2J, T2v, T2y, T2F, T2D, T2E, T2M, T2H, T2I, T2L; { V T1u, T5, Tg, T1c, TV, T13, Ta, T1w, TQ, T11, TI, T1j, Tx, T18, Tl; V T1e, TD, T1h, Ts, T16, T2g, T2h, T14, T19, T1f, T1k, Tb, Tm, Ty, TJ; V T25, T26, T28, T29, T1Y, T1Z, T21, T22; { V T4, T3, T2, T1, Tf, Te, Td, Tc, T1b, TU, TT, TS, TR, T12, T9; V T8, T7, T6, T1v, TP, TO, TN, TM, T10, TH, TG, TF, TE, T1i, Tw; V Tv, Tu, Tt, T17, Tk, Tj, Ti, Th, T1d, TC, TB, TA, Tz, T1g, Tr; V Tq, Tp, To, T15; T4 = LD(&(Rp[0]), ms, &(Rp[0])); T2 = LD(&(Rm[0]), -ms, &(Rm[0])); T3 = VCONJ(T2); T1u = VADD(T4, T3); T1 = LDW(&(W[0])); T5 = VZMULIJ(T1, VSUB(T3, T4)); Tf = LD(&(Rp[WS(rs, 4)]), ms, &(Rp[0])); Td = LD(&(Rm[WS(rs, 4)]), -ms, &(Rm[0])); Te = VCONJ(Td); Tc = LDW(&(W[TWVL * 16])); Tg = VZMULIJ(Tc, VSUB(Te, Tf)); T1b = LDW(&(W[TWVL * 14])); T1c = VZMULJ(T1b, VADD(Te, Tf)); TU = LD(&(Rp[WS(rs, 7)]), ms, &(Rp[WS(rs, 1)])); TS = LD(&(Rm[WS(rs, 7)]), -ms, &(Rm[WS(rs, 1)])); TT = VCONJ(TS); TR = LDW(&(W[TWVL * 28])); TV = VZMULIJ(TR, VSUB(TT, TU)); T12 = LDW(&(W[TWVL * 26])); T13 = VZMULJ(T12, VADD(TT, TU)); T9 = LD(&(Rp[WS(rs, 5)]), ms, &(Rp[WS(rs, 1)])); T7 = LD(&(Rm[WS(rs, 5)]), -ms, &(Rm[WS(rs, 1)])); T8 = VCONJ(T7); T6 = LDW(&(W[TWVL * 20])); Ta = VZMULIJ(T6, VSUB(T8, T9)); T1v = LDW(&(W[TWVL * 18])); T1w = VZMULJ(T1v, VADD(T9, T8)); TP = LD(&(Rp[WS(rs, 2)]), ms, &(Rp[0])); TN = LD(&(Rm[WS(rs, 2)]), -ms, &(Rm[0])); TO = VCONJ(TN); TM = LDW(&(W[TWVL * 8])); TQ = VZMULIJ(TM, VSUB(TO, TP)); T10 = LDW(&(W[TWVL * 6])); T11 = VZMULJ(T10, VADD(TO, TP)); TH = LD(&(Rp[WS(rs, 1)]), ms, &(Rp[WS(rs, 1)])); TF = LD(&(Rm[WS(rs, 1)]), -ms, &(Rm[WS(rs, 1)])); TG = VCONJ(TF); TE = LDW(&(W[TWVL * 4])); TI = VZMULIJ(TE, VSUB(TG, TH)); T1i = LDW(&(W[TWVL * 2])); T1j = VZMULJ(T1i, VADD(TG, TH)); Tw = LD(&(Rp[WS(rs, 3)]), ms, &(Rp[WS(rs, 1)])); Tu = LD(&(Rm[WS(rs, 3)]), -ms, &(Rm[WS(rs, 1)])); Tv = VCONJ(Tu); Tt = LDW(&(W[TWVL * 12])); Tx = VZMULIJ(Tt, VSUB(Tv, Tw)); T17 = LDW(&(W[TWVL * 10])); T18 = VZMULJ(T17, VADD(Tw, Tv)); Tk = LD(&(Rp[WS(rs, 9)]), ms, &(Rp[WS(rs, 1)])); Ti = LD(&(Rm[WS(rs, 9)]), -ms, &(Rm[WS(rs, 1)])); Tj = VCONJ(Ti); Th = LDW(&(W[TWVL * 36])); Tl = VZMULIJ(Th, VSUB(Tj, Tk)); T1d = LDW(&(W[TWVL * 34])); T1e = VZMULJ(T1d, VADD(Tj, Tk)); TC = LD(&(Rp[WS(rs, 6)]), ms, &(Rp[0])); TA = LD(&(Rm[WS(rs, 6)]), -ms, &(Rm[0])); TB = VCONJ(TA); Tz = LDW(&(W[TWVL * 24])); TD = VZMULIJ(Tz, VSUB(TB, TC)); T1g = LDW(&(W[TWVL * 22])); T1h = VZMULJ(T1g, VADD(TB, TC)); Tr = LD(&(Rp[WS(rs, 8)]), ms, &(Rp[0])); Tp = LD(&(Rm[WS(rs, 8)]), -ms, &(Rm[0])); Tq = VCONJ(Tp); To = LDW(&(W[TWVL * 32])); Ts = VZMULIJ(To, VSUB(Tq, Tr)); T15 = LDW(&(W[TWVL * 30])); T16 = VZMULJ(T15, VADD(Tr, Tq)); } TW = VSUB(TQ, TV); T1x = VSUB(T1u, T1w); T2g = VADD(T1u, T1w); T2h = VADD(TQ, TV); T2i = VADD(T2g, T2h); T2A = VSUB(T2g, T2h); T14 = VSUB(T11, T13); T19 = VSUB(T16, T18); T1r = VADD(T14, T19); T1f = VSUB(T1c, T1e); T1k = VSUB(T1h, T1j); T1s = VADD(T1f, T1k); T1a = VSUB(T14, T19); T1y = VADD(T1r, T1s); T1l = VSUB(T1f, T1k); Tb = VSUB(T5, Ta); Tm = VSUB(Tg, Tl); Tn = VADD(Tb, Tm); Ty = VSUB(Ts, Tx); TJ = VSUB(TD, TI); TK = VADD(Ty, TJ); TL = VADD(Tn, TK); T1p = VSUB(Ty, TJ); T1o = VSUB(Tb, Tm); T25 = VADD(T1c, T1e); T26 = VADD(TD, TI); T27 = VADD(T25, T26); T2t = VSUB(T25, T26); T28 = VADD(Ts, Tx); T29 = VADD(T1h, T1j); T2a = VADD(T28, T29); T2u = VSUB(T29, T28); T2e = VADD(T27, T2a); T2C = VADD(T2t, T2u); T1Y = VADD(T11, T13); T1Z = VADD(Tg, Tl); T20 = VADD(T1Y, T1Z); T2w = VSUB(T1Y, T1Z); T21 = VADD(T5, Ta); T22 = VADD(T16, T18); T23 = VADD(T21, T22); T2x = VSUB(T22, T21); T2d = VADD(T20, T23); T2B = VADD(T2w, T2x); } T1U = VADD(T1x, T1y); T1V = VBYI(VADD(TW, TL)); T1W = VMUL(LDK(KP500000000), VSUB(T1U, T1V)); T1X = VCONJ(VMUL(LDK(KP500000000), VADD(T1V, T1U))); ST(&(Rp[WS(rs, 5)]), T1W, ms, &(Rp[WS(rs, 1)])); ST(&(Rm[WS(rs, 4)]), T1X, -ms, &(Rm[0])); T2v = VSUB(T2t, T2u); T2y = VSUB(T2w, T2x); T2z = VMUL(LDK(KP500000000), VBYI(VFNMS(LDK(KP587785252), T2y, VMUL(LDK(KP951056516), T2v)))); T2K = VMUL(LDK(KP500000000), VBYI(VFMA(LDK(KP951056516), T2y, VMUL(LDK(KP587785252), T2v)))); T2F = VMUL(LDK(KP279508497), VSUB(T2B, T2C)); T2D = VADD(T2B, T2C); T2E = VFNMS(LDK(KP125000000), T2D, VMUL(LDK(KP500000000), T2A)); T2G = VSUB(T2E, T2F); T2N = VCONJ(VMUL(LDK(KP500000000), VADD(T2A, T2D))); T2J = VADD(T2F, T2E); ST(&(Rm[WS(rs, 9)]), T2N, -ms, &(Rm[WS(rs, 1)])); T2M = VCONJ(VADD(T2K, T2J)); ST(&(Rm[WS(rs, 5)]), T2M, -ms, &(Rm[WS(rs, 1)])); T2H = VADD(T2z, T2G); ST(&(Rp[WS(rs, 2)]), T2H, ms, &(Rp[0])); T2I = VCONJ(VSUB(T2G, T2z)); ST(&(Rm[WS(rs, 1)]), T2I, -ms, &(Rm[WS(rs, 1)])); T2L = VSUB(T2J, T2K); ST(&(Rp[WS(rs, 6)]), T2L, ms, &(Rp[0])); { V T2c, T2p, T2l, T2s, T2o, T24, T2b, T2f, T2j, T2k, T2r, T2m, T2n, T2q, T1n; V T1Q, T1E, T1K, T1B, T1R, T1F, T1N, T1m, T1J, TZ, T1I, TX, TY, T1q, T1M; V T1A, T1L, T1t, T1z, T1C, T1S, T1T, T1D, T1G, T1O, T1P, T1H; T24 = VSUB(T20, T23); T2b = VSUB(T27, T2a); T2c = VMUL(LDK(KP500000000), VBYI(VFMA(LDK(KP951056516), T24, VMUL(LDK(KP587785252), T2b)))); T2p = VMUL(LDK(KP500000000), VBYI(VFNMS(LDK(KP587785252), T24, VMUL(LDK(KP951056516), T2b)))); T2f = VMUL(LDK(KP279508497), VSUB(T2d, T2e)); T2j = VADD(T2d, T2e); T2k = VFNMS(LDK(KP125000000), T2j, VMUL(LDK(KP500000000), T2i)); T2l = VADD(T2f, T2k); T2s = VMUL(LDK(KP500000000), VADD(T2i, T2j)); T2o = VSUB(T2k, T2f); ST(&(Rp[0]), T2s, ms, &(Rp[0])); T2r = VCONJ(VADD(T2p, T2o)); ST(&(Rm[WS(rs, 7)]), T2r, -ms, &(Rm[WS(rs, 1)])); T2m = VADD(T2c, T2l); ST(&(Rp[WS(rs, 4)]), T2m, ms, &(Rp[0])); T2n = VCONJ(VSUB(T2l, T2c)); ST(&(Rm[WS(rs, 3)]), T2n, -ms, &(Rm[WS(rs, 1)])); T2q = VSUB(T2o, T2p); ST(&(Rp[WS(rs, 8)]), T2q, ms, &(Rp[0])); T1m = VFMA(LDK(KP951056516), T1a, VMUL(LDK(KP587785252), T1l)); T1J = VFNMS(LDK(KP587785252), T1a, VMUL(LDK(KP951056516), T1l)); TX = VFMS(LDK(KP250000000), TL, TW); TY = VMUL(LDK(KP559016994), VSUB(TK, Tn)); TZ = VADD(TX, TY); T1I = VSUB(TY, TX); T1n = VMUL(LDK(KP500000000), VBYI(VSUB(TZ, T1m))); T1Q = VMUL(LDK(KP500000000), VBYI(VADD(T1I, T1J))); T1E = VMUL(LDK(KP500000000), VBYI(VADD(TZ, T1m))); T1K = VMUL(LDK(KP500000000), VBYI(VSUB(T1I, T1J))); T1q = VFMA(LDK(KP475528258), T1o, VMUL(LDK(KP293892626), T1p)); T1M = VFNMS(LDK(KP293892626), T1o, VMUL(LDK(KP475528258), T1p)); T1t = VMUL(LDK(KP279508497), VSUB(T1r, T1s)); T1z = VFNMS(LDK(KP125000000), T1y, VMUL(LDK(KP500000000), T1x)); T1A = VADD(T1t, T1z); T1L = VSUB(T1z, T1t); T1B = VADD(T1q, T1A); T1R = VADD(T1M, T1L); T1F = VSUB(T1A, T1q); T1N = VSUB(T1L, T1M); T1C = VADD(T1n, T1B); ST(&(Rp[WS(rs, 1)]), T1C, ms, &(Rp[WS(rs, 1)])); T1S = VADD(T1Q, T1R); ST(&(Rp[WS(rs, 7)]), T1S, ms, &(Rp[WS(rs, 1)])); T1T = VCONJ(VSUB(T1R, T1Q)); ST(&(Rm[WS(rs, 6)]), T1T, -ms, &(Rm[0])); T1D = VCONJ(VSUB(T1B, T1n)); ST(&(Rm[0]), T1D, -ms, &(Rm[0])); T1G = VADD(T1E, T1F); ST(&(Rp[WS(rs, 9)]), T1G, ms, &(Rp[WS(rs, 1)])); T1O = VADD(T1K, T1N); ST(&(Rp[WS(rs, 3)]), T1O, ms, &(Rp[WS(rs, 1)])); T1P = VCONJ(VSUB(T1N, T1K)); ST(&(Rm[WS(rs, 2)]), T1P, -ms, &(Rm[0])); T1H = VCONJ(VSUB(T1F, T1E)); ST(&(Rm[WS(rs, 8)]), T1H, -ms, &(Rm[0])); } } } VLEAVE(); } static const tw_instr twinstr[] = { VTW(1, 1), VTW(1, 2), VTW(1, 3), VTW(1, 4), VTW(1, 5), VTW(1, 6), VTW(1, 7), VTW(1, 8), VTW(1, 9), VTW(1, 10), VTW(1, 11), VTW(1, 12), VTW(1, 13), VTW(1, 14), VTW(1, 15), VTW(1, 16), VTW(1, 17), VTW(1, 18), VTW(1, 19), {TW_NEXT, VL, 0} }; static const hc2c_desc desc = { 20, XSIMD_STRING("hc2cfdftv_20"), twinstr, &GENUS, {131, 65, 12, 0} }; void XSIMD(codelet_hc2cfdftv_20) (planner *p) { X(khc2c_register) (p, hc2cfdftv_20, &desc, HC2C_VIA_DFT); } #endif /* HAVE_FMA */ fftw-3.3.3/rdft/simd/common/hc2cbdftv_2.c0000644000175000017500000000760412054411065015031 00000000000000/* * Copyright (c) 2003, 2007-11 Matteo Frigo * Copyright (c) 2003, 2007-11 Massachusetts Institute of Technology * * This program is free software; you can redistribute it and/or modify * it under the terms of the GNU General Public License as published by * the Free Software Foundation; either version 2 of the License, or * (at your option) any later version. * * This program is distributed in the hope that it will be useful, * but WITHOUT ANY WARRANTY; without even the implied warranty of * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the * GNU General Public License for more details. * * You should have received a copy of the GNU General Public License * along with this program; if not, write to the Free Software * Foundation, Inc., 51 Franklin Street, Fifth Floor, Boston, MA 02110-1301 USA * */ /* This file was automatically generated --- DO NOT EDIT */ /* Generated on Sun Nov 25 07:42:29 EST 2012 */ #include "codelet-rdft.h" #ifdef HAVE_FMA /* Generated by: ../../../genfft/gen_hc2cdft_c.native -fma -reorder-insns -schedule-for-pipeline -simd -compact -variables 4 -pipeline-latency 8 -trivial-stores -variables 32 -no-generate-bytw -n 2 -dif -sign 1 -name hc2cbdftv_2 -include hc2cbv.h */ /* * This function contains 5 FP additions, 4 FP multiplications, * (or, 3 additions, 2 multiplications, 2 fused multiply/add), * 8 stack variables, 0 constants, and 4 memory accesses */ #include "hc2cbv.h" static void hc2cbdftv_2(R *Rp, R *Ip, R *Rm, R *Im, const R *W, stride rs, INT mb, INT me, INT ms) { { INT m; for (m = mb, W = W + ((mb - 1) * ((TWVL / VL) * 2)); m < me; m = m + VL, Rp = Rp + (VL * ms), Ip = Ip + (VL * ms), Rm = Rm - (VL * ms), Im = Im - (VL * ms), W = W + (TWVL * 2), MAKE_VOLATILE_STRIDE(8, rs)) { V T2, T3, T1, T5, T4, T7, T6; T2 = LD(&(Rp[0]), ms, &(Rp[0])); T3 = LD(&(Rm[0]), -ms, &(Rm[0])); T1 = LDW(&(W[0])); T5 = VFMACONJ(T3, T2); T4 = VZMULI(T1, VFNMSCONJ(T3, T2)); T7 = VCONJ(VSUB(T5, T4)); T6 = VADD(T4, T5); ST(&(Rm[0]), T7, -ms, &(Rm[0])); ST(&(Rp[0]), T6, ms, &(Rp[0])); } } VLEAVE(); } static const tw_instr twinstr[] = { VTW(1, 1), {TW_NEXT, VL, 0} }; static const hc2c_desc desc = { 2, XSIMD_STRING("hc2cbdftv_2"), twinstr, &GENUS, {3, 2, 2, 0} }; void XSIMD(codelet_hc2cbdftv_2) (planner *p) { X(khc2c_register) (p, hc2cbdftv_2, &desc, HC2C_VIA_DFT); } #else /* HAVE_FMA */ /* Generated by: ../../../genfft/gen_hc2cdft_c.native -simd -compact -variables 4 -pipeline-latency 8 -trivial-stores -variables 32 -no-generate-bytw -n 2 -dif -sign 1 -name hc2cbdftv_2 -include hc2cbv.h */ /* * This function contains 5 FP additions, 2 FP multiplications, * (or, 5 additions, 2 multiplications, 0 fused multiply/add), * 9 stack variables, 0 constants, and 4 memory accesses */ #include "hc2cbv.h" static void hc2cbdftv_2(R *Rp, R *Ip, R *Rm, R *Im, const R *W, stride rs, INT mb, INT me, INT ms) { { INT m; for (m = mb, W = W + ((mb - 1) * ((TWVL / VL) * 2)); m < me; m = m + VL, Rp = Rp + (VL * ms), Ip = Ip + (VL * ms), Rm = Rm - (VL * ms), Im = Im - (VL * ms), W = W + (TWVL * 2), MAKE_VOLATILE_STRIDE(8, rs)) { V T6, T5, T2, T4, T3, T1, T7, T8; T2 = LD(&(Rp[0]), ms, &(Rp[0])); T3 = LD(&(Rm[0]), -ms, &(Rm[0])); T4 = VCONJ(T3); T6 = VADD(T2, T4); T1 = LDW(&(W[0])); T5 = VZMULI(T1, VSUB(T2, T4)); T7 = VADD(T5, T6); ST(&(Rp[0]), T7, ms, &(Rp[0])); T8 = VCONJ(VSUB(T6, T5)); ST(&(Rm[0]), T8, -ms, &(Rm[0])); } } VLEAVE(); } static const tw_instr twinstr[] = { VTW(1, 1), {TW_NEXT, VL, 0} }; static const hc2c_desc desc = { 2, XSIMD_STRING("hc2cbdftv_2"), twinstr, &GENUS, {5, 2, 0, 0} }; void XSIMD(codelet_hc2cbdftv_2) (planner *p) { X(khc2c_register) (p, hc2cbdftv_2, &desc, HC2C_VIA_DFT); } #endif /* HAVE_FMA */ fftw-3.3.3/rdft/simd/common/hc2cbdftv_4.c0000644000175000017500000001225312054411065015027 00000000000000/* * Copyright (c) 2003, 2007-11 Matteo Frigo * Copyright (c) 2003, 2007-11 Massachusetts Institute of Technology * * This program is free software; you can redistribute it and/or modify * it under the terms of the GNU General Public License as published by * the Free Software Foundation; either version 2 of the License, or * (at your option) any later version. * * This program is distributed in the hope that it will be useful, * but WITHOUT ANY WARRANTY; without even the implied warranty of * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the * GNU General Public License for more details. * * You should have received a copy of the GNU General Public License * along with this program; if not, write to the Free Software * Foundation, Inc., 51 Franklin Street, Fifth Floor, Boston, MA 02110-1301 USA * */ /* This file was automatically generated --- DO NOT EDIT */ /* Generated on Sun Nov 25 07:42:29 EST 2012 */ #include "codelet-rdft.h" #ifdef HAVE_FMA /* Generated by: ../../../genfft/gen_hc2cdft_c.native -fma -reorder-insns -schedule-for-pipeline -simd -compact -variables 4 -pipeline-latency 8 -trivial-stores -variables 32 -no-generate-bytw -n 4 -dif -sign 1 -name hc2cbdftv_4 -include hc2cbv.h */ /* * This function contains 15 FP additions, 12 FP multiplications, * (or, 9 additions, 6 multiplications, 6 fused multiply/add), * 20 stack variables, 0 constants, and 8 memory accesses */ #include "hc2cbv.h" static void hc2cbdftv_4(R *Rp, R *Ip, R *Rm, R *Im, const R *W, stride rs, INT mb, INT me, INT ms) { { INT m; for (m = mb, W = W + ((mb - 1) * ((TWVL / VL) * 6)); m < me; m = m + VL, Rp = Rp + (VL * ms), Ip = Ip + (VL * ms), Rm = Rm - (VL * ms), Im = Im - (VL * ms), W = W + (TWVL * 6), MAKE_VOLATILE_STRIDE(16, rs)) { V T2, T3, T5, T6, Tf, T1, T9, Ta, T4, Tb, T7, Tc, Th, T8, Tg; V Te, Td, Ti, Tj; T2 = LD(&(Rp[0]), ms, &(Rp[0])); T3 = LD(&(Rm[WS(rs, 1)]), -ms, &(Rm[WS(rs, 1)])); T5 = LD(&(Rp[WS(rs, 1)]), ms, &(Rp[WS(rs, 1)])); T6 = LD(&(Rm[0]), -ms, &(Rm[0])); Tf = LDW(&(W[0])); T1 = LDW(&(W[TWVL * 4])); T9 = LDW(&(W[TWVL * 2])); Ta = VFMACONJ(T3, T2); T4 = VFNMSCONJ(T3, T2); Tb = VFMACONJ(T6, T5); T7 = VFNMSCONJ(T6, T5); Tc = VZMUL(T9, VSUB(Ta, Tb)); Th = VADD(Ta, Tb); T8 = VZMULI(T1, VFNMSI(T7, T4)); Tg = VZMULI(Tf, VFMAI(T7, T4)); Te = VCONJ(VSUB(Tc, T8)); Td = VADD(T8, Tc); Ti = VADD(Tg, Th); Tj = VCONJ(VSUB(Th, Tg)); ST(&(Rm[WS(rs, 1)]), Te, -ms, &(Rm[WS(rs, 1)])); ST(&(Rp[WS(rs, 1)]), Td, ms, &(Rp[WS(rs, 1)])); ST(&(Rp[0]), Ti, ms, &(Rp[0])); ST(&(Rm[0]), Tj, -ms, &(Rm[0])); } } VLEAVE(); } static const tw_instr twinstr[] = { VTW(1, 1), VTW(1, 2), VTW(1, 3), {TW_NEXT, VL, 0} }; static const hc2c_desc desc = { 4, XSIMD_STRING("hc2cbdftv_4"), twinstr, &GENUS, {9, 6, 6, 0} }; void XSIMD(codelet_hc2cbdftv_4) (planner *p) { X(khc2c_register) (p, hc2cbdftv_4, &desc, HC2C_VIA_DFT); } #else /* HAVE_FMA */ /* Generated by: ../../../genfft/gen_hc2cdft_c.native -simd -compact -variables 4 -pipeline-latency 8 -trivial-stores -variables 32 -no-generate-bytw -n 4 -dif -sign 1 -name hc2cbdftv_4 -include hc2cbv.h */ /* * This function contains 15 FP additions, 6 FP multiplications, * (or, 15 additions, 6 multiplications, 0 fused multiply/add), * 22 stack variables, 0 constants, and 8 memory accesses */ #include "hc2cbv.h" static void hc2cbdftv_4(R *Rp, R *Ip, R *Rm, R *Im, const R *W, stride rs, INT mb, INT me, INT ms) { { INT m; for (m = mb, W = W + ((mb - 1) * ((TWVL / VL) * 6)); m < me; m = m + VL, Rp = Rp + (VL * ms), Ip = Ip + (VL * ms), Rm = Rm - (VL * ms), Im = Im - (VL * ms), W = W + (TWVL * 6), MAKE_VOLATILE_STRIDE(16, rs)) { V T5, Tc, T9, Td, T2, T4, T3, T6, T8, T7, Tj, Ti, Th, Tk, Tl; V Ta, Te, T1, Tb, Tf, Tg; T2 = LD(&(Rp[0]), ms, &(Rp[0])); T3 = LD(&(Rm[WS(rs, 1)]), -ms, &(Rm[WS(rs, 1)])); T4 = VCONJ(T3); T5 = VSUB(T2, T4); Tc = VADD(T2, T4); T6 = LD(&(Rp[WS(rs, 1)]), ms, &(Rp[WS(rs, 1)])); T7 = LD(&(Rm[0]), -ms, &(Rm[0])); T8 = VCONJ(T7); T9 = VBYI(VSUB(T6, T8)); Td = VADD(T6, T8); Tj = VADD(Tc, Td); Th = LDW(&(W[0])); Ti = VZMULI(Th, VADD(T5, T9)); Tk = VADD(Ti, Tj); ST(&(Rp[0]), Tk, ms, &(Rp[0])); Tl = VCONJ(VSUB(Tj, Ti)); ST(&(Rm[0]), Tl, -ms, &(Rm[0])); T1 = LDW(&(W[TWVL * 4])); Ta = VZMULI(T1, VSUB(T5, T9)); Tb = LDW(&(W[TWVL * 2])); Te = VZMUL(Tb, VSUB(Tc, Td)); Tf = VADD(Ta, Te); ST(&(Rp[WS(rs, 1)]), Tf, ms, &(Rp[WS(rs, 1)])); Tg = VCONJ(VSUB(Te, Ta)); ST(&(Rm[WS(rs, 1)]), Tg, -ms, &(Rm[WS(rs, 1)])); } } VLEAVE(); } static const tw_instr twinstr[] = { VTW(1, 1), VTW(1, 2), VTW(1, 3), {TW_NEXT, VL, 0} }; static const hc2c_desc desc = { 4, XSIMD_STRING("hc2cbdftv_4"), twinstr, &GENUS, {15, 6, 0, 0} }; void XSIMD(codelet_hc2cbdftv_4) (planner *p) { X(khc2c_register) (p, hc2cbdftv_4, &desc, HC2C_VIA_DFT); } #endif /* HAVE_FMA */ fftw-3.3.3/rdft/simd/common/hc2cbdftv_6.c0000644000175000017500000001613712054411065015036 00000000000000/* * Copyright (c) 2003, 2007-11 Matteo Frigo * Copyright (c) 2003, 2007-11 Massachusetts Institute of Technology * * This program is free software; you can redistribute it and/or modify * it under the terms of the GNU General Public License as published by * the Free Software Foundation; either version 2 of the License, or * (at your option) any later version. * * This program is distributed in the hope that it will be useful, * but WITHOUT ANY WARRANTY; without even the implied warranty of * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the * GNU General Public License for more details. * * You should have received a copy of the GNU General Public License * along with this program; if not, write to the Free Software * Foundation, Inc., 51 Franklin Street, Fifth Floor, Boston, MA 02110-1301 USA * */ /* This file was automatically generated --- DO NOT EDIT */ /* Generated on Sun Nov 25 07:42:29 EST 2012 */ #include "codelet-rdft.h" #ifdef HAVE_FMA /* Generated by: ../../../genfft/gen_hc2cdft_c.native -fma -reorder-insns -schedule-for-pipeline -simd -compact -variables 4 -pipeline-latency 8 -trivial-stores -variables 32 -no-generate-bytw -n 6 -dif -sign 1 -name hc2cbdftv_6 -include hc2cbv.h */ /* * This function contains 29 FP additions, 24 FP multiplications, * (or, 17 additions, 12 multiplications, 12 fused multiply/add), * 38 stack variables, 2 constants, and 12 memory accesses */ #include "hc2cbv.h" static void hc2cbdftv_6(R *Rp, R *Ip, R *Rm, R *Im, const R *W, stride rs, INT mb, INT me, INT ms) { DVK(KP500000000, +0.500000000000000000000000000000000000000000000); DVK(KP866025403, +0.866025403784438646763723170752936183471402627); { INT m; for (m = mb, W = W + ((mb - 1) * ((TWVL / VL) * 10)); m < me; m = m + VL, Rp = Rp + (VL * ms), Ip = Ip + (VL * ms), Rm = Rm - (VL * ms), Im = Im - (VL * ms), W = W + (TWVL * 10), MAKE_VOLATILE_STRIDE(24, rs)) { V Tv, Tn, Tr, Te, T4, Tg, Ta, Tf, T7, T1, Td, T2, T3, T8, T9; V T5, T6, Th, Tj, Tb, Tp, Tx, Ti, Tc, To, Tk, Ts, Tq, Tw, Tm; V Tl, Tu, Tt, Tz, Ty; T2 = LD(&(Rp[0]), ms, &(Rp[0])); T3 = LD(&(Rm[WS(rs, 2)]), -ms, &(Rm[0])); T8 = LD(&(Rp[WS(rs, 1)]), ms, &(Rp[WS(rs, 1)])); T9 = LD(&(Rm[WS(rs, 1)]), -ms, &(Rm[WS(rs, 1)])); T5 = LD(&(Rp[WS(rs, 2)]), ms, &(Rp[0])); T6 = LD(&(Rm[0]), -ms, &(Rm[0])); Tv = LDW(&(W[0])); Tn = LDW(&(W[TWVL * 8])); Tr = LDW(&(W[TWVL * 6])); Te = VFMACONJ(T3, T2); T4 = VFNMSCONJ(T3, T2); Tg = VFMACONJ(T9, T8); Ta = VFMSCONJ(T9, T8); Tf = VFMACONJ(T6, T5); T7 = VFNMSCONJ(T6, T5); T1 = LDW(&(W[TWVL * 4])); Td = LDW(&(W[TWVL * 2])); Th = VADD(Tf, Tg); Tj = VMUL(LDK(KP866025403), VSUB(Tf, Tg)); Tb = VADD(T7, Ta); Tp = VMUL(LDK(KP866025403), VSUB(T7, Ta)); Tx = VADD(Te, Th); Ti = VFNMS(LDK(KP500000000), Th, Te); Tc = VZMULI(T1, VADD(T4, Tb)); To = VFNMS(LDK(KP500000000), Tb, T4); Tk = VZMUL(Td, VFNMSI(Tj, Ti)); Ts = VZMUL(Tr, VFMAI(Tj, Ti)); Tq = VZMULI(Tn, VFNMSI(Tp, To)); Tw = VZMULI(Tv, VFMAI(Tp, To)); Tm = VCONJ(VSUB(Tk, Tc)); Tl = VADD(Tc, Tk); Tu = VCONJ(VSUB(Ts, Tq)); Tt = VADD(Tq, Ts); Tz = VCONJ(VSUB(Tx, Tw)); Ty = VADD(Tw, Tx); ST(&(Rm[WS(rs, 1)]), Tm, -ms, &(Rm[WS(rs, 1)])); ST(&(Rp[WS(rs, 1)]), Tl, ms, &(Rp[WS(rs, 1)])); ST(&(Rm[WS(rs, 2)]), Tu, -ms, &(Rm[0])); ST(&(Rp[WS(rs, 2)]), Tt, ms, &(Rp[0])); ST(&(Rm[0]), Tz, -ms, &(Rm[0])); ST(&(Rp[0]), Ty, ms, &(Rp[0])); } } VLEAVE(); } static const tw_instr twinstr[] = { VTW(1, 1), VTW(1, 2), VTW(1, 3), VTW(1, 4), VTW(1, 5), {TW_NEXT, VL, 0} }; static const hc2c_desc desc = { 6, XSIMD_STRING("hc2cbdftv_6"), twinstr, &GENUS, {17, 12, 12, 0} }; void XSIMD(codelet_hc2cbdftv_6) (planner *p) { X(khc2c_register) (p, hc2cbdftv_6, &desc, HC2C_VIA_DFT); } #else /* HAVE_FMA */ /* Generated by: ../../../genfft/gen_hc2cdft_c.native -simd -compact -variables 4 -pipeline-latency 8 -trivial-stores -variables 32 -no-generate-bytw -n 6 -dif -sign 1 -name hc2cbdftv_6 -include hc2cbv.h */ /* * This function contains 29 FP additions, 14 FP multiplications, * (or, 27 additions, 12 multiplications, 2 fused multiply/add), * 41 stack variables, 2 constants, and 12 memory accesses */ #include "hc2cbv.h" static void hc2cbdftv_6(R *Rp, R *Ip, R *Rm, R *Im, const R *W, stride rs, INT mb, INT me, INT ms) { DVK(KP500000000, +0.500000000000000000000000000000000000000000000); DVK(KP866025403, +0.866025403784438646763723170752936183471402627); { INT m; for (m = mb, W = W + ((mb - 1) * ((TWVL / VL) * 10)); m < me; m = m + VL, Rp = Rp + (VL * ms), Ip = Ip + (VL * ms), Rm = Rm - (VL * ms), Im = Im - (VL * ms), W = W + (TWVL * 10), MAKE_VOLATILE_STRIDE(24, rs)) { V T5, Th, Te, Ts, Tk, Tm, T2, T4, T3, T6, Tc, T8, Tb, T7, Ta; V T9, Td, Ti, Tj, TA, Tf, Tn, Tv, Tt, Tz, T1, Tl, Tg, Tu, Tr; V Tq, Ty, To, Tp, TC, TB, Tx, Tw; T2 = LD(&(Rp[0]), ms, &(Rp[0])); T3 = LD(&(Rm[WS(rs, 2)]), -ms, &(Rm[0])); T4 = VCONJ(T3); T5 = VSUB(T2, T4); Th = VADD(T2, T4); T6 = LD(&(Rp[WS(rs, 2)]), ms, &(Rp[0])); Tc = LD(&(Rp[WS(rs, 1)]), ms, &(Rp[WS(rs, 1)])); T7 = LD(&(Rm[0]), -ms, &(Rm[0])); T8 = VCONJ(T7); Ta = LD(&(Rm[WS(rs, 1)]), -ms, &(Rm[WS(rs, 1)])); Tb = VCONJ(Ta); T9 = VSUB(T6, T8); Td = VSUB(Tb, Tc); Te = VADD(T9, Td); Ts = VBYI(VMUL(LDK(KP866025403), VSUB(T9, Td))); Ti = VADD(T6, T8); Tj = VADD(Tb, Tc); Tk = VADD(Ti, Tj); Tm = VBYI(VMUL(LDK(KP866025403), VSUB(Ti, Tj))); TA = VADD(Th, Tk); T1 = LDW(&(W[TWVL * 4])); Tf = VZMULI(T1, VADD(T5, Te)); Tl = VFNMS(LDK(KP500000000), Tk, Th); Tg = LDW(&(W[TWVL * 2])); Tn = VZMUL(Tg, VSUB(Tl, Tm)); Tu = LDW(&(W[TWVL * 6])); Tv = VZMUL(Tu, VADD(Tm, Tl)); Tr = VFNMS(LDK(KP500000000), Te, T5); Tq = LDW(&(W[TWVL * 8])); Tt = VZMULI(Tq, VSUB(Tr, Ts)); Ty = LDW(&(W[0])); Tz = VZMULI(Ty, VADD(Ts, Tr)); To = VADD(Tf, Tn); ST(&(Rp[WS(rs, 1)]), To, ms, &(Rp[WS(rs, 1)])); Tp = VCONJ(VSUB(Tn, Tf)); ST(&(Rm[WS(rs, 1)]), Tp, -ms, &(Rm[WS(rs, 1)])); TC = VCONJ(VSUB(TA, Tz)); ST(&(Rm[0]), TC, -ms, &(Rm[0])); TB = VADD(Tz, TA); ST(&(Rp[0]), TB, ms, &(Rp[0])); Tx = VCONJ(VSUB(Tv, Tt)); ST(&(Rm[WS(rs, 2)]), Tx, -ms, &(Rm[0])); Tw = VADD(Tt, Tv); ST(&(Rp[WS(rs, 2)]), Tw, ms, &(Rp[0])); } } VLEAVE(); } static const tw_instr twinstr[] = { VTW(1, 1), VTW(1, 2), VTW(1, 3), VTW(1, 4), VTW(1, 5), {TW_NEXT, VL, 0} }; static const hc2c_desc desc = { 6, XSIMD_STRING("hc2cbdftv_6"), twinstr, &GENUS, {27, 12, 2, 0} }; void XSIMD(codelet_hc2cbdftv_6) (planner *p) { X(khc2c_register) (p, hc2cbdftv_6, &desc, HC2C_VIA_DFT); } #endif /* HAVE_FMA */ fftw-3.3.3/rdft/simd/common/hc2cbdftv_8.c0000644000175000017500000002047112054411065015034 00000000000000/* * Copyright (c) 2003, 2007-11 Matteo Frigo * Copyright (c) 2003, 2007-11 Massachusetts Institute of Technology * * This program is free software; you can redistribute it and/or modify * it under the terms of the GNU General Public License as published by * the Free Software Foundation; either version 2 of the License, or * (at your option) any later version. * * This program is distributed in the hope that it will be useful, * but WITHOUT ANY WARRANTY; without even the implied warranty of * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the * GNU General Public License for more details. * * You should have received a copy of the GNU General Public License * along with this program; if not, write to the Free Software * Foundation, Inc., 51 Franklin Street, Fifth Floor, Boston, MA 02110-1301 USA * */ /* This file was automatically generated --- DO NOT EDIT */ /* Generated on Sun Nov 25 07:42:29 EST 2012 */ #include "codelet-rdft.h" #ifdef HAVE_FMA /* Generated by: ../../../genfft/gen_hc2cdft_c.native -fma -reorder-insns -schedule-for-pipeline -simd -compact -variables 4 -pipeline-latency 8 -trivial-stores -variables 32 -no-generate-bytw -n 8 -dif -sign 1 -name hc2cbdftv_8 -include hc2cbv.h */ /* * This function contains 41 FP additions, 32 FP multiplications, * (or, 23 additions, 14 multiplications, 18 fused multiply/add), * 51 stack variables, 1 constants, and 16 memory accesses */ #include "hc2cbv.h" static void hc2cbdftv_8(R *Rp, R *Ip, R *Rm, R *Im, const R *W, stride rs, INT mb, INT me, INT ms) { DVK(KP707106781, +0.707106781186547524400844362104849039284835938); { INT m; for (m = mb, W = W + ((mb - 1) * ((TWVL / VL) * 14)); m < me; m = m + VL, Rp = Rp + (VL * ms), Ip = Ip + (VL * ms), Rm = Rm - (VL * ms), Im = Im - (VL * ms), W = W + (TWVL * 14), MAKE_VOLATILE_STRIDE(32, rs)) { V TJ, T4, Tf, TB, TD, TE, Tm, T1, Tj, TF, Tp, Tb, Tg, Tt, Tx; V T2, T3, Td, Te, T5, T6, T8, T9, Tn, T7, To, Ta, Tk, Tl, TG; V TL, Tq, Tc, Tu, Th, Tv, Ty, Tw, TC, Ti, TK, TA, Tz, TI, TH; V Ts, Tr, TN, TM; T2 = LD(&(Rp[0]), ms, &(Rp[0])); T3 = LD(&(Rm[WS(rs, 3)]), -ms, &(Rm[WS(rs, 1)])); Td = LD(&(Rp[WS(rs, 2)]), ms, &(Rp[0])); Te = LD(&(Rm[WS(rs, 1)]), -ms, &(Rm[WS(rs, 1)])); T5 = LD(&(Rp[WS(rs, 1)]), ms, &(Rp[WS(rs, 1)])); T6 = LD(&(Rm[WS(rs, 2)]), -ms, &(Rm[0])); T8 = LD(&(Rp[WS(rs, 3)]), ms, &(Rp[WS(rs, 1)])); T9 = LD(&(Rm[0]), -ms, &(Rm[0])); TJ = LDW(&(W[0])); Tk = VFMACONJ(T3, T2); T4 = VFNMSCONJ(T3, T2); Tl = VFMACONJ(Te, Td); Tf = VFNMSCONJ(Te, Td); Tn = VFMACONJ(T6, T5); T7 = VFNMSCONJ(T6, T5); To = VFMACONJ(T9, T8); Ta = VFMSCONJ(T9, T8); TB = LDW(&(W[TWVL * 8])); TD = LDW(&(W[TWVL * 6])); TE = VADD(Tk, Tl); Tm = VSUB(Tk, Tl); T1 = LDW(&(W[TWVL * 12])); Tj = LDW(&(W[TWVL * 10])); TF = VADD(Tn, To); Tp = VSUB(Tn, To); Tb = VADD(T7, Ta); Tg = VSUB(T7, Ta); Tt = LDW(&(W[TWVL * 4])); Tx = LDW(&(W[TWVL * 2])); TG = VZMUL(TD, VSUB(TE, TF)); TL = VADD(TE, TF); Tq = VZMUL(Tj, VFNMSI(Tp, Tm)); Tc = VFMA(LDK(KP707106781), Tb, T4); Tu = VFNMS(LDK(KP707106781), Tb, T4); Th = VFMA(LDK(KP707106781), Tg, Tf); Tv = VFNMS(LDK(KP707106781), Tg, Tf); Ty = VZMUL(Tx, VFMAI(Tp, Tm)); Tw = VZMULI(Tt, VFNMSI(Tv, Tu)); TC = VZMULI(TB, VFMAI(Tv, Tu)); Ti = VZMULI(T1, VFNMSI(Th, Tc)); TK = VZMULI(TJ, VFMAI(Th, Tc)); TA = VCONJ(VSUB(Ty, Tw)); Tz = VADD(Tw, Ty); TI = VCONJ(VSUB(TG, TC)); TH = VADD(TC, TG); Ts = VCONJ(VSUB(Tq, Ti)); Tr = VADD(Ti, Tq); TN = VCONJ(VSUB(TL, TK)); TM = VADD(TK, TL); ST(&(Rm[WS(rs, 1)]), TA, -ms, &(Rm[WS(rs, 1)])); ST(&(Rp[WS(rs, 1)]), Tz, ms, &(Rp[WS(rs, 1)])); ST(&(Rm[WS(rs, 2)]), TI, -ms, &(Rm[0])); ST(&(Rp[WS(rs, 2)]), TH, ms, &(Rp[0])); ST(&(Rm[WS(rs, 3)]), Ts, -ms, &(Rm[WS(rs, 1)])); ST(&(Rp[WS(rs, 3)]), Tr, ms, &(Rp[WS(rs, 1)])); ST(&(Rm[0]), TN, -ms, &(Rm[0])); ST(&(Rp[0]), TM, ms, &(Rp[0])); } } VLEAVE(); } static const tw_instr twinstr[] = { VTW(1, 1), VTW(1, 2), VTW(1, 3), VTW(1, 4), VTW(1, 5), VTW(1, 6), VTW(1, 7), {TW_NEXT, VL, 0} }; static const hc2c_desc desc = { 8, XSIMD_STRING("hc2cbdftv_8"), twinstr, &GENUS, {23, 14, 18, 0} }; void XSIMD(codelet_hc2cbdftv_8) (planner *p) { X(khc2c_register) (p, hc2cbdftv_8, &desc, HC2C_VIA_DFT); } #else /* HAVE_FMA */ /* Generated by: ../../../genfft/gen_hc2cdft_c.native -simd -compact -variables 4 -pipeline-latency 8 -trivial-stores -variables 32 -no-generate-bytw -n 8 -dif -sign 1 -name hc2cbdftv_8 -include hc2cbv.h */ /* * This function contains 41 FP additions, 16 FP multiplications, * (or, 41 additions, 16 multiplications, 0 fused multiply/add), * 55 stack variables, 1 constants, and 16 memory accesses */ #include "hc2cbv.h" static void hc2cbdftv_8(R *Rp, R *Ip, R *Rm, R *Im, const R *W, stride rs, INT mb, INT me, INT ms) { DVK(KP707106781, +0.707106781186547524400844362104849039284835938); { INT m; for (m = mb, W = W + ((mb - 1) * ((TWVL / VL) * 14)); m < me; m = m + VL, Rp = Rp + (VL * ms), Ip = Ip + (VL * ms), Rm = Rm - (VL * ms), Im = Im - (VL * ms), W = W + (TWVL * 14), MAKE_VOLATILE_STRIDE(32, rs)) { V T5, Tj, Tq, TI, Te, Tk, Tt, TJ, T2, Tg, T4, Ti, T3, Th, To; V Tp, T6, Tc, T8, Tb, T7, Ta, T9, Td, Tr, Ts, TP, Tu, Tm, TO; V Tn, Tf, Tl, T1, TN, Tv, TR, Tw, TQ, TC, TK, TA, TG, TB, TH; V Ty, Tz, Tx, TF, TD, TM, TE, TL; T2 = LD(&(Rp[0]), ms, &(Rp[0])); Tg = LD(&(Rp[WS(rs, 2)]), ms, &(Rp[0])); T3 = LD(&(Rm[WS(rs, 3)]), -ms, &(Rm[WS(rs, 1)])); T4 = VCONJ(T3); Th = LD(&(Rm[WS(rs, 1)]), -ms, &(Rm[WS(rs, 1)])); Ti = VCONJ(Th); T5 = VSUB(T2, T4); Tj = VSUB(Tg, Ti); To = VADD(T2, T4); Tp = VADD(Tg, Ti); Tq = VSUB(To, Tp); TI = VADD(To, Tp); T6 = LD(&(Rp[WS(rs, 1)]), ms, &(Rp[WS(rs, 1)])); Tc = LD(&(Rp[WS(rs, 3)]), ms, &(Rp[WS(rs, 1)])); T7 = LD(&(Rm[WS(rs, 2)]), -ms, &(Rm[0])); T8 = VCONJ(T7); Ta = LD(&(Rm[0]), -ms, &(Rm[0])); Tb = VCONJ(Ta); T9 = VSUB(T6, T8); Td = VSUB(Tb, Tc); Te = VMUL(LDK(KP707106781), VADD(T9, Td)); Tk = VMUL(LDK(KP707106781), VSUB(T9, Td)); Tr = VADD(T6, T8); Ts = VADD(Tb, Tc); Tt = VBYI(VSUB(Tr, Ts)); TJ = VADD(Tr, Ts); TP = VADD(TI, TJ); Tn = LDW(&(W[TWVL * 10])); Tu = VZMUL(Tn, VSUB(Tq, Tt)); Tf = VADD(T5, Te); Tl = VBYI(VADD(Tj, Tk)); T1 = LDW(&(W[TWVL * 12])); Tm = VZMULI(T1, VSUB(Tf, Tl)); TN = LDW(&(W[0])); TO = VZMULI(TN, VADD(Tl, Tf)); Tv = VADD(Tm, Tu); ST(&(Rp[WS(rs, 3)]), Tv, ms, &(Rp[WS(rs, 1)])); TR = VCONJ(VSUB(TP, TO)); ST(&(Rm[0]), TR, -ms, &(Rm[0])); Tw = VCONJ(VSUB(Tu, Tm)); ST(&(Rm[WS(rs, 3)]), Tw, -ms, &(Rm[WS(rs, 1)])); TQ = VADD(TO, TP); ST(&(Rp[0]), TQ, ms, &(Rp[0])); TB = LDW(&(W[TWVL * 2])); TC = VZMUL(TB, VADD(Tq, Tt)); TH = LDW(&(W[TWVL * 6])); TK = VZMUL(TH, VSUB(TI, TJ)); Ty = VBYI(VSUB(Tk, Tj)); Tz = VSUB(T5, Te); Tx = LDW(&(W[TWVL * 4])); TA = VZMULI(Tx, VADD(Ty, Tz)); TF = LDW(&(W[TWVL * 8])); TG = VZMULI(TF, VSUB(Tz, Ty)); TD = VADD(TA, TC); ST(&(Rp[WS(rs, 1)]), TD, ms, &(Rp[WS(rs, 1)])); TM = VCONJ(VSUB(TK, TG)); ST(&(Rm[WS(rs, 2)]), TM, -ms, &(Rm[0])); TE = VCONJ(VSUB(TC, TA)); ST(&(Rm[WS(rs, 1)]), TE, -ms, &(Rm[WS(rs, 1)])); TL = VADD(TG, TK); ST(&(Rp[WS(rs, 2)]), TL, ms, &(Rp[0])); } } VLEAVE(); } static const tw_instr twinstr[] = { VTW(1, 1), VTW(1, 2), VTW(1, 3), VTW(1, 4), VTW(1, 5), VTW(1, 6), VTW(1, 7), {TW_NEXT, VL, 0} }; static const hc2c_desc desc = { 8, XSIMD_STRING("hc2cbdftv_8"), twinstr, &GENUS, {41, 16, 0, 0} }; void XSIMD(codelet_hc2cbdftv_8) (planner *p) { X(khc2c_register) (p, hc2cbdftv_8, &desc, HC2C_VIA_DFT); } #endif /* HAVE_FMA */ fftw-3.3.3/rdft/simd/common/hc2cbdftv_10.c0000644000175000017500000002566512054411066015120 00000000000000/* * Copyright (c) 2003, 2007-11 Matteo Frigo * Copyright (c) 2003, 2007-11 Massachusetts Institute of Technology * * This program is free software; you can redistribute it and/or modify * it under the terms of the GNU General Public License as published by * the Free Software Foundation; either version 2 of the License, or * (at your option) any later version. * * This program is distributed in the hope that it will be useful, * but WITHOUT ANY WARRANTY; without even the implied warranty of * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the * GNU General Public License for more details. * * You should have received a copy of the GNU General Public License * along with this program; if not, write to the Free Software * Foundation, Inc., 51 Franklin Street, Fifth Floor, Boston, MA 02110-1301 USA * */ /* This file was automatically generated --- DO NOT EDIT */ /* Generated on Sun Nov 25 07:42:29 EST 2012 */ #include "codelet-rdft.h" #ifdef HAVE_FMA /* Generated by: ../../../genfft/gen_hc2cdft_c.native -fma -reorder-insns -schedule-for-pipeline -simd -compact -variables 4 -pipeline-latency 8 -trivial-stores -variables 32 -no-generate-bytw -n 10 -dif -sign 1 -name hc2cbdftv_10 -include hc2cbv.h */ /* * This function contains 61 FP additions, 50 FP multiplications, * (or, 33 additions, 22 multiplications, 28 fused multiply/add), * 76 stack variables, 4 constants, and 20 memory accesses */ #include "hc2cbv.h" static void hc2cbdftv_10(R *Rp, R *Ip, R *Rm, R *Im, const R *W, stride rs, INT mb, INT me, INT ms) { DVK(KP559016994, +0.559016994374947424102293417182819058860154590); DVK(KP250000000, +0.250000000000000000000000000000000000000000000); DVK(KP618033988, +0.618033988749894848204586834365638117720309180); DVK(KP951056516, +0.951056516295153572116439333379382143405698634); { INT m; for (m = mb, W = W + ((mb - 1) * ((TWVL / VL) * 18)); m < me; m = m + VL, Rp = Rp + (VL * ms), Ip = Ip + (VL * ms), Rm = Rm - (VL * ms), Im = Im - (VL * ms), W = W + (TWVL * 18), MAKE_VOLATILE_STRIDE(40, rs)) { V Ts, T4, TR, T1, TZ, TD, Ty, Tn, Ti, TT, T11, TJ, T15, Tr, TN; V TE, Tv, To, Tb, T8, Tw, Te, Tx, Th, Tt, T7, T9, T2, T3, Tc; V Td, Tf, Tg, T5, T6, Tu, Ta; T2 = LD(&(Rp[0]), ms, &(Rp[0])); T3 = LD(&(Rm[WS(rs, 4)]), -ms, &(Rm[0])); Tc = LD(&(Rp[WS(rs, 4)]), ms, &(Rp[0])); Td = LD(&(Rm[0]), -ms, &(Rm[0])); Tf = LD(&(Rp[WS(rs, 1)]), ms, &(Rp[WS(rs, 1)])); Tg = LD(&(Rm[WS(rs, 3)]), -ms, &(Rm[WS(rs, 1)])); T5 = LD(&(Rp[WS(rs, 2)]), ms, &(Rp[0])); T6 = LD(&(Rm[WS(rs, 2)]), -ms, &(Rm[0])); T8 = LD(&(Rp[WS(rs, 3)]), ms, &(Rp[WS(rs, 1)])); Ts = VFMACONJ(T3, T2); T4 = VFNMSCONJ(T3, T2); Tw = VFMACONJ(Td, Tc); Te = VFNMSCONJ(Td, Tc); Tx = VFMACONJ(Tg, Tf); Th = VFMSCONJ(Tg, Tf); Tt = VFMACONJ(T6, T5); T7 = VFNMSCONJ(T6, T5); T9 = LD(&(Rm[WS(rs, 1)]), -ms, &(Rm[WS(rs, 1)])); TR = LDW(&(W[TWVL * 8])); T1 = LDW(&(W[TWVL * 4])); TZ = LDW(&(W[TWVL * 12])); TD = VSUB(Tw, Tx); Ty = VADD(Tw, Tx); Tn = VSUB(Te, Th); Ti = VADD(Te, Th); Tu = VFMACONJ(T9, T8); Ta = VFMSCONJ(T9, T8); TT = LDW(&(W[TWVL * 6])); T11 = LDW(&(W[TWVL * 10])); TJ = LDW(&(W[TWVL * 16])); T15 = LDW(&(W[0])); Tr = LDW(&(W[TWVL * 2])); TN = LDW(&(W[TWVL * 14])); TE = VSUB(Tt, Tu); Tv = VADD(Tt, Tu); To = VSUB(T7, Ta); Tb = VADD(T7, Ta); { V TV, TF, Tz, TB, TL, Tp, Tj, Tl, T17, TA, TS, Tk, TC, TU, TK; V Tm, TO, TG, T12, TW, T16, TM, T10, Tq, TX, TY, T18, T19, TQ, TP; V T13, T14, TI, TH; TV = VMUL(LDK(KP951056516), VFMA(LDK(KP618033988), TD, TE)); TF = VMUL(LDK(KP951056516), VFNMS(LDK(KP618033988), TE, TD)); Tz = VADD(Tv, Ty); TB = VSUB(Tv, Ty); TL = VMUL(LDK(KP951056516), VFMA(LDK(KP618033988), Tn, To)); Tp = VMUL(LDK(KP951056516), VFNMS(LDK(KP618033988), To, Tn)); Tj = VADD(Tb, Ti); Tl = VSUB(Tb, Ti); T17 = VADD(Ts, Tz); TA = VFNMS(LDK(KP250000000), Tz, Ts); TS = VZMULI(TR, VADD(T4, Tj)); Tk = VFNMS(LDK(KP250000000), Tj, T4); TC = VFNMS(LDK(KP559016994), TB, TA); TU = VFMA(LDK(KP559016994), TB, TA); TK = VFMA(LDK(KP559016994), Tl, Tk); Tm = VFNMS(LDK(KP559016994), Tl, Tk); TO = VZMUL(TN, VFMAI(TF, TC)); TG = VZMUL(Tr, VFNMSI(TF, TC)); T12 = VZMUL(T11, VFMAI(TV, TU)); TW = VZMUL(TT, VFNMSI(TV, TU)); T16 = VZMULI(T15, VFMAI(TL, TK)); TM = VZMULI(TJ, VFNMSI(TL, TK)); T10 = VZMULI(TZ, VFNMSI(Tp, Tm)); Tq = VZMULI(T1, VFMAI(Tp, Tm)); TX = VADD(TS, TW); TY = VCONJ(VSUB(TW, TS)); T18 = VADD(T16, T17); T19 = VCONJ(VSUB(T17, T16)); TQ = VCONJ(VSUB(TO, TM)); TP = VADD(TM, TO); T13 = VADD(T10, T12); T14 = VCONJ(VSUB(T12, T10)); TI = VCONJ(VSUB(TG, Tq)); TH = VADD(Tq, TG); ST(&(Rp[WS(rs, 2)]), TX, ms, &(Rp[0])); ST(&(Rm[WS(rs, 2)]), TY, -ms, &(Rm[0])); ST(&(Rp[0]), T18, ms, &(Rp[0])); ST(&(Rm[0]), T19, -ms, &(Rm[0])); ST(&(Rm[WS(rs, 4)]), TQ, -ms, &(Rm[0])); ST(&(Rp[WS(rs, 4)]), TP, ms, &(Rp[0])); ST(&(Rp[WS(rs, 3)]), T13, ms, &(Rp[WS(rs, 1)])); ST(&(Rm[WS(rs, 3)]), T14, -ms, &(Rm[WS(rs, 1)])); ST(&(Rm[WS(rs, 1)]), TI, -ms, &(Rm[WS(rs, 1)])); ST(&(Rp[WS(rs, 1)]), TH, ms, &(Rp[WS(rs, 1)])); } } } VLEAVE(); } static const tw_instr twinstr[] = { VTW(1, 1), VTW(1, 2), VTW(1, 3), VTW(1, 4), VTW(1, 5), VTW(1, 6), VTW(1, 7), VTW(1, 8), VTW(1, 9), {TW_NEXT, VL, 0} }; static const hc2c_desc desc = { 10, XSIMD_STRING("hc2cbdftv_10"), twinstr, &GENUS, {33, 22, 28, 0} }; void XSIMD(codelet_hc2cbdftv_10) (planner *p) { X(khc2c_register) (p, hc2cbdftv_10, &desc, HC2C_VIA_DFT); } #else /* HAVE_FMA */ /* Generated by: ../../../genfft/gen_hc2cdft_c.native -simd -compact -variables 4 -pipeline-latency 8 -trivial-stores -variables 32 -no-generate-bytw -n 10 -dif -sign 1 -name hc2cbdftv_10 -include hc2cbv.h */ /* * This function contains 61 FP additions, 30 FP multiplications, * (or, 55 additions, 24 multiplications, 6 fused multiply/add), * 81 stack variables, 4 constants, and 20 memory accesses */ #include "hc2cbv.h" static void hc2cbdftv_10(R *Rp, R *Ip, R *Rm, R *Im, const R *W, stride rs, INT mb, INT me, INT ms) { DVK(KP250000000, +0.250000000000000000000000000000000000000000000); DVK(KP951056516, +0.951056516295153572116439333379382143405698634); DVK(KP587785252, +0.587785252292473129168705954639072768597652438); DVK(KP559016994, +0.559016994374947424102293417182819058860154590); { INT m; for (m = mb, W = W + ((mb - 1) * ((TWVL / VL) * 18)); m < me; m = m + VL, Rp = Rp + (VL * ms), Ip = Ip + (VL * ms), Rm = Rm - (VL * ms), Im = Im - (VL * ms), W = W + (TWVL * 18), MAKE_VOLATILE_STRIDE(40, rs)) { V T5, TE, Ts, Tt, TC, Tz, TH, TJ, To, Tq, T2, T4, T3, T9, Tx; V Tm, TB, Td, Ty, Ti, TA, T6, T8, T7, Tl, Tk, Tj, Tc, Tb, Ta; V Tf, Th, Tg, TF, TG, Te, Tn; T2 = LD(&(Rp[0]), ms, &(Rp[0])); T3 = LD(&(Rm[WS(rs, 4)]), -ms, &(Rm[0])); T4 = VCONJ(T3); T5 = VSUB(T2, T4); TE = VADD(T2, T4); T6 = LD(&(Rp[WS(rs, 2)]), ms, &(Rp[0])); T7 = LD(&(Rm[WS(rs, 2)]), -ms, &(Rm[0])); T8 = VCONJ(T7); T9 = VSUB(T6, T8); Tx = VADD(T6, T8); Tl = LD(&(Rp[WS(rs, 1)]), ms, &(Rp[WS(rs, 1)])); Tj = LD(&(Rm[WS(rs, 3)]), -ms, &(Rm[WS(rs, 1)])); Tk = VCONJ(Tj); Tm = VSUB(Tk, Tl); TB = VADD(Tk, Tl); Tc = LD(&(Rp[WS(rs, 3)]), ms, &(Rp[WS(rs, 1)])); Ta = LD(&(Rm[WS(rs, 1)]), -ms, &(Rm[WS(rs, 1)])); Tb = VCONJ(Ta); Td = VSUB(Tb, Tc); Ty = VADD(Tb, Tc); Tf = LD(&(Rp[WS(rs, 4)]), ms, &(Rp[0])); Tg = LD(&(Rm[0]), -ms, &(Rm[0])); Th = VCONJ(Tg); Ti = VSUB(Tf, Th); TA = VADD(Tf, Th); Ts = VSUB(T9, Td); Tt = VSUB(Ti, Tm); TC = VSUB(TA, TB); Tz = VSUB(Tx, Ty); TF = VADD(Tx, Ty); TG = VADD(TA, TB); TH = VADD(TF, TG); TJ = VMUL(LDK(KP559016994), VSUB(TF, TG)); Te = VADD(T9, Td); Tn = VADD(Ti, Tm); To = VADD(Te, Tn); Tq = VMUL(LDK(KP559016994), VSUB(Te, Tn)); { V T1c, TX, Tv, T1b, TR, T15, TL, T17, TT, T11, TW, Tu, TQ, Tr, TP; V Tp, T1, T1a, TO, T14, TD, T10, TK, TZ, TI, Tw, T16, TS, TY, TM; V TU, T1e, TN, T1d, T19, T13, TV, T18, T12; T1c = VADD(TE, TH); TW = LDW(&(W[TWVL * 8])); TX = VZMULI(TW, VADD(T5, To)); Tu = VBYI(VFNMS(LDK(KP951056516), Tt, VMUL(LDK(KP587785252), Ts))); TQ = VBYI(VFMA(LDK(KP951056516), Ts, VMUL(LDK(KP587785252), Tt))); Tp = VFNMS(LDK(KP250000000), To, T5); Tr = VSUB(Tp, Tq); TP = VADD(Tq, Tp); T1 = LDW(&(W[TWVL * 4])); Tv = VZMULI(T1, VSUB(Tr, Tu)); T1a = LDW(&(W[0])); T1b = VZMULI(T1a, VADD(TQ, TP)); TO = LDW(&(W[TWVL * 16])); TR = VZMULI(TO, VSUB(TP, TQ)); T14 = LDW(&(W[TWVL * 12])); T15 = VZMULI(T14, VADD(Tu, Tr)); TD = VBYI(VFNMS(LDK(KP951056516), TC, VMUL(LDK(KP587785252), Tz))); T10 = VBYI(VFMA(LDK(KP951056516), Tz, VMUL(LDK(KP587785252), TC))); TI = VFNMS(LDK(KP250000000), TH, TE); TK = VSUB(TI, TJ); TZ = VADD(TJ, TI); Tw = LDW(&(W[TWVL * 2])); TL = VZMUL(Tw, VADD(TD, TK)); T16 = LDW(&(W[TWVL * 10])); T17 = VZMUL(T16, VADD(T10, TZ)); TS = LDW(&(W[TWVL * 14])); TT = VZMUL(TS, VSUB(TK, TD)); TY = LDW(&(W[TWVL * 6])); T11 = VZMUL(TY, VSUB(TZ, T10)); TM = VADD(Tv, TL); ST(&(Rp[WS(rs, 1)]), TM, ms, &(Rp[WS(rs, 1)])); TU = VADD(TR, TT); ST(&(Rp[WS(rs, 4)]), TU, ms, &(Rp[0])); T1e = VCONJ(VSUB(T1c, T1b)); ST(&(Rm[0]), T1e, -ms, &(Rm[0])); TN = VCONJ(VSUB(TL, Tv)); ST(&(Rm[WS(rs, 1)]), TN, -ms, &(Rm[WS(rs, 1)])); T1d = VADD(T1b, T1c); ST(&(Rp[0]), T1d, ms, &(Rp[0])); T19 = VCONJ(VSUB(T17, T15)); ST(&(Rm[WS(rs, 3)]), T19, -ms, &(Rm[WS(rs, 1)])); T13 = VCONJ(VSUB(T11, TX)); ST(&(Rm[WS(rs, 2)]), T13, -ms, &(Rm[0])); TV = VCONJ(VSUB(TT, TR)); ST(&(Rm[WS(rs, 4)]), TV, -ms, &(Rm[0])); T18 = VADD(T15, T17); ST(&(Rp[WS(rs, 3)]), T18, ms, &(Rp[WS(rs, 1)])); T12 = VADD(TX, T11); ST(&(Rp[WS(rs, 2)]), T12, ms, &(Rp[0])); } } } VLEAVE(); } static const tw_instr twinstr[] = { VTW(1, 1), VTW(1, 2), VTW(1, 3), VTW(1, 4), VTW(1, 5), VTW(1, 6), VTW(1, 7), VTW(1, 8), VTW(1, 9), {TW_NEXT, VL, 0} }; static const hc2c_desc desc = { 10, XSIMD_STRING("hc2cbdftv_10"), twinstr, &GENUS, {55, 24, 6, 0} }; void XSIMD(codelet_hc2cbdftv_10) (planner *p) { X(khc2c_register) (p, hc2cbdftv_10, &desc, HC2C_VIA_DFT); } #endif /* HAVE_FMA */ fftw-3.3.3/rdft/simd/common/hc2cbdftv_12.c0000644000175000017500000002744112054411066015114 00000000000000/* * Copyright (c) 2003, 2007-11 Matteo Frigo * Copyright (c) 2003, 2007-11 Massachusetts Institute of Technology * * This program is free software; you can redistribute it and/or modify * it under the terms of the GNU General Public License as published by * the Free Software Foundation; either version 2 of the License, or * (at your option) any later version. * * This program is distributed in the hope that it will be useful, * but WITHOUT ANY WARRANTY; without even the implied warranty of * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the * GNU General Public License for more details. * * You should have received a copy of the GNU General Public License * along with this program; if not, write to the Free Software * Foundation, Inc., 51 Franklin Street, Fifth Floor, Boston, MA 02110-1301 USA * */ /* This file was automatically generated --- DO NOT EDIT */ /* Generated on Sun Nov 25 07:42:30 EST 2012 */ #include "codelet-rdft.h" #ifdef HAVE_FMA /* Generated by: ../../../genfft/gen_hc2cdft_c.native -fma -reorder-insns -schedule-for-pipeline -simd -compact -variables 4 -pipeline-latency 8 -trivial-stores -variables 32 -no-generate-bytw -n 12 -dif -sign 1 -name hc2cbdftv_12 -include hc2cbv.h */ /* * This function contains 71 FP additions, 51 FP multiplications, * (or, 45 additions, 25 multiplications, 26 fused multiply/add), * 88 stack variables, 2 constants, and 24 memory accesses */ #include "hc2cbv.h" static void hc2cbdftv_12(R *Rp, R *Ip, R *Rm, R *Im, const R *W, stride rs, INT mb, INT me, INT ms) { DVK(KP866025403, +0.866025403784438646763723170752936183471402627); DVK(KP500000000, +0.500000000000000000000000000000000000000000000); { INT m; for (m = mb, W = W + ((mb - 1) * ((TWVL / VL) * 22)); m < me; m = m + VL, Rp = Rp + (VL * ms), Ip = Ip + (VL * ms), Rm = Rm - (VL * ms), Im = Im - (VL * ms), W = W + (TWVL * 22), MAKE_VOLATILE_STRIDE(48, rs)) { V Tz, TT, T1, T1j, TN, TF, TP, TL, Tx, T15, TJ, T1b, T1g, T1l, T18; V T12, TO, TC, TK, Tl, T16, TQ, TU, TG, T1c, TM, T1k, Ty, T19, T1a; V T13, T14, T1h, T1i, TS, TR, T1m, T1n, TI, TH; { V T2, Tm, T7, Tp, T8, Tq, T9, Tu, T5, Tr, Tg, Tn, Tj, Ta, T3; V T4, Te, Tf, Th, Ti, TV, T6, TW, Tk, TD, Tt, TB, T11, T1f, Tw; V TE, TX, Tc, Ts, T10, TZ, To, Tb, Tv, T17, T1d, T1e, TY, TA, Td; T2 = LD(&(Rp[0]), ms, &(Rp[0])); Tm = LD(&(Rp[WS(rs, 3)]), ms, &(Rp[WS(rs, 1)])); T7 = LD(&(Rm[WS(rs, 5)]), -ms, &(Rm[WS(rs, 1)])); Tp = LD(&(Rm[WS(rs, 2)]), -ms, &(Rm[0])); T3 = LD(&(Rp[WS(rs, 4)]), ms, &(Rp[0])); T4 = LD(&(Rm[WS(rs, 3)]), -ms, &(Rm[WS(rs, 1)])); Te = LD(&(Rp[WS(rs, 1)]), ms, &(Rp[WS(rs, 1)])); Tf = LD(&(Rp[WS(rs, 5)]), ms, &(Rp[WS(rs, 1)])); Th = LD(&(Rm[0]), -ms, &(Rm[0])); Ti = LD(&(Rm[WS(rs, 4)]), -ms, &(Rm[0])); T8 = VCONJ(T7); Tq = VCONJ(Tp); T9 = LD(&(Rp[WS(rs, 2)]), ms, &(Rp[0])); Tu = VFNMSCONJ(T4, T3); T5 = VFMACONJ(T4, T3); Tr = VADD(Te, Tf); Tg = VSUB(Te, Tf); Tn = VADD(Ti, Th); Tj = VSUB(Th, Ti); Ta = LD(&(Rm[WS(rs, 1)]), -ms, &(Rm[WS(rs, 1)])); TV = LDW(&(W[TWVL * 4])); Tz = LDW(&(W[TWVL * 18])); T6 = VFNMS(LDK(KP500000000), T5, T2); TW = VADD(T2, T5); Ts = VFNMS(LDK(KP500000000), Tr, Tq); T10 = VFMACONJ(Tp, Tr); TZ = VFMACONJ(Tn, Tm); To = VFNMS(LDK(KP500000000), VCONJ(Tn), Tm); Tk = VFMACONJ(Tj, Tg); TD = VFNMSCONJ(Tj, Tg); Tb = VFMACONJ(Ta, T9); Tv = VFMSCONJ(Ta, T9); TT = LDW(&(W[TWVL * 2])); T1 = LDW(&(W[TWVL * 20])); Tt = VSUB(To, Ts); TB = VADD(To, Ts); T11 = VSUB(TZ, T10); T1f = VADD(TZ, T10); Tw = VSUB(Tu, Tv); TE = VADD(Tu, Tv); TX = VFMACONJ(T7, Tb); Tc = VFNMS(LDK(KP500000000), Tb, T8); T1j = LDW(&(W[0])); T17 = LDW(&(W[TWVL * 16])); T1d = LDW(&(W[TWVL * 10])); TN = LDW(&(W[TWVL * 6])); TF = VMUL(LDK(KP866025403), VSUB(TD, TE)); TP = VMUL(LDK(KP866025403), VADD(TE, TD)); TL = VFNMS(LDK(KP866025403), Tw, Tt); Tx = VFMA(LDK(KP866025403), Tw, Tt); T1e = VADD(TW, TX); TY = VSUB(TW, TX); TA = VADD(T6, Tc); Td = VSUB(T6, Tc); T15 = LDW(&(W[TWVL * 14])); TJ = LDW(&(W[TWVL * 8])); T1b = LDW(&(W[TWVL * 12])); T1g = VZMUL(T1d, VSUB(T1e, T1f)); T1l = VADD(T1e, T1f); T18 = VZMULI(T17, VFMAI(T11, TY)); T12 = VZMULI(TV, VFNMSI(T11, TY)); TO = VADD(TA, TB); TC = VSUB(TA, TB); TK = VFNMS(LDK(KP866025403), Tk, Td); Tl = VFMA(LDK(KP866025403), Tk, Td); } T16 = VZMUL(T15, VFNMSI(TP, TO)); TQ = VZMUL(TN, VFMAI(TP, TO)); TU = VZMUL(TT, VFMAI(TF, TC)); TG = VZMUL(Tz, VFNMSI(TF, TC)); T1c = VZMULI(T1b, VFNMSI(TL, TK)); TM = VZMULI(TJ, VFMAI(TL, TK)); T1k = VZMULI(T1j, VFMAI(Tx, Tl)); Ty = VZMULI(T1, VFNMSI(Tx, Tl)); T19 = VCONJ(VSUB(T16, T18)); T1a = VADD(T16, T18); T13 = VCONJ(VSUB(TU, T12)); T14 = VADD(TU, T12); T1h = VADD(T1c, T1g); T1i = VCONJ(VSUB(T1g, T1c)); TS = VCONJ(VSUB(TQ, TM)); TR = VADD(TM, TQ); T1m = VADD(T1k, T1l); T1n = VCONJ(VSUB(T1l, T1k)); TI = VCONJ(VSUB(TG, Ty)); TH = VADD(Ty, TG); ST(&(Rm[WS(rs, 4)]), T19, -ms, &(Rm[0])); ST(&(Rp[WS(rs, 4)]), T1a, ms, &(Rp[0])); ST(&(Rm[WS(rs, 1)]), T13, -ms, &(Rm[WS(rs, 1)])); ST(&(Rp[WS(rs, 1)]), T14, ms, &(Rp[WS(rs, 1)])); ST(&(Rp[WS(rs, 3)]), T1h, ms, &(Rp[WS(rs, 1)])); ST(&(Rm[WS(rs, 3)]), T1i, -ms, &(Rm[WS(rs, 1)])); ST(&(Rm[WS(rs, 2)]), TS, -ms, &(Rm[0])); ST(&(Rp[WS(rs, 2)]), TR, ms, &(Rp[0])); ST(&(Rp[0]), T1m, ms, &(Rp[0])); ST(&(Rm[0]), T1n, -ms, &(Rm[0])); ST(&(Rm[WS(rs, 5)]), TI, -ms, &(Rm[WS(rs, 1)])); ST(&(Rp[WS(rs, 5)]), TH, ms, &(Rp[WS(rs, 1)])); } } VLEAVE(); } static const tw_instr twinstr[] = { VTW(1, 1), VTW(1, 2), VTW(1, 3), VTW(1, 4), VTW(1, 5), VTW(1, 6), VTW(1, 7), VTW(1, 8), VTW(1, 9), VTW(1, 10), VTW(1, 11), {TW_NEXT, VL, 0} }; static const hc2c_desc desc = { 12, XSIMD_STRING("hc2cbdftv_12"), twinstr, &GENUS, {45, 25, 26, 0} }; void XSIMD(codelet_hc2cbdftv_12) (planner *p) { X(khc2c_register) (p, hc2cbdftv_12, &desc, HC2C_VIA_DFT); } #else /* HAVE_FMA */ /* Generated by: ../../../genfft/gen_hc2cdft_c.native -simd -compact -variables 4 -pipeline-latency 8 -trivial-stores -variables 32 -no-generate-bytw -n 12 -dif -sign 1 -name hc2cbdftv_12 -include hc2cbv.h */ /* * This function contains 71 FP additions, 30 FP multiplications, * (or, 67 additions, 26 multiplications, 4 fused multiply/add), * 90 stack variables, 2 constants, and 24 memory accesses */ #include "hc2cbv.h" static void hc2cbdftv_12(R *Rp, R *Ip, R *Rm, R *Im, const R *W, stride rs, INT mb, INT me, INT ms) { DVK(KP866025403, +0.866025403784438646763723170752936183471402627); DVK(KP500000000, +0.500000000000000000000000000000000000000000000); { INT m; for (m = mb, W = W + ((mb - 1) * ((TWVL / VL) * 22)); m < me; m = m + VL, Rp = Rp + (VL * ms), Ip = Ip + (VL * ms), Rm = Rm - (VL * ms), Im = Im - (VL * ms), W = W + (TWVL * 22), MAKE_VOLATILE_STRIDE(48, rs)) { V TY, TZ, Tf, TC, Tq, TG, Tm, TF, Ty, TD, T13, T1h, T2, T9, T3; V T5, T6, Tc, Tb, Td, T8, T4, Ta, T7, Te, To, Tp, Tr, Tv, Ti; V Ts, Tl, Tw, Tu, Tg, Th, Tj, Tk, Tt, Tx, T11, T12; T2 = LD(&(Rp[0]), ms, &(Rp[0])); T8 = LD(&(Rm[WS(rs, 5)]), -ms, &(Rm[WS(rs, 1)])); T9 = VCONJ(T8); T3 = LD(&(Rp[WS(rs, 4)]), ms, &(Rp[0])); T4 = LD(&(Rm[WS(rs, 3)]), -ms, &(Rm[WS(rs, 1)])); T5 = VCONJ(T4); T6 = VADD(T3, T5); Tc = LD(&(Rp[WS(rs, 2)]), ms, &(Rp[0])); Ta = LD(&(Rm[WS(rs, 1)]), -ms, &(Rm[WS(rs, 1)])); Tb = VCONJ(Ta); Td = VADD(Tb, Tc); TY = VADD(T2, T6); TZ = VADD(T9, Td); T7 = VFNMS(LDK(KP500000000), T6, T2); Te = VFNMS(LDK(KP500000000), Td, T9); Tf = VSUB(T7, Te); TC = VADD(T7, Te); To = VSUB(T3, T5); Tp = VSUB(Tb, Tc); Tq = VMUL(LDK(KP866025403), VSUB(To, Tp)); TG = VADD(To, Tp); Tr = LD(&(Rp[WS(rs, 3)]), ms, &(Rp[WS(rs, 1)])); Tu = LD(&(Rm[WS(rs, 2)]), -ms, &(Rm[0])); Tv = VCONJ(Tu); Tg = LD(&(Rm[WS(rs, 4)]), -ms, &(Rm[0])); Th = LD(&(Rm[0]), -ms, &(Rm[0])); Ti = VCONJ(VSUB(Tg, Th)); Ts = VCONJ(VADD(Tg, Th)); Tj = LD(&(Rp[WS(rs, 1)]), ms, &(Rp[WS(rs, 1)])); Tk = LD(&(Rp[WS(rs, 5)]), ms, &(Rp[WS(rs, 1)])); Tl = VSUB(Tj, Tk); Tw = VADD(Tj, Tk); Tm = VMUL(LDK(KP866025403), VSUB(Ti, Tl)); TF = VADD(Ti, Tl); Tt = VFNMS(LDK(KP500000000), Ts, Tr); Tx = VFNMS(LDK(KP500000000), Tw, Tv); Ty = VSUB(Tt, Tx); TD = VADD(Tt, Tx); T11 = VADD(Tr, Ts); T12 = VADD(Tv, Tw); T13 = VBYI(VSUB(T11, T12)); T1h = VADD(T11, T12); { V T1n, T1i, T14, T1a, TA, T1m, TS, T18, TO, T1e, TI, TW, T1g, T1f, T10; V TX, T19, Tn, Tz, T1, T1l, TQ, TR, TP, T17, TM, TN, TL, T1d, TE; V TH, TB, TV, TJ, T1p, T1k, TT, T1o, TK, TU, T1j, T1b, T16, T1c, T15; T1g = VADD(TY, TZ); T1n = VADD(T1g, T1h); T1f = LDW(&(W[TWVL * 10])); T1i = VZMUL(T1f, VSUB(T1g, T1h)); T10 = VSUB(TY, TZ); TX = LDW(&(W[TWVL * 4])); T14 = VZMULI(TX, VSUB(T10, T13)); T19 = LDW(&(W[TWVL * 16])); T1a = VZMULI(T19, VADD(T10, T13)); Tn = VSUB(Tf, Tm); Tz = VBYI(VADD(Tq, Ty)); T1 = LDW(&(W[TWVL * 20])); TA = VZMULI(T1, VSUB(Tn, Tz)); T1l = LDW(&(W[0])); T1m = VZMULI(T1l, VADD(Tn, Tz)); TQ = VBYI(VMUL(LDK(KP866025403), VADD(TG, TF))); TR = VADD(TC, TD); TP = LDW(&(W[TWVL * 6])); TS = VZMUL(TP, VADD(TQ, TR)); T17 = LDW(&(W[TWVL * 14])); T18 = VZMUL(T17, VSUB(TR, TQ)); TM = VADD(Tf, Tm); TN = VBYI(VSUB(Ty, Tq)); TL = LDW(&(W[TWVL * 8])); TO = VZMULI(TL, VADD(TM, TN)); T1d = LDW(&(W[TWVL * 12])); T1e = VZMULI(T1d, VSUB(TM, TN)); TE = VSUB(TC, TD); TH = VBYI(VMUL(LDK(KP866025403), VSUB(TF, TG))); TB = LDW(&(W[TWVL * 18])); TI = VZMUL(TB, VSUB(TE, TH)); TV = LDW(&(W[TWVL * 2])); TW = VZMUL(TV, VADD(TH, TE)); TJ = VADD(TA, TI); ST(&(Rp[WS(rs, 5)]), TJ, ms, &(Rp[WS(rs, 1)])); T1p = VCONJ(VSUB(T1n, T1m)); ST(&(Rm[0]), T1p, -ms, &(Rm[0])); T1k = VCONJ(VSUB(T1i, T1e)); ST(&(Rm[WS(rs, 3)]), T1k, -ms, &(Rm[WS(rs, 1)])); TT = VADD(TO, TS); ST(&(Rp[WS(rs, 2)]), TT, ms, &(Rp[0])); T1o = VADD(T1m, T1n); ST(&(Rp[0]), T1o, ms, &(Rp[0])); TK = VCONJ(VSUB(TI, TA)); ST(&(Rm[WS(rs, 5)]), TK, -ms, &(Rm[WS(rs, 1)])); TU = VCONJ(VSUB(TS, TO)); ST(&(Rm[WS(rs, 2)]), TU, -ms, &(Rm[0])); T1j = VADD(T1e, T1i); ST(&(Rp[WS(rs, 3)]), T1j, ms, &(Rp[WS(rs, 1)])); T1b = VCONJ(VSUB(T18, T1a)); ST(&(Rm[WS(rs, 4)]), T1b, -ms, &(Rm[0])); T16 = VADD(TW, T14); ST(&(Rp[WS(rs, 1)]), T16, ms, &(Rp[WS(rs, 1)])); T1c = VADD(T18, T1a); ST(&(Rp[WS(rs, 4)]), T1c, ms, &(Rp[0])); T15 = VCONJ(VSUB(TW, T14)); ST(&(Rm[WS(rs, 1)]), T15, -ms, &(Rm[WS(rs, 1)])); } } } VLEAVE(); } static const tw_instr twinstr[] = { VTW(1, 1), VTW(1, 2), VTW(1, 3), VTW(1, 4), VTW(1, 5), VTW(1, 6), VTW(1, 7), VTW(1, 8), VTW(1, 9), VTW(1, 10), VTW(1, 11), {TW_NEXT, VL, 0} }; static const hc2c_desc desc = { 12, XSIMD_STRING("hc2cbdftv_12"), twinstr, &GENUS, {67, 26, 4, 0} }; void XSIMD(codelet_hc2cbdftv_12) (planner *p) { X(khc2c_register) (p, hc2cbdftv_12, &desc, HC2C_VIA_DFT); } #endif /* HAVE_FMA */ fftw-3.3.3/rdft/simd/common/hc2cbdftv_16.c0000644000175000017500000003653012054411067015120 00000000000000/* * Copyright (c) 2003, 2007-11 Matteo Frigo * Copyright (c) 2003, 2007-11 Massachusetts Institute of Technology * * This program is free software; you can redistribute it and/or modify * it under the terms of the GNU General Public License as published by * the Free Software Foundation; either version 2 of the License, or * (at your option) any later version. * * This program is distributed in the hope that it will be useful, * but WITHOUT ANY WARRANTY; without even the implied warranty of * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the * GNU General Public License for more details. * * You should have received a copy of the GNU General Public License * along with this program; if not, write to the Free Software * Foundation, Inc., 51 Franklin Street, Fifth Floor, Boston, MA 02110-1301 USA * */ /* This file was automatically generated --- DO NOT EDIT */ /* Generated on Sun Nov 25 07:42:30 EST 2012 */ #include "codelet-rdft.h" #ifdef HAVE_FMA /* Generated by: ../../../genfft/gen_hc2cdft_c.native -fma -reorder-insns -schedule-for-pipeline -simd -compact -variables 4 -pipeline-latency 8 -trivial-stores -variables 32 -no-generate-bytw -n 16 -dif -sign 1 -name hc2cbdftv_16 -include hc2cbv.h */ /* * This function contains 103 FP additions, 80 FP multiplications, * (or, 53 additions, 30 multiplications, 50 fused multiply/add), * 123 stack variables, 3 constants, and 32 memory accesses */ #include "hc2cbv.h" static void hc2cbdftv_16(R *Rp, R *Ip, R *Rm, R *Im, const R *W, stride rs, INT mb, INT me, INT ms) { DVK(KP923879532, +0.923879532511286756128183189396788286822416626); DVK(KP414213562, +0.414213562373095048801688724209698078569671875); DVK(KP707106781, +0.707106781186547524400844362104849039284835938); { INT m; for (m = mb, W = W + ((mb - 1) * ((TWVL / VL) * 30)); m < me; m = m + VL, Rp = Rp + (VL * ms), Ip = Ip + (VL * ms), Rm = Rm - (VL * ms), Im = Im - (VL * ms), W = W + (TWVL * 30), MAKE_VOLATILE_STRIDE(64, rs)) { V T1D, T1F, TV, TW, T17, T18, T1B, T1A, T1H, T1G; { V T8, Tv, Tb, TF, Tl, TJ, TP, T1w, TE, T1t, T10, T1p, TG, Te, Tg; V Th, T2, T3, Ts, Tt, T5, T6, Tp, Tq, T9, TA, T4, TC, Tu, TN; V T7, TB, Tr, Ta, Tj, Tk, Tc, Td, TY, TD, TO, TZ, T1Q, T19, T1I; V T1d, Tf, T11, TH, TQ, Ti, TI, T1k, T1K, T1S, T1r, T14, T16, TU, Ty; V T1z, TX, T1o, T1, TK, TR, Tm, T12, T1C, Tz, T15; T2 = LD(&(Rp[0]), ms, &(Rp[0])); T3 = LD(&(Rm[WS(rs, 7)]), -ms, &(Rm[WS(rs, 1)])); Ts = LD(&(Rp[WS(rs, 6)]), ms, &(Rp[0])); Tt = LD(&(Rm[WS(rs, 1)]), -ms, &(Rm[WS(rs, 1)])); T5 = LD(&(Rp[WS(rs, 4)]), ms, &(Rp[0])); T6 = LD(&(Rm[WS(rs, 3)]), -ms, &(Rm[WS(rs, 1)])); Tp = LD(&(Rp[WS(rs, 2)]), ms, &(Rp[0])); Tq = LD(&(Rm[WS(rs, 5)]), -ms, &(Rm[WS(rs, 1)])); T9 = LD(&(Rp[WS(rs, 1)]), ms, &(Rp[WS(rs, 1)])); TA = VFNMSCONJ(T3, T2); T4 = VFMACONJ(T3, T2); TC = VFMSCONJ(Tt, Ts); Tu = VFMACONJ(Tt, Ts); TN = VFNMSCONJ(T6, T5); T7 = VFMACONJ(T6, T5); TB = VFNMSCONJ(Tq, Tp); Tr = VFMACONJ(Tq, Tp); Ta = LD(&(Rm[WS(rs, 6)]), -ms, &(Rm[0])); Tj = LD(&(Rp[WS(rs, 3)]), ms, &(Rp[WS(rs, 1)])); Tk = LD(&(Rm[WS(rs, 4)]), -ms, &(Rm[0])); Tc = LD(&(Rp[WS(rs, 5)]), ms, &(Rp[WS(rs, 1)])); Td = LD(&(Rm[WS(rs, 2)]), -ms, &(Rm[0])); T8 = VSUB(T4, T7); TY = VADD(T4, T7); TD = VADD(TB, TC); TO = VSUB(TB, TC); Tv = VSUB(Tr, Tu); TZ = VADD(Tr, Tu); Tb = VFMACONJ(Ta, T9); TF = VFNMSCONJ(Ta, T9); Tl = VFMACONJ(Tk, Tj); TJ = VFNMSCONJ(Tk, Tj); TP = VFMA(LDK(KP707106781), TO, TN); T1w = VFNMS(LDK(KP707106781), TO, TN); TE = VFMA(LDK(KP707106781), TD, TA); T1t = VFNMS(LDK(KP707106781), TD, TA); T10 = VADD(TY, TZ); T1p = VSUB(TY, TZ); TG = VFNMSCONJ(Td, Tc); Te = VFMACONJ(Td, Tc); Tg = LD(&(Rp[WS(rs, 7)]), ms, &(Rp[WS(rs, 1)])); Th = LD(&(Rm[0]), -ms, &(Rm[0])); T1Q = LDW(&(W[TWVL * 22])); T19 = LDW(&(W[TWVL * 26])); T1I = LDW(&(W[TWVL * 2])); T1d = LDW(&(W[TWVL * 28])); Tf = VSUB(Tb, Te); T11 = VADD(Tb, Te); TH = VFNMS(LDK(KP414213562), TG, TF); TQ = VFMA(LDK(KP414213562), TF, TG); Ti = VFMACONJ(Th, Tg); TI = VFMSCONJ(Th, Tg); T1k = LDW(&(W[0])); T1K = LDW(&(W[TWVL * 4])); T1S = LDW(&(W[TWVL * 24])); TX = LDW(&(W[TWVL * 14])); T1o = LDW(&(W[TWVL * 6])); T1 = LDW(&(W[TWVL * 10])); TK = VFMA(LDK(KP414213562), TJ, TI); TR = VFNMS(LDK(KP414213562), TI, TJ); Tm = VSUB(Ti, Tl); T12 = VADD(Ti, Tl); T1C = LDW(&(W[TWVL * 18])); Tz = LDW(&(W[TWVL * 12])); T15 = LDW(&(W[TWVL * 16])); { V T1v, T1y, T1N, T1g, T1J, T1c, T1U, T1V, T1m, T1n, T1s, TS, T1u, TL, T1x; V T13, T1q, Tn, Tw, T1L, T1f, TT, T1M, T1e, TM, T1R, T1j, T1b, Tx, T1a; V To, T1T, T1l, T1E, T1O, T1P, T1h, T1i; T1s = LDW(&(W[TWVL * 8])); TS = VADD(TQ, TR); T1u = VSUB(TQ, TR); TL = VADD(TH, TK); T1x = VSUB(TH, TK); T13 = VADD(T11, T12); T1q = VSUB(T11, T12); Tn = VADD(Tf, Tm); Tw = VSUB(Tf, Tm); T1L = VFMA(LDK(KP923879532), T1u, T1t); T1v = VFNMS(LDK(KP923879532), T1u, T1t); T1f = VFMA(LDK(KP923879532), TS, TP); TT = VFNMS(LDK(KP923879532), TS, TP); T1M = VFNMS(LDK(KP923879532), T1x, T1w); T1y = VFMA(LDK(KP923879532), T1x, T1w); T1e = VFMA(LDK(KP923879532), TL, TE); TM = VFNMS(LDK(KP923879532), TL, TE); T1r = VZMUL(T1o, VFMAI(T1q, T1p)); T1R = VZMUL(T1Q, VFNMSI(T1q, T1p)); T14 = VZMUL(TX, VSUB(T10, T13)); T1j = VADD(T10, T13); T1b = VFMA(LDK(KP707106781), Tw, Tv); Tx = VFNMS(LDK(KP707106781), Tw, Tv); T1a = VFMA(LDK(KP707106781), Tn, T8); To = VFNMS(LDK(KP707106781), Tn, T8); T1T = VZMULI(T1S, VFMAI(T1M, T1L)); T1N = VZMULI(T1K, VFNMSI(T1M, T1L)); T16 = VZMULI(T15, VFMAI(TT, TM)); TU = VZMULI(Tz, VFNMSI(TT, TM)); T1l = VZMULI(T1k, VFMAI(T1f, T1e)); T1g = VZMULI(T1d, VFNMSI(T1f, T1e)); T1D = VZMUL(T1C, VFMAI(Tx, To)); Ty = VZMUL(T1, VFNMSI(Tx, To)); T1J = VZMUL(T1I, VFMAI(T1b, T1a)); T1c = VZMUL(T19, VFNMSI(T1b, T1a)); T1U = VCONJ(VSUB(T1R, T1T)); T1V = VADD(T1R, T1T); T1m = VCONJ(VSUB(T1j, T1l)); T1n = VADD(T1j, T1l); T1z = VZMULI(T1s, VFMAI(T1y, T1v)); T1E = LDW(&(W[TWVL * 20])); T1O = VCONJ(VSUB(T1J, T1N)); T1P = VADD(T1J, T1N); T1h = VCONJ(VSUB(T1c, T1g)); T1i = VADD(T1c, T1g); ST(&(Rp[WS(rs, 6)]), T1V, ms, &(Rp[0])); ST(&(Rm[WS(rs, 6)]), T1U, -ms, &(Rm[0])); ST(&(Rp[0]), T1n, ms, &(Rp[0])); ST(&(Rm[0]), T1m, -ms, &(Rm[0])); ST(&(Rp[WS(rs, 1)]), T1P, ms, &(Rp[WS(rs, 1)])); ST(&(Rm[WS(rs, 1)]), T1O, -ms, &(Rm[WS(rs, 1)])); ST(&(Rp[WS(rs, 7)]), T1i, ms, &(Rp[WS(rs, 1)])); ST(&(Rm[WS(rs, 7)]), T1h, -ms, &(Rm[WS(rs, 1)])); T1F = VZMULI(T1E, VFNMSI(T1y, T1v)); } TV = VCONJ(VSUB(Ty, TU)); TW = VADD(Ty, TU); T17 = VCONJ(VSUB(T14, T16)); T18 = VADD(T14, T16); T1B = VADD(T1r, T1z); T1A = VCONJ(VSUB(T1r, T1z)); } T1H = VADD(T1D, T1F); T1G = VCONJ(VSUB(T1D, T1F)); ST(&(Rm[WS(rs, 3)]), TV, -ms, &(Rm[WS(rs, 1)])); ST(&(Rp[WS(rs, 3)]), TW, ms, &(Rp[WS(rs, 1)])); ST(&(Rm[WS(rs, 4)]), T17, -ms, &(Rm[0])); ST(&(Rm[WS(rs, 2)]), T1A, -ms, &(Rm[0])); ST(&(Rp[WS(rs, 2)]), T1B, ms, &(Rp[0])); ST(&(Rp[WS(rs, 4)]), T18, ms, &(Rp[0])); ST(&(Rp[WS(rs, 5)]), T1H, ms, &(Rp[WS(rs, 1)])); ST(&(Rm[WS(rs, 5)]), T1G, -ms, &(Rm[WS(rs, 1)])); } } VLEAVE(); } static const tw_instr twinstr[] = { VTW(1, 1), VTW(1, 2), VTW(1, 3), VTW(1, 4), VTW(1, 5), VTW(1, 6), VTW(1, 7), VTW(1, 8), VTW(1, 9), VTW(1, 10), VTW(1, 11), VTW(1, 12), VTW(1, 13), VTW(1, 14), VTW(1, 15), {TW_NEXT, VL, 0} }; static const hc2c_desc desc = { 16, XSIMD_STRING("hc2cbdftv_16"), twinstr, &GENUS, {53, 30, 50, 0} }; void XSIMD(codelet_hc2cbdftv_16) (planner *p) { X(khc2c_register) (p, hc2cbdftv_16, &desc, HC2C_VIA_DFT); } #else /* HAVE_FMA */ /* Generated by: ../../../genfft/gen_hc2cdft_c.native -simd -compact -variables 4 -pipeline-latency 8 -trivial-stores -variables 32 -no-generate-bytw -n 16 -dif -sign 1 -name hc2cbdftv_16 -include hc2cbv.h */ /* * This function contains 103 FP additions, 42 FP multiplications, * (or, 99 additions, 38 multiplications, 4 fused multiply/add), * 83 stack variables, 3 constants, and 32 memory accesses */ #include "hc2cbv.h" static void hc2cbdftv_16(R *Rp, R *Ip, R *Rm, R *Im, const R *W, stride rs, INT mb, INT me, INT ms) { DVK(KP382683432, +0.382683432365089771728459984030398866761344562); DVK(KP923879532, +0.923879532511286756128183189396788286822416626); DVK(KP707106781, +0.707106781186547524400844362104849039284835938); { INT m; for (m = mb, W = W + ((mb - 1) * ((TWVL / VL) * 30)); m < me; m = m + VL, Rp = Rp + (VL * ms), Ip = Ip + (VL * ms), Rm = Rm - (VL * ms), Im = Im - (VL * ms), W = W + (TWVL * 30), MAKE_VOLATILE_STRIDE(64, rs)) { V Tf, T16, TZ, T1C, TI, T1a, TV, T1D, T1F, T1G, Ty, T19, TC, T17, TS; V T10; { V T2, TD, T4, TF, Tc, Tb, Td, T6, T8, T9, T3, TE, Ta, T7, T5; V Te, TX, TY, TG, TH, TT, TU, Tj, TM, Tw, TQ, Tn, TN, Ts, TP; V Tg, Ti, Th, Tt, Tv, Tu, Tk, Tm, Tl, Tr, Tq, Tp, To, Tx, TA; V TB, TO, TR; T2 = LD(&(Rp[0]), ms, &(Rp[0])); TD = LD(&(Rp[WS(rs, 4)]), ms, &(Rp[0])); T3 = LD(&(Rm[WS(rs, 7)]), -ms, &(Rm[WS(rs, 1)])); T4 = VCONJ(T3); TE = LD(&(Rm[WS(rs, 3)]), -ms, &(Rm[WS(rs, 1)])); TF = VCONJ(TE); Tc = LD(&(Rp[WS(rs, 6)]), ms, &(Rp[0])); Ta = LD(&(Rm[WS(rs, 1)]), -ms, &(Rm[WS(rs, 1)])); Tb = VCONJ(Ta); Td = VSUB(Tb, Tc); T6 = LD(&(Rp[WS(rs, 2)]), ms, &(Rp[0])); T7 = LD(&(Rm[WS(rs, 5)]), -ms, &(Rm[WS(rs, 1)])); T8 = VCONJ(T7); T9 = VSUB(T6, T8); T5 = VSUB(T2, T4); Te = VMUL(LDK(KP707106781), VADD(T9, Td)); Tf = VADD(T5, Te); T16 = VSUB(T5, Te); TX = VADD(T2, T4); TY = VADD(TD, TF); TZ = VSUB(TX, TY); T1C = VADD(TX, TY); TG = VSUB(TD, TF); TH = VMUL(LDK(KP707106781), VSUB(T9, Td)); TI = VADD(TG, TH); T1a = VSUB(TH, TG); TT = VADD(T6, T8); TU = VADD(Tb, Tc); TV = VSUB(TT, TU); T1D = VADD(TT, TU); Tg = LD(&(Rp[WS(rs, 1)]), ms, &(Rp[WS(rs, 1)])); Th = LD(&(Rm[WS(rs, 6)]), -ms, &(Rm[0])); Ti = VCONJ(Th); Tj = VSUB(Tg, Ti); TM = VADD(Tg, Ti); Tt = LD(&(Rp[WS(rs, 3)]), ms, &(Rp[WS(rs, 1)])); Tu = LD(&(Rm[WS(rs, 4)]), -ms, &(Rm[0])); Tv = VCONJ(Tu); Tw = VSUB(Tt, Tv); TQ = VADD(Tt, Tv); Tk = LD(&(Rp[WS(rs, 5)]), ms, &(Rp[WS(rs, 1)])); Tl = LD(&(Rm[WS(rs, 2)]), -ms, &(Rm[0])); Tm = VCONJ(Tl); Tn = VSUB(Tk, Tm); TN = VADD(Tk, Tm); Tr = LD(&(Rp[WS(rs, 7)]), ms, &(Rp[WS(rs, 1)])); Tp = LD(&(Rm[0]), -ms, &(Rm[0])); Tq = VCONJ(Tp); Ts = VSUB(Tq, Tr); TP = VADD(Tq, Tr); T1F = VADD(TM, TN); T1G = VADD(TP, TQ); To = VFNMS(LDK(KP382683432), Tn, VMUL(LDK(KP923879532), Tj)); Tx = VFMA(LDK(KP923879532), Ts, VMUL(LDK(KP382683432), Tw)); Ty = VADD(To, Tx); T19 = VSUB(To, Tx); TA = VFMA(LDK(KP382683432), Tj, VMUL(LDK(KP923879532), Tn)); TB = VFNMS(LDK(KP382683432), Ts, VMUL(LDK(KP923879532), Tw)); TC = VADD(TA, TB); T17 = VSUB(TA, TB); TO = VSUB(TM, TN); TR = VSUB(TP, TQ); TS = VMUL(LDK(KP707106781), VSUB(TO, TR)); T10 = VMUL(LDK(KP707106781), VADD(TO, TR)); } { V T21, T1W, T1u, T20, T1I, T1O, TK, T1S, T12, T1e, T1k, T1A, T1o, T1w, T1c; V T1M, T1U, T1V, T1T, T1s, T1t, T1r, T1Z, T1E, T1H, T1B, T1N, Tz, TJ, T1; V T1R, TW, T11, TL, T1d, T1i, T1j, T1h, T1z, T1m, T1n, T1l, T1v, T18, T1b; V T15, T1L, T13, T1g, T1X, T23, T14, T1f, T1Y, T22, T1p, T1y, T1J, T1Q, T1q; V T1x, T1K, T1P; T1U = VADD(T1C, T1D); T1V = VADD(T1F, T1G); T21 = VADD(T1U, T1V); T1T = LDW(&(W[TWVL * 14])); T1W = VZMUL(T1T, VSUB(T1U, T1V)); T1s = VADD(Tf, Ty); T1t = VBYI(VADD(TI, TC)); T1r = LDW(&(W[TWVL * 28])); T1u = VZMULI(T1r, VSUB(T1s, T1t)); T1Z = LDW(&(W[0])); T20 = VZMULI(T1Z, VADD(T1s, T1t)); T1E = VSUB(T1C, T1D); T1H = VBYI(VSUB(T1F, T1G)); T1B = LDW(&(W[TWVL * 22])); T1I = VZMUL(T1B, VSUB(T1E, T1H)); T1N = LDW(&(W[TWVL * 6])); T1O = VZMUL(T1N, VADD(T1E, T1H)); Tz = VSUB(Tf, Ty); TJ = VBYI(VSUB(TC, TI)); T1 = LDW(&(W[TWVL * 12])); TK = VZMULI(T1, VADD(Tz, TJ)); T1R = LDW(&(W[TWVL * 16])); T1S = VZMULI(T1R, VSUB(Tz, TJ)); TW = VBYI(VSUB(TS, TV)); T11 = VSUB(TZ, T10); TL = LDW(&(W[TWVL * 10])); T12 = VZMUL(TL, VADD(TW, T11)); T1d = LDW(&(W[TWVL * 18])); T1e = VZMUL(T1d, VSUB(T11, TW)); T1i = VBYI(VADD(T1a, T19)); T1j = VADD(T16, T17); T1h = LDW(&(W[TWVL * 4])); T1k = VZMULI(T1h, VADD(T1i, T1j)); T1z = LDW(&(W[TWVL * 24])); T1A = VZMULI(T1z, VSUB(T1j, T1i)); T1m = VBYI(VADD(TV, TS)); T1n = VADD(TZ, T10); T1l = LDW(&(W[TWVL * 2])); T1o = VZMUL(T1l, VADD(T1m, T1n)); T1v = LDW(&(W[TWVL * 26])); T1w = VZMUL(T1v, VSUB(T1n, T1m)); T18 = VSUB(T16, T17); T1b = VBYI(VSUB(T19, T1a)); T15 = LDW(&(W[TWVL * 20])); T1c = VZMULI(T15, VSUB(T18, T1b)); T1L = LDW(&(W[TWVL * 8])); T1M = VZMULI(T1L, VADD(T1b, T18)); T13 = VADD(TK, T12); ST(&(Rp[WS(rs, 3)]), T13, ms, &(Rp[WS(rs, 1)])); T1g = VCONJ(VSUB(T1e, T1c)); ST(&(Rm[WS(rs, 5)]), T1g, -ms, &(Rm[WS(rs, 1)])); T1X = VADD(T1S, T1W); ST(&(Rp[WS(rs, 4)]), T1X, ms, &(Rp[0])); T23 = VCONJ(VSUB(T21, T20)); ST(&(Rm[0]), T23, -ms, &(Rm[0])); T14 = VCONJ(VSUB(T12, TK)); ST(&(Rm[WS(rs, 3)]), T14, -ms, &(Rm[WS(rs, 1)])); T1f = VADD(T1c, T1e); ST(&(Rp[WS(rs, 5)]), T1f, ms, &(Rp[WS(rs, 1)])); T1Y = VCONJ(VSUB(T1W, T1S)); ST(&(Rm[WS(rs, 4)]), T1Y, -ms, &(Rm[0])); T22 = VADD(T20, T21); ST(&(Rp[0]), T22, ms, &(Rp[0])); T1p = VADD(T1k, T1o); ST(&(Rp[WS(rs, 1)]), T1p, ms, &(Rp[WS(rs, 1)])); T1y = VCONJ(VSUB(T1w, T1u)); ST(&(Rm[WS(rs, 7)]), T1y, -ms, &(Rm[WS(rs, 1)])); T1J = VADD(T1A, T1I); ST(&(Rp[WS(rs, 6)]), T1J, ms, &(Rp[0])); T1Q = VCONJ(VSUB(T1O, T1M)); ST(&(Rm[WS(rs, 2)]), T1Q, -ms, &(Rm[0])); T1q = VCONJ(VSUB(T1o, T1k)); ST(&(Rm[WS(rs, 1)]), T1q, -ms, &(Rm[WS(rs, 1)])); T1x = VADD(T1u, T1w); ST(&(Rp[WS(rs, 7)]), T1x, ms, &(Rp[WS(rs, 1)])); T1K = VCONJ(VSUB(T1I, T1A)); ST(&(Rm[WS(rs, 6)]), T1K, -ms, &(Rm[0])); T1P = VADD(T1M, T1O); ST(&(Rp[WS(rs, 2)]), T1P, ms, &(Rp[0])); } } } VLEAVE(); } static const tw_instr twinstr[] = { VTW(1, 1), VTW(1, 2), VTW(1, 3), VTW(1, 4), VTW(1, 5), VTW(1, 6), VTW(1, 7), VTW(1, 8), VTW(1, 9), VTW(1, 10), VTW(1, 11), VTW(1, 12), VTW(1, 13), VTW(1, 14), VTW(1, 15), {TW_NEXT, VL, 0} }; static const hc2c_desc desc = { 16, XSIMD_STRING("hc2cbdftv_16"), twinstr, &GENUS, {99, 38, 4, 0} }; void XSIMD(codelet_hc2cbdftv_16) (planner *p) { X(khc2c_register) (p, hc2cbdftv_16, &desc, HC2C_VIA_DFT); } #endif /* HAVE_FMA */ fftw-3.3.3/rdft/simd/common/hc2cbdftv_32.c0000644000175000017500000010070212054411072015103 00000000000000/* * Copyright (c) 2003, 2007-11 Matteo Frigo * Copyright (c) 2003, 2007-11 Massachusetts Institute of Technology * * This program is free software; you can redistribute it and/or modify * it under the terms of the GNU General Public License as published by * the Free Software Foundation; either version 2 of the License, or * (at your option) any later version. * * This program is distributed in the hope that it will be useful, * but WITHOUT ANY WARRANTY; without even the implied warranty of * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the * GNU General Public License for more details. * * You should have received a copy of the GNU General Public License * along with this program; if not, write to the Free Software * Foundation, Inc., 51 Franklin Street, Fifth Floor, Boston, MA 02110-1301 USA * */ /* This file was automatically generated --- DO NOT EDIT */ /* Generated on Sun Nov 25 07:42:30 EST 2012 */ #include "codelet-rdft.h" #ifdef HAVE_FMA /* Generated by: ../../../genfft/gen_hc2cdft_c.native -fma -reorder-insns -schedule-for-pipeline -simd -compact -variables 4 -pipeline-latency 8 -trivial-stores -variables 32 -no-generate-bytw -n 32 -dif -sign 1 -name hc2cbdftv_32 -include hc2cbv.h */ /* * This function contains 249 FP additions, 192 FP multiplications, * (or, 119 additions, 62 multiplications, 130 fused multiply/add), * 166 stack variables, 7 constants, and 64 memory accesses */ #include "hc2cbv.h" static void hc2cbdftv_32(R *Rp, R *Ip, R *Rm, R *Im, const R *W, stride rs, INT mb, INT me, INT ms) { DVK(KP980785280, +0.980785280403230449126182236134239036973933731); DVK(KP831469612, +0.831469612302545237078788377617905756738560812); DVK(KP668178637, +0.668178637919298919997757686523080761552472251); DVK(KP198912367, +0.198912367379658006911597622644676228597850501); DVK(KP923879532, +0.923879532511286756128183189396788286822416626); DVK(KP414213562, +0.414213562373095048801688724209698078569671875); DVK(KP707106781, +0.707106781186547524400844362104849039284835938); { INT m; for (m = mb, W = W + ((mb - 1) * ((TWVL / VL) * 62)); m < me; m = m + VL, Rp = Rp + (VL * ms), Ip = Ip + (VL * ms), Rm = Rm - (VL * ms), Im = Im - (VL * ms), W = W + (TWVL * 62), MAKE_VOLATILE_STRIDE(128, rs)) { V T3a, T3N; { V T2G, T1o, T2o, T2Y, T1b, T1V, Ts, T1S, T3A, T48, T3p, T45, T31, T2z, T2H; V T1L, Tv, TG, TM, T3q, T1r, TX, TN, T1s, Ty, T1t, TB, TO, TQ, T1y; V T3t, TR, T1H, T1K, TV, T1p, T1q, T1w, TW, Tt, Tu, TE, TF, TK, TL; V Tw, Tx, Tz, TA, T1x; { V T1i, T4, T1j, T15, T1l, T1m, Tb, T16, Tf, T1G, Ti, T1F, Tm, T1J, T1I; V Tp, T2, T3, T13, T14, T5, T6, T8, T9, Td, T7, Ta, Te, Tg, Th; V Tk, Tl, Tn, To, T2m, Tc, T3l, T1k, T3m, T18, Tj, T3y, T1n, Tq, T19; V T3n, T17, T2x, T1a, T2n, T2y, Tr, T3z, T3o; T2 = LD(&(Rp[0]), ms, &(Rp[0])); T3 = LD(&(Rm[WS(rs, 15)]), -ms, &(Rm[WS(rs, 1)])); T13 = LD(&(Rp[WS(rs, 8)]), ms, &(Rp[0])); T14 = LD(&(Rm[WS(rs, 7)]), -ms, &(Rm[WS(rs, 1)])); T5 = LD(&(Rp[WS(rs, 4)]), ms, &(Rp[0])); T6 = LD(&(Rm[WS(rs, 11)]), -ms, &(Rm[WS(rs, 1)])); T8 = LD(&(Rp[WS(rs, 12)]), ms, &(Rp[0])); T9 = LD(&(Rm[WS(rs, 3)]), -ms, &(Rm[WS(rs, 1)])); Td = LD(&(Rp[WS(rs, 10)]), ms, &(Rp[0])); T1i = VFMACONJ(T3, T2); T4 = VFNMSCONJ(T3, T2); T1j = VFMACONJ(T14, T13); T15 = VFNMSCONJ(T14, T13); T1l = VFMACONJ(T6, T5); T7 = VFNMSCONJ(T6, T5); T1m = VFMACONJ(T9, T8); Ta = VFMSCONJ(T9, T8); Te = LD(&(Rm[WS(rs, 5)]), -ms, &(Rm[WS(rs, 1)])); Tg = LD(&(Rp[WS(rs, 2)]), ms, &(Rp[0])); Th = LD(&(Rm[WS(rs, 13)]), -ms, &(Rm[WS(rs, 1)])); Tk = LD(&(Rp[WS(rs, 6)]), ms, &(Rp[0])); Tl = LD(&(Rm[WS(rs, 9)]), -ms, &(Rm[WS(rs, 1)])); Tn = LD(&(Rp[WS(rs, 14)]), ms, &(Rp[0])); To = LD(&(Rm[WS(rs, 1)]), -ms, &(Rm[WS(rs, 1)])); Tb = VADD(T7, Ta); T16 = VSUB(T7, Ta); Tf = VFNMSCONJ(Te, Td); T1G = VFMACONJ(Te, Td); Ti = VFNMSCONJ(Th, Tg); T1F = VFMACONJ(Th, Tg); Tm = VFNMSCONJ(Tl, Tk); T1J = VFMACONJ(Tl, Tk); T1I = VFMACONJ(To, Tn); Tp = VFMSCONJ(To, Tn); T2m = VFMA(LDK(KP707106781), Tb, T4); Tc = VFNMS(LDK(KP707106781), Tb, T4); T3l = VSUB(T1i, T1j); T1k = VADD(T1i, T1j); T1H = VADD(T1F, T1G); T3m = VSUB(T1F, T1G); T18 = VFNMS(LDK(KP414213562), Tf, Ti); Tj = VFMA(LDK(KP414213562), Ti, Tf); T3y = VSUB(T1l, T1m); T1n = VADD(T1l, T1m); Tq = VFNMS(LDK(KP414213562), Tp, Tm); T19 = VFMA(LDK(KP414213562), Tm, Tp); T1K = VADD(T1I, T1J); T3n = VSUB(T1I, T1J); T17 = VFNMS(LDK(KP707106781), T16, T15); T2x = VFMA(LDK(KP707106781), T16, T15); T1a = VSUB(T18, T19); T2n = VADD(T18, T19); T2y = VADD(Tj, Tq); Tr = VSUB(Tj, Tq); T3z = VSUB(T3m, T3n); T3o = VADD(T3m, T3n); T2G = VADD(T1k, T1n); T1o = VSUB(T1k, T1n); T2o = VFNMS(LDK(KP923879532), T2n, T2m); T2Y = VFMA(LDK(KP923879532), T2n, T2m); T1b = VFNMS(LDK(KP923879532), T1a, T17); T1V = VFMA(LDK(KP923879532), T1a, T17); Ts = VFMA(LDK(KP923879532), Tr, Tc); T1S = VFNMS(LDK(KP923879532), Tr, Tc); T3A = VFMA(LDK(KP707106781), T3z, T3y); T48 = VFNMS(LDK(KP707106781), T3z, T3y); T3p = VFMA(LDK(KP707106781), T3o, T3l); T45 = VFNMS(LDK(KP707106781), T3o, T3l); T31 = VFMA(LDK(KP923879532), T2y, T2x); T2z = VFNMS(LDK(KP923879532), T2y, T2x); } Tt = LD(&(Rp[WS(rs, 1)]), ms, &(Rp[WS(rs, 1)])); Tu = LD(&(Rm[WS(rs, 14)]), -ms, &(Rm[0])); TE = LD(&(Rp[WS(rs, 9)]), ms, &(Rp[WS(rs, 1)])); TF = LD(&(Rm[WS(rs, 6)]), -ms, &(Rm[0])); TK = LD(&(Rp[WS(rs, 15)]), ms, &(Rp[WS(rs, 1)])); TL = LD(&(Rm[0]), -ms, &(Rm[0])); TV = LD(&(Rp[WS(rs, 7)]), ms, &(Rp[WS(rs, 1)])); T2H = VADD(T1H, T1K); T1L = VSUB(T1H, T1K); Tv = VFNMSCONJ(Tu, Tt); T1p = VFMACONJ(Tu, Tt); TG = VFNMSCONJ(TF, TE); T1q = VFMACONJ(TF, TE); T1w = VFMACONJ(TL, TK); TM = VFMSCONJ(TL, TK); TW = LD(&(Rm[WS(rs, 8)]), -ms, &(Rm[0])); Tw = LD(&(Rp[WS(rs, 5)]), ms, &(Rp[WS(rs, 1)])); Tx = LD(&(Rm[WS(rs, 10)]), -ms, &(Rm[0])); Tz = LD(&(Rp[WS(rs, 13)]), ms, &(Rp[WS(rs, 1)])); TA = LD(&(Rm[WS(rs, 2)]), -ms, &(Rm[0])); T3q = VSUB(T1p, T1q); T1r = VADD(T1p, T1q); T1x = VFMACONJ(TW, TV); TX = VFNMSCONJ(TW, TV); TN = LD(&(Rp[WS(rs, 3)]), ms, &(Rp[WS(rs, 1)])); T1s = VFMACONJ(Tx, Tw); Ty = VFNMSCONJ(Tx, Tw); T1t = VFMACONJ(TA, Tz); TB = VFMSCONJ(TA, Tz); TO = LD(&(Rm[WS(rs, 12)]), -ms, &(Rm[0])); TQ = LD(&(Rp[WS(rs, 11)]), ms, &(Rp[WS(rs, 1)])); T1y = VADD(T1w, T1x); T3t = VSUB(T1w, T1x); TR = LD(&(Rm[WS(rs, 4)]), -ms, &(Rm[0])); { V T38, T3f, T4p, T4v, T3T, T3Z, T2a, T2i, T4b, T4h, T1O, T20, T2M, T2U, T3F; V T3L, T2g, T3X, T3J, T1g, T4f, T2S, T4l, T2E, T2X, T3O, T3b, T3i, T26, T4t; V T43, T1Y, T3c, T30, T3d, T33; { V T2I, T2A, T2r, T1c, TJ, T2L, T2u, T2B, T10, T1d, T3x, T3E, T1E, T1N, T1h; V T1Z, T4m, T1M, T1D, T4a, T4o, T4n, T47, T4u, T3R, T3S, T3Q, T3Y, T28, T29; V T27, T2h, T44, T4g; { V T36, T1v, T2J, T3s, T3B, T2p, TI, T2q, TD, T1B, T3u, TY, TT, T35, T1u; V T3r, TH, TC, T1z, TP, T1A, TS, T3w, T3D, T1C, T2K, T3v, T3C, T2s, TZ; V T2t, TU, T37, T49, T46; T2I = VSUB(T2G, T2H); T36 = VADD(T2G, T2H); T1u = VADD(T1s, T1t); T3r = VSUB(T1s, T1t); TH = VSUB(Ty, TB); TC = VADD(Ty, TB); T1z = VFMACONJ(TO, TN); TP = VFNMSCONJ(TO, TN); T1A = VFMACONJ(TR, TQ); TS = VFMSCONJ(TR, TQ); T1v = VSUB(T1r, T1u); T2J = VADD(T1r, T1u); T3s = VFNMS(LDK(KP414213562), T3r, T3q); T3B = VFMA(LDK(KP414213562), T3q, T3r); T2p = VFMA(LDK(KP707106781), TH, TG); TI = VFNMS(LDK(KP707106781), TH, TG); T2q = VFMA(LDK(KP707106781), TC, Tv); TD = VFNMS(LDK(KP707106781), TC, Tv); T1B = VADD(T1z, T1A); T3u = VSUB(T1A, T1z); TY = VSUB(TS, TP); TT = VADD(TP, TS); T35 = LDW(&(W[TWVL * 30])); T4m = LDW(&(W[TWVL * 10])); T2A = VFNMS(LDK(KP198912367), T2p, T2q); T2r = VFMA(LDK(KP198912367), T2q, T2p); T1c = VFNMS(LDK(KP668178637), TD, TI); TJ = VFMA(LDK(KP668178637), TI, TD); T1C = VSUB(T1y, T1B); T2K = VADD(T1y, T1B); T3v = VFNMS(LDK(KP414213562), T3u, T3t); T3C = VFMA(LDK(KP414213562), T3t, T3u); T2s = VFNMS(LDK(KP707106781), TY, TX); TZ = VFMA(LDK(KP707106781), TY, TX); T2t = VFMA(LDK(KP707106781), TT, TM); TU = VFNMS(LDK(KP707106781), TT, TM); T1M = VSUB(T1v, T1C); T1D = VADD(T1v, T1C); T37 = VADD(T2J, T2K); T2L = VSUB(T2J, T2K); T3w = VADD(T3s, T3v); T49 = VSUB(T3s, T3v); T3D = VSUB(T3B, T3C); T46 = VADD(T3B, T3C); T2u = VFNMS(LDK(KP198912367), T2t, T2s); T2B = VFMA(LDK(KP198912367), T2s, T2t); T10 = VFNMS(LDK(KP668178637), TZ, TU); T1d = VFMA(LDK(KP668178637), TU, TZ); T38 = VZMUL(T35, VSUB(T36, T37)); T3f = VADD(T36, T37); T4a = VFMA(LDK(KP923879532), T49, T48); T4o = VFNMS(LDK(KP923879532), T49, T48); T4n = VFMA(LDK(KP923879532), T46, T45); T47 = VFNMS(LDK(KP923879532), T46, T45); T4u = LDW(&(W[TWVL * 50])); T3R = VFMA(LDK(KP923879532), T3w, T3p); T3x = VFNMS(LDK(KP923879532), T3w, T3p); T3E = VFNMS(LDK(KP923879532), T3D, T3A); T3S = VFMA(LDK(KP923879532), T3D, T3A); T3Q = LDW(&(W[TWVL * 58])); T3Y = LDW(&(W[TWVL * 2])); } T28 = VFMA(LDK(KP707106781), T1D, T1o); T1E = VFNMS(LDK(KP707106781), T1D, T1o); T1N = VFNMS(LDK(KP707106781), T1M, T1L); T29 = VFMA(LDK(KP707106781), T1M, T1L); T4p = VZMUL(T4m, VFNMSI(T4o, T4n)); T4v = VZMUL(T4u, VFMAI(T4o, T4n)); T27 = LDW(&(W[TWVL * 6])); T2h = LDW(&(W[TWVL * 54])); T3T = VZMUL(T3Q, VFNMSI(T3S, T3R)); T3Z = VZMUL(T3Y, VFMAI(T3S, T3R)); T44 = LDW(&(W[TWVL * 18])); T4g = LDW(&(W[TWVL * 42])); T2a = VZMUL(T27, VFMAI(T29, T28)); T2i = VZMUL(T2h, VFNMSI(T29, T28)); T1h = LDW(&(W[TWVL * 22])); T1Z = LDW(&(W[TWVL * 38])); T4b = VZMUL(T44, VFMAI(T4a, T47)); T4h = VZMUL(T4g, VFNMSI(T4a, T47)); { V T1W, T1T, T1, T3W, T2d, T3I, T2e, T12, T2f, T1f, T2F, T2T, T3k, T3K, T11; V T1e, T32, T2Z, T2l, T4k, T2P, T4e, T2Q, T2w, T2R, T2D, T2v, T2C, T1R, T4s; V T23, T42, T24, T1U, T25, T1X; T2F = LDW(&(W[TWVL * 46])); T2T = LDW(&(W[TWVL * 14])); T1O = VZMUL(T1h, VFNMSI(T1N, T1E)); T20 = VZMUL(T1Z, VFMAI(T1N, T1E)); T3k = LDW(&(W[TWVL * 26])); T3K = LDW(&(W[TWVL * 34])); T2M = VZMUL(T2F, VFNMSI(T2L, T2I)); T2U = VZMUL(T2T, VFMAI(T2L, T2I)); T11 = VADD(TJ, T10); T1W = VSUB(TJ, T10); T1T = VSUB(T1d, T1c); T1e = VADD(T1c, T1d); T1 = LDW(&(W[TWVL * 24])); T3W = LDW(&(W[TWVL * 4])); T3F = VZMUL(T3k, VFNMSI(T3E, T3x)); T3L = VZMUL(T3K, VFMAI(T3E, T3x)); T2d = LDW(&(W[TWVL * 56])); T3I = LDW(&(W[TWVL * 36])); T2e = VFMA(LDK(KP831469612), T11, Ts); T12 = VFNMS(LDK(KP831469612), T11, Ts); T2f = VFMA(LDK(KP831469612), T1e, T1b); T1f = VFNMS(LDK(KP831469612), T1e, T1b); T2v = VSUB(T2r, T2u); T32 = VADD(T2r, T2u); T2Z = VADD(T2A, T2B); T2C = VSUB(T2A, T2B); T2l = LDW(&(W[TWVL * 48])); T4k = LDW(&(W[TWVL * 12])); T2P = LDW(&(W[TWVL * 16])); T4e = LDW(&(W[TWVL * 44])); T2g = VZMULI(T2d, VFMAI(T2f, T2e)); T3X = VZMULI(T3W, VFNMSI(T2f, T2e)); T3J = VZMULI(T3I, VFNMSI(T1f, T12)); T1g = VZMULI(T1, VFMAI(T1f, T12)); T2Q = VFNMS(LDK(KP980785280), T2v, T2o); T2w = VFMA(LDK(KP980785280), T2v, T2o); T2R = VFMA(LDK(KP980785280), T2C, T2z); T2D = VFNMS(LDK(KP980785280), T2C, T2z); T1R = LDW(&(W[TWVL * 40])); T4s = LDW(&(W[TWVL * 52])); T23 = LDW(&(W[TWVL * 8])); T42 = LDW(&(W[TWVL * 20])); T4f = VZMULI(T4e, VFNMSI(T2R, T2Q)); T2S = VZMULI(T2P, VFMAI(T2R, T2Q)); T4l = VZMULI(T4k, VFNMSI(T2D, T2w)); T2E = VZMULI(T2l, VFMAI(T2D, T2w)); T24 = VFMA(LDK(KP831469612), T1T, T1S); T1U = VFNMS(LDK(KP831469612), T1T, T1S); T25 = VFMA(LDK(KP831469612), T1W, T1V); T1X = VFNMS(LDK(KP831469612), T1W, T1V); T2X = LDW(&(W[TWVL * 32])); T3O = LDW(&(W[TWVL * 60])); T3b = LDW(&(W[0])); T3i = LDW(&(W[TWVL * 28])); T26 = VZMULI(T23, VFMAI(T25, T24)); T4t = VZMULI(T4s, VFNMSI(T25, T24)); T43 = VZMULI(T42, VFNMSI(T1X, T1U)); T1Y = VZMULI(T1R, VFMAI(T1X, T1U)); T3c = VFMA(LDK(KP980785280), T2Z, T2Y); T30 = VFNMS(LDK(KP980785280), T2Z, T2Y); T3d = VFMA(LDK(KP980785280), T32, T31); T33 = VFNMS(LDK(KP980785280), T32, T31); } } { V T3e, T3P, T3j, T34, T2c, T4j, T2k, T4d, T1P, T1Q, T4x, T4w, T2j, T4c, T21; V T22, T4r, T4q, T2b, T4i, T3h, T3H, T2N, T2O, T41, T40, T3g, T3G, T2V, T2W; V T3V, T3U, T39, T3M; T1P = VADD(T1g, T1O); T1Q = VCONJ(VSUB(T1O, T1g)); T4x = VCONJ(VSUB(T4v, T4t)); T4w = VADD(T4t, T4v); T2j = VADD(T2g, T2i); T2k = VCONJ(VSUB(T2i, T2g)); T4d = VCONJ(VSUB(T4b, T43)); T4c = VADD(T43, T4b); T3e = VZMULI(T3b, VFMAI(T3d, T3c)); T3P = VZMULI(T3O, VFNMSI(T3d, T3c)); T3j = VZMULI(T3i, VFNMSI(T33, T30)); T34 = VZMULI(T2X, VFMAI(T33, T30)); ST(&(Rp[WS(rs, 6)]), T1P, ms, &(Rp[0])); ST(&(Rp[WS(rs, 13)]), T4w, ms, &(Rp[WS(rs, 1)])); ST(&(Rp[WS(rs, 14)]), T2j, ms, &(Rp[0])); ST(&(Rp[WS(rs, 5)]), T4c, ms, &(Rp[WS(rs, 1)])); ST(&(Rm[WS(rs, 13)]), T4x, -ms, &(Rm[WS(rs, 1)])); ST(&(Rm[WS(rs, 6)]), T1Q, -ms, &(Rm[0])); T21 = VADD(T1Y, T20); T22 = VCONJ(VSUB(T20, T1Y)); T4r = VCONJ(VSUB(T4p, T4l)); T4q = VADD(T4l, T4p); T2b = VADD(T26, T2a); T2c = VCONJ(VSUB(T2a, T26)); T4j = VCONJ(VSUB(T4h, T4f)); T4i = VADD(T4f, T4h); ST(&(Rm[WS(rs, 5)]), T4d, -ms, &(Rm[WS(rs, 1)])); ST(&(Rm[WS(rs, 14)]), T2k, -ms, &(Rm[0])); ST(&(Rp[WS(rs, 10)]), T21, ms, &(Rp[0])); ST(&(Rp[WS(rs, 3)]), T4q, ms, &(Rp[WS(rs, 1)])); ST(&(Rp[WS(rs, 2)]), T2b, ms, &(Rp[0])); ST(&(Rp[WS(rs, 11)]), T4i, ms, &(Rp[WS(rs, 1)])); ST(&(Rm[WS(rs, 3)]), T4r, -ms, &(Rm[WS(rs, 1)])); ST(&(Rm[WS(rs, 10)]), T22, -ms, &(Rm[0])); T2N = VADD(T2E, T2M); T2O = VCONJ(VSUB(T2M, T2E)); T41 = VCONJ(VSUB(T3Z, T3X)); T40 = VADD(T3X, T3Z); T3g = VADD(T3e, T3f); T3h = VCONJ(VSUB(T3f, T3e)); T3H = VCONJ(VSUB(T3F, T3j)); T3G = VADD(T3j, T3F); ST(&(Rm[WS(rs, 11)]), T4j, -ms, &(Rm[WS(rs, 1)])); ST(&(Rm[WS(rs, 2)]), T2c, -ms, &(Rm[0])); ST(&(Rp[WS(rs, 12)]), T2N, ms, &(Rp[0])); ST(&(Rp[WS(rs, 1)]), T40, ms, &(Rp[WS(rs, 1)])); ST(&(Rp[0]), T3g, ms, &(Rp[0])); ST(&(Rp[WS(rs, 7)]), T3G, ms, &(Rp[WS(rs, 1)])); ST(&(Rm[WS(rs, 1)]), T41, -ms, &(Rm[WS(rs, 1)])); ST(&(Rm[WS(rs, 12)]), T2O, -ms, &(Rm[0])); T2V = VADD(T2S, T2U); T2W = VCONJ(VSUB(T2U, T2S)); T3V = VCONJ(VSUB(T3T, T3P)); T3U = VADD(T3P, T3T); T39 = VADD(T34, T38); T3a = VCONJ(VSUB(T38, T34)); T3N = VCONJ(VSUB(T3L, T3J)); T3M = VADD(T3J, T3L); ST(&(Rm[WS(rs, 7)]), T3H, -ms, &(Rm[WS(rs, 1)])); ST(&(Rm[0]), T3h, -ms, &(Rm[0])); ST(&(Rp[WS(rs, 4)]), T2V, ms, &(Rp[0])); ST(&(Rp[WS(rs, 15)]), T3U, ms, &(Rp[WS(rs, 1)])); ST(&(Rp[WS(rs, 8)]), T39, ms, &(Rp[0])); ST(&(Rp[WS(rs, 9)]), T3M, ms, &(Rp[WS(rs, 1)])); ST(&(Rm[WS(rs, 15)]), T3V, -ms, &(Rm[WS(rs, 1)])); ST(&(Rm[WS(rs, 4)]), T2W, -ms, &(Rm[0])); } } } ST(&(Rm[WS(rs, 9)]), T3N, -ms, &(Rm[WS(rs, 1)])); ST(&(Rm[WS(rs, 8)]), T3a, -ms, &(Rm[0])); } } VLEAVE(); } static const tw_instr twinstr[] = { VTW(1, 1), VTW(1, 2), VTW(1, 3), VTW(1, 4), VTW(1, 5), VTW(1, 6), VTW(1, 7), VTW(1, 8), VTW(1, 9), VTW(1, 10), VTW(1, 11), VTW(1, 12), VTW(1, 13), VTW(1, 14), VTW(1, 15), VTW(1, 16), VTW(1, 17), VTW(1, 18), VTW(1, 19), VTW(1, 20), VTW(1, 21), VTW(1, 22), VTW(1, 23), VTW(1, 24), VTW(1, 25), VTW(1, 26), VTW(1, 27), VTW(1, 28), VTW(1, 29), VTW(1, 30), VTW(1, 31), {TW_NEXT, VL, 0} }; static const hc2c_desc desc = { 32, XSIMD_STRING("hc2cbdftv_32"), twinstr, &GENUS, {119, 62, 130, 0} }; void XSIMD(codelet_hc2cbdftv_32) (planner *p) { X(khc2c_register) (p, hc2cbdftv_32, &desc, HC2C_VIA_DFT); } #else /* HAVE_FMA */ /* Generated by: ../../../genfft/gen_hc2cdft_c.native -simd -compact -variables 4 -pipeline-latency 8 -trivial-stores -variables 32 -no-generate-bytw -n 32 -dif -sign 1 -name hc2cbdftv_32 -include hc2cbv.h */ /* * This function contains 249 FP additions, 104 FP multiplications, * (or, 233 additions, 88 multiplications, 16 fused multiply/add), * 161 stack variables, 7 constants, and 64 memory accesses */ #include "hc2cbv.h" static void hc2cbdftv_32(R *Rp, R *Ip, R *Rm, R *Im, const R *W, stride rs, INT mb, INT me, INT ms) { DVK(KP195090322, +0.195090322016128267848284868477022240927691618); DVK(KP980785280, +0.980785280403230449126182236134239036973933731); DVK(KP555570233, +0.555570233019602224742830813948532874374937191); DVK(KP831469612, +0.831469612302545237078788377617905756738560812); DVK(KP923879532, +0.923879532511286756128183189396788286822416626); DVK(KP382683432, +0.382683432365089771728459984030398866761344562); DVK(KP707106781, +0.707106781186547524400844362104849039284835938); { INT m; for (m = mb, W = W + ((mb - 1) * ((TWVL / VL) * 62)); m < me; m = m + VL, Rp = Rp + (VL * ms), Ip = Ip + (VL * ms), Rm = Rm - (VL * ms), Im = Im - (VL * ms), W = W + (TWVL * 62), MAKE_VOLATILE_STRIDE(128, rs)) { V T1W, T21, Tf, T2c, T1t, T2r, T3T, T4m, Ty, T2q, T3P, T4n, T1n, T2d, T1T; V T22, T1E, T24, T3I, T4p, TU, T2n, T1i, T2h, T1L, T25, T3L, T4q, T1f, T2o; V T1j, T2k; { V T2, T4, T1Z, T1p, T1r, T20, T9, T1U, Td, T1V, T3, T1q, T6, T8, T7; V Tc, Tb, Ta, T5, Te, T1o, T1s, T3R, T3S, Tj, T1N, Tw, T1Q, Tn, T1O; V Ts, T1R, Tg, Ti, Th, Tv, Tu, Tt, Tk, Tm, Tl, Tp, Tr, Tq, To; V Tx, T3N, T3O, T1l, T1m, T1P, T1S; T2 = LD(&(Rp[0]), ms, &(Rp[0])); T3 = LD(&(Rm[WS(rs, 15)]), -ms, &(Rm[WS(rs, 1)])); T4 = VCONJ(T3); T1Z = VADD(T2, T4); T1p = LD(&(Rp[WS(rs, 8)]), ms, &(Rp[0])); T1q = LD(&(Rm[WS(rs, 7)]), -ms, &(Rm[WS(rs, 1)])); T1r = VCONJ(T1q); T20 = VADD(T1p, T1r); T6 = LD(&(Rp[WS(rs, 4)]), ms, &(Rp[0])); T7 = LD(&(Rm[WS(rs, 11)]), -ms, &(Rm[WS(rs, 1)])); T8 = VCONJ(T7); T9 = VSUB(T6, T8); T1U = VADD(T6, T8); Tc = LD(&(Rp[WS(rs, 12)]), ms, &(Rp[0])); Ta = LD(&(Rm[WS(rs, 3)]), -ms, &(Rm[WS(rs, 1)])); Tb = VCONJ(Ta); Td = VSUB(Tb, Tc); T1V = VADD(Tb, Tc); T1W = VSUB(T1U, T1V); T21 = VSUB(T1Z, T20); T5 = VSUB(T2, T4); Te = VMUL(LDK(KP707106781), VADD(T9, Td)); Tf = VSUB(T5, Te); T2c = VADD(T5, Te); T1o = VMUL(LDK(KP707106781), VSUB(T9, Td)); T1s = VSUB(T1p, T1r); T1t = VSUB(T1o, T1s); T2r = VADD(T1s, T1o); T3R = VADD(T1Z, T20); T3S = VADD(T1U, T1V); T3T = VSUB(T3R, T3S); T4m = VADD(T3R, T3S); Tg = LD(&(Rp[WS(rs, 2)]), ms, &(Rp[0])); Th = LD(&(Rm[WS(rs, 13)]), -ms, &(Rm[WS(rs, 1)])); Ti = VCONJ(Th); Tj = VSUB(Tg, Ti); T1N = VADD(Tg, Ti); Tv = LD(&(Rp[WS(rs, 14)]), ms, &(Rp[0])); Tt = LD(&(Rm[WS(rs, 1)]), -ms, &(Rm[WS(rs, 1)])); Tu = VCONJ(Tt); Tw = VSUB(Tu, Tv); T1Q = VADD(Tu, Tv); Tk = LD(&(Rp[WS(rs, 10)]), ms, &(Rp[0])); Tl = LD(&(Rm[WS(rs, 5)]), -ms, &(Rm[WS(rs, 1)])); Tm = VCONJ(Tl); Tn = VSUB(Tk, Tm); T1O = VADD(Tk, Tm); Tp = LD(&(Rp[WS(rs, 6)]), ms, &(Rp[0])); Tq = LD(&(Rm[WS(rs, 9)]), -ms, &(Rm[WS(rs, 1)])); Tr = VCONJ(Tq); Ts = VSUB(Tp, Tr); T1R = VADD(Tp, Tr); To = VFMA(LDK(KP382683432), Tj, VMUL(LDK(KP923879532), Tn)); Tx = VFNMS(LDK(KP382683432), Tw, VMUL(LDK(KP923879532), Ts)); Ty = VSUB(To, Tx); T2q = VADD(To, Tx); T3N = VADD(T1N, T1O); T3O = VADD(T1Q, T1R); T3P = VSUB(T3N, T3O); T4n = VADD(T3N, T3O); T1l = VFNMS(LDK(KP382683432), Tn, VMUL(LDK(KP923879532), Tj)); T1m = VFMA(LDK(KP923879532), Tw, VMUL(LDK(KP382683432), Ts)); T1n = VSUB(T1l, T1m); T2d = VADD(T1l, T1m); T1P = VSUB(T1N, T1O); T1S = VSUB(T1Q, T1R); T1T = VMUL(LDK(KP707106781), VSUB(T1P, T1S)); T22 = VMUL(LDK(KP707106781), VADD(T1P, T1S)); } { V TD, T1B, TR, T1y, TH, T1C, TM, T1z, TA, TC, TB, TO, TQ, TP, TG; V TF, TE, TJ, TL, TK, T1A, T1D, T3G, T3H, TN, T2f, TT, T2g, TI, TS; V TY, T1I, T1c, T1F, T12, T1J, T17, T1G, TV, TX, TW, T1b, T1a, T19, T11; V T10, TZ, T14, T16, T15, T1H, T1K, T3J, T3K, T18, T2i, T1e, T2j, T13, T1d; TA = LD(&(Rp[WS(rs, 5)]), ms, &(Rp[WS(rs, 1)])); TB = LD(&(Rm[WS(rs, 10)]), -ms, &(Rm[0])); TC = VCONJ(TB); TD = VSUB(TA, TC); T1B = VADD(TA, TC); TO = LD(&(Rp[WS(rs, 1)]), ms, &(Rp[WS(rs, 1)])); TP = LD(&(Rm[WS(rs, 14)]), -ms, &(Rm[0])); TQ = VCONJ(TP); TR = VSUB(TO, TQ); T1y = VADD(TO, TQ); TG = LD(&(Rp[WS(rs, 13)]), ms, &(Rp[WS(rs, 1)])); TE = LD(&(Rm[WS(rs, 2)]), -ms, &(Rm[0])); TF = VCONJ(TE); TH = VSUB(TF, TG); T1C = VADD(TF, TG); TJ = LD(&(Rp[WS(rs, 9)]), ms, &(Rp[WS(rs, 1)])); TK = LD(&(Rm[WS(rs, 6)]), -ms, &(Rm[0])); TL = VCONJ(TK); TM = VSUB(TJ, TL); T1z = VADD(TJ, TL); T1A = VSUB(T1y, T1z); T1D = VSUB(T1B, T1C); T1E = VFNMS(LDK(KP382683432), T1D, VMUL(LDK(KP923879532), T1A)); T24 = VFMA(LDK(KP382683432), T1A, VMUL(LDK(KP923879532), T1D)); T3G = VADD(T1y, T1z); T3H = VADD(T1B, T1C); T3I = VSUB(T3G, T3H); T4p = VADD(T3G, T3H); TI = VMUL(LDK(KP707106781), VSUB(TD, TH)); TN = VSUB(TI, TM); T2f = VADD(TM, TI); TS = VMUL(LDK(KP707106781), VADD(TD, TH)); TT = VSUB(TR, TS); T2g = VADD(TR, TS); TU = VFMA(LDK(KP831469612), TN, VMUL(LDK(KP555570233), TT)); T2n = VFNMS(LDK(KP195090322), T2f, VMUL(LDK(KP980785280), T2g)); T1i = VFNMS(LDK(KP555570233), TN, VMUL(LDK(KP831469612), TT)); T2h = VFMA(LDK(KP980785280), T2f, VMUL(LDK(KP195090322), T2g)); TV = LD(&(Rp[WS(rs, 3)]), ms, &(Rp[WS(rs, 1)])); TW = LD(&(Rm[WS(rs, 12)]), -ms, &(Rm[0])); TX = VCONJ(TW); TY = VSUB(TV, TX); T1I = VADD(TV, TX); T1b = LD(&(Rp[WS(rs, 15)]), ms, &(Rp[WS(rs, 1)])); T19 = LD(&(Rm[0]), -ms, &(Rm[0])); T1a = VCONJ(T19); T1c = VSUB(T1a, T1b); T1F = VADD(T1a, T1b); T11 = LD(&(Rp[WS(rs, 11)]), ms, &(Rp[WS(rs, 1)])); TZ = LD(&(Rm[WS(rs, 4)]), -ms, &(Rm[0])); T10 = VCONJ(TZ); T12 = VSUB(T10, T11); T1J = VADD(T10, T11); T14 = LD(&(Rp[WS(rs, 7)]), ms, &(Rp[WS(rs, 1)])); T15 = LD(&(Rm[WS(rs, 8)]), -ms, &(Rm[0])); T16 = VCONJ(T15); T17 = VSUB(T14, T16); T1G = VADD(T14, T16); T1H = VSUB(T1F, T1G); T1K = VSUB(T1I, T1J); T1L = VFMA(LDK(KP923879532), T1H, VMUL(LDK(KP382683432), T1K)); T25 = VFNMS(LDK(KP382683432), T1H, VMUL(LDK(KP923879532), T1K)); T3J = VADD(T1F, T1G); T3K = VADD(T1I, T1J); T3L = VSUB(T3J, T3K); T4q = VADD(T3J, T3K); T13 = VMUL(LDK(KP707106781), VSUB(TY, T12)); T18 = VSUB(T13, T17); T2i = VADD(T17, T13); T1d = VMUL(LDK(KP707106781), VADD(TY, T12)); T1e = VSUB(T1c, T1d); T2j = VADD(T1c, T1d); T1f = VFNMS(LDK(KP555570233), T1e, VMUL(LDK(KP831469612), T18)); T2o = VFMA(LDK(KP195090322), T2i, VMUL(LDK(KP980785280), T2j)); T1j = VFMA(LDK(KP555570233), T18, VMUL(LDK(KP831469612), T1e)); T2k = VFNMS(LDK(KP195090322), T2j, VMUL(LDK(KP980785280), T2i)); } { V T4L, T4G, T4s, T4y, T3W, T4g, T42, T4a, T3g, T4e, T3o, T3E, T1w, T46, T2M; V T40, T2u, T4w, T2C, T4k, T36, T3A, T3i, T3s, T28, T2O, T2w, T2G, T2Y, T4K; V T3y, T4C; { V T4E, T4F, T4D, T4o, T4r, T4l, T4x, T3Q, T48, T3V, T49, T3M, T3U, T3F, T4f; V T41, T47, T3c, T3n, T3f, T3m, T3a, T3b, T3d, T3e, T39, T4d, T3l, T3D, T1h; V T2K, T1v, T2L, Tz, T1g, T1k, T1u, T1, T45, T2J, T3Z, T2m, T2A, T2t, T2B; V T2e, T2l, T2p, T2s, T2b, T4v, T2z, T4j; T4E = VADD(T4m, T4n); T4F = VADD(T4p, T4q); T4L = VADD(T4E, T4F); T4D = LDW(&(W[TWVL * 30])); T4G = VZMUL(T4D, VSUB(T4E, T4F)); T4o = VSUB(T4m, T4n); T4r = VBYI(VSUB(T4p, T4q)); T4l = LDW(&(W[TWVL * 46])); T4s = VZMUL(T4l, VSUB(T4o, T4r)); T4x = LDW(&(W[TWVL * 14])); T4y = VZMUL(T4x, VADD(T4o, T4r)); T3M = VMUL(LDK(KP707106781), VSUB(T3I, T3L)); T3Q = VBYI(VSUB(T3M, T3P)); T48 = VBYI(VADD(T3P, T3M)); T3U = VMUL(LDK(KP707106781), VADD(T3I, T3L)); T3V = VSUB(T3T, T3U); T49 = VADD(T3T, T3U); T3F = LDW(&(W[TWVL * 22])); T3W = VZMUL(T3F, VADD(T3Q, T3V)); T4f = LDW(&(W[TWVL * 54])); T4g = VZMUL(T4f, VSUB(T49, T48)); T41 = LDW(&(W[TWVL * 38])); T42 = VZMUL(T41, VSUB(T3V, T3Q)); T47 = LDW(&(W[TWVL * 6])); T4a = VZMUL(T47, VADD(T48, T49)); T3a = VADD(T1t, T1n); T3b = VADD(TU, T1f); T3c = VBYI(VADD(T3a, T3b)); T3n = VBYI(VSUB(T3b, T3a)); T3d = VADD(Tf, Ty); T3e = VADD(T1i, T1j); T3f = VADD(T3d, T3e); T3m = VSUB(T3d, T3e); T39 = LDW(&(W[TWVL * 4])); T3g = VZMULI(T39, VADD(T3c, T3f)); T4d = LDW(&(W[TWVL * 56])); T4e = VZMULI(T4d, VSUB(T3f, T3c)); T3l = LDW(&(W[TWVL * 36])); T3o = VZMULI(T3l, VSUB(T3m, T3n)); T3D = LDW(&(W[TWVL * 24])); T3E = VZMULI(T3D, VADD(T3n, T3m)); Tz = VSUB(Tf, Ty); T1g = VSUB(TU, T1f); T1h = VSUB(Tz, T1g); T2K = VADD(Tz, T1g); T1k = VSUB(T1i, T1j); T1u = VSUB(T1n, T1t); T1v = VBYI(VSUB(T1k, T1u)); T2L = VBYI(VADD(T1u, T1k)); T1 = LDW(&(W[TWVL * 20])); T1w = VZMULI(T1, VADD(T1h, T1v)); T45 = LDW(&(W[TWVL * 8])); T46 = VZMULI(T45, VADD(T2K, T2L)); T2J = LDW(&(W[TWVL * 52])); T2M = VZMULI(T2J, VSUB(T2K, T2L)); T3Z = LDW(&(W[TWVL * 40])); T40 = VZMULI(T3Z, VSUB(T1h, T1v)); T2e = VSUB(T2c, T2d); T2l = VSUB(T2h, T2k); T2m = VSUB(T2e, T2l); T2A = VADD(T2e, T2l); T2p = VSUB(T2n, T2o); T2s = VSUB(T2q, T2r); T2t = VBYI(VSUB(T2p, T2s)); T2B = VBYI(VADD(T2s, T2p)); T2b = LDW(&(W[TWVL * 44])); T2u = VZMULI(T2b, VSUB(T2m, T2t)); T4v = LDW(&(W[TWVL * 16])); T4w = VZMULI(T4v, VADD(T2m, T2t)); T2z = LDW(&(W[TWVL * 12])); T2C = VZMULI(T2z, VADD(T2A, T2B)); T4j = LDW(&(W[TWVL * 48])); T4k = VZMULI(T4j, VSUB(T2A, T2B)); { V T32, T3q, T35, T3r, T30, T31, T33, T34, T2Z, T3z, T3h, T3p, T1Y, T2E, T27; V T2F, T1M, T1X, T23, T26, T1x, T2N, T2v, T2D, T2U, T3x, T2X, T3w, T2S, T2T; V T2V, T2W, T2R, T4J, T3v, T4B; T30 = VADD(T21, T22); T31 = VADD(T1E, T1L); T32 = VADD(T30, T31); T3q = VSUB(T30, T31); T33 = VADD(T1W, T1T); T34 = VADD(T24, T25); T35 = VBYI(VADD(T33, T34)); T3r = VBYI(VSUB(T34, T33)); T2Z = LDW(&(W[TWVL * 58])); T36 = VZMUL(T2Z, VSUB(T32, T35)); T3z = LDW(&(W[TWVL * 26])); T3A = VZMUL(T3z, VADD(T3q, T3r)); T3h = LDW(&(W[TWVL * 2])); T3i = VZMUL(T3h, VADD(T32, T35)); T3p = LDW(&(W[TWVL * 34])); T3s = VZMUL(T3p, VSUB(T3q, T3r)); T1M = VSUB(T1E, T1L); T1X = VSUB(T1T, T1W); T1Y = VBYI(VSUB(T1M, T1X)); T2E = VBYI(VADD(T1X, T1M)); T23 = VSUB(T21, T22); T26 = VSUB(T24, T25); T27 = VSUB(T23, T26); T2F = VADD(T23, T26); T1x = LDW(&(W[TWVL * 18])); T28 = VZMUL(T1x, VADD(T1Y, T27)); T2N = LDW(&(W[TWVL * 50])); T2O = VZMUL(T2N, VSUB(T2F, T2E)); T2v = LDW(&(W[TWVL * 42])); T2w = VZMUL(T2v, VSUB(T27, T1Y)); T2D = LDW(&(W[TWVL * 10])); T2G = VZMUL(T2D, VADD(T2E, T2F)); T2S = VADD(T2c, T2d); T2T = VADD(T2n, T2o); T2U = VADD(T2S, T2T); T3x = VSUB(T2S, T2T); T2V = VADD(T2r, T2q); T2W = VADD(T2h, T2k); T2X = VBYI(VADD(T2V, T2W)); T3w = VBYI(VSUB(T2W, T2V)); T2R = LDW(&(W[TWVL * 60])); T2Y = VZMULI(T2R, VSUB(T2U, T2X)); T4J = LDW(&(W[0])); T4K = VZMULI(T4J, VADD(T2X, T2U)); T3v = LDW(&(W[TWVL * 28])); T3y = VZMULI(T3v, VADD(T3w, T3x)); T4B = LDW(&(W[TWVL * 32])); T4C = VZMULI(T4B, VSUB(T3x, T3w)); } } { V T29, T4M, T2P, T4t, T4N, T2a, T4u, T2Q, T2x, T4H, T2H, T4z, T4I, T2y, T4A; V T2I, T37, T4h, T3B, T3X, T4i, T38, T3Y, T3C, T3j, T4b, T3t, T43, T4c, T3k; V T44, T3u; T29 = VADD(T1w, T28); ST(&(Rp[WS(rs, 5)]), T29, ms, &(Rp[WS(rs, 1)])); T4M = VADD(T4K, T4L); ST(&(Rp[0]), T4M, ms, &(Rp[0])); T2P = VADD(T2M, T2O); ST(&(Rp[WS(rs, 13)]), T2P, ms, &(Rp[WS(rs, 1)])); T4t = VADD(T4k, T4s); ST(&(Rp[WS(rs, 12)]), T4t, ms, &(Rp[0])); T4N = VCONJ(VSUB(T4L, T4K)); ST(&(Rm[0]), T4N, -ms, &(Rm[0])); T2a = VCONJ(VSUB(T28, T1w)); ST(&(Rm[WS(rs, 5)]), T2a, -ms, &(Rm[WS(rs, 1)])); T4u = VCONJ(VSUB(T4s, T4k)); ST(&(Rm[WS(rs, 12)]), T4u, -ms, &(Rm[0])); T2Q = VCONJ(VSUB(T2O, T2M)); ST(&(Rm[WS(rs, 13)]), T2Q, -ms, &(Rm[WS(rs, 1)])); T2x = VADD(T2u, T2w); ST(&(Rp[WS(rs, 11)]), T2x, ms, &(Rp[WS(rs, 1)])); T4H = VADD(T4C, T4G); ST(&(Rp[WS(rs, 8)]), T4H, ms, &(Rp[0])); T2H = VADD(T2C, T2G); ST(&(Rp[WS(rs, 3)]), T2H, ms, &(Rp[WS(rs, 1)])); T4z = VADD(T4w, T4y); ST(&(Rp[WS(rs, 4)]), T4z, ms, &(Rp[0])); T4I = VCONJ(VSUB(T4G, T4C)); ST(&(Rm[WS(rs, 8)]), T4I, -ms, &(Rm[0])); T2y = VCONJ(VSUB(T2w, T2u)); ST(&(Rm[WS(rs, 11)]), T2y, -ms, &(Rm[WS(rs, 1)])); T4A = VCONJ(VSUB(T4y, T4w)); ST(&(Rm[WS(rs, 4)]), T4A, -ms, &(Rm[0])); T2I = VCONJ(VSUB(T2G, T2C)); ST(&(Rm[WS(rs, 3)]), T2I, -ms, &(Rm[WS(rs, 1)])); T37 = VADD(T2Y, T36); ST(&(Rp[WS(rs, 15)]), T37, ms, &(Rp[WS(rs, 1)])); T4h = VADD(T4e, T4g); ST(&(Rp[WS(rs, 14)]), T4h, ms, &(Rp[0])); T3B = VADD(T3y, T3A); ST(&(Rp[WS(rs, 7)]), T3B, ms, &(Rp[WS(rs, 1)])); T3X = VADD(T3E, T3W); ST(&(Rp[WS(rs, 6)]), T3X, ms, &(Rp[0])); T4i = VCONJ(VSUB(T4g, T4e)); ST(&(Rm[WS(rs, 14)]), T4i, -ms, &(Rm[0])); T38 = VCONJ(VSUB(T36, T2Y)); ST(&(Rm[WS(rs, 15)]), T38, -ms, &(Rm[WS(rs, 1)])); T3Y = VCONJ(VSUB(T3W, T3E)); ST(&(Rm[WS(rs, 6)]), T3Y, -ms, &(Rm[0])); T3C = VCONJ(VSUB(T3A, T3y)); ST(&(Rm[WS(rs, 7)]), T3C, -ms, &(Rm[WS(rs, 1)])); T3j = VADD(T3g, T3i); ST(&(Rp[WS(rs, 1)]), T3j, ms, &(Rp[WS(rs, 1)])); T4b = VADD(T46, T4a); ST(&(Rp[WS(rs, 2)]), T4b, ms, &(Rp[0])); T3t = VADD(T3o, T3s); ST(&(Rp[WS(rs, 9)]), T3t, ms, &(Rp[WS(rs, 1)])); T43 = VADD(T40, T42); ST(&(Rp[WS(rs, 10)]), T43, ms, &(Rp[0])); T4c = VCONJ(VSUB(T4a, T46)); ST(&(Rm[WS(rs, 2)]), T4c, -ms, &(Rm[0])); T3k = VCONJ(VSUB(T3i, T3g)); ST(&(Rm[WS(rs, 1)]), T3k, -ms, &(Rm[WS(rs, 1)])); T44 = VCONJ(VSUB(T42, T40)); ST(&(Rm[WS(rs, 10)]), T44, -ms, &(Rm[0])); T3u = VCONJ(VSUB(T3s, T3o)); ST(&(Rm[WS(rs, 9)]), T3u, -ms, &(Rm[WS(rs, 1)])); } } } } VLEAVE(); } static const tw_instr twinstr[] = { VTW(1, 1), VTW(1, 2), VTW(1, 3), VTW(1, 4), VTW(1, 5), VTW(1, 6), VTW(1, 7), VTW(1, 8), VTW(1, 9), VTW(1, 10), VTW(1, 11), VTW(1, 12), VTW(1, 13), VTW(1, 14), VTW(1, 15), VTW(1, 16), VTW(1, 17), VTW(1, 18), VTW(1, 19), VTW(1, 20), VTW(1, 21), VTW(1, 22), VTW(1, 23), VTW(1, 24), VTW(1, 25), VTW(1, 26), VTW(1, 27), VTW(1, 28), VTW(1, 29), VTW(1, 30), VTW(1, 31), {TW_NEXT, VL, 0} }; static const hc2c_desc desc = { 32, XSIMD_STRING("hc2cbdftv_32"), twinstr, &GENUS, {233, 88, 16, 0} }; void XSIMD(codelet_hc2cbdftv_32) (planner *p) { X(khc2c_register) (p, hc2cbdftv_32, &desc, HC2C_VIA_DFT); } #endif /* HAVE_FMA */ fftw-3.3.3/rdft/simd/common/hc2cbdftv_20.c0000644000175000017500000004727412054411067015122 00000000000000/* * Copyright (c) 2003, 2007-11 Matteo Frigo * Copyright (c) 2003, 2007-11 Massachusetts Institute of Technology * * This program is free software; you can redistribute it and/or modify * it under the terms of the GNU General Public License as published by * the Free Software Foundation; either version 2 of the License, or * (at your option) any later version. * * This program is distributed in the hope that it will be useful, * but WITHOUT ANY WARRANTY; without even the implied warranty of * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the * GNU General Public License for more details. * * You should have received a copy of the GNU General Public License * along with this program; if not, write to the Free Software * Foundation, Inc., 51 Franklin Street, Fifth Floor, Boston, MA 02110-1301 USA * */ /* This file was automatically generated --- DO NOT EDIT */ /* Generated on Sun Nov 25 07:42:30 EST 2012 */ #include "codelet-rdft.h" #ifdef HAVE_FMA /* Generated by: ../../../genfft/gen_hc2cdft_c.native -fma -reorder-insns -schedule-for-pipeline -simd -compact -variables 4 -pipeline-latency 8 -trivial-stores -variables 32 -no-generate-bytw -n 20 -dif -sign 1 -name hc2cbdftv_20 -include hc2cbv.h */ /* * This function contains 143 FP additions, 108 FP multiplications, * (or, 77 additions, 42 multiplications, 66 fused multiply/add), * 134 stack variables, 4 constants, and 40 memory accesses */ #include "hc2cbv.h" static void hc2cbdftv_20(R *Rp, R *Ip, R *Rm, R *Im, const R *W, stride rs, INT mb, INT me, INT ms) { DVK(KP559016994, +0.559016994374947424102293417182819058860154590); DVK(KP951056516, +0.951056516295153572116439333379382143405698634); DVK(KP250000000, +0.250000000000000000000000000000000000000000000); DVK(KP618033988, +0.618033988749894848204586834365638117720309180); { INT m; for (m = mb, W = W + ((mb - 1) * ((TWVL / VL) * 38)); m < me; m = m + VL, Rp = Rp + (VL * ms), Ip = Ip + (VL * ms), Rm = Rm - (VL * ms), Im = Im - (VL * ms), W = W + (TWVL * 38), MAKE_VOLATILE_STRIDE(80, rs)) { V T1M, T1T, T4, TF, T12, Te, T16, Ts, Tb, TN, TA, TG, TU, T1Y, T11; V T1e, T29, T21, T15, Th, T13, Tp; { V TS, TT, Tf, T10, T20, T1Z, TX, Tg, Tn, To, T2, T3, TD, TE, T8; V TV, T7, TZ, Tz, T9, Tu, Tv, T5, T6, Tx, Ty, Tc, Td, Tq, Tr; V TY, Ta, TW, Tw; T2 = LD(&(Rp[0]), ms, &(Rp[0])); T3 = LD(&(Rm[WS(rs, 9)]), -ms, &(Rm[WS(rs, 1)])); TD = LD(&(Rp[WS(rs, 5)]), ms, &(Rp[WS(rs, 1)])); TE = LD(&(Rm[WS(rs, 4)]), -ms, &(Rm[0])); T5 = LD(&(Rp[WS(rs, 4)]), ms, &(Rp[0])); T6 = LD(&(Rm[WS(rs, 5)]), -ms, &(Rm[WS(rs, 1)])); Tx = LD(&(Rp[WS(rs, 1)]), ms, &(Rp[WS(rs, 1)])); Ty = LD(&(Rm[WS(rs, 8)]), -ms, &(Rm[0])); T8 = LD(&(Rp[WS(rs, 6)]), ms, &(Rp[0])); TS = VFMACONJ(T3, T2); T4 = VFNMSCONJ(T3, T2); TT = VFMACONJ(TE, TD); TF = VFNMSCONJ(TE, TD); TV = VFMACONJ(T6, T5); T7 = VFNMSCONJ(T6, T5); TZ = VFMACONJ(Ty, Tx); Tz = VFNMSCONJ(Ty, Tx); T9 = LD(&(Rm[WS(rs, 3)]), -ms, &(Rm[WS(rs, 1)])); Tu = LD(&(Rp[WS(rs, 9)]), ms, &(Rp[WS(rs, 1)])); Tv = LD(&(Rm[0]), -ms, &(Rm[0])); Tc = LD(&(Rp[WS(rs, 8)]), ms, &(Rp[0])); Td = LD(&(Rm[WS(rs, 1)]), -ms, &(Rm[WS(rs, 1)])); Tq = LD(&(Rp[WS(rs, 7)]), ms, &(Rp[WS(rs, 1)])); Tr = LD(&(Rm[WS(rs, 2)]), -ms, &(Rm[0])); Tf = LD(&(Rp[WS(rs, 2)]), ms, &(Rp[0])); TY = VFMACONJ(T9, T8); Ta = VFMSCONJ(T9, T8); TW = VFMACONJ(Tv, Tu); Tw = VFNMSCONJ(Tv, Tu); T12 = VFMACONJ(Td, Tc); Te = VFNMSCONJ(Td, Tc); T16 = VFMACONJ(Tr, Tq); Ts = VFMSCONJ(Tr, Tq); T10 = VSUB(TY, TZ); T20 = VADD(TY, TZ); Tb = VADD(T7, Ta); TN = VSUB(T7, Ta); T1Z = VADD(TV, TW); TX = VSUB(TV, TW); TA = VSUB(Tw, Tz); TG = VADD(Tw, Tz); Tg = LD(&(Rm[WS(rs, 7)]), -ms, &(Rm[WS(rs, 1)])); Tn = LD(&(Rp[WS(rs, 3)]), ms, &(Rp[WS(rs, 1)])); To = LD(&(Rm[WS(rs, 6)]), -ms, &(Rm[0])); TU = VSUB(TS, TT); T1Y = VADD(TS, TT); T11 = VADD(TX, T10); T1e = VSUB(TX, T10); T29 = VSUB(T1Z, T20); T21 = VADD(T1Z, T20); T15 = VFMACONJ(Tg, Tf); Th = VFMSCONJ(Tg, Tf); T13 = VFMACONJ(To, Tn); Tp = VFMSCONJ(To, Tn); } { V T1S, T2B, T1W, T1I, T2q, T2w, T2i, T2c, T1C, T1K, T1s, T1g, T1, T2t, T1v; V T1Q, T2A, T1q, T2m, TC, T1w, TP, T1x, T2f, T2r, T2g, T1E, T1D, T2y, T2x; V T1i, T1h, T2D, T2C, T2s, T1t, T1u, T1y, T2u, TQ, T2d, T2e, T1U, T1L, T2j; V T2k; { V T1R, T1F, T1V, T1o, TO, Tl, T1d, T2a, T1l, TB, TK, T1G, Tk, T1b, T19; V T27, T25, T1H, TJ, T17, T23, TM, Ti, T14, T22, Tt, TH, Tj, T18, T24; V TI, T2b, T2p, T1X, T2v, T2h, T2n, T1B, T1f, T28, T2o, T1a, TR, T1J, T1r; V T1z, T26, Tm, TL, T1O, T1m, T1j, T2z, T1N, T1p, T1P, T2l, T1c, T1A, T1n; V T1k; T1R = LDW(&(W[TWVL * 18])); T17 = VSUB(T15, T16); T23 = VADD(T15, T16); TM = VSUB(Te, Th); Ti = VADD(Te, Th); T14 = VSUB(T12, T13); T22 = VADD(T12, T13); Tt = VSUB(Tp, Ts); TH = VADD(Tp, Ts); T1F = LDW(&(W[TWVL * 28])); T1V = LDW(&(W[TWVL * 8])); T1o = VFMA(LDK(KP618033988), TM, TN); TO = VFNMS(LDK(KP618033988), TN, TM); Tj = VADD(Tb, Ti); Tl = VSUB(Tb, Ti); T18 = VADD(T14, T17); T1d = VSUB(T14, T17); T24 = VADD(T22, T23); T2a = VSUB(T22, T23); T1l = VFMA(LDK(KP618033988), Tt, TA); TB = VFNMS(LDK(KP618033988), TA, Tt); TI = VADD(TG, TH); TK = VSUB(TG, TH); T1G = VADD(T4, Tj); Tk = VFNMS(LDK(KP250000000), Tj, T4); T1b = VSUB(T11, T18); T19 = VADD(T11, T18); T27 = VSUB(T21, T24); T25 = VADD(T21, T24); T1H = VADD(TF, TI); TJ = VFNMS(LDK(KP250000000), TI, TF); T2b = VMUL(LDK(KP951056516), VFMA(LDK(KP618033988), T2a, T29)); T2p = VMUL(LDK(KP951056516), VFNMS(LDK(KP618033988), T29, T2a)); T1X = LDW(&(W[TWVL * 6])); T1S = VZMUL(T1R, VADD(TU, T19)); T2v = LDW(&(W[TWVL * 22])); T2B = VADD(T1Y, T25); T26 = VFNMS(LDK(KP250000000), T25, T1Y); T1W = VZMULI(T1V, VFMAI(T1H, T1G)); T1I = VZMULI(T1F, VFNMSI(T1H, T1G)); T2h = LDW(&(W[TWVL * 30])); T2n = LDW(&(W[TWVL * 14])); T1B = VMUL(LDK(KP951056516), VFMA(LDK(KP618033988), T1d, T1e)); T1f = VMUL(LDK(KP951056516), VFNMS(LDK(KP618033988), T1e, T1d)); T28 = VFMA(LDK(KP559016994), T27, T26); T2o = VFNMS(LDK(KP559016994), T27, T26); T1a = VFNMS(LDK(KP250000000), T19, TU); TR = LDW(&(W[TWVL * 2])); T1J = LDW(&(W[TWVL * 26])); T1r = LDW(&(W[TWVL * 34])); T1z = LDW(&(W[TWVL * 10])); T1k = VFMA(LDK(KP559016994), Tl, Tk); Tm = VFNMS(LDK(KP559016994), Tl, Tk); T2q = VZMUL(T2n, VFMAI(T2p, T2o)); T2w = VZMUL(T2v, VFNMSI(T2p, T2o)); T2i = VZMUL(T2h, VFMAI(T2b, T28)); T2c = VZMUL(T1X, VFNMSI(T2b, T28)); T1c = VFNMS(LDK(KP559016994), T1b, T1a); T1A = VFMA(LDK(KP559016994), T1b, T1a); TL = VFNMS(LDK(KP559016994), TK, TJ); T1n = VFMA(LDK(KP559016994), TK, TJ); T1O = VFMA(LDK(KP951056516), T1l, T1k); T1m = VFNMS(LDK(KP951056516), T1l, T1k); T1j = LDW(&(W[TWVL * 36])); T2z = LDW(&(W[0])); T1N = LDW(&(W[TWVL * 20])); T1C = VZMUL(T1z, VFMAI(T1B, T1A)); T1K = VZMUL(T1J, VFNMSI(T1B, T1A)); T1s = VZMUL(T1r, VFMAI(T1f, T1c)); T1g = VZMUL(TR, VFNMSI(T1f, T1c)); T1p = VFMA(LDK(KP951056516), T1o, T1n); T1P = VFNMS(LDK(KP951056516), T1o, T1n); T2l = LDW(&(W[TWVL * 16])); T1 = LDW(&(W[TWVL * 4])); T2t = LDW(&(W[TWVL * 24])); T1v = LDW(&(W[TWVL * 12])); T1Q = VZMULI(T1N, VFNMSI(T1P, T1O)); T2A = VZMULI(T2z, VFMAI(T1p, T1m)); T1q = VZMULI(T1j, VFNMSI(T1p, T1m)); T2m = VZMULI(T2l, VFMAI(T1P, T1O)); TC = VFMA(LDK(KP951056516), TB, Tm); T1w = VFNMS(LDK(KP951056516), TB, Tm); TP = VFNMS(LDK(KP951056516), TO, TL); T1x = VFMA(LDK(KP951056516), TO, TL); T2f = LDW(&(W[TWVL * 32])); } T2D = VCONJ(VSUB(T2B, T2A)); T2C = VADD(T2A, T2B); T2s = VCONJ(VSUB(T2q, T2m)); T2r = VADD(T2m, T2q); T1t = VADD(T1q, T1s); T1u = VCONJ(VSUB(T1s, T1q)); T1y = VZMULI(T1v, VFNMSI(T1x, T1w)); T2u = VZMULI(T2t, VFMAI(T1x, T1w)); TQ = VZMULI(T1, VFNMSI(TP, TC)); T2g = VZMULI(T2f, VFMAI(TP, TC)); ST(&(Rm[0]), T2D, -ms, &(Rm[0])); ST(&(Rp[0]), T2C, ms, &(Rp[0])); ST(&(Rm[WS(rs, 4)]), T2s, -ms, &(Rm[0])); ST(&(Rm[WS(rs, 9)]), T1u, -ms, &(Rm[WS(rs, 1)])); T1E = VCONJ(VSUB(T1C, T1y)); T1D = VADD(T1y, T1C); T2y = VCONJ(VSUB(T2w, T2u)); T2x = VADD(T2u, T2w); T1i = VCONJ(VSUB(T1g, TQ)); T1h = VADD(TQ, T1g); ST(&(Rp[WS(rs, 9)]), T1t, ms, &(Rp[WS(rs, 1)])); T1L = VADD(T1I, T1K); T1M = VCONJ(VSUB(T1K, T1I)); ST(&(Rp[WS(rs, 3)]), T1D, ms, &(Rp[WS(rs, 1)])); ST(&(Rm[WS(rs, 6)]), T2y, -ms, &(Rm[0])); ST(&(Rp[WS(rs, 6)]), T2x, ms, &(Rp[0])); ST(&(Rm[WS(rs, 1)]), T1i, -ms, &(Rm[WS(rs, 1)])); ST(&(Rp[WS(rs, 1)]), T1h, ms, &(Rp[WS(rs, 1)])); T2d = VADD(T1W, T2c); T2e = VCONJ(VSUB(T2c, T1W)); ST(&(Rm[WS(rs, 3)]), T1E, -ms, &(Rm[WS(rs, 1)])); ST(&(Rp[WS(rs, 7)]), T1L, ms, &(Rp[WS(rs, 1)])); T1U = VCONJ(VSUB(T1S, T1Q)); T1T = VADD(T1Q, T1S); T2j = VADD(T2g, T2i); T2k = VCONJ(VSUB(T2i, T2g)); ST(&(Rp[WS(rs, 2)]), T2d, ms, &(Rp[0])); ST(&(Rp[WS(rs, 4)]), T2r, ms, &(Rp[0])); ST(&(Rm[WS(rs, 5)]), T1U, -ms, &(Rm[WS(rs, 1)])); ST(&(Rm[WS(rs, 2)]), T2e, -ms, &(Rm[0])); ST(&(Rp[WS(rs, 8)]), T2j, ms, &(Rp[0])); ST(&(Rm[WS(rs, 8)]), T2k, -ms, &(Rm[0])); } ST(&(Rp[WS(rs, 5)]), T1T, ms, &(Rp[WS(rs, 1)])); ST(&(Rm[WS(rs, 7)]), T1M, -ms, &(Rm[WS(rs, 1)])); } } VLEAVE(); } static const tw_instr twinstr[] = { VTW(1, 1), VTW(1, 2), VTW(1, 3), VTW(1, 4), VTW(1, 5), VTW(1, 6), VTW(1, 7), VTW(1, 8), VTW(1, 9), VTW(1, 10), VTW(1, 11), VTW(1, 12), VTW(1, 13), VTW(1, 14), VTW(1, 15), VTW(1, 16), VTW(1, 17), VTW(1, 18), VTW(1, 19), {TW_NEXT, VL, 0} }; static const hc2c_desc desc = { 20, XSIMD_STRING("hc2cbdftv_20"), twinstr, &GENUS, {77, 42, 66, 0} }; void XSIMD(codelet_hc2cbdftv_20) (planner *p) { X(khc2c_register) (p, hc2cbdftv_20, &desc, HC2C_VIA_DFT); } #else /* HAVE_FMA */ /* Generated by: ../../../genfft/gen_hc2cdft_c.native -simd -compact -variables 4 -pipeline-latency 8 -trivial-stores -variables 32 -no-generate-bytw -n 20 -dif -sign 1 -name hc2cbdftv_20 -include hc2cbv.h */ /* * This function contains 143 FP additions, 62 FP multiplications, * (or, 131 additions, 50 multiplications, 12 fused multiply/add), * 114 stack variables, 4 constants, and 40 memory accesses */ #include "hc2cbv.h" static void hc2cbdftv_20(R *Rp, R *Ip, R *Rm, R *Im, const R *W, stride rs, INT mb, INT me, INT ms) { DVK(KP250000000, +0.250000000000000000000000000000000000000000000); DVK(KP559016994, +0.559016994374947424102293417182819058860154590); DVK(KP951056516, +0.951056516295153572116439333379382143405698634); DVK(KP587785252, +0.587785252292473129168705954639072768597652438); { INT m; for (m = mb, W = W + ((mb - 1) * ((TWVL / VL) * 38)); m < me; m = m + VL, Rp = Rp + (VL * ms), Ip = Ip + (VL * ms), Rm = Rm - (VL * ms), Im = Im - (VL * ms), W = W + (TWVL * 38), MAKE_VOLATILE_STRIDE(80, rs)) { V TK, T1v, TY, T1x, T1j, T2f, TS, TT, TO, TU, T5, To, Tp, Tq, T2a; V T2d, T2g, T2k, T2j, T1k, T1l, T18, T1m, T1f; { V T2, TP, T4, TR, TI, T1d, T9, T12, Td, T15, TE, T1a, Tv, T13, Tm; V T1c, Tz, T16, Ti, T19, T3, TQ, TH, TG, TF, T6, T8, T7, Tc, Tb; V Ta, TD, TC, TB, Ts, Tu, Tt, Tl, Tk, Tj, Tw, Ty, Tx, Tf, Th; V Tg, TA, TJ, TW, TX, T1h, T1i, TM, TN, Te, Tn, T28, T29, T2b, T2c; V T14, T17, T1b, T1e; T2 = LD(&(Rp[0]), ms, &(Rp[0])); TP = LD(&(Rp[WS(rs, 5)]), ms, &(Rp[WS(rs, 1)])); T3 = LD(&(Rm[WS(rs, 9)]), -ms, &(Rm[WS(rs, 1)])); T4 = VCONJ(T3); TQ = LD(&(Rm[WS(rs, 4)]), -ms, &(Rm[0])); TR = VCONJ(TQ); TH = LD(&(Rp[WS(rs, 7)]), ms, &(Rp[WS(rs, 1)])); TF = LD(&(Rm[WS(rs, 2)]), -ms, &(Rm[0])); TG = VCONJ(TF); TI = VSUB(TG, TH); T1d = VADD(TG, TH); T6 = LD(&(Rp[WS(rs, 4)]), ms, &(Rp[0])); T7 = LD(&(Rm[WS(rs, 5)]), -ms, &(Rm[WS(rs, 1)])); T8 = VCONJ(T7); T9 = VSUB(T6, T8); T12 = VADD(T6, T8); Tc = LD(&(Rp[WS(rs, 6)]), ms, &(Rp[0])); Ta = LD(&(Rm[WS(rs, 3)]), -ms, &(Rm[WS(rs, 1)])); Tb = VCONJ(Ta); Td = VSUB(Tb, Tc); T15 = VADD(Tb, Tc); TD = LD(&(Rp[WS(rs, 3)]), ms, &(Rp[WS(rs, 1)])); TB = LD(&(Rm[WS(rs, 6)]), -ms, &(Rm[0])); TC = VCONJ(TB); TE = VSUB(TC, TD); T1a = VADD(TC, TD); Ts = LD(&(Rp[WS(rs, 9)]), ms, &(Rp[WS(rs, 1)])); Tt = LD(&(Rm[0]), -ms, &(Rm[0])); Tu = VCONJ(Tt); Tv = VSUB(Ts, Tu); T13 = VADD(Ts, Tu); Tl = LD(&(Rp[WS(rs, 2)]), ms, &(Rp[0])); Tj = LD(&(Rm[WS(rs, 7)]), -ms, &(Rm[WS(rs, 1)])); Tk = VCONJ(Tj); Tm = VSUB(Tk, Tl); T1c = VADD(Tk, Tl); Tw = LD(&(Rp[WS(rs, 1)]), ms, &(Rp[WS(rs, 1)])); Tx = LD(&(Rm[WS(rs, 8)]), -ms, &(Rm[0])); Ty = VCONJ(Tx); Tz = VSUB(Tw, Ty); T16 = VADD(Tw, Ty); Tf = LD(&(Rp[WS(rs, 8)]), ms, &(Rp[0])); Tg = LD(&(Rm[WS(rs, 1)]), -ms, &(Rm[WS(rs, 1)])); Th = VCONJ(Tg); Ti = VSUB(Tf, Th); T19 = VADD(Tf, Th); TA = VSUB(Tv, Tz); TJ = VSUB(TE, TI); TK = VFNMS(LDK(KP951056516), TJ, VMUL(LDK(KP587785252), TA)); T1v = VFMA(LDK(KP951056516), TA, VMUL(LDK(KP587785252), TJ)); TW = VSUB(T9, Td); TX = VSUB(Ti, Tm); TY = VFNMS(LDK(KP951056516), TX, VMUL(LDK(KP587785252), TW)); T1x = VFMA(LDK(KP951056516), TW, VMUL(LDK(KP587785252), TX)); T1h = VADD(T2, T4); T1i = VADD(TP, TR); T1j = VSUB(T1h, T1i); T2f = VADD(T1h, T1i); TS = VSUB(TP, TR); TM = VADD(Tv, Tz); TN = VADD(TE, TI); TT = VADD(TM, TN); TO = VMUL(LDK(KP559016994), VSUB(TM, TN)); TU = VFNMS(LDK(KP250000000), TT, TS); T5 = VSUB(T2, T4); Te = VADD(T9, Td); Tn = VADD(Ti, Tm); To = VADD(Te, Tn); Tp = VFNMS(LDK(KP250000000), To, T5); Tq = VMUL(LDK(KP559016994), VSUB(Te, Tn)); T28 = VADD(T12, T13); T29 = VADD(T15, T16); T2a = VADD(T28, T29); T2b = VADD(T19, T1a); T2c = VADD(T1c, T1d); T2d = VADD(T2b, T2c); T2g = VADD(T2a, T2d); T2k = VSUB(T2b, T2c); T2j = VSUB(T28, T29); T14 = VSUB(T12, T13); T17 = VSUB(T15, T16); T1k = VADD(T14, T17); T1b = VSUB(T19, T1a); T1e = VSUB(T1c, T1d); T1l = VADD(T1b, T1e); T18 = VSUB(T14, T17); T1m = VADD(T1k, T1l); T1f = VSUB(T1b, T1e); } { V T2L, T22, T1S, T26, T2m, T2G, T2s, T2A, T1q, T1U, T1C, T1M, T10, T2E, T1I; V T2q, T1A, T2K, T20, T2w, T21, T1Q, T1R, T1P, T25, T1r, T1s, T2C, T2N, T1N; V T2H, T2I, T2M, T1E, T1D, T1O, T1V, T2n, T2B, T24, T2o, T2t, T2u, T23, T1W; T2L = VADD(T2f, T2g); T21 = LDW(&(W[TWVL * 18])); T22 = VZMUL(T21, VADD(T1j, T1m)); T1Q = VADD(T5, To); T1R = VBYI(VADD(TS, TT)); T1P = LDW(&(W[TWVL * 28])); T1S = VZMULI(T1P, VSUB(T1Q, T1R)); T25 = LDW(&(W[TWVL * 8])); T26 = VZMULI(T25, VADD(T1Q, T1R)); { V T2l, T2z, T2i, T2y, T2e, T2h, T27, T2F, T2r, T2x, T1g, T1K, T1p, T1L, T1n; V T1o, T11, T1T, T1B, T1J, TL, T1G, TZ, T1H, Tr, TV, T1, T2D, T1F, T2p; V T1w, T1Y, T1z, T1Z, T1u, T1y, T1t, T2J, T1X, T2v; T2l = VBYI(VFMA(LDK(KP951056516), T2j, VMUL(LDK(KP587785252), T2k))); T2z = VBYI(VFNMS(LDK(KP951056516), T2k, VMUL(LDK(KP587785252), T2j))); T2e = VMUL(LDK(KP559016994), VSUB(T2a, T2d)); T2h = VFNMS(LDK(KP250000000), T2g, T2f); T2i = VADD(T2e, T2h); T2y = VSUB(T2h, T2e); T27 = LDW(&(W[TWVL * 6])); T2m = VZMUL(T27, VSUB(T2i, T2l)); T2F = LDW(&(W[TWVL * 22])); T2G = VZMUL(T2F, VADD(T2z, T2y)); T2r = LDW(&(W[TWVL * 30])); T2s = VZMUL(T2r, VADD(T2l, T2i)); T2x = LDW(&(W[TWVL * 14])); T2A = VZMUL(T2x, VSUB(T2y, T2z)); T1g = VBYI(VFNMS(LDK(KP951056516), T1f, VMUL(LDK(KP587785252), T18))); T1K = VBYI(VFMA(LDK(KP951056516), T18, VMUL(LDK(KP587785252), T1f))); T1n = VFNMS(LDK(KP250000000), T1m, T1j); T1o = VMUL(LDK(KP559016994), VSUB(T1k, T1l)); T1p = VSUB(T1n, T1o); T1L = VADD(T1o, T1n); T11 = LDW(&(W[TWVL * 2])); T1q = VZMUL(T11, VADD(T1g, T1p)); T1T = LDW(&(W[TWVL * 26])); T1U = VZMUL(T1T, VSUB(T1L, T1K)); T1B = LDW(&(W[TWVL * 34])); T1C = VZMUL(T1B, VSUB(T1p, T1g)); T1J = LDW(&(W[TWVL * 10])); T1M = VZMUL(T1J, VADD(T1K, T1L)); Tr = VSUB(Tp, Tq); TL = VSUB(Tr, TK); T1G = VADD(Tr, TK); TV = VSUB(TO, TU); TZ = VBYI(VSUB(TV, TY)); T1H = VBYI(VADD(TY, TV)); T1 = LDW(&(W[TWVL * 4])); T10 = VZMULI(T1, VADD(TL, TZ)); T2D = LDW(&(W[TWVL * 24])); T2E = VZMULI(T2D, VSUB(T1G, T1H)); T1F = LDW(&(W[TWVL * 12])); T1I = VZMULI(T1F, VADD(T1G, T1H)); T2p = LDW(&(W[TWVL * 32])); T2q = VZMULI(T2p, VSUB(TL, TZ)); T1u = VADD(Tq, Tp); T1w = VSUB(T1u, T1v); T1Y = VADD(T1u, T1v); T1y = VADD(TO, TU); T1z = VBYI(VADD(T1x, T1y)); T1Z = VBYI(VSUB(T1y, T1x)); T1t = LDW(&(W[TWVL * 36])); T1A = VZMULI(T1t, VSUB(T1w, T1z)); T2J = LDW(&(W[0])); T2K = VZMULI(T2J, VADD(T1w, T1z)); T1X = LDW(&(W[TWVL * 20])); T20 = VZMULI(T1X, VSUB(T1Y, T1Z)); T2v = LDW(&(W[TWVL * 16])); T2w = VZMULI(T2v, VADD(T1Y, T1Z)); } T1r = VADD(T10, T1q); ST(&(Rp[WS(rs, 1)]), T1r, ms, &(Rp[WS(rs, 1)])); T1s = VCONJ(VSUB(T1q, T10)); ST(&(Rm[WS(rs, 1)]), T1s, -ms, &(Rm[WS(rs, 1)])); T2C = VCONJ(VSUB(T2A, T2w)); ST(&(Rm[WS(rs, 4)]), T2C, -ms, &(Rm[0])); T2N = VCONJ(VSUB(T2L, T2K)); ST(&(Rm[0]), T2N, -ms, &(Rm[0])); T1N = VADD(T1I, T1M); ST(&(Rp[WS(rs, 3)]), T1N, ms, &(Rp[WS(rs, 1)])); T2H = VADD(T2E, T2G); ST(&(Rp[WS(rs, 6)]), T2H, ms, &(Rp[0])); T2I = VCONJ(VSUB(T2G, T2E)); ST(&(Rm[WS(rs, 6)]), T2I, -ms, &(Rm[0])); T2M = VADD(T2K, T2L); ST(&(Rp[0]), T2M, ms, &(Rp[0])); T1E = VCONJ(VSUB(T1C, T1A)); ST(&(Rm[WS(rs, 9)]), T1E, -ms, &(Rm[WS(rs, 1)])); T1D = VADD(T1A, T1C); ST(&(Rp[WS(rs, 9)]), T1D, ms, &(Rp[WS(rs, 1)])); T1O = VCONJ(VSUB(T1M, T1I)); ST(&(Rm[WS(rs, 3)]), T1O, -ms, &(Rm[WS(rs, 1)])); T1V = VADD(T1S, T1U); ST(&(Rp[WS(rs, 7)]), T1V, ms, &(Rp[WS(rs, 1)])); T2n = VADD(T26, T2m); ST(&(Rp[WS(rs, 2)]), T2n, ms, &(Rp[0])); T2B = VADD(T2w, T2A); ST(&(Rp[WS(rs, 4)]), T2B, ms, &(Rp[0])); T24 = VCONJ(VSUB(T22, T20)); ST(&(Rm[WS(rs, 5)]), T24, -ms, &(Rm[WS(rs, 1)])); T2o = VCONJ(VSUB(T2m, T26)); ST(&(Rm[WS(rs, 2)]), T2o, -ms, &(Rm[0])); T2t = VADD(T2q, T2s); ST(&(Rp[WS(rs, 8)]), T2t, ms, &(Rp[0])); T2u = VCONJ(VSUB(T2s, T2q)); ST(&(Rm[WS(rs, 8)]), T2u, -ms, &(Rm[0])); T23 = VADD(T20, T22); ST(&(Rp[WS(rs, 5)]), T23, ms, &(Rp[WS(rs, 1)])); T1W = VCONJ(VSUB(T1U, T1S)); ST(&(Rm[WS(rs, 7)]), T1W, -ms, &(Rm[WS(rs, 1)])); } } } VLEAVE(); } static const tw_instr twinstr[] = { VTW(1, 1), VTW(1, 2), VTW(1, 3), VTW(1, 4), VTW(1, 5), VTW(1, 6), VTW(1, 7), VTW(1, 8), VTW(1, 9), VTW(1, 10), VTW(1, 11), VTW(1, 12), VTW(1, 13), VTW(1, 14), VTW(1, 15), VTW(1, 16), VTW(1, 17), VTW(1, 18), VTW(1, 19), {TW_NEXT, VL, 0} }; static const hc2c_desc desc = { 20, XSIMD_STRING("hc2cbdftv_20"), twinstr, &GENUS, {131, 50, 12, 0} }; void XSIMD(codelet_hc2cbdftv_20) (planner *p) { X(khc2c_register) (p, hc2cbdftv_20, &desc, HC2C_VIA_DFT); } #endif /* HAVE_FMA */ fftw-3.3.3/rdft/simd/common/codlist.c0000644000175000017500000000340712054411067014403 00000000000000#include "ifftw.h" #include SIMD_HEADER extern void XSIMD(codelet_hc2cfdftv_2)(planner *); extern void XSIMD(codelet_hc2cfdftv_4)(planner *); extern void XSIMD(codelet_hc2cfdftv_6)(planner *); extern void XSIMD(codelet_hc2cfdftv_8)(planner *); extern void XSIMD(codelet_hc2cfdftv_10)(planner *); extern void XSIMD(codelet_hc2cfdftv_12)(planner *); extern void XSIMD(codelet_hc2cfdftv_16)(planner *); extern void XSIMD(codelet_hc2cfdftv_32)(planner *); extern void XSIMD(codelet_hc2cfdftv_20)(planner *); extern void XSIMD(codelet_hc2cbdftv_2)(planner *); extern void XSIMD(codelet_hc2cbdftv_4)(planner *); extern void XSIMD(codelet_hc2cbdftv_6)(planner *); extern void XSIMD(codelet_hc2cbdftv_8)(planner *); extern void XSIMD(codelet_hc2cbdftv_10)(planner *); extern void XSIMD(codelet_hc2cbdftv_12)(planner *); extern void XSIMD(codelet_hc2cbdftv_16)(planner *); extern void XSIMD(codelet_hc2cbdftv_32)(planner *); extern void XSIMD(codelet_hc2cbdftv_20)(planner *); extern const solvtab XSIMD(solvtab_rdft); const solvtab XSIMD(solvtab_rdft) = { SOLVTAB(XSIMD(codelet_hc2cfdftv_2)), SOLVTAB(XSIMD(codelet_hc2cfdftv_4)), SOLVTAB(XSIMD(codelet_hc2cfdftv_6)), SOLVTAB(XSIMD(codelet_hc2cfdftv_8)), SOLVTAB(XSIMD(codelet_hc2cfdftv_10)), SOLVTAB(XSIMD(codelet_hc2cfdftv_12)), SOLVTAB(XSIMD(codelet_hc2cfdftv_16)), SOLVTAB(XSIMD(codelet_hc2cfdftv_32)), SOLVTAB(XSIMD(codelet_hc2cfdftv_20)), SOLVTAB(XSIMD(codelet_hc2cbdftv_2)), SOLVTAB(XSIMD(codelet_hc2cbdftv_4)), SOLVTAB(XSIMD(codelet_hc2cbdftv_6)), SOLVTAB(XSIMD(codelet_hc2cbdftv_8)), SOLVTAB(XSIMD(codelet_hc2cbdftv_10)), SOLVTAB(XSIMD(codelet_hc2cbdftv_12)), SOLVTAB(XSIMD(codelet_hc2cbdftv_16)), SOLVTAB(XSIMD(codelet_hc2cbdftv_32)), SOLVTAB(XSIMD(codelet_hc2cbdftv_20)), SOLVTAB_END }; fftw-3.3.3/rdft/simd/common/genus.c0000644000175000017500000000371012054410050014047 00000000000000/* * Copyright (c) 2003, 2007-11 Matteo Frigo * Copyright (c) 2003, 2007-11 Massachusetts Institute of Technology * * This program is free software; you can redistribute it and/or modify * it under the terms of the GNU General Public License as published by * the Free Software Foundation; either version 2 of the License, or * (at your option) any later version. * * This program is distributed in the hope that it will be useful, * but WITHOUT ANY WARRANTY; without even the implied warranty of * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the * GNU General Public License for more details. * * You should have received a copy of the GNU General Public License * along with this program; if not, write to the Free Software * Foundation, Inc., 51 Franklin Street, Fifth Floor, Boston, MA 02110-1301 USA * */ #include "codelet-rdft.h" #include SIMD_HEADER #define EXTERN_CONST(t, x) extern const t x; const t x static int hc2cbv_okp(const R *Rp, const R *Ip, const R *Rm, const R *Im, INT rs, INT mb, INT me, INT ms, const planner *plnr) { return (1 && !NO_SIMDP(plnr) && SIMD_STRIDE_OK(rs) && SIMD_VSTRIDE_OK(ms) && ((me - mb) % VL) == 0 && ((mb - 1) % VL) == 0 /* twiddle factors alignment */ && ALIGNED(Rp) && ALIGNED(Rm) && Ip == Rp + 1 && Im == Rm + 1); } EXTERN_CONST(hc2c_genus, XSIMD(rdft_hc2cbv_genus)) = { hc2cbv_okp, HC2R, VL }; static int hc2cfv_okp(const R *Rp, const R *Ip, const R *Rm, const R *Im, INT rs, INT mb, INT me, INT ms, const planner *plnr) { return (1 && !NO_SIMDP(plnr) && SIMD_STRIDE_OK(rs) && SIMD_VSTRIDE_OK(ms) && ((me - mb) % VL) == 0 && ((mb - 1) % VL) == 0 /* twiddle factors alignment */ && ALIGNED(Rp) && ALIGNED(Rm) && Ip == Rp + 1 && Im == Rm + 1); } EXTERN_CONST(hc2c_genus, XSIMD(rdft_hc2cfv_genus)) = { hc2cfv_okp, R2HC, VL }; fftw-3.3.3/rdft/simd/sse2/0000755000175000017500000000000012054411110012223 500000000000000fftw-3.3.3/rdft/simd/sse2/Makefile.am0000644000175000017500000000043712054410050014205 00000000000000AM_CFLAGS = $(SSE2_CFLAGS) SIMD_HEADER=simd-sse2.h include $(top_srcdir)/rdft/simd/codlist.mk include $(top_srcdir)/rdft/simd/simd.mk if HAVE_SSE2 BUILT_SOURCES = $(EXTRA_DIST) noinst_LTLIBRARIES = librdft_sse2_codelets.la librdft_sse2_codelets_la_SOURCES = $(BUILT_SOURCES) endif fftw-3.3.3/rdft/simd/sse2/Makefile.in0000644000175000017500000005012212054410200014207 00000000000000# Makefile.in generated by automake 1.11.6 from Makefile.am. # @configure_input@ # Copyright (C) 1994, 1995, 1996, 1997, 1998, 1999, 2000, 2001, 2002, # 2003, 2004, 2005, 2006, 2007, 2008, 2009, 2010, 2011 Free Software # Foundation, Inc. # This Makefile.in is free software; the Free Software Foundation # gives unlimited permission to copy and/or distribute it, # with or without modifications, as long as this notice is preserved. # This program is distributed in the hope that it will be useful, # but WITHOUT ANY WARRANTY, to the extent permitted by law; without # even the implied warranty of MERCHANTABILITY or FITNESS FOR A # PARTICULAR PURPOSE. @SET_MAKE@ # This file contains a standard list of RDFT SIMD codelets. It is # included by common/Makefile to generate the C files with the actual # codelets in them. It is included by {sse,sse2,...}/Makefile to # generate and compile stub files that include common/*.c # You can customize FFTW for special needs, e.g. to handle certain # sizes more efficiently, by adding new codelets to the lists of those # included by default. If you change the list of codelets, any new # ones you added will be automatically generated when you run the # bootstrap script (see "Generating your own code" in the FFTW # manual). VPATH = @srcdir@ am__make_dryrun = \ { \ am__dry=no; \ case $$MAKEFLAGS in \ *\\[\ \ ]*) \ echo 'am--echo: ; @echo "AM" OK' | $(MAKE) -f - 2>/dev/null \ | grep '^AM OK$$' >/dev/null || am__dry=yes;; \ *) \ for am__flg in $$MAKEFLAGS; do \ case $$am__flg in \ *=*|--*) ;; \ *n*) am__dry=yes; break;; \ esac; \ done;; \ esac; \ test $$am__dry = yes; \ } pkgdatadir = $(datadir)/@PACKAGE@ pkgincludedir = $(includedir)/@PACKAGE@ pkglibdir = $(libdir)/@PACKAGE@ pkglibexecdir = $(libexecdir)/@PACKAGE@ am__cd = CDPATH="$${ZSH_VERSION+.}$(PATH_SEPARATOR)" && cd install_sh_DATA = $(install_sh) -c -m 644 install_sh_PROGRAM = $(install_sh) -c install_sh_SCRIPT = $(install_sh) -c INSTALL_HEADER = $(INSTALL_DATA) transform = $(program_transform_name) NORMAL_INSTALL = : PRE_INSTALL = : POST_INSTALL = : NORMAL_UNINSTALL = : PRE_UNINSTALL = : POST_UNINSTALL = : build_triplet = @build@ host_triplet = @host@ DIST_COMMON = $(srcdir)/Makefile.am $(srcdir)/Makefile.in \ $(top_srcdir)/rdft/simd/codlist.mk \ $(top_srcdir)/rdft/simd/simd.mk subdir = rdft/simd/sse2 ACLOCAL_M4 = $(top_srcdir)/aclocal.m4 am__aclocal_m4_deps = $(top_srcdir)/m4/acx_mpi.m4 \ $(top_srcdir)/m4/acx_pthread.m4 \ $(top_srcdir)/m4/ax_cc_maxopt.m4 \ $(top_srcdir)/m4/ax_check_compiler_flags.m4 \ $(top_srcdir)/m4/ax_compiler_vendor.m4 \ $(top_srcdir)/m4/ax_gcc_aligns_stack.m4 \ $(top_srcdir)/m4/ax_gcc_version.m4 \ $(top_srcdir)/m4/ax_openmp.m4 $(top_srcdir)/m4/libtool.m4 \ $(top_srcdir)/m4/ltoptions.m4 $(top_srcdir)/m4/ltsugar.m4 \ $(top_srcdir)/m4/ltversion.m4 $(top_srcdir)/m4/lt~obsolete.m4 \ $(top_srcdir)/configure.ac am__configure_deps = $(am__aclocal_m4_deps) $(CONFIGURE_DEPENDENCIES) \ $(ACLOCAL_M4) mkinstalldirs = $(install_sh) -d CONFIG_HEADER = $(top_builddir)/config.h CONFIG_CLEAN_FILES = CONFIG_CLEAN_VPATH_FILES = LTLIBRARIES = $(noinst_LTLIBRARIES) librdft_sse2_codelets_la_LIBADD = am__librdft_sse2_codelets_la_SOURCES_DIST = hc2cfdftv_2.c \ hc2cfdftv_4.c hc2cfdftv_6.c hc2cfdftv_8.c hc2cfdftv_10.c \ hc2cfdftv_12.c hc2cfdftv_16.c hc2cfdftv_32.c hc2cfdftv_20.c \ hc2cbdftv_2.c hc2cbdftv_4.c hc2cbdftv_6.c hc2cbdftv_8.c \ hc2cbdftv_10.c hc2cbdftv_12.c hc2cbdftv_16.c hc2cbdftv_32.c \ hc2cbdftv_20.c genus.c codlist.c am__objects_1 = hc2cfdftv_2.lo hc2cfdftv_4.lo hc2cfdftv_6.lo \ hc2cfdftv_8.lo hc2cfdftv_10.lo hc2cfdftv_12.lo hc2cfdftv_16.lo \ hc2cfdftv_32.lo hc2cfdftv_20.lo am__objects_2 = hc2cbdftv_2.lo hc2cbdftv_4.lo hc2cbdftv_6.lo \ hc2cbdftv_8.lo hc2cbdftv_10.lo hc2cbdftv_12.lo hc2cbdftv_16.lo \ hc2cbdftv_32.lo hc2cbdftv_20.lo am__objects_3 = $(am__objects_1) $(am__objects_2) am__objects_4 = $(am__objects_3) genus.lo codlist.lo @HAVE_SSE2_TRUE@am__objects_5 = $(am__objects_4) @HAVE_SSE2_TRUE@am_librdft_sse2_codelets_la_OBJECTS = \ @HAVE_SSE2_TRUE@ $(am__objects_5) librdft_sse2_codelets_la_OBJECTS = \ $(am_librdft_sse2_codelets_la_OBJECTS) @HAVE_SSE2_TRUE@am_librdft_sse2_codelets_la_rpath = DEFAULT_INCLUDES = -I.@am__isrc@ -I$(top_builddir) depcomp = $(SHELL) $(top_srcdir)/depcomp am__depfiles_maybe = depfiles am__mv = mv -f COMPILE = $(CC) $(DEFS) $(DEFAULT_INCLUDES) $(INCLUDES) $(AM_CPPFLAGS) \ $(CPPFLAGS) $(AM_CFLAGS) $(CFLAGS) LTCOMPILE = $(LIBTOOL) --tag=CC $(AM_LIBTOOLFLAGS) $(LIBTOOLFLAGS) \ --mode=compile $(CC) $(DEFS) $(DEFAULT_INCLUDES) $(INCLUDES) \ $(AM_CPPFLAGS) $(CPPFLAGS) $(AM_CFLAGS) $(CFLAGS) CCLD = $(CC) LINK = $(LIBTOOL) --tag=CC $(AM_LIBTOOLFLAGS) $(LIBTOOLFLAGS) \ --mode=link $(CCLD) $(AM_CFLAGS) $(CFLAGS) $(AM_LDFLAGS) \ $(LDFLAGS) -o $@ SOURCES = $(librdft_sse2_codelets_la_SOURCES) DIST_SOURCES = $(am__librdft_sse2_codelets_la_SOURCES_DIST) am__can_run_installinfo = \ case $$AM_UPDATE_INFO_DIR in \ n|no|NO) false;; \ *) (install-info --version) >/dev/null 2>&1;; \ esac ETAGS = etags CTAGS = ctags DISTFILES = $(DIST_COMMON) $(DIST_SOURCES) $(TEXINFOS) $(EXTRA_DIST) ACLOCAL = @ACLOCAL@ ALLOCA = @ALLOCA@ ALTIVEC_CFLAGS = @ALTIVEC_CFLAGS@ AMTAR = @AMTAR@ AR = @AR@ AS = @AS@ AUTOCONF = @AUTOCONF@ AUTOHEADER = @AUTOHEADER@ AUTOMAKE = @AUTOMAKE@ AVX_CFLAGS = @AVX_CFLAGS@ AWK = @AWK@ CC = @CC@ CCDEPMODE = @CCDEPMODE@ CFLAGS = @CFLAGS@ CHECK_PL_OPTS = @CHECK_PL_OPTS@ CPP = @CPP@ CPPFLAGS = @CPPFLAGS@ CYGPATH_W = @CYGPATH_W@ C_FFTW_R2R_KIND = @C_FFTW_R2R_KIND@ C_MPI_FINT = @C_MPI_FINT@ DEFS = @DEFS@ DEPDIR = @DEPDIR@ DLLTOOL = @DLLTOOL@ DSYMUTIL = @DSYMUTIL@ DUMPBIN = @DUMPBIN@ ECHO_C = @ECHO_C@ ECHO_N = @ECHO_N@ ECHO_T = @ECHO_T@ EGREP = @EGREP@ EXEEXT = @EXEEXT@ F77 = @F77@ FFLAGS = @FFLAGS@ FGREP = @FGREP@ FLIBS = @FLIBS@ GREP = @GREP@ INSTALL = @INSTALL@ INSTALL_DATA = @INSTALL_DATA@ INSTALL_PROGRAM = @INSTALL_PROGRAM@ INSTALL_SCRIPT = @INSTALL_SCRIPT@ INSTALL_STRIP_PROGRAM = @INSTALL_STRIP_PROGRAM@ LD = @LD@ LDFLAGS = @LDFLAGS@ LIBOBJS = @LIBOBJS@ LIBQUADMATH = @LIBQUADMATH@ LIBS = @LIBS@ LIBTOOL = @LIBTOOL@ LIPO = @LIPO@ LN_S = @LN_S@ LTLIBOBJS = @LTLIBOBJS@ MAINT = @MAINT@ MAKEINFO = @MAKEINFO@ MANIFEST_TOOL = @MANIFEST_TOOL@ MKDIR_P = @MKDIR_P@ MPICC = @MPICC@ MPILIBS = @MPILIBS@ MPIRUN = @MPIRUN@ NEON_CFLAGS = @NEON_CFLAGS@ NM = @NM@ NMEDIT = @NMEDIT@ OBJDUMP = @OBJDUMP@ OBJEXT = @OBJEXT@ OCAMLBUILD = @OCAMLBUILD@ OPENMP_CFLAGS = @OPENMP_CFLAGS@ OTOOL = @OTOOL@ OTOOL64 = @OTOOL64@ PACKAGE = @PACKAGE@ PACKAGE_BUGREPORT = @PACKAGE_BUGREPORT@ PACKAGE_NAME = @PACKAGE_NAME@ PACKAGE_STRING = @PACKAGE_STRING@ PACKAGE_TARNAME = @PACKAGE_TARNAME@ PACKAGE_URL = @PACKAGE_URL@ PACKAGE_VERSION = @PACKAGE_VERSION@ PATH_SEPARATOR = @PATH_SEPARATOR@ POW_LIB = @POW_LIB@ PRECISION = @PRECISION@ PREC_SUFFIX = @PREC_SUFFIX@ PTHREAD_CC = @PTHREAD_CC@ PTHREAD_CFLAGS = @PTHREAD_CFLAGS@ PTHREAD_LIBS = @PTHREAD_LIBS@ RANLIB = @RANLIB@ SED = @SED@ SET_MAKE = @SET_MAKE@ SHARED_VERSION_INFO = @SHARED_VERSION_INFO@ SHELL = @SHELL@ SSE2_CFLAGS = @SSE2_CFLAGS@ STACK_ALIGN_CFLAGS = @STACK_ALIGN_CFLAGS@ STRIP = @STRIP@ THREADLIBS = @THREADLIBS@ VERSION = @VERSION@ abs_builddir = @abs_builddir@ abs_srcdir = @abs_srcdir@ abs_top_builddir = @abs_top_builddir@ abs_top_srcdir = @abs_top_srcdir@ ac_ct_AR = @ac_ct_AR@ ac_ct_CC = @ac_ct_CC@ ac_ct_DUMPBIN = @ac_ct_DUMPBIN@ ac_ct_F77 = @ac_ct_F77@ acx_pthread_config = @acx_pthread_config@ am__include = @am__include@ am__leading_dot = @am__leading_dot@ am__quote = @am__quote@ am__tar = @am__tar@ am__untar = @am__untar@ bindir = @bindir@ build = @build@ build_alias = @build_alias@ build_cpu = @build_cpu@ build_os = @build_os@ build_vendor = @build_vendor@ builddir = @builddir@ datadir = @datadir@ datarootdir = @datarootdir@ docdir = @docdir@ dvidir = @dvidir@ exec_prefix = @exec_prefix@ host = @host@ host_alias = @host_alias@ host_cpu = @host_cpu@ host_os = @host_os@ host_vendor = @host_vendor@ htmldir = @htmldir@ includedir = @includedir@ infodir = @infodir@ install_sh = @install_sh@ libdir = @libdir@ libexecdir = @libexecdir@ localedir = @localedir@ localstatedir = @localstatedir@ mandir = @mandir@ mkdir_p = @mkdir_p@ oldincludedir = @oldincludedir@ pdfdir = @pdfdir@ prefix = @prefix@ program_transform_name = @program_transform_name@ psdir = @psdir@ sbindir = @sbindir@ sharedstatedir = @sharedstatedir@ srcdir = @srcdir@ sysconfdir = @sysconfdir@ target_alias = @target_alias@ top_build_prefix = @top_build_prefix@ top_builddir = @top_builddir@ top_srcdir = @top_srcdir@ AM_CFLAGS = $(SSE2_CFLAGS) SIMD_HEADER = simd-sse2.h HC2CFDFTV = hc2cfdftv_2.c hc2cfdftv_4.c hc2cfdftv_6.c hc2cfdftv_8.c \ hc2cfdftv_10.c hc2cfdftv_12.c hc2cfdftv_16.c hc2cfdftv_32.c \ hc2cfdftv_20.c HC2CBDFTV = hc2cbdftv_2.c hc2cbdftv_4.c hc2cbdftv_6.c hc2cbdftv_8.c \ hc2cbdftv_10.c hc2cbdftv_12.c hc2cbdftv_16.c hc2cbdftv_32.c \ hc2cbdftv_20.c ########################################################################### SIMD_CODELETS = $(HC2CFDFTV) $(HC2CBDFTV) AM_CPPFLAGS = -I$(top_srcdir)/kernel -I$(top_srcdir)/rdft \ -I$(top_srcdir)/rdft/simd -I$(top_srcdir)/simd-support EXTRA_DIST = $(SIMD_CODELETS) genus.c codlist.c @HAVE_SSE2_TRUE@BUILT_SOURCES = $(EXTRA_DIST) @HAVE_SSE2_TRUE@noinst_LTLIBRARIES = librdft_sse2_codelets.la @HAVE_SSE2_TRUE@librdft_sse2_codelets_la_SOURCES = $(BUILT_SOURCES) all: $(BUILT_SOURCES) $(MAKE) $(AM_MAKEFLAGS) all-am .SUFFIXES: .SUFFIXES: .c .lo .o .obj $(srcdir)/Makefile.in: @MAINTAINER_MODE_TRUE@ $(srcdir)/Makefile.am $(top_srcdir)/rdft/simd/codlist.mk $(top_srcdir)/rdft/simd/simd.mk $(am__configure_deps) @for dep in $?; do \ case '$(am__configure_deps)' in \ *$$dep*) \ ( cd $(top_builddir) && $(MAKE) $(AM_MAKEFLAGS) am--refresh ) \ && { if test -f $@; then exit 0; else break; fi; }; \ exit 1;; \ esac; \ done; \ echo ' cd $(top_srcdir) && $(AUTOMAKE) --gnu rdft/simd/sse2/Makefile'; \ $(am__cd) $(top_srcdir) && \ $(AUTOMAKE) --gnu rdft/simd/sse2/Makefile .PRECIOUS: Makefile Makefile: $(srcdir)/Makefile.in $(top_builddir)/config.status @case '$?' in \ *config.status*) \ cd $(top_builddir) && $(MAKE) $(AM_MAKEFLAGS) am--refresh;; \ *) \ echo ' cd $(top_builddir) && $(SHELL) ./config.status $(subdir)/$@ $(am__depfiles_maybe)'; \ cd $(top_builddir) && $(SHELL) ./config.status $(subdir)/$@ $(am__depfiles_maybe);; \ esac; $(top_srcdir)/rdft/simd/codlist.mk $(top_srcdir)/rdft/simd/simd.mk: $(top_builddir)/config.status: $(top_srcdir)/configure $(CONFIG_STATUS_DEPENDENCIES) cd $(top_builddir) && $(MAKE) $(AM_MAKEFLAGS) am--refresh $(top_srcdir)/configure: @MAINTAINER_MODE_TRUE@ $(am__configure_deps) cd $(top_builddir) && $(MAKE) $(AM_MAKEFLAGS) am--refresh $(ACLOCAL_M4): @MAINTAINER_MODE_TRUE@ $(am__aclocal_m4_deps) cd $(top_builddir) && $(MAKE) $(AM_MAKEFLAGS) am--refresh $(am__aclocal_m4_deps): clean-noinstLTLIBRARIES: -test -z "$(noinst_LTLIBRARIES)" || rm -f $(noinst_LTLIBRARIES) @list='$(noinst_LTLIBRARIES)'; for p in $$list; do \ dir="`echo $$p | sed -e 's|/[^/]*$$||'`"; \ test "$$dir" != "$$p" || dir=.; \ echo "rm -f \"$${dir}/so_locations\""; \ rm -f "$${dir}/so_locations"; \ done librdft_sse2_codelets.la: $(librdft_sse2_codelets_la_OBJECTS) $(librdft_sse2_codelets_la_DEPENDENCIES) $(EXTRA_librdft_sse2_codelets_la_DEPENDENCIES) $(LINK) $(am_librdft_sse2_codelets_la_rpath) $(librdft_sse2_codelets_la_OBJECTS) $(librdft_sse2_codelets_la_LIBADD) $(LIBS) mostlyclean-compile: -rm -f *.$(OBJEXT) distclean-compile: -rm -f *.tab.c @AMDEP_TRUE@@am__include@ @am__quote@./$(DEPDIR)/codlist.Plo@am__quote@ @AMDEP_TRUE@@am__include@ @am__quote@./$(DEPDIR)/genus.Plo@am__quote@ @AMDEP_TRUE@@am__include@ @am__quote@./$(DEPDIR)/hc2cbdftv_10.Plo@am__quote@ @AMDEP_TRUE@@am__include@ @am__quote@./$(DEPDIR)/hc2cbdftv_12.Plo@am__quote@ @AMDEP_TRUE@@am__include@ @am__quote@./$(DEPDIR)/hc2cbdftv_16.Plo@am__quote@ @AMDEP_TRUE@@am__include@ @am__quote@./$(DEPDIR)/hc2cbdftv_2.Plo@am__quote@ @AMDEP_TRUE@@am__include@ @am__quote@./$(DEPDIR)/hc2cbdftv_20.Plo@am__quote@ @AMDEP_TRUE@@am__include@ @am__quote@./$(DEPDIR)/hc2cbdftv_32.Plo@am__quote@ @AMDEP_TRUE@@am__include@ @am__quote@./$(DEPDIR)/hc2cbdftv_4.Plo@am__quote@ @AMDEP_TRUE@@am__include@ @am__quote@./$(DEPDIR)/hc2cbdftv_6.Plo@am__quote@ @AMDEP_TRUE@@am__include@ @am__quote@./$(DEPDIR)/hc2cbdftv_8.Plo@am__quote@ @AMDEP_TRUE@@am__include@ @am__quote@./$(DEPDIR)/hc2cfdftv_10.Plo@am__quote@ @AMDEP_TRUE@@am__include@ @am__quote@./$(DEPDIR)/hc2cfdftv_12.Plo@am__quote@ @AMDEP_TRUE@@am__include@ @am__quote@./$(DEPDIR)/hc2cfdftv_16.Plo@am__quote@ @AMDEP_TRUE@@am__include@ @am__quote@./$(DEPDIR)/hc2cfdftv_2.Plo@am__quote@ @AMDEP_TRUE@@am__include@ @am__quote@./$(DEPDIR)/hc2cfdftv_20.Plo@am__quote@ @AMDEP_TRUE@@am__include@ @am__quote@./$(DEPDIR)/hc2cfdftv_32.Plo@am__quote@ @AMDEP_TRUE@@am__include@ @am__quote@./$(DEPDIR)/hc2cfdftv_4.Plo@am__quote@ @AMDEP_TRUE@@am__include@ @am__quote@./$(DEPDIR)/hc2cfdftv_6.Plo@am__quote@ @AMDEP_TRUE@@am__include@ @am__quote@./$(DEPDIR)/hc2cfdftv_8.Plo@am__quote@ .c.o: @am__fastdepCC_TRUE@ $(COMPILE) -MT $@ -MD -MP -MF $(DEPDIR)/$*.Tpo -c -o $@ $< @am__fastdepCC_TRUE@ $(am__mv) $(DEPDIR)/$*.Tpo $(DEPDIR)/$*.Po @AMDEP_TRUE@@am__fastdepCC_FALSE@ source='$<' object='$@' libtool=no @AMDEPBACKSLASH@ @AMDEP_TRUE@@am__fastdepCC_FALSE@ DEPDIR=$(DEPDIR) $(CCDEPMODE) $(depcomp) @AMDEPBACKSLASH@ @am__fastdepCC_FALSE@ $(COMPILE) -c $< .c.obj: @am__fastdepCC_TRUE@ $(COMPILE) -MT $@ -MD -MP -MF $(DEPDIR)/$*.Tpo -c -o $@ `$(CYGPATH_W) '$<'` @am__fastdepCC_TRUE@ $(am__mv) $(DEPDIR)/$*.Tpo $(DEPDIR)/$*.Po @AMDEP_TRUE@@am__fastdepCC_FALSE@ source='$<' object='$@' libtool=no @AMDEPBACKSLASH@ @AMDEP_TRUE@@am__fastdepCC_FALSE@ DEPDIR=$(DEPDIR) $(CCDEPMODE) $(depcomp) @AMDEPBACKSLASH@ @am__fastdepCC_FALSE@ $(COMPILE) -c `$(CYGPATH_W) '$<'` .c.lo: @am__fastdepCC_TRUE@ $(LTCOMPILE) -MT $@ -MD -MP -MF $(DEPDIR)/$*.Tpo -c -o $@ $< @am__fastdepCC_TRUE@ $(am__mv) $(DEPDIR)/$*.Tpo $(DEPDIR)/$*.Plo @AMDEP_TRUE@@am__fastdepCC_FALSE@ source='$<' object='$@' libtool=yes @AMDEPBACKSLASH@ @AMDEP_TRUE@@am__fastdepCC_FALSE@ DEPDIR=$(DEPDIR) $(CCDEPMODE) $(depcomp) @AMDEPBACKSLASH@ @am__fastdepCC_FALSE@ $(LTCOMPILE) -c -o $@ $< mostlyclean-libtool: -rm -f *.lo clean-libtool: -rm -rf .libs _libs ID: $(HEADERS) $(SOURCES) $(LISP) $(TAGS_FILES) list='$(SOURCES) $(HEADERS) $(LISP) $(TAGS_FILES)'; \ unique=`for i in $$list; do \ if test -f "$$i"; then echo $$i; else echo $(srcdir)/$$i; fi; \ done | \ $(AWK) '{ files[$$0] = 1; nonempty = 1; } \ END { if (nonempty) { for (i in files) print i; }; }'`; \ mkid -fID $$unique tags: TAGS TAGS: $(HEADERS) $(SOURCES) $(TAGS_DEPENDENCIES) \ $(TAGS_FILES) $(LISP) set x; \ here=`pwd`; \ list='$(SOURCES) $(HEADERS) $(LISP) $(TAGS_FILES)'; \ unique=`for i in $$list; do \ if test -f "$$i"; then echo $$i; else echo $(srcdir)/$$i; fi; \ done | \ $(AWK) '{ files[$$0] = 1; nonempty = 1; } \ END { if (nonempty) { for (i in files) print i; }; }'`; \ shift; \ if test -z "$(ETAGS_ARGS)$$*$$unique"; then :; else \ test -n "$$unique" || unique=$$empty_fix; \ if test $$# -gt 0; then \ $(ETAGS) $(ETAGSFLAGS) $(AM_ETAGSFLAGS) $(ETAGS_ARGS) \ "$$@" $$unique; \ else \ $(ETAGS) $(ETAGSFLAGS) $(AM_ETAGSFLAGS) $(ETAGS_ARGS) \ $$unique; \ fi; \ fi ctags: CTAGS CTAGS: $(HEADERS) $(SOURCES) $(TAGS_DEPENDENCIES) \ $(TAGS_FILES) $(LISP) list='$(SOURCES) $(HEADERS) $(LISP) $(TAGS_FILES)'; \ unique=`for i in $$list; do \ if test -f "$$i"; then echo $$i; else echo $(srcdir)/$$i; fi; \ done | \ $(AWK) '{ files[$$0] = 1; nonempty = 1; } \ END { if (nonempty) { for (i in files) print i; }; }'`; \ test -z "$(CTAGS_ARGS)$$unique" \ || $(CTAGS) $(CTAGSFLAGS) $(AM_CTAGSFLAGS) $(CTAGS_ARGS) \ $$unique GTAGS: here=`$(am__cd) $(top_builddir) && pwd` \ && $(am__cd) $(top_srcdir) \ && gtags -i $(GTAGS_ARGS) "$$here" distclean-tags: -rm -f TAGS ID GTAGS GRTAGS GSYMS GPATH tags distdir: $(DISTFILES) @srcdirstrip=`echo "$(srcdir)" | sed 's/[].[^$$\\*]/\\\\&/g'`; \ topsrcdirstrip=`echo "$(top_srcdir)" | sed 's/[].[^$$\\*]/\\\\&/g'`; \ list='$(DISTFILES)'; \ dist_files=`for file in $$list; do echo $$file; done | \ sed -e "s|^$$srcdirstrip/||;t" \ -e "s|^$$topsrcdirstrip/|$(top_builddir)/|;t"`; \ case $$dist_files in \ */*) $(MKDIR_P) `echo "$$dist_files" | \ sed '/\//!d;s|^|$(distdir)/|;s,/[^/]*$$,,' | \ sort -u` ;; \ esac; \ for file in $$dist_files; do \ if test -f $$file || test -d $$file; then d=.; else d=$(srcdir); fi; \ if test -d $$d/$$file; then \ dir=`echo "/$$file" | sed -e 's,/[^/]*$$,,'`; \ if test -d "$(distdir)/$$file"; then \ find "$(distdir)/$$file" -type d ! -perm -700 -exec chmod u+rwx {} \;; \ fi; \ if test -d $(srcdir)/$$file && test $$d != $(srcdir); then \ cp -fpR $(srcdir)/$$file "$(distdir)$$dir" || exit 1; \ find "$(distdir)/$$file" -type d ! -perm -700 -exec chmod u+rwx {} \;; \ fi; \ cp -fpR $$d/$$file "$(distdir)$$dir" || exit 1; \ else \ test -f "$(distdir)/$$file" \ || cp -p $$d/$$file "$(distdir)/$$file" \ || exit 1; \ fi; \ done check-am: all-am check: $(BUILT_SOURCES) $(MAKE) $(AM_MAKEFLAGS) check-am all-am: Makefile $(LTLIBRARIES) installdirs: install: $(BUILT_SOURCES) $(MAKE) $(AM_MAKEFLAGS) install-am install-exec: install-exec-am install-data: install-data-am uninstall: uninstall-am install-am: all-am @$(MAKE) $(AM_MAKEFLAGS) install-exec-am install-data-am installcheck: installcheck-am install-strip: if test -z '$(STRIP)'; then \ $(MAKE) $(AM_MAKEFLAGS) INSTALL_PROGRAM="$(INSTALL_STRIP_PROGRAM)" \ install_sh_PROGRAM="$(INSTALL_STRIP_PROGRAM)" INSTALL_STRIP_FLAG=-s \ install; \ else \ $(MAKE) $(AM_MAKEFLAGS) INSTALL_PROGRAM="$(INSTALL_STRIP_PROGRAM)" \ install_sh_PROGRAM="$(INSTALL_STRIP_PROGRAM)" INSTALL_STRIP_FLAG=-s \ "INSTALL_PROGRAM_ENV=STRIPPROG='$(STRIP)'" install; \ fi mostlyclean-generic: clean-generic: distclean-generic: -test -z "$(CONFIG_CLEAN_FILES)" || rm -f $(CONFIG_CLEAN_FILES) -test . = "$(srcdir)" || test -z "$(CONFIG_CLEAN_VPATH_FILES)" || rm -f $(CONFIG_CLEAN_VPATH_FILES) maintainer-clean-generic: @echo "This command is intended for maintainers to use" @echo "it deletes files that may require special tools to rebuild." -test -z "$(BUILT_SOURCES)" || rm -f $(BUILT_SOURCES) clean: clean-am clean-am: clean-generic clean-libtool clean-noinstLTLIBRARIES \ mostlyclean-am distclean: distclean-am -rm -rf ./$(DEPDIR) -rm -f Makefile distclean-am: clean-am distclean-compile distclean-generic \ distclean-tags dvi: dvi-am dvi-am: html: html-am html-am: info: info-am info-am: install-data-am: install-dvi: install-dvi-am install-dvi-am: install-exec-am: install-html: install-html-am install-html-am: install-info: install-info-am install-info-am: install-man: install-pdf: install-pdf-am install-pdf-am: install-ps: install-ps-am install-ps-am: installcheck-am: maintainer-clean: maintainer-clean-am -rm -rf ./$(DEPDIR) -rm -f Makefile maintainer-clean-am: distclean-am maintainer-clean-generic mostlyclean: mostlyclean-am mostlyclean-am: mostlyclean-compile mostlyclean-generic \ mostlyclean-libtool pdf: pdf-am pdf-am: ps: ps-am ps-am: uninstall-am: .MAKE: all check install install-am install-strip .PHONY: CTAGS GTAGS all all-am check check-am clean clean-generic \ clean-libtool clean-noinstLTLIBRARIES ctags distclean \ distclean-compile distclean-generic distclean-libtool \ distclean-tags distdir dvi dvi-am html html-am info info-am \ install install-am install-data install-data-am install-dvi \ install-dvi-am install-exec install-exec-am install-html \ install-html-am install-info install-info-am install-man \ install-pdf install-pdf-am install-ps install-ps-am \ install-strip installcheck installcheck-am installdirs \ maintainer-clean maintainer-clean-generic mostlyclean \ mostlyclean-compile mostlyclean-generic mostlyclean-libtool \ pdf pdf-am ps ps-am tags uninstall uninstall-am $(EXTRA_DIST): Makefile ( \ echo "/* Generated automatically. DO NOT EDIT! */"; \ echo "#define SIMD_HEADER \"$(SIMD_HEADER)\""; \ echo "#include \"../common/"$*".c\""; \ ) >$@ # Tell versions [3.59,3.63) of GNU make to not export all variables. # Otherwise a system limit (for SysV at least) may be exceeded. .NOEXPORT: fftw-3.3.3/rdft/simd/sse2/hc2cfdftv_2.c0000644000175000017500000000016212054411072014407 00000000000000/* Generated automatically. DO NOT EDIT! */ #define SIMD_HEADER "simd-sse2.h" #include "../common/hc2cfdftv_2.c" fftw-3.3.3/rdft/simd/sse2/hc2cfdftv_4.c0000644000175000017500000000016212054411072014411 00000000000000/* Generated automatically. DO NOT EDIT! */ #define SIMD_HEADER "simd-sse2.h" #include "../common/hc2cfdftv_4.c" fftw-3.3.3/rdft/simd/sse2/hc2cfdftv_6.c0000644000175000017500000000016212054411072014413 00000000000000/* Generated automatically. DO NOT EDIT! */ #define SIMD_HEADER "simd-sse2.h" #include "../common/hc2cfdftv_6.c" fftw-3.3.3/rdft/simd/sse2/hc2cfdftv_8.c0000644000175000017500000000016212054411072014415 00000000000000/* Generated automatically. DO NOT EDIT! */ #define SIMD_HEADER "simd-sse2.h" #include "../common/hc2cfdftv_8.c" fftw-3.3.3/rdft/simd/sse2/hc2cfdftv_10.c0000644000175000017500000000016312054411072014467 00000000000000/* Generated automatically. DO NOT EDIT! */ #define SIMD_HEADER "simd-sse2.h" #include "../common/hc2cfdftv_10.c" fftw-3.3.3/rdft/simd/sse2/hc2cfdftv_12.c0000644000175000017500000000016312054411072014471 00000000000000/* Generated automatically. DO NOT EDIT! */ #define SIMD_HEADER "simd-sse2.h" #include "../common/hc2cfdftv_12.c" fftw-3.3.3/rdft/simd/sse2/hc2cfdftv_16.c0000644000175000017500000000016312054411072014475 00000000000000/* Generated automatically. DO NOT EDIT! */ #define SIMD_HEADER "simd-sse2.h" #include "../common/hc2cfdftv_16.c" fftw-3.3.3/rdft/simd/sse2/hc2cfdftv_32.c0000644000175000017500000000016312054411072014473 00000000000000/* Generated automatically. DO NOT EDIT! */ #define SIMD_HEADER "simd-sse2.h" #include "../common/hc2cfdftv_32.c" fftw-3.3.3/rdft/simd/sse2/hc2cfdftv_20.c0000644000175000017500000000016312054411072014470 00000000000000/* Generated automatically. DO NOT EDIT! */ #define SIMD_HEADER "simd-sse2.h" #include "../common/hc2cfdftv_20.c" fftw-3.3.3/rdft/simd/sse2/hc2cbdftv_2.c0000644000175000017500000000016212054411072014403 00000000000000/* Generated automatically. DO NOT EDIT! */ #define SIMD_HEADER "simd-sse2.h" #include "../common/hc2cbdftv_2.c" fftw-3.3.3/rdft/simd/sse2/hc2cbdftv_4.c0000644000175000017500000000016212054411072014405 00000000000000/* Generated automatically. DO NOT EDIT! */ #define SIMD_HEADER "simd-sse2.h" #include "../common/hc2cbdftv_4.c" fftw-3.3.3/rdft/simd/sse2/hc2cbdftv_6.c0000644000175000017500000000016212054411072014407 00000000000000/* Generated automatically. DO NOT EDIT! */ #define SIMD_HEADER "simd-sse2.h" #include "../common/hc2cbdftv_6.c" fftw-3.3.3/rdft/simd/sse2/hc2cbdftv_8.c0000644000175000017500000000016212054411072014411 00000000000000/* Generated automatically. DO NOT EDIT! */ #define SIMD_HEADER "simd-sse2.h" #include "../common/hc2cbdftv_8.c" fftw-3.3.3/rdft/simd/sse2/hc2cbdftv_10.c0000644000175000017500000000016312054411072014463 00000000000000/* Generated automatically. DO NOT EDIT! */ #define SIMD_HEADER "simd-sse2.h" #include "../common/hc2cbdftv_10.c" fftw-3.3.3/rdft/simd/sse2/hc2cbdftv_12.c0000644000175000017500000000016312054411072014465 00000000000000/* Generated automatically. DO NOT EDIT! */ #define SIMD_HEADER "simd-sse2.h" #include "../common/hc2cbdftv_12.c" fftw-3.3.3/rdft/simd/sse2/hc2cbdftv_16.c0000644000175000017500000000016312054411072014471 00000000000000/* Generated automatically. DO NOT EDIT! */ #define SIMD_HEADER "simd-sse2.h" #include "../common/hc2cbdftv_16.c" fftw-3.3.3/rdft/simd/sse2/hc2cbdftv_32.c0000644000175000017500000000016312054411072014467 00000000000000/* Generated automatically. DO NOT EDIT! */ #define SIMD_HEADER "simd-sse2.h" #include "../common/hc2cbdftv_32.c" fftw-3.3.3/rdft/simd/sse2/hc2cbdftv_20.c0000644000175000017500000000016312054411072014464 00000000000000/* Generated automatically. DO NOT EDIT! */ #define SIMD_HEADER "simd-sse2.h" #include "../common/hc2cbdftv_20.c" fftw-3.3.3/rdft/simd/sse2/genus.c0000644000175000017500000000015412054411072013437 00000000000000/* Generated automatically. DO NOT EDIT! */ #define SIMD_HEADER "simd-sse2.h" #include "../common/genus.c" fftw-3.3.3/rdft/simd/sse2/codlist.c0000644000175000017500000000015612054411072013761 00000000000000/* Generated automatically. DO NOT EDIT! */ #define SIMD_HEADER "simd-sse2.h" #include "../common/codlist.c" fftw-3.3.3/rdft/simd/avx/0000755000175000017500000000000012054411110012145 500000000000000fftw-3.3.3/rdft/simd/avx/Makefile.am0000644000175000017500000000043212054410050014122 00000000000000AM_CFLAGS = $(AVX_CFLAGS) SIMD_HEADER=simd-avx.h include $(top_srcdir)/rdft/simd/codlist.mk include $(top_srcdir)/rdft/simd/simd.mk if HAVE_AVX noinst_LTLIBRARIES = librdft_avx_codelets.la BUILT_SOURCES = $(EXTRA_DIST) librdft_avx_codelets_la_SOURCES = $(BUILT_SOURCES) endif fftw-3.3.3/rdft/simd/avx/Makefile.in0000644000175000017500000005004212054410200014132 00000000000000# Makefile.in generated by automake 1.11.6 from Makefile.am. # @configure_input@ # Copyright (C) 1994, 1995, 1996, 1997, 1998, 1999, 2000, 2001, 2002, # 2003, 2004, 2005, 2006, 2007, 2008, 2009, 2010, 2011 Free Software # Foundation, Inc. # This Makefile.in is free software; the Free Software Foundation # gives unlimited permission to copy and/or distribute it, # with or without modifications, as long as this notice is preserved. # This program is distributed in the hope that it will be useful, # but WITHOUT ANY WARRANTY, to the extent permitted by law; without # even the implied warranty of MERCHANTABILITY or FITNESS FOR A # PARTICULAR PURPOSE. @SET_MAKE@ # This file contains a standard list of RDFT SIMD codelets. It is # included by common/Makefile to generate the C files with the actual # codelets in them. It is included by {sse,sse2,...}/Makefile to # generate and compile stub files that include common/*.c # You can customize FFTW for special needs, e.g. to handle certain # sizes more efficiently, by adding new codelets to the lists of those # included by default. If you change the list of codelets, any new # ones you added will be automatically generated when you run the # bootstrap script (see "Generating your own code" in the FFTW # manual). VPATH = @srcdir@ am__make_dryrun = \ { \ am__dry=no; \ case $$MAKEFLAGS in \ *\\[\ \ ]*) \ echo 'am--echo: ; @echo "AM" OK' | $(MAKE) -f - 2>/dev/null \ | grep '^AM OK$$' >/dev/null || am__dry=yes;; \ *) \ for am__flg in $$MAKEFLAGS; do \ case $$am__flg in \ *=*|--*) ;; \ *n*) am__dry=yes; break;; \ esac; \ done;; \ esac; \ test $$am__dry = yes; \ } pkgdatadir = $(datadir)/@PACKAGE@ pkgincludedir = $(includedir)/@PACKAGE@ pkglibdir = $(libdir)/@PACKAGE@ pkglibexecdir = $(libexecdir)/@PACKAGE@ am__cd = CDPATH="$${ZSH_VERSION+.}$(PATH_SEPARATOR)" && cd install_sh_DATA = $(install_sh) -c -m 644 install_sh_PROGRAM = $(install_sh) -c install_sh_SCRIPT = $(install_sh) -c INSTALL_HEADER = $(INSTALL_DATA) transform = $(program_transform_name) NORMAL_INSTALL = : PRE_INSTALL = : POST_INSTALL = : NORMAL_UNINSTALL = : PRE_UNINSTALL = : POST_UNINSTALL = : build_triplet = @build@ host_triplet = @host@ DIST_COMMON = $(srcdir)/Makefile.am $(srcdir)/Makefile.in \ $(top_srcdir)/rdft/simd/codlist.mk \ $(top_srcdir)/rdft/simd/simd.mk subdir = rdft/simd/avx ACLOCAL_M4 = $(top_srcdir)/aclocal.m4 am__aclocal_m4_deps = $(top_srcdir)/m4/acx_mpi.m4 \ $(top_srcdir)/m4/acx_pthread.m4 \ $(top_srcdir)/m4/ax_cc_maxopt.m4 \ $(top_srcdir)/m4/ax_check_compiler_flags.m4 \ $(top_srcdir)/m4/ax_compiler_vendor.m4 \ $(top_srcdir)/m4/ax_gcc_aligns_stack.m4 \ $(top_srcdir)/m4/ax_gcc_version.m4 \ $(top_srcdir)/m4/ax_openmp.m4 $(top_srcdir)/m4/libtool.m4 \ $(top_srcdir)/m4/ltoptions.m4 $(top_srcdir)/m4/ltsugar.m4 \ $(top_srcdir)/m4/ltversion.m4 $(top_srcdir)/m4/lt~obsolete.m4 \ $(top_srcdir)/configure.ac am__configure_deps = $(am__aclocal_m4_deps) $(CONFIGURE_DEPENDENCIES) \ $(ACLOCAL_M4) mkinstalldirs = $(install_sh) -d CONFIG_HEADER = $(top_builddir)/config.h CONFIG_CLEAN_FILES = CONFIG_CLEAN_VPATH_FILES = LTLIBRARIES = $(noinst_LTLIBRARIES) librdft_avx_codelets_la_LIBADD = am__librdft_avx_codelets_la_SOURCES_DIST = hc2cfdftv_2.c hc2cfdftv_4.c \ hc2cfdftv_6.c hc2cfdftv_8.c hc2cfdftv_10.c hc2cfdftv_12.c \ hc2cfdftv_16.c hc2cfdftv_32.c hc2cfdftv_20.c hc2cbdftv_2.c \ hc2cbdftv_4.c hc2cbdftv_6.c hc2cbdftv_8.c hc2cbdftv_10.c \ hc2cbdftv_12.c hc2cbdftv_16.c hc2cbdftv_32.c hc2cbdftv_20.c \ genus.c codlist.c am__objects_1 = hc2cfdftv_2.lo hc2cfdftv_4.lo hc2cfdftv_6.lo \ hc2cfdftv_8.lo hc2cfdftv_10.lo hc2cfdftv_12.lo hc2cfdftv_16.lo \ hc2cfdftv_32.lo hc2cfdftv_20.lo am__objects_2 = hc2cbdftv_2.lo hc2cbdftv_4.lo hc2cbdftv_6.lo \ hc2cbdftv_8.lo hc2cbdftv_10.lo hc2cbdftv_12.lo hc2cbdftv_16.lo \ hc2cbdftv_32.lo hc2cbdftv_20.lo am__objects_3 = $(am__objects_1) $(am__objects_2) am__objects_4 = $(am__objects_3) genus.lo codlist.lo @HAVE_AVX_TRUE@am__objects_5 = $(am__objects_4) @HAVE_AVX_TRUE@am_librdft_avx_codelets_la_OBJECTS = $(am__objects_5) librdft_avx_codelets_la_OBJECTS = \ $(am_librdft_avx_codelets_la_OBJECTS) @HAVE_AVX_TRUE@am_librdft_avx_codelets_la_rpath = DEFAULT_INCLUDES = -I.@am__isrc@ -I$(top_builddir) depcomp = $(SHELL) $(top_srcdir)/depcomp am__depfiles_maybe = depfiles am__mv = mv -f COMPILE = $(CC) $(DEFS) $(DEFAULT_INCLUDES) $(INCLUDES) $(AM_CPPFLAGS) \ $(CPPFLAGS) $(AM_CFLAGS) $(CFLAGS) LTCOMPILE = $(LIBTOOL) --tag=CC $(AM_LIBTOOLFLAGS) $(LIBTOOLFLAGS) \ --mode=compile $(CC) $(DEFS) $(DEFAULT_INCLUDES) $(INCLUDES) \ $(AM_CPPFLAGS) $(CPPFLAGS) $(AM_CFLAGS) $(CFLAGS) CCLD = $(CC) LINK = $(LIBTOOL) --tag=CC $(AM_LIBTOOLFLAGS) $(LIBTOOLFLAGS) \ --mode=link $(CCLD) $(AM_CFLAGS) $(CFLAGS) $(AM_LDFLAGS) \ $(LDFLAGS) -o $@ SOURCES = $(librdft_avx_codelets_la_SOURCES) DIST_SOURCES = $(am__librdft_avx_codelets_la_SOURCES_DIST) am__can_run_installinfo = \ case $$AM_UPDATE_INFO_DIR in \ n|no|NO) false;; \ *) (install-info --version) >/dev/null 2>&1;; \ esac ETAGS = etags CTAGS = ctags DISTFILES = $(DIST_COMMON) $(DIST_SOURCES) $(TEXINFOS) $(EXTRA_DIST) ACLOCAL = @ACLOCAL@ ALLOCA = @ALLOCA@ ALTIVEC_CFLAGS = @ALTIVEC_CFLAGS@ AMTAR = @AMTAR@ AR = @AR@ AS = @AS@ AUTOCONF = @AUTOCONF@ AUTOHEADER = @AUTOHEADER@ AUTOMAKE = @AUTOMAKE@ AVX_CFLAGS = @AVX_CFLAGS@ AWK = @AWK@ CC = @CC@ CCDEPMODE = @CCDEPMODE@ CFLAGS = @CFLAGS@ CHECK_PL_OPTS = @CHECK_PL_OPTS@ CPP = @CPP@ CPPFLAGS = @CPPFLAGS@ CYGPATH_W = @CYGPATH_W@ C_FFTW_R2R_KIND = @C_FFTW_R2R_KIND@ C_MPI_FINT = @C_MPI_FINT@ DEFS = @DEFS@ DEPDIR = @DEPDIR@ DLLTOOL = @DLLTOOL@ DSYMUTIL = @DSYMUTIL@ DUMPBIN = @DUMPBIN@ ECHO_C = @ECHO_C@ ECHO_N = @ECHO_N@ ECHO_T = @ECHO_T@ EGREP = @EGREP@ EXEEXT = @EXEEXT@ F77 = @F77@ FFLAGS = @FFLAGS@ FGREP = @FGREP@ FLIBS = @FLIBS@ GREP = @GREP@ INSTALL = @INSTALL@ INSTALL_DATA = @INSTALL_DATA@ INSTALL_PROGRAM = @INSTALL_PROGRAM@ INSTALL_SCRIPT = @INSTALL_SCRIPT@ INSTALL_STRIP_PROGRAM = @INSTALL_STRIP_PROGRAM@ LD = @LD@ LDFLAGS = @LDFLAGS@ LIBOBJS = @LIBOBJS@ LIBQUADMATH = @LIBQUADMATH@ LIBS = @LIBS@ LIBTOOL = @LIBTOOL@ LIPO = @LIPO@ LN_S = @LN_S@ LTLIBOBJS = @LTLIBOBJS@ MAINT = @MAINT@ MAKEINFO = @MAKEINFO@ MANIFEST_TOOL = @MANIFEST_TOOL@ MKDIR_P = @MKDIR_P@ MPICC = @MPICC@ MPILIBS = @MPILIBS@ MPIRUN = @MPIRUN@ NEON_CFLAGS = @NEON_CFLAGS@ NM = @NM@ NMEDIT = @NMEDIT@ OBJDUMP = @OBJDUMP@ OBJEXT = @OBJEXT@ OCAMLBUILD = @OCAMLBUILD@ OPENMP_CFLAGS = @OPENMP_CFLAGS@ OTOOL = @OTOOL@ OTOOL64 = @OTOOL64@ PACKAGE = @PACKAGE@ PACKAGE_BUGREPORT = @PACKAGE_BUGREPORT@ PACKAGE_NAME = @PACKAGE_NAME@ PACKAGE_STRING = @PACKAGE_STRING@ PACKAGE_TARNAME = @PACKAGE_TARNAME@ PACKAGE_URL = @PACKAGE_URL@ PACKAGE_VERSION = @PACKAGE_VERSION@ PATH_SEPARATOR = @PATH_SEPARATOR@ POW_LIB = @POW_LIB@ PRECISION = @PRECISION@ PREC_SUFFIX = @PREC_SUFFIX@ PTHREAD_CC = @PTHREAD_CC@ PTHREAD_CFLAGS = @PTHREAD_CFLAGS@ PTHREAD_LIBS = @PTHREAD_LIBS@ RANLIB = @RANLIB@ SED = @SED@ SET_MAKE = @SET_MAKE@ SHARED_VERSION_INFO = @SHARED_VERSION_INFO@ SHELL = @SHELL@ SSE2_CFLAGS = @SSE2_CFLAGS@ STACK_ALIGN_CFLAGS = @STACK_ALIGN_CFLAGS@ STRIP = @STRIP@ THREADLIBS = @THREADLIBS@ VERSION = @VERSION@ abs_builddir = @abs_builddir@ abs_srcdir = @abs_srcdir@ abs_top_builddir = @abs_top_builddir@ abs_top_srcdir = @abs_top_srcdir@ ac_ct_AR = @ac_ct_AR@ ac_ct_CC = @ac_ct_CC@ ac_ct_DUMPBIN = @ac_ct_DUMPBIN@ ac_ct_F77 = @ac_ct_F77@ acx_pthread_config = @acx_pthread_config@ am__include = @am__include@ am__leading_dot = @am__leading_dot@ am__quote = @am__quote@ am__tar = @am__tar@ am__untar = @am__untar@ bindir = @bindir@ build = @build@ build_alias = @build_alias@ build_cpu = @build_cpu@ build_os = @build_os@ build_vendor = @build_vendor@ builddir = @builddir@ datadir = @datadir@ datarootdir = @datarootdir@ docdir = @docdir@ dvidir = @dvidir@ exec_prefix = @exec_prefix@ host = @host@ host_alias = @host_alias@ host_cpu = @host_cpu@ host_os = @host_os@ host_vendor = @host_vendor@ htmldir = @htmldir@ includedir = @includedir@ infodir = @infodir@ install_sh = @install_sh@ libdir = @libdir@ libexecdir = @libexecdir@ localedir = @localedir@ localstatedir = @localstatedir@ mandir = @mandir@ mkdir_p = @mkdir_p@ oldincludedir = @oldincludedir@ pdfdir = @pdfdir@ prefix = @prefix@ program_transform_name = @program_transform_name@ psdir = @psdir@ sbindir = @sbindir@ sharedstatedir = @sharedstatedir@ srcdir = @srcdir@ sysconfdir = @sysconfdir@ target_alias = @target_alias@ top_build_prefix = @top_build_prefix@ top_builddir = @top_builddir@ top_srcdir = @top_srcdir@ AM_CFLAGS = $(AVX_CFLAGS) SIMD_HEADER = simd-avx.h HC2CFDFTV = hc2cfdftv_2.c hc2cfdftv_4.c hc2cfdftv_6.c hc2cfdftv_8.c \ hc2cfdftv_10.c hc2cfdftv_12.c hc2cfdftv_16.c hc2cfdftv_32.c \ hc2cfdftv_20.c HC2CBDFTV = hc2cbdftv_2.c hc2cbdftv_4.c hc2cbdftv_6.c hc2cbdftv_8.c \ hc2cbdftv_10.c hc2cbdftv_12.c hc2cbdftv_16.c hc2cbdftv_32.c \ hc2cbdftv_20.c ########################################################################### SIMD_CODELETS = $(HC2CFDFTV) $(HC2CBDFTV) AM_CPPFLAGS = -I$(top_srcdir)/kernel -I$(top_srcdir)/rdft \ -I$(top_srcdir)/rdft/simd -I$(top_srcdir)/simd-support EXTRA_DIST = $(SIMD_CODELETS) genus.c codlist.c @HAVE_AVX_TRUE@noinst_LTLIBRARIES = librdft_avx_codelets.la @HAVE_AVX_TRUE@BUILT_SOURCES = $(EXTRA_DIST) @HAVE_AVX_TRUE@librdft_avx_codelets_la_SOURCES = $(BUILT_SOURCES) all: $(BUILT_SOURCES) $(MAKE) $(AM_MAKEFLAGS) all-am .SUFFIXES: .SUFFIXES: .c .lo .o .obj $(srcdir)/Makefile.in: @MAINTAINER_MODE_TRUE@ $(srcdir)/Makefile.am $(top_srcdir)/rdft/simd/codlist.mk $(top_srcdir)/rdft/simd/simd.mk $(am__configure_deps) @for dep in $?; do \ case '$(am__configure_deps)' in \ *$$dep*) \ ( cd $(top_builddir) && $(MAKE) $(AM_MAKEFLAGS) am--refresh ) \ && { if test -f $@; then exit 0; else break; fi; }; \ exit 1;; \ esac; \ done; \ echo ' cd $(top_srcdir) && $(AUTOMAKE) --gnu rdft/simd/avx/Makefile'; \ $(am__cd) $(top_srcdir) && \ $(AUTOMAKE) --gnu rdft/simd/avx/Makefile .PRECIOUS: Makefile Makefile: $(srcdir)/Makefile.in $(top_builddir)/config.status @case '$?' in \ *config.status*) \ cd $(top_builddir) && $(MAKE) $(AM_MAKEFLAGS) am--refresh;; \ *) \ echo ' cd $(top_builddir) && $(SHELL) ./config.status $(subdir)/$@ $(am__depfiles_maybe)'; \ cd $(top_builddir) && $(SHELL) ./config.status $(subdir)/$@ $(am__depfiles_maybe);; \ esac; $(top_srcdir)/rdft/simd/codlist.mk $(top_srcdir)/rdft/simd/simd.mk: $(top_builddir)/config.status: $(top_srcdir)/configure $(CONFIG_STATUS_DEPENDENCIES) cd $(top_builddir) && $(MAKE) $(AM_MAKEFLAGS) am--refresh $(top_srcdir)/configure: @MAINTAINER_MODE_TRUE@ $(am__configure_deps) cd $(top_builddir) && $(MAKE) $(AM_MAKEFLAGS) am--refresh $(ACLOCAL_M4): @MAINTAINER_MODE_TRUE@ $(am__aclocal_m4_deps) cd $(top_builddir) && $(MAKE) $(AM_MAKEFLAGS) am--refresh $(am__aclocal_m4_deps): clean-noinstLTLIBRARIES: -test -z "$(noinst_LTLIBRARIES)" || rm -f $(noinst_LTLIBRARIES) @list='$(noinst_LTLIBRARIES)'; for p in $$list; do \ dir="`echo $$p | sed -e 's|/[^/]*$$||'`"; \ test "$$dir" != "$$p" || dir=.; \ echo "rm -f \"$${dir}/so_locations\""; \ rm -f "$${dir}/so_locations"; \ done librdft_avx_codelets.la: $(librdft_avx_codelets_la_OBJECTS) $(librdft_avx_codelets_la_DEPENDENCIES) $(EXTRA_librdft_avx_codelets_la_DEPENDENCIES) $(LINK) $(am_librdft_avx_codelets_la_rpath) $(librdft_avx_codelets_la_OBJECTS) $(librdft_avx_codelets_la_LIBADD) $(LIBS) mostlyclean-compile: -rm -f *.$(OBJEXT) distclean-compile: -rm -f *.tab.c @AMDEP_TRUE@@am__include@ @am__quote@./$(DEPDIR)/codlist.Plo@am__quote@ @AMDEP_TRUE@@am__include@ @am__quote@./$(DEPDIR)/genus.Plo@am__quote@ @AMDEP_TRUE@@am__include@ @am__quote@./$(DEPDIR)/hc2cbdftv_10.Plo@am__quote@ @AMDEP_TRUE@@am__include@ @am__quote@./$(DEPDIR)/hc2cbdftv_12.Plo@am__quote@ @AMDEP_TRUE@@am__include@ @am__quote@./$(DEPDIR)/hc2cbdftv_16.Plo@am__quote@ @AMDEP_TRUE@@am__include@ @am__quote@./$(DEPDIR)/hc2cbdftv_2.Plo@am__quote@ @AMDEP_TRUE@@am__include@ @am__quote@./$(DEPDIR)/hc2cbdftv_20.Plo@am__quote@ @AMDEP_TRUE@@am__include@ @am__quote@./$(DEPDIR)/hc2cbdftv_32.Plo@am__quote@ @AMDEP_TRUE@@am__include@ @am__quote@./$(DEPDIR)/hc2cbdftv_4.Plo@am__quote@ @AMDEP_TRUE@@am__include@ @am__quote@./$(DEPDIR)/hc2cbdftv_6.Plo@am__quote@ @AMDEP_TRUE@@am__include@ @am__quote@./$(DEPDIR)/hc2cbdftv_8.Plo@am__quote@ @AMDEP_TRUE@@am__include@ @am__quote@./$(DEPDIR)/hc2cfdftv_10.Plo@am__quote@ @AMDEP_TRUE@@am__include@ @am__quote@./$(DEPDIR)/hc2cfdftv_12.Plo@am__quote@ @AMDEP_TRUE@@am__include@ @am__quote@./$(DEPDIR)/hc2cfdftv_16.Plo@am__quote@ @AMDEP_TRUE@@am__include@ @am__quote@./$(DEPDIR)/hc2cfdftv_2.Plo@am__quote@ @AMDEP_TRUE@@am__include@ @am__quote@./$(DEPDIR)/hc2cfdftv_20.Plo@am__quote@ @AMDEP_TRUE@@am__include@ @am__quote@./$(DEPDIR)/hc2cfdftv_32.Plo@am__quote@ @AMDEP_TRUE@@am__include@ @am__quote@./$(DEPDIR)/hc2cfdftv_4.Plo@am__quote@ @AMDEP_TRUE@@am__include@ @am__quote@./$(DEPDIR)/hc2cfdftv_6.Plo@am__quote@ @AMDEP_TRUE@@am__include@ @am__quote@./$(DEPDIR)/hc2cfdftv_8.Plo@am__quote@ .c.o: @am__fastdepCC_TRUE@ $(COMPILE) -MT $@ -MD -MP -MF $(DEPDIR)/$*.Tpo -c -o $@ $< @am__fastdepCC_TRUE@ $(am__mv) $(DEPDIR)/$*.Tpo $(DEPDIR)/$*.Po @AMDEP_TRUE@@am__fastdepCC_FALSE@ source='$<' object='$@' libtool=no @AMDEPBACKSLASH@ @AMDEP_TRUE@@am__fastdepCC_FALSE@ DEPDIR=$(DEPDIR) $(CCDEPMODE) $(depcomp) @AMDEPBACKSLASH@ @am__fastdepCC_FALSE@ $(COMPILE) -c $< .c.obj: @am__fastdepCC_TRUE@ $(COMPILE) -MT $@ -MD -MP -MF $(DEPDIR)/$*.Tpo -c -o $@ `$(CYGPATH_W) '$<'` @am__fastdepCC_TRUE@ $(am__mv) $(DEPDIR)/$*.Tpo $(DEPDIR)/$*.Po @AMDEP_TRUE@@am__fastdepCC_FALSE@ source='$<' object='$@' libtool=no @AMDEPBACKSLASH@ @AMDEP_TRUE@@am__fastdepCC_FALSE@ DEPDIR=$(DEPDIR) $(CCDEPMODE) $(depcomp) @AMDEPBACKSLASH@ @am__fastdepCC_FALSE@ $(COMPILE) -c `$(CYGPATH_W) '$<'` .c.lo: @am__fastdepCC_TRUE@ $(LTCOMPILE) -MT $@ -MD -MP -MF $(DEPDIR)/$*.Tpo -c -o $@ $< @am__fastdepCC_TRUE@ $(am__mv) $(DEPDIR)/$*.Tpo $(DEPDIR)/$*.Plo @AMDEP_TRUE@@am__fastdepCC_FALSE@ source='$<' object='$@' libtool=yes @AMDEPBACKSLASH@ @AMDEP_TRUE@@am__fastdepCC_FALSE@ DEPDIR=$(DEPDIR) $(CCDEPMODE) $(depcomp) @AMDEPBACKSLASH@ @am__fastdepCC_FALSE@ $(LTCOMPILE) -c -o $@ $< mostlyclean-libtool: -rm -f *.lo clean-libtool: -rm -rf .libs _libs ID: $(HEADERS) $(SOURCES) $(LISP) $(TAGS_FILES) list='$(SOURCES) $(HEADERS) $(LISP) $(TAGS_FILES)'; \ unique=`for i in $$list; do \ if test -f "$$i"; then echo $$i; else echo $(srcdir)/$$i; fi; \ done | \ $(AWK) '{ files[$$0] = 1; nonempty = 1; } \ END { if (nonempty) { for (i in files) print i; }; }'`; \ mkid -fID $$unique tags: TAGS TAGS: $(HEADERS) $(SOURCES) $(TAGS_DEPENDENCIES) \ $(TAGS_FILES) $(LISP) set x; \ here=`pwd`; \ list='$(SOURCES) $(HEADERS) $(LISP) $(TAGS_FILES)'; \ unique=`for i in $$list; do \ if test -f "$$i"; then echo $$i; else echo $(srcdir)/$$i; fi; \ done | \ $(AWK) '{ files[$$0] = 1; nonempty = 1; } \ END { if (nonempty) { for (i in files) print i; }; }'`; \ shift; \ if test -z "$(ETAGS_ARGS)$$*$$unique"; then :; else \ test -n "$$unique" || unique=$$empty_fix; \ if test $$# -gt 0; then \ $(ETAGS) $(ETAGSFLAGS) $(AM_ETAGSFLAGS) $(ETAGS_ARGS) \ "$$@" $$unique; \ else \ $(ETAGS) $(ETAGSFLAGS) $(AM_ETAGSFLAGS) $(ETAGS_ARGS) \ $$unique; \ fi; \ fi ctags: CTAGS CTAGS: $(HEADERS) $(SOURCES) $(TAGS_DEPENDENCIES) \ $(TAGS_FILES) $(LISP) list='$(SOURCES) $(HEADERS) $(LISP) $(TAGS_FILES)'; \ unique=`for i in $$list; do \ if test -f "$$i"; then echo $$i; else echo $(srcdir)/$$i; fi; \ done | \ $(AWK) '{ files[$$0] = 1; nonempty = 1; } \ END { if (nonempty) { for (i in files) print i; }; }'`; \ test -z "$(CTAGS_ARGS)$$unique" \ || $(CTAGS) $(CTAGSFLAGS) $(AM_CTAGSFLAGS) $(CTAGS_ARGS) \ $$unique GTAGS: here=`$(am__cd) $(top_builddir) && pwd` \ && $(am__cd) $(top_srcdir) \ && gtags -i $(GTAGS_ARGS) "$$here" distclean-tags: -rm -f TAGS ID GTAGS GRTAGS GSYMS GPATH tags distdir: $(DISTFILES) @srcdirstrip=`echo "$(srcdir)" | sed 's/[].[^$$\\*]/\\\\&/g'`; \ topsrcdirstrip=`echo "$(top_srcdir)" | sed 's/[].[^$$\\*]/\\\\&/g'`; \ list='$(DISTFILES)'; \ dist_files=`for file in $$list; do echo $$file; done | \ sed -e "s|^$$srcdirstrip/||;t" \ -e "s|^$$topsrcdirstrip/|$(top_builddir)/|;t"`; \ case $$dist_files in \ */*) $(MKDIR_P) `echo "$$dist_files" | \ sed '/\//!d;s|^|$(distdir)/|;s,/[^/]*$$,,' | \ sort -u` ;; \ esac; \ for file in $$dist_files; do \ if test -f $$file || test -d $$file; then d=.; else d=$(srcdir); fi; \ if test -d $$d/$$file; then \ dir=`echo "/$$file" | sed -e 's,/[^/]*$$,,'`; \ if test -d "$(distdir)/$$file"; then \ find "$(distdir)/$$file" -type d ! -perm -700 -exec chmod u+rwx {} \;; \ fi; \ if test -d $(srcdir)/$$file && test $$d != $(srcdir); then \ cp -fpR $(srcdir)/$$file "$(distdir)$$dir" || exit 1; \ find "$(distdir)/$$file" -type d ! -perm -700 -exec chmod u+rwx {} \;; \ fi; \ cp -fpR $$d/$$file "$(distdir)$$dir" || exit 1; \ else \ test -f "$(distdir)/$$file" \ || cp -p $$d/$$file "$(distdir)/$$file" \ || exit 1; \ fi; \ done check-am: all-am check: $(BUILT_SOURCES) $(MAKE) $(AM_MAKEFLAGS) check-am all-am: Makefile $(LTLIBRARIES) installdirs: install: $(BUILT_SOURCES) $(MAKE) $(AM_MAKEFLAGS) install-am install-exec: install-exec-am install-data: install-data-am uninstall: uninstall-am install-am: all-am @$(MAKE) $(AM_MAKEFLAGS) install-exec-am install-data-am installcheck: installcheck-am install-strip: if test -z '$(STRIP)'; then \ $(MAKE) $(AM_MAKEFLAGS) INSTALL_PROGRAM="$(INSTALL_STRIP_PROGRAM)" \ install_sh_PROGRAM="$(INSTALL_STRIP_PROGRAM)" INSTALL_STRIP_FLAG=-s \ install; \ else \ $(MAKE) $(AM_MAKEFLAGS) INSTALL_PROGRAM="$(INSTALL_STRIP_PROGRAM)" \ install_sh_PROGRAM="$(INSTALL_STRIP_PROGRAM)" INSTALL_STRIP_FLAG=-s \ "INSTALL_PROGRAM_ENV=STRIPPROG='$(STRIP)'" install; \ fi mostlyclean-generic: clean-generic: distclean-generic: -test -z "$(CONFIG_CLEAN_FILES)" || rm -f $(CONFIG_CLEAN_FILES) -test . = "$(srcdir)" || test -z "$(CONFIG_CLEAN_VPATH_FILES)" || rm -f $(CONFIG_CLEAN_VPATH_FILES) maintainer-clean-generic: @echo "This command is intended for maintainers to use" @echo "it deletes files that may require special tools to rebuild." -test -z "$(BUILT_SOURCES)" || rm -f $(BUILT_SOURCES) clean: clean-am clean-am: clean-generic clean-libtool clean-noinstLTLIBRARIES \ mostlyclean-am distclean: distclean-am -rm -rf ./$(DEPDIR) -rm -f Makefile distclean-am: clean-am distclean-compile distclean-generic \ distclean-tags dvi: dvi-am dvi-am: html: html-am html-am: info: info-am info-am: install-data-am: install-dvi: install-dvi-am install-dvi-am: install-exec-am: install-html: install-html-am install-html-am: install-info: install-info-am install-info-am: install-man: install-pdf: install-pdf-am install-pdf-am: install-ps: install-ps-am install-ps-am: installcheck-am: maintainer-clean: maintainer-clean-am -rm -rf ./$(DEPDIR) -rm -f Makefile maintainer-clean-am: distclean-am maintainer-clean-generic mostlyclean: mostlyclean-am mostlyclean-am: mostlyclean-compile mostlyclean-generic \ mostlyclean-libtool pdf: pdf-am pdf-am: ps: ps-am ps-am: uninstall-am: .MAKE: all check install install-am install-strip .PHONY: CTAGS GTAGS all all-am check check-am clean clean-generic \ clean-libtool clean-noinstLTLIBRARIES ctags distclean \ distclean-compile distclean-generic distclean-libtool \ distclean-tags distdir dvi dvi-am html html-am info info-am \ install install-am install-data install-data-am install-dvi \ install-dvi-am install-exec install-exec-am install-html \ install-html-am install-info install-info-am install-man \ install-pdf install-pdf-am install-ps install-ps-am \ install-strip installcheck installcheck-am installdirs \ maintainer-clean maintainer-clean-generic mostlyclean \ mostlyclean-compile mostlyclean-generic mostlyclean-libtool \ pdf pdf-am ps ps-am tags uninstall uninstall-am $(EXTRA_DIST): Makefile ( \ echo "/* Generated automatically. DO NOT EDIT! */"; \ echo "#define SIMD_HEADER \"$(SIMD_HEADER)\""; \ echo "#include \"../common/"$*".c\""; \ ) >$@ # Tell versions [3.59,3.63) of GNU make to not export all variables. # Otherwise a system limit (for SysV at least) may be exceeded. .NOEXPORT: fftw-3.3.3/rdft/simd/avx/hc2cfdftv_2.c0000644000175000017500000000016112054411110014321 00000000000000/* Generated automatically. DO NOT EDIT! */ #define SIMD_HEADER "simd-avx.h" #include "../common/hc2cfdftv_2.c" fftw-3.3.3/rdft/simd/avx/hc2cfdftv_4.c0000644000175000017500000000016112054411110014323 00000000000000/* Generated automatically. DO NOT EDIT! */ #define SIMD_HEADER "simd-avx.h" #include "../common/hc2cfdftv_4.c" fftw-3.3.3/rdft/simd/avx/hc2cfdftv_6.c0000644000175000017500000000016112054411110014325 00000000000000/* Generated automatically. DO NOT EDIT! */ #define SIMD_HEADER "simd-avx.h" #include "../common/hc2cfdftv_6.c" fftw-3.3.3/rdft/simd/avx/hc2cfdftv_8.c0000644000175000017500000000016112054411110014327 00000000000000/* Generated automatically. DO NOT EDIT! */ #define SIMD_HEADER "simd-avx.h" #include "../common/hc2cfdftv_8.c" fftw-3.3.3/rdft/simd/avx/hc2cfdftv_10.c0000644000175000017500000000016212054411110014401 00000000000000/* Generated automatically. DO NOT EDIT! */ #define SIMD_HEADER "simd-avx.h" #include "../common/hc2cfdftv_10.c" fftw-3.3.3/rdft/simd/avx/hc2cfdftv_12.c0000644000175000017500000000016212054411110014403 00000000000000/* Generated automatically. DO NOT EDIT! */ #define SIMD_HEADER "simd-avx.h" #include "../common/hc2cfdftv_12.c" fftw-3.3.3/rdft/simd/avx/hc2cfdftv_16.c0000644000175000017500000000016212054411110014407 00000000000000/* Generated automatically. DO NOT EDIT! */ #define SIMD_HEADER "simd-avx.h" #include "../common/hc2cfdftv_16.c" fftw-3.3.3/rdft/simd/avx/hc2cfdftv_32.c0000644000175000017500000000016212054411110014405 00000000000000/* Generated automatically. DO NOT EDIT! */ #define SIMD_HEADER "simd-avx.h" #include "../common/hc2cfdftv_32.c" fftw-3.3.3/rdft/simd/avx/hc2cfdftv_20.c0000644000175000017500000000016212054411110014402 00000000000000/* Generated automatically. DO NOT EDIT! */ #define SIMD_HEADER "simd-avx.h" #include "../common/hc2cfdftv_20.c" fftw-3.3.3/rdft/simd/avx/hc2cbdftv_2.c0000644000175000017500000000016112054411110014315 00000000000000/* Generated automatically. DO NOT EDIT! */ #define SIMD_HEADER "simd-avx.h" #include "../common/hc2cbdftv_2.c" fftw-3.3.3/rdft/simd/avx/hc2cbdftv_4.c0000644000175000017500000000016112054411110014317 00000000000000/* Generated automatically. DO NOT EDIT! */ #define SIMD_HEADER "simd-avx.h" #include "../common/hc2cbdftv_4.c" fftw-3.3.3/rdft/simd/avx/hc2cbdftv_6.c0000644000175000017500000000016112054411110014321 00000000000000/* Generated automatically. DO NOT EDIT! */ #define SIMD_HEADER "simd-avx.h" #include "../common/hc2cbdftv_6.c" fftw-3.3.3/rdft/simd/avx/hc2cbdftv_8.c0000644000175000017500000000016112054411110014323 00000000000000/* Generated automatically. DO NOT EDIT! */ #define SIMD_HEADER "simd-avx.h" #include "../common/hc2cbdftv_8.c" fftw-3.3.3/rdft/simd/avx/hc2cbdftv_10.c0000644000175000017500000000016212054411110014375 00000000000000/* Generated automatically. DO NOT EDIT! */ #define SIMD_HEADER "simd-avx.h" #include "../common/hc2cbdftv_10.c" fftw-3.3.3/rdft/simd/avx/hc2cbdftv_12.c0000644000175000017500000000016212054411110014377 00000000000000/* Generated automatically. DO NOT EDIT! */ #define SIMD_HEADER "simd-avx.h" #include "../common/hc2cbdftv_12.c" fftw-3.3.3/rdft/simd/avx/hc2cbdftv_16.c0000644000175000017500000000016212054411110014403 00000000000000/* Generated automatically. DO NOT EDIT! */ #define SIMD_HEADER "simd-avx.h" #include "../common/hc2cbdftv_16.c" fftw-3.3.3/rdft/simd/avx/hc2cbdftv_32.c0000644000175000017500000000016212054411110014401 00000000000000/* Generated automatically. DO NOT EDIT! */ #define SIMD_HEADER "simd-avx.h" #include "../common/hc2cbdftv_32.c" fftw-3.3.3/rdft/simd/avx/hc2cbdftv_20.c0000644000175000017500000000016212054411110014376 00000000000000/* Generated automatically. DO NOT EDIT! */ #define SIMD_HEADER "simd-avx.h" #include "../common/hc2cbdftv_20.c" fftw-3.3.3/rdft/simd/avx/genus.c0000644000175000017500000000015312054411110013351 00000000000000/* Generated automatically. DO NOT EDIT! */ #define SIMD_HEADER "simd-avx.h" #include "../common/genus.c" fftw-3.3.3/rdft/simd/avx/codlist.c0000644000175000017500000000015512054411110013673 00000000000000/* Generated automatically. DO NOT EDIT! */ #define SIMD_HEADER "simd-avx.h" #include "../common/codlist.c" fftw-3.3.3/rdft/simd/altivec/0000755000175000017500000000000012054411110012776 500000000000000fftw-3.3.3/rdft/simd/altivec/Makefile.am0000644000175000017500000000045412054410050014757 00000000000000AM_CFLAGS = $(ALTIVEC_CFLAGS) SIMD_HEADER=simd-altivec.h include $(top_srcdir)/rdft/simd/codlist.mk include $(top_srcdir)/rdft/simd/simd.mk if HAVE_ALTIVEC noinst_LTLIBRARIES = librdft_altivec_codelets.la BUILT_SOURCES = $(EXTRA_DIST) librdft_altivec_codelets_la_SOURCES = $(BUILT_SOURCES) endif fftw-3.3.3/rdft/simd/altivec/Makefile.in0000644000175000017500000005025112054410200014765 00000000000000# Makefile.in generated by automake 1.11.6 from Makefile.am. # @configure_input@ # Copyright (C) 1994, 1995, 1996, 1997, 1998, 1999, 2000, 2001, 2002, # 2003, 2004, 2005, 2006, 2007, 2008, 2009, 2010, 2011 Free Software # Foundation, Inc. # This Makefile.in is free software; the Free Software Foundation # gives unlimited permission to copy and/or distribute it, # with or without modifications, as long as this notice is preserved. # This program is distributed in the hope that it will be useful, # but WITHOUT ANY WARRANTY, to the extent permitted by law; without # even the implied warranty of MERCHANTABILITY or FITNESS FOR A # PARTICULAR PURPOSE. @SET_MAKE@ # This file contains a standard list of RDFT SIMD codelets. It is # included by common/Makefile to generate the C files with the actual # codelets in them. It is included by {sse,sse2,...}/Makefile to # generate and compile stub files that include common/*.c # You can customize FFTW for special needs, e.g. to handle certain # sizes more efficiently, by adding new codelets to the lists of those # included by default. If you change the list of codelets, any new # ones you added will be automatically generated when you run the # bootstrap script (see "Generating your own code" in the FFTW # manual). VPATH = @srcdir@ am__make_dryrun = \ { \ am__dry=no; \ case $$MAKEFLAGS in \ *\\[\ \ ]*) \ echo 'am--echo: ; @echo "AM" OK' | $(MAKE) -f - 2>/dev/null \ | grep '^AM OK$$' >/dev/null || am__dry=yes;; \ *) \ for am__flg in $$MAKEFLAGS; do \ case $$am__flg in \ *=*|--*) ;; \ *n*) am__dry=yes; break;; \ esac; \ done;; \ esac; \ test $$am__dry = yes; \ } pkgdatadir = $(datadir)/@PACKAGE@ pkgincludedir = $(includedir)/@PACKAGE@ pkglibdir = $(libdir)/@PACKAGE@ pkglibexecdir = $(libexecdir)/@PACKAGE@ am__cd = CDPATH="$${ZSH_VERSION+.}$(PATH_SEPARATOR)" && cd install_sh_DATA = $(install_sh) -c -m 644 install_sh_PROGRAM = $(install_sh) -c install_sh_SCRIPT = $(install_sh) -c INSTALL_HEADER = $(INSTALL_DATA) transform = $(program_transform_name) NORMAL_INSTALL = : PRE_INSTALL = : POST_INSTALL = : NORMAL_UNINSTALL = : PRE_UNINSTALL = : POST_UNINSTALL = : build_triplet = @build@ host_triplet = @host@ DIST_COMMON = $(srcdir)/Makefile.am $(srcdir)/Makefile.in \ $(top_srcdir)/rdft/simd/codlist.mk \ $(top_srcdir)/rdft/simd/simd.mk subdir = rdft/simd/altivec ACLOCAL_M4 = $(top_srcdir)/aclocal.m4 am__aclocal_m4_deps = $(top_srcdir)/m4/acx_mpi.m4 \ $(top_srcdir)/m4/acx_pthread.m4 \ $(top_srcdir)/m4/ax_cc_maxopt.m4 \ $(top_srcdir)/m4/ax_check_compiler_flags.m4 \ $(top_srcdir)/m4/ax_compiler_vendor.m4 \ $(top_srcdir)/m4/ax_gcc_aligns_stack.m4 \ $(top_srcdir)/m4/ax_gcc_version.m4 \ $(top_srcdir)/m4/ax_openmp.m4 $(top_srcdir)/m4/libtool.m4 \ $(top_srcdir)/m4/ltoptions.m4 $(top_srcdir)/m4/ltsugar.m4 \ $(top_srcdir)/m4/ltversion.m4 $(top_srcdir)/m4/lt~obsolete.m4 \ $(top_srcdir)/configure.ac am__configure_deps = $(am__aclocal_m4_deps) $(CONFIGURE_DEPENDENCIES) \ $(ACLOCAL_M4) mkinstalldirs = $(install_sh) -d CONFIG_HEADER = $(top_builddir)/config.h CONFIG_CLEAN_FILES = CONFIG_CLEAN_VPATH_FILES = LTLIBRARIES = $(noinst_LTLIBRARIES) librdft_altivec_codelets_la_LIBADD = am__librdft_altivec_codelets_la_SOURCES_DIST = hc2cfdftv_2.c \ hc2cfdftv_4.c hc2cfdftv_6.c hc2cfdftv_8.c hc2cfdftv_10.c \ hc2cfdftv_12.c hc2cfdftv_16.c hc2cfdftv_32.c hc2cfdftv_20.c \ hc2cbdftv_2.c hc2cbdftv_4.c hc2cbdftv_6.c hc2cbdftv_8.c \ hc2cbdftv_10.c hc2cbdftv_12.c hc2cbdftv_16.c hc2cbdftv_32.c \ hc2cbdftv_20.c genus.c codlist.c am__objects_1 = hc2cfdftv_2.lo hc2cfdftv_4.lo hc2cfdftv_6.lo \ hc2cfdftv_8.lo hc2cfdftv_10.lo hc2cfdftv_12.lo hc2cfdftv_16.lo \ hc2cfdftv_32.lo hc2cfdftv_20.lo am__objects_2 = hc2cbdftv_2.lo hc2cbdftv_4.lo hc2cbdftv_6.lo \ hc2cbdftv_8.lo hc2cbdftv_10.lo hc2cbdftv_12.lo hc2cbdftv_16.lo \ hc2cbdftv_32.lo hc2cbdftv_20.lo am__objects_3 = $(am__objects_1) $(am__objects_2) am__objects_4 = $(am__objects_3) genus.lo codlist.lo @HAVE_ALTIVEC_TRUE@am__objects_5 = $(am__objects_4) @HAVE_ALTIVEC_TRUE@am_librdft_altivec_codelets_la_OBJECTS = \ @HAVE_ALTIVEC_TRUE@ $(am__objects_5) librdft_altivec_codelets_la_OBJECTS = \ $(am_librdft_altivec_codelets_la_OBJECTS) @HAVE_ALTIVEC_TRUE@am_librdft_altivec_codelets_la_rpath = DEFAULT_INCLUDES = -I.@am__isrc@ -I$(top_builddir) depcomp = $(SHELL) $(top_srcdir)/depcomp am__depfiles_maybe = depfiles am__mv = mv -f COMPILE = $(CC) $(DEFS) $(DEFAULT_INCLUDES) $(INCLUDES) $(AM_CPPFLAGS) \ $(CPPFLAGS) $(AM_CFLAGS) $(CFLAGS) LTCOMPILE = $(LIBTOOL) --tag=CC $(AM_LIBTOOLFLAGS) $(LIBTOOLFLAGS) \ --mode=compile $(CC) $(DEFS) $(DEFAULT_INCLUDES) $(INCLUDES) \ $(AM_CPPFLAGS) $(CPPFLAGS) $(AM_CFLAGS) $(CFLAGS) CCLD = $(CC) LINK = $(LIBTOOL) --tag=CC $(AM_LIBTOOLFLAGS) $(LIBTOOLFLAGS) \ --mode=link $(CCLD) $(AM_CFLAGS) $(CFLAGS) $(AM_LDFLAGS) \ $(LDFLAGS) -o $@ SOURCES = $(librdft_altivec_codelets_la_SOURCES) DIST_SOURCES = $(am__librdft_altivec_codelets_la_SOURCES_DIST) am__can_run_installinfo = \ case $$AM_UPDATE_INFO_DIR in \ n|no|NO) false;; \ *) (install-info --version) >/dev/null 2>&1;; \ esac ETAGS = etags CTAGS = ctags DISTFILES = $(DIST_COMMON) $(DIST_SOURCES) $(TEXINFOS) $(EXTRA_DIST) ACLOCAL = @ACLOCAL@ ALLOCA = @ALLOCA@ ALTIVEC_CFLAGS = @ALTIVEC_CFLAGS@ AMTAR = @AMTAR@ AR = @AR@ AS = @AS@ AUTOCONF = @AUTOCONF@ AUTOHEADER = @AUTOHEADER@ AUTOMAKE = @AUTOMAKE@ AVX_CFLAGS = @AVX_CFLAGS@ AWK = @AWK@ CC = @CC@ CCDEPMODE = @CCDEPMODE@ CFLAGS = @CFLAGS@ CHECK_PL_OPTS = @CHECK_PL_OPTS@ CPP = @CPP@ CPPFLAGS = @CPPFLAGS@ CYGPATH_W = @CYGPATH_W@ C_FFTW_R2R_KIND = @C_FFTW_R2R_KIND@ C_MPI_FINT = @C_MPI_FINT@ DEFS = @DEFS@ DEPDIR = @DEPDIR@ DLLTOOL = @DLLTOOL@ DSYMUTIL = @DSYMUTIL@ DUMPBIN = @DUMPBIN@ ECHO_C = @ECHO_C@ ECHO_N = @ECHO_N@ ECHO_T = @ECHO_T@ EGREP = @EGREP@ EXEEXT = @EXEEXT@ F77 = @F77@ FFLAGS = @FFLAGS@ FGREP = @FGREP@ FLIBS = @FLIBS@ GREP = @GREP@ INSTALL = @INSTALL@ INSTALL_DATA = @INSTALL_DATA@ INSTALL_PROGRAM = @INSTALL_PROGRAM@ INSTALL_SCRIPT = @INSTALL_SCRIPT@ INSTALL_STRIP_PROGRAM = @INSTALL_STRIP_PROGRAM@ LD = @LD@ LDFLAGS = @LDFLAGS@ LIBOBJS = @LIBOBJS@ LIBQUADMATH = @LIBQUADMATH@ LIBS = @LIBS@ LIBTOOL = @LIBTOOL@ LIPO = @LIPO@ LN_S = @LN_S@ LTLIBOBJS = @LTLIBOBJS@ MAINT = @MAINT@ MAKEINFO = @MAKEINFO@ MANIFEST_TOOL = @MANIFEST_TOOL@ MKDIR_P = @MKDIR_P@ MPICC = @MPICC@ MPILIBS = @MPILIBS@ MPIRUN = @MPIRUN@ NEON_CFLAGS = @NEON_CFLAGS@ NM = @NM@ NMEDIT = @NMEDIT@ OBJDUMP = @OBJDUMP@ OBJEXT = @OBJEXT@ OCAMLBUILD = @OCAMLBUILD@ OPENMP_CFLAGS = @OPENMP_CFLAGS@ OTOOL = @OTOOL@ OTOOL64 = @OTOOL64@ PACKAGE = @PACKAGE@ PACKAGE_BUGREPORT = @PACKAGE_BUGREPORT@ PACKAGE_NAME = @PACKAGE_NAME@ PACKAGE_STRING = @PACKAGE_STRING@ PACKAGE_TARNAME = @PACKAGE_TARNAME@ PACKAGE_URL = @PACKAGE_URL@ PACKAGE_VERSION = @PACKAGE_VERSION@ PATH_SEPARATOR = @PATH_SEPARATOR@ POW_LIB = @POW_LIB@ PRECISION = @PRECISION@ PREC_SUFFIX = @PREC_SUFFIX@ PTHREAD_CC = @PTHREAD_CC@ PTHREAD_CFLAGS = @PTHREAD_CFLAGS@ PTHREAD_LIBS = @PTHREAD_LIBS@ RANLIB = @RANLIB@ SED = @SED@ SET_MAKE = @SET_MAKE@ SHARED_VERSION_INFO = @SHARED_VERSION_INFO@ SHELL = @SHELL@ SSE2_CFLAGS = @SSE2_CFLAGS@ STACK_ALIGN_CFLAGS = @STACK_ALIGN_CFLAGS@ STRIP = @STRIP@ THREADLIBS = @THREADLIBS@ VERSION = @VERSION@ abs_builddir = @abs_builddir@ abs_srcdir = @abs_srcdir@ abs_top_builddir = @abs_top_builddir@ abs_top_srcdir = @abs_top_srcdir@ ac_ct_AR = @ac_ct_AR@ ac_ct_CC = @ac_ct_CC@ ac_ct_DUMPBIN = @ac_ct_DUMPBIN@ ac_ct_F77 = @ac_ct_F77@ acx_pthread_config = @acx_pthread_config@ am__include = @am__include@ am__leading_dot = @am__leading_dot@ am__quote = @am__quote@ am__tar = @am__tar@ am__untar = @am__untar@ bindir = @bindir@ build = @build@ build_alias = @build_alias@ build_cpu = @build_cpu@ build_os = @build_os@ build_vendor = @build_vendor@ builddir = @builddir@ datadir = @datadir@ datarootdir = @datarootdir@ docdir = @docdir@ dvidir = @dvidir@ exec_prefix = @exec_prefix@ host = @host@ host_alias = @host_alias@ host_cpu = @host_cpu@ host_os = @host_os@ host_vendor = @host_vendor@ htmldir = @htmldir@ includedir = @includedir@ infodir = @infodir@ install_sh = @install_sh@ libdir = @libdir@ libexecdir = @libexecdir@ localedir = @localedir@ localstatedir = @localstatedir@ mandir = @mandir@ mkdir_p = @mkdir_p@ oldincludedir = @oldincludedir@ pdfdir = @pdfdir@ prefix = @prefix@ program_transform_name = @program_transform_name@ psdir = @psdir@ sbindir = @sbindir@ sharedstatedir = @sharedstatedir@ srcdir = @srcdir@ sysconfdir = @sysconfdir@ target_alias = @target_alias@ top_build_prefix = @top_build_prefix@ top_builddir = @top_builddir@ top_srcdir = @top_srcdir@ AM_CFLAGS = $(ALTIVEC_CFLAGS) SIMD_HEADER = simd-altivec.h HC2CFDFTV = hc2cfdftv_2.c hc2cfdftv_4.c hc2cfdftv_6.c hc2cfdftv_8.c \ hc2cfdftv_10.c hc2cfdftv_12.c hc2cfdftv_16.c hc2cfdftv_32.c \ hc2cfdftv_20.c HC2CBDFTV = hc2cbdftv_2.c hc2cbdftv_4.c hc2cbdftv_6.c hc2cbdftv_8.c \ hc2cbdftv_10.c hc2cbdftv_12.c hc2cbdftv_16.c hc2cbdftv_32.c \ hc2cbdftv_20.c ########################################################################### SIMD_CODELETS = $(HC2CFDFTV) $(HC2CBDFTV) AM_CPPFLAGS = -I$(top_srcdir)/kernel -I$(top_srcdir)/rdft \ -I$(top_srcdir)/rdft/simd -I$(top_srcdir)/simd-support EXTRA_DIST = $(SIMD_CODELETS) genus.c codlist.c @HAVE_ALTIVEC_TRUE@noinst_LTLIBRARIES = librdft_altivec_codelets.la @HAVE_ALTIVEC_TRUE@BUILT_SOURCES = $(EXTRA_DIST) @HAVE_ALTIVEC_TRUE@librdft_altivec_codelets_la_SOURCES = $(BUILT_SOURCES) all: $(BUILT_SOURCES) $(MAKE) $(AM_MAKEFLAGS) all-am .SUFFIXES: .SUFFIXES: .c .lo .o .obj $(srcdir)/Makefile.in: @MAINTAINER_MODE_TRUE@ $(srcdir)/Makefile.am $(top_srcdir)/rdft/simd/codlist.mk $(top_srcdir)/rdft/simd/simd.mk $(am__configure_deps) @for dep in $?; do \ case '$(am__configure_deps)' in \ *$$dep*) \ ( cd $(top_builddir) && $(MAKE) $(AM_MAKEFLAGS) am--refresh ) \ && { if test -f $@; then exit 0; else break; fi; }; \ exit 1;; \ esac; \ done; \ echo ' cd $(top_srcdir) && $(AUTOMAKE) --gnu rdft/simd/altivec/Makefile'; \ $(am__cd) $(top_srcdir) && \ $(AUTOMAKE) --gnu rdft/simd/altivec/Makefile .PRECIOUS: Makefile Makefile: $(srcdir)/Makefile.in $(top_builddir)/config.status @case '$?' in \ *config.status*) \ cd $(top_builddir) && $(MAKE) $(AM_MAKEFLAGS) am--refresh;; \ *) \ echo ' cd $(top_builddir) && $(SHELL) ./config.status $(subdir)/$@ $(am__depfiles_maybe)'; \ cd $(top_builddir) && $(SHELL) ./config.status $(subdir)/$@ $(am__depfiles_maybe);; \ esac; $(top_srcdir)/rdft/simd/codlist.mk $(top_srcdir)/rdft/simd/simd.mk: $(top_builddir)/config.status: $(top_srcdir)/configure $(CONFIG_STATUS_DEPENDENCIES) cd $(top_builddir) && $(MAKE) $(AM_MAKEFLAGS) am--refresh $(top_srcdir)/configure: @MAINTAINER_MODE_TRUE@ $(am__configure_deps) cd $(top_builddir) && $(MAKE) $(AM_MAKEFLAGS) am--refresh $(ACLOCAL_M4): @MAINTAINER_MODE_TRUE@ $(am__aclocal_m4_deps) cd $(top_builddir) && $(MAKE) $(AM_MAKEFLAGS) am--refresh $(am__aclocal_m4_deps): clean-noinstLTLIBRARIES: -test -z "$(noinst_LTLIBRARIES)" || rm -f $(noinst_LTLIBRARIES) @list='$(noinst_LTLIBRARIES)'; for p in $$list; do \ dir="`echo $$p | sed -e 's|/[^/]*$$||'`"; \ test "$$dir" != "$$p" || dir=.; \ echo "rm -f \"$${dir}/so_locations\""; \ rm -f "$${dir}/so_locations"; \ done librdft_altivec_codelets.la: $(librdft_altivec_codelets_la_OBJECTS) $(librdft_altivec_codelets_la_DEPENDENCIES) $(EXTRA_librdft_altivec_codelets_la_DEPENDENCIES) $(LINK) $(am_librdft_altivec_codelets_la_rpath) $(librdft_altivec_codelets_la_OBJECTS) $(librdft_altivec_codelets_la_LIBADD) $(LIBS) mostlyclean-compile: -rm -f *.$(OBJEXT) distclean-compile: -rm -f *.tab.c @AMDEP_TRUE@@am__include@ @am__quote@./$(DEPDIR)/codlist.Plo@am__quote@ @AMDEP_TRUE@@am__include@ @am__quote@./$(DEPDIR)/genus.Plo@am__quote@ @AMDEP_TRUE@@am__include@ @am__quote@./$(DEPDIR)/hc2cbdftv_10.Plo@am__quote@ @AMDEP_TRUE@@am__include@ @am__quote@./$(DEPDIR)/hc2cbdftv_12.Plo@am__quote@ @AMDEP_TRUE@@am__include@ @am__quote@./$(DEPDIR)/hc2cbdftv_16.Plo@am__quote@ @AMDEP_TRUE@@am__include@ @am__quote@./$(DEPDIR)/hc2cbdftv_2.Plo@am__quote@ @AMDEP_TRUE@@am__include@ @am__quote@./$(DEPDIR)/hc2cbdftv_20.Plo@am__quote@ @AMDEP_TRUE@@am__include@ @am__quote@./$(DEPDIR)/hc2cbdftv_32.Plo@am__quote@ @AMDEP_TRUE@@am__include@ @am__quote@./$(DEPDIR)/hc2cbdftv_4.Plo@am__quote@ @AMDEP_TRUE@@am__include@ @am__quote@./$(DEPDIR)/hc2cbdftv_6.Plo@am__quote@ @AMDEP_TRUE@@am__include@ @am__quote@./$(DEPDIR)/hc2cbdftv_8.Plo@am__quote@ @AMDEP_TRUE@@am__include@ @am__quote@./$(DEPDIR)/hc2cfdftv_10.Plo@am__quote@ @AMDEP_TRUE@@am__include@ @am__quote@./$(DEPDIR)/hc2cfdftv_12.Plo@am__quote@ @AMDEP_TRUE@@am__include@ @am__quote@./$(DEPDIR)/hc2cfdftv_16.Plo@am__quote@ @AMDEP_TRUE@@am__include@ @am__quote@./$(DEPDIR)/hc2cfdftv_2.Plo@am__quote@ @AMDEP_TRUE@@am__include@ @am__quote@./$(DEPDIR)/hc2cfdftv_20.Plo@am__quote@ @AMDEP_TRUE@@am__include@ @am__quote@./$(DEPDIR)/hc2cfdftv_32.Plo@am__quote@ @AMDEP_TRUE@@am__include@ @am__quote@./$(DEPDIR)/hc2cfdftv_4.Plo@am__quote@ @AMDEP_TRUE@@am__include@ @am__quote@./$(DEPDIR)/hc2cfdftv_6.Plo@am__quote@ @AMDEP_TRUE@@am__include@ @am__quote@./$(DEPDIR)/hc2cfdftv_8.Plo@am__quote@ .c.o: @am__fastdepCC_TRUE@ $(COMPILE) -MT $@ -MD -MP -MF $(DEPDIR)/$*.Tpo -c -o $@ $< @am__fastdepCC_TRUE@ $(am__mv) $(DEPDIR)/$*.Tpo $(DEPDIR)/$*.Po @AMDEP_TRUE@@am__fastdepCC_FALSE@ source='$<' object='$@' libtool=no @AMDEPBACKSLASH@ @AMDEP_TRUE@@am__fastdepCC_FALSE@ DEPDIR=$(DEPDIR) $(CCDEPMODE) $(depcomp) @AMDEPBACKSLASH@ @am__fastdepCC_FALSE@ $(COMPILE) -c $< .c.obj: @am__fastdepCC_TRUE@ $(COMPILE) -MT $@ -MD -MP -MF $(DEPDIR)/$*.Tpo -c -o $@ `$(CYGPATH_W) '$<'` @am__fastdepCC_TRUE@ $(am__mv) $(DEPDIR)/$*.Tpo $(DEPDIR)/$*.Po @AMDEP_TRUE@@am__fastdepCC_FALSE@ source='$<' object='$@' libtool=no @AMDEPBACKSLASH@ @AMDEP_TRUE@@am__fastdepCC_FALSE@ DEPDIR=$(DEPDIR) $(CCDEPMODE) $(depcomp) @AMDEPBACKSLASH@ @am__fastdepCC_FALSE@ $(COMPILE) -c `$(CYGPATH_W) '$<'` .c.lo: @am__fastdepCC_TRUE@ $(LTCOMPILE) -MT $@ -MD -MP -MF $(DEPDIR)/$*.Tpo -c -o $@ $< @am__fastdepCC_TRUE@ $(am__mv) $(DEPDIR)/$*.Tpo $(DEPDIR)/$*.Plo @AMDEP_TRUE@@am__fastdepCC_FALSE@ source='$<' object='$@' libtool=yes @AMDEPBACKSLASH@ @AMDEP_TRUE@@am__fastdepCC_FALSE@ DEPDIR=$(DEPDIR) $(CCDEPMODE) $(depcomp) @AMDEPBACKSLASH@ @am__fastdepCC_FALSE@ $(LTCOMPILE) -c -o $@ $< mostlyclean-libtool: -rm -f *.lo clean-libtool: -rm -rf .libs _libs ID: $(HEADERS) $(SOURCES) $(LISP) $(TAGS_FILES) list='$(SOURCES) $(HEADERS) $(LISP) $(TAGS_FILES)'; \ unique=`for i in $$list; do \ if test -f "$$i"; then echo $$i; else echo $(srcdir)/$$i; fi; \ done | \ $(AWK) '{ files[$$0] = 1; nonempty = 1; } \ END { if (nonempty) { for (i in files) print i; }; }'`; \ mkid -fID $$unique tags: TAGS TAGS: $(HEADERS) $(SOURCES) $(TAGS_DEPENDENCIES) \ $(TAGS_FILES) $(LISP) set x; \ here=`pwd`; \ list='$(SOURCES) $(HEADERS) $(LISP) $(TAGS_FILES)'; \ unique=`for i in $$list; do \ if test -f "$$i"; then echo $$i; else echo $(srcdir)/$$i; fi; \ done | \ $(AWK) '{ files[$$0] = 1; nonempty = 1; } \ END { if (nonempty) { for (i in files) print i; }; }'`; \ shift; \ if test -z "$(ETAGS_ARGS)$$*$$unique"; then :; else \ test -n "$$unique" || unique=$$empty_fix; \ if test $$# -gt 0; then \ $(ETAGS) $(ETAGSFLAGS) $(AM_ETAGSFLAGS) $(ETAGS_ARGS) \ "$$@" $$unique; \ else \ $(ETAGS) $(ETAGSFLAGS) $(AM_ETAGSFLAGS) $(ETAGS_ARGS) \ $$unique; \ fi; \ fi ctags: CTAGS CTAGS: $(HEADERS) $(SOURCES) $(TAGS_DEPENDENCIES) \ $(TAGS_FILES) $(LISP) list='$(SOURCES) $(HEADERS) $(LISP) $(TAGS_FILES)'; \ unique=`for i in $$list; do \ if test -f "$$i"; then echo $$i; else echo $(srcdir)/$$i; fi; \ done | \ $(AWK) '{ files[$$0] = 1; nonempty = 1; } \ END { if (nonempty) { for (i in files) print i; }; }'`; \ test -z "$(CTAGS_ARGS)$$unique" \ || $(CTAGS) $(CTAGSFLAGS) $(AM_CTAGSFLAGS) $(CTAGS_ARGS) \ $$unique GTAGS: here=`$(am__cd) $(top_builddir) && pwd` \ && $(am__cd) $(top_srcdir) \ && gtags -i $(GTAGS_ARGS) "$$here" distclean-tags: -rm -f TAGS ID GTAGS GRTAGS GSYMS GPATH tags distdir: $(DISTFILES) @srcdirstrip=`echo "$(srcdir)" | sed 's/[].[^$$\\*]/\\\\&/g'`; \ topsrcdirstrip=`echo "$(top_srcdir)" | sed 's/[].[^$$\\*]/\\\\&/g'`; \ list='$(DISTFILES)'; \ dist_files=`for file in $$list; do echo $$file; done | \ sed -e "s|^$$srcdirstrip/||;t" \ -e "s|^$$topsrcdirstrip/|$(top_builddir)/|;t"`; \ case $$dist_files in \ */*) $(MKDIR_P) `echo "$$dist_files" | \ sed '/\//!d;s|^|$(distdir)/|;s,/[^/]*$$,,' | \ sort -u` ;; \ esac; \ for file in $$dist_files; do \ if test -f $$file || test -d $$file; then d=.; else d=$(srcdir); fi; \ if test -d $$d/$$file; then \ dir=`echo "/$$file" | sed -e 's,/[^/]*$$,,'`; \ if test -d "$(distdir)/$$file"; then \ find "$(distdir)/$$file" -type d ! -perm -700 -exec chmod u+rwx {} \;; \ fi; \ if test -d $(srcdir)/$$file && test $$d != $(srcdir); then \ cp -fpR $(srcdir)/$$file "$(distdir)$$dir" || exit 1; \ find "$(distdir)/$$file" -type d ! -perm -700 -exec chmod u+rwx {} \;; \ fi; \ cp -fpR $$d/$$file "$(distdir)$$dir" || exit 1; \ else \ test -f "$(distdir)/$$file" \ || cp -p $$d/$$file "$(distdir)/$$file" \ || exit 1; \ fi; \ done check-am: all-am check: $(BUILT_SOURCES) $(MAKE) $(AM_MAKEFLAGS) check-am all-am: Makefile $(LTLIBRARIES) installdirs: install: $(BUILT_SOURCES) $(MAKE) $(AM_MAKEFLAGS) install-am install-exec: install-exec-am install-data: install-data-am uninstall: uninstall-am install-am: all-am @$(MAKE) $(AM_MAKEFLAGS) install-exec-am install-data-am installcheck: installcheck-am install-strip: if test -z '$(STRIP)'; then \ $(MAKE) $(AM_MAKEFLAGS) INSTALL_PROGRAM="$(INSTALL_STRIP_PROGRAM)" \ install_sh_PROGRAM="$(INSTALL_STRIP_PROGRAM)" INSTALL_STRIP_FLAG=-s \ install; \ else \ $(MAKE) $(AM_MAKEFLAGS) INSTALL_PROGRAM="$(INSTALL_STRIP_PROGRAM)" \ install_sh_PROGRAM="$(INSTALL_STRIP_PROGRAM)" INSTALL_STRIP_FLAG=-s \ "INSTALL_PROGRAM_ENV=STRIPPROG='$(STRIP)'" install; \ fi mostlyclean-generic: clean-generic: distclean-generic: -test -z "$(CONFIG_CLEAN_FILES)" || rm -f $(CONFIG_CLEAN_FILES) -test . = "$(srcdir)" || test -z "$(CONFIG_CLEAN_VPATH_FILES)" || rm -f $(CONFIG_CLEAN_VPATH_FILES) maintainer-clean-generic: @echo "This command is intended for maintainers to use" @echo "it deletes files that may require special tools to rebuild." -test -z "$(BUILT_SOURCES)" || rm -f $(BUILT_SOURCES) clean: clean-am clean-am: clean-generic clean-libtool clean-noinstLTLIBRARIES \ mostlyclean-am distclean: distclean-am -rm -rf ./$(DEPDIR) -rm -f Makefile distclean-am: clean-am distclean-compile distclean-generic \ distclean-tags dvi: dvi-am dvi-am: html: html-am html-am: info: info-am info-am: install-data-am: install-dvi: install-dvi-am install-dvi-am: install-exec-am: install-html: install-html-am install-html-am: install-info: install-info-am install-info-am: install-man: install-pdf: install-pdf-am install-pdf-am: install-ps: install-ps-am install-ps-am: installcheck-am: maintainer-clean: maintainer-clean-am -rm -rf ./$(DEPDIR) -rm -f Makefile maintainer-clean-am: distclean-am maintainer-clean-generic mostlyclean: mostlyclean-am mostlyclean-am: mostlyclean-compile mostlyclean-generic \ mostlyclean-libtool pdf: pdf-am pdf-am: ps: ps-am ps-am: uninstall-am: .MAKE: all check install install-am install-strip .PHONY: CTAGS GTAGS all all-am check check-am clean clean-generic \ clean-libtool clean-noinstLTLIBRARIES ctags distclean \ distclean-compile distclean-generic distclean-libtool \ distclean-tags distdir dvi dvi-am html html-am info info-am \ install install-am install-data install-data-am install-dvi \ install-dvi-am install-exec install-exec-am install-html \ install-html-am install-info install-info-am install-man \ install-pdf install-pdf-am install-ps install-ps-am \ install-strip installcheck installcheck-am installdirs \ maintainer-clean maintainer-clean-generic mostlyclean \ mostlyclean-compile mostlyclean-generic mostlyclean-libtool \ pdf pdf-am ps ps-am tags uninstall uninstall-am $(EXTRA_DIST): Makefile ( \ echo "/* Generated automatically. DO NOT EDIT! */"; \ echo "#define SIMD_HEADER \"$(SIMD_HEADER)\""; \ echo "#include \"../common/"$*".c\""; \ ) >$@ # Tell versions [3.59,3.63) of GNU make to not export all variables. # Otherwise a system limit (for SysV at least) may be exceeded. .NOEXPORT: fftw-3.3.3/rdft/simd/altivec/hc2cfdftv_2.c0000644000175000017500000000016512054411110015156 00000000000000/* Generated automatically. DO NOT EDIT! */ #define SIMD_HEADER "simd-altivec.h" #include "../common/hc2cfdftv_2.c" fftw-3.3.3/rdft/simd/altivec/hc2cfdftv_4.c0000644000175000017500000000016512054411110015160 00000000000000/* Generated automatically. DO NOT EDIT! */ #define SIMD_HEADER "simd-altivec.h" #include "../common/hc2cfdftv_4.c" fftw-3.3.3/rdft/simd/altivec/hc2cfdftv_6.c0000644000175000017500000000016512054411110015162 00000000000000/* Generated automatically. DO NOT EDIT! */ #define SIMD_HEADER "simd-altivec.h" #include "../common/hc2cfdftv_6.c" fftw-3.3.3/rdft/simd/altivec/hc2cfdftv_8.c0000644000175000017500000000016512054411110015164 00000000000000/* Generated automatically. DO NOT EDIT! */ #define SIMD_HEADER "simd-altivec.h" #include "../common/hc2cfdftv_8.c" fftw-3.3.3/rdft/simd/altivec/hc2cfdftv_10.c0000644000175000017500000000016612054411110015236 00000000000000/* Generated automatically. DO NOT EDIT! */ #define SIMD_HEADER "simd-altivec.h" #include "../common/hc2cfdftv_10.c" fftw-3.3.3/rdft/simd/altivec/hc2cfdftv_12.c0000644000175000017500000000016612054411110015240 00000000000000/* Generated automatically. DO NOT EDIT! */ #define SIMD_HEADER "simd-altivec.h" #include "../common/hc2cfdftv_12.c" fftw-3.3.3/rdft/simd/altivec/hc2cfdftv_16.c0000644000175000017500000000016612054411110015244 00000000000000/* Generated automatically. DO NOT EDIT! */ #define SIMD_HEADER "simd-altivec.h" #include "../common/hc2cfdftv_16.c" fftw-3.3.3/rdft/simd/altivec/hc2cfdftv_32.c0000644000175000017500000000016612054411110015242 00000000000000/* Generated automatically. DO NOT EDIT! */ #define SIMD_HEADER "simd-altivec.h" #include "../common/hc2cfdftv_32.c" fftw-3.3.3/rdft/simd/altivec/hc2cfdftv_20.c0000644000175000017500000000016612054411110015237 00000000000000/* Generated automatically. DO NOT EDIT! */ #define SIMD_HEADER "simd-altivec.h" #include "../common/hc2cfdftv_20.c" fftw-3.3.3/rdft/simd/altivec/hc2cbdftv_2.c0000644000175000017500000000016512054411110015152 00000000000000/* Generated automatically. DO NOT EDIT! */ #define SIMD_HEADER "simd-altivec.h" #include "../common/hc2cbdftv_2.c" fftw-3.3.3/rdft/simd/altivec/hc2cbdftv_4.c0000644000175000017500000000016512054411110015154 00000000000000/* Generated automatically. DO NOT EDIT! */ #define SIMD_HEADER "simd-altivec.h" #include "../common/hc2cbdftv_4.c" fftw-3.3.3/rdft/simd/altivec/hc2cbdftv_6.c0000644000175000017500000000016512054411110015156 00000000000000/* Generated automatically. DO NOT EDIT! */ #define SIMD_HEADER "simd-altivec.h" #include "../common/hc2cbdftv_6.c" fftw-3.3.3/rdft/simd/altivec/hc2cbdftv_8.c0000644000175000017500000000016512054411110015160 00000000000000/* Generated automatically. DO NOT EDIT! */ #define SIMD_HEADER "simd-altivec.h" #include "../common/hc2cbdftv_8.c" fftw-3.3.3/rdft/simd/altivec/hc2cbdftv_10.c0000644000175000017500000000016612054411110015232 00000000000000/* Generated automatically. DO NOT EDIT! */ #define SIMD_HEADER "simd-altivec.h" #include "../common/hc2cbdftv_10.c" fftw-3.3.3/rdft/simd/altivec/hc2cbdftv_12.c0000644000175000017500000000016612054411110015234 00000000000000/* Generated automatically. DO NOT EDIT! */ #define SIMD_HEADER "simd-altivec.h" #include "../common/hc2cbdftv_12.c" fftw-3.3.3/rdft/simd/altivec/hc2cbdftv_16.c0000644000175000017500000000016612054411110015240 00000000000000/* Generated automatically. DO NOT EDIT! */ #define SIMD_HEADER "simd-altivec.h" #include "../common/hc2cbdftv_16.c" fftw-3.3.3/rdft/simd/altivec/hc2cbdftv_32.c0000644000175000017500000000016612054411110015236 00000000000000/* Generated automatically. DO NOT EDIT! */ #define SIMD_HEADER "simd-altivec.h" #include "../common/hc2cbdftv_32.c" fftw-3.3.3/rdft/simd/altivec/hc2cbdftv_20.c0000644000175000017500000000016612054411110015233 00000000000000/* Generated automatically. DO NOT EDIT! */ #define SIMD_HEADER "simd-altivec.h" #include "../common/hc2cbdftv_20.c" fftw-3.3.3/rdft/simd/altivec/genus.c0000644000175000017500000000015712054411110014206 00000000000000/* Generated automatically. DO NOT EDIT! */ #define SIMD_HEADER "simd-altivec.h" #include "../common/genus.c" fftw-3.3.3/rdft/simd/altivec/codlist.c0000644000175000017500000000016112054411110014521 00000000000000/* Generated automatically. DO NOT EDIT! */ #define SIMD_HEADER "simd-altivec.h" #include "../common/codlist.c" fftw-3.3.3/rdft/simd/neon/0000755000175000017500000000000012054411110012306 500000000000000fftw-3.3.3/rdft/simd/neon/Makefile.am0000644000175000017500000000043512054410050014266 00000000000000AM_CFLAGS = $(NEON_CFLAGS) SIMD_HEADER=simd-neon.h include $(top_srcdir)/rdft/simd/codlist.mk include $(top_srcdir)/rdft/simd/simd.mk if HAVE_NEON noinst_LTLIBRARIES = librdft_neon_codelets.la BUILT_SOURCES = $(EXTRA_DIST) librdft_neon_codelets_la_SOURCES = $(BUILT_SOURCES) endif fftw-3.3.3/rdft/simd/neon/Makefile.in0000644000175000017500000005012212054410200014272 00000000000000# Makefile.in generated by automake 1.11.6 from Makefile.am. # @configure_input@ # Copyright (C) 1994, 1995, 1996, 1997, 1998, 1999, 2000, 2001, 2002, # 2003, 2004, 2005, 2006, 2007, 2008, 2009, 2010, 2011 Free Software # Foundation, Inc. # This Makefile.in is free software; the Free Software Foundation # gives unlimited permission to copy and/or distribute it, # with or without modifications, as long as this notice is preserved. # This program is distributed in the hope that it will be useful, # but WITHOUT ANY WARRANTY, to the extent permitted by law; without # even the implied warranty of MERCHANTABILITY or FITNESS FOR A # PARTICULAR PURPOSE. @SET_MAKE@ # This file contains a standard list of RDFT SIMD codelets. It is # included by common/Makefile to generate the C files with the actual # codelets in them. It is included by {sse,sse2,...}/Makefile to # generate and compile stub files that include common/*.c # You can customize FFTW for special needs, e.g. to handle certain # sizes more efficiently, by adding new codelets to the lists of those # included by default. If you change the list of codelets, any new # ones you added will be automatically generated when you run the # bootstrap script (see "Generating your own code" in the FFTW # manual). VPATH = @srcdir@ am__make_dryrun = \ { \ am__dry=no; \ case $$MAKEFLAGS in \ *\\[\ \ ]*) \ echo 'am--echo: ; @echo "AM" OK' | $(MAKE) -f - 2>/dev/null \ | grep '^AM OK$$' >/dev/null || am__dry=yes;; \ *) \ for am__flg in $$MAKEFLAGS; do \ case $$am__flg in \ *=*|--*) ;; \ *n*) am__dry=yes; break;; \ esac; \ done;; \ esac; \ test $$am__dry = yes; \ } pkgdatadir = $(datadir)/@PACKAGE@ pkgincludedir = $(includedir)/@PACKAGE@ pkglibdir = $(libdir)/@PACKAGE@ pkglibexecdir = $(libexecdir)/@PACKAGE@ am__cd = CDPATH="$${ZSH_VERSION+.}$(PATH_SEPARATOR)" && cd install_sh_DATA = $(install_sh) -c -m 644 install_sh_PROGRAM = $(install_sh) -c install_sh_SCRIPT = $(install_sh) -c INSTALL_HEADER = $(INSTALL_DATA) transform = $(program_transform_name) NORMAL_INSTALL = : PRE_INSTALL = : POST_INSTALL = : NORMAL_UNINSTALL = : PRE_UNINSTALL = : POST_UNINSTALL = : build_triplet = @build@ host_triplet = @host@ DIST_COMMON = $(srcdir)/Makefile.am $(srcdir)/Makefile.in \ $(top_srcdir)/rdft/simd/codlist.mk \ $(top_srcdir)/rdft/simd/simd.mk subdir = rdft/simd/neon ACLOCAL_M4 = $(top_srcdir)/aclocal.m4 am__aclocal_m4_deps = $(top_srcdir)/m4/acx_mpi.m4 \ $(top_srcdir)/m4/acx_pthread.m4 \ $(top_srcdir)/m4/ax_cc_maxopt.m4 \ $(top_srcdir)/m4/ax_check_compiler_flags.m4 \ $(top_srcdir)/m4/ax_compiler_vendor.m4 \ $(top_srcdir)/m4/ax_gcc_aligns_stack.m4 \ $(top_srcdir)/m4/ax_gcc_version.m4 \ $(top_srcdir)/m4/ax_openmp.m4 $(top_srcdir)/m4/libtool.m4 \ $(top_srcdir)/m4/ltoptions.m4 $(top_srcdir)/m4/ltsugar.m4 \ $(top_srcdir)/m4/ltversion.m4 $(top_srcdir)/m4/lt~obsolete.m4 \ $(top_srcdir)/configure.ac am__configure_deps = $(am__aclocal_m4_deps) $(CONFIGURE_DEPENDENCIES) \ $(ACLOCAL_M4) mkinstalldirs = $(install_sh) -d CONFIG_HEADER = $(top_builddir)/config.h CONFIG_CLEAN_FILES = CONFIG_CLEAN_VPATH_FILES = LTLIBRARIES = $(noinst_LTLIBRARIES) librdft_neon_codelets_la_LIBADD = am__librdft_neon_codelets_la_SOURCES_DIST = hc2cfdftv_2.c \ hc2cfdftv_4.c hc2cfdftv_6.c hc2cfdftv_8.c hc2cfdftv_10.c \ hc2cfdftv_12.c hc2cfdftv_16.c hc2cfdftv_32.c hc2cfdftv_20.c \ hc2cbdftv_2.c hc2cbdftv_4.c hc2cbdftv_6.c hc2cbdftv_8.c \ hc2cbdftv_10.c hc2cbdftv_12.c hc2cbdftv_16.c hc2cbdftv_32.c \ hc2cbdftv_20.c genus.c codlist.c am__objects_1 = hc2cfdftv_2.lo hc2cfdftv_4.lo hc2cfdftv_6.lo \ hc2cfdftv_8.lo hc2cfdftv_10.lo hc2cfdftv_12.lo hc2cfdftv_16.lo \ hc2cfdftv_32.lo hc2cfdftv_20.lo am__objects_2 = hc2cbdftv_2.lo hc2cbdftv_4.lo hc2cbdftv_6.lo \ hc2cbdftv_8.lo hc2cbdftv_10.lo hc2cbdftv_12.lo hc2cbdftv_16.lo \ hc2cbdftv_32.lo hc2cbdftv_20.lo am__objects_3 = $(am__objects_1) $(am__objects_2) am__objects_4 = $(am__objects_3) genus.lo codlist.lo @HAVE_NEON_TRUE@am__objects_5 = $(am__objects_4) @HAVE_NEON_TRUE@am_librdft_neon_codelets_la_OBJECTS = \ @HAVE_NEON_TRUE@ $(am__objects_5) librdft_neon_codelets_la_OBJECTS = \ $(am_librdft_neon_codelets_la_OBJECTS) @HAVE_NEON_TRUE@am_librdft_neon_codelets_la_rpath = DEFAULT_INCLUDES = -I.@am__isrc@ -I$(top_builddir) depcomp = $(SHELL) $(top_srcdir)/depcomp am__depfiles_maybe = depfiles am__mv = mv -f COMPILE = $(CC) $(DEFS) $(DEFAULT_INCLUDES) $(INCLUDES) $(AM_CPPFLAGS) \ $(CPPFLAGS) $(AM_CFLAGS) $(CFLAGS) LTCOMPILE = $(LIBTOOL) --tag=CC $(AM_LIBTOOLFLAGS) $(LIBTOOLFLAGS) \ --mode=compile $(CC) $(DEFS) $(DEFAULT_INCLUDES) $(INCLUDES) \ $(AM_CPPFLAGS) $(CPPFLAGS) $(AM_CFLAGS) $(CFLAGS) CCLD = $(CC) LINK = $(LIBTOOL) --tag=CC $(AM_LIBTOOLFLAGS) $(LIBTOOLFLAGS) \ --mode=link $(CCLD) $(AM_CFLAGS) $(CFLAGS) $(AM_LDFLAGS) \ $(LDFLAGS) -o $@ SOURCES = $(librdft_neon_codelets_la_SOURCES) DIST_SOURCES = $(am__librdft_neon_codelets_la_SOURCES_DIST) am__can_run_installinfo = \ case $$AM_UPDATE_INFO_DIR in \ n|no|NO) false;; \ *) (install-info --version) >/dev/null 2>&1;; \ esac ETAGS = etags CTAGS = ctags DISTFILES = $(DIST_COMMON) $(DIST_SOURCES) $(TEXINFOS) $(EXTRA_DIST) ACLOCAL = @ACLOCAL@ ALLOCA = @ALLOCA@ ALTIVEC_CFLAGS = @ALTIVEC_CFLAGS@ AMTAR = @AMTAR@ AR = @AR@ AS = @AS@ AUTOCONF = @AUTOCONF@ AUTOHEADER = @AUTOHEADER@ AUTOMAKE = @AUTOMAKE@ AVX_CFLAGS = @AVX_CFLAGS@ AWK = @AWK@ CC = @CC@ CCDEPMODE = @CCDEPMODE@ CFLAGS = @CFLAGS@ CHECK_PL_OPTS = @CHECK_PL_OPTS@ CPP = @CPP@ CPPFLAGS = @CPPFLAGS@ CYGPATH_W = @CYGPATH_W@ C_FFTW_R2R_KIND = @C_FFTW_R2R_KIND@ C_MPI_FINT = @C_MPI_FINT@ DEFS = @DEFS@ DEPDIR = @DEPDIR@ DLLTOOL = @DLLTOOL@ DSYMUTIL = @DSYMUTIL@ DUMPBIN = @DUMPBIN@ ECHO_C = @ECHO_C@ ECHO_N = @ECHO_N@ ECHO_T = @ECHO_T@ EGREP = @EGREP@ EXEEXT = @EXEEXT@ F77 = @F77@ FFLAGS = @FFLAGS@ FGREP = @FGREP@ FLIBS = @FLIBS@ GREP = @GREP@ INSTALL = @INSTALL@ INSTALL_DATA = @INSTALL_DATA@ INSTALL_PROGRAM = @INSTALL_PROGRAM@ INSTALL_SCRIPT = @INSTALL_SCRIPT@ INSTALL_STRIP_PROGRAM = @INSTALL_STRIP_PROGRAM@ LD = @LD@ LDFLAGS = @LDFLAGS@ LIBOBJS = @LIBOBJS@ LIBQUADMATH = @LIBQUADMATH@ LIBS = @LIBS@ LIBTOOL = @LIBTOOL@ LIPO = @LIPO@ LN_S = @LN_S@ LTLIBOBJS = @LTLIBOBJS@ MAINT = @MAINT@ MAKEINFO = @MAKEINFO@ MANIFEST_TOOL = @MANIFEST_TOOL@ MKDIR_P = @MKDIR_P@ MPICC = @MPICC@ MPILIBS = @MPILIBS@ MPIRUN = @MPIRUN@ NEON_CFLAGS = @NEON_CFLAGS@ NM = @NM@ NMEDIT = @NMEDIT@ OBJDUMP = @OBJDUMP@ OBJEXT = @OBJEXT@ OCAMLBUILD = @OCAMLBUILD@ OPENMP_CFLAGS = @OPENMP_CFLAGS@ OTOOL = @OTOOL@ OTOOL64 = @OTOOL64@ PACKAGE = @PACKAGE@ PACKAGE_BUGREPORT = @PACKAGE_BUGREPORT@ PACKAGE_NAME = @PACKAGE_NAME@ PACKAGE_STRING = @PACKAGE_STRING@ PACKAGE_TARNAME = @PACKAGE_TARNAME@ PACKAGE_URL = @PACKAGE_URL@ PACKAGE_VERSION = @PACKAGE_VERSION@ PATH_SEPARATOR = @PATH_SEPARATOR@ POW_LIB = @POW_LIB@ PRECISION = @PRECISION@ PREC_SUFFIX = @PREC_SUFFIX@ PTHREAD_CC = @PTHREAD_CC@ PTHREAD_CFLAGS = @PTHREAD_CFLAGS@ PTHREAD_LIBS = @PTHREAD_LIBS@ RANLIB = @RANLIB@ SED = @SED@ SET_MAKE = @SET_MAKE@ SHARED_VERSION_INFO = @SHARED_VERSION_INFO@ SHELL = @SHELL@ SSE2_CFLAGS = @SSE2_CFLAGS@ STACK_ALIGN_CFLAGS = @STACK_ALIGN_CFLAGS@ STRIP = @STRIP@ THREADLIBS = @THREADLIBS@ VERSION = @VERSION@ abs_builddir = @abs_builddir@ abs_srcdir = @abs_srcdir@ abs_top_builddir = @abs_top_builddir@ abs_top_srcdir = @abs_top_srcdir@ ac_ct_AR = @ac_ct_AR@ ac_ct_CC = @ac_ct_CC@ ac_ct_DUMPBIN = @ac_ct_DUMPBIN@ ac_ct_F77 = @ac_ct_F77@ acx_pthread_config = @acx_pthread_config@ am__include = @am__include@ am__leading_dot = @am__leading_dot@ am__quote = @am__quote@ am__tar = @am__tar@ am__untar = @am__untar@ bindir = @bindir@ build = @build@ build_alias = @build_alias@ build_cpu = @build_cpu@ build_os = @build_os@ build_vendor = @build_vendor@ builddir = @builddir@ datadir = @datadir@ datarootdir = @datarootdir@ docdir = @docdir@ dvidir = @dvidir@ exec_prefix = @exec_prefix@ host = @host@ host_alias = @host_alias@ host_cpu = @host_cpu@ host_os = @host_os@ host_vendor = @host_vendor@ htmldir = @htmldir@ includedir = @includedir@ infodir = @infodir@ install_sh = @install_sh@ libdir = @libdir@ libexecdir = @libexecdir@ localedir = @localedir@ localstatedir = @localstatedir@ mandir = @mandir@ mkdir_p = @mkdir_p@ oldincludedir = @oldincludedir@ pdfdir = @pdfdir@ prefix = @prefix@ program_transform_name = @program_transform_name@ psdir = @psdir@ sbindir = @sbindir@ sharedstatedir = @sharedstatedir@ srcdir = @srcdir@ sysconfdir = @sysconfdir@ target_alias = @target_alias@ top_build_prefix = @top_build_prefix@ top_builddir = @top_builddir@ top_srcdir = @top_srcdir@ AM_CFLAGS = $(NEON_CFLAGS) SIMD_HEADER = simd-neon.h HC2CFDFTV = hc2cfdftv_2.c hc2cfdftv_4.c hc2cfdftv_6.c hc2cfdftv_8.c \ hc2cfdftv_10.c hc2cfdftv_12.c hc2cfdftv_16.c hc2cfdftv_32.c \ hc2cfdftv_20.c HC2CBDFTV = hc2cbdftv_2.c hc2cbdftv_4.c hc2cbdftv_6.c hc2cbdftv_8.c \ hc2cbdftv_10.c hc2cbdftv_12.c hc2cbdftv_16.c hc2cbdftv_32.c \ hc2cbdftv_20.c ########################################################################### SIMD_CODELETS = $(HC2CFDFTV) $(HC2CBDFTV) AM_CPPFLAGS = -I$(top_srcdir)/kernel -I$(top_srcdir)/rdft \ -I$(top_srcdir)/rdft/simd -I$(top_srcdir)/simd-support EXTRA_DIST = $(SIMD_CODELETS) genus.c codlist.c @HAVE_NEON_TRUE@noinst_LTLIBRARIES = librdft_neon_codelets.la @HAVE_NEON_TRUE@BUILT_SOURCES = $(EXTRA_DIST) @HAVE_NEON_TRUE@librdft_neon_codelets_la_SOURCES = $(BUILT_SOURCES) all: $(BUILT_SOURCES) $(MAKE) $(AM_MAKEFLAGS) all-am .SUFFIXES: .SUFFIXES: .c .lo .o .obj $(srcdir)/Makefile.in: @MAINTAINER_MODE_TRUE@ $(srcdir)/Makefile.am $(top_srcdir)/rdft/simd/codlist.mk $(top_srcdir)/rdft/simd/simd.mk $(am__configure_deps) @for dep in $?; do \ case '$(am__configure_deps)' in \ *$$dep*) \ ( cd $(top_builddir) && $(MAKE) $(AM_MAKEFLAGS) am--refresh ) \ && { if test -f $@; then exit 0; else break; fi; }; \ exit 1;; \ esac; \ done; \ echo ' cd $(top_srcdir) && $(AUTOMAKE) --gnu rdft/simd/neon/Makefile'; \ $(am__cd) $(top_srcdir) && \ $(AUTOMAKE) --gnu rdft/simd/neon/Makefile .PRECIOUS: Makefile Makefile: $(srcdir)/Makefile.in $(top_builddir)/config.status @case '$?' in \ *config.status*) \ cd $(top_builddir) && $(MAKE) $(AM_MAKEFLAGS) am--refresh;; \ *) \ echo ' cd $(top_builddir) && $(SHELL) ./config.status $(subdir)/$@ $(am__depfiles_maybe)'; \ cd $(top_builddir) && $(SHELL) ./config.status $(subdir)/$@ $(am__depfiles_maybe);; \ esac; $(top_srcdir)/rdft/simd/codlist.mk $(top_srcdir)/rdft/simd/simd.mk: $(top_builddir)/config.status: $(top_srcdir)/configure $(CONFIG_STATUS_DEPENDENCIES) cd $(top_builddir) && $(MAKE) $(AM_MAKEFLAGS) am--refresh $(top_srcdir)/configure: @MAINTAINER_MODE_TRUE@ $(am__configure_deps) cd $(top_builddir) && $(MAKE) $(AM_MAKEFLAGS) am--refresh $(ACLOCAL_M4): @MAINTAINER_MODE_TRUE@ $(am__aclocal_m4_deps) cd $(top_builddir) && $(MAKE) $(AM_MAKEFLAGS) am--refresh $(am__aclocal_m4_deps): clean-noinstLTLIBRARIES: -test -z "$(noinst_LTLIBRARIES)" || rm -f $(noinst_LTLIBRARIES) @list='$(noinst_LTLIBRARIES)'; for p in $$list; do \ dir="`echo $$p | sed -e 's|/[^/]*$$||'`"; \ test "$$dir" != "$$p" || dir=.; \ echo "rm -f \"$${dir}/so_locations\""; \ rm -f "$${dir}/so_locations"; \ done librdft_neon_codelets.la: $(librdft_neon_codelets_la_OBJECTS) $(librdft_neon_codelets_la_DEPENDENCIES) $(EXTRA_librdft_neon_codelets_la_DEPENDENCIES) $(LINK) $(am_librdft_neon_codelets_la_rpath) $(librdft_neon_codelets_la_OBJECTS) $(librdft_neon_codelets_la_LIBADD) $(LIBS) mostlyclean-compile: -rm -f *.$(OBJEXT) distclean-compile: -rm -f *.tab.c @AMDEP_TRUE@@am__include@ @am__quote@./$(DEPDIR)/codlist.Plo@am__quote@ @AMDEP_TRUE@@am__include@ @am__quote@./$(DEPDIR)/genus.Plo@am__quote@ @AMDEP_TRUE@@am__include@ @am__quote@./$(DEPDIR)/hc2cbdftv_10.Plo@am__quote@ @AMDEP_TRUE@@am__include@ @am__quote@./$(DEPDIR)/hc2cbdftv_12.Plo@am__quote@ @AMDEP_TRUE@@am__include@ @am__quote@./$(DEPDIR)/hc2cbdftv_16.Plo@am__quote@ @AMDEP_TRUE@@am__include@ @am__quote@./$(DEPDIR)/hc2cbdftv_2.Plo@am__quote@ @AMDEP_TRUE@@am__include@ @am__quote@./$(DEPDIR)/hc2cbdftv_20.Plo@am__quote@ @AMDEP_TRUE@@am__include@ @am__quote@./$(DEPDIR)/hc2cbdftv_32.Plo@am__quote@ @AMDEP_TRUE@@am__include@ @am__quote@./$(DEPDIR)/hc2cbdftv_4.Plo@am__quote@ @AMDEP_TRUE@@am__include@ @am__quote@./$(DEPDIR)/hc2cbdftv_6.Plo@am__quote@ @AMDEP_TRUE@@am__include@ @am__quote@./$(DEPDIR)/hc2cbdftv_8.Plo@am__quote@ @AMDEP_TRUE@@am__include@ @am__quote@./$(DEPDIR)/hc2cfdftv_10.Plo@am__quote@ @AMDEP_TRUE@@am__include@ @am__quote@./$(DEPDIR)/hc2cfdftv_12.Plo@am__quote@ @AMDEP_TRUE@@am__include@ @am__quote@./$(DEPDIR)/hc2cfdftv_16.Plo@am__quote@ @AMDEP_TRUE@@am__include@ @am__quote@./$(DEPDIR)/hc2cfdftv_2.Plo@am__quote@ @AMDEP_TRUE@@am__include@ @am__quote@./$(DEPDIR)/hc2cfdftv_20.Plo@am__quote@ @AMDEP_TRUE@@am__include@ @am__quote@./$(DEPDIR)/hc2cfdftv_32.Plo@am__quote@ @AMDEP_TRUE@@am__include@ @am__quote@./$(DEPDIR)/hc2cfdftv_4.Plo@am__quote@ @AMDEP_TRUE@@am__include@ @am__quote@./$(DEPDIR)/hc2cfdftv_6.Plo@am__quote@ @AMDEP_TRUE@@am__include@ @am__quote@./$(DEPDIR)/hc2cfdftv_8.Plo@am__quote@ .c.o: @am__fastdepCC_TRUE@ $(COMPILE) -MT $@ -MD -MP -MF $(DEPDIR)/$*.Tpo -c -o $@ $< @am__fastdepCC_TRUE@ $(am__mv) $(DEPDIR)/$*.Tpo $(DEPDIR)/$*.Po @AMDEP_TRUE@@am__fastdepCC_FALSE@ source='$<' object='$@' libtool=no @AMDEPBACKSLASH@ @AMDEP_TRUE@@am__fastdepCC_FALSE@ DEPDIR=$(DEPDIR) $(CCDEPMODE) $(depcomp) @AMDEPBACKSLASH@ @am__fastdepCC_FALSE@ $(COMPILE) -c $< .c.obj: @am__fastdepCC_TRUE@ $(COMPILE) -MT $@ -MD -MP -MF $(DEPDIR)/$*.Tpo -c -o $@ `$(CYGPATH_W) '$<'` @am__fastdepCC_TRUE@ $(am__mv) $(DEPDIR)/$*.Tpo $(DEPDIR)/$*.Po @AMDEP_TRUE@@am__fastdepCC_FALSE@ source='$<' object='$@' libtool=no @AMDEPBACKSLASH@ @AMDEP_TRUE@@am__fastdepCC_FALSE@ DEPDIR=$(DEPDIR) $(CCDEPMODE) $(depcomp) @AMDEPBACKSLASH@ @am__fastdepCC_FALSE@ $(COMPILE) -c `$(CYGPATH_W) '$<'` .c.lo: @am__fastdepCC_TRUE@ $(LTCOMPILE) -MT $@ -MD -MP -MF $(DEPDIR)/$*.Tpo -c -o $@ $< @am__fastdepCC_TRUE@ $(am__mv) $(DEPDIR)/$*.Tpo $(DEPDIR)/$*.Plo @AMDEP_TRUE@@am__fastdepCC_FALSE@ source='$<' object='$@' libtool=yes @AMDEPBACKSLASH@ @AMDEP_TRUE@@am__fastdepCC_FALSE@ DEPDIR=$(DEPDIR) $(CCDEPMODE) $(depcomp) @AMDEPBACKSLASH@ @am__fastdepCC_FALSE@ $(LTCOMPILE) -c -o $@ $< mostlyclean-libtool: -rm -f *.lo clean-libtool: -rm -rf .libs _libs ID: $(HEADERS) $(SOURCES) $(LISP) $(TAGS_FILES) list='$(SOURCES) $(HEADERS) $(LISP) $(TAGS_FILES)'; \ unique=`for i in $$list; do \ if test -f "$$i"; then echo $$i; else echo $(srcdir)/$$i; fi; \ done | \ $(AWK) '{ files[$$0] = 1; nonempty = 1; } \ END { if (nonempty) { for (i in files) print i; }; }'`; \ mkid -fID $$unique tags: TAGS TAGS: $(HEADERS) $(SOURCES) $(TAGS_DEPENDENCIES) \ $(TAGS_FILES) $(LISP) set x; \ here=`pwd`; \ list='$(SOURCES) $(HEADERS) $(LISP) $(TAGS_FILES)'; \ unique=`for i in $$list; do \ if test -f "$$i"; then echo $$i; else echo $(srcdir)/$$i; fi; \ done | \ $(AWK) '{ files[$$0] = 1; nonempty = 1; } \ END { if (nonempty) { for (i in files) print i; }; }'`; \ shift; \ if test -z "$(ETAGS_ARGS)$$*$$unique"; then :; else \ test -n "$$unique" || unique=$$empty_fix; \ if test $$# -gt 0; then \ $(ETAGS) $(ETAGSFLAGS) $(AM_ETAGSFLAGS) $(ETAGS_ARGS) \ "$$@" $$unique; \ else \ $(ETAGS) $(ETAGSFLAGS) $(AM_ETAGSFLAGS) $(ETAGS_ARGS) \ $$unique; \ fi; \ fi ctags: CTAGS CTAGS: $(HEADERS) $(SOURCES) $(TAGS_DEPENDENCIES) \ $(TAGS_FILES) $(LISP) list='$(SOURCES) $(HEADERS) $(LISP) $(TAGS_FILES)'; \ unique=`for i in $$list; do \ if test -f "$$i"; then echo $$i; else echo $(srcdir)/$$i; fi; \ done | \ $(AWK) '{ files[$$0] = 1; nonempty = 1; } \ END { if (nonempty) { for (i in files) print i; }; }'`; \ test -z "$(CTAGS_ARGS)$$unique" \ || $(CTAGS) $(CTAGSFLAGS) $(AM_CTAGSFLAGS) $(CTAGS_ARGS) \ $$unique GTAGS: here=`$(am__cd) $(top_builddir) && pwd` \ && $(am__cd) $(top_srcdir) \ && gtags -i $(GTAGS_ARGS) "$$here" distclean-tags: -rm -f TAGS ID GTAGS GRTAGS GSYMS GPATH tags distdir: $(DISTFILES) @srcdirstrip=`echo "$(srcdir)" | sed 's/[].[^$$\\*]/\\\\&/g'`; \ topsrcdirstrip=`echo "$(top_srcdir)" | sed 's/[].[^$$\\*]/\\\\&/g'`; \ list='$(DISTFILES)'; \ dist_files=`for file in $$list; do echo $$file; done | \ sed -e "s|^$$srcdirstrip/||;t" \ -e "s|^$$topsrcdirstrip/|$(top_builddir)/|;t"`; \ case $$dist_files in \ */*) $(MKDIR_P) `echo "$$dist_files" | \ sed '/\//!d;s|^|$(distdir)/|;s,/[^/]*$$,,' | \ sort -u` ;; \ esac; \ for file in $$dist_files; do \ if test -f $$file || test -d $$file; then d=.; else d=$(srcdir); fi; \ if test -d $$d/$$file; then \ dir=`echo "/$$file" | sed -e 's,/[^/]*$$,,'`; \ if test -d "$(distdir)/$$file"; then \ find "$(distdir)/$$file" -type d ! -perm -700 -exec chmod u+rwx {} \;; \ fi; \ if test -d $(srcdir)/$$file && test $$d != $(srcdir); then \ cp -fpR $(srcdir)/$$file "$(distdir)$$dir" || exit 1; \ find "$(distdir)/$$file" -type d ! -perm -700 -exec chmod u+rwx {} \;; \ fi; \ cp -fpR $$d/$$file "$(distdir)$$dir" || exit 1; \ else \ test -f "$(distdir)/$$file" \ || cp -p $$d/$$file "$(distdir)/$$file" \ || exit 1; \ fi; \ done check-am: all-am check: $(BUILT_SOURCES) $(MAKE) $(AM_MAKEFLAGS) check-am all-am: Makefile $(LTLIBRARIES) installdirs: install: $(BUILT_SOURCES) $(MAKE) $(AM_MAKEFLAGS) install-am install-exec: install-exec-am install-data: install-data-am uninstall: uninstall-am install-am: all-am @$(MAKE) $(AM_MAKEFLAGS) install-exec-am install-data-am installcheck: installcheck-am install-strip: if test -z '$(STRIP)'; then \ $(MAKE) $(AM_MAKEFLAGS) INSTALL_PROGRAM="$(INSTALL_STRIP_PROGRAM)" \ install_sh_PROGRAM="$(INSTALL_STRIP_PROGRAM)" INSTALL_STRIP_FLAG=-s \ install; \ else \ $(MAKE) $(AM_MAKEFLAGS) INSTALL_PROGRAM="$(INSTALL_STRIP_PROGRAM)" \ install_sh_PROGRAM="$(INSTALL_STRIP_PROGRAM)" INSTALL_STRIP_FLAG=-s \ "INSTALL_PROGRAM_ENV=STRIPPROG='$(STRIP)'" install; \ fi mostlyclean-generic: clean-generic: distclean-generic: -test -z "$(CONFIG_CLEAN_FILES)" || rm -f $(CONFIG_CLEAN_FILES) -test . = "$(srcdir)" || test -z "$(CONFIG_CLEAN_VPATH_FILES)" || rm -f $(CONFIG_CLEAN_VPATH_FILES) maintainer-clean-generic: @echo "This command is intended for maintainers to use" @echo "it deletes files that may require special tools to rebuild." -test -z "$(BUILT_SOURCES)" || rm -f $(BUILT_SOURCES) clean: clean-am clean-am: clean-generic clean-libtool clean-noinstLTLIBRARIES \ mostlyclean-am distclean: distclean-am -rm -rf ./$(DEPDIR) -rm -f Makefile distclean-am: clean-am distclean-compile distclean-generic \ distclean-tags dvi: dvi-am dvi-am: html: html-am html-am: info: info-am info-am: install-data-am: install-dvi: install-dvi-am install-dvi-am: install-exec-am: install-html: install-html-am install-html-am: install-info: install-info-am install-info-am: install-man: install-pdf: install-pdf-am install-pdf-am: install-ps: install-ps-am install-ps-am: installcheck-am: maintainer-clean: maintainer-clean-am -rm -rf ./$(DEPDIR) -rm -f Makefile maintainer-clean-am: distclean-am maintainer-clean-generic mostlyclean: mostlyclean-am mostlyclean-am: mostlyclean-compile mostlyclean-generic \ mostlyclean-libtool pdf: pdf-am pdf-am: ps: ps-am ps-am: uninstall-am: .MAKE: all check install install-am install-strip .PHONY: CTAGS GTAGS all all-am check check-am clean clean-generic \ clean-libtool clean-noinstLTLIBRARIES ctags distclean \ distclean-compile distclean-generic distclean-libtool \ distclean-tags distdir dvi dvi-am html html-am info info-am \ install install-am install-data install-data-am install-dvi \ install-dvi-am install-exec install-exec-am install-html \ install-html-am install-info install-info-am install-man \ install-pdf install-pdf-am install-ps install-ps-am \ install-strip installcheck installcheck-am installdirs \ maintainer-clean maintainer-clean-generic mostlyclean \ mostlyclean-compile mostlyclean-generic mostlyclean-libtool \ pdf pdf-am ps ps-am tags uninstall uninstall-am $(EXTRA_DIST): Makefile ( \ echo "/* Generated automatically. DO NOT EDIT! */"; \ echo "#define SIMD_HEADER \"$(SIMD_HEADER)\""; \ echo "#include \"../common/"$*".c\""; \ ) >$@ # Tell versions [3.59,3.63) of GNU make to not export all variables. # Otherwise a system limit (for SysV at least) may be exceeded. .NOEXPORT: fftw-3.3.3/rdft/simd/neon/hc2cfdftv_2.c0000644000175000017500000000016212054411110014463 00000000000000/* Generated automatically. DO NOT EDIT! */ #define SIMD_HEADER "simd-neon.h" #include "../common/hc2cfdftv_2.c" fftw-3.3.3/rdft/simd/neon/hc2cfdftv_4.c0000644000175000017500000000016212054411110014465 00000000000000/* Generated automatically. DO NOT EDIT! */ #define SIMD_HEADER "simd-neon.h" #include "../common/hc2cfdftv_4.c" fftw-3.3.3/rdft/simd/neon/hc2cfdftv_6.c0000644000175000017500000000016212054411110014467 00000000000000/* Generated automatically. DO NOT EDIT! */ #define SIMD_HEADER "simd-neon.h" #include "../common/hc2cfdftv_6.c" fftw-3.3.3/rdft/simd/neon/hc2cfdftv_8.c0000644000175000017500000000016212054411110014471 00000000000000/* Generated automatically. DO NOT EDIT! */ #define SIMD_HEADER "simd-neon.h" #include "../common/hc2cfdftv_8.c" fftw-3.3.3/rdft/simd/neon/hc2cfdftv_10.c0000644000175000017500000000016312054411110014543 00000000000000/* Generated automatically. DO NOT EDIT! */ #define SIMD_HEADER "simd-neon.h" #include "../common/hc2cfdftv_10.c" fftw-3.3.3/rdft/simd/neon/hc2cfdftv_12.c0000644000175000017500000000016312054411110014545 00000000000000/* Generated automatically. DO NOT EDIT! */ #define SIMD_HEADER "simd-neon.h" #include "../common/hc2cfdftv_12.c" fftw-3.3.3/rdft/simd/neon/hc2cfdftv_16.c0000644000175000017500000000016312054411110014551 00000000000000/* Generated automatically. DO NOT EDIT! */ #define SIMD_HEADER "simd-neon.h" #include "../common/hc2cfdftv_16.c" fftw-3.3.3/rdft/simd/neon/hc2cfdftv_32.c0000644000175000017500000000016312054411110014547 00000000000000/* Generated automatically. DO NOT EDIT! */ #define SIMD_HEADER "simd-neon.h" #include "../common/hc2cfdftv_32.c" fftw-3.3.3/rdft/simd/neon/hc2cfdftv_20.c0000644000175000017500000000016312054411110014544 00000000000000/* Generated automatically. DO NOT EDIT! */ #define SIMD_HEADER "simd-neon.h" #include "../common/hc2cfdftv_20.c" fftw-3.3.3/rdft/simd/neon/hc2cbdftv_2.c0000644000175000017500000000016212054411110014457 00000000000000/* Generated automatically. DO NOT EDIT! */ #define SIMD_HEADER "simd-neon.h" #include "../common/hc2cbdftv_2.c" fftw-3.3.3/rdft/simd/neon/hc2cbdftv_4.c0000644000175000017500000000016212054411110014461 00000000000000/* Generated automatically. DO NOT EDIT! */ #define SIMD_HEADER "simd-neon.h" #include "../common/hc2cbdftv_4.c" fftw-3.3.3/rdft/simd/neon/hc2cbdftv_6.c0000644000175000017500000000016212054411110014463 00000000000000/* Generated automatically. DO NOT EDIT! */ #define SIMD_HEADER "simd-neon.h" #include "../common/hc2cbdftv_6.c" fftw-3.3.3/rdft/simd/neon/hc2cbdftv_8.c0000644000175000017500000000016212054411110014465 00000000000000/* Generated automatically. DO NOT EDIT! */ #define SIMD_HEADER "simd-neon.h" #include "../common/hc2cbdftv_8.c" fftw-3.3.3/rdft/simd/neon/hc2cbdftv_10.c0000644000175000017500000000016312054411110014537 00000000000000/* Generated automatically. DO NOT EDIT! */ #define SIMD_HEADER "simd-neon.h" #include "../common/hc2cbdftv_10.c" fftw-3.3.3/rdft/simd/neon/hc2cbdftv_12.c0000644000175000017500000000016312054411110014541 00000000000000/* Generated automatically. DO NOT EDIT! */ #define SIMD_HEADER "simd-neon.h" #include "../common/hc2cbdftv_12.c" fftw-3.3.3/rdft/simd/neon/hc2cbdftv_16.c0000644000175000017500000000016312054411110014545 00000000000000/* Generated automatically. DO NOT EDIT! */ #define SIMD_HEADER "simd-neon.h" #include "../common/hc2cbdftv_16.c" fftw-3.3.3/rdft/simd/neon/hc2cbdftv_32.c0000644000175000017500000000016312054411110014543 00000000000000/* Generated automatically. DO NOT EDIT! */ #define SIMD_HEADER "simd-neon.h" #include "../common/hc2cbdftv_32.c" fftw-3.3.3/rdft/simd/neon/hc2cbdftv_20.c0000644000175000017500000000016312054411110014540 00000000000000/* Generated automatically. DO NOT EDIT! */ #define SIMD_HEADER "simd-neon.h" #include "../common/hc2cbdftv_20.c" fftw-3.3.3/rdft/simd/neon/genus.c0000644000175000017500000000015412054411110013513 00000000000000/* Generated automatically. DO NOT EDIT! */ #define SIMD_HEADER "simd-neon.h" #include "../common/genus.c" fftw-3.3.3/rdft/simd/neon/codlist.c0000644000175000017500000000015612054411110014035 00000000000000/* Generated automatically. DO NOT EDIT! */ #define SIMD_HEADER "simd-neon.h" #include "../common/codlist.c" fftw-3.3.3/reodft/0000755000175000017500000000000012054411110010737 500000000000000fftw-3.3.3/reodft/Makefile.am0000644000175000017500000000077612054410050012727 00000000000000AM_CPPFLAGS = -I$(top_srcdir)/kernel -I$(top_srcdir)/rdft SUBDIRS = noinst_LTLIBRARIES = libreodft.la # pkgincludedir = $(includedir)/fftw3@PREC_SUFFIX@ # pkginclude_HEADERS = reodft.h # no longer used due to numerical problems EXTRA_DIST = reodft11e-r2hc.c redft00e-r2hc.c rodft00e-r2hc.c libreodft_la_SOURCES = conf.c reodft.h reodft010e-r2hc.c \ reodft11e-radix2.c reodft11e-r2hc-odd.c redft00e-r2hc-pad.c \ rodft00e-r2hc-pad.c reodft00e-splitradix.c # redft00e-r2hc.c rodft00e-r2hc.c reodft11e-r2hc.c fftw-3.3.3/reodft/Makefile.in0000644000175000017500000005372612054410200012740 00000000000000# Makefile.in generated by automake 1.11.6 from Makefile.am. # @configure_input@ # Copyright (C) 1994, 1995, 1996, 1997, 1998, 1999, 2000, 2001, 2002, # 2003, 2004, 2005, 2006, 2007, 2008, 2009, 2010, 2011 Free Software # Foundation, Inc. # This Makefile.in is free software; the Free Software Foundation # gives unlimited permission to copy and/or distribute it, # with or without modifications, as long as this notice is preserved. # This program is distributed in the hope that it will be useful, # but WITHOUT ANY WARRANTY, to the extent permitted by law; without # even the implied warranty of MERCHANTABILITY or FITNESS FOR A # PARTICULAR PURPOSE. @SET_MAKE@ VPATH = @srcdir@ am__make_dryrun = \ { \ am__dry=no; \ case $$MAKEFLAGS in \ *\\[\ \ ]*) \ echo 'am--echo: ; @echo "AM" OK' | $(MAKE) -f - 2>/dev/null \ | grep '^AM OK$$' >/dev/null || am__dry=yes;; \ *) \ for am__flg in $$MAKEFLAGS; do \ case $$am__flg in \ *=*|--*) ;; \ *n*) am__dry=yes; break;; \ esac; \ done;; \ esac; \ test $$am__dry = yes; \ } pkgdatadir = $(datadir)/@PACKAGE@ pkgincludedir = $(includedir)/@PACKAGE@ pkglibdir = $(libdir)/@PACKAGE@ pkglibexecdir = $(libexecdir)/@PACKAGE@ am__cd = CDPATH="$${ZSH_VERSION+.}$(PATH_SEPARATOR)" && cd install_sh_DATA = $(install_sh) -c -m 644 install_sh_PROGRAM = $(install_sh) -c install_sh_SCRIPT = $(install_sh) -c INSTALL_HEADER = $(INSTALL_DATA) transform = $(program_transform_name) NORMAL_INSTALL = : PRE_INSTALL = : POST_INSTALL = : NORMAL_UNINSTALL = : PRE_UNINSTALL = : POST_UNINSTALL = : build_triplet = @build@ host_triplet = @host@ subdir = reodft DIST_COMMON = $(srcdir)/Makefile.am $(srcdir)/Makefile.in ACLOCAL_M4 = $(top_srcdir)/aclocal.m4 am__aclocal_m4_deps = $(top_srcdir)/m4/acx_mpi.m4 \ $(top_srcdir)/m4/acx_pthread.m4 \ $(top_srcdir)/m4/ax_cc_maxopt.m4 \ $(top_srcdir)/m4/ax_check_compiler_flags.m4 \ $(top_srcdir)/m4/ax_compiler_vendor.m4 \ $(top_srcdir)/m4/ax_gcc_aligns_stack.m4 \ $(top_srcdir)/m4/ax_gcc_version.m4 \ $(top_srcdir)/m4/ax_openmp.m4 $(top_srcdir)/m4/libtool.m4 \ $(top_srcdir)/m4/ltoptions.m4 $(top_srcdir)/m4/ltsugar.m4 \ $(top_srcdir)/m4/ltversion.m4 $(top_srcdir)/m4/lt~obsolete.m4 \ $(top_srcdir)/configure.ac am__configure_deps = $(am__aclocal_m4_deps) $(CONFIGURE_DEPENDENCIES) \ $(ACLOCAL_M4) mkinstalldirs = $(install_sh) -d CONFIG_HEADER = $(top_builddir)/config.h CONFIG_CLEAN_FILES = CONFIG_CLEAN_VPATH_FILES = LTLIBRARIES = $(noinst_LTLIBRARIES) libreodft_la_LIBADD = am_libreodft_la_OBJECTS = conf.lo reodft010e-r2hc.lo \ reodft11e-radix2.lo reodft11e-r2hc-odd.lo redft00e-r2hc-pad.lo \ rodft00e-r2hc-pad.lo reodft00e-splitradix.lo libreodft_la_OBJECTS = $(am_libreodft_la_OBJECTS) DEFAULT_INCLUDES = -I.@am__isrc@ -I$(top_builddir) depcomp = $(SHELL) $(top_srcdir)/depcomp am__depfiles_maybe = depfiles am__mv = mv -f COMPILE = $(CC) $(DEFS) $(DEFAULT_INCLUDES) $(INCLUDES) $(AM_CPPFLAGS) \ $(CPPFLAGS) $(AM_CFLAGS) $(CFLAGS) LTCOMPILE = $(LIBTOOL) --tag=CC $(AM_LIBTOOLFLAGS) $(LIBTOOLFLAGS) \ --mode=compile $(CC) $(DEFS) $(DEFAULT_INCLUDES) $(INCLUDES) \ $(AM_CPPFLAGS) $(CPPFLAGS) $(AM_CFLAGS) $(CFLAGS) CCLD = $(CC) LINK = $(LIBTOOL) --tag=CC $(AM_LIBTOOLFLAGS) $(LIBTOOLFLAGS) \ --mode=link $(CCLD) $(AM_CFLAGS) $(CFLAGS) $(AM_LDFLAGS) \ $(LDFLAGS) -o $@ SOURCES = $(libreodft_la_SOURCES) DIST_SOURCES = $(libreodft_la_SOURCES) RECURSIVE_TARGETS = all-recursive check-recursive dvi-recursive \ html-recursive info-recursive install-data-recursive \ install-dvi-recursive install-exec-recursive \ install-html-recursive install-info-recursive \ install-pdf-recursive install-ps-recursive install-recursive \ installcheck-recursive installdirs-recursive pdf-recursive \ ps-recursive uninstall-recursive am__can_run_installinfo = \ case $$AM_UPDATE_INFO_DIR in \ n|no|NO) false;; \ *) (install-info --version) >/dev/null 2>&1;; \ esac RECURSIVE_CLEAN_TARGETS = mostlyclean-recursive clean-recursive \ distclean-recursive maintainer-clean-recursive AM_RECURSIVE_TARGETS = $(RECURSIVE_TARGETS:-recursive=) \ $(RECURSIVE_CLEAN_TARGETS:-recursive=) tags TAGS ctags CTAGS \ distdir ETAGS = etags CTAGS = ctags DIST_SUBDIRS = $(SUBDIRS) DISTFILES = $(DIST_COMMON) $(DIST_SOURCES) $(TEXINFOS) $(EXTRA_DIST) am__relativize = \ dir0=`pwd`; \ sed_first='s,^\([^/]*\)/.*$$,\1,'; \ sed_rest='s,^[^/]*/*,,'; \ sed_last='s,^.*/\([^/]*\)$$,\1,'; \ sed_butlast='s,/*[^/]*$$,,'; \ while test -n "$$dir1"; do \ first=`echo "$$dir1" | sed -e "$$sed_first"`; \ if test "$$first" != "."; then \ if test "$$first" = ".."; then \ dir2=`echo "$$dir0" | sed -e "$$sed_last"`/"$$dir2"; \ dir0=`echo "$$dir0" | sed -e "$$sed_butlast"`; \ else \ first2=`echo "$$dir2" | sed -e "$$sed_first"`; \ if test "$$first2" = "$$first"; then \ dir2=`echo "$$dir2" | sed -e "$$sed_rest"`; \ else \ dir2="../$$dir2"; \ fi; \ dir0="$$dir0"/"$$first"; \ fi; \ fi; \ dir1=`echo "$$dir1" | sed -e "$$sed_rest"`; \ done; \ reldir="$$dir2" ACLOCAL = @ACLOCAL@ ALLOCA = @ALLOCA@ ALTIVEC_CFLAGS = @ALTIVEC_CFLAGS@ AMTAR = @AMTAR@ AR = @AR@ AS = @AS@ AUTOCONF = @AUTOCONF@ AUTOHEADER = @AUTOHEADER@ AUTOMAKE = @AUTOMAKE@ AVX_CFLAGS = @AVX_CFLAGS@ AWK = @AWK@ CC = @CC@ CCDEPMODE = @CCDEPMODE@ CFLAGS = @CFLAGS@ CHECK_PL_OPTS = @CHECK_PL_OPTS@ CPP = @CPP@ CPPFLAGS = @CPPFLAGS@ CYGPATH_W = @CYGPATH_W@ C_FFTW_R2R_KIND = @C_FFTW_R2R_KIND@ C_MPI_FINT = @C_MPI_FINT@ DEFS = @DEFS@ DEPDIR = @DEPDIR@ DLLTOOL = @DLLTOOL@ DSYMUTIL = @DSYMUTIL@ DUMPBIN = @DUMPBIN@ ECHO_C = @ECHO_C@ ECHO_N = @ECHO_N@ ECHO_T = @ECHO_T@ EGREP = @EGREP@ EXEEXT = @EXEEXT@ F77 = @F77@ FFLAGS = @FFLAGS@ FGREP = @FGREP@ FLIBS = @FLIBS@ GREP = @GREP@ INSTALL = @INSTALL@ INSTALL_DATA = @INSTALL_DATA@ INSTALL_PROGRAM = @INSTALL_PROGRAM@ INSTALL_SCRIPT = @INSTALL_SCRIPT@ INSTALL_STRIP_PROGRAM = @INSTALL_STRIP_PROGRAM@ LD = @LD@ LDFLAGS = @LDFLAGS@ LIBOBJS = @LIBOBJS@ LIBQUADMATH = @LIBQUADMATH@ LIBS = @LIBS@ LIBTOOL = @LIBTOOL@ LIPO = @LIPO@ LN_S = @LN_S@ LTLIBOBJS = @LTLIBOBJS@ MAINT = @MAINT@ MAKEINFO = @MAKEINFO@ MANIFEST_TOOL = @MANIFEST_TOOL@ MKDIR_P = @MKDIR_P@ MPICC = @MPICC@ MPILIBS = @MPILIBS@ MPIRUN = @MPIRUN@ NEON_CFLAGS = @NEON_CFLAGS@ NM = @NM@ NMEDIT = @NMEDIT@ OBJDUMP = @OBJDUMP@ OBJEXT = @OBJEXT@ OCAMLBUILD = @OCAMLBUILD@ OPENMP_CFLAGS = @OPENMP_CFLAGS@ OTOOL = @OTOOL@ OTOOL64 = @OTOOL64@ PACKAGE = @PACKAGE@ PACKAGE_BUGREPORT = @PACKAGE_BUGREPORT@ PACKAGE_NAME = @PACKAGE_NAME@ PACKAGE_STRING = @PACKAGE_STRING@ PACKAGE_TARNAME = @PACKAGE_TARNAME@ PACKAGE_URL = @PACKAGE_URL@ PACKAGE_VERSION = @PACKAGE_VERSION@ PATH_SEPARATOR = @PATH_SEPARATOR@ POW_LIB = @POW_LIB@ PRECISION = @PRECISION@ PREC_SUFFIX = @PREC_SUFFIX@ PTHREAD_CC = @PTHREAD_CC@ PTHREAD_CFLAGS = @PTHREAD_CFLAGS@ PTHREAD_LIBS = @PTHREAD_LIBS@ RANLIB = @RANLIB@ SED = @SED@ SET_MAKE = @SET_MAKE@ SHARED_VERSION_INFO = @SHARED_VERSION_INFO@ SHELL = @SHELL@ SSE2_CFLAGS = @SSE2_CFLAGS@ STACK_ALIGN_CFLAGS = @STACK_ALIGN_CFLAGS@ STRIP = @STRIP@ THREADLIBS = @THREADLIBS@ VERSION = @VERSION@ abs_builddir = @abs_builddir@ abs_srcdir = @abs_srcdir@ abs_top_builddir = @abs_top_builddir@ abs_top_srcdir = @abs_top_srcdir@ ac_ct_AR = @ac_ct_AR@ ac_ct_CC = @ac_ct_CC@ ac_ct_DUMPBIN = @ac_ct_DUMPBIN@ ac_ct_F77 = @ac_ct_F77@ acx_pthread_config = @acx_pthread_config@ am__include = @am__include@ am__leading_dot = @am__leading_dot@ am__quote = @am__quote@ am__tar = @am__tar@ am__untar = @am__untar@ bindir = @bindir@ build = @build@ build_alias = @build_alias@ build_cpu = @build_cpu@ build_os = @build_os@ build_vendor = @build_vendor@ builddir = @builddir@ datadir = @datadir@ datarootdir = @datarootdir@ docdir = @docdir@ dvidir = @dvidir@ exec_prefix = @exec_prefix@ host = @host@ host_alias = @host_alias@ host_cpu = @host_cpu@ host_os = @host_os@ host_vendor = @host_vendor@ htmldir = @htmldir@ includedir = @includedir@ infodir = @infodir@ install_sh = @install_sh@ libdir = @libdir@ libexecdir = @libexecdir@ localedir = @localedir@ localstatedir = @localstatedir@ mandir = @mandir@ mkdir_p = @mkdir_p@ oldincludedir = @oldincludedir@ pdfdir = @pdfdir@ prefix = @prefix@ program_transform_name = @program_transform_name@ psdir = @psdir@ sbindir = @sbindir@ sharedstatedir = @sharedstatedir@ srcdir = @srcdir@ sysconfdir = @sysconfdir@ target_alias = @target_alias@ top_build_prefix = @top_build_prefix@ top_builddir = @top_builddir@ top_srcdir = @top_srcdir@ AM_CPPFLAGS = -I$(top_srcdir)/kernel -I$(top_srcdir)/rdft SUBDIRS = noinst_LTLIBRARIES = libreodft.la # pkgincludedir = $(includedir)/fftw3@PREC_SUFFIX@ # pkginclude_HEADERS = reodft.h # no longer used due to numerical problems EXTRA_DIST = reodft11e-r2hc.c redft00e-r2hc.c rodft00e-r2hc.c libreodft_la_SOURCES = conf.c reodft.h reodft010e-r2hc.c \ reodft11e-radix2.c reodft11e-r2hc-odd.c redft00e-r2hc-pad.c \ rodft00e-r2hc-pad.c reodft00e-splitradix.c all: all-recursive .SUFFIXES: .SUFFIXES: .c .lo .o .obj $(srcdir)/Makefile.in: @MAINTAINER_MODE_TRUE@ $(srcdir)/Makefile.am $(am__configure_deps) @for dep in $?; do \ case '$(am__configure_deps)' in \ *$$dep*) \ ( cd $(top_builddir) && $(MAKE) $(AM_MAKEFLAGS) am--refresh ) \ && { if test -f $@; then exit 0; else break; fi; }; \ exit 1;; \ esac; \ done; \ echo ' cd $(top_srcdir) && $(AUTOMAKE) --gnu reodft/Makefile'; \ $(am__cd) $(top_srcdir) && \ $(AUTOMAKE) --gnu reodft/Makefile .PRECIOUS: Makefile Makefile: $(srcdir)/Makefile.in $(top_builddir)/config.status @case '$?' in \ *config.status*) \ cd $(top_builddir) && $(MAKE) $(AM_MAKEFLAGS) am--refresh;; \ *) \ echo ' cd $(top_builddir) && $(SHELL) ./config.status $(subdir)/$@ $(am__depfiles_maybe)'; \ cd $(top_builddir) && $(SHELL) ./config.status $(subdir)/$@ $(am__depfiles_maybe);; \ esac; $(top_builddir)/config.status: $(top_srcdir)/configure $(CONFIG_STATUS_DEPENDENCIES) cd $(top_builddir) && $(MAKE) $(AM_MAKEFLAGS) am--refresh $(top_srcdir)/configure: @MAINTAINER_MODE_TRUE@ $(am__configure_deps) cd $(top_builddir) && $(MAKE) $(AM_MAKEFLAGS) am--refresh $(ACLOCAL_M4): @MAINTAINER_MODE_TRUE@ $(am__aclocal_m4_deps) cd $(top_builddir) && $(MAKE) $(AM_MAKEFLAGS) am--refresh $(am__aclocal_m4_deps): clean-noinstLTLIBRARIES: -test -z "$(noinst_LTLIBRARIES)" || rm -f $(noinst_LTLIBRARIES) @list='$(noinst_LTLIBRARIES)'; for p in $$list; do \ dir="`echo $$p | sed -e 's|/[^/]*$$||'`"; \ test "$$dir" != "$$p" || dir=.; \ echo "rm -f \"$${dir}/so_locations\""; \ rm -f "$${dir}/so_locations"; \ done libreodft.la: $(libreodft_la_OBJECTS) $(libreodft_la_DEPENDENCIES) $(EXTRA_libreodft_la_DEPENDENCIES) $(LINK) $(libreodft_la_OBJECTS) $(libreodft_la_LIBADD) $(LIBS) mostlyclean-compile: -rm -f *.$(OBJEXT) distclean-compile: -rm -f *.tab.c @AMDEP_TRUE@@am__include@ @am__quote@./$(DEPDIR)/conf.Plo@am__quote@ @AMDEP_TRUE@@am__include@ @am__quote@./$(DEPDIR)/redft00e-r2hc-pad.Plo@am__quote@ @AMDEP_TRUE@@am__include@ @am__quote@./$(DEPDIR)/reodft00e-splitradix.Plo@am__quote@ @AMDEP_TRUE@@am__include@ @am__quote@./$(DEPDIR)/reodft010e-r2hc.Plo@am__quote@ @AMDEP_TRUE@@am__include@ @am__quote@./$(DEPDIR)/reodft11e-r2hc-odd.Plo@am__quote@ @AMDEP_TRUE@@am__include@ @am__quote@./$(DEPDIR)/reodft11e-radix2.Plo@am__quote@ @AMDEP_TRUE@@am__include@ @am__quote@./$(DEPDIR)/rodft00e-r2hc-pad.Plo@am__quote@ .c.o: @am__fastdepCC_TRUE@ $(COMPILE) -MT $@ -MD -MP -MF $(DEPDIR)/$*.Tpo -c -o $@ $< @am__fastdepCC_TRUE@ $(am__mv) $(DEPDIR)/$*.Tpo $(DEPDIR)/$*.Po @AMDEP_TRUE@@am__fastdepCC_FALSE@ source='$<' object='$@' libtool=no @AMDEPBACKSLASH@ @AMDEP_TRUE@@am__fastdepCC_FALSE@ DEPDIR=$(DEPDIR) $(CCDEPMODE) $(depcomp) @AMDEPBACKSLASH@ @am__fastdepCC_FALSE@ $(COMPILE) -c $< .c.obj: @am__fastdepCC_TRUE@ $(COMPILE) -MT $@ -MD -MP -MF $(DEPDIR)/$*.Tpo -c -o $@ `$(CYGPATH_W) '$<'` @am__fastdepCC_TRUE@ $(am__mv) $(DEPDIR)/$*.Tpo $(DEPDIR)/$*.Po @AMDEP_TRUE@@am__fastdepCC_FALSE@ source='$<' object='$@' libtool=no @AMDEPBACKSLASH@ @AMDEP_TRUE@@am__fastdepCC_FALSE@ DEPDIR=$(DEPDIR) $(CCDEPMODE) $(depcomp) @AMDEPBACKSLASH@ @am__fastdepCC_FALSE@ $(COMPILE) -c `$(CYGPATH_W) '$<'` .c.lo: @am__fastdepCC_TRUE@ $(LTCOMPILE) -MT $@ -MD -MP -MF $(DEPDIR)/$*.Tpo -c -o $@ $< @am__fastdepCC_TRUE@ $(am__mv) $(DEPDIR)/$*.Tpo $(DEPDIR)/$*.Plo @AMDEP_TRUE@@am__fastdepCC_FALSE@ source='$<' object='$@' libtool=yes @AMDEPBACKSLASH@ @AMDEP_TRUE@@am__fastdepCC_FALSE@ DEPDIR=$(DEPDIR) $(CCDEPMODE) $(depcomp) @AMDEPBACKSLASH@ @am__fastdepCC_FALSE@ $(LTCOMPILE) -c -o $@ $< mostlyclean-libtool: -rm -f *.lo clean-libtool: -rm -rf .libs _libs # This directory's subdirectories are mostly independent; you can cd # into them and run `make' without going through this Makefile. # To change the values of `make' variables: instead of editing Makefiles, # (1) if the variable is set in `config.status', edit `config.status' # (which will cause the Makefiles to be regenerated when you run `make'); # (2) otherwise, pass the desired values on the `make' command line. $(RECURSIVE_TARGETS): @fail= failcom='exit 1'; \ for f in x $$MAKEFLAGS; do \ case $$f in \ *=* | --[!k]*);; \ *k*) failcom='fail=yes';; \ esac; \ done; \ dot_seen=no; \ target=`echo $@ | sed s/-recursive//`; \ list='$(SUBDIRS)'; for subdir in $$list; do \ echo "Making $$target in $$subdir"; \ if test "$$subdir" = "."; then \ dot_seen=yes; \ local_target="$$target-am"; \ else \ local_target="$$target"; \ fi; \ ($(am__cd) $$subdir && $(MAKE) $(AM_MAKEFLAGS) $$local_target) \ || eval $$failcom; \ done; \ if test "$$dot_seen" = "no"; then \ $(MAKE) $(AM_MAKEFLAGS) "$$target-am" || exit 1; \ fi; test -z "$$fail" $(RECURSIVE_CLEAN_TARGETS): @fail= failcom='exit 1'; \ for f in x $$MAKEFLAGS; do \ case $$f in \ *=* | --[!k]*);; \ *k*) failcom='fail=yes';; \ esac; \ done; \ dot_seen=no; \ case "$@" in \ distclean-* | maintainer-clean-*) list='$(DIST_SUBDIRS)' ;; \ *) list='$(SUBDIRS)' ;; \ esac; \ rev=''; for subdir in $$list; do \ if test "$$subdir" = "."; then :; else \ rev="$$subdir $$rev"; \ fi; \ done; \ rev="$$rev ."; \ target=`echo $@ | sed s/-recursive//`; \ for subdir in $$rev; do \ echo "Making $$target in $$subdir"; \ if test "$$subdir" = "."; then \ local_target="$$target-am"; \ else \ local_target="$$target"; \ fi; \ ($(am__cd) $$subdir && $(MAKE) $(AM_MAKEFLAGS) $$local_target) \ || eval $$failcom; \ done && test -z "$$fail" tags-recursive: list='$(SUBDIRS)'; for subdir in $$list; do \ test "$$subdir" = . || ($(am__cd) $$subdir && $(MAKE) $(AM_MAKEFLAGS) tags); \ done ctags-recursive: list='$(SUBDIRS)'; for subdir in $$list; do \ test "$$subdir" = . || ($(am__cd) $$subdir && $(MAKE) $(AM_MAKEFLAGS) ctags); \ done ID: $(HEADERS) $(SOURCES) $(LISP) $(TAGS_FILES) list='$(SOURCES) $(HEADERS) $(LISP) $(TAGS_FILES)'; \ unique=`for i in $$list; do \ if test -f "$$i"; then echo $$i; else echo $(srcdir)/$$i; fi; \ done | \ $(AWK) '{ files[$$0] = 1; nonempty = 1; } \ END { if (nonempty) { for (i in files) print i; }; }'`; \ mkid -fID $$unique tags: TAGS TAGS: tags-recursive $(HEADERS) $(SOURCES) $(TAGS_DEPENDENCIES) \ $(TAGS_FILES) $(LISP) set x; \ here=`pwd`; \ if ($(ETAGS) --etags-include --version) >/dev/null 2>&1; then \ include_option=--etags-include; \ empty_fix=.; \ else \ include_option=--include; \ empty_fix=; \ fi; \ list='$(SUBDIRS)'; for subdir in $$list; do \ if test "$$subdir" = .; then :; else \ test ! -f $$subdir/TAGS || \ set "$$@" "$$include_option=$$here/$$subdir/TAGS"; \ fi; \ done; \ list='$(SOURCES) $(HEADERS) $(LISP) $(TAGS_FILES)'; \ unique=`for i in $$list; do \ if test -f "$$i"; then echo $$i; else echo $(srcdir)/$$i; fi; \ done | \ $(AWK) '{ files[$$0] = 1; nonempty = 1; } \ END { if (nonempty) { for (i in files) print i; }; }'`; \ shift; \ if test -z "$(ETAGS_ARGS)$$*$$unique"; then :; else \ test -n "$$unique" || unique=$$empty_fix; \ if test $$# -gt 0; then \ $(ETAGS) $(ETAGSFLAGS) $(AM_ETAGSFLAGS) $(ETAGS_ARGS) \ "$$@" $$unique; \ else \ $(ETAGS) $(ETAGSFLAGS) $(AM_ETAGSFLAGS) $(ETAGS_ARGS) \ $$unique; \ fi; \ fi ctags: CTAGS CTAGS: ctags-recursive $(HEADERS) $(SOURCES) $(TAGS_DEPENDENCIES) \ $(TAGS_FILES) $(LISP) list='$(SOURCES) $(HEADERS) $(LISP) $(TAGS_FILES)'; \ unique=`for i in $$list; do \ if test -f "$$i"; then echo $$i; else echo $(srcdir)/$$i; fi; \ done | \ $(AWK) '{ files[$$0] = 1; nonempty = 1; } \ END { if (nonempty) { for (i in files) print i; }; }'`; \ test -z "$(CTAGS_ARGS)$$unique" \ || $(CTAGS) $(CTAGSFLAGS) $(AM_CTAGSFLAGS) $(CTAGS_ARGS) \ $$unique GTAGS: here=`$(am__cd) $(top_builddir) && pwd` \ && $(am__cd) $(top_srcdir) \ && gtags -i $(GTAGS_ARGS) "$$here" distclean-tags: -rm -f TAGS ID GTAGS GRTAGS GSYMS GPATH tags distdir: $(DISTFILES) @srcdirstrip=`echo "$(srcdir)" | sed 's/[].[^$$\\*]/\\\\&/g'`; \ topsrcdirstrip=`echo "$(top_srcdir)" | sed 's/[].[^$$\\*]/\\\\&/g'`; \ list='$(DISTFILES)'; \ dist_files=`for file in $$list; do echo $$file; done | \ sed -e "s|^$$srcdirstrip/||;t" \ -e "s|^$$topsrcdirstrip/|$(top_builddir)/|;t"`; \ case $$dist_files in \ */*) $(MKDIR_P) `echo "$$dist_files" | \ sed '/\//!d;s|^|$(distdir)/|;s,/[^/]*$$,,' | \ sort -u` ;; \ esac; \ for file in $$dist_files; do \ if test -f $$file || test -d $$file; then d=.; else d=$(srcdir); fi; \ if test -d $$d/$$file; then \ dir=`echo "/$$file" | sed -e 's,/[^/]*$$,,'`; \ if test -d "$(distdir)/$$file"; then \ find "$(distdir)/$$file" -type d ! -perm -700 -exec chmod u+rwx {} \;; \ fi; \ if test -d $(srcdir)/$$file && test $$d != $(srcdir); then \ cp -fpR $(srcdir)/$$file "$(distdir)$$dir" || exit 1; \ find "$(distdir)/$$file" -type d ! -perm -700 -exec chmod u+rwx {} \;; \ fi; \ cp -fpR $$d/$$file "$(distdir)$$dir" || exit 1; \ else \ test -f "$(distdir)/$$file" \ || cp -p $$d/$$file "$(distdir)/$$file" \ || exit 1; \ fi; \ done @list='$(DIST_SUBDIRS)'; for subdir in $$list; do \ if test "$$subdir" = .; then :; else \ $(am__make_dryrun) \ || test -d "$(distdir)/$$subdir" \ || $(MKDIR_P) "$(distdir)/$$subdir" \ || exit 1; \ dir1=$$subdir; dir2="$(distdir)/$$subdir"; \ $(am__relativize); \ new_distdir=$$reldir; \ dir1=$$subdir; dir2="$(top_distdir)"; \ $(am__relativize); \ new_top_distdir=$$reldir; \ echo " (cd $$subdir && $(MAKE) $(AM_MAKEFLAGS) top_distdir="$$new_top_distdir" distdir="$$new_distdir" \\"; \ echo " am__remove_distdir=: am__skip_length_check=: am__skip_mode_fix=: distdir)"; \ ($(am__cd) $$subdir && \ $(MAKE) $(AM_MAKEFLAGS) \ top_distdir="$$new_top_distdir" \ distdir="$$new_distdir" \ am__remove_distdir=: \ am__skip_length_check=: \ am__skip_mode_fix=: \ distdir) \ || exit 1; \ fi; \ done check-am: all-am check: check-recursive all-am: Makefile $(LTLIBRARIES) installdirs: installdirs-recursive installdirs-am: install: install-recursive install-exec: install-exec-recursive install-data: install-data-recursive uninstall: uninstall-recursive install-am: all-am @$(MAKE) $(AM_MAKEFLAGS) install-exec-am install-data-am installcheck: installcheck-recursive install-strip: if test -z '$(STRIP)'; then \ $(MAKE) $(AM_MAKEFLAGS) INSTALL_PROGRAM="$(INSTALL_STRIP_PROGRAM)" \ install_sh_PROGRAM="$(INSTALL_STRIP_PROGRAM)" INSTALL_STRIP_FLAG=-s \ install; \ else \ $(MAKE) $(AM_MAKEFLAGS) INSTALL_PROGRAM="$(INSTALL_STRIP_PROGRAM)" \ install_sh_PROGRAM="$(INSTALL_STRIP_PROGRAM)" INSTALL_STRIP_FLAG=-s \ "INSTALL_PROGRAM_ENV=STRIPPROG='$(STRIP)'" install; \ fi mostlyclean-generic: clean-generic: distclean-generic: -test -z "$(CONFIG_CLEAN_FILES)" || rm -f $(CONFIG_CLEAN_FILES) -test . = "$(srcdir)" || test -z "$(CONFIG_CLEAN_VPATH_FILES)" || rm -f $(CONFIG_CLEAN_VPATH_FILES) maintainer-clean-generic: @echo "This command is intended for maintainers to use" @echo "it deletes files that may require special tools to rebuild." clean: clean-recursive clean-am: clean-generic clean-libtool clean-noinstLTLIBRARIES \ mostlyclean-am distclean: distclean-recursive -rm -rf ./$(DEPDIR) -rm -f Makefile distclean-am: clean-am distclean-compile distclean-generic \ distclean-tags dvi: dvi-recursive dvi-am: html: html-recursive html-am: info: info-recursive info-am: install-data-am: install-dvi: install-dvi-recursive install-dvi-am: install-exec-am: install-html: install-html-recursive install-html-am: install-info: install-info-recursive install-info-am: install-man: install-pdf: install-pdf-recursive install-pdf-am: install-ps: install-ps-recursive install-ps-am: installcheck-am: maintainer-clean: maintainer-clean-recursive -rm -rf ./$(DEPDIR) -rm -f Makefile maintainer-clean-am: distclean-am maintainer-clean-generic mostlyclean: mostlyclean-recursive mostlyclean-am: mostlyclean-compile mostlyclean-generic \ mostlyclean-libtool pdf: pdf-recursive pdf-am: ps: ps-recursive ps-am: uninstall-am: .MAKE: $(RECURSIVE_CLEAN_TARGETS) $(RECURSIVE_TARGETS) ctags-recursive \ install-am install-strip tags-recursive .PHONY: $(RECURSIVE_CLEAN_TARGETS) $(RECURSIVE_TARGETS) CTAGS GTAGS \ all all-am check check-am clean clean-generic clean-libtool \ clean-noinstLTLIBRARIES ctags ctags-recursive distclean \ distclean-compile distclean-generic distclean-libtool \ distclean-tags distdir dvi dvi-am html html-am info info-am \ install install-am install-data install-data-am install-dvi \ install-dvi-am install-exec install-exec-am install-html \ install-html-am install-info install-info-am install-man \ install-pdf install-pdf-am install-ps install-ps-am \ install-strip installcheck installcheck-am installdirs \ installdirs-am maintainer-clean maintainer-clean-generic \ mostlyclean mostlyclean-compile mostlyclean-generic \ mostlyclean-libtool pdf pdf-am ps ps-am tags tags-recursive \ uninstall uninstall-am # redft00e-r2hc.c rodft00e-r2hc.c reodft11e-r2hc.c # Tell versions [3.59,3.63) of GNU make to not export all variables. # Otherwise a system limit (for SysV at least) may be exceeded. .NOEXPORT: fftw-3.3.3/reodft/conf.c0000644000175000017500000000300212054410050011745 00000000000000/* * Copyright (c) 2003, 2007-11 Matteo Frigo * Copyright (c) 2003, 2007-11 Massachusetts Institute of Technology * * This program is free software; you can redistribute it and/or modify * it under the terms of the GNU General Public License as published by * the Free Software Foundation; either version 2 of the License, or * (at your option) any later version. * * This program is distributed in the hope that it will be useful, * but WITHOUT ANY WARRANTY; without even the implied warranty of * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the * GNU General Public License for more details. * * You should have received a copy of the GNU General Public License * along with this program; if not, write to the Free Software * Foundation, Inc., 51 Franklin Street, Fifth Floor, Boston, MA 02110-1301 USA * */ #include "reodft.h" static const solvtab s = { #if 0 /* 1 to enable "standard" algorithms with substandard accuracy; you must also add them to Makefile.am to compile these files*/ SOLVTAB(X(redft00e_r2hc_register)), SOLVTAB(X(rodft00e_r2hc_register)), SOLVTAB(X(reodft11e_r2hc_register)), #endif SOLVTAB(X(redft00e_r2hc_pad_register)), SOLVTAB(X(rodft00e_r2hc_pad_register)), SOLVTAB(X(reodft00e_splitradix_register)), SOLVTAB(X(reodft010e_r2hc_register)), SOLVTAB(X(reodft11e_radix2_r2hc_register)), SOLVTAB(X(reodft11e_r2hc_odd_register)), SOLVTAB_END }; void X(reodft_conf_standard)(planner *p) { X(solvtab_exec)(s, p); } fftw-3.3.3/reodft/reodft.h0000644000175000017500000000274012054410050012320 00000000000000/* * Copyright (c) 2003, 2007-11 Matteo Frigo * Copyright (c) 2003, 2007-11 Massachusetts Institute of Technology * * This program is free software; you can redistribute it and/or modify * it under the terms of the GNU General Public License as published by * the Free Software Foundation; either version 2 of the License, or * (at your option) any later version. * * This program is distributed in the hope that it will be useful, * but WITHOUT ANY WARRANTY; without even the implied warranty of * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the * GNU General Public License for more details. * * You should have received a copy of the GNU General Public License * along with this program; if not, write to the Free Software * Foundation, Inc., 51 Franklin Street, Fifth Floor, Boston, MA 02110-1301 USA * */ #ifndef __REODFT_H__ #define __REODFT_H__ #include "ifftw.h" #include "rdft.h" #define REODFT_KINDP(k) ((k) >= REDFT00 && (k) <= RODFT11) void X(redft00e_r2hc_register)(planner *p); void X(redft00e_r2hc_pad_register)(planner *p); void X(rodft00e_r2hc_register)(planner *p); void X(rodft00e_r2hc_pad_register)(planner *p); void X(reodft00e_splitradix_register)(planner *p); void X(reodft010e_r2hc_register)(planner *p); void X(reodft11e_r2hc_register)(planner *p); void X(reodft11e_radix2_r2hc_register)(planner *p); void X(reodft11e_r2hc_odd_register)(planner *p); /* configurations */ void X(reodft_conf_standard)(planner *p); #endif /* __REODFT_H__ */ fftw-3.3.3/reodft/reodft010e-r2hc.c0000644000175000017500000002625012054410050013537 00000000000000/* * Copyright (c) 2003, 2007-11 Matteo Frigo * Copyright (c) 2003, 2007-11 Massachusetts Institute of Technology * * This program is free software; you can redistribute it and/or modify * it under the terms of the GNU General Public License as published by * the Free Software Foundation; either version 2 of the License, or * (at your option) any later version. * * This program is distributed in the hope that it will be useful, * but WITHOUT ANY WARRANTY; without even the implied warranty of * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the * GNU General Public License for more details. * * You should have received a copy of the GNU General Public License * along with this program; if not, write to the Free Software * Foundation, Inc., 51 Franklin Street, Fifth Floor, Boston, MA 02110-1301 USA * */ /* Do an R{E,O}DFT{01,10} problem via an R2HC problem, with some pre/post-processing ala FFTPACK. */ #include "reodft.h" typedef struct { solver super; } S; typedef struct { plan_rdft super; plan *cld; twid *td; INT is, os; INT n; INT vl; INT ivs, ovs; rdft_kind kind; } P; /* A real-even-01 DFT operates logically on a size-4N array: I 0 -r(I*) -I 0 r(I*), where r denotes reversal and * denotes deletion of the 0th element. To compute the transform of this, we imagine performing a radix-4 (real-input) DIF step, which turns the size-4N DFT into 4 size-N (contiguous) DFTs, two of which are zero and two of which are conjugates. The non-redundant size-N DFT has halfcomplex input, so we can do it with a size-N hc2r transform. (In order to share plans with the re10 (inverse) transform, however, we use the DHT trick to re-express the hc2r problem as r2hc. This has little cost since we are already pre- and post-processing the data in {i,n-i} order.) Finally, we have to write out the data in the correct order...the two size-N redundant (conjugate) hc2r DFTs correspond to the even and odd outputs in O (i.e. the usual interleaved output of DIF transforms); since this data has even symmetry, we only write the first half of it. The real-even-10 DFT is just the reverse of these steps, i.e. a radix-4 DIT transform. There, however, we just use the r2hc transform naturally without resorting to the DHT trick. A real-odd-01 DFT is very similar, except that the input is 0 I (rI)* 0 -I -(rI)*. This format, however, can be transformed into precisely the real-even-01 format above by sending I -> rI and shifting the array by N. The former swap is just another transformation on the input during preprocessing; the latter multiplies the even/odd outputs by i/-i, which combines with the factor of -i (to take the imaginary part) to simply flip the sign of the odd outputs. Vice-versa for real-odd-10. The FFTPACK source code was very helpful in working this out. (They do unnecessary passes over the array, though.) The same algorithm is also described in: John Makhoul, "A fast cosine transform in one and two dimensions," IEEE Trans. on Acoust. Speech and Sig. Proc., ASSP-28 (1), 27--34 (1980). Note that Numerical Recipes suggests a different algorithm that requires more operations and uses trig. functions for both the pre- and post-processing passes. */ static void apply_re01(const plan *ego_, R *I, R *O) { const P *ego = (const P *) ego_; INT is = ego->is, os = ego->os; INT i, n = ego->n; INT iv, vl = ego->vl; INT ivs = ego->ivs, ovs = ego->ovs; R *W = ego->td->W; R *buf; buf = (R *) MALLOC(sizeof(R) * n, BUFFERS); for (iv = 0; iv < vl; ++iv, I += ivs, O += ovs) { buf[0] = I[0]; for (i = 1; i < n - i; ++i) { E a, b, apb, amb, wa, wb; a = I[is * i]; b = I[is * (n - i)]; apb = a + b; amb = a - b; wa = W[2*i]; wb = W[2*i + 1]; buf[i] = wa * amb + wb * apb; buf[n - i] = wa * apb - wb * amb; } if (i == n - i) { buf[i] = K(2.0) * I[is * i] * W[2*i]; } { plan_rdft *cld = (plan_rdft *) ego->cld; cld->apply((plan *) cld, buf, buf); } O[0] = buf[0]; for (i = 1; i < n - i; ++i) { E a, b; INT k; a = buf[i]; b = buf[n - i]; k = i + i; O[os * (k - 1)] = a - b; O[os * k] = a + b; } if (i == n - i) { O[os * (n - 1)] = buf[i]; } } X(ifree)(buf); } /* ro01 is same as re01, but with i <-> n - 1 - i in the input and the sign of the odd output elements flipped. */ static void apply_ro01(const plan *ego_, R *I, R *O) { const P *ego = (const P *) ego_; INT is = ego->is, os = ego->os; INT i, n = ego->n; INT iv, vl = ego->vl; INT ivs = ego->ivs, ovs = ego->ovs; R *W = ego->td->W; R *buf; buf = (R *) MALLOC(sizeof(R) * n, BUFFERS); for (iv = 0; iv < vl; ++iv, I += ivs, O += ovs) { buf[0] = I[is * (n - 1)]; for (i = 1; i < n - i; ++i) { E a, b, apb, amb, wa, wb; a = I[is * (n - 1 - i)]; b = I[is * (i - 1)]; apb = a + b; amb = a - b; wa = W[2*i]; wb = W[2*i+1]; buf[i] = wa * amb + wb * apb; buf[n - i] = wa * apb - wb * amb; } if (i == n - i) { buf[i] = K(2.0) * I[is * (i - 1)] * W[2*i]; } { plan_rdft *cld = (plan_rdft *) ego->cld; cld->apply((plan *) cld, buf, buf); } O[0] = buf[0]; for (i = 1; i < n - i; ++i) { E a, b; INT k; a = buf[i]; b = buf[n - i]; k = i + i; O[os * (k - 1)] = b - a; O[os * k] = a + b; } if (i == n - i) { O[os * (n - 1)] = -buf[i]; } } X(ifree)(buf); } static void apply_re10(const plan *ego_, R *I, R *O) { const P *ego = (const P *) ego_; INT is = ego->is, os = ego->os; INT i, n = ego->n; INT iv, vl = ego->vl; INT ivs = ego->ivs, ovs = ego->ovs; R *W = ego->td->W; R *buf; buf = (R *) MALLOC(sizeof(R) * n, BUFFERS); for (iv = 0; iv < vl; ++iv, I += ivs, O += ovs) { buf[0] = I[0]; for (i = 1; i < n - i; ++i) { E u, v; INT k = i + i; u = I[is * (k - 1)]; v = I[is * k]; buf[n - i] = u; buf[i] = v; } if (i == n - i) { buf[i] = I[is * (n - 1)]; } { plan_rdft *cld = (plan_rdft *) ego->cld; cld->apply((plan *) cld, buf, buf); } O[0] = K(2.0) * buf[0]; for (i = 1; i < n - i; ++i) { E a, b, wa, wb; a = K(2.0) * buf[i]; b = K(2.0) * buf[n - i]; wa = W[2*i]; wb = W[2*i + 1]; O[os * i] = wa * a + wb * b; O[os * (n - i)] = wb * a - wa * b; } if (i == n - i) { O[os * i] = K(2.0) * buf[i] * W[2*i]; } } X(ifree)(buf); } /* ro10 is same as re10, but with i <-> n - 1 - i in the output and the sign of the odd input elements flipped. */ static void apply_ro10(const plan *ego_, R *I, R *O) { const P *ego = (const P *) ego_; INT is = ego->is, os = ego->os; INT i, n = ego->n; INT iv, vl = ego->vl; INT ivs = ego->ivs, ovs = ego->ovs; R *W = ego->td->W; R *buf; buf = (R *) MALLOC(sizeof(R) * n, BUFFERS); for (iv = 0; iv < vl; ++iv, I += ivs, O += ovs) { buf[0] = I[0]; for (i = 1; i < n - i; ++i) { E u, v; INT k = i + i; u = -I[is * (k - 1)]; v = I[is * k]; buf[n - i] = u; buf[i] = v; } if (i == n - i) { buf[i] = -I[is * (n - 1)]; } { plan_rdft *cld = (plan_rdft *) ego->cld; cld->apply((plan *) cld, buf, buf); } O[os * (n - 1)] = K(2.0) * buf[0]; for (i = 1; i < n - i; ++i) { E a, b, wa, wb; a = K(2.0) * buf[i]; b = K(2.0) * buf[n - i]; wa = W[2*i]; wb = W[2*i + 1]; O[os * (n - 1 - i)] = wa * a + wb * b; O[os * (i - 1)] = wb * a - wa * b; } if (i == n - i) { O[os * (i - 1)] = K(2.0) * buf[i] * W[2*i]; } } X(ifree)(buf); } static void awake(plan *ego_, enum wakefulness wakefulness) { P *ego = (P *) ego_; static const tw_instr reodft010e_tw[] = { { TW_COS, 0, 1 }, { TW_SIN, 0, 1 }, { TW_NEXT, 1, 0 } }; X(plan_awake)(ego->cld, wakefulness); X(twiddle_awake)(wakefulness, &ego->td, reodft010e_tw, 4*ego->n, 1, ego->n/2+1); } static void destroy(plan *ego_) { P *ego = (P *) ego_; X(plan_destroy_internal)(ego->cld); } static void print(const plan *ego_, printer *p) { const P *ego = (const P *) ego_; p->print(p, "(%se-r2hc-%D%v%(%p%))", X(rdft_kind_str)(ego->kind), ego->n, ego->vl, ego->cld); } static int applicable0(const solver *ego_, const problem *p_) { const problem_rdft *p = (const problem_rdft *) p_; UNUSED(ego_); return (1 && p->sz->rnk == 1 && p->vecsz->rnk <= 1 && (p->kind[0] == REDFT01 || p->kind[0] == REDFT10 || p->kind[0] == RODFT01 || p->kind[0] == RODFT10) ); } static int applicable(const solver *ego, const problem *p, const planner *plnr) { return (!NO_SLOWP(plnr) && applicable0(ego, p)); } static plan *mkplan(const solver *ego_, const problem *p_, planner *plnr) { P *pln; const problem_rdft *p; plan *cld; R *buf; INT n; opcnt ops; static const plan_adt padt = { X(rdft_solve), awake, print, destroy }; if (!applicable(ego_, p_, plnr)) return (plan *)0; p = (const problem_rdft *) p_; n = p->sz->dims[0].n; buf = (R *) MALLOC(sizeof(R) * n, BUFFERS); cld = X(mkplan_d)(plnr, X(mkproblem_rdft_1_d)(X(mktensor_1d)(n, 1, 1), X(mktensor_0d)(), buf, buf, R2HC)); X(ifree)(buf); if (!cld) return (plan *)0; switch (p->kind[0]) { case REDFT01: pln = MKPLAN_RDFT(P, &padt, apply_re01); break; case REDFT10: pln = MKPLAN_RDFT(P, &padt, apply_re10); break; case RODFT01: pln = MKPLAN_RDFT(P, &padt, apply_ro01); break; case RODFT10: pln = MKPLAN_RDFT(P, &padt, apply_ro10); break; default: A(0); return (plan*)0; } pln->n = n; pln->is = p->sz->dims[0].is; pln->os = p->sz->dims[0].os; pln->cld = cld; pln->td = 0; pln->kind = p->kind[0]; X(tensor_tornk1)(p->vecsz, &pln->vl, &pln->ivs, &pln->ovs); X(ops_zero)(&ops); ops.other = 4 + (n-1)/2 * 10 + (1 - n % 2) * 5; if (p->kind[0] == REDFT01 || p->kind[0] == RODFT01) { ops.add = (n-1)/2 * 6; ops.mul = (n-1)/2 * 4 + (1 - n % 2) * 2; } else { /* 10 transforms */ ops.add = (n-1)/2 * 2; ops.mul = 1 + (n-1)/2 * 6 + (1 - n % 2) * 2; } X(ops_zero)(&pln->super.super.ops); X(ops_madd2)(pln->vl, &ops, &pln->super.super.ops); X(ops_madd2)(pln->vl, &cld->ops, &pln->super.super.ops); return &(pln->super.super); } /* constructor */ static solver *mksolver(void) { static const solver_adt sadt = { PROBLEM_RDFT, mkplan, 0 }; S *slv = MKSOLVER(S, &sadt); return &(slv->super); } void X(reodft010e_r2hc_register)(planner *p) { REGISTER_SOLVER(p, mksolver()); } fftw-3.3.3/reodft/reodft11e-radix2.c0000644000175000017500000003133412054410050014012 00000000000000/* * Copyright (c) 2003, 2007-11 Matteo Frigo * Copyright (c) 2003, 2007-11 Massachusetts Institute of Technology * * This program is free software; you can redistribute it and/or modify * it under the terms of the GNU General Public License as published by * the Free Software Foundation; either version 2 of the License, or * (at your option) any later version. * * This program is distributed in the hope that it will be useful, * but WITHOUT ANY WARRANTY; without even the implied warranty of * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the * GNU General Public License for more details. * * You should have received a copy of the GNU General Public License * along with this program; if not, write to the Free Software * Foundation, Inc., 51 Franklin Street, Fifth Floor, Boston, MA 02110-1301 USA * */ /* Do an R{E,O}DFT11 problem of *even* size by a pair of R2HC problems of half the size, plus some pre/post-processing. Use a trick from: Zhongde Wang, "On computing the discrete Fourier and cosine transforms," IEEE Trans. Acoust. Speech Sig. Proc. ASSP-33 (4), 1341--1344 (1985). to re-express as a pair of half-size REDFT01 (DCT-III) problems. Our implementation looks quite a bit different from the algorithm described in the paper because we combined the paper's pre/post-processing with the pre/post-processing used to turn REDFT01 into R2HC. (Also, the paper uses a DCT/DST pair, but we turn the DST into a DCT via the usual reordering/sign-flip trick. We additionally combined a couple of the matrices/transformations of the paper into a single pass.) NOTE: We originally used a simpler method by S. C. Chan and K. L. Ho that turned out to have numerical problems; see reodft11e-r2hc.c. (For odd sizes, see reodft11e-r2hc-odd.c.) */ #include "reodft.h" typedef struct { solver super; } S; typedef struct { plan_rdft super; plan *cld; twid *td, *td2; INT is, os; INT n; INT vl; INT ivs, ovs; rdft_kind kind; } P; static void apply_re11(const plan *ego_, R *I, R *O) { const P *ego = (const P *) ego_; INT is = ego->is, os = ego->os; INT i, n = ego->n, n2 = n/2; INT iv, vl = ego->vl; INT ivs = ego->ivs, ovs = ego->ovs; R *W = ego->td->W; R *W2; R *buf; buf = (R *) MALLOC(sizeof(R) * n, BUFFERS); for (iv = 0; iv < vl; ++iv, I += ivs, O += ovs) { buf[0] = K(2.0) * I[0]; buf[n2] = K(2.0) * I[is * (n - 1)]; for (i = 1; i + i < n2; ++i) { INT k = i + i; E a, b, a2, b2; { E u, v; u = I[is * (k - 1)]; v = I[is * k]; a = u + v; b2 = u - v; } { E u, v; u = I[is * (n - k - 1)]; v = I[is * (n - k)]; b = u + v; a2 = u - v; } { E wa, wb; wa = W[2*i]; wb = W[2*i + 1]; { E apb, amb; apb = a + b; amb = a - b; buf[i] = wa * amb + wb * apb; buf[n2 - i] = wa * apb - wb * amb; } { E apb, amb; apb = a2 + b2; amb = a2 - b2; buf[n2 + i] = wa * amb + wb * apb; buf[n - i] = wa * apb - wb * amb; } } } if (i + i == n2) { E u, v; u = I[is * (n2 - 1)]; v = I[is * n2]; buf[i] = (u + v) * (W[2*i] * K(2.0)); buf[n - i] = (u - v) * (W[2*i] * K(2.0)); } /* child plan: two r2hc's of size n/2 */ { plan_rdft *cld = (plan_rdft *) ego->cld; cld->apply((plan *) cld, buf, buf); } W2 = ego->td2->W; { /* i == 0 case */ E wa, wb; E a, b; wa = W2[0]; /* cos */ wb = W2[1]; /* sin */ a = buf[0]; b = buf[n2]; O[0] = wa * a + wb * b; O[os * (n - 1)] = wb * a - wa * b; } W2 += 2; for (i = 1; i + i < n2; ++i, W2 += 2) { INT k; E u, v, u2, v2; u = buf[i]; v = buf[n2 - i]; u2 = buf[n2 + i]; v2 = buf[n - i]; k = (i + i) - 1; { E wa, wb; E a, b; wa = W2[0]; /* cos */ wb = W2[1]; /* sin */ a = u - v; b = v2 - u2; O[os * k] = wa * a + wb * b; O[os * (n - 1 - k)] = wb * a - wa * b; } ++k; W2 += 2; { E wa, wb; E a, b; wa = W2[0]; /* cos */ wb = W2[1]; /* sin */ a = u + v; b = u2 + v2; O[os * k] = wa * a + wb * b; O[os * (n - 1 - k)] = wb * a - wa * b; } } if (i + i == n2) { INT k = (i + i) - 1; E wa, wb; E a, b; wa = W2[0]; /* cos */ wb = W2[1]; /* sin */ a = buf[i]; b = buf[n2 + i]; O[os * k] = wa * a - wb * b; O[os * (n - 1 - k)] = wb * a + wa * b; } } X(ifree)(buf); } #if 0 /* This version of apply_re11 uses REDFT01 child plans, more similar to the original paper by Z. Wang. We keep it around for reference (it is simpler) and because it may become more efficient if we ever implement REDFT01 codelets. */ static void apply_re11(const plan *ego_, R *I, R *O) { const P *ego = (const P *) ego_; INT is = ego->is, os = ego->os; INT i, n = ego->n; INT iv, vl = ego->vl; INT ivs = ego->ivs, ovs = ego->ovs; R *W; R *buf; buf = (R *) MALLOC(sizeof(R) * n, BUFFERS); for (iv = 0; iv < vl; ++iv, I += ivs, O += ovs) { buf[0] = K(2.0) * I[0]; buf[n/2] = K(2.0) * I[is * (n - 1)]; for (i = 1; i + i < n; ++i) { INT k = i + i; E a, b; a = I[is * (k - 1)]; b = I[is * k]; buf[i] = a + b; buf[n - i] = a - b; } /* child plan: two redft01's (DCT-III) */ { plan_rdft *cld = (plan_rdft *) ego->cld; cld->apply((plan *) cld, buf, buf); } W = ego->td2->W; for (i = 0; i + 1 < n/2; ++i, W += 2) { { E wa, wb; E a, b; wa = W[0]; /* cos */ wb = W[1]; /* sin */ a = buf[i]; b = buf[n/2 + i]; O[os * i] = wa * a + wb * b; O[os * (n - 1 - i)] = wb * a - wa * b; } ++i; W += 2; { E wa, wb; E a, b; wa = W[0]; /* cos */ wb = W[1]; /* sin */ a = buf[i]; b = buf[n/2 + i]; O[os * i] = wa * a - wb * b; O[os * (n - 1 - i)] = wb * a + wa * b; } } if (i < n/2) { E wa, wb; E a, b; wa = W[0]; /* cos */ wb = W[1]; /* sin */ a = buf[i]; b = buf[n/2 + i]; O[os * i] = wa * a + wb * b; O[os * (n - 1 - i)] = wb * a - wa * b; } } X(ifree)(buf); } #endif /* 0 */ /* like for rodft01, rodft11 is obtained from redft11 by reversing the input and flipping the sign of every other output. */ static void apply_ro11(const plan *ego_, R *I, R *O) { const P *ego = (const P *) ego_; INT is = ego->is, os = ego->os; INT i, n = ego->n, n2 = n/2; INT iv, vl = ego->vl; INT ivs = ego->ivs, ovs = ego->ovs; R *W = ego->td->W; R *W2; R *buf; buf = (R *) MALLOC(sizeof(R) * n, BUFFERS); for (iv = 0; iv < vl; ++iv, I += ivs, O += ovs) { buf[0] = K(2.0) * I[is * (n - 1)]; buf[n2] = K(2.0) * I[0]; for (i = 1; i + i < n2; ++i) { INT k = i + i; E a, b, a2, b2; { E u, v; u = I[is * (n - k)]; v = I[is * (n - 1 - k)]; a = u + v; b2 = u - v; } { E u, v; u = I[is * (k)]; v = I[is * (k - 1)]; b = u + v; a2 = u - v; } { E wa, wb; wa = W[2*i]; wb = W[2*i + 1]; { E apb, amb; apb = a + b; amb = a - b; buf[i] = wa * amb + wb * apb; buf[n2 - i] = wa * apb - wb * amb; } { E apb, amb; apb = a2 + b2; amb = a2 - b2; buf[n2 + i] = wa * amb + wb * apb; buf[n - i] = wa * apb - wb * amb; } } } if (i + i == n2) { E u, v; u = I[is * n2]; v = I[is * (n2 - 1)]; buf[i] = (u + v) * (W[2*i] * K(2.0)); buf[n - i] = (u - v) * (W[2*i] * K(2.0)); } /* child plan: two r2hc's of size n/2 */ { plan_rdft *cld = (plan_rdft *) ego->cld; cld->apply((plan *) cld, buf, buf); } W2 = ego->td2->W; { /* i == 0 case */ E wa, wb; E a, b; wa = W2[0]; /* cos */ wb = W2[1]; /* sin */ a = buf[0]; b = buf[n2]; O[0] = wa * a + wb * b; O[os * (n - 1)] = wa * b - wb * a; } W2 += 2; for (i = 1; i + i < n2; ++i, W2 += 2) { INT k; E u, v, u2, v2; u = buf[i]; v = buf[n2 - i]; u2 = buf[n2 + i]; v2 = buf[n - i]; k = (i + i) - 1; { E wa, wb; E a, b; wa = W2[0]; /* cos */ wb = W2[1]; /* sin */ a = v - u; b = u2 - v2; O[os * k] = wa * a + wb * b; O[os * (n - 1 - k)] = wa * b - wb * a; } ++k; W2 += 2; { E wa, wb; E a, b; wa = W2[0]; /* cos */ wb = W2[1]; /* sin */ a = u + v; b = u2 + v2; O[os * k] = wa * a + wb * b; O[os * (n - 1 - k)] = wa * b - wb * a; } } if (i + i == n2) { INT k = (i + i) - 1; E wa, wb; E a, b; wa = W2[0]; /* cos */ wb = W2[1]; /* sin */ a = buf[i]; b = buf[n2 + i]; O[os * k] = wb * b - wa * a; O[os * (n - 1 - k)] = wa * b + wb * a; } } X(ifree)(buf); } static void awake(plan *ego_, enum wakefulness wakefulness) { P *ego = (P *) ego_; static const tw_instr reodft010e_tw[] = { { TW_COS, 0, 1 }, { TW_SIN, 0, 1 }, { TW_NEXT, 1, 0 } }; static const tw_instr reodft11e_tw[] = { { TW_COS, 1, 1 }, { TW_SIN, 1, 1 }, { TW_NEXT, 2, 0 } }; X(plan_awake)(ego->cld, wakefulness); X(twiddle_awake)(wakefulness, &ego->td, reodft010e_tw, 2*ego->n, 1, ego->n/4+1); X(twiddle_awake)(wakefulness, &ego->td2, reodft11e_tw, 8*ego->n, 1, ego->n); } static void destroy(plan *ego_) { P *ego = (P *) ego_; X(plan_destroy_internal)(ego->cld); } static void print(const plan *ego_, printer *p) { const P *ego = (const P *) ego_; p->print(p, "(%se-radix2-r2hc-%D%v%(%p%))", X(rdft_kind_str)(ego->kind), ego->n, ego->vl, ego->cld); } static int applicable0(const solver *ego_, const problem *p_) { const problem_rdft *p = (const problem_rdft *) p_; UNUSED(ego_); return (1 && p->sz->rnk == 1 && p->vecsz->rnk <= 1 && p->sz->dims[0].n % 2 == 0 && (p->kind[0] == REDFT11 || p->kind[0] == RODFT11) ); } static int applicable(const solver *ego, const problem *p, const planner *plnr) { return (!NO_SLOWP(plnr) && applicable0(ego, p)); } static plan *mkplan(const solver *ego_, const problem *p_, planner *plnr) { P *pln; const problem_rdft *p; plan *cld; R *buf; INT n; opcnt ops; static const plan_adt padt = { X(rdft_solve), awake, print, destroy }; if (!applicable(ego_, p_, plnr)) return (plan *)0; p = (const problem_rdft *) p_; n = p->sz->dims[0].n; buf = (R *) MALLOC(sizeof(R) * n, BUFFERS); cld = X(mkplan_d)(plnr, X(mkproblem_rdft_1_d)(X(mktensor_1d)(n/2, 1, 1), X(mktensor_1d)(2, n/2, n/2), buf, buf, R2HC)); X(ifree)(buf); if (!cld) return (plan *)0; pln = MKPLAN_RDFT(P, &padt, p->kind[0]==REDFT11 ? apply_re11:apply_ro11); pln->n = n; pln->is = p->sz->dims[0].is; pln->os = p->sz->dims[0].os; pln->cld = cld; pln->td = pln->td2 = 0; pln->kind = p->kind[0]; X(tensor_tornk1)(p->vecsz, &pln->vl, &pln->ivs, &pln->ovs); X(ops_zero)(&ops); ops.add = 2 + (n/2 - 1)/2 * 20; ops.mul = 6 + (n/2 - 1)/2 * 16; ops.other = 4*n + 2 + (n/2 - 1)/2 * 6; if ((n/2) % 2 == 0) { ops.add += 4; ops.mul += 8; ops.other += 4; } X(ops_zero)(&pln->super.super.ops); X(ops_madd2)(pln->vl, &ops, &pln->super.super.ops); X(ops_madd2)(pln->vl, &cld->ops, &pln->super.super.ops); return &(pln->super.super); } /* constructor */ static solver *mksolver(void) { static const solver_adt sadt = { PROBLEM_RDFT, mkplan, 0 }; S *slv = MKSOLVER(S, &sadt); return &(slv->super); } void X(reodft11e_radix2_r2hc_register)(planner *p) { REGISTER_SOLVER(p, mksolver()); } fftw-3.3.3/reodft/reodft11e-r2hc-odd.c0000644000175000017500000002111512054410050014217 00000000000000/* * Copyright (c) 2003, 2007-11 Matteo Frigo * Copyright (c) 2003, 2007-11 Massachusetts Institute of Technology * * This program is free software; you can redistribute it and/or modify * it under the terms of the GNU General Public License as published by * the Free Software Foundation; either version 2 of the License, or * (at your option) any later version. * * This program is distributed in the hope that it will be useful, * but WITHOUT ANY WARRANTY; without even the implied warranty of * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the * GNU General Public License for more details. * * You should have received a copy of the GNU General Public License * along with this program; if not, write to the Free Software * Foundation, Inc., 51 Franklin Street, Fifth Floor, Boston, MA 02110-1301 USA * */ /* Do an R{E,O}DFT11 problem via an R2HC problem of the same *odd* size, with some permutations and post-processing, as described in: S. C. Chan and K. L. Ho, "Fast algorithms for computing the discrete cosine transform," IEEE Trans. Circuits Systems II: Analog & Digital Sig. Proc. 39 (3), 185--190 (1992). (For even sizes, see reodft11e-radix2.c.) This algorithm is related to the 8 x n prime-factor-algorithm (PFA) decomposition of the size 8n "logical" DFT corresponding to the R{EO}DFT11. Aside from very confusing notation (several symbols are redefined from one line to the next), be aware that this paper has some errors. In particular, the signs are wrong in Eqs. (34-35). Also, Eqs. (36-37) should be simply C(k) = C(2k + 1 mod N), and similarly for S (or, equivalently, the second cases should have 2*N - 2*k - 1 instead of N - k - 1). Note also that in their definition of the DFT, similarly to FFTW's, the exponent's sign is -1, but they forgot to correspondingly multiply S (the sine terms) by -1. */ #include "reodft.h" typedef struct { solver super; } S; typedef struct { plan_rdft super; plan *cld; INT is, os; INT n; INT vl; INT ivs, ovs; rdft_kind kind; } P; static DK(SQRT2, +1.4142135623730950488016887242096980785696718753769); #define SGN_SET(x, i) ((i) % 2 ? -(x) : (x)) static void apply_re11(const plan *ego_, R *I, R *O) { const P *ego = (const P *) ego_; INT is = ego->is, os = ego->os; INT i, n = ego->n, n2 = n/2; INT iv, vl = ego->vl; INT ivs = ego->ivs, ovs = ego->ovs; R *buf; buf = (R *) MALLOC(sizeof(R) * n, BUFFERS); for (iv = 0; iv < vl; ++iv, I += ivs, O += ovs) { { INT m; for (i = 0, m = n2; m < n; ++i, m += 4) buf[i] = I[is * m]; for (; m < 2 * n; ++i, m += 4) buf[i] = -I[is * (2*n - m - 1)]; for (; m < 3 * n; ++i, m += 4) buf[i] = -I[is * (m - 2*n)]; for (; m < 4 * n; ++i, m += 4) buf[i] = I[is * (4*n - m - 1)]; m -= 4 * n; for (; i < n; ++i, m += 4) buf[i] = I[is * m]; } { /* child plan: R2HC of size n */ plan_rdft *cld = (plan_rdft *) ego->cld; cld->apply((plan *) cld, buf, buf); } /* FIXME: strength-reduce loop by 4 to eliminate ugly sgn_set? */ for (i = 0; i + i + 1 < n2; ++i) { INT k = i + i + 1; E c1, s1; E c2, s2; c1 = buf[k]; c2 = buf[k + 1]; s2 = buf[n - (k + 1)]; s1 = buf[n - k]; O[os * i] = SQRT2 * (SGN_SET(c1, (i+1)/2) + SGN_SET(s1, i/2)); O[os * (n - (i+1))] = SQRT2 * (SGN_SET(c1, (n-i)/2) - SGN_SET(s1, (n-(i+1))/2)); O[os * (n2 - (i+1))] = SQRT2 * (SGN_SET(c2, (n2-i)/2) - SGN_SET(s2, (n2-(i+1))/2)); O[os * (n2 + (i+1))] = SQRT2 * (SGN_SET(c2, (n2+i+2)/2) + SGN_SET(s2, (n2+(i+1))/2)); } if (i + i + 1 == n2) { E c, s; c = buf[n2]; s = buf[n - n2]; O[os * i] = SQRT2 * (SGN_SET(c, (i+1)/2) + SGN_SET(s, i/2)); O[os * (n - (i+1))] = SQRT2 * (SGN_SET(c, (i+2)/2) + SGN_SET(s, (i+1)/2)); } O[os * n2] = SQRT2 * SGN_SET(buf[0], (n2+1)/2); } X(ifree)(buf); } /* like for rodft01, rodft11 is obtained from redft11 by reversing the input and flipping the sign of every other output. */ static void apply_ro11(const plan *ego_, R *I, R *O) { const P *ego = (const P *) ego_; INT is = ego->is, os = ego->os; INT i, n = ego->n, n2 = n/2; INT iv, vl = ego->vl; INT ivs = ego->ivs, ovs = ego->ovs; R *buf; buf = (R *) MALLOC(sizeof(R) * n, BUFFERS); for (iv = 0; iv < vl; ++iv, I += ivs, O += ovs) { { INT m; for (i = 0, m = n2; m < n; ++i, m += 4) buf[i] = I[is * (n - 1 - m)]; for (; m < 2 * n; ++i, m += 4) buf[i] = -I[is * (m - n)]; for (; m < 3 * n; ++i, m += 4) buf[i] = -I[is * (3*n - 1 - m)]; for (; m < 4 * n; ++i, m += 4) buf[i] = I[is * (m - 3*n)]; m -= 4 * n; for (; i < n; ++i, m += 4) buf[i] = I[is * (n - 1 - m)]; } { /* child plan: R2HC of size n */ plan_rdft *cld = (plan_rdft *) ego->cld; cld->apply((plan *) cld, buf, buf); } /* FIXME: strength-reduce loop by 4 to eliminate ugly sgn_set? */ for (i = 0; i + i + 1 < n2; ++i) { INT k = i + i + 1; INT j; E c1, s1; E c2, s2; c1 = buf[k]; c2 = buf[k + 1]; s2 = buf[n - (k + 1)]; s1 = buf[n - k]; O[os * i] = SQRT2 * (SGN_SET(c1, (i+1)/2 + i) + SGN_SET(s1, i/2 + i)); O[os * (n - (i+1))] = SQRT2 * (SGN_SET(c1, (n-i)/2 + i) - SGN_SET(s1, (n-(i+1))/2 + i)); j = n2 - (i+1); O[os * j] = SQRT2 * (SGN_SET(c2, (n2-i)/2 + j) - SGN_SET(s2, (n2-(i+1))/2 + j)); O[os * (n2 + (i+1))] = SQRT2 * (SGN_SET(c2, (n2+i+2)/2 + j) + SGN_SET(s2, (n2+(i+1))/2 + j)); } if (i + i + 1 == n2) { E c, s; c = buf[n2]; s = buf[n - n2]; O[os * i] = SQRT2 * (SGN_SET(c, (i+1)/2 + i) + SGN_SET(s, i/2 + i)); O[os * (n - (i+1))] = SQRT2 * (SGN_SET(c, (i+2)/2 + i) + SGN_SET(s, (i+1)/2 + i)); } O[os * n2] = SQRT2 * SGN_SET(buf[0], (n2+1)/2 + n2); } X(ifree)(buf); } static void awake(plan *ego_, enum wakefulness wakefulness) { P *ego = (P *) ego_; X(plan_awake)(ego->cld, wakefulness); } static void destroy(plan *ego_) { P *ego = (P *) ego_; X(plan_destroy_internal)(ego->cld); } static void print(const plan *ego_, printer *p) { const P *ego = (const P *) ego_; p->print(p, "(%se-r2hc-odd-%D%v%(%p%))", X(rdft_kind_str)(ego->kind), ego->n, ego->vl, ego->cld); } static int applicable0(const solver *ego_, const problem *p_) { const problem_rdft *p = (const problem_rdft *) p_; UNUSED(ego_); return (1 && p->sz->rnk == 1 && p->vecsz->rnk <= 1 && p->sz->dims[0].n % 2 == 1 && (p->kind[0] == REDFT11 || p->kind[0] == RODFT11) ); } static int applicable(const solver *ego, const problem *p, const planner *plnr) { return (!NO_SLOWP(plnr) && applicable0(ego, p)); } static plan *mkplan(const solver *ego_, const problem *p_, planner *plnr) { P *pln; const problem_rdft *p; plan *cld; R *buf; INT n; opcnt ops; static const plan_adt padt = { X(rdft_solve), awake, print, destroy }; if (!applicable(ego_, p_, plnr)) return (plan *)0; p = (const problem_rdft *) p_; n = p->sz->dims[0].n; buf = (R *) MALLOC(sizeof(R) * n, BUFFERS); cld = X(mkplan_d)(plnr, X(mkproblem_rdft_1_d)(X(mktensor_1d)(n, 1, 1), X(mktensor_0d)(), buf, buf, R2HC)); X(ifree)(buf); if (!cld) return (plan *)0; pln = MKPLAN_RDFT(P, &padt, p->kind[0]==REDFT11 ? apply_re11:apply_ro11); pln->n = n; pln->is = p->sz->dims[0].is; pln->os = p->sz->dims[0].os; pln->cld = cld; pln->kind = p->kind[0]; X(tensor_tornk1)(p->vecsz, &pln->vl, &pln->ivs, &pln->ovs); X(ops_zero)(&ops); ops.add = n - 1; ops.mul = n; ops.other = 4*n; X(ops_zero)(&pln->super.super.ops); X(ops_madd2)(pln->vl, &ops, &pln->super.super.ops); X(ops_madd2)(pln->vl, &cld->ops, &pln->super.super.ops); return &(pln->super.super); } /* constructor */ static solver *mksolver(void) { static const solver_adt sadt = { PROBLEM_RDFT, mkplan, 0 }; S *slv = MKSOLVER(S, &sadt); return &(slv->super); } void X(reodft11e_r2hc_odd_register)(planner *p) { REGISTER_SOLVER(p, mksolver()); } fftw-3.3.3/reodft/redft00e-r2hc-pad.c0000644000175000017500000001171212054410050014036 00000000000000/* * Copyright (c) 2003, 2007-11 Matteo Frigo * Copyright (c) 2003, 2007-11 Massachusetts Institute of Technology * * This program is free software; you can redistribute it and/or modify * it under the terms of the GNU General Public License as published by * the Free Software Foundation; either version 2 of the License, or * (at your option) any later version. * * This program is distributed in the hope that it will be useful, * but WITHOUT ANY WARRANTY; without even the implied warranty of * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the * GNU General Public License for more details. * * You should have received a copy of the GNU General Public License * along with this program; if not, write to the Free Software * Foundation, Inc., 51 Franklin Street, Fifth Floor, Boston, MA 02110-1301 USA * */ /* Do a REDFT00 problem via an R2HC problem, padded symmetrically to twice the size. This is asymptotically a factor of ~2 worse than redft00e-r2hc.c (the algorithm used in e.g. FFTPACK and Numerical Recipes), but we abandoned the latter after we discovered that it has intrinsic accuracy problems. */ #include "reodft.h" typedef struct { solver super; } S; typedef struct { plan_rdft super; plan *cld, *cldcpy; INT is; INT n; INT vl; INT ivs, ovs; } P; static void apply(const plan *ego_, R *I, R *O) { const P *ego = (const P *) ego_; INT is = ego->is; INT i, n = ego->n; INT iv, vl = ego->vl; INT ivs = ego->ivs, ovs = ego->ovs; R *buf; buf = (R *) MALLOC(sizeof(R) * (2*n), BUFFERS); for (iv = 0; iv < vl; ++iv, I += ivs, O += ovs) { buf[0] = I[0]; for (i = 1; i < n; ++i) { R a = I[i * is]; buf[i] = a; buf[2*n - i] = a; } buf[i] = I[i * is]; /* i == n, Nyquist */ /* r2hc transform of size 2*n */ { plan_rdft *cld = (plan_rdft *) ego->cld; cld->apply((plan *) cld, buf, buf); } /* copy n+1 real numbers (real parts of hc array) from buf to O */ { plan_rdft *cldcpy = (plan_rdft *) ego->cldcpy; cldcpy->apply((plan *) cldcpy, buf, O); } } X(ifree)(buf); } static void awake(plan *ego_, enum wakefulness wakefulness) { P *ego = (P *) ego_; X(plan_awake)(ego->cld, wakefulness); X(plan_awake)(ego->cldcpy, wakefulness); } static void destroy(plan *ego_) { P *ego = (P *) ego_; X(plan_destroy_internal)(ego->cldcpy); X(plan_destroy_internal)(ego->cld); } static void print(const plan *ego_, printer *p) { const P *ego = (const P *) ego_; p->print(p, "(redft00e-r2hc-pad-%D%v%(%p%)%(%p%))", ego->n + 1, ego->vl, ego->cld, ego->cldcpy); } static int applicable0(const solver *ego_, const problem *p_) { const problem_rdft *p = (const problem_rdft *) p_; UNUSED(ego_); return (1 && p->sz->rnk == 1 && p->vecsz->rnk <= 1 && p->kind[0] == REDFT00 && p->sz->dims[0].n > 1 /* n == 1 is not well-defined */ ); } static int applicable(const solver *ego, const problem *p, const planner *plnr) { return (!NO_SLOWP(plnr) && applicable0(ego, p)); } static plan *mkplan(const solver *ego_, const problem *p_, planner *plnr) { P *pln; const problem_rdft *p; plan *cld = (plan *) 0, *cldcpy; R *buf = (R *) 0; INT n; INT vl, ivs, ovs; opcnt ops; static const plan_adt padt = { X(rdft_solve), awake, print, destroy }; if (!applicable(ego_, p_, plnr)) goto nada; p = (const problem_rdft *) p_; n = p->sz->dims[0].n - 1; A(n > 0); buf = (R *) MALLOC(sizeof(R) * (2*n), BUFFERS); cld = X(mkplan_d)(plnr,X(mkproblem_rdft_1_d)(X(mktensor_1d)(2*n,1,1), X(mktensor_0d)(), buf, buf, R2HC)); if (!cld) goto nada; X(tensor_tornk1)(p->vecsz, &vl, &ivs, &ovs); cldcpy = X(mkplan_d)(plnr, X(mkproblem_rdft_1_d)(X(mktensor_0d)(), X(mktensor_1d)(n+1,1, p->sz->dims[0].os), buf, TAINT(p->O, ovs), R2HC)); if (!cldcpy) goto nada; X(ifree)(buf); pln = MKPLAN_RDFT(P, &padt, apply); pln->n = n; pln->is = p->sz->dims[0].is; pln->cld = cld; pln->cldcpy = cldcpy; pln->vl = vl; pln->ivs = ivs; pln->ovs = ovs; X(ops_zero)(&ops); ops.other = n + 2*n; /* loads + stores (input -> buf) */ X(ops_zero)(&pln->super.super.ops); X(ops_madd2)(pln->vl, &ops, &pln->super.super.ops); X(ops_madd2)(pln->vl, &cld->ops, &pln->super.super.ops); X(ops_madd2)(pln->vl, &cldcpy->ops, &pln->super.super.ops); return &(pln->super.super); nada: X(ifree0)(buf); if (cld) X(plan_destroy_internal)(cld); return (plan *)0; } /* constructor */ static solver *mksolver(void) { static const solver_adt sadt = { PROBLEM_RDFT, mkplan, 0 }; S *slv = MKSOLVER(S, &sadt); return &(slv->super); } void X(redft00e_r2hc_pad_register)(planner *p) { REGISTER_SOLVER(p, mksolver()); } fftw-3.3.3/reodft/rodft00e-r2hc-pad.c0000644000175000017500000001164012054410050014050 00000000000000/* * Copyright (c) 2003, 2007-11 Matteo Frigo * Copyright (c) 2003, 2007-11 Massachusetts Institute of Technology * * This program is free software; you can redistribute it and/or modify * it under the terms of the GNU General Public License as published by * the Free Software Foundation; either version 2 of the License, or * (at your option) any later version. * * This program is distributed in the hope that it will be useful, * but WITHOUT ANY WARRANTY; without even the implied warranty of * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the * GNU General Public License for more details. * * You should have received a copy of the GNU General Public License * along with this program; if not, write to the Free Software * Foundation, Inc., 51 Franklin Street, Fifth Floor, Boston, MA 02110-1301 USA * */ /* Do a RODFT00 problem via an R2HC problem, padded antisymmetrically to twice the size. This is asymptotically a factor of ~2 worse than rodft00e-r2hc.c (the algorithm used in e.g. FFTPACK and Numerical Recipes), but we abandoned the latter after we discovered that it has intrinsic accuracy problems. */ #include "reodft.h" typedef struct { solver super; } S; typedef struct { plan_rdft super; plan *cld, *cldcpy; INT is; INT n; INT vl; INT ivs, ovs; } P; static void apply(const plan *ego_, R *I, R *O) { const P *ego = (const P *) ego_; INT is = ego->is; INT i, n = ego->n; INT iv, vl = ego->vl; INT ivs = ego->ivs, ovs = ego->ovs; R *buf; buf = (R *) MALLOC(sizeof(R) * (2*n), BUFFERS); for (iv = 0; iv < vl; ++iv, I += ivs, O += ovs) { buf[0] = K(0.0); for (i = 1; i < n; ++i) { R a = I[(i-1) * is]; buf[i] = -a; buf[2*n - i] = a; } buf[i] = K(0.0); /* i == n, Nyquist */ /* r2hc transform of size 2*n */ { plan_rdft *cld = (plan_rdft *) ego->cld; cld->apply((plan *) cld, buf, buf); } /* copy n-1 real numbers (imag. parts of hc array) from buf to O */ { plan_rdft *cldcpy = (plan_rdft *) ego->cldcpy; cldcpy->apply((plan *) cldcpy, buf+2*n-1, O); } } X(ifree)(buf); } static void awake(plan *ego_, enum wakefulness wakefulness) { P *ego = (P *) ego_; X(plan_awake)(ego->cld, wakefulness); X(plan_awake)(ego->cldcpy, wakefulness); } static void destroy(plan *ego_) { P *ego = (P *) ego_; X(plan_destroy_internal)(ego->cldcpy); X(plan_destroy_internal)(ego->cld); } static void print(const plan *ego_, printer *p) { const P *ego = (const P *) ego_; p->print(p, "(rodft00e-r2hc-pad-%D%v%(%p%)%(%p%))", ego->n - 1, ego->vl, ego->cld, ego->cldcpy); } static int applicable0(const solver *ego_, const problem *p_) { const problem_rdft *p = (const problem_rdft *) p_; UNUSED(ego_); return (1 && p->sz->rnk == 1 && p->vecsz->rnk <= 1 && p->kind[0] == RODFT00 ); } static int applicable(const solver *ego, const problem *p, const planner *plnr) { return (!NO_SLOWP(plnr) && applicable0(ego, p)); } static plan *mkplan(const solver *ego_, const problem *p_, planner *plnr) { P *pln; const problem_rdft *p; plan *cld = (plan *) 0, *cldcpy; R *buf = (R *) 0; INT n; INT vl, ivs, ovs; opcnt ops; static const plan_adt padt = { X(rdft_solve), awake, print, destroy }; if (!applicable(ego_, p_, plnr)) goto nada; p = (const problem_rdft *) p_; n = p->sz->dims[0].n + 1; A(n > 0); buf = (R *) MALLOC(sizeof(R) * (2*n), BUFFERS); cld = X(mkplan_d)(plnr,X(mkproblem_rdft_1_d)(X(mktensor_1d)(2*n,1,1), X(mktensor_0d)(), buf, buf, R2HC)); if (!cld) goto nada; X(tensor_tornk1)(p->vecsz, &vl, &ivs, &ovs); cldcpy = X(mkplan_d)(plnr, X(mkproblem_rdft_1_d)(X(mktensor_0d)(), X(mktensor_1d)(n-1,-1, p->sz->dims[0].os), buf+2*n-1,TAINT(p->O, ovs), R2HC)); if (!cldcpy) goto nada; X(ifree)(buf); pln = MKPLAN_RDFT(P, &padt, apply); pln->n = n; pln->is = p->sz->dims[0].is; pln->cld = cld; pln->cldcpy = cldcpy; pln->vl = vl; pln->ivs = ivs; pln->ovs = ovs; X(ops_zero)(&ops); ops.other = n-1 + 2*n; /* loads + stores (input -> buf) */ X(ops_zero)(&pln->super.super.ops); X(ops_madd2)(pln->vl, &ops, &pln->super.super.ops); X(ops_madd2)(pln->vl, &cld->ops, &pln->super.super.ops); X(ops_madd2)(pln->vl, &cldcpy->ops, &pln->super.super.ops); return &(pln->super.super); nada: X(ifree0)(buf); if (cld) X(plan_destroy_internal)(cld); return (plan *)0; } /* constructor */ static solver *mksolver(void) { static const solver_adt sadt = { PROBLEM_RDFT, mkplan, 0 }; S *slv = MKSOLVER(S, &sadt); return &(slv->super); } void X(rodft00e_r2hc_pad_register)(planner *p) { REGISTER_SOLVER(p, mksolver()); } fftw-3.3.3/reodft/reodft00e-splitradix.c0000644000175000017500000002366512054410050015012 00000000000000/* * Copyright (c) 2005 Matteo Frigo * Copyright (c) 2005 Massachusetts Institute of Technology * * This program is free software; you can redistribute it and/or modify * it under the terms of the GNU General Public License as published by * the Free Software Foundation; either version 2 of the License, or * (at your option) any later version. * * This program is distributed in the hope that it will be useful, * but WITHOUT ANY WARRANTY; without even the implied warranty of * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the * GNU General Public License for more details. * * You should have received a copy of the GNU General Public License * along with this program; if not, write to the Free Software * Foundation, Inc., 51 Franklin Street, Fifth Floor, Boston, MA 02110-1301 USA * */ /* Do an R{E,O}DFT00 problem (of an odd length n) recursively via an R{E,O}DFT00 problem and an RDFT problem of half the length. This works by "logically" expanding the array to a real-even/odd DFT of length 2n-/+2 and then applying the split-radix algorithm. In this way, we can avoid having to pad to twice the length (ala redft00-r2hc-pad), saving a factor of ~2 for n=2^m+/-1, but don't incur the accuracy loss that the "ordinary" algorithm sacrifices (ala redft00-r2hc.c). */ #include "reodft.h" typedef struct { solver super; } S; typedef struct { plan_rdft super; plan *clde, *cldo; twid *td; INT is, os; INT n; INT vl; INT ivs, ovs; } P; /* redft00 */ static void apply_e(const plan *ego_, R *I, R *O) { const P *ego = (const P *) ego_; INT is = ego->is, os = ego->os; INT i, j, n = ego->n + 1, n2 = (n-1)/2; INT iv, vl = ego->vl; INT ivs = ego->ivs, ovs = ego->ovs; R *W = ego->td->W - 2; R *buf; buf = (R *) MALLOC(sizeof(R) * n2, BUFFERS); for (iv = 0; iv < vl; ++iv, I += ivs, O += ovs) { /* do size (n-1)/2 r2hc transform of odd-indexed elements with stride 4, "wrapping around" end of array with even boundary conditions */ for (j = 0, i = 1; i < n; i += 4) buf[j++] = I[is * i]; for (i = 2*n-2-i; i > 0; i -= 4) buf[j++] = I[is * i]; { plan_rdft *cld = (plan_rdft *) ego->cldo; cld->apply((plan *) cld, buf, buf); } /* do size (n+1)/2 redft00 of the even-indexed elements, writing to O: */ { plan_rdft *cld = (plan_rdft *) ego->clde; cld->apply((plan *) cld, I, O); } /* combine the results with the twiddle factors to get output */ { /* DC element */ E b20 = O[0], b0 = K(2.0) * buf[0]; O[0] = b20 + b0; O[2*(n2*os)] = b20 - b0; /* O[n2*os] = O[n2*os]; */ } for (i = 1; i < n2 - i; ++i) { E ap, am, br, bi, wr, wi, wbr, wbi; br = buf[i]; bi = buf[n2 - i]; wr = W[2*i]; wi = W[2*i+1]; #if FFT_SIGN == -1 wbr = K(2.0) * (wr*br + wi*bi); wbi = K(2.0) * (wr*bi - wi*br); #else wbr = K(2.0) * (wr*br - wi*bi); wbi = K(2.0) * (wr*bi + wi*br); #endif ap = O[i*os]; O[i*os] = ap + wbr; O[(2*n2 - i)*os] = ap - wbr; am = O[(n2 - i)*os]; #if FFT_SIGN == -1 O[(n2 - i)*os] = am - wbi; O[(n2 + i)*os] = am + wbi; #else O[(n2 - i)*os] = am + wbi; O[(n2 + i)*os] = am - wbi; #endif } if (i == n2 - i) { /* Nyquist element */ E ap, wbr; wbr = K(2.0) * (W[2*i] * buf[i]); ap = O[i*os]; O[i*os] = ap + wbr; O[(2*n2 - i)*os] = ap - wbr; } } X(ifree)(buf); } /* rodft00 */ static void apply_o(const plan *ego_, R *I, R *O) { const P *ego = (const P *) ego_; INT is = ego->is, os = ego->os; INT i, j, n = ego->n - 1, n2 = (n+1)/2; INT iv, vl = ego->vl; INT ivs = ego->ivs, ovs = ego->ovs; R *W = ego->td->W - 2; R *buf; buf = (R *) MALLOC(sizeof(R) * n2, BUFFERS); for (iv = 0; iv < vl; ++iv, I += ivs, O += ovs) { /* do size (n+1)/2 r2hc transform of even-indexed elements with stride 4, "wrapping around" end of array with odd boundary conditions */ for (j = 0, i = 0; i < n; i += 4) buf[j++] = I[is * i]; for (i = 2*n-i; i > 0; i -= 4) buf[j++] = -I[is * i]; { plan_rdft *cld = (plan_rdft *) ego->cldo; cld->apply((plan *) cld, buf, buf); } /* do size (n-1)/2 rodft00 of the odd-indexed elements, writing to O: */ { plan_rdft *cld = (plan_rdft *) ego->clde; if (I == O) { /* can't use I+is and I, subplan would lose in-placeness */ cld->apply((plan *) cld, I + is, I + is); /* we could maybe avoid this copy by modifying the twiddle loop, but currently I can't be bothered. */ A(is >= os); for (i = 0; i < n2-1; ++i) O[os*i] = I[is*(i+1)]; } else cld->apply((plan *) cld, I + is, O); } /* combine the results with the twiddle factors to get output */ O[(n2-1)*os] = K(2.0) * buf[0]; for (i = 1; i < n2 - i; ++i) { E ap, am, br, bi, wr, wi, wbr, wbi; br = buf[i]; bi = buf[n2 - i]; wr = W[2*i]; wi = W[2*i+1]; #if FFT_SIGN == -1 wbr = K(2.0) * (wr*br + wi*bi); wbi = K(2.0) * (wi*br - wr*bi); #else wbr = K(2.0) * (wr*br - wi*bi); wbi = K(2.0) * (wr*bi + wi*br); #endif ap = O[(i-1)*os]; O[(i-1)*os] = wbi + ap; O[(2*n2-1 - i)*os] = wbi - ap; am = O[(n2-1 - i)*os]; #if FFT_SIGN == -1 O[(n2-1 - i)*os] = wbr + am; O[(n2-1 + i)*os] = wbr - am; #else O[(n2-1 - i)*os] = wbr + am; O[(n2-1 + i)*os] = wbr - am; #endif } if (i == n2 - i) { /* Nyquist element */ E ap, wbi; wbi = K(2.0) * (W[2*i+1] * buf[i]); ap = O[(i-1)*os]; O[(i-1)*os] = wbi + ap; O[(2*n2-1 - i)*os] = wbi - ap; } } X(ifree)(buf); } static void awake(plan *ego_, enum wakefulness wakefulness) { P *ego = (P *) ego_; static const tw_instr reodft00e_tw[] = { { TW_COS, 1, 1 }, { TW_SIN, 1, 1 }, { TW_NEXT, 1, 0 } }; X(plan_awake)(ego->clde, wakefulness); X(plan_awake)(ego->cldo, wakefulness); X(twiddle_awake)(wakefulness, &ego->td, reodft00e_tw, 2*ego->n, 1, ego->n/4); } static void destroy(plan *ego_) { P *ego = (P *) ego_; X(plan_destroy_internal)(ego->cldo); X(plan_destroy_internal)(ego->clde); } static void print(const plan *ego_, printer *p) { const P *ego = (const P *) ego_; if (ego->super.apply == apply_e) p->print(p, "(redft00e-splitradix-%D%v%(%p%)%(%p%))", ego->n + 1, ego->vl, ego->clde, ego->cldo); else p->print(p, "(rodft00e-splitradix-%D%v%(%p%)%(%p%))", ego->n - 1, ego->vl, ego->clde, ego->cldo); } static int applicable0(const solver *ego_, const problem *p_) { const problem_rdft *p = (const problem_rdft *) p_; UNUSED(ego_); return (1 && p->sz->rnk == 1 && p->vecsz->rnk <= 1 && (p->kind[0] == REDFT00 || p->kind[0] == RODFT00) && p->sz->dims[0].n > 1 /* don't create size-0 sub-plans */ && p->sz->dims[0].n % 2 /* odd: 4 divides "logical" DFT */ && (p->I != p->O || p->vecsz->rnk == 0 || p->vecsz->dims[0].is == p->vecsz->dims[0].os) && (p->kind[0] != RODFT00 || p->I != p->O || p->sz->dims[0].is >= p->sz->dims[0].os) /* laziness */ ); } static int applicable(const solver *ego, const problem *p, const planner *plnr) { return (!NO_SLOWP(plnr) && applicable0(ego, p)); } static plan *mkplan(const solver *ego_, const problem *p_, planner *plnr) { P *pln; const problem_rdft *p; plan *clde, *cldo; R *buf; INT n, n0; opcnt ops; int inplace_odd; static const plan_adt padt = { X(rdft_solve), awake, print, destroy }; if (!applicable(ego_, p_, plnr)) return (plan *)0; p = (const problem_rdft *) p_; n = (n0 = p->sz->dims[0].n) + (p->kind[0] == REDFT00 ? (INT)-1 : (INT)1); A(n > 0 && n % 2 == 0); buf = (R *) MALLOC(sizeof(R) * (n/2), BUFFERS); inplace_odd = p->kind[0]==RODFT00 && p->I == p->O; clde = X(mkplan_d)(plnr, X(mkproblem_rdft_1_d)( X(mktensor_1d)(n0-n/2, 2*p->sz->dims[0].is, inplace_odd ? p->sz->dims[0].is : p->sz->dims[0].os), X(mktensor_0d)(), TAINT(p->I + p->sz->dims[0].is * (p->kind[0]==RODFT00), p->vecsz->rnk ? p->vecsz->dims[0].is : 0), TAINT(p->O + p->sz->dims[0].is * inplace_odd, p->vecsz->rnk ? p->vecsz->dims[0].os : 0), p->kind[0])); if (!clde) { X(ifree)(buf); return (plan *)0; } cldo = X(mkplan_d)(plnr, X(mkproblem_rdft_1_d)( X(mktensor_1d)(n/2, 1, 1), X(mktensor_0d)(), buf, buf, R2HC)); X(ifree)(buf); if (!cldo) return (plan *)0; pln = MKPLAN_RDFT(P, &padt, p->kind[0] == REDFT00 ? apply_e : apply_o); pln->n = n; pln->is = p->sz->dims[0].is; pln->os = p->sz->dims[0].os; pln->clde = clde; pln->cldo = cldo; pln->td = 0; X(tensor_tornk1)(p->vecsz, &pln->vl, &pln->ivs, &pln->ovs); X(ops_zero)(&ops); ops.other = n/2; ops.add = (p->kind[0]==REDFT00 ? (INT)2 : (INT)0) + (n/2-1)/2 * 6 + ((n/2)%2==0) * 2; ops.mul = 1 + (n/2-1)/2 * 6 + ((n/2)%2==0) * 2; /* tweak ops.other so that r2hc-pad is used for small sizes, which seems to be a lot faster on my machine: */ ops.other += 256; X(ops_zero)(&pln->super.super.ops); X(ops_madd2)(pln->vl, &ops, &pln->super.super.ops); X(ops_madd2)(pln->vl, &clde->ops, &pln->super.super.ops); X(ops_madd2)(pln->vl, &cldo->ops, &pln->super.super.ops); return &(pln->super.super); } /* constructor */ static solver *mksolver(void) { static const solver_adt sadt = { PROBLEM_RDFT, mkplan, 0 }; S *slv = MKSOLVER(S, &sadt); return &(slv->super); } void X(reodft00e_splitradix_register)(planner *p) { REGISTER_SOLVER(p, mksolver()); } fftw-3.3.3/reodft/reodft11e-r2hc.c0000644000175000017500000001723412054410050013462 00000000000000/* * Copyright (c) 2003, 2007-11 Matteo Frigo * Copyright (c) 2003, 2007-11 Massachusetts Institute of Technology * * This program is free software; you can redistribute it and/or modify * it under the terms of the GNU General Public License as published by * the Free Software Foundation; either version 2 of the License, or * (at your option) any later version. * * This program is distributed in the hope that it will be useful, * but WITHOUT ANY WARRANTY; without even the implied warranty of * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the * GNU General Public License for more details. * * You should have received a copy of the GNU General Public License * along with this program; if not, write to the Free Software * Foundation, Inc., 51 Franklin Street, Fifth Floor, Boston, MA 02110-1301 USA * */ /* Do an R{E,O}DFT11 problem via an R2HC problem, with some pre/post-processing ala FFTPACK. Use a trick from: S. C. Chan and K. L. Ho, "Direct methods for computing discrete sinusoidal transforms," IEE Proceedings F 137 (6), 433--442 (1990). to re-express as an REDFT01 (DCT-III) problem. NOTE: We no longer use this algorithm, because it turns out to suffer a catastrophic loss of accuracy for certain inputs, apparently because its post-processing multiplies the output by a cosine. Near the zero of the cosine, the REDFT01 must produce a near-singular output. */ #include "reodft.h" typedef struct { solver super; } S; typedef struct { plan_rdft super; plan *cld; twid *td, *td2; INT is, os; INT n; INT vl; INT ivs, ovs; rdft_kind kind; } P; static void apply_re11(const plan *ego_, R *I, R *O) { const P *ego = (const P *) ego_; INT is = ego->is, os = ego->os; INT i, n = ego->n; INT iv, vl = ego->vl; INT ivs = ego->ivs, ovs = ego->ovs; R *W; R *buf; E cur; buf = (R *) MALLOC(sizeof(R) * n, BUFFERS); for (iv = 0; iv < vl; ++iv, I += ivs, O += ovs) { /* I wish that this didn't require an extra pass. */ /* FIXME: use recursive/cascade summation for better stability? */ buf[n - 1] = cur = K(2.0) * I[is * (n - 1)]; for (i = n - 1; i > 0; --i) { E curnew; buf[(i - 1)] = curnew = K(2.0) * I[is * (i - 1)] - cur; cur = curnew; } W = ego->td->W; for (i = 1; i < n - i; ++i) { E a, b, apb, amb, wa, wb; a = buf[i]; b = buf[n - i]; apb = a + b; amb = a - b; wa = W[2*i]; wb = W[2*i + 1]; buf[i] = wa * amb + wb * apb; buf[n - i] = wa * apb - wb * amb; } if (i == n - i) { buf[i] = K(2.0) * buf[i] * W[2*i]; } { plan_rdft *cld = (plan_rdft *) ego->cld; cld->apply((plan *) cld, buf, buf); } W = ego->td2->W; O[0] = W[0] * buf[0]; for (i = 1; i < n - i; ++i) { E a, b; INT k; a = buf[i]; b = buf[n - i]; k = i + i; O[os * (k - 1)] = W[k - 1] * (a - b); O[os * k] = W[k] * (a + b); } if (i == n - i) { O[os * (n - 1)] = W[n - 1] * buf[i]; } } X(ifree)(buf); } /* like for rodft01, rodft11 is obtained from redft11 by reversing the input and flipping the sign of every other output. */ static void apply_ro11(const plan *ego_, R *I, R *O) { const P *ego = (const P *) ego_; INT is = ego->is, os = ego->os; INT i, n = ego->n; INT iv, vl = ego->vl; INT ivs = ego->ivs, ovs = ego->ovs; R *W; R *buf; E cur; buf = (R *) MALLOC(sizeof(R) * n, BUFFERS); for (iv = 0; iv < vl; ++iv, I += ivs, O += ovs) { /* I wish that this didn't require an extra pass. */ /* FIXME: use recursive/cascade summation for better stability? */ buf[n - 1] = cur = K(2.0) * I[0]; for (i = n - 1; i > 0; --i) { E curnew; buf[(i - 1)] = curnew = K(2.0) * I[is * (n - i)] - cur; cur = curnew; } W = ego->td->W; for (i = 1; i < n - i; ++i) { E a, b, apb, amb, wa, wb; a = buf[i]; b = buf[n - i]; apb = a + b; amb = a - b; wa = W[2*i]; wb = W[2*i + 1]; buf[i] = wa * amb + wb * apb; buf[n - i] = wa * apb - wb * amb; } if (i == n - i) { buf[i] = K(2.0) * buf[i] * W[2*i]; } { plan_rdft *cld = (plan_rdft *) ego->cld; cld->apply((plan *) cld, buf, buf); } W = ego->td2->W; O[0] = W[0] * buf[0]; for (i = 1; i < n - i; ++i) { E a, b; INT k; a = buf[i]; b = buf[n - i]; k = i + i; O[os * (k - 1)] = W[k - 1] * (b - a); O[os * k] = W[k] * (a + b); } if (i == n - i) { O[os * (n - 1)] = -W[n - 1] * buf[i]; } } X(ifree)(buf); } static void awake(plan *ego_, enum wakefulness wakefulness) { P *ego = (P *) ego_; static const tw_instr reodft010e_tw[] = { { TW_COS, 0, 1 }, { TW_SIN, 0, 1 }, { TW_NEXT, 1, 0 } }; static const tw_instr reodft11e_tw[] = { { TW_COS, 1, 1 }, { TW_NEXT, 2, 0 } }; X(plan_awake)(ego->cld, wakefulness); X(twiddle_awake)(wakefulness, &ego->td, reodft010e_tw, 4*ego->n, 1, ego->n/2+1); X(twiddle_awake)(wakefulness, &ego->td2, reodft11e_tw, 8*ego->n, 1, ego->n * 2); } static void destroy(plan *ego_) { P *ego = (P *) ego_; X(plan_destroy_internal)(ego->cld); } static void print(const plan *ego_, printer *p) { const P *ego = (const P *) ego_; p->print(p, "(%se-r2hc-%D%v%(%p%))", X(rdft_kind_str)(ego->kind), ego->n, ego->vl, ego->cld); } static int applicable0(const solver *ego_, const problem *p_) { const problem_rdft *p = (const problem_rdft *) p_; UNUSED(ego_); return (1 && p->sz->rnk == 1 && p->vecsz->rnk <= 1 && (p->kind[0] == REDFT11 || p->kind[0] == RODFT11) ); } static int applicable(const solver *ego, const problem *p, const planner *plnr) { return (!NO_SLOWP(plnr) && applicable0(ego, p)); } static plan *mkplan(const solver *ego_, const problem *p_, planner *plnr) { P *pln; const problem_rdft *p; plan *cld; R *buf; INT n; opcnt ops; static const plan_adt padt = { X(rdft_solve), awake, print, destroy }; if (!applicable(ego_, p_, plnr)) return (plan *)0; p = (const problem_rdft *) p_; n = p->sz->dims[0].n; buf = (R *) MALLOC(sizeof(R) * n, BUFFERS); cld = X(mkplan_d)(plnr, X(mkproblem_rdft_1_d)(X(mktensor_1d)(n, 1, 1), X(mktensor_0d)(), buf, buf, R2HC)); X(ifree)(buf); if (!cld) return (plan *)0; pln = MKPLAN_RDFT(P, &padt, p->kind[0]==REDFT11 ? apply_re11:apply_ro11); pln->n = n; pln->is = p->sz->dims[0].is; pln->os = p->sz->dims[0].os; pln->cld = cld; pln->td = pln->td2 = 0; pln->kind = p->kind[0]; X(tensor_tornk1)(p->vecsz, &pln->vl, &pln->ivs, &pln->ovs); X(ops_zero)(&ops); ops.other = 5 + (n-1) * 2 + (n-1)/2 * 12 + (1 - n % 2) * 6; ops.add = (n - 1) * 1 + (n-1)/2 * 6; ops.mul = 2 + (n-1) * 1 + (n-1)/2 * 6 + (1 - n % 2) * 3; X(ops_zero)(&pln->super.super.ops); X(ops_madd2)(pln->vl, &ops, &pln->super.super.ops); X(ops_madd2)(pln->vl, &cld->ops, &pln->super.super.ops); return &(pln->super.super); } /* constructor */ static solver *mksolver(void) { static const solver_adt sadt = { PROBLEM_RDFT, mkplan, 0 }; S *slv = MKSOLVER(S, &sadt); return &(slv->super); } void X(reodft11e_r2hc_register)(planner *p) { REGISTER_SOLVER(p, mksolver()); } fftw-3.3.3/reodft/redft00e-r2hc.c0000644000175000017500000001356512054410050013304 00000000000000/* * Copyright (c) 2003, 2007-11 Matteo Frigo * Copyright (c) 2003, 2007-11 Massachusetts Institute of Technology * * This program is free software; you can redistribute it and/or modify * it under the terms of the GNU General Public License as published by * the Free Software Foundation; either version 2 of the License, or * (at your option) any later version. * * This program is distributed in the hope that it will be useful, * but WITHOUT ANY WARRANTY; without even the implied warranty of * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the * GNU General Public License for more details. * * You should have received a copy of the GNU General Public License * along with this program; if not, write to the Free Software * Foundation, Inc., 51 Franklin Street, Fifth Floor, Boston, MA 02110-1301 USA * */ /* Do a REDFT00 problem via an R2HC problem, with some pre/post-processing. This code uses the trick from FFTPACK, also documented in a similar form by Numerical Recipes. Unfortunately, this algorithm seems to have intrinsic numerical problems (similar to those in reodft11e-r2hc.c), possibly due to the fact that it multiplies its input by a cosine, causing a loss of precision near the zero. For transforms of 16k points, it has already lost three or four decimal places of accuracy, which we deem unacceptable. So, we have abandoned this algorithm in favor of the one in redft00-r2hc-pad.c, which unfortunately sacrifices 30-50% in speed. The only other alternative in the literature that does not have similar numerical difficulties seems to be the direct adaptation of the Cooley-Tukey decomposition for symmetric data, but this would require a whole new set of codelets and it's not clear that it's worth it at this point. However, we did implement the latter algorithm for the specific case of odd n (logically adapting the split-radix algorithm); see reodft00e-splitradix.c. */ #include "reodft.h" typedef struct { solver super; } S; typedef struct { plan_rdft super; plan *cld; twid *td; INT is, os; INT n; INT vl; INT ivs, ovs; } P; static void apply(const plan *ego_, R *I, R *O) { const P *ego = (const P *) ego_; INT is = ego->is, os = ego->os; INT i, n = ego->n; INT iv, vl = ego->vl; INT ivs = ego->ivs, ovs = ego->ovs; R *W = ego->td->W; R *buf; E csum; buf = (R *) MALLOC(sizeof(R) * n, BUFFERS); for (iv = 0; iv < vl; ++iv, I += ivs, O += ovs) { buf[0] = I[0] + I[is * n]; csum = I[0] - I[is * n]; for (i = 1; i < n - i; ++i) { E a, b, apb, amb; a = I[is * i]; b = I[is * (n - i)]; csum += W[2*i] * (amb = K(2.0)*(a - b)); amb = W[2*i+1] * amb; apb = (a + b); buf[i] = apb - amb; buf[n - i] = apb + amb; } if (i == n - i) { buf[i] = K(2.0) * I[is * i]; } { plan_rdft *cld = (plan_rdft *) ego->cld; cld->apply((plan *) cld, buf, buf); } /* FIXME: use recursive/cascade summation for better stability? */ O[0] = buf[0]; O[os] = csum; for (i = 1; i + i < n; ++i) { INT k = i + i; O[os * k] = buf[i]; O[os * (k + 1)] = O[os * (k - 1)] - buf[n - i]; } if (i + i == n) { O[os * n] = buf[i]; } } X(ifree)(buf); } static void awake(plan *ego_, enum wakefulness wakefulness) { P *ego = (P *) ego_; static const tw_instr redft00e_tw[] = { { TW_COS, 0, 1 }, { TW_SIN, 0, 1 }, { TW_NEXT, 1, 0 } }; X(plan_awake)(ego->cld, wakefulness); X(twiddle_awake)(wakefulness, &ego->td, redft00e_tw, 2*ego->n, 1, (ego->n+1)/2); } static void destroy(plan *ego_) { P *ego = (P *) ego_; X(plan_destroy_internal)(ego->cld); } static void print(const plan *ego_, printer *p) { const P *ego = (const P *) ego_; p->print(p, "(redft00e-r2hc-%D%v%(%p%))", ego->n + 1, ego->vl, ego->cld); } static int applicable0(const solver *ego_, const problem *p_) { const problem_rdft *p = (const problem_rdft *) p_; UNUSED(ego_); return (1 && p->sz->rnk == 1 && p->vecsz->rnk <= 1 && p->kind[0] == REDFT00 && p->sz->dims[0].n > 1 /* n == 1 is not well-defined */ ); } static int applicable(const solver *ego, const problem *p, const planner *plnr) { return (!NO_SLOWP(plnr) && applicable0(ego, p)); } static plan *mkplan(const solver *ego_, const problem *p_, planner *plnr) { P *pln; const problem_rdft *p; plan *cld; R *buf; INT n; opcnt ops; static const plan_adt padt = { X(rdft_solve), awake, print, destroy }; if (!applicable(ego_, p_, plnr)) return (plan *)0; p = (const problem_rdft *) p_; n = p->sz->dims[0].n - 1; A(n > 0); buf = (R *) MALLOC(sizeof(R) * n, BUFFERS); cld = X(mkplan_d)(plnr, X(mkproblem_rdft_1_d)(X(mktensor_1d)(n, 1, 1), X(mktensor_0d)(), buf, buf, R2HC)); X(ifree)(buf); if (!cld) return (plan *)0; pln = MKPLAN_RDFT(P, &padt, apply); pln->n = n; pln->is = p->sz->dims[0].is; pln->os = p->sz->dims[0].os; pln->cld = cld; pln->td = 0; X(tensor_tornk1)(p->vecsz, &pln->vl, &pln->ivs, &pln->ovs); X(ops_zero)(&ops); ops.other = 8 + (n-1)/2 * 11 + (1 - n % 2) * 5; ops.add = 2 + (n-1)/2 * 5; ops.mul = (n-1)/2 * 3 + (1 - n % 2) * 1; X(ops_zero)(&pln->super.super.ops); X(ops_madd2)(pln->vl, &ops, &pln->super.super.ops); X(ops_madd2)(pln->vl, &cld->ops, &pln->super.super.ops); return &(pln->super.super); } /* constructor */ static solver *mksolver(void) { static const solver_adt sadt = { PROBLEM_RDFT, mkplan, 0 }; S *slv = MKSOLVER(S, &sadt); return &(slv->super); } void X(redft00e_r2hc_register)(planner *p) { REGISTER_SOLVER(p, mksolver()); } fftw-3.3.3/reodft/rodft00e-r2hc.c0000644000175000017500000001347512054410050013316 00000000000000/* * Copyright (c) 2003, 2007-11 Matteo Frigo * Copyright (c) 2003, 2007-11 Massachusetts Institute of Technology * * This program is free software; you can redistribute it and/or modify * it under the terms of the GNU General Public License as published by * the Free Software Foundation; either version 2 of the License, or * (at your option) any later version. * * This program is distributed in the hope that it will be useful, * but WITHOUT ANY WARRANTY; without even the implied warranty of * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the * GNU General Public License for more details. * * You should have received a copy of the GNU General Public License * along with this program; if not, write to the Free Software * Foundation, Inc., 51 Franklin Street, Fifth Floor, Boston, MA 02110-1301 USA * */ /* Do a RODFT00 problem via an R2HC problem, with some pre/post-processing. This code uses the trick from FFTPACK, also documented in a similar form by Numerical Recipes. Unfortunately, this algorithm seems to have intrinsic numerical problems (similar to those in reodft11e-r2hc.c), possibly due to the fact that it multiplies its input by a sine, causing a loss of precision near the zero. For transforms of 16k points, it has already lost three or four decimal places of accuracy, which we deem unacceptable. So, we have abandoned this algorithm in favor of the one in rodft00-r2hc-pad.c, which unfortunately sacrifices 30-50% in speed. The only other alternative in the literature that does not have similar numerical difficulties seems to be the direct adaptation of the Cooley-Tukey decomposition for antisymmetric data, but this would require a whole new set of codelets and it's not clear that it's worth it at this point. However, we did implement the latter algorithm for the specific case of odd n (logically adapting the split-radix algorithm); see reodft00e-splitradix.c. */ #include "reodft.h" typedef struct { solver super; } S; typedef struct { plan_rdft super; plan *cld; twid *td; INT is, os; INT n; INT vl; INT ivs, ovs; } P; static void apply(const plan *ego_, R *I, R *O) { const P *ego = (const P *) ego_; INT is = ego->is, os = ego->os; INT i, n = ego->n; INT iv, vl = ego->vl; INT ivs = ego->ivs, ovs = ego->ovs; R *W = ego->td->W; R *buf; buf = (R *) MALLOC(sizeof(R) * n, BUFFERS); for (iv = 0; iv < vl; ++iv, I += ivs, O += ovs) { buf[0] = 0; for (i = 1; i < n - i; ++i) { E a, b, apb, amb; a = I[is * (i - 1)]; b = I[is * ((n - i) - 1)]; apb = K(2.0) * W[i] * (a + b); amb = (a - b); buf[i] = apb + amb; buf[n - i] = apb - amb; } if (i == n - i) { buf[i] = K(4.0) * I[is * (i - 1)]; } { plan_rdft *cld = (plan_rdft *) ego->cld; cld->apply((plan *) cld, buf, buf); } /* FIXME: use recursive/cascade summation for better stability? */ O[0] = buf[0] * 0.5; for (i = 1; i + i < n - 1; ++i) { INT k = i + i; O[os * (k - 1)] = -buf[n - i]; O[os * k] = O[os * (k - 2)] + buf[i]; } if (i + i == n - 1) { O[os * (n - 2)] = -buf[n - i]; } } X(ifree)(buf); } static void awake(plan *ego_, enum wakefulness wakefulness) { P *ego = (P *) ego_; static const tw_instr rodft00e_tw[] = { { TW_SIN, 0, 1 }, { TW_NEXT, 1, 0 } }; X(plan_awake)(ego->cld, wakefulness); X(twiddle_awake)(wakefulness, &ego->td, rodft00e_tw, 2*ego->n, 1, (ego->n+1)/2); } static void destroy(plan *ego_) { P *ego = (P *) ego_; X(plan_destroy_internal)(ego->cld); } static void print(const plan *ego_, printer *p) { const P *ego = (const P *) ego_; p->print(p, "(rodft00e-r2hc-%D%v%(%p%))", ego->n - 1, ego->vl, ego->cld); } static int applicable0(const solver *ego_, const problem *p_) { const problem_rdft *p = (const problem_rdft *) p_; UNUSED(ego_); return (1 && p->sz->rnk == 1 && p->vecsz->rnk <= 1 && p->kind[0] == RODFT00 ); } static int applicable(const solver *ego, const problem *p, const planner *plnr) { return (!NO_SLOWP(plnr) && applicable0(ego, p)); } static plan *mkplan(const solver *ego_, const problem *p_, planner *plnr) { P *pln; const problem_rdft *p; plan *cld; R *buf; INT n; opcnt ops; static const plan_adt padt = { X(rdft_solve), awake, print, destroy }; if (!applicable(ego_, p_, plnr)) return (plan *)0; p = (const problem_rdft *) p_; n = p->sz->dims[0].n + 1; buf = (R *) MALLOC(sizeof(R) * n, BUFFERS); cld = X(mkplan_d)(plnr, X(mkproblem_rdft_1_d)(X(mktensor_1d)(n, 1, 1), X(mktensor_0d)(), buf, buf, R2HC)); X(ifree)(buf); if (!cld) return (plan *)0; pln = MKPLAN_RDFT(P, &padt, apply); pln->n = n; pln->is = p->sz->dims[0].is; pln->os = p->sz->dims[0].os; pln->cld = cld; pln->td = 0; X(tensor_tornk1)(p->vecsz, &pln->vl, &pln->ivs, &pln->ovs); X(ops_zero)(&ops); ops.other = 4 + (n-1)/2 * 5 + (n-2)/2 * 5; ops.add = (n-1)/2 * 4 + (n-2)/2 * 1; ops.mul = 1 + (n-1)/2 * 2; if (n % 2 == 0) ops.mul += 1; X(ops_zero)(&pln->super.super.ops); X(ops_madd2)(pln->vl, &ops, &pln->super.super.ops); X(ops_madd2)(pln->vl, &cld->ops, &pln->super.super.ops); return &(pln->super.super); } /* constructor */ static solver *mksolver(void) { static const solver_adt sadt = { PROBLEM_RDFT, mkplan, 0 }; S *slv = MKSOLVER(S, &sadt); return &(slv->super); } void X(rodft00e_r2hc_register)(planner *p) { REGISTER_SOLVER(p, mksolver()); } fftw-3.3.3/api/0000755000175000017500000000000012054411110010225 500000000000000fftw-3.3.3/api/fftw3.h0000644000175000017500000004250612054410050011360 00000000000000/* * Copyright (c) 2003, 2007-11 Matteo Frigo * Copyright (c) 2003, 2007-11 Massachusetts Institute of Technology * * The following statement of license applies *only* to this header file, * and *not* to the other files distributed with FFTW or derived therefrom: * * Redistribution and use in source and binary forms, with or without * modification, are permitted provided that the following conditions * are met: * * 1. Redistributions of source code must retain the above copyright * notice, this list of conditions and the following disclaimer. * * 2. Redistributions in binary form must reproduce the above copyright * notice, this list of conditions and the following disclaimer in the * documentation and/or other materials provided with the distribution. * * THIS SOFTWARE IS PROVIDED BY THE AUTHOR ``AS IS'' AND ANY EXPRESS * OR IMPLIED WARRANTIES, INCLUDING, BUT NOT LIMITED TO, THE IMPLIED * WARRANTIES OF MERCHANTABILITY AND FITNESS FOR A PARTICULAR PURPOSE * ARE DISCLAIMED. IN NO EVENT SHALL THE AUTHOR BE LIABLE FOR ANY * DIRECT, INDIRECT, INCIDENTAL, SPECIAL, EXEMPLARY, OR CONSEQUENTIAL * DAMAGES (INCLUDING, BUT NOT LIMITED TO, PROCUREMENT OF SUBSTITUTE * GOODS OR SERVICES; LOSS OF USE, DATA, OR PROFITS; OR BUSINESS * INTERRUPTION) HOWEVER CAUSED AND ON ANY THEORY OF LIABILITY, * WHETHER IN CONTRACT, STRICT LIABILITY, OR TORT (INCLUDING * NEGLIGENCE OR OTHERWISE) ARISING IN ANY WAY OUT OF THE USE OF THIS * SOFTWARE, EVEN IF ADVISED OF THE POSSIBILITY OF SUCH DAMAGE. */ /***************************** NOTE TO USERS ********************************* * * THIS IS A HEADER FILE, NOT A MANUAL * * If you want to know how to use FFTW, please read the manual, * online at http://www.fftw.org/doc/ and also included with FFTW. * For a quick start, see the manual's tutorial section. * * (Reading header files to learn how to use a library is a habit * stemming from code lacking a proper manual. Arguably, it's a * *bad* habit in most cases, because header files can contain * interfaces that are not part of the public, stable API.) * ****************************************************************************/ #ifndef FFTW3_H #define FFTW3_H #include #ifdef __cplusplus extern "C" { #endif /* __cplusplus */ /* If is included, use the C99 complex type. Otherwise define a type bit-compatible with C99 complex */ #if !defined(FFTW_NO_Complex) && defined(_Complex_I) && defined(complex) && defined(I) # define FFTW_DEFINE_COMPLEX(R, C) typedef R _Complex C #else # define FFTW_DEFINE_COMPLEX(R, C) typedef R C[2] #endif #define FFTW_CONCAT(prefix, name) prefix ## name #define FFTW_MANGLE_DOUBLE(name) FFTW_CONCAT(fftw_, name) #define FFTW_MANGLE_FLOAT(name) FFTW_CONCAT(fftwf_, name) #define FFTW_MANGLE_LONG_DOUBLE(name) FFTW_CONCAT(fftwl_, name) #define FFTW_MANGLE_QUAD(name) FFTW_CONCAT(fftwq_, name) /* IMPORTANT: for Windows compilers, you should add a line #define FFTW_DLL here and in kernel/ifftw.h if you are compiling/using FFTW as a DLL, in order to do the proper importing/exporting, or alternatively compile with -DFFTW_DLL or the equivalent command-line flag. This is not necessary under MinGW/Cygwin, where libtool does the imports/exports automatically. */ #if defined(FFTW_DLL) && (defined(_WIN32) || defined(__WIN32__)) /* annoying Windows syntax for shared-library declarations */ # if defined(COMPILING_FFTW) /* defined in api.h when compiling FFTW */ # define FFTW_EXTERN extern __declspec(dllexport) # else /* user is calling FFTW; import symbol */ # define FFTW_EXTERN extern __declspec(dllimport) # endif #else # define FFTW_EXTERN extern #endif enum fftw_r2r_kind_do_not_use_me { FFTW_R2HC=0, FFTW_HC2R=1, FFTW_DHT=2, FFTW_REDFT00=3, FFTW_REDFT01=4, FFTW_REDFT10=5, FFTW_REDFT11=6, FFTW_RODFT00=7, FFTW_RODFT01=8, FFTW_RODFT10=9, FFTW_RODFT11=10 }; struct fftw_iodim_do_not_use_me { int n; /* dimension size */ int is; /* input stride */ int os; /* output stride */ }; #include /* for ptrdiff_t */ struct fftw_iodim64_do_not_use_me { ptrdiff_t n; /* dimension size */ ptrdiff_t is; /* input stride */ ptrdiff_t os; /* output stride */ }; typedef void (*fftw_write_char_func_do_not_use_me)(char c, void *); typedef int (*fftw_read_char_func_do_not_use_me)(void *); /* huge second-order macro that defines prototypes for all API functions. We expand this macro for each supported precision X: name-mangling macro R: real data type C: complex data type */ #define FFTW_DEFINE_API(X, R, C) \ \ FFTW_DEFINE_COMPLEX(R, C); \ \ typedef struct X(plan_s) *X(plan); \ \ typedef struct fftw_iodim_do_not_use_me X(iodim); \ typedef struct fftw_iodim64_do_not_use_me X(iodim64); \ \ typedef enum fftw_r2r_kind_do_not_use_me X(r2r_kind); \ \ typedef fftw_write_char_func_do_not_use_me X(write_char_func); \ typedef fftw_read_char_func_do_not_use_me X(read_char_func); \ \ FFTW_EXTERN void X(execute)(const X(plan) p); \ \ FFTW_EXTERN X(plan) X(plan_dft)(int rank, const int *n, \ C *in, C *out, int sign, unsigned flags); \ \ FFTW_EXTERN X(plan) X(plan_dft_1d)(int n, C *in, C *out, int sign, \ unsigned flags); \ FFTW_EXTERN X(plan) X(plan_dft_2d)(int n0, int n1, \ C *in, C *out, int sign, unsigned flags); \ FFTW_EXTERN X(plan) X(plan_dft_3d)(int n0, int n1, int n2, \ C *in, C *out, int sign, unsigned flags); \ \ FFTW_EXTERN X(plan) X(plan_many_dft)(int rank, const int *n, \ int howmany, \ C *in, const int *inembed, \ int istride, int idist, \ C *out, const int *onembed, \ int ostride, int odist, \ int sign, unsigned flags); \ \ FFTW_EXTERN X(plan) X(plan_guru_dft)(int rank, const X(iodim) *dims, \ int howmany_rank, \ const X(iodim) *howmany_dims, \ C *in, C *out, \ int sign, unsigned flags); \ FFTW_EXTERN X(plan) X(plan_guru_split_dft)(int rank, const X(iodim) *dims, \ int howmany_rank, \ const X(iodim) *howmany_dims, \ R *ri, R *ii, R *ro, R *io, \ unsigned flags); \ \ FFTW_EXTERN X(plan) X(plan_guru64_dft)(int rank, \ const X(iodim64) *dims, \ int howmany_rank, \ const X(iodim64) *howmany_dims, \ C *in, C *out, \ int sign, unsigned flags); \ FFTW_EXTERN X(plan) X(plan_guru64_split_dft)(int rank, \ const X(iodim64) *dims, \ int howmany_rank, \ const X(iodim64) *howmany_dims, \ R *ri, R *ii, R *ro, R *io, \ unsigned flags); \ \ FFTW_EXTERN void X(execute_dft)(const X(plan) p, C *in, C *out); \ FFTW_EXTERN void X(execute_split_dft)(const X(plan) p, R *ri, R *ii, \ R *ro, R *io); \ \ FFTW_EXTERN X(plan) X(plan_many_dft_r2c)(int rank, const int *n, \ int howmany, \ R *in, const int *inembed, \ int istride, int idist, \ C *out, const int *onembed, \ int ostride, int odist, \ unsigned flags); \ \ FFTW_EXTERN X(plan) X(plan_dft_r2c)(int rank, const int *n, \ R *in, C *out, unsigned flags); \ \ FFTW_EXTERN X(plan) X(plan_dft_r2c_1d)(int n,R *in,C *out,unsigned flags); \ FFTW_EXTERN X(plan) X(plan_dft_r2c_2d)(int n0, int n1, \ R *in, C *out, unsigned flags); \ FFTW_EXTERN X(plan) X(plan_dft_r2c_3d)(int n0, int n1, \ int n2, \ R *in, C *out, unsigned flags); \ \ \ FFTW_EXTERN X(plan) X(plan_many_dft_c2r)(int rank, const int *n, \ int howmany, \ C *in, const int *inembed, \ int istride, int idist, \ R *out, const int *onembed, \ int ostride, int odist, \ unsigned flags); \ \ FFTW_EXTERN X(plan) X(plan_dft_c2r)(int rank, const int *n, \ C *in, R *out, unsigned flags); \ \ FFTW_EXTERN X(plan) X(plan_dft_c2r_1d)(int n,C *in,R *out,unsigned flags); \ FFTW_EXTERN X(plan) X(plan_dft_c2r_2d)(int n0, int n1, \ C *in, R *out, unsigned flags); \ FFTW_EXTERN X(plan) X(plan_dft_c2r_3d)(int n0, int n1, \ int n2, \ C *in, R *out, unsigned flags); \ \ FFTW_EXTERN X(plan) X(plan_guru_dft_r2c)(int rank, const X(iodim) *dims, \ int howmany_rank, \ const X(iodim) *howmany_dims, \ R *in, C *out, \ unsigned flags); \ FFTW_EXTERN X(plan) X(plan_guru_dft_c2r)(int rank, const X(iodim) *dims, \ int howmany_rank, \ const X(iodim) *howmany_dims, \ C *in, R *out, \ unsigned flags); \ \ FFTW_EXTERN X(plan) X(plan_guru_split_dft_r2c)( \ int rank, const X(iodim) *dims, \ int howmany_rank, \ const X(iodim) *howmany_dims, \ R *in, R *ro, R *io, \ unsigned flags); \ FFTW_EXTERN X(plan) X(plan_guru_split_dft_c2r)( \ int rank, const X(iodim) *dims, \ int howmany_rank, \ const X(iodim) *howmany_dims, \ R *ri, R *ii, R *out, \ unsigned flags); \ \ FFTW_EXTERN X(plan) X(plan_guru64_dft_r2c)(int rank, \ const X(iodim64) *dims, \ int howmany_rank, \ const X(iodim64) *howmany_dims, \ R *in, C *out, \ unsigned flags); \ FFTW_EXTERN X(plan) X(plan_guru64_dft_c2r)(int rank, \ const X(iodim64) *dims, \ int howmany_rank, \ const X(iodim64) *howmany_dims, \ C *in, R *out, \ unsigned flags); \ \ FFTW_EXTERN X(plan) X(plan_guru64_split_dft_r2c)( \ int rank, const X(iodim64) *dims, \ int howmany_rank, \ const X(iodim64) *howmany_dims, \ R *in, R *ro, R *io, \ unsigned flags); \ FFTW_EXTERN X(plan) X(plan_guru64_split_dft_c2r)( \ int rank, const X(iodim64) *dims, \ int howmany_rank, \ const X(iodim64) *howmany_dims, \ R *ri, R *ii, R *out, \ unsigned flags); \ \ FFTW_EXTERN void X(execute_dft_r2c)(const X(plan) p, R *in, C *out); \ FFTW_EXTERN void X(execute_dft_c2r)(const X(plan) p, C *in, R *out); \ \ FFTW_EXTERN void X(execute_split_dft_r2c)(const X(plan) p, \ R *in, R *ro, R *io); \ FFTW_EXTERN void X(execute_split_dft_c2r)(const X(plan) p, \ R *ri, R *ii, R *out); \ \ FFTW_EXTERN X(plan) X(plan_many_r2r)(int rank, const int *n, \ int howmany, \ R *in, const int *inembed, \ int istride, int idist, \ R *out, const int *onembed, \ int ostride, int odist, \ const X(r2r_kind) *kind, unsigned flags); \ \ FFTW_EXTERN X(plan) X(plan_r2r)(int rank, const int *n, R *in, R *out, \ const X(r2r_kind) *kind, unsigned flags); \ \ FFTW_EXTERN X(plan) X(plan_r2r_1d)(int n, R *in, R *out, \ X(r2r_kind) kind, unsigned flags); \ FFTW_EXTERN X(plan) X(plan_r2r_2d)(int n0, int n1, R *in, R *out, \ X(r2r_kind) kind0, X(r2r_kind) kind1, \ unsigned flags); \ FFTW_EXTERN X(plan) X(plan_r2r_3d)(int n0, int n1, int n2, \ R *in, R *out, X(r2r_kind) kind0, \ X(r2r_kind) kind1, X(r2r_kind) kind2, \ unsigned flags); \ \ FFTW_EXTERN X(plan) X(plan_guru_r2r)(int rank, const X(iodim) *dims, \ int howmany_rank, \ const X(iodim) *howmany_dims, \ R *in, R *out, \ const X(r2r_kind) *kind, unsigned flags); \ \ FFTW_EXTERN X(plan) X(plan_guru64_r2r)(int rank, const X(iodim64) *dims, \ int howmany_rank, \ const X(iodim64) *howmany_dims, \ R *in, R *out, \ const X(r2r_kind) *kind, unsigned flags); \ \ FFTW_EXTERN void X(execute_r2r)(const X(plan) p, R *in, R *out); \ \ FFTW_EXTERN void X(destroy_plan)(X(plan) p); \ FFTW_EXTERN void X(forget_wisdom)(void); \ FFTW_EXTERN void X(cleanup)(void); \ \ FFTW_EXTERN void X(set_timelimit)(double t); \ \ FFTW_EXTERN void X(plan_with_nthreads)(int nthreads); \ FFTW_EXTERN int X(init_threads)(void); \ FFTW_EXTERN void X(cleanup_threads)(void); \ \ FFTW_EXTERN int X(export_wisdom_to_filename)(const char *filename); \ FFTW_EXTERN void X(export_wisdom_to_file)(FILE *output_file); \ FFTW_EXTERN char *X(export_wisdom_to_string)(void); \ FFTW_EXTERN void X(export_wisdom)(X(write_char_func) write_char, \ void *data); \ FFTW_EXTERN int X(import_system_wisdom)(void); \ FFTW_EXTERN int X(import_wisdom_from_filename)(const char *filename); \ FFTW_EXTERN int X(import_wisdom_from_file)(FILE *input_file); \ FFTW_EXTERN int X(import_wisdom_from_string)(const char *input_string); \ FFTW_EXTERN int X(import_wisdom)(X(read_char_func) read_char, void *data); \ \ FFTW_EXTERN void X(fprint_plan)(const X(plan) p, FILE *output_file); \ FFTW_EXTERN void X(print_plan)(const X(plan) p); \ \ FFTW_EXTERN void *X(malloc)(size_t n); \ FFTW_EXTERN R *X(alloc_real)(size_t n); \ FFTW_EXTERN C *X(alloc_complex)(size_t n); \ FFTW_EXTERN void X(free)(void *p); \ \ FFTW_EXTERN void X(flops)(const X(plan) p, \ double *add, double *mul, double *fmas); \ FFTW_EXTERN double X(estimate_cost)(const X(plan) p); \ FFTW_EXTERN double X(cost)(const X(plan) p); \ \ FFTW_EXTERN const char X(version)[]; \ FFTW_EXTERN const char X(cc)[]; \ FFTW_EXTERN const char X(codelet_optim)[]; /* end of FFTW_DEFINE_API macro */ FFTW_DEFINE_API(FFTW_MANGLE_DOUBLE, double, fftw_complex) FFTW_DEFINE_API(FFTW_MANGLE_FLOAT, float, fftwf_complex) FFTW_DEFINE_API(FFTW_MANGLE_LONG_DOUBLE, long double, fftwl_complex) /* __float128 (quad precision) is a gcc extension on i386, x86_64, and ia64 for gcc >= 4.6 (compiled in FFTW with --enable-quad-precision) */ #if (__GNUC__ > 4 || (__GNUC__ == 4 && __GNUC_MINOR__ >= 6)) \ && !(defined(__ICC) || defined(__INTEL_COMPILER)) \ && (defined(__i386__) || defined(__x86_64__) || defined(__ia64__)) # if !defined(FFTW_NO_Complex) && defined(_Complex_I) && defined(complex) && defined(I) /* note: __float128 is a typedef, which is not supported with the _Complex keyword in gcc, so instead we use this ugly __attribute__ version. However, we can't simply pass the __attribute__ version to FFTW_DEFINE_API because the __attribute__ confuses gcc in pointer types. Hence redefining FFTW_DEFINE_COMPLEX. Ugh. */ # undef FFTW_DEFINE_COMPLEX # define FFTW_DEFINE_COMPLEX(R, C) typedef _Complex float __attribute__((mode(TC))) C # endif FFTW_DEFINE_API(FFTW_MANGLE_QUAD, __float128, fftwq_complex) #endif #define FFTW_FORWARD (-1) #define FFTW_BACKWARD (+1) #define FFTW_NO_TIMELIMIT (-1.0) /* documented flags */ #define FFTW_MEASURE (0U) #define FFTW_DESTROY_INPUT (1U << 0) #define FFTW_UNALIGNED (1U << 1) #define FFTW_CONSERVE_MEMORY (1U << 2) #define FFTW_EXHAUSTIVE (1U << 3) /* NO_EXHAUSTIVE is default */ #define FFTW_PRESERVE_INPUT (1U << 4) /* cancels FFTW_DESTROY_INPUT */ #define FFTW_PATIENT (1U << 5) /* IMPATIENT is default */ #define FFTW_ESTIMATE (1U << 6) #define FFTW_WISDOM_ONLY (1U << 21) /* undocumented beyond-guru flags */ #define FFTW_ESTIMATE_PATIENT (1U << 7) #define FFTW_BELIEVE_PCOST (1U << 8) #define FFTW_NO_DFT_R2HC (1U << 9) #define FFTW_NO_NONTHREADED (1U << 10) #define FFTW_NO_BUFFERING (1U << 11) #define FFTW_NO_INDIRECT_OP (1U << 12) #define FFTW_ALLOW_LARGE_GENERIC (1U << 13) /* NO_LARGE_GENERIC is default */ #define FFTW_NO_RANK_SPLITS (1U << 14) #define FFTW_NO_VRANK_SPLITS (1U << 15) #define FFTW_NO_VRECURSE (1U << 16) #define FFTW_NO_SIMD (1U << 17) #define FFTW_NO_SLOW (1U << 18) #define FFTW_NO_FIXED_RADIX_LARGE_N (1U << 19) #define FFTW_ALLOW_PRUNING (1U << 20) #ifdef __cplusplus } /* extern "C" */ #endif /* __cplusplus */ #endif /* FFTW3_H */ fftw-3.3.3/api/fftw3.f0000644000175000017500000000461712054411075011367 00000000000000 INTEGER FFTW_R2HC PARAMETER (FFTW_R2HC=0) INTEGER FFTW_HC2R PARAMETER (FFTW_HC2R=1) INTEGER FFTW_DHT PARAMETER (FFTW_DHT=2) INTEGER FFTW_REDFT00 PARAMETER (FFTW_REDFT00=3) INTEGER FFTW_REDFT01 PARAMETER (FFTW_REDFT01=4) INTEGER FFTW_REDFT10 PARAMETER (FFTW_REDFT10=5) INTEGER FFTW_REDFT11 PARAMETER (FFTW_REDFT11=6) INTEGER FFTW_RODFT00 PARAMETER (FFTW_RODFT00=7) INTEGER FFTW_RODFT01 PARAMETER (FFTW_RODFT01=8) INTEGER FFTW_RODFT10 PARAMETER (FFTW_RODFT10=9) INTEGER FFTW_RODFT11 PARAMETER (FFTW_RODFT11=10) INTEGER FFTW_FORWARD PARAMETER (FFTW_FORWARD=-1) INTEGER FFTW_BACKWARD PARAMETER (FFTW_BACKWARD=+1) INTEGER FFTW_MEASURE PARAMETER (FFTW_MEASURE=0) INTEGER FFTW_DESTROY_INPUT PARAMETER (FFTW_DESTROY_INPUT=1) INTEGER FFTW_UNALIGNED PARAMETER (FFTW_UNALIGNED=2) INTEGER FFTW_CONSERVE_MEMORY PARAMETER (FFTW_CONSERVE_MEMORY=4) INTEGER FFTW_EXHAUSTIVE PARAMETER (FFTW_EXHAUSTIVE=8) INTEGER FFTW_PRESERVE_INPUT PARAMETER (FFTW_PRESERVE_INPUT=16) INTEGER FFTW_PATIENT PARAMETER (FFTW_PATIENT=32) INTEGER FFTW_ESTIMATE PARAMETER (FFTW_ESTIMATE=64) INTEGER FFTW_WISDOM_ONLY PARAMETER (FFTW_WISDOM_ONLY=2097152) INTEGER FFTW_ESTIMATE_PATIENT PARAMETER (FFTW_ESTIMATE_PATIENT=128) INTEGER FFTW_BELIEVE_PCOST PARAMETER (FFTW_BELIEVE_PCOST=256) INTEGER FFTW_NO_DFT_R2HC PARAMETER (FFTW_NO_DFT_R2HC=512) INTEGER FFTW_NO_NONTHREADED PARAMETER (FFTW_NO_NONTHREADED=1024) INTEGER FFTW_NO_BUFFERING PARAMETER (FFTW_NO_BUFFERING=2048) INTEGER FFTW_NO_INDIRECT_OP PARAMETER (FFTW_NO_INDIRECT_OP=4096) INTEGER FFTW_ALLOW_LARGE_GENERIC PARAMETER (FFTW_ALLOW_LARGE_GENERIC=8192) INTEGER FFTW_NO_RANK_SPLITS PARAMETER (FFTW_NO_RANK_SPLITS=16384) INTEGER FFTW_NO_VRANK_SPLITS PARAMETER (FFTW_NO_VRANK_SPLITS=32768) INTEGER FFTW_NO_VRECURSE PARAMETER (FFTW_NO_VRECURSE=65536) INTEGER FFTW_NO_SIMD PARAMETER (FFTW_NO_SIMD=131072) INTEGER FFTW_NO_SLOW PARAMETER (FFTW_NO_SLOW=262144) INTEGER FFTW_NO_FIXED_RADIX_LARGE_N PARAMETER (FFTW_NO_FIXED_RADIX_LARGE_N=524288) INTEGER FFTW_ALLOW_PRUNING PARAMETER (FFTW_ALLOW_PRUNING=1048576) fftw-3.3.3/api/fftw3l.f030000644000175000017500000006353412054411075011711 00000000000000! Generated automatically. DO NOT EDIT! type, bind(C) :: fftwl_iodim integer(C_INT) n, is, os end type fftwl_iodim type, bind(C) :: fftwl_iodim64 integer(C_INTPTR_T) n, is, os end type fftwl_iodim64 interface type(C_PTR) function fftwl_plan_dft(rank,n,in,out,sign,flags) bind(C, name='fftwl_plan_dft') import integer(C_INT), value :: rank integer(C_INT), dimension(*), intent(in) :: n complex(C_LONG_DOUBLE_COMPLEX), dimension(*), intent(out) :: in complex(C_LONG_DOUBLE_COMPLEX), dimension(*), intent(out) :: out integer(C_INT), value :: sign integer(C_INT), value :: flags end function fftwl_plan_dft type(C_PTR) function fftwl_plan_dft_1d(n,in,out,sign,flags) bind(C, name='fftwl_plan_dft_1d') import integer(C_INT), value :: n complex(C_LONG_DOUBLE_COMPLEX), dimension(*), intent(out) :: in complex(C_LONG_DOUBLE_COMPLEX), dimension(*), intent(out) :: out integer(C_INT), value :: sign integer(C_INT), value :: flags end function fftwl_plan_dft_1d type(C_PTR) function fftwl_plan_dft_2d(n0,n1,in,out,sign,flags) bind(C, name='fftwl_plan_dft_2d') import integer(C_INT), value :: n0 integer(C_INT), value :: n1 complex(C_LONG_DOUBLE_COMPLEX), dimension(*), intent(out) :: in complex(C_LONG_DOUBLE_COMPLEX), dimension(*), intent(out) :: out integer(C_INT), value :: sign integer(C_INT), value :: flags end function fftwl_plan_dft_2d type(C_PTR) function fftwl_plan_dft_3d(n0,n1,n2,in,out,sign,flags) bind(C, name='fftwl_plan_dft_3d') import integer(C_INT), value :: n0 integer(C_INT), value :: n1 integer(C_INT), value :: n2 complex(C_LONG_DOUBLE_COMPLEX), dimension(*), intent(out) :: in complex(C_LONG_DOUBLE_COMPLEX), dimension(*), intent(out) :: out integer(C_INT), value :: sign integer(C_INT), value :: flags end function fftwl_plan_dft_3d type(C_PTR) function fftwl_plan_many_dft(rank,n,howmany,in,inembed,istride,idist,out,onembed,ostride,odist,sign,flags) & bind(C, name='fftwl_plan_many_dft') import integer(C_INT), value :: rank integer(C_INT), dimension(*), intent(in) :: n integer(C_INT), value :: howmany complex(C_LONG_DOUBLE_COMPLEX), dimension(*), intent(out) :: in integer(C_INT), dimension(*), intent(in) :: inembed integer(C_INT), value :: istride integer(C_INT), value :: idist complex(C_LONG_DOUBLE_COMPLEX), dimension(*), intent(out) :: out integer(C_INT), dimension(*), intent(in) :: onembed integer(C_INT), value :: ostride integer(C_INT), value :: odist integer(C_INT), value :: sign integer(C_INT), value :: flags end function fftwl_plan_many_dft type(C_PTR) function fftwl_plan_guru_dft(rank,dims,howmany_rank,howmany_dims,in,out,sign,flags) & bind(C, name='fftwl_plan_guru_dft') import integer(C_INT), value :: rank type(fftwl_iodim), dimension(*), intent(in) :: dims integer(C_INT), value :: howmany_rank type(fftwl_iodim), dimension(*), intent(in) :: howmany_dims complex(C_LONG_DOUBLE_COMPLEX), dimension(*), intent(out) :: in complex(C_LONG_DOUBLE_COMPLEX), dimension(*), intent(out) :: out integer(C_INT), value :: sign integer(C_INT), value :: flags end function fftwl_plan_guru_dft type(C_PTR) function fftwl_plan_guru_split_dft(rank,dims,howmany_rank,howmany_dims,ri,ii,ro,io,flags) & bind(C, name='fftwl_plan_guru_split_dft') import integer(C_INT), value :: rank type(fftwl_iodim), dimension(*), intent(in) :: dims integer(C_INT), value :: howmany_rank type(fftwl_iodim), dimension(*), intent(in) :: howmany_dims real(C_LONG_DOUBLE), dimension(*), intent(out) :: ri real(C_LONG_DOUBLE), dimension(*), intent(out) :: ii real(C_LONG_DOUBLE), dimension(*), intent(out) :: ro real(C_LONG_DOUBLE), dimension(*), intent(out) :: io integer(C_INT), value :: flags end function fftwl_plan_guru_split_dft type(C_PTR) function fftwl_plan_guru64_dft(rank,dims,howmany_rank,howmany_dims,in,out,sign,flags) & bind(C, name='fftwl_plan_guru64_dft') import integer(C_INT), value :: rank type(fftwl_iodim64), dimension(*), intent(in) :: dims integer(C_INT), value :: howmany_rank type(fftwl_iodim64), dimension(*), intent(in) :: howmany_dims complex(C_LONG_DOUBLE_COMPLEX), dimension(*), intent(out) :: in complex(C_LONG_DOUBLE_COMPLEX), dimension(*), intent(out) :: out integer(C_INT), value :: sign integer(C_INT), value :: flags end function fftwl_plan_guru64_dft type(C_PTR) function fftwl_plan_guru64_split_dft(rank,dims,howmany_rank,howmany_dims,ri,ii,ro,io,flags) & bind(C, name='fftwl_plan_guru64_split_dft') import integer(C_INT), value :: rank type(fftwl_iodim64), dimension(*), intent(in) :: dims integer(C_INT), value :: howmany_rank type(fftwl_iodim64), dimension(*), intent(in) :: howmany_dims real(C_LONG_DOUBLE), dimension(*), intent(out) :: ri real(C_LONG_DOUBLE), dimension(*), intent(out) :: ii real(C_LONG_DOUBLE), dimension(*), intent(out) :: ro real(C_LONG_DOUBLE), dimension(*), intent(out) :: io integer(C_INT), value :: flags end function fftwl_plan_guru64_split_dft subroutine fftwl_execute_dft(p,in,out) bind(C, name='fftwl_execute_dft') import type(C_PTR), value :: p complex(C_LONG_DOUBLE_COMPLEX), dimension(*), intent(inout) :: in complex(C_LONG_DOUBLE_COMPLEX), dimension(*), intent(out) :: out end subroutine fftwl_execute_dft subroutine fftwl_execute_split_dft(p,ri,ii,ro,io) bind(C, name='fftwl_execute_split_dft') import type(C_PTR), value :: p real(C_LONG_DOUBLE), dimension(*), intent(inout) :: ri real(C_LONG_DOUBLE), dimension(*), intent(inout) :: ii real(C_LONG_DOUBLE), dimension(*), intent(out) :: ro real(C_LONG_DOUBLE), dimension(*), intent(out) :: io end subroutine fftwl_execute_split_dft type(C_PTR) function fftwl_plan_many_dft_r2c(rank,n,howmany,in,inembed,istride,idist,out,onembed,ostride,odist,flags) & bind(C, name='fftwl_plan_many_dft_r2c') import integer(C_INT), value :: rank integer(C_INT), dimension(*), intent(in) :: n integer(C_INT), value :: howmany real(C_LONG_DOUBLE), dimension(*), intent(out) :: in integer(C_INT), dimension(*), intent(in) :: inembed integer(C_INT), value :: istride integer(C_INT), value :: idist complex(C_LONG_DOUBLE_COMPLEX), dimension(*), intent(out) :: out integer(C_INT), dimension(*), intent(in) :: onembed integer(C_INT), value :: ostride integer(C_INT), value :: odist integer(C_INT), value :: flags end function fftwl_plan_many_dft_r2c type(C_PTR) function fftwl_plan_dft_r2c(rank,n,in,out,flags) bind(C, name='fftwl_plan_dft_r2c') import integer(C_INT), value :: rank integer(C_INT), dimension(*), intent(in) :: n real(C_LONG_DOUBLE), dimension(*), intent(out) :: in complex(C_LONG_DOUBLE_COMPLEX), dimension(*), intent(out) :: out integer(C_INT), value :: flags end function fftwl_plan_dft_r2c type(C_PTR) function fftwl_plan_dft_r2c_1d(n,in,out,flags) bind(C, name='fftwl_plan_dft_r2c_1d') import integer(C_INT), value :: n real(C_LONG_DOUBLE), dimension(*), intent(out) :: in complex(C_LONG_DOUBLE_COMPLEX), dimension(*), intent(out) :: out integer(C_INT), value :: flags end function fftwl_plan_dft_r2c_1d type(C_PTR) function fftwl_plan_dft_r2c_2d(n0,n1,in,out,flags) bind(C, name='fftwl_plan_dft_r2c_2d') import integer(C_INT), value :: n0 integer(C_INT), value :: n1 real(C_LONG_DOUBLE), dimension(*), intent(out) :: in complex(C_LONG_DOUBLE_COMPLEX), dimension(*), intent(out) :: out integer(C_INT), value :: flags end function fftwl_plan_dft_r2c_2d type(C_PTR) function fftwl_plan_dft_r2c_3d(n0,n1,n2,in,out,flags) bind(C, name='fftwl_plan_dft_r2c_3d') import integer(C_INT), value :: n0 integer(C_INT), value :: n1 integer(C_INT), value :: n2 real(C_LONG_DOUBLE), dimension(*), intent(out) :: in complex(C_LONG_DOUBLE_COMPLEX), dimension(*), intent(out) :: out integer(C_INT), value :: flags end function fftwl_plan_dft_r2c_3d type(C_PTR) function fftwl_plan_many_dft_c2r(rank,n,howmany,in,inembed,istride,idist,out,onembed,ostride,odist,flags) & bind(C, name='fftwl_plan_many_dft_c2r') import integer(C_INT), value :: rank integer(C_INT), dimension(*), intent(in) :: n integer(C_INT), value :: howmany complex(C_LONG_DOUBLE_COMPLEX), dimension(*), intent(out) :: in integer(C_INT), dimension(*), intent(in) :: inembed integer(C_INT), value :: istride integer(C_INT), value :: idist real(C_LONG_DOUBLE), dimension(*), intent(out) :: out integer(C_INT), dimension(*), intent(in) :: onembed integer(C_INT), value :: ostride integer(C_INT), value :: odist integer(C_INT), value :: flags end function fftwl_plan_many_dft_c2r type(C_PTR) function fftwl_plan_dft_c2r(rank,n,in,out,flags) bind(C, name='fftwl_plan_dft_c2r') import integer(C_INT), value :: rank integer(C_INT), dimension(*), intent(in) :: n complex(C_LONG_DOUBLE_COMPLEX), dimension(*), intent(out) :: in real(C_LONG_DOUBLE), dimension(*), intent(out) :: out integer(C_INT), value :: flags end function fftwl_plan_dft_c2r type(C_PTR) function fftwl_plan_dft_c2r_1d(n,in,out,flags) bind(C, name='fftwl_plan_dft_c2r_1d') import integer(C_INT), value :: n complex(C_LONG_DOUBLE_COMPLEX), dimension(*), intent(out) :: in real(C_LONG_DOUBLE), dimension(*), intent(out) :: out integer(C_INT), value :: flags end function fftwl_plan_dft_c2r_1d type(C_PTR) function fftwl_plan_dft_c2r_2d(n0,n1,in,out,flags) bind(C, name='fftwl_plan_dft_c2r_2d') import integer(C_INT), value :: n0 integer(C_INT), value :: n1 complex(C_LONG_DOUBLE_COMPLEX), dimension(*), intent(out) :: in real(C_LONG_DOUBLE), dimension(*), intent(out) :: out integer(C_INT), value :: flags end function fftwl_plan_dft_c2r_2d type(C_PTR) function fftwl_plan_dft_c2r_3d(n0,n1,n2,in,out,flags) bind(C, name='fftwl_plan_dft_c2r_3d') import integer(C_INT), value :: n0 integer(C_INT), value :: n1 integer(C_INT), value :: n2 complex(C_LONG_DOUBLE_COMPLEX), dimension(*), intent(out) :: in real(C_LONG_DOUBLE), dimension(*), intent(out) :: out integer(C_INT), value :: flags end function fftwl_plan_dft_c2r_3d type(C_PTR) function fftwl_plan_guru_dft_r2c(rank,dims,howmany_rank,howmany_dims,in,out,flags) & bind(C, name='fftwl_plan_guru_dft_r2c') import integer(C_INT), value :: rank type(fftwl_iodim), dimension(*), intent(in) :: dims integer(C_INT), value :: howmany_rank type(fftwl_iodim), dimension(*), intent(in) :: howmany_dims real(C_LONG_DOUBLE), dimension(*), intent(out) :: in complex(C_LONG_DOUBLE_COMPLEX), dimension(*), intent(out) :: out integer(C_INT), value :: flags end function fftwl_plan_guru_dft_r2c type(C_PTR) function fftwl_plan_guru_dft_c2r(rank,dims,howmany_rank,howmany_dims,in,out,flags) & bind(C, name='fftwl_plan_guru_dft_c2r') import integer(C_INT), value :: rank type(fftwl_iodim), dimension(*), intent(in) :: dims integer(C_INT), value :: howmany_rank type(fftwl_iodim), dimension(*), intent(in) :: howmany_dims complex(C_LONG_DOUBLE_COMPLEX), dimension(*), intent(out) :: in real(C_LONG_DOUBLE), dimension(*), intent(out) :: out integer(C_INT), value :: flags end function fftwl_plan_guru_dft_c2r type(C_PTR) function fftwl_plan_guru_split_dft_r2c(rank,dims,howmany_rank,howmany_dims,in,ro,io,flags) & bind(C, name='fftwl_plan_guru_split_dft_r2c') import integer(C_INT), value :: rank type(fftwl_iodim), dimension(*), intent(in) :: dims integer(C_INT), value :: howmany_rank type(fftwl_iodim), dimension(*), intent(in) :: howmany_dims real(C_LONG_DOUBLE), dimension(*), intent(out) :: in real(C_LONG_DOUBLE), dimension(*), intent(out) :: ro real(C_LONG_DOUBLE), dimension(*), intent(out) :: io integer(C_INT), value :: flags end function fftwl_plan_guru_split_dft_r2c type(C_PTR) function fftwl_plan_guru_split_dft_c2r(rank,dims,howmany_rank,howmany_dims,ri,ii,out,flags) & bind(C, name='fftwl_plan_guru_split_dft_c2r') import integer(C_INT), value :: rank type(fftwl_iodim), dimension(*), intent(in) :: dims integer(C_INT), value :: howmany_rank type(fftwl_iodim), dimension(*), intent(in) :: howmany_dims real(C_LONG_DOUBLE), dimension(*), intent(out) :: ri real(C_LONG_DOUBLE), dimension(*), intent(out) :: ii real(C_LONG_DOUBLE), dimension(*), intent(out) :: out integer(C_INT), value :: flags end function fftwl_plan_guru_split_dft_c2r type(C_PTR) function fftwl_plan_guru64_dft_r2c(rank,dims,howmany_rank,howmany_dims,in,out,flags) & bind(C, name='fftwl_plan_guru64_dft_r2c') import integer(C_INT), value :: rank type(fftwl_iodim64), dimension(*), intent(in) :: dims integer(C_INT), value :: howmany_rank type(fftwl_iodim64), dimension(*), intent(in) :: howmany_dims real(C_LONG_DOUBLE), dimension(*), intent(out) :: in complex(C_LONG_DOUBLE_COMPLEX), dimension(*), intent(out) :: out integer(C_INT), value :: flags end function fftwl_plan_guru64_dft_r2c type(C_PTR) function fftwl_plan_guru64_dft_c2r(rank,dims,howmany_rank,howmany_dims,in,out,flags) & bind(C, name='fftwl_plan_guru64_dft_c2r') import integer(C_INT), value :: rank type(fftwl_iodim64), dimension(*), intent(in) :: dims integer(C_INT), value :: howmany_rank type(fftwl_iodim64), dimension(*), intent(in) :: howmany_dims complex(C_LONG_DOUBLE_COMPLEX), dimension(*), intent(out) :: in real(C_LONG_DOUBLE), dimension(*), intent(out) :: out integer(C_INT), value :: flags end function fftwl_plan_guru64_dft_c2r type(C_PTR) function fftwl_plan_guru64_split_dft_r2c(rank,dims,howmany_rank,howmany_dims,in,ro,io,flags) & bind(C, name='fftwl_plan_guru64_split_dft_r2c') import integer(C_INT), value :: rank type(fftwl_iodim64), dimension(*), intent(in) :: dims integer(C_INT), value :: howmany_rank type(fftwl_iodim64), dimension(*), intent(in) :: howmany_dims real(C_LONG_DOUBLE), dimension(*), intent(out) :: in real(C_LONG_DOUBLE), dimension(*), intent(out) :: ro real(C_LONG_DOUBLE), dimension(*), intent(out) :: io integer(C_INT), value :: flags end function fftwl_plan_guru64_split_dft_r2c type(C_PTR) function fftwl_plan_guru64_split_dft_c2r(rank,dims,howmany_rank,howmany_dims,ri,ii,out,flags) & bind(C, name='fftwl_plan_guru64_split_dft_c2r') import integer(C_INT), value :: rank type(fftwl_iodim64), dimension(*), intent(in) :: dims integer(C_INT), value :: howmany_rank type(fftwl_iodim64), dimension(*), intent(in) :: howmany_dims real(C_LONG_DOUBLE), dimension(*), intent(out) :: ri real(C_LONG_DOUBLE), dimension(*), intent(out) :: ii real(C_LONG_DOUBLE), dimension(*), intent(out) :: out integer(C_INT), value :: flags end function fftwl_plan_guru64_split_dft_c2r subroutine fftwl_execute_dft_r2c(p,in,out) bind(C, name='fftwl_execute_dft_r2c') import type(C_PTR), value :: p real(C_LONG_DOUBLE), dimension(*), intent(inout) :: in complex(C_LONG_DOUBLE_COMPLEX), dimension(*), intent(out) :: out end subroutine fftwl_execute_dft_r2c subroutine fftwl_execute_dft_c2r(p,in,out) bind(C, name='fftwl_execute_dft_c2r') import type(C_PTR), value :: p complex(C_LONG_DOUBLE_COMPLEX), dimension(*), intent(inout) :: in real(C_LONG_DOUBLE), dimension(*), intent(out) :: out end subroutine fftwl_execute_dft_c2r subroutine fftwl_execute_split_dft_r2c(p,in,ro,io) bind(C, name='fftwl_execute_split_dft_r2c') import type(C_PTR), value :: p real(C_LONG_DOUBLE), dimension(*), intent(inout) :: in real(C_LONG_DOUBLE), dimension(*), intent(out) :: ro real(C_LONG_DOUBLE), dimension(*), intent(out) :: io end subroutine fftwl_execute_split_dft_r2c subroutine fftwl_execute_split_dft_c2r(p,ri,ii,out) bind(C, name='fftwl_execute_split_dft_c2r') import type(C_PTR), value :: p real(C_LONG_DOUBLE), dimension(*), intent(inout) :: ri real(C_LONG_DOUBLE), dimension(*), intent(inout) :: ii real(C_LONG_DOUBLE), dimension(*), intent(out) :: out end subroutine fftwl_execute_split_dft_c2r type(C_PTR) function fftwl_plan_many_r2r(rank,n,howmany,in,inembed,istride,idist,out,onembed,ostride,odist,kind,flags) & bind(C, name='fftwl_plan_many_r2r') import integer(C_INT), value :: rank integer(C_INT), dimension(*), intent(in) :: n integer(C_INT), value :: howmany real(C_LONG_DOUBLE), dimension(*), intent(out) :: in integer(C_INT), dimension(*), intent(in) :: inembed integer(C_INT), value :: istride integer(C_INT), value :: idist real(C_LONG_DOUBLE), dimension(*), intent(out) :: out integer(C_INT), dimension(*), intent(in) :: onembed integer(C_INT), value :: ostride integer(C_INT), value :: odist integer(C_FFTW_R2R_KIND), dimension(*), intent(in) :: kind integer(C_INT), value :: flags end function fftwl_plan_many_r2r type(C_PTR) function fftwl_plan_r2r(rank,n,in,out,kind,flags) bind(C, name='fftwl_plan_r2r') import integer(C_INT), value :: rank integer(C_INT), dimension(*), intent(in) :: n real(C_LONG_DOUBLE), dimension(*), intent(out) :: in real(C_LONG_DOUBLE), dimension(*), intent(out) :: out integer(C_FFTW_R2R_KIND), dimension(*), intent(in) :: kind integer(C_INT), value :: flags end function fftwl_plan_r2r type(C_PTR) function fftwl_plan_r2r_1d(n,in,out,kind,flags) bind(C, name='fftwl_plan_r2r_1d') import integer(C_INT), value :: n real(C_LONG_DOUBLE), dimension(*), intent(out) :: in real(C_LONG_DOUBLE), dimension(*), intent(out) :: out integer(C_FFTW_R2R_KIND), value :: kind integer(C_INT), value :: flags end function fftwl_plan_r2r_1d type(C_PTR) function fftwl_plan_r2r_2d(n0,n1,in,out,kind0,kind1,flags) bind(C, name='fftwl_plan_r2r_2d') import integer(C_INT), value :: n0 integer(C_INT), value :: n1 real(C_LONG_DOUBLE), dimension(*), intent(out) :: in real(C_LONG_DOUBLE), dimension(*), intent(out) :: out integer(C_FFTW_R2R_KIND), value :: kind0 integer(C_FFTW_R2R_KIND), value :: kind1 integer(C_INT), value :: flags end function fftwl_plan_r2r_2d type(C_PTR) function fftwl_plan_r2r_3d(n0,n1,n2,in,out,kind0,kind1,kind2,flags) bind(C, name='fftwl_plan_r2r_3d') import integer(C_INT), value :: n0 integer(C_INT), value :: n1 integer(C_INT), value :: n2 real(C_LONG_DOUBLE), dimension(*), intent(out) :: in real(C_LONG_DOUBLE), dimension(*), intent(out) :: out integer(C_FFTW_R2R_KIND), value :: kind0 integer(C_FFTW_R2R_KIND), value :: kind1 integer(C_FFTW_R2R_KIND), value :: kind2 integer(C_INT), value :: flags end function fftwl_plan_r2r_3d type(C_PTR) function fftwl_plan_guru_r2r(rank,dims,howmany_rank,howmany_dims,in,out,kind,flags) & bind(C, name='fftwl_plan_guru_r2r') import integer(C_INT), value :: rank type(fftwl_iodim), dimension(*), intent(in) :: dims integer(C_INT), value :: howmany_rank type(fftwl_iodim), dimension(*), intent(in) :: howmany_dims real(C_LONG_DOUBLE), dimension(*), intent(out) :: in real(C_LONG_DOUBLE), dimension(*), intent(out) :: out integer(C_FFTW_R2R_KIND), dimension(*), intent(in) :: kind integer(C_INT), value :: flags end function fftwl_plan_guru_r2r type(C_PTR) function fftwl_plan_guru64_r2r(rank,dims,howmany_rank,howmany_dims,in,out,kind,flags) & bind(C, name='fftwl_plan_guru64_r2r') import integer(C_INT), value :: rank type(fftwl_iodim64), dimension(*), intent(in) :: dims integer(C_INT), value :: howmany_rank type(fftwl_iodim64), dimension(*), intent(in) :: howmany_dims real(C_LONG_DOUBLE), dimension(*), intent(out) :: in real(C_LONG_DOUBLE), dimension(*), intent(out) :: out integer(C_FFTW_R2R_KIND), dimension(*), intent(in) :: kind integer(C_INT), value :: flags end function fftwl_plan_guru64_r2r subroutine fftwl_execute_r2r(p,in,out) bind(C, name='fftwl_execute_r2r') import type(C_PTR), value :: p real(C_LONG_DOUBLE), dimension(*), intent(inout) :: in real(C_LONG_DOUBLE), dimension(*), intent(out) :: out end subroutine fftwl_execute_r2r subroutine fftwl_destroy_plan(p) bind(C, name='fftwl_destroy_plan') import type(C_PTR), value :: p end subroutine fftwl_destroy_plan subroutine fftwl_forget_wisdom() bind(C, name='fftwl_forget_wisdom') import end subroutine fftwl_forget_wisdom subroutine fftwl_cleanup() bind(C, name='fftwl_cleanup') import end subroutine fftwl_cleanup subroutine fftwl_set_timelimit(t) bind(C, name='fftwl_set_timelimit') import real(C_DOUBLE), value :: t end subroutine fftwl_set_timelimit subroutine fftwl_plan_with_nthreads(nthreads) bind(C, name='fftwl_plan_with_nthreads') import integer(C_INT), value :: nthreads end subroutine fftwl_plan_with_nthreads integer(C_INT) function fftwl_init_threads() bind(C, name='fftwl_init_threads') import end function fftwl_init_threads subroutine fftwl_cleanup_threads() bind(C, name='fftwl_cleanup_threads') import end subroutine fftwl_cleanup_threads integer(C_INT) function fftwl_export_wisdom_to_filename(filename) bind(C, name='fftwl_export_wisdom_to_filename') import character(C_CHAR), dimension(*), intent(in) :: filename end function fftwl_export_wisdom_to_filename subroutine fftwl_export_wisdom_to_file(output_file) bind(C, name='fftwl_export_wisdom_to_file') import type(C_PTR), value :: output_file end subroutine fftwl_export_wisdom_to_file type(C_PTR) function fftwl_export_wisdom_to_string() bind(C, name='fftwl_export_wisdom_to_string') import end function fftwl_export_wisdom_to_string subroutine fftwl_export_wisdom(write_char,data) bind(C, name='fftwl_export_wisdom') import type(C_FUNPTR), value :: write_char type(C_PTR), value :: data end subroutine fftwl_export_wisdom integer(C_INT) function fftwl_import_system_wisdom() bind(C, name='fftwl_import_system_wisdom') import end function fftwl_import_system_wisdom integer(C_INT) function fftwl_import_wisdom_from_filename(filename) bind(C, name='fftwl_import_wisdom_from_filename') import character(C_CHAR), dimension(*), intent(in) :: filename end function fftwl_import_wisdom_from_filename integer(C_INT) function fftwl_import_wisdom_from_file(input_file) bind(C, name='fftwl_import_wisdom_from_file') import type(C_PTR), value :: input_file end function fftwl_import_wisdom_from_file integer(C_INT) function fftwl_import_wisdom_from_string(input_string) bind(C, name='fftwl_import_wisdom_from_string') import character(C_CHAR), dimension(*), intent(in) :: input_string end function fftwl_import_wisdom_from_string integer(C_INT) function fftwl_import_wisdom(read_char,data) bind(C, name='fftwl_import_wisdom') import type(C_FUNPTR), value :: read_char type(C_PTR), value :: data end function fftwl_import_wisdom subroutine fftwl_fprint_plan(p,output_file) bind(C, name='fftwl_fprint_plan') import type(C_PTR), value :: p type(C_PTR), value :: output_file end subroutine fftwl_fprint_plan subroutine fftwl_print_plan(p) bind(C, name='fftwl_print_plan') import type(C_PTR), value :: p end subroutine fftwl_print_plan type(C_PTR) function fftwl_malloc(n) bind(C, name='fftwl_malloc') import integer(C_SIZE_T), value :: n end function fftwl_malloc type(C_PTR) function fftwl_alloc_real(n) bind(C, name='fftwl_alloc_real') import integer(C_SIZE_T), value :: n end function fftwl_alloc_real type(C_PTR) function fftwl_alloc_complex(n) bind(C, name='fftwl_alloc_complex') import integer(C_SIZE_T), value :: n end function fftwl_alloc_complex subroutine fftwl_free(p) bind(C, name='fftwl_free') import type(C_PTR), value :: p end subroutine fftwl_free subroutine fftwl_flops(p,add,mul,fmas) bind(C, name='fftwl_flops') import type(C_PTR), value :: p real(C_DOUBLE), intent(out) :: add real(C_DOUBLE), intent(out) :: mul real(C_DOUBLE), intent(out) :: fmas end subroutine fftwl_flops real(C_DOUBLE) function fftwl_estimate_cost(p) bind(C, name='fftwl_estimate_cost') import type(C_PTR), value :: p end function fftwl_estimate_cost real(C_DOUBLE) function fftwl_cost(p) bind(C, name='fftwl_cost') import type(C_PTR), value :: p end function fftwl_cost end interface fftw-3.3.3/api/fftw3q.f030000644000175000017500000006075012054411075011713 00000000000000! Generated automatically. DO NOT EDIT! type, bind(C) :: fftwq_iodim integer(C_INT) n, is, os end type fftwq_iodim type, bind(C) :: fftwq_iodim64 integer(C_INTPTR_T) n, is, os end type fftwq_iodim64 interface type(C_PTR) function fftwq_plan_dft(rank,n,in,out,sign,flags) bind(C, name='fftwq_plan_dft') import integer(C_INT), value :: rank integer(C_INT), dimension(*), intent(in) :: n complex(16), dimension(*), intent(out) :: in complex(16), dimension(*), intent(out) :: out integer(C_INT), value :: sign integer(C_INT), value :: flags end function fftwq_plan_dft type(C_PTR) function fftwq_plan_dft_1d(n,in,out,sign,flags) bind(C, name='fftwq_plan_dft_1d') import integer(C_INT), value :: n complex(16), dimension(*), intent(out) :: in complex(16), dimension(*), intent(out) :: out integer(C_INT), value :: sign integer(C_INT), value :: flags end function fftwq_plan_dft_1d type(C_PTR) function fftwq_plan_dft_2d(n0,n1,in,out,sign,flags) bind(C, name='fftwq_plan_dft_2d') import integer(C_INT), value :: n0 integer(C_INT), value :: n1 complex(16), dimension(*), intent(out) :: in complex(16), dimension(*), intent(out) :: out integer(C_INT), value :: sign integer(C_INT), value :: flags end function fftwq_plan_dft_2d type(C_PTR) function fftwq_plan_dft_3d(n0,n1,n2,in,out,sign,flags) bind(C, name='fftwq_plan_dft_3d') import integer(C_INT), value :: n0 integer(C_INT), value :: n1 integer(C_INT), value :: n2 complex(16), dimension(*), intent(out) :: in complex(16), dimension(*), intent(out) :: out integer(C_INT), value :: sign integer(C_INT), value :: flags end function fftwq_plan_dft_3d type(C_PTR) function fftwq_plan_many_dft(rank,n,howmany,in,inembed,istride,idist,out,onembed,ostride,odist,sign,flags) & bind(C, name='fftwq_plan_many_dft') import integer(C_INT), value :: rank integer(C_INT), dimension(*), intent(in) :: n integer(C_INT), value :: howmany complex(16), dimension(*), intent(out) :: in integer(C_INT), dimension(*), intent(in) :: inembed integer(C_INT), value :: istride integer(C_INT), value :: idist complex(16), dimension(*), intent(out) :: out integer(C_INT), dimension(*), intent(in) :: onembed integer(C_INT), value :: ostride integer(C_INT), value :: odist integer(C_INT), value :: sign integer(C_INT), value :: flags end function fftwq_plan_many_dft type(C_PTR) function fftwq_plan_guru_dft(rank,dims,howmany_rank,howmany_dims,in,out,sign,flags) & bind(C, name='fftwq_plan_guru_dft') import integer(C_INT), value :: rank type(fftwq_iodim), dimension(*), intent(in) :: dims integer(C_INT), value :: howmany_rank type(fftwq_iodim), dimension(*), intent(in) :: howmany_dims complex(16), dimension(*), intent(out) :: in complex(16), dimension(*), intent(out) :: out integer(C_INT), value :: sign integer(C_INT), value :: flags end function fftwq_plan_guru_dft type(C_PTR) function fftwq_plan_guru_split_dft(rank,dims,howmany_rank,howmany_dims,ri,ii,ro,io,flags) & bind(C, name='fftwq_plan_guru_split_dft') import integer(C_INT), value :: rank type(fftwq_iodim), dimension(*), intent(in) :: dims integer(C_INT), value :: howmany_rank type(fftwq_iodim), dimension(*), intent(in) :: howmany_dims real(16), dimension(*), intent(out) :: ri real(16), dimension(*), intent(out) :: ii real(16), dimension(*), intent(out) :: ro real(16), dimension(*), intent(out) :: io integer(C_INT), value :: flags end function fftwq_plan_guru_split_dft type(C_PTR) function fftwq_plan_guru64_dft(rank,dims,howmany_rank,howmany_dims,in,out,sign,flags) & bind(C, name='fftwq_plan_guru64_dft') import integer(C_INT), value :: rank type(fftwq_iodim64), dimension(*), intent(in) :: dims integer(C_INT), value :: howmany_rank type(fftwq_iodim64), dimension(*), intent(in) :: howmany_dims complex(16), dimension(*), intent(out) :: in complex(16), dimension(*), intent(out) :: out integer(C_INT), value :: sign integer(C_INT), value :: flags end function fftwq_plan_guru64_dft type(C_PTR) function fftwq_plan_guru64_split_dft(rank,dims,howmany_rank,howmany_dims,ri,ii,ro,io,flags) & bind(C, name='fftwq_plan_guru64_split_dft') import integer(C_INT), value :: rank type(fftwq_iodim64), dimension(*), intent(in) :: dims integer(C_INT), value :: howmany_rank type(fftwq_iodim64), dimension(*), intent(in) :: howmany_dims real(16), dimension(*), intent(out) :: ri real(16), dimension(*), intent(out) :: ii real(16), dimension(*), intent(out) :: ro real(16), dimension(*), intent(out) :: io integer(C_INT), value :: flags end function fftwq_plan_guru64_split_dft subroutine fftwq_execute_dft(p,in,out) bind(C, name='fftwq_execute_dft') import type(C_PTR), value :: p complex(16), dimension(*), intent(inout) :: in complex(16), dimension(*), intent(out) :: out end subroutine fftwq_execute_dft subroutine fftwq_execute_split_dft(p,ri,ii,ro,io) bind(C, name='fftwq_execute_split_dft') import type(C_PTR), value :: p real(16), dimension(*), intent(inout) :: ri real(16), dimension(*), intent(inout) :: ii real(16), dimension(*), intent(out) :: ro real(16), dimension(*), intent(out) :: io end subroutine fftwq_execute_split_dft type(C_PTR) function fftwq_plan_many_dft_r2c(rank,n,howmany,in,inembed,istride,idist,out,onembed,ostride,odist,flags) & bind(C, name='fftwq_plan_many_dft_r2c') import integer(C_INT), value :: rank integer(C_INT), dimension(*), intent(in) :: n integer(C_INT), value :: howmany real(16), dimension(*), intent(out) :: in integer(C_INT), dimension(*), intent(in) :: inembed integer(C_INT), value :: istride integer(C_INT), value :: idist complex(16), dimension(*), intent(out) :: out integer(C_INT), dimension(*), intent(in) :: onembed integer(C_INT), value :: ostride integer(C_INT), value :: odist integer(C_INT), value :: flags end function fftwq_plan_many_dft_r2c type(C_PTR) function fftwq_plan_dft_r2c(rank,n,in,out,flags) bind(C, name='fftwq_plan_dft_r2c') import integer(C_INT), value :: rank integer(C_INT), dimension(*), intent(in) :: n real(16), dimension(*), intent(out) :: in complex(16), dimension(*), intent(out) :: out integer(C_INT), value :: flags end function fftwq_plan_dft_r2c type(C_PTR) function fftwq_plan_dft_r2c_1d(n,in,out,flags) bind(C, name='fftwq_plan_dft_r2c_1d') import integer(C_INT), value :: n real(16), dimension(*), intent(out) :: in complex(16), dimension(*), intent(out) :: out integer(C_INT), value :: flags end function fftwq_plan_dft_r2c_1d type(C_PTR) function fftwq_plan_dft_r2c_2d(n0,n1,in,out,flags) bind(C, name='fftwq_plan_dft_r2c_2d') import integer(C_INT), value :: n0 integer(C_INT), value :: n1 real(16), dimension(*), intent(out) :: in complex(16), dimension(*), intent(out) :: out integer(C_INT), value :: flags end function fftwq_plan_dft_r2c_2d type(C_PTR) function fftwq_plan_dft_r2c_3d(n0,n1,n2,in,out,flags) bind(C, name='fftwq_plan_dft_r2c_3d') import integer(C_INT), value :: n0 integer(C_INT), value :: n1 integer(C_INT), value :: n2 real(16), dimension(*), intent(out) :: in complex(16), dimension(*), intent(out) :: out integer(C_INT), value :: flags end function fftwq_plan_dft_r2c_3d type(C_PTR) function fftwq_plan_many_dft_c2r(rank,n,howmany,in,inembed,istride,idist,out,onembed,ostride,odist,flags) & bind(C, name='fftwq_plan_many_dft_c2r') import integer(C_INT), value :: rank integer(C_INT), dimension(*), intent(in) :: n integer(C_INT), value :: howmany complex(16), dimension(*), intent(out) :: in integer(C_INT), dimension(*), intent(in) :: inembed integer(C_INT), value :: istride integer(C_INT), value :: idist real(16), dimension(*), intent(out) :: out integer(C_INT), dimension(*), intent(in) :: onembed integer(C_INT), value :: ostride integer(C_INT), value :: odist integer(C_INT), value :: flags end function fftwq_plan_many_dft_c2r type(C_PTR) function fftwq_plan_dft_c2r(rank,n,in,out,flags) bind(C, name='fftwq_plan_dft_c2r') import integer(C_INT), value :: rank integer(C_INT), dimension(*), intent(in) :: n complex(16), dimension(*), intent(out) :: in real(16), dimension(*), intent(out) :: out integer(C_INT), value :: flags end function fftwq_plan_dft_c2r type(C_PTR) function fftwq_plan_dft_c2r_1d(n,in,out,flags) bind(C, name='fftwq_plan_dft_c2r_1d') import integer(C_INT), value :: n complex(16), dimension(*), intent(out) :: in real(16), dimension(*), intent(out) :: out integer(C_INT), value :: flags end function fftwq_plan_dft_c2r_1d type(C_PTR) function fftwq_plan_dft_c2r_2d(n0,n1,in,out,flags) bind(C, name='fftwq_plan_dft_c2r_2d') import integer(C_INT), value :: n0 integer(C_INT), value :: n1 complex(16), dimension(*), intent(out) :: in real(16), dimension(*), intent(out) :: out integer(C_INT), value :: flags end function fftwq_plan_dft_c2r_2d type(C_PTR) function fftwq_plan_dft_c2r_3d(n0,n1,n2,in,out,flags) bind(C, name='fftwq_plan_dft_c2r_3d') import integer(C_INT), value :: n0 integer(C_INT), value :: n1 integer(C_INT), value :: n2 complex(16), dimension(*), intent(out) :: in real(16), dimension(*), intent(out) :: out integer(C_INT), value :: flags end function fftwq_plan_dft_c2r_3d type(C_PTR) function fftwq_plan_guru_dft_r2c(rank,dims,howmany_rank,howmany_dims,in,out,flags) & bind(C, name='fftwq_plan_guru_dft_r2c') import integer(C_INT), value :: rank type(fftwq_iodim), dimension(*), intent(in) :: dims integer(C_INT), value :: howmany_rank type(fftwq_iodim), dimension(*), intent(in) :: howmany_dims real(16), dimension(*), intent(out) :: in complex(16), dimension(*), intent(out) :: out integer(C_INT), value :: flags end function fftwq_plan_guru_dft_r2c type(C_PTR) function fftwq_plan_guru_dft_c2r(rank,dims,howmany_rank,howmany_dims,in,out,flags) & bind(C, name='fftwq_plan_guru_dft_c2r') import integer(C_INT), value :: rank type(fftwq_iodim), dimension(*), intent(in) :: dims integer(C_INT), value :: howmany_rank type(fftwq_iodim), dimension(*), intent(in) :: howmany_dims complex(16), dimension(*), intent(out) :: in real(16), dimension(*), intent(out) :: out integer(C_INT), value :: flags end function fftwq_plan_guru_dft_c2r type(C_PTR) function fftwq_plan_guru_split_dft_r2c(rank,dims,howmany_rank,howmany_dims,in,ro,io,flags) & bind(C, name='fftwq_plan_guru_split_dft_r2c') import integer(C_INT), value :: rank type(fftwq_iodim), dimension(*), intent(in) :: dims integer(C_INT), value :: howmany_rank type(fftwq_iodim), dimension(*), intent(in) :: howmany_dims real(16), dimension(*), intent(out) :: in real(16), dimension(*), intent(out) :: ro real(16), dimension(*), intent(out) :: io integer(C_INT), value :: flags end function fftwq_plan_guru_split_dft_r2c type(C_PTR) function fftwq_plan_guru_split_dft_c2r(rank,dims,howmany_rank,howmany_dims,ri,ii,out,flags) & bind(C, name='fftwq_plan_guru_split_dft_c2r') import integer(C_INT), value :: rank type(fftwq_iodim), dimension(*), intent(in) :: dims integer(C_INT), value :: howmany_rank type(fftwq_iodim), dimension(*), intent(in) :: howmany_dims real(16), dimension(*), intent(out) :: ri real(16), dimension(*), intent(out) :: ii real(16), dimension(*), intent(out) :: out integer(C_INT), value :: flags end function fftwq_plan_guru_split_dft_c2r type(C_PTR) function fftwq_plan_guru64_dft_r2c(rank,dims,howmany_rank,howmany_dims,in,out,flags) & bind(C, name='fftwq_plan_guru64_dft_r2c') import integer(C_INT), value :: rank type(fftwq_iodim64), dimension(*), intent(in) :: dims integer(C_INT), value :: howmany_rank type(fftwq_iodim64), dimension(*), intent(in) :: howmany_dims real(16), dimension(*), intent(out) :: in complex(16), dimension(*), intent(out) :: out integer(C_INT), value :: flags end function fftwq_plan_guru64_dft_r2c type(C_PTR) function fftwq_plan_guru64_dft_c2r(rank,dims,howmany_rank,howmany_dims,in,out,flags) & bind(C, name='fftwq_plan_guru64_dft_c2r') import integer(C_INT), value :: rank type(fftwq_iodim64), dimension(*), intent(in) :: dims integer(C_INT), value :: howmany_rank type(fftwq_iodim64), dimension(*), intent(in) :: howmany_dims complex(16), dimension(*), intent(out) :: in real(16), dimension(*), intent(out) :: out integer(C_INT), value :: flags end function fftwq_plan_guru64_dft_c2r type(C_PTR) function fftwq_plan_guru64_split_dft_r2c(rank,dims,howmany_rank,howmany_dims,in,ro,io,flags) & bind(C, name='fftwq_plan_guru64_split_dft_r2c') import integer(C_INT), value :: rank type(fftwq_iodim64), dimension(*), intent(in) :: dims integer(C_INT), value :: howmany_rank type(fftwq_iodim64), dimension(*), intent(in) :: howmany_dims real(16), dimension(*), intent(out) :: in real(16), dimension(*), intent(out) :: ro real(16), dimension(*), intent(out) :: io integer(C_INT), value :: flags end function fftwq_plan_guru64_split_dft_r2c type(C_PTR) function fftwq_plan_guru64_split_dft_c2r(rank,dims,howmany_rank,howmany_dims,ri,ii,out,flags) & bind(C, name='fftwq_plan_guru64_split_dft_c2r') import integer(C_INT), value :: rank type(fftwq_iodim64), dimension(*), intent(in) :: dims integer(C_INT), value :: howmany_rank type(fftwq_iodim64), dimension(*), intent(in) :: howmany_dims real(16), dimension(*), intent(out) :: ri real(16), dimension(*), intent(out) :: ii real(16), dimension(*), intent(out) :: out integer(C_INT), value :: flags end function fftwq_plan_guru64_split_dft_c2r subroutine fftwq_execute_dft_r2c(p,in,out) bind(C, name='fftwq_execute_dft_r2c') import type(C_PTR), value :: p real(16), dimension(*), intent(inout) :: in complex(16), dimension(*), intent(out) :: out end subroutine fftwq_execute_dft_r2c subroutine fftwq_execute_dft_c2r(p,in,out) bind(C, name='fftwq_execute_dft_c2r') import type(C_PTR), value :: p complex(16), dimension(*), intent(inout) :: in real(16), dimension(*), intent(out) :: out end subroutine fftwq_execute_dft_c2r subroutine fftwq_execute_split_dft_r2c(p,in,ro,io) bind(C, name='fftwq_execute_split_dft_r2c') import type(C_PTR), value :: p real(16), dimension(*), intent(inout) :: in real(16), dimension(*), intent(out) :: ro real(16), dimension(*), intent(out) :: io end subroutine fftwq_execute_split_dft_r2c subroutine fftwq_execute_split_dft_c2r(p,ri,ii,out) bind(C, name='fftwq_execute_split_dft_c2r') import type(C_PTR), value :: p real(16), dimension(*), intent(inout) :: ri real(16), dimension(*), intent(inout) :: ii real(16), dimension(*), intent(out) :: out end subroutine fftwq_execute_split_dft_c2r type(C_PTR) function fftwq_plan_many_r2r(rank,n,howmany,in,inembed,istride,idist,out,onembed,ostride,odist,kind,flags) & bind(C, name='fftwq_plan_many_r2r') import integer(C_INT), value :: rank integer(C_INT), dimension(*), intent(in) :: n integer(C_INT), value :: howmany real(16), dimension(*), intent(out) :: in integer(C_INT), dimension(*), intent(in) :: inembed integer(C_INT), value :: istride integer(C_INT), value :: idist real(16), dimension(*), intent(out) :: out integer(C_INT), dimension(*), intent(in) :: onembed integer(C_INT), value :: ostride integer(C_INT), value :: odist integer(C_FFTW_R2R_KIND), dimension(*), intent(in) :: kind integer(C_INT), value :: flags end function fftwq_plan_many_r2r type(C_PTR) function fftwq_plan_r2r(rank,n,in,out,kind,flags) bind(C, name='fftwq_plan_r2r') import integer(C_INT), value :: rank integer(C_INT), dimension(*), intent(in) :: n real(16), dimension(*), intent(out) :: in real(16), dimension(*), intent(out) :: out integer(C_FFTW_R2R_KIND), dimension(*), intent(in) :: kind integer(C_INT), value :: flags end function fftwq_plan_r2r type(C_PTR) function fftwq_plan_r2r_1d(n,in,out,kind,flags) bind(C, name='fftwq_plan_r2r_1d') import integer(C_INT), value :: n real(16), dimension(*), intent(out) :: in real(16), dimension(*), intent(out) :: out integer(C_FFTW_R2R_KIND), value :: kind integer(C_INT), value :: flags end function fftwq_plan_r2r_1d type(C_PTR) function fftwq_plan_r2r_2d(n0,n1,in,out,kind0,kind1,flags) bind(C, name='fftwq_plan_r2r_2d') import integer(C_INT), value :: n0 integer(C_INT), value :: n1 real(16), dimension(*), intent(out) :: in real(16), dimension(*), intent(out) :: out integer(C_FFTW_R2R_KIND), value :: kind0 integer(C_FFTW_R2R_KIND), value :: kind1 integer(C_INT), value :: flags end function fftwq_plan_r2r_2d type(C_PTR) function fftwq_plan_r2r_3d(n0,n1,n2,in,out,kind0,kind1,kind2,flags) bind(C, name='fftwq_plan_r2r_3d') import integer(C_INT), value :: n0 integer(C_INT), value :: n1 integer(C_INT), value :: n2 real(16), dimension(*), intent(out) :: in real(16), dimension(*), intent(out) :: out integer(C_FFTW_R2R_KIND), value :: kind0 integer(C_FFTW_R2R_KIND), value :: kind1 integer(C_FFTW_R2R_KIND), value :: kind2 integer(C_INT), value :: flags end function fftwq_plan_r2r_3d type(C_PTR) function fftwq_plan_guru_r2r(rank,dims,howmany_rank,howmany_dims,in,out,kind,flags) & bind(C, name='fftwq_plan_guru_r2r') import integer(C_INT), value :: rank type(fftwq_iodim), dimension(*), intent(in) :: dims integer(C_INT), value :: howmany_rank type(fftwq_iodim), dimension(*), intent(in) :: howmany_dims real(16), dimension(*), intent(out) :: in real(16), dimension(*), intent(out) :: out integer(C_FFTW_R2R_KIND), dimension(*), intent(in) :: kind integer(C_INT), value :: flags end function fftwq_plan_guru_r2r type(C_PTR) function fftwq_plan_guru64_r2r(rank,dims,howmany_rank,howmany_dims,in,out,kind,flags) & bind(C, name='fftwq_plan_guru64_r2r') import integer(C_INT), value :: rank type(fftwq_iodim64), dimension(*), intent(in) :: dims integer(C_INT), value :: howmany_rank type(fftwq_iodim64), dimension(*), intent(in) :: howmany_dims real(16), dimension(*), intent(out) :: in real(16), dimension(*), intent(out) :: out integer(C_FFTW_R2R_KIND), dimension(*), intent(in) :: kind integer(C_INT), value :: flags end function fftwq_plan_guru64_r2r subroutine fftwq_execute_r2r(p,in,out) bind(C, name='fftwq_execute_r2r') import type(C_PTR), value :: p real(16), dimension(*), intent(inout) :: in real(16), dimension(*), intent(out) :: out end subroutine fftwq_execute_r2r subroutine fftwq_destroy_plan(p) bind(C, name='fftwq_destroy_plan') import type(C_PTR), value :: p end subroutine fftwq_destroy_plan subroutine fftwq_forget_wisdom() bind(C, name='fftwq_forget_wisdom') import end subroutine fftwq_forget_wisdom subroutine fftwq_cleanup() bind(C, name='fftwq_cleanup') import end subroutine fftwq_cleanup subroutine fftwq_set_timelimit(t) bind(C, name='fftwq_set_timelimit') import real(C_DOUBLE), value :: t end subroutine fftwq_set_timelimit subroutine fftwq_plan_with_nthreads(nthreads) bind(C, name='fftwq_plan_with_nthreads') import integer(C_INT), value :: nthreads end subroutine fftwq_plan_with_nthreads integer(C_INT) function fftwq_init_threads() bind(C, name='fftwq_init_threads') import end function fftwq_init_threads subroutine fftwq_cleanup_threads() bind(C, name='fftwq_cleanup_threads') import end subroutine fftwq_cleanup_threads integer(C_INT) function fftwq_export_wisdom_to_filename(filename) bind(C, name='fftwq_export_wisdom_to_filename') import character(C_CHAR), dimension(*), intent(in) :: filename end function fftwq_export_wisdom_to_filename subroutine fftwq_export_wisdom_to_file(output_file) bind(C, name='fftwq_export_wisdom_to_file') import type(C_PTR), value :: output_file end subroutine fftwq_export_wisdom_to_file type(C_PTR) function fftwq_export_wisdom_to_string() bind(C, name='fftwq_export_wisdom_to_string') import end function fftwq_export_wisdom_to_string subroutine fftwq_export_wisdom(write_char,data) bind(C, name='fftwq_export_wisdom') import type(C_FUNPTR), value :: write_char type(C_PTR), value :: data end subroutine fftwq_export_wisdom integer(C_INT) function fftwq_import_system_wisdom() bind(C, name='fftwq_import_system_wisdom') import end function fftwq_import_system_wisdom integer(C_INT) function fftwq_import_wisdom_from_filename(filename) bind(C, name='fftwq_import_wisdom_from_filename') import character(C_CHAR), dimension(*), intent(in) :: filename end function fftwq_import_wisdom_from_filename integer(C_INT) function fftwq_import_wisdom_from_file(input_file) bind(C, name='fftwq_import_wisdom_from_file') import type(C_PTR), value :: input_file end function fftwq_import_wisdom_from_file integer(C_INT) function fftwq_import_wisdom_from_string(input_string) bind(C, name='fftwq_import_wisdom_from_string') import character(C_CHAR), dimension(*), intent(in) :: input_string end function fftwq_import_wisdom_from_string integer(C_INT) function fftwq_import_wisdom(read_char,data) bind(C, name='fftwq_import_wisdom') import type(C_FUNPTR), value :: read_char type(C_PTR), value :: data end function fftwq_import_wisdom subroutine fftwq_fprint_plan(p,output_file) bind(C, name='fftwq_fprint_plan') import type(C_PTR), value :: p type(C_PTR), value :: output_file end subroutine fftwq_fprint_plan subroutine fftwq_print_plan(p) bind(C, name='fftwq_print_plan') import type(C_PTR), value :: p end subroutine fftwq_print_plan type(C_PTR) function fftwq_malloc(n) bind(C, name='fftwq_malloc') import integer(C_SIZE_T), value :: n end function fftwq_malloc ! Unable to generate Fortran interface for fftwq_alloc_real type(C_PTR) function fftwq_alloc_complex(n) bind(C, name='fftwq_alloc_complex') import integer(C_SIZE_T), value :: n end function fftwq_alloc_complex subroutine fftwq_free(p) bind(C, name='fftwq_free') import type(C_PTR), value :: p end subroutine fftwq_free subroutine fftwq_flops(p,add,mul,fmas) bind(C, name='fftwq_flops') import type(C_PTR), value :: p real(C_DOUBLE), intent(out) :: add real(C_DOUBLE), intent(out) :: mul real(C_DOUBLE), intent(out) :: fmas end subroutine fftwq_flops real(C_DOUBLE) function fftwq_estimate_cost(p) bind(C, name='fftwq_estimate_cost') import type(C_PTR), value :: p end function fftwq_estimate_cost real(C_DOUBLE) function fftwq_cost(p) bind(C, name='fftwq_cost') import type(C_PTR), value :: p end function fftwq_cost end interface fftw-3.3.3/api/Makefile.am0000644000175000017500000000565712054410050012220 00000000000000AM_CPPFLAGS = -I$(top_srcdir)/kernel -I$(top_srcdir)/dft \ -I$(top_srcdir)/rdft -I$(top_srcdir)/reodft AM_CFLAGS = $(STACK_ALIGN_CFLAGS) EXTRA_DIST = f03api.sh genf03.pl fftw3.f03.in include_HEADERS = fftw3.h fftw3.f fftw3l.f03 fftw3q.f03 nodist_include_HEADERS = fftw3.f03 noinst_LTLIBRARIES = libapi.la # pkgincludedir = $(includedir)/fftw3@PREC_SUFFIX@ # pkginclude_HEADERS = api.h x77.h guru.h guru64.h libapi_la_SOURCES = apiplan.c configure.c execute-dft-c2r.c \ execute-dft-r2c.c execute-dft.c execute-r2r.c execute-split-dft-c2r.c \ execute-split-dft-r2c.c execute-split-dft.c execute.c \ export-wisdom-to-file.c export-wisdom-to-string.c export-wisdom.c \ f77api.c flops.c forget-wisdom.c import-system-wisdom.c \ import-wisdom-from-file.c import-wisdom-from-string.c import-wisdom.c \ malloc.c map-r2r-kind.c mapflags.c mkprinter-file.c mktensor-iodims.c \ mktensor-rowmajor.c plan-dft-1d.c plan-dft-2d.c plan-dft-3d.c \ plan-dft-c2r-1d.c plan-dft-c2r-2d.c plan-dft-c2r-3d.c plan-dft-c2r.c \ plan-dft-r2c-1d.c plan-dft-r2c-2d.c plan-dft-r2c-3d.c plan-dft-r2c.c \ plan-dft.c plan-guru-dft-c2r.c plan-guru-dft-r2c.c plan-guru-dft.c \ plan-guru-r2r.c plan-guru-split-dft-c2r.c plan-guru-split-dft-r2c.c \ plan-guru-split-dft.c plan-many-dft-c2r.c plan-many-dft-r2c.c \ plan-many-dft.c plan-many-r2r.c plan-r2r-1d.c plan-r2r-2d.c \ plan-r2r-3d.c plan-r2r.c print-plan.c rdft2-pad.c the-planner.c \ version.c api.h f77funcs.h fftw3.h x77.h guru.h guru64.h \ mktensor-iodims.h plan-guru-dft-c2r.h plan-guru-dft-r2c.h \ plan-guru-dft.h plan-guru-r2r.h plan-guru-split-dft-c2r.h \ plan-guru-split-dft-r2c.h plan-guru-split-dft.h plan-guru64-dft-c2r.c \ plan-guru64-dft-r2c.c plan-guru64-dft.c plan-guru64-r2r.c \ plan-guru64-split-dft-c2r.c plan-guru64-split-dft-r2c.c \ plan-guru64-split-dft.c mktensor-iodims64.c BUILT_SOURCES = fftw3.f fftw3.f03.in fftw3.f03 fftw3l.f03 fftw3q.f03 CLEANFILES = fftw3.f03 fftw3.f03: fftw3.f03.in (echo "! Generated automatically. DO NOT EDIT!"; echo; \ echo " integer, parameter :: C_FFTW_R2R_KIND = @C_FFTW_R2R_KIND@"; \ grep -v "Generated automatically" $(srcdir)/fftw3.f03.in) > $@ if MAINTAINER_MODE # convert constants to F77 PARAMETER statements fftw3.f: fftw3.h rm -f $@ perl -pe 's/([A-Z0-9_]+)=([+-]?[0-9]+)/\n INTEGER \1\n PARAMETER (\1=\2)\n/g' $< |egrep 'PARAMETER|INTEGER' > $@ perl -pe 's/#define +([A-Z0-9_]+) +\(([+-]?[0-9]+)U?\)/\n INTEGER \1\n PARAMETER (\1=\2)\n/g' $< |egrep 'PARAMETER|INTEGER' >> $@ perl -pe 'if (/#define +([A-Z0-9_]+) +\(([0-9]+)U? *<< *([0-9]+)\)/) { print "\n INTEGER $$1\n PARAMETER ($$1=",$$2 << $$3,")\n"; }' $< |egrep 'PARAMETER|INTEGER' >> $@ fftw3.f03.in: fftw3.h f03api.sh genf03.pl sh $(srcdir)/f03api.sh d f > $@ fftw3l.f03: fftw3.h f03api.sh genf03.pl sh $(srcdir)/f03api.sh l | grep -v parameter > $@ fftw3q.f03: fftw3.h f03api.sh genf03.pl sh $(srcdir)/f03api.sh q | grep -v parameter > $@ endif # MAINTAINER_MODE fftw-3.3.3/api/Makefile.in0000644000175000017500000007060712054410177012240 00000000000000# Makefile.in generated by automake 1.11.6 from Makefile.am. # @configure_input@ # Copyright (C) 1994, 1995, 1996, 1997, 1998, 1999, 2000, 2001, 2002, # 2003, 2004, 2005, 2006, 2007, 2008, 2009, 2010, 2011 Free Software # Foundation, Inc. # This Makefile.in is free software; the Free Software Foundation # gives unlimited permission to copy and/or distribute it, # with or without modifications, as long as this notice is preserved. # This program is distributed in the hope that it will be useful, # but WITHOUT ANY WARRANTY, to the extent permitted by law; without # even the implied warranty of MERCHANTABILITY or FITNESS FOR A # PARTICULAR PURPOSE. @SET_MAKE@ VPATH = @srcdir@ am__make_dryrun = \ { \ am__dry=no; \ case $$MAKEFLAGS in \ *\\[\ \ ]*) \ echo 'am--echo: ; @echo "AM" OK' | $(MAKE) -f - 2>/dev/null \ | grep '^AM OK$$' >/dev/null || am__dry=yes;; \ *) \ for am__flg in $$MAKEFLAGS; do \ case $$am__flg in \ *=*|--*) ;; \ *n*) am__dry=yes; break;; \ esac; \ done;; \ esac; \ test $$am__dry = yes; \ } pkgdatadir = $(datadir)/@PACKAGE@ pkgincludedir = $(includedir)/@PACKAGE@ pkglibdir = $(libdir)/@PACKAGE@ pkglibexecdir = $(libexecdir)/@PACKAGE@ am__cd = CDPATH="$${ZSH_VERSION+.}$(PATH_SEPARATOR)" && cd install_sh_DATA = $(install_sh) -c -m 644 install_sh_PROGRAM = $(install_sh) -c install_sh_SCRIPT = $(install_sh) -c INSTALL_HEADER = $(INSTALL_DATA) transform = $(program_transform_name) NORMAL_INSTALL = : PRE_INSTALL = : POST_INSTALL = : NORMAL_UNINSTALL = : PRE_UNINSTALL = : POST_UNINSTALL = : build_triplet = @build@ host_triplet = @host@ subdir = api DIST_COMMON = $(include_HEADERS) $(srcdir)/Makefile.am \ $(srcdir)/Makefile.in ACLOCAL_M4 = $(top_srcdir)/aclocal.m4 am__aclocal_m4_deps = $(top_srcdir)/m4/acx_mpi.m4 \ $(top_srcdir)/m4/acx_pthread.m4 \ $(top_srcdir)/m4/ax_cc_maxopt.m4 \ $(top_srcdir)/m4/ax_check_compiler_flags.m4 \ $(top_srcdir)/m4/ax_compiler_vendor.m4 \ $(top_srcdir)/m4/ax_gcc_aligns_stack.m4 \ $(top_srcdir)/m4/ax_gcc_version.m4 \ $(top_srcdir)/m4/ax_openmp.m4 $(top_srcdir)/m4/libtool.m4 \ $(top_srcdir)/m4/ltoptions.m4 $(top_srcdir)/m4/ltsugar.m4 \ $(top_srcdir)/m4/ltversion.m4 $(top_srcdir)/m4/lt~obsolete.m4 \ $(top_srcdir)/configure.ac am__configure_deps = $(am__aclocal_m4_deps) $(CONFIGURE_DEPENDENCIES) \ $(ACLOCAL_M4) mkinstalldirs = $(install_sh) -d CONFIG_HEADER = $(top_builddir)/config.h CONFIG_CLEAN_FILES = CONFIG_CLEAN_VPATH_FILES = LTLIBRARIES = $(noinst_LTLIBRARIES) libapi_la_LIBADD = am_libapi_la_OBJECTS = apiplan.lo configure.lo execute-dft-c2r.lo \ execute-dft-r2c.lo execute-dft.lo execute-r2r.lo \ execute-split-dft-c2r.lo execute-split-dft-r2c.lo \ execute-split-dft.lo execute.lo export-wisdom-to-file.lo \ export-wisdom-to-string.lo export-wisdom.lo f77api.lo flops.lo \ forget-wisdom.lo import-system-wisdom.lo \ import-wisdom-from-file.lo import-wisdom-from-string.lo \ import-wisdom.lo malloc.lo map-r2r-kind.lo mapflags.lo \ mkprinter-file.lo mktensor-iodims.lo mktensor-rowmajor.lo \ plan-dft-1d.lo plan-dft-2d.lo plan-dft-3d.lo \ plan-dft-c2r-1d.lo plan-dft-c2r-2d.lo plan-dft-c2r-3d.lo \ plan-dft-c2r.lo plan-dft-r2c-1d.lo plan-dft-r2c-2d.lo \ plan-dft-r2c-3d.lo plan-dft-r2c.lo plan-dft.lo \ plan-guru-dft-c2r.lo plan-guru-dft-r2c.lo plan-guru-dft.lo \ plan-guru-r2r.lo plan-guru-split-dft-c2r.lo \ plan-guru-split-dft-r2c.lo plan-guru-split-dft.lo \ plan-many-dft-c2r.lo plan-many-dft-r2c.lo plan-many-dft.lo \ plan-many-r2r.lo plan-r2r-1d.lo plan-r2r-2d.lo plan-r2r-3d.lo \ plan-r2r.lo print-plan.lo rdft2-pad.lo the-planner.lo \ version.lo plan-guru64-dft-c2r.lo plan-guru64-dft-r2c.lo \ plan-guru64-dft.lo plan-guru64-r2r.lo \ plan-guru64-split-dft-c2r.lo plan-guru64-split-dft-r2c.lo \ plan-guru64-split-dft.lo mktensor-iodims64.lo libapi_la_OBJECTS = $(am_libapi_la_OBJECTS) DEFAULT_INCLUDES = -I.@am__isrc@ -I$(top_builddir) depcomp = $(SHELL) $(top_srcdir)/depcomp am__depfiles_maybe = depfiles am__mv = mv -f COMPILE = $(CC) $(DEFS) $(DEFAULT_INCLUDES) $(INCLUDES) $(AM_CPPFLAGS) \ $(CPPFLAGS) $(AM_CFLAGS) $(CFLAGS) LTCOMPILE = $(LIBTOOL) --tag=CC $(AM_LIBTOOLFLAGS) $(LIBTOOLFLAGS) \ --mode=compile $(CC) $(DEFS) $(DEFAULT_INCLUDES) $(INCLUDES) \ $(AM_CPPFLAGS) $(CPPFLAGS) $(AM_CFLAGS) $(CFLAGS) CCLD = $(CC) LINK = $(LIBTOOL) --tag=CC $(AM_LIBTOOLFLAGS) $(LIBTOOLFLAGS) \ --mode=link $(CCLD) $(AM_CFLAGS) $(CFLAGS) $(AM_LDFLAGS) \ $(LDFLAGS) -o $@ SOURCES = $(libapi_la_SOURCES) DIST_SOURCES = $(libapi_la_SOURCES) am__can_run_installinfo = \ case $$AM_UPDATE_INFO_DIR in \ n|no|NO) false;; \ *) (install-info --version) >/dev/null 2>&1;; \ esac am__vpath_adj_setup = srcdirstrip=`echo "$(srcdir)" | sed 's|.|.|g'`; am__vpath_adj = case $$p in \ $(srcdir)/*) f=`echo "$$p" | sed "s|^$$srcdirstrip/||"`;; \ *) f=$$p;; \ esac; am__strip_dir = f=`echo $$p | sed -e 's|^.*/||'`; am__install_max = 40 am__nobase_strip_setup = \ srcdirstrip=`echo "$(srcdir)" | sed 's/[].[^$$\\*|]/\\\\&/g'` am__nobase_strip = \ for p in $$list; do echo "$$p"; done | sed -e "s|$$srcdirstrip/||" am__nobase_list = $(am__nobase_strip_setup); \ for p in $$list; do echo "$$p $$p"; done | \ sed "s| $$srcdirstrip/| |;"' / .*\//!s/ .*/ ./; s,\( .*\)/[^/]*$$,\1,' | \ $(AWK) 'BEGIN { files["."] = "" } { files[$$2] = files[$$2] " " $$1; \ if (++n[$$2] == $(am__install_max)) \ { print $$2, files[$$2]; n[$$2] = 0; files[$$2] = "" } } \ END { for (dir in files) print dir, files[dir] }' am__base_list = \ sed '$$!N;$$!N;$$!N;$$!N;$$!N;$$!N;$$!N;s/\n/ /g' | \ sed '$$!N;$$!N;$$!N;$$!N;s/\n/ /g' am__uninstall_files_from_dir = { \ test -z "$$files" \ || { test ! -d "$$dir" && test ! -f "$$dir" && test ! -r "$$dir"; } \ || { echo " ( cd '$$dir' && rm -f" $$files ")"; \ $(am__cd) "$$dir" && rm -f $$files; }; \ } am__installdirs = "$(DESTDIR)$(includedir)" "$(DESTDIR)$(includedir)" HEADERS = $(include_HEADERS) $(nodist_include_HEADERS) ETAGS = etags CTAGS = ctags DISTFILES = $(DIST_COMMON) $(DIST_SOURCES) $(TEXINFOS) $(EXTRA_DIST) ACLOCAL = @ACLOCAL@ ALLOCA = @ALLOCA@ ALTIVEC_CFLAGS = @ALTIVEC_CFLAGS@ AMTAR = @AMTAR@ AR = @AR@ AS = @AS@ AUTOCONF = @AUTOCONF@ AUTOHEADER = @AUTOHEADER@ AUTOMAKE = @AUTOMAKE@ AVX_CFLAGS = @AVX_CFLAGS@ AWK = @AWK@ CC = @CC@ CCDEPMODE = @CCDEPMODE@ CFLAGS = @CFLAGS@ CHECK_PL_OPTS = @CHECK_PL_OPTS@ CPP = @CPP@ CPPFLAGS = @CPPFLAGS@ CYGPATH_W = @CYGPATH_W@ C_FFTW_R2R_KIND = @C_FFTW_R2R_KIND@ C_MPI_FINT = @C_MPI_FINT@ DEFS = @DEFS@ DEPDIR = @DEPDIR@ DLLTOOL = @DLLTOOL@ DSYMUTIL = @DSYMUTIL@ DUMPBIN = @DUMPBIN@ ECHO_C = @ECHO_C@ ECHO_N = @ECHO_N@ ECHO_T = @ECHO_T@ EGREP = @EGREP@ EXEEXT = @EXEEXT@ F77 = @F77@ FFLAGS = @FFLAGS@ FGREP = @FGREP@ FLIBS = @FLIBS@ GREP = @GREP@ INSTALL = @INSTALL@ INSTALL_DATA = @INSTALL_DATA@ INSTALL_PROGRAM = @INSTALL_PROGRAM@ INSTALL_SCRIPT = @INSTALL_SCRIPT@ INSTALL_STRIP_PROGRAM = @INSTALL_STRIP_PROGRAM@ LD = @LD@ LDFLAGS = @LDFLAGS@ LIBOBJS = @LIBOBJS@ LIBQUADMATH = @LIBQUADMATH@ LIBS = @LIBS@ LIBTOOL = @LIBTOOL@ LIPO = @LIPO@ LN_S = @LN_S@ LTLIBOBJS = @LTLIBOBJS@ MAINT = @MAINT@ MAKEINFO = @MAKEINFO@ MANIFEST_TOOL = @MANIFEST_TOOL@ MKDIR_P = @MKDIR_P@ MPICC = @MPICC@ MPILIBS = @MPILIBS@ MPIRUN = @MPIRUN@ NEON_CFLAGS = @NEON_CFLAGS@ NM = @NM@ NMEDIT = @NMEDIT@ OBJDUMP = @OBJDUMP@ OBJEXT = @OBJEXT@ OCAMLBUILD = @OCAMLBUILD@ OPENMP_CFLAGS = @OPENMP_CFLAGS@ OTOOL = @OTOOL@ OTOOL64 = @OTOOL64@ PACKAGE = @PACKAGE@ PACKAGE_BUGREPORT = @PACKAGE_BUGREPORT@ PACKAGE_NAME = @PACKAGE_NAME@ PACKAGE_STRING = @PACKAGE_STRING@ PACKAGE_TARNAME = @PACKAGE_TARNAME@ PACKAGE_URL = @PACKAGE_URL@ PACKAGE_VERSION = @PACKAGE_VERSION@ PATH_SEPARATOR = @PATH_SEPARATOR@ POW_LIB = @POW_LIB@ PRECISION = @PRECISION@ PREC_SUFFIX = @PREC_SUFFIX@ PTHREAD_CC = @PTHREAD_CC@ PTHREAD_CFLAGS = @PTHREAD_CFLAGS@ PTHREAD_LIBS = @PTHREAD_LIBS@ RANLIB = @RANLIB@ SED = @SED@ SET_MAKE = @SET_MAKE@ SHARED_VERSION_INFO = @SHARED_VERSION_INFO@ SHELL = @SHELL@ SSE2_CFLAGS = @SSE2_CFLAGS@ STACK_ALIGN_CFLAGS = @STACK_ALIGN_CFLAGS@ STRIP = @STRIP@ THREADLIBS = @THREADLIBS@ VERSION = @VERSION@ abs_builddir = @abs_builddir@ abs_srcdir = @abs_srcdir@ abs_top_builddir = @abs_top_builddir@ abs_top_srcdir = @abs_top_srcdir@ ac_ct_AR = @ac_ct_AR@ ac_ct_CC = @ac_ct_CC@ ac_ct_DUMPBIN = @ac_ct_DUMPBIN@ ac_ct_F77 = @ac_ct_F77@ acx_pthread_config = @acx_pthread_config@ am__include = @am__include@ am__leading_dot = @am__leading_dot@ am__quote = @am__quote@ am__tar = @am__tar@ am__untar = @am__untar@ bindir = @bindir@ build = @build@ build_alias = @build_alias@ build_cpu = @build_cpu@ build_os = @build_os@ build_vendor = @build_vendor@ builddir = @builddir@ datadir = @datadir@ datarootdir = @datarootdir@ docdir = @docdir@ dvidir = @dvidir@ exec_prefix = @exec_prefix@ host = @host@ host_alias = @host_alias@ host_cpu = @host_cpu@ host_os = @host_os@ host_vendor = @host_vendor@ htmldir = @htmldir@ includedir = @includedir@ infodir = @infodir@ install_sh = @install_sh@ libdir = @libdir@ libexecdir = @libexecdir@ localedir = @localedir@ localstatedir = @localstatedir@ mandir = @mandir@ mkdir_p = @mkdir_p@ oldincludedir = @oldincludedir@ pdfdir = @pdfdir@ prefix = @prefix@ program_transform_name = @program_transform_name@ psdir = @psdir@ sbindir = @sbindir@ sharedstatedir = @sharedstatedir@ srcdir = @srcdir@ sysconfdir = @sysconfdir@ target_alias = @target_alias@ top_build_prefix = @top_build_prefix@ top_builddir = @top_builddir@ top_srcdir = @top_srcdir@ AM_CPPFLAGS = -I$(top_srcdir)/kernel -I$(top_srcdir)/dft \ -I$(top_srcdir)/rdft -I$(top_srcdir)/reodft AM_CFLAGS = $(STACK_ALIGN_CFLAGS) EXTRA_DIST = f03api.sh genf03.pl fftw3.f03.in include_HEADERS = fftw3.h fftw3.f fftw3l.f03 fftw3q.f03 nodist_include_HEADERS = fftw3.f03 noinst_LTLIBRARIES = libapi.la # pkgincludedir = $(includedir)/fftw3@PREC_SUFFIX@ # pkginclude_HEADERS = api.h x77.h guru.h guru64.h libapi_la_SOURCES = apiplan.c configure.c execute-dft-c2r.c \ execute-dft-r2c.c execute-dft.c execute-r2r.c execute-split-dft-c2r.c \ execute-split-dft-r2c.c execute-split-dft.c execute.c \ export-wisdom-to-file.c export-wisdom-to-string.c export-wisdom.c \ f77api.c flops.c forget-wisdom.c import-system-wisdom.c \ import-wisdom-from-file.c import-wisdom-from-string.c import-wisdom.c \ malloc.c map-r2r-kind.c mapflags.c mkprinter-file.c mktensor-iodims.c \ mktensor-rowmajor.c plan-dft-1d.c plan-dft-2d.c plan-dft-3d.c \ plan-dft-c2r-1d.c plan-dft-c2r-2d.c plan-dft-c2r-3d.c plan-dft-c2r.c \ plan-dft-r2c-1d.c plan-dft-r2c-2d.c plan-dft-r2c-3d.c plan-dft-r2c.c \ plan-dft.c plan-guru-dft-c2r.c plan-guru-dft-r2c.c plan-guru-dft.c \ plan-guru-r2r.c plan-guru-split-dft-c2r.c plan-guru-split-dft-r2c.c \ plan-guru-split-dft.c plan-many-dft-c2r.c plan-many-dft-r2c.c \ plan-many-dft.c plan-many-r2r.c plan-r2r-1d.c plan-r2r-2d.c \ plan-r2r-3d.c plan-r2r.c print-plan.c rdft2-pad.c the-planner.c \ version.c api.h f77funcs.h fftw3.h x77.h guru.h guru64.h \ mktensor-iodims.h plan-guru-dft-c2r.h plan-guru-dft-r2c.h \ plan-guru-dft.h plan-guru-r2r.h plan-guru-split-dft-c2r.h \ plan-guru-split-dft-r2c.h plan-guru-split-dft.h plan-guru64-dft-c2r.c \ plan-guru64-dft-r2c.c plan-guru64-dft.c plan-guru64-r2r.c \ plan-guru64-split-dft-c2r.c plan-guru64-split-dft-r2c.c \ plan-guru64-split-dft.c mktensor-iodims64.c BUILT_SOURCES = fftw3.f fftw3.f03.in fftw3.f03 fftw3l.f03 fftw3q.f03 CLEANFILES = fftw3.f03 all: $(BUILT_SOURCES) $(MAKE) $(AM_MAKEFLAGS) all-am .SUFFIXES: .SUFFIXES: .c .lo .o .obj $(srcdir)/Makefile.in: @MAINTAINER_MODE_TRUE@ $(srcdir)/Makefile.am $(am__configure_deps) @for dep in $?; do \ case '$(am__configure_deps)' in \ *$$dep*) \ ( cd $(top_builddir) && $(MAKE) $(AM_MAKEFLAGS) am--refresh ) \ && { if test -f $@; then exit 0; else break; fi; }; \ exit 1;; \ esac; \ done; \ echo ' cd $(top_srcdir) && $(AUTOMAKE) --gnu api/Makefile'; \ $(am__cd) $(top_srcdir) && \ $(AUTOMAKE) --gnu api/Makefile .PRECIOUS: Makefile Makefile: $(srcdir)/Makefile.in $(top_builddir)/config.status @case '$?' in \ *config.status*) \ cd $(top_builddir) && $(MAKE) $(AM_MAKEFLAGS) am--refresh;; \ *) \ echo ' cd $(top_builddir) && $(SHELL) ./config.status $(subdir)/$@ $(am__depfiles_maybe)'; \ cd $(top_builddir) && $(SHELL) ./config.status $(subdir)/$@ $(am__depfiles_maybe);; \ esac; $(top_builddir)/config.status: $(top_srcdir)/configure $(CONFIG_STATUS_DEPENDENCIES) cd $(top_builddir) && $(MAKE) $(AM_MAKEFLAGS) am--refresh $(top_srcdir)/configure: @MAINTAINER_MODE_TRUE@ $(am__configure_deps) cd $(top_builddir) && $(MAKE) $(AM_MAKEFLAGS) am--refresh $(ACLOCAL_M4): @MAINTAINER_MODE_TRUE@ $(am__aclocal_m4_deps) cd $(top_builddir) && $(MAKE) $(AM_MAKEFLAGS) am--refresh $(am__aclocal_m4_deps): clean-noinstLTLIBRARIES: -test -z "$(noinst_LTLIBRARIES)" || rm -f $(noinst_LTLIBRARIES) @list='$(noinst_LTLIBRARIES)'; for p in $$list; do \ dir="`echo $$p | sed -e 's|/[^/]*$$||'`"; \ test "$$dir" != "$$p" || dir=.; \ echo "rm -f \"$${dir}/so_locations\""; \ rm -f "$${dir}/so_locations"; \ done libapi.la: $(libapi_la_OBJECTS) $(libapi_la_DEPENDENCIES) $(EXTRA_libapi_la_DEPENDENCIES) $(LINK) $(libapi_la_OBJECTS) $(libapi_la_LIBADD) $(LIBS) mostlyclean-compile: -rm -f *.$(OBJEXT) distclean-compile: -rm -f *.tab.c @AMDEP_TRUE@@am__include@ @am__quote@./$(DEPDIR)/apiplan.Plo@am__quote@ @AMDEP_TRUE@@am__include@ @am__quote@./$(DEPDIR)/configure.Plo@am__quote@ @AMDEP_TRUE@@am__include@ @am__quote@./$(DEPDIR)/execute-dft-c2r.Plo@am__quote@ @AMDEP_TRUE@@am__include@ @am__quote@./$(DEPDIR)/execute-dft-r2c.Plo@am__quote@ @AMDEP_TRUE@@am__include@ @am__quote@./$(DEPDIR)/execute-dft.Plo@am__quote@ @AMDEP_TRUE@@am__include@ @am__quote@./$(DEPDIR)/execute-r2r.Plo@am__quote@ @AMDEP_TRUE@@am__include@ @am__quote@./$(DEPDIR)/execute-split-dft-c2r.Plo@am__quote@ @AMDEP_TRUE@@am__include@ @am__quote@./$(DEPDIR)/execute-split-dft-r2c.Plo@am__quote@ @AMDEP_TRUE@@am__include@ @am__quote@./$(DEPDIR)/execute-split-dft.Plo@am__quote@ @AMDEP_TRUE@@am__include@ @am__quote@./$(DEPDIR)/execute.Plo@am__quote@ @AMDEP_TRUE@@am__include@ @am__quote@./$(DEPDIR)/export-wisdom-to-file.Plo@am__quote@ @AMDEP_TRUE@@am__include@ @am__quote@./$(DEPDIR)/export-wisdom-to-string.Plo@am__quote@ @AMDEP_TRUE@@am__include@ @am__quote@./$(DEPDIR)/export-wisdom.Plo@am__quote@ @AMDEP_TRUE@@am__include@ @am__quote@./$(DEPDIR)/f77api.Plo@am__quote@ @AMDEP_TRUE@@am__include@ @am__quote@./$(DEPDIR)/flops.Plo@am__quote@ @AMDEP_TRUE@@am__include@ @am__quote@./$(DEPDIR)/forget-wisdom.Plo@am__quote@ @AMDEP_TRUE@@am__include@ @am__quote@./$(DEPDIR)/import-system-wisdom.Plo@am__quote@ @AMDEP_TRUE@@am__include@ @am__quote@./$(DEPDIR)/import-wisdom-from-file.Plo@am__quote@ @AMDEP_TRUE@@am__include@ @am__quote@./$(DEPDIR)/import-wisdom-from-string.Plo@am__quote@ @AMDEP_TRUE@@am__include@ @am__quote@./$(DEPDIR)/import-wisdom.Plo@am__quote@ @AMDEP_TRUE@@am__include@ @am__quote@./$(DEPDIR)/malloc.Plo@am__quote@ @AMDEP_TRUE@@am__include@ @am__quote@./$(DEPDIR)/map-r2r-kind.Plo@am__quote@ @AMDEP_TRUE@@am__include@ @am__quote@./$(DEPDIR)/mapflags.Plo@am__quote@ @AMDEP_TRUE@@am__include@ @am__quote@./$(DEPDIR)/mkprinter-file.Plo@am__quote@ @AMDEP_TRUE@@am__include@ @am__quote@./$(DEPDIR)/mktensor-iodims.Plo@am__quote@ @AMDEP_TRUE@@am__include@ @am__quote@./$(DEPDIR)/mktensor-iodims64.Plo@am__quote@ @AMDEP_TRUE@@am__include@ @am__quote@./$(DEPDIR)/mktensor-rowmajor.Plo@am__quote@ @AMDEP_TRUE@@am__include@ @am__quote@./$(DEPDIR)/plan-dft-1d.Plo@am__quote@ @AMDEP_TRUE@@am__include@ @am__quote@./$(DEPDIR)/plan-dft-2d.Plo@am__quote@ @AMDEP_TRUE@@am__include@ @am__quote@./$(DEPDIR)/plan-dft-3d.Plo@am__quote@ @AMDEP_TRUE@@am__include@ @am__quote@./$(DEPDIR)/plan-dft-c2r-1d.Plo@am__quote@ @AMDEP_TRUE@@am__include@ @am__quote@./$(DEPDIR)/plan-dft-c2r-2d.Plo@am__quote@ @AMDEP_TRUE@@am__include@ @am__quote@./$(DEPDIR)/plan-dft-c2r-3d.Plo@am__quote@ @AMDEP_TRUE@@am__include@ @am__quote@./$(DEPDIR)/plan-dft-c2r.Plo@am__quote@ @AMDEP_TRUE@@am__include@ @am__quote@./$(DEPDIR)/plan-dft-r2c-1d.Plo@am__quote@ @AMDEP_TRUE@@am__include@ @am__quote@./$(DEPDIR)/plan-dft-r2c-2d.Plo@am__quote@ @AMDEP_TRUE@@am__include@ @am__quote@./$(DEPDIR)/plan-dft-r2c-3d.Plo@am__quote@ @AMDEP_TRUE@@am__include@ @am__quote@./$(DEPDIR)/plan-dft-r2c.Plo@am__quote@ @AMDEP_TRUE@@am__include@ @am__quote@./$(DEPDIR)/plan-dft.Plo@am__quote@ @AMDEP_TRUE@@am__include@ @am__quote@./$(DEPDIR)/plan-guru-dft-c2r.Plo@am__quote@ @AMDEP_TRUE@@am__include@ @am__quote@./$(DEPDIR)/plan-guru-dft-r2c.Plo@am__quote@ @AMDEP_TRUE@@am__include@ @am__quote@./$(DEPDIR)/plan-guru-dft.Plo@am__quote@ @AMDEP_TRUE@@am__include@ @am__quote@./$(DEPDIR)/plan-guru-r2r.Plo@am__quote@ @AMDEP_TRUE@@am__include@ @am__quote@./$(DEPDIR)/plan-guru-split-dft-c2r.Plo@am__quote@ @AMDEP_TRUE@@am__include@ @am__quote@./$(DEPDIR)/plan-guru-split-dft-r2c.Plo@am__quote@ @AMDEP_TRUE@@am__include@ @am__quote@./$(DEPDIR)/plan-guru-split-dft.Plo@am__quote@ @AMDEP_TRUE@@am__include@ @am__quote@./$(DEPDIR)/plan-guru64-dft-c2r.Plo@am__quote@ @AMDEP_TRUE@@am__include@ @am__quote@./$(DEPDIR)/plan-guru64-dft-r2c.Plo@am__quote@ @AMDEP_TRUE@@am__include@ @am__quote@./$(DEPDIR)/plan-guru64-dft.Plo@am__quote@ @AMDEP_TRUE@@am__include@ @am__quote@./$(DEPDIR)/plan-guru64-r2r.Plo@am__quote@ @AMDEP_TRUE@@am__include@ @am__quote@./$(DEPDIR)/plan-guru64-split-dft-c2r.Plo@am__quote@ @AMDEP_TRUE@@am__include@ @am__quote@./$(DEPDIR)/plan-guru64-split-dft-r2c.Plo@am__quote@ @AMDEP_TRUE@@am__include@ @am__quote@./$(DEPDIR)/plan-guru64-split-dft.Plo@am__quote@ @AMDEP_TRUE@@am__include@ @am__quote@./$(DEPDIR)/plan-many-dft-c2r.Plo@am__quote@ @AMDEP_TRUE@@am__include@ @am__quote@./$(DEPDIR)/plan-many-dft-r2c.Plo@am__quote@ @AMDEP_TRUE@@am__include@ @am__quote@./$(DEPDIR)/plan-many-dft.Plo@am__quote@ @AMDEP_TRUE@@am__include@ @am__quote@./$(DEPDIR)/plan-many-r2r.Plo@am__quote@ @AMDEP_TRUE@@am__include@ @am__quote@./$(DEPDIR)/plan-r2r-1d.Plo@am__quote@ @AMDEP_TRUE@@am__include@ @am__quote@./$(DEPDIR)/plan-r2r-2d.Plo@am__quote@ @AMDEP_TRUE@@am__include@ @am__quote@./$(DEPDIR)/plan-r2r-3d.Plo@am__quote@ @AMDEP_TRUE@@am__include@ @am__quote@./$(DEPDIR)/plan-r2r.Plo@am__quote@ @AMDEP_TRUE@@am__include@ @am__quote@./$(DEPDIR)/print-plan.Plo@am__quote@ @AMDEP_TRUE@@am__include@ @am__quote@./$(DEPDIR)/rdft2-pad.Plo@am__quote@ @AMDEP_TRUE@@am__include@ @am__quote@./$(DEPDIR)/the-planner.Plo@am__quote@ @AMDEP_TRUE@@am__include@ @am__quote@./$(DEPDIR)/version.Plo@am__quote@ .c.o: @am__fastdepCC_TRUE@ $(COMPILE) -MT $@ -MD -MP -MF $(DEPDIR)/$*.Tpo -c -o $@ $< @am__fastdepCC_TRUE@ $(am__mv) $(DEPDIR)/$*.Tpo $(DEPDIR)/$*.Po @AMDEP_TRUE@@am__fastdepCC_FALSE@ source='$<' object='$@' libtool=no @AMDEPBACKSLASH@ @AMDEP_TRUE@@am__fastdepCC_FALSE@ DEPDIR=$(DEPDIR) $(CCDEPMODE) $(depcomp) @AMDEPBACKSLASH@ @am__fastdepCC_FALSE@ $(COMPILE) -c $< .c.obj: @am__fastdepCC_TRUE@ $(COMPILE) -MT $@ -MD -MP -MF $(DEPDIR)/$*.Tpo -c -o $@ `$(CYGPATH_W) '$<'` @am__fastdepCC_TRUE@ $(am__mv) $(DEPDIR)/$*.Tpo $(DEPDIR)/$*.Po @AMDEP_TRUE@@am__fastdepCC_FALSE@ source='$<' object='$@' libtool=no @AMDEPBACKSLASH@ @AMDEP_TRUE@@am__fastdepCC_FALSE@ DEPDIR=$(DEPDIR) $(CCDEPMODE) $(depcomp) @AMDEPBACKSLASH@ @am__fastdepCC_FALSE@ $(COMPILE) -c `$(CYGPATH_W) '$<'` .c.lo: @am__fastdepCC_TRUE@ $(LTCOMPILE) -MT $@ -MD -MP -MF $(DEPDIR)/$*.Tpo -c -o $@ $< @am__fastdepCC_TRUE@ $(am__mv) $(DEPDIR)/$*.Tpo $(DEPDIR)/$*.Plo @AMDEP_TRUE@@am__fastdepCC_FALSE@ source='$<' object='$@' libtool=yes @AMDEPBACKSLASH@ @AMDEP_TRUE@@am__fastdepCC_FALSE@ DEPDIR=$(DEPDIR) $(CCDEPMODE) $(depcomp) @AMDEPBACKSLASH@ @am__fastdepCC_FALSE@ $(LTCOMPILE) -c -o $@ $< mostlyclean-libtool: -rm -f *.lo clean-libtool: -rm -rf .libs _libs install-includeHEADERS: $(include_HEADERS) @$(NORMAL_INSTALL) @list='$(include_HEADERS)'; test -n "$(includedir)" || list=; \ if test -n "$$list"; then \ echo " $(MKDIR_P) '$(DESTDIR)$(includedir)'"; \ $(MKDIR_P) "$(DESTDIR)$(includedir)" || exit 1; \ fi; \ for p in $$list; do \ if test -f "$$p"; then d=; else d="$(srcdir)/"; fi; \ echo "$$d$$p"; \ done | $(am__base_list) | \ while read files; do \ echo " $(INSTALL_HEADER) $$files '$(DESTDIR)$(includedir)'"; \ $(INSTALL_HEADER) $$files "$(DESTDIR)$(includedir)" || exit $$?; \ done uninstall-includeHEADERS: @$(NORMAL_UNINSTALL) @list='$(include_HEADERS)'; test -n "$(includedir)" || list=; \ files=`for p in $$list; do echo $$p; done | sed -e 's|^.*/||'`; \ dir='$(DESTDIR)$(includedir)'; $(am__uninstall_files_from_dir) install-nodist_includeHEADERS: $(nodist_include_HEADERS) @$(NORMAL_INSTALL) @list='$(nodist_include_HEADERS)'; test -n "$(includedir)" || list=; \ if test -n "$$list"; then \ echo " $(MKDIR_P) '$(DESTDIR)$(includedir)'"; \ $(MKDIR_P) "$(DESTDIR)$(includedir)" || exit 1; \ fi; \ for p in $$list; do \ if test -f "$$p"; then d=; else d="$(srcdir)/"; fi; \ echo "$$d$$p"; \ done | $(am__base_list) | \ while read files; do \ echo " $(INSTALL_HEADER) $$files '$(DESTDIR)$(includedir)'"; \ $(INSTALL_HEADER) $$files "$(DESTDIR)$(includedir)" || exit $$?; \ done uninstall-nodist_includeHEADERS: @$(NORMAL_UNINSTALL) @list='$(nodist_include_HEADERS)'; test -n "$(includedir)" || list=; \ files=`for p in $$list; do echo $$p; done | sed -e 's|^.*/||'`; \ dir='$(DESTDIR)$(includedir)'; $(am__uninstall_files_from_dir) ID: $(HEADERS) $(SOURCES) $(LISP) $(TAGS_FILES) list='$(SOURCES) $(HEADERS) $(LISP) $(TAGS_FILES)'; \ unique=`for i in $$list; do \ if test -f "$$i"; then echo $$i; else echo $(srcdir)/$$i; fi; \ done | \ $(AWK) '{ files[$$0] = 1; nonempty = 1; } \ END { if (nonempty) { for (i in files) print i; }; }'`; \ mkid -fID $$unique tags: TAGS TAGS: $(HEADERS) $(SOURCES) $(TAGS_DEPENDENCIES) \ $(TAGS_FILES) $(LISP) set x; \ here=`pwd`; \ list='$(SOURCES) $(HEADERS) $(LISP) $(TAGS_FILES)'; \ unique=`for i in $$list; do \ if test -f "$$i"; then echo $$i; else echo $(srcdir)/$$i; fi; \ done | \ $(AWK) '{ files[$$0] = 1; nonempty = 1; } \ END { if (nonempty) { for (i in files) print i; }; }'`; \ shift; \ if test -z "$(ETAGS_ARGS)$$*$$unique"; then :; else \ test -n "$$unique" || unique=$$empty_fix; \ if test $$# -gt 0; then \ $(ETAGS) $(ETAGSFLAGS) $(AM_ETAGSFLAGS) $(ETAGS_ARGS) \ "$$@" $$unique; \ else \ $(ETAGS) $(ETAGSFLAGS) $(AM_ETAGSFLAGS) $(ETAGS_ARGS) \ $$unique; \ fi; \ fi ctags: CTAGS CTAGS: $(HEADERS) $(SOURCES) $(TAGS_DEPENDENCIES) \ $(TAGS_FILES) $(LISP) list='$(SOURCES) $(HEADERS) $(LISP) $(TAGS_FILES)'; \ unique=`for i in $$list; do \ if test -f "$$i"; then echo $$i; else echo $(srcdir)/$$i; fi; \ done | \ $(AWK) '{ files[$$0] = 1; nonempty = 1; } \ END { if (nonempty) { for (i in files) print i; }; }'`; \ test -z "$(CTAGS_ARGS)$$unique" \ || $(CTAGS) $(CTAGSFLAGS) $(AM_CTAGSFLAGS) $(CTAGS_ARGS) \ $$unique GTAGS: here=`$(am__cd) $(top_builddir) && pwd` \ && $(am__cd) $(top_srcdir) \ && gtags -i $(GTAGS_ARGS) "$$here" distclean-tags: -rm -f TAGS ID GTAGS GRTAGS GSYMS GPATH tags distdir: $(DISTFILES) @srcdirstrip=`echo "$(srcdir)" | sed 's/[].[^$$\\*]/\\\\&/g'`; \ topsrcdirstrip=`echo "$(top_srcdir)" | sed 's/[].[^$$\\*]/\\\\&/g'`; \ list='$(DISTFILES)'; \ dist_files=`for file in $$list; do echo $$file; done | \ sed -e "s|^$$srcdirstrip/||;t" \ -e "s|^$$topsrcdirstrip/|$(top_builddir)/|;t"`; \ case $$dist_files in \ */*) $(MKDIR_P) `echo "$$dist_files" | \ sed '/\//!d;s|^|$(distdir)/|;s,/[^/]*$$,,' | \ sort -u` ;; \ esac; \ for file in $$dist_files; do \ if test -f $$file || test -d $$file; then d=.; else d=$(srcdir); fi; \ if test -d $$d/$$file; then \ dir=`echo "/$$file" | sed -e 's,/[^/]*$$,,'`; \ if test -d "$(distdir)/$$file"; then \ find "$(distdir)/$$file" -type d ! -perm -700 -exec chmod u+rwx {} \;; \ fi; \ if test -d $(srcdir)/$$file && test $$d != $(srcdir); then \ cp -fpR $(srcdir)/$$file "$(distdir)$$dir" || exit 1; \ find "$(distdir)/$$file" -type d ! -perm -700 -exec chmod u+rwx {} \;; \ fi; \ cp -fpR $$d/$$file "$(distdir)$$dir" || exit 1; \ else \ test -f "$(distdir)/$$file" \ || cp -p $$d/$$file "$(distdir)/$$file" \ || exit 1; \ fi; \ done check-am: all-am check: $(BUILT_SOURCES) $(MAKE) $(AM_MAKEFLAGS) check-am all-am: Makefile $(LTLIBRARIES) $(HEADERS) installdirs: for dir in "$(DESTDIR)$(includedir)" "$(DESTDIR)$(includedir)"; do \ test -z "$$dir" || $(MKDIR_P) "$$dir"; \ done install: $(BUILT_SOURCES) $(MAKE) $(AM_MAKEFLAGS) install-am install-exec: install-exec-am install-data: install-data-am uninstall: uninstall-am install-am: all-am @$(MAKE) $(AM_MAKEFLAGS) install-exec-am install-data-am installcheck: installcheck-am install-strip: if test -z '$(STRIP)'; then \ $(MAKE) $(AM_MAKEFLAGS) INSTALL_PROGRAM="$(INSTALL_STRIP_PROGRAM)" \ install_sh_PROGRAM="$(INSTALL_STRIP_PROGRAM)" INSTALL_STRIP_FLAG=-s \ install; \ else \ $(MAKE) $(AM_MAKEFLAGS) INSTALL_PROGRAM="$(INSTALL_STRIP_PROGRAM)" \ install_sh_PROGRAM="$(INSTALL_STRIP_PROGRAM)" INSTALL_STRIP_FLAG=-s \ "INSTALL_PROGRAM_ENV=STRIPPROG='$(STRIP)'" install; \ fi mostlyclean-generic: clean-generic: -test -z "$(CLEANFILES)" || rm -f $(CLEANFILES) distclean-generic: -test -z "$(CONFIG_CLEAN_FILES)" || rm -f $(CONFIG_CLEAN_FILES) -test . = "$(srcdir)" || test -z "$(CONFIG_CLEAN_VPATH_FILES)" || rm -f $(CONFIG_CLEAN_VPATH_FILES) maintainer-clean-generic: @echo "This command is intended for maintainers to use" @echo "it deletes files that may require special tools to rebuild." -test -z "$(BUILT_SOURCES)" || rm -f $(BUILT_SOURCES) clean: clean-am clean-am: clean-generic clean-libtool clean-noinstLTLIBRARIES \ mostlyclean-am distclean: distclean-am -rm -rf ./$(DEPDIR) -rm -f Makefile distclean-am: clean-am distclean-compile distclean-generic \ distclean-tags dvi: dvi-am dvi-am: html: html-am html-am: info: info-am info-am: install-data-am: install-includeHEADERS install-nodist_includeHEADERS install-dvi: install-dvi-am install-dvi-am: install-exec-am: install-html: install-html-am install-html-am: install-info: install-info-am install-info-am: install-man: install-pdf: install-pdf-am install-pdf-am: install-ps: install-ps-am install-ps-am: installcheck-am: maintainer-clean: maintainer-clean-am -rm -rf ./$(DEPDIR) -rm -f Makefile maintainer-clean-am: distclean-am maintainer-clean-generic mostlyclean: mostlyclean-am mostlyclean-am: mostlyclean-compile mostlyclean-generic \ mostlyclean-libtool pdf: pdf-am pdf-am: ps: ps-am ps-am: uninstall-am: uninstall-includeHEADERS uninstall-nodist_includeHEADERS .MAKE: all check install install-am install-strip .PHONY: CTAGS GTAGS all all-am check check-am clean clean-generic \ clean-libtool clean-noinstLTLIBRARIES ctags distclean \ distclean-compile distclean-generic distclean-libtool \ distclean-tags distdir dvi dvi-am html html-am info info-am \ install install-am install-data install-data-am install-dvi \ install-dvi-am install-exec install-exec-am install-html \ install-html-am install-includeHEADERS install-info \ install-info-am install-man install-nodist_includeHEADERS \ install-pdf install-pdf-am install-ps install-ps-am \ install-strip installcheck installcheck-am installdirs \ maintainer-clean maintainer-clean-generic mostlyclean \ mostlyclean-compile mostlyclean-generic mostlyclean-libtool \ pdf pdf-am ps ps-am tags uninstall uninstall-am \ uninstall-includeHEADERS uninstall-nodist_includeHEADERS fftw3.f03: fftw3.f03.in (echo "! Generated automatically. DO NOT EDIT!"; echo; \ echo " integer, parameter :: C_FFTW_R2R_KIND = @C_FFTW_R2R_KIND@"; \ grep -v "Generated automatically" $(srcdir)/fftw3.f03.in) > $@ # convert constants to F77 PARAMETER statements @MAINTAINER_MODE_TRUE@fftw3.f: fftw3.h @MAINTAINER_MODE_TRUE@ rm -f $@ @MAINTAINER_MODE_TRUE@ perl -pe 's/([A-Z0-9_]+)=([+-]?[0-9]+)/\n INTEGER \1\n PARAMETER (\1=\2)\n/g' $< |egrep 'PARAMETER|INTEGER' > $@ @MAINTAINER_MODE_TRUE@ perl -pe 's/#define +([A-Z0-9_]+) +\(([+-]?[0-9]+)U?\)/\n INTEGER \1\n PARAMETER (\1=\2)\n/g' $< |egrep 'PARAMETER|INTEGER' >> $@ @MAINTAINER_MODE_TRUE@ perl -pe 'if (/#define +([A-Z0-9_]+) +\(([0-9]+)U? *<< *([0-9]+)\)/) { print "\n INTEGER $$1\n PARAMETER ($$1=",$$2 << $$3,")\n"; }' $< |egrep 'PARAMETER|INTEGER' >> $@ @MAINTAINER_MODE_TRUE@fftw3.f03.in: fftw3.h f03api.sh genf03.pl @MAINTAINER_MODE_TRUE@ sh $(srcdir)/f03api.sh d f > $@ @MAINTAINER_MODE_TRUE@fftw3l.f03: fftw3.h f03api.sh genf03.pl @MAINTAINER_MODE_TRUE@ sh $(srcdir)/f03api.sh l | grep -v parameter > $@ @MAINTAINER_MODE_TRUE@fftw3q.f03: fftw3.h f03api.sh genf03.pl @MAINTAINER_MODE_TRUE@ sh $(srcdir)/f03api.sh q | grep -v parameter > $@ # Tell versions [3.59,3.63) of GNU make to not export all variables. # Otherwise a system limit (for SysV at least) may be exceeded. .NOEXPORT: fftw-3.3.3/api/apiplan.c0000644000175000017500000001217312054410050011743 00000000000000/* * Copyright (c) 2003, 2007-11 Matteo Frigo * Copyright (c) 2003, 2007-11 Massachusetts Institute of Technology * * This program is free software; you can redistribute it and/or modify * it under the terms of the GNU General Public License as published by * the Free Software Foundation; either version 2 of the License, or * (at your option) any later version. * * This program is distributed in the hope that it will be useful, * but WITHOUT ANY WARRANTY; without even the implied warranty of * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the * GNU General Public License for more details. * * You should have received a copy of the GNU General Public License * along with this program; if not, write to the Free Software * Foundation, Inc., 51 Franklin Street, Fifth Floor, Boston, MA 02110-1301 USA * */ #include "api.h" static plan *mkplan0(planner *plnr, unsigned flags, const problem *prb, int hash_info, wisdom_state_t wisdom_state) { /* map API flags into FFTW flags */ X(mapflags)(plnr, flags); plnr->flags.hash_info = hash_info; plnr->wisdom_state = wisdom_state; /* create plan */ return plnr->adt->mkplan(plnr, prb); } static unsigned force_estimator(unsigned flags) { flags &= ~(FFTW_MEASURE | FFTW_PATIENT | FFTW_EXHAUSTIVE); return (flags | FFTW_ESTIMATE); } static plan *mkplan(planner *plnr, unsigned flags, const problem *prb, int hash_info) { plan *pln; pln = mkplan0(plnr, flags, prb, hash_info, WISDOM_NORMAL); if (plnr->wisdom_state == WISDOM_NORMAL && !pln) { /* maybe the planner failed because of inconsistent wisdom; plan again ignoring infeasible wisdom */ pln = mkplan0(plnr, force_estimator(flags), prb, hash_info, WISDOM_IGNORE_INFEASIBLE); } if (plnr->wisdom_state == WISDOM_IS_BOGUS) { /* if the planner detected a wisdom inconsistency, forget all wisdom and plan again */ plnr->adt->forget(plnr, FORGET_EVERYTHING); A(!pln); pln = mkplan0(plnr, flags, prb, hash_info, WISDOM_NORMAL); if (plnr->wisdom_state == WISDOM_IS_BOGUS) { /* if it still fails, plan without wisdom */ plnr->adt->forget(plnr, FORGET_EVERYTHING); A(!pln); pln = mkplan0(plnr, force_estimator(flags), prb, hash_info, WISDOM_IGNORE_ALL); } } return pln; } apiplan *X(mkapiplan)(int sign, unsigned flags, problem *prb) { apiplan *p = 0; plan *pln; unsigned flags_used_for_planning; planner *plnr = X(the_planner)(); unsigned int pats[] = {FFTW_ESTIMATE, FFTW_MEASURE, FFTW_PATIENT, FFTW_EXHAUSTIVE}; int pat, pat_max; double pcost = 0; if (flags & FFTW_WISDOM_ONLY) { /* Special mode that returns a plan only if wisdom is present, and returns 0 otherwise. This is now documented in the manual, as a way to detect whether wisdom is available for a problem. */ flags_used_for_planning = flags; pln = mkplan0(plnr, flags, prb, 0, WISDOM_ONLY); } else { pat_max = flags & FFTW_ESTIMATE ? 0 : (flags & FFTW_EXHAUSTIVE ? 3 : (flags & FFTW_PATIENT ? 2 : 1)); pat = plnr->timelimit >= 0 ? 0 : pat_max; flags &= ~(FFTW_ESTIMATE | FFTW_MEASURE | FFTW_PATIENT | FFTW_EXHAUSTIVE); plnr->start_time = X(get_crude_time)(); /* plan at incrementally increasing patience until we run out of time */ for (pln = 0, flags_used_for_planning = 0; pat <= pat_max; ++pat) { plan *pln1; unsigned tmpflags = flags | pats[pat]; pln1 = mkplan(plnr, tmpflags, prb, 0); if (!pln1) { /* don't bother continuing if planner failed or timed out */ A(!pln || plnr->timed_out); break; } X(plan_destroy_internal)(pln); pln = pln1; flags_used_for_planning = tmpflags; pcost = pln->pcost; } } if (pln) { /* build apiplan */ p = (apiplan *) MALLOC(sizeof(apiplan), PLANS); p->prb = prb; p->sign = sign; /* cache for execute_dft */ /* re-create plan from wisdom, adding blessing */ p->pln = mkplan(plnr, flags_used_for_planning, prb, BLESSING); /* record pcost from most recent measurement for use in X(cost) */ p->pln->pcost = pcost; if (sizeof(trigreal) > sizeof(R)) { /* this is probably faster, and we have enough trigreal bits to maintain accuracy */ X(plan_awake)(p->pln, AWAKE_SQRTN_TABLE); } else { /* more accurate */ X(plan_awake)(p->pln, AWAKE_SINCOS); } /* we don't use pln for p->pln, above, since by re-creating the plan we might use more patient wisdom from a timed-out mkplan */ X(plan_destroy_internal)(pln); } else X(problem_destroy)(prb); /* discard all information not necessary to reconstruct the plan */ plnr->adt->forget(plnr, FORGET_ACCURSED); #ifdef FFTW_RANDOM_ESTIMATOR X(random_estimate_seed)++; /* subsequent "random" plans are distinct */ #endif return p; } void X(destroy_plan)(X(plan) p) { if (p) { X(plan_awake)(p->pln, SLEEPY); X(plan_destroy_internal)(p->pln); X(problem_destroy)(p->prb); X(ifree)(p); } } fftw-3.3.3/api/configure.c0000644000175000017500000000205212054410050012273 00000000000000/* * Copyright (c) 2003, 2007-11 Matteo Frigo * Copyright (c) 2003, 2007-11 Massachusetts Institute of Technology * * This program is free software; you can redistribute it and/or modify * it under the terms of the GNU General Public License as published by * the Free Software Foundation; either version 2 of the License, or * (at your option) any later version. * * This program is distributed in the hope that it will be useful, * but WITHOUT ANY WARRANTY; without even the implied warranty of * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the * GNU General Public License for more details. * * You should have received a copy of the GNU General Public License * along with this program; if not, write to the Free Software * Foundation, Inc., 51 Franklin Street, Fifth Floor, Boston, MA 02110-1301 USA * */ #include "api.h" #include "dft.h" #include "rdft.h" #include "reodft.h" void X(configure_planner)(planner *plnr) { X(dft_conf_standard)(plnr); X(rdft_conf_standard)(plnr); X(reodft_conf_standard)(plnr); } fftw-3.3.3/api/execute-dft-c2r.c0000644000175000017500000000224112054410050013213 00000000000000/* * Copyright (c) 2003, 2007-11 Matteo Frigo * Copyright (c) 2003, 2007-11 Massachusetts Institute of Technology * * This program is free software; you can redistribute it and/or modify * it under the terms of the GNU General Public License as published by * the Free Software Foundation; either version 2 of the License, or * (at your option) any later version. * * This program is distributed in the hope that it will be useful, * but WITHOUT ANY WARRANTY; without even the implied warranty of * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the * GNU General Public License for more details. * * You should have received a copy of the GNU General Public License * along with this program; if not, write to the Free Software * Foundation, Inc., 51 Franklin Street, Fifth Floor, Boston, MA 02110-1301 USA * */ #include "api.h" #include "rdft.h" /* guru interface: requires care in alignment, r - i, etcetera. */ void X(execute_dft_c2r)(const X(plan) p, C *in, R *out) { plan_rdft2 *pln = (plan_rdft2 *) p->pln; problem_rdft2 *prb = (problem_rdft2 *) p->prb; pln->apply((plan *) pln, out, out + (prb->r1 - prb->r0), in[0], in[0]+1); } fftw-3.3.3/api/execute-dft-r2c.c0000644000175000017500000000224112054410050013213 00000000000000/* * Copyright (c) 2003, 2007-11 Matteo Frigo * Copyright (c) 2003, 2007-11 Massachusetts Institute of Technology * * This program is free software; you can redistribute it and/or modify * it under the terms of the GNU General Public License as published by * the Free Software Foundation; either version 2 of the License, or * (at your option) any later version. * * This program is distributed in the hope that it will be useful, * but WITHOUT ANY WARRANTY; without even the implied warranty of * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the * GNU General Public License for more details. * * You should have received a copy of the GNU General Public License * along with this program; if not, write to the Free Software * Foundation, Inc., 51 Franklin Street, Fifth Floor, Boston, MA 02110-1301 USA * */ #include "api.h" #include "rdft.h" /* guru interface: requires care in alignment, r - i, etcetera. */ void X(execute_dft_r2c)(const X(plan) p, R *in, C *out) { plan_rdft2 *pln = (plan_rdft2 *) p->pln; problem_rdft2 *prb = (problem_rdft2 *) p->prb; pln->apply((plan *) pln, in, in + (prb->r1 - prb->r0), out[0], out[0]+1); } fftw-3.3.3/api/execute-dft.c0000644000175000017500000000226312054410050012533 00000000000000/* * Copyright (c) 2003, 2007-11 Matteo Frigo * Copyright (c) 2003, 2007-11 Massachusetts Institute of Technology * * This program is free software; you can redistribute it and/or modify * it under the terms of the GNU General Public License as published by * the Free Software Foundation; either version 2 of the License, or * (at your option) any later version. * * This program is distributed in the hope that it will be useful, * but WITHOUT ANY WARRANTY; without even the implied warranty of * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the * GNU General Public License for more details. * * You should have received a copy of the GNU General Public License * along with this program; if not, write to the Free Software * Foundation, Inc., 51 Franklin Street, Fifth Floor, Boston, MA 02110-1301 USA * */ #include "api.h" #include "dft.h" /* guru interface: requires care in alignment etcetera. */ void X(execute_dft)(const X(plan) p, C *in, C *out) { plan_dft *pln = (plan_dft *) p->pln; if (p->sign == FFT_SIGN) pln->apply((plan *) pln, in[0], in[0]+1, out[0], out[0]+1); else pln->apply((plan *) pln, in[0]+1, in[0], out[0]+1, out[0]); } fftw-3.3.3/api/execute-r2r.c0000644000175000017500000000207112054410050012460 00000000000000/* * Copyright (c) 2003, 2007-11 Matteo Frigo * Copyright (c) 2003, 2007-11 Massachusetts Institute of Technology * * This program is free software; you can redistribute it and/or modify * it under the terms of the GNU General Public License as published by * the Free Software Foundation; either version 2 of the License, or * (at your option) any later version. * * This program is distributed in the hope that it will be useful, * but WITHOUT ANY WARRANTY; without even the implied warranty of * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the * GNU General Public License for more details. * * You should have received a copy of the GNU General Public License * along with this program; if not, write to the Free Software * Foundation, Inc., 51 Franklin Street, Fifth Floor, Boston, MA 02110-1301 USA * */ #include "api.h" #include "rdft.h" /* guru interface: requires care in alignment, etcetera. */ void X(execute_r2r)(const X(plan) p, R *in, R *out) { plan_rdft *pln = (plan_rdft *) p->pln; pln->apply((plan *) pln, in, out); } fftw-3.3.3/api/execute-split-dft-c2r.c0000644000175000017500000000224612054410050014351 00000000000000/* * Copyright (c) 2003, 2007-11 Matteo Frigo * Copyright (c) 2003, 2007-11 Massachusetts Institute of Technology * * This program is free software; you can redistribute it and/or modify * it under the terms of the GNU General Public License as published by * the Free Software Foundation; either version 2 of the License, or * (at your option) any later version. * * This program is distributed in the hope that it will be useful, * but WITHOUT ANY WARRANTY; without even the implied warranty of * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the * GNU General Public License for more details. * * You should have received a copy of the GNU General Public License * along with this program; if not, write to the Free Software * Foundation, Inc., 51 Franklin Street, Fifth Floor, Boston, MA 02110-1301 USA * */ #include "api.h" #include "rdft.h" /* guru interface: requires care in alignment, r - i, etcetera. */ void X(execute_split_dft_c2r)(const X(plan) p, R *ri, R *ii, R *out) { plan_rdft2 *pln = (plan_rdft2 *) p->pln; problem_rdft2 *prb = (problem_rdft2 *) p->prb; pln->apply((plan *) pln, out, out + (prb->r1 - prb->r0), ri, ii); } fftw-3.3.3/api/execute-split-dft-r2c.c0000644000175000017500000000224312054410050014346 00000000000000/* * Copyright (c) 2003, 2007-11 Matteo Frigo * Copyright (c) 2003, 2007-11 Massachusetts Institute of Technology * * This program is free software; you can redistribute it and/or modify * it under the terms of the GNU General Public License as published by * the Free Software Foundation; either version 2 of the License, or * (at your option) any later version. * * This program is distributed in the hope that it will be useful, * but WITHOUT ANY WARRANTY; without even the implied warranty of * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the * GNU General Public License for more details. * * You should have received a copy of the GNU General Public License * along with this program; if not, write to the Free Software * Foundation, Inc., 51 Franklin Street, Fifth Floor, Boston, MA 02110-1301 USA * */ #include "api.h" #include "rdft.h" /* guru interface: requires care in alignment, r - i, etcetera. */ void X(execute_split_dft_r2c)(const X(plan) p, R *in, R *ro, R *io) { plan_rdft2 *pln = (plan_rdft2 *) p->pln; problem_rdft2 *prb = (problem_rdft2 *) p->prb; pln->apply((plan *) pln, in, in + (prb->r1 - prb->r0), ro, io); } fftw-3.3.3/api/execute-split-dft.c0000644000175000017500000000212712054410050013663 00000000000000/* * Copyright (c) 2003, 2007-11 Matteo Frigo * Copyright (c) 2003, 2007-11 Massachusetts Institute of Technology * * This program is free software; you can redistribute it and/or modify * it under the terms of the GNU General Public License as published by * the Free Software Foundation; either version 2 of the License, or * (at your option) any later version. * * This program is distributed in the hope that it will be useful, * but WITHOUT ANY WARRANTY; without even the implied warranty of * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the * GNU General Public License for more details. * * You should have received a copy of the GNU General Public License * along with this program; if not, write to the Free Software * Foundation, Inc., 51 Franklin Street, Fifth Floor, Boston, MA 02110-1301 USA * */ #include "api.h" #include "dft.h" /* guru interface: requires care in alignment, r - i, etcetera. */ void X(execute_split_dft)(const X(plan) p, R *ri, R *ii, R *ro, R *io) { plan_dft *pln = (plan_dft *) p->pln; pln->apply((plan *) pln, ri, ii, ro, io); } fftw-3.3.3/api/execute.c0000644000175000017500000000170012054410050011753 00000000000000/* * Copyright (c) 2003, 2007-11 Matteo Frigo * Copyright (c) 2003, 2007-11 Massachusetts Institute of Technology * * This program is free software; you can redistribute it and/or modify * it under the terms of the GNU General Public License as published by * the Free Software Foundation; either version 2 of the License, or * (at your option) any later version. * * This program is distributed in the hope that it will be useful, * but WITHOUT ANY WARRANTY; without even the implied warranty of * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the * GNU General Public License for more details. * * You should have received a copy of the GNU General Public License * along with this program; if not, write to the Free Software * Foundation, Inc., 51 Franklin Street, Fifth Floor, Boston, MA 02110-1301 USA * */ #include "api.h" void X(execute)(const X(plan) p) { plan *pln = p->pln; pln->adt->solve(pln, p->prb); } fftw-3.3.3/api/export-wisdom-to-file.c0000644000175000017500000000251012054410050014467 00000000000000/* * Copyright (c) 2003, 2007-11 Matteo Frigo * Copyright (c) 2003, 2007-11 Massachusetts Institute of Technology * * This program is free software; you can redistribute it and/or modify * it under the terms of the GNU General Public License as published by * the Free Software Foundation; either version 2 of the License, or * (at your option) any later version. * * This program is distributed in the hope that it will be useful, * but WITHOUT ANY WARRANTY; without even the implied warranty of * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the * GNU General Public License for more details. * * You should have received a copy of the GNU General Public License * along with this program; if not, write to the Free Software * Foundation, Inc., 51 Franklin Street, Fifth Floor, Boston, MA 02110-1301 USA * */ #include "api.h" void X(export_wisdom_to_file)(FILE *output_file) { printer *p = X(mkprinter_file)(output_file); planner *plnr = X(the_planner)(); plnr->adt->exprt(plnr, p); X(printer_destroy)(p); } int X(export_wisdom_to_filename)(const char *filename) { FILE *f = fopen(filename, "w"); int ret; if (!f) return 0; /* error opening file */ X(export_wisdom_to_file)(f); ret = !ferror(f); if (fclose(f)) ret = 0; /* error closing file */ return ret; } fftw-3.3.3/api/export-wisdom-to-string.c0000644000175000017500000000364712054410050015072 00000000000000/* * Copyright (c) 2003, 2007-11 Matteo Frigo * Copyright (c) 2003, 2007-11 Massachusetts Institute of Technology * * This program is free software; you can redistribute it and/or modify * it under the terms of the GNU General Public License as published by * the Free Software Foundation; either version 2 of the License, or * (at your option) any later version. * * This program is distributed in the hope that it will be useful, * but WITHOUT ANY WARRANTY; without even the implied warranty of * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the * GNU General Public License for more details. * * You should have received a copy of the GNU General Public License * along with this program; if not, write to the Free Software * Foundation, Inc., 51 Franklin Street, Fifth Floor, Boston, MA 02110-1301 USA * */ #include "api.h" typedef struct { printer super; int *cnt; } P_cnt; static void putchr_cnt(printer * p_, char c) { P_cnt *p = (P_cnt *) p_; UNUSED(c); ++*p->cnt; } static printer *mkprinter_cnt(int *cnt) { P_cnt *p = (P_cnt *) X(mkprinter)(sizeof(P_cnt), putchr_cnt, 0); p->cnt = cnt; *cnt = 0; return &p->super; } typedef struct { printer super; char *s; } P_str; static void putchr_str(printer * p_, char c) { P_str *p = (P_str *) p_; *p->s++ = c; *p->s = 0; } static printer *mkprinter_str(char *s) { P_str *p = (P_str *) X(mkprinter)(sizeof(P_str), putchr_str, 0); p->s = s; *s = 0; return &p->super; } char *X(export_wisdom_to_string)(void) { printer *p; planner *plnr = X(the_planner)(); int cnt; char *s; p = mkprinter_cnt(&cnt); plnr->adt->exprt(plnr, p); X(printer_destroy)(p); s = (char *) malloc(sizeof(char) * (cnt + 1)); if (s) { p = mkprinter_str(s); plnr->adt->exprt(plnr, p); X(printer_destroy)(p); } return s; } fftw-3.3.3/api/export-wisdom.c0000644000175000017500000000255412054410050013142 00000000000000/* * Copyright (c) 2003, 2007-11 Matteo Frigo * Copyright (c) 2003, 2007-11 Massachusetts Institute of Technology * * This program is free software; you can redistribute it and/or modify * it under the terms of the GNU General Public License as published by * the Free Software Foundation; either version 2 of the License, or * (at your option) any later version. * * This program is distributed in the hope that it will be useful, * but WITHOUT ANY WARRANTY; without even the implied warranty of * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the * GNU General Public License for more details. * * You should have received a copy of the GNU General Public License * along with this program; if not, write to the Free Software * Foundation, Inc., 51 Franklin Street, Fifth Floor, Boston, MA 02110-1301 USA * */ #include "api.h" typedef struct { printer super; void (*write_char)(char c, void *); void *data; } P; static void putchr_generic(printer * p_, char c) { P *p = (P *) p_; (p->write_char)(c, p->data); } void X(export_wisdom)(void (*write_char)(char c, void *), void *data) { P *p = (P *) X(mkprinter)(sizeof(P), putchr_generic, 0); planner *plnr = X(the_planner)(); p->write_char = write_char; p->data = data; plnr->adt->exprt(plnr, (printer *) p); X(printer_destroy)((printer *) p); } fftw-3.3.3/api/f77api.c0000644000175000017500000001145112054410050011412 00000000000000/* * Copyright (c) 2003, 2007-11 Matteo Frigo * Copyright (c) 2003, 2007-11 Massachusetts Institute of Technology * * This program is free software; you can redistribute it and/or modify * it under the terms of the GNU General Public License as published by * the Free Software Foundation; either version 2 of the License, or * (at your option) any later version. * * This program is distributed in the hope that it will be useful, * but WITHOUT ANY WARRANTY; without even the implied warranty of * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the * GNU General Public License for more details. * * You should have received a copy of the GNU General Public License * along with this program; if not, write to the Free Software * Foundation, Inc., 51 Franklin Street, Fifth Floor, Boston, MA 02110-1301 USA * */ #include "api.h" #include "dft.h" #include "rdft.h" #include "x77.h" /* if F77_FUNC is not defined, then we don't know how to mangle identifiers for the Fortran linker, and we must omit the f77 API. */ #if defined(F77_FUNC) || defined(WINDOWS_F77_MANGLING) /*-----------------------------------------------------------------------*/ /* some internal functions used by the f77 api */ /* in fortran, the natural array ordering is column-major, which corresponds to reversing the dimensions relative to C's row-major */ static int *reverse_n(int rnk, const int *n) { int *nrev; int i; A(FINITE_RNK(rnk)); nrev = (int *) MALLOC(sizeof(int) * rnk, PROBLEMS); for (i = 0; i < rnk; ++i) nrev[rnk - i - 1] = n[i]; return nrev; } /* f77 doesn't have data structures, so we have to pass iodims as parallel arrays */ static X(iodim) *make_dims(int rnk, const int *n, const int *is, const int *os) { X(iodim) *dims; int i; A(FINITE_RNK(rnk)); dims = (X(iodim) *) MALLOC(sizeof(X(iodim)) * rnk, PROBLEMS); for (i = 0; i < rnk; ++i) { dims[i].n = n[i]; dims[i].is = is[i]; dims[i].os = os[i]; } return dims; } typedef struct { void (*f77_write_char)(char *, void *); void *data; } write_char_data; static void write_char(char c, void *d) { write_char_data *ad = (write_char_data *) d; ad->f77_write_char(&c, ad->data); } typedef struct { void (*f77_read_char)(int *, void *); void *data; } read_char_data; static int read_char(void *d) { read_char_data *ed = (read_char_data *) d; int c; ed->f77_read_char(&c, ed->data); return (c < 0 ? EOF : c); } static X(r2r_kind) *ints2kinds(int rnk, const int *ik) { if (!FINITE_RNK(rnk) || rnk == 0) return 0; else { int i; X(r2r_kind) *k; k = (X(r2r_kind) *) MALLOC(sizeof(X(r2r_kind)) * rnk, PROBLEMS); /* reverse order for Fortran -> C */ for (i = 0; i < rnk; ++i) k[i] = (X(r2r_kind)) ik[rnk - 1 - i]; return k; } } /*-----------------------------------------------------------------------*/ #define F77(a, A) F77x(x77(a), X77(A)) #ifndef WINDOWS_F77_MANGLING #if defined(F77_FUNC) # define F77x(a, A) F77_FUNC(a, A) # include "f77funcs.h" #endif /* If identifiers with underscores are mangled differently than those without underscores, then we include *both* mangling versions. The reason is that the only Fortran compiler that does such differing mangling is currently g77 (which adds an extra underscore to names with underscores), whereas other compilers running on the same machine are likely to use non-underscored mangling. (I'm sick of users complaining that FFTW works with g77 but not with e.g. pgf77 or ifc on the same machine.) Note that all FFTW identifiers contain underscores, and configure picks g77 by default. */ #if defined(F77_FUNC_) && !defined(F77_FUNC_EQUIV) # undef F77x # define F77x(a, A) F77_FUNC_(a, A) # include "f77funcs.h" #endif #else /* WINDOWS_F77_MANGLING */ /* Various mangling conventions common (?) under Windows. */ /* g77 */ # define WINDOWS_F77_FUNC(a, A) a ## __ # define F77x(a, A) WINDOWS_F77_FUNC(a, A) # include "f77funcs.h" /* Intel, etc. */ # undef WINDOWS_F77_FUNC # define WINDOWS_F77_FUNC(a, A) a ## _ # include "f77funcs.h" /* Digital/Compaq/HP Visual Fortran, Intel Fortran. stdcall attribute is apparently required to adjust for calling conventions (callee pops stack in stdcall). See also: http://msdn.microsoft.com/library/en-us/vccore98/html/_core_mixed.2d.language_programming.3a_.overview.asp */ # undef WINDOWS_F77_FUNC # if defined(__GNUC__) # define WINDOWS_F77_FUNC(a, A) __attribute__((stdcall)) A # elif defined(_MSC_VER) || defined(_ICC) || defined(_STDCALL_SUPPORTED) # define WINDOWS_F77_FUNC(a, A) __stdcall A # else # define WINDOWS_F77_FUNC(a, A) A /* oh well */ # endif # include "f77funcs.h" #endif /* WINDOWS_F77_MANGLING */ #endif /* F77_FUNC */ fftw-3.3.3/api/flops.c0000644000175000017500000000261312054410050011440 00000000000000/* * Copyright (c) 2003, 2007-11 Matteo Frigo * Copyright (c) 2003, 2007-11 Massachusetts Institute of Technology * * This program is free software; you can redistribute it and/or modify * it under the terms of the GNU General Public License as published by * the Free Software Foundation; either version 2 of the License, or * (at your option) any later version. * * This program is distributed in the hope that it will be useful, * but WITHOUT ANY WARRANTY; without even the implied warranty of * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the * GNU General Public License for more details. * * You should have received a copy of the GNU General Public License * along with this program; if not, write to the Free Software * Foundation, Inc., 51 Franklin Street, Fifth Floor, Boston, MA 02110-1301 USA * */ #include "api.h" void X(flops)(const X(plan) p, double *add, double *mul, double *fma) { planner *plnr = X(the_planner)(); opcnt *o = &p->pln->ops; *add = o->add; *mul = o->mul; *fma = o->fma; if (plnr->cost_hook) { *add = plnr->cost_hook(p->prb, *add, COST_SUM); *mul = plnr->cost_hook(p->prb, *mul, COST_SUM); *fma = plnr->cost_hook(p->prb, *fma, COST_SUM); } } double X(estimate_cost)(const X(plan) p) { return X(iestimate_cost)(X(the_planner)(), p->pln, p->prb); } double X(cost)(const X(plan) p) { return p->pln->pcost; } fftw-3.3.3/api/forget-wisdom.c0000644000175000017500000000172712054410050013110 00000000000000/* * Copyright (c) 2003, 2007-11 Matteo Frigo * Copyright (c) 2003, 2007-11 Massachusetts Institute of Technology * * This program is free software; you can redistribute it and/or modify * it under the terms of the GNU General Public License as published by * the Free Software Foundation; either version 2 of the License, or * (at your option) any later version. * * This program is distributed in the hope that it will be useful, * but WITHOUT ANY WARRANTY; without even the implied warranty of * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the * GNU General Public License for more details. * * You should have received a copy of the GNU General Public License * along with this program; if not, write to the Free Software * Foundation, Inc., 51 Franklin Street, Fifth Floor, Boston, MA 02110-1301 USA * */ #include "api.h" void X(forget_wisdom)(void) { planner *plnr = X(the_planner)(); plnr->adt->forget(plnr, FORGET_EVERYTHING); } fftw-3.3.3/api/import-system-wisdom.c0000644000175000017500000000277312054410050014460 00000000000000/* * Copyright (c) 2003, 2007-11 Matteo Frigo * Copyright (c) 2003, 2007-11 Massachusetts Institute of Technology * * This program is free software; you can redistribute it and/or modify * it under the terms of the GNU General Public License as published by * the Free Software Foundation; either version 2 of the License, or * (at your option) any later version. * * This program is distributed in the hope that it will be useful, * but WITHOUT ANY WARRANTY; without even the implied warranty of * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the * GNU General Public License for more details. * * You should have received a copy of the GNU General Public License * along with this program; if not, write to the Free Software * Foundation, Inc., 51 Franklin Street, Fifth Floor, Boston, MA 02110-1301 USA * */ #include "api.h" #if defined(FFTW_SINGLE) # define WISDOM_NAME "wisdomf" #elif defined(FFTW_LDOUBLE) # define WISDOM_NAME "wisdoml" #else # define WISDOM_NAME "wisdom" #endif /* OS-specific configuration-file directory */ #if defined(__DJGPP__) # define WISDOM_DIR "/dev/env/DJDIR/etc/fftw/" #else # define WISDOM_DIR "/etc/fftw/" #endif int X(import_system_wisdom)(void) { #if defined(__WIN32__) || defined(WIN32) || defined(_WINDOWS) return 0; /* TODO? */ #else FILE *f; f = fopen(WISDOM_DIR WISDOM_NAME, "r"); if (f) { int ret = X(import_wisdom_from_file)(f); fclose(f); return ret; } else return 0; #endif } fftw-3.3.3/api/import-wisdom-from-file.c0000644000175000017500000000440112054410050015002 00000000000000/* * Copyright (c) 2003, 2007-11 Matteo Frigo * Copyright (c) 2003, 2007-11 Massachusetts Institute of Technology * * This program is free software; you can redistribute it and/or modify * it under the terms of the GNU General Public License as published by * the Free Software Foundation; either version 2 of the License, or * (at your option) any later version. * * This program is distributed in the hope that it will be useful, * but WITHOUT ANY WARRANTY; without even the implied warranty of * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the * GNU General Public License for more details. * * You should have received a copy of the GNU General Public License * along with this program; if not, write to the Free Software * Foundation, Inc., 51 Franklin Street, Fifth Floor, Boston, MA 02110-1301 USA * */ #include "api.h" #include /* getc()/putc() are *unbelievably* slow on linux. Looks like glibc is grabbing a lock for each call to getc()/putc(), or something like that. You pay the price for these idiotic posix threads whether you use them or not. So, we do our own buffering. This completely defeats the purpose of having stdio in the first place, of course. */ #define BUFSZ 256 typedef struct { scanner super; FILE *f; char buf[BUFSZ]; char *bufr, *bufw; } S; static int getchr_file(scanner * sc_) { S *sc = (S *) sc_; if (sc->bufr >= sc->bufw) { sc->bufr = sc->buf; sc->bufw = sc->buf + fread(sc->buf, 1, BUFSZ, sc->f); if (sc->bufr >= sc->bufw) return EOF; } return *(sc->bufr++); } static scanner *mkscanner_file(FILE *f) { S *sc = (S *) X(mkscanner)(sizeof(S), getchr_file); sc->f = f; sc->bufr = sc->bufw = sc->buf; return &sc->super; } int X(import_wisdom_from_file)(FILE *input_file) { scanner *s = mkscanner_file(input_file); planner *plnr = X(the_planner)(); int ret = plnr->adt->imprt(plnr, s); X(scanner_destroy)(s); return ret; } int X(import_wisdom_from_filename)(const char *filename) { FILE *f = fopen(filename, "r"); int ret; if (!f) return 0; /* error opening file */ ret = X(import_wisdom_from_file)(f); if (fclose(f)) ret = 0; /* error closing file */ return ret; } fftw-3.3.3/api/import-wisdom-from-string.c0000644000175000017500000000266412054410050015402 00000000000000/* * Copyright (c) 2003, 2007-11 Matteo Frigo * Copyright (c) 2003, 2007-11 Massachusetts Institute of Technology * * This program is free software; you can redistribute it and/or modify * it under the terms of the GNU General Public License as published by * the Free Software Foundation; either version 2 of the License, or * (at your option) any later version. * * This program is distributed in the hope that it will be useful, * but WITHOUT ANY WARRANTY; without even the implied warranty of * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the * GNU General Public License for more details. * * You should have received a copy of the GNU General Public License * along with this program; if not, write to the Free Software * Foundation, Inc., 51 Franklin Street, Fifth Floor, Boston, MA 02110-1301 USA * */ #include "api.h" typedef struct { scanner super; const char *s; } S_str; static int getchr_str(scanner * sc_) { S_str *sc = (S_str *) sc_; if (!*sc->s) return EOF; return *sc->s++; } static scanner *mkscanner_str(const char *s) { S_str *sc = (S_str *) X(mkscanner)(sizeof(S_str), getchr_str); sc->s = s; return &sc->super; } int X(import_wisdom_from_string)(const char *input_string) { scanner *s = mkscanner_str(input_string); planner *plnr = X(the_planner)(); int ret = plnr->adt->imprt(plnr, s); X(scanner_destroy)(s); return ret; } fftw-3.3.3/api/import-wisdom.c0000644000175000017500000000256112054410050013131 00000000000000/* * Copyright (c) 2003, 2007-11 Matteo Frigo * Copyright (c) 2003, 2007-11 Massachusetts Institute of Technology * * This program is free software; you can redistribute it and/or modify * it under the terms of the GNU General Public License as published by * the Free Software Foundation; either version 2 of the License, or * (at your option) any later version. * * This program is distributed in the hope that it will be useful, * but WITHOUT ANY WARRANTY; without even the implied warranty of * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the * GNU General Public License for more details. * * You should have received a copy of the GNU General Public License * along with this program; if not, write to the Free Software * Foundation, Inc., 51 Franklin Street, Fifth Floor, Boston, MA 02110-1301 USA * */ #include "api.h" typedef struct { scanner super; int (*read_char)(void *); void *data; } S; static int getchr_generic(scanner * s_) { S *s = (S *) s_; return (s->read_char)(s->data); } int X(import_wisdom)(int (*read_char)(void *), void *data) { S *s = (S *) X(mkscanner)(sizeof(S), getchr_generic); planner *plnr = X(the_planner)(); int ret; s->read_char = read_char; s->data = data; ret = plnr->adt->imprt(plnr, (scanner *) s); X(scanner_destroy)((scanner *) s); return ret; } fftw-3.3.3/api/malloc.c0000644000175000017500000000307312054410050011565 00000000000000/* * Copyright (c) 2003, 2007-11 Matteo Frigo * Copyright (c) 2003, 2007-11 Massachusetts Institute of Technology * * This program is free software; you can redistribute it and/or modify * it under the terms of the GNU General Public License as published by * the Free Software Foundation; either version 2 of the License, or * (at your option) any later version. * * This program is distributed in the hope that it will be useful, * but WITHOUT ANY WARRANTY; without even the implied warranty of * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the * GNU General Public License for more details. * * You should have received a copy of the GNU General Public License * along with this program; if not, write to the Free Software * Foundation, Inc., 51 Franklin Street, Fifth Floor, Boston, MA 02110-1301 USA * */ #include "api.h" void *X(malloc)(size_t n) { return X(kernel_malloc)(n); } void X(free)(void *p) { X(kernel_free)(p); } /* The following two routines are mainly for the convenience of the Fortran 2003 API, although C users may find them convienent as well. The problem is that, although Fortran 2003 has a c_sizeof intrinsic that is equivalent to sizeof, it is broken in some gfortran versions, and in any case is a bit unnatural in a Fortran context. So we provide routines to allocate real and complex arrays, which are all that are really needed by FFTW. */ R *X(alloc_real)(size_t n) { return (R *) X(malloc)(sizeof(R) * n); } C *X(alloc_complex)(size_t n) { return (C *) X(malloc)(sizeof(C) * n); } fftw-3.3.3/api/map-r2r-kind.c0000644000175000017500000000326512054410050012524 00000000000000/* * Copyright (c) 2003, 2007-11 Matteo Frigo * Copyright (c) 2003, 2007-11 Massachusetts Institute of Technology * * This program is free software; you can redistribute it and/or modify * it under the terms of the GNU General Public License as published by * the Free Software Foundation; either version 2 of the License, or * (at your option) any later version. * * This program is distributed in the hope that it will be useful, * but WITHOUT ANY WARRANTY; without even the implied warranty of * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the * GNU General Public License for more details. * * You should have received a copy of the GNU General Public License * along with this program; if not, write to the Free Software * Foundation, Inc., 51 Franklin Street, Fifth Floor, Boston, MA 02110-1301 USA * */ #include "api.h" #include "rdft.h" rdft_kind *X(map_r2r_kind)(int rank, const X(r2r_kind) * kind) { int i; rdft_kind *k; A(FINITE_RNK(rank)); k = (rdft_kind *) MALLOC(rank * sizeof(rdft_kind), PROBLEMS); for (i = 0; i < rank; ++i) { rdft_kind m; switch (kind[i]) { case FFTW_R2HC: m = R2HC; break; case FFTW_HC2R: m = HC2R; break; case FFTW_DHT: m = DHT; break; case FFTW_REDFT00: m = REDFT00; break; case FFTW_REDFT01: m = REDFT01; break; case FFTW_REDFT10: m = REDFT10; break; case FFTW_REDFT11: m = REDFT11; break; case FFTW_RODFT00: m = RODFT00; break; case FFTW_RODFT01: m = RODFT01; break; case FFTW_RODFT10: m = RODFT10; break; case FFTW_RODFT11: m = RODFT11; break; default: m = R2HC; A(0); } k[i] = m; } return k; } fftw-3.3.3/api/mapflags.c0000644000175000017500000001262612054410050012114 00000000000000/* * Copyright (c) 2003, 2007-11 Matteo Frigo * Copyright (c) 2003, 2007-11 Massachusetts Institute of Technology * * This program is free software; you can redistribute it and/or modify * it under the terms of the GNU General Public License as published by * the Free Software Foundation; either version 2 of the License, or * (at your option) any later version. * * This program is distributed in the hope that it will be useful, * but WITHOUT ANY WARRANTY; without even the implied warranty of * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the * GNU General Public License for more details. * * You should have received a copy of the GNU General Public License * along with this program; if not, write to the Free Software * Foundation, Inc., 51 Franklin Street, Fifth Floor, Boston, MA 02110-1301 USA * */ #include "api.h" #include /* a flag operation: x is either a flag, in which case xm == 0, or a mask, in which case xm == x; using this we can compactly code the various bit operations via (flags & x) ^ xm or (flags | x) ^ xm. */ typedef struct { unsigned x, xm; } flagmask; typedef struct { flagmask flag; flagmask op; } flagop; #define FLAGP(f, msk)(((f) & (msk).x) ^ (msk).xm) #define OP(f, msk)(((f) | (msk).x) ^ (msk).xm) #define YES(x) {x, 0} #define NO(x) {x, x} #define IMPLIES(predicate, consequence) { predicate, consequence } #define EQV(a, b) IMPLIES(YES(a), YES(b)), IMPLIES(NO(a), NO(b)) #define NEQV(a, b) IMPLIES(YES(a), NO(b)), IMPLIES(NO(a), YES(b)) static void map_flags(unsigned *iflags, unsigned *oflags, const flagop flagmap[], int nmap) { int i; for (i = 0; i < nmap; ++i) if (FLAGP(*iflags, flagmap[i].flag)) *oflags = OP(*oflags, flagmap[i].op); } /* encoding of the planner timelimit into a BITS_FOR_TIMELIMIT-bits nonnegative integer, such that we can still view the integer as ``impatience'': higher means *lower* time limit, and 0 is the highest possible value (about 1 year of calendar time) */ static unsigned timelimit_to_flags(double timelimit) { const double tmax = 365 * 24 * 3600; const double tstep = 1.05; const int nsteps = (1 << BITS_FOR_TIMELIMIT); int x; if (timelimit < 0 || timelimit >= tmax) return 0; if (timelimit <= 1.0e-10) return nsteps - 1; x = (int) (0.5 + (log(tmax / timelimit) / log(tstep))); if (x < 0) x = 0; if (x >= nsteps) x = nsteps - 1; return x; } void X(mapflags)(planner *plnr, unsigned flags) { unsigned l, u, t; /* map of api flags -> api flags, to implement consistency rules and combination flags */ const flagop self_flagmap[] = { /* in some cases (notably for halfcomplex->real transforms), DESTROY_INPUT is the default, so we need to support an inverse flag to disable it. (PRESERVE, DESTROY) -> (PRESERVE, DESTROY) (0, 0) (1, 0) (0, 1) (0, 1) (1, 0) (1, 0) (1, 1) (1, 0) */ IMPLIES(YES(FFTW_PRESERVE_INPUT), NO(FFTW_DESTROY_INPUT)), IMPLIES(NO(FFTW_DESTROY_INPUT), YES(FFTW_PRESERVE_INPUT)), IMPLIES(YES(FFTW_EXHAUSTIVE), YES(FFTW_PATIENT)), IMPLIES(YES(FFTW_ESTIMATE), NO(FFTW_PATIENT)), IMPLIES(YES(FFTW_ESTIMATE), YES(FFTW_ESTIMATE_PATIENT | FFTW_NO_INDIRECT_OP | FFTW_ALLOW_PRUNING)), IMPLIES(NO(FFTW_EXHAUSTIVE), YES(FFTW_NO_SLOW)), /* a canonical set of fftw2-like impatience flags */ IMPLIES(NO(FFTW_PATIENT), YES(FFTW_NO_VRECURSE | FFTW_NO_RANK_SPLITS | FFTW_NO_VRANK_SPLITS | FFTW_NO_NONTHREADED | FFTW_NO_DFT_R2HC | FFTW_NO_FIXED_RADIX_LARGE_N | FFTW_BELIEVE_PCOST)) }; /* map of (processed) api flags to internal problem/planner flags */ const flagop l_flagmap[] = { EQV(FFTW_PRESERVE_INPUT, NO_DESTROY_INPUT), EQV(FFTW_NO_SIMD, NO_SIMD), EQV(FFTW_CONSERVE_MEMORY, CONSERVE_MEMORY), EQV(FFTW_NO_BUFFERING, NO_BUFFERING), NEQV(FFTW_ALLOW_LARGE_GENERIC, NO_LARGE_GENERIC) }; const flagop u_flagmap[] = { IMPLIES(YES(FFTW_EXHAUSTIVE), NO(0xFFFFFFFF)), IMPLIES(NO(FFTW_EXHAUSTIVE), YES(NO_UGLY)), /* the following are undocumented, "beyond-guru" flags that require some understanding of FFTW internals */ EQV(FFTW_ESTIMATE_PATIENT, ESTIMATE), EQV(FFTW_ALLOW_PRUNING, ALLOW_PRUNING), EQV(FFTW_BELIEVE_PCOST, BELIEVE_PCOST), EQV(FFTW_NO_DFT_R2HC, NO_DFT_R2HC), EQV(FFTW_NO_NONTHREADED, NO_NONTHREADED), EQV(FFTW_NO_INDIRECT_OP, NO_INDIRECT_OP), EQV(FFTW_NO_RANK_SPLITS, NO_RANK_SPLITS), EQV(FFTW_NO_VRANK_SPLITS, NO_VRANK_SPLITS), EQV(FFTW_NO_VRECURSE, NO_VRECURSE), EQV(FFTW_NO_SLOW, NO_SLOW), EQV(FFTW_NO_FIXED_RADIX_LARGE_N, NO_FIXED_RADIX_LARGE_N) }; map_flags(&flags, &flags, self_flagmap, NELEM(self_flagmap)); l = u = 0; map_flags(&flags, &l, l_flagmap, NELEM(l_flagmap)); map_flags(&flags, &u, u_flagmap, NELEM(u_flagmap)); /* enforce l <= u */ PLNR_L(plnr) = l; PLNR_U(plnr) = u | l; /* assert that the conversion didn't lose bits */ A(PLNR_L(plnr) == l); A(PLNR_U(plnr) == (u | l)); /* compute flags representation of the timelimit */ t = timelimit_to_flags(plnr->timelimit); PLNR_TIMELIMIT_IMPATIENCE(plnr) = t; A(PLNR_TIMELIMIT_IMPATIENCE(plnr) == t); } fftw-3.3.3/api/mkprinter-file.c0000644000175000017500000000271412054410050013247 00000000000000/* * Copyright (c) 2003, 2007-11 Matteo Frigo * Copyright (c) 2003, 2007-11 Massachusetts Institute of Technology * * This program is free software; you can redistribute it and/or modify * it under the terms of the GNU General Public License as published by * the Free Software Foundation; either version 2 of the License, or * (at your option) any later version. * * This program is distributed in the hope that it will be useful, * but WITHOUT ANY WARRANTY; without even the implied warranty of * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the * GNU General Public License for more details. * * You should have received a copy of the GNU General Public License * along with this program; if not, write to the Free Software * Foundation, Inc., 51 Franklin Street, Fifth Floor, Boston, MA 02110-1301 USA * */ #include "api.h" #include #define BUFSZ 256 typedef struct { printer super; FILE *f; char buf[BUFSZ]; char *bufw; } P; static void myflush(P *p) { fwrite(p->buf, 1, p->bufw - p->buf, p->f); p->bufw = p->buf; } static void myputchr(printer *p_, char c) { P *p = (P *) p_; if (p->bufw >= p->buf + BUFSZ) myflush(p); *p->bufw++ = c; } static void mycleanup(printer *p_) { P *p = (P *) p_; myflush(p); } printer *X(mkprinter_file)(FILE *f) { P *p = (P *) X(mkprinter)(sizeof(P), myputchr, mycleanup); p->f = f; p->bufw = p->buf; return &p->super; } fftw-3.3.3/api/mktensor-iodims.c0000644000175000017500000000005712054410050013441 00000000000000#include "guru.h" #include "mktensor-iodims.h" fftw-3.3.3/api/mktensor-rowmajor.c0000644000175000017500000000341412054410050014015 00000000000000/* * Copyright (c) 2003, 2007-11 Matteo Frigo * Copyright (c) 2003, 2007-11 Massachusetts Institute of Technology * * This program is free software; you can redistribute it and/or modify * it under the terms of the GNU General Public License as published by * the Free Software Foundation; either version 2 of the License, or * (at your option) any later version. * * This program is distributed in the hope that it will be useful, * but WITHOUT ANY WARRANTY; without even the implied warranty of * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the * GNU General Public License for more details. * * You should have received a copy of the GNU General Public License * along with this program; if not, write to the Free Software * Foundation, Inc., 51 Franklin Street, Fifth Floor, Boston, MA 02110-1301 USA * */ #include "api.h" tensor *X(mktensor_rowmajor)(int rnk, const int *n, const int *niphys, const int *nophys, int is, int os) { tensor *x = X(mktensor)(rnk); if (FINITE_RNK(rnk) && rnk > 0) { int i; A(n && niphys && nophys); x->dims[rnk - 1].is = is; x->dims[rnk - 1].os = os; x->dims[rnk - 1].n = n[rnk - 1]; for (i = rnk - 1; i > 0; --i) { x->dims[i - 1].is = x->dims[i].is * niphys[i]; x->dims[i - 1].os = x->dims[i].os * nophys[i]; x->dims[i - 1].n = n[i - 1]; } } return x; } static int rowmajor_kosherp(int rnk, const int *n) { int i; if (!FINITE_RNK(rnk)) return 0; if (rnk < 0) return 0; for (i = 0; i < rnk; ++i) if (n[i] <= 0) return 0; return 1; } int X(many_kosherp)(int rnk, const int *n, int howmany) { return (howmany >= 0) && rowmajor_kosherp(rnk, n); } fftw-3.3.3/api/plan-dft-1d.c0000644000175000017500000000176112054410050012327 00000000000000/* * Copyright (c) 2003, 2007-11 Matteo Frigo * Copyright (c) 2003, 2007-11 Massachusetts Institute of Technology * * This program is free software; you can redistribute it and/or modify * it under the terms of the GNU General Public License as published by * the Free Software Foundation; either version 2 of the License, or * (at your option) any later version. * * This program is distributed in the hope that it will be useful, * but WITHOUT ANY WARRANTY; without even the implied warranty of * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the * GNU General Public License for more details. * * You should have received a copy of the GNU General Public License * along with this program; if not, write to the Free Software * Foundation, Inc., 51 Franklin Street, Fifth Floor, Boston, MA 02110-1301 USA * */ #include "api.h" #include "dft.h" X(plan) X(plan_dft_1d)(int n, C *in, C *out, int sign, unsigned flags) { return X(plan_dft)(1, &n, in, out, sign, flags); } fftw-3.3.3/api/plan-dft-2d.c0000644000175000017500000000205012054410050012320 00000000000000/* * Copyright (c) 2003, 2007-11 Matteo Frigo * Copyright (c) 2003, 2007-11 Massachusetts Institute of Technology * * This program is free software; you can redistribute it and/or modify * it under the terms of the GNU General Public License as published by * the Free Software Foundation; either version 2 of the License, or * (at your option) any later version. * * This program is distributed in the hope that it will be useful, * but WITHOUT ANY WARRANTY; without even the implied warranty of * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the * GNU General Public License for more details. * * You should have received a copy of the GNU General Public License * along with this program; if not, write to the Free Software * Foundation, Inc., 51 Franklin Street, Fifth Floor, Boston, MA 02110-1301 USA * */ #include "api.h" #include "dft.h" X(plan) X(plan_dft_2d)(int nx, int ny, C *in, C *out, int sign, unsigned flags) { int n[2]; n[0] = nx; n[1] = ny; return X(plan_dft)(2, n, in, out, sign, flags); } fftw-3.3.3/api/plan-dft-3d.c0000644000175000017500000000211112054410050012317 00000000000000/* * Copyright (c) 2003, 2007-11 Matteo Frigo * Copyright (c) 2003, 2007-11 Massachusetts Institute of Technology * * This program is free software; you can redistribute it and/or modify * it under the terms of the GNU General Public License as published by * the Free Software Foundation; either version 2 of the License, or * (at your option) any later version. * * This program is distributed in the hope that it will be useful, * but WITHOUT ANY WARRANTY; without even the implied warranty of * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the * GNU General Public License for more details. * * You should have received a copy of the GNU General Public License * along with this program; if not, write to the Free Software * Foundation, Inc., 51 Franklin Street, Fifth Floor, Boston, MA 02110-1301 USA * */ #include "api.h" #include "dft.h" X(plan) X(plan_dft_3d)(int nx, int ny, int nz, C *in, C *out, int sign, unsigned flags) { int n[3]; n[0] = nx; n[1] = ny; n[2] = nz; return X(plan_dft)(3, n, in, out, sign, flags); } fftw-3.3.3/api/plan-dft-c2r-1d.c0000644000175000017500000000173012054410050013007 00000000000000/* * Copyright (c) 2003, 2007-11 Matteo Frigo * Copyright (c) 2003, 2007-11 Massachusetts Institute of Technology * * This program is free software; you can redistribute it and/or modify * it under the terms of the GNU General Public License as published by * the Free Software Foundation; either version 2 of the License, or * (at your option) any later version. * * This program is distributed in the hope that it will be useful, * but WITHOUT ANY WARRANTY; without even the implied warranty of * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the * GNU General Public License for more details. * * You should have received a copy of the GNU General Public License * along with this program; if not, write to the Free Software * Foundation, Inc., 51 Franklin Street, Fifth Floor, Boston, MA 02110-1301 USA * */ #include "api.h" X(plan) X(plan_dft_c2r_1d)(int n, C *in, R *out, unsigned flags) { return X(plan_dft_c2r)(1, &n, in, out, flags); } fftw-3.3.3/api/plan-dft-c2r-2d.c0000644000175000017500000000201712054410050013007 00000000000000/* * Copyright (c) 2003, 2007-11 Matteo Frigo * Copyright (c) 2003, 2007-11 Massachusetts Institute of Technology * * This program is free software; you can redistribute it and/or modify * it under the terms of the GNU General Public License as published by * the Free Software Foundation; either version 2 of the License, or * (at your option) any later version. * * This program is distributed in the hope that it will be useful, * but WITHOUT ANY WARRANTY; without even the implied warranty of * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the * GNU General Public License for more details. * * You should have received a copy of the GNU General Public License * along with this program; if not, write to the Free Software * Foundation, Inc., 51 Franklin Street, Fifth Floor, Boston, MA 02110-1301 USA * */ #include "api.h" X(plan) X(plan_dft_c2r_2d)(int nx, int ny, C *in, R *out, unsigned flags) { int n[2]; n[0] = nx; n[1] = ny; return X(plan_dft_c2r)(2, n, in, out, flags); } fftw-3.3.3/api/plan-dft-c2r-3d.c0000644000175000017500000000205512054410050013012 00000000000000/* * Copyright (c) 2003, 2007-11 Matteo Frigo * Copyright (c) 2003, 2007-11 Massachusetts Institute of Technology * * This program is free software; you can redistribute it and/or modify * it under the terms of the GNU General Public License as published by * the Free Software Foundation; either version 2 of the License, or * (at your option) any later version. * * This program is distributed in the hope that it will be useful, * but WITHOUT ANY WARRANTY; without even the implied warranty of * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the * GNU General Public License for more details. * * You should have received a copy of the GNU General Public License * along with this program; if not, write to the Free Software * Foundation, Inc., 51 Franklin Street, Fifth Floor, Boston, MA 02110-1301 USA * */ #include "api.h" X(plan) X(plan_dft_c2r_3d)(int nx, int ny, int nz, C *in, R *out, unsigned flags) { int n[3]; n[0] = nx; n[1] = ny; n[2] = nz; return X(plan_dft_c2r)(3, n, in, out, flags); } fftw-3.3.3/api/plan-dft-c2r.c0000644000175000017500000000200712054410050012503 00000000000000/* * Copyright (c) 2003, 2007-11 Matteo Frigo * Copyright (c) 2003, 2007-11 Massachusetts Institute of Technology * * This program is free software; you can redistribute it and/or modify * it under the terms of the GNU General Public License as published by * the Free Software Foundation; either version 2 of the License, or * (at your option) any later version. * * This program is distributed in the hope that it will be useful, * but WITHOUT ANY WARRANTY; without even the implied warranty of * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the * GNU General Public License for more details. * * You should have received a copy of the GNU General Public License * along with this program; if not, write to the Free Software * Foundation, Inc., 51 Franklin Street, Fifth Floor, Boston, MA 02110-1301 USA * */ #include "api.h" X(plan) X(plan_dft_c2r)(int rank, const int *n, C *in, R *out, unsigned flags) { return X(plan_many_dft_c2r)(rank, n, 1, in, 0, 1, 1, out, 0, 1, 1, flags); } fftw-3.3.3/api/plan-dft-r2c-1d.c0000644000175000017500000000173012054410050013007 00000000000000/* * Copyright (c) 2003, 2007-11 Matteo Frigo * Copyright (c) 2003, 2007-11 Massachusetts Institute of Technology * * This program is free software; you can redistribute it and/or modify * it under the terms of the GNU General Public License as published by * the Free Software Foundation; either version 2 of the License, or * (at your option) any later version. * * This program is distributed in the hope that it will be useful, * but WITHOUT ANY WARRANTY; without even the implied warranty of * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the * GNU General Public License for more details. * * You should have received a copy of the GNU General Public License * along with this program; if not, write to the Free Software * Foundation, Inc., 51 Franklin Street, Fifth Floor, Boston, MA 02110-1301 USA * */ #include "api.h" X(plan) X(plan_dft_r2c_1d)(int n, R *in, C *out, unsigned flags) { return X(plan_dft_r2c)(1, &n, in, out, flags); } fftw-3.3.3/api/plan-dft-r2c-2d.c0000644000175000017500000000201712054410050013007 00000000000000/* * Copyright (c) 2003, 2007-11 Matteo Frigo * Copyright (c) 2003, 2007-11 Massachusetts Institute of Technology * * This program is free software; you can redistribute it and/or modify * it under the terms of the GNU General Public License as published by * the Free Software Foundation; either version 2 of the License, or * (at your option) any later version. * * This program is distributed in the hope that it will be useful, * but WITHOUT ANY WARRANTY; without even the implied warranty of * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the * GNU General Public License for more details. * * You should have received a copy of the GNU General Public License * along with this program; if not, write to the Free Software * Foundation, Inc., 51 Franklin Street, Fifth Floor, Boston, MA 02110-1301 USA * */ #include "api.h" X(plan) X(plan_dft_r2c_2d)(int nx, int ny, R *in, C *out, unsigned flags) { int n[2]; n[0] = nx; n[1] = ny; return X(plan_dft_r2c)(2, n, in, out, flags); } fftw-3.3.3/api/plan-dft-r2c-3d.c0000644000175000017500000000205512054410050013012 00000000000000/* * Copyright (c) 2003, 2007-11 Matteo Frigo * Copyright (c) 2003, 2007-11 Massachusetts Institute of Technology * * This program is free software; you can redistribute it and/or modify * it under the terms of the GNU General Public License as published by * the Free Software Foundation; either version 2 of the License, or * (at your option) any later version. * * This program is distributed in the hope that it will be useful, * but WITHOUT ANY WARRANTY; without even the implied warranty of * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the * GNU General Public License for more details. * * You should have received a copy of the GNU General Public License * along with this program; if not, write to the Free Software * Foundation, Inc., 51 Franklin Street, Fifth Floor, Boston, MA 02110-1301 USA * */ #include "api.h" X(plan) X(plan_dft_r2c_3d)(int nx, int ny, int nz, R *in, C *out, unsigned flags) { int n[3]; n[0] = nx; n[1] = ny; n[2] = nz; return X(plan_dft_r2c)(3, n, in, out, flags); } fftw-3.3.3/api/plan-dft-r2c.c0000644000175000017500000000202312054410050012501 00000000000000/* * Copyright (c) 2003, 2007-11 Matteo Frigo * Copyright (c) 2003, 2007-11 Massachusetts Institute of Technology * * This program is free software; you can redistribute it and/or modify * it under the terms of the GNU General Public License as published by * the Free Software Foundation; either version 2 of the License, or * (at your option) any later version. * * This program is distributed in the hope that it will be useful, * but WITHOUT ANY WARRANTY; without even the implied warranty of * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the * GNU General Public License for more details. * * You should have received a copy of the GNU General Public License * along with this program; if not, write to the Free Software * Foundation, Inc., 51 Franklin Street, Fifth Floor, Boston, MA 02110-1301 USA * */ #include "api.h" X(plan) X(plan_dft_r2c)(int rank, const int *n, R *in, C *out, unsigned flags) { return X(plan_many_dft_r2c)(rank, n, 1, in, 0, 1, 1, out, 0, 1, 1, flags); } fftw-3.3.3/api/plan-dft.c0000644000175000017500000000205212054410050012017 00000000000000/* * Copyright (c) 2003, 2007-11 Matteo Frigo * Copyright (c) 2003, 2007-11 Massachusetts Institute of Technology * * This program is free software; you can redistribute it and/or modify * it under the terms of the GNU General Public License as published by * the Free Software Foundation; either version 2 of the License, or * (at your option) any later version. * * This program is distributed in the hope that it will be useful, * but WITHOUT ANY WARRANTY; without even the implied warranty of * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the * GNU General Public License for more details. * * You should have received a copy of the GNU General Public License * along with this program; if not, write to the Free Software * Foundation, Inc., 51 Franklin Street, Fifth Floor, Boston, MA 02110-1301 USA * */ #include "api.h" X(plan) X(plan_dft)(int rank, const int *n, C *in, C *out, int sign, unsigned flags) { return X(plan_many_dft)(rank, n, 1, in, 0, 1, 1, out, 0, 1, 1, sign, flags); } fftw-3.3.3/api/plan-guru-dft-c2r.c0000644000175000017500000000006112054410050013461 00000000000000#include "guru.h" #include "plan-guru-dft-c2r.h" fftw-3.3.3/api/plan-guru-dft-r2c.c0000644000175000017500000000006112054410050013461 00000000000000#include "guru.h" #include "plan-guru-dft-r2c.h" fftw-3.3.3/api/plan-guru-dft.c0000644000175000017500000000005512054410050013000 00000000000000#include "guru.h" #include "plan-guru-dft.h" fftw-3.3.3/api/plan-guru-r2r.c0000644000175000017500000000005512054410050012730 00000000000000#include "guru.h" #include "plan-guru-r2r.h" fftw-3.3.3/api/plan-guru-split-dft-c2r.c0000644000175000017500000000006712054410050014620 00000000000000#include "guru.h" #include "plan-guru-split-dft-c2r.h" fftw-3.3.3/api/plan-guru-split-dft-r2c.c0000644000175000017500000000006712054410050014620 00000000000000#include "guru.h" #include "plan-guru-split-dft-r2c.h" fftw-3.3.3/api/plan-guru-split-dft.c0000644000175000017500000000006312054410050014130 00000000000000#include "guru.h" #include "plan-guru-split-dft.h" fftw-3.3.3/api/plan-many-dft-c2r.c0000644000175000017500000000346412054410050013455 00000000000000/* * Copyright (c) 2003, 2007-11 Matteo Frigo * Copyright (c) 2003, 2007-11 Massachusetts Institute of Technology * * This program is free software; you can redistribute it and/or modify * it under the terms of the GNU General Public License as published by * the Free Software Foundation; either version 2 of the License, or * (at your option) any later version. * * This program is distributed in the hope that it will be useful, * but WITHOUT ANY WARRANTY; without even the implied warranty of * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the * GNU General Public License for more details. * * You should have received a copy of the GNU General Public License * along with this program; if not, write to the Free Software * Foundation, Inc., 51 Franklin Street, Fifth Floor, Boston, MA 02110-1301 USA * */ #include "api.h" #include "rdft.h" X(plan) X(plan_many_dft_c2r)(int rank, const int *n, int howmany, C *in, const int *inembed, int istride, int idist, R *out, const int *onembed, int ostride, int odist, unsigned flags) { R *ri, *ii; int *nfi, *nfo; int inplace; X(plan) p; if (!X(many_kosherp)(rank, n, howmany)) return 0; EXTRACT_REIM(FFT_SIGN, in, &ri, &ii); inplace = out == ri; if (!inplace) flags |= FFTW_DESTROY_INPUT; p = X(mkapiplan)( 0, flags, X(mkproblem_rdft2_d_3pointers)( X(mktensor_rowmajor)( rank, n, X(rdft2_pad)(rank, n, inembed, inplace, 1, &nfi), X(rdft2_pad)(rank, n, onembed, inplace, 0, &nfo), 2 * istride, ostride), X(mktensor_1d)(howmany, 2 * idist, odist), TAINT_UNALIGNED(out, flags), TAINT_UNALIGNED(ri, flags), TAINT_UNALIGNED(ii, flags), HC2R)); X(ifree0)(nfi); X(ifree0)(nfo); return p; } fftw-3.3.3/api/plan-many-dft-r2c.c0000644000175000017500000000340112054410050013444 00000000000000/* * Copyright (c) 2003, 2007-11 Matteo Frigo * Copyright (c) 2003, 2007-11 Massachusetts Institute of Technology * * This program is free software; you can redistribute it and/or modify * it under the terms of the GNU General Public License as published by * the Free Software Foundation; either version 2 of the License, or * (at your option) any later version. * * This program is distributed in the hope that it will be useful, * but WITHOUT ANY WARRANTY; without even the implied warranty of * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the * GNU General Public License for more details. * * You should have received a copy of the GNU General Public License * along with this program; if not, write to the Free Software * Foundation, Inc., 51 Franklin Street, Fifth Floor, Boston, MA 02110-1301 USA * */ #include "api.h" #include "rdft.h" X(plan) X(plan_many_dft_r2c)(int rank, const int *n, int howmany, R *in, const int *inembed, int istride, int idist, C *out, const int *onembed, int ostride, int odist, unsigned flags) { R *ro, *io; int *nfi, *nfo; int inplace; X(plan) p; if (!X(many_kosherp)(rank, n, howmany)) return 0; EXTRACT_REIM(FFT_SIGN, out, &ro, &io); inplace = in == ro; p = X(mkapiplan)( 0, flags, X(mkproblem_rdft2_d_3pointers)( X(mktensor_rowmajor)( rank, n, X(rdft2_pad)(rank, n, inembed, inplace, 0, &nfi), X(rdft2_pad)(rank, n, onembed, inplace, 1, &nfo), istride, 2 * ostride), X(mktensor_1d)(howmany, idist, 2 * odist), TAINT_UNALIGNED(in, flags), TAINT_UNALIGNED(ro, flags), TAINT_UNALIGNED(io, flags), R2HC)); X(ifree0)(nfi); X(ifree0)(nfo); return p; } fftw-3.3.3/api/plan-many-dft.c0000644000175000017500000000321412054410050012762 00000000000000/* * Copyright (c) 2003, 2007-11 Matteo Frigo * Copyright (c) 2003, 2007-11 Massachusetts Institute of Technology * * This program is free software; you can redistribute it and/or modify * it under the terms of the GNU General Public License as published by * the Free Software Foundation; either version 2 of the License, or * (at your option) any later version. * * This program is distributed in the hope that it will be useful, * but WITHOUT ANY WARRANTY; without even the implied warranty of * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the * GNU General Public License for more details. * * You should have received a copy of the GNU General Public License * along with this program; if not, write to the Free Software * Foundation, Inc., 51 Franklin Street, Fifth Floor, Boston, MA 02110-1301 USA * */ #include "api.h" #include "dft.h" #define N0(nembed)((nembed) ? (nembed) : n) X(plan) X(plan_many_dft)(int rank, const int *n, int howmany, C *in, const int *inembed, int istride, int idist, C *out, const int *onembed, int ostride, int odist, int sign, unsigned flags) { R *ri, *ii, *ro, *io; if (!X(many_kosherp)(rank, n, howmany)) return 0; EXTRACT_REIM(sign, in, &ri, &ii); EXTRACT_REIM(sign, out, &ro, &io); return X(mkapiplan)(sign, flags, X(mkproblem_dft_d)( X(mktensor_rowmajor)(rank, n, N0(inembed), N0(onembed), 2 * istride, 2 * ostride), X(mktensor_1d)(howmany, 2 * idist, 2 * odist), TAINT_UNALIGNED(ri, flags), TAINT_UNALIGNED(ii, flags), TAINT_UNALIGNED(ro, flags), TAINT_UNALIGNED(io, flags))); } fftw-3.3.3/api/plan-many-r2r.c0000644000175000017500000000310712054410050012713 00000000000000/* * Copyright (c) 2003, 2007-11 Matteo Frigo * Copyright (c) 2003, 2007-11 Massachusetts Institute of Technology * * This program is free software; you can redistribute it and/or modify * it under the terms of the GNU General Public License as published by * the Free Software Foundation; either version 2 of the License, or * (at your option) any later version. * * This program is distributed in the hope that it will be useful, * but WITHOUT ANY WARRANTY; without even the implied warranty of * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the * GNU General Public License for more details. * * You should have received a copy of the GNU General Public License * along with this program; if not, write to the Free Software * Foundation, Inc., 51 Franklin Street, Fifth Floor, Boston, MA 02110-1301 USA * */ #include "api.h" #include "rdft.h" #define N0(nembed)((nembed) ? (nembed) : n) X(plan) X(plan_many_r2r)(int rank, const int *n, int howmany, R *in, const int *inembed, int istride, int idist, R *out, const int *onembed, int ostride, int odist, const X(r2r_kind) * kind, unsigned flags) { X(plan) p; rdft_kind *k; if (!X(many_kosherp)(rank, n, howmany)) return 0; k = X(map_r2r_kind)(rank, kind); p = X(mkapiplan)( 0, flags, X(mkproblem_rdft_d)(X(mktensor_rowmajor)(rank, n, N0(inembed), N0(onembed), istride, ostride), X(mktensor_1d)(howmany, idist, odist), TAINT_UNALIGNED(in, flags), TAINT_UNALIGNED(out, flags), k)); X(ifree0)(k); return p; } fftw-3.3.3/api/plan-r2r-1d.c0000644000175000017500000000175112054410050012256 00000000000000/* * Copyright (c) 2003, 2007-11 Matteo Frigo * Copyright (c) 2003, 2007-11 Massachusetts Institute of Technology * * This program is free software; you can redistribute it and/or modify * it under the terms of the GNU General Public License as published by * the Free Software Foundation; either version 2 of the License, or * (at your option) any later version. * * This program is distributed in the hope that it will be useful, * but WITHOUT ANY WARRANTY; without even the implied warranty of * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the * GNU General Public License for more details. * * You should have received a copy of the GNU General Public License * along with this program; if not, write to the Free Software * Foundation, Inc., 51 Franklin Street, Fifth Floor, Boston, MA 02110-1301 USA * */ #include "api.h" X(plan) X(plan_r2r_1d)(int n, R *in, R *out, X(r2r_kind) kind, unsigned flags) { return X(plan_r2r)(1, &n, in, out, &kind, flags); } fftw-3.3.3/api/plan-r2r-2d.c0000644000175000017500000000220212054410050012247 00000000000000/* * Copyright (c) 2003, 2007-11 Matteo Frigo * Copyright (c) 2003, 2007-11 Massachusetts Institute of Technology * * This program is free software; you can redistribute it and/or modify * it under the terms of the GNU General Public License as published by * the Free Software Foundation; either version 2 of the License, or * (at your option) any later version. * * This program is distributed in the hope that it will be useful, * but WITHOUT ANY WARRANTY; without even the implied warranty of * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the * GNU General Public License for more details. * * You should have received a copy of the GNU General Public License * along with this program; if not, write to the Free Software * Foundation, Inc., 51 Franklin Street, Fifth Floor, Boston, MA 02110-1301 USA * */ #include "api.h" X(plan) X(plan_r2r_2d)(int nx, int ny, R *in, R *out, X(r2r_kind) kindx, X(r2r_kind) kindy, unsigned flags) { int n[2]; X(r2r_kind) kind[2]; n[0] = nx; n[1] = ny; kind[0] = kindx; kind[1] = kindy; return X(plan_r2r)(2, n, in, out, kind, flags); } fftw-3.3.3/api/plan-r2r-3d.c0000644000175000017500000000231412054410050012254 00000000000000/* * Copyright (c) 2003, 2007-11 Matteo Frigo * Copyright (c) 2003, 2007-11 Massachusetts Institute of Technology * * This program is free software; you can redistribute it and/or modify * it under the terms of the GNU General Public License as published by * the Free Software Foundation; either version 2 of the License, or * (at your option) any later version. * * This program is distributed in the hope that it will be useful, * but WITHOUT ANY WARRANTY; without even the implied warranty of * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the * GNU General Public License for more details. * * You should have received a copy of the GNU General Public License * along with this program; if not, write to the Free Software * Foundation, Inc., 51 Franklin Street, Fifth Floor, Boston, MA 02110-1301 USA * */ #include "api.h" X(plan) X(plan_r2r_3d)(int nx, int ny, int nz, R *in, R *out, X(r2r_kind) kindx, X(r2r_kind) kindy, X(r2r_kind) kindz, unsigned flags) { int n[3]; X(r2r_kind) kind[3]; n[0] = nx; n[1] = ny; n[2] = nz; kind[0] = kindx; kind[1] = kindy; kind[2] = kindz; return X(plan_r2r)(3, n, in, out, kind, flags); } fftw-3.3.3/api/plan-r2r.c0000644000175000017500000000205012054410050011745 00000000000000/* * Copyright (c) 2003, 2007-11 Matteo Frigo * Copyright (c) 2003, 2007-11 Massachusetts Institute of Technology * * This program is free software; you can redistribute it and/or modify * it under the terms of the GNU General Public License as published by * the Free Software Foundation; either version 2 of the License, or * (at your option) any later version. * * This program is distributed in the hope that it will be useful, * but WITHOUT ANY WARRANTY; without even the implied warranty of * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the * GNU General Public License for more details. * * You should have received a copy of the GNU General Public License * along with this program; if not, write to the Free Software * Foundation, Inc., 51 Franklin Street, Fifth Floor, Boston, MA 02110-1301 USA * */ #include "api.h" X(plan) X(plan_r2r)(int rank, const int *n, R *in, R *out, const X(r2r_kind) * kind, unsigned flags) { return X(plan_many_r2r)(rank, n, 1, in, 0, 1, 1, out, 0, 1, 1, kind, flags); } fftw-3.3.3/api/print-plan.c0000644000175000017500000000215412054410050012401 00000000000000/* * Copyright (c) 2003, 2007-11 Matteo Frigo * Copyright (c) 2003, 2007-11 Massachusetts Institute of Technology * * This program is free software; you can redistribute it and/or modify * it under the terms of the GNU General Public License as published by * the Free Software Foundation; either version 2 of the License, or * (at your option) any later version. * * This program is distributed in the hope that it will be useful, * but WITHOUT ANY WARRANTY; without even the implied warranty of * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the * GNU General Public License for more details. * * You should have received a copy of the GNU General Public License * along with this program; if not, write to the Free Software * Foundation, Inc., 51 Franklin Street, Fifth Floor, Boston, MA 02110-1301 USA * */ #include "api.h" void X(fprint_plan)(const X(plan) p, FILE *output_file) { printer *pr = X(mkprinter_file)(output_file); plan *pln = p->pln; pln->adt->print(pln, pr); X(printer_destroy)(pr); } void X(print_plan)(const X(plan) p) { X(fprint_plan)(p, stdout); } fftw-3.3.3/api/rdft2-pad.c0000644000175000017500000000256112054410050012102 00000000000000/* * Copyright (c) 2003, 2007-11 Matteo Frigo * Copyright (c) 2003, 2007-11 Massachusetts Institute of Technology * * This program is free software; you can redistribute it and/or modify * it under the terms of the GNU General Public License as published by * the Free Software Foundation; either version 2 of the License, or * (at your option) any later version. * * This program is distributed in the hope that it will be useful, * but WITHOUT ANY WARRANTY; without even the implied warranty of * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the * GNU General Public License for more details. * * You should have received a copy of the GNU General Public License * along with this program; if not, write to the Free Software * Foundation, Inc., 51 Franklin Street, Fifth Floor, Boston, MA 02110-1301 USA * */ #include #include "api.h" const int *X(rdft2_pad)(int rnk, const int *n, const int *nembed, int inplace, int cmplx, int **nfree) { A(FINITE_RNK(rnk)); *nfree = 0; if (!nembed && rnk > 0) { if (inplace || cmplx) { int *np = (int *) MALLOC(sizeof(int) * rnk, PROBLEMS); memcpy(np, n, sizeof(int) * rnk); np[rnk - 1] = (n[rnk - 1] / 2 + 1) * (1 + !cmplx); nembed = *nfree = np; } else nembed = n; } return nembed; } fftw-3.3.3/api/the-planner.c0000644000175000017500000000254412054410050012535 00000000000000/* * Copyright (c) 2003, 2007-11 Matteo Frigo * Copyright (c) 2003, 2007-11 Massachusetts Institute of Technology * * This program is free software; you can redistribute it and/or modify * it under the terms of the GNU General Public License as published by * the Free Software Foundation; either version 2 of the License, or * (at your option) any later version. * * This program is distributed in the hope that it will be useful, * but WITHOUT ANY WARRANTY; without even the implied warranty of * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the * GNU General Public License for more details. * * You should have received a copy of the GNU General Public License * along with this program; if not, write to the Free Software * Foundation, Inc., 51 Franklin Street, Fifth Floor, Boston, MA 02110-1301 USA * */ #include "api.h" static planner *plnr = 0; /* create the planner for the rest of the API */ planner *X(the_planner)(void) { if (!plnr) { plnr = X(mkplanner)(); X(configure_planner)(plnr); } return plnr; } void X(cleanup)(void) { if (plnr) { X(planner_destroy)(plnr); plnr = 0; } } void X(set_timelimit)(double tlim) { /* PLNR is not necessarily initialized when this function is called, so use X(the_planner)() */ X(the_planner)()->timelimit = tlim; } fftw-3.3.3/api/version.c0000644000175000017500000000243712054410050012006 00000000000000/* * Copyright (c) 2003, 2007-11 Matteo Frigo * Copyright (c) 2003, 2007-11 Massachusetts Institute of Technology * * This program is free software; you can redistribute it and/or modify * it under the terms of the GNU General Public License as published by * the Free Software Foundation; either version 2 of the License, or * (at your option) any later version. * * This program is distributed in the hope that it will be useful, * but WITHOUT ANY WARRANTY; without even the implied warranty of * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the * GNU General Public License for more details. * * You should have received a copy of the GNU General Public License * along with this program; if not, write to the Free Software * Foundation, Inc., 51 Franklin Street, Fifth Floor, Boston, MA 02110-1301 USA * */ #include "api.h" const char X(cc)[] = FFTW_CC; /* fftw <= 3.2.2 had special compiler flags for codelets, which are not used anymore. We keep this variable around because it is part of the ABI */ const char X(codelet_optim)[] = ""; const char X(version)[] = PACKAGE "-" PACKAGE_VERSION #if HAVE_FMA "-fma" #endif #if HAVE_SSE2 "-sse2" #endif #if HAVE_AVX "-avx" #endif #if HAVE_ALTIVEC "-altivec" #endif #if HAVE_NEON "-neon" #endif ; fftw-3.3.3/api/api.h0000644000175000017500000000724712054410050011103 00000000000000/* * Copyright (c) 2003, 2007-11 Matteo Frigo * Copyright (c) 2003, 2007-11 Massachusetts Institute of Technology * * This program is free software; you can redistribute it and/or modify * it under the terms of the GNU General Public License as published by * the Free Software Foundation; either version 2 of the License, or * (at your option) any later version. * * This program is distributed in the hope that it will be useful, * but WITHOUT ANY WARRANTY; without even the implied warranty of * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the * GNU General Public License for more details. * * You should have received a copy of the GNU General Public License * along with this program; if not, write to the Free Software * Foundation, Inc., 51 Franklin Street, Fifth Floor, Boston, MA 02110-1301 USA * */ /* internal API definitions */ #ifndef __API_H__ #define __API_H__ #ifndef CALLING_FFTW /* defined in hook.c, when calling internal functions */ # define COMPILING_FFTW /* used for DLL symbol exporting in fftw3.h */ #endif /* When compiling with GNU libtool on Windows, DLL_EXPORT is #defined for compiling the shared-library code. In this case, we'll #define FFTW_DLL to add dllexport attributes to the specified functions in fftw3.h. If we don't specify dllexport explicitly, then libtool automatically exports all symbols. However, if we specify dllexport explicitly for any functions, then libtool apparently doesn't do any automatic exporting. (Not documented, grrr, but this is the observed behavior with libtool 1.5.8.) Thus, using this forces us to correctly dllexport every exported symbol, or linking bench.exe will fail. This has the advantage of forcing us to mark things correctly, which is necessary for other compilers (such as MS VC++). */ #ifdef DLL_EXPORT # define FFTW_DLL #endif /* just in case: force not to use C99 complex numbers (we need this for IBM xlc because _Complex_I is treated specially and is defined even if is not included) */ #define FFTW_NO_Complex #include "fftw3.h" #include "ifftw.h" #include "rdft.h" #ifdef __cplusplus extern "C" { #endif /* __cplusplus */ /* the API ``plan'' contains both the kernel plan and problem */ struct X(plan_s) { plan *pln; problem *prb; int sign; }; /* shorthand */ typedef struct X(plan_s) apiplan; /* complex type for internal use */ typedef R C[2]; #define EXTRACT_REIM(sign, c, r, i) X(extract_reim)(sign, (c)[0], r, i) #define TAINT_UNALIGNED(p, flg) TAINT(p, ((flg) & FFTW_UNALIGNED) != 0) tensor *X(mktensor_rowmajor)(int rnk, const int *n, const int *niphys, const int *nophys, int is, int os); tensor *X(mktensor_iodims)(int rank, const X(iodim) *dims, int is, int os); tensor *X(mktensor_iodims64)(int rank, const X(iodim64) *dims, int is, int os); const int *X(rdft2_pad)(int rnk, const int *n, const int *nembed, int inplace, int cmplx, int **nfree); int X(many_kosherp)(int rnk, const int *n, int howmany); int X(guru_kosherp)(int rank, const X(iodim) *dims, int howmany_rank, const X(iodim) *howmany_dims); int X(guru64_kosherp)(int rank, const X(iodim64) *dims, int howmany_rank, const X(iodim64) *howmany_dims); /* Note: FFTW_EXTERN is used for "internal" functions used in tests/hook.c */ FFTW_EXTERN printer *X(mkprinter_file)(FILE *f); FFTW_EXTERN planner *X(the_planner)(void); void X(configure_planner)(planner *plnr); void X(mapflags)(planner *, unsigned); apiplan *X(mkapiplan)(int sign, unsigned flags, problem *prb); rdft_kind *X(map_r2r_kind)(int rank, const X(r2r_kind) * kind); #ifdef __cplusplus } /* extern "C" */ #endif /* __cplusplus */ #endif /* __API_H__ */ fftw-3.3.3/api/f77funcs.h0000644000175000017500000003462012054410050011767 00000000000000/* * Copyright (c) 2003, 2007-11 Matteo Frigo * Copyright (c) 2003, 2007-11 Massachusetts Institute of Technology * * This program is free software; you can redistribute it and/or modify * it under the terms of the GNU General Public License as published by * the Free Software Foundation; either version 2 of the License, or * (at your option) any later version. * * This program is distributed in the hope that it will be useful, * but WITHOUT ANY WARRANTY; without even the implied warranty of * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the * GNU General Public License for more details. * * You should have received a copy of the GNU General Public License * along with this program; if not, write to the Free Software * Foundation, Inc., 51 Franklin Street, Fifth Floor, Boston, MA 02110-1301 USA * */ /* Functions in the FFTW Fortran API, mangled according to the F77(...) macro. This file is designed to be #included by f77api.c, possibly multiple times in order to support multiple compiler manglings (via redefinition of F77). */ FFTW_VOIDFUNC F77(execute, EXECUTE)(X(plan) * const p) { plan *pln = (*p)->pln; pln->adt->solve(pln, (*p)->prb); } FFTW_VOIDFUNC F77(destroy_plan, DESTROY_PLAN)(X(plan) *p) { X(destroy_plan)(*p); } FFTW_VOIDFUNC F77(cleanup, CLEANUP)(void) { X(cleanup)(); } FFTW_VOIDFUNC F77(forget_wisdom, FORGET_WISDOM)(void) { X(forget_wisdom)(); } FFTW_VOIDFUNC F77(export_wisdom, EXPORT_WISDOM)(void (*f77_write_char)(char *, void *), void *data) { write_char_data ad; ad.f77_write_char = f77_write_char; ad.data = data; X(export_wisdom)(write_char, (void *) &ad); } FFTW_VOIDFUNC F77(import_wisdom, IMPORT_WISDOM)(int *isuccess, void (*f77_read_char)(int *, void *), void *data) { read_char_data ed; ed.f77_read_char = f77_read_char; ed.data = data; *isuccess = X(import_wisdom)(read_char, (void *) &ed); } FFTW_VOIDFUNC F77(import_system_wisdom, IMPORT_SYSTEM_WISDOM)(int *isuccess) { *isuccess = X(import_system_wisdom)(); } FFTW_VOIDFUNC F77(print_plan, PRINT_PLAN)(X(plan) * const p) { X(print_plan)(*p); fflush(stdout); } FFTW_VOIDFUNC F77(flops,FLOPS)(X(plan) *p, double *add, double *mul, double *fma) { X(flops)(*p, add, mul, fma); } FFTW_VOIDFUNC F77(estimate_cost,ESTIMATE_COST)(double *cost, X(plan) * const p) { *cost = X(estimate_cost)(*p); } FFTW_VOIDFUNC F77(cost,COST)(double *cost, X(plan) * const p) { *cost = X(cost)(*p); } FFTW_VOIDFUNC F77(set_timelimit,SET_TIMELIMIT)(double *t) { X(set_timelimit)(*t); } /******************************** DFT ***********************************/ FFTW_VOIDFUNC F77(plan_dft, PLAN_DFT)(X(plan) *p, int *rank, const int *n, C *in, C *out, int *sign, int *flags) { int *nrev = reverse_n(*rank, n); *p = X(plan_dft)(*rank, nrev, in, out, *sign, *flags); X(ifree0)(nrev); } FFTW_VOIDFUNC F77(plan_dft_1d, PLAN_DFT_1D)(X(plan) *p, int *n, C *in, C *out, int *sign, int *flags) { *p = X(plan_dft_1d)(*n, in, out, *sign, *flags); } FFTW_VOIDFUNC F77(plan_dft_2d, PLAN_DFT_2D)(X(plan) *p, int *nx, int *ny, C *in, C *out, int *sign, int *flags) { *p = X(plan_dft_2d)(*ny, *nx, in, out, *sign, *flags); } FFTW_VOIDFUNC F77(plan_dft_3d, PLAN_DFT_3D)(X(plan) *p, int *nx, int *ny, int *nz, C *in, C *out, int *sign, int *flags) { *p = X(plan_dft_3d)(*nz, *ny, *nx, in, out, *sign, *flags); } FFTW_VOIDFUNC F77(plan_many_dft, PLAN_MANY_DFT)(X(plan) *p, int *rank, const int *n, int *howmany, C *in, const int *inembed, int *istride, int *idist, C *out, const int *onembed, int *ostride, int *odist, int *sign, int *flags) { int *nrev = reverse_n(*rank, n); int *inembedrev = reverse_n(*rank, inembed); int *onembedrev = reverse_n(*rank, onembed); *p = X(plan_many_dft)(*rank, nrev, *howmany, in, inembedrev, *istride, *idist, out, onembedrev, *ostride, *odist, *sign, *flags); X(ifree0)(onembedrev); X(ifree0)(inembedrev); X(ifree0)(nrev); } FFTW_VOIDFUNC F77(plan_guru_dft, PLAN_GURU_DFT)(X(plan) *p, int *rank, const int *n, const int *is, const int *os, int *howmany_rank, const int *h_n, const int *h_is, const int *h_os, C *in, C *out, int *sign, int *flags) { X(iodim) *dims = make_dims(*rank, n, is, os); X(iodim) *howmany_dims = make_dims(*howmany_rank, h_n, h_is, h_os); *p = X(plan_guru_dft)(*rank, dims, *howmany_rank, howmany_dims, in, out, *sign, *flags); X(ifree0)(howmany_dims); X(ifree0)(dims); } FFTW_VOIDFUNC F77(plan_guru_split_dft, PLAN_GURU_SPLIT_DFT)(X(plan) *p, int *rank, const int *n, const int *is, const int *os, int *howmany_rank, const int *h_n, const int *h_is, const int *h_os, R *ri, R *ii, R *ro, R *io, int *flags) { X(iodim) *dims = make_dims(*rank, n, is, os); X(iodim) *howmany_dims = make_dims(*howmany_rank, h_n, h_is, h_os); *p = X(plan_guru_split_dft)(*rank, dims, *howmany_rank, howmany_dims, ri, ii, ro, io, *flags); X(ifree0)(howmany_dims); X(ifree0)(dims); } FFTW_VOIDFUNC F77(execute_dft, EXECUTE_DFT)(X(plan) * const p, C *in, C *out) { plan_dft *pln = (plan_dft *) (*p)->pln; if ((*p)->sign == FFT_SIGN) pln->apply((plan *) pln, in[0], in[0]+1, out[0], out[0]+1); else pln->apply((plan *) pln, in[0]+1, in[0], out[0]+1, out[0]); } FFTW_VOIDFUNC F77(execute_split_dft, EXECUTE_SPLIT_DFT)(X(plan) * const p, R *ri, R *ii, R *ro, R *io) { plan_dft *pln = (plan_dft *) (*p)->pln; pln->apply((plan *) pln, ri, ii, ro, io); } /****************************** DFT r2c *********************************/ FFTW_VOIDFUNC F77(plan_dft_r2c, PLAN_DFT_R2C)(X(plan) *p, int *rank, const int *n, R *in, C *out, int *flags) { int *nrev = reverse_n(*rank, n); *p = X(plan_dft_r2c)(*rank, nrev, in, out, *flags); X(ifree0)(nrev); } FFTW_VOIDFUNC F77(plan_dft_r2c_1d, PLAN_DFT_R2C_1D)(X(plan) *p, int *n, R *in, C *out, int *flags) { *p = X(plan_dft_r2c_1d)(*n, in, out, *flags); } FFTW_VOIDFUNC F77(plan_dft_r2c_2d, PLAN_DFT_R2C_2D)(X(plan) *p, int *nx, int *ny, R *in, C *out, int *flags) { *p = X(plan_dft_r2c_2d)(*ny, *nx, in, out, *flags); } FFTW_VOIDFUNC F77(plan_dft_r2c_3d, PLAN_DFT_R2C_3D)(X(plan) *p, int *nx, int *ny, int *nz, R *in, C *out, int *flags) { *p = X(plan_dft_r2c_3d)(*nz, *ny, *nx, in, out, *flags); } FFTW_VOIDFUNC F77(plan_many_dft_r2c, PLAN_MANY_DFT_R2C)( X(plan) *p, int *rank, const int *n, int *howmany, R *in, const int *inembed, int *istride, int *idist, C *out, const int *onembed, int *ostride, int *odist, int *flags) { int *nrev = reverse_n(*rank, n); int *inembedrev = reverse_n(*rank, inembed); int *onembedrev = reverse_n(*rank, onembed); *p = X(plan_many_dft_r2c)(*rank, nrev, *howmany, in, inembedrev, *istride, *idist, out, onembedrev, *ostride, *odist, *flags); X(ifree0)(onembedrev); X(ifree0)(inembedrev); X(ifree0)(nrev); } FFTW_VOIDFUNC F77(plan_guru_dft_r2c, PLAN_GURU_DFT_R2C)( X(plan) *p, int *rank, const int *n, const int *is, const int *os, int *howmany_rank, const int *h_n, const int *h_is, const int *h_os, R *in, C *out, int *flags) { X(iodim) *dims = make_dims(*rank, n, is, os); X(iodim) *howmany_dims = make_dims(*howmany_rank, h_n, h_is, h_os); *p = X(plan_guru_dft_r2c)(*rank, dims, *howmany_rank, howmany_dims, in, out, *flags); X(ifree0)(howmany_dims); X(ifree0)(dims); } FFTW_VOIDFUNC F77(plan_guru_split_dft_r2c, PLAN_GURU_SPLIT_DFT_R2C)( X(plan) *p, int *rank, const int *n, const int *is, const int *os, int *howmany_rank, const int *h_n, const int *h_is, const int *h_os, R *in, R *ro, R *io, int *flags) { X(iodim) *dims = make_dims(*rank, n, is, os); X(iodim) *howmany_dims = make_dims(*howmany_rank, h_n, h_is, h_os); *p = X(plan_guru_split_dft_r2c)(*rank, dims, *howmany_rank, howmany_dims, in, ro, io, *flags); X(ifree0)(howmany_dims); X(ifree0)(dims); } FFTW_VOIDFUNC F77(execute_dft_r2c, EXECUTE_DFT_R2C)(X(plan) * const p, R *in, C *out) { plan_rdft2 *pln = (plan_rdft2 *) (*p)->pln; problem_rdft2 *prb = (problem_rdft2 *) (*p)->prb; pln->apply((plan *) pln, in, in + (prb->r1 - prb->r0), out[0], out[0]+1); } FFTW_VOIDFUNC F77(execute_split_dft_r2c, EXECUTE_SPLIT_DFT_R2C)(X(plan) * const p, R *in, R *ro, R *io) { plan_rdft2 *pln = (plan_rdft2 *) (*p)->pln; problem_rdft2 *prb = (problem_rdft2 *) (*p)->prb; pln->apply((plan *) pln, in, in + (prb->r1 - prb->r0), ro, io); } /****************************** DFT c2r *********************************/ FFTW_VOIDFUNC F77(plan_dft_c2r, PLAN_DFT_C2R)(X(plan) *p, int *rank, const int *n, C *in, R *out, int *flags) { int *nrev = reverse_n(*rank, n); *p = X(plan_dft_c2r)(*rank, nrev, in, out, *flags); X(ifree0)(nrev); } FFTW_VOIDFUNC F77(plan_dft_c2r_1d, PLAN_DFT_C2R_1D)(X(plan) *p, int *n, C *in, R *out, int *flags) { *p = X(plan_dft_c2r_1d)(*n, in, out, *flags); } FFTW_VOIDFUNC F77(plan_dft_c2r_2d, PLAN_DFT_C2R_2D)(X(plan) *p, int *nx, int *ny, C *in, R *out, int *flags) { *p = X(plan_dft_c2r_2d)(*ny, *nx, in, out, *flags); } FFTW_VOIDFUNC F77(plan_dft_c2r_3d, PLAN_DFT_C2R_3D)(X(plan) *p, int *nx, int *ny, int *nz, C *in, R *out, int *flags) { *p = X(plan_dft_c2r_3d)(*nz, *ny, *nx, in, out, *flags); } FFTW_VOIDFUNC F77(plan_many_dft_c2r, PLAN_MANY_DFT_C2R)( X(plan) *p, int *rank, const int *n, int *howmany, C *in, const int *inembed, int *istride, int *idist, R *out, const int *onembed, int *ostride, int *odist, int *flags) { int *nrev = reverse_n(*rank, n); int *inembedrev = reverse_n(*rank, inembed); int *onembedrev = reverse_n(*rank, onembed); *p = X(plan_many_dft_c2r)(*rank, nrev, *howmany, in, inembedrev, *istride, *idist, out, onembedrev, *ostride, *odist, *flags); X(ifree0)(onembedrev); X(ifree0)(inembedrev); X(ifree0)(nrev); } FFTW_VOIDFUNC F77(plan_guru_dft_c2r, PLAN_GURU_DFT_C2R)( X(plan) *p, int *rank, const int *n, const int *is, const int *os, int *howmany_rank, const int *h_n, const int *h_is, const int *h_os, C *in, R *out, int *flags) { X(iodim) *dims = make_dims(*rank, n, is, os); X(iodim) *howmany_dims = make_dims(*howmany_rank, h_n, h_is, h_os); *p = X(plan_guru_dft_c2r)(*rank, dims, *howmany_rank, howmany_dims, in, out, *flags); X(ifree0)(howmany_dims); X(ifree0)(dims); } FFTW_VOIDFUNC F77(plan_guru_split_dft_c2r, PLAN_GURU_SPLIT_DFT_C2R)( X(plan) *p, int *rank, const int *n, const int *is, const int *os, int *howmany_rank, const int *h_n, const int *h_is, const int *h_os, R *ri, R *ii, R *out, int *flags) { X(iodim) *dims = make_dims(*rank, n, is, os); X(iodim) *howmany_dims = make_dims(*howmany_rank, h_n, h_is, h_os); *p = X(plan_guru_split_dft_c2r)(*rank, dims, *howmany_rank, howmany_dims, ri, ii, out, *flags); X(ifree0)(howmany_dims); X(ifree0)(dims); } FFTW_VOIDFUNC F77(execute_dft_c2r, EXECUTE_DFT_C2R)(X(plan) * const p, C *in, R *out) { plan_rdft2 *pln = (plan_rdft2 *) (*p)->pln; problem_rdft2 *prb = (problem_rdft2 *) (*p)->prb; pln->apply((plan *) pln, out, out + (prb->r1 - prb->r0), in[0], in[0]+1); } FFTW_VOIDFUNC F77(execute_split_dft_c2r, EXECUTE_SPLIT_DFT_C2R)(X(plan) * const p, R *ri, R *ii, R *out) { plan_rdft2 *pln = (plan_rdft2 *) (*p)->pln; problem_rdft2 *prb = (problem_rdft2 *) (*p)->prb; pln->apply((plan *) pln, out, out + (prb->r1 - prb->r0), ri, ii); } /****************************** r2r *********************************/ FFTW_VOIDFUNC F77(plan_r2r, PLAN_R2R)(X(plan) *p, int *rank, const int *n, R *in, R *out, int *kind, int *flags) { int *nrev = reverse_n(*rank, n); X(r2r_kind) *k = ints2kinds(*rank, kind); *p = X(plan_r2r)(*rank, nrev, in, out, k, *flags); X(ifree0)(k); X(ifree0)(nrev); } FFTW_VOIDFUNC F77(plan_r2r_1d, PLAN_R2R_1D)(X(plan) *p, int *n, R *in, R *out, int *kind, int *flags) { *p = X(plan_r2r_1d)(*n, in, out, (X(r2r_kind)) *kind, *flags); } FFTW_VOIDFUNC F77(plan_r2r_2d, PLAN_R2R_2D)(X(plan) *p, int *nx, int *ny, R *in, R *out, int *kindx, int *kindy, int *flags) { *p = X(plan_r2r_2d)(*ny, *nx, in, out, (X(r2r_kind)) *kindy, (X(r2r_kind)) *kindx, *flags); } FFTW_VOIDFUNC F77(plan_r2r_3d, PLAN_R2R_3D)(X(plan) *p, int *nx, int *ny, int *nz, R *in, R *out, int *kindx, int *kindy, int *kindz, int *flags) { *p = X(plan_r2r_3d)(*nz, *ny, *nx, in, out, (X(r2r_kind)) *kindz, (X(r2r_kind)) *kindy, (X(r2r_kind)) *kindx, *flags); } FFTW_VOIDFUNC F77(plan_many_r2r, PLAN_MANY_R2R)( X(plan) *p, int *rank, const int *n, int *howmany, R *in, const int *inembed, int *istride, int *idist, R *out, const int *onembed, int *ostride, int *odist, int *kind, int *flags) { int *nrev = reverse_n(*rank, n); int *inembedrev = reverse_n(*rank, inembed); int *onembedrev = reverse_n(*rank, onembed); X(r2r_kind) *k = ints2kinds(*rank, kind); *p = X(plan_many_r2r)(*rank, nrev, *howmany, in, inembedrev, *istride, *idist, out, onembedrev, *ostride, *odist, k, *flags); X(ifree0)(k); X(ifree0)(onembedrev); X(ifree0)(inembedrev); X(ifree0)(nrev); } FFTW_VOIDFUNC F77(plan_guru_r2r, PLAN_GURU_R2R)( X(plan) *p, int *rank, const int *n, const int *is, const int *os, int *howmany_rank, const int *h_n, const int *h_is, const int *h_os, R *in, R *out, int *kind, int *flags) { X(iodim) *dims = make_dims(*rank, n, is, os); X(iodim) *howmany_dims = make_dims(*howmany_rank, h_n, h_is, h_os); X(r2r_kind) *k = ints2kinds(*rank, kind); *p = X(plan_guru_r2r)(*rank, dims, *howmany_rank, howmany_dims, in, out, k, *flags); X(ifree0)(k); X(ifree0)(howmany_dims); X(ifree0)(dims); } FFTW_VOIDFUNC F77(execute_r2r, EXECUTE_R2R)(X(plan) * const p, R *in, R *out) { plan_rdft *pln = (plan_rdft *) (*p)->pln; pln->apply((plan *) pln, in, out); } fftw-3.3.3/api/x77.h0000644000175000017500000000515012054410050010746 00000000000000/* * Copyright (c) 2003, 2007-11 Matteo Frigo * Copyright (c) 2003, 2007-11 Massachusetts Institute of Technology * * This program is free software; you can redistribute it and/or modify * it under the terms of the GNU General Public License as published by * the Free Software Foundation; either version 2 of the License, or * (at your option) any later version. * * This program is distributed in the hope that it will be useful, * but WITHOUT ANY WARRANTY; without even the implied warranty of * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the * GNU General Public License for more details. * * You should have received a copy of the GNU General Public License * along with this program; if not, write to the Free Software * Foundation, Inc., 51 Franklin Street, Fifth Floor, Boston, MA 02110-1301 USA * */ /* Fortran-like (e.g. as in BLAS) type prefixes for F77 interface */ #if defined(FFTW_SINGLE) # define x77(name) CONCAT(sfftw_, name) # define X77(NAME) CONCAT(SFFTW_, NAME) #elif defined(FFTW_LDOUBLE) /* FIXME: what is best? BLAS uses D..._X, apparently. Ugh. */ # define x77(name) CONCAT(lfftw_, name) # define X77(NAME) CONCAT(LFFTW_, NAME) #elif defined(FFTW_QUAD) # define x77(name) CONCAT(qfftw_, name) # define X77(NAME) CONCAT(QFFTW_, NAME) #else # define x77(name) CONCAT(dfftw_, name) # define X77(NAME) CONCAT(DFFTW_, NAME) #endif /* If F77_FUNC is not defined and the user didn't explicitly specify --disable-fortran, then make our best guess at default wrappers (since F77_FUNC_EQUIV should not be defined in this case, we will use both double-underscored g77 wrappers and single- or non-underscored wrappers). This saves us from dealing with complaints in the cases where the user failed to specify an F77 compiler or wrapper detection failed for some reason. */ #if !defined(F77_FUNC) && !defined(DISABLE_FORTRAN) # if (defined(_WIN32) || defined(__WIN32__)) && !defined(WINDOWS_F77_MANGLING) # define WINDOWS_F77_MANGLING 1 # endif # if defined(_AIX) || defined(__hpux) || defined(hpux) # define F77_FUNC(a, A) a # elif defined(CRAY) || defined(_CRAY) || defined(_UNICOS) # define F77_FUNC(a, A) A # else # define F77_FUNC(a, A) a ## _ # endif # define F77_FUNC_(a, A) a ## __ #endif #if defined(WITH_G77_WRAPPERS) && !defined(DISABLE_FORTRAN) # undef F77_FUNC_ # define F77_FUNC_(a, A) a ## __ # undef F77_FUNC_EQUIV #endif /* annoying Windows syntax for shared-library declarations */ #if defined(FFTW_DLL) && (defined(_WIN32) || defined(__WIN32__)) # define FFTW_VOIDFUNC __declspec(dllexport) void #else # define FFTW_VOIDFUNC void #endif fftw-3.3.3/api/guru.h0000644000175000017500000000022112054410050011275 00000000000000#define XGURU(name) X(plan_guru_ ## name) #define IODIM X(iodim) #define MKTENSOR_IODIMS X(mktensor_iodims) #define GURU_KOSHERP X(guru_kosherp) fftw-3.3.3/api/guru64.h0000644000175000017500000000023112054410050011450 00000000000000#define XGURU(name) X(plan_guru64_ ## name) #define IODIM X(iodim64) #define MKTENSOR_IODIMS X(mktensor_iodims64) #define GURU_KOSHERP X(guru64_kosherp) fftw-3.3.3/api/mktensor-iodims.h0000644000175000017500000000342512054410050013450 00000000000000/* * Copyright (c) 2003, 2007-11 Matteo Frigo * Copyright (c) 2003, 2007-11 Massachusetts Institute of Technology * * This program is free software; you can redistribute it and/or modify * it under the terms of the GNU General Public License as published by * the Free Software Foundation; either version 2 of the License, or * (at your option) any later version. * * This program is distributed in the hope that it will be useful, * but WITHOUT ANY WARRANTY; without even the implied warranty of * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the * GNU General Public License for more details. * * You should have received a copy of the GNU General Public License * along with this program; if not, write to the Free Software * Foundation, Inc., 51 Franklin Street, Fifth Floor, Boston, MA 02110-1301 USA * */ #include "api.h" tensor *MKTENSOR_IODIMS(int rank, const IODIM *dims, int is, int os) { int i; tensor *x = X(mktensor)(rank); if (FINITE_RNK(rank)) { for (i = 0; i < rank; ++i) { x->dims[i].n = dims[i].n; x->dims[i].is = dims[i].is * is; x->dims[i].os = dims[i].os * os; } } return x; } static int iodims_kosherp(int rank, const IODIM *dims, int allow_minfty) { int i; if (rank < 0) return 0; if (allow_minfty) { if (!FINITE_RNK(rank)) return 1; for (i = 0; i < rank; ++i) if (dims[i].n < 0) return 0; } else { if (!FINITE_RNK(rank)) return 0; for (i = 0; i < rank; ++i) if (dims[i].n <= 0) return 0; } return 1; } int GURU_KOSHERP(int rank, const IODIM *dims, int howmany_rank, const IODIM *howmany_dims) { return (iodims_kosherp(rank, dims, 0) && iodims_kosherp(howmany_rank, howmany_dims, 1)); } fftw-3.3.3/api/plan-guru-dft-c2r.h0000644000175000017500000000275212054410050013477 00000000000000/* * Copyright (c) 2003, 2007-11 Matteo Frigo * Copyright (c) 2003, 2007-11 Massachusetts Institute of Technology * * This program is free software; you can redistribute it and/or modify * it under the terms of the GNU General Public License as published by * the Free Software Foundation; either version 2 of the License, or * (at your option) any later version. * * This program is distributed in the hope that it will be useful, * but WITHOUT ANY WARRANTY; without even the implied warranty of * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the * GNU General Public License for more details. * * You should have received a copy of the GNU General Public License * along with this program; if not, write to the Free Software * Foundation, Inc., 51 Franklin Street, Fifth Floor, Boston, MA 02110-1301 USA * */ #include "api.h" #include "rdft.h" X(plan) XGURU(dft_c2r)(int rank, const IODIM *dims, int howmany_rank, const IODIM *howmany_dims, C *in, R *out, unsigned flags) { R *ri, *ii; if (!GURU_KOSHERP(rank, dims, howmany_rank, howmany_dims)) return 0; EXTRACT_REIM(FFT_SIGN, in, &ri, &ii); if (out != ri) flags |= FFTW_DESTROY_INPUT; return X(mkapiplan)( 0, flags, X(mkproblem_rdft2_d_3pointers)( MKTENSOR_IODIMS(rank, dims, 2, 1), MKTENSOR_IODIMS(howmany_rank, howmany_dims, 2, 1), TAINT_UNALIGNED(out, flags), TAINT_UNALIGNED(ri, flags), TAINT_UNALIGNED(ii, flags), HC2R)); } fftw-3.3.3/api/plan-guru-dft-r2c.h0000644000175000017500000000267612054410050013504 00000000000000/* * Copyright (c) 2003, 2007-11 Matteo Frigo * Copyright (c) 2003, 2007-11 Massachusetts Institute of Technology * * This program is free software; you can redistribute it and/or modify * it under the terms of the GNU General Public License as published by * the Free Software Foundation; either version 2 of the License, or * (at your option) any later version. * * This program is distributed in the hope that it will be useful, * but WITHOUT ANY WARRANTY; without even the implied warranty of * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the * GNU General Public License for more details. * * You should have received a copy of the GNU General Public License * along with this program; if not, write to the Free Software * Foundation, Inc., 51 Franklin Street, Fifth Floor, Boston, MA 02110-1301 USA * */ #include "api.h" #include "rdft.h" X(plan) XGURU(dft_r2c)(int rank, const IODIM *dims, int howmany_rank, const IODIM *howmany_dims, R *in, C *out, unsigned flags) { R *ro, *io; if (!GURU_KOSHERP(rank, dims, howmany_rank, howmany_dims)) return 0; EXTRACT_REIM(FFT_SIGN, out, &ro, &io); return X(mkapiplan)( 0, flags, X(mkproblem_rdft2_d_3pointers)( MKTENSOR_IODIMS(rank, dims, 1, 2), MKTENSOR_IODIMS(howmany_rank, howmany_dims, 1, 2), TAINT_UNALIGNED(in, flags), TAINT_UNALIGNED(ro, flags), TAINT_UNALIGNED(io, flags), R2HC)); } fftw-3.3.3/api/plan-guru-dft.h0000644000175000017500000000276012054410050013012 00000000000000/* * Copyright (c) 2003, 2007-11 Matteo Frigo * Copyright (c) 2003, 2007-11 Massachusetts Institute of Technology * * This program is free software; you can redistribute it and/or modify * it under the terms of the GNU General Public License as published by * the Free Software Foundation; either version 2 of the License, or * (at your option) any later version. * * This program is distributed in the hope that it will be useful, * but WITHOUT ANY WARRANTY; without even the implied warranty of * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the * GNU General Public License for more details. * * You should have received a copy of the GNU General Public License * along with this program; if not, write to the Free Software * Foundation, Inc., 51 Franklin Street, Fifth Floor, Boston, MA 02110-1301 USA * */ #include "api.h" #include "dft.h" X(plan) XGURU(dft)(int rank, const IODIM *dims, int howmany_rank, const IODIM *howmany_dims, C *in, C *out, int sign, unsigned flags) { R *ri, *ii, *ro, *io; if (!GURU_KOSHERP(rank, dims, howmany_rank, howmany_dims)) return 0; EXTRACT_REIM(sign, in, &ri, &ii); EXTRACT_REIM(sign, out, &ro, &io); return X(mkapiplan)( sign, flags, X(mkproblem_dft_d)(MKTENSOR_IODIMS(rank, dims, 2, 2), MKTENSOR_IODIMS(howmany_rank, howmany_dims, 2, 2), TAINT_UNALIGNED(ri, flags), TAINT_UNALIGNED(ii, flags), TAINT_UNALIGNED(ro, flags), TAINT_UNALIGNED(io, flags))); } fftw-3.3.3/api/plan-guru-r2r.h0000644000175000017500000000270412054410050012740 00000000000000/* * Copyright (c) 2003, 2007-11 Matteo Frigo * Copyright (c) 2003, 2007-11 Massachusetts Institute of Technology * * This program is free software; you can redistribute it and/or modify * it under the terms of the GNU General Public License as published by * the Free Software Foundation; either version 2 of the License, or * (at your option) any later version. * * This program is distributed in the hope that it will be useful, * but WITHOUT ANY WARRANTY; without even the implied warranty of * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the * GNU General Public License for more details. * * You should have received a copy of the GNU General Public License * along with this program; if not, write to the Free Software * Foundation, Inc., 51 Franklin Street, Fifth Floor, Boston, MA 02110-1301 USA * */ #include "api.h" #include "rdft.h" X(plan) XGURU(r2r)(int rank, const IODIM *dims, int howmany_rank, const IODIM *howmany_dims, R *in, R *out, const X(r2r_kind) * kind, unsigned flags) { X(plan) p; rdft_kind *k; if (!GURU_KOSHERP(rank, dims, howmany_rank, howmany_dims)) return 0; k = X(map_r2r_kind)(rank, kind); p = X(mkapiplan)( 0, flags, X(mkproblem_rdft_d)(MKTENSOR_IODIMS(rank, dims, 1, 1), MKTENSOR_IODIMS(howmany_rank, howmany_dims, 1, 1), TAINT_UNALIGNED(in, flags), TAINT_UNALIGNED(out, flags), k)); X(ifree0)(k); return p; } fftw-3.3.3/api/plan-guru-split-dft-c2r.h0000644000175000017500000000266712054410050014635 00000000000000/* * Copyright (c) 2003, 2007-11 Matteo Frigo * Copyright (c) 2003, 2007-11 Massachusetts Institute of Technology * * This program is free software; you can redistribute it and/or modify * it under the terms of the GNU General Public License as published by * the Free Software Foundation; either version 2 of the License, or * (at your option) any later version. * * This program is distributed in the hope that it will be useful, * but WITHOUT ANY WARRANTY; without even the implied warranty of * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the * GNU General Public License for more details. * * You should have received a copy of the GNU General Public License * along with this program; if not, write to the Free Software * Foundation, Inc., 51 Franklin Street, Fifth Floor, Boston, MA 02110-1301 USA * */ #include "api.h" #include "rdft.h" X(plan) XGURU(split_dft_c2r)(int rank, const IODIM *dims, int howmany_rank, const IODIM *howmany_dims, R *ri, R *ii, R *out, unsigned flags) { if (!GURU_KOSHERP(rank, dims, howmany_rank, howmany_dims)) return 0; if (out != ri) flags |= FFTW_DESTROY_INPUT; return X(mkapiplan)( 0, flags, X(mkproblem_rdft2_d_3pointers)( MKTENSOR_IODIMS(rank, dims, 1, 1), MKTENSOR_IODIMS(howmany_rank, howmany_dims, 1, 1), TAINT_UNALIGNED(out, flags), TAINT_UNALIGNED(ri, flags), TAINT_UNALIGNED(ii, flags), HC2R)); } fftw-3.3.3/api/plan-guru-split-dft-r2c.h0000644000175000017500000000261012054410050014621 00000000000000/* * Copyright (c) 2003, 2007-11 Matteo Frigo * Copyright (c) 2003, 2007-11 Massachusetts Institute of Technology * * This program is free software; you can redistribute it and/or modify * it under the terms of the GNU General Public License as published by * the Free Software Foundation; either version 2 of the License, or * (at your option) any later version. * * This program is distributed in the hope that it will be useful, * but WITHOUT ANY WARRANTY; without even the implied warranty of * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the * GNU General Public License for more details. * * You should have received a copy of the GNU General Public License * along with this program; if not, write to the Free Software * Foundation, Inc., 51 Franklin Street, Fifth Floor, Boston, MA 02110-1301 USA * */ #include "api.h" #include "rdft.h" X(plan) XGURU(split_dft_r2c)(int rank, const IODIM *dims, int howmany_rank, const IODIM *howmany_dims, R *in, R *ro, R *io, unsigned flags) { if (!GURU_KOSHERP(rank, dims, howmany_rank, howmany_dims)) return 0; return X(mkapiplan)( 0, flags, X(mkproblem_rdft2_d_3pointers)( MKTENSOR_IODIMS(rank, dims, 1, 1), MKTENSOR_IODIMS(howmany_rank, howmany_dims, 1, 1), TAINT_UNALIGNED(in, flags), TAINT_UNALIGNED(ro, flags), TAINT_UNALIGNED(io, flags), R2HC)); } fftw-3.3.3/api/plan-guru-split-dft.h0000644000175000017500000000271012054410050014136 00000000000000/* * Copyright (c) 2003, 2007-11 Matteo Frigo * Copyright (c) 2003, 2007-11 Massachusetts Institute of Technology * * This program is free software; you can redistribute it and/or modify * it under the terms of the GNU General Public License as published by * the Free Software Foundation; either version 2 of the License, or * (at your option) any later version. * * This program is distributed in the hope that it will be useful, * but WITHOUT ANY WARRANTY; without even the implied warranty of * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the * GNU General Public License for more details. * * You should have received a copy of the GNU General Public License * along with this program; if not, write to the Free Software * Foundation, Inc., 51 Franklin Street, Fifth Floor, Boston, MA 02110-1301 USA * */ #include "api.h" #include "dft.h" X(plan) XGURU(split_dft)(int rank, const IODIM *dims, int howmany_rank, const IODIM *howmany_dims, R *ri, R *ii, R *ro, R *io, unsigned flags) { if (!GURU_KOSHERP(rank, dims, howmany_rank, howmany_dims)) return 0; return X(mkapiplan)( ii - ri == 1 && io - ro == 1 ? FFT_SIGN : -FFT_SIGN, flags, X(mkproblem_dft_d)(MKTENSOR_IODIMS(rank, dims, 1, 1), MKTENSOR_IODIMS(howmany_rank, howmany_dims, 1, 1), TAINT_UNALIGNED(ri, flags), TAINT_UNALIGNED(ii, flags), TAINT_UNALIGNED(ro, flags), TAINT_UNALIGNED(io, flags))); } fftw-3.3.3/api/plan-guru64-dft-c2r.c0000644000175000017500000000006312054410050013635 00000000000000#include "guru64.h" #include "plan-guru-dft-c2r.h" fftw-3.3.3/api/plan-guru64-dft-r2c.c0000644000175000017500000000006312054410050013635 00000000000000#include "guru64.h" #include "plan-guru-dft-r2c.h" fftw-3.3.3/api/plan-guru64-dft.c0000644000175000017500000000005712054410050013154 00000000000000#include "guru64.h" #include "plan-guru-dft.h" fftw-3.3.3/api/plan-guru64-r2r.c0000644000175000017500000000005712054410050013104 00000000000000#include "guru64.h" #include "plan-guru-r2r.h" fftw-3.3.3/api/plan-guru64-split-dft-c2r.c0000644000175000017500000000007112054410050014765 00000000000000#include "guru64.h" #include "plan-guru-split-dft-c2r.h" fftw-3.3.3/api/plan-guru64-split-dft-r2c.c0000644000175000017500000000007112054410050014765 00000000000000#include "guru64.h" #include "plan-guru-split-dft-r2c.h" fftw-3.3.3/api/plan-guru64-split-dft.c0000644000175000017500000000006512054410050014304 00000000000000#include "guru64.h" #include "plan-guru-split-dft.h" fftw-3.3.3/api/mktensor-iodims64.c0000644000175000017500000000006112054410050013606 00000000000000#include "guru64.h" #include "mktensor-iodims.h" fftw-3.3.3/api/f03api.sh0000644000175000017500000000266212054410050011573 00000000000000#! /bin/sh # Script to generate Fortran 2003 interface declarations for FFTW from # the fftw3.h header file. # This is designed so that the Fortran caller can do: # use, intrinsic :: iso_c_binding # implicit none # include 'fftw3.f03' # and then call the C FFTW functions directly, with type checking. echo "! Generated automatically. DO NOT EDIT!" echo # C_FFTW_R2R_KIND is determined by configure and inserted by the Makefile # echo " integer, parameter :: C_FFTW_R2R_KIND = @C_FFTW_R2R_KIND@" # Extract constants perl -pe 's/([A-Z0-9_]+)=([+-]?[0-9]+)/\n integer\(C_INT\), parameter :: \1 = \2\n/g' < fftw3.h | grep 'integer(C_INT)' perl -pe 's/#define +([A-Z0-9_]+) +\(([+-]?[0-9]+)U?\)/\n integer\(C_INT\), parameter :: \1 = \2\n/g' < fftw3.h | grep 'integer(C_INT)' perl -pe 'if (/#define +([A-Z0-9_]+) +\(([0-9]+)U? *<< *([0-9]+)\)/) { print "\n integer\(C_INT\), parameter :: $1 = ",$2 << $3,"\n"; }' < fftw3.h | grep 'integer(C_INT)' # Extract function declarations for p in $*; do if test "$p" = "d"; then p=""; fi echo cat < (...args...) # extern (...args...) # ... # with no line breaks within a given function. (It's too much work to # write a general parser, since we just have to handle FFTW's header files.) sub canonicalize_type { my($type); ($type) = @_; $type =~ s/ +/ /g; $type =~ s/^ //; $type =~ s/ $//; $type =~ s/([^\* ])\*/$1 \*/g; return $type; } # C->Fortran map of supported return types %return_types = ( "int" => "integer(C_INT)", "ptrdiff_t" => "integer(C_INTPTR_T)", "size_t" => "integer(C_SIZE_T)", "double" => "real(C_DOUBLE)", "float" => "real(C_FLOAT)", "long double" => "real(C_LONG_DOUBLE)", "float128__" => "real(16)", "fftw_plan" => "type(C_PTR)", "fftwf_plan" => "type(C_PTR)", "fftwl_plan" => "type(C_PTR)", "fftwq_plan" => "type(C_PTR)", "void *" => "type(C_PTR)", "char *" => "type(C_PTR)", "double *" => "type(C_PTR)", "float *" => "type(C_PTR)", "long double *" => "type(C_PTR)", "float128__ *" => "type(C_PTR)", "fftw_complex *" => "type(C_PTR)", "fftwf_complex *" => "type(C_PTR)", "fftwl_complex *" => "type(C_PTR)", "fftwq_complex *" => "type(C_PTR)", ); # C->Fortran map of supported argument types %arg_types = ( "int" => "integer(C_INT), value", "unsigned" => "integer(C_INT), value", "size_t" => "integer(C_SIZE_T), value", "ptrdiff_t" => "integer(C_INTPTR_T), value", "fftw_r2r_kind" => "integer(C_FFTW_R2R_KIND), value", "fftwf_r2r_kind" => "integer(C_FFTW_R2R_KIND), value", "fftwl_r2r_kind" => "integer(C_FFTW_R2R_KIND), value", "fftwq_r2r_kind" => "integer(C_FFTW_R2R_KIND), value", "double" => "real(C_DOUBLE), value", "float" => "real(C_FLOAT), value", "long double" => "real(C_LONG_DOUBLE), value", "__float128" => "real(16), value", "fftw_complex" => "complex(C_DOUBLE_COMPLEX), value", "fftwf_complex" => "complex(C_DOUBLE_COMPLEX), value", "fftwl_complex" => "complex(C_LONG_DOUBLE), value", "fftwq_complex" => "complex(16), value", "fftw_plan" => "type(C_PTR), value", "fftwf_plan" => "type(C_PTR), value", "fftwl_plan" => "type(C_PTR), value", "fftwq_plan" => "type(C_PTR), value", "const fftw_plan" => "type(C_PTR), value", "const fftwf_plan" => "type(C_PTR), value", "const fftwl_plan" => "type(C_PTR), value", "const fftwq_plan" => "type(C_PTR), value", "const int *" => "integer(C_INT), dimension(*), intent(in)", "ptrdiff_t *" => "integer(C_INTPTR_T), intent(out)", "const ptrdiff_t *" => "integer(C_INTPTR_T), dimension(*), intent(in)", "const fftw_r2r_kind *" => "integer(C_FFTW_R2R_KIND), dimension(*), intent(in)", "const fftwf_r2r_kind *" => "integer(C_FFTW_R2R_KIND), dimension(*), intent(in)", "const fftwl_r2r_kind *" => "integer(C_FFTW_R2R_KIND), dimension(*), intent(in)", "const fftwq_r2r_kind *" => "integer(C_FFTW_R2R_KIND), dimension(*), intent(in)", "double *" => "real(C_DOUBLE), dimension(*), intent(out)", "float *" => "real(C_FLOAT), dimension(*), intent(out)", "long double *" => "real(C_LONG_DOUBLE), dimension(*), intent(out)", "__float128 *" => "real(16), dimension(*), intent(out)", "fftw_complex *" => "complex(C_DOUBLE_COMPLEX), dimension(*), intent(out)", "fftwf_complex *" => "complex(C_FLOAT_COMPLEX), dimension(*), intent(out)", "fftwl_complex *" => "complex(C_LONG_DOUBLE_COMPLEX), dimension(*), intent(out)", "fftwq_complex *" => "complex(16), dimension(*), intent(out)", "const fftw_iodim *" => "type(fftw_iodim), dimension(*), intent(in)", "const fftwf_iodim *" => "type(fftwf_iodim), dimension(*), intent(in)", "const fftwl_iodim *" => "type(fftwl_iodim), dimension(*), intent(in)", "const fftwq_iodim *" => "type(fftwq_iodim), dimension(*), intent(in)", "const fftw_iodim64 *" => "type(fftw_iodim64), dimension(*), intent(in)", "const fftwf_iodim64 *" => "type(fftwf_iodim64), dimension(*), intent(in)", "const fftwl_iodim64 *" => "type(fftwl_iodim64), dimension(*), intent(in)", "const fftwq_iodim64 *" => "type(fftwq_iodim64), dimension(*), intent(in)", "void *" => "type(C_PTR), value", "FILE *" => "type(C_PTR), value", "const char *" => "character(C_CHAR), dimension(*), intent(in)", "fftw_write_char_func" => "type(C_FUNPTR), value", "fftwf_write_char_func" => "type(C_FUNPTR), value", "fftwl_write_char_func" => "type(C_FUNPTR), value", "fftwq_write_char_func" => "type(C_FUNPTR), value", "fftw_read_char_func" => "type(C_FUNPTR), value", "fftwf_read_char_func" => "type(C_FUNPTR), value", "fftwl_read_char_func" => "type(C_FUNPTR), value", "fftwq_read_char_func" => "type(C_FUNPTR), value", # Although the MPI standard defines this type as simply "integer", # if we use integer without a 'C_' kind in a bind(C) interface then # gfortran complains. Instead, since MPI also requires the C type # MPI_Fint to match Fortran integers, we use the size of this type # (extracted by configure and substituted by the Makefile). "MPI_Comm" => "integer(C_MPI_FINT), value" ); while (<>) { next if /^ *$/; if (/^ *extern +([a-zA-Z_0-9 ]+[ \*]) *([a-zA-Z_0-9]+) *\((.*)\) *$/) { $ret = &canonicalize_type($1); $name = $2; $args = $3; $args =~ s/^ *void *$//; $bad = ($ret ne "void") && !exists($return_types{$ret}); foreach $arg (split(/ *, */, $args)) { $arg =~ /^([a-zA-Z_0-9 ]+[ \*]) *([a-zA-Z_0-9]+) *$/; $argtype = &canonicalize_type($1); $bad = 1 if !exists($arg_types{$argtype}); } if ($bad) { print "! Unable to generate Fortran interface for $name\n"; next; } # any function taking an MPI_Comm arg needs a C wrapper (grr). if ($args =~ /MPI_Comm/) { $cname = $name . "_f03"; } else { $cname = $name; } # Fortran has a 132-character line-length limit by default (grr) $len = 0; print " "; $len = $len + length(" "); if ($ret eq "void") { $kind = "subroutine" } else { print "$return_types{$ret} "; $len = $len + length("$return_types{$ret} "); $kind = "function" } print "$kind $name("; $len = $len + length("$kind $name("); $len0 = $len; $argnames = $args; $argnames =~ s/([a-zA-Z_0-9 ]+[ \*]) *([a-zA-Z_0-9]+) */$2/g; $comma = ""; foreach $argname (split(/ *, */, $argnames)) { if ($len + length("$comma$argname") + 3 > 132) { printf ", &\n%*s", $len0, ""; $len = $len0; $comma = ""; } print "$comma$argname"; $len = $len + length("$comma$argname"); $comma = ","; } print ") "; $len = $len + 2; if ($len + length("bind(C, name='$cname')") > 132) { printf "&\n%*s", $len0 - length("$name("), ""; } print "bind(C, name='$cname')\n"; print " import\n"; foreach $arg (split(/ *, */, $args)) { $arg =~ /^([a-zA-Z_0-9 ]+[ \*]) *([a-zA-Z_0-9]+) *$/; $argtype = &canonicalize_type($1); $argname = $2; $ftype = $arg_types{$argtype}; # Various special cases for argument types: if ($name =~ /_flops$/ && $argtype eq "double *") { $ftype = "real(C_DOUBLE), intent(out)" } if ($name =~ /_execute/ && ($argname eq "ri" || $argname eq "ii" || $argname eq "in")) { $ftype =~ s/intent\(out\)/intent(inout)/; } print " $ftype :: $argname\n" } print " end $kind $name\n"; print " \n"; } } fftw-3.3.3/api/fftw3.f03.in0000644000175000017500000015041512054411075012135 00000000000000! Generated automatically. DO NOT EDIT! integer(C_INT), parameter :: FFTW_R2HC = 0 integer(C_INT), parameter :: FFTW_HC2R = 1 integer(C_INT), parameter :: FFTW_DHT = 2 integer(C_INT), parameter :: FFTW_REDFT00 = 3 integer(C_INT), parameter :: FFTW_REDFT01 = 4 integer(C_INT), parameter :: FFTW_REDFT10 = 5 integer(C_INT), parameter :: FFTW_REDFT11 = 6 integer(C_INT), parameter :: FFTW_RODFT00 = 7 integer(C_INT), parameter :: FFTW_RODFT01 = 8 integer(C_INT), parameter :: FFTW_RODFT10 = 9 integer(C_INT), parameter :: FFTW_RODFT11 = 10 integer(C_INT), parameter :: FFTW_FORWARD = -1 integer(C_INT), parameter :: FFTW_BACKWARD = +1 integer(C_INT), parameter :: FFTW_MEASURE = 0 integer(C_INT), parameter :: FFTW_DESTROY_INPUT = 1 integer(C_INT), parameter :: FFTW_UNALIGNED = 2 integer(C_INT), parameter :: FFTW_CONSERVE_MEMORY = 4 integer(C_INT), parameter :: FFTW_EXHAUSTIVE = 8 integer(C_INT), parameter :: FFTW_PRESERVE_INPUT = 16 integer(C_INT), parameter :: FFTW_PATIENT = 32 integer(C_INT), parameter :: FFTW_ESTIMATE = 64 integer(C_INT), parameter :: FFTW_WISDOM_ONLY = 2097152 integer(C_INT), parameter :: FFTW_ESTIMATE_PATIENT = 128 integer(C_INT), parameter :: FFTW_BELIEVE_PCOST = 256 integer(C_INT), parameter :: FFTW_NO_DFT_R2HC = 512 integer(C_INT), parameter :: FFTW_NO_NONTHREADED = 1024 integer(C_INT), parameter :: FFTW_NO_BUFFERING = 2048 integer(C_INT), parameter :: FFTW_NO_INDIRECT_OP = 4096 integer(C_INT), parameter :: FFTW_ALLOW_LARGE_GENERIC = 8192 integer(C_INT), parameter :: FFTW_NO_RANK_SPLITS = 16384 integer(C_INT), parameter :: FFTW_NO_VRANK_SPLITS = 32768 integer(C_INT), parameter :: FFTW_NO_VRECURSE = 65536 integer(C_INT), parameter :: FFTW_NO_SIMD = 131072 integer(C_INT), parameter :: FFTW_NO_SLOW = 262144 integer(C_INT), parameter :: FFTW_NO_FIXED_RADIX_LARGE_N = 524288 integer(C_INT), parameter :: FFTW_ALLOW_PRUNING = 1048576 type, bind(C) :: fftw_iodim integer(C_INT) n, is, os end type fftw_iodim type, bind(C) :: fftw_iodim64 integer(C_INTPTR_T) n, is, os end type fftw_iodim64 interface type(C_PTR) function fftw_plan_dft(rank,n,in,out,sign,flags) bind(C, name='fftw_plan_dft') import integer(C_INT), value :: rank integer(C_INT), dimension(*), intent(in) :: n complex(C_DOUBLE_COMPLEX), dimension(*), intent(out) :: in complex(C_DOUBLE_COMPLEX), dimension(*), intent(out) :: out integer(C_INT), value :: sign integer(C_INT), value :: flags end function fftw_plan_dft type(C_PTR) function fftw_plan_dft_1d(n,in,out,sign,flags) bind(C, name='fftw_plan_dft_1d') import integer(C_INT), value :: n complex(C_DOUBLE_COMPLEX), dimension(*), intent(out) :: in complex(C_DOUBLE_COMPLEX), dimension(*), intent(out) :: out integer(C_INT), value :: sign integer(C_INT), value :: flags end function fftw_plan_dft_1d type(C_PTR) function fftw_plan_dft_2d(n0,n1,in,out,sign,flags) bind(C, name='fftw_plan_dft_2d') import integer(C_INT), value :: n0 integer(C_INT), value :: n1 complex(C_DOUBLE_COMPLEX), dimension(*), intent(out) :: in complex(C_DOUBLE_COMPLEX), dimension(*), intent(out) :: out integer(C_INT), value :: sign integer(C_INT), value :: flags end function fftw_plan_dft_2d type(C_PTR) function fftw_plan_dft_3d(n0,n1,n2,in,out,sign,flags) bind(C, name='fftw_plan_dft_3d') import integer(C_INT), value :: n0 integer(C_INT), value :: n1 integer(C_INT), value :: n2 complex(C_DOUBLE_COMPLEX), dimension(*), intent(out) :: in complex(C_DOUBLE_COMPLEX), dimension(*), intent(out) :: out integer(C_INT), value :: sign integer(C_INT), value :: flags end function fftw_plan_dft_3d type(C_PTR) function fftw_plan_many_dft(rank,n,howmany,in,inembed,istride,idist,out,onembed,ostride,odist,sign,flags) & bind(C, name='fftw_plan_many_dft') import integer(C_INT), value :: rank integer(C_INT), dimension(*), intent(in) :: n integer(C_INT), value :: howmany complex(C_DOUBLE_COMPLEX), dimension(*), intent(out) :: in integer(C_INT), dimension(*), intent(in) :: inembed integer(C_INT), value :: istride integer(C_INT), value :: idist complex(C_DOUBLE_COMPLEX), dimension(*), intent(out) :: out integer(C_INT), dimension(*), intent(in) :: onembed integer(C_INT), value :: ostride integer(C_INT), value :: odist integer(C_INT), value :: sign integer(C_INT), value :: flags end function fftw_plan_many_dft type(C_PTR) function fftw_plan_guru_dft(rank,dims,howmany_rank,howmany_dims,in,out,sign,flags) & bind(C, name='fftw_plan_guru_dft') import integer(C_INT), value :: rank type(fftw_iodim), dimension(*), intent(in) :: dims integer(C_INT), value :: howmany_rank type(fftw_iodim), dimension(*), intent(in) :: howmany_dims complex(C_DOUBLE_COMPLEX), dimension(*), intent(out) :: in complex(C_DOUBLE_COMPLEX), dimension(*), intent(out) :: out integer(C_INT), value :: sign integer(C_INT), value :: flags end function fftw_plan_guru_dft type(C_PTR) function fftw_plan_guru_split_dft(rank,dims,howmany_rank,howmany_dims,ri,ii,ro,io,flags) & bind(C, name='fftw_plan_guru_split_dft') import integer(C_INT), value :: rank type(fftw_iodim), dimension(*), intent(in) :: dims integer(C_INT), value :: howmany_rank type(fftw_iodim), dimension(*), intent(in) :: howmany_dims real(C_DOUBLE), dimension(*), intent(out) :: ri real(C_DOUBLE), dimension(*), intent(out) :: ii real(C_DOUBLE), dimension(*), intent(out) :: ro real(C_DOUBLE), dimension(*), intent(out) :: io integer(C_INT), value :: flags end function fftw_plan_guru_split_dft type(C_PTR) function fftw_plan_guru64_dft(rank,dims,howmany_rank,howmany_dims,in,out,sign,flags) & bind(C, name='fftw_plan_guru64_dft') import integer(C_INT), value :: rank type(fftw_iodim64), dimension(*), intent(in) :: dims integer(C_INT), value :: howmany_rank type(fftw_iodim64), dimension(*), intent(in) :: howmany_dims complex(C_DOUBLE_COMPLEX), dimension(*), intent(out) :: in complex(C_DOUBLE_COMPLEX), dimension(*), intent(out) :: out integer(C_INT), value :: sign integer(C_INT), value :: flags end function fftw_plan_guru64_dft type(C_PTR) function fftw_plan_guru64_split_dft(rank,dims,howmany_rank,howmany_dims,ri,ii,ro,io,flags) & bind(C, name='fftw_plan_guru64_split_dft') import integer(C_INT), value :: rank type(fftw_iodim64), dimension(*), intent(in) :: dims integer(C_INT), value :: howmany_rank type(fftw_iodim64), dimension(*), intent(in) :: howmany_dims real(C_DOUBLE), dimension(*), intent(out) :: ri real(C_DOUBLE), dimension(*), intent(out) :: ii real(C_DOUBLE), dimension(*), intent(out) :: ro real(C_DOUBLE), dimension(*), intent(out) :: io integer(C_INT), value :: flags end function fftw_plan_guru64_split_dft subroutine fftw_execute_dft(p,in,out) bind(C, name='fftw_execute_dft') import type(C_PTR), value :: p complex(C_DOUBLE_COMPLEX), dimension(*), intent(inout) :: in complex(C_DOUBLE_COMPLEX), dimension(*), intent(out) :: out end subroutine fftw_execute_dft subroutine fftw_execute_split_dft(p,ri,ii,ro,io) bind(C, name='fftw_execute_split_dft') import type(C_PTR), value :: p real(C_DOUBLE), dimension(*), intent(inout) :: ri real(C_DOUBLE), dimension(*), intent(inout) :: ii real(C_DOUBLE), dimension(*), intent(out) :: ro real(C_DOUBLE), dimension(*), intent(out) :: io end subroutine fftw_execute_split_dft type(C_PTR) function fftw_plan_many_dft_r2c(rank,n,howmany,in,inembed,istride,idist,out,onembed,ostride,odist,flags) & bind(C, name='fftw_plan_many_dft_r2c') import integer(C_INT), value :: rank integer(C_INT), dimension(*), intent(in) :: n integer(C_INT), value :: howmany real(C_DOUBLE), dimension(*), intent(out) :: in integer(C_INT), dimension(*), intent(in) :: inembed integer(C_INT), value :: istride integer(C_INT), value :: idist complex(C_DOUBLE_COMPLEX), dimension(*), intent(out) :: out integer(C_INT), dimension(*), intent(in) :: onembed integer(C_INT), value :: ostride integer(C_INT), value :: odist integer(C_INT), value :: flags end function fftw_plan_many_dft_r2c type(C_PTR) function fftw_plan_dft_r2c(rank,n,in,out,flags) bind(C, name='fftw_plan_dft_r2c') import integer(C_INT), value :: rank integer(C_INT), dimension(*), intent(in) :: n real(C_DOUBLE), dimension(*), intent(out) :: in complex(C_DOUBLE_COMPLEX), dimension(*), intent(out) :: out integer(C_INT), value :: flags end function fftw_plan_dft_r2c type(C_PTR) function fftw_plan_dft_r2c_1d(n,in,out,flags) bind(C, name='fftw_plan_dft_r2c_1d') import integer(C_INT), value :: n real(C_DOUBLE), dimension(*), intent(out) :: in complex(C_DOUBLE_COMPLEX), dimension(*), intent(out) :: out integer(C_INT), value :: flags end function fftw_plan_dft_r2c_1d type(C_PTR) function fftw_plan_dft_r2c_2d(n0,n1,in,out,flags) bind(C, name='fftw_plan_dft_r2c_2d') import integer(C_INT), value :: n0 integer(C_INT), value :: n1 real(C_DOUBLE), dimension(*), intent(out) :: in complex(C_DOUBLE_COMPLEX), dimension(*), intent(out) :: out integer(C_INT), value :: flags end function fftw_plan_dft_r2c_2d type(C_PTR) function fftw_plan_dft_r2c_3d(n0,n1,n2,in,out,flags) bind(C, name='fftw_plan_dft_r2c_3d') import integer(C_INT), value :: n0 integer(C_INT), value :: n1 integer(C_INT), value :: n2 real(C_DOUBLE), dimension(*), intent(out) :: in complex(C_DOUBLE_COMPLEX), dimension(*), intent(out) :: out integer(C_INT), value :: flags end function fftw_plan_dft_r2c_3d type(C_PTR) function fftw_plan_many_dft_c2r(rank,n,howmany,in,inembed,istride,idist,out,onembed,ostride,odist,flags) & bind(C, name='fftw_plan_many_dft_c2r') import integer(C_INT), value :: rank integer(C_INT), dimension(*), intent(in) :: n integer(C_INT), value :: howmany complex(C_DOUBLE_COMPLEX), dimension(*), intent(out) :: in integer(C_INT), dimension(*), intent(in) :: inembed integer(C_INT), value :: istride integer(C_INT), value :: idist real(C_DOUBLE), dimension(*), intent(out) :: out integer(C_INT), dimension(*), intent(in) :: onembed integer(C_INT), value :: ostride integer(C_INT), value :: odist integer(C_INT), value :: flags end function fftw_plan_many_dft_c2r type(C_PTR) function fftw_plan_dft_c2r(rank,n,in,out,flags) bind(C, name='fftw_plan_dft_c2r') import integer(C_INT), value :: rank integer(C_INT), dimension(*), intent(in) :: n complex(C_DOUBLE_COMPLEX), dimension(*), intent(out) :: in real(C_DOUBLE), dimension(*), intent(out) :: out integer(C_INT), value :: flags end function fftw_plan_dft_c2r type(C_PTR) function fftw_plan_dft_c2r_1d(n,in,out,flags) bind(C, name='fftw_plan_dft_c2r_1d') import integer(C_INT), value :: n complex(C_DOUBLE_COMPLEX), dimension(*), intent(out) :: in real(C_DOUBLE), dimension(*), intent(out) :: out integer(C_INT), value :: flags end function fftw_plan_dft_c2r_1d type(C_PTR) function fftw_plan_dft_c2r_2d(n0,n1,in,out,flags) bind(C, name='fftw_plan_dft_c2r_2d') import integer(C_INT), value :: n0 integer(C_INT), value :: n1 complex(C_DOUBLE_COMPLEX), dimension(*), intent(out) :: in real(C_DOUBLE), dimension(*), intent(out) :: out integer(C_INT), value :: flags end function fftw_plan_dft_c2r_2d type(C_PTR) function fftw_plan_dft_c2r_3d(n0,n1,n2,in,out,flags) bind(C, name='fftw_plan_dft_c2r_3d') import integer(C_INT), value :: n0 integer(C_INT), value :: n1 integer(C_INT), value :: n2 complex(C_DOUBLE_COMPLEX), dimension(*), intent(out) :: in real(C_DOUBLE), dimension(*), intent(out) :: out integer(C_INT), value :: flags end function fftw_plan_dft_c2r_3d type(C_PTR) function fftw_plan_guru_dft_r2c(rank,dims,howmany_rank,howmany_dims,in,out,flags) & bind(C, name='fftw_plan_guru_dft_r2c') import integer(C_INT), value :: rank type(fftw_iodim), dimension(*), intent(in) :: dims integer(C_INT), value :: howmany_rank type(fftw_iodim), dimension(*), intent(in) :: howmany_dims real(C_DOUBLE), dimension(*), intent(out) :: in complex(C_DOUBLE_COMPLEX), dimension(*), intent(out) :: out integer(C_INT), value :: flags end function fftw_plan_guru_dft_r2c type(C_PTR) function fftw_plan_guru_dft_c2r(rank,dims,howmany_rank,howmany_dims,in,out,flags) & bind(C, name='fftw_plan_guru_dft_c2r') import integer(C_INT), value :: rank type(fftw_iodim), dimension(*), intent(in) :: dims integer(C_INT), value :: howmany_rank type(fftw_iodim), dimension(*), intent(in) :: howmany_dims complex(C_DOUBLE_COMPLEX), dimension(*), intent(out) :: in real(C_DOUBLE), dimension(*), intent(out) :: out integer(C_INT), value :: flags end function fftw_plan_guru_dft_c2r type(C_PTR) function fftw_plan_guru_split_dft_r2c(rank,dims,howmany_rank,howmany_dims,in,ro,io,flags) & bind(C, name='fftw_plan_guru_split_dft_r2c') import integer(C_INT), value :: rank type(fftw_iodim), dimension(*), intent(in) :: dims integer(C_INT), value :: howmany_rank type(fftw_iodim), dimension(*), intent(in) :: howmany_dims real(C_DOUBLE), dimension(*), intent(out) :: in real(C_DOUBLE), dimension(*), intent(out) :: ro real(C_DOUBLE), dimension(*), intent(out) :: io integer(C_INT), value :: flags end function fftw_plan_guru_split_dft_r2c type(C_PTR) function fftw_plan_guru_split_dft_c2r(rank,dims,howmany_rank,howmany_dims,ri,ii,out,flags) & bind(C, name='fftw_plan_guru_split_dft_c2r') import integer(C_INT), value :: rank type(fftw_iodim), dimension(*), intent(in) :: dims integer(C_INT), value :: howmany_rank type(fftw_iodim), dimension(*), intent(in) :: howmany_dims real(C_DOUBLE), dimension(*), intent(out) :: ri real(C_DOUBLE), dimension(*), intent(out) :: ii real(C_DOUBLE), dimension(*), intent(out) :: out integer(C_INT), value :: flags end function fftw_plan_guru_split_dft_c2r type(C_PTR) function fftw_plan_guru64_dft_r2c(rank,dims,howmany_rank,howmany_dims,in,out,flags) & bind(C, name='fftw_plan_guru64_dft_r2c') import integer(C_INT), value :: rank type(fftw_iodim64), dimension(*), intent(in) :: dims integer(C_INT), value :: howmany_rank type(fftw_iodim64), dimension(*), intent(in) :: howmany_dims real(C_DOUBLE), dimension(*), intent(out) :: in complex(C_DOUBLE_COMPLEX), dimension(*), intent(out) :: out integer(C_INT), value :: flags end function fftw_plan_guru64_dft_r2c type(C_PTR) function fftw_plan_guru64_dft_c2r(rank,dims,howmany_rank,howmany_dims,in,out,flags) & bind(C, name='fftw_plan_guru64_dft_c2r') import integer(C_INT), value :: rank type(fftw_iodim64), dimension(*), intent(in) :: dims integer(C_INT), value :: howmany_rank type(fftw_iodim64), dimension(*), intent(in) :: howmany_dims complex(C_DOUBLE_COMPLEX), dimension(*), intent(out) :: in real(C_DOUBLE), dimension(*), intent(out) :: out integer(C_INT), value :: flags end function fftw_plan_guru64_dft_c2r type(C_PTR) function fftw_plan_guru64_split_dft_r2c(rank,dims,howmany_rank,howmany_dims,in,ro,io,flags) & bind(C, name='fftw_plan_guru64_split_dft_r2c') import integer(C_INT), value :: rank type(fftw_iodim64), dimension(*), intent(in) :: dims integer(C_INT), value :: howmany_rank type(fftw_iodim64), dimension(*), intent(in) :: howmany_dims real(C_DOUBLE), dimension(*), intent(out) :: in real(C_DOUBLE), dimension(*), intent(out) :: ro real(C_DOUBLE), dimension(*), intent(out) :: io integer(C_INT), value :: flags end function fftw_plan_guru64_split_dft_r2c type(C_PTR) function fftw_plan_guru64_split_dft_c2r(rank,dims,howmany_rank,howmany_dims,ri,ii,out,flags) & bind(C, name='fftw_plan_guru64_split_dft_c2r') import integer(C_INT), value :: rank type(fftw_iodim64), dimension(*), intent(in) :: dims integer(C_INT), value :: howmany_rank type(fftw_iodim64), dimension(*), intent(in) :: howmany_dims real(C_DOUBLE), dimension(*), intent(out) :: ri real(C_DOUBLE), dimension(*), intent(out) :: ii real(C_DOUBLE), dimension(*), intent(out) :: out integer(C_INT), value :: flags end function fftw_plan_guru64_split_dft_c2r subroutine fftw_execute_dft_r2c(p,in,out) bind(C, name='fftw_execute_dft_r2c') import type(C_PTR), value :: p real(C_DOUBLE), dimension(*), intent(inout) :: in complex(C_DOUBLE_COMPLEX), dimension(*), intent(out) :: out end subroutine fftw_execute_dft_r2c subroutine fftw_execute_dft_c2r(p,in,out) bind(C, name='fftw_execute_dft_c2r') import type(C_PTR), value :: p complex(C_DOUBLE_COMPLEX), dimension(*), intent(inout) :: in real(C_DOUBLE), dimension(*), intent(out) :: out end subroutine fftw_execute_dft_c2r subroutine fftw_execute_split_dft_r2c(p,in,ro,io) bind(C, name='fftw_execute_split_dft_r2c') import type(C_PTR), value :: p real(C_DOUBLE), dimension(*), intent(inout) :: in real(C_DOUBLE), dimension(*), intent(out) :: ro real(C_DOUBLE), dimension(*), intent(out) :: io end subroutine fftw_execute_split_dft_r2c subroutine fftw_execute_split_dft_c2r(p,ri,ii,out) bind(C, name='fftw_execute_split_dft_c2r') import type(C_PTR), value :: p real(C_DOUBLE), dimension(*), intent(inout) :: ri real(C_DOUBLE), dimension(*), intent(inout) :: ii real(C_DOUBLE), dimension(*), intent(out) :: out end subroutine fftw_execute_split_dft_c2r type(C_PTR) function fftw_plan_many_r2r(rank,n,howmany,in,inembed,istride,idist,out,onembed,ostride,odist,kind,flags) & bind(C, name='fftw_plan_many_r2r') import integer(C_INT), value :: rank integer(C_INT), dimension(*), intent(in) :: n integer(C_INT), value :: howmany real(C_DOUBLE), dimension(*), intent(out) :: in integer(C_INT), dimension(*), intent(in) :: inembed integer(C_INT), value :: istride integer(C_INT), value :: idist real(C_DOUBLE), dimension(*), intent(out) :: out integer(C_INT), dimension(*), intent(in) :: onembed integer(C_INT), value :: ostride integer(C_INT), value :: odist integer(C_FFTW_R2R_KIND), dimension(*), intent(in) :: kind integer(C_INT), value :: flags end function fftw_plan_many_r2r type(C_PTR) function fftw_plan_r2r(rank,n,in,out,kind,flags) bind(C, name='fftw_plan_r2r') import integer(C_INT), value :: rank integer(C_INT), dimension(*), intent(in) :: n real(C_DOUBLE), dimension(*), intent(out) :: in real(C_DOUBLE), dimension(*), intent(out) :: out integer(C_FFTW_R2R_KIND), dimension(*), intent(in) :: kind integer(C_INT), value :: flags end function fftw_plan_r2r type(C_PTR) function fftw_plan_r2r_1d(n,in,out,kind,flags) bind(C, name='fftw_plan_r2r_1d') import integer(C_INT), value :: n real(C_DOUBLE), dimension(*), intent(out) :: in real(C_DOUBLE), dimension(*), intent(out) :: out integer(C_FFTW_R2R_KIND), value :: kind integer(C_INT), value :: flags end function fftw_plan_r2r_1d type(C_PTR) function fftw_plan_r2r_2d(n0,n1,in,out,kind0,kind1,flags) bind(C, name='fftw_plan_r2r_2d') import integer(C_INT), value :: n0 integer(C_INT), value :: n1 real(C_DOUBLE), dimension(*), intent(out) :: in real(C_DOUBLE), dimension(*), intent(out) :: out integer(C_FFTW_R2R_KIND), value :: kind0 integer(C_FFTW_R2R_KIND), value :: kind1 integer(C_INT), value :: flags end function fftw_plan_r2r_2d type(C_PTR) function fftw_plan_r2r_3d(n0,n1,n2,in,out,kind0,kind1,kind2,flags) bind(C, name='fftw_plan_r2r_3d') import integer(C_INT), value :: n0 integer(C_INT), value :: n1 integer(C_INT), value :: n2 real(C_DOUBLE), dimension(*), intent(out) :: in real(C_DOUBLE), dimension(*), intent(out) :: out integer(C_FFTW_R2R_KIND), value :: kind0 integer(C_FFTW_R2R_KIND), value :: kind1 integer(C_FFTW_R2R_KIND), value :: kind2 integer(C_INT), value :: flags end function fftw_plan_r2r_3d type(C_PTR) function fftw_plan_guru_r2r(rank,dims,howmany_rank,howmany_dims,in,out,kind,flags) & bind(C, name='fftw_plan_guru_r2r') import integer(C_INT), value :: rank type(fftw_iodim), dimension(*), intent(in) :: dims integer(C_INT), value :: howmany_rank type(fftw_iodim), dimension(*), intent(in) :: howmany_dims real(C_DOUBLE), dimension(*), intent(out) :: in real(C_DOUBLE), dimension(*), intent(out) :: out integer(C_FFTW_R2R_KIND), dimension(*), intent(in) :: kind integer(C_INT), value :: flags end function fftw_plan_guru_r2r type(C_PTR) function fftw_plan_guru64_r2r(rank,dims,howmany_rank,howmany_dims,in,out,kind,flags) & bind(C, name='fftw_plan_guru64_r2r') import integer(C_INT), value :: rank type(fftw_iodim64), dimension(*), intent(in) :: dims integer(C_INT), value :: howmany_rank type(fftw_iodim64), dimension(*), intent(in) :: howmany_dims real(C_DOUBLE), dimension(*), intent(out) :: in real(C_DOUBLE), dimension(*), intent(out) :: out integer(C_FFTW_R2R_KIND), dimension(*), intent(in) :: kind integer(C_INT), value :: flags end function fftw_plan_guru64_r2r subroutine fftw_execute_r2r(p,in,out) bind(C, name='fftw_execute_r2r') import type(C_PTR), value :: p real(C_DOUBLE), dimension(*), intent(inout) :: in real(C_DOUBLE), dimension(*), intent(out) :: out end subroutine fftw_execute_r2r subroutine fftw_destroy_plan(p) bind(C, name='fftw_destroy_plan') import type(C_PTR), value :: p end subroutine fftw_destroy_plan subroutine fftw_forget_wisdom() bind(C, name='fftw_forget_wisdom') import end subroutine fftw_forget_wisdom subroutine fftw_cleanup() bind(C, name='fftw_cleanup') import end subroutine fftw_cleanup subroutine fftw_set_timelimit(t) bind(C, name='fftw_set_timelimit') import real(C_DOUBLE), value :: t end subroutine fftw_set_timelimit subroutine fftw_plan_with_nthreads(nthreads) bind(C, name='fftw_plan_with_nthreads') import integer(C_INT), value :: nthreads end subroutine fftw_plan_with_nthreads integer(C_INT) function fftw_init_threads() bind(C, name='fftw_init_threads') import end function fftw_init_threads subroutine fftw_cleanup_threads() bind(C, name='fftw_cleanup_threads') import end subroutine fftw_cleanup_threads integer(C_INT) function fftw_export_wisdom_to_filename(filename) bind(C, name='fftw_export_wisdom_to_filename') import character(C_CHAR), dimension(*), intent(in) :: filename end function fftw_export_wisdom_to_filename subroutine fftw_export_wisdom_to_file(output_file) bind(C, name='fftw_export_wisdom_to_file') import type(C_PTR), value :: output_file end subroutine fftw_export_wisdom_to_file type(C_PTR) function fftw_export_wisdom_to_string() bind(C, name='fftw_export_wisdom_to_string') import end function fftw_export_wisdom_to_string subroutine fftw_export_wisdom(write_char,data) bind(C, name='fftw_export_wisdom') import type(C_FUNPTR), value :: write_char type(C_PTR), value :: data end subroutine fftw_export_wisdom integer(C_INT) function fftw_import_system_wisdom() bind(C, name='fftw_import_system_wisdom') import end function fftw_import_system_wisdom integer(C_INT) function fftw_import_wisdom_from_filename(filename) bind(C, name='fftw_import_wisdom_from_filename') import character(C_CHAR), dimension(*), intent(in) :: filename end function fftw_import_wisdom_from_filename integer(C_INT) function fftw_import_wisdom_from_file(input_file) bind(C, name='fftw_import_wisdom_from_file') import type(C_PTR), value :: input_file end function fftw_import_wisdom_from_file integer(C_INT) function fftw_import_wisdom_from_string(input_string) bind(C, name='fftw_import_wisdom_from_string') import character(C_CHAR), dimension(*), intent(in) :: input_string end function fftw_import_wisdom_from_string integer(C_INT) function fftw_import_wisdom(read_char,data) bind(C, name='fftw_import_wisdom') import type(C_FUNPTR), value :: read_char type(C_PTR), value :: data end function fftw_import_wisdom subroutine fftw_fprint_plan(p,output_file) bind(C, name='fftw_fprint_plan') import type(C_PTR), value :: p type(C_PTR), value :: output_file end subroutine fftw_fprint_plan subroutine fftw_print_plan(p) bind(C, name='fftw_print_plan') import type(C_PTR), value :: p end subroutine fftw_print_plan type(C_PTR) function fftw_malloc(n) bind(C, name='fftw_malloc') import integer(C_SIZE_T), value :: n end function fftw_malloc type(C_PTR) function fftw_alloc_real(n) bind(C, name='fftw_alloc_real') import integer(C_SIZE_T), value :: n end function fftw_alloc_real type(C_PTR) function fftw_alloc_complex(n) bind(C, name='fftw_alloc_complex') import integer(C_SIZE_T), value :: n end function fftw_alloc_complex subroutine fftw_free(p) bind(C, name='fftw_free') import type(C_PTR), value :: p end subroutine fftw_free subroutine fftw_flops(p,add,mul,fmas) bind(C, name='fftw_flops') import type(C_PTR), value :: p real(C_DOUBLE), intent(out) :: add real(C_DOUBLE), intent(out) :: mul real(C_DOUBLE), intent(out) :: fmas end subroutine fftw_flops real(C_DOUBLE) function fftw_estimate_cost(p) bind(C, name='fftw_estimate_cost') import type(C_PTR), value :: p end function fftw_estimate_cost real(C_DOUBLE) function fftw_cost(p) bind(C, name='fftw_cost') import type(C_PTR), value :: p end function fftw_cost end interface type, bind(C) :: fftwf_iodim integer(C_INT) n, is, os end type fftwf_iodim type, bind(C) :: fftwf_iodim64 integer(C_INTPTR_T) n, is, os end type fftwf_iodim64 interface type(C_PTR) function fftwf_plan_dft(rank,n,in,out,sign,flags) bind(C, name='fftwf_plan_dft') import integer(C_INT), value :: rank integer(C_INT), dimension(*), intent(in) :: n complex(C_FLOAT_COMPLEX), dimension(*), intent(out) :: in complex(C_FLOAT_COMPLEX), dimension(*), intent(out) :: out integer(C_INT), value :: sign integer(C_INT), value :: flags end function fftwf_plan_dft type(C_PTR) function fftwf_plan_dft_1d(n,in,out,sign,flags) bind(C, name='fftwf_plan_dft_1d') import integer(C_INT), value :: n complex(C_FLOAT_COMPLEX), dimension(*), intent(out) :: in complex(C_FLOAT_COMPLEX), dimension(*), intent(out) :: out integer(C_INT), value :: sign integer(C_INT), value :: flags end function fftwf_plan_dft_1d type(C_PTR) function fftwf_plan_dft_2d(n0,n1,in,out,sign,flags) bind(C, name='fftwf_plan_dft_2d') import integer(C_INT), value :: n0 integer(C_INT), value :: n1 complex(C_FLOAT_COMPLEX), dimension(*), intent(out) :: in complex(C_FLOAT_COMPLEX), dimension(*), intent(out) :: out integer(C_INT), value :: sign integer(C_INT), value :: flags end function fftwf_plan_dft_2d type(C_PTR) function fftwf_plan_dft_3d(n0,n1,n2,in,out,sign,flags) bind(C, name='fftwf_plan_dft_3d') import integer(C_INT), value :: n0 integer(C_INT), value :: n1 integer(C_INT), value :: n2 complex(C_FLOAT_COMPLEX), dimension(*), intent(out) :: in complex(C_FLOAT_COMPLEX), dimension(*), intent(out) :: out integer(C_INT), value :: sign integer(C_INT), value :: flags end function fftwf_plan_dft_3d type(C_PTR) function fftwf_plan_many_dft(rank,n,howmany,in,inembed,istride,idist,out,onembed,ostride,odist,sign,flags) & bind(C, name='fftwf_plan_many_dft') import integer(C_INT), value :: rank integer(C_INT), dimension(*), intent(in) :: n integer(C_INT), value :: howmany complex(C_FLOAT_COMPLEX), dimension(*), intent(out) :: in integer(C_INT), dimension(*), intent(in) :: inembed integer(C_INT), value :: istride integer(C_INT), value :: idist complex(C_FLOAT_COMPLEX), dimension(*), intent(out) :: out integer(C_INT), dimension(*), intent(in) :: onembed integer(C_INT), value :: ostride integer(C_INT), value :: odist integer(C_INT), value :: sign integer(C_INT), value :: flags end function fftwf_plan_many_dft type(C_PTR) function fftwf_plan_guru_dft(rank,dims,howmany_rank,howmany_dims,in,out,sign,flags) & bind(C, name='fftwf_plan_guru_dft') import integer(C_INT), value :: rank type(fftwf_iodim), dimension(*), intent(in) :: dims integer(C_INT), value :: howmany_rank type(fftwf_iodim), dimension(*), intent(in) :: howmany_dims complex(C_FLOAT_COMPLEX), dimension(*), intent(out) :: in complex(C_FLOAT_COMPLEX), dimension(*), intent(out) :: out integer(C_INT), value :: sign integer(C_INT), value :: flags end function fftwf_plan_guru_dft type(C_PTR) function fftwf_plan_guru_split_dft(rank,dims,howmany_rank,howmany_dims,ri,ii,ro,io,flags) & bind(C, name='fftwf_plan_guru_split_dft') import integer(C_INT), value :: rank type(fftwf_iodim), dimension(*), intent(in) :: dims integer(C_INT), value :: howmany_rank type(fftwf_iodim), dimension(*), intent(in) :: howmany_dims real(C_FLOAT), dimension(*), intent(out) :: ri real(C_FLOAT), dimension(*), intent(out) :: ii real(C_FLOAT), dimension(*), intent(out) :: ro real(C_FLOAT), dimension(*), intent(out) :: io integer(C_INT), value :: flags end function fftwf_plan_guru_split_dft type(C_PTR) function fftwf_plan_guru64_dft(rank,dims,howmany_rank,howmany_dims,in,out,sign,flags) & bind(C, name='fftwf_plan_guru64_dft') import integer(C_INT), value :: rank type(fftwf_iodim64), dimension(*), intent(in) :: dims integer(C_INT), value :: howmany_rank type(fftwf_iodim64), dimension(*), intent(in) :: howmany_dims complex(C_FLOAT_COMPLEX), dimension(*), intent(out) :: in complex(C_FLOAT_COMPLEX), dimension(*), intent(out) :: out integer(C_INT), value :: sign integer(C_INT), value :: flags end function fftwf_plan_guru64_dft type(C_PTR) function fftwf_plan_guru64_split_dft(rank,dims,howmany_rank,howmany_dims,ri,ii,ro,io,flags) & bind(C, name='fftwf_plan_guru64_split_dft') import integer(C_INT), value :: rank type(fftwf_iodim64), dimension(*), intent(in) :: dims integer(C_INT), value :: howmany_rank type(fftwf_iodim64), dimension(*), intent(in) :: howmany_dims real(C_FLOAT), dimension(*), intent(out) :: ri real(C_FLOAT), dimension(*), intent(out) :: ii real(C_FLOAT), dimension(*), intent(out) :: ro real(C_FLOAT), dimension(*), intent(out) :: io integer(C_INT), value :: flags end function fftwf_plan_guru64_split_dft subroutine fftwf_execute_dft(p,in,out) bind(C, name='fftwf_execute_dft') import type(C_PTR), value :: p complex(C_FLOAT_COMPLEX), dimension(*), intent(inout) :: in complex(C_FLOAT_COMPLEX), dimension(*), intent(out) :: out end subroutine fftwf_execute_dft subroutine fftwf_execute_split_dft(p,ri,ii,ro,io) bind(C, name='fftwf_execute_split_dft') import type(C_PTR), value :: p real(C_FLOAT), dimension(*), intent(inout) :: ri real(C_FLOAT), dimension(*), intent(inout) :: ii real(C_FLOAT), dimension(*), intent(out) :: ro real(C_FLOAT), dimension(*), intent(out) :: io end subroutine fftwf_execute_split_dft type(C_PTR) function fftwf_plan_many_dft_r2c(rank,n,howmany,in,inembed,istride,idist,out,onembed,ostride,odist,flags) & bind(C, name='fftwf_plan_many_dft_r2c') import integer(C_INT), value :: rank integer(C_INT), dimension(*), intent(in) :: n integer(C_INT), value :: howmany real(C_FLOAT), dimension(*), intent(out) :: in integer(C_INT), dimension(*), intent(in) :: inembed integer(C_INT), value :: istride integer(C_INT), value :: idist complex(C_FLOAT_COMPLEX), dimension(*), intent(out) :: out integer(C_INT), dimension(*), intent(in) :: onembed integer(C_INT), value :: ostride integer(C_INT), value :: odist integer(C_INT), value :: flags end function fftwf_plan_many_dft_r2c type(C_PTR) function fftwf_plan_dft_r2c(rank,n,in,out,flags) bind(C, name='fftwf_plan_dft_r2c') import integer(C_INT), value :: rank integer(C_INT), dimension(*), intent(in) :: n real(C_FLOAT), dimension(*), intent(out) :: in complex(C_FLOAT_COMPLEX), dimension(*), intent(out) :: out integer(C_INT), value :: flags end function fftwf_plan_dft_r2c type(C_PTR) function fftwf_plan_dft_r2c_1d(n,in,out,flags) bind(C, name='fftwf_plan_dft_r2c_1d') import integer(C_INT), value :: n real(C_FLOAT), dimension(*), intent(out) :: in complex(C_FLOAT_COMPLEX), dimension(*), intent(out) :: out integer(C_INT), value :: flags end function fftwf_plan_dft_r2c_1d type(C_PTR) function fftwf_plan_dft_r2c_2d(n0,n1,in,out,flags) bind(C, name='fftwf_plan_dft_r2c_2d') import integer(C_INT), value :: n0 integer(C_INT), value :: n1 real(C_FLOAT), dimension(*), intent(out) :: in complex(C_FLOAT_COMPLEX), dimension(*), intent(out) :: out integer(C_INT), value :: flags end function fftwf_plan_dft_r2c_2d type(C_PTR) function fftwf_plan_dft_r2c_3d(n0,n1,n2,in,out,flags) bind(C, name='fftwf_plan_dft_r2c_3d') import integer(C_INT), value :: n0 integer(C_INT), value :: n1 integer(C_INT), value :: n2 real(C_FLOAT), dimension(*), intent(out) :: in complex(C_FLOAT_COMPLEX), dimension(*), intent(out) :: out integer(C_INT), value :: flags end function fftwf_plan_dft_r2c_3d type(C_PTR) function fftwf_plan_many_dft_c2r(rank,n,howmany,in,inembed,istride,idist,out,onembed,ostride,odist,flags) & bind(C, name='fftwf_plan_many_dft_c2r') import integer(C_INT), value :: rank integer(C_INT), dimension(*), intent(in) :: n integer(C_INT), value :: howmany complex(C_FLOAT_COMPLEX), dimension(*), intent(out) :: in integer(C_INT), dimension(*), intent(in) :: inembed integer(C_INT), value :: istride integer(C_INT), value :: idist real(C_FLOAT), dimension(*), intent(out) :: out integer(C_INT), dimension(*), intent(in) :: onembed integer(C_INT), value :: ostride integer(C_INT), value :: odist integer(C_INT), value :: flags end function fftwf_plan_many_dft_c2r type(C_PTR) function fftwf_plan_dft_c2r(rank,n,in,out,flags) bind(C, name='fftwf_plan_dft_c2r') import integer(C_INT), value :: rank integer(C_INT), dimension(*), intent(in) :: n complex(C_FLOAT_COMPLEX), dimension(*), intent(out) :: in real(C_FLOAT), dimension(*), intent(out) :: out integer(C_INT), value :: flags end function fftwf_plan_dft_c2r type(C_PTR) function fftwf_plan_dft_c2r_1d(n,in,out,flags) bind(C, name='fftwf_plan_dft_c2r_1d') import integer(C_INT), value :: n complex(C_FLOAT_COMPLEX), dimension(*), intent(out) :: in real(C_FLOAT), dimension(*), intent(out) :: out integer(C_INT), value :: flags end function fftwf_plan_dft_c2r_1d type(C_PTR) function fftwf_plan_dft_c2r_2d(n0,n1,in,out,flags) bind(C, name='fftwf_plan_dft_c2r_2d') import integer(C_INT), value :: n0 integer(C_INT), value :: n1 complex(C_FLOAT_COMPLEX), dimension(*), intent(out) :: in real(C_FLOAT), dimension(*), intent(out) :: out integer(C_INT), value :: flags end function fftwf_plan_dft_c2r_2d type(C_PTR) function fftwf_plan_dft_c2r_3d(n0,n1,n2,in,out,flags) bind(C, name='fftwf_plan_dft_c2r_3d') import integer(C_INT), value :: n0 integer(C_INT), value :: n1 integer(C_INT), value :: n2 complex(C_FLOAT_COMPLEX), dimension(*), intent(out) :: in real(C_FLOAT), dimension(*), intent(out) :: out integer(C_INT), value :: flags end function fftwf_plan_dft_c2r_3d type(C_PTR) function fftwf_plan_guru_dft_r2c(rank,dims,howmany_rank,howmany_dims,in,out,flags) & bind(C, name='fftwf_plan_guru_dft_r2c') import integer(C_INT), value :: rank type(fftwf_iodim), dimension(*), intent(in) :: dims integer(C_INT), value :: howmany_rank type(fftwf_iodim), dimension(*), intent(in) :: howmany_dims real(C_FLOAT), dimension(*), intent(out) :: in complex(C_FLOAT_COMPLEX), dimension(*), intent(out) :: out integer(C_INT), value :: flags end function fftwf_plan_guru_dft_r2c type(C_PTR) function fftwf_plan_guru_dft_c2r(rank,dims,howmany_rank,howmany_dims,in,out,flags) & bind(C, name='fftwf_plan_guru_dft_c2r') import integer(C_INT), value :: rank type(fftwf_iodim), dimension(*), intent(in) :: dims integer(C_INT), value :: howmany_rank type(fftwf_iodim), dimension(*), intent(in) :: howmany_dims complex(C_FLOAT_COMPLEX), dimension(*), intent(out) :: in real(C_FLOAT), dimension(*), intent(out) :: out integer(C_INT), value :: flags end function fftwf_plan_guru_dft_c2r type(C_PTR) function fftwf_plan_guru_split_dft_r2c(rank,dims,howmany_rank,howmany_dims,in,ro,io,flags) & bind(C, name='fftwf_plan_guru_split_dft_r2c') import integer(C_INT), value :: rank type(fftwf_iodim), dimension(*), intent(in) :: dims integer(C_INT), value :: howmany_rank type(fftwf_iodim), dimension(*), intent(in) :: howmany_dims real(C_FLOAT), dimension(*), intent(out) :: in real(C_FLOAT), dimension(*), intent(out) :: ro real(C_FLOAT), dimension(*), intent(out) :: io integer(C_INT), value :: flags end function fftwf_plan_guru_split_dft_r2c type(C_PTR) function fftwf_plan_guru_split_dft_c2r(rank,dims,howmany_rank,howmany_dims,ri,ii,out,flags) & bind(C, name='fftwf_plan_guru_split_dft_c2r') import integer(C_INT), value :: rank type(fftwf_iodim), dimension(*), intent(in) :: dims integer(C_INT), value :: howmany_rank type(fftwf_iodim), dimension(*), intent(in) :: howmany_dims real(C_FLOAT), dimension(*), intent(out) :: ri real(C_FLOAT), dimension(*), intent(out) :: ii real(C_FLOAT), dimension(*), intent(out) :: out integer(C_INT), value :: flags end function fftwf_plan_guru_split_dft_c2r type(C_PTR) function fftwf_plan_guru64_dft_r2c(rank,dims,howmany_rank,howmany_dims,in,out,flags) & bind(C, name='fftwf_plan_guru64_dft_r2c') import integer(C_INT), value :: rank type(fftwf_iodim64), dimension(*), intent(in) :: dims integer(C_INT), value :: howmany_rank type(fftwf_iodim64), dimension(*), intent(in) :: howmany_dims real(C_FLOAT), dimension(*), intent(out) :: in complex(C_FLOAT_COMPLEX), dimension(*), intent(out) :: out integer(C_INT), value :: flags end function fftwf_plan_guru64_dft_r2c type(C_PTR) function fftwf_plan_guru64_dft_c2r(rank,dims,howmany_rank,howmany_dims,in,out,flags) & bind(C, name='fftwf_plan_guru64_dft_c2r') import integer(C_INT), value :: rank type(fftwf_iodim64), dimension(*), intent(in) :: dims integer(C_INT), value :: howmany_rank type(fftwf_iodim64), dimension(*), intent(in) :: howmany_dims complex(C_FLOAT_COMPLEX), dimension(*), intent(out) :: in real(C_FLOAT), dimension(*), intent(out) :: out integer(C_INT), value :: flags end function fftwf_plan_guru64_dft_c2r type(C_PTR) function fftwf_plan_guru64_split_dft_r2c(rank,dims,howmany_rank,howmany_dims,in,ro,io,flags) & bind(C, name='fftwf_plan_guru64_split_dft_r2c') import integer(C_INT), value :: rank type(fftwf_iodim64), dimension(*), intent(in) :: dims integer(C_INT), value :: howmany_rank type(fftwf_iodim64), dimension(*), intent(in) :: howmany_dims real(C_FLOAT), dimension(*), intent(out) :: in real(C_FLOAT), dimension(*), intent(out) :: ro real(C_FLOAT), dimension(*), intent(out) :: io integer(C_INT), value :: flags end function fftwf_plan_guru64_split_dft_r2c type(C_PTR) function fftwf_plan_guru64_split_dft_c2r(rank,dims,howmany_rank,howmany_dims,ri,ii,out,flags) & bind(C, name='fftwf_plan_guru64_split_dft_c2r') import integer(C_INT), value :: rank type(fftwf_iodim64), dimension(*), intent(in) :: dims integer(C_INT), value :: howmany_rank type(fftwf_iodim64), dimension(*), intent(in) :: howmany_dims real(C_FLOAT), dimension(*), intent(out) :: ri real(C_FLOAT), dimension(*), intent(out) :: ii real(C_FLOAT), dimension(*), intent(out) :: out integer(C_INT), value :: flags end function fftwf_plan_guru64_split_dft_c2r subroutine fftwf_execute_dft_r2c(p,in,out) bind(C, name='fftwf_execute_dft_r2c') import type(C_PTR), value :: p real(C_FLOAT), dimension(*), intent(inout) :: in complex(C_FLOAT_COMPLEX), dimension(*), intent(out) :: out end subroutine fftwf_execute_dft_r2c subroutine fftwf_execute_dft_c2r(p,in,out) bind(C, name='fftwf_execute_dft_c2r') import type(C_PTR), value :: p complex(C_FLOAT_COMPLEX), dimension(*), intent(inout) :: in real(C_FLOAT), dimension(*), intent(out) :: out end subroutine fftwf_execute_dft_c2r subroutine fftwf_execute_split_dft_r2c(p,in,ro,io) bind(C, name='fftwf_execute_split_dft_r2c') import type(C_PTR), value :: p real(C_FLOAT), dimension(*), intent(inout) :: in real(C_FLOAT), dimension(*), intent(out) :: ro real(C_FLOAT), dimension(*), intent(out) :: io end subroutine fftwf_execute_split_dft_r2c subroutine fftwf_execute_split_dft_c2r(p,ri,ii,out) bind(C, name='fftwf_execute_split_dft_c2r') import type(C_PTR), value :: p real(C_FLOAT), dimension(*), intent(inout) :: ri real(C_FLOAT), dimension(*), intent(inout) :: ii real(C_FLOAT), dimension(*), intent(out) :: out end subroutine fftwf_execute_split_dft_c2r type(C_PTR) function fftwf_plan_many_r2r(rank,n,howmany,in,inembed,istride,idist,out,onembed,ostride,odist,kind,flags) & bind(C, name='fftwf_plan_many_r2r') import integer(C_INT), value :: rank integer(C_INT), dimension(*), intent(in) :: n integer(C_INT), value :: howmany real(C_FLOAT), dimension(*), intent(out) :: in integer(C_INT), dimension(*), intent(in) :: inembed integer(C_INT), value :: istride integer(C_INT), value :: idist real(C_FLOAT), dimension(*), intent(out) :: out integer(C_INT), dimension(*), intent(in) :: onembed integer(C_INT), value :: ostride integer(C_INT), value :: odist integer(C_FFTW_R2R_KIND), dimension(*), intent(in) :: kind integer(C_INT), value :: flags end function fftwf_plan_many_r2r type(C_PTR) function fftwf_plan_r2r(rank,n,in,out,kind,flags) bind(C, name='fftwf_plan_r2r') import integer(C_INT), value :: rank integer(C_INT), dimension(*), intent(in) :: n real(C_FLOAT), dimension(*), intent(out) :: in real(C_FLOAT), dimension(*), intent(out) :: out integer(C_FFTW_R2R_KIND), dimension(*), intent(in) :: kind integer(C_INT), value :: flags end function fftwf_plan_r2r type(C_PTR) function fftwf_plan_r2r_1d(n,in,out,kind,flags) bind(C, name='fftwf_plan_r2r_1d') import integer(C_INT), value :: n real(C_FLOAT), dimension(*), intent(out) :: in real(C_FLOAT), dimension(*), intent(out) :: out integer(C_FFTW_R2R_KIND), value :: kind integer(C_INT), value :: flags end function fftwf_plan_r2r_1d type(C_PTR) function fftwf_plan_r2r_2d(n0,n1,in,out,kind0,kind1,flags) bind(C, name='fftwf_plan_r2r_2d') import integer(C_INT), value :: n0 integer(C_INT), value :: n1 real(C_FLOAT), dimension(*), intent(out) :: in real(C_FLOAT), dimension(*), intent(out) :: out integer(C_FFTW_R2R_KIND), value :: kind0 integer(C_FFTW_R2R_KIND), value :: kind1 integer(C_INT), value :: flags end function fftwf_plan_r2r_2d type(C_PTR) function fftwf_plan_r2r_3d(n0,n1,n2,in,out,kind0,kind1,kind2,flags) bind(C, name='fftwf_plan_r2r_3d') import integer(C_INT), value :: n0 integer(C_INT), value :: n1 integer(C_INT), value :: n2 real(C_FLOAT), dimension(*), intent(out) :: in real(C_FLOAT), dimension(*), intent(out) :: out integer(C_FFTW_R2R_KIND), value :: kind0 integer(C_FFTW_R2R_KIND), value :: kind1 integer(C_FFTW_R2R_KIND), value :: kind2 integer(C_INT), value :: flags end function fftwf_plan_r2r_3d type(C_PTR) function fftwf_plan_guru_r2r(rank,dims,howmany_rank,howmany_dims,in,out,kind,flags) & bind(C, name='fftwf_plan_guru_r2r') import integer(C_INT), value :: rank type(fftwf_iodim), dimension(*), intent(in) :: dims integer(C_INT), value :: howmany_rank type(fftwf_iodim), dimension(*), intent(in) :: howmany_dims real(C_FLOAT), dimension(*), intent(out) :: in real(C_FLOAT), dimension(*), intent(out) :: out integer(C_FFTW_R2R_KIND), dimension(*), intent(in) :: kind integer(C_INT), value :: flags end function fftwf_plan_guru_r2r type(C_PTR) function fftwf_plan_guru64_r2r(rank,dims,howmany_rank,howmany_dims,in,out,kind,flags) & bind(C, name='fftwf_plan_guru64_r2r') import integer(C_INT), value :: rank type(fftwf_iodim64), dimension(*), intent(in) :: dims integer(C_INT), value :: howmany_rank type(fftwf_iodim64), dimension(*), intent(in) :: howmany_dims real(C_FLOAT), dimension(*), intent(out) :: in real(C_FLOAT), dimension(*), intent(out) :: out integer(C_FFTW_R2R_KIND), dimension(*), intent(in) :: kind integer(C_INT), value :: flags end function fftwf_plan_guru64_r2r subroutine fftwf_execute_r2r(p,in,out) bind(C, name='fftwf_execute_r2r') import type(C_PTR), value :: p real(C_FLOAT), dimension(*), intent(inout) :: in real(C_FLOAT), dimension(*), intent(out) :: out end subroutine fftwf_execute_r2r subroutine fftwf_destroy_plan(p) bind(C, name='fftwf_destroy_plan') import type(C_PTR), value :: p end subroutine fftwf_destroy_plan subroutine fftwf_forget_wisdom() bind(C, name='fftwf_forget_wisdom') import end subroutine fftwf_forget_wisdom subroutine fftwf_cleanup() bind(C, name='fftwf_cleanup') import end subroutine fftwf_cleanup subroutine fftwf_set_timelimit(t) bind(C, name='fftwf_set_timelimit') import real(C_DOUBLE), value :: t end subroutine fftwf_set_timelimit subroutine fftwf_plan_with_nthreads(nthreads) bind(C, name='fftwf_plan_with_nthreads') import integer(C_INT), value :: nthreads end subroutine fftwf_plan_with_nthreads integer(C_INT) function fftwf_init_threads() bind(C, name='fftwf_init_threads') import end function fftwf_init_threads subroutine fftwf_cleanup_threads() bind(C, name='fftwf_cleanup_threads') import end subroutine fftwf_cleanup_threads integer(C_INT) function fftwf_export_wisdom_to_filename(filename) bind(C, name='fftwf_export_wisdom_to_filename') import character(C_CHAR), dimension(*), intent(in) :: filename end function fftwf_export_wisdom_to_filename subroutine fftwf_export_wisdom_to_file(output_file) bind(C, name='fftwf_export_wisdom_to_file') import type(C_PTR), value :: output_file end subroutine fftwf_export_wisdom_to_file type(C_PTR) function fftwf_export_wisdom_to_string() bind(C, name='fftwf_export_wisdom_to_string') import end function fftwf_export_wisdom_to_string subroutine fftwf_export_wisdom(write_char,data) bind(C, name='fftwf_export_wisdom') import type(C_FUNPTR), value :: write_char type(C_PTR), value :: data end subroutine fftwf_export_wisdom integer(C_INT) function fftwf_import_system_wisdom() bind(C, name='fftwf_import_system_wisdom') import end function fftwf_import_system_wisdom integer(C_INT) function fftwf_import_wisdom_from_filename(filename) bind(C, name='fftwf_import_wisdom_from_filename') import character(C_CHAR), dimension(*), intent(in) :: filename end function fftwf_import_wisdom_from_filename integer(C_INT) function fftwf_import_wisdom_from_file(input_file) bind(C, name='fftwf_import_wisdom_from_file') import type(C_PTR), value :: input_file end function fftwf_import_wisdom_from_file integer(C_INT) function fftwf_import_wisdom_from_string(input_string) bind(C, name='fftwf_import_wisdom_from_string') import character(C_CHAR), dimension(*), intent(in) :: input_string end function fftwf_import_wisdom_from_string integer(C_INT) function fftwf_import_wisdom(read_char,data) bind(C, name='fftwf_import_wisdom') import type(C_FUNPTR), value :: read_char type(C_PTR), value :: data end function fftwf_import_wisdom subroutine fftwf_fprint_plan(p,output_file) bind(C, name='fftwf_fprint_plan') import type(C_PTR), value :: p type(C_PTR), value :: output_file end subroutine fftwf_fprint_plan subroutine fftwf_print_plan(p) bind(C, name='fftwf_print_plan') import type(C_PTR), value :: p end subroutine fftwf_print_plan type(C_PTR) function fftwf_malloc(n) bind(C, name='fftwf_malloc') import integer(C_SIZE_T), value :: n end function fftwf_malloc type(C_PTR) function fftwf_alloc_real(n) bind(C, name='fftwf_alloc_real') import integer(C_SIZE_T), value :: n end function fftwf_alloc_real type(C_PTR) function fftwf_alloc_complex(n) bind(C, name='fftwf_alloc_complex') import integer(C_SIZE_T), value :: n end function fftwf_alloc_complex subroutine fftwf_free(p) bind(C, name='fftwf_free') import type(C_PTR), value :: p end subroutine fftwf_free subroutine fftwf_flops(p,add,mul,fmas) bind(C, name='fftwf_flops') import type(C_PTR), value :: p real(C_DOUBLE), intent(out) :: add real(C_DOUBLE), intent(out) :: mul real(C_DOUBLE), intent(out) :: fmas end subroutine fftwf_flops real(C_DOUBLE) function fftwf_estimate_cost(p) bind(C, name='fftwf_estimate_cost') import type(C_PTR), value :: p end function fftwf_estimate_cost real(C_DOUBLE) function fftwf_cost(p) bind(C, name='fftwf_cost') import type(C_PTR), value :: p end function fftwf_cost end interface fftw-3.3.3/libbench2/0000755000175000017500000000000012054411110011304 500000000000000fftw-3.3.3/libbench2/Makefile.am0000644000175000017500000000112512054410050013261 00000000000000noinst_LIBRARIES=libbench2.a libbench2_a_SOURCES=after-ccopy-from.c after-ccopy-to.c \ after-hccopy-from.c after-hccopy-to.c after-rcopy-from.c \ after-rcopy-to.c allocate.c aset.c bench-cost-postprocess.c \ bench-exit.c bench-main.c can-do.c caset.c dotens2.c info.c main.c \ mflops.c mp.c ovtpvt.c pow2.c problem.c report.c speed.c tensor.c \ timer.c useropt.c util.c verify-dft.c verify-lib.c verify-r2r.c \ verify-rdft2.c verify.c zero.c bench-user.h bench.h verify.h \ my-getopt.c my-getopt.h benchmark: all @echo "nothing to benchmark" accuracy: all @echo "nothing to benchmark" fftw-3.3.3/libbench2/Makefile.in0000644000175000017500000004607512054410200013304 00000000000000# Makefile.in generated by automake 1.11.6 from Makefile.am. # @configure_input@ # Copyright (C) 1994, 1995, 1996, 1997, 1998, 1999, 2000, 2001, 2002, # 2003, 2004, 2005, 2006, 2007, 2008, 2009, 2010, 2011 Free Software # Foundation, Inc. # This Makefile.in is free software; the Free Software Foundation # gives unlimited permission to copy and/or distribute it, # with or without modifications, as long as this notice is preserved. # This program is distributed in the hope that it will be useful, # but WITHOUT ANY WARRANTY, to the extent permitted by law; without # even the implied warranty of MERCHANTABILITY or FITNESS FOR A # PARTICULAR PURPOSE. @SET_MAKE@ VPATH = @srcdir@ am__make_dryrun = \ { \ am__dry=no; \ case $$MAKEFLAGS in \ *\\[\ \ ]*) \ echo 'am--echo: ; @echo "AM" OK' | $(MAKE) -f - 2>/dev/null \ | grep '^AM OK$$' >/dev/null || am__dry=yes;; \ *) \ for am__flg in $$MAKEFLAGS; do \ case $$am__flg in \ *=*|--*) ;; \ *n*) am__dry=yes; break;; \ esac; \ done;; \ esac; \ test $$am__dry = yes; \ } pkgdatadir = $(datadir)/@PACKAGE@ pkgincludedir = $(includedir)/@PACKAGE@ pkglibdir = $(libdir)/@PACKAGE@ pkglibexecdir = $(libexecdir)/@PACKAGE@ am__cd = CDPATH="$${ZSH_VERSION+.}$(PATH_SEPARATOR)" && cd install_sh_DATA = $(install_sh) -c -m 644 install_sh_PROGRAM = $(install_sh) -c install_sh_SCRIPT = $(install_sh) -c INSTALL_HEADER = $(INSTALL_DATA) transform = $(program_transform_name) NORMAL_INSTALL = : PRE_INSTALL = : POST_INSTALL = : NORMAL_UNINSTALL = : PRE_UNINSTALL = : POST_UNINSTALL = : build_triplet = @build@ host_triplet = @host@ subdir = libbench2 DIST_COMMON = $(srcdir)/Makefile.am $(srcdir)/Makefile.in ACLOCAL_M4 = $(top_srcdir)/aclocal.m4 am__aclocal_m4_deps = $(top_srcdir)/m4/acx_mpi.m4 \ $(top_srcdir)/m4/acx_pthread.m4 \ $(top_srcdir)/m4/ax_cc_maxopt.m4 \ $(top_srcdir)/m4/ax_check_compiler_flags.m4 \ $(top_srcdir)/m4/ax_compiler_vendor.m4 \ $(top_srcdir)/m4/ax_gcc_aligns_stack.m4 \ $(top_srcdir)/m4/ax_gcc_version.m4 \ $(top_srcdir)/m4/ax_openmp.m4 $(top_srcdir)/m4/libtool.m4 \ $(top_srcdir)/m4/ltoptions.m4 $(top_srcdir)/m4/ltsugar.m4 \ $(top_srcdir)/m4/ltversion.m4 $(top_srcdir)/m4/lt~obsolete.m4 \ $(top_srcdir)/configure.ac am__configure_deps = $(am__aclocal_m4_deps) $(CONFIGURE_DEPENDENCIES) \ $(ACLOCAL_M4) mkinstalldirs = $(install_sh) -d CONFIG_HEADER = $(top_builddir)/config.h CONFIG_CLEAN_FILES = CONFIG_CLEAN_VPATH_FILES = LIBRARIES = $(noinst_LIBRARIES) ARFLAGS = cru libbench2_a_AR = $(AR) $(ARFLAGS) libbench2_a_LIBADD = am_libbench2_a_OBJECTS = after-ccopy-from.$(OBJEXT) \ after-ccopy-to.$(OBJEXT) after-hccopy-from.$(OBJEXT) \ after-hccopy-to.$(OBJEXT) after-rcopy-from.$(OBJEXT) \ after-rcopy-to.$(OBJEXT) allocate.$(OBJEXT) aset.$(OBJEXT) \ bench-cost-postprocess.$(OBJEXT) bench-exit.$(OBJEXT) \ bench-main.$(OBJEXT) can-do.$(OBJEXT) caset.$(OBJEXT) \ dotens2.$(OBJEXT) info.$(OBJEXT) main.$(OBJEXT) \ mflops.$(OBJEXT) mp.$(OBJEXT) ovtpvt.$(OBJEXT) pow2.$(OBJEXT) \ problem.$(OBJEXT) report.$(OBJEXT) speed.$(OBJEXT) \ tensor.$(OBJEXT) timer.$(OBJEXT) useropt.$(OBJEXT) \ util.$(OBJEXT) verify-dft.$(OBJEXT) verify-lib.$(OBJEXT) \ verify-r2r.$(OBJEXT) verify-rdft2.$(OBJEXT) verify.$(OBJEXT) \ zero.$(OBJEXT) my-getopt.$(OBJEXT) libbench2_a_OBJECTS = $(am_libbench2_a_OBJECTS) DEFAULT_INCLUDES = -I.@am__isrc@ -I$(top_builddir) depcomp = $(SHELL) $(top_srcdir)/depcomp am__depfiles_maybe = depfiles am__mv = mv -f COMPILE = $(CC) $(DEFS) $(DEFAULT_INCLUDES) $(INCLUDES) $(AM_CPPFLAGS) \ $(CPPFLAGS) $(AM_CFLAGS) $(CFLAGS) LTCOMPILE = $(LIBTOOL) --tag=CC $(AM_LIBTOOLFLAGS) $(LIBTOOLFLAGS) \ --mode=compile $(CC) $(DEFS) $(DEFAULT_INCLUDES) $(INCLUDES) \ $(AM_CPPFLAGS) $(CPPFLAGS) $(AM_CFLAGS) $(CFLAGS) CCLD = $(CC) LINK = $(LIBTOOL) --tag=CC $(AM_LIBTOOLFLAGS) $(LIBTOOLFLAGS) \ --mode=link $(CCLD) $(AM_CFLAGS) $(CFLAGS) $(AM_LDFLAGS) \ $(LDFLAGS) -o $@ SOURCES = $(libbench2_a_SOURCES) DIST_SOURCES = $(libbench2_a_SOURCES) am__can_run_installinfo = \ case $$AM_UPDATE_INFO_DIR in \ n|no|NO) false;; \ *) (install-info --version) >/dev/null 2>&1;; \ esac ETAGS = etags CTAGS = ctags DISTFILES = $(DIST_COMMON) $(DIST_SOURCES) $(TEXINFOS) $(EXTRA_DIST) ACLOCAL = @ACLOCAL@ ALLOCA = @ALLOCA@ ALTIVEC_CFLAGS = @ALTIVEC_CFLAGS@ AMTAR = @AMTAR@ AR = @AR@ AS = @AS@ AUTOCONF = @AUTOCONF@ AUTOHEADER = @AUTOHEADER@ AUTOMAKE = @AUTOMAKE@ AVX_CFLAGS = @AVX_CFLAGS@ AWK = @AWK@ CC = @CC@ CCDEPMODE = @CCDEPMODE@ CFLAGS = @CFLAGS@ CHECK_PL_OPTS = @CHECK_PL_OPTS@ CPP = @CPP@ CPPFLAGS = @CPPFLAGS@ CYGPATH_W = @CYGPATH_W@ C_FFTW_R2R_KIND = @C_FFTW_R2R_KIND@ C_MPI_FINT = @C_MPI_FINT@ DEFS = @DEFS@ DEPDIR = @DEPDIR@ DLLTOOL = @DLLTOOL@ DSYMUTIL = @DSYMUTIL@ DUMPBIN = @DUMPBIN@ ECHO_C = @ECHO_C@ ECHO_N = @ECHO_N@ ECHO_T = @ECHO_T@ EGREP = @EGREP@ EXEEXT = @EXEEXT@ F77 = @F77@ FFLAGS = @FFLAGS@ FGREP = @FGREP@ FLIBS = @FLIBS@ GREP = @GREP@ INSTALL = @INSTALL@ INSTALL_DATA = @INSTALL_DATA@ INSTALL_PROGRAM = @INSTALL_PROGRAM@ INSTALL_SCRIPT = @INSTALL_SCRIPT@ INSTALL_STRIP_PROGRAM = @INSTALL_STRIP_PROGRAM@ LD = @LD@ LDFLAGS = @LDFLAGS@ LIBOBJS = @LIBOBJS@ LIBQUADMATH = @LIBQUADMATH@ LIBS = @LIBS@ LIBTOOL = @LIBTOOL@ LIPO = @LIPO@ LN_S = @LN_S@ LTLIBOBJS = @LTLIBOBJS@ MAINT = @MAINT@ MAKEINFO = @MAKEINFO@ MANIFEST_TOOL = @MANIFEST_TOOL@ MKDIR_P = @MKDIR_P@ MPICC = @MPICC@ MPILIBS = @MPILIBS@ MPIRUN = @MPIRUN@ NEON_CFLAGS = @NEON_CFLAGS@ NM = @NM@ NMEDIT = @NMEDIT@ OBJDUMP = @OBJDUMP@ OBJEXT = @OBJEXT@ OCAMLBUILD = @OCAMLBUILD@ OPENMP_CFLAGS = @OPENMP_CFLAGS@ OTOOL = @OTOOL@ OTOOL64 = @OTOOL64@ PACKAGE = @PACKAGE@ PACKAGE_BUGREPORT = @PACKAGE_BUGREPORT@ PACKAGE_NAME = @PACKAGE_NAME@ PACKAGE_STRING = @PACKAGE_STRING@ PACKAGE_TARNAME = @PACKAGE_TARNAME@ PACKAGE_URL = @PACKAGE_URL@ PACKAGE_VERSION = @PACKAGE_VERSION@ PATH_SEPARATOR = @PATH_SEPARATOR@ POW_LIB = @POW_LIB@ PRECISION = @PRECISION@ PREC_SUFFIX = @PREC_SUFFIX@ PTHREAD_CC = @PTHREAD_CC@ PTHREAD_CFLAGS = @PTHREAD_CFLAGS@ PTHREAD_LIBS = @PTHREAD_LIBS@ RANLIB = @RANLIB@ SED = @SED@ SET_MAKE = @SET_MAKE@ SHARED_VERSION_INFO = @SHARED_VERSION_INFO@ SHELL = @SHELL@ SSE2_CFLAGS = @SSE2_CFLAGS@ STACK_ALIGN_CFLAGS = @STACK_ALIGN_CFLAGS@ STRIP = @STRIP@ THREADLIBS = @THREADLIBS@ VERSION = @VERSION@ abs_builddir = @abs_builddir@ abs_srcdir = @abs_srcdir@ abs_top_builddir = @abs_top_builddir@ abs_top_srcdir = @abs_top_srcdir@ ac_ct_AR = @ac_ct_AR@ ac_ct_CC = @ac_ct_CC@ ac_ct_DUMPBIN = @ac_ct_DUMPBIN@ ac_ct_F77 = @ac_ct_F77@ acx_pthread_config = @acx_pthread_config@ am__include = @am__include@ am__leading_dot = @am__leading_dot@ am__quote = @am__quote@ am__tar = @am__tar@ am__untar = @am__untar@ bindir = @bindir@ build = @build@ build_alias = @build_alias@ build_cpu = @build_cpu@ build_os = @build_os@ build_vendor = @build_vendor@ builddir = @builddir@ datadir = @datadir@ datarootdir = @datarootdir@ docdir = @docdir@ dvidir = @dvidir@ exec_prefix = @exec_prefix@ host = @host@ host_alias = @host_alias@ host_cpu = @host_cpu@ host_os = @host_os@ host_vendor = @host_vendor@ htmldir = @htmldir@ includedir = @includedir@ infodir = @infodir@ install_sh = @install_sh@ libdir = @libdir@ libexecdir = @libexecdir@ localedir = @localedir@ localstatedir = @localstatedir@ mandir = @mandir@ mkdir_p = @mkdir_p@ oldincludedir = @oldincludedir@ pdfdir = @pdfdir@ prefix = @prefix@ program_transform_name = @program_transform_name@ psdir = @psdir@ sbindir = @sbindir@ sharedstatedir = @sharedstatedir@ srcdir = @srcdir@ sysconfdir = @sysconfdir@ target_alias = @target_alias@ top_build_prefix = @top_build_prefix@ top_builddir = @top_builddir@ top_srcdir = @top_srcdir@ noinst_LIBRARIES = libbench2.a libbench2_a_SOURCES = after-ccopy-from.c after-ccopy-to.c \ after-hccopy-from.c after-hccopy-to.c after-rcopy-from.c \ after-rcopy-to.c allocate.c aset.c bench-cost-postprocess.c \ bench-exit.c bench-main.c can-do.c caset.c dotens2.c info.c main.c \ mflops.c mp.c ovtpvt.c pow2.c problem.c report.c speed.c tensor.c \ timer.c useropt.c util.c verify-dft.c verify-lib.c verify-r2r.c \ verify-rdft2.c verify.c zero.c bench-user.h bench.h verify.h \ my-getopt.c my-getopt.h all: all-am .SUFFIXES: .SUFFIXES: .c .lo .o .obj $(srcdir)/Makefile.in: @MAINTAINER_MODE_TRUE@ $(srcdir)/Makefile.am $(am__configure_deps) @for dep in $?; do \ case '$(am__configure_deps)' in \ *$$dep*) \ ( cd $(top_builddir) && $(MAKE) $(AM_MAKEFLAGS) am--refresh ) \ && { if test -f $@; then exit 0; else break; fi; }; \ exit 1;; \ esac; \ done; \ echo ' cd $(top_srcdir) && $(AUTOMAKE) --gnu libbench2/Makefile'; \ $(am__cd) $(top_srcdir) && \ $(AUTOMAKE) --gnu libbench2/Makefile .PRECIOUS: Makefile Makefile: $(srcdir)/Makefile.in $(top_builddir)/config.status @case '$?' in \ *config.status*) \ cd $(top_builddir) && $(MAKE) $(AM_MAKEFLAGS) am--refresh;; \ *) \ echo ' cd $(top_builddir) && $(SHELL) ./config.status $(subdir)/$@ $(am__depfiles_maybe)'; \ cd $(top_builddir) && $(SHELL) ./config.status $(subdir)/$@ $(am__depfiles_maybe);; \ esac; $(top_builddir)/config.status: $(top_srcdir)/configure $(CONFIG_STATUS_DEPENDENCIES) cd $(top_builddir) && $(MAKE) $(AM_MAKEFLAGS) am--refresh $(top_srcdir)/configure: @MAINTAINER_MODE_TRUE@ $(am__configure_deps) cd $(top_builddir) && $(MAKE) $(AM_MAKEFLAGS) am--refresh $(ACLOCAL_M4): @MAINTAINER_MODE_TRUE@ $(am__aclocal_m4_deps) cd $(top_builddir) && $(MAKE) $(AM_MAKEFLAGS) am--refresh $(am__aclocal_m4_deps): clean-noinstLIBRARIES: -test -z "$(noinst_LIBRARIES)" || rm -f $(noinst_LIBRARIES) libbench2.a: $(libbench2_a_OBJECTS) $(libbench2_a_DEPENDENCIES) $(EXTRA_libbench2_a_DEPENDENCIES) -rm -f libbench2.a $(libbench2_a_AR) libbench2.a $(libbench2_a_OBJECTS) $(libbench2_a_LIBADD) $(RANLIB) libbench2.a mostlyclean-compile: -rm -f *.$(OBJEXT) distclean-compile: -rm -f *.tab.c @AMDEP_TRUE@@am__include@ @am__quote@./$(DEPDIR)/after-ccopy-from.Po@am__quote@ @AMDEP_TRUE@@am__include@ @am__quote@./$(DEPDIR)/after-ccopy-to.Po@am__quote@ @AMDEP_TRUE@@am__include@ @am__quote@./$(DEPDIR)/after-hccopy-from.Po@am__quote@ @AMDEP_TRUE@@am__include@ @am__quote@./$(DEPDIR)/after-hccopy-to.Po@am__quote@ @AMDEP_TRUE@@am__include@ @am__quote@./$(DEPDIR)/after-rcopy-from.Po@am__quote@ @AMDEP_TRUE@@am__include@ @am__quote@./$(DEPDIR)/after-rcopy-to.Po@am__quote@ @AMDEP_TRUE@@am__include@ @am__quote@./$(DEPDIR)/allocate.Po@am__quote@ @AMDEP_TRUE@@am__include@ @am__quote@./$(DEPDIR)/aset.Po@am__quote@ @AMDEP_TRUE@@am__include@ @am__quote@./$(DEPDIR)/bench-cost-postprocess.Po@am__quote@ @AMDEP_TRUE@@am__include@ @am__quote@./$(DEPDIR)/bench-exit.Po@am__quote@ @AMDEP_TRUE@@am__include@ @am__quote@./$(DEPDIR)/bench-main.Po@am__quote@ @AMDEP_TRUE@@am__include@ @am__quote@./$(DEPDIR)/can-do.Po@am__quote@ @AMDEP_TRUE@@am__include@ @am__quote@./$(DEPDIR)/caset.Po@am__quote@ @AMDEP_TRUE@@am__include@ @am__quote@./$(DEPDIR)/dotens2.Po@am__quote@ @AMDEP_TRUE@@am__include@ @am__quote@./$(DEPDIR)/info.Po@am__quote@ @AMDEP_TRUE@@am__include@ @am__quote@./$(DEPDIR)/main.Po@am__quote@ @AMDEP_TRUE@@am__include@ @am__quote@./$(DEPDIR)/mflops.Po@am__quote@ @AMDEP_TRUE@@am__include@ @am__quote@./$(DEPDIR)/mp.Po@am__quote@ @AMDEP_TRUE@@am__include@ @am__quote@./$(DEPDIR)/my-getopt.Po@am__quote@ @AMDEP_TRUE@@am__include@ @am__quote@./$(DEPDIR)/ovtpvt.Po@am__quote@ @AMDEP_TRUE@@am__include@ @am__quote@./$(DEPDIR)/pow2.Po@am__quote@ @AMDEP_TRUE@@am__include@ @am__quote@./$(DEPDIR)/problem.Po@am__quote@ @AMDEP_TRUE@@am__include@ @am__quote@./$(DEPDIR)/report.Po@am__quote@ @AMDEP_TRUE@@am__include@ @am__quote@./$(DEPDIR)/speed.Po@am__quote@ @AMDEP_TRUE@@am__include@ @am__quote@./$(DEPDIR)/tensor.Po@am__quote@ @AMDEP_TRUE@@am__include@ @am__quote@./$(DEPDIR)/timer.Po@am__quote@ @AMDEP_TRUE@@am__include@ @am__quote@./$(DEPDIR)/useropt.Po@am__quote@ @AMDEP_TRUE@@am__include@ @am__quote@./$(DEPDIR)/util.Po@am__quote@ @AMDEP_TRUE@@am__include@ @am__quote@./$(DEPDIR)/verify-dft.Po@am__quote@ @AMDEP_TRUE@@am__include@ @am__quote@./$(DEPDIR)/verify-lib.Po@am__quote@ @AMDEP_TRUE@@am__include@ @am__quote@./$(DEPDIR)/verify-r2r.Po@am__quote@ @AMDEP_TRUE@@am__include@ @am__quote@./$(DEPDIR)/verify-rdft2.Po@am__quote@ @AMDEP_TRUE@@am__include@ @am__quote@./$(DEPDIR)/verify.Po@am__quote@ @AMDEP_TRUE@@am__include@ @am__quote@./$(DEPDIR)/zero.Po@am__quote@ .c.o: @am__fastdepCC_TRUE@ $(COMPILE) -MT $@ -MD -MP -MF $(DEPDIR)/$*.Tpo -c -o $@ $< @am__fastdepCC_TRUE@ $(am__mv) $(DEPDIR)/$*.Tpo $(DEPDIR)/$*.Po @AMDEP_TRUE@@am__fastdepCC_FALSE@ source='$<' object='$@' libtool=no @AMDEPBACKSLASH@ @AMDEP_TRUE@@am__fastdepCC_FALSE@ DEPDIR=$(DEPDIR) $(CCDEPMODE) $(depcomp) @AMDEPBACKSLASH@ @am__fastdepCC_FALSE@ $(COMPILE) -c $< .c.obj: @am__fastdepCC_TRUE@ $(COMPILE) -MT $@ -MD -MP -MF $(DEPDIR)/$*.Tpo -c -o $@ `$(CYGPATH_W) '$<'` @am__fastdepCC_TRUE@ $(am__mv) $(DEPDIR)/$*.Tpo $(DEPDIR)/$*.Po @AMDEP_TRUE@@am__fastdepCC_FALSE@ source='$<' object='$@' libtool=no @AMDEPBACKSLASH@ @AMDEP_TRUE@@am__fastdepCC_FALSE@ DEPDIR=$(DEPDIR) $(CCDEPMODE) $(depcomp) @AMDEPBACKSLASH@ @am__fastdepCC_FALSE@ $(COMPILE) -c `$(CYGPATH_W) '$<'` .c.lo: @am__fastdepCC_TRUE@ $(LTCOMPILE) -MT $@ -MD -MP -MF $(DEPDIR)/$*.Tpo -c -o $@ $< @am__fastdepCC_TRUE@ $(am__mv) $(DEPDIR)/$*.Tpo $(DEPDIR)/$*.Plo @AMDEP_TRUE@@am__fastdepCC_FALSE@ source='$<' object='$@' libtool=yes @AMDEPBACKSLASH@ @AMDEP_TRUE@@am__fastdepCC_FALSE@ DEPDIR=$(DEPDIR) $(CCDEPMODE) $(depcomp) @AMDEPBACKSLASH@ @am__fastdepCC_FALSE@ $(LTCOMPILE) -c -o $@ $< mostlyclean-libtool: -rm -f *.lo clean-libtool: -rm -rf .libs _libs ID: $(HEADERS) $(SOURCES) $(LISP) $(TAGS_FILES) list='$(SOURCES) $(HEADERS) $(LISP) $(TAGS_FILES)'; \ unique=`for i in $$list; do \ if test -f "$$i"; then echo $$i; else echo $(srcdir)/$$i; fi; \ done | \ $(AWK) '{ files[$$0] = 1; nonempty = 1; } \ END { if (nonempty) { for (i in files) print i; }; }'`; \ mkid -fID $$unique tags: TAGS TAGS: $(HEADERS) $(SOURCES) $(TAGS_DEPENDENCIES) \ $(TAGS_FILES) $(LISP) set x; \ here=`pwd`; \ list='$(SOURCES) $(HEADERS) $(LISP) $(TAGS_FILES)'; \ unique=`for i in $$list; do \ if test -f "$$i"; then echo $$i; else echo $(srcdir)/$$i; fi; \ done | \ $(AWK) '{ files[$$0] = 1; nonempty = 1; } \ END { if (nonempty) { for (i in files) print i; }; }'`; \ shift; \ if test -z "$(ETAGS_ARGS)$$*$$unique"; then :; else \ test -n "$$unique" || unique=$$empty_fix; \ if test $$# -gt 0; then \ $(ETAGS) $(ETAGSFLAGS) $(AM_ETAGSFLAGS) $(ETAGS_ARGS) \ "$$@" $$unique; \ else \ $(ETAGS) $(ETAGSFLAGS) $(AM_ETAGSFLAGS) $(ETAGS_ARGS) \ $$unique; \ fi; \ fi ctags: CTAGS CTAGS: $(HEADERS) $(SOURCES) $(TAGS_DEPENDENCIES) \ $(TAGS_FILES) $(LISP) list='$(SOURCES) $(HEADERS) $(LISP) $(TAGS_FILES)'; \ unique=`for i in $$list; do \ if test -f "$$i"; then echo $$i; else echo $(srcdir)/$$i; fi; \ done | \ $(AWK) '{ files[$$0] = 1; nonempty = 1; } \ END { if (nonempty) { for (i in files) print i; }; }'`; \ test -z "$(CTAGS_ARGS)$$unique" \ || $(CTAGS) $(CTAGSFLAGS) $(AM_CTAGSFLAGS) $(CTAGS_ARGS) \ $$unique GTAGS: here=`$(am__cd) $(top_builddir) && pwd` \ && $(am__cd) $(top_srcdir) \ && gtags -i $(GTAGS_ARGS) "$$here" distclean-tags: -rm -f TAGS ID GTAGS GRTAGS GSYMS GPATH tags distdir: $(DISTFILES) @srcdirstrip=`echo "$(srcdir)" | sed 's/[].[^$$\\*]/\\\\&/g'`; \ topsrcdirstrip=`echo "$(top_srcdir)" | sed 's/[].[^$$\\*]/\\\\&/g'`; \ list='$(DISTFILES)'; \ dist_files=`for file in $$list; do echo $$file; done | \ sed -e "s|^$$srcdirstrip/||;t" \ -e "s|^$$topsrcdirstrip/|$(top_builddir)/|;t"`; \ case $$dist_files in \ */*) $(MKDIR_P) `echo "$$dist_files" | \ sed '/\//!d;s|^|$(distdir)/|;s,/[^/]*$$,,' | \ sort -u` ;; \ esac; \ for file in $$dist_files; do \ if test -f $$file || test -d $$file; then d=.; else d=$(srcdir); fi; \ if test -d $$d/$$file; then \ dir=`echo "/$$file" | sed -e 's,/[^/]*$$,,'`; \ if test -d "$(distdir)/$$file"; then \ find "$(distdir)/$$file" -type d ! -perm -700 -exec chmod u+rwx {} \;; \ fi; \ if test -d $(srcdir)/$$file && test $$d != $(srcdir); then \ cp -fpR $(srcdir)/$$file "$(distdir)$$dir" || exit 1; \ find "$(distdir)/$$file" -type d ! -perm -700 -exec chmod u+rwx {} \;; \ fi; \ cp -fpR $$d/$$file "$(distdir)$$dir" || exit 1; \ else \ test -f "$(distdir)/$$file" \ || cp -p $$d/$$file "$(distdir)/$$file" \ || exit 1; \ fi; \ done check-am: all-am check: check-am all-am: Makefile $(LIBRARIES) installdirs: install: install-am install-exec: install-exec-am install-data: install-data-am uninstall: uninstall-am install-am: all-am @$(MAKE) $(AM_MAKEFLAGS) install-exec-am install-data-am installcheck: installcheck-am install-strip: if test -z '$(STRIP)'; then \ $(MAKE) $(AM_MAKEFLAGS) INSTALL_PROGRAM="$(INSTALL_STRIP_PROGRAM)" \ install_sh_PROGRAM="$(INSTALL_STRIP_PROGRAM)" INSTALL_STRIP_FLAG=-s \ install; \ else \ $(MAKE) $(AM_MAKEFLAGS) INSTALL_PROGRAM="$(INSTALL_STRIP_PROGRAM)" \ install_sh_PROGRAM="$(INSTALL_STRIP_PROGRAM)" INSTALL_STRIP_FLAG=-s \ "INSTALL_PROGRAM_ENV=STRIPPROG='$(STRIP)'" install; \ fi mostlyclean-generic: clean-generic: distclean-generic: -test -z "$(CONFIG_CLEAN_FILES)" || rm -f $(CONFIG_CLEAN_FILES) -test . = "$(srcdir)" || test -z "$(CONFIG_CLEAN_VPATH_FILES)" || rm -f $(CONFIG_CLEAN_VPATH_FILES) maintainer-clean-generic: @echo "This command is intended for maintainers to use" @echo "it deletes files that may require special tools to rebuild." clean: clean-am clean-am: clean-generic clean-libtool clean-noinstLIBRARIES \ mostlyclean-am distclean: distclean-am -rm -rf ./$(DEPDIR) -rm -f Makefile distclean-am: clean-am distclean-compile distclean-generic \ distclean-tags dvi: dvi-am dvi-am: html: html-am html-am: info: info-am info-am: install-data-am: install-dvi: install-dvi-am install-dvi-am: install-exec-am: install-html: install-html-am install-html-am: install-info: install-info-am install-info-am: install-man: install-pdf: install-pdf-am install-pdf-am: install-ps: install-ps-am install-ps-am: installcheck-am: maintainer-clean: maintainer-clean-am -rm -rf ./$(DEPDIR) -rm -f Makefile maintainer-clean-am: distclean-am maintainer-clean-generic mostlyclean: mostlyclean-am mostlyclean-am: mostlyclean-compile mostlyclean-generic \ mostlyclean-libtool pdf: pdf-am pdf-am: ps: ps-am ps-am: uninstall-am: .MAKE: install-am install-strip .PHONY: CTAGS GTAGS all all-am check check-am clean clean-generic \ clean-libtool clean-noinstLIBRARIES ctags distclean \ distclean-compile distclean-generic distclean-libtool \ distclean-tags distdir dvi dvi-am html html-am info info-am \ install install-am install-data install-data-am install-dvi \ install-dvi-am install-exec install-exec-am install-html \ install-html-am install-info install-info-am install-man \ install-pdf install-pdf-am install-ps install-ps-am \ install-strip installcheck installcheck-am installdirs \ maintainer-clean maintainer-clean-generic mostlyclean \ mostlyclean-compile mostlyclean-generic mostlyclean-libtool \ pdf pdf-am ps ps-am tags uninstall uninstall-am benchmark: all @echo "nothing to benchmark" accuracy: all @echo "nothing to benchmark" # Tell versions [3.59,3.63) of GNU make to not export all variables. # Otherwise a system limit (for SysV at least) may be exceeded. .NOEXPORT: fftw-3.3.3/libbench2/after-ccopy-from.c0000644000175000017500000000035012054410050014545 00000000000000/* not worth copyrighting */ #include "bench.h" /* default routine, can be overridden by user */ void after_problem_ccopy_from(bench_problem *p, bench_real *ri, bench_real *ii) { UNUSED(p); UNUSED(ri); UNUSED(ii); } fftw-3.3.3/libbench2/after-ccopy-to.c0000644000175000017500000000034612054410050014231 00000000000000/* not worth copyrighting */ #include "bench.h" /* default routine, can be overridden by user */ void after_problem_ccopy_to(bench_problem *p, bench_real *ro, bench_real *io) { UNUSED(p); UNUSED(ro); UNUSED(io); } fftw-3.3.3/libbench2/after-hccopy-from.c0000644000175000017500000000035112054410050014716 00000000000000/* not worth copyrighting */ #include "bench.h" /* default routine, can be overridden by user */ void after_problem_hccopy_from(bench_problem *p, bench_real *ri, bench_real *ii) { UNUSED(p); UNUSED(ri); UNUSED(ii); } fftw-3.3.3/libbench2/after-hccopy-to.c0000644000175000017500000000034712054410050014402 00000000000000/* not worth copyrighting */ #include "bench.h" /* default routine, can be overridden by user */ void after_problem_hccopy_to(bench_problem *p, bench_real *ro, bench_real *io) { UNUSED(p); UNUSED(ro); UNUSED(io); } fftw-3.3.3/libbench2/after-rcopy-from.c0000644000175000017500000000030712054410050014566 00000000000000/* not worth copyrighting */ #include "bench.h" /* default routine, can be overridden by user */ void after_problem_rcopy_from(bench_problem *p, bench_real *ri) { UNUSED(p); UNUSED(ri); } fftw-3.3.3/libbench2/after-rcopy-to.c0000644000175000017500000000030512054410050014243 00000000000000/* not worth copyrighting */ #include "bench.h" /* default routine, can be overridden by user */ void after_problem_rcopy_to(bench_problem *p, bench_real *ro) { UNUSED(p); UNUSED(ro); } fftw-3.3.3/libbench2/allocate.c0000644000175000017500000000550512054410050013163 00000000000000/* not worth copyrighting */ #include "bench.h" static void bounds(bench_problem *p, int *ilb, int *iub, int *olb, int *oub) { bench_tensor *t = tensor_append(p->sz, p->vecsz); tensor_ibounds(t, ilb, iub); tensor_obounds(t, olb, oub); tensor_destroy(t); } /* * Allocate I/O arrays for a problem. * * This is the default routine that can be overridden by the user in * complicated cases. */ void problem_alloc(bench_problem *p) { int ilb, iub, olb, oub; int isz, osz; bounds(p, &ilb, &iub, &olb, &oub); isz = iub - ilb; osz = oub - olb; if (p->kind == PROBLEM_COMPLEX) { bench_complex *in, *out; p->iphyssz = isz; p->inphys = in = (bench_complex *) bench_malloc(isz * sizeof(bench_complex)); p->in = in - ilb; if (p->in_place) { p->out = p->in; p->outphys = p->inphys; p->ophyssz = p->iphyssz; } else { p->ophyssz = osz; p->outphys = out = (bench_complex *) bench_malloc(osz * sizeof(bench_complex)); p->out = out - olb; } } else if (p->kind == PROBLEM_R2R) { bench_real *in, *out; p->iphyssz = isz; p->inphys = in = (bench_real *) bench_malloc(isz * sizeof(bench_real)); p->in = in - ilb; if (p->in_place) { p->out = p->in; p->outphys = p->inphys; p->ophyssz = p->iphyssz; } else { p->ophyssz = osz; p->outphys = out = (bench_real *) bench_malloc(osz * sizeof(bench_real)); p->out = out - olb; } } else if (p->kind == PROBLEM_REAL && p->sign < 0) { /* R2HC */ bench_real *in; bench_complex *out; isz = isz > osz*2 ? isz : osz*2; p->iphyssz = isz; p->inphys = in = (bench_real *) bench_malloc(p->iphyssz * sizeof(bench_real)); p->in = in - ilb; if (p->in_place) { p->out = p->in; p->outphys = p->inphys; p->ophyssz = p->iphyssz / 2; } else { p->ophyssz = osz; p->outphys = out = (bench_complex *) bench_malloc(osz * sizeof(bench_complex)); p->out = out - olb; } } else if (p->kind == PROBLEM_REAL && p->sign > 0) { /* HC2R */ bench_real *out; bench_complex *in; osz = osz > isz*2 ? osz : isz*2; p->ophyssz = osz; p->outphys = out = (bench_real *) bench_malloc(p->ophyssz * sizeof(bench_real)); p->out = out - olb; if (p->in_place) { p->in = p->out; p->inphys = p->outphys; p->iphyssz = p->ophyssz / 2; } else { p->iphyssz = isz; p->inphys = in = (bench_complex *) bench_malloc(isz * sizeof(bench_complex)); p->in = in - ilb; } } else { BENCH_ASSERT(0); /* TODO */ } } void problem_free(bench_problem *p) { if (p->outphys && p->outphys != p->inphys) bench_free(p->outphys); if (p->inphys) bench_free(p->inphys); tensor_destroy(p->sz); tensor_destroy(p->vecsz); } fftw-3.3.3/libbench2/aset.c0000644000175000017500000000023212054410050012323 00000000000000/* not worth copyrighting */ #include "bench.h" void aset(bench_real *A, int n, bench_real x) { int i; for (i = 0; i < n; ++i) A[i] = x; } fftw-3.3.3/libbench2/bench-cost-postprocess.c0000644000175000017500000000024312054410050016000 00000000000000/* not worth copyrighting */ #include "bench.h" /* default routine, can be overridden by user */ double bench_cost_postprocess(double cost) { return cost; } fftw-3.3.3/libbench2/bench-exit.c0000644000175000017500000000022512054410050013417 00000000000000/* not worth copyrighting */ #include "bench.h" /* default routine, can be overridden by user */ void bench_exit(int status) { exit(status); } fftw-3.3.3/libbench2/bench-main.c0000644000175000017500000001115112054410050013372 00000000000000/* * Copyright (c) 2001 Matteo Frigo * Copyright (c) 2001 Massachusetts Institute of Technology * * This program is free software; you can redistribute it and/or modify * it under the terms of the GNU General Public License as published by * the Free Software Foundation; either version 2 of the License, or * (at your option) any later version. * * This program is distributed in the hope that it will be useful, * but WITHOUT ANY WARRANTY; without even the implied warranty of * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the * GNU General Public License for more details. * * You should have received a copy of the GNU General Public License * along with this program; if not, write to the Free Software * Foundation, Inc., 51 Franklin Street, Fifth Floor, Boston, MA 02110-1301 USA * */ #include "bench.h" #include "my-getopt.h" #include #include int verbose; static const struct my_option options[] = { {"accuracy", REQARG, 'a'}, {"accuracy-rounds", REQARG, 405}, {"impulse-accuracy-rounds", REQARG, 406}, {"can-do", REQARG, 'd'}, {"help", NOARG, 'h'}, {"info", REQARG, 'i'}, {"info-all", NOARG, 'I'}, {"print-precision", NOARG, 402}, {"print-time-min", NOARG, 400}, {"random-seed", REQARG, 404}, {"report-benchmark", NOARG, 320}, {"report-mflops", NOARG, 300}, {"report-time", NOARG, 310}, {"report-verbose", NOARG, 330}, {"speed", REQARG, 's'}, {"setup-speed", REQARG, 'S'}, {"time-min", REQARG, 't'}, {"time-repeat", REQARG, 'r'}, {"user-option", REQARG, 'o'}, {"verbose", OPTARG, 'v'}, {"verify", REQARG, 'y'}, {"verify-rounds", REQARG, 401}, {"verify-tolerance", REQARG, 403}, {0, NOARG, 0} }; int bench_main(int argc, char *argv[]) { double tmin = 0.0; double tol; int repeat = 0; int rounds = 10; int iarounds = 0; int arounds = 1; /* this is too low for precise results */ int c; report = report_verbose; /* default */ verbose = 0; tol = SINGLE_PRECISION ? 1.0e-3 : (QUAD_PRECISION ? 1e-29 : 1.0e-10); main_init(&argc, &argv); bench_srand(1); while ((c = my_getopt (argc, argv, options)) != -1) { switch (c) { case 't' : tmin = strtod(my_optarg, 0); break; case 'r': repeat = atoi(my_optarg); break; case 's': timer_init(tmin, repeat); speed(my_optarg, 0); break; case 'S': timer_init(tmin, repeat); speed(my_optarg, 1); break; case 'd': report_can_do(my_optarg); break; case 'o': useropt(my_optarg); break; case 'v': if (verbose >= 0) { /* verbose < 0 disables output */ if (my_optarg) verbose = atoi(my_optarg); else ++verbose; } break; case 'y': verify(my_optarg, rounds, tol); break; case 'a': accuracy(my_optarg, arounds, iarounds); break; case 'i': report_info(my_optarg); break; case 'I': report_info_all(); break; case 'h': if (verbose >= 0) my_usage(argv[0], options); break; case 300: /* --report-mflops */ report = report_mflops; break; case 310: /* --report-time */ report = report_time; break; case 320: /* --report-benchmark */ report = report_benchmark; break; case 330: /* --report-verbose */ report = report_verbose; break; case 400: /* --print-time-min */ timer_init(tmin, repeat); ovtpvt("%g\n", time_min); break; case 401: /* --verify-rounds */ rounds = atoi(my_optarg); break; case 402: /* --print-precision */ if (SINGLE_PRECISION) ovtpvt("single\n"); else if (QUAD_PRECISION) ovtpvt("quad\n"); else if (LDOUBLE_PRECISION) ovtpvt("long-double\n"); else if (DOUBLE_PRECISION) ovtpvt("double\n"); else ovtpvt("unknown %d\n", sizeof(bench_real)); break; case 403: /* --verify-tolerance */ tol = strtod(my_optarg, 0); break; case 404: /* --random-seed */ bench_srand(atoi(my_optarg)); break; case 405: /* --accuracy-rounds */ arounds = atoi(my_optarg); break; case 406: /* --impulse-accuracy-rounds */ iarounds = atoi(my_optarg); break; case '?': /* my_getopt() already printed an error message. */ cleanup(); return 1; default: abort (); } } /* assume that any remaining arguments are problems to be benchmarked */ while (my_optind < argc) { timer_init(tmin, repeat); speed(argv[my_optind++], 0); } cleanup(); return 0; } fftw-3.3.3/libbench2/can-do.c0000644000175000017500000000201112054410050012525 00000000000000/* * Copyright (c) 2001 Matteo Frigo * Copyright (c) 2001 Massachusetts Institute of Technology * * This program is free software; you can redistribute it and/or modify * it under the terms of the GNU General Public License as published by * the Free Software Foundation; either version 2 of the License, or * (at your option) any later version. * * This program is distributed in the hope that it will be useful, * but WITHOUT ANY WARRANTY; without even the implied warranty of * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the * GNU General Public License for more details. * * You should have received a copy of the GNU General Public License * along with this program; if not, write to the Free Software * Foundation, Inc., 51 Franklin Street, Fifth Floor, Boston, MA 02110-1301 USA * */ #include "bench.h" #include void report_can_do(const char *param) { bench_problem *p; p = problem_parse(param); ovtpvt("#%c\n", can_do(p) ? 't' : 'f'); problem_destroy(p); } fftw-3.3.3/libbench2/caset.c0000644000175000017500000000031712054410050012472 00000000000000/* not worth copyrighting */ #include "bench.h" void caset(bench_complex *A, int n, bench_complex x) { int i; for (i = 0; i < n; ++i) { c_re(A[i]) = c_re(x); c_im(A[i]) = c_im(x); } } fftw-3.3.3/libbench2/dotens2.c0000644000175000017500000000334212054410050012752 00000000000000/* * Copyright (c) 2003, 2007-11 Matteo Frigo * Copyright (c) 2003, 2007-11 Massachusetts Institute of Technology * * This program is free software; you can redistribute it and/or modify * it under the terms of the GNU General Public License as published by * the Free Software Foundation; either version 2 of the License, or * (at your option) any later version. * * This program is distributed in the hope that it will be useful, * but WITHOUT ANY WARRANTY; without even the implied warranty of * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the * GNU General Public License for more details. * * You should have received a copy of the GNU General Public License * along with this program; if not, write to the Free Software * Foundation, Inc., 51 Franklin Street, Fifth Floor, Boston, MA 02110-1301 USA * */ #include "verify.h" static void recur(int rnk, const bench_iodim *dims0, const bench_iodim *dims1, dotens2_closure *k, int indx0, int ondx0, int indx1, int ondx1) { if (rnk == 0) k->apply(k, indx0, ondx0, indx1, ondx1); else { int i, n = dims0[0].n; int is0 = dims0[0].is; int os0 = dims0[0].os; int is1 = dims1[0].is; int os1 = dims1[0].os; BENCH_ASSERT(n == dims1[0].n); for (i = 0; i < n; ++i) { recur(rnk - 1, dims0 + 1, dims1 + 1, k, indx0, ondx0, indx1, ondx1); indx0 += is0; ondx0 += os0; indx1 += is1; ondx1 += os1; } } } void bench_dotens2(const bench_tensor *sz0, const bench_tensor *sz1, dotens2_closure *k) { BENCH_ASSERT(sz0->rnk == sz1->rnk); if (sz0->rnk == RNK_MINFTY) return; recur(sz0->rnk, sz0->dims, sz1->dims, k, 0, 0, 0, 0); } fftw-3.3.3/libbench2/info.c0000644000175000017500000000312312054410050012324 00000000000000/* * Copyright (c) 2001 Matteo Frigo * Copyright (c) 2001 Massachusetts Institute of Technology * * This program is free software; you can redistribute it and/or modify * it under the terms of the GNU General Public License as published by * the Free Software Foundation; either version 2 of the License, or * (at your option) any later version. * * This program is distributed in the hope that it will be useful, * but WITHOUT ANY WARRANTY; without even the implied warranty of * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the * GNU General Public License for more details. * * You should have received a copy of the GNU General Public License * along with this program; if not, write to the Free Software * Foundation, Inc., 51 Franklin Street, Fifth Floor, Boston, MA 02110-1301 USA * */ #include "bench.h" #include #include void report_info(const char *param) { struct bench_doc *p; for (p = bench_doc; p->key; ++p) { if (!strcmp(param, p->key)) { if (!p->val) p->val = p->f(); ovtpvt("%s\n", p->val); } } } void report_info_all(void) { struct bench_doc *p; /* * TODO: escape quotes? The format is not unambigously * parseable if the info string contains double quotes. */ for (p = bench_doc; p->key; ++p) { if (!p->val) p->val = p->f(); ovtpvt("(%s \"%s\")\n", p->key, p->val); } ovtpvt("(benchmark-precision \"%s\")\n", SINGLE_PRECISION ? "single" : (LDOUBLE_PRECISION ? "long-double" : (QUAD_PRECISION ? "quad" : "double"))); } fftw-3.3.3/libbench2/main.c0000644000175000017500000000250012054410050012313 00000000000000/* * Copyright (c) 2001 Matteo Frigo * Copyright (c) 2001 Massachusetts Institute of Technology * * This program is free software; you can redistribute it and/or modify * it under the terms of the GNU General Public License as published by * the Free Software Foundation; either version 2 of the License, or * (at your option) any later version. * * This program is distributed in the hope that it will be useful, * but WITHOUT ANY WARRANTY; without even the implied warranty of * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the * GNU General Public License for more details. * * You should have received a copy of the GNU General Public License * along with this program; if not, write to the Free Software * Foundation, Inc., 51 Franklin Street, Fifth Floor, Boston, MA 02110-1301 USA * */ #include "bench.h" /* On some systems, we are required to define a dummy main-like routine (called "MAIN__" or something similar in order to link a C main() with the Fortran libraries). This is detected by autoconf; see the autoconf 2.52 or later manual. */ #ifdef F77_DUMMY_MAIN # ifdef __cplusplus extern "C" # endif int F77_DUMMY_MAIN() { return 1; } #endif /* in a separate file so that the user can override it */ int main(int argc, char *argv[]) { return bench_main(argc, argv); } fftw-3.3.3/libbench2/mflops.c0000644000175000017500000000143712054410050012677 00000000000000/* not worth copyrighting */ #include "bench.h" #include double mflops(const bench_problem *p, double t) { int size = tensor_sz(p->sz); int vsize = tensor_sz(p->vecsz); if (size <= 1) /* a copy: just return reals copied / time */ switch (p->kind) { case PROBLEM_COMPLEX: return (2.0 * size * vsize / (t * 1.0e6)); case PROBLEM_REAL: case PROBLEM_R2R: return (1.0 * size * vsize / (t * 1.0e6)); } switch (p->kind) { case PROBLEM_COMPLEX: return (5.0 * size * vsize * log((double)size) / (log(2.0) * t * 1.0e6)); case PROBLEM_REAL: case PROBLEM_R2R: return (2.5 * vsize * size * log((double) size) / (log(2.0) * t * 1.0e6)); } BENCH_ASSERT(0 /* can't happen */); return 0.0; } fftw-3.3.3/libbench2/mp.c0000644000175000017500000003136412054410050012015 00000000000000#include "config.h" #include "bench.h" #include #define DG unsigned short #define ACC unsigned long #define REAL bench_real #define BITS_IN_REAL 53 /* mantissa */ #define SHFT 16 #define RADIX 65536L #define IRADIX (1.0 / RADIX) #define LO(x) ((x) & (RADIX - 1)) #define HI(x) ((x) >> SHFT) #define HI_SIGNED(x) \ ((((x) + (ACC)(RADIX >> 1) * RADIX) >> SHFT) - (RADIX >> 1)) #define ZEROEXP (-32768) #define LEN 10 typedef struct { short sign; short expt; DG d[LEN]; } N[1]; #define EXA a->expt #define EXB b->expt #define EXC c->expt #define AD a->d #define BD b->d #define SGNA a->sign #define SGNB b->sign static const N zero = {{ 1, ZEROEXP, {0} }}; static void cpy(const N a, N b) { *b = *a; } static void fromreal(REAL x, N a) { int i, e; cpy(zero, a); if (x == 0.0) return; if (x >= 0) { SGNA = 1; } else { SGNA = -1; x = -x; } e = 0; while (x >= 1.0) { x *= IRADIX; ++e; } while (x < IRADIX) { x *= RADIX; --e; } EXA = e; for (i = LEN - 1; i >= 0 && x != 0.0; --i) { REAL y; x *= RADIX; y = (REAL) ((int) x); AD[i] = (DG)y; x -= y; } } static void fromshort(int x, N a) { cpy(zero, a); if (x < 0) { x = -x; SGNA = -1; } else { SGNA = 1; } EXA = 1; AD[LEN - 1] = x; } static void pack(DG *d, int e, int s, int l, N a) { int i, j; for (i = l - 1; i >= 0; --i, --e) if (d[i] != 0) break; if (i < 0) { /* number is zero */ cpy(zero, a); } else { EXA = e; SGNA = s; if (i >= LEN - 1) { for (j = LEN - 1; j >= 0; --i, --j) AD[j] = d[i]; } else { for (j = LEN - 1; i >= 0; --i, --j) AD[j] = d[i]; for ( ; j >= 0; --j) AD[j] = 0; } } } /* compare absolute values */ static int abscmp(const N a, const N b) { int i; if (EXA > EXB) return 1; if (EXA < EXB) return -1; for (i = LEN - 1; i >= 0; --i) { if (AD[i] > BD[i]) return 1; if (AD[i] < BD[i]) return -1; } return 0; } static int eq(const N a, const N b) { return (SGNA == SGNB) && (abscmp(a, b) == 0); } /* add magnitudes, for |a| >= |b| */ static void addmag0(int s, const N a, const N b, N c) { int ia, ib; ACC r = 0; DG d[LEN + 1]; for (ia = 0, ib = EXA - EXB; ib < LEN; ++ia, ++ib) { r += (ACC)AD[ia] + (ACC)BD[ib]; d[ia] = LO(r); r = HI(r); } for (; ia < LEN; ++ia) { r += (ACC)AD[ia]; d[ia] = LO(r); r = HI(r); } d[ia] = LO(r); pack(d, EXA + 1, s * SGNA, LEN + 1, c); } static void addmag(int s, const N a, const N b, N c) { if (abscmp(a, b) > 0) addmag0(1, a, b, c); else addmag0(s, b, a, c); } /* subtract magnitudes, for |a| >= |b| */ static void submag0(int s, const N a, const N b, N c) { int ia, ib; ACC r = 0; DG d[LEN]; for (ia = 0, ib = EXA - EXB; ib < LEN; ++ia, ++ib) { r += (ACC)AD[ia] - (ACC)BD[ib]; d[ia] = LO(r); r = HI_SIGNED(r); } for (; ia < LEN; ++ia) { r += (ACC)AD[ia]; d[ia] = LO(r); r = HI_SIGNED(r); } pack(d, EXA, s * SGNA, LEN, c); } static void submag(int s, const N a, const N b, N c) { if (abscmp(a, b) > 0) submag0(1, a, b, c); else submag0(s, b, a, c); } /* c = a + b */ static void add(const N a, const N b, N c) { if (SGNA == SGNB) addmag(1, a, b, c); else submag(1, a, b, c); } static void sub(const N a, const N b, N c) { if (SGNA == SGNB) submag(-1, a, b, c); else addmag(-1, a, b, c); } static void mul(const N a, const N b, N c) { DG d[2 * LEN]; int i, j, k; ACC r; for (i = 0; i < LEN; ++i) d[2 * i] = d[2 * i + 1] = 0; for (i = 0; i < LEN; ++i) { ACC ai = AD[i]; if (ai) { r = 0; for (j = 0, k = i; j < LEN; ++j, ++k) { r += ai * (ACC)BD[j] + (ACC)d[k]; d[k] = LO(r); r = HI(r); } d[k] = LO(r); } } pack(d, EXA + EXB, SGNA * SGNB, 2 * LEN, c); } static REAL toreal(const N a) { REAL h, l, f; int i, bits; ACC r; DG sticky; if (EXA != ZEROEXP) { f = IRADIX; i = LEN; bits = 0; h = (r = AD[--i]) * f; f *= IRADIX; for (bits = 0; r > 0; ++bits) r >>= 1; /* first digit */ while (bits + SHFT <= BITS_IN_REAL) { h += AD[--i] * f; f *= IRADIX; bits += SHFT; } /* guard digit (leave one bit for sticky bit, hence `<' instead of `<=') */ bits = 0; l = 0.0; while (bits + SHFT < BITS_IN_REAL) { l += AD[--i] * f; f *= IRADIX; bits += SHFT; } /* sticky bit */ sticky = 0; while (i > 0) sticky |= AD[--i]; if (sticky) l += (RADIX / 2) * f; h += l; for (i = 0; i < EXA; ++i) h *= (REAL)RADIX; for (i = 0; i > EXA; --i) h *= IRADIX; if (SGNA == -1) h = -h; return h; } else { return 0.0; } } static void neg(N a) { SGNA = -SGNA; } static void inv(const N a, N x) { N w, z, one, two; fromreal(1.0 / toreal(a), x); /* initial guess */ fromshort(1, one); fromshort(2, two); for (;;) { /* Newton */ mul(a, x, w); sub(two, w, z); if (eq(one, z)) break; mul(x, z, x); } } /* 2 pi */ static const N n2pi = {{ 1, 1, {18450, 59017, 1760, 5212, 9779, 4518, 2886, 54545, 18558, 6} }}; /* 1 / 31! */ static const N i31fac = {{ 1, -7, {28087, 45433, 51357, 24545, 14291, 3954, 57879, 8109, 38716, 41382} }}; /* 1 / 32! */ static const N i32fac = {{ 1, -7, {52078, 60811, 3652, 39679, 37310, 47227, 28432, 57597, 13497, 1293} }}; static void msin(const N a, N b) { N a2, g, k; int i; cpy(i31fac, g); cpy(g, b); mul(a, a, a2); /* Taylor */ for (i = 31; i > 1; i -= 2) { fromshort(i * (i - 1), k); mul(k, g, g); mul(a2, b, k); sub(g, k, b); } mul(a, b, b); } static void mcos(const N a, N b) { N a2, g, k; int i; cpy(i32fac, g); cpy(g, b); mul(a, a, a2); /* Taylor */ for (i = 32; i > 0; i -= 2) { fromshort(i * (i - 1), k); mul(k, g, g); mul(a2, b, k); sub(g, k, b); } } static void by2pi(REAL m, REAL n, N a) { N b; fromreal(n, b); inv(b, a); fromreal(m, b); mul(a, b, a); mul(n2pi, a, a); } static void sin2pi(REAL m, REAL n, N a); static void cos2pi(REAL m, REAL n, N a) { N b; if (m < 0) cos2pi(-m, n, a); else if (m > n * 0.5) cos2pi(n - m, n, a); else if (m > n * 0.25) {sin2pi(m - n * 0.25, n, a); neg(a);} else if (m > n * 0.125) sin2pi(n * 0.25 - m, n, a); else { by2pi(m, n, b); mcos(b, a); } } static void sin2pi(REAL m, REAL n, N a) { N b; if (m < 0) {sin2pi(-m, n, a); neg(a);} else if (m > n * 0.5) {sin2pi(n - m, n, a); neg(a);} else if (m > n * 0.25) {cos2pi(m - n * 0.25, n, a);} else if (m > n * 0.125) {cos2pi(n * 0.25 - m, n, a);} else {by2pi(m, n, b); msin(b, a);} } /*----------------------------------------------------------------------*/ /* FFT stuff */ /* (r0 + i i0)(r1 + i i1) */ static void cmul(N r0, N i0, N r1, N i1, N r2, N i2) { N s, t, q; mul(r0, r1, s); mul(i0, i1, t); sub(s, t, q); mul(r0, i1, s); mul(i0, r1, t); add(s, t, i2); cpy(q, r2); } /* (r0 - i i0)(r1 + i i1) */ static void cmulj(N r0, N i0, N r1, N i1, N r2, N i2) { N s, t, q; mul(r0, r1, s); mul(i0, i1, t); add(s, t, q); mul(r0, i1, s); mul(i0, r1, t); sub(s, t, i2); cpy(q, r2); } static void mcexp(int m, int n, N r, N i) { static int cached_n = -1; static N w[64][2]; int k, j; if (n != cached_n) { for (j = 1, k = 0; j < n; j += j, ++k) { cos2pi(j, n, w[k][0]); sin2pi(j, n, w[k][1]); } cached_n = n; } fromshort(1, r); fromshort(0, i); if (m > 0) { for (k = 0; m; ++k, m >>= 1) if (m & 1) cmul(w[k][0], w[k][1], r, i, r, i); } else { m = -m; for (k = 0; m; ++k, m >>= 1) if (m & 1) cmulj(w[k][0], w[k][1], r, i, r, i); } } static void bitrev(int n, N *a) { int i, j, m; for (i = j = 0; i < n - 1; ++i) { if (i < j) { N t; cpy(a[2*i], t); cpy(a[2*j], a[2*i]); cpy(t, a[2*j]); cpy(a[2*i+1], t); cpy(a[2*j+1], a[2*i+1]); cpy(t, a[2*j+1]); } /* bit reversed counter */ m = n; do { m >>= 1; j ^= m; } while (!(j & m)); } } static void fft0(int n, N *a, int sign) { int i, j, k; bitrev(n, a); for (i = 1; i < n; i = 2 * i) { for (j = 0; j < i; ++j) { N wr, wi; mcexp(sign * (int)j, 2 * i, wr, wi); for (k = j; k < n; k += 2 * i) { N *a0 = a + 2 * k; N *a1 = a0 + 2 * i; N r0, i0, r1, i1, t0, t1, xr, xi; cpy(a0[0], r0); cpy(a0[1], i0); cpy(a1[0], r1); cpy(a1[1], i1); mul(r1, wr, t0); mul(i1, wi, t1); sub(t0, t1, xr); mul(r1, wi, t0); mul(i1, wr, t1); add(t0, t1, xi); add(r0, xr, a0[0]); add(i0, xi, a0[1]); sub(r0, xr, a1[0]); sub(i0, xi, a1[1]); } } } } /* a[2*k]+i*a[2*k+1] = exp(2*pi*i*k^2/(2*n)) */ static void bluestein_sequence(int n, N *a) { int k, ksq, n2 = 2 * n; ksq = 1; /* (-1)^2 */ for (k = 0; k < n; ++k) { /* careful with overflow */ ksq = ksq + 2*k - 1; while (ksq > n2) ksq -= n2; mcexp(ksq, n2, a[2*k], a[2*k+1]); } } static int pow2_atleast(int x) { int h; for (h = 1; h < x; h = 2 * h) ; return h; } static N *cached_bluestein_w = 0; static N *cached_bluestein_y = 0; static int cached_bluestein_n = -1; static void bluestein(int n, N *a) { int nb = pow2_atleast(2 * n); N *b = (N *)bench_malloc(2 * nb * sizeof(N)); N *w = cached_bluestein_w; N *y = cached_bluestein_y; N nbinv; int i; fromreal(1.0 / nb, nbinv); /* exact because nb = 2^k */ if (cached_bluestein_n != n) { if (w) bench_free(w); if (y) bench_free(y); w = (N *)bench_malloc(2 * n * sizeof(N)); y = (N *)bench_malloc(2 * nb * sizeof(N)); cached_bluestein_n = n; cached_bluestein_w = w; cached_bluestein_y = y; bluestein_sequence(n, w); for (i = 0; i < 2*nb; ++i) cpy(zero, y[i]); for (i = 0; i < n; ++i) { cpy(w[2*i], y[2*i]); cpy(w[2*i+1], y[2*i+1]); } for (i = 1; i < n; ++i) { cpy(w[2*i], y[2*(nb-i)]); cpy(w[2*i+1], y[2*(nb-i)+1]); } fft0(nb, y, -1); } for (i = 0; i < 2*nb; ++i) cpy(zero, b[i]); for (i = 0; i < n; ++i) cmulj(w[2*i], w[2*i+1], a[2*i], a[2*i+1], b[2*i], b[2*i+1]); /* scaled convolution b * y */ fft0(nb, b, -1); for (i = 0; i < nb; ++i) cmul(b[2*i], b[2*i+1], y[2*i], y[2*i+1], b[2*i], b[2*i+1]); fft0(nb, b, 1); for (i = 0; i < n; ++i) { cmulj(w[2*i], w[2*i+1], b[2*i], b[2*i+1], a[2*i], a[2*i+1]); mul(nbinv, a[2*i], a[2*i]); mul(nbinv, a[2*i+1], a[2*i+1]); } bench_free(b); } static void swapri(int n, N *a) { int i; for (i = 0; i < n; ++i) { N t; cpy(a[2 * i], t); cpy(a[2 * i + 1], a[2 * i]); cpy(t, a[2 * i + 1]); } } static void fft1(int n, N *a, int sign) { if (power_of_two(n)) { fft0(n, a, sign); } else { if (sign == 1) swapri(n, a); bluestein(n, a); if (sign == 1) swapri(n, a); } } static void fromrealv(int n, bench_complex *a, N *b) { int i; for (i = 0; i < n; ++i) { fromreal(c_re(a[i]), b[2 * i]); fromreal(c_im(a[i]), b[2 * i + 1]); } } static void compare(int n, N *a, N *b, double *err) { int i; double e1, e2, einf; double n1, n2, ninf; e1 = e2 = einf = 0.0; n1 = n2 = ninf = 0.0; # define DO(x1, x2, xinf, var) { \ double d = var; \ if (d < 0) d = -d; \ x1 += d; x2 += d * d; if (d > xinf) xinf = d; \ } for (i = 0; i < 2 * n; ++i) { N dd; sub(a[i], b[i], dd); DO(n1, n2, ninf, toreal(a[i])); DO(e1, e2, einf, toreal(dd)); } # undef DO err[0] = e1 / n1; err[1] = sqrt(e2 / n2); err[2] = einf / ninf; } void fftaccuracy(int n, bench_complex *a, bench_complex *ffta, int sign, double err[6]) { N *b = (N *)bench_malloc(2 * n * sizeof(N)); N *fftb = (N *)bench_malloc(2 * n * sizeof(N)); N mn, ninv; int i; fromreal(n, mn); inv(mn, ninv); /* forward error */ fromrealv(n, a, b); fromrealv(n, ffta, fftb); fft1(n, b, sign); compare(n, b, fftb, err); /* backward error */ fromrealv(n, a, b); fromrealv(n, ffta, fftb); for (i = 0; i < 2 * n; ++i) mul(fftb[i], ninv, fftb[i]); fft1(n, fftb, -sign); compare(n, b, fftb, err + 3); bench_free(fftb); bench_free(b); } void fftaccuracy_done(void) { if (cached_bluestein_w) bench_free(cached_bluestein_w); if (cached_bluestein_y) bench_free(cached_bluestein_y); cached_bluestein_w = 0; cached_bluestein_y = 0; cached_bluestein_n = -1; } fftw-3.3.3/libbench2/ovtpvt.c0000644000175000017500000000074212054410050012737 00000000000000#include #include #include #include "bench.h" void ovtpvt(const char *format, ...) { va_list ap; va_start(ap, format); if (verbose >= 0) vfprintf(stdout, format, ap); va_end(ap); fflush(stdout); } void ovtpvt_err(const char *format, ...) { va_list ap; va_start(ap, format); if (verbose >= 0) { fflush(stdout); vfprintf(stderr, format, ap); } va_end(ap); fflush(stdout); } fftw-3.3.3/libbench2/pow2.c0000644000175000017500000000014512054410050012261 00000000000000#include "bench.h" int power_of_two(int n) { return (((n) > 0) && (((n) & ((n) - 1)) == 0)); } fftw-3.3.3/libbench2/problem.c0000644000175000017500000001673412054410050013045 00000000000000/* * Copyright (c) 2001 Matteo Frigo * Copyright (c) 2001 Massachusetts Institute of Technology * * This program is free software; you can redistribute it and/or modify * it under the terms of the GNU General Public License as published by * the Free Software Foundation; either version 2 of the License, or * (at your option) any later version. * * This program is distributed in the hope that it will be useful, * but WITHOUT ANY WARRANTY; without even the implied warranty of * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the * GNU General Public License for more details. * * You should have received a copy of the GNU General Public License * along with this program; if not, write to the Free Software * Foundation, Inc., 51 Franklin Street, Fifth Floor, Boston, MA 02110-1301 USA * */ #include "config.h" #include "bench.h" #include #include #include #include int always_pad_real = 0; /* by default, only pad in-place case */ typedef enum { SAME, PADDED, HALFISH } n_transform; /* funny transformations for last dimension of PROBLEM_REAL */ static int transform_n(int n, n_transform nt) { switch (nt) { case SAME: return n; case PADDED: return 2*(n/2+1); case HALFISH: return (n/2+1); default: BENCH_ASSERT(0); return 0; } } /* do what I mean */ static bench_tensor *dwim(bench_tensor *t, bench_iodim **last_iodim, n_transform nti, n_transform nto, bench_iodim *dt) { int i; bench_iodim *d, *d1; if (!FINITE_RNK(t->rnk) || t->rnk < 1) return t; i = t->rnk; d1 = *last_iodim; while (--i >= 0) { d = t->dims + i; if (!d->is) d->is = d1->is * transform_n(d1->n, d1==dt ? nti : SAME); if (!d->os) d->os = d1->os * transform_n(d1->n, d1==dt ? nto : SAME); d1 = d; } *last_iodim = d1; return t; } static void transpose_tensor(bench_tensor *t) { if (!FINITE_RNK(t->rnk) || t->rnk < 2) return; t->dims[0].os = t->dims[1].os; t->dims[1].os = t->dims[0].os * t->dims[0].n; } static const char *parseint(const char *s, int *n) { int sign = 1; *n = 0; if (*s == '-') { sign = -1; ++s; } else if (*s == '+') { sign = +1; ++s; } BENCH_ASSERT(isdigit(*s)); while (isdigit(*s)) { *n = *n * 10 + (*s - '0'); ++s; } *n *= sign; if (*s == 'k' || *s == 'K') { *n *= 1024; ++s; } if (*s == 'm' || *s == 'M') { *n *= 1024 * 1024; ++s; } return s; } struct dimlist { bench_iodim car; r2r_kind_t k; struct dimlist *cdr; }; static const char *parsetensor(const char *s, bench_tensor **tp, r2r_kind_t **k) { struct dimlist *l = 0, *m; bench_tensor *t; int rnk = 0; L1: m = (struct dimlist *)bench_malloc(sizeof(struct dimlist)); /* nconc onto l */ m->cdr = l; l = m; ++rnk; s = parseint(s, &m->car.n); if (*s == ':') { /* read input stride */ ++s; s = parseint(s, &m->car.is); if (*s == ':') { /* read output stride */ ++s; s = parseint(s, &m->car.os); } else { /* default */ m->car.os = m->car.is; } } else { m->car.is = 0; m->car.os = 0; } if (*s == 'f' || *s == 'F') { m->k = R2R_R2HC; ++s; } else if (*s == 'b' || *s == 'B') { m->k = R2R_HC2R; ++s; } else if (*s == 'h' || *s == 'H') { m->k = R2R_DHT; ++s; } else if (*s == 'e' || *s == 'E' || *s == 'o' || *s == 'O') { char c = *(s++); int ab; s = parseint(s, &ab); if (c == 'e' || c == 'E') { if (ab == 0) m->k = R2R_REDFT00; else if (ab == 1) m->k = R2R_REDFT01; else if (ab == 10) m->k = R2R_REDFT10; else if (ab == 11) m->k = R2R_REDFT11; else BENCH_ASSERT(0); } else { if (ab == 0) m->k = R2R_RODFT00; else if (ab == 1) m->k = R2R_RODFT01; else if (ab == 10) m->k = R2R_RODFT10; else if (ab == 11) m->k = R2R_RODFT11; else BENCH_ASSERT(0); } } else m->k = R2R_R2HC; if (*s == 'x' || *s == 'X') { ++s; goto L1; } /* now we have a dimlist. Build bench_tensor, etc. */ if (k && rnk > 0) { int i; *k = (r2r_kind_t *) bench_malloc(sizeof(r2r_kind_t) * rnk); for (m = l, i = rnk - 1; i >= 0; --i, m = m->cdr) { BENCH_ASSERT(m); (*k)[i] = m->k; } } t = mktensor(rnk); while (--rnk >= 0) { bench_iodim *d = t->dims + rnk; BENCH_ASSERT(l); m = l; l = m->cdr; d->n = m->car.n; d->is = m->car.is; d->os = m->car.os; bench_free(m); } *tp = t; return s; } /* parse a problem description, return a problem */ bench_problem *problem_parse(const char *s) { bench_problem *p; bench_iodim last_iodim0 = {1,1,1}, *last_iodim = &last_iodim0; bench_iodim *sz_last_iodim; bench_tensor *sz; n_transform nti = SAME, nto = SAME; int transpose = 0; p = (bench_problem *) bench_malloc(sizeof(bench_problem)); p->kind = PROBLEM_COMPLEX; p->k = 0; p->sign = -1; p->in = p->out = 0; p->inphys = p->outphys = 0; p->iphyssz = p->ophyssz = 0; p->in_place = 0; p->destroy_input = 0; p->split = 0; p->userinfo = 0; p->scrambled_in = p->scrambled_out = 0; p->sz = p->vecsz = 0; p->ini = p->outi = 0; p->pstring = (char *) bench_malloc(sizeof(char) * (strlen(s) + 1)); strcpy(p->pstring, s); L1: switch (tolower(*s)) { case 'i': p->in_place = 1; ++s; goto L1; case 'o': p->in_place = 0; ++s; goto L1; case 'd': p->destroy_input = 1; ++s; goto L1; case '/': p->split = 1; ++s; goto L1; case 'f': case '-': p->sign = -1; ++s; goto L1; case 'b': case '+': p->sign = 1; ++s; goto L1; case 'r': p->kind = PROBLEM_REAL; ++s; goto L1; case 'c': p->kind = PROBLEM_COMPLEX; ++s; goto L1; case 'k': p->kind = PROBLEM_R2R; ++s; goto L1; case 't': transpose = 1; ++s; goto L1; /* hack for MPI: */ case '[': p->scrambled_in = 1; ++s; goto L1; case ']': p->scrambled_out = 1; ++s; goto L1; default : ; } s = parsetensor(s, &sz, p->kind == PROBLEM_R2R ? &p->k : 0); if (p->kind == PROBLEM_REAL) { if (p->sign < 0) { nti = p->in_place || always_pad_real ? PADDED : SAME; nto = HALFISH; } else { nti = HALFISH; nto = p->in_place || always_pad_real ? PADDED : SAME; } } sz_last_iodim = sz->dims + sz->rnk - 1; if (*s == '*') { /* "external" vector */ ++s; p->sz = dwim(sz, &last_iodim, nti, nto, sz_last_iodim); s = parsetensor(s, &sz, 0); p->vecsz = dwim(sz, &last_iodim, nti, nto, sz_last_iodim); } else if (*s == 'v' || *s == 'V') { /* "internal" vector */ bench_tensor *vecsz; ++s; s = parsetensor(s, &vecsz, 0); p->vecsz = dwim(vecsz, &last_iodim, nti, nto, sz_last_iodim); p->sz = dwim(sz, &last_iodim, nti, nto, sz_last_iodim); } else { p->sz = dwim(sz, &last_iodim, nti, nto, sz_last_iodim); p->vecsz = mktensor(0); } if (transpose) { transpose_tensor(p->sz); transpose_tensor(p->vecsz); } if (!p->in_place) p->out = ((bench_real *) p->in) + (1 << 20); /* whatever */ BENCH_ASSERT(p->sz && p->vecsz); BENCH_ASSERT(!*s); return p; } void problem_destroy(bench_problem *p) { BENCH_ASSERT(p); problem_free(p); bench_free0(p->k); bench_free0(p->pstring); bench_free(p); } fftw-3.3.3/libbench2/report.c0000644000175000017500000000674712054410050012723 00000000000000/* * Copyright (c) 2001 Matteo Frigo * Copyright (c) 2001 Massachusetts Institute of Technology * * This program is free software; you can redistribute it and/or modify * it under the terms of the GNU General Public License as published by * the Free Software Foundation; either version 2 of the License, or * (at your option) any later version. * * This program is distributed in the hope that it will be useful, * but WITHOUT ANY WARRANTY; without even the implied warranty of * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the * GNU General Public License for more details. * * You should have received a copy of the GNU General Public License * along with this program; if not, write to the Free Software * Foundation, Inc., 51 Franklin Street, Fifth Floor, Boston, MA 02110-1301 USA * */ #include "bench.h" #include #include #include void (*report)(const bench_problem *p, double *t, int st); #undef min #undef max /* you never know */ struct stats { double min; double max; double avg; double median; }; static void mkstat(double *t, int st, struct stats *a) { int i, j; a->min = t[0]; a->max = t[0]; a->avg = 0.0; for (i = 0; i < st; ++i) { if (t[i] < a->min) a->min = t[i]; if (t[i] > a->max) a->max = t[i]; a->avg += t[i]; } a->avg /= (double)st; /* compute median --- silly bubblesort algorithm */ for (i = st - 1; i > 1; --i) { for (j = 0; j < i - 1; ++j) { double t0, t1; if ((t0 = t[j]) > (t1 = t[j + 1])) { t[j] = t1; t[j + 1] = t0; } } } a->median = t[st / 2]; } void report_mflops(const bench_problem *p, double *t, int st) { struct stats s; mkstat(t, st, &s); ovtpvt("(%g %g %g %g)\n", mflops(p, s.max), mflops(p, s.avg), mflops(p, s.min), mflops(p, s.median)); } void report_time(const bench_problem *p, double *t, int st) { struct stats s; UNUSED(p); mkstat(t, st, &s); ovtpvt("(%g %g %g %g)\n", s.min, s.avg, s.max, s.median); } void report_benchmark(const bench_problem *p, double *t, int st) { struct stats s; mkstat(t, st, &s); ovtpvt("%.5g %.8g %g\n", mflops(p, s.min), s.min, p->setup_time); } static void sprintf_time(double x, char *buf, int buflen) { #ifdef HAVE_SNPRINTF # define MY_SPRINTF(a, b) snprintf(buf, buflen, a, b) #else # define MY_SPRINTF(a, b) sprintf(buf, a, b) #endif if (x < 1.0E-6) MY_SPRINTF("%.2f ns", x * 1.0E9); else if (x < 1.0E-3) MY_SPRINTF("%.2f us", x * 1.0E6); else if (x < 1.0) MY_SPRINTF("%.2f ms", x * 1.0E3); else MY_SPRINTF("%.2f s", x); #undef MY_SPRINTF } void report_verbose(const bench_problem *p, double *t, int st) { struct stats s; char bmin[64], bmax[64], bavg[64], bmedian[64], btmin[64]; char bsetup[64]; int copyp = tensor_sz(p->sz) == 1; mkstat(t, st, &s); sprintf_time(s.min, bmin, 64); sprintf_time(s.max, bmax, 64); sprintf_time(s.avg, bavg, 64); sprintf_time(s.median, bmedian, 64); sprintf_time(time_min, btmin, 64); sprintf_time(p->setup_time, bsetup, 64); ovtpvt("Problem: %s, setup: %s, time: %s, %s: %.5g\n", p->pstring, bsetup, bmin, copyp ? "fp-move/us" : "``mflops''", mflops(p, s.min)); if (verbose) { ovtpvt("Took %d measurements for at least %s each.\n", st, btmin); ovtpvt("Time: min %s, max %s, avg %s, median %s\n", bmin, bmax, bavg, bmedian); } } fftw-3.3.3/libbench2/speed.c0000644000175000017500000000460212054410050012474 00000000000000/* * Copyright (c) 2001 Matteo Frigo * Copyright (c) 2001 Massachusetts Institute of Technology * * This program is free software; you can redistribute it and/or modify * it under the terms of the GNU General Public License as published by * the Free Software Foundation; either version 2 of the License, or * (at your option) any later version. * * This program is distributed in the hope that it will be useful, * but WITHOUT ANY WARRANTY; without even the implied warranty of * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the * GNU General Public License for more details. * * You should have received a copy of the GNU General Public License * along with this program; if not, write to the Free Software * Foundation, Inc., 51 Franklin Street, Fifth Floor, Boston, MA 02110-1301 USA * */ #include "bench.h" int no_speed_allocation = 0; /* 1 to not allocate array data in speed() */ void speed(const char *param, int setup_only) { double *t; int iter = 0, k; bench_problem *p; double tmin, y; t = (double *) bench_malloc(time_repeat * sizeof(double)); for (k = 0; k < time_repeat; ++k) t[k] = 0; p = problem_parse(param); BENCH_ASSERT(can_do(p)); if (!no_speed_allocation) { problem_alloc(p); problem_zero(p); } timer_start(LIBBENCH_TIMER); setup(p); p->setup_time = bench_cost_postprocess(timer_stop(LIBBENCH_TIMER)); /* reset the input to zero again, because the planner in paranoid mode sets it to random values, thus making the benchmark diverge. */ if (!no_speed_allocation) problem_zero(p); if (setup_only) goto done; start_over: for (iter = 1; iter < (1<<30); iter *= 2) { tmin = 1.0e20; for (k = 0; k < time_repeat; ++k) { timer_start(LIBBENCH_TIMER); doit(iter, p); y = bench_cost_postprocess(timer_stop(LIBBENCH_TIMER)); if (y < 0) /* yes, it happens */ goto start_over; t[k] = y; if (y < tmin) tmin = y; } if (tmin >= time_min) goto done; } goto start_over; /* this also happens */ done: done(p); if (iter) for (k = 0; k < time_repeat; ++k) t[k] /= iter; else for (k = 0; k < time_repeat; ++k) t[k] = 0; report(p, t, time_repeat); if (!no_speed_allocation) problem_destroy(p); bench_free(t); return; } fftw-3.3.3/libbench2/tensor.c0000644000175000017500000001323012054410050012703 00000000000000/* * Copyright (c) 2001 Matteo Frigo * Copyright (c) 2001 Massachusetts Institute of Technology * * This program is free software; you can redistribute it and/or modify * it under the terms of the GNU General Public License as published by * the Free Software Foundation; either version 2 of the License, or * (at your option) any later version. * * This program is distributed in the hope that it will be useful, * but WITHOUT ANY WARRANTY; without even the implied warranty of * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the * GNU General Public License for more details. * * You should have received a copy of the GNU General Public License * along with this program; if not, write to the Free Software * Foundation, Inc., 51 Franklin Street, Fifth Floor, Boston, MA 02110-1301 USA * */ #include "bench.h" #include bench_tensor *mktensor(int rnk) { bench_tensor *x; BENCH_ASSERT(rnk >= 0); x = (bench_tensor *)bench_malloc(sizeof(bench_tensor)); if (FINITE_RNK(rnk) && rnk > 0) x->dims = (bench_iodim *)bench_malloc(sizeof(bench_iodim) * rnk); else x->dims = 0; x->rnk = rnk; return x; } void tensor_destroy(bench_tensor *sz) { bench_free0(sz->dims); bench_free(sz); } int tensor_sz(const bench_tensor *sz) { int i, n = 1; if (!FINITE_RNK(sz->rnk)) return 0; for (i = 0; i < sz->rnk; ++i) n *= sz->dims[i].n; return n; } /* total order among bench_iodim's */ static int dimcmp(const bench_iodim *a, const bench_iodim *b) { if (b->is != a->is) return (b->is - a->is); /* shorter strides go later */ if (b->os != a->os) return (b->os - a->os); /* shorter strides go later */ return (int)(a->n - b->n); /* larger n's go later */ } bench_tensor *tensor_compress(const bench_tensor *sz) { int i, rnk; bench_tensor *x; BENCH_ASSERT(FINITE_RNK(sz->rnk)); for (i = rnk = 0; i < sz->rnk; ++i) { BENCH_ASSERT(sz->dims[i].n > 0); if (sz->dims[i].n != 1) ++rnk; } x = mktensor(rnk); for (i = rnk = 0; i < sz->rnk; ++i) { if (sz->dims[i].n != 1) x->dims[rnk++] = sz->dims[i]; } if (rnk) { /* God knows how qsort() behaves if n==0 */ qsort(x->dims, (size_t)x->rnk, sizeof(bench_iodim), (int (*)(const void *, const void *))dimcmp); } return x; } int tensor_unitstridep(bench_tensor *t) { BENCH_ASSERT(FINITE_RNK(t->rnk)); return (t->rnk == 0 || (t->dims[t->rnk - 1].is == 1 && t->dims[t->rnk - 1].os == 1)); } /* detect screwy real padded rowmajor... ugh */ int tensor_real_rowmajorp(bench_tensor *t, int sign, int in_place) { int i; BENCH_ASSERT(FINITE_RNK(t->rnk)); i = t->rnk - 1; if (--i >= 0) { bench_iodim *d = t->dims + i; if (sign < 0) { if (d[0].is != d[1].is * (in_place ? 2*(d[1].n/2 + 1) : d[1].n)) return 0; if (d[0].os != d[1].os * (d[1].n/2 + 1)) return 0; } else { if (d[0].is != d[1].is * (d[1].n/2 + 1)) return 0; if (d[0].os != d[1].os * (in_place ? 2*(d[1].n/2 + 1) : d[1].n)) return 0; } } while (--i >= 0) { bench_iodim *d = t->dims + i; if (d[0].is != d[1].is * d[1].n) return 0; if (d[0].os != d[1].os * d[1].n) return 0; } return 1; } int tensor_rowmajorp(bench_tensor *t) { int i; BENCH_ASSERT(FINITE_RNK(t->rnk)); i = t->rnk - 1; while (--i >= 0) { bench_iodim *d = t->dims + i; if (d[0].is != d[1].is * d[1].n) return 0; if (d[0].os != d[1].os * d[1].n) return 0; } return 1; } static void dimcpy(bench_iodim *dst, const bench_iodim *src, int rnk) { int i; if (FINITE_RNK(rnk)) for (i = 0; i < rnk; ++i) dst[i] = src[i]; } bench_tensor *tensor_append(const bench_tensor *a, const bench_tensor *b) { if (!FINITE_RNK(a->rnk) || !FINITE_RNK(b->rnk)) { return mktensor(RNK_MINFTY); } else { bench_tensor *x = mktensor(a->rnk + b->rnk); dimcpy(x->dims, a->dims, a->rnk); dimcpy(x->dims + a->rnk, b->dims, b->rnk); return x; } } static int imax(int a, int b) { return (a > b) ? a : b; } static int imin(int a, int b) { return (a < b) ? a : b; } #define DEFBOUNDS(name, xs) \ void name(bench_tensor *t, int *lbp, int *ubp) \ { \ int lb = 0; \ int ub = 1; \ int i; \ \ BENCH_ASSERT(FINITE_RNK(t->rnk)); \ \ for (i = 0; i < t->rnk; ++i) { \ bench_iodim *d = t->dims + i; \ int n = d->n; \ int s = d->xs; \ lb = imin(lb, lb + s * (n - 1)); \ ub = imax(ub, ub + s * (n - 1)); \ } \ \ *lbp = lb; \ *ubp = ub; \ } DEFBOUNDS(tensor_ibounds, is) DEFBOUNDS(tensor_obounds, os) bench_tensor *tensor_copy(const bench_tensor *sz) { bench_tensor *x = mktensor(sz->rnk); dimcpy(x->dims, sz->dims, sz->rnk); return x; } /* Like tensor_copy, but copy only rnk dimensions starting with start_dim. */ bench_tensor *tensor_copy_sub(const bench_tensor *sz, int start_dim, int rnk) { bench_tensor *x; BENCH_ASSERT(FINITE_RNK(sz->rnk) && start_dim + rnk <= sz->rnk); x = mktensor(rnk); dimcpy(x->dims, sz->dims + start_dim, rnk); return x; } bench_tensor *tensor_copy_swapio(const bench_tensor *sz) { bench_tensor *x = tensor_copy(sz); int i; if (FINITE_RNK(x->rnk)) for (i = 0; i < x->rnk; ++i) { int s; s = x->dims[i].is; x->dims[i].is = x->dims[i].os; x->dims[i].os = s; } return x; } fftw-3.3.3/libbench2/timer.c0000644000175000017500000000537712054410050012526 00000000000000/* * Copyright (c) 2001 Matteo Frigo * Copyright (c) 2001 Massachusetts Institute of Technology * * This program is free software; you can redistribute it and/or modify * it under the terms of the GNU General Public License as published by * the Free Software Foundation; either version 2 of the License, or * (at your option) any later version. * * This program is distributed in the hope that it will be useful, * but WITHOUT ANY WARRANTY; without even the implied warranty of * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the * GNU General Public License for more details. * * You should have received a copy of the GNU General Public License * along with this program; if not, write to the Free Software * Foundation, Inc., 51 Franklin Street, Fifth Floor, Boston, MA 02110-1301 USA * */ #include "bench.h" #include /* * System-dependent timing functions: */ #ifdef HAVE_SYS_TIME_H #include #endif #ifdef HAVE_UNISTD_H #include #endif #ifdef HAVE_BSDGETTIMEOFDAY #ifndef HAVE_GETTIMEOFDAY #define gettimeofday BSDgettimeofday #define HAVE_GETTIMEOFDAY 1 #endif #endif double time_min; int time_repeat; #if !defined(HAVE_TIMER) && (defined(__WIN32__) || defined(_WIN32) || defined(_WINDOWS) || defined(__CYGWIN__)) #include typedef LARGE_INTEGER mytime; static mytime get_time(void) { mytime tv; QueryPerformanceCounter(&tv); return tv; } static double elapsed(mytime t1, mytime t0) { LARGE_INTEGER freq; QueryPerformanceFrequency(&freq); return (((double) t1.QuadPart - (double) t0.QuadPart)) / ((double) freq.QuadPart); } #define HAVE_TIMER #endif #if defined(HAVE_GETTIMEOFDAY) && !defined(HAVE_TIMER) typedef struct timeval mytime; static mytime get_time(void) { struct timeval tv; gettimeofday(&tv, 0); return tv; } static double elapsed(mytime t1, mytime t0) { return ((double) t1.tv_sec - (double) t0.tv_sec) + ((double) t1.tv_usec - (double) t0.tv_usec) * 1.0E-6; } #define HAVE_TIMER #endif #ifndef HAVE_TIMER #error "timer not defined" #endif static double calibrate(void) { /* there seems to be no reasonable way to calibrate the clock automatically any longer. Grrr... */ return 0.01; } void timer_init(double tmin, int repeat) { static int inited = 0; if (inited) return; inited = 1; if (!repeat) repeat = 8; time_repeat = repeat; if (tmin > 0) time_min = tmin; else time_min = calibrate(); } static mytime t0[BENCH_NTIMERS]; void timer_start(int n) { BENCH_ASSERT(n >= 0 && n < BENCH_NTIMERS); t0[n] = get_time(); } double timer_stop(int n) { mytime t1; BENCH_ASSERT(n >= 0 && n < BENCH_NTIMERS); t1 = get_time(); return elapsed(t1, t0[n]); } fftw-3.3.3/libbench2/useropt.c0000644000175000017500000000173012054410050013074 00000000000000/* * Copyright (c) 2000 Matteo Frigo * Copyright (c) 2000 Massachusetts Institute of Technology * * This program is free software; you can redistribute it and/or modify * it under the terms of the GNU General Public License as published by * the Free Software Foundation; either version 2 of the License, or * (at your option) any later version. * * This program is distributed in the hope that it will be useful, * but WITHOUT ANY WARRANTY; without even the implied warranty of * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the * GNU General Public License for more details. * * You should have received a copy of the GNU General Public License * along with this program; if not, write to the Free Software * Foundation, Inc., 51 Franklin Street, Fifth Floor, Boston, MA 02110-1301 USA * */ #include #include #include "bench.h" void useropt(const char *arg) { ovtpvt_err("unknown user option: %s. Ignoring.\n", arg); } fftw-3.3.3/libbench2/util.c0000644000175000017500000001355012054410050012353 00000000000000/* * Copyright (c) 2000 Matteo Frigo * Copyright (c) 2000 Massachusetts Institute of Technology * * This program is free software; you can redistribute it and/or modify * it under the terms of the GNU General Public License as published by * the Free Software Foundation; either version 2 of the License, or * (at your option) any later version. * * This program is distributed in the hope that it will be useful, * but WITHOUT ANY WARRANTY; without even the implied warranty of * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the * GNU General Public License for more details. * * You should have received a copy of the GNU General Public License * along with this program; if not, write to the Free Software * Foundation, Inc., 51 Franklin Street, Fifth Floor, Boston, MA 02110-1301 USA * */ #include "bench.h" #include #include #include #include #if defined(HAVE_DECL_MEMALIGN) && !HAVE_DECL_MEMALIGN # if defined(HAVE_MALLOC_H) # include # else extern void *memalign(size_t, size_t); # endif #endif #if defined(HAVE_DECL_POSIX_MEMALIGN) && !HAVE_DECL_POSIX_MEMALIGN extern int posix_memalign(void **, size_t, size_t); #endif void bench_assertion_failed(const char *s, int line, const char *file) { ovtpvt_err("bench: %s:%d: assertion failed: %s\n", file, line, s); bench_exit(EXIT_FAILURE); } #ifdef HAVE_DRAND48 # if defined(HAVE_DECL_DRAND48) && !HAVE_DECL_DRAND48 extern double drand48(void); # endif double bench_drand(void) { return drand48() - 0.5; } # if defined(HAVE_DECL_SRAND48) && !HAVE_DECL_SRAND48 extern void srand48(long); # endif void bench_srand(int seed) { srand48(seed); } #else double bench_drand(void) { double d = rand(); return (d / (double) RAND_MAX) - 0.5; } void bench_srand(int seed) { srand(seed); } #endif /********************************************************** * DEBUGGING CODE **********************************************************/ #ifdef BENCH_DEBUG static int bench_malloc_cnt = 0; /* * debugging malloc/free. Initialize every malloced and freed area to * random values, just to make sure we are not using uninitialized * pointers. Also check for writes past the ends of allocated blocks, * and a couple of other things. * * This code is a quick and dirty hack -- use at your own risk. */ static int bench_malloc_total = 0, bench_malloc_max = 0, bench_malloc_cnt_max = 0; #define MAGIC ((size_t)0xABadCafe) #define PAD_FACTOR 2 #define TWO_SIZE_T (2 * sizeof(size_t)) #define VERBOSE_ALLOCATION 0 #if VERBOSE_ALLOCATION #define WHEN_VERBOSE(a) a #else #define WHEN_VERBOSE(a) #endif void *bench_malloc(size_t n) { char *p; size_t i; bench_malloc_total += n; if (bench_malloc_total > bench_malloc_max) bench_malloc_max = bench_malloc_total; p = (char *) malloc(PAD_FACTOR * n + TWO_SIZE_T); BENCH_ASSERT(p); /* store the size in a known position */ ((size_t *) p)[0] = n; ((size_t *) p)[1] = MAGIC; for (i = 0; i < PAD_FACTOR * n; i++) p[i + TWO_SIZE_T] = (char) (i ^ 0xDEADBEEF); ++bench_malloc_cnt; if (bench_malloc_cnt > bench_malloc_cnt_max) bench_malloc_cnt_max = bench_malloc_cnt; /* skip the size we stored previously */ return (void *) (p + TWO_SIZE_T); } void bench_free(void *p) { char *q; BENCH_ASSERT(p); q = ((char *) p) - TWO_SIZE_T; BENCH_ASSERT(q); { size_t n = ((size_t *) q)[0]; size_t magic = ((size_t *) q)[1]; size_t i; ((size_t *) q)[0] = 0; /* set to zero to detect duplicate free's */ BENCH_ASSERT(magic == MAGIC); ((size_t *) q)[1] = ~MAGIC; bench_malloc_total -= n; BENCH_ASSERT(bench_malloc_total >= 0); /* check for writing past end of array: */ for (i = n; i < PAD_FACTOR * n; ++i) if (q[i + TWO_SIZE_T] != (char) (i ^ 0xDEADBEEF)) { BENCH_ASSERT(0 /* array bounds overwritten */); } for (i = 0; i < PAD_FACTOR * n; ++i) q[i + TWO_SIZE_T] = (char) (i ^ 0xBEEFDEAD); --bench_malloc_cnt; BENCH_ASSERT(bench_malloc_cnt >= 0); BENCH_ASSERT( (bench_malloc_cnt == 0 && bench_malloc_total == 0) || (bench_malloc_cnt > 0 && bench_malloc_total > 0)); free(q); } } #else /********************************************************** * NON DEBUGGING CODE **********************************************************/ /* production version, no hacks */ #define MIN_ALIGNMENT 128 /* must be power of two */ #define real_free free /* memalign and malloc use ordinary free */ void *bench_malloc(size_t n) { void *p; if (n == 0) n = 1; #if defined(WITH_OUR_MALLOC) /* Our own aligned malloc/free. Assumes sizeof(void*) is a power of two <= 8 and that malloc is at least sizeof(void*)-aligned. Assumes size_t = uintptr_t. */ { void *p0; if ((p0 = malloc(n + MIN_ALIGNMENT))) { p = (void *) (((size_t) p0 + MIN_ALIGNMENT) & (~((size_t) (MIN_ALIGNMENT - 1)))); *((void **) p - 1) = p0; } else p = (void *) 0; } #elif defined(HAVE_MEMALIGN) p = memalign(MIN_ALIGNMENT, n); #elif defined(HAVE_POSIX_MEMALIGN) /* note: posix_memalign is broken in glibc 2.2.5: it constrains the size, not the alignment, to be (power of two) * sizeof(void*). The bug seems to have been fixed as of glibc 2.3.1. */ if (posix_memalign(&p, MIN_ALIGNMENT, n)) p = (void*) 0; #elif defined(__ICC) || defined(__INTEL_COMPILER) || defined(HAVE__MM_MALLOC) /* Intel's C compiler defines _mm_malloc and _mm_free intrinsics */ p = (void *) _mm_malloc(n, MIN_ALIGNMENT); # undef real_free # define real_free _mm_free #else p = malloc(n); #endif BENCH_ASSERT(p); return p; } void bench_free(void *p) { #ifdef WITH_OUR_MALLOC if (p) free(*((void **) p - 1)); #else real_free(p); #endif } #endif void bench_free0(void *p) { if (p) bench_free(p); } fftw-3.3.3/libbench2/verify-dft.c0000644000175000017500000001227312054410050013456 00000000000000/* * Copyright (c) 2003, 2007-11 Matteo Frigo * Copyright (c) 2003, 2007-11 Massachusetts Institute of Technology * * This program is free software; you can redistribute it and/or modify * it under the terms of the GNU General Public License as published by * the Free Software Foundation; either version 2 of the License, or * (at your option) any later version. * * This program is distributed in the hope that it will be useful, * but WITHOUT ANY WARRANTY; without even the implied warranty of * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the * GNU General Public License for more details. * * You should have received a copy of the GNU General Public License * along with this program; if not, write to the Free Software * Foundation, Inc., 51 Franklin Street, Fifth Floor, Boston, MA 02110-1301 USA * */ #include "verify.h" /* copy A into B, using output stride of A and input stride of B */ typedef struct { dotens2_closure k; R *ra; R *ia; R *rb; R *ib; int scalea, scaleb; } cpy_closure; static void cpy0(dotens2_closure *k_, int indxa, int ondxa, int indxb, int ondxb) { cpy_closure *k = (cpy_closure *)k_; k->rb[indxb * k->scaleb] = k->ra[ondxa * k->scalea]; k->ib[indxb * k->scaleb] = k->ia[ondxa * k->scalea]; UNUSED(indxa); UNUSED(ondxb); } static void cpy(R *ra, R *ia, const bench_tensor *sza, int scalea, R *rb, R *ib, const bench_tensor *szb, int scaleb) { cpy_closure k; k.k.apply = cpy0; k.ra = ra; k.ia = ia; k.rb = rb; k.ib = ib; k.scalea = scalea; k.scaleb = scaleb; bench_dotens2(sza, szb, &k.k); } typedef struct { dofft_closure k; bench_problem *p; } dofft_dft_closure; static void dft_apply(dofft_closure *k_, bench_complex *in, bench_complex *out) { dofft_dft_closure *k = (dofft_dft_closure *)k_; bench_problem *p = k->p; bench_tensor *totalsz, *pckdsz; bench_tensor *totalsz_swap, *pckdsz_swap; bench_real *ri, *ii, *ro, *io; int totalscale; totalsz = tensor_append(p->vecsz, p->sz); pckdsz = verify_pack(totalsz, 2); ri = (bench_real *) p->in; ro = (bench_real *) p->out; totalsz_swap = tensor_copy_swapio(totalsz); pckdsz_swap = tensor_copy_swapio(pckdsz); /* confusion: the stride is the distance between complex elements when using interleaved format, but it is the distance between real elements when using split format */ if (p->split) { ii = p->ini ? (bench_real *) p->ini : ri + p->iphyssz; io = p->outi ? (bench_real *) p->outi : ro + p->ophyssz; totalscale = 1; } else { ii = p->ini ? (bench_real *) p->ini : ri + 1; io = p->outi ? (bench_real *) p->outi : ro + 1; totalscale = 2; } cpy(&c_re(in[0]), &c_im(in[0]), pckdsz, 1, ri, ii, totalsz, totalscale); after_problem_ccopy_from(p, ri, ii); doit(1, p); after_problem_ccopy_to(p, ro, io); if (k->k.recopy_input) cpy(ri, ii, totalsz_swap, totalscale, &c_re(in[0]), &c_im(in[0]), pckdsz_swap, 1); cpy(ro, io, totalsz, totalscale, &c_re(out[0]), &c_im(out[0]), pckdsz, 1); tensor_destroy(totalsz); tensor_destroy(pckdsz); tensor_destroy(totalsz_swap); tensor_destroy(pckdsz_swap); } void verify_dft(bench_problem *p, int rounds, double tol, errors *e) { C *inA, *inB, *inC, *outA, *outB, *outC, *tmp; int n, vecn, N; dofft_dft_closure k; BENCH_ASSERT(p->kind == PROBLEM_COMPLEX); k.k.apply = dft_apply; k.k.recopy_input = 0; k.p = p; if (rounds == 0) rounds = 20; /* default value */ n = tensor_sz(p->sz); vecn = tensor_sz(p->vecsz); N = n * vecn; inA = (C *) bench_malloc(N * sizeof(C)); inB = (C *) bench_malloc(N * sizeof(C)); inC = (C *) bench_malloc(N * sizeof(C)); outA = (C *) bench_malloc(N * sizeof(C)); outB = (C *) bench_malloc(N * sizeof(C)); outC = (C *) bench_malloc(N * sizeof(C)); tmp = (C *) bench_malloc(N * sizeof(C)); e->i = impulse(&k.k, n, vecn, inA, inB, inC, outA, outB, outC, tmp, rounds, tol); e->l = linear(&k.k, 0, N, inA, inB, inC, outA, outB, outC, tmp, rounds, tol); e->s = 0.0; e->s = dmax(e->s, tf_shift(&k.k, 0, p->sz, n, vecn, p->sign, inA, inB, outA, outB, tmp, rounds, tol, TIME_SHIFT)); e->s = dmax(e->s, tf_shift(&k.k, 0, p->sz, n, vecn, p->sign, inA, inB, outA, outB, tmp, rounds, tol, FREQ_SHIFT)); if (!p->in_place && !p->destroy_input) preserves_input(&k.k, 0, N, inA, inB, outB, rounds); bench_free(tmp); bench_free(outC); bench_free(outB); bench_free(outA); bench_free(inC); bench_free(inB); bench_free(inA); } void accuracy_dft(bench_problem *p, int rounds, int impulse_rounds, double t[6]) { dofft_dft_closure k; int n; C *a, *b; BENCH_ASSERT(p->kind == PROBLEM_COMPLEX); BENCH_ASSERT(p->sz->rnk == 1); BENCH_ASSERT(p->vecsz->rnk == 0); k.k.apply = dft_apply; k.k.recopy_input = 0; k.p = p; n = tensor_sz(p->sz); a = (C *) bench_malloc(n * sizeof(C)); b = (C *) bench_malloc(n * sizeof(C)); accuracy_test(&k.k, 0, p->sign, n, a, b, rounds, impulse_rounds, t); bench_free(b); bench_free(a); } fftw-3.3.3/libbench2/verify-lib.c0000644000175000017500000003134612054410050013451 00000000000000/* * Copyright (c) 2003, 2007-11 Matteo Frigo * Copyright (c) 2003, 2007-11 Massachusetts Institute of Technology * * This program is free software; you can redistribute it and/or modify * it under the terms of the GNU General Public License as published by * the Free Software Foundation; either version 2 of the License, or * (at your option) any later version. * * This program is distributed in the hope that it will be useful, * but WITHOUT ANY WARRANTY; without even the implied warranty of * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the * GNU General Public License for more details. * * You should have received a copy of the GNU General Public License * along with this program; if not, write to the Free Software * Foundation, Inc., 51 Franklin Street, Fifth Floor, Boston, MA 02110-1301 USA * */ #include "verify.h" #include #include #include /* * Utility functions: */ static double dabs(double x) { return (x < 0.0) ? -x : x; } static double dmin(double x, double y) { return (x < y) ? x : y; } static double norm2(double x, double y) { return dmax(dabs(x), dabs(y)); } double dmax(double x, double y) { return (x > y) ? x : y; } static double aerror(C *a, C *b, int n) { if (n > 0) { /* compute the relative Linf error */ double e = 0.0, mag = 0.0; int i; for (i = 0; i < n; ++i) { e = dmax(e, norm2(c_re(a[i]) - c_re(b[i]), c_im(a[i]) - c_im(b[i]))); mag = dmax(mag, dmin(norm2(c_re(a[i]), c_im(a[i])), norm2(c_re(b[i]), c_im(b[i])))); } e /= mag; #ifdef HAVE_ISNAN BENCH_ASSERT(!isnan(e)); #endif return e; } else return 0.0; } #ifdef HAVE_DRAND48 # if defined(HAVE_DECL_DRAND48) && !HAVE_DECL_DRAND48 extern double drand48(void); # endif double mydrand(void) { return drand48() - 0.5; } #else double mydrand(void) { double d = rand(); return (d / (double) RAND_MAX) - 0.5; } #endif void arand(C *a, int n) { int i; /* generate random inputs */ for (i = 0; i < n; ++i) { c_re(a[i]) = mydrand(); c_im(a[i]) = mydrand(); } } /* make array real */ void mkreal(C *A, int n) { int i; for (i = 0; i < n; ++i) { c_im(A[i]) = 0.0; } } static void assign_conj(C *Ac, C *A, int rank, const bench_iodim *dim, int stride) { if (rank == 0) { c_re(*Ac) = c_re(*A); c_im(*Ac) = -c_im(*A); } else { int i, n0 = dim[rank - 1].n, s = stride; rank -= 1; stride *= n0; assign_conj(Ac, A, rank, dim, stride); for (i = 1; i < n0; ++i) assign_conj(Ac + (n0 - i) * s, A + i * s, rank, dim, stride); } } /* make array hermitian */ void mkhermitian(C *A, int rank, const bench_iodim *dim, int stride) { if (rank == 0) c_im(*A) = 0.0; else { int i, n0 = dim[rank - 1].n, s = stride; rank -= 1; stride *= n0; mkhermitian(A, rank, dim, stride); for (i = 1; 2*i < n0; ++i) assign_conj(A + (n0 - i) * s, A + i * s, rank, dim, stride); if (2*i == n0) mkhermitian(A + i * s, rank, dim, stride); } } void mkhermitian1(C *a, int n) { bench_iodim d; d.n = n; d.is = d.os = 1; mkhermitian(a, 1, &d, 1); } /* C = A */ void acopy(C *c, C *a, int n) { int i; for (i = 0; i < n; ++i) { c_re(c[i]) = c_re(a[i]); c_im(c[i]) = c_im(a[i]); } } /* C = A + B */ void aadd(C *c, C *a, C *b, int n) { int i; for (i = 0; i < n; ++i) { c_re(c[i]) = c_re(a[i]) + c_re(b[i]); c_im(c[i]) = c_im(a[i]) + c_im(b[i]); } } /* C = A - B */ void asub(C *c, C *a, C *b, int n) { int i; for (i = 0; i < n; ++i) { c_re(c[i]) = c_re(a[i]) - c_re(b[i]); c_im(c[i]) = c_im(a[i]) - c_im(b[i]); } } /* B = rotate left A (complex) */ void arol(C *b, C *a, int n, int nb, int na) { int i, ib, ia; for (ib = 0; ib < nb; ++ib) { for (i = 0; i < n - 1; ++i) for (ia = 0; ia < na; ++ia) { C *pb = b + (ib * n + i) * na + ia; C *pa = a + (ib * n + i + 1) * na + ia; c_re(*pb) = c_re(*pa); c_im(*pb) = c_im(*pa); } for (ia = 0; ia < na; ++ia) { C *pb = b + (ib * n + n - 1) * na + ia; C *pa = a + ib * n * na + ia; c_re(*pb) = c_re(*pa); c_im(*pb) = c_im(*pa); } } } void aphase_shift(C *b, C *a, int n, int nb, int na, double sign) { int j, jb, ja; trigreal twopin; twopin = K2PI / n; for (jb = 0; jb < nb; ++jb) for (j = 0; j < n; ++j) { trigreal s = sign * SIN(j * twopin); trigreal c = COS(j * twopin); for (ja = 0; ja < na; ++ja) { int k = (jb * n + j) * na + ja; c_re(b[k]) = c_re(a[k]) * c - c_im(a[k]) * s; c_im(b[k]) = c_re(a[k]) * s + c_im(a[k]) * c; } } } /* A = alpha * A (complex, in place) */ void ascale(C *a, C alpha, int n) { int i; for (i = 0; i < n; ++i) { R xr = c_re(a[i]), xi = c_im(a[i]); c_re(a[i]) = xr * c_re(alpha) - xi * c_im(alpha); c_im(a[i]) = xr * c_im(alpha) + xi * c_re(alpha); } } double acmp(C *a, C *b, int n, const char *test, double tol) { double d = aerror(a, b, n); if (d > tol) { ovtpvt_err("Found relative error %e (%s)\n", d, test); { int i, N; N = n > 300 && verbose <= 2 ? 300 : n; for (i = 0; i < N; ++i) ovtpvt_err("%8d %16.12f %16.12f %16.12f %16.12f\n", i, (double) c_re(a[i]), (double) c_im(a[i]), (double) c_re(b[i]), (double) c_im(b[i])); } bench_exit(EXIT_FAILURE); } return d; } /* * Implementation of the FFT tester described in * * Funda Ergün. Testing multivariate linear functions: Overcoming the * generator bottleneck. In Proceedings of the Twenty-Seventh Annual * ACM Symposium on the Theory of Computing, pages 407-416, Las Vegas, * Nevada, 29 May--1 June 1995. * * Also: F. Ergun, S. R. Kumar, and D. Sivakumar, "Self-testing without * the generator bottleneck," SIAM J. on Computing 29 (5), 1630-51 (2000). */ static double impulse0(dofft_closure *k, int n, int vecn, C *inA, C *inB, C *inC, C *outA, C *outB, C *outC, C *tmp, int rounds, double tol) { int N = n * vecn; double e = 0.0; int j; k->apply(k, inA, tmp); e = dmax(e, acmp(tmp, outA, N, "impulse 1", tol)); for (j = 0; j < rounds; ++j) { arand(inB, N); asub(inC, inA, inB, N); k->apply(k, inB, outB); k->apply(k, inC, outC); aadd(tmp, outB, outC, N); e = dmax(e, acmp(tmp, outA, N, "impulse", tol)); } return e; } double impulse(dofft_closure *k, int n, int vecn, C *inA, C *inB, C *inC, C *outA, C *outB, C *outC, C *tmp, int rounds, double tol) { int i, j; double e = 0.0; /* check impulsive input */ for (i = 0; i < vecn; ++i) { R x = (sqrt(n)*(i+1)) / (double)(vecn+1); for (j = 0; j < n; ++j) { c_re(inA[j + i * n]) = 0; c_im(inA[j + i * n]) = 0; c_re(outA[j + i * n]) = x; c_im(outA[j + i * n]) = 0; } c_re(inA[i * n]) = x; c_im(inA[i * n]) = 0; } e = dmax(e, impulse0(k, n, vecn, inA, inB, inC, outA, outB, outC, tmp, rounds, tol)); /* check constant input */ for (i = 0; i < vecn; ++i) { R x = (i+1) / ((double)(vecn+1) * sqrt(n)); for (j = 0; j < n; ++j) { c_re(inA[j + i * n]) = x; c_im(inA[j + i * n]) = 0; c_re(outA[j + i * n]) = 0; c_im(outA[j + i * n]) = 0; } c_re(outA[i * n]) = n * x; c_im(outA[i * n]) = 0; } e = dmax(e, impulse0(k, n, vecn, inA, inB, inC, outA, outB, outC, tmp, rounds, tol)); return e; } double linear(dofft_closure *k, int realp, int n, C *inA, C *inB, C *inC, C *outA, C *outB, C *outC, C *tmp, int rounds, double tol) { int j; double e = 0.0; for (j = 0; j < rounds; ++j) { C alpha, beta; c_re(alpha) = mydrand(); c_im(alpha) = realp ? 0.0 : mydrand(); c_re(beta) = mydrand(); c_im(beta) = realp ? 0.0 : mydrand(); arand(inA, n); arand(inB, n); k->apply(k, inA, outA); k->apply(k, inB, outB); ascale(outA, alpha, n); ascale(outB, beta, n); aadd(tmp, outA, outB, n); ascale(inA, alpha, n); ascale(inB, beta, n); aadd(inC, inA, inB, n); k->apply(k, inC, outC); e = dmax(e, acmp(outC, tmp, n, "linear", tol)); } return e; } double tf_shift(dofft_closure *k, int realp, const bench_tensor *sz, int n, int vecn, double sign, C *inA, C *inB, C *outA, C *outB, C *tmp, int rounds, double tol, int which_shift) { int nb, na, dim, N = n * vecn; int i, j; double e = 0.0; /* test 3: check the time-shift property */ /* the paper performs more tests, but this code should be fine too */ nb = 1; na = n; /* check shifts across all SZ dimensions */ for (dim = 0; dim < sz->rnk; ++dim) { int ncur = sz->dims[dim].n; na /= ncur; for (j = 0; j < rounds; ++j) { arand(inA, N); if (which_shift == TIME_SHIFT) { for (i = 0; i < vecn; ++i) { if (realp) mkreal(inA + i * n, n); arol(inB + i * n, inA + i * n, ncur, nb, na); } k->apply(k, inA, outA); k->apply(k, inB, outB); for (i = 0; i < vecn; ++i) aphase_shift(tmp + i * n, outB + i * n, ncur, nb, na, sign); e = dmax(e, acmp(tmp, outA, N, "time shift", tol)); } else { for (i = 0; i < vecn; ++i) { if (realp) mkhermitian(inA + i * n, sz->rnk, sz->dims, 1); aphase_shift(inB + i * n, inA + i * n, ncur, nb, na, -sign); } k->apply(k, inA, outA); k->apply(k, inB, outB); for (i = 0; i < vecn; ++i) arol(tmp + i * n, outB + i * n, ncur, nb, na); e = dmax(e, acmp(tmp, outA, N, "freq shift", tol)); } } nb *= ncur; } return e; } void preserves_input(dofft_closure *k, aconstrain constrain, int n, C *inA, C *inB, C *outB, int rounds) { int j; int recopy_input = k->recopy_input; k->recopy_input = 1; for (j = 0; j < rounds; ++j) { arand(inA, n); if (constrain) constrain(inA, n); acopy(inB, inA, n); k->apply(k, inB, outB); acmp(inB, inA, n, "preserves_input", 0.0); } k->recopy_input = recopy_input; } /* Make a copy of the size tensor, with the same dimensions, but with the strides corresponding to a "packed" row-major array with the given stride. */ bench_tensor *verify_pack(const bench_tensor *sz, int s) { bench_tensor *x = tensor_copy(sz); if (FINITE_RNK(x->rnk) && x->rnk > 0) { int i; x->dims[x->rnk - 1].is = s; x->dims[x->rnk - 1].os = s; for (i = x->rnk - 1; i > 0; --i) { x->dims[i - 1].is = x->dims[i].is * x->dims[i].n; x->dims[i - 1].os = x->dims[i].os * x->dims[i].n; } } return x; } static int all_zero(C *a, int n) { int i; for (i = 0; i < n; ++i) if (c_re(a[i]) != 0.0 || c_im(a[i]) != 0.0) return 0; return 1; } static int one_accuracy_test(dofft_closure *k, aconstrain constrain, int sign, int n, C *a, C *b, double t[6]) { double err[6]; if (constrain) constrain(a, n); if (all_zero(a, n)) return 0; k->apply(k, a, b); fftaccuracy(n, a, b, sign, err); t[0] += err[0]; t[1] += err[1] * err[1]; t[2] = dmax(t[2], err[2]); t[3] += err[3]; t[4] += err[4] * err[4]; t[5] = dmax(t[5], err[5]); return 1; } void accuracy_test(dofft_closure *k, aconstrain constrain, int sign, int n, C *a, C *b, int rounds, int impulse_rounds, double t[6]) { int r, i; int ntests = 0; bench_complex czero = {0, 0}; for (i = 0; i < 6; ++i) t[i] = 0.0; for (r = 0; r < rounds; ++r) { arand(a, n); if (one_accuracy_test(k, constrain, sign, n, a, b, t)) ++ntests; } /* impulses at beginning of array */ for (r = 0; r < impulse_rounds; ++r) { if (r > n - r - 1) continue; caset(a, n, czero); c_re(a[r]) = c_im(a[r]) = 1.0; if (one_accuracy_test(k, constrain, sign, n, a, b, t)) ++ntests; } /* impulses at end of array */ for (r = 0; r < impulse_rounds; ++r) { if (r <= n - r - 1) continue; caset(a, n, czero); c_re(a[n - r - 1]) = c_im(a[n - r - 1]) = 1.0; if (one_accuracy_test(k, constrain, sign, n, a, b, t)) ++ntests; } /* randomly-located impulses */ for (r = 0; r < impulse_rounds; ++r) { caset(a, n, czero); i = rand() % n; c_re(a[i]) = c_im(a[i]) = 1.0; if (one_accuracy_test(k, constrain, sign, n, a, b, t)) ++ntests; } t[0] /= ntests; t[1] = sqrt(t[1] / ntests); t[3] /= ntests; t[4] = sqrt(t[4] / ntests); fftaccuracy_done(); } fftw-3.3.3/libbench2/verify-r2r.c0000644000175000017500000005565312054410050013417 00000000000000/* * Copyright (c) 2003, 2007-11 Matteo Frigo * Copyright (c) 2003, 2007-11 Massachusetts Institute of Technology * * This program is free software; you can redistribute it and/or modify * it under the terms of the GNU General Public License as published by * the Free Software Foundation; either version 2 of the License, or * (at your option) any later version. * * This program is distributed in the hope that it will be useful, * but WITHOUT ANY WARRANTY; without even the implied warranty of * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the * GNU General Public License for more details. * * You should have received a copy of the GNU General Public License * along with this program; if not, write to the Free Software * Foundation, Inc., 51 Franklin Street, Fifth Floor, Boston, MA 02110-1301 USA * */ /* Lots of ugly duplication from verify-lib.c, plus lots of ugliness in general for all of the r2r variants...oh well, for now */ #include "verify.h" #include #include #include typedef struct { bench_problem *p; bench_tensor *probsz; bench_tensor *totalsz; bench_tensor *pckdsz; bench_tensor *pckdvecsz; } info; /* * Utility functions: */ static double dabs(double x) { return (x < 0.0) ? -x : x; } static double dmin(double x, double y) { return (x < y) ? x : y; } static double raerror(R *a, R *b, int n) { if (n > 0) { /* compute the relative Linf error */ double e = 0.0, mag = 0.0; int i; for (i = 0; i < n; ++i) { e = dmax(e, dabs(a[i] - b[i])); mag = dmax(mag, dmin(dabs(a[i]), dabs(b[i]))); } if (dabs(mag) < 1e-14 && dabs(e) < 1e-14) e = 0.0; else e /= mag; #ifdef HAVE_ISNAN BENCH_ASSERT(!isnan(e)); #endif return e; } else return 0.0; } #define by2pi(m, n) ((K2PI * (m)) / (n)) /* * Improve accuracy by reducing x to range [0..1/8] * before multiplication by 2 * PI. */ static trigreal bench_sincos(trigreal m, trigreal n, int sinp) { /* waiting for C to get tail recursion... */ trigreal half_n = n * 0.5; trigreal quarter_n = half_n * 0.5; trigreal eighth_n = quarter_n * 0.5; trigreal sgn = 1.0; if (sinp) goto sin; cos: if (m < 0) { m = -m; /* goto cos; */ } if (m > half_n) { m = n - m; goto cos; } if (m > eighth_n) { m = quarter_n - m; goto sin; } return sgn * COS(by2pi(m, n)); msin: sgn = -sgn; sin: if (m < 0) { m = -m; goto msin; } if (m > half_n) { m = n - m; goto msin; } if (m > eighth_n) { m = quarter_n - m; goto cos; } return sgn * SIN(by2pi(m, n)); } static trigreal cos2pi(int m, int n) { return bench_sincos((trigreal)m, (trigreal)n, 0); } static trigreal sin2pi(int m, int n) { return bench_sincos((trigreal)m, (trigreal)n, 1); } static trigreal cos00(int i, int j, int n) { return cos2pi(i * j, n); } static trigreal cos01(int i, int j, int n) { return cos00(i, 2*j + 1, 2*n); } static trigreal cos10(int i, int j, int n) { return cos00(2*i + 1, j, 2*n); } static trigreal cos11(int i, int j, int n) { return cos00(2*i + 1, 2*j + 1, 4*n); } static trigreal sin00(int i, int j, int n) { return sin2pi(i * j, n); } static trigreal sin01(int i, int j, int n) { return sin00(i, 2*j + 1, 2*n); } static trigreal sin10(int i, int j, int n) { return sin00(2*i + 1, j, 2*n); } static trigreal sin11(int i, int j, int n) { return sin00(2*i + 1, 2*j + 1, 4*n); } static trigreal realhalf(int i, int j, int n) { UNUSED(i); if (j <= n - j) return 1.0; else return 0.0; } static trigreal coshalf(int i, int j, int n) { if (j <= n - j) return cos00(i, j, n); else return cos00(i, n - j, n); } static trigreal unity(int i, int j, int n) { UNUSED(i); UNUSED(j); UNUSED(n); return 1.0; } typedef trigreal (*trigfun)(int, int, int); static void rarand(R *a, int n) { int i; /* generate random inputs */ for (i = 0; i < n; ++i) { a[i] = mydrand(); } } /* C = A + B */ static void raadd(R *c, R *a, R *b, int n) { int i; for (i = 0; i < n; ++i) { c[i] = a[i] + b[i]; } } /* C = A - B */ static void rasub(R *c, R *a, R *b, int n) { int i; for (i = 0; i < n; ++i) { c[i] = a[i] - b[i]; } } /* B = rotate left A + rotate right A */ static void rarolr(R *b, R *a, int n, int nb, int na, r2r_kind_t k) { int isL0 = 0, isL1 = 0, isR0 = 0, isR1 = 0; int i, ib, ia; for (ib = 0; ib < nb; ++ib) { for (i = 0; i < n - 1; ++i) for (ia = 0; ia < na; ++ia) b[(ib * n + i) * na + ia] = a[(ib * n + i + 1) * na + ia]; /* ugly switch to do boundary conditions for various r2r types */ switch (k) { /* periodic boundaries */ case R2R_DHT: case R2R_R2HC: for (ia = 0; ia < na; ++ia) { b[(ib * n + n - 1) * na + ia] = a[(ib * n + 0) * na + ia]; b[(ib * n + 0) * na + ia] += a[(ib * n + n - 1) * na + ia]; } break; case R2R_HC2R: /* ugh (hermitian halfcomplex boundaries) */ if (n > 2) { if (n % 2 == 0) for (ia = 0; ia < na; ++ia) { b[(ib * n + n - 1) * na + ia] = 0.0; b[(ib * n + 0) * na + ia] += a[(ib * n + 1) * na + ia]; b[(ib * n + n/2) * na + ia] += + a[(ib * n + n/2 - 1) * na + ia] - a[(ib * n + n/2 + 1) * na + ia]; b[(ib * n + n/2 + 1) * na + ia] += - a[(ib * n + n/2) * na + ia]; } else for (ia = 0; ia < na; ++ia) { b[(ib * n + n - 1) * na + ia] = 0.0; b[(ib * n + 0) * na + ia] += a[(ib * n + 1) * na + ia]; b[(ib * n + n/2) * na + ia] += + a[(ib * n + n/2) * na + ia] - a[(ib * n + n/2 + 1) * na + ia]; b[(ib * n + n/2 + 1) * na + ia] += - a[(ib * n + n/2 + 1) * na + ia] - a[(ib * n + n/2) * na + ia]; } } else /* n <= 2 */ { for (ia = 0; ia < na; ++ia) { b[(ib * n + n - 1) * na + ia] = a[(ib * n + 0) * na + ia]; b[(ib * n + 0) * na + ia] += a[(ib * n + n - 1) * na + ia]; } } break; /* various even/odd boundary conditions */ case R2R_REDFT00: isL1 = isR1 = 1; goto mirrors; case R2R_REDFT01: isL1 = 1; goto mirrors; case R2R_REDFT10: isL0 = isR0 = 1; goto mirrors; case R2R_REDFT11: isL0 = 1; isR0 = -1; goto mirrors; case R2R_RODFT00: goto mirrors; case R2R_RODFT01: isR1 = 1; goto mirrors; case R2R_RODFT10: isL0 = isR0 = -1; goto mirrors; case R2R_RODFT11: isL0 = -1; isR0 = 1; goto mirrors; mirrors: for (ia = 0; ia < na; ++ia) b[(ib * n + n - 1) * na + ia] = isR0 * a[(ib * n + n - 1) * na + ia] + (n > 1 ? isR1 * a[(ib * n + n - 2) * na + ia] : 0); for (ia = 0; ia < na; ++ia) b[(ib * n) * na + ia] += isL0 * a[(ib * n) * na + ia] + (n > 1 ? isL1 * a[(ib * n + 1) * na + ia] : 0); } for (i = 1; i < n; ++i) for (ia = 0; ia < na; ++ia) b[(ib * n + i) * na + ia] += a[(ib * n + i - 1) * na + ia]; } } static void raphase_shift(R *b, R *a, int n, int nb, int na, int n0, int k0, trigfun t) { int j, jb, ja; for (jb = 0; jb < nb; ++jb) for (j = 0; j < n; ++j) { trigreal c = 2.0 * t(1, j + k0, n0); for (ja = 0; ja < na; ++ja) { int k = (jb * n + j) * na + ja; b[k] = a[k] * c; } } } /* A = alpha * A (real, in place) */ static void rascale(R *a, R alpha, int n) { int i; for (i = 0; i < n; ++i) { a[i] *= alpha; } } /* * compute rdft: */ /* copy real A into real B, using output stride of A and input stride of B */ typedef struct { dotens2_closure k; R *ra; R *rb; } cpyr_closure; static void cpyr0(dotens2_closure *k_, int indxa, int ondxa, int indxb, int ondxb) { cpyr_closure *k = (cpyr_closure *)k_; k->rb[indxb] = k->ra[ondxa]; UNUSED(indxa); UNUSED(ondxb); } static void cpyr(R *ra, bench_tensor *sza, R *rb, bench_tensor *szb) { cpyr_closure k; k.k.apply = cpyr0; k.ra = ra; k.rb = rb; bench_dotens2(sza, szb, &k.k); } static void dofft(info *nfo, R *in, R *out) { cpyr(in, nfo->pckdsz, (R *) nfo->p->in, nfo->totalsz); after_problem_rcopy_from(nfo->p, (bench_real *)nfo->p->in); doit(1, nfo->p); after_problem_rcopy_to(nfo->p, (bench_real *)nfo->p->out); cpyr((R *) nfo->p->out, nfo->totalsz, out, nfo->pckdsz); } static double racmp(R *a, R *b, int n, const char *test, double tol) { double d = raerror(a, b, n); if (d > tol) { ovtpvt_err("Found relative error %e (%s)\n", d, test); { int i, N; N = n > 300 && verbose <= 2 ? 300 : n; for (i = 0; i < N; ++i) ovtpvt_err("%8d %16.12f %16.12f\n", i, (double) a[i], (double) b[i]); } bench_exit(EXIT_FAILURE); } return d; } /***********************************************************************/ typedef struct { int n; /* physical size */ int n0; /* "logical" transform size */ int i0, k0; /* shifts of input/output */ trigfun ti, ts; /* impulse/shift trig functions */ } dim_stuff; static void impulse_response(int rnk, dim_stuff *d, R impulse_amp, R *A, int N) { if (rnk == 0) A[0] = impulse_amp; else { int i; N /= d->n; for (i = 0; i < d->n; ++i) { impulse_response(rnk - 1, d + 1, impulse_amp * d->ti(d->i0, d->k0 + i, d->n0), A + i * N, N); } } } /***************************************************************************/ /* * Implementation of the FFT tester described in * * Funda Ergün. Testing multivariate linear functions: Overcoming the * generator bottleneck. In Proceedings of the Twenty-Seventh Annual * ACM Symposium on the Theory of Computing, pages 407-416, Las Vegas, * Nevada, 29 May--1 June 1995. * * Also: F. Ergun, S. R. Kumar, and D. Sivakumar, "Self-testing without * the generator bottleneck," SIAM J. on Computing 29 (5), 1630-51 (2000). */ static double rlinear(int n, info *nfo, R *inA, R *inB, R *inC, R *outA, R *outB, R *outC, R *tmp, int rounds, double tol) { double e = 0.0; int j; for (j = 0; j < rounds; ++j) { R alpha, beta; alpha = mydrand(); beta = mydrand(); rarand(inA, n); rarand(inB, n); dofft(nfo, inA, outA); dofft(nfo, inB, outB); rascale(outA, alpha, n); rascale(outB, beta, n); raadd(tmp, outA, outB, n); rascale(inA, alpha, n); rascale(inB, beta, n); raadd(inC, inA, inB, n); dofft(nfo, inC, outC); e = dmax(e, racmp(outC, tmp, n, "linear", tol)); } return e; } static double rimpulse(dim_stuff *d, R impulse_amp, int n, int vecn, info *nfo, R *inA, R *inB, R *inC, R *outA, R *outB, R *outC, R *tmp, int rounds, double tol) { double e = 0.0; int N = n * vecn; int i; int j; /* test 2: check that the unit impulse is transformed properly */ for (i = 0; i < N; ++i) { /* pls */ inA[i] = 0.0; } for (i = 0; i < vecn; ++i) { inA[i * n] = (i+1) / (double)(vecn+1); /* transform of the pls */ impulse_response(nfo->probsz->rnk, d, impulse_amp * inA[i * n], outA + i * n, n); } dofft(nfo, inA, tmp); e = dmax(e, racmp(tmp, outA, N, "impulse 1", tol)); for (j = 0; j < rounds; ++j) { rarand(inB, N); rasub(inC, inA, inB, N); dofft(nfo, inB, outB); dofft(nfo, inC, outC); raadd(tmp, outB, outC, N); e = dmax(e, racmp(tmp, outA, N, "impulse", tol)); } return e; } static double t_shift(int n, int vecn, info *nfo, R *inA, R *inB, R *outA, R *outB, R *tmp, int rounds, double tol, dim_stuff *d) { double e = 0.0; int nb, na, dim, N = n * vecn; int i, j; bench_tensor *sz = nfo->probsz; /* test 3: check the time-shift property */ /* the paper performs more tests, but this code should be fine too */ nb = 1; na = n; /* check shifts across all SZ dimensions */ for (dim = 0; dim < sz->rnk; ++dim) { int ncur = sz->dims[dim].n; na /= ncur; for (j = 0; j < rounds; ++j) { rarand(inA, N); for (i = 0; i < vecn; ++i) { rarolr(inB + i * n, inA + i*n, ncur, nb,na, nfo->p->k[dim]); } dofft(nfo, inA, outA); dofft(nfo, inB, outB); for (i = 0; i < vecn; ++i) raphase_shift(tmp + i * n, outA + i * n, ncur, nb, na, d[dim].n0, d[dim].k0, d[dim].ts); e = dmax(e, racmp(tmp, outB, N, "time shift", tol)); } nb *= ncur; } return e; } /***********************************************************************/ void verify_r2r(bench_problem *p, int rounds, double tol, errors *e) { R *inA, *inB, *inC, *outA, *outB, *outC, *tmp; info nfo; int n, vecn, N; double impulse_amp = 1.0; dim_stuff *d; int i; if (rounds == 0) rounds = 20; /* default value */ n = tensor_sz(p->sz); vecn = tensor_sz(p->vecsz); N = n * vecn; d = (dim_stuff *) bench_malloc(sizeof(dim_stuff) * p->sz->rnk); for (i = 0; i < p->sz->rnk; ++i) { int n0, i0, k0; trigfun ti, ts; d[i].n = n0 = p->sz->dims[i].n; if (p->k[i] > R2R_DHT) n0 = 2 * (n0 + (p->k[i] == R2R_REDFT00 ? -1 : (p->k[i] == R2R_RODFT00 ? 1 : 0))); switch (p->k[i]) { case R2R_R2HC: i0 = k0 = 0; ti = realhalf; ts = coshalf; break; case R2R_DHT: i0 = k0 = 0; ti = unity; ts = cos00; break; case R2R_HC2R: i0 = k0 = 0; ti = unity; ts = cos00; break; case R2R_REDFT00: i0 = k0 = 0; ti = ts = cos00; break; case R2R_REDFT01: i0 = k0 = 0; ti = ts = cos01; break; case R2R_REDFT10: i0 = k0 = 0; ti = cos10; impulse_amp *= 2.0; ts = cos00; break; case R2R_REDFT11: i0 = k0 = 0; ti = cos11; impulse_amp *= 2.0; ts = cos01; break; case R2R_RODFT00: i0 = k0 = 1; ti = sin00; impulse_amp *= 2.0; ts = cos00; break; case R2R_RODFT01: i0 = 1; k0 = 0; ti = sin01; impulse_amp *= n == 1 ? 1.0 : 2.0; ts = cos01; break; case R2R_RODFT10: i0 = 0; k0 = 1; ti = sin10; impulse_amp *= 2.0; ts = cos00; break; case R2R_RODFT11: i0 = k0 = 0; ti = sin11; impulse_amp *= 2.0; ts = cos01; break; default: BENCH_ASSERT(0); return; } d[i].n0 = n0; d[i].i0 = i0; d[i].k0 = k0; d[i].ti = ti; d[i].ts = ts; } inA = (R *) bench_malloc(N * sizeof(R)); inB = (R *) bench_malloc(N * sizeof(R)); inC = (R *) bench_malloc(N * sizeof(R)); outA = (R *) bench_malloc(N * sizeof(R)); outB = (R *) bench_malloc(N * sizeof(R)); outC = (R *) bench_malloc(N * sizeof(R)); tmp = (R *) bench_malloc(N * sizeof(R)); nfo.p = p; nfo.probsz = p->sz; nfo.totalsz = tensor_append(p->vecsz, nfo.probsz); nfo.pckdsz = verify_pack(nfo.totalsz, 1); nfo.pckdvecsz = verify_pack(p->vecsz, tensor_sz(nfo.probsz)); e->i = rimpulse(d, impulse_amp, n, vecn, &nfo, inA, inB, inC, outA, outB, outC, tmp, rounds, tol); e->l = rlinear(N, &nfo, inA, inB, inC, outA, outB, outC, tmp, rounds,tol); e->s = t_shift(n, vecn, &nfo, inA, inB, outA, outB, tmp, rounds, tol, d); /* grr, verify-lib.c:preserves_input() only works for complex */ if (!p->in_place && !p->destroy_input) { bench_tensor *totalsz_swap, *pckdsz_swap; totalsz_swap = tensor_copy_swapio(nfo.totalsz); pckdsz_swap = tensor_copy_swapio(nfo.pckdsz); for (i = 0; i < rounds; ++i) { rarand(inA, N); dofft(&nfo, inA, outB); cpyr((R *) nfo.p->in, totalsz_swap, inB, pckdsz_swap); racmp(inB, inA, N, "preserves_input", 0.0); } tensor_destroy(totalsz_swap); tensor_destroy(pckdsz_swap); } tensor_destroy(nfo.totalsz); tensor_destroy(nfo.pckdsz); tensor_destroy(nfo.pckdvecsz); bench_free(tmp); bench_free(outC); bench_free(outB); bench_free(outA); bench_free(inC); bench_free(inB); bench_free(inA); bench_free(d); } typedef struct { dofft_closure k; bench_problem *p; int n0; } dofft_r2r_closure; static void cpyr1(int n, R *in, int is, R *out, int os, R scale) { int i; for (i = 0; i < n; ++i) out[i * os] = in[i * is] * scale; } static void mke00(C *a, int n, int c) { int i; for (i = 1; i + i < n; ++i) a[n - i][c] = a[i][c]; } static void mkre00(C *a, int n) { mkreal(a, n); mke00(a, n, 0); } static void mkimag(C *a, int n) { int i; for (i = 0; i < n; ++i) c_re(a[i]) = 0.0; } static void mko00(C *a, int n, int c) { int i; a[0][c] = 0.0; for (i = 1; i + i < n; ++i) a[n - i][c] = -a[i][c]; if (i + i == n) a[i][c] = 0.0; } static void mkro00(C *a, int n) { mkreal(a, n); mko00(a, n, 0); } static void mkio00(C *a, int n) { mkimag(a, n); mko00(a, n, 1); } static void mkre01(C *a, int n) /* n should be be multiple of 4 */ { R a0; a0 = c_re(a[0]); mko00(a, n/2, 0); c_re(a[n/2]) = -(c_re(a[0]) = a0); mkre00(a, n); } static void mkro01(C *a, int n) /* n should be be multiple of 4 */ { c_re(a[0]) = c_im(a[0]) = 0.0; mkre00(a, n/2); mkro00(a, n); } static void mkoddonly(C *a, int n) { int i; for (i = 0; i < n; i += 2) c_re(a[i]) = c_im(a[i]) = 0.0; } static void mkre10(C *a, int n) { mkoddonly(a, n); mkre00(a, n); } static void mkio10(C *a, int n) { mkoddonly(a, n); mkio00(a, n); } static void mkre11(C *a, int n) { mkoddonly(a, n); mko00(a, n/2, 0); mkre00(a, n); } static void mkro11(C *a, int n) { mkoddonly(a, n); mkre00(a, n/2); mkro00(a, n); } static void mkio11(C *a, int n) { mkoddonly(a, n); mke00(a, n/2, 1); mkio00(a, n); } static void r2r_apply(dofft_closure *k_, bench_complex *in, bench_complex *out) { dofft_r2r_closure *k = (dofft_r2r_closure *)k_; bench_problem *p = k->p; bench_real *ri, *ro; int n, is, os; n = p->sz->dims[0].n; is = p->sz->dims[0].is; os = p->sz->dims[0].os; ri = (bench_real *) p->in; ro = (bench_real *) p->out; switch (p->k[0]) { case R2R_R2HC: cpyr1(n, &c_re(in[0]), 2, ri, is, 1.0); break; case R2R_HC2R: cpyr1(n/2 + 1, &c_re(in[0]), 2, ri, is, 1.0); cpyr1((n+1)/2 - 1, &c_im(in[n-1]), -2, ri + is*(n-1), -is, 1.0); break; case R2R_REDFT00: cpyr1(n, &c_re(in[0]), 2, ri, is, 1.0); break; case R2R_RODFT00: cpyr1(n, &c_re(in[1]), 2, ri, is, 1.0); break; case R2R_REDFT01: cpyr1(n, &c_re(in[0]), 2, ri, is, 1.0); break; case R2R_REDFT10: cpyr1(n, &c_re(in[1]), 4, ri, is, 1.0); break; case R2R_RODFT01: cpyr1(n, &c_re(in[1]), 2, ri, is, 1.0); break; case R2R_RODFT10: cpyr1(n, &c_im(in[1]), 4, ri, is, 1.0); break; case R2R_REDFT11: cpyr1(n, &c_re(in[1]), 4, ri, is, 1.0); break; case R2R_RODFT11: cpyr1(n, &c_re(in[1]), 4, ri, is, 1.0); break; default: BENCH_ASSERT(0); /* not yet implemented */ } after_problem_rcopy_from(p, ri); doit(1, p); after_problem_rcopy_to(p, ro); switch (p->k[0]) { case R2R_R2HC: if (k->k.recopy_input) cpyr1(n, ri, is, &c_re(in[0]), 2, 1.0); cpyr1(n/2 + 1, ro, os, &c_re(out[0]), 2, 1.0); cpyr1((n+1)/2 - 1, ro + os*(n-1), -os, &c_im(out[1]), 2, 1.0); c_im(out[0]) = 0.0; if (n % 2 == 0) c_im(out[n/2]) = 0.0; mkhermitian1(out, n); break; case R2R_HC2R: if (k->k.recopy_input) { cpyr1(n/2 + 1, ri, is, &c_re(in[0]), 2, 1.0); cpyr1((n+1)/2 - 1, ri + is*(n-1), -is, &c_im(in[1]), 2,1.0); } cpyr1(n, ro, os, &c_re(out[0]), 2, 1.0); mkreal(out, n); break; case R2R_REDFT00: if (k->k.recopy_input) cpyr1(n, ri, is, &c_re(in[0]), 2, 1.0); cpyr1(n, ro, os, &c_re(out[0]), 2, 1.0); mkre00(out, k->n0); break; case R2R_RODFT00: if (k->k.recopy_input) cpyr1(n, ri, is, &c_im(in[1]), 2, -1.0); cpyr1(n, ro, os, &c_im(out[1]), 2, -1.0); mkio00(out, k->n0); break; case R2R_REDFT01: if (k->k.recopy_input) cpyr1(n, ri, is, &c_re(in[0]), 2, 1.0); cpyr1(n, ro, os, &c_re(out[1]), 4, 2.0); mkre10(out, k->n0); break; case R2R_REDFT10: if (k->k.recopy_input) cpyr1(n, ri, is, &c_re(in[1]), 4, 2.0); cpyr1(n, ro, os, &c_re(out[0]), 2, 1.0); mkre01(out, k->n0); break; case R2R_RODFT01: if (k->k.recopy_input) cpyr1(n, ri, is, &c_re(in[1]), 2, 1.0); cpyr1(n, ro, os, &c_im(out[1]), 4, -2.0); mkio10(out, k->n0); break; case R2R_RODFT10: if (k->k.recopy_input) cpyr1(n, ri, is, &c_im(in[1]), 4, -2.0); cpyr1(n, ro, os, &c_re(out[1]), 2, 1.0); mkro01(out, k->n0); break; case R2R_REDFT11: if (k->k.recopy_input) cpyr1(n, ri, is, &c_re(in[1]), 4, 2.0); cpyr1(n, ro, os, &c_re(out[1]), 4, 2.0); mkre11(out, k->n0); break; case R2R_RODFT11: if (k->k.recopy_input) cpyr1(n, ri, is, &c_im(in[1]), 4, -2.0); cpyr1(n, ro, os, &c_im(out[1]), 4, -2.0); mkio11(out, k->n0); break; default: BENCH_ASSERT(0); /* not yet implemented */ } } void accuracy_r2r(bench_problem *p, int rounds, int impulse_rounds, double t[6]) { dofft_r2r_closure k; int n, n0 = 1; C *a, *b; aconstrain constrain = 0; BENCH_ASSERT(p->kind == PROBLEM_R2R); BENCH_ASSERT(p->sz->rnk == 1); BENCH_ASSERT(p->vecsz->rnk == 0); k.k.apply = r2r_apply; k.k.recopy_input = 0; k.p = p; n = tensor_sz(p->sz); switch (p->k[0]) { case R2R_R2HC: constrain = mkreal; n0 = n; break; case R2R_HC2R: constrain = mkhermitian1; n0 = n; break; case R2R_REDFT00: constrain = mkre00; n0 = 2*(n-1); break; case R2R_RODFT00: constrain = mkro00; n0 = 2*(n+1); break; case R2R_REDFT01: constrain = mkre01; n0 = 4*n; break; case R2R_REDFT10: constrain = mkre10; n0 = 4*n; break; case R2R_RODFT01: constrain = mkro01; n0 = 4*n; break; case R2R_RODFT10: constrain = mkio10; n0 = 4*n; break; case R2R_REDFT11: constrain = mkre11; n0 = 8*n; break; case R2R_RODFT11: constrain = mkro11; n0 = 8*n; break; default: BENCH_ASSERT(0); /* not yet implemented */ } k.n0 = n0; a = (C *) bench_malloc(n0 * sizeof(C)); b = (C *) bench_malloc(n0 * sizeof(C)); accuracy_test(&k.k, constrain, -1, n0, a, b, rounds, impulse_rounds, t); bench_free(b); bench_free(a); } fftw-3.3.3/libbench2/verify-rdft2.c0000644000175000017500000002170512054410050013722 00000000000000/* * Copyright (c) 2003, 2007-11 Matteo Frigo * Copyright (c) 2003, 2007-11 Massachusetts Institute of Technology * * This program is free software; you can redistribute it and/or modify * it under the terms of the GNU General Public License as published by * the Free Software Foundation; either version 2 of the License, or * (at your option) any later version. * * This program is distributed in the hope that it will be useful, * but WITHOUT ANY WARRANTY; without even the implied warranty of * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the * GNU General Public License for more details. * * You should have received a copy of the GNU General Public License * along with this program; if not, write to the Free Software * Foundation, Inc., 51 Franklin Street, Fifth Floor, Boston, MA 02110-1301 USA * */ #include "verify.h" /* copy real A into real B, using output stride of A and input stride of B */ typedef struct { dotens2_closure k; R *ra; R *rb; } cpyr_closure; static void cpyr0(dotens2_closure *k_, int indxa, int ondxa, int indxb, int ondxb) { cpyr_closure *k = (cpyr_closure *)k_; k->rb[indxb] = k->ra[ondxa]; UNUSED(indxa); UNUSED(ondxb); } static void cpyr(R *ra, const bench_tensor *sza, R *rb, const bench_tensor *szb) { cpyr_closure k; k.k.apply = cpyr0; k.ra = ra; k.rb = rb; bench_dotens2(sza, szb, &k.k); } /* copy unpacked halfcomplex A[n] into packed-complex B[n], using output stride of A and input stride of B. Only copies non-redundant half; other half must be copied via mkhermitian. */ typedef struct { dotens2_closure k; int n; int as; int scalea; R *ra, *ia; R *rb, *ib; } cpyhc2_closure; static void cpyhc20(dotens2_closure *k_, int indxa, int ondxa, int indxb, int ondxb) { cpyhc2_closure *k = (cpyhc2_closure *)k_; int i, n = k->n; int scalea = k->scalea; int as = k->as * scalea; R *ra = k->ra + ondxa * scalea, *ia = k->ia + ondxa * scalea; R *rb = k->rb + indxb, *ib = k->ib + indxb; UNUSED(indxa); UNUSED(ondxb); for (i = 0; i < n/2 + 1; ++i) { rb[2*i] = ra[as*i]; ib[2*i] = ia[as*i]; } } static void cpyhc2(R *ra, R *ia, const bench_tensor *sza, const bench_tensor *vecsza, int scalea, R *rb, R *ib, const bench_tensor *szb) { cpyhc2_closure k; BENCH_ASSERT(sza->rnk <= 1); k.k.apply = cpyhc20; k.n = tensor_sz(sza); k.scalea = scalea; if (!FINITE_RNK(sza->rnk) || sza->rnk == 0) k.as = 0; else k.as = sza->dims[0].os; k.ra = ra; k.ia = ia; k.rb = rb; k.ib = ib; bench_dotens2(vecsza, szb, &k.k); } /* icpyhc2 is the inverse of cpyhc2 */ static void icpyhc20(dotens2_closure *k_, int indxa, int ondxa, int indxb, int ondxb) { cpyhc2_closure *k = (cpyhc2_closure *)k_; int i, n = k->n; int scalea = k->scalea; int as = k->as * scalea; R *ra = k->ra + indxa * scalea, *ia = k->ia + indxa * scalea; R *rb = k->rb + ondxb, *ib = k->ib + ondxb; UNUSED(ondxa); UNUSED(indxb); for (i = 0; i < n/2 + 1; ++i) { ra[as*i] = rb[2*i]; ia[as*i] = ib[2*i]; } } static void icpyhc2(R *ra, R *ia, const bench_tensor *sza, const bench_tensor *vecsza, int scalea, R *rb, R *ib, const bench_tensor *szb) { cpyhc2_closure k; BENCH_ASSERT(sza->rnk <= 1); k.k.apply = icpyhc20; k.n = tensor_sz(sza); k.scalea = scalea; if (!FINITE_RNK(sza->rnk) || sza->rnk == 0) k.as = 0; else k.as = sza->dims[0].is; k.ra = ra; k.ia = ia; k.rb = rb; k.ib = ib; bench_dotens2(vecsza, szb, &k.k); } typedef struct { dofft_closure k; bench_problem *p; } dofft_rdft2_closure; static void rdft2_apply(dofft_closure *k_, bench_complex *in, bench_complex *out) { dofft_rdft2_closure *k = (dofft_rdft2_closure *)k_; bench_problem *p = k->p; bench_tensor *totalsz, *pckdsz, *totalsz_swap, *pckdsz_swap; bench_tensor *probsz2, *totalsz2, *pckdsz2; bench_tensor *probsz2_swap, *totalsz2_swap, *pckdsz2_swap; bench_real *ri, *ii, *ro, *io; int n2, totalscale; totalsz = tensor_append(p->vecsz, p->sz); pckdsz = verify_pack(totalsz, 2); n2 = tensor_sz(totalsz); if (FINITE_RNK(p->sz->rnk) && p->sz->rnk > 0) n2 = (n2 / p->sz->dims[p->sz->rnk - 1].n) * (p->sz->dims[p->sz->rnk - 1].n / 2 + 1); ri = (bench_real *) p->in; ro = (bench_real *) p->out; if (FINITE_RNK(p->sz->rnk) && p->sz->rnk > 0 && n2 > 0) { probsz2 = tensor_copy_sub(p->sz, p->sz->rnk - 1, 1); totalsz2 = tensor_copy_sub(totalsz, 0, totalsz->rnk - 1); pckdsz2 = tensor_copy_sub(pckdsz, 0, pckdsz->rnk - 1); } else { probsz2 = mktensor(0); totalsz2 = tensor_copy(totalsz); pckdsz2 = tensor_copy(pckdsz); } totalsz_swap = tensor_copy_swapio(totalsz); pckdsz_swap = tensor_copy_swapio(pckdsz); totalsz2_swap = tensor_copy_swapio(totalsz2); pckdsz2_swap = tensor_copy_swapio(pckdsz2); probsz2_swap = tensor_copy_swapio(probsz2); /* confusion: the stride is the distance between complex elements when using interleaved format, but it is the distance between real elements when using split format */ if (p->split) { ii = p->ini ? (bench_real *) p->ini : ri + n2; io = p->outi ? (bench_real *) p->outi : ro + n2; totalscale = 1; } else { ii = p->ini ? (bench_real *) p->ini : ri + 1; io = p->outi ? (bench_real *) p->outi : ro + 1; totalscale = 2; } if (p->sign < 0) { /* R2HC */ int N, vN, i; cpyr(&c_re(in[0]), pckdsz, ri, totalsz); after_problem_rcopy_from(p, ri); doit(1, p); after_problem_hccopy_to(p, ro, io); if (k->k.recopy_input) cpyr(ri, totalsz_swap, &c_re(in[0]), pckdsz_swap); cpyhc2(ro, io, probsz2, totalsz2, totalscale, &c_re(out[0]), &c_im(out[0]), pckdsz2); N = tensor_sz(p->sz); vN = tensor_sz(p->vecsz); for (i = 0; i < vN; ++i) mkhermitian(out + i*N, p->sz->rnk, p->sz->dims, 1); } else { /* HC2R */ icpyhc2(ri, ii, probsz2, totalsz2, totalscale, &c_re(in[0]), &c_im(in[0]), pckdsz2); after_problem_hccopy_from(p, ri, ii); doit(1, p); after_problem_rcopy_to(p, ro); if (k->k.recopy_input) cpyhc2(ri, ii, probsz2_swap, totalsz2_swap, totalscale, &c_re(in[0]), &c_im(in[0]), pckdsz2_swap); mkreal(out, tensor_sz(pckdsz)); cpyr(ro, totalsz, &c_re(out[0]), pckdsz); } tensor_destroy(totalsz); tensor_destroy(pckdsz); tensor_destroy(totalsz_swap); tensor_destroy(pckdsz_swap); tensor_destroy(probsz2); tensor_destroy(totalsz2); tensor_destroy(pckdsz2); tensor_destroy(probsz2_swap); tensor_destroy(totalsz2_swap); tensor_destroy(pckdsz2_swap); } void verify_rdft2(bench_problem *p, int rounds, double tol, errors *e) { C *inA, *inB, *inC, *outA, *outB, *outC, *tmp; int n, vecn, N; dofft_rdft2_closure k; BENCH_ASSERT(p->kind == PROBLEM_REAL); if (!FINITE_RNK(p->sz->rnk) || !FINITE_RNK(p->vecsz->rnk)) return; /* give up */ k.k.apply = rdft2_apply; k.k.recopy_input = 0; k.p = p; if (rounds == 0) rounds = 20; /* default value */ n = tensor_sz(p->sz); vecn = tensor_sz(p->vecsz); N = n * vecn; inA = (C *) bench_malloc(N * sizeof(C)); inB = (C *) bench_malloc(N * sizeof(C)); inC = (C *) bench_malloc(N * sizeof(C)); outA = (C *) bench_malloc(N * sizeof(C)); outB = (C *) bench_malloc(N * sizeof(C)); outC = (C *) bench_malloc(N * sizeof(C)); tmp = (C *) bench_malloc(N * sizeof(C)); e->i = impulse(&k.k, n, vecn, inA, inB, inC, outA, outB, outC, tmp, rounds, tol); e->l = linear(&k.k, 1, N, inA, inB, inC, outA, outB, outC, tmp, rounds, tol); e->s = 0.0; if (p->sign < 0) e->s = dmax(e->s, tf_shift(&k.k, 1, p->sz, n, vecn, p->sign, inA, inB, outA, outB, tmp, rounds, tol, TIME_SHIFT)); else e->s = dmax(e->s, tf_shift(&k.k, 1, p->sz, n, vecn, p->sign, inA, inB, outA, outB, tmp, rounds, tol, FREQ_SHIFT)); if (!p->in_place && !p->destroy_input) preserves_input(&k.k, p->sign < 0 ? mkreal : mkhermitian1, N, inA, inB, outB, rounds); bench_free(tmp); bench_free(outC); bench_free(outB); bench_free(outA); bench_free(inC); bench_free(inB); bench_free(inA); } void accuracy_rdft2(bench_problem *p, int rounds, int impulse_rounds, double t[6]) { dofft_rdft2_closure k; int n; C *a, *b; BENCH_ASSERT(p->kind == PROBLEM_REAL); BENCH_ASSERT(p->sz->rnk == 1); BENCH_ASSERT(p->vecsz->rnk == 0); k.k.apply = rdft2_apply; k.k.recopy_input = 0; k.p = p; n = tensor_sz(p->sz); a = (C *) bench_malloc(n * sizeof(C)); b = (C *) bench_malloc(n * sizeof(C)); accuracy_test(&k.k, p->sign < 0 ? mkreal : mkhermitian1, p->sign, n, a, b, rounds, impulse_rounds, t); bench_free(b); bench_free(a); } fftw-3.3.3/libbench2/verify.c0000644000175000017500000000477012054410050012706 00000000000000/* * Copyright (c) 2000 Matteo Frigo * Copyright (c) 2000 Massachusetts Institute of Technology * * This program is free software; you can redistribute it and/or modify * it under the terms of the GNU General Public License as published by * the Free Software Foundation; either version 2 of the License, or * (at your option) any later version. * * This program is distributed in the hope that it will be useful, * but WITHOUT ANY WARRANTY; without even the implied warranty of * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the * GNU General Public License for more details. * * You should have received a copy of the GNU General Public License * along with this program; if not, write to the Free Software * Foundation, Inc., 51 Franklin Street, Fifth Floor, Boston, MA 02110-1301 USA * */ #include #include #include "verify.h" void verify_problem(bench_problem *p, int rounds, double tol) { errors e; const char *pstring = p->pstring ? p->pstring : ""; switch (p->kind) { case PROBLEM_COMPLEX: verify_dft(p, rounds, tol, &e); break; case PROBLEM_REAL: verify_rdft2(p, rounds, tol, &e); break; case PROBLEM_R2R: verify_r2r(p, rounds, tol, &e); break; } if (verbose) ovtpvt("%s %g %g %g\n", pstring, e.l, e.i, e.s); } void verify(const char *param, int rounds, double tol) { bench_problem *p; p = problem_parse(param); problem_alloc(p); if (!can_do(p)) { ovtpvt_err("No can_do for %s\n", p->pstring); BENCH_ASSERT(0); } problem_zero(p); setup(p); verify_problem(p, rounds, tol); done(p); problem_destroy(p); } static void do_accuracy(bench_problem *p, int rounds, int impulse_rounds) { double t[6]; switch (p->kind) { case PROBLEM_COMPLEX: accuracy_dft(p, rounds, impulse_rounds, t); break; case PROBLEM_REAL: accuracy_rdft2(p, rounds, impulse_rounds, t); break; case PROBLEM_R2R: accuracy_r2r(p, rounds, impulse_rounds, t); break; } /* t[0] : L1 error t[1] : L2 error t[2] : Linf error t[3..5]: L1, L2, Linf backward error */ ovtpvt("%6.2e %6.2e %6.2e %6.2e %6.2e %6.2e\n", t[0], t[1], t[2], t[3], t[4], t[5]); } void accuracy(const char *param, int rounds, int impulse_rounds) { bench_problem *p; p = problem_parse(param); BENCH_ASSERT(can_do(p)); problem_alloc(p); problem_zero(p); setup(p); do_accuracy(p, rounds, impulse_rounds); done(p); problem_destroy(p); } fftw-3.3.3/libbench2/zero.c0000644000175000017500000000317712054410050012361 00000000000000/* * Copyright (c) 2001 Matteo Frigo * Copyright (c) 2001 Massachusetts Institute of Technology * * This program is free software; you can redistribute it and/or modify * it under the terms of the GNU General Public License as published by * the Free Software Foundation; either version 2 of the License, or * (at your option) any later version. * * This program is distributed in the hope that it will be useful, * but WITHOUT ANY WARRANTY; without even the implied warranty of * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the * GNU General Public License for more details. * * You should have received a copy of the GNU General Public License * along with this program; if not, write to the Free Software * Foundation, Inc., 51 Franklin Street, Fifth Floor, Boston, MA 02110-1301 USA * */ #include "bench.h" /* set I/O arrays to zero. Default routine */ void problem_zero(bench_problem *p) { bench_complex czero = {0, 0}; if (p->kind == PROBLEM_COMPLEX) { caset((bench_complex *) p->inphys, p->iphyssz, czero); caset((bench_complex *) p->outphys, p->ophyssz, czero); } else if (p->kind == PROBLEM_R2R) { aset((bench_real *) p->inphys, p->iphyssz, 0.0); aset((bench_real *) p->outphys, p->ophyssz, 0.0); } else if (p->kind == PROBLEM_REAL && p->sign < 0) { aset((bench_real *) p->inphys, p->iphyssz, 0.0); caset((bench_complex *) p->outphys, p->ophyssz, czero); } else if (p->kind == PROBLEM_REAL && p->sign > 0) { caset((bench_complex *) p->inphys, p->iphyssz, czero); aset((bench_real *) p->outphys, p->ophyssz, 0.0); } else { BENCH_ASSERT(0); /* TODO */ } } fftw-3.3.3/libbench2/bench-user.h0000644000175000017500000001761112054410050013440 00000000000000/* * Copyright (c) 2001 Matteo Frigo * Copyright (c) 2001 Massachusetts Institute of Technology * * This program is free software; you can redistribute it and/or modify * it under the terms of the GNU General Public License as published by * the Free Software Foundation; either version 2 of the License, or * (at your option) any later version. * * This program is distributed in the hope that it will be useful, * but WITHOUT ANY WARRANTY; without even the implied warranty of * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the * GNU General Public License for more details. * * You should have received a copy of the GNU General Public License * along with this program; if not, write to the Free Software * Foundation, Inc., 51 Franklin Street, Fifth Floor, Boston, MA 02110-1301 USA * */ #ifndef __BENCH_USER_H__ #define __BENCH_USER_H__ #ifdef __cplusplus extern "C" { #endif /* __cplusplus */ /* benchmark program definitions for user code */ #include "config.h" #if HAVE_STDDEF_H #include #endif #if HAVE_STDLIB_H #include #endif #if defined(BENCHFFT_SINGLE) typedef float bench_real; #elif defined(BENCHFFT_LDOUBLE) typedef long double bench_real; #elif defined(BENCHFFT_QUAD) typedef __float128 bench_real; #else typedef double bench_real; #endif typedef bench_real bench_complex[2]; #define c_re(c) ((c)[0]) #define c_im(c) ((c)[1]) #undef DOUBLE_PRECISION #define DOUBLE_PRECISION (sizeof(bench_real) == sizeof(double)) #undef SINGLE_PRECISION #define SINGLE_PRECISION (!DOUBLE_PRECISION && sizeof(bench_real) == sizeof(float)) #undef LDOUBLE_PRECISION #define LDOUBLE_PRECISION (!DOUBLE_PRECISION && sizeof(bench_real) == sizeof(long double)) #undef QUAD_PRECISION #ifdef BENCHFFT_QUAD #define QUAD_PRECISION (!LDOUBLE_PRECISION && sizeof(bench_real) == sizeof(__float128)) #else #define QUAD_PRECISION 0 #endif typedef enum { PROBLEM_COMPLEX, PROBLEM_REAL, PROBLEM_R2R } problem_kind_t; typedef enum { R2R_R2HC, R2R_HC2R, R2R_DHT, R2R_REDFT00, R2R_REDFT01, R2R_REDFT10, R2R_REDFT11, R2R_RODFT00, R2R_RODFT01, R2R_RODFT10, R2R_RODFT11 } r2r_kind_t; typedef struct { int n; int is; /* input stride */ int os; /* output stride */ } bench_iodim; typedef struct { int rnk; bench_iodim *dims; } bench_tensor; bench_tensor *mktensor(int rnk); void tensor_destroy(bench_tensor *sz); int tensor_sz(const bench_tensor *sz); bench_tensor *tensor_compress(const bench_tensor *sz); int tensor_unitstridep(bench_tensor *t); int tensor_rowmajorp(bench_tensor *t); int tensor_real_rowmajorp(bench_tensor *t, int sign, int in_place); bench_tensor *tensor_append(const bench_tensor *a, const bench_tensor *b); bench_tensor *tensor_copy(const bench_tensor *sz); bench_tensor *tensor_copy_sub(const bench_tensor *sz, int start_dim, int rnk); bench_tensor *tensor_copy_swapio(const bench_tensor *sz); void tensor_ibounds(bench_tensor *t, int *lbp, int *ubp); void tensor_obounds(bench_tensor *t, int *lbp, int *ubp); /* Definition of rank -infinity. This definition has the property that if you want rank 0 or 1, you can simply test for rank <= 1. This is a common case. A tensor of rank -infinity has size 0. */ #define RNK_MINFTY ((int)(((unsigned) -1) >> 1)) #define FINITE_RNK(rnk) ((rnk) != RNK_MINFTY) typedef struct { problem_kind_t kind; r2r_kind_t *k; bench_tensor *sz; bench_tensor *vecsz; int sign; int in_place; int destroy_input; int split; void *in, *out; void *inphys, *outphys; int iphyssz, ophyssz; char *pstring; void *userinfo; /* user can store whatever */ int scrambled_in, scrambled_out; /* hack for MPI */ /* internal hack so that we can use verifier in FFTW test program */ void *ini, *outi; /* if nonzero, point to imag. parts for dft */ /* another internal hack to avoid passing around too many parameters */ double setup_time; } bench_problem; extern int verbose; extern int no_speed_allocation; extern int always_pad_real; #define LIBBENCH_TIMER 0 #define USER_TIMER 1 #define BENCH_NTIMERS 2 extern void timer_start(int which_timer); extern double timer_stop(int which_timer); extern int can_do(bench_problem *p); extern void setup(bench_problem *p); extern void doit(int iter, bench_problem *p); extern void done(bench_problem *p); extern void main_init(int *argc, char ***argv); extern void cleanup(void); extern void verify(const char *param, int rounds, double tol); extern void useropt(const char *arg); extern void verify_problem(bench_problem *p, int rounds, double tol); extern void problem_alloc(bench_problem *p); extern void problem_free(bench_problem *p); extern void problem_zero(bench_problem *p); extern void problem_destroy(bench_problem *p); extern int power_of_two(int n); extern int log_2(int n); #define CASSIGN(out, in) (c_re(out) = c_re(in), c_im(out) = c_im(in)) bench_tensor *verify_pack(const bench_tensor *sz, int s); typedef struct { double l; double i; double s; } errors; void verify_dft(bench_problem *p, int rounds, double tol, errors *e); void verify_rdft2(bench_problem *p, int rounds, double tol, errors *e); void verify_r2r(bench_problem *p, int rounds, double tol, errors *e); /**************************************************************/ /* routines to override */ extern void after_problem_ccopy_from(bench_problem *p, bench_real *ri, bench_real *ii); extern void after_problem_ccopy_to(bench_problem *p, bench_real *ro, bench_real *io); extern void after_problem_hccopy_from(bench_problem *p, bench_real *ri, bench_real *ii); extern void after_problem_hccopy_to(bench_problem *p, bench_real *ro, bench_real *io); extern void after_problem_rcopy_from(bench_problem *p, bench_real *ri); extern void after_problem_rcopy_to(bench_problem *p, bench_real *ro); extern void bench_exit(int status); extern double bench_cost_postprocess(double cost); /************************************************************** * malloc **************************************************************/ extern void *bench_malloc(size_t size); extern void bench_free(void *ptr); extern void bench_free0(void *ptr); /************************************************************** * alloca **************************************************************/ #ifdef HAVE_ALLOCA_H #include #endif /************************************************************** * assert **************************************************************/ extern void bench_assertion_failed(const char *s, int line, const char *file); #define BENCH_ASSERT(ex) \ (void)((ex) || (bench_assertion_failed(#ex, __LINE__, __FILE__), 0)) #define UNUSED(x) (void)x /*************************************** * Documentation strings ***************************************/ struct bench_doc { const char *key; const char *val; const char *(*f)(void); }; extern struct bench_doc bench_doc[]; #ifdef CC #define CC_DOC BENCH_DOC("cc", CC) #elif defined(BENCH_CC) #define CC_DOC BENCH_DOC("cc", BENCH_CC) #else #define CC_DOC /* none */ #endif #ifdef CXX #define CXX_DOC BENCH_DOC("cxx", CXX) #elif defined(BENCH_CXX) #define CXX_DOC BENCH_DOC("cxx", BENCH_CXX) #else #define CXX_DOC /* none */ #endif #ifdef F77 #define F77_DOC BENCH_DOC("f77", F77) #elif defined(BENCH_F77) #define F77_DOC BENCH_DOC("f77", BENCH_F77) #else #define F77_DOC /* none */ #endif #ifdef F90 #define F90_DOC BENCH_DOC("f90", F90) #elif defined(BENCH_F90) #define F90_DOC BENCH_DOC("f90", BENCH_F90) #else #define F90_DOC /* none */ #endif #define BEGIN_BENCH_DOC \ struct bench_doc bench_doc[] = { \ CC_DOC \ CXX_DOC \ F77_DOC \ F90_DOC #define BENCH_DOC(key, val) { key, val, 0 }, #define BENCH_DOCF(key, f) { key, 0, f }, #define END_BENCH_DOC \ {0, 0, 0}}; #ifdef __cplusplus } /* extern "C" */ #endif /* __cplusplus */ #endif /* __BENCH_USER_H__ */ fftw-3.3.3/libbench2/bench.h0000644000175000017500000000421712054410050012462 00000000000000/* * Copyright (c) 2001 Matteo Frigo * Copyright (c) 2001 Massachusetts Institute of Technology * * This program is free software; you can redistribute it and/or modify * it under the terms of the GNU General Public License as published by * the Free Software Foundation; either version 2 of the License, or * (at your option) any later version. * * This program is distributed in the hope that it will be useful, * but WITHOUT ANY WARRANTY; without even the implied warranty of * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the * GNU General Public License for more details. * * You should have received a copy of the GNU General Public License * along with this program; if not, write to the Free Software * Foundation, Inc., 51 Franklin Street, Fifth Floor, Boston, MA 02110-1301 USA * */ /* benchmark program definitions */ #include "bench-user.h" extern double time_min; extern int time_repeat; extern void timer_init(double tmin, int repeat); /* report functions */ extern void (*report)(const bench_problem *p, double *t, int st); void report_mflops(const bench_problem *p, double *t, int st); void report_time(const bench_problem *p, double *t, int st); void report_benchmark(const bench_problem *p, double *t, int st); void report_verbose(const bench_problem *p, double *t, int st); void report_can_do(const char *param); void report_info(const char *param); void report_info_all(void); extern int aligned_main(int argc, char *argv[]); extern int bench_main(int argc, char *argv[]); extern void speed(const char *param, int setup_only); extern void accuracy(const char *param, int rounds, int impulse_rounds); extern double mflops(const bench_problem *p, double t); extern double bench_drand(void); extern void bench_srand(int seed); extern bench_problem *problem_parse(const char *desc); extern void ovtpvt(const char *format, ...); extern void ovtpvt_err(const char *format, ...); extern void fftaccuracy(int n, bench_complex *a, bench_complex *ffta, int sign, double err[6]); extern void fftaccuracy_done(void); extern void caset(bench_complex *A, int n, bench_complex x); extern void aset(bench_real *A, int n, bench_real x); fftw-3.3.3/libbench2/verify.h0000644000175000017500000000672312054410050012713 00000000000000/* * Copyright (c) 2003, 2007-11 Matteo Frigo * Copyright (c) 2003, 2007-11 Massachusetts Institute of Technology * * This program is free software; you can redistribute it and/or modify * it under the terms of the GNU General Public License as published by * the Free Software Foundation; either version 2 of the License, or * (at your option) any later version. * * This program is distributed in the hope that it will be useful, * but WITHOUT ANY WARRANTY; without even the implied warranty of * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the * GNU General Public License for more details. * * You should have received a copy of the GNU General Public License * along with this program; if not, write to the Free Software * Foundation, Inc., 51 Franklin Street, Fifth Floor, Boston, MA 02110-1301 USA * */ #include "bench.h" typedef bench_real R; typedef bench_complex C; typedef struct dofft_closure_s { void (*apply)(struct dofft_closure_s *k, bench_complex *in, bench_complex *out); int recopy_input; } dofft_closure; double dmax(double x, double y); typedef void (*aconstrain)(C *a, int n); void arand(C *a, int n); void mkreal(C *A, int n); void mkhermitian(C *A, int rank, const bench_iodim *dim, int stride); void mkhermitian1(C *a, int n); void aadd(C *c, C *a, C *b, int n); void asub(C *c, C *a, C *b, int n); void arol(C *b, C *a, int n, int nb, int na); void aphase_shift(C *b, C *a, int n, int nb, int na, double sign); void ascale(C *a, C alpha, int n); double acmp(C *a, C *b, int n, const char *test, double tol); double mydrand(void); double impulse(dofft_closure *k, int n, int vecn, C *inA, C *inB, C *inC, C *outA, C *outB, C *outC, C *tmp, int rounds, double tol); double linear(dofft_closure *k, int realp, int n, C *inA, C *inB, C *inC, C *outA, C *outB, C *outC, C *tmp, int rounds, double tol); void preserves_input(dofft_closure *k, aconstrain constrain, int n, C *inA, C *inB, C *outB, int rounds); enum { TIME_SHIFT, FREQ_SHIFT }; double tf_shift(dofft_closure *k, int realp, const bench_tensor *sz, int n, int vecn, double sign, C *inA, C *inB, C *outA, C *outB, C *tmp, int rounds, double tol, int which_shift); typedef struct dotens2_closure_s { void (*apply)(struct dotens2_closure_s *k, int indx0, int ondx0, int indx1, int ondx1); } dotens2_closure; void bench_dotens2(const bench_tensor *sz0, const bench_tensor *sz1, dotens2_closure *k); void accuracy_test(dofft_closure *k, aconstrain constrain, int sign, int n, C *a, C *b, int rounds, int impulse_rounds, double t[6]); void accuracy_dft(bench_problem *p, int rounds, int impulse_rounds, double t[6]); void accuracy_rdft2(bench_problem *p, int rounds, int impulse_rounds, double t[6]); void accuracy_r2r(bench_problem *p, int rounds, int impulse_rounds, double t[6]); #if defined(BENCHFFT_LDOUBLE) && HAVE_COSL typedef long double trigreal; # define COS cosl # define SIN sinl # define TAN tanl # define KTRIG(x) (x##L) #elif defined(BENCHFFT_QUAD) && HAVE_LIBQUADMATH typedef __float128 trigreal; # define COS cosq # define SIN sinq # define TAN tanq # define KTRIG(x) (x##Q) extern trigreal cosq(trigreal); extern trigreal sinq(trigreal); extern trigreal tanq(trigreal); #else typedef double trigreal; # define COS cos # define SIN sin # define TAN tan # define KTRIG(x) (x) #endif #define K2PI KTRIG(6.2831853071795864769252867665590057683943388) fftw-3.3.3/libbench2/my-getopt.c0000644000175000017500000000757612054410050013336 00000000000000/* * Copyright (c) 2003, 2007-11 Matteo Frigo * Copyright (c) 2003, 2007-11 Massachusetts Institute of Technology * * This program is free software; you can redistribute it and/or modify * it under the terms of the GNU General Public License as published by * the Free Software Foundation; either version 2 of the License, or * (at your option) any later version. * * This program is distributed in the hope that it will be useful, * but WITHOUT ANY WARRANTY; without even the implied warranty of * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the * GNU General Public License for more details. * * You should have received a copy of the GNU General Public License * along with this program; if not, write to the Free Software * Foundation, Inc., 51 Franklin Street, Fifth Floor, Boston, MA 02110-1301 USA * */ #include #include #include "config.h" #include "my-getopt.h" int my_optind = 1; const char *my_optarg = 0; static const char *scan_pointer = 0; void my_usage(const char *progname, const struct my_option *opt) { int i; size_t col = 0; fprintf(stdout, "Usage: %s", progname); col += (strlen(progname) + 7); for (i = 0; opt[i].long_name; i++) { size_t option_len; option_len = strlen(opt[i].long_name); if (col >= 80 - (option_len + 16)) { fputs("\n\t", stdout); col = 8; } fprintf(stdout, " [--%s", opt[i].long_name); col += (option_len + 4); if (opt[i].short_name < 128) { fprintf(stdout, " | -%c", opt[i].short_name); col += 5; } switch (opt[i].argtype) { case REQARG: fputs(" arg]", stdout); col += 5; break; case OPTARG: fputs(" [arg]]", stdout); col += 10; break; default: fputs("]", stdout); col++; } } fputs ("\n", stdout); } int my_getopt(int argc, char *argv[], const struct my_option *optarray) { const char *p; const struct my_option *l; if (scan_pointer && *scan_pointer) { /* continue a previously scanned argv[] element */ p = scan_pointer; goto short_option; } else { /* new argv[] element */ if (my_optind >= argc) return -1; /* no more options */ p = argv[my_optind]; if (*p++ != '-') return (-1); /* not an option */ if (!*p) return (-1); /* string is exactly '-' */ ++my_optind; } if (*p == '-') { /* long option */ scan_pointer = 0; my_optarg = 0; ++p; for (l = optarray; l->short_name; ++l) { size_t len = strlen(l->long_name); if (!strncmp(l->long_name, p, len) && (!p[len] || p[len] == '=')) { switch (l->argtype) { case NOARG: goto ok; case OPTARG: if (p[len] == '=') my_optarg = p + len + 1; goto ok; case REQARG: if (p[len] == '=') { my_optarg = p + len + 1; goto ok; } if (my_optind >= argc) { fprintf(stderr, "option --%s requires an argument\n", l->long_name); return '?'; } my_optarg = argv[my_optind]; ++my_optind; goto ok; } } } } else { short_option: scan_pointer = 0; my_optarg = 0; for (l = optarray; l->short_name; ++l) { if (l->short_name == (char)l->short_name && *p == l->short_name) { ++p; switch (l->argtype) { case NOARG: scan_pointer = p; goto ok; case OPTARG: if (*p) my_optarg = p; goto ok; case REQARG: if (*p) { my_optarg = p; } else { if (my_optind >= argc) { fprintf(stderr, "option -%c requires an argument\n", l->short_name); return '?'; } my_optarg = argv[my_optind]; ++my_optind; } goto ok; } } } } fprintf(stderr, "unrecognized option %s\n", argv[my_optind - 1]); return '?'; ok: return l->short_name; } fftw-3.3.3/libbench2/my-getopt.h0000644000175000017500000000265712054410050013336 00000000000000/* * Copyright (c) 2003, 2007-11 Matteo Frigo * Copyright (c) 2003, 2007-11 Massachusetts Institute of Technology * * This program is free software; you can redistribute it and/or modify * it under the terms of the GNU General Public License as published by * the Free Software Foundation; either version 2 of the License, or * (at your option) any later version. * * This program is distributed in the hope that it will be useful, * but WITHOUT ANY WARRANTY; without even the implied warranty of * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the * GNU General Public License for more details. * * You should have received a copy of the GNU General Public License * along with this program; if not, write to the Free Software * Foundation, Inc., 51 Franklin Street, Fifth Floor, Boston, MA 02110-1301 USA * */ #ifndef __MY_GETOPT_H__ #define __MY_GETOPT_H__ #ifdef __cplusplus extern "C" { #endif /* __cplusplus */ enum { REQARG, OPTARG, NOARG }; struct my_option { const char *long_name; int argtype; int short_name; }; extern int my_optind; extern const char *my_optarg; extern void my_usage(const char *progname, const struct my_option *opt); extern int my_getopt(int argc, char *argv[], const struct my_option *optarray); #ifdef __cplusplus } /* extern "C" */ #endif /* __cplusplus */ #endif /* __MY_GETOPT_H__ */ fftw-3.3.3/threads/0000755000175000017500000000000012054411110011106 500000000000000fftw-3.3.3/threads/Makefile.am0000644000175000017500000000244712054410050013073 00000000000000AM_CPPFLAGS = -I$(top_srcdir)/kernel -I$(top_srcdir)/dft \ -I$(top_srcdir)/rdft -I$(top_srcdir)/api AM_CFLAGS = $(STACK_ALIGN_CFLAGS) if OPENMP FFTWOMPLIB = libfftw3@PREC_SUFFIX@_omp.la else FFTWOMPLIB = endif if THREADS if COMBINED_THREADS noinst_LTLIBRARIES = libfftw3@PREC_SUFFIX@_threads.la else lib_LTLIBRARIES = libfftw3@PREC_SUFFIX@_threads.la $(FFTWOMPLIB) endif else lib_LTLIBRARIES = $(FFTWOMPLIB) endif # pkgincludedir = $(includedir)/fftw3@PREC_SUFFIX@ # pkginclude_HEADERS = threads.h libfftw3@PREC_SUFFIX@_threads_la_SOURCES = api.c conf.c threads.c \ threads.h dft-vrank-geq1.c ct.c rdft-vrank-geq1.c hc2hc.c \ vrank-geq1-rdft2.c f77api.c f77funcs.h libfftw3@PREC_SUFFIX@_threads_la_CFLAGS = $(AM_CFLAGS) $(PTHREAD_CFLAGS) libfftw3@PREC_SUFFIX@_threads_la_LDFLAGS = -version-info @SHARED_VERSION_INFO@ if !COMBINED_THREADS libfftw3@PREC_SUFFIX@_threads_la_LIBADD = ../libfftw3@PREC_SUFFIX@.la endif libfftw3@PREC_SUFFIX@_omp_la_SOURCES = api.c conf.c openmp.c \ threads.h dft-vrank-geq1.c ct.c rdft-vrank-geq1.c hc2hc.c \ vrank-geq1-rdft2.c f77api.c f77funcs.h libfftw3@PREC_SUFFIX@_omp_la_CFLAGS = $(AM_CFLAGS) $(OPENMP_CFLAGS) libfftw3@PREC_SUFFIX@_omp_la_LDFLAGS = -version-info @SHARED_VERSION_INFO@ if !COMBINED_THREADS libfftw3@PREC_SUFFIX@_omp_la_LIBADD = ../libfftw3@PREC_SUFFIX@.la endif fftw-3.3.3/threads/Makefile.in0000644000175000017500000013157412054410201013106 00000000000000# Makefile.in generated by automake 1.11.6 from Makefile.am. # @configure_input@ # Copyright (C) 1994, 1995, 1996, 1997, 1998, 1999, 2000, 2001, 2002, # 2003, 2004, 2005, 2006, 2007, 2008, 2009, 2010, 2011 Free Software # Foundation, Inc. # This Makefile.in is free software; the Free Software Foundation # gives unlimited permission to copy and/or distribute it, # with or without modifications, as long as this notice is preserved. # This program is distributed in the hope that it will be useful, # but WITHOUT ANY WARRANTY, to the extent permitted by law; without # even the implied warranty of MERCHANTABILITY or FITNESS FOR A # PARTICULAR PURPOSE. @SET_MAKE@ VPATH = @srcdir@ am__make_dryrun = \ { \ am__dry=no; \ case $$MAKEFLAGS in \ *\\[\ \ ]*) \ echo 'am--echo: ; @echo "AM" OK' | $(MAKE) -f - 2>/dev/null \ | grep '^AM OK$$' >/dev/null || am__dry=yes;; \ *) \ for am__flg in $$MAKEFLAGS; do \ case $$am__flg in \ *=*|--*) ;; \ *n*) am__dry=yes; break;; \ esac; \ done;; \ esac; \ test $$am__dry = yes; \ } pkgdatadir = $(datadir)/@PACKAGE@ pkgincludedir = $(includedir)/@PACKAGE@ pkglibdir = $(libdir)/@PACKAGE@ pkglibexecdir = $(libexecdir)/@PACKAGE@ am__cd = CDPATH="$${ZSH_VERSION+.}$(PATH_SEPARATOR)" && cd install_sh_DATA = $(install_sh) -c -m 644 install_sh_PROGRAM = $(install_sh) -c install_sh_SCRIPT = $(install_sh) -c INSTALL_HEADER = $(INSTALL_DATA) transform = $(program_transform_name) NORMAL_INSTALL = : PRE_INSTALL = : POST_INSTALL = : NORMAL_UNINSTALL = : PRE_UNINSTALL = : POST_UNINSTALL = : build_triplet = @build@ host_triplet = @host@ subdir = threads DIST_COMMON = $(srcdir)/Makefile.am $(srcdir)/Makefile.in ACLOCAL_M4 = $(top_srcdir)/aclocal.m4 am__aclocal_m4_deps = $(top_srcdir)/m4/acx_mpi.m4 \ $(top_srcdir)/m4/acx_pthread.m4 \ $(top_srcdir)/m4/ax_cc_maxopt.m4 \ $(top_srcdir)/m4/ax_check_compiler_flags.m4 \ $(top_srcdir)/m4/ax_compiler_vendor.m4 \ $(top_srcdir)/m4/ax_gcc_aligns_stack.m4 \ $(top_srcdir)/m4/ax_gcc_version.m4 \ $(top_srcdir)/m4/ax_openmp.m4 $(top_srcdir)/m4/libtool.m4 \ $(top_srcdir)/m4/ltoptions.m4 $(top_srcdir)/m4/ltsugar.m4 \ $(top_srcdir)/m4/ltversion.m4 $(top_srcdir)/m4/lt~obsolete.m4 \ $(top_srcdir)/configure.ac am__configure_deps = $(am__aclocal_m4_deps) $(CONFIGURE_DEPENDENCIES) \ $(ACLOCAL_M4) mkinstalldirs = $(install_sh) -d CONFIG_HEADER = $(top_builddir)/config.h CONFIG_CLEAN_FILES = CONFIG_CLEAN_VPATH_FILES = am__vpath_adj_setup = srcdirstrip=`echo "$(srcdir)" | sed 's|.|.|g'`; am__vpath_adj = case $$p in \ $(srcdir)/*) f=`echo "$$p" | sed "s|^$$srcdirstrip/||"`;; \ *) f=$$p;; \ esac; am__strip_dir = f=`echo $$p | sed -e 's|^.*/||'`; am__install_max = 40 am__nobase_strip_setup = \ srcdirstrip=`echo "$(srcdir)" | sed 's/[].[^$$\\*|]/\\\\&/g'` am__nobase_strip = \ for p in $$list; do echo "$$p"; done | sed -e "s|$$srcdirstrip/||" am__nobase_list = $(am__nobase_strip_setup); \ for p in $$list; do echo "$$p $$p"; done | \ sed "s| $$srcdirstrip/| |;"' / .*\//!s/ .*/ ./; s,\( .*\)/[^/]*$$,\1,' | \ $(AWK) 'BEGIN { files["."] = "" } { files[$$2] = files[$$2] " " $$1; \ if (++n[$$2] == $(am__install_max)) \ { print $$2, files[$$2]; n[$$2] = 0; files[$$2] = "" } } \ END { for (dir in files) print dir, files[dir] }' am__base_list = \ sed '$$!N;$$!N;$$!N;$$!N;$$!N;$$!N;$$!N;s/\n/ /g' | \ sed '$$!N;$$!N;$$!N;$$!N;s/\n/ /g' am__uninstall_files_from_dir = { \ test -z "$$files" \ || { test ! -d "$$dir" && test ! -f "$$dir" && test ! -r "$$dir"; } \ || { echo " ( cd '$$dir' && rm -f" $$files ")"; \ $(am__cd) "$$dir" && rm -f $$files; }; \ } am__installdirs = "$(DESTDIR)$(libdir)" LTLIBRARIES = $(lib_LTLIBRARIES) $(noinst_LTLIBRARIES) @COMBINED_THREADS_FALSE@libfftw3@PREC_SUFFIX@_omp_la_DEPENDENCIES = \ @COMBINED_THREADS_FALSE@ ../libfftw3@PREC_SUFFIX@.la am_libfftw3@PREC_SUFFIX@_omp_la_OBJECTS = \ libfftw3@PREC_SUFFIX@_omp_la-api.lo \ libfftw3@PREC_SUFFIX@_omp_la-conf.lo \ libfftw3@PREC_SUFFIX@_omp_la-openmp.lo \ libfftw3@PREC_SUFFIX@_omp_la-dft-vrank-geq1.lo \ libfftw3@PREC_SUFFIX@_omp_la-ct.lo \ libfftw3@PREC_SUFFIX@_omp_la-rdft-vrank-geq1.lo \ libfftw3@PREC_SUFFIX@_omp_la-hc2hc.lo \ libfftw3@PREC_SUFFIX@_omp_la-vrank-geq1-rdft2.lo \ libfftw3@PREC_SUFFIX@_omp_la-f77api.lo libfftw3@PREC_SUFFIX@_omp_la_OBJECTS = \ $(am_libfftw3@PREC_SUFFIX@_omp_la_OBJECTS) libfftw3@PREC_SUFFIX@_omp_la_LINK = $(LIBTOOL) --tag=CC \ $(AM_LIBTOOLFLAGS) $(LIBTOOLFLAGS) --mode=link $(CCLD) \ $(libfftw3@PREC_SUFFIX@_omp_la_CFLAGS) $(CFLAGS) \ $(libfftw3@PREC_SUFFIX@_omp_la_LDFLAGS) $(LDFLAGS) -o $@ @COMBINED_THREADS_FALSE@@OPENMP_TRUE@@THREADS_TRUE@am_libfftw3@PREC_SUFFIX@_omp_la_rpath = \ @COMBINED_THREADS_FALSE@@OPENMP_TRUE@@THREADS_TRUE@ -rpath \ @COMBINED_THREADS_FALSE@@OPENMP_TRUE@@THREADS_TRUE@ $(libdir) @OPENMP_TRUE@@THREADS_FALSE@am_libfftw3@PREC_SUFFIX@_omp_la_rpath = \ @OPENMP_TRUE@@THREADS_FALSE@ -rpath $(libdir) @COMBINED_THREADS_FALSE@libfftw3@PREC_SUFFIX@_threads_la_DEPENDENCIES = \ @COMBINED_THREADS_FALSE@ ../libfftw3@PREC_SUFFIX@.la am_libfftw3@PREC_SUFFIX@_threads_la_OBJECTS = \ libfftw3@PREC_SUFFIX@_threads_la-api.lo \ libfftw3@PREC_SUFFIX@_threads_la-conf.lo \ libfftw3@PREC_SUFFIX@_threads_la-threads.lo \ libfftw3@PREC_SUFFIX@_threads_la-dft-vrank-geq1.lo \ libfftw3@PREC_SUFFIX@_threads_la-ct.lo \ libfftw3@PREC_SUFFIX@_threads_la-rdft-vrank-geq1.lo \ libfftw3@PREC_SUFFIX@_threads_la-hc2hc.lo \ libfftw3@PREC_SUFFIX@_threads_la-vrank-geq1-rdft2.lo \ libfftw3@PREC_SUFFIX@_threads_la-f77api.lo libfftw3@PREC_SUFFIX@_threads_la_OBJECTS = \ $(am_libfftw3@PREC_SUFFIX@_threads_la_OBJECTS) libfftw3@PREC_SUFFIX@_threads_la_LINK = $(LIBTOOL) --tag=CC \ $(AM_LIBTOOLFLAGS) $(LIBTOOLFLAGS) --mode=link $(CCLD) \ $(libfftw3@PREC_SUFFIX@_threads_la_CFLAGS) $(CFLAGS) \ $(libfftw3@PREC_SUFFIX@_threads_la_LDFLAGS) $(LDFLAGS) -o $@ @COMBINED_THREADS_FALSE@@THREADS_TRUE@am_libfftw3@PREC_SUFFIX@_threads_la_rpath = \ @COMBINED_THREADS_FALSE@@THREADS_TRUE@ -rpath $(libdir) @COMBINED_THREADS_TRUE@@THREADS_TRUE@am_libfftw3@PREC_SUFFIX@_threads_la_rpath = DEFAULT_INCLUDES = -I.@am__isrc@ -I$(top_builddir) depcomp = $(SHELL) $(top_srcdir)/depcomp am__depfiles_maybe = depfiles am__mv = mv -f COMPILE = $(CC) $(DEFS) $(DEFAULT_INCLUDES) $(INCLUDES) $(AM_CPPFLAGS) \ $(CPPFLAGS) $(AM_CFLAGS) $(CFLAGS) LTCOMPILE = $(LIBTOOL) --tag=CC $(AM_LIBTOOLFLAGS) $(LIBTOOLFLAGS) \ --mode=compile $(CC) $(DEFS) $(DEFAULT_INCLUDES) $(INCLUDES) \ $(AM_CPPFLAGS) $(CPPFLAGS) $(AM_CFLAGS) $(CFLAGS) CCLD = $(CC) LINK = $(LIBTOOL) --tag=CC $(AM_LIBTOOLFLAGS) $(LIBTOOLFLAGS) \ --mode=link $(CCLD) $(AM_CFLAGS) $(CFLAGS) $(AM_LDFLAGS) \ $(LDFLAGS) -o $@ SOURCES = $(libfftw3@PREC_SUFFIX@_omp_la_SOURCES) \ $(libfftw3@PREC_SUFFIX@_threads_la_SOURCES) DIST_SOURCES = $(libfftw3@PREC_SUFFIX@_omp_la_SOURCES) \ $(libfftw3@PREC_SUFFIX@_threads_la_SOURCES) am__can_run_installinfo = \ case $$AM_UPDATE_INFO_DIR in \ n|no|NO) false;; \ *) (install-info --version) >/dev/null 2>&1;; \ esac ETAGS = etags CTAGS = ctags DISTFILES = $(DIST_COMMON) $(DIST_SOURCES) $(TEXINFOS) $(EXTRA_DIST) ACLOCAL = @ACLOCAL@ ALLOCA = @ALLOCA@ ALTIVEC_CFLAGS = @ALTIVEC_CFLAGS@ AMTAR = @AMTAR@ AR = @AR@ AS = @AS@ AUTOCONF = @AUTOCONF@ AUTOHEADER = @AUTOHEADER@ AUTOMAKE = @AUTOMAKE@ AVX_CFLAGS = @AVX_CFLAGS@ AWK = @AWK@ CC = @CC@ CCDEPMODE = @CCDEPMODE@ CFLAGS = @CFLAGS@ CHECK_PL_OPTS = @CHECK_PL_OPTS@ CPP = @CPP@ CPPFLAGS = @CPPFLAGS@ CYGPATH_W = @CYGPATH_W@ C_FFTW_R2R_KIND = @C_FFTW_R2R_KIND@ C_MPI_FINT = @C_MPI_FINT@ DEFS = @DEFS@ DEPDIR = @DEPDIR@ DLLTOOL = @DLLTOOL@ DSYMUTIL = @DSYMUTIL@ DUMPBIN = @DUMPBIN@ ECHO_C = @ECHO_C@ ECHO_N = @ECHO_N@ ECHO_T = @ECHO_T@ EGREP = @EGREP@ EXEEXT = @EXEEXT@ F77 = @F77@ FFLAGS = @FFLAGS@ FGREP = @FGREP@ FLIBS = @FLIBS@ GREP = @GREP@ INSTALL = @INSTALL@ INSTALL_DATA = @INSTALL_DATA@ INSTALL_PROGRAM = @INSTALL_PROGRAM@ INSTALL_SCRIPT = @INSTALL_SCRIPT@ INSTALL_STRIP_PROGRAM = @INSTALL_STRIP_PROGRAM@ LD = @LD@ LDFLAGS = @LDFLAGS@ LIBOBJS = @LIBOBJS@ LIBQUADMATH = @LIBQUADMATH@ LIBS = @LIBS@ LIBTOOL = @LIBTOOL@ LIPO = @LIPO@ LN_S = @LN_S@ LTLIBOBJS = @LTLIBOBJS@ MAINT = @MAINT@ MAKEINFO = @MAKEINFO@ MANIFEST_TOOL = @MANIFEST_TOOL@ MKDIR_P = @MKDIR_P@ MPICC = @MPICC@ MPILIBS = @MPILIBS@ MPIRUN = @MPIRUN@ NEON_CFLAGS = @NEON_CFLAGS@ NM = @NM@ NMEDIT = @NMEDIT@ OBJDUMP = @OBJDUMP@ OBJEXT = @OBJEXT@ OCAMLBUILD = @OCAMLBUILD@ OPENMP_CFLAGS = @OPENMP_CFLAGS@ OTOOL = @OTOOL@ OTOOL64 = @OTOOL64@ PACKAGE = @PACKAGE@ PACKAGE_BUGREPORT = @PACKAGE_BUGREPORT@ PACKAGE_NAME = @PACKAGE_NAME@ PACKAGE_STRING = @PACKAGE_STRING@ PACKAGE_TARNAME = @PACKAGE_TARNAME@ PACKAGE_URL = @PACKAGE_URL@ PACKAGE_VERSION = @PACKAGE_VERSION@ PATH_SEPARATOR = @PATH_SEPARATOR@ POW_LIB = @POW_LIB@ PRECISION = @PRECISION@ PREC_SUFFIX = @PREC_SUFFIX@ PTHREAD_CC = @PTHREAD_CC@ PTHREAD_CFLAGS = @PTHREAD_CFLAGS@ PTHREAD_LIBS = @PTHREAD_LIBS@ RANLIB = @RANLIB@ SED = @SED@ SET_MAKE = @SET_MAKE@ SHARED_VERSION_INFO = @SHARED_VERSION_INFO@ SHELL = @SHELL@ SSE2_CFLAGS = @SSE2_CFLAGS@ STACK_ALIGN_CFLAGS = @STACK_ALIGN_CFLAGS@ STRIP = @STRIP@ THREADLIBS = @THREADLIBS@ VERSION = @VERSION@ abs_builddir = @abs_builddir@ abs_srcdir = @abs_srcdir@ abs_top_builddir = @abs_top_builddir@ abs_top_srcdir = @abs_top_srcdir@ ac_ct_AR = @ac_ct_AR@ ac_ct_CC = @ac_ct_CC@ ac_ct_DUMPBIN = @ac_ct_DUMPBIN@ ac_ct_F77 = @ac_ct_F77@ acx_pthread_config = @acx_pthread_config@ am__include = @am__include@ am__leading_dot = @am__leading_dot@ am__quote = @am__quote@ am__tar = @am__tar@ am__untar = @am__untar@ bindir = @bindir@ build = @build@ build_alias = @build_alias@ build_cpu = @build_cpu@ build_os = @build_os@ build_vendor = @build_vendor@ builddir = @builddir@ datadir = @datadir@ datarootdir = @datarootdir@ docdir = @docdir@ dvidir = @dvidir@ exec_prefix = @exec_prefix@ host = @host@ host_alias = @host_alias@ host_cpu = @host_cpu@ host_os = @host_os@ host_vendor = @host_vendor@ htmldir = @htmldir@ includedir = @includedir@ infodir = @infodir@ install_sh = @install_sh@ libdir = @libdir@ libexecdir = @libexecdir@ localedir = @localedir@ localstatedir = @localstatedir@ mandir = @mandir@ mkdir_p = @mkdir_p@ oldincludedir = @oldincludedir@ pdfdir = @pdfdir@ prefix = @prefix@ program_transform_name = @program_transform_name@ psdir = @psdir@ sbindir = @sbindir@ sharedstatedir = @sharedstatedir@ srcdir = @srcdir@ sysconfdir = @sysconfdir@ target_alias = @target_alias@ top_build_prefix = @top_build_prefix@ top_builddir = @top_builddir@ top_srcdir = @top_srcdir@ AM_CPPFLAGS = -I$(top_srcdir)/kernel -I$(top_srcdir)/dft \ -I$(top_srcdir)/rdft -I$(top_srcdir)/api AM_CFLAGS = $(STACK_ALIGN_CFLAGS) @OPENMP_FALSE@FFTWOMPLIB = @OPENMP_TRUE@FFTWOMPLIB = libfftw3@PREC_SUFFIX@_omp.la @COMBINED_THREADS_TRUE@@THREADS_TRUE@noinst_LTLIBRARIES = libfftw3@PREC_SUFFIX@_threads.la @COMBINED_THREADS_FALSE@@THREADS_TRUE@lib_LTLIBRARIES = libfftw3@PREC_SUFFIX@_threads.la $(FFTWOMPLIB) @THREADS_FALSE@lib_LTLIBRARIES = $(FFTWOMPLIB) # pkgincludedir = $(includedir)/fftw3@PREC_SUFFIX@ # pkginclude_HEADERS = threads.h libfftw3@PREC_SUFFIX@_threads_la_SOURCES = api.c conf.c threads.c \ threads.h dft-vrank-geq1.c ct.c rdft-vrank-geq1.c hc2hc.c \ vrank-geq1-rdft2.c f77api.c f77funcs.h libfftw3@PREC_SUFFIX@_threads_la_CFLAGS = $(AM_CFLAGS) $(PTHREAD_CFLAGS) libfftw3@PREC_SUFFIX@_threads_la_LDFLAGS = -version-info @SHARED_VERSION_INFO@ @COMBINED_THREADS_FALSE@libfftw3@PREC_SUFFIX@_threads_la_LIBADD = ../libfftw3@PREC_SUFFIX@.la libfftw3@PREC_SUFFIX@_omp_la_SOURCES = api.c conf.c openmp.c \ threads.h dft-vrank-geq1.c ct.c rdft-vrank-geq1.c hc2hc.c \ vrank-geq1-rdft2.c f77api.c f77funcs.h libfftw3@PREC_SUFFIX@_omp_la_CFLAGS = $(AM_CFLAGS) $(OPENMP_CFLAGS) libfftw3@PREC_SUFFIX@_omp_la_LDFLAGS = -version-info @SHARED_VERSION_INFO@ @COMBINED_THREADS_FALSE@libfftw3@PREC_SUFFIX@_omp_la_LIBADD = ../libfftw3@PREC_SUFFIX@.la all: all-am .SUFFIXES: .SUFFIXES: .c .lo .o .obj $(srcdir)/Makefile.in: @MAINTAINER_MODE_TRUE@ $(srcdir)/Makefile.am $(am__configure_deps) @for dep in $?; do \ case '$(am__configure_deps)' in \ *$$dep*) \ ( cd $(top_builddir) && $(MAKE) $(AM_MAKEFLAGS) am--refresh ) \ && { if test -f $@; then exit 0; else break; fi; }; \ exit 1;; \ esac; \ done; \ echo ' cd $(top_srcdir) && $(AUTOMAKE) --gnu threads/Makefile'; \ $(am__cd) $(top_srcdir) && \ $(AUTOMAKE) --gnu threads/Makefile .PRECIOUS: Makefile Makefile: $(srcdir)/Makefile.in $(top_builddir)/config.status @case '$?' in \ *config.status*) \ cd $(top_builddir) && $(MAKE) $(AM_MAKEFLAGS) am--refresh;; \ *) \ echo ' cd $(top_builddir) && $(SHELL) ./config.status $(subdir)/$@ $(am__depfiles_maybe)'; \ cd $(top_builddir) && $(SHELL) ./config.status $(subdir)/$@ $(am__depfiles_maybe);; \ esac; $(top_builddir)/config.status: $(top_srcdir)/configure $(CONFIG_STATUS_DEPENDENCIES) cd $(top_builddir) && $(MAKE) $(AM_MAKEFLAGS) am--refresh $(top_srcdir)/configure: @MAINTAINER_MODE_TRUE@ $(am__configure_deps) cd $(top_builddir) && $(MAKE) $(AM_MAKEFLAGS) am--refresh $(ACLOCAL_M4): @MAINTAINER_MODE_TRUE@ $(am__aclocal_m4_deps) cd $(top_builddir) && $(MAKE) $(AM_MAKEFLAGS) am--refresh $(am__aclocal_m4_deps): install-libLTLIBRARIES: $(lib_LTLIBRARIES) @$(NORMAL_INSTALL) @list='$(lib_LTLIBRARIES)'; test -n "$(libdir)" || list=; \ list2=; for p in $$list; do \ if test -f $$p; then \ list2="$$list2 $$p"; \ else :; fi; \ done; \ test -z "$$list2" || { \ echo " $(MKDIR_P) '$(DESTDIR)$(libdir)'"; \ $(MKDIR_P) "$(DESTDIR)$(libdir)" || exit 1; \ echo " $(LIBTOOL) $(AM_LIBTOOLFLAGS) $(LIBTOOLFLAGS) --mode=install $(INSTALL) $(INSTALL_STRIP_FLAG) $$list2 '$(DESTDIR)$(libdir)'"; \ $(LIBTOOL) $(AM_LIBTOOLFLAGS) $(LIBTOOLFLAGS) --mode=install $(INSTALL) $(INSTALL_STRIP_FLAG) $$list2 "$(DESTDIR)$(libdir)"; \ } uninstall-libLTLIBRARIES: @$(NORMAL_UNINSTALL) @list='$(lib_LTLIBRARIES)'; test -n "$(libdir)" || list=; \ for p in $$list; do \ $(am__strip_dir) \ echo " $(LIBTOOL) $(AM_LIBTOOLFLAGS) $(LIBTOOLFLAGS) --mode=uninstall rm -f '$(DESTDIR)$(libdir)/$$f'"; \ $(LIBTOOL) $(AM_LIBTOOLFLAGS) $(LIBTOOLFLAGS) --mode=uninstall rm -f "$(DESTDIR)$(libdir)/$$f"; \ done clean-libLTLIBRARIES: -test -z "$(lib_LTLIBRARIES)" || rm -f $(lib_LTLIBRARIES) @list='$(lib_LTLIBRARIES)'; for p in $$list; do \ dir="`echo $$p | sed -e 's|/[^/]*$$||'`"; \ test "$$dir" != "$$p" || dir=.; \ echo "rm -f \"$${dir}/so_locations\""; \ rm -f "$${dir}/so_locations"; \ done clean-noinstLTLIBRARIES: -test -z "$(noinst_LTLIBRARIES)" || rm -f $(noinst_LTLIBRARIES) @list='$(noinst_LTLIBRARIES)'; for p in $$list; do \ dir="`echo $$p | sed -e 's|/[^/]*$$||'`"; \ test "$$dir" != "$$p" || dir=.; \ echo "rm -f \"$${dir}/so_locations\""; \ rm -f "$${dir}/so_locations"; \ done libfftw3@PREC_SUFFIX@_omp.la: $(libfftw3@PREC_SUFFIX@_omp_la_OBJECTS) $(libfftw3@PREC_SUFFIX@_omp_la_DEPENDENCIES) $(EXTRA_libfftw3@PREC_SUFFIX@_omp_la_DEPENDENCIES) $(libfftw3@PREC_SUFFIX@_omp_la_LINK) $(am_libfftw3@PREC_SUFFIX@_omp_la_rpath) $(libfftw3@PREC_SUFFIX@_omp_la_OBJECTS) $(libfftw3@PREC_SUFFIX@_omp_la_LIBADD) $(LIBS) libfftw3@PREC_SUFFIX@_threads.la: $(libfftw3@PREC_SUFFIX@_threads_la_OBJECTS) $(libfftw3@PREC_SUFFIX@_threads_la_DEPENDENCIES) $(EXTRA_libfftw3@PREC_SUFFIX@_threads_la_DEPENDENCIES) $(libfftw3@PREC_SUFFIX@_threads_la_LINK) $(am_libfftw3@PREC_SUFFIX@_threads_la_rpath) $(libfftw3@PREC_SUFFIX@_threads_la_OBJECTS) $(libfftw3@PREC_SUFFIX@_threads_la_LIBADD) $(LIBS) mostlyclean-compile: -rm -f *.$(OBJEXT) distclean-compile: -rm -f *.tab.c @AMDEP_TRUE@@am__include@ @am__quote@./$(DEPDIR)/libfftw3@PREC_SUFFIX@_omp_la-api.Plo@am__quote@ @AMDEP_TRUE@@am__include@ @am__quote@./$(DEPDIR)/libfftw3@PREC_SUFFIX@_omp_la-conf.Plo@am__quote@ @AMDEP_TRUE@@am__include@ @am__quote@./$(DEPDIR)/libfftw3@PREC_SUFFIX@_omp_la-ct.Plo@am__quote@ @AMDEP_TRUE@@am__include@ @am__quote@./$(DEPDIR)/libfftw3@PREC_SUFFIX@_omp_la-dft-vrank-geq1.Plo@am__quote@ @AMDEP_TRUE@@am__include@ @am__quote@./$(DEPDIR)/libfftw3@PREC_SUFFIX@_omp_la-f77api.Plo@am__quote@ @AMDEP_TRUE@@am__include@ @am__quote@./$(DEPDIR)/libfftw3@PREC_SUFFIX@_omp_la-hc2hc.Plo@am__quote@ @AMDEP_TRUE@@am__include@ @am__quote@./$(DEPDIR)/libfftw3@PREC_SUFFIX@_omp_la-openmp.Plo@am__quote@ @AMDEP_TRUE@@am__include@ @am__quote@./$(DEPDIR)/libfftw3@PREC_SUFFIX@_omp_la-rdft-vrank-geq1.Plo@am__quote@ @AMDEP_TRUE@@am__include@ @am__quote@./$(DEPDIR)/libfftw3@PREC_SUFFIX@_omp_la-vrank-geq1-rdft2.Plo@am__quote@ @AMDEP_TRUE@@am__include@ @am__quote@./$(DEPDIR)/libfftw3@PREC_SUFFIX@_threads_la-api.Plo@am__quote@ @AMDEP_TRUE@@am__include@ @am__quote@./$(DEPDIR)/libfftw3@PREC_SUFFIX@_threads_la-conf.Plo@am__quote@ @AMDEP_TRUE@@am__include@ @am__quote@./$(DEPDIR)/libfftw3@PREC_SUFFIX@_threads_la-ct.Plo@am__quote@ @AMDEP_TRUE@@am__include@ @am__quote@./$(DEPDIR)/libfftw3@PREC_SUFFIX@_threads_la-dft-vrank-geq1.Plo@am__quote@ @AMDEP_TRUE@@am__include@ @am__quote@./$(DEPDIR)/libfftw3@PREC_SUFFIX@_threads_la-f77api.Plo@am__quote@ @AMDEP_TRUE@@am__include@ @am__quote@./$(DEPDIR)/libfftw3@PREC_SUFFIX@_threads_la-hc2hc.Plo@am__quote@ @AMDEP_TRUE@@am__include@ @am__quote@./$(DEPDIR)/libfftw3@PREC_SUFFIX@_threads_la-rdft-vrank-geq1.Plo@am__quote@ @AMDEP_TRUE@@am__include@ @am__quote@./$(DEPDIR)/libfftw3@PREC_SUFFIX@_threads_la-threads.Plo@am__quote@ @AMDEP_TRUE@@am__include@ @am__quote@./$(DEPDIR)/libfftw3@PREC_SUFFIX@_threads_la-vrank-geq1-rdft2.Plo@am__quote@ .c.o: @am__fastdepCC_TRUE@ $(COMPILE) -MT $@ -MD -MP -MF $(DEPDIR)/$*.Tpo -c -o $@ $< @am__fastdepCC_TRUE@ $(am__mv) $(DEPDIR)/$*.Tpo $(DEPDIR)/$*.Po @AMDEP_TRUE@@am__fastdepCC_FALSE@ source='$<' object='$@' libtool=no @AMDEPBACKSLASH@ @AMDEP_TRUE@@am__fastdepCC_FALSE@ DEPDIR=$(DEPDIR) $(CCDEPMODE) $(depcomp) @AMDEPBACKSLASH@ @am__fastdepCC_FALSE@ $(COMPILE) -c $< .c.obj: @am__fastdepCC_TRUE@ $(COMPILE) -MT $@ -MD -MP -MF $(DEPDIR)/$*.Tpo -c -o $@ `$(CYGPATH_W) '$<'` @am__fastdepCC_TRUE@ $(am__mv) $(DEPDIR)/$*.Tpo $(DEPDIR)/$*.Po @AMDEP_TRUE@@am__fastdepCC_FALSE@ source='$<' object='$@' libtool=no @AMDEPBACKSLASH@ @AMDEP_TRUE@@am__fastdepCC_FALSE@ DEPDIR=$(DEPDIR) $(CCDEPMODE) $(depcomp) @AMDEPBACKSLASH@ @am__fastdepCC_FALSE@ $(COMPILE) -c `$(CYGPATH_W) '$<'` .c.lo: @am__fastdepCC_TRUE@ $(LTCOMPILE) -MT $@ -MD -MP -MF $(DEPDIR)/$*.Tpo -c -o $@ $< @am__fastdepCC_TRUE@ $(am__mv) $(DEPDIR)/$*.Tpo $(DEPDIR)/$*.Plo @AMDEP_TRUE@@am__fastdepCC_FALSE@ source='$<' object='$@' libtool=yes @AMDEPBACKSLASH@ @AMDEP_TRUE@@am__fastdepCC_FALSE@ DEPDIR=$(DEPDIR) $(CCDEPMODE) $(depcomp) @AMDEPBACKSLASH@ @am__fastdepCC_FALSE@ $(LTCOMPILE) -c -o $@ $< libfftw3@PREC_SUFFIX@_omp_la-api.lo: api.c @am__fastdepCC_TRUE@ $(LIBTOOL) --tag=CC $(AM_LIBTOOLFLAGS) $(LIBTOOLFLAGS) --mode=compile $(CC) $(DEFS) $(DEFAULT_INCLUDES) $(INCLUDES) $(AM_CPPFLAGS) $(CPPFLAGS) $(libfftw3@PREC_SUFFIX@_omp_la_CFLAGS) $(CFLAGS) -MT libfftw3@PREC_SUFFIX@_omp_la-api.lo -MD -MP -MF $(DEPDIR)/libfftw3@PREC_SUFFIX@_omp_la-api.Tpo -c -o libfftw3@PREC_SUFFIX@_omp_la-api.lo `test -f 'api.c' || echo '$(srcdir)/'`api.c @am__fastdepCC_TRUE@ $(am__mv) $(DEPDIR)/libfftw3@PREC_SUFFIX@_omp_la-api.Tpo $(DEPDIR)/libfftw3@PREC_SUFFIX@_omp_la-api.Plo @AMDEP_TRUE@@am__fastdepCC_FALSE@ source='api.c' object='libfftw3@PREC_SUFFIX@_omp_la-api.lo' libtool=yes @AMDEPBACKSLASH@ @AMDEP_TRUE@@am__fastdepCC_FALSE@ DEPDIR=$(DEPDIR) $(CCDEPMODE) $(depcomp) @AMDEPBACKSLASH@ @am__fastdepCC_FALSE@ $(LIBTOOL) --tag=CC $(AM_LIBTOOLFLAGS) $(LIBTOOLFLAGS) --mode=compile $(CC) $(DEFS) $(DEFAULT_INCLUDES) $(INCLUDES) $(AM_CPPFLAGS) $(CPPFLAGS) $(libfftw3@PREC_SUFFIX@_omp_la_CFLAGS) $(CFLAGS) -c -o libfftw3@PREC_SUFFIX@_omp_la-api.lo `test -f 'api.c' || echo '$(srcdir)/'`api.c libfftw3@PREC_SUFFIX@_omp_la-conf.lo: conf.c @am__fastdepCC_TRUE@ $(LIBTOOL) --tag=CC $(AM_LIBTOOLFLAGS) $(LIBTOOLFLAGS) --mode=compile $(CC) $(DEFS) $(DEFAULT_INCLUDES) $(INCLUDES) $(AM_CPPFLAGS) $(CPPFLAGS) $(libfftw3@PREC_SUFFIX@_omp_la_CFLAGS) $(CFLAGS) -MT libfftw3@PREC_SUFFIX@_omp_la-conf.lo -MD -MP -MF $(DEPDIR)/libfftw3@PREC_SUFFIX@_omp_la-conf.Tpo -c -o libfftw3@PREC_SUFFIX@_omp_la-conf.lo `test -f 'conf.c' || echo '$(srcdir)/'`conf.c @am__fastdepCC_TRUE@ $(am__mv) $(DEPDIR)/libfftw3@PREC_SUFFIX@_omp_la-conf.Tpo $(DEPDIR)/libfftw3@PREC_SUFFIX@_omp_la-conf.Plo @AMDEP_TRUE@@am__fastdepCC_FALSE@ source='conf.c' object='libfftw3@PREC_SUFFIX@_omp_la-conf.lo' libtool=yes @AMDEPBACKSLASH@ @AMDEP_TRUE@@am__fastdepCC_FALSE@ DEPDIR=$(DEPDIR) $(CCDEPMODE) $(depcomp) @AMDEPBACKSLASH@ @am__fastdepCC_FALSE@ $(LIBTOOL) --tag=CC $(AM_LIBTOOLFLAGS) $(LIBTOOLFLAGS) --mode=compile $(CC) $(DEFS) $(DEFAULT_INCLUDES) $(INCLUDES) $(AM_CPPFLAGS) $(CPPFLAGS) $(libfftw3@PREC_SUFFIX@_omp_la_CFLAGS) $(CFLAGS) -c -o libfftw3@PREC_SUFFIX@_omp_la-conf.lo `test -f 'conf.c' || echo '$(srcdir)/'`conf.c libfftw3@PREC_SUFFIX@_omp_la-openmp.lo: openmp.c @am__fastdepCC_TRUE@ $(LIBTOOL) --tag=CC $(AM_LIBTOOLFLAGS) $(LIBTOOLFLAGS) --mode=compile $(CC) $(DEFS) $(DEFAULT_INCLUDES) $(INCLUDES) $(AM_CPPFLAGS) $(CPPFLAGS) $(libfftw3@PREC_SUFFIX@_omp_la_CFLAGS) $(CFLAGS) -MT libfftw3@PREC_SUFFIX@_omp_la-openmp.lo -MD -MP -MF $(DEPDIR)/libfftw3@PREC_SUFFIX@_omp_la-openmp.Tpo -c -o libfftw3@PREC_SUFFIX@_omp_la-openmp.lo `test -f 'openmp.c' || echo '$(srcdir)/'`openmp.c @am__fastdepCC_TRUE@ $(am__mv) $(DEPDIR)/libfftw3@PREC_SUFFIX@_omp_la-openmp.Tpo $(DEPDIR)/libfftw3@PREC_SUFFIX@_omp_la-openmp.Plo @AMDEP_TRUE@@am__fastdepCC_FALSE@ source='openmp.c' object='libfftw3@PREC_SUFFIX@_omp_la-openmp.lo' libtool=yes @AMDEPBACKSLASH@ @AMDEP_TRUE@@am__fastdepCC_FALSE@ DEPDIR=$(DEPDIR) $(CCDEPMODE) $(depcomp) @AMDEPBACKSLASH@ @am__fastdepCC_FALSE@ $(LIBTOOL) --tag=CC $(AM_LIBTOOLFLAGS) $(LIBTOOLFLAGS) --mode=compile $(CC) $(DEFS) $(DEFAULT_INCLUDES) $(INCLUDES) $(AM_CPPFLAGS) $(CPPFLAGS) $(libfftw3@PREC_SUFFIX@_omp_la_CFLAGS) $(CFLAGS) -c -o libfftw3@PREC_SUFFIX@_omp_la-openmp.lo `test -f 'openmp.c' || echo '$(srcdir)/'`openmp.c libfftw3@PREC_SUFFIX@_omp_la-dft-vrank-geq1.lo: dft-vrank-geq1.c @am__fastdepCC_TRUE@ $(LIBTOOL) --tag=CC $(AM_LIBTOOLFLAGS) $(LIBTOOLFLAGS) --mode=compile $(CC) $(DEFS) $(DEFAULT_INCLUDES) $(INCLUDES) $(AM_CPPFLAGS) $(CPPFLAGS) $(libfftw3@PREC_SUFFIX@_omp_la_CFLAGS) $(CFLAGS) -MT libfftw3@PREC_SUFFIX@_omp_la-dft-vrank-geq1.lo -MD -MP -MF $(DEPDIR)/libfftw3@PREC_SUFFIX@_omp_la-dft-vrank-geq1.Tpo -c -o libfftw3@PREC_SUFFIX@_omp_la-dft-vrank-geq1.lo `test -f 'dft-vrank-geq1.c' || echo '$(srcdir)/'`dft-vrank-geq1.c @am__fastdepCC_TRUE@ $(am__mv) $(DEPDIR)/libfftw3@PREC_SUFFIX@_omp_la-dft-vrank-geq1.Tpo $(DEPDIR)/libfftw3@PREC_SUFFIX@_omp_la-dft-vrank-geq1.Plo @AMDEP_TRUE@@am__fastdepCC_FALSE@ source='dft-vrank-geq1.c' object='libfftw3@PREC_SUFFIX@_omp_la-dft-vrank-geq1.lo' libtool=yes @AMDEPBACKSLASH@ @AMDEP_TRUE@@am__fastdepCC_FALSE@ DEPDIR=$(DEPDIR) $(CCDEPMODE) $(depcomp) @AMDEPBACKSLASH@ @am__fastdepCC_FALSE@ $(LIBTOOL) --tag=CC $(AM_LIBTOOLFLAGS) $(LIBTOOLFLAGS) --mode=compile $(CC) $(DEFS) $(DEFAULT_INCLUDES) $(INCLUDES) $(AM_CPPFLAGS) $(CPPFLAGS) $(libfftw3@PREC_SUFFIX@_omp_la_CFLAGS) $(CFLAGS) -c -o libfftw3@PREC_SUFFIX@_omp_la-dft-vrank-geq1.lo `test -f 'dft-vrank-geq1.c' || echo '$(srcdir)/'`dft-vrank-geq1.c libfftw3@PREC_SUFFIX@_omp_la-ct.lo: ct.c @am__fastdepCC_TRUE@ $(LIBTOOL) --tag=CC $(AM_LIBTOOLFLAGS) $(LIBTOOLFLAGS) --mode=compile $(CC) $(DEFS) $(DEFAULT_INCLUDES) $(INCLUDES) $(AM_CPPFLAGS) $(CPPFLAGS) $(libfftw3@PREC_SUFFIX@_omp_la_CFLAGS) $(CFLAGS) -MT libfftw3@PREC_SUFFIX@_omp_la-ct.lo -MD -MP -MF $(DEPDIR)/libfftw3@PREC_SUFFIX@_omp_la-ct.Tpo -c -o libfftw3@PREC_SUFFIX@_omp_la-ct.lo `test -f 'ct.c' || echo '$(srcdir)/'`ct.c @am__fastdepCC_TRUE@ $(am__mv) $(DEPDIR)/libfftw3@PREC_SUFFIX@_omp_la-ct.Tpo $(DEPDIR)/libfftw3@PREC_SUFFIX@_omp_la-ct.Plo @AMDEP_TRUE@@am__fastdepCC_FALSE@ source='ct.c' object='libfftw3@PREC_SUFFIX@_omp_la-ct.lo' libtool=yes @AMDEPBACKSLASH@ @AMDEP_TRUE@@am__fastdepCC_FALSE@ DEPDIR=$(DEPDIR) $(CCDEPMODE) $(depcomp) @AMDEPBACKSLASH@ @am__fastdepCC_FALSE@ $(LIBTOOL) --tag=CC $(AM_LIBTOOLFLAGS) $(LIBTOOLFLAGS) --mode=compile $(CC) $(DEFS) $(DEFAULT_INCLUDES) $(INCLUDES) $(AM_CPPFLAGS) $(CPPFLAGS) $(libfftw3@PREC_SUFFIX@_omp_la_CFLAGS) $(CFLAGS) -c -o libfftw3@PREC_SUFFIX@_omp_la-ct.lo `test -f 'ct.c' || echo '$(srcdir)/'`ct.c libfftw3@PREC_SUFFIX@_omp_la-rdft-vrank-geq1.lo: rdft-vrank-geq1.c @am__fastdepCC_TRUE@ $(LIBTOOL) --tag=CC $(AM_LIBTOOLFLAGS) $(LIBTOOLFLAGS) --mode=compile $(CC) $(DEFS) $(DEFAULT_INCLUDES) $(INCLUDES) $(AM_CPPFLAGS) $(CPPFLAGS) $(libfftw3@PREC_SUFFIX@_omp_la_CFLAGS) $(CFLAGS) -MT libfftw3@PREC_SUFFIX@_omp_la-rdft-vrank-geq1.lo -MD -MP -MF $(DEPDIR)/libfftw3@PREC_SUFFIX@_omp_la-rdft-vrank-geq1.Tpo -c -o libfftw3@PREC_SUFFIX@_omp_la-rdft-vrank-geq1.lo `test -f 'rdft-vrank-geq1.c' || echo '$(srcdir)/'`rdft-vrank-geq1.c @am__fastdepCC_TRUE@ $(am__mv) $(DEPDIR)/libfftw3@PREC_SUFFIX@_omp_la-rdft-vrank-geq1.Tpo $(DEPDIR)/libfftw3@PREC_SUFFIX@_omp_la-rdft-vrank-geq1.Plo @AMDEP_TRUE@@am__fastdepCC_FALSE@ source='rdft-vrank-geq1.c' object='libfftw3@PREC_SUFFIX@_omp_la-rdft-vrank-geq1.lo' libtool=yes @AMDEPBACKSLASH@ @AMDEP_TRUE@@am__fastdepCC_FALSE@ DEPDIR=$(DEPDIR) $(CCDEPMODE) $(depcomp) @AMDEPBACKSLASH@ @am__fastdepCC_FALSE@ $(LIBTOOL) --tag=CC $(AM_LIBTOOLFLAGS) $(LIBTOOLFLAGS) --mode=compile $(CC) $(DEFS) $(DEFAULT_INCLUDES) $(INCLUDES) $(AM_CPPFLAGS) $(CPPFLAGS) $(libfftw3@PREC_SUFFIX@_omp_la_CFLAGS) $(CFLAGS) -c -o libfftw3@PREC_SUFFIX@_omp_la-rdft-vrank-geq1.lo `test -f 'rdft-vrank-geq1.c' || echo '$(srcdir)/'`rdft-vrank-geq1.c libfftw3@PREC_SUFFIX@_omp_la-hc2hc.lo: hc2hc.c @am__fastdepCC_TRUE@ $(LIBTOOL) --tag=CC $(AM_LIBTOOLFLAGS) $(LIBTOOLFLAGS) --mode=compile $(CC) $(DEFS) $(DEFAULT_INCLUDES) $(INCLUDES) $(AM_CPPFLAGS) $(CPPFLAGS) $(libfftw3@PREC_SUFFIX@_omp_la_CFLAGS) $(CFLAGS) -MT libfftw3@PREC_SUFFIX@_omp_la-hc2hc.lo -MD -MP -MF $(DEPDIR)/libfftw3@PREC_SUFFIX@_omp_la-hc2hc.Tpo -c -o libfftw3@PREC_SUFFIX@_omp_la-hc2hc.lo `test -f 'hc2hc.c' || echo '$(srcdir)/'`hc2hc.c @am__fastdepCC_TRUE@ $(am__mv) $(DEPDIR)/libfftw3@PREC_SUFFIX@_omp_la-hc2hc.Tpo $(DEPDIR)/libfftw3@PREC_SUFFIX@_omp_la-hc2hc.Plo @AMDEP_TRUE@@am__fastdepCC_FALSE@ source='hc2hc.c' object='libfftw3@PREC_SUFFIX@_omp_la-hc2hc.lo' libtool=yes @AMDEPBACKSLASH@ @AMDEP_TRUE@@am__fastdepCC_FALSE@ DEPDIR=$(DEPDIR) $(CCDEPMODE) $(depcomp) @AMDEPBACKSLASH@ @am__fastdepCC_FALSE@ $(LIBTOOL) --tag=CC $(AM_LIBTOOLFLAGS) $(LIBTOOLFLAGS) --mode=compile $(CC) $(DEFS) $(DEFAULT_INCLUDES) $(INCLUDES) $(AM_CPPFLAGS) $(CPPFLAGS) $(libfftw3@PREC_SUFFIX@_omp_la_CFLAGS) $(CFLAGS) -c -o libfftw3@PREC_SUFFIX@_omp_la-hc2hc.lo `test -f 'hc2hc.c' || echo '$(srcdir)/'`hc2hc.c libfftw3@PREC_SUFFIX@_omp_la-vrank-geq1-rdft2.lo: vrank-geq1-rdft2.c @am__fastdepCC_TRUE@ $(LIBTOOL) --tag=CC $(AM_LIBTOOLFLAGS) $(LIBTOOLFLAGS) --mode=compile $(CC) $(DEFS) $(DEFAULT_INCLUDES) $(INCLUDES) $(AM_CPPFLAGS) $(CPPFLAGS) $(libfftw3@PREC_SUFFIX@_omp_la_CFLAGS) $(CFLAGS) -MT libfftw3@PREC_SUFFIX@_omp_la-vrank-geq1-rdft2.lo -MD -MP -MF $(DEPDIR)/libfftw3@PREC_SUFFIX@_omp_la-vrank-geq1-rdft2.Tpo -c -o libfftw3@PREC_SUFFIX@_omp_la-vrank-geq1-rdft2.lo `test -f 'vrank-geq1-rdft2.c' || echo '$(srcdir)/'`vrank-geq1-rdft2.c @am__fastdepCC_TRUE@ $(am__mv) $(DEPDIR)/libfftw3@PREC_SUFFIX@_omp_la-vrank-geq1-rdft2.Tpo $(DEPDIR)/libfftw3@PREC_SUFFIX@_omp_la-vrank-geq1-rdft2.Plo @AMDEP_TRUE@@am__fastdepCC_FALSE@ source='vrank-geq1-rdft2.c' object='libfftw3@PREC_SUFFIX@_omp_la-vrank-geq1-rdft2.lo' libtool=yes @AMDEPBACKSLASH@ @AMDEP_TRUE@@am__fastdepCC_FALSE@ DEPDIR=$(DEPDIR) $(CCDEPMODE) $(depcomp) @AMDEPBACKSLASH@ @am__fastdepCC_FALSE@ $(LIBTOOL) --tag=CC $(AM_LIBTOOLFLAGS) $(LIBTOOLFLAGS) --mode=compile $(CC) $(DEFS) $(DEFAULT_INCLUDES) $(INCLUDES) $(AM_CPPFLAGS) $(CPPFLAGS) $(libfftw3@PREC_SUFFIX@_omp_la_CFLAGS) $(CFLAGS) -c -o libfftw3@PREC_SUFFIX@_omp_la-vrank-geq1-rdft2.lo `test -f 'vrank-geq1-rdft2.c' || echo '$(srcdir)/'`vrank-geq1-rdft2.c libfftw3@PREC_SUFFIX@_omp_la-f77api.lo: f77api.c @am__fastdepCC_TRUE@ $(LIBTOOL) --tag=CC $(AM_LIBTOOLFLAGS) $(LIBTOOLFLAGS) --mode=compile $(CC) $(DEFS) $(DEFAULT_INCLUDES) $(INCLUDES) $(AM_CPPFLAGS) $(CPPFLAGS) $(libfftw3@PREC_SUFFIX@_omp_la_CFLAGS) $(CFLAGS) -MT libfftw3@PREC_SUFFIX@_omp_la-f77api.lo -MD -MP -MF $(DEPDIR)/libfftw3@PREC_SUFFIX@_omp_la-f77api.Tpo -c -o libfftw3@PREC_SUFFIX@_omp_la-f77api.lo `test -f 'f77api.c' || echo '$(srcdir)/'`f77api.c @am__fastdepCC_TRUE@ $(am__mv) $(DEPDIR)/libfftw3@PREC_SUFFIX@_omp_la-f77api.Tpo $(DEPDIR)/libfftw3@PREC_SUFFIX@_omp_la-f77api.Plo @AMDEP_TRUE@@am__fastdepCC_FALSE@ source='f77api.c' object='libfftw3@PREC_SUFFIX@_omp_la-f77api.lo' libtool=yes @AMDEPBACKSLASH@ @AMDEP_TRUE@@am__fastdepCC_FALSE@ DEPDIR=$(DEPDIR) $(CCDEPMODE) $(depcomp) @AMDEPBACKSLASH@ @am__fastdepCC_FALSE@ $(LIBTOOL) --tag=CC $(AM_LIBTOOLFLAGS) $(LIBTOOLFLAGS) --mode=compile $(CC) $(DEFS) $(DEFAULT_INCLUDES) $(INCLUDES) $(AM_CPPFLAGS) $(CPPFLAGS) $(libfftw3@PREC_SUFFIX@_omp_la_CFLAGS) $(CFLAGS) -c -o libfftw3@PREC_SUFFIX@_omp_la-f77api.lo `test -f 'f77api.c' || echo '$(srcdir)/'`f77api.c libfftw3@PREC_SUFFIX@_threads_la-api.lo: api.c @am__fastdepCC_TRUE@ $(LIBTOOL) --tag=CC $(AM_LIBTOOLFLAGS) $(LIBTOOLFLAGS) --mode=compile $(CC) $(DEFS) $(DEFAULT_INCLUDES) $(INCLUDES) $(AM_CPPFLAGS) $(CPPFLAGS) $(libfftw3@PREC_SUFFIX@_threads_la_CFLAGS) $(CFLAGS) -MT libfftw3@PREC_SUFFIX@_threads_la-api.lo -MD -MP -MF $(DEPDIR)/libfftw3@PREC_SUFFIX@_threads_la-api.Tpo -c -o libfftw3@PREC_SUFFIX@_threads_la-api.lo `test -f 'api.c' || echo '$(srcdir)/'`api.c @am__fastdepCC_TRUE@ $(am__mv) $(DEPDIR)/libfftw3@PREC_SUFFIX@_threads_la-api.Tpo $(DEPDIR)/libfftw3@PREC_SUFFIX@_threads_la-api.Plo @AMDEP_TRUE@@am__fastdepCC_FALSE@ source='api.c' object='libfftw3@PREC_SUFFIX@_threads_la-api.lo' libtool=yes @AMDEPBACKSLASH@ @AMDEP_TRUE@@am__fastdepCC_FALSE@ DEPDIR=$(DEPDIR) $(CCDEPMODE) $(depcomp) @AMDEPBACKSLASH@ @am__fastdepCC_FALSE@ $(LIBTOOL) --tag=CC $(AM_LIBTOOLFLAGS) $(LIBTOOLFLAGS) --mode=compile $(CC) $(DEFS) $(DEFAULT_INCLUDES) $(INCLUDES) $(AM_CPPFLAGS) $(CPPFLAGS) $(libfftw3@PREC_SUFFIX@_threads_la_CFLAGS) $(CFLAGS) -c -o libfftw3@PREC_SUFFIX@_threads_la-api.lo `test -f 'api.c' || echo '$(srcdir)/'`api.c libfftw3@PREC_SUFFIX@_threads_la-conf.lo: conf.c @am__fastdepCC_TRUE@ $(LIBTOOL) --tag=CC $(AM_LIBTOOLFLAGS) $(LIBTOOLFLAGS) --mode=compile $(CC) $(DEFS) $(DEFAULT_INCLUDES) $(INCLUDES) $(AM_CPPFLAGS) $(CPPFLAGS) $(libfftw3@PREC_SUFFIX@_threads_la_CFLAGS) $(CFLAGS) -MT libfftw3@PREC_SUFFIX@_threads_la-conf.lo -MD -MP -MF $(DEPDIR)/libfftw3@PREC_SUFFIX@_threads_la-conf.Tpo -c -o libfftw3@PREC_SUFFIX@_threads_la-conf.lo `test -f 'conf.c' || echo '$(srcdir)/'`conf.c @am__fastdepCC_TRUE@ $(am__mv) $(DEPDIR)/libfftw3@PREC_SUFFIX@_threads_la-conf.Tpo $(DEPDIR)/libfftw3@PREC_SUFFIX@_threads_la-conf.Plo @AMDEP_TRUE@@am__fastdepCC_FALSE@ source='conf.c' object='libfftw3@PREC_SUFFIX@_threads_la-conf.lo' libtool=yes @AMDEPBACKSLASH@ @AMDEP_TRUE@@am__fastdepCC_FALSE@ DEPDIR=$(DEPDIR) $(CCDEPMODE) $(depcomp) @AMDEPBACKSLASH@ @am__fastdepCC_FALSE@ $(LIBTOOL) --tag=CC $(AM_LIBTOOLFLAGS) $(LIBTOOLFLAGS) --mode=compile $(CC) $(DEFS) $(DEFAULT_INCLUDES) $(INCLUDES) $(AM_CPPFLAGS) $(CPPFLAGS) $(libfftw3@PREC_SUFFIX@_threads_la_CFLAGS) $(CFLAGS) -c -o libfftw3@PREC_SUFFIX@_threads_la-conf.lo `test -f 'conf.c' || echo '$(srcdir)/'`conf.c libfftw3@PREC_SUFFIX@_threads_la-threads.lo: threads.c @am__fastdepCC_TRUE@ $(LIBTOOL) --tag=CC $(AM_LIBTOOLFLAGS) $(LIBTOOLFLAGS) --mode=compile $(CC) $(DEFS) $(DEFAULT_INCLUDES) $(INCLUDES) $(AM_CPPFLAGS) $(CPPFLAGS) $(libfftw3@PREC_SUFFIX@_threads_la_CFLAGS) $(CFLAGS) -MT libfftw3@PREC_SUFFIX@_threads_la-threads.lo -MD -MP -MF $(DEPDIR)/libfftw3@PREC_SUFFIX@_threads_la-threads.Tpo -c -o libfftw3@PREC_SUFFIX@_threads_la-threads.lo `test -f 'threads.c' || echo '$(srcdir)/'`threads.c @am__fastdepCC_TRUE@ $(am__mv) $(DEPDIR)/libfftw3@PREC_SUFFIX@_threads_la-threads.Tpo $(DEPDIR)/libfftw3@PREC_SUFFIX@_threads_la-threads.Plo @AMDEP_TRUE@@am__fastdepCC_FALSE@ source='threads.c' object='libfftw3@PREC_SUFFIX@_threads_la-threads.lo' libtool=yes @AMDEPBACKSLASH@ @AMDEP_TRUE@@am__fastdepCC_FALSE@ DEPDIR=$(DEPDIR) $(CCDEPMODE) $(depcomp) @AMDEPBACKSLASH@ @am__fastdepCC_FALSE@ $(LIBTOOL) --tag=CC $(AM_LIBTOOLFLAGS) $(LIBTOOLFLAGS) --mode=compile $(CC) $(DEFS) $(DEFAULT_INCLUDES) $(INCLUDES) $(AM_CPPFLAGS) $(CPPFLAGS) $(libfftw3@PREC_SUFFIX@_threads_la_CFLAGS) $(CFLAGS) -c -o libfftw3@PREC_SUFFIX@_threads_la-threads.lo `test -f 'threads.c' || echo '$(srcdir)/'`threads.c libfftw3@PREC_SUFFIX@_threads_la-dft-vrank-geq1.lo: dft-vrank-geq1.c @am__fastdepCC_TRUE@ $(LIBTOOL) --tag=CC $(AM_LIBTOOLFLAGS) $(LIBTOOLFLAGS) --mode=compile $(CC) $(DEFS) $(DEFAULT_INCLUDES) $(INCLUDES) $(AM_CPPFLAGS) $(CPPFLAGS) $(libfftw3@PREC_SUFFIX@_threads_la_CFLAGS) $(CFLAGS) -MT libfftw3@PREC_SUFFIX@_threads_la-dft-vrank-geq1.lo -MD -MP -MF $(DEPDIR)/libfftw3@PREC_SUFFIX@_threads_la-dft-vrank-geq1.Tpo -c -o libfftw3@PREC_SUFFIX@_threads_la-dft-vrank-geq1.lo `test -f 'dft-vrank-geq1.c' || echo '$(srcdir)/'`dft-vrank-geq1.c @am__fastdepCC_TRUE@ $(am__mv) $(DEPDIR)/libfftw3@PREC_SUFFIX@_threads_la-dft-vrank-geq1.Tpo $(DEPDIR)/libfftw3@PREC_SUFFIX@_threads_la-dft-vrank-geq1.Plo @AMDEP_TRUE@@am__fastdepCC_FALSE@ source='dft-vrank-geq1.c' object='libfftw3@PREC_SUFFIX@_threads_la-dft-vrank-geq1.lo' libtool=yes @AMDEPBACKSLASH@ @AMDEP_TRUE@@am__fastdepCC_FALSE@ DEPDIR=$(DEPDIR) $(CCDEPMODE) $(depcomp) @AMDEPBACKSLASH@ @am__fastdepCC_FALSE@ $(LIBTOOL) --tag=CC $(AM_LIBTOOLFLAGS) $(LIBTOOLFLAGS) --mode=compile $(CC) $(DEFS) $(DEFAULT_INCLUDES) $(INCLUDES) $(AM_CPPFLAGS) $(CPPFLAGS) $(libfftw3@PREC_SUFFIX@_threads_la_CFLAGS) $(CFLAGS) -c -o libfftw3@PREC_SUFFIX@_threads_la-dft-vrank-geq1.lo `test -f 'dft-vrank-geq1.c' || echo '$(srcdir)/'`dft-vrank-geq1.c libfftw3@PREC_SUFFIX@_threads_la-ct.lo: ct.c @am__fastdepCC_TRUE@ $(LIBTOOL) --tag=CC $(AM_LIBTOOLFLAGS) $(LIBTOOLFLAGS) --mode=compile $(CC) $(DEFS) $(DEFAULT_INCLUDES) $(INCLUDES) $(AM_CPPFLAGS) $(CPPFLAGS) $(libfftw3@PREC_SUFFIX@_threads_la_CFLAGS) $(CFLAGS) -MT libfftw3@PREC_SUFFIX@_threads_la-ct.lo -MD -MP -MF $(DEPDIR)/libfftw3@PREC_SUFFIX@_threads_la-ct.Tpo -c -o libfftw3@PREC_SUFFIX@_threads_la-ct.lo `test -f 'ct.c' || echo '$(srcdir)/'`ct.c @am__fastdepCC_TRUE@ $(am__mv) $(DEPDIR)/libfftw3@PREC_SUFFIX@_threads_la-ct.Tpo $(DEPDIR)/libfftw3@PREC_SUFFIX@_threads_la-ct.Plo @AMDEP_TRUE@@am__fastdepCC_FALSE@ source='ct.c' object='libfftw3@PREC_SUFFIX@_threads_la-ct.lo' libtool=yes @AMDEPBACKSLASH@ @AMDEP_TRUE@@am__fastdepCC_FALSE@ DEPDIR=$(DEPDIR) $(CCDEPMODE) $(depcomp) @AMDEPBACKSLASH@ @am__fastdepCC_FALSE@ $(LIBTOOL) --tag=CC $(AM_LIBTOOLFLAGS) $(LIBTOOLFLAGS) --mode=compile $(CC) $(DEFS) $(DEFAULT_INCLUDES) $(INCLUDES) $(AM_CPPFLAGS) $(CPPFLAGS) $(libfftw3@PREC_SUFFIX@_threads_la_CFLAGS) $(CFLAGS) -c -o libfftw3@PREC_SUFFIX@_threads_la-ct.lo `test -f 'ct.c' || echo '$(srcdir)/'`ct.c libfftw3@PREC_SUFFIX@_threads_la-rdft-vrank-geq1.lo: rdft-vrank-geq1.c @am__fastdepCC_TRUE@ $(LIBTOOL) --tag=CC $(AM_LIBTOOLFLAGS) $(LIBTOOLFLAGS) --mode=compile $(CC) $(DEFS) $(DEFAULT_INCLUDES) $(INCLUDES) $(AM_CPPFLAGS) $(CPPFLAGS) $(libfftw3@PREC_SUFFIX@_threads_la_CFLAGS) $(CFLAGS) -MT libfftw3@PREC_SUFFIX@_threads_la-rdft-vrank-geq1.lo -MD -MP -MF $(DEPDIR)/libfftw3@PREC_SUFFIX@_threads_la-rdft-vrank-geq1.Tpo -c -o libfftw3@PREC_SUFFIX@_threads_la-rdft-vrank-geq1.lo `test -f 'rdft-vrank-geq1.c' || echo '$(srcdir)/'`rdft-vrank-geq1.c @am__fastdepCC_TRUE@ $(am__mv) $(DEPDIR)/libfftw3@PREC_SUFFIX@_threads_la-rdft-vrank-geq1.Tpo $(DEPDIR)/libfftw3@PREC_SUFFIX@_threads_la-rdft-vrank-geq1.Plo @AMDEP_TRUE@@am__fastdepCC_FALSE@ source='rdft-vrank-geq1.c' object='libfftw3@PREC_SUFFIX@_threads_la-rdft-vrank-geq1.lo' libtool=yes @AMDEPBACKSLASH@ @AMDEP_TRUE@@am__fastdepCC_FALSE@ DEPDIR=$(DEPDIR) $(CCDEPMODE) $(depcomp) @AMDEPBACKSLASH@ @am__fastdepCC_FALSE@ $(LIBTOOL) --tag=CC $(AM_LIBTOOLFLAGS) $(LIBTOOLFLAGS) --mode=compile $(CC) $(DEFS) $(DEFAULT_INCLUDES) $(INCLUDES) $(AM_CPPFLAGS) $(CPPFLAGS) $(libfftw3@PREC_SUFFIX@_threads_la_CFLAGS) $(CFLAGS) -c -o libfftw3@PREC_SUFFIX@_threads_la-rdft-vrank-geq1.lo `test -f 'rdft-vrank-geq1.c' || echo '$(srcdir)/'`rdft-vrank-geq1.c libfftw3@PREC_SUFFIX@_threads_la-hc2hc.lo: hc2hc.c @am__fastdepCC_TRUE@ $(LIBTOOL) --tag=CC $(AM_LIBTOOLFLAGS) $(LIBTOOLFLAGS) --mode=compile $(CC) $(DEFS) $(DEFAULT_INCLUDES) $(INCLUDES) $(AM_CPPFLAGS) $(CPPFLAGS) $(libfftw3@PREC_SUFFIX@_threads_la_CFLAGS) $(CFLAGS) -MT libfftw3@PREC_SUFFIX@_threads_la-hc2hc.lo -MD -MP -MF $(DEPDIR)/libfftw3@PREC_SUFFIX@_threads_la-hc2hc.Tpo -c -o libfftw3@PREC_SUFFIX@_threads_la-hc2hc.lo `test -f 'hc2hc.c' || echo '$(srcdir)/'`hc2hc.c @am__fastdepCC_TRUE@ $(am__mv) $(DEPDIR)/libfftw3@PREC_SUFFIX@_threads_la-hc2hc.Tpo $(DEPDIR)/libfftw3@PREC_SUFFIX@_threads_la-hc2hc.Plo @AMDEP_TRUE@@am__fastdepCC_FALSE@ source='hc2hc.c' object='libfftw3@PREC_SUFFIX@_threads_la-hc2hc.lo' libtool=yes @AMDEPBACKSLASH@ @AMDEP_TRUE@@am__fastdepCC_FALSE@ DEPDIR=$(DEPDIR) $(CCDEPMODE) $(depcomp) @AMDEPBACKSLASH@ @am__fastdepCC_FALSE@ $(LIBTOOL) --tag=CC $(AM_LIBTOOLFLAGS) $(LIBTOOLFLAGS) --mode=compile $(CC) $(DEFS) $(DEFAULT_INCLUDES) $(INCLUDES) $(AM_CPPFLAGS) $(CPPFLAGS) $(libfftw3@PREC_SUFFIX@_threads_la_CFLAGS) $(CFLAGS) -c -o libfftw3@PREC_SUFFIX@_threads_la-hc2hc.lo `test -f 'hc2hc.c' || echo '$(srcdir)/'`hc2hc.c libfftw3@PREC_SUFFIX@_threads_la-vrank-geq1-rdft2.lo: vrank-geq1-rdft2.c @am__fastdepCC_TRUE@ $(LIBTOOL) --tag=CC $(AM_LIBTOOLFLAGS) $(LIBTOOLFLAGS) --mode=compile $(CC) $(DEFS) $(DEFAULT_INCLUDES) $(INCLUDES) $(AM_CPPFLAGS) $(CPPFLAGS) $(libfftw3@PREC_SUFFIX@_threads_la_CFLAGS) $(CFLAGS) -MT libfftw3@PREC_SUFFIX@_threads_la-vrank-geq1-rdft2.lo -MD -MP -MF $(DEPDIR)/libfftw3@PREC_SUFFIX@_threads_la-vrank-geq1-rdft2.Tpo -c -o libfftw3@PREC_SUFFIX@_threads_la-vrank-geq1-rdft2.lo `test -f 'vrank-geq1-rdft2.c' || echo '$(srcdir)/'`vrank-geq1-rdft2.c @am__fastdepCC_TRUE@ $(am__mv) $(DEPDIR)/libfftw3@PREC_SUFFIX@_threads_la-vrank-geq1-rdft2.Tpo $(DEPDIR)/libfftw3@PREC_SUFFIX@_threads_la-vrank-geq1-rdft2.Plo @AMDEP_TRUE@@am__fastdepCC_FALSE@ source='vrank-geq1-rdft2.c' object='libfftw3@PREC_SUFFIX@_threads_la-vrank-geq1-rdft2.lo' libtool=yes @AMDEPBACKSLASH@ @AMDEP_TRUE@@am__fastdepCC_FALSE@ DEPDIR=$(DEPDIR) $(CCDEPMODE) $(depcomp) @AMDEPBACKSLASH@ @am__fastdepCC_FALSE@ $(LIBTOOL) --tag=CC $(AM_LIBTOOLFLAGS) $(LIBTOOLFLAGS) --mode=compile $(CC) $(DEFS) $(DEFAULT_INCLUDES) $(INCLUDES) $(AM_CPPFLAGS) $(CPPFLAGS) $(libfftw3@PREC_SUFFIX@_threads_la_CFLAGS) $(CFLAGS) -c -o libfftw3@PREC_SUFFIX@_threads_la-vrank-geq1-rdft2.lo `test -f 'vrank-geq1-rdft2.c' || echo '$(srcdir)/'`vrank-geq1-rdft2.c libfftw3@PREC_SUFFIX@_threads_la-f77api.lo: f77api.c @am__fastdepCC_TRUE@ $(LIBTOOL) --tag=CC $(AM_LIBTOOLFLAGS) $(LIBTOOLFLAGS) --mode=compile $(CC) $(DEFS) $(DEFAULT_INCLUDES) $(INCLUDES) $(AM_CPPFLAGS) $(CPPFLAGS) $(libfftw3@PREC_SUFFIX@_threads_la_CFLAGS) $(CFLAGS) -MT libfftw3@PREC_SUFFIX@_threads_la-f77api.lo -MD -MP -MF $(DEPDIR)/libfftw3@PREC_SUFFIX@_threads_la-f77api.Tpo -c -o libfftw3@PREC_SUFFIX@_threads_la-f77api.lo `test -f 'f77api.c' || echo '$(srcdir)/'`f77api.c @am__fastdepCC_TRUE@ $(am__mv) $(DEPDIR)/libfftw3@PREC_SUFFIX@_threads_la-f77api.Tpo $(DEPDIR)/libfftw3@PREC_SUFFIX@_threads_la-f77api.Plo @AMDEP_TRUE@@am__fastdepCC_FALSE@ source='f77api.c' object='libfftw3@PREC_SUFFIX@_threads_la-f77api.lo' libtool=yes @AMDEPBACKSLASH@ @AMDEP_TRUE@@am__fastdepCC_FALSE@ DEPDIR=$(DEPDIR) $(CCDEPMODE) $(depcomp) @AMDEPBACKSLASH@ @am__fastdepCC_FALSE@ $(LIBTOOL) --tag=CC $(AM_LIBTOOLFLAGS) $(LIBTOOLFLAGS) --mode=compile $(CC) $(DEFS) $(DEFAULT_INCLUDES) $(INCLUDES) $(AM_CPPFLAGS) $(CPPFLAGS) $(libfftw3@PREC_SUFFIX@_threads_la_CFLAGS) $(CFLAGS) -c -o libfftw3@PREC_SUFFIX@_threads_la-f77api.lo `test -f 'f77api.c' || echo '$(srcdir)/'`f77api.c mostlyclean-libtool: -rm -f *.lo clean-libtool: -rm -rf .libs _libs ID: $(HEADERS) $(SOURCES) $(LISP) $(TAGS_FILES) list='$(SOURCES) $(HEADERS) $(LISP) $(TAGS_FILES)'; \ unique=`for i in $$list; do \ if test -f "$$i"; then echo $$i; else echo $(srcdir)/$$i; fi; \ done | \ $(AWK) '{ files[$$0] = 1; nonempty = 1; } \ END { if (nonempty) { for (i in files) print i; }; }'`; \ mkid -fID $$unique tags: TAGS TAGS: $(HEADERS) $(SOURCES) $(TAGS_DEPENDENCIES) \ $(TAGS_FILES) $(LISP) set x; \ here=`pwd`; \ list='$(SOURCES) $(HEADERS) $(LISP) $(TAGS_FILES)'; \ unique=`for i in $$list; do \ if test -f "$$i"; then echo $$i; else echo $(srcdir)/$$i; fi; \ done | \ $(AWK) '{ files[$$0] = 1; nonempty = 1; } \ END { if (nonempty) { for (i in files) print i; }; }'`; \ shift; \ if test -z "$(ETAGS_ARGS)$$*$$unique"; then :; else \ test -n "$$unique" || unique=$$empty_fix; \ if test $$# -gt 0; then \ $(ETAGS) $(ETAGSFLAGS) $(AM_ETAGSFLAGS) $(ETAGS_ARGS) \ "$$@" $$unique; \ else \ $(ETAGS) $(ETAGSFLAGS) $(AM_ETAGSFLAGS) $(ETAGS_ARGS) \ $$unique; \ fi; \ fi ctags: CTAGS CTAGS: $(HEADERS) $(SOURCES) $(TAGS_DEPENDENCIES) \ $(TAGS_FILES) $(LISP) list='$(SOURCES) $(HEADERS) $(LISP) $(TAGS_FILES)'; \ unique=`for i in $$list; do \ if test -f "$$i"; then echo $$i; else echo $(srcdir)/$$i; fi; \ done | \ $(AWK) '{ files[$$0] = 1; nonempty = 1; } \ END { if (nonempty) { for (i in files) print i; }; }'`; \ test -z "$(CTAGS_ARGS)$$unique" \ || $(CTAGS) $(CTAGSFLAGS) $(AM_CTAGSFLAGS) $(CTAGS_ARGS) \ $$unique GTAGS: here=`$(am__cd) $(top_builddir) && pwd` \ && $(am__cd) $(top_srcdir) \ && gtags -i $(GTAGS_ARGS) "$$here" distclean-tags: -rm -f TAGS ID GTAGS GRTAGS GSYMS GPATH tags distdir: $(DISTFILES) @srcdirstrip=`echo "$(srcdir)" | sed 's/[].[^$$\\*]/\\\\&/g'`; \ topsrcdirstrip=`echo "$(top_srcdir)" | sed 's/[].[^$$\\*]/\\\\&/g'`; \ list='$(DISTFILES)'; \ dist_files=`for file in $$list; do echo $$file; done | \ sed -e "s|^$$srcdirstrip/||;t" \ -e "s|^$$topsrcdirstrip/|$(top_builddir)/|;t"`; \ case $$dist_files in \ */*) $(MKDIR_P) `echo "$$dist_files" | \ sed '/\//!d;s|^|$(distdir)/|;s,/[^/]*$$,,' | \ sort -u` ;; \ esac; \ for file in $$dist_files; do \ if test -f $$file || test -d $$file; then d=.; else d=$(srcdir); fi; \ if test -d $$d/$$file; then \ dir=`echo "/$$file" | sed -e 's,/[^/]*$$,,'`; \ if test -d "$(distdir)/$$file"; then \ find "$(distdir)/$$file" -type d ! -perm -700 -exec chmod u+rwx {} \;; \ fi; \ if test -d $(srcdir)/$$file && test $$d != $(srcdir); then \ cp -fpR $(srcdir)/$$file "$(distdir)$$dir" || exit 1; \ find "$(distdir)/$$file" -type d ! -perm -700 -exec chmod u+rwx {} \;; \ fi; \ cp -fpR $$d/$$file "$(distdir)$$dir" || exit 1; \ else \ test -f "$(distdir)/$$file" \ || cp -p $$d/$$file "$(distdir)/$$file" \ || exit 1; \ fi; \ done check-am: all-am check: check-am all-am: Makefile $(LTLIBRARIES) installdirs: for dir in "$(DESTDIR)$(libdir)"; do \ test -z "$$dir" || $(MKDIR_P) "$$dir"; \ done install: install-am install-exec: install-exec-am install-data: install-data-am uninstall: uninstall-am install-am: all-am @$(MAKE) $(AM_MAKEFLAGS) install-exec-am install-data-am installcheck: installcheck-am install-strip: if test -z '$(STRIP)'; then \ $(MAKE) $(AM_MAKEFLAGS) INSTALL_PROGRAM="$(INSTALL_STRIP_PROGRAM)" \ install_sh_PROGRAM="$(INSTALL_STRIP_PROGRAM)" INSTALL_STRIP_FLAG=-s \ install; \ else \ $(MAKE) $(AM_MAKEFLAGS) INSTALL_PROGRAM="$(INSTALL_STRIP_PROGRAM)" \ install_sh_PROGRAM="$(INSTALL_STRIP_PROGRAM)" INSTALL_STRIP_FLAG=-s \ "INSTALL_PROGRAM_ENV=STRIPPROG='$(STRIP)'" install; \ fi mostlyclean-generic: clean-generic: distclean-generic: -test -z "$(CONFIG_CLEAN_FILES)" || rm -f $(CONFIG_CLEAN_FILES) -test . = "$(srcdir)" || test -z "$(CONFIG_CLEAN_VPATH_FILES)" || rm -f $(CONFIG_CLEAN_VPATH_FILES) maintainer-clean-generic: @echo "This command is intended for maintainers to use" @echo "it deletes files that may require special tools to rebuild." clean: clean-am clean-am: clean-generic clean-libLTLIBRARIES clean-libtool \ clean-noinstLTLIBRARIES mostlyclean-am distclean: distclean-am -rm -rf ./$(DEPDIR) -rm -f Makefile distclean-am: clean-am distclean-compile distclean-generic \ distclean-tags dvi: dvi-am dvi-am: html: html-am html-am: info: info-am info-am: install-data-am: install-dvi: install-dvi-am install-dvi-am: install-exec-am: install-libLTLIBRARIES install-html: install-html-am install-html-am: install-info: install-info-am install-info-am: install-man: install-pdf: install-pdf-am install-pdf-am: install-ps: install-ps-am install-ps-am: installcheck-am: maintainer-clean: maintainer-clean-am -rm -rf ./$(DEPDIR) -rm -f Makefile maintainer-clean-am: distclean-am maintainer-clean-generic mostlyclean: mostlyclean-am mostlyclean-am: mostlyclean-compile mostlyclean-generic \ mostlyclean-libtool pdf: pdf-am pdf-am: ps: ps-am ps-am: uninstall-am: uninstall-libLTLIBRARIES .MAKE: install-am install-strip .PHONY: CTAGS GTAGS all all-am check check-am clean clean-generic \ clean-libLTLIBRARIES clean-libtool clean-noinstLTLIBRARIES \ ctags distclean distclean-compile distclean-generic \ distclean-libtool distclean-tags distdir dvi dvi-am html \ html-am info info-am install install-am install-data \ install-data-am install-dvi install-dvi-am install-exec \ install-exec-am install-html install-html-am install-info \ install-info-am install-libLTLIBRARIES install-man install-pdf \ install-pdf-am install-ps install-ps-am install-strip \ installcheck installcheck-am installdirs maintainer-clean \ maintainer-clean-generic mostlyclean mostlyclean-compile \ mostlyclean-generic mostlyclean-libtool pdf pdf-am ps ps-am \ tags uninstall uninstall-am uninstall-libLTLIBRARIES # Tell versions [3.59,3.63) of GNU make to not export all variables. # Otherwise a system limit (for SysV at least) may be exceeded. .NOEXPORT: fftw-3.3.3/threads/api.c0000644000175000017500000000376312054410050011756 00000000000000/* * Copyright (c) 2003, 2007-11 Matteo Frigo * Copyright (c) 2003, 2007-11 Massachusetts Institute of Technology * * This program is free software; you can redistribute it and/or modify * it under the terms of the GNU General Public License as published by * the Free Software Foundation; either version 2 of the License, or * (at your option) any later version. * * This program is distributed in the hope that it will be useful, * but WITHOUT ANY WARRANTY; without even the implied warranty of * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the * GNU General Public License for more details. * * You should have received a copy of the GNU General Public License * along with this program; if not, write to the Free Software * Foundation, Inc., 51 Franklin Street, Fifth Floor, Boston, MA 02110-1301 USA * */ #include "api.h" #include "threads.h" static int threads_inited = 0; static void threads_register_hooks(void) { X(mksolver_ct_hook) = X(mksolver_ct_threads); X(mksolver_hc2hc_hook) = X(mksolver_hc2hc_threads); } static void threads_unregister_hooks(void) { X(mksolver_ct_hook) = 0; X(mksolver_hc2hc_hook) = 0; } /* should be called before all other FFTW functions! */ int X(init_threads)(void) { if (!threads_inited) { planner *plnr; if (X(ithreads_init)()) return 0; threads_register_hooks(); /* this should be the first time the_planner is called, and hence the time it is configured */ plnr = X(the_planner)(); X(threads_conf_standard)(plnr); threads_inited = 1; } return 1; } void X(cleanup_threads)(void) { X(cleanup)(); if (threads_inited) { X(threads_cleanup)(); threads_unregister_hooks(); threads_inited = 0; } } void X(plan_with_nthreads)(int nthreads) { planner *plnr; if (!threads_inited) { X(cleanup)(); X(init_threads)(); } A(threads_inited); plnr = X(the_planner)(); plnr->nthr = X(imax)(1, nthreads); } fftw-3.3.3/threads/conf.c0000644000175000017500000000215412054410050012123 00000000000000/* * Copyright (c) 2003, 2007-11 Matteo Frigo * Copyright (c) 2003, 2007-11 Massachusetts Institute of Technology * * This program is free software; you can redistribute it and/or modify * it under the terms of the GNU General Public License as published by * the Free Software Foundation; either version 2 of the License, or * (at your option) any later version. * * This program is distributed in the hope that it will be useful, * but WITHOUT ANY WARRANTY; without even the implied warranty of * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the * GNU General Public License for more details. * * You should have received a copy of the GNU General Public License * along with this program; if not, write to the Free Software * Foundation, Inc., 51 Franklin Street, Fifth Floor, Boston, MA 02110-1301 USA * */ #include "threads.h" static const solvtab s = { SOLVTAB(X(dft_thr_vrank_geq1_register)), SOLVTAB(X(rdft_thr_vrank_geq1_register)), SOLVTAB(X(rdft2_thr_vrank_geq1_register)), SOLVTAB_END }; void X(threads_conf_standard)(planner *p) { X(solvtab_exec)(s, p); } fftw-3.3.3/threads/openmp.c0000644000175000017500000000466712054410050012507 00000000000000/* * Copyright (c) 2003, 2007-11 Matteo Frigo * Copyright (c) 2003, 2007-11 Massachusetts Institute of Technology * * This program is free software; you can redistribute it and/or modify * it under the terms of the GNU General Public License as published by * the Free Software Foundation; either version 2 of the License, or * (at your option) any later version. * * This program is distributed in the hope that it will be useful, * but WITHOUT ANY WARRANTY; without even the implied warranty of * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the * GNU General Public License for more details. * * You should have received a copy of the GNU General Public License * along with this program; if not, write to the Free Software * Foundation, Inc., 51 Franklin Street, Fifth Floor, Boston, MA 02110-1301 USA * */ /* openmp.c: thread spawning via OpenMP */ #include "threads.h" #if !defined(_OPENMP) #error OpenMP enabled but not using an OpenMP compiler #endif int X(ithreads_init)(void) { return 0; /* no error */ } /* Distribute a loop from 0 to loopmax-1 over nthreads threads. proc(d) is called to execute a block of iterations from d->min to d->max-1. d->thr_num indicate the number of the thread that is executing proc (from 0 to nthreads-1), and d->data is the same as the data parameter passed to X(spawn_loop). This function returns only after all the threads have completed. */ void X(spawn_loop)(int loopmax, int nthr, spawn_function proc, void *data) { int block_size; spawn_data d; int i; A(loopmax >= 0); A(nthr > 0); A(proc); if (!loopmax) return; /* Choose the block size and number of threads in order to (1) minimize the critical path and (2) use the fewest threads that achieve the same critical path (to minimize overhead). e.g. if loopmax is 5 and nthr is 4, we should use only 3 threads with block sizes of 2, 2, and 1. */ block_size = (loopmax + nthr - 1) / nthr; nthr = (loopmax + block_size - 1) / block_size; THREAD_ON; /* prevent debugging mode from failing under threads */ #pragma omp parallel for private(d) for (i = 0; i < nthr; ++i) { d.max = (d.min = i * block_size) + block_size; if (d.max > loopmax) d.max = loopmax; d.thr_num = i; d.data = data; proc(&d); } THREAD_OFF; /* prevent debugging mode from failing under threads */ } void X(threads_cleanup)(void) { } fftw-3.3.3/threads/threads.h0000644000175000017500000000334212054410050012635 00000000000000/* * Copyright (c) 2003, 2007-11 Matteo Frigo * Copyright (c) 2003, 2007-11 Massachusetts Institute of Technology * * This program is free software; you can redistribute it and/or modify * it under the terms of the GNU General Public License as published by * the Free Software Foundation; either version 2 of the License, or * (at your option) any later version. * * This program is distributed in the hope that it will be useful, * but WITHOUT ANY WARRANTY; without even the implied warranty of * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the * GNU General Public License for more details. * * You should have received a copy of the GNU General Public License * along with this program; if not, write to the Free Software * Foundation, Inc., 51 Franklin Street, Fifth Floor, Boston, MA 02110-1301 USA * */ #ifndef __THREADS_H__ #define __THREADS_H__ #include "ifftw.h" #include "ct.h" #include "hc2hc.h" typedef struct { int min, max, thr_num; void *data; } spawn_data; typedef void *(*spawn_function) (spawn_data *); void X(spawn_loop)(int loopmax, int nthreads, spawn_function proc, void *data); int X(ithreads_init)(void); void X(threads_cleanup)(void); /* configurations */ void X(dft_thr_vrank_geq1_register)(planner *p); void X(rdft_thr_vrank_geq1_register)(planner *p); void X(rdft2_thr_vrank_geq1_register)(planner *p); ct_solver *X(mksolver_ct_threads)(size_t size, INT r, int dec, ct_mkinferior mkcldw, ct_force_vrecursion force_vrecursionp); hc2hc_solver *X(mksolver_hc2hc_threads)(size_t size, INT r, hc2hc_mkinferior mkcldw); void X(threads_conf_standard)(planner *p); void X(threads_register_hooks)(void); void X(threads_unregister_hooks)(void); #endif /* __THREADS_H__ */ fftw-3.3.3/threads/dft-vrank-geq1.c0000644000175000017500000001360512054410050013730 00000000000000/* * Copyright (c) 2003, 2007-11 Matteo Frigo * Copyright (c) 2003, 2007-11 Massachusetts Institute of Technology * * This program is free software; you can redistribute it and/or modify * it under the terms of the GNU General Public License as published by * the Free Software Foundation; either version 2 of the License, or * (at your option) any later version. * * This program is distributed in the hope that it will be useful, * but WITHOUT ANY WARRANTY; without even the implied warranty of * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the * GNU General Public License for more details. * * You should have received a copy of the GNU General Public License * along with this program; if not, write to the Free Software * Foundation, Inc., 51 Franklin Street, Fifth Floor, Boston, MA 02110-1301 USA * */ #include "threads.h" typedef struct { solver super; int vecloop_dim; const int *buddies; int nbuddies; } S; typedef struct { plan_dft super; plan **cldrn; INT its, ots; int nthr; const S *solver; } P; typedef struct { INT its, ots; R *ri, *ii, *ro, *io; plan **cldrn; } PD; static void *spawn_apply(spawn_data *d) { PD *ego = (PD *) d->data; INT its = ego->its; INT ots = ego->ots; int thr_num = d->thr_num; plan_dft *cld = (plan_dft *) ego->cldrn[thr_num]; cld->apply((plan *) cld, ego->ri + thr_num * its, ego->ii + thr_num * its, ego->ro + thr_num * ots, ego->io + thr_num * ots); return 0; } static void apply(const plan *ego_, R *ri, R *ii, R *ro, R *io) { const P *ego = (const P *) ego_; PD d; d.its = ego->its; d.ots = ego->ots; d.cldrn = ego->cldrn; d.ri = ri; d.ii = ii; d.ro = ro; d.io = io; X(spawn_loop)(ego->nthr, ego->nthr, spawn_apply, (void*) &d); } static void awake(plan *ego_, enum wakefulness wakefulness) { P *ego = (P *) ego_; int i; for (i = 0; i < ego->nthr; ++i) X(plan_awake)(ego->cldrn[i], wakefulness); } static void destroy(plan *ego_) { P *ego = (P *) ego_; int i; for (i = 0; i < ego->nthr; ++i) X(plan_destroy_internal)(ego->cldrn[i]); X(ifree)(ego->cldrn); } static void print(const plan *ego_, printer *p) { const P *ego = (const P *) ego_; const S *s = ego->solver; int i; p->print(p, "(dft-thr-vrank>=1-x%d/%d", ego->nthr, s->vecloop_dim); for (i = 0; i < ego->nthr; ++i) if (i == 0 || (ego->cldrn[i] != ego->cldrn[i-1] && (i <= 1 || ego->cldrn[i] != ego->cldrn[i-2]))) p->print(p, "%(%p%)", ego->cldrn[i]); p->putchr(p, ')'); } static int pickdim(const S *ego, const tensor *vecsz, int oop, int *dp) { return X(pickdim)(ego->vecloop_dim, ego->buddies, ego->nbuddies, vecsz, oop, dp); } static int applicable0(const solver *ego_, const problem *p_, const planner *plnr, int *dp) { const S *ego = (const S *) ego_; const problem_dft *p = (const problem_dft *) p_; return (1 && plnr->nthr > 1 && FINITE_RNK(p->vecsz->rnk) && p->vecsz->rnk > 0 && pickdim(ego, p->vecsz, p->ri != p->ro, dp) ); } static int applicable(const solver *ego_, const problem *p_, const planner *plnr, int *dp) { const S *ego = (const S *)ego_; if (!applicable0(ego_, p_, plnr, dp)) return 0; /* fftw2 behavior */ if (NO_VRANK_SPLITSP(plnr) && (ego->vecloop_dim != ego->buddies[0])) return 0; return 1; } static plan *mkplan(const solver *ego_, const problem *p_, planner *plnr) { const S *ego = (const S *) ego_; const problem_dft *p; P *pln; problem *cldp; int vdim; iodim *d; plan **cldrn = (plan **) 0; int i, nthr; INT its, ots, block_size; tensor *vecsz = 0; static const plan_adt padt = { X(dft_solve), awake, print, destroy }; if (!applicable(ego_, p_, plnr, &vdim)) return (plan *) 0; p = (const problem_dft *) p_; d = p->vecsz->dims + vdim; block_size = (d->n + plnr->nthr - 1) / plnr->nthr; nthr = (int)((d->n + block_size - 1) / block_size); plnr->nthr = (plnr->nthr + nthr - 1) / nthr; its = d->is * block_size; ots = d->os * block_size; cldrn = (plan **)MALLOC(sizeof(plan *) * nthr, PLANS); for (i = 0; i < nthr; ++i) cldrn[i] = (plan *) 0; vecsz = X(tensor_copy)(p->vecsz); for (i = 0; i < nthr; ++i) { vecsz->dims[vdim].n = (i == nthr - 1) ? (d->n - i*block_size) : block_size; cldp = X(mkproblem_dft)(p->sz, vecsz, p->ri + i*its, p->ii + i*its, p->ro + i*ots, p->io + i*ots); cldrn[i] = X(mkplan_d)(plnr, cldp); if (!cldrn[i]) goto nada; } X(tensor_destroy)(vecsz); pln = MKPLAN_DFT(P, &padt, apply); pln->cldrn = cldrn; pln->its = its; pln->ots = ots; pln->nthr = nthr; pln->solver = ego; X(ops_zero)(&pln->super.super.ops); pln->super.super.pcost = 0; for (i = 0; i < nthr; ++i) { X(ops_add2)(&cldrn[i]->ops, &pln->super.super.ops); pln->super.super.pcost += cldrn[i]->pcost; } return &(pln->super.super); nada: if (cldrn) { for (i = 0; i < nthr; ++i) X(plan_destroy_internal)(cldrn[i]); X(ifree)(cldrn); } X(tensor_destroy)(vecsz); return (plan *) 0; } static solver *mksolver(int vecloop_dim, const int *buddies, int nbuddies) { static const solver_adt sadt = { PROBLEM_DFT, mkplan, 0 }; S *slv = MKSOLVER(S, &sadt); slv->vecloop_dim = vecloop_dim; slv->buddies = buddies; slv->nbuddies = nbuddies; return &(slv->super); } void X(dft_thr_vrank_geq1_register)(planner *p) { int i; /* FIXME: Should we try other vecloop_dim values? */ static const int buddies[] = { 1, -1 }; const int nbuddies = (int)(sizeof(buddies) / sizeof(buddies[0])); for (i = 0; i < nbuddies; ++i) REGISTER_SOLVER(p, mksolver(buddies[i], buddies, nbuddies)); } fftw-3.3.3/threads/ct.c0000644000175000017500000001524612054410050011612 00000000000000/* * Copyright (c) 2003, 2007-11 Matteo Frigo * Copyright (c) 2003, 2007-11 Massachusetts Institute of Technology * * This program is free software; you can redistribute it and/or modify * it under the terms of the GNU General Public License as published by * the Free Software Foundation; either version 2 of the License, or * (at your option) any later version. * * This program is distributed in the hope that it will be useful, * but WITHOUT ANY WARRANTY; without even the implied warranty of * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the * GNU General Public License for more details. * * You should have received a copy of the GNU General Public License * along with this program; if not, write to the Free Software * Foundation, Inc., 51 Franklin Street, Fifth Floor, Boston, MA 02110-1301 USA * */ #include "threads.h" typedef struct { plan_dft super; plan *cld; plan **cldws; int nthr; INT r; } P; typedef struct { plan **cldws; R *r, *i; } PD; static void *spawn_apply(spawn_data *d) { PD *ego = (PD *) d->data; INT thr_num = d->thr_num; plan_dftw *cldw = (plan_dftw *) (ego->cldws[thr_num]); cldw->apply((plan *) cldw, ego->r, ego->i); return 0; } static void apply_dit(const plan *ego_, R *ri, R *ii, R *ro, R *io) { const P *ego = (const P *) ego_; plan_dft *cld; cld = (plan_dft *) ego->cld; cld->apply(ego->cld, ri, ii, ro, io); { PD d; d.r = ro; d.i = io; d.cldws = ego->cldws; X(spawn_loop)(ego->nthr, ego->nthr, spawn_apply, (void*)&d); } } static void apply_dif(const plan *ego_, R *ri, R *ii, R *ro, R *io) { const P *ego = (const P *) ego_; plan_dft *cld; { PD d; d.r = ri; d.i = ii; d.cldws = ego->cldws; X(spawn_loop)(ego->nthr, ego->nthr, spawn_apply, (void*)&d); } cld = (plan_dft *) ego->cld; cld->apply(ego->cld, ri, ii, ro, io); } static void awake(plan *ego_, enum wakefulness wakefulness) { P *ego = (P *) ego_; int i; X(plan_awake)(ego->cld, wakefulness); for (i = 0; i < ego->nthr; ++i) X(plan_awake)(ego->cldws[i], wakefulness); } static void destroy(plan *ego_) { P *ego = (P *) ego_; int i; X(plan_destroy_internal)(ego->cld); for (i = 0; i < ego->nthr; ++i) X(plan_destroy_internal)(ego->cldws[i]); X(ifree)(ego->cldws); } static void print(const plan *ego_, printer *p) { const P *ego = (const P *) ego_; int i; p->print(p, "(dft-thr-ct-%s-x%d/%D", ego->super.apply == apply_dit ? "dit" : "dif", ego->nthr, ego->r); for (i = 0; i < ego->nthr; ++i) if (i == 0 || (ego->cldws[i] != ego->cldws[i-1] && (i <= 1 || ego->cldws[i] != ego->cldws[i-2]))) p->print(p, "%(%p%)", ego->cldws[i]); p->print(p, "%(%p%))", ego->cld); } static plan *mkplan(const solver *ego_, const problem *p_, planner *plnr) { const ct_solver *ego = (const ct_solver *) ego_; const problem_dft *p; P *pln = 0; plan *cld = 0, **cldws = 0; INT n, r, m, v, ivs, ovs; INT block_size; int i, nthr, plnr_nthr_save; iodim *d; static const plan_adt padt = { X(dft_solve), awake, print, destroy }; if (plnr->nthr <= 1 || !X(ct_applicable)(ego, p_, plnr)) return (plan *) 0; p = (const problem_dft *) p_; d = p->sz->dims; n = d[0].n; r = X(choose_radix)(ego->r, n); m = n / r; X(tensor_tornk1)(p->vecsz, &v, &ivs, &ovs); block_size = (m + plnr->nthr - 1) / plnr->nthr; nthr = (int)((m + block_size - 1) / block_size); plnr_nthr_save = plnr->nthr; plnr->nthr = (plnr->nthr + nthr - 1) / nthr; cldws = (plan **) MALLOC(sizeof(plan *) * nthr, PLANS); for (i = 0; i < nthr; ++i) cldws[i] = (plan *) 0; switch (ego->dec) { case DECDIT: { for (i = 0; i < nthr; ++i) { cldws[i] = ego->mkcldw(ego, r, m * d[0].os, m * d[0].os, m, d[0].os, v, ovs, ovs, i*block_size, (i == nthr - 1) ? (m - i*block_size) : block_size, p->ro, p->io, plnr); if (!cldws[i]) goto nada; } plnr->nthr = plnr_nthr_save; cld = X(mkplan_d)(plnr, X(mkproblem_dft_d)( X(mktensor_1d)(m, r * d[0].is, d[0].os), X(mktensor_2d)(r, d[0].is, m * d[0].os, v, ivs, ovs), p->ri, p->ii, p->ro, p->io) ); if (!cld) goto nada; pln = MKPLAN_DFT(P, &padt, apply_dit); break; } case DECDIF: case DECDIF+TRANSPOSE: { INT cors, covs; /* cldw ors, ovs */ if (ego->dec == DECDIF+TRANSPOSE) { cors = ivs; covs = m * d[0].is; /* ensure that we generate well-formed dftw subproblems */ /* FIXME: too conservative */ if (!(1 && r == v && d[0].is == r * cors)) goto nada; /* FIXME: allow in-place only for now, like in fftw-3.[01] */ if (!(1 && p->ri == p->ro && d[0].is == r * d[0].os && cors == d[0].os && covs == ovs )) goto nada; } else { cors = m * d[0].is; covs = ivs; } for (i = 0; i < nthr; ++i) { cldws[i] = ego->mkcldw(ego, r, m * d[0].is, cors, m, d[0].is, v, ivs, covs, i*block_size, (i == nthr - 1) ? (m - i*block_size) : block_size, p->ri, p->ii, plnr); if (!cldws[i]) goto nada; } plnr->nthr = plnr_nthr_save; cld = X(mkplan_d)(plnr, X(mkproblem_dft_d)( X(mktensor_1d)(m, d[0].is, r * d[0].os), X(mktensor_2d)(r, cors, d[0].os, v, covs, ovs), p->ri, p->ii, p->ro, p->io) ); if (!cld) goto nada; pln = MKPLAN_DFT(P, &padt, apply_dif); break; } default: A(0); } pln->cld = cld; pln->cldws = cldws; pln->nthr = nthr; pln->r = r; X(ops_zero)(&pln->super.super.ops); for (i = 0; i < nthr; ++i) { X(ops_add2)(&cldws[i]->ops, &pln->super.super.ops); pln->super.super.could_prune_now_p |= cldws[i]->could_prune_now_p; } X(ops_add2)(&cld->ops, &pln->super.super.ops); return &(pln->super.super); nada: if (cldws) { for (i = 0; i < nthr; ++i) X(plan_destroy_internal)(cldws[i]); X(ifree)(cldws); } X(plan_destroy_internal)(cld); return (plan *) 0; } ct_solver *X(mksolver_ct_threads)(size_t size, INT r, int dec, ct_mkinferior mkcldw, ct_force_vrecursion force_vrecursionp) { static const solver_adt sadt = { PROBLEM_DFT, mkplan, 0 }; ct_solver *slv = (ct_solver *) X(mksolver)(size, &sadt); slv->r = r; slv->dec = dec; slv->mkcldw = mkcldw; slv->force_vrecursionp = force_vrecursionp; return slv; } fftw-3.3.3/threads/rdft-vrank-geq1.c0000644000175000017500000001332112054410050014105 00000000000000/* * Copyright (c) 2003, 2007-11 Matteo Frigo * Copyright (c) 2003, 2007-11 Massachusetts Institute of Technology * * This program is free software; you can redistribute it and/or modify * it under the terms of the GNU General Public License as published by * the Free Software Foundation; either version 2 of the License, or * (at your option) any later version. * * This program is distributed in the hope that it will be useful, * but WITHOUT ANY WARRANTY; without even the implied warranty of * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the * GNU General Public License for more details. * * You should have received a copy of the GNU General Public License * along with this program; if not, write to the Free Software * Foundation, Inc., 51 Franklin Street, Fifth Floor, Boston, MA 02110-1301 USA * */ #include "threads.h" typedef struct { solver super; int vecloop_dim; const int *buddies; int nbuddies; } S; typedef struct { plan_rdft super; plan **cldrn; INT its, ots; int nthr; const S *solver; } P; typedef struct { INT its, ots; R *I, *O; plan **cldrn; } PD; static void *spawn_apply(spawn_data *d) { PD *ego = (PD *) d->data; int thr_num = d->thr_num; plan_rdft *cld = (plan_rdft *) ego->cldrn[d->thr_num]; cld->apply((plan *) cld, ego->I + thr_num * ego->its, ego->O + thr_num * ego->ots); return 0; } static void apply(const plan *ego_, R *I, R *O) { const P *ego = (const P *) ego_; PD d; d.its = ego->its; d.ots = ego->ots; d.cldrn = ego->cldrn; d.I = I; d.O = O; X(spawn_loop)(ego->nthr, ego->nthr, spawn_apply, (void*) &d); } static void awake(plan *ego_, enum wakefulness wakefulness) { P *ego = (P *) ego_; int i; for (i = 0; i < ego->nthr; ++i) X(plan_awake)(ego->cldrn[i], wakefulness); } static void destroy(plan *ego_) { P *ego = (P *) ego_; int i; for (i = 0; i < ego->nthr; ++i) X(plan_destroy_internal)(ego->cldrn[i]); X(ifree)(ego->cldrn); } static void print(const plan *ego_, printer *p) { const P *ego = (const P *) ego_; const S *s = ego->solver; int i; p->print(p, "(rdft-thr-vrank>=1-x%d/%d", ego->nthr, s->vecloop_dim); for (i = 0; i < ego->nthr; ++i) if (i == 0 || (ego->cldrn[i] != ego->cldrn[i-1] && (i <= 1 || ego->cldrn[i] != ego->cldrn[i-2]))) p->print(p, "%(%p%)", ego->cldrn[i]); p->putchr(p, ')'); } static int pickdim(const S *ego, const tensor *vecsz, int oop, int *dp) { return X(pickdim)(ego->vecloop_dim, ego->buddies, ego->nbuddies, vecsz, oop, dp); } static int applicable0(const solver *ego_, const problem *p_, const planner *plnr, int *dp) { const S *ego = (const S *) ego_; const problem_rdft *p = (const problem_rdft *) p_; return (1 && plnr->nthr > 1 && FINITE_RNK(p->vecsz->rnk) && p->vecsz->rnk > 0 && pickdim(ego, p->vecsz, p->I != p->O, dp) ); } static int applicable(const solver *ego_, const problem *p_, const planner *plnr, int *dp) { const S *ego = (const S *)ego_; if (!applicable0(ego_, p_, plnr, dp)) return 0; /* fftw2 behavior */ if (NO_VRANK_SPLITSP(plnr) && (ego->vecloop_dim != ego->buddies[0])) return 0; return 1; } static plan *mkplan(const solver *ego_, const problem *p_, planner *plnr) { const S *ego = (const S *) ego_; const problem_rdft *p; P *pln; problem *cldp; int vdim; iodim *d; plan **cldrn = (plan **) 0; int i, nthr; INT its, ots, block_size; tensor *vecsz; static const plan_adt padt = { X(rdft_solve), awake, print, destroy }; if (!applicable(ego_, p_, plnr, &vdim)) return (plan *) 0; p = (const problem_rdft *) p_; d = p->vecsz->dims + vdim; block_size = (d->n + plnr->nthr - 1) / plnr->nthr; nthr = (int)((d->n + block_size - 1) / block_size); plnr->nthr = (plnr->nthr + nthr - 1) / nthr; its = d->is * block_size; ots = d->os * block_size; cldrn = (plan **)MALLOC(sizeof(plan *) * nthr, PLANS); for (i = 0; i < nthr; ++i) cldrn[i] = (plan *) 0; vecsz = X(tensor_copy)(p->vecsz); for (i = 0; i < nthr; ++i) { vecsz->dims[vdim].n = (i == nthr - 1) ? (d->n - i*block_size) : block_size; cldp = X(mkproblem_rdft)(p->sz, vecsz, p->I + i*its, p->O + i*ots, p->kind); cldrn[i] = X(mkplan_d)(plnr, cldp); if (!cldrn[i]) goto nada; } X(tensor_destroy)(vecsz); pln = MKPLAN_RDFT(P, &padt, apply); pln->cldrn = cldrn; pln->its = its; pln->ots = ots; pln->nthr = nthr; pln->solver = ego; X(ops_zero)(&pln->super.super.ops); pln->super.super.pcost = 0; for (i = 0; i < nthr; ++i) { X(ops_add2)(&cldrn[i]->ops, &pln->super.super.ops); pln->super.super.pcost += cldrn[i]->pcost; } return &(pln->super.super); nada: if (cldrn) { for (i = 0; i < nthr; ++i) X(plan_destroy_internal)(cldrn[i]); X(ifree)(cldrn); } X(tensor_destroy)(vecsz); return (plan *) 0; } static solver *mksolver(int vecloop_dim, const int *buddies, int nbuddies) { static const solver_adt sadt = { PROBLEM_RDFT, mkplan, 0 }; S *slv = MKSOLVER(S, &sadt); slv->vecloop_dim = vecloop_dim; slv->buddies = buddies; slv->nbuddies = nbuddies; return &(slv->super); } void X(rdft_thr_vrank_geq1_register)(planner *p) { int i; /* FIXME: Should we try other vecloop_dim values? */ static const int buddies[] = { 1, -1 }; const int nbuddies = (int)(sizeof(buddies) / sizeof(buddies[0])); for (i = 0; i < nbuddies; ++i) REGISTER_SOLVER(p, mksolver(buddies[i], buddies, nbuddies)); } fftw-3.3.3/threads/hc2hc.c0000644000175000017500000001357612054410050012177 00000000000000/* * Copyright (c) 2003, 2007-11 Matteo Frigo * Copyright (c) 2003, 2007-11 Massachusetts Institute of Technology * * This program is free software; you can redistribute it and/or modify * it under the terms of the GNU General Public License as published by * the Free Software Foundation; either version 2 of the License, or * (at your option) any later version. * * This program is distributed in the hope that it will be useful, * but WITHOUT ANY WARRANTY; without even the implied warranty of * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the * GNU General Public License for more details. * * You should have received a copy of the GNU General Public License * along with this program; if not, write to the Free Software * Foundation, Inc., 51 Franklin Street, Fifth Floor, Boston, MA 02110-1301 USA * */ #include "threads.h" typedef struct { plan_rdft super; plan *cld; plan **cldws; int nthr; INT r; } P; typedef struct { plan **cldws; R *IO; } PD; static void *spawn_apply(spawn_data *d) { PD *ego = (PD *) d->data; plan_hc2hc *cldw = (plan_hc2hc *) (ego->cldws[d->thr_num]); cldw->apply((plan *) cldw, ego->IO); return 0; } static void apply_dit(const plan *ego_, R *I, R *O) { const P *ego = (const P *) ego_; plan_rdft *cld; cld = (plan_rdft *) ego->cld; cld->apply((plan *) cld, I, O); { PD d; d.IO = O; d.cldws = ego->cldws; X(spawn_loop)(ego->nthr, ego->nthr, spawn_apply, (void*)&d); } } static void apply_dif(const plan *ego_, R *I, R *O) { const P *ego = (const P *) ego_; plan_rdft *cld; { PD d; d.IO = I; d.cldws = ego->cldws; X(spawn_loop)(ego->nthr, ego->nthr, spawn_apply, (void*)&d); } cld = (plan_rdft *) ego->cld; cld->apply((plan *) cld, I, O); } static void awake(plan *ego_, enum wakefulness wakefulness) { P *ego = (P *) ego_; int i; X(plan_awake)(ego->cld, wakefulness); for (i = 0; i < ego->nthr; ++i) X(plan_awake)(ego->cldws[i], wakefulness); } static void destroy(plan *ego_) { P *ego = (P *) ego_; int i; X(plan_destroy_internal)(ego->cld); for (i = 0; i < ego->nthr; ++i) X(plan_destroy_internal)(ego->cldws[i]); X(ifree)(ego->cldws); } static void print(const plan *ego_, printer *p) { const P *ego = (const P *) ego_; int i; p->print(p, "(rdft-thr-ct-%s-x%d/%D", ego->super.apply == apply_dit ? "dit" : "dif", ego->nthr, ego->r); for (i = 0; i < ego->nthr; ++i) if (i == 0 || (ego->cldws[i] != ego->cldws[i-1] && (i <= 1 || ego->cldws[i] != ego->cldws[i-2]))) p->print(p, "%(%p%)", ego->cldws[i]); p->print(p, "%(%p%))", ego->cld); } static plan *mkplan(const solver *ego_, const problem *p_, planner *plnr) { const hc2hc_solver *ego = (const hc2hc_solver *) ego_; const problem_rdft *p; P *pln = 0; plan *cld = 0, **cldws = 0; INT n, r, m, v, ivs, ovs, mcount; int i, nthr, plnr_nthr_save; INT block_size; iodim *d; static const plan_adt padt = { X(rdft_solve), awake, print, destroy }; if (plnr->nthr <= 1 || !X(hc2hc_applicable)(ego, p_, plnr)) return (plan *) 0; p = (const problem_rdft *) p_; d = p->sz->dims; n = d[0].n; r = X(choose_radix)(ego->r, n); m = n / r; mcount = (m + 2) / 2; X(tensor_tornk1)(p->vecsz, &v, &ivs, &ovs); block_size = (mcount + plnr->nthr - 1) / plnr->nthr; nthr = (int)((mcount + block_size - 1) / block_size); plnr_nthr_save = plnr->nthr; plnr->nthr = (plnr->nthr + nthr - 1) / nthr; cldws = (plan **) MALLOC(sizeof(plan *) * nthr, PLANS); for (i = 0; i < nthr; ++i) cldws[i] = (plan *) 0; switch (p->kind[0]) { case R2HC: for (i = 0; i < nthr; ++i) { cldws[i] = ego->mkcldw(ego, R2HC, r, m, d[0].os, v, ovs, i*block_size, (i == nthr - 1) ? (mcount - i*block_size) : block_size, p->O, plnr); if (!cldws[i]) goto nada; } plnr->nthr = plnr_nthr_save; cld = X(mkplan_d)(plnr, X(mkproblem_rdft_d)( X(mktensor_1d)(m, r * d[0].is, d[0].os), X(mktensor_2d)(r, d[0].is, m * d[0].os, v, ivs, ovs), p->I, p->O, p->kind) ); if (!cld) goto nada; pln = MKPLAN_RDFT(P, &padt, apply_dit); break; case HC2R: for (i = 0; i < nthr; ++i) { cldws[i] = ego->mkcldw(ego, HC2R, r, m, d[0].is, v, ivs, i*block_size, (i == nthr - 1) ? (mcount - i*block_size) : block_size, p->I, plnr); if (!cldws[i]) goto nada; } plnr->nthr = plnr_nthr_save; cld = X(mkplan_d)(plnr, X(mkproblem_rdft_d)( X(mktensor_1d)(m, d[0].is, r * d[0].os), X(mktensor_2d)(r, m * d[0].is, d[0].os, v, ivs, ovs), p->I, p->O, p->kind) ); if (!cld) goto nada; pln = MKPLAN_RDFT(P, &padt, apply_dif); break; default: A(0); } pln->cld = cld; pln->cldws = cldws; pln->nthr = nthr; pln->r = r; X(ops_zero)(&pln->super.super.ops); for (i = 0; i < nthr; ++i) { X(ops_add2)(&cldws[i]->ops, &pln->super.super.ops); pln->super.super.could_prune_now_p |= cldws[i]->could_prune_now_p; } X(ops_add2)(&cld->ops, &pln->super.super.ops); return &(pln->super.super); nada: if (cldws) { for (i = 0; i < nthr; ++i) X(plan_destroy_internal)(cldws[i]); X(ifree)(cldws); } X(plan_destroy_internal)(cld); return (plan *) 0; } hc2hc_solver *X(mksolver_hc2hc_threads)(size_t size, INT r, hc2hc_mkinferior mkcldw) { static const solver_adt sadt = { PROBLEM_RDFT, mkplan, 0 }; hc2hc_solver *slv = (hc2hc_solver *)X(mksolver)(size, &sadt); slv->r = r; slv->mkcldw = mkcldw; return slv; } fftw-3.3.3/threads/vrank-geq1-rdft2.c0000644000175000017500000001407412054410050014175 00000000000000/* * Copyright (c) 2003, 2007-11 Matteo Frigo * Copyright (c) 2003, 2007-11 Massachusetts Institute of Technology * * This program is free software; you can redistribute it and/or modify * it under the terms of the GNU General Public License as published by * the Free Software Foundation; either version 2 of the License, or * (at your option) any later version. * * This program is distributed in the hope that it will be useful, * but WITHOUT ANY WARRANTY; without even the implied warranty of * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the * GNU General Public License for more details. * * You should have received a copy of the GNU General Public License * along with this program; if not, write to the Free Software * Foundation, Inc., 51 Franklin Street, Fifth Floor, Boston, MA 02110-1301 USA * */ #include "threads.h" typedef struct { solver super; int vecloop_dim; const int *buddies; int nbuddies; } S; typedef struct { plan_rdft2 super; plan **cldrn; INT its, ots; int nthr; const S *solver; } P; typedef struct { INT its, ots; R *r0, *r1, *cr, *ci; plan **cldrn; } PD; static void *spawn_apply(spawn_data *d) { PD *ego = (PD *) d->data; INT its = ego->its; INT ots = ego->ots; int thr_num = d->thr_num; plan_rdft2 *cld = (plan_rdft2 *) ego->cldrn[d->thr_num]; cld->apply((plan *) cld, ego->r0 + thr_num * its, ego->r1 + thr_num * its, ego->cr + thr_num * ots, ego->ci + thr_num * ots); return 0; } static void apply(const plan *ego_, R *r0, R *r1, R *cr, R *ci) { const P *ego = (const P *) ego_; PD d; d.its = ego->its; d.ots = ego->ots; d.cldrn = ego->cldrn; d.r0 = r0; d.r1 = r1; d.cr = cr; d.ci = ci; X(spawn_loop)(ego->nthr, ego->nthr, spawn_apply, (void*) &d); } static void awake(plan *ego_, enum wakefulness wakefulness) { P *ego = (P *) ego_; int i; for (i = 0; i < ego->nthr; ++i) X(plan_awake)(ego->cldrn[i], wakefulness); } static void destroy(plan *ego_) { P *ego = (P *) ego_; int i; for (i = 0; i < ego->nthr; ++i) X(plan_destroy_internal)(ego->cldrn[i]); X(ifree)(ego->cldrn); } static void print(const plan *ego_, printer *p) { const P *ego = (const P *) ego_; const S *s = ego->solver; int i; p->print(p, "(rdft2-thr-vrank>=1-x%d/%d)", ego->nthr, s->vecloop_dim); for (i = 0; i < ego->nthr; ++i) if (i == 0 || (ego->cldrn[i] != ego->cldrn[i-1] && (i <= 1 || ego->cldrn[i] != ego->cldrn[i-2]))) p->print(p, "%(%p%)", ego->cldrn[i]); p->putchr(p, ')'); } static int pickdim(const S *ego, const tensor *vecsz, int oop, int *dp) { return X(pickdim)(ego->vecloop_dim, ego->buddies, ego->nbuddies, vecsz, oop, dp); } static int applicable0(const solver *ego_, const problem *p_, const planner *plnr, int *dp) { const S *ego = (const S *) ego_; const problem_rdft2 *p = (const problem_rdft2 *) p_; if (FINITE_RNK(p->vecsz->rnk) && p->vecsz->rnk > 0 && plnr->nthr > 1 && pickdim(ego, p->vecsz, p->r0 != p->cr, dp)) { if (p->r0 != p->cr) return 1; /* can always operate out-of-place */ return(X(rdft2_inplace_strides)(p, *dp)); } return 0; } static int applicable(const solver *ego_, const problem *p_, const planner *plnr, int *dp) { const S *ego = (const S *)ego_; if (!applicable0(ego_, p_, plnr, dp)) return 0; /* fftw2 behavior */ if (NO_VRANK_SPLITSP(plnr) && (ego->vecloop_dim != ego->buddies[0])) return 0; return 1; } static plan *mkplan(const solver *ego_, const problem *p_, planner *plnr) { const S *ego = (const S *) ego_; const problem_rdft2 *p; P *pln; problem *cldp; int vdim; iodim *d; plan **cldrn = (plan **) 0; int i, nthr; INT its, ots, block_size; tensor *vecsz; static const plan_adt padt = { X(rdft2_solve), awake, print, destroy }; if (!applicable(ego_, p_, plnr, &vdim)) return (plan *) 0; p = (const problem_rdft2 *) p_; d = p->vecsz->dims + vdim; block_size = (d->n + plnr->nthr - 1) / plnr->nthr; nthr = (int)((d->n + block_size - 1) / block_size); plnr->nthr = (plnr->nthr + nthr - 1) / nthr; X(rdft2_strides)(p->kind, d, &its, &ots); its *= block_size; ots *= block_size; cldrn = (plan **)MALLOC(sizeof(plan *) * nthr, PLANS); for (i = 0; i < nthr; ++i) cldrn[i] = (plan *) 0; vecsz = X(tensor_copy)(p->vecsz); for (i = 0; i < nthr; ++i) { vecsz->dims[vdim].n = (i == nthr - 1) ? (d->n - i*block_size) : block_size; cldp = X(mkproblem_rdft2)(p->sz, vecsz, p->r0 + i*its, p->r1 + i*its, p->cr + i*ots, p->ci + i*ots, p->kind); cldrn[i] = X(mkplan_d)(plnr, cldp); if (!cldrn[i]) goto nada; } X(tensor_destroy)(vecsz); pln = MKPLAN_RDFT2(P, &padt, apply); pln->cldrn = cldrn; pln->its = its; pln->ots = ots; pln->nthr = nthr; pln->solver = ego; X(ops_zero)(&pln->super.super.ops); pln->super.super.pcost = 0; for (i = 0; i < nthr; ++i) { X(ops_add2)(&cldrn[i]->ops, &pln->super.super.ops); pln->super.super.pcost += cldrn[i]->pcost; } return &(pln->super.super); nada: if (cldrn) { for (i = 0; i < nthr; ++i) X(plan_destroy_internal)(cldrn[i]); X(ifree)(cldrn); } X(tensor_destroy)(vecsz); return (plan *) 0; } static solver *mksolver(int vecloop_dim, const int *buddies, int nbuddies) { static const solver_adt sadt = { PROBLEM_RDFT2, mkplan, 0 }; S *slv = MKSOLVER(S, &sadt); slv->vecloop_dim = vecloop_dim; slv->buddies = buddies; slv->nbuddies = nbuddies; return &(slv->super); } void X(rdft2_thr_vrank_geq1_register)(planner *p) { int i; /* FIXME: Should we try other vecloop_dim values? */ static const int buddies[] = { 1, -1 }; const int nbuddies = (int)(sizeof(buddies) / sizeof(buddies[0])); for (i = 0; i < nbuddies; ++i) REGISTER_SOLVER(p, mksolver(buddies[i], buddies, nbuddies)); } fftw-3.3.3/threads/f77api.c0000644000175000017500000000450412054410050012274 00000000000000/* * Copyright (c) 2003, 2007-11 Matteo Frigo * Copyright (c) 2003, 2007-11 Massachusetts Institute of Technology * * This program is free software; you can redistribute it and/or modify * it under the terms of the GNU General Public License as published by * the Free Software Foundation; either version 2 of the License, or * (at your option) any later version. * * This program is distributed in the hope that it will be useful, * but WITHOUT ANY WARRANTY; without even the implied warranty of * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the * GNU General Public License for more details. * * You should have received a copy of the GNU General Public License * along with this program; if not, write to the Free Software * Foundation, Inc., 51 Franklin Street, Fifth Floor, Boston, MA 02110-1301 USA * */ #include "api.h" /* if F77_FUNC is not defined, then we don't know how to mangle identifiers for the Fortran linker, and we must omit the f77 API. */ #if defined(F77_FUNC) || defined(WINDOWS_F77_MANGLING) #include "x77.h" #define F77(a, A) F77x(x77(a), X77(A)) #ifndef WINDOWS_F77_MANGLING #if defined(F77_FUNC) # define F77x(a, A) F77_FUNC(a, A) # include "f77funcs.h" #endif #if defined(F77_FUNC_) && !defined(F77_FUNC_EQUIV) # undef F77x # define F77x(a, A) F77_FUNC_(a, A) # include "f77funcs.h" #endif #else /* WINDOWS_F77_MANGLING */ /* Various mangling conventions common (?) under Windows. */ /* g77 */ # define WINDOWS_F77_FUNC(a, A) a ## __ # define F77x(a, A) WINDOWS_F77_FUNC(a, A) # include "f77funcs.h" /* Intel, etc. */ # undef WINDOWS_F77_FUNC # define WINDOWS_F77_FUNC(a, A) a ## _ # include "f77funcs.h" /* Digital/Compaq/HP Visual Fortran, Intel Fortran. stdcall attribute is apparently required to adjust for calling conventions (callee pops stack in stdcall). See also: http://msdn.microsoft.com/library/en-us/vccore98/html/_core_mixed.2d.language_programming.3a_.overview.asp */ # undef WINDOWS_F77_FUNC # if defined(__GNUC__) # define WINDOWS_F77_FUNC(a, A) __attribute__((stdcall)) A # elif defined(_MSC_VER) || defined(_ICC) || defined(_STDCALL_SUPPORTED) # define WINDOWS_F77_FUNC(a, A) __stdcall A # else # define WINDOWS_F77_FUNC(a, A) A /* oh well */ # endif # include "f77funcs.h" #endif /* WINDOWS_F77_MANGLING */ #endif /* F77_FUNC */ fftw-3.3.3/threads/f77funcs.h0000644000175000017500000000255412054410050012651 00000000000000/* * Copyright (c) 2003, 2007-11 Matteo Frigo * Copyright (c) 2003, 2007-11 Massachusetts Institute of Technology * * This program is free software; you can redistribute it and/or modify * it under the terms of the GNU General Public License as published by * the Free Software Foundation; either version 2 of the License, or * (at your option) any later version. * * This program is distributed in the hope that it will be useful, * but WITHOUT ANY WARRANTY; without even the implied warranty of * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the * GNU General Public License for more details. * * You should have received a copy of the GNU General Public License * along with this program; if not, write to the Free Software * Foundation, Inc., 51 Franklin Street, Fifth Floor, Boston, MA 02110-1301 USA * */ /* Functions in the FFTW Fortran API, mangled according to the F77(...) macro. This file is designed to be #included by f77api.c, possibly multiple times in order to support multiple compiler manglings (via redefinition of F77). */ FFTW_VOIDFUNC F77(plan_with_nthreads, PLAN_WITH_NTHREADS)(int *nthreads) { X(plan_with_nthreads)(*nthreads); } FFTW_VOIDFUNC F77(init_threads, INIT_THREADS)(int *okay) { *okay = X(init_threads)(); } FFTW_VOIDFUNC F77(cleanup_threads, CLEANUP_THREADS)(void) { X(cleanup_threads)(); } fftw-3.3.3/threads/threads.c0000644000175000017500000002503612054410050012634 00000000000000/* * Copyright (c) 2003, 2007-11 Matteo Frigo * Copyright (c) 2003, 2007-11 Massachusetts Institute of Technology * * This program is free software; you can redistribute it and/or modify * it under the terms of the GNU General Public License as published by * the Free Software Foundation; either version 2 of the License, or * (at your option) any later version. * * This program is distributed in the hope that it will be useful, * but WITHOUT ANY WARRANTY; without even the implied warranty of * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the * GNU General Public License for more details. * * You should have received a copy of the GNU General Public License * along with this program; if not, write to the Free Software * Foundation, Inc., 51 Franklin Street, Fifth Floor, Boston, MA 02110-1301 USA * */ /* threads.c: Portable thread spawning for loops, via the X(spawn_loop) function. The first portion of this file is a set of macros to spawn and join threads on various systems. */ #include "threads.h" #if defined(USING_POSIX_THREADS) #include #ifdef HAVE_UNISTD_H # include #endif /* imlementation of semaphores and mutexes: */ #if (defined(_POSIX_SEMAPHORES) && (_POSIX_SEMAPHORES >= 200112L)) /* If optional POSIX semaphores are supported, use them to implement both semaphores and mutexes. */ # include # include typedef sem_t os_sem_t; static void os_sem_init(os_sem_t *s) { sem_init(s, 0, 0); } static void os_sem_destroy(os_sem_t *s) { sem_destroy(s); } static void os_sem_down(os_sem_t *s) { int err; do { err = sem_wait(s); } while (err == -1 && errno == EINTR); CK(err == 0); } static void os_sem_up(os_sem_t *s) { sem_post(s); } /* The reason why we use sem_t to implement mutexes is that I have seen mysterious hangs with glibc-2.7 and linux-2.6.22 when using pthread_mutex_t, but no hangs with sem_t or with linux >= 2.6.24. For lack of better information, sem_t looks like the safest choice. */ typedef sem_t os_mutex_t; static void os_mutex_init(os_mutex_t *s) { sem_init(s, 0, 1); } #define os_mutex_destroy os_sem_destroy #define os_mutex_lock os_sem_down #define os_mutex_unlock os_sem_up #else /* If optional POSIX semaphores are not defined, use pthread mutexes for mutexes, and simulate semaphores with condition variables */ typedef pthread_mutex_t os_mutex_t; static void os_mutex_init(os_mutex_t *s) { pthread_mutex_init(s, (pthread_mutexattr_t *)0); } static void os_mutex_destroy(os_mutex_t *s) { pthread_mutex_destroy(s); } static void os_mutex_lock(os_mutex_t *s) { pthread_mutex_lock(s); } static void os_mutex_unlock(os_mutex_t *s) { pthread_mutex_unlock(s); } typedef struct { pthread_mutex_t m; pthread_cond_t c; volatile int x; } os_sem_t; static void os_sem_init(os_sem_t *s) { pthread_mutex_init(&s->m, (pthread_mutexattr_t *)0); pthread_cond_init(&s->c, (pthread_condattr_t *)0); /* wrap initialization in lock to exploit the release semantics of pthread_mutex_unlock() */ pthread_mutex_lock(&s->m); s->x = 0; pthread_mutex_unlock(&s->m); } static void os_sem_destroy(os_sem_t *s) { pthread_mutex_destroy(&s->m); pthread_cond_destroy(&s->c); } static void os_sem_down(os_sem_t *s) { pthread_mutex_lock(&s->m); while (s->x <= 0) pthread_cond_wait(&s->c, &s->m); --s->x; pthread_mutex_unlock(&s->m); } static void os_sem_up(os_sem_t *s) { pthread_mutex_lock(&s->m); ++s->x; pthread_cond_signal(&s->c); pthread_mutex_unlock(&s->m); } #endif #define FFTW_WORKER void * static void os_create_thread(FFTW_WORKER (*worker)(void *arg), void *arg) { pthread_attr_t attr; pthread_t tid; pthread_attr_init(&attr); pthread_attr_setscope(&attr, PTHREAD_SCOPE_SYSTEM); pthread_attr_setdetachstate(&attr, PTHREAD_CREATE_DETACHED); pthread_create(&tid, &attr, worker, (void *)arg); pthread_attr_destroy(&attr); } static void os_destroy_thread(void) { pthread_exit((void *)0); } #elif defined(__WIN32__) || defined(_WIN32) || defined(_WINDOWS) /* hack: windef.h defines INT for its own purposes and this causes a conflict with our own INT in ifftw.h. Divert the windows definition into another name unlikely to cause a conflict */ #define INT magnus_ab_INTegro_seclorum_nascitur_ordo #include #include #undef INT typedef HANDLE os_mutex_t; static void os_mutex_init(os_mutex_t *s) { *s = CreateMutex(NULL, FALSE, NULL); } static void os_mutex_destroy(os_mutex_t *s) { CloseHandle(*s); } static void os_mutex_lock(os_mutex_t *s) { WaitForSingleObject(*s, INFINITE); } static void os_mutex_unlock(os_mutex_t *s) { ReleaseMutex(*s); } typedef HANDLE os_sem_t; static void os_sem_init(os_sem_t *s) { *s = CreateSemaphore(NULL, 0, 0x7FFFFFFFL, NULL); } static void os_sem_destroy(os_sem_t *s) { CloseHandle(*s); } static void os_sem_down(os_sem_t *s) { WaitForSingleObject(*s, INFINITE); } static void os_sem_up(os_sem_t *s) { ReleaseSemaphore(*s, 1, NULL); } #define FFTW_WORKER unsigned __stdcall typedef unsigned (__stdcall *winthread_start) (void *); static void os_create_thread(winthread_start worker, void *arg) { _beginthreadex((void *)NULL, /* security attrib */ 0, /* stack size */ worker, /* start address */ arg, /* parameters */ 0, /* creation flags */ (unsigned *)NULL); /* tid */ } static void os_destroy_thread(void) { _endthreadex(0); } #else #error "No threading layer defined" #endif /************************************************************************/ /* Main code: */ struct worker { os_sem_t ready; os_sem_t done; struct work *w; struct worker *cdr; }; static struct worker *make_worker(void) { struct worker *q = (struct worker *)MALLOC(sizeof(*q), OTHER); os_sem_init(&q->ready); os_sem_init(&q->done); return q; } static void unmake_worker(struct worker *q) { os_sem_destroy(&q->done); os_sem_destroy(&q->ready); X(ifree)(q); } struct work { spawn_function proc; spawn_data d; struct worker *q; /* the worker responsible for performing this work */ }; static os_mutex_t queue_lock; static os_sem_t termination_semaphore; static struct worker *worker_queue; #define WITH_QUEUE_LOCK(what) \ { \ os_mutex_lock(&queue_lock); \ what; \ os_mutex_unlock(&queue_lock); \ } static FFTW_WORKER worker(void *arg) { struct worker *ego = (struct worker *)arg; struct work *w; for (;;) { /* wait until work becomes available */ os_sem_down(&ego->ready); w = ego->w; /* !w->proc ==> terminate worker */ if (!w->proc) break; /* do the work */ w->proc(&w->d); /* signal that work is done */ os_sem_up(&ego->done); } /* termination protocol */ os_sem_up(&termination_semaphore); os_destroy_thread(); /* UNREACHABLE */ return 0; } static void enqueue(struct worker *q) { WITH_QUEUE_LOCK({ q->cdr = worker_queue; worker_queue = q; }); } static struct worker *dequeue(void) { struct worker *q; WITH_QUEUE_LOCK({ q = worker_queue; if (q) worker_queue = q->cdr; }); if (!q) { /* no worker is available. Create one */ q = make_worker(); os_create_thread(worker, q); } return q; } static void kill_workforce(void) { struct work w; w.proc = 0; THREAD_ON; /* needed for debugging mode: since make_worker is called from dequeue which is only called in thread_on mode, we need to unmake_worker in thread_on. */ WITH_QUEUE_LOCK({ /* tell all workers that they must terminate. Because workers enqueue themselves before signaling the completion of the work, all workers belong to the worker queue if we get here. Also, all workers are waiting at os_sem_down(ready), so we can hold the queue lock without deadlocking */ while (worker_queue) { struct worker *q = worker_queue; worker_queue = q->cdr; q->w = &w; os_sem_up(&q->ready); os_sem_down(&termination_semaphore); unmake_worker(q); } }); THREAD_OFF; } int X(ithreads_init)(void) { os_mutex_init(&queue_lock); os_sem_init(&termination_semaphore); WITH_QUEUE_LOCK({ worker_queue = 0; }) return 0; /* no error */ } /* Distribute a loop from 0 to loopmax-1 over nthreads threads. proc(d) is called to execute a block of iterations from d->min to d->max-1. d->thr_num indicate the number of the thread that is executing proc (from 0 to nthreads-1), and d->data is the same as the data parameter passed to X(spawn_loop). This function returns only after all the threads have completed. */ void X(spawn_loop)(int loopmax, int nthr, spawn_function proc, void *data) { int block_size; struct work *r; int i; A(loopmax >= 0); A(nthr > 0); A(proc); if (!loopmax) return; /* Choose the block size and number of threads in order to (1) minimize the critical path and (2) use the fewest threads that achieve the same critical path (to minimize overhead). e.g. if loopmax is 5 and nthr is 4, we should use only 3 threads with block sizes of 2, 2, and 1. */ block_size = (loopmax + nthr - 1) / nthr; nthr = (loopmax + block_size - 1) / block_size; THREAD_ON; /* prevent debugging mode from failing under threads */ STACK_MALLOC(struct work *, r, sizeof(struct work) * nthr); /* distribute work: */ for (i = 0; i < nthr; ++i) { struct work *w = &r[i]; spawn_data *d = &w->d; d->max = (d->min = i * block_size) + block_size; if (d->max > loopmax) d->max = loopmax; d->thr_num = i; d->data = data; w->proc = proc; if (i == nthr - 1) { /* do the work ourselves */ proc(d); } else { /* assign a worker to W */ w->q = dequeue(); /* tell worker w->q to do it */ w->q->w = w; /* Dirac could have written this */ os_sem_up(&w->q->ready); } } for (i = 0; i < nthr - 1; ++i) { struct work *w = &r[i]; os_sem_down(&w->q->done); enqueue(w->q); } STACK_FREE(r); THREAD_OFF; /* prevent debugging mode from failing under threads */ } void X(threads_cleanup)(void) { kill_workforce(); os_mutex_destroy(&queue_lock); os_sem_destroy(&termination_semaphore); } fftw-3.3.3/tests/0000755000175000017500000000000012054411110010616 500000000000000fftw-3.3.3/tests/README0000644000175000017500000000372612054410050011430 00000000000000This directory contains a benchmarking and testing program for fftw3. The `bench' program has a zillion options, because we use it for benchmarking other FFT libraries as well. This file only documents the basic usage of bench. Usage: bench where each command is as follows: -s --speed Benchmarks the speed of . The syntax for problems is [i|o][r|c][f|b], where i/o means in-place or out-of-place. Out of place is the default. r/c means real or complex transform. Complex is the default. f/b means forward or backward transform. Forward is the default. is an arbitrary multidimensional sequence of integers separated by the character 'x'. (The syntax for problems is actually richer, but we do not document it here. See the man page for fftw-wisdom for more information.) Example: ib256 : in-place backward complex transform of size 256 32x64 : out-of-place forward complex 2D transform of 32 rows and 64 columns. -y --verify Verify that FFTW is computing the correct answer. The program does not output anything unless an error occurs or verbosity is at least one. -v Set verbosity to , or 1 if is omitted. -v2 will output the created plans with fftw_print_plan. -oestimate -opatient -oexhaustive Plan with FFTW_ESTIMATE, FFTW_PATIENT, or FFTW_EXHAUSTIVE, respectively. The default is FFTW_MEASURE. If you benchmark FFTW, please use -opatient. -onthreads=N Use N threads, if FFTW was compiled with --enable-threads. N must be a positive integer; the default is N=1. -onosimd Disable SIMD instructions (e.g. SSE or SSE2). -ounaligned Plan with the FFTW_UNALIGNED flag. -owisdom On startup, read wisdom from a file wis.dat in the current directory (if it exists). On completion, write accumulated wisdom to wis.dat (overwriting any existing file of that name). fftw-3.3.3/tests/Makefile.am0000644000175000017500000000667112054410050012606 00000000000000AM_CPPFLAGS = -I$(top_srcdir)/kernel -I$(top_srcdir)/libbench2 \ -I$(top_srcdir)/dft -I$(top_srcdir)/rdft -I$(top_srcdir)/reodft \ -I$(top_srcdir)/threads -I$(top_srcdir)/api noinst_PROGRAMS = bench EXTRA_DIST = check.pl README if THREADS bench_CFLAGS = $(PTHREAD_CFLAGS) if !COMBINED_THREADS LIBFFTWTHREADS = $(top_builddir)/threads/libfftw3@PREC_SUFFIX@_threads.la endif else if OPENMP bench_CFLAGS = $(OPENMP_CFLAGS) LIBFFTWTHREADS = $(top_builddir)/threads/libfftw3@PREC_SUFFIX@_omp.la endif endif bench_SOURCES = bench.c hook.c fftw-bench.c fftw-bench.h bench_LDADD = $(LIBFFTWTHREADS) \ $(top_builddir)/libfftw3@PREC_SUFFIX@.la \ $(top_builddir)/libbench2/libbench2.a $(THREADLIBS) check-local: bench$(EXEEXT) perl -w $(srcdir)/check.pl $(CHECK_PL_OPTS) -r -c=30 -v `pwd`/bench @echo "--------------------------------------------------------------" @echo " FFTW transforms passed basic tests!" @echo "--------------------------------------------------------------" if SMP perl -w $(srcdir)/check.pl $(CHECK_PL_OPTS) -r -c=30 -v --nthreads=2 `pwd`/bench @echo "--------------------------------------------------------------" @echo " FFTW threaded transforms passed basic tests!" @echo "--------------------------------------------------------------" endif bigcheck: bench$(EXEEXT) perl -w $(srcdir)/check.pl $(CHECK_PL_OPTS) -a -v `pwd`/bench @echo "--------------------------------------------------------------" @echo " FFTW transforms passed big tests!" @echo "--------------------------------------------------------------" if SMP perl -w $(srcdir)/check.pl $(CHECK_PL_OPTS) -a -v --nthreads=2 `pwd`/bench perl -w $(srcdir)/check.pl $(CHECK_PL_OPTS) -a -v --nthreads=3 `pwd`/bench perl -w $(srcdir)/check.pl $(CHECK_PL_OPTS) -a -v --nthreads=10 `pwd`/bench @echo "--------------------------------------------------------------" @echo " FFTW threaded transforms passed big tests!" @echo "--------------------------------------------------------------" endif smallcheck: bench$(EXEEXT) perl -w $(srcdir)/check.pl -r -c=1 -v `pwd`/bench perl -w $(srcdir)/check.pl -r --estimate -c=5 -v `pwd`/bench @echo "--------------------------------------------------------------" @echo " FFTW transforms passed a few tests!" @echo "--------------------------------------------------------------" if SMP perl -w $(srcdir)/check.pl -r --estimate -c=2 -v --nthreads=2 `pwd`/bench @echo "--------------------------------------------------------------" @echo " FFTW threaded transforms passed a few tests!" @echo "--------------------------------------------------------------" endif paranoid-check: bench$(EXEEXT) if SMP perl -w $(srcdir)/check.pl -a --patient --nthreads=10 --paranoid `pwd`/bench perl -w $(srcdir)/check.pl -a --patient --nthreads=7 --paranoid `pwd`/bench perl -w $(srcdir)/check.pl -a --patient --nthreads=3 --paranoid `pwd`/bench perl -w $(srcdir)/check.pl -a --patient --nthreads=2 --paranoid `pwd`/bench endif perl -w $(srcdir)/check.pl -a --patient --paranoid `pwd`/bench exhaustive-check: bench$(EXEEXT) if SMP perl -w $(srcdir)/check.pl -a --exhaustive --nthreads=10 --paranoid `pwd`/bench perl -w $(srcdir)/check.pl -a --exhaustive --nthreads=7 --paranoid `pwd`/bench perl -w $(srcdir)/check.pl -a --exhaustive --nthreads=3 --paranoid `pwd`/bench perl -w $(srcdir)/check.pl -a --exhaustive --nthreads=2 --paranoid `pwd`/bench endif perl -w $(srcdir)/check.pl -a --exhaustive --paranoid `pwd`/bench fftw-3.3.3/tests/Makefile.in0000644000175000017500000006066212054410201012615 00000000000000# Makefile.in generated by automake 1.11.6 from Makefile.am. # @configure_input@ # Copyright (C) 1994, 1995, 1996, 1997, 1998, 1999, 2000, 2001, 2002, # 2003, 2004, 2005, 2006, 2007, 2008, 2009, 2010, 2011 Free Software # Foundation, Inc. # This Makefile.in is free software; the Free Software Foundation # gives unlimited permission to copy and/or distribute it, # with or without modifications, as long as this notice is preserved. # This program is distributed in the hope that it will be useful, # but WITHOUT ANY WARRANTY, to the extent permitted by law; without # even the implied warranty of MERCHANTABILITY or FITNESS FOR A # PARTICULAR PURPOSE. @SET_MAKE@ VPATH = @srcdir@ am__make_dryrun = \ { \ am__dry=no; \ case $$MAKEFLAGS in \ *\\[\ \ ]*) \ echo 'am--echo: ; @echo "AM" OK' | $(MAKE) -f - 2>/dev/null \ | grep '^AM OK$$' >/dev/null || am__dry=yes;; \ *) \ for am__flg in $$MAKEFLAGS; do \ case $$am__flg in \ *=*|--*) ;; \ *n*) am__dry=yes; break;; \ esac; \ done;; \ esac; \ test $$am__dry = yes; \ } pkgdatadir = $(datadir)/@PACKAGE@ pkgincludedir = $(includedir)/@PACKAGE@ pkglibdir = $(libdir)/@PACKAGE@ pkglibexecdir = $(libexecdir)/@PACKAGE@ am__cd = CDPATH="$${ZSH_VERSION+.}$(PATH_SEPARATOR)" && cd install_sh_DATA = $(install_sh) -c -m 644 install_sh_PROGRAM = $(install_sh) -c install_sh_SCRIPT = $(install_sh) -c INSTALL_HEADER = $(INSTALL_DATA) transform = $(program_transform_name) NORMAL_INSTALL = : PRE_INSTALL = : POST_INSTALL = : NORMAL_UNINSTALL = : PRE_UNINSTALL = : POST_UNINSTALL = : build_triplet = @build@ host_triplet = @host@ noinst_PROGRAMS = bench$(EXEEXT) subdir = tests DIST_COMMON = README $(srcdir)/Makefile.am $(srcdir)/Makefile.in ACLOCAL_M4 = $(top_srcdir)/aclocal.m4 am__aclocal_m4_deps = $(top_srcdir)/m4/acx_mpi.m4 \ $(top_srcdir)/m4/acx_pthread.m4 \ $(top_srcdir)/m4/ax_cc_maxopt.m4 \ $(top_srcdir)/m4/ax_check_compiler_flags.m4 \ $(top_srcdir)/m4/ax_compiler_vendor.m4 \ $(top_srcdir)/m4/ax_gcc_aligns_stack.m4 \ $(top_srcdir)/m4/ax_gcc_version.m4 \ $(top_srcdir)/m4/ax_openmp.m4 $(top_srcdir)/m4/libtool.m4 \ $(top_srcdir)/m4/ltoptions.m4 $(top_srcdir)/m4/ltsugar.m4 \ $(top_srcdir)/m4/ltversion.m4 $(top_srcdir)/m4/lt~obsolete.m4 \ $(top_srcdir)/configure.ac am__configure_deps = $(am__aclocal_m4_deps) $(CONFIGURE_DEPENDENCIES) \ $(ACLOCAL_M4) mkinstalldirs = $(install_sh) -d CONFIG_HEADER = $(top_builddir)/config.h CONFIG_CLEAN_FILES = CONFIG_CLEAN_VPATH_FILES = PROGRAMS = $(noinst_PROGRAMS) am_bench_OBJECTS = bench-bench.$(OBJEXT) bench-hook.$(OBJEXT) \ bench-fftw-bench.$(OBJEXT) bench_OBJECTS = $(am_bench_OBJECTS) am__DEPENDENCIES_1 = bench_DEPENDENCIES = $(LIBFFTWTHREADS) \ $(top_builddir)/libfftw3@PREC_SUFFIX@.la \ $(top_builddir)/libbench2/libbench2.a $(am__DEPENDENCIES_1) bench_LINK = $(LIBTOOL) --tag=CC $(AM_LIBTOOLFLAGS) $(LIBTOOLFLAGS) \ --mode=link $(CCLD) $(bench_CFLAGS) $(CFLAGS) $(AM_LDFLAGS) \ $(LDFLAGS) -o $@ DEFAULT_INCLUDES = -I.@am__isrc@ -I$(top_builddir) depcomp = $(SHELL) $(top_srcdir)/depcomp am__depfiles_maybe = depfiles am__mv = mv -f COMPILE = $(CC) $(DEFS) $(DEFAULT_INCLUDES) $(INCLUDES) $(AM_CPPFLAGS) \ $(CPPFLAGS) $(AM_CFLAGS) $(CFLAGS) LTCOMPILE = $(LIBTOOL) --tag=CC $(AM_LIBTOOLFLAGS) $(LIBTOOLFLAGS) \ --mode=compile $(CC) $(DEFS) $(DEFAULT_INCLUDES) $(INCLUDES) \ $(AM_CPPFLAGS) $(CPPFLAGS) $(AM_CFLAGS) $(CFLAGS) CCLD = $(CC) LINK = $(LIBTOOL) --tag=CC $(AM_LIBTOOLFLAGS) $(LIBTOOLFLAGS) \ --mode=link $(CCLD) $(AM_CFLAGS) $(CFLAGS) $(AM_LDFLAGS) \ $(LDFLAGS) -o $@ SOURCES = $(bench_SOURCES) DIST_SOURCES = $(bench_SOURCES) am__can_run_installinfo = \ case $$AM_UPDATE_INFO_DIR in \ n|no|NO) false;; \ *) (install-info --version) >/dev/null 2>&1;; \ esac ETAGS = etags CTAGS = ctags DISTFILES = $(DIST_COMMON) $(DIST_SOURCES) $(TEXINFOS) $(EXTRA_DIST) ACLOCAL = @ACLOCAL@ ALLOCA = @ALLOCA@ ALTIVEC_CFLAGS = @ALTIVEC_CFLAGS@ AMTAR = @AMTAR@ AR = @AR@ AS = @AS@ AUTOCONF = @AUTOCONF@ AUTOHEADER = @AUTOHEADER@ AUTOMAKE = @AUTOMAKE@ AVX_CFLAGS = @AVX_CFLAGS@ AWK = @AWK@ CC = @CC@ CCDEPMODE = @CCDEPMODE@ CFLAGS = @CFLAGS@ CHECK_PL_OPTS = @CHECK_PL_OPTS@ CPP = @CPP@ CPPFLAGS = @CPPFLAGS@ CYGPATH_W = @CYGPATH_W@ C_FFTW_R2R_KIND = @C_FFTW_R2R_KIND@ C_MPI_FINT = @C_MPI_FINT@ DEFS = @DEFS@ DEPDIR = @DEPDIR@ DLLTOOL = @DLLTOOL@ DSYMUTIL = @DSYMUTIL@ DUMPBIN = @DUMPBIN@ ECHO_C = @ECHO_C@ ECHO_N = @ECHO_N@ ECHO_T = @ECHO_T@ EGREP = @EGREP@ EXEEXT = @EXEEXT@ F77 = @F77@ FFLAGS = @FFLAGS@ FGREP = @FGREP@ FLIBS = @FLIBS@ GREP = @GREP@ INSTALL = @INSTALL@ INSTALL_DATA = @INSTALL_DATA@ INSTALL_PROGRAM = @INSTALL_PROGRAM@ INSTALL_SCRIPT = @INSTALL_SCRIPT@ INSTALL_STRIP_PROGRAM = @INSTALL_STRIP_PROGRAM@ LD = @LD@ LDFLAGS = @LDFLAGS@ LIBOBJS = @LIBOBJS@ LIBQUADMATH = @LIBQUADMATH@ LIBS = @LIBS@ LIBTOOL = @LIBTOOL@ LIPO = @LIPO@ LN_S = @LN_S@ LTLIBOBJS = @LTLIBOBJS@ MAINT = @MAINT@ MAKEINFO = @MAKEINFO@ MANIFEST_TOOL = @MANIFEST_TOOL@ MKDIR_P = @MKDIR_P@ MPICC = @MPICC@ MPILIBS = @MPILIBS@ MPIRUN = @MPIRUN@ NEON_CFLAGS = @NEON_CFLAGS@ NM = @NM@ NMEDIT = @NMEDIT@ OBJDUMP = @OBJDUMP@ OBJEXT = @OBJEXT@ OCAMLBUILD = @OCAMLBUILD@ OPENMP_CFLAGS = @OPENMP_CFLAGS@ OTOOL = @OTOOL@ OTOOL64 = @OTOOL64@ PACKAGE = @PACKAGE@ PACKAGE_BUGREPORT = @PACKAGE_BUGREPORT@ PACKAGE_NAME = @PACKAGE_NAME@ PACKAGE_STRING = @PACKAGE_STRING@ PACKAGE_TARNAME = @PACKAGE_TARNAME@ PACKAGE_URL = @PACKAGE_URL@ PACKAGE_VERSION = @PACKAGE_VERSION@ PATH_SEPARATOR = @PATH_SEPARATOR@ POW_LIB = @POW_LIB@ PRECISION = @PRECISION@ PREC_SUFFIX = @PREC_SUFFIX@ PTHREAD_CC = @PTHREAD_CC@ PTHREAD_CFLAGS = @PTHREAD_CFLAGS@ PTHREAD_LIBS = @PTHREAD_LIBS@ RANLIB = @RANLIB@ SED = @SED@ SET_MAKE = @SET_MAKE@ SHARED_VERSION_INFO = @SHARED_VERSION_INFO@ SHELL = @SHELL@ SSE2_CFLAGS = @SSE2_CFLAGS@ STACK_ALIGN_CFLAGS = @STACK_ALIGN_CFLAGS@ STRIP = @STRIP@ THREADLIBS = @THREADLIBS@ VERSION = @VERSION@ abs_builddir = @abs_builddir@ abs_srcdir = @abs_srcdir@ abs_top_builddir = @abs_top_builddir@ abs_top_srcdir = @abs_top_srcdir@ ac_ct_AR = @ac_ct_AR@ ac_ct_CC = @ac_ct_CC@ ac_ct_DUMPBIN = @ac_ct_DUMPBIN@ ac_ct_F77 = @ac_ct_F77@ acx_pthread_config = @acx_pthread_config@ am__include = @am__include@ am__leading_dot = @am__leading_dot@ am__quote = @am__quote@ am__tar = @am__tar@ am__untar = @am__untar@ bindir = @bindir@ build = @build@ build_alias = @build_alias@ build_cpu = @build_cpu@ build_os = @build_os@ build_vendor = @build_vendor@ builddir = @builddir@ datadir = @datadir@ datarootdir = @datarootdir@ docdir = @docdir@ dvidir = @dvidir@ exec_prefix = @exec_prefix@ host = @host@ host_alias = @host_alias@ host_cpu = @host_cpu@ host_os = @host_os@ host_vendor = @host_vendor@ htmldir = @htmldir@ includedir = @includedir@ infodir = @infodir@ install_sh = @install_sh@ libdir = @libdir@ libexecdir = @libexecdir@ localedir = @localedir@ localstatedir = @localstatedir@ mandir = @mandir@ mkdir_p = @mkdir_p@ oldincludedir = @oldincludedir@ pdfdir = @pdfdir@ prefix = @prefix@ program_transform_name = @program_transform_name@ psdir = @psdir@ sbindir = @sbindir@ sharedstatedir = @sharedstatedir@ srcdir = @srcdir@ sysconfdir = @sysconfdir@ target_alias = @target_alias@ top_build_prefix = @top_build_prefix@ top_builddir = @top_builddir@ top_srcdir = @top_srcdir@ AM_CPPFLAGS = -I$(top_srcdir)/kernel -I$(top_srcdir)/libbench2 \ -I$(top_srcdir)/dft -I$(top_srcdir)/rdft -I$(top_srcdir)/reodft \ -I$(top_srcdir)/threads -I$(top_srcdir)/api EXTRA_DIST = check.pl README @OPENMP_TRUE@@THREADS_FALSE@bench_CFLAGS = $(OPENMP_CFLAGS) @THREADS_TRUE@bench_CFLAGS = $(PTHREAD_CFLAGS) @COMBINED_THREADS_FALSE@@THREADS_TRUE@LIBFFTWTHREADS = $(top_builddir)/threads/libfftw3@PREC_SUFFIX@_threads.la @OPENMP_TRUE@@THREADS_FALSE@LIBFFTWTHREADS = $(top_builddir)/threads/libfftw3@PREC_SUFFIX@_omp.la bench_SOURCES = bench.c hook.c fftw-bench.c fftw-bench.h bench_LDADD = $(LIBFFTWTHREADS) \ $(top_builddir)/libfftw3@PREC_SUFFIX@.la \ $(top_builddir)/libbench2/libbench2.a $(THREADLIBS) all: all-am .SUFFIXES: .SUFFIXES: .c .lo .o .obj $(srcdir)/Makefile.in: @MAINTAINER_MODE_TRUE@ $(srcdir)/Makefile.am $(am__configure_deps) @for dep in $?; do \ case '$(am__configure_deps)' in \ *$$dep*) \ ( cd $(top_builddir) && $(MAKE) $(AM_MAKEFLAGS) am--refresh ) \ && { if test -f $@; then exit 0; else break; fi; }; \ exit 1;; \ esac; \ done; \ echo ' cd $(top_srcdir) && $(AUTOMAKE) --gnu tests/Makefile'; \ $(am__cd) $(top_srcdir) && \ $(AUTOMAKE) --gnu tests/Makefile .PRECIOUS: Makefile Makefile: $(srcdir)/Makefile.in $(top_builddir)/config.status @case '$?' in \ *config.status*) \ cd $(top_builddir) && $(MAKE) $(AM_MAKEFLAGS) am--refresh;; \ *) \ echo ' cd $(top_builddir) && $(SHELL) ./config.status $(subdir)/$@ $(am__depfiles_maybe)'; \ cd $(top_builddir) && $(SHELL) ./config.status $(subdir)/$@ $(am__depfiles_maybe);; \ esac; $(top_builddir)/config.status: $(top_srcdir)/configure $(CONFIG_STATUS_DEPENDENCIES) cd $(top_builddir) && $(MAKE) $(AM_MAKEFLAGS) am--refresh $(top_srcdir)/configure: @MAINTAINER_MODE_TRUE@ $(am__configure_deps) cd $(top_builddir) && $(MAKE) $(AM_MAKEFLAGS) am--refresh $(ACLOCAL_M4): @MAINTAINER_MODE_TRUE@ $(am__aclocal_m4_deps) cd $(top_builddir) && $(MAKE) $(AM_MAKEFLAGS) am--refresh $(am__aclocal_m4_deps): clean-noinstPROGRAMS: @list='$(noinst_PROGRAMS)'; test -n "$$list" || exit 0; \ echo " rm -f" $$list; \ rm -f $$list || exit $$?; \ test -n "$(EXEEXT)" || exit 0; \ list=`for p in $$list; do echo "$$p"; done | sed 's/$(EXEEXT)$$//'`; \ echo " rm -f" $$list; \ rm -f $$list bench$(EXEEXT): $(bench_OBJECTS) $(bench_DEPENDENCIES) $(EXTRA_bench_DEPENDENCIES) @rm -f bench$(EXEEXT) $(bench_LINK) $(bench_OBJECTS) $(bench_LDADD) $(LIBS) mostlyclean-compile: -rm -f *.$(OBJEXT) distclean-compile: -rm -f *.tab.c @AMDEP_TRUE@@am__include@ @am__quote@./$(DEPDIR)/bench-bench.Po@am__quote@ @AMDEP_TRUE@@am__include@ @am__quote@./$(DEPDIR)/bench-fftw-bench.Po@am__quote@ @AMDEP_TRUE@@am__include@ @am__quote@./$(DEPDIR)/bench-hook.Po@am__quote@ .c.o: @am__fastdepCC_TRUE@ $(COMPILE) -MT $@ -MD -MP -MF $(DEPDIR)/$*.Tpo -c -o $@ $< @am__fastdepCC_TRUE@ $(am__mv) $(DEPDIR)/$*.Tpo $(DEPDIR)/$*.Po @AMDEP_TRUE@@am__fastdepCC_FALSE@ source='$<' object='$@' libtool=no @AMDEPBACKSLASH@ @AMDEP_TRUE@@am__fastdepCC_FALSE@ DEPDIR=$(DEPDIR) $(CCDEPMODE) $(depcomp) @AMDEPBACKSLASH@ @am__fastdepCC_FALSE@ $(COMPILE) -c $< .c.obj: @am__fastdepCC_TRUE@ $(COMPILE) -MT $@ -MD -MP -MF $(DEPDIR)/$*.Tpo -c -o $@ `$(CYGPATH_W) '$<'` @am__fastdepCC_TRUE@ $(am__mv) $(DEPDIR)/$*.Tpo $(DEPDIR)/$*.Po @AMDEP_TRUE@@am__fastdepCC_FALSE@ source='$<' object='$@' libtool=no @AMDEPBACKSLASH@ @AMDEP_TRUE@@am__fastdepCC_FALSE@ DEPDIR=$(DEPDIR) $(CCDEPMODE) $(depcomp) @AMDEPBACKSLASH@ @am__fastdepCC_FALSE@ $(COMPILE) -c `$(CYGPATH_W) '$<'` .c.lo: @am__fastdepCC_TRUE@ $(LTCOMPILE) -MT $@ -MD -MP -MF $(DEPDIR)/$*.Tpo -c -o $@ $< @am__fastdepCC_TRUE@ $(am__mv) $(DEPDIR)/$*.Tpo $(DEPDIR)/$*.Plo @AMDEP_TRUE@@am__fastdepCC_FALSE@ source='$<' object='$@' libtool=yes @AMDEPBACKSLASH@ @AMDEP_TRUE@@am__fastdepCC_FALSE@ DEPDIR=$(DEPDIR) $(CCDEPMODE) $(depcomp) @AMDEPBACKSLASH@ @am__fastdepCC_FALSE@ $(LTCOMPILE) -c -o $@ $< bench-bench.o: bench.c @am__fastdepCC_TRUE@ $(CC) $(DEFS) $(DEFAULT_INCLUDES) $(INCLUDES) $(AM_CPPFLAGS) $(CPPFLAGS) $(bench_CFLAGS) $(CFLAGS) -MT bench-bench.o -MD -MP -MF $(DEPDIR)/bench-bench.Tpo -c -o bench-bench.o `test -f 'bench.c' || echo '$(srcdir)/'`bench.c @am__fastdepCC_TRUE@ $(am__mv) $(DEPDIR)/bench-bench.Tpo $(DEPDIR)/bench-bench.Po @AMDEP_TRUE@@am__fastdepCC_FALSE@ source='bench.c' object='bench-bench.o' libtool=no @AMDEPBACKSLASH@ @AMDEP_TRUE@@am__fastdepCC_FALSE@ DEPDIR=$(DEPDIR) $(CCDEPMODE) $(depcomp) @AMDEPBACKSLASH@ @am__fastdepCC_FALSE@ $(CC) $(DEFS) $(DEFAULT_INCLUDES) $(INCLUDES) $(AM_CPPFLAGS) $(CPPFLAGS) $(bench_CFLAGS) $(CFLAGS) -c -o bench-bench.o `test -f 'bench.c' || echo '$(srcdir)/'`bench.c bench-bench.obj: bench.c @am__fastdepCC_TRUE@ $(CC) $(DEFS) $(DEFAULT_INCLUDES) $(INCLUDES) $(AM_CPPFLAGS) $(CPPFLAGS) $(bench_CFLAGS) $(CFLAGS) -MT bench-bench.obj -MD -MP -MF $(DEPDIR)/bench-bench.Tpo -c -o bench-bench.obj `if test -f 'bench.c'; then $(CYGPATH_W) 'bench.c'; else $(CYGPATH_W) '$(srcdir)/bench.c'; fi` @am__fastdepCC_TRUE@ $(am__mv) $(DEPDIR)/bench-bench.Tpo $(DEPDIR)/bench-bench.Po @AMDEP_TRUE@@am__fastdepCC_FALSE@ source='bench.c' object='bench-bench.obj' libtool=no @AMDEPBACKSLASH@ @AMDEP_TRUE@@am__fastdepCC_FALSE@ DEPDIR=$(DEPDIR) $(CCDEPMODE) $(depcomp) @AMDEPBACKSLASH@ @am__fastdepCC_FALSE@ $(CC) $(DEFS) $(DEFAULT_INCLUDES) $(INCLUDES) $(AM_CPPFLAGS) $(CPPFLAGS) $(bench_CFLAGS) $(CFLAGS) -c -o bench-bench.obj `if test -f 'bench.c'; then $(CYGPATH_W) 'bench.c'; else $(CYGPATH_W) '$(srcdir)/bench.c'; fi` bench-hook.o: hook.c @am__fastdepCC_TRUE@ $(CC) $(DEFS) $(DEFAULT_INCLUDES) $(INCLUDES) $(AM_CPPFLAGS) $(CPPFLAGS) $(bench_CFLAGS) $(CFLAGS) -MT bench-hook.o -MD -MP -MF $(DEPDIR)/bench-hook.Tpo -c -o bench-hook.o `test -f 'hook.c' || echo '$(srcdir)/'`hook.c @am__fastdepCC_TRUE@ $(am__mv) $(DEPDIR)/bench-hook.Tpo $(DEPDIR)/bench-hook.Po @AMDEP_TRUE@@am__fastdepCC_FALSE@ source='hook.c' object='bench-hook.o' libtool=no @AMDEPBACKSLASH@ @AMDEP_TRUE@@am__fastdepCC_FALSE@ DEPDIR=$(DEPDIR) $(CCDEPMODE) $(depcomp) @AMDEPBACKSLASH@ @am__fastdepCC_FALSE@ $(CC) $(DEFS) $(DEFAULT_INCLUDES) $(INCLUDES) $(AM_CPPFLAGS) $(CPPFLAGS) $(bench_CFLAGS) $(CFLAGS) -c -o bench-hook.o `test -f 'hook.c' || echo '$(srcdir)/'`hook.c bench-hook.obj: hook.c @am__fastdepCC_TRUE@ $(CC) $(DEFS) $(DEFAULT_INCLUDES) $(INCLUDES) $(AM_CPPFLAGS) $(CPPFLAGS) $(bench_CFLAGS) $(CFLAGS) -MT bench-hook.obj -MD -MP -MF $(DEPDIR)/bench-hook.Tpo -c -o bench-hook.obj `if test -f 'hook.c'; then $(CYGPATH_W) 'hook.c'; else $(CYGPATH_W) '$(srcdir)/hook.c'; fi` @am__fastdepCC_TRUE@ $(am__mv) $(DEPDIR)/bench-hook.Tpo $(DEPDIR)/bench-hook.Po @AMDEP_TRUE@@am__fastdepCC_FALSE@ source='hook.c' object='bench-hook.obj' libtool=no @AMDEPBACKSLASH@ @AMDEP_TRUE@@am__fastdepCC_FALSE@ DEPDIR=$(DEPDIR) $(CCDEPMODE) $(depcomp) @AMDEPBACKSLASH@ @am__fastdepCC_FALSE@ $(CC) $(DEFS) $(DEFAULT_INCLUDES) $(INCLUDES) $(AM_CPPFLAGS) $(CPPFLAGS) $(bench_CFLAGS) $(CFLAGS) -c -o bench-hook.obj `if test -f 'hook.c'; then $(CYGPATH_W) 'hook.c'; else $(CYGPATH_W) '$(srcdir)/hook.c'; fi` bench-fftw-bench.o: fftw-bench.c @am__fastdepCC_TRUE@ $(CC) $(DEFS) $(DEFAULT_INCLUDES) $(INCLUDES) $(AM_CPPFLAGS) $(CPPFLAGS) $(bench_CFLAGS) $(CFLAGS) -MT bench-fftw-bench.o -MD -MP -MF $(DEPDIR)/bench-fftw-bench.Tpo -c -o bench-fftw-bench.o `test -f 'fftw-bench.c' || echo '$(srcdir)/'`fftw-bench.c @am__fastdepCC_TRUE@ $(am__mv) $(DEPDIR)/bench-fftw-bench.Tpo $(DEPDIR)/bench-fftw-bench.Po @AMDEP_TRUE@@am__fastdepCC_FALSE@ source='fftw-bench.c' object='bench-fftw-bench.o' libtool=no @AMDEPBACKSLASH@ @AMDEP_TRUE@@am__fastdepCC_FALSE@ DEPDIR=$(DEPDIR) $(CCDEPMODE) $(depcomp) @AMDEPBACKSLASH@ @am__fastdepCC_FALSE@ $(CC) $(DEFS) $(DEFAULT_INCLUDES) $(INCLUDES) $(AM_CPPFLAGS) $(CPPFLAGS) $(bench_CFLAGS) $(CFLAGS) -c -o bench-fftw-bench.o `test -f 'fftw-bench.c' || echo '$(srcdir)/'`fftw-bench.c bench-fftw-bench.obj: fftw-bench.c @am__fastdepCC_TRUE@ $(CC) $(DEFS) $(DEFAULT_INCLUDES) $(INCLUDES) $(AM_CPPFLAGS) $(CPPFLAGS) $(bench_CFLAGS) $(CFLAGS) -MT bench-fftw-bench.obj -MD -MP -MF $(DEPDIR)/bench-fftw-bench.Tpo -c -o bench-fftw-bench.obj `if test -f 'fftw-bench.c'; then $(CYGPATH_W) 'fftw-bench.c'; else $(CYGPATH_W) '$(srcdir)/fftw-bench.c'; fi` @am__fastdepCC_TRUE@ $(am__mv) $(DEPDIR)/bench-fftw-bench.Tpo $(DEPDIR)/bench-fftw-bench.Po @AMDEP_TRUE@@am__fastdepCC_FALSE@ source='fftw-bench.c' object='bench-fftw-bench.obj' libtool=no @AMDEPBACKSLASH@ @AMDEP_TRUE@@am__fastdepCC_FALSE@ DEPDIR=$(DEPDIR) $(CCDEPMODE) $(depcomp) @AMDEPBACKSLASH@ @am__fastdepCC_FALSE@ $(CC) $(DEFS) $(DEFAULT_INCLUDES) $(INCLUDES) $(AM_CPPFLAGS) $(CPPFLAGS) $(bench_CFLAGS) $(CFLAGS) -c -o bench-fftw-bench.obj `if test -f 'fftw-bench.c'; then $(CYGPATH_W) 'fftw-bench.c'; else $(CYGPATH_W) '$(srcdir)/fftw-bench.c'; fi` mostlyclean-libtool: -rm -f *.lo clean-libtool: -rm -rf .libs _libs ID: $(HEADERS) $(SOURCES) $(LISP) $(TAGS_FILES) list='$(SOURCES) $(HEADERS) $(LISP) $(TAGS_FILES)'; \ unique=`for i in $$list; do \ if test -f "$$i"; then echo $$i; else echo $(srcdir)/$$i; fi; \ done | \ $(AWK) '{ files[$$0] = 1; nonempty = 1; } \ END { if (nonempty) { for (i in files) print i; }; }'`; \ mkid -fID $$unique tags: TAGS TAGS: $(HEADERS) $(SOURCES) $(TAGS_DEPENDENCIES) \ $(TAGS_FILES) $(LISP) set x; \ here=`pwd`; \ list='$(SOURCES) $(HEADERS) $(LISP) $(TAGS_FILES)'; \ unique=`for i in $$list; do \ if test -f "$$i"; then echo $$i; else echo $(srcdir)/$$i; fi; \ done | \ $(AWK) '{ files[$$0] = 1; nonempty = 1; } \ END { if (nonempty) { for (i in files) print i; }; }'`; \ shift; \ if test -z "$(ETAGS_ARGS)$$*$$unique"; then :; else \ test -n "$$unique" || unique=$$empty_fix; \ if test $$# -gt 0; then \ $(ETAGS) $(ETAGSFLAGS) $(AM_ETAGSFLAGS) $(ETAGS_ARGS) \ "$$@" $$unique; \ else \ $(ETAGS) $(ETAGSFLAGS) $(AM_ETAGSFLAGS) $(ETAGS_ARGS) \ $$unique; \ fi; \ fi ctags: CTAGS CTAGS: $(HEADERS) $(SOURCES) $(TAGS_DEPENDENCIES) \ $(TAGS_FILES) $(LISP) list='$(SOURCES) $(HEADERS) $(LISP) $(TAGS_FILES)'; \ unique=`for i in $$list; do \ if test -f "$$i"; then echo $$i; else echo $(srcdir)/$$i; fi; \ done | \ $(AWK) '{ files[$$0] = 1; nonempty = 1; } \ END { if (nonempty) { for (i in files) print i; }; }'`; \ test -z "$(CTAGS_ARGS)$$unique" \ || $(CTAGS) $(CTAGSFLAGS) $(AM_CTAGSFLAGS) $(CTAGS_ARGS) \ $$unique GTAGS: here=`$(am__cd) $(top_builddir) && pwd` \ && $(am__cd) $(top_srcdir) \ && gtags -i $(GTAGS_ARGS) "$$here" distclean-tags: -rm -f TAGS ID GTAGS GRTAGS GSYMS GPATH tags distdir: $(DISTFILES) @srcdirstrip=`echo "$(srcdir)" | sed 's/[].[^$$\\*]/\\\\&/g'`; \ topsrcdirstrip=`echo "$(top_srcdir)" | sed 's/[].[^$$\\*]/\\\\&/g'`; \ list='$(DISTFILES)'; \ dist_files=`for file in $$list; do echo $$file; done | \ sed -e "s|^$$srcdirstrip/||;t" \ -e "s|^$$topsrcdirstrip/|$(top_builddir)/|;t"`; \ case $$dist_files in \ */*) $(MKDIR_P) `echo "$$dist_files" | \ sed '/\//!d;s|^|$(distdir)/|;s,/[^/]*$$,,' | \ sort -u` ;; \ esac; \ for file in $$dist_files; do \ if test -f $$file || test -d $$file; then d=.; else d=$(srcdir); fi; \ if test -d $$d/$$file; then \ dir=`echo "/$$file" | sed -e 's,/[^/]*$$,,'`; \ if test -d "$(distdir)/$$file"; then \ find "$(distdir)/$$file" -type d ! -perm -700 -exec chmod u+rwx {} \;; \ fi; \ if test -d $(srcdir)/$$file && test $$d != $(srcdir); then \ cp -fpR $(srcdir)/$$file "$(distdir)$$dir" || exit 1; \ find "$(distdir)/$$file" -type d ! -perm -700 -exec chmod u+rwx {} \;; \ fi; \ cp -fpR $$d/$$file "$(distdir)$$dir" || exit 1; \ else \ test -f "$(distdir)/$$file" \ || cp -p $$d/$$file "$(distdir)/$$file" \ || exit 1; \ fi; \ done check-am: all-am $(MAKE) $(AM_MAKEFLAGS) check-local check: check-am all-am: Makefile $(PROGRAMS) installdirs: install: install-am install-exec: install-exec-am install-data: install-data-am uninstall: uninstall-am install-am: all-am @$(MAKE) $(AM_MAKEFLAGS) install-exec-am install-data-am installcheck: installcheck-am install-strip: if test -z '$(STRIP)'; then \ $(MAKE) $(AM_MAKEFLAGS) INSTALL_PROGRAM="$(INSTALL_STRIP_PROGRAM)" \ install_sh_PROGRAM="$(INSTALL_STRIP_PROGRAM)" INSTALL_STRIP_FLAG=-s \ install; \ else \ $(MAKE) $(AM_MAKEFLAGS) INSTALL_PROGRAM="$(INSTALL_STRIP_PROGRAM)" \ install_sh_PROGRAM="$(INSTALL_STRIP_PROGRAM)" INSTALL_STRIP_FLAG=-s \ "INSTALL_PROGRAM_ENV=STRIPPROG='$(STRIP)'" install; \ fi mostlyclean-generic: clean-generic: distclean-generic: -test -z "$(CONFIG_CLEAN_FILES)" || rm -f $(CONFIG_CLEAN_FILES) -test . = "$(srcdir)" || test -z "$(CONFIG_CLEAN_VPATH_FILES)" || rm -f $(CONFIG_CLEAN_VPATH_FILES) maintainer-clean-generic: @echo "This command is intended for maintainers to use" @echo "it deletes files that may require special tools to rebuild." clean: clean-am clean-am: clean-generic clean-libtool clean-noinstPROGRAMS \ mostlyclean-am distclean: distclean-am -rm -rf ./$(DEPDIR) -rm -f Makefile distclean-am: clean-am distclean-compile distclean-generic \ distclean-tags dvi: dvi-am dvi-am: html: html-am html-am: info: info-am info-am: install-data-am: install-dvi: install-dvi-am install-dvi-am: install-exec-am: install-html: install-html-am install-html-am: install-info: install-info-am install-info-am: install-man: install-pdf: install-pdf-am install-pdf-am: install-ps: install-ps-am install-ps-am: installcheck-am: maintainer-clean: maintainer-clean-am -rm -rf ./$(DEPDIR) -rm -f Makefile maintainer-clean-am: distclean-am maintainer-clean-generic mostlyclean: mostlyclean-am mostlyclean-am: mostlyclean-compile mostlyclean-generic \ mostlyclean-libtool pdf: pdf-am pdf-am: ps: ps-am ps-am: uninstall-am: .MAKE: check-am install-am install-strip .PHONY: CTAGS GTAGS all all-am check check-am check-local clean \ clean-generic clean-libtool clean-noinstPROGRAMS ctags \ distclean distclean-compile distclean-generic \ distclean-libtool distclean-tags distdir dvi dvi-am html \ html-am info info-am install install-am install-data \ install-data-am install-dvi install-dvi-am install-exec \ install-exec-am install-html install-html-am install-info \ install-info-am install-man install-pdf install-pdf-am \ install-ps install-ps-am install-strip installcheck \ installcheck-am installdirs maintainer-clean \ maintainer-clean-generic mostlyclean mostlyclean-compile \ mostlyclean-generic mostlyclean-libtool pdf pdf-am ps ps-am \ tags uninstall uninstall-am check-local: bench$(EXEEXT) perl -w $(srcdir)/check.pl $(CHECK_PL_OPTS) -r -c=30 -v `pwd`/bench @echo "--------------------------------------------------------------" @echo " FFTW transforms passed basic tests!" @echo "--------------------------------------------------------------" @SMP_TRUE@ perl -w $(srcdir)/check.pl $(CHECK_PL_OPTS) -r -c=30 -v --nthreads=2 `pwd`/bench @SMP_TRUE@ @echo "--------------------------------------------------------------" @SMP_TRUE@ @echo " FFTW threaded transforms passed basic tests!" @SMP_TRUE@ @echo "--------------------------------------------------------------" bigcheck: bench$(EXEEXT) perl -w $(srcdir)/check.pl $(CHECK_PL_OPTS) -a -v `pwd`/bench @echo "--------------------------------------------------------------" @echo " FFTW transforms passed big tests!" @echo "--------------------------------------------------------------" @SMP_TRUE@ perl -w $(srcdir)/check.pl $(CHECK_PL_OPTS) -a -v --nthreads=2 `pwd`/bench @SMP_TRUE@ perl -w $(srcdir)/check.pl $(CHECK_PL_OPTS) -a -v --nthreads=3 `pwd`/bench @SMP_TRUE@ perl -w $(srcdir)/check.pl $(CHECK_PL_OPTS) -a -v --nthreads=10 `pwd`/bench @SMP_TRUE@ @echo "--------------------------------------------------------------" @SMP_TRUE@ @echo " FFTW threaded transforms passed big tests!" @SMP_TRUE@ @echo "--------------------------------------------------------------" smallcheck: bench$(EXEEXT) perl -w $(srcdir)/check.pl -r -c=1 -v `pwd`/bench perl -w $(srcdir)/check.pl -r --estimate -c=5 -v `pwd`/bench @echo "--------------------------------------------------------------" @echo " FFTW transforms passed a few tests!" @echo "--------------------------------------------------------------" @SMP_TRUE@ perl -w $(srcdir)/check.pl -r --estimate -c=2 -v --nthreads=2 `pwd`/bench @SMP_TRUE@ @echo "--------------------------------------------------------------" @SMP_TRUE@ @echo " FFTW threaded transforms passed a few tests!" @SMP_TRUE@ @echo "--------------------------------------------------------------" paranoid-check: bench$(EXEEXT) @SMP_TRUE@ perl -w $(srcdir)/check.pl -a --patient --nthreads=10 --paranoid `pwd`/bench @SMP_TRUE@ perl -w $(srcdir)/check.pl -a --patient --nthreads=7 --paranoid `pwd`/bench @SMP_TRUE@ perl -w $(srcdir)/check.pl -a --patient --nthreads=3 --paranoid `pwd`/bench @SMP_TRUE@ perl -w $(srcdir)/check.pl -a --patient --nthreads=2 --paranoid `pwd`/bench perl -w $(srcdir)/check.pl -a --patient --paranoid `pwd`/bench exhaustive-check: bench$(EXEEXT) @SMP_TRUE@ perl -w $(srcdir)/check.pl -a --exhaustive --nthreads=10 --paranoid `pwd`/bench @SMP_TRUE@ perl -w $(srcdir)/check.pl -a --exhaustive --nthreads=7 --paranoid `pwd`/bench @SMP_TRUE@ perl -w $(srcdir)/check.pl -a --exhaustive --nthreads=3 --paranoid `pwd`/bench @SMP_TRUE@ perl -w $(srcdir)/check.pl -a --exhaustive --nthreads=2 --paranoid `pwd`/bench perl -w $(srcdir)/check.pl -a --exhaustive --paranoid `pwd`/bench # Tell versions [3.59,3.63) of GNU make to not export all variables. # Otherwise a system limit (for SysV at least) may be exceeded. .NOEXPORT: fftw-3.3.3/tests/bench.c0000644000175000017500000003573512054410050012000 00000000000000/**************************************************************************/ /* NOTE to users: this is the FFTW self-test and benchmark program. It is probably NOT a good place to learn FFTW usage, since it has a lot of added complexity in order to exercise and test the full API, etcetera. We suggest reading the manual. (Some of the self-test code is split off into fftw-bench.c and hook.c.) */ /**************************************************************************/ #include #include #include #include "fftw-bench.h" static const char *mkversion(void) { return FFTW(version); } static const char *mkcc(void) { return FFTW(cc); } static const char *mkcodelet_optim(void) { return FFTW(codelet_optim); } BEGIN_BENCH_DOC BENCH_DOC("name", "fftw3") BENCH_DOCF("version", mkversion) BENCH_DOCF("cc", mkcc) BENCH_DOCF("codelet-optim", mkcodelet_optim) END_BENCH_DOC static FFTW(iodim) *bench_tensor_to_fftw_iodim(bench_tensor *t) { FFTW(iodim) *d; int i; BENCH_ASSERT(t->rnk >= 0); if (t->rnk == 0) return 0; d = (FFTW(iodim) *)bench_malloc(sizeof(FFTW(iodim)) * t->rnk); for (i = 0; i < t->rnk; ++i) { d[i].n = t->dims[i].n; d[i].is = t->dims[i].is; d[i].os = t->dims[i].os; } return d; } static void extract_reim_split(int sign, int size, bench_real *p, bench_real **r, bench_real **i) { if (sign == FFTW_FORWARD) { *r = p + 0; *i = p + size; } else { *r = p + size; *i = p + 0; } } static int sizeof_problem(bench_problem *p) { return tensor_sz(p->sz) * tensor_sz(p->vecsz); } /* ouch */ static int expressible_as_api_many(bench_tensor *t) { int i; BENCH_ASSERT(FINITE_RNK(t->rnk)); i = t->rnk - 1; while (--i >= 0) { bench_iodim *d = t->dims + i; if (d[0].is % d[1].is) return 0; if (d[0].os % d[1].os) return 0; } return 1; } static int *mkn(bench_tensor *t) { int *n = (int *) bench_malloc(sizeof(int *) * t->rnk); int i; for (i = 0; i < t->rnk; ++i) n[i] = t->dims[i].n; return n; } static void mknembed_many(bench_tensor *t, int **inembedp, int **onembedp) { int i; bench_iodim *d; int *inembed = (int *) bench_malloc(sizeof(int *) * t->rnk); int *onembed = (int *) bench_malloc(sizeof(int *) * t->rnk); BENCH_ASSERT(FINITE_RNK(t->rnk)); *inembedp = inembed; *onembedp = onembed; i = t->rnk - 1; while (--i >= 0) { d = t->dims + i; inembed[i+1] = d[0].is / d[1].is; onembed[i+1] = d[0].os / d[1].os; } } /* try to use the most appropriate API function. Big mess. */ static int imax(int a, int b) { return (a > b ? a : b); } static int halfish_sizeof_problem(bench_problem *p) { int n2 = sizeof_problem(p); if (FINITE_RNK(p->sz->rnk) && p->sz->rnk > 0) n2 = (n2 / imax(p->sz->dims[p->sz->rnk - 1].n, 1)) * (p->sz->dims[p->sz->rnk - 1].n / 2 + 1); return n2; } static FFTW(plan) mkplan_real_split(bench_problem *p, unsigned flags) { FFTW(plan) pln; bench_tensor *sz = p->sz, *vecsz = p->vecsz; FFTW(iodim) *dims, *howmany_dims; bench_real *ri, *ii, *ro, *io; int n2 = halfish_sizeof_problem(p); extract_reim_split(FFTW_FORWARD, n2, (bench_real *) p->in, &ri, &ii); extract_reim_split(FFTW_FORWARD, n2, (bench_real *) p->out, &ro, &io); dims = bench_tensor_to_fftw_iodim(sz); howmany_dims = bench_tensor_to_fftw_iodim(vecsz); if (p->sign < 0) { if (verbose > 2) printf("using plan_guru_split_dft_r2c\n"); pln = FFTW(plan_guru_split_dft_r2c)(sz->rnk, dims, vecsz->rnk, howmany_dims, ri, ro, io, flags); } else { if (verbose > 2) printf("using plan_guru_split_dft_c2r\n"); pln = FFTW(plan_guru_split_dft_c2r)(sz->rnk, dims, vecsz->rnk, howmany_dims, ri, ii, ro, flags); } bench_free(dims); bench_free(howmany_dims); return pln; } static FFTW(plan) mkplan_real_interleaved(bench_problem *p, unsigned flags) { FFTW(plan) pln; bench_tensor *sz = p->sz, *vecsz = p->vecsz; if (vecsz->rnk == 0 && tensor_unitstridep(sz) && tensor_real_rowmajorp(sz, p->sign, p->in_place)) goto api_simple; if (vecsz->rnk == 1 && expressible_as_api_many(sz)) goto api_many; goto api_guru; api_simple: switch (sz->rnk) { case 1: if (p->sign < 0) { if (verbose > 2) printf("using plan_dft_r2c_1d\n"); return FFTW(plan_dft_r2c_1d)(sz->dims[0].n, (bench_real *) p->in, (bench_complex *) p->out, flags); } else { if (verbose > 2) printf("using plan_dft_c2r_1d\n"); return FFTW(plan_dft_c2r_1d)(sz->dims[0].n, (bench_complex *) p->in, (bench_real *) p->out, flags); } break; case 2: if (p->sign < 0) { if (verbose > 2) printf("using plan_dft_r2c_2d\n"); return FFTW(plan_dft_r2c_2d)(sz->dims[0].n, sz->dims[1].n, (bench_real *) p->in, (bench_complex *) p->out, flags); } else { if (verbose > 2) printf("using plan_dft_c2r_2d\n"); return FFTW(plan_dft_c2r_2d)(sz->dims[0].n, sz->dims[1].n, (bench_complex *) p->in, (bench_real *) p->out, flags); } break; case 3: if (p->sign < 0) { if (verbose > 2) printf("using plan_dft_r2c_3d\n"); return FFTW(plan_dft_r2c_3d)( sz->dims[0].n, sz->dims[1].n, sz->dims[2].n, (bench_real *) p->in, (bench_complex *) p->out, flags); } else { if (verbose > 2) printf("using plan_dft_c2r_3d\n"); return FFTW(plan_dft_c2r_3d)( sz->dims[0].n, sz->dims[1].n, sz->dims[2].n, (bench_complex *) p->in, (bench_real *) p->out, flags); } break; default: { int *n = mkn(sz); if (p->sign < 0) { if (verbose > 2) printf("using plan_dft_r2c\n"); pln = FFTW(plan_dft_r2c)(sz->rnk, n, (bench_real *) p->in, (bench_complex *) p->out, flags); } else { if (verbose > 2) printf("using plan_dft_c2r\n"); pln = FFTW(plan_dft_c2r)(sz->rnk, n, (bench_complex *) p->in, (bench_real *) p->out, flags); } bench_free(n); return pln; } } api_many: { int *n, *inembed, *onembed; BENCH_ASSERT(vecsz->rnk == 1); n = mkn(sz); mknembed_many(sz, &inembed, &onembed); if (p->sign < 0) { if (verbose > 2) printf("using plan_many_dft_r2c\n"); pln = FFTW(plan_many_dft_r2c)( sz->rnk, n, vecsz->dims[0].n, (bench_real *) p->in, inembed, sz->dims[sz->rnk - 1].is, vecsz->dims[0].is, (bench_complex *) p->out, onembed, sz->dims[sz->rnk - 1].os, vecsz->dims[0].os, flags); } else { if (verbose > 2) printf("using plan_many_dft_c2r\n"); pln = FFTW(plan_many_dft_c2r)( sz->rnk, n, vecsz->dims[0].n, (bench_complex *) p->in, inembed, sz->dims[sz->rnk - 1].is, vecsz->dims[0].is, (bench_real *) p->out, onembed, sz->dims[sz->rnk - 1].os, vecsz->dims[0].os, flags); } bench_free(n); bench_free(inembed); bench_free(onembed); return pln; } api_guru: { FFTW(iodim) *dims, *howmany_dims; if (p->sign < 0) { dims = bench_tensor_to_fftw_iodim(sz); howmany_dims = bench_tensor_to_fftw_iodim(vecsz); if (verbose > 2) printf("using plan_guru_dft_r2c\n"); pln = FFTW(plan_guru_dft_r2c)(sz->rnk, dims, vecsz->rnk, howmany_dims, (bench_real *) p->in, (bench_complex *) p->out, flags); } else { dims = bench_tensor_to_fftw_iodim(sz); howmany_dims = bench_tensor_to_fftw_iodim(vecsz); if (verbose > 2) printf("using plan_guru_dft_c2r\n"); pln = FFTW(plan_guru_dft_c2r)(sz->rnk, dims, vecsz->rnk, howmany_dims, (bench_complex *) p->in, (bench_real *) p->out, flags); } bench_free(dims); bench_free(howmany_dims); return pln; } } static FFTW(plan) mkplan_real(bench_problem *p, unsigned flags) { if (p->split) return mkplan_real_split(p, flags); else return mkplan_real_interleaved(p, flags); } static FFTW(plan) mkplan_complex_split(bench_problem *p, unsigned flags) { FFTW(plan) pln; bench_tensor *sz = p->sz, *vecsz = p->vecsz; FFTW(iodim) *dims, *howmany_dims; bench_real *ri, *ii, *ro, *io; extract_reim_split(p->sign, p->iphyssz, (bench_real *) p->in, &ri, &ii); extract_reim_split(p->sign, p->ophyssz, (bench_real *) p->out, &ro, &io); dims = bench_tensor_to_fftw_iodim(sz); howmany_dims = bench_tensor_to_fftw_iodim(vecsz); if (verbose > 2) printf("using plan_guru_split_dft\n"); pln = FFTW(plan_guru_split_dft)(sz->rnk, dims, vecsz->rnk, howmany_dims, ri, ii, ro, io, flags); bench_free(dims); bench_free(howmany_dims); return pln; } static FFTW(plan) mkplan_complex_interleaved(bench_problem *p, unsigned flags) { FFTW(plan) pln; bench_tensor *sz = p->sz, *vecsz = p->vecsz; if (vecsz->rnk == 0 && tensor_unitstridep(sz) && tensor_rowmajorp(sz)) goto api_simple; if (vecsz->rnk == 1 && expressible_as_api_many(sz)) goto api_many; goto api_guru; api_simple: switch (sz->rnk) { case 1: if (verbose > 2) printf("using plan_dft_1d\n"); return FFTW(plan_dft_1d)(sz->dims[0].n, (bench_complex *) p->in, (bench_complex *) p->out, p->sign, flags); break; case 2: if (verbose > 2) printf("using plan_dft_2d\n"); return FFTW(plan_dft_2d)(sz->dims[0].n, sz->dims[1].n, (bench_complex *) p->in, (bench_complex *) p->out, p->sign, flags); break; case 3: if (verbose > 2) printf("using plan_dft_3d\n"); return FFTW(plan_dft_3d)( sz->dims[0].n, sz->dims[1].n, sz->dims[2].n, (bench_complex *) p->in, (bench_complex *) p->out, p->sign, flags); break; default: { int *n = mkn(sz); if (verbose > 2) printf("using plan_dft\n"); pln = FFTW(plan_dft)(sz->rnk, n, (bench_complex *) p->in, (bench_complex *) p->out, p->sign, flags); bench_free(n); return pln; } } api_many: { int *n, *inembed, *onembed; BENCH_ASSERT(vecsz->rnk == 1); n = mkn(sz); mknembed_many(sz, &inembed, &onembed); if (verbose > 2) printf("using plan_many_dft\n"); pln = FFTW(plan_many_dft)( sz->rnk, n, vecsz->dims[0].n, (bench_complex *) p->in, inembed, sz->dims[sz->rnk - 1].is, vecsz->dims[0].is, (bench_complex *) p->out, onembed, sz->dims[sz->rnk - 1].os, vecsz->dims[0].os, p->sign, flags); bench_free(n); bench_free(inembed); bench_free(onembed); return pln; } api_guru: { FFTW(iodim) *dims, *howmany_dims; dims = bench_tensor_to_fftw_iodim(sz); howmany_dims = bench_tensor_to_fftw_iodim(vecsz); if (verbose > 2) printf("using plan_guru_dft\n"); pln = FFTW(plan_guru_dft)(sz->rnk, dims, vecsz->rnk, howmany_dims, (bench_complex *) p->in, (bench_complex *) p->out, p->sign, flags); bench_free(dims); bench_free(howmany_dims); return pln; } } static FFTW(plan) mkplan_complex(bench_problem *p, unsigned flags) { if (p->split) return mkplan_complex_split(p, flags); else return mkplan_complex_interleaved(p, flags); } static FFTW(plan) mkplan_r2r(bench_problem *p, unsigned flags) { FFTW(plan) pln; bench_tensor *sz = p->sz, *vecsz = p->vecsz; FFTW(r2r_kind) *k; k = (FFTW(r2r_kind) *) bench_malloc(sizeof(FFTW(r2r_kind)) * sz->rnk); { int i; for (i = 0; i < sz->rnk; ++i) switch (p->k[i]) { case R2R_R2HC: k[i] = FFTW_R2HC; break; case R2R_HC2R: k[i] = FFTW_HC2R; break; case R2R_DHT: k[i] = FFTW_DHT; break; case R2R_REDFT00: k[i] = FFTW_REDFT00; break; case R2R_REDFT01: k[i] = FFTW_REDFT01; break; case R2R_REDFT10: k[i] = FFTW_REDFT10; break; case R2R_REDFT11: k[i] = FFTW_REDFT11; break; case R2R_RODFT00: k[i] = FFTW_RODFT00; break; case R2R_RODFT01: k[i] = FFTW_RODFT01; break; case R2R_RODFT10: k[i] = FFTW_RODFT10; break; case R2R_RODFT11: k[i] = FFTW_RODFT11; break; default: BENCH_ASSERT(0); } } if (vecsz->rnk == 0 && tensor_unitstridep(sz) && tensor_rowmajorp(sz)) goto api_simple; if (vecsz->rnk == 1 && expressible_as_api_many(sz)) goto api_many; goto api_guru; api_simple: switch (sz->rnk) { case 1: if (verbose > 2) printf("using plan_r2r_1d\n"); pln = FFTW(plan_r2r_1d)(sz->dims[0].n, (bench_real *) p->in, (bench_real *) p->out, k[0], flags); goto done; case 2: if (verbose > 2) printf("using plan_r2r_2d\n"); pln = FFTW(plan_r2r_2d)(sz->dims[0].n, sz->dims[1].n, (bench_real *) p->in, (bench_real *) p->out, k[0], k[1], flags); goto done; case 3: if (verbose > 2) printf("using plan_r2r_3d\n"); pln = FFTW(plan_r2r_3d)( sz->dims[0].n, sz->dims[1].n, sz->dims[2].n, (bench_real *) p->in, (bench_real *) p->out, k[0], k[1], k[2], flags); goto done; default: { int *n = mkn(sz); if (verbose > 2) printf("using plan_r2r\n"); pln = FFTW(plan_r2r)(sz->rnk, n, (bench_real *) p->in, (bench_real *) p->out, k, flags); bench_free(n); goto done; } } api_many: { int *n, *inembed, *onembed; BENCH_ASSERT(vecsz->rnk == 1); n = mkn(sz); mknembed_many(sz, &inembed, &onembed); if (verbose > 2) printf("using plan_many_r2r\n"); pln = FFTW(plan_many_r2r)( sz->rnk, n, vecsz->dims[0].n, (bench_real *) p->in, inembed, sz->dims[sz->rnk - 1].is, vecsz->dims[0].is, (bench_real *) p->out, onembed, sz->dims[sz->rnk - 1].os, vecsz->dims[0].os, k, flags); bench_free(n); bench_free(inembed); bench_free(onembed); goto done; } api_guru: { FFTW(iodim) *dims, *howmany_dims; dims = bench_tensor_to_fftw_iodim(sz); howmany_dims = bench_tensor_to_fftw_iodim(vecsz); if (verbose > 2) printf("using plan_guru_r2r\n"); pln = FFTW(plan_guru_r2r)(sz->rnk, dims, vecsz->rnk, howmany_dims, (bench_real *) p->in, (bench_real *) p->out, k, flags); bench_free(dims); bench_free(howmany_dims); goto done; } done: bench_free(k); return pln; } FFTW(plan) mkplan(bench_problem *p, unsigned flags) { switch (p->kind) { case PROBLEM_COMPLEX: return mkplan_complex(p, flags); case PROBLEM_REAL: return mkplan_real(p, flags); case PROBLEM_R2R: return mkplan_r2r(p, flags); default: BENCH_ASSERT(0); return 0; } } void main_init(int *argc, char ***argv) { UNUSED(argc); UNUSED(argv); } void initial_cleanup(void) { } void final_cleanup(void) { } int import_wisdom(FILE *f) { return FFTW(import_wisdom_from_file)(f); } void export_wisdom(FILE *f) { FFTW(export_wisdom_to_file)(f); } fftw-3.3.3/tests/hook.c0000644000175000017500000001550312054410050011650 00000000000000/* fftw hook to be used in the benchmark program. We keep it in a separate file because 1) bench.c is supposed to test the API---we do not want to #include "ifftw.h" and accidentally use internal symbols/macros. 2) this code is a royal mess. The messiness is due to A) confusion between internal fftw tensors and bench_tensor's (which we want to keep separate because the benchmark program tests other routines too) B) despite A), our desire to recycle the libbench verifier. */ #include #include "bench-user.h" #define CALLING_FFTW /* hack for Windows DLL nonsense */ #include "api.h" #include "dft.h" #include "rdft.h" extern int paranoid; /* in bench.c */ extern X(plan) the_plan; /* in bench.c */ /* transform an fftw tensor into a bench_tensor. */ static bench_tensor *fftw_tensor_to_bench_tensor(tensor *t) { bench_tensor *bt = mktensor(t->rnk); if (FINITE_RNK(t->rnk)) { int i; for (i = 0; i < t->rnk; ++i) { /* FIXME: 64-bit unclean because of INT -> int conversion */ bt->dims[i].n = t->dims[i].n; bt->dims[i].is = t->dims[i].is; bt->dims[i].os = t->dims[i].os; BENCH_ASSERT(bt->dims[i].n == t->dims[i].n); BENCH_ASSERT(bt->dims[i].is == t->dims[i].is); BENCH_ASSERT(bt->dims[i].os == t->dims[i].os); } } return bt; } /* transform an fftw problem into a bench_problem. */ static bench_problem *fftw_problem_to_bench_problem(planner *plnr, const problem *p_) { bench_problem *bp = 0; switch (p_->adt->problem_kind) { case PROBLEM_DFT: { const problem_dft *p = (const problem_dft *) p_; if (!p->ri || !p->ii) abort(); bp = (bench_problem *) bench_malloc(sizeof(bench_problem)); bp->kind = PROBLEM_COMPLEX; bp->sign = FFT_SIGN; bp->split = 1; /* tensor strides are in R's, not C's */ bp->in = UNTAINT(p->ri); bp->out = UNTAINT(p->ro); bp->ini = UNTAINT(p->ii); bp->outi = UNTAINT(p->io); bp->inphys = bp->outphys = 0; bp->iphyssz = bp->ophyssz = 0; bp->in_place = p->ri == p->ro; bp->sz = fftw_tensor_to_bench_tensor(p->sz); bp->vecsz = fftw_tensor_to_bench_tensor(p->vecsz); bp->k = 0; break; } case PROBLEM_RDFT: { const problem_rdft *p = (const problem_rdft *) p_; int i; if (!p->I || !p->O) abort(); for (i = 0; i < p->sz->rnk; ++i) switch (p->kind[i]) { case R2HC01: case R2HC10: case R2HC11: case HC2R01: case HC2R10: case HC2R11: return bp; default: ; } bp = (bench_problem *) bench_malloc(sizeof(bench_problem)); bp->kind = PROBLEM_R2R; bp->sign = FFT_SIGN; bp->split = 0; bp->in = UNTAINT(p->I); bp->out = UNTAINT(p->O); bp->ini = bp->outi = 0; bp->inphys = bp->outphys = 0; bp->iphyssz = bp->ophyssz = 0; bp->in_place = p->I == p->O; bp->sz = fftw_tensor_to_bench_tensor(p->sz); bp->vecsz = fftw_tensor_to_bench_tensor(p->vecsz); bp->k = (r2r_kind_t *) bench_malloc(sizeof(r2r_kind_t) * p->sz->rnk); for (i = 0; i < p->sz->rnk; ++i) switch (p->kind[i]) { case R2HC: bp->k[i] = R2R_R2HC; break; case HC2R: bp->k[i] = R2R_HC2R; break; case DHT: bp->k[i] = R2R_DHT; break; case REDFT00: bp->k[i] = R2R_REDFT00; break; case REDFT01: bp->k[i] = R2R_REDFT01; break; case REDFT10: bp->k[i] = R2R_REDFT10; break; case REDFT11: bp->k[i] = R2R_REDFT11; break; case RODFT00: bp->k[i] = R2R_RODFT00; break; case RODFT01: bp->k[i] = R2R_RODFT01; break; case RODFT10: bp->k[i] = R2R_RODFT10; break; case RODFT11: bp->k[i] = R2R_RODFT11; break; default: CK(0); } break; } case PROBLEM_RDFT2: { const problem_rdft2 *p = (const problem_rdft2 *) p_; int rnk = p->sz->rnk; if (!p->r0 || !p->r1 || !p->cr || !p->ci) abort(); /* give up verifying rdft2 R2HCII */ if (p->kind != R2HC && p->kind != HC2R) return bp; if (rnk > 0) { /* can't verify separate even/odd arrays for now */ if (2 * (p->r1 - p->r0) != ((p->kind == R2HC) ? p->sz->dims[rnk-1].is : p->sz->dims[rnk-1].os)) return bp; } bp = (bench_problem *) bench_malloc(sizeof(bench_problem)); bp->kind = PROBLEM_REAL; bp->sign = p->kind == R2HC ? FFT_SIGN : -FFT_SIGN; bp->split = 1; /* tensor strides are in R's, not C's */ if (p->kind == R2HC) { bp->sign = FFT_SIGN; bp->in = UNTAINT(p->r0); bp->out = UNTAINT(p->cr); bp->ini = 0; bp->outi = UNTAINT(p->ci); } else { bp->sign = -FFT_SIGN; bp->out = UNTAINT(p->r0); bp->in = UNTAINT(p->cr); bp->outi = 0; bp->ini = UNTAINT(p->ci); } bp->inphys = bp->outphys = 0; bp->iphyssz = bp->ophyssz = 0; bp->in_place = p->r0 == p->cr; bp->sz = fftw_tensor_to_bench_tensor(p->sz); if (rnk > 0) { if (p->kind == R2HC) bp->sz->dims[rnk-1].is /= 2; else bp->sz->dims[rnk-1].os /= 2; } bp->vecsz = fftw_tensor_to_bench_tensor(p->vecsz); bp->k = 0; break; } default: abort(); } bp->userinfo = 0; bp->pstring = 0; bp->destroy_input = !NO_DESTROY_INPUTP(plnr); return bp; } static void hook(planner *plnr, plan *pln, const problem *p_, int optimalp) { int rounds = 5; double tol = SINGLE_PRECISION ? 1.0e-3 : 1.0e-10; UNUSED(optimalp); if (verbose > 5) { printer *pr = X(mkprinter_file)(stdout); pr->print(pr, "%P:%(%p%)\n", p_, pln); X(printer_destroy)(pr); printf("cost %g \n\n", pln->pcost); } if (paranoid) { bench_problem *bp; bp = fftw_problem_to_bench_problem(plnr, p_); if (bp) { X(plan) the_plan_save = the_plan; the_plan = (apiplan *) MALLOC(sizeof(apiplan), PLANS); the_plan->pln = pln; the_plan->prb = (problem *) p_; X(plan_awake)(pln, AWAKE_SQRTN_TABLE); verify_problem(bp, rounds, tol); X(plan_awake)(pln, SLEEPY); X(ifree)(the_plan); the_plan = the_plan_save; problem_destroy(bp); } } } static void paranoid_checks(void) { /* FIXME: assumes char = 8 bits, which is false on at least one DSP I know of. */ #if 0 /* if flags_t is not 64 bits i want to know it. */ CK(sizeof(flags_t) == 8); CK(sizeof(md5uint) >= 4); #endif CK(sizeof(uintptr_t) >= sizeof(R *)); CK(sizeof(INT) >= sizeof(R *)); } void install_hook(void) { planner *plnr = X(the_planner)(); plnr->hook = hook; paranoid_checks(); } void uninstall_hook(void) { planner *plnr = X(the_planner)(); plnr->hook = 0; } fftw-3.3.3/tests/fftw-bench.c0000644000175000017500000001300712054410050012730 00000000000000/* See bench.c. We keep a few common subroutines in this file so that they can be re-used in the MPI test program. */ #include #include #include #include "fftw-bench.h" #ifdef _OPENMP # include #endif #ifdef HAVE_SMP int threads_ok = 1; #endif FFTW(plan) the_plan = 0; static const char *wisdat = "wis.dat"; unsigned the_flags = 0; int paranoid = 0; int usewisdom = 0; int havewisdom = 0; int nthreads = 1; int amnesia = 0; extern void install_hook(void); /* in hook.c */ extern void uninstall_hook(void); /* in hook.c */ #ifdef FFTW_RANDOM_ESTIMATOR extern unsigned FFTW(random_estimate_seed); #endif void useropt(const char *arg) { int x; double y; if (!strcmp(arg, "patient")) the_flags |= FFTW_PATIENT; else if (!strcmp(arg, "estimate")) the_flags |= FFTW_ESTIMATE; else if (!strcmp(arg, "estimatepat")) the_flags |= FFTW_ESTIMATE_PATIENT; else if (!strcmp(arg, "exhaustive")) the_flags |= FFTW_EXHAUSTIVE; else if (!strcmp(arg, "unaligned")) the_flags |= FFTW_UNALIGNED; else if (!strcmp(arg, "nosimd")) the_flags |= FFTW_NO_SIMD; else if (!strcmp(arg, "noindirectop")) the_flags |= FFTW_NO_INDIRECT_OP; else if (!strcmp(arg, "wisdom-only")) the_flags |= FFTW_WISDOM_ONLY; else if (sscanf(arg, "flag=%d", &x) == 1) the_flags |= x; else if (sscanf(arg, "bflag=%d", &x) == 1) the_flags |= 1U << x; else if (!strcmp(arg, "paranoid")) paranoid = 1; else if (!strcmp(arg, "wisdom")) usewisdom = 1; else if (!strcmp(arg, "amnesia")) amnesia = 1; else if (sscanf(arg, "nthreads=%d", &x) == 1) nthreads = x; #ifdef FFTW_RANDOM_ESTIMATOR else if (sscanf(arg, "eseed=%d", &x) == 1) FFTW(random_estimate_seed) = x; #endif else if (sscanf(arg, "timelimit=%lg", &y) == 1) { FFTW(set_timelimit)(y); } else fprintf(stderr, "unknown user option: %s. Ignoring.\n", arg); } void rdwisdom(void) { FILE *f; double tim; int success = 0; if (havewisdom) return; #ifdef HAVE_SMP if (threads_ok) { BENCH_ASSERT(FFTW(init_threads)()); FFTW(plan_with_nthreads)(nthreads); #ifdef _OPENMP omp_set_num_threads(nthreads); #endif } else if (nthreads > 1 && verbose > 1) { fprintf(stderr, "bench: WARNING - nthreads = %d, but threads not supported\n", nthreads); nthreads = 1; } #endif if (!usewisdom) return; timer_start(USER_TIMER); if ((f = fopen(wisdat, "r"))) { if (!import_wisdom(f)) fprintf(stderr, "bench: ERROR reading wisdom\n"); else success = 1; fclose(f); } tim = timer_stop(USER_TIMER); if (success) { if (verbose > 1) printf("READ WISDOM (%g seconds): ", tim); if (verbose > 3) export_wisdom(stdout); if (verbose > 1) printf("\n"); } havewisdom = 1; } void wrwisdom(void) { FILE *f; double tim; if (!havewisdom) return; timer_start(USER_TIMER); if ((f = fopen(wisdat, "w"))) { export_wisdom(f); fclose(f); } tim = timer_stop(USER_TIMER); if (verbose > 1) printf("write wisdom took %g seconds\n", tim); } static unsigned preserve_input_flags(bench_problem *p) { /* * fftw3 cannot preserve input for multidimensional c2r transforms. * Enforce FFTW_DESTROY_INPUT */ if (p->kind == PROBLEM_REAL && p->sign > 0 && !p->in_place && p->sz->rnk > 1) p->destroy_input = 1; if (p->destroy_input) return FFTW_DESTROY_INPUT; else return FFTW_PRESERVE_INPUT; } int can_do(bench_problem *p) { double tim; if (verbose > 2 && p->pstring) printf("Planning %s...\n", p->pstring); rdwisdom(); timer_start(USER_TIMER); the_plan = mkplan(p, preserve_input_flags(p) | the_flags | FFTW_ESTIMATE); tim = timer_stop(USER_TIMER); if (verbose > 2) printf("estimate-planner time: %g s\n", tim); if (the_plan) { FFTW(destroy_plan)(the_plan); return 1; } return 0; } void setup(bench_problem *p) { double tim; if (amnesia) { FFTW(forget_wisdom)(); havewisdom = 0; } /* Regression test: check that fftw_malloc exists and links * properly */ FFTW(free(FFTW(malloc(42)))); rdwisdom(); install_hook(); #ifdef HAVE_SMP if (verbose > 1 && nthreads > 1) printf("NTHREADS = %d\n", nthreads); #endif timer_start(USER_TIMER); the_plan = mkplan(p, preserve_input_flags(p) | the_flags); tim = timer_stop(USER_TIMER); if (verbose > 1) printf("planner time: %g s\n", tim); BENCH_ASSERT(the_plan); { double add, mul, nfma, cost, pcost; FFTW(flops)(the_plan, &add, &mul, &nfma); cost = FFTW(estimate_cost)(the_plan); pcost = FFTW(cost)(the_plan); if (verbose > 1) { FFTW(print_plan)(the_plan); printf("\n"); printf("flops: %0.0f add, %0.0f mul, %0.0f fma\n", add, mul, nfma); printf("estimated cost: %f, pcost = %f\n", cost, pcost); } } } void doit(int iter, bench_problem *p) { int i; FFTW(plan) q = the_plan; UNUSED(p); for (i = 0; i < iter; ++i) FFTW(execute)(q); } void done(bench_problem *p) { UNUSED(p); FFTW(destroy_plan)(the_plan); uninstall_hook(); } void cleanup(void) { initial_cleanup(); wrwisdom(); #ifdef HAVE_SMP FFTW(cleanup_threads)(); #else FFTW(cleanup)(); #endif # ifdef FFTW_DEBUG_MALLOC { /* undocumented memory checker */ FFTW_EXTERN void FFTW(malloc_print_minfo)(int v); FFTW(malloc_print_minfo)(verbose); } # endif final_cleanup(); } fftw-3.3.3/tests/fftw-bench.h0000644000175000017500000000154112054410050012735 00000000000000/* declarations of common subroutines, etc. for use with FFTW self-test/benchmark program (see bench.c). */ #include "bench-user.h" #include "fftw3.h" #define CONCAT(prefix, name) prefix ## name #if defined(BENCHFFT_SINGLE) #define FFTW(x) CONCAT(fftwf_, x) #elif defined(BENCHFFT_LDOUBLE) #define FFTW(x) CONCAT(fftwl_, x) #elif defined(BENCHFFT_QUAD) #define FFTW(x) CONCAT(fftwq_, x) #else #define FFTW(x) CONCAT(fftw_, x) #endif #ifdef __cplusplus extern "C" { #endif /* __cplusplus */ extern FFTW(plan) mkplan(bench_problem *p, unsigned flags); extern void initial_cleanup(void); extern void final_cleanup(void); extern int import_wisdom(FILE *f); extern void export_wisdom(FILE *f); #if defined(HAVE_THREADS) || defined(HAVE_OPENMP) # define HAVE_SMP extern int threads_ok; #endif #ifdef __cplusplus } /* extern "C" */ #endif /* __cplusplus */ fftw-3.3.3/tests/check.pl0000644000175000017500000001763512054410050012166 00000000000000#! /usr/bin/perl -w $program = "./bench"; $default_options = ""; $verbose = 0; $paranoid = 0; $exhaustive = 0; $patient = 0; $estimate = 0; $wisdom = 0; $nthreads = 1; $rounds = 0; $maxsize = 60000; $maxcount = 100; $do_0d = 0; $do_1d = 0; $do_2d = 0; $do_random = 0; $keepgoing = 0; $flushcount = 42; $mpi = 0; $mpi_transposed_in = 0; $mpi_transposed_out = 0; sub make_options { my $options = $default_options; $options = "--verify-rounds=$rounds $options" if $rounds; $options = "--verbose=$verbose $options" if $verbose; $options = "-o paranoid $options" if $paranoid; $options = "-o exhaustive $options" if $exhaustive; $options = "-o patient $options" if $patient; $options = "-o estimate $options" if $estimate; $options = "-o wisdom $options" if $wisdom; $options = "-o nthreads=$nthreads $options" if ($nthreads > 1); $options = "-obflag=30 $options" if $mpi_transposed_in; $options = "-obflag=31 $options" if $mpi_transposed_out; return $options; } @list_of_problems = (); sub flush_problems { my $options = shift; my $problist = ""; if ($#list_of_problems >= 0) { for (@list_of_problems) { $problist = "$problist --verify '$_'"; } print "Executing \"$program $options $problist\"\n" if $verbose; system("$program $options $problist"); $exit_value = $? >> 8; $signal_num = $? & 127; $dumped_core = $? & 128; if ($signal_num == 1) { print "hangup\n"; exit 0; } if ($signal_num == 2) { print "interrupted\n"; exit 0; } if ($signal_num == 9) { print "killed\n"; exit 0; } if ($exit_value != 0 || $dumped_core || $signal_num) { print "FAILED $program: $problist\n"; if ($signal_num) { print "received signal $signal_num\n"; } exit 1 unless $keepgoing; } @list_of_problems = (); } } sub do_problem { my $problem = shift; my $doablep = shift; my $options = &make_options; if ($problem =~ /\// && $problem =~ /r/ && ($problem =~ /i.*x/ || $problem =~ /v/ || $problem =~ /\*/)) { return; # cannot do real split inplace-multidimensional or vector } # in --mpi mode, restrict to problems supported by MPI code if ($mpi) { if ($problem =~ /\//) { return; } # no split if ($problem =~ /\*/) { return; } # no non-contiguous vectors if ($problem =~ /r/ && $problem !~ /x/) { return; } # no 1d r2c if ($problem =~ /k/ && $problem !~ /x/) { return; } # no 1d r2r if ($mpi_transposed_in || $problem =~ /\[/) { if ($problem !~ /x/) { return; } # no 1d transposed_in if ($problem =~ /r/ && $problem !~ /b/) { return; } # only c2r } if ($mpi_transposed_out || $problem =~ /\]/) { if ($problem !~ /x/) { return; } # no 1d transposed_out if ($problem =~ /r/ && $problem =~ /b/) { return; } # only r2c } } # size-1 redft00 is not defined/doable return if ($problem =~ /[^0-9]1e00/); if ($doablep) { @list_of_problems = ($problem, @list_of_problems); &flush_problems($options) if ($#list_of_problems > $flushcount); } else { print "Executing \"$program $options --can-do $problem\"\n" if $verbose; $result=`$program $options --can-do $problem`; if ($result ne "#f\n" && $result ne "#f\r\n") { print "FAILED $program: $problem is not undoable\n"; exit 1 unless $keepgoing; } } } # given geometry, try both directions and in place/out of place sub do_geometry { my $geom = shift; my $doablep = shift; do_problem("if$geom", $doablep); do_problem("of$geom", $doablep); do_problem("ib$geom", $doablep); do_problem("ob$geom", $doablep); do_problem("//if$geom", $doablep); do_problem("//of$geom", $doablep); do_problem("//ib$geom", $doablep); do_problem("//ob$geom", $doablep); } # given size, try all transform kinds (complex, real, etc.) sub do_size { my $size = shift; my $doablep = shift; do_geometry("c$size", $doablep); do_geometry("r$size", $doablep); } sub small_0d { for ($i = 0; $i <= 16; ++$i) { for ($j = 0; $j <= 16; ++$j) { for ($vl = 1; $vl <= 5; ++$vl) { my $ivl = $i * $vl; my $jvl = $j * $vl; do_problem("o1v${i}:${vl}:${jvl}x${j}:${ivl}:${vl}x${vl}:1:1", 1); do_problem("i1v${i}:${vl}:${jvl}x${j}:${ivl}:${vl}x${vl}:1:1", 1); do_problem("ok1v${i}:${vl}:${jvl}x${j}:${ivl}:${vl}x${vl}:1:1", 1); do_problem("ik1v${i}:${vl}:${jvl}x${j}:${ivl}:${vl}x${vl}:1:1", 1); } } } } sub small_1d { do_size (0, 0); for ($i = 1; $i <= 100; ++$i) { do_size ($i, 1); } do_size (128, 1); do_size (256, 1); do_size (512, 1); do_size (1024, 1); do_size (2048, 1); do_size (4096, 1); } sub small_2d { do_size ("0x0", 0); for ($i = 1; $i <= 100; ++$i) { my $ub = 900/$i; $ub = 100 if $ub > 100; for ($j = 1; $j <= $ub; ++$j) { do_size ("${i}x${j}", 1); } } } sub rand_small_factors { my $l = shift; my $n = 1; my $maxfactor = 13; my $f = int(rand($maxfactor) + 1); while ($n * $f < $l) { $n *= $f; $f = int(rand($maxfactor) + 1); }; return $n; } # way too complicated... sub one_random_test { my $q = int(2 + rand($maxsize)); my $rnk = int(1 + rand(4)); my $vtype = int(rand(3)); my $g = int(2 + exp(log($q) / ($rnk + ($vtype > 0)))); my $first = 1; my $sz = ""; my $is_r2r = shift; my @r2r_kinds = ("f", "b", "h", "e00", "e01", "e10", "e11", "o00", "o01", "o10", "o11"); while ($q > 1 && $rnk > 0) { my $r = rand_small_factors(int(rand($g) + 10)); if ($r > 1) { $sz = "${sz}x" if (!$first); $first = 0; $sz = "${sz}${r}"; if ($is_r2r) { my $k = $r2r_kinds[int(1 + rand($#r2r_kinds))]; $sz = "${sz}${k}"; } $q = int($q / $r); if ($g > $q) { $g = $q; } --$rnk; } } if ($vtype > 0 && $g > 1) { my $v = int(1 + rand($g)); $sz = "${sz}*${v}" if ($vtype == 1); $sz = "${sz}v${v}" if ($vtype == 2); } if ($mpi) { my $stype = int(rand(3)); $sz = "]${sz}" if ($stype == 1); $sz = "[${sz}" if ($stype == 2); } $sz = "d$sz" if (int(rand(3)) == 0); if ($is_r2r) { do_problem("ik$sz", 1); do_problem("ok$sz", 1); } else { do_size($sz, 1); } } sub random_tests { my $i; for ($i = 0; $i < $maxcount; ++$i) { &one_random_test(0); &one_random_test(1); } } sub parse_arguments (@) { local (@arglist) = @_; while (@arglist) { if ($arglist[0] eq '-v') { ++$verbose; } elsif ($arglist[0] eq '--verbose') { ++$verbose; } elsif ($arglist[0] eq '-p') { ++$paranoid; } elsif ($arglist[0] eq '--paranoid') { ++$paranoid; } elsif ($arglist[0] eq '--exhaustive') { ++$exhaustive; } elsif ($arglist[0] eq '--patient') { ++$patient; } elsif ($arglist[0] eq '--estimate') { ++$estimate; } elsif ($arglist[0] eq '--wisdom') { ++$wisdom; } elsif ($arglist[0] =~ /^--nthreads=(.+)$/) { $nthreads = $1; } elsif ($arglist[0] eq '-k') { ++$keepgoing; } elsif ($arglist[0] eq '--keep-going') { ++$keepgoing; } elsif ($arglist[0] =~ /^--verify-rounds=(.+)$/) { $rounds = $1; } elsif ($arglist[0] =~ /^--count=(.+)$/) { $maxcount = $1; } elsif ($arglist[0] =~ /^-c=(.+)$/) { $maxcount = $1; } elsif ($arglist[0] =~ /^--flushcount=(.+)$/) { $flushcount = $1; } elsif ($arglist[0] =~ /^--maxsize=(.+)$/) { $maxsize = $1; } elsif ($arglist[0] eq '--mpi') { ++$mpi; } elsif ($arglist[0] eq '--mpi-transposed-in') { ++$mpi; ++$mpi_transposed_in; } elsif ($arglist[0] eq '--mpi-transposed-out') { ++$mpi; ++$mpi_transposed_out; } elsif ($arglist[0] eq '-0d') { ++$do_0d; } elsif ($arglist[0] eq '-1d') { ++$do_1d; } elsif ($arglist[0] eq '-2d') { ++$do_2d; } elsif ($arglist[0] eq '-r') { ++$do_random; } elsif ($arglist[0] eq '--random') { ++$do_random; } elsif ($arglist[0] eq '-a') { ++$do_0d; ++$do_1d; ++$do_2d; ++$do_random; } else { $program=$arglist[0]; } shift (@arglist); } } # MAIN PROGRAM: &parse_arguments (@ARGV); &random_tests if $do_random; &small_0d if $do_0d; &small_1d if $do_1d; &small_2d if $do_2d; { my $options = &make_options; &flush_problems($options); } fftw-3.3.3/mpi/0000755000175000017500000000000012054411110010241 500000000000000fftw-3.3.3/mpi/fftw3-mpi.h0000644000175000017500000002263012054410050012153 00000000000000/* * Copyright (c) 2003, 2007-11 Matteo Frigo * Copyright (c) 2003, 2007-11 Massachusetts Institute of Technology * * The following statement of license applies *only* to this header file, * and *not* to the other files distributed with FFTW or derived therefrom: * * Redistribution and use in source and binary forms, with or without * modification, are permitted provided that the following conditions * are met: * * 1. Redistributions of source code must retain the above copyright * notice, this list of conditions and the following disclaimer. * * 2. Redistributions in binary form must reproduce the above copyright * notice, this list of conditions and the following disclaimer in the * documentation and/or other materials provided with the distribution. * * THIS SOFTWARE IS PROVIDED BY THE AUTHOR ``AS IS'' AND ANY EXPRESS * OR IMPLIED WARRANTIES, INCLUDING, BUT NOT LIMITED TO, THE IMPLIED * WARRANTIES OF MERCHANTABILITY AND FITNESS FOR A PARTICULAR PURPOSE * ARE DISCLAIMED. IN NO EVENT SHALL THE AUTHOR BE LIABLE FOR ANY * DIRECT, INDIRECT, INCIDENTAL, SPECIAL, EXEMPLARY, OR CONSEQUENTIAL * DAMAGES (INCLUDING, BUT NOT LIMITED TO, PROCUREMENT OF SUBSTITUTE * GOODS OR SERVICES; LOSS OF USE, DATA, OR PROFITS; OR BUSINESS * INTERRUPTION) HOWEVER CAUSED AND ON ANY THEORY OF LIABILITY, * WHETHER IN CONTRACT, STRICT LIABILITY, OR TORT (INCLUDING * NEGLIGENCE OR OTHERWISE) ARISING IN ANY WAY OUT OF THE USE OF THIS * SOFTWARE, EVEN IF ADVISED OF THE POSSIBILITY OF SUCH DAMAGE. */ /***************************** NOTE TO USERS ********************************* * * THIS IS A HEADER FILE, NOT A MANUAL * * If you want to know how to use FFTW, please read the manual, * online at http://www.fftw.org/doc/ and also included with FFTW. * For a quick start, see the manual's tutorial section. * * (Reading header files to learn how to use a library is a habit * stemming from code lacking a proper manual. Arguably, it's a * *bad* habit in most cases, because header files can contain * interfaces that are not part of the public, stable API.) * ****************************************************************************/ #ifndef FFTW3_MPI_H #define FFTW3_MPI_H #include "fftw3.h" #include #ifdef __cplusplus extern "C" { #endif /* __cplusplus */ struct fftw_mpi_ddim_do_not_use_me { ptrdiff_t n; /* dimension size */ ptrdiff_t ib; /* input block */ ptrdiff_t ob; /* output block */ }; /* huge second-order macro that defines prototypes for all API functions. We expand this macro for each supported precision XM: name-mangling macro (MPI) X: name-mangling macro (serial) R: real data type C: complex data type */ #define FFTW_MPI_DEFINE_API(XM, X, R, C) \ \ typedef struct fftw_mpi_ddim_do_not_use_me XM(ddim); \ \ FFTW_EXTERN void XM(init)(void); \ FFTW_EXTERN void XM(cleanup)(void); \ \ FFTW_EXTERN ptrdiff_t XM(local_size_many_transposed) \ (int rnk, const ptrdiff_t *n, ptrdiff_t howmany, \ ptrdiff_t block0, ptrdiff_t block1, MPI_Comm comm, \ ptrdiff_t *local_n0, ptrdiff_t *local_0_start, \ ptrdiff_t *local_n1, ptrdiff_t *local_1_start); \ FFTW_EXTERN ptrdiff_t XM(local_size_many) \ (int rnk, const ptrdiff_t *n, ptrdiff_t howmany, \ ptrdiff_t block0, MPI_Comm comm, \ ptrdiff_t *local_n0, ptrdiff_t *local_0_start); \ FFTW_EXTERN ptrdiff_t XM(local_size_transposed) \ (int rnk, const ptrdiff_t *n, MPI_Comm comm, \ ptrdiff_t *local_n0, ptrdiff_t *local_0_start, \ ptrdiff_t *local_n1, ptrdiff_t *local_1_start); \ FFTW_EXTERN ptrdiff_t XM(local_size) \ (int rnk, const ptrdiff_t *n, MPI_Comm comm, \ ptrdiff_t *local_n0, ptrdiff_t *local_0_start); \ FFTW_EXTERN ptrdiff_t XM(local_size_many_1d)( \ ptrdiff_t n0, ptrdiff_t howmany, \ MPI_Comm comm, int sign, unsigned flags, \ ptrdiff_t *local_ni, ptrdiff_t *local_i_start, \ ptrdiff_t *local_no, ptrdiff_t *local_o_start); \ FFTW_EXTERN ptrdiff_t XM(local_size_1d)( \ ptrdiff_t n0, MPI_Comm comm, int sign, unsigned flags, \ ptrdiff_t *local_ni, ptrdiff_t *local_i_start, \ ptrdiff_t *local_no, ptrdiff_t *local_o_start); \ FFTW_EXTERN ptrdiff_t XM(local_size_2d)( \ ptrdiff_t n0, ptrdiff_t n1, MPI_Comm comm, \ ptrdiff_t *local_n0, ptrdiff_t *local_0_start); \ FFTW_EXTERN ptrdiff_t XM(local_size_2d_transposed)( \ ptrdiff_t n0, ptrdiff_t n1, MPI_Comm comm, \ ptrdiff_t *local_n0, ptrdiff_t *local_0_start, \ ptrdiff_t *local_n1, ptrdiff_t *local_1_start); \ FFTW_EXTERN ptrdiff_t XM(local_size_3d)( \ ptrdiff_t n0, ptrdiff_t n1, ptrdiff_t n2, MPI_Comm comm, \ ptrdiff_t *local_n0, ptrdiff_t *local_0_start); \ FFTW_EXTERN ptrdiff_t XM(local_size_3d_transposed)( \ ptrdiff_t n0, ptrdiff_t n1, ptrdiff_t n2, MPI_Comm comm, \ ptrdiff_t *local_n0, ptrdiff_t *local_0_start, \ ptrdiff_t *local_n1, ptrdiff_t *local_1_start); \ \ FFTW_EXTERN X(plan) XM(plan_many_transpose) \ (ptrdiff_t n0, ptrdiff_t n1, \ ptrdiff_t howmany, ptrdiff_t block0, ptrdiff_t block1, \ R *in, R *out, MPI_Comm comm, unsigned flags); \ FFTW_EXTERN X(plan) XM(plan_transpose) \ (ptrdiff_t n0, ptrdiff_t n1, \ R *in, R *out, MPI_Comm comm, unsigned flags); \ \ FFTW_EXTERN X(plan) XM(plan_many_dft) \ (int rnk, const ptrdiff_t *n, ptrdiff_t howmany, \ ptrdiff_t block, ptrdiff_t tblock, C *in, C *out, \ MPI_Comm comm, int sign, unsigned flags); \ FFTW_EXTERN X(plan) XM(plan_dft) \ (int rnk, const ptrdiff_t *n, C *in, C *out, \ MPI_Comm comm, int sign, unsigned flags); \ FFTW_EXTERN X(plan) XM(plan_dft_1d) \ (ptrdiff_t n0, C *in, C *out, \ MPI_Comm comm, int sign, unsigned flags); \ FFTW_EXTERN X(plan) XM(plan_dft_2d) \ (ptrdiff_t n0, ptrdiff_t n1, C *in, C *out, \ MPI_Comm comm, int sign, unsigned flags); \ FFTW_EXTERN X(plan) XM(plan_dft_3d) \ (ptrdiff_t n0, ptrdiff_t n1, ptrdiff_t n2, C *in, C *out, \ MPI_Comm comm, int sign, unsigned flags); \ \ FFTW_EXTERN X(plan) XM(plan_many_r2r) \ (int rnk, const ptrdiff_t *n, ptrdiff_t howmany, \ ptrdiff_t iblock, ptrdiff_t oblock, R *in, R *out, \ MPI_Comm comm, const X(r2r_kind) *kind, unsigned flags); \ FFTW_EXTERN X(plan) XM(plan_r2r) \ (int rnk, const ptrdiff_t *n, R *in, R *out, \ MPI_Comm comm, const X(r2r_kind) *kind, unsigned flags); \ FFTW_EXTERN X(plan) XM(plan_r2r_2d) \ (ptrdiff_t n0, ptrdiff_t n1, R *in, R *out, MPI_Comm comm, \ X(r2r_kind) kind0, X(r2r_kind) kind1, unsigned flags); \ FFTW_EXTERN X(plan) XM(plan_r2r_3d) \ (ptrdiff_t n0, ptrdiff_t n1, ptrdiff_t n2, \ R *in, R *out, MPI_Comm comm, X(r2r_kind) kind0, \ X(r2r_kind) kind1, X(r2r_kind) kind2, unsigned flags); \ \ FFTW_EXTERN X(plan) XM(plan_many_dft_r2c) \ (int rnk, const ptrdiff_t *n, ptrdiff_t howmany, \ ptrdiff_t iblock, ptrdiff_t oblock, R *in, C *out, \ MPI_Comm comm, unsigned flags); \ FFTW_EXTERN X(plan) XM(plan_dft_r2c) \ (int rnk, const ptrdiff_t *n, R *in, C *out, \ MPI_Comm comm, unsigned flags); \ FFTW_EXTERN X(plan) XM(plan_dft_r2c_2d) \ (ptrdiff_t n0, ptrdiff_t n1, R *in, C *out, \ MPI_Comm comm, unsigned flags); \ FFTW_EXTERN X(plan) XM(plan_dft_r2c_3d) \ (ptrdiff_t n0, ptrdiff_t n1, ptrdiff_t n2, R *in, C *out, \ MPI_Comm comm, unsigned flags); \ \ FFTW_EXTERN X(plan) XM(plan_many_dft_c2r) \ (int rnk, const ptrdiff_t *n, ptrdiff_t howmany, \ ptrdiff_t iblock, ptrdiff_t oblock, C *in, R *out, \ MPI_Comm comm, unsigned flags); \ FFTW_EXTERN X(plan) XM(plan_dft_c2r) \ (int rnk, const ptrdiff_t *n, C *in, R *out, \ MPI_Comm comm, unsigned flags); \ FFTW_EXTERN X(plan) XM(plan_dft_c2r_2d) \ (ptrdiff_t n0, ptrdiff_t n1, C *in, R *out, \ MPI_Comm comm, unsigned flags); \ FFTW_EXTERN X(plan) XM(plan_dft_c2r_3d) \ (ptrdiff_t n0, ptrdiff_t n1, ptrdiff_t n2, C *in, R *out, \ MPI_Comm comm, unsigned flags); \ \ FFTW_EXTERN void XM(gather_wisdom)(MPI_Comm comm_); \ FFTW_EXTERN void XM(broadcast_wisdom)(MPI_Comm comm_); \ \ FFTW_EXTERN void XM(execute_dft)(X(plan) p, C *in, C *out); \ FFTW_EXTERN void XM(execute_dft_r2c)(X(plan) p, R *in, C *out); \ FFTW_EXTERN void XM(execute_dft_c2r)(X(plan) p, C *in, R *out); \ FFTW_EXTERN void XM(execute_r2r)(X(plan) p, R *in, R *out); /* end of FFTW_MPI_DEFINE_API macro */ #define FFTW_MPI_MANGLE_DOUBLE(name) FFTW_MANGLE_DOUBLE(FFTW_CONCAT(mpi_,name)) #define FFTW_MPI_MANGLE_FLOAT(name) FFTW_MANGLE_FLOAT(FFTW_CONCAT(mpi_,name)) #define FFTW_MPI_MANGLE_LONG_DOUBLE(name) FFTW_MANGLE_LONG_DOUBLE(FFTW_CONCAT(mpi_,name)) FFTW_MPI_DEFINE_API(FFTW_MPI_MANGLE_DOUBLE, FFTW_MANGLE_DOUBLE, double, fftw_complex) FFTW_MPI_DEFINE_API(FFTW_MPI_MANGLE_FLOAT, FFTW_MANGLE_FLOAT, float, fftwf_complex) FFTW_MPI_DEFINE_API(FFTW_MPI_MANGLE_LONG_DOUBLE, FFTW_MANGLE_LONG_DOUBLE, long double, fftwl_complex) #define FFTW_MPI_DEFAULT_BLOCK (0) /* MPI-specific flags */ #define FFTW_MPI_SCRAMBLED_IN (1U << 27) #define FFTW_MPI_SCRAMBLED_OUT (1U << 28) #define FFTW_MPI_TRANSPOSED_IN (1U << 29) #define FFTW_MPI_TRANSPOSED_OUT (1U << 30) #ifdef __cplusplus } /* extern "C" */ #endif /* __cplusplus */ #endif /* FFTW3_MPI_H */ fftw-3.3.3/mpi/Makefile.am0000644000175000017500000001044612054410050012224 00000000000000AM_CPPFLAGS = -I$(top_srcdir)/kernel -I$(top_srcdir)/dft \ -I$(top_srcdir)/rdft -I$(top_srcdir)/api -I$(top_srcdir)/tests \ -I$(top_srcdir)/libbench2 if MPI lib_LTLIBRARIES = libfftw3@PREC_SUFFIX@_mpi.la include_HEADERS = fftw3-mpi.h nodist_include_HEADERS = fftw3-mpi.f03 fftw3l-mpi.f03 noinst_PROGRAMS = mpi-bench endif CC=@MPICC@ EXTRA_DIST = testsched.c f03api.sh f03-wrap.sh genf03-wrap.pl fftw3-mpi.f03.in fftw3l-mpi.f03.in BUILT_SOURCES = fftw3-mpi.f03.in fftw3-mpi.f03 fftw3l-mpi.f03.in fftw3l-mpi.f03 f03-wrap.c CLEANFILES = fftw3-mpi.f03 fftw3l-mpi.f03 TRANSPOSE_SRC = transpose-alltoall.c transpose-pairwise.c transpose-recurse.c transpose-problem.c transpose-solve.c mpi-transpose.h DFT_SRC = dft-serial.c dft-rank-geq2.c dft-rank-geq2-transposed.c dft-rank1.c dft-rank1-bigvec.c dft-problem.c dft-solve.c mpi-dft.h RDFT_SRC = rdft-serial.c rdft-rank-geq2.c rdft-rank-geq2-transposed.c rdft-rank1-bigvec.c rdft-problem.c rdft-solve.c mpi-rdft.h RDFT2_SRC = rdft2-serial.c rdft2-rank-geq2.c rdft2-rank-geq2-transposed.c rdft2-problem.c rdft2-solve.c mpi-rdft2.h SRC = any-true.c api.c block.c choose-radix.c conf.c dtensor.c fftw3-mpi.h ifftw-mpi.h rearrange.c wisdom-api.c f03-wrap.c libfftw3@PREC_SUFFIX@_mpi_la_SOURCES = $(SRC) $(TRANSPOSE_SRC) $(DFT_SRC) $(RDFT_SRC) $(RDFT2_SRC) libfftw3@PREC_SUFFIX@_mpi_la_LDFLAGS = -version-info @SHARED_VERSION_INFO@ libfftw3@PREC_SUFFIX@_mpi_la_LIBADD = ../libfftw3@PREC_SUFFIX@.la @MPILIBS@ if THREADS mpi_bench_CFLAGS = $(PTHREAD_CFLAGS) if !COMBINED_THREADS LIBFFTWTHREADS = $(top_builddir)/threads/libfftw3@PREC_SUFFIX@_threads.la endif else if OPENMP mpi_bench_CFLAGS = $(OPENMP_CFLAGS) LIBFFTWTHREADS = $(top_builddir)/threads/libfftw3@PREC_SUFFIX@_omp.la endif endif mpi_bench_SOURCES = mpi-bench.c $(top_srcdir)/tests/fftw-bench.c $(top_srcdir)/tests/hook.c mpi_bench_LDADD = libfftw3@PREC_SUFFIX@_mpi.la $(LIBFFTWTHREADS) $(top_builddir)/libfftw3@PREC_SUFFIX@.la $(top_builddir)/libbench2/libbench2.a $(MPILIBS) $(THREADLIBS) CHECK = $(top_srcdir)/tests/check.pl NUMCHECK=10 CHECKSIZE=10000 CHECKOPTS = --verbose --random --maxsize=$(CHECKSIZE) -c=$(NUMCHECK) $(CHECK_PL_OPTS) if MPI check-local: mpi-bench$(EXEEXT) perl -w $(CHECK) $(CHECKOPTS) --mpi "$(MPIRUN) -np 1 `pwd`/mpi-bench" @echo "--------------------------------------------------------------" @echo " MPI FFTW transforms passed "$(NUMCHECK)" tests, 1 CPU" @echo "--------------------------------------------------------------" perl -w $(CHECK) $(CHECKOPTS) --mpi "$(MPIRUN) -np 2 `pwd`/mpi-bench" @echo "--------------------------------------------------------------" @echo " MPI FFTW transforms passed "$(NUMCHECK)" tests, 2 CPUs" @echo "--------------------------------------------------------------" perl -w $(CHECK) $(CHECKOPTS) --mpi "$(MPIRUN) -np 3 `pwd`/mpi-bench" @echo "--------------------------------------------------------------" @echo " MPI FFTW transforms passed "$(NUMCHECK)" tests, 3 CPUs" @echo "--------------------------------------------------------------" perl -w $(CHECK) $(CHECKOPTS) --mpi "$(MPIRUN) -np 4 `pwd`/mpi-bench" @echo "--------------------------------------------------------------" @echo " MPI FFTW transforms passed "$(NUMCHECK)" tests, 4 CPUs" @echo "--------------------------------------------------------------" if SMP perl -w $(CHECK) $(CHECKOPTS) --mpi --nthreads=2 "$(MPIRUN) -np 3 `pwd`/mpi-bench" @echo "--------------------------------------------------------------" @echo " MPI FFTW threaded transforms passed "$(NUMCHECK)" tests!" @echo "--------------------------------------------------------------" endif bigcheck: mpi-bench$(EXEEXT) $(MAKE) $(AM_MAKEFLAGS) NUMCHECK=100 CHECKSIZE=60000 check-local smallcheck: mpi-bench$(EXEEXT) $(MAKE) $(AM_MAKEFLAGS) NUMCHECK=2 check-local endif fftw3-mpi.f03: fftw3-mpi.f03.in sed 's/C_MPI_FINT/@C_MPI_FINT@/' $(srcdir)/fftw3-mpi.f03.in > $@ fftw3l-mpi.f03: fftw3l-mpi.f03.in sed 's/C_MPI_FINT/@C_MPI_FINT@/' $(srcdir)/fftw3l-mpi.f03.in > $@ if MAINTAINER_MODE fftw3-mpi.f03.in: fftw3-mpi.h f03api.sh $(top_srcdir)/api/genf03.pl sh $(srcdir)/f03api.sh d f > $@ fftw3l-mpi.f03.in: fftw3-mpi.h f03api.sh $(top_srcdir)/api/genf03.pl sh $(srcdir)/f03api.sh l | grep -v parameter | sed 's/fftw3.f03/fftw3l.f03/' > $@ f03-wrap.c: fftw3-mpi.h f03-wrap.sh genf03-wrap.pl sh $(srcdir)/f03-wrap.sh > $@ endif fftw-3.3.3/mpi/Makefile.in0000644000175000017500000010721212054410200012230 00000000000000# Makefile.in generated by automake 1.11.6 from Makefile.am. # @configure_input@ # Copyright (C) 1994, 1995, 1996, 1997, 1998, 1999, 2000, 2001, 2002, # 2003, 2004, 2005, 2006, 2007, 2008, 2009, 2010, 2011 Free Software # Foundation, Inc. # This Makefile.in is free software; the Free Software Foundation # gives unlimited permission to copy and/or distribute it, # with or without modifications, as long as this notice is preserved. # This program is distributed in the hope that it will be useful, # but WITHOUT ANY WARRANTY, to the extent permitted by law; without # even the implied warranty of MERCHANTABILITY or FITNESS FOR A # PARTICULAR PURPOSE. @SET_MAKE@ VPATH = @srcdir@ am__make_dryrun = \ { \ am__dry=no; \ case $$MAKEFLAGS in \ *\\[\ \ ]*) \ echo 'am--echo: ; @echo "AM" OK' | $(MAKE) -f - 2>/dev/null \ | grep '^AM OK$$' >/dev/null || am__dry=yes;; \ *) \ for am__flg in $$MAKEFLAGS; do \ case $$am__flg in \ *=*|--*) ;; \ *n*) am__dry=yes; break;; \ esac; \ done;; \ esac; \ test $$am__dry = yes; \ } pkgdatadir = $(datadir)/@PACKAGE@ pkgincludedir = $(includedir)/@PACKAGE@ pkglibdir = $(libdir)/@PACKAGE@ pkglibexecdir = $(libexecdir)/@PACKAGE@ am__cd = CDPATH="$${ZSH_VERSION+.}$(PATH_SEPARATOR)" && cd install_sh_DATA = $(install_sh) -c -m 644 install_sh_PROGRAM = $(install_sh) -c install_sh_SCRIPT = $(install_sh) -c INSTALL_HEADER = $(INSTALL_DATA) transform = $(program_transform_name) NORMAL_INSTALL = : PRE_INSTALL = : POST_INSTALL = : NORMAL_UNINSTALL = : PRE_UNINSTALL = : POST_UNINSTALL = : build_triplet = @build@ host_triplet = @host@ @MPI_TRUE@noinst_PROGRAMS = mpi-bench$(EXEEXT) subdir = mpi DIST_COMMON = $(am__include_HEADERS_DIST) $(srcdir)/Makefile.am \ $(srcdir)/Makefile.in ACLOCAL_M4 = $(top_srcdir)/aclocal.m4 am__aclocal_m4_deps = $(top_srcdir)/m4/acx_mpi.m4 \ $(top_srcdir)/m4/acx_pthread.m4 \ $(top_srcdir)/m4/ax_cc_maxopt.m4 \ $(top_srcdir)/m4/ax_check_compiler_flags.m4 \ $(top_srcdir)/m4/ax_compiler_vendor.m4 \ $(top_srcdir)/m4/ax_gcc_aligns_stack.m4 \ $(top_srcdir)/m4/ax_gcc_version.m4 \ $(top_srcdir)/m4/ax_openmp.m4 $(top_srcdir)/m4/libtool.m4 \ $(top_srcdir)/m4/ltoptions.m4 $(top_srcdir)/m4/ltsugar.m4 \ $(top_srcdir)/m4/ltversion.m4 $(top_srcdir)/m4/lt~obsolete.m4 \ $(top_srcdir)/configure.ac am__configure_deps = $(am__aclocal_m4_deps) $(CONFIGURE_DEPENDENCIES) \ $(ACLOCAL_M4) mkinstalldirs = $(install_sh) -d CONFIG_HEADER = $(top_builddir)/config.h CONFIG_CLEAN_FILES = CONFIG_CLEAN_VPATH_FILES = am__vpath_adj_setup = srcdirstrip=`echo "$(srcdir)" | sed 's|.|.|g'`; am__vpath_adj = case $$p in \ $(srcdir)/*) f=`echo "$$p" | sed "s|^$$srcdirstrip/||"`;; \ *) f=$$p;; \ esac; am__strip_dir = f=`echo $$p | sed -e 's|^.*/||'`; am__install_max = 40 am__nobase_strip_setup = \ srcdirstrip=`echo "$(srcdir)" | sed 's/[].[^$$\\*|]/\\\\&/g'` am__nobase_strip = \ for p in $$list; do echo "$$p"; done | sed -e "s|$$srcdirstrip/||" am__nobase_list = $(am__nobase_strip_setup); \ for p in $$list; do echo "$$p $$p"; done | \ sed "s| $$srcdirstrip/| |;"' / .*\//!s/ .*/ ./; s,\( .*\)/[^/]*$$,\1,' | \ $(AWK) 'BEGIN { files["."] = "" } { files[$$2] = files[$$2] " " $$1; \ if (++n[$$2] == $(am__install_max)) \ { print $$2, files[$$2]; n[$$2] = 0; files[$$2] = "" } } \ END { for (dir in files) print dir, files[dir] }' am__base_list = \ sed '$$!N;$$!N;$$!N;$$!N;$$!N;$$!N;$$!N;s/\n/ /g' | \ sed '$$!N;$$!N;$$!N;$$!N;s/\n/ /g' am__uninstall_files_from_dir = { \ test -z "$$files" \ || { test ! -d "$$dir" && test ! -f "$$dir" && test ! -r "$$dir"; } \ || { echo " ( cd '$$dir' && rm -f" $$files ")"; \ $(am__cd) "$$dir" && rm -f $$files; }; \ } am__installdirs = "$(DESTDIR)$(libdir)" "$(DESTDIR)$(includedir)" \ "$(DESTDIR)$(includedir)" LTLIBRARIES = $(lib_LTLIBRARIES) libfftw3@PREC_SUFFIX@_mpi_la_DEPENDENCIES = \ ../libfftw3@PREC_SUFFIX@.la am__objects_1 = any-true.lo api.lo block.lo choose-radix.lo conf.lo \ dtensor.lo rearrange.lo wisdom-api.lo f03-wrap.lo am__objects_2 = transpose-alltoall.lo transpose-pairwise.lo \ transpose-recurse.lo transpose-problem.lo transpose-solve.lo am__objects_3 = dft-serial.lo dft-rank-geq2.lo \ dft-rank-geq2-transposed.lo dft-rank1.lo dft-rank1-bigvec.lo \ dft-problem.lo dft-solve.lo am__objects_4 = rdft-serial.lo rdft-rank-geq2.lo \ rdft-rank-geq2-transposed.lo rdft-rank1-bigvec.lo \ rdft-problem.lo rdft-solve.lo am__objects_5 = rdft2-serial.lo rdft2-rank-geq2.lo \ rdft2-rank-geq2-transposed.lo rdft2-problem.lo rdft2-solve.lo am_libfftw3@PREC_SUFFIX@_mpi_la_OBJECTS = $(am__objects_1) \ $(am__objects_2) $(am__objects_3) $(am__objects_4) \ $(am__objects_5) libfftw3@PREC_SUFFIX@_mpi_la_OBJECTS = \ $(am_libfftw3@PREC_SUFFIX@_mpi_la_OBJECTS) libfftw3@PREC_SUFFIX@_mpi_la_LINK = $(LIBTOOL) --tag=CC \ $(AM_LIBTOOLFLAGS) $(LIBTOOLFLAGS) --mode=link $(CCLD) \ $(AM_CFLAGS) $(CFLAGS) $(libfftw3@PREC_SUFFIX@_mpi_la_LDFLAGS) \ $(LDFLAGS) -o $@ @MPI_TRUE@am_libfftw3@PREC_SUFFIX@_mpi_la_rpath = -rpath $(libdir) PROGRAMS = $(noinst_PROGRAMS) am_mpi_bench_OBJECTS = mpi_bench-mpi-bench.$(OBJEXT) \ mpi_bench-fftw-bench.$(OBJEXT) mpi_bench-hook.$(OBJEXT) mpi_bench_OBJECTS = $(am_mpi_bench_OBJECTS) am__DEPENDENCIES_1 = mpi_bench_DEPENDENCIES = libfftw3@PREC_SUFFIX@_mpi.la \ $(LIBFFTWTHREADS) $(top_builddir)/libfftw3@PREC_SUFFIX@.la \ $(top_builddir)/libbench2/libbench2.a $(am__DEPENDENCIES_1) \ $(am__DEPENDENCIES_1) mpi_bench_LINK = $(LIBTOOL) --tag=CC $(AM_LIBTOOLFLAGS) \ $(LIBTOOLFLAGS) --mode=link $(CCLD) $(mpi_bench_CFLAGS) \ $(CFLAGS) $(AM_LDFLAGS) $(LDFLAGS) -o $@ DEFAULT_INCLUDES = -I.@am__isrc@ -I$(top_builddir) depcomp = $(SHELL) $(top_srcdir)/depcomp am__depfiles_maybe = depfiles am__mv = mv -f COMPILE = $(CC) $(DEFS) $(DEFAULT_INCLUDES) $(INCLUDES) $(AM_CPPFLAGS) \ $(CPPFLAGS) $(AM_CFLAGS) $(CFLAGS) LTCOMPILE = $(LIBTOOL) --tag=CC $(AM_LIBTOOLFLAGS) $(LIBTOOLFLAGS) \ --mode=compile $(CC) $(DEFS) $(DEFAULT_INCLUDES) $(INCLUDES) \ $(AM_CPPFLAGS) $(CPPFLAGS) $(AM_CFLAGS) $(CFLAGS) CCLD = $(CC) LINK = $(LIBTOOL) --tag=CC $(AM_LIBTOOLFLAGS) $(LIBTOOLFLAGS) \ --mode=link $(CCLD) $(AM_CFLAGS) $(CFLAGS) $(AM_LDFLAGS) \ $(LDFLAGS) -o $@ SOURCES = $(libfftw3@PREC_SUFFIX@_mpi_la_SOURCES) $(mpi_bench_SOURCES) DIST_SOURCES = $(libfftw3@PREC_SUFFIX@_mpi_la_SOURCES) \ $(mpi_bench_SOURCES) am__can_run_installinfo = \ case $$AM_UPDATE_INFO_DIR in \ n|no|NO) false;; \ *) (install-info --version) >/dev/null 2>&1;; \ esac am__include_HEADERS_DIST = fftw3-mpi.h HEADERS = $(include_HEADERS) $(nodist_include_HEADERS) ETAGS = etags CTAGS = ctags DISTFILES = $(DIST_COMMON) $(DIST_SOURCES) $(TEXINFOS) $(EXTRA_DIST) ACLOCAL = @ACLOCAL@ ALLOCA = @ALLOCA@ ALTIVEC_CFLAGS = @ALTIVEC_CFLAGS@ AMTAR = @AMTAR@ AR = @AR@ AS = @AS@ AUTOCONF = @AUTOCONF@ AUTOHEADER = @AUTOHEADER@ AUTOMAKE = @AUTOMAKE@ AVX_CFLAGS = @AVX_CFLAGS@ AWK = @AWK@ CC = @MPICC@ CCDEPMODE = @CCDEPMODE@ CFLAGS = @CFLAGS@ CHECK_PL_OPTS = @CHECK_PL_OPTS@ CPP = @CPP@ CPPFLAGS = @CPPFLAGS@ CYGPATH_W = @CYGPATH_W@ C_FFTW_R2R_KIND = @C_FFTW_R2R_KIND@ C_MPI_FINT = @C_MPI_FINT@ DEFS = @DEFS@ DEPDIR = @DEPDIR@ DLLTOOL = @DLLTOOL@ DSYMUTIL = @DSYMUTIL@ DUMPBIN = @DUMPBIN@ ECHO_C = @ECHO_C@ ECHO_N = @ECHO_N@ ECHO_T = @ECHO_T@ EGREP = @EGREP@ EXEEXT = @EXEEXT@ F77 = @F77@ FFLAGS = @FFLAGS@ FGREP = @FGREP@ FLIBS = @FLIBS@ GREP = @GREP@ INSTALL = @INSTALL@ INSTALL_DATA = @INSTALL_DATA@ INSTALL_PROGRAM = @INSTALL_PROGRAM@ INSTALL_SCRIPT = @INSTALL_SCRIPT@ INSTALL_STRIP_PROGRAM = @INSTALL_STRIP_PROGRAM@ LD = @LD@ LDFLAGS = @LDFLAGS@ LIBOBJS = @LIBOBJS@ LIBQUADMATH = @LIBQUADMATH@ LIBS = @LIBS@ LIBTOOL = @LIBTOOL@ LIPO = @LIPO@ LN_S = @LN_S@ LTLIBOBJS = @LTLIBOBJS@ MAINT = @MAINT@ MAKEINFO = @MAKEINFO@ MANIFEST_TOOL = @MANIFEST_TOOL@ MKDIR_P = @MKDIR_P@ MPICC = @MPICC@ MPILIBS = @MPILIBS@ MPIRUN = @MPIRUN@ NEON_CFLAGS = @NEON_CFLAGS@ NM = @NM@ NMEDIT = @NMEDIT@ OBJDUMP = @OBJDUMP@ OBJEXT = @OBJEXT@ OCAMLBUILD = @OCAMLBUILD@ OPENMP_CFLAGS = @OPENMP_CFLAGS@ OTOOL = @OTOOL@ OTOOL64 = @OTOOL64@ PACKAGE = @PACKAGE@ PACKAGE_BUGREPORT = @PACKAGE_BUGREPORT@ PACKAGE_NAME = @PACKAGE_NAME@ PACKAGE_STRING = @PACKAGE_STRING@ PACKAGE_TARNAME = @PACKAGE_TARNAME@ PACKAGE_URL = @PACKAGE_URL@ PACKAGE_VERSION = @PACKAGE_VERSION@ PATH_SEPARATOR = @PATH_SEPARATOR@ POW_LIB = @POW_LIB@ PRECISION = @PRECISION@ PREC_SUFFIX = @PREC_SUFFIX@ PTHREAD_CC = @PTHREAD_CC@ PTHREAD_CFLAGS = @PTHREAD_CFLAGS@ PTHREAD_LIBS = @PTHREAD_LIBS@ RANLIB = @RANLIB@ SED = @SED@ SET_MAKE = @SET_MAKE@ SHARED_VERSION_INFO = @SHARED_VERSION_INFO@ SHELL = @SHELL@ SSE2_CFLAGS = @SSE2_CFLAGS@ STACK_ALIGN_CFLAGS = @STACK_ALIGN_CFLAGS@ STRIP = @STRIP@ THREADLIBS = @THREADLIBS@ VERSION = @VERSION@ abs_builddir = @abs_builddir@ abs_srcdir = @abs_srcdir@ abs_top_builddir = @abs_top_builddir@ abs_top_srcdir = @abs_top_srcdir@ ac_ct_AR = @ac_ct_AR@ ac_ct_CC = @ac_ct_CC@ ac_ct_DUMPBIN = @ac_ct_DUMPBIN@ ac_ct_F77 = @ac_ct_F77@ acx_pthread_config = @acx_pthread_config@ am__include = @am__include@ am__leading_dot = @am__leading_dot@ am__quote = @am__quote@ am__tar = @am__tar@ am__untar = @am__untar@ bindir = @bindir@ build = @build@ build_alias = @build_alias@ build_cpu = @build_cpu@ build_os = @build_os@ build_vendor = @build_vendor@ builddir = @builddir@ datadir = @datadir@ datarootdir = @datarootdir@ docdir = @docdir@ dvidir = @dvidir@ exec_prefix = @exec_prefix@ host = @host@ host_alias = @host_alias@ host_cpu = @host_cpu@ host_os = @host_os@ host_vendor = @host_vendor@ htmldir = @htmldir@ includedir = @includedir@ infodir = @infodir@ install_sh = @install_sh@ libdir = @libdir@ libexecdir = @libexecdir@ localedir = @localedir@ localstatedir = @localstatedir@ mandir = @mandir@ mkdir_p = @mkdir_p@ oldincludedir = @oldincludedir@ pdfdir = @pdfdir@ prefix = @prefix@ program_transform_name = @program_transform_name@ psdir = @psdir@ sbindir = @sbindir@ sharedstatedir = @sharedstatedir@ srcdir = @srcdir@ sysconfdir = @sysconfdir@ target_alias = @target_alias@ top_build_prefix = @top_build_prefix@ top_builddir = @top_builddir@ top_srcdir = @top_srcdir@ AM_CPPFLAGS = -I$(top_srcdir)/kernel -I$(top_srcdir)/dft \ -I$(top_srcdir)/rdft -I$(top_srcdir)/api -I$(top_srcdir)/tests \ -I$(top_srcdir)/libbench2 @MPI_TRUE@lib_LTLIBRARIES = libfftw3@PREC_SUFFIX@_mpi.la @MPI_TRUE@include_HEADERS = fftw3-mpi.h @MPI_TRUE@nodist_include_HEADERS = fftw3-mpi.f03 fftw3l-mpi.f03 EXTRA_DIST = testsched.c f03api.sh f03-wrap.sh genf03-wrap.pl fftw3-mpi.f03.in fftw3l-mpi.f03.in BUILT_SOURCES = fftw3-mpi.f03.in fftw3-mpi.f03 fftw3l-mpi.f03.in fftw3l-mpi.f03 f03-wrap.c CLEANFILES = fftw3-mpi.f03 fftw3l-mpi.f03 TRANSPOSE_SRC = transpose-alltoall.c transpose-pairwise.c transpose-recurse.c transpose-problem.c transpose-solve.c mpi-transpose.h DFT_SRC = dft-serial.c dft-rank-geq2.c dft-rank-geq2-transposed.c dft-rank1.c dft-rank1-bigvec.c dft-problem.c dft-solve.c mpi-dft.h RDFT_SRC = rdft-serial.c rdft-rank-geq2.c rdft-rank-geq2-transposed.c rdft-rank1-bigvec.c rdft-problem.c rdft-solve.c mpi-rdft.h RDFT2_SRC = rdft2-serial.c rdft2-rank-geq2.c rdft2-rank-geq2-transposed.c rdft2-problem.c rdft2-solve.c mpi-rdft2.h SRC = any-true.c api.c block.c choose-radix.c conf.c dtensor.c fftw3-mpi.h ifftw-mpi.h rearrange.c wisdom-api.c f03-wrap.c libfftw3@PREC_SUFFIX@_mpi_la_SOURCES = $(SRC) $(TRANSPOSE_SRC) $(DFT_SRC) $(RDFT_SRC) $(RDFT2_SRC) libfftw3@PREC_SUFFIX@_mpi_la_LDFLAGS = -version-info @SHARED_VERSION_INFO@ libfftw3@PREC_SUFFIX@_mpi_la_LIBADD = ../libfftw3@PREC_SUFFIX@.la @MPILIBS@ @OPENMP_TRUE@@THREADS_FALSE@mpi_bench_CFLAGS = $(OPENMP_CFLAGS) @THREADS_TRUE@mpi_bench_CFLAGS = $(PTHREAD_CFLAGS) @COMBINED_THREADS_FALSE@@THREADS_TRUE@LIBFFTWTHREADS = $(top_builddir)/threads/libfftw3@PREC_SUFFIX@_threads.la @OPENMP_TRUE@@THREADS_FALSE@LIBFFTWTHREADS = $(top_builddir)/threads/libfftw3@PREC_SUFFIX@_omp.la mpi_bench_SOURCES = mpi-bench.c $(top_srcdir)/tests/fftw-bench.c $(top_srcdir)/tests/hook.c mpi_bench_LDADD = libfftw3@PREC_SUFFIX@_mpi.la $(LIBFFTWTHREADS) $(top_builddir)/libfftw3@PREC_SUFFIX@.la $(top_builddir)/libbench2/libbench2.a $(MPILIBS) $(THREADLIBS) CHECK = $(top_srcdir)/tests/check.pl NUMCHECK = 10 CHECKSIZE = 10000 CHECKOPTS = --verbose --random --maxsize=$(CHECKSIZE) -c=$(NUMCHECK) $(CHECK_PL_OPTS) all: $(BUILT_SOURCES) $(MAKE) $(AM_MAKEFLAGS) all-am .SUFFIXES: .SUFFIXES: .c .lo .o .obj $(srcdir)/Makefile.in: @MAINTAINER_MODE_TRUE@ $(srcdir)/Makefile.am $(am__configure_deps) @for dep in $?; do \ case '$(am__configure_deps)' in \ *$$dep*) \ ( cd $(top_builddir) && $(MAKE) $(AM_MAKEFLAGS) am--refresh ) \ && { if test -f $@; then exit 0; else break; fi; }; \ exit 1;; \ esac; \ done; \ echo ' cd $(top_srcdir) && $(AUTOMAKE) --gnu mpi/Makefile'; \ $(am__cd) $(top_srcdir) && \ $(AUTOMAKE) --gnu mpi/Makefile .PRECIOUS: Makefile Makefile: $(srcdir)/Makefile.in $(top_builddir)/config.status @case '$?' in \ *config.status*) \ cd $(top_builddir) && $(MAKE) $(AM_MAKEFLAGS) am--refresh;; \ *) \ echo ' cd $(top_builddir) && $(SHELL) ./config.status $(subdir)/$@ $(am__depfiles_maybe)'; \ cd $(top_builddir) && $(SHELL) ./config.status $(subdir)/$@ $(am__depfiles_maybe);; \ esac; $(top_builddir)/config.status: $(top_srcdir)/configure $(CONFIG_STATUS_DEPENDENCIES) cd $(top_builddir) && $(MAKE) $(AM_MAKEFLAGS) am--refresh $(top_srcdir)/configure: @MAINTAINER_MODE_TRUE@ $(am__configure_deps) cd $(top_builddir) && $(MAKE) $(AM_MAKEFLAGS) am--refresh $(ACLOCAL_M4): @MAINTAINER_MODE_TRUE@ $(am__aclocal_m4_deps) cd $(top_builddir) && $(MAKE) $(AM_MAKEFLAGS) am--refresh $(am__aclocal_m4_deps): install-libLTLIBRARIES: $(lib_LTLIBRARIES) @$(NORMAL_INSTALL) @list='$(lib_LTLIBRARIES)'; test -n "$(libdir)" || list=; \ list2=; for p in $$list; do \ if test -f $$p; then \ list2="$$list2 $$p"; \ else :; fi; \ done; \ test -z "$$list2" || { \ echo " $(MKDIR_P) '$(DESTDIR)$(libdir)'"; \ $(MKDIR_P) "$(DESTDIR)$(libdir)" || exit 1; \ echo " $(LIBTOOL) $(AM_LIBTOOLFLAGS) $(LIBTOOLFLAGS) --mode=install $(INSTALL) $(INSTALL_STRIP_FLAG) $$list2 '$(DESTDIR)$(libdir)'"; \ $(LIBTOOL) $(AM_LIBTOOLFLAGS) $(LIBTOOLFLAGS) --mode=install $(INSTALL) $(INSTALL_STRIP_FLAG) $$list2 "$(DESTDIR)$(libdir)"; \ } uninstall-libLTLIBRARIES: @$(NORMAL_UNINSTALL) @list='$(lib_LTLIBRARIES)'; test -n "$(libdir)" || list=; \ for p in $$list; do \ $(am__strip_dir) \ echo " $(LIBTOOL) $(AM_LIBTOOLFLAGS) $(LIBTOOLFLAGS) --mode=uninstall rm -f '$(DESTDIR)$(libdir)/$$f'"; \ $(LIBTOOL) $(AM_LIBTOOLFLAGS) $(LIBTOOLFLAGS) --mode=uninstall rm -f "$(DESTDIR)$(libdir)/$$f"; \ done clean-libLTLIBRARIES: -test -z "$(lib_LTLIBRARIES)" || rm -f $(lib_LTLIBRARIES) @list='$(lib_LTLIBRARIES)'; for p in $$list; do \ dir="`echo $$p | sed -e 's|/[^/]*$$||'`"; \ test "$$dir" != "$$p" || dir=.; \ echo "rm -f \"$${dir}/so_locations\""; \ rm -f "$${dir}/so_locations"; \ done libfftw3@PREC_SUFFIX@_mpi.la: $(libfftw3@PREC_SUFFIX@_mpi_la_OBJECTS) $(libfftw3@PREC_SUFFIX@_mpi_la_DEPENDENCIES) $(EXTRA_libfftw3@PREC_SUFFIX@_mpi_la_DEPENDENCIES) $(libfftw3@PREC_SUFFIX@_mpi_la_LINK) $(am_libfftw3@PREC_SUFFIX@_mpi_la_rpath) $(libfftw3@PREC_SUFFIX@_mpi_la_OBJECTS) $(libfftw3@PREC_SUFFIX@_mpi_la_LIBADD) $(LIBS) clean-noinstPROGRAMS: @list='$(noinst_PROGRAMS)'; test -n "$$list" || exit 0; \ echo " rm -f" $$list; \ rm -f $$list || exit $$?; \ test -n "$(EXEEXT)" || exit 0; \ list=`for p in $$list; do echo "$$p"; done | sed 's/$(EXEEXT)$$//'`; \ echo " rm -f" $$list; \ rm -f $$list mpi-bench$(EXEEXT): $(mpi_bench_OBJECTS) $(mpi_bench_DEPENDENCIES) $(EXTRA_mpi_bench_DEPENDENCIES) @rm -f mpi-bench$(EXEEXT) $(mpi_bench_LINK) $(mpi_bench_OBJECTS) $(mpi_bench_LDADD) $(LIBS) mostlyclean-compile: -rm -f *.$(OBJEXT) distclean-compile: -rm -f *.tab.c @AMDEP_TRUE@@am__include@ @am__quote@./$(DEPDIR)/any-true.Plo@am__quote@ @AMDEP_TRUE@@am__include@ @am__quote@./$(DEPDIR)/api.Plo@am__quote@ @AMDEP_TRUE@@am__include@ @am__quote@./$(DEPDIR)/block.Plo@am__quote@ @AMDEP_TRUE@@am__include@ @am__quote@./$(DEPDIR)/choose-radix.Plo@am__quote@ @AMDEP_TRUE@@am__include@ @am__quote@./$(DEPDIR)/conf.Plo@am__quote@ @AMDEP_TRUE@@am__include@ @am__quote@./$(DEPDIR)/dft-problem.Plo@am__quote@ @AMDEP_TRUE@@am__include@ @am__quote@./$(DEPDIR)/dft-rank-geq2-transposed.Plo@am__quote@ @AMDEP_TRUE@@am__include@ @am__quote@./$(DEPDIR)/dft-rank-geq2.Plo@am__quote@ @AMDEP_TRUE@@am__include@ @am__quote@./$(DEPDIR)/dft-rank1-bigvec.Plo@am__quote@ @AMDEP_TRUE@@am__include@ @am__quote@./$(DEPDIR)/dft-rank1.Plo@am__quote@ @AMDEP_TRUE@@am__include@ @am__quote@./$(DEPDIR)/dft-serial.Plo@am__quote@ @AMDEP_TRUE@@am__include@ @am__quote@./$(DEPDIR)/dft-solve.Plo@am__quote@ @AMDEP_TRUE@@am__include@ @am__quote@./$(DEPDIR)/dtensor.Plo@am__quote@ @AMDEP_TRUE@@am__include@ @am__quote@./$(DEPDIR)/f03-wrap.Plo@am__quote@ @AMDEP_TRUE@@am__include@ @am__quote@./$(DEPDIR)/mpi_bench-fftw-bench.Po@am__quote@ @AMDEP_TRUE@@am__include@ @am__quote@./$(DEPDIR)/mpi_bench-hook.Po@am__quote@ @AMDEP_TRUE@@am__include@ @am__quote@./$(DEPDIR)/mpi_bench-mpi-bench.Po@am__quote@ @AMDEP_TRUE@@am__include@ @am__quote@./$(DEPDIR)/rdft-problem.Plo@am__quote@ @AMDEP_TRUE@@am__include@ @am__quote@./$(DEPDIR)/rdft-rank-geq2-transposed.Plo@am__quote@ @AMDEP_TRUE@@am__include@ @am__quote@./$(DEPDIR)/rdft-rank-geq2.Plo@am__quote@ @AMDEP_TRUE@@am__include@ @am__quote@./$(DEPDIR)/rdft-rank1-bigvec.Plo@am__quote@ @AMDEP_TRUE@@am__include@ @am__quote@./$(DEPDIR)/rdft-serial.Plo@am__quote@ @AMDEP_TRUE@@am__include@ @am__quote@./$(DEPDIR)/rdft-solve.Plo@am__quote@ @AMDEP_TRUE@@am__include@ @am__quote@./$(DEPDIR)/rdft2-problem.Plo@am__quote@ @AMDEP_TRUE@@am__include@ @am__quote@./$(DEPDIR)/rdft2-rank-geq2-transposed.Plo@am__quote@ @AMDEP_TRUE@@am__include@ @am__quote@./$(DEPDIR)/rdft2-rank-geq2.Plo@am__quote@ @AMDEP_TRUE@@am__include@ @am__quote@./$(DEPDIR)/rdft2-serial.Plo@am__quote@ @AMDEP_TRUE@@am__include@ @am__quote@./$(DEPDIR)/rdft2-solve.Plo@am__quote@ @AMDEP_TRUE@@am__include@ @am__quote@./$(DEPDIR)/rearrange.Plo@am__quote@ @AMDEP_TRUE@@am__include@ @am__quote@./$(DEPDIR)/transpose-alltoall.Plo@am__quote@ @AMDEP_TRUE@@am__include@ @am__quote@./$(DEPDIR)/transpose-pairwise.Plo@am__quote@ @AMDEP_TRUE@@am__include@ @am__quote@./$(DEPDIR)/transpose-problem.Plo@am__quote@ @AMDEP_TRUE@@am__include@ @am__quote@./$(DEPDIR)/transpose-recurse.Plo@am__quote@ @AMDEP_TRUE@@am__include@ @am__quote@./$(DEPDIR)/transpose-solve.Plo@am__quote@ @AMDEP_TRUE@@am__include@ @am__quote@./$(DEPDIR)/wisdom-api.Plo@am__quote@ .c.o: @am__fastdepCC_TRUE@ $(COMPILE) -MT $@ -MD -MP -MF $(DEPDIR)/$*.Tpo -c -o $@ $< @am__fastdepCC_TRUE@ $(am__mv) $(DEPDIR)/$*.Tpo $(DEPDIR)/$*.Po @AMDEP_TRUE@@am__fastdepCC_FALSE@ source='$<' object='$@' libtool=no @AMDEPBACKSLASH@ @AMDEP_TRUE@@am__fastdepCC_FALSE@ DEPDIR=$(DEPDIR) $(CCDEPMODE) $(depcomp) @AMDEPBACKSLASH@ @am__fastdepCC_FALSE@ $(COMPILE) -c $< .c.obj: @am__fastdepCC_TRUE@ $(COMPILE) -MT $@ -MD -MP -MF $(DEPDIR)/$*.Tpo -c -o $@ `$(CYGPATH_W) '$<'` @am__fastdepCC_TRUE@ $(am__mv) $(DEPDIR)/$*.Tpo $(DEPDIR)/$*.Po @AMDEP_TRUE@@am__fastdepCC_FALSE@ source='$<' object='$@' libtool=no @AMDEPBACKSLASH@ @AMDEP_TRUE@@am__fastdepCC_FALSE@ DEPDIR=$(DEPDIR) $(CCDEPMODE) $(depcomp) @AMDEPBACKSLASH@ @am__fastdepCC_FALSE@ $(COMPILE) -c `$(CYGPATH_W) '$<'` .c.lo: @am__fastdepCC_TRUE@ $(LTCOMPILE) -MT $@ -MD -MP -MF $(DEPDIR)/$*.Tpo -c -o $@ $< @am__fastdepCC_TRUE@ $(am__mv) $(DEPDIR)/$*.Tpo $(DEPDIR)/$*.Plo @AMDEP_TRUE@@am__fastdepCC_FALSE@ source='$<' object='$@' libtool=yes @AMDEPBACKSLASH@ @AMDEP_TRUE@@am__fastdepCC_FALSE@ DEPDIR=$(DEPDIR) $(CCDEPMODE) $(depcomp) @AMDEPBACKSLASH@ @am__fastdepCC_FALSE@ $(LTCOMPILE) -c -o $@ $< mpi_bench-mpi-bench.o: mpi-bench.c @am__fastdepCC_TRUE@ $(CC) $(DEFS) $(DEFAULT_INCLUDES) $(INCLUDES) $(AM_CPPFLAGS) $(CPPFLAGS) $(mpi_bench_CFLAGS) $(CFLAGS) -MT mpi_bench-mpi-bench.o -MD -MP -MF $(DEPDIR)/mpi_bench-mpi-bench.Tpo -c -o mpi_bench-mpi-bench.o `test -f 'mpi-bench.c' || echo '$(srcdir)/'`mpi-bench.c @am__fastdepCC_TRUE@ $(am__mv) $(DEPDIR)/mpi_bench-mpi-bench.Tpo $(DEPDIR)/mpi_bench-mpi-bench.Po @AMDEP_TRUE@@am__fastdepCC_FALSE@ source='mpi-bench.c' object='mpi_bench-mpi-bench.o' libtool=no @AMDEPBACKSLASH@ @AMDEP_TRUE@@am__fastdepCC_FALSE@ DEPDIR=$(DEPDIR) $(CCDEPMODE) $(depcomp) @AMDEPBACKSLASH@ @am__fastdepCC_FALSE@ $(CC) $(DEFS) $(DEFAULT_INCLUDES) $(INCLUDES) $(AM_CPPFLAGS) $(CPPFLAGS) $(mpi_bench_CFLAGS) $(CFLAGS) -c -o mpi_bench-mpi-bench.o `test -f 'mpi-bench.c' || echo '$(srcdir)/'`mpi-bench.c mpi_bench-mpi-bench.obj: mpi-bench.c @am__fastdepCC_TRUE@ $(CC) $(DEFS) $(DEFAULT_INCLUDES) $(INCLUDES) $(AM_CPPFLAGS) $(CPPFLAGS) $(mpi_bench_CFLAGS) $(CFLAGS) -MT mpi_bench-mpi-bench.obj -MD -MP -MF $(DEPDIR)/mpi_bench-mpi-bench.Tpo -c -o mpi_bench-mpi-bench.obj `if test -f 'mpi-bench.c'; then $(CYGPATH_W) 'mpi-bench.c'; else $(CYGPATH_W) '$(srcdir)/mpi-bench.c'; fi` @am__fastdepCC_TRUE@ $(am__mv) $(DEPDIR)/mpi_bench-mpi-bench.Tpo $(DEPDIR)/mpi_bench-mpi-bench.Po @AMDEP_TRUE@@am__fastdepCC_FALSE@ source='mpi-bench.c' object='mpi_bench-mpi-bench.obj' libtool=no @AMDEPBACKSLASH@ @AMDEP_TRUE@@am__fastdepCC_FALSE@ DEPDIR=$(DEPDIR) $(CCDEPMODE) $(depcomp) @AMDEPBACKSLASH@ @am__fastdepCC_FALSE@ $(CC) $(DEFS) $(DEFAULT_INCLUDES) $(INCLUDES) $(AM_CPPFLAGS) $(CPPFLAGS) $(mpi_bench_CFLAGS) $(CFLAGS) -c -o mpi_bench-mpi-bench.obj `if test -f 'mpi-bench.c'; then $(CYGPATH_W) 'mpi-bench.c'; else $(CYGPATH_W) '$(srcdir)/mpi-bench.c'; fi` mpi_bench-fftw-bench.o: $(top_srcdir)/tests/fftw-bench.c @am__fastdepCC_TRUE@ $(CC) $(DEFS) $(DEFAULT_INCLUDES) $(INCLUDES) $(AM_CPPFLAGS) $(CPPFLAGS) $(mpi_bench_CFLAGS) $(CFLAGS) -MT mpi_bench-fftw-bench.o -MD -MP -MF $(DEPDIR)/mpi_bench-fftw-bench.Tpo -c -o mpi_bench-fftw-bench.o `test -f '$(top_srcdir)/tests/fftw-bench.c' || echo '$(srcdir)/'`$(top_srcdir)/tests/fftw-bench.c @am__fastdepCC_TRUE@ $(am__mv) $(DEPDIR)/mpi_bench-fftw-bench.Tpo $(DEPDIR)/mpi_bench-fftw-bench.Po @AMDEP_TRUE@@am__fastdepCC_FALSE@ source='$(top_srcdir)/tests/fftw-bench.c' object='mpi_bench-fftw-bench.o' libtool=no @AMDEPBACKSLASH@ @AMDEP_TRUE@@am__fastdepCC_FALSE@ DEPDIR=$(DEPDIR) $(CCDEPMODE) $(depcomp) @AMDEPBACKSLASH@ @am__fastdepCC_FALSE@ $(CC) $(DEFS) $(DEFAULT_INCLUDES) $(INCLUDES) $(AM_CPPFLAGS) $(CPPFLAGS) $(mpi_bench_CFLAGS) $(CFLAGS) -c -o mpi_bench-fftw-bench.o `test -f '$(top_srcdir)/tests/fftw-bench.c' || echo '$(srcdir)/'`$(top_srcdir)/tests/fftw-bench.c mpi_bench-fftw-bench.obj: $(top_srcdir)/tests/fftw-bench.c @am__fastdepCC_TRUE@ $(CC) $(DEFS) $(DEFAULT_INCLUDES) $(INCLUDES) $(AM_CPPFLAGS) $(CPPFLAGS) $(mpi_bench_CFLAGS) $(CFLAGS) -MT mpi_bench-fftw-bench.obj -MD -MP -MF $(DEPDIR)/mpi_bench-fftw-bench.Tpo -c -o mpi_bench-fftw-bench.obj `if test -f '$(top_srcdir)/tests/fftw-bench.c'; then $(CYGPATH_W) '$(top_srcdir)/tests/fftw-bench.c'; else $(CYGPATH_W) '$(srcdir)/$(top_srcdir)/tests/fftw-bench.c'; fi` @am__fastdepCC_TRUE@ $(am__mv) $(DEPDIR)/mpi_bench-fftw-bench.Tpo $(DEPDIR)/mpi_bench-fftw-bench.Po @AMDEP_TRUE@@am__fastdepCC_FALSE@ source='$(top_srcdir)/tests/fftw-bench.c' object='mpi_bench-fftw-bench.obj' libtool=no @AMDEPBACKSLASH@ @AMDEP_TRUE@@am__fastdepCC_FALSE@ DEPDIR=$(DEPDIR) $(CCDEPMODE) $(depcomp) @AMDEPBACKSLASH@ @am__fastdepCC_FALSE@ $(CC) $(DEFS) $(DEFAULT_INCLUDES) $(INCLUDES) $(AM_CPPFLAGS) $(CPPFLAGS) $(mpi_bench_CFLAGS) $(CFLAGS) -c -o mpi_bench-fftw-bench.obj `if test -f '$(top_srcdir)/tests/fftw-bench.c'; then $(CYGPATH_W) '$(top_srcdir)/tests/fftw-bench.c'; else $(CYGPATH_W) '$(srcdir)/$(top_srcdir)/tests/fftw-bench.c'; fi` mpi_bench-hook.o: $(top_srcdir)/tests/hook.c @am__fastdepCC_TRUE@ $(CC) $(DEFS) $(DEFAULT_INCLUDES) $(INCLUDES) $(AM_CPPFLAGS) $(CPPFLAGS) $(mpi_bench_CFLAGS) $(CFLAGS) -MT mpi_bench-hook.o -MD -MP -MF $(DEPDIR)/mpi_bench-hook.Tpo -c -o mpi_bench-hook.o `test -f '$(top_srcdir)/tests/hook.c' || echo '$(srcdir)/'`$(top_srcdir)/tests/hook.c @am__fastdepCC_TRUE@ $(am__mv) $(DEPDIR)/mpi_bench-hook.Tpo $(DEPDIR)/mpi_bench-hook.Po @AMDEP_TRUE@@am__fastdepCC_FALSE@ source='$(top_srcdir)/tests/hook.c' object='mpi_bench-hook.o' libtool=no @AMDEPBACKSLASH@ @AMDEP_TRUE@@am__fastdepCC_FALSE@ DEPDIR=$(DEPDIR) $(CCDEPMODE) $(depcomp) @AMDEPBACKSLASH@ @am__fastdepCC_FALSE@ $(CC) $(DEFS) $(DEFAULT_INCLUDES) $(INCLUDES) $(AM_CPPFLAGS) $(CPPFLAGS) $(mpi_bench_CFLAGS) $(CFLAGS) -c -o mpi_bench-hook.o `test -f '$(top_srcdir)/tests/hook.c' || echo '$(srcdir)/'`$(top_srcdir)/tests/hook.c mpi_bench-hook.obj: $(top_srcdir)/tests/hook.c @am__fastdepCC_TRUE@ $(CC) $(DEFS) $(DEFAULT_INCLUDES) $(INCLUDES) $(AM_CPPFLAGS) $(CPPFLAGS) $(mpi_bench_CFLAGS) $(CFLAGS) -MT mpi_bench-hook.obj -MD -MP -MF $(DEPDIR)/mpi_bench-hook.Tpo -c -o mpi_bench-hook.obj `if test -f '$(top_srcdir)/tests/hook.c'; then $(CYGPATH_W) '$(top_srcdir)/tests/hook.c'; else $(CYGPATH_W) '$(srcdir)/$(top_srcdir)/tests/hook.c'; fi` @am__fastdepCC_TRUE@ $(am__mv) $(DEPDIR)/mpi_bench-hook.Tpo $(DEPDIR)/mpi_bench-hook.Po @AMDEP_TRUE@@am__fastdepCC_FALSE@ source='$(top_srcdir)/tests/hook.c' object='mpi_bench-hook.obj' libtool=no @AMDEPBACKSLASH@ @AMDEP_TRUE@@am__fastdepCC_FALSE@ DEPDIR=$(DEPDIR) $(CCDEPMODE) $(depcomp) @AMDEPBACKSLASH@ @am__fastdepCC_FALSE@ $(CC) $(DEFS) $(DEFAULT_INCLUDES) $(INCLUDES) $(AM_CPPFLAGS) $(CPPFLAGS) $(mpi_bench_CFLAGS) $(CFLAGS) -c -o mpi_bench-hook.obj `if test -f '$(top_srcdir)/tests/hook.c'; then $(CYGPATH_W) '$(top_srcdir)/tests/hook.c'; else $(CYGPATH_W) '$(srcdir)/$(top_srcdir)/tests/hook.c'; fi` mostlyclean-libtool: -rm -f *.lo clean-libtool: -rm -rf .libs _libs install-includeHEADERS: $(include_HEADERS) @$(NORMAL_INSTALL) @list='$(include_HEADERS)'; test -n "$(includedir)" || list=; \ if test -n "$$list"; then \ echo " $(MKDIR_P) '$(DESTDIR)$(includedir)'"; \ $(MKDIR_P) "$(DESTDIR)$(includedir)" || exit 1; \ fi; \ for p in $$list; do \ if test -f "$$p"; then d=; else d="$(srcdir)/"; fi; \ echo "$$d$$p"; \ done | $(am__base_list) | \ while read files; do \ echo " $(INSTALL_HEADER) $$files '$(DESTDIR)$(includedir)'"; \ $(INSTALL_HEADER) $$files "$(DESTDIR)$(includedir)" || exit $$?; \ done uninstall-includeHEADERS: @$(NORMAL_UNINSTALL) @list='$(include_HEADERS)'; test -n "$(includedir)" || list=; \ files=`for p in $$list; do echo $$p; done | sed -e 's|^.*/||'`; \ dir='$(DESTDIR)$(includedir)'; $(am__uninstall_files_from_dir) install-nodist_includeHEADERS: $(nodist_include_HEADERS) @$(NORMAL_INSTALL) @list='$(nodist_include_HEADERS)'; test -n "$(includedir)" || list=; \ if test -n "$$list"; then \ echo " $(MKDIR_P) '$(DESTDIR)$(includedir)'"; \ $(MKDIR_P) "$(DESTDIR)$(includedir)" || exit 1; \ fi; \ for p in $$list; do \ if test -f "$$p"; then d=; else d="$(srcdir)/"; fi; \ echo "$$d$$p"; \ done | $(am__base_list) | \ while read files; do \ echo " $(INSTALL_HEADER) $$files '$(DESTDIR)$(includedir)'"; \ $(INSTALL_HEADER) $$files "$(DESTDIR)$(includedir)" || exit $$?; \ done uninstall-nodist_includeHEADERS: @$(NORMAL_UNINSTALL) @list='$(nodist_include_HEADERS)'; test -n "$(includedir)" || list=; \ files=`for p in $$list; do echo $$p; done | sed -e 's|^.*/||'`; \ dir='$(DESTDIR)$(includedir)'; $(am__uninstall_files_from_dir) ID: $(HEADERS) $(SOURCES) $(LISP) $(TAGS_FILES) list='$(SOURCES) $(HEADERS) $(LISP) $(TAGS_FILES)'; \ unique=`for i in $$list; do \ if test -f "$$i"; then echo $$i; else echo $(srcdir)/$$i; fi; \ done | \ $(AWK) '{ files[$$0] = 1; nonempty = 1; } \ END { if (nonempty) { for (i in files) print i; }; }'`; \ mkid -fID $$unique tags: TAGS TAGS: $(HEADERS) $(SOURCES) $(TAGS_DEPENDENCIES) \ $(TAGS_FILES) $(LISP) set x; \ here=`pwd`; \ list='$(SOURCES) $(HEADERS) $(LISP) $(TAGS_FILES)'; \ unique=`for i in $$list; do \ if test -f "$$i"; then echo $$i; else echo $(srcdir)/$$i; fi; \ done | \ $(AWK) '{ files[$$0] = 1; nonempty = 1; } \ END { if (nonempty) { for (i in files) print i; }; }'`; \ shift; \ if test -z "$(ETAGS_ARGS)$$*$$unique"; then :; else \ test -n "$$unique" || unique=$$empty_fix; \ if test $$# -gt 0; then \ $(ETAGS) $(ETAGSFLAGS) $(AM_ETAGSFLAGS) $(ETAGS_ARGS) \ "$$@" $$unique; \ else \ $(ETAGS) $(ETAGSFLAGS) $(AM_ETAGSFLAGS) $(ETAGS_ARGS) \ $$unique; \ fi; \ fi ctags: CTAGS CTAGS: $(HEADERS) $(SOURCES) $(TAGS_DEPENDENCIES) \ $(TAGS_FILES) $(LISP) list='$(SOURCES) $(HEADERS) $(LISP) $(TAGS_FILES)'; \ unique=`for i in $$list; do \ if test -f "$$i"; then echo $$i; else echo $(srcdir)/$$i; fi; \ done | \ $(AWK) '{ files[$$0] = 1; nonempty = 1; } \ END { if (nonempty) { for (i in files) print i; }; }'`; \ test -z "$(CTAGS_ARGS)$$unique" \ || $(CTAGS) $(CTAGSFLAGS) $(AM_CTAGSFLAGS) $(CTAGS_ARGS) \ $$unique GTAGS: here=`$(am__cd) $(top_builddir) && pwd` \ && $(am__cd) $(top_srcdir) \ && gtags -i $(GTAGS_ARGS) "$$here" distclean-tags: -rm -f TAGS ID GTAGS GRTAGS GSYMS GPATH tags distdir: $(DISTFILES) @srcdirstrip=`echo "$(srcdir)" | sed 's/[].[^$$\\*]/\\\\&/g'`; \ topsrcdirstrip=`echo "$(top_srcdir)" | sed 's/[].[^$$\\*]/\\\\&/g'`; \ list='$(DISTFILES)'; \ dist_files=`for file in $$list; do echo $$file; done | \ sed -e "s|^$$srcdirstrip/||;t" \ -e "s|^$$topsrcdirstrip/|$(top_builddir)/|;t"`; \ case $$dist_files in \ */*) $(MKDIR_P) `echo "$$dist_files" | \ sed '/\//!d;s|^|$(distdir)/|;s,/[^/]*$$,,' | \ sort -u` ;; \ esac; \ for file in $$dist_files; do \ if test -f $$file || test -d $$file; then d=.; else d=$(srcdir); fi; \ if test -d $$d/$$file; then \ dir=`echo "/$$file" | sed -e 's,/[^/]*$$,,'`; \ if test -d "$(distdir)/$$file"; then \ find "$(distdir)/$$file" -type d ! -perm -700 -exec chmod u+rwx {} \;; \ fi; \ if test -d $(srcdir)/$$file && test $$d != $(srcdir); then \ cp -fpR $(srcdir)/$$file "$(distdir)$$dir" || exit 1; \ find "$(distdir)/$$file" -type d ! -perm -700 -exec chmod u+rwx {} \;; \ fi; \ cp -fpR $$d/$$file "$(distdir)$$dir" || exit 1; \ else \ test -f "$(distdir)/$$file" \ || cp -p $$d/$$file "$(distdir)/$$file" \ || exit 1; \ fi; \ done @MPI_FALSE@check-local: check-am: all-am $(MAKE) $(AM_MAKEFLAGS) check-local check: $(BUILT_SOURCES) $(MAKE) $(AM_MAKEFLAGS) check-am all-am: Makefile $(LTLIBRARIES) $(PROGRAMS) $(HEADERS) installdirs: for dir in "$(DESTDIR)$(libdir)" "$(DESTDIR)$(includedir)" "$(DESTDIR)$(includedir)"; do \ test -z "$$dir" || $(MKDIR_P) "$$dir"; \ done install: $(BUILT_SOURCES) $(MAKE) $(AM_MAKEFLAGS) install-am install-exec: install-exec-am install-data: install-data-am uninstall: uninstall-am install-am: all-am @$(MAKE) $(AM_MAKEFLAGS) install-exec-am install-data-am installcheck: installcheck-am install-strip: if test -z '$(STRIP)'; then \ $(MAKE) $(AM_MAKEFLAGS) INSTALL_PROGRAM="$(INSTALL_STRIP_PROGRAM)" \ install_sh_PROGRAM="$(INSTALL_STRIP_PROGRAM)" INSTALL_STRIP_FLAG=-s \ install; \ else \ $(MAKE) $(AM_MAKEFLAGS) INSTALL_PROGRAM="$(INSTALL_STRIP_PROGRAM)" \ install_sh_PROGRAM="$(INSTALL_STRIP_PROGRAM)" INSTALL_STRIP_FLAG=-s \ "INSTALL_PROGRAM_ENV=STRIPPROG='$(STRIP)'" install; \ fi mostlyclean-generic: clean-generic: -test -z "$(CLEANFILES)" || rm -f $(CLEANFILES) distclean-generic: -test -z "$(CONFIG_CLEAN_FILES)" || rm -f $(CONFIG_CLEAN_FILES) -test . = "$(srcdir)" || test -z "$(CONFIG_CLEAN_VPATH_FILES)" || rm -f $(CONFIG_CLEAN_VPATH_FILES) maintainer-clean-generic: @echo "This command is intended for maintainers to use" @echo "it deletes files that may require special tools to rebuild." -test -z "$(BUILT_SOURCES)" || rm -f $(BUILT_SOURCES) clean: clean-am clean-am: clean-generic clean-libLTLIBRARIES clean-libtool \ clean-noinstPROGRAMS mostlyclean-am distclean: distclean-am -rm -rf ./$(DEPDIR) -rm -f Makefile distclean-am: clean-am distclean-compile distclean-generic \ distclean-tags dvi: dvi-am dvi-am: html: html-am html-am: info: info-am info-am: install-data-am: install-includeHEADERS install-nodist_includeHEADERS install-dvi: install-dvi-am install-dvi-am: install-exec-am: install-libLTLIBRARIES install-html: install-html-am install-html-am: install-info: install-info-am install-info-am: install-man: install-pdf: install-pdf-am install-pdf-am: install-ps: install-ps-am install-ps-am: installcheck-am: maintainer-clean: maintainer-clean-am -rm -rf ./$(DEPDIR) -rm -f Makefile maintainer-clean-am: distclean-am maintainer-clean-generic mostlyclean: mostlyclean-am mostlyclean-am: mostlyclean-compile mostlyclean-generic \ mostlyclean-libtool pdf: pdf-am pdf-am: ps: ps-am ps-am: uninstall-am: uninstall-includeHEADERS uninstall-libLTLIBRARIES \ uninstall-nodist_includeHEADERS .MAKE: all check check-am install install-am install-strip .PHONY: CTAGS GTAGS all all-am check check-am check-local clean \ clean-generic clean-libLTLIBRARIES clean-libtool \ clean-noinstPROGRAMS ctags distclean distclean-compile \ distclean-generic distclean-libtool distclean-tags distdir dvi \ dvi-am html html-am info info-am install install-am \ install-data install-data-am install-dvi install-dvi-am \ install-exec install-exec-am install-html install-html-am \ install-includeHEADERS install-info install-info-am \ install-libLTLIBRARIES install-man \ install-nodist_includeHEADERS install-pdf install-pdf-am \ install-ps install-ps-am install-strip installcheck \ installcheck-am installdirs maintainer-clean \ maintainer-clean-generic mostlyclean mostlyclean-compile \ mostlyclean-generic mostlyclean-libtool pdf pdf-am ps ps-am \ tags uninstall uninstall-am uninstall-includeHEADERS \ uninstall-libLTLIBRARIES uninstall-nodist_includeHEADERS @MPI_TRUE@check-local: mpi-bench$(EXEEXT) @MPI_TRUE@ perl -w $(CHECK) $(CHECKOPTS) --mpi "$(MPIRUN) -np 1 `pwd`/mpi-bench" @MPI_TRUE@ @echo "--------------------------------------------------------------" @MPI_TRUE@ @echo " MPI FFTW transforms passed "$(NUMCHECK)" tests, 1 CPU" @MPI_TRUE@ @echo "--------------------------------------------------------------" @MPI_TRUE@ perl -w $(CHECK) $(CHECKOPTS) --mpi "$(MPIRUN) -np 2 `pwd`/mpi-bench" @MPI_TRUE@ @echo "--------------------------------------------------------------" @MPI_TRUE@ @echo " MPI FFTW transforms passed "$(NUMCHECK)" tests, 2 CPUs" @MPI_TRUE@ @echo "--------------------------------------------------------------" @MPI_TRUE@ perl -w $(CHECK) $(CHECKOPTS) --mpi "$(MPIRUN) -np 3 `pwd`/mpi-bench" @MPI_TRUE@ @echo "--------------------------------------------------------------" @MPI_TRUE@ @echo " MPI FFTW transforms passed "$(NUMCHECK)" tests, 3 CPUs" @MPI_TRUE@ @echo "--------------------------------------------------------------" @MPI_TRUE@ perl -w $(CHECK) $(CHECKOPTS) --mpi "$(MPIRUN) -np 4 `pwd`/mpi-bench" @MPI_TRUE@ @echo "--------------------------------------------------------------" @MPI_TRUE@ @echo " MPI FFTW transforms passed "$(NUMCHECK)" tests, 4 CPUs" @MPI_TRUE@ @echo "--------------------------------------------------------------" @MPI_TRUE@@SMP_TRUE@ perl -w $(CHECK) $(CHECKOPTS) --mpi --nthreads=2 "$(MPIRUN) -np 3 `pwd`/mpi-bench" @MPI_TRUE@@SMP_TRUE@ @echo "--------------------------------------------------------------" @MPI_TRUE@@SMP_TRUE@ @echo " MPI FFTW threaded transforms passed "$(NUMCHECK)" tests!" @MPI_TRUE@@SMP_TRUE@ @echo "--------------------------------------------------------------" @MPI_TRUE@bigcheck: mpi-bench$(EXEEXT) @MPI_TRUE@ $(MAKE) $(AM_MAKEFLAGS) NUMCHECK=100 CHECKSIZE=60000 check-local @MPI_TRUE@smallcheck: mpi-bench$(EXEEXT) @MPI_TRUE@ $(MAKE) $(AM_MAKEFLAGS) NUMCHECK=2 check-local fftw3-mpi.f03: fftw3-mpi.f03.in sed 's/C_MPI_FINT/@C_MPI_FINT@/' $(srcdir)/fftw3-mpi.f03.in > $@ fftw3l-mpi.f03: fftw3l-mpi.f03.in sed 's/C_MPI_FINT/@C_MPI_FINT@/' $(srcdir)/fftw3l-mpi.f03.in > $@ @MAINTAINER_MODE_TRUE@fftw3-mpi.f03.in: fftw3-mpi.h f03api.sh $(top_srcdir)/api/genf03.pl @MAINTAINER_MODE_TRUE@ sh $(srcdir)/f03api.sh d f > $@ @MAINTAINER_MODE_TRUE@fftw3l-mpi.f03.in: fftw3-mpi.h f03api.sh $(top_srcdir)/api/genf03.pl @MAINTAINER_MODE_TRUE@ sh $(srcdir)/f03api.sh l | grep -v parameter | sed 's/fftw3.f03/fftw3l.f03/' > $@ @MAINTAINER_MODE_TRUE@f03-wrap.c: fftw3-mpi.h f03-wrap.sh genf03-wrap.pl @MAINTAINER_MODE_TRUE@ sh $(srcdir)/f03-wrap.sh > $@ # Tell versions [3.59,3.63) of GNU make to not export all variables. # Otherwise a system limit (for SysV at least) may be exceeded. .NOEXPORT: fftw-3.3.3/mpi/any-true.c0000644000175000017500000000357612054410050012106 00000000000000/* * Copyright (c) 2003, 2007-11 Matteo Frigo * Copyright (c) 2003, 2007-11 Massachusetts Institute of Technology * * This program is free software; you can redistribute it and/or modify * it under the terms of the GNU General Public License as published by * the Free Software Foundation; either version 2 of the License, or * (at your option) any later version. * * This program is distributed in the hope that it will be useful, * but WITHOUT ANY WARRANTY; without even the implied warranty of * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the * GNU General Public License for more details. * * You should have received a copy of the GNU General Public License * along with this program; if not, write to the Free Software * Foundation, Inc., 51 Franklin Street, Fifth Floor, Boston, MA 02110-1301 USA * */ #include "ifftw-mpi.h" /* During planning, if any process fails to create a plan then all of the processes must fail. This synchronization is implemented by the following routine. Instead of if (failure) goto nada; we instead do: if (any_true(failure, comm)) goto nada; */ int XM(any_true)(int condition, MPI_Comm comm) { int result; MPI_Allreduce(&condition, &result, 1, MPI_INT, MPI_LOR, comm); return result; } /***********************************************************************/ #if defined(FFTW_DEBUG) /* for debugging, we include an assertion to make sure that MPI problems all produce equal hashes, as checked by this routine: */ int XM(md5_equal)(md5 m, MPI_Comm comm) { unsigned long s0[4]; int i, eq_me, eq_all; X(md5end)(&m); for (i = 0; i < 4; ++i) s0[i] = m.s[i]; MPI_Bcast(s0, 4, MPI_UNSIGNED_LONG, 0, comm); for (i = 0; i < 4 && s0[i] == m.s[i]; ++i) ; eq_me = i == 4; MPI_Allreduce(&eq_me, &eq_all, 1, MPI_INT, MPI_LAND, comm); return eq_all; } #endif fftw-3.3.3/mpi/api.c0000644000175000017500000006451012054410050011106 00000000000000/* * Copyright (c) 2003, 2007-11 Matteo Frigo * Copyright (c) 2003, 2007-11 Massachusetts Institute of Technology * * This program is free software; you can redistribute it and/or modify * it under the terms of the GNU General Public License as published by * the Free Software Foundation; either version 2 of the License, or * (at your option) any later version. * * This program is distributed in the hope that it will be useful, * but WITHOUT ANY WARRANTY; without even the implied warranty of * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the * GNU General Public License for more details. * * You should have received a copy of the GNU General Public License * along with this program; if not, write to the Free Software * Foundation, Inc., 51 Franklin Street, Fifth Floor, Boston, MA 02110-1301 USA * */ #include "api.h" #include "fftw3-mpi.h" #include "ifftw-mpi.h" #include "mpi-transpose.h" #include "mpi-dft.h" #include "mpi-rdft.h" #include "mpi-rdft2.h" /* Convert API flags to internal MPI flags. */ #define MPI_FLAGS(f) ((f) >> 27) /*************************************************************************/ static int mpi_inited = 0; static MPI_Comm problem_comm(const problem *p) { switch (p->adt->problem_kind) { case PROBLEM_MPI_DFT: return ((const problem_mpi_dft *) p)->comm; case PROBLEM_MPI_RDFT: return ((const problem_mpi_rdft *) p)->comm; case PROBLEM_MPI_RDFT2: return ((const problem_mpi_rdft2 *) p)->comm; case PROBLEM_MPI_TRANSPOSE: return ((const problem_mpi_transpose *) p)->comm; default: return MPI_COMM_NULL; } } /* used to synchronize cost measurements (timing or estimation) across all processes for an MPI problem, which is critical to ensure that all processes decide to use the same MPI plans (whereas serial plans need not be syncronized). */ static double cost_hook(const problem *p, double t, cost_kind k) { MPI_Comm comm = problem_comm(p); double tsum; if (comm == MPI_COMM_NULL) return t; MPI_Allreduce(&t, &tsum, 1, MPI_DOUBLE, k == COST_SUM ? MPI_SUM : MPI_MAX, comm); return tsum; } /* Used to reject wisdom that is not in sync across all processes for an MPI problem, which is critical to ensure that all processes decide to use the same MPI plans. (Even though costs are synchronized, above, out-of-sync wisdom may result from plans being produced by communicators that do not span all processes, either from a user-specified communicator or e.g. from transpose-recurse. */ static int wisdom_ok_hook(const problem *p, flags_t flags) { MPI_Comm comm = problem_comm(p); int eq_me, eq_all; /* unpack flags bitfield, since MPI communications may involve byte-order changes and MPI cannot do this for bit fields */ #if SIZEOF_UNSIGNED_INT >= 4 /* must be big enough to hold 20-bit fields */ unsigned int f[5]; #else unsigned long f[5]; /* at least 32 bits as per C standard */ #endif if (comm == MPI_COMM_NULL) return 1; /* non-MPI wisdom is always ok */ if (XM(any_true)(0, comm)) return 0; /* some process had nowisdom_hook */ /* otherwise, check that the flags and solver index are identical on all processes in this problem's communicator. TO DO: possibly we can relax strict equality, but it is critical to ensure that any flags which affect what plan is created (and whether the solver is applicable) are the same, e.g. DESTROY_INPUT, NO_UGLY, etcetera. (If the MPI algorithm differs between processes, deadlocks/crashes generally result.) */ f[0] = flags.l; f[1] = flags.hash_info; f[2] = flags.timelimit_impatience; f[3] = flags.u; f[4] = flags.slvndx; MPI_Bcast(f, 5, SIZEOF_UNSIGNED_INT >= 4 ? MPI_UNSIGNED : MPI_UNSIGNED_LONG, 0, comm); eq_me = f[0] == flags.l && f[1] == flags.hash_info && f[2] == flags.timelimit_impatience && f[3] == flags.u && f[4] == flags.slvndx; MPI_Allreduce(&eq_me, &eq_all, 1, MPI_INT, MPI_LAND, comm); return eq_all; } /* This hook is called when wisdom is not found. The any_true here matches up with the any_true in wisdom_ok_hook, in order to handle the case where some processes had wisdom (and called wisdom_ok_hook) and some processes didn't have wisdom (and called nowisdom_hook). */ static void nowisdom_hook(const problem *p) { MPI_Comm comm = problem_comm(p); if (comm == MPI_COMM_NULL) return; /* nothing to do for non-MPI p */ XM(any_true)(1, comm); /* signal nowisdom to any wisdom_ok_hook */ } /* needed to synchronize planner bogosity flag, in case non-MPI problems on a subset of processes encountered bogus wisdom */ static wisdom_state_t bogosity_hook(wisdom_state_t state, const problem *p) { MPI_Comm comm = problem_comm(p); if (comm != MPI_COMM_NULL /* an MPI problem */ && XM(any_true)(state == WISDOM_IS_BOGUS, comm)) /* bogus somewhere */ return WISDOM_IS_BOGUS; return state; } void XM(init)(void) { if (!mpi_inited) { planner *plnr = X(the_planner)(); plnr->cost_hook = cost_hook; plnr->wisdom_ok_hook = wisdom_ok_hook; plnr->nowisdom_hook = nowisdom_hook; plnr->bogosity_hook = bogosity_hook; XM(conf_standard)(plnr); mpi_inited = 1; } } void XM(cleanup)(void) { X(cleanup)(); mpi_inited = 0; } /*************************************************************************/ static dtensor *mkdtensor_api(int rnk, const XM(ddim) *dims0) { dtensor *x = XM(mkdtensor)(rnk); int i; for (i = 0; i < rnk; ++i) { x->dims[i].n = dims0[i].n; x->dims[i].b[IB] = dims0[i].ib; x->dims[i].b[OB] = dims0[i].ob; } return x; } static dtensor *default_sz(int rnk, const XM(ddim) *dims0, int n_pes, int rdft2) { dtensor *sz = XM(mkdtensor)(rnk); dtensor *sz0 = mkdtensor_api(rnk, dims0); block_kind k; int i; for (i = 0; i < rnk; ++i) sz->dims[i].n = dims0[i].n; if (rdft2) sz->dims[rnk-1].n = dims0[rnk-1].n / 2 + 1; for (i = 0; i < rnk; ++i) { sz->dims[i].b[IB] = dims0[i].ib ? dims0[i].ib : sz->dims[i].n; sz->dims[i].b[OB] = dims0[i].ob ? dims0[i].ob : sz->dims[i].n; } /* If we haven't used all of the processes yet, and some of the block sizes weren't specified (i.e. 0), then set the unspecified blocks so as to use as many processes as possible with as few distributed dimensions as possible. */ FORALL_BLOCK_KIND(k) { INT nb = XM(num_blocks_total)(sz, k); INT np = n_pes / nb; for (i = 0; i < rnk && np > 1; ++i) if (!sz0->dims[i].b[k]) { sz->dims[i].b[k] = XM(default_block)(sz->dims[i].n, np); nb *= XM(num_blocks)(sz->dims[i].n, sz->dims[i].b[k]); np = n_pes / nb; } } if (rdft2) sz->dims[rnk-1].n = dims0[rnk-1].n; /* punt for 1d prime */ if (rnk == 1 && X(is_prime)(sz->dims[0].n)) sz->dims[0].b[IB] = sz->dims[0].b[OB] = sz->dims[0].n; XM(dtensor_destroy)(sz0); sz0 = XM(dtensor_canonical)(sz, 0); XM(dtensor_destroy)(sz); return sz0; } /* allocate simple local (serial) dims array corresponding to n[rnk] */ static XM(ddim) *simple_dims(int rnk, const ptrdiff_t *n) { XM(ddim) *dims = (XM(ddim) *) MALLOC(sizeof(XM(ddim)) * rnk, TENSORS); int i; for (i = 0; i < rnk; ++i) dims[i].n = dims[i].ib = dims[i].ob = n[i]; return dims; } /*************************************************************************/ static void local_size(int my_pe, const dtensor *sz, block_kind k, ptrdiff_t *local_n, ptrdiff_t *local_start) { int i; if (my_pe >= XM(num_blocks_total)(sz, k)) for (i = 0; i < sz->rnk; ++i) local_n[i] = local_start[i] = 0; else { XM(block_coords)(sz, k, my_pe, local_start); for (i = 0; i < sz->rnk; ++i) { local_n[i] = XM(block)(sz->dims[i].n, sz->dims[i].b[k], local_start[i]); local_start[i] *= sz->dims[i].b[k]; } } } static INT prod(int rnk, const ptrdiff_t *local_n) { int i; INT N = 1; for (i = 0; i < rnk; ++i) N *= local_n[i]; return N; } ptrdiff_t XM(local_size_guru)(int rnk, const XM(ddim) *dims0, ptrdiff_t howmany, MPI_Comm comm, ptrdiff_t *local_n_in, ptrdiff_t *local_start_in, ptrdiff_t *local_n_out, ptrdiff_t *local_start_out, int sign, unsigned flags) { INT N; int my_pe, n_pes, i; dtensor *sz; if (rnk == 0) return howmany; MPI_Comm_rank(comm, &my_pe); MPI_Comm_size(comm, &n_pes); sz = default_sz(rnk, dims0, n_pes, 0); /* Now, we must figure out how much local space the user should allocate (or at least an upper bound). This depends strongly on the exact algorithms we employ...ugh! FIXME: get this info from the solvers somehow? */ N = 1; /* never return zero allocation size */ if (rnk > 1 && XM(is_block1d)(sz, IB) && XM(is_block1d)(sz, OB)) { INT Nafter; ddim odims[2]; /* dft-rank-geq2-transposed */ odims[0] = sz->dims[0]; odims[1] = sz->dims[1]; /* save */ /* we may need extra space for transposed intermediate data */ for (i = 0; i < 2; ++i) if (XM(num_blocks)(sz->dims[i].n, sz->dims[i].b[IB]) == 1 && XM(num_blocks)(sz->dims[i].n, sz->dims[i].b[OB]) == 1) { sz->dims[i].b[IB] = XM(default_block)(sz->dims[i].n, n_pes); sz->dims[1-i].b[IB] = sz->dims[1-i].n; local_size(my_pe, sz, IB, local_n_in, local_start_in); N = X(imax)(N, prod(rnk, local_n_in)); sz->dims[i] = odims[i]; sz->dims[1-i] = odims[1-i]; break; } /* dft-rank-geq2 */ Nafter = howmany; for (i = 1; i < sz->rnk; ++i) Nafter *= sz->dims[i].n; N = X(imax)(N, (sz->dims[0].n * XM(block)(Nafter, XM(default_block)(Nafter, n_pes), my_pe) + howmany - 1) / howmany); /* dft-rank-geq2 with dimensions swapped */ Nafter = howmany * sz->dims[0].n; for (i = 2; i < sz->rnk; ++i) Nafter *= sz->dims[i].n; N = X(imax)(N, (sz->dims[1].n * XM(block)(Nafter, XM(default_block)(Nafter, n_pes), my_pe) + howmany - 1) / howmany); } else if (rnk == 1) { if (howmany >= n_pes && !MPI_FLAGS(flags)) { /* dft-rank1-bigvec */ ptrdiff_t n[2], start[2]; dtensor *sz2 = XM(mkdtensor)(2); sz2->dims[0] = sz->dims[0]; sz2->dims[0].b[IB] = sz->dims[0].n; sz2->dims[1].n = sz2->dims[1].b[OB] = howmany; sz2->dims[1].b[IB] = XM(default_block)(howmany, n_pes); local_size(my_pe, sz2, IB, n, start); XM(dtensor_destroy)(sz2); N = X(imax)(N, (prod(2, n) + howmany - 1) / howmany); } else { /* dft-rank1 */ INT r, m, rblock[2], mblock[2]; /* Since the 1d transforms are so different, we require the user to call local_size_1d for this case. Ugh. */ CK(sign == FFTW_FORWARD || sign == FFTW_BACKWARD); if ((r = XM(choose_radix)(sz->dims[0], n_pes, flags, sign, rblock, mblock))) { m = sz->dims[0].n / r; if (flags & FFTW_MPI_SCRAMBLED_IN) sz->dims[0].b[IB] = rblock[IB] * m; else { /* !SCRAMBLED_IN */ sz->dims[0].b[IB] = r * mblock[IB]; N = X(imax)(N, rblock[IB] * m); } if (flags & FFTW_MPI_SCRAMBLED_OUT) sz->dims[0].b[OB] = r * mblock[OB]; else { /* !SCRAMBLED_OUT */ N = X(imax)(N, r * mblock[OB]); sz->dims[0].b[OB] = rblock[OB] * m; } } } } local_size(my_pe, sz, IB, local_n_in, local_start_in); local_size(my_pe, sz, OB, local_n_out, local_start_out); /* at least, make sure we have enough space to store input & output */ N = X(imax)(N, X(imax)(prod(rnk, local_n_in), prod(rnk, local_n_out))); XM(dtensor_destroy)(sz); return N * howmany; } ptrdiff_t XM(local_size_many_transposed)(int rnk, const ptrdiff_t *n, ptrdiff_t howmany, ptrdiff_t xblock, ptrdiff_t yblock, MPI_Comm comm, ptrdiff_t *local_nx, ptrdiff_t *local_x_start, ptrdiff_t *local_ny, ptrdiff_t *local_y_start) { ptrdiff_t N; XM(ddim) *dims; ptrdiff_t *local; if (rnk == 0) { *local_nx = *local_ny = 1; *local_x_start = *local_y_start = 0; return howmany; } dims = simple_dims(rnk, n); local = (ptrdiff_t *) MALLOC(sizeof(ptrdiff_t) * rnk * 4, TENSORS); /* default 1d block distribution, with transposed output if yblock < n[1] */ dims[0].ib = xblock; if (rnk > 1) { if (yblock < n[1]) dims[1].ob = yblock; else dims[0].ob = xblock; } else dims[0].ob = xblock; /* FIXME: 1d not really supported here since we don't have flags/sign */ N = XM(local_size_guru)(rnk, dims, howmany, comm, local, local + rnk, local + 2*rnk, local + 3*rnk, 0, 0); *local_nx = local[0]; *local_x_start = local[rnk]; if (rnk > 1) { *local_ny = local[2*rnk + 1]; *local_y_start = local[3*rnk + 1]; } else { *local_ny = *local_nx; *local_y_start = *local_x_start; } X(ifree)(local); X(ifree)(dims); return N; } ptrdiff_t XM(local_size_many)(int rnk, const ptrdiff_t *n, ptrdiff_t howmany, ptrdiff_t xblock, MPI_Comm comm, ptrdiff_t *local_nx, ptrdiff_t *local_x_start) { ptrdiff_t local_ny, local_y_start; return XM(local_size_many_transposed)(rnk, n, howmany, xblock, rnk > 1 ? n[1] : FFTW_MPI_DEFAULT_BLOCK, comm, local_nx, local_x_start, &local_ny, &local_y_start); } ptrdiff_t XM(local_size_transposed)(int rnk, const ptrdiff_t *n, MPI_Comm comm, ptrdiff_t *local_nx, ptrdiff_t *local_x_start, ptrdiff_t *local_ny, ptrdiff_t *local_y_start) { return XM(local_size_many_transposed)(rnk, n, 1, FFTW_MPI_DEFAULT_BLOCK, FFTW_MPI_DEFAULT_BLOCK, comm, local_nx, local_x_start, local_ny, local_y_start); } ptrdiff_t XM(local_size)(int rnk, const ptrdiff_t *n, MPI_Comm comm, ptrdiff_t *local_nx, ptrdiff_t *local_x_start) { return XM(local_size_many)(rnk, n, 1, FFTW_MPI_DEFAULT_BLOCK, comm, local_nx, local_x_start); } ptrdiff_t XM(local_size_many_1d)(ptrdiff_t nx, ptrdiff_t howmany, MPI_Comm comm, int sign, unsigned flags, ptrdiff_t *local_nx, ptrdiff_t *local_x_start, ptrdiff_t *local_ny, ptrdiff_t *local_y_start) { XM(ddim) d; d.n = nx; d.ib = d.ob = FFTW_MPI_DEFAULT_BLOCK; return XM(local_size_guru)(1, &d, howmany, comm, local_nx, local_x_start, local_ny, local_y_start, sign, flags); } ptrdiff_t XM(local_size_1d)(ptrdiff_t nx, MPI_Comm comm, int sign, unsigned flags, ptrdiff_t *local_nx, ptrdiff_t *local_x_start, ptrdiff_t *local_ny, ptrdiff_t *local_y_start) { return XM(local_size_many_1d)(nx, 1, comm, sign, flags, local_nx, local_x_start, local_ny, local_y_start); } ptrdiff_t XM(local_size_2d_transposed)(ptrdiff_t nx, ptrdiff_t ny, MPI_Comm comm, ptrdiff_t *local_nx, ptrdiff_t *local_x_start, ptrdiff_t *local_ny, ptrdiff_t *local_y_start) { ptrdiff_t n[2]; n[0] = nx; n[1] = ny; return XM(local_size_transposed)(2, n, comm, local_nx, local_x_start, local_ny, local_y_start); } ptrdiff_t XM(local_size_2d)(ptrdiff_t nx, ptrdiff_t ny, MPI_Comm comm, ptrdiff_t *local_nx, ptrdiff_t *local_x_start) { ptrdiff_t n[2]; n[0] = nx; n[1] = ny; return XM(local_size)(2, n, comm, local_nx, local_x_start); } ptrdiff_t XM(local_size_3d_transposed)(ptrdiff_t nx, ptrdiff_t ny, ptrdiff_t nz, MPI_Comm comm, ptrdiff_t *local_nx, ptrdiff_t *local_x_start, ptrdiff_t *local_ny, ptrdiff_t *local_y_start) { ptrdiff_t n[3]; n[0] = nx; n[1] = ny; n[2] = nz; return XM(local_size_transposed)(3, n, comm, local_nx, local_x_start, local_ny, local_y_start); } ptrdiff_t XM(local_size_3d)(ptrdiff_t nx, ptrdiff_t ny, ptrdiff_t nz, MPI_Comm comm, ptrdiff_t *local_nx, ptrdiff_t *local_x_start) { ptrdiff_t n[3]; n[0] = nx; n[1] = ny; n[2] = nz; return XM(local_size)(3, n, comm, local_nx, local_x_start); } /*************************************************************************/ /* Transpose API */ X(plan) XM(plan_many_transpose)(ptrdiff_t nx, ptrdiff_t ny, ptrdiff_t howmany, ptrdiff_t xblock, ptrdiff_t yblock, R *in, R *out, MPI_Comm comm, unsigned flags) { int n_pes; XM(init)(); if (howmany < 0 || xblock < 0 || yblock < 0 || nx <= 0 || ny <= 0) return 0; MPI_Comm_size(comm, &n_pes); if (!xblock) xblock = XM(default_block)(nx, n_pes); if (!yblock) yblock = XM(default_block)(ny, n_pes); if (n_pes < XM(num_blocks)(nx, xblock) || n_pes < XM(num_blocks)(ny, yblock)) return 0; return X(mkapiplan)(FFTW_FORWARD, flags, XM(mkproblem_transpose)(nx, ny, howmany, in, out, xblock, yblock, comm, MPI_FLAGS(flags))); } X(plan) XM(plan_transpose)(ptrdiff_t nx, ptrdiff_t ny, R *in, R *out, MPI_Comm comm, unsigned flags) { return XM(plan_many_transpose)(nx, ny, 1, FFTW_MPI_DEFAULT_BLOCK, FFTW_MPI_DEFAULT_BLOCK, in, out, comm, flags); } /*************************************************************************/ /* Complex DFT API */ X(plan) XM(plan_guru_dft)(int rnk, const XM(ddim) *dims0, ptrdiff_t howmany, C *in, C *out, MPI_Comm comm, int sign, unsigned flags) { int n_pes, i; dtensor *sz; XM(init)(); if (howmany < 0 || rnk < 1) return 0; for (i = 0; i < rnk; ++i) if (dims0[i].n < 1 || dims0[i].ib < 0 || dims0[i].ob < 0) return 0; MPI_Comm_size(comm, &n_pes); sz = default_sz(rnk, dims0, n_pes, 0); if (XM(num_blocks_total)(sz, IB) > n_pes || XM(num_blocks_total)(sz, OB) > n_pes) { XM(dtensor_destroy)(sz); return 0; } return X(mkapiplan)(sign, flags, XM(mkproblem_dft_d)(sz, howmany, (R *) in, (R *) out, comm, sign, MPI_FLAGS(flags))); } X(plan) XM(plan_many_dft)(int rnk, const ptrdiff_t *n, ptrdiff_t howmany, ptrdiff_t iblock, ptrdiff_t oblock, C *in, C *out, MPI_Comm comm, int sign, unsigned flags) { XM(ddim) *dims = simple_dims(rnk, n); X(plan) pln; if (rnk == 1) { dims[0].ib = iblock; dims[0].ob = oblock; } else if (rnk > 1) { dims[0 != (flags & FFTW_MPI_TRANSPOSED_IN)].ib = iblock; dims[0 != (flags & FFTW_MPI_TRANSPOSED_OUT)].ob = oblock; } pln = XM(plan_guru_dft)(rnk,dims,howmany, in,out, comm, sign, flags); X(ifree)(dims); return pln; } X(plan) XM(plan_dft)(int rnk, const ptrdiff_t *n, C *in, C *out, MPI_Comm comm, int sign, unsigned flags) { return XM(plan_many_dft)(rnk, n, 1, FFTW_MPI_DEFAULT_BLOCK, FFTW_MPI_DEFAULT_BLOCK, in, out, comm, sign, flags); } X(plan) XM(plan_dft_1d)(ptrdiff_t nx, C *in, C *out, MPI_Comm comm, int sign, unsigned flags) { return XM(plan_dft)(1, &nx, in, out, comm, sign, flags); } X(plan) XM(plan_dft_2d)(ptrdiff_t nx, ptrdiff_t ny, C *in, C *out, MPI_Comm comm, int sign, unsigned flags) { ptrdiff_t n[2]; n[0] = nx; n[1] = ny; return XM(plan_dft)(2, n, in, out, comm, sign, flags); } X(plan) XM(plan_dft_3d)(ptrdiff_t nx, ptrdiff_t ny, ptrdiff_t nz, C *in, C *out, MPI_Comm comm, int sign, unsigned flags) { ptrdiff_t n[3]; n[0] = nx; n[1] = ny; n[2] = nz; return XM(plan_dft)(3, n, in, out, comm, sign, flags); } /*************************************************************************/ /* R2R API */ X(plan) XM(plan_guru_r2r)(int rnk, const XM(ddim) *dims0, ptrdiff_t howmany, R *in, R *out, MPI_Comm comm, const X(r2r_kind) *kind, unsigned flags) { int n_pes, i; dtensor *sz; rdft_kind *k; X(plan) pln; XM(init)(); if (howmany < 0 || rnk < 1) return 0; for (i = 0; i < rnk; ++i) if (dims0[i].n < 1 || dims0[i].ib < 0 || dims0[i].ob < 0) return 0; k = X(map_r2r_kind)(rnk, kind); MPI_Comm_size(comm, &n_pes); sz = default_sz(rnk, dims0, n_pes, 0); if (XM(num_blocks_total)(sz, IB) > n_pes || XM(num_blocks_total)(sz, OB) > n_pes) { XM(dtensor_destroy)(sz); return 0; } pln = X(mkapiplan)(0, flags, XM(mkproblem_rdft_d)(sz, howmany, in, out, comm, k, MPI_FLAGS(flags))); X(ifree0)(k); return pln; } X(plan) XM(plan_many_r2r)(int rnk, const ptrdiff_t *n, ptrdiff_t howmany, ptrdiff_t iblock, ptrdiff_t oblock, R *in, R *out, MPI_Comm comm, const X(r2r_kind) *kind, unsigned flags) { XM(ddim) *dims = simple_dims(rnk, n); X(plan) pln; if (rnk == 1) { dims[0].ib = iblock; dims[0].ob = oblock; } else if (rnk > 1) { dims[0 != (flags & FFTW_MPI_TRANSPOSED_IN)].ib = iblock; dims[0 != (flags & FFTW_MPI_TRANSPOSED_OUT)].ob = oblock; } pln = XM(plan_guru_r2r)(rnk,dims,howmany, in,out, comm, kind, flags); X(ifree)(dims); return pln; } X(plan) XM(plan_r2r)(int rnk, const ptrdiff_t *n, R *in, R *out, MPI_Comm comm, const X(r2r_kind) *kind, unsigned flags) { return XM(plan_many_r2r)(rnk, n, 1, FFTW_MPI_DEFAULT_BLOCK, FFTW_MPI_DEFAULT_BLOCK, in, out, comm, kind, flags); } X(plan) XM(plan_r2r_2d)(ptrdiff_t nx, ptrdiff_t ny, R *in, R *out, MPI_Comm comm, X(r2r_kind) kindx, X(r2r_kind) kindy, unsigned flags) { ptrdiff_t n[2]; X(r2r_kind) kind[2]; n[0] = nx; n[1] = ny; kind[0] = kindx; kind[1] = kindy; return XM(plan_r2r)(2, n, in, out, comm, kind, flags); } X(plan) XM(plan_r2r_3d)(ptrdiff_t nx, ptrdiff_t ny, ptrdiff_t nz, R *in, R *out, MPI_Comm comm, X(r2r_kind) kindx, X(r2r_kind) kindy, X(r2r_kind) kindz, unsigned flags) { ptrdiff_t n[3]; X(r2r_kind) kind[3]; n[0] = nx; n[1] = ny; n[2] = nz; kind[0] = kindx; kind[1] = kindy; kind[2] = kindz; return XM(plan_r2r)(3, n, in, out, comm, kind, flags); } /*************************************************************************/ /* R2C/C2R API */ static X(plan) plan_guru_rdft2(int rnk, const XM(ddim) *dims0, ptrdiff_t howmany, R *r, C *c, MPI_Comm comm, rdft_kind kind, unsigned flags) { int n_pes, i; dtensor *sz; R *cr = (R *) c; XM(init)(); if (howmany < 0 || rnk < 2) return 0; for (i = 0; i < rnk; ++i) if (dims0[i].n < 1 || dims0[i].ib < 0 || dims0[i].ob < 0) return 0; MPI_Comm_size(comm, &n_pes); sz = default_sz(rnk, dims0, n_pes, 1); sz->dims[rnk-1].n = dims0[rnk-1].n / 2 + 1; if (XM(num_blocks_total)(sz, IB) > n_pes || XM(num_blocks_total)(sz, OB) > n_pes) { XM(dtensor_destroy)(sz); return 0; } sz->dims[rnk-1].n = dims0[rnk-1].n; if (kind == R2HC) return X(mkapiplan)(0, flags, XM(mkproblem_rdft2_d)(sz, howmany, r, cr, comm, R2HC, MPI_FLAGS(flags))); else return X(mkapiplan)(0, flags, XM(mkproblem_rdft2_d)(sz, howmany, cr, r, comm, HC2R, MPI_FLAGS(flags))); } X(plan) XM(plan_many_dft_r2c)(int rnk, const ptrdiff_t *n, ptrdiff_t howmany, ptrdiff_t iblock, ptrdiff_t oblock, R *in, C *out, MPI_Comm comm, unsigned flags) { XM(ddim) *dims = simple_dims(rnk, n); X(plan) pln; if (rnk == 1) { dims[0].ib = iblock; dims[0].ob = oblock; } else if (rnk > 1) { dims[0 != (flags & FFTW_MPI_TRANSPOSED_IN)].ib = iblock; dims[0 != (flags & FFTW_MPI_TRANSPOSED_OUT)].ob = oblock; } pln = plan_guru_rdft2(rnk,dims,howmany, in,out, comm, R2HC, flags); X(ifree)(dims); return pln; } X(plan) XM(plan_many_dft_c2r)(int rnk, const ptrdiff_t *n, ptrdiff_t howmany, ptrdiff_t iblock, ptrdiff_t oblock, C *in, R *out, MPI_Comm comm, unsigned flags) { XM(ddim) *dims = simple_dims(rnk, n); X(plan) pln; if (rnk == 1) { dims[0].ib = iblock; dims[0].ob = oblock; } else if (rnk > 1) { dims[0 != (flags & FFTW_MPI_TRANSPOSED_IN)].ib = iblock; dims[0 != (flags & FFTW_MPI_TRANSPOSED_OUT)].ob = oblock; } pln = plan_guru_rdft2(rnk,dims,howmany, out,in, comm, HC2R, flags); X(ifree)(dims); return pln; } X(plan) XM(plan_dft_r2c)(int rnk, const ptrdiff_t *n, R *in, C *out, MPI_Comm comm, unsigned flags) { return XM(plan_many_dft_r2c)(rnk, n, 1, FFTW_MPI_DEFAULT_BLOCK, FFTW_MPI_DEFAULT_BLOCK, in, out, comm, flags); } X(plan) XM(plan_dft_r2c_2d)(ptrdiff_t nx, ptrdiff_t ny, R *in, C *out, MPI_Comm comm, unsigned flags) { ptrdiff_t n[2]; n[0] = nx; n[1] = ny; return XM(plan_dft_r2c)(2, n, in, out, comm, flags); } X(plan) XM(plan_dft_r2c_3d)(ptrdiff_t nx, ptrdiff_t ny, ptrdiff_t nz, R *in, C *out, MPI_Comm comm, unsigned flags) { ptrdiff_t n[3]; n[0] = nx; n[1] = ny; n[2] = nz; return XM(plan_dft_r2c)(3, n, in, out, comm, flags); } X(plan) XM(plan_dft_c2r)(int rnk, const ptrdiff_t *n, C *in, R *out, MPI_Comm comm, unsigned flags) { return XM(plan_many_dft_c2r)(rnk, n, 1, FFTW_MPI_DEFAULT_BLOCK, FFTW_MPI_DEFAULT_BLOCK, in, out, comm, flags); } X(plan) XM(plan_dft_c2r_2d)(ptrdiff_t nx, ptrdiff_t ny, C *in, R *out, MPI_Comm comm, unsigned flags) { ptrdiff_t n[2]; n[0] = nx; n[1] = ny; return XM(plan_dft_c2r)(2, n, in, out, comm, flags); } X(plan) XM(plan_dft_c2r_3d)(ptrdiff_t nx, ptrdiff_t ny, ptrdiff_t nz, C *in, R *out, MPI_Comm comm, unsigned flags) { ptrdiff_t n[3]; n[0] = nx; n[1] = ny; n[2] = nz; return XM(plan_dft_c2r)(3, n, in, out, comm, flags); } /*************************************************************************/ /* New-array execute functions */ void XM(execute_dft)(const X(plan) p, C *in, C *out) { /* internally, MPI plans are just rdft plans */ X(execute_r2r)(p, (R*) in, (R*) out); } void XM(execute_dft_r2c)(const X(plan) p, R *in, C *out) { /* internally, MPI plans are just rdft plans */ X(execute_r2r)(p, in, (R*) out); } void XM(execute_dft_c2r)(const X(plan) p, C *in, R *out) { /* internally, MPI plans are just rdft plans */ X(execute_r2r)(p, (R*) in, out); } void XM(execute_r2r)(const X(plan) p, R *in, R *out) { /* internally, MPI plans are just rdft plans */ X(execute_r2r)(p, in, out); } fftw-3.3.3/mpi/block.c0000644000175000017500000000746512054410050011435 00000000000000/* * Copyright (c) 2003, 2007-11 Matteo Frigo * Copyright (c) 2003, 2007-11 Massachusetts Institute of Technology * * This program is free software; you can redistribute it and/or modify * it under the terms of the GNU General Public License as published by * the Free Software Foundation; either version 2 of the License, or * (at your option) any later version. * * This program is distributed in the hope that it will be useful, * but WITHOUT ANY WARRANTY; without even the implied warranty of * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the * GNU General Public License for more details. * * You should have received a copy of the GNU General Public License * along with this program; if not, write to the Free Software * Foundation, Inc., 51 Franklin Street, Fifth Floor, Boston, MA 02110-1301 USA * */ #include "ifftw-mpi.h" INT XM(num_blocks)(INT n, INT block) { return (n + block - 1) / block; } int XM(num_blocks_ok)(INT n, INT block, MPI_Comm comm) { int n_pes; MPI_Comm_size(comm, &n_pes); return n_pes >= XM(num_blocks)(n, block); } /* Pick a default block size for dividing a problem of size n among n_pes processes. Divide as equally as possible, while minimizing the maximum block size among the processes as well as the number of processes with nonzero blocks. */ INT XM(default_block)(INT n, int n_pes) { return ((n + n_pes - 1) / n_pes); } /* For a given block size and dimension n, compute the block size on the given process. */ INT XM(block)(INT n, INT block, int which_block) { INT d = n - which_block * block; return d <= 0 ? 0 : (d > block ? block : d); } static INT num_blocks_kind(const ddim *dim, block_kind k) { return XM(num_blocks)(dim->n, dim->b[k]); } INT XM(num_blocks_total)(const dtensor *sz, block_kind k) { if (FINITE_RNK(sz->rnk)) { int i; INT ntot = 1; for (i = 0; i < sz->rnk; ++i) ntot *= num_blocks_kind(sz->dims + i, k); return ntot; } else return 0; } int XM(idle_process)(const dtensor *sz, block_kind k, int which_pe) { return (which_pe >= XM(num_blocks_total)(sz, k)); } /* Given a non-idle process which_pe, computes the coordinate vector coords[rnk] giving the coordinates of a block in the matrix of blocks. k specifies whether we are talking about the input or output data distribution. */ void XM(block_coords)(const dtensor *sz, block_kind k, int which_pe, INT *coords) { int i; A(!XM(idle_process)(sz, k, which_pe) && FINITE_RNK(sz->rnk)); for (i = sz->rnk - 1; i >= 0; --i) { INT nb = num_blocks_kind(sz->dims + i, k); coords[i] = which_pe % nb; which_pe /= nb; } } INT XM(total_block)(const dtensor *sz, block_kind k, int which_pe) { if (XM(idle_process)(sz, k, which_pe)) return 0; else { int i; INT N = 1, *coords; STACK_MALLOC(INT*, coords, sizeof(INT) * sz->rnk); XM(block_coords)(sz, k, which_pe, coords); for (i = 0; i < sz->rnk; ++i) N *= XM(block)(sz->dims[i].n, sz->dims[i].b[k], coords[i]); STACK_FREE(coords); return N; } } /* returns whether sz is local for dims >= dim */ int XM(is_local_after)(int dim, const dtensor *sz, block_kind k) { if (FINITE_RNK(sz->rnk)) for (; dim < sz->rnk; ++dim) if (XM(num_blocks)(sz->dims[dim].n, sz->dims[dim].b[k]) > 1) return 0; return 1; } int XM(is_local)(const dtensor *sz, block_kind k) { return XM(is_local_after)(0, sz, k); } /* Return whether sz is distributed for k according to a simple 1d block distribution in the first or second dimensions */ int XM(is_block1d)(const dtensor *sz, block_kind k) { int i; if (!FINITE_RNK(sz->rnk)) return 0; for (i = 0; i < sz->rnk && num_blocks_kind(sz->dims + i, k) == 1; ++i) ; return(i < sz->rnk && i < 2 && XM(is_local_after)(i + 1, sz, k)); } fftw-3.3.3/mpi/choose-radix.c0000644000175000017500000000666412054410050012730 00000000000000/* * Copyright (c) 2003, 2007-11 Matteo Frigo * Copyright (c) 2003, 2007-11 Massachusetts Institute of Technology * * This program is free software; you can redistribute it and/or modify * it under the terms of the GNU General Public License as published by * the Free Software Foundation; either version 2 of the License, or * (at your option) any later version. * * This program is distributed in the hope that it will be useful, * but WITHOUT ANY WARRANTY; without even the implied warranty of * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the * GNU General Public License for more details. * * You should have received a copy of the GNU General Public License * along with this program; if not, write to the Free Software * Foundation, Inc., 51 Franklin Street, Fifth Floor, Boston, MA 02110-1301 USA * */ #include "ifftw-mpi.h" /* Return the radix r for a 1d MPI transform of a distributed dimension d, with the given flags and transform size. That is, decomposes d.n as r * m, Cooley-Tukey style. Also computes the block sizes rblock and mblock. Returns 0 if such a decomposition is not feasible. This is unfortunately somewhat complicated. A distributed Cooley-Tukey algorithm works as follows (see dft-rank1.c): d.n is initially distributed as an m x r array with block size mblock[IB]. Then it is internally transposed to an r x m array with block size rblock[IB]. Then it is internally transposed to m x r again with block size mblock[OB]. Finally, it is transposed to r x m with block size rblock[IB]. If flags & SCRAMBLED_IN, then the first transpose is skipped (the array starts out as r x m). If flags & SCRAMBLED_OUT, then the last transpose is skipped (the array ends up as m x r). To make sure the forward and backward transforms use the same "scrambling" format, we swap r and m when sign != FFT_SIGN. There are some downsides to this, especially in the case where either m or r is not divisible by n_pes. For one thing, it means that in general we can't use the same block size for the input and output. For another thing, it means that we can't in general honor a user's "requested" block sizes in d.b[]. Therefore, for simplicity, we simply ignore d.b[] for now. */ INT XM(choose_radix)(ddim d, int n_pes, unsigned flags, int sign, INT rblock[2], INT mblock[2]) { INT r, m; UNUSED(flags); /* we would need this if we paid attention to d.b[*] */ /* If n_pes is a factor of d.n, then choose r to be d.n / n_pes. This not only ensures that the input (the m dimension) is equally distributed if possible, and at the r dimension is maximally equally distributed (if d.n/n_pes >= n_pes), it also makes one of the local transpositions in the algorithm trivial. */ if (d.n % n_pes == 0 /* it's good if n_pes divides d.n ...*/ && d.n / n_pes >= n_pes /* .. unless we can't use n_pes processes */) r = d.n / n_pes; else { /* n_pes does not divide d.n, pick a factor close to sqrt(d.n) */ for (r = X(isqrt)(d.n); d.n % r != 0; ++r) ; } if (r == 1 || r == d.n) return 0; /* punt if we can't reduce size */ if (sign != FFT_SIGN) { /* swap {m,r} so that scrambling is reversible */ m = r; r = d.n / m; } else m = d.n / r; rblock[IB] = rblock[OB] = XM(default_block)(r, n_pes); mblock[IB] = mblock[OB] = XM(default_block)(m, n_pes); return r; } fftw-3.3.3/mpi/conf.c0000644000175000017500000000330712054410050011257 00000000000000/* * Copyright (c) 2003, 2007-11 Matteo Frigo * Copyright (c) 2003, 2007-11 Massachusetts Institute of Technology * * This program is free software; you can redistribute it and/or modify * it under the terms of the GNU General Public License as published by * the Free Software Foundation; either version 2 of the License, or * (at your option) any later version. * * This program is distributed in the hope that it will be useful, * but WITHOUT ANY WARRANTY; without even the implied warranty of * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the * GNU General Public License for more details. * * You should have received a copy of the GNU General Public License * along with this program; if not, write to the Free Software * Foundation, Inc., 51 Franklin Street, Fifth Floor, Boston, MA 02110-1301 USA * */ #include "mpi-transpose.h" #include "mpi-dft.h" #include "mpi-rdft.h" #include "mpi-rdft2.h" static const solvtab s = { SOLVTAB(XM(transpose_pairwise_register)), SOLVTAB(XM(transpose_alltoall_register)), SOLVTAB(XM(transpose_recurse_register)), SOLVTAB(XM(dft_rank_geq2_register)), SOLVTAB(XM(dft_rank_geq2_transposed_register)), SOLVTAB(XM(dft_serial_register)), SOLVTAB(XM(dft_rank1_bigvec_register)), SOLVTAB(XM(dft_rank1_register)), SOLVTAB(XM(rdft_rank_geq2_register)), SOLVTAB(XM(rdft_rank_geq2_transposed_register)), SOLVTAB(XM(rdft_serial_register)), SOLVTAB(XM(rdft_rank1_bigvec_register)), SOLVTAB(XM(rdft2_rank_geq2_register)), SOLVTAB(XM(rdft2_rank_geq2_transposed_register)), SOLVTAB(XM(rdft2_serial_register)), SOLVTAB_END }; void XM(conf_standard)(planner *p) { X(solvtab_exec)(s, p); } fftw-3.3.3/mpi/dtensor.c0000644000175000017500000000743112054410050012012 00000000000000/* * Copyright (c) 2003, 2007-11 Matteo Frigo * Copyright (c) 2003, 2007-11 Massachusetts Institute of Technology * * This program is free software; you can redistribute it and/or modify * it under the terms of the GNU General Public License as published by * the Free Software Foundation; either version 2 of the License, or * (at your option) any later version. * * This program is distributed in the hope that it will be useful, * but WITHOUT ANY WARRANTY; without even the implied warranty of * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the * GNU General Public License for more details. * * You should have received a copy of the GNU General Public License * along with this program; if not, write to the Free Software * Foundation, Inc., 51 Franklin Street, Fifth Floor, Boston, MA 02110-1301 USA * */ #include "ifftw-mpi.h" dtensor *XM(mkdtensor)(int rnk) { dtensor *x; A(rnk >= 0); #if defined(STRUCT_HACK_KR) if (FINITE_RNK(rnk) && rnk > 1) x = (dtensor *)MALLOC(sizeof(dtensor) + (rnk - 1) * sizeof(ddim), TENSORS); else x = (dtensor *)MALLOC(sizeof(dtensor), TENSORS); #elif defined(STRUCT_HACK_C99) if (FINITE_RNK(rnk)) x = (dtensor *)MALLOC(sizeof(dtensor) + rnk * sizeof(ddim), TENSORS); else x = (dtensor *)MALLOC(sizeof(dtensor), TENSORS); #else x = (dtensor *)MALLOC(sizeof(dtensor), TENSORS); if (FINITE_RNK(rnk) && rnk > 0) x->dims = (ddim *)MALLOC(sizeof(ddim) * rnk, TENSORS); else x->dims = 0; #endif x->rnk = rnk; return x; } void XM(dtensor_destroy)(dtensor *sz) { #if !defined(STRUCT_HACK_C99) && !defined(STRUCT_HACK_KR) X(ifree0)(sz->dims); #endif X(ifree)(sz); } void XM(dtensor_md5)(md5 *p, const dtensor *t) { int i; X(md5int)(p, t->rnk); if (FINITE_RNK(t->rnk)) { for (i = 0; i < t->rnk; ++i) { const ddim *q = t->dims + i; X(md5INT)(p, q->n); X(md5INT)(p, q->b[IB]); X(md5INT)(p, q->b[OB]); } } } dtensor *XM(dtensor_copy)(const dtensor *sz) { dtensor *x = XM(mkdtensor)(sz->rnk); int i; if (FINITE_RNK(sz->rnk)) for (i = 0; i < sz->rnk; ++i) x->dims[i] = sz->dims[i]; return x; } dtensor *XM(dtensor_canonical)(const dtensor *sz, int compress) { int i, rnk; dtensor *x; block_kind k; if (!FINITE_RNK(sz->rnk)) return XM(mkdtensor)(sz->rnk); for (i = rnk = 0; i < sz->rnk; ++i) { if (sz->dims[i].n <= 0) return XM(mkdtensor)(RNK_MINFTY); else if (!compress || sz->dims[i].n > 1) ++rnk; } x = XM(mkdtensor)(rnk); for (i = rnk = 0; i < sz->rnk; ++i) { if (!compress || sz->dims[i].n > 1) { x->dims[rnk].n = sz->dims[i].n; FORALL_BLOCK_KIND(k) { if (XM(num_blocks)(sz->dims[i].n, sz->dims[i].b[k]) == 1) x->dims[rnk].b[k] = sz->dims[i].n; else x->dims[rnk].b[k] = sz->dims[i].b[k]; } ++rnk; } } return x; } int XM(dtensor_validp)(const dtensor *sz) { int i; if (sz->rnk < 0) return 0; if (FINITE_RNK(sz->rnk)) for (i = 0; i < sz->rnk; ++i) if (sz->dims[i].n < 0 || sz->dims[i].b[IB] <= 0 || sz->dims[i].b[OB] <= 0) return 0; return 1; } void XM(dtensor_print)(const dtensor *t, printer *p) { if (FINITE_RNK(t->rnk)) { int i; int first = 1; p->print(p, "("); for (i = 0; i < t->rnk; ++i) { const ddim *d = t->dims + i; p->print(p, "%s(%D %D %D)", first ? "" : " ", d->n, d->b[IB], d->b[OB]); first = 0; } p->print(p, ")"); } else { p->print(p, "rank-minfty"); } } fftw-3.3.3/mpi/ifftw-mpi.h0000644000175000017500000001245712054410050012247 00000000000000/* * Copyright (c) 2003, 2007-11 Matteo Frigo * Copyright (c) 2003, 2007-11 Massachusetts Institute of Technology * * This program is free software; you can redistribute it and/or modify * it under the terms of the GNU General Public License as published by * the Free Software Foundation; either version 2 of the License, or * (at your option) any later version. * * This program is distributed in the hope that it will be useful, * but WITHOUT ANY WARRANTY; without even the implied warranty of * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the * GNU General Public License for more details. * * You should have received a copy of the GNU General Public License * along with this program; if not, write to the Free Software * Foundation, Inc., 51 Franklin Street, Fifth Floor, Boston, MA 02110-1301 USA * */ /* FFTW-MPI internal header file */ #ifndef __IFFTW_MPI_H__ #define __IFFTW_MPI_H__ #include "ifftw.h" #include "rdft.h" #include /* mpi problem flags: problem-dependent meaning, but in general SCRAMBLED means some reordering *within* the dimensions, while TRANSPOSED means some reordering *of* the dimensions */ #define SCRAMBLED_IN (1 << 0) #define SCRAMBLED_OUT (1 << 1) #define TRANSPOSED_IN (1 << 2) #define TRANSPOSED_OUT (1 << 3) #define RANK1_BIGVEC_ONLY (1 << 4) /* for rank=1, allow only bigvec solver */ #define ONLY_SCRAMBLEDP(flags) (!((flags) & ~(SCRAMBLED_IN|SCRAMBLED_OUT))) #define ONLY_TRANSPOSEDP(flags) (!((flags) & ~(TRANSPOSED_IN|TRANSPOSED_OUT))) #if defined(FFTW_SINGLE) # define FFTW_MPI_TYPE MPI_FLOAT #elif defined(FFTW_LDOUBLE) # define FFTW_MPI_TYPE MPI_LONG_DOUBLE #elif defined(FFTW_QUAD) # error MPI quad-precision type is unknown #else # define FFTW_MPI_TYPE MPI_DOUBLE #endif /* all fftw-mpi identifiers start with fftw_mpi (or fftwf_mpi etc.) */ #define XM(name) X(CONCAT(mpi_, name)) /***********************************************************************/ /* block distributions */ /* a distributed dimension of length n with input and output block sizes ib and ob, respectively. */ typedef enum { IB = 0, OB } block_kind; typedef struct { INT n; INT b[2]; /* b[IB], b[OB] */ } ddim; /* Loop over k in {IB, OB}. Note: need explicit casts for C++. */ #define FORALL_BLOCK_KIND(k) for (k = IB; k <= OB; k = (block_kind) (((int) k) + 1)) /* unlike tensors in the serial FFTW, the ordering of the dtensor dimensions matters - both the array and the block layout are row-major order. */ typedef struct { int rnk; #if defined(STRUCT_HACK_KR) ddim dims[1]; #elif defined(STRUCT_HACK_C99) ddim dims[]; #else ddim *dims; #endif } dtensor; /* dtensor.c: */ dtensor *XM(mkdtensor)(int rnk); void XM(dtensor_destroy)(dtensor *sz); dtensor *XM(dtensor_copy)(const dtensor *sz); dtensor *XM(dtensor_canonical)(const dtensor *sz, int compress); int XM(dtensor_validp)(const dtensor *sz); void XM(dtensor_md5)(md5 *p, const dtensor *t); void XM(dtensor_print)(const dtensor *t, printer *p); /* block.c: */ /* for a single distributed dimension: */ INT XM(num_blocks)(INT n, INT block); int XM(num_blocks_ok)(INT n, INT block, MPI_Comm comm); INT XM(default_block)(INT n, int n_pes); INT XM(block)(INT n, INT block, int which_block); /* for multiple distributed dimensions: */ INT XM(num_blocks_total)(const dtensor *sz, block_kind k); int XM(idle_process)(const dtensor *sz, block_kind k, int which_pe); void XM(block_coords)(const dtensor *sz, block_kind k, int which_pe, INT *coords); INT XM(total_block)(const dtensor *sz, block_kind k, int which_pe); int XM(is_local_after)(int dim, const dtensor *sz, block_kind k); int XM(is_local)(const dtensor *sz, block_kind k); int XM(is_block1d)(const dtensor *sz, block_kind k); /* choose-radix.c */ INT XM(choose_radix)(ddim d, int n_pes, unsigned flags, int sign, INT rblock[2], INT mblock[2]); /***********************************************************************/ /* any_true.c */ int XM(any_true)(int condition, MPI_Comm comm); int XM(md5_equal)(md5 m, MPI_Comm comm); /* conf.c */ void XM(conf_standard)(planner *p); /***********************************************************************/ /* rearrange.c */ /* Different ways to rearrange the vector dimension vn during transposition, reflecting different tradeoffs between ease of transposition and contiguity during the subsequent DFTs. TODO: can we pare this down to CONTIG and DISCONTIG, at least in MEASURE mode? SQUARE_MIDDLE is also used for 1d destroy-input DFTs. */ typedef enum { CONTIG = 0, /* vn x 1: make subsequent DFTs contiguous */ DISCONTIG, /* P x (vn/P) for P processes */ SQUARE_BEFORE, /* try to get square transpose at beginning */ SQUARE_MIDDLE, /* try to get square transpose in the middle */ SQUARE_AFTER /* try to get square transpose at end */ } rearrangement; /* skipping SQUARE_AFTER since it doesn't seem to offer any advantage over SQUARE_BEFORE */ #define FORALL_REARRANGE(rearrange) for (rearrange = CONTIG; rearrange <= SQUARE_MIDDLE; rearrange = (rearrangement) (((int) rearrange) + 1)) int XM(rearrange_applicable)(rearrangement rearrange, ddim dim0, INT vn, int n_pes); INT XM(rearrange_ny)(rearrangement rearrange, ddim dim0, INT vn, int n_pes); /***********************************************************************/ #endif /* __IFFTW_MPI_H__ */ fftw-3.3.3/mpi/rearrange.c0000644000175000017500000000410312054410050012273 00000000000000/* * Copyright (c) 2003, 2007-11 Matteo Frigo * Copyright (c) 2003, 2007-11 Massachusetts Institute of Technology * * This program is free software; you can redistribute it and/or modify * it under the terms of the GNU General Public License as published by * the Free Software Foundation; either version 2 of the License, or * (at your option) any later version. * * This program is distributed in the hope that it will be useful, * but WITHOUT ANY WARRANTY; without even the implied warranty of * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the * GNU General Public License for more details. * * You should have received a copy of the GNU General Public License * along with this program; if not, write to the Free Software * Foundation, Inc., 51 Franklin Street, Fifth Floor, Boston, MA 02110-1301 USA * */ #include "ifftw-mpi.h" /* common functions for rearrangements of the data for the *-rank1-bigvec solvers */ static int div_mult(INT b, INT a) { return (a > b && a % b == 0); } static int div_mult2(INT b, INT a, INT n) { return (div_mult(b, a) && div_mult(n, b)); } int XM(rearrange_applicable)(rearrangement rearrange, ddim dim0, INT vn, int n_pes) { /* note: it is important that cases other than CONTIG be applicable only when the resulting transpose dimension is divisible by n_pes; otherwise, the allocation size returned by the API will be incorrect */ return ((rearrange != DISCONTIG || div_mult(n_pes, vn)) && (rearrange != SQUARE_BEFORE || div_mult2(dim0.b[IB], vn, n_pes)) && (rearrange != SQUARE_AFTER || (dim0.b[IB] != dim0.b[OB] && div_mult2(dim0.b[OB], vn, n_pes))) && (rearrange != SQUARE_MIDDLE || div_mult(dim0.n * n_pes, vn))); } INT XM(rearrange_ny)(rearrangement rearrange, ddim dim0, INT vn, int n_pes) { switch (rearrange) { case CONTIG: return vn; case DISCONTIG: return n_pes; case SQUARE_BEFORE: return dim0.b[IB]; case SQUARE_AFTER: return dim0.b[OB]; case SQUARE_MIDDLE: return dim0.n * n_pes; } return 0; } fftw-3.3.3/mpi/wisdom-api.c0000644000175000017500000000745512054410050012413 00000000000000/* * Copyright (c) 2003, 2007-11 Matteo Frigo * Copyright (c) 2003, 2007-11 Massachusetts Institute of Technology * * This program is free software; you can redistribute it and/or modify * it under the terms of the GNU General Public License as published by * the Free Software Foundation; either version 2 of the License, or * (at your option) any later version. * * This program is distributed in the hope that it will be useful, * but WITHOUT ANY WARRANTY; without even the implied warranty of * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the * GNU General Public License for more details. * * You should have received a copy of the GNU General Public License * along with this program; if not, write to the Free Software * Foundation, Inc., 51 Franklin Street, Fifth Floor, Boston, MA 02110-1301 USA * */ #include "fftw3-mpi.h" #include "ifftw-mpi.h" #include #if SIZEOF_SIZE_T == SIZEOF_UNSIGNED_INT # define FFTW_MPI_SIZE_T MPI_UNSIGNED #elif SIZEOF_SIZE_T == SIZEOF_UNSIGNED_LONG # define FFTW_MPI_SIZE_T MPI_UNSIGNED_LONG #elif SIZEOF_SIZE_T == SIZEOF_UNSIGNED_LONG_LONG # define FFTW_MPI_SIZE_T MPI_UNSIGNED_LONG_LONG #else # error MPI type for size_t is unknown # define FFTW_MPI_SIZE_T MPI_UNSIGNED_LONG #endif /* Import wisdom from all processes to process 0, as prelude to exporting a single wisdom file (this is convenient when we are running on identical processors, to avoid the annoyance of having per-process wisdom files). In order to make the time for this operation logarithmic in the number of processors (rather than linear), we employ a tree reduction algorithm. This means that the wisdom is modified on processes other than root, which shouldn't matter in practice. */ void XM(gather_wisdom)(MPI_Comm comm_) { MPI_Comm comm, comm2; int my_pe, n_pes; char *wis; size_t wislen; MPI_Status status; MPI_Comm_dup(comm_, &comm); MPI_Comm_rank(comm, &my_pe); MPI_Comm_size(comm, &n_pes); if (n_pes > 2) { /* recursively split into even/odd processes */ MPI_Comm_split(comm, my_pe % 2, my_pe, &comm2); XM(gather_wisdom)(comm2); MPI_Comm_free(&comm2); } if (n_pes > 1 && my_pe < 2) { /* import process 1 -> 0 */ if (my_pe == 1) { wis = X(export_wisdom_to_string)(); wislen = strlen(wis) + 1; MPI_Send(&wislen, 1, FFTW_MPI_SIZE_T, 0, 111, comm); MPI_Send(wis, wislen, MPI_CHAR, 0, 222, comm); free(wis); } else /* my_pe == 0 */ { MPI_Recv(&wislen, 1, FFTW_MPI_SIZE_T, 1, 111, comm, &status); wis = (char *) MALLOC(wislen * sizeof(char), OTHER); MPI_Recv(wis, wislen, MPI_CHAR, 1, 222, comm, &status); if (!X(import_wisdom_from_string)(wis)) MPI_Abort(comm, 1); X(ifree)(wis); } } MPI_Comm_free(&comm); } /* broadcast wisdom from process 0 to all other processes; this is useful so that we can import wisdom once and not worry about parallel I/O or process-specific wisdom, although of course it assumes that all the processes have identical performance characteristics (i.e. identical hardware). */ void XM(broadcast_wisdom)(MPI_Comm comm_) { MPI_Comm comm; int my_pe; char *wis; size_t wislen; MPI_Comm_dup(comm_, &comm); MPI_Comm_rank(comm, &my_pe); if (my_pe != 0) { MPI_Bcast(&wislen, 1, FFTW_MPI_SIZE_T, 0, comm); wis = (char *) MALLOC(wislen * sizeof(char), OTHER); MPI_Bcast(wis, wislen, MPI_CHAR, 0, comm); if (!X(import_wisdom_from_string)(wis)) MPI_Abort(comm, 1); X(ifree)(wis); } else /* my_pe == 0 */ { wis = X(export_wisdom_to_string)(); wislen = strlen(wis) + 1; MPI_Bcast(&wislen, 1, FFTW_MPI_SIZE_T, 0, comm); MPI_Bcast(wis, wislen, MPI_CHAR, 0, comm); X(free)(wis); } MPI_Comm_free(&comm); } fftw-3.3.3/mpi/f03-wrap.c0000644000175000017500000003275512054411104011703 00000000000000/* Generated automatically. DO NOT EDIT! */ #include "fftw3-mpi.h" #include "ifftw-mpi.h" FFTW_EXTERN ptrdiff_t XM(local_size_many_transposed_f03)(int rnk, const ptrdiff_t * n, ptrdiff_t howmany, ptrdiff_t block0, ptrdiff_t block1, MPI_Fint f_comm, ptrdiff_t * local_n0, ptrdiff_t * local_0_start, ptrdiff_t * local_n1, ptrdiff_t * local_1_start); FFTW_EXTERN ptrdiff_t XM(local_size_many_f03)(int rnk, const ptrdiff_t * n, ptrdiff_t howmany, ptrdiff_t block0, MPI_Fint f_comm, ptrdiff_t * local_n0, ptrdiff_t * local_0_start); FFTW_EXTERN ptrdiff_t XM(local_size_transposed_f03)(int rnk, const ptrdiff_t * n, MPI_Fint f_comm, ptrdiff_t * local_n0, ptrdiff_t * local_0_start, ptrdiff_t * local_n1, ptrdiff_t * local_1_start); FFTW_EXTERN ptrdiff_t XM(local_size_f03)(int rnk, const ptrdiff_t * n, MPI_Fint f_comm, ptrdiff_t * local_n0, ptrdiff_t * local_0_start); FFTW_EXTERN ptrdiff_t XM(local_size_many_1d_f03)(ptrdiff_t n0, ptrdiff_t howmany, MPI_Fint f_comm, int sign, unsigned flags, ptrdiff_t * local_ni, ptrdiff_t * local_i_start, ptrdiff_t * local_no, ptrdiff_t * local_o_start); FFTW_EXTERN ptrdiff_t XM(local_size_1d_f03)(ptrdiff_t n0, MPI_Fint f_comm, int sign, unsigned flags, ptrdiff_t * local_ni, ptrdiff_t * local_i_start, ptrdiff_t * local_no, ptrdiff_t * local_o_start); FFTW_EXTERN ptrdiff_t XM(local_size_2d_f03)(ptrdiff_t n0, ptrdiff_t n1, MPI_Fint f_comm, ptrdiff_t * local_n0, ptrdiff_t * local_0_start); FFTW_EXTERN ptrdiff_t XM(local_size_2d_transposed_f03)(ptrdiff_t n0, ptrdiff_t n1, MPI_Fint f_comm, ptrdiff_t * local_n0, ptrdiff_t * local_0_start, ptrdiff_t * local_n1, ptrdiff_t * local_1_start); FFTW_EXTERN ptrdiff_t XM(local_size_3d_f03)(ptrdiff_t n0, ptrdiff_t n1, ptrdiff_t n2, MPI_Fint f_comm, ptrdiff_t * local_n0, ptrdiff_t * local_0_start); FFTW_EXTERN ptrdiff_t XM(local_size_3d_transposed_f03)(ptrdiff_t n0, ptrdiff_t n1, ptrdiff_t n2, MPI_Fint f_comm, ptrdiff_t * local_n0, ptrdiff_t * local_0_start, ptrdiff_t * local_n1, ptrdiff_t * local_1_start); FFTW_EXTERN X(plan) XM(plan_many_transpose_f03)(ptrdiff_t n0, ptrdiff_t n1, ptrdiff_t howmany, ptrdiff_t block0, ptrdiff_t block1, R * in, R * out, MPI_Fint f_comm, unsigned flags); FFTW_EXTERN X(plan) XM(plan_transpose_f03)(ptrdiff_t n0, ptrdiff_t n1, R * in, R * out, MPI_Fint f_comm, unsigned flags); FFTW_EXTERN X(plan) XM(plan_many_dft_f03)(int rnk, const ptrdiff_t * n, ptrdiff_t howmany, ptrdiff_t block, ptrdiff_t tblock, X(complex) * in, X(complex) * out, MPI_Fint f_comm, int sign, unsigned flags); FFTW_EXTERN X(plan) XM(plan_dft_f03)(int rnk, const ptrdiff_t * n, X(complex) * in, X(complex) * out, MPI_Fint f_comm, int sign, unsigned flags); FFTW_EXTERN X(plan) XM(plan_dft_1d_f03)(ptrdiff_t n0, X(complex) * in, X(complex) * out, MPI_Fint f_comm, int sign, unsigned flags); FFTW_EXTERN X(plan) XM(plan_dft_2d_f03)(ptrdiff_t n0, ptrdiff_t n1, X(complex) * in, X(complex) * out, MPI_Fint f_comm, int sign, unsigned flags); FFTW_EXTERN X(plan) XM(plan_dft_3d_f03)(ptrdiff_t n0, ptrdiff_t n1, ptrdiff_t n2, X(complex) * in, X(complex) * out, MPI_Fint f_comm, int sign, unsigned flags); FFTW_EXTERN X(plan) XM(plan_many_r2r_f03)(int rnk, const ptrdiff_t * n, ptrdiff_t howmany, ptrdiff_t iblock, ptrdiff_t oblock, R * in, R * out, MPI_Fint f_comm, const X(r2r_kind) * kind, unsigned flags); FFTW_EXTERN X(plan) XM(plan_r2r_f03)(int rnk, const ptrdiff_t * n, R * in, R * out, MPI_Fint f_comm, const X(r2r_kind) * kind, unsigned flags); FFTW_EXTERN X(plan) XM(plan_r2r_2d_f03)(ptrdiff_t n0, ptrdiff_t n1, R * in, R * out, MPI_Fint f_comm, X(r2r_kind) kind0, X(r2r_kind) kind1, unsigned flags); FFTW_EXTERN X(plan) XM(plan_r2r_3d_f03)(ptrdiff_t n0, ptrdiff_t n1, ptrdiff_t n2, R * in, R * out, MPI_Fint f_comm, X(r2r_kind) kind0, X(r2r_kind) kind1, X(r2r_kind) kind2, unsigned flags); FFTW_EXTERN X(plan) XM(plan_many_dft_r2c_f03)(int rnk, const ptrdiff_t * n, ptrdiff_t howmany, ptrdiff_t iblock, ptrdiff_t oblock, R * in, X(complex) * out, MPI_Fint f_comm, unsigned flags); FFTW_EXTERN X(plan) XM(plan_dft_r2c_f03)(int rnk, const ptrdiff_t * n, R * in, X(complex) * out, MPI_Fint f_comm, unsigned flags); FFTW_EXTERN X(plan) XM(plan_dft_r2c_2d_f03)(ptrdiff_t n0, ptrdiff_t n1, R * in, X(complex) * out, MPI_Fint f_comm, unsigned flags); FFTW_EXTERN X(plan) XM(plan_dft_r2c_3d_f03)(ptrdiff_t n0, ptrdiff_t n1, ptrdiff_t n2, R * in, X(complex) * out, MPI_Fint f_comm, unsigned flags); FFTW_EXTERN X(plan) XM(plan_many_dft_c2r_f03)(int rnk, const ptrdiff_t * n, ptrdiff_t howmany, ptrdiff_t iblock, ptrdiff_t oblock, X(complex) * in, R * out, MPI_Fint f_comm, unsigned flags); FFTW_EXTERN X(plan) XM(plan_dft_c2r_f03)(int rnk, const ptrdiff_t * n, X(complex) * in, R * out, MPI_Fint f_comm, unsigned flags); FFTW_EXTERN X(plan) XM(plan_dft_c2r_2d_f03)(ptrdiff_t n0, ptrdiff_t n1, X(complex) * in, R * out, MPI_Fint f_comm, unsigned flags); FFTW_EXTERN X(plan) XM(plan_dft_c2r_3d_f03)(ptrdiff_t n0, ptrdiff_t n1, ptrdiff_t n2, X(complex) * in, R * out, MPI_Fint f_comm, unsigned flags); FFTW_EXTERN void XM(gather_wisdom_f03)(MPI_Fint f_comm_); FFTW_EXTERN void XM(broadcast_wisdom_f03)(MPI_Fint f_comm_); ptrdiff_t XM(local_size_many_transposed_f03)(int rnk, const ptrdiff_t * n, ptrdiff_t howmany, ptrdiff_t block0, ptrdiff_t block1, MPI_Fint f_comm, ptrdiff_t * local_n0, ptrdiff_t * local_0_start, ptrdiff_t * local_n1, ptrdiff_t * local_1_start) { MPI_Comm comm; comm = MPI_Comm_f2c(f_comm); return XM(local_size_many_transposed)(rnk,n,howmany,block0,block1,comm,local_n0,local_0_start,local_n1,local_1_start); } ptrdiff_t XM(local_size_many_f03)(int rnk, const ptrdiff_t * n, ptrdiff_t howmany, ptrdiff_t block0, MPI_Fint f_comm, ptrdiff_t * local_n0, ptrdiff_t * local_0_start) { MPI_Comm comm; comm = MPI_Comm_f2c(f_comm); return XM(local_size_many)(rnk,n,howmany,block0,comm,local_n0,local_0_start); } ptrdiff_t XM(local_size_transposed_f03)(int rnk, const ptrdiff_t * n, MPI_Fint f_comm, ptrdiff_t * local_n0, ptrdiff_t * local_0_start, ptrdiff_t * local_n1, ptrdiff_t * local_1_start) { MPI_Comm comm; comm = MPI_Comm_f2c(f_comm); return XM(local_size_transposed)(rnk,n,comm,local_n0,local_0_start,local_n1,local_1_start); } ptrdiff_t XM(local_size_f03)(int rnk, const ptrdiff_t * n, MPI_Fint f_comm, ptrdiff_t * local_n0, ptrdiff_t * local_0_start) { MPI_Comm comm; comm = MPI_Comm_f2c(f_comm); return XM(local_size)(rnk,n,comm,local_n0,local_0_start); } ptrdiff_t XM(local_size_many_1d_f03)(ptrdiff_t n0, ptrdiff_t howmany, MPI_Fint f_comm, int sign, unsigned flags, ptrdiff_t * local_ni, ptrdiff_t * local_i_start, ptrdiff_t * local_no, ptrdiff_t * local_o_start) { MPI_Comm comm; comm = MPI_Comm_f2c(f_comm); return XM(local_size_many_1d)(n0,howmany,comm,sign,flags,local_ni,local_i_start,local_no,local_o_start); } ptrdiff_t XM(local_size_1d_f03)(ptrdiff_t n0, MPI_Fint f_comm, int sign, unsigned flags, ptrdiff_t * local_ni, ptrdiff_t * local_i_start, ptrdiff_t * local_no, ptrdiff_t * local_o_start) { MPI_Comm comm; comm = MPI_Comm_f2c(f_comm); return XM(local_size_1d)(n0,comm,sign,flags,local_ni,local_i_start,local_no,local_o_start); } ptrdiff_t XM(local_size_2d_f03)(ptrdiff_t n0, ptrdiff_t n1, MPI_Fint f_comm, ptrdiff_t * local_n0, ptrdiff_t * local_0_start) { MPI_Comm comm; comm = MPI_Comm_f2c(f_comm); return XM(local_size_2d)(n0,n1,comm,local_n0,local_0_start); } ptrdiff_t XM(local_size_2d_transposed_f03)(ptrdiff_t n0, ptrdiff_t n1, MPI_Fint f_comm, ptrdiff_t * local_n0, ptrdiff_t * local_0_start, ptrdiff_t * local_n1, ptrdiff_t * local_1_start) { MPI_Comm comm; comm = MPI_Comm_f2c(f_comm); return XM(local_size_2d_transposed)(n0,n1,comm,local_n0,local_0_start,local_n1,local_1_start); } ptrdiff_t XM(local_size_3d_f03)(ptrdiff_t n0, ptrdiff_t n1, ptrdiff_t n2, MPI_Fint f_comm, ptrdiff_t * local_n0, ptrdiff_t * local_0_start) { MPI_Comm comm; comm = MPI_Comm_f2c(f_comm); return XM(local_size_3d)(n0,n1,n2,comm,local_n0,local_0_start); } ptrdiff_t XM(local_size_3d_transposed_f03)(ptrdiff_t n0, ptrdiff_t n1, ptrdiff_t n2, MPI_Fint f_comm, ptrdiff_t * local_n0, ptrdiff_t * local_0_start, ptrdiff_t * local_n1, ptrdiff_t * local_1_start) { MPI_Comm comm; comm = MPI_Comm_f2c(f_comm); return XM(local_size_3d_transposed)(n0,n1,n2,comm,local_n0,local_0_start,local_n1,local_1_start); } X(plan) XM(plan_many_transpose_f03)(ptrdiff_t n0, ptrdiff_t n1, ptrdiff_t howmany, ptrdiff_t block0, ptrdiff_t block1, R * in, R * out, MPI_Fint f_comm, unsigned flags) { MPI_Comm comm; comm = MPI_Comm_f2c(f_comm); return XM(plan_many_transpose)(n0,n1,howmany,block0,block1,in,out,comm,flags); } X(plan) XM(plan_transpose_f03)(ptrdiff_t n0, ptrdiff_t n1, R * in, R * out, MPI_Fint f_comm, unsigned flags) { MPI_Comm comm; comm = MPI_Comm_f2c(f_comm); return XM(plan_transpose)(n0,n1,in,out,comm,flags); } X(plan) XM(plan_many_dft_f03)(int rnk, const ptrdiff_t * n, ptrdiff_t howmany, ptrdiff_t block, ptrdiff_t tblock, X(complex) * in, X(complex) * out, MPI_Fint f_comm, int sign, unsigned flags) { MPI_Comm comm; comm = MPI_Comm_f2c(f_comm); return XM(plan_many_dft)(rnk,n,howmany,block,tblock,in,out,comm,sign,flags); } X(plan) XM(plan_dft_f03)(int rnk, const ptrdiff_t * n, X(complex) * in, X(complex) * out, MPI_Fint f_comm, int sign, unsigned flags) { MPI_Comm comm; comm = MPI_Comm_f2c(f_comm); return XM(plan_dft)(rnk,n,in,out,comm,sign,flags); } X(plan) XM(plan_dft_1d_f03)(ptrdiff_t n0, X(complex) * in, X(complex) * out, MPI_Fint f_comm, int sign, unsigned flags) { MPI_Comm comm; comm = MPI_Comm_f2c(f_comm); return XM(plan_dft_1d)(n0,in,out,comm,sign,flags); } X(plan) XM(plan_dft_2d_f03)(ptrdiff_t n0, ptrdiff_t n1, X(complex) * in, X(complex) * out, MPI_Fint f_comm, int sign, unsigned flags) { MPI_Comm comm; comm = MPI_Comm_f2c(f_comm); return XM(plan_dft_2d)(n0,n1,in,out,comm,sign,flags); } X(plan) XM(plan_dft_3d_f03)(ptrdiff_t n0, ptrdiff_t n1, ptrdiff_t n2, X(complex) * in, X(complex) * out, MPI_Fint f_comm, int sign, unsigned flags) { MPI_Comm comm; comm = MPI_Comm_f2c(f_comm); return XM(plan_dft_3d)(n0,n1,n2,in,out,comm,sign,flags); } X(plan) XM(plan_many_r2r_f03)(int rnk, const ptrdiff_t * n, ptrdiff_t howmany, ptrdiff_t iblock, ptrdiff_t oblock, R * in, R * out, MPI_Fint f_comm, const X(r2r_kind) * kind, unsigned flags) { MPI_Comm comm; comm = MPI_Comm_f2c(f_comm); return XM(plan_many_r2r)(rnk,n,howmany,iblock,oblock,in,out,comm,kind,flags); } X(plan) XM(plan_r2r_f03)(int rnk, const ptrdiff_t * n, R * in, R * out, MPI_Fint f_comm, const X(r2r_kind) * kind, unsigned flags) { MPI_Comm comm; comm = MPI_Comm_f2c(f_comm); return XM(plan_r2r)(rnk,n,in,out,comm,kind,flags); } X(plan) XM(plan_r2r_2d_f03)(ptrdiff_t n0, ptrdiff_t n1, R * in, R * out, MPI_Fint f_comm, X(r2r_kind) kind0, X(r2r_kind) kind1, unsigned flags) { MPI_Comm comm; comm = MPI_Comm_f2c(f_comm); return XM(plan_r2r_2d)(n0,n1,in,out,comm,kind0,kind1,flags); } X(plan) XM(plan_r2r_3d_f03)(ptrdiff_t n0, ptrdiff_t n1, ptrdiff_t n2, R * in, R * out, MPI_Fint f_comm, X(r2r_kind) kind0, X(r2r_kind) kind1, X(r2r_kind) kind2, unsigned flags) { MPI_Comm comm; comm = MPI_Comm_f2c(f_comm); return XM(plan_r2r_3d)(n0,n1,n2,in,out,comm,kind0,kind1,kind2,flags); } X(plan) XM(plan_many_dft_r2c_f03)(int rnk, const ptrdiff_t * n, ptrdiff_t howmany, ptrdiff_t iblock, ptrdiff_t oblock, R * in, X(complex) * out, MPI_Fint f_comm, unsigned flags) { MPI_Comm comm; comm = MPI_Comm_f2c(f_comm); return XM(plan_many_dft_r2c)(rnk,n,howmany,iblock,oblock,in,out,comm,flags); } X(plan) XM(plan_dft_r2c_f03)(int rnk, const ptrdiff_t * n, R * in, X(complex) * out, MPI_Fint f_comm, unsigned flags) { MPI_Comm comm; comm = MPI_Comm_f2c(f_comm); return XM(plan_dft_r2c)(rnk,n,in,out,comm,flags); } X(plan) XM(plan_dft_r2c_2d_f03)(ptrdiff_t n0, ptrdiff_t n1, R * in, X(complex) * out, MPI_Fint f_comm, unsigned flags) { MPI_Comm comm; comm = MPI_Comm_f2c(f_comm); return XM(plan_dft_r2c_2d)(n0,n1,in,out,comm,flags); } X(plan) XM(plan_dft_r2c_3d_f03)(ptrdiff_t n0, ptrdiff_t n1, ptrdiff_t n2, R * in, X(complex) * out, MPI_Fint f_comm, unsigned flags) { MPI_Comm comm; comm = MPI_Comm_f2c(f_comm); return XM(plan_dft_r2c_3d)(n0,n1,n2,in,out,comm,flags); } X(plan) XM(plan_many_dft_c2r_f03)(int rnk, const ptrdiff_t * n, ptrdiff_t howmany, ptrdiff_t iblock, ptrdiff_t oblock, X(complex) * in, R * out, MPI_Fint f_comm, unsigned flags) { MPI_Comm comm; comm = MPI_Comm_f2c(f_comm); return XM(plan_many_dft_c2r)(rnk,n,howmany,iblock,oblock,in,out,comm,flags); } X(plan) XM(plan_dft_c2r_f03)(int rnk, const ptrdiff_t * n, X(complex) * in, R * out, MPI_Fint f_comm, unsigned flags) { MPI_Comm comm; comm = MPI_Comm_f2c(f_comm); return XM(plan_dft_c2r)(rnk,n,in,out,comm,flags); } X(plan) XM(plan_dft_c2r_2d_f03)(ptrdiff_t n0, ptrdiff_t n1, X(complex) * in, R * out, MPI_Fint f_comm, unsigned flags) { MPI_Comm comm; comm = MPI_Comm_f2c(f_comm); return XM(plan_dft_c2r_2d)(n0,n1,in,out,comm,flags); } X(plan) XM(plan_dft_c2r_3d_f03)(ptrdiff_t n0, ptrdiff_t n1, ptrdiff_t n2, X(complex) * in, R * out, MPI_Fint f_comm, unsigned flags) { MPI_Comm comm; comm = MPI_Comm_f2c(f_comm); return XM(plan_dft_c2r_3d)(n0,n1,n2,in,out,comm,flags); } void XM(gather_wisdom_f03)(MPI_Fint f_comm_) { MPI_Comm comm_; comm_ = MPI_Comm_f2c(f_comm_); XM(gather_wisdom)(comm_); } void XM(broadcast_wisdom_f03)(MPI_Fint f_comm_) { MPI_Comm comm_; comm_ = MPI_Comm_f2c(f_comm_); XM(broadcast_wisdom)(comm_); } fftw-3.3.3/mpi/transpose-alltoall.c0000644000175000017500000002017512054410050014154 00000000000000/* * Copyright (c) 2003, 2007-11 Matteo Frigo * Copyright (c) 2003, 2007-11 Massachusetts Institute of Technology * * This program is free software; you can redistribute it and/or modify * it under the terms of the GNU General Public License as published by * the Free Software Foundation; either version 2 of the License, or * (at your option) any later version. * * This program is distributed in the hope that it will be useful, * but WITHOUT ANY WARRANTY; without even the implied warranty of * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the * GNU General Public License for more details. * * You should have received a copy of the GNU General Public License * along with this program; if not, write to the Free Software * Foundation, Inc., 51 Franklin Street, Fifth Floor, Boston, MA 02110-1301 USA * */ /* plans for distributed out-of-place transpose using MPI_Alltoall, and which destroy the input array (unless TRANSPOSED_IN is used) */ #include "mpi-transpose.h" #include typedef struct { solver super; int copy_transposed_in; /* whether to copy the input for TRANSPOSED_IN, which makes the final transpose out-of-place but costs an extra copy and requires us to destroy the input */ } S; typedef struct { plan_mpi_transpose super; plan *cld1, *cld2, *cld2rest, *cld3; MPI_Comm comm; int *send_block_sizes, *send_block_offsets; int *recv_block_sizes, *recv_block_offsets; INT rest_Ioff, rest_Ooff; int equal_blocks; } P; static void apply(const plan *ego_, R *I, R *O) { const P *ego = (const P *) ego_; plan_rdft *cld1, *cld2, *cld2rest, *cld3; /* transpose locally to get contiguous chunks */ cld1 = (plan_rdft *) ego->cld1; if (cld1) { cld1->apply(ego->cld1, I, O); /* transpose chunks globally */ if (ego->equal_blocks) MPI_Alltoall(O, ego->send_block_sizes[0], FFTW_MPI_TYPE, I, ego->recv_block_sizes[0], FFTW_MPI_TYPE, ego->comm); else MPI_Alltoallv(O, ego->send_block_sizes, ego->send_block_offsets, FFTW_MPI_TYPE, I, ego->recv_block_sizes, ego->recv_block_offsets, FFTW_MPI_TYPE, ego->comm); } else { /* TRANSPOSED_IN, no need to destroy input */ /* transpose chunks globally */ if (ego->equal_blocks) MPI_Alltoall(I, ego->send_block_sizes[0], FFTW_MPI_TYPE, O, ego->recv_block_sizes[0], FFTW_MPI_TYPE, ego->comm); else MPI_Alltoallv(I, ego->send_block_sizes, ego->send_block_offsets, FFTW_MPI_TYPE, O, ego->recv_block_sizes, ego->recv_block_offsets, FFTW_MPI_TYPE, ego->comm); I = O; /* final transpose (if any) is in-place */ } /* transpose locally, again, to get ordinary row-major */ cld2 = (plan_rdft *) ego->cld2; if (cld2) { cld2->apply(ego->cld2, I, O); cld2rest = (plan_rdft *) ego->cld2rest; if (cld2rest) { /* leftover from unequal block sizes */ cld2rest->apply(ego->cld2rest, I + ego->rest_Ioff, O + ego->rest_Ooff); cld3 = (plan_rdft *) ego->cld3; if (cld3) cld3->apply(ego->cld3, O, O); /* else TRANSPOSED_OUT is true and user wants O transposed */ } } } static int applicable(const S *ego, const problem *p_, const planner *plnr) { const problem_mpi_transpose *p = (const problem_mpi_transpose *) p_; return (1 && p->I != p->O && (!NO_DESTROY_INPUTP(plnr) || ((p->flags & TRANSPOSED_IN) && !ego->copy_transposed_in)) && ((p->flags & TRANSPOSED_IN) || !ego->copy_transposed_in) && ONLY_TRANSPOSEDP(p->flags) ); } static void awake(plan *ego_, enum wakefulness wakefulness) { P *ego = (P *) ego_; X(plan_awake)(ego->cld1, wakefulness); X(plan_awake)(ego->cld2, wakefulness); X(plan_awake)(ego->cld2rest, wakefulness); X(plan_awake)(ego->cld3, wakefulness); } static void destroy(plan *ego_) { P *ego = (P *) ego_; X(ifree0)(ego->send_block_sizes); MPI_Comm_free(&ego->comm); X(plan_destroy_internal)(ego->cld3); X(plan_destroy_internal)(ego->cld2rest); X(plan_destroy_internal)(ego->cld2); X(plan_destroy_internal)(ego->cld1); } static void print(const plan *ego_, printer *p) { const P *ego = (const P *) ego_; p->print(p, "(mpi-transpose-alltoall%s%(%p%)%(%p%)%(%p%)%(%p%))", ego->equal_blocks ? "/e" : "", ego->cld1, ego->cld2, ego->cld2rest, ego->cld3); } static plan *mkplan(const solver *ego_, const problem *p_, planner *plnr) { const S *ego = (const S *) ego_; const problem_mpi_transpose *p; P *pln; plan *cld1 = 0, *cld2 = 0, *cld2rest = 0, *cld3 = 0; INT b, bt, vn, rest_Ioff, rest_Ooff; R *I; int *sbs, *sbo, *rbs, *rbo; int pe, my_pe, n_pes; int equal_blocks = 1; static const plan_adt padt = { XM(transpose_solve), awake, print, destroy }; if (!applicable(ego, p_, plnr)) return (plan *) 0; p = (const problem_mpi_transpose *) p_; vn = p->vn; MPI_Comm_rank(p->comm, &my_pe); MPI_Comm_size(p->comm, &n_pes); b = XM(block)(p->nx, p->block, my_pe); if (p->flags & TRANSPOSED_IN) { /* I is already transposed */ if (ego->copy_transposed_in) { cld1 = X(mkplan_f_d)(plnr, X(mkproblem_rdft_0_d)(X(mktensor_1d) (b * p->ny * vn, 1, 1), I = p->I, p->O), 0, 0, NO_SLOW); if (XM(any_true)(!cld1, p->comm)) goto nada; } else I = p->O; /* final transpose is in-place */ } else { /* transpose b x ny x vn -> ny x b x vn */ cld1 = X(mkplan_f_d)(plnr, X(mkproblem_rdft_0_d)(X(mktensor_3d) (b, p->ny * vn, vn, p->ny, vn, b * vn, vn, 1, 1), I = p->I, p->O), 0, 0, NO_SLOW); if (XM(any_true)(!cld1, p->comm)) goto nada; } if (XM(any_true)(!XM(mkplans_posttranspose)(p, plnr, I, p->O, my_pe, &cld2, &cld2rest, &cld3, &rest_Ioff, &rest_Ooff), p->comm)) goto nada; pln = MKPLAN_MPI_TRANSPOSE(P, &padt, apply); pln->cld1 = cld1; pln->cld2 = cld2; pln->cld2rest = cld2rest; pln->rest_Ioff = rest_Ioff; pln->rest_Ooff = rest_Ooff; pln->cld3 = cld3; MPI_Comm_dup(p->comm, &pln->comm); /* Compute sizes/offsets of blocks to send for all-to-all command. */ sbs = (int *) MALLOC(4 * n_pes * sizeof(int), PLANS); sbo = sbs + n_pes; rbs = sbo + n_pes; rbo = rbs + n_pes; b = XM(block)(p->nx, p->block, my_pe); bt = XM(block)(p->ny, p->tblock, my_pe); for (pe = 0; pe < n_pes; ++pe) { INT db, dbt; /* destination block sizes */ db = XM(block)(p->nx, p->block, pe); dbt = XM(block)(p->ny, p->tblock, pe); if (db != p->block || dbt != p->tblock) equal_blocks = 0; /* MPI requires type "int" here; apparently it has no 64-bit API? Grrr. */ sbs[pe] = (int) (b * dbt * vn); sbo[pe] = (int) (pe * (b * p->tblock) * vn); rbs[pe] = (int) (db * bt * vn); rbo[pe] = (int) (pe * (p->block * bt) * vn); } pln->send_block_sizes = sbs; pln->send_block_offsets = sbo; pln->recv_block_sizes = rbs; pln->recv_block_offsets = rbo; pln->equal_blocks = equal_blocks; X(ops_zero)(&pln->super.super.ops); if (cld1) X(ops_add2)(&cld1->ops, &pln->super.super.ops); if (cld2) X(ops_add2)(&cld2->ops, &pln->super.super.ops); if (cld2rest) X(ops_add2)(&cld2rest->ops, &pln->super.super.ops); if (cld3) X(ops_add2)(&cld3->ops, &pln->super.super.ops); /* FIXME: should MPI operations be counted in "other" somehow? */ return &(pln->super.super); nada: X(plan_destroy_internal)(cld3); X(plan_destroy_internal)(cld2rest); X(plan_destroy_internal)(cld2); X(plan_destroy_internal)(cld1); return (plan *) 0; } static solver *mksolver(int copy_transposed_in) { static const solver_adt sadt = { PROBLEM_MPI_TRANSPOSE, mkplan, 0 }; S *slv = MKSOLVER(S, &sadt); slv->copy_transposed_in = copy_transposed_in; return &(slv->super); } void XM(transpose_alltoall_register)(planner *p) { int cti; for (cti = 0; cti <= 1; ++cti) REGISTER_SOLVER(p, mksolver(cti)); } fftw-3.3.3/mpi/transpose-pairwise.c0000644000175000017500000003615212054410050014175 00000000000000/* * Copyright (c) 2003, 2007-11 Matteo Frigo * Copyright (c) 2003, 2007-11 Massachusetts Institute of Technology * * This program is free software; you can redistribute it and/or modify * it under the terms of the GNU General Public License as published by * the Free Software Foundation; either version 2 of the License, or * (at your option) any later version. * * This program is distributed in the hope that it will be useful, * but WITHOUT ANY WARRANTY; without even the implied warranty of * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the * GNU General Public License for more details. * * You should have received a copy of the GNU General Public License * along with this program; if not, write to the Free Software * Foundation, Inc., 51 Franklin Street, Fifth Floor, Boston, MA 02110-1301 USA * */ /* Distributed transposes using a sequence of carefully scheduled pairwise exchanges. This has the advantage that it can be done in-place, or out-of-place while preserving the input, using buffer space proportional to the local size divided by the number of processes (i.e. to the total array size divided by the number of processes squared). */ #include "mpi-transpose.h" #include typedef struct { solver super; int preserve_input; /* preserve input even if DESTROY_INPUT was passed */ } S; typedef struct { plan_mpi_transpose super; plan *cld1, *cld2, *cld2rest, *cld3; INT rest_Ioff, rest_Ooff; int n_pes, my_pe, *sched; INT *send_block_sizes, *send_block_offsets; INT *recv_block_sizes, *recv_block_offsets; MPI_Comm comm; int preserve_input; } P; static void transpose_chunks(int *sched, int n_pes, int my_pe, INT *sbs, INT *sbo, INT *rbs, INT *rbo, MPI_Comm comm, R *I, R *O) { if (sched) { int i; MPI_Status status; /* TODO: explore non-synchronous send/recv? */ if (I == O) { R *buf = (R*) MALLOC(sizeof(R) * sbs[0], BUFFERS); for (i = 0; i < n_pes; ++i) { int pe = sched[i]; if (my_pe == pe) { if (rbo[pe] != sbo[pe]) memmove(O + rbo[pe], O + sbo[pe], sbs[pe] * sizeof(R)); } else { memcpy(buf, O + sbo[pe], sbs[pe] * sizeof(R)); MPI_Sendrecv(buf, (int) (sbs[pe]), FFTW_MPI_TYPE, pe, (my_pe * n_pes + pe) & 0xffff, O + rbo[pe], (int) (rbs[pe]), FFTW_MPI_TYPE, pe, (pe * n_pes + my_pe) & 0xffff, comm, &status); } } X(ifree)(buf); } else { /* I != O */ for (i = 0; i < n_pes; ++i) { int pe = sched[i]; if (my_pe == pe) memcpy(O + rbo[pe], I + sbo[pe], sbs[pe] * sizeof(R)); else MPI_Sendrecv(I + sbo[pe], (int) (sbs[pe]), FFTW_MPI_TYPE, pe, (my_pe * n_pes + pe) & 0xffff, O + rbo[pe], (int) (rbs[pe]), FFTW_MPI_TYPE, pe, (pe * n_pes + my_pe) & 0xffff, comm, &status); } } } } static void apply(const plan *ego_, R *I, R *O) { const P *ego = (const P *) ego_; plan_rdft *cld1, *cld2, *cld2rest, *cld3; /* transpose locally to get contiguous chunks */ cld1 = (plan_rdft *) ego->cld1; if (cld1) { cld1->apply(ego->cld1, I, O); if (ego->preserve_input) I = O; /* transpose chunks globally */ transpose_chunks(ego->sched, ego->n_pes, ego->my_pe, ego->send_block_sizes, ego->send_block_offsets, ego->recv_block_sizes, ego->recv_block_offsets, ego->comm, O, I); } else if (ego->preserve_input) { /* transpose chunks globally */ transpose_chunks(ego->sched, ego->n_pes, ego->my_pe, ego->send_block_sizes, ego->send_block_offsets, ego->recv_block_sizes, ego->recv_block_offsets, ego->comm, I, O); I = O; } else { /* transpose chunks globally */ transpose_chunks(ego->sched, ego->n_pes, ego->my_pe, ego->send_block_sizes, ego->send_block_offsets, ego->recv_block_sizes, ego->recv_block_offsets, ego->comm, I, I); } /* transpose locally, again, to get ordinary row-major; this may take two transposes if the block sizes are unequal (3 subplans, two of which operate on disjoint data) */ cld2 = (plan_rdft *) ego->cld2; cld2->apply(ego->cld2, I, O); cld2rest = (plan_rdft *) ego->cld2rest; if (cld2rest) { cld2rest->apply(ego->cld2rest, I + ego->rest_Ioff, O + ego->rest_Ooff); cld3 = (plan_rdft *) ego->cld3; if (cld3) cld3->apply(ego->cld3, O, O); /* else TRANSPOSED_OUT is true and user wants O transposed */ } } static int applicable(const S *ego, const problem *p_, const planner *plnr) { const problem_mpi_transpose *p = (const problem_mpi_transpose *) p_; /* Note: this is *not* UGLY for out-of-place, destroy-input plans; the planner often prefers transpose-pairwise to transpose-alltoall, at least with LAM MPI on my machine. */ return (1 && (!ego->preserve_input || (!NO_DESTROY_INPUTP(plnr) && p->I != p->O)) && ONLY_TRANSPOSEDP(p->flags)); } static void awake(plan *ego_, enum wakefulness wakefulness) { P *ego = (P *) ego_; X(plan_awake)(ego->cld1, wakefulness); X(plan_awake)(ego->cld2, wakefulness); X(plan_awake)(ego->cld2rest, wakefulness); X(plan_awake)(ego->cld3, wakefulness); } static void destroy(plan *ego_) { P *ego = (P *) ego_; X(ifree0)(ego->sched); X(ifree0)(ego->send_block_sizes); MPI_Comm_free(&ego->comm); X(plan_destroy_internal)(ego->cld3); X(plan_destroy_internal)(ego->cld2rest); X(plan_destroy_internal)(ego->cld2); X(plan_destroy_internal)(ego->cld1); } static void print(const plan *ego_, printer *p) { const P *ego = (const P *) ego_; p->print(p, "(mpi-transpose-pairwise%s%(%p%)%(%p%)%(%p%)%(%p%))", ego->preserve_input==2 ?"/p":"", ego->cld1, ego->cld2, ego->cld2rest, ego->cld3); } /* Given a process which_pe and a number of processes npes, fills the array sched[npes] with a sequence of processes to communicate with for a deadlock-free, optimum-overlap all-to-all communication. (All processes must call this routine to get their own schedules.) The schedule can be re-ordered arbitrarily as long as all processes apply the same permutation to their schedules. The algorithm here is based upon the one described in: J. A. M. Schreuder, "Constructing timetables for sport competitions," Mathematical Programming Study 13, pp. 58-67 (1980). In a sport competition, you have N teams and want every team to play every other team in as short a time as possible (maximum overlap between games). This timetabling problem is therefore identical to that of an all-to-all communications problem. In our case, there is one wrinkle: as part of the schedule, the process must do some data transfer with itself (local data movement), analogous to a requirement that each team "play itself" in addition to other teams. With this wrinkle, it turns out that an optimal timetable (N parallel games) can be constructed for any N, not just for even N as in the original problem described by Schreuder. */ static void fill1_comm_sched(int *sched, int which_pe, int npes) { int pe, i, n, s = 0; A(which_pe >= 0 && which_pe < npes); if (npes % 2 == 0) { n = npes; sched[s++] = which_pe; } else n = npes + 1; for (pe = 0; pe < n - 1; ++pe) { if (npes % 2 == 0) { if (pe == which_pe) sched[s++] = npes - 1; else if (npes - 1 == which_pe) sched[s++] = pe; } else if (pe == which_pe) sched[s++] = pe; if (pe != which_pe && which_pe < n - 1) { i = (pe - which_pe + (n - 1)) % (n - 1); if (i < n/2) sched[s++] = (pe + i) % (n - 1); i = (which_pe - pe + (n - 1)) % (n - 1); if (i < n/2) sched[s++] = (pe - i + (n - 1)) % (n - 1); } } A(s == npes); } /* Sort the communication schedule sched for npes so that the schedule on process sortpe is ascending or descending (!ascending). This is necessary to allow in-place transposes when the problem does not divide equally among the processes. In this case there is one process where the incoming blocks are bigger/smaller than the outgoing blocks and thus have to be received in descending/ascending order, respectively, to avoid overwriting data before it is sent. */ static void sort1_comm_sched(int *sched, int npes, int sortpe, int ascending) { int *sortsched, i; sortsched = (int *) MALLOC(npes * sizeof(int) * 2, OTHER); fill1_comm_sched(sortsched, sortpe, npes); if (ascending) for (i = 0; i < npes; ++i) sortsched[npes + sortsched[i]] = sched[i]; else for (i = 0; i < npes; ++i) sortsched[2*npes - 1 - sortsched[i]] = sched[i]; for (i = 0; i < npes; ++i) sched[i] = sortsched[npes + i]; X(ifree)(sortsched); } /* make the plans to do the post-MPI transpositions (shared with transpose-alltoall) */ int XM(mkplans_posttranspose)(const problem_mpi_transpose *p, planner *plnr, R *I, R *O, int my_pe, plan **cld2, plan **cld2rest, plan **cld3, INT *rest_Ioff, INT *rest_Ooff) { INT vn = p->vn; INT b = p->block; INT bt = XM(block)(p->ny, p->tblock, my_pe); INT nxb = p->nx / b; /* number of equal-sized blocks */ INT nxr = p->nx - nxb * b; /* leftover rows after equal blocks */ *cld2 = *cld2rest = *cld3 = NULL; *rest_Ioff = *rest_Ooff = 0; if (!(p->flags & TRANSPOSED_OUT) && (nxr == 0 || I != O)) { INT nx = p->nx * vn; b *= vn; *cld2 = X(mkplan_f_d)(plnr, X(mkproblem_rdft_0_d)(X(mktensor_3d) (nxb, bt * b, b, bt, b, nx, b, 1, 1), I, O), 0, 0, NO_SLOW); if (!*cld2) goto nada; if (nxr > 0) { *rest_Ioff = nxb * bt * b; *rest_Ooff = nxb * b; b = nxr * vn; *cld2rest = X(mkplan_f_d)(plnr, X(mkproblem_rdft_0_d)(X(mktensor_2d) (bt, b, nx, b, 1, 1), I + *rest_Ioff, O + *rest_Ooff), 0, 0, NO_SLOW); if (!*cld2rest) goto nada; } } else { *cld2 = X(mkplan_f_d)(plnr, X(mkproblem_rdft_0_d)( X(mktensor_4d) (nxb, bt * b * vn, bt * b * vn, bt, b * vn, vn, b, vn, bt * vn, vn, 1, 1), I, O), 0, 0, NO_SLOW); if (!*cld2) goto nada; *rest_Ioff = *rest_Ooff = nxb * bt * b * vn; *cld2rest = X(mkplan_f_d)(plnr, X(mkproblem_rdft_0_d)( X(mktensor_3d) (bt, nxr * vn, vn, nxr, vn, bt * vn, vn, 1, 1), I + *rest_Ioff, O + *rest_Ooff), 0, 0, NO_SLOW); if (!*cld2rest) goto nada; if (!(p->flags & TRANSPOSED_OUT)) { *cld3 = X(mkplan_f_d)(plnr, X(mkproblem_rdft_0_d)( X(mktensor_3d) (p->nx, bt * vn, vn, bt, vn, p->nx * vn, vn, 1, 1), O, O), 0, 0, NO_SLOW); if (!*cld3) goto nada; } } return 1; nada: X(plan_destroy_internal)(*cld3); X(plan_destroy_internal)(*cld2rest); X(plan_destroy_internal)(*cld2); return 0; } static plan *mkplan(const solver *ego_, const problem *p_, planner *plnr) { const S *ego = (const S *) ego_; const problem_mpi_transpose *p; P *pln; plan *cld1 = 0, *cld2 = 0, *cld2rest = 0, *cld3 = 0; INT b, bt, vn, rest_Ioff, rest_Ooff; INT *sbs, *sbo, *rbs, *rbo; int pe, my_pe, n_pes, sort_pe = -1, ascending = 1; R *I, *O; static const plan_adt padt = { XM(transpose_solve), awake, print, destroy }; UNUSED(ego); if (!applicable(ego, p_, plnr)) return (plan *) 0; p = (const problem_mpi_transpose *) p_; vn = p->vn; I = p->I; O = p->O; MPI_Comm_rank(p->comm, &my_pe); MPI_Comm_size(p->comm, &n_pes); b = XM(block)(p->nx, p->block, my_pe); if (!(p->flags & TRANSPOSED_IN)) { /* b x ny x vn -> ny x b x vn */ cld1 = X(mkplan_f_d)(plnr, X(mkproblem_rdft_0_d)(X(mktensor_3d) (b, p->ny * vn, vn, p->ny, vn, b * vn, vn, 1, 1), I, O), 0, 0, NO_SLOW); if (XM(any_true)(!cld1, p->comm)) goto nada; } if (ego->preserve_input || NO_DESTROY_INPUTP(plnr)) I = O; if (XM(any_true)(!XM(mkplans_posttranspose)(p, plnr, I, O, my_pe, &cld2, &cld2rest, &cld3, &rest_Ioff, &rest_Ooff), p->comm)) goto nada; pln = MKPLAN_MPI_TRANSPOSE(P, &padt, apply); pln->cld1 = cld1; pln->cld2 = cld2; pln->cld2rest = cld2rest; pln->rest_Ioff = rest_Ioff; pln->rest_Ooff = rest_Ooff; pln->cld3 = cld3; pln->preserve_input = ego->preserve_input ? 2 : NO_DESTROY_INPUTP(plnr); MPI_Comm_dup(p->comm, &pln->comm); n_pes = (int) X(imax)(XM(num_blocks)(p->nx, p->block), XM(num_blocks)(p->ny, p->tblock)); /* Compute sizes/offsets of blocks to exchange between processors */ sbs = (INT *) MALLOC(4 * n_pes * sizeof(INT), PLANS); sbo = sbs + n_pes; rbs = sbo + n_pes; rbo = rbs + n_pes; b = XM(block)(p->nx, p->block, my_pe); bt = XM(block)(p->ny, p->tblock, my_pe); for (pe = 0; pe < n_pes; ++pe) { INT db, dbt; /* destination block sizes */ db = XM(block)(p->nx, p->block, pe); dbt = XM(block)(p->ny, p->tblock, pe); sbs[pe] = b * dbt * vn; sbo[pe] = pe * (b * p->tblock) * vn; rbs[pe] = db * bt * vn; rbo[pe] = pe * (p->block * bt) * vn; if (db * dbt > 0 && db * p->tblock != p->block * dbt) { A(sort_pe == -1); /* only one process should need sorting */ sort_pe = pe; ascending = db * p->tblock > p->block * dbt; } } pln->n_pes = n_pes; pln->my_pe = my_pe; pln->send_block_sizes = sbs; pln->send_block_offsets = sbo; pln->recv_block_sizes = rbs; pln->recv_block_offsets = rbo; if (my_pe >= n_pes) { pln->sched = 0; /* this process is not doing anything */ } else { pln->sched = (int *) MALLOC(n_pes * sizeof(int), PLANS); fill1_comm_sched(pln->sched, my_pe, n_pes); if (sort_pe >= 0) sort1_comm_sched(pln->sched, n_pes, sort_pe, ascending); } X(ops_zero)(&pln->super.super.ops); if (cld1) X(ops_add2)(&cld1->ops, &pln->super.super.ops); if (cld2) X(ops_add2)(&cld2->ops, &pln->super.super.ops); if (cld2rest) X(ops_add2)(&cld2rest->ops, &pln->super.super.ops); if (cld3) X(ops_add2)(&cld3->ops, &pln->super.super.ops); /* FIXME: should MPI operations be counted in "other" somehow? */ return &(pln->super.super); nada: X(plan_destroy_internal)(cld3); X(plan_destroy_internal)(cld2rest); X(plan_destroy_internal)(cld2); X(plan_destroy_internal)(cld1); return (plan *) 0; } static solver *mksolver(int preserve_input) { static const solver_adt sadt = { PROBLEM_MPI_TRANSPOSE, mkplan, 0 }; S *slv = MKSOLVER(S, &sadt); slv->preserve_input = preserve_input; return &(slv->super); } void XM(transpose_pairwise_register)(planner *p) { int preserve_input; for (preserve_input = 0; preserve_input <= 1; ++preserve_input) REGISTER_SOLVER(p, mksolver(preserve_input)); } fftw-3.3.3/mpi/transpose-recurse.c0000644000175000017500000002327012054410050014017 00000000000000/* * Copyright (c) 2003, 2007-11 Matteo Frigo * Copyright (c) 2003, 2007-11 Massachusetts Institute of Technology * * This program is free software; you can redistribute it and/or modify * it under the terms of the GNU General Public License as published by * the Free Software Foundation; either version 2 of the License, or * (at your option) any later version. * * This program is distributed in the hope that it will be useful, * but WITHOUT ANY WARRANTY; without even the implied warranty of * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the * GNU General Public License for more details. * * You should have received a copy of the GNU General Public License * along with this program; if not, write to the Free Software * Foundation, Inc., 51 Franklin Street, Fifth Floor, Boston, MA 02110-1301 USA * */ /* Recursive "radix-r" distributed transpose, which breaks a transpose over p processes into p/r transposes over r processes plus r transposes over p/r processes. If performed recursively, this produces a total of O(p log p) messages vs. O(p^2) messages for a direct approach. However, this is not necessarily an improvement. The total size of all the messages is actually increased from O(N) to O(N log p) where N is the total data size. Also, the amount of local data rearrangement is increased. So, it's not clear, a priori, what the best algorithm will be, and we'll leave it to the planner. (In theory and practice, it looks like this becomes advantageous for large p, in the limit where the message sizes are small and latency-dominated.) */ #include "mpi-transpose.h" #include typedef struct { solver super; int (*radix)(int np); const char *nam; int preserve_input; /* preserve input even if DESTROY_INPUT was passed */ } S; typedef struct { plan_mpi_transpose super; plan *cld1, *cldtr, *cldtm; int preserve_input; int r; /* "radix" */ const char *nam; } P; static void apply(const plan *ego_, R *I, R *O) { const P *ego = (const P *) ego_; plan_rdft *cld1, *cldtr, *cldtm; cld1 = (plan_rdft *) ego->cld1; if (cld1) cld1->apply((plan *) cld1, I, O); if (ego->preserve_input) I = O; cldtr = (plan_rdft *) ego->cldtr; if (cldtr) cldtr->apply((plan *) cldtr, O, I); cldtm = (plan_rdft *) ego->cldtm; if (cldtm) cldtm->apply((plan *) cldtm, I, O); } static int radix_sqrt(int np) { int r; for (r = (int) (X(isqrt)(np)); np % r != 0; ++r) ; return r; } static int radix_first(int np) { int r = (int) (X(first_divisor)(np)); return (r >= (int) (X(isqrt)(np)) ? 0 : r); } /* the local allocated space on process pe required for the given transpose dimensions and block sizes */ static INT transpose_space(INT nx, INT ny, INT block, INT tblock, int pe) { return X(imax)(XM(block)(nx, block, pe) * ny, nx * XM(block)(ny, tblock, pe)); } /* check whether the recursive transposes fit within the space that must have been allocated on each process for this transpose; this must be modified if the subdivision in mkplan is changed! */ static int enough_space(INT nx, INT ny, INT block, INT tblock, int r, int n_pes) { int pe; int m = n_pes / r; for (pe = 0; pe < n_pes; ++pe) { INT space = transpose_space(nx, ny, block, tblock, pe); INT b1 = XM(block)(nx, r * block, pe / r); INT b2 = XM(block)(ny, m * tblock, pe % r); if (transpose_space(b1, ny, block, m*tblock, pe % r) > space || transpose_space(nx, b2, r*block, tblock, pe / r) > space) return 0; } return 1; } /* In theory, transpose-recurse becomes advantageous for message sizes below some minimum, assuming that the time is dominated by communications. In practice, we want to constrain the minimum message size for transpose-recurse to keep the planning time down. I've set this conservatively according to some simple experiments on a Cray XT3 where the crossover message size was 128, although on a larger-latency machine the crossover will be larger. */ #define SMALL_MESSAGE 2048 static int applicable(const S *ego, const problem *p_, const planner *plnr, int *r) { const problem_mpi_transpose *p = (const problem_mpi_transpose *) p_; int n_pes; MPI_Comm_size(p->comm, &n_pes); return (1 && p->tblock * n_pes == p->ny && (!ego->preserve_input || (!NO_DESTROY_INPUTP(plnr) && p->I != p->O)) && (*r = ego->radix(n_pes)) && *r < n_pes && *r > 1 && enough_space(p->nx, p->ny, p->block, p->tblock, *r, n_pes) && (!CONSERVE_MEMORYP(plnr) || *r > 8 || !X(toobig)((p->nx * (p->ny / n_pes) * p->vn) / *r)) && (!NO_SLOWP(plnr) || (p->nx * (p->ny / n_pes) * p->vn) / n_pes <= SMALL_MESSAGE) && ONLY_TRANSPOSEDP(p->flags) ); } static void awake(plan *ego_, enum wakefulness wakefulness) { P *ego = (P *) ego_; X(plan_awake)(ego->cld1, wakefulness); X(plan_awake)(ego->cldtr, wakefulness); X(plan_awake)(ego->cldtm, wakefulness); } static void destroy(plan *ego_) { P *ego = (P *) ego_; X(plan_destroy_internal)(ego->cldtm); X(plan_destroy_internal)(ego->cldtr); X(plan_destroy_internal)(ego->cld1); } static void print(const plan *ego_, printer *p) { const P *ego = (const P *) ego_; p->print(p, "(mpi-transpose-recurse/%s/%d%s%(%p%)%(%p%)%(%p%))", ego->nam, ego->r, ego->preserve_input==2 ?"/p":"", ego->cld1, ego->cldtr, ego->cldtm); } static plan *mkplan(const solver *ego_, const problem *p_, planner *plnr) { const S *ego = (const S *) ego_; const problem_mpi_transpose *p; P *pln; plan *cld1 = 0, *cldtr = 0, *cldtm = 0; R *I, *O; int me, np, r, m; INT b; MPI_Comm comm2; static const plan_adt padt = { XM(transpose_solve), awake, print, destroy }; UNUSED(ego); if (!applicable(ego, p_, plnr, &r)) return (plan *) 0; p = (const problem_mpi_transpose *) p_; MPI_Comm_size(p->comm, &np); MPI_Comm_rank(p->comm, &me); m = np / r; A(r * m == np); I = p->I; O = p->O; b = XM(block)(p->nx, p->block, me); A(p->tblock * np == p->ny); /* this is currently required for cld1 */ if (p->flags & TRANSPOSED_IN) { /* m x r x (bt x b x vn) -> r x m x (bt x b x vn) */ INT vn = p->vn * b * p->tblock; cld1 = X(mkplan_f_d)(plnr, X(mkproblem_rdft_0_d)(X(mktensor_3d) (m, r*vn, vn, r, vn, m*vn, vn, 1, 1), I, O), 0, 0, NO_SLOW); } else if (I != O) { /* combine cld1 with TRANSPOSED_IN permutation */ /* b x m x r x bt x vn -> r x m x bt x b x vn */ INT vn = p->vn; INT bt = p->tblock; cld1 = X(mkplan_f_d)(plnr, X(mkproblem_rdft_0_d)(X(mktensor_5d) (b, m*r*bt*vn, vn, m, r*bt*vn, bt*b*vn, r, bt*vn, m*bt*b*vn, bt, vn, b*vn, vn, 1, 1), I, O), 0, 0, NO_SLOW); } else { /* TRANSPOSED_IN permutation must be separate for in-place */ /* b x (m x r) x bt x vn -> b x (r x m) x bt x vn */ INT vn = p->vn * p->tblock; cld1 = X(mkplan_f_d)(plnr, X(mkproblem_rdft_0_d)(X(mktensor_4d) (m, r*vn, vn, r, vn, m*vn, vn, 1, 1, b, np*vn, np*vn), I, O), 0, 0, NO_SLOW); } if (XM(any_true)(!cld1, p->comm)) goto nada; if (ego->preserve_input || NO_DESTROY_INPUTP(plnr)) I = O; b = XM(block)(p->nx, r * p->block, me / r); MPI_Comm_split(p->comm, me / r, me, &comm2); if (b) cldtr = X(mkplan_d)(plnr, XM(mkproblem_transpose) (b, p->ny, p->vn, O, I, p->block, m * p->tblock, comm2, p->I != p->O ? TRANSPOSED_IN : (p->flags & TRANSPOSED_IN))); MPI_Comm_free(&comm2); if (XM(any_true)(b && !cldtr, p->comm)) goto nada; b = XM(block)(p->ny, m * p->tblock, me % r); MPI_Comm_split(p->comm, me % r, me, &comm2); if (b) cldtm = X(mkplan_d)(plnr, XM(mkproblem_transpose) (p->nx, b, p->vn, I, O, r * p->block, p->tblock, comm2, TRANSPOSED_IN | (p->flags & TRANSPOSED_OUT))); MPI_Comm_free(&comm2); if (XM(any_true)(b && !cldtm, p->comm)) goto nada; pln = MKPLAN_MPI_TRANSPOSE(P, &padt, apply); pln->cld1 = cld1; pln->cldtr = cldtr; pln->cldtm = cldtm; pln->preserve_input = ego->preserve_input ? 2 : NO_DESTROY_INPUTP(plnr); pln->r = r; pln->nam = ego->nam; pln->super.super.ops = cld1->ops; if (cldtr) X(ops_add2)(&cldtr->ops, &pln->super.super.ops); if (cldtm) X(ops_add2)(&cldtm->ops, &pln->super.super.ops); return &(pln->super.super); nada: X(plan_destroy_internal)(cldtm); X(plan_destroy_internal)(cldtr); X(plan_destroy_internal)(cld1); return (plan *) 0; } static solver *mksolver(int preserve_input, int (*radix)(int np), const char *nam) { static const solver_adt sadt = { PROBLEM_MPI_TRANSPOSE, mkplan, 0 }; S *slv = MKSOLVER(S, &sadt); slv->preserve_input = preserve_input; slv->radix = radix; slv->nam = nam; return &(slv->super); } void XM(transpose_recurse_register)(planner *p) { int preserve_input; for (preserve_input = 0; preserve_input <= 1; ++preserve_input) { REGISTER_SOLVER(p, mksolver(preserve_input, radix_sqrt, "sqrt")); REGISTER_SOLVER(p, mksolver(preserve_input, radix_first, "first")); } } fftw-3.3.3/mpi/transpose-problem.c0000644000175000017500000000705312054410050014010 00000000000000/* * Copyright (c) 2003, 2007-11 Matteo Frigo * Copyright (c) 2003, 2007-11 Massachusetts Institute of Technology * * This program is free software; you can redistribute it and/or modify * it under the terms of the GNU General Public License as published by * the Free Software Foundation; either version 2 of the License, or * (at your option) any later version. * * This program is distributed in the hope that it will be useful, * but WITHOUT ANY WARRANTY; without even the implied warranty of * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the * GNU General Public License for more details. * * You should have received a copy of the GNU General Public License * along with this program; if not, write to the Free Software * Foundation, Inc., 51 Franklin Street, Fifth Floor, Boston, MA 02110-1301 USA * */ #include "mpi-transpose.h" static void destroy(problem *ego_) { problem_mpi_transpose *ego = (problem_mpi_transpose *) ego_; MPI_Comm_free(&ego->comm); X(ifree)(ego_); } static void hash(const problem *p_, md5 *m) { const problem_mpi_transpose *p = (const problem_mpi_transpose *) p_; int i; X(md5puts)(m, "mpi-transpose"); X(md5int)(m, p->I == p->O); /* don't include alignment -- may differ between processes X(md5int)(m, X(alignment_of)(p->I)); X(md5int)(m, X(alignment_of)(p->O)); ... note that applicability of MPI plans does not depend on alignment (although optimality may, in principle). */ X(md5INT)(m, p->vn); X(md5INT)(m, p->nx); X(md5INT)(m, p->ny); X(md5INT)(m, p->block); X(md5INT)(m, p->tblock); MPI_Comm_size(p->comm, &i); X(md5int)(m, i); A(XM(md5_equal)(*m, p->comm)); } static void print(const problem *ego_, printer *p) { const problem_mpi_transpose *ego = (const problem_mpi_transpose *) ego_; int i; MPI_Comm_size(ego->comm, &i); p->print(p, "(mpi-transpose %d %d %d %D %D %D %D %D %d)", ego->I == ego->O, X(alignment_of)(ego->I), X(alignment_of)(ego->O), ego->vn, ego->nx, ego->ny, ego->block, ego->tblock, i); } static void zero(const problem *ego_) { const problem_mpi_transpose *ego = (const problem_mpi_transpose *) ego_; R *I = ego->I; INT i, N = ego->vn * ego->ny; int my_pe; MPI_Comm_rank(ego->comm, &my_pe); N *= XM(block)(ego->nx, ego->block, my_pe); for (i = 0; i < N; ++i) I[i] = K(0.0); } static const problem_adt padt = { PROBLEM_MPI_TRANSPOSE, hash, zero, print, destroy }; problem *XM(mkproblem_transpose)(INT nx, INT ny, INT vn, R *I, R *O, INT block, INT tblock, MPI_Comm comm, unsigned flags) { problem_mpi_transpose *ego = (problem_mpi_transpose *)X(mkproblem)(sizeof(problem_mpi_transpose), &padt); A(nx > 0 && ny > 0 && vn > 0); A(block > 0 && XM(num_blocks_ok)(nx, block, comm) && tblock > 0 && XM(num_blocks_ok)(ny, tblock, comm)); /* enforce pointer equality if untainted pointers are equal */ if (UNTAINT(I) == UNTAINT(O)) I = O = JOIN_TAINT(I, O); ego->nx = nx; ego->ny = ny; ego->vn = vn; ego->I = I; ego->O = O; ego->block = block > nx ? nx : block; ego->tblock = tblock > ny ? ny : tblock; /* canonicalize flags: we can freely assume that the data is "transposed" if one of the dimensions is 1. */ if (ego->block == 1) flags |= TRANSPOSED_IN; if (ego->tblock == 1) flags |= TRANSPOSED_OUT; ego->flags = flags; MPI_Comm_dup(comm, &ego->comm); return &(ego->super); } fftw-3.3.3/mpi/transpose-solve.c0000644000175000017500000000225512054410050013477 00000000000000/* * Copyright (c) 2003, 2007-11 Matteo Frigo * Copyright (c) 2003, 2007-11 Massachusetts Institute of Technology * * This program is free software; you can redistribute it and/or modify * it under the terms of the GNU General Public License as published by * the Free Software Foundation; either version 2 of the License, or * (at your option) any later version. * * This program is distributed in the hope that it will be useful, * but WITHOUT ANY WARRANTY; without even the implied warranty of * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the * GNU General Public License for more details. * * You should have received a copy of the GNU General Public License * along with this program; if not, write to the Free Software * Foundation, Inc., 51 Franklin Street, Fifth Floor, Boston, MA 02110-1301 USA * */ #include "mpi-transpose.h" /* use the apply() operation for MPI_TRANSPOSE problems */ void XM(transpose_solve)(const plan *ego_, const problem *p_) { const plan_mpi_transpose *ego = (const plan_mpi_transpose *) ego_; const problem_mpi_transpose *p = (const problem_mpi_transpose *) p_; ego->apply(ego_, UNTAINT(p->I), UNTAINT(p->O)); } fftw-3.3.3/mpi/mpi-transpose.h0000644000175000017500000000422612054410050013141 00000000000000/* * Copyright (c) 2003, 2007-11 Matteo Frigo * Copyright (c) 2003, 2007-11 Massachusetts Institute of Technology * * This program is free software; you can redistribute it and/or modify * it under the terms of the GNU General Public License as published by * the Free Software Foundation; either version 2 of the License, or * (at your option) any later version. * * This program is distributed in the hope that it will be useful, * but WITHOUT ANY WARRANTY; without even the implied warranty of * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the * GNU General Public License for more details. * * You should have received a copy of the GNU General Public License * along with this program; if not, write to the Free Software * Foundation, Inc., 51 Franklin Street, Fifth Floor, Boston, MA 02110-1301 USA * */ #include "ifftw-mpi.h" /* tproblem.c: */ typedef struct { problem super; INT vn; /* vector length (vector stride 1) */ INT nx, ny; /* nx x ny transposed to ny x nx */ R *I, *O; /* contiguous real arrays (both same size!) */ unsigned flags; /* TRANSPOSED_IN: input is *locally* transposed TRANSPOSED_OUT: output is *locally* transposed */ INT block, tblock; /* block size, slab decomposition; tblock is for transposed blocks on output */ MPI_Comm comm; } problem_mpi_transpose; problem *XM(mkproblem_transpose)(INT nx, INT ny, INT vn, R *I, R *O, INT block, INT tblock, MPI_Comm comm, unsigned flags); /* tsolve.c: */ void XM(transpose_solve)(const plan *ego_, const problem *p_); /* plans have same operands as rdft plans, so just re-use */ typedef plan_rdft plan_mpi_transpose; #define MKPLAN_MPI_TRANSPOSE(type, adt, apply) \ (type *)X(mkplan_rdft)(sizeof(type), adt, apply) /* transpose-pairwise.c: */ int XM(mkplans_posttranspose)(const problem_mpi_transpose *p, planner *plnr, R *I, R *O, int my_pe, plan **cld2, plan **cld2rest, plan **cld3, INT *rest_Ioff, INT *rest_Ooff); /* various solvers */ void XM(transpose_pairwise_register)(planner *p); void XM(transpose_alltoall_register)(planner *p); void XM(transpose_recurse_register)(planner *p); fftw-3.3.3/mpi/dft-serial.c0000644000175000017500000000716212054410050012367 00000000000000/* * Copyright (c) 2003, 2007-11 Matteo Frigo * Copyright (c) 2003, 2007-11 Massachusetts Institute of Technology * * This program is free software; you can redistribute it and/or modify * it under the terms of the GNU General Public License as published by * the Free Software Foundation; either version 2 of the License, or * (at your option) any later version. * * This program is distributed in the hope that it will be useful, * but WITHOUT ANY WARRANTY; without even the implied warranty of * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the * GNU General Public License for more details. * * You should have received a copy of the GNU General Public License * along with this program; if not, write to the Free Software * Foundation, Inc., 51 Franklin Street, Fifth Floor, Boston, MA 02110-1301 USA * */ /* "MPI" DFTs where all of the data is on one processor...just call through to serial API. */ #include "mpi-dft.h" #include "dft.h" typedef struct { plan_mpi_dft super; plan *cld; INT roff, ioff; } P; static void apply(const plan *ego_, R *I, R *O) { const P *ego = (const P *) ego_; plan_dft *cld; INT roff = ego->roff, ioff = ego->ioff; cld = (plan_dft *) ego->cld; cld->apply(ego->cld, I+roff, I+ioff, O+roff, O+ioff); } static void awake(plan *ego_, enum wakefulness wakefulness) { P *ego = (P *) ego_; X(plan_awake)(ego->cld, wakefulness); } static void destroy(plan *ego_) { P *ego = (P *) ego_; X(plan_destroy_internal)(ego->cld); } static void print(const plan *ego_, printer *p) { const P *ego = (const P *) ego_; p->print(p, "(mpi-dft-serial %(%p%))", ego->cld); } int XM(dft_serial_applicable)(const problem_mpi_dft *p) { return (1 && p->flags == 0 /* TRANSPOSED/SCRAMBLED_IN/OUT not supported */ && ((XM(is_local)(p->sz, IB) && XM(is_local)(p->sz, OB)) || p->vn == 0)); } static plan *mkplan(const solver *ego, const problem *p_, planner *plnr) { const problem_mpi_dft *p = (const problem_mpi_dft *) p_; P *pln; plan *cld; int my_pe; R *ri, *ii, *ro, *io; static const plan_adt padt = { XM(dft_solve), awake, print, destroy }; UNUSED(ego); /* check whether applicable: */ if (!XM(dft_serial_applicable)(p)) return (plan *) 0; X(extract_reim)(p->sign, p->I, &ri, &ii); X(extract_reim)(p->sign, p->O, &ro, &io); MPI_Comm_rank(p->comm, &my_pe); if (my_pe == 0 && p->vn > 0) { int i, rnk = p->sz->rnk; tensor *sz = X(mktensor)(p->sz->rnk); sz->dims[rnk - 1].is = sz->dims[rnk - 1].os = 2 * p->vn; sz->dims[rnk - 1].n = p->sz->dims[rnk - 1].n; for (i = rnk - 1; i > 0; --i) { sz->dims[i - 1].is = sz->dims[i - 1].os = sz->dims[i].is * sz->dims[i].n; sz->dims[i - 1].n = p->sz->dims[i - 1].n; } cld = X(mkplan_d)(plnr, X(mkproblem_dft_d)(sz, X(mktensor_1d)(p->vn, 2, 2), ri, ii, ro, io)); } else { /* idle process: make nop plan */ cld = X(mkplan_d)(plnr, X(mkproblem_dft_d)(X(mktensor_0d)(), X(mktensor_1d)(0,0,0), ri, ii, ro, io)); } if (XM(any_true)(!cld, p->comm)) return (plan *) 0; pln = MKPLAN_MPI_DFT(P, &padt, apply); pln->cld = cld; pln->roff = ro - p->O; pln->ioff = io - p->O; X(ops_cpy)(&cld->ops, &pln->super.super.ops); return &(pln->super.super); } static solver *mksolver(void) { static const solver_adt sadt = { PROBLEM_MPI_DFT, mkplan, 0 }; return MKSOLVER(solver, &sadt); } void XM(dft_serial_register)(planner *p) { REGISTER_SOLVER(p, mksolver()); } fftw-3.3.3/mpi/dft-rank-geq2.c0000644000175000017500000001314312054410050012673 00000000000000/* * Copyright (c) 2003, 2007-11 Matteo Frigo * Copyright (c) 2003, 2007-11 Massachusetts Institute of Technology * * This program is free software; you can redistribute it and/or modify * it under the terms of the GNU General Public License as published by * the Free Software Foundation; either version 2 of the License, or * (at your option) any later version. * * This program is distributed in the hope that it will be useful, * but WITHOUT ANY WARRANTY; without even the implied warranty of * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the * GNU General Public License for more details. * * You should have received a copy of the GNU General Public License * along with this program; if not, write to the Free Software * Foundation, Inc., 51 Franklin Street, Fifth Floor, Boston, MA 02110-1301 USA * */ /* Complex DFTs of rank >= 2, for the case where we are distributed across the first dimension only, and the output is not transposed. */ #include "mpi-dft.h" #include "dft.h" typedef struct { solver super; int preserve_input; /* preserve input even if DESTROY_INPUT was passed */ } S; typedef struct { plan_mpi_dft super; plan *cld1, *cld2; INT roff, ioff; int preserve_input; } P; static void apply(const plan *ego_, R *I, R *O) { const P *ego = (const P *) ego_; plan_dft *cld1; plan_rdft *cld2; INT roff = ego->roff, ioff = ego->ioff; /* DFT local dimensions */ cld1 = (plan_dft *) ego->cld1; if (ego->preserve_input) { cld1->apply(ego->cld1, I+roff, I+ioff, O+roff, O+ioff); I = O; } else cld1->apply(ego->cld1, I+roff, I+ioff, I+roff, I+ioff); /* DFT non-local dimension (via dft-rank1-bigvec, usually): */ cld2 = (plan_rdft *) ego->cld2; cld2->apply(ego->cld2, I, O); } static int applicable(const S *ego, const problem *p_, const planner *plnr) { const problem_mpi_dft *p = (const problem_mpi_dft *) p_; return (1 && p->sz->rnk > 1 && p->flags == 0 /* TRANSPOSED/SCRAMBLED_IN/OUT not supported */ && (!ego->preserve_input || (!NO_DESTROY_INPUTP(plnr) && p->I != p->O)) && XM(is_local_after)(1, p->sz, IB) && XM(is_local_after)(1, p->sz, OB) && (!NO_SLOWP(plnr) /* slow if dft-serial is applicable */ || !XM(dft_serial_applicable)(p)) ); } static void awake(plan *ego_, enum wakefulness wakefulness) { P *ego = (P *) ego_; X(plan_awake)(ego->cld1, wakefulness); X(plan_awake)(ego->cld2, wakefulness); } static void destroy(plan *ego_) { P *ego = (P *) ego_; X(plan_destroy_internal)(ego->cld2); X(plan_destroy_internal)(ego->cld1); } static void print(const plan *ego_, printer *p) { const P *ego = (const P *) ego_; p->print(p, "(mpi-dft-rank-geq2%s%(%p%)%(%p%))", ego->preserve_input==2 ?"/p":"", ego->cld1, ego->cld2); } static plan *mkplan(const solver *ego_, const problem *p_, planner *plnr) { const S *ego = (const S *) ego_; const problem_mpi_dft *p; P *pln; plan *cld1 = 0, *cld2 = 0; R *ri, *ii, *ro, *io, *I, *O; tensor *sz; dtensor *sz2; int i, my_pe, n_pes; INT nrest; static const plan_adt padt = { XM(dft_solve), awake, print, destroy }; UNUSED(ego); if (!applicable(ego, p_, plnr)) return (plan *) 0; p = (const problem_mpi_dft *) p_; X(extract_reim)(p->sign, I = p->I, &ri, &ii); X(extract_reim)(p->sign, O = p->O, &ro, &io); if (ego->preserve_input || NO_DESTROY_INPUTP(plnr)) I = O; else { ro = ri; io = ii; } MPI_Comm_rank(p->comm, &my_pe); MPI_Comm_size(p->comm, &n_pes); sz = X(mktensor)(p->sz->rnk - 1); /* tensor of last rnk-1 dimensions */ i = p->sz->rnk - 2; A(i >= 0); sz->dims[i].n = p->sz->dims[i+1].n; sz->dims[i].is = sz->dims[i].os = 2 * p->vn; for (--i; i >= 0; --i) { sz->dims[i].n = p->sz->dims[i+1].n; sz->dims[i].is = sz->dims[i].os = sz->dims[i+1].n * sz->dims[i+1].is; } nrest = X(tensor_sz)(sz); { INT is = sz->dims[0].n * sz->dims[0].is; INT b = XM(block)(p->sz->dims[0].n, p->sz->dims[0].b[IB], my_pe); cld1 = X(mkplan_d)(plnr, X(mkproblem_dft_d)(sz, X(mktensor_2d)(b, is, is, p->vn, 2, 2), ri, ii, ro, io)); if (XM(any_true)(!cld1, p->comm)) goto nada; } sz2 = XM(mkdtensor)(1); /* tensor for first (distributed) dimension */ sz2->dims[0] = p->sz->dims[0]; cld2 = X(mkplan_d)(plnr, XM(mkproblem_dft_d)(sz2, nrest * p->vn, I, O, p->comm, p->sign, RANK1_BIGVEC_ONLY)); if (XM(any_true)(!cld2, p->comm)) goto nada; pln = MKPLAN_MPI_DFT(P, &padt, apply); pln->cld1 = cld1; pln->cld2 = cld2; pln->preserve_input = ego->preserve_input ? 2 : NO_DESTROY_INPUTP(plnr); pln->roff = ri - p->I; pln->ioff = ii - p->I; X(ops_add)(&cld1->ops, &cld2->ops, &pln->super.super.ops); return &(pln->super.super); nada: X(plan_destroy_internal)(cld2); X(plan_destroy_internal)(cld1); return (plan *) 0; } static solver *mksolver(int preserve_input) { static const solver_adt sadt = { PROBLEM_MPI_DFT, mkplan, 0 }; S *slv = MKSOLVER(S, &sadt); slv->preserve_input = preserve_input; return &(slv->super); } void XM(dft_rank_geq2_register)(planner *p) { int preserve_input; for (preserve_input = 0; preserve_input <= 1; ++preserve_input) REGISTER_SOLVER(p, mksolver(preserve_input)); } fftw-3.3.3/mpi/dft-rank-geq2-transposed.c0000644000175000017500000001536012054410050015056 00000000000000/* * Copyright (c) 2003, 2007-11 Matteo Frigo * Copyright (c) 2003, 2007-11 Massachusetts Institute of Technology * * This program is free software; you can redistribute it and/or modify * it under the terms of the GNU General Public License as published by * the Free Software Foundation; either version 2 of the License, or * (at your option) any later version. * * This program is distributed in the hope that it will be useful, * but WITHOUT ANY WARRANTY; without even the implied warranty of * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the * GNU General Public License for more details. * * You should have received a copy of the GNU General Public License * along with this program; if not, write to the Free Software * Foundation, Inc., 51 Franklin Street, Fifth Floor, Boston, MA 02110-1301 USA * */ /* Complex DFTs of rank >= 2, for the case where we are distributed across the first dimension only, and the output is transposed both in data distribution and in ordering (for the first 2 dimensions). (Note that we don't have to handle the case where the input is transposed, since this is equivalent to transposed output with the first two dimensions swapped, and is automatically canonicalized as such by dft-problem.c. */ #include "mpi-dft.h" #include "mpi-transpose.h" #include "dft.h" typedef struct { solver super; int preserve_input; /* preserve input even if DESTROY_INPUT was passed */ } S; typedef struct { plan_mpi_dft super; plan *cld1, *cldt, *cld2; INT roff, ioff; int preserve_input; } P; static void apply(const plan *ego_, R *I, R *O) { const P *ego = (const P *) ego_; plan_dft *cld1, *cld2; plan_rdft *cldt; INT roff = ego->roff, ioff = ego->ioff; /* DFT local dimensions */ cld1 = (plan_dft *) ego->cld1; if (ego->preserve_input) { cld1->apply(ego->cld1, I+roff, I+ioff, O+roff, O+ioff); I = O; } else cld1->apply(ego->cld1, I+roff, I+ioff, I+roff, I+ioff); /* global transpose */ cldt = (plan_rdft *) ego->cldt; cldt->apply(ego->cldt, I, O); /* DFT final local dimension */ cld2 = (plan_dft *) ego->cld2; cld2->apply(ego->cld2, O+roff, O+ioff, O+roff, O+ioff); } static int applicable(const S *ego, const problem *p_, const planner *plnr) { const problem_mpi_dft *p = (const problem_mpi_dft *) p_; return (1 && p->sz->rnk > 1 && p->flags == TRANSPOSED_OUT && (!ego->preserve_input || (!NO_DESTROY_INPUTP(plnr) && p->I != p->O)) && XM(is_local_after)(1, p->sz, IB) && XM(is_local_after)(2, p->sz, OB) && XM(num_blocks)(p->sz->dims[0].n, p->sz->dims[0].b[OB]) == 1 && (!NO_SLOWP(plnr) /* slow if dft-serial is applicable */ || !XM(dft_serial_applicable)(p)) ); } static void awake(plan *ego_, enum wakefulness wakefulness) { P *ego = (P *) ego_; X(plan_awake)(ego->cld1, wakefulness); X(plan_awake)(ego->cldt, wakefulness); X(plan_awake)(ego->cld2, wakefulness); } static void destroy(plan *ego_) { P *ego = (P *) ego_; X(plan_destroy_internal)(ego->cld2); X(plan_destroy_internal)(ego->cldt); X(plan_destroy_internal)(ego->cld1); } static void print(const plan *ego_, printer *p) { const P *ego = (const P *) ego_; p->print(p, "(mpi-dft-rank-geq2-transposed%s%(%p%)%(%p%)%(%p%))", ego->preserve_input==2 ?"/p":"", ego->cld1, ego->cldt, ego->cld2); } static plan *mkplan(const solver *ego_, const problem *p_, planner *plnr) { const S *ego = (const S *) ego_; const problem_mpi_dft *p; P *pln; plan *cld1 = 0, *cldt = 0, *cld2 = 0; R *ri, *ii, *ro, *io, *I, *O; tensor *sz; int i, my_pe, n_pes; INT nrest; static const plan_adt padt = { XM(dft_solve), awake, print, destroy }; UNUSED(ego); if (!applicable(ego, p_, plnr)) return (plan *) 0; p = (const problem_mpi_dft *) p_; X(extract_reim)(p->sign, I = p->I, &ri, &ii); X(extract_reim)(p->sign, O = p->O, &ro, &io); if (ego->preserve_input || NO_DESTROY_INPUTP(plnr)) I = O; else { ro = ri; io = ii; } MPI_Comm_rank(p->comm, &my_pe); MPI_Comm_size(p->comm, &n_pes); sz = X(mktensor)(p->sz->rnk - 1); /* tensor of last rnk-1 dimensions */ i = p->sz->rnk - 2; A(i >= 0); sz->dims[i].n = p->sz->dims[i+1].n; sz->dims[i].is = sz->dims[i].os = 2 * p->vn; for (--i; i >= 0; --i) { sz->dims[i].n = p->sz->dims[i+1].n; sz->dims[i].is = sz->dims[i].os = sz->dims[i+1].n * sz->dims[i+1].is; } nrest = 1; for (i = 1; i < sz->rnk; ++i) nrest *= sz->dims[i].n; { INT is = sz->dims[0].n * sz->dims[0].is; INT b = XM(block)(p->sz->dims[0].n, p->sz->dims[0].b[IB], my_pe); cld1 = X(mkplan_d)(plnr, X(mkproblem_dft_d)(sz, X(mktensor_2d)(b, is, is, p->vn, 2, 2), ri, ii, ro, io)); if (XM(any_true)(!cld1, p->comm)) goto nada; } nrest *= p->vn; cldt = X(mkplan_d)(plnr, XM(mkproblem_transpose)( p->sz->dims[0].n, p->sz->dims[1].n, nrest * 2, I, O, p->sz->dims[0].b[IB], p->sz->dims[1].b[OB], p->comm, 0)); if (XM(any_true)(!cldt, p->comm)) goto nada; X(extract_reim)(p->sign, O, &ro, &io); { INT is = p->sz->dims[0].n * nrest * 2; INT b = XM(block)(p->sz->dims[1].n, p->sz->dims[1].b[OB], my_pe); cld2 = X(mkplan_d)(plnr, X(mkproblem_dft_d)(X(mktensor_1d)( p->sz->dims[0].n, nrest * 2, nrest * 2), X(mktensor_2d)(b, is, is, nrest, 2, 2), ro, io, ro, io)); if (XM(any_true)(!cld2, p->comm)) goto nada; } pln = MKPLAN_MPI_DFT(P, &padt, apply); pln->cld1 = cld1; pln->cldt = cldt; pln->cld2 = cld2; pln->preserve_input = ego->preserve_input ? 2 : NO_DESTROY_INPUTP(plnr); pln->roff = ri - p->I; pln->ioff = ii - p->I; X(ops_add)(&cld1->ops, &cld2->ops, &pln->super.super.ops); X(ops_add2)(&cldt->ops, &pln->super.super.ops); return &(pln->super.super); nada: X(plan_destroy_internal)(cld2); X(plan_destroy_internal)(cldt); X(plan_destroy_internal)(cld1); return (plan *) 0; } static solver *mksolver(int preserve_input) { static const solver_adt sadt = { PROBLEM_MPI_DFT, mkplan, 0 }; S *slv = MKSOLVER(S, &sadt); slv->preserve_input = preserve_input; return &(slv->super); } void XM(dft_rank_geq2_transposed_register)(planner *p) { int preserve_input; for (preserve_input = 0; preserve_input <= 1; ++preserve_input) REGISTER_SOLVER(p, mksolver(preserve_input)); } fftw-3.3.3/mpi/dft-rank1.c0000644000175000017500000002620412054410050012122 00000000000000/* * Copyright (c) 2003, 2007-11 Matteo Frigo * Copyright (c) 2003, 2007-11 Massachusetts Institute of Technology * * This program is free software; you can redistribute it and/or modify * it under the terms of the GNU General Public License as published by * the Free Software Foundation; either version 2 of the License, or * (at your option) any later version. * * This program is distributed in the hope that it will be useful, * but WITHOUT ANY WARRANTY; without even the implied warranty of * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the * GNU General Public License for more details. * * You should have received a copy of the GNU General Public License * along with this program; if not, write to the Free Software * Foundation, Inc., 51 Franklin Street, Fifth Floor, Boston, MA 02110-1301 USA * */ /* Complex DFTs of rank == 1 via six-step algorithm. */ #include "mpi-dft.h" #include "mpi-transpose.h" #include "dft.h" typedef struct { solver super; rdftapply apply; /* apply_ddft_first or apply_ddft_last */ int preserve_input; /* preserve input even if DESTROY_INPUT was passed */ } S; typedef struct { plan_mpi_dft super; triggen *t; plan *cldt, *cld_ddft, *cld_dft; INT roff, ioff; int preserve_input; INT vn, xmin, xmax, xs, m, r; } P; static void do_twiddle(triggen *t, INT ir, INT m, INT vn, R *xr, R *xi) { void (*rotate)(triggen *, INT, R, R, R *) = t->rotate; INT im, iv; for (im = 0; im < m; ++im) for (iv = 0; iv < vn; ++iv) { /* TODO: modify/inline rotate function so that it can do whole vn vector at once? */ R c[2]; rotate(t, ir * im, *xr, *xi, c); *xr = c[0]; *xi = c[1]; xr += 2; xi += 2; } } /* radix-r DFT of size r*m. This is equivalent to an m x r 2d DFT, plus twiddle factors between the size-m and size-r 1d DFTs, where the m dimension is initially distributed. The output is transposed to r x m where the r dimension is distributed. This algorithm follows the general sequence: global transpose (m x r -> r x m) DFTs of size m multiply by twiddles + global transpose (r x m -> m x r) DFTs of size r global transpose (m x r -> r x m) where the multiplication by twiddles can come before or after the middle transpose. The first/last transposes are omitted for SCRAMBLED_IN/OUT formats, respectively. However, we wish to exploit our dft-rank1-bigvec solver, which solves a vector of distributed DFTs via transpose+dft+transpose. Therefore, we can group *either* the DFTs of size m *or* the DFTs of size r with their surrounding transposes as a single distributed-DFT (ddft) plan. These two variations correspond to apply_ddft_first or apply_ddft_last, respectively. */ static void apply_ddft_first(const plan *ego_, R *I, R *O) { const P *ego = (const P *) ego_; plan_dft *cld_dft; plan_rdft *cldt, *cld_ddft; INT roff, ioff, im, mmax, ms, r, vn; triggen *t; R *dI, *dO; /* distributed size-m DFTs, with output in m x r format */ cld_ddft = (plan_rdft *) ego->cld_ddft; cld_ddft->apply(ego->cld_ddft, I, O); cldt = (plan_rdft *) ego->cldt; if (ego->preserve_input || !cldt) I = O; /* twiddle multiplications, followed by 1d DFTs of size-r */ cld_dft = (plan_dft *) ego->cld_dft; roff = ego->roff; ioff = ego->ioff; mmax = ego->xmax; ms = ego->xs; t = ego->t; r = ego->r; vn = ego->vn; dI = O; dO = I; for (im = ego->xmin; im <= mmax; ++im) { do_twiddle(t, im, r, vn, dI+roff, dI+ioff); cld_dft->apply((plan *) cld_dft, dI+roff, dI+ioff, dO+roff, dO+ioff); dI += ms; dO += ms; } /* final global transpose (m x r -> r x m), if not SCRAMBLED_OUT */ if (cldt) cldt->apply((plan *) cldt, I, O); } static void apply_ddft_last(const plan *ego_, R *I, R *O) { const P *ego = (const P *) ego_; plan_dft *cld_dft; plan_rdft *cldt, *cld_ddft; INT roff, ioff, ir, rmax, rs, m, vn; triggen *t; R *dI, *dO0, *dO; /* initial global transpose (m x r -> r x m), if not SCRAMBLED_IN */ cldt = (plan_rdft *) ego->cldt; if (cldt) { cldt->apply((plan *) cldt, I, O); dI = O; } else dI = I; if (ego->preserve_input) dO = O; else dO = I; dO0 = dO; /* 1d DFTs of size m, followed by twiddle multiplications */ cld_dft = (plan_dft *) ego->cld_dft; roff = ego->roff; ioff = ego->ioff; rmax = ego->xmax; rs = ego->xs; t = ego->t; m = ego->m; vn = ego->vn; for (ir = ego->xmin; ir <= rmax; ++ir) { cld_dft->apply((plan *) cld_dft, dI+roff, dI+ioff, dO+roff, dO+ioff); do_twiddle(t, ir, m, vn, dO+roff, dO+ioff); dI += rs; dO += rs; } /* distributed size-r DFTs, with output in r x m format */ cld_ddft = (plan_rdft *) ego->cld_ddft; cld_ddft->apply(ego->cld_ddft, dO0, O); } static int applicable(const S *ego, const problem *p_, const planner *plnr, INT *r, INT rblock[2], INT mblock[2]) { const problem_mpi_dft *p = (const problem_mpi_dft *) p_; int n_pes; MPI_Comm_size(p->comm, &n_pes); return (1 && p->sz->rnk == 1 && ONLY_SCRAMBLEDP(p->flags) && (!ego->preserve_input || (!NO_DESTROY_INPUTP(plnr) && p->I != p->O)) && (!(p->flags & SCRAMBLED_IN) || ego->apply == apply_ddft_last) && (!(p->flags & SCRAMBLED_OUT) || ego->apply == apply_ddft_first) && (!NO_SLOWP(plnr) /* slow if dft-serial is applicable */ || !XM(dft_serial_applicable)(p)) /* disallow if dft-rank1-bigvec is applicable since the data distribution may be slightly different (ugh!) */ && (p->vn < n_pes || p->flags) && (*r = XM(choose_radix)(p->sz->dims[0], n_pes, p->flags, p->sign, rblock, mblock)) /* ddft_first or last has substantial advantages in the bigvec transpositions for the common case where n_pes == n/r or r, respectively */ && (!NO_UGLYP(plnr) || !(*r == n_pes && ego->apply == apply_ddft_first) || !(p->sz->dims[0].n / *r == n_pes && ego->apply == apply_ddft_last)) ); } static void awake(plan *ego_, enum wakefulness wakefulness) { P *ego = (P *) ego_; X(plan_awake)(ego->cldt, wakefulness); X(plan_awake)(ego->cld_dft, wakefulness); X(plan_awake)(ego->cld_ddft, wakefulness); switch (wakefulness) { case SLEEPY: X(triggen_destroy)(ego->t); ego->t = 0; break; default: ego->t = X(mktriggen)(AWAKE_SQRTN_TABLE, ego->r * ego->m); break; } } static void destroy(plan *ego_) { P *ego = (P *) ego_; X(plan_destroy_internal)(ego->cldt); X(plan_destroy_internal)(ego->cld_dft); X(plan_destroy_internal)(ego->cld_ddft); } static void print(const plan *ego_, printer *p) { const P *ego = (const P *) ego_; p->print(p, "(mpi-dft-rank1/%D%s%s%(%p%)%(%p%)%(%p%))", ego->r, ego->super.apply == apply_ddft_first ? "/first" : "/last", ego->preserve_input==2 ?"/p":"", ego->cld_ddft, ego->cld_dft, ego->cldt); } static plan *mkplan(const solver *ego_, const problem *p_, planner *plnr) { const S *ego = (const S *) ego_; const problem_mpi_dft *p; P *pln; plan *cld_dft = 0, *cld_ddft = 0, *cldt = 0; R *ri, *ii, *ro, *io, *I, *O; INT r, rblock[2], m, mblock[2], rp, mp, mpblock[2], mpb; int my_pe, n_pes, preserve_input, ddft_first; dtensor *sz; static const plan_adt padt = { XM(dft_solve), awake, print, destroy }; UNUSED(ego); if (!applicable(ego, p_, plnr, &r, rblock, mblock)) return (plan *) 0; p = (const problem_mpi_dft *) p_; MPI_Comm_rank(p->comm, &my_pe); MPI_Comm_size(p->comm, &n_pes); m = p->sz->dims[0].n / r; /* some hackery so that we can plan both ddft_first and ddft_last as if they were ddft_first */ if ((ddft_first = (ego->apply == apply_ddft_first))) { rp = r; mp = m; mpblock[IB] = mblock[IB]; mpblock[OB] = mblock[OB]; mpb = XM(block)(mp, mpblock[OB], my_pe); } else { rp = m; mp = r; mpblock[IB] = rblock[IB]; mpblock[OB] = rblock[OB]; mpb = XM(block)(mp, mpblock[IB], my_pe); } preserve_input = ego->preserve_input ? 2 : NO_DESTROY_INPUTP(plnr); sz = XM(mkdtensor)(1); sz->dims[0].n = mp; sz->dims[0].b[IB] = mpblock[IB]; sz->dims[0].b[OB] = mpblock[OB]; I = (ddft_first || !preserve_input) ? p->I : p->O; O = p->O; cld_ddft = X(mkplan_d)(plnr, XM(mkproblem_dft_d)(sz, rp * p->vn, I, O, p->comm, p->sign, RANK1_BIGVEC_ONLY)); if (XM(any_true)(!cld_ddft, p->comm)) goto nada; I = TAINT((ddft_first || !p->flags) ? p->O : p->I, rp * p->vn * 2); O = TAINT((preserve_input || (ddft_first && p->flags)) ? p->O : p->I, rp * p->vn * 2); X(extract_reim)(p->sign, I, &ri, &ii); X(extract_reim)(p->sign, O, &ro, &io); cld_dft = X(mkplan_d)(plnr, X(mkproblem_dft_d)(X(mktensor_1d)(rp, p->vn*2,p->vn*2), X(mktensor_1d)(p->vn, 2, 2), ri, ii, ro, io)); if (XM(any_true)(!cld_dft, p->comm)) goto nada; if (!p->flags) { /* !(SCRAMBLED_IN or SCRAMBLED_OUT) */ I = (ddft_first && preserve_input) ? p->O : p->I; O = p->O; cldt = X(mkplan_d)(plnr, XM(mkproblem_transpose)( m, r, p->vn * 2, I, O, ddft_first ? mblock[OB] : mblock[IB], ddft_first ? rblock[OB] : rblock[IB], p->comm, 0)); if (XM(any_true)(!cldt, p->comm)) goto nada; } pln = MKPLAN_MPI_DFT(P, &padt, ego->apply); pln->cld_ddft = cld_ddft; pln->cld_dft = cld_dft; pln->cldt = cldt; pln->preserve_input = preserve_input; X(extract_reim)(p->sign, p->O, &ro, &io); pln->roff = ro - p->O; pln->ioff = io - p->O; pln->vn = p->vn; pln->m = m; pln->r = r; pln->xmin = (ddft_first ? mblock[OB] : rblock[IB]) * my_pe; pln->xmax = pln->xmin + mpb - 1; pln->xs = rp * p->vn * 2; pln->t = 0; X(ops_add)(&cld_ddft->ops, &cld_dft->ops, &pln->super.super.ops); if (cldt) X(ops_add2)(&cldt->ops, &pln->super.super.ops); { double n0 = (1 + pln->xmax - pln->xmin) * (mp - 1) * pln->vn; pln->super.super.ops.mul += 8 * n0; pln->super.super.ops.add += 4 * n0; pln->super.super.ops.other += 8 * n0; } return &(pln->super.super); nada: X(plan_destroy_internal)(cldt); X(plan_destroy_internal)(cld_dft); X(plan_destroy_internal)(cld_ddft); return (plan *) 0; } static solver *mksolver(rdftapply apply, int preserve_input) { static const solver_adt sadt = { PROBLEM_MPI_DFT, mkplan, 0 }; S *slv = MKSOLVER(S, &sadt); slv->apply = apply; slv->preserve_input = preserve_input; return &(slv->super); } void XM(dft_rank1_register)(planner *p) { rdftapply apply[] = { apply_ddft_first, apply_ddft_last }; unsigned int iapply; int preserve_input; for (iapply = 0; iapply < sizeof(apply) / sizeof(apply[0]); ++iapply) for (preserve_input = 0; preserve_input <= 1; ++preserve_input) REGISTER_SOLVER(p, mksolver(apply[iapply], preserve_input)); } fftw-3.3.3/mpi/dft-rank1-bigvec.c0000644000175000017500000001457112054410050013363 00000000000000/* * Copyright (c) 2003, 2007-11 Matteo Frigo * Copyright (c) 2003, 2007-11 Massachusetts Institute of Technology * * This program is free software; you can redistribute it and/or modify * it under the terms of the GNU General Public License as published by * the Free Software Foundation; either version 2 of the License, or * (at your option) any later version. * * This program is distributed in the hope that it will be useful, * but WITHOUT ANY WARRANTY; without even the implied warranty of * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the * GNU General Public License for more details. * * You should have received a copy of the GNU General Public License * along with this program; if not, write to the Free Software * Foundation, Inc., 51 Franklin Street, Fifth Floor, Boston, MA 02110-1301 USA * */ /* Complex DFTs of rank == 1 when the vector length vn is >= # processes. In this case, we don't need to use a six-step type algorithm, and can instead transpose the DFT dimension with the vector dimension to make the DFT local. */ #include "mpi-dft.h" #include "mpi-transpose.h" #include "dft.h" typedef struct { solver super; int preserve_input; /* preserve input even if DESTROY_INPUT was passed */ rearrangement rearrange; } S; typedef struct { plan_mpi_dft super; plan *cldt_before, *cld, *cldt_after; INT roff, ioff; int preserve_input; rearrangement rearrange; } P; static void apply(const plan *ego_, R *I, R *O) { const P *ego = (const P *) ego_; plan_dft *cld; plan_rdft *cldt_before, *cldt_after; INT roff = ego->roff, ioff = ego->ioff; /* global transpose */ cldt_before = (plan_rdft *) ego->cldt_before; cldt_before->apply(ego->cldt_before, I, O); if (ego->preserve_input) I = O; /* 1d DFT(s) */ cld = (plan_dft *) ego->cld; cld->apply(ego->cld, O+roff, O+ioff, I+roff, I+ioff); /* global transpose */ cldt_after = (plan_rdft *) ego->cldt_after; cldt_after->apply(ego->cldt_after, I, O); } static int applicable(const S *ego, const problem *p_, const planner *plnr) { const problem_mpi_dft *p = (const problem_mpi_dft *) p_; int n_pes; MPI_Comm_size(p->comm, &n_pes); return (1 && p->sz->rnk == 1 && !(p->flags & ~RANK1_BIGVEC_ONLY) && (!ego->preserve_input || (!NO_DESTROY_INPUTP(plnr) && p->I != p->O)) && (p->vn >= n_pes /* TODO: relax this, using more memory? */ || (p->flags & RANK1_BIGVEC_ONLY)) && XM(rearrange_applicable)(ego->rearrange, p->sz->dims[0], p->vn, n_pes) && (!NO_SLOWP(plnr) /* slow if dft-serial is applicable */ || !XM(dft_serial_applicable)(p)) ); } static void awake(plan *ego_, enum wakefulness wakefulness) { P *ego = (P *) ego_; X(plan_awake)(ego->cldt_before, wakefulness); X(plan_awake)(ego->cld, wakefulness); X(plan_awake)(ego->cldt_after, wakefulness); } static void destroy(plan *ego_) { P *ego = (P *) ego_; X(plan_destroy_internal)(ego->cldt_after); X(plan_destroy_internal)(ego->cld); X(plan_destroy_internal)(ego->cldt_before); } static void print(const plan *ego_, printer *p) { const P *ego = (const P *) ego_; const char descrip[][16] = { "contig", "discontig", "square-after", "square-middle", "square-before" }; p->print(p, "(mpi-dft-rank1-bigvec/%s%s %(%p%) %(%p%) %(%p%))", descrip[ego->rearrange], ego->preserve_input==2 ?"/p":"", ego->cldt_before, ego->cld, ego->cldt_after); } static plan *mkplan(const solver *ego_, const problem *p_, planner *plnr) { const S *ego = (const S *) ego_; const problem_mpi_dft *p; P *pln; plan *cld = 0, *cldt_before = 0, *cldt_after = 0; R *ri, *ii, *ro, *io, *I, *O; INT yblock, yb, nx, ny, vn; int my_pe, n_pes; static const plan_adt padt = { XM(dft_solve), awake, print, destroy }; UNUSED(ego); if (!applicable(ego, p_, plnr)) return (plan *) 0; p = (const problem_mpi_dft *) p_; MPI_Comm_rank(p->comm, &my_pe); MPI_Comm_size(p->comm, &n_pes); nx = p->sz->dims[0].n; if (!(ny = XM(rearrange_ny)(ego->rearrange, p->sz->dims[0],p->vn,n_pes))) return (plan *) 0; vn = p->vn / ny; A(ny * vn == p->vn); yblock = XM(default_block)(ny, n_pes); cldt_before = X(mkplan_d)(plnr, XM(mkproblem_transpose)( nx, ny, vn*2, I = p->I, O = p->O, p->sz->dims[0].b[IB], yblock, p->comm, 0)); if (XM(any_true)(!cldt_before, p->comm)) goto nada; if (ego->preserve_input || NO_DESTROY_INPUTP(plnr)) { I = O; } X(extract_reim)(p->sign, I, &ri, &ii); X(extract_reim)(p->sign, O, &ro, &io); yb = XM(block)(ny, yblock, my_pe); cld = X(mkplan_d)(plnr, X(mkproblem_dft_d)(X(mktensor_1d)(nx, vn*2, vn*2), X(mktensor_2d)(yb, vn*2*nx, vn*2*nx, vn, 2, 2), ro, io, ri, ii)); if (XM(any_true)(!cld, p->comm)) goto nada; cldt_after = X(mkplan_d)(plnr, XM(mkproblem_transpose)( ny, nx, vn*2, I, O, yblock, p->sz->dims[0].b[OB], p->comm, 0)); if (XM(any_true)(!cldt_after, p->comm)) goto nada; pln = MKPLAN_MPI_DFT(P, &padt, apply); pln->cldt_before = cldt_before; pln->cld = cld; pln->cldt_after = cldt_after; pln->preserve_input = ego->preserve_input ? 2 : NO_DESTROY_INPUTP(plnr); pln->roff = ro - p->O; pln->ioff = io - p->O; pln->rearrange = ego->rearrange; X(ops_add)(&cldt_before->ops, &cld->ops, &pln->super.super.ops); X(ops_add2)(&cldt_after->ops, &pln->super.super.ops); return &(pln->super.super); nada: X(plan_destroy_internal)(cldt_after); X(plan_destroy_internal)(cld); X(plan_destroy_internal)(cldt_before); return (plan *) 0; } static solver *mksolver(rearrangement rearrange, int preserve_input) { static const solver_adt sadt = { PROBLEM_MPI_DFT, mkplan, 0 }; S *slv = MKSOLVER(S, &sadt); slv->rearrange = rearrange; slv->preserve_input = preserve_input; return &(slv->super); } void XM(dft_rank1_bigvec_register)(planner *p) { rearrangement rearrange; int preserve_input; FORALL_REARRANGE(rearrange) for (preserve_input = 0; preserve_input <= 1; ++preserve_input) REGISTER_SOLVER(p, mksolver(rearrange, preserve_input)); } fftw-3.3.3/mpi/dft-problem.c0000644000175000017500000000757712054410050012562 00000000000000/* * Copyright (c) 2003, 2007-11 Matteo Frigo * Copyright (c) 2003, 2007-11 Massachusetts Institute of Technology * * This program is free software; you can redistribute it and/or modify * it under the terms of the GNU General Public License as published by * the Free Software Foundation; either version 2 of the License, or * (at your option) any later version. * * This program is distributed in the hope that it will be useful, * but WITHOUT ANY WARRANTY; without even the implied warranty of * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the * GNU General Public License for more details. * * You should have received a copy of the GNU General Public License * along with this program; if not, write to the Free Software * Foundation, Inc., 51 Franklin Street, Fifth Floor, Boston, MA 02110-1301 USA * */ #include "mpi-dft.h" static void destroy(problem *ego_) { problem_mpi_dft *ego = (problem_mpi_dft *) ego_; XM(dtensor_destroy)(ego->sz); MPI_Comm_free(&ego->comm); X(ifree)(ego_); } static void hash(const problem *p_, md5 *m) { const problem_mpi_dft *p = (const problem_mpi_dft *) p_; int i; X(md5puts)(m, "mpi-dft"); X(md5int)(m, p->I == p->O); /* don't include alignment -- may differ between processes X(md5int)(m, X(alignment_of)(p->I)); X(md5int)(m, X(alignment_of)(p->O)); ... note that applicability of MPI plans does not depend on alignment (although optimality may, in principle). */ XM(dtensor_md5)(m, p->sz); X(md5INT)(m, p->vn); X(md5int)(m, p->sign); X(md5int)(m, p->flags); MPI_Comm_size(p->comm, &i); X(md5int)(m, i); A(XM(md5_equal)(*m, p->comm)); } static void print(const problem *ego_, printer *p) { const problem_mpi_dft *ego = (const problem_mpi_dft *) ego_; int i; p->print(p, "(mpi-dft %d %d %d ", ego->I == ego->O, X(alignment_of)(ego->I), X(alignment_of)(ego->O)); XM(dtensor_print)(ego->sz, p); p->print(p, " %D %d %d", ego->vn, ego->sign, ego->flags); MPI_Comm_size(ego->comm, &i); p->print(p, " %d)", i); } static void zero(const problem *ego_) { const problem_mpi_dft *ego = (const problem_mpi_dft *) ego_; R *I = ego->I; INT i, N; int my_pe; MPI_Comm_rank(ego->comm, &my_pe); N = 2 * ego->vn * XM(total_block)(ego->sz, IB, my_pe); for (i = 0; i < N; ++i) I[i] = K(0.0); } static const problem_adt padt = { PROBLEM_MPI_DFT, hash, zero, print, destroy }; problem *XM(mkproblem_dft)(const dtensor *sz, INT vn, R *I, R *O, MPI_Comm comm, int sign, unsigned flags) { problem_mpi_dft *ego = (problem_mpi_dft *)X(mkproblem)(sizeof(problem_mpi_dft), &padt); int n_pes; A(XM(dtensor_validp)(sz) && FINITE_RNK(sz->rnk)); MPI_Comm_size(comm, &n_pes); A(n_pes >= XM(num_blocks_total)(sz, IB) && n_pes >= XM(num_blocks_total)(sz, OB)); A(vn >= 0); A(sign == -1 || sign == 1); /* enforce pointer equality if untainted pointers are equal */ if (UNTAINT(I) == UNTAINT(O)) I = O = JOIN_TAINT(I, O); ego->sz = XM(dtensor_canonical)(sz, 1); ego->vn = vn; ego->I = I; ego->O = O; ego->sign = sign; /* canonicalize: replace TRANSPOSED_IN with TRANSPOSED_OUT by swapping the first two dimensions (for rnk > 1) */ if ((flags & TRANSPOSED_IN) && ego->sz->rnk > 1) { ddim dim0 = ego->sz->dims[0]; ego->sz->dims[0] = ego->sz->dims[1]; ego->sz->dims[1] = dim0; flags &= ~TRANSPOSED_IN; flags ^= TRANSPOSED_OUT; } ego->flags = flags; MPI_Comm_dup(comm, &ego->comm); return &(ego->super); } problem *XM(mkproblem_dft_d)(dtensor *sz, INT vn, R *I, R *O, MPI_Comm comm, int sign, unsigned flags) { problem *p = XM(mkproblem_dft)(sz, vn, I, O, comm, sign, flags); XM(dtensor_destroy)(sz); return p; } fftw-3.3.3/mpi/dft-solve.c0000644000175000017500000000220312054410050012227 00000000000000/* * Copyright (c) 2003, 2007-11 Matteo Frigo * Copyright (c) 2003, 2007-11 Massachusetts Institute of Technology * * This program is free software; you can redistribute it and/or modify * it under the terms of the GNU General Public License as published by * the Free Software Foundation; either version 2 of the License, or * (at your option) any later version. * * This program is distributed in the hope that it will be useful, * but WITHOUT ANY WARRANTY; without even the implied warranty of * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the * GNU General Public License for more details. * * You should have received a copy of the GNU General Public License * along with this program; if not, write to the Free Software * Foundation, Inc., 51 Franklin Street, Fifth Floor, Boston, MA 02110-1301 USA * */ #include "mpi-dft.h" /* use the apply() operation for MPI_DFT problems */ void XM(dft_solve)(const plan *ego_, const problem *p_) { const plan_mpi_dft *ego = (const plan_mpi_dft *) ego_; const problem_mpi_dft *p = (const problem_mpi_dft *) p_; ego->apply(ego_, UNTAINT(p->I), UNTAINT(p->O)); } fftw-3.3.3/mpi/mpi-dft.h0000644000175000017500000000404012054410050011672 00000000000000/* * Copyright (c) 2003, 2007-11 Matteo Frigo * Copyright (c) 2003, 2007-11 Massachusetts Institute of Technology * * This program is free software; you can redistribute it and/or modify * it under the terms of the GNU General Public License as published by * the Free Software Foundation; either version 2 of the License, or * (at your option) any later version. * * This program is distributed in the hope that it will be useful, * but WITHOUT ANY WARRANTY; without even the implied warranty of * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the * GNU General Public License for more details. * * You should have received a copy of the GNU General Public License * along with this program; if not, write to the Free Software * Foundation, Inc., 51 Franklin Street, Fifth Floor, Boston, MA 02110-1301 USA * */ #include "ifftw-mpi.h" /* problem.c: */ typedef struct { problem super; dtensor *sz; INT vn; /* vector length (vector stride 1) */ R *I, *O; /* contiguous interleaved arrays */ int sign; /* FFTW_FORWARD / FFTW_BACKWARD */ unsigned flags; /* TRANSPOSED_IN/OUT meaningful for rnk>1 only SCRAMBLED_IN/OUT meaningful for 1d transforms only */ MPI_Comm comm; } problem_mpi_dft; problem *XM(mkproblem_dft)(const dtensor *sz, INT vn, R *I, R *O, MPI_Comm comm, int sign, unsigned flags); problem *XM(mkproblem_dft_d)(dtensor *sz, INT vn, R *I, R *O, MPI_Comm comm, int sign, unsigned flags); /* solve.c: */ void XM(dft_solve)(const plan *ego_, const problem *p_); /* plans have same operands as rdft plans, so just re-use */ typedef plan_rdft plan_mpi_dft; #define MKPLAN_MPI_DFT(type, adt, apply) \ (type *)X(mkplan_rdft)(sizeof(type), adt, apply) int XM(dft_serial_applicable)(const problem_mpi_dft *p); /* various solvers */ void XM(dft_rank_geq2_register)(planner *p); void XM(dft_rank_geq2_transposed_register)(planner *p); void XM(dft_serial_register)(planner *p); void XM(dft_rank1_bigvec_register)(planner *p); void XM(dft_rank1_register)(planner *p); fftw-3.3.3/mpi/rdft-serial.c0000644000175000017500000000670712054410050012555 00000000000000/* * Copyright (c) 2003, 2007-11 Matteo Frigo * Copyright (c) 2003, 2007-11 Massachusetts Institute of Technology * * This program is free software; you can redistribute it and/or modify * it under the terms of the GNU General Public License as published by * the Free Software Foundation; either version 2 of the License, or * (at your option) any later version. * * This program is distributed in the hope that it will be useful, * but WITHOUT ANY WARRANTY; without even the implied warranty of * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the * GNU General Public License for more details. * * You should have received a copy of the GNU General Public License * along with this program; if not, write to the Free Software * Foundation, Inc., 51 Franklin Street, Fifth Floor, Boston, MA 02110-1301 USA * */ /* "MPI" RDFTs where all of the data is on one processor...just call through to serial API. */ #include "mpi-rdft.h" typedef struct { plan_mpi_rdft super; plan *cld; } P; static void apply(const plan *ego_, R *I, R *O) { const P *ego = (const P *) ego_; plan_rdft *cld = (plan_rdft *) ego->cld; cld->apply(ego->cld, I, O); } static void awake(plan *ego_, enum wakefulness wakefulness) { P *ego = (P *) ego_; X(plan_awake)(ego->cld, wakefulness); } static void destroy(plan *ego_) { P *ego = (P *) ego_; X(plan_destroy_internal)(ego->cld); } static void print(const plan *ego_, printer *p) { const P *ego = (const P *) ego_; p->print(p, "(mpi-rdft-serial %(%p%))", ego->cld); } int XM(rdft_serial_applicable)(const problem_mpi_rdft *p) { return (1 && p->flags == 0 /* TRANSPOSED/SCRAMBLED_IN/OUT not supported */ && ((XM(is_local)(p->sz, IB) && XM(is_local)(p->sz, OB)) || p->vn == 0)); } static plan *mkplan(const solver *ego, const problem *p_, planner *plnr) { const problem_mpi_rdft *p = (const problem_mpi_rdft *) p_; P *pln; plan *cld; int my_pe; static const plan_adt padt = { XM(rdft_solve), awake, print, destroy }; UNUSED(ego); /* check whether applicable: */ if (!XM(rdft_serial_applicable)(p)) return (plan *) 0; MPI_Comm_rank(p->comm, &my_pe); if (my_pe == 0 && p->vn > 0) { int i, rnk = p->sz->rnk; tensor *sz = X(mktensor)(rnk); rdft_kind *kind = (rdft_kind *) MALLOC(sizeof(rdft_kind) * rnk, PROBLEMS); sz->dims[rnk - 1].is = sz->dims[rnk - 1].os = p->vn; sz->dims[rnk - 1].n = p->sz->dims[rnk - 1].n; for (i = rnk - 1; i > 0; --i) { sz->dims[i - 1].is = sz->dims[i - 1].os = sz->dims[i].is * sz->dims[i].n; sz->dims[i - 1].n = p->sz->dims[i - 1].n; } for (i = 0; i < rnk; ++i) kind[i] = p->kind[i]; cld = X(mkplan_d)(plnr, X(mkproblem_rdft_d)(sz, X(mktensor_1d)(p->vn, 1, 1), p->I, p->O, kind)); X(ifree0)(kind); } else { /* idle process: make nop plan */ cld = X(mkplan_d)(plnr, X(mkproblem_rdft_0_d)(X(mktensor_1d)(0,0,0), p->I, p->O)); } if (XM(any_true)(!cld, p->comm)) return (plan *) 0; pln = MKPLAN_MPI_RDFT(P, &padt, apply); pln->cld = cld; X(ops_cpy)(&cld->ops, &pln->super.super.ops); return &(pln->super.super); } static solver *mksolver(void) { static const solver_adt sadt = { PROBLEM_MPI_RDFT, mkplan, 0 }; return MKSOLVER(solver, &sadt); } void XM(rdft_serial_register)(planner *p) { REGISTER_SOLVER(p, mksolver()); } fftw-3.3.3/mpi/rdft-rank-geq2.c0000644000175000017500000001225512054410050013060 00000000000000/* * Copyright (c) 2003, 2007-11 Matteo Frigo * Copyright (c) 2003, 2007-11 Massachusetts Institute of Technology * * This program is free software; you can redistribute it and/or modify * it under the terms of the GNU General Public License as published by * the Free Software Foundation; either version 2 of the License, or * (at your option) any later version. * * This program is distributed in the hope that it will be useful, * but WITHOUT ANY WARRANTY; without even the implied warranty of * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the * GNU General Public License for more details. * * You should have received a copy of the GNU General Public License * along with this program; if not, write to the Free Software * Foundation, Inc., 51 Franklin Street, Fifth Floor, Boston, MA 02110-1301 USA * */ /* Complex RDFTs of rank >= 2, for the case where we are distributed across the first dimension only, and the output is not transposed. */ #include "mpi-rdft.h" typedef struct { solver super; int preserve_input; /* preserve input even if DESTROY_INPUT was passed */ } S; typedef struct { plan_mpi_rdft super; plan *cld1, *cld2; int preserve_input; } P; static void apply(const plan *ego_, R *I, R *O) { const P *ego = (const P *) ego_; plan_rdft *cld1, *cld2; /* RDFT local dimensions */ cld1 = (plan_rdft *) ego->cld1; if (ego->preserve_input) { cld1->apply(ego->cld1, I, O); I = O; } else cld1->apply(ego->cld1, I, I); /* RDFT non-local dimension (via rdft-rank1-bigvec, usually): */ cld2 = (plan_rdft *) ego->cld2; cld2->apply(ego->cld2, I, O); } static int applicable(const S *ego, const problem *p_, const planner *plnr) { const problem_mpi_rdft *p = (const problem_mpi_rdft *) p_; return (1 && p->sz->rnk > 1 && p->flags == 0 /* TRANSPOSED/SCRAMBLED_IN/OUT not supported */ && (!ego->preserve_input || (!NO_DESTROY_INPUTP(plnr) && p->I != p->O)) && XM(is_local_after)(1, p->sz, IB) && XM(is_local_after)(1, p->sz, OB) && (!NO_SLOWP(plnr) /* slow if rdft-serial is applicable */ || !XM(rdft_serial_applicable)(p)) ); } static void awake(plan *ego_, enum wakefulness wakefulness) { P *ego = (P *) ego_; X(plan_awake)(ego->cld1, wakefulness); X(plan_awake)(ego->cld2, wakefulness); } static void destroy(plan *ego_) { P *ego = (P *) ego_; X(plan_destroy_internal)(ego->cld2); X(plan_destroy_internal)(ego->cld1); } static void print(const plan *ego_, printer *p) { const P *ego = (const P *) ego_; p->print(p, "(mpi-rdft-rank-geq2%s%(%p%)%(%p%))", ego->preserve_input==2 ?"/p":"", ego->cld1, ego->cld2); } static plan *mkplan(const solver *ego_, const problem *p_, planner *plnr) { const S *ego = (const S *) ego_; const problem_mpi_rdft *p; P *pln; plan *cld1 = 0, *cld2 = 0; R *I, *O, *I2; tensor *sz; dtensor *sz2; int i, my_pe, n_pes; INT nrest; static const plan_adt padt = { XM(rdft_solve), awake, print, destroy }; UNUSED(ego); if (!applicable(ego, p_, plnr)) return (plan *) 0; p = (const problem_mpi_rdft *) p_; I2 = I = p->I; O = p->O; if (ego->preserve_input || NO_DESTROY_INPUTP(plnr)) I = O; MPI_Comm_rank(p->comm, &my_pe); MPI_Comm_size(p->comm, &n_pes); sz = X(mktensor)(p->sz->rnk - 1); /* tensor of last rnk-1 dimensions */ i = p->sz->rnk - 2; A(i >= 0); sz->dims[i].n = p->sz->dims[i+1].n; sz->dims[i].is = sz->dims[i].os = p->vn; for (--i; i >= 0; --i) { sz->dims[i].n = p->sz->dims[i+1].n; sz->dims[i].is = sz->dims[i].os = sz->dims[i+1].n * sz->dims[i+1].is; } nrest = X(tensor_sz)(sz); { INT is = sz->dims[0].n * sz->dims[0].is; INT b = XM(block)(p->sz->dims[0].n, p->sz->dims[0].b[IB], my_pe); cld1 = X(mkplan_d)(plnr, X(mkproblem_rdft_d)(sz, X(mktensor_2d)(b, is, is, p->vn, 1, 1), I2, I, p->kind + 1)); if (XM(any_true)(!cld1, p->comm)) goto nada; } sz2 = XM(mkdtensor)(1); /* tensor for first (distributed) dimension */ sz2->dims[0] = p->sz->dims[0]; cld2 = X(mkplan_d)(plnr, XM(mkproblem_rdft_d)(sz2, nrest * p->vn, I, O, p->comm, p->kind, RANK1_BIGVEC_ONLY)); if (XM(any_true)(!cld2, p->comm)) goto nada; pln = MKPLAN_MPI_RDFT(P, &padt, apply); pln->cld1 = cld1; pln->cld2 = cld2; pln->preserve_input = ego->preserve_input ? 2 : NO_DESTROY_INPUTP(plnr); X(ops_add)(&cld1->ops, &cld2->ops, &pln->super.super.ops); return &(pln->super.super); nada: X(plan_destroy_internal)(cld2); X(plan_destroy_internal)(cld1); return (plan *) 0; } static solver *mksolver(int preserve_input) { static const solver_adt sadt = { PROBLEM_MPI_RDFT, mkplan, 0 }; S *slv = MKSOLVER(S, &sadt); slv->preserve_input = preserve_input; return &(slv->super); } void XM(rdft_rank_geq2_register)(planner *p) { int preserve_input; for (preserve_input = 0; preserve_input <= 1; ++preserve_input) REGISTER_SOLVER(p, mksolver(preserve_input)); } fftw-3.3.3/mpi/rdft-rank-geq2-transposed.c0000644000175000017500000001435612054410050015244 00000000000000/* * Copyright (c) 2003, 2007-11 Matteo Frigo * Copyright (c) 2003, 2007-11 Massachusetts Institute of Technology * * This program is free software; you can redistribute it and/or modify * it under the terms of the GNU General Public License as published by * the Free Software Foundation; either version 2 of the License, or * (at your option) any later version. * * This program is distributed in the hope that it will be useful, * but WITHOUT ANY WARRANTY; without even the implied warranty of * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the * GNU General Public License for more details. * * You should have received a copy of the GNU General Public License * along with this program; if not, write to the Free Software * Foundation, Inc., 51 Franklin Street, Fifth Floor, Boston, MA 02110-1301 USA * */ /* Complex RDFTs of rank >= 2, for the case where we are distributed across the first dimension only, and the output is transposed both in data distribution and in ordering (for the first 2 dimensions). (Note that we don't have to handle the case where the input is transposed, since this is equivalent to transposed output with the first two dimensions swapped, and is automatically canonicalized as such by rdft-problem.c. */ #include "mpi-rdft.h" #include "mpi-transpose.h" typedef struct { solver super; int preserve_input; /* preserve input even if DESTROY_INPUT was passed */ } S; typedef struct { plan_mpi_rdft super; plan *cld1, *cldt, *cld2; INT roff, ioff; int preserve_input; } P; static void apply(const plan *ego_, R *I, R *O) { const P *ego = (const P *) ego_; plan_rdft *cld1, *cld2, *cldt; /* RDFT local dimensions */ cld1 = (plan_rdft *) ego->cld1; if (ego->preserve_input) { cld1->apply(ego->cld1, I, O); I = O; } else cld1->apply(ego->cld1, I, I); /* global transpose */ cldt = (plan_rdft *) ego->cldt; cldt->apply(ego->cldt, I, O); /* RDFT final local dimension */ cld2 = (plan_rdft *) ego->cld2; cld2->apply(ego->cld2, O, O); } static int applicable(const S *ego, const problem *p_, const planner *plnr) { const problem_mpi_rdft *p = (const problem_mpi_rdft *) p_; return (1 && p->sz->rnk > 1 && p->flags == TRANSPOSED_OUT && (!ego->preserve_input || (!NO_DESTROY_INPUTP(plnr) && p->I != p->O)) && XM(is_local_after)(1, p->sz, IB) && XM(is_local_after)(2, p->sz, OB) && XM(num_blocks)(p->sz->dims[0].n, p->sz->dims[0].b[OB]) == 1 && (!NO_SLOWP(plnr) /* slow if rdft-serial is applicable */ || !XM(rdft_serial_applicable)(p)) ); } static void awake(plan *ego_, enum wakefulness wakefulness) { P *ego = (P *) ego_; X(plan_awake)(ego->cld1, wakefulness); X(plan_awake)(ego->cldt, wakefulness); X(plan_awake)(ego->cld2, wakefulness); } static void destroy(plan *ego_) { P *ego = (P *) ego_; X(plan_destroy_internal)(ego->cld2); X(plan_destroy_internal)(ego->cldt); X(plan_destroy_internal)(ego->cld1); } static void print(const plan *ego_, printer *p) { const P *ego = (const P *) ego_; p->print(p, "(mpi-rdft-rank-geq2-transposed%s%(%p%)%(%p%)%(%p%))", ego->preserve_input==2 ?"/p":"", ego->cld1, ego->cldt, ego->cld2); } static plan *mkplan(const solver *ego_, const problem *p_, planner *plnr) { const S *ego = (const S *) ego_; const problem_mpi_rdft *p; P *pln; plan *cld1 = 0, *cldt = 0, *cld2 = 0; R *I, *O, *I2; tensor *sz; int i, my_pe, n_pes; INT nrest; static const plan_adt padt = { XM(rdft_solve), awake, print, destroy }; UNUSED(ego); if (!applicable(ego, p_, plnr)) return (plan *) 0; p = (const problem_mpi_rdft *) p_; I2 = I = p->I; O = p->O; if (ego->preserve_input || NO_DESTROY_INPUTP(plnr)) I = O; MPI_Comm_rank(p->comm, &my_pe); MPI_Comm_size(p->comm, &n_pes); sz = X(mktensor)(p->sz->rnk - 1); /* tensor of last rnk-1 dimensions */ i = p->sz->rnk - 2; A(i >= 0); sz->dims[i].n = p->sz->dims[i+1].n; sz->dims[i].is = sz->dims[i].os = p->vn; for (--i; i >= 0; --i) { sz->dims[i].n = p->sz->dims[i+1].n; sz->dims[i].is = sz->dims[i].os = sz->dims[i+1].n * sz->dims[i+1].is; } nrest = 1; for (i = 1; i < sz->rnk; ++i) nrest *= sz->dims[i].n; { INT is = sz->dims[0].n * sz->dims[0].is; INT b = XM(block)(p->sz->dims[0].n, p->sz->dims[0].b[IB], my_pe); cld1 = X(mkplan_d)(plnr, X(mkproblem_rdft_d)(sz, X(mktensor_2d)(b, is, is, p->vn, 1, 1), I2, I, p->kind + 1)); if (XM(any_true)(!cld1, p->comm)) goto nada; } nrest *= p->vn; cldt = X(mkplan_d)(plnr, XM(mkproblem_transpose)( p->sz->dims[0].n, p->sz->dims[1].n, nrest, I, O, p->sz->dims[0].b[IB], p->sz->dims[1].b[OB], p->comm, 0)); if (XM(any_true)(!cldt, p->comm)) goto nada; { INT is = p->sz->dims[0].n * nrest; INT b = XM(block)(p->sz->dims[1].n, p->sz->dims[1].b[OB], my_pe); cld2 = X(mkplan_d)(plnr, X(mkproblem_rdft_1_d)(X(mktensor_1d)( p->sz->dims[0].n, nrest, nrest), X(mktensor_2d)(b, is, is, nrest, 1, 1), O, O, p->kind[0])); if (XM(any_true)(!cld2, p->comm)) goto nada; } pln = MKPLAN_MPI_RDFT(P, &padt, apply); pln->cld1 = cld1; pln->cldt = cldt; pln->cld2 = cld2; pln->preserve_input = ego->preserve_input ? 2 : NO_DESTROY_INPUTP(plnr); X(ops_add)(&cld1->ops, &cld2->ops, &pln->super.super.ops); X(ops_add2)(&cldt->ops, &pln->super.super.ops); return &(pln->super.super); nada: X(plan_destroy_internal)(cld2); X(plan_destroy_internal)(cldt); X(plan_destroy_internal)(cld1); return (plan *) 0; } static solver *mksolver(int preserve_input) { static const solver_adt sadt = { PROBLEM_MPI_RDFT, mkplan, 0 }; S *slv = MKSOLVER(S, &sadt); slv->preserve_input = preserve_input; return &(slv->super); } void XM(rdft_rank_geq2_transposed_register)(planner *p) { int preserve_input; for (preserve_input = 0; preserve_input <= 1; ++preserve_input) REGISTER_SOLVER(p, mksolver(preserve_input)); } fftw-3.3.3/mpi/rdft-rank1-bigvec.c0000644000175000017500000001427112054410050013542 00000000000000/* * Copyright (c) 2003, 2007-11 Matteo Frigo * Copyright (c) 2003, 2007-11 Massachusetts Institute of Technology * * This program is free software; you can redistribute it and/or modify * it under the terms of the GNU General Public License as published by * the Free Software Foundation; either version 2 of the License, or * (at your option) any later version. * * This program is distributed in the hope that it will be useful, * but WITHOUT ANY WARRANTY; without even the implied warranty of * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the * GNU General Public License for more details. * * You should have received a copy of the GNU General Public License * along with this program; if not, write to the Free Software * Foundation, Inc., 51 Franklin Street, Fifth Floor, Boston, MA 02110-1301 USA * */ /* Complex RDFTs of rank == 1 when the vector length vn is >= # processes. In this case, we don't need to use a six-step type algorithm, and can instead transpose the RDFT dimension with the vector dimension to make the RDFT local. */ #include "mpi-rdft.h" #include "mpi-transpose.h" typedef struct { solver super; int preserve_input; /* preserve input even if DESTROY_INPUT was passed */ rearrangement rearrange; } S; typedef struct { plan_mpi_rdft super; plan *cldt_before, *cld, *cldt_after; int preserve_input; rearrangement rearrange; } P; static void apply(const plan *ego_, R *I, R *O) { const P *ego = (const P *) ego_; plan_rdft *cld, *cldt_before, *cldt_after; /* global transpose */ cldt_before = (plan_rdft *) ego->cldt_before; cldt_before->apply(ego->cldt_before, I, O); if (ego->preserve_input) I = O; /* 1d RDFT(s) */ cld = (plan_rdft *) ego->cld; cld->apply(ego->cld, O, I); /* global transpose */ cldt_after = (plan_rdft *) ego->cldt_after; cldt_after->apply(ego->cldt_after, I, O); } static int applicable(const S *ego, const problem *p_, const planner *plnr) { const problem_mpi_rdft *p = (const problem_mpi_rdft *) p_; int n_pes; MPI_Comm_size(p->comm, &n_pes); return (1 && p->sz->rnk == 1 && !(p->flags & ~RANK1_BIGVEC_ONLY) && (!ego->preserve_input || (!NO_DESTROY_INPUTP(plnr) && p->I != p->O)) #if 0 /* don't need this check since no other rank-1 rdft solver */ && (p->vn >= n_pes /* TODO: relax this, using more memory? */ || (p->flags & RANK1_BIGVEC_ONLY)) #endif && XM(rearrange_applicable)(ego->rearrange, p->sz->dims[0], p->vn, n_pes) && (!NO_SLOWP(plnr) /* slow if rdft-serial is applicable */ || !XM(rdft_serial_applicable)(p)) ); } static void awake(plan *ego_, enum wakefulness wakefulness) { P *ego = (P *) ego_; X(plan_awake)(ego->cldt_before, wakefulness); X(plan_awake)(ego->cld, wakefulness); X(plan_awake)(ego->cldt_after, wakefulness); } static void destroy(plan *ego_) { P *ego = (P *) ego_; X(plan_destroy_internal)(ego->cldt_after); X(plan_destroy_internal)(ego->cld); X(plan_destroy_internal)(ego->cldt_before); } static void print(const plan *ego_, printer *p) { const P *ego = (const P *) ego_; const char descrip[][16] = { "contig", "discontig", "square-after", "square-middle", "square-before" }; p->print(p, "(mpi-rdft-rank1-bigvec/%s%s %(%p%) %(%p%) %(%p%))", descrip[ego->rearrange], ego->preserve_input==2 ?"/p":"", ego->cldt_before, ego->cld, ego->cldt_after); } static plan *mkplan(const solver *ego_, const problem *p_, planner *plnr) { const S *ego = (const S *) ego_; const problem_mpi_rdft *p; P *pln; plan *cld = 0, *cldt_before = 0, *cldt_after = 0; R *I, *O; INT yblock, yb, nx, ny, vn; int my_pe, n_pes; static const plan_adt padt = { XM(rdft_solve), awake, print, destroy }; UNUSED(ego); if (!applicable(ego, p_, plnr)) return (plan *) 0; p = (const problem_mpi_rdft *) p_; MPI_Comm_rank(p->comm, &my_pe); MPI_Comm_size(p->comm, &n_pes); nx = p->sz->dims[0].n; if (!(ny = XM(rearrange_ny)(ego->rearrange, p->sz->dims[0],p->vn,n_pes))) return (plan *) 0; vn = p->vn / ny; A(ny * vn == p->vn); yblock = XM(default_block)(ny, n_pes); cldt_before = X(mkplan_d)(plnr, XM(mkproblem_transpose)( nx, ny, vn, I = p->I, O = p->O, p->sz->dims[0].b[IB], yblock, p->comm, 0)); if (XM(any_true)(!cldt_before, p->comm)) goto nada; if (ego->preserve_input || NO_DESTROY_INPUTP(plnr)) { I = O; } yb = XM(block)(ny, yblock, my_pe); cld = X(mkplan_d)(plnr, X(mkproblem_rdft_1_d)(X(mktensor_1d)(nx, vn, vn), X(mktensor_2d)(yb, vn*nx, vn*nx, vn, 1, 1), O, I, p->kind[0])); if (XM(any_true)(!cld, p->comm)) goto nada; cldt_after = X(mkplan_d)(plnr, XM(mkproblem_transpose)( ny, nx, vn, I, O, yblock, p->sz->dims[0].b[OB], p->comm, 0)); if (XM(any_true)(!cldt_after, p->comm)) goto nada; pln = MKPLAN_MPI_RDFT(P, &padt, apply); pln->cldt_before = cldt_before; pln->cld = cld; pln->cldt_after = cldt_after; pln->preserve_input = ego->preserve_input ? 2 : NO_DESTROY_INPUTP(plnr); pln->rearrange = ego->rearrange; X(ops_add)(&cldt_before->ops, &cld->ops, &pln->super.super.ops); X(ops_add2)(&cldt_after->ops, &pln->super.super.ops); return &(pln->super.super); nada: X(plan_destroy_internal)(cldt_after); X(plan_destroy_internal)(cld); X(plan_destroy_internal)(cldt_before); return (plan *) 0; } static solver *mksolver(rearrangement rearrange, int preserve_input) { static const solver_adt sadt = { PROBLEM_MPI_RDFT, mkplan, 0 }; S *slv = MKSOLVER(S, &sadt); slv->rearrange = rearrange; slv->preserve_input = preserve_input; return &(slv->super); } void XM(rdft_rank1_bigvec_register)(planner *p) { rearrangement rearrange; int preserve_input; FORALL_REARRANGE(rearrange) for (preserve_input = 0; preserve_input <= 1; ++preserve_input) REGISTER_SOLVER(p, mksolver(rearrange, preserve_input)); } fftw-3.3.3/mpi/rdft-problem.c0000644000175000017500000001117612054410050012732 00000000000000/* * Copyright (c) 2003, 2007-11 Matteo Frigo * Copyright (c) 2003, 2007-11 Massachusetts Institute of Technology * * This program is free software; you can redistribute it and/or modify * it under the terms of the GNU General Public License as published by * the Free Software Foundation; either version 2 of the License, or * (at your option) any later version. * * This program is distributed in the hope that it will be useful, * but WITHOUT ANY WARRANTY; without even the implied warranty of * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the * GNU General Public License for more details. * * You should have received a copy of the GNU General Public License * along with this program; if not, write to the Free Software * Foundation, Inc., 51 Franklin Street, Fifth Floor, Boston, MA 02110-1301 USA * */ #include "mpi-rdft.h" static void destroy(problem *ego_) { problem_mpi_rdft *ego = (problem_mpi_rdft *) ego_; XM(dtensor_destroy)(ego->sz); MPI_Comm_free(&ego->comm); #if !defined(STRUCT_HACK_C99) && !defined(STRUCT_HACK_KR) X(ifree0)(ego->kind); #endif X(ifree)(ego_); } static void hash(const problem *p_, md5 *m) { const problem_mpi_rdft *p = (const problem_mpi_rdft *) p_; int i; X(md5puts)(m, "mpi-dft"); X(md5int)(m, p->I == p->O); /* don't include alignment -- may differ between processes X(md5int)(m, X(alignment_of)(p->I)); X(md5int)(m, X(alignment_of)(p->O)); ... note that applicability of MPI plans does not depend on alignment (although optimality may, in principle). */ XM(dtensor_md5)(m, p->sz); X(md5INT)(m, p->vn); for (i = 0; i < p->sz->rnk; ++i) X(md5int)(m, p->kind[i]); X(md5int)(m, p->flags); MPI_Comm_size(p->comm, &i); X(md5int)(m, i); A(XM(md5_equal)(*m, p->comm)); } static void print(const problem *ego_, printer *p) { const problem_mpi_rdft *ego = (const problem_mpi_rdft *) ego_; int i; p->print(p, "(mpi-rdft %d %d %d ", ego->I == ego->O, X(alignment_of)(ego->I), X(alignment_of)(ego->O)); XM(dtensor_print)(ego->sz, p); for (i = 0; i < ego->sz->rnk; ++i) p->print(p, " %d", (int)ego->kind[i]); p->print(p, " %D %d", ego->vn, ego->flags); MPI_Comm_size(ego->comm, &i); p->print(p, " %d)", i); } static void zero(const problem *ego_) { const problem_mpi_rdft *ego = (const problem_mpi_rdft *) ego_; R *I = ego->I; INT i, N; int my_pe; MPI_Comm_rank(ego->comm, &my_pe); N = ego->vn * XM(total_block)(ego->sz, IB, my_pe); for (i = 0; i < N; ++i) I[i] = K(0.0); } static const problem_adt padt = { PROBLEM_MPI_RDFT, hash, zero, print, destroy }; problem *XM(mkproblem_rdft)(const dtensor *sz, INT vn, R *I, R *O, MPI_Comm comm, const rdft_kind *kind, unsigned flags) { problem_mpi_rdft *ego; int i, rnk = sz->rnk; int n_pes; A(XM(dtensor_validp)(sz) && FINITE_RNK(sz->rnk)); MPI_Comm_size(comm, &n_pes); A(n_pes >= XM(num_blocks_total)(sz, IB) && n_pes >= XM(num_blocks_total)(sz, OB)); A(vn >= 0); #if defined(STRUCT_HACK_KR) ego = (problem_mpi_rdft *) X(mkproblem)(sizeof(problem_mpi_rdft) + sizeof(rdft_kind) * (rnk > 0 ? rnk - 1 : 0), &padt); #elif defined(STRUCT_HACK_C99) ego = (problem_mpi_rdft *) X(mkproblem)(sizeof(problem_mpi_rdft) + sizeof(rdft_kind) * rnk, &padt); #else ego = (problem_mpi_rdft *) X(mkproblem)(sizeof(problem_mpi_rdft), &padt); ego->kind = (rdft_kind *) MALLOC(sizeof(rdft_kind) * rnk, PROBLEMS); #endif /* enforce pointer equality if untainted pointers are equal */ if (UNTAINT(I) == UNTAINT(O)) I = O = JOIN_TAINT(I, O); ego->sz = XM(dtensor_canonical)(sz, 0); ego->vn = vn; ego->I = I; ego->O = O; for (i = 0; i< ego->sz->rnk; ++i) ego->kind[i] = kind[i]; /* canonicalize: replace TRANSPOSED_IN with TRANSPOSED_OUT by swapping the first two dimensions (for rnk > 1) */ if ((flags & TRANSPOSED_IN) && ego->sz->rnk > 1) { rdft_kind k = ego->kind[0]; ddim dim0 = ego->sz->dims[0]; ego->sz->dims[0] = ego->sz->dims[1]; ego->sz->dims[1] = dim0; ego->kind[0] = ego->kind[1]; ego->kind[1] = k; flags &= ~TRANSPOSED_IN; flags ^= TRANSPOSED_OUT; } ego->flags = flags; MPI_Comm_dup(comm, &ego->comm); return &(ego->super); } problem *XM(mkproblem_rdft_d)(dtensor *sz, INT vn, R *I, R *O, MPI_Comm comm, const rdft_kind *kind, unsigned flags) { problem *p = XM(mkproblem_rdft)(sz, vn, I, O, comm, kind, flags); XM(dtensor_destroy)(sz); return p; } fftw-3.3.3/mpi/rdft-solve.c0000644000175000017500000000221212054410050012411 00000000000000/* * Copyright (c) 2003, 2007-11 Matteo Frigo * Copyright (c) 2003, 2007-11 Massachusetts Institute of Technology * * This program is free software; you can redistribute it and/or modify * it under the terms of the GNU General Public License as published by * the Free Software Foundation; either version 2 of the License, or * (at your option) any later version. * * This program is distributed in the hope that it will be useful, * but WITHOUT ANY WARRANTY; without even the implied warranty of * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the * GNU General Public License for more details. * * You should have received a copy of the GNU General Public License * along with this program; if not, write to the Free Software * Foundation, Inc., 51 Franklin Street, Fifth Floor, Boston, MA 02110-1301 USA * */ #include "mpi-rdft.h" /* use the apply() operation for MPI_RDFT problems */ void XM(rdft_solve)(const plan *ego_, const problem *p_) { const plan_mpi_rdft *ego = (const plan_mpi_rdft *) ego_; const problem_mpi_rdft *p = (const problem_mpi_rdft *) p_; ego->apply(ego_, UNTAINT(p->I), UNTAINT(p->O)); } fftw-3.3.3/mpi/mpi-rdft.h0000644000175000017500000000417712054410050012067 00000000000000/* * Copyright (c) 2003, 2007-11 Matteo Frigo * Copyright (c) 2003, 2007-11 Massachusetts Institute of Technology * * This program is free software; you can redistribute it and/or modify * it under the terms of the GNU General Public License as published by * the Free Software Foundation; either version 2 of the License, or * (at your option) any later version. * * This program is distributed in the hope that it will be useful, * but WITHOUT ANY WARRANTY; without even the implied warranty of * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the * GNU General Public License for more details. * * You should have received a copy of the GNU General Public License * along with this program; if not, write to the Free Software * Foundation, Inc., 51 Franklin Street, Fifth Floor, Boston, MA 02110-1301 USA * */ #include "ifftw-mpi.h" /* problem.c: */ typedef struct { problem super; dtensor *sz; INT vn; /* vector length (vector stride 1) */ R *I, *O; /* contiguous interleaved arrays */ unsigned flags; /* TRANSPOSED_IN/OUT meaningful for rnk>1 only SCRAMBLED_IN/OUT meaningful for 1d transforms only */ MPI_Comm comm; #if defined(STRUCT_HACK_KR) rdft_kind kind[1]; #elif defined(STRUCT_HACK_C99) rdft_kind kind[]; #else rdft_kind *kind; #endif } problem_mpi_rdft; problem *XM(mkproblem_rdft)(const dtensor *sz, INT vn, R *I, R *O, MPI_Comm comm, const rdft_kind *kind, unsigned flags); problem *XM(mkproblem_rdft_d)(dtensor *sz, INT vn, R *I, R *O, MPI_Comm comm, const rdft_kind *kind, unsigned flags); /* solve.c: */ void XM(rdft_solve)(const plan *ego_, const problem *p_); /* plans have same operands as rdft plans, so just re-use */ typedef plan_rdft plan_mpi_rdft; #define MKPLAN_MPI_RDFT(type, adt, apply) \ (type *)X(mkplan_rdft)(sizeof(type), adt, apply) int XM(rdft_serial_applicable)(const problem_mpi_rdft *p); /* various solvers */ void XM(rdft_rank_geq2_register)(planner *p); void XM(rdft_rank_geq2_transposed_register)(planner *p); void XM(rdft_serial_register)(planner *p); void XM(rdft_rank1_bigvec_register)(planner *p); fftw-3.3.3/mpi/rdft2-serial.c0000644000175000017500000000773212054410050012636 00000000000000/* * Copyright (c) 2003, 2007-11 Matteo Frigo * Copyright (c) 2003, 2007-11 Massachusetts Institute of Technology * * This program is free software; you can redistribute it and/or modify * it under the terms of the GNU General Public License as published by * the Free Software Foundation; either version 2 of the License, or * (at your option) any later version. * * This program is distributed in the hope that it will be useful, * but WITHOUT ANY WARRANTY; without even the implied warranty of * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the * GNU General Public License for more details. * * You should have received a copy of the GNU General Public License * along with this program; if not, write to the Free Software * Foundation, Inc., 51 Franklin Street, Fifth Floor, Boston, MA 02110-1301 USA * */ /* "MPI" DFTs where all of the data is on one processor...just call through to serial API. */ #include "mpi-rdft2.h" #include "rdft.h" typedef struct { plan_mpi_rdft2 super; plan *cld; INT vn; } P; static void apply_r2c(const plan *ego_, R *I, R *O) { const P *ego = (const P *) ego_; plan_rdft2 *cld; cld = (plan_rdft2 *) ego->cld; cld->apply(ego->cld, I, I+ego->vn, O, O+1); } static void apply_c2r(const plan *ego_, R *I, R *O) { const P *ego = (const P *) ego_; plan_rdft2 *cld; cld = (plan_rdft2 *) ego->cld; cld->apply(ego->cld, O, O+ego->vn, I, I+1); } static void awake(plan *ego_, enum wakefulness wakefulness) { P *ego = (P *) ego_; X(plan_awake)(ego->cld, wakefulness); } static void destroy(plan *ego_) { P *ego = (P *) ego_; X(plan_destroy_internal)(ego->cld); } static void print(const plan *ego_, printer *p) { const P *ego = (const P *) ego_; p->print(p, "(mpi-rdft2-serial %(%p%))", ego->cld); } int XM(rdft2_serial_applicable)(const problem_mpi_rdft2 *p) { return (1 && p->flags == 0 /* TRANSPOSED/SCRAMBLED_IN/OUT not supported */ && ((XM(is_local)(p->sz, IB) && XM(is_local)(p->sz, OB)) || p->vn == 0)); } static plan *mkplan(const solver *ego, const problem *p_, planner *plnr) { const problem_mpi_rdft2 *p = (const problem_mpi_rdft2 *) p_; P *pln; plan *cld; int my_pe; R *r0, *r1, *cr, *ci; static const plan_adt padt = { XM(rdft2_solve), awake, print, destroy }; UNUSED(ego); /* check whether applicable: */ if (!XM(rdft2_serial_applicable)(p)) return (plan *) 0; if (p->kind == R2HC) { r1 = (r0 = p->I) + p->vn; ci = (cr = p->O) + 1; } else { r1 = (r0 = p->O) + p->vn; ci = (cr = p->I) + 1; } MPI_Comm_rank(p->comm, &my_pe); if (my_pe == 0 && p->vn > 0) { INT ivs = 1 + (p->kind == HC2R), ovs = 1 + (p->kind == R2HC); int i, rnk = p->sz->rnk; tensor *sz = X(mktensor)(p->sz->rnk); sz->dims[rnk - 1].is = sz->dims[rnk - 1].os = 2 * p->vn; sz->dims[rnk - 1].n = p->sz->dims[rnk - 1].n / 2 + 1; for (i = rnk - 1; i > 0; --i) { sz->dims[i - 1].is = sz->dims[i - 1].os = sz->dims[i].is * sz->dims[i].n; sz->dims[i - 1].n = p->sz->dims[i - 1].n; } sz->dims[rnk - 1].n = p->sz->dims[rnk - 1].n; cld = X(mkplan_d)(plnr, X(mkproblem_rdft2_d)(sz, X(mktensor_1d)(p->vn,ivs,ovs), r0, r1, cr, ci, p->kind)); } else { /* idle process: make nop plan */ cld = X(mkplan_d)(plnr, X(mkproblem_rdft2_d)(X(mktensor_0d)(), X(mktensor_1d)(0,0,0), cr, ci, cr, ci, HC2R)); } if (XM(any_true)(!cld, p->comm)) return (plan *) 0; pln = MKPLAN_MPI_RDFT2(P, &padt, p->kind == R2HC ? apply_r2c : apply_c2r); pln->cld = cld; pln->vn = p->vn; X(ops_cpy)(&cld->ops, &pln->super.super.ops); return &(pln->super.super); } static solver *mksolver(void) { static const solver_adt sadt = { PROBLEM_MPI_RDFT2, mkplan, 0 }; return MKSOLVER(solver, &sadt); } void XM(rdft2_serial_register)(planner *p) { REGISTER_SOLVER(p, mksolver()); } fftw-3.3.3/mpi/rdft2-rank-geq2.c0000644000175000017500000001430312054410050013136 00000000000000/* * Copyright (c) 2003, 2007-11 Matteo Frigo * Copyright (c) 2003, 2007-11 Massachusetts Institute of Technology * * This program is free software; you can redistribute it and/or modify * it under the terms of the GNU General Public License as published by * the Free Software Foundation; either version 2 of the License, or * (at your option) any later version. * * This program is distributed in the hope that it will be useful, * but WITHOUT ANY WARRANTY; without even the implied warranty of * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the * GNU General Public License for more details. * * You should have received a copy of the GNU General Public License * along with this program; if not, write to the Free Software * Foundation, Inc., 51 Franklin Street, Fifth Floor, Boston, MA 02110-1301 USA * */ /* Complex RDFT2s of rank >= 2, for the case where we are distributed across the first dimension only, and the output is not transposed. */ #include "mpi-dft.h" #include "mpi-rdft2.h" #include "rdft.h" typedef struct { solver super; int preserve_input; /* preserve input even if DESTROY_INPUT was passed */ } S; typedef struct { plan_mpi_rdft2 super; plan *cld1, *cld2; INT vn; int preserve_input; } P; static void apply_r2c(const plan *ego_, R *I, R *O) { const P *ego = (const P *) ego_; plan_rdft2 *cld1; plan_rdft *cld2; /* RDFT2 local dimensions */ cld1 = (plan_rdft2 *) ego->cld1; if (ego->preserve_input) { cld1->apply(ego->cld1, I, I+ego->vn, O, O+1); I = O; } else cld1->apply(ego->cld1, I, I+ego->vn, I, I+1); /* DFT non-local dimension (via dft-rank1-bigvec, usually): */ cld2 = (plan_rdft *) ego->cld2; cld2->apply(ego->cld2, I, O); } static void apply_c2r(const plan *ego_, R *I, R *O) { const P *ego = (const P *) ego_; plan_rdft2 *cld1; plan_rdft *cld2; /* DFT non-local dimension (via dft-rank1-bigvec, usually): */ cld2 = (plan_rdft *) ego->cld2; cld2->apply(ego->cld2, I, O); /* RDFT2 local dimensions */ cld1 = (plan_rdft2 *) ego->cld1; cld1->apply(ego->cld1, O, O+ego->vn, O, O+1); } static int applicable(const S *ego, const problem *p_, const planner *plnr) { const problem_mpi_rdft2 *p = (const problem_mpi_rdft2 *) p_; return (1 && p->sz->rnk > 1 && p->flags == 0 /* TRANSPOSED/SCRAMBLED_IN/OUT not supported */ && (!ego->preserve_input || (!NO_DESTROY_INPUTP(plnr) && p->I != p->O && p->kind == R2HC)) && XM(is_local_after)(1, p->sz, IB) && XM(is_local_after)(1, p->sz, OB) && (!NO_SLOWP(plnr) /* slow if rdft2-serial is applicable */ || !XM(rdft2_serial_applicable)(p)) ); } static void awake(plan *ego_, enum wakefulness wakefulness) { P *ego = (P *) ego_; X(plan_awake)(ego->cld1, wakefulness); X(plan_awake)(ego->cld2, wakefulness); } static void destroy(plan *ego_) { P *ego = (P *) ego_; X(plan_destroy_internal)(ego->cld2); X(plan_destroy_internal)(ego->cld1); } static void print(const plan *ego_, printer *p) { const P *ego = (const P *) ego_; p->print(p, "(mpi-rdft2-rank-geq2%s%(%p%)%(%p%))", ego->preserve_input==2 ?"/p":"", ego->cld1, ego->cld2); } static plan *mkplan(const solver *ego_, const problem *p_, planner *plnr) { const S *ego = (const S *) ego_; const problem_mpi_rdft2 *p; P *pln; plan *cld1 = 0, *cld2 = 0; R *r0, *r1, *cr, *ci, *I, *O; tensor *sz; dtensor *sz2; int i, my_pe, n_pes; INT nrest; static const plan_adt padt = { XM(rdft2_solve), awake, print, destroy }; UNUSED(ego); if (!applicable(ego, p_, plnr)) return (plan *) 0; p = (const problem_mpi_rdft2 *) p_; I = p->I; O = p->O; if (p->kind == R2HC) { r1 = (r0 = p->I) + p->vn; if (ego->preserve_input || NO_DESTROY_INPUTP(plnr)) { ci = (cr = p->O) + 1; I = O; } else ci = (cr = p->I) + 1; } else { r1 = (r0 = p->O) + p->vn; ci = (cr = p->O) + 1; } MPI_Comm_rank(p->comm, &my_pe); MPI_Comm_size(p->comm, &n_pes); sz = X(mktensor)(p->sz->rnk - 1); /* tensor of last rnk-1 dimensions */ i = p->sz->rnk - 2; A(i >= 0); sz->dims[i].is = sz->dims[i].os = 2 * p->vn; sz->dims[i].n = p->sz->dims[i+1].n / 2 + 1; for (--i; i >= 0; --i) { sz->dims[i].n = p->sz->dims[i+1].n; sz->dims[i].is = sz->dims[i].os = sz->dims[i+1].n * sz->dims[i+1].is; } nrest = X(tensor_sz)(sz); { INT ivs = 1 + (p->kind == HC2R), ovs = 1 + (p->kind == R2HC); INT is = sz->dims[0].n * sz->dims[0].is; INT b = XM(block)(p->sz->dims[0].n, p->sz->dims[0].b[IB], my_pe); sz->dims[p->sz->rnk - 2].n = p->sz->dims[p->sz->rnk - 1].n; cld1 = X(mkplan_d)(plnr, X(mkproblem_rdft2_d)(sz, X(mktensor_2d)(b, is, is, p->vn,ivs,ovs), r0, r1, cr, ci, p->kind)); if (XM(any_true)(!cld1, p->comm)) goto nada; } sz2 = XM(mkdtensor)(1); /* tensor for first (distributed) dimension */ sz2->dims[0] = p->sz->dims[0]; cld2 = X(mkplan_d)(plnr, XM(mkproblem_dft_d)(sz2, nrest * p->vn, I, O, p->comm, p->kind == R2HC ? FFT_SIGN : -FFT_SIGN, RANK1_BIGVEC_ONLY)); if (XM(any_true)(!cld2, p->comm)) goto nada; pln = MKPLAN_MPI_RDFT2(P, &padt, p->kind == R2HC ? apply_r2c : apply_c2r); pln->cld1 = cld1; pln->cld2 = cld2; pln->preserve_input = ego->preserve_input ? 2 : NO_DESTROY_INPUTP(plnr); pln->vn = p->vn; X(ops_add)(&cld1->ops, &cld2->ops, &pln->super.super.ops); return &(pln->super.super); nada: X(plan_destroy_internal)(cld2); X(plan_destroy_internal)(cld1); return (plan *) 0; } static solver *mksolver(int preserve_input) { static const solver_adt sadt = { PROBLEM_MPI_RDFT2, mkplan, 0 }; S *slv = MKSOLVER(S, &sadt); slv->preserve_input = preserve_input; return &(slv->super); } void XM(rdft2_rank_geq2_register)(planner *p) { int preserve_input; for (preserve_input = 0; preserve_input <= 1; ++preserve_input) REGISTER_SOLVER(p, mksolver(preserve_input)); } fftw-3.3.3/mpi/rdft2-rank-geq2-transposed.c0000644000175000017500000002023112054410050015313 00000000000000/* * Copyright (c) 2003, 2007-11 Matteo Frigo * Copyright (c) 2003, 2007-11 Massachusetts Institute of Technology * * This program is free software; you can redistribute it and/or modify * it under the terms of the GNU General Public License as published by * the Free Software Foundation; either version 2 of the License, or * (at your option) any later version. * * This program is distributed in the hope that it will be useful, * but WITHOUT ANY WARRANTY; without even the implied warranty of * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the * GNU General Public License for more details. * * You should have received a copy of the GNU General Public License * along with this program; if not, write to the Free Software * Foundation, Inc., 51 Franklin Street, Fifth Floor, Boston, MA 02110-1301 USA * */ /* Real-input (r2c) DFTs of rank >= 2, for the case where we are distributed across the first dimension only, and the output is transposed both in data distribution and in ordering (for the first 2 dimensions). Conversely, real-output (c2r) DFTs where the input is transposed. We don't currently support transposed-input r2c or transposed-output c2r transforms. */ #include "mpi-rdft2.h" #include "mpi-transpose.h" #include "rdft.h" #include "dft.h" typedef struct { solver super; int preserve_input; /* preserve input even if DESTROY_INPUT was passed */ } S; typedef struct { plan_mpi_rdft2 super; plan *cld1, *cldt, *cld2; INT vn; int preserve_input; } P; static void apply_r2c(const plan *ego_, R *I, R *O) { const P *ego = (const P *) ego_; plan_rdft2 *cld1; plan_dft *cld2; plan_rdft *cldt; /* RDFT2 local dimensions */ cld1 = (plan_rdft2 *) ego->cld1; if (ego->preserve_input) { cld1->apply(ego->cld1, I, I+ego->vn, O, O+1); I = O; } else cld1->apply(ego->cld1, I, I+ego->vn, I, I+1); /* global transpose */ cldt = (plan_rdft *) ego->cldt; cldt->apply(ego->cldt, I, O); /* DFT final local dimension */ cld2 = (plan_dft *) ego->cld2; cld2->apply(ego->cld2, O, O+1, O, O+1); } static void apply_c2r(const plan *ego_, R *I, R *O) { const P *ego = (const P *) ego_; plan_rdft2 *cld1; plan_dft *cld2; plan_rdft *cldt; /* IDFT local dimensions */ cld2 = (plan_dft *) ego->cld2; if (ego->preserve_input) { cld2->apply(ego->cld2, I+1, I, O+1, O); I = O; } else cld2->apply(ego->cld2, I+1, I, I+1, I); /* global transpose */ cldt = (plan_rdft *) ego->cldt; cldt->apply(ego->cldt, I, O); /* RDFT2 final local dimension */ cld1 = (plan_rdft2 *) ego->cld1; cld1->apply(ego->cld1, O, O+ego->vn, O, O+1); } static int applicable(const S *ego, const problem *p_, const planner *plnr) { const problem_mpi_rdft2 *p = (const problem_mpi_rdft2 *) p_; return (1 && p->sz->rnk > 1 && (!ego->preserve_input || (!NO_DESTROY_INPUTP(plnr) && p->I != p->O)) && ((p->flags == TRANSPOSED_OUT && p->kind == R2HC && XM(is_local_after)(1, p->sz, IB) && XM(is_local_after)(2, p->sz, OB) && XM(num_blocks)(p->sz->dims[0].n, p->sz->dims[0].b[OB]) == 1) || (p->flags == TRANSPOSED_IN && p->kind == HC2R && XM(is_local_after)(1, p->sz, OB) && XM(is_local_after)(2, p->sz, IB) && XM(num_blocks)(p->sz->dims[0].n, p->sz->dims[0].b[IB]) == 1)) && (!NO_SLOWP(plnr) /* slow if rdft2-serial is applicable */ || !XM(rdft2_serial_applicable)(p)) ); } static void awake(plan *ego_, enum wakefulness wakefulness) { P *ego = (P *) ego_; X(plan_awake)(ego->cld1, wakefulness); X(plan_awake)(ego->cldt, wakefulness); X(plan_awake)(ego->cld2, wakefulness); } static void destroy(plan *ego_) { P *ego = (P *) ego_; X(plan_destroy_internal)(ego->cld2); X(plan_destroy_internal)(ego->cldt); X(plan_destroy_internal)(ego->cld1); } static void print(const plan *ego_, printer *p) { const P *ego = (const P *) ego_; p->print(p, "(mpi-rdft2-rank-geq2-transposed%s%(%p%)%(%p%)%(%p%))", ego->preserve_input==2 ?"/p":"", ego->cld1, ego->cldt, ego->cld2); } static plan *mkplan(const solver *ego_, const problem *p_, planner *plnr) { const S *ego = (const S *) ego_; const problem_mpi_rdft2 *p; P *pln; plan *cld1 = 0, *cldt = 0, *cld2 = 0; R *r0, *r1, *cr, *ci, *ri, *ii, *ro, *io, *I, *O; tensor *sz; int i, my_pe, n_pes; INT nrest, n1, b1; static const plan_adt padt = { XM(rdft2_solve), awake, print, destroy }; block_kind k1, k2; UNUSED(ego); if (!applicable(ego, p_, plnr)) return (plan *) 0; p = (const problem_mpi_rdft2 *) p_; I = p->I; O = p->O; if (p->kind == R2HC) { k1 = IB; k2 = OB; r1 = (r0 = I) + p->vn; if (ego->preserve_input || NO_DESTROY_INPUTP(plnr)) { ci = (cr = O) + 1; I = O; } else ci = (cr = I) + 1; io = ii = (ro = ri = O) + 1; } else { k1 = OB; k2 = IB; r1 = (r0 = O) + p->vn; ci = (cr = O) + 1; if (ego->preserve_input || NO_DESTROY_INPUTP(plnr)) { ri = (ii = I) + 1; ro = (io = O) + 1; I = O; } else ro = ri = (io = ii = I) + 1; } MPI_Comm_rank(p->comm, &my_pe); MPI_Comm_size(p->comm, &n_pes); sz = X(mktensor)(p->sz->rnk - 1); /* tensor of last rnk-1 dimensions */ i = p->sz->rnk - 2; A(i >= 0); sz->dims[i].n = p->sz->dims[i+1].n / 2 + 1; sz->dims[i].is = sz->dims[i].os = 2 * p->vn; for (--i; i >= 0; --i) { sz->dims[i].n = p->sz->dims[i+1].n; sz->dims[i].is = sz->dims[i].os = sz->dims[i+1].n * sz->dims[i+1].is; } nrest = 1; for (i = 1; i < sz->rnk; ++i) nrest *= sz->dims[i].n; { INT ivs = 1 + (p->kind == HC2R), ovs = 1 + (p->kind == R2HC); INT is = sz->dims[0].n * sz->dims[0].is; INT b = XM(block)(p->sz->dims[0].n, p->sz->dims[0].b[k1], my_pe); sz->dims[p->sz->rnk - 2].n = p->sz->dims[p->sz->rnk - 1].n; cld1 = X(mkplan_d)(plnr, X(mkproblem_rdft2_d)(sz, X(mktensor_2d)(b, is, is, p->vn,ivs,ovs), r0, r1, cr, ci, p->kind)); if (XM(any_true)(!cld1, p->comm)) goto nada; } nrest *= p->vn; n1 = p->sz->dims[1].n; b1 = p->sz->dims[1].b[k2]; if (p->sz->rnk == 2) { /* n1 dimension is cut in ~half */ n1 = n1 / 2 + 1; b1 = b1 == p->sz->dims[1].n ? n1 : b1; } if (p->kind == R2HC) cldt = X(mkplan_d)(plnr, XM(mkproblem_transpose)( p->sz->dims[0].n, n1, nrest * 2, I, O, p->sz->dims[0].b[IB], b1, p->comm, 0)); else cldt = X(mkplan_d)(plnr, XM(mkproblem_transpose)( n1, p->sz->dims[0].n, nrest * 2, I, O, b1, p->sz->dims[0].b[OB], p->comm, 0)); if (XM(any_true)(!cldt, p->comm)) goto nada; { INT is = p->sz->dims[0].n * nrest * 2; INT b = XM(block)(n1, b1, my_pe); cld2 = X(mkplan_d)(plnr, X(mkproblem_dft_d)(X(mktensor_1d)( p->sz->dims[0].n, nrest * 2, nrest * 2), X(mktensor_2d)(b, is, is, nrest, 2, 2), ri, ii, ro, io)); if (XM(any_true)(!cld2, p->comm)) goto nada; } pln = MKPLAN_MPI_RDFT2(P, &padt, p->kind == R2HC ? apply_r2c : apply_c2r); pln->cld1 = cld1; pln->cldt = cldt; pln->cld2 = cld2; pln->preserve_input = ego->preserve_input ? 2 : NO_DESTROY_INPUTP(plnr); pln->vn = p->vn; X(ops_add)(&cld1->ops, &cld2->ops, &pln->super.super.ops); X(ops_add2)(&cldt->ops, &pln->super.super.ops); return &(pln->super.super); nada: X(plan_destroy_internal)(cld2); X(plan_destroy_internal)(cldt); X(plan_destroy_internal)(cld1); return (plan *) 0; } static solver *mksolver(int preserve_input) { static const solver_adt sadt = { PROBLEM_MPI_RDFT2, mkplan, 0 }; S *slv = MKSOLVER(S, &sadt); slv->preserve_input = preserve_input; return &(slv->super); } void XM(rdft2_rank_geq2_transposed_register)(planner *p) { int preserve_input; for (preserve_input = 0; preserve_input <= 1; ++preserve_input) REGISTER_SOLVER(p, mksolver(preserve_input)); } fftw-3.3.3/mpi/rdft2-problem.c0000644000175000017500000000777712054410050013030 00000000000000/* * Copyright (c) 2003, 2007-11 Matteo Frigo * Copyright (c) 2003, 2007-11 Massachusetts Institute of Technology * * This program is free software; you can redistribute it and/or modify * it under the terms of the GNU General Public License as published by * the Free Software Foundation; either version 2 of the License, or * (at your option) any later version. * * This program is distributed in the hope that it will be useful, * but WITHOUT ANY WARRANTY; without even the implied warranty of * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the * GNU General Public License for more details. * * You should have received a copy of the GNU General Public License * along with this program; if not, write to the Free Software * Foundation, Inc., 51 Franklin Street, Fifth Floor, Boston, MA 02110-1301 USA * */ #include "mpi-rdft2.h" static void destroy(problem *ego_) { problem_mpi_rdft2 *ego = (problem_mpi_rdft2 *) ego_; XM(dtensor_destroy)(ego->sz); MPI_Comm_free(&ego->comm); X(ifree)(ego_); } static void hash(const problem *p_, md5 *m) { const problem_mpi_rdft2 *p = (const problem_mpi_rdft2 *) p_; int i; X(md5puts)(m, "mpi-rdft2"); X(md5int)(m, p->I == p->O); /* don't include alignment -- may differ between processes X(md5int)(m, X(alignment_of)(p->I)); X(md5int)(m, X(alignment_of)(p->O)); ... note that applicability of MPI plans does not depend on alignment (although optimality may, in principle). */ XM(dtensor_md5)(m, p->sz); X(md5INT)(m, p->vn); X(md5int)(m, p->kind); X(md5int)(m, p->flags); MPI_Comm_size(p->comm, &i); X(md5int)(m, i); A(XM(md5_equal)(*m, p->comm)); } static void print(const problem *ego_, printer *p) { const problem_mpi_rdft2 *ego = (const problem_mpi_rdft2 *) ego_; int i; p->print(p, "(mpi-rdft2 %d %d %d ", ego->I == ego->O, X(alignment_of)(ego->I), X(alignment_of)(ego->O)); XM(dtensor_print)(ego->sz, p); p->print(p, " %D %d %d", ego->vn, (int) ego->kind, ego->flags); MPI_Comm_size(ego->comm, &i); p->print(p, " %d)", i); } static void zero(const problem *ego_) { const problem_mpi_rdft2 *ego = (const problem_mpi_rdft2 *) ego_; R *I = ego->I; dtensor *sz; INT i, N; int my_pe; sz = XM(dtensor_copy)(ego->sz); sz->dims[sz->rnk - 1].n = sz->dims[sz->rnk - 1].n / 2 + 1; MPI_Comm_rank(ego->comm, &my_pe); N = 2 * ego->vn * XM(total_block)(sz, IB, my_pe); XM(dtensor_destroy)(sz); for (i = 0; i < N; ++i) I[i] = K(0.0); } static const problem_adt padt = { PROBLEM_MPI_RDFT2, hash, zero, print, destroy }; problem *XM(mkproblem_rdft2)(const dtensor *sz, INT vn, R *I, R *O, MPI_Comm comm, rdft_kind kind, unsigned flags) { problem_mpi_rdft2 *ego = (problem_mpi_rdft2 *)X(mkproblem)(sizeof(problem_mpi_rdft2), &padt); int n_pes; A(XM(dtensor_validp)(sz) && FINITE_RNK(sz->rnk) && sz->rnk > 1); MPI_Comm_size(comm, &n_pes); A(vn >= 0); A(kind == R2HC || kind == HC2R); /* enforce pointer equality if untainted pointers are equal */ if (UNTAINT(I) == UNTAINT(O)) I = O = JOIN_TAINT(I, O); ego->sz = XM(dtensor_canonical)(sz, 0); #ifdef FFTW_DEBUG ego->sz->dims[sz->rnk - 1].n = sz->dims[sz->rnk - 1].n / 2 + 1; A(n_pes >= XM(num_blocks_total)(ego->sz, IB) && n_pes >= XM(num_blocks_total)(ego->sz, OB)); ego->sz->dims[sz->rnk - 1].n = sz->dims[sz->rnk - 1].n; #endif ego->vn = vn; ego->I = I; ego->O = O; ego->kind = kind; /* We only support TRANSPOSED_OUT for r2c and TRANSPOSED_IN for c2r transforms. */ ego->flags = flags; MPI_Comm_dup(comm, &ego->comm); return &(ego->super); } problem *XM(mkproblem_rdft2_d)(dtensor *sz, INT vn, R *I, R *O, MPI_Comm comm, rdft_kind kind, unsigned flags) { problem *p = XM(mkproblem_rdft2)(sz, vn, I, O, comm, kind, flags); XM(dtensor_destroy)(sz); return p; } fftw-3.3.3/mpi/rdft2-solve.c0000644000175000017500000000222112054410050012473 00000000000000/* * Copyright (c) 2003, 2007-11 Matteo Frigo * Copyright (c) 2003, 2007-11 Massachusetts Institute of Technology * * This program is free software; you can redistribute it and/or modify * it under the terms of the GNU General Public License as published by * the Free Software Foundation; either version 2 of the License, or * (at your option) any later version. * * This program is distributed in the hope that it will be useful, * but WITHOUT ANY WARRANTY; without even the implied warranty of * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the * GNU General Public License for more details. * * You should have received a copy of the GNU General Public License * along with this program; if not, write to the Free Software * Foundation, Inc., 51 Franklin Street, Fifth Floor, Boston, MA 02110-1301 USA * */ #include "mpi-rdft2.h" /* use the apply() operation for MPI_RDFT2 problems */ void XM(rdft2_solve)(const plan *ego_, const problem *p_) { const plan_mpi_rdft2 *ego = (const plan_mpi_rdft2 *) ego_; const problem_mpi_rdft2 *p = (const problem_mpi_rdft2 *) p_; ego->apply(ego_, UNTAINT(p->I), UNTAINT(p->O)); } fftw-3.3.3/mpi/mpi-rdft2.h0000644000175000017500000000455612054410050012152 00000000000000/* * Copyright (c) 2003, 2007-11 Matteo Frigo * Copyright (c) 2003, 2007-11 Massachusetts Institute of Technology * * This program is free software; you can redistribute it and/or modify * it under the terms of the GNU General Public License as published by * the Free Software Foundation; either version 2 of the License, or * (at your option) any later version. * * This program is distributed in the hope that it will be useful, * but WITHOUT ANY WARRANTY; without even the implied warranty of * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the * GNU General Public License for more details. * * You should have received a copy of the GNU General Public License * along with this program; if not, write to the Free Software * Foundation, Inc., 51 Franklin Street, Fifth Floor, Boston, MA 02110-1301 USA * */ #include "ifftw-mpi.h" /* r2c and c2r transforms. The sz dtensor, as usual, gives the size of the "logical" complex array. For the last dimension N, however, only N/2+1 complex numbers are stored for the complex data. Moreover, for the real data, the last dimension is *always* padded to a size 2*(N/2+1). (Contrast this with the serial API, where there is only padding for in-place plans.) */ /* problem.c: */ typedef struct { problem super; dtensor *sz; INT vn; /* vector length (vector stride 1) */ R *I, *O; /* contiguous interleaved arrays */ rdft_kind kind; /* assert(kind < DHT) */ unsigned flags; /* TRANSPOSED_IN/OUT meaningful for rnk>1 only SCRAMBLED_IN/OUT meaningful for 1d transforms only */ MPI_Comm comm; } problem_mpi_rdft2; problem *XM(mkproblem_rdft2)(const dtensor *sz, INT vn, R *I, R *O, MPI_Comm comm, rdft_kind kind, unsigned flags); problem *XM(mkproblem_rdft2_d)(dtensor *sz, INT vn, R *I, R *O, MPI_Comm comm, rdft_kind kind, unsigned flags); /* solve.c: */ void XM(rdft2_solve)(const plan *ego_, const problem *p_); /* plans have same operands as rdft plans, so just re-use */ typedef plan_rdft plan_mpi_rdft2; #define MKPLAN_MPI_RDFT2(type, adt, apply) \ (type *)X(mkplan_rdft)(sizeof(type), adt, apply) int XM(rdft2_serial_applicable)(const problem_mpi_rdft2 *p); /* various solvers */ void XM(rdft2_rank_geq2_register)(planner *p); void XM(rdft2_rank_geq2_transposed_register)(planner *p); void XM(rdft2_serial_register)(planner *p); fftw-3.3.3/mpi/mpi-bench.c0000644000175000017500000005657512054410050012213 00000000000000/**************************************************************************/ /* NOTE to users: this is the FFTW-MPI self-test and benchmark program. It is probably NOT a good place to learn FFTW usage, since it has a lot of added complexity in order to exercise and test the full API, etcetera. We suggest reading the manual. */ /**************************************************************************/ #include #include #include #include "fftw3-mpi.h" #include "fftw-bench.h" #if defined(BENCHFFT_SINGLE) # define BENCH_MPI_TYPE MPI_FLOAT #elif defined(BENCHFFT_LDOUBLE) # define BENCH_MPI_TYPE MPI_LONG_DOUBLE #elif defined(BENCHFFT_QUAD) # error MPI quad-precision type is unknown #else # define BENCH_MPI_TYPE MPI_DOUBLE #endif #if SIZEOF_PTRDIFF_T == SIZEOF_INT # define FFTW_MPI_PTRDIFF_T MPI_INT #elif SIZEOF_PTRDIFF_T == SIZEOF_LONG # define FFTW_MPI_PTRDIFF_T MPI_LONG #elif SIZEOF_PTRDIFF_T == SIZEOF_LONG_LONG # define FFTW_MPI_PTRDIFF_T MPI_LONG_LONG #else # error MPI type for ptrdiff_t is unknown # define FFTW_MPI_PTRDIFF_T MPI_LONG #endif static const char *mkversion(void) { return FFTW(version); } static const char *mkcc(void) { return FFTW(cc); } static const char *mkcodelet_optim(void) { return FFTW(codelet_optim); } static const char *mknproc(void) { static char buf[32]; int ncpus; MPI_Comm_size(MPI_COMM_WORLD, &ncpus); #ifdef HAVE_SNPRINTF snprintf(buf, 32, "%d", ncpus); #else sprintf(buf, "%d", ncpus); #endif return buf; } BEGIN_BENCH_DOC BENCH_DOC("name", "fftw3_mpi") BENCH_DOCF("version", mkversion) BENCH_DOCF("cc", mkcc) BENCH_DOCF("codelet-optim", mkcodelet_optim) BENCH_DOCF("nproc", mknproc) END_BENCH_DOC static int n_pes = 1, my_pe = 0; /* global variables describing the shape of the data and its distribution */ static int rnk; static ptrdiff_t vn, iNtot, oNtot; static ptrdiff_t *local_ni=0, *local_starti=0; static ptrdiff_t *local_no=0, *local_starto=0; static ptrdiff_t *all_local_ni=0, *all_local_starti=0; /* n_pes x rnk arrays */ static ptrdiff_t *all_local_no=0, *all_local_starto=0; /* n_pes x rnk arrays */ static ptrdiff_t *istrides = 0, *ostrides = 0; static ptrdiff_t *total_ni=0, *total_no=0; static int *isend_cnt = 0, *isend_off = 0; /* for MPI_Scatterv */ static int *orecv_cnt = 0, *orecv_off = 0; /* for MPI_Gatherv */ static bench_real *local_in = 0, *local_out = 0; static bench_real *all_local_in = 0, *all_local_out = 0; static int all_local_in_alloc = 0, all_local_out_alloc = 0; static FFTW(plan) plan_scramble_in = 0, plan_unscramble_out = 0; static void alloc_rnk(int rnk_) { rnk = rnk_; bench_free(local_ni); if (rnk == 0) local_ni = 0; else local_ni = (ptrdiff_t *) bench_malloc(sizeof(ptrdiff_t) * rnk * (8 + n_pes * 4)); local_starti = local_ni + rnk; local_no = local_ni + 2 * rnk; local_starto = local_ni + 3 * rnk; istrides = local_ni + 4 * rnk; ostrides = local_ni + 5 * rnk; total_ni = local_ni + 6 * rnk; total_no = local_ni + 7 * rnk; all_local_ni = local_ni + 8 * rnk; all_local_starti = local_ni + (8 + n_pes) * rnk; all_local_no = local_ni + (8 + 2 * n_pes) * rnk; all_local_starto = local_ni + (8 + 3 * n_pes) * rnk; } static void setup_gather_scatter(void) { int i, j; ptrdiff_t off; MPI_Gather(local_ni, rnk, FFTW_MPI_PTRDIFF_T, all_local_ni, rnk, FFTW_MPI_PTRDIFF_T, 0, MPI_COMM_WORLD); MPI_Bcast(all_local_ni, rnk*n_pes, FFTW_MPI_PTRDIFF_T, 0, MPI_COMM_WORLD); MPI_Gather(local_starti, rnk, FFTW_MPI_PTRDIFF_T, all_local_starti, rnk, FFTW_MPI_PTRDIFF_T, 0, MPI_COMM_WORLD); MPI_Bcast(all_local_starti, rnk*n_pes, FFTW_MPI_PTRDIFF_T, 0, MPI_COMM_WORLD); MPI_Gather(local_no, rnk, FFTW_MPI_PTRDIFF_T, all_local_no, rnk, FFTW_MPI_PTRDIFF_T, 0, MPI_COMM_WORLD); MPI_Bcast(all_local_no, rnk*n_pes, FFTW_MPI_PTRDIFF_T, 0, MPI_COMM_WORLD); MPI_Gather(local_starto, rnk, FFTW_MPI_PTRDIFF_T, all_local_starto, rnk, FFTW_MPI_PTRDIFF_T, 0, MPI_COMM_WORLD); MPI_Bcast(all_local_starto, rnk*n_pes, FFTW_MPI_PTRDIFF_T, 0, MPI_COMM_WORLD); off = 0; for (i = 0; i < n_pes; ++i) { ptrdiff_t N = vn; for (j = 0; j < rnk; ++j) N *= all_local_ni[i * rnk + j]; isend_cnt[i] = N; isend_off[i] = off; off += N; } iNtot = off; all_local_in_alloc = 1; istrides[rnk - 1] = vn; for (j = rnk - 2; j >= 0; --j) istrides[j] = total_ni[j + 1] * istrides[j + 1]; off = 0; for (i = 0; i < n_pes; ++i) { ptrdiff_t N = vn; for (j = 0; j < rnk; ++j) N *= all_local_no[i * rnk + j]; orecv_cnt[i] = N; orecv_off[i] = off; off += N; } oNtot = off; all_local_out_alloc = 1; ostrides[rnk - 1] = vn; for (j = rnk - 2; j >= 0; --j) ostrides[j] = total_no[j + 1] * ostrides[j + 1]; } static void copy_block_out(const bench_real *in, int rnk, ptrdiff_t *n, ptrdiff_t *start, ptrdiff_t is, ptrdiff_t *os, ptrdiff_t vn, bench_real *out) { ptrdiff_t i; if (rnk == 0) { for (i = 0; i < vn; ++i) out[i] = in[i]; } else if (rnk == 1) { /* this case is just an optimization */ ptrdiff_t j; out += start[0] * os[0]; for (j = 0; j < n[0]; ++j) { for (i = 0; i < vn; ++i) out[i] = in[i]; in += is; out += os[0]; } } else { /* we should do n[0] for locality, but this way is simpler to code */ for (i = 0; i < n[rnk - 1]; ++i) copy_block_out(in + i * is, rnk - 1, n, start, is * n[rnk - 1], os, vn, out + (start[rnk - 1] + i) * os[rnk - 1]); } } static void copy_block_in(bench_real *in, int rnk, ptrdiff_t *n, ptrdiff_t *start, ptrdiff_t is, ptrdiff_t *os, ptrdiff_t vn, const bench_real *out) { ptrdiff_t i; if (rnk == 0) { for (i = 0; i < vn; ++i) in[i] = out[i]; } else if (rnk == 1) { /* this case is just an optimization */ ptrdiff_t j; out += start[0] * os[0]; for (j = 0; j < n[0]; ++j) { for (i = 0; i < vn; ++i) in[i] = out[i]; in += is; out += os[0]; } } else { /* we should do n[0] for locality, but this way is simpler to code */ for (i = 0; i < n[rnk - 1]; ++i) copy_block_in(in + i * is, rnk - 1, n, start, is * n[rnk - 1], os, vn, out + (start[rnk - 1] + i) * os[rnk - 1]); } } static void do_scatter_in(bench_real *in) { bench_real *ali; int i; if (all_local_in_alloc) { bench_free(all_local_in); all_local_in = (bench_real*) bench_malloc(iNtot*sizeof(bench_real)); all_local_in_alloc = 0; } ali = all_local_in; for (i = 0; i < n_pes; ++i) { copy_block_in(ali, rnk, all_local_ni + i * rnk, all_local_starti + i * rnk, vn, istrides, vn, in); ali += isend_cnt[i]; } MPI_Scatterv(all_local_in, isend_cnt, isend_off, BENCH_MPI_TYPE, local_in, isend_cnt[my_pe], BENCH_MPI_TYPE, 0, MPI_COMM_WORLD); } static void do_gather_out(bench_real *out) { bench_real *alo; int i; if (all_local_out_alloc) { bench_free(all_local_out); all_local_out = (bench_real*) bench_malloc(oNtot*sizeof(bench_real)); all_local_out_alloc = 0; } MPI_Gatherv(local_out, orecv_cnt[my_pe], BENCH_MPI_TYPE, all_local_out, orecv_cnt, orecv_off, BENCH_MPI_TYPE, 0, MPI_COMM_WORLD); MPI_Bcast(all_local_out, oNtot, BENCH_MPI_TYPE, 0, MPI_COMM_WORLD); alo = all_local_out; for (i = 0; i < n_pes; ++i) { copy_block_out(alo, rnk, all_local_no + i * rnk, all_local_starto + i * rnk, vn, ostrides, vn, out); alo += orecv_cnt[i]; } } static void alloc_local(ptrdiff_t nreal, int inplace) { bench_free(local_in); if (local_out != local_in) bench_free(local_out); local_in = local_out = 0; if (nreal > 0) { ptrdiff_t i; local_in = (bench_real*) bench_malloc(nreal * sizeof(bench_real)); if (inplace) local_out = local_in; else local_out = (bench_real*) bench_malloc(nreal * sizeof(bench_real)); for (i = 0; i < nreal; ++i) local_in[i] = local_out[i] = 0.0; } } void after_problem_rcopy_from(bench_problem *p, bench_real *ri) { UNUSED(p); do_scatter_in(ri); if (plan_scramble_in) FFTW(execute)(plan_scramble_in); } void after_problem_rcopy_to(bench_problem *p, bench_real *ro) { UNUSED(p); if (plan_unscramble_out) FFTW(execute)(plan_unscramble_out); do_gather_out(ro); } void after_problem_ccopy_from(bench_problem *p, bench_real *ri, bench_real *ii) { UNUSED(ii); after_problem_rcopy_from(p, ri); } void after_problem_ccopy_to(bench_problem *p, bench_real *ro, bench_real *io) { UNUSED(io); after_problem_rcopy_to(p, ro); } void after_problem_hccopy_from(bench_problem *p, bench_real *ri, bench_real *ii) { UNUSED(ii); after_problem_rcopy_from(p, ri); } void after_problem_hccopy_to(bench_problem *p, bench_real *ro, bench_real *io) { UNUSED(io); after_problem_rcopy_to(p, ro); } static FFTW(plan) mkplan_transpose_local(ptrdiff_t nx, ptrdiff_t ny, ptrdiff_t vn, bench_real *in, bench_real *out) { FFTW(iodim64) hdims[3]; FFTW(r2r_kind) k[3]; FFTW(plan) pln; hdims[0].n = nx; hdims[0].is = ny * vn; hdims[0].os = vn; hdims[1].n = ny; hdims[1].is = vn; hdims[1].os = nx * vn; hdims[2].n = vn; hdims[2].is = 1; hdims[2].os = 1; k[0] = k[1] = k[2] = FFTW_R2HC; pln = FFTW(plan_guru64_r2r)(0, 0, 3, hdims, in, out, k, FFTW_ESTIMATE); BENCH_ASSERT(pln != 0); return pln; } static int tensor_rowmajor_transposedp(bench_tensor *t) { bench_iodim *d; int i; BENCH_ASSERT(FINITE_RNK(t->rnk)); if (t->rnk < 2) return 0; d = t->dims; if (d[0].is != d[1].is * d[1].n || d[0].os != d[1].is || d[1].os != d[0].os * d[0].n) return 0; if (t->rnk > 2 && d[1].is != d[2].is * d[2].n) return 0; for (i = 2; i + 1 < t->rnk; ++i) { d = t->dims + i; if (d[0].is != d[1].is * d[1].n || d[0].os != d[1].os * d[1].n) return 0; } if (t->rnk > 2 && t->dims[t->rnk-1].is != t->dims[t->rnk-1].os) return 0; return 1; } static int tensor_contiguousp(bench_tensor *t, int s) { return (t->dims[t->rnk-1].is == s && ((tensor_rowmajorp(t) && t->dims[t->rnk-1].is == t->dims[t->rnk-1].os) || tensor_rowmajor_transposedp(t))); } static FFTW(plan) mkplan_complex(bench_problem *p, unsigned flags) { FFTW(plan) pln = 0; int i; ptrdiff_t ntot; vn = p->vecsz->rnk == 1 ? p->vecsz->dims[0].n : 1; if (p->sz->rnk < 1 || p->split || !tensor_contiguousp(p->sz, vn) || tensor_rowmajor_transposedp(p->sz) || p->vecsz->rnk > 1 || (p->vecsz->rnk == 1 && (p->vecsz->dims[0].is != 1 || p->vecsz->dims[0].os != 1))) return 0; alloc_rnk(p->sz->rnk); for (i = 0; i < rnk; ++i) { total_ni[i] = total_no[i] = p->sz->dims[i].n; local_ni[i] = local_no[i] = total_ni[i]; local_starti[i] = local_starto[i] = 0; } if (rnk > 1) { ptrdiff_t n, start, nT, startT; ntot = FFTW(mpi_local_size_many_transposed) (p->sz->rnk, total_ni, vn, FFTW_MPI_DEFAULT_BLOCK, FFTW_MPI_DEFAULT_BLOCK, MPI_COMM_WORLD, &n, &start, &nT, &startT); if (flags & FFTW_MPI_TRANSPOSED_IN) { local_ni[1] = nT; local_starti[1] = startT; } else { local_ni[0] = n; local_starti[0] = start; } if (flags & FFTW_MPI_TRANSPOSED_OUT) { local_no[1] = nT; local_starto[1] = startT; } else { local_no[0] = n; local_starto[0] = start; } } else if (rnk == 1) { ntot = FFTW(mpi_local_size_many_1d) (total_ni[0], vn, MPI_COMM_WORLD, p->sign, flags, local_ni, local_starti, local_no, local_starto); } alloc_local(ntot * 2, p->in == p->out); pln = FFTW(mpi_plan_many_dft)(p->sz->rnk, total_ni, vn, FFTW_MPI_DEFAULT_BLOCK, FFTW_MPI_DEFAULT_BLOCK, (FFTW(complex) *) local_in, (FFTW(complex) *) local_out, MPI_COMM_WORLD, p->sign, flags); vn *= 2; if (rnk > 1) { ptrdiff_t nrest = 1; for (i = 2; i < rnk; ++i) nrest *= p->sz->dims[i].n; if (flags & FFTW_MPI_TRANSPOSED_IN) plan_scramble_in = mkplan_transpose_local( p->sz->dims[0].n, local_ni[1], vn * nrest, local_in, local_in); if (flags & FFTW_MPI_TRANSPOSED_OUT) plan_unscramble_out = mkplan_transpose_local( local_no[1], p->sz->dims[0].n, vn * nrest, local_out, local_out); } return pln; } static int tensor_real_contiguousp(bench_tensor *t, int sign, int s) { return (t->dims[t->rnk-1].is == s && ((tensor_real_rowmajorp(t, sign, 1) && t->dims[t->rnk-1].is == t->dims[t->rnk-1].os))); } static FFTW(plan) mkplan_real(bench_problem *p, unsigned flags) { FFTW(plan) pln = 0; int i; ptrdiff_t ntot; vn = p->vecsz->rnk == 1 ? p->vecsz->dims[0].n : 1; if (p->sz->rnk < 2 || p->split || !tensor_real_contiguousp(p->sz, p->sign, vn) || tensor_rowmajor_transposedp(p->sz) || p->vecsz->rnk > 1 || (p->vecsz->rnk == 1 && (p->vecsz->dims[0].is != 1 || p->vecsz->dims[0].os != 1))) return 0; alloc_rnk(p->sz->rnk); for (i = 0; i < rnk; ++i) { total_ni[i] = total_no[i] = p->sz->dims[i].n; local_ni[i] = local_no[i] = total_ni[i]; local_starti[i] = local_starto[i] = 0; } local_ni[rnk-1] = local_no[rnk-1] = total_ni[rnk-1] = total_no[rnk-1] = p->sz->dims[rnk-1].n / 2 + 1; { ptrdiff_t n, start, nT, startT; ntot = FFTW(mpi_local_size_many_transposed) (p->sz->rnk, total_ni, vn, FFTW_MPI_DEFAULT_BLOCK, FFTW_MPI_DEFAULT_BLOCK, MPI_COMM_WORLD, &n, &start, &nT, &startT); if (flags & FFTW_MPI_TRANSPOSED_IN) { local_ni[1] = nT; local_starti[1] = startT; } else { local_ni[0] = n; local_starti[0] = start; } if (flags & FFTW_MPI_TRANSPOSED_OUT) { local_no[1] = nT; local_starto[1] = startT; } else { local_no[0] = n; local_starto[0] = start; } } alloc_local(ntot * 2, p->in == p->out); total_ni[rnk - 1] = p->sz->dims[rnk - 1].n; if (p->sign < 0) pln = FFTW(mpi_plan_many_dft_r2c)(p->sz->rnk, total_ni, vn, FFTW_MPI_DEFAULT_BLOCK, FFTW_MPI_DEFAULT_BLOCK, local_in, (FFTW(complex) *) local_out, MPI_COMM_WORLD, flags); else pln = FFTW(mpi_plan_many_dft_c2r)(p->sz->rnk, total_ni, vn, FFTW_MPI_DEFAULT_BLOCK, FFTW_MPI_DEFAULT_BLOCK, (FFTW(complex) *) local_in, local_out, MPI_COMM_WORLD, flags); total_ni[rnk - 1] = p->sz->dims[rnk - 1].n / 2 + 1; vn *= 2; { ptrdiff_t nrest = 1; for (i = 2; i < rnk; ++i) nrest *= total_ni[i]; if (flags & FFTW_MPI_TRANSPOSED_IN) plan_scramble_in = mkplan_transpose_local( total_ni[0], local_ni[1], vn * nrest, local_in, local_in); if (flags & FFTW_MPI_TRANSPOSED_OUT) plan_unscramble_out = mkplan_transpose_local( local_no[1], total_ni[0], vn * nrest, local_out, local_out); } return pln; } static FFTW(plan) mkplan_transpose(bench_problem *p, unsigned flags) { ptrdiff_t ntot, nx, ny; int ix=0, iy=1, i; const bench_iodim *d = p->vecsz->dims; FFTW(plan) pln; if (p->vecsz->rnk == 3) { for (i = 0; i < 3; ++i) if (d[i].is == 1 && d[i].os == 1) { vn = d[i].n; ix = (i + 1) % 3; iy = (i + 2) % 3; break; } if (i == 3) return 0; } else { vn = 1; ix = 0; iy = 1; } if (d[ix].is == d[iy].n * vn && d[ix].os == vn && d[iy].os == d[ix].n * vn && d[iy].is == vn) { nx = d[ix].n; ny = d[iy].n; } else if (d[iy].is == d[ix].n * vn && d[iy].os == vn && d[ix].os == d[iy].n * vn && d[ix].is == vn) { nx = d[iy].n; ny = d[ix].n; } else return 0; alloc_rnk(2); ntot = vn * FFTW(mpi_local_size_2d_transposed)(nx, ny, MPI_COMM_WORLD, &local_ni[0], &local_starti[0], &local_no[0], &local_starto[0]); local_ni[1] = ny; local_starti[1] = 0; local_no[1] = nx; local_starto[1] = 0; total_ni[0] = nx; total_ni[1] = ny; total_no[1] = nx; total_no[0] = ny; alloc_local(ntot, p->in == p->out); pln = FFTW(mpi_plan_many_transpose)(nx, ny, vn, FFTW_MPI_DEFAULT_BLOCK, FFTW_MPI_DEFAULT_BLOCK, local_in, local_out, MPI_COMM_WORLD, flags); if (flags & FFTW_MPI_TRANSPOSED_IN) plan_scramble_in = mkplan_transpose_local(local_ni[0], ny, vn, local_in, local_in); if (flags & FFTW_MPI_TRANSPOSED_OUT) plan_unscramble_out = mkplan_transpose_local (nx, local_no[0], vn, local_out, local_out); #if 0 if (pln && vn == 1) { int i, j; bench_real *ri = (bench_real *) p->in; bench_real *ro = (bench_real *) p->out; if (!ri || !ro) return pln; setup_gather_scatter(); for (i = 0; i < nx * ny; ++i) ri[i] = i; after_problem_rcopy_from(p, ri); FFTW(execute)(pln); after_problem_rcopy_to(p, ro); if (my_pe == 0) { for (i = 0; i < nx; ++i) { for (j = 0; j < ny; ++j) printf(" %3g", ro[j * nx + i]); printf("\n"); } } } #endif return pln; } static FFTW(plan) mkplan_r2r(bench_problem *p, unsigned flags) { FFTW(plan) pln = 0; int i; ptrdiff_t ntot; FFTW(r2r_kind) *k; if ((p->sz->rnk == 0 || (p->sz->rnk == 1 && p->sz->dims[0].n == 1)) && p->vecsz->rnk >= 2 && p->vecsz->rnk <= 3) return mkplan_transpose(p, flags); vn = p->vecsz->rnk == 1 ? p->vecsz->dims[0].n : 1; if (p->sz->rnk < 1 || p->split || !tensor_contiguousp(p->sz, vn) || tensor_rowmajor_transposedp(p->sz) || p->vecsz->rnk > 1 || (p->vecsz->rnk == 1 && (p->vecsz->dims[0].is != 1 || p->vecsz->dims[0].os != 1))) return 0; alloc_rnk(p->sz->rnk); for (i = 0; i < rnk; ++i) { total_ni[i] = total_no[i] = p->sz->dims[i].n; local_ni[i] = local_no[i] = total_ni[i]; local_starti[i] = local_starto[i] = 0; } if (rnk > 1) { ptrdiff_t n, start, nT, startT; ntot = FFTW(mpi_local_size_many_transposed) (p->sz->rnk, total_ni, vn, FFTW_MPI_DEFAULT_BLOCK, FFTW_MPI_DEFAULT_BLOCK, MPI_COMM_WORLD, &n, &start, &nT, &startT); if (flags & FFTW_MPI_TRANSPOSED_IN) { local_ni[1] = nT; local_starti[1] = startT; } else { local_ni[0] = n; local_starti[0] = start; } if (flags & FFTW_MPI_TRANSPOSED_OUT) { local_no[1] = nT; local_starto[1] = startT; } else { local_no[0] = n; local_starto[0] = start; } } else if (rnk == 1) { ntot = FFTW(mpi_local_size_many_1d) (total_ni[0], vn, MPI_COMM_WORLD, p->sign, flags, local_ni, local_starti, local_no, local_starto); } alloc_local(ntot, p->in == p->out); k = (FFTW(r2r_kind) *) bench_malloc(sizeof(FFTW(r2r_kind)) * p->sz->rnk); for (i = 0; i < p->sz->rnk; ++i) switch (p->k[i]) { case R2R_R2HC: k[i] = FFTW_R2HC; break; case R2R_HC2R: k[i] = FFTW_HC2R; break; case R2R_DHT: k[i] = FFTW_DHT; break; case R2R_REDFT00: k[i] = FFTW_REDFT00; break; case R2R_REDFT01: k[i] = FFTW_REDFT01; break; case R2R_REDFT10: k[i] = FFTW_REDFT10; break; case R2R_REDFT11: k[i] = FFTW_REDFT11; break; case R2R_RODFT00: k[i] = FFTW_RODFT00; break; case R2R_RODFT01: k[i] = FFTW_RODFT01; break; case R2R_RODFT10: k[i] = FFTW_RODFT10; break; case R2R_RODFT11: k[i] = FFTW_RODFT11; break; default: BENCH_ASSERT(0); } pln = FFTW(mpi_plan_many_r2r)(p->sz->rnk, total_ni, vn, FFTW_MPI_DEFAULT_BLOCK, FFTW_MPI_DEFAULT_BLOCK, local_in, local_out, MPI_COMM_WORLD, k, flags); bench_free(k); if (rnk > 1) { ptrdiff_t nrest = 1; for (i = 2; i < rnk; ++i) nrest *= p->sz->dims[i].n; if (flags & FFTW_MPI_TRANSPOSED_IN) plan_scramble_in = mkplan_transpose_local( p->sz->dims[0].n, local_ni[1], vn * nrest, local_in, local_in); if (flags & FFTW_MPI_TRANSPOSED_OUT) plan_unscramble_out = mkplan_transpose_local( local_no[1], p->sz->dims[0].n, vn * nrest, local_out, local_out); } return pln; } FFTW(plan) mkplan(bench_problem *p, unsigned flags) { FFTW(plan) pln = 0; FFTW(destroy_plan)(plan_scramble_in); plan_scramble_in = 0; FFTW(destroy_plan)(plan_unscramble_out); plan_unscramble_out = 0; if (p->scrambled_in) { if (p->sz->rnk == 1 && p->sz->dims[0].n != 1) flags |= FFTW_MPI_SCRAMBLED_IN; else flags |= FFTW_MPI_TRANSPOSED_IN; } if (p->scrambled_out) { if (p->sz->rnk == 1 && p->sz->dims[0].n != 1) flags |= FFTW_MPI_SCRAMBLED_OUT; else flags |= FFTW_MPI_TRANSPOSED_OUT; } switch (p->kind) { case PROBLEM_COMPLEX: pln =mkplan_complex(p, flags); break; case PROBLEM_REAL: pln = mkplan_real(p, flags); break; case PROBLEM_R2R: pln = mkplan_r2r(p, flags); break; default: BENCH_ASSERT(0); } if (pln) setup_gather_scatter(); return pln; } void main_init(int *argc, char ***argv) { #ifdef HAVE_SMP # if MPI_VERSION >= 2 /* for MPI_Init_thread */ int provided; MPI_Init_thread(argc, argv, MPI_THREAD_FUNNELED, &provided); threads_ok = provided >= MPI_THREAD_FUNNELED; # else MPI_Init(argc, argv); threads_ok = 0; # endif #else MPI_Init(argc, argv); #endif MPI_Comm_rank(MPI_COMM_WORLD, &my_pe); MPI_Comm_size(MPI_COMM_WORLD, &n_pes); if (my_pe != 0) verbose = -999; no_speed_allocation = 1; /* so we can benchmark transforms > memory */ always_pad_real = 1; /* out-of-place real transforms are padded */ isend_cnt = (int *) bench_malloc(sizeof(int) * n_pes); isend_off = (int *) bench_malloc(sizeof(int) * n_pes); orecv_cnt = (int *) bench_malloc(sizeof(int) * n_pes); orecv_off = (int *) bench_malloc(sizeof(int) * n_pes); /* init_threads must be called before any other FFTW function, including mpi_init, because it has to register the threads hooks before the planner is initalized */ #ifdef HAVE_SMP if (threads_ok) { BENCH_ASSERT(FFTW(init_threads)()); } #endif FFTW(mpi_init)(); } void initial_cleanup(void) { alloc_rnk(0); alloc_local(0, 0); bench_free(all_local_in); all_local_in = 0; bench_free(all_local_out); all_local_out = 0; bench_free(isend_off); isend_off = 0; bench_free(isend_cnt); isend_cnt = 0; bench_free(orecv_off); orecv_off = 0; bench_free(orecv_cnt); orecv_cnt = 0; FFTW(destroy_plan)(plan_scramble_in); plan_scramble_in = 0; FFTW(destroy_plan)(plan_unscramble_out); plan_unscramble_out = 0; } void final_cleanup(void) { MPI_Finalize(); } void bench_exit(int status) { MPI_Abort(MPI_COMM_WORLD, status); } double bench_cost_postprocess(double cost) { double cost_max; MPI_Allreduce(&cost, &cost_max, 1, MPI_DOUBLE, MPI_MAX, MPI_COMM_WORLD); return cost_max; } int import_wisdom(FILE *f) { int success = 1, sall; if (my_pe == 0) success = FFTW(import_wisdom_from_file)(f); FFTW(mpi_broadcast_wisdom)(MPI_COMM_WORLD); MPI_Allreduce(&success, &sall, 1, MPI_INT, MPI_LAND, MPI_COMM_WORLD); return sall; } void export_wisdom(FILE *f) { FFTW(mpi_gather_wisdom)(MPI_COMM_WORLD); if (my_pe == 0) FFTW(export_wisdom_to_file)(f); } fftw-3.3.3/mpi/testsched.c0000644000175000017500000004067212054410050012326 00000000000000/* * Copyright (c) 2003, 2007-11 Matteo Frigo * Copyright (c) 1999-2003, 2007-8 Massachusetts Institute of Technology * * This program is free software; you can redistribute it and/or modify * it under the terms of the GNU General Public License as published by * the Free Software Foundation; either version 2 of the License, or * (at your option) any later version. * * This program is distributed in the hope that it will be useful, * but WITHOUT ANY WARRANTY; without even the implied warranty of * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the * GNU General Public License for more details. * * You should have received a copy of the GNU General Public License * along with this program; if not, write to the Free Software * Foundation, Inc., 51 Franklin Street, Fifth Floor, Boston, MA 02110-1301 USA * */ /**********************************************************************/ /* This is a modified and combined version of the sched.c and test_sched.c files shipped with FFTW 2, written to implement and test various all-to-all communications scheduling patterns. It is not used in FFTW 3, but I keep it around in case we ever want to play with this again or to change algorithms. In particular, I used it to implement and test the fill1_comm_sched routine in transpose-pairwise.c, which allows us to create a schedule for one process at a time and is much more compact than the FFTW 2 code. Note that the scheduling algorithm is somewhat modified from that of FFTW 2. Originally, I thought that one "stall" in the schedule was unavoidable for odd numbers of processes, since this is the case for the soccer-timetabling problem. However, because of the self-communication step, we can use the self-communication to fill in the stalls. (Thanks to Ralf Wildenhues for pointing this out.) This greatly simplifies the process re-sorting algorithm. */ /**********************************************************************/ #include #include /* This file contains routines to compute communications schedules for all-to-all communications (complete exchanges) that are performed in-place. (That is, the block that processor x sends to processor y gets replaced on processor x by a block received from processor y.) A schedule, int **sched, is a two-dimensional array where sched[pe][i] is the processor that pe expects to exchange a message with on the i-th step of the exchange. sched[pe][i] == -1 for the i after the last exchange scheduled on pe. Here, processors (pe's, for processing elements), are numbered from 0 to npes-1. There are a couple of constraints that a schedule should satisfy (besides the obvious one that every processor has to communicate with every other processor exactly once). * First, and most importantly, there must be no deadlocks. * Second, we would like to overlap communications as much as possible, so that all exchanges occur in parallel. It turns out that perfect overlap is possible for all number of processes (npes). It turns out that this scheduling problem is actually well-studied, and good solutions are known. The problem is known as a "time-tabling" problem, and is specifically the problem of scheduling a sports competition (where n teams must compete exactly once with every other team). The problem is discussed and algorithms are presented in: [1] J. A. M. Schreuder, "Constructing Timetables for Sport Competitions," Mathematical Programming Study 13, pp. 58-67 (1980). [2] A. Schaerf, "Scheduling Sport Tournaments using Constraint Logic Programming," Proc. of 12th Europ. Conf. on Artif. Intell. (ECAI-96), pp. 634-639 (Budapest 1996). http://hermes.dis.uniromal.it/~aschaerf/publications.html (These people actually impose a lot of additional constraints that we don't care about, so they are solving harder problems. [1] gives a simple enough algorithm for our purposes, though.) In the timetabling problem, N teams can all play one another in N-1 steps if N is even, and N steps if N is odd. Here, however, there is a "self-communication" step (a team must also "play itself") and so we can always make an optimal N-step schedule regardless of N. However, we have to do more: for a particular processor, the communications schedule must be sorted in ascending or descending order of processor index. (This is necessary so that the data coming in for the transpose does not overwrite data that will be sent later; for that processor the incoming and outgoing blocks are of different non-zero sizes.) Fortunately, because the schedule is stall free, each parallel step of the schedule is independent of every other step, and we can reorder the steps arbitrarily to achieve any desired order on a particular process. */ void free_comm_schedule(int **sched, int npes) { if (sched) { int i; for (i = 0; i < npes; ++i) free(sched[i]); free(sched); } } void empty_comm_schedule(int **sched, int npes) { int i; for (i = 0; i < npes; ++i) sched[i][0] = -1; } extern void fill_comm_schedule(int **sched, int npes); /* Create a new communications schedule for a given number of processors. The schedule is initialized to a deadlock-free, maximum overlap schedule. Returns NULL on an error (may print a message to stderr if there is a program bug detected). */ int **make_comm_schedule(int npes) { int **sched; int i; sched = (int **) malloc(sizeof(int *) * npes); if (!sched) return NULL; for (i = 0; i < npes; ++i) sched[i] = NULL; for (i = 0; i < npes; ++i) { sched[i] = (int *) malloc(sizeof(int) * 10 * (npes + 1)); if (!sched[i]) { free_comm_schedule(sched,npes); return NULL; } } empty_comm_schedule(sched,npes); fill_comm_schedule(sched,npes); if (!check_comm_schedule(sched,npes)) { free_comm_schedule(sched,npes); return NULL; } return sched; } static void add_dest_to_comm_schedule(int **sched, int pe, int dest) { int i; for (i = 0; sched[pe][i] != -1; ++i) ; sched[pe][i] = dest; sched[pe][i+1] = -1; } static void add_pair_to_comm_schedule(int **sched, int pe1, int pe2) { add_dest_to_comm_schedule(sched, pe1, pe2); if (pe1 != pe2) add_dest_to_comm_schedule(sched, pe2, pe1); } /* Simplification of algorithm presented in [1] (we have fewer constraints). Produces a perfect schedule (npes steps). */ void fill_comm_schedule(int **sched, int npes) { int pe, i, n; if (npes % 2 == 0) { n = npes; for (pe = 0; pe < npes; ++pe) add_pair_to_comm_schedule(sched,pe,pe); } else n = npes + 1; for (pe = 0; pe < n - 1; ++pe) { add_pair_to_comm_schedule(sched, pe, npes % 2 == 0 ? npes - 1 : pe); for (i = 1; i < n/2; ++i) { int pe_a, pe_b; pe_a = pe - i; if (pe_a < 0) pe_a += n - 1; pe_b = (pe + i) % (n - 1); add_pair_to_comm_schedule(sched,pe_a,pe_b); } } } /* given an array sched[npes], fills it with the communications schedule for process pe. */ void fill1_comm_sched(int *sched, int which_pe, int npes) { int pe, i, n, s = 0; if (npes % 2 == 0) { n = npes; sched[s++] = which_pe; } else n = npes + 1; for (pe = 0; pe < n - 1; ++pe) { if (npes % 2 == 0) { if (pe == which_pe) sched[s++] = npes - 1; else if (npes - 1 == which_pe) sched[s++] = pe; } else if (pe == which_pe) sched[s++] = pe; if (pe != which_pe && which_pe < n - 1) { i = (pe - which_pe + (n - 1)) % (n - 1); if (i < n/2) sched[s++] = (pe + i) % (n - 1); i = (which_pe - pe + (n - 1)) % (n - 1); if (i < n/2) sched[s++] = (pe - i + (n - 1)) % (n - 1); } } if (s != npes) { fprintf(stderr, "bug in fill1_com_schedule (%d, %d/%d)\n", s, which_pe, npes); exit(EXIT_FAILURE); } } /* sort the communication schedule sched for npes so that the schedule on process sortpe is ascending or descending (!ascending). */ static void sort1_comm_sched(int *sched, int npes, int sortpe, int ascending) { int *sortsched, i; sortsched = (int *) malloc(npes * sizeof(int) * 2); fill1_comm_sched(sortsched, sortpe, npes); if (ascending) for (i = 0; i < npes; ++i) sortsched[npes + sortsched[i]] = sched[i]; else for (i = 0; i < npes; ++i) sortsched[2*npes - 1 - sortsched[i]] = sched[i]; for (i = 0; i < npes; ++i) sched[i] = sortsched[npes + i]; free(sortsched); } /* Below, we have various checks in case of bugs: */ /* check for deadlocks by simulating the schedule and looking for cycles in the dependency list; returns 0 if there are deadlocks (or other errors) */ static int check_schedule_deadlock(int **sched, int npes) { int *step, *depend, *visited, pe, pe2, period, done = 0; int counter = 0; /* step[pe] is the step in the schedule that a given pe is on */ step = (int *) malloc(sizeof(int) * npes); /* depend[pe] is the pe' that pe is currently waiting for a message from (-1 if none) */ depend = (int *) malloc(sizeof(int) * npes); /* visited[pe] tells whether we have visited the current pe already when we are looking for cycles. */ visited = (int *) malloc(sizeof(int) * npes); if (!step || !depend || !visited) { free(step); free(depend); free(visited); return 0; } for (pe = 0; pe < npes; ++pe) step[pe] = 0; while (!done) { ++counter; for (pe = 0; pe < npes; ++pe) depend[pe] = sched[pe][step[pe]]; /* now look for cycles in the dependencies with period > 2: */ for (pe = 0; pe < npes; ++pe) if (depend[pe] != -1) { for (pe2 = 0; pe2 < npes; ++pe2) visited[pe2] = 0; period = 0; pe2 = pe; do { visited[pe2] = period + 1; pe2 = depend[pe2]; period++; } while (pe2 != -1 && !visited[pe2]); if (pe2 == -1) { fprintf(stderr, "BUG: unterminated cycle in schedule!\n"); free(step); free(depend); free(visited); return 0; } if (period - (visited[pe2] - 1) > 2) { fprintf(stderr,"BUG: deadlock in schedule!\n"); free(step); free(depend); free(visited); return 0; } if (pe2 == pe) step[pe]++; } done = 1; for (pe = 0; pe < npes; ++pe) if (sched[pe][step[pe]] != -1) { done = 0; break; } } free(step); free(depend); free(visited); return (counter > 0 ? counter : 1); } /* sanity checks; prints message and returns 0 on failure. undocumented feature: the return value on success is actually the number of steps required for the schedule to complete, counting stalls. */ int check_comm_schedule(int **sched, int npes) { int pe, i, comm_pe; for (pe = 0; pe < npes; ++pe) { for (comm_pe = 0; comm_pe < npes; ++comm_pe) { for (i = 0; sched[pe][i] != -1 && sched[pe][i] != comm_pe; ++i) ; if (sched[pe][i] == -1) { fprintf(stderr,"BUG: schedule never sends message from " "%d to %d.\n",pe,comm_pe); return 0; /* never send message to comm_pe */ } } for (i = 0; sched[pe][i] != -1; ++i) ; if (i != npes) { fprintf(stderr,"BUG: schedule sends too many messages from " "%d\n",pe); return 0; } } return check_schedule_deadlock(sched,npes); } /* invert the order of all the schedules; this has no effect on its required properties. */ void invert_comm_schedule(int **sched, int npes) { int pe, i; for (pe = 0; pe < npes; ++pe) for (i = 0; i < npes/2; ++i) { int dummy = sched[pe][i]; sched[pe][i] = sched[pe][npes-1-i]; sched[pe][npes-1-i] = dummy; } } /* Sort the schedule for sort_pe in ascending order of processor index. Unfortunately, for odd npes (when schedule has a stall to begin with) this will introduce an extra stall due to the motion of the self-communication past a stall. We could fix this if it were really important. Actually, we don't get an extra stall when sort_pe == 0 or npes-1, which is sufficient for our purposes. */ void sort_comm_schedule(int **sched, int npes, int sort_pe) { int i,j,pe; /* Note that we can do this sort in O(npes) swaps because we know that the numbers we are sorting are just 0...npes-1. But we'll just do a bubble sort for simplicity here. */ for (i = 0; i < npes - 1; ++i) for (j = i + 1; j < npes; ++j) if (sched[sort_pe][i] > sched[sort_pe][j]) { for (pe = 0; pe < npes; ++pe) { int s = sched[pe][i]; sched[pe][i] = sched[pe][j]; sched[pe][j] = s; } } } /* print the schedule (for debugging purposes) */ void print_comm_schedule(int **sched, int npes) { int pe, i, width; if (npes < 10) width = 1; else if (npes < 100) width = 2; else width = 3; for (pe = 0; pe < npes; ++pe) { printf("pe %*d schedule:", width, pe); for (i = 0; sched[pe][i] != -1; ++i) printf(" %*d",width,sched[pe][i]); printf("\n"); } } int main(int argc, char **argv) { int **sched; int npes = -1, sortpe = -1, steps, i; if (argc >= 2) { npes = atoi(argv[1]); if (npes <= 0) { fprintf(stderr,"npes must be positive!"); return 1; } } if (argc >= 3) { sortpe = atoi(argv[2]); if (sortpe < 0 || sortpe >= npes) { fprintf(stderr,"sortpe must be between 0 and npes-1.\n"); return 1; } } if (npes != -1) { printf("Computing schedule for npes = %d:\n",npes); sched = make_comm_schedule(npes); if (!sched) { fprintf(stderr,"Out of memory!"); return 6; } if (steps = check_comm_schedule(sched,npes)) printf("schedule OK (takes %d steps to complete).\n", steps); else printf("schedule not OK.\n"); print_comm_schedule(sched, npes); if (sortpe != -1) { printf("\nRe-creating schedule for pe = %d...\n", sortpe); int *sched1 = (int*) malloc(sizeof(int) * npes); for (i = 0; i < npes; ++i) sched1[i] = -1; fill1_comm_sched(sched1, sortpe, npes); printf(" ="); for (i = 0; i < npes; ++i) printf(" %*d", npes < 10 ? 1 : (npes < 100 ? 2 : 3), sched1[i]); printf("\n"); printf("\nSorting schedule for sortpe = %d...\n", sortpe); sort_comm_schedule(sched,npes,sortpe); if (steps = check_comm_schedule(sched,npes)) printf("schedule OK (takes %d steps to complete).\n", steps); else printf("schedule not OK.\n"); print_comm_schedule(sched, npes); printf("\nInverting schedule...\n"); invert_comm_schedule(sched,npes); if (steps = check_comm_schedule(sched,npes)) printf("schedule OK (takes %d steps to complete).\n", steps); else printf("schedule not OK.\n"); print_comm_schedule(sched, npes); free_comm_schedule(sched,npes); free(sched1); } } else { printf("Doing infinite tests...\n"); for (npes = 1; ; ++npes) { int *sched1 = (int*) malloc(sizeof(int) * npes); printf("npes = %d...",npes); sched = make_comm_schedule(npes); if (!sched) { fprintf(stderr,"Out of memory!\n"); return 5; } for (sortpe = 0; sortpe < npes; ++sortpe) { empty_comm_schedule(sched,npes); fill_comm_schedule(sched,npes); if (!check_comm_schedule(sched,npes)) { fprintf(stderr, "\n -- fill error for sortpe = %d!\n",sortpe); return 2; } for (i = 0; i < npes; ++i) sched1[i] = -1; fill1_comm_sched(sched1, sortpe, npes); for (i = 0; i < npes; ++i) if (sched1[i] != sched[sortpe][i]) fprintf(stderr, "\n -- fill1 error for pe = %d!\n", sortpe); sort_comm_schedule(sched,npes,sortpe); if (!check_comm_schedule(sched,npes)) { fprintf(stderr, "\n -- sort error for sortpe = %d!\n",sortpe); return 3; } invert_comm_schedule(sched,npes); if (!check_comm_schedule(sched,npes)) { fprintf(stderr, "\n -- invert error for sortpe = %d!\n", sortpe); return 4; } } free_comm_schedule(sched,npes); printf("OK\n"); if (npes % 50 == 0) printf("(...Hit Ctrl-C to stop...)\n"); free(sched1); } } return 0; } fftw-3.3.3/mpi/f03api.sh0000644000175000017500000000266612054410050011613 00000000000000#! /bin/sh # Script to generate Fortran 2003 interface declarations for FFTW's MPI # interface from the fftw3-mpi.h header file. # This is designed so that the Fortran caller can do: # use, intrinsic :: iso_c_binding # implicit none # include 'fftw3-mpi.f03' # and then call the C FFTW MPI functions directly, with type checking. # # One caveat: because there is no standard way to conver MPI_Comm objects # from Fortran (= integer) to C (= opaque type), the Fortran interface # technically calls C wrapper functions (also auto-generated) which # call MPI_Comm_f2c to convert the communicators as needed. echo "! Generated automatically. DO NOT EDIT!" echo echo " include 'fftw3.f03'" echo # Extract constants perl -pe 's/#define +([A-Z0-9_]+) +\(([+-]?[0-9]+)U?\)/\n integer\(C_INTPTR_T\), parameter :: \1 = \2\n/g' < fftw3-mpi.h | grep 'integer(C_INTPTR_T)' perl -pe 'if (/#define +([A-Z0-9_]+) +\(([0-9]+)U? *<< *([0-9]+)\)/) { print "\n integer\(C_INT\), parameter :: $1 = ",$2 << $3,"\n"; }' < fftw3-mpi.h | grep 'integer(C_INT)' # Extract function declarations for p in $*; do if test "$p" = "d"; then p=""; fi echo cat < fftw_mpi_(...args...) # extern fftw_mpi_(...args...) # ... # with no line breaks within a given function. (It's too much work to # write a general parser, since we just have to handle FFTW's header files.) # Each declaration has at least one MPI_Comm argument. sub canonicalize_type { my($type); ($type) = @_; $type =~ s/ +/ /g; $type =~ s/^ //; $type =~ s/ $//; $type =~ s/([^\* ])\*/$1 \*/g; $type =~ s/double/R/; $type =~ s/fftw_([A-Za-z0-9_]+)/X(\1)/; return $type; } while (<>) { next if /^ *$/; if (/^ *extern +([a-zA-Z_0-9 ]+[ \*]) *fftw_mpi_([a-zA-Z_0-9]+) *\((.*)\) *$/) { $ret = &canonicalize_type($1); $name = $2; $args = $3; print "\n$ret XM(${name}_f03)("; $comma = ""; foreach $arg (split(/ *, */, $args)) { $arg =~ /^([a-zA-Z_0-9 ]+[ \*]) *([a-zA-Z_0-9]+) *$/; $argtype = &canonicalize_type($1); $argname = $2; print $comma; if ($argtype eq "MPI_Comm") { print "MPI_Fint f_$argname"; } else { print "$argtype $argname"; } $comma = ", "; } print ")\n{\n"; print " MPI_Comm "; $comma = ""; foreach $arg (split(/ *, */, $args)) { $arg =~ /^([a-zA-Z_0-9 ]+[ \*]) *([a-zA-Z_0-9]+) *$/; $argtype = &canonicalize_type($1); $argname = $2; if ($argtype eq "MPI_Comm") { print "$comma$argname"; $comma = ", "; } } print ";\n\n"; foreach $arg (split(/ *, */, $args)) { $arg =~ /^([a-zA-Z_0-9 ]+[ \*]) *([a-zA-Z_0-9]+) *$/; $argtype = &canonicalize_type($1); $argname = $2; if ($argtype eq "MPI_Comm") { print " $argname = MPI_Comm_f2c(f_$argname);\n"; } } $argnames = $args; $argnames =~ s/([a-zA-Z_0-9 ]+[ \*]) *([a-zA-Z_0-9]+) */$2/g; print " "; print "return " if ($ret ne "void"); print "XM($name)($argnames);\n}\n"; } } fftw-3.3.3/mpi/fftw3-mpi.f03.in0000644000175000017500000011024712054411104012724 00000000000000! Generated automatically. DO NOT EDIT! include 'fftw3.f03' integer(C_INTPTR_T), parameter :: FFTW_MPI_DEFAULT_BLOCK = 0 integer(C_INT), parameter :: FFTW_MPI_SCRAMBLED_IN = 134217728 integer(C_INT), parameter :: FFTW_MPI_SCRAMBLED_OUT = 268435456 integer(C_INT), parameter :: FFTW_MPI_TRANSPOSED_IN = 536870912 integer(C_INT), parameter :: FFTW_MPI_TRANSPOSED_OUT = 1073741824 type, bind(C) :: fftw_mpi_ddim integer(C_INTPTR_T) n, ib, ob end type fftw_mpi_ddim interface subroutine fftw_mpi_init() bind(C, name='fftw_mpi_init') import end subroutine fftw_mpi_init subroutine fftw_mpi_cleanup() bind(C, name='fftw_mpi_cleanup') import end subroutine fftw_mpi_cleanup integer(C_INTPTR_T) function fftw_mpi_local_size_many_transposed(rnk,n,howmany,block0,block1,comm,local_n0,local_0_start, & local_n1,local_1_start) & bind(C, name='fftw_mpi_local_size_many_transposed_f03') import integer(C_INT), value :: rnk integer(C_INTPTR_T), dimension(*), intent(in) :: n integer(C_INTPTR_T), value :: howmany integer(C_INTPTR_T), value :: block0 integer(C_INTPTR_T), value :: block1 integer(C_MPI_FINT), value :: comm integer(C_INTPTR_T), intent(out) :: local_n0 integer(C_INTPTR_T), intent(out) :: local_0_start integer(C_INTPTR_T), intent(out) :: local_n1 integer(C_INTPTR_T), intent(out) :: local_1_start end function fftw_mpi_local_size_many_transposed integer(C_INTPTR_T) function fftw_mpi_local_size_many(rnk,n,howmany,block0,comm,local_n0,local_0_start) & bind(C, name='fftw_mpi_local_size_many_f03') import integer(C_INT), value :: rnk integer(C_INTPTR_T), dimension(*), intent(in) :: n integer(C_INTPTR_T), value :: howmany integer(C_INTPTR_T), value :: block0 integer(C_MPI_FINT), value :: comm integer(C_INTPTR_T), intent(out) :: local_n0 integer(C_INTPTR_T), intent(out) :: local_0_start end function fftw_mpi_local_size_many integer(C_INTPTR_T) function fftw_mpi_local_size_transposed(rnk,n,comm,local_n0,local_0_start,local_n1,local_1_start) & bind(C, name='fftw_mpi_local_size_transposed_f03') import integer(C_INT), value :: rnk integer(C_INTPTR_T), dimension(*), intent(in) :: n integer(C_MPI_FINT), value :: comm integer(C_INTPTR_T), intent(out) :: local_n0 integer(C_INTPTR_T), intent(out) :: local_0_start integer(C_INTPTR_T), intent(out) :: local_n1 integer(C_INTPTR_T), intent(out) :: local_1_start end function fftw_mpi_local_size_transposed integer(C_INTPTR_T) function fftw_mpi_local_size(rnk,n,comm,local_n0,local_0_start) bind(C, name='fftw_mpi_local_size_f03') import integer(C_INT), value :: rnk integer(C_INTPTR_T), dimension(*), intent(in) :: n integer(C_MPI_FINT), value :: comm integer(C_INTPTR_T), intent(out) :: local_n0 integer(C_INTPTR_T), intent(out) :: local_0_start end function fftw_mpi_local_size integer(C_INTPTR_T) function fftw_mpi_local_size_many_1d(n0,howmany,comm,sign,flags,local_ni,local_i_start,local_no, & local_o_start) bind(C, name='fftw_mpi_local_size_many_1d_f03') import integer(C_INTPTR_T), value :: n0 integer(C_INTPTR_T), value :: howmany integer(C_MPI_FINT), value :: comm integer(C_INT), value :: sign integer(C_INT), value :: flags integer(C_INTPTR_T), intent(out) :: local_ni integer(C_INTPTR_T), intent(out) :: local_i_start integer(C_INTPTR_T), intent(out) :: local_no integer(C_INTPTR_T), intent(out) :: local_o_start end function fftw_mpi_local_size_many_1d integer(C_INTPTR_T) function fftw_mpi_local_size_1d(n0,comm,sign,flags,local_ni,local_i_start,local_no,local_o_start) & bind(C, name='fftw_mpi_local_size_1d_f03') import integer(C_INTPTR_T), value :: n0 integer(C_MPI_FINT), value :: comm integer(C_INT), value :: sign integer(C_INT), value :: flags integer(C_INTPTR_T), intent(out) :: local_ni integer(C_INTPTR_T), intent(out) :: local_i_start integer(C_INTPTR_T), intent(out) :: local_no integer(C_INTPTR_T), intent(out) :: local_o_start end function fftw_mpi_local_size_1d integer(C_INTPTR_T) function fftw_mpi_local_size_2d(n0,n1,comm,local_n0,local_0_start) & bind(C, name='fftw_mpi_local_size_2d_f03') import integer(C_INTPTR_T), value :: n0 integer(C_INTPTR_T), value :: n1 integer(C_MPI_FINT), value :: comm integer(C_INTPTR_T), intent(out) :: local_n0 integer(C_INTPTR_T), intent(out) :: local_0_start end function fftw_mpi_local_size_2d integer(C_INTPTR_T) function fftw_mpi_local_size_2d_transposed(n0,n1,comm,local_n0,local_0_start,local_n1,local_1_start) & bind(C, name='fftw_mpi_local_size_2d_transposed_f03') import integer(C_INTPTR_T), value :: n0 integer(C_INTPTR_T), value :: n1 integer(C_MPI_FINT), value :: comm integer(C_INTPTR_T), intent(out) :: local_n0 integer(C_INTPTR_T), intent(out) :: local_0_start integer(C_INTPTR_T), intent(out) :: local_n1 integer(C_INTPTR_T), intent(out) :: local_1_start end function fftw_mpi_local_size_2d_transposed integer(C_INTPTR_T) function fftw_mpi_local_size_3d(n0,n1,n2,comm,local_n0,local_0_start) & bind(C, name='fftw_mpi_local_size_3d_f03') import integer(C_INTPTR_T), value :: n0 integer(C_INTPTR_T), value :: n1 integer(C_INTPTR_T), value :: n2 integer(C_MPI_FINT), value :: comm integer(C_INTPTR_T), intent(out) :: local_n0 integer(C_INTPTR_T), intent(out) :: local_0_start end function fftw_mpi_local_size_3d integer(C_INTPTR_T) function fftw_mpi_local_size_3d_transposed(n0,n1,n2,comm,local_n0,local_0_start,local_n1,local_1_start) & bind(C, name='fftw_mpi_local_size_3d_transposed_f03') import integer(C_INTPTR_T), value :: n0 integer(C_INTPTR_T), value :: n1 integer(C_INTPTR_T), value :: n2 integer(C_MPI_FINT), value :: comm integer(C_INTPTR_T), intent(out) :: local_n0 integer(C_INTPTR_T), intent(out) :: local_0_start integer(C_INTPTR_T), intent(out) :: local_n1 integer(C_INTPTR_T), intent(out) :: local_1_start end function fftw_mpi_local_size_3d_transposed type(C_PTR) function fftw_mpi_plan_many_transpose(n0,n1,howmany,block0,block1,in,out,comm,flags) & bind(C, name='fftw_mpi_plan_many_transpose_f03') import integer(C_INTPTR_T), value :: n0 integer(C_INTPTR_T), value :: n1 integer(C_INTPTR_T), value :: howmany integer(C_INTPTR_T), value :: block0 integer(C_INTPTR_T), value :: block1 real(C_DOUBLE), dimension(*), intent(out) :: in real(C_DOUBLE), dimension(*), intent(out) :: out integer(C_MPI_FINT), value :: comm integer(C_INT), value :: flags end function fftw_mpi_plan_many_transpose type(C_PTR) function fftw_mpi_plan_transpose(n0,n1,in,out,comm,flags) bind(C, name='fftw_mpi_plan_transpose_f03') import integer(C_INTPTR_T), value :: n0 integer(C_INTPTR_T), value :: n1 real(C_DOUBLE), dimension(*), intent(out) :: in real(C_DOUBLE), dimension(*), intent(out) :: out integer(C_MPI_FINT), value :: comm integer(C_INT), value :: flags end function fftw_mpi_plan_transpose type(C_PTR) function fftw_mpi_plan_many_dft(rnk,n,howmany,block,tblock,in,out,comm,sign,flags) & bind(C, name='fftw_mpi_plan_many_dft_f03') import integer(C_INT), value :: rnk integer(C_INTPTR_T), dimension(*), intent(in) :: n integer(C_INTPTR_T), value :: howmany integer(C_INTPTR_T), value :: block integer(C_INTPTR_T), value :: tblock complex(C_DOUBLE_COMPLEX), dimension(*), intent(out) :: in complex(C_DOUBLE_COMPLEX), dimension(*), intent(out) :: out integer(C_MPI_FINT), value :: comm integer(C_INT), value :: sign integer(C_INT), value :: flags end function fftw_mpi_plan_many_dft type(C_PTR) function fftw_mpi_plan_dft(rnk,n,in,out,comm,sign,flags) bind(C, name='fftw_mpi_plan_dft_f03') import integer(C_INT), value :: rnk integer(C_INTPTR_T), dimension(*), intent(in) :: n complex(C_DOUBLE_COMPLEX), dimension(*), intent(out) :: in complex(C_DOUBLE_COMPLEX), dimension(*), intent(out) :: out integer(C_MPI_FINT), value :: comm integer(C_INT), value :: sign integer(C_INT), value :: flags end function fftw_mpi_plan_dft type(C_PTR) function fftw_mpi_plan_dft_1d(n0,in,out,comm,sign,flags) bind(C, name='fftw_mpi_plan_dft_1d_f03') import integer(C_INTPTR_T), value :: n0 complex(C_DOUBLE_COMPLEX), dimension(*), intent(out) :: in complex(C_DOUBLE_COMPLEX), dimension(*), intent(out) :: out integer(C_MPI_FINT), value :: comm integer(C_INT), value :: sign integer(C_INT), value :: flags end function fftw_mpi_plan_dft_1d type(C_PTR) function fftw_mpi_plan_dft_2d(n0,n1,in,out,comm,sign,flags) bind(C, name='fftw_mpi_plan_dft_2d_f03') import integer(C_INTPTR_T), value :: n0 integer(C_INTPTR_T), value :: n1 complex(C_DOUBLE_COMPLEX), dimension(*), intent(out) :: in complex(C_DOUBLE_COMPLEX), dimension(*), intent(out) :: out integer(C_MPI_FINT), value :: comm integer(C_INT), value :: sign integer(C_INT), value :: flags end function fftw_mpi_plan_dft_2d type(C_PTR) function fftw_mpi_plan_dft_3d(n0,n1,n2,in,out,comm,sign,flags) bind(C, name='fftw_mpi_plan_dft_3d_f03') import integer(C_INTPTR_T), value :: n0 integer(C_INTPTR_T), value :: n1 integer(C_INTPTR_T), value :: n2 complex(C_DOUBLE_COMPLEX), dimension(*), intent(out) :: in complex(C_DOUBLE_COMPLEX), dimension(*), intent(out) :: out integer(C_MPI_FINT), value :: comm integer(C_INT), value :: sign integer(C_INT), value :: flags end function fftw_mpi_plan_dft_3d type(C_PTR) function fftw_mpi_plan_many_r2r(rnk,n,howmany,iblock,oblock,in,out,comm,kind,flags) & bind(C, name='fftw_mpi_plan_many_r2r_f03') import integer(C_INT), value :: rnk integer(C_INTPTR_T), dimension(*), intent(in) :: n integer(C_INTPTR_T), value :: howmany integer(C_INTPTR_T), value :: iblock integer(C_INTPTR_T), value :: oblock real(C_DOUBLE), dimension(*), intent(out) :: in real(C_DOUBLE), dimension(*), intent(out) :: out integer(C_MPI_FINT), value :: comm integer(C_FFTW_R2R_KIND), dimension(*), intent(in) :: kind integer(C_INT), value :: flags end function fftw_mpi_plan_many_r2r type(C_PTR) function fftw_mpi_plan_r2r(rnk,n,in,out,comm,kind,flags) bind(C, name='fftw_mpi_plan_r2r_f03') import integer(C_INT), value :: rnk integer(C_INTPTR_T), dimension(*), intent(in) :: n real(C_DOUBLE), dimension(*), intent(out) :: in real(C_DOUBLE), dimension(*), intent(out) :: out integer(C_MPI_FINT), value :: comm integer(C_FFTW_R2R_KIND), dimension(*), intent(in) :: kind integer(C_INT), value :: flags end function fftw_mpi_plan_r2r type(C_PTR) function fftw_mpi_plan_r2r_2d(n0,n1,in,out,comm,kind0,kind1,flags) bind(C, name='fftw_mpi_plan_r2r_2d_f03') import integer(C_INTPTR_T), value :: n0 integer(C_INTPTR_T), value :: n1 real(C_DOUBLE), dimension(*), intent(out) :: in real(C_DOUBLE), dimension(*), intent(out) :: out integer(C_MPI_FINT), value :: comm integer(C_FFTW_R2R_KIND), value :: kind0 integer(C_FFTW_R2R_KIND), value :: kind1 integer(C_INT), value :: flags end function fftw_mpi_plan_r2r_2d type(C_PTR) function fftw_mpi_plan_r2r_3d(n0,n1,n2,in,out,comm,kind0,kind1,kind2,flags) bind(C, name='fftw_mpi_plan_r2r_3d_f03') import integer(C_INTPTR_T), value :: n0 integer(C_INTPTR_T), value :: n1 integer(C_INTPTR_T), value :: n2 real(C_DOUBLE), dimension(*), intent(out) :: in real(C_DOUBLE), dimension(*), intent(out) :: out integer(C_MPI_FINT), value :: comm integer(C_FFTW_R2R_KIND), value :: kind0 integer(C_FFTW_R2R_KIND), value :: kind1 integer(C_FFTW_R2R_KIND), value :: kind2 integer(C_INT), value :: flags end function fftw_mpi_plan_r2r_3d type(C_PTR) function fftw_mpi_plan_many_dft_r2c(rnk,n,howmany,iblock,oblock,in,out,comm,flags) & bind(C, name='fftw_mpi_plan_many_dft_r2c_f03') import integer(C_INT), value :: rnk integer(C_INTPTR_T), dimension(*), intent(in) :: n integer(C_INTPTR_T), value :: howmany integer(C_INTPTR_T), value :: iblock integer(C_INTPTR_T), value :: oblock real(C_DOUBLE), dimension(*), intent(out) :: in complex(C_DOUBLE_COMPLEX), dimension(*), intent(out) :: out integer(C_MPI_FINT), value :: comm integer(C_INT), value :: flags end function fftw_mpi_plan_many_dft_r2c type(C_PTR) function fftw_mpi_plan_dft_r2c(rnk,n,in,out,comm,flags) bind(C, name='fftw_mpi_plan_dft_r2c_f03') import integer(C_INT), value :: rnk integer(C_INTPTR_T), dimension(*), intent(in) :: n real(C_DOUBLE), dimension(*), intent(out) :: in complex(C_DOUBLE_COMPLEX), dimension(*), intent(out) :: out integer(C_MPI_FINT), value :: comm integer(C_INT), value :: flags end function fftw_mpi_plan_dft_r2c type(C_PTR) function fftw_mpi_plan_dft_r2c_2d(n0,n1,in,out,comm,flags) bind(C, name='fftw_mpi_plan_dft_r2c_2d_f03') import integer(C_INTPTR_T), value :: n0 integer(C_INTPTR_T), value :: n1 real(C_DOUBLE), dimension(*), intent(out) :: in complex(C_DOUBLE_COMPLEX), dimension(*), intent(out) :: out integer(C_MPI_FINT), value :: comm integer(C_INT), value :: flags end function fftw_mpi_plan_dft_r2c_2d type(C_PTR) function fftw_mpi_plan_dft_r2c_3d(n0,n1,n2,in,out,comm,flags) bind(C, name='fftw_mpi_plan_dft_r2c_3d_f03') import integer(C_INTPTR_T), value :: n0 integer(C_INTPTR_T), value :: n1 integer(C_INTPTR_T), value :: n2 real(C_DOUBLE), dimension(*), intent(out) :: in complex(C_DOUBLE_COMPLEX), dimension(*), intent(out) :: out integer(C_MPI_FINT), value :: comm integer(C_INT), value :: flags end function fftw_mpi_plan_dft_r2c_3d type(C_PTR) function fftw_mpi_plan_many_dft_c2r(rnk,n,howmany,iblock,oblock,in,out,comm,flags) & bind(C, name='fftw_mpi_plan_many_dft_c2r_f03') import integer(C_INT), value :: rnk integer(C_INTPTR_T), dimension(*), intent(in) :: n integer(C_INTPTR_T), value :: howmany integer(C_INTPTR_T), value :: iblock integer(C_INTPTR_T), value :: oblock complex(C_DOUBLE_COMPLEX), dimension(*), intent(out) :: in real(C_DOUBLE), dimension(*), intent(out) :: out integer(C_MPI_FINT), value :: comm integer(C_INT), value :: flags end function fftw_mpi_plan_many_dft_c2r type(C_PTR) function fftw_mpi_plan_dft_c2r(rnk,n,in,out,comm,flags) bind(C, name='fftw_mpi_plan_dft_c2r_f03') import integer(C_INT), value :: rnk integer(C_INTPTR_T), dimension(*), intent(in) :: n complex(C_DOUBLE_COMPLEX), dimension(*), intent(out) :: in real(C_DOUBLE), dimension(*), intent(out) :: out integer(C_MPI_FINT), value :: comm integer(C_INT), value :: flags end function fftw_mpi_plan_dft_c2r type(C_PTR) function fftw_mpi_plan_dft_c2r_2d(n0,n1,in,out,comm,flags) bind(C, name='fftw_mpi_plan_dft_c2r_2d_f03') import integer(C_INTPTR_T), value :: n0 integer(C_INTPTR_T), value :: n1 complex(C_DOUBLE_COMPLEX), dimension(*), intent(out) :: in real(C_DOUBLE), dimension(*), intent(out) :: out integer(C_MPI_FINT), value :: comm integer(C_INT), value :: flags end function fftw_mpi_plan_dft_c2r_2d type(C_PTR) function fftw_mpi_plan_dft_c2r_3d(n0,n1,n2,in,out,comm,flags) bind(C, name='fftw_mpi_plan_dft_c2r_3d_f03') import integer(C_INTPTR_T), value :: n0 integer(C_INTPTR_T), value :: n1 integer(C_INTPTR_T), value :: n2 complex(C_DOUBLE_COMPLEX), dimension(*), intent(out) :: in real(C_DOUBLE), dimension(*), intent(out) :: out integer(C_MPI_FINT), value :: comm integer(C_INT), value :: flags end function fftw_mpi_plan_dft_c2r_3d subroutine fftw_mpi_gather_wisdom(comm_) bind(C, name='fftw_mpi_gather_wisdom_f03') import integer(C_MPI_FINT), value :: comm_ end subroutine fftw_mpi_gather_wisdom subroutine fftw_mpi_broadcast_wisdom(comm_) bind(C, name='fftw_mpi_broadcast_wisdom_f03') import integer(C_MPI_FINT), value :: comm_ end subroutine fftw_mpi_broadcast_wisdom subroutine fftw_mpi_execute_dft(p,in,out) bind(C, name='fftw_mpi_execute_dft') import type(C_PTR), value :: p complex(C_DOUBLE_COMPLEX), dimension(*), intent(inout) :: in complex(C_DOUBLE_COMPLEX), dimension(*), intent(out) :: out end subroutine fftw_mpi_execute_dft subroutine fftw_mpi_execute_dft_r2c(p,in,out) bind(C, name='fftw_mpi_execute_dft_r2c') import type(C_PTR), value :: p real(C_DOUBLE), dimension(*), intent(inout) :: in complex(C_DOUBLE_COMPLEX), dimension(*), intent(out) :: out end subroutine fftw_mpi_execute_dft_r2c subroutine fftw_mpi_execute_dft_c2r(p,in,out) bind(C, name='fftw_mpi_execute_dft_c2r') import type(C_PTR), value :: p complex(C_DOUBLE_COMPLEX), dimension(*), intent(inout) :: in real(C_DOUBLE), dimension(*), intent(out) :: out end subroutine fftw_mpi_execute_dft_c2r subroutine fftw_mpi_execute_r2r(p,in,out) bind(C, name='fftw_mpi_execute_r2r') import type(C_PTR), value :: p real(C_DOUBLE), dimension(*), intent(inout) :: in real(C_DOUBLE), dimension(*), intent(out) :: out end subroutine fftw_mpi_execute_r2r end interface type, bind(C) :: fftwf_mpi_ddim integer(C_INTPTR_T) n, ib, ob end type fftwf_mpi_ddim interface subroutine fftwf_mpi_init() bind(C, name='fftwf_mpi_init') import end subroutine fftwf_mpi_init subroutine fftwf_mpi_cleanup() bind(C, name='fftwf_mpi_cleanup') import end subroutine fftwf_mpi_cleanup integer(C_INTPTR_T) function fftwf_mpi_local_size_many_transposed(rnk,n,howmany,block0,block1,comm,local_n0,local_0_start, & local_n1,local_1_start) & bind(C, name='fftwf_mpi_local_size_many_transposed_f03') import integer(C_INT), value :: rnk integer(C_INTPTR_T), dimension(*), intent(in) :: n integer(C_INTPTR_T), value :: howmany integer(C_INTPTR_T), value :: block0 integer(C_INTPTR_T), value :: block1 integer(C_MPI_FINT), value :: comm integer(C_INTPTR_T), intent(out) :: local_n0 integer(C_INTPTR_T), intent(out) :: local_0_start integer(C_INTPTR_T), intent(out) :: local_n1 integer(C_INTPTR_T), intent(out) :: local_1_start end function fftwf_mpi_local_size_many_transposed integer(C_INTPTR_T) function fftwf_mpi_local_size_many(rnk,n,howmany,block0,comm,local_n0,local_0_start) & bind(C, name='fftwf_mpi_local_size_many_f03') import integer(C_INT), value :: rnk integer(C_INTPTR_T), dimension(*), intent(in) :: n integer(C_INTPTR_T), value :: howmany integer(C_INTPTR_T), value :: block0 integer(C_MPI_FINT), value :: comm integer(C_INTPTR_T), intent(out) :: local_n0 integer(C_INTPTR_T), intent(out) :: local_0_start end function fftwf_mpi_local_size_many integer(C_INTPTR_T) function fftwf_mpi_local_size_transposed(rnk,n,comm,local_n0,local_0_start,local_n1,local_1_start) & bind(C, name='fftwf_mpi_local_size_transposed_f03') import integer(C_INT), value :: rnk integer(C_INTPTR_T), dimension(*), intent(in) :: n integer(C_MPI_FINT), value :: comm integer(C_INTPTR_T), intent(out) :: local_n0 integer(C_INTPTR_T), intent(out) :: local_0_start integer(C_INTPTR_T), intent(out) :: local_n1 integer(C_INTPTR_T), intent(out) :: local_1_start end function fftwf_mpi_local_size_transposed integer(C_INTPTR_T) function fftwf_mpi_local_size(rnk,n,comm,local_n0,local_0_start) bind(C, name='fftwf_mpi_local_size_f03') import integer(C_INT), value :: rnk integer(C_INTPTR_T), dimension(*), intent(in) :: n integer(C_MPI_FINT), value :: comm integer(C_INTPTR_T), intent(out) :: local_n0 integer(C_INTPTR_T), intent(out) :: local_0_start end function fftwf_mpi_local_size integer(C_INTPTR_T) function fftwf_mpi_local_size_many_1d(n0,howmany,comm,sign,flags,local_ni,local_i_start,local_no, & local_o_start) bind(C, name='fftwf_mpi_local_size_many_1d_f03') import integer(C_INTPTR_T), value :: n0 integer(C_INTPTR_T), value :: howmany integer(C_MPI_FINT), value :: comm integer(C_INT), value :: sign integer(C_INT), value :: flags integer(C_INTPTR_T), intent(out) :: local_ni integer(C_INTPTR_T), intent(out) :: local_i_start integer(C_INTPTR_T), intent(out) :: local_no integer(C_INTPTR_T), intent(out) :: local_o_start end function fftwf_mpi_local_size_many_1d integer(C_INTPTR_T) function fftwf_mpi_local_size_1d(n0,comm,sign,flags,local_ni,local_i_start,local_no,local_o_start) & bind(C, name='fftwf_mpi_local_size_1d_f03') import integer(C_INTPTR_T), value :: n0 integer(C_MPI_FINT), value :: comm integer(C_INT), value :: sign integer(C_INT), value :: flags integer(C_INTPTR_T), intent(out) :: local_ni integer(C_INTPTR_T), intent(out) :: local_i_start integer(C_INTPTR_T), intent(out) :: local_no integer(C_INTPTR_T), intent(out) :: local_o_start end function fftwf_mpi_local_size_1d integer(C_INTPTR_T) function fftwf_mpi_local_size_2d(n0,n1,comm,local_n0,local_0_start) & bind(C, name='fftwf_mpi_local_size_2d_f03') import integer(C_INTPTR_T), value :: n0 integer(C_INTPTR_T), value :: n1 integer(C_MPI_FINT), value :: comm integer(C_INTPTR_T), intent(out) :: local_n0 integer(C_INTPTR_T), intent(out) :: local_0_start end function fftwf_mpi_local_size_2d integer(C_INTPTR_T) function fftwf_mpi_local_size_2d_transposed(n0,n1,comm,local_n0,local_0_start,local_n1,local_1_start) & bind(C, name='fftwf_mpi_local_size_2d_transposed_f03') import integer(C_INTPTR_T), value :: n0 integer(C_INTPTR_T), value :: n1 integer(C_MPI_FINT), value :: comm integer(C_INTPTR_T), intent(out) :: local_n0 integer(C_INTPTR_T), intent(out) :: local_0_start integer(C_INTPTR_T), intent(out) :: local_n1 integer(C_INTPTR_T), intent(out) :: local_1_start end function fftwf_mpi_local_size_2d_transposed integer(C_INTPTR_T) function fftwf_mpi_local_size_3d(n0,n1,n2,comm,local_n0,local_0_start) & bind(C, name='fftwf_mpi_local_size_3d_f03') import integer(C_INTPTR_T), value :: n0 integer(C_INTPTR_T), value :: n1 integer(C_INTPTR_T), value :: n2 integer(C_MPI_FINT), value :: comm integer(C_INTPTR_T), intent(out) :: local_n0 integer(C_INTPTR_T), intent(out) :: local_0_start end function fftwf_mpi_local_size_3d integer(C_INTPTR_T) function fftwf_mpi_local_size_3d_transposed(n0,n1,n2,comm,local_n0,local_0_start,local_n1,local_1_start) & bind(C, name='fftwf_mpi_local_size_3d_transposed_f03') import integer(C_INTPTR_T), value :: n0 integer(C_INTPTR_T), value :: n1 integer(C_INTPTR_T), value :: n2 integer(C_MPI_FINT), value :: comm integer(C_INTPTR_T), intent(out) :: local_n0 integer(C_INTPTR_T), intent(out) :: local_0_start integer(C_INTPTR_T), intent(out) :: local_n1 integer(C_INTPTR_T), intent(out) :: local_1_start end function fftwf_mpi_local_size_3d_transposed type(C_PTR) function fftwf_mpi_plan_many_transpose(n0,n1,howmany,block0,block1,in,out,comm,flags) & bind(C, name='fftwf_mpi_plan_many_transpose_f03') import integer(C_INTPTR_T), value :: n0 integer(C_INTPTR_T), value :: n1 integer(C_INTPTR_T), value :: howmany integer(C_INTPTR_T), value :: block0 integer(C_INTPTR_T), value :: block1 real(C_FLOAT), dimension(*), intent(out) :: in real(C_FLOAT), dimension(*), intent(out) :: out integer(C_MPI_FINT), value :: comm integer(C_INT), value :: flags end function fftwf_mpi_plan_many_transpose type(C_PTR) function fftwf_mpi_plan_transpose(n0,n1,in,out,comm,flags) bind(C, name='fftwf_mpi_plan_transpose_f03') import integer(C_INTPTR_T), value :: n0 integer(C_INTPTR_T), value :: n1 real(C_FLOAT), dimension(*), intent(out) :: in real(C_FLOAT), dimension(*), intent(out) :: out integer(C_MPI_FINT), value :: comm integer(C_INT), value :: flags end function fftwf_mpi_plan_transpose type(C_PTR) function fftwf_mpi_plan_many_dft(rnk,n,howmany,block,tblock,in,out,comm,sign,flags) & bind(C, name='fftwf_mpi_plan_many_dft_f03') import integer(C_INT), value :: rnk integer(C_INTPTR_T), dimension(*), intent(in) :: n integer(C_INTPTR_T), value :: howmany integer(C_INTPTR_T), value :: block integer(C_INTPTR_T), value :: tblock complex(C_FLOAT_COMPLEX), dimension(*), intent(out) :: in complex(C_FLOAT_COMPLEX), dimension(*), intent(out) :: out integer(C_MPI_FINT), value :: comm integer(C_INT), value :: sign integer(C_INT), value :: flags end function fftwf_mpi_plan_many_dft type(C_PTR) function fftwf_mpi_plan_dft(rnk,n,in,out,comm,sign,flags) bind(C, name='fftwf_mpi_plan_dft_f03') import integer(C_INT), value :: rnk integer(C_INTPTR_T), dimension(*), intent(in) :: n complex(C_FLOAT_COMPLEX), dimension(*), intent(out) :: in complex(C_FLOAT_COMPLEX), dimension(*), intent(out) :: out integer(C_MPI_FINT), value :: comm integer(C_INT), value :: sign integer(C_INT), value :: flags end function fftwf_mpi_plan_dft type(C_PTR) function fftwf_mpi_plan_dft_1d(n0,in,out,comm,sign,flags) bind(C, name='fftwf_mpi_plan_dft_1d_f03') import integer(C_INTPTR_T), value :: n0 complex(C_FLOAT_COMPLEX), dimension(*), intent(out) :: in complex(C_FLOAT_COMPLEX), dimension(*), intent(out) :: out integer(C_MPI_FINT), value :: comm integer(C_INT), value :: sign integer(C_INT), value :: flags end function fftwf_mpi_plan_dft_1d type(C_PTR) function fftwf_mpi_plan_dft_2d(n0,n1,in,out,comm,sign,flags) bind(C, name='fftwf_mpi_plan_dft_2d_f03') import integer(C_INTPTR_T), value :: n0 integer(C_INTPTR_T), value :: n1 complex(C_FLOAT_COMPLEX), dimension(*), intent(out) :: in complex(C_FLOAT_COMPLEX), dimension(*), intent(out) :: out integer(C_MPI_FINT), value :: comm integer(C_INT), value :: sign integer(C_INT), value :: flags end function fftwf_mpi_plan_dft_2d type(C_PTR) function fftwf_mpi_plan_dft_3d(n0,n1,n2,in,out,comm,sign,flags) bind(C, name='fftwf_mpi_plan_dft_3d_f03') import integer(C_INTPTR_T), value :: n0 integer(C_INTPTR_T), value :: n1 integer(C_INTPTR_T), value :: n2 complex(C_FLOAT_COMPLEX), dimension(*), intent(out) :: in complex(C_FLOAT_COMPLEX), dimension(*), intent(out) :: out integer(C_MPI_FINT), value :: comm integer(C_INT), value :: sign integer(C_INT), value :: flags end function fftwf_mpi_plan_dft_3d type(C_PTR) function fftwf_mpi_plan_many_r2r(rnk,n,howmany,iblock,oblock,in,out,comm,kind,flags) & bind(C, name='fftwf_mpi_plan_many_r2r_f03') import integer(C_INT), value :: rnk integer(C_INTPTR_T), dimension(*), intent(in) :: n integer(C_INTPTR_T), value :: howmany integer(C_INTPTR_T), value :: iblock integer(C_INTPTR_T), value :: oblock real(C_FLOAT), dimension(*), intent(out) :: in real(C_FLOAT), dimension(*), intent(out) :: out integer(C_MPI_FINT), value :: comm integer(C_FFTW_R2R_KIND), dimension(*), intent(in) :: kind integer(C_INT), value :: flags end function fftwf_mpi_plan_many_r2r type(C_PTR) function fftwf_mpi_plan_r2r(rnk,n,in,out,comm,kind,flags) bind(C, name='fftwf_mpi_plan_r2r_f03') import integer(C_INT), value :: rnk integer(C_INTPTR_T), dimension(*), intent(in) :: n real(C_FLOAT), dimension(*), intent(out) :: in real(C_FLOAT), dimension(*), intent(out) :: out integer(C_MPI_FINT), value :: comm integer(C_FFTW_R2R_KIND), dimension(*), intent(in) :: kind integer(C_INT), value :: flags end function fftwf_mpi_plan_r2r type(C_PTR) function fftwf_mpi_plan_r2r_2d(n0,n1,in,out,comm,kind0,kind1,flags) bind(C, name='fftwf_mpi_plan_r2r_2d_f03') import integer(C_INTPTR_T), value :: n0 integer(C_INTPTR_T), value :: n1 real(C_FLOAT), dimension(*), intent(out) :: in real(C_FLOAT), dimension(*), intent(out) :: out integer(C_MPI_FINT), value :: comm integer(C_FFTW_R2R_KIND), value :: kind0 integer(C_FFTW_R2R_KIND), value :: kind1 integer(C_INT), value :: flags end function fftwf_mpi_plan_r2r_2d type(C_PTR) function fftwf_mpi_plan_r2r_3d(n0,n1,n2,in,out,comm,kind0,kind1,kind2,flags) & bind(C, name='fftwf_mpi_plan_r2r_3d_f03') import integer(C_INTPTR_T), value :: n0 integer(C_INTPTR_T), value :: n1 integer(C_INTPTR_T), value :: n2 real(C_FLOAT), dimension(*), intent(out) :: in real(C_FLOAT), dimension(*), intent(out) :: out integer(C_MPI_FINT), value :: comm integer(C_FFTW_R2R_KIND), value :: kind0 integer(C_FFTW_R2R_KIND), value :: kind1 integer(C_FFTW_R2R_KIND), value :: kind2 integer(C_INT), value :: flags end function fftwf_mpi_plan_r2r_3d type(C_PTR) function fftwf_mpi_plan_many_dft_r2c(rnk,n,howmany,iblock,oblock,in,out,comm,flags) & bind(C, name='fftwf_mpi_plan_many_dft_r2c_f03') import integer(C_INT), value :: rnk integer(C_INTPTR_T), dimension(*), intent(in) :: n integer(C_INTPTR_T), value :: howmany integer(C_INTPTR_T), value :: iblock integer(C_INTPTR_T), value :: oblock real(C_FLOAT), dimension(*), intent(out) :: in complex(C_FLOAT_COMPLEX), dimension(*), intent(out) :: out integer(C_MPI_FINT), value :: comm integer(C_INT), value :: flags end function fftwf_mpi_plan_many_dft_r2c type(C_PTR) function fftwf_mpi_plan_dft_r2c(rnk,n,in,out,comm,flags) bind(C, name='fftwf_mpi_plan_dft_r2c_f03') import integer(C_INT), value :: rnk integer(C_INTPTR_T), dimension(*), intent(in) :: n real(C_FLOAT), dimension(*), intent(out) :: in complex(C_FLOAT_COMPLEX), dimension(*), intent(out) :: out integer(C_MPI_FINT), value :: comm integer(C_INT), value :: flags end function fftwf_mpi_plan_dft_r2c type(C_PTR) function fftwf_mpi_plan_dft_r2c_2d(n0,n1,in,out,comm,flags) bind(C, name='fftwf_mpi_plan_dft_r2c_2d_f03') import integer(C_INTPTR_T), value :: n0 integer(C_INTPTR_T), value :: n1 real(C_FLOAT), dimension(*), intent(out) :: in complex(C_FLOAT_COMPLEX), dimension(*), intent(out) :: out integer(C_MPI_FINT), value :: comm integer(C_INT), value :: flags end function fftwf_mpi_plan_dft_r2c_2d type(C_PTR) function fftwf_mpi_plan_dft_r2c_3d(n0,n1,n2,in,out,comm,flags) bind(C, name='fftwf_mpi_plan_dft_r2c_3d_f03') import integer(C_INTPTR_T), value :: n0 integer(C_INTPTR_T), value :: n1 integer(C_INTPTR_T), value :: n2 real(C_FLOAT), dimension(*), intent(out) :: in complex(C_FLOAT_COMPLEX), dimension(*), intent(out) :: out integer(C_MPI_FINT), value :: comm integer(C_INT), value :: flags end function fftwf_mpi_plan_dft_r2c_3d type(C_PTR) function fftwf_mpi_plan_many_dft_c2r(rnk,n,howmany,iblock,oblock,in,out,comm,flags) & bind(C, name='fftwf_mpi_plan_many_dft_c2r_f03') import integer(C_INT), value :: rnk integer(C_INTPTR_T), dimension(*), intent(in) :: n integer(C_INTPTR_T), value :: howmany integer(C_INTPTR_T), value :: iblock integer(C_INTPTR_T), value :: oblock complex(C_FLOAT_COMPLEX), dimension(*), intent(out) :: in real(C_FLOAT), dimension(*), intent(out) :: out integer(C_MPI_FINT), value :: comm integer(C_INT), value :: flags end function fftwf_mpi_plan_many_dft_c2r type(C_PTR) function fftwf_mpi_plan_dft_c2r(rnk,n,in,out,comm,flags) bind(C, name='fftwf_mpi_plan_dft_c2r_f03') import integer(C_INT), value :: rnk integer(C_INTPTR_T), dimension(*), intent(in) :: n complex(C_FLOAT_COMPLEX), dimension(*), intent(out) :: in real(C_FLOAT), dimension(*), intent(out) :: out integer(C_MPI_FINT), value :: comm integer(C_INT), value :: flags end function fftwf_mpi_plan_dft_c2r type(C_PTR) function fftwf_mpi_plan_dft_c2r_2d(n0,n1,in,out,comm,flags) bind(C, name='fftwf_mpi_plan_dft_c2r_2d_f03') import integer(C_INTPTR_T), value :: n0 integer(C_INTPTR_T), value :: n1 complex(C_FLOAT_COMPLEX), dimension(*), intent(out) :: in real(C_FLOAT), dimension(*), intent(out) :: out integer(C_MPI_FINT), value :: comm integer(C_INT), value :: flags end function fftwf_mpi_plan_dft_c2r_2d type(C_PTR) function fftwf_mpi_plan_dft_c2r_3d(n0,n1,n2,in,out,comm,flags) bind(C, name='fftwf_mpi_plan_dft_c2r_3d_f03') import integer(C_INTPTR_T), value :: n0 integer(C_INTPTR_T), value :: n1 integer(C_INTPTR_T), value :: n2 complex(C_FLOAT_COMPLEX), dimension(*), intent(out) :: in real(C_FLOAT), dimension(*), intent(out) :: out integer(C_MPI_FINT), value :: comm integer(C_INT), value :: flags end function fftwf_mpi_plan_dft_c2r_3d subroutine fftwf_mpi_gather_wisdom(comm_) bind(C, name='fftwf_mpi_gather_wisdom_f03') import integer(C_MPI_FINT), value :: comm_ end subroutine fftwf_mpi_gather_wisdom subroutine fftwf_mpi_broadcast_wisdom(comm_) bind(C, name='fftwf_mpi_broadcast_wisdom_f03') import integer(C_MPI_FINT), value :: comm_ end subroutine fftwf_mpi_broadcast_wisdom subroutine fftwf_mpi_execute_dft(p,in,out) bind(C, name='fftwf_mpi_execute_dft') import type(C_PTR), value :: p complex(C_FLOAT_COMPLEX), dimension(*), intent(inout) :: in complex(C_FLOAT_COMPLEX), dimension(*), intent(out) :: out end subroutine fftwf_mpi_execute_dft subroutine fftwf_mpi_execute_dft_r2c(p,in,out) bind(C, name='fftwf_mpi_execute_dft_r2c') import type(C_PTR), value :: p real(C_FLOAT), dimension(*), intent(inout) :: in complex(C_FLOAT_COMPLEX), dimension(*), intent(out) :: out end subroutine fftwf_mpi_execute_dft_r2c subroutine fftwf_mpi_execute_dft_c2r(p,in,out) bind(C, name='fftwf_mpi_execute_dft_c2r') import type(C_PTR), value :: p complex(C_FLOAT_COMPLEX), dimension(*), intent(inout) :: in real(C_FLOAT), dimension(*), intent(out) :: out end subroutine fftwf_mpi_execute_dft_c2r subroutine fftwf_mpi_execute_r2r(p,in,out) bind(C, name='fftwf_mpi_execute_r2r') import type(C_PTR), value :: p real(C_FLOAT), dimension(*), intent(inout) :: in real(C_FLOAT), dimension(*), intent(out) :: out end subroutine fftwf_mpi_execute_r2r end interface fftw-3.3.3/mpi/fftw3l-mpi.f03.in0000644000175000017500000004442512054411104013104 00000000000000! Generated automatically. DO NOT EDIT! include 'fftw3l.f03' type, bind(C) :: fftwl_mpi_ddim integer(C_INTPTR_T) n, ib, ob end type fftwl_mpi_ddim interface subroutine fftwl_mpi_init() bind(C, name='fftwl_mpi_init') import end subroutine fftwl_mpi_init subroutine fftwl_mpi_cleanup() bind(C, name='fftwl_mpi_cleanup') import end subroutine fftwl_mpi_cleanup integer(C_INTPTR_T) function fftwl_mpi_local_size_many_transposed(rnk,n,howmany,block0,block1,comm,local_n0,local_0_start, & local_n1,local_1_start) & bind(C, name='fftwl_mpi_local_size_many_transposed_f03') import integer(C_INT), value :: rnk integer(C_INTPTR_T), dimension(*), intent(in) :: n integer(C_INTPTR_T), value :: howmany integer(C_INTPTR_T), value :: block0 integer(C_INTPTR_T), value :: block1 integer(C_MPI_FINT), value :: comm integer(C_INTPTR_T), intent(out) :: local_n0 integer(C_INTPTR_T), intent(out) :: local_0_start integer(C_INTPTR_T), intent(out) :: local_n1 integer(C_INTPTR_T), intent(out) :: local_1_start end function fftwl_mpi_local_size_many_transposed integer(C_INTPTR_T) function fftwl_mpi_local_size_many(rnk,n,howmany,block0,comm,local_n0,local_0_start) & bind(C, name='fftwl_mpi_local_size_many_f03') import integer(C_INT), value :: rnk integer(C_INTPTR_T), dimension(*), intent(in) :: n integer(C_INTPTR_T), value :: howmany integer(C_INTPTR_T), value :: block0 integer(C_MPI_FINT), value :: comm integer(C_INTPTR_T), intent(out) :: local_n0 integer(C_INTPTR_T), intent(out) :: local_0_start end function fftwl_mpi_local_size_many integer(C_INTPTR_T) function fftwl_mpi_local_size_transposed(rnk,n,comm,local_n0,local_0_start,local_n1,local_1_start) & bind(C, name='fftwl_mpi_local_size_transposed_f03') import integer(C_INT), value :: rnk integer(C_INTPTR_T), dimension(*), intent(in) :: n integer(C_MPI_FINT), value :: comm integer(C_INTPTR_T), intent(out) :: local_n0 integer(C_INTPTR_T), intent(out) :: local_0_start integer(C_INTPTR_T), intent(out) :: local_n1 integer(C_INTPTR_T), intent(out) :: local_1_start end function fftwl_mpi_local_size_transposed integer(C_INTPTR_T) function fftwl_mpi_local_size(rnk,n,comm,local_n0,local_0_start) bind(C, name='fftwl_mpi_local_size_f03') import integer(C_INT), value :: rnk integer(C_INTPTR_T), dimension(*), intent(in) :: n integer(C_MPI_FINT), value :: comm integer(C_INTPTR_T), intent(out) :: local_n0 integer(C_INTPTR_T), intent(out) :: local_0_start end function fftwl_mpi_local_size integer(C_INTPTR_T) function fftwl_mpi_local_size_many_1d(n0,howmany,comm,sign,flags,local_ni,local_i_start,local_no, & local_o_start) bind(C, name='fftwl_mpi_local_size_many_1d_f03') import integer(C_INTPTR_T), value :: n0 integer(C_INTPTR_T), value :: howmany integer(C_MPI_FINT), value :: comm integer(C_INT), value :: sign integer(C_INT), value :: flags integer(C_INTPTR_T), intent(out) :: local_ni integer(C_INTPTR_T), intent(out) :: local_i_start integer(C_INTPTR_T), intent(out) :: local_no integer(C_INTPTR_T), intent(out) :: local_o_start end function fftwl_mpi_local_size_many_1d integer(C_INTPTR_T) function fftwl_mpi_local_size_1d(n0,comm,sign,flags,local_ni,local_i_start,local_no,local_o_start) & bind(C, name='fftwl_mpi_local_size_1d_f03') import integer(C_INTPTR_T), value :: n0 integer(C_MPI_FINT), value :: comm integer(C_INT), value :: sign integer(C_INT), value :: flags integer(C_INTPTR_T), intent(out) :: local_ni integer(C_INTPTR_T), intent(out) :: local_i_start integer(C_INTPTR_T), intent(out) :: local_no integer(C_INTPTR_T), intent(out) :: local_o_start end function fftwl_mpi_local_size_1d integer(C_INTPTR_T) function fftwl_mpi_local_size_2d(n0,n1,comm,local_n0,local_0_start) & bind(C, name='fftwl_mpi_local_size_2d_f03') import integer(C_INTPTR_T), value :: n0 integer(C_INTPTR_T), value :: n1 integer(C_MPI_FINT), value :: comm integer(C_INTPTR_T), intent(out) :: local_n0 integer(C_INTPTR_T), intent(out) :: local_0_start end function fftwl_mpi_local_size_2d integer(C_INTPTR_T) function fftwl_mpi_local_size_2d_transposed(n0,n1,comm,local_n0,local_0_start,local_n1,local_1_start) & bind(C, name='fftwl_mpi_local_size_2d_transposed_f03') import integer(C_INTPTR_T), value :: n0 integer(C_INTPTR_T), value :: n1 integer(C_MPI_FINT), value :: comm integer(C_INTPTR_T), intent(out) :: local_n0 integer(C_INTPTR_T), intent(out) :: local_0_start integer(C_INTPTR_T), intent(out) :: local_n1 integer(C_INTPTR_T), intent(out) :: local_1_start end function fftwl_mpi_local_size_2d_transposed integer(C_INTPTR_T) function fftwl_mpi_local_size_3d(n0,n1,n2,comm,local_n0,local_0_start) & bind(C, name='fftwl_mpi_local_size_3d_f03') import integer(C_INTPTR_T), value :: n0 integer(C_INTPTR_T), value :: n1 integer(C_INTPTR_T), value :: n2 integer(C_MPI_FINT), value :: comm integer(C_INTPTR_T), intent(out) :: local_n0 integer(C_INTPTR_T), intent(out) :: local_0_start end function fftwl_mpi_local_size_3d integer(C_INTPTR_T) function fftwl_mpi_local_size_3d_transposed(n0,n1,n2,comm,local_n0,local_0_start,local_n1,local_1_start) & bind(C, name='fftwl_mpi_local_size_3d_transposed_f03') import integer(C_INTPTR_T), value :: n0 integer(C_INTPTR_T), value :: n1 integer(C_INTPTR_T), value :: n2 integer(C_MPI_FINT), value :: comm integer(C_INTPTR_T), intent(out) :: local_n0 integer(C_INTPTR_T), intent(out) :: local_0_start integer(C_INTPTR_T), intent(out) :: local_n1 integer(C_INTPTR_T), intent(out) :: local_1_start end function fftwl_mpi_local_size_3d_transposed type(C_PTR) function fftwl_mpi_plan_many_transpose(n0,n1,howmany,block0,block1,in,out,comm,flags) & bind(C, name='fftwl_mpi_plan_many_transpose_f03') import integer(C_INTPTR_T), value :: n0 integer(C_INTPTR_T), value :: n1 integer(C_INTPTR_T), value :: howmany integer(C_INTPTR_T), value :: block0 integer(C_INTPTR_T), value :: block1 real(C_LONG_DOUBLE), dimension(*), intent(out) :: in real(C_LONG_DOUBLE), dimension(*), intent(out) :: out integer(C_MPI_FINT), value :: comm integer(C_INT), value :: flags end function fftwl_mpi_plan_many_transpose type(C_PTR) function fftwl_mpi_plan_transpose(n0,n1,in,out,comm,flags) bind(C, name='fftwl_mpi_plan_transpose_f03') import integer(C_INTPTR_T), value :: n0 integer(C_INTPTR_T), value :: n1 real(C_LONG_DOUBLE), dimension(*), intent(out) :: in real(C_LONG_DOUBLE), dimension(*), intent(out) :: out integer(C_MPI_FINT), value :: comm integer(C_INT), value :: flags end function fftwl_mpi_plan_transpose type(C_PTR) function fftwl_mpi_plan_many_dft(rnk,n,howmany,block,tblock,in,out,comm,sign,flags) & bind(C, name='fftwl_mpi_plan_many_dft_f03') import integer(C_INT), value :: rnk integer(C_INTPTR_T), dimension(*), intent(in) :: n integer(C_INTPTR_T), value :: howmany integer(C_INTPTR_T), value :: block integer(C_INTPTR_T), value :: tblock complex(C_LONG_DOUBLE_COMPLEX), dimension(*), intent(out) :: in complex(C_LONG_DOUBLE_COMPLEX), dimension(*), intent(out) :: out integer(C_MPI_FINT), value :: comm integer(C_INT), value :: sign integer(C_INT), value :: flags end function fftwl_mpi_plan_many_dft type(C_PTR) function fftwl_mpi_plan_dft(rnk,n,in,out,comm,sign,flags) bind(C, name='fftwl_mpi_plan_dft_f03') import integer(C_INT), value :: rnk integer(C_INTPTR_T), dimension(*), intent(in) :: n complex(C_LONG_DOUBLE_COMPLEX), dimension(*), intent(out) :: in complex(C_LONG_DOUBLE_COMPLEX), dimension(*), intent(out) :: out integer(C_MPI_FINT), value :: comm integer(C_INT), value :: sign integer(C_INT), value :: flags end function fftwl_mpi_plan_dft type(C_PTR) function fftwl_mpi_plan_dft_1d(n0,in,out,comm,sign,flags) bind(C, name='fftwl_mpi_plan_dft_1d_f03') import integer(C_INTPTR_T), value :: n0 complex(C_LONG_DOUBLE_COMPLEX), dimension(*), intent(out) :: in complex(C_LONG_DOUBLE_COMPLEX), dimension(*), intent(out) :: out integer(C_MPI_FINT), value :: comm integer(C_INT), value :: sign integer(C_INT), value :: flags end function fftwl_mpi_plan_dft_1d type(C_PTR) function fftwl_mpi_plan_dft_2d(n0,n1,in,out,comm,sign,flags) bind(C, name='fftwl_mpi_plan_dft_2d_f03') import integer(C_INTPTR_T), value :: n0 integer(C_INTPTR_T), value :: n1 complex(C_LONG_DOUBLE_COMPLEX), dimension(*), intent(out) :: in complex(C_LONG_DOUBLE_COMPLEX), dimension(*), intent(out) :: out integer(C_MPI_FINT), value :: comm integer(C_INT), value :: sign integer(C_INT), value :: flags end function fftwl_mpi_plan_dft_2d type(C_PTR) function fftwl_mpi_plan_dft_3d(n0,n1,n2,in,out,comm,sign,flags) bind(C, name='fftwl_mpi_plan_dft_3d_f03') import integer(C_INTPTR_T), value :: n0 integer(C_INTPTR_T), value :: n1 integer(C_INTPTR_T), value :: n2 complex(C_LONG_DOUBLE_COMPLEX), dimension(*), intent(out) :: in complex(C_LONG_DOUBLE_COMPLEX), dimension(*), intent(out) :: out integer(C_MPI_FINT), value :: comm integer(C_INT), value :: sign integer(C_INT), value :: flags end function fftwl_mpi_plan_dft_3d type(C_PTR) function fftwl_mpi_plan_many_r2r(rnk,n,howmany,iblock,oblock,in,out,comm,kind,flags) & bind(C, name='fftwl_mpi_plan_many_r2r_f03') import integer(C_INT), value :: rnk integer(C_INTPTR_T), dimension(*), intent(in) :: n integer(C_INTPTR_T), value :: howmany integer(C_INTPTR_T), value :: iblock integer(C_INTPTR_T), value :: oblock real(C_LONG_DOUBLE), dimension(*), intent(out) :: in real(C_LONG_DOUBLE), dimension(*), intent(out) :: out integer(C_MPI_FINT), value :: comm integer(C_FFTW_R2R_KIND), dimension(*), intent(in) :: kind integer(C_INT), value :: flags end function fftwl_mpi_plan_many_r2r type(C_PTR) function fftwl_mpi_plan_r2r(rnk,n,in,out,comm,kind,flags) bind(C, name='fftwl_mpi_plan_r2r_f03') import integer(C_INT), value :: rnk integer(C_INTPTR_T), dimension(*), intent(in) :: n real(C_LONG_DOUBLE), dimension(*), intent(out) :: in real(C_LONG_DOUBLE), dimension(*), intent(out) :: out integer(C_MPI_FINT), value :: comm integer(C_FFTW_R2R_KIND), dimension(*), intent(in) :: kind integer(C_INT), value :: flags end function fftwl_mpi_plan_r2r type(C_PTR) function fftwl_mpi_plan_r2r_2d(n0,n1,in,out,comm,kind0,kind1,flags) bind(C, name='fftwl_mpi_plan_r2r_2d_f03') import integer(C_INTPTR_T), value :: n0 integer(C_INTPTR_T), value :: n1 real(C_LONG_DOUBLE), dimension(*), intent(out) :: in real(C_LONG_DOUBLE), dimension(*), intent(out) :: out integer(C_MPI_FINT), value :: comm integer(C_FFTW_R2R_KIND), value :: kind0 integer(C_FFTW_R2R_KIND), value :: kind1 integer(C_INT), value :: flags end function fftwl_mpi_plan_r2r_2d type(C_PTR) function fftwl_mpi_plan_r2r_3d(n0,n1,n2,in,out,comm,kind0,kind1,kind2,flags) & bind(C, name='fftwl_mpi_plan_r2r_3d_f03') import integer(C_INTPTR_T), value :: n0 integer(C_INTPTR_T), value :: n1 integer(C_INTPTR_T), value :: n2 real(C_LONG_DOUBLE), dimension(*), intent(out) :: in real(C_LONG_DOUBLE), dimension(*), intent(out) :: out integer(C_MPI_FINT), value :: comm integer(C_FFTW_R2R_KIND), value :: kind0 integer(C_FFTW_R2R_KIND), value :: kind1 integer(C_FFTW_R2R_KIND), value :: kind2 integer(C_INT), value :: flags end function fftwl_mpi_plan_r2r_3d type(C_PTR) function fftwl_mpi_plan_many_dft_r2c(rnk,n,howmany,iblock,oblock,in,out,comm,flags) & bind(C, name='fftwl_mpi_plan_many_dft_r2c_f03') import integer(C_INT), value :: rnk integer(C_INTPTR_T), dimension(*), intent(in) :: n integer(C_INTPTR_T), value :: howmany integer(C_INTPTR_T), value :: iblock integer(C_INTPTR_T), value :: oblock real(C_LONG_DOUBLE), dimension(*), intent(out) :: in complex(C_LONG_DOUBLE_COMPLEX), dimension(*), intent(out) :: out integer(C_MPI_FINT), value :: comm integer(C_INT), value :: flags end function fftwl_mpi_plan_many_dft_r2c type(C_PTR) function fftwl_mpi_plan_dft_r2c(rnk,n,in,out,comm,flags) bind(C, name='fftwl_mpi_plan_dft_r2c_f03') import integer(C_INT), value :: rnk integer(C_INTPTR_T), dimension(*), intent(in) :: n real(C_LONG_DOUBLE), dimension(*), intent(out) :: in complex(C_LONG_DOUBLE_COMPLEX), dimension(*), intent(out) :: out integer(C_MPI_FINT), value :: comm integer(C_INT), value :: flags end function fftwl_mpi_plan_dft_r2c type(C_PTR) function fftwl_mpi_plan_dft_r2c_2d(n0,n1,in,out,comm,flags) bind(C, name='fftwl_mpi_plan_dft_r2c_2d_f03') import integer(C_INTPTR_T), value :: n0 integer(C_INTPTR_T), value :: n1 real(C_LONG_DOUBLE), dimension(*), intent(out) :: in complex(C_LONG_DOUBLE_COMPLEX), dimension(*), intent(out) :: out integer(C_MPI_FINT), value :: comm integer(C_INT), value :: flags end function fftwl_mpi_plan_dft_r2c_2d type(C_PTR) function fftwl_mpi_plan_dft_r2c_3d(n0,n1,n2,in,out,comm,flags) bind(C, name='fftwl_mpi_plan_dft_r2c_3d_f03') import integer(C_INTPTR_T), value :: n0 integer(C_INTPTR_T), value :: n1 integer(C_INTPTR_T), value :: n2 real(C_LONG_DOUBLE), dimension(*), intent(out) :: in complex(C_LONG_DOUBLE_COMPLEX), dimension(*), intent(out) :: out integer(C_MPI_FINT), value :: comm integer(C_INT), value :: flags end function fftwl_mpi_plan_dft_r2c_3d type(C_PTR) function fftwl_mpi_plan_many_dft_c2r(rnk,n,howmany,iblock,oblock,in,out,comm,flags) & bind(C, name='fftwl_mpi_plan_many_dft_c2r_f03') import integer(C_INT), value :: rnk integer(C_INTPTR_T), dimension(*), intent(in) :: n integer(C_INTPTR_T), value :: howmany integer(C_INTPTR_T), value :: iblock integer(C_INTPTR_T), value :: oblock complex(C_LONG_DOUBLE_COMPLEX), dimension(*), intent(out) :: in real(C_LONG_DOUBLE), dimension(*), intent(out) :: out integer(C_MPI_FINT), value :: comm integer(C_INT), value :: flags end function fftwl_mpi_plan_many_dft_c2r type(C_PTR) function fftwl_mpi_plan_dft_c2r(rnk,n,in,out,comm,flags) bind(C, name='fftwl_mpi_plan_dft_c2r_f03') import integer(C_INT), value :: rnk integer(C_INTPTR_T), dimension(*), intent(in) :: n complex(C_LONG_DOUBLE_COMPLEX), dimension(*), intent(out) :: in real(C_LONG_DOUBLE), dimension(*), intent(out) :: out integer(C_MPI_FINT), value :: comm integer(C_INT), value :: flags end function fftwl_mpi_plan_dft_c2r type(C_PTR) function fftwl_mpi_plan_dft_c2r_2d(n0,n1,in,out,comm,flags) bind(C, name='fftwl_mpi_plan_dft_c2r_2d_f03') import integer(C_INTPTR_T), value :: n0 integer(C_INTPTR_T), value :: n1 complex(C_LONG_DOUBLE_COMPLEX), dimension(*), intent(out) :: in real(C_LONG_DOUBLE), dimension(*), intent(out) :: out integer(C_MPI_FINT), value :: comm integer(C_INT), value :: flags end function fftwl_mpi_plan_dft_c2r_2d type(C_PTR) function fftwl_mpi_plan_dft_c2r_3d(n0,n1,n2,in,out,comm,flags) bind(C, name='fftwl_mpi_plan_dft_c2r_3d_f03') import integer(C_INTPTR_T), value :: n0 integer(C_INTPTR_T), value :: n1 integer(C_INTPTR_T), value :: n2 complex(C_LONG_DOUBLE_COMPLEX), dimension(*), intent(out) :: in real(C_LONG_DOUBLE), dimension(*), intent(out) :: out integer(C_MPI_FINT), value :: comm integer(C_INT), value :: flags end function fftwl_mpi_plan_dft_c2r_3d subroutine fftwl_mpi_gather_wisdom(comm_) bind(C, name='fftwl_mpi_gather_wisdom_f03') import integer(C_MPI_FINT), value :: comm_ end subroutine fftwl_mpi_gather_wisdom subroutine fftwl_mpi_broadcast_wisdom(comm_) bind(C, name='fftwl_mpi_broadcast_wisdom_f03') import integer(C_MPI_FINT), value :: comm_ end subroutine fftwl_mpi_broadcast_wisdom subroutine fftwl_mpi_execute_dft(p,in,out) bind(C, name='fftwl_mpi_execute_dft') import type(C_PTR), value :: p complex(C_LONG_DOUBLE_COMPLEX), dimension(*), intent(inout) :: in complex(C_LONG_DOUBLE_COMPLEX), dimension(*), intent(out) :: out end subroutine fftwl_mpi_execute_dft subroutine fftwl_mpi_execute_dft_r2c(p,in,out) bind(C, name='fftwl_mpi_execute_dft_r2c') import type(C_PTR), value :: p real(C_LONG_DOUBLE), dimension(*), intent(inout) :: in complex(C_LONG_DOUBLE_COMPLEX), dimension(*), intent(out) :: out end subroutine fftwl_mpi_execute_dft_r2c subroutine fftwl_mpi_execute_dft_c2r(p,in,out) bind(C, name='fftwl_mpi_execute_dft_c2r') import type(C_PTR), value :: p complex(C_LONG_DOUBLE_COMPLEX), dimension(*), intent(inout) :: in real(C_LONG_DOUBLE), dimension(*), intent(out) :: out end subroutine fftwl_mpi_execute_dft_c2r subroutine fftwl_mpi_execute_r2r(p,in,out) bind(C, name='fftwl_mpi_execute_r2r') import type(C_PTR), value :: p real(C_LONG_DOUBLE), dimension(*), intent(inout) :: in real(C_LONG_DOUBLE), dimension(*), intent(out) :: out end subroutine fftwl_mpi_execute_r2r end interface fftw-3.3.3/doc/0000755000175000017500000000000012054411114010225 500000000000000fftw-3.3.3/doc/acknowledgements.texi0000644000175000017500000000720512054410050014374 00000000000000@node Acknowledgments, License and Copyright, Installation and Customization, Top @chapter Acknowledgments Matteo Frigo was supported in part by the Special Research Program SFB F011 ``AURORA'' of the Austrian Science Fund FWF and by MIT Lincoln Laboratory. For previous versions of FFTW, he was supported in part by the Defense Advanced Research Projects Agency (DARPA), under Grants N00014-94-1-0985 and F30602-97-1-0270, and by a Digital Equipment Corporation Fellowship. Steven G. Johnson was supported in part by a Dept.@ of Defense NDSEG Fellowship, an MIT Karl Taylor Compton Fellowship, and by the Materials Research Science and Engineering Center program of the National Science Foundation under award DMR-9400334. Code for the Cell Broadband Engine was graciously donated to the FFTW project by the IBM Austin Research Lab and included in fftw-3.2. (This code was removed in fftw-3.3.) Code for the MIPS paired-single SIMD support was graciously donated to the FFTW project by CodeSourcery, Inc. We are grateful to Sun Microsystems Inc.@ for its donation of a cluster of 9 8-processor Ultra HPC 5000 SMPs (24 Gflops peak). These machines served as the primary platform for the development of early versions of FFTW. We thank Intel Corporation for donating a four-processor Pentium Pro machine. We thank the GNU/Linux community for giving us a decent OS to run on that machine. We are thankful to the AMD corporation for donating an AMD Athlon XP 1700+ computer to the FFTW project. We thank the Compaq/HP testdrive program and VA Software Corporation (SourceForge.net) for providing remote access to machines that were used to test FFTW. The @code{genfft} suite of code generators was written using Objective Caml, a dialect of ML. Objective Caml is a small and elegant language developed by Xavier Leroy. The implementation is available from @uref{http://caml.inria.fr/, @code{http://caml.inria.fr/}}. In previous releases of FFTW, @code{genfft} was written in Caml Light, by the same authors. An even earlier implementation of @code{genfft} was written in Scheme, but Caml is definitely better for this kind of application. @cindex Caml @cindex LISP FFTW uses many tools from the GNU project, including @code{automake}, @code{texinfo}, and @code{libtool}. Prof.@ Charles E.@ Leiserson of MIT provided continuous support and encouragement. This program would not exist without him. Charles also proposed the name ``codelets'' for the basic FFT blocks. @cindex codelet Prof.@ John D.@ Joannopoulos of MIT demonstrated continuing tolerance of Steven's ``extra-curricular'' computer-science activities, as well as remarkable creativity in working them into his grant proposals. Steven's physics degree would not exist without him. Franz Franchetti wrote SIMD extensions to FFTW 2, which eventually led to the SIMD support in FFTW 3. Stefan Kral wrote most of the K7 code generator distributed with FFTW 3.0.x and 3.1.x. Andrew Sterian contributed the Windows timing code in FFTW 2. Didier Miras reported a bug in the test procedure used in FFTW 1.2. We now use a completely different test algorithm by Funda Ergun that does not require a separate FFT program to compare against. Wolfgang Reimer contributed the Pentium cycle counter and a few fixes that help portability. Ming-Chang Liu uncovered a well-hidden bug in the complex transforms of FFTW 2.0 and supplied a patch to correct it. The FFTW FAQ was written in @code{bfnn} (Bizarre Format With No Name) and formatted using the tools developed by Ian Jackson for the Linux FAQ. @emph{We are especially thankful to all of our users for their continuing support, feedback, and interest during our development of FFTW.} fftw-3.3.3/doc/cindex.texi0000644000175000017500000000014512054410050012310 00000000000000@node Concept Index, Library Index, License and Copyright, Top @chapter Concept Index @printindex cp fftw-3.3.3/doc/fftw3.texi0000644000175000017500000001157412054410050012077 00000000000000\input texinfo @c -*-texinfo-*- @c Update by C-x C-e on: (texinfo-multiple-files-update "fftw3.texi" nil t) @setfilename fftw3.info @include version.texi @settitle FFTW @value{VERSION} @setchapternewpage odd @c define constant index (ct) @defcodeindex ct @syncodeindex ct fn @syncodeindex vr fn @syncodeindex pg fn @syncodeindex tp fn @c define foreign function index (ff) @defcodeindex ff @syncodeindex ff cp @c define foreign constant index (fc) @defcodeindex fc @syncodeindex fc cp @c define foreign program index (fp) @defcodeindex fp @syncodeindex fp cp @comment %**end of header @iftex @paragraphindent 0 @parskip=@medskipamount @end iftex @macro Onlogn @ifinfo O(n log n) @end ifinfo @html O(n log n) @end html @tex $O(n \\log n)$ @end tex @end macro @macro ndims @ifinfo n[0] x n[1] x n[2] x ... x n[d-1] @end ifinfo @html n0 × n1 × n2 × … × nd-1 @end html @tex $n_0 \\times n_1 \\times n_2 \\times \\cdots \\times n_{d-1}$ @end tex @end macro @macro ndimshalf @ifinfo n[0] x n[1] x n[2] x ... x (n[d-1]/2 + 1) @end ifinfo @html n0 × n1 × n2 × … × (nd-1/2 + 1) @end html @tex $n_0 \\times n_1 \\times n_2 \\times \\cdots \\times (n_{d-1}/2 + 1)$ @end tex @end macro @macro ndimspad @ifinfo n[0] x n[1] x n[2] x ... x [2 (n[d-1]/2 + 1)] @end ifinfo @html n0 × n1 × n2 × … × [2 (nd-1/2 + 1)] @end html @tex $n_0 \\times n_1 \\times n_2 \\times \\cdots \\times [2(n_{d-1}/2 + 1)]$ @end tex @end macro @macro twodims{d1, d2} @ifinfo \d1\ x \d2\ @end ifinfo @html \d1\ × \d2\ @end html @tex $\d1\ \\times \d2\$ @end tex @end macro @macro threedims{d1, d2, d3} @ifinfo \d1\ x \d2\ x \d3\ @end ifinfo @html \d1\ × \d2\ × \d3\ @end html @tex $\d1\ \\times \d2\ \\times \d3\$ @end tex @end macro @macro dimk{k} @ifinfo n[\k\] @end ifinfo @html n\k\ @end html @tex $n_\k\$ @end tex @end macro @macro ndimstrans @ifinfo n[1] x n[0] x n[2] x ... x n[d-1] @end ifinfo @html n1 × n0 × n2 ×…× nd-1 @end html @tex $n_1 \\times n_0 \\times n_2 \\times \\cdots \\times n_{d-1}$ @end tex @end macro @copying This manual is for FFTW (version @value{VERSION}, @value{UPDATED}). Copyright @copyright{} 2003 Matteo Frigo. Copyright @copyright{} 2003 Massachusetts Institute of Technology. @quotation Permission is granted to make and distribute verbatim copies of this manual provided the copyright notice and this permission notice are preserved on all copies. Permission is granted to copy and distribute modified versions of this manual under the conditions for verbatim copying, provided that the entire resulting derived work is distributed under the terms of a permission notice identical to this one. Permission is granted to copy and distribute translations of this manual into another language, under the above conditions for modified versions, except that this permission notice may be stated in a translation approved by the Free Software Foundation. @end quotation @end copying @dircategory Texinfo documentation system @direntry * fftw3: (fftw3). FFTW User's Manual. @end direntry @titlepage @title FFTW @subtitle for version @value{VERSION}, @value{UPDATED} @author{Matteo Frigo} @author{Steven G. Johnson} @page @vskip 0pt plus 1filll @insertcopying @end titlepage @contents @ifnottex @node Top, Introduction, (dir), (dir) @top FFTW User Manual Welcome to FFTW, the Fastest Fourier Transform in the West. FFTW is a collection of fast C routines to compute the discrete Fourier transform. This manual documents FFTW version @value{VERSION}. @end ifnottex @menu * Introduction:: * Tutorial:: * Other Important Topics:: * FFTW Reference:: * Multi-threaded FFTW:: * Distributed-memory FFTW with MPI:: * Calling FFTW from Modern Fortran:: * Calling FFTW from Legacy Fortran:: * Upgrading from FFTW version 2:: * Installation and Customization:: * Acknowledgments:: * License and Copyright:: * Concept Index:: * Library Index:: @end menu @c ************************************************************ @include intro.texi @include tutorial.texi @include other.texi @include reference.texi @include threads.texi @include mpi.texi @include modern-fortran.texi @include legacy-fortran.texi @include upgrading.texi @include install.texi @include acknowledgements.texi @include license.texi @include cindex.texi @include findex.texi @c ************************************************************ @bye fftw-3.3.3/doc/findex.texi0000644000175000017500000000012012054410050012304 00000000000000@node Library Index, , Concept Index, Top @chapter Library Index @printindex fn fftw-3.3.3/doc/install.texi0000644000175000017500000004126412054410050012513 00000000000000@node Installation and Customization, Acknowledgments, Upgrading from FFTW version 2, Top @chapter Installation and Customization @cindex installation This chapter describes the installation and customization of FFTW, the latest version of which may be downloaded from @uref{http://www.fftw.org, the FFTW home page}. In principle, FFTW should work on any system with an ANSI C compiler (@code{gcc} is fine). However, planner time is drastically reduced if FFTW can exploit a hardware cycle counter; FFTW comes with cycle-counter support for all modern general-purpose CPUs, but you may need to add a couple of lines of code if your compiler is not yet supported (@pxref{Cycle Counters}). (On Unix, there will be a warning at the end of the @code{configure} output if no cycle counter is found.) @cindex cycle counter @cindex compiler @cindex portability Installation of FFTW is simplest if you have a Unix or a GNU system, such as GNU/Linux, and we describe this case in the first section below, including the use of special configuration options to e.g. install different precisions or exploit optimizations for particular architectures (e.g. SIMD). Compilation on non-Unix systems is a more manual process, but we outline the procedure in the second section. It is also likely that pre-compiled binaries will be available for popular systems. Finally, we describe how you can customize FFTW for particular needs by generating @emph{codelets} for fast transforms of sizes not supported efficiently by the standard FFTW distribution. @cindex codelet @menu * Installation on Unix:: * Installation on non-Unix systems:: * Cycle Counters:: * Generating your own code:: @end menu @c ------------------------------------------------------------ @node Installation on Unix, Installation on non-Unix systems, Installation and Customization, Installation and Customization @section Installation on Unix FFTW comes with a @code{configure} program in the GNU style. Installation can be as simple as: @fpindex configure @example ./configure make make install @end example This will build the uniprocessor complex and real transform libraries along with the test programs. (We recommend that you use GNU @code{make} if it is available; on some systems it is called @code{gmake}.) The ``@code{make install}'' command installs the fftw and rfftw libraries in standard places, and typically requires root privileges (unless you specify a different install directory with the @code{--prefix} flag to @code{configure}). You can also type ``@code{make check}'' to put the FFTW test programs through their paces. If you have problems during configuration or compilation, you may want to run ``@code{make distclean}'' before trying again; this ensures that you don't have any stale files left over from previous compilation attempts. The @code{configure} script chooses the @code{gcc} compiler by default, if it is available; you can select some other compiler with: @example ./configure CC="@r{@i{}}" @end example The @code{configure} script knows good @code{CFLAGS} (C compiler flags) @cindex compiler flags for a few systems. If your system is not known, the @code{configure} script will print out a warning. In this case, you should re-configure FFTW with the command @example ./configure CFLAGS="@r{@i{}}" @end example and then compile as usual. If you do find an optimal set of @code{CFLAGS} for your system, please let us know what they are (along with the output of @code{config.guess}) so that we can include them in future releases. @code{configure} supports all the standard flags defined by the GNU Coding Standards; see the @code{INSTALL} file in FFTW or @uref{http://www.gnu.org/prep/standards/html_node/index.html, the GNU web page}. Note especially @code{--help} to list all flags and @code{--enable-shared} to create shared, rather than static, libraries. @code{configure} also accepts a few FFTW-specific flags, particularly: @itemize @bullet @item @cindex precision @code{--enable-float}: Produces a single-precision version of FFTW (@code{float}) instead of the default double-precision (@code{double}). @xref{Precision}. @item @cindex precision @code{--enable-long-double}: Produces a long-double precision version of FFTW (@code{long double}) instead of the default double-precision (@code{double}). The @code{configure} script will halt with an error message if @code{long double} is the same size as @code{double} on your machine/compiler. @xref{Precision}. @item @cindex precision @code{--enable-quad-precision}: Produces a quadruple-precision version of FFTW using the nonstandard @code{__float128} type provided by @code{gcc} 4.6 or later on x86, x86-64, and Itanium architectures, instead of the default double-precision (@code{double}). The @code{configure} script will halt with an error message if the compiler is not @code{gcc} version 4.6 or later or if @code{gcc}'s @code{libquadmath} library is not installed. @xref{Precision}. @item @cindex threads @code{--enable-threads}: Enables compilation and installation of the FFTW threads library (@pxref{Multi-threaded FFTW}), which provides a simple interface to parallel transforms for SMP systems. By default, the threads routines are not compiled. @item @code{--enable-openmp}: Like @code{--enable-threads}, but using OpenMP compiler directives in order to induce parallelism rather than spawning its own threads directly, and installing an @samp{fftw3_omp} library rather than an @samp{fftw3_threads} library (@pxref{Multi-threaded FFTW}). You can use both @code{--enable-openmp} and @code{--enable-threads} since they compile/install libraries with different names. By default, the OpenMP routines are not compiled. @item @code{--with-combined-threads}: By default, if @code{--enable-threads} is used, the threads support is compiled into a separate library that must be linked in addition to the main FFTW library. This is so that users of the serial library do not need to link the system threads libraries. If @code{--with-combined-threads} is specified, however, then no separate threads library is created, and threads are included in the main FFTW library. This is mainly useful under Windows, where no system threads library is required and inter-library dependencies are problematic. @item @cindex MPI @code{--enable-mpi}: Enables compilation and installation of the FFTW MPI library (@pxref{Distributed-memory FFTW with MPI}), which provides parallel transforms for distributed-memory systems with MPI. (By default, the MPI routines are not compiled.) @xref{FFTW MPI Installation}. @item @cindex Fortran-callable wrappers @code{--disable-fortran}: Disables inclusion of legacy-Fortran wrapper routines (@pxref{Calling FFTW from Legacy Fortran}) in the standard FFTW libraries. These wrapper routines increase the library size by only a negligible amount, so they are included by default as long as the @code{configure} script finds a Fortran compiler on your system. (To specify a particular Fortran compiler @i{foo}, pass @code{F77=}@i{foo} to @code{configure}.) @item @code{--with-g77-wrappers}: By default, when Fortran wrappers are included, the wrappers employ the linking conventions of the Fortran compiler detected by the @code{configure} script. If this compiler is GNU @code{g77}, however, then @emph{two} versions of the wrappers are included: one with @code{g77}'s idiosyncratic convention of appending two underscores to identifiers, and one with the more common convention of appending only a single underscore. This way, the same FFTW library will work with both @code{g77} and other Fortran compilers, such as GNU @code{gfortran}. However, the converse is not true: if you configure with a different compiler, then the @code{g77}-compatible wrappers are not included. By specifying @code{--with-g77-wrappers}, the @code{g77}-compatible wrappers are included in addition to wrappers for whatever Fortran compiler @code{configure} finds. @fpindex g77 @item @code{--with-slow-timer}: Disables the use of hardware cycle counters, and falls back on @code{gettimeofday} or @code{clock}. This greatly worsens performance, and should generally not be used (unless you don't have a cycle counter but still really want an optimized plan regardless of the time). @xref{Cycle Counters}. @item @code{--enable-sse}, @code{--enable-sse2}, @code{--enable-avx}, @code{--enable-altivec}, @code{--enable-neon}: Enable the compilation of SIMD code for SSE (Pentium III+), SSE2 (Pentium IV+), AVX (Sandy Bridge, Interlagos), AltiVec (PowerPC G4+), NEON (some ARM processors). SSE, AltiVec, and NEON only work with @code{--enable-float} (above). SSE2 works in both single and double precision (and is simply SSE in single precision). The resulting code will @emph{still work} on earlier CPUs lacking the SIMD extensions (SIMD is automatically disabled, although the FFTW library is still larger). @itemize @minus @item These options require a compiler supporting SIMD extensions, and compiler support is always a bit flaky: see the FFTW FAQ for a list of compiler versions that have problems compiling FFTW. @item With AltiVec and @code{gcc}, you may have to use the @code{-mabi=altivec} option when compiling any code that links to FFTW, in order to properly align the stack; otherwise, FFTW could crash when it tries to use an AltiVec feature. (This is not necessary on MacOS X.) @item With SSE/SSE2 and @code{gcc}, you should use a version of gcc that properly aligns the stack when compiling any code that links to FFTW. By default, @code{gcc} 2.95 and later versions align the stack as needed, but you should not compile FFTW with the @code{-Os} option or the @code{-mpreferred-stack-boundary} option with an argument less than 4. @item Because of the large variety of ARM processors and ABIs, FFTW does not attempt to guess the correct @code{gcc} flags for generating NEON code. In general, you will have to provide them on the command line. This command line is known to have worked at least once: @example ./configure --with-slow-timer --host=arm-linux-gnueabi \ --enable-single --enable-neon \ "CC=arm-linux-gnueabi-gcc -march=armv7-a -mfloat-abi=softfp" @end example @end itemize @end itemize @cindex compiler To force @code{configure} to use a particular C compiler @i{foo} (instead of the default, usually @code{gcc}), pass @code{CC=}@i{foo} to the @code{configure} script; you may also need to set the flags via the variable @code{CFLAGS} as described above. @cindex compiler flags @c ------------------------------------------------------------ @node Installation on non-Unix systems, Cycle Counters, Installation on Unix, Installation and Customization @section Installation on non-Unix systems It should be relatively straightforward to compile FFTW even on non-Unix systems lacking the niceties of a @code{configure} script. Basically, you need to edit the @code{config.h} header (copy it from @code{config.h.in}) to @code{#define} the various options and compiler characteristics, and then compile all the @samp{.c} files in the relevant directories. The @code{config.h} header contains about 100 options to set, each one initially an @code{#undef}, each documented with a comment, and most of them fairly obvious. For most of the options, you should simply @code{#define} them to @code{1} if they are applicable, although a few options require a particular value (e.g. @code{SIZEOF_LONG_LONG} should be defined to the size of the @code{long long} type, in bytes, or zero if it is not supported). We will likely post some sample @code{config.h} files for various operating systems and compilers for you to use (at least as a starting point). Please let us know if you have to hand-create a configuration file (and/or a pre-compiled binary) that you want to share. To create the FFTW library, you will then need to compile all of the @samp{.c} files in the @code{kernel}, @code{dft}, @code{dft/scalar}, @code{dft/scalar/codelets}, @code{rdft}, @code{rdft/scalar}, @code{rdft/scalar/r2cf}, @code{rdft/scalar/r2cb}, @code{rdft/scalar/r2r}, @code{reodft}, and @code{api} directories. If you are compiling with SIMD support (e.g. you defined @code{HAVE_SSE2} in @code{config.h}), then you also need to compile the @code{.c} files in the @code{simd-support}, @code{@{dft,rdft@}/simd}, @code{@{dft,rdft@}/simd/*} directories. Once these files are all compiled, link them into a library, or a shared library, or directly into your program. To compile the FFTW test program, additionally compile the code in the @code{libbench2/} directory, and link it into a library. Then compile the code in the @code{tests/} directory and link it to the @code{libbench2} and FFTW libraries. To compile the @code{fftw-wisdom} (command-line) tool (@pxref{Wisdom Utilities}), compile @code{tools/fftw-wisdom.c} and link it to the @code{libbench2} and FFTW libraries @c ------------------------------------------------------------ @node Cycle Counters, Generating your own code, Installation on non-Unix systems, Installation and Customization @section Cycle Counters @cindex cycle counter FFTW's planner actually executes and times different possible FFT algorithms in order to pick the fastest plan for a given @math{n}. In order to do this in as short a time as possible, however, the timer must have a very high resolution, and to accomplish this we employ the hardware @dfn{cycle counters} that are available on most CPUs. Currently, FFTW supports the cycle counters on x86, PowerPC/POWER, Alpha, UltraSPARC (SPARC v9), IA64, PA-RISC, and MIPS processors. @cindex compiler Access to the cycle counters, unfortunately, is a compiler and/or operating-system dependent task, often requiring inline assembly language, and it may be that your compiler is not supported. If you are @emph{not} supported, FFTW will by default fall back on its estimator (effectively using @code{FFTW_ESTIMATE} for all plans). @ctindex FFTW_ESTIMATE You can add support by editing the file @code{kernel/cycle.h}; normally, this will involve adapting one of the examples already present in order to use the inline-assembler syntax for your C compiler, and will only require a couple of lines of code. Anyone adding support for a new system to @code{cycle.h} is encouraged to email us at @email{fftw@@fftw.org}. If a cycle counter is not available on your system (e.g. some embedded processor), and you don't want to use estimated plans, as a last resort you can use the @code{--with-slow-timer} option to @code{configure} (on Unix) or @code{#define WITH_SLOW_TIMER} in @code{config.h} (elsewhere). This will use the much lower-resolution @code{gettimeofday} function, or even @code{clock} if the former is unavailable, and planning will be extremely slow. @c ------------------------------------------------------------ @node Generating your own code, , Cycle Counters, Installation and Customization @section Generating your own code @cindex code generator The directory @code{genfft} contains the programs that were used to generate FFTW's ``codelets,'' which are hard-coded transforms of small sizes. @cindex codelet We do not expect casual users to employ the generator, which is a rather sophisticated program that generates directed acyclic graphs of FFT algorithms and performs algebraic simplifications on them. It was written in Objective Caml, a dialect of ML, which is available at @uref{http://caml.inria.fr/ocaml/index.en.html}. @cindex Caml If you have Objective Caml installed (along with recent versions of GNU @code{autoconf}, @code{automake}, and @code{libtool}), then you can change the set of codelets that are generated or play with the generation options. The set of generated codelets is specified by the @code{@{dft,rdft@}/@{codelets,simd@}/*/Makefile.am} files. For example, you can add efficient REDFT codelets of small sizes by modifying @code{rdft/codelets/r2r/Makefile.am}. @cindex REDFT After you modify any @code{Makefile.am} files, you can type @code{sh bootstrap.sh} in the top-level directory followed by @code{make} to re-generate the files. We do not provide more details about the code-generation process, since we do not expect that most users will need to generate their own code. However, feel free to contact us at @email{fftw@@fftw.org} if you are interested in the subject. @cindex monadic programming You might find it interesting to learn Caml and/or some modern programming techniques that we used in the generator (including monadic programming), especially if you heard the rumor that Java and object-oriented programming are the latest advancement in the field. The internal operation of the codelet generator is described in the paper, ``A Fast Fourier Transform Compiler,'' by M. Frigo, which is available from the @uref{http://www.fftw.org,FFTW home page} and also appeared in the @cite{Proceedings of the 1999 ACM SIGPLAN Conference on Programming Language Design and Implementation (PLDI)}. fftw-3.3.3/doc/intro.texi0000644000175000017500000001704512054410050012200 00000000000000@node Introduction, Tutorial, Top, Top @chapter Introduction This manual documents version @value{VERSION} of FFTW, the @emph{Fastest Fourier Transform in the West}. FFTW is a comprehensive collection of fast C routines for computing the discrete Fourier transform (DFT) and various special cases thereof. @cindex discrete Fourier transform @cindex DFT @itemize @bullet @item FFTW computes the DFT of complex data, real data, even- or odd-symmetric real data (these symmetric transforms are usually known as the discrete cosine or sine transform, respectively), and the discrete Hartley transform (DHT) of real data. @item The input data can have arbitrary length. FFTW employs @Onlogn{} algorithms for all lengths, including prime numbers. @item FFTW supports arbitrary multi-dimensional data. @item FFTW supports the SSE, SSE2, AVX, Altivec, and MIPS PS instruction sets. @item FFTW includes parallel (multi-threaded) transforms for shared-memory systems. @item Starting with version 3.3, FFTW includes distributed-memory parallel transforms using MPI. @end itemize We assume herein that you are familiar with the properties and uses of the DFT that are relevant to your application. Otherwise, see e.g. @cite{The Fast Fourier Transform and Its Applications} by E. O. Brigham (Prentice-Hall, Englewood Cliffs, NJ, 1988). @uref{http://www.fftw.org, Our web page} also has links to FFT-related information online. @cindex FFTW @c TODO: revise. We don't need to brag any longer @c @c FFTW is usually faster (and sometimes much faster) than all other @c freely-available Fourier transform programs found on the Net. It is @c competitive with (and often faster than) the FFT codes in Sun's @c Performance Library, IBM's ESSL library, HP's CXML library, and @c Intel's MKL library, which are targeted at specific machines. @c Moreover, FFTW's performance is @emph{portable}. Indeed, FFTW is @c unique in that it automatically adapts itself to your machine, your @c cache, the size of your memory, your number of registers, and all the @c other factors that normally make it impossible to optimize a program @c for more than one machine. An extensive comparison of FFTW's @c performance with that of other Fourier transform codes has been made, @c and the results are available on the Web at @c @uref{http://fftw.org/benchfft, the benchFFT home page}. @c @cindex benchmark @c @fpindex benchfft In order to use FFTW effectively, you need to learn one basic concept of FFTW's internal structure: FFTW does not use a fixed algorithm for computing the transform, but instead it adapts the DFT algorithm to details of the underlying hardware in order to maximize performance. Hence, the computation of the transform is split into two phases. First, FFTW's @dfn{planner} ``learns'' the fastest way to compute the transform on your machine. The planner @cindex planner produces a data structure called a @dfn{plan} that contains this @cindex plan information. Subsequently, the plan is @dfn{executed} @cindex execute to transform the array of input data as dictated by the plan. The plan can be reused as many times as needed. In typical high-performance applications, many transforms of the same size are computed and, consequently, a relatively expensive initialization of this sort is acceptable. On the other hand, if you need a single transform of a given size, the one-time cost of the planner becomes significant. For this case, FFTW provides fast planners based on heuristics or on previously computed plans. FFTW supports transforms of data with arbitrary length, rank, multiplicity, and a general memory layout. In simple cases, however, this generality may be unnecessary and confusing. Consequently, we organized the interface to FFTW into three levels of increasing generality. @itemize @bullet @item The @dfn{basic interface} computes a single transform of contiguous data. @item The @dfn{advanced interface} computes transforms of multiple or strided arrays. @item The @dfn{guru interface} supports the most general data layouts, multiplicities, and strides. @end itemize We expect that most users will be best served by the basic interface, whereas the guru interface requires careful attention to the documentation to avoid problems. @cindex basic interface @cindex advanced interface @cindex guru interface Besides the automatic performance adaptation performed by the planner, it is also possible for advanced users to customize FFTW manually. For example, if code space is a concern, we provide a tool that links only the subset of FFTW needed by your application. Conversely, you may need to extend FFTW because the standard distribution is not sufficient for your needs. For example, the standard FFTW distribution works most efficiently for arrays whose size can be factored into small primes (@math{2}, @math{3}, @math{5}, and @math{7}), and otherwise it uses a slower general-purpose routine. If you need efficient transforms of other sizes, you can use FFTW's code generator, which produces fast C programs (``codelets'') for any particular array size you may care about. @cindex code generator @cindex codelet For example, if you need transforms of size @ifinfo @math{513 = 19 x 3^3}, @end ifinfo @tex $513 = 19 \cdot 3^3$, @end tex @html 513 = 19*33, @end html you can customize FFTW to support the factor @math{19} efficiently. For more information regarding FFTW, see the paper, ``The Design and Implementation of FFTW3,'' by M. Frigo and S. G. Johnson, which was an invited paper in @cite{Proc. IEEE} @b{93} (2), p. 216 (2005). The code generator is described in the paper ``A fast Fourier transform compiler'', @cindex compiler by M. Frigo, in the @cite{Proceedings of the 1999 ACM SIGPLAN Conference on Programming Language Design and Implementation (PLDI), Atlanta, Georgia, May 1999}. These papers, along with the latest version of FFTW, the FAQ, benchmarks, and other links, are available at @uref{http://www.fftw.org, the FFTW home page}. The current version of FFTW incorporates many good ideas from the past thirty years of FFT literature. In one way or another, FFTW uses the Cooley-Tukey algorithm, the prime factor algorithm, Rader's algorithm for prime sizes, and a split-radix algorithm (with a ``conjugate-pair'' variation pointed out to us by Dan Bernstein). FFTW's code generator also produces new algorithms that we do not completely understand. @cindex algorithm The reader is referred to the cited papers for the appropriate references. The rest of this manual is organized as follows. We first discuss the sequential (single-processor) implementation. We start by describing the basic interface/features of FFTW in @ref{Tutorial}. Next, @ref{Other Important Topics} discusses data alignment (@pxref{SIMD alignment and fftw_malloc}), the storage scheme of multi-dimensional arrays (@pxref{Multi-dimensional Array Format}), and FFTW's mechanism for storing plans on disk (@pxref{Words of Wisdom-Saving Plans}). Next, @ref{FFTW Reference} provides comprehensive documentation of all FFTW's features. Parallel transforms are discussed in their own chapters: @ref{Multi-threaded FFTW} and @ref{Distributed-memory FFTW with MPI}. Fortran programmers can also use FFTW, as described in @ref{Calling FFTW from Legacy Fortran} and @ref{Calling FFTW from Modern Fortran}. @ref{Installation and Customization} explains how to install FFTW in your computer system and how to adapt FFTW to your needs. License and copyright information is given in @ref{License and Copyright}. Finally, we thank all the people who helped us in @ref{Acknowledgments}. fftw-3.3.3/doc/legacy-fortran.texi0000644000175000017500000003572712054410050013771 00000000000000@node Calling FFTW from Legacy Fortran, Upgrading from FFTW version 2, Calling FFTW from Modern Fortran, Top @chapter Calling FFTW from Legacy Fortran @cindex Fortran interface This chapter describes the interface to FFTW callable by Fortran code in older compilers not supporting the Fortran 2003 C interoperability features (@pxref{Calling FFTW from Modern Fortran}). This interface has the major disadvantage that it is not type-checked, so if you mistake the argument types or ordering then your program will not have any compiler errors, and will likely crash at runtime. So, greater care is needed. Also, technically interfacing older Fortran versions to C is nonstandard, but in practice we have found that the techniques used in this chapter have worked with all known Fortran compilers for many years. The legacy Fortran interface differs from the C interface only in the prefix (@samp{dfftw_} instead of @samp{fftw_} in double precision) and a few other minor details. This Fortran interface is included in the FFTW libraries by default, unless a Fortran compiler isn't found on your system or @code{--disable-fortran} is included in the @code{configure} flags. We assume here that the reader is already familiar with the usage of FFTW in C, as described elsewhere in this manual. The MPI parallel interface to FFTW is @emph{not} currently available to legacy Fortran. @menu * Fortran-interface routines:: * FFTW Constants in Fortran:: * FFTW Execution in Fortran:: * Fortran Examples:: * Wisdom of Fortran?:: @end menu @c ------------------------------------------------------- @node Fortran-interface routines, FFTW Constants in Fortran, Calling FFTW from Legacy Fortran, Calling FFTW from Legacy Fortran @section Fortran-interface routines Nearly all of the FFTW functions have Fortran-callable equivalents. The name of the legacy Fortran routine is the same as that of the corresponding C routine, but with the @samp{fftw_} prefix replaced by @samp{dfftw_}.@footnote{Technically, Fortran 77 identifiers are not allowed to have more than 6 characters, nor may they contain underscores. Any compiler that enforces this limitation doesn't deserve to link to FFTW.} The single and long-double precision versions use @samp{sfftw_} and @samp{lfftw_}, respectively, instead of @samp{fftwf_} and @samp{fftwl_}; quadruple precision (@code{real*16}) is available on some systems as @samp{fftwq_} (@pxref{Precision}). (Note that @code{long double} on x86 hardware is usually at most 80-bit extended precision, @emph{not} quadruple precision.) For the most part, all of the arguments to the functions are the same, with the following exceptions: @itemize @bullet @item @code{plan} variables (what would be of type @code{fftw_plan} in C), must be declared as a type that is at least as big as a pointer (address) on your machine. We recommend using @code{integer*8} everywhere, since this should always be big enough. @cindex portability @item Any function that returns a value (e.g. @code{fftw_plan_dft}) is converted into a @emph{subroutine}. The return value is converted into an additional @emph{first} parameter of this subroutine.@footnote{The reason for this is that some Fortran implementations seem to have trouble with C function return values, and vice versa.} @item @cindex column-major The Fortran routines expect multi-dimensional arrays to be in @emph{column-major} order, which is the ordinary format of Fortran arrays (@pxref{Multi-dimensional Array Format}). They do this transparently and costlessly simply by reversing the order of the dimensions passed to FFTW, but this has one important consequence for multi-dimensional real-complex transforms, discussed below. @item Wisdom import and export is somewhat more tricky because one cannot easily pass files or strings between C and Fortran; see @ref{Wisdom of Fortran?}. @item Legacy Fortran cannot use the @code{fftw_malloc} dynamic-allocation routine. If you want to exploit the SIMD FFTW (@pxref{SIMD alignment and fftw_malloc}), you'll need to figure out some other way to ensure that your arrays are at least 16-byte aligned. @item @tindex fftw_iodim @cindex guru interface Since Fortran 77 does not have data structures, the @code{fftw_iodim} structure from the guru interface (@pxref{Guru vector and transform sizes}) must be split into separate arguments. In particular, any @code{fftw_iodim} array arguments in the C guru interface become three integer array arguments (@code{n}, @code{is}, and @code{os}) in the Fortran guru interface, all of whose lengths should be equal to the corresponding @code{rank} argument. @item The guru planner interface in Fortran does @emph{not} do any automatic translation between column-major and row-major; you are responsible for setting the strides etcetera to correspond to your Fortran arrays. However, as a slight bug that we are preserving for backwards compatibility, the @samp{plan_guru_r2r} in Fortran @emph{does} reverse the order of its @code{kind} array parameter, so the @code{kind} array of that routine should be in the reverse of the order of the iodim arrays (see above). @end itemize In general, you should take care to use Fortran data types that correspond to (i.e. are the same size as) the C types used by FFTW. In practice, this correspondence is usually straightforward (i.e. @code{integer} corresponds to @code{int}, @code{real} corresponds to @code{float}, etcetera). The native Fortran double/single-precision complex type should be compatible with @code{fftw_complex}/@code{fftwf_complex}. Such simple correspondences are assumed in the examples below. @cindex portability @c ------------------------------------------------------- @node FFTW Constants in Fortran, FFTW Execution in Fortran, Fortran-interface routines, Calling FFTW from Legacy Fortran @section FFTW Constants in Fortran When creating plans in FFTW, a number of constants are used to specify options, such as @code{FFTW_MEASURE} or @code{FFTW_ESTIMATE}. The same constants must be used with the wrapper routines, but of course the C header files where the constants are defined can't be incorporated directly into Fortran code. Instead, we have placed Fortran equivalents of the FFTW constant definitions in the file @code{fftw3.f}, which can be found in the same directory as @code{fftw3.h}. If your Fortran compiler supports a preprocessor of some sort, you should be able to @code{include} or @code{#include} this file; otherwise, you can paste it directly into your code. @cindex flags In C, you combine different flags (like @code{FFTW_PRESERVE_INPUT} and @code{FFTW_MEASURE}) using the @samp{@code{|}} operator; in Fortran you should just use @samp{@code{+}}. (Take care not to add in the same flag more than once, though. Alternatively, you can use the @code{ior} intrinsic function standardized in Fortran 95.) @c ------------------------------------------------------- @node FFTW Execution in Fortran, Fortran Examples, FFTW Constants in Fortran, Calling FFTW from Legacy Fortran @section FFTW Execution in Fortran In C, in order to use a plan, one normally calls @code{fftw_execute}, which executes the plan to perform the transform on the input/output arrays passed when the plan was created (@pxref{Using Plans}). The corresponding subroutine call in legacy Fortran is: @example call dfftw_execute(plan) @end example @findex dfftw_execute However, we have had reports that this causes problems with some recent optimizing Fortran compilers. The problem is, because the input/output arrays are not passed as explicit arguments to @code{dfftw_execute}, the semantics of Fortran (unlike C) allow the compiler to assume that the input/output arrays are not changed by @code{dfftw_execute}. As a consequence, certain compilers end up optimizing out or repositioning the call to @code{dfftw_execute}, assuming incorrectly that it does nothing. There are various workarounds to this, but the safest and simplest thing is to not use @code{dfftw_execute} in Fortran. Instead, use the functions described in @ref{New-array Execute Functions}, which take the input/output arrays as explicit arguments. For example, if the plan is for a complex-data DFT and was created for the arrays @code{in} and @code{out}, you would do: @example call dfftw_execute_dft(plan, in, out) @end example @findex dfftw_execute_dft There are a few things to be careful of, however: @itemize @bullet @item You must use the correct type of execute function, matching the way the plan was created. Complex DFT plans should use @code{dfftw_execute_dft}, Real-input (r2c) DFT plans should use use @code{dfftw_execute_dft_r2c}, and real-output (c2r) DFT plans should use @code{dfftw_execute_dft_c2r}. The various r2r plans should use @code{dfftw_execute_r2r}. @item You should normally pass the same input/output arrays that were used when creating the plan. This is always safe. @item @emph{If} you pass @emph{different} input/output arrays compared to those used when creating the plan, you must abide by all the restrictions of the new-array execute functions (@pxref{New-array Execute Functions}). The most difficult of these, in Fortran, is the requirement that the new arrays have the same alignment as the original arrays, because there seems to be no way in legacy Fortran to obtain guaranteed-aligned arrays (analogous to @code{fftw_malloc} in C). You can, of course, use the @code{FFTW_UNALIGNED} flag when creating the plan, in which case the plan does not depend on the alignment, but this may sacrifice substantial performance on architectures (like x86) with SIMD instructions (@pxref{SIMD alignment and fftw_malloc}). @ctindex FFTW_UNALIGNED @end itemize @c ------------------------------------------------------- @node Fortran Examples, Wisdom of Fortran?, FFTW Execution in Fortran, Calling FFTW from Legacy Fortran @section Fortran Examples In C, you might have something like the following to transform a one-dimensional complex array: @example fftw_complex in[N], out[N]; fftw_plan plan; plan = fftw_plan_dft_1d(N,in,out,FFTW_FORWARD,FFTW_ESTIMATE); fftw_execute(plan); fftw_destroy_plan(plan); @end example In Fortran, you would use the following to accomplish the same thing: @example double complex in, out dimension in(N), out(N) integer*8 plan call dfftw_plan_dft_1d(plan,N,in,out,FFTW_FORWARD,FFTW_ESTIMATE) call dfftw_execute_dft(plan, in, out) call dfftw_destroy_plan(plan) @end example @findex dfftw_plan_dft_1d @findex dfftw_execute_dft @findex dfftw_destroy_plan Notice how all routines are called as Fortran subroutines, and the plan is returned via the first argument to @code{dfftw_plan_dft_1d}. Notice also that we changed @code{fftw_execute} to @code{dfftw_execute_dft} (@pxref{FFTW Execution in Fortran}). To do the same thing, but using 8 threads in parallel (@pxref{Multi-threaded FFTW}), you would simply prefix these calls with: @example integer iret call dfftw_init_threads(iret) call dfftw_plan_with_nthreads(8) @end example @findex dfftw_init_threads @findex dfftw_plan_with_nthreads (You might want to check the value of @code{iret}: if it is zero, it indicates an unlikely error during thread initialization.) To transform a three-dimensional array in-place with C, you might do: @example fftw_complex arr[L][M][N]; fftw_plan plan; plan = fftw_plan_dft_3d(L,M,N, arr,arr, FFTW_FORWARD, FFTW_ESTIMATE); fftw_execute(plan); fftw_destroy_plan(plan); @end example In Fortran, you would use this instead: @example double complex arr dimension arr(L,M,N) integer*8 plan call dfftw_plan_dft_3d(plan, L,M,N, arr,arr, & FFTW_FORWARD, FFTW_ESTIMATE) call dfftw_execute_dft(plan, arr, arr) call dfftw_destroy_plan(plan) @end example @findex dfftw_plan_dft_3d Note that we pass the array dimensions in the ``natural'' order in both C and Fortran. To transform a one-dimensional real array in Fortran, you might do: @example double precision in dimension in(N) double complex out dimension out(N/2 + 1) integer*8 plan call dfftw_plan_dft_r2c_1d(plan,N,in,out,FFTW_ESTIMATE) call dfftw_execute_dft_r2c(plan, in, out) call dfftw_destroy_plan(plan) @end example @findex dfftw_plan_dft_r2c_1d @findex dfftw_execute_dft_r2c To transform a two-dimensional real array, out of place, you might use the following: @example double precision in dimension in(M,N) double complex out dimension out(M/2 + 1, N) integer*8 plan call dfftw_plan_dft_r2c_2d(plan,M,N,in,out,FFTW_ESTIMATE) call dfftw_execute_dft_r2c(plan, in, out) call dfftw_destroy_plan(plan) @end example @findex dfftw_plan_dft_r2c_2d @strong{Important:} Notice that it is the @emph{first} dimension of the complex output array that is cut in half in Fortran, rather than the last dimension as in C. This is a consequence of the interface routines reversing the order of the array dimensions passed to FFTW so that the Fortran program can use its ordinary column-major order. @cindex column-major @cindex r2c/c2r multi-dimensional array format @c ------------------------------------------------------- @node Wisdom of Fortran?, , Fortran Examples, Calling FFTW from Legacy Fortran @section Wisdom of Fortran? In this section, we discuss how one can import/export FFTW wisdom (saved plans) to/from a Fortran program; we assume that the reader is already familiar with wisdom, as described in @ref{Words of Wisdom-Saving Plans}. @cindex portability The basic problem is that is difficult to (portably) pass files and strings between Fortran and C, so we cannot provide a direct Fortran equivalent to the @code{fftw_export_wisdom_to_file}, etcetera, functions. Fortran interfaces @emph{are} provided for the functions that do not take file/string arguments, however: @code{dfftw_import_system_wisdom}, @code{dfftw_import_wisdom}, @code{dfftw_export_wisdom}, and @code{dfftw_forget_wisdom}. @findex dfftw_import_system_wisdom @findex dfftw_import_wisdom @findex dfftw_export_wisdom @findex dfftw_forget_wisdom So, for example, to import the system-wide wisdom, you would do: @example integer isuccess call dfftw_import_system_wisdom(isuccess) @end example As usual, the C return value is turned into a first parameter; @code{isuccess} is non-zero on success and zero on failure (e.g. if there is no system wisdom installed). If you want to import/export wisdom from/to an arbitrary file or elsewhere, you can employ the generic @code{dfftw_import_wisdom} and @code{dfftw_export_wisdom} functions, for which you must supply a subroutine to read/write one character at a time. The FFTW package contains an example file @code{doc/f77_wisdom.f} demonstrating how to implement @code{import_wisdom_from_file} and @code{export_wisdom_to_file} subroutines in this way. (These routines cannot be compiled into the FFTW library itself, lest all FFTW-using programs be required to link with the Fortran I/O library.) fftw-3.3.3/doc/license.texi0000644000175000017500000000346212054410050012465 00000000000000@node License and Copyright, Concept Index, Acknowledgments, Top @chapter License and Copyright FFTW is Copyright @copyright{} 2003, 2007-11 Matteo Frigo, Copyright @copyright{} 2003, 2007-11 Massachusetts Institute of Technology. FFTW is free software; you can redistribute it and/or modify it under the terms of the GNU General Public License as published by the Free Software Foundation; either version 2 of the License, or (at your option) any later version. This program is distributed in the hope that it will be useful, but WITHOUT ANY WARRANTY; without even the implied warranty of MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the GNU General Public License for more details. You should have received a copy of the GNU General Public License along with this program; if not, write to the Free Software Foundation, Inc., 51 Franklin Street, Fifth Floor, Boston, MA 02110-1301 USA You can also find the @uref{http://www.gnu.org/licenses/gpl-2.0.html, GPL on the GNU web site}. In addition, we kindly ask you to acknowledge FFTW and its authors in any program or publication in which you use FFTW. (You are not @emph{required} to do so; it is up to your common sense to decide whether you want to comply with this request or not.) For general publications, we suggest referencing: Matteo Frigo and Steven G. Johnson, ``The design and implementation of FFTW3,'' @i{Proc. IEEE} @b{93} (2), 216--231 (2005). Non-free versions of FFTW are available under terms different from those of the General Public License. (e.g. they do not require you to accompany any object code using FFTW with the corresponding source code.) For these alternative terms you must purchase a license from MIT's Technology Licensing Office. Users interested in such a license should contact us (@email{fftw@@fftw.org}) for more information. fftw-3.3.3/doc/modern-fortran.texi0000644000175000017500000007306012054410050014001 00000000000000@node Calling FFTW from Modern Fortran, Calling FFTW from Legacy Fortran, Distributed-memory FFTW with MPI, Top @chapter Calling FFTW from Modern Fortran @cindex Fortran interface Fortran 2003 standardized ways for Fortran code to call C libraries, and this allows us to support a direct translation of the FFTW C API into Fortran. Compared to the legacy Fortran 77 interface (@pxref{Calling FFTW from Legacy Fortran}), this direct interface offers many advantages, especially compile-time type-checking and aligned memory allocation. As of this writing, support for these C interoperability features seems widespread, having been implemented in nearly all major Fortran compilers (e.g. GNU, Intel, IBM, Oracle/Solaris, Portland Group, NAG). @cindex portability This chapter documents that interface. For the most part, since this interface allows Fortran to call the C interface directly, the usage is identical to C translated to Fortran syntax. However, there are a few subtle points such as memory allocation, wisdom, and data types that deserve closer attention. @menu * Overview of Fortran interface:: * Reversing array dimensions:: * FFTW Fortran type reference:: * Plan execution in Fortran:: * Allocating aligned memory in Fortran:: * Accessing the wisdom API from Fortran:: * Defining an FFTW module:: @end menu @c ------------------------------------------------------- @node Overview of Fortran interface, Reversing array dimensions, Calling FFTW from Modern Fortran, Calling FFTW from Modern Fortran @section Overview of Fortran interface FFTW provides a file @code{fftw3.f03} that defines Fortran 2003 interfaces for all of its C routines, except for the MPI routines described elsewhere, which can be found in the same directory as @code{fftw3.h} (the C header file). In any Fortran subroutine where you want to use FFTW functions, you should begin with: @cindex iso_c_binding @example use, intrinsic :: iso_c_binding include 'fftw3.f03' @end example This includes the interface definitions and the standard @code{iso_c_binding} module (which defines the equivalents of C types). You can also put the FFTW functions into a module if you prefer (@pxref{Defining an FFTW module}). At this point, you can now call anything in the FFTW C interface directly, almost exactly as in C other than minor changes in syntax. For example: @findex fftw_plan_dft_2d @findex fftw_execute_dft @findex fftw_destroy_plan @example type(C_PTR) :: plan complex(C_DOUBLE_COMPLEX), dimension(1024,1000) :: in, out plan = fftw_plan_dft_2d(1000,1024, in,out, FFTW_FORWARD,FFTW_ESTIMATE) ... call fftw_execute_dft(plan, in, out) ... call fftw_destroy_plan(plan) @end example A few important things to keep in mind are: @itemize @bullet @item @tindex fftw_complex @ctindex C_PTR @ctindex C_INT @ctindex C_DOUBLE @ctindex C_DOUBLE_COMPLEX FFTW plans are @code{type(C_PTR)}. Other C types are mapped in the obvious way via the @code{iso_c_binding} standard: @code{int} turns into @code{integer(C_INT)}, @code{fftw_complex} turns into @code{complex(C_DOUBLE_COMPLEX)}, @code{double} turns into @code{real(C_DOUBLE)}, and so on. @xref{FFTW Fortran type reference}. @item Functions in C become functions in Fortran if they have a return value, and subroutines in Fortran otherwise. @item The ordering of the Fortran array dimensions must be @emph{reversed} when they are passed to the FFTW plan creation, thanks to differences in array indexing conventions (@pxref{Multi-dimensional Array Format}). This is @emph{unlike} the legacy Fortran interface (@pxref{Fortran-interface routines}), which reversed the dimensions for you. @xref{Reversing array dimensions}. @item @cindex alignment @cindex SIMD Using ordinary Fortran array declarations like this works, but may yield suboptimal performance because the data may not be not aligned to exploit SIMD instructions on modern proessors (@pxref{SIMD alignment and fftw_malloc}). Better performance will often be obtained by allocating with @samp{fftw_alloc}. @xref{Allocating aligned memory in Fortran}. @item @findex fftw_execute Similar to the legacy Fortran interface (@pxref{FFTW Execution in Fortran}), we currently recommend @emph{not} using @code{fftw_execute} but rather using the more specialized functions like @code{fftw_execute_dft} (@pxref{New-array Execute Functions}). However, you should execute the plan on the @code{same arrays} as the ones for which you created the plan, unless you are especially careful. @xref{Plan execution in Fortran}. To prevent you from using @code{fftw_execute} by mistake, the @code{fftw3.f03} file does not provide an @code{fftw_execute} interface declaration. @item @cindex flags Multiple planner flags are combined with @code{ior} (equivalent to @samp{|} in C). e.g. @code{FFTW_MEASURE | FFTW_DESTROY_INPUT} becomes @code{ior(FFTW_MEASURE, FFTW_DESTROY_INPUT)}. (You can also use @samp{+} as long as you don't try to include a given flag more than once.) @end itemize @menu * Extended and quadruple precision in Fortran:: @end menu @node Extended and quadruple precision in Fortran, , Overview of Fortran interface, Overview of Fortran interface @subsection Extended and quadruple precision in Fortran @cindex precision If FFTW is compiled in @code{long double} (extended) precision (@pxref{Installation and Customization}), you may be able to call the resulting @code{fftwl_} routines (@pxref{Precision}) from Fortran if your compiler supports the @code{C_LONG_DOUBLE_COMPLEX} type code. Because some Fortran compilers do not support @code{C_LONG_DOUBLE_COMPLEX}, the @code{fftwl_} declarations are segregated into a separate interface file @code{fftw3l.f03}, which you should include @emph{in addition} to @code{fftw3.f03} (which declares precision-independent @samp{FFTW_} constants): @cindex iso_c_binding @example use, intrinsic :: iso_c_binding include 'fftw3.f03' include 'fftw3l.f03' @end example We also support using the nonstandard @code{__float128} quadruple-precision type provided by recent versions of @code{gcc} on 32- and 64-bit x86 hardware (@pxref{Installation and Customization}), using the corresponding @code{real(16)} and @code{complex(16)} types supported by @code{gfortran}. The quadruple-precision @samp{fftwq_} functions (@pxref{Precision}) are declared in a @code{fftw3q.f03} interface file, which should be included in addition to @code{fftw3l.f03}, as above. You should also link with @code{-lfftw3q -lquadmath -lm} as in C. @c ------------------------------------------------------- @node Reversing array dimensions, FFTW Fortran type reference, Overview of Fortran interface, Calling FFTW from Modern Fortran @section Reversing array dimensions @cindex row-major @cindex column-major A minor annoyance in calling FFTW from Fortran is that FFTW's array dimensions are defined in the C convention (row-major order), while Fortran's array dimensions are the opposite convention (column-major order). @xref{Multi-dimensional Array Format}. This is just a bookkeeping difference, with no effect on performance. The only consequence of this is that, whenever you create an FFTW plan for a multi-dimensional transform, you must always @emph{reverse the ordering of the dimensions}. For example, consider the three-dimensional (@threedims{L,M,N}) arrays: @example complex(C_DOUBLE_COMPLEX), dimension(L,M,N) :: in, out @end example To plan a DFT for these arrays using @code{fftw_plan_dft_3d}, you could do: @findex fftw_plan_dft_3d @example plan = fftw_plan_dft_3d(N,M,L, in,out, FFTW_FORWARD,FFTW_ESTIMATE) @end example That is, from FFTW's perspective this is a @threedims{N,M,L} array. @emph{No data transposition need occur}, as this is @emph{only notation}. Similarly, to use the more generic routine @code{fftw_plan_dft} with the same arrays, you could do: @example integer(C_INT), dimension(3) :: n = [N,M,L] plan = fftw_plan_dft_3d(3, n, in,out, FFTW_FORWARD,FFTW_ESTIMATE) @end example Note, by the way, that this is different from the legacy Fortran interface (@pxref{Fortran-interface routines}), which automatically reverses the order of the array dimension for you. Here, you are calling the C interface directly, so there is no ``translation'' layer. @cindex r2c/c2r multi-dimensional array format An important thing to keep in mind is the implication of this for multidimensional real-to-complex transforms (@pxref{Multi-Dimensional DFTs of Real Data}). In C, a multidimensional real-to-complex DFT chops the last dimension roughly in half (@threedims{N,M,L} real input goes to @threedims{N,M,L/2+1} complex output). In Fortran, because the array dimension notation is reversed, the @emph{first} dimension of the complex data is chopped roughly in half. For example consider the @samp{r2c} transform of @threedims{L,M,N} real input in Fortran: @findex fftw_plan_dft_r2c_3d @findex fftw_execute_dft_r2c @example type(C_PTR) :: plan real(C_DOUBLE), dimension(L,M,N) :: in complex(C_DOUBLE_COMPLEX), dimension(L/2+1,M,N) :: out plan = fftw_plan_dft_r2c_3d(N,M,L, in,out, FFTW_ESTIMATE) ... call fftw_execute_dft_r2c(plan, in, out) @end example @cindex in-place @cindex padding Alternatively, for an in-place r2c transform, as described in the C documentation we must @emph{pad} the @emph{first} dimension of the real input with an extra two entries (which are ignored by FFTW) so as to leave enough space for the complex output. The input is @emph{allocated} as a @threedims{2[L/2+1],M,N} array, even though only @threedims{L,M,N} of it is actually used. In this example, we will allocate the array as a pointer type, using @samp{fftw_alloc} to ensure aligned memory for maximum performance (@pxref{Allocating aligned memory in Fortran}); this also makes it easy to reference the same memory as both a real array and a complex array. @findex fftw_alloc_complex @findex c_f_pointer @example real(C_DOUBLE), pointer :: in(:,:,:) complex(C_DOUBLE_COMPLEX), pointer :: out(:,:,:) type(C_PTR) :: plan, data data = fftw_alloc_complex(int((L/2+1) * M * N, C_SIZE_T)) call c_f_pointer(data, in, [2*(L/2+1),M,N]) call c_f_pointer(data, out, [L/2+1,M,N]) plan = fftw_plan_dft_r2c_3d(N,M,L, in,out, FFTW_ESTIMATE) ... call fftw_execute_dft_r2c(plan, in, out) ... call fftw_destroy_plan(plan) call fftw_free(data) @end example @c ------------------------------------------------------- @node FFTW Fortran type reference, Plan execution in Fortran, Reversing array dimensions, Calling FFTW from Modern Fortran @section FFTW Fortran type reference The following are the most important type correspondences between the C interface and Fortran: @itemize @bullet @item @tindex fftw_plan Plans (@code{fftw_plan} and variants) are @code{type(C_PTR)} (i.e. an opaque pointer). @item @tindex fftw_complex @cindex precision @ctindex C_DOUBLE @ctindex C_FLOAT @ctindex C_LONG_DOUBLE @ctindex C_DOUBLE_COMPLEX @ctindex C_FLOAT_COMPLEX @ctindex C_LONG_DOUBLE_COMPLEX The C floating-point types @code{double}, @code{float}, and @code{long double} correspond to @code{real(C_DOUBLE)}, @code{real(C_FLOAT)}, and @code{real(C_LONG_DOUBLE)}, respectively. The C complex types @code{fftw_complex}, @code{fftwf_complex}, and @code{fftwl_complex} correspond in Fortran to @code{complex(C_DOUBLE_COMPLEX)}, @code{complex(C_FLOAT_COMPLEX)}, and @code{complex(C_LONG_DOUBLE_COMPLEX)}, respectively. Just as in C (@pxref{Precision}), the FFTW subroutines and types are prefixed with @samp{fftw_}, @code{fftwf_}, and @code{fftwl_} for the different precisions, and link to different libraries (@code{-lfftw3}, @code{-lfftw3f}, and @code{-lfftw3l} on Unix), but use the @emph{same} include file @code{fftw3.f03} and the @emph{same} constants (all of which begin with @samp{FFTW_}). The exception is @code{long double} precision, for which you should @emph{also} include @code{fftw3l.f03} (@pxref{Extended and quadruple precision in Fortran}). @item @tindex ptrdiff_t @ctindex C_INT @ctindex C_INTPTR_T @ctindex C_SIZE_T @findex fftw_malloc The C integer types @code{int} and @code{unsigned} (used for planner flags) become @code{integer(C_INT)}. The C integer type @code{ptrdiff_t} (e.g. in the @ref{64-bit Guru Interface}) becomes @code{integer(C_INTPTR_T)}, and @code{size_t} (in @code{fftw_malloc} etc.) becomes @code{integer(C_SIZE_T)}. @item @tindex fftw_r2r_kind @ctindex C_FFTW_R2R_KIND The @code{fftw_r2r_kind} type (@pxref{Real-to-Real Transform Kinds}) becomes @code{integer(C_FFTW_R2R_KIND)}. The various constant values of the C enumerated type (@code{FFTW_R2HC} etc.) become simply integer constants of the same names in Fortran. @item @ctindex FFTW_DESTROY_INPUT @cindex in-place @findex fftw_flops Numeric array pointer arguments (e.g. @code{double *}) become @code{dimension(*), intent(out)} arrays of the same type, or @code{dimension(*), intent(in)} if they are pointers to constant data (e.g. @code{const int *}). There are a few exceptions where numeric pointers refer to scalar outputs (e.g. for @code{fftw_flops}), in which case they are @code{intent(out)} scalar arguments in Fortran too. For the new-array execute functions (@pxref{New-array Execute Functions}), the input arrays are declared @code{dimension(*), intent(inout)}, since they can be modified in the case of in-place or @code{FFTW_DESTROY_INPUT} transforms. @item @findex fftw_alloc_real @findex c_f_pointer Pointer @emph{return} values (e.g @code{double *}) become @code{type(C_PTR)}. (If they are pointers to arrays, as for @code{fftw_alloc_real}, you can convert them back to Fortran array pointers with the standard intrinsic function @code{c_f_pointer}.) @item @cindex guru interface @tindex fftw_iodim @tindex fftw_iodim64 @cindex 64-bit architecture The @code{fftw_iodim} type in the guru interface (@pxref{Guru vector and transform sizes}) becomes @code{type(fftw_iodim)} in Fortran, a derived data type (the Fortran analogue of C's @code{struct}) with three @code{integer(C_INT)} components: @code{n}, @code{is}, and @code{os}, with the same meanings as in C. The @code{fftw_iodim64} type in the 64-bit guru interface (@pxref{64-bit Guru Interface}) is the same, except that its components are of type @code{integer(C_INTPTR_T)}. @item @ctindex C_FUNPTR Using the wisdom import/export functions from Fortran is a bit tricky, and is discussed in @ref{Accessing the wisdom API from Fortran}. In brief, the @code{FILE *} arguments map to @code{type(C_PTR)}, @code{const char *} to @code{character(C_CHAR), dimension(*), intent(in)} (null-terminated!), and the generic read-char/write-char functions map to @code{type(C_FUNPTR)}. @end itemize @cindex portability You may be wondering if you need to search-and-replace @code{real(kind(0.0d0))} (or whatever your favorite Fortran spelling of ``double precision'' is) with @code{real(C_DOUBLE)} everywhere in your program, and similarly for @code{complex} and @code{integer} types. The answer is no; you can still use your existing types. As long as these types match their C counterparts, things should work without a hitch. The worst that can happen, e.g. in the (unlikely) event of a system where @code{real(kind(0.0d0))} is different from @code{real(C_DOUBLE)}, is that the compiler will give you a type-mismatch error. That is, if you don't use the @code{iso_c_binding} kinds you need to accept at least the theoretical possibility of having to change your code in response to compiler errors on some future machine, but you don't need to worry about silently compiling incorrect code that yields runtime errors. @c ------------------------------------------------------- @node Plan execution in Fortran, Allocating aligned memory in Fortran, FFTW Fortran type reference, Calling FFTW from Modern Fortran @section Plan execution in Fortran In C, in order to use a plan, one normally calls @code{fftw_execute}, which executes the plan to perform the transform on the input/output arrays passed when the plan was created (@pxref{Using Plans}). The corresponding subroutine call in modern Fortran is: @example call fftw_execute(plan) @end example @findex fftw_execute However, we have had reports that this causes problems with some recent optimizing Fortran compilers. The problem is, because the input/output arrays are not passed as explicit arguments to @code{fftw_execute}, the semantics of Fortran (unlike C) allow the compiler to assume that the input/output arrays are not changed by @code{fftw_execute}. As a consequence, certain compilers end up repositioning the call to @code{fftw_execute}, assuming incorrectly that it does nothing to the arrays. There are various workarounds to this, but the safest and simplest thing is to not use @code{fftw_execute} in Fortran. Instead, use the functions described in @ref{New-array Execute Functions}, which take the input/output arrays as explicit arguments. For example, if the plan is for a complex-data DFT and was created for the arrays @code{in} and @code{out}, you would do: @example call fftw_execute_dft(plan, in, out) @end example @findex fftw_execute_dft There are a few things to be careful of, however: @itemize @bullet @item @findex fftw_execute_dft_r2c @findex fftw_execute_dft_c2r @findex fftw_execute_r2r You must use the correct type of execute function, matching the way the plan was created. Complex DFT plans should use @code{fftw_execute_dft}, Real-input (r2c) DFT plans should use use @code{fftw_execute_dft_r2c}, and real-output (c2r) DFT plans should use @code{fftw_execute_dft_c2r}. The various r2r plans should use @code{fftw_execute_r2r}. Fortunately, if you use the wrong one you will get a compile-time type-mismatch error (unlike legacy Fortran). @item You should normally pass the same input/output arrays that were used when creating the plan. This is always safe. @item @emph{If} you pass @emph{different} input/output arrays compared to those used when creating the plan, you must abide by all the restrictions of the new-array execute functions (@pxref{New-array Execute Functions}). The most tricky of these is the requirement that the new arrays have the same alignment as the original arrays; the best (and possibly only) way to guarantee this is to use the @samp{fftw_alloc} functions to allocate your arrays (@pxref{Allocating aligned memory in Fortran}). Alternatively, you can use the @code{FFTW_UNALIGNED} flag when creating the plan, in which case the plan does not depend on the alignment, but this may sacrifice substantial performance on architectures (like x86) with SIMD instructions (@pxref{SIMD alignment and fftw_malloc}). @ctindex FFTW_UNALIGNED @end itemize @c ------------------------------------------------------- @node Allocating aligned memory in Fortran, Accessing the wisdom API from Fortran, Plan execution in Fortran, Calling FFTW from Modern Fortran @section Allocating aligned memory in Fortran @cindex alignment @findex fftw_alloc_real @findex fftw_alloc_complex In order to obtain maximum performance in FFTW, you should store your data in arrays that have been specially aligned in memory (@pxref{SIMD alignment and fftw_malloc}). Enforcing alignment also permits you to safely use the new-array execute functions (@pxref{New-array Execute Functions}) to apply a given plan to more than one pair of in/out arrays. Unfortunately, standard Fortran arrays do @emph{not} provide any alignment guarantees. The @emph{only} way to allocate aligned memory in standard Fortran is to allocate it with an external C function, like the @code{fftw_alloc_real} and @code{fftw_alloc_complex} functions. Fortunately, Fortran 2003 provides a simple way to associate such allocated memory with a standard Fortran array pointer that you can then use normally. We therefore recommend allocating all your input/output arrays using the following technique: @enumerate @item Declare a @code{pointer}, @code{arr}, to your array of the desired type and dimensions. For example, @code{real(C_DOUBLE), pointer :: a(:,:)} for a 2d real array, or @code{complex(C_DOUBLE_COMPLEX), pointer :: a(:,:,:)} for a 3d complex array. @item The number of elements to allocate must be an @code{integer(C_SIZE_T)}. You can either declare a variable of this type, e.g. @code{integer(C_SIZE_T) :: sz}, to store the number of elements to allocate, or you can use the @code{int(..., C_SIZE_T)} intrinsic function. e.g. set @code{sz = L * M * N} or use @code{int(L * M * N, C_SIZE_T)} for an @threedims{L,M,N} array. @item Declare a @code{type(C_PTR) :: p} to hold the return value from FFTW's allocation routine. Set @code{p = fftw_alloc_real(sz)} for a real array, or @code{p = fftw_alloc_complex(sz)} for a complex array. @item @findex c_f_pointer Associate your pointer @code{arr} with the allocated memory @code{p} using the standard @code{c_f_pointer} subroutine: @code{call c_f_pointer(p, arr, [...dimensions...])}, where @code{[...dimensions...])} are an array of the dimensions of the array (in the usual Fortran order). e.g. @code{call c_f_pointer(p, arr, [L,M,N])} for an @threedims{L,M,N} array. (Alternatively, you can omit the dimensions argument if you specified the shape explicitly when declaring @code{arr}.) You can now use @code{arr} as a usual multidimensional array. @item When you are done using the array, deallocate the memory by @code{call fftw_free(p)} on @code{p}. @end enumerate For example, here is how we would allocate an @twodims{L,M} 2d real array: @example real(C_DOUBLE), pointer :: arr(:,:) type(C_PTR) :: p p = fftw_alloc_real(int(L * M, C_SIZE_T)) call c_f_pointer(p, arr, [L,M]) @emph{...use arr and arr(i,j) as usual...} call fftw_free(p) @end example and here is an @threedims{L,M,N} 3d complex array: @example complex(C_DOUBLE_COMPLEX), pointer :: arr(:,:,:) type(C_PTR) :: p p = fftw_alloc_complex(int(L * M * N, C_SIZE_T)) call c_f_pointer(p, arr, [L,M,N]) @emph{...use arr and arr(i,j,k) as usual...} call fftw_free(p) @end example See @ref{Reversing array dimensions} for an example allocating a single array and associating both real and complex array pointers with it, for in-place real-to-complex transforms. @c ------------------------------------------------------- @node Accessing the wisdom API from Fortran, Defining an FFTW module, Allocating aligned memory in Fortran, Calling FFTW from Modern Fortran @section Accessing the wisdom API from Fortran @cindex wisdom @cindex saving plans to disk As explained in @ref{Words of Wisdom-Saving Plans}, FFTW provides a ``wisdom'' API for saving plans to disk so that they can be recreated quickly. The C API for exporting (@pxref{Wisdom Export}) and importing (@pxref{Wisdom Import}) wisdom is somewhat tricky to use from Fortran, however, because of differences in file I/O and string types between C and Fortran. @menu * Wisdom File Export/Import from Fortran:: * Wisdom String Export/Import from Fortran:: * Wisdom Generic Export/Import from Fortran:: @end menu @c =========> @node Wisdom File Export/Import from Fortran, Wisdom String Export/Import from Fortran, Accessing the wisdom API from Fortran, Accessing the wisdom API from Fortran @subsection Wisdom File Export/Import from Fortran @findex fftw_import wisdom_from_filename @findex fftw_export_wisdom_to_filename The easiest way to export and import wisdom is to do so using @code{fftw_export_wisdom_to_filename} and @code{fftw_wisdom_from_filename}. The only trick is that these require you to pass a C string, which is an array of type @code{CHARACTER(C_CHAR)} that is terminated by @code{C_NULL_CHAR}. You can call them like this: @example integer(C_INT) :: ret ret = fftw_export_wisdom_to_filename(C_CHAR_'my_wisdom.dat' // C_NULL_CHAR) if (ret .eq. 0) stop 'error exporting wisdom to file' ret = fftw_import_wisdom_from_filename(C_CHAR_'my_wisdom.dat' // C_NULL_CHAR) if (ret .eq. 0) stop 'error importing wisdom from file' @end example Note that prepending @samp{C_CHAR_} is needed to specify that the literal string is of kind @code{C_CHAR}, and we null-terminate the string by appending @samp{// C_NULL_CHAR}. These functions return an @code{integer(C_INT)} (@code{ret}) which is @code{0} if an error occurred during export/import and nonzero otherwise. It is also possible to use the lower-level routines @code{fftw_export_wisdom_to_file} and @code{fftw_import_wisdom_from_file}, which accept parameters of the C type @code{FILE*}, expressed in Fortran as @code{type(C_PTR)}. However, you are then responsible for creating the @code{FILE*} yourself. You can do this by using @code{iso_c_binding} to define Fortran intefaces for the C library functions @code{fopen} and @code{fclose}, which is a bit strange in Fortran but workable. @c =========> @node Wisdom String Export/Import from Fortran, Wisdom Generic Export/Import from Fortran, Wisdom File Export/Import from Fortran, Accessing the wisdom API from Fortran @subsection Wisdom String Export/Import from Fortran @findex fftw_export_wisdom_to_string Dealing with FFTW's C string export/import is a bit more painful. In particular, the @code{fftw_export_wisdom_to_string} function requires you to deal with a dynamically allocated C string. To get its length, you must define an interface to the C @code{strlen} function, and to deallocate it you must define an interface to C @code{free}: @example use, intrinsic :: iso_c_binding interface integer(C_INT) function strlen(s) bind(C, name='strlen') import type(C_PTR), value :: s end function strlen subroutine free(p) bind(C, name='free') import type(C_PTR), value :: p end subroutine free end interface @end example Given these definitions, you can then export wisdom to a Fortran character array: @example character(C_CHAR), pointer :: s(:) integer(C_SIZE_T) :: slen type(C_PTR) :: p p = fftw_export_wisdom_to_string() if (.not. c_associated(p)) stop 'error exporting wisdom' slen = strlen(p) call c_f_pointer(p, s, [slen+1]) ... call free(p) @end example @findex c_associated @findex c_f_pointer Note that @code{slen} is the length of the C string, but the length of the array is @code{slen+1} because it includes the terminating null character. (You can omit the @samp{+1} if you don't want Fortran to know about the null character.) The standard @code{c_associated} function checks whether @code{p} is a null pointer, which is returned by @code{fftw_export_wisdom_to_string} if there was an error. @findex fftw_import_wisdom_from_string To import wisdom from a string, use @code{fftw_import_wisdom_from_string} as usual; note that the argument of this function must be a @code{character(C_CHAR)} that is terminated by the @code{C_NULL_CHAR} character, like the @code{s} array above. @c =========> @node Wisdom Generic Export/Import from Fortran, , Wisdom String Export/Import from Fortran, Accessing the wisdom API from Fortran @subsection Wisdom Generic Export/Import from Fortran The most generic wisdom export/import functions allow you to provide an arbitrary callback function to read/write one character at a time in any way you want. However, your callback function must be written in a special way, using the @code{bind(C)} attribute to be passed to a C interface. @findex fftw_export_wisdom In particular, to call the generic wisdom export function @code{fftw_export_wisdom}, you would write a callback subroutine of the form: @example subroutine my_write_char(c, p) bind(C) use, intrinsic :: iso_c_binding character(C_CHAR), value :: c type(C_PTR), value :: p @emph{...write c...} end subroutine my_write_char @end example Given such a subroutine (along with the corresponding interface definition), you could then export wisdom using: @findex c_funloc @example call fftw_export_wisdom(c_funloc(my_write_char), p) @end example @findex c_loc @findex c_f_pointer The standard @code{c_funloc} intrinsic converts a Fortran @code{bind(C)} subroutine into a C function pointer. The parameter @code{p} is a @code{type(C_PTR)} to any arbitrary data that you want to pass to @code{my_write_char} (or @code{C_NULL_PTR} if none). (Note that you can get a C pointer to Fortran data using the intrinsic @code{c_loc}, and convert it back to a Fortran pointer in @code{my_write_char} using @code{c_f_pointer}.) Similarly, to use the generic @code{fftw_import_wisdom}, you would define a callback function of the form: @findex fftw_import_wisdom @example integer(C_INT) function my_read_char(p) bind(C) use, intrinsic :: iso_c_binding type(C_PTR), value :: p character :: c @emph{...read a character c...} my_read_char = ichar(c, C_INT) end function my_read_char .... integer(C_INT) :: ret ret = fftw_import_wisdom(c_funloc(my_read_char), p) if (ret .eq. 0) stop 'error importing wisdom' @end example Your function can return @code{-1} if the end of the input is reached. Again, @code{p} is an arbitrary @code{type(C_PTR} that is passed through to your function. @code{fftw_import_wisdom} returns @code{0} if an error occurred and nonzero otherwise. @c ------------------------------------------------------- @node Defining an FFTW module, , Accessing the wisdom API from Fortran, Calling FFTW from Modern Fortran @section Defining an FFTW module Rather than using the @code{include} statement to include the @code{fftw3.f03} interface file in any subroutine where you want to use FFTW, you might prefer to define an FFTW Fortran module. FFTW does not install itself as a module, primarily because @code{fftw3.f03} can be shared between different Fortran compilers while modules (in general) cannot. However, it is trivial to define your own FFTW module if you want. Just create a file containing: @example module FFTW3 use, intrinsic :: iso_c_binding include 'fftw3.f03' end module @end example Compile this file into a module as usual for your compiler (e.g. with @code{gfortran -c} you will get a file @code{fftw3.mod}). Now, instead of @code{include 'fftw3.f03'}, whenever you want to use FFTW routines you can just do: @example use FFTW3 @end example as usual for Fortran modules. (You still need to link to the FFTW library, of course.) fftw-3.3.3/doc/mpi.texi0000644000175000017500000023421712054410050011634 00000000000000@node Distributed-memory FFTW with MPI, Calling FFTW from Modern Fortran, Multi-threaded FFTW, Top @chapter Distributed-memory FFTW with MPI @cindex MPI @cindex parallel transform In this chapter we document the parallel FFTW routines for parallel systems supporting the MPI message-passing interface. Unlike the shared-memory threads described in the previous chapter, MPI allows you to use @emph{distributed-memory} parallelism, where each CPU has its own separate memory, and which can scale up to clusters of many thousands of processors. This capability comes at a price, however: each process only stores a @emph{portion} of the data to be transformed, which means that the data structures and programming-interface are quite different from the serial or threads versions of FFTW. @cindex data distribution Distributed-memory parallelism is especially useful when you are transforming arrays so large that they do not fit into the memory of a single processor. The storage per-process required by FFTW's MPI routines is proportional to the total array size divided by the number of processes. Conversely, distributed-memory parallelism can easily pose an unacceptably high communications overhead for small problems; the threshold problem size for which parallelism becomes advantageous will depend on the precise problem you are interested in, your hardware, and your MPI implementation. A note on terminology: in MPI, you divide the data among a set of ``processes'' which each run in their own memory address space. Generally, each process runs on a different physical processor, but this is not required. A set of processes in MPI is described by an opaque data structure called a ``communicator,'' the most common of which is the predefined communicator @code{MPI_COMM_WORLD} which refers to @emph{all} processes. For more information on these and other concepts common to all MPI programs, we refer the reader to the documentation at @uref{http://www.mcs.anl.gov/research/projects/mpi/, the MPI home page}. @cindex MPI communicator @ctindex MPI_COMM_WORLD We assume in this chapter that the reader is familiar with the usage of the serial (uniprocessor) FFTW, and focus only on the concepts new to the MPI interface. @menu * FFTW MPI Installation:: * Linking and Initializing MPI FFTW:: * 2d MPI example:: * MPI Data Distribution:: * Multi-dimensional MPI DFTs of Real Data:: * Other Multi-dimensional Real-data MPI Transforms:: * FFTW MPI Transposes:: * FFTW MPI Wisdom:: * Avoiding MPI Deadlocks:: * FFTW MPI Performance Tips:: * Combining MPI and Threads:: * FFTW MPI Reference:: * FFTW MPI Fortran Interface:: @end menu @c ------------------------------------------------------------ @node FFTW MPI Installation, Linking and Initializing MPI FFTW, Distributed-memory FFTW with MPI, Distributed-memory FFTW with MPI @section FFTW MPI Installation All of the FFTW MPI code is located in the @code{mpi} subdirectory of the FFTW package. On Unix systems, the FFTW MPI libraries and header files are automatically configured, compiled, and installed along with the uniprocessor FFTW libraries simply by including @code{--enable-mpi} in the flags to the @code{configure} script (@pxref{Installation on Unix}). @fpindex configure Any implementation of the MPI standard, version 1 or later, should work with FFTW. The @code{configure} script will attempt to automatically detect how to compile and link code using your MPI implementation. In some cases, especially if you have multiple different MPI implementations installed or have an unusual MPI software package, you may need to provide this information explicitly. Most commonly, one compiles MPI code by invoking a special compiler command, typically @code{mpicc} for C code. The @code{configure} script knows the most common names for this command, but you can specify the MPI compilation command explicitly by setting the @code{MPICC} variable, as in @samp{./configure MPICC=mpicc ...}. @fpindex mpicc If, instead of a special compiler command, you need to link a certain library, you can specify the link command via the @code{MPILIBS} variable, as in @samp{./configure MPILIBS=-lmpi ...}. Note that if your MPI library is installed in a non-standard location (one the compiler does not know about by default), you may also have to specify the location of the library and header files via @code{LDFLAGS} and @code{CPPFLAGS} variables, respectively, as in @samp{./configure LDFLAGS=-L/path/to/mpi/libs CPPFLAGS=-I/path/to/mpi/include ...}. @c ------------------------------------------------------------ @node Linking and Initializing MPI FFTW, 2d MPI example, FFTW MPI Installation, Distributed-memory FFTW with MPI @section Linking and Initializing MPI FFTW Programs using the MPI FFTW routines should be linked with @code{-lfftw3_mpi -lfftw3 -lm} on Unix in double precision, @code{-lfftw3f_mpi -lfftw3f -lm} in single precision, and so on (@pxref{Precision}). You will also need to link with whatever library is responsible for MPI on your system; in most MPI implementations, there is a special compiler alias named @code{mpicc} to compile and link MPI code. @fpindex mpicc @cindex linking on Unix @cindex precision @findex fftw_init_threads Before calling any FFTW routines except possibly @code{fftw_init_threads} (@pxref{Combining MPI and Threads}), but after calling @code{MPI_Init}, you should call the function: @example void fftw_mpi_init(void); @end example @findex fftw_mpi_init If, at the end of your program, you want to get rid of all memory and other resources allocated internally by FFTW, for both the serial and MPI routines, you can call: @example void fftw_mpi_cleanup(void); @end example @findex fftw_mpi_cleanup which is much like the @code{fftw_cleanup()} function except that it also gets rid of FFTW's MPI-related data. You must @emph{not} execute any previously created plans after calling this function. @c ------------------------------------------------------------ @node 2d MPI example, MPI Data Distribution, Linking and Initializing MPI FFTW, Distributed-memory FFTW with MPI @section 2d MPI example Before we document the FFTW MPI interface in detail, we begin with a simple example outlining how one would perform a two-dimensional @code{N0} by @code{N1} complex DFT. @example #include int main(int argc, char **argv) @{ const ptrdiff_t N0 = ..., N1 = ...; fftw_plan plan; fftw_complex *data; ptrdiff_t alloc_local, local_n0, local_0_start, i, j; MPI_Init(&argc, &argv); fftw_mpi_init(); /* @r{get local data size and allocate} */ alloc_local = fftw_mpi_local_size_2d(N0, N1, MPI_COMM_WORLD, &local_n0, &local_0_start); data = fftw_alloc_complex(alloc_local); /* @r{create plan for in-place forward DFT} */ plan = fftw_mpi_plan_dft_2d(N0, N1, data, data, MPI_COMM_WORLD, FFTW_FORWARD, FFTW_ESTIMATE); /* @r{initialize data to some function} my_function(x,y) */ for (i = 0; i < local_n0; ++i) for (j = 0; j < N1; ++j) data[i*N1 + j] = my_function(local_0_start + i, j); /* @r{compute transforms, in-place, as many times as desired} */ fftw_execute(plan); fftw_destroy_plan(plan); MPI_Finalize(); @} @end example As can be seen above, the MPI interface follows the same basic style of allocate/plan/execute/destroy as the serial FFTW routines. All of the MPI-specific routines are prefixed with @samp{fftw_mpi_} instead of @samp{fftw_}. There are a few important differences, however: First, we must call @code{fftw_mpi_init()} after calling @code{MPI_Init} (required in all MPI programs) and before calling any other @samp{fftw_mpi_} routine. @findex MPI_Init @findex fftw_mpi_init Second, when we create the plan with @code{fftw_mpi_plan_dft_2d}, analogous to @code{fftw_plan_dft_2d}, we pass an additional argument: the communicator, indicating which processes will participate in the transform (here @code{MPI_COMM_WORLD}, indicating all processes). Whenever you create, execute, or destroy a plan for an MPI transform, you must call the corresponding FFTW routine on @emph{all} processes in the communicator for that transform. (That is, these are @emph{collective} calls.) Note that the plan for the MPI transform uses the standard @code{fftw_execute} and @code{fftw_destroy} routines (on the other hand, there are MPI-specific new-array execute functions documented below). @cindex collective function @findex fftw_mpi_plan_dft_2d @ctindex MPI_COMM_WORLD Third, all of the FFTW MPI routines take @code{ptrdiff_t} arguments instead of @code{int} as for the serial FFTW. @code{ptrdiff_t} is a standard C integer type which is (at least) 32 bits wide on a 32-bit machine and 64 bits wide on a 64-bit machine. This is to make it easy to specify very large parallel transforms on a 64-bit machine. (You can specify 64-bit transform sizes in the serial FFTW, too, but only by using the @samp{guru64} planner interface. @xref{64-bit Guru Interface}.) @tindex ptrdiff_t @cindex 64-bit architecture Fourth, and most importantly, you don't allocate the entire two-dimensional array on each process. Instead, you call @code{fftw_mpi_local_size_2d} to find out what @emph{portion} of the array resides on each processor, and how much space to allocate. Here, the portion of the array on each process is a @code{local_n0} by @code{N1} slice of the total array, starting at index @code{local_0_start}. The total number of @code{fftw_complex} numbers to allocate is given by the @code{alloc_local} return value, which @emph{may} be greater than @code{local_n0 * N1} (in case some intermediate calculations require additional storage). The data distribution in FFTW's MPI interface is described in more detail by the next section. @findex fftw_mpi_local_size_2d @cindex data distribution Given the portion of the array that resides on the local process, it is straightforward to initialize the data (here to a function @code{myfunction}) and otherwise manipulate it. Of course, at the end of the program you may want to output the data somehow, but synchronizing this output is up to you and is beyond the scope of this manual. (One good way to output a large multi-dimensional distributed array in MPI to a portable binary file is to use the free HDF5 library; see the @uref{http://www.hdfgroup.org/, HDF home page}.) @cindex HDF5 @cindex MPI I/O @c ------------------------------------------------------------ @node MPI Data Distribution, Multi-dimensional MPI DFTs of Real Data, 2d MPI example, Distributed-memory FFTW with MPI @section MPI Data Distribution @cindex data distribution The most important concept to understand in using FFTW's MPI interface is the data distribution. With a serial or multithreaded FFT, all of the inputs and outputs are stored as a single contiguous chunk of memory. With a distributed-memory FFT, the inputs and outputs are broken into disjoint blocks, one per process. In particular, FFTW uses a @emph{1d block distribution} of the data, distributed along the @emph{first dimension}. For example, if you want to perform a @twodims{100,200} complex DFT, distributed over 4 processes, each process will get a @twodims{25,200} slice of the data. That is, process 0 will get rows 0 through 24, process 1 will get rows 25 through 49, process 2 will get rows 50 through 74, and process 3 will get rows 75 through 99. If you take the same array but distribute it over 3 processes, then it is not evenly divisible so the different processes will have unequal chunks. FFTW's default choice in this case is to assign 34 rows to processes 0 and 1, and 32 rows to process 2. @cindex block distribution FFTW provides several @samp{fftw_mpi_local_size} routines that you can call to find out what portion of an array is stored on the current process. In most cases, you should use the default block sizes picked by FFTW, but it is also possible to specify your own block size. For example, with a @twodims{100,200} array on three processes, you can tell FFTW to use a block size of 40, which would assign 40 rows to processes 0 and 1, and 20 rows to process 2. FFTW's default is to divide the data equally among the processes if possible, and as best it can otherwise. The rows are always assigned in ``rank order,'' i.e. process 0 gets the first block of rows, then process 1, and so on. (You can change this by using @code{MPI_Comm_split} to create a new communicator with re-ordered processes.) However, you should always call the @samp{fftw_mpi_local_size} routines, if possible, rather than trying to predict FFTW's distribution choices. In particular, it is critical that you allocate the storage size that is returned by @samp{fftw_mpi_local_size}, which is @emph{not} necessarily the size of the local slice of the array. The reason is that intermediate steps of FFTW's algorithms involve transposing the array and redistributing the data, so at these intermediate steps FFTW may require more local storage space (albeit always proportional to the total size divided by the number of processes). The @samp{fftw_mpi_local_size} functions know how much storage is required for these intermediate steps and tell you the correct amount to allocate. @menu * Basic and advanced distribution interfaces:: * Load balancing:: * Transposed distributions:: * One-dimensional distributions:: @end menu @node Basic and advanced distribution interfaces, Load balancing, MPI Data Distribution, MPI Data Distribution @subsection Basic and advanced distribution interfaces As with the planner interface, the @samp{fftw_mpi_local_size} distribution interface is broken into basic and advanced (@samp{_many}) interfaces, where the latter allows you to specify the block size manually and also to request block sizes when computing multiple transforms simultaneously. These functions are documented more exhaustively by the FFTW MPI Reference, but we summarize the basic ideas here using a couple of two-dimensional examples. For the @twodims{100,200} complex-DFT example, above, we would find the distribution by calling the following function in the basic interface: @example ptrdiff_t fftw_mpi_local_size_2d(ptrdiff_t n0, ptrdiff_t n1, MPI_Comm comm, ptrdiff_t *local_n0, ptrdiff_t *local_0_start); @end example @findex fftw_mpi_local_size_2d Given the total size of the data to be transformed (here, @code{n0 = 100} and @code{n1 = 200}) and an MPI communicator (@code{comm}), this function provides three numbers. First, it describes the shape of the local data: the current process should store a @code{local_n0} by @code{n1} slice of the overall dataset, in row-major order (@code{n1} dimension contiguous), starting at index @code{local_0_start}. That is, if the total dataset is viewed as a @code{n0} by @code{n1} matrix, the current process should store the rows @code{local_0_start} to @code{local_0_start+local_n0-1}. Obviously, if you are running with only a single MPI process, that process will store the entire array: @code{local_0_start} will be zero and @code{local_n0} will be @code{n0}. @xref{Row-major Format}. @cindex row-major Second, the return value is the total number of data elements (e.g., complex numbers for a complex DFT) that should be allocated for the input and output arrays on the current process (ideally with @code{fftw_malloc} or an @samp{fftw_alloc} function, to ensure optimal alignment). It might seem that this should always be equal to @code{local_n0 * n1}, but this is @emph{not} the case. FFTW's distributed FFT algorithms require data redistributions at intermediate stages of the transform, and in some circumstances this may require slightly larger local storage. This is discussed in more detail below, under @ref{Load balancing}. @findex fftw_malloc @findex fftw_alloc_complex @cindex advanced interface The advanced-interface @samp{local_size} function for multidimensional transforms returns the same three things (@code{local_n0}, @code{local_0_start}, and the total number of elements to allocate), but takes more inputs: @example ptrdiff_t fftw_mpi_local_size_many(int rnk, const ptrdiff_t *n, ptrdiff_t howmany, ptrdiff_t block0, MPI_Comm comm, ptrdiff_t *local_n0, ptrdiff_t *local_0_start); @end example @findex fftw_mpi_local_size_many The two-dimensional case above corresponds to @code{rnk = 2} and an array @code{n} of length 2 with @code{n[0] = n0} and @code{n[1] = n1}. This routine is for any @code{rnk > 1}; one-dimensional transforms have their own interface because they work slightly differently, as discussed below. First, the advanced interface allows you to perform multiple transforms at once, of interleaved data, as specified by the @code{howmany} parameter. (@code{hoamany} is 1 for a single transform.) Second, here you can specify your desired block size in the @code{n0} dimension, @code{block0}. To use FFTW's default block size, pass @code{FFTW_MPI_DEFAULT_BLOCK} (0) for @code{block0}. Otherwise, on @code{P} processes, FFTW will return @code{local_n0} equal to @code{block0} on the first @code{P / block0} processes (rounded down), return @code{local_n0} equal to @code{n0 - block0 * (P / block0)} on the next process, and @code{local_n0} equal to zero on any remaining processes. In general, we recommend using the default block size (which corresponds to @code{n0 / P}, rounded up). @ctindex FFTW_MPI_DEFAULT_BLOCK @cindex block distribution For example, suppose you have @code{P = 4} processes and @code{n0 = 21}. The default will be a block size of @code{6}, which will give @code{local_n0 = 6} on the first three processes and @code{local_n0 = 3} on the last process. Instead, however, you could specify @code{block0 = 5} if you wanted, which would give @code{local_n0 = 5} on processes 0 to 2, @code{local_n0 = 6} on process 3. (This choice, while it may look superficially more ``balanced,'' has the same critical path as FFTW's default but requires more communications.) @node Load balancing, Transposed distributions, Basic and advanced distribution interfaces, MPI Data Distribution @subsection Load balancing @cindex load balancing Ideally, when you parallelize a transform over some @math{P} processes, each process should end up with work that takes equal time. Otherwise, all of the processes end up waiting on whichever process is slowest. This goal is known as ``load balancing.'' In this section, we describe the circumstances under which FFTW is able to load-balance well, and in particular how you should choose your transform size in order to load balance. Load balancing is especially difficult when you are parallelizing over heterogeneous machines; for example, if one of your processors is a old 486 and another is a Pentium IV, obviously you should give the Pentium more work to do than the 486 since the latter is much slower. FFTW does not deal with this problem, however---it assumes that your processes run on hardware of comparable speed, and that the goal is therefore to divide the problem as equally as possible. For a multi-dimensional complex DFT, FFTW can divide the problem equally among the processes if: (i) the @emph{first} dimension @code{n0} is divisible by @math{P}; and (ii), the @emph{product} of the subsequent dimensions is divisible by @math{P}. (For the advanced interface, where you can specify multiple simultaneous transforms via some ``vector'' length @code{howmany}, a factor of @code{howmany} is included in the product of the subsequent dimensions.) For a one-dimensional complex DFT, the length @code{N} of the data should be divisible by @math{P} @emph{squared} to be able to divide the problem equally among the processes. @node Transposed distributions, One-dimensional distributions, Load balancing, MPI Data Distribution @subsection Transposed distributions Internally, FFTW's MPI transform algorithms work by first computing transforms of the data local to each process, then by globally @emph{transposing} the data in some fashion to redistribute the data among the processes, transforming the new data local to each process, and transposing back. For example, a two-dimensional @code{n0} by @code{n1} array, distributed across the @code{n0} dimension, is transformd by: (i) transforming the @code{n1} dimension, which are local to each process; (ii) transposing to an @code{n1} by @code{n0} array, distributed across the @code{n1} dimension; (iii) transforming the @code{n0} dimension, which is now local to each process; (iv) transposing back. @cindex transpose However, in many applications it is acceptable to compute a multidimensional DFT whose results are produced in transposed order (e.g., @code{n1} by @code{n0} in two dimensions). This provides a significant performance advantage, because it means that the final transposition step can be omitted. FFTW supports this optimization, which you specify by passing the flag @code{FFTW_MPI_TRANSPOSED_OUT} to the planner routines. To compute the inverse transform of transposed output, you specify @code{FFTW_MPI_TRANSPOSED_IN} to tell it that the input is transposed. In this section, we explain how to interpret the output format of such a transform. @ctindex FFTW_MPI_TRANSPOSED_OUT @ctindex FFTW_MPI_TRANSPOSED_IN Suppose you have are transforming multi-dimensional data with (at least two) dimensions @ndims{}. As always, it is distributed along the first dimension @dimk{0}. Now, if we compute its DFT with the @code{FFTW_MPI_TRANSPOSED_OUT} flag, the resulting output data are stored with the first @emph{two} dimensions transposed: @ndimstrans{}, distributed along the @dimk{1} dimension. Conversely, if we take the @ndimstrans{} data and transform it with the @code{FFTW_MPI_TRANSPOSED_IN} flag, then the format goes back to the original @ndims{} array. There are two ways to find the portion of the transposed array that resides on the current process. First, you can simply call the appropriate @samp{local_size} function, passing @ndimstrans{} (the transposed dimensions). This would mean calling the @samp{local_size} function twice, once for the transposed and once for the non-transposed dimensions. Alternatively, you can call one of the @samp{local_size_transposed} functions, which returns both the non-transposed and transposed data distribution from a single call. For example, for a 3d transform with transposed output (or input), you might call: @example ptrdiff_t fftw_mpi_local_size_3d_transposed( ptrdiff_t n0, ptrdiff_t n1, ptrdiff_t n2, MPI_Comm comm, ptrdiff_t *local_n0, ptrdiff_t *local_0_start, ptrdiff_t *local_n1, ptrdiff_t *local_1_start); @end example @findex fftw_mpi_local_size_3d_transposed Here, @code{local_n0} and @code{local_0_start} give the size and starting index of the @code{n0} dimension for the @emph{non}-transposed data, as in the previous sections. For @emph{transposed} data (e.g. the output for @code{FFTW_MPI_TRANSPOSED_OUT}), @code{local_n1} and @code{local_1_start} give the size and starting index of the @code{n1} dimension, which is the first dimension of the transposed data (@code{n1} by @code{n0} by @code{n2}). (Note that @code{FFTW_MPI_TRANSPOSED_IN} is completely equivalent to performing @code{FFTW_MPI_TRANSPOSED_OUT} and passing the first two dimensions to the planner in reverse order, or vice versa. If you pass @emph{both} the @code{FFTW_MPI_TRANSPOSED_IN} and @code{FFTW_MPI_TRANSPOSED_OUT} flags, it is equivalent to swapping the first two dimensions passed to the planner and passing @emph{neither} flag.) @node One-dimensional distributions, , Transposed distributions, MPI Data Distribution @subsection One-dimensional distributions For one-dimensional distributed DFTs using FFTW, matters are slightly more complicated because the data distribution is more closely tied to how the algorithm works. In particular, you can no longer pass an arbitrary block size and must accept FFTW's default; also, the block sizes may be different for input and output. Also, the data distribution depends on the flags and transform direction, in order for forward and backward transforms to work correctly. @example ptrdiff_t fftw_mpi_local_size_1d(ptrdiff_t n0, MPI_Comm comm, int sign, unsigned flags, ptrdiff_t *local_ni, ptrdiff_t *local_i_start, ptrdiff_t *local_no, ptrdiff_t *local_o_start); @end example @findex fftw_mpi_local_size_1d This function computes the data distribution for a 1d transform of size @code{n0} with the given transform @code{sign} and @code{flags}. Both input and output data use block distributions. The input on the current process will consist of @code{local_ni} numbers starting at index @code{local_i_start}; e.g. if only a single process is used, then @code{local_ni} will be @code{n0} and @code{local_i_start} will be @code{0}. Similarly for the output, with @code{local_no} numbers starting at index @code{local_o_start}. The return value of @code{fftw_mpi_local_size_1d} will be the total number of elements to allocate on the current process (which might be slightly larger than the local size due to intermediate steps in the algorithm). As mentioned above (@pxref{Load balancing}), the data will be divided equally among the processes if @code{n0} is divisible by the @emph{square} of the number of processes. In this case, @code{local_ni} will equal @code{local_no}. Otherwise, they may be different. For some applications, such as convolutions, the order of the output data is irrelevant. In this case, performance can be improved by specifying that the output data be stored in an FFTW-defined ``scrambled'' format. (In particular, this is the analogue of transposed output in the multidimensional case: scrambled output saves a communications step.) If you pass @code{FFTW_MPI_SCRAMBLED_OUT} in the flags, then the output is stored in this (undocumented) scrambled order. Conversely, to perform the inverse transform of data in scrambled order, pass the @code{FFTW_MPI_SCRAMBLED_IN} flag. @ctindex FFTW_MPI_SCRAMBLED_OUT @ctindex FFTW_MPI_SCRAMBLED_IN In MPI FFTW, only composite sizes @code{n0} can be parallelized; we have not yet implemented a parallel algorithm for large prime sizes. @c ------------------------------------------------------------ @node Multi-dimensional MPI DFTs of Real Data, Other Multi-dimensional Real-data MPI Transforms, MPI Data Distribution, Distributed-memory FFTW with MPI @section Multi-dimensional MPI DFTs of Real Data FFTW's MPI interface also supports multi-dimensional DFTs of real data, similar to the serial r2c and c2r interfaces. (Parallel one-dimensional real-data DFTs are not currently supported; you must use a complex transform and set the imaginary parts of the inputs to zero.) The key points to understand for r2c and c2r MPI transforms (compared to the MPI complex DFTs or the serial r2c/c2r transforms), are: @itemize @bullet @item Just as for serial transforms, r2c/c2r DFTs transform @ndims{} real data to/from @ndimshalf{} complex data: the last dimension of the complex data is cut in half (rounded down), plus one. As for the serial transforms, the sizes you pass to the @samp{plan_dft_r2c} and @samp{plan_dft_c2r} are the @ndims{} dimensions of the real data. @item @cindex padding Although the real data is @emph{conceptually} @ndims{}, it is @emph{physically} stored as an @ndimspad{} array, where the last dimension has been @emph{padded} to make it the same size as the complex output. This is much like the in-place serial r2c/c2r interface (@pxref{Multi-Dimensional DFTs of Real Data}), except that in MPI the padding is required even for out-of-place data. The extra padding numbers are ignored by FFTW (they are @emph{not} like zero-padding the transform to a larger size); they are only used to determine the data layout. @item @cindex data distribution The data distribution in MPI for @emph{both} the real and complex data is determined by the shape of the @emph{complex} data. That is, you call the appropriate @samp{local size} function for the @ndimshalf{} complex data, and then use the @emph{same} distribution for the real data except that the last complex dimension is replaced by a (padded) real dimension of twice the length. @end itemize For example suppose we are performing an out-of-place r2c transform of @threedims{L,M,N} real data [padded to @threedims{L,M,2(N/2+1)}], resulting in @threedims{L,M,N/2+1} complex data. Similar to the example in @ref{2d MPI example}, we might do something like: @example #include int main(int argc, char **argv) @{ const ptrdiff_t L = ..., M = ..., N = ...; fftw_plan plan; double *rin; fftw_complex *cout; ptrdiff_t alloc_local, local_n0, local_0_start, i, j, k; MPI_Init(&argc, &argv); fftw_mpi_init(); /* @r{get local data size and allocate} */ alloc_local = fftw_mpi_local_size_3d(L, M, N/2+1, MPI_COMM_WORLD, &local_n0, &local_0_start); rin = fftw_alloc_real(2 * alloc_local); cout = fftw_alloc_complex(alloc_local); /* @r{create plan for out-of-place r2c DFT} */ plan = fftw_mpi_plan_dft_r2c_3d(L, M, N, rin, cout, MPI_COMM_WORLD, FFTW_MEASURE); /* @r{initialize rin to some function} my_func(x,y,z) */ for (i = 0; i < local_n0; ++i) for (j = 0; j < M; ++j) for (k = 0; k < N; ++k) rin[(i*M + j) * (2*(N/2+1)) + k] = my_func(local_0_start+i, j, k); /* @r{compute transforms as many times as desired} */ fftw_execute(plan); fftw_destroy_plan(plan); MPI_Finalize(); @} @end example @findex fftw_alloc_real @cindex row-major Note that we allocated @code{rin} using @code{fftw_alloc_real} with an argument of @code{2 * alloc_local}: since @code{alloc_local} is the number of @emph{complex} values to allocate, the number of @emph{real} values is twice as many. The @code{rin} array is then @threedims{local_n0,M,2(N/2+1)} in row-major order, so its @code{(i,j,k)} element is at the index @code{(i*M + j) * (2*(N/2+1)) + k} (@pxref{Multi-dimensional Array Format }). @cindex transpose @ctindex FFTW_TRANSPOSED_OUT @ctindex FFTW_TRANSPOSED_IN As for the complex transforms, improved performance can be obtained by specifying that the output is the transpose of the input or vice versa (@pxref{Transposed distributions}). In our @threedims{L,M,N} r2c example, including @code{FFTW_TRANSPOSED_OUT} in the flags means that the input would be a padded @threedims{L,M,2(N/2+1)} real array distributed over the @code{L} dimension, while the output would be a @threedims{M,L,N/2+1} complex array distributed over the @code{M} dimension. To perform the inverse c2r transform with the same data distributions, you would use the @code{FFTW_TRANSPOSED_IN} flag. @c ------------------------------------------------------------ @node Other Multi-dimensional Real-data MPI Transforms, FFTW MPI Transposes, Multi-dimensional MPI DFTs of Real Data, Distributed-memory FFTW with MPI @section Other multi-dimensional Real-Data MPI Transforms @cindex r2r FFTW's MPI interface also supports multi-dimensional @samp{r2r} transforms of all kinds supported by the serial interface (e.g. discrete cosine and sine transforms, discrete Hartley transforms, etc.). Only multi-dimensional @samp{r2r} transforms, not one-dimensional transforms, are currently parallelized. @tindex fftw_r2r_kind These are used much like the multidimensional complex DFTs discussed above, except that the data is real rather than complex, and one needs to pass an r2r transform kind (@code{fftw_r2r_kind}) for each dimension as in the serial FFTW (@pxref{More DFTs of Real Data}). For example, one might perform a two-dimensional @twodims{L,M} that is an REDFT10 (DCT-II) in the first dimension and an RODFT10 (DST-II) in the second dimension with code like: @example const ptrdiff_t L = ..., M = ...; fftw_plan plan; double *data; ptrdiff_t alloc_local, local_n0, local_0_start, i, j; /* @r{get local data size and allocate} */ alloc_local = fftw_mpi_local_size_2d(L, M, MPI_COMM_WORLD, &local_n0, &local_0_start); data = fftw_alloc_real(alloc_local); /* @r{create plan for in-place REDFT10 x RODFT10} */ plan = fftw_mpi_plan_r2r_2d(L, M, data, data, MPI_COMM_WORLD, FFTW_REDFT10, FFTW_RODFT10, FFTW_MEASURE); /* @r{initialize data to some function} my_function(x,y) */ for (i = 0; i < local_n0; ++i) for (j = 0; j < M; ++j) data[i*M + j] = my_function(local_0_start + i, j); /* @r{compute transforms, in-place, as many times as desired} */ fftw_execute(plan); fftw_destroy_plan(plan); @end example @findex fftw_alloc_real Notice that we use the same @samp{local_size} functions as we did for complex data, only now we interpret the sizes in terms of real rather than complex values, and correspondingly use @code{fftw_alloc_real}. @c ------------------------------------------------------------ @node FFTW MPI Transposes, FFTW MPI Wisdom, Other Multi-dimensional Real-data MPI Transforms, Distributed-memory FFTW with MPI @section FFTW MPI Transposes @cindex transpose The FFTW's MPI Fourier transforms rely on one or more @emph{global transposition} step for their communications. For example, the multidimensional transforms work by transforming along some dimensions, then transposing to make the first dimension local and transforming that, then transposing back. Because global transposition of a block-distributed matrix has many other potential uses besides FFTs, FFTW's transpose routines can be called directly, as documented in this section. @menu * Basic distributed-transpose interface:: * Advanced distributed-transpose interface:: * An improved replacement for MPI_Alltoall:: @end menu @node Basic distributed-transpose interface, Advanced distributed-transpose interface, FFTW MPI Transposes, FFTW MPI Transposes @subsection Basic distributed-transpose interface In particular, suppose that we have an @code{n0} by @code{n1} array in row-major order, block-distributed across the @code{n0} dimension. To transpose this into an @code{n1} by @code{n0} array block-distributed across the @code{n1} dimension, we would create a plan by calling the following function: @example fftw_plan fftw_mpi_plan_transpose(ptrdiff_t n0, ptrdiff_t n1, double *in, double *out, MPI_Comm comm, unsigned flags); @end example @findex fftw_mpi_plan_transpose The input and output arrays (@code{in} and @code{out}) can be the same. The transpose is actually executed by calling @code{fftw_execute} on the plan, as usual. @findex fftw_execute The @code{flags} are the usual FFTW planner flags, but support two additional flags: @code{FFTW_MPI_TRANSPOSED_OUT} and/or @code{FFTW_MPI_TRANSPOSED_IN}. What these flags indicate, for transpose plans, is that the output and/or input, respectively, are @emph{locally} transposed. That is, on each process input data is normally stored as a @code{local_n0} by @code{n1} array in row-major order, but for an @code{FFTW_MPI_TRANSPOSED_IN} plan the input data is stored as @code{n1} by @code{local_n0} in row-major order. Similarly, @code{FFTW_MPI_TRANSPOSED_OUT} means that the output is @code{n0} by @code{local_n1} instead of @code{local_n1} by @code{n0}. @ctindex FFTW_MPI_TRANSPOSED_OUT @ctindex FFTW_MPI_TRANSPOSED_IN To determine the local size of the array on each process before and after the transpose, as well as the amount of storage that must be allocated, one should call @code{fftw_mpi_local_size_2d_transposed}, just as for a 2d DFT as described in the previous section: @cindex data distribution @example ptrdiff_t fftw_mpi_local_size_2d_transposed (ptrdiff_t n0, ptrdiff_t n1, MPI_Comm comm, ptrdiff_t *local_n0, ptrdiff_t *local_0_start, ptrdiff_t *local_n1, ptrdiff_t *local_1_start); @end example @findex fftw_mpi_local_size_2d_transposed Again, the return value is the local storage to allocate, which in this case is the number of @emph{real} (@code{double}) values rather than complex numbers as in the previous examples. @node Advanced distributed-transpose interface, An improved replacement for MPI_Alltoall, Basic distributed-transpose interface, FFTW MPI Transposes @subsection Advanced distributed-transpose interface The above routines are for a transpose of a matrix of numbers (of type @code{double}), using FFTW's default block sizes. More generally, one can perform transposes of @emph{tuples} of numbers, with user-specified block sizes for the input and output: @example fftw_plan fftw_mpi_plan_many_transpose (ptrdiff_t n0, ptrdiff_t n1, ptrdiff_t howmany, ptrdiff_t block0, ptrdiff_t block1, double *in, double *out, MPI_Comm comm, unsigned flags); @end example @findex fftw_mpi_plan_many_transpose In this case, one is transposing an @code{n0} by @code{n1} matrix of @code{howmany}-tuples (e.g. @code{howmany = 2} for complex numbers). The input is distributed along the @code{n0} dimension with block size @code{block0}, and the @code{n1} by @code{n0} output is distributed along the @code{n1} dimension with block size @code{block1}. If @code{FFTW_MPI_DEFAULT_BLOCK} (0) is passed for a block size then FFTW uses its default block size. To get the local size of the data on each process, you should then call @code{fftw_mpi_local_size_many_transposed}. @ctindex FFTW_MPI_DEFAULT_BLOCK @findex fftw_mpi_local_size_many_transposed @node An improved replacement for MPI_Alltoall, , Advanced distributed-transpose interface, FFTW MPI Transposes @subsection An improved replacement for MPI_Alltoall We close this section by noting that FFTW's MPI transpose routines can be thought of as a generalization for the @code{MPI_Alltoall} function (albeit only for floating-point types), and in some circumstances can function as an improved replacement. @findex MPI_Alltoall @code{MPI_Alltoall} is defined by the MPI standard as: @example int MPI_Alltoall(void *sendbuf, int sendcount, MPI_Datatype sendtype, void *recvbuf, int recvcnt, MPI_Datatype recvtype, MPI_Comm comm); @end example In particular, for @code{double*} arrays @code{in} and @code{out}, consider the call: @example MPI_Alltoall(in, howmany, MPI_DOUBLE, out, howmany MPI_DOUBLE, comm); @end example This is completely equivalent to: @example MPI_Comm_size(comm, &P); plan = fftw_mpi_plan_many_transpose(P, P, howmany, 1, 1, in, out, comm, FFTW_ESTIMATE); fftw_execute(plan); fftw_destroy_plan(plan); @end example That is, computing a @twodims{P,P} transpose on @code{P} processes, with a block size of 1, is just a standard all-to-all communication. However, using the FFTW routine instead of @code{MPI_Alltoall} may have certain advantages. First of all, FFTW's routine can operate in-place (@code{in == out}) whereas @code{MPI_Alltoall} can only operate out-of-place. @cindex in-place Second, even for out-of-place plans, FFTW's routine may be faster, especially if you need to perform the all-to-all communication many times and can afford to use @code{FFTW_MEASURE} or @code{FFTW_PATIENT}. It should certainly be no slower, not including the time to create the plan, since one of the possible algorithms that FFTW uses for an out-of-place transpose @emph{is} simply to call @code{MPI_Alltoall}. However, FFTW also considers several other possible algorithms that, depending on your MPI implementation and your hardware, may be faster. @ctindex FFTW_MEASURE @ctindex FFTW_PATIENT @c ------------------------------------------------------------ @node FFTW MPI Wisdom, Avoiding MPI Deadlocks, FFTW MPI Transposes, Distributed-memory FFTW with MPI @section FFTW MPI Wisdom @cindex wisdom @cindex saving plans to disk FFTW's ``wisdom'' facility (@pxref{Words of Wisdom-Saving Plans}) can be used to save MPI plans as well as to save uniprocessor plans. However, for MPI there are several unavoidable complications. @cindex MPI I/O First, the MPI standard does not guarantee that every process can perform file I/O (at least, not using C stdio routines)---in general, we may only assume that process 0 is capable of I/O.@footnote{In fact, even this assumption is not technically guaranteed by the standard, although it seems to be universal in actual MPI implementations and is widely assumed by MPI-using software. Technically, you need to query the @code{MPI_IO} attribute of @code{MPI_COMM_WORLD} with @code{MPI_Attr_get}. If this attribute is @code{MPI_PROC_NULL}, no I/O is possible. If it is @code{MPI_ANY_SOURCE}, any process can perform I/O. Otherwise, it is the rank of a process that can perform I/O ... but since it is not guaranteed to yield the @emph{same} rank on all processes, you have to do an @code{MPI_Allreduce} of some kind if you want all processes to agree about which is going to do I/O. And even then, the standard only guarantees that this process can perform output, but not input. See e.g. @cite{Parallel Programming with MPI} by P. S. Pacheco, section 8.1.3. Needless to say, in our experience virtually no MPI programmers worry about this.} So, if we want to export the wisdom from a single process to a file, we must first export the wisdom to a string, then send it to process 0, then write it to a file. Second, in principle we may want to have separate wisdom for every process, since in general the processes may run on different hardware even for a single MPI program. However, in practice FFTW's MPI code is designed for the case of homogeneous hardware (@pxref{Load balancing}), and in this case it is convenient to use the same wisdom for every process. Thus, we need a mechanism to synchronize the wisdom. To address both of these problems, FFTW provides the following two functions: @example void fftw_mpi_broadcast_wisdom(MPI_Comm comm); void fftw_mpi_gather_wisdom(MPI_Comm comm); @end example @findex fftw_mpi_gather_wisdom @findex fftw_mpi_broadcast_wisdom Given a communicator @code{comm}, @code{fftw_mpi_broadcast_wisdom} will broadcast the wisdom from process 0 to all other processes. Conversely, @code{fftw_mpi_gather_wisdom} will collect wisdom from all processes onto process 0. (If the plans created for the same problem by different processes are not the same, @code{fftw_mpi_gather_wisdom} will arbitrarily choose one of the plans.) Both of these functions may result in suboptimal plans for different processes if the processes are running on non-identical hardware. Both of these functions are @emph{collective} calls, which means that they must be executed by all processes in the communicator. @cindex collective function So, for example, a typical code snippet to import wisdom from a file and use it on all processes would be: @example @{ int rank; fftw_mpi_init(); MPI_Comm_rank(MPI_COMM_WORLD, &rank); if (rank == 0) fftw_import_wisdom_from_filename("mywisdom"); fftw_mpi_broadcast_wisdom(MPI_COMM_WORLD); @} @end example (Note that we must call @code{fftw_mpi_init} before importing any wisdom that might contain MPI plans.) Similarly, a typical code snippet to export wisdom from all processes to a file is: @findex fftw_mpi_init @example @{ int rank; fftw_mpi_gather_wisdom(MPI_COMM_WORLD); MPI_Comm_rank(MPI_COMM_WORLD, &rank); if (rank == 0) fftw_export_wisdom_to_filename("mywisdom"); @} @end example @c ------------------------------------------------------------ @node Avoiding MPI Deadlocks, FFTW MPI Performance Tips, FFTW MPI Wisdom, Distributed-memory FFTW with MPI @section Avoiding MPI Deadlocks @cindex deadlock An MPI program can @emph{deadlock} if one process is waiting for a message from another process that never gets sent. To avoid deadlocks when using FFTW's MPI routines, it is important to know which functions are @emph{collective}: that is, which functions must @emph{always} be called in the @emph{same order} from @emph{every} process in a given communicator. (For example, @code{MPI_Barrier} is the canonical example of a collective function in the MPI standard.) @cindex collective function @findex MPI_Barrier The functions in FFTW that are @emph{always} collective are: every function beginning with @samp{fftw_mpi_plan}, as well as @code{fftw_mpi_broadcast_wisdom} and @code{fftw_mpi_gather_wisdom}. Also, the following functions from the ordinary FFTW interface are collective when they are applied to a plan created by an @samp{fftw_mpi_plan} function: @code{fftw_execute}, @code{fftw_destroy_plan}, and @code{fftw_flops}. @findex fftw_execute @findex fftw_destroy_plan @findex fftw_flops @c ------------------------------------------------------------ @node FFTW MPI Performance Tips, Combining MPI and Threads, Avoiding MPI Deadlocks, Distributed-memory FFTW with MPI @section FFTW MPI Performance Tips In this section, we collect a few tips on getting the best performance out of FFTW's MPI transforms. First, because of the 1d block distribution, FFTW's parallelization is currently limited by the size of the first dimension. (Multidimensional block distributions may be supported by a future version.) More generally, you should ideally arrange the dimensions so that FFTW can divide them equally among the processes. @xref{Load balancing}. @cindex block distribution @cindex load balancing Second, if it is not too inconvenient, you should consider working with transposed output for multidimensional plans, as this saves a considerable amount of communications. @xref{Transposed distributions}. @cindex transpose Third, the fastest choices are generally either an in-place transform or an out-of-place transform with the @code{FFTW_DESTROY_INPUT} flag (which allows the input array to be used as scratch space). In-place is especially beneficial if the amount of data per process is large. @ctindex FFTW_DESTROY_INPUT Fourth, if you have multiple arrays to transform at once, rather than calling FFTW's MPI transforms several times it usually seems to be faster to interleave the data and use the advanced interface. (This groups the communications together instead of requiring separate messages for each transform.) @c ------------------------------------------------------------ @node Combining MPI and Threads, FFTW MPI Reference, FFTW MPI Performance Tips, Distributed-memory FFTW with MPI @section Combining MPI and Threads @cindex threads In certain cases, it may be advantageous to combine MPI (distributed-memory) and threads (shared-memory) parallelization. FFTW supports this, with certain caveats. For example, if you have a cluster of 4-processor shared-memory nodes, you may want to use threads within the nodes and MPI between the nodes, instead of MPI for all parallelization. In particular, it is possible to seamlessly combine the MPI FFTW routines with the multi-threaded FFTW routines (@pxref{Multi-threaded FFTW}). However, some care must be taken in the initialization code, which should look something like this: @example int threads_ok; int main(int argc, char **argv) @{ int provided; MPI_Init_thread(&argc, &argv, MPI_THREAD_FUNNELED, &provided); threads_ok = provided >= MPI_THREAD_FUNNELED; if (threads_ok) threads_ok = fftw_init_threads(); fftw_mpi_init(); ... if (threads_ok) fftw_plan_with_nthreads(...); ... MPI_Finalize(); @} @end example @findex fftw_mpi_init @findex fftw_init_threads @findex fftw_plan_with_nthreads First, note that instead of calling @code{MPI_Init}, you should call @code{MPI_Init_threads}, which is the initialization routine defined by the MPI-2 standard to indicate to MPI that your program will be multithreaded. We pass @code{MPI_THREAD_FUNNELED}, which indicates that we will only call MPI routines from the main thread. (FFTW will launch additional threads internally, but the extra threads will not call MPI code.) (You may also pass @code{MPI_THREAD_SERIALIZED} or @code{MPI_THREAD_MULTIPLE}, which requests additional multithreading support from the MPI implementation, but this is not required by FFTW.) The @code{provided} parameter returns what level of threads support is actually supported by your MPI implementation; this @emph{must} be at least @code{MPI_THREAD_FUNNELED} if you want to call the FFTW threads routines, so we define a global variable @code{threads_ok} to record this. You should only call @code{fftw_init_threads} or @code{fftw_plan_with_nthreads} if @code{threads_ok} is true. For more information on thread safety in MPI, see the @uref{http://www.mpi-forum.org/docs/mpi-20-html/node162.htm, MPI and Threads} section of the MPI-2 standard. @cindex thread safety Second, we must call @code{fftw_init_threads} @emph{before} @code{fftw_mpi_init}. This is critical for technical reasons having to do with how FFTW initializes its list of algorithms. Then, if you call @code{fftw_plan_with_nthreads(N)}, @emph{every} MPI process will launch (up to) @code{N} threads to parallelize its transforms. For example, in the hypothetical cluster of 4-processor nodes, you might wish to launch only a single MPI process per node, and then call @code{fftw_plan_with_nthreads(4)} on each process to use all processors in the nodes. This may or may not be faster than simply using as many MPI processes as you have processors, however. On the one hand, using threads within a node eliminates the need for explicit message passing within the node. On the other hand, FFTW's transpose routines are not multi-threaded, and this means that the communications that do take place will not benefit from parallelization within the node. Moreover, many MPI implementations already have optimizations to exploit shared memory when it is available, so adding the multithreaded FFTW on top of this may be superfluous. @cindex transpose @c ------------------------------------------------------------ @node FFTW MPI Reference, FFTW MPI Fortran Interface, Combining MPI and Threads, Distributed-memory FFTW with MPI @section FFTW MPI Reference This chapter provides a complete reference to all FFTW MPI functions, datatypes, and constants. See also @ref{FFTW Reference} for information on functions and types in common with the serial interface. @menu * MPI Files and Data Types:: * MPI Initialization:: * Using MPI Plans:: * MPI Data Distribution Functions:: * MPI Plan Creation:: * MPI Wisdom Communication:: @end menu @node MPI Files and Data Types, MPI Initialization, FFTW MPI Reference, FFTW MPI Reference @subsection MPI Files and Data Types All programs using FFTW's MPI support should include its header file: @example #include @end example Note that this header file includes the serial-FFTW @code{fftw3.h} header file, and also the @code{mpi.h} header file for MPI, so you need not include those files separately. You must also link to @emph{both} the FFTW MPI library and to the serial FFTW library. On Unix, this means adding @code{-lfftw3_mpi -lfftw3 -lm} at the end of the link command. @cindex precision Different precisions are handled as in the serial interface: @xref{Precision}. That is, @samp{fftw_} functions become @code{fftwf_} (in single precision) etcetera, and the libraries become @code{-lfftw3f_mpi -lfftw3f -lm} etcetera on Unix. Long-double precision is supported in MPI, but quad precision (@samp{fftwq_}) is not due to the lack of MPI support for this type. @node MPI Initialization, Using MPI Plans, MPI Files and Data Types, FFTW MPI Reference @subsection MPI Initialization Before calling any other FFTW MPI (@samp{fftw_mpi_}) function, and before importing any wisdom for MPI problems, you must call: @findex fftw_mpi_init @example void fftw_mpi_init(void); @end example @findex fftw_init_threads If FFTW threads support is used, however, @code{fftw_mpi_init} should be called @emph{after} @code{fftw_init_threads} (@pxref{Combining MPI and Threads}). Calling @code{fftw_mpi_init} additional times (before @code{fftw_mpi_cleanup}) has no effect. If you want to deallocate all persistent data and reset FFTW to the pristine state it was in when you started your program, you can call: @findex fftw_mpi_cleanup @example void fftw_mpi_cleanup(void); @end example @findex fftw_cleanup (This calls @code{fftw_cleanup}, so you need not call the serial cleanup routine too, although it is safe to do so.) After calling @code{fftw_mpi_cleanup}, all existing plans become undefined, and you should not attempt to execute or destroy them. You must call @code{fftw_mpi_init} again after @code{fftw_mpi_cleanup} if you want to resume using the MPI FFTW routines. @node Using MPI Plans, MPI Data Distribution Functions, MPI Initialization, FFTW MPI Reference @subsection Using MPI Plans Once an MPI plan is created, you can execute and destroy it using @code{fftw_execute}, @code{fftw_destroy_plan}, and the other functions in the serial interface that operate on generic plans (@pxref{Using Plans}). @cindex collective function @cindex MPI communicator The @code{fftw_execute} and @code{fftw_destroy_plan} functions, applied to MPI plans, are @emph{collective} calls: they must be called for all processes in the communicator that was used to create the plan. @cindex new-array execution You must @emph{not} use the serial new-array plan-execution functions @code{fftw_execute_dft} and so on (@pxref{New-array Execute Functions}) with MPI plans. Such functions are specialized to the problem type, and there are specific new-array execute functions for MPI plans: @findex fftw_mpi_execute_dft @findex fftw_mpi_execute_dft_r2c @findex fftw_mpi_execute_dft_c2r @findex fftw_mpi_execute_r2r @example void fftw_mpi_execute_dft(fftw_plan p, fftw_complex *in, fftw_complex *out); void fftw_mpi_execute_dft_r2c(fftw_plan p, double *in, fftw_complex *out); void fftw_mpi_execute_dft_c2r(fftw_plan p, fftw_complex *in, double *out); void fftw_mpi_execute_r2r(fftw_plan p, double *in, double *out); @end example @cindex alignment @findex fftw_malloc These functions have the same restrictions as those of the serial new-array execute functions. They are @emph{always} safe to apply to the @emph{same} @code{in} and @code{out} arrays that were used to create the plan. They can only be applied to new arrarys if those arrays have the same types, dimensions, in-placeness, and alignment as the original arrays, where the best way to ensure the same alignment is to use FFTW's @code{fftw_malloc} and related allocation functions for all arrays (@pxref{Memory Allocation}). Note that distributed transposes (@pxref{FFTW MPI Transposes}) use @code{fftw_mpi_execute_r2r}, since they count as rank-zero r2r plans from FFTW's perspective. @node MPI Data Distribution Functions, MPI Plan Creation, Using MPI Plans, FFTW MPI Reference @subsection MPI Data Distribution Functions @cindex data distribution As described above (@pxref{MPI Data Distribution}), in order to allocate your arrays, @emph{before} creating a plan, you must first call one of the following routines to determine the required allocation size and the portion of the array locally stored on a given process. The @code{MPI_Comm} communicator passed here must be equivalent to the communicator used below for plan creation. The basic interface for multidimensional transforms consists of the functions: @findex fftw_mpi_local_size_2d @findex fftw_mpi_local_size_3d @findex fftw_mpi_local_size @findex fftw_mpi_local_size_2d_transposed @findex fftw_mpi_local_size_3d_transposed @findex fftw_mpi_local_size_transposed @example ptrdiff_t fftw_mpi_local_size_2d(ptrdiff_t n0, ptrdiff_t n1, MPI_Comm comm, ptrdiff_t *local_n0, ptrdiff_t *local_0_start); ptrdiff_t fftw_mpi_local_size_3d(ptrdiff_t n0, ptrdiff_t n1, ptrdiff_t n2, MPI_Comm comm, ptrdiff_t *local_n0, ptrdiff_t *local_0_start); ptrdiff_t fftw_mpi_local_size(int rnk, const ptrdiff_t *n, MPI_Comm comm, ptrdiff_t *local_n0, ptrdiff_t *local_0_start); ptrdiff_t fftw_mpi_local_size_2d_transposed(ptrdiff_t n0, ptrdiff_t n1, MPI_Comm comm, ptrdiff_t *local_n0, ptrdiff_t *local_0_start, ptrdiff_t *local_n1, ptrdiff_t *local_1_start); ptrdiff_t fftw_mpi_local_size_3d_transposed(ptrdiff_t n0, ptrdiff_t n1, ptrdiff_t n2, MPI_Comm comm, ptrdiff_t *local_n0, ptrdiff_t *local_0_start, ptrdiff_t *local_n1, ptrdiff_t *local_1_start); ptrdiff_t fftw_mpi_local_size_transposed(int rnk, const ptrdiff_t *n, MPI_Comm comm, ptrdiff_t *local_n0, ptrdiff_t *local_0_start, ptrdiff_t *local_n1, ptrdiff_t *local_1_start); @end example These functions return the number of elements to allocate (complex numbers for DFT/r2c/c2r plans, real numbers for r2r plans), whereas the @code{local_n0} and @code{local_0_start} return the portion (@code{local_0_start} to @code{local_0_start + local_n0 - 1}) of the first dimension of an @ndims{} array that is stored on the local process. @xref{Basic and advanced distribution interfaces}. For @code{FFTW_MPI_TRANSPOSED_OUT} plans, the @samp{_transposed} variants are useful in order to also return the local portion of the first dimension in the @ndimstrans{} transposed output. @xref{Transposed distributions}. The advanced interface for multidimensional transforms is: @cindex advanced interface @findex fftw_mpi_local_size_many @findex fftw_mpi_local_size_many_transposed @example ptrdiff_t fftw_mpi_local_size_many(int rnk, const ptrdiff_t *n, ptrdiff_t howmany, ptrdiff_t block0, MPI_Comm comm, ptrdiff_t *local_n0, ptrdiff_t *local_0_start); ptrdiff_t fftw_mpi_local_size_many_transposed(int rnk, const ptrdiff_t *n, ptrdiff_t howmany, ptrdiff_t block0, ptrdiff_t block1, MPI_Comm comm, ptrdiff_t *local_n0, ptrdiff_t *local_0_start, ptrdiff_t *local_n1, ptrdiff_t *local_1_start); @end example These differ from the basic interface in only two ways. First, they allow you to specify block sizes @code{block0} and @code{block1} (the latter for the transposed output); you can pass @code{FFTW_MPI_DEFAULT_BLOCK} to use FFTW's default block size as in the basic interface. Second, you can pass a @code{howmany} parameter, corresponding to the advanced planning interface below: this is for transforms of contiguous @code{howmany}-tuples of numbers (@code{howmany = 1} in the basic interface). The corresponding basic and advanced routines for one-dimensional transforms (currently only complex DFTs) are: @findex fftw_mpi_local_size_1d @findex fftw_mpi_local_size_many_1d @example ptrdiff_t fftw_mpi_local_size_1d( ptrdiff_t n0, MPI_Comm comm, int sign, unsigned flags, ptrdiff_t *local_ni, ptrdiff_t *local_i_start, ptrdiff_t *local_no, ptrdiff_t *local_o_start); ptrdiff_t fftw_mpi_local_size_many_1d( ptrdiff_t n0, ptrdiff_t howmany, MPI_Comm comm, int sign, unsigned flags, ptrdiff_t *local_ni, ptrdiff_t *local_i_start, ptrdiff_t *local_no, ptrdiff_t *local_o_start); @end example @ctindex FFTW_MPI_SCRAMBLED_OUT @ctindex FFTW_MPI_SCRAMBLED_IN As above, the return value is the number of elements to allocate (complex numbers, for complex DFTs). The @code{local_ni} and @code{local_i_start} arguments return the portion (@code{local_i_start} to @code{local_i_start + local_ni - 1}) of the 1d array that is stored on this process for the transform @emph{input}, and @code{local_no} and @code{local_o_start} are the corresponding quantities for the input. The @code{sign} (@code{FFTW_FORWARD} or @code{FFTW_BACKWARD}) and @code{flags} must match the arguments passed when creating a plan. Although the inputs and outputs have different data distributions in general, it is guaranteed that the @emph{output} data distribution of an @code{FFTW_FORWARD} plan will match the @emph{input} data distribution of an @code{FFTW_BACKWARD} plan and vice versa; similarly for the @code{FFTW_MPI_SCRAMBLED_OUT} and @code{FFTW_MPI_SCRAMBLED_IN} flags. @xref{One-dimensional distributions}. @node MPI Plan Creation, MPI Wisdom Communication, MPI Data Distribution Functions, FFTW MPI Reference @subsection MPI Plan Creation @subsubheading Complex-data MPI DFTs Plans for complex-data DFTs (@pxref{2d MPI example}) are created by: @findex fftw_mpi_plan_dft_1d @findex fftw_mpi_plan_dft_2d @findex fftw_mpi_plan_dft_3d @findex fftw_mpi_plan_dft @findex fftw_mpi_plan_many_dft @example fftw_plan fftw_mpi_plan_dft_1d(ptrdiff_t n0, fftw_complex *in, fftw_complex *out, MPI_Comm comm, int sign, unsigned flags); fftw_plan fftw_mpi_plan_dft_2d(ptrdiff_t n0, ptrdiff_t n1, fftw_complex *in, fftw_complex *out, MPI_Comm comm, int sign, unsigned flags); fftw_plan fftw_mpi_plan_dft_3d(ptrdiff_t n0, ptrdiff_t n1, ptrdiff_t n2, fftw_complex *in, fftw_complex *out, MPI_Comm comm, int sign, unsigned flags); fftw_plan fftw_mpi_plan_dft(int rnk, const ptrdiff_t *n, fftw_complex *in, fftw_complex *out, MPI_Comm comm, int sign, unsigned flags); fftw_plan fftw_mpi_plan_many_dft(int rnk, const ptrdiff_t *n, ptrdiff_t howmany, ptrdiff_t block, ptrdiff_t tblock, fftw_complex *in, fftw_complex *out, MPI_Comm comm, int sign, unsigned flags); @end example @cindex MPI communicator @cindex collective function These are similar to their serial counterparts (@pxref{Complex DFTs}) in specifying the dimensions, sign, and flags of the transform. The @code{comm} argument gives an MPI communicator that specifies the set of processes to participate in the transform; plan creation is a collective function that must be called for all processes in the communicator. The @code{in} and @code{out} pointers refer only to a portion of the overall transform data (@pxref{MPI Data Distribution}) as specified by the @samp{local_size} functions in the previous section. Unless @code{flags} contains @code{FFTW_ESTIMATE}, these arrays are overwritten during plan creation as for the serial interface. For multi-dimensional transforms, any dimensions @code{> 1} are supported; for one-dimensional transforms, only composite (non-prime) @code{n0} are currently supported (unlike the serial FFTW). Requesting an unsupported transform size will yield a @code{NULL} plan. (As in the serial interface, highly composite sizes generally yield the best performance.) @cindex advanced interface @ctindex FFTW_MPI_DEFAULT_BLOCK @cindex stride The advanced-interface @code{fftw_mpi_plan_many_dft} additionally allows you to specify the block sizes for the first dimension (@code{block}) of the @ndims{} input data and the first dimension (@code{tblock}) of the @ndimstrans{} transposed data (at intermediate steps of the transform, and for the output if @code{FFTW_TRANSPOSED_OUT} is specified in @code{flags}). These must be the same block sizes as were passed to the corresponding @samp{local_size} function; you can pass @code{FFTW_MPI_DEFAULT_BLOCK} to use FFTW's default block size as in the basic interface. Also, the @code{howmany} parameter specifies that the transform is of contiguous @code{howmany}-tuples rather than individual complex numbers; this corresponds to the same parameter in the serial advanced interface (@pxref{Advanced Complex DFTs}) with @code{stride = howmany} and @code{dist = 1}. @subsubheading MPI flags The @code{flags} can be any of those for the serial FFTW (@pxref{Planner Flags}), and in addition may include one or more of the following MPI-specific flags, which improve performance at the cost of changing the output or input data formats. @itemize @bullet @item @ctindex FFTW_MPI_SCRAMBLED_OUT @ctindex FFTW_MPI_SCRAMBLED_IN @code{FFTW_MPI_SCRAMBLED_OUT}, @code{FFTW_MPI_SCRAMBLED_IN}: valid for 1d transforms only, these flags indicate that the output/input of the transform are in an undocumented ``scrambled'' order. A forward @code{FFTW_MPI_SCRAMBLED_OUT} transform can be inverted by a backward @code{FFTW_MPI_SCRAMBLED_IN} (times the usual 1/@i{N} normalization). @xref{One-dimensional distributions}. @item @ctindex FFTW_MPI_TRANSPOSED_OUT @ctindex FFTW_MPI_TRANSPOSED_IN @code{FFTW_MPI_TRANSPOSED_OUT}, @code{FFTW_MPI_TRANSPOSED_IN}: valid for multidimensional (@code{rnk > 1}) transforms only, these flags specify that the output or input of an @ndims{} transform is transposed to @ndimstrans{}. @xref{Transposed distributions}. @end itemize @subsubheading Real-data MPI DFTs @cindex r2c Plans for real-input/output (r2c/c2r) DFTs (@pxref{Multi-dimensional MPI DFTs of Real Data}) are created by: @findex fftw_mpi_plan_dft_r2c_2d @findex fftw_mpi_plan_dft_r2c_2d @findex fftw_mpi_plan_dft_r2c_3d @findex fftw_mpi_plan_dft_r2c @findex fftw_mpi_plan_dft_c2r_2d @findex fftw_mpi_plan_dft_c2r_2d @findex fftw_mpi_plan_dft_c2r_3d @findex fftw_mpi_plan_dft_c2r @example fftw_plan fftw_mpi_plan_dft_r2c_2d(ptrdiff_t n0, ptrdiff_t n1, double *in, fftw_complex *out, MPI_Comm comm, unsigned flags); fftw_plan fftw_mpi_plan_dft_r2c_2d(ptrdiff_t n0, ptrdiff_t n1, double *in, fftw_complex *out, MPI_Comm comm, unsigned flags); fftw_plan fftw_mpi_plan_dft_r2c_3d(ptrdiff_t n0, ptrdiff_t n1, ptrdiff_t n2, double *in, fftw_complex *out, MPI_Comm comm, unsigned flags); fftw_plan fftw_mpi_plan_dft_r2c(int rnk, const ptrdiff_t *n, double *in, fftw_complex *out, MPI_Comm comm, unsigned flags); fftw_plan fftw_mpi_plan_dft_c2r_2d(ptrdiff_t n0, ptrdiff_t n1, fftw_complex *in, double *out, MPI_Comm comm, unsigned flags); fftw_plan fftw_mpi_plan_dft_c2r_2d(ptrdiff_t n0, ptrdiff_t n1, fftw_complex *in, double *out, MPI_Comm comm, unsigned flags); fftw_plan fftw_mpi_plan_dft_c2r_3d(ptrdiff_t n0, ptrdiff_t n1, ptrdiff_t n2, fftw_complex *in, double *out, MPI_Comm comm, unsigned flags); fftw_plan fftw_mpi_plan_dft_c2r(int rnk, const ptrdiff_t *n, fftw_complex *in, double *out, MPI_Comm comm, unsigned flags); @end example Similar to the serial interface (@pxref{Real-data DFTs}), these transform logically @ndims{} real data to/from @ndimshalf{} complex data, representing the non-redundant half of the conjugate-symmetry output of a real-input DFT (@pxref{Multi-dimensional Transforms}). However, the real array must be stored within a padded @ndimspad{} array (much like the in-place serial r2c transforms, but here for out-of-place transforms as well). Currently, only multi-dimensional (@code{rnk > 1}) r2c/c2r transforms are supported (requesting a plan for @code{rnk = 1} will yield @code{NULL}). As explained above (@pxref{Multi-dimensional MPI DFTs of Real Data}), the data distribution of both the real and complex arrays is given by the @samp{local_size} function called for the dimensions of the @emph{complex} array. Similar to the other planning functions, the input and output arrays are overwritten when the plan is created except in @code{FFTW_ESTIMATE} mode. As for the complex DFTs above, there is an advance interface that allows you to manually specify block sizes and to transform contiguous @code{howmany}-tuples of real/complex numbers: @findex fftw_mpi_plan_many_dft_r2c @findex fftw_mpi_plan_many_dft_c2r @example fftw_plan fftw_mpi_plan_many_dft_r2c (int rnk, const ptrdiff_t *n, ptrdiff_t howmany, ptrdiff_t iblock, ptrdiff_t oblock, double *in, fftw_complex *out, MPI_Comm comm, unsigned flags); fftw_plan fftw_mpi_plan_many_dft_c2r (int rnk, const ptrdiff_t *n, ptrdiff_t howmany, ptrdiff_t iblock, ptrdiff_t oblock, fftw_complex *in, double *out, MPI_Comm comm, unsigned flags); @end example @subsubheading MPI r2r transforms @cindex r2r There are corresponding plan-creation routines for r2r transforms (@pxref{More DFTs of Real Data}), currently supporting multidimensional (@code{rnk > 1}) transforms only (@code{rnk = 1} will yield a @code{NULL} plan): @example fftw_plan fftw_mpi_plan_r2r_2d(ptrdiff_t n0, ptrdiff_t n1, double *in, double *out, MPI_Comm comm, fftw_r2r_kind kind0, fftw_r2r_kind kind1, unsigned flags); fftw_plan fftw_mpi_plan_r2r_3d(ptrdiff_t n0, ptrdiff_t n1, ptrdiff_t n2, double *in, double *out, MPI_Comm comm, fftw_r2r_kind kind0, fftw_r2r_kind kind1, fftw_r2r_kind kind2, unsigned flags); fftw_plan fftw_mpi_plan_r2r(int rnk, const ptrdiff_t *n, double *in, double *out, MPI_Comm comm, const fftw_r2r_kind *kind, unsigned flags); fftw_plan fftw_mpi_plan_many_r2r(int rnk, const ptrdiff_t *n, ptrdiff_t iblock, ptrdiff_t oblock, double *in, double *out, MPI_Comm comm, const fftw_r2r_kind *kind, unsigned flags); @end example The parameters are much the same as for the complex DFTs above, except that the arrays are of real numbers (and hence the outputs of the @samp{local_size} data-distribution functions should be interpreted as counts of real rather than complex numbers). Also, the @code{kind} parameters specify the r2r kinds along each dimension as for the serial interface (@pxref{Real-to-Real Transform Kinds}). @xref{Other Multi-dimensional Real-data MPI Transforms}. @subsubheading MPI transposition @cindex transpose FFTW also provides routines to plan a transpose of a distributed @code{n0} by @code{n1} array of real numbers, or an array of @code{howmany}-tuples of real numbers with specified block sizes (@pxref{FFTW MPI Transposes}): @findex fftw_mpi_plan_transpose @findex fftw_mpi_plan_many_transpose @example fftw_plan fftw_mpi_plan_transpose(ptrdiff_t n0, ptrdiff_t n1, double *in, double *out, MPI_Comm comm, unsigned flags); fftw_plan fftw_mpi_plan_many_transpose (ptrdiff_t n0, ptrdiff_t n1, ptrdiff_t howmany, ptrdiff_t block0, ptrdiff_t block1, double *in, double *out, MPI_Comm comm, unsigned flags); @end example @cindex new-array execution @findex fftw_mpi_execute_r2r These plans are used with the @code{fftw_mpi_execute_r2r} new-array execute function (@pxref{Using MPI Plans }), since they count as (rank zero) r2r plans from FFTW's perspective. @node MPI Wisdom Communication, , MPI Plan Creation, FFTW MPI Reference @subsection MPI Wisdom Communication To facilitate synchronizing wisdom among the different MPI processes, we provide two functions: @findex fftw_mpi_gather_wisdom @findex fftw_mpi_broadcast_wisdom @example void fftw_mpi_gather_wisdom(MPI_Comm comm); void fftw_mpi_broadcast_wisdom(MPI_Comm comm); @end example The @code{fftw_mpi_gather_wisdom} function gathers all wisdom in the given communicator @code{comm} to the process of rank 0 in the communicator: that process obtains the union of all wisdom on all the processes. As a side effect, some other processes will gain additional wisdom from other processes, but only process 0 will gain the complete union. The @code{fftw_mpi_broadcast_wisdom} does the reverse: it exports wisdom from process 0 in @code{comm} to all other processes in the communicator, replacing any wisdom they currently have. @xref{FFTW MPI Wisdom}. @c ------------------------------------------------------------ @node FFTW MPI Fortran Interface, , FFTW MPI Reference, Distributed-memory FFTW with MPI @section FFTW MPI Fortran Interface @cindex Fortran interface @cindex iso_c_binding The FFTW MPI interface is callable from modern Fortran compilers supporting the Fortran 2003 @code{iso_c_binding} standard for calling C functions. As described in @ref{Calling FFTW from Modern Fortran}, this means that you can directly call FFTW's C interface from Fortran with only minor changes in syntax. There are, however, a few things specific to the MPI interface to keep in mind: @itemize @bullet @item Instead of including @code{fftw3.f03} as in @ref{Overview of Fortran interface }, you should @code{include 'fftw3-mpi.f03'} (after @code{use, intrinsic :: iso_c_binding} as before). The @code{fftw3-mpi.f03} file includes @code{fftw3.f03}, so you should @emph{not} @code{include} them both yourself. (You will also want to include the MPI header file, usually via @code{include 'mpif.h'} or similar, although though this is not needed by @code{fftw3-mpi.f03} @i{per se}.) (To use the @samp{fftwl_} @code{long double} extended-precision routines in supporting compilers, you should include @code{fftw3f-mpi.f03} in @emph{addition} to @code{fftw3-mpi.f03}. @xref{Extended and quadruple precision in Fortran}.) @item Because of the different storage conventions between C and Fortran, you reverse the order of your array dimensions when passing them to FFTW (@pxref{Reversing array dimensions}). This is merely a difference in notation and incurs no performance overhead. However, it means that, whereas in C the @emph{first} dimension is distributed, in Fortran the @emph{last} dimension of your array is distributed. @item @cindex MPI communicator In Fortran, communicators are stored as @code{integer} types; there is no @code{MPI_Comm} type, nor is there any way to access a C @code{MPI_Comm}. Fortunately, this is taken care of for you by the FFTW Fortran interface: whenever the C interface expects an @code{MPI_Comm} type, you should pass the Fortran communicator as an @code{integer}.@footnote{Technically, this is because you aren't actually calling the C functions directly. You are calling wrapper functions that translate the communicator with @code{MPI_Comm_f2c} before calling the ordinary C interface. This is all done transparently, however, since the @code{fftw3-mpi.f03} interface file renames the wrappers so that they are called in Fortran with the same names as the C interface functions.} @item Because you need to call the @samp{local_size} function to find out how much space to allocate, and this may be @emph{larger} than the local portion of the array (@pxref{MPI Data Distribution}), you should @emph{always} allocate your arrays dynamically using FFTW's allocation routines as described in @ref{Allocating aligned memory in Fortran}. (Coincidentally, this also provides the best performance by guaranteeding proper data alignment.) @item Because all sizes in the MPI FFTW interface are declared as @code{ptrdiff_t} in C, you should use @code{integer(C_INTPTR_T)} in Fortran (@pxref{FFTW Fortran type reference}). @item @findex fftw_execute_dft @findex fftw_mpi_execute_dft @cindex new-array execution In Fortran, because of the language semantics, we generally recommend using the new-array execute functions for all plans, even in the common case where you are executing the plan on the same arrays for which the plan was created (@pxref{Plan execution in Fortran}). However, note that in the MPI interface these functions are changed: @code{fftw_execute_dft} becomes @code{fftw_mpi_execute_dft}, etcetera. @xref{Using MPI Plans}. @end itemize For example, here is a Fortran code snippet to perform a distributed @twodims{L,M} complex DFT in-place. (This assumes you have already initialized MPI with @code{MPI_init} and have also performed @code{call fftw_mpi_init}.) @example use, intrinsic :: iso_c_binding include 'fftw3-mpi.f03' integer(C_INTPTR_T), parameter :: L = ... integer(C_INTPTR_T), parameter :: M = ... type(C_PTR) :: plan, cdata complex(C_DOUBLE_COMPLEX), pointer :: data(:,:) integer(C_INTPTR_T) :: i, j, alloc_local, local_M, local_j_offset ! @r{get local data size and allocate (note dimension reversal)} alloc_local = fftw_mpi_local_size_2d(M, L, MPI_COMM_WORLD, & local_M, local_j_offset) cdata = fftw_alloc_complex(alloc_local) call c_f_pointer(cdata, data, [L,local_M]) ! @r{create MPI plan for in-place forward DFT (note dimension reversal)} plan = fftw_mpi_plan_dft_2d(M, L, data, data, MPI_COMM_WORLD, & FFTW_FORWARD, FFTW_MEASURE) ! @r{initialize data to some function} my_function(i,j) do j = 1, local_M do i = 1, L data(i, j) = my_function(i, j + local_j_offset) end do end do ! @r{compute transform (as many times as desired)} call fftw_mpi_execute_dft(plan, data, data) call fftw_destroy_plan(plan) call fftw_free(cdata) @end example Note that when we called @code{fftw_mpi_local_size_2d} and @code{fftw_mpi_plan_dft_2d} with the dimensions in reversed order, since a @twodims{L,M} Fortran array is viewed by FFTW in C as a @twodims{M, L} array. This means that the array was distributed over the @code{M} dimension, the local portion of which is a @twodims{L,local_M} array in Fortran. (You must @emph{not} use an @code{allocate} statement to allocate an @twodims{L,local_M} array, however; you must allocate @code{alloc_local} complex numbers, which may be greater than @code{L * local_M}, in order to reserve space for intermediate steps of the transform.) Finally, we mention that because C's array indices are zero-based, the @code{local_j_offset} argument can conveniently be interpreted as an offset in the 1-based @code{j} index (rather than as a starting index as in C). If instead you had used the @code{ior(FFTW_MEASURE, FFTW_MPI_TRANSPOSED_OUT)} flag, the output of the transform would be a transposed @twodims{M,local_L} array, associated with the @emph{same} @code{cdata} allocation (since the transform is in-place), and which you could declare with: @example complex(C_DOUBLE_COMPLEX), pointer :: tdata(:,:) ... call c_f_pointer(cdata, tdata, [M,local_L]) @end example where @code{local_L} would have been obtained by changing the @code{fftw_mpi_local_size_2d} call to: @example alloc_local = fftw_mpi_local_size_2d_transposed(M, L, MPI_COMM_WORLD, & local_M, local_j_offset, local_L, local_i_offset) @end example fftw-3.3.3/doc/other.texi0000644000175000017500000004055212054410050012165 00000000000000@node Other Important Topics, FFTW Reference, Tutorial, Top @chapter Other Important Topics @menu * SIMD alignment and fftw_malloc:: * Multi-dimensional Array Format:: * Words of Wisdom-Saving Plans:: * Caveats in Using Wisdom:: @end menu @c ------------------------------------------------------------ @node SIMD alignment and fftw_malloc, Multi-dimensional Array Format, Other Important Topics, Other Important Topics @section SIMD alignment and fftw_malloc SIMD, which stands for ``Single Instruction Multiple Data,'' is a set of special operations supported by some processors to perform a single operation on several numbers (usually 2 or 4) simultaneously. SIMD floating-point instructions are available on several popular CPUs: SSE/SSE2/AVX on recent x86/x86-64 processors, AltiVec (single precision) on some PowerPCs (Apple G4 and higher), NEON on some ARM models, and MIPS Paired Single (currently only in FFTW 3.2.x). FFTW can be compiled to support the SIMD instructions on any of these systems. @cindex SIMD @cindex SSE @cindex SSE2 @cindex AVX @cindex AltiVec @cindex MIPS PS @cindex precision A program linking to an FFTW library compiled with SIMD support can obtain a nonnegligible speedup for most complex and r2c/c2r transforms. In order to obtain this speedup, however, the arrays of complex (or real) data passed to FFTW must be specially aligned in memory (typically 16-byte aligned), and often this alignment is more stringent than that provided by the usual @code{malloc} (etc.) allocation routines. @cindex portability In order to guarantee proper alignment for SIMD, therefore, in case your program is ever linked against a SIMD-using FFTW, we recommend allocating your transform data with @code{fftw_malloc} and de-allocating it with @code{fftw_free}. @findex fftw_malloc @findex fftw_free These have exactly the same interface and behavior as @code{malloc}/@code{free}, except that for a SIMD FFTW they ensure that the returned pointer has the necessary alignment (by calling @code{memalign} or its equivalent on your OS). You are not @emph{required} to use @code{fftw_malloc}. You can allocate your data in any way that you like, from @code{malloc} to @code{new} (in C++) to a fixed-size array declaration. If the array happens not to be properly aligned, FFTW will not use the SIMD extensions. @cindex C++ @findex fftw_alloc_real @findex fftw_alloc_complex Since @code{fftw_malloc} only ever needs to be used for real and complex arrays, we provide two convenient wrapper routines @code{fftw_alloc_real(N)} and @code{fftw_alloc_complex(N)} that are equivalent to @code{(double*)fftw_malloc(sizeof(double) * N)} and @code{(fftw_complex*)fftw_malloc(sizeof(fftw_complex) * N)}, respectively (or their equivalents in other precisions). @c ------------------------------------------------------------ @node Multi-dimensional Array Format, Words of Wisdom-Saving Plans, SIMD alignment and fftw_malloc, Other Important Topics @section Multi-dimensional Array Format This section describes the format in which multi-dimensional arrays are stored in FFTW. We felt that a detailed discussion of this topic was necessary. Since several different formats are common, this topic is often a source of confusion. @menu * Row-major Format:: * Column-major Format:: * Fixed-size Arrays in C:: * Dynamic Arrays in C:: * Dynamic Arrays in C-The Wrong Way:: @end menu @c =========> @node Row-major Format, Column-major Format, Multi-dimensional Array Format, Multi-dimensional Array Format @subsection Row-major Format @cindex row-major The multi-dimensional arrays passed to @code{fftw_plan_dft} etcetera are expected to be stored as a single contiguous block in @dfn{row-major} order (sometimes called ``C order''). Basically, this means that as you step through adjacent memory locations, the first dimension's index varies most slowly and the last dimension's index varies most quickly. To be more explicit, let us consider an array of rank @math{d} whose dimensions are @ndims{}. Now, we specify a location in the array by a sequence of @math{d} (zero-based) indices, one for each dimension: @tex $(i_0, i_1, i_2, \ldots, i_{d-1})$. @end tex @ifinfo (i[0], i[1], ..., i[d-1]). @end ifinfo @html (i0, i1, i2,..., id-1). @end html If the array is stored in row-major order, then this element is located at the position @tex $i_{d-1} + n_{d-1} (i_{d-2} + n_{d-2} (\ldots + n_1 i_0))$. @end tex @ifinfo i[d-1] + n[d-1] * (i[d-2] + n[d-2] * (... + n[1] * i[0])). @end ifinfo @html id-1 + nd-1 * (id-2 + nd-2 * (... + n1 * i0)). @end html Note that, for the ordinary complex DFT, each element of the array must be of type @code{fftw_complex}; i.e. a (real, imaginary) pair of (double-precision) numbers. In the advanced FFTW interface, the physical dimensions @math{n} from which the indices are computed can be different from (larger than) the logical dimensions of the transform to be computed, in order to transform a subset of a larger array. @cindex advanced interface Note also that, in the advanced interface, the expression above is multiplied by a @dfn{stride} to get the actual array index---this is useful in situations where each element of the multi-dimensional array is actually a data structure (or another array), and you just want to transform a single field. In the basic interface, however, the stride is 1. @cindex stride @c =========> @node Column-major Format, Fixed-size Arrays in C, Row-major Format, Multi-dimensional Array Format @subsection Column-major Format @cindex column-major Readers from the Fortran world are used to arrays stored in @dfn{column-major} order (sometimes called ``Fortran order''). This is essentially the exact opposite of row-major order in that, here, the @emph{first} dimension's index varies most quickly. If you have an array stored in column-major order and wish to transform it using FFTW, it is quite easy to do. When creating the plan, simply pass the dimensions of the array to the planner in @emph{reverse order}. For example, if your array is a rank three @code{N x M x L} matrix in column-major order, you should pass the dimensions of the array as if it were an @code{L x M x N} matrix (which it is, from the perspective of FFTW). This is done for you @emph{automatically} by the FFTW legacy-Fortran interface (@pxref{Calling FFTW from Legacy Fortran}), but you must do it manually with the modern Fortran interface (@pxref{Reversing array dimensions}). @cindex Fortran interface @c =========> @node Fixed-size Arrays in C, Dynamic Arrays in C, Column-major Format, Multi-dimensional Array Format @subsection Fixed-size Arrays in C @cindex C multi-dimensional arrays A multi-dimensional array whose size is declared at compile time in C is @emph{already} in row-major order. You don't have to do anything special to transform it. For example: @example @{ fftw_complex data[N0][N1][N2]; fftw_plan plan; ... plan = fftw_plan_dft_3d(N0, N1, N2, &data[0][0][0], &data[0][0][0], FFTW_FORWARD, FFTW_ESTIMATE); ... @} @end example This will plan a 3d in-place transform of size @code{N0 x N1 x N2}. Notice how we took the address of the zero-th element to pass to the planner (we could also have used a typecast). However, we tend to @emph{discourage} users from declaring their arrays in this way, for two reasons. First, this allocates the array on the stack (``automatic'' storage), which has a very limited size on most operating systems (declaring an array with more than a few thousand elements will often cause a crash). (You can get around this limitation on many systems by declaring the array as @code{static} and/or global, but that has its own drawbacks.) Second, it may not optimally align the array for use with a SIMD FFTW (@pxref{SIMD alignment and fftw_malloc}). Instead, we recommend using @code{fftw_malloc}, as described below. @c =========> @node Dynamic Arrays in C, Dynamic Arrays in C-The Wrong Way, Fixed-size Arrays in C, Multi-dimensional Array Format @subsection Dynamic Arrays in C We recommend allocating most arrays dynamically, with @code{fftw_malloc}. This isn't too hard to do, although it is not as straightforward for multi-dimensional arrays as it is for one-dimensional arrays. Creating the array is simple: using a dynamic-allocation routine like @code{fftw_malloc}, allocate an array big enough to store N @code{fftw_complex} values (for a complex DFT), where N is the product of the sizes of the array dimensions (i.e. the total number of complex values in the array). For example, here is code to allocate a @threedims{5,12,27} rank-3 array: @findex fftw_malloc @example fftw_complex *an_array; an_array = (fftw_complex*) fftw_malloc(5*12*27 * sizeof(fftw_complex)); @end example Accessing the array elements, however, is more tricky---you can't simply use multiple applications of the @samp{[]} operator like you could for fixed-size arrays. Instead, you have to explicitly compute the offset into the array using the formula given earlier for row-major arrays. For example, to reference the @math{(i,j,k)}-th element of the array allocated above, you would use the expression @code{an_array[k + 27 * (j + 12 * i)]}. This pain can be alleviated somewhat by defining appropriate macros, or, in C++, creating a class and overloading the @samp{()} operator. The recent C99 standard provides a way to reinterpret the dynamic array as a ``variable-length'' multi-dimensional array amenable to @samp{[]}, but this feature is not yet widely supported by compilers. @cindex C99 @cindex C++ @c =========> @node Dynamic Arrays in C-The Wrong Way, , Dynamic Arrays in C, Multi-dimensional Array Format @subsection Dynamic Arrays in C---The Wrong Way A different method for allocating multi-dimensional arrays in C is often suggested that is incompatible with FFTW: @emph{using it will cause FFTW to die a painful death}. We discuss the technique here, however, because it is so commonly known and used. This method is to create arrays of pointers of arrays of pointers of @dots{}etcetera. For example, the analogue in this method to the example above is: @example int i,j; fftw_complex ***a_bad_array; /* @r{another way to make a 5x12x27 array} */ a_bad_array = (fftw_complex ***) malloc(5 * sizeof(fftw_complex **)); for (i = 0; i < 5; ++i) @{ a_bad_array[i] = (fftw_complex **) malloc(12 * sizeof(fftw_complex *)); for (j = 0; j < 12; ++j) a_bad_array[i][j] = (fftw_complex *) malloc(27 * sizeof(fftw_complex)); @} @end example As you can see, this sort of array is inconvenient to allocate (and deallocate). On the other hand, it has the advantage that the @math{(i,j,k)}-th element can be referenced simply by @code{a_bad_array[i][j][k]}. If you like this technique and want to maximize convenience in accessing the array, but still want to pass the array to FFTW, you can use a hybrid method. Allocate the array as one contiguous block, but also declare an array of arrays of pointers that point to appropriate places in the block. That sort of trick is beyond the scope of this documentation; for more information on multi-dimensional arrays in C, see the @code{comp.lang.c} @uref{http://c-faq.com/aryptr/dynmuldimary.html, FAQ}. @c ------------------------------------------------------------ @node Words of Wisdom-Saving Plans, Caveats in Using Wisdom, Multi-dimensional Array Format, Other Important Topics @section Words of Wisdom---Saving Plans @cindex wisdom @cindex saving plans to disk FFTW implements a method for saving plans to disk and restoring them. In fact, what FFTW does is more general than just saving and loading plans. The mechanism is called @dfn{wisdom}. Here, we describe this feature at a high level. @xref{FFTW Reference}, for a less casual but more complete discussion of how to use wisdom in FFTW. Plans created with the @code{FFTW_MEASURE}, @code{FFTW_PATIENT}, or @code{FFTW_EXHAUSTIVE} options produce near-optimal FFT performance, but may require a long time to compute because FFTW must measure the runtime of many possible plans and select the best one. This setup is designed for the situations where so many transforms of the same size must be computed that the start-up time is irrelevant. For short initialization times, but slower transforms, we have provided @code{FFTW_ESTIMATE}. The @code{wisdom} mechanism is a way to get the best of both worlds: you compute a good plan once, save it to disk, and later reload it as many times as necessary. The wisdom mechanism can actually save and reload many plans at once, not just one. @ctindex FFTW_MEASURE @ctindex FFTW_PATIENT @ctindex FFTW_EXHAUSTIVE @ctindex FFTW_ESTIMATE Whenever you create a plan, the FFTW planner accumulates wisdom, which is information sufficient to reconstruct the plan. After planning, you can save this information to disk by means of the function: @example int fftw_export_wisdom_to_filename(const char *filename); @end example @findex fftw_export_wisdom_to_filename (This function returns non-zero on success.) The next time you run the program, you can restore the wisdom with @code{fftw_import_wisdom_from_filename} (which also returns non-zero on success), and then recreate the plan using the same flags as before. @example int fftw_import_wisdom_from_filename(const char *filename); @end example @findex fftw_import_wisdom_from_filename Wisdom is automatically used for any size to which it is applicable, as long as the planner flags are not more ``patient'' than those with which the wisdom was created. For example, wisdom created with @code{FFTW_MEASURE} can be used if you later plan with @code{FFTW_ESTIMATE} or @code{FFTW_MEASURE}, but not with @code{FFTW_PATIENT}. The @code{wisdom} is cumulative, and is stored in a global, private data structure managed internally by FFTW. The storage space required is minimal, proportional to the logarithm of the sizes the wisdom was generated from. If memory usage is a concern, however, the wisdom can be forgotten and its associated memory freed by calling: @example void fftw_forget_wisdom(void); @end example @findex fftw_forget_wisdom Wisdom can be exported to a file, a string, or any other medium. For details, see @ref{Wisdom}. @node Caveats in Using Wisdom, , Words of Wisdom-Saving Plans, Other Important Topics @section Caveats in Using Wisdom @cindex wisdom, problems with @quotation @html @end html For in much wisdom is much grief, and he that increaseth knowledge increaseth sorrow. @html @end html [Ecclesiastes 1:18] @cindex Ecclesiastes @end quotation @iftex @medskip @end iftex @cindex portability There are pitfalls to using wisdom, in that it can negate FFTW's ability to adapt to changing hardware and other conditions. For example, it would be perfectly possible to export wisdom from a program running on one processor and import it into a program running on another processor. Doing so, however, would mean that the second program would use plans optimized for the first processor, instead of the one it is running on. It should be safe to reuse wisdom as long as the hardware and program binaries remain unchanged. (Actually, the optimal plan may change even between runs of the same binary on identical hardware, due to differences in the virtual memory environment, etcetera. Users seriously interested in performance should worry about this problem, too.) It is likely that, if the same wisdom is used for two different program binaries, even running on the same machine, the plans may be sub-optimal because of differing code alignments. It is therefore wise to recreate wisdom every time an application is recompiled. The more the underlying hardware and software changes between the creation of wisdom and its use, the greater grows the risk of sub-optimal plans. Nevertheless, if the choice is between using @code{FFTW_ESTIMATE} or using possibly-suboptimal wisdom (created on the same machine, but for a different binary), the wisdom is likely to be better. For this reason, we provide a function to import wisdom from a standard system-wide location (@code{/etc/fftw/wisdom} on Unix): @cindex wisdom, system-wide @example int fftw_import_system_wisdom(void); @end example @findex fftw_import_system_wisdom FFTW also provides a standalone program, @code{fftw-wisdom} (described by its own @code{man} page on Unix) with which users can create wisdom, e.g. for a canonical set of sizes to store in the system wisdom file. @xref{Wisdom Utilities}. @cindex fftw-wisdom utility fftw-3.3.3/doc/reference.texi0000644000175000017500000026263212054410050013007 00000000000000@node FFTW Reference, Multi-threaded FFTW, Other Important Topics, Top @chapter FFTW Reference This chapter provides a complete reference for all sequential (i.e., one-processor) FFTW functions. Parallel transforms are described in later chapters. @menu * Data Types and Files:: * Using Plans:: * Basic Interface:: * Advanced Interface:: * Guru Interface:: * New-array Execute Functions:: * Wisdom:: * What FFTW Really Computes:: @end menu @c ------------------------------------------------------------ @node Data Types and Files, Using Plans, FFTW Reference, FFTW Reference @section Data Types and Files All programs using FFTW should include its header file: @example #include @end example You must also link to the FFTW library. On Unix, this means adding @code{-lfftw3 -lm} at the @emph{end} of the link command. @menu * Complex numbers:: * Precision:: * Memory Allocation:: @end menu @c =========> @node Complex numbers, Precision, Data Types and Files, Data Types and Files @subsection Complex numbers The default FFTW interface uses @code{double} precision for all floating-point numbers, and defines a @code{fftw_complex} type to hold complex numbers as: @example typedef double fftw_complex[2]; @end example @tindex fftw_complex Here, the @code{[0]} element holds the real part and the @code{[1]} element holds the imaginary part. Alternatively, if you have a C compiler (such as @code{gcc}) that supports the C99 revision of the ANSI C standard, you can use C's new native complex type (which is binary-compatible with the typedef above). In particular, if you @code{#include } @emph{before} @code{}, then @code{fftw_complex} is defined to be the native complex type and you can manipulate it with ordinary arithmetic (e.g. @code{x = y * (3+4*I)}, where @code{x} and @code{y} are @code{fftw_complex} and @code{I} is the standard symbol for the imaginary unit); @cindex C99 C++ has its own @code{complex} template class, defined in the standard @code{} header file. Reportedly, the C++ standards committee has recently agreed to mandate that the storage format used for this type be binary-compatible with the C99 type, i.e. an array @code{T[2]} with consecutive real @code{[0]} and imaginary @code{[1]} parts. (See report @uref{http://www.open-std.org/jtc1/sc22/WG21/docs/papers/2002/n1388.pdf WG21/N1388}.) Although not part of the official standard as of this writing, the proposal stated that: ``This solution has been tested with all current major implementations of the standard library and shown to be working.'' To the extent that this is true, if you have a variable @code{complex *x}, you can pass it directly to FFTW via @code{reinterpret_cast(x)}. @cindex C++ @cindex portability @c =========> @node Precision, Memory Allocation, Complex numbers, Data Types and Files @subsection Precision @cindex precision You can install single and long-double precision versions of FFTW, which replace @code{double} with @code{float} and @code{long double}, respectively (@pxref{Installation and Customization}). To use these interfaces, you: @itemize @bullet @item Link to the single/long-double libraries; on Unix, @code{-lfftw3f} or @code{-lfftw3l} instead of (or in addition to) @code{-lfftw3}. (You can link to the different-precision libraries simultaneously.) @item Include the @emph{same} @code{} header file. @item Replace all lowercase instances of @samp{fftw_} with @samp{fftwf_} or @samp{fftwl_} for single or long-double precision, respectively. (@code{fftw_complex} becomes @code{fftwf_complex}, @code{fftw_execute} becomes @code{fftwf_execute}, etcetera.) @item Uppercase names, i.e. names beginning with @samp{FFTW_}, remain the same. @item Replace @code{double} with @code{float} or @code{long double} for subroutine parameters. @end itemize Depending upon your compiler and/or hardware, @code{long double} may not be any more precise than @code{double} (or may not be supported at all, although it is standard in C99). @cindex C99 We also support using the nonstandard @code{__float128} quadruple-precision type provided by recent versions of @code{gcc} on 32- and 64-bit x86 hardware (@pxref{Installation and Customization}). To use this type, link with @code{-lfftw3q -lquadmath -lm} (the @code{libquadmath} library provided by @code{gcc} is needed for quadruple-precision trigonometric functions) and use @samp{fftwq_} identifiers. @c =========> @node Memory Allocation, , Precision, Data Types and Files @subsection Memory Allocation @example void *fftw_malloc(size_t n); void fftw_free(void *p); @end example @findex fftw_malloc @findex fftw_free These are functions that behave identically to @code{malloc} and @code{free}, except that they guarantee that the returned pointer obeys any special alignment restrictions imposed by any algorithm in FFTW (e.g. for SIMD acceleration). @xref{SIMD alignment and fftw_malloc}. @cindex alignment Data allocated by @code{fftw_malloc} @emph{must} be deallocated by @code{fftw_free} and not by the ordinary @code{free}. These routines simply call through to your operating system's @code{malloc} or, if necessary, its aligned equivalent (e.g. @code{memalign}), so you normally need not worry about any significant time or space overhead. You are @emph{not required} to use them to allocate your data, but we strongly recommend it. Note: in C++, just as with ordinary @code{malloc}, you must typecast the output of @code{fftw_malloc} to whatever pointer type you are allocating. @cindex C++ We also provide the following two convenience functions to allocate real and complex arrays with @code{n} elements, which are equivalent to @code{(double *) fftw_malloc(sizeof(double) * n)} and @code{(fftw_complex *) fftw_malloc(sizeof(fftw_complex) * n)}, respectively: @example double *fftw_alloc_real(size_t n); fftw_complex *fftw_alloc_complex(size_t n); @end example @findex fftw_alloc_real @findex fftw_alloc_complex The equivalent functions in other precisions allocate arrays of @code{n} elements in that precision. e.g. @code{fftwf_alloc_real(n)} is equivalent to @code{(float *) fftwf_malloc(sizeof(float) * n)}. @cindex precision @c ------------------------------------------------------------ @node Using Plans, Basic Interface, Data Types and Files, FFTW Reference @section Using Plans Plans for all transform types in FFTW are stored as type @code{fftw_plan} (an opaque pointer type), and are created by one of the various planning routines described in the following sections. @tindex fftw_plan An @code{fftw_plan} contains all information necessary to compute the transform, including the pointers to the input and output arrays. @example void fftw_execute(const fftw_plan plan); @end example @findex fftw_execute This executes the @code{plan}, to compute the corresponding transform on the arrays for which it was planned (which must still exist). The plan is not modified, and @code{fftw_execute} can be called as many times as desired. To apply a given plan to a different array, you can use the new-array execute interface. @xref{New-array Execute Functions}. @code{fftw_execute} (and equivalents) is the only function in FFTW guaranteed to be thread-safe; see @ref{Thread safety}. This function: @example void fftw_destroy_plan(fftw_plan plan); @end example @findex fftw_destroy_plan deallocates the @code{plan} and all its associated data. FFTW's planner saves some other persistent data, such as the accumulated wisdom and a list of algorithms available in the current configuration. If you want to deallocate all of that and reset FFTW to the pristine state it was in when you started your program, you can call: @example void fftw_cleanup(void); @end example @findex fftw_cleanup After calling @code{fftw_cleanup}, all existing plans become undefined, and you should not attempt to execute them nor to destroy them. You can however create and execute/destroy new plans, in which case FFTW starts accumulating wisdom information again. @code{fftw_cleanup} does not deallocate your plans, however. To prevent memory leaks, you must still call @code{fftw_destroy_plan} before executing @code{fftw_cleanup}. Occasionally, it may useful to know FFTW's internal ``cost'' metric that it uses to compare plans to one another; this cost is proportional to an execution time of the plan, in undocumented units, if the plan was created with the @code{FFTW_MEASURE} or other timing-based options, or alternatively is a heuristic cost function for @code{FFTW_ESTIMATE} plans. (The cost values of measured and estimated plans are not comparable, being in different units. Also, costs from different FFTW versions or the same version compiled differently may not be in the same units. Plans created from wisdom have a cost of 0 since no timing measurement is performed for them. Finally, certain problems for which only one top-level algorithm was possible may have required no measurements of the cost of the whole plan, in which case @code{fftw_cost} will also return 0.) The cost metric for a given plan is returned by: @example double fftw_cost(const fftw_plan plan); @end example @findex fftw_cost The following two routines are provided purely for academic purposes (that is, for entertainment). @example void fftw_flops(const fftw_plan plan, double *add, double *mul, double *fma); @end example @findex fftw_flops Given a @code{plan}, set @code{add}, @code{mul}, and @code{fma} to an exact count of the number of floating-point additions, multiplications, and fused multiply-add operations involved in the plan's execution. The total number of floating-point operations (flops) is @code{add + mul + 2*fma}, or @code{add + mul + fma} if the hardware supports fused multiply-add instructions (although the number of FMA operations is only approximate because of compiler voodoo). (The number of operations should be an integer, but we use @code{double} to avoid overflowing @code{int} for large transforms; the arguments are of type @code{double} even for single and long-double precision versions of FFTW.) @example void fftw_fprint_plan(const fftw_plan plan, FILE *output_file); void fftw_print_plan(const fftw_plan plan); @end example @findex fftw_fprint_plan @findex fftw_print_plan This outputs a ``nerd-readable'' representation of the @code{plan} to the given file or to @code{stdout}, respectively. @c ------------------------------------------------------------ @node Basic Interface, Advanced Interface, Using Plans, FFTW Reference @section Basic Interface @cindex basic interface Recall that the FFTW API is divided into three parts@footnote{@i{Gallia est omnis divisa in partes tres} (Julius Caesar).}: the @dfn{basic interface} computes a single transform of contiguous data, the @dfn{advanced interface} computes transforms of multiple or strided arrays, and the @dfn{guru interface} supports the most general data layouts, multiplicities, and strides. This section describes the the basic interface, which we expect to satisfy the needs of most users. @menu * Complex DFTs:: * Planner Flags:: * Real-data DFTs:: * Real-data DFT Array Format:: * Real-to-Real Transforms:: * Real-to-Real Transform Kinds:: @end menu @c =========> @node Complex DFTs, Planner Flags, Basic Interface, Basic Interface @subsection Complex DFTs @example fftw_plan fftw_plan_dft_1d(int n0, fftw_complex *in, fftw_complex *out, int sign, unsigned flags); fftw_plan fftw_plan_dft_2d(int n0, int n1, fftw_complex *in, fftw_complex *out, int sign, unsigned flags); fftw_plan fftw_plan_dft_3d(int n0, int n1, int n2, fftw_complex *in, fftw_complex *out, int sign, unsigned flags); fftw_plan fftw_plan_dft(int rank, const int *n, fftw_complex *in, fftw_complex *out, int sign, unsigned flags); @end example @findex fftw_plan_dft_1d @findex fftw_plan_dft_2d @findex fftw_plan_dft_3d @findex fftw_plan_dft Plan a complex input/output discrete Fourier transform (DFT) in zero or more dimensions, returning an @code{fftw_plan} (@pxref{Using Plans}). Once you have created a plan for a certain transform type and parameters, then creating another plan of the same type and parameters, but for different arrays, is fast and shares constant data with the first plan (if it still exists). The planner returns @code{NULL} if the plan cannot be created. In the standard FFTW distribution, the basic interface is guaranteed to return a non-@code{NULL} plan. A plan may be @code{NULL}, however, if you are using a customized FFTW configuration supporting a restricted set of transforms. @subsubheading Arguments @itemize @bullet @item @code{rank} is the rank of the transform (it should be the size of the array @code{*n}), and can be any non-negative integer. (@xref{Complex Multi-Dimensional DFTs}, for the definition of ``rank''.) The @samp{_1d}, @samp{_2d}, and @samp{_3d} planners correspond to a @code{rank} of @code{1}, @code{2}, and @code{3}, respectively. The rank may be zero, which is equivalent to a rank-1 transform of size 1, i.e. a copy of one number from input to output. @item @code{n0}, @code{n1}, @code{n2}, or @code{n[0..rank-1]} (as appropriate for each routine) specify the size of the transform dimensions. They can be any positive integer. @itemize @minus @item @cindex row-major Multi-dimensional arrays are stored in row-major order with dimensions: @code{n0} x @code{n1}; or @code{n0} x @code{n1} x @code{n2}; or @code{n[0]} x @code{n[1]} x ... x @code{n[rank-1]}. @xref{Multi-dimensional Array Format}. @item FFTW is best at handling sizes of the form @ifinfo @math{2^a 3^b 5^c 7^d 11^e 13^f}, @end ifinfo @tex $2^a 3^b 5^c 7^d 11^e 13^f$, @end tex @html 2a 3b 5c 7d 11e 13f, @end html where @math{e+f} is either @math{0} or @math{1}, and the other exponents are arbitrary. Other sizes are computed by means of a slow, general-purpose algorithm (which nevertheless retains @Onlogn{} performance even for prime sizes). It is possible to customize FFTW for different array sizes; see @ref{Installation and Customization}. Transforms whose sizes are powers of @math{2} are especially fast. @end itemize @item @code{in} and @code{out} point to the input and output arrays of the transform, which may be the same (yielding an in-place transform). @cindex in-place These arrays are overwritten during planning, unless @code{FFTW_ESTIMATE} is used in the flags. (The arrays need not be initialized, but they must be allocated.) If @code{in == out}, the transform is @dfn{in-place} and the input array is overwritten. If @code{in != out}, the two arrays must not overlap (but FFTW does not check for this condition). @item @ctindex FFTW_FORWARD @ctindex FFTW_BACKWARD @code{sign} is the sign of the exponent in the formula that defines the Fourier transform. It can be @math{-1} (= @code{FFTW_FORWARD}) or @math{+1} (= @code{FFTW_BACKWARD}). @item @cindex flags @code{flags} is a bitwise OR (@samp{|}) of zero or more planner flags, as defined in @ref{Planner Flags}. @end itemize FFTW computes an unnormalized transform: computing a forward followed by a backward transform (or vice versa) will result in the original data multiplied by the size of the transform (the product of the dimensions). @cindex normalization For more information, see @ref{What FFTW Really Computes}. @c =========> @node Planner Flags, Real-data DFTs, Complex DFTs, Basic Interface @subsection Planner Flags All of the planner routines in FFTW accept an integer @code{flags} argument, which is a bitwise OR (@samp{|}) of zero or more of the flag constants defined below. These flags control the rigor (and time) of the planning process, and can also impose (or lift) restrictions on the type of transform algorithm that is employed. @emph{Important:} the planner overwrites the input array during planning unless a saved plan (@pxref{Wisdom}) is available for that problem, so you should initialize your input data after creating the plan. The only exceptions to this are the @code{FFTW_ESTIMATE} and @code{FFTW_WISDOM_ONLY} flags, as mentioned below. In all cases, if wisdom is available for the given problem that was created with equal-or-greater planning rigor, then the more rigorous wisdom is used. For example, in @code{FFTW_ESTIMATE} mode any available wisdom is used, whereas in @code{FFTW_PATIENT} mode only wisdom created in patient or exhaustive mode can be used. @xref{Words of Wisdom-Saving Plans}. @subsubheading Planning-rigor flags @itemize @bullet @item @ctindex FFTW_ESTIMATE @code{FFTW_ESTIMATE} specifies that, instead of actual measurements of different algorithms, a simple heuristic is used to pick a (probably sub-optimal) plan quickly. With this flag, the input/output arrays are not overwritten during planning. @item @ctindex FFTW_MEASURE @code{FFTW_MEASURE} tells FFTW to find an optimized plan by actually @emph{computing} several FFTs and measuring their execution time. Depending on your machine, this can take some time (often a few seconds). @code{FFTW_MEASURE} is the default planning option. @item @ctindex FFTW_PATIENT @code{FFTW_PATIENT} is like @code{FFTW_MEASURE}, but considers a wider range of algorithms and often produces a ``more optimal'' plan (especially for large transforms), but at the expense of several times longer planning time (especially for large transforms). @item @ctindex FFTW_EXHAUSTIVE @code{FFTW_EXHAUSTIVE} is like @code{FFTW_PATIENT}, but considers an even wider range of algorithms, including many that we think are unlikely to be fast, to produce the most optimal plan but with a substantially increased planning time. @item @ctindex FFTW_WISDOM_ONLY @code{FFTW_WISDOM_ONLY} is a special planning mode in which the plan is only created if wisdom is available for the given problem, and otherwise a @code{NULL} plan is returned. This can be combined with other flags, e.g. @samp{FFTW_WISDOM_ONLY | FFTW_PATIENT} creates a plan only if wisdom is available that was created in @code{FFTW_PATIENT} or @code{FFTW_EXHAUSTIVE} mode. The @code{FFTW_WISDOM_ONLY} flag is intended for users who need to detect whether wisdom is available; for example, if wisdom is not available one may wish to allocate new arrays for planning so that user data is not overwritten. @end itemize @subsubheading Algorithm-restriction flags @itemize @bullet @item @ctindex FFTW_DESTROY_INPUT @code{FFTW_DESTROY_INPUT} specifies that an out-of-place transform is allowed to @emph{overwrite its input} array with arbitrary data; this can sometimes allow more efficient algorithms to be employed. @cindex out-of-place @item @ctindex FFTW_PRESERVE_INPUT @code{FFTW_PRESERVE_INPUT} specifies that an out-of-place transform must @emph{not change its input} array. This is ordinarily the @emph{default}, except for c2r and hc2r (i.e. complex-to-real) transforms for which @code{FFTW_DESTROY_INPUT} is the default. In the latter cases, passing @code{FFTW_PRESERVE_INPUT} will attempt to use algorithms that do not destroy the input, at the expense of worse performance; for multi-dimensional c2r transforms, however, no input-preserving algorithms are implemented and the planner will return @code{NULL} if one is requested. @cindex c2r @cindex hc2r @item @ctindex FFTW_UNALIGNED @cindex alignment @code{FFTW_UNALIGNED} specifies that the algorithm may not impose any unusual alignment requirements on the input/output arrays (i.e. no SIMD may be used). This flag is normally @emph{not necessary}, since the planner automatically detects misaligned arrays. The only use for this flag is if you want to use the new-array execute interface to execute a given plan on a different array that may not be aligned like the original. (Using @code{fftw_malloc} makes this flag unnecessary even then.) @end itemize @subsubheading Limiting planning time @example extern void fftw_set_timelimit(double seconds); @end example @findex fftw_set_timelimit This function instructs FFTW to spend at most @code{seconds} seconds (approximately) in the planner. If @code{seconds == FFTW_NO_TIMELIMIT} (the default value, which is negative), then planning time is unbounded. Otherwise, FFTW plans with a progressively wider range of algorithms until the the given time limit is reached or the given range of algorithms is explored, returning the best available plan. @ctindex FFTW_NO_TIMELIMIT For example, specifying @code{FFTW_PATIENT} first plans in @code{FFTW_ESTIMATE} mode, then in @code{FFTW_MEASURE} mode, then finally (time permitting) in @code{FFTW_PATIENT}. If @code{FFTW_EXHAUSTIVE} is specified instead, the planner will further progress to @code{FFTW_EXHAUSTIVE} mode. Note that the @code{seconds} argument specifies only a rough limit; in practice, the planner may use somewhat more time if the time limit is reached when the planner is in the middle of an operation that cannot be interrupted. At the very least, the planner will complete planning in @code{FFTW_ESTIMATE} mode (which is thus equivalent to a time limit of 0). @c =========> @node Real-data DFTs, Real-data DFT Array Format, Planner Flags, Basic Interface @subsection Real-data DFTs @example fftw_plan fftw_plan_dft_r2c_1d(int n0, double *in, fftw_complex *out, unsigned flags); fftw_plan fftw_plan_dft_r2c_2d(int n0, int n1, double *in, fftw_complex *out, unsigned flags); fftw_plan fftw_plan_dft_r2c_3d(int n0, int n1, int n2, double *in, fftw_complex *out, unsigned flags); fftw_plan fftw_plan_dft_r2c(int rank, const int *n, double *in, fftw_complex *out, unsigned flags); @end example @findex fftw_plan_dft_r2c_1d @findex fftw_plan_dft_r2c_2d @findex fftw_plan_dft_r2c_3d @findex fftw_plan_dft_r2c @cindex r2c Plan a real-input/complex-output discrete Fourier transform (DFT) in zero or more dimensions, returning an @code{fftw_plan} (@pxref{Using Plans}). Once you have created a plan for a certain transform type and parameters, then creating another plan of the same type and parameters, but for different arrays, is fast and shares constant data with the first plan (if it still exists). The planner returns @code{NULL} if the plan cannot be created. A non-@code{NULL} plan is always returned by the basic interface unless you are using a customized FFTW configuration supporting a restricted set of transforms, or if you use the @code{FFTW_PRESERVE_INPUT} flag with a multi-dimensional out-of-place c2r transform (see below). @subsubheading Arguments @itemize @bullet @item @code{rank} is the rank of the transform (it should be the size of the array @code{*n}), and can be any non-negative integer. (@xref{Complex Multi-Dimensional DFTs}, for the definition of ``rank''.) The @samp{_1d}, @samp{_2d}, and @samp{_3d} planners correspond to a @code{rank} of @code{1}, @code{2}, and @code{3}, respectively. The rank may be zero, which is equivalent to a rank-1 transform of size 1, i.e. a copy of one real number (with zero imaginary part) from input to output. @item @code{n0}, @code{n1}, @code{n2}, or @code{n[0..rank-1]}, (as appropriate for each routine) specify the size of the transform dimensions. They can be any positive integer. This is different in general from the @emph{physical} array dimensions, which are described in @ref{Real-data DFT Array Format}. @itemize @minus @item FFTW is best at handling sizes of the form @ifinfo @math{2^a 3^b 5^c 7^d 11^e 13^f}, @end ifinfo @tex $2^a 3^b 5^c 7^d 11^e 13^f$, @end tex @html 2a 3b 5c 7d 11e 13f, @end html where @math{e+f} is either @math{0} or @math{1}, and the other exponents are arbitrary. Other sizes are computed by means of a slow, general-purpose algorithm (which nevertheless retains @Onlogn{} performance even for prime sizes). (It is possible to customize FFTW for different array sizes; see @ref{Installation and Customization}.) Transforms whose sizes are powers of @math{2} are especially fast, and it is generally beneficial for the @emph{last} dimension of an r2c/c2r transform to be @emph{even}. @end itemize @item @code{in} and @code{out} point to the input and output arrays of the transform, which may be the same (yielding an in-place transform). @cindex in-place These arrays are overwritten during planning, unless @code{FFTW_ESTIMATE} is used in the flags. (The arrays need not be initialized, but they must be allocated.) For an in-place transform, it is important to remember that the real array will require padding, described in @ref{Real-data DFT Array Format}. @cindex padding @item @cindex flags @code{flags} is a bitwise OR (@samp{|}) of zero or more planner flags, as defined in @ref{Planner Flags}. @end itemize The inverse transforms, taking complex input (storing the non-redundant half of a logically Hermitian array) to real output, are given by: @example fftw_plan fftw_plan_dft_c2r_1d(int n0, fftw_complex *in, double *out, unsigned flags); fftw_plan fftw_plan_dft_c2r_2d(int n0, int n1, fftw_complex *in, double *out, unsigned flags); fftw_plan fftw_plan_dft_c2r_3d(int n0, int n1, int n2, fftw_complex *in, double *out, unsigned flags); fftw_plan fftw_plan_dft_c2r(int rank, const int *n, fftw_complex *in, double *out, unsigned flags); @end example @findex fftw_plan_dft_c2r_1d @findex fftw_plan_dft_c2r_2d @findex fftw_plan_dft_c2r_3d @findex fftw_plan_dft_c2r @cindex c2r The arguments are the same as for the r2c transforms, except that the input and output data formats are reversed. FFTW computes an unnormalized transform: computing an r2c followed by a c2r transform (or vice versa) will result in the original data multiplied by the size of the transform (the product of the logical dimensions). @cindex normalization An r2c transform produces the same output as a @code{FFTW_FORWARD} complex DFT of the same input, and a c2r transform is correspondingly equivalent to @code{FFTW_BACKWARD}. For more information, see @ref{What FFTW Really Computes}. @c =========> @node Real-data DFT Array Format, Real-to-Real Transforms, Real-data DFTs, Basic Interface @subsection Real-data DFT Array Format @cindex r2c/c2r multi-dimensional array format The output of a DFT of real data (r2c) contains symmetries that, in principle, make half of the outputs redundant (@pxref{What FFTW Really Computes}). (Similarly for the input of an inverse c2r transform.) In practice, it is not possible to entirely realize these savings in an efficient and understandable format that generalizes to multi-dimensional transforms. Instead, the output of the r2c transforms is @emph{slightly} over half of the output of the corresponding complex transform. We do not ``pack'' the data in any way, but store it as an ordinary array of @code{fftw_complex} values. In fact, this data is simply a subsection of what would be the array in the corresponding complex transform. Specifically, for a real transform of @math{d} (= @code{rank}) dimensions @ndims{}, the complex data is an @ndimshalf array of @code{fftw_complex} values in row-major order (with the division rounded down). That is, we only store the @emph{lower} half (non-negative frequencies), plus one element, of the last dimension of the data from the ordinary complex transform. (We could have instead taken half of any other dimension, but implementation turns out to be simpler if the last, contiguous, dimension is used.) @cindex out-of-place For an out-of-place transform, the real data is simply an array with physical dimensions @ndims in row-major order. @cindex in-place @cindex padding For an in-place transform, some complications arise since the complex data is slightly larger than the real data. In this case, the final dimension of the real data must be @emph{padded} with extra values to accommodate the size of the complex data---two extra if the last dimension is even and one if it is odd. That is, the last dimension of the real data must physically contain @tex $2 (n_{d-1}/2+1)$ @end tex @ifinfo 2 * (n[d-1]/2+1) @end ifinfo @html 2 * (nd-1/2+1) @end html @code{double} values (exactly enough to hold the complex data). This physical array size does not, however, change the @emph{logical} array size---only @tex $n_{d-1}$ @end tex @ifinfo n[d-1] @end ifinfo @html nd-1 @end html values are actually stored in the last dimension, and @tex $n_{d-1}$ @end tex @ifinfo n[d-1] @end ifinfo @html nd-1 @end html is the last dimension passed to the planner. @c =========> @node Real-to-Real Transforms, Real-to-Real Transform Kinds, Real-data DFT Array Format, Basic Interface @subsection Real-to-Real Transforms @cindex r2r @example fftw_plan fftw_plan_r2r_1d(int n, double *in, double *out, fftw_r2r_kind kind, unsigned flags); fftw_plan fftw_plan_r2r_2d(int n0, int n1, double *in, double *out, fftw_r2r_kind kind0, fftw_r2r_kind kind1, unsigned flags); fftw_plan fftw_plan_r2r_3d(int n0, int n1, int n2, double *in, double *out, fftw_r2r_kind kind0, fftw_r2r_kind kind1, fftw_r2r_kind kind2, unsigned flags); fftw_plan fftw_plan_r2r(int rank, const int *n, double *in, double *out, const fftw_r2r_kind *kind, unsigned flags); @end example @findex fftw_plan_r2r_1d @findex fftw_plan_r2r_2d @findex fftw_plan_r2r_3d @findex fftw_plan_r2r Plan a real input/output (r2r) transform of various kinds in zero or more dimensions, returning an @code{fftw_plan} (@pxref{Using Plans}). Once you have created a plan for a certain transform type and parameters, then creating another plan of the same type and parameters, but for different arrays, is fast and shares constant data with the first plan (if it still exists). The planner returns @code{NULL} if the plan cannot be created. A non-@code{NULL} plan is always returned by the basic interface unless you are using a customized FFTW configuration supporting a restricted set of transforms, or for size-1 @code{FFTW_REDFT00} kinds (which are not defined). @ctindex FFTW_REDFT00 @subsubheading Arguments @itemize @bullet @item @code{rank} is the dimensionality of the transform (it should be the size of the arrays @code{*n} and @code{*kind}), and can be any non-negative integer. The @samp{_1d}, @samp{_2d}, and @samp{_3d} planners correspond to a @code{rank} of @code{1}, @code{2}, and @code{3}, respectively. A @code{rank} of zero is equivalent to a copy of one number from input to output. @item @code{n}, or @code{n0}/@code{n1}/@code{n2}, or @code{n[rank]}, respectively, gives the (physical) size of the transform dimensions. They can be any positive integer. @itemize @minus @item @cindex row-major Multi-dimensional arrays are stored in row-major order with dimensions: @code{n0} x @code{n1}; or @code{n0} x @code{n1} x @code{n2}; or @code{n[0]} x @code{n[1]} x ... x @code{n[rank-1]}. @xref{Multi-dimensional Array Format}. @item FFTW is generally best at handling sizes of the form @ifinfo @math{2^a 3^b 5^c 7^d 11^e 13^f}, @end ifinfo @tex $2^a 3^b 5^c 7^d 11^e 13^f$, @end tex @html 2a 3b 5c 7d 11e 13f, @end html where @math{e+f} is either @math{0} or @math{1}, and the other exponents are arbitrary. Other sizes are computed by means of a slow, general-purpose algorithm (which nevertheless retains @Onlogn{} performance even for prime sizes). (It is possible to customize FFTW for different array sizes; see @ref{Installation and Customization}.) Transforms whose sizes are powers of @math{2} are especially fast. @item For a @code{REDFT00} or @code{RODFT00} transform kind in a dimension of size @math{n}, it is @math{n-1} or @math{n+1}, respectively, that should be factorizable in the above form. @end itemize @item @code{in} and @code{out} point to the input and output arrays of the transform, which may be the same (yielding an in-place transform). @cindex in-place These arrays are overwritten during planning, unless @code{FFTW_ESTIMATE} is used in the flags. (The arrays need not be initialized, but they must be allocated.) @item @code{kind}, or @code{kind0}/@code{kind1}/@code{kind2}, or @code{kind[rank]}, is the kind of r2r transform used for the corresponding dimension. The valid kind constants are described in @ref{Real-to-Real Transform Kinds}. In a multi-dimensional transform, what is computed is the separable product formed by taking each transform kind along the corresponding dimension, one dimension after another. @item @cindex flags @code{flags} is a bitwise OR (@samp{|}) of zero or more planner flags, as defined in @ref{Planner Flags}. @end itemize @c =========> @node Real-to-Real Transform Kinds, , Real-to-Real Transforms, Basic Interface @subsection Real-to-Real Transform Kinds @cindex kind (r2r) FFTW currently supports 11 different r2r transform kinds, specified by one of the constants below. For the precise definitions of these transforms, see @ref{What FFTW Really Computes}. For a more colloquial introduction to these transform kinds, see @ref{More DFTs of Real Data}. For dimension of size @code{n}, there is a corresponding ``logical'' dimension @code{N} that determines the normalization (and the optimal factorization); the formula for @code{N} is given for each kind below. Also, with each transform kind is listed its corrsponding inverse transform. FFTW computes unnormalized transforms: a transform followed by its inverse will result in the original data multiplied by @code{N} (or the product of the @code{N}'s for each dimension, in multi-dimensions). @cindex normalization @itemize @bullet @item @ctindex FFTW_R2HC @code{FFTW_R2HC} computes a real-input DFT with output in ``halfcomplex'' format, i.e. real and imaginary parts for a transform of size @code{n} stored as: @tex $$ r_0, r_1, r_2, \ldots, r_{n/2}, i_{(n+1)/2-1}, \ldots, i_2, i_1 $$ @end tex @ifinfo r0, r1, r2, r(n/2), i((n+1)/2-1), ..., i2, i1 @end ifinfo @html

r0, r1, r2, ..., rn/2, i(n+1)/2-1, ..., i2, i1

@end html (Logical @code{N=n}, inverse is @code{FFTW_HC2R}.) @item @ctindex FFTW_HC2R @code{FFTW_HC2R} computes the reverse of @code{FFTW_R2HC}, above. (Logical @code{N=n}, inverse is @code{FFTW_R2HC}.) @item @ctindex FFTW_DHT @code{FFTW_DHT} computes a discrete Hartley transform. (Logical @code{N=n}, inverse is @code{FFTW_DHT}.) @cindex discrete Hartley transform @item @ctindex FFTW_REDFT00 @code{FFTW_REDFT00} computes an REDFT00 transform, i.e. a DCT-I. (Logical @code{N=2*(n-1)}, inverse is @code{FFTW_REDFT00}.) @cindex discrete cosine transform @cindex DCT @item @ctindex FFTW_REDFT10 @code{FFTW_REDFT10} computes an REDFT10 transform, i.e. a DCT-II (sometimes called ``the'' DCT). (Logical @code{N=2*n}, inverse is @code{FFTW_REDFT01}.) @item @ctindex FFTW_REDFT01 @code{FFTW_REDFT01} computes an REDFT01 transform, i.e. a DCT-III (sometimes called ``the'' IDCT, being the inverse of DCT-II). (Logical @code{N=2*n}, inverse is @code{FFTW_REDFT=10}.) @cindex IDCT @item @ctindex FFTW_REDFT11 @code{FFTW_REDFT11} computes an REDFT11 transform, i.e. a DCT-IV. (Logical @code{N=2*n}, inverse is @code{FFTW_REDFT11}.) @item @ctindex FFTW_RODFT00 @code{FFTW_RODFT00} computes an RODFT00 transform, i.e. a DST-I. (Logical @code{N=2*(n+1)}, inverse is @code{FFTW_RODFT00}.) @cindex discrete sine transform @cindex DST @item @ctindex FFTW_RODFT10 @code{FFTW_RODFT10} computes an RODFT10 transform, i.e. a DST-II. (Logical @code{N=2*n}, inverse is @code{FFTW_RODFT01}.) @item @ctindex FFTW_RODFT01 @code{FFTW_RODFT01} computes an RODFT01 transform, i.e. a DST-III. (Logical @code{N=2*n}, inverse is @code{FFTW_RODFT=10}.) @item @ctindex FFTW_RODFT11 @code{FFTW_RODFT11} computes an RODFT11 transform, i.e. a DST-IV. (Logical @code{N=2*n}, inverse is @code{FFTW_RODFT11}.) @end itemize @c ------------------------------------------------------------ @node Advanced Interface, Guru Interface, Basic Interface, FFTW Reference @section Advanced Interface @cindex advanced interface FFTW's ``advanced'' interface supplements the basic interface with four new planner routines, providing a new level of flexibility: you can plan a transform of multiple arrays simultaneously, operate on non-contiguous (strided) data, and transform a subset of a larger multi-dimensional array. Other than these additional features, the planner operates in the same fashion as in the basic interface, and the resulting @code{fftw_plan} is used in the same way (@pxref{Using Plans}). @menu * Advanced Complex DFTs:: * Advanced Real-data DFTs:: * Advanced Real-to-real Transforms:: @end menu @c =========> @node Advanced Complex DFTs, Advanced Real-data DFTs, Advanced Interface, Advanced Interface @subsection Advanced Complex DFTs @example fftw_plan fftw_plan_many_dft(int rank, const int *n, int howmany, fftw_complex *in, const int *inembed, int istride, int idist, fftw_complex *out, const int *onembed, int ostride, int odist, int sign, unsigned flags); @end example @findex fftw_plan_many_dft This routine plans multiple multidimensional complex DFTs, and it extends the @code{fftw_plan_dft} routine (@pxref{Complex DFTs}) to compute @code{howmany} transforms, each having rank @code{rank} and size @code{n}. In addition, the transform data need not be contiguous, but it may be laid out in memory with an arbitrary stride. To account for these possibilities, @code{fftw_plan_many_dft} adds the new parameters @code{howmany}, @{@code{i},@code{o}@}@code{nembed}, @{@code{i},@code{o}@}@code{stride}, and @{@code{i},@code{o}@}@code{dist}. The FFTW basic interface (@pxref{Complex DFTs}) provides routines specialized for ranks 1, 2, and@tie{}3, but the advanced interface handles only the general-rank case. @code{howmany} is the number of transforms to compute. The resulting plan computes @code{howmany} transforms, where the input of the @code{k}-th transform is at location @code{in+k*idist} (in C pointer arithmetic), and its output is at location @code{out+k*odist}. Plans obtained in this way can often be faster than calling FFTW multiple times for the individual transforms. The basic @code{fftw_plan_dft} interface corresponds to @code{howmany=1} (in which case the @code{dist} parameters are ignored). @cindex howmany parameter @cindex dist Each of the @code{howmany} transforms has rank @code{rank} and size @code{n}, as in the basic interface. In addition, the advanced interface allows the input and output arrays of each transform to be row-major subarrays of larger rank-@code{rank} arrays, described by @code{inembed} and @code{onembed} parameters, respectively. @{@code{i},@code{o}@}@code{nembed} must be arrays of length @code{rank}, and @code{n} should be elementwise less than or equal to @{@code{i},@code{o}@}@code{nembed}. Passing @code{NULL} for an @code{nembed} parameter is equivalent to passing @code{n} (i.e. same physical and logical dimensions, as in the basic interface.) The @code{stride} parameters indicate that the @code{j}-th element of the input or output arrays is located at @code{j*istride} or @code{j*ostride}, respectively. (For a multi-dimensional array, @code{j} is the ordinary row-major index.) When combined with the @code{k}-th transform in a @code{howmany} loop, from above, this means that the (@code{j},@code{k})-th element is at @code{j*stride+k*dist}. (The basic @code{fftw_plan_dft} interface corresponds to a stride of 1.) @cindex stride For in-place transforms, the input and output @code{stride} and @code{dist} parameters should be the same; otherwise, the planner may return @code{NULL}. Arrays @code{n}, @code{inembed}, and @code{onembed} are not used after this function returns. You can safely free or reuse them. @strong{Examples}: One transform of one 5 by 6 array contiguous in memory: @example int rank = 2; int n[] = @{5, 6@}; int howmany = 1; int idist = odist = 0; /* unused because howmany = 1 */ int istride = ostride = 1; /* array is contiguous in memory */ int *inembed = n, *onembed = n; @end example Transform of three 5 by 6 arrays, each contiguous in memory, stored in memory one after another: @example int rank = 2; int n[] = @{5, 6@}; int howmany = 3; int idist = odist = n[0]*n[1]; /* = 30, the distance in memory between the first element of the first array and the first element of the second array */ int istride = ostride = 1; /* array is contiguous in memory */ int *inembed = n, *onembed = n; @end example Transform each column of a 2d array with 10 rows and 3 columns: @example int rank = 1; /* not 2: we are computing 1d transforms */ int n[] = @{10@}; /* 1d transforms of length 10 */ int howmany = 3; int idist = odist = 1; int istride = ostride = 3; /* distance between two elements in the same column */ int *inembed = n, *onembed = n; @end example @c =========> @node Advanced Real-data DFTs, Advanced Real-to-real Transforms, Advanced Complex DFTs, Advanced Interface @subsection Advanced Real-data DFTs @example fftw_plan fftw_plan_many_dft_r2c(int rank, const int *n, int howmany, double *in, const int *inembed, int istride, int idist, fftw_complex *out, const int *onembed, int ostride, int odist, unsigned flags); fftw_plan fftw_plan_many_dft_c2r(int rank, const int *n, int howmany, fftw_complex *in, const int *inembed, int istride, int idist, double *out, const int *onembed, int ostride, int odist, unsigned flags); @end example @findex fftw_plan_many_dft_r2c @findex fftw_plan_many_dft_c2r Like @code{fftw_plan_many_dft}, these two functions add @code{howmany}, @code{nembed}, @code{stride}, and @code{dist} parameters to the @code{fftw_plan_dft_r2c} and @code{fftw_plan_dft_c2r} functions, but otherwise behave the same as the basic interface. The interpretation of @code{howmany}, @code{stride}, and @code{dist} are the same as for @code{fftw_plan_many_dft}, above. Note that the @code{stride} and @code{dist} for the real array are in units of @code{double}, and for the complex array are in units of @code{fftw_complex}. If an @code{nembed} parameter is @code{NULL}, it is interpreted as what it would be in the basic interface, as described in @ref{Real-data DFT Array Format}. That is, for the complex array the size is assumed to be the same as @code{n}, but with the last dimension cut roughly in half. For the real array, the size is assumed to be @code{n} if the transform is out-of-place, or @code{n} with the last dimension ``padded'' if the transform is in-place. If an @code{nembed} parameter is non-@code{NULL}, it is interpreted as the physical size of the corresponding array, in row-major order, just as for @code{fftw_plan_many_dft}. In this case, each dimension of @code{nembed} should be @code{>=} what it would be in the basic interface (e.g. the halved or padded @code{n}). Arrays @code{n}, @code{inembed}, and @code{onembed} are not used after this function returns. You can safely free or reuse them. @c =========> @node Advanced Real-to-real Transforms, , Advanced Real-data DFTs, Advanced Interface @subsection Advanced Real-to-real Transforms @example fftw_plan fftw_plan_many_r2r(int rank, const int *n, int howmany, double *in, const int *inembed, int istride, int idist, double *out, const int *onembed, int ostride, int odist, const fftw_r2r_kind *kind, unsigned flags); @end example @findex fftw_plan_many_r2r Like @code{fftw_plan_many_dft}, this functions adds @code{howmany}, @code{nembed}, @code{stride}, and @code{dist} parameters to the @code{fftw_plan_r2r} function, but otherwise behave the same as the basic interface. The interpretation of those additional parameters are the same as for @code{fftw_plan_many_dft}. (Of course, the @code{stride} and @code{dist} parameters are now in units of @code{double}, not @code{fftw_complex}.) Arrays @code{n}, @code{inembed}, @code{onembed}, and @code{kind} are not used after this function returns. You can safely free or reuse them. @c ------------------------------------------------------------ @node Guru Interface, New-array Execute Functions, Advanced Interface, FFTW Reference @section Guru Interface @cindex guru interface The ``guru'' interface to FFTW is intended to expose as much as possible of the flexibility in the underlying FFTW architecture. It allows one to compute multi-dimensional ``vectors'' (loops) of multi-dimensional transforms, where each vector/transform dimension has an independent size and stride. @cindex vector One can also use more general complex-number formats, e.g. separate real and imaginary arrays. For those users who require the flexibility of the guru interface, it is important that they pay special attention to the documentation lest they shoot themselves in the foot. @menu * Interleaved and split arrays:: * Guru vector and transform sizes:: * Guru Complex DFTs:: * Guru Real-data DFTs:: * Guru Real-to-real Transforms:: * 64-bit Guru Interface:: @end menu @c =========> @node Interleaved and split arrays, Guru vector and transform sizes, Guru Interface, Guru Interface @subsection Interleaved and split arrays The guru interface supports two representations of complex numbers, which we call the interleaved and the split format. The @dfn{interleaved} format is the same one used by the basic and advanced interfaces, and it is documented in @ref{Complex numbers}. In the interleaved format, you provide pointers to the real part of a complex number, and the imaginary part understood to be stored in the next memory location. @cindex interleaved format The @dfn{split} format allows separate pointers to the real and imaginary parts of a complex array. @cindex split format Technically, the interleaved format is redundant, because you can always express an interleaved array in terms of a split array with appropriate pointers and strides. On the other hand, the interleaved format is simpler to use, and it is common in practice. Hence, FFTW supports it as a special case. @c =========> @node Guru vector and transform sizes, Guru Complex DFTs, Interleaved and split arrays, Guru Interface @subsection Guru vector and transform sizes The guru interface introduces one basic new data structure, @code{fftw_iodim}, that is used to specify sizes and strides for multi-dimensional transforms and vectors: @example typedef struct @{ int n; int is; int os; @} fftw_iodim; @end example @tindex fftw_iodim Here, @code{n} is the size of the dimension, and @code{is} and @code{os} are the strides of that dimension for the input and output arrays. (The stride is the separation of consecutive elements along this dimension.) The meaning of the stride parameter depends on the type of the array that the stride refers to. @emph{If the array is interleaved complex, strides are expressed in units of complex numbers (@code{fftw_complex}). If the array is split complex or real, strides are expressed in units of real numbers (@code{double}).} This convention is consistent with the usual pointer arithmetic in the C language. An interleaved array is denoted by a pointer @code{p} to @code{fftw_complex}, so that @code{p+1} points to the next complex number. Split arrays are denoted by pointers to @code{double}, in which case pointer arithmetic operates in units of @code{sizeof(double)}. @cindex stride The guru planner interfaces all take a (@code{rank}, @code{dims[rank]}) pair describing the transform size, and a (@code{howmany_rank}, @code{howmany_dims[howmany_rank]}) pair describing the ``vector'' size (a multi-dimensional loop of transforms to perform), where @code{dims} and @code{howmany_dims} are arrays of @code{fftw_iodim}. For example, the @code{howmany} parameter in the advanced complex-DFT interface corresponds to @code{howmany_rank} = 1, @code{howmany_dims[0].n} = @code{howmany}, @code{howmany_dims[0].is} = @code{idist}, and @code{howmany_dims[0].os} = @code{odist}. @cindex howmany loop @cindex dist (To compute a single transform, you can just use @code{howmany_rank} = 0.) A row-major multidimensional array with dimensions @code{n[rank]} (@pxref{Row-major Format}) corresponds to @code{dims[i].n} = @code{n[i]} and the recurrence @code{dims[i].is} = @code{n[i+1] * dims[i+1].is} (similarly for @code{os}). The stride of the last (@code{i=rank-1}) dimension is the overall stride of the array. e.g. to be equivalent to the advanced complex-DFT interface, you would have @code{dims[rank-1].is} = @code{istride} and @code{dims[rank-1].os} = @code{ostride}. @cindex row-major In general, we only guarantee FFTW to return a non-@code{NULL} plan if the vector and transform dimensions correspond to a set of distinct indices, and for in-place transforms the input/output strides should be the same. @c =========> @node Guru Complex DFTs, Guru Real-data DFTs, Guru vector and transform sizes, Guru Interface @subsection Guru Complex DFTs @example fftw_plan fftw_plan_guru_dft( int rank, const fftw_iodim *dims, int howmany_rank, const fftw_iodim *howmany_dims, fftw_complex *in, fftw_complex *out, int sign, unsigned flags); fftw_plan fftw_plan_guru_split_dft( int rank, const fftw_iodim *dims, int howmany_rank, const fftw_iodim *howmany_dims, double *ri, double *ii, double *ro, double *io, unsigned flags); @end example @findex fftw_plan_guru_dft @findex fftw_plan_guru_split_dft These two functions plan a complex-data, multi-dimensional DFT for the interleaved and split format, respectively. Transform dimensions are given by (@code{rank}, @code{dims}) over a multi-dimensional vector (loop) of dimensions (@code{howmany_rank}, @code{howmany_dims}). @code{dims} and @code{howmany_dims} should point to @code{fftw_iodim} arrays of length @code{rank} and @code{howmany_rank}, respectively. @cindex flags @code{flags} is a bitwise OR (@samp{|}) of zero or more planner flags, as defined in @ref{Planner Flags}. In the @code{fftw_plan_guru_dft} function, the pointers @code{in} and @code{out} point to the interleaved input and output arrays, respectively. The sign can be either @math{-1} (= @code{FFTW_FORWARD}) or @math{+1} (= @code{FFTW_BACKWARD}). If the pointers are equal, the transform is in-place. In the @code{fftw_plan_guru_split_dft} function, @code{ri} and @code{ii} point to the real and imaginary input arrays, and @code{ro} and @code{io} point to the real and imaginary output arrays. The input and output pointers may be the same, indicating an in-place transform. For example, for @code{fftw_complex} pointers @code{in} and @code{out}, the corresponding parameters are: @example ri = (double *) in; ii = (double *) in + 1; ro = (double *) out; io = (double *) out + 1; @end example Because @code{fftw_plan_guru_split_dft} accepts split arrays, strides are expressed in units of @code{double}. For a contiguous @code{fftw_complex} array, the overall stride of the transform should be 2, the distance between consecutive real parts or between consecutive imaginary parts; see @ref{Guru vector and transform sizes}. Note that the dimension strides are applied equally to the real and imaginary parts; real and imaginary arrays with different strides are not supported. There is no @code{sign} parameter in @code{fftw_plan_guru_split_dft}. This function always plans for an @code{FFTW_FORWARD} transform. To plan for an @code{FFTW_BACKWARD} transform, you can exploit the identity that the backwards DFT is equal to the forwards DFT with the real and imaginary parts swapped. For example, in the case of the @code{fftw_complex} arrays above, the @code{FFTW_BACKWARD} transform is computed by the parameters: @example ri = (double *) in + 1; ii = (double *) in; ro = (double *) out + 1; io = (double *) out; @end example @c =========> @node Guru Real-data DFTs, Guru Real-to-real Transforms, Guru Complex DFTs, Guru Interface @subsection Guru Real-data DFTs @example fftw_plan fftw_plan_guru_dft_r2c( int rank, const fftw_iodim *dims, int howmany_rank, const fftw_iodim *howmany_dims, double *in, fftw_complex *out, unsigned flags); fftw_plan fftw_plan_guru_split_dft_r2c( int rank, const fftw_iodim *dims, int howmany_rank, const fftw_iodim *howmany_dims, double *in, double *ro, double *io, unsigned flags); fftw_plan fftw_plan_guru_dft_c2r( int rank, const fftw_iodim *dims, int howmany_rank, const fftw_iodim *howmany_dims, fftw_complex *in, double *out, unsigned flags); fftw_plan fftw_plan_guru_split_dft_c2r( int rank, const fftw_iodim *dims, int howmany_rank, const fftw_iodim *howmany_dims, double *ri, double *ii, double *out, unsigned flags); @end example @findex fftw_plan_guru_dft_r2c @findex fftw_plan_guru_split_dft_r2c @findex fftw_plan_guru_dft_c2r @findex fftw_plan_guru_split_dft_c2r Plan a real-input (r2c) or real-output (c2r), multi-dimensional DFT with transform dimensions given by (@code{rank}, @code{dims}) over a multi-dimensional vector (loop) of dimensions (@code{howmany_rank}, @code{howmany_dims}). @code{dims} and @code{howmany_dims} should point to @code{fftw_iodim} arrays of length @code{rank} and @code{howmany_rank}, respectively. As for the basic and advanced interfaces, an r2c transform is @code{FFTW_FORWARD} and a c2r transform is @code{FFTW_BACKWARD}. The @emph{last} dimension of @code{dims} is interpreted specially: that dimension of the real array has size @code{dims[rank-1].n}, but that dimension of the complex array has size @code{dims[rank-1].n/2+1} (division rounded down). The strides, on the other hand, are taken to be exactly as specified. It is up to the user to specify the strides appropriately for the peculiar dimensions of the data, and we do not guarantee that the planner will succeed (return non-@code{NULL}) for any dimensions other than those described in @ref{Real-data DFT Array Format} and generalized in @ref{Advanced Real-data DFTs}. (That is, for an in-place transform, each individual dimension should be able to operate in place.) @cindex in-place @code{in} and @code{out} point to the input and output arrays for r2c and c2r transforms, respectively. For split arrays, @code{ri} and @code{ii} point to the real and imaginary input arrays for a c2r transform, and @code{ro} and @code{io} point to the real and imaginary output arrays for an r2c transform. @code{in} and @code{ro} or @code{ri} and @code{out} may be the same, indicating an in-place transform. (In-place transforms where @code{in} and @code{io} or @code{ii} and @code{out} are the same are not currently supported.) @cindex flags @code{flags} is a bitwise OR (@samp{|}) of zero or more planner flags, as defined in @ref{Planner Flags}. In-place transforms of rank greater than 1 are currently only supported for interleaved arrays. For split arrays, the planner will return @code{NULL}. @cindex in-place @c =========> @node Guru Real-to-real Transforms, 64-bit Guru Interface, Guru Real-data DFTs, Guru Interface @subsection Guru Real-to-real Transforms @example fftw_plan fftw_plan_guru_r2r(int rank, const fftw_iodim *dims, int howmany_rank, const fftw_iodim *howmany_dims, double *in, double *out, const fftw_r2r_kind *kind, unsigned flags); @end example @findex fftw_plan_guru_r2r Plan a real-to-real (r2r) multi-dimensional @code{FFTW_FORWARD} transform with transform dimensions given by (@code{rank}, @code{dims}) over a multi-dimensional vector (loop) of dimensions (@code{howmany_rank}, @code{howmany_dims}). @code{dims} and @code{howmany_dims} should point to @code{fftw_iodim} arrays of length @code{rank} and @code{howmany_rank}, respectively. The transform kind of each dimension is given by the @code{kind} parameter, which should point to an array of length @code{rank}. Valid @code{fftw_r2r_kind} constants are given in @ref{Real-to-Real Transform Kinds}. @code{in} and @code{out} point to the real input and output arrays; they may be the same, indicating an in-place transform. @cindex flags @code{flags} is a bitwise OR (@samp{|}) of zero or more planner flags, as defined in @ref{Planner Flags}. @c =========> @node 64-bit Guru Interface, , Guru Real-to-real Transforms, Guru Interface @subsection 64-bit Guru Interface @cindex 64-bit architecture When compiled in 64-bit mode on a 64-bit architecture (where addresses are 64 bits wide), FFTW uses 64-bit quantities internally for all transform sizes, strides, and so on---you don't have to do anything special to exploit this. However, in the ordinary FFTW interfaces, you specify the transform size by an @code{int} quantity, which is normally only 32 bits wide. This means that, even though FFTW is using 64-bit sizes internally, you cannot specify a single transform dimension larger than @ifinfo 2^31-1 @end ifinfo @html 231−1 @end html @tex $2^31-1$ @end tex numbers. We expect that few users will require transforms larger than this, but, for those who do, we provide a 64-bit version of the guru interface in which all sizes are specified as integers of type @code{ptrdiff_t} instead of @code{int}. (@code{ptrdiff_t} is a signed integer type defined by the C standard to be wide enough to represent address differences, and thus must be at least 64 bits wide on a 64-bit machine.) We stress that there is @emph{no performance advantage} to using this interface---the same internal FFTW code is employed regardless---and it is only necessary if you want to specify very large transform sizes. @tindex ptrdiff_t In particular, the 64-bit guru interface is a set of planner routines that are exactly the same as the guru planner routines, except that they are named with @samp{guru64} instead of @samp{guru} and they take arguments of type @code{fftw_iodim64} instead of @code{fftw_iodim}. For example, instead of @code{fftw_plan_guru_dft}, we have @code{fftw_plan_guru64_dft}. @example fftw_plan fftw_plan_guru64_dft( int rank, const fftw_iodim64 *dims, int howmany_rank, const fftw_iodim64 *howmany_dims, fftw_complex *in, fftw_complex *out, int sign, unsigned flags); @end example @findex fftw_plan_guru64_dft The @code{fftw_iodim64} type is similar to @code{fftw_iodim}, with the same interpretation, except that it uses type @code{ptrdiff_t} instead of type @code{int}. @example typedef struct @{ ptrdiff_t n; ptrdiff_t is; ptrdiff_t os; @} fftw_iodim64; @end example @tindex fftw_iodim64 Every other @samp{fftw_plan_guru} function also has a @samp{fftw_plan_guru64} equivalent, but we do not repeat their documentation here since they are identical to the 32-bit versions except as noted above. @c ----------------------------------------------------------- @node New-array Execute Functions, Wisdom, Guru Interface, FFTW Reference @section New-array Execute Functions @cindex execute @cindex new-array execution Normally, one executes a plan for the arrays with which the plan was created, by calling @code{fftw_execute(plan)} as described in @ref{Using Plans}. @findex fftw_execute However, it is possible for sophisticated users to apply a given plan to a @emph{different} array using the ``new-array execute'' functions detailed below, provided that the following conditions are met: @itemize @bullet @item The array size, strides, etcetera are the same (since those are set by the plan). @item The input and output arrays are the same (in-place) or different (out-of-place) if the plan was originally created to be in-place or out-of-place, respectively. @item For split arrays, the separations between the real and imaginary parts, @code{ii-ri} and @code{io-ro}, are the same as they were for the input and output arrays when the plan was created. (This condition is automatically satisfied for interleaved arrays.) @item The @dfn{alignment} of the new input/output arrays is the same as that of the input/output arrays when the plan was created, unless the plan was created with the @code{FFTW_UNALIGNED} flag. @ctindex FFTW_UNALIGNED Here, the alignment is a platform-dependent quantity (for example, it is the address modulo 16 if SSE SIMD instructions are used, but the address modulo 4 for non-SIMD single-precision FFTW on the same machine). In general, only arrays allocated with @code{fftw_malloc} are guaranteed to be equally aligned (@pxref{SIMD alignment and fftw_malloc}). @end itemize @cindex alignment The alignment issue is especially critical, because if you don't use @code{fftw_malloc} then you may have little control over the alignment of arrays in memory. For example, neither the C++ @code{new} function nor the Fortran @code{allocate} statement provide strong enough guarantees about data alignment. If you don't use @code{fftw_malloc}, therefore, you probably have to use @code{FFTW_UNALIGNED} (which disables most SIMD support). If possible, it is probably better for you to simply create multiple plans (creating a new plan is quick once one exists for a given size), or better yet re-use the same array for your transforms. If you are tempted to use the new-array execute interface because you want to transform a known bunch of arrays of the same size, you should probably go use the advanced interface instead (@pxref{Advanced Interface})). The new-array execute functions are: @example void fftw_execute_dft( const fftw_plan p, fftw_complex *in, fftw_complex *out); void fftw_execute_split_dft( const fftw_plan p, double *ri, double *ii, double *ro, double *io); void fftw_execute_dft_r2c( const fftw_plan p, double *in, fftw_complex *out); void fftw_execute_split_dft_r2c( const fftw_plan p, double *in, double *ro, double *io); void fftw_execute_dft_c2r( const fftw_plan p, fftw_complex *in, double *out); void fftw_execute_split_dft_c2r( const fftw_plan p, double *ri, double *ii, double *out); void fftw_execute_r2r( const fftw_plan p, double *in, double *out); @end example @findex fftw_execute_dft @findex fftw_execute_split_dft @findex fftw_execute_dft_r2c @findex fftw_execute_split_dft_r2c @findex fftw_execute_dft_c2r @findex fftw_execute_split_dft_c2r @findex fftw_execute_r2r These execute the @code{plan} to compute the corresponding transform on the input/output arrays specified by the subsequent arguments. The input/output array arguments have the same meanings as the ones passed to the guru planner routines in the preceding sections. The @code{plan} is not modified, and these routines can be called as many times as desired, or intermixed with calls to the ordinary @code{fftw_execute}. The @code{plan} @emph{must} have been created for the transform type corresponding to the execute function, e.g. it must be a complex-DFT plan for @code{fftw_execute_dft}. Any of the planner routines for that transform type, from the basic to the guru interface, could have been used to create the plan, however. @c ------------------------------------------------------------ @node Wisdom, What FFTW Really Computes, New-array Execute Functions, FFTW Reference @section Wisdom @cindex wisdom @cindex saving plans to disk This section documents the FFTW mechanism for saving and restoring plans from disk. This mechanism is called @dfn{wisdom}. @menu * Wisdom Export:: * Wisdom Import:: * Forgetting Wisdom:: * Wisdom Utilities:: @end menu @c =========> @node Wisdom Export, Wisdom Import, Wisdom, Wisdom @subsection Wisdom Export @example int fftw_export_wisdom_to_filename(const char *filename); void fftw_export_wisdom_to_file(FILE *output_file); char *fftw_export_wisdom_to_string(void); void fftw_export_wisdom(void (*write_char)(char c, void *), void *data); @end example @findex fftw_export_wisdom @findex fftw_export_wisdom_to_filename @findex fftw_export_wisdom_to_file @findex fftw_export_wisdom_to_string These functions allow you to export all currently accumulated wisdom in a form from which it can be later imported and restored, even during a separate run of the program. (@xref{Words of Wisdom-Saving Plans}.) The current store of wisdom is not affected by calling any of these routines. @code{fftw_export_wisdom} exports the wisdom to any output medium, as specified by the callback function @code{write_char}. @code{write_char} is a @code{putc}-like function that writes the character @code{c} to some output; its second parameter is the @code{data} pointer passed to @code{fftw_export_wisdom}. For convenience, the following three ``wrapper'' routines are provided: @code{fftw_export_wisdom_to_filename} writes wisdom to a file named @code{filename} (which is created or overwritten), returning @code{1} on success and @code{0} on failure. A lower-level function, which requires you to open and close the file yourself (e.g. if you want to write wisdom to a portion of a larger file) is @code{fftw_export_wisdom_to_file}. This writes the wisdom to the current position in @code{output_file}, which should be open with write permission; upon exit, the file remains open and is positioned at the end of the wisdom data. @code{fftw_export_wisdom_to_string} returns a pointer to a @code{NULL}-terminated string holding the wisdom data. This string is dynamically allocated, and it is the responsibility of the caller to deallocate it with @code{free} when it is no longer needed. All of these routines export the wisdom in the same format, which we will not document here except to say that it is LISP-like ASCII text that is insensitive to white space. @c =========> @node Wisdom Import, Forgetting Wisdom, Wisdom Export, Wisdom @subsection Wisdom Import @example int fftw_import_system_wisdom(void); int fftw_import_wisdom_from_filename(const char *filename); int fftw_import_wisdom_from_string(const char *input_string); int fftw_import_wisdom(int (*read_char)(void *), void *data); @end example @findex fftw_import_wisdom @findex fftw_import_system_wisdom @findex fftw_import_wisdom_from_filename @findex fftw_import_wisdom_from_file @findex fftw_import_wisdom_from_string These functions import wisdom into a program from data stored by the @code{fftw_export_wisdom} functions above. (@xref{Words of Wisdom-Saving Plans}.) The imported wisdom replaces any wisdom already accumulated by the running program. @code{fftw_import_wisdom} imports wisdom from any input medium, as specified by the callback function @code{read_char}. @code{read_char} is a @code{getc}-like function that returns the next character in the input; its parameter is the @code{data} pointer passed to @code{fftw_import_wisdom}. If the end of the input data is reached (which should never happen for valid data), @code{read_char} should return @code{EOF} (as defined in @code{}). For convenience, the following three ``wrapper'' routines are provided: @code{fftw_import_wisdom_from_filename} reads wisdom from a file named @code{filename}. A lower-level function, which requires you to open and close the file yourself (e.g. if you want to read wisdom from a portion of a larger file) is @code{fftw_import_wisdom_from_file}. This reads wisdom from the current position in @code{input_file} (which should be open with read permission); upon exit, the file remains open, but the position of the read pointer is unspecified. @code{fftw_import_wisdom_from_string} reads wisdom from the @code{NULL}-terminated string @code{input_string}. @code{fftw_import_system_wisdom} reads wisdom from an implementation-defined standard file (@code{/etc/fftw/wisdom} on Unix and GNU systems). @cindex wisdom, system-wide The return value of these import routines is @code{1} if the wisdom was read successfully and @code{0} otherwise. Note that, in all of these functions, any data in the input stream past the end of the wisdom data is simply ignored. @c =========> @node Forgetting Wisdom, Wisdom Utilities, Wisdom Import, Wisdom @subsection Forgetting Wisdom @example void fftw_forget_wisdom(void); @end example @findex fftw_forget_wisdom Calling @code{fftw_forget_wisdom} causes all accumulated @code{wisdom} to be discarded and its associated memory to be freed. (New @code{wisdom} can still be gathered subsequently, however.) @c =========> @node Wisdom Utilities, , Forgetting Wisdom, Wisdom @subsection Wisdom Utilities FFTW includes two standalone utility programs that deal with wisdom. We merely summarize them here, since they come with their own @code{man} pages for Unix and GNU systems (with HTML versions on our web site). The first program is @code{fftw-wisdom} (or @code{fftwf-wisdom} in single precision, etcetera), which can be used to create a wisdom file containing plans for any of the transform sizes and types supported by FFTW. It is preferable to create wisdom directly from your executable (@pxref{Caveats in Using Wisdom}), but this program is useful for creating global wisdom files for @code{fftw_import_system_wisdom}. @cindex fftw-wisdom utility The second program is @code{fftw-wisdom-to-conf}, which takes a wisdom file as input and produces a @dfn{configuration routine} as output. The latter is a C subroutine that you can compile and link into your program, replacing a routine of the same name in the FFTW library, that determines which parts of FFTW are callable by your program. @code{fftw-wisdom-to-conf} produces a configuration routine that links to only those parts of FFTW needed by the saved plans in the wisdom, greatly reducing the size of statically linked executables (which should only attempt to create plans corresponding to those in the wisdom, however). @cindex fftw-wisdom-to-conf utility @cindex configuration routines @c ------------------------------------------------------------ @node What FFTW Really Computes, , Wisdom, FFTW Reference @section What FFTW Really Computes In this section, we provide precise mathematical definitions for the transforms that FFTW computes. These transform definitions are fairly standard, but some authors follow slightly different conventions for the normalization of the transform (the constant factor in front) and the sign of the complex exponent. We begin by presenting the one-dimensional (1d) transform definitions, and then give the straightforward extension to multi-dimensional transforms. @menu * The 1d Discrete Fourier Transform (DFT):: * The 1d Real-data DFT:: * 1d Real-even DFTs (DCTs):: * 1d Real-odd DFTs (DSTs):: * 1d Discrete Hartley Transforms (DHTs):: * Multi-dimensional Transforms:: @end menu @c =========> @node The 1d Discrete Fourier Transform (DFT), The 1d Real-data DFT, What FFTW Really Computes, What FFTW Really Computes @subsection The 1d Discrete Fourier Transform (DFT) @cindex discrete Fourier transform @cindex DFT The forward (@code{FFTW_FORWARD}) discrete Fourier transform (DFT) of a 1d complex array @math{X} of size @math{n} computes an array @math{Y}, where: @tex $$ Y_k = \sum_{j = 0}^{n - 1} X_j e^{-2\pi j k \sqrt{-1}/n} \ . $$ @end tex @ifinfo @center Y[k] = sum for j = 0 to (n - 1) of X[j] * exp(-2 pi j k sqrt(-1)/n) . @end ifinfo @html
.
@end html The backward (@code{FFTW_BACKWARD}) DFT computes: @tex $$ Y_k = \sum_{j = 0}^{n - 1} X_j e^{2\pi j k \sqrt{-1}/n} \ . $$ @end tex @ifinfo @center Y[k] = sum for j = 0 to (n - 1) of X[j] * exp(2 pi j k sqrt(-1)/n) . @end ifinfo @html
.
@end html @cindex normalization FFTW computes an unnormalized transform, in that there is no coefficient in front of the summation in the DFT. In other words, applying the forward and then the backward transform will multiply the input by @math{n}. @cindex frequency From above, an @code{FFTW_FORWARD} transform corresponds to a sign of @math{-1} in the exponent of the DFT. Note also that we use the standard ``in-order'' output ordering---the @math{k}-th output corresponds to the frequency @math{k/n} (or @math{k/T}, where @math{T} is your total sampling period). For those who like to think in terms of positive and negative frequencies, this means that the positive frequencies are stored in the first half of the output and the negative frequencies are stored in backwards order in the second half of the output. (The frequency @math{-k/n} is the same as the frequency @math{(n-k)/n}.) @c =========> @node The 1d Real-data DFT, 1d Real-even DFTs (DCTs), The 1d Discrete Fourier Transform (DFT), What FFTW Really Computes @subsection The 1d Real-data DFT The real-input (r2c) DFT in FFTW computes the @emph{forward} transform @math{Y} of the size @code{n} real array @math{X}, exactly as defined above, i.e. @tex $$ Y_k = \sum_{j = 0}^{n - 1} X_j e^{-2\pi j k \sqrt{-1}/n} \ . $$ @end tex @ifinfo @center Y[k] = sum for j = 0 to (n - 1) of X[j] * exp(-2 pi j k sqrt(-1)/n) . @end ifinfo @html
.
@end html This output array @math{Y} can easily be shown to possess the ``Hermitian'' symmetry @cindex Hermitian @tex $Y_k = Y_{n-k}^*$, @end tex @ifinfo Y[k] = Y[n-k]*, @end ifinfo @html Yk = Yn-k*, @end html where we take @math{Y} to be periodic so that @tex $Y_n = Y_0$. @end tex @ifinfo Y[n] = Y[0]. @end ifinfo @html Yn = Y0. @end html As a result of this symmetry, half of the output @math{Y} is redundant (being the complex conjugate of the other half), and so the 1d r2c transforms only output elements @math{0}@dots{}@math{n/2} of @math{Y} (@math{n/2+1} complex numbers), where the division by @math{2} is rounded down. Moreover, the Hermitian symmetry implies that @tex $Y_0$ @end tex @ifinfo Y[0] @end ifinfo @html Y0 @end html and, if @math{n} is even, the @tex $Y_{n/2}$ @end tex @ifinfo Y[n/2] @end ifinfo @html Yn/2 @end html element, are purely real. So, for the @code{R2HC} r2r transform, these elements are not stored in the halfcomplex output format. @cindex r2r @ctindex R2HC @cindex halfcomplex format The c2r and @code{H2RC} r2r transforms compute the backward DFT of the @emph{complex} array @math{X} with Hermitian symmetry, stored in the r2c/@code{R2HC} output formats, respectively, where the backward transform is defined exactly as for the complex case: @tex $$ Y_k = \sum_{j = 0}^{n - 1} X_j e^{2\pi j k \sqrt{-1}/n} \ . $$ @end tex @ifinfo @center Y[k] = sum for j = 0 to (n - 1) of X[j] * exp(2 pi j k sqrt(-1)/n) . @end ifinfo @html
.
@end html The outputs @code{Y} of this transform can easily be seen to be purely real, and are stored as an array of real numbers. @cindex normalization Like FFTW's complex DFT, these transforms are unnormalized. In other words, applying the real-to-complex (forward) and then the complex-to-real (backward) transform will multiply the input by @math{n}. @c =========> @node 1d Real-even DFTs (DCTs), 1d Real-odd DFTs (DSTs), The 1d Real-data DFT, What FFTW Really Computes @subsection 1d Real-even DFTs (DCTs) The Real-even symmetry DFTs in FFTW are exactly equivalent to the unnormalized forward (and backward) DFTs as defined above, where the input array @math{X} of length @math{N} is purely real and is also @dfn{even} symmetry. In this case, the output array is likewise real and even symmetry. @cindex real-even DFT @cindex REDFT @ctindex REDFT00 For the case of @code{REDFT00}, this even symmetry means that @tex $X_j = X_{N-j}$, @end tex @ifinfo X[j] = X[N-j], @end ifinfo @html Xj = XN-j, @end html where we take @math{X} to be periodic so that @tex $X_N = X_0$. @end tex @ifinfo X[N] = X[0]. @end ifinfo @html XN = X0. @end html Because of this redundancy, only the first @math{n} real numbers are actually stored, where @math{N = 2(n-1)}. The proper definition of even symmetry for @code{REDFT10}, @code{REDFT01}, and @code{REDFT11} transforms is somewhat more intricate because of the shifts by @math{1/2} of the input and/or output, although the corresponding boundary conditions are given in @ref{Real even/odd DFTs (cosine/sine transforms)}. Because of the even symmetry, however, the sine terms in the DFT all cancel and the remaining cosine terms are written explicitly below. This formulation often leads people to call such a transform a @dfn{discrete cosine transform} (DCT), although it is really just a special case of the DFT. @cindex discrete cosine transform @cindex DCT In each of the definitions below, we transform a real array @math{X} of length @math{n} to a real array @math{Y} of length @math{n}: @subsubheading REDFT00 (DCT-I) @ctindex REDFT00 An @code{REDFT00} transform (type-I DCT) in FFTW is defined by: @tex $$ Y_k = X_0 + (-1)^k X_{n-1} + 2 \sum_{j=1}^{n-2} X_j \cos [ \pi j k / (n-1)]. $$ @end tex @ifinfo Y[k] = X[0] + (-1)^k X[n-1] + 2 (sum for j = 1 to n-2 of X[j] cos(pi jk /(n-1))). @end ifinfo @html
.
@end html Note that this transform is not defined for @math{n=1}. For @math{n=2}, the summation term above is dropped as you might expect. @subsubheading REDFT10 (DCT-II) @ctindex REDFT10 An @code{REDFT10} transform (type-II DCT, sometimes called ``the'' DCT) in FFTW is defined by: @tex $$ Y_k = 2 \sum_{j=0}^{n-1} X_j \cos [\pi (j+1/2) k / n]. $$ @end tex @ifinfo Y[k] = 2 (sum for j = 0 to n-1 of X[j] cos(pi (j+1/2) k / n)). @end ifinfo @html
.
@end html @subsubheading REDFT01 (DCT-III) @ctindex REDFT01 An @code{REDFT01} transform (type-III DCT) in FFTW is defined by: @tex $$ Y_k = X_0 + 2 \sum_{j=1}^{n-1} X_j \cos [\pi j (k+1/2) / n]. $$ @end tex @ifinfo Y[k] = X[0] + 2 (sum for j = 1 to n-1 of X[j] cos(pi j (k+1/2) / n)). @end ifinfo @html
.
@end html In the case of @math{n=1}, this reduces to @tex $Y_0 = X_0$. @end tex @ifinfo Y[0] = X[0]. @end ifinfo @html Y0 = X0. @end html Up to a scale factor (see below), this is the inverse of @code{REDFT10} (``the'' DCT), and so the @code{REDFT01} (DCT-III) is sometimes called the ``IDCT''. @cindex IDCT @subsubheading REDFT11 (DCT-IV) @ctindex REDFT11 An @code{REDFT11} transform (type-IV DCT) in FFTW is defined by: @tex $$ Y_k = 2 \sum_{j=0}^{n-1} X_j \cos [\pi (j+1/2) (k+1/2) / n]. $$ @end tex @ifinfo Y[k] = 2 (sum for j = 0 to n-1 of X[j] cos(pi (j+1/2) (k+1/2) / n)). @end ifinfo @html
.
@end html @subsubheading Inverses and Normalization These definitions correspond directly to the unnormalized DFTs used elsewhere in FFTW (hence the factors of @math{2} in front of the summations). The unnormalized inverse of @code{REDFT00} is @code{REDFT00}, of @code{REDFT10} is @code{REDFT01} and vice versa, and of @code{REDFT11} is @code{REDFT11}. Each unnormalized inverse results in the original array multiplied by @math{N}, where @math{N} is the @emph{logical} DFT size. For @code{REDFT00}, @math{N=2(n-1)} (note that @math{n=1} is not defined); otherwise, @math{N=2n}. @cindex normalization In defining the discrete cosine transform, some authors also include additional factors of @ifinfo sqrt(2) @end ifinfo @html √2 @end html @tex $\sqrt{2}$ @end tex (or its inverse) multiplying selected inputs and/or outputs. This is a mostly cosmetic change that makes the transform orthogonal, but sacrifices the direct equivalence to a symmetric DFT. @c =========> @node 1d Real-odd DFTs (DSTs), 1d Discrete Hartley Transforms (DHTs), 1d Real-even DFTs (DCTs), What FFTW Really Computes @subsection 1d Real-odd DFTs (DSTs) The Real-odd symmetry DFTs in FFTW are exactly equivalent to the unnormalized forward (and backward) DFTs as defined above, where the input array @math{X} of length @math{N} is purely real and is also @dfn{odd} symmetry. In this case, the output is odd symmetry and purely imaginary. @cindex real-odd DFT @cindex RODFT @ctindex RODFT00 For the case of @code{RODFT00}, this odd symmetry means that @tex $X_j = -X_{N-j}$, @end tex @ifinfo X[j] = -X[N-j], @end ifinfo @html Xj = -XN-j, @end html where we take @math{X} to be periodic so that @tex $X_N = X_0$. @end tex @ifinfo X[N] = X[0]. @end ifinfo @html XN = X0. @end html Because of this redundancy, only the first @math{n} real numbers starting at @math{j=1} are actually stored (the @math{j=0} element is zero), where @math{N = 2(n+1)}. The proper definition of odd symmetry for @code{RODFT10}, @code{RODFT01}, and @code{RODFT11} transforms is somewhat more intricate because of the shifts by @math{1/2} of the input and/or output, although the corresponding boundary conditions are given in @ref{Real even/odd DFTs (cosine/sine transforms)}. Because of the odd symmetry, however, the cosine terms in the DFT all cancel and the remaining sine terms are written explicitly below. This formulation often leads people to call such a transform a @dfn{discrete sine transform} (DST), although it is really just a special case of the DFT. @cindex discrete sine transform @cindex DST In each of the definitions below, we transform a real array @math{X} of length @math{n} to a real array @math{Y} of length @math{n}: @subsubheading RODFT00 (DST-I) @ctindex RODFT00 An @code{RODFT00} transform (type-I DST) in FFTW is defined by: @tex $$ Y_k = 2 \sum_{j=0}^{n-1} X_j \sin [ \pi (j+1) (k+1) / (n+1)]. $$ @end tex @ifinfo Y[k] = 2 (sum for j = 0 to n-1 of X[j] sin(pi (j+1)(k+1) / (n+1))). @end ifinfo @html
.
@end html @subsubheading RODFT10 (DST-II) @ctindex RODFT10 An @code{RODFT10} transform (type-II DST) in FFTW is defined by: @tex $$ Y_k = 2 \sum_{j=0}^{n-1} X_j \sin [\pi (j+1/2) (k+1) / n]. $$ @end tex @ifinfo Y[k] = 2 (sum for j = 0 to n-1 of X[j] sin(pi (j+1/2) (k+1) / n)). @end ifinfo @html
.
@end html @subsubheading RODFT01 (DST-III) @ctindex RODFT01 An @code{RODFT01} transform (type-III DST) in FFTW is defined by: @tex $$ Y_k = (-1)^k X_{n-1} + 2 \sum_{j=0}^{n-2} X_j \sin [\pi (j+1) (k+1/2) / n]. $$ @end tex @ifinfo Y[k] = (-1)^k X[n-1] + 2 (sum for j = 0 to n-2 of X[j] sin(pi (j+1) (k+1/2) / n)). @end ifinfo @html
.
@end html In the case of @math{n=1}, this reduces to @tex $Y_0 = X_0$. @end tex @ifinfo Y[0] = X[0]. @end ifinfo @html Y0 = X0. @end html @subsubheading RODFT11 (DST-IV) @ctindex RODFT11 An @code{RODFT11} transform (type-IV DST) in FFTW is defined by: @tex $$ Y_k = 2 \sum_{j=0}^{n-1} X_j \sin [\pi (j+1/2) (k+1/2) / n]. $$ @end tex @ifinfo Y[k] = 2 (sum for j = 0 to n-1 of X[j] sin(pi (j+1/2) (k+1/2) / n)). @end ifinfo @html
.
@end html @subsubheading Inverses and Normalization These definitions correspond directly to the unnormalized DFTs used elsewhere in FFTW (hence the factors of @math{2} in front of the summations). The unnormalized inverse of @code{RODFT00} is @code{RODFT00}, of @code{RODFT10} is @code{RODFT01} and vice versa, and of @code{RODFT11} is @code{RODFT11}. Each unnormalized inverse results in the original array multiplied by @math{N}, where @math{N} is the @emph{logical} DFT size. For @code{RODFT00}, @math{N=2(n+1)}; otherwise, @math{N=2n}. @cindex normalization In defining the discrete sine transform, some authors also include additional factors of @ifinfo sqrt(2) @end ifinfo @html √2 @end html @tex $\sqrt{2}$ @end tex (or its inverse) multiplying selected inputs and/or outputs. This is a mostly cosmetic change that makes the transform orthogonal, but sacrifices the direct equivalence to an antisymmetric DFT. @c =========> @node 1d Discrete Hartley Transforms (DHTs), Multi-dimensional Transforms, 1d Real-odd DFTs (DSTs), What FFTW Really Computes @subsection 1d Discrete Hartley Transforms (DHTs) @cindex discrete Hartley transform @cindex DHT The discrete Hartley transform (DHT) of a 1d real array @math{X} of size @math{n} computes a real array @math{Y} of the same size, where: @tex $$ Y_k = \sum_{j = 0}^{n - 1} X_j [ \cos(2\pi j k / n) + \sin(2\pi j k / n)]. $$ @end tex @ifinfo @center Y[k] = sum for j = 0 to (n - 1) of X[j] * [cos(2 pi j k / n) + sin(2 pi j k / n)]. @end ifinfo @html
.
@end html @cindex normalization FFTW computes an unnormalized transform, in that there is no coefficient in front of the summation in the DHT. In other words, applying the transform twice (the DHT is its own inverse) will multiply the input by @math{n}. @c =========> @node Multi-dimensional Transforms, , 1d Discrete Hartley Transforms (DHTs), What FFTW Really Computes @subsection Multi-dimensional Transforms The multi-dimensional transforms of FFTW, in general, compute simply the separable product of the given 1d transform along each dimension of the array. Since each of these transforms is unnormalized, computing the forward followed by the backward/inverse multi-dimensional transform will result in the original array scaled by the product of the normalization factors for each dimension (e.g. the product of the dimension sizes, for a multi-dimensional DFT). @tex As an explicit example, consider the following exact mathematical definition of our multi-dimensional DFT. Let $X$ be a $d$-dimensional complex array whose elements are $X[j_1, j_2, \ldots, j_d]$, where $0 \leq j_s < n_s$ for all~$s \in \{ 1, 2, \ldots, d \}$. Let also $\omega_s = e^{2\pi \sqrt{-1}/n_s}$, for all ~$s \in \{ 1, 2, \ldots, d \}$. The forward transform computes a complex array~$Y$, whose structure is the same as that of~$X$, defined by $$ Y[k_1, k_2, \ldots, k_d] = \sum_{j_1 = 0}^{n_1 - 1} \sum_{j_2 = 0}^{n_2 - 1} \cdots \sum_{j_d = 0}^{n_d - 1} X[j_1, j_2, \ldots, j_d] \omega_1^{-j_1 k_1} \omega_2^{-j_2 k_2} \cdots \omega_d^{-j_d k_d} \ . $$ The backward transform computes $$ Y[k_1, k_2, \ldots, k_d] = \sum_{j_1 = 0}^{n_1 - 1} \sum_{j_2 = 0}^{n_2 - 1} \cdots \sum_{j_d = 0}^{n_d - 1} X[j_1, j_2, \ldots, j_d] \omega_1^{j_1 k_1} \omega_2^{j_2 k_2} \cdots \omega_d^{j_d k_d} \ . $$ Computing the forward transform followed by the backward transform will multiply the array by $\prod_{s=1}^{d} n_d$. @end tex @cindex r2c The definition of FFTW's multi-dimensional DFT of real data (r2c) deserves special attention. In this case, we logically compute the full multi-dimensional DFT of the input data; since the input data are purely real, the output data have the Hermitian symmetry and therefore only one non-redundant half need be stored. More specifically, for an @ndims multi-dimensional real-input DFT, the full (logical) complex output array @tex $Y[k_0, k_1, \ldots, k_{d-1}]$ @end tex @html Y[k0, k1, ..., kd-1] @end html @ifinfo Y[k[0], k[1], ..., k[d-1]] @end ifinfo has the symmetry: @tex $$ Y[k_0, k_1, \ldots, k_{d-1}] = Y[n_0 - k_0, n_1 - k_1, \ldots, n_{d-1} - k_{d-1}]^* $$ @end tex @html Y[k0, k1, ..., kd-1] = Y[n0 - k0, n1 - k1, ..., nd-1 - kd-1]* @end html @ifinfo Y[k[0], k[1], ..., k[d-1]] = Y[n[0] - k[0], n[1] - k[1], ..., n[d-1] - k[d-1]]* @end ifinfo (where each dimension is periodic). Because of this symmetry, we only store the @tex $k_{d-1} = 0 \cdots n_{d-1}/2$ @end tex @html kd-1 = 0...nd-1/2+1 @end html @ifinfo k[d-1] = 0...n[d-1]/2 @end ifinfo elements of the @emph{last} dimension (division by @math{2} is rounded down). (We could instead have cut any other dimension in half, but the last dimension proved computationally convenient.) This results in the peculiar array format described in more detail by @ref{Real-data DFT Array Format}. The multi-dimensional c2r transform is simply the unnormalized inverse of the r2c transform. i.e. it is the same as FFTW's complex backward multi-dimensional DFT, operating on a Hermitian input array in the peculiar format mentioned above and outputting a real array (since the DFT output is purely real). We should remind the user that the separable product of 1d transforms along each dimension, as computed by FFTW, is not always the same thing as the usual multi-dimensional transform. A multi-dimensional @code{R2HC} (or @code{HC2R}) transform is not identical to the multi-dimensional DFT, requiring some post-processing to combine the requisite real and imaginary parts, as was described in @ref{The Halfcomplex-format DFT}. Likewise, FFTW's multidimensional @code{FFTW_DHT} r2r transform is not the same thing as the logical multi-dimensional discrete Hartley transform defined in the literature, as discussed in @ref{The Discrete Hartley Transform}. fftw-3.3.3/doc/threads.texi0000644000175000017500000002416512054410050012500 00000000000000@node Multi-threaded FFTW, Distributed-memory FFTW with MPI, FFTW Reference, Top @chapter Multi-threaded FFTW @cindex parallel transform In this chapter we document the parallel FFTW routines for shared-memory parallel hardware. These routines, which support parallel one- and multi-dimensional transforms of both real and complex data, are the easiest way to take advantage of multiple processors with FFTW. They work just like the corresponding uniprocessor transform routines, except that you have an extra initialization routine to call, and there is a routine to set the number of threads to employ. Any program that uses the uniprocessor FFTW can therefore be trivially modified to use the multi-threaded FFTW. A shared-memory machine is one in which all CPUs can directly access the same main memory, and such machines are now common due to the ubiquity of multi-core CPUs. FFTW's multi-threading support allows you to utilize these additional CPUs transparently from a single program. However, this does not necessarily translate into performance gains---when multiple threads/CPUs are employed, there is an overhead required for synchronization that may outweigh the computatational parallelism. Therefore, you can only benefit from threads if your problem is sufficiently large. @cindex shared-memory @cindex threads @menu * Installation and Supported Hardware/Software:: * Usage of Multi-threaded FFTW:: * How Many Threads to Use?:: * Thread safety:: @end menu @c ------------------------------------------------------------ @node Installation and Supported Hardware/Software, Usage of Multi-threaded FFTW, Multi-threaded FFTW, Multi-threaded FFTW @section Installation and Supported Hardware/Software All of the FFTW threads code is located in the @code{threads} subdirectory of the FFTW package. On Unix systems, the FFTW threads libraries and header files can be automatically configured, compiled, and installed along with the uniprocessor FFTW libraries simply by including @code{--enable-threads} in the flags to the @code{configure} script (@pxref{Installation on Unix}), or @code{--enable-openmp} to use @uref{http://www.openmp.org,OpenMP} threads. @fpindex configure @cindex portability @cindex OpenMP The threads routines require your operating system to have some sort of shared-memory threads support. Specifically, the FFTW threads package works with POSIX threads (available on most Unix variants, from GNU/Linux to MacOS X) and Win32 threads. OpenMP threads, which are supported in many common compilers (e.g. gcc) are also supported, and may give better performance on some systems. (OpenMP threads are also useful if you are employing OpenMP in your own code, in order to minimize conflicts between threading models.) If you have a shared-memory machine that uses a different threads API, it should be a simple matter of programming to include support for it; see the file @code{threads/threads.c} for more detail. You can compile FFTW with @emph{both} @code{--enable-threads} and @code{--enable-openmp} at the same time, since they install libraries with different names (@samp{fftw3_threads} and @samp{fftw3_omp}, as described below). However, your programs may only link to @emph{one} of these two libraries at a time. Ideally, of course, you should also have multiple processors in order to get any benefit from the threaded transforms. @c ------------------------------------------------------------ @node Usage of Multi-threaded FFTW, How Many Threads to Use?, Installation and Supported Hardware/Software, Multi-threaded FFTW @section Usage of Multi-threaded FFTW Here, it is assumed that the reader is already familiar with the usage of the uniprocessor FFTW routines, described elsewhere in this manual. We only describe what one has to change in order to use the multi-threaded routines. @cindex OpenMP First, programs using the parallel complex transforms should be linked with @code{-lfftw3_threads -lfftw3 -lm} on Unix, or @code{-lfftw3_omp -lfftw3 -lm} if you compiled with OpenMP. You will also need to link with whatever library is responsible for threads on your system (e.g. @code{-lpthread} on GNU/Linux) or include whatever compiler flag enables OpenMP (e.g. @code{-fopenmp} with gcc). @cindex linking on Unix Second, before calling @emph{any} FFTW routines, you should call the function: @example int fftw_init_threads(void); @end example @findex fftw_init_threads This function, which need only be called once, performs any one-time initialization required to use threads on your system. It returns zero if there was some error (which should not happen under normal circumstances) and a non-zero value otherwise. Third, before creating a plan that you want to parallelize, you should call: @example void fftw_plan_with_nthreads(int nthreads); @end example @findex fftw_plan_with_nthreads The @code{nthreads} argument indicates the number of threads you want FFTW to use (or actually, the maximum number). All plans subsequently created with any planner routine will use that many threads. You can call @code{fftw_plan_with_nthreads}, create some plans, call @code{fftw_plan_with_nthreads} again with a different argument, and create some more plans for a new number of threads. Plans already created before a call to @code{fftw_plan_with_nthreads} are unaffected. If you pass an @code{nthreads} argument of @code{1} (the default), threads are disabled for subsequent plans. @cindex OpenMP With OpenMP, to configure FFTW to use all of the currently running OpenMP threads (set by @code{omp_set_num_threads(nthreads)} or by the @code{OMP_NUM_THREADS} environment variable), you can do: @code{fftw_plan_with_nthreads(omp_get_max_threads())}. (The @samp{omp_} OpenMP functions are declared via @code{#include }.) @cindex thread safety Given a plan, you then execute it as usual with @code{fftw_execute(plan)}, and the execution will use the number of threads specified when the plan was created. When done, you destroy it as usual with @code{fftw_destroy_plan}. As described in @ref{Thread safety}, plan @emph{execution} is thread-safe, but plan creation and destruction are @emph{not}: you should create/destroy plans only from a single thread, but can safely execute multiple plans in parallel. There is one additional routine: if you want to get rid of all memory and other resources allocated internally by FFTW, you can call: @example void fftw_cleanup_threads(void); @end example @findex fftw_cleanup_threads which is much like the @code{fftw_cleanup()} function except that it also gets rid of threads-related data. You must @emph{not} execute any previously created plans after calling this function. We should also mention one other restriction: if you save wisdom from a program using the multi-threaded FFTW, that wisdom @emph{cannot be used} by a program using only the single-threaded FFTW (i.e. not calling @code{fftw_init_threads}). @xref{Words of Wisdom-Saving Plans}. @c ------------------------------------------------------------ @node How Many Threads to Use?, Thread safety, Usage of Multi-threaded FFTW, Multi-threaded FFTW @section How Many Threads to Use? @cindex number of threads There is a fair amount of overhead involved in synchronizing threads, so the optimal number of threads to use depends upon the size of the transform as well as on the number of processors you have. As a general rule, you don't want to use more threads than you have processors. (Using more threads will work, but there will be extra overhead with no benefit.) In fact, if the problem size is too small, you may want to use fewer threads than you have processors. You will have to experiment with your system to see what level of parallelization is best for your problem size. Typically, the problem will have to involve at least a few thousand data points before threads become beneficial. If you plan with @code{FFTW_PATIENT}, it will automatically disable threads for sizes that don't benefit from parallelization. @ctindex FFTW_PATIENT @c ------------------------------------------------------------ @node Thread safety, , How Many Threads to Use?, Multi-threaded FFTW @section Thread safety @cindex threads @cindex OpenMP @cindex thread safety Users writing multi-threaded programs (including OpenMP) must concern themselves with the @dfn{thread safety} of the libraries they use---that is, whether it is safe to call routines in parallel from multiple threads. FFTW can be used in such an environment, but some care must be taken because the planner routines share data (e.g. wisdom and trigonometric tables) between calls and plans. The upshot is that the only thread-safe (re-entrant) routine in FFTW is @code{fftw_execute} (and the new-array variants thereof). All other routines (e.g. the planner) should only be called from one thread at a time. So, for example, you can wrap a semaphore lock around any calls to the planner; even more simply, you can just create all of your plans from one thread. We do not think this should be an important restriction (FFTW is designed for the situation where the only performance-sensitive code is the actual execution of the transform), and the benefits of shared data between plans are great. Note also that, since the plan is not modified by @code{fftw_execute}, it is safe to execute the @emph{same plan} in parallel by multiple threads. However, since a given plan operates by default on a fixed array, you need to use one of the new-array execute functions (@pxref{New-array Execute Functions}) so that different threads compute the transform of different data. (Users should note that these comments only apply to programs using shared-memory threads or OpenMP. Parallelism using MPI or forked processes involves a separate address-space and global variables for each process, and is not susceptible to problems of this sort.) If you are configured FFTW with the @code{--enable-debug} or @code{--enable-debug-malloc} flags (@pxref{Installation on Unix}), then @code{fftw_execute} is not thread-safe. These flags are not documented because they are intended only for developing and debugging FFTW, but if you must use @code{--enable-debug} then you should also specifically pass @code{--disable-debug-malloc} for @code{fftw_execute} to be thread-safe. fftw-3.3.3/doc/tutorial.texi0000644000175000017500000011301112054410050012676 00000000000000@node Tutorial, Other Important Topics, Introduction, Top @chapter Tutorial @menu * Complex One-Dimensional DFTs:: * Complex Multi-Dimensional DFTs:: * One-Dimensional DFTs of Real Data:: * Multi-Dimensional DFTs of Real Data:: * More DFTs of Real Data:: @end menu This chapter describes the basic usage of FFTW, i.e., how to compute @cindex basic interface the Fourier transform of a single array. This chapter tells the truth, but not the @emph{whole} truth. Specifically, FFTW implements additional routines and flags that are not documented here, although in many cases we try to indicate where added capabilities exist. For more complete information, see @ref{FFTW Reference}. (Note that you need to compile and install FFTW before you can use it in a program. For the details of the installation, see @ref{Installation and Customization}.) We recommend that you read this tutorial in order.@footnote{You can read the tutorial in bit-reversed order after computing your first transform.} At the least, read the first section (@pxref{Complex One-Dimensional DFTs}) before reading any of the others, even if your main interest lies in one of the other transform types. Users of FFTW version 2 and earlier may also want to read @ref{Upgrading from FFTW version 2}. @c ------------------------------------------------------------ @node Complex One-Dimensional DFTs, Complex Multi-Dimensional DFTs, Tutorial, Tutorial @section Complex One-Dimensional DFTs @quotation Plan: To bother about the best method of accomplishing an accidental result. [Ambrose Bierce, @cite{The Enlarged Devil's Dictionary}.] @cindex Devil @end quotation @iftex @medskip @end iftex The basic usage of FFTW to compute a one-dimensional DFT of size @code{N} is simple, and it typically looks something like this code: @example #include ... @{ fftw_complex *in, *out; fftw_plan p; ... in = (fftw_complex*) fftw_malloc(sizeof(fftw_complex) * N); out = (fftw_complex*) fftw_malloc(sizeof(fftw_complex) * N); p = fftw_plan_dft_1d(N, in, out, FFTW_FORWARD, FFTW_ESTIMATE); ... fftw_execute(p); /* @r{repeat as needed} */ ... fftw_destroy_plan(p); fftw_free(in); fftw_free(out); @} @end example You must link this code with the @code{fftw3} library. On Unix systems, link with @code{-lfftw3 -lm}. The example code first allocates the input and output arrays. You can allocate them in any way that you like, but we recommend using @code{fftw_malloc}, which behaves like @findex fftw_malloc @code{malloc} except that it properly aligns the array when SIMD instructions (such as SSE and Altivec) are available (@pxref{SIMD alignment and fftw_malloc}). [Alternatively, we provide a convenient wrapper function @code{fftw_alloc_complex(N)} which has the same effect.] @findex fftw_alloc_complex @cindex SIMD The data is an array of type @code{fftw_complex}, which is by default a @code{double[2]} composed of the real (@code{in[i][0]}) and imaginary (@code{in[i][1]}) parts of a complex number. @tindex fftw_complex The next step is to create a @dfn{plan}, which is an object @cindex plan that contains all the data that FFTW needs to compute the FFT. This function creates the plan: @example fftw_plan fftw_plan_dft_1d(int n, fftw_complex *in, fftw_complex *out, int sign, unsigned flags); @end example @findex fftw_plan_dft_1d @tindex fftw_plan The first argument, @code{n}, is the size of the transform you are trying to compute. The size @code{n} can be any positive integer, but sizes that are products of small factors are transformed most efficiently (although prime sizes still use an @Onlogn{} algorithm). The next two arguments are pointers to the input and output arrays of the transform. These pointers can be equal, indicating an @dfn{in-place} transform. @cindex in-place The fourth argument, @code{sign}, can be either @code{FFTW_FORWARD} (@code{-1}) or @code{FFTW_BACKWARD} (@code{+1}), @ctindex FFTW_FORWARD @ctindex FFTW_BACKWARD and indicates the direction of the transform you are interested in; technically, it is the sign of the exponent in the transform. The @code{flags} argument is usually either @code{FFTW_MEASURE} or @cindex flags @code{FFTW_ESTIMATE}. @code{FFTW_MEASURE} instructs FFTW to run @ctindex FFTW_MEASURE and measure the execution time of several FFTs in order to find the best way to compute the transform of size @code{n}. This process takes some time (usually a few seconds), depending on your machine and on the size of the transform. @code{FFTW_ESTIMATE}, on the contrary, does not run any computation and just builds a @ctindex FFTW_ESTIMATE reasonable plan that is probably sub-optimal. In short, if your program performs many transforms of the same size and initialization time is not important, use @code{FFTW_MEASURE}; otherwise use the estimate. @emph{You must create the plan before initializing the input}, because @code{FFTW_MEASURE} overwrites the @code{in}/@code{out} arrays. (Technically, @code{FFTW_ESTIMATE} does not touch your arrays, but you should always create plans first just to be sure.) Once the plan has been created, you can use it as many times as you like for transforms on the specified @code{in}/@code{out} arrays, computing the actual transforms via @code{fftw_execute(plan)}: @example void fftw_execute(const fftw_plan plan); @end example @findex fftw_execute The DFT results are stored in-order in the array @code{out}, with the zero-frequency (DC) component in @code{out[0]}. @cindex frequency If @code{in != out}, the transform is @dfn{out-of-place} and the input array @code{in} is not modified. Otherwise, the input array is overwritten with the transform. @cindex execute If you want to transform a @emph{different} array of the same size, you can create a new plan with @code{fftw_plan_dft_1d} and FFTW automatically reuses the information from the previous plan, if possible. Alternatively, with the ``guru'' interface you can apply a given plan to a different array, if you are careful. @xref{FFTW Reference}. When you are done with the plan, you deallocate it by calling @code{fftw_destroy_plan(plan)}: @example void fftw_destroy_plan(fftw_plan plan); @end example @findex fftw_destroy_plan If you allocate an array with @code{fftw_malloc()} you must deallocate it with @code{fftw_free()}. Do not use @code{free()} or, heaven forbid, @code{delete}. @findex fftw_free FFTW computes an @emph{unnormalized} DFT. Thus, computing a forward followed by a backward transform (or vice versa) results in the original array scaled by @code{n}. For the definition of the DFT, see @ref{What FFTW Really Computes}. @cindex DFT @cindex normalization If you have a C compiler, such as @code{gcc}, that supports the C99 standard, and you @code{#include } @emph{before} @code{}, then @code{fftw_complex} is the native double-precision complex type and you can manipulate it with ordinary arithmetic. Otherwise, FFTW defines its own complex type, which is bit-compatible with the C99 complex type. @xref{Complex numbers}. (The C++ @code{} template class may also be usable via a typecast.) @cindex C++ To use single or long-double precision versions of FFTW, replace the @code{fftw_} prefix by @code{fftwf_} or @code{fftwl_} and link with @code{-lfftw3f} or @code{-lfftw3l}, but use the @emph{same} @code{} header file. @cindex precision Many more flags exist besides @code{FFTW_MEASURE} and @code{FFTW_ESTIMATE}. For example, use @code{FFTW_PATIENT} if you're willing to wait even longer for a possibly even faster plan (@pxref{FFTW Reference}). @ctindex FFTW_PATIENT You can also save plans for future use, as described by @ref{Words of Wisdom-Saving Plans}. @c ------------------------------------------------------------ @node Complex Multi-Dimensional DFTs, One-Dimensional DFTs of Real Data, Complex One-Dimensional DFTs, Tutorial @section Complex Multi-Dimensional DFTs Multi-dimensional transforms work much the same way as one-dimensional transforms: you allocate arrays of @code{fftw_complex} (preferably using @code{fftw_malloc}), create an @code{fftw_plan}, execute it as many times as you want with @code{fftw_execute(plan)}, and clean up with @code{fftw_destroy_plan(plan)} (and @code{fftw_free}). FFTW provides two routines for creating plans for 2d and 3d transforms, and one routine for creating plans of arbitrary dimensionality. The 2d and 3d routines have the following signature: @example fftw_plan fftw_plan_dft_2d(int n0, int n1, fftw_complex *in, fftw_complex *out, int sign, unsigned flags); fftw_plan fftw_plan_dft_3d(int n0, int n1, int n2, fftw_complex *in, fftw_complex *out, int sign, unsigned flags); @end example @findex fftw_plan_dft_2d @findex fftw_plan_dft_3d These routines create plans for @code{n0} by @code{n1} two-dimensional (2d) transforms and @code{n0} by @code{n1} by @code{n2} 3d transforms, respectively. All of these transforms operate on contiguous arrays in the C-standard @dfn{row-major} order, so that the last dimension has the fastest-varying index in the array. This layout is described further in @ref{Multi-dimensional Array Format}. FFTW can also compute transforms of higher dimensionality. In order to avoid confusion between the various meanings of the the word ``dimension'', we use the term @emph{rank} @cindex rank to denote the number of independent indices in an array.@footnote{The term ``rank'' is commonly used in the APL, FORTRAN, and Common Lisp traditions, although it is not so common in the C@tie{}world.} For example, we say that a 2d transform has rank@tie{}2, a 3d transform has rank@tie{}3, and so on. You can plan transforms of arbitrary rank by means of the following function: @example fftw_plan fftw_plan_dft(int rank, const int *n, fftw_complex *in, fftw_complex *out, int sign, unsigned flags); @end example @findex fftw_plan_dft Here, @code{n} is a pointer to an array @code{n[rank]} denoting an @code{n[0]} by @code{n[1]} by @dots{} by @code{n[rank-1]} transform. Thus, for example, the call @example fftw_plan_dft_2d(n0, n1, in, out, sign, flags); @end example is equivalent to the following code fragment: @example int n[2]; n[0] = n0; n[1] = n1; fftw_plan_dft(2, n, in, out, sign, flags); @end example @code{fftw_plan_dft} is not restricted to 2d and 3d transforms, however, but it can plan transforms of arbitrary rank. You may have noticed that all the planner routines described so far have overlapping functionality. For example, you can plan a 1d or 2d transform by using @code{fftw_plan_dft} with a @code{rank} of @code{1} or @code{2}, or even by calling @code{fftw_plan_dft_3d} with @code{n0} and/or @code{n1} equal to @code{1} (with no loss in efficiency). This pattern continues, and FFTW's planning routines in general form a ``partial order,'' sequences of @cindex partial order interfaces with strictly increasing generality but correspondingly greater complexity. @code{fftw_plan_dft} is the most general complex-DFT routine that we describe in this tutorial, but there are also the advanced and guru interfaces, @cindex advanced interface @cindex guru interface which allow one to efficiently combine multiple/strided transforms into a single FFTW plan, transform a subset of a larger multi-dimensional array, and/or to handle more general complex-number formats. For more information, see @ref{FFTW Reference}. @c ------------------------------------------------------------ @node One-Dimensional DFTs of Real Data, Multi-Dimensional DFTs of Real Data, Complex Multi-Dimensional DFTs, Tutorial @section One-Dimensional DFTs of Real Data In many practical applications, the input data @code{in[i]} are purely real numbers, in which case the DFT output satisfies the ``Hermitian'' @cindex Hermitian redundancy: @code{out[i]} is the conjugate of @code{out[n-i]}. It is possible to take advantage of these circumstances in order to achieve roughly a factor of two improvement in both speed and memory usage. In exchange for these speed and space advantages, the user sacrifices some of the simplicity of FFTW's complex transforms. First of all, the input and output arrays are of @emph{different sizes and types}: the input is @code{n} real numbers, while the output is @code{n/2+1} complex numbers (the non-redundant outputs); this also requires slight ``padding'' of the input array for @cindex padding in-place transforms. Second, the inverse transform (complex to real) has the side-effect of @emph{overwriting its input array}, by default. Neither of these inconveniences should pose a serious problem for users, but it is important to be aware of them. The routines to perform real-data transforms are almost the same as those for complex transforms: you allocate arrays of @code{double} and/or @code{fftw_complex} (preferably using @code{fftw_malloc} or @code{fftw_alloc_complex}), create an @code{fftw_plan}, execute it as many times as you want with @code{fftw_execute(plan)}, and clean up with @code{fftw_destroy_plan(plan)} (and @code{fftw_free}). The only differences are that the input (or output) is of type @code{double} and there are new routines to create the plan. In one dimension: @example fftw_plan fftw_plan_dft_r2c_1d(int n, double *in, fftw_complex *out, unsigned flags); fftw_plan fftw_plan_dft_c2r_1d(int n, fftw_complex *in, double *out, unsigned flags); @end example @findex fftw_plan_dft_r2c_1d @findex fftw_plan_dft_c2r_1d for the real input to complex-Hermitian output (@dfn{r2c}) and complex-Hermitian input to real output (@dfn{c2r}) transforms. @cindex r2c @cindex c2r Unlike the complex DFT planner, there is no @code{sign} argument. Instead, r2c DFTs are always @code{FFTW_FORWARD} and c2r DFTs are always @code{FFTW_BACKWARD}. @ctindex FFTW_FORWARD @ctindex FFTW_BACKWARD (For single/long-double precision @code{fftwf} and @code{fftwl}, @code{double} should be replaced by @code{float} and @code{long double}, respectively.) @cindex precision Here, @code{n} is the ``logical'' size of the DFT, not necessarily the physical size of the array. In particular, the real (@code{double}) array has @code{n} elements, while the complex (@code{fftw_complex}) array has @code{n/2+1} elements (where the division is rounded down). For an in-place transform, @cindex in-place @code{in} and @code{out} are aliased to the same array, which must be big enough to hold both; so, the real array would actually have @code{2*(n/2+1)} elements, where the elements beyond the first @code{n} are unused padding. (Note that this is very different from the concept of ``zero-padding'' a transform to a larger length, which changes the logical size of the DFT by actually adding new input data.) The @math{k}th element of the complex array is exactly the same as the @math{k}th element of the corresponding complex DFT. All positive @code{n} are supported; products of small factors are most efficient, but an @Onlogn algorithm is used even for prime sizes. As noted above, the c2r transform destroys its input array even for out-of-place transforms. This can be prevented, if necessary, by including @code{FFTW_PRESERVE_INPUT} in the @code{flags}, with unfortunately some sacrifice in performance. @cindex flags @ctindex FFTW_PRESERVE_INPUT This flag is also not currently supported for multi-dimensional real DFTs (next section). Readers familiar with DFTs of real data will recall that the 0th (the ``DC'') and @code{n/2}-th (the ``Nyquist'' frequency, when @code{n} is even) elements of the complex output are purely real. Some implementations therefore store the Nyquist element where the DC imaginary part would go, in order to make the input and output arrays the same size. Such packing, however, does not generalize well to multi-dimensional transforms, and the space savings are miniscule in any case; FFTW does not support it. An alternative interface for one-dimensional r2c and c2r DFTs can be found in the @samp{r2r} interface (@pxref{The Halfcomplex-format DFT}), with ``halfcomplex''-format output that @emph{is} the same size (and type) as the input array. @cindex halfcomplex format That interface, although it is not very useful for multi-dimensional transforms, may sometimes yield better performance. @c ------------------------------------------------------------ @node Multi-Dimensional DFTs of Real Data, More DFTs of Real Data, One-Dimensional DFTs of Real Data, Tutorial @section Multi-Dimensional DFTs of Real Data Multi-dimensional DFTs of real data use the following planner routines: @example fftw_plan fftw_plan_dft_r2c_2d(int n0, int n1, double *in, fftw_complex *out, unsigned flags); fftw_plan fftw_plan_dft_r2c_3d(int n0, int n1, int n2, double *in, fftw_complex *out, unsigned flags); fftw_plan fftw_plan_dft_r2c(int rank, const int *n, double *in, fftw_complex *out, unsigned flags); @end example @findex fftw_plan_dft_r2c_2d @findex fftw_plan_dft_r2c_3d @findex fftw_plan_dft_r2c as well as the corresponding @code{c2r} routines with the input/output types swapped. These routines work similarly to their complex analogues, except for the fact that here the complex output array is cut roughly in half and the real array requires padding for in-place transforms (as in 1d, above). As before, @code{n} is the logical size of the array, and the consequences of this on the the format of the complex arrays deserve careful attention. @cindex r2c/c2r multi-dimensional array format Suppose that the real data has dimensions @ndims (in row-major order). Then, after an r2c transform, the output is an @ndimshalf array of @code{fftw_complex} values in row-major order, corresponding to slightly over half of the output of the corresponding complex DFT. (The division is rounded down.) The ordering of the data is otherwise exactly the same as in the complex-DFT case. For out-of-place transforms, this is the end of the story: the real data is stored as a row-major array of size @ndims and the complex data is stored as a row-major array of size @ndimshalf{}. For in-place transforms, however, extra padding of the real-data array is necessary because the complex array is larger than the real array, and the two arrays share the same memory locations. Thus, for in-place transforms, the final dimension of the real-data array must be padded with extra values to accommodate the size of the complex data---two values if the last dimension is even and one if it is odd. @cindex padding That is, the last dimension of the real data must physically contain @tex $2 (n_{d-1}/2+1)$ @end tex @ifinfo 2 * (n[d-1]/2+1) @end ifinfo @html 2 * (nd-1/2+1) @end html @code{double} values (exactly enough to hold the complex data). This physical array size does not, however, change the @emph{logical} array size---only @tex $n_{d-1}$ @end tex @ifinfo n[d-1] @end ifinfo @html nd-1 @end html values are actually stored in the last dimension, and @tex $n_{d-1}$ @end tex @ifinfo n[d-1] @end ifinfo @html nd-1 @end html is the last dimension passed to the plan-creation routine. For example, consider the transform of a two-dimensional real array of size @code{n0} by @code{n1}. The output of the r2c transform is a two-dimensional complex array of size @code{n0} by @code{n1/2+1}, where the @code{y} dimension has been cut nearly in half because of redundancies in the output. Because @code{fftw_complex} is twice the size of @code{double}, the output array is slightly bigger than the input array. Thus, if we want to compute the transform in place, we must @emph{pad} the input array so that it is of size @code{n0} by @code{2*(n1/2+1)}. If @code{n1} is even, then there are two padding elements at the end of each row (which need not be initialized, as they are only used for output). @ifhtml The following illustration depicts the input and output arrays just described, for both the out-of-place and in-place transforms (with the arrows indicating consecutive memory locations): @image{rfftwnd-for-html} @end ifhtml @ifnotinfo @ifnothtml @float Figure,fig:rfftwnd @center @image{rfftwnd} @caption{Illustration of the data layout for a 2d @code{nx} by @code{ny} real-to-complex transform.} @end float @ref{fig:rfftwnd} depicts the input and output arrays just described, for both the out-of-place and in-place transforms (with the arrows indicating consecutive memory locations): @end ifnothtml @end ifnotinfo These transforms are unnormalized, so an r2c followed by a c2r transform (or vice versa) will result in the original data scaled by the number of real data elements---that is, the product of the (logical) dimensions of the real data. @cindex normalization (Because the last dimension is treated specially, if it is equal to @code{1} the transform is @emph{not} equivalent to a lower-dimensional r2c/c2r transform. In that case, the last complex dimension also has size @code{1} (@code{=1/2+1}), and no advantage is gained over the complex transforms.) @c ------------------------------------------------------------ @node More DFTs of Real Data, , Multi-Dimensional DFTs of Real Data, Tutorial @section More DFTs of Real Data @menu * The Halfcomplex-format DFT:: * Real even/odd DFTs (cosine/sine transforms):: * The Discrete Hartley Transform:: @end menu FFTW supports several other transform types via a unified @dfn{r2r} (real-to-real) interface, @cindex r2r so called because it takes a real (@code{double}) array and outputs a real array of the same size. These r2r transforms currently fall into three categories: DFTs of real input and complex-Hermitian output in halfcomplex format, DFTs of real input with even/odd symmetry (a.k.a. discrete cosine/sine transforms, DCTs/DSTs), and discrete Hartley transforms (DHTs), all described in more detail by the following sections. The r2r transforms follow the by now familiar interface of creating an @code{fftw_plan}, executing it with @code{fftw_execute(plan)}, and destroying it with @code{fftw_destroy_plan(plan)}. Furthermore, all r2r transforms share the same planner interface: @example fftw_plan fftw_plan_r2r_1d(int n, double *in, double *out, fftw_r2r_kind kind, unsigned flags); fftw_plan fftw_plan_r2r_2d(int n0, int n1, double *in, double *out, fftw_r2r_kind kind0, fftw_r2r_kind kind1, unsigned flags); fftw_plan fftw_plan_r2r_3d(int n0, int n1, int n2, double *in, double *out, fftw_r2r_kind kind0, fftw_r2r_kind kind1, fftw_r2r_kind kind2, unsigned flags); fftw_plan fftw_plan_r2r(int rank, const int *n, double *in, double *out, const fftw_r2r_kind *kind, unsigned flags); @end example @findex fftw_plan_r2r_1d @findex fftw_plan_r2r_2d @findex fftw_plan_r2r_3d @findex fftw_plan_r2r Just as for the complex DFT, these plan 1d/2d/3d/multi-dimensional transforms for contiguous arrays in row-major order, transforming (real) input to output of the same size, where @code{n} specifies the @emph{physical} dimensions of the arrays. All positive @code{n} are supported (with the exception of @code{n=1} for the @code{FFTW_REDFT00} kind, noted in the real-even subsection below); products of small factors are most efficient (factorizing @code{n-1} and @code{n+1} for @code{FFTW_REDFT00} and @code{FFTW_RODFT00} kinds, described below), but an @Onlogn algorithm is used even for prime sizes. Each dimension has a @dfn{kind} parameter, of type @code{fftw_r2r_kind}, specifying the kind of r2r transform to be used for that dimension. @cindex kind (r2r) @tindex fftw_r2r_kind (In the case of @code{fftw_plan_r2r}, this is an array @code{kind[rank]} where @code{kind[i]} is the transform kind for the dimension @code{n[i]}.) The kind can be one of a set of predefined constants, defined in the following subsections. In other words, FFTW computes the separable product of the specified r2r transforms over each dimension, which can be used e.g. for partial differential equations with mixed boundary conditions. (For some r2r kinds, notably the halfcomplex DFT and the DHT, such a separable product is somewhat problematic in more than one dimension, however, as is described below.) In the current version of FFTW, all r2r transforms except for the halfcomplex type are computed via pre- or post-processing of halfcomplex transforms, and they are therefore not as fast as they could be. Since most other general DCT/DST codes employ a similar algorithm, however, FFTW's implementation should provide at least competitive performance. @c =========> @node The Halfcomplex-format DFT, Real even/odd DFTs (cosine/sine transforms), More DFTs of Real Data, More DFTs of Real Data @subsection The Halfcomplex-format DFT An r2r kind of @code{FFTW_R2HC} (@dfn{r2hc}) corresponds to an r2c DFT @ctindex FFTW_R2HC @cindex r2c @cindex r2hc (@pxref{One-Dimensional DFTs of Real Data}) but with ``halfcomplex'' format output, and may sometimes be faster and/or more convenient than the latter. @cindex halfcomplex format The inverse @dfn{hc2r} transform is of kind @code{FFTW_HC2R}. @ctindex FFTW_HC2R @cindex hc2r This consists of the non-redundant half of the complex output for a 1d real-input DFT of size @code{n}, stored as a sequence of @code{n} real numbers (@code{double}) in the format: @tex $$ r_0, r_1, r_2, \ldots, r_{n/2}, i_{(n+1)/2-1}, \ldots, i_2, i_1 $$ @end tex @ifinfo r0, r1, r2, r(n/2), i((n+1)/2-1), ..., i2, i1 @end ifinfo @html

r0, r1, r2, ..., rn/2, i(n+1)/2-1, ..., i2, i1

@end html Here, @ifinfo rk @end ifinfo @tex $r_k$ @end tex @html rk @end html is the real part of the @math{k}th output, and @ifinfo ik @end ifinfo @tex $i_k$ @end tex @html ik @end html is the imaginary part. (Division by 2 is rounded down.) For a halfcomplex array @code{hc[n]}, the @math{k}th component thus has its real part in @code{hc[k]} and its imaginary part in @code{hc[n-k]}, with the exception of @code{k} @code{==} @code{0} or @code{n/2} (the latter only if @code{n} is even)---in these two cases, the imaginary part is zero due to symmetries of the real-input DFT, and is not stored. Thus, the r2hc transform of @code{n} real values is a halfcomplex array of length @code{n}, and vice versa for hc2r. @cindex normalization Aside from the differing format, the output of @code{FFTW_R2HC}/@code{FFTW_HC2R} is otherwise exactly the same as for the corresponding 1d r2c/c2r transform (i.e. @code{FFTW_FORWARD}/@code{FFTW_BACKWARD} transforms, respectively). Recall that these transforms are unnormalized, so r2hc followed by hc2r will result in the original data multiplied by @code{n}. Furthermore, like the c2r transform, an out-of-place hc2r transform will @emph{destroy its input} array. Although these halfcomplex transforms can be used with the multi-dimensional r2r interface, the interpretation of such a separable product of transforms along each dimension is problematic. For example, consider a two-dimensional @code{n0} by @code{n1}, r2hc by r2hc transform planned by @code{fftw_plan_r2r_2d(n0, n1, in, out, FFTW_R2HC, FFTW_R2HC, FFTW_MEASURE)}. Conceptually, FFTW first transforms the rows (of size @code{n1}) to produce halfcomplex rows, and then transforms the columns (of size @code{n0}). Half of these column transforms, however, are of imaginary parts, and should therefore be multiplied by @math{i} and combined with the r2hc transforms of the real columns to produce the 2d DFT amplitudes; FFTW's r2r transform does @emph{not} perform this combination for you. Thus, if a multi-dimensional real-input/output DFT is required, we recommend using the ordinary r2c/c2r interface (@pxref{Multi-Dimensional DFTs of Real Data}). @c =========> @node Real even/odd DFTs (cosine/sine transforms), The Discrete Hartley Transform, The Halfcomplex-format DFT, More DFTs of Real Data @subsection Real even/odd DFTs (cosine/sine transforms) The Fourier transform of a real-even function @math{f(-x) = f(x)} is real-even, and @math{i} times the Fourier transform of a real-odd function @math{f(-x) = -f(x)} is real-odd. Similar results hold for a discrete Fourier transform, and thus for these symmetries the need for complex inputs/outputs is entirely eliminated. Moreover, one gains a factor of two in speed/space from the fact that the data are real, and an additional factor of two from the even/odd symmetry: only the non-redundant (first) half of the array need be stored. The result is the real-even DFT (@dfn{REDFT}) and the real-odd DFT (@dfn{RODFT}), also known as the discrete cosine and sine transforms (@dfn{DCT} and @dfn{DST}), respectively. @cindex real-even DFT @cindex REDFT @cindex real-odd DFT @cindex RODFT @cindex discrete cosine transform @cindex DCT @cindex discrete sine transform @cindex DST (In this section, we describe the 1d transforms; multi-dimensional transforms are just a separable product of these transforms operating along each dimension.) Because of the discrete sampling, one has an additional choice: is the data even/odd around a sampling point, or around the point halfway between two samples? The latter corresponds to @emph{shifting} the samples by @emph{half} an interval, and gives rise to several transform variants denoted by REDFT@math{ab} and RODFT@math{ab}: @math{a} and @math{b} are @math{0} or @math{1}, and indicate whether the input (@math{a}) and/or output (@math{b}) are shifted by half a sample (@math{1} means it is shifted). These are also known as types I-IV of the DCT and DST, and all four types are supported by FFTW's r2r interface.@footnote{There are also type V-VIII transforms, which correspond to a logical DFT of @emph{odd} size @math{N}, independent of whether the physical size @code{n} is odd, but we do not support these variants.} The r2r kinds for the various REDFT and RODFT types supported by FFTW, along with the boundary conditions at both ends of the @emph{input} array (@code{n} real numbers @code{in[j=0..n-1]}), are: @itemize @bullet @item @code{FFTW_REDFT00} (DCT-I): even around @math{j=0} and even around @math{j=n-1}. @ctindex FFTW_REDFT00 @item @code{FFTW_REDFT10} (DCT-II, ``the'' DCT): even around @math{j=-0.5} and even around @math{j=n-0.5}. @ctindex FFTW_REDFT10 @item @code{FFTW_REDFT01} (DCT-III, ``the'' IDCT): even around @math{j=0} and odd around @math{j=n}. @ctindex FFTW_REDFT01 @cindex IDCT @item @code{FFTW_REDFT11} (DCT-IV): even around @math{j=-0.5} and odd around @math{j=n-0.5}. @ctindex FFTW_REDFT11 @item @code{FFTW_RODFT00} (DST-I): odd around @math{j=-1} and odd around @math{j=n}. @ctindex FFTW_RODFT00 @item @code{FFTW_RODFT10} (DST-II): odd around @math{j=-0.5} and odd around @math{j=n-0.5}. @ctindex FFTW_RODFT10 @item @code{FFTW_RODFT01} (DST-III): odd around @math{j=-1} and even around @math{j=n-1}. @ctindex FFTW_RODFT01 @item @code{FFTW_RODFT11} (DST-IV): odd around @math{j=-0.5} and even around @math{j=n-0.5}. @ctindex FFTW_RODFT11 @end itemize Note that these symmetries apply to the ``logical'' array being transformed; @strong{there are no constraints on your physical input data}. So, for example, if you specify a size-5 REDFT00 (DCT-I) of the data @math{abcde}, it corresponds to the DFT of the logical even array @math{abcdedcb} of size 8. A size-4 REDFT10 (DCT-II) of the data @math{abcd} corresponds to the size-8 logical DFT of the even array @math{abcddcba}, shifted by half a sample. All of these transforms are invertible. The inverse of R*DFT00 is R*DFT00; of R*DFT10 is R*DFT01 and vice versa (these are often called simply ``the'' DCT and IDCT, respectively); and of R*DFT11 is R*DFT11. However, the transforms computed by FFTW are unnormalized, exactly like the corresponding real and complex DFTs, so computing a transform followed by its inverse yields the original array scaled by @math{N}, where @math{N} is the @emph{logical} DFT size. For REDFT00, @math{N=2(n-1)}; for RODFT00, @math{N=2(n+1)}; otherwise, @math{N=2n}. @cindex normalization @cindex IDCT Note that the boundary conditions of the transform output array are given by the input boundary conditions of the inverse transform. Thus, the above transforms are all inequivalent in terms of input/output boundary conditions, even neglecting the 0.5 shift difference. FFTW is most efficient when @math{N} is a product of small factors; note that this @emph{differs} from the factorization of the physical size @code{n} for REDFT00 and RODFT00! There is another oddity: @code{n=1} REDFT00 transforms correspond to @math{N=0}, and so are @emph{not defined} (the planner will return @code{NULL}). Otherwise, any positive @code{n} is supported. For the precise mathematical definitions of these transforms as used by FFTW, see @ref{What FFTW Really Computes}. (For people accustomed to the DCT/DST, FFTW's definitions have a coefficient of @math{2} in front of the cos/sin functions so that they correspond precisely to an even/odd DFT of size @math{N}. Some authors also include additional multiplicative factors of @ifinfo sqrt(2) @end ifinfo @html √2 @end html @tex $\sqrt{2}$ @end tex for selected inputs and outputs; this makes the transform orthogonal, but sacrifices the direct equivalence to a symmetric DFT.) @subsubheading Which type do you need? Since the required flavor of even/odd DFT depends upon your problem, you are the best judge of this choice, but we can make a few comments on relative efficiency to help you in your selection. In particular, R*DFT01 and R*DFT10 tend to be slightly faster than R*DFT11 (especially for odd sizes), while the R*DFT00 transforms are sometimes significantly slower (especially for even sizes).@footnote{R*DFT00 is sometimes slower in FFTW because we discovered that the standard algorithm for computing this by a pre/post-processed real DFT---the algorithm used in FFTPACK, Numerical Recipes, and other sources for decades now---has serious numerical problems: it already loses several decimal places of accuracy for 16k sizes. There seem to be only two alternatives in the literature that do not suffer similarly: a recursive decomposition into smaller DCTs, which would require a large set of codelets for efficiency and generality, or sacrificing a factor of @tex $\sim 2$ @end tex @ifnottex 2 @end ifnottex in speed to use a real DFT of twice the size. We currently employ the latter technique for general @math{n}, as well as a limited form of the former method: a split-radix decomposition when @math{n} is odd (@math{N} a multiple of 4). For @math{N} containing many factors of 2, the split-radix method seems to recover most of the speed of the standard algorithm without the accuracy tradeoff.} Thus, if only the boundary conditions on the transform inputs are specified, we generally recommend R*DFT10 over R*DFT00 and R*DFT01 over R*DFT11 (unless the half-sample shift or the self-inverse property is significant for your problem). If performance is important to you and you are using only small sizes (say @math{n<200}), e.g. for multi-dimensional transforms, then you might consider generating hard-coded transforms of those sizes and types that you are interested in (@pxref{Generating your own code}). We are interested in hearing what types of symmetric transforms you find most useful. @c =========> @node The Discrete Hartley Transform, , Real even/odd DFTs (cosine/sine transforms), More DFTs of Real Data @subsection The Discrete Hartley Transform If you are planning to use the DHT because you've heard that it is ``faster'' than the DFT (FFT), @strong{stop here}. The DHT is not faster than the DFT. That story is an old but enduring misconception that was debunked in 1987. The discrete Hartley transform (DHT) is an invertible linear transform closely related to the DFT. In the DFT, one multiplies each input by @math{cos - i * sin} (a complex exponential), whereas in the DHT each input is multiplied by simply @math{cos + sin}. Thus, the DHT transforms @code{n} real numbers to @code{n} real numbers, and has the convenient property of being its own inverse. In FFTW, a DHT (of any positive @code{n}) can be specified by an r2r kind of @code{FFTW_DHT}. @ctindex FFTW_DHT @cindex discrete Hartley transform @cindex DHT Like the DFT, in FFTW the DHT is unnormalized, so computing a DHT of size @code{n} followed by another DHT of the same size will result in the original array multiplied by @code{n}. @cindex normalization The DHT was originally proposed as a more efficient alternative to the DFT for real data, but it was subsequently shown that a specialized DFT (such as FFTW's r2hc or r2c transforms) could be just as fast. In FFTW, the DHT is actually computed by post-processing an r2hc transform, so there is ordinarily no reason to prefer it from a performance perspective.@footnote{We provide the DHT mainly as a byproduct of some internal algorithms. FFTW computes a real input/output DFT of @emph{prime} size by re-expressing it as a DHT plus post/pre-processing and then using Rader's prime-DFT algorithm adapted to the DHT.} However, we have heard rumors that the DHT might be the most appropriate transform in its own right for certain applications, and we would be very interested to hear from anyone who finds it useful. If @code{FFTW_DHT} is specified for multiple dimensions of a multi-dimensional transform, FFTW computes the separable product of 1d DHTs along each dimension. Unfortunately, this is not quite the same thing as a true multi-dimensional DHT; you can compute the latter, if necessary, with at most @code{rank-1} post-processing passes [see e.g. H. Hao and R. N. Bracewell, @i{Proc. IEEE} @b{75}, 264--266 (1987)]. For the precise mathematical definition of the DHT as used by FFTW, see @ref{What FFTW Really Computes}. fftw-3.3.3/doc/upgrading.texi0000644000175000017500000002364512054410050013030 00000000000000@node Upgrading from FFTW version 2, Installation and Customization, Calling FFTW from Legacy Fortran, Top @chapter Upgrading from FFTW version 2 In this chapter, we outline the process for updating codes designed for the older FFTW 2 interface to work with FFTW 3. The interface for FFTW 3 is not backwards-compatible with the interface for FFTW 2 and earlier versions; codes written to use those versions will fail to link with FFTW 3. Nor is it possible to write ``compatibility wrappers'' to bridge the gap (at least not efficiently), because FFTW 3 has different semantics from previous versions. However, upgrading should be a straightforward process because the data formats are identical and the overall style of planning/execution is essentially the same. Unlike FFTW 2, there are no separate header files for real and complex transforms (or even for different precisions) in FFTW 3; all interfaces are defined in the @code{} header file. @heading Numeric Types The main difference in data types is that @code{fftw_complex} in FFTW 2 was defined as a @code{struct} with macros @code{c_re} and @code{c_im} for accessing the real/imaginary parts. (This is binary-compatible with FFTW 3 on any machine except perhaps for some older Crays in single precision.) The equivalent macros for FFTW 3 are: @example #define c_re(c) ((c)[0]) #define c_im(c) ((c)[1]) @end example This does not work if you are using the C99 complex type, however, unless you insert a @code{double*} typecast into the above macros (@pxref{Complex numbers}). Also, FFTW 2 had an @code{fftw_real} typedef that was an alias for @code{double} (in double precision). In FFTW 3 you should just use @code{double} (or whatever precision you are employing). @heading Plans The major difference between FFTW 2 and FFTW 3 is in the planning/execution division of labor. In FFTW 2, plans were found for a given transform size and type, and then could be applied to @emph{any} arrays and for @emph{any} multiplicity/stride parameters. In FFTW 3, you specify the particular arrays, stride parameters, etcetera when creating the plan, and the plan is then executed for @emph{those} arrays (unless the guru interface is used) and @emph{those} parameters @emph{only}. (FFTW 2 had ``specific planner'' routines that planned for a particular array and stride, but the plan could still be used for other arrays and strides.) That is, much of the information that was formerly specified at execution time is now specified at planning time. Like FFTW 2's specific planner routines, the FFTW 3 planner overwrites the input/output arrays unless you use @code{FFTW_ESTIMATE}. FFTW 2 had separate data types @code{fftw_plan}, @code{fftwnd_plan}, @code{rfftw_plan}, and @code{rfftwnd_plan} for complex and real one- and multi-dimensional transforms, and each type had its own @samp{destroy} function. In FFTW 3, all plans are of type @code{fftw_plan} and all are destroyed by @code{fftw_destroy_plan(plan)}. Where you formerly used @code{fftw_create_plan} and @code{fftw_one} to plan and compute a single 1d transform, you would now use @code{fftw_plan_dft_1d} to plan the transform. If you used the generic @code{fftw} function to execute the transform with multiplicity (@code{howmany}) and stride parameters, you would now use the advanced interface @code{fftw_plan_many_dft} to specify those parameters. The plans are now executed with @code{fftw_execute(plan)}, which takes all of its parameters (including the input/output arrays) from the plan. In-place transforms no longer interpret their output argument as scratch space, nor is there an @code{FFTW_IN_PLACE} flag. You simply pass the same pointer for both the input and output arguments. (Previously, the output @code{ostride} and @code{odist} parameters were ignored for in-place transforms; now, if they are specified via the advanced interface, they are significant even in the in-place case, although they should normally equal the corresponding input parameters.) The @code{FFTW_ESTIMATE} and @code{FFTW_MEASURE} flags have the same meaning as before, although the planning time will differ. You may also consider using @code{FFTW_PATIENT}, which is like @code{FFTW_MEASURE} except that it takes more time in order to consider a wider variety of algorithms. For multi-dimensional complex DFTs, instead of @code{fftwnd_create_plan} (or @code{fftw2d_create_plan} or @code{fftw3d_create_plan}), followed by @code{fftwnd_one}, you would use @code{fftw_plan_dft} (or @code{fftw_plan_dft_2d} or @code{fftw_plan_dft_3d}). followed by @code{fftw_execute}. If you used @code{fftwnd} to to specify strides etcetera, you would instead specify these via @code{fftw_plan_many_dft}. The analogues to @code{rfftw_create_plan} and @code{rfftw_one} with @code{FFTW_REAL_TO_COMPLEX} or @code{FFTW_COMPLEX_TO_REAL} directions are @code{fftw_plan_r2r_1d} with kind @code{FFTW_R2HC} or @code{FFTW_HC2R}, followed by @code{fftw_execute}. The stride etcetera arguments of @code{rfftw} are now in @code{fftw_plan_many_r2r}. Instead of @code{rfftwnd_create_plan} (or @code{rfftw2d_create_plan} or @code{rfftw3d_create_plan}) followed by @code{rfftwnd_one_real_to_complex} or @code{rfftwnd_one_complex_to_real}, you now use @code{fftw_plan_dft_r2c} (or @code{fftw_plan_dft_r2c_2d} or @code{fftw_plan_dft_r2c_3d}) or @code{fftw_plan_dft_c2r} (or @code{fftw_plan_dft_c2r_2d} or @code{fftw_plan_dft_c2r_3d}), respectively, followed by @code{fftw_execute}. As usual, the strides etcetera of @code{rfftwnd_real_to_complex} or @code{rfftwnd_complex_to_real} are no specified in the advanced planner routines, @code{fftw_plan_many_dft_r2c} or @code{fftw_plan_many_dft_c2r}. @heading Wisdom In FFTW 2, you had to supply the @code{FFTW_USE_WISDOM} flag in order to use wisdom; in FFTW 3, wisdom is always used. (You could simulate the FFTW 2 wisdom-less behavior by calling @code{fftw_forget_wisdom} after every planner call.) The FFTW 3 wisdom import/export routines are almost the same as before (although the storage format is entirely different). There is one significant difference, however. In FFTW 2, the import routines would never read past the end of the wisdom, so you could store extra data beyond the wisdom in the same file, for example. In FFTW 3, the file-import routine may read up to a few hundred bytes past the end of the wisdom, so you cannot store other data just beyond it.@footnote{We do our own buffering because GNU libc I/O routines are horribly slow for single-character I/O, apparently for thread-safety reasons (whether you are using threads or not).} Wisdom has been enhanced by additional humility in FFTW 3: whereas FFTW 2 would re-use wisdom for a given transform size regardless of the stride etc., in FFTW 3 wisdom is only used with the strides etc. for which it was created. Unfortunately, this means FFTW 3 has to create new plans from scratch more often than FFTW 2 (in FFTW 2, planning e.g. one transform of size 1024 also created wisdom for all smaller powers of 2, but this no longer occurs). FFTW 3 also has the new routine @code{fftw_import_system_wisdom} to import wisdom from a standard system-wide location. @heading Memory allocation In FFTW 3, we recommend allocating your arrays with @code{fftw_malloc} and deallocating them with @code{fftw_free}; this is not required, but allows optimal performance when SIMD acceleration is used. (Those two functions actually existed in FFTW 2, and worked the same way, but were not documented.) In FFTW 2, there were @code{fftw_malloc_hook} and @code{fftw_free_hook} functions that allowed the user to replace FFTW's memory-allocation routines (e.g. to implement different error-handling, since by default FFTW prints an error message and calls @code{exit} to abort the program if @code{malloc} returns @code{NULL}). These hooks are not supported in FFTW 3; those few users who require this functionality can just directly modify the memory-allocation routines in FFTW (they are defined in @code{kernel/alloc.c}). @heading Fortran interface In FFTW 2, the subroutine names were obtained by replacing @samp{fftw_} with @samp{fftw_f77}; in FFTW 3, you replace @samp{fftw_} with @samp{dfftw_} (or @samp{sfftw_} or @samp{lfftw_}, depending upon the precision). In FFTW 3, we have begun recommending that you always declare the type used to store plans as @code{integer*8}. (Too many people didn't notice our instruction to switch from @code{integer} to @code{integer*8} for 64-bit machines.) In FFTW 3, we provide a @code{fftw3.f} ``header file'' to include in your code (and which is officially installed on Unix systems). (In FFTW 2, we supplied a @code{fftw_f77.i} file, but it was not installed.) Otherwise, the C-Fortran interface relationship is much the same as it was before (e.g. return values become initial parameters, and multi-dimensional arrays are in column-major order). Unlike FFTW 2, we do provide some support for wisdom import/export in Fortran (@pxref{Wisdom of Fortran?}). @heading Threads Like FFTW 2, only the execution routines are thread-safe. All planner routines, etcetera, should be called by only a single thread at a time (@pxref{Thread safety}). @emph{Unlike} FFTW 2, there is no special @code{FFTW_THREADSAFE} flag for the planner to allow a given plan to be usable by multiple threads in parallel; this is now the case by default. The multi-threaded version of FFTW 2 required you to pass the number of threads each time you execute the transform. The number of threads is now stored in the plan, and is specified before the planner is called by @code{fftw_plan_with_nthreads}. The threads initialization routine used to be called @code{fftw_threads_init} and would return zero on success; the new routine is called @code{fftw_init_threads} and returns zero on failure. @xref{Multi-threaded FFTW}. There is no separate threads header file in FFTW 3; all the function prototypes are in @code{}. However, you still have to link to a separate library (@code{-lfftw3_threads -lfftw3 -lm} on Unix), as well as to the threading library (e.g. POSIX threads on Unix). fftw-3.3.3/doc/version.texi0000644000175000017500000000014512054411104012524 00000000000000@set UPDATED 25 November 2012 @set UPDATED-MONTH November 2012 @set EDITION 3.3.3 @set VERSION 3.3.3 fftw-3.3.3/doc/rfftwnd.pdf0000644000175000017500000003017612054411105012321 00000000000000%PDF-1.4 %Çì¢ 5 0 obj <> stream xœµ\Ë®ä¶Ý÷Wô&€çÊâK¢Ù²Žï, Ží1ÌËÎ bÿ}Šd=¥Û’ºÇÏeñÏ!YâC"ûçë8¸ëXþã¿{{ùêëùúã.îZþûåÇKÌóý5cöc¸z——!ç«Çiþú˗׺„Á¥ë/ãç4.™Ê‹yY–«¾þû‰‚ž©€%̉ÜÊn”4ÈH}uÓ<¤™þMó°¸Ì*1Ï—)¥!ÄK_Î1èù’½VžCÌóeYÆ!¥e_Ï1èùBFj¬É0k1û* ‡Üå¿T£cÐóÅ—Mq¿JÇ *ÈÍqp}~ z–¸˜–\›±ÃxßÏÈ¢§‰l-éJãg_Ó1È"h_Ó ”ÅÖ‚Eûõ:²8Ú¯Ø ”EÒ¾ª(%7Mûqt`1¤€›ñs¡±s[Ã>Àb涆„ÅJÜŠ“r-Fnë?@XlÜ®ÀÂbⶊÄózFÛŸ”nƒ`VBÐíié6Êæ¥]MÇ ˜™v5@Áܤ ›“Ó˘vëuóÓnÅN `†ÚUuE+Ÿºþ¢…ÎÏ×þæ¾{{ýë+Z•åë2Ät}õúÒ–jôÿaçìÂ5Žaˆ“¿¾z{ùâúÇWÿ¾ä@ZGw} ~fªÅ«ï/_ ÃPòþöêò‹s)·¥I¦è®oÉ2/hySÝ»8äˆ(ZŠúnî@‹CÅ?­øB7%±ª2»N•XP XUªªTUuV VÕ¡XÕ<º!UM¬2Sè÷º£µ&[|ÜÔ.Ѻ—º PiÊÕò†; ã¤q‰A©ÄLl2"1Â#+ åƒòˆxØdÚË唦2'Ó>…¥¾îö!úã—Ž‡ gżs:FYQ;¯*öVÄÁNõEE%7·Ð!È :h¡c”µÓBû¨Ö~ £d/ÿõßK{Ö¹n,i,k3$JÑ°R‡±4»úVgq´b\–šhB¢GM,oj‘šS='²Ç!¶y>Ñ@/´ìX¦2,ªE}‹è‡ŸÁoª(óËK¨%Þãhm1§YVÆ‘ÆI-UVLô“i¢ç¶6êäË’œÖàc(‰A}Š¤»€h©s¢%FòGD§¬Þžš„Û‹ë-CY(­$O1èÔÁœLÏ=^@5UŒ9ÑTß:k—ê¼—Õ6cƒ­îbA9•6ƒþsãBÓš³¢ÍÉÝãeT™ c6§)¶Rö©îð:õÒžŽùä[¿’ûT`~¯Ë@{‘¼ÔÜo¾øðí÷ßÿôîǶ&øêkWˆh»³\ÒgÊqóÊ*® ~z÷áÓÇ?_z÷ôáÍ·ßý°Z>Ëz¾ÎÖc¦¥ÎïœÕƒâ–.–ÑÖ’³öÈqðkºôz’œµØøÄÎ#´õçlúø™3”ñò µsr‡,*aúœ© <4dUa;ÉÊ0ö¨t_ú¯¸»²¸/ë¤ò¾ÈÑ–¥ÄøƒØ¾tÄL僓—¯›ºÏ -bgò"PþœÉ I‹„]•ÚëÊ”ÉKYæ¢é‘¹ë 5¹ÛUæƒs×ãÍ)cí¾NkmNk»è|h¬ídÞk•©­)êݽ®­¹û°Kk>8$ikâ”ðXk¢Ì›­¹'óð°7ÁK£^i[êêdYÆåÿyâÉM¡¼Þ¿yâ‰êS>ÑÚ§D6à·DÃðÇÄS¿&dq–Æï‰áŠˆ5¥4jš¡S£Qƒ˜¦Æ M DBD `L 5#~½ôe‡¦ÊI ²Á-Óº^nr…A¾ ŒMML&iãR‹P©Á˜ä4ƒõgÀ¾àó †á ˆá € Ó:zøLƒaøPCÇÅz´7Dô†èQŒèŒè1L˜Öñ#z#z‹§Síx1@¿‹IºÓÐëbÑNƒõ„®„I Ƥ&f’´1©E˜Ô°ísZsÇ®ÏÙПQŒ¡Œœ¢1L˜ÀÐ¥QŒ§A.Ö“‰ûœ ňÀˆðÄ°ňäjûÿœs×[nkC6ň}N‘V}>Žc9£w«þ4&1“š˜IÓÊ$]¬Lj0&=¢ý T 䌊bä˜ bÚIƒ,ÎÒÝy…È‘ÀˆíQ½ j#jÓÔ¤©„¨Qˆ¨ŒªÑ ö¸Ú¤ËÕ`}.c¹uºZ¬/dg®tb651™¦•Ks¥Rƒ1ý'@\Û[gBùºÁïË Šönœó1»??áÊ1q¼y~Qí-¹— õØÂ/ÂÛ—ÉÂœÆYö U,ä±í-9çnÜô ´vD*ÉÂœþ…§-›ätÇ(t÷¢#º`D“Œèœ†],:¢‹æ~Çgãp“Œ#’¶q„ ¶qD QúåÞ¾ÈÒÊrÉðU^,ë¯òd÷Ãìá[7m#§ÁGøÖm¾ð‘ü>ÇgôK~‘qö'ïp„¯ÓjôØlédÚµS†RVˆåÔ–®~(ë.G$ôå”úù%—’ ïq´v ç"Ø l@Qd¢ÛØå2’`Ùêšîòëè–·™:y‰Çtçý ¦4rËI°¥UÖ9#6-…|¹T…«ªºË “+(ôó¹ußá=Ž{×”ïsÍ3ó­ÃQ»;çãNmÏÏ€åLÈÊË•yl[ç¿\¿ÿé_o~8ù%•ísO õg™‚î~wââÒ¾öH çßñ¦±ŽôaŒÃË/Ñ~ÏÖre=]1v­õúõÇÿþó»÷o?¼ùá×í‹'ñB§g+¨û3 Î…ðý§>}|üHµL2Sr £? ®; ±ö('êYNÈ( >» Ùk7?Ö}ãŇZîÚÉê3MµWK4­dó‰‚ip/ˆ´—B«9ð9•¡ÎeZ e|[uôÐùŸøeƒ0ÓÈïVüúVñÈn‘v ›·‹wPpþ{OÓ1Èn‘îk:²[¤ºuëøÂn‘î×ëÊn‘îWìÊn‘î«:ÒÉ!żóžü±„ ›7’wPK»šŽAK»šN  –tëFò ÄÒn½N  –v+v±´«êÊbitõ…Ìî°t‘˜ÛƒÒMÅÑžžC DÑžžcÄ`nG/ ‚öjt ‚øÙ«Ò1¢gOÑ1èy=$õ‘s aqƒˆì»2Ž ö鶎„ÅËŽŽ#ˆÅŠ"z{Ù';µ8‚XŒìTãbñ±£äòÂ;–0Óæ¡n¾êòX ú†Å—·§e>ž"_£ ©eßfŠ¢zUÐ,ìZ¤<â÷ŒnÓPo%(Ö{‡tçýþo?üh)¼ý º»ˆµ Œ´l ¯[²¥»pi(¹rÙ¡Ú¥Kµk—ˆ‘‹—’«—ˆU)ÎòúšU±¥Se(QÕ¡š*5UˆUUˆRUå^Þ÷LaýY5ÄEÛS,ô¬ëÃR­t—Ks±à¥O¡Ó´²éMQ!Sƒq©I¨ìz©2éJŒ™4­Lz_T˜Ô`Lj&»dªLrçS˜4­LzkT˜Ô`Lj&»jªLz©Ý"M®µc¤ÉÅvCÉÕvDÉåv@…ió”ÈwCÉ÷Ž‘uY¬‰.Œ5Ñe(Ñ…(Ѩ0mžÑe(ÑÕ1¶»ôÜ’«ÇÚb€È“¶^á·žáõÊ$ic’;ùÊ$`“2éE~e’{õ¤ieÒ›ù¤cR“0Ùu~câÛõÊ$ic’ûùÊ$`“2é¥þm\·¥„O3tÑ£ ÀHì†C19Š‘ÀA.ÖS¤â§n6tz#z#z ÃzÃz#z«\9‰Ÿºƒ[Æu¬{/MÈé¸:Þ|víSê«Ê>ŽI À$?7!L’6&±(€IN™Ô°¬[(“¦ãê …1‰™8àŒI À$?=!L’6&±(°õdÆÓØ’)bKfxÅȘ6¿¤Mï€Ù]!2¹FÔhd‰ˆ,Q£Qƒ˜¦Æ M DBD `T†–Z ¶ô'*$¸Ôׇ³ ¼Ä}!K¥Sƒ±éÏT0™¦•K-B%`’%2©aYS&MÇÕ10c0É’@™Ô`Lú“̤ieR‹0‰˜~Ÿ# yÈå+*'ÜŒ7á$³úã´ùðåCÂþqDõ?çàh0îŽHdð=ÄÒ,Q@´ZJnIêV›¢B¹–Z=T),€RKG†ïBè=Íkù $|ýí—T¹¸ÌuÄTCY8α«©Mí(”TÝ^,IªÑTMÙCízºV;£ãê‰ÛËEIõõËÿqz,}endstream endobj 6 0 obj 4558 endobj 4 0 obj <> /Contents 5 0 R >> endobj 3 0 obj << /Type /Pages /Kids [ 4 0 R ] /Count 1 >> endobj 1 0 obj <> endobj 7 0 obj <>endobj 12 0 obj <> endobj 13 0 obj <> endobj 8 0 obj <> endobj 16 0 obj <> endobj 10 0 obj <> endobj 17 0 obj <> endobj 9 0 obj <> endobj 14 0 obj <>stream xœUU TSW~Ï÷^-BMú›D¨ ²/Ö­ D.€(T„° Rvp©m/¸©Öj¡µªµ(*ÈÖZK¥˜ÒjE‡íØÎ37ž™lOÛsÞ¹ï.ÿö}ÿýÿKS“(š¦'Gi´ÅšÂì´TûÊ »Ðxæ$ü²¨ˆäZYóÄ/S1ÇîMAŽ"äèP;sr‚îL…cÎPú%¢éÒê£á›óÊò³3³ •î«ãÖxxzzý¾ã¿hÑ"妲ßN”š‚ìÌ\åaR¬ÑnÎÓir —(Ãi­6;M™©-ËË*P¦¦§kÒíj‰©ZMŽri¶6;/os±Ò=ÜCàççï- ó—gë6(ãSs ”1Ê8Mf‘65ÿO›E¹ænÎ+,*I-ÝT––®ÉÈÌÖêf»{{ÌóôÙàë礤¨ÔJjµšJ¤ÖRaTN­£"(5µ”Š¤¢¨h*–ZN1”œzŽ:IM¦ž§)g*”zšJI()õ"ÅS4å&p)EP•”…žE¯§Lòž”6©MĈâDç¦9¬u;Š]ÅÛÅ„qaâ˜,FÏ\g³4+c}ٷٓÜsÜ»Ü%®N!'Ün(Æ}£I5 {‡dÒ1#6ðR“Í0 M캌wzšÒbääCXÉZ¢»æÈ¥cÁ(I—Æ°‚v¡ {™èÓh·ˆà#|š's\I‰q˜;ú3„² 'ÄSQ¥ât©ÈKÄyU°¯Oœ ¤ ùÒ<¦°1ãyfI‡=„+™ÔÔArùa¨g.¬n)èE¸Ü–@Ä|˜D éXŠËIåT,”Çð÷¿^L¦’ç—/öSHM«Áœ¿Q8Á©Â><¯Wò>¼M‚há €m>¼²¼ù½MÉ$žÖ]×ë«÷7ÈìŽÝÛ÷lA\fEÍ;àúpn‰äúì í¨ÁÅ/IcÃBÒÊ*ãÕAËRÏuË1»ÐæÎ]ÿI`Æ„ZN¶ÜàÂX'k2à=¯¼>Ø öfÜl^¯`/ñlf®ÍK)L|ð²Ä˜ÇØ÷›¯ØîϨîÃ~&I“Žñ6ãOZø·*w¡w—»åh½®±c‘­„‰)JÏ’äUè*×pÃÌÁoÏ6š×ß´9IQÄ¢¬â­Ëv’É[Ëvål_ž¯MF‘œ´Ô«gÅ/=íµWåÒæÄõ.ÇБê“92"y´ygy~¡ ¦ÍÞ´eK´æT{癆»G#‡ïk8Â9áÝÏà@3Ôñ@DBâˆâÄ^ „“Kô⇠L‡‘ñ83‘ÙŽ1ƒJ/éè‚6!½ÅVÿTÇJËû¿c¤ç—ŸkÏísù(ˆ! Tó&òðµy+ÓÐÈÂbàÇžå7v±ß“‹w“ÜÛÖ†§ I%³m^d†Àjj[mÎÂí¥Û²Ðô¿¢mÕeû¸PöØ[Çv@'Qã¡ÔxôÄM°Íê4M0ó쟲ˆp<æw]ÙñiIcfKLcâÈ,_â@ÂHð¨\aê@?¼X£XÀl[°15q¾ ýðÈ:Ì÷Œ6…Õ(þ ·ÞmC¯ß”IËñN¯´e!©c²ÚãÔ‚Ý™„&KˆÚBh˜¥–ÿµžl¾Î-½É’rœ*~¹DÀßÐõf'êA.õpy,Šxkcqn‘vcÙZ¤AÙ5…ÇKî|ÿÝSÜ«Ì!wÓJpFfôÝGkâ¤ç.íç8Z0LJcùÑö"#SW‡ú%L´˜.ó]S¾Ðå %PTB÷ ‹`žÉB›aˆ±ê¸3Â92X#ôt¯pè%dkÁS°k½ù‡Þ˜gÿdÒñ‰öxËÞo³?t^ëýæ3M”œ<ýmyítú2û¿Ì>XÝæ¡N-^¹N®íJ­‹Dœt\RÞ\ÃÝdþ;¥Ô)FƒÖëJèÆ°á¸ÛG¼a}÷Q©¼½Mªc&ÓýñÐ<j‡/›×Ç%¤¬OˆO¹ØÙu±¹S}ð6ÅS¸‡Á «N|Ä&³ëùØõî›M÷ÿñ»Þê ½æ®N»ž¾"ãOFº—Šð)ø…·•ªq)3ÛáÙß îX“ôtß-ìȘKd1$DºIô#"ã:™@p[¯!BØÁ#àÊÙÍ&és¸J„¿=o«RãªßŒAø-ÑÉ⬢ø#ˆè¹ ã˜âÖB^C$‘Àu$8ØMŸP ØOf-] ”ni@kf÷Ø7>WN«ïõqéÕYšýà þÕ3Ôípôx àóA†APñD5ªŸ\ñÆh–ôÛâÄ>,´ ð$Ÿ|è )r*¬Å=ÄêcjóäŸ7ïwtüñãŠú?ÕY  endstream endobj 11 0 obj <> endobj 15 0 obj <>stream xœURmLSW>—Bï•ò14×)íe›ÀpÓœä£ àêæS©¥–Ž~ *RÜœ Y„-¶"à ’0cTÎ92™Ìpafü21À{ÉaËN™fÛŸ“sžó<ÏyÎû¾ òõA çêŒE:›A«‰H°s½P¸¸–×ùˆ¡LÌ⽓_(ÊhœÄàëZç[»ZC "J^@†)®p$Z¬öƒ>Ï&lØ÷NæÆÍ›ÃÿE¶ÄÄÄGìÏo„$]¡Ao”tS¤3Z¬&Ùö¦HÙF£A+èvk^¡ ÉÍÕåzeû5F]¾ 2 V«¥HظQx-*jK]¶½e09V(ìÕ˜ … Á›ÿB(Èl±ÚŽi´º£cD¸€íAo£}h?J@IH…RPÚ\(1(˜VùR¨=f”LóÈg‡ÏeŸ ’¬‘DH²$ųA"`§ 2³sùà0«VôÜÝü¢›ÇÙãC=ׇÛÞO—“IŠüÅ‚l×0a²¬j­|Eô}–ZØÀ~?¦“jAÀMñ+žÄ( C’‰j† ØÑO¾ažÄ**‰?Þ§ Sf&ÇÇ}ú´tLAˆÙ ° Â2æ65;å5Ë¢iˆðg=wJ¯ª=–aÌÁÊÇÑcÇÆô} }êÕ4œ‚³Í5þþ^¯ê•Y*UröðÌŒgdT“¶9˜›fü“î  ë‰@Ö«"GÈ/Ò7z4Û[ªÏ·ÈŸ°•Ÿ®(ÅÜѲ‹ô>ûžLÜá\ Dë#¯­ÝT<ŦÒ¤¨MßËá",°Dv7Øo¸»äéT9ð¼0­Tª†P‰ØÔÈ4Ÿù·r0À‚Oü-¢$ʸÝ$¤Lk¥µ·š?ÁÜM—]o/;ñáiEq)ƪS{KW öÌwñ&|°^×bmÏï-öàAüõ•ýØR]Z‚ƒp)‰¿]{0é=sK×õËÎÛ¹§Î]{®ªáüšÿ$¹êýÀç0Å—Âê×"k1G¶E„‘ ¢š~¶Âª© àëäE©=Ùb݇¹¤œ¡Ùþ žÆEC[{Cþ»¬ Iœ×q^š†àyFl‚"¾Òkp-¾yºïä5ý“­ß†:¿N|É.²ó·0ˆ€À'#sí "—–ääÀZœßx¼í¤ó“/Êû¹ªi¾îçÎîïp'î(qZêŸ+ª>ÊÑ®ÑáÜ:ýh–ùÕÛ·»´oÑÒVÇ¥+Ž†Šr‡|–ý *¯²s‘‡G)Òâ"ÇSAS°AÐN¥ÂüàUj—FÈ;ÒÒ“/í]Ý-q’Ï( &ï”™¼T2Fì†/ùÅJ"+—È d†`“d”>stream 2012-11-25T07:42:45-05:00 2012-11-25T07:42:45-05:00 fig2dev Version 3.2 Patchlevel 5d ./rfftwnd.fig endstream endobj 2 0 obj <>endobj xref 0 19 0000000000 65535 f 0000004873 00000 n 0000011691 00000 n 0000004814 00000 n 0000004663 00000 n 0000000015 00000 n 0000004643 00000 n 0000004938 00000 n 0000005050 00000 n 0000005967 00000 n 0000005529 00000 n 0000008809 00000 n 0000004979 00000 n 0000005009 00000 n 0000006326 00000 n 0000009066 00000 n 0000005442 00000 n 0000005880 00000 n 0000010250 00000 n trailer << /Size 19 /Root 1 0 R /Info 2 0 R /ID [<98E2D5EBE99A0340445FD4962CA4A0FB><98E2D5EBE99A0340445FD4962CA4A0FB>] >> startxref 11880 %%EOF fftw-3.3.3/doc/rfftwnd.eps0000644000175000017500000014773312054411104012346 00000000000000%!PS-Adobe-3.0 EPSF-3.0 %%Title: ./rfftwnd.fig %%Creator: fig2dev Version 3.2 Patchlevel 5d %%CreationDate: Sun Nov 25 07:42:44 2012 %%BoundingBox: 0 0 270 405 %Magnification: 0.7000 %%EndComments %%BeginProlog /$F2psDict 200 dict def $F2psDict begin $F2psDict /mtrx matrix put /col-1 {0 setgray} bind def /col0 {0.000 0.000 0.000 srgb} bind def /col1 {0.000 0.000 1.000 srgb} bind def /col2 {0.000 1.000 0.000 srgb} bind def /col3 {0.000 1.000 1.000 srgb} bind def /col4 {1.000 0.000 0.000 srgb} bind def /col5 {1.000 0.000 1.000 srgb} bind def /col6 {1.000 1.000 0.000 srgb} bind def /col7 {1.000 1.000 1.000 srgb} bind def /col8 {0.000 0.000 0.560 srgb} bind def /col9 {0.000 0.000 0.690 srgb} bind def /col10 {0.000 0.000 0.820 srgb} bind def /col11 {0.530 0.810 1.000 srgb} bind def /col12 {0.000 0.560 0.000 srgb} bind def /col13 {0.000 0.690 0.000 srgb} bind def /col14 {0.000 0.820 0.000 srgb} bind def /col15 {0.000 0.560 0.560 srgb} bind def /col16 {0.000 0.690 0.690 srgb} bind def /col17 {0.000 0.820 0.820 srgb} bind def /col18 {0.560 0.000 0.000 srgb} bind def /col19 {0.690 0.000 0.000 srgb} bind def /col20 {0.820 0.000 0.000 srgb} bind def /col21 {0.560 0.000 0.560 srgb} bind def /col22 {0.690 0.000 0.690 srgb} bind def /col23 {0.820 0.000 0.820 srgb} bind def /col24 {0.500 0.190 0.000 srgb} bind def /col25 {0.630 0.250 0.000 srgb} bind def /col26 {0.750 0.380 0.000 srgb} bind def /col27 {1.000 0.500 0.500 srgb} bind def /col28 {1.000 0.630 0.630 srgb} bind def /col29 {1.000 0.750 0.750 srgb} bind def /col30 {1.000 0.880 0.880 srgb} bind def /col31 {1.000 0.840 0.000 srgb} bind def /col32 {0.475 0.490 0.475 srgb} bind def /col33 {0.937 0.922 0.937 srgb} bind def /col34 {0.906 0.188 0.125 srgb} bind def /col35 {0.969 0.557 0.525 srgb} bind def /col36 {0.412 0.588 0.780 srgb} bind def /col37 {0.525 0.667 0.843 srgb} bind def /col38 {0.875 0.859 0.000 srgb} bind def end /cp {closepath} bind def /ef {eofill} bind def /gr {grestore} bind def /gs {gsave} bind def /sa {save} bind def /rs {restore} bind def /l {lineto} bind def /m {moveto} bind def /rm {rmoveto} bind def /n {newpath} bind def /s {stroke} bind def /sh {show} bind def /slc {setlinecap} bind def /slj {setlinejoin} bind def /slw {setlinewidth} bind def /srgb {setrgbcolor} bind def /rot {rotate} bind def /sc {scale} bind def /sd {setdash} bind def /ff {findfont} bind def /sf {setfont} bind def /scf {scalefont} bind def /sw {stringwidth} bind def /tr {translate} bind def /tnt {dup dup currentrgbcolor 4 -2 roll dup 1 exch sub 3 -1 roll mul add 4 -2 roll dup 1 exch sub 3 -1 roll mul add 4 -2 roll dup 1 exch sub 3 -1 roll mul add srgb} bind def /shd {dup dup currentrgbcolor 4 -2 roll mul 4 -2 roll mul 4 -2 roll mul srgb} bind def /reencdict 12 dict def /ReEncode { reencdict begin /newcodesandnames exch def /newfontname exch def /basefontname exch def /basefontdict basefontname findfont def /newfont basefontdict maxlength dict def basefontdict { exch dup /FID ne { dup /Encoding eq { exch dup length array copy newfont 3 1 roll put } { exch newfont 3 1 roll put } ifelse } { pop pop } ifelse } forall newfont /FontName newfontname put newcodesandnames aload pop 128 1 255 { newfont /Encoding get exch /.notdef put } for newcodesandnames length 2 idiv { newfont /Encoding get 3 1 roll put } repeat newfontname newfont definefont pop end } def /isovec [ 8#055 /minus 8#200 /grave 8#201 /acute 8#202 /circumflex 8#203 /tilde 8#204 /macron 8#205 /breve 8#206 /dotaccent 8#207 /dieresis 8#210 /ring 8#211 /cedilla 8#212 /hungarumlaut 8#213 /ogonek 8#214 /caron 8#220 /dotlessi 8#230 /oe 8#231 /OE 8#240 /space 8#241 /exclamdown 8#242 /cent 8#243 /sterling 8#244 /currency 8#245 /yen 8#246 /brokenbar 8#247 /section 8#250 /dieresis 8#251 /copyright 8#252 /ordfeminine 8#253 /guillemotleft 8#254 /logicalnot 8#255 /hyphen 8#256 /registered 8#257 /macron 8#260 /degree 8#261 /plusminus 8#262 /twosuperior 8#263 /threesuperior 8#264 /acute 8#265 /mu 8#266 /paragraph 8#267 /periodcentered 8#270 /cedilla 8#271 /onesuperior 8#272 /ordmasculine 8#273 /guillemotright 8#274 /onequarter 8#275 /onehalf 8#276 /threequarters 8#277 /questiondown 8#300 /Agrave 8#301 /Aacute 8#302 /Acircumflex 8#303 /Atilde 8#304 /Adieresis 8#305 /Aring 8#306 /AE 8#307 /Ccedilla 8#310 /Egrave 8#311 /Eacute 8#312 /Ecircumflex 8#313 /Edieresis 8#314 /Igrave 8#315 /Iacute 8#316 /Icircumflex 8#317 /Idieresis 8#320 /Eth 8#321 /Ntilde 8#322 /Ograve 8#323 /Oacute 8#324 /Ocircumflex 8#325 /Otilde 8#326 /Odieresis 8#327 /multiply 8#330 /Oslash 8#331 /Ugrave 8#332 /Uacute 8#333 /Ucircumflex 8#334 /Udieresis 8#335 /Yacute 8#336 /Thorn 8#337 /germandbls 8#340 /agrave 8#341 /aacute 8#342 /acircumflex 8#343 /atilde 8#344 /adieresis 8#345 /aring 8#346 /ae 8#347 /ccedilla 8#350 /egrave 8#351 /eacute 8#352 /ecircumflex 8#353 /edieresis 8#354 /igrave 8#355 /iacute 8#356 /icircumflex 8#357 /idieresis 8#360 /eth 8#361 /ntilde 8#362 /ograve 8#363 /oacute 8#364 /ocircumflex 8#365 /otilde 8#366 /odieresis 8#367 /divide 8#370 /oslash 8#371 /ugrave 8#372 /uacute 8#373 /ucircumflex 8#374 /udieresis 8#375 /yacute 8#376 /thorn 8#377 /ydieresis] def /Helvetica /Helvetica-iso isovec ReEncode /Helvetica-Bold /Helvetica-Bold-iso isovec ReEncode /$F2psBegin {$F2psDict begin /$F2psEnteredState save def} def /$F2psEnd {$F2psEnteredState restore end} def /pageheader { save newpath 0 405 moveto 0 0 lineto 270 0 lineto 270 405 lineto closepath clip newpath -2.5 407.2 translate 1 -1 scale $F2psBegin 10 setmiterlimit 0 slj 0 slc 0.04200 0.04200 sc } bind def /pagefooter { $F2psEnd restore } bind def %%EndProlog pageheader % % Fig objects follow % % % here starts figure with depth 998 % Polyline 0 slj 0 slc 0.000 slw n 1221 7280 m 6435 7280 l 6435 9676 l 1221 9676 l 1221 7280 l cp gs col7 1.00 shd ef gr % Polyline 7.500 slw n 1221 7280 m 6435 7280 l 6435 9676 l 1221 9676 l 1221 7280 l cp gs col32 s gr % Polyline 0.000 slw n 1221 7280 m 1620 7280 l 1620 7656 l 1221 7656 l 1221 7280 l cp gs col33 1.00 shd ef gr % Polyline 7.500 slw n 1221 7280 m 1620 7280 l 1620 7656 l 1221 7656 l 1221 7280 l cp gs col32 s gr % Polyline 0.000 slw n 1620 7280 m 2019 7280 l 2019 7656 l 1620 7656 l 1620 7280 l cp gs col33 1.00 shd ef gr % Polyline 7.500 slw n 1620 7280 m 2019 7280 l 2019 7656 l 1620 7656 l 1620 7280 l cp gs col32 s gr % Polyline 0.000 slw n 2019 7280 m 2418 7280 l 2418 7656 l 2019 7656 l 2019 7280 l cp gs col33 1.00 shd ef gr % Polyline 7.500 slw n 2019 7280 m 2418 7280 l 2418 7656 l 2019 7656 l 2019 7280 l cp gs col32 s gr % Polyline 0.000 slw n 2418 7280 m 2817 7280 l 2817 7656 l 2418 7656 l 2418 7280 l cp gs col33 1.00 shd ef gr % Polyline 7.500 slw n 2418 7280 m 2817 7280 l 2817 7656 l 2418 7656 l 2418 7280 l cp gs col32 s gr % Polyline 0.000 slw n 4038 7280 m 4438 7280 l 4438 7656 l 4038 7656 l 4038 7280 l cp gs col33 1.00 shd ef gr % Polyline 7.500 slw n 4038 7280 m 4438 7280 l 4438 7656 l 4038 7656 l 4038 7280 l cp gs col32 s gr % Polyline 0.000 slw n 4438 7280 m 4837 7280 l 4837 7656 l 4438 7656 l 4438 7280 l cp gs col33 1.00 shd ef gr % Polyline 7.500 slw n 4438 7280 m 4837 7280 l 4837 7656 l 4438 7656 l 4438 7280 l cp gs col32 s gr % Polyline 0.000 slw n 4837 7280 m 5236 7280 l 5236 7656 l 4837 7656 l 4837 7280 l cp gs col33 1.00 shd ef gr % Polyline 7.500 slw n 4837 7280 m 5236 7280 l 5236 7656 l 4837 7656 l 4837 7280 l cp gs col32 s gr % Polyline 0.000 slw n 5236 7280 m 5635 7280 l 5635 7656 l 5236 7656 l 5236 7280 l cp gs col33 1.00 shd ef gr % Polyline 7.500 slw n 5236 7280 m 5635 7280 l 5635 7656 l 5236 7656 l 5236 7280 l cp gs col32 s gr % Polyline 0.000 slw n 1221 7656 m 1620 7656 l 1620 8032 l 1221 8032 l 1221 7656 l cp gs col33 1.00 shd ef gr % Polyline 7.500 slw n 1221 7656 m 1620 7656 l 1620 8032 l 1221 8032 l 1221 7656 l cp gs col32 s gr % Polyline 0.000 slw n 1620 7656 m 2019 7656 l 2019 8032 l 1620 8032 l 1620 7656 l cp gs col33 1.00 shd ef gr % Polyline 7.500 slw n 1620 7656 m 2019 7656 l 2019 8032 l 1620 8032 l 1620 7656 l cp gs col32 s gr % Polyline 0.000 slw n 2019 7656 m 2418 7656 l 2418 8032 l 2019 8032 l 2019 7656 l cp gs col33 1.00 shd ef gr % Polyline 7.500 slw n 2019 7656 m 2418 7656 l 2418 8032 l 2019 8032 l 2019 7656 l cp gs col32 s gr % Polyline 0.000 slw n 2418 7656 m 2817 7656 l 2817 8032 l 2418 8032 l 2418 7656 l cp gs col33 1.00 shd ef gr % Polyline 7.500 slw n 2418 7656 m 2817 7656 l 2817 8032 l 2418 8032 l 2418 7656 l cp gs col32 s gr % Polyline 0.000 slw n 4038 7656 m 4438 7656 l 4438 8032 l 4038 8032 l 4038 7656 l cp gs col33 1.00 shd ef gr % Polyline 7.500 slw n 4038 7656 m 4438 7656 l 4438 8032 l 4038 8032 l 4038 7656 l cp gs col32 s gr % Polyline 0.000 slw n 4438 7656 m 4837 7656 l 4837 8032 l 4438 8032 l 4438 7656 l cp gs col33 1.00 shd ef gr % Polyline 7.500 slw n 4438 7656 m 4837 7656 l 4837 8032 l 4438 8032 l 4438 7656 l cp gs col32 s gr % Polyline 0.000 slw n 4837 7656 m 5236 7656 l 5236 8032 l 4837 8032 l 4837 7656 l cp gs col33 1.00 shd ef gr % Polyline 7.500 slw n 4837 7656 m 5236 7656 l 5236 8032 l 4837 8032 l 4837 7656 l cp gs col32 s gr % Polyline 0.000 slw n 5236 7656 m 5635 7656 l 5635 8032 l 5236 8032 l 5236 7656 l cp gs col33 1.00 shd ef gr % Polyline 7.500 slw n 5236 7656 m 5635 7656 l 5635 8032 l 5236 8032 l 5236 7656 l cp gs col32 s gr % Polyline 0.000 slw n 1221 8924 m 1620 8924 l 1620 9300 l 1221 9300 l 1221 8924 l cp gs col33 1.00 shd ef gr % Polyline 7.500 slw n 1221 8924 m 1620 8924 l 1620 9300 l 1221 9300 l 1221 8924 l cp gs col32 s gr % Polyline 0.000 slw n 1620 8924 m 2019 8924 l 2019 9300 l 1620 9300 l 1620 8924 l cp gs col33 1.00 shd ef gr % Polyline 7.500 slw n 1620 8924 m 2019 8924 l 2019 9300 l 1620 9300 l 1620 8924 l cp gs col32 s gr % Polyline 0.000 slw n 2019 8924 m 2418 8924 l 2418 9300 l 2019 9300 l 2019 8924 l cp gs col33 1.00 shd ef gr % Polyline 7.500 slw n 2019 8924 m 2418 8924 l 2418 9300 l 2019 9300 l 2019 8924 l cp gs col32 s gr % Polyline 0.000 slw n 2418 8924 m 2817 8924 l 2817 9300 l 2418 9300 l 2418 8924 l cp gs col33 1.00 shd ef gr % Polyline 7.500 slw n 2418 8924 m 2817 8924 l 2817 9300 l 2418 9300 l 2418 8924 l cp gs col32 s gr % Polyline 0.000 slw n 4038 8924 m 4438 8924 l 4438 9300 l 4038 9300 l 4038 8924 l cp gs col33 1.00 shd ef gr % Polyline 7.500 slw n 4038 8924 m 4438 8924 l 4438 9300 l 4038 9300 l 4038 8924 l cp gs col32 s gr % Polyline 0.000 slw n 4438 8924 m 4837 8924 l 4837 9300 l 4438 9300 l 4438 8924 l cp gs col33 1.00 shd ef gr % Polyline 7.500 slw n 4438 8924 m 4837 8924 l 4837 9300 l 4438 9300 l 4438 8924 l cp gs col32 s gr % Polyline 0.000 slw n 4837 8924 m 5236 8924 l 5236 9300 l 4837 9300 l 4837 8924 l cp gs col33 1.00 shd ef gr % Polyline 7.500 slw n 4837 8924 m 5236 8924 l 5236 9300 l 4837 9300 l 4837 8924 l cp gs col32 s gr % Polyline 0.000 slw n 5236 8924 m 5635 8924 l 5635 9300 l 5236 9300 l 5236 8924 l cp gs col33 1.00 shd ef gr % Polyline 7.500 slw n 5236 8924 m 5635 8924 l 5635 9300 l 5236 9300 l 5236 8924 l cp gs col32 s gr % Polyline 0.000 slw n 1221 9300 m 1620 9300 l 1620 9676 l 1221 9676 l 1221 9300 l cp gs col33 1.00 shd ef gr % Polyline 7.500 slw n 1221 9300 m 1620 9300 l 1620 9676 l 1221 9676 l 1221 9300 l cp gs col32 s gr % Polyline 0.000 slw n 1620 9300 m 2019 9300 l 2019 9676 l 1620 9676 l 1620 9300 l cp gs col33 1.00 shd ef gr % Polyline 7.500 slw n 1620 9300 m 2019 9300 l 2019 9676 l 1620 9676 l 1620 9300 l cp gs col32 s gr % Polyline 0.000 slw n 2019 9300 m 2418 9300 l 2418 9676 l 2019 9676 l 2019 9300 l cp gs col33 1.00 shd ef gr % Polyline 7.500 slw n 2019 9300 m 2418 9300 l 2418 9676 l 2019 9676 l 2019 9300 l cp gs col32 s gr % Polyline 0.000 slw n 2418 9300 m 2817 9300 l 2817 9676 l 2418 9676 l 2418 9300 l cp gs col33 1.00 shd ef gr % Polyline 7.500 slw n 2418 9300 m 2817 9300 l 2817 9676 l 2418 9676 l 2418 9300 l cp gs col32 s gr % Polyline 0.000 slw n 4038 9300 m 4438 9300 l 4438 9676 l 4038 9676 l 4038 9300 l cp gs col33 1.00 shd ef gr % Polyline 7.500 slw n 4038 9300 m 4438 9300 l 4438 9676 l 4038 9676 l 4038 9300 l cp gs col32 s gr % Polyline 0.000 slw n 4438 9300 m 4837 9300 l 4837 9676 l 4438 9676 l 4438 9300 l cp gs col33 1.00 shd ef gr % Polyline 7.500 slw n 4438 9300 m 4837 9300 l 4837 9676 l 4438 9676 l 4438 9300 l cp gs col32 s gr % Polyline 0.000 slw n 4837 9300 m 5236 9300 l 5236 9676 l 4837 9676 l 4837 9300 l cp gs col33 1.00 shd ef gr % Polyline 7.500 slw n 4837 9300 m 5236 9300 l 5236 9676 l 4837 9676 l 4837 9300 l cp gs col32 s gr % Polyline 0.000 slw n 5236 9300 m 5635 9300 l 5635 9676 l 5236 9676 l 5236 9300 l cp gs col33 1.00 shd ef gr % Polyline 7.500 slw n 5236 9300 m 5635 9300 l 5635 9676 l 5236 9676 l 5236 9300 l cp gs col32 s gr /Helvetica-iso ff 225.00 scf sf 74 89 m gs 1 -1 sc ( ) col0 sh gr /Helvetica-iso ff 225.00 scf sf 2064 7726 m gs 1 -1 sc (...) col0 sh gr % Polyline n 2819 7658 m 2869 7658 l gs col32 s gr % Polyline n 2952 7658 m 3002 7658 l gs col32 s gr % Polyline n 3085 7658 m 3135 7658 l gs col32 s gr % Polyline n 3219 7658 m 3252 7658 l gs col32 s gr % Polyline n 2819 8033 m 2869 8033 l gs col32 s gr % Polyline n 2952 8033 m 3002 8033 l gs col32 s gr % Polyline n 3085 8033 m 3135 8033 l gs col32 s gr % Polyline n 3219 8033 m 3252 8033 l gs col32 s gr % Polyline n 2819 8025 m 2819 8075 l gs col32 s gr % Polyline n 2819 8158 m 2819 8208 l gs col32 s gr % Polyline n 2819 8291 m 2819 8341 l gs col32 s gr % Polyline n 2419 8025 m 2419 8075 l gs col32 s gr % Polyline n 2419 8158 m 2419 8208 l gs col32 s gr % Polyline n 2419 8291 m 2419 8341 l gs col32 s gr % Polyline n 2019 8025 m 2019 8075 l gs col32 s gr % Polyline n 2019 8158 m 2019 8208 l gs col32 s gr % Polyline n 2019 8291 m 2019 8341 l gs col32 s gr % Polyline n 1619 8025 m 1619 8075 l gs col32 s gr % Polyline n 1619 8158 m 1619 8208 l gs col32 s gr % Polyline n 1619 8291 m 1619 8341 l gs col32 s gr % Polyline n 4036 7658 m 3986 7658 l gs col32 s gr % Polyline n 3902 7658 m 3852 7658 l gs col32 s gr % Polyline n 3769 7658 m 3719 7658 l gs col32 s gr % Polyline n 3636 7658 m 3602 7658 l gs col32 s gr % Polyline n 4036 8033 m 3986 8033 l gs col32 s gr % Polyline n 3902 8033 m 3852 8033 l gs col32 s gr % Polyline n 3769 8033 m 3719 8033 l gs col32 s gr % Polyline n 3636 8033 m 3602 8033 l gs col32 s gr % Polyline n 4035 8025 m 4035 8075 l gs col32 s gr % Polyline n 4035 8158 m 4035 8208 l gs col32 s gr % Polyline n 4035 8291 m 4035 8341 l gs col32 s gr % Polyline n 4435 8025 m 4435 8075 l gs col32 s gr % Polyline n 4435 8158 m 4435 8208 l gs col32 s gr % Polyline n 4435 8291 m 4435 8341 l gs col32 s gr % Polyline n 4835 8025 m 4835 8075 l gs col32 s gr % Polyline n 4835 8158 m 4835 8208 l gs col32 s gr % Polyline n 4835 8291 m 4835 8341 l gs col32 s gr % Polyline n 5235 8025 m 5235 8075 l gs col32 s gr % Polyline n 5235 8158 m 5235 8208 l gs col32 s gr % Polyline n 5235 8291 m 5235 8341 l gs col32 s gr % Polyline n 4036 9300 m 3986 9300 l gs col32 s gr % Polyline n 3902 9300 m 3852 9300 l gs col32 s gr % Polyline n 3769 9300 m 3719 9300 l gs col32 s gr % Polyline n 3636 9300 m 3602 9300 l gs col32 s gr % Polyline n 4036 8925 m 3986 8925 l gs col32 s gr % Polyline n 3902 8925 m 3852 8925 l gs col32 s gr % Polyline n 3769 8925 m 3719 8925 l gs col32 s gr % Polyline n 3636 8925 m 3602 8925 l gs col32 s gr % Polyline n 4035 8933 m 4035 8883 l gs col32 s gr % Polyline n 4035 8800 m 4035 8750 l gs col32 s gr % Polyline n 4035 8666 m 4035 8616 l gs col32 s gr % Polyline n 4435 8933 m 4435 8883 l gs col32 s gr % Polyline n 4435 8800 m 4435 8750 l gs col32 s gr % Polyline n 4435 8666 m 4435 8616 l gs col32 s gr % Polyline n 4835 8933 m 4835 8883 l gs col32 s gr % Polyline n 4835 8800 m 4835 8750 l gs col32 s gr % Polyline n 4835 8666 m 4835 8616 l gs col32 s gr % Polyline n 5235 8933 m 5235 8883 l gs col32 s gr % Polyline n 5235 8800 m 5235 8750 l gs col32 s gr % Polyline n 5235 8666 m 5235 8616 l gs col32 s gr % Polyline n 2819 9300 m 2869 9300 l gs col32 s gr % Polyline n 2952 9300 m 3002 9300 l gs col32 s gr % Polyline n 3085 9300 m 3135 9300 l gs col32 s gr % Polyline n 3219 9300 m 3252 9300 l gs col32 s gr % Polyline n 2819 8925 m 2869 8925 l gs col32 s gr % Polyline n 2952 8925 m 3002 8925 l gs col32 s gr % Polyline n 3085 8925 m 3135 8925 l gs col32 s gr % Polyline n 3219 8925 m 3252 8925 l gs col32 s gr % Polyline n 2819 8933 m 2819 8883 l gs col32 s gr % Polyline n 2819 8800 m 2819 8750 l gs col32 s gr % Polyline n 2819 8666 m 2819 8616 l gs col32 s gr % Polyline n 2419 8933 m 2419 8883 l gs col32 s gr % Polyline n 2419 8800 m 2419 8750 l gs col32 s gr % Polyline n 2419 8666 m 2419 8616 l gs col32 s gr % Polyline n 2019 8933 m 2019 8883 l gs col32 s gr % Polyline n 2019 8800 m 2019 8750 l gs col32 s gr % Polyline n 2019 8666 m 2019 8616 l gs col32 s gr % Polyline n 1619 8933 m 1619 8883 l gs col32 s gr % Polyline n 1619 8800 m 1619 8750 l gs col32 s gr % Polyline n 1619 8666 m 1619 8616 l gs col32 s gr /Helvetica-iso ff 195.00 scf sf 74 89 m gs 1 -1 sc ( ) col0 sh gr /Helvetica-iso ff 195.00 scf sf 2338 7001 m gs 1 -1 sc (ny ) col0 sh gr /Helvetica-iso ff 195.00 scf sf 2605 7001 m gs 1 -1 sc (+ 2-ny%2) col34 sh gr /Helvetica-iso ff 195.00 scf sf 3500 7001 m gs 1 -1 sc ( = 2*\(ny/2+1\)) col0 sh gr /Helvetica-iso ff 195.00 scf sf 74 89 m gs 1 -1 sc ( ) col0 sh gr /Helvetica-iso ff 195.00 scf sf 681 8451 m gs 1 -1 sc (nx) col0 sh gr /Helvetica-iso ff 150.00 scf sf 74 89 m gs 1 -1 sc ( ) col0 sh gr /Helvetica-iso ff 150.00 scf sf 1364 7179 m gs 1 -1 sc (0) col0 sh gr /Helvetica-iso ff 150.00 scf sf 74 89 m gs 1 -1 sc ( ) col0 sh gr /Helvetica-iso ff 150.00 scf sf 6097 7179 m gs 1 -1 sc (ny+1) col0 sh gr /Helvetica-iso ff 150.00 scf sf 74 89 m gs 1 -1 sc ( ) col0 sh gr /Helvetica-iso ff 150.00 scf sf 1064 7479 m gs 1 -1 sc (0) col0 sh gr /Helvetica-iso ff 150.00 scf sf 74 89 m gs 1 -1 sc ( ) col0 sh gr /Helvetica-iso ff 150.00 scf sf 864 9479 m gs 1 -1 sc (nx-1) col0 sh gr % Polyline 0.000 slw n 5636 7280 m 6035 7280 l 6035 7656 l 5636 7656 l 5636 7280 l cp gs col35 1.00 shd ef gr % Polyline 7.500 slw n 5636 7280 m 6035 7280 l 6035 7656 l 5636 7656 l 5636 7280 l cp gs col32 s gr % Polyline 0.000 slw n 5636 7656 m 6035 7656 l 6035 8032 l 5636 8032 l 5636 7656 l cp gs col35 1.00 shd ef gr % Polyline 7.500 slw n 5636 7656 m 6035 7656 l 6035 8032 l 5636 8032 l 5636 7656 l cp gs col32 s gr % Polyline 0.000 slw n 5636 8924 m 6035 8924 l 6035 9300 l 5636 9300 l 5636 8924 l cp gs col35 1.00 shd ef gr % Polyline 7.500 slw n 5636 8924 m 6035 8924 l 6035 9300 l 5636 9300 l 5636 8924 l cp gs col32 s gr % Polyline 0.000 slw n 5636 9300 m 6035 9300 l 6035 9676 l 5636 9676 l 5636 9300 l cp gs col35 1.00 shd ef gr % Polyline 7.500 slw n 5636 9300 m 6035 9300 l 6035 9676 l 5636 9676 l 5636 9300 l cp gs col32 s gr % Polyline 0.000 slw n 6036 7280 m 6435 7280 l 6435 7656 l 6036 7656 l 6036 7280 l cp gs col35 1.00 shd ef gr % Polyline 7.500 slw n 6036 7280 m 6435 7280 l 6435 7656 l 6036 7656 l 6036 7280 l cp gs col32 s gr % Polyline 0.000 slw n 6036 7656 m 6435 7656 l 6435 8032 l 6036 8032 l 6036 7656 l cp gs col35 1.00 shd ef gr % Polyline 7.500 slw n 6036 7656 m 6435 7656 l 6435 8032 l 6036 8032 l 6036 7656 l cp gs col32 s gr % Polyline 0.000 slw n 6036 8924 m 6435 8924 l 6435 9300 l 6036 9300 l 6036 8924 l cp gs col35 1.00 shd ef gr % Polyline 7.500 slw n 6036 8924 m 6435 8924 l 6435 9300 l 6036 9300 l 6036 8924 l cp gs col32 s gr % Polyline 0.000 slw n 6036 9300 m 6435 9300 l 6435 9676 l 6036 9676 l 6036 9300 l cp gs col35 1.00 shd ef gr % Polyline 7.500 slw n 6036 9300 m 6435 9300 l 6435 9676 l 6036 9676 l 6036 9300 l cp gs col32 s gr % Polyline n 5635 7283 m 5635 9683 l gs col0 s gr % Polyline n 1420 7515 m 6312 7515 l gs col0 s gr % Polyline 0.000 slw n 6348 7518 m 6117 7462 l 6117 7515 l gs 0.00 setgray ef gr % Polyline 7.500 slw n 6348 7518 m 6117 7462 l 6117 7515 l gs col0 s gr % Polyline 0.000 slw n 6348 7512 m 6117 7568 l 6117 7515 l gs 0.00 setgray ef gr % Polyline 7.500 slw n 6348 7512 m 6117 7568 l 6117 7515 l gs col0 s gr % Polyline n 1420 7891 m 5863 7891 l gs col0 s gr % Polyline 0.000 slw n 5895 7894 m 5685 7838 l 5685 7891 l gs 0.00 setgray ef gr % Polyline 7.500 slw n 5895 7894 m 5685 7838 l 5685 7891 l gs col0 s gr % Polyline 0.000 slw n 5895 7888 m 5685 7944 l 5685 7891 l gs 0.00 setgray ef gr % Polyline 7.500 slw n 5895 7888 m 5685 7944 l 5685 7891 l gs col0 s gr % Polyline n 1420 9112 m 5863 9112 l gs col0 s gr % Polyline 0.000 slw n 5895 9115 m 5685 9059 l 5685 9112 l gs 0.00 setgray ef gr % Polyline 7.500 slw n 5895 9115 m 5685 9059 l 5685 9112 l gs col0 s gr % Polyline 0.000 slw n 5895 9109 m 5685 9165 l 5685 9112 l gs 0.00 setgray ef gr % Polyline 7.500 slw n 5895 9109 m 5685 9165 l 5685 9112 l gs col0 s gr % Polyline n 1420 9488 m 5863 9488 l gs col0 s gr % Polyline 0.000 slw n 5895 9491 m 5685 9435 l 5685 9488 l gs 0.00 setgray ef gr % Polyline 7.500 slw n 5895 9491 m 5685 9435 l 5685 9488 l gs col0 s gr % Polyline 0.000 slw n 5895 9485 m 5685 9541 l 5685 9488 l gs 0.00 setgray ef gr % Polyline 7.500 slw n 5895 9485 m 5685 9541 l 5685 9488 l gs col0 s gr /Helvetica-iso ff 375.00 scf sf 74 89 m gs 1 -1 sc ( ) col0 sh gr /Helvetica-iso ff 375.00 scf sf 6250 9461 m gs 1 -1 sc 90.0 rot (\(padding\)) col0 sh gr /Helvetica-Bold-iso ff 240.00 scf sf 74 89 m gs 1 -1 sc ( ) col0 sh gr /Helvetica-Bold-iso ff 240.00 scf sf 428 9283 m gs 1 -1 sc 90.0 rot (input, in-place) col0 sh gr /Helvetica-iso ff 150.00 scf sf 74 89 m gs 1 -1 sc ( ) col0 sh gr /Helvetica-iso ff 150.00 scf sf 1264 7429 m gs 1 -1 sc (0) col0 sh gr /Helvetica-iso ff 150.00 scf sf 74 89 m gs 1 -1 sc ( ) col0 sh gr /Helvetica-iso ff 150.00 scf sf 1681 7429 m gs 1 -1 sc (1) col0 sh gr /Helvetica-iso ff 150.00 scf sf 74 89 m gs 1 -1 sc ( ) col0 sh gr /Helvetica-iso ff 150.00 scf sf 2081 7429 m gs 1 -1 sc (2) col0 sh gr /Helvetica-iso ff 150.00 scf sf 74 89 m gs 1 -1 sc ( ) col0 sh gr /Helvetica-iso ff 150.00 scf sf 2481 7429 m gs 1 -1 sc (3) col0 sh gr /Helvetica-iso ff 150.00 scf sf 74 89 m gs 1 -1 sc ( ) col0 sh gr /Helvetica-iso ff 150.00 scf sf 4081 7429 m gs 1 -1 sc (ny-4) col0 sh gr /Helvetica-iso ff 150.00 scf sf 74 89 m gs 1 -1 sc ( ) col0 sh gr /Helvetica-iso ff 150.00 scf sf 4897 7429 m gs 1 -1 sc (ny-2) col0 sh gr /Helvetica-iso ff 150.00 scf sf 74 89 m gs 1 -1 sc ( ) col0 sh gr /Helvetica-iso ff 150.00 scf sf 5297 7429 m gs 1 -1 sc (ny-1) col0 sh gr /Helvetica-iso ff 150.00 scf sf 74 89 m gs 1 -1 sc ( ) col0 sh gr /Helvetica-iso ff 150.00 scf sf 4497 7429 m gs 1 -1 sc (ny-3) col0 sh gr /Helvetica-iso ff 150.00 scf sf 74 89 m gs 1 -1 sc ( ) col0 sh gr /Helvetica-iso ff 150.00 scf sf 1264 7795 m gs 1 -1 sc (ny+2) col0 sh gr /Helvetica-iso ff 150.00 scf sf 74 89 m gs 1 -1 sc ( ) col0 sh gr /Helvetica-iso ff 150.00 scf sf 1664 7795 m gs 1 -1 sc (ny+3) col0 sh gr /Helvetica-iso ff 150.00 scf sf 74 89 m gs 1 -1 sc ( ) col0 sh gr /Helvetica-iso ff 150.00 scf sf 5681 7429 m gs 1 -1 sc (ny) col0 sh gr /Helvetica-iso ff 150.00 scf sf 74 89 m gs 1 -1 sc ( ) col0 sh gr /Helvetica-iso ff 150.00 scf sf 6081 7429 m gs 1 -1 sc (ny+1) col0 sh gr % Polyline 0.000 slw n 5226 5196 m 5623 5196 l 5623 5572 l 5226 5572 l 5226 5196 l cp gs col36 1.00 shd ef gr % Polyline n 4827 5196 m 5226 5196 l 5226 5572 l 4827 5572 l 4827 5196 l cp gs col37 1.00 shd ef gr % Polyline 7.500 slw n 4835 5194 m 5631 5194 l 5631 5569 l 4835 5569 l 4835 5194 l cp gs col38 s gr % Polyline 0.000 slw n 4434 5196 m 4832 5196 l 4832 5572 l 4434 5572 l 4434 5196 l cp gs col36 1.00 shd ef gr % Polyline n 4035 5196 m 4434 5196 l 4434 5572 l 4035 5572 l 4035 5196 l cp gs col37 1.00 shd ef gr % Polyline 7.500 slw n 4044 5194 m 4840 5194 l 4840 5569 l 4044 5569 l 4044 5194 l cp gs col38 s gr % Polyline 0.000 slw n 6026 5196 m 6440 5196 l 6440 5572 l 6026 5572 l 6026 5196 l cp gs col36 1.00 shd ef gr % Polyline n 5627 5196 m 6026 5196 l 6026 5572 l 5627 5572 l 5627 5196 l cp gs col37 1.00 shd ef gr % Polyline 7.500 slw n 5635 5194 m 6440 5194 l 6440 5569 l 5635 5569 l 5635 5194 l cp gs col38 s gr % Polyline 0.000 slw n 5226 5571 m 5623 5571 l 5623 5947 l 5226 5947 l 5226 5571 l cp gs col36 1.00 shd ef gr % Polyline n 4827 5571 m 5226 5571 l 5226 5947 l 4827 5947 l 4827 5571 l cp gs col37 1.00 shd ef gr % Polyline 7.500 slw n 4835 5569 m 5631 5569 l 5631 5944 l 4835 5944 l 4835 5569 l cp gs col38 s gr % Polyline 0.000 slw n 4434 5571 m 4832 5571 l 4832 5947 l 4434 5947 l 4434 5571 l cp gs col36 1.00 shd ef gr % Polyline n 4035 5571 m 4434 5571 l 4434 5947 l 4035 5947 l 4035 5571 l cp gs col37 1.00 shd ef gr % Polyline 7.500 slw n 4044 5569 m 4840 5569 l 4840 5944 l 4044 5944 l 4044 5569 l cp gs col38 s gr % Polyline 0.000 slw n 6026 5571 m 6440 5571 l 6440 5947 l 6026 5947 l 6026 5571 l cp gs col36 1.00 shd ef gr % Polyline n 5627 5571 m 6026 5571 l 6026 5947 l 5627 5947 l 5627 5571 l cp gs col37 1.00 shd ef gr % Polyline 7.500 slw n 5635 5569 m 6440 5569 l 6440 5944 l 5635 5944 l 5635 5569 l cp gs col38 s gr % Polyline 0.000 slw n 2409 5571 m 2807 5571 l 2807 5947 l 2409 5947 l 2409 5571 l cp gs col36 1.00 shd ef gr % Polyline n 2010 5571 m 2409 5571 l 2409 5947 l 2010 5947 l 2010 5571 l cp gs col37 1.00 shd ef gr % Polyline 7.500 slw n 2019 5561 m 2815 5561 l 2815 5936 l 2019 5936 l 2019 5561 l cp gs col38 s gr % Polyline 0.000 slw n 1618 5571 m 2015 5571 l 2015 5947 l 1618 5947 l 1618 5571 l cp gs col36 1.00 shd ef gr % Polyline n 1219 5571 m 1618 5571 l 1618 5947 l 1219 5947 l 1219 5571 l cp gs col37 1.00 shd ef gr % Polyline 7.500 slw n 1227 5561 m 2023 5561 l 2023 5939 l 1227 5939 l 1227 5561 l cp gs col38 s gr % Polyline 0.000 slw n 2409 5196 m 2807 5196 l 2807 5572 l 2409 5572 l 2409 5196 l cp gs col36 1.00 shd ef gr % Polyline n 2010 5196 m 2409 5196 l 2409 5572 l 2010 5572 l 2010 5196 l cp gs col37 1.00 shd ef gr % Polyline 7.500 slw n 2019 5186 m 2815 5186 l 2815 5561 l 2019 5561 l 2019 5186 l cp gs col38 s gr % Polyline 0.000 slw n 1618 5196 m 2015 5196 l 2015 5572 l 1618 5572 l 1618 5196 l cp gs col36 1.00 shd ef gr % Polyline n 1219 5196 m 1618 5196 l 1618 5572 l 1219 5572 l 1219 5196 l cp gs col37 1.00 shd ef gr % Polyline 7.500 slw n 1227 5186 m 2023 5186 l 2023 5561 l 1227 5561 l 1227 5186 l cp gs col38 s gr % Polyline 0.000 slw n 5226 3546 m 5623 3546 l 5623 3922 l 5226 3922 l 5226 3546 l cp gs col36 1.00 shd ef gr % Polyline n 4827 3546 m 5226 3546 l 5226 3922 l 4827 3922 l 4827 3546 l cp gs col37 1.00 shd ef gr % Polyline 7.500 slw n 4835 3544 m 5631 3544 l 5631 3919 l 4835 3919 l 4835 3544 l cp gs col32 s gr % Polyline 0.000 slw n 4434 3546 m 4832 3546 l 4832 3922 l 4434 3922 l 4434 3546 l cp gs col36 1.00 shd ef gr % Polyline n 4035 3546 m 4434 3546 l 4434 3922 l 4035 3922 l 4035 3546 l cp gs col37 1.00 shd ef gr % Polyline 7.500 slw n 4044 3544 m 4840 3544 l 4840 3919 l 4044 3919 l 4044 3544 l cp gs col38 s gr % Polyline 0.000 slw n 5990 3546 m 6432 3546 l 6432 3955 l 5990 3955 l 5990 3546 l cp gs col36 1.00 shd ef gr % Polyline n 5627 3546 m 6026 3546 l 6026 3922 l 5627 3922 l 5627 3546 l cp gs col37 1.00 shd ef gr % Polyline 7.500 slw n 5635 3544 m 6440 3544 l 6440 3919 l 5635 3919 l 5635 3544 l cp gs col38 s gr % Polyline 0.000 slw n 5226 3921 m 5623 3921 l 5623 4297 l 5226 4297 l 5226 3921 l cp gs col36 1.00 shd ef gr % Polyline n 4827 3921 m 5226 3921 l 5226 4297 l 4827 4297 l 4827 3921 l cp gs col37 1.00 shd ef gr % Polyline 7.500 slw n 4835 3919 m 5631 3919 l 5631 4294 l 4835 4294 l 4835 3919 l cp gs col38 s gr % Polyline 0.000 slw n 4434 3921 m 4832 3921 l 4832 4297 l 4434 4297 l 4434 3921 l cp gs col36 1.00 shd ef gr % Polyline n 4035 3921 m 4434 3921 l 4434 4297 l 4035 4297 l 4035 3921 l cp gs col37 1.00 shd ef gr % Polyline 7.500 slw n 4044 3919 m 4840 3919 l 4840 4294 l 4044 4294 l 4044 3919 l cp gs col38 s gr % Polyline 0.000 slw n 6026 3921 m 6432 3921 l 6432 4297 l 6026 4297 l 6026 3921 l cp gs col36 1.00 shd ef gr % Polyline n 5627 3921 m 6026 3921 l 6026 4297 l 5627 4297 l 5627 3921 l cp gs col37 1.00 shd ef gr % Polyline 7.500 slw n 5635 3919 m 6440 3919 l 6440 4294 l 5635 4294 l 5635 3919 l cp gs col38 s gr % Polyline 0.000 slw n 2409 3921 m 2807 3921 l 2807 4297 l 2409 4297 l 2409 3921 l cp gs col36 1.00 shd ef gr % Polyline n 2010 3921 m 2409 3921 l 2409 4297 l 2010 4297 l 2010 3921 l cp gs col37 1.00 shd ef gr % Polyline 7.500 slw n 2019 3919 m 2815 3919 l 2815 4294 l 2019 4294 l 2019 3919 l cp gs col38 s gr % Polyline 0.000 slw n 1618 3921 m 2015 3921 l 2015 4297 l 1618 4297 l 1618 3921 l cp gs col36 1.00 shd ef gr % Polyline n 1219 3921 m 1618 3921 l 1618 4297 l 1219 4297 l 1219 3921 l cp gs col37 1.00 shd ef gr % Polyline 7.500 slw n 1227 3919 m 2023 3919 l 2023 4294 l 1227 4294 l 1227 3919 l cp gs col38 s gr % Polyline 0.000 slw n 2409 3546 m 2815 3546 l 2815 3922 l 2409 3922 l 2409 3546 l cp gs col36 1.00 shd ef gr % Polyline n 2010 3546 m 2409 3546 l 2409 3922 l 2010 3922 l 2010 3546 l cp gs col37 1.00 shd ef gr % Polyline 7.500 slw n 2019 3544 m 2815 3544 l 2815 3919 l 2019 3919 l 2019 3544 l cp gs col38 s gr % Polyline 0.000 slw n 1618 3546 m 2015 3546 l 2015 3922 l 1618 3922 l 1618 3546 l cp gs col36 1.00 shd ef gr % Polyline n 1219 3546 m 1618 3546 l 1618 3922 l 1219 3922 l 1219 3546 l cp gs col37 1.00 shd ef gr % Polyline 7.500 slw n 1227 3544 m 2023 3544 l 2023 3919 l 1227 3919 l 1227 3544 l cp gs col38 s gr % Polyline n 1221 3546 m 6440 3546 l 6440 5941 l 1221 5941 l 1221 3546 l cp gs col32 s gr /Helvetica-iso ff 225.00 scf sf 74 89 m gs 1 -1 sc ( ) col0 sh gr /Helvetica-iso ff 225.00 scf sf 2064 3993 m gs 1 -1 sc (...) col0 sh gr % Polyline n 2819 3915 m 2869 3915 l gs col32 s gr % Polyline n 2952 3915 m 3002 3915 l gs col32 s gr % Polyline n 3085 3915 m 3135 3915 l gs col32 s gr % Polyline n 3219 3915 m 3252 3915 l gs col32 s gr % Polyline n 2819 4290 m 2869 4290 l gs col32 s gr % Polyline n 2952 4290 m 3002 4290 l gs col32 s gr % Polyline n 3085 4290 m 3135 4290 l gs col32 s gr % Polyline n 3219 4290 m 3252 4290 l gs col32 s gr % Polyline n 2819 4282 m 2819 4332 l gs col32 s gr % Polyline n 2819 4415 m 2819 4465 l gs col32 s gr % Polyline n 2819 4548 m 2819 4598 l gs col32 s gr % Polyline n 2019 4282 m 2019 4332 l gs col32 s gr % Polyline n 2019 4415 m 2019 4465 l gs col32 s gr % Polyline n 2019 4548 m 2019 4598 l gs col32 s gr % Polyline n 4036 3915 m 3986 3915 l gs col32 s gr % Polyline n 3902 3915 m 3852 3915 l gs col32 s gr % Polyline n 3769 3915 m 3719 3915 l gs col32 s gr % Polyline n 3636 3915 m 3602 3915 l gs col32 s gr % Polyline n 4036 4290 m 3986 4290 l gs col32 s gr % Polyline n 3902 4290 m 3852 4290 l gs col32 s gr % Polyline n 3769 4290 m 3719 4290 l gs col32 s gr % Polyline n 3636 4290 m 3602 4290 l gs col32 s gr % Polyline n 4035 4282 m 4035 4332 l gs col32 s gr % Polyline n 4035 4415 m 4035 4465 l gs col32 s gr % Polyline n 4035 4548 m 4035 4598 l gs col32 s gr % Polyline n 4835 4282 m 4835 4332 l gs col32 s gr % Polyline n 4835 4415 m 4835 4465 l gs col32 s gr % Polyline n 4835 4548 m 4835 4598 l gs col32 s gr % Polyline n 4036 5565 m 3986 5565 l gs col32 s gr % Polyline n 3902 5565 m 3852 5565 l gs col32 s gr % Polyline n 3769 5565 m 3719 5565 l gs col32 s gr % Polyline n 3636 5565 m 3602 5565 l gs col32 s gr % Polyline n 4036 5190 m 3986 5190 l gs col32 s gr % Polyline n 3902 5190 m 3852 5190 l gs col32 s gr % Polyline n 3769 5190 m 3719 5190 l gs col32 s gr % Polyline n 3636 5190 m 3602 5190 l gs col32 s gr % Polyline n 4035 5198 m 4035 5148 l gs col32 s gr % Polyline n 4035 5065 m 4035 5015 l gs col32 s gr % Polyline n 4035 4932 m 4035 4882 l gs col32 s gr % Polyline n 4835 5198 m 4835 5148 l gs col32 s gr % Polyline n 4835 5065 m 4835 5015 l gs col32 s gr % Polyline n 4835 4932 m 4835 4882 l gs col32 s gr % Polyline n 2819 5565 m 2869 5565 l gs col32 s gr % Polyline n 2952 5565 m 3002 5565 l gs col32 s gr % Polyline n 3085 5565 m 3135 5565 l gs col32 s gr % Polyline n 3219 5565 m 3252 5565 l gs col32 s gr % Polyline n 2819 5190 m 2869 5190 l gs col32 s gr % Polyline n 2952 5190 m 3002 5190 l gs col32 s gr % Polyline n 3085 5190 m 3135 5190 l gs col32 s gr % Polyline n 3219 5190 m 3252 5190 l gs col32 s gr % Polyline n 2819 5198 m 2819 5148 l gs col32 s gr % Polyline n 2819 5065 m 2819 5015 l gs col32 s gr % Polyline n 2819 4932 m 2819 4882 l gs col32 s gr % Polyline n 2019 5198 m 2019 5148 l gs col32 s gr % Polyline n 2019 5065 m 2019 5015 l gs col32 s gr % Polyline n 2019 4932 m 2019 4882 l gs col32 s gr /Helvetica-iso ff 195.00 scf sf 74 89 m gs 1 -1 sc ( ) col0 sh gr /Helvetica-iso ff 195.00 scf sf 3181 3267 m gs 1 -1 sc (ny/2+1) col0 sh gr /Helvetica-iso ff 195.00 scf sf 74 89 m gs 1 -1 sc ( ) col0 sh gr /Helvetica-iso ff 195.00 scf sf 681 4717 m gs 1 -1 sc (nx) col0 sh gr /Helvetica-iso ff 150.00 scf sf 74 89 m gs 1 -1 sc ( ) col0 sh gr /Helvetica-iso ff 150.00 scf sf 1564 3445 m gs 1 -1 sc (0) col0 sh gr /Helvetica-iso ff 150.00 scf sf 74 89 m gs 1 -1 sc ( ) col0 sh gr /Helvetica-iso ff 150.00 scf sf 5831 3445 m gs 1 -1 sc (ny/2) col0 sh gr /Helvetica-iso ff 150.00 scf sf 74 89 m gs 1 -1 sc ( ) col0 sh gr /Helvetica-iso ff 150.00 scf sf 1064 3745 m gs 1 -1 sc (0) col0 sh gr /Helvetica-iso ff 150.00 scf sf 74 89 m gs 1 -1 sc ( ) col0 sh gr /Helvetica-iso ff 150.00 scf sf 864 5745 m gs 1 -1 sc (nx-1) col0 sh gr % Polyline n 5635 4282 m 5635 4332 l gs col32 s gr % Polyline n 5635 4415 m 5635 4465 l gs col32 s gr % Polyline n 5635 4548 m 5635 4598 l gs col32 s gr % Polyline n 5635 5198 m 5635 5148 l gs col32 s gr % Polyline n 5635 5065 m 5635 5015 l gs col32 s gr % Polyline n 5635 4932 m 5635 4882 l gs col32 s gr % Polyline n 1420 3781 m 6312 3781 l gs col0 s gr % Polyline 0.000 slw n 6348 3784 m 6117 3728 l 6117 3781 l gs 0.00 setgray ef gr % Polyline 7.500 slw n 6348 3784 m 6117 3728 l 6117 3781 l gs col0 s gr % Polyline 0.000 slw n 6348 3778 m 6117 3834 l 6117 3781 l gs 0.00 setgray ef gr % Polyline 7.500 slw n 6348 3778 m 6117 3834 l 6117 3781 l gs col0 s gr % Polyline n 1420 4169 m 6312 4169 l gs col0 s gr % Polyline 0.000 slw n 6348 4172 m 6117 4116 l 6117 4169 l gs 0.00 setgray ef gr % Polyline 7.500 slw n 6348 4172 m 6117 4116 l 6117 4169 l gs col0 s gr % Polyline 0.000 slw n 6348 4166 m 6117 4222 l 6117 4169 l gs 0.00 setgray ef gr % Polyline 7.500 slw n 6348 4166 m 6117 4222 l 6117 4169 l gs col0 s gr % Polyline n 1420 5390 m 6312 5390 l gs col0 s gr % Polyline 0.000 slw n 6348 5393 m 6117 5337 l 6117 5390 l gs 0.00 setgray ef gr % Polyline 7.500 slw n 6348 5393 m 6117 5337 l 6117 5390 l gs col0 s gr % Polyline 0.000 slw n 6348 5387 m 6117 5443 l 6117 5390 l gs 0.00 setgray ef gr % Polyline 7.500 slw n 6348 5387 m 6117 5443 l 6117 5390 l gs col0 s gr % Polyline n 1420 5766 m 6312 5766 l gs col0 s gr % Polyline 0.000 slw n 6348 5769 m 6117 5713 l 6117 5766 l gs 0.00 setgray ef gr % Polyline 7.500 slw n 6348 5769 m 6117 5713 l 6117 5766 l gs col0 s gr % Polyline 0.000 slw n 6348 5763 m 6117 5819 l 6117 5766 l gs 0.00 setgray ef gr % Polyline 7.500 slw n 6348 5763 m 6117 5819 l 6117 5766 l gs col0 s gr % Polyline 0.000 slw n 1469 6215 m 1868 6215 l 1868 6591 l 1469 6591 l 1469 6215 l cp gs col33 1.00 shd ef gr % Polyline 7.500 slw n 1469 6215 m 1868 6215 l 1868 6591 l 1469 6591 l 1469 6215 l cp gs col32 s gr /Helvetica-iso ff 195.00 scf sf 74 89 m gs 1 -1 sc ( ) col0 sh gr /Helvetica-iso ff 195.00 scf sf 1981 6463 m gs 1 -1 sc (= double) col0 sh gr % Polyline 0.000 slw n 4026 6217 m 4432 6217 l 4432 6593 l 4026 6593 l 4026 6217 l cp gs col36 1.00 shd ef gr % Polyline n 3627 6217 m 4026 6217 l 4026 6593 l 3627 6593 l 3627 6217 l cp gs col37 1.00 shd ef gr % Polyline 7.500 slw n 3635 6215 m 4440 6215 l 4440 6590 l 3635 6590 l 3635 6215 l cp gs col38 s gr /Helvetica-iso ff 195.00 scf sf 74 89 m gs 1 -1 sc ( ) col0 sh gr /Helvetica-iso ff 195.00 scf sf 4547 6463 m gs 1 -1 sc (= fftw_complex) col0 sh gr /Helvetica-Bold-iso ff 240.00 scf sf 74 89 m gs 1 -1 sc ( ) col0 sh gr /Helvetica-Bold-iso ff 240.00 scf sf 428 5128 m gs 1 -1 sc 90.0 rot (output) col0 sh gr /Helvetica-iso ff 150.00 scf sf 74 89 m gs 1 -1 sc ( ) col0 sh gr /Helvetica-iso ff 150.00 scf sf 1264 3679 m gs 1 -1 sc (0) col0 sh gr /Helvetica-iso ff 150.00 scf sf 74 89 m gs 1 -1 sc ( ) col0 sh gr /Helvetica-iso ff 150.00 scf sf 2081 3679 m gs 1 -1 sc (1) col0 sh gr /Helvetica-iso ff 150.00 scf sf 74 89 m gs 1 -1 sc ( ) col0 sh gr /Helvetica-iso ff 150.00 scf sf 4097 3679 m gs 1 -1 sc (ny/2-2) col0 sh gr /Helvetica-iso ff 150.00 scf sf 74 89 m gs 1 -1 sc ( ) col0 sh gr /Helvetica-iso ff 150.00 scf sf 4914 3679 m gs 1 -1 sc (ny/2-1) col0 sh gr /Helvetica-iso ff 150.00 scf sf 74 89 m gs 1 -1 sc ( ) col0 sh gr /Helvetica-iso ff 150.00 scf sf 1264 4062 m gs 1 -1 sc (ny/2+1) col0 sh gr /Helvetica-iso ff 150.00 scf sf 74 89 m gs 1 -1 sc ( ) col0 sh gr /Helvetica-iso ff 150.00 scf sf 5697 3679 m gs 1 -1 sc (ny/2) col0 sh gr % Polyline 0.000 slw n 1221 495 m 5635 495 l 5635 2890 l 1221 2890 l 1221 495 l cp gs col7 1.00 shd ef gr % Polyline 7.500 slw n 1221 495 m 5635 495 l 5635 2890 l 1221 2890 l 1221 495 l cp gs col32 s gr % Polyline 0.000 slw n 1221 495 m 1620 495 l 1620 871 l 1221 871 l 1221 495 l cp gs col33 1.00 shd ef gr % Polyline 7.500 slw n 1221 495 m 1620 495 l 1620 871 l 1221 871 l 1221 495 l cp gs col32 s gr % Polyline 0.000 slw n 1620 495 m 2019 495 l 2019 871 l 1620 871 l 1620 495 l cp gs col33 1.00 shd ef gr % Polyline 7.500 slw n 1620 495 m 2019 495 l 2019 871 l 1620 871 l 1620 495 l cp gs col32 s gr % Polyline 0.000 slw n 2019 495 m 2418 495 l 2418 871 l 2019 871 l 2019 495 l cp gs col33 1.00 shd ef gr % Polyline 7.500 slw n 2019 495 m 2418 495 l 2418 871 l 2019 871 l 2019 495 l cp gs col32 s gr % Polyline 0.000 slw n 2418 495 m 2817 495 l 2817 871 l 2418 871 l 2418 495 l cp gs col33 1.00 shd ef gr % Polyline 7.500 slw n 2418 495 m 2817 495 l 2817 871 l 2418 871 l 2418 495 l cp gs col32 s gr % Polyline 0.000 slw n 4038 495 m 4438 495 l 4438 871 l 4038 871 l 4038 495 l cp gs col33 1.00 shd ef gr % Polyline 7.500 slw n 4038 495 m 4438 495 l 4438 871 l 4038 871 l 4038 495 l cp gs col32 s gr % Polyline 0.000 slw n 4438 495 m 4837 495 l 4837 871 l 4438 871 l 4438 495 l cp gs col33 1.00 shd ef gr % Polyline 7.500 slw n 4438 495 m 4837 495 l 4837 871 l 4438 871 l 4438 495 l cp gs col32 s gr % Polyline 0.000 slw n 4837 495 m 5236 495 l 5236 871 l 4837 871 l 4837 495 l cp gs col33 1.00 shd ef gr % Polyline 7.500 slw n 4837 495 m 5236 495 l 5236 871 l 4837 871 l 4837 495 l cp gs col32 s gr % Polyline 0.000 slw n 5236 495 m 5635 495 l 5635 871 l 5236 871 l 5236 495 l cp gs col33 1.00 shd ef gr % Polyline 7.500 slw n 5236 495 m 5635 495 l 5635 871 l 5236 871 l 5236 495 l cp gs col32 s gr % Polyline 0.000 slw n 1221 871 m 1620 871 l 1620 1247 l 1221 1247 l 1221 871 l cp gs col33 1.00 shd ef gr % Polyline 7.500 slw n 1221 871 m 1620 871 l 1620 1247 l 1221 1247 l 1221 871 l cp gs col32 s gr % Polyline 0.000 slw n 1620 871 m 2019 871 l 2019 1247 l 1620 1247 l 1620 871 l cp gs col33 1.00 shd ef gr % Polyline 7.500 slw n 1620 871 m 2019 871 l 2019 1247 l 1620 1247 l 1620 871 l cp gs col32 s gr % Polyline 0.000 slw n 2019 871 m 2418 871 l 2418 1247 l 2019 1247 l 2019 871 l cp gs col33 1.00 shd ef gr % Polyline 7.500 slw n 2019 871 m 2418 871 l 2418 1247 l 2019 1247 l 2019 871 l cp gs col32 s gr % Polyline 0.000 slw n 2418 871 m 2817 871 l 2817 1247 l 2418 1247 l 2418 871 l cp gs col33 1.00 shd ef gr % Polyline 7.500 slw n 2418 871 m 2817 871 l 2817 1247 l 2418 1247 l 2418 871 l cp gs col32 s gr % Polyline 0.000 slw n 4038 871 m 4438 871 l 4438 1247 l 4038 1247 l 4038 871 l cp gs col33 1.00 shd ef gr % Polyline 7.500 slw n 4038 871 m 4438 871 l 4438 1247 l 4038 1247 l 4038 871 l cp gs col32 s gr % Polyline 0.000 slw n 4438 871 m 4837 871 l 4837 1247 l 4438 1247 l 4438 871 l cp gs col33 1.00 shd ef gr % Polyline 7.500 slw n 4438 871 m 4837 871 l 4837 1247 l 4438 1247 l 4438 871 l cp gs col32 s gr % Polyline 0.000 slw n 4837 871 m 5236 871 l 5236 1247 l 4837 1247 l 4837 871 l cp gs col33 1.00 shd ef gr % Polyline 7.500 slw n 4837 871 m 5236 871 l 5236 1247 l 4837 1247 l 4837 871 l cp gs col32 s gr % Polyline 0.000 slw n 5236 871 m 5635 871 l 5635 1247 l 5236 1247 l 5236 871 l cp gs col33 1.00 shd ef gr % Polyline 7.500 slw n 5236 871 m 5635 871 l 5635 1247 l 5236 1247 l 5236 871 l cp gs col32 s gr % Polyline 0.000 slw n 1221 2139 m 1620 2139 l 1620 2515 l 1221 2515 l 1221 2139 l cp gs col33 1.00 shd ef gr % Polyline 7.500 slw n 1221 2139 m 1620 2139 l 1620 2515 l 1221 2515 l 1221 2139 l cp gs col32 s gr % Polyline 0.000 slw n 1620 2139 m 2019 2139 l 2019 2515 l 1620 2515 l 1620 2139 l cp gs col33 1.00 shd ef gr % Polyline 7.500 slw n 1620 2139 m 2019 2139 l 2019 2515 l 1620 2515 l 1620 2139 l cp gs col32 s gr % Polyline 0.000 slw n 2019 2139 m 2418 2139 l 2418 2515 l 2019 2515 l 2019 2139 l cp gs col33 1.00 shd ef gr % Polyline 7.500 slw n 2019 2139 m 2418 2139 l 2418 2515 l 2019 2515 l 2019 2139 l cp gs col32 s gr % Polyline 0.000 slw n 2418 2139 m 2817 2139 l 2817 2515 l 2418 2515 l 2418 2139 l cp gs col33 1.00 shd ef gr % Polyline 7.500 slw n 2418 2139 m 2817 2139 l 2817 2515 l 2418 2515 l 2418 2139 l cp gs col32 s gr % Polyline 0.000 slw n 4038 2139 m 4438 2139 l 4438 2515 l 4038 2515 l 4038 2139 l cp gs col33 1.00 shd ef gr % Polyline 7.500 slw n 4038 2139 m 4438 2139 l 4438 2515 l 4038 2515 l 4038 2139 l cp gs col32 s gr % Polyline 0.000 slw n 4438 2139 m 4837 2139 l 4837 2515 l 4438 2515 l 4438 2139 l cp gs col33 1.00 shd ef gr % Polyline 7.500 slw n 4438 2139 m 4837 2139 l 4837 2515 l 4438 2515 l 4438 2139 l cp gs col32 s gr % Polyline 0.000 slw n 4837 2139 m 5236 2139 l 5236 2515 l 4837 2515 l 4837 2139 l cp gs col33 1.00 shd ef gr % Polyline 7.500 slw n 4837 2139 m 5236 2139 l 5236 2515 l 4837 2515 l 4837 2139 l cp gs col32 s gr % Polyline 0.000 slw n 5236 2139 m 5635 2139 l 5635 2515 l 5236 2515 l 5236 2139 l cp gs col33 1.00 shd ef gr % Polyline 7.500 slw n 5236 2139 m 5635 2139 l 5635 2515 l 5236 2515 l 5236 2139 l cp gs col32 s gr % Polyline 0.000 slw n 1221 2515 m 1620 2515 l 1620 2890 l 1221 2890 l 1221 2515 l cp gs col33 1.00 shd ef gr % Polyline 7.500 slw n 1221 2515 m 1620 2515 l 1620 2890 l 1221 2890 l 1221 2515 l cp gs col32 s gr % Polyline 0.000 slw n 1620 2515 m 2019 2515 l 2019 2890 l 1620 2890 l 1620 2515 l cp gs col33 1.00 shd ef gr % Polyline 7.500 slw n 1620 2515 m 2019 2515 l 2019 2890 l 1620 2890 l 1620 2515 l cp gs col32 s gr % Polyline 0.000 slw n 2019 2515 m 2418 2515 l 2418 2890 l 2019 2890 l 2019 2515 l cp gs col33 1.00 shd ef gr % Polyline 7.500 slw n 2019 2515 m 2418 2515 l 2418 2890 l 2019 2890 l 2019 2515 l cp gs col32 s gr % Polyline 0.000 slw n 2418 2515 m 2817 2515 l 2817 2890 l 2418 2890 l 2418 2515 l cp gs col33 1.00 shd ef gr % Polyline 7.500 slw n 2418 2515 m 2817 2515 l 2817 2890 l 2418 2890 l 2418 2515 l cp gs col32 s gr % Polyline 0.000 slw n 4038 2515 m 4438 2515 l 4438 2890 l 4038 2890 l 4038 2515 l cp gs col33 1.00 shd ef gr % Polyline 7.500 slw n 4038 2515 m 4438 2515 l 4438 2890 l 4038 2890 l 4038 2515 l cp gs col32 s gr % Polyline 0.000 slw n 4438 2515 m 4837 2515 l 4837 2890 l 4438 2890 l 4438 2515 l cp gs col33 1.00 shd ef gr % Polyline 7.500 slw n 4438 2515 m 4837 2515 l 4837 2890 l 4438 2890 l 4438 2515 l cp gs col32 s gr % Polyline 0.000 slw n 4837 2515 m 5236 2515 l 5236 2890 l 4837 2890 l 4837 2515 l cp gs col33 1.00 shd ef gr % Polyline 7.500 slw n 4837 2515 m 5236 2515 l 5236 2890 l 4837 2890 l 4837 2515 l cp gs col32 s gr % Polyline 0.000 slw n 5236 2515 m 5635 2515 l 5635 2890 l 5236 2890 l 5236 2515 l cp gs col33 1.00 shd ef gr % Polyline 7.500 slw n 5236 2515 m 5635 2515 l 5635 2890 l 5236 2890 l 5236 2515 l cp gs col32 s gr % Polyline n 1420 730 m 5459 730 l gs col0 s gr % Polyline 0.000 slw n 5488 733 m 5298 677 l 5298 730 l gs 0.00 setgray ef gr % Polyline 7.500 slw n 5488 733 m 5298 677 l 5298 730 l gs col0 s gr % Polyline 0.000 slw n 5488 727 m 5298 783 l 5298 730 l gs 0.00 setgray ef gr % Polyline 7.500 slw n 5488 727 m 5298 783 l 5298 730 l gs col0 s gr /Helvetica-iso ff 225.00 scf sf 74 89 m gs 1 -1 sc ( ) col0 sh gr /Helvetica-iso ff 225.00 scf sf 2064 943 m gs 1 -1 sc (...) col0 sh gr % Polyline n 2819 873 m 2869 873 l gs col32 s gr % Polyline n 2952 873 m 3002 873 l gs col32 s gr % Polyline n 3085 873 m 3135 873 l gs col32 s gr % Polyline n 3219 873 m 3252 873 l gs col32 s gr % Polyline n 2819 1248 m 2869 1248 l gs col32 s gr % Polyline n 2952 1248 m 3002 1248 l gs col32 s gr % Polyline n 3085 1248 m 3135 1248 l gs col32 s gr % Polyline n 3219 1248 m 3252 1248 l gs col32 s gr % Polyline n 2819 1240 m 2819 1290 l gs col32 s gr % Polyline n 2819 1373 m 2819 1423 l gs col32 s gr % Polyline n 2819 1506 m 2819 1556 l gs col32 s gr % Polyline n 2419 1240 m 2419 1290 l gs col32 s gr % Polyline n 2419 1373 m 2419 1423 l gs col32 s gr % Polyline n 2419 1506 m 2419 1556 l gs col32 s gr % Polyline n 2019 1240 m 2019 1290 l gs col32 s gr % Polyline n 2019 1373 m 2019 1423 l gs col32 s gr % Polyline n 2019 1506 m 2019 1556 l gs col32 s gr % Polyline n 1619 1240 m 1619 1290 l gs col32 s gr % Polyline n 1619 1373 m 1619 1423 l gs col32 s gr % Polyline n 1619 1506 m 1619 1556 l gs col32 s gr % Polyline n 4036 873 m 3986 873 l gs col32 s gr % Polyline n 3902 873 m 3852 873 l gs col32 s gr % Polyline n 3769 873 m 3719 873 l gs col32 s gr % Polyline n 3636 873 m 3602 873 l gs col32 s gr % Polyline n 4036 1248 m 3986 1248 l gs col32 s gr % Polyline n 3902 1248 m 3852 1248 l gs col32 s gr % Polyline n 3769 1248 m 3719 1248 l gs col32 s gr % Polyline n 3636 1248 m 3602 1248 l gs col32 s gr % Polyline n 4035 1240 m 4035 1290 l gs col32 s gr % Polyline n 4035 1373 m 4035 1423 l gs col32 s gr % Polyline n 4035 1506 m 4035 1556 l gs col32 s gr % Polyline n 4435 1240 m 4435 1290 l gs col32 s gr % Polyline n 4435 1373 m 4435 1423 l gs col32 s gr % Polyline n 4435 1506 m 4435 1556 l gs col32 s gr % Polyline n 4835 1240 m 4835 1290 l gs col32 s gr % Polyline n 4835 1373 m 4835 1423 l gs col32 s gr % Polyline n 4835 1506 m 4835 1556 l gs col32 s gr % Polyline n 5235 1240 m 5235 1290 l gs col32 s gr % Polyline n 5235 1373 m 5235 1423 l gs col32 s gr % Polyline n 5235 1506 m 5235 1556 l gs col32 s gr % Polyline n 4036 2515 m 3986 2515 l gs col32 s gr % Polyline n 3902 2515 m 3852 2515 l gs col32 s gr % Polyline n 3769 2515 m 3719 2515 l gs col32 s gr % Polyline n 3636 2515 m 3602 2515 l gs col32 s gr % Polyline n 4036 2140 m 3986 2140 l gs col32 s gr % Polyline n 3902 2140 m 3852 2140 l gs col32 s gr % Polyline n 3769 2140 m 3719 2140 l gs col32 s gr % Polyline n 3636 2140 m 3602 2140 l gs col32 s gr % Polyline n 4035 2148 m 4035 2098 l gs col32 s gr % Polyline n 4035 2015 m 4035 1965 l gs col32 s gr % Polyline n 4035 1881 m 4035 1831 l gs col32 s gr % Polyline n 4435 2148 m 4435 2098 l gs col32 s gr % Polyline n 4435 2015 m 4435 1965 l gs col32 s gr % Polyline n 4435 1881 m 4435 1831 l gs col32 s gr % Polyline n 4835 2148 m 4835 2098 l gs col32 s gr % Polyline n 4835 2015 m 4835 1965 l gs col32 s gr % Polyline n 4835 1881 m 4835 1831 l gs col32 s gr % Polyline n 5235 2148 m 5235 2098 l gs col32 s gr % Polyline n 5235 2015 m 5235 1965 l gs col32 s gr % Polyline n 5235 1881 m 5235 1831 l gs col32 s gr % Polyline n 2819 2515 m 2869 2515 l gs col32 s gr % Polyline n 2952 2515 m 3002 2515 l gs col32 s gr % Polyline n 3085 2515 m 3135 2515 l gs col32 s gr % Polyline n 3219 2515 m 3252 2515 l gs col32 s gr % Polyline n 2819 2140 m 2869 2140 l gs col32 s gr % Polyline n 2952 2140 m 3002 2140 l gs col32 s gr % Polyline n 3085 2140 m 3135 2140 l gs col32 s gr % Polyline n 3219 2140 m 3252 2140 l gs col32 s gr % Polyline n 2819 2148 m 2819 2098 l gs col32 s gr % Polyline n 2819 2015 m 2819 1965 l gs col32 s gr % Polyline n 2819 1881 m 2819 1831 l gs col32 s gr % Polyline n 2419 2148 m 2419 2098 l gs col32 s gr % Polyline n 2419 2015 m 2419 1965 l gs col32 s gr % Polyline n 2419 1881 m 2419 1831 l gs col32 s gr % Polyline n 2019 2148 m 2019 2098 l gs col32 s gr % Polyline n 2019 2015 m 2019 1965 l gs col32 s gr % Polyline n 2019 1881 m 2019 1831 l gs col32 s gr % Polyline n 1619 2148 m 1619 2098 l gs col32 s gr % Polyline n 1619 2015 m 1619 1965 l gs col32 s gr % Polyline n 1619 1881 m 1619 1831 l gs col32 s gr /Helvetica-iso ff 195.00 scf sf 74 89 m gs 1 -1 sc ( ) col0 sh gr /Helvetica-iso ff 195.00 scf sf 3381 217 m gs 1 -1 sc (ny) col0 sh gr /Helvetica-iso ff 195.00 scf sf 74 89 m gs 1 -1 sc ( ) col0 sh gr /Helvetica-iso ff 195.00 scf sf 681 1667 m gs 1 -1 sc (nx) col0 sh gr /Helvetica-iso ff 150.00 scf sf 74 89 m gs 1 -1 sc ( ) col0 sh gr /Helvetica-iso ff 150.00 scf sf 1364 395 m gs 1 -1 sc (0) col0 sh gr /Helvetica-iso ff 150.00 scf sf 74 89 m gs 1 -1 sc ( ) col0 sh gr /Helvetica-iso ff 150.00 scf sf 5281 395 m gs 1 -1 sc (ny-1) col0 sh gr /Helvetica-iso ff 150.00 scf sf 74 89 m gs 1 -1 sc ( ) col0 sh gr /Helvetica-iso ff 150.00 scf sf 1064 695 m gs 1 -1 sc (0) col0 sh gr /Helvetica-iso ff 150.00 scf sf 74 89 m gs 1 -1 sc ( ) col0 sh gr /Helvetica-iso ff 150.00 scf sf 864 2695 m gs 1 -1 sc (nx-1) col0 sh gr % Polyline n 1420 1106 m 5459 1106 l gs col0 s gr % Polyline 0.000 slw n 5488 1109 m 5298 1053 l 5298 1106 l gs 0.00 setgray ef gr % Polyline 7.500 slw n 5488 1109 m 5298 1053 l 5298 1106 l gs col0 s gr % Polyline 0.000 slw n 5488 1103 m 5298 1159 l 5298 1106 l gs 0.00 setgray ef gr % Polyline 7.500 slw n 5488 1103 m 5298 1159 l 5298 1106 l gs col0 s gr % Polyline n 1420 2327 m 5459 2327 l gs col0 s gr % Polyline 0.000 slw n 5488 2330 m 5298 2274 l 5298 2327 l gs 0.00 setgray ef gr % Polyline 7.500 slw n 5488 2330 m 5298 2274 l 5298 2327 l gs col0 s gr % Polyline 0.000 slw n 5488 2324 m 5298 2380 l 5298 2327 l gs 0.00 setgray ef gr % Polyline 7.500 slw n 5488 2324 m 5298 2380 l 5298 2327 l gs col0 s gr % Polyline n 1420 2703 m 5459 2703 l gs col0 s gr % Polyline 0.000 slw n 5488 2706 m 5298 2650 l 5298 2703 l gs 0.00 setgray ef gr % Polyline 7.500 slw n 5488 2706 m 5298 2650 l 5298 2703 l gs col0 s gr % Polyline 0.000 slw n 5488 2700 m 5298 2755 l 5298 2703 l gs 0.00 setgray ef gr % Polyline 7.500 slw n 5488 2700 m 5298 2755 l 5298 2703 l gs col0 s gr /Helvetica-Bold-iso ff 240.00 scf sf 74 89 m gs 1 -1 sc ( ) col0 sh gr /Helvetica-Bold-iso ff 240.00 scf sf 428 2734 m gs 1 -1 sc 90.0 rot (input, out-of-place) col0 sh gr /Helvetica-iso ff 150.00 scf sf 74 89 m gs 1 -1 sc ( ) col0 sh gr /Helvetica-iso ff 150.00 scf sf 1264 629 m gs 1 -1 sc (0) col0 sh gr /Helvetica-iso ff 150.00 scf sf 74 89 m gs 1 -1 sc ( ) col0 sh gr /Helvetica-iso ff 150.00 scf sf 1681 629 m gs 1 -1 sc (1) col0 sh gr /Helvetica-iso ff 150.00 scf sf 74 89 m gs 1 -1 sc ( ) col0 sh gr /Helvetica-iso ff 150.00 scf sf 2081 629 m gs 1 -1 sc (2) col0 sh gr /Helvetica-iso ff 150.00 scf sf 74 89 m gs 1 -1 sc ( ) col0 sh gr /Helvetica-iso ff 150.00 scf sf 2481 629 m gs 1 -1 sc (3) col0 sh gr /Helvetica-iso ff 150.00 scf sf 74 89 m gs 1 -1 sc ( ) col0 sh gr /Helvetica-iso ff 150.00 scf sf 4081 629 m gs 1 -1 sc (ny-4) col0 sh gr /Helvetica-iso ff 150.00 scf sf 74 89 m gs 1 -1 sc ( ) col0 sh gr /Helvetica-iso ff 150.00 scf sf 4897 629 m gs 1 -1 sc (ny-2) col0 sh gr /Helvetica-iso ff 150.00 scf sf 74 89 m gs 1 -1 sc ( ) col0 sh gr /Helvetica-iso ff 150.00 scf sf 5297 629 m gs 1 -1 sc (ny-1) col0 sh gr /Helvetica-iso ff 150.00 scf sf 74 89 m gs 1 -1 sc ( ) col0 sh gr /Helvetica-iso ff 150.00 scf sf 4497 629 m gs 1 -1 sc (ny-3) col0 sh gr /Helvetica-iso ff 150.00 scf sf 74 89 m gs 1 -1 sc ( ) col0 sh gr /Helvetica-iso ff 150.00 scf sf 1264 1012 m gs 1 -1 sc (ny) col0 sh gr /Helvetica-iso ff 150.00 scf sf 1425 4800 m gs 1 -1 sc ( ) col0 sh gr /Helvetica-iso ff 150.00 scf sf 1664 1012 m gs 1 -1 sc (ny+1) col0 sh gr % Polyline n 273 3662 m 273 3039 l gs col0 s gr % Polyline n 382 3920 m 156 3662 l gs col0 s gr % Polyline n 273 3662 m 148 3662 l gs col0 s gr % Polyline n 487 3662 m 487 3039 l gs col0 s gr % Polyline n 378 3920 m 604 3662 l gs col0 s gr % Polyline n 487 3662 m 612 3662 l gs col0 s gr % Polyline n 273 6130 m 273 6753 l gs col0 s gr % Polyline n 382 5872 m 156 6130 l gs col0 s gr % Polyline n 273 6130 m 148 6130 l gs col0 s gr % Polyline n 487 6129 m 487 6753 l gs col0 s gr % Polyline n 378 5872 m 604 6129 l gs col0 s gr % Polyline n 487 6129 m 612 6129 l gs col0 s gr % here ends figure; pagefooter showpage %%Trailer %EOF fftw-3.3.3/doc/Makefile.am0000644000175000017500000000256412054410050012206 00000000000000SUBDIRS = FAQ info_TEXINFOS = fftw3.texi fftw3_TEXINFOS = acknowledgements.texi cindex.texi fftw3.texi findex.texi install.texi intro.texi legacy-fortran.texi license.texi modern-fortran.texi mpi.texi other.texi reference.texi threads.texi tutorial.texi upgrading.texi version.texi rfftwnd.pdf rfftwnd.eps DVIPS = dvips -Pwww EQN_IMAGES = equation-dft.png equation-dht.png equation-idft.png \ equation-redft00.png equation-redft01.png equation-redft10.png \ equation-redft11.png equation-rodft00.png equation-rodft01.png \ equation-rodft10.png equation-rodft11.png EXTRA_DIST = f77_wisdom.f fftw3.pdf html rfftwnd.fig rfftwnd.eps \ rfftwnd.pdf rfftwnd-for-html.png $(EQN_IMAGES) html: $(fftw3_TEXINFOS) $(EQN_IMAGES) rfftwnd-for-html.png $(MAKEINFO) $(AM_MAKEINFOFLAGS) $(MAKEINFOFLAGS) -I $(srcdir) \ --html --number-sections -o html fftw3.texi for i in $(EQN_IMAGES); do cp -f ${srcdir}/$$i html; done cp -f rfftwnd-for-html.png html maintainer-clean-local: rm -rf html if MAINTAINER_MODE # generate the figure for the manual and distribute the binaries, so that # people don't need to have fig2dev installed. rfftwnd.eps: rfftwnd.fig fig2dev -L eps -m .7 ${srcdir}/rfftwnd.fig rfftwnd.eps rfftwnd-for-html.png: rfftwnd.fig fig2dev -L png -m 1 ${srcdir}/rfftwnd.fig rfftwnd-for-html.png rfftwnd.pdf: rfftwnd.fig fig2dev -L pdf -m .7 ${srcdir}/rfftwnd.fig rfftwnd.pdf endif fftw-3.3.3/doc/Makefile.in0000644000175000017500000007611112054410200012213 00000000000000# Makefile.in generated by automake 1.11.6 from Makefile.am. # @configure_input@ # Copyright (C) 1994, 1995, 1996, 1997, 1998, 1999, 2000, 2001, 2002, # 2003, 2004, 2005, 2006, 2007, 2008, 2009, 2010, 2011 Free Software # Foundation, Inc. # This Makefile.in is free software; the Free Software Foundation # gives unlimited permission to copy and/or distribute it, # with or without modifications, as long as this notice is preserved. # This program is distributed in the hope that it will be useful, # but WITHOUT ANY WARRANTY, to the extent permitted by law; without # even the implied warranty of MERCHANTABILITY or FITNESS FOR A # PARTICULAR PURPOSE. @SET_MAKE@ VPATH = @srcdir@ am__make_dryrun = \ { \ am__dry=no; \ case $$MAKEFLAGS in \ *\\[\ \ ]*) \ echo 'am--echo: ; @echo "AM" OK' | $(MAKE) -f - 2>/dev/null \ | grep '^AM OK$$' >/dev/null || am__dry=yes;; \ *) \ for am__flg in $$MAKEFLAGS; do \ case $$am__flg in \ *=*|--*) ;; \ *n*) am__dry=yes; break;; \ esac; \ done;; \ esac; \ test $$am__dry = yes; \ } pkgdatadir = $(datadir)/@PACKAGE@ pkgincludedir = $(includedir)/@PACKAGE@ pkglibdir = $(libdir)/@PACKAGE@ pkglibexecdir = $(libexecdir)/@PACKAGE@ am__cd = CDPATH="$${ZSH_VERSION+.}$(PATH_SEPARATOR)" && cd install_sh_DATA = $(install_sh) -c -m 644 install_sh_PROGRAM = $(install_sh) -c install_sh_SCRIPT = $(install_sh) -c INSTALL_HEADER = $(INSTALL_DATA) transform = $(program_transform_name) NORMAL_INSTALL = : PRE_INSTALL = : POST_INSTALL = : NORMAL_UNINSTALL = : PRE_UNINSTALL = : POST_UNINSTALL = : build_triplet = @build@ host_triplet = @host@ subdir = doc DIST_COMMON = $(fftw3_TEXINFOS) $(srcdir)/Makefile.am \ $(srcdir)/Makefile.in $(srcdir)/stamp-vti \ $(srcdir)/version.texi mdate-sh texinfo.tex ACLOCAL_M4 = $(top_srcdir)/aclocal.m4 am__aclocal_m4_deps = $(top_srcdir)/m4/acx_mpi.m4 \ $(top_srcdir)/m4/acx_pthread.m4 \ $(top_srcdir)/m4/ax_cc_maxopt.m4 \ $(top_srcdir)/m4/ax_check_compiler_flags.m4 \ $(top_srcdir)/m4/ax_compiler_vendor.m4 \ $(top_srcdir)/m4/ax_gcc_aligns_stack.m4 \ $(top_srcdir)/m4/ax_gcc_version.m4 \ $(top_srcdir)/m4/ax_openmp.m4 $(top_srcdir)/m4/libtool.m4 \ $(top_srcdir)/m4/ltoptions.m4 $(top_srcdir)/m4/ltsugar.m4 \ $(top_srcdir)/m4/ltversion.m4 $(top_srcdir)/m4/lt~obsolete.m4 \ $(top_srcdir)/configure.ac am__configure_deps = $(am__aclocal_m4_deps) $(CONFIGURE_DEPENDENCIES) \ $(ACLOCAL_M4) mkinstalldirs = $(install_sh) -d CONFIG_HEADER = $(top_builddir)/config.h CONFIG_CLEAN_FILES = CONFIG_CLEAN_VPATH_FILES = SOURCES = DIST_SOURCES = INFO_DEPS = $(srcdir)/fftw3.info am__TEXINFO_TEX_DIR = $(srcdir) DVIS = fftw3.dvi PDFS = fftw3.pdf PSS = fftw3.ps HTMLS = fftw3.html TEXINFOS = fftw3.texi TEXI2DVI = texi2dvi TEXI2PDF = $(TEXI2DVI) --pdf --batch MAKEINFOHTML = $(MAKEINFO) --html AM_MAKEINFOHTMLFLAGS = $(AM_MAKEINFOFLAGS) RECURSIVE_TARGETS = all-recursive check-recursive dvi-recursive \ html-recursive info-recursive install-data-recursive \ install-dvi-recursive install-exec-recursive \ install-html-recursive install-info-recursive \ install-pdf-recursive install-ps-recursive install-recursive \ installcheck-recursive installdirs-recursive pdf-recursive \ ps-recursive uninstall-recursive am__can_run_installinfo = \ case $$AM_UPDATE_INFO_DIR in \ n|no|NO) false;; \ *) (install-info --version) >/dev/null 2>&1;; \ esac am__installdirs = "$(DESTDIR)$(infodir)" am__vpath_adj_setup = srcdirstrip=`echo "$(srcdir)" | sed 's|.|.|g'`; am__vpath_adj = case $$p in \ $(srcdir)/*) f=`echo "$$p" | sed "s|^$$srcdirstrip/||"`;; \ *) f=$$p;; \ esac; am__strip_dir = f=`echo $$p | sed -e 's|^.*/||'`; am__install_max = 40 am__nobase_strip_setup = \ srcdirstrip=`echo "$(srcdir)" | sed 's/[].[^$$\\*|]/\\\\&/g'` am__nobase_strip = \ for p in $$list; do echo "$$p"; done | sed -e "s|$$srcdirstrip/||" am__nobase_list = $(am__nobase_strip_setup); \ for p in $$list; do echo "$$p $$p"; done | \ sed "s| $$srcdirstrip/| |;"' / .*\//!s/ .*/ ./; s,\( .*\)/[^/]*$$,\1,' | \ $(AWK) 'BEGIN { files["."] = "" } { files[$$2] = files[$$2] " " $$1; \ if (++n[$$2] == $(am__install_max)) \ { print $$2, files[$$2]; n[$$2] = 0; files[$$2] = "" } } \ END { for (dir in files) print dir, files[dir] }' am__base_list = \ sed '$$!N;$$!N;$$!N;$$!N;$$!N;$$!N;$$!N;s/\n/ /g' | \ sed '$$!N;$$!N;$$!N;$$!N;s/\n/ /g' am__uninstall_files_from_dir = { \ test -z "$$files" \ || { test ! -d "$$dir" && test ! -f "$$dir" && test ! -r "$$dir"; } \ || { echo " ( cd '$$dir' && rm -f" $$files ")"; \ $(am__cd) "$$dir" && rm -f $$files; }; \ } RECURSIVE_CLEAN_TARGETS = mostlyclean-recursive clean-recursive \ distclean-recursive maintainer-clean-recursive AM_RECURSIVE_TARGETS = $(RECURSIVE_TARGETS:-recursive=) \ $(RECURSIVE_CLEAN_TARGETS:-recursive=) tags TAGS ctags CTAGS \ distdir ETAGS = etags CTAGS = ctags DIST_SUBDIRS = $(SUBDIRS) DISTFILES = $(DIST_COMMON) $(DIST_SOURCES) $(TEXINFOS) $(EXTRA_DIST) am__relativize = \ dir0=`pwd`; \ sed_first='s,^\([^/]*\)/.*$$,\1,'; \ sed_rest='s,^[^/]*/*,,'; \ sed_last='s,^.*/\([^/]*\)$$,\1,'; \ sed_butlast='s,/*[^/]*$$,,'; \ while test -n "$$dir1"; do \ first=`echo "$$dir1" | sed -e "$$sed_first"`; \ if test "$$first" != "."; then \ if test "$$first" = ".."; then \ dir2=`echo "$$dir0" | sed -e "$$sed_last"`/"$$dir2"; \ dir0=`echo "$$dir0" | sed -e "$$sed_butlast"`; \ else \ first2=`echo "$$dir2" | sed -e "$$sed_first"`; \ if test "$$first2" = "$$first"; then \ dir2=`echo "$$dir2" | sed -e "$$sed_rest"`; \ else \ dir2="../$$dir2"; \ fi; \ dir0="$$dir0"/"$$first"; \ fi; \ fi; \ dir1=`echo "$$dir1" | sed -e "$$sed_rest"`; \ done; \ reldir="$$dir2" ACLOCAL = @ACLOCAL@ ALLOCA = @ALLOCA@ ALTIVEC_CFLAGS = @ALTIVEC_CFLAGS@ AMTAR = @AMTAR@ AR = @AR@ AS = @AS@ AUTOCONF = @AUTOCONF@ AUTOHEADER = @AUTOHEADER@ AUTOMAKE = @AUTOMAKE@ AVX_CFLAGS = @AVX_CFLAGS@ AWK = @AWK@ CC = @CC@ CCDEPMODE = @CCDEPMODE@ CFLAGS = @CFLAGS@ CHECK_PL_OPTS = @CHECK_PL_OPTS@ CPP = @CPP@ CPPFLAGS = @CPPFLAGS@ CYGPATH_W = @CYGPATH_W@ C_FFTW_R2R_KIND = @C_FFTW_R2R_KIND@ C_MPI_FINT = @C_MPI_FINT@ DEFS = @DEFS@ DEPDIR = @DEPDIR@ DLLTOOL = @DLLTOOL@ DSYMUTIL = @DSYMUTIL@ DUMPBIN = @DUMPBIN@ ECHO_C = @ECHO_C@ ECHO_N = @ECHO_N@ ECHO_T = @ECHO_T@ EGREP = @EGREP@ EXEEXT = @EXEEXT@ F77 = @F77@ FFLAGS = @FFLAGS@ FGREP = @FGREP@ FLIBS = @FLIBS@ GREP = @GREP@ INSTALL = @INSTALL@ INSTALL_DATA = @INSTALL_DATA@ INSTALL_PROGRAM = @INSTALL_PROGRAM@ INSTALL_SCRIPT = @INSTALL_SCRIPT@ INSTALL_STRIP_PROGRAM = @INSTALL_STRIP_PROGRAM@ LD = @LD@ LDFLAGS = @LDFLAGS@ LIBOBJS = @LIBOBJS@ LIBQUADMATH = @LIBQUADMATH@ LIBS = @LIBS@ LIBTOOL = @LIBTOOL@ LIPO = @LIPO@ LN_S = @LN_S@ LTLIBOBJS = @LTLIBOBJS@ MAINT = @MAINT@ MAKEINFO = @MAKEINFO@ MANIFEST_TOOL = @MANIFEST_TOOL@ MKDIR_P = @MKDIR_P@ MPICC = @MPICC@ MPILIBS = @MPILIBS@ MPIRUN = @MPIRUN@ NEON_CFLAGS = @NEON_CFLAGS@ NM = @NM@ NMEDIT = @NMEDIT@ OBJDUMP = @OBJDUMP@ OBJEXT = @OBJEXT@ OCAMLBUILD = @OCAMLBUILD@ OPENMP_CFLAGS = @OPENMP_CFLAGS@ OTOOL = @OTOOL@ OTOOL64 = @OTOOL64@ PACKAGE = @PACKAGE@ PACKAGE_BUGREPORT = @PACKAGE_BUGREPORT@ PACKAGE_NAME = @PACKAGE_NAME@ PACKAGE_STRING = @PACKAGE_STRING@ PACKAGE_TARNAME = @PACKAGE_TARNAME@ PACKAGE_URL = @PACKAGE_URL@ PACKAGE_VERSION = @PACKAGE_VERSION@ PATH_SEPARATOR = @PATH_SEPARATOR@ POW_LIB = @POW_LIB@ PRECISION = @PRECISION@ PREC_SUFFIX = @PREC_SUFFIX@ PTHREAD_CC = @PTHREAD_CC@ PTHREAD_CFLAGS = @PTHREAD_CFLAGS@ PTHREAD_LIBS = @PTHREAD_LIBS@ RANLIB = @RANLIB@ SED = @SED@ SET_MAKE = @SET_MAKE@ SHARED_VERSION_INFO = @SHARED_VERSION_INFO@ SHELL = @SHELL@ SSE2_CFLAGS = @SSE2_CFLAGS@ STACK_ALIGN_CFLAGS = @STACK_ALIGN_CFLAGS@ STRIP = @STRIP@ THREADLIBS = @THREADLIBS@ VERSION = @VERSION@ abs_builddir = @abs_builddir@ abs_srcdir = @abs_srcdir@ abs_top_builddir = @abs_top_builddir@ abs_top_srcdir = @abs_top_srcdir@ ac_ct_AR = @ac_ct_AR@ ac_ct_CC = @ac_ct_CC@ ac_ct_DUMPBIN = @ac_ct_DUMPBIN@ ac_ct_F77 = @ac_ct_F77@ acx_pthread_config = @acx_pthread_config@ am__include = @am__include@ am__leading_dot = @am__leading_dot@ am__quote = @am__quote@ am__tar = @am__tar@ am__untar = @am__untar@ bindir = @bindir@ build = @build@ build_alias = @build_alias@ build_cpu = @build_cpu@ build_os = @build_os@ build_vendor = @build_vendor@ builddir = @builddir@ datadir = @datadir@ datarootdir = @datarootdir@ docdir = @docdir@ dvidir = @dvidir@ exec_prefix = @exec_prefix@ host = @host@ host_alias = @host_alias@ host_cpu = @host_cpu@ host_os = @host_os@ host_vendor = @host_vendor@ htmldir = @htmldir@ includedir = @includedir@ infodir = @infodir@ install_sh = @install_sh@ libdir = @libdir@ libexecdir = @libexecdir@ localedir = @localedir@ localstatedir = @localstatedir@ mandir = @mandir@ mkdir_p = @mkdir_p@ oldincludedir = @oldincludedir@ pdfdir = @pdfdir@ prefix = @prefix@ program_transform_name = @program_transform_name@ psdir = @psdir@ sbindir = @sbindir@ sharedstatedir = @sharedstatedir@ srcdir = @srcdir@ sysconfdir = @sysconfdir@ target_alias = @target_alias@ top_build_prefix = @top_build_prefix@ top_builddir = @top_builddir@ top_srcdir = @top_srcdir@ SUBDIRS = FAQ info_TEXINFOS = fftw3.texi fftw3_TEXINFOS = acknowledgements.texi cindex.texi fftw3.texi findex.texi install.texi intro.texi legacy-fortran.texi license.texi modern-fortran.texi mpi.texi other.texi reference.texi threads.texi tutorial.texi upgrading.texi version.texi rfftwnd.pdf rfftwnd.eps DVIPS = dvips -Pwww EQN_IMAGES = equation-dft.png equation-dht.png equation-idft.png \ equation-redft00.png equation-redft01.png equation-redft10.png \ equation-redft11.png equation-rodft00.png equation-rodft01.png \ equation-rodft10.png equation-rodft11.png EXTRA_DIST = f77_wisdom.f fftw3.pdf html rfftwnd.fig rfftwnd.eps \ rfftwnd.pdf rfftwnd-for-html.png $(EQN_IMAGES) all: all-recursive .SUFFIXES: .SUFFIXES: .dvi .html .info .pdf .ps .texi $(srcdir)/Makefile.in: @MAINTAINER_MODE_TRUE@ $(srcdir)/Makefile.am $(am__configure_deps) @for dep in $?; do \ case '$(am__configure_deps)' in \ *$$dep*) \ ( cd $(top_builddir) && $(MAKE) $(AM_MAKEFLAGS) am--refresh ) \ && { if test -f $@; then exit 0; else break; fi; }; \ exit 1;; \ esac; \ done; \ echo ' cd $(top_srcdir) && $(AUTOMAKE) --gnu doc/Makefile'; \ $(am__cd) $(top_srcdir) && \ $(AUTOMAKE) --gnu doc/Makefile .PRECIOUS: Makefile Makefile: $(srcdir)/Makefile.in $(top_builddir)/config.status @case '$?' in \ *config.status*) \ cd $(top_builddir) && $(MAKE) $(AM_MAKEFLAGS) am--refresh;; \ *) \ echo ' cd $(top_builddir) && $(SHELL) ./config.status $(subdir)/$@ $(am__depfiles_maybe)'; \ cd $(top_builddir) && $(SHELL) ./config.status $(subdir)/$@ $(am__depfiles_maybe);; \ esac; $(top_builddir)/config.status: $(top_srcdir)/configure $(CONFIG_STATUS_DEPENDENCIES) cd $(top_builddir) && $(MAKE) $(AM_MAKEFLAGS) am--refresh $(top_srcdir)/configure: @MAINTAINER_MODE_TRUE@ $(am__configure_deps) cd $(top_builddir) && $(MAKE) $(AM_MAKEFLAGS) am--refresh $(ACLOCAL_M4): @MAINTAINER_MODE_TRUE@ $(am__aclocal_m4_deps) cd $(top_builddir) && $(MAKE) $(AM_MAKEFLAGS) am--refresh $(am__aclocal_m4_deps): mostlyclean-libtool: -rm -f *.lo clean-libtool: -rm -rf .libs _libs .texi.info: restore=: && backupdir="$(am__leading_dot)am$$$$" && \ am__cwd=`pwd` && $(am__cd) $(srcdir) && \ rm -rf $$backupdir && mkdir $$backupdir && \ if ($(MAKEINFO) --version) >/dev/null 2>&1; then \ for f in $@ $@-[0-9] $@-[0-9][0-9] $(@:.info=).i[0-9] $(@:.info=).i[0-9][0-9]; do \ if test -f $$f; then mv $$f $$backupdir; restore=mv; else :; fi; \ done; \ else :; fi && \ cd "$$am__cwd"; \ if $(MAKEINFO) $(AM_MAKEINFOFLAGS) $(MAKEINFOFLAGS) -I $(srcdir) \ -o $@ $<; \ then \ rc=0; \ $(am__cd) $(srcdir); \ else \ rc=$$?; \ $(am__cd) $(srcdir) && \ $$restore $$backupdir/* `echo "./$@" | sed 's|[^/]*$$||'`; \ fi; \ rm -rf $$backupdir; exit $$rc .texi.dvi: TEXINPUTS="$(am__TEXINFO_TEX_DIR)$(PATH_SEPARATOR)$$TEXINPUTS" \ MAKEINFO='$(MAKEINFO) $(AM_MAKEINFOFLAGS) $(MAKEINFOFLAGS) -I $(srcdir)' \ $(TEXI2DVI) $< .texi.pdf: TEXINPUTS="$(am__TEXINFO_TEX_DIR)$(PATH_SEPARATOR)$$TEXINPUTS" \ MAKEINFO='$(MAKEINFO) $(AM_MAKEINFOFLAGS) $(MAKEINFOFLAGS) -I $(srcdir)' \ $(TEXI2PDF) $< .texi.html: rm -rf $(@:.html=.htp) if $(MAKEINFOHTML) $(AM_MAKEINFOHTMLFLAGS) $(MAKEINFOFLAGS) -I $(srcdir) \ -o $(@:.html=.htp) $<; \ then \ rm -rf $@; \ if test ! -d $(@:.html=.htp) && test -d $(@:.html=); then \ mv $(@:.html=) $@; else mv $(@:.html=.htp) $@; fi; \ else \ if test ! -d $(@:.html=.htp) && test -d $(@:.html=); then \ rm -rf $(@:.html=); else rm -Rf $(@:.html=.htp) $@; fi; \ exit 1; \ fi $(srcdir)/fftw3.info: fftw3.texi $(srcdir)/version.texi $(fftw3_TEXINFOS) fftw3.dvi: fftw3.texi $(srcdir)/version.texi $(fftw3_TEXINFOS) fftw3.pdf: fftw3.texi $(srcdir)/version.texi $(fftw3_TEXINFOS) fftw3.html: fftw3.texi $(srcdir)/version.texi $(fftw3_TEXINFOS) $(srcdir)/version.texi: @MAINTAINER_MODE_TRUE@ $(srcdir)/stamp-vti $(srcdir)/stamp-vti: fftw3.texi $(top_srcdir)/configure @(dir=.; test -f ./fftw3.texi || dir=$(srcdir); \ set `$(SHELL) $(srcdir)/mdate-sh $$dir/fftw3.texi`; \ echo "@set UPDATED $$1 $$2 $$3"; \ echo "@set UPDATED-MONTH $$2 $$3"; \ echo "@set EDITION $(VERSION)"; \ echo "@set VERSION $(VERSION)") > vti.tmp @cmp -s vti.tmp $(srcdir)/version.texi \ || (echo "Updating $(srcdir)/version.texi"; \ cp vti.tmp $(srcdir)/version.texi) -@rm -f vti.tmp @cp $(srcdir)/version.texi $@ mostlyclean-vti: -rm -f vti.tmp maintainer-clean-vti: @MAINTAINER_MODE_TRUE@ -rm -f $(srcdir)/stamp-vti $(srcdir)/version.texi .dvi.ps: TEXINPUTS="$(am__TEXINFO_TEX_DIR)$(PATH_SEPARATOR)$$TEXINPUTS" \ $(DVIPS) -o $@ $< uninstall-dvi-am: @$(NORMAL_UNINSTALL) @list='$(DVIS)'; test -n "$(dvidir)" || list=; \ for p in $$list; do \ $(am__strip_dir) \ echo " rm -f '$(DESTDIR)$(dvidir)/$$f'"; \ rm -f "$(DESTDIR)$(dvidir)/$$f"; \ done uninstall-html-am: @$(NORMAL_UNINSTALL) @list='$(HTMLS)'; test -n "$(htmldir)" || list=; \ for p in $$list; do \ $(am__strip_dir) \ echo " rm -rf '$(DESTDIR)$(htmldir)/$$f'"; \ rm -rf "$(DESTDIR)$(htmldir)/$$f"; \ done uninstall-info-am: @$(PRE_UNINSTALL) @if test -d '$(DESTDIR)$(infodir)' && $(am__can_run_installinfo); then \ list='$(INFO_DEPS)'; \ for file in $$list; do \ relfile=`echo "$$file" | sed 's|^.*/||'`; \ echo " install-info --info-dir='$(DESTDIR)$(infodir)' --remove '$(DESTDIR)$(infodir)/$$relfile'"; \ if install-info --info-dir="$(DESTDIR)$(infodir)" --remove "$(DESTDIR)$(infodir)/$$relfile"; \ then :; else test ! -f "$(DESTDIR)$(infodir)/$$relfile" || exit 1; fi; \ done; \ else :; fi @$(NORMAL_UNINSTALL) @list='$(INFO_DEPS)'; \ for file in $$list; do \ relfile=`echo "$$file" | sed 's|^.*/||'`; \ relfile_i=`echo "$$relfile" | sed 's|\.info$$||;s|$$|.i|'`; \ (if test -d "$(DESTDIR)$(infodir)" && cd "$(DESTDIR)$(infodir)"; then \ echo " cd '$(DESTDIR)$(infodir)' && rm -f $$relfile $$relfile-[0-9] $$relfile-[0-9][0-9] $$relfile_i[0-9] $$relfile_i[0-9][0-9]"; \ rm -f $$relfile $$relfile-[0-9] $$relfile-[0-9][0-9] $$relfile_i[0-9] $$relfile_i[0-9][0-9]; \ else :; fi); \ done uninstall-pdf-am: @$(NORMAL_UNINSTALL) @list='$(PDFS)'; test -n "$(pdfdir)" || list=; \ for p in $$list; do \ $(am__strip_dir) \ echo " rm -f '$(DESTDIR)$(pdfdir)/$$f'"; \ rm -f "$(DESTDIR)$(pdfdir)/$$f"; \ done uninstall-ps-am: @$(NORMAL_UNINSTALL) @list='$(PSS)'; test -n "$(psdir)" || list=; \ for p in $$list; do \ $(am__strip_dir) \ echo " rm -f '$(DESTDIR)$(psdir)/$$f'"; \ rm -f "$(DESTDIR)$(psdir)/$$f"; \ done dist-info: $(INFO_DEPS) @srcdirstrip=`echo "$(srcdir)" | sed 's|.|.|g'`; \ list='$(INFO_DEPS)'; \ for base in $$list; do \ case $$base in \ $(srcdir)/*) base=`echo "$$base" | sed "s|^$$srcdirstrip/||"`;; \ esac; \ if test -f $$base; then d=.; else d=$(srcdir); fi; \ base_i=`echo "$$base" | sed 's|\.info$$||;s|$$|.i|'`; \ for file in $$d/$$base $$d/$$base-[0-9] $$d/$$base-[0-9][0-9] $$d/$$base_i[0-9] $$d/$$base_i[0-9][0-9]; do \ if test -f $$file; then \ relfile=`expr "$$file" : "$$d/\(.*\)"`; \ test -f "$(distdir)/$$relfile" || \ cp -p $$file "$(distdir)/$$relfile"; \ else :; fi; \ done; \ done mostlyclean-aminfo: -rm -rf fftw3.aux fftw3.cp fftw3.cps fftw3.ct fftw3.fc fftw3.ff fftw3.fn \ fftw3.fns fftw3.fp fftw3.ky fftw3.kys fftw3.log fftw3.pg \ fftw3.tmp fftw3.toc fftw3.tp fftw3.vr clean-aminfo: -test -z "fftw3.dvi fftw3.pdf fftw3.ps fftw3.html" \ || rm -rf fftw3.dvi fftw3.pdf fftw3.ps fftw3.html maintainer-clean-aminfo: @list='$(INFO_DEPS)'; for i in $$list; do \ i_i=`echo "$$i" | sed 's|\.info$$||;s|$$|.i|'`; \ echo " rm -f $$i $$i-[0-9] $$i-[0-9][0-9] $$i_i[0-9] $$i_i[0-9][0-9]"; \ rm -f $$i $$i-[0-9] $$i-[0-9][0-9] $$i_i[0-9] $$i_i[0-9][0-9]; \ done # This directory's subdirectories are mostly independent; you can cd # into them and run `make' without going through this Makefile. # To change the values of `make' variables: instead of editing Makefiles, # (1) if the variable is set in `config.status', edit `config.status' # (which will cause the Makefiles to be regenerated when you run `make'); # (2) otherwise, pass the desired values on the `make' command line. $(RECURSIVE_TARGETS): @fail= failcom='exit 1'; \ for f in x $$MAKEFLAGS; do \ case $$f in \ *=* | --[!k]*);; \ *k*) failcom='fail=yes';; \ esac; \ done; \ dot_seen=no; \ target=`echo $@ | sed s/-recursive//`; \ list='$(SUBDIRS)'; for subdir in $$list; do \ echo "Making $$target in $$subdir"; \ if test "$$subdir" = "."; then \ dot_seen=yes; \ local_target="$$target-am"; \ else \ local_target="$$target"; \ fi; \ ($(am__cd) $$subdir && $(MAKE) $(AM_MAKEFLAGS) $$local_target) \ || eval $$failcom; \ done; \ if test "$$dot_seen" = "no"; then \ $(MAKE) $(AM_MAKEFLAGS) "$$target-am" || exit 1; \ fi; test -z "$$fail" $(RECURSIVE_CLEAN_TARGETS): @fail= failcom='exit 1'; \ for f in x $$MAKEFLAGS; do \ case $$f in \ *=* | --[!k]*);; \ *k*) failcom='fail=yes';; \ esac; \ done; \ dot_seen=no; \ case "$@" in \ distclean-* | maintainer-clean-*) list='$(DIST_SUBDIRS)' ;; \ *) list='$(SUBDIRS)' ;; \ esac; \ rev=''; for subdir in $$list; do \ if test "$$subdir" = "."; then :; else \ rev="$$subdir $$rev"; \ fi; \ done; \ rev="$$rev ."; \ target=`echo $@ | sed s/-recursive//`; \ for subdir in $$rev; do \ echo "Making $$target in $$subdir"; \ if test "$$subdir" = "."; then \ local_target="$$target-am"; \ else \ local_target="$$target"; \ fi; \ ($(am__cd) $$subdir && $(MAKE) $(AM_MAKEFLAGS) $$local_target) \ || eval $$failcom; \ done && test -z "$$fail" tags-recursive: list='$(SUBDIRS)'; for subdir in $$list; do \ test "$$subdir" = . || ($(am__cd) $$subdir && $(MAKE) $(AM_MAKEFLAGS) tags); \ done ctags-recursive: list='$(SUBDIRS)'; for subdir in $$list; do \ test "$$subdir" = . || ($(am__cd) $$subdir && $(MAKE) $(AM_MAKEFLAGS) ctags); \ done ID: $(HEADERS) $(SOURCES) $(LISP) $(TAGS_FILES) list='$(SOURCES) $(HEADERS) $(LISP) $(TAGS_FILES)'; \ unique=`for i in $$list; do \ if test -f "$$i"; then echo $$i; else echo $(srcdir)/$$i; fi; \ done | \ $(AWK) '{ files[$$0] = 1; nonempty = 1; } \ END { if (nonempty) { for (i in files) print i; }; }'`; \ mkid -fID $$unique tags: TAGS TAGS: tags-recursive $(HEADERS) $(SOURCES) $(TAGS_DEPENDENCIES) \ $(TAGS_FILES) $(LISP) set x; \ here=`pwd`; \ if ($(ETAGS) --etags-include --version) >/dev/null 2>&1; then \ include_option=--etags-include; \ empty_fix=.; \ else \ include_option=--include; \ empty_fix=; \ fi; \ list='$(SUBDIRS)'; for subdir in $$list; do \ if test "$$subdir" = .; then :; else \ test ! -f $$subdir/TAGS || \ set "$$@" "$$include_option=$$here/$$subdir/TAGS"; \ fi; \ done; \ list='$(SOURCES) $(HEADERS) $(LISP) $(TAGS_FILES)'; \ unique=`for i in $$list; do \ if test -f "$$i"; then echo $$i; else echo $(srcdir)/$$i; fi; \ done | \ $(AWK) '{ files[$$0] = 1; nonempty = 1; } \ END { if (nonempty) { for (i in files) print i; }; }'`; \ shift; \ if test -z "$(ETAGS_ARGS)$$*$$unique"; then :; else \ test -n "$$unique" || unique=$$empty_fix; \ if test $$# -gt 0; then \ $(ETAGS) $(ETAGSFLAGS) $(AM_ETAGSFLAGS) $(ETAGS_ARGS) \ "$$@" $$unique; \ else \ $(ETAGS) $(ETAGSFLAGS) $(AM_ETAGSFLAGS) $(ETAGS_ARGS) \ $$unique; \ fi; \ fi ctags: CTAGS CTAGS: ctags-recursive $(HEADERS) $(SOURCES) $(TAGS_DEPENDENCIES) \ $(TAGS_FILES) $(LISP) list='$(SOURCES) $(HEADERS) $(LISP) $(TAGS_FILES)'; \ unique=`for i in $$list; do \ if test -f "$$i"; then echo $$i; else echo $(srcdir)/$$i; fi; \ done | \ $(AWK) '{ files[$$0] = 1; nonempty = 1; } \ END { if (nonempty) { for (i in files) print i; }; }'`; \ test -z "$(CTAGS_ARGS)$$unique" \ || $(CTAGS) $(CTAGSFLAGS) $(AM_CTAGSFLAGS) $(CTAGS_ARGS) \ $$unique GTAGS: here=`$(am__cd) $(top_builddir) && pwd` \ && $(am__cd) $(top_srcdir) \ && gtags -i $(GTAGS_ARGS) "$$here" distclean-tags: -rm -f TAGS ID GTAGS GRTAGS GSYMS GPATH tags distdir: $(DISTFILES) @srcdirstrip=`echo "$(srcdir)" | sed 's/[].[^$$\\*]/\\\\&/g'`; \ topsrcdirstrip=`echo "$(top_srcdir)" | sed 's/[].[^$$\\*]/\\\\&/g'`; \ list='$(DISTFILES)'; \ dist_files=`for file in $$list; do echo $$file; done | \ sed -e "s|^$$srcdirstrip/||;t" \ -e "s|^$$topsrcdirstrip/|$(top_builddir)/|;t"`; \ case $$dist_files in \ */*) $(MKDIR_P) `echo "$$dist_files" | \ sed '/\//!d;s|^|$(distdir)/|;s,/[^/]*$$,,' | \ sort -u` ;; \ esac; \ for file in $$dist_files; do \ if test -f $$file || test -d $$file; then d=.; else d=$(srcdir); fi; \ if test -d $$d/$$file; then \ dir=`echo "/$$file" | sed -e 's,/[^/]*$$,,'`; \ if test -d "$(distdir)/$$file"; then \ find "$(distdir)/$$file" -type d ! -perm -700 -exec chmod u+rwx {} \;; \ fi; \ if test -d $(srcdir)/$$file && test $$d != $(srcdir); then \ cp -fpR $(srcdir)/$$file "$(distdir)$$dir" || exit 1; \ find "$(distdir)/$$file" -type d ! -perm -700 -exec chmod u+rwx {} \;; \ fi; \ cp -fpR $$d/$$file "$(distdir)$$dir" || exit 1; \ else \ test -f "$(distdir)/$$file" \ || cp -p $$d/$$file "$(distdir)/$$file" \ || exit 1; \ fi; \ done @list='$(DIST_SUBDIRS)'; for subdir in $$list; do \ if test "$$subdir" = .; then :; else \ $(am__make_dryrun) \ || test -d "$(distdir)/$$subdir" \ || $(MKDIR_P) "$(distdir)/$$subdir" \ || exit 1; \ dir1=$$subdir; dir2="$(distdir)/$$subdir"; \ $(am__relativize); \ new_distdir=$$reldir; \ dir1=$$subdir; dir2="$(top_distdir)"; \ $(am__relativize); \ new_top_distdir=$$reldir; \ echo " (cd $$subdir && $(MAKE) $(AM_MAKEFLAGS) top_distdir="$$new_top_distdir" distdir="$$new_distdir" \\"; \ echo " am__remove_distdir=: am__skip_length_check=: am__skip_mode_fix=: distdir)"; \ ($(am__cd) $$subdir && \ $(MAKE) $(AM_MAKEFLAGS) \ top_distdir="$$new_top_distdir" \ distdir="$$new_distdir" \ am__remove_distdir=: \ am__skip_length_check=: \ am__skip_mode_fix=: \ distdir) \ || exit 1; \ fi; \ done $(MAKE) $(AM_MAKEFLAGS) \ top_distdir="$(top_distdir)" distdir="$(distdir)" \ dist-info check-am: all-am check: check-recursive all-am: Makefile $(INFO_DEPS) installdirs: installdirs-recursive installdirs-am: for dir in "$(DESTDIR)$(infodir)"; do \ test -z "$$dir" || $(MKDIR_P) "$$dir"; \ done install: install-recursive install-exec: install-exec-recursive install-data: install-data-recursive uninstall: uninstall-recursive install-am: all-am @$(MAKE) $(AM_MAKEFLAGS) install-exec-am install-data-am installcheck: installcheck-recursive install-strip: if test -z '$(STRIP)'; then \ $(MAKE) $(AM_MAKEFLAGS) INSTALL_PROGRAM="$(INSTALL_STRIP_PROGRAM)" \ install_sh_PROGRAM="$(INSTALL_STRIP_PROGRAM)" INSTALL_STRIP_FLAG=-s \ install; \ else \ $(MAKE) $(AM_MAKEFLAGS) INSTALL_PROGRAM="$(INSTALL_STRIP_PROGRAM)" \ install_sh_PROGRAM="$(INSTALL_STRIP_PROGRAM)" INSTALL_STRIP_FLAG=-s \ "INSTALL_PROGRAM_ENV=STRIPPROG='$(STRIP)'" install; \ fi mostlyclean-generic: clean-generic: distclean-generic: -test -z "$(CONFIG_CLEAN_FILES)" || rm -f $(CONFIG_CLEAN_FILES) -test . = "$(srcdir)" || test -z "$(CONFIG_CLEAN_VPATH_FILES)" || rm -f $(CONFIG_CLEAN_VPATH_FILES) maintainer-clean-generic: @echo "This command is intended for maintainers to use" @echo "it deletes files that may require special tools to rebuild." clean: clean-recursive clean-am: clean-aminfo clean-generic clean-libtool mostlyclean-am distclean: distclean-recursive -rm -f Makefile distclean-am: clean-am distclean-generic distclean-tags dvi: dvi-recursive dvi-am: $(DVIS) html-am: $(HTMLS) info: info-recursive info-am: $(INFO_DEPS) install-data-am: install-info-am install-dvi: install-dvi-recursive install-dvi-am: $(DVIS) @$(NORMAL_INSTALL) @list='$(DVIS)'; test -n "$(dvidir)" || list=; \ if test -n "$$list"; then \ echo " $(MKDIR_P) '$(DESTDIR)$(dvidir)'"; \ $(MKDIR_P) "$(DESTDIR)$(dvidir)" || exit 1; \ fi; \ for p in $$list; do \ if test -f "$$p"; then d=; else d="$(srcdir)/"; fi; \ echo "$$d$$p"; \ done | $(am__base_list) | \ while read files; do \ echo " $(INSTALL_DATA) $$files '$(DESTDIR)$(dvidir)'"; \ $(INSTALL_DATA) $$files "$(DESTDIR)$(dvidir)" || exit $$?; \ done install-exec-am: install-html: install-html-recursive install-html-am: $(HTMLS) @$(NORMAL_INSTALL) @list='$(HTMLS)'; list2=; test -n "$(htmldir)" || list=; \ if test -n "$$list"; then \ echo " $(MKDIR_P) '$(DESTDIR)$(htmldir)'"; \ $(MKDIR_P) "$(DESTDIR)$(htmldir)" || exit 1; \ fi; \ for p in $$list; do \ if test -f "$$p" || test -d "$$p"; then d=; else d="$(srcdir)/"; fi; \ $(am__strip_dir) \ d2=$$d$$p; \ if test -d "$$d2"; then \ echo " $(MKDIR_P) '$(DESTDIR)$(htmldir)/$$f'"; \ $(MKDIR_P) "$(DESTDIR)$(htmldir)/$$f" || exit 1; \ echo " $(INSTALL_DATA) '$$d2'/* '$(DESTDIR)$(htmldir)/$$f'"; \ $(INSTALL_DATA) "$$d2"/* "$(DESTDIR)$(htmldir)/$$f" || exit $$?; \ else \ list2="$$list2 $$d2"; \ fi; \ done; \ test -z "$$list2" || { echo "$$list2" | $(am__base_list) | \ while read files; do \ echo " $(INSTALL_DATA) $$files '$(DESTDIR)$(htmldir)'"; \ $(INSTALL_DATA) $$files "$(DESTDIR)$(htmldir)" || exit $$?; \ done; } install-info: install-info-recursive install-info-am: $(INFO_DEPS) @$(NORMAL_INSTALL) @srcdirstrip=`echo "$(srcdir)" | sed 's|.|.|g'`; \ list='$(INFO_DEPS)'; test -n "$(infodir)" || list=; \ if test -n "$$list"; then \ echo " $(MKDIR_P) '$(DESTDIR)$(infodir)'"; \ $(MKDIR_P) "$(DESTDIR)$(infodir)" || exit 1; \ fi; \ for file in $$list; do \ case $$file in \ $(srcdir)/*) file=`echo "$$file" | sed "s|^$$srcdirstrip/||"`;; \ esac; \ if test -f $$file; then d=.; else d=$(srcdir); fi; \ file_i=`echo "$$file" | sed 's|\.info$$||;s|$$|.i|'`; \ for ifile in $$d/$$file $$d/$$file-[0-9] $$d/$$file-[0-9][0-9] \ $$d/$$file_i[0-9] $$d/$$file_i[0-9][0-9] ; do \ if test -f $$ifile; then \ echo "$$ifile"; \ else : ; fi; \ done; \ done | $(am__base_list) | \ while read files; do \ echo " $(INSTALL_DATA) $$files '$(DESTDIR)$(infodir)'"; \ $(INSTALL_DATA) $$files "$(DESTDIR)$(infodir)" || exit $$?; done @$(POST_INSTALL) @if $(am__can_run_installinfo); then \ list='$(INFO_DEPS)'; test -n "$(infodir)" || list=; \ for file in $$list; do \ relfile=`echo "$$file" | sed 's|^.*/||'`; \ echo " install-info --info-dir='$(DESTDIR)$(infodir)' '$(DESTDIR)$(infodir)/$$relfile'";\ install-info --info-dir="$(DESTDIR)$(infodir)" "$(DESTDIR)$(infodir)/$$relfile" || :;\ done; \ else : ; fi install-man: install-pdf: install-pdf-recursive install-pdf-am: $(PDFS) @$(NORMAL_INSTALL) @list='$(PDFS)'; test -n "$(pdfdir)" || list=; \ if test -n "$$list"; then \ echo " $(MKDIR_P) '$(DESTDIR)$(pdfdir)'"; \ $(MKDIR_P) "$(DESTDIR)$(pdfdir)" || exit 1; \ fi; \ for p in $$list; do \ if test -f "$$p"; then d=; else d="$(srcdir)/"; fi; \ echo "$$d$$p"; \ done | $(am__base_list) | \ while read files; do \ echo " $(INSTALL_DATA) $$files '$(DESTDIR)$(pdfdir)'"; \ $(INSTALL_DATA) $$files "$(DESTDIR)$(pdfdir)" || exit $$?; done install-ps: install-ps-recursive install-ps-am: $(PSS) @$(NORMAL_INSTALL) @list='$(PSS)'; test -n "$(psdir)" || list=; \ if test -n "$$list"; then \ echo " $(MKDIR_P) '$(DESTDIR)$(psdir)'"; \ $(MKDIR_P) "$(DESTDIR)$(psdir)" || exit 1; \ fi; \ for p in $$list; do \ if test -f "$$p"; then d=; else d="$(srcdir)/"; fi; \ echo "$$d$$p"; \ done | $(am__base_list) | \ while read files; do \ echo " $(INSTALL_DATA) $$files '$(DESTDIR)$(psdir)'"; \ $(INSTALL_DATA) $$files "$(DESTDIR)$(psdir)" || exit $$?; done installcheck-am: maintainer-clean: maintainer-clean-recursive -rm -f Makefile maintainer-clean-am: distclean-am maintainer-clean-aminfo \ maintainer-clean-generic maintainer-clean-local \ maintainer-clean-vti mostlyclean: mostlyclean-recursive mostlyclean-am: mostlyclean-aminfo mostlyclean-generic \ mostlyclean-libtool mostlyclean-vti pdf: pdf-recursive pdf-am: $(PDFS) ps: ps-recursive ps-am: $(PSS) uninstall-am: uninstall-dvi-am uninstall-html-am uninstall-info-am \ uninstall-pdf-am uninstall-ps-am .MAKE: $(RECURSIVE_CLEAN_TARGETS) $(RECURSIVE_TARGETS) ctags-recursive \ install-am install-strip tags-recursive .PHONY: $(RECURSIVE_CLEAN_TARGETS) $(RECURSIVE_TARGETS) CTAGS GTAGS \ all all-am check check-am clean clean-aminfo clean-generic \ clean-libtool ctags ctags-recursive dist-info distclean \ distclean-generic distclean-libtool distclean-tags distdir dvi \ dvi-am html html-am info info-am install install-am \ install-data install-data-am install-dvi install-dvi-am \ install-exec install-exec-am install-html install-html-am \ install-info install-info-am install-man install-pdf \ install-pdf-am install-ps install-ps-am install-strip \ installcheck installcheck-am installdirs installdirs-am \ maintainer-clean maintainer-clean-aminfo \ maintainer-clean-generic maintainer-clean-local \ maintainer-clean-vti mostlyclean mostlyclean-aminfo \ mostlyclean-generic mostlyclean-libtool mostlyclean-vti pdf \ pdf-am ps ps-am tags tags-recursive uninstall uninstall-am \ uninstall-dvi-am uninstall-html-am uninstall-info-am \ uninstall-pdf-am uninstall-ps-am html: $(fftw3_TEXINFOS) $(EQN_IMAGES) rfftwnd-for-html.png $(MAKEINFO) $(AM_MAKEINFOFLAGS) $(MAKEINFOFLAGS) -I $(srcdir) \ --html --number-sections -o html fftw3.texi for i in $(EQN_IMAGES); do cp -f ${srcdir}/$$i html; done cp -f rfftwnd-for-html.png html maintainer-clean-local: rm -rf html # generate the figure for the manual and distribute the binaries, so that # people don't need to have fig2dev installed. @MAINTAINER_MODE_TRUE@rfftwnd.eps: rfftwnd.fig @MAINTAINER_MODE_TRUE@ fig2dev -L eps -m .7 ${srcdir}/rfftwnd.fig rfftwnd.eps @MAINTAINER_MODE_TRUE@rfftwnd-for-html.png: rfftwnd.fig @MAINTAINER_MODE_TRUE@ fig2dev -L png -m 1 ${srcdir}/rfftwnd.fig rfftwnd-for-html.png @MAINTAINER_MODE_TRUE@rfftwnd.pdf: rfftwnd.fig @MAINTAINER_MODE_TRUE@ fig2dev -L pdf -m .7 ${srcdir}/rfftwnd.fig rfftwnd.pdf # Tell versions [3.59,3.63) of GNU make to not export all variables. # Otherwise a system limit (for SysV at least) may be exceeded. .NOEXPORT: fftw-3.3.3/doc/stamp-vti0000644000175000017500000000014512054411104012013 00000000000000@set UPDATED 25 November 2012 @set UPDATED-MONTH November 2012 @set EDITION 3.3.3 @set VERSION 3.3.3 fftw-3.3.3/doc/mdate-sh0000755000175000017500000001371711777117217011630 00000000000000#!/bin/sh # Get modification time of a file or directory and pretty-print it. scriptversion=2010-08-21.06; # UTC # Copyright (C) 1995, 1996, 1997, 2003, 2004, 2005, 2007, 2009, 2010 # Free Software Foundation, Inc. # written by Ulrich Drepper , June 1995 # # This program is free software; you can redistribute it and/or modify # it under the terms of the GNU General Public License as published by # the Free Software Foundation; either version 2, or (at your option) # any later version. # # This program is distributed in the hope that it will be useful, # but WITHOUT ANY WARRANTY; without even the implied warranty of # MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the # GNU General Public License for more details. # # You should have received a copy of the GNU General Public License # along with this program. If not, see . # As a special exception to the GNU General Public License, if you # distribute this file as part of a program that contains a # configuration script generated by Autoconf, you may include it under # the same distribution terms that you use for the rest of that program. # This file is maintained in Automake, please report # bugs to or send patches to # . if test -n "${ZSH_VERSION+set}" && (emulate sh) >/dev/null 2>&1; then emulate sh NULLCMD=: # Pre-4.2 versions of Zsh do word splitting on ${1+"$@"}, which # is contrary to our usage. Disable this feature. alias -g '${1+"$@"}'='"$@"' setopt NO_GLOB_SUBST fi case $1 in '') echo "$0: No file. Try \`$0 --help' for more information." 1>&2 exit 1; ;; -h | --h*) cat <<\EOF Usage: mdate-sh [--help] [--version] FILE Pretty-print the modification day of FILE, in the format: 1 January 1970 Report bugs to . EOF exit $? ;; -v | --v*) echo "mdate-sh $scriptversion" exit $? ;; esac error () { echo "$0: $1" >&2 exit 1 } # Prevent date giving response in another language. LANG=C export LANG LC_ALL=C export LC_ALL LC_TIME=C export LC_TIME # GNU ls changes its time format in response to the TIME_STYLE # variable. Since we cannot assume `unset' works, revert this # variable to its documented default. if test "${TIME_STYLE+set}" = set; then TIME_STYLE=posix-long-iso export TIME_STYLE fi save_arg1=$1 # Find out how to get the extended ls output of a file or directory. if ls -L /dev/null 1>/dev/null 2>&1; then ls_command='ls -L -l -d' else ls_command='ls -l -d' fi # Avoid user/group names that might have spaces, when possible. if ls -n /dev/null 1>/dev/null 2>&1; then ls_command="$ls_command -n" fi # A `ls -l' line looks as follows on OS/2. # drwxrwx--- 0 Aug 11 2001 foo # This differs from Unix, which adds ownership information. # drwxrwx--- 2 root root 4096 Aug 11 2001 foo # # To find the date, we split the line on spaces and iterate on words # until we find a month. This cannot work with files whose owner is a # user named `Jan', or `Feb', etc. However, it's unlikely that `/' # will be owned by a user whose name is a month. So we first look at # the extended ls output of the root directory to decide how many # words should be skipped to get the date. # On HPUX /bin/sh, "set" interprets "-rw-r--r--" as options, so the "x" below. set x`$ls_command /` # Find which argument is the month. month= command= until test $month do test $# -gt 0 || error "failed parsing \`$ls_command /' output" shift # Add another shift to the command. command="$command shift;" case $1 in Jan) month=January; nummonth=1;; Feb) month=February; nummonth=2;; Mar) month=March; nummonth=3;; Apr) month=April; nummonth=4;; May) month=May; nummonth=5;; Jun) month=June; nummonth=6;; Jul) month=July; nummonth=7;; Aug) month=August; nummonth=8;; Sep) month=September; nummonth=9;; Oct) month=October; nummonth=10;; Nov) month=November; nummonth=11;; Dec) month=December; nummonth=12;; esac done test -n "$month" || error "failed parsing \`$ls_command /' output" # Get the extended ls output of the file or directory. set dummy x`eval "$ls_command \"\\\$save_arg1\""` # Remove all preceding arguments eval $command # Because of the dummy argument above, month is in $2. # # On a POSIX system, we should have # # $# = 5 # $1 = file size # $2 = month # $3 = day # $4 = year or time # $5 = filename # # On Darwin 7.7.0 and 7.6.0, we have # # $# = 4 # $1 = day # $2 = month # $3 = year or time # $4 = filename # Get the month. case $2 in Jan) month=January; nummonth=1;; Feb) month=February; nummonth=2;; Mar) month=March; nummonth=3;; Apr) month=April; nummonth=4;; May) month=May; nummonth=5;; Jun) month=June; nummonth=6;; Jul) month=July; nummonth=7;; Aug) month=August; nummonth=8;; Sep) month=September; nummonth=9;; Oct) month=October; nummonth=10;; Nov) month=November; nummonth=11;; Dec) month=December; nummonth=12;; esac case $3 in ???*) day=$1;; *) day=$3; shift;; esac # Here we have to deal with the problem that the ls output gives either # the time of day or the year. case $3 in *:*) set `date`; eval year=\$$# case $2 in Jan) nummonthtod=1;; Feb) nummonthtod=2;; Mar) nummonthtod=3;; Apr) nummonthtod=4;; May) nummonthtod=5;; Jun) nummonthtod=6;; Jul) nummonthtod=7;; Aug) nummonthtod=8;; Sep) nummonthtod=9;; Oct) nummonthtod=10;; Nov) nummonthtod=11;; Dec) nummonthtod=12;; esac # For the first six month of the year the time notation can also # be used for files modified in the last year. if (expr $nummonth \> $nummonthtod) > /dev/null; then year=`expr $year - 1` fi;; *) year=$3;; esac # The result. echo $day $month $year # Local Variables: # mode: shell-script # sh-indentation: 2 # eval: (add-hook 'write-file-hooks 'time-stamp) # time-stamp-start: "scriptversion=" # time-stamp-format: "%:y-%02m-%02d.%02H" # time-stamp-time-zone: "UTC" # time-stamp-end: "; # UTC" # End: fftw-3.3.3/doc/texinfo.tex0000644000175000017500000116130511777117217012374 00000000000000% texinfo.tex -- TeX macros to handle Texinfo files. % % Load plain if necessary, i.e., if running under initex. \expandafter\ifx\csname fmtname\endcsname\relax\input plain\fi % \def\texinfoversion{2012-03-11.15} % % Copyright 1985, 1986, 1988, 1990, 1991, 1992, 1993, 1994, 1995, % 1996, 1997, 1998, 1999, 2000, 2001, 2002, 2003, 2004, 2005, 2006, % 2007, 2008, 2009, 2010, 2011, 2012 Free Software Foundation, Inc. % % This texinfo.tex file is free software: you can redistribute it and/or % modify it under the terms of the GNU General Public License as % published by the Free Software Foundation, either version 3 of the % License, or (at your option) any later version. % % This texinfo.tex file is distributed in the hope that it will be % useful, but WITHOUT ANY WARRANTY; without even the implied warranty % of MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the GNU % General Public License for more details. % % You should have received a copy of the GNU General Public License % along with this program. If not, see . % % As a special exception, when this file is read by TeX when processing % a Texinfo source document, you may use the result without % restriction. (This has been our intent since Texinfo was invented.) % % Please try the latest version of texinfo.tex before submitting bug % reports; you can get the latest version from: % http://www.gnu.org/software/texinfo/ (the Texinfo home page), or % ftp://tug.org/tex/texinfo.tex % (and all CTAN mirrors, see http://www.ctan.org). % The texinfo.tex in any given distribution could well be out % of date, so if that's what you're using, please check. % % Send bug reports to bug-texinfo@gnu.org. Please include including a % complete document in each bug report with which we can reproduce the % problem. Patches are, of course, greatly appreciated. % % To process a Texinfo manual with TeX, it's most reliable to use the % texi2dvi shell script that comes with the distribution. For a simple % manual foo.texi, however, you can get away with this: % tex foo.texi % texindex foo.?? % tex foo.texi % tex foo.texi % dvips foo.dvi -o # or whatever; this makes foo.ps. % The extra TeX runs get the cross-reference information correct. % Sometimes one run after texindex suffices, and sometimes you need more % than two; texi2dvi does it as many times as necessary. % % It is possible to adapt texinfo.tex for other languages, to some % extent. You can get the existing language-specific files from the % full Texinfo distribution. % % The GNU Texinfo home page is http://www.gnu.org/software/texinfo. \message{Loading texinfo [version \texinfoversion]:} % If in a .fmt file, print the version number % and turn on active characters that we couldn't do earlier because % they might have appeared in the input file name. \everyjob{\message{[Texinfo version \texinfoversion]}% \catcode`+=\active \catcode`\_=\active} \chardef\other=12 % We never want plain's \outer definition of \+ in Texinfo. % For @tex, we can use \tabalign. \let\+ = \relax % Save some plain tex macros whose names we will redefine. \let\ptexb=\b \let\ptexbullet=\bullet \let\ptexc=\c \let\ptexcomma=\, \let\ptexdot=\. \let\ptexdots=\dots \let\ptexend=\end \let\ptexequiv=\equiv \let\ptexexclam=\! \let\ptexfootnote=\footnote \let\ptexgtr=> \let\ptexhat=^ \let\ptexi=\i \let\ptexindent=\indent \let\ptexinsert=\insert \let\ptexlbrace=\{ \let\ptexless=< \let\ptexnewwrite\newwrite \let\ptexnoindent=\noindent \let\ptexplus=+ \let\ptexraggedright=\raggedright \let\ptexrbrace=\} \let\ptexslash=\/ \let\ptexstar=\* \let\ptext=\t \let\ptextop=\top {\catcode`\'=\active \global\let\ptexquoteright'}% active in plain's math mode % If this character appears in an error message or help string, it % starts a new line in the output. \newlinechar = `^^J % Use TeX 3.0's \inputlineno to get the line number, for better error % messages, but if we're using an old version of TeX, don't do anything. % \ifx\inputlineno\thisisundefined \let\linenumber = \empty % Pre-3.0. \else \def\linenumber{l.\the\inputlineno:\space} \fi % Set up fixed words for English if not already set. \ifx\putwordAppendix\undefined \gdef\putwordAppendix{Appendix}\fi \ifx\putwordChapter\undefined \gdef\putwordChapter{Chapter}\fi \ifx\putworderror\undefined \gdef\putworderror{error}\fi \ifx\putwordfile\undefined \gdef\putwordfile{file}\fi \ifx\putwordin\undefined \gdef\putwordin{in}\fi \ifx\putwordIndexIsEmpty\undefined \gdef\putwordIndexIsEmpty{(Index is empty)}\fi \ifx\putwordIndexNonexistent\undefined \gdef\putwordIndexNonexistent{(Index is nonexistent)}\fi \ifx\putwordInfo\undefined \gdef\putwordInfo{Info}\fi \ifx\putwordInstanceVariableof\undefined \gdef\putwordInstanceVariableof{Instance Variable of}\fi \ifx\putwordMethodon\undefined \gdef\putwordMethodon{Method on}\fi \ifx\putwordNoTitle\undefined \gdef\putwordNoTitle{No Title}\fi \ifx\putwordof\undefined \gdef\putwordof{of}\fi \ifx\putwordon\undefined \gdef\putwordon{on}\fi \ifx\putwordpage\undefined \gdef\putwordpage{page}\fi \ifx\putwordsection\undefined \gdef\putwordsection{section}\fi \ifx\putwordSection\undefined \gdef\putwordSection{Section}\fi \ifx\putwordsee\undefined \gdef\putwordsee{see}\fi \ifx\putwordSee\undefined \gdef\putwordSee{See}\fi \ifx\putwordShortTOC\undefined \gdef\putwordShortTOC{Short Contents}\fi \ifx\putwordTOC\undefined \gdef\putwordTOC{Table of Contents}\fi % \ifx\putwordMJan\undefined \gdef\putwordMJan{January}\fi \ifx\putwordMFeb\undefined \gdef\putwordMFeb{February}\fi \ifx\putwordMMar\undefined \gdef\putwordMMar{March}\fi \ifx\putwordMApr\undefined \gdef\putwordMApr{April}\fi \ifx\putwordMMay\undefined \gdef\putwordMMay{May}\fi \ifx\putwordMJun\undefined \gdef\putwordMJun{June}\fi \ifx\putwordMJul\undefined \gdef\putwordMJul{July}\fi \ifx\putwordMAug\undefined \gdef\putwordMAug{August}\fi \ifx\putwordMSep\undefined \gdef\putwordMSep{September}\fi \ifx\putwordMOct\undefined \gdef\putwordMOct{October}\fi \ifx\putwordMNov\undefined \gdef\putwordMNov{November}\fi \ifx\putwordMDec\undefined \gdef\putwordMDec{December}\fi % \ifx\putwordDefmac\undefined \gdef\putwordDefmac{Macro}\fi \ifx\putwordDefspec\undefined \gdef\putwordDefspec{Special Form}\fi \ifx\putwordDefvar\undefined \gdef\putwordDefvar{Variable}\fi \ifx\putwordDefopt\undefined \gdef\putwordDefopt{User Option}\fi \ifx\putwordDeffunc\undefined \gdef\putwordDeffunc{Function}\fi % Since the category of space is not known, we have to be careful. \chardef\spacecat = 10 \def\spaceisspace{\catcode`\ =\spacecat} % sometimes characters are active, so we need control sequences. \chardef\ampChar = `\& \chardef\colonChar = `\: \chardef\commaChar = `\, \chardef\dashChar = `\- \chardef\dotChar = `\. \chardef\exclamChar= `\! \chardef\hashChar = `\# \chardef\lquoteChar= `\` \chardef\questChar = `\? \chardef\rquoteChar= `\' \chardef\semiChar = `\; \chardef\slashChar = `\/ \chardef\underChar = `\_ % Ignore a token. % \def\gobble#1{} % The following is used inside several \edef's. \def\makecsname#1{\expandafter\noexpand\csname#1\endcsname} % Hyphenation fixes. \hyphenation{ Flor-i-da Ghost-script Ghost-view Mac-OS Post-Script ap-pen-dix bit-map bit-maps data-base data-bases eshell fall-ing half-way long-est man-u-script man-u-scripts mini-buf-fer mini-buf-fers over-view par-a-digm par-a-digms rath-er rec-tan-gu-lar ro-bot-ics se-vere-ly set-up spa-ces spell-ing spell-ings stand-alone strong-est time-stamp time-stamps which-ever white-space wide-spread wrap-around } % Margin to add to right of even pages, to left of odd pages. \newdimen\bindingoffset \newdimen\normaloffset \newdimen\pagewidth \newdimen\pageheight % For a final copy, take out the rectangles % that mark overfull boxes (in case you have decided % that the text looks ok even though it passes the margin). % \def\finalout{\overfullrule=0pt } % Sometimes it is convenient to have everything in the transcript file % and nothing on the terminal. We don't just call \tracingall here, % since that produces some useless output on the terminal. We also make % some effort to order the tracing commands to reduce output in the log % file; cf. trace.sty in LaTeX. % \def\gloggingall{\begingroup \globaldefs = 1 \loggingall \endgroup}% \def\loggingall{% \tracingstats2 \tracingpages1 \tracinglostchars2 % 2 gives us more in etex \tracingparagraphs1 \tracingoutput1 \tracingmacros2 \tracingrestores1 \showboxbreadth\maxdimen \showboxdepth\maxdimen \ifx\eTeXversion\thisisundefined\else % etex gives us more logging \tracingscantokens1 \tracingifs1 \tracinggroups1 \tracingnesting2 \tracingassigns1 \fi \tracingcommands3 % 3 gives us more in etex \errorcontextlines16 }% % @errormsg{MSG}. Do the index-like expansions on MSG, but if things % aren't perfect, it's not the end of the world, being an error message, % after all. % \def\errormsg{\begingroup \indexnofonts \doerrormsg} \def\doerrormsg#1{\errmessage{#1}} % add check for \lastpenalty to plain's definitions. If the last thing % we did was a \nobreak, we don't want to insert more space. % \def\smallbreak{\ifnum\lastpenalty<10000\par\ifdim\lastskip<\smallskipamount \removelastskip\penalty-50\smallskip\fi\fi} \def\medbreak{\ifnum\lastpenalty<10000\par\ifdim\lastskip<\medskipamount \removelastskip\penalty-100\medskip\fi\fi} \def\bigbreak{\ifnum\lastpenalty<10000\par\ifdim\lastskip<\bigskipamount \removelastskip\penalty-200\bigskip\fi\fi} % Do @cropmarks to get crop marks. % \newif\ifcropmarks \let\cropmarks = \cropmarkstrue % % Dimensions to add cropmarks at corners. % Added by P. A. MacKay, 12 Nov. 1986 % \newdimen\outerhsize \newdimen\outervsize % set by the paper size routines \newdimen\cornerlong \cornerlong=1pc \newdimen\cornerthick \cornerthick=.3pt \newdimen\topandbottommargin \topandbottommargin=.75in % Output a mark which sets \thischapter, \thissection and \thiscolor. % We dump everything together because we only have one kind of mark. % This works because we only use \botmark / \topmark, not \firstmark. % % A mark contains a subexpression of the \ifcase ... \fi construct. % \get*marks macros below extract the needed part using \ifcase. % % Another complication is to let the user choose whether \thischapter % (\thissection) refers to the chapter (section) in effect at the top % of a page, or that at the bottom of a page. The solution is % described on page 260 of The TeXbook. It involves outputting two % marks for the sectioning macros, one before the section break, and % one after. I won't pretend I can describe this better than DEK... \def\domark{% \toks0=\expandafter{\lastchapterdefs}% \toks2=\expandafter{\lastsectiondefs}% \toks4=\expandafter{\prevchapterdefs}% \toks6=\expandafter{\prevsectiondefs}% \toks8=\expandafter{\lastcolordefs}% \mark{% \the\toks0 \the\toks2 \noexpand\or \the\toks4 \the\toks6 \noexpand\else \the\toks8 }% } % \topmark doesn't work for the very first chapter (after the title % page or the contents), so we use \firstmark there -- this gets us % the mark with the chapter defs, unless the user sneaks in, e.g., % @setcolor (or @url, or @link, etc.) between @contents and the very % first @chapter. \def\gettopheadingmarks{% \ifcase0\topmark\fi \ifx\thischapter\empty \ifcase0\firstmark\fi \fi } \def\getbottomheadingmarks{\ifcase1\botmark\fi} \def\getcolormarks{\ifcase2\topmark\fi} % Avoid "undefined control sequence" errors. \def\lastchapterdefs{} \def\lastsectiondefs{} \def\prevchapterdefs{} \def\prevsectiondefs{} \def\lastcolordefs{} % Main output routine. \chardef\PAGE = 255 \output = {\onepageout{\pagecontents\PAGE}} \newbox\headlinebox \newbox\footlinebox % \onepageout takes a vbox as an argument. Note that \pagecontents % does insertions, but you have to call it yourself. \def\onepageout#1{% \ifcropmarks \hoffset=0pt \else \hoffset=\normaloffset \fi % \ifodd\pageno \advance\hoffset by \bindingoffset \else \advance\hoffset by -\bindingoffset\fi % % Do this outside of the \shipout so @code etc. will be expanded in % the headline as they should be, not taken literally (outputting ''code). \ifodd\pageno \getoddheadingmarks \else \getevenheadingmarks \fi \setbox\headlinebox = \vbox{\let\hsize=\pagewidth \makeheadline}% \ifodd\pageno \getoddfootingmarks \else \getevenfootingmarks \fi \setbox\footlinebox = \vbox{\let\hsize=\pagewidth \makefootline}% % {% % Have to do this stuff outside the \shipout because we want it to % take effect in \write's, yet the group defined by the \vbox ends % before the \shipout runs. % \indexdummies % don't expand commands in the output. \normalturnoffactive % \ in index entries must not stay \, e.g., if % the page break happens to be in the middle of an example. % We don't want .vr (or whatever) entries like this: % \entry{{\tt \indexbackslash }acronym}{32}{\code {\acronym}} % "\acronym" won't work when it's read back in; % it needs to be % {\code {{\tt \backslashcurfont }acronym} \shipout\vbox{% % Do this early so pdf references go to the beginning of the page. \ifpdfmakepagedest \pdfdest name{\the\pageno} xyz\fi % \ifcropmarks \vbox to \outervsize\bgroup \hsize = \outerhsize \vskip-\topandbottommargin \vtop to0pt{% \line{\ewtop\hfil\ewtop}% \nointerlineskip \line{% \vbox{\moveleft\cornerthick\nstop}% \hfill \vbox{\moveright\cornerthick\nstop}% }% \vss}% \vskip\topandbottommargin \line\bgroup \hfil % center the page within the outer (page) hsize. \ifodd\pageno\hskip\bindingoffset\fi \vbox\bgroup \fi % \unvbox\headlinebox \pagebody{#1}% \ifdim\ht\footlinebox > 0pt % Only leave this space if the footline is nonempty. % (We lessened \vsize for it in \oddfootingyyy.) % The \baselineskip=24pt in plain's \makefootline has no effect. \vskip 24pt \unvbox\footlinebox \fi % \ifcropmarks \egroup % end of \vbox\bgroup \hfil\egroup % end of (centering) \line\bgroup \vskip\topandbottommargin plus1fill minus1fill \boxmaxdepth = \cornerthick \vbox to0pt{\vss \line{% \vbox{\moveleft\cornerthick\nsbot}% \hfill \vbox{\moveright\cornerthick\nsbot}% }% \nointerlineskip \line{\ewbot\hfil\ewbot}% }% \egroup % \vbox from first cropmarks clause \fi }% end of \shipout\vbox }% end of group with \indexdummies \advancepageno \ifnum\outputpenalty>-20000 \else\dosupereject\fi } \newinsert\margin \dimen\margin=\maxdimen \def\pagebody#1{\vbox to\pageheight{\boxmaxdepth=\maxdepth #1}} {\catcode`\@ =11 \gdef\pagecontents#1{\ifvoid\topins\else\unvbox\topins\fi % marginal hacks, juha@viisa.uucp (Juha Takala) \ifvoid\margin\else % marginal info is present \rlap{\kern\hsize\vbox to\z@{\kern1pt\box\margin \vss}}\fi \dimen@=\dp#1\relax \unvbox#1\relax \ifvoid\footins\else\vskip\skip\footins\footnoterule \unvbox\footins\fi \ifr@ggedbottom \kern-\dimen@ \vfil \fi} } % Here are the rules for the cropmarks. Note that they are % offset so that the space between them is truly \outerhsize or \outervsize % (P. A. MacKay, 12 November, 1986) % \def\ewtop{\vrule height\cornerthick depth0pt width\cornerlong} \def\nstop{\vbox {\hrule height\cornerthick depth\cornerlong width\cornerthick}} \def\ewbot{\vrule height0pt depth\cornerthick width\cornerlong} \def\nsbot{\vbox {\hrule height\cornerlong depth\cornerthick width\cornerthick}} % Parse an argument, then pass it to #1. The argument is the rest of % the input line (except we remove a trailing comment). #1 should be a % macro which expects an ordinary undelimited TeX argument. % \def\parsearg{\parseargusing{}} \def\parseargusing#1#2{% \def\argtorun{#2}% \begingroup \obeylines \spaceisspace #1% \parseargline\empty% Insert the \empty token, see \finishparsearg below. } {\obeylines % \gdef\parseargline#1^^M{% \endgroup % End of the group started in \parsearg. \argremovecomment #1\comment\ArgTerm% }% } % First remove any @comment, then any @c comment. \def\argremovecomment#1\comment#2\ArgTerm{\argremovec #1\c\ArgTerm} \def\argremovec#1\c#2\ArgTerm{\argcheckspaces#1\^^M\ArgTerm} % Each occurrence of `\^^M' or `\^^M' is replaced by a single space. % % \argremovec might leave us with trailing space, e.g., % @end itemize @c foo % This space token undergoes the same procedure and is eventually removed % by \finishparsearg. % \def\argcheckspaces#1\^^M{\argcheckspacesX#1\^^M \^^M} \def\argcheckspacesX#1 \^^M{\argcheckspacesY#1\^^M} \def\argcheckspacesY#1\^^M#2\^^M#3\ArgTerm{% \def\temp{#3}% \ifx\temp\empty % Do not use \next, perhaps the caller of \parsearg uses it; reuse \temp: \let\temp\finishparsearg \else \let\temp\argcheckspaces \fi % Put the space token in: \temp#1 #3\ArgTerm } % If a _delimited_ argument is enclosed in braces, they get stripped; so % to get _exactly_ the rest of the line, we had to prevent such situation. % We prepended an \empty token at the very beginning and we expand it now, % just before passing the control to \argtorun. % (Similarly, we have to think about #3 of \argcheckspacesY above: it is % either the null string, or it ends with \^^M---thus there is no danger % that a pair of braces would be stripped. % % But first, we have to remove the trailing space token. % \def\finishparsearg#1 \ArgTerm{\expandafter\argtorun\expandafter{#1}} % \parseargdef\foo{...} % is roughly equivalent to % \def\foo{\parsearg\Xfoo} % \def\Xfoo#1{...} % % Actually, I use \csname\string\foo\endcsname, ie. \\foo, as it is my % favourite TeX trick. --kasal, 16nov03 \def\parseargdef#1{% \expandafter \doparseargdef \csname\string#1\endcsname #1% } \def\doparseargdef#1#2{% \def#2{\parsearg#1}% \def#1##1% } % Several utility definitions with active space: { \obeyspaces \gdef\obeyedspace{ } % Make each space character in the input produce a normal interword % space in the output. Don't allow a line break at this space, as this % is used only in environments like @example, where each line of input % should produce a line of output anyway. % \gdef\sepspaces{\obeyspaces\let =\tie} % If an index command is used in an @example environment, any spaces % therein should become regular spaces in the raw index file, not the % expansion of \tie (\leavevmode \penalty \@M \ ). \gdef\unsepspaces{\let =\space} } \def\flushcr{\ifx\par\lisppar \def\next##1{}\else \let\next=\relax \fi \next} % Define the framework for environments in texinfo.tex. It's used like this: % % \envdef\foo{...} % \def\Efoo{...} % % It's the responsibility of \envdef to insert \begingroup before the % actual body; @end closes the group after calling \Efoo. \envdef also % defines \thisenv, so the current environment is known; @end checks % whether the environment name matches. The \checkenv macro can also be % used to check whether the current environment is the one expected. % % Non-false conditionals (@iftex, @ifset) don't fit into this, so they % are not treated as environments; they don't open a group. (The % implementation of @end takes care not to call \endgroup in this % special case.) % At run-time, environments start with this: \def\startenvironment#1{\begingroup\def\thisenv{#1}} % initialize \let\thisenv\empty % ... but they get defined via ``\envdef\foo{...}'': \long\def\envdef#1#2{\def#1{\startenvironment#1#2}} \def\envparseargdef#1#2{\parseargdef#1{\startenvironment#1#2}} % Check whether we're in the right environment: \def\checkenv#1{% \def\temp{#1}% \ifx\thisenv\temp \else \badenverr \fi } % Environment mismatch, #1 expected: \def\badenverr{% \errhelp = \EMsimple \errmessage{This command can appear only \inenvironment\temp, not \inenvironment\thisenv}% } \def\inenvironment#1{% \ifx#1\empty outside of any environment% \else in environment \expandafter\string#1% \fi } % @end foo executes the definition of \Efoo. % But first, it executes a specialized version of \checkenv % \parseargdef\end{% \if 1\csname iscond.#1\endcsname \else % The general wording of \badenverr may not be ideal. \expandafter\checkenv\csname#1\endcsname \csname E#1\endcsname \endgroup \fi } \newhelp\EMsimple{Press RETURN to continue.} % Be sure we're in horizontal mode when doing a tie, since we make space % equivalent to this in @example-like environments. Otherwise, a space % at the beginning of a line will start with \penalty -- and % since \penalty is valid in vertical mode, we'd end up putting the % penalty on the vertical list instead of in the new paragraph. {\catcode`@ = 11 % Avoid using \@M directly, because that causes trouble % if the definition is written into an index file. \global\let\tiepenalty = \@M \gdef\tie{\leavevmode\penalty\tiepenalty\ } } % @: forces normal size whitespace following. \def\:{\spacefactor=1000 } % @* forces a line break. \def\*{\hfil\break\hbox{}\ignorespaces} % @/ allows a line break. \let\/=\allowbreak % @. is an end-of-sentence period. \def\.{.\spacefactor=\endofsentencespacefactor\space} % @! is an end-of-sentence bang. \def\!{!\spacefactor=\endofsentencespacefactor\space} % @? is an end-of-sentence query. \def\?{?\spacefactor=\endofsentencespacefactor\space} % @frenchspacing on|off says whether to put extra space after punctuation. % \def\onword{on} \def\offword{off} % \parseargdef\frenchspacing{% \def\temp{#1}% \ifx\temp\onword \plainfrenchspacing \else\ifx\temp\offword \plainnonfrenchspacing \else \errhelp = \EMsimple \errmessage{Unknown @frenchspacing option `\temp', must be on|off}% \fi\fi } % @w prevents a word break. Without the \leavevmode, @w at the % beginning of a paragraph, when TeX is still in vertical mode, would % produce a whole line of output instead of starting the paragraph. \def\w#1{\leavevmode\hbox{#1}} % @group ... @end group forces ... to be all on one page, by enclosing % it in a TeX vbox. We use \vtop instead of \vbox to construct the box % to keep its height that of a normal line. According to the rules for % \topskip (p.114 of the TeXbook), the glue inserted is % max (\topskip - \ht (first item), 0). If that height is large, % therefore, no glue is inserted, and the space between the headline and % the text is small, which looks bad. % % Another complication is that the group might be very large. This can % cause the glue on the previous page to be unduly stretched, because it % does not have much material. In this case, it's better to add an % explicit \vfill so that the extra space is at the bottom. The % threshold for doing this is if the group is more than \vfilllimit % percent of a page (\vfilllimit can be changed inside of @tex). % \newbox\groupbox \def\vfilllimit{0.7} % \envdef\group{% \ifnum\catcode`\^^M=\active \else \errhelp = \groupinvalidhelp \errmessage{@group invalid in context where filling is enabled}% \fi \startsavinginserts % \setbox\groupbox = \vtop\bgroup % Do @comment since we are called inside an environment such as % @example, where each end-of-line in the input causes an % end-of-line in the output. We don't want the end-of-line after % the `@group' to put extra space in the output. Since @group % should appear on a line by itself (according to the Texinfo % manual), we don't worry about eating any user text. \comment } % % The \vtop produces a box with normal height and large depth; thus, TeX puts % \baselineskip glue before it, and (when the next line of text is done) % \lineskip glue after it. Thus, space below is not quite equal to space % above. But it's pretty close. \def\Egroup{% % To get correct interline space between the last line of the group % and the first line afterwards, we have to propagate \prevdepth. \endgraf % Not \par, as it may have been set to \lisppar. \global\dimen1 = \prevdepth \egroup % End the \vtop. % \dimen0 is the vertical size of the group's box. \dimen0 = \ht\groupbox \advance\dimen0 by \dp\groupbox % \dimen2 is how much space is left on the page (more or less). \dimen2 = \pageheight \advance\dimen2 by -\pagetotal % if the group doesn't fit on the current page, and it's a big big % group, force a page break. \ifdim \dimen0 > \dimen2 \ifdim \pagetotal < \vfilllimit\pageheight \page \fi \fi \box\groupbox \prevdepth = \dimen1 \checkinserts } % % TeX puts in an \escapechar (i.e., `@') at the beginning of the help % message, so this ends up printing `@group can only ...'. % \newhelp\groupinvalidhelp{% group can only be used in environments such as @example,^^J% where each line of input produces a line of output.} % @need space-in-mils % forces a page break if there is not space-in-mils remaining. \newdimen\mil \mil=0.001in \parseargdef\need{% % Ensure vertical mode, so we don't make a big box in the middle of a % paragraph. \par % % If the @need value is less than one line space, it's useless. \dimen0 = #1\mil \dimen2 = \ht\strutbox \advance\dimen2 by \dp\strutbox \ifdim\dimen0 > \dimen2 % % Do a \strut just to make the height of this box be normal, so the % normal leading is inserted relative to the preceding line. % And a page break here is fine. \vtop to #1\mil{\strut\vfil}% % % TeX does not even consider page breaks if a penalty added to the % main vertical list is 10000 or more. But in order to see if the % empty box we just added fits on the page, we must make it consider % page breaks. On the other hand, we don't want to actually break the % page after the empty box. So we use a penalty of 9999. % % There is an extremely small chance that TeX will actually break the % page at this \penalty, if there are no other feasible breakpoints in % sight. (If the user is using lots of big @group commands, which % almost-but-not-quite fill up a page, TeX will have a hard time doing % good page breaking, for example.) However, I could not construct an % example where a page broke at this \penalty; if it happens in a real % document, then we can reconsider our strategy. \penalty9999 % % Back up by the size of the box, whether we did a page break or not. \kern -#1\mil % % Do not allow a page break right after this kern. \nobreak \fi } % @br forces paragraph break (and is undocumented). \let\br = \par % @page forces the start of a new page. % \def\page{\par\vfill\supereject} % @exdent text.... % outputs text on separate line in roman font, starting at standard page margin % This records the amount of indent in the innermost environment. % That's how much \exdent should take out. \newskip\exdentamount % This defn is used inside fill environments such as @defun. \parseargdef\exdent{\hfil\break\hbox{\kern -\exdentamount{\rm#1}}\hfil\break} % This defn is used inside nofill environments such as @example. \parseargdef\nofillexdent{{\advance \leftskip by -\exdentamount \leftline{\hskip\leftskip{\rm#1}}}} % @inmargin{WHICH}{TEXT} puts TEXT in the WHICH margin next to the current % paragraph. For more general purposes, use the \margin insertion % class. WHICH is `l' or `r'. Not documented, written for gawk manual. % \newskip\inmarginspacing \inmarginspacing=1cm \def\strutdepth{\dp\strutbox} % \def\doinmargin#1#2{\strut\vadjust{% \nobreak \kern-\strutdepth \vtop to \strutdepth{% \baselineskip=\strutdepth \vss % if you have multiple lines of stuff to put here, you'll need to % make the vbox yourself of the appropriate size. \ifx#1l% \llap{\ignorespaces #2\hskip\inmarginspacing}% \else \rlap{\hskip\hsize \hskip\inmarginspacing \ignorespaces #2}% \fi \null }% }} \def\inleftmargin{\doinmargin l} \def\inrightmargin{\doinmargin r} % % @inmargin{TEXT [, RIGHT-TEXT]} % (if RIGHT-TEXT is given, use TEXT for left page, RIGHT-TEXT for right; % else use TEXT for both). % \def\inmargin#1{\parseinmargin #1,,\finish} \def\parseinmargin#1,#2,#3\finish{% not perfect, but better than nothing. \setbox0 = \hbox{\ignorespaces #2}% \ifdim\wd0 > 0pt \def\lefttext{#1}% have both texts \def\righttext{#2}% \else \def\lefttext{#1}% have only one text \def\righttext{#1}% \fi % \ifodd\pageno \def\temp{\inrightmargin\righttext}% odd page -> outside is right margin \else \def\temp{\inleftmargin\lefttext}% \fi \temp } % @| inserts a changebar to the left of the current line. It should % surround any changed text. This approach does *not* work if the % change spans more than two lines of output. To handle that, we would % have adopt a much more difficult approach (putting marks into the main % vertical list for the beginning and end of each change). This command % is not documented, not supported, and doesn't work. % \def\|{% % \vadjust can only be used in horizontal mode. \leavevmode % % Append this vertical mode material after the current line in the output. \vadjust{% % We want to insert a rule with the height and depth of the current % leading; that is exactly what \strutbox is supposed to record. \vskip-\baselineskip % % \vadjust-items are inserted at the left edge of the type. So % the \llap here moves out into the left-hand margin. \llap{% % % For a thicker or thinner bar, change the `1pt'. \vrule height\baselineskip width1pt % % This is the space between the bar and the text. \hskip 12pt }% }% } % @include FILE -- \input text of FILE. % \def\include{\parseargusing\filenamecatcodes\includezzz} \def\includezzz#1{% \pushthisfilestack \def\thisfile{#1}% {% \makevalueexpandable % we want to expand any @value in FILE. \turnoffactive % and allow special characters in the expansion \indexnofonts % Allow `@@' and other weird things in file names. \wlog{texinfo.tex: doing @include of #1^^J}% \edef\temp{\noexpand\input #1 }% % % This trickery is to read FILE outside of a group, in case it makes % definitions, etc. \expandafter }\temp \popthisfilestack } \def\filenamecatcodes{% \catcode`\\=\other \catcode`~=\other \catcode`^=\other \catcode`_=\other \catcode`|=\other \catcode`<=\other \catcode`>=\other \catcode`+=\other \catcode`-=\other \catcode`\`=\other \catcode`\'=\other } \def\pushthisfilestack{% \expandafter\pushthisfilestackX\popthisfilestack\StackTerm } \def\pushthisfilestackX{% \expandafter\pushthisfilestackY\thisfile\StackTerm } \def\pushthisfilestackY #1\StackTerm #2\StackTerm {% \gdef\popthisfilestack{\gdef\thisfile{#1}\gdef\popthisfilestack{#2}}% } \def\popthisfilestack{\errthisfilestackempty} \def\errthisfilestackempty{\errmessage{Internal error: the stack of filenames is empty.}} % \def\thisfile{} % @center line % outputs that line, centered. % \parseargdef\center{% \ifhmode \let\centersub\centerH \else \let\centersub\centerV \fi \centersub{\hfil \ignorespaces#1\unskip \hfil}% \let\centersub\relax % don't let the definition persist, just in case } \def\centerH#1{{% \hfil\break \advance\hsize by -\leftskip \advance\hsize by -\rightskip \line{#1}% \break }} % \newcount\centerpenalty \def\centerV#1{% % The idea here is the same as in \startdefun, \cartouche, etc.: if % @center is the first thing after a section heading, we need to wipe % out the negative parskip inserted by \sectionheading, but still % prevent a page break here. \centerpenalty = \lastpenalty \ifnum\centerpenalty>10000 \vskip\parskip \fi \ifnum\centerpenalty>9999 \penalty\centerpenalty \fi \line{\kern\leftskip #1\kern\rightskip}% } % @sp n outputs n lines of vertical space % \parseargdef\sp{\vskip #1\baselineskip} % @comment ...line which is ignored... % @c is the same as @comment % @ignore ... @end ignore is another way to write a comment % \def\comment{\begingroup \catcode`\^^M=\other% \catcode`\@=\other \catcode`\{=\other \catcode`\}=\other% \commentxxx} {\catcode`\^^M=\other \gdef\commentxxx#1^^M{\endgroup}} % \let\c=\comment % @paragraphindent NCHARS % We'll use ems for NCHARS, close enough. % NCHARS can also be the word `asis' or `none'. % We cannot feasibly implement @paragraphindent asis, though. % \def\asisword{asis} % no translation, these are keywords \def\noneword{none} % \parseargdef\paragraphindent{% \def\temp{#1}% \ifx\temp\asisword \else \ifx\temp\noneword \defaultparindent = 0pt \else \defaultparindent = #1em \fi \fi \parindent = \defaultparindent } % @exampleindent NCHARS % We'll use ems for NCHARS like @paragraphindent. % It seems @exampleindent asis isn't necessary, but % I preserve it to make it similar to @paragraphindent. \parseargdef\exampleindent{% \def\temp{#1}% \ifx\temp\asisword \else \ifx\temp\noneword \lispnarrowing = 0pt \else \lispnarrowing = #1em \fi \fi } % @firstparagraphindent WORD % If WORD is `none', then suppress indentation of the first paragraph % after a section heading. If WORD is `insert', then do indent at such % paragraphs. % % The paragraph indentation is suppressed or not by calling % \suppressfirstparagraphindent, which the sectioning commands do. % We switch the definition of this back and forth according to WORD. % By default, we suppress indentation. % \def\suppressfirstparagraphindent{\dosuppressfirstparagraphindent} \def\insertword{insert} % \parseargdef\firstparagraphindent{% \def\temp{#1}% \ifx\temp\noneword \let\suppressfirstparagraphindent = \dosuppressfirstparagraphindent \else\ifx\temp\insertword \let\suppressfirstparagraphindent = \relax \else \errhelp = \EMsimple \errmessage{Unknown @firstparagraphindent option `\temp'}% \fi\fi } % Here is how we actually suppress indentation. Redefine \everypar to % \kern backwards by \parindent, and then reset itself to empty. % % We also make \indent itself not actually do anything until the next % paragraph. % \gdef\dosuppressfirstparagraphindent{% \gdef\indent{% \restorefirstparagraphindent \indent }% \gdef\noindent{% \restorefirstparagraphindent \noindent }% \global\everypar = {% \kern -\parindent \restorefirstparagraphindent }% } \gdef\restorefirstparagraphindent{% \global \let \indent = \ptexindent \global \let \noindent = \ptexnoindent \global \everypar = {}% } % @refill is a no-op. \let\refill=\relax % If working on a large document in chapters, it is convenient to % be able to disable indexing, cross-referencing, and contents, for test runs. % This is done with @novalidate (before @setfilename). % \newif\iflinks \linkstrue % by default we want the aux files. \let\novalidate = \linksfalse % @setfilename is done at the beginning of every texinfo file. % So open here the files we need to have open while reading the input. % This makes it possible to make a .fmt file for texinfo. \def\setfilename{% \fixbackslash % Turn off hack to swallow `\input texinfo'. \iflinks \tryauxfile % Open the new aux file. TeX will close it automatically at exit. \immediate\openout\auxfile=\jobname.aux \fi % \openindices needs to do some work in any case. \openindices \let\setfilename=\comment % Ignore extra @setfilename cmds. % % If texinfo.cnf is present on the system, read it. % Useful for site-wide @afourpaper, etc. \openin 1 texinfo.cnf \ifeof 1 \else \input texinfo.cnf \fi \closein 1 % \comment % Ignore the actual filename. } % Called from \setfilename. % \def\openindices{% \newindex{cp}% \newcodeindex{fn}% \newcodeindex{vr}% \newcodeindex{tp}% \newcodeindex{ky}% \newcodeindex{pg}% } % @bye. \outer\def\bye{\pagealignmacro\tracingstats=1\ptexend} \message{pdf,} % adobe `portable' document format \newcount\tempnum \newcount\lnkcount \newtoks\filename \newcount\filenamelength \newcount\pgn \newtoks\toksA \newtoks\toksB \newtoks\toksC \newtoks\toksD \newbox\boxA \newcount\countA \newif\ifpdf \newif\ifpdfmakepagedest % when pdftex is run in dvi mode, \pdfoutput is defined (so \pdfoutput=1 % can be set). So we test for \relax and 0 as well as being undefined. \ifx\pdfoutput\thisisundefined \else \ifx\pdfoutput\relax \else \ifcase\pdfoutput \else \pdftrue \fi \fi \fi % PDF uses PostScript string constants for the names of xref targets, % for display in the outlines, and in other places. Thus, we have to % double any backslashes. Otherwise, a name like "\node" will be % interpreted as a newline (\n), followed by o, d, e. Not good. % % See http://www.ntg.nl/pipermail/ntg-pdftex/2004-July/000654.html and % related messages. The final outcome is that it is up to the TeX user % to double the backslashes and otherwise make the string valid, so % that's what we do. pdftex 1.30.0 (ca.2005) introduced a primitive to % do this reliably, so we use it. % #1 is a control sequence in which to do the replacements, % which we \xdef. \def\txiescapepdf#1{% \ifx\pdfescapestring\relax % No primitive available; should we give a warning or log? % Many times it won't matter. \else % The expandable \pdfescapestring primitive escapes parentheses, % backslashes, and other special chars. \xdef#1{\pdfescapestring{#1}}% \fi } \newhelp\nopdfimagehelp{Texinfo supports .png, .jpg, .jpeg, and .pdf images with PDF output, and none of those formats could be found. (.eps cannot be supported due to the design of the PDF format; use regular TeX (DVI output) for that.)} \ifpdf % % Color manipulation macros based on pdfcolor.tex, % except using rgb instead of cmyk; the latter is said to render as a % very dark gray on-screen and a very dark halftone in print, instead % of actual black. \def\rgbDarkRed{0.50 0.09 0.12} \def\rgbBlack{0 0 0} % % k sets the color for filling (usual text, etc.); % K sets the color for stroking (thin rules, e.g., normal _'s). \def\pdfsetcolor#1{\pdfliteral{#1 rg #1 RG}} % % Set color, and create a mark which defines \thiscolor accordingly, % so that \makeheadline knows which color to restore. \def\setcolor#1{% \xdef\lastcolordefs{\gdef\noexpand\thiscolor{#1}}% \domark \pdfsetcolor{#1}% } % \def\maincolor{\rgbBlack} \pdfsetcolor{\maincolor} \edef\thiscolor{\maincolor} \def\lastcolordefs{} % \def\makefootline{% \baselineskip24pt \line{\pdfsetcolor{\maincolor}\the\footline}% } % \def\makeheadline{% \vbox to 0pt{% \vskip-22.5pt \line{% \vbox to8.5pt{}% % Extract \thiscolor definition from the marks. \getcolormarks % Typeset the headline with \maincolor, then restore the color. \pdfsetcolor{\maincolor}\the\headline\pdfsetcolor{\thiscolor}% }% \vss }% \nointerlineskip } % % \pdfcatalog{/PageMode /UseOutlines} % % #1 is image name, #2 width (might be empty/whitespace), #3 height (ditto). \def\dopdfimage#1#2#3{% \def\pdfimagewidth{#2}\setbox0 = \hbox{\ignorespaces #2}% \def\pdfimageheight{#3}\setbox2 = \hbox{\ignorespaces #3}% % % pdftex (and the PDF format) support .pdf, .png, .jpg (among % others). Let's try in that order, PDF first since if % someone has a scalable image, presumably better to use that than a % bitmap. \let\pdfimgext=\empty \begingroup \openin 1 #1.pdf \ifeof 1 \openin 1 #1.PDF \ifeof 1 \openin 1 #1.png \ifeof 1 \openin 1 #1.jpg \ifeof 1 \openin 1 #1.jpeg \ifeof 1 \openin 1 #1.JPG \ifeof 1 \errhelp = \nopdfimagehelp \errmessage{Could not find image file #1 for pdf}% \else \gdef\pdfimgext{JPG}% \fi \else \gdef\pdfimgext{jpeg}% \fi \else \gdef\pdfimgext{jpg}% \fi \else \gdef\pdfimgext{png}% \fi \else \gdef\pdfimgext{PDF}% \fi \else \gdef\pdfimgext{pdf}% \fi \closein 1 \endgroup % % without \immediate, ancient pdftex seg faults when the same image is % included twice. (Version 3.14159-pre-1.0-unofficial-20010704.) \ifnum\pdftexversion < 14 \immediate\pdfimage \else \immediate\pdfximage \fi \ifdim \wd0 >0pt width \pdfimagewidth \fi \ifdim \wd2 >0pt height \pdfimageheight \fi \ifnum\pdftexversion<13 #1.\pdfimgext \else {#1.\pdfimgext}% \fi \ifnum\pdftexversion < 14 \else \pdfrefximage \pdflastximage \fi} % \def\pdfmkdest#1{{% % We have to set dummies so commands such as @code, and characters % such as \, aren't expanded when present in a section title. \indexnofonts \turnoffactive \makevalueexpandable \def\pdfdestname{#1}% \txiescapepdf\pdfdestname \safewhatsit{\pdfdest name{\pdfdestname} xyz}% }} % % used to mark target names; must be expandable. \def\pdfmkpgn#1{#1} % % by default, use a color that is dark enough to print on paper as % nearly black, but still distinguishable for online viewing. \def\urlcolor{\rgbDarkRed} \def\linkcolor{\rgbDarkRed} \def\endlink{\setcolor{\maincolor}\pdfendlink} % % Adding outlines to PDF; macros for calculating structure of outlines % come from Petr Olsak \def\expnumber#1{\expandafter\ifx\csname#1\endcsname\relax 0% \else \csname#1\endcsname \fi} \def\advancenumber#1{\tempnum=\expnumber{#1}\relax \advance\tempnum by 1 \expandafter\xdef\csname#1\endcsname{\the\tempnum}} % % #1 is the section text, which is what will be displayed in the % outline by the pdf viewer. #2 is the pdf expression for the number % of subentries (or empty, for subsubsections). #3 is the node text, % which might be empty if this toc entry had no corresponding node. % #4 is the page number % \def\dopdfoutline#1#2#3#4{% % Generate a link to the node text if that exists; else, use the % page number. We could generate a destination for the section % text in the case where a section has no node, but it doesn't % seem worth the trouble, since most documents are normally structured. \edef\pdfoutlinedest{#3}% \ifx\pdfoutlinedest\empty \def\pdfoutlinedest{#4}% \else \txiescapepdf\pdfoutlinedest \fi % % Also escape PDF chars in the display string. \edef\pdfoutlinetext{#1}% \txiescapepdf\pdfoutlinetext % \pdfoutline goto name{\pdfmkpgn{\pdfoutlinedest}}#2{\pdfoutlinetext}% } % \def\pdfmakeoutlines{% \begingroup % Read toc silently, to get counts of subentries for \pdfoutline. \def\partentry##1##2##3##4{}% ignore parts in the outlines \def\numchapentry##1##2##3##4{% \def\thischapnum{##2}% \def\thissecnum{0}% \def\thissubsecnum{0}% }% \def\numsecentry##1##2##3##4{% \advancenumber{chap\thischapnum}% \def\thissecnum{##2}% \def\thissubsecnum{0}% }% \def\numsubsecentry##1##2##3##4{% \advancenumber{sec\thissecnum}% \def\thissubsecnum{##2}% }% \def\numsubsubsecentry##1##2##3##4{% \advancenumber{subsec\thissubsecnum}% }% \def\thischapnum{0}% \def\thissecnum{0}% \def\thissubsecnum{0}% % % use \def rather than \let here because we redefine \chapentry et % al. a second time, below. \def\appentry{\numchapentry}% \def\appsecentry{\numsecentry}% \def\appsubsecentry{\numsubsecentry}% \def\appsubsubsecentry{\numsubsubsecentry}% \def\unnchapentry{\numchapentry}% \def\unnsecentry{\numsecentry}% \def\unnsubsecentry{\numsubsecentry}% \def\unnsubsubsecentry{\numsubsubsecentry}% \readdatafile{toc}% % % Read toc second time, this time actually producing the outlines. % The `-' means take the \expnumber as the absolute number of % subentries, which we calculated on our first read of the .toc above. % % We use the node names as the destinations. \def\numchapentry##1##2##3##4{% \dopdfoutline{##1}{count-\expnumber{chap##2}}{##3}{##4}}% \def\numsecentry##1##2##3##4{% \dopdfoutline{##1}{count-\expnumber{sec##2}}{##3}{##4}}% \def\numsubsecentry##1##2##3##4{% \dopdfoutline{##1}{count-\expnumber{subsec##2}}{##3}{##4}}% \def\numsubsubsecentry##1##2##3##4{% count is always zero \dopdfoutline{##1}{}{##3}{##4}}% % % PDF outlines are displayed using system fonts, instead of % document fonts. Therefore we cannot use special characters, % since the encoding is unknown. For example, the eogonek from % Latin 2 (0xea) gets translated to a | character. Info from % Staszek Wawrykiewicz, 19 Jan 2004 04:09:24 +0100. % % TODO this right, we have to translate 8-bit characters to % their "best" equivalent, based on the @documentencoding. Too % much work for too little return. Just use the ASCII equivalents % we use for the index sort strings. % \indexnofonts \setupdatafile % We can have normal brace characters in the PDF outlines, unlike % Texinfo index files. So set that up. \def\{{\lbracecharliteral}% \def\}{\rbracecharliteral}% \catcode`\\=\active \otherbackslash \input \tocreadfilename \endgroup } {\catcode`[=1 \catcode`]=2 \catcode`{=\other \catcode`}=\other \gdef\lbracecharliteral[{]% \gdef\rbracecharliteral[}]% ] % \def\skipspaces#1{\def\PP{#1}\def\D{|}% \ifx\PP\D\let\nextsp\relax \else\let\nextsp\skipspaces \ifx\p\space\else\addtokens{\filename}{\PP}% \advance\filenamelength by 1 \fi \fi \nextsp} \def\getfilename#1{% \filenamelength=0 % If we don't expand the argument now, \skipspaces will get % snagged on things like "@value{foo}". \edef\temp{#1}% \expandafter\skipspaces\temp|\relax } \ifnum\pdftexversion < 14 \let \startlink \pdfannotlink \else \let \startlink \pdfstartlink \fi % make a live url in pdf output. \def\pdfurl#1{% \begingroup % it seems we really need yet another set of dummies; have not % tried to figure out what each command should do in the context % of @url. for now, just make @/ a no-op, that's the only one % people have actually reported a problem with. % \normalturnoffactive \def\@{@}% \let\/=\empty \makevalueexpandable % do we want to go so far as to use \indexnofonts instead of just % special-casing \var here? \def\var##1{##1}% % \leavevmode\setcolor{\urlcolor}% \startlink attr{/Border [0 0 0]}% user{/Subtype /Link /A << /S /URI /URI (#1) >>}% \endgroup} \def\pdfgettoks#1.{\setbox\boxA=\hbox{\toksA={#1.}\toksB={}\maketoks}} \def\addtokens#1#2{\edef\addtoks{\noexpand#1={\the#1#2}}\addtoks} \def\adn#1{\addtokens{\toksC}{#1}\global\countA=1\let\next=\maketoks} \def\poptoks#1#2|ENDTOKS|{\let\first=#1\toksD={#1}\toksA={#2}} \def\maketoks{% \expandafter\poptoks\the\toksA|ENDTOKS|\relax \ifx\first0\adn0 \else\ifx\first1\adn1 \else\ifx\first2\adn2 \else\ifx\first3\adn3 \else\ifx\first4\adn4 \else\ifx\first5\adn5 \else\ifx\first6\adn6 \else\ifx\first7\adn7 \else\ifx\first8\adn8 \else\ifx\first9\adn9 \else \ifnum0=\countA\else\makelink\fi \ifx\first.\let\next=\done\else \let\next=\maketoks \addtokens{\toksB}{\the\toksD} \ifx\first,\addtokens{\toksB}{\space}\fi \fi \fi\fi\fi\fi\fi\fi\fi\fi\fi\fi \next} \def\makelink{\addtokens{\toksB}% {\noexpand\pdflink{\the\toksC}}\toksC={}\global\countA=0} \def\pdflink#1{% \startlink attr{/Border [0 0 0]} goto name{\pdfmkpgn{#1}} \setcolor{\linkcolor}#1\endlink} \def\done{\edef\st{\global\noexpand\toksA={\the\toksB}}\st} \else % non-pdf mode \let\pdfmkdest = \gobble \let\pdfurl = \gobble \let\endlink = \relax \let\setcolor = \gobble \let\pdfsetcolor = \gobble \let\pdfmakeoutlines = \relax \fi % \ifx\pdfoutput \message{fonts,} % Change the current font style to #1, remembering it in \curfontstyle. % For now, we do not accumulate font styles: @b{@i{foo}} prints foo in % italics, not bold italics. % \def\setfontstyle#1{% \def\curfontstyle{#1}% not as a control sequence, because we are \edef'd. \csname ten#1\endcsname % change the current font } % Select #1 fonts with the current style. % \def\selectfonts#1{\csname #1fonts\endcsname \csname\curfontstyle\endcsname} \def\rm{\fam=0 \setfontstyle{rm}} \def\it{\fam=\itfam \setfontstyle{it}} \def\sl{\fam=\slfam \setfontstyle{sl}} \def\bf{\fam=\bffam \setfontstyle{bf}}\def\bfstylename{bf} \def\tt{\fam=\ttfam \setfontstyle{tt}} % Unfortunately, we have to override this for titles and the like, since % in those cases "rm" is bold. Sigh. \def\rmisbold{\rm\def\curfontstyle{bf}} % Texinfo sort of supports the sans serif font style, which plain TeX does not. % So we set up a \sf. \newfam\sffam \def\sf{\fam=\sffam \setfontstyle{sf}} \let\li = \sf % Sometimes we call it \li, not \sf. % We don't need math for this font style. \def\ttsl{\setfontstyle{ttsl}} % Default leading. \newdimen\textleading \textleading = 13.2pt % Set the baselineskip to #1, and the lineskip and strut size % correspondingly. There is no deep meaning behind these magic numbers % used as factors; they just match (closely enough) what Knuth defined. % \def\lineskipfactor{.08333} \def\strutheightpercent{.70833} \def\strutdepthpercent {.29167} % % can get a sort of poor man's double spacing by redefining this. \def\baselinefactor{1} % \def\setleading#1{% \dimen0 = #1\relax \normalbaselineskip = \baselinefactor\dimen0 \normallineskip = \lineskipfactor\normalbaselineskip \normalbaselines \setbox\strutbox =\hbox{% \vrule width0pt height\strutheightpercent\baselineskip depth \strutdepthpercent \baselineskip }% } % PDF CMaps. See also LaTeX's t1.cmap. % % do nothing with this by default. \expandafter\let\csname cmapOT1\endcsname\gobble \expandafter\let\csname cmapOT1IT\endcsname\gobble \expandafter\let\csname cmapOT1TT\endcsname\gobble % if we are producing pdf, and we have \pdffontattr, then define cmaps. % (\pdffontattr was introduced many years ago, but people still run % older pdftex's; it's easy to conditionalize, so we do.) \ifpdf \ifx\pdffontattr\thisisundefined \else \begingroup \catcode`\^^M=\active \def^^M{^^J}% Output line endings as the ^^J char. \catcode`\%=12 \immediate\pdfobj stream {%!PS-Adobe-3.0 Resource-CMap %%DocumentNeededResources: ProcSet (CIDInit) %%IncludeResource: ProcSet (CIDInit) %%BeginResource: CMap (TeX-OT1-0) %%Title: (TeX-OT1-0 TeX OT1 0) %%Version: 1.000 %%EndComments /CIDInit /ProcSet findresource begin 12 dict begin begincmap /CIDSystemInfo << /Registry (TeX) /Ordering (OT1) /Supplement 0 >> def /CMapName /TeX-OT1-0 def /CMapType 2 def 1 begincodespacerange <00> <7F> endcodespacerange 8 beginbfrange <00> <01> <0393> <09> <0A> <03A8> <23> <26> <0023> <28> <3B> <0028> <3F> <5B> <003F> <5D> <5E> <005D> <61> <7A> <0061> <7B> <7C> <2013> endbfrange 40 beginbfchar <02> <0398> <03> <039B> <04> <039E> <05> <03A0> <06> <03A3> <07> <03D2> <08> <03A6> <0B> <00660066> <0C> <00660069> <0D> <0066006C> <0E> <006600660069> <0F> <00660066006C> <10> <0131> <11> <0237> <12> <0060> <13> <00B4> <14> <02C7> <15> <02D8> <16> <00AF> <17> <02DA> <18> <00B8> <19> <00DF> <1A> <00E6> <1B> <0153> <1C> <00F8> <1D> <00C6> <1E> <0152> <1F> <00D8> <21> <0021> <22> <201D> <27> <2019> <3C> <00A1> <3D> <003D> <3E> <00BF> <5C> <201C> <5F> <02D9> <60> <2018> <7D> <02DD> <7E> <007E> <7F> <00A8> endbfchar endcmap CMapName currentdict /CMap defineresource pop end end %%EndResource %%EOF }\endgroup \expandafter\edef\csname cmapOT1\endcsname#1{% \pdffontattr#1{/ToUnicode \the\pdflastobj\space 0 R}% }% % % \cmapOT1IT \begingroup \catcode`\^^M=\active \def^^M{^^J}% Output line endings as the ^^J char. \catcode`\%=12 \immediate\pdfobj stream {%!PS-Adobe-3.0 Resource-CMap %%DocumentNeededResources: ProcSet (CIDInit) %%IncludeResource: ProcSet (CIDInit) %%BeginResource: CMap (TeX-OT1IT-0) %%Title: (TeX-OT1IT-0 TeX OT1IT 0) %%Version: 1.000 %%EndComments /CIDInit /ProcSet findresource begin 12 dict begin begincmap /CIDSystemInfo << /Registry (TeX) /Ordering (OT1IT) /Supplement 0 >> def /CMapName /TeX-OT1IT-0 def /CMapType 2 def 1 begincodespacerange <00> <7F> endcodespacerange 8 beginbfrange <00> <01> <0393> <09> <0A> <03A8> <25> <26> <0025> <28> <3B> <0028> <3F> <5B> <003F> <5D> <5E> <005D> <61> <7A> <0061> <7B> <7C> <2013> endbfrange 42 beginbfchar <02> <0398> <03> <039B> <04> <039E> <05> <03A0> <06> <03A3> <07> <03D2> <08> <03A6> <0B> <00660066> <0C> <00660069> <0D> <0066006C> <0E> <006600660069> <0F> <00660066006C> <10> <0131> <11> <0237> <12> <0060> <13> <00B4> <14> <02C7> <15> <02D8> <16> <00AF> <17> <02DA> <18> <00B8> <19> <00DF> <1A> <00E6> <1B> <0153> <1C> <00F8> <1D> <00C6> <1E> <0152> <1F> <00D8> <21> <0021> <22> <201D> <23> <0023> <24> <00A3> <27> <2019> <3C> <00A1> <3D> <003D> <3E> <00BF> <5C> <201C> <5F> <02D9> <60> <2018> <7D> <02DD> <7E> <007E> <7F> <00A8> endbfchar endcmap CMapName currentdict /CMap defineresource pop end end %%EndResource %%EOF }\endgroup \expandafter\edef\csname cmapOT1IT\endcsname#1{% \pdffontattr#1{/ToUnicode \the\pdflastobj\space 0 R}% }% % % \cmapOT1TT \begingroup \catcode`\^^M=\active \def^^M{^^J}% Output line endings as the ^^J char. \catcode`\%=12 \immediate\pdfobj stream {%!PS-Adobe-3.0 Resource-CMap %%DocumentNeededResources: ProcSet (CIDInit) %%IncludeResource: ProcSet (CIDInit) %%BeginResource: CMap (TeX-OT1TT-0) %%Title: (TeX-OT1TT-0 TeX OT1TT 0) %%Version: 1.000 %%EndComments /CIDInit /ProcSet findresource begin 12 dict begin begincmap /CIDSystemInfo << /Registry (TeX) /Ordering (OT1TT) /Supplement 0 >> def /CMapName /TeX-OT1TT-0 def /CMapType 2 def 1 begincodespacerange <00> <7F> endcodespacerange 5 beginbfrange <00> <01> <0393> <09> <0A> <03A8> <21> <26> <0021> <28> <5F> <0028> <61> <7E> <0061> endbfrange 32 beginbfchar <02> <0398> <03> <039B> <04> <039E> <05> <03A0> <06> <03A3> <07> <03D2> <08> <03A6> <0B> <2191> <0C> <2193> <0D> <0027> <0E> <00A1> <0F> <00BF> <10> <0131> <11> <0237> <12> <0060> <13> <00B4> <14> <02C7> <15> <02D8> <16> <00AF> <17> <02DA> <18> <00B8> <19> <00DF> <1A> <00E6> <1B> <0153> <1C> <00F8> <1D> <00C6> <1E> <0152> <1F> <00D8> <20> <2423> <27> <2019> <60> <2018> <7F> <00A8> endbfchar endcmap CMapName currentdict /CMap defineresource pop end end %%EndResource %%EOF }\endgroup \expandafter\edef\csname cmapOT1TT\endcsname#1{% \pdffontattr#1{/ToUnicode \the\pdflastobj\space 0 R}% }% \fi\fi % Set the font macro #1 to the font named #2, adding on the % specified font prefix (normally `cm'). % #3 is the font's design size, #4 is a scale factor, #5 is the CMap % encoding (currently only OT1, OT1IT and OT1TT are allowed, pass % empty to omit). \def\setfont#1#2#3#4#5{% \font#1=\fontprefix#2#3 scaled #4 \csname cmap#5\endcsname#1% } % This is what gets called when #5 of \setfont is empty. \let\cmap\gobble % emacs-page end of cmaps % Use cm as the default font prefix. % To specify the font prefix, you must define \fontprefix % before you read in texinfo.tex. \ifx\fontprefix\thisisundefined \def\fontprefix{cm} \fi % Support font families that don't use the same naming scheme as CM. \def\rmshape{r} \def\rmbshape{bx} %where the normal face is bold \def\bfshape{b} \def\bxshape{bx} \def\ttshape{tt} \def\ttbshape{tt} \def\ttslshape{sltt} \def\itshape{ti} \def\itbshape{bxti} \def\slshape{sl} \def\slbshape{bxsl} \def\sfshape{ss} \def\sfbshape{ss} \def\scshape{csc} \def\scbshape{csc} % Definitions for a main text size of 11pt. This is the default in % Texinfo. % \def\definetextfontsizexi{% % Text fonts (11.2pt, magstep1). \def\textnominalsize{11pt} \edef\mainmagstep{\magstephalf} \setfont\textrm\rmshape{10}{\mainmagstep}{OT1} \setfont\texttt\ttshape{10}{\mainmagstep}{OT1TT} \setfont\textbf\bfshape{10}{\mainmagstep}{OT1} \setfont\textit\itshape{10}{\mainmagstep}{OT1IT} \setfont\textsl\slshape{10}{\mainmagstep}{OT1} \setfont\textsf\sfshape{10}{\mainmagstep}{OT1} \setfont\textsc\scshape{10}{\mainmagstep}{OT1} \setfont\textttsl\ttslshape{10}{\mainmagstep}{OT1TT} \font\texti=cmmi10 scaled \mainmagstep \font\textsy=cmsy10 scaled \mainmagstep \def\textecsize{1095} % A few fonts for @defun names and args. \setfont\defbf\bfshape{10}{\magstep1}{OT1} \setfont\deftt\ttshape{10}{\magstep1}{OT1TT} \setfont\defttsl\ttslshape{10}{\magstep1}{OT1TT} \def\df{\let\tentt=\deftt \let\tenbf = \defbf \let\tenttsl=\defttsl \bf} % Fonts for indices, footnotes, small examples (9pt). \def\smallnominalsize{9pt} \setfont\smallrm\rmshape{9}{1000}{OT1} \setfont\smalltt\ttshape{9}{1000}{OT1TT} \setfont\smallbf\bfshape{10}{900}{OT1} \setfont\smallit\itshape{9}{1000}{OT1IT} \setfont\smallsl\slshape{9}{1000}{OT1} \setfont\smallsf\sfshape{9}{1000}{OT1} \setfont\smallsc\scshape{10}{900}{OT1} \setfont\smallttsl\ttslshape{10}{900}{OT1TT} \font\smalli=cmmi9 \font\smallsy=cmsy9 \def\smallecsize{0900} % Fonts for small examples (8pt). \def\smallernominalsize{8pt} \setfont\smallerrm\rmshape{8}{1000}{OT1} \setfont\smallertt\ttshape{8}{1000}{OT1TT} \setfont\smallerbf\bfshape{10}{800}{OT1} \setfont\smallerit\itshape{8}{1000}{OT1IT} \setfont\smallersl\slshape{8}{1000}{OT1} \setfont\smallersf\sfshape{8}{1000}{OT1} \setfont\smallersc\scshape{10}{800}{OT1} \setfont\smallerttsl\ttslshape{10}{800}{OT1TT} \font\smalleri=cmmi8 \font\smallersy=cmsy8 \def\smallerecsize{0800} % Fonts for title page (20.4pt): \def\titlenominalsize{20pt} \setfont\titlerm\rmbshape{12}{\magstep3}{OT1} \setfont\titleit\itbshape{10}{\magstep4}{OT1IT} \setfont\titlesl\slbshape{10}{\magstep4}{OT1} \setfont\titlett\ttbshape{12}{\magstep3}{OT1TT} \setfont\titlettsl\ttslshape{10}{\magstep4}{OT1TT} \setfont\titlesf\sfbshape{17}{\magstep1}{OT1} \let\titlebf=\titlerm \setfont\titlesc\scbshape{10}{\magstep4}{OT1} \font\titlei=cmmi12 scaled \magstep3 \font\titlesy=cmsy10 scaled \magstep4 \def\titleecsize{2074} % Chapter (and unnumbered) fonts (17.28pt). \def\chapnominalsize{17pt} \setfont\chaprm\rmbshape{12}{\magstep2}{OT1} \setfont\chapit\itbshape{10}{\magstep3}{OT1IT} \setfont\chapsl\slbshape{10}{\magstep3}{OT1} \setfont\chaptt\ttbshape{12}{\magstep2}{OT1TT} \setfont\chapttsl\ttslshape{10}{\magstep3}{OT1TT} \setfont\chapsf\sfbshape{17}{1000}{OT1} \let\chapbf=\chaprm \setfont\chapsc\scbshape{10}{\magstep3}{OT1} \font\chapi=cmmi12 scaled \magstep2 \font\chapsy=cmsy10 scaled \magstep3 \def\chapecsize{1728} % Section fonts (14.4pt). \def\secnominalsize{14pt} \setfont\secrm\rmbshape{12}{\magstep1}{OT1} \setfont\secit\itbshape{10}{\magstep2}{OT1IT} \setfont\secsl\slbshape{10}{\magstep2}{OT1} \setfont\sectt\ttbshape{12}{\magstep1}{OT1TT} \setfont\secttsl\ttslshape{10}{\magstep2}{OT1TT} \setfont\secsf\sfbshape{12}{\magstep1}{OT1} \let\secbf\secrm \setfont\secsc\scbshape{10}{\magstep2}{OT1} \font\seci=cmmi12 scaled \magstep1 \font\secsy=cmsy10 scaled \magstep2 \def\sececsize{1440} % Subsection fonts (13.15pt). \def\ssecnominalsize{13pt} \setfont\ssecrm\rmbshape{12}{\magstephalf}{OT1} \setfont\ssecit\itbshape{10}{1315}{OT1IT} \setfont\ssecsl\slbshape{10}{1315}{OT1} \setfont\ssectt\ttbshape{12}{\magstephalf}{OT1TT} \setfont\ssecttsl\ttslshape{10}{1315}{OT1TT} \setfont\ssecsf\sfbshape{12}{\magstephalf}{OT1} \let\ssecbf\ssecrm \setfont\ssecsc\scbshape{10}{1315}{OT1} \font\sseci=cmmi12 scaled \magstephalf \font\ssecsy=cmsy10 scaled 1315 \def\ssececsize{1200} % Reduced fonts for @acro in text (10pt). \def\reducednominalsize{10pt} \setfont\reducedrm\rmshape{10}{1000}{OT1} \setfont\reducedtt\ttshape{10}{1000}{OT1TT} \setfont\reducedbf\bfshape{10}{1000}{OT1} \setfont\reducedit\itshape{10}{1000}{OT1IT} \setfont\reducedsl\slshape{10}{1000}{OT1} \setfont\reducedsf\sfshape{10}{1000}{OT1} \setfont\reducedsc\scshape{10}{1000}{OT1} \setfont\reducedttsl\ttslshape{10}{1000}{OT1TT} \font\reducedi=cmmi10 \font\reducedsy=cmsy10 \def\reducedecsize{1000} \textleading = 13.2pt % line spacing for 11pt CM \textfonts % reset the current fonts \rm } % end of 11pt text font size definitions % Definitions to make the main text be 10pt Computer Modern, with % section, chapter, etc., sizes following suit. This is for the GNU % Press printing of the Emacs 22 manual. Maybe other manuals in the % future. Used with @smallbook, which sets the leading to 12pt. % \def\definetextfontsizex{% % Text fonts (10pt). \def\textnominalsize{10pt} \edef\mainmagstep{1000} \setfont\textrm\rmshape{10}{\mainmagstep}{OT1} \setfont\texttt\ttshape{10}{\mainmagstep}{OT1TT} \setfont\textbf\bfshape{10}{\mainmagstep}{OT1} \setfont\textit\itshape{10}{\mainmagstep}{OT1IT} \setfont\textsl\slshape{10}{\mainmagstep}{OT1} \setfont\textsf\sfshape{10}{\mainmagstep}{OT1} \setfont\textsc\scshape{10}{\mainmagstep}{OT1} \setfont\textttsl\ttslshape{10}{\mainmagstep}{OT1TT} \font\texti=cmmi10 scaled \mainmagstep \font\textsy=cmsy10 scaled \mainmagstep \def\textecsize{1000} % A few fonts for @defun names and args. \setfont\defbf\bfshape{10}{\magstephalf}{OT1} \setfont\deftt\ttshape{10}{\magstephalf}{OT1TT} \setfont\defttsl\ttslshape{10}{\magstephalf}{OT1TT} \def\df{\let\tentt=\deftt \let\tenbf = \defbf \let\tenttsl=\defttsl \bf} % Fonts for indices, footnotes, small examples (9pt). \def\smallnominalsize{9pt} \setfont\smallrm\rmshape{9}{1000}{OT1} \setfont\smalltt\ttshape{9}{1000}{OT1TT} \setfont\smallbf\bfshape{10}{900}{OT1} \setfont\smallit\itshape{9}{1000}{OT1IT} \setfont\smallsl\slshape{9}{1000}{OT1} \setfont\smallsf\sfshape{9}{1000}{OT1} \setfont\smallsc\scshape{10}{900}{OT1} \setfont\smallttsl\ttslshape{10}{900}{OT1TT} \font\smalli=cmmi9 \font\smallsy=cmsy9 \def\smallecsize{0900} % Fonts for small examples (8pt). \def\smallernominalsize{8pt} \setfont\smallerrm\rmshape{8}{1000}{OT1} \setfont\smallertt\ttshape{8}{1000}{OT1TT} \setfont\smallerbf\bfshape{10}{800}{OT1} \setfont\smallerit\itshape{8}{1000}{OT1IT} \setfont\smallersl\slshape{8}{1000}{OT1} \setfont\smallersf\sfshape{8}{1000}{OT1} \setfont\smallersc\scshape{10}{800}{OT1} \setfont\smallerttsl\ttslshape{10}{800}{OT1TT} \font\smalleri=cmmi8 \font\smallersy=cmsy8 \def\smallerecsize{0800} % Fonts for title page (20.4pt): \def\titlenominalsize{20pt} \setfont\titlerm\rmbshape{12}{\magstep3}{OT1} \setfont\titleit\itbshape{10}{\magstep4}{OT1IT} \setfont\titlesl\slbshape{10}{\magstep4}{OT1} \setfont\titlett\ttbshape{12}{\magstep3}{OT1TT} \setfont\titlettsl\ttslshape{10}{\magstep4}{OT1TT} \setfont\titlesf\sfbshape{17}{\magstep1}{OT1} \let\titlebf=\titlerm \setfont\titlesc\scbshape{10}{\magstep4}{OT1} \font\titlei=cmmi12 scaled \magstep3 \font\titlesy=cmsy10 scaled \magstep4 \def\titleecsize{2074} % Chapter fonts (14.4pt). \def\chapnominalsize{14pt} \setfont\chaprm\rmbshape{12}{\magstep1}{OT1} \setfont\chapit\itbshape{10}{\magstep2}{OT1IT} \setfont\chapsl\slbshape{10}{\magstep2}{OT1} \setfont\chaptt\ttbshape{12}{\magstep1}{OT1TT} \setfont\chapttsl\ttslshape{10}{\magstep2}{OT1TT} \setfont\chapsf\sfbshape{12}{\magstep1}{OT1} \let\chapbf\chaprm \setfont\chapsc\scbshape{10}{\magstep2}{OT1} \font\chapi=cmmi12 scaled \magstep1 \font\chapsy=cmsy10 scaled \magstep2 \def\chapecsize{1440} % Section fonts (12pt). \def\secnominalsize{12pt} \setfont\secrm\rmbshape{12}{1000}{OT1} \setfont\secit\itbshape{10}{\magstep1}{OT1IT} \setfont\secsl\slbshape{10}{\magstep1}{OT1} \setfont\sectt\ttbshape{12}{1000}{OT1TT} \setfont\secttsl\ttslshape{10}{\magstep1}{OT1TT} \setfont\secsf\sfbshape{12}{1000}{OT1} \let\secbf\secrm \setfont\secsc\scbshape{10}{\magstep1}{OT1} \font\seci=cmmi12 \font\secsy=cmsy10 scaled \magstep1 \def\sececsize{1200} % Subsection fonts (10pt). \def\ssecnominalsize{10pt} \setfont\ssecrm\rmbshape{10}{1000}{OT1} \setfont\ssecit\itbshape{10}{1000}{OT1IT} \setfont\ssecsl\slbshape{10}{1000}{OT1} \setfont\ssectt\ttbshape{10}{1000}{OT1TT} \setfont\ssecttsl\ttslshape{10}{1000}{OT1TT} \setfont\ssecsf\sfbshape{10}{1000}{OT1} \let\ssecbf\ssecrm \setfont\ssecsc\scbshape{10}{1000}{OT1} \font\sseci=cmmi10 \font\ssecsy=cmsy10 \def\ssececsize{1000} % Reduced fonts for @acro in text (9pt). \def\reducednominalsize{9pt} \setfont\reducedrm\rmshape{9}{1000}{OT1} \setfont\reducedtt\ttshape{9}{1000}{OT1TT} \setfont\reducedbf\bfshape{10}{900}{OT1} \setfont\reducedit\itshape{9}{1000}{OT1IT} \setfont\reducedsl\slshape{9}{1000}{OT1} \setfont\reducedsf\sfshape{9}{1000}{OT1} \setfont\reducedsc\scshape{10}{900}{OT1} \setfont\reducedttsl\ttslshape{10}{900}{OT1TT} \font\reducedi=cmmi9 \font\reducedsy=cmsy9 \def\reducedecsize{0900} \divide\parskip by 2 % reduce space between paragraphs \textleading = 12pt % line spacing for 10pt CM \textfonts % reset the current fonts \rm } % end of 10pt text font size definitions % We provide the user-level command % @fonttextsize 10 % (or 11) to redefine the text font size. pt is assumed. % \def\xiword{11} \def\xword{10} \def\xwordpt{10pt} % \parseargdef\fonttextsize{% \def\textsizearg{#1}% %\wlog{doing @fonttextsize \textsizearg}% % % Set \globaldefs so that documents can use this inside @tex, since % makeinfo 4.8 does not support it, but we need it nonetheless. % \begingroup \globaldefs=1 \ifx\textsizearg\xword \definetextfontsizex \else \ifx\textsizearg\xiword \definetextfontsizexi \else \errhelp=\EMsimple \errmessage{@fonttextsize only supports `10' or `11', not `\textsizearg'} \fi\fi \endgroup } % In order for the font changes to affect most math symbols and letters, % we have to define the \textfont of the standard families. Since % texinfo doesn't allow for producing subscripts and superscripts except % in the main text, we don't bother to reset \scriptfont and % \scriptscriptfont (which would also require loading a lot more fonts). % \def\resetmathfonts{% \textfont0=\tenrm \textfont1=\teni \textfont2=\tensy \textfont\itfam=\tenit \textfont\slfam=\tensl \textfont\bffam=\tenbf \textfont\ttfam=\tentt \textfont\sffam=\tensf } % The font-changing commands redefine the meanings of \tenSTYLE, instead % of just \STYLE. We do this because \STYLE needs to also set the % current \fam for math mode. Our \STYLE (e.g., \rm) commands hardwire % \tenSTYLE to set the current font. % % Each font-changing command also sets the names \lsize (one size lower) % and \lllsize (three sizes lower). These relative commands are used in % the LaTeX logo and acronyms. % % This all needs generalizing, badly. % \def\textfonts{% \let\tenrm=\textrm \let\tenit=\textit \let\tensl=\textsl \let\tenbf=\textbf \let\tentt=\texttt \let\smallcaps=\textsc \let\tensf=\textsf \let\teni=\texti \let\tensy=\textsy \let\tenttsl=\textttsl \def\curfontsize{text}% \def\lsize{reduced}\def\lllsize{smaller}% \resetmathfonts \setleading{\textleading}} \def\titlefonts{% \let\tenrm=\titlerm \let\tenit=\titleit \let\tensl=\titlesl \let\tenbf=\titlebf \let\tentt=\titlett \let\smallcaps=\titlesc \let\tensf=\titlesf \let\teni=\titlei \let\tensy=\titlesy \let\tenttsl=\titlettsl \def\curfontsize{title}% \def\lsize{chap}\def\lllsize{subsec}% \resetmathfonts \setleading{27pt}} \def\titlefont#1{{\titlefonts\rmisbold #1}} \def\chapfonts{% \let\tenrm=\chaprm \let\tenit=\chapit \let\tensl=\chapsl \let\tenbf=\chapbf \let\tentt=\chaptt \let\smallcaps=\chapsc \let\tensf=\chapsf \let\teni=\chapi \let\tensy=\chapsy \let\tenttsl=\chapttsl \def\curfontsize{chap}% \def\lsize{sec}\def\lllsize{text}% \resetmathfonts \setleading{19pt}} \def\secfonts{% \let\tenrm=\secrm \let\tenit=\secit \let\tensl=\secsl \let\tenbf=\secbf \let\tentt=\sectt \let\smallcaps=\secsc \let\tensf=\secsf \let\teni=\seci \let\tensy=\secsy \let\tenttsl=\secttsl \def\curfontsize{sec}% \def\lsize{subsec}\def\lllsize{reduced}% \resetmathfonts \setleading{16pt}} \def\subsecfonts{% \let\tenrm=\ssecrm \let\tenit=\ssecit \let\tensl=\ssecsl \let\tenbf=\ssecbf \let\tentt=\ssectt \let\smallcaps=\ssecsc \let\tensf=\ssecsf \let\teni=\sseci \let\tensy=\ssecsy \let\tenttsl=\ssecttsl \def\curfontsize{ssec}% \def\lsize{text}\def\lllsize{small}% \resetmathfonts \setleading{15pt}} \let\subsubsecfonts = \subsecfonts \def\reducedfonts{% \let\tenrm=\reducedrm \let\tenit=\reducedit \let\tensl=\reducedsl \let\tenbf=\reducedbf \let\tentt=\reducedtt \let\reducedcaps=\reducedsc \let\tensf=\reducedsf \let\teni=\reducedi \let\tensy=\reducedsy \let\tenttsl=\reducedttsl \def\curfontsize{reduced}% \def\lsize{small}\def\lllsize{smaller}% \resetmathfonts \setleading{10.5pt}} \def\smallfonts{% \let\tenrm=\smallrm \let\tenit=\smallit \let\tensl=\smallsl \let\tenbf=\smallbf \let\tentt=\smalltt \let\smallcaps=\smallsc \let\tensf=\smallsf \let\teni=\smalli \let\tensy=\smallsy \let\tenttsl=\smallttsl \def\curfontsize{small}% \def\lsize{smaller}\def\lllsize{smaller}% \resetmathfonts \setleading{10.5pt}} \def\smallerfonts{% \let\tenrm=\smallerrm \let\tenit=\smallerit \let\tensl=\smallersl \let\tenbf=\smallerbf \let\tentt=\smallertt \let\smallcaps=\smallersc \let\tensf=\smallersf \let\teni=\smalleri \let\tensy=\smallersy \let\tenttsl=\smallerttsl \def\curfontsize{smaller}% \def\lsize{smaller}\def\lllsize{smaller}% \resetmathfonts \setleading{9.5pt}} % Fonts for short table of contents. \setfont\shortcontrm\rmshape{12}{1000}{OT1} \setfont\shortcontbf\bfshape{10}{\magstep1}{OT1} % no cmb12 \setfont\shortcontsl\slshape{12}{1000}{OT1} \setfont\shortconttt\ttshape{12}{1000}{OT1TT} % Define these just so they can be easily changed for other fonts. \def\angleleft{$\langle$} \def\angleright{$\rangle$} % Set the fonts to use with the @small... environments. \let\smallexamplefonts = \smallfonts % About \smallexamplefonts. If we use \smallfonts (9pt), @smallexample % can fit this many characters: % 8.5x11=86 smallbook=72 a4=90 a5=69 % If we use \scriptfonts (8pt), then we can fit this many characters: % 8.5x11=90+ smallbook=80 a4=90+ a5=77 % For me, subjectively, the few extra characters that fit aren't worth % the additional smallness of 8pt. So I'm making the default 9pt. % % By the way, for comparison, here's what fits with @example (10pt): % 8.5x11=71 smallbook=60 a4=75 a5=58 % --karl, 24jan03. % Set up the default fonts, so we can use them for creating boxes. % \definetextfontsizexi \message{markup,} % Check if we are currently using a typewriter font. Since all the % Computer Modern typewriter fonts have zero interword stretch (and % shrink), and it is reasonable to expect all typewriter fonts to have % this property, we can check that font parameter. % \def\ifmonospace{\ifdim\fontdimen3\font=0pt } % Markup style infrastructure. \defmarkupstylesetup\INITMACRO will % define and register \INITMACRO to be called on markup style changes. % \INITMACRO can check \currentmarkupstyle for the innermost % style and the set of \ifmarkupSTYLE switches for all styles % currently in effect. \newif\ifmarkupvar \newif\ifmarkupsamp \newif\ifmarkupkey %\newif\ifmarkupfile % @file == @samp. %\newif\ifmarkupoption % @option == @samp. \newif\ifmarkupcode \newif\ifmarkupkbd %\newif\ifmarkupenv % @env == @code. %\newif\ifmarkupcommand % @command == @code. \newif\ifmarkuptex % @tex (and part of @math, for now). \newif\ifmarkupexample \newif\ifmarkupverb \newif\ifmarkupverbatim \let\currentmarkupstyle\empty \def\setupmarkupstyle#1{% \csname markup#1true\endcsname \def\currentmarkupstyle{#1}% \markupstylesetup } \let\markupstylesetup\empty \def\defmarkupstylesetup#1{% \expandafter\def\expandafter\markupstylesetup \expandafter{\markupstylesetup #1}% \def#1% } % Markup style setup for left and right quotes. \defmarkupstylesetup\markupsetuplq{% \expandafter\let\expandafter \temp \csname markupsetuplq\currentmarkupstyle\endcsname \ifx\temp\relax \markupsetuplqdefault \else \temp \fi } \defmarkupstylesetup\markupsetuprq{% \expandafter\let\expandafter \temp \csname markupsetuprq\currentmarkupstyle\endcsname \ifx\temp\relax \markupsetuprqdefault \else \temp \fi } { \catcode`\'=\active \catcode`\`=\active \gdef\markupsetuplqdefault{\let`\lq} \gdef\markupsetuprqdefault{\let'\rq} \gdef\markupsetcodequoteleft{\let`\codequoteleft} \gdef\markupsetcodequoteright{\let'\codequoteright} \gdef\markupsetnoligaturesquoteleft{\let`\noligaturesquoteleft} } \let\markupsetuplqcode \markupsetcodequoteleft \let\markupsetuprqcode \markupsetcodequoteright % \let\markupsetuplqexample \markupsetcodequoteleft \let\markupsetuprqexample \markupsetcodequoteright % \let\markupsetuplqsamp \markupsetcodequoteleft \let\markupsetuprqsamp \markupsetcodequoteright % \let\markupsetuplqverb \markupsetcodequoteleft \let\markupsetuprqverb \markupsetcodequoteright % \let\markupsetuplqverbatim \markupsetcodequoteleft \let\markupsetuprqverbatim \markupsetcodequoteright \let\markupsetuplqkbd \markupsetnoligaturesquoteleft % Allow an option to not use regular directed right quote/apostrophe % (char 0x27), but instead the undirected quote from cmtt (char 0x0d). % The undirected quote is ugly, so don't make it the default, but it % works for pasting with more pdf viewers (at least evince), the % lilypond developers report. xpdf does work with the regular 0x27. % \def\codequoteright{% \expandafter\ifx\csname SETtxicodequoteundirected\endcsname\relax \expandafter\ifx\csname SETcodequoteundirected\endcsname\relax '% \else \char'15 \fi \else \char'15 \fi } % % and a similar option for the left quote char vs. a grave accent. % Modern fonts display ASCII 0x60 as a grave accent, so some people like % the code environments to do likewise. % \def\codequoteleft{% \expandafter\ifx\csname SETtxicodequotebacktick\endcsname\relax \expandafter\ifx\csname SETcodequotebacktick\endcsname\relax % [Knuth] pp. 380,381,391 % \relax disables Spanish ligatures ?` and !` of \tt font. \relax`% \else \char'22 \fi \else \char'22 \fi } % Commands to set the quote options. % \parseargdef\codequoteundirected{% \def\temp{#1}% \ifx\temp\onword \expandafter\let\csname SETtxicodequoteundirected\endcsname = t% \else\ifx\temp\offword \expandafter\let\csname SETtxicodequoteundirected\endcsname = \relax \else \errhelp = \EMsimple \errmessage{Unknown @codequoteundirected value `\temp', must be on|off}% \fi\fi } % \parseargdef\codequotebacktick{% \def\temp{#1}% \ifx\temp\onword \expandafter\let\csname SETtxicodequotebacktick\endcsname = t% \else\ifx\temp\offword \expandafter\let\csname SETtxicodequotebacktick\endcsname = \relax \else \errhelp = \EMsimple \errmessage{Unknown @codequotebacktick value `\temp', must be on|off}% \fi\fi } % [Knuth] pp. 380,381,391, disable Spanish ligatures ?` and !` of \tt font. \def\noligaturesquoteleft{\relax\lq} % Count depth in font-changes, for error checks \newcount\fontdepth \fontdepth=0 % Font commands. % #1 is the font command (\sl or \it), #2 is the text to slant. % If we are in a monospaced environment, however, 1) always use \ttsl, % and 2) do not add an italic correction. \def\dosmartslant#1#2{% \ifusingtt {{\ttsl #2}\let\next=\relax}% {\def\next{{#1#2}\futurelet\next\smartitaliccorrection}}% \next } \def\smartslanted{\dosmartslant\sl} \def\smartitalic{\dosmartslant\it} % Output an italic correction unless \next (presumed to be the following % character) is such as not to need one. \def\smartitaliccorrection{% \ifx\next,% \else\ifx\next-% \else\ifx\next.% \else\ptexslash \fi\fi\fi \aftersmartic } % like \smartslanted except unconditionally uses \ttsl, and no ic. % @var is set to this for defun arguments. \def\ttslanted#1{{\ttsl #1}} % @cite is like \smartslanted except unconditionally use \sl. We never want % ttsl for book titles, do we? \def\cite#1{{\sl #1}\futurelet\next\smartitaliccorrection} \def\aftersmartic{} \def\var#1{% \let\saveaftersmartic = \aftersmartic \def\aftersmartic{\null\let\aftersmartic=\saveaftersmartic}% \smartslanted{#1}% } \let\i=\smartitalic \let\slanted=\smartslanted \let\dfn=\smartslanted \let\emph=\smartitalic % Explicit font changes: @r, @sc, undocumented @ii. \def\r#1{{\rm #1}} % roman font \def\sc#1{{\smallcaps#1}} % smallcaps font \def\ii#1{{\it #1}} % italic font % @b, explicit bold. Also @strong. \def\b#1{{\bf #1}} \let\strong=\b % @sansserif, explicit sans. \def\sansserif#1{{\sf #1}} % We can't just use \exhyphenpenalty, because that only has effect at % the end of a paragraph. Restore normal hyphenation at the end of the % group within which \nohyphenation is presumably called. % \def\nohyphenation{\hyphenchar\font = -1 \aftergroup\restorehyphenation} \def\restorehyphenation{\hyphenchar\font = `- } % Set sfcode to normal for the chars that usually have another value. % Can't use plain's \frenchspacing because it uses the `\x notation, and % sometimes \x has an active definition that messes things up. % \catcode`@=11 \def\plainfrenchspacing{% \sfcode\dotChar =\@m \sfcode\questChar=\@m \sfcode\exclamChar=\@m \sfcode\colonChar=\@m \sfcode\semiChar =\@m \sfcode\commaChar =\@m \def\endofsentencespacefactor{1000}% for @. and friends } \def\plainnonfrenchspacing{% \sfcode`\.3000\sfcode`\?3000\sfcode`\!3000 \sfcode`\:2000\sfcode`\;1500\sfcode`\,1250 \def\endofsentencespacefactor{3000}% for @. and friends } \catcode`@=\other \def\endofsentencespacefactor{3000}% default % @t, explicit typewriter. \def\t#1{% {\tt \rawbackslash \plainfrenchspacing #1}% \null } % @samp. \def\samp#1{{\setupmarkupstyle{samp}\lq\tclose{#1}\rq\null}} % definition of @key that produces a lozenge. Doesn't adjust to text size. %\setfont\keyrm\rmshape{8}{1000}{OT1} %\font\keysy=cmsy9 %\def\key#1{{\keyrm\textfont2=\keysy \leavevmode\hbox{% % \raise0.4pt\hbox{\angleleft}\kern-.08em\vtop{% % \vbox{\hrule\kern-0.4pt % \hbox{\raise0.4pt\hbox{\vphantom{\angleleft}}#1}}% % \kern-0.4pt\hrule}% % \kern-.06em\raise0.4pt\hbox{\angleright}}}} % definition of @key with no lozenge. If the current font is already % monospace, don't change it; that way, we respect @kbdinputstyle. But % if it isn't monospace, then use \tt. % \def\key#1{{\setupmarkupstyle{key}% \nohyphenation \ifmonospace\else\tt\fi #1}\null} % ctrl is no longer a Texinfo command. \def\ctrl #1{{\tt \rawbackslash \hat}#1} % @file, @option are the same as @samp. \let\file=\samp \let\option=\samp % @code is a modification of @t, % which makes spaces the same size as normal in the surrounding text. \def\tclose#1{% {% % Change normal interword space to be same as for the current font. \spaceskip = \fontdimen2\font % % Switch to typewriter. \tt % % But `\ ' produces the large typewriter interword space. \def\ {{\spaceskip = 0pt{} }}% % % Turn off hyphenation. \nohyphenation % \rawbackslash \plainfrenchspacing #1% }% \null % reset spacefactor to 1000 } % We *must* turn on hyphenation at `-' and `_' in @code. % Otherwise, it is too hard to avoid overfull hboxes % in the Emacs manual, the Library manual, etc. % Unfortunately, TeX uses one parameter (\hyphenchar) to control % both hyphenation at - and hyphenation within words. % We must therefore turn them both off (\tclose does that) % and arrange explicitly to hyphenate at a dash. % -- rms. { \catcode`\-=\active \catcode`\_=\active \catcode`\'=\active \catcode`\`=\active \global\let'=\rq \global\let`=\lq % default definitions % \global\def\code{\begingroup \setupmarkupstyle{code}% % The following should really be moved into \setupmarkupstyle handlers. \catcode\dashChar=\active \catcode\underChar=\active \ifallowcodebreaks \let-\codedash \let_\codeunder \else \let-\realdash \let_\realunder \fi \codex } } \def\codex #1{\tclose{#1}\endgroup} \def\realdash{-} \def\codedash{-\discretionary{}{}{}} \def\codeunder{% % this is all so @math{@code{var_name}+1} can work. In math mode, _ % is "active" (mathcode"8000) and \normalunderscore (or \char95, etc.) % will therefore expand the active definition of _, which is us % (inside @code that is), therefore an endless loop. \ifusingtt{\ifmmode \mathchar"075F % class 0=ordinary, family 7=ttfam, pos 0x5F=_. \else\normalunderscore \fi \discretionary{}{}{}}% {\_}% } % An additional complication: the above will allow breaks after, e.g., % each of the four underscores in __typeof__. This is undesirable in % some manuals, especially if they don't have long identifiers in % general. @allowcodebreaks provides a way to control this. % \newif\ifallowcodebreaks \allowcodebreakstrue \def\keywordtrue{true} \def\keywordfalse{false} \parseargdef\allowcodebreaks{% \def\txiarg{#1}% \ifx\txiarg\keywordtrue \allowcodebreakstrue \else\ifx\txiarg\keywordfalse \allowcodebreaksfalse \else \errhelp = \EMsimple \errmessage{Unknown @allowcodebreaks option `\txiarg', must be true|false}% \fi\fi } % @uref (abbreviation for `urlref') takes an optional (comma-separated) % second argument specifying the text to display and an optional third % arg as text to display instead of (rather than in addition to) the url % itself. First (mandatory) arg is the url. % (This \urefnobreak definition isn't used now, leaving it for a while % for comparison.) \def\urefnobreak#1{\dourefnobreak #1,,,\finish} \def\dourefnobreak#1,#2,#3,#4\finish{\begingroup \unsepspaces \pdfurl{#1}% \setbox0 = \hbox{\ignorespaces #3}% \ifdim\wd0 > 0pt \unhbox0 % third arg given, show only that \else \setbox0 = \hbox{\ignorespaces #2}% \ifdim\wd0 > 0pt \ifpdf \unhbox0 % PDF: 2nd arg given, show only it \else \unhbox0\ (\code{#1})% DVI: 2nd arg given, show both it and url \fi \else \code{#1}% only url given, so show it \fi \fi \endlink \endgroup} % This \urefbreak definition is the active one. \def\urefbreak{\begingroup \urefcatcodes \dourefbreak} \let\uref=\urefbreak \def\dourefbreak#1{\urefbreakfinish #1,,,\finish} \def\urefbreakfinish#1,#2,#3,#4\finish{% doesn't work in @example \unsepspaces \pdfurl{#1}% \setbox0 = \hbox{\ignorespaces #3}% \ifdim\wd0 > 0pt \unhbox0 % third arg given, show only that \else \setbox0 = \hbox{\ignorespaces #2}% \ifdim\wd0 > 0pt \ifpdf \unhbox0 % PDF: 2nd arg given, show only it \else \unhbox0\ (\urefcode{#1})% DVI: 2nd arg given, show both it and url \fi \else \urefcode{#1}% only url given, so show it \fi \fi \endlink \endgroup} % Allow line breaks around only a few characters (only). \def\urefcatcodes{% \catcode\ampChar=\active \catcode\dotChar=\active \catcode\hashChar=\active \catcode\questChar=\active \catcode\slashChar=\active } { \urefcatcodes % \global\def\urefcode{\begingroup \setupmarkupstyle{code}% \urefcatcodes \let&\urefcodeamp \let.\urefcodedot \let#\urefcodehash \let?\urefcodequest \let/\urefcodeslash \codex } % % By default, they are just regular characters. \global\def&{\normalamp} \global\def.{\normaldot} \global\def#{\normalhash} \global\def?{\normalquest} \global\def/{\normalslash} } % we put a little stretch before and after the breakable chars, to help % line breaking of long url's. The unequal skips make look better in % cmtt at least, especially for dots. \def\urefprestretch{\urefprebreak \hskip0pt plus.13em } \def\urefpoststretch{\urefpostbreak \hskip0pt plus.1em } % \def\urefcodeamp{\urefprestretch \&\urefpoststretch} \def\urefcodedot{\urefprestretch .\urefpoststretch} \def\urefcodehash{\urefprestretch \#\urefpoststretch} \def\urefcodequest{\urefprestretch ?\urefpoststretch} \def\urefcodeslash{\futurelet\next\urefcodeslashfinish} { \catcode`\/=\active \global\def\urefcodeslashfinish{% \urefprestretch \slashChar % Allow line break only after the final / in a sequence of % slashes, to avoid line break between the slashes in http://. \ifx\next/\else \urefpoststretch \fi } } % One more complication: by default we'll break after the special % characters, but some people like to break before the special chars, so % allow that. Also allow no breaking at all, for manual control. % \parseargdef\urefbreakstyle{% \def\txiarg{#1}% \ifx\txiarg\wordnone \def\urefprebreak{\nobreak}\def\urefpostbreak{\nobreak} \else\ifx\txiarg\wordbefore \def\urefprebreak{\allowbreak}\def\urefpostbreak{\nobreak} \else\ifx\txiarg\wordafter \def\urefprebreak{\nobreak}\def\urefpostbreak{\allowbreak} \else \errhelp = \EMsimple \errmessage{Unknown @urefbreakstyle setting `\txiarg'}% \fi\fi\fi } \def\wordafter{after} \def\wordbefore{before} \def\wordnone{none} \urefbreakstyle after % @url synonym for @uref, since that's how everyone uses it. % \let\url=\uref % rms does not like angle brackets --karl, 17may97. % So now @email is just like @uref, unless we are pdf. % %\def\email#1{\angleleft{\tt #1}\angleright} \ifpdf \def\email#1{\doemail#1,,\finish} \def\doemail#1,#2,#3\finish{\begingroup \unsepspaces \pdfurl{mailto:#1}% \setbox0 = \hbox{\ignorespaces #2}% \ifdim\wd0>0pt\unhbox0\else\code{#1}\fi \endlink \endgroup} \else \let\email=\uref \fi % @kbd is like @code, except that if the argument is just one @key command, % then @kbd has no effect. \def\kbd#1{{\setupmarkupstyle{kbd}\def\look{#1}\expandafter\kbdfoo\look??\par}} % @kbdinputstyle -- arg is `distinct' (@kbd uses slanted tty font always), % `example' (@kbd uses ttsl only inside of @example and friends), % or `code' (@kbd uses normal tty font always). \parseargdef\kbdinputstyle{% \def\txiarg{#1}% \ifx\txiarg\worddistinct \gdef\kbdexamplefont{\ttsl}\gdef\kbdfont{\ttsl}% \else\ifx\txiarg\wordexample \gdef\kbdexamplefont{\ttsl}\gdef\kbdfont{\tt}% \else\ifx\txiarg\wordcode \gdef\kbdexamplefont{\tt}\gdef\kbdfont{\tt}% \else \errhelp = \EMsimple \errmessage{Unknown @kbdinputstyle setting `\txiarg'}% \fi\fi\fi } \def\worddistinct{distinct} \def\wordexample{example} \def\wordcode{code} % Default is `distinct'. \kbdinputstyle distinct \def\xkey{\key} \def\kbdfoo#1#2#3\par{\def\one{#1}\def\three{#3}\def\threex{??}% \ifx\one\xkey\ifx\threex\three \key{#2}% \else{\tclose{\kbdfont\setupmarkupstyle{kbd}\look}}\fi \else{\tclose{\kbdfont\setupmarkupstyle{kbd}\look}}\fi} % For @indicateurl, @env, @command quotes seem unnecessary, so use \code. \let\indicateurl=\code \let\env=\code \let\command=\code % @clicksequence{File @click{} Open ...} \def\clicksequence#1{\begingroup #1\endgroup} % @clickstyle @arrow (by default) \parseargdef\clickstyle{\def\click{#1}} \def\click{\arrow} % Typeset a dimension, e.g., `in' or `pt'. The only reason for the % argument is to make the input look right: @dmn{pt} instead of @dmn{}pt. % \def\dmn#1{\thinspace #1} % @l was never documented to mean ``switch to the Lisp font'', % and it is not used as such in any manual I can find. We need it for % Polish suppressed-l. --karl, 22sep96. %\def\l#1{{\li #1}\null} % @acronym for "FBI", "NATO", and the like. % We print this one point size smaller, since it's intended for % all-uppercase. % \def\acronym#1{\doacronym #1,,\finish} \def\doacronym#1,#2,#3\finish{% {\selectfonts\lsize #1}% \def\temp{#2}% \ifx\temp\empty \else \space ({\unsepspaces \ignorespaces \temp \unskip})% \fi \null % reset \spacefactor=1000 } % @abbr for "Comput. J." and the like. % No font change, but don't do end-of-sentence spacing. % \def\abbr#1{\doabbr #1,,\finish} \def\doabbr#1,#2,#3\finish{% {\plainfrenchspacing #1}% \def\temp{#2}% \ifx\temp\empty \else \space ({\unsepspaces \ignorespaces \temp \unskip})% \fi \null % reset \spacefactor=1000 } % @asis just yields its argument. Used with @table, for example. % \def\asis#1{#1} % @math outputs its argument in math mode. % % One complication: _ usually means subscripts, but it could also mean % an actual _ character, as in @math{@var{some_variable} + 1}. So make % _ active, and distinguish by seeing if the current family is \slfam, % which is what @var uses. { \catcode`\_ = \active \gdef\mathunderscore{% \catcode`\_=\active \def_{\ifnum\fam=\slfam \_\else\sb\fi}% } } % Another complication: we want \\ (and @\) to output a math (or tt) \. % FYI, plain.tex uses \\ as a temporary control sequence (for no % particular reason), but this is not advertised and we don't care. % % The \mathchar is class=0=ordinary, family=7=ttfam, position=5C=\. \def\mathbackslash{\ifnum\fam=\ttfam \mathchar"075C \else\backslash \fi} % \def\math{% \tex \mathunderscore \let\\ = \mathbackslash \mathactive % make the texinfo accent commands work in math mode \let\"=\ddot \let\'=\acute \let\==\bar \let\^=\hat \let\`=\grave \let\u=\breve \let\v=\check \let\~=\tilde \let\dotaccent=\dot $\finishmath } \def\finishmath#1{#1$\endgroup} % Close the group opened by \tex. % Some active characters (such as <) are spaced differently in math. % We have to reset their definitions in case the @math was an argument % to a command which sets the catcodes (such as @item or @section). % { \catcode`^ = \active \catcode`< = \active \catcode`> = \active \catcode`+ = \active \catcode`' = \active \gdef\mathactive{% \let^ = \ptexhat \let< = \ptexless \let> = \ptexgtr \let+ = \ptexplus \let' = \ptexquoteright } } % @inlinefmt{FMTNAME,PROCESSED-TEXT} and @inlineraw{FMTNAME,RAW-TEXT}. % Ignore unless FMTNAME == tex; then it is like @iftex and @tex, % except specified as a normal braced arg, so no newlines to worry about. % \def\outfmtnametex{tex} % \long\def\inlinefmt#1{\doinlinefmt #1,\finish} \long\def\doinlinefmt#1,#2,\finish{% \def\inlinefmtname{#1}% \ifx\inlinefmtname\outfmtnametex \ignorespaces #2\fi } % For raw, must switch into @tex before parsing the argument, to avoid % setting catcodes prematurely. Doing it this way means that, for % example, @inlineraw{html, foo{bar} gets a parse error instead of being % ignored. But this isn't important because if people want a literal % *right* brace they would have to use a command anyway, so they may as % well use a command to get a left brace too. We could re-use the % delimiter character idea from \verb, but it seems like overkill. % \long\def\inlineraw{\tex \doinlineraw} \long\def\doinlineraw#1{\doinlinerawtwo #1,\finish} \def\doinlinerawtwo#1,#2,\finish{% \def\inlinerawname{#1}% \ifx\inlinerawname\outfmtnametex \ignorespaces #2\fi \endgroup % close group opened by \tex. } \message{glyphs,} % and logos. % @@ prints an @, as does @atchar{}. \def\@{\char64 } \let\atchar=\@ % @{ @} @lbracechar{} @rbracechar{} all generate brace characters. % Unless we're in typewriter, use \ecfont because the CM text fonts do % not have braces, and we don't want to switch into math. \def\mylbrace{{\ifmonospace\else\ecfont\fi \char123}} \def\myrbrace{{\ifmonospace\else\ecfont\fi \char125}} \let\{=\mylbrace \let\lbracechar=\{ \let\}=\myrbrace \let\rbracechar=\} \begingroup % Definitions to produce \{ and \} commands for indices, % and @{ and @} for the aux/toc files. \catcode`\{ = \other \catcode`\} = \other \catcode`\[ = 1 \catcode`\] = 2 \catcode`\! = 0 \catcode`\\ = \other !gdef!lbracecmd[\{]% !gdef!rbracecmd[\}]% !gdef!lbraceatcmd[@{]% !gdef!rbraceatcmd[@}]% !endgroup % @comma{} to avoid , parsing problems. \let\comma = , % Accents: @, @dotaccent @ringaccent @ubaraccent @udotaccent % Others are defined by plain TeX: @` @' @" @^ @~ @= @u @v @H. \let\, = \ptexc \let\dotaccent = \ptexdot \def\ringaccent#1{{\accent23 #1}} \let\tieaccent = \ptext \let\ubaraccent = \ptexb \let\udotaccent = \d % Other special characters: @questiondown @exclamdown @ordf @ordm % Plain TeX defines: @AA @AE @O @OE @L (plus lowercase versions) @ss. \def\questiondown{?`} \def\exclamdown{!`} \def\ordf{\leavevmode\raise1ex\hbox{\selectfonts\lllsize \underbar{a}}} \def\ordm{\leavevmode\raise1ex\hbox{\selectfonts\lllsize \underbar{o}}} % Dotless i and dotless j, used for accents. \def\imacro{i} \def\jmacro{j} \def\dotless#1{% \def\temp{#1}% \ifx\temp\imacro \ifmmode\imath \else\ptexi \fi \else\ifx\temp\jmacro \ifmmode\jmath \else\j \fi \else \errmessage{@dotless can be used only with i or j}% \fi\fi } % The \TeX{} logo, as in plain, but resetting the spacing so that a % period following counts as ending a sentence. (Idea found in latex.) % \edef\TeX{\TeX \spacefactor=1000 } % @LaTeX{} logo. Not quite the same results as the definition in % latex.ltx, since we use a different font for the raised A; it's most % convenient for us to use an explicitly smaller font, rather than using % the \scriptstyle font (since we don't reset \scriptstyle and % \scriptscriptstyle). % \def\LaTeX{% L\kern-.36em {\setbox0=\hbox{T}% \vbox to \ht0{\hbox{% \ifx\textnominalsize\xwordpt % for 10pt running text, \lllsize (8pt) is too small for the A in LaTeX. % Revert to plain's \scriptsize, which is 7pt. \count255=\the\fam $\fam\count255 \scriptstyle A$% \else % For 11pt, we can use our lllsize. \selectfonts\lllsize A% \fi }% \vss }}% \kern-.15em \TeX } % Some math mode symbols. \def\bullet{$\ptexbullet$} \def\geq{\ifmmode \ge\else $\ge$\fi} \def\leq{\ifmmode \le\else $\le$\fi} \def\minus{\ifmmode -\else $-$\fi} % @dots{} outputs an ellipsis using the current font. % We do .5em per period so that it has the same spacing in the cm % typewriter fonts as three actual period characters; on the other hand, % in other typewriter fonts three periods are wider than 1.5em. So do % whichever is larger. % \def\dots{% \leavevmode \setbox0=\hbox{...}% get width of three periods \ifdim\wd0 > 1.5em \dimen0 = \wd0 \else \dimen0 = 1.5em \fi \hbox to \dimen0{% \hskip 0pt plus.25fil .\hskip 0pt plus1fil .\hskip 0pt plus1fil .\hskip 0pt plus.5fil }% } % @enddots{} is an end-of-sentence ellipsis. % \def\enddots{% \dots \spacefactor=\endofsentencespacefactor } % @point{}, @result{}, @expansion{}, @print{}, @equiv{}. % % Since these characters are used in examples, they should be an even number of % \tt widths. Each \tt character is 1en, so two makes it 1em. % \def\point{$\star$} \def\arrow{\leavevmode\raise.05ex\hbox to 1em{\hfil$\rightarrow$\hfil}} \def\result{\leavevmode\raise.05ex\hbox to 1em{\hfil$\Rightarrow$\hfil}} \def\expansion{\leavevmode\hbox to 1em{\hfil$\mapsto$\hfil}} \def\print{\leavevmode\lower.1ex\hbox to 1em{\hfil$\dashv$\hfil}} \def\equiv{\leavevmode\hbox to 1em{\hfil$\ptexequiv$\hfil}} % The @error{} command. % Adapted from the TeXbook's \boxit. % \newbox\errorbox % {\tentt \global\dimen0 = 3em}% Width of the box. \dimen2 = .55pt % Thickness of rules % The text. (`r' is open on the right, `e' somewhat less so on the left.) \setbox0 = \hbox{\kern-.75pt \reducedsf \putworderror\kern-1.5pt} % \setbox\errorbox=\hbox to \dimen0{\hfil \hsize = \dimen0 \advance\hsize by -5.8pt % Space to left+right. \advance\hsize by -2\dimen2 % Rules. \vbox{% \hrule height\dimen2 \hbox{\vrule width\dimen2 \kern3pt % Space to left of text. \vtop{\kern2.4pt \box0 \kern2.4pt}% Space above/below. \kern3pt\vrule width\dimen2}% Space to right. \hrule height\dimen2} \hfil} % \def\error{\leavevmode\lower.7ex\copy\errorbox} % @pounds{} is a sterling sign, which Knuth put in the CM italic font. % \def\pounds{{\it\$}} % @euro{} comes from a separate font, depending on the current style. % We use the free feym* fonts from the eurosym package by Henrik % Theiling, which support regular, slanted, bold and bold slanted (and % "outlined" (blackboard board, sort of) versions, which we don't need). % It is available from http://www.ctan.org/tex-archive/fonts/eurosym. % % Although only regular is the truly official Euro symbol, we ignore % that. The Euro is designed to be slightly taller than the regular % font height. % % feymr - regular % feymo - slanted % feybr - bold % feybo - bold slanted % % There is no good (free) typewriter version, to my knowledge. % A feymr10 euro is ~7.3pt wide, while a normal cmtt10 char is ~5.25pt wide. % Hmm. % % Also doesn't work in math. Do we need to do math with euro symbols? % Hope not. % % \def\euro{{\eurofont e}} \def\eurofont{% % We set the font at each command, rather than predefining it in % \textfonts and the other font-switching commands, so that % installations which never need the symbol don't have to have the % font installed. % % There is only one designed size (nominal 10pt), so we always scale % that to the current nominal size. % % By the way, simply using "at 1em" works for cmr10 and the like, but % does not work for cmbx10 and other extended/shrunken fonts. % \def\eurosize{\csname\curfontsize nominalsize\endcsname}% % \ifx\curfontstyle\bfstylename % bold: \font\thiseurofont = \ifusingit{feybo10}{feybr10} at \eurosize \else % regular: \font\thiseurofont = \ifusingit{feymo10}{feymr10} at \eurosize \fi \thiseurofont } % Glyphs from the EC fonts. We don't use \let for the aliases, because % sometimes we redefine the original macro, and the alias should reflect % the redefinition. % % Use LaTeX names for the Icelandic letters. \def\DH{{\ecfont \char"D0}} % Eth \def\dh{{\ecfont \char"F0}} % eth \def\TH{{\ecfont \char"DE}} % Thorn \def\th{{\ecfont \char"FE}} % thorn % \def\guillemetleft{{\ecfont \char"13}} \def\guillemotleft{\guillemetleft} \def\guillemetright{{\ecfont \char"14}} \def\guillemotright{\guillemetright} \def\guilsinglleft{{\ecfont \char"0E}} \def\guilsinglright{{\ecfont \char"0F}} \def\quotedblbase{{\ecfont \char"12}} \def\quotesinglbase{{\ecfont \char"0D}} % % This positioning is not perfect (see the ogonek LaTeX package), but % we have the precomposed glyphs for the most common cases. We put the % tests to use those glyphs in the single \ogonek macro so we have fewer % dummy definitions to worry about for index entries, etc. % % ogonek is also used with other letters in Lithuanian (IOU), but using % the precomposed glyphs for those is not so easy since they aren't in % the same EC font. \def\ogonek#1{{% \def\temp{#1}% \ifx\temp\macrocharA\Aogonek \else\ifx\temp\macrochara\aogonek \else\ifx\temp\macrocharE\Eogonek \else\ifx\temp\macrochare\eogonek \else \ecfont \setbox0=\hbox{#1}% \ifdim\ht0=1ex\accent"0C #1% \else\ooalign{\unhbox0\crcr\hidewidth\char"0C \hidewidth}% \fi \fi\fi\fi\fi }% } \def\Aogonek{{\ecfont \char"81}}\def\macrocharA{A} \def\aogonek{{\ecfont \char"A1}}\def\macrochara{a} \def\Eogonek{{\ecfont \char"86}}\def\macrocharE{E} \def\eogonek{{\ecfont \char"A6}}\def\macrochare{e} % % Use the ec* fonts (cm-super in outline format) for non-CM glyphs. \def\ecfont{% % We can't distinguish serif/sans and italic/slanted, but this % is used for crude hacks anyway (like adding French and German % quotes to documents typeset with CM, where we lose kerning), so % hopefully nobody will notice/care. \edef\ecsize{\csname\curfontsize ecsize\endcsname}% \edef\nominalsize{\csname\curfontsize nominalsize\endcsname}% \ifx\curfontstyle\bfstylename % bold: \font\thisecfont = ecb\ifusingit{i}{x}\ecsize \space at \nominalsize \else % regular: \font\thisecfont = ec\ifusingit{ti}{rm}\ecsize \space at \nominalsize \fi \thisecfont } % @registeredsymbol - R in a circle. The font for the R should really % be smaller yet, but lllsize is the best we can do for now. % Adapted from the plain.tex definition of \copyright. % \def\registeredsymbol{% $^{{\ooalign{\hfil\raise.07ex\hbox{\selectfonts\lllsize R}% \hfil\crcr\Orb}}% }$% } % @textdegree - the normal degrees sign. % \def\textdegree{$^\circ$} % Laurent Siebenmann reports \Orb undefined with: % Textures 1.7.7 (preloaded format=plain 93.10.14) (68K) 16 APR 2004 02:38 % so we'll define it if necessary. % \ifx\Orb\thisisundefined \def\Orb{\mathhexbox20D} \fi % Quotes. \chardef\quotedblleft="5C \chardef\quotedblright=`\" \chardef\quoteleft=`\` \chardef\quoteright=`\' \message{page headings,} \newskip\titlepagetopglue \titlepagetopglue = 1.5in \newskip\titlepagebottomglue \titlepagebottomglue = 2pc % First the title page. Must do @settitle before @titlepage. \newif\ifseenauthor \newif\iffinishedtitlepage % Do an implicit @contents or @shortcontents after @end titlepage if the % user says @setcontentsaftertitlepage or @setshortcontentsaftertitlepage. % \newif\ifsetcontentsaftertitlepage \let\setcontentsaftertitlepage = \setcontentsaftertitlepagetrue \newif\ifsetshortcontentsaftertitlepage \let\setshortcontentsaftertitlepage = \setshortcontentsaftertitlepagetrue \parseargdef\shorttitlepage{% \begingroup \hbox{}\vskip 1.5in \chaprm \centerline{#1}% \endgroup\page\hbox{}\page} \envdef\titlepage{% % Open one extra group, as we want to close it in the middle of \Etitlepage. \begingroup \parindent=0pt \textfonts % Leave some space at the very top of the page. \vglue\titlepagetopglue % No rule at page bottom unless we print one at the top with @title. \finishedtitlepagetrue % % Most title ``pages'' are actually two pages long, with space % at the top of the second. We don't want the ragged left on the second. \let\oldpage = \page \def\page{% \iffinishedtitlepage\else \finishtitlepage \fi \let\page = \oldpage \page \null }% } \def\Etitlepage{% \iffinishedtitlepage\else \finishtitlepage \fi % It is important to do the page break before ending the group, % because the headline and footline are only empty inside the group. % If we use the new definition of \page, we always get a blank page % after the title page, which we certainly don't want. \oldpage \endgroup % % Need this before the \...aftertitlepage checks so that if they are % in effect the toc pages will come out with page numbers. \HEADINGSon % % If they want short, they certainly want long too. \ifsetshortcontentsaftertitlepage \shortcontents \contents \global\let\shortcontents = \relax \global\let\contents = \relax \fi % \ifsetcontentsaftertitlepage \contents \global\let\contents = \relax \global\let\shortcontents = \relax \fi } \def\finishtitlepage{% \vskip4pt \hrule height 2pt width \hsize \vskip\titlepagebottomglue \finishedtitlepagetrue } % Macros to be used within @titlepage: \let\subtitlerm=\tenrm \def\subtitlefont{\subtitlerm \normalbaselineskip = 13pt \normalbaselines} \parseargdef\title{% \checkenv\titlepage \leftline{\titlefonts\rmisbold #1} % print a rule at the page bottom also. \finishedtitlepagefalse \vskip4pt \hrule height 4pt width \hsize \vskip4pt } \parseargdef\subtitle{% \checkenv\titlepage {\subtitlefont \rightline{#1}}% } % @author should come last, but may come many times. % It can also be used inside @quotation. % \parseargdef\author{% \def\temp{\quotation}% \ifx\thisenv\temp \def\quotationauthor{#1}% printed in \Equotation. \else \checkenv\titlepage \ifseenauthor\else \vskip 0pt plus 1filll \seenauthortrue \fi {\secfonts\rmisbold \leftline{#1}}% \fi } % Set up page headings and footings. \let\thispage=\folio \newtoks\evenheadline % headline on even pages \newtoks\oddheadline % headline on odd pages \newtoks\evenfootline % footline on even pages \newtoks\oddfootline % footline on odd pages % Now make TeX use those variables \headline={{\textfonts\rm \ifodd\pageno \the\oddheadline \else \the\evenheadline \fi}} \footline={{\textfonts\rm \ifodd\pageno \the\oddfootline \else \the\evenfootline \fi}\HEADINGShook} \let\HEADINGShook=\relax % Commands to set those variables. % For example, this is what @headings on does % @evenheading @thistitle|@thispage|@thischapter % @oddheading @thischapter|@thispage|@thistitle % @evenfooting @thisfile|| % @oddfooting ||@thisfile \def\evenheading{\parsearg\evenheadingxxx} \def\evenheadingxxx #1{\evenheadingyyy #1\|\|\|\|\finish} \def\evenheadingyyy #1\|#2\|#3\|#4\finish{% \global\evenheadline={\rlap{\centerline{#2}}\line{#1\hfil#3}}} \def\oddheading{\parsearg\oddheadingxxx} \def\oddheadingxxx #1{\oddheadingyyy #1\|\|\|\|\finish} \def\oddheadingyyy #1\|#2\|#3\|#4\finish{% \global\oddheadline={\rlap{\centerline{#2}}\line{#1\hfil#3}}} \parseargdef\everyheading{\oddheadingxxx{#1}\evenheadingxxx{#1}}% \def\evenfooting{\parsearg\evenfootingxxx} \def\evenfootingxxx #1{\evenfootingyyy #1\|\|\|\|\finish} \def\evenfootingyyy #1\|#2\|#3\|#4\finish{% \global\evenfootline={\rlap{\centerline{#2}}\line{#1\hfil#3}}} \def\oddfooting{\parsearg\oddfootingxxx} \def\oddfootingxxx #1{\oddfootingyyy #1\|\|\|\|\finish} \def\oddfootingyyy #1\|#2\|#3\|#4\finish{% \global\oddfootline = {\rlap{\centerline{#2}}\line{#1\hfil#3}}% % % Leave some space for the footline. Hopefully ok to assume % @evenfooting will not be used by itself. \global\advance\pageheight by -12pt \global\advance\vsize by -12pt } \parseargdef\everyfooting{\oddfootingxxx{#1}\evenfootingxxx{#1}} % @evenheadingmarks top \thischapter <- chapter at the top of a page % @evenheadingmarks bottom \thischapter <- chapter at the bottom of a page % % The same set of arguments for: % % @oddheadingmarks % @evenfootingmarks % @oddfootingmarks % @everyheadingmarks % @everyfootingmarks \def\evenheadingmarks{\headingmarks{even}{heading}} \def\oddheadingmarks{\headingmarks{odd}{heading}} \def\evenfootingmarks{\headingmarks{even}{footing}} \def\oddfootingmarks{\headingmarks{odd}{footing}} \def\everyheadingmarks#1 {\headingmarks{even}{heading}{#1} \headingmarks{odd}{heading}{#1} } \def\everyfootingmarks#1 {\headingmarks{even}{footing}{#1} \headingmarks{odd}{footing}{#1} } % #1 = even/odd, #2 = heading/footing, #3 = top/bottom. \def\headingmarks#1#2#3 {% \expandafter\let\expandafter\temp \csname get#3headingmarks\endcsname \global\expandafter\let\csname get#1#2marks\endcsname \temp } \everyheadingmarks bottom \everyfootingmarks bottom % @headings double turns headings on for double-sided printing. % @headings single turns headings on for single-sided printing. % @headings off turns them off. % @headings on same as @headings double, retained for compatibility. % @headings after turns on double-sided headings after this page. % @headings doubleafter turns on double-sided headings after this page. % @headings singleafter turns on single-sided headings after this page. % By default, they are off at the start of a document, % and turned `on' after @end titlepage. \def\headings #1 {\csname HEADINGS#1\endcsname} \def\headingsoff{% non-global headings elimination \evenheadline={\hfil}\evenfootline={\hfil}% \oddheadline={\hfil}\oddfootline={\hfil}% } \def\HEADINGSoff{{\globaldefs=1 \headingsoff}} % global setting \HEADINGSoff % it's the default % When we turn headings on, set the page number to 1. % For double-sided printing, put current file name in lower left corner, % chapter name on inside top of right hand pages, document % title on inside top of left hand pages, and page numbers on outside top % edge of all pages. \def\HEADINGSdouble{% \global\pageno=1 \global\evenfootline={\hfil} \global\oddfootline={\hfil} \global\evenheadline={\line{\folio\hfil\thistitle}} \global\oddheadline={\line{\thischapter\hfil\folio}} \global\let\contentsalignmacro = \chapoddpage } \let\contentsalignmacro = \chappager % For single-sided printing, chapter title goes across top left of page, % page number on top right. \def\HEADINGSsingle{% \global\pageno=1 \global\evenfootline={\hfil} \global\oddfootline={\hfil} \global\evenheadline={\line{\thischapter\hfil\folio}} \global\oddheadline={\line{\thischapter\hfil\folio}} \global\let\contentsalignmacro = \chappager } \def\HEADINGSon{\HEADINGSdouble} \def\HEADINGSafter{\let\HEADINGShook=\HEADINGSdoublex} \let\HEADINGSdoubleafter=\HEADINGSafter \def\HEADINGSdoublex{% \global\evenfootline={\hfil} \global\oddfootline={\hfil} \global\evenheadline={\line{\folio\hfil\thistitle}} \global\oddheadline={\line{\thischapter\hfil\folio}} \global\let\contentsalignmacro = \chapoddpage } \def\HEADINGSsingleafter{\let\HEADINGShook=\HEADINGSsinglex} \def\HEADINGSsinglex{% \global\evenfootline={\hfil} \global\oddfootline={\hfil} \global\evenheadline={\line{\thischapter\hfil\folio}} \global\oddheadline={\line{\thischapter\hfil\folio}} \global\let\contentsalignmacro = \chappager } % Subroutines used in generating headings % This produces Day Month Year style of output. % Only define if not already defined, in case a txi-??.tex file has set % up a different format (e.g., txi-cs.tex does this). \ifx\today\thisisundefined \def\today{% \number\day\space \ifcase\month \or\putwordMJan\or\putwordMFeb\or\putwordMMar\or\putwordMApr \or\putwordMMay\or\putwordMJun\or\putwordMJul\or\putwordMAug \or\putwordMSep\or\putwordMOct\or\putwordMNov\or\putwordMDec \fi \space\number\year} \fi % @settitle line... specifies the title of the document, for headings. % It generates no output of its own. \def\thistitle{\putwordNoTitle} \def\settitle{\parsearg{\gdef\thistitle}} \message{tables,} % Tables -- @table, @ftable, @vtable, @item(x). % default indentation of table text \newdimen\tableindent \tableindent=.8in % default indentation of @itemize and @enumerate text \newdimen\itemindent \itemindent=.3in % margin between end of table item and start of table text. \newdimen\itemmargin \itemmargin=.1in % used internally for \itemindent minus \itemmargin \newdimen\itemmax % Note @table, @ftable, and @vtable define @item, @itemx, etc., with % these defs. % They also define \itemindex % to index the item name in whatever manner is desired (perhaps none). \newif\ifitemxneedsnegativevskip \def\itemxpar{\par\ifitemxneedsnegativevskip\nobreak\vskip-\parskip\nobreak\fi} \def\internalBitem{\smallbreak \parsearg\itemzzz} \def\internalBitemx{\itemxpar \parsearg\itemzzz} \def\itemzzz #1{\begingroup % \advance\hsize by -\rightskip \advance\hsize by -\tableindent \setbox0=\hbox{\itemindicate{#1}}% \itemindex{#1}% \nobreak % This prevents a break before @itemx. % % If the item text does not fit in the space we have, put it on a line % by itself, and do not allow a page break either before or after that % line. We do not start a paragraph here because then if the next % command is, e.g., @kindex, the whatsit would get put into the % horizontal list on a line by itself, resulting in extra blank space. \ifdim \wd0>\itemmax % % Make this a paragraph so we get the \parskip glue and wrapping, % but leave it ragged-right. \begingroup \advance\leftskip by-\tableindent \advance\hsize by\tableindent \advance\rightskip by0pt plus1fil\relax \leavevmode\unhbox0\par \endgroup % % We're going to be starting a paragraph, but we don't want the % \parskip glue -- logically it's part of the @item we just started. \nobreak \vskip-\parskip % % Stop a page break at the \parskip glue coming up. However, if % what follows is an environment such as @example, there will be no % \parskip glue; then the negative vskip we just inserted would % cause the example and the item to crash together. So we use this % bizarre value of 10001 as a signal to \aboveenvbreak to insert % \parskip glue after all. Section titles are handled this way also. % \penalty 10001 \endgroup \itemxneedsnegativevskipfalse \else % The item text fits into the space. Start a paragraph, so that the % following text (if any) will end up on the same line. \noindent % Do this with kerns and \unhbox so that if there is a footnote in % the item text, it can migrate to the main vertical list and % eventually be printed. \nobreak\kern-\tableindent \dimen0 = \itemmax \advance\dimen0 by \itemmargin \advance\dimen0 by -\wd0 \unhbox0 \nobreak\kern\dimen0 \endgroup \itemxneedsnegativevskiptrue \fi } \def\item{\errmessage{@item while not in a list environment}} \def\itemx{\errmessage{@itemx while not in a list environment}} % @table, @ftable, @vtable. \envdef\table{% \let\itemindex\gobble \tablecheck{table}% } \envdef\ftable{% \def\itemindex ##1{\doind {fn}{\code{##1}}}% \tablecheck{ftable}% } \envdef\vtable{% \def\itemindex ##1{\doind {vr}{\code{##1}}}% \tablecheck{vtable}% } \def\tablecheck#1{% \ifnum \the\catcode`\^^M=\active \endgroup \errmessage{This command won't work in this context; perhaps the problem is that we are \inenvironment\thisenv}% \def\next{\doignore{#1}}% \else \let\next\tablex \fi \next } \def\tablex#1{% \def\itemindicate{#1}% \parsearg\tabley } \def\tabley#1{% {% \makevalueexpandable \edef\temp{\noexpand\tablez #1\space\space\space}% \expandafter }\temp \endtablez } \def\tablez #1 #2 #3 #4\endtablez{% \aboveenvbreak \ifnum 0#1>0 \advance \leftskip by #1\mil \fi \ifnum 0#2>0 \tableindent=#2\mil \fi \ifnum 0#3>0 \advance \rightskip by #3\mil \fi \itemmax=\tableindent \advance \itemmax by -\itemmargin \advance \leftskip by \tableindent \exdentamount=\tableindent \parindent = 0pt \parskip = \smallskipamount \ifdim \parskip=0pt \parskip=2pt \fi \let\item = \internalBitem \let\itemx = \internalBitemx } \def\Etable{\endgraf\afterenvbreak} \let\Eftable\Etable \let\Evtable\Etable \let\Eitemize\Etable \let\Eenumerate\Etable % This is the counter used by @enumerate, which is really @itemize \newcount \itemno \envdef\itemize{\parsearg\doitemize} \def\doitemize#1{% \aboveenvbreak \itemmax=\itemindent \advance\itemmax by -\itemmargin \advance\leftskip by \itemindent \exdentamount=\itemindent \parindent=0pt \parskip=\smallskipamount \ifdim\parskip=0pt \parskip=2pt \fi % % Try typesetting the item mark that if the document erroneously says % something like @itemize @samp (intending @table), there's an error % right away at the @itemize. It's not the best error message in the % world, but it's better than leaving it to the @item. This means if % the user wants an empty mark, they have to say @w{} not just @w. \def\itemcontents{#1}% \setbox0 = \hbox{\itemcontents}% % % @itemize with no arg is equivalent to @itemize @bullet. \ifx\itemcontents\empty\def\itemcontents{\bullet}\fi % \let\item=\itemizeitem } % Definition of @item while inside @itemize and @enumerate. % \def\itemizeitem{% \advance\itemno by 1 % for enumerations {\let\par=\endgraf \smallbreak}% reasonable place to break {% % If the document has an @itemize directly after a section title, a % \nobreak will be last on the list, and \sectionheading will have % done a \vskip-\parskip. In that case, we don't want to zero % parskip, or the item text will crash with the heading. On the % other hand, when there is normal text preceding the item (as there % usually is), we do want to zero parskip, or there would be too much % space. In that case, we won't have a \nobreak before. At least % that's the theory. \ifnum\lastpenalty<10000 \parskip=0in \fi \noindent \hbox to 0pt{\hss \itemcontents \kern\itemmargin}% % \vadjust{\penalty 1200}}% not good to break after first line of item. \flushcr } % \splitoff TOKENS\endmark defines \first to be the first token in % TOKENS, and \rest to be the remainder. % \def\splitoff#1#2\endmark{\def\first{#1}\def\rest{#2}}% % Allow an optional argument of an uppercase letter, lowercase letter, % or number, to specify the first label in the enumerated list. No % argument is the same as `1'. % \envparseargdef\enumerate{\enumeratey #1 \endenumeratey} \def\enumeratey #1 #2\endenumeratey{% % If we were given no argument, pretend we were given `1'. \def\thearg{#1}% \ifx\thearg\empty \def\thearg{1}\fi % % Detect if the argument is a single token. If so, it might be a % letter. Otherwise, the only valid thing it can be is a number. % (We will always have one token, because of the test we just made. % This is a good thing, since \splitoff doesn't work given nothing at % all -- the first parameter is undelimited.) \expandafter\splitoff\thearg\endmark \ifx\rest\empty % Only one token in the argument. It could still be anything. % A ``lowercase letter'' is one whose \lccode is nonzero. % An ``uppercase letter'' is one whose \lccode is both nonzero, and % not equal to itself. % Otherwise, we assume it's a number. % % We need the \relax at the end of the \ifnum lines to stop TeX from % continuing to look for a . % \ifnum\lccode\expandafter`\thearg=0\relax \numericenumerate % a number (we hope) \else % It's a letter. \ifnum\lccode\expandafter`\thearg=\expandafter`\thearg\relax \lowercaseenumerate % lowercase letter \else \uppercaseenumerate % uppercase letter \fi \fi \else % Multiple tokens in the argument. We hope it's a number. \numericenumerate \fi } % An @enumerate whose labels are integers. The starting integer is % given in \thearg. % \def\numericenumerate{% \itemno = \thearg \startenumeration{\the\itemno}% } % The starting (lowercase) letter is in \thearg. \def\lowercaseenumerate{% \itemno = \expandafter`\thearg \startenumeration{% % Be sure we're not beyond the end of the alphabet. \ifnum\itemno=0 \errmessage{No more lowercase letters in @enumerate; get a bigger alphabet}% \fi \char\lccode\itemno }% } % The starting (uppercase) letter is in \thearg. \def\uppercaseenumerate{% \itemno = \expandafter`\thearg \startenumeration{% % Be sure we're not beyond the end of the alphabet. \ifnum\itemno=0 \errmessage{No more uppercase letters in @enumerate; get a bigger alphabet} \fi \char\uccode\itemno }% } % Call \doitemize, adding a period to the first argument and supplying the % common last two arguments. Also subtract one from the initial value in % \itemno, since @item increments \itemno. % \def\startenumeration#1{% \advance\itemno by -1 \doitemize{#1.}\flushcr } % @alphaenumerate and @capsenumerate are abbreviations for giving an arg % to @enumerate. % \def\alphaenumerate{\enumerate{a}} \def\capsenumerate{\enumerate{A}} \def\Ealphaenumerate{\Eenumerate} \def\Ecapsenumerate{\Eenumerate} % @multitable macros % Amy Hendrickson, 8/18/94, 3/6/96 % % @multitable ... @end multitable will make as many columns as desired. % Contents of each column will wrap at width given in preamble. Width % can be specified either with sample text given in a template line, % or in percent of \hsize, the current width of text on page. % Table can continue over pages but will only break between lines. % To make preamble: % % Either define widths of columns in terms of percent of \hsize: % @multitable @columnfractions .25 .3 .45 % @item ... % % Numbers following @columnfractions are the percent of the total % current hsize to be used for each column. You may use as many % columns as desired. % Or use a template: % @multitable {Column 1 template} {Column 2 template} {Column 3 template} % @item ... % using the widest term desired in each column. % Each new table line starts with @item, each subsequent new column % starts with @tab. Empty columns may be produced by supplying @tab's % with nothing between them for as many times as empty columns are needed, % ie, @tab@tab@tab will produce two empty columns. % @item, @tab do not need to be on their own lines, but it will not hurt % if they are. % Sample multitable: % @multitable {Column 1 template} {Column 2 template} {Column 3 template} % @item first col stuff @tab second col stuff @tab third col % @item % first col stuff % @tab % second col stuff % @tab % third col % @item first col stuff @tab second col stuff % @tab Many paragraphs of text may be used in any column. % % They will wrap at the width determined by the template. % @item@tab@tab This will be in third column. % @end multitable % Default dimensions may be reset by user. % @multitableparskip is vertical space between paragraphs in table. % @multitableparindent is paragraph indent in table. % @multitablecolmargin is horizontal space to be left between columns. % @multitablelinespace is space to leave between table items, baseline % to baseline. % 0pt means it depends on current normal line spacing. % \newskip\multitableparskip \newskip\multitableparindent \newdimen\multitablecolspace \newskip\multitablelinespace \multitableparskip=0pt \multitableparindent=6pt \multitablecolspace=12pt \multitablelinespace=0pt % Macros used to set up halign preamble: % \let\endsetuptable\relax \def\xendsetuptable{\endsetuptable} \let\columnfractions\relax \def\xcolumnfractions{\columnfractions} \newif\ifsetpercent % #1 is the @columnfraction, usually a decimal number like .5, but might % be just 1. We just use it, whatever it is. % \def\pickupwholefraction#1 {% \global\advance\colcount by 1 \expandafter\xdef\csname col\the\colcount\endcsname{#1\hsize}% \setuptable } \newcount\colcount \def\setuptable#1{% \def\firstarg{#1}% \ifx\firstarg\xendsetuptable \let\go = \relax \else \ifx\firstarg\xcolumnfractions \global\setpercenttrue \else \ifsetpercent \let\go\pickupwholefraction \else \global\advance\colcount by 1 \setbox0=\hbox{#1\unskip\space}% Add a normal word space as a % separator; typically that is always in the input, anyway. \expandafter\xdef\csname col\the\colcount\endcsname{\the\wd0}% \fi \fi \ifx\go\pickupwholefraction % Put the argument back for the \pickupwholefraction call, so % we'll always have a period there to be parsed. \def\go{\pickupwholefraction#1}% \else \let\go = \setuptable \fi% \fi \go } % multitable-only commands. % % @headitem starts a heading row, which we typeset in bold. % Assignments have to be global since we are inside the implicit group % of an alignment entry. \everycr resets \everytab so we don't have to % undo it ourselves. \def\headitemfont{\b}% for people to use in the template row; not changeable \def\headitem{% \checkenv\multitable \crcr \global\everytab={\bf}% can't use \headitemfont since the parsing differs \the\everytab % for the first item }% % % A \tab used to include \hskip1sp. But then the space in a template % line is not enough. That is bad. So let's go back to just `&' until % we again encounter the problem the 1sp was intended to solve. % --karl, nathan@acm.org, 20apr99. \def\tab{\checkenv\multitable &\the\everytab}% % @multitable ... @end multitable definitions: % \newtoks\everytab % insert after every tab. % \envdef\multitable{% \vskip\parskip \startsavinginserts % % @item within a multitable starts a normal row. % We use \def instead of \let so that if one of the multitable entries % contains an @itemize, we don't choke on the \item (seen as \crcr aka % \endtemplate) expanding \doitemize. \def\item{\crcr}% % \tolerance=9500 \hbadness=9500 \setmultitablespacing \parskip=\multitableparskip \parindent=\multitableparindent \overfullrule=0pt \global\colcount=0 % \everycr = {% \noalign{% \global\everytab={}% \global\colcount=0 % Reset the column counter. % Check for saved footnotes, etc. \checkinserts % Keeps underfull box messages off when table breaks over pages. %\filbreak % Maybe so, but it also creates really weird page breaks when the % table breaks over pages. Wouldn't \vfil be better? Wait until the % problem manifests itself, so it can be fixed for real --karl. }% }% % \parsearg\domultitable } \def\domultitable#1{% % To parse everything between @multitable and @item: \setuptable#1 \endsetuptable % % This preamble sets up a generic column definition, which will % be used as many times as user calls for columns. % \vtop will set a single line and will also let text wrap and % continue for many paragraphs if desired. \halign\bgroup &% \global\advance\colcount by 1 \multistrut \vtop{% % Use the current \colcount to find the correct column width: \hsize=\expandafter\csname col\the\colcount\endcsname % % In order to keep entries from bumping into each other % we will add a \leftskip of \multitablecolspace to all columns after % the first one. % % If a template has been used, we will add \multitablecolspace % to the width of each template entry. % % If the user has set preamble in terms of percent of \hsize we will % use that dimension as the width of the column, and the \leftskip % will keep entries from bumping into each other. Table will start at % left margin and final column will justify at right margin. % % Make sure we don't inherit \rightskip from the outer environment. \rightskip=0pt \ifnum\colcount=1 % The first column will be indented with the surrounding text. \advance\hsize by\leftskip \else \ifsetpercent \else % If user has not set preamble in terms of percent of \hsize % we will advance \hsize by \multitablecolspace. \advance\hsize by \multitablecolspace \fi % In either case we will make \leftskip=\multitablecolspace: \leftskip=\multitablecolspace \fi % Ignoring space at the beginning and end avoids an occasional spurious % blank line, when TeX decides to break the line at the space before the % box from the multistrut, so the strut ends up on a line by itself. % For example: % @multitable @columnfractions .11 .89 % @item @code{#} % @tab Legal holiday which is valid in major parts of the whole country. % Is automatically provided with highlighting sequences respectively % marking characters. \noindent\ignorespaces##\unskip\multistrut }\cr } \def\Emultitable{% \crcr \egroup % end the \halign \global\setpercentfalse } \def\setmultitablespacing{% \def\multistrut{\strut}% just use the standard line spacing % % Compute \multitablelinespace (if not defined by user) for use in % \multitableparskip calculation. We used define \multistrut based on % this, but (ironically) that caused the spacing to be off. % See bug-texinfo report from Werner Lemberg, 31 Oct 2004 12:52:20 +0100. \ifdim\multitablelinespace=0pt \setbox0=\vbox{X}\global\multitablelinespace=\the\baselineskip \global\advance\multitablelinespace by-\ht0 \fi % Test to see if parskip is larger than space between lines of % table. If not, do nothing. % If so, set to same dimension as multitablelinespace. \ifdim\multitableparskip>\multitablelinespace \global\multitableparskip=\multitablelinespace \global\advance\multitableparskip-7pt % to keep parskip somewhat smaller % than skip between lines in the table. \fi% \ifdim\multitableparskip=0pt \global\multitableparskip=\multitablelinespace \global\advance\multitableparskip-7pt % to keep parskip somewhat smaller % than skip between lines in the table. \fi} \message{conditionals,} % @iftex, @ifnotdocbook, @ifnothtml, @ifnotinfo, @ifnotplaintext, % @ifnotxml always succeed. They currently do nothing; we don't % attempt to check whether the conditionals are properly nested. But we % have to remember that they are conditionals, so that @end doesn't % attempt to close an environment group. % \def\makecond#1{% \expandafter\let\csname #1\endcsname = \relax \expandafter\let\csname iscond.#1\endcsname = 1 } \makecond{iftex} \makecond{ifnotdocbook} \makecond{ifnothtml} \makecond{ifnotinfo} \makecond{ifnotplaintext} \makecond{ifnotxml} % Ignore @ignore, @ifhtml, @ifinfo, and the like. % \def\direntry{\doignore{direntry}} \def\documentdescription{\doignore{documentdescription}} \def\docbook{\doignore{docbook}} \def\html{\doignore{html}} \def\ifdocbook{\doignore{ifdocbook}} \def\ifhtml{\doignore{ifhtml}} \def\ifinfo{\doignore{ifinfo}} \def\ifnottex{\doignore{ifnottex}} \def\ifplaintext{\doignore{ifplaintext}} \def\ifxml{\doignore{ifxml}} \def\ignore{\doignore{ignore}} \def\menu{\doignore{menu}} \def\xml{\doignore{xml}} % Ignore text until a line `@end #1', keeping track of nested conditionals. % % A count to remember the depth of nesting. \newcount\doignorecount \def\doignore#1{\begingroup % Scan in ``verbatim'' mode: \obeylines \catcode`\@ = \other \catcode`\{ = \other \catcode`\} = \other % % Make sure that spaces turn into tokens that match what \doignoretext wants. \spaceisspace % % Count number of #1's that we've seen. \doignorecount = 0 % % Swallow text until we reach the matching `@end #1'. \dodoignore{#1}% } { \catcode`_=11 % We want to use \_STOP_ which cannot appear in texinfo source. \obeylines % % \gdef\dodoignore#1{% % #1 contains the command name as a string, e.g., `ifinfo'. % % Define a command to find the next `@end #1'. \long\def\doignoretext##1^^M@end #1{% \doignoretextyyy##1^^M@#1\_STOP_}% % % And this command to find another #1 command, at the beginning of a % line. (Otherwise, we would consider a line `@c @ifset', for % example, to count as an @ifset for nesting.) \long\def\doignoretextyyy##1^^M@#1##2\_STOP_{\doignoreyyy{##2}\_STOP_}% % % And now expand that command. \doignoretext ^^M% }% } \def\doignoreyyy#1{% \def\temp{#1}% \ifx\temp\empty % Nothing found. \let\next\doignoretextzzz \else % Found a nested condition, ... \advance\doignorecount by 1 \let\next\doignoretextyyy % ..., look for another. % If we're here, #1 ends with ^^M\ifinfo (for example). \fi \next #1% the token \_STOP_ is present just after this macro. } % We have to swallow the remaining "\_STOP_". % \def\doignoretextzzz#1{% \ifnum\doignorecount = 0 % We have just found the outermost @end. \let\next\enddoignore \else % Still inside a nested condition. \advance\doignorecount by -1 \let\next\doignoretext % Look for the next @end. \fi \next } % Finish off ignored text. { \obeylines% % Ignore anything after the last `@end #1'; this matters in verbatim % environments, where otherwise the newline after an ignored conditional % would result in a blank line in the output. \gdef\enddoignore#1^^M{\endgroup\ignorespaces}% } % @set VAR sets the variable VAR to an empty value. % @set VAR REST-OF-LINE sets VAR to the value REST-OF-LINE. % % Since we want to separate VAR from REST-OF-LINE (which might be % empty), we can't just use \parsearg; we have to insert a space of our % own to delimit the rest of the line, and then take it out again if we % didn't need it. % We rely on the fact that \parsearg sets \catcode`\ =10. % \parseargdef\set{\setyyy#1 \endsetyyy} \def\setyyy#1 #2\endsetyyy{% {% \makevalueexpandable \def\temp{#2}% \edef\next{\gdef\makecsname{SET#1}}% \ifx\temp\empty \next{}% \else \setzzz#2\endsetzzz \fi }% } % Remove the trailing space \setxxx inserted. \def\setzzz#1 \endsetzzz{\next{#1}} % @clear VAR clears (i.e., unsets) the variable VAR. % \parseargdef\clear{% {% \makevalueexpandable \global\expandafter\let\csname SET#1\endcsname=\relax }% } % @value{foo} gets the text saved in variable foo. \def\value{\begingroup\makevalueexpandable\valuexxx} \def\valuexxx#1{\expandablevalue{#1}\endgroup} { \catcode`\- = \active \catcode`\_ = \active % \gdef\makevalueexpandable{% \let\value = \expandablevalue % We don't want these characters active, ... \catcode`\-=\other \catcode`\_=\other % ..., but we might end up with active ones in the argument if % we're called from @code, as @code{@value{foo-bar_}}, though. % So \let them to their normal equivalents. \let-\realdash \let_\normalunderscore } } % We have this subroutine so that we can handle at least some @value's % properly in indexes (we call \makevalueexpandable in \indexdummies). % The command has to be fully expandable (if the variable is set), since % the result winds up in the index file. This means that if the % variable's value contains other Texinfo commands, it's almost certain % it will fail (although perhaps we could fix that with sufficient work % to do a one-level expansion on the result, instead of complete). % \def\expandablevalue#1{% \expandafter\ifx\csname SET#1\endcsname\relax {[No value for ``#1'']}% \message{Variable `#1', used in @value, is not set.}% \else \csname SET#1\endcsname \fi } % @ifset VAR ... @end ifset reads the `...' iff VAR has been defined % with @set. % % To get special treatment of `@end ifset,' call \makeond and the redefine. % \makecond{ifset} \def\ifset{\parsearg{\doifset{\let\next=\ifsetfail}}} \def\doifset#1#2{% {% \makevalueexpandable \let\next=\empty \expandafter\ifx\csname SET#2\endcsname\relax #1% If not set, redefine \next. \fi \expandafter }\next } \def\ifsetfail{\doignore{ifset}} % @ifclear VAR ... @end ifclear reads the `...' iff VAR has never been % defined with @set, or has been undefined with @clear. % % The `\else' inside the `\doifset' parameter is a trick to reuse the % above code: if the variable is not set, do nothing, if it is set, % then redefine \next to \ifclearfail. % \makecond{ifclear} \def\ifclear{\parsearg{\doifset{\else \let\next=\ifclearfail}}} \def\ifclearfail{\doignore{ifclear}} % @dircategory CATEGORY -- specify a category of the dir file % which this file should belong to. Ignore this in TeX. \let\dircategory=\comment % @defininfoenclose. \let\definfoenclose=\comment \message{indexing,} % Index generation facilities % Define \newwrite to be identical to plain tex's \newwrite % except not \outer, so it can be used within macros and \if's. \edef\newwrite{\makecsname{ptexnewwrite}} % \newindex {foo} defines an index named foo. % It automatically defines \fooindex such that % \fooindex ...rest of line... puts an entry in the index foo. % It also defines \fooindfile to be the number of the output channel for % the file that accumulates this index. The file's extension is foo. % The name of an index should be no more than 2 characters long % for the sake of vms. % \def\newindex#1{% \iflinks \expandafter\newwrite \csname#1indfile\endcsname \openout \csname#1indfile\endcsname \jobname.#1 % Open the file \fi \expandafter\xdef\csname#1index\endcsname{% % Define @#1index \noexpand\doindex{#1}} } % @defindex foo == \newindex{foo} % \def\defindex{\parsearg\newindex} % Define @defcodeindex, like @defindex except put all entries in @code. % \def\defcodeindex{\parsearg\newcodeindex} % \def\newcodeindex#1{% \iflinks \expandafter\newwrite \csname#1indfile\endcsname \openout \csname#1indfile\endcsname \jobname.#1 \fi \expandafter\xdef\csname#1index\endcsname{% \noexpand\docodeindex{#1}}% } % @synindex foo bar makes index foo feed into index bar. % Do this instead of @defindex foo if you don't want it as a separate index. % % @syncodeindex foo bar similar, but put all entries made for index foo % inside @code. % \def\synindex#1 #2 {\dosynindex\doindex{#1}{#2}} \def\syncodeindex#1 #2 {\dosynindex\docodeindex{#1}{#2}} % #1 is \doindex or \docodeindex, #2 the index getting redefined (foo), % #3 the target index (bar). \def\dosynindex#1#2#3{% % Only do \closeout if we haven't already done it, else we'll end up % closing the target index. \expandafter \ifx\csname donesynindex#2\endcsname \relax % The \closeout helps reduce unnecessary open files; the limit on the % Acorn RISC OS is a mere 16 files. \expandafter\closeout\csname#2indfile\endcsname \expandafter\let\csname donesynindex#2\endcsname = 1 \fi % redefine \fooindfile: \expandafter\let\expandafter\temp\expandafter=\csname#3indfile\endcsname \expandafter\let\csname#2indfile\endcsname=\temp % redefine \fooindex: \expandafter\xdef\csname#2index\endcsname{\noexpand#1{#3}}% } % Define \doindex, the driver for all \fooindex macros. % Argument #1 is generated by the calling \fooindex macro, % and it is "foo", the name of the index. % \doindex just uses \parsearg; it calls \doind for the actual work. % This is because \doind is more useful to call from other macros. % There is also \dosubind {index}{topic}{subtopic} % which makes an entry in a two-level index such as the operation index. \def\doindex#1{\edef\indexname{#1}\parsearg\singleindexer} \def\singleindexer #1{\doind{\indexname}{#1}} % like the previous two, but they put @code around the argument. \def\docodeindex#1{\edef\indexname{#1}\parsearg\singlecodeindexer} \def\singlecodeindexer #1{\doind{\indexname}{\code{#1}}} % Take care of Texinfo commands that can appear in an index entry. % Since there are some commands we want to expand, and others we don't, % we have to laboriously prevent expansion for those that we don't. % \def\indexdummies{% \escapechar = `\\ % use backslash in output files. \def\@{@}% change to @@ when we switch to @ as escape char in index files. \def\ {\realbackslash\space }% % % Need these unexpandable (because we define \tt as a dummy) % definitions when @{ or @} appear in index entry text. Also, more % complicated, when \tex is in effect and \{ is a \delimiter again. % We can't use \lbracecmd and \rbracecmd because texindex assumes % braces and backslashes are used only as delimiters. Perhaps we % should define @lbrace and @rbrace commands a la @comma. \def\{{{\tt\char123}}% \def\}{{\tt\char125}}% % % I don't entirely understand this, but when an index entry is % generated from a macro call, the \endinput which \scanmacro inserts % causes processing to be prematurely terminated. This is, % apparently, because \indexsorttmp is fully expanded, and \endinput % is an expandable command. The redefinition below makes \endinput % disappear altogether for that purpose -- although logging shows that % processing continues to some further point. On the other hand, it % seems \endinput does not hurt in the printed index arg, since that % is still getting written without apparent harm. % % Sample source (mac-idx3.tex, reported by Graham Percival to % help-texinfo, 22may06): % @macro funindex {WORD} % @findex xyz % @end macro % ... % @funindex commtest % % The above is not enough to reproduce the bug, but it gives the flavor. % % Sample whatsit resulting: % .@write3{\entry{xyz}{@folio }{@code {xyz@endinput }}} % % So: \let\endinput = \empty % % Do the redefinitions. \commondummies } % For the aux and toc files, @ is the escape character. So we want to % redefine everything using @ as the escape character (instead of % \realbackslash, still used for index files). When everything uses @, % this will be simpler. % \def\atdummies{% \def\@{@@}% \def\ {@ }% \let\{ = \lbraceatcmd \let\} = \rbraceatcmd % % Do the redefinitions. \commondummies \otherbackslash } % Called from \indexdummies and \atdummies. % \def\commondummies{% % % \definedummyword defines \#1 as \string\#1\space, thus effectively % preventing its expansion. This is used only for control words, % not control letters, because the \space would be incorrect for % control characters, but is needed to separate the control word % from whatever follows. % % For control letters, we have \definedummyletter, which omits the % space. % % These can be used both for control words that take an argument and % those that do not. If it is followed by {arg} in the input, then % that will dutifully get written to the index (or wherever). % \def\definedummyword ##1{\def##1{\string##1\space}}% \def\definedummyletter##1{\def##1{\string##1}}% \let\definedummyaccent\definedummyletter % \commondummiesnofonts % \definedummyletter\_% \definedummyletter\-% % % Non-English letters. \definedummyword\AA \definedummyword\AE \definedummyword\DH \definedummyword\L \definedummyword\O \definedummyword\OE \definedummyword\TH \definedummyword\aa \definedummyword\ae \definedummyword\dh \definedummyword\exclamdown \definedummyword\l \definedummyword\o \definedummyword\oe \definedummyword\ordf \definedummyword\ordm \definedummyword\questiondown \definedummyword\ss \definedummyword\th % % Although these internal commands shouldn't show up, sometimes they do. \definedummyword\bf \definedummyword\gtr \definedummyword\hat \definedummyword\less \definedummyword\sf \definedummyword\sl \definedummyword\tclose \definedummyword\tt % \definedummyword\LaTeX \definedummyword\TeX % % Assorted special characters. \definedummyword\arrow \definedummyword\bullet \definedummyword\comma \definedummyword\copyright \definedummyword\registeredsymbol \definedummyword\dots \definedummyword\enddots \definedummyword\entrybreak \definedummyword\equiv \definedummyword\error \definedummyword\euro \definedummyword\expansion \definedummyword\geq \definedummyword\guillemetleft \definedummyword\guillemetright \definedummyword\guilsinglleft \definedummyword\guilsinglright \definedummyword\leq \definedummyword\minus \definedummyword\ogonek \definedummyword\pounds \definedummyword\point \definedummyword\print \definedummyword\quotedblbase \definedummyword\quotedblleft \definedummyword\quotedblright \definedummyword\quoteleft \definedummyword\quoteright \definedummyword\quotesinglbase \definedummyword\result \definedummyword\textdegree % % We want to disable all macros so that they are not expanded by \write. \macrolist % \normalturnoffactive % % Handle some cases of @value -- where it does not contain any % (non-fully-expandable) commands. \makevalueexpandable } % \commondummiesnofonts: common to \commondummies and \indexnofonts. % \def\commondummiesnofonts{% % Control letters and accents. \definedummyletter\!% \definedummyaccent\"% \definedummyaccent\'% \definedummyletter\*% \definedummyaccent\,% \definedummyletter\.% \definedummyletter\/% \definedummyletter\:% \definedummyaccent\=% \definedummyletter\?% \definedummyaccent\^% \definedummyaccent\`% \definedummyaccent\~% \definedummyword\u \definedummyword\v \definedummyword\H \definedummyword\dotaccent \definedummyword\ogonek \definedummyword\ringaccent \definedummyword\tieaccent \definedummyword\ubaraccent \definedummyword\udotaccent \definedummyword\dotless % % Texinfo font commands. \definedummyword\b \definedummyword\i \definedummyword\r \definedummyword\sansserif \definedummyword\sc \definedummyword\slanted \definedummyword\t % % Commands that take arguments. \definedummyword\acronym \definedummyword\anchor \definedummyword\cite \definedummyword\code \definedummyword\command \definedummyword\dfn \definedummyword\dmn \definedummyword\email \definedummyword\emph \definedummyword\env \definedummyword\file \definedummyword\indicateurl \definedummyword\kbd \definedummyword\key \definedummyword\math \definedummyword\option \definedummyword\pxref \definedummyword\ref \definedummyword\samp \definedummyword\strong \definedummyword\tie \definedummyword\uref \definedummyword\url \definedummyword\var \definedummyword\verb \definedummyword\w \definedummyword\xref } % \indexnofonts is used when outputting the strings to sort the index % by, and when constructing control sequence names. It eliminates all % control sequences and just writes whatever the best ASCII sort string % would be for a given command (usually its argument). % \def\indexnofonts{% % Accent commands should become @asis. \def\definedummyaccent##1{\let##1\asis}% % We can just ignore other control letters. \def\definedummyletter##1{\let##1\empty}% % All control words become @asis by default; overrides below. \let\definedummyword\definedummyaccent % \commondummiesnofonts % % Don't no-op \tt, since it isn't a user-level command % and is used in the definitions of the active chars like <, >, |, etc. % Likewise with the other plain tex font commands. %\let\tt=\asis % \def\ { }% \def\@{@}% \def\_{\normalunderscore}% \def\-{}% @- shouldn't affect sorting % % Unfortunately, texindex is not prepared to handle braces in the % content at all. So for index sorting, we map @{ and @} to strings % starting with |, since that ASCII character is between ASCII { and }. \def\{{|a}% \def\}{|b}% % % Non-English letters. \def\AA{AA}% \def\AE{AE}% \def\DH{DZZ}% \def\L{L}% \def\OE{OE}% \def\O{O}% \def\TH{ZZZ}% \def\aa{aa}% \def\ae{ae}% \def\dh{dzz}% \def\exclamdown{!}% \def\l{l}% \def\oe{oe}% \def\ordf{a}% \def\ordm{o}% \def\o{o}% \def\questiondown{?}% \def\ss{ss}% \def\th{zzz}% % \def\LaTeX{LaTeX}% \def\TeX{TeX}% % % Assorted special characters. % (The following {} will end up in the sort string, but that's ok.) \def\arrow{->}% \def\bullet{bullet}% \def\comma{,}% \def\copyright{copyright}% \def\dots{...}% \def\enddots{...}% \def\equiv{==}% \def\error{error}% \def\euro{euro}% \def\expansion{==>}% \def\geq{>=}% \def\guillemetleft{<<}% \def\guillemetright{>>}% \def\guilsinglleft{<}% \def\guilsinglright{>}% \def\leq{<=}% \def\minus{-}% \def\point{.}% \def\pounds{pounds}% \def\print{-|}% \def\quotedblbase{"}% \def\quotedblleft{"}% \def\quotedblright{"}% \def\quoteleft{`}% \def\quoteright{'}% \def\quotesinglbase{,}% \def\registeredsymbol{R}% \def\result{=>}% \def\textdegree{o}% % \expandafter\ifx\csname SETtxiindexlquoteignore\endcsname\relax \else \indexlquoteignore \fi % % We need to get rid of all macros, leaving only the arguments (if present). % Of course this is not nearly correct, but it is the best we can do for now. % makeinfo does not expand macros in the argument to @deffn, which ends up % writing an index entry, and texindex isn't prepared for an index sort entry % that starts with \. % % Since macro invocations are followed by braces, we can just redefine them % to take a single TeX argument. The case of a macro invocation that % goes to end-of-line is not handled. % \macrolist } % Undocumented (for FSFS 2nd ed.): @set txiindexlquoteignore makes us % ignore left quotes in the sort term. {\catcode`\`=\active \gdef\indexlquoteignore{\let`=\empty}} \let\indexbackslash=0 %overridden during \printindex. \let\SETmarginindex=\relax % put index entries in margin (undocumented)? % Most index entries go through here, but \dosubind is the general case. % #1 is the index name, #2 is the entry text. \def\doind#1#2{\dosubind{#1}{#2}{}} % Workhorse for all \fooindexes. % #1 is name of index, #2 is stuff to put there, #3 is subentry -- % empty if called from \doind, as we usually are (the main exception % is with most defuns, which call us directly). % \def\dosubind#1#2#3{% \iflinks {% % Store the main index entry text (including the third arg). \toks0 = {#2}% % If third arg is present, precede it with a space. \def\thirdarg{#3}% \ifx\thirdarg\empty \else \toks0 = \expandafter{\the\toks0 \space #3}% \fi % \edef\writeto{\csname#1indfile\endcsname}% % \safewhatsit\dosubindwrite }% \fi } % Write the entry in \toks0 to the index file: % \def\dosubindwrite{% % Put the index entry in the margin if desired. \ifx\SETmarginindex\relax\else \insert\margin{\hbox{\vrule height8pt depth3pt width0pt \the\toks0}}% \fi % % Remember, we are within a group. \indexdummies % Must do this here, since \bf, etc expand at this stage \def\backslashcurfont{\indexbackslash}% \indexbackslash isn't defined now % so it will be output as is; and it will print as backslash. % % Process the index entry with all font commands turned off, to % get the string to sort by. {\indexnofonts \edef\temp{\the\toks0}% need full expansion \xdef\indexsorttmp{\temp}% }% % % Set up the complete index entry, with both the sort key and % the original text, including any font commands. We write % three arguments to \entry to the .?? file (four in the % subentry case), texindex reduces to two when writing the .??s % sorted result. \edef\temp{% \write\writeto{% \string\entry{\indexsorttmp}{\noexpand\folio}{\the\toks0}}% }% \temp } % Take care of unwanted page breaks/skips around a whatsit: % % If a skip is the last thing on the list now, preserve it % by backing up by \lastskip, doing the \write, then inserting % the skip again. Otherwise, the whatsit generated by the % \write or \pdfdest will make \lastskip zero. The result is that % sequences like this: % @end defun % @tindex whatever % @defun ... % will have extra space inserted, because the \medbreak in the % start of the @defun won't see the skip inserted by the @end of % the previous defun. % % But don't do any of this if we're not in vertical mode. We % don't want to do a \vskip and prematurely end a paragraph. % % Avoid page breaks due to these extra skips, too. % % But wait, there is a catch there: % We'll have to check whether \lastskip is zero skip. \ifdim is not % sufficient for this purpose, as it ignores stretch and shrink parts % of the skip. The only way seems to be to check the textual % representation of the skip. % % The following is almost like \def\zeroskipmacro{0.0pt} except that % the ``p'' and ``t'' characters have catcode \other, not 11 (letter). % \edef\zeroskipmacro{\expandafter\the\csname z@skip\endcsname} % \newskip\whatsitskip \newcount\whatsitpenalty % % ..., ready, GO: % \def\safewhatsit#1{\ifhmode #1% \else % \lastskip and \lastpenalty cannot both be nonzero simultaneously. \whatsitskip = \lastskip \edef\lastskipmacro{\the\lastskip}% \whatsitpenalty = \lastpenalty % % If \lastskip is nonzero, that means the last item was a % skip. And since a skip is discardable, that means this % -\whatsitskip glue we're inserting is preceded by a % non-discardable item, therefore it is not a potential % breakpoint, therefore no \nobreak needed. \ifx\lastskipmacro\zeroskipmacro \else \vskip-\whatsitskip \fi % #1% % \ifx\lastskipmacro\zeroskipmacro % If \lastskip was zero, perhaps the last item was a penalty, and % perhaps it was >=10000, e.g., a \nobreak. In that case, we want % to re-insert the same penalty (values >10000 are used for various % signals); since we just inserted a non-discardable item, any % following glue (such as a \parskip) would be a breakpoint. For example: % @deffn deffn-whatever % @vindex index-whatever % Description. % would allow a break between the index-whatever whatsit % and the "Description." paragraph. \ifnum\whatsitpenalty>9999 \penalty\whatsitpenalty \fi \else % On the other hand, if we had a nonzero \lastskip, % this make-up glue would be preceded by a non-discardable item % (the whatsit from the \write), so we must insert a \nobreak. \nobreak\vskip\whatsitskip \fi \fi} % The index entry written in the file actually looks like % \entry {sortstring}{page}{topic} % or % \entry {sortstring}{page}{topic}{subtopic} % The texindex program reads in these files and writes files % containing these kinds of lines: % \initial {c} % before the first topic whose initial is c % \entry {topic}{pagelist} % for a topic that is used without subtopics % \primary {topic} % for the beginning of a topic that is used with subtopics % \secondary {subtopic}{pagelist} % for each subtopic. % Define the user-accessible indexing commands % @findex, @vindex, @kindex, @cindex. \def\findex {\fnindex} \def\kindex {\kyindex} \def\cindex {\cpindex} \def\vindex {\vrindex} \def\tindex {\tpindex} \def\pindex {\pgindex} \def\cindexsub {\begingroup\obeylines\cindexsub} {\obeylines % \gdef\cindexsub "#1" #2^^M{\endgroup % \dosubind{cp}{#2}{#1}}} % Define the macros used in formatting output of the sorted index material. % @printindex causes a particular index (the ??s file) to get printed. % It does not print any chapter heading (usually an @unnumbered). % \parseargdef\printindex{\begingroup \dobreak \chapheadingskip{10000}% % \smallfonts \rm \tolerance = 9500 \plainfrenchspacing \everypar = {}% don't want the \kern\-parindent from indentation suppression. % % See if the index file exists and is nonempty. % Change catcode of @ here so that if the index file contains % \initial {@} % as its first line, TeX doesn't complain about mismatched braces % (because it thinks @} is a control sequence). \catcode`\@ = 11 \openin 1 \jobname.#1s \ifeof 1 % \enddoublecolumns gets confused if there is no text in the index, % and it loses the chapter title and the aux file entries for the % index. The easiest way to prevent this problem is to make sure % there is some text. \putwordIndexNonexistent \else % % If the index file exists but is empty, then \openin leaves \ifeof % false. We have to make TeX try to read something from the file, so % it can discover if there is anything in it. \read 1 to \temp \ifeof 1 \putwordIndexIsEmpty \else % Index files are almost Texinfo source, but we use \ as the escape % character. It would be better to use @, but that's too big a change % to make right now. \def\indexbackslash{\backslashcurfont}% \catcode`\\ = 0 \escapechar = `\\ \begindoublecolumns \input \jobname.#1s \enddoublecolumns \fi \fi \closein 1 \endgroup} % These macros are used by the sorted index file itself. % Change them to control the appearance of the index. \def\initial#1{{% % Some minor font changes for the special characters. \let\tentt=\sectt \let\tt=\sectt \let\sf=\sectt % % Remove any glue we may have, we'll be inserting our own. \removelastskip % % We like breaks before the index initials, so insert a bonus. \nobreak \vskip 0pt plus 3\baselineskip \penalty 0 \vskip 0pt plus -3\baselineskip % % Typeset the initial. Making this add up to a whole number of % baselineskips increases the chance of the dots lining up from column % to column. It still won't often be perfect, because of the stretch % we need before each entry, but it's better. % % No shrink because it confuses \balancecolumns. \vskip 1.67\baselineskip plus .5\baselineskip \leftline{\secbf #1}% % Do our best not to break after the initial. \nobreak \vskip .33\baselineskip plus .1\baselineskip }} % \entry typesets a paragraph consisting of the text (#1), dot leaders, and % then page number (#2) flushed to the right margin. It is used for index % and table of contents entries. The paragraph is indented by \leftskip. % % A straightforward implementation would start like this: % \def\entry#1#2{... % But this freezes the catcodes in the argument, and can cause problems to % @code, which sets - active. This problem was fixed by a kludge--- % ``-'' was active throughout whole index, but this isn't really right. % The right solution is to prevent \entry from swallowing the whole text. % --kasal, 21nov03 \def\entry{% \begingroup % % Start a new paragraph if necessary, so our assignments below can't % affect previous text. \par % % Do not fill out the last line with white space. \parfillskip = 0in % % No extra space above this paragraph. \parskip = 0in % % Do not prefer a separate line ending with a hyphen to fewer lines. \finalhyphendemerits = 0 % % \hangindent is only relevant when the entry text and page number % don't both fit on one line. In that case, bob suggests starting the % dots pretty far over on the line. Unfortunately, a large % indentation looks wrong when the entry text itself is broken across % lines. So we use a small indentation and put up with long leaders. % % \hangafter is reset to 1 (which is the value we want) at the start % of each paragraph, so we need not do anything with that. \hangindent = 2em % % When the entry text needs to be broken, just fill out the first line % with blank space. \rightskip = 0pt plus1fil % % A bit of stretch before each entry for the benefit of balancing % columns. \vskip 0pt plus1pt % % When reading the text of entry, convert explicit line breaks % from @* into spaces. The user might give these in long section % titles, for instance. \def\*{\unskip\space\ignorespaces}% \def\entrybreak{\hfil\break}% % % Swallow the left brace of the text (first parameter): \afterassignment\doentry \let\temp = } \def\entrybreak{\unskip\space\ignorespaces}% \def\doentry{% \bgroup % Instead of the swallowed brace. \noindent \aftergroup\finishentry % And now comes the text of the entry. } \def\finishentry#1{% % #1 is the page number. % % The following is kludged to not output a line of dots in the index if % there are no page numbers. The next person who breaks this will be % cursed by a Unix daemon. \setbox\boxA = \hbox{#1}% \ifdim\wd\boxA = 0pt \ % \else % % If we must, put the page number on a line of its own, and fill out % this line with blank space. (The \hfil is overwhelmed with the % fill leaders glue in \indexdotfill if the page number does fit.) \hfil\penalty50 \null\nobreak\indexdotfill % Have leaders before the page number. % % The `\ ' here is removed by the implicit \unskip that TeX does as % part of (the primitive) \par. Without it, a spurious underfull % \hbox ensues. \ifpdf \pdfgettoks#1.% \ \the\toksA \else \ #1% \fi \fi \par \endgroup } % Like plain.tex's \dotfill, except uses up at least 1 em. \def\indexdotfill{\cleaders \hbox{$\mathsurround=0pt \mkern1.5mu.\mkern1.5mu$}\hskip 1em plus 1fill} \def\primary #1{\line{#1\hfil}} \newskip\secondaryindent \secondaryindent=0.5cm \def\secondary#1#2{{% \parfillskip=0in \parskip=0in \hangindent=1in \hangafter=1 \noindent\hskip\secondaryindent\hbox{#1}\indexdotfill \ifpdf \pdfgettoks#2.\ \the\toksA % The page number ends the paragraph. \else #2 \fi \par }} % Define two-column mode, which we use to typeset indexes. % Adapted from the TeXbook, page 416, which is to say, % the manmac.tex format used to print the TeXbook itself. \catcode`\@=11 \newbox\partialpage \newdimen\doublecolumnhsize \def\begindoublecolumns{\begingroup % ended by \enddoublecolumns % Grab any single-column material above us. \output = {% % % Here is a possibility not foreseen in manmac: if we accumulate a % whole lot of material, we might end up calling this \output % routine twice in a row (see the doublecol-lose test, which is % essentially a couple of indexes with @setchapternewpage off). In % that case we just ship out what is in \partialpage with the normal % output routine. Generally, \partialpage will be empty when this % runs and this will be a no-op. See the indexspread.tex test case. \ifvoid\partialpage \else \onepageout{\pagecontents\partialpage}% \fi % \global\setbox\partialpage = \vbox{% % Unvbox the main output page. \unvbox\PAGE \kern-\topskip \kern\baselineskip }% }% \eject % run that output routine to set \partialpage % % Use the double-column output routine for subsequent pages. \output = {\doublecolumnout}% % % Change the page size parameters. We could do this once outside this % routine, in each of @smallbook, @afourpaper, and the default 8.5x11 % format, but then we repeat the same computation. Repeating a couple % of assignments once per index is clearly meaningless for the % execution time, so we may as well do it in one place. % % First we halve the line length, less a little for the gutter between % the columns. We compute the gutter based on the line length, so it % changes automatically with the paper format. The magic constant % below is chosen so that the gutter has the same value (well, +-<1pt) % as it did when we hard-coded it. % % We put the result in a separate register, \doublecolumhsize, so we % can restore it in \pagesofar, after \hsize itself has (potentially) % been clobbered. % \doublecolumnhsize = \hsize \advance\doublecolumnhsize by -.04154\hsize \divide\doublecolumnhsize by 2 \hsize = \doublecolumnhsize % % Double the \vsize as well. (We don't need a separate register here, % since nobody clobbers \vsize.) \vsize = 2\vsize } % The double-column output routine for all double-column pages except % the last. % \def\doublecolumnout{% \splittopskip=\topskip \splitmaxdepth=\maxdepth % Get the available space for the double columns -- the normal % (undoubled) page height minus any material left over from the % previous page. \dimen@ = \vsize \divide\dimen@ by 2 \advance\dimen@ by -\ht\partialpage % % box0 will be the left-hand column, box2 the right. \setbox0=\vsplit255 to\dimen@ \setbox2=\vsplit255 to\dimen@ \onepageout\pagesofar \unvbox255 \penalty\outputpenalty } % % Re-output the contents of the output page -- any previous material, % followed by the two boxes we just split, in box0 and box2. \def\pagesofar{% \unvbox\partialpage % \hsize = \doublecolumnhsize \wd0=\hsize \wd2=\hsize \hbox to\pagewidth{\box0\hfil\box2}% } % % All done with double columns. \def\enddoublecolumns{% % The following penalty ensures that the page builder is exercised % _before_ we change the output routine. This is necessary in the % following situation: % % The last section of the index consists only of a single entry. % Before this section, \pagetotal is less than \pagegoal, so no % break occurs before the last section starts. However, the last % section, consisting of \initial and the single \entry, does not % fit on the page and has to be broken off. Without the following % penalty the page builder will not be exercised until \eject % below, and by that time we'll already have changed the output % routine to the \balancecolumns version, so the next-to-last % double-column page will be processed with \balancecolumns, which % is wrong: The two columns will go to the main vertical list, with % the broken-off section in the recent contributions. As soon as % the output routine finishes, TeX starts reconsidering the page % break. The two columns and the broken-off section both fit on the % page, because the two columns now take up only half of the page % goal. When TeX sees \eject from below which follows the final % section, it invokes the new output routine that we've set after % \balancecolumns below; \onepageout will try to fit the two columns % and the final section into the vbox of \pageheight (see % \pagebody), causing an overfull box. % % Note that glue won't work here, because glue does not exercise the % page builder, unlike penalties (see The TeXbook, pp. 280-281). \penalty0 % \output = {% % Split the last of the double-column material. Leave it on the % current page, no automatic page break. \balancecolumns % % If we end up splitting too much material for the current page, % though, there will be another page break right after this \output % invocation ends. Having called \balancecolumns once, we do not % want to call it again. Therefore, reset \output to its normal % definition right away. (We hope \balancecolumns will never be % called on to balance too much material, but if it is, this makes % the output somewhat more palatable.) \global\output = {\onepageout{\pagecontents\PAGE}}% }% \eject \endgroup % started in \begindoublecolumns % % \pagegoal was set to the doubled \vsize above, since we restarted % the current page. We're now back to normal single-column % typesetting, so reset \pagegoal to the normal \vsize (after the % \endgroup where \vsize got restored). \pagegoal = \vsize } % % Called at the end of the double column material. \def\balancecolumns{% \setbox0 = \vbox{\unvbox255}% like \box255 but more efficient, see p.120. \dimen@ = \ht0 \advance\dimen@ by \topskip \advance\dimen@ by-\baselineskip \divide\dimen@ by 2 % target to split to %debug\message{final 2-column material height=\the\ht0, target=\the\dimen@.}% \splittopskip = \topskip % Loop until we get a decent breakpoint. {% \vbadness = 10000 \loop \global\setbox3 = \copy0 \global\setbox1 = \vsplit3 to \dimen@ \ifdim\ht3>\dimen@ \global\advance\dimen@ by 1pt \repeat }% %debug\message{split to \the\dimen@, column heights: \the\ht1, \the\ht3.}% \setbox0=\vbox to\dimen@{\unvbox1}% \setbox2=\vbox to\dimen@{\unvbox3}% % \pagesofar } \catcode`\@ = \other \message{sectioning,} % Chapters, sections, etc. % Let's start with @part. \outer\parseargdef\part{\partzzz{#1}} \def\partzzz#1{% \chapoddpage \null \vskip.3\vsize % move it down on the page a bit \begingroup \noindent \titlefonts\rmisbold #1\par % the text \let\lastnode=\empty % no node to associate with \writetocentry{part}{#1}{}% but put it in the toc \headingsoff % no headline or footline on the part page \chapoddpage \endgroup } % \unnumberedno is an oxymoron. But we count the unnumbered % sections so that we can refer to them unambiguously in the pdf % outlines by their "section number". We avoid collisions with chapter % numbers by starting them at 10000. (If a document ever has 10000 % chapters, we're in trouble anyway, I'm sure.) \newcount\unnumberedno \unnumberedno = 10000 \newcount\chapno \newcount\secno \secno=0 \newcount\subsecno \subsecno=0 \newcount\subsubsecno \subsubsecno=0 % This counter is funny since it counts through charcodes of letters A, B, ... \newcount\appendixno \appendixno = `\@ % % \def\appendixletter{\char\the\appendixno} % We do the following ugly conditional instead of the above simple % construct for the sake of pdftex, which needs the actual % letter in the expansion, not just typeset. % \def\appendixletter{% \ifnum\appendixno=`A A% \else\ifnum\appendixno=`B B% \else\ifnum\appendixno=`C C% \else\ifnum\appendixno=`D D% \else\ifnum\appendixno=`E E% \else\ifnum\appendixno=`F F% \else\ifnum\appendixno=`G G% \else\ifnum\appendixno=`H H% \else\ifnum\appendixno=`I I% \else\ifnum\appendixno=`J J% \else\ifnum\appendixno=`K K% \else\ifnum\appendixno=`L L% \else\ifnum\appendixno=`M M% \else\ifnum\appendixno=`N N% \else\ifnum\appendixno=`O O% \else\ifnum\appendixno=`P P% \else\ifnum\appendixno=`Q Q% \else\ifnum\appendixno=`R R% \else\ifnum\appendixno=`S S% \else\ifnum\appendixno=`T T% \else\ifnum\appendixno=`U U% \else\ifnum\appendixno=`V V% \else\ifnum\appendixno=`W W% \else\ifnum\appendixno=`X X% \else\ifnum\appendixno=`Y Y% \else\ifnum\appendixno=`Z Z% % The \the is necessary, despite appearances, because \appendixletter is % expanded while writing the .toc file. \char\appendixno is not % expandable, thus it is written literally, thus all appendixes come out % with the same letter (or @) in the toc without it. \else\char\the\appendixno \fi\fi\fi\fi\fi\fi\fi\fi\fi\fi\fi\fi\fi \fi\fi\fi\fi\fi\fi\fi\fi\fi\fi\fi\fi\fi} % Each @chapter defines these (using marks) as the number+name, number % and name of the chapter. Page headings and footings can use % these. @section does likewise. \def\thischapter{} \def\thischapternum{} \def\thischaptername{} \def\thissection{} \def\thissectionnum{} \def\thissectionname{} \newcount\absseclevel % used to calculate proper heading level \newcount\secbase\secbase=0 % @raisesections/@lowersections modify this count % @raisesections: treat @section as chapter, @subsection as section, etc. \def\raisesections{\global\advance\secbase by -1} \let\up=\raisesections % original BFox name % @lowersections: treat @chapter as section, @section as subsection, etc. \def\lowersections{\global\advance\secbase by 1} \let\down=\lowersections % original BFox name % we only have subsub. \chardef\maxseclevel = 3 % % A numbered section within an unnumbered changes to unnumbered too. % To achieve this, remember the "biggest" unnum. sec. we are currently in: \chardef\unnlevel = \maxseclevel % % Trace whether the current chapter is an appendix or not: % \chapheadtype is "N" or "A", unnumbered chapters are ignored. \def\chapheadtype{N} % Choose a heading macro % #1 is heading type % #2 is heading level % #3 is text for heading \def\genhead#1#2#3{% % Compute the abs. sec. level: \absseclevel=#2 \advance\absseclevel by \secbase % Make sure \absseclevel doesn't fall outside the range: \ifnum \absseclevel < 0 \absseclevel = 0 \else \ifnum \absseclevel > 3 \absseclevel = 3 \fi \fi % The heading type: \def\headtype{#1}% \if \headtype U% \ifnum \absseclevel < \unnlevel \chardef\unnlevel = \absseclevel \fi \else % Check for appendix sections: \ifnum \absseclevel = 0 \edef\chapheadtype{\headtype}% \else \if \headtype A\if \chapheadtype N% \errmessage{@appendix... within a non-appendix chapter}% \fi\fi \fi % Check for numbered within unnumbered: \ifnum \absseclevel > \unnlevel \def\headtype{U}% \else \chardef\unnlevel = 3 \fi \fi % Now print the heading: \if \headtype U% \ifcase\absseclevel \unnumberedzzz{#3}% \or \unnumberedseczzz{#3}% \or \unnumberedsubseczzz{#3}% \or \unnumberedsubsubseczzz{#3}% \fi \else \if \headtype A% \ifcase\absseclevel \appendixzzz{#3}% \or \appendixsectionzzz{#3}% \or \appendixsubseczzz{#3}% \or \appendixsubsubseczzz{#3}% \fi \else \ifcase\absseclevel \chapterzzz{#3}% \or \seczzz{#3}% \or \numberedsubseczzz{#3}% \or \numberedsubsubseczzz{#3}% \fi \fi \fi \suppressfirstparagraphindent } % an interface: \def\numhead{\genhead N} \def\apphead{\genhead A} \def\unnmhead{\genhead U} % @chapter, @appendix, @unnumbered. Increment top-level counter, reset % all lower-level sectioning counters to zero. % % Also set \chaplevelprefix, which we prepend to @float sequence numbers % (e.g., figures), q.v. By default (before any chapter), that is empty. \let\chaplevelprefix = \empty % \outer\parseargdef\chapter{\numhead0{#1}} % normally numhead0 calls chapterzzz \def\chapterzzz#1{% % section resetting is \global in case the chapter is in a group, such % as an @include file. \global\secno=0 \global\subsecno=0 \global\subsubsecno=0 \global\advance\chapno by 1 % % Used for \float. \gdef\chaplevelprefix{\the\chapno.}% \resetallfloatnos % % \putwordChapter can contain complex things in translations. \toks0=\expandafter{\putwordChapter}% \message{\the\toks0 \space \the\chapno}% % % Write the actual heading. \chapmacro{#1}{Ynumbered}{\the\chapno}% % % So @section and the like are numbered underneath this chapter. \global\let\section = \numberedsec \global\let\subsection = \numberedsubsec \global\let\subsubsection = \numberedsubsubsec } \outer\parseargdef\appendix{\apphead0{#1}} % normally calls appendixzzz % \def\appendixzzz#1{% \global\secno=0 \global\subsecno=0 \global\subsubsecno=0 \global\advance\appendixno by 1 \gdef\chaplevelprefix{\appendixletter.}% \resetallfloatnos % % \putwordAppendix can contain complex things in translations. \toks0=\expandafter{\putwordAppendix}% \message{\the\toks0 \space \appendixletter}% % \chapmacro{#1}{Yappendix}{\appendixletter}% % \global\let\section = \appendixsec \global\let\subsection = \appendixsubsec \global\let\subsubsection = \appendixsubsubsec } % normally unnmhead0 calls unnumberedzzz: \outer\parseargdef\unnumbered{\unnmhead0{#1}} \def\unnumberedzzz#1{% \global\secno=0 \global\subsecno=0 \global\subsubsecno=0 \global\advance\unnumberedno by 1 % % Since an unnumbered has no number, no prefix for figures. \global\let\chaplevelprefix = \empty \resetallfloatnos % % This used to be simply \message{#1}, but TeX fully expands the % argument to \message. Therefore, if #1 contained @-commands, TeX % expanded them. For example, in `@unnumbered The @cite{Book}', TeX % expanded @cite (which turns out to cause errors because \cite is meant % to be executed, not expanded). % % Anyway, we don't want the fully-expanded definition of @cite to appear % as a result of the \message, we just want `@cite' itself. We use % \the to achieve this: TeX expands \the only once, % simply yielding the contents of . (We also do this for % the toc entries.) \toks0 = {#1}% \message{(\the\toks0)}% % \chapmacro{#1}{Ynothing}{\the\unnumberedno}% % \global\let\section = \unnumberedsec \global\let\subsection = \unnumberedsubsec \global\let\subsubsection = \unnumberedsubsubsec } % @centerchap is like @unnumbered, but the heading is centered. \outer\parseargdef\centerchap{% % Well, we could do the following in a group, but that would break % an assumption that \chapmacro is called at the outermost level. % Thus we are safer this way: --kasal, 24feb04 \let\centerparametersmaybe = \centerparameters \unnmhead0{#1}% \let\centerparametersmaybe = \relax } % @top is like @unnumbered. \let\top\unnumbered % Sections. % \outer\parseargdef\numberedsec{\numhead1{#1}} % normally calls seczzz \def\seczzz#1{% \global\subsecno=0 \global\subsubsecno=0 \global\advance\secno by 1 \sectionheading{#1}{sec}{Ynumbered}{\the\chapno.\the\secno}% } % normally calls appendixsectionzzz: \outer\parseargdef\appendixsection{\apphead1{#1}} \def\appendixsectionzzz#1{% \global\subsecno=0 \global\subsubsecno=0 \global\advance\secno by 1 \sectionheading{#1}{sec}{Yappendix}{\appendixletter.\the\secno}% } \let\appendixsec\appendixsection % normally calls unnumberedseczzz: \outer\parseargdef\unnumberedsec{\unnmhead1{#1}} \def\unnumberedseczzz#1{% \global\subsecno=0 \global\subsubsecno=0 \global\advance\secno by 1 \sectionheading{#1}{sec}{Ynothing}{\the\unnumberedno.\the\secno}% } % Subsections. % % normally calls numberedsubseczzz: \outer\parseargdef\numberedsubsec{\numhead2{#1}} \def\numberedsubseczzz#1{% \global\subsubsecno=0 \global\advance\subsecno by 1 \sectionheading{#1}{subsec}{Ynumbered}{\the\chapno.\the\secno.\the\subsecno}% } % normally calls appendixsubseczzz: \outer\parseargdef\appendixsubsec{\apphead2{#1}} \def\appendixsubseczzz#1{% \global\subsubsecno=0 \global\advance\subsecno by 1 \sectionheading{#1}{subsec}{Yappendix}% {\appendixletter.\the\secno.\the\subsecno}% } % normally calls unnumberedsubseczzz: \outer\parseargdef\unnumberedsubsec{\unnmhead2{#1}} \def\unnumberedsubseczzz#1{% \global\subsubsecno=0 \global\advance\subsecno by 1 \sectionheading{#1}{subsec}{Ynothing}% {\the\unnumberedno.\the\secno.\the\subsecno}% } % Subsubsections. % % normally numberedsubsubseczzz: \outer\parseargdef\numberedsubsubsec{\numhead3{#1}} \def\numberedsubsubseczzz#1{% \global\advance\subsubsecno by 1 \sectionheading{#1}{subsubsec}{Ynumbered}% {\the\chapno.\the\secno.\the\subsecno.\the\subsubsecno}% } % normally appendixsubsubseczzz: \outer\parseargdef\appendixsubsubsec{\apphead3{#1}} \def\appendixsubsubseczzz#1{% \global\advance\subsubsecno by 1 \sectionheading{#1}{subsubsec}{Yappendix}% {\appendixletter.\the\secno.\the\subsecno.\the\subsubsecno}% } % normally unnumberedsubsubseczzz: \outer\parseargdef\unnumberedsubsubsec{\unnmhead3{#1}} \def\unnumberedsubsubseczzz#1{% \global\advance\subsubsecno by 1 \sectionheading{#1}{subsubsec}{Ynothing}% {\the\unnumberedno.\the\secno.\the\subsecno.\the\subsubsecno}% } % These macros control what the section commands do, according % to what kind of chapter we are in (ordinary, appendix, or unnumbered). % Define them by default for a numbered chapter. \let\section = \numberedsec \let\subsection = \numberedsubsec \let\subsubsection = \numberedsubsubsec % Define @majorheading, @heading and @subheading % NOTE on use of \vbox for chapter headings, section headings, and such: % 1) We use \vbox rather than the earlier \line to permit % overlong headings to fold. % 2) \hyphenpenalty is set to 10000 because hyphenation in a % heading is obnoxious; this forbids it. % 3) Likewise, headings look best if no \parindent is used, and % if justification is not attempted. Hence \raggedright. \def\majorheading{% {\advance\chapheadingskip by 10pt \chapbreak }% \parsearg\chapheadingzzz } \def\chapheading{\chapbreak \parsearg\chapheadingzzz} \def\chapheadingzzz#1{% {\chapfonts \vbox{\hyphenpenalty=10000\tolerance=5000 \parindent=0pt\ptexraggedright \rmisbold #1\hfill}}% \bigskip \par\penalty 200\relax \suppressfirstparagraphindent } % @heading, @subheading, @subsubheading. \parseargdef\heading{\sectionheading{#1}{sec}{Yomitfromtoc}{} \suppressfirstparagraphindent} \parseargdef\subheading{\sectionheading{#1}{subsec}{Yomitfromtoc}{} \suppressfirstparagraphindent} \parseargdef\subsubheading{\sectionheading{#1}{subsubsec}{Yomitfromtoc}{} \suppressfirstparagraphindent} % These macros generate a chapter, section, etc. heading only % (including whitespace, linebreaking, etc. around it), % given all the information in convenient, parsed form. % Args are the skip and penalty (usually negative) \def\dobreak#1#2{\par\ifdim\lastskip<#1\removelastskip\penalty#2\vskip#1\fi} % Parameter controlling skip before chapter headings (if needed) \newskip\chapheadingskip % Define plain chapter starts, and page on/off switching for it. \def\chapbreak{\dobreak \chapheadingskip {-4000}} \def\chappager{\par\vfill\supereject} % Because \domark is called before \chapoddpage, the filler page will % get the headings for the next chapter, which is wrong. But we don't % care -- we just disable all headings on the filler page. \def\chapoddpage{% \chappager \ifodd\pageno \else \begingroup \headingsoff \null \chappager \endgroup \fi } \def\setchapternewpage #1 {\csname CHAPPAG#1\endcsname} \def\CHAPPAGoff{% \global\let\contentsalignmacro = \chappager \global\let\pchapsepmacro=\chapbreak \global\let\pagealignmacro=\chappager} \def\CHAPPAGon{% \global\let\contentsalignmacro = \chappager \global\let\pchapsepmacro=\chappager \global\let\pagealignmacro=\chappager \global\def\HEADINGSon{\HEADINGSsingle}} \def\CHAPPAGodd{% \global\let\contentsalignmacro = \chapoddpage \global\let\pchapsepmacro=\chapoddpage \global\let\pagealignmacro=\chapoddpage \global\def\HEADINGSon{\HEADINGSdouble}} \CHAPPAGon % Chapter opening. % % #1 is the text, #2 is the section type (Ynumbered, Ynothing, % Yappendix, Yomitfromtoc), #3 the chapter number. % % To test against our argument. \def\Ynothingkeyword{Ynothing} \def\Yomitfromtockeyword{Yomitfromtoc} \def\Yappendixkeyword{Yappendix} % \def\chapmacro#1#2#3{% % Insert the first mark before the heading break (see notes for \domark). \let\prevchapterdefs=\lastchapterdefs \let\prevsectiondefs=\lastsectiondefs \gdef\lastsectiondefs{\gdef\thissectionname{}\gdef\thissectionnum{}% \gdef\thissection{}}% % \def\temptype{#2}% \ifx\temptype\Ynothingkeyword \gdef\lastchapterdefs{\gdef\thischaptername{#1}\gdef\thischapternum{}% \gdef\thischapter{\thischaptername}}% \else\ifx\temptype\Yomitfromtockeyword \gdef\lastchapterdefs{\gdef\thischaptername{#1}\gdef\thischapternum{}% \gdef\thischapter{}}% \else\ifx\temptype\Yappendixkeyword \toks0={#1}% \xdef\lastchapterdefs{% \gdef\noexpand\thischaptername{\the\toks0}% \gdef\noexpand\thischapternum{\appendixletter}% % \noexpand\putwordAppendix avoids expanding indigestible % commands in some of the translations. \gdef\noexpand\thischapter{\noexpand\putwordAppendix{} \noexpand\thischapternum: \noexpand\thischaptername}% }% \else \toks0={#1}% \xdef\lastchapterdefs{% \gdef\noexpand\thischaptername{\the\toks0}% \gdef\noexpand\thischapternum{\the\chapno}% % \noexpand\putwordChapter avoids expanding indigestible % commands in some of the translations. \gdef\noexpand\thischapter{\noexpand\putwordChapter{} \noexpand\thischapternum: \noexpand\thischaptername}% }% \fi\fi\fi % % Output the mark. Pass it through \safewhatsit, to take care of % the preceding space. \safewhatsit\domark % % Insert the chapter heading break. \pchapsepmacro % % Now the second mark, after the heading break. No break points % between here and the heading. \let\prevchapterdefs=\lastchapterdefs \let\prevsectiondefs=\lastsectiondefs \domark % {% \chapfonts \rmisbold % % Have to define \lastsection before calling \donoderef, because the % xref code eventually uses it. On the other hand, it has to be called % after \pchapsepmacro, or the headline will change too soon. \gdef\lastsection{#1}% % % Only insert the separating space if we have a chapter/appendix % number, and don't print the unnumbered ``number''. \ifx\temptype\Ynothingkeyword \setbox0 = \hbox{}% \def\toctype{unnchap}% \else\ifx\temptype\Yomitfromtockeyword \setbox0 = \hbox{}% contents like unnumbered, but no toc entry \def\toctype{omit}% \else\ifx\temptype\Yappendixkeyword \setbox0 = \hbox{\putwordAppendix{} #3\enspace}% \def\toctype{app}% \else \setbox0 = \hbox{#3\enspace}% \def\toctype{numchap}% \fi\fi\fi % % Write the toc entry for this chapter. Must come before the % \donoderef, because we include the current node name in the toc % entry, and \donoderef resets it to empty. \writetocentry{\toctype}{#1}{#3}% % % For pdftex, we have to write out the node definition (aka, make % the pdfdest) after any page break, but before the actual text has % been typeset. If the destination for the pdf outline is after the % text, then jumping from the outline may wind up with the text not % being visible, for instance under high magnification. \donoderef{#2}% % % Typeset the actual heading. \nobreak % Avoid page breaks at the interline glue. \vbox{\hyphenpenalty=10000 \tolerance=5000 \parindent=0pt \ptexraggedright \hangindent=\wd0 \centerparametersmaybe \unhbox0 #1\par}% }% \nobreak\bigskip % no page break after a chapter title \nobreak } % @centerchap -- centered and unnumbered. \let\centerparametersmaybe = \relax \def\centerparameters{% \advance\rightskip by 3\rightskip \leftskip = \rightskip \parfillskip = 0pt } % I don't think this chapter style is supported any more, so I'm not % updating it with the new noderef stuff. We'll see. --karl, 11aug03. % \def\setchapterstyle #1 {\csname CHAPF#1\endcsname} % \def\unnchfopen #1{% \chapoddpage {\chapfonts \vbox{\hyphenpenalty=10000\tolerance=5000 \parindent=0pt\ptexraggedright \rmisbold #1\hfill}}\bigskip \par\nobreak } \def\chfopen #1#2{\chapoddpage {\chapfonts \vbox to 3in{\vfil \hbox to\hsize{\hfil #2} \hbox to\hsize{\hfil #1} \vfil}}% \par\penalty 5000 % } \def\centerchfopen #1{% \chapoddpage {\chapfonts \vbox{\hyphenpenalty=10000\tolerance=5000 \parindent=0pt \hfill {\rmisbold #1}\hfill}}\bigskip \par\nobreak } \def\CHAPFopen{% \global\let\chapmacro=\chfopen \global\let\centerchapmacro=\centerchfopen} % Section titles. These macros combine the section number parts and % call the generic \sectionheading to do the printing. % \newskip\secheadingskip \def\secheadingbreak{\dobreak \secheadingskip{-1000}} % Subsection titles. \newskip\subsecheadingskip \def\subsecheadingbreak{\dobreak \subsecheadingskip{-500}} % Subsubsection titles. \def\subsubsecheadingskip{\subsecheadingskip} \def\subsubsecheadingbreak{\subsecheadingbreak} % Print any size, any type, section title. % % #1 is the text, #2 is the section level (sec/subsec/subsubsec), #3 is % the section type for xrefs (Ynumbered, Ynothing, Yappendix), #4 is the % section number. % \def\seckeyword{sec} % \def\sectionheading#1#2#3#4{% {% \checkenv{}% should not be in an environment. % % Switch to the right set of fonts. \csname #2fonts\endcsname \rmisbold % \def\sectionlevel{#2}% \def\temptype{#3}% % % Insert first mark before the heading break (see notes for \domark). \let\prevsectiondefs=\lastsectiondefs \ifx\temptype\Ynothingkeyword \ifx\sectionlevel\seckeyword \gdef\lastsectiondefs{\gdef\thissectionname{#1}\gdef\thissectionnum{}% \gdef\thissection{\thissectionname}}% \fi \else\ifx\temptype\Yomitfromtockeyword % Don't redefine \thissection. \else\ifx\temptype\Yappendixkeyword \ifx\sectionlevel\seckeyword \toks0={#1}% \xdef\lastsectiondefs{% \gdef\noexpand\thissectionname{\the\toks0}% \gdef\noexpand\thissectionnum{#4}% % \noexpand\putwordSection avoids expanding indigestible % commands in some of the translations. \gdef\noexpand\thissection{\noexpand\putwordSection{} \noexpand\thissectionnum: \noexpand\thissectionname}% }% \fi \else \ifx\sectionlevel\seckeyword \toks0={#1}% \xdef\lastsectiondefs{% \gdef\noexpand\thissectionname{\the\toks0}% \gdef\noexpand\thissectionnum{#4}% % \noexpand\putwordSection avoids expanding indigestible % commands in some of the translations. \gdef\noexpand\thissection{\noexpand\putwordSection{} \noexpand\thissectionnum: \noexpand\thissectionname}% }% \fi \fi\fi\fi % % Go into vertical mode. Usually we'll already be there, but we % don't want the following whatsit to end up in a preceding paragraph % if the document didn't happen to have a blank line. \par % % Output the mark. Pass it through \safewhatsit, to take care of % the preceding space. \safewhatsit\domark % % Insert space above the heading. \csname #2headingbreak\endcsname % % Now the second mark, after the heading break. No break points % between here and the heading. \let\prevsectiondefs=\lastsectiondefs \domark % % Only insert the space after the number if we have a section number. \ifx\temptype\Ynothingkeyword \setbox0 = \hbox{}% \def\toctype{unn}% \gdef\lastsection{#1}% \else\ifx\temptype\Yomitfromtockeyword % for @headings -- no section number, don't include in toc, % and don't redefine \lastsection. \setbox0 = \hbox{}% \def\toctype{omit}% \let\sectionlevel=\empty \else\ifx\temptype\Yappendixkeyword \setbox0 = \hbox{#4\enspace}% \def\toctype{app}% \gdef\lastsection{#1}% \else \setbox0 = \hbox{#4\enspace}% \def\toctype{num}% \gdef\lastsection{#1}% \fi\fi\fi % % Write the toc entry (before \donoderef). See comments in \chapmacro. \writetocentry{\toctype\sectionlevel}{#1}{#4}% % % Write the node reference (= pdf destination for pdftex). % Again, see comments in \chapmacro. \donoderef{#3}% % % Interline glue will be inserted when the vbox is completed. % That glue will be a valid breakpoint for the page, since it'll be % preceded by a whatsit (usually from the \donoderef, or from the % \writetocentry if there was no node). We don't want to allow that % break, since then the whatsits could end up on page n while the % section is on page n+1, thus toc/etc. are wrong. Debian bug 276000. \nobreak % % Output the actual section heading. \vbox{\hyphenpenalty=10000 \tolerance=5000 \parindent=0pt \ptexraggedright \hangindent=\wd0 % zero if no section number \unhbox0 #1}% }% % Add extra space after the heading -- half of whatever came above it. % Don't allow stretch, though. \kern .5 \csname #2headingskip\endcsname % % Do not let the kern be a potential breakpoint, as it would be if it % was followed by glue. \nobreak % % We'll almost certainly start a paragraph next, so don't let that % glue accumulate. (Not a breakpoint because it's preceded by a % discardable item.) However, when a paragraph is not started next % (\startdefun, \cartouche, \center, etc.), this needs to be wiped out % or the negative glue will cause weirdly wrong output, typically % obscuring the section heading with something else. \vskip-\parskip % % This is so the last item on the main vertical list is a known % \penalty > 10000, so \startdefun, etc., can recognize the situation % and do the needful. \penalty 10001 } \message{toc,} % Table of contents. \newwrite\tocfile % Write an entry to the toc file, opening it if necessary. % Called from @chapter, etc. % % Example usage: \writetocentry{sec}{Section Name}{\the\chapno.\the\secno} % We append the current node name (if any) and page number as additional % arguments for the \{chap,sec,...}entry macros which will eventually % read this. The node name is used in the pdf outlines as the % destination to jump to. % % We open the .toc file for writing here instead of at @setfilename (or % any other fixed time) so that @contents can be anywhere in the document. % But if #1 is `omit', then we don't do anything. This is used for the % table of contents chapter openings themselves. % \newif\iftocfileopened \def\omitkeyword{omit}% % \def\writetocentry#1#2#3{% \edef\writetoctype{#1}% \ifx\writetoctype\omitkeyword \else \iftocfileopened\else \immediate\openout\tocfile = \jobname.toc \global\tocfileopenedtrue \fi % \iflinks {\atdummies \edef\temp{% \write\tocfile{@#1entry{#2}{#3}{\lastnode}{\noexpand\folio}}}% \temp }% \fi \fi % % Tell \shipout to create a pdf destination on each page, if we're % writing pdf. These are used in the table of contents. We can't % just write one on every page because the title pages are numbered % 1 and 2 (the page numbers aren't printed), and so are the first % two pages of the document. Thus, we'd have two destinations named % `1', and two named `2'. \ifpdf \global\pdfmakepagedesttrue \fi } % These characters do not print properly in the Computer Modern roman % fonts, so we must take special care. This is more or less redundant % with the Texinfo input format setup at the end of this file. % \def\activecatcodes{% \catcode`\"=\active \catcode`\$=\active \catcode`\<=\active \catcode`\>=\active \catcode`\\=\active \catcode`\^=\active \catcode`\_=\active \catcode`\|=\active \catcode`\~=\active } % Read the toc file, which is essentially Texinfo input. \def\readtocfile{% \setupdatafile \activecatcodes \input \tocreadfilename } \newskip\contentsrightmargin \contentsrightmargin=1in \newcount\savepageno \newcount\lastnegativepageno \lastnegativepageno = -1 % Prepare to read what we've written to \tocfile. % \def\startcontents#1{% % If @setchapternewpage on, and @headings double, the contents should % start on an odd page, unlike chapters. Thus, we maintain % \contentsalignmacro in parallel with \pagealignmacro. % From: Torbjorn Granlund \contentsalignmacro \immediate\closeout\tocfile % % Don't need to put `Contents' or `Short Contents' in the headline. % It is abundantly clear what they are. \chapmacro{#1}{Yomitfromtoc}{}% % \savepageno = \pageno \begingroup % Set up to handle contents files properly. \raggedbottom % Worry more about breakpoints than the bottom. \advance\hsize by -\contentsrightmargin % Don't use the full line length. % % Roman numerals for page numbers. \ifnum \pageno>0 \global\pageno = \lastnegativepageno \fi } % redefined for the two-volume lispref. We always output on % \jobname.toc even if this is redefined. % \def\tocreadfilename{\jobname.toc} % Normal (long) toc. % \def\contents{% \startcontents{\putwordTOC}% \openin 1 \tocreadfilename\space \ifeof 1 \else \readtocfile \fi \vfill \eject \contentsalignmacro % in case @setchapternewpage odd is in effect \ifeof 1 \else \pdfmakeoutlines \fi \closein 1 \endgroup \lastnegativepageno = \pageno \global\pageno = \savepageno } % And just the chapters. \def\summarycontents{% \startcontents{\putwordShortTOC}% % \let\partentry = \shortpartentry \let\numchapentry = \shortchapentry \let\appentry = \shortchapentry \let\unnchapentry = \shortunnchapentry % We want a true roman here for the page numbers. \secfonts \let\rm=\shortcontrm \let\bf=\shortcontbf \let\sl=\shortcontsl \let\tt=\shortconttt \rm \hyphenpenalty = 10000 \advance\baselineskip by 1pt % Open it up a little. \def\numsecentry##1##2##3##4{} \let\appsecentry = \numsecentry \let\unnsecentry = \numsecentry \let\numsubsecentry = \numsecentry \let\appsubsecentry = \numsecentry \let\unnsubsecentry = \numsecentry \let\numsubsubsecentry = \numsecentry \let\appsubsubsecentry = \numsecentry \let\unnsubsubsecentry = \numsecentry \openin 1 \tocreadfilename\space \ifeof 1 \else \readtocfile \fi \closein 1 \vfill \eject \contentsalignmacro % in case @setchapternewpage odd is in effect \endgroup \lastnegativepageno = \pageno \global\pageno = \savepageno } \let\shortcontents = \summarycontents % Typeset the label for a chapter or appendix for the short contents. % The arg is, e.g., `A' for an appendix, or `3' for a chapter. % \def\shortchaplabel#1{% % This space should be enough, since a single number is .5em, and the % widest letter (M) is 1em, at least in the Computer Modern fonts. % But use \hss just in case. % (This space doesn't include the extra space that gets added after % the label; that gets put in by \shortchapentry above.) % % We'd like to right-justify chapter numbers, but that looks strange % with appendix letters. And right-justifying numbers and % left-justifying letters looks strange when there is less than 10 % chapters. Have to read the whole toc once to know how many chapters % there are before deciding ... \hbox to 1em{#1\hss}% } % These macros generate individual entries in the table of contents. % The first argument is the chapter or section name. % The last argument is the page number. % The arguments in between are the chapter number, section number, ... % Parts, in the main contents. Replace the part number, which doesn't % exist, with an empty box. Let's hope all the numbers have the same width. % Also ignore the page number, which is conventionally not printed. \def\numeralbox{\setbox0=\hbox{8}\hbox to \wd0{\hfil}} \def\partentry#1#2#3#4{\dochapentry{\numeralbox\labelspace#1}{}} % % Parts, in the short toc. \def\shortpartentry#1#2#3#4{% \penalty-300 \vskip.5\baselineskip plus.15\baselineskip minus.1\baselineskip \shortchapentry{{\bf #1}}{\numeralbox}{}{}% } % Chapters, in the main contents. \def\numchapentry#1#2#3#4{\dochapentry{#2\labelspace#1}{#4}} % % Chapters, in the short toc. % See comments in \dochapentry re vbox and related settings. \def\shortchapentry#1#2#3#4{% \tocentry{\shortchaplabel{#2}\labelspace #1}{\doshortpageno\bgroup#4\egroup}% } % Appendices, in the main contents. % Need the word Appendix, and a fixed-size box. % \def\appendixbox#1{% % We use M since it's probably the widest letter. \setbox0 = \hbox{\putwordAppendix{} M}% \hbox to \wd0{\putwordAppendix{} #1\hss}} % \def\appentry#1#2#3#4{\dochapentry{\appendixbox{#2}\labelspace#1}{#4}} % Unnumbered chapters. \def\unnchapentry#1#2#3#4{\dochapentry{#1}{#4}} \def\shortunnchapentry#1#2#3#4{\tocentry{#1}{\doshortpageno\bgroup#4\egroup}} % Sections. \def\numsecentry#1#2#3#4{\dosecentry{#2\labelspace#1}{#4}} \let\appsecentry=\numsecentry \def\unnsecentry#1#2#3#4{\dosecentry{#1}{#4}} % Subsections. \def\numsubsecentry#1#2#3#4{\dosubsecentry{#2\labelspace#1}{#4}} \let\appsubsecentry=\numsubsecentry \def\unnsubsecentry#1#2#3#4{\dosubsecentry{#1}{#4}} % And subsubsections. \def\numsubsubsecentry#1#2#3#4{\dosubsubsecentry{#2\labelspace#1}{#4}} \let\appsubsubsecentry=\numsubsubsecentry \def\unnsubsubsecentry#1#2#3#4{\dosubsubsecentry{#1}{#4}} % This parameter controls the indentation of the various levels. % Same as \defaultparindent. \newdimen\tocindent \tocindent = 15pt % Now for the actual typesetting. In all these, #1 is the text and #2 is the % page number. % % If the toc has to be broken over pages, we want it to be at chapters % if at all possible; hence the \penalty. \def\dochapentry#1#2{% \penalty-300 \vskip1\baselineskip plus.33\baselineskip minus.25\baselineskip \begingroup \chapentryfonts \tocentry{#1}{\dopageno\bgroup#2\egroup}% \endgroup \nobreak\vskip .25\baselineskip plus.1\baselineskip } \def\dosecentry#1#2{\begingroup \secentryfonts \leftskip=\tocindent \tocentry{#1}{\dopageno\bgroup#2\egroup}% \endgroup} \def\dosubsecentry#1#2{\begingroup \subsecentryfonts \leftskip=2\tocindent \tocentry{#1}{\dopageno\bgroup#2\egroup}% \endgroup} \def\dosubsubsecentry#1#2{\begingroup \subsubsecentryfonts \leftskip=3\tocindent \tocentry{#1}{\dopageno\bgroup#2\egroup}% \endgroup} % We use the same \entry macro as for the index entries. \let\tocentry = \entry % Space between chapter (or whatever) number and the title. \def\labelspace{\hskip1em \relax} \def\dopageno#1{{\rm #1}} \def\doshortpageno#1{{\rm #1}} \def\chapentryfonts{\secfonts \rm} \def\secentryfonts{\textfonts} \def\subsecentryfonts{\textfonts} \def\subsubsecentryfonts{\textfonts} \message{environments,} % @foo ... @end foo. % @tex ... @end tex escapes into raw TeX temporarily. % One exception: @ is still an escape character, so that @end tex works. % But \@ or @@ will get a plain @ character. \envdef\tex{% \setupmarkupstyle{tex}% \catcode `\\=0 \catcode `\{=1 \catcode `\}=2 \catcode `\$=3 \catcode `\&=4 \catcode `\#=6 \catcode `\^=7 \catcode `\_=8 \catcode `\~=\active \let~=\tie \catcode `\%=14 \catcode `\+=\other \catcode `\"=\other \catcode `\|=\other \catcode `\<=\other \catcode `\>=\other \catcode`\`=\other \catcode`\'=\other \escapechar=`\\ % % ' is active in math mode (mathcode"8000). So reset it, and all our % other math active characters (just in case), to plain's definitions. \mathactive % \let\b=\ptexb \let\bullet=\ptexbullet \let\c=\ptexc \let\,=\ptexcomma \let\.=\ptexdot \let\dots=\ptexdots \let\equiv=\ptexequiv \let\!=\ptexexclam \let\i=\ptexi \let\indent=\ptexindent \let\noindent=\ptexnoindent \let\{=\ptexlbrace \let\+=\tabalign \let\}=\ptexrbrace \let\/=\ptexslash \let\*=\ptexstar \let\t=\ptext \expandafter \let\csname top\endcsname=\ptextop % outer \let\frenchspacing=\plainfrenchspacing % \def\endldots{\mathinner{\ldots\ldots\ldots\ldots}}% \def\enddots{\relax\ifmmode\endldots\else$\mathsurround=0pt \endldots\,$\fi}% \def\@{@}% } % There is no need to define \Etex. % Define @lisp ... @end lisp. % @lisp environment forms a group so it can rebind things, % including the definition of @end lisp (which normally is erroneous). % Amount to narrow the margins by for @lisp. \newskip\lispnarrowing \lispnarrowing=0.4in % This is the definition that ^^M gets inside @lisp, @example, and other % such environments. \null is better than a space, since it doesn't % have any width. \def\lisppar{\null\endgraf} % This space is always present above and below environments. \newskip\envskipamount \envskipamount = 0pt % Make spacing and below environment symmetrical. We use \parskip here % to help in doing that, since in @example-like environments \parskip % is reset to zero; thus the \afterenvbreak inserts no space -- but the % start of the next paragraph will insert \parskip. % \def\aboveenvbreak{{% % =10000 instead of <10000 because of a special case in \itemzzz and % \sectionheading, q.v. \ifnum \lastpenalty=10000 \else \advance\envskipamount by \parskip \endgraf \ifdim\lastskip<\envskipamount \removelastskip % it's not a good place to break if the last penalty was \nobreak % or better ... \ifnum\lastpenalty<10000 \penalty-50 \fi \vskip\envskipamount \fi \fi }} \let\afterenvbreak = \aboveenvbreak % \nonarrowing is a flag. If "set", @lisp etc don't narrow margins; it will % also clear it, so that its embedded environments do the narrowing again. \let\nonarrowing=\relax % @cartouche ... @end cartouche: draw rectangle w/rounded corners around % environment contents. \font\circle=lcircle10 \newdimen\circthick \newdimen\cartouter\newdimen\cartinner \newskip\normbskip\newskip\normpskip\newskip\normlskip \circthick=\fontdimen8\circle % \def\ctl{{\circle\char'013\hskip -6pt}}% 6pt from pl file: 1/2charwidth \def\ctr{{\hskip 6pt\circle\char'010}} \def\cbl{{\circle\char'012\hskip -6pt}} \def\cbr{{\hskip 6pt\circle\char'011}} \def\carttop{\hbox to \cartouter{\hskip\lskip \ctl\leaders\hrule height\circthick\hfil\ctr \hskip\rskip}} \def\cartbot{\hbox to \cartouter{\hskip\lskip \cbl\leaders\hrule height\circthick\hfil\cbr \hskip\rskip}} % \newskip\lskip\newskip\rskip \envdef\cartouche{% \ifhmode\par\fi % can't be in the midst of a paragraph. \startsavinginserts \lskip=\leftskip \rskip=\rightskip \leftskip=0pt\rightskip=0pt % we want these *outside*. \cartinner=\hsize \advance\cartinner by-\lskip \advance\cartinner by-\rskip \cartouter=\hsize \advance\cartouter by 18.4pt % allow for 3pt kerns on either % side, and for 6pt waste from % each corner char, and rule thickness \normbskip=\baselineskip \normpskip=\parskip \normlskip=\lineskip % Flag to tell @lisp, etc., not to narrow margin. \let\nonarrowing = t% % % If this cartouche directly follows a sectioning command, we need the % \parskip glue (backspaced over by default) or the cartouche can % collide with the section heading. \ifnum\lastpenalty>10000 \vskip\parskip \penalty\lastpenalty \fi % \vbox\bgroup \baselineskip=0pt\parskip=0pt\lineskip=0pt \carttop \hbox\bgroup \hskip\lskip \vrule\kern3pt \vbox\bgroup \kern3pt \hsize=\cartinner \baselineskip=\normbskip \lineskip=\normlskip \parskip=\normpskip \vskip -\parskip \comment % For explanation, see the end of def\group. } \def\Ecartouche{% \ifhmode\par\fi \kern3pt \egroup \kern3pt\vrule \hskip\rskip \egroup \cartbot \egroup \checkinserts } % This macro is called at the beginning of all the @example variants, % inside a group. \newdimen\nonfillparindent \def\nonfillstart{% \aboveenvbreak \hfuzz = 12pt % Don't be fussy \sepspaces % Make spaces be word-separators rather than space tokens. \let\par = \lisppar % don't ignore blank lines \obeylines % each line of input is a line of output \parskip = 0pt % Turn off paragraph indentation but redefine \indent to emulate % the normal \indent. \nonfillparindent=\parindent \parindent = 0pt \let\indent\nonfillindent % \emergencystretch = 0pt % don't try to avoid overfull boxes \ifx\nonarrowing\relax \advance \leftskip by \lispnarrowing \exdentamount=\lispnarrowing \else \let\nonarrowing = \relax \fi \let\exdent=\nofillexdent } \begingroup \obeyspaces % We want to swallow spaces (but not other tokens) after the fake % @indent in our nonfill-environments, where spaces are normally % active and set to @tie, resulting in them not being ignored after % @indent. \gdef\nonfillindent{\futurelet\temp\nonfillindentcheck}% \gdef\nonfillindentcheck{% \ifx\temp % \expandafter\nonfillindentgobble% \else% \leavevmode\nonfillindentbox% \fi% }% \endgroup \def\nonfillindentgobble#1{\nonfillindent} \def\nonfillindentbox{\hbox to \nonfillparindent{\hss}} % If you want all examples etc. small: @set dispenvsize small. % If you want even small examples the full size: @set dispenvsize nosmall. % This affects the following displayed environments: % @example, @display, @format, @lisp % \def\smallword{small} \def\nosmallword{nosmall} \let\SETdispenvsize\relax \def\setnormaldispenv{% \ifx\SETdispenvsize\smallword % end paragraph for sake of leading, in case document has no blank % line. This is redundant with what happens in \aboveenvbreak, but % we need to do it before changing the fonts, and it's inconvenient % to change the fonts afterward. \ifnum \lastpenalty=10000 \else \endgraf \fi \smallexamplefonts \rm \fi } \def\setsmalldispenv{% \ifx\SETdispenvsize\nosmallword \else \ifnum \lastpenalty=10000 \else \endgraf \fi \smallexamplefonts \rm \fi } % We often define two environments, @foo and @smallfoo. % Let's do it in one command. #1 is the env name, #2 the definition. \def\makedispenvdef#1#2{% \expandafter\envdef\csname#1\endcsname {\setnormaldispenv #2}% \expandafter\envdef\csname small#1\endcsname {\setsmalldispenv #2}% \expandafter\let\csname E#1\endcsname \afterenvbreak \expandafter\let\csname Esmall#1\endcsname \afterenvbreak } % Define two environment synonyms (#1 and #2) for an environment. \def\maketwodispenvdef#1#2#3{% \makedispenvdef{#1}{#3}% \makedispenvdef{#2}{#3}% } % % @lisp: indented, narrowed, typewriter font; % @example: same as @lisp. % % @smallexample and @smalllisp: use smaller fonts. % Originally contributed by Pavel@xerox. % \maketwodispenvdef{lisp}{example}{% \nonfillstart \tt\setupmarkupstyle{example}% \let\kbdfont = \kbdexamplefont % Allow @kbd to do something special. \gobble % eat return } % @display/@smalldisplay: same as @lisp except keep current font. % \makedispenvdef{display}{% \nonfillstart \gobble } % @format/@smallformat: same as @display except don't narrow margins. % \makedispenvdef{format}{% \let\nonarrowing = t% \nonfillstart \gobble } % @flushleft: same as @format, but doesn't obey \SETdispenvsize. \envdef\flushleft{% \let\nonarrowing = t% \nonfillstart \gobble } \let\Eflushleft = \afterenvbreak % @flushright. % \envdef\flushright{% \let\nonarrowing = t% \nonfillstart \advance\leftskip by 0pt plus 1fill\relax \gobble } \let\Eflushright = \afterenvbreak % @raggedright does more-or-less normal line breaking but no right % justification. From plain.tex. \envdef\raggedright{% \rightskip0pt plus2em \spaceskip.3333em \xspaceskip.5em\relax } \let\Eraggedright\par \envdef\raggedleft{% \parindent=0pt \leftskip0pt plus2em \spaceskip.3333em \xspaceskip.5em \parfillskip=0pt \hbadness=10000 % Last line will usually be underfull, so turn off % badness reporting. } \let\Eraggedleft\par \envdef\raggedcenter{% \parindent=0pt \rightskip0pt plus1em \leftskip0pt plus1em \spaceskip.3333em \xspaceskip.5em \parfillskip=0pt \hbadness=10000 % Last line will usually be underfull, so turn off % badness reporting. } \let\Eraggedcenter\par % @quotation does normal linebreaking (hence we can't use \nonfillstart) % and narrows the margins. We keep \parskip nonzero in general, since % we're doing normal filling. So, when using \aboveenvbreak and % \afterenvbreak, temporarily make \parskip 0. % \makedispenvdef{quotation}{\quotationstart} % \def\quotationstart{% {\parskip=0pt \aboveenvbreak}% because \aboveenvbreak inserts \parskip \parindent=0pt % % @cartouche defines \nonarrowing to inhibit narrowing at next level down. \ifx\nonarrowing\relax \advance\leftskip by \lispnarrowing \advance\rightskip by \lispnarrowing \exdentamount = \lispnarrowing \else \let\nonarrowing = \relax \fi \parsearg\quotationlabel } % We have retained a nonzero parskip for the environment, since we're % doing normal filling. % \def\Equotation{% \par \ifx\quotationauthor\thisisundefined\else % indent a bit. \leftline{\kern 2\leftskip \sl ---\quotationauthor}% \fi {\parskip=0pt \afterenvbreak}% } \def\Esmallquotation{\Equotation} % If we're given an argument, typeset it in bold with a colon after. \def\quotationlabel#1{% \def\temp{#1}% \ifx\temp\empty \else {\bf #1: }% \fi } % LaTeX-like @verbatim...@end verbatim and @verb{...} % If we want to allow any as delimiter, % we need the curly braces so that makeinfo sees the @verb command, eg: % `@verbx...x' would look like the '@verbx' command. --janneke@gnu.org % % [Knuth]: Donald Ervin Knuth, 1996. The TeXbook. % % [Knuth] p.344; only we need to do the other characters Texinfo sets % active too. Otherwise, they get lost as the first character on a % verbatim line. \def\dospecials{% \do\ \do\\\do\{\do\}\do\$\do\&% \do\#\do\^\do\^^K\do\_\do\^^A\do\%\do\~% \do\<\do\>\do\|\do\@\do+\do\"% % Don't do the quotes -- if we do, @set txicodequoteundirected and % @set txicodequotebacktick will not have effect on @verb and % @verbatim, and ?` and !` ligatures won't get disabled. %\do\`\do\'% } % % [Knuth] p. 380 \def\uncatcodespecials{% \def\do##1{\catcode`##1=\other}\dospecials} % % Setup for the @verb command. % % Eight spaces for a tab \begingroup \catcode`\^^I=\active \gdef\tabeightspaces{\catcode`\^^I=\active\def^^I{\ \ \ \ \ \ \ \ }} \endgroup % \def\setupverb{% \tt % easiest (and conventionally used) font for verbatim \def\par{\leavevmode\endgraf}% \setupmarkupstyle{verb}% \tabeightspaces % Respect line breaks, % print special symbols as themselves, and % make each space count % must do in this order: \obeylines \uncatcodespecials \sepspaces } % Setup for the @verbatim environment % % Real tab expansion. \newdimen\tabw \setbox0=\hbox{\tt\space} \tabw=8\wd0 % tab amount % % We typeset each line of the verbatim in an \hbox, so we can handle % tabs. The \global is in case the verbatim line starts with an accent, % or some other command that starts with a begin-group. Otherwise, the % entire \verbbox would disappear at the corresponding end-group, before % it is typeset. Meanwhile, we can't have nested verbatim commands % (can we?), so the \global won't be overwriting itself. \newbox\verbbox \def\starttabbox{\global\setbox\verbbox=\hbox\bgroup} % \begingroup \catcode`\^^I=\active \gdef\tabexpand{% \catcode`\^^I=\active \def^^I{\leavevmode\egroup \dimen\verbbox=\wd\verbbox % the width so far, or since the previous tab \divide\dimen\verbbox by\tabw \multiply\dimen\verbbox by\tabw % compute previous multiple of \tabw \advance\dimen\verbbox by\tabw % advance to next multiple of \tabw \wd\verbbox=\dimen\verbbox \box\verbbox \starttabbox }% } \endgroup % start the verbatim environment. \def\setupverbatim{% \let\nonarrowing = t% \nonfillstart \tt % easiest (and conventionally used) font for verbatim % The \leavevmode here is for blank lines. Otherwise, we would % never \starttabox and the \egroup would end verbatim mode. \def\par{\leavevmode\egroup\box\verbbox\endgraf}% \tabexpand \setupmarkupstyle{verbatim}% % Respect line breaks, % print special symbols as themselves, and % make each space count. % Must do in this order: \obeylines \uncatcodespecials \sepspaces \everypar{\starttabbox}% } % Do the @verb magic: verbatim text is quoted by unique % delimiter characters. Before first delimiter expect a % right brace, after last delimiter expect closing brace: % % \def\doverb'{'#1'}'{#1} % % [Knuth] p. 382; only eat outer {} \begingroup \catcode`[=1\catcode`]=2\catcode`\{=\other\catcode`\}=\other \gdef\doverb{#1[\def\next##1#1}[##1\endgroup]\next] \endgroup % \def\verb{\begingroup\setupverb\doverb} % % % Do the @verbatim magic: define the macro \doverbatim so that % the (first) argument ends when '@end verbatim' is reached, ie: % % \def\doverbatim#1@end verbatim{#1} % % For Texinfo it's a lot easier than for LaTeX, % because texinfo's \verbatim doesn't stop at '\end{verbatim}': % we need not redefine '\', '{' and '}'. % % Inspired by LaTeX's verbatim command set [latex.ltx] % \begingroup \catcode`\ =\active \obeylines % % ignore everything up to the first ^^M, that's the newline at the end % of the @verbatim input line itself. Otherwise we get an extra blank % line in the output. \xdef\doverbatim#1^^M#2@end verbatim{#2\noexpand\end\gobble verbatim}% % We really want {...\end verbatim} in the body of the macro, but % without the active space; thus we have to use \xdef and \gobble. \endgroup % \envdef\verbatim{% \setupverbatim\doverbatim } \let\Everbatim = \afterenvbreak % @verbatiminclude FILE - insert text of file in verbatim environment. % \def\verbatiminclude{\parseargusing\filenamecatcodes\doverbatiminclude} % \def\doverbatiminclude#1{% {% \makevalueexpandable \setupverbatim \indexnofonts % Allow `@@' and other weird things in file names. \wlog{texinfo.tex: doing @verbatiminclude of #1^^J}% \input #1 \afterenvbreak }% } % @copying ... @end copying. % Save the text away for @insertcopying later. % % We save the uninterpreted tokens, rather than creating a box. % Saving the text in a box would be much easier, but then all the % typesetting commands (@smallbook, font changes, etc.) have to be done % beforehand -- and a) we want @copying to be done first in the source % file; b) letting users define the frontmatter in as flexible order as % possible is very desirable. % \def\copying{\checkenv{}\begingroup\scanargctxt\docopying} \def\docopying#1@end copying{\endgroup\def\copyingtext{#1}} % \def\insertcopying{% \begingroup \parindent = 0pt % paragraph indentation looks wrong on title page \scanexp\copyingtext \endgroup } \message{defuns,} % @defun etc. \newskip\defbodyindent \defbodyindent=.4in \newskip\defargsindent \defargsindent=50pt \newskip\deflastargmargin \deflastargmargin=18pt \newcount\defunpenalty % Start the processing of @deffn: \def\startdefun{% \ifnum\lastpenalty<10000 \medbreak \defunpenalty=10003 % Will keep this @deffn together with the % following @def command, see below. \else % If there are two @def commands in a row, we'll have a \nobreak, % which is there to keep the function description together with its % header. But if there's nothing but headers, we need to allow a % break somewhere. Check specifically for penalty 10002, inserted % by \printdefunline, instead of 10000, since the sectioning % commands also insert a nobreak penalty, and we don't want to allow % a break between a section heading and a defun. % % As a further refinement, we avoid "club" headers by signalling % with penalty of 10003 after the very first @deffn in the % sequence (see above), and penalty of 10002 after any following % @def command. \ifnum\lastpenalty=10002 \penalty2000 \else \defunpenalty=10002 \fi % % Similarly, after a section heading, do not allow a break. % But do insert the glue. \medskip % preceded by discardable penalty, so not a breakpoint \fi % \parindent=0in \advance\leftskip by \defbodyindent \exdentamount=\defbodyindent } \def\dodefunx#1{% % First, check whether we are in the right environment: \checkenv#1% % % As above, allow line break if we have multiple x headers in a row. % It's not a great place, though. \ifnum\lastpenalty=10002 \penalty3000 \else \defunpenalty=10002 \fi % % And now, it's time to reuse the body of the original defun: \expandafter\gobbledefun#1% } \def\gobbledefun#1\startdefun{} % \printdefunline \deffnheader{text} % \def\printdefunline#1#2{% \begingroup % call \deffnheader: #1#2 \endheader % common ending: \interlinepenalty = 10000 \advance\rightskip by 0pt plus 1fil\relax \endgraf \nobreak\vskip -\parskip \penalty\defunpenalty % signal to \startdefun and \dodefunx % Some of the @defun-type tags do not enable magic parentheses, % rendering the following check redundant. But we don't optimize. \checkparencounts \endgroup } \def\Edefun{\endgraf\medbreak} % \makedefun{deffn} creates \deffn, \deffnx and \Edeffn; % the only thing remaining is to define \deffnheader. % \def\makedefun#1{% \expandafter\let\csname E#1\endcsname = \Edefun \edef\temp{\noexpand\domakedefun \makecsname{#1}\makecsname{#1x}\makecsname{#1header}}% \temp } % \domakedefun \deffn \deffnx \deffnheader % % Define \deffn and \deffnx, without parameters. % \deffnheader has to be defined explicitly. % \def\domakedefun#1#2#3{% \envdef#1{% \startdefun \doingtypefnfalse % distinguish typed functions from all else \parseargusing\activeparens{\printdefunline#3}% }% \def#2{\dodefunx#1}% \def#3% } \newif\ifdoingtypefn % doing typed function? \newif\ifrettypeownline % typeset return type on its own line? % @deftypefnnewline on|off says whether the return type of typed functions % are printed on their own line. This affects @deftypefn, @deftypefun, % @deftypeop, and @deftypemethod. % \parseargdef\deftypefnnewline{% \def\temp{#1}% \ifx\temp\onword \expandafter\let\csname SETtxideftypefnnl\endcsname = \empty \else\ifx\temp\offword \expandafter\let\csname SETtxideftypefnnl\endcsname = \relax \else \errhelp = \EMsimple \errmessage{Unknown @txideftypefnnl value `\temp', must be on|off}% \fi\fi } % Untyped functions: % @deffn category name args \makedefun{deffn}{\deffngeneral{}} % @deffn category class name args \makedefun{defop}#1 {\defopon{#1\ \putwordon}} % \defopon {category on}class name args \def\defopon#1#2 {\deffngeneral{\putwordon\ \code{#2}}{#1\ \code{#2}} } % \deffngeneral {subind}category name args % \def\deffngeneral#1#2 #3 #4\endheader{% % Remember that \dosubind{fn}{foo}{} is equivalent to \doind{fn}{foo}. \dosubind{fn}{\code{#3}}{#1}% \defname{#2}{}{#3}\magicamp\defunargs{#4\unskip}% } % Typed functions: % @deftypefn category type name args \makedefun{deftypefn}{\deftypefngeneral{}} % @deftypeop category class type name args \makedefun{deftypeop}#1 {\deftypeopon{#1\ \putwordon}} % \deftypeopon {category on}class type name args \def\deftypeopon#1#2 {\deftypefngeneral{\putwordon\ \code{#2}}{#1\ \code{#2}} } % \deftypefngeneral {subind}category type name args % \def\deftypefngeneral#1#2 #3 #4 #5\endheader{% \dosubind{fn}{\code{#4}}{#1}% \doingtypefntrue \defname{#2}{#3}{#4}\defunargs{#5\unskip}% } % Typed variables: % @deftypevr category type var args \makedefun{deftypevr}{\deftypecvgeneral{}} % @deftypecv category class type var args \makedefun{deftypecv}#1 {\deftypecvof{#1\ \putwordof}} % \deftypecvof {category of}class type var args \def\deftypecvof#1#2 {\deftypecvgeneral{\putwordof\ \code{#2}}{#1\ \code{#2}} } % \deftypecvgeneral {subind}category type var args % \def\deftypecvgeneral#1#2 #3 #4 #5\endheader{% \dosubind{vr}{\code{#4}}{#1}% \defname{#2}{#3}{#4}\defunargs{#5\unskip}% } % Untyped variables: % @defvr category var args \makedefun{defvr}#1 {\deftypevrheader{#1} {} } % @defcv category class var args \makedefun{defcv}#1 {\defcvof{#1\ \putwordof}} % \defcvof {category of}class var args \def\defcvof#1#2 {\deftypecvof{#1}#2 {} } % Types: % @deftp category name args \makedefun{deftp}#1 #2 #3\endheader{% \doind{tp}{\code{#2}}% \defname{#1}{}{#2}\defunargs{#3\unskip}% } % Remaining @defun-like shortcuts: \makedefun{defun}{\deffnheader{\putwordDeffunc} } \makedefun{defmac}{\deffnheader{\putwordDefmac} } \makedefun{defspec}{\deffnheader{\putwordDefspec} } \makedefun{deftypefun}{\deftypefnheader{\putwordDeffunc} } \makedefun{defvar}{\defvrheader{\putwordDefvar} } \makedefun{defopt}{\defvrheader{\putwordDefopt} } \makedefun{deftypevar}{\deftypevrheader{\putwordDefvar} } \makedefun{defmethod}{\defopon\putwordMethodon} \makedefun{deftypemethod}{\deftypeopon\putwordMethodon} \makedefun{defivar}{\defcvof\putwordInstanceVariableof} \makedefun{deftypeivar}{\deftypecvof\putwordInstanceVariableof} % \defname, which formats the name of the @def (not the args). % #1 is the category, such as "Function". % #2 is the return type, if any. % #3 is the function name. % % We are followed by (but not passed) the arguments, if any. % \def\defname#1#2#3{% \par % Get the values of \leftskip and \rightskip as they were outside the @def... \advance\leftskip by -\defbodyindent % % Determine if we are typesetting the return type of a typed function % on a line by itself. \rettypeownlinefalse \ifdoingtypefn % doing a typed function specifically? % then check user option for putting return type on its own line: \expandafter\ifx\csname SETtxideftypefnnl\endcsname\relax \else \rettypeownlinetrue \fi \fi % % How we'll format the category name. Putting it in brackets helps % distinguish it from the body text that may end up on the next line % just below it. \def\temp{#1}% \setbox0=\hbox{\kern\deflastargmargin \ifx\temp\empty\else [\rm\temp]\fi} % % Figure out line sizes for the paragraph shape. We'll always have at % least two. \tempnum = 2 % % The first line needs space for \box0; but if \rightskip is nonzero, % we need only space for the part of \box0 which exceeds it: \dimen0=\hsize \advance\dimen0 by -\wd0 \advance\dimen0 by \rightskip % % If doing a return type on its own line, we'll have another line. \ifrettypeownline \advance\tempnum by 1 \def\maybeshapeline{0in \hsize}% \else \def\maybeshapeline{}% \fi % % The continuations: \dimen2=\hsize \advance\dimen2 by -\defargsindent % % The final paragraph shape: \parshape \tempnum 0in \dimen0 \maybeshapeline \defargsindent \dimen2 % % Put the category name at the right margin. \noindent \hbox to 0pt{% \hfil\box0 \kern-\hsize % \hsize has to be shortened this way: \kern\leftskip % Intentionally do not respect \rightskip, since we need the space. }% % % Allow all lines to be underfull without complaint: \tolerance=10000 \hbadness=10000 \exdentamount=\defbodyindent {% % defun fonts. We use typewriter by default (used to be bold) because: % . we're printing identifiers, they should be in tt in principle. % . in languages with many accents, such as Czech or French, it's % common to leave accents off identifiers. The result looks ok in % tt, but exceedingly strange in rm. % . we don't want -- and --- to be treated as ligatures. % . this still does not fix the ?` and !` ligatures, but so far no % one has made identifiers using them :). \df \tt \def\temp{#2}% text of the return type \ifx\temp\empty\else \tclose{\temp}% typeset the return type \ifrettypeownline % put return type on its own line; prohibit line break following: \hfil\vadjust{\nobreak}\break \else \space % type on same line, so just followed by a space \fi \fi % no return type #3% output function name }% {\rm\enskip}% hskip 0.5 em of \tenrm % \boldbrax % arguments will be output next, if any. } % Print arguments in slanted roman (not ttsl), inconsistently with using % tt for the name. This is because literal text is sometimes needed in % the argument list (groff manual), and ttsl and tt are not very % distinguishable. Prevent hyphenation at `-' chars. % \def\defunargs#1{% % use sl by default (not ttsl), % tt for the names. \df \sl \hyphenchar\font=0 % % On the other hand, if an argument has two dashes (for instance), we % want a way to get ttsl. Let's try @var for that. \def\var##1{{\setupmarkupstyle{var}\ttslanted{##1}}}% #1% \sl\hyphenchar\font=45 } % We want ()&[] to print specially on the defun line. % \def\activeparens{% \catcode`\(=\active \catcode`\)=\active \catcode`\[=\active \catcode`\]=\active \catcode`\&=\active } % Make control sequences which act like normal parenthesis chars. \let\lparen = ( \let\rparen = ) % Be sure that we always have a definition for `(', etc. For example, % if the fn name has parens in it, \boldbrax will not be in effect yet, % so TeX would otherwise complain about undefined control sequence. { \activeparens \global\let(=\lparen \global\let)=\rparen \global\let[=\lbrack \global\let]=\rbrack \global\let& = \& \gdef\boldbrax{\let(=\opnr\let)=\clnr\let[=\lbrb\let]=\rbrb} \gdef\magicamp{\let&=\amprm} } \newcount\parencount % If we encounter &foo, then turn on ()-hacking afterwards \newif\ifampseen \def\amprm#1 {\ampseentrue{\bf\ }} \def\parenfont{% \ifampseen % At the first level, print parens in roman, % otherwise use the default font. \ifnum \parencount=1 \rm \fi \else % The \sf parens (in \boldbrax) actually are a little bolder than % the contained text. This is especially needed for [ and ] . \sf \fi } \def\infirstlevel#1{% \ifampseen \ifnum\parencount=1 #1% \fi \fi } \def\bfafterword#1 {#1 \bf} \def\opnr{% \global\advance\parencount by 1 {\parenfont(}% \infirstlevel \bfafterword } \def\clnr{% {\parenfont)}% \infirstlevel \sl \global\advance\parencount by -1 } \newcount\brackcount \def\lbrb{% \global\advance\brackcount by 1 {\bf[}% } \def\rbrb{% {\bf]}% \global\advance\brackcount by -1 } \def\checkparencounts{% \ifnum\parencount=0 \else \badparencount \fi \ifnum\brackcount=0 \else \badbrackcount \fi } % these should not use \errmessage; the glibc manual, at least, actually % has such constructs (when documenting function pointers). \def\badparencount{% \message{Warning: unbalanced parentheses in @def...}% \global\parencount=0 } \def\badbrackcount{% \message{Warning: unbalanced square brackets in @def...}% \global\brackcount=0 } \message{macros,} % @macro. % To do this right we need a feature of e-TeX, \scantokens, % which we arrange to emulate with a temporary file in ordinary TeX. \ifx\eTeXversion\thisisundefined \newwrite\macscribble \def\scantokens#1{% \toks0={#1}% \immediate\openout\macscribble=\jobname.tmp \immediate\write\macscribble{\the\toks0}% \immediate\closeout\macscribble \input \jobname.tmp } \fi \def\scanmacro#1{\begingroup \newlinechar`\^^M \let\xeatspaces\eatspaces % % Undo catcode changes of \startcontents and \doprintindex % When called from @insertcopying or (short)caption, we need active % backslash to get it printed correctly. Previously, we had % \catcode`\\=\other instead. We'll see whether a problem appears % with macro expansion. --kasal, 19aug04 \catcode`\@=0 \catcode`\\=\active \escapechar=`\@ % % ... and for \example: \spaceisspace % % The \empty here causes a following catcode 5 newline to be eaten as % part of reading whitespace after a control sequence. It does not % eat a catcode 13 newline. There's no good way to handle the two % cases (untried: maybe e-TeX's \everyeof could help, though plain TeX % would then have different behavior). See the Macro Details node in % the manual for the workaround we recommend for macros and % line-oriented commands. % \scantokens{#1\empty}% \endgroup} \def\scanexp#1{% \edef\temp{\noexpand\scanmacro{#1}}% \temp } \newcount\paramno % Count of parameters \newtoks\macname % Macro name \newif\ifrecursive % Is it recursive? % List of all defined macros in the form % \definedummyword\macro1\definedummyword\macro2... % Currently is also contains all @aliases; the list can be split % if there is a need. \def\macrolist{} % Add the macro to \macrolist \def\addtomacrolist#1{\expandafter \addtomacrolistxxx \csname#1\endcsname} \def\addtomacrolistxxx#1{% \toks0 = \expandafter{\macrolist\definedummyword#1}% \xdef\macrolist{\the\toks0}% } % Utility routines. % This does \let #1 = #2, with \csnames; that is, % \let \csname#1\endcsname = \csname#2\endcsname % (except of course we have to play expansion games). % \def\cslet#1#2{% \expandafter\let \csname#1\expandafter\endcsname \csname#2\endcsname } % Trim leading and trailing spaces off a string. % Concepts from aro-bend problem 15 (see CTAN). {\catcode`\@=11 \gdef\eatspaces #1{\expandafter\trim@\expandafter{#1 }} \gdef\trim@ #1{\trim@@ @#1 @ #1 @ @@} \gdef\trim@@ #1@ #2@ #3@@{\trim@@@\empty #2 @} \def\unbrace#1{#1} \unbrace{\gdef\trim@@@ #1 } #2@{#1} } % Trim a single trailing ^^M off a string. {\catcode`\^^M=\other \catcode`\Q=3% \gdef\eatcr #1{\eatcra #1Q^^MQ}% \gdef\eatcra#1^^MQ{\eatcrb#1Q}% \gdef\eatcrb#1Q#2Q{#1}% } % Macro bodies are absorbed as an argument in a context where % all characters are catcode 10, 11 or 12, except \ which is active % (as in normal texinfo). It is necessary to change the definition of \ % to recognize macro arguments; this is the job of \mbodybackslash. % % Non-ASCII encodings make 8-bit characters active, so un-activate % them to avoid their expansion. Must do this non-globally, to % confine the change to the current group. % % It's necessary to have hard CRs when the macro is executed. This is % done by making ^^M (\endlinechar) catcode 12 when reading the macro % body, and then making it the \newlinechar in \scanmacro. % \def\scanctxt{% used as subroutine \catcode`\"=\other \catcode`\+=\other \catcode`\<=\other \catcode`\>=\other \catcode`\@=\other \catcode`\^=\other \catcode`\_=\other \catcode`\|=\other \catcode`\~=\other \ifx\declaredencoding\ascii \else \setnonasciicharscatcodenonglobal\other \fi } \def\scanargctxt{% used for copying and captions, not macros. \scanctxt \catcode`\\=\other \catcode`\^^M=\other } \def\macrobodyctxt{% used for @macro definitions \scanctxt \catcode`\{=\other \catcode`\}=\other \catcode`\^^M=\other \usembodybackslash } \def\macroargctxt{% used when scanning invocations \scanctxt \catcode`\\=0 } % why catcode 0 for \ in the above? To recognize \\ \{ \} as "escapes" % for the single characters \ { }. Thus, we end up with the "commands" % that would be written @\ @{ @} in a Texinfo document. % % We already have @{ and @}. For @\, we define it here, and only for % this purpose, to produce a typewriter backslash (so, the @\ that we % define for @math can't be used with @macro calls): % \def\\{\normalbackslash}% % % We would like to do this for \, too, since that is what makeinfo does. % But it is not possible, because Texinfo already has a command @, for a % cedilla accent. Documents must use @comma{} instead. % % \anythingelse will almost certainly be an error of some kind. % \mbodybackslash is the definition of \ in @macro bodies. % It maps \foo\ => \csname macarg.foo\endcsname => #N % where N is the macro parameter number. % We define \csname macarg.\endcsname to be \realbackslash, so % \\ in macro replacement text gets you a backslash. % {\catcode`@=0 @catcode`@\=@active @gdef@usembodybackslash{@let\=@mbodybackslash} @gdef@mbodybackslash#1\{@csname macarg.#1@endcsname} } \expandafter\def\csname macarg.\endcsname{\realbackslash} \def\margbackslash#1{\char`\#1 } \def\macro{\recursivefalse\parsearg\macroxxx} \def\rmacro{\recursivetrue\parsearg\macroxxx} \def\macroxxx#1{% \getargs{#1}% now \macname is the macname and \argl the arglist \ifx\argl\empty % no arguments \paramno=0\relax \else \expandafter\parsemargdef \argl;% \if\paramno>256\relax \ifx\eTeXversion\thisisundefined \errhelp = \EMsimple \errmessage{You need eTeX to compile a file with macros with more than 256 arguments} \fi \fi \fi \if1\csname ismacro.\the\macname\endcsname \message{Warning: redefining \the\macname}% \else \expandafter\ifx\csname \the\macname\endcsname \relax \else \errmessage{Macro name \the\macname\space already defined}\fi \global\cslet{macsave.\the\macname}{\the\macname}% \global\expandafter\let\csname ismacro.\the\macname\endcsname=1% \addtomacrolist{\the\macname}% \fi \begingroup \macrobodyctxt \ifrecursive \expandafter\parsermacbody \else \expandafter\parsemacbody \fi} \parseargdef\unmacro{% \if1\csname ismacro.#1\endcsname \global\cslet{#1}{macsave.#1}% \global\expandafter\let \csname ismacro.#1\endcsname=0% % Remove the macro name from \macrolist: \begingroup \expandafter\let\csname#1\endcsname \relax \let\definedummyword\unmacrodo \xdef\macrolist{\macrolist}% \endgroup \else \errmessage{Macro #1 not defined}% \fi } % Called by \do from \dounmacro on each macro. The idea is to omit any % macro definitions that have been changed to \relax. % \def\unmacrodo#1{% \ifx #1\relax % remove this \else \noexpand\definedummyword \noexpand#1% \fi } % This makes use of the obscure feature that if the last token of a % is #, then the preceding argument is delimited by % an opening brace, and that opening brace is not consumed. \def\getargs#1{\getargsxxx#1{}} \def\getargsxxx#1#{\getmacname #1 \relax\getmacargs} \def\getmacname#1 #2\relax{\macname={#1}} \def\getmacargs#1{\def\argl{#1}} % For macro processing make @ a letter so that we can make Texinfo private macro names. \edef\texiatcatcode{\the\catcode`\@} \catcode `@=11\relax % Parse the optional {params} list. Set up \paramno and \paramlist % so \defmacro knows what to do. Define \macarg.BLAH for each BLAH % in the params list to some hook where the argument si to be expanded. If % there are less than 10 arguments that hook is to be replaced by ##N where N % is the position in that list, that is to say the macro arguments are to be % defined `a la TeX in the macro body. % % That gets used by \mbodybackslash (above). % % We need to get `macro parameter char #' into several definitions. % The technique used is stolen from LaTeX: let \hash be something % unexpandable, insert that wherever you need a #, and then redefine % it to # just before using the token list produced. % % The same technique is used to protect \eatspaces till just before % the macro is used. % % If there are 10 or more arguments, a different technique is used, where the % hook remains in the body, and when macro is to be expanded the body is % processed again to replace the arguments. % % In that case, the hook is \the\toks N-1, and we simply set \toks N-1 to the % argument N value and then \edef the body (nothing else will expand because of % the catcode regime underwhich the body was input). % % If you compile with TeX (not eTeX), and you have macros with 10 or more % arguments, you need that no macro has more than 256 arguments, otherwise an % error is produced. \def\parsemargdef#1;{% \paramno=0\def\paramlist{}% \let\hash\relax \let\xeatspaces\relax \parsemargdefxxx#1,;,% % In case that there are 10 or more arguments we parse again the arguments % list to set new definitions for the \macarg.BLAH macros corresponding to % each BLAH argument. It was anyhow needed to parse already once this list % in order to count the arguments, and as macros with at most 9 arguments % are by far more frequent than macro with 10 or more arguments, defining % twice the \macarg.BLAH macros does not cost too much processing power. \ifnum\paramno<10\relax\else \paramno0\relax \parsemmanyargdef@@#1,;,% 10 or more arguments \fi } \def\parsemargdefxxx#1,{% \if#1;\let\next=\relax \else \let\next=\parsemargdefxxx \advance\paramno by 1 \expandafter\edef\csname macarg.\eatspaces{#1}\endcsname {\xeatspaces{\hash\the\paramno}}% \edef\paramlist{\paramlist\hash\the\paramno,}% \fi\next} \def\parsemmanyargdef@@#1,{% \if#1;\let\next=\relax \else \let\next=\parsemmanyargdef@@ \edef\tempb{\eatspaces{#1}}% \expandafter\def\expandafter\tempa \expandafter{\csname macarg.\tempb\endcsname}% % Note that we need some extra \noexpand\noexpand, this is because we % don't want \the to be expanded in the \parsermacbody as it uses an % \xdef . \expandafter\edef\tempa {\noexpand\noexpand\noexpand\the\toks\the\paramno}% \advance\paramno by 1\relax \fi\next} % These two commands read recursive and nonrecursive macro bodies. % (They're different since rec and nonrec macros end differently.) % \catcode `\@\texiatcatcode \long\def\parsemacbody#1@end macro% {\xdef\temp{\eatcr{#1}}\endgroup\defmacro}% \long\def\parsermacbody#1@end rmacro% {\xdef\temp{\eatcr{#1}}\endgroup\defmacro}% \catcode `\@=11\relax \let\endargs@\relax \let\nil@\relax \def\nilm@{\nil@}% \long\def\nillm@{\nil@}% % This macro is expanded during the Texinfo macro expansion, not during its % definition. It gets all the arguments values and assigns them to macros % macarg.ARGNAME % % #1 is the macro name % #2 is the list of argument names % #3 is the list of argument values \def\getargvals@#1#2#3{% \def\macargdeflist@{}% \def\saveparamlist@{#2}% Need to keep a copy for parameter expansion. \def\paramlist{#2,\nil@}% \def\macroname{#1}% \begingroup \macroargctxt \def\argvaluelist{#3,\nil@}% \def\@tempa{#3}% \ifx\@tempa\empty \setemptyargvalues@ \else \getargvals@@ \fi } % \def\getargvals@@{% \ifx\paramlist\nilm@ % Some sanity check needed here that \argvaluelist is also empty. \ifx\argvaluelist\nillm@ \else \errhelp = \EMsimple \errmessage{Too many arguments in macro `\macroname'!}% \fi \let\next\macargexpandinbody@ \else \ifx\argvaluelist\nillm@ % No more arguments values passed to macro. Set remaining named-arg % macros to empty. \let\next\setemptyargvalues@ \else % pop current arg name into \@tempb \def\@tempa##1{\pop@{\@tempb}{\paramlist}##1\endargs@}% \expandafter\@tempa\expandafter{\paramlist}% % pop current argument value into \@tempc \def\@tempa##1{\longpop@{\@tempc}{\argvaluelist}##1\endargs@}% \expandafter\@tempa\expandafter{\argvaluelist}% % Here \@tempb is the current arg name and \@tempc is the current arg value. % First place the new argument macro definition into \@tempd \expandafter\macname\expandafter{\@tempc}% \expandafter\let\csname macarg.\@tempb\endcsname\relax \expandafter\def\expandafter\@tempe\expandafter{% \csname macarg.\@tempb\endcsname}% \edef\@tempd{\long\def\@tempe{\the\macname}}% \push@\@tempd\macargdeflist@ \let\next\getargvals@@ \fi \fi \next } \def\push@#1#2{% \expandafter\expandafter\expandafter\def \expandafter\expandafter\expandafter#2% \expandafter\expandafter\expandafter{% \expandafter#1#2}% } % Replace arguments by their values in the macro body, and place the result % in macro \@tempa \def\macvalstoargs@{% % To do this we use the property that token registers that are \the'ed % within an \edef expand only once. So we are going to place all argument % values into respective token registers. % % First we save the token context, and initialize argument numbering. \begingroup \paramno0\relax % Then, for each argument number #N, we place the corresponding argument % value into a new token list register \toks#N \expandafter\putargsintokens@\saveparamlist@,;,% % Then, we expand the body so that argument are replaced by their % values. The trick for values not to be expanded themselves is that they % are within tokens and that tokens expand only once in an \edef . \edef\@tempc{\csname mac.\macroname .body\endcsname}% % Now we restore the token stack pointer to free the token list registers % which we have used, but we make sure that expanded body is saved after % group. \expandafter \endgroup \expandafter\def\expandafter\@tempa\expandafter{\@tempc}% } \def\macargexpandinbody@{% %% Define the named-macro outside of this group and then close this group. \expandafter \endgroup \macargdeflist@ % First the replace in body the macro arguments by their values, the result % is in \@tempa . \macvalstoargs@ % Then we point at the \norecurse or \gobble (for recursive) macro value % with \@tempb . \expandafter\let\expandafter\@tempb\csname mac.\macroname .recurse\endcsname % Depending on whether it is recursive or not, we need some tailing % \egroup . \ifx\@tempb\gobble \let\@tempc\relax \else \let\@tempc\egroup \fi % And now we do the real job: \edef\@tempd{\noexpand\@tempb{\macroname}\noexpand\scanmacro{\@tempa}\@tempc}% \@tempd } \def\putargsintokens@#1,{% \if#1;\let\next\relax \else \let\next\putargsintokens@ % First we allocate the new token list register, and give it a temporary % alias \@tempb . \toksdef\@tempb\the\paramno % Then we place the argument value into that token list register. \expandafter\let\expandafter\@tempa\csname macarg.#1\endcsname \expandafter\@tempb\expandafter{\@tempa}% \advance\paramno by 1\relax \fi \next } % Save the token stack pointer into macro #1 \def\texisavetoksstackpoint#1{\edef#1{\the\@cclvi}} % Restore the token stack pointer from number in macro #1 \def\texirestoretoksstackpoint#1{\expandafter\mathchardef\expandafter\@cclvi#1\relax} % newtoks that can be used non \outer . \def\texinonouternewtoks{\alloc@ 5\toks \toksdef \@cclvi} % Tailing missing arguments are set to empty \def\setemptyargvalues@{% \ifx\paramlist\nilm@ \let\next\macargexpandinbody@ \else \expandafter\setemptyargvaluesparser@\paramlist\endargs@ \let\next\setemptyargvalues@ \fi \next } \def\setemptyargvaluesparser@#1,#2\endargs@{% \expandafter\def\expandafter\@tempa\expandafter{% \expandafter\def\csname macarg.#1\endcsname{}}% \push@\@tempa\macargdeflist@ \def\paramlist{#2}% } % #1 is the element target macro % #2 is the list macro % #3,#4\endargs@ is the list value \def\pop@#1#2#3,#4\endargs@{% \def#1{#3}% \def#2{#4}% } \long\def\longpop@#1#2#3,#4\endargs@{% \long\def#1{#3}% \long\def#2{#4}% } % This defines a Texinfo @macro. There are eight cases: recursive and % nonrecursive macros of zero, one, up to nine, and many arguments. % Much magic with \expandafter here. % \xdef is used so that macro definitions will survive the file % they're defined in; @include reads the file inside a group. % \def\defmacro{% \let\hash=##% convert placeholders to macro parameter chars \ifrecursive \ifcase\paramno % 0 \expandafter\xdef\csname\the\macname\endcsname{% \noexpand\scanmacro{\temp}}% \or % 1 \expandafter\xdef\csname\the\macname\endcsname{% \bgroup\noexpand\macroargctxt \noexpand\braceorline \expandafter\noexpand\csname\the\macname xxx\endcsname}% \expandafter\xdef\csname\the\macname xxx\endcsname##1{% \egroup\noexpand\scanmacro{\temp}}% \else \ifnum\paramno<10\relax % at most 9 \expandafter\xdef\csname\the\macname\endcsname{% \bgroup\noexpand\macroargctxt \noexpand\csname\the\macname xx\endcsname}% \expandafter\xdef\csname\the\macname xx\endcsname##1{% \expandafter\noexpand\csname\the\macname xxx\endcsname ##1,}% \expandafter\expandafter \expandafter\xdef \expandafter\expandafter \csname\the\macname xxx\endcsname \paramlist{\egroup\noexpand\scanmacro{\temp}}% \else % 10 or more \expandafter\xdef\csname\the\macname\endcsname{% \noexpand\getargvals@{\the\macname}{\argl}% }% \global\expandafter\let\csname mac.\the\macname .body\endcsname\temp \global\expandafter\let\csname mac.\the\macname .recurse\endcsname\gobble \fi \fi \else \ifcase\paramno % 0 \expandafter\xdef\csname\the\macname\endcsname{% \noexpand\norecurse{\the\macname}% \noexpand\scanmacro{\temp}\egroup}% \or % 1 \expandafter\xdef\csname\the\macname\endcsname{% \bgroup\noexpand\macroargctxt \noexpand\braceorline \expandafter\noexpand\csname\the\macname xxx\endcsname}% \expandafter\xdef\csname\the\macname xxx\endcsname##1{% \egroup \noexpand\norecurse{\the\macname}% \noexpand\scanmacro{\temp}\egroup}% \else % at most 9 \ifnum\paramno<10\relax \expandafter\xdef\csname\the\macname\endcsname{% \bgroup\noexpand\macroargctxt \expandafter\noexpand\csname\the\macname xx\endcsname}% \expandafter\xdef\csname\the\macname xx\endcsname##1{% \expandafter\noexpand\csname\the\macname xxx\endcsname ##1,}% \expandafter\expandafter \expandafter\xdef \expandafter\expandafter \csname\the\macname xxx\endcsname \paramlist{% \egroup \noexpand\norecurse{\the\macname}% \noexpand\scanmacro{\temp}\egroup}% \else % 10 or more: \expandafter\xdef\csname\the\macname\endcsname{% \noexpand\getargvals@{\the\macname}{\argl}% }% \global\expandafter\let\csname mac.\the\macname .body\endcsname\temp \global\expandafter\let\csname mac.\the\macname .recurse\endcsname\norecurse \fi \fi \fi} \catcode `\@\texiatcatcode\relax \def\norecurse#1{\bgroup\cslet{#1}{macsave.#1}} % \braceorline decides whether the next nonwhitespace character is a % {. If so it reads up to the closing }, if not, it reads the whole % line. Whatever was read is then fed to the next control sequence % as an argument (by \parsebrace or \parsearg). % \def\braceorline#1{\let\macnamexxx=#1\futurelet\nchar\braceorlinexxx} \def\braceorlinexxx{% \ifx\nchar\bgroup\else \expandafter\parsearg \fi \macnamexxx} % @alias. % We need some trickery to remove the optional spaces around the equal % sign. Make them active and then expand them all to nothing. % \def\alias{\parseargusing\obeyspaces\aliasxxx} \def\aliasxxx #1{\aliasyyy#1\relax} \def\aliasyyy #1=#2\relax{% {% \expandafter\let\obeyedspace=\empty \addtomacrolist{#1}% \xdef\next{\global\let\makecsname{#1}=\makecsname{#2}}% }% \next } \message{cross references,} \newwrite\auxfile \newif\ifhavexrefs % True if xref values are known. \newif\ifwarnedxrefs % True if we warned once that they aren't known. % @inforef is relatively simple. \def\inforef #1{\inforefzzz #1,,,,**} \def\inforefzzz #1,#2,#3,#4**{% \putwordSee{} \putwordInfo{} \putwordfile{} \file{\ignorespaces #3{}}, node \samp{\ignorespaces#1{}}} % @node's only job in TeX is to define \lastnode, which is used in % cross-references. The @node line might or might not have commas, and % might or might not have spaces before the first comma, like: % @node foo , bar , ... % We don't want such trailing spaces in the node name. % \parseargdef\node{\checkenv{}\donode #1 ,\finishnodeparse} % % also remove a trailing comma, in case of something like this: % @node Help-Cross, , , Cross-refs \def\donode#1 ,#2\finishnodeparse{\dodonode #1,\finishnodeparse} \def\dodonode#1,#2\finishnodeparse{\gdef\lastnode{#1}} \let\nwnode=\node \let\lastnode=\empty % Write a cross-reference definition for the current node. #1 is the % type (Ynumbered, Yappendix, Ynothing). % \def\donoderef#1{% \ifx\lastnode\empty\else \setref{\lastnode}{#1}% \global\let\lastnode=\empty \fi } % @anchor{NAME} -- define xref target at arbitrary point. % \newcount\savesfregister % \def\savesf{\relax \ifhmode \savesfregister=\spacefactor \fi} \def\restoresf{\relax \ifhmode \spacefactor=\savesfregister \fi} \def\anchor#1{\savesf \setref{#1}{Ynothing}\restoresf \ignorespaces} % \setref{NAME}{SNT} defines a cross-reference point NAME (a node or an % anchor), which consists of three parts: % 1) NAME-title - the current sectioning name taken from \lastsection, % or the anchor name. % 2) NAME-snt - section number and type, passed as the SNT arg, or % empty for anchors. % 3) NAME-pg - the page number. % % This is called from \donoderef, \anchor, and \dofloat. In the case of % floats, there is an additional part, which is not written here: % 4) NAME-lof - the text as it should appear in a @listoffloats. % \def\setref#1#2{% \pdfmkdest{#1}% \iflinks {% \atdummies % preserve commands, but don't expand them \edef\writexrdef##1##2{% \write\auxfile{@xrdef{#1-% #1 of \setref, expanded by the \edef ##1}{##2}}% these are parameters of \writexrdef }% \toks0 = \expandafter{\lastsection}% \immediate \writexrdef{title}{\the\toks0 }% \immediate \writexrdef{snt}{\csname #2\endcsname}% \Ynumbered etc. \safewhatsit{\writexrdef{pg}{\folio}}% will be written later, at \shipout }% \fi } % @xrefautosectiontitle on|off says whether @section(ing) names are used % automatically in xrefs, if the third arg is not explicitly specified. % This was provided as a "secret" @set xref-automatic-section-title % variable, now it's official. % \parseargdef\xrefautomaticsectiontitle{% \def\temp{#1}% \ifx\temp\onword \expandafter\let\csname SETxref-automatic-section-title\endcsname = \empty \else\ifx\temp\offword \expandafter\let\csname SETxref-automatic-section-title\endcsname = \relax \else \errhelp = \EMsimple \errmessage{Unknown @xrefautomaticsectiontitle value `\temp', must be on|off}% \fi\fi } % @xref, @pxref, and @ref generate cross-references. For \xrefX, #1 is % the node name, #2 the name of the Info cross-reference, #3 the printed % node name, #4 the name of the Info file, #5 the name of the printed % manual. All but the node name can be omitted. % \def\pxref#1{\putwordsee{} \xrefX[#1,,,,,,,]} \def\xref#1{\putwordSee{} \xrefX[#1,,,,,,,]} \def\ref#1{\xrefX[#1,,,,,,,]} % \newbox\topbox \newbox\printedrefnamebox \newbox\printedmanualbox % \def\xrefX[#1,#2,#3,#4,#5,#6]{\begingroup \unsepspaces % \def\printedrefname{\ignorespaces #3}% \setbox\printedrefnamebox = \hbox{\printedrefname\unskip}% % \def\printedmanual{\ignorespaces #5}% \setbox\printedmanualbox = \hbox{\printedmanual\unskip}% % % If the printed reference name (arg #3) was not explicitly given in % the @xref, figure out what we want to use. \ifdim \wd\printedrefnamebox = 0pt % No printed node name was explicitly given. \expandafter\ifx\csname SETxref-automatic-section-title\endcsname \relax % Not auto section-title: use node name inside the square brackets. \def\printedrefname{\ignorespaces #1}% \else % Auto section-title: use chapter/section title inside % the square brackets if we have it. \ifdim \wd\printedmanualbox > 0pt % It is in another manual, so we don't have it; use node name. \def\printedrefname{\ignorespaces #1}% \else \ifhavexrefs % We (should) know the real title if we have the xref values. \def\printedrefname{\refx{#1-title}{}}% \else % Otherwise just copy the Info node name. \def\printedrefname{\ignorespaces #1}% \fi% \fi \fi \fi % % Make link in pdf output. \ifpdf {\indexnofonts \turnoffactive \makevalueexpandable % This expands tokens, so do it after making catcode changes, so _ % etc. don't get their TeX definitions. \getfilename{#4}% % \edef\pdfxrefdest{#1}% \txiescapepdf\pdfxrefdest % \leavevmode \startlink attr{/Border [0 0 0]}% \ifnum\filenamelength>0 goto file{\the\filename.pdf} name{\pdfxrefdest}% \else goto name{\pdfmkpgn{\pdfxrefdest}}% \fi }% \setcolor{\linkcolor}% \fi % % Float references are printed completely differently: "Figure 1.2" % instead of "[somenode], p.3". We distinguish them by the % LABEL-title being set to a magic string. {% % Have to otherify everything special to allow the \csname to % include an _ in the xref name, etc. \indexnofonts \turnoffactive \expandafter\global\expandafter\let\expandafter\Xthisreftitle \csname XR#1-title\endcsname }% \iffloat\Xthisreftitle % If the user specified the print name (third arg) to the ref, % print it instead of our usual "Figure 1.2". \ifdim\wd\printedrefnamebox = 0pt \refx{#1-snt}{}% \else \printedrefname \fi % % if the user also gave the printed manual name (fifth arg), append % "in MANUALNAME". \ifdim \wd\printedmanualbox > 0pt \space \putwordin{} \cite{\printedmanual}% \fi \else % node/anchor (non-float) references. % % If we use \unhbox to print the node names, TeX does not insert % empty discretionaries after hyphens, which means that it will not % find a line break at a hyphen in a node names. Since some manuals % are best written with fairly long node names, containing hyphens, % this is a loss. Therefore, we give the text of the node name % again, so it is as if TeX is seeing it for the first time. % % Cross-manual reference. Only include the "Section ``foo'' in" if % the foo is neither missing or Top. Thus, @xref{,,,foo,The Foo Manual} % outputs simply "see The Foo Manual". \ifdim \wd\printedmanualbox > 0pt % What is the 7sp about? The idea is that we also want to omit % the Section part if we would be printing "Top", since they are % clearly trying to refer to the whole manual. But, this being % TeX, we can't easily compare strings while ignoring the possible % spaces before and after in the input. By adding the arbitrary % 7sp, we make it much less likely that a real node name would % happen to have the same width as "Top" (e.g., in a monospaced font). % I hope it will never happen in practice. % % For the same basic reason, we retypeset the "Top" at every % reference, since the current font is indeterminate. % \setbox\topbox = \hbox{Top\kern7sp}% \setbox2 = \hbox{\ignorespaces \printedrefname \unskip \kern7sp}% \ifdim \wd2 > 7sp \ifdim \wd2 = \wd\topbox \else \putwordSection{} ``\printedrefname'' \putwordin{}\space \fi \fi \cite{\printedmanual}% \else % Reference in this manual. % % _ (for example) has to be the character _ for the purposes of the % control sequence corresponding to the node, but it has to expand % into the usual \leavevmode...\vrule stuff for purposes of % printing. So we \turnoffactive for the \refx-snt, back on for the % printing, back off for the \refx-pg. {\turnoffactive % Only output a following space if the -snt ref is nonempty; for % @unnumbered and @anchor, it won't be. \setbox2 = \hbox{\ignorespaces \refx{#1-snt}{}}% \ifdim \wd2 > 0pt \refx{#1-snt}\space\fi }% % output the `[mynode]' via the macro below so it can be overridden. \xrefprintnodename\printedrefname % % But we always want a comma and a space: ,\space % % output the `page 3'. \turnoffactive \putwordpage\tie\refx{#1-pg}{}% \fi \fi \endlink \endgroup} % This macro is called from \xrefX for the `[nodename]' part of xref % output. It's a separate macro only so it can be changed more easily, % since square brackets don't work well in some documents. Particularly % one that Bob is working on :). % \def\xrefprintnodename#1{[#1]} % Things referred to by \setref. % \def\Ynothing{} \def\Yomitfromtoc{} \def\Ynumbered{% \ifnum\secno=0 \putwordChapter@tie \the\chapno \else \ifnum\subsecno=0 \putwordSection@tie \the\chapno.\the\secno \else \ifnum\subsubsecno=0 \putwordSection@tie \the\chapno.\the\secno.\the\subsecno \else \putwordSection@tie \the\chapno.\the\secno.\the\subsecno.\the\subsubsecno \fi\fi\fi } \def\Yappendix{% \ifnum\secno=0 \putwordAppendix@tie @char\the\appendixno{}% \else \ifnum\subsecno=0 \putwordSection@tie @char\the\appendixno.\the\secno \else \ifnum\subsubsecno=0 \putwordSection@tie @char\the\appendixno.\the\secno.\the\subsecno \else \putwordSection@tie @char\the\appendixno.\the\secno.\the\subsecno.\the\subsubsecno \fi\fi\fi } % Define \refx{NAME}{SUFFIX} to reference a cross-reference string named NAME. % If its value is nonempty, SUFFIX is output afterward. % \def\refx#1#2{% {% \indexnofonts \otherbackslash \expandafter\global\expandafter\let\expandafter\thisrefX \csname XR#1\endcsname }% \ifx\thisrefX\relax % If not defined, say something at least. \angleleft un\-de\-fined\angleright \iflinks \ifhavexrefs {\toks0 = {#1}% avoid expansion of possibly-complex value \message{\linenumber Undefined cross reference `\the\toks0'.}}% \else \ifwarnedxrefs\else \global\warnedxrefstrue \message{Cross reference values unknown; you must run TeX again.}% \fi \fi \fi \else % It's defined, so just use it. \thisrefX \fi #2% Output the suffix in any case. } % This is the macro invoked by entries in the aux file. Usually it's % just a \def (we prepend XR to the control sequence name to avoid % collisions). But if this is a float type, we have more work to do. % \def\xrdef#1#2{% {% The node name might contain 8-bit characters, which in our current % implementation are changed to commands like @'e. Don't let these % mess up the control sequence name. \indexnofonts \turnoffactive \xdef\safexrefname{#1}% }% % \expandafter\gdef\csname XR\safexrefname\endcsname{#2}% remember this xref % % Was that xref control sequence that we just defined for a float? \expandafter\iffloat\csname XR\safexrefname\endcsname % it was a float, and we have the (safe) float type in \iffloattype. \expandafter\let\expandafter\floatlist \csname floatlist\iffloattype\endcsname % % Is this the first time we've seen this float type? \expandafter\ifx\floatlist\relax \toks0 = {\do}% yes, so just \do \else % had it before, so preserve previous elements in list. \toks0 = \expandafter{\floatlist\do}% \fi % % Remember this xref in the control sequence \floatlistFLOATTYPE, % for later use in \listoffloats. \expandafter\xdef\csname floatlist\iffloattype\endcsname{\the\toks0 {\safexrefname}}% \fi } % Read the last existing aux file, if any. No error if none exists. % \def\tryauxfile{% \openin 1 \jobname.aux \ifeof 1 \else \readdatafile{aux}% \global\havexrefstrue \fi \closein 1 } \def\setupdatafile{% \catcode`\^^@=\other \catcode`\^^A=\other \catcode`\^^B=\other \catcode`\^^C=\other \catcode`\^^D=\other \catcode`\^^E=\other \catcode`\^^F=\other \catcode`\^^G=\other \catcode`\^^H=\other \catcode`\^^K=\other \catcode`\^^L=\other \catcode`\^^N=\other \catcode`\^^P=\other \catcode`\^^Q=\other \catcode`\^^R=\other \catcode`\^^S=\other \catcode`\^^T=\other \catcode`\^^U=\other \catcode`\^^V=\other \catcode`\^^W=\other \catcode`\^^X=\other \catcode`\^^Z=\other \catcode`\^^[=\other \catcode`\^^\=\other \catcode`\^^]=\other \catcode`\^^^=\other \catcode`\^^_=\other % It was suggested to set the catcode of ^ to 7, which would allow ^^e4 etc. % in xref tags, i.e., node names. But since ^^e4 notation isn't % supported in the main text, it doesn't seem desirable. Furthermore, % that is not enough: for node names that actually contain a ^ % character, we would end up writing a line like this: 'xrdef {'hat % b-title}{'hat b} and \xrdef does a \csname...\endcsname on the first % argument, and \hat is not an expandable control sequence. It could % all be worked out, but why? Either we support ^^ or we don't. % % The other change necessary for this was to define \auxhat: % \def\auxhat{\def^{'hat }}% extra space so ok if followed by letter % and then to call \auxhat in \setq. % \catcode`\^=\other % % Special characters. Should be turned off anyway, but... \catcode`\~=\other \catcode`\[=\other \catcode`\]=\other \catcode`\"=\other \catcode`\_=\other \catcode`\|=\other \catcode`\<=\other \catcode`\>=\other \catcode`\$=\other \catcode`\#=\other \catcode`\&=\other \catcode`\%=\other \catcode`+=\other % avoid \+ for paranoia even though we've turned it off % % This is to support \ in node names and titles, since the \ % characters end up in a \csname. It's easier than % leaving it active and making its active definition an actual \ % character. What I don't understand is why it works in the *value* % of the xrdef. Seems like it should be a catcode12 \, and that % should not typeset properly. But it works, so I'm moving on for % now. --karl, 15jan04. \catcode`\\=\other % % Make the characters 128-255 be printing characters. {% \count1=128 \def\loop{% \catcode\count1=\other \advance\count1 by 1 \ifnum \count1<256 \loop \fi }% }% % % @ is our escape character in .aux files, and we need braces. \catcode`\{=1 \catcode`\}=2 \catcode`\@=0 } \def\readdatafile#1{% \begingroup \setupdatafile \input\jobname.#1 \endgroup} \message{insertions,} % including footnotes. \newcount \footnoteno % The trailing space in the following definition for supereject is % vital for proper filling; pages come out unaligned when you do a % pagealignmacro call if that space before the closing brace is % removed. (Generally, numeric constants should always be followed by a % space to prevent strange expansion errors.) \def\supereject{\par\penalty -20000\footnoteno =0 } % @footnotestyle is meaningful for Info output only. \let\footnotestyle=\comment {\catcode `\@=11 % % Auto-number footnotes. Otherwise like plain. \gdef\footnote{% \let\indent=\ptexindent \let\noindent=\ptexnoindent \global\advance\footnoteno by \@ne \edef\thisfootno{$^{\the\footnoteno}$}% % % In case the footnote comes at the end of a sentence, preserve the % extra spacing after we do the footnote number. \let\@sf\empty \ifhmode\edef\@sf{\spacefactor\the\spacefactor}\ptexslash\fi % % Remove inadvertent blank space before typesetting the footnote number. \unskip \thisfootno\@sf \dofootnote }% % Don't bother with the trickery in plain.tex to not require the % footnote text as a parameter. Our footnotes don't need to be so general. % % Oh yes, they do; otherwise, @ifset (and anything else that uses % \parseargline) fails inside footnotes because the tokens are fixed when % the footnote is read. --karl, 16nov96. % \gdef\dofootnote{% \insert\footins\bgroup % We want to typeset this text as a normal paragraph, even if the % footnote reference occurs in (for example) a display environment. % So reset some parameters. \hsize=\pagewidth \interlinepenalty\interfootnotelinepenalty \splittopskip\ht\strutbox % top baseline for broken footnotes \splitmaxdepth\dp\strutbox \floatingpenalty\@MM \leftskip\z@skip \rightskip\z@skip \spaceskip\z@skip \xspaceskip\z@skip \parindent\defaultparindent % \smallfonts \rm % % Because we use hanging indentation in footnotes, a @noindent appears % to exdent this text, so make it be a no-op. makeinfo does not use % hanging indentation so @noindent can still be needed within footnote % text after an @example or the like (not that this is good style). \let\noindent = \relax % % Hang the footnote text off the number. Use \everypar in case the % footnote extends for more than one paragraph. \everypar = {\hang}% \textindent{\thisfootno}% % % Don't crash into the line above the footnote text. Since this % expands into a box, it must come within the paragraph, lest it % provide a place where TeX can split the footnote. \footstrut % % Invoke rest of plain TeX footnote routine. \futurelet\next\fo@t } }%end \catcode `\@=11 % In case a @footnote appears in a vbox, save the footnote text and create % the real \insert just after the vbox finished. Otherwise, the insertion % would be lost. % Similarly, if a @footnote appears inside an alignment, save the footnote % text to a box and make the \insert when a row of the table is finished. % And the same can be done for other insert classes. --kasal, 16nov03. % Replace the \insert primitive by a cheating macro. % Deeper inside, just make sure that the saved insertions are not spilled % out prematurely. % \def\startsavinginserts{% \ifx \insert\ptexinsert \let\insert\saveinsert \else \let\checkinserts\relax \fi } % This \insert replacement works for both \insert\footins{foo} and % \insert\footins\bgroup foo\egroup, but it doesn't work for \insert27{foo}. % \def\saveinsert#1{% \edef\next{\noexpand\savetobox \makeSAVEname#1}% \afterassignment\next % swallow the left brace \let\temp = } \def\makeSAVEname#1{\makecsname{SAVE\expandafter\gobble\string#1}} \def\savetobox#1{\global\setbox#1 = \vbox\bgroup \unvbox#1} \def\checksaveins#1{\ifvoid#1\else \placesaveins#1\fi} \def\placesaveins#1{% \ptexinsert \csname\expandafter\gobblesave\string#1\endcsname {\box#1}% } % eat @SAVE -- beware, all of them have catcode \other: { \def\dospecials{\do S\do A\do V\do E} \uncatcodespecials % ;-) \gdef\gobblesave @SAVE{} } % initialization: \def\newsaveins #1{% \edef\next{\noexpand\newsaveinsX \makeSAVEname#1}% \next } \def\newsaveinsX #1{% \csname newbox\endcsname #1% \expandafter\def\expandafter\checkinserts\expandafter{\checkinserts \checksaveins #1}% } % initialize: \let\checkinserts\empty \newsaveins\footins \newsaveins\margin % @image. We use the macros from epsf.tex to support this. % If epsf.tex is not installed and @image is used, we complain. % % Check for and read epsf.tex up front. If we read it only at @image % time, we might be inside a group, and then its definitions would get % undone and the next image would fail. \openin 1 = epsf.tex \ifeof 1 \else % Do not bother showing banner with epsf.tex v2.7k (available in % doc/epsf.tex and on ctan). \def\epsfannounce{\toks0 = }% \input epsf.tex \fi \closein 1 % % We will only complain once about lack of epsf.tex. \newif\ifwarnednoepsf \newhelp\noepsfhelp{epsf.tex must be installed for images to work. It is also included in the Texinfo distribution, or you can get it from ftp://tug.org/tex/epsf.tex.} % \def\image#1{% \ifx\epsfbox\thisisundefined \ifwarnednoepsf \else \errhelp = \noepsfhelp \errmessage{epsf.tex not found, images will be ignored}% \global\warnednoepsftrue \fi \else \imagexxx #1,,,,,\finish \fi } % % Arguments to @image: % #1 is (mandatory) image filename; we tack on .eps extension. % #2 is (optional) width, #3 is (optional) height. % #4 is (ignored optional) html alt text. % #5 is (ignored optional) extension. % #6 is just the usual extra ignored arg for parsing stuff. \newif\ifimagevmode \def\imagexxx#1,#2,#3,#4,#5,#6\finish{\begingroup \catcode`\^^M = 5 % in case we're inside an example \normalturnoffactive % allow _ et al. in names % If the image is by itself, center it. \ifvmode \imagevmodetrue \else \ifx\centersub\centerV % for @center @image, we need a vbox so we can have our vertical space \imagevmodetrue \vbox\bgroup % vbox has better behavior than vtop herev \fi\fi % \ifimagevmode \nobreak\medskip % Usually we'll have text after the image which will insert % \parskip glue, so insert it here too to equalize the space % above and below. \nobreak\vskip\parskip \nobreak \fi % % Leave vertical mode so that indentation from an enclosing % environment such as @quotation is respected. % However, if we're at the top level, we don't want the % normal paragraph indentation. % On the other hand, if we are in the case of @center @image, we don't % want to start a paragraph, which will create a hsize-width box and % eradicate the centering. \ifx\centersub\centerV\else \noindent \fi % % Output the image. \ifpdf \dopdfimage{#1}{#2}{#3}% \else % \epsfbox itself resets \epsf?size at each figure. \setbox0 = \hbox{\ignorespaces #2}\ifdim\wd0 > 0pt \epsfxsize=#2\relax \fi \setbox0 = \hbox{\ignorespaces #3}\ifdim\wd0 > 0pt \epsfysize=#3\relax \fi \epsfbox{#1.eps}% \fi % \ifimagevmode \medskip % space after a standalone image \fi \ifx\centersub\centerV \egroup \fi \endgroup} % @float FLOATTYPE,LABEL,LOC ... @end float for displayed figures, tables, % etc. We don't actually implement floating yet, we always include the % float "here". But it seemed the best name for the future. % \envparseargdef\float{\eatcommaspace\eatcommaspace\dofloat#1, , ,\finish} % There may be a space before second and/or third parameter; delete it. \def\eatcommaspace#1, {#1,} % #1 is the optional FLOATTYPE, the text label for this float, typically % "Figure", "Table", "Example", etc. Can't contain commas. If omitted, % this float will not be numbered and cannot be referred to. % % #2 is the optional xref label. Also must be present for the float to % be referable. % % #3 is the optional positioning argument; for now, it is ignored. It % will somehow specify the positions allowed to float to (here, top, bottom). % % We keep a separate counter for each FLOATTYPE, which we reset at each % chapter-level command. \let\resetallfloatnos=\empty % \def\dofloat#1,#2,#3,#4\finish{% \let\thiscaption=\empty \let\thisshortcaption=\empty % % don't lose footnotes inside @float. % % BEWARE: when the floats start float, we have to issue warning whenever an % insert appears inside a float which could possibly float. --kasal, 26may04 % \startsavinginserts % % We can't be used inside a paragraph. \par % \vtop\bgroup \def\floattype{#1}% \def\floatlabel{#2}% \def\floatloc{#3}% we do nothing with this yet. % \ifx\floattype\empty \let\safefloattype=\empty \else {% % the floattype might have accents or other special characters, % but we need to use it in a control sequence name. \indexnofonts \turnoffactive \xdef\safefloattype{\floattype}% }% \fi % % If label is given but no type, we handle that as the empty type. \ifx\floatlabel\empty \else % We want each FLOATTYPE to be numbered separately (Figure 1, % Table 1, Figure 2, ...). (And if no label, no number.) % \expandafter\getfloatno\csname\safefloattype floatno\endcsname \global\advance\floatno by 1 % {% % This magic value for \lastsection is output by \setref as the % XREFLABEL-title value. \xrefX uses it to distinguish float % labels (which have a completely different output format) from % node and anchor labels. And \xrdef uses it to construct the % lists of floats. % \edef\lastsection{\floatmagic=\safefloattype}% \setref{\floatlabel}{Yfloat}% }% \fi % % start with \parskip glue, I guess. \vskip\parskip % % Don't suppress indentation if a float happens to start a section. \restorefirstparagraphindent } % we have these possibilities: % @float Foo,lbl & @caption{Cap}: Foo 1.1: Cap % @float Foo,lbl & no caption: Foo 1.1 % @float Foo & @caption{Cap}: Foo: Cap % @float Foo & no caption: Foo % @float ,lbl & Caption{Cap}: 1.1: Cap % @float ,lbl & no caption: 1.1 % @float & @caption{Cap}: Cap % @float & no caption: % \def\Efloat{% \let\floatident = \empty % % In all cases, if we have a float type, it comes first. \ifx\floattype\empty \else \def\floatident{\floattype}\fi % % If we have an xref label, the number comes next. \ifx\floatlabel\empty \else \ifx\floattype\empty \else % if also had float type, need tie first. \appendtomacro\floatident{\tie}% \fi % the number. \appendtomacro\floatident{\chaplevelprefix\the\floatno}% \fi % % Start the printed caption with what we've constructed in % \floatident, but keep it separate; we need \floatident again. \let\captionline = \floatident % \ifx\thiscaption\empty \else \ifx\floatident\empty \else \appendtomacro\captionline{: }% had ident, so need a colon between \fi % % caption text. \appendtomacro\captionline{\scanexp\thiscaption}% \fi % % If we have anything to print, print it, with space before. % Eventually this needs to become an \insert. \ifx\captionline\empty \else \vskip.5\parskip \captionline % % Space below caption. \vskip\parskip \fi % % If have an xref label, write the list of floats info. Do this % after the caption, to avoid chance of it being a breakpoint. \ifx\floatlabel\empty \else % Write the text that goes in the lof to the aux file as % \floatlabel-lof. Besides \floatident, we include the short % caption if specified, else the full caption if specified, else nothing. {% \atdummies % % since we read the caption text in the macro world, where ^^M % is turned into a normal character, we have to scan it back, so % we don't write the literal three characters "^^M" into the aux file. \scanexp{% \xdef\noexpand\gtemp{% \ifx\thisshortcaption\empty \thiscaption \else \thisshortcaption \fi }% }% \immediate\write\auxfile{@xrdef{\floatlabel-lof}{\floatident \ifx\gtemp\empty \else : \gtemp \fi}}% }% \fi \egroup % end of \vtop % % place the captured inserts % % BEWARE: when the floats start floating, we have to issue warning % whenever an insert appears inside a float which could possibly % float. --kasal, 26may04 % \checkinserts } % Append the tokens #2 to the definition of macro #1, not expanding either. % \def\appendtomacro#1#2{% \expandafter\def\expandafter#1\expandafter{#1#2}% } % @caption, @shortcaption % \def\caption{\docaption\thiscaption} \def\shortcaption{\docaption\thisshortcaption} \def\docaption{\checkenv\float \bgroup\scanargctxt\defcaption} \def\defcaption#1#2{\egroup \def#1{#2}} % The parameter is the control sequence identifying the counter we are % going to use. Create it if it doesn't exist and assign it to \floatno. \def\getfloatno#1{% \ifx#1\relax % Haven't seen this figure type before. \csname newcount\endcsname #1% % % Remember to reset this floatno at the next chap. \expandafter\gdef\expandafter\resetallfloatnos \expandafter{\resetallfloatnos #1=0 }% \fi \let\floatno#1% } % \setref calls this to get the XREFLABEL-snt value. We want an @xref % to the FLOATLABEL to expand to "Figure 3.1". We call \setref when we % first read the @float command. % \def\Yfloat{\floattype@tie \chaplevelprefix\the\floatno}% % Magic string used for the XREFLABEL-title value, so \xrefX can % distinguish floats from other xref types. \def\floatmagic{!!float!!} % #1 is the control sequence we are passed; we expand into a conditional % which is true if #1 represents a float ref. That is, the magic % \lastsection value which we \setref above. % \def\iffloat#1{\expandafter\doiffloat#1==\finish} % % #1 is (maybe) the \floatmagic string. If so, #2 will be the % (safe) float type for this float. We set \iffloattype to #2. % \def\doiffloat#1=#2=#3\finish{% \def\temp{#1}% \def\iffloattype{#2}% \ifx\temp\floatmagic } % @listoffloats FLOATTYPE - print a list of floats like a table of contents. % \parseargdef\listoffloats{% \def\floattype{#1}% floattype {% % the floattype might have accents or other special characters, % but we need to use it in a control sequence name. \indexnofonts \turnoffactive \xdef\safefloattype{\floattype}% }% % % \xrdef saves the floats as a \do-list in \floatlistSAFEFLOATTYPE. \expandafter\ifx\csname floatlist\safefloattype\endcsname \relax \ifhavexrefs % if the user said @listoffloats foo but never @float foo. \message{\linenumber No `\safefloattype' floats to list.}% \fi \else \begingroup \leftskip=\tocindent % indent these entries like a toc \let\do=\listoffloatsdo \csname floatlist\safefloattype\endcsname \endgroup \fi } % This is called on each entry in a list of floats. We're passed the % xref label, in the form LABEL-title, which is how we save it in the % aux file. We strip off the -title and look up \XRLABEL-lof, which % has the text we're supposed to typeset here. % % Figures without xref labels will not be included in the list (since % they won't appear in the aux file). % \def\listoffloatsdo#1{\listoffloatsdoentry#1\finish} \def\listoffloatsdoentry#1-title\finish{{% % Can't fully expand XR#1-lof because it can contain anything. Just % pass the control sequence. On the other hand, XR#1-pg is just the % page number, and we want to fully expand that so we can get a link % in pdf output. \toksA = \expandafter{\csname XR#1-lof\endcsname}% % % use the same \entry macro we use to generate the TOC and index. \edef\writeentry{\noexpand\entry{\the\toksA}{\csname XR#1-pg\endcsname}}% \writeentry }} \message{localization,} % For single-language documents, @documentlanguage is usually given very % early, just after @documentencoding. Single argument is the language % (de) or locale (de_DE) abbreviation. % { \catcode`\_ = \active \globaldefs=1 \parseargdef\documentlanguage{\begingroup \let_=\normalunderscore % normal _ character for filenames \tex % read txi-??.tex file in plain TeX. % Read the file by the name they passed if it exists. \openin 1 txi-#1.tex \ifeof 1 \documentlanguagetrywithoutunderscore{#1_\finish}% \else \globaldefs = 1 % everything in the txi-LL files needs to persist \input txi-#1.tex \fi \closein 1 \endgroup % end raw TeX \endgroup} % % If they passed de_DE, and txi-de_DE.tex doesn't exist, % try txi-de.tex. % \gdef\documentlanguagetrywithoutunderscore#1_#2\finish{% \openin 1 txi-#1.tex \ifeof 1 \errhelp = \nolanghelp \errmessage{Cannot read language file txi-#1.tex}% \else \globaldefs = 1 % everything in the txi-LL files needs to persist \input txi-#1.tex \fi \closein 1 } }% end of special _ catcode % \newhelp\nolanghelp{The given language definition file cannot be found or is empty. Maybe you need to install it? Putting it in the current directory should work if nowhere else does.} % This macro is called from txi-??.tex files; the first argument is the % \language name to set (without the "\lang@" prefix), the second and % third args are \{left,right}hyphenmin. % % The language names to pass are determined when the format is built. % See the etex.log file created at that time, e.g., % /usr/local/texlive/2008/texmf-var/web2c/pdftex/etex.log. % % With TeX Live 2008, etex now includes hyphenation patterns for all % available languages. This means we can support hyphenation in % Texinfo, at least to some extent. (This still doesn't solve the % accented characters problem.) % \catcode`@=11 \def\txisetlanguage#1#2#3{% % do not set the language if the name is undefined in the current TeX. \expandafter\ifx\csname lang@#1\endcsname \relax \message{no patterns for #1}% \else \global\language = \csname lang@#1\endcsname \fi % but there is no harm in adjusting the hyphenmin values regardless. \global\lefthyphenmin = #2\relax \global\righthyphenmin = #3\relax } % Helpers for encodings. % Set the catcode of characters 128 through 255 to the specified number. % \def\setnonasciicharscatcode#1{% \count255=128 \loop\ifnum\count255<256 \global\catcode\count255=#1\relax \advance\count255 by 1 \repeat } \def\setnonasciicharscatcodenonglobal#1{% \count255=128 \loop\ifnum\count255<256 \catcode\count255=#1\relax \advance\count255 by 1 \repeat } % @documentencoding sets the definition of non-ASCII characters % according to the specified encoding. % \parseargdef\documentencoding{% % Encoding being declared for the document. \def\declaredencoding{\csname #1.enc\endcsname}% % % Supported encodings: names converted to tokens in order to be able % to compare them with \ifx. \def\ascii{\csname US-ASCII.enc\endcsname}% \def\latnine{\csname ISO-8859-15.enc\endcsname}% \def\latone{\csname ISO-8859-1.enc\endcsname}% \def\lattwo{\csname ISO-8859-2.enc\endcsname}% \def\utfeight{\csname UTF-8.enc\endcsname}% % \ifx \declaredencoding \ascii \asciichardefs % \else \ifx \declaredencoding \lattwo \setnonasciicharscatcode\active \lattwochardefs % \else \ifx \declaredencoding \latone \setnonasciicharscatcode\active \latonechardefs % \else \ifx \declaredencoding \latnine \setnonasciicharscatcode\active \latninechardefs % \else \ifx \declaredencoding \utfeight \setnonasciicharscatcode\active \utfeightchardefs % \else \message{Unknown document encoding #1, ignoring.}% % \fi % utfeight \fi % latnine \fi % latone \fi % lattwo \fi % ascii } % A message to be logged when using a character that isn't available % the default font encoding (OT1). % \def\missingcharmsg#1{\message{Character missing in OT1 encoding: #1.}} % Take account of \c (plain) vs. \, (Texinfo) difference. \def\cedilla#1{\ifx\c\ptexc\c{#1}\else\,{#1}\fi} % First, make active non-ASCII characters in order for them to be % correctly categorized when TeX reads the replacement text of % macros containing the character definitions. \setnonasciicharscatcode\active % % Latin1 (ISO-8859-1) character definitions. \def\latonechardefs{% \gdef^^a0{\tie} \gdef^^a1{\exclamdown} \gdef^^a2{\missingcharmsg{CENT SIGN}} \gdef^^a3{{\pounds}} \gdef^^a4{\missingcharmsg{CURRENCY SIGN}} \gdef^^a5{\missingcharmsg{YEN SIGN}} \gdef^^a6{\missingcharmsg{BROKEN BAR}} \gdef^^a7{\S} \gdef^^a8{\"{}} \gdef^^a9{\copyright} \gdef^^aa{\ordf} \gdef^^ab{\guillemetleft} \gdef^^ac{$\lnot$} \gdef^^ad{\-} \gdef^^ae{\registeredsymbol} \gdef^^af{\={}} % \gdef^^b0{\textdegree} \gdef^^b1{$\pm$} \gdef^^b2{$^2$} \gdef^^b3{$^3$} \gdef^^b4{\'{}} \gdef^^b5{$\mu$} \gdef^^b6{\P} % \gdef^^b7{$^.$} \gdef^^b8{\cedilla\ } \gdef^^b9{$^1$} \gdef^^ba{\ordm} % \gdef^^bb{\guillemetright} \gdef^^bc{$1\over4$} \gdef^^bd{$1\over2$} \gdef^^be{$3\over4$} \gdef^^bf{\questiondown} % \gdef^^c0{\`A} \gdef^^c1{\'A} \gdef^^c2{\^A} \gdef^^c3{\~A} \gdef^^c4{\"A} \gdef^^c5{\ringaccent A} \gdef^^c6{\AE} \gdef^^c7{\cedilla C} \gdef^^c8{\`E} \gdef^^c9{\'E} \gdef^^ca{\^E} \gdef^^cb{\"E} \gdef^^cc{\`I} \gdef^^cd{\'I} \gdef^^ce{\^I} \gdef^^cf{\"I} % \gdef^^d0{\DH} \gdef^^d1{\~N} \gdef^^d2{\`O} \gdef^^d3{\'O} \gdef^^d4{\^O} \gdef^^d5{\~O} \gdef^^d6{\"O} \gdef^^d7{$\times$} \gdef^^d8{\O} \gdef^^d9{\`U} \gdef^^da{\'U} \gdef^^db{\^U} \gdef^^dc{\"U} \gdef^^dd{\'Y} \gdef^^de{\TH} \gdef^^df{\ss} % \gdef^^e0{\`a} \gdef^^e1{\'a} \gdef^^e2{\^a} \gdef^^e3{\~a} \gdef^^e4{\"a} \gdef^^e5{\ringaccent a} \gdef^^e6{\ae} \gdef^^e7{\cedilla c} \gdef^^e8{\`e} \gdef^^e9{\'e} \gdef^^ea{\^e} \gdef^^eb{\"e} \gdef^^ec{\`{\dotless i}} \gdef^^ed{\'{\dotless i}} \gdef^^ee{\^{\dotless i}} \gdef^^ef{\"{\dotless i}} % \gdef^^f0{\dh} \gdef^^f1{\~n} \gdef^^f2{\`o} \gdef^^f3{\'o} \gdef^^f4{\^o} \gdef^^f5{\~o} \gdef^^f6{\"o} \gdef^^f7{$\div$} \gdef^^f8{\o} \gdef^^f9{\`u} \gdef^^fa{\'u} \gdef^^fb{\^u} \gdef^^fc{\"u} \gdef^^fd{\'y} \gdef^^fe{\th} \gdef^^ff{\"y} } % Latin9 (ISO-8859-15) encoding character definitions. \def\latninechardefs{% % Encoding is almost identical to Latin1. \latonechardefs % \gdef^^a4{\euro} \gdef^^a6{\v S} \gdef^^a8{\v s} \gdef^^b4{\v Z} \gdef^^b8{\v z} \gdef^^bc{\OE} \gdef^^bd{\oe} \gdef^^be{\"Y} } % Latin2 (ISO-8859-2) character definitions. \def\lattwochardefs{% \gdef^^a0{\tie} \gdef^^a1{\ogonek{A}} \gdef^^a2{\u{}} \gdef^^a3{\L} \gdef^^a4{\missingcharmsg{CURRENCY SIGN}} \gdef^^a5{\v L} \gdef^^a6{\'S} \gdef^^a7{\S} \gdef^^a8{\"{}} \gdef^^a9{\v S} \gdef^^aa{\cedilla S} \gdef^^ab{\v T} \gdef^^ac{\'Z} \gdef^^ad{\-} \gdef^^ae{\v Z} \gdef^^af{\dotaccent Z} % \gdef^^b0{\textdegree} \gdef^^b1{\ogonek{a}} \gdef^^b2{\ogonek{ }} \gdef^^b3{\l} \gdef^^b4{\'{}} \gdef^^b5{\v l} \gdef^^b6{\'s} \gdef^^b7{\v{}} \gdef^^b8{\cedilla\ } \gdef^^b9{\v s} \gdef^^ba{\cedilla s} \gdef^^bb{\v t} \gdef^^bc{\'z} \gdef^^bd{\H{}} \gdef^^be{\v z} \gdef^^bf{\dotaccent z} % \gdef^^c0{\'R} \gdef^^c1{\'A} \gdef^^c2{\^A} \gdef^^c3{\u A} \gdef^^c4{\"A} \gdef^^c5{\'L} \gdef^^c6{\'C} \gdef^^c7{\cedilla C} \gdef^^c8{\v C} \gdef^^c9{\'E} \gdef^^ca{\ogonek{E}} \gdef^^cb{\"E} \gdef^^cc{\v E} \gdef^^cd{\'I} \gdef^^ce{\^I} \gdef^^cf{\v D} % \gdef^^d0{\DH} \gdef^^d1{\'N} \gdef^^d2{\v N} \gdef^^d3{\'O} \gdef^^d4{\^O} \gdef^^d5{\H O} \gdef^^d6{\"O} \gdef^^d7{$\times$} \gdef^^d8{\v R} \gdef^^d9{\ringaccent U} \gdef^^da{\'U} \gdef^^db{\H U} \gdef^^dc{\"U} \gdef^^dd{\'Y} \gdef^^de{\cedilla T} \gdef^^df{\ss} % \gdef^^e0{\'r} \gdef^^e1{\'a} \gdef^^e2{\^a} \gdef^^e3{\u a} \gdef^^e4{\"a} \gdef^^e5{\'l} \gdef^^e6{\'c} \gdef^^e7{\cedilla c} \gdef^^e8{\v c} \gdef^^e9{\'e} \gdef^^ea{\ogonek{e}} \gdef^^eb{\"e} \gdef^^ec{\v e} \gdef^^ed{\'{\dotless{i}}} \gdef^^ee{\^{\dotless{i}}} \gdef^^ef{\v d} % \gdef^^f0{\dh} \gdef^^f1{\'n} \gdef^^f2{\v n} \gdef^^f3{\'o} \gdef^^f4{\^o} \gdef^^f5{\H o} \gdef^^f6{\"o} \gdef^^f7{$\div$} \gdef^^f8{\v r} \gdef^^f9{\ringaccent u} \gdef^^fa{\'u} \gdef^^fb{\H u} \gdef^^fc{\"u} \gdef^^fd{\'y} \gdef^^fe{\cedilla t} \gdef^^ff{\dotaccent{}} } % UTF-8 character definitions. % % This code to support UTF-8 is based on LaTeX's utf8.def, with some % changes for Texinfo conventions. It is included here under the GPL by % permission from Frank Mittelbach and the LaTeX team. % \newcount\countUTFx \newcount\countUTFy \newcount\countUTFz \gdef\UTFviiiTwoOctets#1#2{\expandafter \UTFviiiDefined\csname u8:#1\string #2\endcsname} % \gdef\UTFviiiThreeOctets#1#2#3{\expandafter \UTFviiiDefined\csname u8:#1\string #2\string #3\endcsname} % \gdef\UTFviiiFourOctets#1#2#3#4{\expandafter \UTFviiiDefined\csname u8:#1\string #2\string #3\string #4\endcsname} \gdef\UTFviiiDefined#1{% \ifx #1\relax \message{\linenumber Unicode char \string #1 not defined for Texinfo}% \else \expandafter #1% \fi } \begingroup \catcode`\~13 \catcode`\"12 \def\UTFviiiLoop{% \global\catcode\countUTFx\active \uccode`\~\countUTFx \uppercase\expandafter{\UTFviiiTmp}% \advance\countUTFx by 1 \ifnum\countUTFx < \countUTFy \expandafter\UTFviiiLoop \fi} \countUTFx = "C2 \countUTFy = "E0 \def\UTFviiiTmp{% \xdef~{\noexpand\UTFviiiTwoOctets\string~}} \UTFviiiLoop \countUTFx = "E0 \countUTFy = "F0 \def\UTFviiiTmp{% \xdef~{\noexpand\UTFviiiThreeOctets\string~}} \UTFviiiLoop \countUTFx = "F0 \countUTFy = "F4 \def\UTFviiiTmp{% \xdef~{\noexpand\UTFviiiFourOctets\string~}} \UTFviiiLoop \endgroup \begingroup \catcode`\"=12 \catcode`\<=12 \catcode`\.=12 \catcode`\,=12 \catcode`\;=12 \catcode`\!=12 \catcode`\~=13 \gdef\DeclareUnicodeCharacter#1#2{% \countUTFz = "#1\relax %\wlog{\space\space defining Unicode char U+#1 (decimal \the\countUTFz)}% \begingroup \parseXMLCharref \def\UTFviiiTwoOctets##1##2{% \csname u8:##1\string ##2\endcsname}% \def\UTFviiiThreeOctets##1##2##3{% \csname u8:##1\string ##2\string ##3\endcsname}% \def\UTFviiiFourOctets##1##2##3##4{% \csname u8:##1\string ##2\string ##3\string ##4\endcsname}% \expandafter\expandafter\expandafter\expandafter \expandafter\expandafter\expandafter \gdef\UTFviiiTmp{#2}% \endgroup} \gdef\parseXMLCharref{% \ifnum\countUTFz < "A0\relax \errhelp = \EMsimple \errmessage{Cannot define Unicode char value < 00A0}% \else\ifnum\countUTFz < "800\relax \parseUTFviiiA,% \parseUTFviiiB C\UTFviiiTwoOctets.,% \else\ifnum\countUTFz < "10000\relax \parseUTFviiiA;% \parseUTFviiiA,% \parseUTFviiiB E\UTFviiiThreeOctets.{,;}% \else \parseUTFviiiA;% \parseUTFviiiA,% \parseUTFviiiA!% \parseUTFviiiB F\UTFviiiFourOctets.{!,;}% \fi\fi\fi } \gdef\parseUTFviiiA#1{% \countUTFx = \countUTFz \divide\countUTFz by 64 \countUTFy = \countUTFz \multiply\countUTFz by 64 \advance\countUTFx by -\countUTFz \advance\countUTFx by 128 \uccode `#1\countUTFx \countUTFz = \countUTFy} \gdef\parseUTFviiiB#1#2#3#4{% \advance\countUTFz by "#10\relax \uccode `#3\countUTFz \uppercase{\gdef\UTFviiiTmp{#2#3#4}}} \endgroup \def\utfeightchardefs{% \DeclareUnicodeCharacter{00A0}{\tie} \DeclareUnicodeCharacter{00A1}{\exclamdown} \DeclareUnicodeCharacter{00A3}{\pounds} \DeclareUnicodeCharacter{00A8}{\"{ }} \DeclareUnicodeCharacter{00A9}{\copyright} \DeclareUnicodeCharacter{00AA}{\ordf} \DeclareUnicodeCharacter{00AB}{\guillemetleft} \DeclareUnicodeCharacter{00AD}{\-} \DeclareUnicodeCharacter{00AE}{\registeredsymbol} \DeclareUnicodeCharacter{00AF}{\={ }} \DeclareUnicodeCharacter{00B0}{\ringaccent{ }} \DeclareUnicodeCharacter{00B4}{\'{ }} \DeclareUnicodeCharacter{00B8}{\cedilla{ }} \DeclareUnicodeCharacter{00BA}{\ordm} \DeclareUnicodeCharacter{00BB}{\guillemetright} \DeclareUnicodeCharacter{00BF}{\questiondown} \DeclareUnicodeCharacter{00C0}{\`A} \DeclareUnicodeCharacter{00C1}{\'A} \DeclareUnicodeCharacter{00C2}{\^A} \DeclareUnicodeCharacter{00C3}{\~A} \DeclareUnicodeCharacter{00C4}{\"A} \DeclareUnicodeCharacter{00C5}{\AA} \DeclareUnicodeCharacter{00C6}{\AE} \DeclareUnicodeCharacter{00C7}{\cedilla{C}} \DeclareUnicodeCharacter{00C8}{\`E} \DeclareUnicodeCharacter{00C9}{\'E} \DeclareUnicodeCharacter{00CA}{\^E} \DeclareUnicodeCharacter{00CB}{\"E} \DeclareUnicodeCharacter{00CC}{\`I} \DeclareUnicodeCharacter{00CD}{\'I} \DeclareUnicodeCharacter{00CE}{\^I} \DeclareUnicodeCharacter{00CF}{\"I} \DeclareUnicodeCharacter{00D0}{\DH} \DeclareUnicodeCharacter{00D1}{\~N} \DeclareUnicodeCharacter{00D2}{\`O} \DeclareUnicodeCharacter{00D3}{\'O} \DeclareUnicodeCharacter{00D4}{\^O} \DeclareUnicodeCharacter{00D5}{\~O} \DeclareUnicodeCharacter{00D6}{\"O} \DeclareUnicodeCharacter{00D8}{\O} \DeclareUnicodeCharacter{00D9}{\`U} \DeclareUnicodeCharacter{00DA}{\'U} \DeclareUnicodeCharacter{00DB}{\^U} \DeclareUnicodeCharacter{00DC}{\"U} \DeclareUnicodeCharacter{00DD}{\'Y} \DeclareUnicodeCharacter{00DE}{\TH} \DeclareUnicodeCharacter{00DF}{\ss} \DeclareUnicodeCharacter{00E0}{\`a} \DeclareUnicodeCharacter{00E1}{\'a} \DeclareUnicodeCharacter{00E2}{\^a} \DeclareUnicodeCharacter{00E3}{\~a} \DeclareUnicodeCharacter{00E4}{\"a} \DeclareUnicodeCharacter{00E5}{\aa} \DeclareUnicodeCharacter{00E6}{\ae} \DeclareUnicodeCharacter{00E7}{\cedilla{c}} \DeclareUnicodeCharacter{00E8}{\`e} \DeclareUnicodeCharacter{00E9}{\'e} \DeclareUnicodeCharacter{00EA}{\^e} \DeclareUnicodeCharacter{00EB}{\"e} \DeclareUnicodeCharacter{00EC}{\`{\dotless{i}}} \DeclareUnicodeCharacter{00ED}{\'{\dotless{i}}} \DeclareUnicodeCharacter{00EE}{\^{\dotless{i}}} \DeclareUnicodeCharacter{00EF}{\"{\dotless{i}}} \DeclareUnicodeCharacter{00F0}{\dh} \DeclareUnicodeCharacter{00F1}{\~n} \DeclareUnicodeCharacter{00F2}{\`o} \DeclareUnicodeCharacter{00F3}{\'o} \DeclareUnicodeCharacter{00F4}{\^o} \DeclareUnicodeCharacter{00F5}{\~o} \DeclareUnicodeCharacter{00F6}{\"o} \DeclareUnicodeCharacter{00F8}{\o} \DeclareUnicodeCharacter{00F9}{\`u} \DeclareUnicodeCharacter{00FA}{\'u} \DeclareUnicodeCharacter{00FB}{\^u} \DeclareUnicodeCharacter{00FC}{\"u} \DeclareUnicodeCharacter{00FD}{\'y} \DeclareUnicodeCharacter{00FE}{\th} \DeclareUnicodeCharacter{00FF}{\"y} \DeclareUnicodeCharacter{0100}{\=A} \DeclareUnicodeCharacter{0101}{\=a} \DeclareUnicodeCharacter{0102}{\u{A}} \DeclareUnicodeCharacter{0103}{\u{a}} \DeclareUnicodeCharacter{0104}{\ogonek{A}} \DeclareUnicodeCharacter{0105}{\ogonek{a}} \DeclareUnicodeCharacter{0106}{\'C} \DeclareUnicodeCharacter{0107}{\'c} \DeclareUnicodeCharacter{0108}{\^C} \DeclareUnicodeCharacter{0109}{\^c} \DeclareUnicodeCharacter{0118}{\ogonek{E}} \DeclareUnicodeCharacter{0119}{\ogonek{e}} \DeclareUnicodeCharacter{010A}{\dotaccent{C}} \DeclareUnicodeCharacter{010B}{\dotaccent{c}} \DeclareUnicodeCharacter{010C}{\v{C}} \DeclareUnicodeCharacter{010D}{\v{c}} \DeclareUnicodeCharacter{010E}{\v{D}} \DeclareUnicodeCharacter{0112}{\=E} \DeclareUnicodeCharacter{0113}{\=e} \DeclareUnicodeCharacter{0114}{\u{E}} \DeclareUnicodeCharacter{0115}{\u{e}} \DeclareUnicodeCharacter{0116}{\dotaccent{E}} \DeclareUnicodeCharacter{0117}{\dotaccent{e}} \DeclareUnicodeCharacter{011A}{\v{E}} \DeclareUnicodeCharacter{011B}{\v{e}} \DeclareUnicodeCharacter{011C}{\^G} \DeclareUnicodeCharacter{011D}{\^g} \DeclareUnicodeCharacter{011E}{\u{G}} \DeclareUnicodeCharacter{011F}{\u{g}} \DeclareUnicodeCharacter{0120}{\dotaccent{G}} \DeclareUnicodeCharacter{0121}{\dotaccent{g}} \DeclareUnicodeCharacter{0124}{\^H} \DeclareUnicodeCharacter{0125}{\^h} \DeclareUnicodeCharacter{0128}{\~I} \DeclareUnicodeCharacter{0129}{\~{\dotless{i}}} \DeclareUnicodeCharacter{012A}{\=I} \DeclareUnicodeCharacter{012B}{\={\dotless{i}}} \DeclareUnicodeCharacter{012C}{\u{I}} \DeclareUnicodeCharacter{012D}{\u{\dotless{i}}} \DeclareUnicodeCharacter{0130}{\dotaccent{I}} \DeclareUnicodeCharacter{0131}{\dotless{i}} \DeclareUnicodeCharacter{0132}{IJ} \DeclareUnicodeCharacter{0133}{ij} \DeclareUnicodeCharacter{0134}{\^J} \DeclareUnicodeCharacter{0135}{\^{\dotless{j}}} \DeclareUnicodeCharacter{0139}{\'L} \DeclareUnicodeCharacter{013A}{\'l} \DeclareUnicodeCharacter{0141}{\L} \DeclareUnicodeCharacter{0142}{\l} \DeclareUnicodeCharacter{0143}{\'N} \DeclareUnicodeCharacter{0144}{\'n} \DeclareUnicodeCharacter{0147}{\v{N}} \DeclareUnicodeCharacter{0148}{\v{n}} \DeclareUnicodeCharacter{014C}{\=O} \DeclareUnicodeCharacter{014D}{\=o} \DeclareUnicodeCharacter{014E}{\u{O}} \DeclareUnicodeCharacter{014F}{\u{o}} \DeclareUnicodeCharacter{0150}{\H{O}} \DeclareUnicodeCharacter{0151}{\H{o}} \DeclareUnicodeCharacter{0152}{\OE} \DeclareUnicodeCharacter{0153}{\oe} \DeclareUnicodeCharacter{0154}{\'R} \DeclareUnicodeCharacter{0155}{\'r} \DeclareUnicodeCharacter{0158}{\v{R}} \DeclareUnicodeCharacter{0159}{\v{r}} \DeclareUnicodeCharacter{015A}{\'S} \DeclareUnicodeCharacter{015B}{\'s} \DeclareUnicodeCharacter{015C}{\^S} \DeclareUnicodeCharacter{015D}{\^s} \DeclareUnicodeCharacter{015E}{\cedilla{S}} \DeclareUnicodeCharacter{015F}{\cedilla{s}} \DeclareUnicodeCharacter{0160}{\v{S}} \DeclareUnicodeCharacter{0161}{\v{s}} \DeclareUnicodeCharacter{0162}{\cedilla{t}} \DeclareUnicodeCharacter{0163}{\cedilla{T}} \DeclareUnicodeCharacter{0164}{\v{T}} \DeclareUnicodeCharacter{0168}{\~U} \DeclareUnicodeCharacter{0169}{\~u} \DeclareUnicodeCharacter{016A}{\=U} \DeclareUnicodeCharacter{016B}{\=u} \DeclareUnicodeCharacter{016C}{\u{U}} \DeclareUnicodeCharacter{016D}{\u{u}} \DeclareUnicodeCharacter{016E}{\ringaccent{U}} \DeclareUnicodeCharacter{016F}{\ringaccent{u}} \DeclareUnicodeCharacter{0170}{\H{U}} \DeclareUnicodeCharacter{0171}{\H{u}} \DeclareUnicodeCharacter{0174}{\^W} \DeclareUnicodeCharacter{0175}{\^w} \DeclareUnicodeCharacter{0176}{\^Y} \DeclareUnicodeCharacter{0177}{\^y} \DeclareUnicodeCharacter{0178}{\"Y} \DeclareUnicodeCharacter{0179}{\'Z} \DeclareUnicodeCharacter{017A}{\'z} \DeclareUnicodeCharacter{017B}{\dotaccent{Z}} \DeclareUnicodeCharacter{017C}{\dotaccent{z}} \DeclareUnicodeCharacter{017D}{\v{Z}} \DeclareUnicodeCharacter{017E}{\v{z}} \DeclareUnicodeCharacter{01C4}{D\v{Z}} \DeclareUnicodeCharacter{01C5}{D\v{z}} \DeclareUnicodeCharacter{01C6}{d\v{z}} \DeclareUnicodeCharacter{01C7}{LJ} \DeclareUnicodeCharacter{01C8}{Lj} \DeclareUnicodeCharacter{01C9}{lj} \DeclareUnicodeCharacter{01CA}{NJ} \DeclareUnicodeCharacter{01CB}{Nj} \DeclareUnicodeCharacter{01CC}{nj} \DeclareUnicodeCharacter{01CD}{\v{A}} \DeclareUnicodeCharacter{01CE}{\v{a}} \DeclareUnicodeCharacter{01CF}{\v{I}} \DeclareUnicodeCharacter{01D0}{\v{\dotless{i}}} \DeclareUnicodeCharacter{01D1}{\v{O}} \DeclareUnicodeCharacter{01D2}{\v{o}} \DeclareUnicodeCharacter{01D3}{\v{U}} \DeclareUnicodeCharacter{01D4}{\v{u}} \DeclareUnicodeCharacter{01E2}{\={\AE}} \DeclareUnicodeCharacter{01E3}{\={\ae}} \DeclareUnicodeCharacter{01E6}{\v{G}} \DeclareUnicodeCharacter{01E7}{\v{g}} \DeclareUnicodeCharacter{01E8}{\v{K}} \DeclareUnicodeCharacter{01E9}{\v{k}} \DeclareUnicodeCharacter{01F0}{\v{\dotless{j}}} \DeclareUnicodeCharacter{01F1}{DZ} \DeclareUnicodeCharacter{01F2}{Dz} \DeclareUnicodeCharacter{01F3}{dz} \DeclareUnicodeCharacter{01F4}{\'G} \DeclareUnicodeCharacter{01F5}{\'g} \DeclareUnicodeCharacter{01F8}{\`N} \DeclareUnicodeCharacter{01F9}{\`n} \DeclareUnicodeCharacter{01FC}{\'{\AE}} \DeclareUnicodeCharacter{01FD}{\'{\ae}} \DeclareUnicodeCharacter{01FE}{\'{\O}} \DeclareUnicodeCharacter{01FF}{\'{\o}} \DeclareUnicodeCharacter{021E}{\v{H}} \DeclareUnicodeCharacter{021F}{\v{h}} \DeclareUnicodeCharacter{0226}{\dotaccent{A}} \DeclareUnicodeCharacter{0227}{\dotaccent{a}} \DeclareUnicodeCharacter{0228}{\cedilla{E}} \DeclareUnicodeCharacter{0229}{\cedilla{e}} \DeclareUnicodeCharacter{022E}{\dotaccent{O}} \DeclareUnicodeCharacter{022F}{\dotaccent{o}} \DeclareUnicodeCharacter{0232}{\=Y} \DeclareUnicodeCharacter{0233}{\=y} \DeclareUnicodeCharacter{0237}{\dotless{j}} \DeclareUnicodeCharacter{02DB}{\ogonek{ }} \DeclareUnicodeCharacter{1E02}{\dotaccent{B}} \DeclareUnicodeCharacter{1E03}{\dotaccent{b}} \DeclareUnicodeCharacter{1E04}{\udotaccent{B}} \DeclareUnicodeCharacter{1E05}{\udotaccent{b}} \DeclareUnicodeCharacter{1E06}{\ubaraccent{B}} \DeclareUnicodeCharacter{1E07}{\ubaraccent{b}} \DeclareUnicodeCharacter{1E0A}{\dotaccent{D}} \DeclareUnicodeCharacter{1E0B}{\dotaccent{d}} \DeclareUnicodeCharacter{1E0C}{\udotaccent{D}} \DeclareUnicodeCharacter{1E0D}{\udotaccent{d}} \DeclareUnicodeCharacter{1E0E}{\ubaraccent{D}} \DeclareUnicodeCharacter{1E0F}{\ubaraccent{d}} \DeclareUnicodeCharacter{1E1E}{\dotaccent{F}} \DeclareUnicodeCharacter{1E1F}{\dotaccent{f}} \DeclareUnicodeCharacter{1E20}{\=G} \DeclareUnicodeCharacter{1E21}{\=g} \DeclareUnicodeCharacter{1E22}{\dotaccent{H}} \DeclareUnicodeCharacter{1E23}{\dotaccent{h}} \DeclareUnicodeCharacter{1E24}{\udotaccent{H}} \DeclareUnicodeCharacter{1E25}{\udotaccent{h}} \DeclareUnicodeCharacter{1E26}{\"H} \DeclareUnicodeCharacter{1E27}{\"h} \DeclareUnicodeCharacter{1E30}{\'K} \DeclareUnicodeCharacter{1E31}{\'k} \DeclareUnicodeCharacter{1E32}{\udotaccent{K}} \DeclareUnicodeCharacter{1E33}{\udotaccent{k}} \DeclareUnicodeCharacter{1E34}{\ubaraccent{K}} \DeclareUnicodeCharacter{1E35}{\ubaraccent{k}} \DeclareUnicodeCharacter{1E36}{\udotaccent{L}} \DeclareUnicodeCharacter{1E37}{\udotaccent{l}} \DeclareUnicodeCharacter{1E3A}{\ubaraccent{L}} \DeclareUnicodeCharacter{1E3B}{\ubaraccent{l}} \DeclareUnicodeCharacter{1E3E}{\'M} \DeclareUnicodeCharacter{1E3F}{\'m} \DeclareUnicodeCharacter{1E40}{\dotaccent{M}} \DeclareUnicodeCharacter{1E41}{\dotaccent{m}} \DeclareUnicodeCharacter{1E42}{\udotaccent{M}} \DeclareUnicodeCharacter{1E43}{\udotaccent{m}} \DeclareUnicodeCharacter{1E44}{\dotaccent{N}} \DeclareUnicodeCharacter{1E45}{\dotaccent{n}} \DeclareUnicodeCharacter{1E46}{\udotaccent{N}} \DeclareUnicodeCharacter{1E47}{\udotaccent{n}} \DeclareUnicodeCharacter{1E48}{\ubaraccent{N}} \DeclareUnicodeCharacter{1E49}{\ubaraccent{n}} \DeclareUnicodeCharacter{1E54}{\'P} \DeclareUnicodeCharacter{1E55}{\'p} \DeclareUnicodeCharacter{1E56}{\dotaccent{P}} \DeclareUnicodeCharacter{1E57}{\dotaccent{p}} \DeclareUnicodeCharacter{1E58}{\dotaccent{R}} \DeclareUnicodeCharacter{1E59}{\dotaccent{r}} \DeclareUnicodeCharacter{1E5A}{\udotaccent{R}} \DeclareUnicodeCharacter{1E5B}{\udotaccent{r}} \DeclareUnicodeCharacter{1E5E}{\ubaraccent{R}} \DeclareUnicodeCharacter{1E5F}{\ubaraccent{r}} \DeclareUnicodeCharacter{1E60}{\dotaccent{S}} \DeclareUnicodeCharacter{1E61}{\dotaccent{s}} \DeclareUnicodeCharacter{1E62}{\udotaccent{S}} \DeclareUnicodeCharacter{1E63}{\udotaccent{s}} \DeclareUnicodeCharacter{1E6A}{\dotaccent{T}} \DeclareUnicodeCharacter{1E6B}{\dotaccent{t}} \DeclareUnicodeCharacter{1E6C}{\udotaccent{T}} \DeclareUnicodeCharacter{1E6D}{\udotaccent{t}} \DeclareUnicodeCharacter{1E6E}{\ubaraccent{T}} \DeclareUnicodeCharacter{1E6F}{\ubaraccent{t}} \DeclareUnicodeCharacter{1E7C}{\~V} \DeclareUnicodeCharacter{1E7D}{\~v} \DeclareUnicodeCharacter{1E7E}{\udotaccent{V}} \DeclareUnicodeCharacter{1E7F}{\udotaccent{v}} \DeclareUnicodeCharacter{1E80}{\`W} \DeclareUnicodeCharacter{1E81}{\`w} \DeclareUnicodeCharacter{1E82}{\'W} \DeclareUnicodeCharacter{1E83}{\'w} \DeclareUnicodeCharacter{1E84}{\"W} \DeclareUnicodeCharacter{1E85}{\"w} \DeclareUnicodeCharacter{1E86}{\dotaccent{W}} \DeclareUnicodeCharacter{1E87}{\dotaccent{w}} \DeclareUnicodeCharacter{1E88}{\udotaccent{W}} \DeclareUnicodeCharacter{1E89}{\udotaccent{w}} \DeclareUnicodeCharacter{1E8A}{\dotaccent{X}} \DeclareUnicodeCharacter{1E8B}{\dotaccent{x}} \DeclareUnicodeCharacter{1E8C}{\"X} \DeclareUnicodeCharacter{1E8D}{\"x} \DeclareUnicodeCharacter{1E8E}{\dotaccent{Y}} \DeclareUnicodeCharacter{1E8F}{\dotaccent{y}} \DeclareUnicodeCharacter{1E90}{\^Z} \DeclareUnicodeCharacter{1E91}{\^z} \DeclareUnicodeCharacter{1E92}{\udotaccent{Z}} \DeclareUnicodeCharacter{1E93}{\udotaccent{z}} \DeclareUnicodeCharacter{1E94}{\ubaraccent{Z}} \DeclareUnicodeCharacter{1E95}{\ubaraccent{z}} \DeclareUnicodeCharacter{1E96}{\ubaraccent{h}} \DeclareUnicodeCharacter{1E97}{\"t} \DeclareUnicodeCharacter{1E98}{\ringaccent{w}} \DeclareUnicodeCharacter{1E99}{\ringaccent{y}} \DeclareUnicodeCharacter{1EA0}{\udotaccent{A}} \DeclareUnicodeCharacter{1EA1}{\udotaccent{a}} \DeclareUnicodeCharacter{1EB8}{\udotaccent{E}} \DeclareUnicodeCharacter{1EB9}{\udotaccent{e}} \DeclareUnicodeCharacter{1EBC}{\~E} \DeclareUnicodeCharacter{1EBD}{\~e} \DeclareUnicodeCharacter{1ECA}{\udotaccent{I}} \DeclareUnicodeCharacter{1ECB}{\udotaccent{i}} \DeclareUnicodeCharacter{1ECC}{\udotaccent{O}} \DeclareUnicodeCharacter{1ECD}{\udotaccent{o}} \DeclareUnicodeCharacter{1EE4}{\udotaccent{U}} \DeclareUnicodeCharacter{1EE5}{\udotaccent{u}} \DeclareUnicodeCharacter{1EF2}{\`Y} \DeclareUnicodeCharacter{1EF3}{\`y} \DeclareUnicodeCharacter{1EF4}{\udotaccent{Y}} \DeclareUnicodeCharacter{1EF8}{\~Y} \DeclareUnicodeCharacter{1EF9}{\~y} \DeclareUnicodeCharacter{2013}{--} \DeclareUnicodeCharacter{2014}{---} \DeclareUnicodeCharacter{2018}{\quoteleft} \DeclareUnicodeCharacter{2019}{\quoteright} \DeclareUnicodeCharacter{201A}{\quotesinglbase} \DeclareUnicodeCharacter{201C}{\quotedblleft} \DeclareUnicodeCharacter{201D}{\quotedblright} \DeclareUnicodeCharacter{201E}{\quotedblbase} \DeclareUnicodeCharacter{2022}{\bullet} \DeclareUnicodeCharacter{2026}{\dots} \DeclareUnicodeCharacter{2039}{\guilsinglleft} \DeclareUnicodeCharacter{203A}{\guilsinglright} \DeclareUnicodeCharacter{20AC}{\euro} \DeclareUnicodeCharacter{2192}{\expansion} \DeclareUnicodeCharacter{21D2}{\result} \DeclareUnicodeCharacter{2212}{\minus} \DeclareUnicodeCharacter{2217}{\point} \DeclareUnicodeCharacter{2261}{\equiv} }% end of \utfeightchardefs % US-ASCII character definitions. \def\asciichardefs{% nothing need be done \relax } % Make non-ASCII characters printable again for compatibility with % existing Texinfo documents that may use them, even without declaring a % document encoding. % \setnonasciicharscatcode \other \message{formatting,} \newdimen\defaultparindent \defaultparindent = 15pt \chapheadingskip = 15pt plus 4pt minus 2pt \secheadingskip = 12pt plus 3pt minus 2pt \subsecheadingskip = 9pt plus 2pt minus 2pt % Prevent underfull vbox error messages. \vbadness = 10000 % Don't be very finicky about underfull hboxes, either. \hbadness = 6666 % Following George Bush, get rid of widows and orphans. \widowpenalty=10000 \clubpenalty=10000 % Use TeX 3.0's \emergencystretch to help line breaking, but if we're % using an old version of TeX, don't do anything. We want the amount of % stretch added to depend on the line length, hence the dependence on % \hsize. We call this whenever the paper size is set. % \def\setemergencystretch{% \ifx\emergencystretch\thisisundefined % Allow us to assign to \emergencystretch anyway. \def\emergencystretch{\dimen0}% \else \emergencystretch = .15\hsize \fi } % Parameters in order: 1) textheight; 2) textwidth; % 3) voffset; 4) hoffset; 5) binding offset; 6) topskip; % 7) physical page height; 8) physical page width. % % We also call \setleading{\textleading}, so the caller should define % \textleading. The caller should also set \parskip. % \def\internalpagesizes#1#2#3#4#5#6#7#8{% \voffset = #3\relax \topskip = #6\relax \splittopskip = \topskip % \vsize = #1\relax \advance\vsize by \topskip \outervsize = \vsize \advance\outervsize by 2\topandbottommargin \pageheight = \vsize % \hsize = #2\relax \outerhsize = \hsize \advance\outerhsize by 0.5in \pagewidth = \hsize % \normaloffset = #4\relax \bindingoffset = #5\relax % \ifpdf \pdfpageheight #7\relax \pdfpagewidth #8\relax % if we don't reset these, they will remain at "1 true in" of % whatever layout pdftex was dumped with. \pdfhorigin = 1 true in \pdfvorigin = 1 true in \fi % \setleading{\textleading} % \parindent = \defaultparindent \setemergencystretch } % @letterpaper (the default). \def\letterpaper{{\globaldefs = 1 \parskip = 3pt plus 2pt minus 1pt \textleading = 13.2pt % % If page is nothing but text, make it come out even. \internalpagesizes{607.2pt}{6in}% that's 46 lines {\voffset}{.25in}% {\bindingoffset}{36pt}% {11in}{8.5in}% }} % Use @smallbook to reset parameters for 7x9.25 trim size. \def\smallbook{{\globaldefs = 1 \parskip = 2pt plus 1pt \textleading = 12pt % \internalpagesizes{7.5in}{5in}% {-.2in}{0in}% {\bindingoffset}{16pt}% {9.25in}{7in}% % \lispnarrowing = 0.3in \tolerance = 700 \hfuzz = 1pt \contentsrightmargin = 0pt \defbodyindent = .5cm }} % Use @smallerbook to reset parameters for 6x9 trim size. % (Just testing, parameters still in flux.) \def\smallerbook{{\globaldefs = 1 \parskip = 1.5pt plus 1pt \textleading = 12pt % \internalpagesizes{7.4in}{4.8in}% {-.2in}{-.4in}% {0pt}{14pt}% {9in}{6in}% % \lispnarrowing = 0.25in \tolerance = 700 \hfuzz = 1pt \contentsrightmargin = 0pt \defbodyindent = .4cm }} % Use @afourpaper to print on European A4 paper. \def\afourpaper{{\globaldefs = 1 \parskip = 3pt plus 2pt minus 1pt \textleading = 13.2pt % % Double-side printing via postscript on Laserjet 4050 % prints double-sided nicely when \bindingoffset=10mm and \hoffset=-6mm. % To change the settings for a different printer or situation, adjust % \normaloffset until the front-side and back-side texts align. Then % do the same for \bindingoffset. You can set these for testing in % your texinfo source file like this: % @tex % \global\normaloffset = -6mm % \global\bindingoffset = 10mm % @end tex \internalpagesizes{673.2pt}{160mm}% that's 51 lines {\voffset}{\hoffset}% {\bindingoffset}{44pt}% {297mm}{210mm}% % \tolerance = 700 \hfuzz = 1pt \contentsrightmargin = 0pt \defbodyindent = 5mm }} % Use @afivepaper to print on European A5 paper. % From romildo@urano.iceb.ufop.br, 2 July 2000. % He also recommends making @example and @lisp be small. \def\afivepaper{{\globaldefs = 1 \parskip = 2pt plus 1pt minus 0.1pt \textleading = 12.5pt % \internalpagesizes{160mm}{120mm}% {\voffset}{\hoffset}% {\bindingoffset}{8pt}% {210mm}{148mm}% % \lispnarrowing = 0.2in \tolerance = 800 \hfuzz = 1.2pt \contentsrightmargin = 0pt \defbodyindent = 2mm \tableindent = 12mm }} % A specific text layout, 24x15cm overall, intended for A4 paper. \def\afourlatex{{\globaldefs = 1 \afourpaper \internalpagesizes{237mm}{150mm}% {\voffset}{4.6mm}% {\bindingoffset}{7mm}% {297mm}{210mm}% % % Must explicitly reset to 0 because we call \afourpaper. \globaldefs = 0 }} % Use @afourwide to print on A4 paper in landscape format. \def\afourwide{{\globaldefs = 1 \afourpaper \internalpagesizes{241mm}{165mm}% {\voffset}{-2.95mm}% {\bindingoffset}{7mm}% {297mm}{210mm}% \globaldefs = 0 }} % @pagesizes TEXTHEIGHT[,TEXTWIDTH] % Perhaps we should allow setting the margins, \topskip, \parskip, % and/or leading, also. Or perhaps we should compute them somehow. % \parseargdef\pagesizes{\pagesizesyyy #1,,\finish} \def\pagesizesyyy#1,#2,#3\finish{{% \setbox0 = \hbox{\ignorespaces #2}\ifdim\wd0 > 0pt \hsize=#2\relax \fi \globaldefs = 1 % \parskip = 3pt plus 2pt minus 1pt \setleading{\textleading}% % \dimen0 = #1\relax \advance\dimen0 by \voffset % \dimen2 = \hsize \advance\dimen2 by \normaloffset % \internalpagesizes{#1}{\hsize}% {\voffset}{\normaloffset}% {\bindingoffset}{44pt}% {\dimen0}{\dimen2}% }} % Set default to letter. % \letterpaper \message{and turning on texinfo input format.} \def^^L{\par} % remove \outer, so ^L can appear in an @comment % DEL is a comment character, in case @c does not suffice. \catcode`\^^? = 14 % Define macros to output various characters with catcode for normal text. \catcode`\"=\other \def\normaldoublequote{"} \catcode`\$=\other \def\normaldollar{$}%$ font-lock fix \catcode`\+=\other \def\normalplus{+} \catcode`\<=\other \def\normalless{<} \catcode`\>=\other \def\normalgreater{>} \catcode`\^=\other \def\normalcaret{^} \catcode`\_=\other \def\normalunderscore{_} \catcode`\|=\other \def\normalverticalbar{|} \catcode`\~=\other \def\normaltilde{~} % This macro is used to make a character print one way in \tt % (where it can probably be output as-is), and another way in other fonts, % where something hairier probably needs to be done. % % #1 is what to print if we are indeed using \tt; #2 is what to print % otherwise. Since all the Computer Modern typewriter fonts have zero % interword stretch (and shrink), and it is reasonable to expect all % typewriter fonts to have this, we can check that font parameter. % \def\ifusingtt#1#2{\ifdim \fontdimen3\font=0pt #1\else #2\fi} % Same as above, but check for italic font. Actually this also catches % non-italic slanted fonts since it is impossible to distinguish them from % italic fonts. But since this is only used by $ and it uses \sl anyway % this is not a problem. \def\ifusingit#1#2{\ifdim \fontdimen1\font>0pt #1\else #2\fi} % Turn off all special characters except @ % (and those which the user can use as if they were ordinary). % Most of these we simply print from the \tt font, but for some, we can % use math or other variants that look better in normal text. \catcode`\"=\active \def\activedoublequote{{\tt\char34}} \let"=\activedoublequote \catcode`\~=\active \def~{{\tt\char126}} \chardef\hat=`\^ \catcode`\^=\active \def^{{\tt \hat}} \catcode`\_=\active \def_{\ifusingtt\normalunderscore\_} \let\realunder=_ % Subroutine for the previous macro. \def\_{\leavevmode \kern.07em \vbox{\hrule width.3em height.1ex}\kern .07em } \catcode`\|=\active \def|{{\tt\char124}} \chardef \less=`\< \catcode`\<=\active \def<{{\tt \less}} \chardef \gtr=`\> \catcode`\>=\active \def>{{\tt \gtr}} \catcode`\+=\active \def+{{\tt \char 43}} \catcode`\$=\active \def${\ifusingit{{\sl\$}}\normaldollar}%$ font-lock fix % If a .fmt file is being used, characters that might appear in a file % name cannot be active until we have parsed the command line. % So turn them off again, and have \everyjob (or @setfilename) turn them on. % \otherifyactive is called near the end of this file. \def\otherifyactive{\catcode`+=\other \catcode`\_=\other} % Used sometimes to turn off (effectively) the active characters even after % parsing them. \def\turnoffactive{% \normalturnoffactive \otherbackslash } \catcode`\@=0 % \backslashcurfont outputs one backslash character in current font, % as in \char`\\. \global\chardef\backslashcurfont=`\\ \global\let\rawbackslashxx=\backslashcurfont % let existing .??s files work % \realbackslash is an actual character `\' with catcode other, and % \doublebackslash is two of them (for the pdf outlines). {\catcode`\\=\other @gdef@realbackslash{\} @gdef@doublebackslash{\\}} % In texinfo, backslash is an active character; it prints the backslash % in fixed width font. \catcode`\\=\active % @ for escape char from now on. % The story here is that in math mode, the \char of \backslashcurfont % ends up printing the roman \ from the math symbol font (because \char % in math mode uses the \mathcode, and plain.tex sets % \mathcode`\\="026E). It seems better for @backslashchar{} to always % print a typewriter backslash, hence we use an explicit \mathchar, % which is the decimal equivalent of "715c (class 7, e.g., use \fam; % ignored family value; char position "5C). We can't use " for the % usual hex value because it has already been made active. @def@normalbackslash{{@tt @ifmmode @mathchar29020 @else @backslashcurfont @fi}} @let@backslashchar = @normalbackslash % @backslashchar{} is for user documents. % On startup, @fixbackslash assigns: % @let \ = @normalbackslash % \rawbackslash defines an active \ to do \backslashcurfont. % \otherbackslash defines an active \ to be a literal `\' character with % catcode other. We switch back and forth between these. @gdef@rawbackslash{@let\=@backslashcurfont} @gdef@otherbackslash{@let\=@realbackslash} % Same as @turnoffactive except outputs \ as {\tt\char`\\} instead of % the literal character `\'. % @def@normalturnoffactive{% @let"=@normaldoublequote @let$=@normaldollar %$ font-lock fix @let+=@normalplus @let<=@normalless @let>=@normalgreater @let\=@normalbackslash @let^=@normalcaret @let_=@normalunderscore @let|=@normalverticalbar @let~=@normaltilde @markupsetuplqdefault @markupsetuprqdefault @unsepspaces } % Make _ and + \other characters, temporarily. % This is canceled by @fixbackslash. @otherifyactive % If a .fmt file is being used, we don't want the `\input texinfo' to show up. % That is what \eatinput is for; after that, the `\' should revert to printing % a backslash. % @gdef@eatinput input texinfo{@fixbackslash} @global@let\ = @eatinput % On the other hand, perhaps the file did not have a `\input texinfo'. Then % the first `\' in the file would cause an error. This macro tries to fix % that, assuming it is called before the first `\' could plausibly occur. % Also turn back on active characters that might appear in the input % file name, in case not using a pre-dumped format. % @gdef@fixbackslash{% @ifx\@eatinput @let\ = @normalbackslash @fi @catcode`+=@active @catcode`@_=@active } % Say @foo, not \foo, in error messages. @escapechar = `@@ % These (along with & and #) are made active for url-breaking, so need % active definitions as the normal characters. @def@normaldot{.} @def@normalquest{?} @def@normalslash{/} % These look ok in all fonts, so just make them not special. % @hashchar{} gets its own user-level command, because of #line. @catcode`@& = @other @def@normalamp{&} @catcode`@# = @other @def@normalhash{#} @catcode`@% = @other @def@normalpercent{%} @let @hashchar = @normalhash @c Finally, make ` and ' active, so that txicodequoteundirected and @c txicodequotebacktick work right in, e.g., @w{@code{`foo'}}. If we @c don't make ` and ' active, @code will not get them as active chars. @c Do this last of all since we use ` in the previous @catcode assignments. @catcode`@'=@active @catcode`@`=@active @markupsetuplqdefault @markupsetuprqdefault @c Local variables: @c eval: (add-hook 'write-file-hooks 'time-stamp) @c page-delimiter: "^\\\\message" @c time-stamp-start: "def\\\\texinfoversion{" @c time-stamp-format: "%:y-%02m-%02d.%02H" @c time-stamp-end: "}" @c End: @c vim:sw=2: @ignore arch-tag: e1b36e32-c96e-4135-a41a-0b2efa2ea115 @end ignore fftw-3.3.3/doc/f77_wisdom.f0000644000175000017500000000555412054410050012310 00000000000000c Copyright (c) 2003, 2007-11 Matteo Frigo c Copyright (c) 2003, 2007-11 Massachusetts Institute of Technology c c This program is free software; you can redistribute it and/or modify c it under the terms of the GNU General Public License as published by c the Free Software Foundation; either version 2 of the License, or c (at your option) any later version. c c This program is distributed in the hope that it will be useful, c but WITHOUT ANY WARRANTY; without even the implied warranty of c MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the c GNU General Public License for more details. c c You should have received a copy of the GNU General Public License c along with this program; if not, write to the Free Software c Foundation, Inc., 51 Franklin Street, Fifth Floor, Boston, MA 02110-1301 USA c cccccccccccccccccccccccccccccccccccccccccccccccccccccccccccccccccccccccccccccc c c This is an example implementation of Fortran wisdom export/import c to/from a Fortran unit (file), exploiting the generic c dfftw_export_wisdom/dfftw_import_wisdom functions. c c We cannot compile this file into the FFTW library itself, lest all c FFTW-calling programs be required to link to the Fortran I/O c libraries. c cccccccccccccccccccccccccccccccccccccccccccccccccccccccccccccccccccccccccccccc c Strictly speaking, the '$' format specifier, which allows us to c write a character without a trailing newline, is not standard F77. c However, it seems to be a nearly universal extension. subroutine write_char(c, iunit) character c integer iunit write(iunit,321) c 321 format(a,$) end subroutine export_wisdom_to_file(iunit) integer iunit external write_char call dfftw_export_wisdom(write_char, iunit) end c Fortran 77 does not have any portable way to read an arbitrary c file one character at a time. The best alternative seems to be to c read a whole line into a buffer, since for fftw-exported wisdom we c can bound the line length. (If the file contains longer lines, c then the lines will be truncated and the wisdom import should c simply fail.) Ugh. subroutine read_char(ic, iunit) integer ic integer iunit character*256 buf save buf integer ibuf data ibuf/257/ save ibuf if (ibuf .lt. 257) then ic = ichar(buf(ibuf:ibuf)) ibuf = ibuf + 1 return endif read(iunit,123,end=666) buf ic = ichar(buf(1:1)) ibuf = 2 return 666 ic = -1 ibuf = 257 123 format(a256) end subroutine import_wisdom_from_file(isuccess, iunit) integer isuccess integer iunit external read_char call dfftw_import_wisdom(isuccess, read_char, iunit) end fftw-3.3.3/doc/fftw3.pdf0000644000175000017500000241174512054411114011707 00000000000000%PDF-1.5 %ÐÔÅØ 1 0 obj << /Length 587 /Filter /FlateDecode >> stream xÚmTM¢@½ó+z&ÎÁ±?tBL$ñ°ãd4›½*´.‰<øï·_•èÌf’W¯_wÕ«îrðãc;Šòê`GæUŠOÛV×&³£øç¾öƒ¤Ê®[vïÖæ6ïWÛ7ñÑTÙÖvb¯“uYt/N¼.³ó5·½êÿ¢¥=åS‚> stream xÚmTM¢@½ó+z&ÎÁ±?tBL0ñ°ãd4›½*´.‰<Ì¿ß~U¢Îf’W¯_u½ªîvðãc;ZäÕÁŽÌ«Ÿ¶­®MfGñÏ}í I•]/¶ìÞ­ÍmÞ¯¶o⣩²­íÄ0^'ë²è^œx]fçkn{ÕÿEK{*ʇuÄpg6;µÞ$4»¢;»µgZ8, ’ü²M[Tå›P¯RJG¤eWxm½ñ­ž÷ŽE™7·¢â žÒ"/²îÑ7»¸¦‘¼ýj;{Y—ÇÊ‹"1þt‹m×|‘£o¼irÛåI É‘c¶×º>[TÒ›ÏEnn#×ÛûþbÅø¹‘ûÒBS¬ØEV嶭÷™möåÉz‘”s…«¹gËüŸµ)gŽÏR©ð133wÄ xAÄbêí;¬ÒaGL6K& 0+‡}&ö"?‘á°(Ò¦Òa/ ¡cì,•!£½¥‰î-fö3¤Ù*IÃx {aªùð”sIC%ÒðhSô¢¨7å£Å}­HÏ=ŤIYƒ¹(îƒêjŧ ÿZóéàü4{ÖØSOØá5˜‡áZ ä®ekxvKº·Ǭü÷…Ü@2aÂ> stream xÚmSÁnâ0½ç+¼$z Ø¨"¤€ÄaKU¢Õ^C<ÐHàDN8ð÷õÌŠV{Hôüæç=üúØS`¾Jñ m}u%ŒÒßE Y]^/`»w¦¶oâÃÕå:1L·ÙÖVÝ‹omy¾èUÿ­àTÙ ÖÃþŽv¹Êó‘DM^ug{¦…Ç‚° ÉpmUÛ7¡^¥”žX[“ÖôÚã{=1î+kܽ¨8 …@iaª²»¯è_^|Ó˜¼¿µ\¶öXq,ÆŸ>ØvîFŽ^‚ñÎp•=‰!9òÌþÚ4gÀêBË¥0pôùÞÞ‹ ˆñs#P~k@hZ+vQÖÚ¦(ÁöA,åRÄÑf€5ÿĦœq8>K¥Â_¸—žX NˆHæžÐÔ3$¤Çž˜{<Ý0Š*¢5cÕ~ÿP÷õʯÂùÝ5WÂ42^!ž0^#žrq‰xƘœE„3xÎü ñ ªz“)cÒgl1BÌîÒ°õ•?ŸXqû!òŠNA‡¨Wš»A*dý1ùÔ)iȧΰÅç“Оó â9ç’†NVf¤¡–kô¯VäaŠžUJü†ôì?%Íš5Ø»bÿTW£=ј«±®–¾Œ¿É5ëñ2éfè&p2pj³V^ócH£Mc†VYxLS7˜E=›þ1âj· ¾gÈÈ endstream endobj 6 0 obj << /Length 285 /Filter /FlateDecode >> stream xÚPMKÄ0½çW䘀‰™™¤m®‚]XPPăíº‡m±[ö÷;iºâAD2“äÍ›7¤ã¬‰lô|;·¼N;YŠû€ç 9ÿ 1càd†§š¾«$.Û:Ht¶ M©ÏšB6V2½Ë'Õ¶éQ?§­¸Nâó›#ƒÏŠçQy:ß™bêd/î sœ.Bf¦=ÊàÑz*ôý8iCDꤱQ6 & ꘋý8”?²¥ÆPòí¨±^»ù|Õ&×+:À,{Ù¼%_aÖ`°*"Œ‡Ú/°È¸y™çnäÖ:¨VG¯¦ýnÌý¼šÀ¦ÜÃÜ4ꆂÝØ’·ãÇpd¹¿Ž,ýtŠítvê ŸØqS endstream endobj 13 0 obj << /Length 843 /Filter /FlateDecode >> stream xÚUKÓ0¾÷Wä˜J4ëÄvWE TâÀrpokmW‰³°ÿž“MB+T©öŒÇóøæ'üÒ¨bQÁyR‰2ª æµÝ.¢Í·‹4Ø žÁúO“•Ì²h5óöv³¸[K¥,©X•F› —•,a2Ú4ÑÏx³7ýrÅ9ª]fE<¨É£þÁv´Y¯7?p'â{&ÙÓ2+cÝõƶtÌø½!ƒL’î‹E+2=¢÷írÿº f,ÍÀWš,m>A«´JÒ¬¤ÌÞÙÞ{îÌnç¤H²‚G,á£]ú»u.æU®xÂ82³»g©$Ós@R‘H‘Cto•1Æ)÷ÏÊ9m)Ñõ²1$b)ÓU.“¢,C¾Õ+ó­^™oõ¿ùö½ªd~èµs¡…ÛÞ78MÕØRo–%uÖûÖìîË %ò<©¦„_©çGÓSÛ+=M+â]Güqº!…³dpTþIª Çé]g¶>)<|Ú*gŽ¤©íÉèàSFÛS¼’èZzºŽäáI*SÊôÔóLƒé𪂛>C·}òç­u¦&à:ÚSD´8wÏÊŸ_ýêÂýS§{ÝQE!<Á®àú<å@ _íù$T·açyEÓY°{I°ãÂŽ«/³ˆŸIò°ãf;ÊGë«j ÐMŒ>f³Ý“ÆS¦˜ÐÀ‹·ñ§Œ‡¶Á'ü:µmãÈ%øw7—i‘iwð ˆ4¿è'ùUn7š0c  ÃÁ“[ ÍK‡‹£`»G’|™"ñ‚¦ 1è¨@”H¾FÜC—zràƒUÑrMPNäÁ€Í˜~í_`PA+o$î¦Û¶úæËy=¦2Œ©¸SÆTN˜“ÆT\Œ©Œ]媶?¨©•2L§œ¦SÌÙA᩶›)€aïéRåñÁÏÞnP; ­—¼˜¸TÉÐi°R[çìëBú3ŽÁâ\ÑOŸ–þ[äiîƒå±þSë“»Áh¢[Vâ b(\µ•c[³² ¢@Yk²êòMÀnªâ äptg`$-j·¡g”Xˆ¹Æw½ÓúØßíƒÃ;žíþÁš¾oö=©ï7‹¿¶‡Aå endstream endobj 58 0 obj << /Length 2546 /Filter /FlateDecode >> stream xÚíœKs7Çïú<’ƒ×`°7ÇZ%Ê–+Y›[>ds˜#‡[|h‡”oåÃ/f&…‡+ŠE©\%QCNw ÿŸèÆxT¨x$‹‘ IVæë³¢¿Ú~ /Þ}{†õç¦êƒSðÉofg¯.8áÉBâÑìjÄq…¤°æf‹ÑOãåäçÙ÷gŸY;œ;:ì>yË£P"˜³Î£Š©¬” ÞfÉÆõ/«f2¥‚·WÃ÷7ÛÍ„âñ¾¾íº˜S QV’ΔúíF+Kx2ŘðñåpK»U–ðxq3ß/•±þ~\ÐV¯yÉ”¡ÞÀß”R ßʃŸî~1÷¾G»üÈ‘èL ­¤£’QTTƒîˆ«·P!ÕLz¯@ôíSp¿…Á päDËx‰rÃã×}2á±ÃOñ/ðàŒhfz oöÛvY¯¼¨T ‚’”¯Ÿ/1€iùl–ì픀¸ÔK’‹!Ò¡»Ãĉ ¦’qÉDÓ„ºT…J_ëëUó»Ê@”ŽØ4ÓóåºÙìT"R|©‹l|~1ÓI­¤†+Ž8ËÍI_7¡L£²0ãv(ïÇpÿð(vì&Œ $ÁHÜ¥…¤Â*h“rñAòöfµ_ÞIQ%ÄW;ueÝC_>Î@ ¨÷‚IuiÁ(E·üÒ\PÍ…'_Є‰~‰¤®¼kì{õ¾öAB”›’ò§žKBX ƒ—O…½?P¨ô2á¢H!óg‰à UX$˜F›2¡P¯Ÿò¢A\>Î@Œ¨ðIui‘P+›²´HpƒÄ¶mò)À¬TFùiÍ_èáXL*Q7ZfüïM–¾ øÈ:r3A…-I ­¨OŒ%ªTÄI…8®,Zvý:ûU³õ]½ºšë”éÕ¶]×{».ñb%$bey"«×zÔòpb@@q°·,†‘¤€ˆú´@`Žh)f­ê2JóiBĸټÚN¦êÅbªýþ]ðb¾Ý-7Í«îËpqßÖ›]ÐN½½ó/Q¡–Ò:!)õï›/¥3“k@JIJʨO+e¡* ¤¤ÇÛçËݼmöÍ àwu»_5Ÿ‡·f“ŠŽdþ2UýfŒœÌ*Â¥-Ÿg ÆÔû‹PIŠ‡#Ÿ·:^LPpiJÞ¡åõÃþצš­—ëë¾Wºm÷õÐ<®÷]±íõr¾ó·OUñ+ ñšb·äVÕs‰íÀÜ_îáþ)0àëI¹äö––0’„ÜÇ>ý«F¦ÒiYiĨØß_¾=þÄëÕòãFÕRõ¤^o&…cºïÞøÍÄú_á*™(÷IúÖÇO?£…zó{5RLŒ~ë?¹QDU¯V£÷gÿôÅÚ¥|®†ê Öu½ZuSK5žûRòôKÚû¬?Íèdç$` ’“D@ "IAõi!eqj!%Õîâ¸Ú}ݶuåçÖ‹I¥²S¿$õNPêoO’§µ·ó… èÑÌňE €"Iõi3ŒˆV¥Åæ³wÛ>MM×õdÊËñ¶íÀÊE·€ ÃAJÄ8>å&ûCmѵùl91¶ Î¶@$)¶¢>-[Ýî´šR,[®?¿ºYo,Z“nt§ÔÃbÅ³í­„ÒÃœO3#ª‰½-[IŠ ¨OKšÎŠ dJ­‹åïÍbº[þ¯3˜Ö¥òr3|ãÇ#AåóHR÷ÎGzÈóirb4Ae4HR4E}š¨T×Ìñ˜Ž&³ pþyS¯—ó£ÅnýÆQ¢¥*D^¦»Hz2£ž 0ê@\ìÝ@€‘$€Šû´@ DEPü^@ýaFú®Ðvóü¨W來éÜþõM!Áõ¨ä î ćƒD’<êÓ”U´ìö}­àf6ú0¬‰zò1›E–»ÅvýÇû^ÉOËA]6þqUo'Tª(yN%¸Ð|Vœ+P7\yY‘¤X‰ú´¬pŒXeŽÃعæÍ@C÷¥©÷G)á_»¥Í=<^LxJþ2çÄ°ÒcŸ•3à JÀ D’ÂêÈç­n3%j,õ9¦»Í³CSù]sÕ´ÍfÞø{ÊLªßûÏï)…ÑÐ6ã•Ý„"MèYˆw F’¢àÈg ¹I®³M›þP‚Þgꪡë~'±Ùu¡/–«Æ?©º^|’½½ÇhYòóŽ3Ë;Pýq ’qQŸf­K$AUA,qþ¹ý¦Çͺƒï—½ÖÁ'tœá÷Q Ù„ÂÔö#I÷i R­k°#Ì´ l›ùrgŸ59$ITˆϽ—óøg·ŒXù:1!A$)£>-‡e©âš2ïm³Þ¶úÇk»—Zïdv{!•|9xçÖ£ø|¦œSP_B¼LHRLE}šõá™Þ9³‰ ”{Á¦@¥JGB_(úKëI-_>˜öþ—€‘–.Œ•1‡J¦"–ÒPiÒÜ7õδ./7ÝJnß´Wµ)+é”1)_–q¢–,ŸDg †"$ƒ0/‹ ’ŒQŸvÚ%ÌÑè/0‚mU˜¿¬¤…o´ÈÇ̈a%`"Iaõi1ÃÝñ¥Êaf« 5osÊeU p†‘(_Vuº `4ÉÇ͈á¥'Þ300’nQŸ·¢@‚ ‡çþ§ Û¯ æ5L©Ø o_öd‰‘!Ÿ0g FT›x÷$a$)¢> a¸ˆK@ 6´N{&ô‰*¿ˆ}"p˜̆ˆÀq ñîÁHpÄ}Z8CÌü?.±ßNÝãGO¦øsLj>¿žmˆ=´ùÔ81j ‚Dz©‘¤¨‰ú´Ô”É5åݨ.ýc¹YìBÏ0rFN<™è±ËLja%¢Þ‡Za$),¢>M¿³ af±0óÌëŧ /ÇõfÞèG/‡g_ÂmŒUµÆÄ˪æ!WÑFž|òœyêmÉÃHRäE}Ú„D’Yòloà6{wëpÙŸà9ý©,ÄŠÐ|Vœ+P·+ ’+QŸ–‚Un¡Ždå®5X©€+È)g«&z,ó1qb˜@ɨ÷©]I “¨O‹IQ¡RLhµêiï³V–ªnO©B@S>Î@ (GI ¨O³ž‘¼_ŽjLeôíM{ãv?ªÄî‘/½™G+¹´`Ù(ºû#$B*¨·' ÂHpuh2QOu‚ÿ*Æâ·jÌ™a1nô>{–ow½Zîõ5÷”wã¨Äâ”Ïö°©rjlªØ‰Ì#'lªÈÐÿÇzû endstream endobj 132 0 obj << /Length 2997 /Filter /FlateDecode >> stream xÚí[sÓÈÇßù~txxÊù g¨ÌJ<[žCSËÕì÷yÓœü¹üyFs‚D&¤•áèééò·“¥tN‘ü׿åÁÓ¥ñà ùÄ€úwF"¸DEQÌry,£ùC‘“βlþÓU{5óþ„óú¬Û¶ýÏl^mVã ][mvçÛöbüq×ü]ïú ž :ñU”ˆJê×ü—ô òñ‹˜üä<ø où̃±·ìéÀh‰úßL§®×qùš<*ÿÃdezª£?¿ í»rGÙA4ÒŒÄOEÿΈOM…(…ü ¨ T<Þ^\®ë#ON—nå©@BºHSþ–âj̃‘Îu2FÀ@£‰¤”»0‚‘D0 û4åTXŒØF/ëj½XU]‰aùûóïvù2|©4§d „‚jRá$D#(èÓ”Ëø n ân‚ºí¢•_Gˆ–'›ëk“§¼@T^$ïN>,TîÒ±°BX@‰hîÄDÃ"èÓ`ÁsÄ(ÀB(,[¼iº ‹ÈóÍ Éç]ÝžWgµï:ÅŠò~ /‰åŽÎy:NÖ@'(­'I § Oœ1Y"³™` ‘‚iœ4Lÿ©¯UÛV}ýûqZž~8ÁóZ¾>?»êê‘®Ó“‚ίäñÍYÿB'¿k¶÷¸#J„3~8󱤞Β5b êJ 'K ’KAŸ†%*+J5K¹bé·f·Ú^¸˜9"jÛþ¶–)wBÕ¶r۵͛+YÕ,.ê‹mûÔ·Ãw×M÷nüî—Ï%/éçY7¼ÝÁStê“‹N` PtNRÁ=(I$óû>݃å•B @Âöî§>½ÃõÃÔ¡ã–Jw³[ÅqÒ'W-:ûɧ408¥'"{À‘ÄÀ ú4`I lÀÒî¿›Í_fÿ‡™È<ß4]S­›¿ÕKLcç/uI–!¿û« •½t0¬P$îlaÂHb`}0ˆ—…ðM9KV{ãMý¡êojsŽ3™@¢uïžfQA `¦ªú±Ú5g“K›W«÷'\ÈŸÏtsg¥‹_Óðkà-S;ß‘·§u§s.¯5’¦Ú#/ˆ$&oЧ–—”²òÀ@^SÛlõüùMµ–‚öÅŒóê•#~_n2ùÇï Ñj$ƒ @›ˆÎw/ÁH" …}Ðòe[Ðt­dûúî>ßîj5˜L†l £=×…P‚2ü [e½uR’õzOsïÔFÑ;ìÓè]p”eÂê­gtgéâ²ö‘Œ7$¨öE[_®¥ˆ²>$õ=߶ö¤‚ÿŸ‰Šâ±|pHØP–þþg6[É×~–™“¯\o¼˜QDòþNÌõìÕƒÿºb§„Èœ‹iìÖën[­×.®hröU[_>TÄé Y!&¢:{æ0’HAŸº0Á9F7 þÂ$tO·ä’cßë3‡+%E:eÖ@ˆ²‰ânÊ@$1Ê‚> e\žÊ™Y3.õ`5<ôṯ۬¡íI]­ÖÛ¡ vÖ`yª`q¬‚?­øѤÓe „èš(í¼¥F£+èÓÐÅ8âÜ,<ãÌ?ˆ½ž²ÛöÞ¾$RóàæÒÛcÙ~(Š•ÑtX¬,ᜋ<0’,AŸ ·¥!¬§JåµM²ñ¦ÙFf;‹Þë„…É"˜Ýƒ¡ÈÇ MÝK§™¡¡}mûnBÌÐOÞKöi˜‘U7áfc&þæe}^·õÆóL–7ØÞl R™O‡ÊA5Øù8LI ª O3…ËÄ°úb 2]»E®Y×»½!Èöq—Ããd½ö>“ѻɛ•Õt`¬0ñÜÀ€HbÀ}j`J‚Jy!û¼˜Ý–•wWˆ%ÌwúÑ• TÖ“y²Ÿà4‘Ö¹” ˆÀthþ …ü’aÀ5÷©–?/ú…¥Ê³†LÊŠüÔÿ®Å¨@úŸµ0ŸýU (³“+ÆWÿÈ ¿î endstream endobj 193 0 obj << /Length 2211 /Filter /FlateDecode >> stream xÚí›]sÛ6†ïó+t)]!> îÍNÖ±;î´ÓN×\ìî…*Ñ.§éRrbí¯_€ÀC l&öÖq&3ÃPç÷ÁÇ@-"êbÀ‡ü¼]•íÕY¯ý}bSŒd6Që§à/D“Mùdš€M=eÓÄGŒ$BSا£IR”RhCš>ûMµk{³j·[©:‘b=–dˆpö‚z‘GÁ`26†Î@(Lʼ0€Hb0}â„©ñ‡ÍDªâ—ÜÁ@ WZ f(©êCm{šËrAÒù!¯¯WëÜKEÊMÈK§â7ÇÀ1ÙNg qI œÏTùdˆ2A´O5c #–1ãN¥>?[m·Ey£ÈH¹¥G]]×M?¢®~¬Ô'žoòº4-2Úð¤¦,+Ó«(ö 3"%‚Æ€u" ER`÷/Lûý%0àIÒÐM›©Ô+ ˆ$&ÌÀçH‹&%Üö"›ýÓG݉çõÇ"ÿÔ¶çêºmƃvÍæE© µk¡¸#Ù+hבæmR=½ywBÍ»§¨Ÿ"IŒ¢ O7IÀXÝãŽ"ÇÑùý!/7ù¦egUš‹?ïV›úîv›·ÝêF­®ñ|]ìM1B5X–8j‰óàE8FB§¦ŽIj~÷tI;!I{é•^IA$1Iƒ>mÇÀ3‰Rî$%FÐ_rÓ3ìÛî[©·ª›>YîÇV±M±ËËýh Éò/[V¼Ý®NˆöŒŽº8»Ûª]ñ_pŠw¨ç(yê=‹1‘ÀËÓD¼R1"ÌZæ­aèƒ^ãhE ¿Æa»pœªø3«”+‰ZéžÜ|þZ÷þ<ýŒ½Æ¯óÿ„é‚dz×ÝuÝP÷Ö@$1Ö‚>kê“òŽ5e­¬ÊeÃ[3‡Ø÷‡|çŸÀÊ ‰¯å4·§T¸ôMé…ô÷—À€—Œ¾£’÷¨ %†FЩEƒd‘Ä ®>;®·¹]N¹³‡úý$\];Vó¥g)š +Ïdô€z ¼èÁP"è…:ôR½/š:ôlÝl׳8G½WÝÕ愺þ×'ÓM­Í>ÐÈD™À¯®¶Cɤ{:JJ}U½µ %†ÒÀéÉŒWï®gÂúÃfÆûn­K?ÊJ|Úæ››]^.(Ñ Â¾I.I³×ü4%ȳQ5Ì5á(UòØ=Zzóý%0à›FÜé½Å %&}Щ– Ä¥íµˆ‘þ‡b—û|XçT·…c]Üü®/~ 2‚Ää]ãÿϹ›|“ŒéwB÷ržx7ua(1ƒN•ÆÍK¤®L¢vW·*×ùí¡•ö²Üä÷~9©~¿ôVý˜¦nó5` €A_–Ä[ÆÀP"„j 8F’š“E˜¹¦þ[½ªq tAž}•Dh0i›NCg DCOÄ»C9¡ák†é÷ endstream endobj 9 0 obj << /Type /ObjStm /N 100 /First 837 /Length 2215 /Filter /FlateDecode >> stream xÚ½ZM·½Ï¯à1¹pX¬"Á€A¹$€aéD؃#/#Æn Àù÷ylÖØÜmØZ¶»ì™!ßcÉ*ò±k*IS/‰8QÕD”X-UO‚GªÉZKdÉ; GEO„§‚gê¨ÍzáÑX41ê‹ÖÄ@R° ½uà •ÄšÔG•hnh:þ¡*k¹ªJet`tE“ IuK"hŠÞ ªVJPut[ðS‰Œ>¨ÂŠM¬\UÕ))¡I“¤£I¯IUa¢i‡1èºÂ|%ü¨ø‰+¾l¨jø€*êíRá±Úzª¨ê…ºª$U4éèGÅO­pc18 ˜•à¸·á_…á…†²I»4Tì@mž¬'Øhä’ÀaŒJ“çS²Zz‚Íàí.ÉÞéŠQ‚° vÙ¥c¸1š8Áñ:‡§Rïɵ`tà<¯è†1¹ñÀ·î2V’7ô…0@ÞmT® ^Ò ô˜Ð‚ÏœJKMÑ!Â_«c¬0ØK˜<ÍáDÌØ‹‰†Ap#јo.Ì®N­_“ KÍa¤Ž™ˆ9Öë˜8—nTzw  Û9Ë»ó˜|ä†Ú¼Ù?ИlEŠ_h3OÇdåÑå:zÇÃRÃD má < £½y=¬n>H·)>ü2I†clX§Ã3ã㪗gÏ.×WÿýÏ}º~ýí¿î/×?=>¼¿xÿ.V×7—ë7÷ï?¼}sÿKmûâ¯÷ß}ÿíW?¦×_,Ƹ» ù[´ûVí‹/6ä@KÏž¥ë ¯ÉÇ/x„GÛöˆJׯß>¾yyÿ>½Fž¿H×W÷?¾Ow·ö»=Ãh|Ô5ŒØSúöS‡®/°fIŸÒ¹Ú>î\µßÓ¹ë—Ày=ØèÙ,¦÷hvSb\fA³àYÈ,¦rüæ¨Ëõ凾ß>ÿåû‡_®_=¾ýîþíôq¹ãðã§Jy„•\¢4Vñ\ U¾ÜfÁËtýóã«Çt}žþ@Ãÿd6*yÄîAÆ%›ó.™<•Ls…gkçŒ)³±ÊMýT6ïUm-[ߧª‹¨Ì3"öÆæ_Ù.›-b«š…qÆYzÝ¥ó'ÒIÏy­bª˜ö̺OGe[–›7ÅðîóÑ">⌘ºÑ±ÀÃý\ºRPÚä+=Ï»Ã÷Ôe ’›m›¨ŒÚƧm .û|u}ŠÙY-øL3›œË‡é©|•2öçòqƒM3Š©8¢ËG5MÝÆÇcåðÙ">DæëAi,ýsùáyìïŸô–¥¬¿EáL ±¿Ÿ|­f©'ó!B#ÄL>çÌå`ýµE|Ñb3§Ã6l Îâ›ñExd à<×ýpÍ‹â§ðJø“¯¹ŸÊÇû¢ö!v‹·sù¼g«“›gœ¤ÎåCŒ–¹ßd×\ü€ŽÑ!D3ÏÝ[ÉOæÓ’©Ÿzvßß½°.âCˆî„h÷àc¤B?à[”ŽKÀ5ø Â5ï§[^> !Z}¦[êH…G^> k@un'¨q.|À×ñÖCû ©Ð÷Ó­,Ú]“¶L|µæ£ߢèBÈÍgv'E&ôséx+ƒŽ;6†ûÉAd_gËà#ÃÆ°Ê×Çáræ>B$: [\ZÉÊZǦÐÙóôº†Šö·¿ÿcˆ«Î%=|øᇻ_UµþO|ª•›Ú5DŠ§Iqü -Nʧé]uŠDuŠD-Š©µÙý6 jS+êÓŒ>-èS+êS+ê¬O”>QúDé¥O”áYR”¥D©Q†WB+¡È•äJàQàQàQàQà…¾H7}ï&ðݾ›Ä…ÈG¡òQÈ|:…ÐG¡ôQH}Z…Øw÷ôë„ÓµÌ9ëäÙNƒ«VˆuͧA/HvO¥_à Þã4o²Ðߢәyɵ…fçH§AY”î­zV >«¹œL‡”D&¡¶b'%uÑî„ò¸Ùø¤e/ûùIe{£:E MàEò88 .ã+œ‚8k9—¯"/ÕÐÒzÏ|°½ÐUÚVÀ¸áš²²žÍ–ñ™À¦^s¯|¼Jê…; ‡²\°}¢sùFé-¤eùWÏåcø³Äøápvdž¬R–‘Œj(Ë8ê(Mº(i·Üâ¨[‹çv´øV)“X꡼"–ºŸ‹Ô)ÙÈ ¹H Ï÷Gºèh¦Ð÷,Š•mʹ|Ðæ7e¹f98›-ãC€oLe™3óþr¨«”;Dh‰³ –±òO¢ aÒz¶Î'y±ýàY­vÁiºÇVPl\$ѹ|Z~:{J-Yö£K]]„&x>е–Ñá4]o׌àrp-Pm¬,·kÄí~°\ÅÇ­ÅuYIÜ=ûÁµ\m«deÍ]CUÖlâçÒaܬsÊHût}•ª\rd"®mÜwî_À/Úv2æ¿•›¨\3µ}QÙV‰ØãŽEn¢2Á³r.¢3µà£–[;à[%£açRâTË¥âv2ÒAk7QY°ígZÓU¢2åçòŽ.ËW<›†¨l™Û_]%* l ´J>šž¶JT.0)dP|×½ŸËGãhyS•±É>¸r´UwÈE=®‰›Âv*_'Ë›ª\°'ÜÏ ¶(78‚ç<ó5lâý·=O¯)^þû}²òÏòñ‘Äü©²rÿå+ž>ñ5ÊxË‘¢K]¯ÀÎ2tT Õnï]†Žj¡£Zè²xxxxxxx~{‘3ð<ð<ðZàµÀkׯ^(ãÒ8µÛ›¡ê8…<þu_ë«t_„È¿¦ûzù¼º¯ëçÕ}æûÈ=»k endstream endobj 221 0 obj << /Length 49 /Filter /FlateDecode >> stream xÚ3T0BCKscc=K …ä\.°hQº„äÎeUgbl¤±*OS endstream endobj 198 0 obj << /Type /ObjStm /N 100 /First 889 /Length 1726 /Filter /FlateDecode >> stream xÚµWÛnÛF}×Wìcò ÕÞ/€À‰ 6hƒ± ·M‚‚‘h‡­DåØýúž!i[)ÉI$À‡ËÙsvæÌÞ¤ÓL0é “Rái™R°cZ;<=3Æ㘵ÏÈœ‡?þ}€¿—,Dø{ØB¸ô@ž­J“«eRòž±ä @ &é ‘Þ|¤ütˆBÂL 0°(©ì@¡Ñ˜‚aJt ¬qä옲D¹ ²Ÿ/†ÏF‡˜tÃ'×ú¸tA#¤†ÎFÍeè¥ â@t^¡BZº ¸P±ŸOˆÏzN+xÍç~K|ê@|Æp¡[>«9«~>} >-x³r‚ÏHĪË'‘OÕê§bíŸ}ÁˆOh}Ë' á¸|Ø0¸ÁSó äSÙ~>û£|Íê‰CwØIˆÏxîþÕ%h5Þ‹9×ÐY˱Ø—NC²»ðŒâZ™~¾­fH÷²åS‘ãx\>,Ðt*­ù¤ãÒl‰/†OGËéXó Åcì/ÊÔ§vš[Ýðiï¸ ýËuÔâÃ3šfúáTÃcÿlˆ*Oº9Î +‚=ŸÆžîè”Éé$^óIÉ£îçÃæ0„Ð A5úáÐK™#z‹¨šÓ™ –N2[´B!—‡„¸´pºÆÕ„6ð¸å@!…>¡Ä® qã8Ûo!´‡!ÄñÛB͇ ‚U[øüø jÆ4G‰íÉÇm>ZÔÆìŒ ˆoÙè?ÿb¸†y\góõbñáÎgRäU3ÁÝ#4¾\ð0óW?ßš(Iœëô•Åì<Å0Ùèl> stream xÚWÛnÛ8}÷Wð±y"Þ$(ºèÖë&À^Š$E_,‹N„Ê’!ɹôë{F—D6%ÙÉ‹I‰gÎ ç ‡²Ô!ó˜ô=&ÉA1îÑ0®i ±p $FÁDý¬™X|&•˜ÉÀ0å-ô˜R4J¦=à€Õšž¦ =‡Ìx6œù>l`>½7šÊ`ôYú3i ¥ª9CŸFÉ 1*f4=Ì„cÈŒOÎ)î‰ò1ј Å}Ʊ6SÜ`‚M(À9¼+!M°m!À .È \¨$˜¥³³ÒK0k²’`Ö2œ) fÚmû*¸á>6­˜AK`°=¥(¥È‰Ò`6H¶B¢¹™Ò`6”J¸ž§fJLF2qÊ„iSH‹ä Z!vXA,¼˜Ô’¾Ær‘ z Ei`Bo¦àF x¦ ™Ð>MÀì“ ?36 $IA69tíÔP!о œÃ;ö&Œög £© }šM&Ù.m”žÆQó¹ÃŒU~Z`¤çë"ÊÊU^¬Ç¨‡±Ž¯Ûb»›¦šÎ}íXÖK©§QVå&Mªú¹(¢ç6¸ã £±=Øe•=˪ÛPM“ü²ånÔÇŒz¬„±ÅQ–‘ _žfšPþœÃí«ÓÛ&÷#úO¶íãi­!áþz²ËmeiºØfË*ɳ6ÐÃ0‡ùGRÆùº1oç#˜†s“UÝ{;aw±²kß:v‹¼¸³U•dw„êG8¸2á÷{•à$]õ-¸Ö÷Q-Ëbqý£S=}îJíÈãîëûZ^™yR. Û*”o‹Ä;UE7Pº7'Çw[ODZRœ9‡±‡³6ë]ÅÌÚMœbÏãxÏæj| 7ÂÝOçyTT©}Þ=â-Éù¾³·:ÞÿÙ¦Ur'k›•8¡Ã èh„µºG§Š›ª´Oæ¬ Ü>e…ŠŽ¨qôýÕvC§¶±Ä†ãǨ°gWùª¢Iw½ÇÔ‰à{ÝÕyÍWô;µ©ã Ž‡óü±†GY#ZmP+Våu›(탣çŒ`õ•­lõÜ­½·ŽŠª*’[4‘øtm×yñÜï4Iu_óí¢!<îöÚ„ÅæzU®m¹‡íï$ûÙ6çVó‹ m6J“_íë–æU¹·™8E܃اˆ>âÁ÷¤4ëó®·u™|ÙÿÀaû3*“eo+Qï›1îYÒsÒ]÷í!§­«AÞÔÖm”Â9lóì¼v«•:Ê&/]ŸmˆS‡í¿Ìî·¬ʃ¨ã:f§RÛí›.pÙ~§‘vô°­»·ê¾¹n wnË–}¿¥¿0¬Gæ£ ‡Nñ‹Nåàî-WpÜë)Ug°S{{e{„Áäÿ¯c¾ÑÆõÙœ õ¦ÈÛK…ݤ€BúƒnÕü±A²þÿœ¦UŽv×z~å!­v>o—Ü<äI¼Ûç¸PÒ|ù³û¯:…8Ñ7deFI®«%Ù —ÑÎáÆ·ðm’í†Ò6¸öm¸ÀŠûÒ®la³®NFWÇî€E’Ú²ç·;×Ï›î …ã¹×^o×ßçþ endstream endobj 696 0 obj << /Length 3034 /Filter /FlateDecode >> stream xÚµYÝܶ÷_qèKµÀIõ±’ú–¸¾ÚRõ¡)Ðô'qw…h¥ )ù|ùë;_”´kÅi‹œÈшœg~3«îbøSwU|W¤iTeå]}~Õïxð÷¿¾RÂc¸âüúñÕŸòüNÅQWêîñ°^ê±¹ûWðú¤/£±»0MÓ@ýyfY¼ëwIŒvØ…ðl¦zl‡ǹRÚýûñ›Woç]ó$ù/ÅCÎÏä+@¾"JTž‰|û2‹Tš±|jæû%JKO/r« M"UíùóÇSë@ðBg JÁ´Sî” êWõ.TD?ãÀ°îòÕGœëXø&àß ~><<~ÏÃñdD¨r-T’í£*ÙH¦‡]QÚÆ M¬˜0L¶¥sÂ#ì.Q{wì™ÉmϘv!Â÷¸Ä†T™Dqî÷‹å.Ì’2ªŠ»P¥‘Ê•ˆ¢ÃRù>@3áSã#êá|±ædz×~D³Oî:#Þ€Üh|4é£×Ìh‡il{#‹‚ <ÀeñÅ‘¹X 7­«­eö°+So=¡YæÑ.FŸý!Îã¿<<ÂC!!“5üæãõ±í09~ã.tÞ¦nÑRkG27ÈbÍpˆØžûlmO0\–ƒU$2ü«bÓðI´ÏgóuáðEoÜ*É VÉ 7È•„¯3Ÿ˜­Ñ£F¿Ú§5ä³×DÃ'Ó‡²†•'_Ú&t/ç³m[3ýz â%A4‡Âí‹ÀámX-¤ÙôŽçÚoLnÒ]÷ÂÔûoÍsÏï´0³Ö0àÃÆÕG¸t²B=8ðf M€¶Pæï¯ä/¶†N.*¸&ßØîâžV<"ÊÊß`öVÛ±3/³á^og÷JÙåϲìMhM°ê¦÷„â¡ÚGiò?yÐ#_®>øåôˆXëžI'J/Oæ©Mì sv¦?Ž§b©*æKfo£³zq"xz¶Ê"JãÊ ó~KÞ2Ú'™ç@cm-”EI–z¦~{™t¿÷ÝpÜZEå8üâ2û(‡Ì²HgV¨Žè8Øv<ëñˆbbšÂ·UuíNàÉ|¢l5wÏÞöu75; ÃG¦\l{¢¤1ñÿ»¢u¿µ3ȹÁnnº°Ãv”ðµ>p˜ž)×uc6 cùë÷ôÓ/Š&·. >|x#7F‰ ¿Ú)ÿø§Ÿvþ ×÷7ñüÛwß}à‘¶½­G'áÍø;š=ÀH6¹h‹ž"Q [ÂyÙ~œ€Ï-áO®Ñë`²ÉSŸËæ{@¹ fI!?]ºÖIKTëÚþGüåÄ|]CðécŸ]àèQŇ0Ñ|oª(Ýi³wžÅ6\´Ê]Òä |À΃º…èÆ÷Ì5Gdì ‹µ¬Ömgqdá'íз¨púÚ\F¦{HˆÿÑ1­åël,æÝÀ‰k²KîØ ä™¯J)À´F~8ÐüݲO(6Ñ<Âá)¥¬ .œ2ϳ5 †Íb¸oÖd#̪¶d²¾}¡ô—„Ó»ª8è!e% Òɤ@n̨ÛÎ1‘l¶| ³‡í^\äC[ti9Zª’+`üšÞQâµ²ŽlrÖŸÚsû³ˆ$e 7’ j¯á!¤dI¼50FkT¾T) 1Ÿöí‚ µrű*pJ݆Jv£Øá§ÒÈG" \NXƒŠ­u# ±÷Þ´ÄoãÒé¾çRù³¸®ž£ý*ÍȯÝÀ<¾M³’êwnG•Mî0² Ì%UìÍ·ãf^€tó…´Å¤iùÅY׸ú‰.y˜8ïJˆC,ýµ²KSŠ€XA÷!-r_Áh¾[ü¾FÒð+½eOµ¯¢¤œÍ„2l3É¢}5—1Œ1pMˆtÂ.ï8R× ßQ”;Ï)ŠÚÓ“3?MÒfZBR™ûîF¼4ªð¬Ý•¤ÇHÑÇ6”H3‚síc>™zÂX»åy¤Ò¹Ž"ǬÒ+G®´'UwÏj<ÀžÒ@!¸ Å!GTêSÓÖ#Gz¤ú\¾Z6%uððSŇïi¼@Í'pû÷7M+)°Àfð°âbÃcJ?p‹ÎŒ¥^˜:B$/˜¡¤poÜ>K9¥ôÂo.€Q:æ>Æ= ôÑBZ™ûÏ7…˜‡© nEOÿ•tTâÃ`8ÍZfÉR !÷@ðÂM¿R]ø’Òï’ é{óIB$7úJ¿qÛ·c«»ögí­7Ñ“Ÿ]F*>µ¢œ& †yRc~é(^Ó¡ÁûÞox0 ”åáÉkÿÀKùBÉÖŠ¯>nÝ"¾“Wý\öpÅy-{ÿáÏf³¥ÄH>FLb‘Êü,FãcÄAþX-Œ>rÑÛ'i4¢Â¹²¸öØ#ÆÌjû`‹}VI¿Óú['›iÂõ˜Ñ/F´ÆǶñ SûuC ‘I À¾ib3òÝØÖÂÂÕèòþbÍGìŸvR’®œ/åûê6aš@0pËÛ~®j`%Md57¶€ÂŽÖm ˜rG†n\ŒõÈ÷œççvšn&ä^ YP”3)ŠÜð80Ì64ó5x’'A'y°"žW‚bÄÀ—®Ån-Žw’ï¹oyæ$o¤¤–7|¸È/û·£Ž°ÞY¯&âB<™à!ß9í墔Ûðnþ*éhãž¾Þ IéËÜñôQ÷ppɇé"]`ìA×ÂJ £˜; Ó ïOÖÈרß9Ùåà?¨­ÑŽa)ÌoŒ"þ–møÇùG™ëR8‰²¥u9c{V§˜uÞX¾H¢DÍ©véíSNQ>"¥émGÙW 6´ÇI~¡øbo..¢„~¾ù¿šÈ¿ª¿n¤ Q/Úúè7v©`ý¶nÛ\¬3µ%¹5vñÆñ»H,Œ9°-ýKͯ"ÊÓêw3Æq²Óâ _4BuS¿ÜýÝ)w ®>öÌG.ífÁYkÜýg&øÖ7÷E9,Ö“X*X*÷ÍÒ›¾Y–,Y¿™ 3±”8\…™‡Ï-¶Äq4c¯ÕÄü|Ýÿš£oà¿á{†Ã«{†‰n_ÏØÆÓîæ#:H}®~–ù ŽÖúnm ê0‰Íõ8J@œ;ǾO±üDÄuF=…U{ætnøŸ]àÀZŸíðÔù®ímZüÚ8NÚIUy( `r,jžúšuAœÄ„·/D…ËãRÚÕ h¯V €Nï*JÚŠܼY¶œkAþ_@­À‚¥áU¨@ µ&Ù×õÉ[Ÿ(G(§’3`$p™6¤€H¦’Σld>iL°˜®%!¤%¹?ÒÞÉ]83%R|Sí*Œ ¡íéóÒg;éŒý³ú¸ð endstream endobj 716 0 obj << /Length 3598 /Filter /FlateDecode >> stream xÚËnÛHòž¯r09lvóµÀ2Î$³™‰œ©-qC‘’Š“ýú­WÓ$ÅŒâÀ€Yì®®®®w5¥V!ü©U®R­ƒÜd«õþIH£ívÅÀ_¯ž(ÁóÑaþvóä×—q¼Ra‡¹ZÝÜIÝlV·^tññæ÷•N£  Bƒ/_Þ¼¿ðµÖžàQž¼¸¶‰£èùAÌó %™ ”6¼÷¡m.¢ÌûRn,°¦^Ìë› ?J½¦’×]Ñó|UÖŸ;›ºúÆP¿³Œ×?uV0›;~Êñª- )os¡<hør ”¾!Ü[&WU ØkÄ.ú²©á%¸ðN¼+xAÎ/€MÛvxÉŒwI½(¨˜Ï(4dyûÂí/ÀІ!í×ÞÖ4¦ã0ú „‘yv];+ø;º¾¨7E+K6e×·å§#±Jóe'[5½,8~U².-±/ƒwM˾áqðüãc(>rÛÁé3y//2íá"“ÄÀt±?TöÞòœY3I2b ßø4 “ JáïIH-ê_÷M×3d'ì¢ÆqôŽ÷N¼¢mY¢ÝË÷»†$¦àÜåÿZ(¥œXyô®X÷MKú€©²FAVà­ÛUÅh‡¶ÜÛŽáaF—'ž–g,Ï¢J)`©é`ä5 ¤ö¾t¼•=ã‚z…t!LWä$Û.ØÖÖÖ¶-*ÿplt:o”h¯m@ºµEk_ßñè`‹QbÄü²3›À±¾-êä¼ïø½Ä;¡L»KôÔlJ™DŒÈGfÇúYÔѺ!Æ)Ä9œ­v>Ú^ ç5-J0IA¡å7Ú12†Z|\“ä’ôˆ–ƒ“W<8Û¶ØË4êì„·«Û§r@ß=E1 ö XVÔÎm‘b\õåúXnÞ™!¿’±-œu’Xƒ¡#½¬‹Ö2T°Y‚ú@w zÎàk0;òµ(ñÊ;ÑÍD­8v}³g‡‡´æœa‡†‘îxwi{ဠÇFiÐØI …ƒ~ó@EÂé ª(7ÃP¥QC4A%p¡Ön!0–õ–1‘ÁKžé¬¬!†pòP0«–| Ôšï›}n»r+T)¼ p&²vGÛ‚þOCn®/ŸíL=äÂ4òÞü|‰‰­-· ¿ràÍ+ï•àýÞìꮩ/™ÒØSa’BXÑ9ü,9…–=…] õpÖ-$Ÿ˜cœi¬œ¥½• °>bpÚë/^Ȳt¼,‰(;dQ¾h{`g*@xâµç] ê`P5Ú‹T‚#„aŒ82wà :›„7xrÀf ƒWNÏ8Ý­1'Òá72Wó³w¤&ÒÁmÁž1(Q˜#mA/Ö–ƒïþPV¶}Š> ?¹Ò?>ɨQÿ‚ƒ|á8òµ Ç øD;•Àà;^K‘Hh ò$‰‡Œ²kö2v@Ëš1«C…'Ì~¿rØgX››L¤„e¥l°Ô ”eǶj<˜š¿4ÆUë½}ݸڲE âáýPáÛ–}ÜÕT¼pc A¾k›=C½cå@‹‡JH´(ÍÛ¢],‰?E @À€1õO-VºÐZz×5O4µeÔ{ iÎq€J)À(j2Q´ÛA¥8ÏU8b0·0t%ç«ì7ÿÃáñ3q)‹jÛ´àp{¢5DÆY¢¥Öé$C‘eéd1&é¿Šméf“²Ž%´Ò‡"<çÃù‚ô¯½ß›òë"= i)DM-Hìq±\ÿ÷¸åû‡¢lŸò;k[º Cb&œÖ9©Ã(T² p;ƒ€ù9?/„Æo¶­»Þ–µK¦:sÍÍó)¼>äÓ¥#Tag£FÁµóV$\Û{FDÓñ;ß9 ÄíÃ!Hý4˜A+ÛcÿOôŽ5¨Z_b>dïÙ֢fyU)›´2v§æâjU :à½R'éIè—ª•KÖvq€3à´¨!£Ý0v‹Õ+lj8гÑ}" É½’¸]¤…±ßí‹þ#]’%âlA÷Ø+Ó-\d‚(3gtï°ýúRî˜ëâö:rï-ësp]v{¹q…WªeF«ÀÏ/”kaÅmYL‚„'Ä¢Ï ‘Ê©°;Q;x¸U>9ócÔîÖŸ‘Â|›™ÚGüºFÀí{V覓KL#;Ì{ Õïû²Û4{ÿ])ñj¢sï- „´/¢u€TvTã(È5¤%h<Ïä`ÁöGèK!tFtÐ:¾Iø< a"ÏÒ ' ï²þo³mfy¤jXV·®•Ï¼¿\uL5á ÑÌ‹Ô‰M¥'ú’¨Ã>õɢ Æ>æzô‰ÔYKÎh“+ohCZ»Ã(òåÂ}Õrí $åÙm0Α}©Ü“o[ã6+ó\ÕJM X]߶€jyüm¨´nK.à¹ûˤ —¹ Jl… \ã¾ æ¾O"5uÿ:‰]äì,ƒ" r Ž_ ËzD`)šÍ¶™Úróã<‘ãw¿£ænÃã(G´!„Ù†2ùI:Ê2(¹óÉyþávH°Ïp¯s„`ùî1Ï lLÿ´(yõV¦[œŠ11>Ç O±»ÇkøýÚn$×a.Õ³Lg÷M‹Ý6&Fò^„îq=Â{ÏR蛷׬„xPBìŧíIœ@9²ŠR$™: Ùa/Æ I×›`Ÿ$‰GX“{jêäq`ífäòNÂQå´r¤¢ã¹‡©'×¾J˜ËQxXj+£@§`åqÄZýD[ÉëýŶrºÍ,‡ÊËðVpÁeHna4ünCÉõ!B¯í¶X“yúÐA¢¤€f.`#ÑôÄÙRèoá ãóþƒ³ ö™Óé4 ’<ŸžnÁÙtžqý´°Ýúsì̶9v:ö–‡»ü ï3ö7{oøW9ÛÖ<02g.>‡¤ “é‰"sÀ3q ÏýšÈ!ûØK¾6%ùp>½‚È‚P·}TÉÁ«Ïð1Ùb&x† ùy{]Á®ÀT.#DZ«“cïJ>`ÓÜ©dóô´ù×A†QÿG$ëÿùD1ØB =ëøHöëÐùóäÄP >/kº:½üp3ݳ5 BO"¹;÷¿)Âóá÷Wø†EÍ‘DˆK»o]o÷<Ã]÷tŸ®L³ÀŸu€ðùíažírzmè~äÇÆ{]®¡žB™§¡ûìL¨3kËíÎ]Ö¤áô? °žÜ–ü31œIò…`¬M¨Ð¬"•!ôŽ²ÞXŠ³mff œªˆŸ·³³_ð¯å¾šúú—Ò•ÆÂÃÄ$8ó¦è§½(ô±*KVQytÎj¶?B_J83¢Cö \•Ð`óÏ7.ùN~¸ÚOñEý™A²Ð(MäW©û´a›C%÷»†Qv¶:¸ôKøy1Ô Až¨É¥fYFómf•*C1«·Ï¨àþ\³Uv³uB7è5“MOõŠT¨—ôš©l¥ OIÓôôŠØþý;zFúÿCèiƒ endstream endobj 733 0 obj << /Length 2692 /Filter /FlateDecode >> stream xÚ½YëoÛÈÿž¿B@?” ,†Ëåó®-‹ã".×Æ ‚Â4µ²áC%©8¾¢ÿ{çµ%ÓÈ¥ ærv¸óØ™ßάÔ"€?µÈƒEªµŸGÙ¢¬ŸDíîàV·&7†6P3Zà÷ÄVò+PmÓL=f…ÊDÍIVœmi—£XâK ñ@ò^¦í φŸÏï»ö®+j0& " M„„8ؘ¡°ˆ&ø‚…R*‹Ú.¦"Xd&¦¢$õu˜-âb&ùþÀr߯& ÌlÖ¹˜ÓCT `“Ê‚Ól½y31†FÙí½:ôC[ÛßÉ`1§á ŒÆ Ô^ž> Ê åÑ"úÖ‘é¸Wö¹ <[”ƒÂQIÐ䲌Ê|•ŠupƒñhÊR¯3„5ÚbÐ@$qø†YvŒ&|éL12Ð) ´a¬ð £ çÛnc:ŸÏ‡h‘øyª3R6 ý,ÏPÄ¢ªš;ErØ(&l/ öE&ª0¨LÑxçé‰r2gL×ËG½)y+‘³uîâ(ö“<ùŸànòí<Ü9²èÚ)ö†¾âÁÍ+Bé¯üöKcV—Ž¡^¹HÞåÕºÇp‹F DÖGÅ™Ž-t~bÑÓÑ渿aÄù¢d )qŠF@è O6Kĵ;& ×0ôBPâ "À¨ë/8~ÃD›f§ùB'hÃ9g·œq¬RÉÖ…uó\3ÀéѵzÇìØy¬w²ó¯›ªèî¸îŠ¼KóÅVJú_Zá¢{˜s¢ŠcðŸrkú$e~Š‚®y8˜¤Æ¶•;I¤’sà)¡ “T擆_ á†ÝÆÍ'Ь)Vš6»—ŠWý?ÐÛß\c–œäb8ªÆîí¬½íÁÈBM%­ˆ±x1Á& …$WOx"„ÑéÆ „Q{ËC.ÇJûY<ß·@!„ÈŸ]SDàk…­”3?Ì%Q>Ü?ئ¬X'cïOÛíp¯ýÝ_æ¢Ë÷OƒŽ©ÿ&Z‚Œ³’×úTºã—n±Î¥¤ËsBè£}… €|ûgDž+"ŸZùæÏüÀšeªÂs>vqŠÈ5øº-©²h·çüGöçüx‹”Y¥)÷ÿÏ¢YòþDîè¼O›íðIm¨É®ee„¹õéê—w^¾»œ’^_¯ßüürýúI{ŸØo¾šÒEïyZøÅóÙ„‡ÒÈePgöŒ™…ÔEÏOh]zË`&¸38 Ô¸Öó>$iò{"™Ôß^wíyqbÃSÖn;C¦ÚfbëÉ ¸û¸Äj.Oþ3ç™ÕãTý'B©Ö0û\ŸEU=Íg~q` `À¯÷vØ Ë/g木åÚ9‘à`öÚ*ô¡»wl•½íðHÈ"nïñøe‘ïû•GýC?ÐÞÕý…ƒ0ÖX›S&Õ~œ&N̪b}è›UUÏ)–äà`RJ@ ¦~š¥§î¤!šË†2OZs(°&žC–±3â"²’ ¯‚-ê™ip Úfp¼\šRy ÄTî$¡†x•$ÉqgI4ÝA [%gË5sØÍÄäØ'è8‘wB[–8žÜ²Ù%¸-‰’+N÷½á17½Ôðâ÷¼b=`t Ù¦¦ñ#h7»Yc‰ìûå+Y–*ÔŽŒàÒµ?ÕyN¨€'Aó´t8–S?ÑéY«ôµ4{ºÛ àŲBÙw­@TG5öf•½kz÷áÁñÚ™^ïwÔ{÷õ›Ÿ/™†·?ÝÊ)ùš„Ãè\[f®¯_˺´0ó²ì "SŽw§pÆ÷´IäQ¬}Y‚× [·t{–ÄOõô*ý+EÈ™8˾»Õq߯& ÌTþçbN[|T[cmŠÞˆÏÐôµ»ÓæYöH‚7ŠJS‡yïLúפ·‡Â>eiw#7ƒÅ&Ám~”.^hÁµjqýì³=ôסJN¯¹J½Tº^ 4Tñ£+½ÚXŸøøé^Êqã狎W 2±†žBÆtMAq\ÑO"®*u2&©à Ò9>%¡Ý¡àGÙ6’œŠÍØ|5°ÜwÅ^šþŽ)ÛCÃ=‡ 0†#êA(…'ÅQðÖÅü£° Ž¿ºŒp‚bw˜N8`¤†A_ÔFÀÅ×~S,À˜‚˜äFæ4¤ò`‘+_Ùy0é<£IEÑi,M/÷2Pb1Ó~à~›ÂEæçIÂ?óùAšâÕ^äc0xGrcM?åÐ}’¤ 'wHæß\´wk‡@ƒ¯SŒüäC—ÌSl7ävË]F!ÅtÅ‚„ñt$éîÎÇw>û/mPC endstream endobj 560 0 obj << /Type /ObjStm /N 100 /First 872 /Length 2120 /Filter /FlateDecode >> stream xÚÅYÛnÛH}×WôcæÁTß/‹Á,2v<0 í 3› ŒDÛÄH¤—¢b{¾~O‘MY2%Y²ãö­ºêtUuuWÓ˜À83V0©Qhf- Ãuz&Œ@˜u’IåQ*&-µ-S tLi?0ž3ÍÑïÓZ¢D;” 1Ôï™ e`–ÚA2§ÐŠ9â,óôÁ1ïåÀr΂-,8ÌoÁ1ÁrƒŠã¨¡hHQ³’ iU#H(A-*$R8&€k`%§uf¥@…ø`íÂj,^‚³ãÔÎ΀gGÈ8{êˆç`¨ÎkÎ$GËjè“;0S)ÀÃjƒ ”kµgRrªT ÇèUB©Ü¥"ÎÆ¢b%*váv`-8kZYÊp ÁÒ€‡…µ¤!ýÀ^’ XLZhß’Åœ1™Ìiª€³ƒ¢-Œ&½fXMzLµ0›ôC˜)ƒÂtNò ¬¸*èÏÂvŠ pµ°¢4QR $”ªTCB22t®È#HÃÊA=ç°P'9¦ÃkàKœÉ$“¡±7XÀd+±^Ð&ÑèQ¤OM²hùCÒÚ„*b…:Øtzðæç³áY‘×y:ÉÿJë¼,~üü3^°áoåeɆ'ìÍçY^\a¤ÿ4I‹ÙOì—_†zâèIZ§M™Ïê*ÿ>'qÔ>#ªG¦;So’C(¨<®²vIËl{ƒ›¸|ÉgãrÚ–Óé¼ÈG}f›hz-öÏ]í@×ãýv2)ÉÛãr'ùuÑ*{šMËêaƒ¤½gõåŽFÙ¬Ór}Ó¨àn± ߶ۦóåUÙ/™Ù“ÿ¸åOóÖ“ÞÝß‚xx6¥b£ôÌÛ"û!±]ȾÒ÷˜¹EþoY‘Uùè%ö™ÚCp’]åEç=Årøš–ãù$î–g©ö‰ ï³ëtô°OíÏèïüvàh›\bóY 'ÛI6Æ’cDš:-êÙ†¸ÝFÞ#;ÐmÒD;-":¯®e`‹G®œ-ÿ|êtOFûwœÛë*G{.¼0.§ ëíbd¼÷ìLÞ“tFjŸLÒN=ñl:žÏêrÚÝÁ!;Q>Ë¿ý~.òûõ\—ÆwäU”ÅÑS}ö0«³élëuäýý÷0jƒâq9'ûݽ™M8Yœ(å¼jDß5²G¸Ë´œž'[sÒüY”w“l|¼žu‡ÈjgoÖû|„ó>[6XyûPå×7uËaA_/%öÛº½ÿ³hÃ^ïß«´=O—æõz1o0¼ÄE‚ ?¥×Ù`Î5­ ɃE~>žg3( 'g“45]²qžþZÞ³¯œµù­ òÛ,p©AZº·EQ‚Õ×&Cû6ŠG‘Íø`økYáÆÙ2ãßHâ¨f_µ1I'-y")—Bö¤•H„° ¹˜¯ïóâOpbQŸÏÏÚÏ››º¾ýÇpxww—\]ÕwIY]7JˆNØWÊ Õð÷?þE ”ÕóÉäÛ>4§PX#üÔæ[ÚS‡t+V ²ØÐ5· ÝÕéA #BÊÚÉ`;üT•£‹ :€UNNÙð2»¯Ÿêœè™Ë‰—š‹raÂB©p[ÊXªX¶‹ e´mÛ½¥‹¥ãŸÐ–"Η‘¯ˆóE¤“QŽˆãÒìíF"˜D"3ÖA$I±â()ÿFi¥ú¸Ò ŠnAq•hm˜²"á#5O½Z8€t ‹}9¯Ë iúa’¥ç‰‡º;ɆóÄŠí’?â®ÜDËöž†;5.ËÛ|4;TIО)í¢ƒNè NŸ8ëÖ¢¹8ûp²È(/R²Î¦)å‡SR$ÎÓ«‘N\ >J¨j=¨óI-R½tÒd] ˆkÇ4­C„ØC ŒqL 2":Ò“V¢­}%@Ò¥ TŠ¶ ú.³Ç«X{5;ºHÄ#ùñUé qŸh„idTóü—(a^ ‘„Cï°´óíöÛ«»Fž¯¼¼<Öxì1ï™ô2ÔÊ ½ó•p¡·xO}Sei|A!Pú‹‰Z  oÙ†`ñè—Ÿ:ÕÜåõM|$;Øc¤õI0ôpm{è%:ñV½&(Áe¢,†}îqä*‡ÓG¢´6ÁѸÚÞùëËCcp‰±r®±äË­›<ȘZ&í/‹Ð:¹Qö•Iíhªgpú{…màvÎ_nGœ¹‚Ó¯ ›(ú¥&›_>‰\¾ŸKk²"q×t½÷ÞBãšÿ÷Âh.@ô#+biÔ±ǺÜðˆ€{˜± ¼ðh# q·¯ld,ôGkÿŒe93iþõ34èOS8(MQª—¦(ñÒ4EÆ4CÆ4CuiI—nÄ4CÅôDñ-éî–²¸1ÓURT¦ÿ8×LüG¹¶.QXœÁÎu{å2ÇNëW "ôKÞéDúO pôËÜ#ázE8²ÁâZŠàª”Þ­OlŽKzUl¦>ÙÑÉêmùäôr±·þS(ä, endstream endobj 744 0 obj << /Length 3479 /Filter /FlateDecode >> stream xÚµZmoܸþž_áö“Üzy¢HQÒú!——CZ$.‚"ò.wWwZɧ—8ɯï g¨7kEaÀ+Q9>3óÌPò"„?y‘…‰R"ÓéÅöô$t­Íá‚.ÞüòDr¿ tÜLzþ|ý䇗q|!C‘…™¼¸ÞO‡ºÞ]¼ôå‡ë^¨$&40ˆk|ùòúÝåF)(ØåÉ‹ëaš8Š)öü¶@&ÕB*Ms_-N»¼Ëéªhñ×yE÷yÓä—Q|¡ÛzO¿k»½ÜÀE™a*3JªLÈDûUî÷ÝÝÇm}º-ígî>“̤"‰\á$ipw,¶8Ïq.ÚÍe”<&ØÙ}Þ—˺*H˜‰,‹ýлº¿)íûèÚq,"™ù®(®[`ÝÚÝ°|ÜÂJ”R_l$üÆÔ½#eª ±y‰W:ø-ŒÃ5™´Q¢ü\•HÇ"•‰ï ãIš#¯v4EqÊE•7_¨ýÜŒ°ëqbæ3Êó3¦÷g¼Í›®¥9  ýømuÏ*Ü°þ„ÿo EQ(tŠ:K…–rŠ@i‚Ê~Æ]”qÐvö–Ú܆ÃoWÓï4ÛYêå÷:›ƒx[ŠW–‚L Á&³Ø&s;;€ùjXMl‚ßí¶c¡Ž9 ¼­+Dd—UË•¥ïÄËcƒgþ=C–ïT#c-R3œX5•µ»–4ìÊEM÷ÃlÀŒ#èùõ±à7ö}µíŠºârŠãÝÞCõü¸–(é|›œù:u‚›‰ÇÛ»}÷QîuEÕÑÃêjÒiÀ¶ü­xàYÝwWN¨ £æ¶5 Þ?F_á Z¦±Ì-¢õ§µ‡Å(¡Õ*ø(—ºiq[ôx‡þdŽI¢¥z²T(9Xï*ÂbðfÉ`)»¯Äc&j‹¯<¿3(hé¼@]“Wí¾nNtû…©{/ßÐéKQøº¦‡ŒÀCœ(¿@žje)*ŠD¤^ËFGƹ«:·ˆ-·rºÉ+ŽxãgÑŸ.}ЩÖlšÑàUoúŽÞ@1[º$3qc6<ømS»wý¶ã^¨8÷âÉ™¿Ï·]Ý´‹w…^dá|5§Úí>…,˜íåFc œ¸Ö¼ìŽu8²5ʼnm©uo|µl~mW8—}kÉî¼/2jfkÊ“ Xy½¶©0‘žxäpm ð|úáØšÁë•õamãåƒÃëE|0`Yê¦èŽ'lò.? @Åê´‘0ºDLÈ=¿6!‘Š;‡uê15ÅÖ7YêÌèòhjøyWóX~–¢ºíy|4ñ| 7Ê‘àðRÓ:Ó£GÌÀ3¬§õõ¾LÎmçì‡[qHû³ÏË+º.ª]±Í;2ç Væá- Áe4Fñ øà­]âz¨êæòo"e«¡\iÝ7Ý‘#Ù·œ¡™’‘Y K^ ¹\¦‹»õÁuÔ ÜX‘mÖæJ$˜J:%Î_¾~óîé›çkSÍ–fÅlT ÌfÐèF® $¥Ðq¼`DH V…r‘j=òç§ÏþuVÊD¤™ü.)ÿ~^Êér%kQïv]‡J˜DÏ=#Ï9®$ƒÓ®h¬gоˆPéÏ5pK ÁtÜ4üeÜÐÝ)§t­¨,5PtŸô§ÆL´uÈ|qúhnù÷è1¤"±4ßgzŒräv!Ê’ùÂIDéå”C:¯Úx7¨%PÝšÑC½ªº£ {¼ wè!”ÌééŽ.~ï[~ó¦/Ê—ó``1u•ß”, åy$^ίí ó¬ÜÀk_PFà¼ýͦ¾Dä%¢ Äzå Gí±n€=l¢ ˜ÎžÚÆÍÄ;éÐä'ˆÐපôü4¬ïpÓ”.ìábìWóèd÷(êh¦»c³YÚ(WS‡A²’—Å×|Lžã7˜Ý»mpÝ}¨érÏ‘\O$ýkæ “ê4~¬KtE±ÁT¢9k w…Ï*úvYW{€9 ™Î¼l”FŽCÎ"Ü.ŽÞR'‡¼Ú6—`Ùö­™ê.~©%JFàá ƒ…Ýr¿ôÔ)òÒALå%æFDn§C5_c‹ÀOGÆ壽#Ýæg”«Mòè(E:ê¶ÆÍãpYŒuxŠ½O_N±*ZߣRßæe?¬ Oùe}F%0L¦Gá“05£Bç~¯ÑqXòŠ3FµJÛ½dÙã=@qL'nIe1Û\tu?ÔÐ23±q|8Ê윛¤´ŸŒå x œE_îÆ’abÒ™ÓÉKB)gs –/ rU«õ»/ë@rc5m‘9gÒcªç¸—€+ÇôWøàëjëÌôB…Ù` apÌ[jâÑ-·“ˆ;ÇÉõdÅøÈ¿Û{§Æ£: í~Hç‰8Àú¥vÒ~ÇaáyYüq鳪0@ó$™}ÄãAëêA ݲ7Þ¨Mªž,Ád"‘¤é˜–þ?Í œV§gr…NÔFA¤ÎC6=,&ßv}^.8?cèS‘ŸsÙF%³s "“m7ÞãäÞ’ÁÆÂt°±×=I ÈÒrŽ°Ou±›J'óOh»E–n½$ëõÏû%]J÷M<yˆü¶/]k%.‡ƒÖ¶«—»AcQm˜+WY¡ÖÎ4ž­¦K™ò[û+!ÓhZ'0IpW¸rÖbŽT3Â믶©7ûÆþÙÛjû…$@=Æ)ºqŒhî$6"ZÒ°ûÙa˜ŽøG—TÄ# @ú36J Hd¦ôZ™àÕ~Õãnd³q¸üË?øe]A'ã_M#j8ÏÊq)íZ)§¡æÙÔûó•$ŠP  bJa:NÉõ5¼|pÿ3#â(û†“Ð`݃dŽnÁ°^¸€#-˜Fƒ—„Ù;¡¾µ^{Žt5ãFz}Áô`S1·Ó÷˜@ç8<`>¾lðjOÇ<_o8ý£ó¢ùá– Ž»'ÇãÌ“Q{Б[¸²D½ªuÄ(‘M6p²l. Ïª€0û+ ˨¿(œ/ÃÕô8θlÏâÝÑ…ˆ†ô¶æùd$Zz-O–ÖHÊL4C¤NS.àUÞw50`¦@ДoÖØÒcJºR„.nžÏ— aßÔ'ê?ô¹m짢î)ýÙðÔs'Ò¢žé’°Îáä£- ©Â³ÀÏÓ²³M•û³±Úïy籿I¥}ÓÿoõXIÛç[î1nŠñ%fhÍoo]nèx~8#$ïÜwÆ0‚wó•ÔÎãjV7ËW 'Y';9ó²¾øÝØ}_²e¾µvù…C„e] î2ƒÌ()~å bx&ð B@®‚Ÿ;Ì[¦ß=ø÷7“V¾}XNã–ùì˜ßv.žÈš~?~‘ñÆîQ[ûáÊ‚xY‘\®C»¯¢Ù:έá{Cêå cµ’{©Ü»£å´lb´|l‚­»ºâ+ôÕ&hòPž{-xÉ‚%q]ˆM^2;ÅôÏ»“(@‹CÚµÿ³DDzþAÈrÕ¦þò‘¹ËÃlJÉé'"ëlJ‡äA2µœó¤Qèå£LŽúÂÖVÂÚrMÿŽþoÏyG© NÒ™¶ðŒ³Þ¢Ìçt”$ÂŒ*š‰ä¾Èp)ÞÝÛPl,ÜÃð™ƒcõÂÉüË>$¤Ù˜f::Œ³æx.ñ‹EMp(®œIóU:= ùÖü*«ÞlÃGë<¹a&]nŠÝj¦œáOÈw¶´=Küô¬ ŽÆ@ð]´q[e¾æÛÒÝpè7 øaPz_U²J,dAÈ_8ôx0ñÜ}‚ã_ XO‰SžÜ%InvúMGiv¾¡ÄúM:sO,?ðv?yù†j² #Œ‰ŠæKMˆªŠow *åO…‹pÐÊ4¬ÍýWH&hü™±OU 1çØó2<˜+𛨒¿d›½8hÁGÑç\Æ/a e»M>¼z\!=ƒ—Õi%ƒ@ø•š;((|Q1æO¨†Ó¦â"^V”`ï´íý¸‰.Â(s‘d" Õw‡Mÿúf|%þ,&qR½µ“O‹€±sØ|G6ÑICåèðÌçå´Ë CEË%ªLºÊÂd‰ç#ªïüð‚CŽÆŠóþî•" endstream endobj 751 0 obj << /Length 3059 /Filter /FlateDecode >> stream xÚÍZYÜÆ~ׯ8áÚËûàå6²(°lGšÀdAà9;Œ9䆇WûïSG7/qµk @ ;}TWWW×ñUSrÀ?¹Kƒ]¬µHM²;œŸ4ÚÞì¸ñæoϤ¥óПQþuÿìO/Ãp'‘©ÜísVû|÷Î{~Ênû¢½ðµÖžúúÂ7&ôö‰ñ†¾iˬ‚™À„‘^¼ßÿýÙ‹ý¸Y¨Ô¥BÊÇÅŠ#¤6,Ö«#J¤¼û •xÍÀâ2ìýŠ žÎxâ9÷Íù¶¬Šö’»Ýp¸P±w²¤öæûÊH ““¶½9,ÕB: ™(GdÙ÷§¬çí»áöö‡­š¶ïx¨?]H¤KS+PŸÕyÖæ–t˜ÀrC>D"”‘Ûúe}¨†¼@Ò¯“ˆäò¥2dŠoP UñQœþbù%s~¬µcw}‚”Ǧ¥ß­ýµFêùæxìïôÄz¡¤PŠÐ„3%©4=õ&ß@Df<²ý`ßbK‘$Ž¸«$EÖܨ³¾tFý¼®«Â¿m‹CÙ•MÍ£3ö;_°h†5'™q<îñ&‰•‘_‘‘ñd‡8zÈj=guy =TY_ðPÙ3Í] &@­¦ÍË:kïy>kaâ\ôåA\øQª½àí]Ù 1£cïåËýOLš?ÒÔEÇ|Êž,Ô°ºAÉî@ ëñÐQËó Í%Ê»;•>1 ê×]—½kA“ :žã£SÁdlɱúÌ>pª0RÞ[6kUó°a"¸×(ÝEƈŒC‡aN)ü‘ŠbÈrdLÜzÆ`# ¬·!]½-=Š ÑFHëH±å9s¤wÏÇ3FÆ«ñŒÃÿ^óAÛî=ê4–Þmvƒú‰ ~Êõi•Š„–8R¨Õ#ÒQû3ò³­™’¼¨xz?a°§û©Ÿo¹žÑ"6cøê+&uqcÓµQ=zŒ}”dð¸ðPe]ÇÇ?go﹓U]Ã×-ùÐelaÐþµÌœ;ÂéÒ®mhï‰%‡×âu½€ÊÑR,â´4xƒ¯ ÜT‡±À¤^WÖ7UÁƒæè·jꟃÍ#…q ¦ÅŽ š#Ï o^r³-@Ëd+È)%t.¢Ü–zᵉHƒ~ÿ‘÷¹&çzàS½`~Üäîë4‘ŽWa£ÝL‡Zh•ÎyVÛÇà/ãÞ6‘^UÖ¿p‹‚ÇÆQ,L5·0àwŧi”*£,MVâßžfízÆ`+Í®¶YÕ ŸqXÕò›ßÁ; v’Ð{Cháè`C[Ô‡l¨´M°&‰6,D”]¨S€5<–^‘ÖŸo&×9C3Xe(ï_XL„ç:ÔØA,AÌ*(ÐFÿlmt!ÍΩŸrÎÕˆúýü“BÕpóg…SÑY¸?³<è§òCâwkøˆÔ(WCdSÙW›ñ Ÿ™^h*ÿ‘.CN-çEïtZ–_«„â Ù¯ä#ÌQöˆ‘2N¡xÔÿ¯#=Ì Òi 9/M¤Vñœê=m3œ Šœ)ô^òx[tö­óàÞ@±Š…+DÌ«ªbB ŠÐØø¾0×*’؇­–­Gjþ=4œªË›¡:fÄ°*&X…ý²^¼Z 'ÓAæŽüÜw_ìÉÓEÞJæY–“ÑÎâzÿ­x™ëŒu4=ÌäôÁÅãQ¤Ü—£$Çbª²®çÖhyÜ=QÕ£”#•\æv½ÿëE ©½§Ì‚%˜ Mƒ™´¬í¹¥€ÉWÏù ‚õ„FÝ%A{âÏ$ã9‚J=¹¹u%†cÇ¡Å{KXR´<™2hì·?o»õþŒÁV³ÚfY¡¡h l¼Ûª+`üjT uí{Ë9ëß[@ª¤e®UûØ ¥œüä‹ Äwi€fJèŸ/ص?#ß8éšéT°²,‡y¼¶à*Œ¹<×a➢q ãép‌üÆOå ]0¶WXe2¢PFÞ+ËŸ|Àò´M¨¤)s·w}Èð7r—µ±)Døê£RëB* €ødÖ–pæ `œ£ãiz£NG/ÇOØæ¨ßæLò³Ôf<Øô¥Îˆ–œŸÏ'fÐàW™óÖ먆’i Ÿô-1«Ù €¨¥)QWÈ;/ê¦?iÆ"RrÓQÑ_a–W!¢°ÁZc±0SŠÎuì-9YF’ÜìÀ³bð›P(´IvZ¤?ù`»¶€óØŸÑMϘ°ßì32£ÆQ‚ι"vì—eŒT¾²Xë *”U)DñT§‰×fô-…áY†Ä º”ßðÓìÈŒº 5-4‚ùQŠoøñR'ø¦¹ÂLëoа“ó=,oº§4¥Š¬“‹‚ w°2@”¹vzц Þ~_wÏLl0wd¤Ÿ0Ò§¹ãPS<ü@ûÞEá,¼ŸîúÔý²¶PWÇô¬ý{€º¥ÿ36”qDÿãÝû`—Ã8l"àŽvwDuÆïð[íÞ>ûÇå¹G$ I¥w)f@sP»D¤X%ñ"ˆct‡$HG4W_°áq4B•~ÁôI:Et>75½„sDÊí|Íó½ãtõãw—<ôïÿ‡7XF¾¹úþ’ç­õøbw&xƒß•âˆO›å%š â;à‡ËþÔ 7'ž¦GËÝàÊÈáɦç2™<Ã(+5‘¤Æ}>7üßT oƒq• w5ÿ¸2ý­ endstream endobj 758 0 obj << /Length 2993 /Filter /FlateDecode >> stream xÚYYÛÈ~÷¯ìK8‰E³Ú /ÉÆÙÍC8m 8bKb–"’²=ùõ©«yhzfŒ@lv—ºª««¾:¨nø©›mr“omq³;½Ih¶?ÜðàÃ_Þ(¡ÛáfAùÇ»7ïÞ§éJâm²U7wûåVwÕÍÇ(»ý|÷דë8K2Ø„&ß¿¿ûçíÆ™~HòæÏw›Tëï”)_(+l¬ŒeÞ?»Þ½EŽ@˜- õ6Ö‰ñ"¶B²Þ+Îrå)êŽé¨Ä‡ŠÎ·]D]ÝÞÂct=ÏŽµü^ö}y«óè1$RYœZ=‰ð±/Û_?‡±6Nà@BX¹¶ëö ,ÚÐÞYg*·N‚ûê,.ÒIx”°¤ilÓí¼›úþÝ6Vë(eYÃz³…y]ªb­Óµ^6ažS$ [z³Q&†Ó?Føðïúr²6º;â\nUw¨¢·l‹@À÷­<'ÓãÑáÀF»²i‚†SÄ°ÛÂÝeÌo¿¿ÞŸ›²½¯öã½®>%iÒ&°_ªÓ¨U2¨[t—QFC}ð“û¦< ðGõûà1Ÿ2%{Déÿs©¿Ü¦iT6ŽQÎÑñ1äÉ»°RLoõœÚ¨×÷‚ R6|pÀ«À–6U^¼x]é³XçfmFˆÐtæÛn;~6Ý0ð¨nùé>%*ÛÕ®Ý=":SäÖ¹k¡;—#dYB¼ë8Òqöuq„œ0blÌ8ãÿ‰ðÂîo„/ .÷,:8ô4t´4¸•c¶šÐGŠ|&>A¢y.û±f –U\ÿâ'Y4€^áDÄ ˆ pá9%ûrZÁ>šGO»ë]9°ü°zp-ŠãqŠh0,"å®ë!I¢ÚV°m È€-9g¥$!íîˆàЇ, y!ç ‚ á\Ð5ªÌƒ~ºaä)9O‹0›ŸÞßñªÜ¯ú(’at¨Ïׂ—é^‰3†¿^Æ®‡[z+‘‘¥[é‹£«¦°w<,§Aãñ˜LXËIp¥âô«å`‡3hH2>\ú Ö·N›F_õ§ž‹¤hüڵƒ’xN^B[ÑujºA|@==µ^Z»4c Š}‡ÆUqU‹Ìß/aÇ«%O¢é5Ž§¤ƒY¼y<‚Jç]Ö»< nä)²{˜+ùµ)û‡{5 · YpUC::P`‡‚ˆËõ”hÄr“{¤u”ŸG˜D¡q|êzM–†/ÞÒBNhühB€D´WM‘Î4H[£çl`Ú9 vÝ¢:JÌ(¤¨œ»®zU¶¥ k ç)¾q Kq¨+MðzfY û¿oæÿªá+&$ÝŸŽåyr~Ë ßǹDÿàö`ý`ÌŸEösyŠH«ëC˜êy¨L‡x¾z÷Ä/‹|µ%‰,€¯ïÍ4×D™ìùFªE÷:6`ÿÝm~‚K\š‘ÉÓÐeà™&¿´h,`e+HŒoç¾ÜAÖ‹{éLó|î L - Ïë¢`ä@êº=#äà°š¹®ÓÄf;åRuû±ÖÁFAj7§½p8_z׈l½c±²(`ò^¶ZŽµÀ%|Ý•}Nj¦Tlj5ëô™àŠmFn‹eÂHÇÃÉU—²É1€þìúS !´ý{W]Ú õ1\…f9øÅœ^Æç”RÄJ­+5ë b‹°Ùþûr€@Èóáû­9œV} ÝÆŠev”ÁëÀêq”¨ULÊé N Xr›ËªX¯éì²&Á"ŽOX¶lIíoWg0½!¡äNŽâ%¼3?Z€.só-ät Ù!ر™Åe °Å¿•§ pÖ‡3ÝI¨HYÌ^–@K€>Ñ „ú „£)ÙÅ*”£/ÌñjÙ4z_÷”ur9X£6T³Ì¹0–h•N\,œ¹0sžœ#ñžlüÅÊ Á…Òtr× 5`@%)Øw;z_bšÕù†ú¿œ§" ÆÇómª#˜¥œ jÊ©†ú‘>|¬|>zõì™P6éW:ÂÅÜ3î%fØP‹HÒPTÔ )eO¹%™”ù¬(*Ka§ì¹}§¬¹ ÄH5‰<_þ32‰²m[»Bo¬ÒH<›P¤µÁ☂K-½³Mš&’]ã_Ê[28ê±Ù;™šúpœö(¤\ª*ŒÉíáž$ÄÊ%.â@n ™Î ­ì¹à}špm !%—Ì’+³ø‡¨¯ðB²\n!ãl×CP?\ÿ—_Q/³V§Bwšà·c9ÈßýædÙÂ*²öÝÈÓa¹öë/®ÿÚ×Üâ×ä 〃tJ ʄуÊÇ`¼Ñ&ÞšEÏCçjênh(V*·/!ëÆ*[Ñß\%/ámàSp‡PƒrÁ-Ý’ZjZ\ŽÝ¥©x,¥¦ÿ[)®¯»‹ØžH¶´=ˆ;­šò‹ö—¸¬q}†!Â?¦¤dÎýX>m›h. 1hà­S4bÜZª#Ç\ÈúƒÝÏ;2P«M|£Àiý'!¨ïÈ|ˆ¬dC©.ê-|'hÐp9LDžÉPžü²gt$Ýâô„o¯˜Ù!fÅ¢_‰Š¥&†ÔW8N¦±è‘V¾Ï4ð.‹b†õî·øL–åú»ºVI¶î«-÷ëMœ˜‰ýôÜcåR>pç$…¸ûLÿ®Hã<)VŒ¤š@A‘]e4'TÔêÕþdbÓ/Ú北)#Ù.Ž†_,Ë;jÛp´~ö³ŸÕO{“Á¶¨¢ÌhÆäà¾92ô>ω¼ 9üœ+Ê ¶ÖÞ̸V€d=/ò•Ç³÷yOÕ ?×µÛ8KÔê<(âettãp4áW•ˆ˜IˆüvÃ&½¶`&ç—EتxîË’~áéï…ýËbÛ$]Ú) ªj€pûÝV¹ï]ÐÙ´ÆáÒvãSÍØeLu-w³)óJ“yHÌ|§Íø´ÐLÕ*ñ Ô™¤ ‘L¶bPM#n+Jöôû$Íì/I¢—íïI@/u)ýˆµE —zõ…Ñ}e¼_ô€õ%JLÞµüŒw+_‘©|0¾)kSwê;¿•ÎîÈ_"W_z½»Wôåxþ¤*Ÿ,E#4þíôus¼F<ñü*‡ Ù^uâ/-~ÅZæéwPìéü¤ÖÃ8 endstream endobj 762 0 obj << /Length 3052 /Filter /FlateDecode >> stream xÚÕYY“Û6~÷¯˜òËj²#„Á+~rlÏ&ÙÊQ“Éæ!IÍÒ$±L‘ O&¿~û…²½/[µ¥á"ÐÝh|ý5 ¯øé«,¸JÂPe6½ÚŸÔÚ쯸p÷gZÆ­aàz2òËûgŸßFÑ•Tdúê~7ê~{õËêÕ!?u®¹^‡a¸2_\¯­V÷ש]õ]Ýy =âUrýÛý7ÏÞÜ‹EÆ|¢T8ò¯bÅS±´NUz§Véвd»]÷øp*óê&ˆÆêÃv×=lLó ·¿QPT¨n&7õñTº?¸å³Â÷mëþm餵î»TêJ'‰ŠlxµÖ¡Ò¯ÞWm±¯ÜVæ,ó} «éøÁ¹M×&ˆUH¤*Š2‘¿F«štÕGæ4Ùª¨N}'½5ÿ‹Äë¯\s,ºÕÆfR†f+TWÖϦë0mhÑÜ´pc6KRêH™0õƒP^!¯¶ƒ×k½:]ë ‚5ZXžEÉI-´ãÚXÐ?ÉÄšWòØ0¹¨Õ°ïÆË›{A0:S'íš¼jÁÚÇV]¯pÕŸª²xw ’9pÂàXy}{ÏtªÊ5à"6qp#Ë–ÿ«ZÄ™¹¬É2›Áêè.KR›DÅIâGå;?º …ë-LC•¤éÜh_Wmçò-Ê”$´£p03”¹åRNBZؾòñÚ$«§|j—ÕQ "ÜÞÞÿüpûýÝÏ/ï^/ œ„SµÈ;pAÜZð/2d,CúAÂ@é ÉðåËWÿ¼(„U6²~¸XÊšTeÉ옢CÝ^§á › B؆j_ºÏ˺گýiÇöSã6E[ÔÕ¢t±™êŒ²[’ ÎYÚ©i–üBƒÑt®rÑ/R‡±v³4xÎp&D—%¡R•ŒãÚCÝ—[ÖùíõZT~¾qCnÕ¢- œèA°]YçÝ"èÁ4*ŒíÜo/ÛÄèAFÜ8—µŠ©8š‡†7®=±N›®xO'¼|ÂÝW„ä$©Q¶rÅ€©îfù «7o¦Ò4ÂD ¿B´*ë=äÑ2/ŸcG nø§ ©wgßÀù¹áAUÝqSå6®mó¦mæƒO>SņB´œÏƒó¦á#H&È•D$‚2™tžoÖ)oºbÓ—9bŸI$R™8’H…¥ bÏ,—E€ôÑ'x§' ç óŽ²Rõ/ÂFX4šqƒàÞêJçѵ%²Õã¡(E«A½!|HC & ì Z±ä°p Mü!Ee2•EÙ Ä.(o2GÕªÏÍßõ¢…ÊFPšZ¼"³¤ß#5¬²·@a[¼g@¤Z!£›º¯¶ˆ4¤Æ‰+Ôü)†8s‹Ô˜ ôçþãjMà"+øx¼xäÂ8Q6Ô*ªepå#¸eâ*€n,­Via˜B|¶È[E¨ ¡±¡ö|!XµùÑ›-ŒÊQ„žµÁ¶9b¹o;®Ø ¾|(ö\vUÝïÓƒÕ¡.·ÓÏêîð#¶jkZŒ·ƒ PdÍ #ÄÒêé± m°"jÀ«ù¦ëó’@Æh$Ãèâ6'ÃÙ5Ÿ¡'‘z´=7rs Eg‡òÊû¡E#ƒQ¤|æ´V{ó=a[MÕʾ ¿Ú6m·ì F… ð®X™dNøÉ%Îöo¤Z|Ê·[  ©NH§é»ºsžÚçž èAÉÿ¿GÛ»æ‰kÛ„A{V“w RîYŽ°©«;I?!<üc˜ùÓ5õZdy.Ì<"˜ûÂpú0dZ&é!ÀWë÷œíE@ò«=í ×'þŒU.çÕÞµ2ɈBÄãx­ ™¤Ž#†-o½K¢ £B볤GåÁâcÆÅmÞåê!4«ûƒø-˜ƒ ÕÈß-‚µ ÌàØÝgŸx!7 Bø?àÁ˜?˜lzÔHÌVfù4,Ÿ¦_BA?kåì/ o²P_ÿWÒ 9Ö*>cÇ3LÆ”VH˜aFÔ ¤ B^ÑvH—’ðl¥âæ—eÉMòQ[x*¶xãL…™þHÏb=Cf\§í1•EšÎm_àMDD¾¦¶m¿A°@IX+øäŽ%™h’B²hæ.µƒ©›–¹$ŸwpÁc˜MŽh¼)Äb7⻽ôæÕ¢³E”“ˆôß_`(ÆNˆA°4zf>b&#…Þ/K3eÍÕ²óDöüµ+÷uStB¬$0ˆÆ‘M–[q _p„˜[GÇEDƒÇHÊfȬâ9IÛ¦‘þâä&ƒÀ$qo HÈÞ²Ì;ªæô |Ã/·®íšZOPt¾€þLص)ᄪ¨Sñ,¤ 6{X×;Ïfüz0)µƒÞ3Œ žÖ=G "¯Xƒ`+d¡~]t1‡ ~$«bǃ)ÀÙ!xB¦%ýoÉ7eïQó/ç Ù ž¥Ù?ܽùñÍÝ¿Þ<|ýÝ?Ý/^¬h£‚ñ‚ ¨XØ¡ÎÞ&˜1?Ì÷í§ä·0e¼z,¸ Ã×ÑÙìC×Wyç8‚<ª&ÏÂR¾i0†Úd;˜tU’%!Õ‡X‹n‘Cü¤Ü'ô»yZ £|Ïq[ ßÖ³<,^múÆhŸ‰ ñì„A< ½eaó™–]±ÞÂy¨[s =AƒƒÅ·'ØFlÊý!èÒÒÉ‚´¶&¶}=u°ÉÙ¹sùÖ!„™¼6?À_ItжÔÎk`‰ú$Ÿƒ ¡\z,É©w“—ÒÏŒ†KŽ»?/JÌ­XNòúÕsIt`ØvÛ1?þ¸è‚‰Òéà·ëó%uŠ ×;£¾(ÆwO¿÷EÛ=ÇoT¹ß{Wmé@tI Tâ$Š†±ã0jð|-qÆÈ:³ÜA´£|ÃCùËp=žÎîGCÛl6y—ÊL:&rŽ§ç\‘J„XB¿ º•Žª\†K‚æ¼åw#bbßÙí°¥ ÚëÄl Ž†¸#6L ºÝWqô ðÖ7™[¦ ò¤…WžMi!ò  ™Ä®/¹Ì:¡;³d ø k|ÀŽ9~â^ä…eDͱé©ptÏ• ·G?l'zJ×3àæƤ‘¶(-²*´±Á- 0•T¡‰åLY –±Y}‹ “à¯ñn¸ e¯Â?‰\Àÿú‡-ø –Á¨üv™OziåÒed˜¡0Œœ-©?Q£ðS4:k1ÿ_:΄ƒøN¤Ú·ìÎtû¬Ýâ@…±þß醑ä?ƒ° endstream endobj 764 0 obj << /Type /XObject /Subtype /Form /FormType 1 /PTEX.FileName (./rfftwnd.pdf) /PTEX.PageNumber 1 /PTEX.InfoDict 765 0 R /BBox [0 0 270 405] /Resources << /ProcSet [ /PDF /Text ] /ExtGState << /R7 766 0 R >>/Font << /R8 767 0 R/R10 768 0 R>> >> /Length 4558 /Filter /FlateDecode >> stream xœµ\Ë®ä¶Ý÷Wô&€çÊâK¢Ù²Žï, Ží1ÌËÎ bÿ}Šd=¥Û’ºÇÏeñÏ!YâC"ûçë8¸ëXþã¿{{ùêëùúã.îZþûåÇKÌóý5cöc¸z——!ç«Çiþú˗׺„Á¥ë/ãç4.™Ê‹yY–«¾þû‰‚ž©€%̉ÜÊn”4ÈH}uÓ<¤™þMó°¸Ì*1Ï—)¥!ÄK_Î1èù’½VžCÌóeYÆ!¥e_Ï1èùBFj¬É0k1û* ‡Üå¿T£cÐóÅ—Mq¿JÇ *ÈÍqp}~ z–¸˜–\›±ÃxßÏÈ¢§‰l-éJãg_Ó1È"h_Ó ”ÅÖ‚Eûõ:²8Ú¯Ø ”EÒ¾ª(%7Mûqt`1¤€›ñs¡±s[Ã>Àb涆„ÅJÜŠ“r-Fnë?@XlÜ®ÀÂbⶊÄózFÛŸ”nƒ`VBÐíié6Êæ¥]MÇ ˜™v5@Áܤ ›“Ó˘vëuóÓnÅN `†ÚUuE+Ÿºþ¢…ÎÏ×þæ¾{{ýë+Z•åë2Ät}õúÒ–jôÿaçìÂ5Žaˆ“¿¾z{ùâúÇWÿ¾ä@ZGw} ~fªÅ«ï/_ ÃPòþöêò‹s)·¥I¦è®oÉ2/hySÝ»8äˆ(ZŠúnî@‹CÅ?­øB7%±ª2»N•XP XUªªTUuV VÕ¡XÕ<º!UM¬2Sè÷º£µ&[|ÜÔ.Ѻ—º PiÊÕò†; ã¤q‰A©ÄLl2"1Â#+ åƒòˆxØdÚË唦2'Ó>…¥¾îö!úã—Ž‡ gżs:FYQ;¯*öVÄÁNõEE%7·Ð!È :h¡c”µÓBû¨Ö~ £d/ÿõßK{Ö¹n,i,k3$JÑ°R‡±4»úVgq´b\–šhB¢GM,oj‘šS='²Ç!¶y>Ñ@/´ìX¦2,ªE}‹è‡ŸÁoª(óËK¨%Þãhm1§YVÆ‘ÆI-UVLô“i¢ç¶6êäË’œÖàc(‰A}Š¤»€h©s¢%FòGD§¬Þžš„Û‹ë-CY(­$O1èÔÁœLÏ=^@5UŒ9ÑTß:k—ê¼—Õ6cƒ­îbA9•6ƒþsãBÓš³¢ÍÉÝãeT™ c6§)¶Rö©îð:õÒžŽùä[¿’ûT`~¯Ë@{‘¼ÔÜo¾øðí÷ßÿôîǶ&øêkWˆh»³\ÒgÊqóÊ*® ~z÷áÓÇ?_z÷ôáÍ·ßý°Z>Ëz¾ÎÖc¦¥ÎïœÕƒâ–.–ÑÖ’³öÈqðkºôz’œµØøÄÎ#´õçlúø™3”ñò µsr‡,*aúœ© <4dUa;ÉÊ0ö¨t_ú¯¸»²¸/ë¤ò¾ÈÑ–¥ÄøƒØ¾tÄL僓—¯›ºÏ -bgò"PþœÉ I‹„]•ÚëÊ”ÉKYæ¢é‘¹ë 5¹ÛUæƒs×ãÍ)cí¾NkmNk»è|h¬ídÞk•©­)êݽ®­¹û°Kk>8$ikâ”ðXk¢Ì›­¹'óð°7ÁK£^i[êêdYÆåÿyâÉM¡¼Þ¿yâ‰êS>ÑÚ§D6à·DÃðÇÄS¿&dq–Æï‰áŠˆ5¥4jš¡S£Qƒ˜¦Æ M DBD `L 5#~½ôe‡¦ÊI ²Á-Óº^nr…A¾ ŒMML&iãR‹P©Á˜ä4ƒõgÀ¾àó †á ˆá € Ó:zøLƒaøPCÇÅz´7Dô†èQŒèŒè1L˜Öñ#z#z‹§Síx1@¿‹IºÓÐëbÑNƒõ„®„I Ƥ&f’´1©E˜Ô°ísZsÇ®ÏÙПQŒ¡Œœ¢1L˜ÀÐ¥QŒ§A.Ö“‰ûœ ňÀˆðÄ°ňäjûÿœs×[nkC6ň}N‘V}>Žc9£w«þ4&1“š˜IÓÊ$]¬Lj0&=¢ý T 䌊bä˜ bÚIƒ,ÎÒÝy…È‘ÀˆíQ½ j#jÓÔ¤©„¨Qˆ¨ŒªÑ ö¸Ú¤ËÕ`}.c¹uºZ¬/dg®tb651™¦•Ks¥Rƒ1ý'@\Û[gBùºÁïË Šönœó1»??áÊ1q¼y~Qí-¹— õØÂ/ÂÛ—ÉÂœÆYö U,ä±í-9çnÜô ´vD*ÉÂœþ…§-›ätÇ(t÷¢#º`D“Œèœ†],:¢‹æ~Çgãp“Œ#’¶q„ ¶qD QúåÞ¾ÈÒÊrÉðU^,ë¯òd÷Ãìá[7m#§ÁGøÖm¾ð‘ü>ÇgôK~‘qö'ïp„¯ÓjôØlédÚµS†RVˆåÔ–®~(ë.G$ôå”úù%—’ ïq´v ç"Ø l@Qd¢ÛØå2’`Ùêšîòëè–·™:y‰Çtçý ¦4rËI°¥UÖ9#6-…|¹T…«ªºË “+(ôó¹ußá=Ž{×”ïsÍ3ó­ÃQ»;çãNmÏÏ€åLÈÊË•yl[ç¿\¿ÿé_o~8ù%•ísO õg™‚î~wââÒ¾öH çßñ¦±ŽôaŒÃË/Ñ~ÏÖre=]1v­õúõÇÿþó»÷o?¼ùá×í‹'ñB§g+¨û3 Î…ðý§>}|üHµL2Sr £? ®; ±ö('êYNÈ( >» Ùk7?Ö}ãŇZîÚÉê3MµWK4­dó‰‚ip/ˆ´—B«9ð9•¡ÎeZ e|[uôÐùŸøeƒ0ÓÈïVüúVñÈn‘v ›·‹wPpþ{OÓ1Èn‘îk:²[¤ºuëøÂn‘î×ëÊn‘îWìÊn‘î«:ÒÉ!żóžü±„ ›7’wPK»šŽAK»šN  –tëFò ÄÒn½N  –v+v±´«êÊbitõ…Ìî°t‘˜ÛƒÒMÅÑžžC DÑžžcÄ`nG/ ‚öjt ‚øÙ«Ò1¢gOÑ1èy=$õ‘s aqƒˆì»2Ž ö鶎„ÅËŽŽ#ˆÅŠ"z{Ù';µ8‚XŒìTãbñ±£äòÂ;–0Óæ¡n¾êòX ú†Å—·§e>ž"_£ ©eßfŠ¢zUÐ,ìZ¤<â÷ŒnÓPo%(Ö{‡tçýþo?üh)¼ý º»ˆµ Œ´l ¯[²¥»pi(¹rÙ¡Ú¥Kµk—ˆ‘‹—’«—ˆU)ÎòúšU±¥Se(QÕ¡š*5UˆUUˆRUå^Þ÷LaýY5ÄEÛS,ô¬ëÃR­t—Ks±à¥O¡Ó´²éMQ!Sƒq©I¨ìz©2éJŒ™4­Lz_T˜Ô`Lj&»dªLrçS˜4­LzkT˜Ô`Lj&»jªLz©Ý"M®µc¤ÉÅvCÉÕvDÉåv@…ió”ÈwCÉ÷Ž‘uY¬‰.Œ5Ñe(Ñ…(Ѩ0mžÑe(ÑÕ1¶»ôÜ’«ÇÚb€È“¶^á·žáõÊ$ic’;ùÊ$`“2éE~e’{õ¤ieÒ›ù¤cR“0Ùu~câÛõÊ$ic’ûùÊ$`“2é¥þm\·¥„O3tÑ£ ÀHì†C19Š‘ÀA.ÖS¤â§n6tz#z#z ÃzÃz#z«\9‰Ÿºƒ[Æu¬{/MÈé¸:Þ|víSê«Ê>ŽI À$?7!L’6&±(€IN™Ô°¬[(“¦ãê …1‰™8àŒI À$?=!L’6&±(°õdÆÓØ’)bKfxÅȘ6¿¤Mï€Ù]!2¹FÔhd‰ˆ,Q£Qƒ˜¦Æ M DBD `T†–Z ¶ô'*$¸Ôׇ³ ¼Ä}!K¥Sƒ±éÏT0™¦•K-B%`’%2©aYS&MÇÕ10c0É’@™Ô`Lú“̤ieR‹0‰˜~Ÿ# yÈå+*'ÜŒ7á$³úã´ùðåCÂþqDõ?çàh0îŽHdð=ÄÒ,Q@´ZJnIêV›¢B¹–Z=T),€RKG†ïBè=Íkù $|ýí—T¹¸ÌuÄTCY8α«©Mí(”TÝ^,IªÑTMÙCízºV;£ãê‰ÛËEIõõËÿqz,} endstream endobj 773 0 obj << /Filter /FlateDecode /Subtype /Type1C /Length 2398 >> stream xœUU TSW~Ï÷^-BMú›D¨ ²/Ö­ D.€(T„° Rvp©m/¸©Öj¡µªµ(*ÈÖZK¥˜ÒjE‡íØÎ37ž™lOÛsÞ¹ï.ÿö}ÿýÿKS“(š¦'Gi´ÅšÂì´TûÊ »Ðxæ$ü²¨ˆäZYóÄ/S1ÇîMAŽ"äèP;sr‚îL…cÎPú%¢éÒê£á›óÊò³3³ •î«ãÖxxzzý¾ã¿hÑ"妲ßN”š‚ìÌ\åaR¬ÑnÎÓir —(Ãi­6;M™©-ËË*P¦¦§kÒíj‰©ZMŽri¶6;/os±Ò=ÜCàççï- ó—gë6(ãSs ”1Ê8Mf‘65ÿO›E¹ænÎ+,*I-ÝT––®ÉÈÌÖêf»{{ÌóôÙàë礤¨ÔJjµšJ¤ÖRaTN­£"(5µ”Š¤¢¨h*–ZN1”œzŽ:IM¦ž§)g*”zšJI()õ"ÅS4å&p)EP•”…žE¯§Lòž”6©MĈâDç¦9¬u;Š]ÅÛÅ„qaâ˜,FÏ\g³4+c}ٷٓÜsÜ»Ü%®N!'Ün(Æ}£I5 {‡dÒ1#6ðR“Í0 M캌wzšÒbääCXÉZ¢»æÈ¥cÁ(I—Æ°‚v¡ {™èÓh·ˆà#|š's\I‰q˜;ú3„² 'ÄSQ¥ât©ÈKÄyU°¯Oœ ¤ ùÒ<¦°1ãyfI‡=„+™ÔÔArùa¨g.¬n)èE¸Ü–@Ä|˜D éXŠËIåT,”Çð÷¿^L¦’ç—/öSHM«Áœ¿Q8Á©Â><¯Wò>¼M‚há €m>¼²¼ù½MÉ$žÖ]×ë«÷7ÈìŽÝÛ÷lA\fEÍ;àúpn‰äúì í¨ÁÅ/IcÃBÒÊ*ãÕAËRÏuË1»ÐæÎ]ÿI`Æ„ZN¶ÜàÂX'k2à=¯¼>Ø öfÜl^¯`/ñlf®ÍK)L|ð²Ä˜ÇØ÷›¯ØîϨîÃ~&I“Žñ6ãOZø·*w¡w—»åh½®±c‘­„‰)JÏ’äUè*×pÃÌÁoÏ6š×ß´9IQÄ¢¬â­Ëv’É[Ëvål_ž¯MF‘œ´Ô«gÅ/=íµWåÒæÄõ.ÇБê“92"y´ygy~¡ ¦ÍÞ´eK´æT{癆»G#‡ïk8Â9áÝÏà@3Ôñ@DBâˆâÄ^ „“Kô⇠L‡‘ñ83‘ÙŽ1ƒJ/éè‚6!½ÅVÿTÇJËû¿c¤ç—ŸkÏísù(ˆ! Tó&òðµy+ÓÐÈÂbàÇžå7v±ß“‹w“ÜÛÖ†§ I%³m^d†Àjj[mÎÂí¥Û²Ðô¿¢mÕeû¸PöØ[Çv@'Qã¡ÔxôÄM°Íê4M0ó쟲ˆp<æw]ÙñiIcfKLcâÈ,_â@ÂHð¨\aê@?¼X£XÀl[°15q¾ ýðÈ:Ì÷Œ6…Õ(þ ·ÞmC¯ß”IËñN¯´e!©c²ÚãÔ‚Ý™„&KˆÚBh˜¥–ÿµžl¾Î-½É’rœ*~¹DÀßÐõf'êA.õpy,Šxkcqn‘vcÙZ¤AÙ5…ÇKî|ÿÝSÜ«Ì!wÓJpFfôÝGkâ¤ç.íç8Z0LJcùÑö"#SW‡ú%L´˜.ó]S¾Ðå %PTB÷ ‹`žÉB›aˆ±ê¸3Â92X#ôt¯pè%dkÁS°k½ù‡Þ˜gÿdÒñ‰öxËÞo³?t^ëýæ3M”œ<ýmyítú2û¿Ì>XÝæ¡N-^¹N®íJ­‹Dœt\RÞ\ÃÝdþ;¥Ô)FƒÖëJèÆ°á¸ÛG¼a}÷Q©¼½Mªc&ÓýñÐ<j‡/›×Ç%¤¬OˆO¹ØÙu±¹S}ð6ÅS¸‡Á «N|Ä&³ëùØõî›M÷ÿñ»Þê ½æ®N»ž¾"ãOFº—Šð)ø…·•ªq)3ÛáÙß îX“ôtß-ìȘKd1$DºIô#"ã:™@p[¯!BØÁ#àÊÙÍ&és¸J„¿=o«RãªßŒAø-ÑÉ⬢ø#ˆè¹ ã˜âÖB^C$‘Àu$8ØMŸP ØOf-] ”ni@kf÷Ø7>WN«ïõqéÕYšýà þÕ3Ôípôx àóA†APñD5ªŸ\ñÆh–ôÛâÄ>,´ ð$Ÿ|è )r*¬Å=ÄêcjóäŸ7ïwtüñãŠú?ÕY  endstream endobj 774 0 obj << /Filter /FlateDecode /Subtype /Type1C /Length 1099 >> stream xœURmLSW>—Bï•ò14×)íe›ÀpÓœä£ àêæS©¥–Ž~ *RÜœ Y„-¶"à ’0cTÎ92™Ìpafü21À{ÉaËN™fÛŸ“sžó<ÏyÎû¾ òõA çêŒE:›A«‰H°s½P¸¸–×ùˆ¡LÌ⽓_(ÊhœÄàëZç[»ZC "J^@†)®p$Z¬öƒ>Ï&lØ÷NæÆÍ›ÃÿE¶ÄÄÄGìÏo„$]¡Ao”tS¤3Z¬&Ùö¦HÙF£A+èvk^¡ ÉÍÕåzeû5F]¾ 2 V«¥HظQx-*jK]¶½e09V(ìÕ˜ … Á›ÿB(Èl±ÚŽi´º£cD¸€íAo£}h?J@IH…RPÚ\(1(˜VùR¨=f”LóÈg‡ÏeŸ ’¬‘DH²$ųA"`§ 2³sùà0«VôÜÝü¢›ÇÙãC=ׇÛÞO—“IŠüÅ‚l×0a²¬j­|Eô}–ZØÀ~?¦“jAÀMñ+žÄ( C’‰j† ØÑO¾ažÄ**‰?Þ§ Sf&ÇÇ}ú´tLAˆÙ ° Â2æ65;å5Ë¢iˆðg=wJ¯ª=–aÌÁÊÇÑcÇÆô} }êÕ4œ‚³Í5þþ^¯ê•Y*UröðÌŒgdT“¶9˜›fü“î  ë‰@Ö«"GÈ/Ò7z4Û[ªÏ·ÈŸ°•Ÿ®(ÅÜѲ‹ô>ûžLÜá\ Dë#¯­ÝT<ŦÒ¤¨MßËá",°Dv7Øo¸»äéT9ð¼0­Tª†P‰ØÔÈ4Ÿù·r0À‚Oü-¢$ʸÝ$¤Lk¥µ·š?ÁÜM—]o/;ñáiEq)ƪS{KW öÌwñ&|°^×bmÏï-öàAüõ•ýØR]Z‚ƒp)‰¿]{0é=sK×õËÎÛ¹§Î]{®ªáüšÿ$¹êýÀç0Å—Âê×"k1G¶E„‘ ¢š~¶Âª© àëäE©=Ùb݇¹¤œ¡Ùþ žÆEC[{Cþ»¬ Iœ×q^š†àyFl‚"¾Òkp-¾yºïä5ý“­ß†:¿N|É.²ó·0ˆ€À'#sí "—–ääÀZœßx¼í¤ó“/Êû¹ªi¾îçÎîïp'î(qZêŸ+ª>ÊÑ®ÑáÜ:ýh–ùÕÛ·»´oÑÒVÇ¥+Ž†Šr‡|–ý *¯²s‘‡G)Òâ"ÇSAS°AÐN¥ÂüàUj—FÈ;ÒÒ“/í]Ý-q’Ï( &ï”™¼T2Fì†/ùÅJ"+—È d†`“d”> stream xÚÕZ_s㸠ßOáG¥kùŸTgúЛv;íc›™>ônn´²«µ¥T’7››ûðR¦9q»½»vü ŠA@ó ƒßlc¥Ì å6Õñó½ýÆþÃ;è¶@¸M(¿¹{÷á£ÖÎò‚|swŸ²ºÛmþ–¹›ïîþ´‘Vä†`â;?~¼ûëÍVJ™É~Hòî÷wÓ4Zˆ+åAÊ·2Nå\*š»n¶JÚìéF¸¬>è-öŽûšU×÷õðx³ª®Ý5íJ ÌMÊœ+™+ãâº*Ѫ™‚ƒ|6õÝilÚ:ŠÑŒûÅÔMûxºáÙøOcøŠâ>“†Þ]~³µZgwûz¨½Ú·Üè\q ™Ãc1¿ÐŒÔÐõÿ€7UdCslex¦cGÝ ZÓSWÕõz)ÛòÐ= ° :φ[è6&«¿TõãH$÷]?ñ=e5Æ®2Píë¾^ʼnp,,€Ó ¢^„äYÙ÷¥W ½6=«øVû°÷ë‘ 4Kûòp†·;jм8 .«¬ûúŸ§¦¯ÃåÎ…¹ øu¢íöñPVÛØ—í:8k ù–i†–‡'e“Re|wKå'2>á³7V çyäcŠœ©Àç·ž‡ÎhD óÕ·kæ à í°]3Õ"—ŽGŠ&°õºÁlvS¡zðeh~ÝÝý‚î¬9'³ÛЇšžQU];€N붪‡%£8u×.M T*˜7sÃs)ùÌÊ=+.=4&Õ\mö<ÐÇÝdÂuOºƒÊ¾¾? q¬[56]›ÓüË)œ¾n¨ã„hÔH?MM&åç)Ç’Zû2NÝëvŽ퇑é~, &ì™Ú˜¼°Ò!…É5ø¾-Ï4DÄ#•2²àù&¡ú–1±6#ç9Wfa$¯OÈî Åõ‚SQØâ°»ÆN-Záëª`Òææ|‚`œ½*Ù.P™„ ‚ÏÙûÃdlm‰BL$|í`:ˆ3‹Õiæ}¬¥÷Nâi{ãÒ&û;ú"ìïú]Ý“×Øj%02´x­ÊÊû±Teä"*jD§‡âˆ‡" !{F¶><¡ ¦ #L. 9·T€µCæ+ª…æ6¿Æª™5W+Äê™Õ?Ñ„âúBl)^³jú*Ö ¬Zss6GÍÖÄW¹Pò3ý¥,Ê© ~³¶.:zŠmACïéÁñs<Çt© Œ+aÒÂêí¤¢ûûñéû,'Fh+'êÏ7pËÃɇ4˜ÁQ˜xõˆjGGô–’dÔ> c+0t+•ôÈ ŸÃ¡yØ#ç1ö$x¡§B;þ[xÒIÆŽ€¤$8‡/Í}Ž¤ðò»w9µÐ”ÀÁ¬áœ]ó¹Á„Lk`¡Y»«wÔ¹ë0²>A\åm—’ÊÒg¯&’€…u°aÙWæ|럚!Œ¯¿êô*bq…Ð;”ÇÀ  ÚöÌye!aé[X5´ª¥ÕY[CF?”}èU‰ª< «A5p·iŠnç¼màm³CÙ?x`ñbæSnÏ,bªm_Ö  ~&¥'•t‹aa¾a_öõbP€ Ð:ú8{„¼¶°CçW^•X0&§Ò êO¸) |Ž¯).“ÍSËÍC:ªrÌV¹ŒÔâ:ñ}ªÜ2Ï`O% 5ß5$I4‹_Ø> ãl´€°s À2€%6x¥Â%¶¢õ@s†~¥tô,+ØÛ#©©#†!="¡j’ÕK„£³XæÇù®­ÌÛÜ_64¬c•ÃH­]“ܨD¥îüʲ,dKÖ¥W™8%TS &Q(Tz¡VKÂ=—º#¼ÜØå‚ÑùÐëÌP‹IÆyM×k¹yðI‰\¶ á›ö˜N×1…ÊÒsmͲ¡pw2‘M|.hÓB–ð”\ƒÒÿ 7†[ÎŽî8qNc½¸kM="ù°Æ½3®¼/å6a/´Ú˜,ñ…Ñtò=®Ø­þË‚å–Mf>ÉÿMAo³¬Ük"Är¾r~ÞÖwÎð¼äQl‘Z¾éœåu¾Y‰ÉÏ‹_!Ê$·’ Ù9¶¹°"ñÔšÙWš¹3æ ÇŸÊó"ÜKãËá–c¤?rÐMKM7 ódM»¤Ä€¬õ±&ËŠ7†ÁŽ“¹—;ħ$ :Ê ¹ìCÇdøB6}E½=í›”¤­=”€‡©qÎwý2š±)°ÇÓ?2 B„Š!-¥@$æ!¼€ œ}vïK‹ÞÕMàÀ;ý ¿D{ endstream endobj 784 0 obj << /Length 983 /Filter /FlateDecode >> stream xÚ…VKã6 ¾çWäh£c%¿â½lÛ-¦·¹m{ÐÚN¢B±²–<³SôÇ—exv1@,S||$?ÒÃ÷)üñ}î«,cu~Ø7×]ê¥ÃyO‡?Ûñ —€b²ÐüpÜ=~,Š=OYÖ|<-]Ûý§è狼¹nˆ“,Ë"ñcœäyãCÎ Jj¸Ió¢Œêøïãï»_s°Bˆo£úò†++à'¯«ò•Ÿ®|ÿ‹Ùýñ¦‚î¾]€O+ÂþQÇ¡ ÐàŸ´­¤S¦§;s¢§»åV:‰§<Ò2‡èÌèèödº”Áu‹©˜r FäK3„ï±ô_ƒÒ 1/XZO*Ÿ)Ô†/Pã…˜]½¾ëj7tR'Î$¹Þt÷5ä7ÈÞú+»ïV.åP9j+@ì‘1.|ÏÖ’{J-lßiKPXµEÔ¾-÷XxV±Z¬ð¼Oè ü<‡[ž­a´ÝM5ÎßzQ‹Hõ7ì3Êdß’ :ÿ&¢D0ü¸DZmg›8áÑóH}Ù!êÚN”Au”W‘q—UØ C$æ”Ü4Kï&àAP à?¥(Õƒlº»îZbç_i‘¾(Œ´â6$`0— ¦úV50 ý™îÓÛ®zF­Ž”®ÝÕ ¯¤ Ï­ñd! "DXZ•îxé,ZWõ¼Ë!\Œ}R©Õ¿T©:¬ :==ÑÐád´&àWKÒ03dBF waéëaÂÕ³jÊr°™òð5æ”Ç‹Ò°ãò ÛÙQ;<§P5’ù²¢váYõ2¨ÒA¹m¤F°(À® {Ž×hAäÀe‹¸•(,yMg¯@Ý];oéìî"‰•…*æEÞcdÐF:·Ú׎ÛÜ~DmÎÀ í‹B+QA@‹_ZåS$&¤Óe[¬@ߺFŽÈ Q–aà ¥$8Í¡èUÙ  î–_ì¬/~Yô€Å«#u VnmÝ}œwd6÷u]1HbZ |kÇX)æu=#_ð,D$ÓÃÒ´ªX E˜V¸q›;¼f¢št±zNhøóT ÒjúWÄ(9|ŽΡðS¯ËY’¹¤¾Ap cõHƒ&‹ïBRÔUôÔ—žXxj¤íü€V&Á7ÌßÎpöÖ½ öIíçr‘aÿXþ­~”â›­¨YVÎÍBVmùÈ™Èg7?ñGñæ¯,gÙ»¿²¨eûÃÿ:’ /Ïø* ÅYªÞo&œè°]îlû£lÙ´‚ðkø?¿µ|3 endstream endobj 790 0 obj << /Length 2638 /Filter /FlateDecode >> stream xÚ­ÙnäÈí}¾ÂoQOÜeÕ¡kƒ<$ë™L˜m ³ CnU»•iIãñ~}È"uõ–½ž$0`QT‹Å›lyŸ¼Ê«Dk‘™ôj_½ ¶}¸"àãßÞH^·……ÛÅÊ¿îÞܼ¢+Š,ÌäÕî°$µ+®>2Üü²ûáJ'JÄa TöýûÝ?7[­u üá’7ïvÓ9‘R¯dWþ†£82B›X1Gqj„Ô†ÎV"Úl£XÿØleдILpû~×!Ížm~"è6ïsäñò² ™ ©ÒÕ­â$è†óy³U)ï‘jžf¿li[GÖ5ýѶöm^w‡¦ÝÈ bÌF%ÁÒIÛîK™À¡.¥±³—-ÙS™¡š„ÞªÖw ™Š$ýFa ßöÍŸð.騲FÞ{Ûò½½&ÁtÓîV‡™H%¨Bj!#¢´ÏO'àkktÜ;aØ}>t–0eOÏ>ÿìDÒÑkNÖ‰ÞèØñÃ\Ç+®#)t4±]4ÃýÉú®§Áb½¸žä“Ú6Ç“ŸøµfV›¡?½ŸqW2îBSq·8òµº¼¡òW+H:2D"3–Ž$NvG‹²Ð:qŠ žl êèÃ~h[Kb?=êR%ˆõÑðÖck™Ü>ïíCÓ–¶ûX‰ähÚðÉ™6žH¦ ûÊnKH'G ©Î'ûuë¸×l°mUöe^éT²´.wÌO&A¼QÞƒÉD2afÜÞ=YºŽÂDì±ì±ÏÔ73¢¢ t÷TU¶oŸ|L¢Ñäâ³ÈÅf›e2(ÊnßÚÞ"1\¯+k{ƒÿÑ£ÓÍ’—²ƒÜ¾ßu7·?í:´Æ9!E2º ù!oû“}éMÄ–ÌͼÝ~˜hb d•š °@³¼'/‡RÅEE*€ Ûç%ï¸gF¼3BÍéÔ •>–õ¡:»ï˦îÄÈ‘Ø9š"JJ†ˆÀÒñýšˆò’#àxßýèQÔó"´Öª<•9S]EB9€'ˆ2ï·øÖåqy)‘eÙèʇCÿxw>MkW^Ñ!2ѸôM& ìWK6½5¡ïf¯f@eàé¬Ðà d3¥äŠ ,ª™rÆmC-ÂlÉX¤9øÀy`}Û8¾š©¤È´^1Ctžî˜“—9ÒR$s–à˜eT*²díVï7©†sZãÚzõÆôE3AwÌGÛl”%~EÞjË[W&òïºÀWº6ßÙ ˆæ×;`èNxÿ²îéc}MOÎ~[z±Ú®,ddD¬×ÆâAúŸK`BLg¨»ò¡¶üápÊœ»ÿ‰ÓÁŠž|Å%ÔÅ%B>gÆÈÿù^òÅ‹'>óY^{"Ýÿ[ úÕR˜1ê…«~‹¬Â×h?¼ö%£×‰ìåÅÊ»ø[üûò] <ø3 di£¿ëÛÿÊt" Ú­ï° î¹ûÛW:ÕeLÛÊ8 Ïr‚ÏúaÀ£4”þ¹+§1S¶„ ¸ÀT¸àg¨U®§ï®Z$‰‘²¸QÅ.n* Zé/·EYY೩©ÂWëX¸<.î-°úòahÌ»JÎE*¿»K…åˆG¬U¶©Qü 㚦-l‹‰Å,Nçl"§úž a5]@‹IGb¬êWiøÍ €à!(q½†ùÂzS¸“LÆTãÍÜØÈ”¸»37Ôå  )‹v¦Ë±JNyð||êÊýÊ_P…ç³Êš“ÞøW(ÉvˆYPSš$ þâÒ,áF¯ƒª˜*Uo’NQ<ÑëoNæV2ÁVÒÕ„€C%ŽåÀ–ñ"…öwmF”hc,ÑdðuoÏXÿ .ça1 '‡ö àNi¤¦Z£—£_Ý‘ˆ"³lúï>¾§ Cm(ªt6{>ö#uC7§¦g:ЮWuwä6ÐÝpÏ%/íãš×°–u5ê"XªÊ0^ËsËÇ°wM¼ŒÉ, ‰é*ªz…Jß´áó–Šb„«ÆE€ (”ñ¾äNŽ¶¡ ikù+:§¯Ò‰µ0 oÙ>§Œò£ÙÔ…—Ò‰Ò_A uë3äXDR}ƒJe’ýs[#0ê";é_´–D/­¥»ž»(T4ê¢äÉC1¶KîíDásjÁ¢àžš`3µ±^O4ô Óµô1”ŠX™åÅGÈeô‹Q :žŠñSóàgÁÉ€šLºžz¸ëp2Ð+n.;ºþв 7Š¨ïk >·#Ù[âLÄa´Vfƒ‹þ’¿¼Ë÷HôH3£)ÔÒÜíèr1¹o„%ÃL„z¼¦Í¡?o!Gõ. êt訥4P{&JgZÄfÕHMåˆÏÁjçèÆguç9ÒìËÃÚ kž!#Tà̳F3Îæp’¦\Ì\ó1ãbI3HAlÍ„dš¢2®;æ=A“ÌÖ }þ^«˜Â>w%Pð'•`ž1¿é¿ïž,zäƒe„³98Žža3ó@Î#4tô*»Hr(ÇOXªþâu)t8™ÎK5IfVäSé¥gŒjJÐ#û,¾p¥¤8›t<éc±tv_„Í H$Óp¡~†¥–1@‹ƒåÐ4Oa€—=E5”¢ëŠVin¶"è¿›z0ÆùQ¿ÆŸ[[`^35EM•|N9nÅn %E0éÉ~`kxl0åpÿØ Ø$z¢#ø¸q3w— ¢0å‰<~Áº~èݘ9á“è, ׬àë2å3ñÃåŽuIZ–&g Í>à nŒHÑÓ;t ˆ @Ç€À;Ix›,™Å`½åòóÞ9MªÊ,‡@[ñ€žÉø÷ìKו8òÀ¼¶ãp™ñ°ïßCÞs Ëh‹PU~¥ß.K}b ê"oq(]gS”¬¶mœR¹ú~T\‹"A†k!`-É Pc7Za½*aj$xX±+Ã( è Ï–ko?ì¼I7Ì™(s¶ … àëÚS’BÝÚ¦r¿ã(]#ýŸèY׺R§LGoÑôâ»uýÁÎ W‡®뾬ò4ÑŽ6•·mFçvÝ‘æßÈ z³Ýäœ1ÔI9ïìA«[ΰÎ÷¶ë(]uæÂð1_µgøÇ#÷刿 Äý›CµöÐŒ¯`Û¼À9ö ʧe|Eeðæ¶.½qCƒÍ¨Œ‚ŸÊ'ê¸Ú„8.âê[ÓŽˆ¿ý~‡¿jÐËžÝòyîÙð Žz8‰•ôÛ$§œi%åøÑëxKûþ‚‘žŠ30U4¤?Œ&B®Ð“ ÏÇî âÅîΦ}¦ÙÇ—²àä‘så~²Nx”Š*¶ž~îÊ)m‘ØZ÷“ÈLŒ?=ÿ/<Ý endstream endobj 794 0 obj << /Length 3750 /Filter /FlateDecode >> stream xÚ­ZYoÜ8~ϯ0öeÕX7-’)%ÈC&Çfw1Àñb2ƒ@n©cmÚ’WRçìß:¨ƒ2;qfÜI‹Å:¾*JžÅð'ÏòøÌj-ò$;ÛÝ>Š©·{ÆË¿?’nÞ&n3¸ztñ*MÏd,ò8—gWû%©«òìmôü¦¸ªn³ÕZGêñf›$itµÉ’è8´]]6[•ç©‰¤ÜüzõÏG/¯¦ÕR¥ÈμǗ¾´ˆ1Ç—É!uÂ|)‘ ¹Ù¦F7[ÝTÀ¢M¢×øPldtØck×ÞÞªÏÛ}ÛÝNI£¯®ÕõÎA4Ò £5/ð¬ÙlmGê¸ñ¡nJnµ{÷¾ñ$'Ež¥@†^õêêçw—êõóÐR&1¼ì¦þ§±›•/g%B%jœÔ©›]ˆ”’BåÙ‚”dwm×UýœNµMÙsïÐòoA›“Q‚R;n£X¶ÖæÄO«ªõ‘&&qœœcDK:VhœÁ~rø'¯ß³<èñýí‚@@ ×ËÐÞÞT»¡nÝ©(¡‘·3Õà ·~¸){§ÚaVG§í‡»ãpÎRÚq±Pªœ(ÝP/üз·ÕRëùñ”ÇF?싞ܶáí‹–ÚqtÛvnÆÎh}ÜàKM]ÑÓÀcà jÎnÜôC1AÇÛèŠ{ã¨nðõøjÕõUÐ,¤Ú$£ÊßìÀ@†±çñÆyòÐM¢ÚàpT÷üKú¿dâÓ–±IêÛöëçê2´n*EšLSaÆf°¿q)R_÷Ãja– 4š¶ÙvUylÊb’tãyŸxc‡–Šª¦\aû3Ÿ»Lµæ¬àÒl¤à)éY–üØDŠÃŽBFî vš4ÀŽ¾þ­ Ê*ËDlä(€&(#a•gœ#=õN*ÇFÑóZ…[ªúï±jvÕÄ@hY0£…V—ÍE2;èŽl ’ÊoQÜ×ä/AóxdᙽÅL.l:m lׇ*´ N–<‘UÓÙ$ÊÌ?7Y:Y­G“~Ì„öw› cA·a…L9åvfnSäVgBr ­šáIqˆ\"{¶˜ôTÀØAظõèÉôLþ; ªï`ð1þþ±ãÏîtkÉx±˜Ê…$0Sjž†xŠã):?õ}ZpN±¿ÁY©<Þ@¤^©ïú('åý›•iEj‚ OCDâL­v&ÇEn嬰q|Ÿ†j6^ù'V@r*{^íA!ƒaДÈÀÜ!ö‰à M}]uÕyˆÐ(›LÒ¹û&ø!ì›”õvB#}¬œÃ®»¢¸EÑÀM ¹ hëYk‚ «Iïc ]‡ ÂdÄ­:™gTivrxëõmñžÂQÑ}áÆÖ$#{zQ¬{‡øÞõˆi ­ Jrp˜k‹ëhEQ¯…‹áŸË |j„ƒÓ:^m2ÀUˆy4¦4íšp÷]çU l(™ ËryÛüÄôÉbŽ äW'Ž´4›CÌOXSb„qþŽ0^ÛÌh 6sàáè¤sSôüJ=ŒòBMd±f1ˆÕx‹µSë#Z@ ÚˆD-àÛÛA!h%T6ù¬D™ÀèÈ:A ýßZU*LÍRöÛð ¦.r‰N€(ÃolMá¹ú¼«î8#Áþ°@hš/·æ>}Ê$âàÆ5œèä¤Û.Hr°™1Z\¨ Å"Óz™e:Bl¬Àß‘giq@Á¦YT÷eÀóæé7ЖÑ‹hjH·bhß eý,lq¾éÉú”µÜ±+úªÇƒ0£øÒòüáñŽ¼ÃàÆÜj¿U#QÝkCË¿ý—[HvºV&‰X®-øˆ™°ö˜³Ä9)>àbÅ8 øœG8¡‚.Æù1ùG×R@À–…o©|A³{^k™¡œ*5hðf2ý:Ø…¦Ìv‘âÇ 8“âp¬úø—O§9¦¡Kƒ¾Y:9D¨³“#-¦ŒºUó¬$¨® ¶¿/'À…Øê“ècMèºÆÔ°˜02÷c(\æŒþs  ‰p9§¯K$aT´UéÐýáë“‘QYÿK]uuóÞM#øMj—ÓÔœEyâp”JE2C¹¯‚üdñ"DˆÍ'ýõ̦f¾Å!“Àz÷©îÿÕçb7¾¸ÃÖÞ–¾ÖÓVUª¢¾¸u-ŒøKÒV©œç¬ÊL$<ìÇ„'vjw :u.tÑ#Õ¢A!fàጟq¿úéòçg—/‚!.[jÏåøóçÿ:EìÀÂјpõ€œQ éò¸u,M±f¾ kXI²Ñeµ+˜e’ȨþÈ£ÓÀæD‘Ÿ‹Î ›ÕïIvéë[b_­}{8´ÎYÂÒ„Ò¨Œ¹8Í ©,áR µ>Õȶ` ÇÃÀmôÇø˧ ¶«ÑǺ©e1Ü¢\^«ï5‚&ì»> {À(·}”"ÄuìP{±ÆtŽufð1õ‡ ¥‚Y5sJÐÙ´UAßâ1…$râq¿>½Crï@ÛÞ©G;ŒÐÈs—4ç“Y»Ö{ÝzeZwýøòÙ›_¾œk*+D6jf$MXNEÏÛfëÂh ›”^¬ÀÜàΟ ×Ê­­LBè“I×£šf±¯ž™ ®Øèع÷< wM #§Š¯*Nædï+‡ëW%yÝ–Z’cÄG€x)6q&Ç͹*5ÞnY{Ïà]r¾Üí"ÉÇŒôp¼m\ï¸YlŸ®4+¡5³øa›ueƒ×\R÷‹(½ÇËšaW¹YÄ\¾íèÆJ ×Ï…æZXcC…Z&/øˆÉ+Q×rÌ$~?¸6FÁjÏW-ð8ßÌÀÃ:ó„Ù²Võ‚̈ø^Ùf’À3/óFšX1ÀyÝŒ«¸;)©ÆÛ…Þ€œ5kBž}A¡õÅÎCÇc:úO#"˜ÓUî> .¹ëØ»¬`Ðh»Ò©?Qž“ën0€À.º\Fãå‡9C£LöýwÑüúvñ~è*Ú_Ä¿‰Fæ•H¸ñöG ‰ÝE…„é/Ôß»’ÆÅWÒË$ú²šÆÜ•4Žò•4öÚµ 2ÃÎ$¡l)¿õ¥Å8{»˜ØõšèÂy’’ûßf€¯›¨Š\åóŠ¿Ð¸¤ê}•ÁŸ`€ß#¬š ´k•%¼áÄðÈw-hNuÿ¸sö§"õø ‡ûH„¯•s—L´Ç®¦+ìÜze'¿œ“á™ÉXªâ®ý±áCùT­…š­|.¦2½ÿŇIÖ_|èû·/þjÙòósx-¥Wñ=…|ûi÷DX›ü.ÖõIÖÊ –êOk™ƒÚ§:?uó‘Éä«!,r®áò7 ìébC⻟`ôòÕD*¿§¦ð]¸`‡,»€äBâWÕ‚•æ›²6Qõç¨Iº:8ŸÔ$Y}fd–0êkþ1]RßÍ:ª>p0ˆ÷þ©oëCÑ°êÑ0EÑ-™WólzaâÎ!k¦ü\ÖýrÕŠŸ8çsƒ‹ÄV,an0XòØt¯² lçzátw®LSŸ«xÛ¨©ª5O'>þˆ×_@¤ï]ÄïÙÕÖî·âp[O_~PÝ 0À"T»N~øÙº/|2Ƽ²Â¶q_Þ½/êÆ-Ø_CˆSëâüúÅ}ËE'ü+iUyÑßaèldOeÛDÚÑd-­Áà› lÞ W¬°U0tf7CH;‘­¥zý–e=ŒŒ9Útç']„@ÚöoÝ$ÇS6/;z±‹É7,÷Â[q÷_£oÆû–}>&€yú»Ñn\é²·|üD‡[÷ëÈH¯^TYòÅ^ø0_TÈ”#'öŽõA|±îO×rTì€aº¸¥€?Pe\®äÉÏC˜ùۃ˗'>ÓÔ¹H­Ye¯*vwJ.ùñÜ3Ž9&â“LŠr®ç]â~:ÁI íݯI‡ÇIdÊ-¯ö¡q7Ên¼çnÇ©œ}ŠóGFX@Qž” ­®Eæ.?±Ò§÷‡ÙÎU¢ϯN\@/ÂÐ…}Røå©œÜô‹7aJ€ÍZDÀ_°r¾¬yƆjÂ^Í7õ´ÕD»Ä)I Of4ŽFmã)ÁHýv]=•9ù5×À œ3 î ãÙ“?—i°SÆŽXËÁ)EÈB¼B¥¾W¨Ôs¡ÒE‚À¶cîÙ!/õ\ÇL¼:¦žë˜bʘ1ü¼Ey” endstream endobj 800 0 obj << /Length 3587 /Filter /FlateDecode >> stream xÚÝksܶñ»…š/¥RL<ø€3žNÇ­;{j«ÎÇÓ¡ï(‹É©<+ʯï>@‚¤p’¥8$£X‚Àîb±»ØÝ“G1üÉ#eZ kò£õîALÐöý7^þãtãV0p5ùÕ郇O“äHÆÂÆVžM§:ݽ‰¤:~{ú¯#)‘Æ)ÌBЧOO¿;^i­#-à‡<øæt\'Qê#‘·c”æFHmxí¯Êu±ïÊã•ÊdÔœá3ŽúsØ˨êÖÇ+µe_òÛ®Ø]l«úý ts5µƒŸUÔ (6›ª¯šºØ2|}¬ò輩Öå£ã•‘¦^®WôƒÊÇ*‹Êúaƒýh³qS·Í¾ÞùR ™0øY¢7DöðFh ˆFMUãú=¢Ú¨iù 7# d<8ÿ‚açÅöì‘*xÅÀw4°ìt‰ÿJ\¾¬ù­7¼±®ìþŽäçÑé°Ü¶èû²å!ë¦mËnêM7£U2±}ƒPØÚ|ºµÒˆÔŒBÕWg=²€G΄ÀÄ"Kõ0’¨6;ì e€ÀUh,Ÿ /²%´†2"ñØ HàŽ£eûá8I£b ›ar¯7üþ}õ¸Øq·­º’Ñ’Ç’´(TèÛ¢îΚvÇ]ž·­ ^ˆÙ·’I.’Tι¸)ë¦/aa($—6U'2zùÍ“§§Lpizx‘$r¤é]ˆnØ•O·›ÿ%ÎÿâÀ¼)4“ì–ycaàT»1BÓäBeãˆ"4‰…ÓSܳ()Œ‰ âÌ”~¡¶dcæ,ìÊ|¦žUMê„´IÑ—<úò¼ùsŸ$"´ª/ö=oRVÄÉì¬'qsà¼Q7ÒŸ€Ü|†i$,—%ˆâCBÚ;¿Øž=w­•'"ÕêF.ÁqÌU`1â4è„’üAg”?hÓ‘‚–A­F#é`z„ÜL»DŸ›•C·êfsãªâx•€¾eÓ• Šm׸ÇB+=? ?Ö )08»:Ï#Tïø$­vÅŠÔÁ³èÙêÙkv„Fž#Kn?ùúÔMCò㟼:=™ ±Ýò»³fß.Wcµ@ÈÖí/œªn™“ðrä$¼GËú×  mUË&wÔHgź¼æ(6Ó9M)4œ°©‰åOuh³WnÜJia2·)x-­[ ?V¨Ò©yÖ8X?ŒtW³wCXQ“Y„0ÔË/Fø5öÌq%¹0§Î©\9qÓn´’ؼ¬`»BƆÐK•³uh-‹öŠAk°P`âa—’ÄìiïG0'uK¦Z$)ö€uQ2y<ÒQ\³/”×™m;˜eœÛŸÚ$~®‘ŒGÝVP‘Ù8m[¢™ÁIJö;üïl~Û…VH‘[ë±óÃãXˆz%߆Ks‘©tª#Nø<0Ÿ™Êæ.N[jß̵±5¨ ‘!z˜If!ì$ø€ÁÄïü/ÉX‡°/U{7y ‡xõ Ñ|ćgðҸ糥¶”1ø ^óþ6*ÒŒ#£ôädRpbþ;­§ïº^h MÒr÷ÙDLJùÈÍa”¡‰VÚ$"óµ’©ÐZÞoex&»¶$ÆÏЋû=ÜàD~Æ=xïv{Î媹w×3+²\ßÈuhªüÛg¬‘Ÿwˆ©¹0ÞÍ šèD?Í£ÐZxo(qÔ±'úñDÛlJÐhžÍ²2±V&sÕû©DîS°',µFŠXI'´êžÊGÞAù°ÐŽ¢‹jÀ‹.öžyÙÕÚL¶ñ‘Ëé}åÞª!Ô6›Û58\Zÿjab€¬ÑW?°‰ {põü•ªç.»øú>6°Ë½;ÇÌaŽÝ¤\’[O¹õôȉv™ˆ‰‰|T{H4øPaºó\äþ¶õ»2e¿^Û‰ÐäÕÜS±¼¸‹WójîÕ|üÖS®þ‰Lïdîä­Òtƒ(Ùß…Î=#2¤{뜇Ýð³Í¸×6g"Kò?«â©1ÎûGP<ò0æ6½“¦Sx_½#ï,–÷N©þÿ‚¹Ô?ëdY¾­úCÝÊp:Ò_©¥î"¯ï%R¤6û)¨;]»þl¢ôɼ"¿\ ë¯È–u)»ç Ï-†Ð(ÚF-ŽîB³»Úíʾ­(h‡˜ »Ø^¹ CÇЬå Ô¶y_­‹íg¾Úœ áR-”Ts†âÀòqå¸Å° _r—öÆ`Ûí‹ŒùÅ"Êà6F­0€iÍÁP®õY2¥ä@rT¦#žóùÀyX‚8?]“ÐÎèy /\VñóÇày›Ùa"ÔµvÓ4ìpÒC ™fz‘Ø•à‘2h’©dÁÜ>ÁËŽi­‡.ïN 4§ô¤cËyDûKÌ ©ÔÕ(¤£Æ£¦SG b;—‘¤kð=+—7î«w¸ˆLìðËô¼’Œ^✟?ÙQôe§˜àT½ã^~NGû Ü¿dÄ Ò½^²ÀßILû»,©J²èCµ.4 Sð ÎhÊ• o€Ç5Ã@"¶¥›§«v¤ôî#dØ£Lqk¦3;áf‹fCF.»·î+>ÇÛ+< _„òC.y—¸¬_â(—Ü¡„¤b2ÒÚ蟫ŒÈqlQflîÄ>mŒ›`Ýì@±—n-ŸäK†ò™|HlBc_×M» a»+¶pŒ6'£†_÷[§¹·Õ®‚{ãyYœN6‚w¹!m¼ó‚8nËŸ¹ôvn™®ñï÷ý8ƒ~_ÁÀ÷¬Ùn_]‚ ÙØ“íŠ,$6–BŒÀ«ªÜnÜÇ[¬}©ÞWT–ƒà™±£¤h±NHÑevzÉN¥Y#Ã<èªKòj˜Ïû°Ï•VxuTM¨ äAÛ|T±ÀÃDF 3ñ„‡ü/š{ôYFj\êjzrQ“×Âwœ.ymæŒüÉ¡€ï$|&PÎfaV‡J”$›w›÷ªïxâƒNôŒžÐX½r€8 ék~+Ú–îý„¶¿Á|JÍoÐí¼¤"±Îr‘g÷Ù‘àþbÉÀ£b(]˜ì²s[Ǿ‰cçàsËYxa+k°Ö®cêâÙgÓ2)|U.nêÉùç®ûd¨Åª‡…¹‚`6-xÛ°²«Õ™›·e¹€L†r®ÎbÈÁÅÆu&'–€¼8¬d¡\|w2+ŸH°nŠVôS,¾w«²•ÀU›Ð²åO{ 6XÐjlË¡P+¬Nâg_òœJ:gXŽüx80KNh¶eäˆÏ3Q—ï·hÉXàbìfƒû!vªÉ½bئú>–4kíjXˆ)5-1¥íj.m´kºž[%|›®«‘D„š®ƒRnba|Š÷à™±353ü¸h]r¿vKqÑ©ºí6ÏÀ*7m‡•Í¢ºáÊS뎷ª.X=›é)øÒ>=5=k›Ý0wéî¦VRDgn ·ê—·5Û|=”¯o¥ÙäV*ù.Š%™ˆ•¾¥"E›Q9•,Ç %­ì<髃âø/h 4ú¸$ï0Š,&Æ‚Wvý!ÎTq1Ñ£¢+™‚W>ÚÁ!Aý8Ûí­½©ó˜æjáÖ©¥w…@9Ö·.Ë r‘å7»€\î/V±šÐF1ù–ƒŸ‹’?–s4"TàŠ®ÈÈq?Ôu:ÚàA1uyœ¨(x«ƒ½´Óùë¼ÒŶ¨kâ7t.«íÖIX¨©ÞlË~ßÖª—´Çëù¾ý6È|55ç®Ð½ âÉà¶N®dnºªwZ?ą́*éiMÍÌGÛà¬No¸çY¦gfûÀŸÒäO%ÖÙl\´åš,9vvÀáŸ.„¸L ‚ðLº)ø»4tMÄ…{î;ôz±5x×ØvµrØìÊrYšoáx¥FYªÌ €2_€âÅý9dùãÉ·áRýa±æšÔ9ÆÊȈ*zóûc%­—p7Ák øš/NÝ[$•Sñ¾ts¨%=îNÁgŸuøçÃè[èXN:Ý,o<$£ غxZs±åø$ŠõzßõÍ÷û=ª/×<÷ä§Ù®ÄRd)ýâÍÛøhp°‰ nùÑ%Ú)C1ݽzðoÆzZõ™ú –-sêYÙ§BÇ3'Y3Ј sb;V|r„Ó¡NËU¶Ôýj¹XÓFîåëÕkŸ²øŸŽZò„)qyWxÎï½Ö´NkâGn•‚>DŽÚYë"fŸ™žp!„ôÕÌpWÒ˜v¸F¯Äú˜…ó6m:—%˜.ïqÓIr¡õ4I?©7¥c ¥#½»5¢Œ¹¹D$FMÁIòõãf0À`ôÐ[o€â‘ÕrŠ-ø qš.4Ú]ÐxþŽ@úÌøÌÌ wß‘§ ?Áö¦á—d?à” ^¿ûßÎa Îg6ün¶²ƒXÿ/&±S endstream endobj 808 0 obj << /Length 3764 /Filter /FlateDecode >> stream xÚ¥Z[ܶ~÷¯ô%ÚÖ#‹¤®m¢µë$m´ñyH‚‚+qvÔh¤±.Ùl‘ßs£.cÅP,°¢HŠ<<×ïŽ:Dð§EtÈŒ ‹8?”—õön|ýù %óŽ0ñ¸šù—û¯Þ&ÉAEaêpZ/u_¾ ^ŸíutýÝÑèßßã8 îïò8˜Æ®¯mswÔE‘¤2wßßÿíÅ_ïçÝ­%Y8óÓt¥y*3]o^ß¿zóîþ%–æÁÛ·÷ß|6`; *÷]¤â¶ë®xølïtüˆÿ÷X~”ŸøEZÖ0¦‚g¼TwâišuËݧ¾[fåó¬ñìx¼ì†WO΃ÓÔ–3)Y0ಠŽ~T&T h<[Z/Âuž¹Uv}ï†+л´w^û;©e=¸F¦Ž?mËOǧm_uôußÈ÷oÞÞsƒéŽ‚¡þ¯C‚^½MÍšå:ÏÃHç@'øϹKX@‡Ìï@’qð®»8!fÏ]?ÈK3u[6SE{@{²°ÈðBñJ¶ªH~¨^q^<ÌÔŒõµ©K;Ö^”q'[Ž´ N„3ñIâ5•*KÂ(b£H6¸ze}?k¡Î@½2uHÓ,Œzøí÷Ñ¡‚AU›ôðDS/‡$ŒéÔÍáÝ‹íië¼V’…¹1¼§fO]Ï„ƒð\9ºŠ»ëö:r ‹²Æ)ÝrqàÀ[Ò±Zæ\ìÄ„g¢ò?UÎ[oø9ö¶`ã Ûq׃\‘½/¹ãaBÝ3q0ز¯Ñ~J\‡X«¡QÕ=ËÓÜû dVo×–Ž;Ia¢åÇð|¹¸±¯Kµ ¿‹ÅòÉ€]&Œ€_È.°‡E&>ç›3|dTp†/³ÌÌèà,ÁîcwÕñó§£à%Zçª?í)*[¦Þ}¼«™ì,“fyÐã©z”v’Åm ÌpÙ ŒÞš—|Ķó*ÇVëZ´¼;'oÉ<ù™žz~»öÝCã.(®Œñí…B!5 xñaÜ8ö¦êmCi"Ÿ¢6Ð*ïÀߦ®.ɉÀ†104'<áù¼Do‚ ¯lüfùqrOØ0à¦@Òìe—N¾ë]³¶V˜ë]m[> i?Ï®¹rËŸýC‰>5ŽâçâȈ·ÿ“‚N~ÙrÿÕöc]NíñŒZ_ÿä)^ƒm Ô«"þft¾›è‚a5wMýx¦ƒ¢ßÅž“(<ÒÇgà׎@xØ×Äô*‰¼Ow%Pòà0@~+•‚7ôÎjÄÞ§sÝ8bu€, „ëÍ}à.V \ \óX_ȶiíÇ–¬ý¸G6Jß•|CÓáûISÁþi`€Oƒ®B,Eœ‚?MÈFR°{““—ÎÁAƒµ H &!Þ3壟ˆNC%¢÷$—i 2AM†ŠšHeÞ‘AK´›ÚÊö2VBt­ÑÕJ †.¶ÒÕgk ýÞ_ã(oærÕUDŒ–ñ”G׺žø´Ãrð°dK {‹jâK‡†ù£gþ<É°Ý|„Zþ‘” £ø¦¶qÃÀïtXDä!ls:öÂŽWÛ±Èá\ŸFT$ ø€vy:Öí½úA"øº+¯~DEgêÁ+‰¨¥×@þJTʬ'¼‰ãœ•Š¢Iª"c¿<áVI ¢éQ„–¼?öâžô¼ˆsîÇySì;~ÎYg)‚,‚SØ^ $¬ؘÿ=r“5[Ãä.M´^e0P¤yÞCa&‰»(±Zü(þ¸gZY?SG‘¸ §Ãǽ`Bl$¤`2H!ÞÈsZÇ ÜD;3Eºè?š²ê"[E+˜wñ.’ÇÀ´†º"ƒ¬ý#1ß϶¯Ž¥x< ½›öl„Bp‚ºÚ‘>AS¸‰M/šAL¹þ+‚ÙÉ:Äâg½,Åú .‚««æ>xæ,)çn󓤡Ž²C Ï8#à… Œ4àŸÒ”ôl{ÖÙÿþ¸Z`SÞnÃ8†#Ó “cn}û¹ç3pãqsÞžßØ•·‚ÁS¡`ãþ½T¾xµbÙ*R·LP)À9sˆc­>‘ïÉäãjöÎo–$¢+‡|‘Ö SO8ÞçY¢èUZt1+‰r=ÏÎö¬‹ðòĪ.ìVqнV°}âˆy˜G·ã:â..Š)•/Ý þrÜijÂáa‰‚:LBƒø_Cæ ;sDH]eïFÁÃ_Üi(sÌ0C/ÌŒ>†‡Íâ$Ñkº1¬q¦‰íkcÛVX¥ê+<ƒtœ¥ñæ‹{n<°ƒ/íÝȧâ¾AœØö5¿ ’<4iþHn„YûI¿Û[X¤ÒOñ1 c½ö(*[ç3ÚW‡W7( –L7I’.ò0Mò­šï…žÏl§’®áBèD”‹l8Â&t»c–…:/œúÿí„–)Éj`ÎÙú͉Ð(©’ºòmífä^Ÿdäœdä$ò‚ã8>e'†éð^²GFÜë×Ï(}èQ$¡27péKqXŠ~É>S a9Yf塉ÛÇPϵ’þª4²ˆ?Ê`PÇ%”‰_…õKŽ•±Ê’Ïß$«R \Ñå?ëµäZ?ÔÇøBë R¤f¦yño<þ^MÂ\e‹Ò³/(T˜@pÝD–Ô¾ød"_ƒ†¸ãHêòQFÛÝN"Þ㜟SÛbþ×@Æ@Ù¹D×ñ:-ª·2̉ø–`½ žKÅ|{|“Çal>*¡’¹ÙÑŸºf]S©<À!‚'ܨ×V,PÒíY»<@<„µÜ‹[ª.åkÌ@±à08Æ-ž?íúú±æÌ toWÛ_ÐVNu¥5{<ÉŠ0ÏŠ_¯¶^ò$L£td¬mŠÊ•¶x²}j®3ã§ZÇ}š0&WÛbC¨ÁEÌÌç¾Ò=—:¸î»”½Œx7ÂRv´/÷0#7-xž ÓŠ/ŠP:ï'yVU9œwž¾  ƒ–•ñUq•ž{™`hPN;IU˜¿æ;5h÷ú\r‹ÎG=ÒAah·~H±Iœa‚‰0+!!©©â˦A’…H£`®gæ½ê‘ Æ’ïl9N¾žÊdbØ·€/j{w;AA¨]ºaðÐy[ࢠ¸.„Øšã.–AvÔO 5©ØÏ®Çê?hd_S½ºZ™ð¦“µGé»öîDµd\bäç©ï.ܲ2i)cÝ!À9*®yP8tË XÇo¹ ø—An äæ‚Ëù|Á¥âx{½µ®Óæ`¸0Nš†i‘o µC| h“=]1Ò„yT¬êà¾L]KmzS’6\d.VEæXÜSl¼çàí“ËØ,%VhÓ!Õ‘Ð]ÀB¬s¾{ÛÅG‹•`™`›Gð"ãù¯R”7¢Mk‰Â8³¾^Q³šaÛò¤úêVã Ò8ºúSJ|¶€??aôýB ¾Nó|ºQd¡ÿ¼Ë“àÏxÆ×ɽ_M××¥_òk°ü+mé¨p‡.¦•…8Œ¬$0µ›zºÔEÆ,4 Iai+ßÛ²˜~\1Ñá½CÝMòÞâøB ~#•ÚE€)åÈ3m ¨žyæm²Å W½ÿ¨/óbsÁtÊrêmù¼'#>F®•þ€ ¾÷(O1°PéÆç.tm°§ŽçÜPóÒä¥`Ypù"9þfR+Ãë\ÝLa}kìo—é˜OõÍ uFr!ðd‚tƒø{·Bx? y-YèóÍÍqcÇQlþ¨à™Ñ7¾Ì±´Ûúý$7_, Ly10CòCòû¦…Sj¹Ã!'dB@9‹š¥$*ö‹ €XÇ#ÝVõOÜ·oKí‡=6Osµñ–;ÇÍÃlI…kùŒ$?ÃAàõáúF…f)|µ£z ¥s=ÁòZ«2‡Ûþ¶'æ‚90! Þ¢ÊðåÏôÒ0[¤óÉMËŽý†­¥ œËmëµ-]ý$g!Eš~«åïS%Òð¢ÀcˆÜrþ©¹bÄ8¿ùŸ8á%)næ¯5Y—ñšLßè²ÛcIŽ–+s³µÒÝ +X oaÁ4;®ë­>™ÃwÃÉ\‡þ;ô0ìò,$p endstream endobj 815 0 obj << /Length 1315 /Filter /FlateDecode >> stream xÚVKoã6¾çWèVˆñ!JBoÛMší¡h³.rÈ#ÑŽP=¼zlêöÏw†Cie×ÅÌÑÇá¼9CÄðãA©”,WYP4W±Cû}@ÄÃWÜóEÀ­8ßm¯nî’$à1ËãœÛÝZÔ¶ žB®6ÏÛŸ™ ¦c Rzw·}ÜDRÊP2ø!ËÕívÑ“ñ? BÎo[¤3ŸT¤{8l"‘…¶«/$ª¿¹Sfy*3<£àx€Ô$§C ñœJæœÅ€D+¾ûE¾9¹›ˆ‡¤¡¿góñq™éðÕàÖ—5bM_ÙOM×H§áøjFBÇWKÐûû-!MµE1#á/à™—ÎÜÀÔ ~ß}wè+3Z—‹xÀ%ã y0ö¦v]ßÀ‘8«Ö¯ã@9èÑ~¥>á…íG3­uµáa1#’];`H¸ M[×3O§a7Õn+sn-{>¢ÇÅÆÔ¹{'9y¹ë»† Ð8 G§±µ„½½vD|Š¹jKô\*Ó`wSͱ9ãÂWû‡¯}V'’¯KÿwLà…‚R9Ë’å’T¨7¿l…æØ’PŠ5 Ú?Õcu¨-!eÕØv¨ ÈôÝíp…ØŸˆFSÓÖ’|ÌNšÓ=u•¢9ãZŸ±èšÃ4ZÔ¢çêÔI8¸D ͦ7/Î(؆Êsn”S1Ÿ3 vxI+Å‹s°FöôeM¿Ò×b3ÛD‰áo-ð‚ÉãÔBY×ÇM&Cg½‹\A¬Š íЕ†Ÿ',J, ü"G€L³@°_9ƒàË ~Åï'Ïö_‘UîÊ~OìGdê&‚ Óz‚bé%ÎÔf©{$qXíNœð>؃ég¿dí­{#~cçÚ·€‹•)X¢ò¹â ûDüReÊ”‰dᣂ™ÑœU4ÂŵÌ0`U æ§ÁZB-ÛCÊ”á=#äÞtÄäš"Œ€Ÿ=ǻޖ:C]_{òµaÇKg§™ý¥ß$PAøW R‡noo½€t- z½÷,½Üè9ó\SZý ¶Roø'1ϳVþL­!‚‘Çc}Ú î0MxsU")Í*Qp7àr–ÐÆUµƒÛ¥u}¨¡žˆ¯î,ÜDÀ¬M I•„¼Ps(Þçk‚¼¤Öž^¥3–è4H4NCåÆ/K`aÀ7‡O‘õ@žÏG+†ò¹œ8“ÉE*–Qeê%®üôHSò0¿%Tø`M] ýÁ÷¦gŸ­ƒÙ[âRâÜg‘eLæ*HÉ2!¿ñ䘹£ûÏ…:³Ù¬úó"ž'x …ÈQÐÓs”€Ã5‡ÓYð渚€+k|¼ú•”­_+h <†‡ÈüÊñu,‚ŒåZ+÷bqšâk%‹ýkåqϬ > stream xÚÅZYÛF~÷¯ò²T2¢ÙÍÛûä8và>63›ì"ŽÔ’S¤Â¦23ùõ©«yÈ{ d³2ê‹ÝÕu~Umµà?µÈƒE†~e‹õáQ@£ínÁï¿}¤dÝ ®F+¿¾zôøE/TàçA®WÛñVW›ÅOÞ³}qìL»\…aè…O–«(Š½7ÝÞ ½<—+yMÛõ_-³ÐkŽåÚ´ÎÓØSñòç«ï=¿êɉµ~ ݸòÂS <õµŠ#!<É"_….Wq¢½7Ë•ò„Ü4brC-äBÉâ®–¹#ˆä½#?Œ{×"åç©ÛÚW¼ùåËW¸ý7¼wQ•»ú``[å¶-ê 7Þ*ìpâÆ1á·þvaøp^i?Js¼ÞO?‹ Ì}·ü$Ì7´ò°ˆü,ÒЪ—þ50a 4 ?^á^*ajEU5@…òÖ|·©ÀWZñ7+•ûJ‹Øáfß\ÀG Ä›}¹^êÔÛs×ã6–ÛÛÙšEÞ;`ûeYï*Ã/kÛµ§uW65¼:U]yäéÈû¦èŠ‹/ '^){Ó•amkŠJœ$ô6uxͱ°–e#·œc‹<óÉv¼V\4OƒQu'V¨¤»Úm¬øPKæÐnd)QØÑ•oÈ>0¬’Õ5²àvë(Pf )ÌF\D„§‘¸`y³íLÍMa:´z">†Üœo¸E’éÚ%šö®w¤¼QÑoYtsC׊Ž¿ÜÐuÁ®„fgnÜ  &h#£ ¤~˜D°í€MfaI˜úyž»uÈ2Ó­}&BЃZ%±["/Z~Ka"š3_©(†#鵓:C@p꜊"ÀÐîT´eá)¸µÃÎPªÊ8@êÕ€»-7‰ *1A!Ä%ØÓŒ]`3q …Æ,ÀüÆ‘O:”—WŽ°!¯îzpß° òÈOÓp"§uSA´.ç Áä¦-Ž®¸‚#éÝÍÄÒ '\$&þ"™@ðš4uNŸ˜ª<ãŒ9))?Éâ™ý…K?BCäíO¿—*Z*Ù‡Â<=H¦nå,0ØÖfhDåÐÉØØ7Í Ä/‘®‘bqéãÓl‡5bÇ Ñ_²:ß{p`h>Í®lâvèxáÕ@\]¥àRή>Gêx“‡¬Òܲ ^‰rPø>¥ëJVBª{è!‘Ô€÷¦”¦#‚³å ª“Àw W›q¨¯&Ù>DŸWlWŽ. oì*Ìš+ÌT_ÆÚéjSÐáAÒÖÔd¤iì=Es yÆËT<Ý‚i…÷Š“Æku®P ±ë¶ìýU!aW ¶|µKùˆKÄ•ˆ¡Ë ô@´£8‰§Ù|âꉰ¿í.V¶E'ìÏ¥=?R i̼­©!¦©ÃŠiF03ÅYWHÝ {¥]Ÿ¬¥›bŸ i&Ù3µð€› d|€‰\CóGNÓ“Z'Ÿ„å~Óöàè$¾ÍV«™ ˆ.!kê ‡RI@ˆ.ˆÖ¸8û¹pŽõ-@8k3-sÝž,…Ö^õB?@ØÎ )ãtνk|Oõc|¹Y`ÿ$ô~åôö!J–ú‰{„¹"­£påi,•¬p÷ jònRME£¹v )íàñ9 ³à f àærÛÈ×j”‚p$ç7¬:w ËÇbÒÃM/=ϸI}!uàDkïëÂJA+‹¼ a¾HE“/—'#µ4Åx±ä #nSÇvæèØí¹Sl~EYÓ3Kom´3×ÛpÃI•ˆ€KºZsBµµòYï¯þ!D”õ°ÏŒÍrÔÆ 8«©˜%©^$å‡" ý4ÎYöéÔ'Eåæ‹Ñª²"Îì˜ùINvTùŽúsv|–òdÔKy.q§MÌÐ?Êr|Þ}ö~ÚÒÚÃ4»·\€~T} s¬ò‰6‡ò*æ^DZšgP¸p„¦v4Hx¢©³0¯©Ì´ë&Få=‰«EwVÖ•ò¯-ÉHç4ÿ¡Vg:‘ÇHE“ÿ±8ð#»§¯îñÿ£rûÿÙ£‚Åø Ö‘þ›ÿzšõ}lŽþV6ŽfèêøÜcƒ'gh®“͟ˈ‡œðcªrO8˜ÅfŸæ†ê +øoêþrTx endstream endobj 826 0 obj << /Length 3163 /Filter /FlateDecode >> stream xÚ•YYÜÆ~ׯø!áKšÝÍ3FÙ›(°7ˆ´€¬¡wÈÙ¡Å!Ç$G£u’ÿžºšÇ˜ŽlìÑwwuuÕWÕ&„µÉÃMjLGÙfw|Ro÷´áÊ›¿¾P2χ‰þlæ_î_|yÇy˜«Íý~¾Õ}±yðT²}wÿ÷Iu„ ìB½··÷ßo}cŒgøÁ)/¾¹ωµþáÌÏS”dQ LÄgßµC g‡Úv¸ÁjîíÛÎõÑ ñÚ®¨š­òl÷Ì»öxªËO<íëÛûî.ín«3ïÀýe]Ëf«SoàŽvµ¯í:‹ ž¹ÿˆõs?ðàãÖ‡fyµ·{>ÁP CÀ*¸a2¿a“È1w¿.ï‰ë¾ñ• TÌÃî¼Q¼Ü(s·ÍWü@UP[?ŠbÏrÇav¥­odühŸª¸ý »"ïd«Žé ²¦hÏu韺rWõUÛ¸ùÆ#~×G¬<ÊE}åu}àî„ùZiš0y¯ Ê$ÌX¬Øâã6†¢Ù•÷°”a ^ø:”ÝÞîJ 1WŒŽaKz"xÇšCo<_4t¶éA¡ä¨¡]’‡× å>jy!)•âkAI*HôA•6Qz±7t[.úóc_²h¿/ ‹Yõ@—¶Yä¡P›Ì¡.¨ûV\IqꋳPÊb‘²•«\KšÑ‰ã `Å'þyÇmË#Œp}¥Õ'à¹9‹”Àzi¨–¶æŸÐÙ¢Y͵÷ý¡l¸k¾â Çä"CP9Õ¶Á—1ï+pHe£“íû«™3gŒÚ$_ó æ-ókˆuc¡è܆½5©ðB¤Ò0ˆÍ()Ý6µùS:Žcr}M—A’Ѧ’ã:â3AyÐÿ¾a{Zí¹WäDæ,¬¬k¥Ø>úÚæ׆CW®F:Ì‚8Éw,È"ßßÍZÆûv53ž. qW}š³ @(KÓ³k?îî`MÆ2zÃÆuR(ûC{® —GÌá·Ê–/Ëö*e^–^sA¡×¥óÉó@ý™DŠ›íϦ¯ÜùzS±ºäç‚Ìxg13×͓d$¸×Y–#û½Î‡‘K5 [´$þ$ w,å·»’Xý+òÇ•ªhÁ™ß#nýgxu} ñê-*p+$§à 3š¦²ãüqÞ8ÿ[Ì´¡ø sRÅM ùN<% ˜f¿P  )SÀ]ä™úŒP¸Ùþlúš"\m: Åj<î»ìz n1ÁP~w­~.9,x‰`¡úâØ&ÇÊW¿!Jx †8 §xz âД æåТ½ÇjO` ¡Ë¢ÜÕ–4Z‰[¨(‰QÕ2ÝuYÔpùjÜcnÁ‡ŒGëckö:m‹Uœ5YD£ßëq>·HU,£k“43ÎFÜqrT"ï_¤Sg,pŽ®ˆ5£Ù1’¡I"—¤|ÏÃA$¼:g¯*ŠÁÇ5S~ˆ³IØS BÀä)ã<¥?­:8`l—÷¿éÎП˜+¤|蘽1“þÃ]øîáNÁŸ~÷ÕšãKËÐ]ä5X[ÁŠÃ9-ü3ãvï‹ýðÞˆMwá Þ)WÑRùÑ$Òï¯tâ±Ê¤A›%MhcÞßþãÍ÷/ß|-k©ë›·÷¯¿{yÿ *çWbWÖÖ»;ù ~òØWÝŸÙƒH‚–]Ççu͵¼91– SpY5>°¡[ç‹$"¢›ƒ%)ãš<¨(ÍèwÜ… ÿíN-8½ê¹„Ä¢s×ÝhLVŽ¤ƒ$F¨q‘tÑʶ©ýÀ4râ/X˜îN¦xTËD®K.«òsÙµ>Ù>”ÿEÂËD£Ò‰«…yc¿ä„nî=Õí#/Ì•DtWôœ ܉¥ ºªA*Œ´ 7 VÇË#øÐü¥*¶‚;Ý6…l]ÉÎÇ1)›ã§…vµ à ½2gí |5=tªAݪ§†«Œã©ž§°Iñ3Vøñ¡âO³hïíëï¾æG{Xû•`Ä$1Ðor¯8úݱˆ[îOëW<ô«C–‘üàÊÃD:±b²@ÔÕΚɱs¹Â/Ž|rÉŸ¦h47J6àûG†>Õ?¼ 7ŒxQº¹ÐÄãÂðzóöÅ?WCP·óLãN:4Ωe”̼Ýud<ÿ"4ÎÃ@î²Åg"c™ìO³×¨Zn9 ȱ}Ý€¢ÚB>¤\fÖ¶ >rŽEU,/gfW´QÅ4©xñ­Ÿn¿ª“ »y>†Â+ŸkŠ²ßuÕìs»øò¡¶f[8æý¿Ãšù endstream endobj 833 0 obj << /Length 2683 /Filter /FlateDecode >> stream xÚ­YÝsÛ¸Ï_áéËQ¶ÉÁKûæšNúp7ÓñLœLJ‹["U’Šíöîï~!Ó9_¦“‰,–‹ÅbñÛ](9‹á_rVÆg¹”Q™gËÝ«˜¨ÝúŒ;ÿüû«Äò…ÀN8ÿzõêõ{¥Î’8*ã29»ZME]Õg×Á»Mµt·¥”üq¦© ~6Žôa·_„¢Ún¨št¦_- ´{³ìaZ”¹ ’|ñùê¯þv5ª£„x¡ÞÈùDñ—Q š[ų"™²â2QºU&‚Ÿa4Õ" vf ú6o»®ZÈ$xìylnß¡’§F«%y”IÉ¢?âÞ4ì«,ƒN/ÛÝN75«í¶%{,«Á4k&îÚ~°Ó¸*Ìâª8®› T‚¯Qæ%ègÁ½6Vlª†ÈE$’ Ô!5V«áþËW\Î)‰¨,KÇÁÉá±lŒ]ÚôÍV«Uny´©ºåÁ9*XOÁæe”¨„!sšÈ n/Q¤€-›ö°Þ Yfài\ÇMk UÏm?t•“€D «¶»G‹à¢ô˜q‡äÃv0amÀÀ½i›jkeu -™“%ñ3'ž×v}éĉ mtHÛŠýýxÂÑsgüe”9þwvG[€é6äññ+¦³•ã 7»ýVãµ"8ôãwOÛó=·iæéÚ,¤y°5w(^Ïù…Y«ä»üâ)<¿Õöô‹¤úƪð–ˆrº[)ãàƬ™®ö죣àd?´fÒÏsÚƒpX§ô´‡;f{˜Sî€LFî¯ •òÝóbŸbÓ‘“–Ü€0¼ÿ{ð”I|?½¿Öä’9îÎFiÏ6±±ÖBF2/¬9xñ}Ç6«KD½$Ú·äØéÍHµ§s Tb/æ©Ñí'¸iº»ùQèÐä¬Ð%È=ÐU¹!]—'«9KÒ¬¥FkÁ˜@Ϫ4s5ÆSFSE ýïE´ü5Úõ¡‚`#4¸ä kOÉ8 e z€zyPk²sÈ¿cs(>ü,õPŠ(W…;ýOq,æ|„»Ž+s¢`£°Ú“$òE˜e9@OsʸøqιC!Ë( †x>ÁMçUó%=¾™=;ɼá†\}"èO‰§&€\ê<ç¨;­Ä :f»:•"ðÿ›9[À1E…¿‘·Ë¥îÕdá\»ðЇz«Á±);èÑCÊ$Ø´è÷ýú+ýíì y |´cØ@±Dm`¹{üõÛöÀôeE! »„³vµCßå%Fœ"l)Õ~6ì·†ƒFÊ—[Úvþ5w¶E£k\žuŸ$Jby~°Ò÷ö¦V' èN£GÔ7‰ƒe{ØÖÜ]9öOq’>è:ÄCdÊ4h…y¦‚M?èªÆ!¾X³!ï† çëdA¸reÉôòõZÇ dÝR è®Â€:-¨${=ðÈðY·Sb•d´ƒ¡ ‡žØ¢ûÖ†mx¨«nkãJ›*`pxîÃ]… _p ôé)Ÿ šH‰¡L2”)X€`L£‚õHu vzHÖ,Q;e£=ÒñÞXP‘Þy£Oäî¼Í€D°þ-"£íß͹ Ä–òèJxÃacõÚ®ºëÃD÷©>Óû¦ü,T×–åÆ&yð‡Mj÷ê|ƒÃhU¨Òw{æ@7DÄãû… ¾± >Ò!@æ2à…ˆâi–]ßñ·, á ‡çÜ oyæ‚)€âG†40h¥ù;—Á’ü˜øòÆ iîÃgÃ>ûŠb!ô–•íÜPÀÒ<‹ê¯ÆZÆ}»Ó÷›jp¬Ö«Xk·³a×Æ/÷&ì;Ãñ »jÙµƒxË õÀ¼nùw³ ^®¢\Ž ÞÅÅïc•¿tÙªMéò\ô$¸ DÀ¾·Ußs·Âj;gé¶mUÓ¦p‚:³À˜¤hëdâÏ1Ŧç”N§‰ËN.k‘ 8*»"\ã]Àñ»²äNhÍeŒöŒ _M­Ý®¸ájÃíÅ‚_¢ÒH”'è×i‹eºÇãÎZeP=.°®ä¢2“Þ5̸&¡–g?AmÊù*xÃÍV‡[ݬ‡ÍŸ˜é™r‡¾Ÿà&a¿·Ê´smö\ää¾(^Ùlîæ`‹úaãÒ¹¸ÕÁÏñ$×z8K¢íG÷p<`¢?ì÷t­ÚŽ¯ÐnÆBBRp1[ÝqéuZâ‡I¥œ€ÖãXç«ï¨ó%ÂîÇE™B¡ ËCLB¡²}üÖ[€}fx‹e¹cÄXá`ˆ; •1WLà€•gOŸ ˆý@þÉ»Žyˆ<´äv5P Ìs°ôz­{6rŽ¶FÞ§±Ñ4hyÐ…Ý*.ù ‚zïß_}t¹máãzýñòºp.áRâC诠ìbòrýʆŒóWhEÍ„ŠÇû…w7Íê°µ šDŒÏ#~ ÒÿÅ#kp+µé—D9Œ¥Ä‹¯äÉ3‡¦”sÓ˜y⌊̧DúLöŠ3¶ [ò6ñ3z‘ˆù°í[n—è¬ín×6x;r×pÓða: ªŽ°ò6b!Ñs®©üy@ @`ã¶çM0¦tZû’7"ÞÉô÷|V*<NÄ&P,K ¬U•Ù$ƒ'©Ø1É€åá6´ëƒ™fn— äÉÝCìjŸ$-´Œ­âüìˆã>AÛlY÷´2ÍÀ…•¹¼­áfj¾óóêËMU»Â/L P¯Ïg¹€»VæÇ$ª=>ÁÞ£¾ÕE]q½«î&[©¸Q‰xàlkr¤³a’YÈÝÜšç¯íud$I EY$c—D7òÍ"uÜù±XÔ©/¨P¯R}bkt%·¼ñŠßØcâæÏÜ(K½¸0GÅþK¢á¨3™ù•Äq³×æ³Ì æW‰¯Ì·vpjLr_jßáÜÒžngípëÙ!£!n]Êöû¸¾õŒ ÓÒaÎü¡ ×T¬ð©ýXoÓàøð#F÷¢Uý^ ‚äŠgR¶Šq&¡/ZnwÕƒÙÑ»ŽÈ÷\HCß[Ú Ì Iâñ5p+Þì~—\àd?˜ív&ôµÂ„C‘‚úšÒ.©\üU^U3怊²ÂKîû¤;r ÚüÑ&‰àäæMgjft‘?0iJ!¤¾=}dµÈü_2—lжÍìm²æÌúÐUAuä„Óå%¿¶pÉ„Ùë¶o™¯ÖPbSÉÚŒò+FòÃŽôS­‘.fS-%Ýí„žÏ`g[«Êä%Âþû?{I¿ö endstream endobj 738 0 obj << /Type /ObjStm /N 100 /First 889 /Length 2779 /Filter /FlateDecode >> stream xÚÅZûSÛHþÝÅü¶I]1ïçVj«HX“½  ØÍÞåRWŠ-À·¶ÅÙ"ûëïkI¿#±I<3RÏLO÷7Ý= “,Ç”SH%3R1o˜u†i™‰i£™ÓŸs©5Ì'rbQQY²d}/X4#éƒÕL)ÔN2EÕƒUL™Z‡ON¼ñLy|èI…àñ&0•$ÞØÈ´Ä÷à,Ó*éžGjŒÇ=Ú[V!ãÓŒç˜ö¨uˆøW:jŒÊ¡RôÈàщÆéÆé}Ï{f´µÈP12Ž+©h\Ã>0ãe )-2‰Yƒ… ™ ã ™Q€G'‰- Ì)‰"q2b@£Î@¾!B®ŽÛΓpÀ$MŸÐŽO”IÌE –s |Q .4)P)A-2RÃy­ñ ÷M„¥:/$ϼ§RÄ?F­Á¾Qö4¤„D‚7ÉQÝ*`BS$úHò0¯S$•:¼ $MƒŒ£Ñ 1 Ö!œ(%†H£¬jÓyDeYÔx4‹ÆãŠMDjÁch‚ŽàÐQ 2Ž 2ÖCLÖ¢a¥íAØIjÊxd¨?YR€Rl`ÁÉjzXr8@­€RPS$v!yTR²1-”T£n¥n4:B®•¥\¤¶(g‰C`\Ig CÊxÔ×ÐrÇDï„ 5Éõ^¼è‰³¯W9ûÓiQöÄéõDz*¿Mÿ쉗Ål˜ÏØ{‰é)?ôÄI>(Ù{ …+‚„RÜcº8M3Ss益)eõsöÓO=ü·c+þ)r piå¸ÃDr2q+Ý÷äJÇ€¨´å4ý­U\“±1Ž‡°•³Ó_Ž¨çl<º˜NòiY¦ÃŠÑóòó¿'Ùx\ –ù:`ï”/Ù üãŸÀ Ùôz<þБÆÆ™ÆUº~1-+¾û0x±¦ïä„& Ìú´(ŽjASMÉ›¦7¦d] K^×@âxV Nsȉãƒ>gù—’}XUÃqv‘÷Ä+°Ía ,U'ÁÏ‹ëÙ ŸW. zu”GÙËâK­O0iÿ8›¡.#ÛVÑUš£Wò*Ä yJ?´G€˜&0˜ ÓÆabZç8™`-%·Qm…À‡'ùyyGšÀ=^Ï-¬“IŠ“Á†fMÜ>;Þ]fåò„8ɹ¯”{UL®®Ë|¾?¸ã{¡µ2+XÁ¼?fËšÚbéMl¤¶Ø°5cEÔñMÙ4åФ±yo»bÈzÉœƒ·0³ ™\ï4,Iܪ>RÒ8ÿBúš^O>šu¶­^GNþUkÏ"—žxPö›toƒåŠb {k‡‰À}‹ÙƒÁ#já†Ü©†1k¿UßÚp…¸ÍX85þ˸Ýu¼C#sê¾8§ßw£ù°˜ìfŸ?w<Φua¤áA#¢U°å*eh˜4·cáèz\Žö†#82r©Ù˜ø؟Ͳʚô‹Ù$+;ËHØX8UÂ&…éVÂÀ©'á‡l›“»Í5§$—J=Ú.;Îe‡ºl—jw·éâ†it¾­iĸ:š:åSåœÚr‰XØ8à BZ“`~ü¤«”vPà*ñ˜êˆ©ÀíAP[Ýx½¡/wÒMðfC7©³nZÆrg•…„°JË;bÿ³Ëœó:Ÿj§°w^M1z{Ð?[WÔÂçߧ¨5XbN+»¿ ÚXRçÃÄ×áõB{vxü†^ór>˜®J– žçPë,ÏJ˜œƒ¬ÌÙ³ƒµÄºVi‡¥˜¶nOº¤ütGÅð!’³Q93ZL‡ü|t±è¡(êaþ‰ý^¯cL6;ÎÊÁå8ÿ”™>_ÁÏ_ÊÃÓöÄÛã#¦__fó¼¥8ì¿éÿúÛß^çãOy9dµˆòzˆè“ö!*ÀÕ-ÒWÐŒfóòÕe6cF÷Ä›¬)`P=ñn4,/ ˜:ê¿S•cê§Úq ø‰~‰’¬47¹öW×êúW·|Ó¼Í/øµ®h)ày“«[Ú@¡ µ"iVý<´­ön*áÝLC’â5ì¿>8Ùÿû­ö^ãá¦.‚Ú]*Ü¡‹Õ¿j O"ÓÇɽþ%,,¸¸M=ìâ‘7¸AiU¶ú>ÙÞŠèV€µ@Í&ù]ÀYY^¶uîi8cÚk •w곋9³0†óAeˆIÚ¯²«×ùèâ²)ROôªöÄ/%Vúƒý馴›e>ùS_ŽFó9†Pi‡†Œ† 22AÏD& Á|6šÿ)>ŠŠ\äÿ½ÎÆâ\œÃWˆ 1c1“Ñôz.¦¢Å4WâŠüÃ8?/ëÜŒ¸WùŒ¢tT ÅÕUæãl~)æWÙ ¥(/g9ÒÏ…¸×Sø†ù ˜åâ³ø"¾Šÿå³ây-˜þhœȼö>«F¦ÑHçý|Ó}Äb·ŸFç'HZWáî»gŽ4j“µÒh.è´Mi:­(Š;(-ú5aš¯êl à'”u^ëçF;W7¢¾^¥ý«E¹u„°u÷¨¨cÉ[ßu_`Ù9/‡¾ô•mò -‡´( v”O…ÅÍ¥¨ÛFaQu E•"O10ÚØI´•¹¿cý…PàÇ&JX¶¢y8ØZ£A/\!o¿¬!%üñöãh(Tç— oØN JrSA±-`cÚA*©¨ïZÎݽ¼E~kÉlB7µ‡n³¯•\“6û`Éwßßrœ³¼7œ|Gµ¿,@½}Á÷všï¬.᱘XÚù ýÊêmVfÝ·¾œät f4–¤Xm{¯¸3ß3kèHi¸³õž ö‘[wϾǷí:€rœŽïh‹È<5c4ÓF¬†p­ôœÖ;-0—}ØN~kÙÕµ4QnšÈÛ‚†¯Ž>L)¤fÆyNGæJk^‡îOq‘ÚXè;ŒïÚÖJl²ƒ¸ª´±/'úëÙœªõ)o õº±{G'Ü»è5*µ¦WÚÞ'^èô¼£~I©ZºÁÁéF'ë-)z«‚ói>ËÊf›ú+†RÍþÏÕ,‚ÙîÛèÖJîà\Œ¤«%°áÒtg`£ûÙ—u«ê‡A¹ŒÃe¸-Ùx.£ðŽV˜ºÓ 7 Ù…Êm P™¶(TõT¡ëÊÈyÑÕa}Ü=Å¡g»ÃÀNÒ …á-]´AàóDg°qG½;mÛŠ5ÜŸµE–›è‰Bϲ ßÛÐÀpq:à}ˆN‡Xµ?H‡%›N©K|±¢ž]‚]Ì­uè7t¨m[  kîé–T6AÇ&5MÚÙOÑY†ïV&X…ÆO!̇¯Ú~²)Øø…Å´\Üz“_dƒÅ!b9˦…Iž®/UÇšcµ2p£¾7cÖWÿ7ÜU"»‡³“¼ºMUó–-Zo`»ÛT ×tIŒ‘á€]åðò߃e,]c ëE_IFI„Æ=ÕE¯Vj®ý¨$!'wßµŸvWÏè~ãƒötÆ!@tëñÏ:C·§l—-ÃÓø¤Ç³Ñ˜ £gT[ǵÃݼ64˜¢Üè. Ùe·ÑêZÖÿJ nŽ endstream endobj 838 0 obj << /Length 2854 /Filter /FlateDecode >> stream xÚ­Y[Û¸~ϯ0ú$cUuÍ>MƒI7v±MœN$82m«+K®(e2A|ϲìÑ$Ù `ÄËáá!Ïí;t¸à/\Á"SÊ/â|Qž4ÚíÜxý·g¡Ð­€p5¡üëúÙ_^&É" ü"(ÂÅz;eµÞ,Þya¾ü°þûBe‘Ÿ)p¡Ñ—/×·Ë•RÊS>ü!ɳ›õ¸OEß)R~[¢4ýPż÷±Ö¥±¸{àU û½áÆ]Ý.WQî•Kø÷»¿\űòÖ{Ýó´m;iµ[YÙUL+ …ñq18Õ6›‹]lÙ™à‚Ùޭ߈ÃÁ4È£×}Õ6?ádèmÛŽ©mgèzA-¡òÄOX5@q |Éî{XF™7Ô}µÚTÀØ®yJwÆ,÷ùfbïÅ÷­1ÜÀCÀ–pËéô–£¤ðÓ4s .ÛÃѯu³óËKåÆaê«@-Ò Vç1)ØO`Î øF¤éó‘©ÊÝúÕ„ÁŒÚ/·a»[†aè]ãAÿñH.•ûI–œÉõ´á9êoIqÁ”¤ðùa8Œ}§R®T,£ÜW‰¨c•¤‘w»,@ÝÕ“%l1ð½­ì¦=ü÷^ªÈûT5;þ ®ÞŠšÎäAC)àF‹©ÆAêU‡c½ Ñ CÏٜũÄÓLq0ýž­rÃ[ GKĶ%ãa°$hyßòЦ²¿sK7žêŒíÛn\¦u—KÓÐ{ÕȺìÁã(öîÑ »„’T9Ÿ!߆#W‚arjíLc:2øíX7Üú÷`{nMO}yÕ­ÞŒ£t63Q„÷ÁP¦•=ð Ôum6¢âRx’rOÊœS›Jü<ÌsáÞ¡ò~6¹bùî—L¸gcË®ºãK’º¸­­Ñý@÷‘dE5ÔüÙW»=·jó‰Ö²Óc.½Ee™ŸÙ")2?,Š?ìÅnýjÂ`Æ.·¡ó¿Øëco:–5æÏ;±h¾6[¸™¦$[þFzEf÷![Nïü¨wØ¢ðòŒa!$ ¤à›©ÈQ¯&ä3'ºdJRH•À“F _+á¸ÔvàP{wC/áœ-§!ÜÖ¦7LžVÖžÂÿ–¿à¾ÿï%Ž·ü¬¬ë›DEÎåóí)9„X‰ïq˜YE“Ê, ¬œ:÷U¿çÖÙ" 2?*Â)øøËÍõ›·¯oæÌ?Íý,ÁÃÕÇÄϲøŒßo×ëW7¿®¿‡ß*J(”ÍIšûAtÆ÷æ_?_¿}³~õÏYQóÜ£ÑSÛ#&a¹¢c'ás(Üi’úEšœÅ³Æèn…ë¤ï<Ã]±‘{‚:Êî`߃PìœÁ™¸3ÿ*²˜ÑYÀi|ÊQƒê‹´µqŽÏÅSè÷CÏÉå"/ö Z‹ZœØ±m{Ýõ+>u䬠9WÁHÕu£?f…÷íñÖÂSS¶'.{ÆÁ¸ ©úJר§/¨‡:'r·WœÇÙFq)lŸ–°;ÝÊÕœÆäF_)COFЫ(ioÆ{æÀyâ«ü<ÒÜ€÷þr½žõß,öã$ždÚXe”åç0o˜n„¼_ÉÞwRåè.‘B(P¾è¦€OéÀoÿÈGa~gzn°½ÀÌÔÊaœìinAóœù¢ÌOòdæZÛ®ÞØçxÊÐãbe 0±,îhþìZvM•=nµ‘p{KÂZBx‰ T¯äÐäÊ0WkNê0 Fˆkº„°üxbF4²‘à±4Š=«»‡e®8´ALyùc—cÒm¸ ²oýÀ={nƒ¹Ã‰ÐåF†'9sà#VôüuwV¤^ÓösÆ/à”Š8ã“$|»XûÉùSþ4*;œ˜±¢qeÁ] <8Æð ‰p¶q¬t %(ÅvÔŽåA¾6áp¿çxÏs•e.ÓtæXvx„iYI¥Á„!I ·XBHí»¡ìO5'Í lT”'Þõ–­EF€ç‚¢ÐøÈŽ•Ó!]h ·“j;~\0;I¸f!Ðź|l`ë;ÅY±ðÜДÈôùlôÈýü\“UçM¢ÄÛnûûæó¢ìG¾ì}ûq[Õ¦üð>Hº &.÷ºãÖŸ'áOs‘hõxSä&)q"1÷:‚Ð=B‚®‘»jÚfõÅtíùË‚JÆÛÖú¸ûœ¡’ÛE*ŸüÜ <¤„c£Ò°ƒHlžŒv>T;8"O-Ä\M<9é1e7G¢{‡*86^*EE Ù|ŒÕ¤('Šä&tSKØv0íT0›[KŽéï}â; êw]j‰;@&rãáÙGrã!ß8ƒ$¼ð+åH¤(C4Ž— 2Îø¡5X,méX,ᙧŽÆ,ÐL\7Ñ;±-_ÉDàHÆÿs?¿å³»üÿüíXçdò\ø˜P/Ž"‚vE˜Rã“…}…Œ€ ´¤L˼S„‡Ž6•Jôȼ3õfœÈ¦›Ⱦ*5 \Ò_.·™9¤ŽTV˜²²1>?E½ÀºN¤ÁÄB#®\̼÷¡Šqoˆx†QÞŸÜ&l0²o­0˜ºÌܹFÉNd*Fd‘â#w#”Tž»BóYcé:ü¿4>Õ—Ø¡úr¶¬T~©ï-+ñMºk5Îi> stream xÚÙnÛFðÝ_¡·R€Esy3}JÛ¸H¦@£ i`ÐäJZ„‡°KÙQ¿¾sì’”L·…svv8÷E‰UbU«,Šü"ÎWU{VïW üùë°t ÜÌ(ÚÞÜÝ'ÉJ~bµÝÍYmëÕïçCy¤^o¢(ò¢7ëM'ÞÃÁ¡Þ·Çõ&̽^e·``üvG^T•ë°ÈOë¯ÛßnÞmGu’0üŸz#å+žæ±/¢˜ß‚Šr½I‚À+pTîlç¡Ç§ðNFu{F=+S÷í-âÐS¥;”CÊ>«Ò^ur_–ùýýöó†Y–ªQkÎ3ï<ÚÔàÑ.7"òEŠWèÅCÙíI«8pÐõ3bÉÄ”]Í@Ï‘@°ê»Z ªïŒ¿ÞdAîݯóâ·ð¢ð¾ƒJe‹Ð †p·ÒLCºgÔ¸?5–ÿ#D/ó¬TŽ´Ô;Y ÍyŽëQ%»²J°Yòû˜((Ñùaáí4CàFu¿×V[<ëS×±?€¦ïÙw’@M¼+i ŒŒÈE@¥Ú+ÁÊ=;´ò,xÁˆI !X ÁpQâJˆDšåÞ/=Ƚz¦GŸGwè1¨äqù´&×Ú›y€i+K+‹s‘!É€‘ùQú^_©ÎšSþ0G,Ï“‘ÄÞ±);øþ8¨Vý--É®×LB’óW bml_zËHÕ™A–5¿×ï9¾OñÂey(+yt+Sù£þ…/‚ xOvÇž98¿Ä³ìÄ‹rg! )<µ$3“ àÒð³é9’Æú4ÆŠƒxÕ.:¤b/G£—YÕ•ZIãD¶%Äú¥ÿOÝXײö1×bðg¼­†´¤3Ö*91tþ²áŠ¡¶DNg>L\ùŒ‰.±¤ÖAÔ†(¤»—+eÇÏQ¬)[ ‘mç¥BÀܱ†jÉ]_U¨(¢æÝ , “Ü«O’ï02ø¬äR º«Ðs!W#“,í“Òà +ÑòmeÛƒ2³P è»Vr[Qr¨$Ì«ÒçP}2Ôä0šfÉ#µêO»Y$ÔÀP òÓ8HI1¼;²/5ÔE[‚Þ|ëÒ)¸Æ´¶ÌÊGî;§ÏÃòYéº%ΙHv ”»ù °þSJvÒǾШo=§'¶Û…,SËÂŲ X¢ýöäªab]àô#C‹Ð|AS+·ÆðM î㈑3å8è‹©åÐÁ¶1pŽíj_NU^póDÖÃamÕ 9ñ‚Ô–œåªÃ„Š‚Ì&IQŒí †éTp`×["szÜLµ4ÝV%õ$¡"¦Äd=qÔspjÉDzQ{N9pÚÚF ™(« N Þ\4b@ÕÅÅ€O½¶ã¹Ò’¶‰(,¦Pgûžj-îð<Ò<‡ÄùÏ8î­ÀÆC{Tí:.F|Ó¢†Ìñ`¥žºZêæŒÖ¿L®«5Ò”;"¦ß ã,¯§Þdaý>åµ#¼á΋ï È#yRa%ÎÑIp¡9ž³vº´>„ÑÔCâ0ãé„ë5Œ–{úÂÑL˽!°ÙôÝ-(qùâÈ÷¤j‹(ù±ƒÕÀf<2êYÞåÖ £øŠvåa,E˜•º^(\s†iÖnžI2¦o¤v{;¯#Kù%‚ÌŠØ%̌ػÝnx¾³*-¤X‘øaŽ)ÖqE}êÔw Ñ›%)‘E~ž¦—[ŸêðS >€SÀlÏ‹G­ŸzU#ã—t3ýü²H±D@¥rª1؅ӜÓAx°»`âk~{ˆ÷W8Ï¿Nð ꕢ̃ԹÕß¼î¯4óƒpôš*TZÙ¬«YÅGû5‰°¬ŠÜƺ×9¸)Ú-‰…cp垧lO/òðåÚA}ºHâ 6KSœðÏÅÐ’µa>™ pCÉÝçv”ç<ý=TW> stream xÚ3T0BCKscc=K …ä\.°hQº„äÎeUgbl¤±*OS endstream endobj 856 0 obj << /Length 3110 /Filter /FlateDecode >> stream xÚ­Yëܶÿî¿âÐ|ˆ6±hñ¥‡íH:uQ¤sEP8†«Óro•h¥­¤õùú×w†Cê±ÇóÙEqÀ-IQÃá<~ó¿Hà_ÉE&%+T~Q%vµ¿¾ Á›q·/†ñbçŸ.=y¥õOX‘üâr·$u¹½x½Ü—ÇÑô›XJ©§›X)½zuù ­¼1;Ó›¶2›Xèž0O¥y=’Š"~2"HjÀͶØ`ªŒ™Ö¸?Ø°?⤢'h¶†]³ ¢¿é ÃÅ™?Ä9ûÆg–:‘ߪo^[ßÉ5[kÞ*¢›=D‡ uä `>8@°`ŠO;Ê)JÏÒ)‘Nª¿½Ìd‡e˜„åT|AW\|g¯¼`=x»W,M³µÖ§ã¡44Ü.2›ÎÕ”#Ý— Ìhë•øl©EPzºKÆòl2’o¿ Æà:›bêBM, MEŠ(ÒÈŸ7A÷Õ)\t’”“ûóË T¤ $¾“ÌF{­ÛàYUSZ?Ú^"–.@^h·Y±à`’eÐ R{&ør\…áK±,ÅWcxUc.y!׊}c¦˜i¶˜(ʦØ)ÂÉ5ã‰xP%Å×ýmmhžbÞ¡GchÉ*MÚ@åræ–VÊëžJ6N6춹*<Â"ŪÀ.ºl@xö1Xw}yíâ;O–¤ú¬ô‡šžŠ,Q{K#gÆzª÷ôÝŒYÏøl'w¢°]ö©½Gh—ÉŒHB6m1 …¬>5â6í*ûÞ¦^·a$SLˆIΗ:óÉRñÉ ,Oäï°‡RÃ9ÐU];˜ê´2‰ l7d TÙCÙ5„ØtY@“/!äõغ’¥É+˜”`'U… ?›‰y’1ØùyG ý"Ÿ–`¯¢¶«Æ4<ƒUø@‹íµõʲÏæߨx¦¦õ1ä¬ãx|Š5sôÄÿW2º¹DQkÛƒQw4m<Œ[»Á-õ×n;e­…ã`iÕ¿¤Èà @“§ÑP ±˜þòƒX>ÝvÕ°˜Ë£-íÜB‰$ÿ‹•–Ë<‡9G–`~ܺVÂåí¾;"‡Ì• g\‰™~w¼Ø赞µR`T)hHò¡ÜÙw§ktÅR¨ÎVè…5Z²Õ@š,ªÂ¼”k šV¶õ'-"Ì,‚Á¯%•p0ºéklØÐ9 €åcß9G³õïƒØËÅ,šKu9Qºæ4R)“xðL¸¯Ì[Í€Ô–åÀ*‘Ã.gƒ…«SßÏu&üJkŠ¿Y4„yqÈ—¢¥í˜Òª¥üÕ`a!Xnê«ž2˜P…Œ{ö>.Û)az¶@U|hë‰þwûC¨®¹´@l+ù¨óqœïÁµ v‹• Œèžõ']P?<§ò—/Ë_¾(³ÈRÒχFŒ®Ë¹¥tVH¦òü<àÞÇ —å3=ŒWEQ,“Zž¯¹ªm(8ö!7—ÃKÈT~–Îmk±\]Ÿ¹ú¡všx’áØsÓ›ºMìÍø¾*‡ñù2oýæâïÇû’u^@Ò"¦pî2Jì‡IåÚŒs—QP—ñ'¼¹íÏöÐn(Î=ã4s¥:”´u ¶jz0Á¸Ü0&;…AÓµ×1è ºò—­ÁÔ·EúÁyDê<~Q®i¸,ßa Q©)«p}R ÛÙÃýÊØï¼[¦›-,_MWŽÁ&(9Oª0 ŸÓ&‘ûr/¯)ó…5‹58¸”ªZ´¢b¡¨æŒ¹>8}¸)Qš«/ØîýxA PÎŽYÁCþxB¿o_“•„ǸdMG(HŸOêÿØgï\sAÏÁ ÈGÌR­‹ìüƲ7™*( à=©˜~w¼Ø¸ß9Ñ©³æ¾}L¸ ,Q7KZ€|Y·ì¾Û®¶œÓµƒ‘ú²y›AÅY¬Û·N³ B€ïë þFß%êŒPh4´@ûÄ»(ù'l°Ÿ~0èÔfx† (" –‰ûštÍò¹•ã>ÖìBü)lŽLüAx &àV¢Ï¨5SC@‚2Ž˜õ¹:¬2–œ£zm;Ú~‹r‰–›Ýƒ×ùµ^~ð¦BÌ‘y¨ŽS™Xnf4Åê—Ÿ©ÿ„dv¿VUðh[ƒ¶¥ñF¼„PØG9Bm¿v`}ŒøŽJ´_N äþK çš)-WþÓja³ùÂgöçl*†·3ñ%æ÷zÙ¨Ã_átÆÄÜpʃ¹ç+¨’âóÚ’zÙò¾û••…ïɆVü¸äb-9}àÂACE7óî¹™m…ºÂ))þþ+üÁ#¯Z¦ Ör9²) |M)O½—¶ÌY>_Iï‚´¥Z»£Ií½”¬æø‹„›Ï'¼ó”}6¡ÁefçM %[ƒÿ ½”Jçì»ë™º'JÒ‡ê,§Pn$%,ŸCùg4§õ¬)—‡ÃþðGUQœ ÿSÄ!ÕjÕg¿ÓL`"Ñw k@βô¬Ä5±IÌ4”ãÏ>û’ɵ}îÞ‚øù=¬Oš±2 Ë }0~ÿ)Ýq endstream endobj 863 0 obj << /Length 2886 /Filter /FlateDecode >> stream xÚ•Yësܶÿî¿â¾…rD˜>ÒO©]w’™tÚX™LÇñ¨§cË#O$ϲò×w»|q¶<šÑKPêêÃÍÏ*‘D ìâ¨oßÞü~j­-ৼøÛÍtŽQê™ áÌÏ8JâG¼4›$‹…Ô1ÿG$S<ôœ{ \…¹üíx¼ UØ®,zKü6ÅÁö×8ŽƒJXqÆqLTšpGK©š{¢=VÞ–ü‡OM–§*™ é¥rn}Ü©L$zÒáw×´ygEÕÐöÞÙì!A{¬µª$&7¡LD ûƒ>~µÇº(­O„8‚Ͳqâ¶=ÝÕÖ·¡NEžçã<§ŸBb‘ÅÓ¬]݃w3%TªÆimçÛJ‚yxNÝŽWr™Å$&ždÞÁ¶¤ÑÓ]מ†ªa ‹t<Ø®wzÅj#t–â@äIBËßX6¡fë¬A%:8©mèñé N=•íáXÕ–ŸŠfûªåñ¾è¶8·èìµWk:Qœ~“¬Z¤é¤èCÛ?ÑqM;à v–Šf1åÐv–FÇΖUoYy*2•€…i!¤½‡=,õ1mD<[Øs çÈDãÅ ×)²äܹ^¸áxwìÉm7Ø-ÑŠ\¦¨kv¤¢öíé~OOoTõ4¯à2à ˜ÚÐïë<^$›¨36àZ*£#N¿Ê´cÃDp@ßÒhÉQN=E ’ Â4Ó©>ÓÎÁ;Õ¤»Û[ç(ReÞ›–"ʧóp*¶ÝéXÛn®Bctg»ë=NÊÒ±k‘ø±Ú:Õ_wdtÙLHfëˆ=­3.V*øxå¢(Ó¥Ýù$ÊrÑÓ%ß—¥7 J‘f“$È8n¨UHÐdÄá]Å |Ê,ˆ(hJ½µç)HgZˆh&Ç0)]ÞáŸT.­)ËÄ4®xÒåù1N¦×ûâ8`@þdD¿ïBc¨ëb¨HféDeãÖ—ðúÔí¡úÓ-ø6n F­{ÛDy§ç"«(F¥<%_É»ãÜpšìn½áè¸à,ažªàæ*‹ð#Ñú™«aþæFKKdîëªù½½”9¤ÉEžÉÑ:Âz·õ9kX£ÝŠ1‡õÁ›ôNDy¶Œ5è’žÓÂX¬!ÅsÜ­î¦c|Ž˜•LI&wE‡ñ4ÏÏ=-Šžv~t‹X§_s•HÄñ$jiÀÚCl¯dÀGv™þàÜ ]kîÛ¦=¸…CUI/w§¦ Ée&®:?^?ÛfK—p‚=zÜcµ!¥Ù’éVñN¼ 2 N¦ØðÇê-£ `N<&ìWoSƒè ‚µî.e”‰Xaô6BÆš¶Ìð44ã~±ï0ͤ&ø±®1’Ë $wô%¶L C¦Ùt~l+Ù(¼D nàÍméâOõ§½è]ƒ:ú‹Sd´²«y¹[½ë¬Åµ‹]ÓÒs½„ÄËK|7{‹N§RM±”»Ã‚4Ò!gô‚³éÞŠæ4eÖm Ò<ñºÖr3¡ôl ôR½žM›ƒÜg»Aì]@ÄÎzQƒR8g]£&°ŸJ{dÁFQXûD´û@;’Ê’arÍE¤ãÏ` ®7Š²6dCH~à £A'FÂѨ ü¸¸Žó[ÖéSOÏ3²‚%=­ï+j¢uuߦ¬Š+:ÛƒGN”êÀ¨¢Ͼ›6=;öEWî Ž<Ææð”c¡&)OZqNÄw?ýò†FEYZ®Î!8ÌãÌwžÜ¡fÀJ€¾=µòòp^ïˬëCœTïlIÙùaI²÷ ÖZv$¨œ¤vIéq”çaâ)Î"‘À¹x›ÀÒûÑf /A³@ù£›yØh(Y0@Õ›w/þåã;†:5…¤¹äÛùŠ»ÙòÃ5Y£ÎiÚÊ)»cÌ“æ\éR%®Šc X!ÿJ~g‡‹éfÏ7u\0Ö¼Ã2è¼)†-; X¢4(ÑÅÀ‰ÉìåÄ—!PRËTp» %ÙY²¢)”y£DéFœ*äckgÕ®v@æ¿ÀÖ€ÙŠ»K± î0Í2̹½©î"¸²Ç\àmÛA‰€Á‡F´ÌÄ+G~^dKX"úò„N£€ëßžžzˆ6õ1øÓhØw\CáCK¿s‰‹O-7 ®æÁÍžúÁ¾ë}ré«Ôì‰#z®Ûëº4ªÑ@UÒ÷¨<(ÉÆwCÏÖ+!Ó3ráÀa1(}íéúxeÐtçô)0úà¨L„‘3ëöà¶ô1‘É%蔎E¸šÎa%ÒCÓvJ¸ku;2Iõ/ ÑzZÌC³`éN† wÈÏc:À‡¸rð¨äÌÇH.Q‹×øv¨ýÅ‘Áíá7l…]!"á±\îím±…$$2ø·Ãú'šŽpÃã«2ð¦&‡p’ º+ƒ±þÁÌ[/¸vg«FóC¦\u'ƒˆD¯¦0DÞM¯fcÅ [ˆX×ÈäRÔÎpºQD«µÝ@^n›ûúilЕí’ ÞjcÂ]úÙ?ÚÁþ€)SOÉ÷µÏ¤L$²Óÿ½·‚‘®1Ã-“à¿` ]2+z:‚;“0úRd‘PÈiõ \Áo>w¶[8å€Ì…\×xe1‘]y: Çý™‚—Éåt°†Á: ɵ=·Ü¤(g?º«å΃©Š± SÏ[$ÚµðRQ3š¦~Òl[è¼·>¶‡Ä"jáhQã?i¸)‚j'¦\(“»¡l›¦7•mÊqñÞçNp:wåHê¬Ãø’òRýÄÔ®£^a¿ÂŽò+m]HŽrÊÆÞ(˜‹l64€“#s-÷,…K‚zDu &a]Ãà,8§.8Ý_Š`¨IæÎ1Fpî<º‹{9–©qà)ÔÚÏ‚%óÜ—Ü‚tŸõI¨ 2uV_ªo¤PrÅ¡ceº…çò9¯B„õéùÜ"~Îç/6Ódž~tס" «Ÿ~ð¢lJi™­º¡£–ç"ر|‹ööŒBX.¼|ÒÄù^üâòvÏ+ŽA^XåMúX‚Ï®„.bÃo †Ï@£d¯©…Â3Ϝ͑fg¢M¼‘»êò+¾£Réó5T²âèØÌ‹#p¹§ë y à€KgsÄÇ[Ø]éE“2†X/»Sq:iÜ,5Žt¿'CÝ•®™¾ùücçw¸,@–žÁ™j:¯Mï Ò,Sî¢Ã)1Nù¿äϹHãxl0)j/ý†}0ÆÃ&øg]4½ït…*bþ¼Jó`IÆ%yšŒÅAs¼ò—‰‹9£€5\ç§ÙR31mç`%{š6¸š„ñ»¿ë…eéº:Ö…×^ÉÜ Å»(:®='KG~–„®ƒ·æç͘¥€Œ Mñ'$å¦? J°NWÍáÃØ#ÁqÛð pBøM(ó"l²«ÚSO„#\HF_y‘0—Gø´µ}ÙU\LòI¨íÕ¦s:}ŽÌ!1AO‹à´‰£àGoÇ1S"MÒçªÞ¨ '3`ŠÊÙRbØ„ h?üÍ PD¹žRŽäübà=¹ –pÅê£rãBB•¦òä¬ÛTÖ§íô¹|úŠ}~÷ü)nDlÓ¼ª!˜ˆŸóþB7aGÉÆ+.·j—ñÿ¬×j?Ù$C-ñ›Üâ•Ó°{ÄÑ”c°óòþ~m& endstream endobj 870 0 obj << /Length 3084 /Filter /FlateDecode >> stream xÚَܸñÝ_1ÈËj6ÓZê ûÉ <‹]ÀI`wë…!«ÙÝÂêèèðxþ>uQר=öÂÀ4Y,V‘źeïFÁ?ï&U7q¸i˜ÜäÕ EÐötÃw?¿ðoˆ»æßö/~º×úÆSnªRïfœ“Ún~sþ~Î.½iowA8áËÛ]jçþ~ÿ†¼3GÓš:7·;_'IäøÁíïû__¼ÙLµïãéóùãEIèzAÈÇÛŸ‹NâùŽùbò¡72ëÏÏÛ£ùö$rƒ$n´ùRfµ`-˜ø>-ÖRŒœ¾aÊyS]€ÏÄž“ßî<§io=§Å#xÎ;M}(ê“à¶YÝ›¶âiSóN!á;YÛf·~â<Ê•ç"Ç…3O‹Ï|³óTêú©†AàzšÏú€xcº^m8‰œJ+&!U8ºžg]_”%o4_Š®‡ ž Ï„ d3QäIvÕMÏ€ ļÅå…æ€BÓÊÉêÃÖ3~BO(>û‡ò~[ïû®V‰ÅÎí>‘ˆ K#TÊB_¤áYÔ²$€¦’T‚Щ`;J™¡}QY`„À9˜®hÍÁ%º`/^ì&a,êv›*‚ÒNv¹”2ÄŸÈ9Ÿ‘´©*ÒR¢; )@»Aô^Öìó#‹;ÖG4ïâ{æÐ!{}“qmv Y‰Ò¦‚y™ö˜åß5ÐÎ{cÖÆŠöåGnh²UûrÑ\Ï'£]BÖ¾e¶÷Šñ Ió½Éû‚ì}‹ñ;üöÕU`ñÍx@¸¿jÚÛý~Ç—ì$ËA²¾Th7‰ÓÅÍ®{!‹ýÌeÖDéFî–¶ïüP(ÎUéY¥÷Ü@VéÑ|ј€<ùÿÐ49Y)*Ô¡µò"Z&þ’_ÁØsPSqr¹ .ý*ñç9 Y›‰ªa‡Ú‹¿ÖÞ„|k²Ã®ËŽæ[ xa€ÝS óTìFAt£SJ·šÙý»çY³yªpÚ Eáöt†â…zÒ¼§š"^]Æp—¹®Yû™£¯‰Nš·O„ÄÂ+q„³Úç}¹¥‡àtÙ(ÁýsSÀµµ¯YÁóõmóø*Ù‹!Ž€½W[꺛Q—c ¸_ yÆ‘=îvdŽC7PÁ³‘9ru¨-ùvŠ\ðBEßYoß çX Ú!ë3+G(†ÅÑ ~¡²Á³åIG¾ˆ»¬So*ón%¸øMÛA EëŒÑÁãr• A’Ý`:.dB’,–y>pp.ùð|(ºCS ‡UŠN3kÃðƒ­Ócå4G†då©i‹þ\u —ëhˆ:E™}* ƒÑ :§$0ÈÎhì5)oj ñ§¡ÍPÓ ”hðD¿§)bÁ„³‘1ÈÅìJÈJ5ì)ÑQñN¾qÁþœ!¡(•§‡¤[FN&Y)‰ þÒE`piA&E-³®ž “)M‰LX / å™®ÅÛ[Ö)^çZ˧9µY…ž#NÆ]ó›ÌsPIÌPþœ±æ¥Éêá‚vŠ+ßc˜¯œÑ§ñÇ uãQè†~°ÈÑ„ç›(qc?žåÌ¡ç‰]¦g”œ à m«ã¡Ä”œ, C}0¨_5åD„ž–¦G€Iwn†R8 MA›û˜w@²^*©.ˆ§Öø2$ŽÙ”çT¼ZS>CgÿhÓM¾¬BP¡ó_Ê †óóÂàLؤ\D™]ˆÍ!ê0ó@.9é'Ë+^–éžD'”åx?QªyÁ@:ÖÙÌi¬ÛH‡mâËî&!wÄDÄßศ±nÉlì„-§¬À´b3ßQ‰ÎW4'Üh,ÉìÈËjŸ_O'ÛÙ•D˜ø ¹‹Ìáåcb‚¬/­± ;»($S™ªiý‡%97æŸyñ",Aò¼ [M‰‰oÚW2xOð–¨0j¥éX™~bYÁÛfÅÊsÅžbgô]öì Jè§nÅËl.T R~¹Pa}DÊÓK]ŽE‚ép™Îf ô ç)¼%òïyvà@¶f"ߺ>”?Qè-®ÿæýþ—·¯÷Û÷‡¨œ,òÝ÷ë¸çql)êÀhNÓOJóÆXŠ®G¶²ySW®Ë*3"Ç‚, ;¸SÁÍ¡+88õu\=˜°8<ÏÞž'+CUŠë¾=Œ¬´RοDÚžŒ &rwÑÆçYõÁLöŽšh J8Ô• (Šò/ÛE5*+ÜKP'!õÇŠsê ¤ÒœÄ¤@!hÅ}1Ü37mŸQ“!õÐ#‚UÝŠÀ,YA¬¦¦VEê³Ó%ÍeWJ¤.ÄÍ->u%dÄs=Õ-qÃ]WH¡£gïGkñ¤Å › “ôp“•Ê\Ú.¼Âe–¶uS$á`ZŠì’†ë6¥`]°Þ!$‘Ñ"y|JÞ®$Y³·âp.^bík´v½dtMcf’•X$cºÒš~h%{T.7“ÂÐv~AnË6æR»,óAžÏ •¸‚ ¤”åÌÇÖãŸ(3x¹Ý3ÃN½·j(4=áT5Ô§VP“âñFøŸèY$¶•BuHWÃ|ó)[ @Àã’ž5 hL¶ÈéÂŒý& .\#|.tQ„€ÞO!­àçÙÁT$J(]E>tLPS  x˜IySP|ƒ 8Ò°Q¾ÝLkÐöl"Àíü¯šñf¡y,›K÷¼”Ù­§ƒUá5¶³Öu8\€«¡Ü«ìúëy¾rCb6;üÏ“ ú)v:¶>?¤n<}}øæ¯@~ µ¿é7…èà FŠpÇÍì=ž›áÝfMàúÉø¡dr•Œ¿:W†ÕØa;1q•?¶Ì@ ›wUn0Ÿ¾(<¢ù’å=óf°±§èÞð—›Í0 E®&%L¶3Ì$¥ötCeæNs1'¢+l E–V«<ÏfyØ#’FâRy6-ú¶A„-ÑŽ{5¡”h%b?îð‘*yºá~VÇ» ‹¤JMÉ%è䧣Àö–ØËýÐñxLçÝ@$ΤÕôYÉÃMyáø]É+fyq:„h68‡§=ZѾ‹|@`ÑmjG~Ö›+.yæ¿Ê·²¡œÍÇÿñŠþ$‹¯¥¥bÜ |L£à¸qªk×W_®+°7}"Ä ÇQË(¥T:k(9òÓ¸Ú —±Dë$ʂçʂÐ\`;ä"l$M_fÊþÜ §ó‚krÅ".îpïýÛ×|U9þª¤YëeÊå%þrš„£ì"ñæ‹äôvyÆß Sf‰¿’÷¶<û,Ñlü+d#_£ ázä,dH÷°$"ã¿q¹6c³Ö˧Ö06𤞠:8áœAÕÿÉ´¶ìDëmC ÇÃv~äéªÄY_‰¢Ê†æ1èäÜóÑÆ”b!YÚ”%¶d\cr°Á=M]¥F]ÔýãQg%—RPä·'©ÀÆÿ2нb³’Ú;1h̪Ç¿Úš•§¦ÌWTkSp*r§:ö«rÓSÃ×,¢®$÷ ‹Sil49ð lÀ™Ù0€Kkò‚K3œ. »ùá±øãìT¾Ãý©ˆ endstream endobj 877 0 obj << /Length 2545 /Filter /FlateDecode >> stream xÚÝY[“Û¶~÷¯Ð䥔g‰xkŸ;›IÆiÓx3}°=.W„´œJ¤R¾ä×÷;8àEZ®wÓ§NfgVÀppîP®"üÉU­2¥D¡óÕæð,rP»[ñà—ïŸI¿/ÄÆp¶óÛ›g__'ÉJF¢ˆ ¹ºÙÎQÝT«7A¬×ïn~\©,i”‹ƒ^_ßük*¥%ðG[ž}w3Þ“Äñ ¢÷)JçI™‹|•æZH¥ùúm]­q6 ¶ÛþãûíÑÖMÿþ¸/›·QmÚ¦ëgËç)®xxýëïxô¼=õÇSÿ~[ï ŽË¿9†!(©„L/ü#÷(/%ÆÄW( ¡‹‚¯¹¹«;ˆ5¦ÉOJþy þc«Ðš²*o÷æ+†[s´¦3Í:΃¾ìë¶ñ8¶üÛßÿ™\ã¤)ÄïUê¨^ 3ÎD”¨aWßN8Ý`W {¿óm$õÞ/µÖïm—®—Q& %Ä]_ç%”*Õþ+§PCxEæu$y B8®C"fÓ{ªöŸ×¹ £Í€¸tÚÈ)@‡±¥qðmÙÕž%ÁÍZÉ 7v[nÌi‘SŸŒ½ú~1›r¿ ª€€ÊžF9‹Š@ì4úæçDê&@U¨+Sy k²Ž[ãK “ptèU þUî4¤ Q¤ù N”xJäµZè´X…³mŸE63âÜ@„”Ù ø[–ËŒÀ‡å’Eä9þà¦=À”MÇgK檫›ÝÞ°.èÏ"}æp½-›nÛÚÃ:Ô›2ýÂÕÜÕõîÔž:†Ue_©µLbD¦™P“©•Õ‡u’e³!‰Š¡u(»–ÁC|T÷##ÅÝ9ÉB}Ú÷õ‘ü­Xæ;N¥ˆµ>7â®·ÞÒ,(­-×q|î®È° ·ZAåÌFpëhtëÝÉžÙ£ZKó3‡<Ù›Z2:‡-·Z xÚD¤rÏÒoØ—t×gú¯fêå\õ¦îkÓyo–Y.¥Î%áx%7d‘tþ™ECŒ$¸Ó×Æ<°icë[\–nÍoä`5ÜNNNbɱ\QÄJ‚wX­w´% >ºXÂÛͧ1Æxœ­§‘·Û~æÞ/²HcLÕqÚt±Yi/I‚œ:c»Y¤R"‚b†H¥8F*!9V½€î§Ÿ˜ˆ—0‹ë›n1Üs¶ÉD:dÑ‹d5NßWÛþ½¬(µ!Íñb±ªd¢‚3M¹“GË'ŸSë!Ï.¬Á"®Î¤ãÑŒ—uõn8~jhb†ä»/wÝ”£ÃsjžÀU|Ÿ+7˜ òOÁ§z ŸøÏÀùÛˆÈÿñ·Ìj´qùyã9VêõÁñ½"Q¦©ˆq³:ñgÇ|œå.f¥"ŠG¸ÿ^7ÈP_s1ɪ1ÒPª£Ùõ:׈ж6–³¼KSçËë"o@È ¿Ûò¨õ'­5¼§ª,BâW A—ŸlƒôÏao,:Ï‚UJÑoLE“Â$” JÉÇ­DpgÌeC«ÈéDIÊq®' –p þÉØ5'çy—2§ó ­ÓÅ%Žž×fÈOàAœ£þ›_»Q¤ËîÝOŽåÎð(Ž/yP…ÚbÎÃÃ}Õ°ùË_ $(…/Ì"dn{ÿ@ÙDÉF é'w®Pù0fJÀÌÊÞT<)ù‡½–F[×L+cû²ö‹3+tSW‡¼îN51jâò{¶ËÝŠR¯Ã¢™’ó‚_–M‹DªOF.-§¾f§AätA@L!8#.pKîF+ÌfJ~‡ÖHëÛ4¾¾Ë¹¾£s®´¡ce×_ÜÑÝ•hp–xq±­ñ¾*ÄW‹ÑäcÝß±M1WP—f;¿ÙsïÀž[×¾,é(šôõž" ï7Ÿê®ï&3‰.{X×­doãâ &ìõ‹…I‘ UŒ/ ÿõÕ«Å64ÁcOATÚ~~6¸µíy|Ë¥š_`{Ô*ÄÔÛÍ1@¾=‰r³¯ í4*A¡SxòP±ÆRä ì¬gqâ•Su)õP]/7[ºaCCã7t|fw‚ ù Ý9ô¥%ö°ðVòOƒ˜"ƒp±ÐS±È“Ç ±k9ïþI2Ôœò Þ,pý{ž "]ìèc!uþèÅ1 §~žBTp×Ò\hs±^Ò*GØL.ê§ÿTO!ˆ&¥K@œ8º:ý ºÄOké͹=Ô¿s¥5c/üˆ“D!༠òXc%½]lzF…½¦çAëÉ›z§ößØÝ ™Ó½9xOIõ\rÒ:9¯áAp¶¨‹X¤Éè.T=àTE_XÞ¹µ­¿Ìµ²¾Žmýq×׊¥TúˆþY¨8ù#ɲ‰Ì(É#v¥¦º¦‘ç¥'`‰ŸB \0ö¯Ê÷ÌS¿GãkÞDB°’ß-¾Û£r̲y/]Ÿ ŽGÛm]öîãŽô VL9>•l¡“º1Miy4óÚ½NŸŽä`6ã—*9>›†R£JÒéCŸHùÓ+ƒ·4¸ÌgolCC3z‰iðe,5Ú\Óuõ,;#›åqŸÜ~3V­P{¨È}©|ó.ZU£Bÿ’¯>ºM‡•Ô¤ßýêõ³r6›ªÉ!ëX­rp‡Ä>bXGy—Š©%Ô*‘G¾«úž2q¢‚}M•ÂA¸AD,>4Ü5fî+RWr»èZiÀŽkœ*­ûLáÚXKãëËxv©LgU<×þ§}MŸY”VÁ *ÊLWÚ±ó#áüpß´³ endstream endobj 885 0 obj << /Length 3596 /Filter /FlateDecode >> stream xÚ­ZÝoÜ6Ï_±÷t2©"EQR‹>¤mrðÝ¥ Á! zò®ÖÖU+¹’¶NŠûão¾¨5ÇÅÁKRÃáœß ­61ü©Mo²$‰ “o¶‡'1öWn¼ùÛ%t!† Êï/ž|ý"M7*ŽŠ¸P›‹ý’ÕÅnó>øẼ«þ,L’$0ßœ…ƤÁ‹ïxäMµ¯úªÝVg¡NóÜ:=ûpñ÷'Ï/¦ES­¿P:¤¼#ž5Kñ”Ê¢T¥››H%†eü9Žc\õt/*lžo¢zylÆ:ÜÕ‡ªê®-ØKšeß—g:> ®_a£†±ë«Ö-ÿö’Þ†‡ò,LmðŸ®çñ®ßUÒ¼­ÇknMK ߈tv)]bm”NºÖ¿™T;š>6a’'QlàäTÁÉ07å妀Ûtß‚”*Ñ}\µŠÒ4{@¸4ÊbõYá Çfb£îe£ÃFÿ7÷>þàc¨38Îôa¹Ò™‘ºŸ‘ÁAEg¡U‰Ÿqg’Ï|û²ý5ô3Oa/fxêàmUZ¡ITTÀ›ÁÜÜ%F)|‹bÐÁHi2ÉõÈÒ6ÝüpÁÀã>N—!©ÞVÛÌ€7žDE¼y¦²I”uê"2\ôXñÔ•¹¢ÃyæÌ»&xq–¸åC9~xÊ#7åUŤêŽ+Jâ4JÀÉØ4‹ò$yÀ9êpAîÙò)S¹º«µó U$q úQÞ‹]®.Ò ø÷~ò Fî–ò{]¶»¦n¯¸7ÔTBßíùw¼®¸±‡sã–”Š7 W–äl$p y²o ^Šd(}bš(‡#ZP%f)ئY²ºô³Šã5¯ôKxmïáß—¼²/áµóó²Æ®x)åaÆ3–Ü*?7 ¶½â–|·½×þ#eÓ7´ƒ8 n¯!0‹&+}Í"c'·íÑFÖN‘é+ô|Ù䑽’Q–LNi+ˆŽ)±³Ý'FÉœØ=V pB=ínw8à´¾tÂƪ7d ][µèÆI(”sã²û²‡óé,OÐM]¯fÎN–³¶Ýáæ8V²ð¥xür¨ÊV¨;Ú;ls~Ktu:† x#… ®ª¶êË&¼9ö"ïP1aÙ\u=œÌ»?Çi|{ ÂÖÛ3×<ÚV¿#·ª©›jx´¯Æ²a|ÇQ’NWðÊwIydõDëúÆšdŠF~6€fEÓ]yU&%…þD&_H£ÎÂÌ‚K‡ÓÊ@SCà‚ß ê€™³VCIp4It 'ÖroߌUÁMQ…ÇèÞq%T c‚ó‘IP_ñ»»¨¡¾l*þ4vüi{„x€ùÜu˜X¹u4¯Ÿc• >&­äÁ o®ôFd'q¾EŽƒánWYqÃniGIlÄÝüpÁÀÑN—Y' HÅüûþ¼ƲiJŽï‚±Xjü çCß>'ñs‘Ý(º€“ЫÝÝõ{9eê`’+¸@ðØjÀH8°ø·×h•H@cÓ5°?@ƒÜI% ƒÕ„¬t` ‚A 4wÉ·`™(§ØÛŽP;F?xHÀ ao'ØAe^`ª#;c׺½_$ÙðÚüPÙ8à*жÉnå,êIé!—"›Á_ôÞب[p­Ü$…ÁpŸ{<ÝÝ&fôyÂctW‡þÕ¦puè/sò—ðù0áDèXš9Ð b+ìIäJò ÀÝ᧺jv„¦p¤lù·nÛ¦Ü Ý$‡ø‘ÔÄÁÅu5Èçe‚É}ùÐáîħ߂û+a¿;öÓ’°NÛBï©Ï][ ¾`öœ›¡oúåùÛ‹ó—Ï.žûn0+ÀæË`m4€óò_hQþ ¿töFcˆRiy5Àn³8§“ºàOjPC¿­—¶¹1Ýó®G00¬ÝStÂIpÉj{=Ö¦XY ˆð‰mè€çwFîÍ7ŒvÖ@èÅþ¶„Q(qç·0Ÿóý=YÎôÂxˆåwß±Þc&ƒ8ERôvy&`%CAmnÉè÷Ðdä˜gfÅ*œðÜv–@´Í·,xû|iüœ[WÌ ›sÜq Ó/ù¬µ’wLï;ˆ vÞ¨;¿<îtèvM £ÙZ£Ù¸áˆ—K¾´s®saF°Ò;Óí“j‘Ã]USÞð êçåQ(\ 6Á®cÿ+'ìE*_¿òÄtëºòm×îj¸K olùÞ”=ñŸðÚC}åõÛÚFÙ W'©rNQñ,hiñ—‹Ow`2Ó.é(/ÄÃmJ÷¥™ç®ÙM[®úQun‹C"â4&p®uªÈ^qh¥=¬n~[¶kCöœ&„ÀÜ3@ô'Îydòé€Ô¬ßy\º,‘£|ñêÍ»go~ôÔ<Êt¶F¨¤u=ëÄW-—B¾–¯ýò÷Ï~øÇ}ëe†ÔÅz~-ÓàÀóøO©Ù¾?î[;ÑQ¾À”odœÚØò)2КS¼zÃßqçÿö­š#¸wÜþëOegûW9\`IŠ KüQõŒôü{è(–Bë¦!ÛV½ƒâ€Ò¥ïÔwÇ]ÔÖö µq”n6FEÙŸÀØn~¸`àÃ¥'ËœʲÀDX*“<§ˆ´ÒkƒyÿºqF¯†×È5°dªO9^ƒíåÎŽRéó(ÛQ‡ rÏnN™Î50”YXhPælUÍ‚´Z’ìÁ% AC)z oë¦áVÏðÿÂ7#Ó“VÝB£ë뫚˞ÐÛ•cÉ„âšÇú¦©I‘U1Õ&f@8pjª9/yÓBßÞܶ؅ÉHƒ{†S»ãväÏÌS;žÉâµÃ¥ÓIîʳL!Ö ­ºÝSͨQc• è<î&¼IQD(M¨#Ë“G£›.øJ¸'Ëœ#H ‡÷ûþC%@‚í3–e8õMå’:H…EÉïÖ¨¾cŸ eÈ7€—"û`ÚQ‡ rŸ}ž0]Ö¨3 Làôàü°GŒŒòÈ®(L­^ãE0Ïɾ0ä‹ûB*ÀE÷N÷Œô_ 4ÑNytÂY±ƒÑ¾C‹&Œ=2 =ÕW Un·¤-7£ôÛ‰QMuUù_}`ïqñEqOÍoneu<^ÂÛÓë<—AKÞ EÆœ#£”,$1kå h™Úû£%\‚͒LJËÔ î›Âe*áFÄ4à¬ÄKn°F8TpŽãÀCËÀ ]hRÞÄ”ÌÁ¥¼·abG§×w ‹3- Žè@}“èâÑpi8—Ù‘²úij?sI²ôÄYm!7~êTD1Ê„FÙ xÞ—a¸ K‡ÞÔûq^üöØ×ä¥=….ÿÓÍ”Œ²¸DÊXç0¥äeÀWÔRIª7|ÞŸðb ©ÈO NÇ‘-$ÌžÃÖRÌ6á:Ço¼J”FÅœëò.â‚Âêd’Ð?ÍÉ8ãbž@PŠ«7q±Ì5±»Ã@'%xu[qîj<‡†rNâv<„s˜oÇS-â69 ¤âÑAÿ)b1×ÿOŽà$4XÎ1ãÆûwõ°ëääm|`í…žj€ê•$Díôl™ú,‡ÝÞ°MeALÀªÆâÓ+=$—2îßÖ k’­vóúÙÅùóŸ.¼æ¢¢<±÷HÌv¼K2ট”êgåЩ”¨[õTßÂt-ð¡úx]‡±æ¸Âß\ÜI—ÐÒ. Ô©»ç ¬Áó5‰ì_è²ÈÛüñ)ŠÌ |)ÊÉ2ë8„2ƒÚ°¼ïßQ5’W r{¦áåþá&s+ÏÆó–u„].8¬1xÒ•ßyš,2H@dPG«úO?G.È}‘'L?—¶ÄÂq‘¶¼ãhR®â ©§ŠvdÒuÆòe%´‡ãC©bBóƒ Ã->‚RFŸ¢ÿÞ0–ñ%Å£+©–è×$O0¡ÜŽG*@ûzì«Ãüÿ8Ê”¾wh³œƒ”ãEÈ“ÿcÀ§Ú¡FÆíëêØ×`E[.›Ô¬cØä×"¤;þ½á<éWî•L†ø á%ÄpAŸdW¹ ;݇²aÜ‚‹rç·£ð„™Ð•‡Üwhékö`‰\´÷Ùlzd øñõôî8½2È£ãòÁVŒïÁ„)¦W»$™ÂNäÌäÛK°¡ endstream endobj 893 0 obj << /Length 3207 /Filter /FlateDecode >> stream xÚµZ[oÜÆ~÷¯úÒàeÈáC"ȃ Ë© I¬uÒ )Tj—»"ˆË*úã{n3K®([*Pèaçr8sæÌw®£àć¿à$õOLz©NNÖÕŸFÛÝ 7>þø&º%.G”Y½ùî}¾—úip²ÚŽ—ZmNþ±Pñé?W; òb?†UhôýûÕ/§Ë0 ¡Hòæ|åö‰”z!CHù„£XO8J`2:‰í¡æýóƒ›÷ñ˜6.á<#.o¯ÎÏn>}<êÉY °nÜ™ú¼,»Ó¥V±[}ÿ°¡®7ÜÎjü5‹f,ú¢*þŸ.ƒ…ÌîKžw§*Y<Ê7ë~ÈÊòQØHÆl¨È÷â$µ|¬O£`ÑTû¡/êÝÛB…–¼Ë?ã>y›•tSËÐ$^$'Ë ô‚ÈIŽê˜ßp£Ê³nhq êö÷yÑr3ÿ’¯aó¦–™¢Ê=8Elïòýé7«7îCKöˆ\4ƒ¬QekìßuþGBØ Ö™]9ûX'¾}a8c5UÎÃý±¥á"¿ÙöyÍÿlóžîòuSo:  ¼Ypľg@0/E‡òÂ؉™˜Gnî…™M¾Í†²çQ¼ôšE½fÒ¦`^*y (Ò2ˆÒš¦•Åø|ôœ¯Ûâ©F:t„_äÿ ßó¿ÿõìÓÍêâçY„¥`M:E°ù‚U˜¾TCž"8-‚“)‚ ›yø•‹@aËÈC!h64AsBh^ xS€Ž&àm D9îD‹¢^—ÛRP]°šµõÐëï AÐzx´¨Ç&(I+cCM2šÒ~ÛðïßÜìùs줔}šš.ˆP4=Š†¦·Œ-an ­‡¢Û4ÈÌ÷Ù_Áo†›}>…‹ÌŠ2»+sfóÅmviÐØŸIÑ€•¦  ¯ÐûF j€¾…­r»ÁŒ¬Uâ{~è\íõ§ËË9«"•ÀR ˆFÌ·y?´u¾ñØ_¬0„àÇ=¤SDr\¡F ÇÎ$‡B\B0‚g€pL Ótnp¢l‡ª¢¼n«Õâ_³hJ/>²@B9s£ØÜ2Øqà£%êÙÛÖ¾ŽîÅqJàEŽ@8¯«Q¼Âs€´§Ó©>8„‹ŒbÄϬ‰ŒgRý*ÃæÆw¦„q÷dBF.þ’!‡0eC3ÖáBcèÈÅ`óá¾áF[²žF ¼¼’Ïéóuo©s+»–¢½ôHœpœè§·ý=Œ§Ž qHVíK Ó“€C›¹¥´;TÝ'sF†›ZU†zt+ÝÏ¢Ùç›Y²æ|KÍ!=P¶­,Ea*“.Ç>¢kxšQŒ#(gnm²>7sX‡ƒQU¾}ï —&ÎÆÀ¥‡ž·Î¡{àAÜ®€O#iè®àÏ—mÞõm±–¬É8c2çµ8Ÿ…È ÈëóÙwç7«~½½¸þéÓ¼5ñFÎö‚?ÌaÉž¤SP˜¦!€/›í‹Qºˆi ùš.%CAÈÅ%›ùäŽ{Юæ3 9·ö,^Õôö‹žEJIðÜ5dÀÆABaS8Û­¬½+à­L üJÙdÙi†Ncã·îˆÔåLérc¼.rVpK,8÷$üMNEผtò U븛á$ýÝ7ƒàõ™ãÇó›ó?Ÿ9ÖaÔóÈÑÆ·Qiœr€ #Í0† Ñ`ƒÝ O5tý,> 8„¤”ÌÐú^ëD-àgÌ´<•xa¢ž‚VGǤO€_Kö·i!ÏÚ‚ÝhbS·c¦H;|‡i[ ˜á! 4`®¿,û¬NVRò)Ù$C¾>‚Ëéq°*›OhÑ9øø°DðZ»{i2©?à>]BCÄ„¾Tæ¢>š.3°¼ÂÛ2ŠÃÀiï³®;Ââi1:~—º^†u奱;âCQb*“@< lTx=Ø!¥…ßÁ¦ýäàÒ±Â3›Shmä v/8€ž¡±ö)±á64Ð!xßÊõ G“ëc+­å£TW `„ªðKIbÓòp´ô¤Å»Ïê5º} Q%ß?³R–}±Ü€ÕƒÌ·©)›ƒ)íêÀƒù¢Ÿ.îÉJq:*õ‰·³œÊ¦+çrÞ1o?sjK©8ErDv%¦•%KüŒs%Å囚2pèðÍá|Ëñ& ílœšØ‹BóͬÄPdeq,ù–ö5¤4ñ´U‹/$¨hó?¸ïYðק«Ÿ®Ï./~¼>7Çu¢=Ç_ñúÆGFÒ{ í]ðxå<)t°ÊÔÈp5´bÓÙëåP³‰B0•ˆî]]IÝ„|%ÉQøœÀÅV´îæ¥ ßÔ Y¹÷˜Ñô¸›½¨²n {g+õÖX@1ZÜ\\½ãywB=ÎqÆ ¾èSöÚ%†ÂC7Á»…;èØ3Õ}yþÁYPi!.–ù©ÖŸwÆ(sõ*p3ÑÄ}$jÖý¬ŠÙ=A#=hŽfCßTY_¬¹¼‚ó›ÓˆŽç)æ(:¸¢]m³Œ|ê„IÎÆ¥LpZ?¡ç—i ˆòÍÐ(”6›V¸ Z‡üH"Hã4CÙWî]É\`…‹4GË n¦ËqàCün!ó6k·ÌŽ³t3v k;:àBHB…ìJmFû I|E០jW˜·.¤¹„jÓúJB&z±àÓ;иcS-ŸãµíƒŽ-ëqÙh‹Ä)%"56ùOÏ9É)R·Ûþá±Ú¬gCNâŒH%ï3M­ì#ˆ»QL³xk0‘åNvò^R{6Fy’R%‰§B›T ‹—EUôœ×!eyh¶°°=[o‚>æTù¸}à" þÜn‘º¼¿Å¥JÜ ¥·iÊb‘bôŒôýl©Ámf*VœE˜€ µMt€àýR=Fí»iHd8CMÕ5ès¸²ë““†‰§#}xý#fŸI’TtLGk㉳ýž£’/ØŠ¼|¤;"& áݾ[ ÕñÌq`)&!E‹‹íle%…«Nžò¸øá‡QmìúÃíêâêüòâêb>X‹¡­Í¡àù¬ ¡9¼Ãa‡ëåKm\m…iû¢Pç;°ŒS<·P#VåÄ KW¸Ÿ˜ƒ­åEk®«á/;Av)Í€•ŽIlÅô-? Éó²ftËל¶âʙ̵ͧNø¤b0ÑñヶOisµùíL˜ }æ°/JîŠ15F†Æù Z¤)Ü,d1H>X¶\ºJ¹žü•Õä­„HŸapÖ»BÈ[Bú½yËÈá¢ðáqÀ>ÈŠå÷ ûìð¤Ò‰”(uÏm„ZJ”…É!'Hɸ.ëQ@µ}|ÆÒ¤²Iøâ2¨"[u("ê¶ëyk6ç+®±öbº¤gWg«ù§l8ã!×qeQ9Áþ+›ÿɱlHŒuëéÛìWÞÁ %~2Ç…æDž×ÿ¯…¼Äʦ(LÊ5´Gtlµpbž÷ 5p½úU„Îÿ¡újõŒmC—„¯|Ü<¬m_2´‚«ŒÍÑ??L£wyC_’gV î]쇣’ÇâïÐÊÓÖJÅ–ˆ#ofÅOþàë'JR/ÒúÉ…zößþ ~mvÁ endstream endobj 901 0 obj << /Length 2997 /Filter /FlateDecode >> stream xÚÕZ[sÛº~ϯÐô¥Tjâ ’ÍSšcwÒ9'É8Jûàd\Z¢$N)R%©8ίï.¼Š¾ñƒpY,À^¾]š/øã‹$XDaÈ/Ö‡­w j\ÿý·t>ú#Ê¿­^ür¥Ô‚, ¾XmǬV›Å÷fŸÛ¬^úazò¯K_Jå]]­þE#×Ù6«³r-}¡âX{"Z~YýãÅåªßT ñLéòiñt,%‰÷®j3”Dzí>m±¥ µä^†bÀj=^­có63k›l]•›ÆN¶‘!‹Â¨#LëÝé•K{-mÖá¼±—­óÏ—YCWeñ@ó)ýÔÕi·§f‘òöÑå%ëtÝæë좓ÛåX¤eIw.½CŠû> ” ŸsÁœÉç!㊤;5xùpõMuÈîÍ-ˆ8òUmÇÛü`[ùÖŽì3"¦Œ|–ª¡É:K׸÷>ÛÐÄý.aÊàXÀ][i±Okµ9âˆP{‡|³)ì¢ÊŠ‘–æT=§óTönë´Í+Ã&±+’À[ÃfUK£wDH9½OV×'PØ M ïµ}4bbI¿. ÿêYÚ´ØŽ’þ pá}^v÷êp,²¶#‚³#]^î\ÉK— F¢W-´¤ÛË«·¿¿^]º41ŠYÐkì¡2'Þdd|ŸÜïsz$Ò–¼¡)°<54šý÷”ƒ¬_—Jyi1Òd ¬FZ…À–UœÃ÷¡6䌤Œ@ÊDá@Jp,`YÑœ‹ƒ¤•,dáÒWZx×KL2-üM r´¸Y¤¼_¯V늘)\eÄtgæÛm{‹o섺·›m{[‹õ-ßàeäeKepa^„GS2œØ˦:Ý¡&"Ý˼¼±¤çýfçªS{áz×SÙä»Â,,Ò]ƒóÊšèdKþLùÅcùMcá°…Ï9ÑlDü±Î89`–ÿ±RAXiÚÙÙ^–öl:`¡æÓþÿÃ=}¶¹Cñyœ0ŽÎa%ø`N)âˆ%:ÿÂÏËã©ýÅŠàÃæÐ%ªMÞ¬kr„@|µ„¡êTçä1ÁÀm4Ûª>Ð<^;ÊDó¤÷=«+©jé–„- æú®§l .4ä“@ˆSm<°é¦NŸ›lv^txVÇõ(ÅxÌ;R¼É²9¤*a‰NR&¹6°†)˜3¾šqaðÍtd tºõþˆìÌ·1"}ÌÖ c‚\ôͧ¦¿|Ë拽Ÿcº³]ˆù1$h@¤¢É1Σ³Žú ¡çL»{¤À±Q,›±ú½72ŒB:›ªuö€}"®i”.mQѱ“ÒY(¶¶¨E£™uV·in'G:iºÈú±FÔ1OKËø˜Öé´»6Z†úlOIàLÀO™ì nÓq'´LøÔ\`£I¶5†Hd0"Š¼;4>œ¡cBcƒð3DÞEtZ­ 2vë &ÀeiÓÎöhöi5®³–Ž1ï‘^k‘í}Þî;¤Ñ£.ë¦ ìRr r :„mŽØ±°å[Þ´Í &ÁT?V—i=ÂeÐ!ÛwBˆ$baØ[ü»O¿ýæ2vƒVz*ƒŽu‡YífÔBÔ Âà Ýxâ©$N!pšÁ²"×õýE’IÍŸ#XO4ˆaÐ5ü¦Å=>Œ}æñe€i•‰¬/™D‰;J$@¡#ûj€ÁïÒ¤ëô65ÖC§²Èš†(Ë„‰´î(¬ßA8O#ëSÓV‡ü»±R¶É"ÎT%*ÉîÔA{œnNGüU}Ö®€À9hj[çk2ÿ$‚ü­¥ c`ˆÞ; GÍ—œ‚ Nä–`8BQÞZÐïB¡ˆÊ@Oú‡ëË—×ÿ¼¼}ûîç•ëå¸\ „«tGûX›­í£mîÛG“š©GØà F¼´ ã0ìW[´bmóŵ°ùùØÅYÓÃèeˆîÈɺœû³H>˜bø×6õ Á^­¡i9>o ɺš‚u8st&óÑÃÍ ^:£úɬ¾ŠHÓ23d²È¨3W¤èâéÁfMV Ýf_Š µï:‡;b  ³Ä’X¦À§¢÷3ÞÚ©+ g€­Ð/K·f0©äY þ±õƸËÚØy4ö/¸w9*„˜ñ™Š ³)³U*Ý›ó.«M2¬Í|t ™!$oBA®ŠŸG3v½?bàB3³mfh$Í@ãæM‡z±÷»1‘_'vØ=”Ãy‡r`T=*?‚)Î'§ûA ÊR?q–9Ss+ Fgz+HŠ“þ­ˆ†ün’@®°Ìíéa•N÷™‡µüOŒð²–‚âýÛiXJ²H÷6É°Sÿ¢1²ý³q‘ú ÃéÁ!9=ËOÌù´8'i¨˜R}À…ñŒ‰ ü ÝѺ2™!Žh>U¸1ΙŠ¬HÝ'ã,I§œ†+&“™Á ¹/+”ýeqgR1.ù\¸˜Dã¤x0èÄ\¶Ó±dð4ás¹A|íÊš')0³÷¡!\%”<¶>XòÈ'Ñ%šNïµ3;ËwT¢Âtéø‹E©¯K¥mE*Bœ‰ãø~È.6ó9Y”ŽYÌeäõ%dEúÉ™ã7b/g$IJƒêhÑxaUÚu˜SËHw2ÑÚž±¦ S€3Q{”Ì"üîò2­-k@ò-Ù2NnAqëêàHãMº=-ËXÈ ÃÌ|}R1ü±þ™ð[gc”zBSÙƒ¢’?éžÁE<‡ åRµSáñž´¼ #µùâb Z =cŒO™šD âüñXWÇ:OÍ·Í»„ K˜kÔ…=õj€by™ÑÓâÀðY`û@#6çãV±Eé ï`'jX¢Îªu˜ˆ6¨Ãé1U è¦öw³ ¼UÚj~ùn¾R44jÌTJkÁаÆ-ë®x°Ð,‰ÂØœÄM´Z„z“.mu)2hã¨B3e<߈èÎÍ*¦¼Ôsx­Ïð‚ù1¯è9¼6n^|û˜çf´bÌ-ssQ<å>‹ÛÖ{8<Ò˜š„4ß6k»¿'5fRÇŠ9EŒ˜ê þââÍAM‚ASh5N{QÓ²Ü)ü0u³f= :à—›D¢÷çÞ†ÆH‘È'¡™}³!¡´~½¡q*ë˜Æ]! ø›>¾Öc4›/ÁÏ'*ÎÉPl‚ö!KKg‘ÑØ}sÀŸ‹±w‡HT]üã©îXfé‹]Uçæ;ÿF†O¡å•ÙWìe5ˆÝU±ÕaE¸q¾4"ù¤2ï]ÏCb4Îá#PÙ¡PºÙ„ºG Eµs*´p²ÑLÉx\QXúÖ-ÉëÕ]YI¡'•“XÎ,5Ew(Ê£”4bD(ó»6…±gÞjG¾ ƒ7ð¶¥%×ÝîUCnõk¼vµAö@ÙWi˜Üöd×9êp‡ ×ošWø]¤tÔ8Èx ÈF Aøó%»ÜÖ»¢Üt“éÿÏ < ß›·X\/Šî?L•ÇÞ â§9Þ‘!ùCô¿º‘0yôŸ6Raý0òŸs,ñ4cIqÛæ2B{+t}Õµ“ç=”î endstream endobj 909 0 obj << /Length 2932 /Filter /FlateDecode >> stream xÚÕYÝoܸÏ_aô¥Ú«—GR”(]q¹»¸M‹â Ç@’ •WZ[È®´•´ñùÐ?¾ó%­$˱Sà  x©ápfHþ8¤9ÓðgÎR}æÃP¥.9Ûì_h¢67gܸüË #|k`\8¸zñíE­Rš³«íXÔU~ö.°ÉêÃÕßÎBoU¬cBÔ‹‹«·«u†A¨àY^¼ºôDÖ>Ó ä|Â"½ö,Nœ2¡cõw·u[€~mƒ¶üµh¹™5D3Áaµ¶IP¯àßÝÊú hZÌigŠ–‡›2Ûíî¹w›µÝ96“ «rf,;î+EåMQM?Ä×,¥*ÞkãP–Hªîïn \*˜a2ž!lŠ •‰xr;Ð+\“u°FÛo@^-ëŠw&:0”¿»ùvcþ蚬jÁ†½|ÖÌ+Æ.dÒHÅ©ï•Ÿ‹ê1‹’¤çRÌ»1ËÚpl Œ•–݃Åñ".žhˆW.ª4‘ ½éyp_].–ª>>¶œ:ŒÈ€RøÁMŽR^(ü½-¸QV‡£ô1" ÒOĦY™ CÈÝ·Ò½É6¡Gårƒ:o¹{ŸáÇ=ô;4HÀ“¸Ž¼Jc#1l~›í –÷^Gú¾,vyYÝ0…0¿eµ>ì²ð vÀ£VëÈéàê¶h¥¦Â–´ý·tÔ8»ÏØU4wMÙu…ˆÏÍ ôT|“ÅSx«]ѶKÛ†ç\›±‹ùøêÍÕë¼¼zµ´>D ˆ'ÒYÛ"çì%ýÒÒ;kwQvÓÂd=œi\¨+î2“éâwUôRªºãÆ°,»ìàxƒïÉa ƒkFL`l]:Ý°öÔ{Tsl;þí0±Ûâ`y7YWä 7d§>¸X%pÄáªø—¶á¶Y­ ‹ÐStE1û+ “¿ÂA{zÓe'¨º šb_…b˜èìn³n²•§iCÁ8òuÉ8üyWîv=Ë¿e#ü‡,Ï ë0M‚¼h7M)*sæàã?*¡KìàY¬­ ¡À¢"èƒs ÿÀ7b„™RÆ¡¦¿ X7s54Ñ7Ŧcg ¦9yŽ›ï.aâë<ë2þþéâŠ/‡e PØ0Vu‚ìê>ë>œ³g>d7·8êŽgoR0œm”¤Ê¤æ‰°Ús¯GìK¡u&ô ~<…9»ððãÛœ¹¥óZ•ŒÜ="4ñ-d\rá$Ü•ä€ðó%÷ã±ý×’Ö$V‘ë¥ýgIa¤¼bÚùx¡HòÑ â×¢©…Òðï¾fÀú€ZUMoù<üZ™ŸSà¯Ã>(aÌ»X« y>H_ ã~üz$`akçjæ0öcËð ]ª,ÊîwÿÜõD^à6ÎêÍçgm¬@Jë'@Ús®…uaca'p¢2¢xb$U%onÓ}Ç)HµB\cKI ºì,ìÜÔûîø?<…yعŽ¿pmW7o׫©êjÝù±Ê_Šn³Ý–jù͸cWß”J\é_‹f1E¸Á La©9o {_‹Yåç’ú5â;nJž¹Hº¦tä»Å<)Qà"L¢R›ÊaÝvw1zCB`£ÓçÇ|Û}„„ò£Éq=ʪc†Jsˆ7ÞÃé §S¢ÑÃÊ"û7euέ¼>^ï ¡Ö8‰¥L¡-oè0‘-@\? Uù%ûíCû©q¢˜ßÙŒÂçÌhF±¿¯9N&Çø“˜³©«¶›Íí›Jæþ/œ§É¿åÜæ® ÎU²`tÀ®8cÂãzs„g×ö$éÒ*Ië±Ox¸”„®®¥žŒš88ð7Å/›âÐõìY7î†j&¡Á’xIA³à¢$ÍŸ²¥´¥9ž¦\n®£Éó­…b÷àØaán©˜âßcU¡`J«c>Ka—¦õÆÚD†²7¥eÀƶÆ4z(ýs¦^KNJCĬí¸$[š4bÖ¨Ï SEôYçTœØ§ê3á^Ø—ê³™Ðq}æ0¬mÂ^`¨¼{¥Pb\Ò%Õª˜Ú¢Çñ}¹ê¥\…¿çÏ‹Uê¸4]‚ z¯âþ2VÒn?œajoùWR_>R#:'±Øbÿ×gÚàEØ¥"â;€¬¬Z¦´÷û}Ñ5e!ßÖ$/+& S߀ -„¾Ï>Q ˜&çdÉÃÏžÉ÷Q°åY’$ª ‚ßÚHim¡°\'þ«Áß_,–ES5³âÎG ~l¼cäÓ®ò¥=´zäc{„|ª‚q ç>^À¹K,8Œp2ÃÇqÞs?1Ÿ¹Ð>òÒÅ$xN\í7+¼¾B¯¥hM{Ù£#Zf–è>WÒ}Ê£O)–¢_®Ó¨ wwT3"‘®Öl/¥÷ær3A²ôZFš C˜—YŠ¾ÔÐ|‰H|ìôÛ¶¤\“ÄÖlˆxû²)– D<1÷¡ì¯•fKuâgHˆZ&н§á{_g\€wñ¦”\“i² p Óíà;#‹ÆN¾W”ÉyêàçêªùwH[Ö jHð¸ãUÉ)K…½] KØvEFרÆËe->Õˆ3Á1¸WN÷7¹@ \ƒ(£ü°eRÙ.½h„ak¸-nwåÍmÇ ý ²$ʆC­G°`ìBÄ.F¢U§ w£ÎF§™`{Ë¿Cç,@ŒcfË]„v<”šÐ¸ƒ«µ·:x»‚D^Då5ÿÊ5ø(ºC¶Ámÿô‡™^öºØâK*jf‘EôÝá È&ätçÒ¥74 #íø¦ š¥(‹Ãߊë&‡¤·¹êèš8Ld%¤2i¹ÌœLª·¥Tü¸%‰:4iþk1d GUìïnK±²¯m’ §´%(ê¯oãD…:™ [J¡öxÝŽª!~‰ ƒ;v»ØÂE­ñFi—Ÿ^Þ("ѵÎm1»òaz)÷7CÿãYlŽ„¶õT«,Cjd‡Byitg±J=8 dP‘Ãý§!áH/]‹{¥ÈýN\ï!,/® ¾6ú¯RhSh#…öù !ÍÄâRÃb›Ø?hIï¢a¡WQÎ 3ú‹– ŠL<ârÊœ^ÿ@™^š¢µ€Ìzzþi’30<¶:ªÄÎ\ûpâ\Ôç¯ÛZ¦dÕÒ¼ÓTiÿÌ ‰¿9Ÿ2äâ'èè«šÿ·Bû|…”Ì?9鵋=Uq4özÉ|§¬{..ã/ãÒ?K?ÃåÜùÄHÓ÷ËZîô€&+ô'ü‰#¯Ç)®`\§ÏåPë!<˜ôÙ‘ŠL<Œ´ŒÿÓÓíS»!v endstream endobj 917 0 obj << /Length 2878 /Filter /FlateDecode >> stream xÚÍÙ’ÛÆñ]_±o•å3ƒ3©<8‰•JÊ•¤”uùAV)’ˆH€Æ¡•\þøô5HÍîR‰s8GOOwOßÐ7!üôMÞ¤Öª<Ên6Çg!­v»¼úÃ3-pk\/ {÷ìÅË8¾Ñ¡ÊÃ\ßÜm—¨îÊ›×ÁïöÅi¨ºÕÚZD¿Z­£(^¾¼ûŽW^UÛª«šMµZ›8Ë’Àä«7wzöõÝtilÌ•Ô!äÓä%Y¤´˜¼®]™,¸_‹Õ:N‚´HišmWV2ü>ŒÃûzØólØW8ȃr¥ƒúCÝ×mà ];6¸ûv ºdô àÐ n°»}1ðfÝß ƒ{ªäâæð‰GýÐvr^ Rf²%36³*Α}âåÐÞÍ wÆ´ UlÔ¾8lèfi„¹i«tÌ»ÈlÓ6ë¦ÚCýA‹Ll»ê‡^ª®zäæ_2 N‡±çý¶ÀêPQUƒ§l·¼KòÃÁ¡è•õ±jXŽ„è²,†ÂÑÐçMd"ê5“ÏV7EŒ´ 6HF{<ª¼0tEÓoÛîo‘€²!·ß­²ˆxÔQ°iÇCÉÃ}ÔO€•ºé‡ª(Sñ~eR䑦$T‚"êa¥hp_iÜŽ÷'fo}ô¿¶‹©>Šá%HÉl c×ô Ò:Ø¡å½w`C)éLz vWor/»(ü[^Û´|I½Û±—ÅÅ“ x-WŽ}U*|ý‰ö\EQÊ´¿\e@²æÈ>ý‘ëv»>Š ‰1œßÈAáFW>$#¼÷ˆ›O Ô°Vt½‘,³‰âÆiOË}½™:vz6Ä.Í#<³€†A¢›Då©Í Q18”µV‰^CÁ-ñ$*Ëaa†ú> ïB:ƒ¼êBýÐ…ö?t¡¹žÃµ1GÌ-I§Q"#c샄Y !¹$L‡RV T²€GPáDTècјÄhŸ›Ì”“%wuÃj ± ´B Šª6HÇVaRV]ü¯g8[Z„pзì?ywƒ Ty õe4]ÝËn@§fgç0î—‡Ñûá„ÍÊd!™>Ô;2– WE·ã r隆=Ro!H²‰S±X\BÔßÀÁýq‚"¯`›¢¯Ð¯$ÆEÕÐQçÓ3o{èB¹‡×ŽHöˆ!×Ñõ¡£öÉò9HžV±Š²¬ð¿ôFËD¥yêàÉ­H¸ UšeçR©>Âë :…X„‡±êyNnYÃÒFâб% ÍUGÁ÷¶Á/ùO»¨ g’3mUœÅîPÙŽï•×*À§eò¥®£Q†ÕÇbC~$JrHŒÚq·çM4\Û·”\%âŸpkÒn\Å·•D9‹-¨NÝ‹_…µ8Óçyåm–.œŠQÁ¨d¥$ÄiÓRR†@9vÒvÖùŽvЛ <ÞNP<›È¨L/2pô*»zCNæà•c¤l4ðüUÈ@¹¹R“Çõ1}ZÓ'b%„íÅEgÎÞ‚ós˜Ã rgoŠ2’‚nÆ\è”<¦«éëÉIèx™9Ã<#Ÿä“ aœþŸˆ)…Q¼”¹É'øãé©è{'%‹S=4ä/^¦p…U!¦xp'$\1E°5¤#Æä|1¼‚Š±Æ5Á+ÊŠÃzhׯس¦Qp·Ê!Ó‘L¤÷¹ “)[ ÜHõ¼Ý÷o‘@Œå™›¾íL÷V—èêfàM|6üCãçµwŠ®ŽtÑeË E— þ÷5%<€#Á3‚wM%ÛC±£jõ×Α,ñé+˜0L„rϼ¢ÿm¾ô£Œ¹ØÖ·ž@þsKÁ^-…yÅ<Âê—È*¼æõCo5}È6ÿùž Lð½r•~¸ëóÉŽl ©ý,rØ=¼?¿Ò¨.=~ÊÐ]‹*è¯Äx”GAÁRìçØV9à õ„ý \Aq°T4/`ä½- ÀAñ5[à¢OŒUQKÍ(X@â{‡Ÿÿ¬ºVNuü¤þŽæþ€W5º Û.u³ã’¯h|Þ1A<öù¹}yZ¬t¦ç؇}U]¶= ‰q?Jbj}ªöVòJꞯ,›¡îüzÀÓ½¼†Hú[µq-' ‘Ãpèõ·ý$|Ðþt‹N¥I02æ’ ›k•ÛüŒ‡;¸ú ¢/‘:9j)Äm(hH¥ýê2C9ÙO#O°Ñ—N>\Ù€rzaRð[.Ž¶Ü"žv6UÇÅN¦R^¦Ô·--u º^€…{Zÿ)-ÎZš¢\ÖÀx`øêcÝý¬%íË;ê™$‰Ë­x¦ïM‡òTY;õ®þüí7ßølr±ln½!•ˆvX^Æ£ ÜŠ¾mà©ëhðk¤ÂŽ© ¾âElÓûèÒ)‹‰¾†°¹+2‘A­¡sy*‹ä™—¨J12cU¤ó‹¶«f4Vyü5Çï (Ô8o¯ùÅ«nK0Ü›CÕ÷¼=§Î©Œ ÕQ\®ñÊf„ºáSÉsù¦¤±éÕ ¢ìÆÎ5Îa»O'iÛ 5Ð5cûÜÈ@_‡®Þ°@]ªDåZ×GŸòä[×d€­k b%·öÜ:Ao¦Ï1HöÛW_ÿþå«0.*£l2AK\sJ¿¯¹PåÛYZ°×´beE¶ªœÔþój!„>O¸Z°|ÍWÝnÄL¨÷5‚3•G±ÔÖ¹_zù5*‰§f7^mLT”OPäNŒô ,ö“\€.`à¢b@²\:`X'Ï0ð¸ßó×›¬ëì°t |xÙã¥h Þ&ŠV&›¬ü¹7êëX…Se*^ÏA4*3¹#åb^©eTYΉ}–Ë {_ÈøЫàÀÌã*;ÝO<?ö°ænŸ|¾Ä0þ°â œÙî Ã&MhØkâöß½#$ùÜ퀒Ð+˜t™ýBXð"ÔØL˜„ÕÙõø&‘xÛDeÉTËCÁsbFG®DATübb¨Åsë:ð*â“%TÖãá£(M(uyÚn¹Ñn½Bƒd(|¼O«tÆrëMo)©Ï>=Šcm¤oìÓ‹Ôb‚µaË.°MÄô”3¶Ã'ü¾B1‡Àè¹+ƒü6¿NŽÙBŽô0\3àˆÂ"üW?Œµôº¦Œèü)éoÓžØÂH#L åg!ÆEï&d#}ÕS•HÂß¼)µiN£"þäK߀aMù\ôZ/dÊäù—8éæšç2,ļ¯kT>ƒ6Þ°:¹èU¿ð¡ÁnöL’þY°˜§±| k¯QÞøpF\”¡|P}i? v5û[,\cùN‚ bËn·Æàt9`! UÂó!³„Ü“À/­ÑE`.{Ñ€¢½ù'V°Ë­çHiqƒòo*)Ú¾j'XK]œàö$ƒÿ¬•0õ endstream endobj 925 0 obj << /Length 3707 /Filter /FlateDecode >> stream xÚ­ZK“ܶ¾ëWLùn,Âăi•Ž‘’ÈJÉSåÃJåpg¸»Œ9ä†äj-W~|ºÑ r1ûP\s ˆ t÷×oøñMžlŒ”,WÙfwx’ØÞîbC}ÂݸÆÞÈ¿lŸ|÷:M7‰UšEE×'"‹¾ôã{‰<ꇶ+÷ÔY5ôìZzŠ“8ÕÑ¿ÛŽúÛn_ºæM5\RkZªÿÞq§}î¤Ö,ÍGîšð4S©Çü"ËL²DÁÉqÉàdˆRã@m:gÀ%7Àzˆªà,MÍ=Ì¥Ì$üNæ ™$j"Ã’!#þØÍ&ŸB…ãLïç+ ñã„ôL$–GŒ±“Xs&,áL²™nW4¿Æaâ)ìEM,Xš"ú©,×V¨$g9LÑæfÊZ"Ká?–€2.¬I.{|ÛçÇb¬—±\ýTî0Ú¸dy¾;ᑵI *±È3fp«Çœ¦.Ì!æ‡Ñ\ñUE¯O2R>ç§ÔsU\”4”§ëCIÊ$€ŒN ˤ¼ŽÆѱ7<°å5Q'+«%xÅ<ƒ!‰†fòqèE +‹zz^”MÙuý…^Ïà‘Ee?Ðkáž—E³¯«æ‚Þúê÷ÒMoÏé9\–Ô8‡c¤–³/žl4Ë̬§)¨„Ü8hY–Š׊epbÞ( –‚©*ŸÔY˜T’,i¥¡µ;B þ÷i™#´RŸÖ>LKƒÅù´8£>µ2LM€©/¨ÉQ;Âãzɘ…02º¹,;·¾–K7ÄŒœ°µ ÍôÐ߆ˆäLè PƒŒå ð“¯¶&/3mm®«ÿ‰¾T %ôhÝ+G&•¦«»vÄ<§üíÊZBÛ” ÅÐS¿õëÔ8«†®è¾œdS)½Ÿç;YLÙµ‡«ë¡t«žþà’%ZÊ¢Á¹œÌ Ÿ=úš"‡“˜Gv|´ßøêºCŽÁYõ¥›S_´Ã^?&irsYíN`Ì%u5åg$WvÀw]önÍ®ŠªéƒRV qr)ïC"ʘ“"à¢!B ®jÒ…&L"›qDÝ^„¨¶åüN*©Ìc†£ÀwgËc'—"XÑìÏr¹#jèí•W¸l:4DÚ¨RØ=¿h˜EYNÊÔ÷ÕYí-=w(Ìkpb=6†öäèÖþ²ûêcÂ%Ø )åB{œÎžMCÌ=ÃÉ “·ý¹¯c@i–±,7öçãüØ#pnëe,§/.‹«íùä =Oß6ýî¨ W]ÖN±ñbyJvȧk BrÙ807·â–4Aç‹}wÙãè{vµ&J.ÛjY¬À·U@ÐÕ£FX% ·AÒ¾æÝ\’ÑB´ò Zì€WÂh8"ëG…ÄQÆ+(`œXÌS@È©õ®"'ÿžýp$´ hák" ÄA ±ÑÁx9¯2!ÇW/_oÄø3Ù~8ø†ãƒu"öþnb—Ã(âó×ÊúhajFœÛÇ0Ò«=Ý”P=D™ ‡"o-¢jp«ÑL¦rEt)08RéáÛAq–ÌGÀÝ–ºÐ’ÀUŽ\.ÓÀ0>CL1{óXAžô«+­•’?¨ý3:Ô ÂØèî%†—øÒ_¶×õž&»tÔÔäÏ`ë6±Ç ç_ºŽâl6Ï'ãTñÇ>¬ç Bdk=ç&¨j‚é9ƒ«š#¶ Í䊱 £š•¤½‚i+ŸDê̾rp’Ë­ã°O{Ш+¨I1 4€úÜÙ!P(,OØ¿ÏW4&ðš™F)d6R€¿S¶/³„‹l·äz•wõ:I¤‡NñKUÖ{›D`OÑгj⫺عqΡ¦*‰¶—eïþöË,ôîþhqw.š¹¨g(ùýu7- ë4 ¼= E]× Bi4(Ë\¡@¯üË«Ÿ¶oßý°}’ ÉY2[:F â´ëÞV e«@ð´g¯ <-.zØ­I(NÛÒ_|YV‚÷¦©4í@ID»è¦öO1&s"5ƒ½‚ •/,Xp.á€çwÝ¿ÑÚ÷]Á+£¨Û¾ƒìä+lÊÂpÈ„ÍVfÌTš+Æ€WŸ^ §rò,ß…Q"èâÿ¡ÅÅý„îÝ¢Ì}Š¶¢®çÈ劬UEéò%Sçþ° ñ{:ѹ1³£Ä~§½ðÇ91·AÀ·²Š¡ O´„\§ëÆØ£mF«—³cEÛåÚŽ>Ÿ¤`Îuµ§Wç™-%Œç$LÑ R+û]W9muã  ÁŸL…˜KéÄV•]´róc@(\-³*ZkŠAÔââ@•Cκ´ÆÓ%$rCãN;1Ë’ºþ'Óc™J‰±L…ÝTT_–©2À-µQÊ°Lñ{ËT4:ö†ËTK¢S½P«è-[bc(äè€" T˽ð6J ·ÁH'9=Ï5ÓÈŽ¯]s­Dî=x·’Ú}yUt,àëUGȵ¿ÞY´¤òÔHÀeàD¤øÕê(¶Ëbt}öïø¡×Ù üQÔí8ct‚ɚ߀HÏ\¼Ô6.fñÃNDÞs›Ùfc‹ìFÞóO@$´^Ѓ†—g::ƒ¸ÜÅMeÝ2t¼ÿ@ÿ£ïúWhÕL{ðüßpuȈ©öó'J¢¤T`‰ßK”¥íéèyh- @ëªÆø¦iÊndÃ:TT®$‰ Çü¾„~ÕXœ€× æ "a˜¿‚„U$ú+ ‚æÇ T,—Y×·… 1C…àb©Z§ÿ¬Ç ƒv¼F1~rÚ4–®U$n•®…Ð`¶r#soàénLGÇÞðPv¿"ê—®!ï¶tãVS¥£(¹«-,Æ)$Dlò룠4 P0— J“ †ôMЀOU‹‚·áÑîºË&¶ØmDÔ__¹šY‡®»8§ÑëB‹í´Þ&à$Ä„ÞV! ŸòmÔ¶rOΦ’Œ kÇN[;7.´,âîAÆ*ÇäõØ✞+ÚÒSî%áÜ8_̱VÜŒUAgªZOÿS‘ÑNGav6U.\¥1P>ÂËGž)ˆ±ÄüxÏ:Î=!§³Zfe.Í%#ÆOFo`óJxswÊ0•©þâ"ãVõ|ï r-ÖˆR¼ÿA*Ö·æ*Rb½q¼‚É%€,÷ùÖqtì lsMtô­9xÈI² L¡  2¸kÇý?וu²¹ „P{<h ÿÂç9‰Û.ÏÍÁ?£JcŒ|žl$p®_1ÄÏæ©á¯Üÿ+1cu‹¥„l§ïÚ±2öòõÖ˨¦"ø_1·a‘ß• ˆ`óžîˆ”Üà»·±"9c"ª#Às–r³®Ÿñeù‰ëÉ@çòÓúY§w–ŸþÕZt‚â·§È5yxNžëpäý¹Tu{QíŠúš°C©¶‰ã`á#gù¼úá›n¦äf ÊžD !СjÊ~5òŠ2ƒ`'[ª°r_»Ê0^B&6"¡ršŒ—”IÔ^ Õ•_¦ÎÂÀãU‘§ÑÖ*0=†­ÖÅÔ®qbªïÜiæg˜t ›DÕ|ÏàÈÛÆ‹âŸØZÃ?I²ª¿üP÷íSš@”œTŒw@†·6v½ô¸E\„ϺêmÔmû`øàúQKVJb‡4s-®C|~®*ÉGÇl©ô†‹2£Líú¹˜éôßS „Yâ;·ˆ8UË÷4hrÃ9Ö^{ê¼Í.tÞTuMãÀ ‰™†V5 zsQÙ´ÿÞÜ/ôÆL誮ʩ̉å¸itr÷Ø d“þ[ëâ@WU.ñ1.ññ/œ݆ì^Å]†ètÏimåij(Å˶¤ÐÝ…¿tàLÛ¢""¥~LfcËtÄ›Á‹©¢7) pH£Ð±² ïàf¨A†„-¬àòhA[4#/‹úÉÖåoßP¯½UìidQÅJ4ƒ4µË¸ú)ñ´P*çâñ Ô³îËèËlÀjËÜ£ì '_/NöÄ~Ìuˆ$gßé:ê6~Ïf—?I[]"pΙá` ¼"rQ¾ÃKå} À+`^cK`vБËb¬”x£ ¹6!Š93†/(ò?€â‚ 8BÐè}M3>žQgú<†®Èüåšç!ž’ù»ÉÇðSݦ•1,78ßÙ/xµäë˨µ0'=ú–Ïeâå(ÖŽƒž‡ˆ$™Xí ìÞç"ŸïYæ /Ÿ†;2óm×#Ä—¸G'2r—{4‚ågŒO˜KÆ<:IÅ,¸Pp2uÅ}¿ò<|e 3~˜(=iœƒ´Á«’T2“™¿y!><à“Ãã× °U.ó¯©nݽ~Êæ[Ïh>V±!.½°;Úóàªplj.ÍßízÛ¶·mÝaÿ ËãBß'Ò,e2Í-SÝz$]„dª˜Y|¥ýËüÙ•Ñ«í±m/¯rŽK[¦,µ4¿NØ/ßlƒ€ KBþžçÖoB.´¯ú]Éu¾[Û»1¾ ]UrŸP„L?ÈÐø" ¡xßwØê62rÔæ¤~Ð)¤,ãÙý2Pø2}"uÿ´Ã`AaÜ…0<Çéöe<{8‰ÃøÕ[–º*Õ$>÷òÅ6~˨}Ÿ|ðK÷Y|.þŒ3šØóYwÛ¬U•&ÍRûo[„`a©)PÁzn:ócïItXÃøChûq endstream endobj 841 0 obj << /Type /ObjStm /N 100 /First 885 /Length 2500 /Filter /FlateDecode >> stream xÚíZ[sZÉ~çWÌãæAÃLÏ­;åÚª]+Úr%ÞrÙr9‰ãJ@6 ŽÖv~}¾>ŒdBH ‹<ä…s›3ÓÓ—¯¿î‡bœáàŒÇh‚ó8Š‰Ä8&“R4ÉÅ™LáhH²arŠW³¾ž8×á®#ï]ÂI6ž&Š'Egã¦ãä±dÐ1¬kcÑ”p"¸“ðz†(œŠñL¾Ã)r&6äq…©ˆ„!NB+6žÈ<Š^O2N’.…1QåÊx”„pG •€‰3q)í6‚#<Ê oBœ‚cðb‚n4@¶@X%¨|ٛЊžñ8};›U¹˜ÀŽ:Œ‚ì*CHÁm.ÙDNÈDŸ B¨ZŽFŸDÝT£î!`HÂRĤ{À°˜Y01Nt.ælR+qa“|»‚˜Ùñ¨˜„qâ`?HÀŒ1šaö&©¦™É$):³˜ì}î@ ™tk°f&Ñi‚ÉzÅPYŽº¦D““ÚJ¼Á~‹š]=Cp'™Ìº+8GˆË˜¢ÀÖqd Ã0©A½ª!8BÑiX'Qçã‚=ŠÃ0nEÕà ­ÀÔLWóꡉ0±ºOÑ1j8ò£j…QÛÕ¼® ‚´ë*¬cu88~Œz·3°@€'JôÜŸpW˜\2“‚ ‰®/P­x¸­sЫÀI¼S/‚¿º œ ¯“ŽË0¡@>ïJIgÏ:ÝÓ¯CÓ}Õû8ìtŸO'ÍpÒ̱+F4¾ît_çÓËY8oc«½õr8õ~ž~1ïndì¸}è`ŠÞ5ŒÍ¶ã~šL¦˜ê}Ô¸Óõâ(õ˜Ǹxþ¡óãK"µïwº?Ogƒál±˜û õó>úl,“3[Ø:ࢹè%F½¹üGÓÎò§Ñä_˜Ë<{fºoL÷íë‹Ÿ>5ÍÅï»ÝþÑYï߶?=ïöf_/šYwður~9ŒÎqm?5çã߈u›d7ÖØ$h(Ř˜žï½cŠ…z8iˆl—{®xlÞs›‹^›îŸÿòWÄ+0ÙÉåxüáŽ19eëcÚ:ÎS„JVÆ îZÑNãOàÖes¸€o—«û¢°Ö^àõî«Ù´ÿfˆ#ˆOL÷tø¥Y ¥ÕèŽi-º$÷ŠîXV£Þ½÷2Q²[EL؀Ľ@GH`BjÀ; §¾ú¶GÍh8_µbŒw[q›ÖwÕnrëÚåiwyr÷&VÜ&Z^-Å] Ÿ†Wr´8†zäÃ9¨ŠuÈéÁ{€Cn"1ÓCÀüóçÏvz1œÍ›Î>vÿÙô}wÞ'ê¾û…|w0íÏ»½‹álÞ£îÄf{183íó_õr#ülÄð? "ÿë@'ƒZ„ý‘ü~™Ju¯dùAҬɀ;ß/&ó¦7÷šÑt¢qØ› ôðürÞLÏGÿiïïåÀpvù:yy¨¸øô(ix¥´éwÚj=Xë¶q ¼EiÚ¶q((¬€âmǯåD±œ@” ©‹ UdÜ ãrX’ìw’Ês凹òÅœöÍ,|š[öå@‚ì+Z áp>vZr`"Qi Ya:@¨Y¬e\lAPcFØËíõÍ‹—Ç­$ãÑÇÉ9,º$ÖÙYóùïçwÚß[K1ˆ­¿Žü—XK& ²LÛ{e ªYJØ:Η`}¹Íû‹wP¾oÁ»kÀ–uR’y׀͹hÙ7@[¬ŒÖ£Ú!Yíd/·ú×áç£ÞlÖûª–ý×aÿ²êéÉ夯¡0ßÏíP·D4IЂ|”*bl* N?͆½ÖÑæ½³aóuÍÇä0$´”5{—´«½Kez%ìMõƒ³^)êw P| NÐæ––l·#ÎÛùhòQ5üjÜÛ×ÂTñÅpö†¶qe“ÖŠÀäoçϧçãá•àå帀1s¸[o¬7ON×j’·CËʘLÁ¨dÛ¸D D~Ÿ 2šây½š¼‰EÚ,uÔNà"a?üáõ’“ïYrŠ[óÇJ¸âPáz¬D‚kE•Hp}ξ«?³ìëÏ1ÀÛ$[ª¤9øØÒ¼p ÁMúé ÑN¦³ó^³wêÌ)QQOkŸIÊø•$/à®”µÛkSÑ~>Ô“´é›md>í‰$ ®¦CÑ~µÒž(–7H¤Ø3ÎZ¥Œ{çûóBº#6~£Ý—²E=ÁâAÄj':$ÚT‹mGòí.òîS¯%Tßšƒ0M늌ÈÀû+#d„‘K†|(  >JŠÉFAù *B‡\ÕùïÖ³ ÈFÒöIźèkŸ÷6“ÌuÃÓ3ý]Hpô¦÷ÛãåI¥°PÒKA˜&Èÿy4­qÜž"WÆÔ(ÈC{j…j–À(P;[|\ê-ëµ²ø¥¾%íÀ'7ªãFWàz×;nTœ_ߨìšã¥~’šã«˜R9 ïÍEm~Þ¶›YêpäðŽš„ŸŒ‹Ò5Rõ5%¤kÚ~å¢;$F‡ÕË‹hf‹EÅYÚðÉHá÷hÐkzuùïAh|Ì`Tšr¨ÏÀÖS±®„ÃÈC`U%‘ôƒ9p1Á‹¸X$ÑÇjLêçìmè¸:š±á²{ap'–¡alî ² úU|7 Ðïó-0xW"Ök_¯S½¦z÷î*¶ RÙ×®¢hP8Xˆ$†;Â2^~¨Oøc¡ÃÈsE®pCÿf¡ J‹{OC®‚´ÅZÉuByæ7´X™\Ãj Ö~E¹£þz6‹¼‘®:©U|8ˆ8ú©×!…{Àex2Ù˜Ôk¢É#ˆÔâ"ß;yåK»÷ÁÎÛ?¤ÜªK,s¹órS=^®ÝŠ]AÕ¯uüôïC;‚ª´NŸ6QµytÕé#hÆü@v¥&óÛ{j«c’[|Ôx¨Y7˜lÙܛ̷Év5+­5΄ÂÎf­¹‘r͉\RïW’M5—R}N5—ÒÕ8úãì» tE45µ{­S*ÓÔÔîÊág`™­4l¹ÔÏ…ú¯»²™k4Ó#=ª$§3Dö´£MÐI‹°+8¸ØŸêÉDœë¶¢Ó?¯ðB=>€’•§a>ú”°ÜVt(Ýž¦­ˆ`ñ¬MŽ–â‚5MÚ—jü¢:¢9 endstream endobj 936 0 obj << /Length 2685 /Filter /FlateDecode >> stream xÚÕ]sÛ¸ñ=¿Bs/¥î,$ØÎ=´Éù&N¯“¸w¹› -A6'é!©8i§ÿ½ûR$ EŽ“>tü@`,û½+ËE r‘Ç‹,ID®íb½´¹YðàÕOϤ߷‚«Ñο\=ûþÒ˜…ŒEçrqµ£ºÚ,ÞDÏo‹»Î5ËU’$‘þãr¥µ‰./¯~eÈ+·u«Ön¹RÆÚ4Jäò÷«¿>ûñj¸Ô(õHêpçòR=!Ï¢Y¤V ™h&ñ·Xfx)<%ï•J¤ð¸˜w!Ío_ýøâòJÆ~÷äáY"²\ö»×õþîйޛꨨðk¢þ8»¦¨ÚmÝì/`žgQ)œX®l’Eoxñüjõøb£— ø-6q[ï]WÖÅnç6H¾œ(XÉDHãŸìnÝ7Ìl@(¤Àcû[}SŠàëãT¤iÖ?èï?¨o«Ð»•i’öÛ.ø¦²ZÙï—*‹\Ó:lC÷ ÚÚ‡\Žeè¶ÔŠL · |o›Šy¥²Xè ÔAf"–O“s˜ಖ: ç$f9«<êp"g£½œAMIΰa"g/m@2—6ìœIé8!íXE/ï¯ ¯+«žÃ& r²½œXoù;¢«W€žQ¥3¡âô jÛ€Ú aµQÂ&ò¡Ð~8aZïÏë–Ôô&‰…LŸ¨8Rž A*PœÜ°â佃<ég)£ýr%‘I&1^rЂ÷‘œ~<9#ž$Ö"íãÄ“ÅC*§ãLp¬rñâ5ÈJðæs–d‘3ž¨¾“GfL)5Z‡„–æ^jò«ˆí4Óž$¶8šÜNâ8àŸof(–An‹“‘©‘–ÒË-M$É mìõà CÎJÏ mõçûAº¾Z±ð’ÝÐTv’Zb„Ü€Ô¾ÌØd.2PÉ/’Ú‰°ªDšOžLžc“¼01¶LõÆÆ n í(ºô†— ¥íB1±ú+Êî C¡µú"O)OÊ.Ù’4óÒ›œä…©ôú´üK/céý"øø#D&óä)"33‘™iNõEûÌЖÁ6-ªÁG¦)¥"2*Ú ²&¥*ú3f›÷ËÒH(Ÿ6ü”—Õ|Ô[ÛJªÀÝP¿ æ#© I)Ê ÀhRÆø ƒ9Tô( ÔîîvnïxÅçœ×E[®Ož½/»[mëCãÊÝóàR]QUÎ/4õEOͨ++×¢¯Œut×Ô˜/¾/7%œ¦ 7÷n;ö˜a°s,HäÚŽA”æÂlÀ¸åu¹+;ÜóŠTcà‹“úÀøÖ:âèn×üƒóÆëOv] œòû›¦ ü-ïlËàÛúƒEåêC»û¸„¬UæQ}Çy|StˆS¦Q]ñ·ª«Õºfn—78Èpª!º¦Ü¸ ©7EW F…ˆaãWÀ´àO{¸n]ç¯ÚN–vEsb?`J¹Ú@áRµe]Èq­ŽÀ7a¾šÉègЛ†—»[b-žÙlÊn„cëŠîÐôð¬u:a¹€üY¼ʘW-ƒÊ#öãÛbO‡Œ¶E{[kaR´C©Ds_;ÉAÏc9Õs_Ë0kÇ€jä ¨hȃ¥PxHÕ{…í¶»Ko åyjŸ~gÙ2öCë65ã‹Õèe÷H§—»<Î…‚ˆ¤5ÎÍ1RJa³|‘¤’RlÆ€gZÄRÈSue¦q{¦?¿!tæ×I¯Ýºc‘`I(öÆoþÙrM À³Úß}ÿ᮸ñM¥æÏP bccüŒÓ=¥~÷¢çH{>BÕ<¸vÈKsLšÐµ[brj×.䧜ûs¬;÷']Úp$–B|Ú··Žª´2ʧo÷EõñífÛ¡¸Ëªãeðï.x¥õÐaùÛêb¹­ïW2MElgÅݸî‰',eõ©;À»ï¯Ýæ"à]†M%»µ95Úî"dÞ*êC÷)2êrFG¦£~HÇì`[ÞôL8T8A1“€vÅM‹ªó§íCøëGƒåÚÔçàW·èTB!£#N$9EŸÆ£ dâ§qeàNPÑ.°;«Ø¥!^”´æ>t®Úø¼[}A~š)ûÀ±½õÈ/y:LÞ`~Á#šh{Òce±È¬Y( Ř²Ÿï±üùÕAÈcÍ®™y, ²:´m¾9š1̻䴀½ÞiTéùK)¦É&/9í´úÝgèž#œSÑÕüõ }H¸´òc›È»‚ìtJ;(Ám´½X!í, ¹b*Šé¢RÑ-…­÷ìïaŽŽ*D’Õ”-÷ZrÜ5Õ¥^]B¹hÈ]Ñ@æi]®‘¡ž>ÉŸR~ö.£¶Ü÷—R&´ÅXÊ°G7˜Ç¿ƒ o(Þ‡åY ò 2cϼ2AòŸ³„pˆ Q“€Ï;ÚÌÿöÕöì«¿ê£Y·O>Zû 9VhþÁ‹©;›lj®ÿ?œÀ,'è^ÇîPjHRк¡»¼†\aŽû+†R ×f-î*O b‚ AÈ*óÙÁ½?¿!eö³k¦ÁéóÁr î¸Ðw•Æ>¸ãž‡Á]g†~m¿ätpïwŸ¡{ŽtÜ‘6j›`<ݸÖÇ`Û㩢ϲðg*(:Z߀X—Å"冡äÀu–R°õ;%¦¶ë¿aq!ñ I ·‹ê=™¾àÑÔ˜›è‚nÑŽÉÁîÇîãI K¿q•kŠÝŠ²*×EëĉPB µ=ïÍ¡PKÔ¤ü#çè¢ù@ɵ¬ /pgÈŽ’?¯ùëÓ,úÝ!gcAð±k@Ó¾ÿd}Ö¡ëLèÌ>"6a+J ÿ5 ægÔf¹¿uë»'>u!c½C1â^§å©*@¦Ø˜ϻ³ñkEmB=M?ñûBþÞ]ML^d¡›ñÿA²c»¤úîÝ·å)¿•Ñ‘‹b¾ã9_í“’A+½Ù¤9ºq&R4á/÷k48úÝßWOøÃ~‰-SÔ`eG,$¸4·çŸ¦b#L:d `ªð¶úäÛ’‘?Î’˜;&ž’ë®(©Å‹Kÿo Ý-“d±o”õ}#¬¹3F‚E&úSo¡.Ä$5g¬6Öø`ì»}0Âÿjði}ÖÿkŒfT<Âÿ A˜ê_,«M Þì@B5ú}¡¥l\y‹B’(Úxü_ÜwY endstream endobj 941 0 obj << /Length 2267 /Filter /FlateDecode >> stream xÚíYYã¸~ï_áG¹Ó戗ŽiìCì ›dÑÁ"˜ ²%·5cKIÞÞÎbþûV±H]M·dä!ð)Š*ÖñÕEóE?¾HÃE,%KU²ØnB³Ú<.hòÃo¸Ý·‚«ÑÎ?<ܼy§õ‚‡, S¾xØŽI=ä‹÷Ë^ÈX°(Œ€ŠY}÷îáÇåJJH?ÜróíCŽâJ†pçKŽ¢GQ¢—ŠÎÞn»§ŸŽû¬ú)ßvxò\„8fz`µ¬–" º¢Ùf›b¹Ilê¦)Ú#>u•·´ÚÕ–ØätÎ5þµ]ýtȪço¸ï`­˜N·õŸ¡ápCûiWne k-+Ý®ðšÄ,ÚQÊËÖ+¨Ð,¸ÛuÌšìP€ Vž¬1¤+q¦…\¬¸d\[­†•bq¢ÜƮɪv[7‡–NÙe-ëK|òš#LÈžìüäÕL hí–U9Жÿò  iO´òÚ Ò‹xä´2Ë8ZGT™¬³ìmd™À-ë:øSE†‰f‰J­a¬æ³íóAœ8pARdÂÕÔ‹}q((J,1íÈD€mK³nÁÖÌÐ)q,>Ÿ²½}é7“\]„Mz6ñW€MòU`Eýa§5·¿¢æ²¶-«Go֨Že¾ÿûwß >aÜs°uZÎ*¯Ä€.¯`7aZŠ™ŸH—­…D CË$>Ÿ 0èR"#Æè/ÈÔ0J uV^r§òU´&ÂIÄ0é)8ð9æPÐqÇ…àrƒXÃMZÂWûúqXÍKp é#œ€`µ´‹*+IlÎŵiÎŇÂQ ¶“JçÁ_Ó9†gÛ5e^œ‹7JÇ5˜ÒÈI²t=AXtîÁ{.‡¬!‡zóãÅ€»êvDpììf"šU‘ÒÂæ]cG¬:z=IÅÈ¼Å‘àŠ…³Ì³7p P6¬&´B¹|îÇY4„ã·åyMjpÕ!nƒ>íH¨°Å@®~…¥J°„ ùax¤ÈI92¦©Œ×‡±í;Ì™ÆUQ2¸õ´Ç‚©\!.‘ª†d‰UNe'ViN•H÷Ο÷¡ŒR¯×ÎéXز%6,nX²Hó©{×M^VYóŒÅ£ *©îŠlÝëÈl•¿o¬"?^nj“5×PÙä´òTŒ…êhE„…l¯ìOW‚ å¤aÌòùNbO§W¶jèÃ,TëãÓ[˜Œ§zÛ6 5äAFU‚ÑUP&ò`%^¶´åPÓ4…ôÙ¹YAÄ»:aa¨/zöõ9ñ²žÑ¾F×ÈØ$@à‚“Çïº &ù ·äj¿ût{®…nfTø˜°Ÿm<8ÕPŒöu˜‡rÁtøÎX³šV½¯ôþ“•Vgz)Õ¨÷m齃ýð6#£ÁFóÊ´A°ÆG‰Äž2É#£íÖ ¸4gB`]†bJö¥ëÌ0›Î &Cgv&r¤ƒ³½’š&B~©ÿà ‚ºxÓ±xVã,‡ìº¢ç®(%Á]¬’@Kö'>ב{Rn ;,Oï¬=Lß /ÐÖ•)làá@;=4Ewj*W3ÁÅÅÂLLáJ¹N„Œ >5íï‡ÄèóÈÚ¼äß¹<ðÎB©kzÊMz ™ê³–MakrEg áRDéЛAUÝÑäÔš5ÚÖ™ÒFCœmOÕ/,è‰ÌÒš Zÿ0Nq¢W›Ìîi³-¦\ªÉ"h€Z³h\Xû‡@4Å©µK€ƒmgu<–c~…õKv8B£s¦5“^7o!T)ü¥².?¾ã²µ{¯éÙÞ˜yDo†¼oV75 òñTŸl¨ÁgÀ\êæù­ÏrÁtV¥–C M]§™}Cƒ¸ŸÜÀ̶Wï?Lvÿªïh}¹÷ÝÜôºô:þ˜¿þIi¢ÇøƒúÅJxOã›[O¡ çëb“+Ÿ=Ÿ†Û7¯Iì*Ì ž5>ãM÷ìHÐæëMgÖ+Ét^fæ:¹u¾<>º²6¸­}/ï½·?€Ø4™Å¥ô®V¡Ä3X…±Û‘ A²Ñ´²v±æ ÃF’HŒ¯úàÃè…O záU©nͧm‡÷Éô’6E½žpœ51Ä£³¬2YÀë>éÿ'\Bþç.Q½?ÜVïù‡ Ý ¡eÏäDœ ¼ö‚¹á©2ŒÅ_ÝSQT3RÛ²qÜPuØùÄAPûlä6¦|qû> stream xÚÍXëã¶ÿ~…‘OrpbùÅ-pm²Á‡H]A,´¶´VΖ’œíõ¯ï I½|´­»MŠÂDó1Îó7ÃV~leèJ AŒLWÛã+jg›Ç•|ÿí+æ÷Å°1žìüËæÕî”Z1J 5lµ)¦¤6»ÕÑ_÷ÙS—7ëXɯֱ”*º»ÛüËÍ|Ÿy“WÛ|s•¦I$ÄúçÍß^}³.Uœ/äw~Äžö¡ÀŸg/I%aB:ö$‘„¯c•ðèÍ:fÑî×u"¢ ÚƒZƒÙ!Þe]æþ~}·i‘Axv2}6OIºŠ™&‰ŽrQtÏ÷O‡¬ê\ïYõá~Wt÷ ßþD-«Îmi²êýk7ÜÖUëg‡å/«×g3ûú‰½F†V,M‰1˜„)æ˜ØÕ§‡CîÏ—Õ5êe•ò#Fçd†]eÛ5å.?g¤ÜÁÂüäDÛúøtÈÿíï©OÝ56ê%lÔa6êÙð'OU[>V¨S«‹CöØ‚èÙíÞx.º¥êÛòæ÷V_@|ÿS%†¬èsô÷{ªï<żP[I’ôÞ•ï×<ò ïJ¢R "sÅšÝÂ(Æ”þ ¼Druû¼1IšDÝšë诬ÝDqª¶] s³Ý.Ä ãAøÀ‹·’ ˆ)5°èÔ¢$$‰¸I)íw8å]¤$Ï„’UÁ—rPθ5T¦¤Æ¾sfOY“sÈ* LaXÔÕþ»ê×P"M@¿>‡ÕË@º?sé Œ¤òaˆ V¢*z8uî)5¼¥y.ÛÜE(ÁØh2ÃäL0é}†fö«3oHªh‡kÖ[”KµYÛ¯ú‰‡¬-·nXVh¯ ÑNf[Ì„9üFQRöëüW]¼œ]SdÛܳ’µMJEMgÖ°ËÛmSzòx™¦Hã¬ÒaÒØä Œ¶²Æj‡(X³žO·eÏ|fZÿôçã @‰v~åðùÖq` ÌžH²èÇI„KP¹Á›Ñ ñïÝ:ƒmŽY÷³µTÿ˜»5žž¿TBdlöÐËeß|ãUg$½µCˆ´¬Þá5e뙳ñysÚ…¥‰ a¾uWÎëÄw¡\tƒ¶ü¡â7kÛÓѸ¯vßÁ¶æ‡]z¶‡Â±S%G· 9†×Ó •$€ìs h`~ç!kýâ®<æ­QíøwÛjê¤ÇýÁ¿Õ™*BkC¸s¹ì³C)† 4ÙÈmh¹Í¹Í$rÃ> yä¸ÍËNZ9âöAŽH®v“½ƒ! ò8W…e aª~GY8šŽ¼Š¸lãèæ-#0 5O\1$iëí’¢©‡c3d£®r&™L¤èvz5AÐ÷‰Cöj‚©©šàïOLÈ'( òݽgÈnú \÷ȯ¬ÜæÈný¿ýÅê^‰þ IM(ø ç\@¢««8tG QÊEi`ZyqÅ]<Ÿ\ñqÜ¥ÁxŒã§½uiìh£8ãÌGuሹÛºiòöÉá«jg“ö£ú x¡M¨£NÇ€ý’:ŒÂD4µ-ã#„U¥¢_¬ÓP°´f—7~ï/'4Üm¡;¬^À“L'Ĥ/”)ÁRÞVî¢no% Wo³6÷üäPÓ{·05Jd<ˆUtªÏ êUðŽP±Ý»Ô, †_q{Hg€gúCþSðµü~H¶ôß ˜ÞæöTîëÜ‹…“[€›Â¦PNQ¢’‡ L:8ç˜&Ñ“Cï,™"žÛ±{"¤¯ ¤]™oozƒ ¦pÝ _’i‚1ºpyYÑgeDYu¥S@ÍÀW}…¦$<1³úL€çUuç§ÖÊFYáZÒ©ð–Ž£¾wÿ nœšª-*#¢,º9¹¥mæ÷´Y‘>ô€“DžUèE“ûˆlµ !¸ÉOm>ú#qï˜÷¨!0ƒödßHƵ¸Ù¢îêØçY˜Ù¬Œúp«_-—4<›ß¶Ù™`oþ¿zÕ/îr¾¸K=¹ÊŠ„~ÿ¾¬|ÛóKzbŸÔeFBT5ŸÐžFÌù™ Qn‹w›FÀ"çb‘Ÿ°´L“*ÕoÔ)bJ,l€Êå”>©ç”\é9i.ôœÌˆh- Š}‰‚‚:ƒâ}O”ú²ƒòK¥ Wk¸ w-µ&Zšó¦¥…©/5઱giÿ†Û“´Ï£t(~(÷‡K}Š¥|žb}zLŒ13ÿu­I¥Â­I˜·xNaA_·~+a‰,øƒÿÓβ]oòþŒø&ŸJœ„™  L`»0}&»ØäÃXü÷¢G¥§Æb4ÂÔ­±S­0Âwu«ŸG £Kº¥#˜É¡xã!ueû”ϽWîë[`ƒ‡¯PA™Ïhà¡Œ„\tìe~b[ £ï‹—Ò/G^†/E^ì&)±p8IéÇF`o“Ú…ð(Tàtà’ÁéÁ%‚Ó3—L\BåÁ%Á¹8bÀ©>w.ë,HëœokDpØÚú/qÆ endstream endobj 950 0 obj << /Length 2727 /Filter /FlateDecode >> stream xÚYKã6¾Ï¯hÌIƶI¤^;§Ý`'É^lØÃd¨mº-Ä–¼¢Ô=ÎbÿûÖ‹z™ž hñQ,«ŠU_ÑñCñC=äJ…¥.vçwv/Üø×ïb¡ÛávFù÷§wß}LÓ‡8 ˨ŒžsVOû‡OÒ›ÏOÿ|PyfQ\hôãǧo¶J©@…ð‡$ïþñ4î“&É7 „”7å ‘•Î‘(+t+Í{ë0ÝlÓ, ~º„ÈÓà§f£â 7Ý¡Úf}*8v\†q"‡z:šÍ6ÉãàWàúlÞs·n6I1ò¡$è[žâ‘]7{³ç¦Móå‚ µÂ¤’%g\1ìðÿ‘Gx& Ü[?ŸdQ{®£°Qœš/õs}ª{dqå@ áU§|BZ¤Ò8@¼ît­›î‹åÒ(¨:$ŽuovýЙp³ÕeüÔ3au:µ¸Ë›å~ÛG<%~wíù2ô2È';õõv_ŸMc붩N<…:~Åiئíìû…´1‹ûk”F'Ô^´ ½„É4é¿wØë, ú®jì¡í6(Ä6ì#Ÿãíh:ëMåTŽ½W<5ÉòÝr-rÙ3í탪ñ‰]ƒr/$5zû: lý‡áVÕìe¨ïê=j9ÊàçF¦wÀ™éN¶åÖ`eîÜvÒz1éX¡))þd¾li¿TóÌBtL':W=*Beq`—Ð'¼5—ª«È€p;Cì•a¨>W/uSuWîV]W‘ÛÙ‰]¦¥D„M¡‚("ð>r}lÂa:ËÍ·cË Ô·A[ýg€VÝ·JGÇat†Ý h-..ž»L‚º—9Ù¾>³©Ú®¯FS»jl9Á®>…]*' hÃ"ßZ³«GÝõ½¸¹ÒuÁ/I 4{ŠÁ ÎkeU5­;Û+¯¼ÔymÛsfÎÖœøvY«e3N3‡ƒ\­>äàˆV…DG Ž ñ<ãks„ãêÉTØzÅí0貟@Ã^N¤kCQ øȽ(œ‡™ ä…µvvÄÖ*ã.Înp²Þòh—ø ‰[èÌ¥3v®F¡¤_¹5Üñ\ŠY^Zs\dJÚCDÙAL”íÆtGéQ]ÐâH#·i¼V£#€ÞÐýÅ–|i—W+žÅê,çêL’°P¥KÉ÷6÷Ø!ÍÂ(s9oŒöãܦ0ë‘Œ0`«³á!Žÿ07zÏ­gÜâÊíqÅseA´D<rÉþu“fÐßñÒdihŠTE<уNP Û¼‹,]ê%ŒŠ`ªc¯Þ g4{>!·¹D¹$Æ9.‰sF™zÐ¥s`ŒØ$Laîü‹)Ë‘9Zqë·3 µÞ†ÿR_n^‡ kŒ× {Ÿ¾w¾‰bÓYV¾ù5L¨^ ¯Iâõá¢Åeº8Ü}àå¨ÿä(k¦tD *Ãè@³éA& ìÿ0Én&ç¸âÛ­œ"Ê"Øv„>^!eÂ’$ $Œ{X¦ðŠß£ÐI"ÃU×s‹2|+üäS(À1O(xd2vÇ9ïyF\l@(ËöféÿÞ—B\2$•á(¡Q…ó ¡‚Æ|‘q6€$ûVÚQ ó¥ã{!C…€¥ÝÍçÀí IQ¼Z— Tˆý D`OPlÍבl„Ã=Åè‚…l#â‡:ÆÆLµÌ ÂA× Žá%Ú?Î~ä=º u±ržª (ñª' £Â¦Æ¸N¤ŒŒYù:ÿŠCÃä7‘Ðòì® ðB¸‰­wã:!½dà@hT'Ê5‚±x * ÈjÒq„£T[Ÿ‹ÝˆštC}ºp%ìgð.gËMº$ð­øã²:ŽŒê¤î[ @‰'.pM/]Í(2K¼—4s‰™¶‚Iüì„8ÊòZ'±j3rÕP©W¶É§'¥mC„–¿¶F¯é˜€®* ‚Mˆ±ÈˆSõjxÛã61¸Â楫  ï¨l‚³ühšcE5–qOP¦·‚±ãU‚Þ* RU8d©ÀM¬ùVô–¬ëcÆmXò¬ ›+€Î‚ä jùf䦠È^k䆓2ÔI<vQa‚¡,rXÆóÆ}ÕW< Î2Peú("e‹UB*’1Hú·ßêê7/âÉC(©-"½X9à‹`m þAÑE¥h?½²ãhƒ«HqËòI|Ý!ðŽgÊî·…2Æ|g »BŒSÅjÿêUÀ¤Ex믳7pŸÓ$5rû¿t* ÏT¶.e…¤ùà¯teº¶¼ÈW¦[™ÞúöøÓL¦úà³Õöæ8?š{>4阧·2ŒÔ‘ ;ºä Uªt¢à7M©é!à‘á Ã#AœB±TÌwðÈvJsGs‡S…E^8¢ö[uf}žÑåG"7Ç3a¹q&ˆÜ@“JXŒhŒJ“ C¶¢FóݸÊeèçˆíÐOƒ³‡‚mZôºÃ!fYFnsid8Á8Ùø»ƒrŽäÞ =JÌ7‚çÌÉœ¥îvÕ©¥ç.âÌ&YûóhÞ›üŠÒë«B…Ñ#Æ~zš+Õ”ŒèP\~Pn€CœMï’Íì‘ÈòˆË'½Û…ß9.z6Qó4Ìk9íÅêÎ0ÿzA¨$’´X8)H-â\?!äŠÅ4Ø@ Ï‚Wî¢éèÛP ¦©Wúìyb‡mjo=y'±C­Ç„ŠëM7ô âF­sC¼‚Tø`ÒÔ„ c†,«½x ÎÏ8æ4°•ÃÍcz¥ï"êPgñ"­ÌðæZiYæS)¾„x ÏH½ÛhÖ_BúÃ.¹~¦&˜¥–§ÀºŽð%ųŠO¼|ºé:©;)‡ðàaíûÃ-{©»AKE¾ôš¹J!EÆ÷õWÀâ1×îÛ_´=ªS:„Ý—ªóÅ<£‚ŸŽüBÏ9ÍôÊ1rËåÅilmûéa6xÍÊÏñ‰r  Í5~åѪƒEpŸù]#b]æÓ+жÌB­jï¬Á©j^¨Úéu= þF¾üZeQ.ð¶tÃ{Ó´=Ó`­ê¨,¡á}YŸTÔèÞÊL“11ºI{*Ôeù•K’Þ\’|†¾\zíÆâØÏt˜¨Ñ=.‰½XACJ,âbq±—zÀP‘H Š_Šg*’‡¦—AèxžÀt¹J‚_ø¦"(§2û†£‰°ãL„»ù*c–ÊstUDašÇÞ+”Þ\!=‡µ©r¿‘jJµòJ~ݤßB¬bUÁÓãùôæ?ó| oåÅK~»|±pEQ`ñ¹žÆSW±W{À Âçsyx}FÐE\Œº–ZB$t“±5@µLô¢ \À££|õêÈü¦‹ÏXÕï‚*°Wñì½ —*ÌÊÑ=¹ÿî}HI°9Èà ʪ ö,±ŸÙç?¯bäW2õRÕ‹ )a×ÕÏT´{šÃ‰©ÐtrNVe X„|HpNX 𹧆 ØdãùŽíÛ¹j®¿ÝÓÆ2YÞÑF©³5?ÒÊœùg?8. ]¬Ô“ŒêÖB=‰Sž£”*ë½{Zý?» endstream endobj 955 0 obj << /Length 2696 /Filter /FlateDecode >> stream xÚÕZKs䶾ï¯Ð-”WoЮ=¬×‘ËŽ+N)Jù om¸3œR&9–וŸn4ø¹|Jé06 ô×_7Å/RøãYza¥d™rëëÔ6» jÜ|óŠ‡y+˜¸šÌüêöÕç×Z_ð”eiÆ/n·SQ·›‹»äÝ}þÐÍåJJ™¨/.WJéäúúöG¹)¶ESTëâr%´s&‘úòýíw¯þz;,ª…x¡v8ó¼zÆ)Æ¥"õÚòwXZ¥"ù)ÕiŽM™.…KŽû®\mÊCQµe]å{šµ¯AQ—ÔÔ­·ôÛ5yÕnëæІ~M¿~vÑà#X€_Á°H“Ç{Ø4îÔ3Sõ$W,å¶ãµƒåÛ0m¶ a™Èl?+¯61Y‚³Ì ¢îëÇC^}ú°$r%dÂÊ‹—Œë ´ôÇ”7M~)lòé’'-)ÜzdQ¢œîÝn»Çe kÆ–Ô° +ú¹ §Àê"eV[hdL KÏ®/Lj¼DÚ&Åoùáa_€)A夻Î83ödó1%Ü7L|È›üPÐ}…¥ÊŠ~qßÈ7¿^j _YרÌo«¯¯o‡wàÈAÄ6_û·Ìiš¢Å»`“ºÚ´ýNË´ ç¤\œÈnàVh#NîÜb[²’éŒ÷³ß Ë™®^ìŽX–¦6vGîÒ÷¬ŠIç©`BØQ|L_ɤ~‰ñS–fý¼¨‚Ž‰T,éW¶q-ØU¿\ÁrS¶]ÔÍ3ÒÌô»ÿ…c³óS[ò@ÁŒ]´n½ ½c<ç´×nЫ~™úìrå”õ0wëÝ î…qþþ;Ä@×›ÆÚ²Úí jÈþ¦S€ðw /ç½øóTðCÇ6ê+·ÛSx9{‘“#‚½!(J)E˜@XÎ[ÐꤩQÑÇÕö¤Mò3¢‡²Û­Çö9´Û)ÆùîcÙÝ“¨af=dwtt·êwó>zÛ5Ãqà9´Eqì”0µuFwÓðÌ; ãÂG¾ùÈ4öï¯&"aðt¯Ò?‹u»¤MC0`œšw7Á–dt)Ó“œZY¸ "DsÈ»÷poD–²î zÆŸDx‘r0¡€€1-ì™(ßÏ^M¦G¶x*´·:'-FLv„É~4¿ Áh> Êwå:jÅlÊŸ÷^ÎŒšÜ—2zYà€R£§!žÄH$²4iŠõ±i CŸ'4Ö؈:_Æ™´îTÿrXØ8Ó à–6§¿|Íß“~FV‚Ñ%¤æŽÙlîå¡ÜçÍœP(—ð,° Ëã,)g ¢ë(›*n’[oBÞvM¹ m$røÛõ÷9"¶P­øöሲ!|•oÐñW|ÁówS]Hô€,Ô-3Âò@Åzfà53:ñÎø+"UÑäû= {Á [úÞ ds€l€ûh® ƒk³óDYqòG¢¬Ô)~9–Äxöñš º>]bFŒpdFŒp`FŒ(˜¦3B ^¡GÜ`}Üo¨{Ÿ‡]»ø·–ikfWœŽcéî9@¥žwT@@1Dâ2˜8Î!íˆìËt\Ì™ý aý§4ä£,k˜úà;§ÚÒi–ɪý-âj“]Qáõò ‹¢Óð‰’´DïôsŽ@—醇”ÙáCJ Uwl‚Ðœ†ªºZÅC©ffä>ÿ×÷ß/d?r4ÌÃÞ“[nÃÊ÷Ar’uçc?ô=ŠúHjB8 I3gCƒ¼ÙNƒ?ôÝIÜÿÙи߮¢-"¶õ}†èå´]Y­;z\V›r]´h]‹zr¶^Y…yÄ ¶†)öf fÀ©©½Ë ÈÜçõ±ƒrºmèÜ÷¥¼Ÿû3Å%88ÒFF·‰f VG£ÃBBÏh!HH˜Ä˜/“o@ùcƒnkuòŽ¼ž:àúqÊäÆ ôTME:I-ôØý°É6Û®ážÔžìº£÷С®¨¹†ë&Ò( õýÏÐùb4È™ÓÈ›2û«Ø±ø7Ö½iü;eu5‘6çw?Ý A[îú×vt½°}¾k1¾|.w°° ™ò—غ}Ø—Ýÿ¿Å7õñ#¦0~vS‘³Ñ2:ÚÔѹut•çŒÿ„XiË”·ãm‹´¸†1Y‘cûHÛcåùxKÝpbÐÊ駰›¼ËAg0½P¸ÂÙ> kÃ`ü‚÷aå!0ï‹|ä›Ø† A!áóׄš[Ïó}zˆ´4`#æ®÷=íp È&=–Ñû3„Å$´ Ùè®$MBŽù‘(Œo/q2)0§i-%–˜gg *;[†›‰!Z‡ÚÕ#WiúÄš".¤[Ò™yê?3i³ ‚âöq×¾ K<Ðr8Œ!øýÌýôxY"c©Ò/-'A~kÇzOÔX– §^ZiôòÌœ˜Ç‹&}I%4•Ùê´_Ï”ÚÎé‰y8S:5°ÜýËžãXM£)ûôã'¡5Ñä]Àíу=1qþpýÃÍoo¾>³!ø ü}Í{Š‘Ÿ@ï¢èc8ð:|õöÝß–”€¨´<­©)•|»=É#ç×/pøŽ¾øåè‹ 88¼ÔÄç”á¥>Žš¥ÌñXÕ‚/~ ”Šqc仲Ê1V} £5¸³HUÞfâ€YF.íÁeÌßžšzš×ç!GAå =1¨í!ûÒ4½1d_‚g½1>Ñx;ó]Ã}·Z¤à2 }ÏU PTÌ? {3ÙbÐŽù B¸{:©Î@‡¦`uæŠÈ¾¯ å]Yíè‘4øËFØ<Š!}“éðpò!Ý.–ÚŸ«dœýd÷t§1¿áÓ,j!Šéiýò¹ «8Ÿe܉‘®NH~©'ëÂøð#Ž}ÿÌ)ñ£þ :øÁúA4žUͱYV_F ?@ý¾¦*E ¯úÿ™ endstream endobj 961 0 obj << /Length 1644 /Filter /FlateDecode >> stream xÚÝXÝoÛ6Ï_áGy‹Y‘Ôç‚=´ë\`¶!3Ї¶0™¶…Ú’'QM³¿~w–úio†,ÏÕA7h‰À3 ‰Ìê:›ŠÄ{h®í;]+Õtoêó¡VM£V4,Jz¶eA2¥W­]úÏx"X"R )'e,€­eÀdÔŸ—Mgˆ¼ù4 ¼ªžÎD{=òªœŠØÓŦr€¾qížÔ>¡—WûÃN}víÖAÚÖ3M¤؈$ñôVÑî¾ù„0Á½TÈv;zGZ¾5=û…ºÎÊf]Õ{Ë»­ÚÝŠè;ÜÃ3¨ {,år ›0&ŠHš”¡·*•¹YÕºGp”*é €Õ¨¼ÕE§2JA¥öÙŽ¸YMÆ ØÏzÅ>ÛeV?ÐFà ÑRÇ¡ù#â‚É 0‘ÏB˜—…?.L Ï'§ÁZwrê v¿«\*…^ÀB&Œ¼wo ’h–N’#$ÚsY¹"b`8\ÖªæÚ#–p΢iRF>ô|Žë¸Ÿ8ܱÐ.X€{¿TZY•·™&µÈa"t˜½*‹CÔǺq9Û½‡.g€ˆR/;@ÎÀðÇú£w Æø³äÖr× Ê¬+íšoà°ó ‘úp?f)ßzKÔ v/Þû\R ‚{›åq{”Ab“‰­¬4Í4íá`¼½ªµZ±^ O/¶ —Ü÷ŠŸ–»ò˜‡÷‰¤)6¥+݈˜É¤ÏÏ€G¶WZÕV~é”±0å«Șá0±Fq¢søÞº-m|à(Û™`ï!Ç9ܨ󔔳HŒªäÚŒt).9“|Ø,ç¿Þ¾}yûÚ™‡Ä|¯jxX ‘,0W¸Ybt¢mŸØŸ‹%iðêå?ŸU!†¢ÅOTÀxçJ€Sµ¤CÞ©%rWäz0¦Š‚É„Gµ üÐV/M%ÆÌR¼IøÈx—åÈð‘Ì{ÕËëùÂʲ&&­ êH è~r²úd9Em¯è…#Na8:"i˜æEÐ× ÚÆF†Ñ,€œ476«‰G}ΰ#–PÛM'!ìîHäYc),žÝ+‡M¡f²4WwÁDÄOª»U<£J7¨ëV¿3›ÏdˆÙRŽ‚ài·JL5;ò*Ê?…m¹ð­¦FKB&ÿ@Ú–ýà1W4ß9u R!jsWjûàâÒ6-rÒÅ7ÅÏvãr¡ú »ê—|v›ÿbƒm$Lbûò8…FUÀ #&©Ø'°,žmâл…ð˜­2Ñ"ÉÙ{vâxÌ¢î2ÒglRå(Cê^Ö"GÝÍAB$ã1î¥|çã5‘؉éÀ¢‚*o ôõ.¼:AÛê~Ÿ•Ë vìgP/Êëœ.8éØáÚáYm‰õÝ߬Ûe›mwc+…DDLúâPûªø…v4[W®Ù¢ú7FŒsQ»1æ_ äc‡£º‘>q_þlTùóü÷?íØS §§_ ë“Ï.!4À27ô¿Ùþ1į øÀÖgV”¬98FÀ)opš 6Ó2‚^#N²7]¢ïí±gkwº˜õ·.sÓöm÷…u_H »ß^Ôfë ¼ÅÂÛØ»wIÃ;ÜïèÎQŽ‹~š‹‚®AÃ:o&ÉŽëÚya\ö­Züi9"j×uY±§jšÉè±Ç§á+{9×UýxÂ]EMÛa zMÏ~½ÁJžZBFýÅ óaÛÄ'pÌ䨉ÿÐB–ðãÎaAû â!vÌy±á,~4èYC@ÝN×ûRW\Zq©~1‡³çî¿cág$2JaoâfN;¯É±d‚£žŠc¿0e‰L\í;Ê7f‡ËáN•ˆ(û æ/¯¸÷ÿ endstream endobj 969 0 obj << /Length 3050 /Filter /FlateDecode >> stream xÚíËrãÆñ¾_¡[Àd1Æ<0âÓÚÎ:Ž]±KQâÃzk‘ˆ20¸²\þøôcðÔPä&U9¥xÀBX)°å®†Üýðöûëß\å#Dd³zN[Ú:Ç ¶Ê ËE'Ë(6²‹£¿xóå·'Ï6ÂÄf|o-K÷΄LàÞ¡yƒtÁs)’(6¨ò®÷ŠO$L4dW>uW6@otoî¼4T"ž¨K:ïÆ1ph¼.Єv¹|h‹Yƒ£`•yU=¡æIŸ¼ÇéÔƒ}I¾p¾-òŠ‡ò¶ÍQ2ŸxbŸ»S»ò×Âw¸~6È‘œnôu#”ïEí»[õåLAt*ƒÛ#à«“slÍ0Ç.b®ŘãÀ¶y8TÅ/Ü™!@„<ŽŸB^TÓl· ëÏÔ¤óP«T@sq២8Ú•KGcms¬wÄèìDë±0)9‰Íu}[îHOU tËûaºVËÍ=¨ò·4 ”BUÿ%¢V6ü½e‰àNñK¾í«'îOñäQh~Š¤)v¬!ÑÂ$|C|0$}:‰ƒãûx }™&8vdød˜š¶¿{Z»;3|~8´Í¡-óÍ!Í“\¬ö:VeîæFÑpû°lðšùE™Ìûa 6TD8nè[7=7î92®/ Ë%·ÜСÊëºhÝ~eUq«;n·EáŽAée=¶õpDúäQÛ˜ðuòø׿÷×.(lP(S¼­ó:`¬ŸœÁKSFs­Ÿ#é NÈ «^̓ý¾é ^mÛÒIÔŽ§Ëzí¼U– \Å©‘Qä¿E S‘R‘#_ŽÌ=º[ÎÖ{bŽÕ!t¿¡CmÚFhaíw×`×Bd)¨.(ÓCX‡ &¶XàT6'Žb‹0s±Õ‘[Zì¶#ñÅaÞCY]Ç+ŸÔßÙmNbX°êk)?IgiG²Å|–ÛÓÖÙA_ìsš_*MÎY–ˆ ¼z“3 ›±Ãì‚´¹Ë Ô~C~Ó’mÞ—õ=Ï( ½Y°Rð˜\È7~ˆŽå;©š± é#DfÞäEƒ5ñyVN¹ú “꼄Od¡¶æŒùt„^`¡œ¼L;8º½shÈ1ç‡9Šì©ìplÛ‚Õ“2ñúÁém IõIÊX;±ŸùÁ»*¿÷fðÈ#©—¼!kgÁUƒ_üXv|Íó((ÿôž¢-)ü›ïÄx^wúç¸'§ü:ø•]?Ú^7Ñò÷¡¡”ZcFƒðÆp9L;“”3GÙ˜1Ö\}ОÐÓy¢Ì\,„ÙO=‡õál_À¶:fz¦˜ (ÆòÝîb°)wÂÂK##ß»ˆŒƒO”? ø­$10q,¬”ç~ÎÀ}ÿjÓyÙ*‘ÏÅn2&&1sc}N‡qœJ”Ôº§ÓsžgÆV}Ž…€õ†íàmLzq±äp ág5u<ô±ì÷ÜZM¡«{$¼EfÁW'R‰ î]ºSs÷û‰ÛxIo£„¢‚“O0 5ó¦BVH­Î–Ì=¥1Ä®™ìPË#9ü¬ )¾ê¶oÚé†UÃÆï0ÛúŽ¿óšäl' ¢ÀÊÇ6]¿$ Å:¸fºª#üD‹ÅTdŸ«ßù7+.UûÂ< ®Ö\ðva’ø|V*º?n5íCÑCéøy²ª)YõI#DSþ3³P¾P*©žE˜³ôU³{UÚUQ߃bùbE ;ÈD^ÈL;ç/ŠŒ?ù’=âËÅ )­0©ç} V'Ëwº$`« T|‡‘y+Y¼Þ@—à;3#ؽÒ/ ÅûgRš)’¦sO1Ù{È[H ¡«§J•s䦲™]KÛBLDzdÉ>LäMâÞS“Å{v9èJ^ m;qåEÝN§¸#Lü££¼*½²aÀÐêx!Ú£ôiS&ìT#ÿé;ºáf7ˆ[ü¾„]O°š¢,àlD¦“OöÝúp¶·:»> stream xÚ­ÉŽãÆõ>_!äb*hÑd×ø”eÆq ‚¤ ´d•…e® :ØL‡iž­ÀÄÒRlÚw´I˜dåj3[ó²,™/ËÃ"/¼‹¢È‡‹A·*fÞ´† ð~BàÔ’"˜*³nZ#¦Šqü€ºc¶μy;8û7íì•Ko;±¸k}<2ÔÙ_.uwë+² ôº·Ýˆ°qPÝ“M'Áãev¹Ù¶t×CË@ÕÞùõJfK’48w-ªà½3pdd05ÚÀl¹= ´;~²¿°¿t†jÒ.æÎlí8DÏë± G„WCÚÀ™5ïdÏd­ï¢8±•lé'"hý{Ôí\F‘Ø+ Š‚Éktê1”ž‡®ªw»o¼IS0·| Jý`MÅT€šu *})ŠÅyÄ-Âà¢3°å4òÆ$TQþ¿pÛ3£†}½o¬ð>S䨲œU–³•çAeñÆ-´BVdÍÙa”Vàµzipý`šÊt°=A“iñ™Mn‡/a-OÛ†ã"ŽòÒ¼åÜÙÞ2«Ô-Tt~õäq˜êçéIùÓ]R„ ´çè½Ù šëYÖKâ2xƒu–3(Xø¾½“ÀÂQeKã›R0¸ÛGRpFY¼à ­ñç¡Úù3rÇXŒzºc§C€uÍ%‘W‹ÂtI5K>J²{:ÝhBf8Ü,Ìóe=:XÃ.õÒsÅJ!·zx'D¬ŸïܶM?Ìð9C¡‘ßÜßyª®Ó¡½BêyzølŒnÃsÌñLÈm‹6ñAöÔÍÝ ßr®ÅÙƒfäK Ù~i\=DÈŽfßc–þÆwH°Ü„8éÿþ൸Ÿfºøw›êµ¥3oTÆÁŸt;©O5Þ½h€õ-u¸”¦Ààç{£š¼Ê3‰²Dà ,p´W™ jˆE.ÝCe6˜.´1C:®ðîm†©4Œtys ’aá’Ôs¦|1¥Å¼ýñR<«®oÆç‘Ë“9¹O•ñêÓ¾çè»éÂv†§³­ìNlrè.[±Ï?w×[yi]óÏÌoÕýç¬jeÕÆCõ?Ï=×ï°·XHøZ®ŽT!AîÀ¢Ádoô‘*ü±ÓG°ˆæ'ücÞ]ší ] ÈǾåñÝqèPüÏÕÇ·‰]=½a/øï×)Þr%cß1q¸¼3p3È w„e|™+¬Šsp5ìà ÷ ¾4mÖƒ—ÐÛËIÈVÎñE!¨Ü·F®ÖÈ¥æCn3ÒFÛbNÚ‘Zq9Š(¤úŶ@ϳâýkcÝk† yð=ÃW@î@2ÄsÀK B­¯P½€HAˆ*rFþÖÑå-UC×é°mcÖ:¦[)Ä ×ìö2X~y³.!¾°Eô¾ãCFË0VÒf}ëÚ‚®‡Á¢û=„Xú4†œ{©çCݘ¢œÚ ÀU Och8¸.ÍÔ ™7NÏÇ55Fy]Edﶃs·•ë}T˜~ñJæ1g_)°*@ÞéNê­„ ™œ‘J|8µÊöÛ®–›sÅcusÛw†T@%›RØ/ÖÔ{S˜ ±úmSz92ïF»ý›OGú– qú+î¬%¡B ~üŒû•üú7¼ÿ •™–pûÂ{+u+I’@@(—’|¼‹îV¿À÷-RWçpwý39G‰žÈfQ¶$uKïY§BlÖédÚ¾¯hSY)v @ßž©Þ£‡ÔòD_ÌŠ±K ‹è¶ Os>SKA~ìë©… ¯bøó-Þ6E¡ ¸ƒ´kÖPcŸN(…§F×èCLÃu2çf¼‹uÒ@(µ@¸ˆòb‘ÁæH´ænï×ÔGoø OìÆHÁo(n꣕¢AŒž{×;ŽbÔB-°…ê–I˜ Ë‹víQv1ó0qUÏHqçÖžìð;_¿»ËdY>€"s¯çª0KÇ„t°Žê w:Àž„\o ôö{‘È[ •Yð5Fz:0„!DR SÒ„žï¢6N>ÎÔ?¢:ÃH }2oD”Mœ†%_,vŽ½â3¥\ €Ša¸#Lãó0Mï²wpHDf€Pf $4`KfÄÍ='tŵScRZ—ß´»Í·ºm)·£„êó땸ÛØ¡’‚¹íê}Ýȇ,´+NóÎU25[5Ek&,•°|º™³$†ÐÙ±ã5ˆëc^L‚ÿ÷Yw­TПt-p:ú¸•Èy(4«³é ÄYã—»žÇDÈa Ÿ©”rÛHDÍGDM­Cê“AýQ… ¯Øˆã;è³öŽ.Ã2óa]oºÚÛYRxd.xÿípêÕ­ \¥U¼ \]ëï„ɤ>j¦®w—ñ½K`éÑ¥üíŸÒ^ˆ•ÔÉlèûc$‡øõí`jípE³W‚]$ßÛhÊÈæg„øé8ñ ‰º)ÿg4•Æò}!¾YhWGÓ¤‹oø2•½* ?Ò§’/Ø)3ÜÐm^÷ŽÀùÊùÿ*¾ûû¿~÷íÛ×rEáû°J endstream endobj 983 0 obj << /Length 1891 /Filter /FlateDecode >> stream xÚ½XYsÛ6~÷¯Ð[©ÆD€gú”æª;mf»Ó‡$ã¡HH„"±Ý_ß],H‘2»©›ñŒ ,‹o{A|áÁ_$Þ"’’%~¼Èv'ž¡Ö› Þ½9á–ÏFwÄùóÅÉÓ×A°àK¼„/.ÖcQùâ½ób›î[U/])¥ã?[º¾8¯__üE”wj­jUfjéŠ ŽCG&Ë¿ž¼º „x :ä¼眅‹0ö—>!üàñ Ý°¥páühNZ(œv  d:i¡7åN•K;-‘tc—è³/Òv]Õ;7W{P$r`'wÊ%wrÜõÛ>w)IÑ-~nˆúÁ <ØMuîöaàŽ>:ï€)wÑN ·Ë%ã©’תN?Hœ]@b'<¤¯^ã7vÎÏ_áüì÷—v©lÚºËZ]•VJZ+tÊ“ŸxΪk‰fÀ®<¯AW£ôäüht¾OTt œò²*]"eà4m·)ŒDw?ÏtØÇ|(pz ·§;;Ú¥z«KVæ Ý/vÎJZݨRÕiqÚ‹)nHPZש¹Ÿ†VÒ¢¨Ü9ÔYÚª/$v®t»Ep¸pìpÏü=ÔlX¯Û«ËÊË,óÄ;Ƴ1>ÊÞtimGõçµ~#ge®Ù2*p°Âh+Æm{vô±F©ãØò¥Ï¤-Bßga,L|±Ö˜—À?.L M)ãˆë÷»#3Qw|ŒQï\O#|`TBýÞxÚÚ•B°…<ü8ñ¶LœÑbYšÐºê5ý<€ž`2ô¡ä, #„úþ£·Èa.niqe8w ÉD$aT,ÎOþ˜ÓGx!#A´Ú˜K5w‘}<¥Ü¶O7ŠF<86¼ˆ% á^Æ×[ÏíŽØç` µé ýÞ”'VŒð˜6ý]˜¨‰â[¦ !4MgWM‚¯jöäq™¶®Ĭ֭Î(’â¤÷É,…¬a­‰ñ…WMòªü¡¥!2Îz@WõÀè‰À=ƒž²A9=”œ*NX˜ˆiïÒ>GÂÙ‚;¥˜´¿,mdEÜ)tÛvœa*ªÈNuU±Â)í¨I¥#XºeXà^dšþ4“l`I—ôÝ©]Uß,cé`i‚ycS&`uT&b¨4 ²²KxîíÒðbξ‹£¤7Ø“'°? ìjꬮ¢+ûèÅ-VÒë ëØóÖÌÁ\2O ™¯réܹfÙs6-°l Gík²½ÎíÙPĪrCcUVÝfKãImÈ™ùÔRràª3Õ7€äwÔ¦4™Ü&^—8gkZ;¸7î²î Ã;Ü›'œÅaòP÷Œ`š]+¨§ÊÎoÐèt­²JWT g›Ü“(T?î„ݘo°M ƒ5÷òÏ·Ï;{óöÕË9üxŠUçj«©ÃyžtrÝ,…Îî è ¦–f¶®¦ÛS#.Ôn(ÄÆÎÀH)¨j½¢€;ÐVˆIT bd Ú”ÔR+ Ë}âZÄãX®Q$Ò ¿†€»¡qV+ôV3Þ!sW´z_X tƒØ?á `˜5:$RRú` Ì4ÒvÏçŽ ö‰¦5ÇfdêZ7­å6ãHðFÓ-—sýJ£ÿ6}-P&j¡j®¨gm[ÓdÙ^ÝAµ4©•K¹üÐfÉ¡Í5L´@Í,É0i…Â^À;å† ø"–HIøÌýòhäÉ0¡6ø{ê·¸í€H°°µl`Zw ÖԵʺ֮k›êušÙ½«CÓÞ-Ba£™cC“tÈ;§²÷:Y¸s†ÿTšuU’¯® {¢Š[ššjà%Ó¾“{Tƒµ8,áMâ-r9¶UWä4EìØTD%õ©°’b~û²Äܾf7O­t»žà[A¥9ÝóÝ%÷À?Ž™ýëæ²ßîöϽ覇µ–øèd>9ßûç9hZÂÚ½Î@KLñ¦°«Ûí›äÇšI?dàŽ4»»{뙿®Ç‘È¡wõo"°‚Æ1sчãÁï£>~QiKvÓǬz6—ùEÌâé _* æ D@Ê ½Ìטš= 6…2œ8H'µ£m”èpº? ÔUØÕ\ߺ< ˜®AqFËüd-cÏ!ƒlÿUØÍz¹ÇŸWÝ Ó¿Vk {BÕ³Ôºšå­îÒßw%—µÈæ5ã Ú7ÞØ<ö¯AÿOö÷Ú?õ÷ ‹ æG0ÿÐùwŽŠoÖ¥þ?5¸ßîÇ9Ý "æ ˜NÅ°ós§ÈÏ8±ï¶ã|‡,ö‡7‚ŒpiÔ°Ð ÞëðŸ6Ã<îÓR´\ý ¼ÿUºšnÕ¨ÏÝè§V¨Êõ¦ëŸ[ ¾’ÐÖ<Ø0…„”qçw¼›H·ß÷×CàÔbÙjÿ§ôm endstream endobj 990 0 obj << /Length 2593 /Filter /FlateDecode >> stream xÚ¥YYoÜÈ~÷¯ä‰cx:¼ì““X#XÄ ü°^Ù£!<$'<,éߧ®n’ã–íu ‡é£Ø]]çW¥`çÃ_°+ü]Eªˆó]Õ¾ðiu¸ßñàßÿxÝ+Ê¿Þ¾øóM’ì_~ìnë£nëÝo^ìï¿ýç.ÊB•ú)œB«77·ö‡(Š¼HÁ’¼xskïIÂðBÊïs”æ± ¢˜ïnuÙ5Ýý¸?„aâ•ò;4b¯ï´,]ÊqÔµl÷²Ä»Ÿ‡™?¸œË®Ó/ý<5ö„¦»:ü2èJ×p;¯ºšš¾Õþû¹w»?ÞixïH×ïˆýBáPDˆ× Ùæ¹Cw"$LÜCQ–z]?ñ Å÷ä^Ý|ôƒXׯ`5O½²«y5—ᬂKipGßk³z>kù´Ê(âIlZs„!¨õØ ruæõ¯6ôÕ¤‡¶y„#Ñv|yJÀoyh¦Û^‹‡E1é—HÒ8ès9<¹$„‰ ³ÈHòxœ>éG]Í“vI4ÍUZÓUÄÓ!(ÀæŠbwåGïÝžœš #•'½å¢L™!jçqr±æªH,Õ©DY}Ù³"â46ZÑO«A—*'G1HF4Ên„åV¦¤¯‹Q,Ѓy©ïØp‰´¿:j%¾Ý!÷U–e<Ι:j;W÷hó ÷ 2μvf¾šf‹uÁ¤ÄŸxi/gýxøûÍ-¯“$i Ÿæ’{(p3—ž?ÕG§€‹@¥Q²èšx|m-îê|;[,‡!€^E¡ÊÓ­Í®ÂB‘°&Âݬœxi¥ ÚX«eæGÞqè[!¦pƒ»rl*ù¢ßì¥&@ÉâRDz2‡Ujúø>×L¿5&\A%d8}u|Ô<º}Ò¸œ¸¤qEÐ+žz<ÿu­ùªA±2PC¬¢8 Q p|F”˜$¢²ý!ICï2Á,ðj†CÈW¡‚07‰‘/RjyR÷ô´jnáa0˜ˆH„‡œŸpÔê ùuÍóØqB3Àƒ‘ŸâÎ|Ìû8{ù04Âðr<‘¤b©^ñïòÉK$úzµ.§Ò2ôŠ`]%7ÄqáÛhŽF]Ø¢ø5í>¡ÑBl¡ e ø…'q“ øˆ¿®fÈ1¢ÎO²WkâÊ|–ütbÉD€ÀŠn~8‡áh`cÿ 4œöc°Km”u gÖ±'xé „­e›N!^Ž{ìå„`â 4M¬gP<øÅ=OKþõ¥8Bâ³SN1Á.C?”­r˜jû½Ö×9’"Ï%ÆÊOc‚Í*=…±""áçíÊH›ï«`úúb齉­Qš" ÇAâýöaŸyN‚¤ÇÃ{Î7$qûW ’¿cžÈ'—÷š©ƒüú¥XJøI´yéó‚¡þλ®嘈îÞ„숟•µ2ß#†»iÁNHÌC—Ћî$l¬ÒsFå2Áƒ„I¦wwÀF|l,Ø¡îÑ&ÜQR®ÁÊ~ý­_q\!ðSU€ÍHºX¹3>"ñÅva`üqáX[øMv`\媡¹%Tse8u%Õ “Za,—$•ƒ ráp#â»’Üþ3Š%µqË%@ùalk‚%º„‘dP?„W©s{ÜÍû°†‘Í×l0\‰Ýv©ç©r p87ŸÔmD@3 8"én„¶TƒÒXÐÏPV“v‚ì Ï Èu•»& Sª¸ÖŒl€ôØ·‚¤Ù0~achÈ°p_Cʯk#­æè [͸`q—¸ð¶ÀªžS^©J „Ä®-h–ZCzÈ0î¸,O!p\×ßuªH-š’Ê2Æ"/Ûšñ ‡RÁ~}·@õ®Ñ£ú5ä>ö&%KȈ)D-àücÅCyOþ´eƒà‡ž‘n­ë¿üTHY£8§$²DùÙÖc¨Z*l<Á1W9¦SüÁðpÖ<Æ“k§Æ±JVšùqLX}`¦]ἤž Î‘¨Qw ت‹AKxŸt†¿¢1À>nbƒ²‚Ï¢ L1Ͷ§œ/I wûÖÿg«H_BAnŽ…Ê\Uzy‚øÅ° j,å€ÿÃgËæ<Kå(÷^óâ™ìå+¼ÃY ÑY¾XõàÉ‹ˆ9SbÑo»ƒþïÜ dI¾àKœE$…×_Áu<%†ûչǤˆk\Ú5X|«ûyõùhv_ú! Í£ãRzXiPÝË./Y»]qYòÔæNÉE×h•2zÎñèpO–˜w&8‹t£\ù7‚‘qG§+ÆülÕIýÄT¥ù*Sä¦ ›ôq±gB»¹ÆL¬ã(ƒ¸w•ªE(c#é) æ;—$ð‘íÇ­ª4gv ©M'ÄÔl 7Õ@yã©ŸÏ5_¹êk†lT‚éa*GøBÔŒ‹&‚¶Í8㘽ŠÐ›MÏŒ‰ôc3ÉÝ,8C¬±K Æ·Fß–M'¡xÍ á¿šã¸áøȹÝ6Ecø¼^CÆÕŽÕ,bžT?å¹âuV µAha‡:ôéŒ ¹P„™m’/®’÷‘+(B—–Ý¿þóîÝ€"j0w§ñháš®;õglhع  ñ…JÓôºm$—ˆäq\Ø.NbÆq#íêú RPƒàõ‰7¨„FVÈÓ+Ydu%\®ÚÏMi΂Ⱥ±¹kÎÍdë)¸ÖÔÔ4¾Õ¿«õ†é9MÆF¦“[¾Šü¥>hý ÌJb«§‡“±esüR ñï¹ï8êÁ^§5@Óz‡“òøª;ñÛ Qîó«3ŠöTšiž®M~ÕÀ yóU“s£÷es,[©ækKrìbTðºéDÃü¦àdnZ&2nKý“6L?Vú2mÿÕ1–«êÐy#ÎÈšÚ»·ïN(‹s£Ï{étÑë÷{K|¼•x Í-(òJ5Žt‡1æ‹­(~àýÆbÆKYigGÑÿº•>ÛJ|‹Í—(üÙ6bÓRhŸÆI·WÍ´g;qÎ#$ºaséÿêEþàKÓÐ}AÓaƳTø!x²Ý¦†â ËzÓOücÝClÏüSï‰ endstream endobj 994 0 obj << /Length 3226 /Filter /FlateDecode >> stream xÚ¥ZYܸ~÷¯äI¸eÔµHŒõfƒìŽá{ah$v·°j©#JO~}ê"%õ¨íIÓY$‹Å:¾*:¼ à_xS7Yû…ÊoªÓ‹€z‡Ã 7~ùñE(t; Ü-(ÿr÷âÕ›$¹ ¿Šðæn¿\ꮾùà½>–çQ·»8Ž=õÝíN©Ä{óæî=÷ü¢÷zÐ]¥owQ’穧ÂÛßîþöâ‡;·iEÏä)¿Í^š+?Œ³wwÔöViäí§®›¾3üÙœÎÀSæõÃÈ©û“ v·Qî=•üsúÃPEèíK[—£˜±tÍí{\à‘ÛãQã¡×tÉkœÆ~”Åp6bu¿>é/g`è’ßì,Á.Œý0a*á’—[=Îý0ÈÝjóic{å=6÷zäê3þÑ>ð–)ïc¿j}y)Q¬|•f7iäŠ.ÆO`Ì øFtCëžåUÙù»Å×u¹ ±þ«&ΙñعñáýmÃmÕr¢~Ï¿ïI »Û]èýzÂ]Á~nºþ³-;óÛKüH¼syÐØR^˜_XŸ¥éêÀ×5ÑRãx—‹Òñ|yÈ¿cÅ€õYn:dBÖÐ5µ ÍÉê(¶}nËJ£8bå•(Üš ÚA—µ ”U5pjKYT9UETU^}ê:–"|ˆæû[Z #ÌüÔtÖbàµøº®†AêÇaf•uqR8NT$ÞÞg÷i°A‡Án¶º¨HçóâGÓ§‘›']7Ó /<€ãî4¼¼®šA¨ðܸØ=ªÊ#·éÜHZ•m{_V¸øï|7–×ÕÝXóÚ’ˆŠým‚)QúŸªc9lÉ! ýD%–tS¾¹ŸEñóKÀ–iƒŠ‘`‹FŸÎ«ôXm-EØ´T»¶ù<¯ë$@_ã±¹5èq:c»‘: ¼N‘q-œ ¬8l@WÓ±¤£ ­ü¯–Lwü=¶3¯Å œa“æ…Fº¯8ÜP0UdDŽ{ëØ©Ÿº;Å”€`7:—Æùàfýæµ%~­\û3Œ"ögq“Ž½Ÿö󑨡»Ú9@‰àCƒh­ŸD^ÅV„Ö_«°EUàu±²)pÁ„‡cÃÜeŽýÔÖLßiAôÛêx|ßËÈçÛ¬´mêygt|/7#¡J €¤ÏÕò"P–TX'úA¼–kãöžä’eþñ抨À¹) º•^­Ñ•t$µÌªðåYägsHþ£ë¦÷ú¶'@Iäy佹͕GB…mª¾“žÄVàê’P’ìû¶¥0ÿ@î»Æã eô# £‡Á]Úðñqyè§J­,o觱é8¾€+4G ˆÄBSëú»ÿ) |B/þiß´ºãÝ´„\n’/õ=y>G6lK0€VÉ?x#­æ6.]o:$ò8Óü  •y˜/ þg^_d̶°Kä™\b+öÒ´4u^ó%bÏÂ[Z”&pÐMÍ€"GÚGé'B,ŠýlnØ]¢7Àþªíè)Å2èsò€NYl0ºÝÛÑ$ÐþŽˆw¿¢c:GÚ‘{˜„®„¬ëî5q°-ºŸÀs€›Â ˆ[´åp@Ï‚]Ì·ø8Í]±€¿ô]M£]5#µ1OŸŒŠà† ¸yø_¿á³ãbîö»ž%X¿¡Ó‰Ñ/« ô>¼.ç®Cö6BöÎpE šŒ=~#ÕËä]IŸ¾ ÅQ `1[íú»¥þÆi.u5?aèÞÖkGªD góã£oñ”à.¸Â+]¶€|ˆŽ–thû{ OElâÊöW<´Ãd³²ýŒ4 xÈW¹ 0˜Äà“¢ g¨Ð„4Øgͯe³°`²í Áq©4ßð†»±ßÁRûmøq~¢Šj°¬Íá†cɯ†?KæÉ=g‰ó*Ø_ ™@ušþDJÕQ”ÁÅ_”L°²Šv_Oôª§’Ø=Þ«Rqæ'©sΖbBu˜†R,æIíöV-bR‰ÑDa6wƤdIJsÊ ÀR€V©V‚D—<ôš¿uÙý¨›ñŽ¸b"~T¥<ó„ø`Q\T#œ¡é"ò¤ÿÎ]ó‹<ö/Tálƒ H*fôx4ŽCüà7²Y&¸ù’ tuÒ %Ž!ã3lr[úl)ò6÷C 9W¨U²WFº{˜H<‹¿¾öz»ñ¬ÁlýûºtÝ–”òƒ9ºPi„™%#Ì•Úb5¡ãþùžVHÁÂÛž»z”¥¬ã±72IdH${îšw™Š¶ïEÌêêdó[¸Õh˜ET©¹ á9JÑ“%9Ëï%з²"$*hc•}Üæ÷õåNŒ®ãHRô±ÿØàMËd Fæàl6ó² ÑÛ•€±2Ñá²ØB#?![øZ82¥}r°èûäÜÔÝôTÚzvW3zZL•kÁ®¦³]Ú>.¬žøQˆy9rZ@g í°3èÿO”„ù endstream endobj 997 0 obj << /Length 3014 /Filter /FlateDecode >> stream xÚÍZÝܶ÷_±È“õ²â—Hðƒãôš¤h‹: ¤Eò®îNñ®t‘´±ôï ‡Ô×i÷6­÷p$5Î g~œ-_¥ðÇW.])™Svµ;>Kýjs·¢Áë??ãn„›åÛg¼ÑzÅSæRÇWÛÛ1«í~õ}¢ÄúÇí7+iËÒ ¸øÕ››íwë”2‘ þäÙŸ¶ý>Zˆ+BÊGH1©2$ʬb\ª ³ëÎDðä>ï@£¢HF'¯‹üpø@ãWõñáÔ-J8WlÁã"húu¯•t÷e‹#™´Å®+ëê9βäÝZؤ š‡¦ÆÙ/å¾_(ve[Ð{Ǽ»/Ž(]Þ•kžìüð@÷Å)WU‰œ[zù¶nè¼dhòª…åceòj‰WO¥—Œk~ôdë’*ÙÞ^˜T+ p£•¹)ŽdrWÒ±ÑÌï-p‹ž<}(gÑÃâ}G‚f5ý?"ÑéЕÉ»þ+Œ¢áC²ÂĬìç>Ý’Ò‚E1?Ë€ZcÂõýéj¿A£K·zçI>4€Sº:¬¾}öÊË&¼R¦”¹ZYþ”²/ªÅ“D@€„qdÙÏ—<~# 6¸˜ä~×t1ðÞä;¼žÞÈ‹«çpJY=AÞ/^¾úË9è5Š)­¦Ð‹ìH J"À,F> µƒ–)8¼ùòSŒ|ótäg-òÍq/?fÜÛ…¸WÓ°7ÎǽºÔâ jw!¨aoOeŠØýÿ êÀKCaµý}‚Z] jHk”4šiÆE(\“£›gxsãÿSKƒbO+ãZsX.}fïù ‹FVC8,ǪŽûP©p¶+û¢Å“>}…@Ó¹ÌskG™MÚÓñë–,”™žt0}Hj†Äêfé¸4Y’?<>PîŸÙÇ$ÑĪQh›jFìÒŽòreà“™òp |ë3ö_BÊXÐJ4óƒ/ƒ³)n xºº˜²eL+)¥…ÜbÿÓB¸’²èX755 ét’SéT…%Mqô†EäGär×æÜë‹^ì±y\>Ó@e†µxCµoi‹üŸ“x¾¤…¢NÖQSëp¦xöˆçùæèÎâ´K‡z0{\lÒr̳¥˜æjÉiEï”0ú[í %á’üÐÖ´[vÔ0qÉ©-âã0ˆÝšýÀ¥*« xsÑ|F¯Ô§îáñÚ…)>ïþrXpˆÀºÞ~dûc. ´{JÁS}ßÞ>yÕF)|4ˆ]Ó%Œ›î‚ Ô¡ƒßµó­,_qïi™:T.¶¼°ÁÅUÓv4¹Ï·4ªogtÁûi“j?~jÓ¹þ3É7cúFb "ⵋc/üŸ_.¡iè#{JIû#›‚Z Õš4…¥ KjcjÒ%ô@1r¶ñÉ<8Õ:_Hq§1AVt-q.‚9“bQHjb¤˜:Ízù1\”yKwis’Ïj• ;+fwçTO;M?—µ¤á“0*'uÎbÒ,†βFÒ´gÄNÓÞ†kOÉ7òÄ…F~Øìó.}½›íµÝ9i qÀ·cÒb¨WÚˆ](êL,êŒ 'ûïbÜG22é¯+¸UÆ>¢2f Å$-oÖ½|ALÅÁWÌã cÁîêH-.¶¤ÀîV¨Qs­Ô††Ú4)Òàa—ûiŽ™!}Cǯ9n–N\nÂݦôÅ¡ó©Y_­xŸï:Ê=Uˆ¤þƒKŠýY&h Eƒ’l9º vb[.»ª-g®*ÎÍUŹý=ûrö“ìË™«úr抾œútûrr?lHŒ¾œàËø'TÂǾ\Ê2¸þÜúÛ‘Û! á˜Àeàp¦L´¦ú)„CTì|/8y[zx€ñ_‘4i!—„Ù;Og(‡äCvØm“@’SÁòä“·åÕgˇã±èš3=°ÌüSÁÇììÍË£Çl»±£à‡ ‹G¬]ï$rA,¶Ü€;éì7ãÐLÑÃ-*\a°i"²‡/?˜¼LRM*`¨¾9ú–Cª£€ Ž’’sž3jAxGð=‹MÈ€a2/`Ê­û X†õ‹™•aiƯtsážµ¾Ñüøþ¸ÂÔÅÝÒå3Q™ïÚ.p–̦|Úvy "”ô •€$ =:Zòß•?°À•>` ìƒXdiÈîò—à{Cþ(¸ütù\9ShD D9ùÒ¿ü›¤£0Ù 'î«Ö‰ýG{\ÝÕÕO'¨Bj/$x.˜|’èx=l¨RpZx8$mÿàÊÑNP†Ûzö¢ÿêŠ]-Xìh<ýeƒÿq†ß£Bœó£€¬8.Å14|]ºÞdÔ«ÆH©Å H*ÞÔãWõâé„^nÏ|uΙ•zÜ9Wg\#–‰ÅEšüþñX”A6ääôä†ßg@t{9ùguÀ5Õ†“ë?ëNŸ ÐNæ—’~Þ€ ýo6€JÐR2Õ¦>Uû˜§îéw4ë«ìQ”ýŠëQÛ *»Ú¤¿™pê†8óKìÆÐgn{¬„„4ŽËëÐ#»€–Á51"÷~NR•·ËŽÆd¦Ÿúl/¤›Ø I~0¸ŸijÀ,þ¯áèUjšš¦óDñ²â£à Ç•77 nõ%õøz{85…a“úRŠQ—ñÛú9¾ÿ!> Ìlj¤IûµøêÕb¯B3«{s7"2m†Ï$‚~Ƀ}7|4Cz`~—TuG+±ûƒc,£œ~€ C‹ú÷Rû‚t ›1#”^×cÞ±˜8ÿä±õ endstream endobj 1002 0 obj << /Length 3006 /Filter /FlateDecode >> stream xÚ½ZßÔ8~ç¯hÝË¥uÓ&þÇAš‡[XnY­nv¤;<„î èNz“4÷×_•ËNœŒgî„D;NÅ.—Ë_}U¾Já_é*—’ʬ¶‡G©ííÞ¯¨qñGÜÉm@pHþtùèñó,[ñ”iÁW—WáP—»Õëäéuyªn½‘R&êÉz£T–<~ù/깨®ª®j¶Õz#2ct¢äúíå¯~¾'Í„x v(ùuõ´QŒKEê]^ÃÌJÉVtÔ(›jêðC­™(`ûÕ/ââ©š. K…öRrèʦ¿j»CïæjÇÓà&¼ïÊíZ˜äãÍZäIÙí¨÷Ùó5O.©Ý^MŸÐìf6{+‡…¹é·ëŒ'0Ó¾úÓUIf¤—-»®ÄÉ¿¤–¡¤LÃUý;6ØF ÁxªW.ÏHò¦®a—µI~©ºC=ÔeƒyÒ9ª¡û²629ƒ®\'ýÐvÕŽ¤k'f-ƒØ>Ží‰àše|Ü• ñK|W4+&©ö4€ñiÜ“rè­yÒUýÜ0OªízÓ¡ËÂͨö ªJœØÍ5ôW(ÑU“¢0ïj#À¨JgNûjh_¥T“OÐ9¨{úÝUoR®šÊIUŸËí³Ûw¥“¯ÜNloÝN[‰mÙWOb[ÉuÊ2)WØSžkRðIr8¬ÈA}tw&QŒ3-Ôǘisp8± ¤Î#cq»Ÿ ¦ ’jœ”¤+Dæ7éMš¦$£fJ ‘{îÆ(V+4ø'º!¨–ðBÁ cfþ:Ó(e2¨œ îTúp{:ÉÆÉÎÓ¨5ËrIëËïž­`©õ‰ÉJb£*&´eu[¥Œ¥y’YêΘˆÌ(‹É’’£—Âœ›Ìø]œYœ5åÆ–…³ÙÑcð#¸J sÃ{-skt}ý6]íàݯhNY¬n¬ä—„ IWûÕïþI <›o 0C‹oÞñåZù¸KMt—à,*ˆh5‡úO¢&´Ó›ÃáàLŠ >ÈL;Ø裡A :^‰W±± ¦S3"Ð 9\ÛSŸe30€[ ”Шʾ¶§dÞ4™ÄiÓW•ÚÈëã©«üw]Uî¹´°¡Íö•ÇÃ.´K§LÙ˜åà݆ϱlŠgìJ|pêiðíé€ÿ;ºžÙñ€9p°¯p†ý­þ¸&…×–üµ§¶_fØñìùåµh{÷Å,¦Â³]6NMƒ¨¾¯ÿSíØz£ ™¼hèU ßwÔ´`Üv;Ä~Øó¤<÷_6£¢Á’ëæ=ÚŇ#8C¸ÚÍÐn&È…·oÒ,u|ð†GN/œÑí÷ àFÊ=hã`6žøÀb­é]†?²™{ Ä¥z¿-ÀnÆi?ÔGHh~Ø·º¡P}ïî ÿÖ™Öã)lbþ qXþÍH"Ï,²€A3<¿þ *f˜`çäh¦<FÖ­>­%Ø¢¡.p€žZhgO/{k„ˆ¸ô4dwR§ã èqvŠ‰„X7ôûH=G…:-Ð5Ecè­þ8ÕŸÖx^ö•õúúía·ÉÅÌ!éìYÔÕ&† ´¶kµÎ£€bĽÅHs”){ê 8>–t[ü’,>Ÿ¥)™Ê­= ÍÑ#À Ñ)ìˆ÷ÒB•A”½Ÿrˆê3ÜÃÉÀ^ïºÄB a:™ÚËèˆ;ÔÝêž°J e}4¤Ÿ#Ët-:ŽÆ1.ìØ÷­›¨˜Ma]çãòFç‰L ^0$·€?™ÃœÊ¾!‚† \dÎO o$¦¨T€³®{š7u_ÅfmÁ!B§ZÏ]+ð}8÷“Ž*‰¢ósÔÞÒL%œ†Ð@­©[ „¹b"-vñ3¸¦ê‹Ü2ŠÑ`h¥¼q`ð@Wx Î)<*À<¯V9Ä9.g*Í~y#ËgŒlÉn¥ÉVÔyt>ôñmÓ½ŒÐ0صÌÜ&À Zð¡¨Î‹Ìgæu‡ wã¯5bùÑf8±…å€ÍÓ6ÅÏ·dzbíCëO£€/å"  Š)’£ãuk»zKý=‚g‘Þ¹·…fÜ<ÔÖú>XÛíÌå6×f1Ÿºwº¸Ã3ÈÛÃÙ™p™üTmË“=T°h˦LáOô7;5»²ÙúTY¥p,‹j…1ðâ}×G-†9rš›{£7fML;HñcdŽvXê‚A|_pÃÊå©ÛáTî‰l(G1ÏèyaÔÐ9ã‚ßý˜ÕÎvŽ©ºI†ÉøÂnˆÅºµ EÌô$ZÊ0gÐóòO!ÃÚÏæÅ‹?ob´$L^<$G•pnÓ2ê2X'qþ¾yA¾íN—;1 ýN×þ Í®y|ñ-zc#DÐ}MnÄÿì¾æVF‹6ßõhfZ©¿±„HÊ®µt»¯rŠ) à“3,Ò@]ä« äÿ1~åRà=P ö°”ò;î¯ôâ6ãa‹]óbb.üc®ÕôDSïºVËèZ­øþk5]Œðœßy±¦ÈÍŸ¸X[5åYhTŸ¯£Óæ¶l»9ËÁ6m£<ßAVdãÎ&Ô×óQù'òÑ·wØ0À¯ø°ûá×ËÖ² ÌŸ(³²- \ÿ”"í…ÐCoLÒ´N8Ä6xØE MÂqoÔ(À Ì"ßç–B3´ Sã;G.–•„3·Äk·þþt8xˆF¨hÉfYQÏ÷ÏU…Ï4fF+w]{t±b7¿Îÿ‚#´'z8Ô@»Üe<—u|jˆÞª¤.^fó2À<^b™|gÔ4LóU°Ò‹š*åñ¨ ¿/è5({F-Ìë‡úPõ>WØïímŽÁ¥‚úËø‰» Ê:¸ ·¶§§ßÉ%íþh0 ™ÿ­Ê=ëéZs‹kã ùÁäqŒÏÁ‘~0ÆóÿÆßNü„×Fx=Cøèpššßq{©qÇãÑGK%i:Ê‹ê¸%=©÷=<öçñ|ÿNm/>àˆÈYðÀâÆxݨç endstream endobj 1008 0 obj << /Length 3169 /Filter /FlateDecode >> stream xÚÍZÝsÛÆ÷_ÁÉ55Ï÷ààNšÆî¤îÔVÓfl?À$$Á&A­(}÷ã_FTë}û>ºlF4½¦ Údg ²Ìí<ÆÑ mu Š—ÖölüQÇuÖšÚ^o1^‰°IÐz¸þSF9Ë ÿ:º9-¤ìåþãŽÉ‚—1pû,¨(‡¨ÞzÕÿ€×;ÑËî¦äƺh}k½ËNdýž¢Â‚‘ªtÐ ¬›/Õsä™Â2ä=€{SnëÒwº}T1N˜~Ó?Å.Iï.².08N˜Qr.1繋‡V³ÂÂÝQ‰‹U"åò_·ýVéYð£][¯ï«bÝín£ý¶¥ŸøÄ®z»G{‡ft¤ÍðÄcĸ>…»š¼F¨D±(àËÃWüW6xÌV%ý1OÃMŽÃšI¸QñýZ‘ªñµûÄø¹ÛKžóZE½áávÏO/ïQ°Ó)X™>'ÚÍ—÷QwˆiÊp1T/¾+»jW¶,œÀ¶ÜLäý¥`]>&úíý |#X”) €0®Øê±ÎæxàÇÇ@€}"P¿ ø‘cx$ò›£Èo§‘ß<ù¸?iˆüúáÈo~Sä_éÌŠÅDÆß'§Çâ„cq~V,VùYÁXþã(»éu|4ŽŸsbÕ„OÉw97ùÿØß ˜§¢ƒ©óP ^T-2™ôØÒ–¯øÂÅüoÙ›‘ÇÆð)Ŷúµèª}}†W»¼) ]Ø4ø ßD¶ñ¾iÊ–}Òž–@ºª)×ÝöžI(zÚÔÃhêÚË@ÞHÀ'¶~®-=“rÛ–w7eã_"ϤäÙÐÝË™Ÿ¹)ë5‘ça±ÜGè–;7!äJð=Ôež@Óì)Êv#ª1ö°Û‘ºZ@±[e%˜¶L§ª‡£Äê\|P>^w+—¯/ª3f ™ðc÷M º‰ejã úæDàÈ]2JÕK „[ô,õƒ7 iz ¦f‹"šòx„ëÖû®ä^wStÑ]bsé# h2–ß•_ƒÎž°7z(‹rüÆ´YîÁ"š» o…Š«VÙtŒ¯Ÿ ÜXaPHGU‰uõ§‹nµÃ×0›¶†É¼v*l nbU_óÝbš«ÚuSvÔ“öªÚÏôYÜ4+J“x¦8t7÷ð•b‹ÉWõz{ØxNÅfC±ýNö±'ûRÂÄZ,øP„@ ߆ríÏCÁ³X›,ŒƒMÂe\ïž½ÿ(˜„ -,àÑ;"Ýõ0o»x÷쟱 rÏ+ÓXV`ÔTÉSG ÓqA¢Gcv›l¼8±Ã!òÛ{V:Æúr Ø}!¿{{| º¼`æuN oP@ªåñ*Ðúö-Á lña »ö=rÿE}]ÆÀ ]¤•I“å®øB‚·ÜeÇŽQzÝ}§}'ùœ}û§ƒgÐë¦BÓZÏx LkkèƬ=uùó¡úz‘¤`1K¥ÀÁ³ðïw°‘v2–¼÷œæÙ¾l€rˆÙ) qKß–Åv… éåÆyTRqöÝ% ‹°ö̺<4 éo`ºDö8ä÷ÇcÂb8ŽAT÷ÕslMÉDå/ãZèÊÃG$Buá|ÖžFrƒ»Ã7Šfs Ë2^—rù‰¡Â—ð‚w¾0ãõ­EÏóI’Ã\…ÚöÖ!3é¶ &S÷ïN8à%!³lVu›ùQØÌIž¶ ¦ëî&š]&B)ý›,µ¶,oMIç;iJŽñà^3ð3ùT§¬O-ùU­Ø'òBùÄç0Ÿ^–Þd""ApuJìÍÉ™¥Ç T¨…%}ÍZ\©UŒt¨à‚0ìOÆÊaiê ÖJ‡ÝЋ»Á^T‘Å’E¯Š×8OUKç­Ô¸$‘åDÑs– 5@‰·ÿ8;­0y4ËŒvåÆwçveAy%‰‡*…v6ÿDô[Š(lÞHÙ"°6‚¦R7sO”éM¤°• ™¸ãµfU·Qeës¼š­òÉÆü„dM¾Ä·;a¢û’B;ýàU‡@¬…ê´_äÈ“PÇŽ]úÂeSyÀPˆDížçO¹„HŸŸ«Þì!0ÖõYáì¯Oý¬`­^~W®‹]6°s¤Ýó½Oè+K½)ê5ÝcòÑÖ|ËCøÖ•ÅÞFµ¦1Wf}B£„„ü'q¦hv xäO°iGž5™}¢h»¢é@ae9~ˆ‰U³NÝGË%Ù-6ÖÝ¡Ø’&0€X õlªß”ñaJ_x[<›-.™3àÁ]ßqÙð-ýײه;Ð=™ä*cD–dtV7ßPžéyÑ“«W|Î@`Àò½ÁO\pyÛìÃeåqáÏÎ+ÿÜù ï¹wO—Ñ2›Fu~Â<‹JBMC’:+$a1¤ÞD¥ƒ|8ɧÂúŠ%3{ô¹Æƒì*\˜6›Ïê ˜¥Ý1ÄW)$ dÞGp¦‚кà”: ÷/8 •²æU(§B£E@qƒ‰ãÕ€‘/áÐt´B­-˜yJ>º8ŠŒÅ#‚.*ëS$"%HóœSöb ùÉáÚ×»4¤â©Ë§!9Zü˜=­(“_±ýÕ=ºF40&­7¡}òظ®¸øZs·ªç¿*2Ö‰LsNÁé—E"9²pœô£éÈø·FáýÕˆA$/C;}ù%Ý#M‹Dhn¾KÐ0vKFìõ‹áŠá0~Öbd˜¥\ xÁUëç…bC¹‚›)'Q/&cn‹kÿ[üXa:#Ô¯ÐEš>òS¬@½‘GÔ3gÚWH!E郤uT•æÜ„A'0<@oörÃÞÉ |êïgz¾ŽBº„Õ½Àã,¼$=âœp(Éh.Re»å‘5DìÒ·IÝn÷¯4宨꾺Š8zXGùLÆëÊš;å/·[¼ßk<´ªÃjí=ÏLf€øBûrÄqnNÕªxÞþj–V±ÊAzVlZõÞ»-y ÛósMÛEêöà+Ø4^x²¡*4E,¿ÒyÀÿÇ2Ô¶·YåF¬¢?²Â …^*¼ó?ZðuG!rš'ÄÆ»ér„?Œ,Ìòó¡õ³…‡^#ëŠ0Løt ZäC¾O¿ÅóEÒÿ^)Ä"B¨ýSßÕÌb-]ZœÆƒÎø É?gÒÆãûãß p­Æ$ÓáÀƒ9½SÂJýXN?„–:y8¥Ïòô1ø9ªþSÎpZò,.y¦…Îg¿7:JP }¼žØoÀ<²!{¼Äû2¸ÌÿÄög¨ endstream endobj 1012 0 obj << /Length 2247 /Filter /FlateDecode >> stream xÚíZKsÛ6¾ûWèHM+oéäÐ6uãÒi¢I“Ir`%Úb+‘)Åãþú.$! ’èZi{èdÆ"Áåb±Øýö[0d„áižP‘¦2ábüqúÓÅÓnRAé@ëŒäžy Ìcƒ}Þ<™rDwæ½3’üüürŠ1¤Dò üüõtrĘ²»@ðQH2æ|[z! Dª4ȺIümœ!EU+¸®³²¹®ê•sŽ±e=¦ir îI“|reðäùë1I¦ã I¬•V¸(ÝoïÛ¢qÒóü&¼Ìçnø7«ñ©³F²ÐÁQÊÁ‡T#,üß9AŽCZ±ÔÈIzÁàé¤þˆ-N!Íè(zËPiB#*IŠTšŽER;aïX")Ž4­·>àÖ§|Ë6Je+âwè‘FZÂ8LÀB• X%Ò©Ò·‹0bØø‚€B'õûþl ÌíöùŽ9–$‚92¸.BŸþó©D˜ˆÐ§MQ¾N+}ÅH\$½Œˆ.€# A¥àš¯`“)Mˆ ÍC*© ´ÎoÑàIç*¦3ª#‰»]Ó'|ÈRY+TÆÍÒŠÆÌúµ‹Ùòâ>ç¶iÂL€Pb. é.‘]\‚[’œD'þ@t"gB'ø½rÁØœøip<Ë žÆ( ®£Üc„QjF©3b?Qj0FéãEF}¤Ôˆb(u¢dèÒGCÖJÜáuÀ*&ùp,¡Ç€Ž2rçäƒp®ßÌ(BÁfbÜAT– …Ø PÂòG{Pb»d‰DAi4=˜8‘sBS ZöÄ÷Šm³'~ ™d£_œ=¡Æ·\'jxöÂ-íR.ˆ³D—ZAN ,¶ ÜÍ9mR 6Œ×œ‘Øí"E ÉÌäÓé–(¶gÆrúH,§°üœ„S #œò¬„ó‹¡ùIÆ™>œq>¢Bˆ­ Säòì  ù¶.ØvZh>¶®<%[/r‡l³¬ÉÝPu%^ôQr Ú{ðöpE¾nçiµÎç›YîùߺŠ:F!B»ý}Ë:дEEð0Ùæv±Ý4yÒíÕÛý¹öiOœ,#PH¡he¥põ!°í³ýÝÂúæìDÿ,ÕôM{1=ól˜ÆíÙ¿ÅíÅY¸½<[=øs{þ/?œ ¸Q{¦ÿÑŠÂÏVQØ£û”A1è¨]IûlþäucZúÆÁYVÎÝÅK•lYü™­‹ªkÓEn‹—-ˆæM£–‹dVÕuÞ8PªìF®¨óÙzyïD ¼ØQWåàbS–Þ G  ØøgMî•äË&¿[äµÉB[{n V[3"‚ô¡»p§ØŒëv2\g³uU7îªê„ ècoN©Á°€u êh6«•uWcö'\Sã÷tw9V/hû<¦ÊlBÞi”¦¡¹#N©'^rkрC‘êûÙc'ßõD8‚`ꀙâ.UCŽ«´"]à·œVf#˜k™|.Ì.Û+ãgpqæïDâ« ×´? 8VŒÿÞ2( Ý Ñ£¡ÿÍŒí Ê°·*Mê7× î„•â}X¥mXÁ dßf¹n: øm % TuqS”ÙÒ gu™ ¼w·+s /·€Ë¢Æqƒxw­Qƒ矺oÃ;Sž¸4Ž)ÔˆönŽê#€;am4 tÂéNû/Ygß² ’Ü3ó>ˆeÇÙXn§줣I 8ùïeãXy E‚5<ßbI©ŽûÃP±CðéÁ–Ë0ú÷NÓ¿ñøº¾+{65Z¥!Å}€ÕÊ°–Ç[ \#íóÉfU) \ H¤aQ]Å5éêVyã Û¦‡}P4³:_û;àQ¹hÙ·ÅxR­¼L¶Y/l9±7˦r/ål¹™·2ó¹-‘&íÌ}W‚Œd×ØñšË»Ì¿m?¿~꾫jà† I 9Ì|X}ÿæð !âLŽî¬äª#OËÑë‹_¢_‡½*žÚ30fbÁDQU»ëÂaŠí)&:Ôñ͆ƒÔÁÇòÞºÖ 6ùJ¿ƒóîí¦ÕØüÄ)XÚâ¬Í3È.!9Ó¢Îgxæ'©G$àzV5«| élwÜ-#DL‡¨YycöC9_dks¥’Uö‡µ¿iän<è²ÌmUÃæÞ˜ƒ}çJ&¿m¼‚&›Õ… ¢ÙžGvÜ`þiS| ‘ᨈ¤r³2Æ[2Ç<Šæ~‹«au®­»œF{W¼Õ´r”"a<)ëéÝó.¤ÍÝ‹¬^/ó{w3kèüýŠ½›m—ûbÚ ìqé!<Ès×OCØVëJº|ÜÀ€¥WÄí0ˆ[É)d¶L™¶L¨ATÁòûã··Q´aˆ‘FÛù ˜G‹ô›=ñ?p¨`ä¬ZAäZ^Ý®¢·>,²‘¹&ŒX ºï@&빊]DxTÔd«¼/Rþ€çÎ`ô¢ýW·Ñ>Ÿ`@LûÏ}þÒé?ð¬<Ò⻃žÿ[üã¾P túò=`ŸI$O »CaŸ?a4O„‚½‚˜;ІRh  šâÊ¿gPÎ T&`*š3ÚþQW§NTÚTbqåîO¾ a Òî²2Þ{Ê­Zß7…ÒW߸v]¨WWVí$î8JƒœyßEÊNÏVs)}s)[Û•”ÓvšÆƒžÏSá³H‘»ºò"–ž9©;3MUϳ k¾í*3å­Pfp¡‡h3b^¼³]•‘´‡­4Ìèmiü¯-èð¤²oE‹ÚN§Òÿ'„»b¹tP²‡h³¬!<§ŒóDFà@¢hXÕ_1‹B\ endstream endobj 930 0 obj << /Type /ObjStm /N 100 /First 893 /Length 2187 /Filter /FlateDecode >> stream xÚÍZ[o7~ׯàcûŠ—ÃC (ÆpQì0b/²Ý&XL­qÖ»²¦ä4Ý_¿ß[ÝI3N²@brfÉsùÎ…¤$ˆqFB0ž¢)Þ„Â&¦`"^Iˆ&2Ú˜Lr­7õ9˜Ì1¤dS0ÚyÅÄèè`öèˆR’ñ^‡3>`!Ǩƒf¼IX< :Ñø”É´9è4ø’SEm æ+XVoÄcÊ&¸L!0ÚÄÑpžñ%@¾¨bÖŠ Ñ# ÄäýH>¥„iÀzଟtà°æ F=dˆ#bIèˆcv¯2@ ‹áM11ª°1R.Š¨!yÒ" ŽX*OÙHÆpi?ECÐ:d(~ò†Tˆ£r0ÄFQ/‚Ï$Þ«a]b'‡¡Q²IÞëÌÚ”ÙdFÆl-ñ‰ 6‚iS n$ÐXJjXX4qQ60sñXë%LŠ3ì=VÇHª:¨—6‘Y±#x`±‡´Ì¾Œ²1'] 4 K’†34&4V05u‰l²s˜Y¼ÉʼVržZË1( Ÿ"—‘àm&µ©0:P”ÔrÂdTdÿ3«"œœUÏ‚OYÁä |p%"¦¸¨v€upqä¡®¢Øƒ%€/×~sèÁ@èá = /èa À P.j Õ8€¯ß !Ϻ¤Ú%Â’ÏŸÆW¿ÿZ›ñ‹Ù¬YŽÆ—÷¿,Ûç¿ÞÎþ3ßÌ'õÜüìà¢îÝhüº¾^šŸ¥ÅœuЗlÕ)c \ƒä…yþÜŒ/Íø‡æª1ã3óÍ›UË·ŽÜùùÕm_×Õtú»ö^6w¿Þ/ëÅ·æ»ïFøן6H}"ÙжŖö²óª™×ºøÙùÕBÛææ©ömµ¬ÖÙ9Ãü°Þk3þûOÿ€Ngf÷Óé»OÐÄ,V½cƒî¼™-[vÎìµ¢?W{tcÏá€T÷€Ð’³ÛW¯êÉmõ}óq¥|öÀ¼¨~/ª9Æ‚®¬èZÃ-°ª¢ByÑP½jcÛ¾ëoh„[²³b³Ç-ÃÉá hy¯µÿ¶¸½WÃ^L«Ù@ ù쬆=Dy›[.¢rز“½«+À§õǸ [¹Àf ¶mëÙ9YF\𙬣p*ŠZDìƒÈÞ¢Až´ F8D‡d"çX:R8”¡›ü.ºå$t¯IKîÓšÛ+馯‰„OH+^†É·ë½ÔÛ{i¸·¼2ÀO„¬¦G$y ô"[„ç½hÕ ül‚ÜáU›óyÕ&‹óf~W-·LG ¸ Š5 å6éF׬ÛׄÉíš°ô6alB,Z ²{ÔÊ6X/ZŠ¤[>qf÷w¿Ôó Cr„9N§IˆÎêS‡èP³YO4-jFÙëÅ(»Y†Å¥”vŒžb_£§.ë&îÚ.û¦Á` ä9Ý&¤‚ÒÀ΂C™µ?õ½n~{vWý»™ï÷Ý^e^Êl±B ÑR<¡öÌî ­N^PðR›í ì â‡‚ßË€–³zµü´z?¼Ä ^‹ÍþÞbƒFyøäµÕ!Òõì •€b‡NÍqk.ôg–ïë'¼›¿“ôö“2Ô ”}–QwGvØvÀ‚³)ì‹?ÜÏïÕ"0xËj6Ñf9GazˆêÃâö¿õŽ½Ø¶× Q²Øö5WoÉŽ‰8g¢LÛ&â.tå•h{V-uï»PÇ©{îè¹ }yp %~F ‰(YV!/¡šë=EùÒ3"Ë©/…dÅ)o؈çøuø Øf;¤×¦(1ðÿIž^L>T³ëzò°ç^çoñ4!š˜­žE©¹Uð— Ñ”‚-…þÑ„Z–ø×MY ¬§ÈN©ËNÁ&GƲyöpêqµƒþr;›bk–Ò©Yc­~‚’wSHîBrœrҔς圧F@‰[ïBN8ñÌI•ž0àMÈZ}…S ³~:¸qȲ×`»œ¾Ö+qÇzÅ÷¶^—ÈK—à»ÓÍ!{˜?ľ˜É[Ýwµ›Ú¼?E\þøê¬-Õ¦·ïgw`s­n»¹YþöÏ»j:m®ÇÅ—Ú`ɨ[8ŠÍþ+ò¤‡7 ‰„ B*ÛHÜFq´¬¯§öˆùMu]o£¿„açf})»'+¥ôfááGÙÑ*z(!šž¬¸ÛP9Yyƒ™Ön,ÞÜ.&ÍݳËêç¢M9â e‹&gõºì ]ˆ–b8H§Xò>ß¼Çܸà\»½ÜºÖÔ‹ñULOÓxç¶mÚûõcŒÂP +iÚûû±@¯ÏôB(2öù˜W+•¨›Y÷øJ}¬þPÏÆÍd²~Ãýö›ë>XõÏÆ‘Óâí·ÃïÞµ~Òßµ$`(êïu¢ÞÎ~>.ËöG‡@¾C”Pe¹2, ðcw ƒOG!lÙƒeî‹åö7%˜å©ÀìÙÙTV`ö*ù³§ŒšFfn3Œ×K~*ŸÌù0oÅ@HÅŸ%?ò«’-(oºÅz`œ3½ó{³?10c…ÿÚÄï endstream endobj 1019 0 obj << /Length 3487 /Filter /FlateDecode >> stream xÚí[Ý“Û¶÷_¡>•š‰â›Lꇤ©û1ÍC’›I:Žh‰wÇZ¢®¤”Äù뻋H‚Nçó9™v<7sp±Xì.~X.ø"‡?¾(ó…•’•ªX¬wÏrWÛÝ,¨ðí_ŸqßoWQÏ/¯ž}úBëÏY™—|qu“ºÚ,^fÊ,_]ýc!­`&7@ÅÕ¾xqõýr%¥Ì$ƒ?ìòì/WÃ8Zˆ2„=O8²À‘d9°ä92…b\*Ï+˜Y®´Ù×ËÏŽÛC³Ú4»ºí—…%„ÒOîÅ®¯éÕ>:­˜Ì„&2Q™OåÍÖ7tuD¨UÅuvÒ…Š}×Ü4QÖ®±_W[œ–_õÑÛƒv€ˆHAä5ýn¡Öw$á®R+ÒüZÜê+ B­Ö‡}×û‡}G…qñ)R|ü1×yÍn@3¬ô¬))œZOtû:ã‘<ô:¡Õã’ƒ ôhYZTôìí>9™éª€»úêÅ0ÈÙй`EîÍÿ Έ<«`tÃËÌ™ã/Èø¶Y7j¬©Àtk`È”Œ¨oêŽ^pÓÀN£ê9=õï­=‰]á_³F®°fC+ƒÂã ÊmCK€m(#÷{ìR3•r2W-ÎíSêìŸõÜŸ‘±ûãJ±ÒªàÔH¹H®™2tyí°&úUŠ¦(˜Ã>±I‘ÔÌØ"ô8eå¶þ…8,Û~¾Ý÷5µÔÛ^ĶCïùéêG+^*fs>•[r²%+ÊA/SÄ`[„ñoê ¤( lVL•‚=ÅP'ž¢ÝL¹ˆz}‹..A±dRñ Eñ.?ÃC?6ü|N•~4žGÃÁòY©'ã%×P1£t<Þ+°©@ãoëŒÒYîU“ÅÏ™àeá¹PIÍä uj&è £§’îS„,t³1Ÿ † ¸M_ ¬Hќν`¹ˆI:¯„®ÀŧtF0¦Ò§d&£­Äàõc,Í`\IÕÒ eèñyzé UŒ£]¦qQ‘6©I‰’•"¹IûhðTÊ–ÎSyYöû¤¢ÙAVH¬¢a¶¼¼†¨E¼ˆÏ/«a}jšåø„R—qN‚fËG…—~Ù¸™h3‚]Ídé ÝHýŸ++H{¡fL°òËWùbM°'0)ËÅÏ®ã'ÀK\áíâ»gߤžŽF”t ž¾òütz† 1W·“¹ ‹ü<•Àù\’ÿÊáØ°™–÷­‡‚€Û ß~7öþÄ—:g\+YVÈrÆÎXìw±•Ù⬕YmÿO­ !ÄÊ‚ÞÂqŠÐŠ?Ò”H[ÒX#Yxô™‘}ŽH,TºMÛÕuö*$ò6‰w3lÿJ;Y«ù¡.Ø"ìþ8DèE;Ïsã¹9„IôÕΗª¡©:Ðì…†]\Ø)Ø_'€`JÜ&˜áB™›\ë°:ÕSôašðöJÂÖ8¹,Ž‹ÈäMÊS™‚Ü¢}0±½L„·´˜’ïBò>U“B&¨éb2Þdâ{½"SžÚñ-ÓÖà*ûh‡÷`&r`PzâÀR\²¢€õ;w¨sÏh¦fÀÑ:°*€B°aéË8“³ñ"ÎäÆt8ó13|: ©/cHœ!¿Œ!k‚%¿†ÔYýß Æ1:~6ò2@´xë‰Æû»?;Èç>C+™_Ú'éîü±ã,{?No¨%dDîó~.y6Ý€1YîÓ °aøpM‰0žØibEœ87Š*Oö%aÎVÚæ^Hß$D^2•—øAŒ» cÚ1¨Ó§6[; ¥çü\¬8W°H6€ó6yJѪxtô‘û"wœŸFî•P!ìì“°Æ%Xí²“þØSÝ™ŒlúêÅõ§÷TÖÕ¡iS**a‚G'Ö˜FA7u_w”+ã‡èï(O`ÝÐë:«ÉüŒƒÿ‚<¸„•ý½ ŠÚôTZW}íò?Ô ‹X½Ýß`)O¥WZZ|tê&ʳ)]ÐD³–¹OG‰^oZ ™JöaÀΦŠ"ë)õI¹.Æ¿æê çÒ§«X— à .‡©«‰w뤌34"аÙþxðDL=ÔÞVc^ÒlпÕÝV¾j©¾»ÛÕi™wnX•U­³Q4ºl´¦ú}»}J5õh÷íª«7ÇvSÑ‚Qûmµ½öêzCuCŠ>ô ºÁ¥UEöõ0¨¨°´Š…ÌÜ*—!ËGc"NÂÎfÓiSøq°1ïU“À ^šø0ì/âÌgàRŸp4 ÿÅ; (>à NMó]5Úi}kûåp¼13ÆÎù7sïA`ü:úðóéB›IjÇ9ÇfÉÂVÁ,­A“ÿ„ZÈ~ p6¹õýÑÇyãýœµÃ7N爹T,Wrâ‚Íâ>5n[ 18¹Žpõ·Š²äOe±ŽÛ<*ÈÂS;åÓkÒœåW´†·îÛ1ø«Cˆ²9¯ ô³d*‚1,ÏABBºÝû·üš›?ý×\þû„ÐNV×^^]{iu§•ÑoiT©à*ÚœÎÃQF“飫²î[ÀBà® ½uîóIU.¢{ ø§ŠÃ£;-ä˜gîe×`®9ÞfXS¾÷Êhž}Y¯«cïÉ9´Ÿ‹pÒ€RðîËB9,ÊK:̆Ú-@„·¾ïaøòçÖ'P‚£àúiÖë½·8àejÏ Ç(ج4×SL}. +#6žy?Ðøk»?QúùŤ9@ÌJ‰Y¹Rþú‚Š6õXäPÅ`¯Ûª?¤†‚ RuZÑUŠMóSjÆÛ+ØêÙrÊ ¿˜s·‡š;‡Áó†¢7­WkJGï{¾JM(m¯1QNËìµ#oá)Ù,(;I¥àõ–ÏûÝ••Ôµ ÀëCûèNOÛø•dï ¥³+ò&ÜÐñî6— (Ü9ÇR»ëHîÌÐTµD¡*|ĨWu ò¦î×]ãÏÆ›˜²ÉväI\¯CÕl©<& Æ7ëJøgJ¦¸tëÀ³.ðHÿ¸p7ì¦5ó»Ñ»é»¡ƒÞwõšBI}áÈÂÁá—ßâqŠ¢þR ¾ˆcu*{ŽÔ ÁÝ€š»êÆ#iQ̧&•e“ù¿8z_˜ÒœhK:8Iƒ-ÂÝ"\ßùR"[‹ŽºÅ_Ø º¿†¿á‚ö=’ñ¥6ß}~Ï û» [ßëÄ:,¥æÈ£SjÃêPò©ôŽ'N1Pðɱ˜eï›B|Ëc*/<$Ò ÔÙ"Rü $†êÀ)ÖßÛ V¡ª\0:—rŠ¦µË(†Ÿ(ž…þðî:t!·Ø·Ð/M !¸t$Ë”tqÌ™"–wcÌÑ-R$ÃÜÇÁ5© EÈ°ƒ;Ùã½CŠ¤»6jqÒÄlfakt½t› ¯w¹Ë’ñ¹Ò[HË⣊–ÚïŽ-Ä2Žxö[pïî°zëü»i×´›0ª¿e@CO¹­¡äûúÎßD%&¦wO¡Æi*üò@9ºÍé‚ þ’)£{¡˜ j“ŒN¢.ùÔRNB»z.>*.ÝŽJŽ‘Êá é¬Óßè-@Gèϱ8CÔë؃Z§5Þ{°ÿø‡l endstream endobj 1024 0 obj << /Length 753 /Filter /FlateDecode >> stream xÚ•UËrÓ0Ýç+¼Cž©…õòƒ¤„ Ã*x†EÛa\GI5Ävœ)ý{®$;qLJ`ºèÕõÑ}ž£ †?äq2†sžU=‹Woo¬>ÍH‹ŠÙÛ…‰qç$(6ãPÅ:¸Có§rßIFŒ1Äß…ç-Å7ïYÉÔ²©dQ‘e âiøP|ž},ŽI¥ÿXE^//É8&Œûò:]6fÓê‡QšRôÞÈPÒ vŠÖª–QmSîla/Ç# Ø9ƒü.ÜŠ.ç=ì,-Mqʲu‹¸ÕÃe8Íó·œÓÕÅhÔš£hÄW}lÆ)RÆ»›¶ó† ZË&¤)êT 9X×ö·Ÿä´y›<ˆ(‹1gy†‰ðIÇc‰XÂÐí¢¸±GZþ<(­H¶õÓÖÒ£ö°æ µ¦‹öºuv%QÍÖ¶¥Ø UëjxTMÏÕv nT×ûµò—ÍÚ#T]n}áq_1ñ%«¦Ô/ÐaBоÔry÷=Û± ‡µ4•V®B¹ö>èhÂL*¦y$"ÅcŽXÀ7l·ƒ u4=÷Œù:ÜF.pvšÆuóUVŒß—F±ÀÄ›w…[cBѲÜm`UÔÖûü7ÀS[OÙy<ìíáf˜ÊVz‹?DÌÌÓ³Nÿ¢C¾ÒÖ$¤ë {ÆqêŒ×÷Eý°œÏÊHË´4wÏÈcíøDÚk‚MSœdGñØßo—Å%™ñg‚HMµÏ4R™=Z•±$÷*sßWÁ0e-—§8˜¥9¡¼ºàmœœ©k×n½>™à¯¼EîÓZM¥œ–Àê|ñ‡q•)ïcÂËe{.ÃÿAWàؤtÙ´›­È}}Šƒ1pñ¢¤¦> stream xÚ3T0BCKscc=K …ä\.°hQº„äÎeUgbl¤±*OS endstream endobj 1034 0 obj << /Length 2987 /Filter /FlateDecode >> stream xÚ•kÛ¸ñ{~Å~;X+¢D½ÚO项lÑ4Awƒ\q=´Z™¶ÙèáÓ#ýñeÙV7,°‡ä çÁ™‘º àOÝäÁME~®³›²~¶ÛÝ0ð÷Ÿ^(¡[ázFù§‡/_Çñ ü<ÈÕÍÃv¾ÕÃææïÇ}qL·ZGQäÅX­µŽ½·c5Øõ°ïL±1ž{ýúáãjF±RžÎW¿>üåÅŸ¦“ã0ü,"å)ð˜ú¡Šµð˜dÚW‘fãÕ:NBïíj­¼KÎRÍœ?—¢Â]À1©Šx—»ÈUä {Û3T®ÂÌ›äÄ" ۤͼr¬MƒèÁ­–ùCÑUe*ñí Ôµã`#Gl[ÙºßÙ¬kS·Ýqi˜ß¬”·FIuù*VÌ:ñë}POz{Óa¢§Óna˜fÞÓÞ²P<Û‡IÑvcNgâ¨mÌ¡Ø+š £j\N—¼± yoÛ¦ê¡+šª{^Ónÿˆg{í°?c9=¡|YÂ'DYê•m}¨ÌÆnŠ¡Þu |Œã+À½5ýÀËøðß‘C+äÅ'§6¯@q†1pcôk¾À‰ Úƶ¨,œöO+°œJÝ1jK0-T&lJÚ¦¹½é„ ÝŽNâŸÅ-z3Lë EŽ5rʆκ *XG`zà®QÂwV‚´¸Ñq•iô˜‚¥¼jC„ €]WÔn«b`hìÛ‘ØB”ÓVJÚÂPc¸ Ç$*éRyYg@‰²XYð—¡³ŸáÊ«#Ýšíacÿ(í"1~¤½Ä¡ƒà÷œ|xµX¯~È©¯à¸Š‹…¨ºàxBŠB*[±AÂ5ü;>èƒUÅÀï?È*º0ÔÙΔÈþ€z‘óŠ’oaö䫨cÇL.†Jä<‹¼[ ªZ‚Ž –¨æ=fRõŒáà@ÓâÌ TÕmÃðf º~üå—@‘ío£Ä’p–5é¢lÝx>[ªä‡~IK%RœˆÒø"¨#.š¹FaÒääé8IìÂ/8[e¿¦€èæ"ˆ‘=6;H°Ç1ëŒ6Àï†<ƒdš°Åvèښł)÷À1Ù@9¿"¡µ÷Fî}à3…ÈõE^Oo3¬ç×È°iQÖ ô¦Ð‡vRt–Øâ²*9Ð’W“ìx<ïÕa-šRhv`LýŸö Øs÷O/…Ž§Ðò’¯FÇ [ N.-¿MñY8Ãìк• [ ÃÈÝvglgÀš:³aZÊ!ݶۮmìWŽÆ œsÜŠ`e]8‹„„[¹v8ÌîÐb h8Ïò8Àsº1†Ù-Ç,Ù_T°£ÇÊÔBÒ3I?ÂÞIiç ÓUÑíŒÏ‰!æ—ÚtJbæmÊ}žâó®éWž9ò ŠÜ“ÓE!:Ë;ßЛˆn°yZE ­áå}»vˆÿ“—ªÜW¡dà¯0LFZBF”‹J!I%¡ÜÕhŒJü(ž³‚®$w(˜AÀàÃÝ0ˆ{2+Éœ•0R¾Ò!°t¦¦5dÒyìûñQ&ˆág´À I ƒ¹ÀÂÞ“ràÈû‰c7¼‚ÚÏ£ôÜ~ TÛ:†èø¹‡LóCc1_ÄŒöئƴ'JC9'Ñ˜#›»,Wö±ƒ0Aq$qèl”öŒ¦YYÀ6,i-§•¸nÚ £¤·ùÚv˶Á]v#x-2§Ù¡l%CgJÌ–­Œ•;Wô”îxD‰'‘ÉFKy!¢Ed€æ"°·5^ø6%WÐ( …¦…Xa›²émY°ádëµi ð¾õ3V’Çx#Û=Æi ¸¡¸Ø!›¡d12»pv–Bèv‚ÛÝZ¼Û¥3ãØ’É0ËÎb"ÌÇÅAoÌe© é©ñá&‚3S)•«~ s~Ã?RÝzŽ™°nýz¶ÁB¡}y ±wÞÂÁøbÅÐ/yÄ)¢H˜+jŠJÙøWƒ[`õå×[¸Ú‚<½: bÒe”*?I’oÔèŽz=#_ñrSbn]Ý ŸÝbÌÑ~˜^[U{0M}XRp–û±ž¬Š³ßsòK!£XÑ…Í…üͺõßûòbëäÍÛ÷Wlåš-áwݽ£þ98H’Ÿ3!žé‹Lr–ê?p]¥f…R fM qúÁƒ#¦ôã¨=H馃43­N…šeó–§23ã2¨ÚZ6î)¥M·‚9«?žIÂD¥ÁÔ ¼{ÑAIå‡j(ô¨7â pX)¬D=8¦×)õ>­0CÛ ¹+ï…†#4BïßÝßýì6ší‚qGäÆŠÆV‹… =GBŽXìQ¯ yïúL¬tVúôð…1gPDðÓß>¼ü«å"ùË<²FÞÛ¢|wÏàÏè• ò;ÀGJ¢P9ÓYk­.­ù:Ã$bj>œ7²œ“bæÂ×qT¾ÁoÍyäê¬0ܳ‰œÂË4þŽ4œ{»²dIr:¨¨úöêHn˜'âÈ)Ê~mVµë*iøÙÙ©)£Ç“µÔ%D¤¸À¬š@Ö¨3³bàq®‚}†™_+‘œœ|<AÛŽØk.1Ð!mŠn$ŸêÈý­$(þ»™ó"¹@<)%Og¹6Žhý“LM9'jzZÒv”Eá.hvˆªmckª4y!&F±-‡žÉäJ©VæZÄÈ^§:w©s T½Ï&€•ä&œA4«B€·ó ƒÓÿ\ö`U³ÇsR<‘4}fË© ™îÔ¢Yf'¿zwËh;¸#Û±Ú, 4¥—!ÈQð¤nTm" Ù&YÂÔib(°ƒ…•4¥xäÀ;ž¢Û´$t²Ç…;0%•“D:üÔ•Fè9DF.#^z½!ÑðÈØÏ+†—.|”KøÚ-š{wÆ©ÇE}8ucˆ–‹ª`S‡ËLåÊÌÐÅ HB«©{ |ds>”ŽýPGŽ÷ÇжéEaäçê;3_}Êj$Ð\íª|­¾3õIõ”sYïj`žcüÖ§±Ÿfê"LÛÚÜr¿¶·ÜöÈiõ‘!©Dx0¯"`(…@s³OÉìÙÀÉB‰íß‹Ò&)pŸ9î·Ûá)ú×3Wµ Ž'aàÝù¹`ñˆ(ôUžŸÐ._%$m³,ò‡[¾7ÆP€æa”¾ôrp“ ª ‰[ž$,YÁF§ÞVvÖÛʧD 3ÞCšbÔfËg½š4wý€*Û|brðð›ŽÈŽ'“n›Å ì.É&Ÿ¥*=¥>uÿpï‰é–'k’Dù¡:w`²Æˆ£#Æ0¶E×½ÛÎĤ++cºïØõF§36 æסóˆ¹øœ˜ëÏ9*äJT…îµBp=wfc\B‚ƒ«®஺VS24õßçÍvlHº=öJù»™òžíH¹†TÈ ©=—ti,J[ø4 ¸g>€R£)çmßîÚ嚟§Lqï(áû~¬Ù¶C÷fÊ}‰ )G¦››¯©{d䶨me‹Ž§$xÌwEù[Ü|nö-…¾gP—ã:Q“+áÙgO­Sx8œ‡fä¡àÕà›½y’ÆkÈ þr‡ $‹ 34•y±X0<Ɉ ÎöfÔ"2Aí Ù•½ôÒFäKs³“—κJ^ rúªÃu-з>çtL¤núó˜a endstream endobj 1041 0 obj << /Length 2924 /Filter /FlateDecode >> stream xÚYëÛÆÿî¿B@?”,šK._nQÀEã4EñFáWµÒ-B‘ ¾“ÿúÎk)R¦zqpÀiwö5;ßÎ Õ*€?µÊƒUE~®³UyxµÝ¯¸ñó÷Ï”ÌÛÀÄÍdæßož½xÇ+øy«ÕÍnºÕÍvõÑ‹ƒõ§›­¢4ô“ ]ˆúúõ͇õ&Š"/òá§<ûîf<'ÃßÉÎ|š£$Ó¾Š´œmÛ®¾†¥±wl›}[ºõFg©7t¶Þc3ñú{ôcÑUe*&—ÍáX™GêÛ¢îvMë–w÷ÍPmyâÝzfžìQÙú×5v·Ü°ý=^8M¦œFYâ«,tBÚT»]ÿÝ’7np£"_Å<£¿oM±íX’2;:9a& ¥@ ¹r'4õ„—yïkûÉtà5íg ½8Õ—Œ0¾áè(öãçùâ½ GÄ0±ÏÌfvÀõ‚‘½¢>-qG&Aö5üµÍÐÛÚtÏÙwÄxÌá‰c€§2!i7Ôeo›úå’Ì6*|¥Ôü’¶î÷Ð]ommû[”ÿçÆnQLYºÄÏd·ˆw»¹G7³`dnæh*¶Ä‹Üó({¶šº:qkÄGìàåxŠBëT^¹Þ(ÏÐn'¦Ò Òâ‚¢ÆíOn…Ùôö {á¥lQÙ/2Ä´Öü6ØÖl—œA ‚9Cg¸qÆÔ4'_AâÙE±Ç. €ÖüÐ3±5ýÐÖ²î‹ie_Ä9ÞÖ´rÂnVt²Ws²i[t<\ŽÚ`!¦(D$Æ€mú¥«ÜGçVl!C½EïÄf Â+*±%”®Ed¢:? ±8 ¿c8ÀùE-FX°⤦Þ|¡ytOû¼Ž@ÙÁðœïû`;ã;N/,§½îk ÿžbܨ`â±*äR=à·Æ§;,T²‹^E8s/¸ýbž —鯼lÑ&Æ/®„Î2ñ%äï綞xÔèoâÓî¥Ì"+€†cåö½†a6¾·E»f”ÊÀS¶¶LGCT¹à ÐiÎpÀÿ¢›–š›Ù‚T?¼7ðb1ù,U˜;ª#uÇ æA‹¿ìrŠÍ½‘CŠ²6ŒÓ©òcÏmtZg‘‡šŒ•0­sïP<Úƒ0„Å[ðXî½"4…i¨·Ž›Ýp×PˆŒ¨JÖHwÝ2A^QÈmT“¤ #¨îÀaSwå›ðíså î@"wÍG0®•bl±„sÍš×ÓÓ°`(*ƒÈú!\1ÓåP Èq%ÈZñ( ãrl,×°÷½˜¹Q˜…c”MJj;»¶9p«àªSX Þ÷•á…,)=‘\èz136'·Ý¦Ø9îT°Uon3æbÁím=¯&\©kp‘'Iuˆ¿©…Pl·•@Ø}ÈéèÚ¼ÄAÆ•¢t6싲P({†_xÅy~‹…¤P~Ž'Qr #IÌIˆµ›…•YiºfhK³xaØ¥!œ-9ùÀË[fÄ´5åáR¿!W¥6ÆšK5–Òoþx¥¬LQÇ?T¯”'eI²†t¨ÅÎ#Ú«ì¯TùçÞ•QaNå3 &]Dµ?h?SchéB>‰`ðq,Á{Ç@Ñ ËB‘¯!YAÇ-2¤g"))¦¨<÷³xþù¨5•$–9àDÑÛ©ó7¤sU¥ëkß°™ÎŸ ™R5@ñÁ\O ¹Ð9¶æ³%¯ /'§Í£Iú›ë%€^ìzʧs=–ç©ÓsY:?WÈ]Ý?cH.ŒâÖT„›¦òXDŒ-Âs99—Æq%l¶gVêñ¬«Ï§uek©}åÁv[Â8h ÚåTõD‚|žä~DÜãžXå¥×²¶ªÞnXý©Ñ3±)æ-9{I±YàŸc÷߬¢Fõ’}Üaßpã¹K) Š‡÷ÿü¢¸´ .øg¼vÜÇWh2þ3¯†8X±å0Ný1˜~ e+ ]Î -ªÂúF ›n€äù÷œ‹Ô3õu6˜§_J®ÆfÓÊ’ßóU–Bž\B` ¿=“å›óú¥l~È<CàGwj|$hÚmÇ}*€ï|°Î1o¨4éÄŸ SÃ{Ç–-uûˆ‹YŸä³GjHVÙW‘ZŽòŽWiî‡Á“šLÞœg/Åió-%`–sÿŽÐ{L endstream endobj 1049 0 obj << /Length 2948 /Filter /FlateDecode >> stream xÚ•YYÛÈ~÷¯˜·P€É°Éæ•<N`gÀ^#+cx‹©%1¦HMzá±u”ê<óRG*ÕÌc¥›0Ë“à‡M¨‚n“&Á0U£¦Ý¤*xäÞ–vÜ:þ¾wöoÈííCÀK©*R‰¼Ãöd{ ÷ËŠ vø-Ã݃©{n™s7¶›¤ îw^ØáØ'ügûp dZ†… ÏÎ3ôuíÇO}×Ö_êöÈÃüîî9ôŠ4pöºËPŸMÃGÓãÿßÃr€—¥Û¥‘ÊørȨ.&ÊÒÙà[Rst–{{a*­_5âÚz ²ÒÕ_¤5§<zÓºCןyÜ¥z‚¦¹ ;BqºTA—*ðRþ¨ååßîÒwÄôÎ:×õŽÕõit#vtp2³„¢‰ÄBú/pW¥Qäø9ÚÖöøÄØéÇÆ‚,t\-ˆÂø¾kÿ4`3Å«Á¼jà$i|ée±qîzË‹'!кl£Öí%ÃÇ#à­ÇÁ/·~ç¼XÔ‹;ö3™›Ý¤å^ùy”Zm7ï/p¿ù*nü+ í!‚{(0q`âuËËf7_õabë+â¿oì™ÅÎj‰2¯E~·NæÁršç¼b¡"ip¦äz˜YžBgXž0p°¬Ï=Ï/Œ‹Y¨ßÒBž~"ÔÕüϦLE/J/Žªz¢!•¨\²îë³u²DÌTAfâ•Â]Ï=÷èzÆ-g­ì=™‡ˇ hýø½0˜Æ6õ3Ô]Ë»H0)ºã†5ùR€uezÁöP¬ÀjCŠ‚P’±tAóË8¶$®K½ƒSñ‰ÐDEI¦ý¤¸‘ß ‡§w#]Ä9¼+ÎLx]2^O ð渠±ÆIÓð•€éÃàu¦EƒÐ€f0Übét|Ü€/W¬ãÀ& ³BéÉövÝÙ^ ´-hW›&b€|}àùYï€Ô¥!uÔb•ì ó¥/L“2Ò9ðNL£§ÿíÝ‹íë—o·kž3/£"™V£1éõJŽ€—Ƈ@ÓT™_ ÛŒCw a±‘ìkgî;Ù‘€8&ñöíüR#Æé‘šW#c‡¾t¸QÒˆow|€“"*«ÌÇšc-ò´à{Ž0œ9ØÃoÄÓƒp„\Tœ²="Áz ߟÆà&ÈõÝD_1!ű7gÇ«öëv׌û¶û½?zßýæݱ!u¦‰ª‹”vC!SËö-ƒáœ•@EÎñúr£ÕÄpÓjyS¸f‘&^/˜PË$–"|]ק¼ðÛ(š€-lÇÐhêûÞô5²¤U…ã<Pü?ÖìÕ/i ‘áÉ¢ÿâÕõ4;3"ô;þ¢"r«ïFƒ••`ž¤Å%XFœÜxQ#TªdR¯dÞ…Õ9YSg ­jP㪻‹dác¡÷Ûs«¦iØ=î–NÜó›ã?ÕDÞC,õ\È‘COåXTän˜±µ°aRV3K:´2 [îzX2?¦•H „nÓ¶$è/zîdz!ɸˆƒ¨Ô6:Âë¹Õs{|W\C JäûúصphìÖĉæÏh@ªG1šµâ˜Qìâ­¢P''-]q5åu+ŠÝ©£$„Ã[°ÉÚˆ ¿²kæx¢I„t€9°bàÞ¼·¡8ë^¾‘Ç“óZ”| Z Ük˜žL+ííêp~³Ÿínìš <@éW#Gø.d–©2yaª ±¡éû)zRèB³ô«6³‡ƒQ *Ѻ‹’°)b‘¾ €æ;¶ZlÎÊ‚½Y/& Ãá «—@$cƒ¢T…<;ŽÎ&”ºÝ¯!dÄIŽ³Üð†m’+~ñ²„È*bœø©£<-g…kìgs¾PÊ€s˜<1î#CÇ5½¹,(‚Ù³¹œÈûc·‘¸7”…ð*­g‡Ìü‘;¬Ê+w¢È·’(ÌCžì¯î` õ-/âdW¹Ø÷!U'‹+T©\6ü— ‚†àË£„¦Ø Ü¸JAŽ"Ž'äÅqik=£øâ$hü¼ÁtÐáf픯£-¡UôJ߬iÕÉ®…Z¦D°”^*Gú~E1{‰³ïVS£·Ý Ù±i\ç³k3H¶yñÎú4›I¥¢ƒÜâWKE“ÈøUk¼±_!Ó‹Àå b Y|/Ä>‰›‘ÉZJµ8B™Kià)ÊE<érIÐ 0ºJ¯Eå ¹ü¬à;¯°”%ÝL¨%¢XÆ7Óý}„SH„#šN ktðC7KWjbŽß£€u'; OëO>¬ ãŸ>w)ÿ¾¸§ÜìíÁ3ì¥ä*We/’³âªNLÕ†Ï,U„Sï³$wŒ³e ˜´~) BùZNÌ€…#lJâîyÓµ7¬ÔR|qp[BÇs-˜bßSu^Eàêï’JE9Ü ªQsQ\Á?•PeõzdYbõûÕRðí1ôf?Y}ÀŸŽrÉå ¡¸T¯oon žý%ÜÅÎQÌáGQ1RPb ‚°zrE¹ÈÌ¡³C'¦(Üý}YŠrQ¥GT•0º|DÌt娔‹«ÈÇu?Ê$u•OÎÛ-¢® Gq 8¢šÀ°¯ 9^ÆŽçÌåâ›”ÁwÎ;±7R©QŽD€Ï‡Çu?!Ï”à=—)*(EÄ£ï¨Ü&©¹;󠜅Í7ï^_“7ý8Ô¡¿¬£Y9r¥ü.3†?Î"Îqàƒû=x~·ZŬ¸\Ǧ»§q¼ˆw1ùàiŽ0È2SÊ_ó‰v ’™"9 ÆÎÜèvö2Ô\Â!¦‚•-YO O¡n„†+ª+9 Uˆ8.+ë¼ËÅP®k%c｟D>ÕTÒ%k×Ü¡* H䧌#„„Ÿ$ÜÛûñ¸æ}Jéb*tý*QŪZ§b1·Û­‘V‰Ž²<™ó•™£›ó™"¹v_Ab¥’(ÜëeÕGb¿?\XÁ¦Ûc®‘XÇ:@÷έ¯[7À½ŸÌ¿è}[F|ÕIBˆD‹ ªâöbi’Fpöò^_GXYü;—¸!9á«0JÓ®æ©E¤«ò»ÓTY:ÅPTÖ‰5Û 6tB ¢"ñ\AëklÚ2&ô kuÈ°A“$£d vô Iñèž'¯ë¸«SW¤¤Îk[ªåáˆÔ€çŒ0´÷EõÚ½Ÿ]?~åG´O|á"ær8uúÁ¦x;‹²ôF_èX·dA0UþK΢<ÛTú£ Ñy5{)˜à€[Nb< jùEˆKTc’C*0^pntª¨LÓ™U±ñ¬Rç׉#š;T©RM*ù`Ê£¼úîØ¿H¢t†£¡»Ê‡–aô¬»h®ÿæçB endstream endobj 1053 0 obj << /Length 49 /Filter /FlateDecode >> stream xÚ3T0BCKscc=K …ä\.°hQº„äÎeUgbl¤±*OS endstream endobj 1059 0 obj << /Length 3125 /Filter /FlateDecode >> stream xÚ•koÜ6ò{~…Ñ/ݼ¬Þ;ôÃÕ½.$häC[ô‰ë¢•TQ²ãûõ7/JZYnsb‡#r8œ÷úWüó¯rï* C•GÙUq~嶿¿bàç^ùBwÂÂò»»Wß¼Žã+ßS¹—ûWwÇåVwåÕ¯»›“îÓïaî’ìQﾯìÐWÇÁ”‡³9·ý¯¿~}÷¡Çj8!íÞ¼Ûû»ÛýÁO’ÔÛÅáþ÷»_ýûnâ)‚/d)ŸqŸ÷© ü8î“,R~1÷Éþ'Á6¿iìøHøèÍ»[dq-\¨0ŒyãÛfâd7œ*ËP±²"a.[³]1žMƒèÁ}-ëîu]›šgÌB};UcäˆcÛo‘Û';˜³Ðرë謶‡/ïÝ9ð¯…÷ðCåÇ>ßo{Âxw6Öê{sè´µü +fÖôG]oï»÷ ìUWŸöAJ·²0'íI÷ )Óò©7º´<)-à)ˆESNÇÀ‰8Ñõæ¡jG»ÅñBÎ×ÀQð5I‹«–çO8iGž -Žán´FÞ7[¾¯§*ÍC8N)'ÙÇþnºÏ†bä©ò³À}ç^¦²gáîñdzÃ,ÍÌóìæÝ{NZø­¾1¼Y¤rÐl¾8ïÌ7I…~´³æh&ÌeáîéN7%¯<ž*w$N í¾/t-Ž!ɇhêÔ©·×Âý"8qO¢dNY8e§0Ln–œv„*€Xkd™~ºrV²WÃk®$£P"kë¼q÷d Dš #h;ýç(9»ä ¼$úoÉi¯hÔ’ðkPQ.ã+,YbW @æÙÚAvÒ¶a¯‰ã2%ÂJº’ÃU¤ ¥ÁàÙÀñ[ÅŠIï’ez—z*¦\ è›·oÞüñáíÏ?}¿•fʼÈÑ/™ŒŸ#å×AäI:¸Î=S_ySIz¹g½uŽ+ð.E]Ç',w^£žSݧ™æ¡j0³$ÏÂK<ú(9.  ÁÕøá*ÌPò˜åàpNTéX´ ÆëfòL-Fàz`O{‰e6Ë°Dò‘ïON—„̽iÓi±_ÃÕu*sC_õ°î>ø^E}x•¿Êó˜:*†5åÁs+? VÄ%fÙ“pßltUÖÇ7SF<¹ý„8Fu˜­y͡ȸdõåf‰ÿ_—[[iàf¼Á2Þ|@eBc¨ ¬Ï“c‰綺è†ð¥¸±T(¸7¥è»pwÔg( ÁhuÏiÌ,¿-'§’·[®MÅÀ¿y±F š¶™0±ó"..Öáùèú6RšHýHЪ¬™gy”õ–Ï{Õ\!ëÆnV¤Â( ¤çd*„áDùÜΚÛV¼·…„¸Ö/Õ§œ8øàÃ8q ,IÂE”ºhŸºÞSäöF `!”BA©: k‘¦ü'’øâ6Ûp¢A(Å]»j‹Ó PáTHÛ‘#R ¯8N%#O]*75òÕq,ÿ´!fÝ57T¢Ä¿poño÷¾©>£Â䮑†Jé’ëj!H –8×ÕÇ^÷ÅVÀ³Êþä´uꉢ7L¡Ç¡E×[på‰k  7A¢Â$º ¢PyQòwñF¨ ò­€³Ú”¸@¬\[Œw¹¨r\ë1 3WaŠ¹¨ŽxÜB躡k6À)©F”P¡%ñ²ÑÄd>9¥K8¡r“9¤öÔŽXá"ž:mÿIflha¶jO]4=·b r?û2UU’&ÏT5Í¥â S(/ð^͡왳¥Ò  K »VêÅ7âlÉŽ °uÕ|bÔèÍRˆQyt¡W#ÿ†€­›¹ÂÅÙÔþÙªpIÓ/+H`)B¨ÐÎ^³ˆŸ÷íp¯ãÅQ<9é¹½ÁmÎòë¡êjÙù ê7ö÷^²`xJ0¹ÀôÒ¥FRjŸxk¨ÓÜ0ÅÈÍ&;R®9¯}¤»IÈÀNöeT“>øt_$9ë©ž@úÁ¼LýæuUÚsó‡ª}NO*ëÒ’Ee€KæsWWE5pgk³ñ†k4(Œ©˜[#éLEvƒP3‘ vQµº‚ Yf8uÁiªf’cû‰œâ— qq@ÏëÈè0q?UQIëï:±ˆ¿ˆ]SšR›-_%ñT¥ños€ÜàÎ×Qܽ{Á¤)„ àK£Uœøk€G~jZ¼þ츬yÒ9Mæš ; šúhˆ¦]¿ÁQ’¸¸H"åO?D¹_)Fzì…þðÌÔÂDìü Ç'ÆH¡¸Çä‰\ù‰_󱼺Ð8šÏºL'ø/š¾” @®¥òØ_Tí77/‡F5—Fƒ˜ÙhdFfÌS6|¢°72{„¨7Ÿù‡¦4Õ´Šy¥s²]6mW¬úK"äK]Š\½…€Ф;8y¨ôL¸ñ"~’«0ËJñÓíw¿l =ŠT¤Ûj)Gq!G㦤‰Jòä¯4ÏÿöPKfŒØÿüUÔ endstream endobj 1065 0 obj << /Length 2212 /Filter /FlateDecode >> stream xÚ•X{oܸÿ?ŸÂ@œ6°hQ¢[7zI}paçŒ;‡"wØÒZ®WVZHZ?®_¾3œ¡V’å&#á‘óüÍpåIòdœ¤Q$–*;Éwo»ÚÜŸñÓo$óùÀè8¿¿}svÇ'2Ë`)On7ãn×'Ÿ½X-~»ýÇI”†" 8Å®^\Üþ²ð£(ò"Èòæï·ý=q~£@ÈùR¢d"Q’)!#Ew !ð¾©à22 €ß‰…¯ÂÈûTwfá‡iàu[ÝUlh|^„™Wš]ß\Qwnž™·å±j;]–fí¦4jª…ôêÊ–j­æ)k3/×]Q3û¯AÔU/y½Û¥i¬¡}Öב1)³¦“ È%¡WÕ_ªUx¤™¾³\õwïpï™èµÙèCÙÁýòW–NyÚÞéE˜:^]¶5Q[LøŸ¡•ŽwÚ=I”þÊjƒÄDm\ª7S61*°ª’t ‚ç²ÌÛ½›Xú×@ªÕÇÉC¡Ùÿ£0‘AtæàêãÅÕß~øy.RT,”ìC/œ9.ÄÓúpúpsóêq±REŽóa'žn }ƒ­c0lcz{u™³|^d‘Çûš5èÂñß³ú©D$Q/‘8ËëjSÜCqNA ÇÖd+¼÷¯ÎöºÛžuõÄÚØ¿¥ôuŠ½÷/GE•—‡µ!¦W2.ŠQ¢hr–)&%"•„Èä;.Ð_È„åJDªC$\-| z)ª{¸+ÉÿQª¼Ëªè ]¿÷[6C‘°è3#†ÒRȡ릩ï½CMeàZ:Hræ!A'ÁxTùST†?j·õ¡\é•züe 2cÎÞ|,ºíœ×¢@ ÷Áæ—›M÷­ÀÄdZ^À‰‚ÉnN-ß2r-&—Š–Þ?«â (E „+ëúÁGô¾m!Oçc •N¤Û¼oóåƒèéRºSœho+ˆ6ÃWIÂYŽL5 â0ƒ(‰²ü!j¶ÆL‹Œ”P’u§RËÐÃÌà?ÚŠ3^–÷½?8`¦ N¯±"ýŒ™l=ÄÔ& g$Þg(7 «j [KÔX~CÕCåíõ½¡ïB9U)´U6©ôzítÜ_Q`z¨³©„™$÷/D"¬(ÓcQ–DQ@¥*ƒQŽk]M£~»gãž¾„újÜÇuDyj ;Œa¬®X¥ Sƒ¶ Ø…ÕMÝqV½$ÚÆcñV!Ô£ç¶3»s .†QXÜÕmÇÐסWìö¥Ù™ S·£ ÅNBï(…ÐÐðõN8ÍWô%O—s…«¯â6q¸tKYSéYÏW,Ⱦ#~Âçy>‹³¡Ӟ͖`¸ï£;5a•<ƒa›e´‚¯`OD‰må+£€sì{Ö¶q ý4:¼—8×w8¹€‹[´ær9L˜™§Üì;b%ƒÕ-ºôy¶Æ*àس!Ô¬ €ýU·m ú·³`H‘†=ؼ‚ *H "hD¡ÖgñÆ÷½?8`&µ¦×Œ±í‘)¹>§|ä0h>¨whÜ;PÛV§lÉŽƒ6€œ Ä-™Ã"H1‚àF’¼@H³tCSBl©¯!sûö9™Ú#ˆ:MÛr‚èzÓÙÈ2 "PiÆíËHôm(»ÂJÿJŸ–ɾ8ñ]ǾîëË3_H«XÞÉä9JŽ{éÍ¡²îùól@f"ã¼H¸«‹56i1E'¤œP ;ÜB#œÏÖìÁQ|óåU¬ÅwˆŠ#jC0¶2ŽpƒÙžúþº7n=â„òÓZÓxox¡)&‡[;!±3»ß9v±¿4-È›v¶O/¾að)HÓT°ûLkwîe4vY§DZlG䢮ªÆ*‚S¥Ç Öš±–ðÔÚêq‡<ÆBâ“®xÂ|ÍÛñ«ÞÎK£«Ãþ9œcçq[ä(HÉõF›ð‡áNY|á7Î8|_T hªU G`9›“ ^Î S0Ž~º©k{«fª°£{â ÄËNÁ$9˜$¹ô»– Ë^ Y2J7p—ߘ’bDÛÖºÓÐx¤*4°¾Ã89´²éÓ'V=fàûwFÓÞaÇ_(Ì“É¡ÃuÅQˆèEŠúЖ<Ï\{ñö¥®P×(v`fYÌì¤ÛNá$I¬Æ‰áL<û ¦Ï ŸA‘ò£çú#÷’&æɆ1$v-ßðàqE<Ê"=V€æßòƒí|,TDY|̵,v L¹ò³y£s>;+ÜZÓ‡z(lJÌÁeFhĬôc„¹§O7Šø±¦…Ööb´dž´SqŠ3å¥+ŽÌÛþð„ý)W)ùKÆîÔ°ÆŽû#Úèècdö×XŸ$Ôν蓨¯>Ÿ‚Ùf$¶Ý±ðï-s©ƒ>}›óI¾zTŸÝØÜ•è”úãÅ­˜;Öˆ¥t aê_îc±Ï7€LlÿêÌBØ¡{¬uĹÓE…8Ò/èæ>?%2ßꆨwï`ùÁaMï+:ë¿v-¤Ÿa&½qÕò±û®Y›ÍŠ§`W;¾§AÁw‚•&çsáaa“–Oêu¾ÞªVD¢ó5&b¡ËWðO—,š¥WU0š«¶ÓMÇkÿ9Ÿ3ºkzÐÞov~ëŒ;'Ö‹þcÄ8<þìÝ×µí}0¸*Î…­Âå·øÝLŠï¨èÏGz$‚¤ãwgî7Í@„*é3°ëÈÓ½ždW”bbÝ >}l0†üðãõõê—ºúxJñ'%ˆ3yh¾øìíÈiGKú³_“M¦2’S(ß@£—¾‘ßî*H e%µ1ÊuLØeù°š›¾‘¢&°Y Å¬g ‡Ïd6ö ¾þD‰Š endstream endobj 1071 0 obj << /Length 3026 /Filter /FlateDecode >> stream xÚY[Ûº~ϯðÛ‘“µV¤îç´inÝ¢Ir E¸Z[^3µ%W’³ëô¿wnÔÅËmƒ"@–"‡Ãáp.ߌÕ,€j–³4 ý<Êfëó€f›»>½{¦„n„‹埖ϮßÆñL~äj¶ÜŽY-7³/Þ«]qìÊf¾ÃÐK~ž/¢(ö^›¶kÌí©+7‹Cy¨›3¯¿}»üÌ£{Óípyï+ïf¾PI’^Ï¿.ÿòìͲ—)Öú…GÊÇÒ'céU¤ü4ÈfIù*Œø Ç}QÍawìý‘ÿl·Ýýêp4+\Ym¶ÝJoþÄÁ‡àŠ >(lŠ®x<|ÿëÍêÕÇ÷ïWŸ?~úëë+¼ÑL%&   }óÁ¨ŽÕÛŸ>¿üôZ¶Òԛߖ7ï_.ßÀ™êÚ½è·ëÄÍû¯Ÿãâå ©ÔW™•©LgŠ½ù½dÅ£˜¬ø®æ™¶>ÈÚöT­;SWÂuª¹8òã4±lç•¥FÍ<\QX¾Ãók‘9eÇÙäÊÛºa"Üe&Z~á¿2ûþ³¯×Å~UÙÅ/ÌpЄÙ7'³ofTÏ污—a}±”ŠÅD-}1Ï?ÈA/„Ù×É:`IƒUÛM7Ùgäq¿TN¥÷T?þžëúp¿â뚢jA ‡öŠßÔT °ÙuyÅëEË`à:õÄ ;s(Û1AämÊÖ4åÆõð: |¥•=ß>®ÎR<œª§|(× !ªýg¸s0½-ÃÁ]SŸWBùıwô«·¦"kFâ‘:µF'¦öo—2An?Í‘0÷S0åKÐBšdÞã@š¤Þíè¼’gÛ²¬xTà|êÕ¨Ëïs¤=giâu;¡ ™ƒ©pã^ƒ×¶õ~_ãì½'»R¯-Bs[´f-ÇvH{ÞËJ½eÚ˜,¬rƺ_]yê»–¸fÝxå\^;×|0Ú²ØÀ‹›a¶©O©ÊÖŸ/­¼—{¡@&ÛᶋöȺZƒ_ªhÍ”–ƒK΢ÝQ{ǦÄ-`vðIJÀ•¸ì0Ëý<¬ö!Úõȱòã(¶¤?1SSµ]YÈYx•'OJ´äñä$×)`II˜ö§ ¶âÜ[îJ¾`4Ü´à?Ûò^D9°ÒjˆÊcßìxÏÕ‹j>ª¼+±ïPûT86ð[âwäQ6?÷GóN%͘¦í®ðÄ@¨a ¦tÀñ 쌾 xíŸÒ{’<Ò;æëƒ.íd™Ÿ©Ìn+¶„ì9¦ºsÞ~ :û âbëžÅhÊ(‚Ñà}ð7Agá òKAáM}×ÒnŸ@9hòi~.ª ‡I‰i%ðÚKPð,*ëc°Rƒwr—ŠÎÿgÇÈU|Ë—4’ƒ](=g¿•ëºÚ`ÈñÅáž;¸v‰PL˜æ½àxÝ”E'cvm0&"RtK×Eà…Uþø"#ÌäÌf pDÔû1 I>ÿíë»ú„1 ЉãP@y6=“Îce„iDv1MåOŠF©Â‘$Bùà Ú;m‹#’fŠÍÆ`¾/ö2ßÜ%Çzľ¨× S£ÆÙË*ñ¹n䆥<œ"#BÎ÷;ü@¹“£!pc®‘(Ж"Ƚa/…˲ô“»y˜µ9Ò{ê0cG1Õ°ÐØ‚þDå}¹«|ì’®uz=`…LU£7n†«Â‰ì•a*—Ìð’pAôE¢‘ ½Ï»²*9Ã6¼çŒõÉC0ö‰§² £1e‰'ù¾ [£7¦±g³!à”¢roa-=a8I+2AQ]Cqš´&l­@´>DRü¢HJŒÄ`ªnšRR%¸¦Ä¿ËÔÈ)Xg±urüH¼gVÅC€…cåí.ûC:›êœO#+ÉÆà˜îÄlyy˼VÑÉÈÚ>œÎɘ–ýªAXªsÒC+—)zK›Üf¡¨È‚ ß]ÏáNµ\­Ä@ÛßKgJHüh„’á ZŸ+…MAP¤³ M“ÌÅíL–9w¤ñÄq`þÄú“ô¯6EãÄÑ¡JyªIìKuUpn_§á£„ä‚è'¦AQ`µ‹qª!FöŠê‘œ—ÐóÕÙva`3 w§6íF ;pãAX¨ÊûEÑ4…Í“0EÙèÁ¦%«…Ù"|_Tc¶Ê’¢e#ÜGÞrɘáy¹'/¿çx2ÆC’—;ÓÐ5Bͱ( ƒÛЂ[Ø^…’0ê±²`º+þI Ëù"Aê§Áàt]³1ÛíÊ fâØWCg”–í‘«Èé Ô¨K§CÑï‰êŠg KúO ³Z0|·¼Ù»®¶;Ä.éËòKŸÇtPüÙ]Öf{æɸgWnÚƒaR9 Ê·)©*ÌG=^%Á|ÁŸö.8¦»ÀÛðE’(¤wùÛ< 9“Â.*Á‘vˆ#Û3/9B1.F2¿“gæd(Ž°gÇÉ`É€MÐ’¯xšõâÕø ìöÔñ|]íÏL|kãWŒÑŸñ?L÷Œ€_CtÖIoèw§æ¦ðÆh?ftR‘½ÇM³bX¿¼ì˜F‰òU¤g±ü@)êšú1¬AY;ÃfµO§3ã>ªÝ¿1pt‚/±¥uIx@ú~b¡¹ð¹Îû¢Ôß^8ŠßŒoùU:YÇâNJ¯0½¼¯†’*BÐtèg‰ú]bK½‘;nwÉ”$ö‡zQÅÂfRi³)7ÝÅÖ‘Tœ²&t¨õ¸øç±ó‚m²cÀ•@»©«Ÿdö}¸"àbMÛTI¦± Òð»ØÈ&­”,( çaFÁðÉ!xˆoN<ç1dÄ@ç†SÊ%#™£èÉ>‚Ò˜ç׋Ü6mÁ{Ÿª5h8ê“ µ­á„¤ÝUÞñä=¢;X†*9Ò1ÁÊfÔêž:"&1KK¸8³²á¬AbØ`ã}cFì°>Äv*Ù"îI8ÛR@“ M‹voP(#½³²%›ÂtБ˨±BÄO!”†KZ¢èx”ºåƒS¶<·=¦¿¸PzäGqO.ÍHýþüXréYq¡bð¨Ÿx"«‘ÆUà nì˜g~NÛø#þ¿H|ÑIUV)ðLj­<ü÷"ná”’;À£âO›]i÷ÎÝHH[‚Þk .Q¨{³þbMÙšJ@<ˆ.ªŒïó;†'tlä PÒP 8ÙÑ¡p–Zñ† ÿMäwÔ¯høJRçOn)Ø}ž<òÊ‚Ï9WºŒ'* ú¥­â£ÖEKDöW?˜ëQÅ¡ÜN.9…ñõi_pÙÅ]—HûY\F:ê»Âž ”îØ“ ¹Lµ]Ý@궽™,—ä—HÙŸ‹9š:¸‚ðì§–ç¸ ÂËßVhªNe»N l1mz~óÚ¡oçnÊ®0{[DÍ\)Á¥Ê‡Ž—[8ÎÂòÝ`ÐØt Ý8˜Æ\œ¡ßÆ$£˜ËóQpÔ§Úvɹù{9i”UÐŒ£Œ \2ÂT]˜»©4õ;©Ò"Ƶ+ÖO~C_(o'cûŠzh$ÂW8[ðçýi9ÍüXÅÃOËÄÓ²1…L®G¶­äÂpµ+îM+òŠÊÑE.Ó‰~Üò¥,/ÊkÚ²ÿÅñåæÐ~ endstream endobj 1077 0 obj << /Length 3335 /Filter /FlateDecode >> stream xÚ¥Zmoä¶þ~¿Âè—È€WEŠ’ÚOM¯n.@Ð 0Mq•w¹^ÕZÉ¥s|¿¾óF½¬å^ÐÀÀ‰Î‡äÌ3/:u•ÀŸº*“«\ë¸4ÅÕþü.!jÿpŃÿöN ßw ÎoîÞ}}›eW*‰Ë¤TWwÇ¥¨»ÃÕ?£Ì^ÿëî»+§±M,H!êííÝO×;­u¤cøC–w½›ÖÉÒô7*„œ_ÖÈ&VÚðÚÕ+ç*NiäÚSºãÅÌSß=ôÕ™©/×iu#Ïœ+|{á™çë4ª)AvÇlÝ8<׊Éi¬¢C5TLòÝÙ:üéó P Ý"Ä¿´{œ8õ][®Û:I¸¥c•ñf†Sí;Õ¼ÔÀã@ŸøIêÀsÞ¼T´í÷ýõ¦3…IVÀ¤¢}a³B£#¦ ãŒ\|cÕÄ×;cTôs’%oå7Ièh1Y„0œèBaÙÔrßJnÊ$jªþÁñðŒBÆf¨w‡Žœ4q­¯»¶j˜ãPû¡¯ñ|î@Ñò@¯ú~Z^ë–Ÿßÿð¤ òáCE|]?T÷¬}_·U/?ÿ9Q†éŠv!aôò<Ö]{çØ/¾}›áÈDM}ßW=¨üò'&øÀ#B–ž“ª"¶Ú^Ycb[¤ä=qsqRÂ?*%7ZS–þ~¿[Øð©ËeHyP™õ:M³ªOÕÃkKªX©ø¶ƒî/)t!”ŠÁð.ÿõmÜ&ÖƦȽK‘šZ`×Ìncs½Ël}¶A—Ÿ›è½¸j½æ&ÅR×:à]–p ‚|wè9Æ€av~àQ}žm'ÀÒ÷]»wOò‚¶‚ϱ=¸Þä¦F—d–D÷4DýÊó4Y+-Â’]¬ö¢š!‰&•€™ðgÃþ‹-¢óÚ,ú©N°Ãæ×£‘cƒw=OÉïÁ‡S廙;0Ô¼aƪi˜DÈ4Ò¤58º—…|P.¹ öÂꇮѯÁ°ò² >ZáœÈ ŸJý0v£°í0N¤rû( ùyvçü¹ÐB1Â$|:)wØ 7MÏ{•­¥x!²”v·qôó^Áƒx¯à@÷èøÝ#jêZž“ûí˜áp(óŸ©9Å`¹o^)|<ú¦vmðJÆð^ÞzFä½ótµñ¤]—ʲvÐøTÝõ~ؽ™M§JÄ6ÏL•xG±ôeø† €"«¶¤jwñȯ‡/ø˜Jtl ’‹¼/\‘ [±a‹f‹µšŽØà21l©iÙÒ<,€`NHG¿„ûp¹p™3G™-uMP[a`K :ºE»"—)Mä~­ÎOCxÁ‘©s .3ŒŽÅ"½€i']$DìþØõñX1JÖÈšP¦0žw—¤›§¬béÀ•‚¤]^”àD¨ê¯,ÿ=:Œ–Mç”IœÛ‹s¢çÓ5é‹wFJì¯ ûsh«&Ë"W‰w¾baÊsH‚£7ð b©i¶¹m•Åen~ÃÆíl£´ñ‚oj‚RO§U€RNåàbs]âW¢KÍû°š'rFige¶Â/Þ”Î5o =§†žcQÂD@×n|8ñKj(oÔ ?F/öž§?¾ Zã±Ñ$Œ‰Å›È”¯Ä{Ñ&ý²è œËpY±Ø¼ä´ ˆ6 ý@YR4>Ê°x9,v y¶±@YÂei}82aö0d­É0Y7}u–gˆ9eˆðz¿“ÎŽÀÇP|PkÓGŠæÇòLùÖä°A…–Yj9ÒZ½í„à>qPh^øýPª}Í)ðù.ØŠ5`™v½›ë”•´¼ó‹;}`OÕ¼ f"§ûeä4@s¸9QÞ>zŒœ°DÈQûàŽ¤KîŽ< ')í6S1”}E鲑Ý>x¯¼¯Z¦iô…,8×è²uk‰„\²<ˆ‡êfE€5å'.óZJ­/ ¤›”eZ0š Oõ!Ô!ž'¼#ÁM‹4ú·`”]cT§ù„ÎÇãðüñüTlº}Õ|ôõg·8 Èx§¸ö¯2Ô­ó¼Ü@ø…ôT½_ö{dЈbCÇ xLHëFùñ3Š³Xe¬Ž›cdÀ2 R1í ŸË1ÆR~JÚÇ?i/Px?ö³e_† ‚çœs˜’|UÍgZëÔ¬E5[XlËëg£Î.2/¦‘kaºü™2gÏÔ§ZxðÆ)ß…d(T¡0ör#B©®‡ùZV©EF՛ÌÔ> Ø`yÉöõQDËV{~c«nÿ‡þ`RñÌy d.s£¤,œ.+.àñF6¢”…(¨¿–“Ø$ë|Ħëº]±QàZX|8¾Ê-P¹úƒŒ›~å(®¨Ò­ã'—íJ…Ýð)å\A<²­+°õ²´kÛ o„l4c#‡§IÐÞò¼¤¼$ú3+EF‡7+Ç?Á§Yà€ÃÙ €#Z"B¨t4 i&×ÁÎ" ™mŠþxËò'h÷™½ 3lŸ ÄwŒM'´“ ¤ÓB(Ž¹:S–Žô‰õrH£T93OØò)‹$ óÌ$­-?l˜19U,¬Û?×ÞÅ\HÜq ,YF˜4Ô£8h8C'S “täשt”àù³Ò¦¯¨ØÄuz(ðoþ€/P$Æa¹‹€Œ ? ßÂ4¢òäZ1Älæ$¯ÚJÛÎÎO-OBÒa_Å4 (áá“ÊÂ>Aq›|ÿ¸. {ÌðyÚè«–uv ðVÏxcnklÄ;)ˆ\SÍòý>þ¥;Ÿ?ú§¦¶Ðb~±:öeЦw%h0®øѺg,p¸…ÑÖûj`hÓ˜Á¹]Ý«ºL.¸Ý¡¢ˆ¾å“%léD[¸) eú bB´!•. Õª6³1~Pdl¦ )2¦¿;K@ùÔƒàD¤ÈÉù0h^:Ÿ)€»BÿáyÈ(Z–1ô/ܶXàâÈêýÀDÆR­âÒ–o÷¦8õš3HÿfÓÂ6-°x —€Å ·/rIÑmÁˆ…¬û¾>Ì¿ð"œŸî ÉprïdI8Ows˜¥`Ó“_þyž%e²@¯Æž©â ¼y¹:N~oˆ·es C\is§p£)’e1¨~Š%ÈÆ©ŠUžML¢êkÄuÚa¸Dˆ‰’ø-§¦cmx"ÔÝohaše–H]:¶›¢ˆóä"3|CþŽïÙt,G( ñ½%´MÏòir6 `óƒ{’_xÿü€V5Îéì—²¨@ëš©\ÒèUëƒÓ?°2=ïi£¹‰È›e€"nö”€4’€éÙ$#„FnPx¼pI³÷zàM"U6‰CNp4}èâ¥ë^ØÏŽoìòu‘Ì¡ó’QjJ#FÕHd¦ \lÄÒ‚â\4 µA4azâ!šÃ•ÈÌÈÙÇ_¦³ü¯"è}z—½±Ã5ì#êSsÁŠ©Yµí´EšÿnDFùÇ{Ê{ŽúpŸ]ëyⱕÈCo§å G¶s"e¾+Ÿª w’;>ÈBÙµJpo­.wsSzmQHÜfQ^Kž­—_#uèShŒÁPÉí¥‘ Éà8õ+´™ŠîÇó'ÀE88ù“¦±-ËéÛN¬ð뎎¾Áƒ~Ôa½ppøtm±;¿wB¹8@]ê4|Zño}Êc«å“ÒŸ¥P“«2îE3”Bý‹ÀWí¶ ±uÝè›é‹™Î  ZíÁ(xå 9žñ,­ò<˜ ÐB#÷ë©ýPs¥3I}’$Ô“IÊvñ‡{üèŽØðDzP¹«¢µ¤éòm}<Ÿ!iùì.PIì— îà*5¶œC4碓Ónä…YÎ@ÜTt»é»,=} ÚNe½+ZÁà®Jfc]–¿¡©uÛº©è&yvïoïx‘Eg [‹–¶LL§?Y43ð5t Wº_ ŽZÜ*)8@ø/;k^.>'ÉÍhˆPä¥p®rf²·)4]Ü%€Ð”Û3q ÿgá¿x’ endstream endobj 1082 0 obj << /Length 3607 /Filter /FlateDecode >> stream xÚZmsã¶þ~¿BÓ/¡R !$“igšg®M&×Ä~¸d\Z‚lö$Ò!©ó]~}w± ¾ ºóu¡ã±z1Á,h`8¦Íl/| ÒâÔI `(0øž:œo »œ"‚=%‚¸­mÈR X³,Õ‰5F¿B¿ÿ :h*=›ÿâô`p*ïg4ëþDÏ\Æ” ÅèX»“sªÐnûêè,NàLªûÚë ž&k·Ò…HÌ":½D”«Mt¬î<èÕiÔY{¤Rìó¸ßŒX£Þ?áBÃùaßè~Øþ~rœá Áðª ÒÅ9Dv^ás΂Q4‹§âº¢õîNgüV*ä3aλyÏÜ#ŒOª8Aæ3n¬‰1CËb=³)¬x|æW\‡ªuÀ 5ÊÃ}ÓVÎMKó -¹ ýNœ‡ÜØaNP ~‹NI!HÇc´˜W`”Ô ôÞ2%ø™€åºJò!­ö\9ðjÈp“$sYÒuÍ‘ßØVíöt„%ê­íˆˆdƒGÒ zhA܆²Ö¹Êö½¡k{¿ràånÀDˆ÷>iùÌ´=u‰\Sæ¿GJÖÀöeu ^V×-L§³˜EDÒqî °Y}röïo&bÔr™EÆ’›H„¢¼þ¡)y›wå΀ÞytMó30Q$"‹˜îårtõÔá|9é4º&1O1Mòo( B ØQÚm%Þ—[úV)¡,25WQz–s…i½så' ”qä¢èéÐWCÚIá;«NÝÀTÚzˆú Š%¤…*_Àâ®D£Qj(ŒPŽ¬£ö¥Œ9Ï…yDME>æ?ÁRXêüÕ'e·ÉÄɪ$å0î‘kÂÁfˆÐ"üõ¸&w¸†Þ¦_ƹÔÕSì‚Ì"›:YvôðrùfíëA`ÞÈ5áŽîËp9Päsœ÷ŒŠð±¬ß;\Q3Q[¿¹ò…ÞºëÃךö °ì–ªÙ¤Í.qòq Ò;`ëM¤¼P^À¾ËUã/~¡†|VÐRªÊmyiñ‰KqîÖ%µ›™™á Xj•TôÔ§˜¢LØ=Ò`½ëxÒ0ÄÐ)ûÁ*àØæEÚ !A.ÓÚ/±ƒáON0kh=ÞoR@ æ|€]â£fO³l}O)| \¹®K¸u£±Ð/Lý:þm¶·°—H!H^âRa€Íà'ê×r1q’ß×fZ KtÌñYåEÔì¯jK¾“h”õ劔4QÏOð¯\¿ñ‡>ÉÖ¿"tÜÔv®m¸ðÔ©ãóƒ;Ú·kwÙ@°¦‘ùfª8³%ì€x© Ç,„8c[ž:F!>E”_•‘9´¼9FH!$ìÙU¿ÆR[Ÿ aHGìËÎÓL°ÏäÎDFŒpF ª¶Ù˜¹æÉ,.SÏrgHt8øšœ{JUøàøåÚ|‹r¦ž!ÊBÊ´ÓèŠ÷ôÛÀòŽ9Ã•ë‘ Ü‚Õ`>x°þ$©Šê é…È:æm‹âMFÉr¥{ ®Á+¡’AßÙ C‹L ü±l € ÑB’äbØ7xi•Ó—œ~ ôpXò=¯¿”¼]¬z\À}áòÿ¡œ‰Æ<Ó“±bÀN¬ µFMNè"ª¦vÑ Ûýû\ö©¥§í %ØÑÇýHò†' B£Æué0TnD®ŠÔÚRpšéÙ­B(%¢o)(äM2­ñaýNËèK¸ 1ŒÏÉ¢i‘ΕtÈS”ƾDVøúX¡×÷Ôa@>c¤Eà vp×KÐI–"k2d$IºÐdTñ:»ô|­ƒmOõZúd™ÿ$-˜pö,IãuW¶”´J.HÚa¶¼þßÌ1Vœ({E£_pøvDÞ=žŽ†ŒBΠr;§)jû®'ç—.$O`Ф‹ËßË_¤R=S³1yó¥8#)(Ã/_Ž@ËåÀ&v ,saZ{,+º_Bš¹" £Ò—5 ÞÛ¯Ž]h``oi¤µ˜…X÷m<ž þ€XøŠ6“pŽæü²ÐAô†Îºp,£ÑRxÐÃ$|dO•s•ü™Ñ˜$ä˜$¬¹ÈvII0‘ff©%l.Aw£c!s9«Aêh°| Ýùã´‹±:ÍæÙ'ߌÐE»}WbMþŠò€îôÈYŽƒE: ¶@¼Äªç¡`vþjâ“`Í3wü…ß­yA/ hÛø¹Ìâõ ŒŽ7¡°ÔÇŸ°×Q`]¾qã/rOSrßxHAî Úx:=$ûûþŒ¼·wèÇ ó”39© Z.@_òìÿPå|£æ¬ Ÿ]AF}ˆé£"®øO$˦âü°@ùR6ið«1*¤Qº˜_Ñ ÷Z — Íz˹*ÚѶâKD àÚ"Œý*urWâ;qÁËèO8Š„ŽÈøÂ8ãÊ/vmÛª¯ø´ ¼û$¯Ðø•!€¾Ñûa1õÔ‡r\¾Té–Wôq`ŽnKwS4ävxð?£«ð endstream endobj 1087 0 obj << /Length 3492 /Filter /FlateDecode >> stream xÚÍZYsãÆ~ß_¡òKÀ*i‚90œ''ö&›rì-[I²)DB$jI€@ÉëÊO_ƒƒµ²c'.Õ.æÂLwO÷ר¯bøÓWy|•Z«r—]­¯bm·WÜø毴¬»…7“•¿¿}õÛ×Ir¥c•Ç¹¾º½Ÿnu»¹úG”d«ÞþùʦFùØÃ.4úúõíßW7ÖÚÈ*øÃ%¯¾¸ÎIŒy!A¸ò E)PdU $ E>sJ[Çg{å”YÝ$ÞD_®ntÔ+m€š4‰îŠ}Q¯«z‹³¼ëTykyŸ7›²Øï?¬2]¯nLæ£Ç]YC öù°2YÔœ¸s„ý‹Ö–ûê‡’Ç ~ômQw÷M{À®‹š•I£|·lyA×J¦ÅÛ)-ÆYå€#‘çÛ%zµV‰D~l¤2Z—]WvBqY¬ñÈ?]"ÇïšÓ~óe½áÁÓ‘®ôÆ$Nå Å*h>å±êq3§£Gܸißs¯ß½´Š÷Ä žàLTþëTìe¦:”juã´‹¾îweûXu%Òéã„Çkšû°_Éï3Í)³…×Ùñ¦@íHOQõx¹¼U-»jíÊ\’Gîâ9_sÑXXWÑ3‹º}ƒoÓöe×#@Öí.ÌoäoúÆûšß¨y¸áw ¡û¦Øpï•F”Q}‚›Úè¼Ñ›w庯šú{9Ò‚å™ ¼ÏòX·Õ‰HfPtS™ÁuÕ®O‡®‡ñjl’E§zƒjh“TdDŠ‚]1_XC„ÀHq·/y¤oø‰¬Ü°=•² ‰h¿Gj½ ¼!Ú£æùcÑöÕú´/Z߉`©,ján‚’"œðMò]5¨ÖºðnËK&6‡ÝŽl—0Ò´Ì7®mxJîF„#1ZeZ°îKZ倕G¨‹Rr *ý‘(+×AMnªw±öÀwÏ}F7ᙆ‹¶äÆqÀÙ?‹HPnà¢]Ù—m³-ë²9Éá‡âfIz,±ª.»ßáº<Ñ ÂU}_Ž{4Agá–îqÔƒÙ”<ÝÜós”.ö¦†Ò´Vò,äM¼-l¸ÌËDÍÌí­ñÍ·Äcÿ÷ÕéÀƒoþv½ÄXsGR©@(j“X$5eX†Î€mÐÞVb;ØcpIÜÓãpöд2MöÆ—°®àsÓð,@^=ÛÏ2¿tvEVÁËdr_ô=Á>´‘m: >ñ-íˆ={ bJœ6êS&ˆ» ^/ ™ Å¨-O 6}¸fǸ#×ĉ¢ý»’}Š®;Âî ôØa·²P6ó@KWÖžáÂf|±»-eòŸ4½÷ì±’¦ƒ•›kX˜&aa:²>t"MSmyß„ó๩ªMÉgo‹˜–¸!`È ‡*ˆHQ]¯/¢ÉkŒ,Á‚= Eu³gYw€ýDâI2çý§¨Šiô.Nb¹Dó¸láJ6‹3¥}ÂÀD×výR\cœ‚½ÂÂA²ÔO—&â¾4,­ãÅ0)Q~Ü®âQHt7ØuàŽ jôB$–Æ*Éí³XqØÀ¢¬ŽE)á Io,Ï.‰ÏqnŒæV (?þ·9­ÅdsŠ¾Èh² öÀQ‡Öƒ×|ÇnÌ<‡=™ h‚ྕÒî¸_ua|"®<"è]—ö©2©ù¨¼²°àÌkMb-Ž c–ãŠÍÃ* @u#t1Ñe{_¬ÉƒÅ+™t–Ž;„#dÐ=ôý–‹N•æsuìðHÇ9 z+ÄçÚ8c é?T…,Ǹž†0ìãÀsÝ7í'<¸/ëm¿[Òh“•ÚáZwÍ㡨?,IÑÅ [²•KÛHŽ©ô$H8 4cÉr2eó—œyŒŒÖC2s`£Þ§s¡ÏÛŸ6¥R!Úb…´cö€*Mƒ¤¶ã*𜧻 ”))%ŒJ©ežSg9˜€ˆ"™9Œâà£ØaÅWè|zæfä[ƒ$¾Z„°áØNM·”½Á!½blÁ;v&vÆsí,ÕÊ™ìÜÎf×·—é³pÜ¡KIèrZÒ29H¶D(çÐ W¦·svéâvæŽÒ>gIEØË(&mžíßm6*?&ü–þÛUn#´É#šf’øoª®‡<éÔ´=“üKáMÀ—zR°yBþö7Š6úËÛ7ܘ¦Ð-öÛ¦…ë ëÆ8{r§Ô^:¨š'Î_g›Ê.ŒðöB~Ǻe¹ 'Ì8IØ ×2NñôLÔJxܸ楉ÐÞî›;’Á‚‚ þwPR*’=º®îB•€GgƒÞJ|búµ¤‹&@(ŒÜÝ®jê9;m w*®Í€ó4°éLÏù¾1ø§ \ÎÁ!QÀÙì“"Š'EœêÉúº|ä‘°m>?ö)m€£øÏ’c0bXˆ¼zQgŠ&Y9r¸!Þí=zO£'±$ÌMòº˜Àû!‘hf ¸t™JG}1ªJíàæ§85ß ×áS­?¾WѶoÇV‡Ló×ÑS"´a®ŠuÛPÁ9⪠îDåé<î¹À“ƒˆêIÜIY†‘ä!µIÐ!ú”ñz ÃÜ\sèíe"ÍÀ—éˆé9ò†*Ÿ#ÉT‡ÚÏx»ØN¿—€C¥*o<ÖÆŽ$` ;àÐDˆÂ(A~ >`þýŠ` ´ºë›–nQأç#Rú駚lþ¯l:ÜZÿØ,~Љ•×O?ètrÎÜO~ºhÚûféÏ& @ÄËÀ$ÉüOQÇÿÝÐ Ôí xéÜ<¬I‹g—Á+QyžýJÀ+U¹›ù‚kŽH¦Ù»|ì²Ê$g€PùäåœYˆ±Œõ?EÍÎrìÙµZ;Ã?4õäG]9™ÖŒDù$,BŠñwUéÐ[¦_g©2Þþ÷v’*Ë^¤¶ZÙ1&ÿévòKh^~ ä•î9/ϳˆ’);ªïÿÙPH³™rx™ûðc¶r éÓÄ…¿f8ÏéóZ<¶zxÿ‚»ƒ¶KÇ?"I±ZÅc`ðvŽ »µLy5Ð"{ÜŽ?Á©P:’7šñMö„N+øwöS¢¶ÚVC5üìZ!¿1îçpYþÒ06ž¸ù;ðRÜœ/ÅÍæ9×#³—âæüWcPZ¥ÎL¹›U‘Uø‰óe7m7 endstream endobj 1091 0 obj << /Length 3164 /Filter /FlateDecode >> stream xÚµ]sÜÆíÝ¿Bo¡:ºí~rÉøɉãÖÆöÄ×éC’Q©#%qrG^Hž÷×XìòK{’êiG3"± ß-}ši&”&é·÷UWÌÒ&E†K™%Ÿñ_KøïKïIü_¸ÐM‰pš ÷þùãåŸí†ºmèV{ë $CW4=ÒÙ¤í'Ϥè:Ú(

ðùíNÝõzØÅ„¡·ëª™! \"éZ'Ü®ê{v¹±™MÞÔ]?\…MaWü„äÀ³hèëÃqÿ…xìŠýž°´7ÅñصǮ.øJŠO犗 õLå  “lߧë¾þw剧d,“VÚoˆíí©Ù¡J½¬Ç¢ïëæŽOÕüñÒ’ !¼.±iˆT_¤,·*C”°”`©J‰Hx~zÎϲ<Í.fT¿p.c ÁòL¹Ÿµ!?·¡ý?m(Ÿÿ†%BÂ9çYù»QÁ4g"—+ÁT²ÒS¥3*pUmØŒÇ^QÊ‘DÄÌIhx=½|=ÃÉ€319c†ÎX²¬UÓƒ¹õ@+ÀW2ð•í}ÝÓm §½§>TÎYBA³$Î÷μÑsfò•AFýD1®¾ÒI”U£“àJb$ƒèUï*ð•Iˆ»ÊÓµN X„¤’EãØ“NÂÃMÛl¨o”5Å¥MÊ™âr‚‚Æ0™¼ÚCiŠ¡þ„º­ âd:¹Bò)0ÉÔP`"ƒBmSyà–®îµˆ#™1›F‚Ñ5½„“?fH’ƒÎÅLç¸EÐyï„…Ôq_ïPñ÷ôæJçŒëåÑwÕpêšOE%7¤3—áèD²sJÍ©d*‰Þ-á\Š¡ð‹ˆÉ&sÛµφ.E÷~KÔ(‡É!U_ÂÎzžU¼ôÕdgT*$ò‚Öª¤«n|ÂÇ,W_¢SùÂàðfålOÃñ4ŒŽÚúâ¢nÞè7þ‚JF£ÀÅ¡¾»÷¹—;dXìa»ý·ñtÄ0øäLiÿRÇ¡+ëÛÛk``$êíðùúp¬¯gÆ¡Ê™} dN99Æ³ÔЊYïXãżôø?¼½þ¾=hµè*v «ÇÿD2ŸÛÕßæ×ýPtÃU¤n8ÃO<ÊO?< —1çÙ ÷h§)T»¤­þ õ×Õ™’€Y¸hÃcŒ58³!:GŒg†¯<Þk ÆÓZ¦²,ßù˜„þž‡Ø’'.2;Èy¤C!CÿqU7eõ˜ò5# ïa§dÎϤ²™\c’"Þ·Þ]7*ÍOÓ¥‘ ã¶Ù‚#x‚7m›è®–ezÔp$(©<)!æ cjž=¡ê†®Ñ8vÕ§º=ùÛ}Eaã*çg-£„r‚«Q€¡»4àÝ(\±€6"‰l£À4F•–aktÞŠÝùÇL²C<…ñ‡4©Àf­]jòv”ru€LLÉôßíO¯Þ}üðþ㯯ßÿcU¯ÌÊt*s0ÄÅ6L‹låÑŠRN–>å·ÄÊ#Ä#‘1§=BYŠŽá"9­|õ &¯ W)—µ€÷Å•ó¼q719K嘂›sŸMÕ/Aóä¹ÏÏY‚yˆëKº°Ÿ°©ƒžˆ3›„ÓñC_À‡ÈÐ8å¯h™ÇÅÜ2yÑêo¨/ŒÙ´VòéÀ1SÉy^PŠÅg«|ÖÈ(O°G£çfË|¹K75ñ®¥;WW¥D8eSió¢·ïâNÄ!ºB]*Z,W†ŠšX“T¿ŸÀ| tÛ{l-5ȆºzŽŠ’ªÃâel1W*Ï%Àé1½…g Ccþt€YOÍ9O®ÇF¾®Eìd¤° ‹ DLÝ‹'hW÷EÓ¸© ,\œ†kç&äØ]ï)Û®¬:ì¹åàÆKº0‘OµëÀ7ƒë2É[ï3c |«h:â÷…ž #üpÓ—T Ç?. ½é4ý s’`»“ç«Š,_òAØJ¬Ú8TSܹ>ºÇzðUD×,˜­¡ºÖUïàm¶w~ÿir„«ÂoO}qLi¸ÓR‚ú‘ö]Ï{µ—¤< ›Û…ÓTÄ8Uth’Ú?BxˆP~º#r§@sc¸³R§¡¯pâ%fÔ¥Ã&ÞKåL¤ÏÔešç“o¹ƒMBšŒ¹¬aùôé¬Ë*õŒq‰6ÙóõÍy>KTLX,W«±]TbÃìôÍËSyò±>ÔPA¹*-Ë(äÉ,”ëYæ]ÒU–Sì0 „„t=djÏÌ]óT>nÙ¸ïl& â e#öËÎìR‚ö ËV+el×7Í1>ØØç¬ÞTâ;ªSE˸¿ÃjÐųý¹.Òj½2=/ìt˜þiªh•ëg&üYÕâX»ã;S?dÏ^†eÓÇ|3ªïq õ¹îéÛ¶ þ—kS[®f«¶œoù`,A endstream endobj 1096 0 obj << /Length 3408 /Filter /FlateDecode >> stream xÚÕkoÜÆñ»…Ð~•ú6û&7’8.Øi¨èÛPé;žŽÈy%y¶•þùÎcù:Qò ²tËÙåÎÌêLŸ:KåYlŒHmr¶Ü=’­¯ÎxðËß©°n £•ß]<úê™sgJŠT¦êìb=Þêbuö*òòüÍÅg&ÖÂK»ôÙ³‹/Œ1‘ð‡KýpÑãqZŸH®ü8E>±BpŸ'&ªj@ïÒ¨©v9²ý~[,³¶¨Êæ1€¼šÃò\'Ñ&,h*£eU"ôþ«¶‡ñ í&lVÕ«<`¨ÖüËsð~uh÷ç*:´ü¸ÊÚŒWMø­ë|›¿;w.ÊU+ΑÑèyIg w ŒPŽÙi7ôžŽ£eÖäH¼±?_ ¯×U½ËÊeÞ-(qࣷ0DZìöuÕs”¯ú®yÜ„í–Åúº(¯Øn²–wc΄œ0°E;õÈàHÛªîPšlÊ–Dd±Ê_KeK|Cïá›eí¸·Û|õ'†›p¢»ó…Š2<,+“èµtŒæ÷YÝËÃ6«á|´M¡áÿ&•Ù¶º:„'¼=š®³’Žî¼éÈ üÎO¬™îhLŇm[¬Š]^6 3Ù–—,‰ê&ÿÈv2 ìÅįèÆM†s|S ƒ2Æ°¬vŒ¤ì¤˜áM›ï†ÂýMô|Íàkß?쳦A6@uüXu¦’†7rùòçç—¿~ÿË·/¿{ñÃÓËüó"¼7Q9e¤0ƒ¶Ói¤.œ àR]v…Zc•Dp·n±[Ós c’pøí¥aôŽíäŸvuòP®*º¥åΙ”'_÷xK‹Ž°ÉeM$ ¶²–õ)öÑ÷#}Ïë&ÇûÚ^Ÿ'6Be³  þšˆÅƒ•Ž·a†aPoA³(SÃb2ð4‡ß¢eÂ&ðñœèÑ]’iEì3—jµpÖ-ñÑ¥>ÿiöN5Y™tïñŠp¬Þ .¤"åyT=ãt[ã@¢Ç «Êí5@hÙ¶TMÑæ kŠßóYiT‰ì}  ”³d:¡SÕ­a‹›¾e Æ`²í6ߢմÂ*zlLo²á²RV-Hkòð†s›RŽ"uÂ(3½“Œ/¤Ã‹O6ʶWU]€Ä·›CÖä•`!ª«<¼Sä¡`šF0Ó10m…±^=Õ©pìs…‡ááðQVÑê,Øì²ÎòÄ.\ ž>»hxD2¿¿äÝ¢§(3ç,鲕N1ú‚6‰»}ã õ-hDF.@Ù¶©xÔm@Ñ>˜Õº ãÌ@xOu=…‰gørXÎ4ÇQhŽIƒPÔ’ÎlWÀi2¼ ¨ƒb™·&¯ ÜÁÊ4ªõ’Y¹âÁR×<˜0×±ð±"Óó3’Ýß0.á"²ÀFç nÐK!Ñ Öx|ƒo‚×9Höˆ‚ÄŒÔgFó›Ãp”,ØäÂ;;¯]ðBMËO‡&8匟Q·ù~'íøpÐäáíàõ=¨AvU”Y}Ípôµ éf†¥h³Á¤‡I¸Œ~ÏëJôöZŽ-Ê{è$ú”ðšÇáVâEkò…`%›–ˆÇgR0Ðýâ Ÿ¢ûÅI/z÷þ¦xÇxDY¯¦hÚŽ²ðnÒå ‡AÙLˆAMç%ÐI ŽÂ¯ˆ&ZÑ“‚Gó˜W_³fz;ÖÌba7>80Õñ¼EÞõŽàG– ïB ë‚9òxç(¸d«ˆ<ïG!5ñ˜^ß?#¦ÔŒ Ð^ ÷ƺä5öÌ‹46 ®KqýB o’ÀcZ¬OÇGðÍ_‚Üž­pBáÅ–ƒ„¤w£:å43»>ÙÝfhiqÔ7üHÉfêûød¡’D$2ž„)ƒ—–q—+HÏ)@–”3É]ÃÄ&Û®‚QWèW<µ¢´ÿ}Ù9¡ñö6ªÊ ã¢o„ýŒƒÃ‚ z‡%ã#[GY]”Bá='9Ä&Ú¤!oA0¹U*&ø÷\6¦QwûmV^®Öí%zö™«ó‰ˆø‚°ó¿ †³÷êtr'#¨}*v$vR¸Ô=ÜFY‘Ø“l²¤n¢>#>}:>H„ïð~<9ïü¤H†»úÃMÏTp ¶w"5 °)á:-ï-C+®bÅíÀm€˜;µEðàP„Óî>Áà·ÛvS®6 ½^EôR £¢ËÕ“évàœ\ï~—çNE¡Ÿ|ßpGe”9Ç ç`üiRß¡ˆ[v¤ñÝRŠ–DÝ Ÿz0¾{¡Ó§£[$ò¥àÉy¥ÐÂèS#Âø3G„±0Fyú-V°\²8/wÚ Èð{£½¹n º#›b³)q^ꊌ(ÜY¢P0ï.MŽª:ålTfElì'‰;]|b(cóIâÎφPŸŽÂ)3ÚdÙº»âNÛËÕ+}[à ¦.ÖÉÿyäé9òty¾ÁÐSB¼QgÜ·I U§¥ŽÞorŠB0® åu: B1©oùi…•Á¥ùX³¢©–Zˆ ^:9y9«yF‰$é-þž´nµÊñw5ëøLœ¯Ž”ŠjtpÛ»ì·óP±…§¢åßÐDK¢&Û…)Œ'yÄu‹¤[Beß5k°“fãè‚›žû?„JcKĹaȶø­+1åbÁe‡#à›Õľb=.„"…“Ž9Ï»ŸK7 YF!•¥¨p0yüSšZ¡Sȸ'Ú½¿m0Ó=FCdÿš/[ºx¤Òr¼zIEä§ãZ,ÂC¹Òqó¹p ÂJ…ã7ØáI,ÄöWíøF«í¥ž2{G»7¬þkÇ›vy¨âŽ „\pQ:žJ^þa ¡ ö|\hnÚ˜{0øK¥=žÊ²é¦n(-kx¶Îÿs(È} 4G2¶ ü  Š‹j݉¾Æ1Ü"ùºèöÏ?@J5ßá ˆ–IÖÛW¡ÿÔð %8(®Ê®é*G}]Ùuáµ$il7]µ§g/sŠn1|*+U;O¦"q½‘é`Ã"ïbÂE¨Ëi½DÙKá“iLëŸS¯$ôMÐà5?é`ס×R‡°9´ž`th¸¿jû W9hè®(ó¡}Fó!ɦ¾ÌùPRoOºý}‚î >„¡’Š¦­‹·ÖIœã¦¹ MÍR5€ÃÉ9ÛßÒ[ ÀáŠg¨%ñ²XÚŽŠa®Pü·ãÁ˜JþÁ–²fÆ5 ð›î¨AÖl²}ÿ™€’œî¨QÞʃ¤Gî< #Ì …·Y½éF·fà©ðƒ®þ„¥–ycç3µ5–míl žÌ›èÖyƒY æ¯Wœ¯³šÔ+T·šþŒ‡6³‘}Àq˜ˆÌnzÓìR¤£¢_ÃW0¬òi,ãfcêÞN ŠP>ÇqÖ¦b馷æÇ÷NQ»÷£ fò¸c4ÓIóÂ0Á¯ôŠ!à °ðæq÷ñÕUàÍÙcv ”ÞÇvnOB»Õ!þxÓ®Ž‰Äô†Æ»âjCå…Ták¯j—·6 ˜z„ êë¹€s¡!ÈMˆ$0~2X’?år{Xåøñ°‹þº^·ïÍb·/Äæo'Ár©ì¿Ymyå.+°%({dXËÇ<\nðc*}ù%€ßM>ÕY±ÿPkúdpÚ¾,›°í¾­WÅz}_ðÏ7ü#„(_ÞÿéþdŽäú›w'Œžt×ÿ?_¬l endstream endobj 1101 0 obj << /Length 2215 /Filter /FlateDecode >> stream xÚµY[oÛ8~ϯðS!'Ë›(if÷¡;mYÔi·õ`ÚÂP,9QbK^Ynšö¿ï9$u5Ý$,ŠFuxxîß!Í&þ±IL'¡$–Ñd¹9¡z¶º™˜Á§ž0Kç¡ß£üÇüäõEL%1Ùd¾ê³š§“/Þo·É¶Îª©/„ðÔ/S_ÊÀ{›ïê*¿Þ×Yêo²MY=šïó?Ìè!¯oq$½ÙÇ)ó.§>S*¤žbÓoó¼›·2œ?Sx¤<”^õ¥g’‘FI„4*¤åþzMa}àVyñ+Š,}& ÍjU?,–åf»Î~XÊe¹¯‡¤ÌÐnë*ÍW«Em“õº\.à²>73z¼(èà•.vuRÕv.·Ï;û¼owâŠjwš}¼\\yý•ôURÝ,-5Ž¿Ã$sʧuÙlóEn—(ûü_Ÿâì8XHXÄVÓÜdµq麜ú<ò@óž&ub\¼ËÿÌÌ\R¤v°nÉëÌn2ôT U¬Ùåôµ–Ïg!%\Š¡B= ýÿn­¢ÆÂ(ÆB¤¨ñ{k¨™}^½æg¬™£þöa6[üñáÓû·çz_ÎH·}5òã«#;£úÃÕf1ÄÙ¡¬F‘*KÖ($·vFG=‹!yŒ« ^d|ŒÕK]¿IkëÙí:)Œ·W¥­ Œ_®|ø²´D_š·s§Ó#êtz:Ýlèò6~Y¤«zÛó·}‚ìM÷“`‘ kD(Ís0P1„’Y<¦À¶]8T(_†ܶáŒD±K@ãjru³¹­5ŠôCs€;ôʇìÊG$<‚§‹Ð›ßº;FŠÑ? Q/IU%½êdä Ðbv±÷K"âÖýÐlrêu¡EÂ#‘)]$ ¨.T*eèòDrØ££¢vcÙ§ ¡´·1ú•Bk臬¥š¹ø0I8•‡ŒFU~Àˆc¹síÕùû Œ¡Ì\ì$ z¹ E1•Š¼ÜvU©cÁÇPØÀù!ä3 ¥ôÊ*ͪsã¢]i‹/” ÁG½K^ïÜg ÑÞiP~~wÞ¢þAÆ …-Øeë¬+eÂP <ý®lšá‡"mSxT« ˜€½V»3£¢í%Úú'Ž÷H~fhîÝfúò#‡]–×R–JÙ =6Ķ ¹hÄ!•ãú¬=œéº›õ~î1†7ÞF‹të´ºnõ¤S~¥L ø¸ªüäNx¬mƒïËl¿®sߎˆ0"…æ¤ØAWÝaß ËÃÅ4 ÜhÃû·s{öIn2g¶A½ˆd[‘ŽªË_¢.èÛw¥  ±„‡—Á·% µß#wÜ÷Œ™6±ÄHs¶µlúmÍL HjsT BªÐk¯PôlÛ2¢!CȜͶšNy‚•Ê-Ô‹à ‰“B'b\aÍÌKy]'yaV†ðÉú¾ì¶†n™¯±²¢;gi#²`xŒ™¶ûÚ¼Œ`ð}jÁŠE ór§ç(bª¥³yÑò0zßó¥å`´­v‰y=’¬ /ð„¶>P/NÖf½ßcàòùhmÏ™9pjù‘ÔæíËѹ±AhlÀ¼Ô(™êË@›u¿{»øðûܹ/Þƒ…½ëŽæè“5=¢$ŠŽîy ÉN:åm2²¦G3e¸^˜j3L™CT¹_§æõÚÔó’˜Ç6IӶΊñ±3N—‚þêYWOvdìYÿ?vdrÜ‘…ÜÓg>m×A—¬lBëÛ}›Ô1`ÃøþH÷q¶¬u àˆ%ÉõåÓÐØ£HØd‹Ò4nóu6€ô|mBÆýP°XaBAÁÕéµXÔÅO¹MvÈ~ÔÿŒ°°qCFÙ³ñ¿àY©ï›8´§àÕœ7à¾cáÛ4G(–ƒ2ÝúÜQغ74î…™#î¡sìÞ8î…ꫨò4¶”f ‹ßº è6…>?tñ¦1¾ê%oäkº ójêóØ%»Pÿv¡-± ¸†v›[¶|4QfÃAõÞ~×Ü‘9¹ :õøX5½tqŠ·¼½Ã–IÒæÿƒË endstream endobj 1110 0 obj << /Length 2236 /Filter /FlateDecode >> stream xÚ•]sܶñ]¿âžž-ÂÀϺ}hb«uǪRù:yp=WŠÄI´yä•äÅR}w±IPLãŒFG`±Xì÷bÁ7üñMl)Y¦›âxhhw¿¡Áí_.¸ÁóÑŸaþ°»xuE°,Èøfw˜“Ú•›^,¶ŸvÛÈD°8ˆŠ†^]í~ÞúRJO2øC”‹·»ñœHˆod1Ÿq”G!“a, Gq2.CË·~ ïfësoxPp’„Þq+¹w®‡Ê/«£júªmò—"ïVåµÿ&rš^ÿôŽ»mz]Þô‡¶;ö(ÅR /ž1.ÒIîï{ÜœX*±W5[‘xƒêy¡”×}KXýùtÚú"õÚn0[˜®1šxÿ6<ÄsD²4‘V÷èÖ8å ã©°Hß½aNóÕž×æÀ/USö#§¾¶µ/2Á îsÉxDG!T¹õC™xw(ÅŽS4{ÕU( Q-é¤\þWŠÝ³­ŸD‰WV}Ñ©Á¬m_5Š¶æ9dM’\ –ÎîÔûkÞ õ–{ }âI‹þ9 àˆ€35 âÀMÆ¥wÓÔ(Mš¡}’ûàÒº}xÌÿ^û ½S™×´­´z~¼F·F‚ ¬Ú¹Qä•À?¥˜T=};¥C/Â4`R @‡ÄmhbxÄ#ã€\ñÛÆÐm”Ò!°¡%Ð)ï ÄA§ kq‚cüÑÆÆšw…$Ø8´Žs8 _÷ävÅÚk¢+.F,1”Ð÷áè0@¯¡Ê­5q6Z‹¦yOßÊÌI±0°AcS`„òôJ-KƒŒ3%ñ&J2–@‚ÇòÀ"XcA?\è:áBæÃî÷gVÊØò-ñU £8‚E4øxÝvFò@éD _¬&Þsð,)bIùtiCêÞÄ–Bóö†ÀM Å‹‹ß¨‰ÛŸ¡¯ˆ¸$:•Ú ÖÆ;®¶©ô´µyÑ“£g#ÿ‚æ©w¬ît² )%~Èâä¶Ìé3 ÞWüid¥}/–N’ còÄ:ß{ƒ:w0{2¹g ÖÁ0ŠG'¾^Mµ)K‚È¢èÄ ÙÕA‚ë „ùí[08XKOÐ_ßü¸óßiyßéè çãY↙°:Äðåa×Gfbãz@È·¨²{8,èÃ?¸‡ë…Êì0çH¶¢µäœcBïk5<ÐBÑj:¥ÙcÓòÖò øK¤Nö‡P’H@&º²:öfúž>¢cì’F×KøëµT¬sשF-hÚ0ZÅ+Ûó]­éE 1æb™ ‰Ë”³¶ØÃ^žôxßÎ4Ø÷\ ¬2ßÏã"f20±ôêÅo×ñ{eL^“Ê Ñ6TjÀ,=Ù¥Ô#ú ÖŒÂ#É‚˜ÛS^¼"WäIÀD(=ÌwŒ •}–adm&ÌDΚ…»Yàfˆš¥Ä\Û°£:§£Êha¦486‹Ðµ+ŸtÚ!kBµIw²Éá—–Á•=ìu» Âž׬ÅãnÂñ ¬Ä{#šÙJ ›g ë·þðÏÛ·£Æ±týgjæB•Ë \W€xl‹WȲ:O†éÔæpñS¤\G)½H &ßnÀª©¸TŒ¡3…“¾ca"lfípn¨®¯R¨«¶!{|Ú[lÔøãå¥ZiÆ –£ÔQ¦öDÂ]•cÍàµÉ%ôù£›|ÌâË—ÕtCìó*±Ï±ë‘Êç±,u=‹°Õ “Œ_RŸœp"ØÙ7æÆY\ÛCo–Ä+ôÙÆ£Óó Em4Ú¾¥§ï1×}µéX¨uý!„òÖW*×[f¼8>ËŸ"M€q7êhTªQ#”Ïs˜!ùC×>í ¦»c©¥¾þ½*J>±½@¯ûÕ´N8>÷ã²èÎ+?šÙjÊ!Páß =åý5¾¢„‰Äy.²Ö3zâ‚;œ•Uip)™&S‡P“¨Þ6Ôm&ÐÔÒ=‘|(EðE×2ž äônpêtq¦“Ä5”¦'˜¾¤/í2M@ƦtèÄ1L~ÙFx{8+íŽx“hà¦[Òž¢…Ö¹7oMY5÷Z&؆Z±ÁÚg*qk†€ÛeŒ÷YF(î ˜op€ Jû0Çz3-×â]Kêw­Þm½¶Wÿ÷qkjúQÐd|屦©ÛšÞsW¡>0[Ây§P78 )} ÁÖl9b§EŒ¤n¸Æp¯{Üûº½ÛFàîhEíx§g(N„ÒT³lï>^B²ãÅ¿Ô‰Î=XÀªŽüPBŠHâÌÉàôVÑTp3Ã@`[?ärꛄÌf}“ˆ2óÊ!³_{噫)h*©cê¾ÐêÉr&xAòº…áÊ%™êŸHÒ© @™îÌZŒÀ ­Î :7’N2]I阯6‰yËÃå±£qΠéü¶‹s}³œÐôòáòWfxð> stream xÚÍZÝsÛ6÷_ÇöÁ€Åâã&Ó™´9÷2Ót2‰;í].ŒL'êÉ¢‡’ç¿¿ß‚RbYT,“š¶/H.±»Øß~KëIå=+ŠÊkTLòŸ”à ¼¢à”õ^± ÌŠ»Qq28R‘³Üa•R<Á àŠœ6Š¨ÈF$Ì…#Ù¨ú1ˆ¢ 4£h Ø)JVØ3–8—5c‹ S«e˜•·VÞe,¿‹ò4ˆÅ ˜zQB{Üöœ,ÂÎ&a #²*Ñ—(R±lʦh‚bËbâ¤Øe°Ã)(B ŠÀ˜‚&eŽ¤²è{^e'ø€w¨L\xeŒryŠ›xòäÉÉä™zŸ³pãWjòÛ¿ÿ£DgÔâf>{òÝw{‰œuÚ@”-³f±ROž¨ÉtOý–-®‡@aÄ;Ý`ÖsžIè€MÖ ‚1º Slf‚«ä¼¹0©õ¼0qî˜@’É˶™¾®Wêš¼|v¦&çõíJ}òüÓuÕûúdò®«¥D Âþdòª^67í´^vqªÜ{Q_̪ï›[õFÄ °lúŒªoƒÐøŽðébÑ`¶7]ðyî±-'“×7ïVåú§Ùâ'“ï›ö¢n»ÙÍ[‘a é±Ê6R ºøYŒš3ƒäiYç×jòcsÞ(˜è›Wu5?½¨VÕ²žËßÓ¶­>Éà¬i¯ªÕ·²<[FM‡Xþ¢‡¬~×Ð[æ¼kô»æÜ@q·¹*ßöááp~È;Ã1ÀEž’ÕF‚Á1i¤5Ö†»ICÝ€€M/ Î?ÔbúUóËisu=¯oO/ Ö¹‹†Ç „•Ñˆ‰X ÖAÒPÈZ‚~pA#´M¢g³å´­WkéÚÕ¼.=o«ÅR$Ü©£¼l†"õ"<[ŠÄ¯+rOȇģ¼Ò¡x¥°ƒW2k¼Ò¸dÇe‡&H(œ42ß(;sî…ÈóÅrUÍçÕjÖ,Ýï/‹Ùí8¸bµ)Àe¤"3\‚)Ŭ‘X%F?_b«¥Ø0§ì»åøÌ|K|õE–_^=ï~¾ù°Z]ÿc2ùøñ£n®ëÅÕµnÚ÷;NBt@,B„òX;&DͤEÌðšàAÜÛ§È#œ"®}ÁÓÆ)ÒæÎèOÙhÄpìHË2ˆ{¸äuðûbz[W‚ÅeuY¯>v €O) á¦)‹ F§”þæY£ZFõª‘3Šæˆ™š©_ó_1ɲ¸â¥üþ:[^4W§¯«?f‹÷rãåÙcôrÄ ½ìÍr,É,!v¥#ÉS<Æ›q™ìH.ÒSëú8ÜEü&]ð&øÑ6¾Ú±xCÒ)v‰Ãå¬ãpþ\<­6åî?oëéMW`œÝ,¦ÅÇcà `ò2rþQœÓ_" 6(ºà0INiqâ 㘌Öô]Á"vˆv¨ÚG$lÜZò3íºÛǹÂ]ÙµÂãOåžÊŸ=5¬Á|ÅS÷•þ´®a®‰[NÚH…GHÿö±%ÍÕt©«Å\¿oþ˜´õ²®Úé‡ÉuÛü^ËÉÕõl22Ï0b:Éi9M+$  ‹éÏ®@ $’sì‘;^‡âйÑþʇ¸â"BQë®_œò8{°À».h¸ ¬Ï"Ê1æz0:Y•x ƒ2ªüb؈k£ÊwÑ÷øe[OgKXw¬|@åï¬uäTàÍD'ö²ý¡¹z7[¬‹”/ŸË_µ(U]WßO”lrÑ^eÌ\Â>Ûpq Hƒ?É÷ˆ-•üƒ„„ú×¥üت¬ßc>öìa¡{˜¸{6{˜Hãë2l¤ÃA^sŠ]8õeÜkààOßÍÊ‘Ó7íü?‡šíe5­ÇÇSY$´#®FéÉÓÆ»1rLÄ¿E•ã."Azt¦w6ii8ou†®.-'Ôž¨;ýcSý‡‹Ë÷msSÎ/&»«ðë{DA¡ýƒ„6£fqcrÙVÃáy-¹]óv§ÞÃÌ7§Éoœ=wvSN­˜±ÏèJ'û}ýUóñôªú½iû[Ã’‰_û8"¸¤Væý>~tþd’f/­k¯ƒ´7³/IÖ3#ÃõçõŸšîŒä]5¯Sd¶ 'ûw/©'¼¤G†—-­qß!DÛ ËyxÓi_IºÝŽÚê@íéG^ìÜ“ÐÓˆ„žÒ膒µrʨJ©†rÔl¤ËLÒríl°ï=¢rêëûºŠ‡˜ñ$>Ãf¸yîy³nÞ¼Ùšdo笳|:⤰•Æ¡/};"9øí·ó‹›ùjvúlvU/dƒRÍ×Ã;çÒm.w«U5ò,KzÚræ½"Ô—Ø»( {“È]ÜÙ-Ô·•47w¡È‡@q›ˆ9`±xDõ~5¾÷Ár(­é 5ùÀP#ß>í`q½w/MÖ L›;£ûžÄ÷Ë·Vš×»WçXû=»×™Ûȼ÷•ƒŒÆo-ŒÕ{XÂ^–àÊò½ Ùð %D¶,»içºbÌ2Rãž#èÒN¿n–uq–‹ÙrÕÎÞÝç(ܺ¬¥à´Ø-{ß%™4JqGùÀéx•ÑW>1ú⥃ÝÌš7CZìfHk§²wZîÿƒâ+  endstream endobj 1115 0 obj << /Length 2704 /Filter /FlateDecode >> stream xÚ­Y[sã¶~ß_¡·R‹@€ ÓéÃ^;Û&Ù$«L’Ž‡–(›Y‰THj½î¯ï¹€WC–ÓÉxÆÁÃœûw¹ð'©XØ( S,6‡‚fëÛ~úç éèV@¸Q¾Z¿øú1 )ÂT¤r±ÞY­·‹_ƒ×wÙ±Íëå*Š¢ þuð¦hÚº¸YÊàÔæÛÕa¹’A~¨ê&{÷ný î‹öŽGßýð¨b›È Ž–ÿYÿëÅÛu¿/£Ô3@ÊGX ˆà$ˆÊH³qhC¹\™X¯²¦ØÀ~¬ ¶,í¿­³²9¼ ª&ç÷E¹„G|—mrÜï\Q IiÃ8Šx•÷år¥l³º-6§}V_ÁD*‚ætÖ*aÖHÒÞe-î—0³‰î2|øÜÍÀÛ¬tëÆãu%È&t ëÓº¥ðmNêÐt7ÈòÁË ¨bÙs’—8eu13Ú_QòÎkðƒj©lp¿:d hü^ÕLRÕÛ¼SÄ;"=lŧÕȸîb¥¨SXPkJ#Ý’›ºj0‰B½å>) GRøõa@gɶ8äeSTe¸\iëe°Ft®ÐÙ‹×-Ü@b¤q´gìcSpÅä‚VÇûyÂ@&”B=C4é1nø ³¬^Ö*B•¨©þÏè\&¡Mž!c*)ýÊeçè˜Nh\ö[~µ©ó¬u¯2ž:îAÙ4ºé¤ƒ—›l¿/Ê[~;¥Á®ÚïÙÁ\2¸eÚÝ©Ü´°ü7>qV°U›‚ ¢0’Š7¼Ûµ÷×¼®Q†Ç‚¦®ÙMÀC~FÛz[ìv×-S–âŠóyyEŠ–©­2S=o«ÓÍ>g¯ŠÒq˜ÌV§–ˆé§Y¯_W‡“m`ä¾>¾oK45 °ÏnØ®ü»ÏX}QhpWc¬Q§:BÇ?žZfå–°Ÿa’ÜÎ ?£Zü-ÃÈôNQ”^ljC5$8\ÏÃI©0ŠmG{ñr‚ÌŸôñƒÂóö6hUÜPt8!Ƀ´†”rÈ *F¯yb ­8i³M{W|pq$`UØÿÄHù—|ãÂŽØ®XvNì‘%Ñi·}ò@ÇÇ'¬EôÂVå$,4…‹½¬áßSÛy×PšC%¥³4Øߧû(ŒUo r,߆ Jk«‡Ì”sºujÖ¼<Ï1bÀIÜf‰x§Òsض¶2¸!oƒé¡¤Ö-ÕbVà4Â$Ùv[`¸óºãäþHƒÃÅýFq/׬-@ð†M9„Ûú§—ßüáÃÇ·o®?ü¼özžH€ç؉¿†šè[܆i4]{¶Âûï½ (¦iïà®Iœ¿´­£–ÉMÊ8Ùy¾(·ÅÒ+æ) YÖÎkÁ}$ð›fÿy5„/(%Ê Ȧþ$5žb`Ã#¤C’ÀI§SBqŸ×¹[#‡ Íb j²{žÃH&šƒc’^ó„BÿÐèð³8㕱ìK×TÂ-"‚(¿ÏZVK¤ÓPÍAIÅ™‰ƒ<Û`DßñÓ±vE7'äbl—>q¸ÍÚŒÉ ÷²¬êezjÚª¦4™£È¼A €Kjɵ"€ V¤Ï‚ZØ‹U=ÒÖ7HNþ`<Žá Ó¦ El&ñÔCDq„+“’[ÂàÆJD)ùÿM&ƒñ¹ºã2䯸°³α¥h²á™Î6´±Æ¨“PGæ"¤VV? žGCjxÊÀ«(‰©ÒMÝ”J@2µFÌàçIóèñ ÎÇâP@Ï‚q§ƒ+ÿn ¬A·ó„ÒÏåÂüJö_r9ÞKQÒ•©dH0.ü:†ÒoGû*–*>¯ã•V"ÔÉ 9]Ë3Á¤LÓæ™+ëÕîL"†þäù¬Ÿn Ì¥ŽG/3”.òR¦ƒ¼®ý2¹ÍxKj騫/`P¢i´q6‚A×T`žÅç¦ø¯{ >=é8Aàû’)a&”0ቦ…„‰KRÒ›A*èx!ò¤ÁŽO$Æ M9ægÝé0së±Gd %YüeX"ͪS‰é¼åg” 0ú³Û¼£ÎÜë’ž÷4Jd½ï5ÐQŒ†N¹to›;nyð3ç ‰HÕbÂ^Ô%U¦™Èˆ&¹VÛ¡QÙúQ ›tÜ:ü„¡æ«é:*wþ“ñ¼rþþæÝz (Á}À[؆<B›2äÕ.Ú´c.ª“û²ÉŸèÎ,$ÌYw6ë­fŠñK¿HeG³Pÿ“  |ò7Ð×Í ÂŒÅW]÷¯à^‹ë¦ÍêY·÷$?ù$?ÉüÎ7RAù€Þaþ_Þfw qg5X­=Õ%?/£ìO9Óž‰Ì@;IðÜQUü•‹8PŠs÷wEŸqÙ1î–ØdcÃ3£ Ò Íé€ÿ],Öü¢OÎÉüod®ácî0¤GS =hÔOz›\ðVݹÞD¡¼O˜I P«!µºè«3®î¢‡ÏA4zÝqŸáiÐÍ42/„_þ%CvMȨ®iªZ@O"ª*ÄQðrûy ?Y¹¡`ÿ+WõøBfñ*wTj7_l(JÔQq«™pâ¢þyÔÄÃ¥w¬üxÈ`ç_¦¯üêE tŽŠzЇ#g=o2G=}æŸñ ê€c(3~ª¡ Ÿee#“B£‘ï²äã}ËÏÓ£@žÃ<ÙÔÁw\g!–n)…—].G,@ý‰ëÈà;ªa8Ø°;ÚÀ‰]ƒÆ<3Õµ S‚°¡ö„®èUŒ “áØ‘À¬ä̳“¦Ó£·ÚÕüZâÔäõªo? h•uW±nºÄ„~ö‰ÉIa“¢¨‡êº< »²ƒ„g½e {‘iÆ}òÈñ•CAûKëÙlþ®ºÇÅ×4o 5m>YÞIï©ås<ÏÜ‹GœxÁôÇpÅÅà;`€Ø„‰6Ý5øJ"Á*‹:_ì^üxæFLj ëõ7=ZtåH‹˜ÊF¨Òp)lpš_Ï )WãÛs·€ÛS}±¹Ilò¬>!.ŸÒxõÙO˜3Ý ö(IŸ›œ«x›1>þ_¹ &Îè£yxúøÇ:ÔjΞBé|ú``N”ýǵÚÏ’9ú øBFw¾z8òÖîœ×¼ß à\¶¯È¢.øâïU†t~ƤÀ<—%̘ӕîsÒÄqÔ}ù`s$bê''qÇáz¹ê ?ÇÖöCÛrþòMÛKWˆx¼úŒ^è"¥.êiÄ«¿‚ЂÜqáDg!-ÎÞfXùâÁ×då±…` g!-[Þ>i!5¿*à¤yÙBè°Lïwg®GŒMÖ¼yûîåÏß®¯_}ûáõ¿ýg5"4“+!ܽ‰r:†ßcÖ4„ ¶üÌ‘†/ÿ> stream xÚ•koÜÆñ»Å¡Z*Ð1\.Ÿ-òAqäVm©– £ˆa}·'±á‘2É“¢ ?¾óZÞòŽv]ð-g‡³³óž¡Ô"‚jQF‹\ë°LŠÅjû""hw·àÅ»¿¿P‚·Ä¥‡ùãÍ‹ï_¥éBEa•jq³ñIݬ¿YròëÍÏ Çae@… ¯^Ý|8Yj­Â?Dyq~3ž“Æñ72„˜GåÀ‘#`I8ÊŠ$T:ŽÂN\¦Yœ,UÐ#yTÛ‡®=Ñ*xÄÿì¡iÐÙ‡Ú¬ìÖ6¸i;^¼¹ºp¬yÒi摦C“8E®~ù5Z¬aóçE¦E¼x"ÔíT«zqýâŸs¼ë´UO/pV×Ckê>ÿ2Îø¥ÊÃLk~ãÃI¡ <ÇI°ªÛž–i0ÜW=¯z»ª¶aŒO'q<óºi‡ª¹sèf`(ªï/ò*Ê€€CgšþádÁxB×îà}ÛËѦað'††ûvww€wÛ ÿšžLL$¢RÅw1(ú2¸³íL]¨àÔ¢þó©èÝ[‘QæËH©,ŒAÎb‹pÛ¯‰,·HGËÝìšÕþ¨Q™Z®S ÌZÛÔÏœ|ŒTÚåR$T5îÆðÆ€ëgÞ±=U§°QFiÖ$ƒ¥*J05EÕ°õíe©AÂU·ÚmûÁ4+:l²Ðao³YÌÀ†2ì ûGd‡< ž ÃáœD#2·4J¿Y¢:ÓN¢•°±¶ ¥¤ásÉ ²BØBEÒ‚ vñŠkÓ ¦ÿë[ËD‡)4a®j@è)îÈ%êñ±­Ö¼ó]o›õ§Ýæ”Ç7¼jwÍpº§ð“Ìðü`÷øtJZ+!ÐdñTgÞ)]=Îœ‚àÕWÎÀýýÞè/Ûí–QW°BkúÛl¨Pqª,žŠçBŒáÁtCµÚÕ¦;eÚòŒŒ æqît¹nwŸjûÝÜqÉÑt!÷sD5¸\RŽÒÌÑS)DÉ‘ûÉ!X…± bÑ,¡®Ì“ŽB½÷pLrH2do4žªáždø°Oµ`a”ú·Ð¤y…% b)yÝ áÿÞõƒ#Cy;Æ›g䮺^íS¥óÈ-ž¹kª5áhž›ý£Eœ§.£”¤lŽÕY°ë¹hÓ© ]ÙXrEƒ«¦¬Yó\k6ŠBí‘dÿG½¤‹Q![3æñ,¸§‡Gá!+Û †ë(øÖzR¬G°Ò0w¶EŠ%ØÖÓñªêPØPIpŘ‘,Qy´¯Pº¿oæÊ!|GÊ­Î ²U5K*uø ÝlVIæiá%'Rã?°ú¾d,’½›¤ŠOyº·5³I0)}‡ø©gŨ#¼%I.4™él*9)Põ‘²–í†1ky×rãÅœå}ÄaI¤‘”»™šá- X=½Txº‰|Ý(ÏVà‹ê<+íÜ¡ìªazµa‚ÏÈP»ã‡ÆÚõ\¶È·³ÀÝß2„Ü%.ò‰[âÆ‘[2xK½G#|…j‹¥7.±*!bdtW·ÝZP[ÞÞõ³}ŠN#h µß3ß¾9?»~ÿî|Öô)c­ÓÍÄq ­ž„#¢{uvsqþöfŽ.4”^áBxK’à‚êôмÕ=âÑ—UÀæ{} 6­¼‚"«½øÅQ+Æ^“1ªfUïÖ” a1‰V@­ Î3Ø4´5.Ž—ÂûDùÑ*¡G EF¬~X¸ÔÓWPÏ2Ž©ïÚÒĶw¸˜SqÅ¡q°Å¢ûYA"—ÀYž4q ¢3Mu¬bbÐüEû$[õ³IN…y4꿇f´À1 ŠŽ¥€L æ®öñäÀøÂ4ýæùÈFrHŠ~Žâ8šN²H†Ç›ºwŒHÅ™Co%¥™š÷[ÐMÇËå0Q9Ü/GIñÄÐQa§®××oœ™éÀµÅÏxn»ë&'uÇõد‹n$‘Oß»‡ÿ„—6•C½T¨}×H$¼Ik $Ô tfÜZÇù¬ ¯ ]I:§!ÜO±>TýºÝΩC äñ¸Ø{=a•NžHåÌw ‹–bÛŸecVU] ãåÇÞÚÃÑ]œÄaRèE uY\jª“!y,"êÖà|šãM!þ@Ͻ¿ôÌŒq©Éé%ÂÁ"3ú M¿0èÒåDõü~På]x‰ËkXƒˆiDÁf¯_Où¥(Lx¥ŠÃÛ'ÐòçPèú·ÿòðÒaÿ»õKàbŸë#!çûùÞÎò Í5%€¤7ûCØà…R6ÃŒü²/S¥êA}ZÅ„$K¿Úæih(íÀŽ‚„KEì†gGìL‹.&ƒ0gLÍ€]³ç©­Ö†âÅØ÷J&ï§uNæ•—¯$BîL˜•=c Oë–‹•ž)‘áâng:©f­eÐ X:^÷™Àì2ïKCŒv¸v[qM±ûʨ”ʨlêCOPàÑ q“‚ ŒWcºK–à±o7Rfì,6 °º9) „®pg”´T§\SK-Æ뎡ËÃïç†]ÿž®B”ú[+Oý* £Tû¥ðÅ屑h&åˆeð–O;-Äã`eBü¼˜ŒU^^¾ysûáòÝ럎ég%}×lšW“ŒU¨²rRµ#·wv8&˜fPAù5{ª²àB&)òiXŸÜDÍÇäØBC±ûÇ^½»|yûöýë×3ËÔŸ·COOÍ!§L†ÇVRÈ뢠ãÄÈx©ÀnÓó©0s$ÉÔ<¯ºÿ„׳·ÿº½¾|ÿîåù ³)$Œø€YvÌg>ì ÊDFLñAÒFe,Z]b1Ñ–£y&<Ç;GdX@øm4"º–ñ.ǃ%»rûƒ_Pòr*5æ Üý€¯Ü‰j­0DY&y—ªŽÛe-áAÄ£ a˜Ô9k¿*„¶¨²µîUø6 †Xäãx¶7[{¬‡Ú¹d†‰LriÚ°å¢TOwuíâŽCÈ2{¹Ê87`¢ê@z¦™ q˜êä Kîìz·š¹C–„ÑÞ–H›jüx «ß*ʤ°ªdoÏ7<> stream xÚ­kÛÆñ»Å! °X.—OùÛuâ"nŠø€|H‚랸’æC ©“Õ¢ÿ½3;³+JâùÜCq€8ûš÷cOÜDð'nÊè&—2,“âfÕ¼ˆìl¿¹!à—^Þ·„ËÉÎ×·/þü.MoD–Q)nn×ST·ÕÍoÁ›­Úº_,¥”Aöj±L’4xk†±7÷ûQWËF7]¤õwïn%è`Æ-BIðá ¼_,E–åQ¥‹?nÿö⯷ž¦4Ž¿’xÜù4õY‘„B&DýG½êÚêåÎæiú¢ v½iWfWkq0Ü(' ªÅß‘fÆ¿E°…myðàÁ z§z5»³ª®!xÝõh:Öó%;ØÞZ˜[éa°¤: ‘Õ 1ÀRÈP¤Ä×F·ºW5IxÜjvDz 9Ïú}K@ÇßÊü ©{í¸ÄÉ­ê«2¨zâEÊÔ‘Íç,7(ú½›Z;…¿÷älzÕ„h3Yðc‡Ë3n±ãdÜË8³Š’1"P«Ñ  p„Æõí@0ÝÀŠ8®xáõJfÓêŠFDmì$…Ç)ŒFÝš¾Û®éP²ÝžÑ°$À zÞú{”FƒÖ—FŒv—$aVÄֆæè„[c>Ÿ¹tÉÉÙGŒš78£Õ`–Y˜„1¿ýÔ©ŠÀ{U+°ñvóŠ4Ï‚ÚhZJ‹Kâã¼@Ú2ð¸òæ'H¾@iÉÑ ¦FµL§a.Æ­Õ@ &ÔŒüµ+`Õ5R¶Cs2Y<ßÑž=¬×¬ÅËÀ «4°»‚˜x“óS \Þ²'~ŠóSß ) ÞnqÇ}Ö†<x¸ÕÖö`VѧÑ+ô¨­jÍÀ7Yšá;[Z³‘ 5ÿÒgn-™¼Ð;NÊ4'Êo… ªªˆ£Á½%¶£ ,É ç =C÷µnõ.tSA9›J32Oɺ«ëi=€iñšõjüé84ì[kžÃ+$ #;3Œ",˜N0©0è¥Áz=îš¹»ïÁÁÆ;b<þîM×4´uÚÓ_fbã Âú¯ÃviÊË É,õŒ†q¡Šã´´X¬A´f¥F0£öETB$ˆb‹È™¹3Žt»^ÎaÊÂ(nÇ£’›C.8{Â~0uM øÃÄ–Õ:Î;÷@xÝRK&‰8äS/¡€qäEëQ+¼ÌÆ\Ôˆ óÓ´eKo¼·çà„ƒ®‹"™E,âP^‰âLç³rˆA2=—唺†KƒFãÂJ€€LDXÈô\V>ã$sš;ör– „¦6ƒV0¸ë¼ É„‘f”uÞ¯iŽœÎï Ä3ªU¯Õh#Nî²].O;mðs´ ÓÓô½« òëBàÚ­.‹‹$‰(-"Ðv#©ÍÞ‰3x笶Dy0{Ž¾d'ß!}÷fìUoê#M¬‡-‘Û š(ëZ¦ c!~‘T²ägŠ´Ò 1*,–y¯m,MÀ„íy¨R8–"ècíhlyx¤Ää}=Ò>[ÂwØs|F­¦ÁJçY¥¥­@“³Ê±¶sªÔ÷•NðŽõ\…Å!<š©YR ¬[ QS™ˆêm—èjˆn5MOË#ë¶ ‹ )åD§”#@L€ÚK gðfRu1Uµ,ÊP&™“‹hKáU&üŽ=­ÝÑ©éÁµÏÒ[ÃÙ–ó29‘ÍrjôY’³>ôaœ&UNÏú³^íG—æ½/Içùrêù¨;'oIUÏ)¯&W©ã<×s¢üؽœ¶ŒõgÕ@ÃÓ 'Ñ4Ðöа€Q¸ÖÚVZL+e­Ùí,u–¸‘wô5ÍŽ ´çßÊL1!Et€u&5ãµ¥ûgÌH»Ð˜ìÏÓá…+ª<èöõ¯kà©Ò‚óô¿í=6™çî`Ú‘R¯ÚO¾ˆˆ³PFâ".™ÖŒ¯Ë Þéj‰;Äå‹‹Ÿ?|¸ûõç_~zû’.ú¯Î㧵»ð8n¤ÑwßÑ7¢0äKÐ(†CåªãnmjÝBÈEß4GZúæ±ééJË“B±œä¾²dBÄïF[cL•Zoƒ¯›žøVTl¿àõ³™Daž]%TØUÐ’ÈÓv¶$ð%M—MLžŠ[˜Ãæßr9)ƒ®hÌfë{‘Èõ)£âÎ})Ê<2»²4éò,Ñ$™>šÆÔªÇzG´’ˆšˆ’jíãÎàò¬û…¡óiàkä}õgòuëÏ8qjyJW³Ä‘«œñÆ.&†~Θ‡çûç“î©lø’ùþ?½ö9N *˜8íØ=é²_ò·,;Á<Oü ä#b~øÊÂKÇ8ø}èa!ãÚ"2ìÉÌÑ–¦«½•ªú Ž³.!¡‰Îó‹lBø±Ü´u!À.€¯…e*Ïâ7 W$”ÐyépïV€ˆz”M¥û9rÒ"Œ¼`\“~‰D”n—~ÀÇ­¹×‡úZìÂÁÒ’Ê×”|6§×éŸArûçËY|»·qò}°êi«¯_ç²…y)`p~­úÞÌsy„‰Ë> stream xÚ•ÛnÛÈõ=_¡§.µ°Xrxo°¶»ëb“‰E‘ê˜I„%Råö:ýùž)R¢í.øÀ¹œ9·9×ñg|þ,ófI¸Y˜ÎòÃVë팟þúƸ.Y¾ùãMÍ|Ïͼ̟-7CTËõì‹Çó¯Ë¿Í‚D¹±Z½¹Yþc¾‚À \øäÍõ²§)õ2„¯s§¡ë!Óþqo««ùB%¡Óì ¢ÔÙTû}5W©óX”[\ÊœM[æMQ•V¦uu`X9”9U½.J]?ñ2‹„ëE‰˜Sot. º–A„ à}@Y{Ü™²Gýt:€z}ûëG̺>÷…Yƒæ|Ïi*üûŽæéq¯KåµÑ CùÎzâà߈ÔÕäG¡YwC›Mó¸:‹áeø‘Z“Ð £°ÿŽ1wJûÓ…(pUì˜ßLÞ6f œº‰ê æj aì†q<·6¶©«§énD‰…±¨Ðg¸çäQ ‡'zdº\OV‰›&cUmöÕÑNáT¢:X—A Ý Œ‚,|/uÁ9*uýXøŒ]ß›/¢XuÎ’DÎûó…ïÜâ$t>Îß#«êƒ.ÑÐbYLsv”¹¾¯¼EKˆ\ayd ]Ê%Σ˜'îˆÉòDóocÒ£Q%(·¦iȉ¿ ¹C‡sàvxz„i:fQ´²]mø’'øß¼y„¨©©uiuûíÌ <ñ˜›¢¶ {ÄÍuk‰`(t¢ŽÍÈñ×¼qÑsÔÃ=ï­ °­â®íÔOü…ÎQ×µ/¾i„à3¬ÝÈÉÛº6öO ¿/{)ì‹—NÉG¬)ß±Å7×ðovsß‘µy~X“va².¦´À„;_„¡›‘÷¾Ý7E¿¡÷ y!'>ɉì+Ï9è9ÜÞODÂU{们jFù/£4‚(Ñ´aü€˜MMÜŸ>p §ßWÝþÖ”û4OôJ•E€ ¶ `wU»_ó¸X¥‰®Á4¶„&– F'ÝX9^ñ¿Ùé†AÉÝ.coŽÑSe`x½)žæÌZ¡ž¥Ž>T”I‚׎µhÛXk,^„˜ÏÆœ§¿@en ' RÉ1¥@7‚=x1åÂñÊ0)vç‰ñœ ‰ù™ãs» iè*ž}ù¥ÒkÞÁº†@šƒ§Egðbp€­È¥ ¢‚ =”è…¤.ЯðŽô`g`D‚‚‚ÁI¶ª\#»IàèEqèè4±¸*üËJ¾¬JÀJ8‹vÇ[â΂­7I<‰& B«Ä)´`.5Ï(´Võ=ÇH\(À‘™"Ʋ#EÊ Cø"‡Àx„Øxi”øÐa|ç€X/\w0ßÙ+ Î,¯qØÇ‘Õ"x´,iþu\뻽‘CÕ²Ð<Ç8Ä&_9Å>4íÄK§LŠ±0 \Oe¿Ûª±öœ®ô:€±-c*‡b M†Î—å< NÙ)*{Í@£Ø÷UŠ6lþºœ*z|/q½Ó!È‘‘Þ2=L‰»"GgÞñ’îZ{b„6ŠòÈ…IF9¥+¨c)Á£laÒ²5Á¶¶rgžïÆب `óZ7|%l\öªÂHY;„­Ó8!…'€2b³&/øÂpSè—ËÜ3þ³$pxèÁ¸Q ÀZ7š!8úP¬Â%¬,Ì’Ö•ÀÙëzk¦«.LÖU[7;t!1'*„I˜ì±‡Wú(vÜËJ¯pËsÒ¸ Fæƒ` ÿ+¨%…$h—Œ˜Ní¨õQJãÈ Húò 6^-/áRüÊäŽ_ß©) â:(®€`BÁk¥às†Q¤$.õFƒrÖšÇö‘ ö‘{3ÒÁìdÀW‡x±W ©*lý0bënLçq:îXÑü@9è!Kʱ­«ö(c®g`pø ‚&€”¯bDÓB -ÿ5KEÍÍ6L­Áò¹™êyA¯°˜?%<4>£O¾30®!'z,ò¹»ú úzh¤îŠ’í‚Ú¬[J©­ÚAO½~±³ÊN•´¹©]t õ½9Äd¸Ö×Ô8‘öˆ't?Hœ.d¯ZË;T’ ¾Šô<Ëâxk}3ëÅÁªú‰ kÂ]®ÕI¢3…# »ƒÜ0> D°øQ›ƒÆÍL÷ž“t A‚ å%¬6®x(a>ˆ%¤ ŽgÉ»ùÀÌuc;ÜLj>f~ÓWSv]P=¢NAÍ{$Òtï=Jq‰jÜ·âu*jF9ábX¦s}íñP!¼T2ÔÚpuEÁ-º|½”'pB5Ÿfwæòqø³¯ªéû¹PiEyÊK«g@¢{)U Ûï)/4”ÖH~S2¬à ÆRPtïÜÁºÑcìýŒ@rw¢Póâu~ÃwÁ¾3™ èý¤†ÃM‘·I®8 S5”cT÷PPY[ÜqJP’ðÅBöpE”•8E"NA€;0÷ªI„³…Š˜åðLŸŠÓ]9Þ…Ϥ³~q‚ÑvÊ°¼MuY « pãd¦âÌ áÿ{Ëa9¾œŸ(%ψó?íô‘Ì™‹˜ñ/ô>°¸”ë«<•Jw av.Id.Jò| ,À¯ð}†² LTArÿ¹:™´¼ \q÷n«ƒ‘׿ zøZÌžL¥µÍÈOÄÅô=áO9ó`"çäS*¤²³Þ¶(A–HE] YU÷o;ËU1ÀùgpÈNh™ýTsùsˆv<úþ{X~è’ê¹ü—éå4=‹ÅV¦ÐF®ßNdðËÕ-hsÅ<#7pã#èòçO×?¾[ÝüúáÃõ/×ï:¨Ž2ùöygEð‰ÆœñìÏ?> stream xÚ•YëoÛÈÿž¿Â@?X{Ü_wE´±[IäÚëÁ`¤•E„"U’Šãú¿w»|™jr0 Í—»³óøÍÌR^…ð'¯²ð*ÑZd&½Ú_„Äm®˜xÿ×ÒÍ[ÃÄõhæŸ7/¾¿¢+Š,ÌäÕf?^j³»ú%øË!?u¶Y­µÖAüÃjmL¼*Ú®)>ž;»[í±nžøùííæg¦‹î€” Þ¼[Éànµ–qœ„Aœ¬~ÝüýÅͦ—)Rê…ǙϥÇÒK#E¦Wqj„Ô† „ÀMa‘µŠ…%³ß¼»»¿-ª¼,~³ÿ£þå4q­¬˜^­¥2rÓÿ‹æÚZ«Tà¼TÄn³Û¢i»ëÕZE&¨êÎJèy‡”Šªíl¾cv½çßm^–Eõàv˜ž(ŠDª2¾ú®*º%a Ì”‰ŸédxZ©4¨Ï¼O{¨ÏånØsiÃ$ pfûÝw‡Än—öͤˆu<Û÷ñPlqëïV´^–U¬)Bíu̯°QöÈ»¢®À}T4õ¹+*˃JšÊîp(ƒ¸Á?Â…‰‰×ŠŸ·]^íÀûòƽÐÕ<§¨vÅ6ïÜÌ•Á˜úîÎÍ$ƒáS§À†G§¦~hò£÷§±ƒ< P8Wr_–G|ÿ\vkÐîÄj%Qðó*Õ~Î)oÛeãÇB…ãoþöþæå«ûÛoßÞ¼¾yµd¦3GÐi:6néUÐòШÇÕ ûp¤º*Ÿ\lÄ‘U<=;9„»D‘PÞp-³÷M}dŠ ŽyQyjF ¶˜CÑãˆôUæç Ð`œïvºI^ŽqÛÁÊp 4° 0+ÌùAá×˱ð€c`°8eÑTœöK×äžç–E>KƒlmfñÉ‘E'Gb‹»‚O¬;+VàdJÓÉþ‰BP8ÂÄcîÊ˶fö%oÐÀM-yÃO7ïï^¾¾ûײ?¬ý‹“ÕÍ¢Çi¦ÑÂo>¼Þܽ{}ó'õÄáÂ4hìζíZyëat–Ìš â!ñÛóéDU7è¢aâ)tC•Oî)•¹Ôô¬ÅñTÚ£e§ pA”K°é„B( øW: /ÄyÌAÝÔÙßØ0¶ÍUªé²^§€ Ÿ €EOE÷HzÊi,%_–£;7•“òã•Tf"åÄ£Kû™¢˜‚WQšÁß!D`Т—žAÿíƒÂMȷݙ†§¬@ø ¼~ÂÐd˜„£.öLó?ƒL²ÂYéÄÉ3%R­½Žçv1שDdQ¯S¸¤£B‘¤³ìMéP°´y¿à4¸ÒâïDZ¬,ê#²Øó&CÎEÜÅA^¡zœ ˜oð×AGê>fyÜÇRôÚ, Úz´´{Ó§Fœbh:Žƒ‡²þH!ôçUCv,òå¢ë¦Z(ÕÛ‰t_ZÒE,…|—r*ìÐØmMÙh49&?£€ACB>yâ<‘‹dQº!êR™"“X•úýöûîñ¾øJ¡"C)ôPÛ,# ø*Ñ“€¢åOe^Ýca{_ý¿Mt%¨ùÄ’ô¡ÐF~»b{O§àŒ£ kÎu©ÁgH— ó¡·LÕ¾ÆÔwDíú‚*†™•[6e^›ïmÇÁÌ9>L„ÊÌ D+.í!6®¹´o­e–«êÆe½Œ!ÁM :‡Œ‚¥½ˆà™ÀzFHE5þ”3.öýûëÑ íÊ|½,–$éÝ«7ÙXN±µÇr^nAüì¯H¥ hz*Uk·Î A²«‡½V©l’ÊV¨Y…‹å˜w?AXU;4tc}ÅÅI´íxt1xB-"•|-xÒyðÀÿ¹‘ ,ºþ."ª¡Ì&{Oí³äæ‰&2~:ƒ‰»‘וðH Ô [0íÉïáqg¹6\û¤ÈÂN¸rïŠá”m]µlƒ•aŸ¹à@o®ùwWósßËêàPãÌGfßÞ‚Om0º\ÏÛ·/\ô£sDYE\ÇGžY>Ôp’ñí{Ó‘y¡’À EÇ!åE>¿òà’]¯¥šªü9haú–ª–%ô‚[ß¾”[ð„Xdqoö³mž3t&’q­Ê’CõC¹zkÛ–9®ÂÊUø\1Â=Ÿ˜îê‘ij)&íåy»XhüL—ØÝPÑU–­83"´U ·l²[*ì¬ô!¾äX÷\cG£ 8r‚ZWN×94>Ú–¼––Ạ™fÍjK¸µÕù*Ö&6¹Ôë´n˾ÁIÇâvìxøX´^ÏŒœ‘âúZ—‹!Ìsþi!RJÔNä =`Î .ÊyIQ)hP†æ‹R[ñ”K^­ 4%¡ù·6Ýp;éW¨Ý†6w %cF‰Û˜IÞŸjä'X9Sáb}@;h¿R7 ªÜ÷ê¸ÿw&VC‹ˆ²úœI}Š‰åpë€Ü}îÜ(F çS-–ãO<ýÜà!;ï7«F “i‘èNÄ^Æ+ø×gƒSÔ«]ò¦™z®ùºÅ+€ßm0è,øjNúNˆºrœjç^wgáy>¨a€R¸rþéÏCôcQ¹{ñ6•¥.¨=׉óƒØ/§²ØˆëQá8ùƒuž < Oz"ID/ž2‹ø”‘& ,4LúsBÐ`{ð]ëÖB8rÒº·ú;—ç—9Ö…F†Î_¤úíµ¿™BðÐ!‡&N妘æÚ¼j=3w 0žk[y1Ì°Xã´~O )•'ÿäŠäBo-? ,° e–VEÞ»+jsïPÛUºÈôšGÚao4Ñ|‚×…P0×ÎeÓ ¿0"ÍÆïrDñ¼«m—„ÏKÔ+¾,ÍîøP œ:p¾ßFdø%9¾`s^ÖEǼöÓ-ÒGšÀï8~$EÊÏ.ÜŠnWìóŠû<2²¡ÒHÀaÌ|ÙBì‚ó<»Ãœ^÷»¦‚Ðñ4¯o Jô¸…3†KR”ó<…=˜7:×g†Åïo€q½R¬Æñ²}è¾MÄPí@ƒ$° “ ™P9ta‚÷vO(ÝØjk—Rž4ƒ…Ò ô*è<)7¸¯È@è¢BqgiJ‚˜B3kô‡Î2³±{·=ä$›p‚@k )çOI°‡üK^àÜn—ƒWŠêüMÄ¡¹àhÐ `Ó@îØpjhðf s‹óî,U"Óé•’‘ˆµüýÝ™{=Z`©;›m3ý¢„ÂüIƒ_ÜÅpÞ{õýŠ ‚Ú Wñ‘’ós@ó*b¨ëÆç¸Ü½ùÙ_‘ZCO™i=•š: ‚úꣃ”³ÁˆNíÎTà±¼Ü8 ¦ÆÖˆ¦>ž†ÖÃUØt7…ocü5w³ E ÔKØ1ºˆÑŠ½!b„ĘüŽqñÎ…ÉmQò§_™ñ ¼©ÍJKWcõy)ŠЖ+`^Žª!ü^âÀçi‡ßµ}ïë:ßéU/rÜ+|»¨ :ÞÍËòRÃ307”¾þa±–œ¯ûC¿jÀþG,.õúTqøÓ×>ø9ÏxË_ù’ÔåK²þ:9íÅCšÄ³<ÃmÛú7›­»¦pX,…#a̤q×â°x•Œ­`?ñ¹t•èr=rÝç'ËRŸ8<4t¤¡åV+Ômâîƒt¯´·ï£;%My ‡~«eëSÕƒ3¬¤Ï Ý« ~‡Ö‡ ÖÅÏ?&X*¶ûO2&T£û8‡S òË¢úÄTW/ߎd"‹&—#PÂ-öÅF˜xt§jyY—B ¼çÇ&§ .ÊBn@Hÿ¦|èW^†³ABˆA¨1ù>IƒrÖ ¨Š/×i Ÿ¯¹ê£öf·»ôZavèï…Ö%9%^õð›ŽÁ+®Ëãò©2í×È»æÙj7.+FOJ,/Ð49AOÂg‰ÿ'é{* endstream endobj 1146 0 obj << /Length 3188 /Filter /FlateDecode >> stream xÚ¥Zëܶÿî¿â¾EWxUñ!Šr?%i\¤hÒ´9£(ÕµÜ[!»ÒZÒæ|ùë;ÃêuÚ;Ÿ^Š>†óúÍðÄUÿÄUž\eJŹ¶WåñEâ{ÛÛ+jüûo/Óm€p3¡üææÅŸ_§é•Hâ<ÉÅÕÍnºÔÍöêmdìõ»›¿_©LÆ&1°Šï}ýúæ?×¥T¤bø‡$/¾»öI¥üÌ!åÓ'2VÇBiÚû¯Õ/‰P®uõµ´Q½ÑYZWV]ÕÔ}­Ã†ˆö×"*êíÁmy £þª¦ï~ï¨Ñ¹¶*Ô®hi×îŠÒ½ºÞdÂF?;·dS'2Öb8r§0'9ü'¤gyÞ3å§o&óWØ_lâ¯àgWöÀ+úcIÍ·?…kx÷Ò‹m£í!T,Rš|*neäR,9ÒÄ™±W&ƒXý„ðõfB¾ÂÁrQŠøz“¦&ºÙ=¦ê^BÃ&ÑÿðP°†™®‘êX%YP¿Ý®¿{Ïd³­¤´ô+Zyw®ýuuø)¢×™Ep"*›£[Û-×q’Èén»Õí”u¢Ý/Iš Záž]Ußm8è&v}é@½ f”úQ½Üd*c!2œ  Õ‡¶h+‡ŒXí±‘»Ä„…ôp›²¡v流,7ôà—†¯ã‡BÃþùpáÜx€4B ă¼©«O(ÏÜFÿhêÛͶ9ðÌÃØÀ<}Vó'ÓX);ç¯;ŸNž§¦íÑ^•ÔÞN•L£~úþ%õ|8÷Ôõñ\l©ujÁÊI¢a/ Dy¬*“ÈÁb¬˜Ê÷ãº|AëŒÔ‰åVmQ7|–íÙQOßP‡÷+Ø8%:“_é«Ù°CSžih×´ã­gÆÛïýÖ ¾þ´øø¦é.&2`@»‡ $¸J,¥ag.}Ejdô^—?E¦£ïëªïWý^x¿²r°’ÈÀ¸Ø|¿qpLôœi•ÅáêN…w÷ôÑÀ-´Ô¤ˆ-Ï:6‘O 53kµ]Q*âT§ áêÁêAk´µdbØÏV3¿:ž¼G̸ÀÛ%9h›LÌuõ®ê¶Í‘DóFDR…žSÛ€½KƒÏ{¼“æLDGü8w= áå½Z»peÖß¹ÖŠ¶ü­©€‡”f¸‘ ä†×ˆCÈú_Ö®h3YŠÅ÷ýŽ,˜ã8f¿o]±í¨›•2c¥´¬ðð{îܙʳhß #wxQî7lºöå*#¤©²Òij¯7>´Û7çÖö7~འùvá÷=m±µz¬,N•~Q0,¢1z.n<ìûpGk‡›Ó@Ð=„ ¦°¥¹JUk«žÂüÍdµp»ØfŽT®Ð  Ëeôö[Ða¸ÔàfÝé14(0á ±þ­)±CÀ€1P‘q’Ù›—1D ~‚©å¢ážÄ#eô-û 5è.Ïš'MÃ)¦ÎcSl·^)bF﫣̊+«w3k.©yhåÁõù´¶=-£Ò‰Ž Úx_ð¶uCÎãã²”°q»?tBL<“HÈ–‹«±ÇG3 Û:¸ßÆ{βè Brj»ªëÝ8¦}Á´^‘ «uãÅÙù@g؆‚¦@8KUµ£î®ö¬xå;ÜÅ3.<–GN“…‡Þ»zp¿»_…‹1º3»oðÚ·ˆ$Š#ÆƇ¾½uM¤Ìm³”¿ÄsãœdN^°» ô&±3õzDµ hÈo‘ñDF]C{ €µÃ«“9CŸpêòâîKa§ßÖ‡¼ ¶Í™Dë§R5¼gqèÁáßê‰(€F!€Æ…gîŠê†ÉH‹Œ‰¶üÛ51YIªuôµž®¼ì+ NÅ6W_n–ÈG–²iÀÜ'¯È·´óéPø,¦àÝwœë-®®)Àâ"õ6Á‡í¨N~Í-5æú„KöôK´ª+€Æҹ݅µ?Ã_­¬˜Å‰M?3èÓyïßû;¸´]65q‘Žƒ–Ù6§% PsÒà yp,6º‚ã\b+G@AI7üÞº)9Àj’?ìÌX.R!b›åæšX*û|@Êó7“Öéb› ‹Ð˜”¢µ¼}3šWêw…pBKÊ%À7b©)—g ~âÐËE'€“b‘åe¦Áùf¿ öÀ”“Ä|®rg"ÎÇŒ–ƒÀƒ%“8[Ä©§´X$ ÊIÕ#¨¨½õOXK8TßSŠÌ9;lx Ī¥^"Óàò‡-ÊëôÔ~?`ÇßV™†œ.á®gvéÑÍ+È8Àõ€éÜã9òIþŒ_G郓BßöI96žäï<(/_먭Sæ6À—Ë‚éê Ð.­cÙ(qœp2ÏÆŒ˜ žî'µCAoMñ.ãa`¢Ac¸¼8pºrÕŒH;(ý5‘çq.† 'Åu „0í¹=x"‰àînƒ7S n *P λ!åòš‰v@4¸¹5U<‘]Sû÷Û]!‹1U#‹5²&øn™ë\Gð4Rç]œÄ%30gðAyþ|Çó7“VKÜómæ.ϧcC·?Â--^¯Ø÷]£øñúÚêábÑÿI@‡äÿpXÙ%Êä‘bÆãeÿ¨Ÿàh¹è$éÄSÜUýžZä´%ÿ¨Q‹,úùì‹žL®t¥$̯ ±Jf|–²ÃÂœÒYÔQ Ê•¾2éS6Í(_Ž¥ÐÁµ6îõJËÑY¥!÷ÔÞ‡ÐÚf± q|UÒ%Ös!݈t4Ö1Âcmv>ñržJõÅ„qb7¨è¾Ÿ’^NÈÁyîõü©ª/ 2OÔ§ãû] ~ÜØ+¬@grÐ(ç?(–»Ú½ø«Í¢®ì`É/IÔç°ó¾•å#,…§„?ÂŒ†\X(8”ŒÓì ˜¯cSùYÌ”²ýBùÌ}š‘ç±–öÙÌ ð0O2Ó>ÊȺT.²ðXYC È©óNO±3~î}¢3õPœ†FW}+Å,¬0<™Vt¸ ãÛ óC™ÚõÒ+§£ÉÏN¯ÙùG'Z™KÈH[ÿ”"Û᮸_MØ fš±Lµ ­ÈëùåO§ïD]Y€°Xš’Ž…ôbý•K ér«uäX˜§)d¨c•Ä¿ú>‹Øv¬Í‡ûî'ˆ°·;zO`æ bMyfˆÅA$‹s°Ï™@èY_HFZ‰ $!)ÅFSûÛj|\À)Æo¿-ƒ’POÞúcÃV ð—U-ŒûHÄ${ÏgPb"æ½H‰}—¿ ŽGÝË™žq¢Z]Ý1h—ÿPa – v×,exO–@꡺­CMV*šâYÁ7žê¶òel?c‰_K`µu‹) º=‡OÂÃÁr¸‡¿pØs» €ùyT2ûJŠåYqžY%?ªŠ€`¥/¦|µŠ:1/“‹ÌÿˆÅër5³7ý²à”Ñðœ|(úÐ܆SŽup†Âót¦Ó{ ‡'Äò5¹:£ú+£.BXúWh‘¤pØçÿ™I˜¿™,°aÛ,²t#ýš(:êÛܱiï©ÿëù•ø”ÿ$‡Rv x˜²«b±Èf<=Y™ú –‹N߈R%£›žÄõ`d[aàøp¦j¨ñ‹0Úâ#5Dr9¸ÃTQI§oÆù**“MÙF®#ȼ°‚¢²£ z8xƒ9&°ê¸üÔ Ü¾ã2> stream xÚÍÙrÛÈñÝ_Á·q<÷±yò¥T’u쬹•{KIÐB™d\EùútÏ @†dA^¥Rªæèéékzº{ÈþØÂÑ…‚8i›ÃêG«O‹ÐøùÏOX„[àªù|ýäé¹R F‰£Ž-Ö»>ªõvñ>yq™^5Yµ\ !ýãr%¥J^æuSåM¶]²CYÝ„ùóóõ?Cë:o.±%“×o—,ùËrÅ´64Ñnùëú¯O^­;šç÷$!¿¢Þõ‚P ?R¯­$LÈ@½&Œ¹\)Í“×@Cò(F&/Ó&ÅV—¼,ÂÈùÒÉäXlp¤FrÇ2!2C´a“g5¬c&Ùfõ0-WÜ&ÙÇt’†n¹„¿-¹I²0þ*ZgÙXœj"õBkMe^ÐXP`þ1î3éK(._õÖO¨x´‰çá]¶‰F4Š ïAu¯½È E†­žÈ~=ƒ1N“«ôSf•s%9Ò,lÝ®æú\Œ‘z6@¬,”GfÊjëÍš °S†fºß—^3›´ñZX€±Æ}Áž˜ŠoPqåH-“´ªR¸©Ï¢]Ø>ILS àÂ/ý¸T,Éveå¿Sv¦,$ká7U–6yñ)n>Wû´†¤ÓH psÄq‘°s¬›õ2Y…ŽH6hçÀ_è•E`Ê]ø6—q`W¢ €›ënÓªY!Ë gˆËe›S8䈜;b£Ê¾ó ­Ÿ;ÛJÙ ”½u!Hôåÿ‰ Òb;BqNLuZ¤@‚ ˆÄw;Uî÷q¤nÊH¬ŽXÒðù”‡sXxæhd+puUELY]wÉú²U¨Z*x«:8#/ÊÃaJÉÒk[%#ìp,r ØlgáðÔµ'Ú—K´Q…ÌZ“xÔ.NG÷¦<Ü—#¬¦”y[0a›2€¹ÁÀW»{ý2ð‘ÎÚ¡‚ž tåqÇh+Á©£UûI´ÏÐ æ[¤•+œO&£`QŠYZ盀#¤fÕ.ÝÄÙˆVFž÷M¾ÍYQÚt@š*-j€;Ôq[ðÑàŽê°-ÅCµûíZ?þ㔹%6/Œ«¦Úæ»ÝHPq•ìvÍõÅá*¿Ø—`]5˜íߢóô,4Æã,Žw&â{¨ˆ³ÖM~9Ý…`,pu)óªõ“´Ã€ìE•-vOþá€A 1Î,”¦D*;ÉË·Q§éEݤUƒŽôOMœÑ ¥,Ø¿œA!c(k%Ñ—Ø{ËZ)jN4ð•ÄZñY 4”y€¡z;ÿÂ_% `eù|•;&î©ÔG^Äùªø|Ör\ÔÍ4Ŭ 4¨¸¡ŠHÁfpÃ…ê…t’ªOÞÊúìà ·l¶¼¥¶DR1ÃÝ\xŸwU‚G~4Ï£CÏ#µcrsNƒ¡þ4H¥WâûÅ>Eœ41¿TœX>¢Î¤äÂg”ßI »“@v§]háü9”ðÕÆη Ž›á¿Ó.îã%5C I†ädsó»`ŠØ6<ÿxɻ̇R¹~˜ùëˆtîÑÍGXC¨Q³ÍGÀ5 ¤»·ù mç1=ºsàÑÙŸ2%Àø`%äUÌÙDZ Hä©õBàdŒœi- ÒqiÏ68” Ñ“æ߈UÆÕÁÁ°Úûâð"cÃ{‘qèVYs¬ŠÐöÑ36|„~ôÁx̪0q6~³}vˆ)Gă9~‡ ¶BC#¹Úgÿ¾}IJ‹¹ÉÈ1¼<_?­øæé†Ê)Ÿ~@.¾Òó%Ìpt/ÎøÕC, –ëË Eðf:ùœÁ]gÛœ®5Á©ìn,Ñeø˜ëNe!Kj7ÄKËLf£F[¾%Õ”¦K5;j5ªí(¹Æ)ÔÊUŒ:HBFLÞ±¹…»Yt\„ÚÁW8!k°b¥O×þÒ¸N¢~p>ljW0 â\G%*2°êE(XÙQõ"Ô@ïN Óåš¡ë &ðM‹°±ƒáX1Ö©6ÂÈãÂútöõ ª‹ˆä@nÞuô >Pʧw3‘³6dß¿¡˜µ!¿mCûÕ†+†-†¥aÓµtlÅÙIÂ$$¹§âK$ üà]”Å“  ò7Dg:”N±Èyr‹NeŸ½®J짹L£åuø¶Õ)oeE MòTÌ ̵óyÝ0Bª¥"y÷u9™ù\eÁ dݳ‹Éqñª[=q¿ 6’sXH& /:š¼*?8ìk~8˜n[* ýMG¶ƒj<ÊD5–›I(7㊉r3dfÔöÙ¾£ØaïärˆÐ³Ù ý|iá6¬¦œÝJÒˆ»¯9|"¹ÀÀiý󳿿{ûæÝ«—o~YOZäÂâdçí]'e[Ä…‹ò_S;C*¦N÷@/Ô›ÚåtÂ@”< •Ä*O»{GÓ* cíŽûÐÆ;Rkì8„÷¿‡ß×±|Ì*Ô.±önöÚZÝÀò}ÿÊ—ŽÛ» GBy¯·¤óßØéùoìæ=üN’U°xõýþT«å½ü)%§ûôáÆ#îÇï¿ßŠ37á¾ÛVœ¤Ë—xØÿ‹ûfÄ€ãëmŽKˆ“â™HÌR€Ðó\›«c®@›t¿ŒP0y. ±ñysžëW=“Nx¸ÍèA_]Á l²äý=VË¢ ,ú™¾ëõ~V×>ëÁ´²cö„¬àyúìÝîi[èo03FÚz[qxE0ñÞPí½#ãÇ ¸Í!VÜÅûº»ï ƒ„[û÷{;8¤ÅÍÃóíÑøeyø¦rnk}õ”à Äø¬rsÚÀypŒÈ[ªe¡ÏôîÀøaíê°‘Gf±³JS¡`´"Œ²Yªû]ª&÷Ö¢†s-ÁsjN´™U{ƒ\Œ«àœú¶§ææUµõeB&!V]ak©¨–G)òHE ! ×ÖîMWDËG,Fmsw–šWãaTùŸ.ôk<Ühpé Í°·«ÊCh…ahÄçXn”÷ æäAq#(œ* ÿ®Ž Ã0ÞaÊ·ão"ü‹¹JÎsˆÈÎâ ôev}=`¯C·û1ƒÇ‘Å[7Û仸ßÇ6&Ä>O8<šõ¤{‰ˆ.º æ>ù K ŸøV…‚cÈØ­ÈT¯AC¤jU²O›ð3*«P؈³úst2 bx±dÉ1ZÌJ2Þþ%,ßø·whà礪 !2èB­.÷8Å4–Ø^?ûå§õÅóŸÞ¼øÛÆ’S½ÅGûLaR¸Åuho³] 7lè|l ‚ŠÿÆjÿ+l¥q‰O( &ü ¯ 6§O6 榭Ã@§,¶˜ Õ‰'ž®ÿ²%ÉÙ endstream endobj 1159 0 obj << /Length 1989 /Filter /FlateDecode >> stream xÚÝXÝsã¶÷_¡GªµPâ‹ “éƒÏw.íõ®wîä!Éhh‰’9¡H…¤Î¹þõÝÅ‚Iƒö]ã‡LÇ3&,»¿]ìø"„?¾HÂ…‘’%*^l¡­÷ |øÛwt+ \ (_Ý^üåFëY&|q»²ºÝ.~ L¸üùöû…4‚Ea\ììÍÍíË•”2 þäâ»Ûþ-Ä „”ÏKÅŠq©èìMZ.WJÈà˜6 ŽD¢°/îÓš£:‰ï«‡CZ~v„£”b&á1­ÓCÖfõ%ðV:ØTu5ÇåJÄAUnórOg¶IÑÞg4H·Ÿ–: Òr“m€EZ–ý†¼\ ã]ºÉhêÎrÍŠj Ÿ‡o,Ò+†L™’qMBµ÷9¨*àb¿<ØU5M´uZ6ðëàªlª¹¶ùþT@ÜìüK Y˜tt«öt,²Éaö¨ÓDLÕŽâ§P‡¾Ã“˜)¡§‡[—ú+~TÀ}¢Dr J  PŒ»´É7„¢4ši€Päý„„L’qf‰áRÁtâÀ¾µì"é5=ÎÓ1v˜–[²¾î¬KuujóÒ"¿È^0¨ÊlµÍYÙäU™´{dD BÔ6'8|¦øÜÉ9Ô¦* ‡Ø¦:€M~#Ü^ßÜ6¨-¥uö³x¬ö±­·ùn·n—p+u°ÛµëÃ1_Õ&-ÖMþŸlÍ·9F1¥åX  ‡2¼¤ÁÛ÷oÖ×Õá@¿@؃[ÈKGÚäûÒÍJü Z)Štß\ú´Ÿö'³Ìÿr¾nÚ´nÇ ŸäW=ɯ"~÷·äw_€‹YôþÿÞɼ»K>ÿà†à/h‰ià€°*™@H‡>eˆ RŠÄ6™þ¦‹î*¨³öT—4¦8_œÜRî˜ô´}4Œ»_Ó†KüfEvÈ(u{+'HÉ·lÒÖq³± »à3ì+nd£ŒÑ~»Hdl‘ÜF;_r¸âØt¡¶3ª7=$,ä}i€ÑoÂa69§1,Ò}NNëýijÑÙÇx‘¨àhsOU·Zii.ýp1Æcy\æHàliFç I¸Sܧ$°£ï4Áxù•·QúϺ{”íäŠ>Þ<¦eIÒKñ^ˆÄº~ ð-}ÓºNÀÏÝzÚÒÝ¿ðá ½äA†C·Ï"i7t„Ǻ²8o2¬¿„ {·[‰0b&Vcï£{F倷KtN©xdt(|b}ÎîÇ“q¡˜–²#»tgÆíð#>ñâÊÇU‡L(ñœ‹1«ê †kÌÙƒ; Œ™ºæ=¥®pkó¬q°B1câx «­(”»¸ZÄán¶ ÕÜÍN"öΈ1Þ ²[ ýË{Ô3l Ö7ï>üpõáµèÂr™Ç÷rx--ÇWWןeIirt ‰óIš^)›Ò|\¡ç€fí©i‰ó!m78qÆž¢†V,ÅŒIÐRah;[&IðpŸ•4Úàmë¯]Úæ0$oEˆ; °j¤dpU´÷ÕiO ®ôM趸c¨(…I¨@iܧ.»™ÀíÙæ?…\fµÒ—9¶i›b$Õ@Ü´u~wÂØÚÐTnƒlì³2«Ón¢ e·nÕQíOÐQ¹Â•Ç.úÀjÛ;é(3£ÎÎbñÚ &êmå—fldL°Aó:ŒÅ9RXÏ#Ë*È„2šÄ´'ÜÚIizFË!Ú2xÈ‹‚F’.}y H& >&,Öz r€3Ô¬ÉY ¬rùåW $«éœOŸòMw7"È z|;¨Êª›ô[ V.û䇼ÀQÛÆF)×5)5Àh"4gÑ$a‘ûñúÃÕÛWÿøîõúÝ¿o½ù`3gsÏ…{Τ0OpóO?sÅB¡Îa‰kˆ4Xo™(ø˜eÓç’$\H N&µ}-a¦m›Ë¸°Ï&ã™éƒÎ`¯ÿù¤#°Ò|Ì6Î% ¶ˆPª=~|÷¨1•c/j~¾¤ÉcºÏh¤“©. +m¢‘BóÏ?õ3:L™ZEAo€,ó…v!IÈ¿–4ؘ†~Zí·èhï߀üFïñG‘bÀÅ \×Yjá9¿ (H–Ë5UÑ+wÇŒmli€5µÏ0šb5éž®ÞÃ-i9òlÀ}3æ*©@§¦Ý汿 ºA$b 'Å_í4ÝþÕ€õé1>÷‘ÎyÄ–fd¿¥¨¹ Væp­ÉkT ÕT#¥¸}j4ï5õ3òO™j÷2â,€6Ïœw¶Fö>™¬$xCdÏ´¶›§ð7jîqj½Ýµ®«ŸkÞ-}ߟن6/ç× ^vàýz~oƒŠo!¢š¶>¨Hb,êl±»ø—ÃhRƒ ØÊ_‰—y&<‡phä´š> stream xÚÕZëoÜ6ÿž¿Â¸/•#>%µ¸n¹Ë ‰ƒû>eWk Ù•ö$m\÷¯¿©×j×v€=ðR$5ä çñ›¡øIü$Nb)Yª’“ÅæQd{ë«j¼ûõwóB˜fþrñèɹÖ'z~ÑíI qÏÍãÌ»woŸT´û‹ë¼Éa* ²Ú6â )6Å:«©·­¨³½Î ×Õäu‘­©½¨vå©€iy½Í궡ɟ"5y>åDʈ ®º 7LËRøÇ…ekÜ3äÏ¿Ìð8]Æ2ú>_´EUÒ®“ „ÎééãÓj³]ç¿ÓÓ³ó‹æ·ÇÐÖ<ØfWN&Bí Y¬“3GŽÆ;cëS¢vë M«'¡Ž —Œk-'§¶= á òE±º-€¹òŠúá਱,6yÙ€ äNDpÎW¥kfå’&}Š¸Î®z¨Vøkzm•Íªª7 Ô\¦¨=¸1àÂŒ€',•8·;\T››6bV&M7+«¯v°A«L´ÚUñŸr·›¬$!H¥™t"àôî«·/ЖZ ÞÙ•Å"k«šzÛ묥–R Bs `V¤r|ÀE¤—ÎÌÝ+Vð»­++àEÞ4vKÒŽê‹b›µ9‘±‡bWvä:Éý„úÛµc's²¨óÌé©Ž‚ÂÚSdô¸¨ÖkTc õ­ve§Ø©çúI {úLl;2Yº×+&hC75ˆÕ˜XõÂê…Ûq$î¸7±Ôy˜9œéþð‹rNA¸bBŠNA@GçELjé'U»v–RÊR°7‰,¥*¼çö„ÖA¯Ð ­‚ª\ßR'1ödôHúSÕ$rìAýÀ V(¶©ú#âA ¶’µÔ¼Ð;ˬutxLG,–à$ˆ DñPéßfÜÎt™±ÇÄý¦œõ%ŽâPk?Zè=‰#Ð. ô§ÂhçOqž6SV˜uJcÀyÈ„ßáOýìp0}†±)ÑÞŸÚ=f í¦÷èÐ6°÷3ªö-µét¡ñŸ9Lcfx§Ìë ,ì²)þÈçTQÇLÄRÿ@T½7´/4´Áª)Xdþµ¨v; Gb| xfžùC¹ãÛk³Äp¿þjÞ~f›R°4é=8™JV”³4•d*íøGsùüýÅ‹WgÏçhÇŠ)­ütÐ Å­ýX$’±Ôu†ëÝ¢ßI=2†7-peõM]´m^ÒÀrW6àA»s­¡ ‹Ó‘7ø×$²ç¿+ëïÙâ4h:Û%¨cÞ]øU¶°m¸NTp~ ÿ<ò¾ë¶»¨ëÉt.À†á”c\#=ÃÑ%®¸±tËSëxá¹¢æœà¹Á¶åÏ„-ùœÐC?w¤5V´Â˜ Ùm;ß–/B{HH.8Z•ù˜´æ'„æ¼&̇ÀàœlS´9MGWVe¸­Ç4{ÐÌ#J:U*£qÐY ‹]];û K‰9SÒL,…XMˆU™Qvs»r]|ñ¡ù»vðÂDTnÜ<£¡wùwyÓ’òÁsVÒ´]9·ÊÀý[ºè"lë¦X;ú·E¾^΄ÑlNZÐigH¯?¼|9‹³4ª>øyÀæã˜Nä¬ÁŒEY—£”$¦±Ã3…UQˆ”¨óh¯1®‹«k“nZeX¤&Îùèå—•‹œu¸%•2©L8€(ø‘å×Smœ¼ G<Ì ì#NMçWíÍåf[\¢À.aÅÛËåj\@¾‘ô/fËeÑZ3!†S ÿÖsÝ4ô|‹ÕŽ`àï¿»BNì ¾¤b¤Ÿ×Ž"­/ˆâµ96É£ ¯Ä ²Wµw)- v&M¨³’1`”²ó2°ôâˬ¢%ÌÈNf‘®…Ðn/ö5#G6Ÿ°DtÎ!AHßXŠ@&¦" Ó$·M£†tàø ëÌúEbnA Ÿ[ZOŸõÄý× E„‡q[AèZ±k¹ÑÙ}I@urº1Ý™ÃàÜ fA®ªâþØ£hŽE!º)³ñ‰cB ‡ìåvçÕÂJl¹,\„‰˜|„á*p©] EÚ½¦Ó𼦧Šõ¥=¨è‘cE·iƒÍäAEyÇéýÁQ< “¤ÌqÅK7Ƀ֋þäõÄ¡õÄÞz€kè9 Ϋ9‡øßSÍÍ÷VsHÔ">d΂>mîä3CRÕ¬¥Þ.c%„hÁ"Ö(J²WS¨Óæ[›¼;'¬|1G àÛcWÿ³5 åcNú4Û`bMÖŠÕ|F¯™âé( ¸xwöúýÛ7ïŸ?»|óábÞìøÜÜn÷32»p9·°°$wf4ûщùªšb1Û¤j)€Še@t$øä€ =|F“mlK¹¸c\Ž1.ÛQ… ™54íÆâÌÚ½¾Í{ò8bñÀö¢Xë¤\Z¨‰ý³™(^Ó'nKE4Pç±ùôÒ ЯÇ/©y±LÌ㪘µâÕÛ—ÏžŸŸ}xyqùËË7Oÿ5ª88ÙN1¬X`•=h ¥PšB-óUÙ L‘öÒÆVÖxL•21 ‚uäÒ@l|ΚbAÍ~‚>ž­› ¬G+Ñ»ü)tpu\\W7ˆ$çøUœPç;¶Y *eïpñ‰94~›èæQƒån¨§p“«•cÜeO£¬Øå÷ÅÕΦ|hؼXˆêBhØî¶kJLPg°ÍšÚ°õ’Z¨ô×b¹£\ƒRG[ŒW*ì¶v6›v†W7¨„ ¼hhRo 1ˆ[ Æò›jf„ž Rjg·Øc¡nƒ~5w÷9ÊÕqõàŸ  M/4¥ÝôAŽa»]…Á„Ž!W,æáW!îýp@`î*d²Ìä*R ÂmõãÙCOQx ¾Lþ»5nù¡T»›)ýÍ {ZÜHðQ°©s¬ï¨ôùÙá`ú §S¢ÃÄ_·ÍùKa]äÃe],szã´ý#ªžrÖ—^U§9¼s±ËÂä{ê³Aña b4XYGXÀÁbŽô ‡‹~ T#^ƒØÄžnqfˆc® Ù$Àþ£Õy¦Žï.JÐÐÎ+,¼I÷1R™¾¼…BÈü«Æ[´¡TìkfPy0îÂ[Ö¬uîS N–''˜€{|¸1¹÷Ã9cš,36&Ü#Þ+ âæ#XÌÛuV–¹cï…æAö“¸Gk3X[zÏf–m%dÆþÞa3nv8˜>g3¢Í`L‹(ײû±Å¡8êJÔ»¡"­Ÿ²Xï–nÿU™Ó ölqjUû¡ý:xƒµW[‰ÁD ÙD¨Çá8â-è%§×}q˜Êð溰qßÝ­y°ÅdÐ.Ó]ž¡—š””¨—. •wúˆÀ¼ÙÒe "l!ìŠ Rö“; ,]9XÊ.“•ØcË.Ú6l.¦´«ju¾ŒÇ³C€©ê=|õþ黳W¿¼<¼…fišöUø¹t˜ñ˜!ýâõ<åpe÷Þ§aÂE@¥¸u*¥yLf î:új2=Ûbr¢o1€@”â{qܺÎû‹kÅ#‹)ö2‡&Q+§ tHOܹàÕKù)|X«ÅG[e&Ú®ÃýîÊ¥Cüt}ñ9wµû‰KÕ,ÖY4³Î—"%Yæ5VdeœQ,k„¬ž *3ÎÓo8s1wÊ2dV%ÿÝC-꥜“¬¦ö¬Á¸dqªÆ%Iït‚×¥JÇ€Ÿ …ß \pi+ê´L€ƒ|.j‹ÅšÚÝœáv !KèãOd¤ãšuîõÛ†$éËhµëâ{‹ä.P×÷ûa‰sä' 1Jôv8x}΃–˜¢;‰×¶L9t÷fz‡cÝÑà–v¼É@§ûÀ5§0fëp ò³ï`cJt|ÆŽAÞ&-.dƒõ@ßxªÖŒ„¹Ã9&qzŒöa½ø]ðŽ)7#ïhRç¡Ñ]$OŽ,ÐŒ‹& 졺¦YuõʺüB§ûó*a—Q2.qšTÜ4>“›–x‹'Ú{dé<2>Œo7,!ë•©åÞð¡Ó¶kúu.Úv­è7+çª|X9Œäñí5ϘE&½OÍ,0êëWß^ÔÅ“ï´ ¸?‡¡É^qµ/¸ ™®³ÆL¥ê¯Rg…ÕÄ{uVwù ÁVuÒ—_cW~µÕ¬ä!|ªø/~Öº6‹ûsx@¹º»*ýîÊ•Ü­\æ¨r%{¸c„Yƒ€È€@¾ô¯nÎ<\ ¤ž^8Ü T “=  ¤°rÄÚ àfßÁÊ”è± ˆöwy¶]Šë¾ ‚¥©cÕ·&ìø Vê?}Iã F’6 =ññ Ç1ÒÖbñd!jwÿ—Ò·Ãý°¿lŸ)Xìw2p2ÑÃ?ëóï‡sÒœ,3ù¬¶ 6C\~|µÿµN ûY®ÚKÐK±D5ضõ²X­.[šZF©1íç”銂Œž|û²¬v1ÒÄ¿åãÁª]½ÜŽŠ>žK>š>ůuí4ün×ÑØ•ø8JÁRÄ"Šê'wÖ£Íü°ŽÊõ?yS5l endstream endobj 1179 0 obj << /Length 2063 /Filter /FlateDecode >> stream xÚíYëoÜÆÿ®¿â¾•—˜ë}ñÕ"ÒÄN]ÄAâ\ÑŽq¥Ž{'Ö<ò‡eõ¯ïÌÎ’GR”-µ6Ô …q³;ùÍìžXqø«„¯"¥X¢ãÕîxÁíh}XQãÕwÂÑù@è(ÿ´¹xú<V‚³„'bµÙ·Úd«×^$×o6Y©H²‡°‹}þ|󷵯”òƒ?$¹x¶øRÞS ¤¼-Q8–H‰„É@®ÂX3¡4 ðòÇÛoªãq ;ÞZO¨Ù•M~(MF½}‘š_x þ`µð%çLÃ^¾PL´Õ~ß^oOEZº%Ø=žr;´Íöí¶–»­Ê`~jë,ßï·-‘–Ü1‹;Æå“ÞR¿žmhÆÃx†1Ź4KbVXÕfµ¿øéƒv "+EÊdUwYb÷E^>éF:æ½›«ºÖ ¼ÈâQ ‹VÍKgŸº|û¤ç[6í¢ ¿(Il+&£©ì¥üD–{꾓õV>Ô©ÛÔÒbÄnªã ÿÄ*ÿ Çþ?®ú燒P"õŽ%(ÅeðÛ´ì'Å’ÇQ~ž})$ã šâ,Œ-~Îy‘Ök_FÂk+ür¯½24И:O jçåZÆ^kê}ºsÓh¤Æ˜yÆ–g |B‹E*²®Ã˜c<BÚô=çñ~½?Ú`¡º˜³!}̮ͫ’äÓ¶ž°*½~eÒÂÏÒ6¥©oŸoš7`D+B2šë¢$ÄdªËÝHOýÉç›ZÉñÜœ@`ÿÆIÔÖiÙì«úHc0rç¢:ä»´(nèÐC5aG®d_j•D¢W!K"#AȨ–|ÁÂ^îöÑã}8.8Sý¹\b˜°@$â'îâ§>?y~~`C¬}†C+r-š\”JI%s±5?$Wæ¨ÂT©@Å‘ø’‚Rµ´X |q"ÇÊÕ†¢š{."‚ÿé¾®ŽKÚ@±œèàÿNt'Òÿ‰MÏmêD¶‹Τ–ŸÆÓ¢{Z4ó´¹,:LÆøjI¯2òàMdŸ/í©L%^Ÿ- õBÁ”œÕjè¹/!?›S ˆ *òZÈSÂ; wë>‘)¯¬J¿6YWf)¥1š¿J‹=TîëVhÌôÿèiküææx4m}ã»öÔµãEÚK©‹Qåç%N/äpÈ8pËÕñ]YS¨„‰HÂí8b Ôƒ³f¿Þm°{æl¦YåÓ,f!5_¿ìŠ6÷³üh Ð¨JD ߬!Kõ™ÉæQ—Gíá\;•€Éd¢ÝÝy´§þˆ.óM‡<ÊÖ¾ŽïÏžõ5þ3ïÐ9LmEtÌJ' Qæ´®S¤¼q>Ç£L{Äù @¥¥w‰ÎçÙ} Rj«Ë0œ¸ÎÛ«¼¤ñ”†Ni–™l)dD2}Æø\aAp/pÅfø ЪÏÄPÞŸ¡/eü|u³ËáÑ:¹^Ë»Êæè |¿IˆE+|‰Ÿˆ@öÍÚ!„Éç#ðyë¬hòñŽ^Ñh‘¿]÷Ž.mäÐÍ* X(ÔäòÒ‡[Þ×ÃõÛµÜQ£¯h¥á>‘ƒ!ÉïQ ®LíÖ#•mèúÕ~¼5î;4¸ÃˆC4“ºïµ ø¢èñ@iï›®® á?TË ?²Ü«J*žçM¶¸€Š²ñôò›(ƃ!ÃÅ‘^6ÿˆ½\ž©š£"S ²‹=…›(tdr¾44™ZA£éN' DUÝZb®6¿v¦lxp èC×^$Bã.( #À£$¸¥ÁWtêš“kMì´¨VA?Ô¥×y†‹^sSdKŒu yIõ ~øë÷ß/í*%6ÇÆ‚³ á ¿FÃæ›÷ ¢­ úFF\Sk›ÞõÞ'wåbZò¾RÃUY>8÷ëýÑKùkÆf–‹Ant1ÁË_^hiª Pá„e°W6Ýi{ßBù„‰[ËÐ%n ù­²$Œ!Ç Á%?’¸{jD¾T„Ì6_€Q ›žQºžØVÞ´u amƒ#X}á6òªöŠºÃÚzÐ5-3j o0v´Oø õs÷=ää!%™1qþ’ ¸æŽ4þ¾äÕ*f‘üµ¨à¢¾mòšEІ*æ¿£]÷]éü{xÏÇ(סãD‚Á;ê[ 9âOb‰CŒ&CºÙ­¡0:—Ý·-à¢ÑÓž‹%@QŒ@ðÀá)ɲ¬ú | œ™ª/PÍÔÔD`* ­P7§µM":¯ òÚ6í©Ú]\InG‡Jýž¦:ØúºÎÛÖ”4q}Õ·&$‘]3É ÎPàY„µ€‹æýÎœZ‚÷¼\r8’ã_»¶Ï~Þ¼xùõæÙ’£˜îXYÜÊ s¿„øº(§oxú@;J—}øàÈàïØq­Ûhˆf–6vj·4ïiKÚ3ÍÞ­ƒº»ž`úˆC‚Pñ¶naQPùí6ºÁNÕ9ÊŠ¾GºŽuýóÕ<7”ÛÌ.ßß½/ 2É×¹ÜdÊ·ÕvÉ1 a¹àwxD³ÇW;+z~誮YL‚2fñ9l¯ªkôf1*øè´ü¶s;)0øÂÐÓóAÀ ©l‹ w‡¨›ß/‰áKØ.lÓãþàó5ÊÙÿf ÃC秄?n/;ÃЛu`«äq Ì¨s8»ÝÛå*7·àõK\LuÙ½9LEÛã™^|Û?æX~ü 3Æ,B endstream endobj 1188 0 obj << /Length 2068 /Filter /FlateDecode >> stream xÚµ]sÛ¸ñÝ¿BoGeB”¼N;Ó^šNÛK›Þù¦¾ŒKKͱDª$uŽï×w P$ Yqr™ÌÄK`±Øì§ø"|Q$‹LJV¤ùbµ»Hìj{» à‡¿^p‡b<ÂüóåÅïÞ*µà +’‚/.7cR—ëÅUôÝ]¹ïM»Œ¥”‘þv§©ŠÞT]ßV7‡Þ¬ãÙ5í#í¿}{ù‚ªþ¡4z÷~É£¿-c®u–D™\~¸üûÅ_.ž”ŸÈá>î%K€}ǽÎSÆeJÜ¿{7ËLE­h èÛ²î6M»ë‘¹ô ž1-%¿¼3­‘y•dѪi[Ó헱ȣ¦^Wõ-!ì·e¯ZSöUSjÛúª6!À¥´ áà;öûçD%1s%¥R³¬Ð ­5S ·Š`‘€äðVcÓ•±êüùxD `üù5V ?š•)S$ÛÕ»¦µl.b™:â`®Ü©7o/A$™è¨ÙÐßL¹%èMÙ—^ÌE´/o ­òd.µšå<_h%™Tg^‡GŽØ g$-« sî¸YZ´pÄ£z î·Ä]wأͳ¨i{ks™dÑQÛ¾ZW;Sw %/"ÚÑ=0=¹?AXËö궾'ù#9 =J©ôG[ºcú”Á™„§;5CSoožä(g<OúÃs L¥Êy¨¶[:óX™íšÎ•¡«”f‰æþÜ?úþûu°Ðˆ:ºŠümˆ`Ìg ¼5°(œÌ›Mÿpç–2}îö•]º†@p-Ö¨Œ}ß®«Íæº'´:yMÀ|Þª—gð5Sïº9Ül !¾ªjGa² 1€$Ó£š®¿kv;B[4EÉ‚<ßWõšPò¼žØæÁðBokãð6Ûò¶CÍþž^OP¾³ª”/Uep]„TüÕ5Ì¿PÅÏm“DÌþ7„(­8é‡ ª2ÎRVä|ANØšÅæâß.nM½´ÐLâaŒ¿–AÑšUíìÁàµW^ÝõAÛ½ªítÂ$8÷»ÇiÛÍ9 ¨þ‚ŸóøCÜ?««]Y?~™ÂòœE6½r†[Ýl›Õ}Øi·÷Tw¿±Ë¼HíÓ¯RÁ4—0R&\ÙòÉO,”ÙBæà/Z7ã<‘Ä‚…Iv€š+˜*¶Ü¨mméÅ}…—¸¬¾ZB²¿£•ÞŸéà„ÃîhÅUuüˆZÛoÍGú 2ȸ¡’‰ÿ‚ÿPm*“È|\™}ï/*ûÉ•ÈV²T["G$ž Ç{.m±… ­-¶²eÌÁÊDÀ°ƒO·DâÇ©WŽTôýþлž8ñÀÃ…bYêó7<Òr{ÝU¿š‰TÆD6ßÕ5”‡ñÚ·Xzº2ê|¨k§>z¨muŠê<êîš– ©Î¼´´QQ5gÚ} &_ÓbéN­šm»o+'üuJD¨™[¢ 6ªý9oiøôL7g¨g|šlçiýiÛ5!gE]ëD&ZG UJD|¨”Æïøì¨^5«jóè9·ÚÐÔ áŠ¥Ì#w¢Ü6¶› SﾆòÖáu´LoGî4mE:ÓV媜Þ/˜– ™ÏÌWÚGÇå©Æ‡Cñ›IÁ `IúòÆÇŸGmÁüšyã#£”Aq ˆ®°“‰û&¦Žw/—à+¾'¬€j»ÖÇ¥ërpY>érd‘3U¨‰ˆ§Û}F 9Ñ¡u€ç¨ÁÇ~ õ˜’)pÆÏUµ?†³ù5OU­EÿËzÝàý–ú$]aïO›/жµF7€n烡ÒÁPÖ<¸êšPœpˆ'æQ)Ks˜byqÎ:9>b‡Œ3%iEaäàÓqF,´§ï3‘³¡†m÷عFMWùHyj¦áF"nFS@øƒ„PíÛ}õ—jí›ßŽ6Ž <Ð7ô—j"Ü/݆có[ghͶýG”õqdŠwR¤`Þ!–ÕIHJ(†ÞùÆ&À1ÀËÕ=k~’Ö€S¶.ŸŽyw ÀõÖ’O1}~¡ÔŠõ‡ê #cVD6Udùø\† B|ƒr­=SwÍV™!îStÁ¡“Žûä¡n motÉ VNT°CS:ÄvÖƒdñsÂS› aJL›Cl.¸w˜ÊÝñ1»Hì˜E(ñâ‚óÉÑÙð|Ò#L#‡õ^–‘C_'‘ä-r£‡‡êü_>ïÿ§8Ü6é™à°ÏH1':„çcŽÚKuÔý=ÛžG‚ŒŸ3é€j=ê«·rgZ²ç[²ApK¥àøf4~ƒ…óL’‰È&Ï°m£vâ»×|ªÆ-øYJž5kR³<×6 %‰˜6k<™vjsŸä‰¯N‡ ³€Ð…M¶GAàÐ!;QüAèḎÔÃ0p°·ùhVJ° wHxé(›˜‡x€ñ>G€>|A_'¢š•x)_Öðl<:lu¨ÓI'4"OÑ?÷¡+Ú¥Cí+k®íúd`0×>ÛN/ÃßPä0Q®B7åLê!=^ýÔÑÌâ¿«>²è=™×lÍ Š£û8„9yœzŠhÏòŸNÉÕ“íܨ5ÓsÛ!ÇGìPÀž’œün‚ƒz°‘A]ÀK~$­°jKûStB´Z:á3ƒXuO_¿š¶q“}ÌÝôkUæÜÇYÇç$¶ÍîøÓÛ7+g}ÿé“z_ÑÄyMý…|" endstream endobj 1204 0 obj << /Length 3755 /Filter /FlateDecode >> stream xÚZYoÜF~÷¯˜7S€‡Ë¾x8Om¼ëŒÍ&‚@1´Ô ¥!Â!’cYùõ[Gwó˜–eÄf³ªê:¾ª±IàOlŠd“):ß쎯êíï7Üø韯„·…ÛÙÈï¯^ýí1‘ÄERˆÍÕÝ|©«ýæ:ÊôÅÇ«oT&ã4Iaê}÷îê—‹­R*R1üáW?\ù}Œ”_IŽ<£(ŠTœI–¢4×±Pš÷Nc´\lM*£[ýøHÉtô ¾ÔÃ…ˆöÝ»LtÙJD§¶Þ•cݵHéše‰ÈâT)^ÿê"WQw±Õ¹‰îÊ]ÝÔc9VøžFÃS»»yt軶þ³nï¹û±hKl—ÇŽ»M4ì´}ý["TÕW-N¹óÒÃzØ úwÕ0TÃàLêèGÚé8Þ>Õ{Û1â+èèt #˜ƒ»S»Cf‡·–ÝtήÌãÆq"¬D?uõw~ïÆÇ›ãC}s_ñý 3ö[b ö†¤IãvЂ^ñ]hûÀr·}Wîwå0~ÝŠëÚžÓ|² 1'b­•SÓ0;¡ìši¢• œ‹VÏðEGeÓp¯;tl×v$8Žº¯?ÑÚ~äîBfV»>HºÉã¢0Žš’…N@ Ð^È|RIÎ`c®TØSDÝéËöwn%ü@âgs Kmî©} Z§$|/Çðâ ‡·cY·ÃŠX ³‹óî‹;¦³Xë‰l%K+Uèá}³h‡–NÖiÐñ÷ÁNáÇ@–‚S+2¼ÝV% øÐíR)YÚ~¬ys<|:\Wî÷5*EÙ¸1VàÛ]¿âŲ» £<³r¥tt{o5r»k›§ó¡Ü“ð¦ ;˜*l±°¡ABéŽM5ÚìÉBë1_¶ðºZœYÏÚzƒ¤³”ÄÍÝ3(NU$–HhôÀo…„²yôCÊeLÕ#¨>?ÐÌ®í\'f•,fê]Éz~Ômˆ3ððI¬3µ8çLL0D/†±[©£W™Bâ¡"[©­ [Í•g&eõ±¯10h ŠO¾±–QÞW<6Õkf”0±JÌ<;Üà/“¾Z’HYc£èXéTr€’v=™ÅY.=RQŒS,c€H>üˆjbË» ÐÐtpÉüõ}{Ž¶êxTÏa•Î!Ÿì[Û ¤±’ĮÅ]?w ËåmöˆlWØ:²Yí«¾åõÞ²4Òdð1uÃñ¾§o±¬È0†õ~5;àêe’¨p,ŒÓÜ£Ízènv7·u»Ç RÉòXIï{†±l÷eD!k:`cˆi"{/¹Ï#%ˆ#‚ÆìßWî¯oɶ«=÷±c™kœÜšHx@YñÍöãæog ´p½ 1yy(Æʲ–1S×—–EòJÛÕgÏêa!fžôÃüÄñÓüÌD„¦§“Ô™LȲµ $@{h…Q¼`{nôv6<ÀözQb€ˆAd]Lì±*[Û´øZOOü²#ãº÷un‡‚,uS,E~A(¯í—,;‘[çʺ6¨B:»–ˆŒ,¥@>¸Å±[ǺE…Ä&ge{O £°QžeÐNåg„7¾\\Ø´ñG•}…^1Ñ¡óù@eã*}I!™Õ#7@níý²Ä¸Úaª¢w[tìlGØõ`¯÷Ì+qиŽ?ýNDTÜë °¾· xÊÙIîÕf¬€ˆ,ŒÓ¦{Œü¾ƪûÔ(ã;~Öí®9Í—²d€ÒEÊâEðìŽÿB7ojE9„ö€Ä\N¢ø²Ï`MüÙJÈZJŠéŒqÃÅç‹\Eé„œÔÂ.Ún "¥p§Z+DÈ8tœäž2hþÓÔJ»ÃH…¯–FHÌš;EZ‘šüJáíÄcl ­²AÞ¨ç•lñÜ3Ú/“’bçÁ68Œ@ãœp ´ù@¬sH5!_ á4œ l?1ÈüT—ÁÓŸça͸‹AÛ(d¬„ŸF1öêcÝ””®eN›Níþà@®k;@2ÇØx^ܨª=bHvËù]0õNb©VŠlM+_¥²…i-U<_ÂE8hï. ¸]”oX=¨àYµôq.F›C“3tðWäÕ‘}a±b°n\fåCãAË„¦žóÓÜ„vWz®À¯ƒNp¬ÑPyOsßn› G©Š½‘WŸÇ BÁ~û诸¨4÷Ý ò‚^c|B6³u±ÉÄÙÊ—‰B/Ÿ àÊfÈ3‘dâ ²fϯ5´ö¹D&¹ˆ¥\*ÀÝ—\YzžšIÏCºp ´Ç;X[À º癜Õõ=‰kÁ7:»d CÔX(âèZ‰b»ÂÙ„ˆÓ"›@ÅXÝûôpNx\AøÉ*èðê2ïË÷lƒ¨"‡dXù-ý­p(¹‚\G†wµ9ƒÊSEëÄ Ï ´—‡”(Ã.á‚/h3嬃2Eµ {Ó„ïü¸ b¤Ô@(¿†˜ù”qP©¯ðaêÔ–#ú8cÔw™¸š§gè+w—ÌÈ×®tl‘®Â“kßb‚¥Œl…ȽÒ­c[»ø¬Ý¥µæµìXÿð~S@mÎÛ¬VÂ`Ì×ä¶"Ä%CøÂ÷}Ýwt߇ìÁ <¥RpÎ_'×"6iþÕjÂÃ&0†DÑk®”¢[æîbý…ë™ã9»åg18 Îû 6ù5Ç!|q ¥!7*2•“„ Iðß#:ô<ÝBB^-ºìý+hN«t˜V†r\n¡Íàg.¯Sý Á,Qh¬/æSvo†úÏ`^gfLIÊk^vú­î‹¥jKUÿ,Yìqã`##½°·ä:<ÏH3-'l. jÞ ¼4½~+Eï+m"0]NàÉÑ#œÄ8¸NÓòy Þ”=fíÏè‚3Q“z¨Üš89CÓÖØ›20ÌZ;;”J¶I:M™ƒ±„¾Ï>Gœ)þúDÓ}¿Iþâ´ó·³BˆsµÍq"`Á.`MO/¬ùšoAó$úG9–¶åb2,ñ‘î߅ŘøÕ¤g3Ыà-Ƽ„1íèílxc®õÓ4•óä %ÂTÇÞ*Ëæ±|ž)­äSµw¡³¼Á,sÉ]æÁbà÷ýS[ëܬlþð ðO‚&ÒíÝõÇd³‡nÈ–ßå›Gt„ÓCn6?¿ú/ cî¯r T‚$ð½Ð —%ñJ(ÇE×I–Á)ÇyRøè¥/Èj[Kô úJþ^‰ŸÊž÷] ù.¬ïBWµ¯GžRîF[`Ä7c¬ÛÇÞK~÷WÇÜë/åÈñWlÏv9[TF}I·çEĶNÙf‚±iùC>Œ ‰»àT€à!æ MÉ?ôq?¯QÂF b¾‡$ÁB€˜$+ Ð"“¯CêÍ܃)â©®}Ë$c…žw˜îÖçÔ@®\·eÿÄݗ܉ p f¬H™•ý኎“$·Rp¯„wà.QåYàO<[; Ëâ¡œý*†•qzûbʃ {3 _ ;%"Ä”/,¥“a°ùÒÍÂ\> stream xÚÍZQs9~ϯÐãÞCdI-µ¤+j«`¹\mݲ¤ [ÜÇÃ`OÀuŽ‡²`ÿý}­ñ'žã1ä(*–eÔÓýu÷×=c­‰Ê(‹Êz¯l åRTÁ’"çñ1~³&)oœ¬”YÆJ¯XÆÎ*ŸAqJ'X@*º,+³J¶ XeJ¸$(쓱Vf«@„Ë‹ß‘‘ÃJr&·Á@¼€óØ…“@˜a1þÄv=6‹¹Œ°&ÓîíL‘ÛEål–0 ü¼Å€q>(çq_X”0òÂæ®ÆË7‰‘ƒNä¼€e >c±¾!9‡›Kùû¢®fe¶ZUQ?ÿ|‚ÿ‡É#.ŤS’HѼˆæ5¬xqžª×c¾P£þëß0&eÔüz6{3xp§3PscáY3_©Ïà©]Æõ>‰×_uøÂz ønQ‚3´W`ÛÑù¢¿¬¡,5:z¦Fõç•zsSççÕ»údô N¯ç«¥DÒ ×‹š—Íõb\—9jçžÕ“iõ¤ù܃ú˜ô}^-pµ„þöâí›ï„»WC‡,BòÑ⇫ñêr¦G]éÛÔÕ‰ƒKyÞîÎ;ílÜC‚·ÏQGDH÷@ïuWØiË/^M—“æêôeõq:'ç³j¾Üñ»øY„Œ©A)¾«ƒŒÎ%l·LCÝ<2\2Ýâ–×_$“8;1~1Î @LŠ˜,¹Öê Å:Ä&DN'îì¿5ÕD0ñ¶(ÆÈ.íaò[‹À€NÚÃæw¹ð¶ õx³àÍ®Lµ,s=l-p‚0â 38i°΂OÏ:Ùøí6;0>Mp rV;è"ù´½§_,=>4˺È0™.W‹éÛëÕ´¹SuDôÜ1ë(ü6kP+†Â>÷Ó€ëÙjzºz¿¨«I+ÍÙÙÅ«]ï³n-bOHGéë )’Ž6~+ÒnÞ7ù=ÁÝFÇà&»Æ4¹{0Ýo9Ë’¼QDDzDhTRÚÁý}7bÑ ¨/¶üãůퟟޯVþ:}úôI_}˜ž^6‹ë+Ý,Þ&Íx9’)gN߯®f£y3©Q)h|8ˆ¬¥>âá{¨ã؀ȵTÔöSOAXK6/k(s\ïÀÂp»µÈ±ÕR;u!ª °dûÀÕ÷`ÏÁ­±çcÂÔ r÷“éP½|½¹’âÓjob©‚8;*¾×öç‹z<]"Ð Ž·¨µ RÞ!Ö¡˜e8Šéïy*jRT•:Á f”4s?‘ü¥¹z;¯Yã³ó_壚—H{Q¢îÀˆº£SÌŠmƒ•` €@äÖõŠôÇò.{Àñ.h)øªaiBx!ÄHF›Ô{üïõ§Ój±¨þþö¹_¯J%zv=!Z4IgÀŒQá´´âàšS?>žÕWÍ¢Èóx6kÆÕj8T`NY3¨¹¤HÃÁAáÞ°¸Ɇ#ì–~Ÿðxk;§mox¼µPš3œSE·ƒ9<<úMx좢σëJ†ÎD /’©l‹È(+9õ»ÞÚ¶Ò^)Ÿ[Ìo ÞBÖˆRÒ‘!XºÅËß!Ë“j9o…£jòTx—’Ê÷) ³¸¬ÆõPç ˆ˜(§¨ºÔ”¼&àȪ• ”ƒÝÃ1¾¶è«uõ6¸³ÈÞ®¥· n‹cî¾àêdzJîñ”x¸§„ÎS‚?BMÁhFMF0°tK ,-ÜQ“=Ÿ×§“éU=—LÝö!Y…¨‰ä”%¶$orÔe¯Mqä&[ñ¸þ\]}˜írÕöÁÜÍE õ‡¥ w‘ÃarºÚ>tµ=wÌ“;æ»w4vvÝ C ûnÍ`šJdPû/<Õ …K÷rUP7üs×%?Ž ä &Lû¸€QŸGël¦.`ƒ–ÇsT)Is :0ýX9å§ðwÊÑæ\âQ5Ú;hãã­ôtTˈ?æ¬ ˜ §ÅüÉ:‡øcå dØòˆS»ÄE!„"Çæ~¾(ì}^/ užUïµôD,Ê›'¤ •q¸°epC±½iä ©)²çÈky !c gy'½ åJ_1üØŽ·–G÷Hœ±ˆ‡`uþž†Þ-ÕtNüý=‚W#›úÒµ.o dI¶ßIÂ’R™ qG£kG/xb‹h\ì’hÜ$Ñ.=';¸ò«ÚîáTŠ02¾:¯³íçRbËÓI—EKKÒo@€”‚ú'ÁÊ[’‚‡ãþÀ+Ž{Ù,®†ú©A°„f¥ý/ÏJô¦»=Ÿ’ié‡-ÏŠ¡/{9Ž£Æô°Žz ïLi×;ïé™v½³cȹãéc¿©sÜÜãÔypê<8oÛ똿4/ ½¯qñýß^a)<:àƒX2=Œ8–Ø:¼Ù0Êð’tÈòr¸™£»#ܪ9í¤ØDùòé|h§×¸L%ÂÜ#ÏóÕû– õúñx¼Ýr> stream xÚ¥XmoÛ8þÞ_áû²•qW”¨·÷¡/ÉmÉ5Ⱥèiah%:VaK†$7õþú›á2¥ÈI¶‡ÀE‡óúÌP|æÁŸ¥Þ,–Šd–o_yj¶¹ŸÑàö߯¸¦sе(ß.^ýz†3î±ÔKùl±²Y-ŠÙóní:ÙÌÝ œè|î :ï˶kÊ?÷,Ü­ÜÖÍÖ//ŸiôPvk çúfÎs—GQì9q8ÿºøÏ«‹E/Sèû/)Ÿ‘žs΢Y”ÆA ìÛ²ºŸ»~)ñ^·8l³©a”8yÖ•uEM½ïÊJ¶ô–iÒB¶9h«¨eAke5Ö#1K¿?ua!¬1/…ÜWJ glíÌ~×b0¡áø¥ãï2?*³ä¾{cëhŒmÊûʨÎsQ‘™ëóñx0[òk⣶êy9O˜íš¬§~=ƒY8g—ÝKZOü±U|?fqÏÀ÷,MÄ36Ô®E>aƒ1S%,›»aê9_¼Ð{W—U^²šû±Ó§óD8JZátë²%i³M[Óh×ÔHúöèµn­UB·ÇŽl;MIQЬêf›UyO“2$ç8&Ö†ä$Üý>C›U'e¡|Yç†4Qd]Fi£Ü´•´‡VÙÿz Û.÷YÆì‚9²_<ñ(9ˆÖ#ª·2Ïö­Ä„0hЖ¡ „ §ã”2®o>Ð ¥9N•Z§f•å†U£…Ì7ðRÐH&)D³³ L»®)ÊÕjÙM‰F,ˆ…!5Ò½;#þt`½'QÀâ˜}Юëý¥‰|ŸŽða¦?¢“÷²QµüðßÅÍâv¹8zbŒ<KŽÖUâE*c„É5ÜZ)¡G„€Î" íø'ÐCïw-Sè1:fˆ(øܹÓ^Ž†i¯Ö:ï:t5~ˆ„…(€…\ÉFB–(¤Àx'¤€ðN¼±xš²D³ˆ ùÏ!…¡v-ò©z0bªó#äl:›<%iò3éôMðbÙ ”AU]8L¶M½Âg™(*i¼É*‰{MÑÊ­†‹2oŸÐy˜+k«õ{YÉF›zÅ”ãN^o0 šÒeO(Ä#yÃ|¨äƒ›5MFØåŠØwä™CQÇ—ÀYíäsä¬Âƒ;­^¨"'Ô€Á¤Wbz•ß•œ­”úI RÅ0ʳVÏ>¬e£‡Ü‰L'4“¢“œFC ÖA¤ä ©º-µÙVÒé;‡–Þ•:8xX—9.¬[ÞkD.Èô¾¼‘Y‡ð†+'ÒZD€(aÌ™à?­Í~×b0æãc†iÂÆLhh …qvžÞ݆Qê¨@úAa Î#ÃX1æ­ÔO ÔTfÇ©ÉlàðG=@‚tñ,Ä££g{MíZäS=Àˆ©•Ù°3p~«ÑWèO‘5qª#õèÖYgi“ÏaBÕ:œU8Ú&[z<`ÍQ±‡ÄQ–Ç8ñtA <©œÖç²8Ÿª>ÆqÚ#ÊjÕ=,uF.‹ÕdML…j†õƒ4õVNV[0/Jü·»ò¹38`g(ú‹Väž#»\‚U2v¬¸AVÜDࣄà0NÅ, 8K#ÿo'„ÙïZ ¦ptÌ8!à*œ±Ôa1ºûÔö-9˜íãÂ%J7œÃÕÄVétxêg3Õ-®.µš…/˜ðõš®;i!àc¶Ým¤rw[q^u¾ðÌè1¨ë8‘Óu¡ÐämUît‘ïh¦«éi·Á–Åñf¨+l0PˇlMúëjª §Ìúhûâyþ#úÇT¹žŒÛ„‰€H °È’òýå„-t|b¶eåäç$Ní DC¬Ö¥jŽS(í~Krb•, ŽoßM±†™l•¢8ÐKY•] Íý_Ô'¦«NèÊ<Õ#Cþñ£Í€~‰L&[d¯§Ìªâ´HpõÑÐòøC Øî5ŽuÊ>Èä÷ÆU&J*§DÄ+M{²n6Ãàå³Ôhõ=E¡¹ÌcPCŸ‰ÍySV-Tlõz~®§Ûz™/ÿ,+uÉšhàf¸Ùc€#õk7pA^¶ò‚×Óô'.ZŽ\ì¶R}!±å¸¢Ç¿èÁðþ¿X_?ÉZ‹Ýv’Weãliô)¹ºsN°ÐC\ÞüôöêbùîãõÍÕŶ5ª1Yâ¾ó³sãß—ê>r§>å›~âW›| ¿l£gÔxy=xû¶¬W«VvæBÎDÇû£§ÑòS¡éûªéבy¯0Bºÿ†²¡ P ô×t¼/™Ê65°¾ºè¢jê¶/ù%´ém_‰ÓŸ´2p wžH ÒŲ¤e’A`ô™H6AA—~’c]é'" 8özùùãíÕ{=ù‹2ž¦Þ¨¿yÊä½»ÝÉ­lÄ$¬€³”:@9*€—«¥ŽBܪŽ8;Æ¡Þ]i¹¿eüÉРÎÐ/ˆþ–˜ EÐ"Xa'ݨŽ4kt¬`!R;BÄé9šã%¡A"MÅ®`¿w""¦MøDœðÈcA4ª"øayùñöó›[C­¦®/ÞüþéöÂr\[½ÀZÉþ#‹©¥ã¯gªO™¶Þê5Ó”O6Þ!t¸q_7·‡¥¡Fã”gß,Q5©mæ¢Ö 5°5™Éöî.'w_÷…£}c{ˆ<¢§æñH˜ÿ|:}Gê[7"œŠ‰z˜>ƒ«ù)"å[þbß"ìM’a‡Úê^SÃ)ÝÂG'9h¿c‹¶E•zÝв-YœJŸqíëÑeê–„ç[t”-6`ÙJx‚@]ST*†°nz몑²‡:³¯$ÿ‚ endstream endobj 1223 0 obj << /Length 1704 /Filter /FlateDecode >> stream xÚX[s›F~÷¯à©EiÃÞ`q§‰cwÒ‘ãÔQ&í$²ÉH ²ëüúž³»¬ÀÁNÒÑÃ^8{ö\¾sYQ/€õâÀ‹8'±P^¶= ônsí™ÉåïGÔÒ-€p1 |±:zv&¥G1õVë!«Uî}ð£pöiõ‡Ç#F .z÷ìlõ~¶àœûœÀIŽNWîÉØw „”ß–(T‚P.Ìݯ뮀»åw7igfw7Eeg3¦|û=K7›"Gá€g8äI™ AäÔY¯»»d»+“M g’¶üR$¬?(ÇcÂeÜL«Iö ¦ükî»MZ%ùº{Œ7`.ûcwewÓëYh',x̱”* Y^n‹ª-ëªbIý²2cSÜjS4-˜@ïÔM^4sœK¿-«¬0Û©‘%ä#Y¢˜D‚÷²,-9†Ä~ 6Ɉ’€G=Õù¤Þ1Ì™ôl¦8ˆÚ5©V$ðÓ¦Ig,òïͲ´zÞ–Ån÷Ú]¡¶÷fná‰ÔÖ'V×ö  üF•´æ¥#‘ǪSA„”ß¡; £6| »$"bNZUTl@f ¢®n´Æaìo‹T;9tȇMĆžÎꥎƒÔžÌ˶kÊ«}§5~µø8ð™B2D^ÆOúâLáЈ8‹T/žò7õlÁtHZ(sIBÊǶÞip>0€© ¬qàe†âÞ˜%šÇI?Æá,|ºQ±ø–û‰>ÞÀ…_[ùŸQÈJ€®´•AÔ‚þ œC ©7 øLkU™B@Ø*ADü­w÷†d‹ÇömgïQ£{¸ R9õªº{$êæÀ¹o kÑj œ‘@8“BZCtÅ[¡ÐŽmÛ ¥í¬ãQq4Jab# 4_˜E¾fn¶ª)Ï„!áÑÓV„ ú-C.èz8zÚÃ2PÞ€lÑsTNø7:úLò²1ø+_ÂoTw.~`›ÙòW7­TÆl9ò“©iSÒ†’(îÔÏêínSüknÓÛk ®tLB™[ÿ‰„’wž"þ¶GµDÏÀ,®›do d±I´Q’àPd—¦Óøá›=i~©ˆ¢{`mÁ˜*¼¬nfÔÏÝÕµ‘§)Ú¢¹Õ®0í.Íìt][âÒà·h¶E^jhjʮصÆ&4D¹ºlm‚©‹G}†Æ Äs L·™BvWP©Ê ¼Ô£ƒÇÅ…9ÑGŽN‡†—Îû0³VÍR¹¸sòsk&ÃJ˲Êˬpß,õ—¢©W)4ú^úXF‘jŒÿ9©×붘L+X/Ô  ]ï]økwdXÖõ¤®E¨*-ÑæÞ|=€Æ¸ÐØ׺„ fJ9~ìùÕÊQ¨ƒ×aÔÆGrj”Ln‚»@ø<¥ð¡õzP¯À´1ˆNÌÐM ×5¦ íF·0°ÐuˆR3@V„W]›&=]iÏ PˆÆú!&”Ù¶üÖÅ»=€aš›Å!}D .·»{ÝâìcOÇÎgeÝ 2ØC%ç§Ïß¾»<±ÌÖ›WÉêòùë·o.Þž¾L.Þ­4š'lÆЮƒLez=7‰N»Eª÷ÝÂ敨ƒye·G¸+i‚I…6˜`ÓYê¨Õ©¿É ¥C~N‡l“áÀ0. ŒBzãâAÃóU÷'£ïéþ˜Rÿ¯X>R]à6Ý?ôù{¾Ö¶Z˜ªrk󖰜ꘂ¬îtjÓí$T ñ)hÄØKYžvé”ÄŸ+lXŽlh3š›þ1Kã_ŒŸ1"\d,à•ˆ¹¹ ®*7ßE—÷z‘Õû%Ë‹lcÒ FÃOi O1ÅÆ1g+$J|’¼¼x÷byšœ\œ¿YžþeÅ‘Lú»º¬t…ÃÅñ±;´<ž÷áâR«µ"!d´m/«Y²N,sä¤>°·óçs“¥—Ÿ Lð"aRXí⾄ͤÏ9…žâÐ%Þ“­¾©Ý“ñÎ$#mõPB::d{Ü1 •N?Ü ¾êÒ²*,¹{ÍÁ> stream xÚ•ioÛ8ö{E¾ ÄQuØizlM[´Ì.:E Xt¬­lytÄñþú}—dÉV/©ÇGòÝ­.<øS‰wií&A|±Ü<óZ=\ðäã›gJðæ€8`¾X<ûýµ1Ês/Q‹Õð¨EvñÙ¹^§»ÆV³¹ÖÚ‰žÏæA`œë´(òí_¿^üɳUUnp8·3唳¹;™­¶ {=‹§¬š*€Šü0r¢höeñ¯g¯=…Æ÷’Ä<ã%^"×W&^Â8p•˜—h67¡? >2õ0êav ¤k¿#‘fIO:|*3ªöÝ¢k€QÝ3ª#Ïñ=OãL9u“n³´ÊòÿÙŒ!û)Ňš!«²âÉÉ9ÊY¢H# ‹?›RÀÀÏ®y(òû*­r[_Âwì9p§ìXç5“[Êœµ—K[ÙÀ r¤´«ŒbŽêv·#m5`^Óå•] 鬋´ÉË-Ê•,¬-OXÖ8»æáêà Oò-J"¼£¥¸³yÊÊÍ.­Pft`yrpaÒåaDù”*²8åB[­Ò¥eÐ_žñjkO-Ò‡ã¼8º}ãJ‘UºÖ\/Ê'óC†vÚퟘð»ÓkNœè‹yø|4_ø󅙘/ÌÞ²4xý(€/` ÒÎ.}Žã3÷ ß5±1ûmì°ÀÚé¡ÄH[ Ad•¨;Ð¥84¾ÎÆ´ñHmÑ@m*ÿò”¶Ú¯QÎ&%Åx-ÍgÆ8 k€oò€ØzG®d—Šœ¢õÀG,ÁÖòÂΛ|#W4xéAvÌ—øµ†pæWV„é|ÌàAùÃ,uÂ6vSVtKÀˆ.µ$;7žq®j^E×ÁQfû*oà2¢?é2b‡ÄuŽ´ÅÖ–§Œ> ­ªô>/ò¦ì¶iÓVVî/ØÔÞ´Ï3\eÓ 5hgMë‘„øʹ'Ž¬Ýòg¾ÙvcÅÕ2ÊâÖ¦ŠçÀp²I1D;ÿEfð[ÌôÃ.Œ0R耠&T€#/rÞ¼ûãrJ7BKqÉ9éæÅí%ç®÷Uº,ìïŸÊ"g-À3Žyk oª²ÝÉæwW¸ü-Ùí/K\‰¸¤·ŠÂ~ ég}2<N‚,‚Ø™AïÀ QtÜÕÖ™hg-c&òÉṤÝÑe6¯ãò(ò# >Hxr§,ãŸ%®R6¶(•GšU¨jC”­Ð©ÐVû +»çIÝÞ7…¬‰sæ½rC~AÀ膮“Àà{§AØçuVnè#X{³´Iy×(DÕS ²i¡ ƒÛê7X,‹²îJ¼´i,ÛFE LTð`a¸:})xüØ5¡XBä*®¬Þ³ä v‚ r TO̸ŠÖ3dZõö1]XC)_jQNv~à‘¥Wd«“¨©&Á5ÈAÑS^Â0†cé¸ÕªÙkwµÙÄÕêÆ°ÃìÜÒs²™bûÀ[fÅj^ÙRÃE.Ü@Ð)\Ñè<A)v#ì\ó7ü˜¹(E*º1d¤[û´´;Ô®2’”x#Bn±ÊBHwƒAnË*—1†-j»G[GKS‰³_ç»xË’4x²«»±EIdü•oylÖ,*œ×éFfìÿhìç%­'•¦Ý(IÆ:[Oi,ÀJ_uˆ˜XPÕJ]¼†ÌEvuV±{n çÜH±~, ¢óüC5|{/‚ä {‰ °t |Ñ œ*{ö¥¾¸¦±­Ÿ šÔj"7Ñc¥®Úí#@—ŒŽ‡Cî©×e[HNe<ä5ÆŠfý|J”çúÐ’Ðç×Tù¶…Óçóç®Ë»åÝ}¾Í0ÍŸk ²LÑbK‚È¿õ>õÛ”ŠæSpš "ÓE_]6!ð8Dh»¨¿mNòaÅѾ®Ùš…CÉ ›XË“§”G‰ê}#½m[ÈUhm{4ø\B<)U‡Ð »Ãr–Ig/x©ý»Í¡Z !"H Ÿ€p-È<¹³µp cç?d«-c‰§‚XŠºdЮm:Ï´ ‘æ]¸ÝΤ²¢þ ¬¼÷fiÌè0Þ7bùjÂ0ö MGeWœb‚o5Z*O7ñ…b×Óæ—­nÿ|pÀD7rz ÑúÉ.¹o¥W7bB?¿3ÃðöÀ°t{úê1ÆqUi±=6§lê$q½PØüv‹Õaÿ€©ÓCûËKô䘡ãQùJ5×5~âH£/Š¿äå$é0ضº•šI ô ¡û G’Ðü„sÝbÐIC3cŒ;?9-éä‚‹]œÛ§”V'ói!—‡§kþ,±ˆã)äsYÛähì¥X¦ÔéÛ{rÄ°x„¢58–∤`§óÓ1·9ì0óx×wQK㈻+Nêm‰ؘö‰w¾|ÿÇ‹·¯î®ßß~xûêßÒMãî ÚVˆâåñ â .•çyç·ä[Ù¹lBˆD­ÿƒŒìw½ËVÍŸññžwIwôgÂiò:¾{ýþãŸW_^ÒÇ«O‹›Û«Å+"f‚A×,ÍÙ²zì“]¶E2$éòœ¥³;~å ÈAMUˆÝîŽîÀŸÊiWò J%0ØtÂ]»~¬ÈÁGÄÖ(ÈÂø•*s»Z FCÍ$U¥…ÁØM#d]PÑÍ ê`€xò:Ð!)½æ)¶9æG{û> stream xÚZ_sÔ8çSÌNÖY²dÉûlØãjY8µ·•ÛÊ™'ñáÛdk?üu«e[öh˜ •ªŒ-·¤V«ÿüº%¾Jà¯òd¥Ó”åÒ¬ÖÛG‰mmoVôðö§GÜÑÅ@{”Ï.ýý…R+ž°<ÉùêâÚêb³ºŒ´9ûý⟫T –%Œb[_¼¸øõ,NÓ4Jü!É£ó‹q%Ä_dHÂÔ~rf¸´´LÁ'–äð ÛiÞâ÷朳lgF2žÊ#+"¢Æ²ÿ®\÷U³;‹¥Î`ttùj_÷U¼©¶å®ƒÏEMOÛ¶8&º'²g&šv[ô¿?9‹Ñ]qS)WKQ¤F0¡ùÊãñâ¨O,i9¨]ÔÅÙYl´Œ.n«Ž‚_`ú¿¿R îa/÷»ºúX:²ù40ºV][’ê3Á#nŠõ½ÕŽ8Í–)Ø ž2®œŽœú¶A§ ªJ±/Ûëb]Rðœte¹šày¾Ê40c®CÿØ Æå4sÝ@þ ãôpéVKÇŸ=_KÛì{hv²éP9ÒD;å°ƒéå:¥ÕLÔlß°G}bUËAGåp }¹­ÖÈü­c»üŒÛR¶]¹q›™‚Tq»™ÜÛ}4øÑ<:|—Ñ5Å~¸Çaš=¨ ”*z7âps¥f\I¦43iúðÍuýco€Ðæ.¦Yl.0¬Ñðq —oGT»úXLF$Óªí¾ÛWÇì+ا6b¶Äoì«£>± å vAŒì6“>%¸‘H4ì`ÆR™ *ÀuÈÊ9xv5Zù{Lœ¶oª]ÑÞSØ€-ÆŠ˜ÞzrP‰ˆ¾X­i?v V)’èþ'’í4²ˆî«²ÞP{·ÿpcŸ;;ÁçOš›ÌüÏ% E_½#+Í#ê\®‹}‡-s–fz•qŸɾÃ}RÿØ è>çÓºÏttŸïH 'y1Yÿéšv›a=<íQô_†%}šižC¬Sv6)8òtù{²ÚÀG“zõÅRnWÐu OõêÝ£-)Æ´˜3¾-êš$¼^ºnˆëRdñeü wÔ'$ºÔë¼Ð³²‡3„\ÃékM@犾TuO`¼×}¹£FgôÒ|è‹Š´Ø~rÚÏN*:Z[C³6g}¤·¿¥§ÿ:Ç’Íô#G•‹Á³\_÷_®p¸uÈ Sz¤} VyÚt Ü2­ãà×ÁЇ‡*÷Ð?ömÅbš…rÚ5Sn'€wÍçàçòé¨M ³1¨Œ¾^¶å¶i]x©\Hò1(ÔH/Ôq€/!Ú™®”1Ìu _:êØ#áËÅ ß5B Ødó=¡æ]µ­ PÐʬ焵Ž8£.-¼<‹m $ê$F)‡Â ÿ—˜ Pžƒš ¦ÒüÁZ3ô½B‚[Ls: dK:¦êük¹ÞO$¤ Ò·  VÕDb˜B6fFOêÁað$ÂiØ¥@Í©ll Ž=òPFµt†0¹Æ˜îâd­÷m[ÒnÕ÷ÔÔ–ëf»µ{‡Z¾ e(<å,1ƒ¦`° ©S€hÌI÷¹„üÈ\˜™*­ÄƒIì—r ¶pÄ2]€f¢®rigÆoèÍá0è¦-Ý6ͤYøæÎ9çunàôQ£ëýnø:¤À:8ÇÍ.äjs”P.hÔÐã˜mHÅR WÒ!@¶ ×?ö:•ù4sÛÀÅK–჉.Aº¿Xù{iÎÝÎÙë@€±{š2ŒxŽ¸0»È3g)°;Qzh){¥X)$1ú”¥8êØ#YÊbÐY@yô†Ðnùgk5­cÕeSÄmwÛìÿâs9 ^¬îáÃ]®ŸšÝô)d ¤?%ÿ]±-§„ç¾ )Q&Y2DÑ-fnve7ÄAÐÏ,k=¥‡}°$—)˵Y¸£!j Ég“ 5¬,‡ Ô§põBÍ8ÖdT&¢s4ò¯}¹ÛØÈ«•«Ãh}ÂOûbƒ1eo­?ƒã^W…íÂü¾8ËÇä, R¬ki– §/¯±>œº$)ed+ÐÊwWÕÈ‹m ;;‘²LŽn”¶„œèþCtx™²¹ŠoènÑVlvjoe8÷œ,ÅØaš+æR©×=žúkS³I,ÓÏo‹;¬ûX^yBL^¾ÜAäªëŠ\ÜÍà›mªI¢çû®o¶Õ¶ƒE 9P|ÎN'Б&™±§MZä§P‚£Ž=òJX :Ë¡ OSÖ$~Á*‰Š šµ{±µø]Tçjº:‡\³ÃHsD@z¤³ÜõUÐ#aDuˆzl¼†©Ž(Kž¬Ò,gJ?\QðÖë>¤–癲(N'š9àÉA±é¨RŽG•y$¨0…lQÍøþ*$ê¬.÷›˜%¨‡üÌ ¡º&§-µ:/ÑÒ·nG°:vÔt4ï=œ3Ÿ_ýüú—Ÿ\ˆ½c‹´ñÇ×ïŸý|~õüõ«7?Ÿÿ;¤xÄÂÇXa«Ö÷w£Š¦èÏ\ŸÑ,®C i€’Î8Ÿx²Ñl]¹yqt*ÆUwD‰› +ÂCg6˜Ä$Ž° r!„Ókå"±çÍ®ßwØqDÒR^óŒ ¢Âæ7åÅKNW f^Œ°(•IrI°ºò¦-o(yÄwhG6Ýa÷r"¡£[W±¦3JûÅ¡òàšrˆ¡z†ôë£PVÊ~B;䥿8Õ”Þ#Czo¹#¨¨‰Á8Sù¢úgð2Q±ÙTC(8à(ó/o€t‚Ø©ü®\F%~rѸ´bÅâîpîMW¹ÀÓ˜ M¼m—ìÝ? *Xy–úŽ,C%Ía-ýÏa¬$6M™Y)X(ºO˜§o«]‹µ¯?üàš»æj}¹ÉÆEšÅQäˆÿøñ(ÛÇ¡ƒ¿m=/—  ‚ %çìþŠÞ=‰J8Æ'ßURËÞv«!^*ÅŠjS´›pÄÛ£½ºº®›¢ç‘‡ð6ø'Ä“&cGâÜ × M^ê»!¾\ÊN„ÍÍ\l0æ¨ERˆh<Чhhi®ƒ ž*Žàf<œ­Í¸haÀTÄô`!–<Êdü¡r |5=Ü‚,m…Ï:-l9V‡…¡ÀäE*(èÃë°®ì ¼<4Ÿf1? 1áw1í q©´ ¹›ÌSóC„©ÆCT•†&€ÝÌd _ìñSˆÃQÇyq,!L¥'µOµGѶC±Ùѵ‡0d„*&$X5n&ÏŽå²7²l´'½ƒè™£za¬¯Ë¯ß7Ë ÈfAôÑÑ’ÀÞ]@E¨mLè \5°<÷ lÉâx¸>ÈõñLOâ­2ãçíRDöÖX’8ûGˆ2ºtÅT_"’°çÇ¢þìùÓÕéªÁcZ•wº‚¯ÖðlÀ?Lî`1µGÇ©²ëOýCÉÝ|’E%°;ãòÍ^ÎöPºñýMœ =dƒ¡Auíy#çÒúoD'7u쑇7ƒzй Ò|ž $@ŠÆ ¯@kÚlHÈ’|B>C!^ÔX–•Ìœ;8ç¤ã#˜ÝjÁ-êžE^H>å—y2Dá‰grz?ÏKÃ#îì Ž`+¬Œgß 9·§KGîê“ Ž6ÀÙ£ ýf†=Q pùqñZëj÷‘ÚŽN¹6ǹãšäO*®Ñ€·…½ñ‘Jxß3…4aFï´ˆ0½LH@®0ôs6¯$’…«aÊ6ˆÙ^ÖòÕ[k¿ŸÏR1Ý×ÃÚ¦Ü àÕ»¥x¤&q_¸tô)íù¶ÚÙÃ*TÎÝ®!‡©‚»Hcµ‚~±Ð@3Ë;LJ'wnm³{vôÛßýÔí±k¥3Nm™Gƒñ¯XÚÏe)nJÔp[xÕwG¿t2ÏéÐï¤/”ûЉ«Î¬C$¬õ%ÞB¶¤Tô?”~kÚMÙº¸)ŒA ªÝÈ^®ŽKÀ&o øê/>— îžB—öš®²Ùš­ûϹ7÷ö¤@%ë¦ÞowÄ|昇O#óŒ Làìi–µ‡ß1ºÇSÿØšO²8óFíís|Ý>Çvïö9¾ÎnŸ“S⎑Ã;!G¯®q¡Ybò•(“œºœ:Çu(̇œÀx=!ÀßÿL¤pWPÃàÕìÿtWÞ¹{¨hx/°lK°Côæx˜ložáÇ]C¿¥¥Y÷ô6 Šñ>zúáJ'ÖÒèbåЫ¾„õ>Þw endstream endobj 1260 0 obj << /Length 2793 /Filter /FlateDecode >> stream xÚÅYësÛ¸ÿž¿ÂÓ/GßI8_é܇œã\Ó±}7‰:שãq ²8G‰*I%ñß}€IC–ÜN§ãÄc]ìã·+yæßüòJÚuSX8í­üyöêÇwax&}‘ú©<›-û¤f‹³[ïb•mSOƒ ðâ×çS­Cï"+Š|óÀƒïÞÍ~çÞ²*×ØÓÞõ¹ôÊó©J¼…©6<öî<Ñ^Y5U2VQìÅéùÝ쯯.gÝ C¥NdWç%J´f^æå¦6ÿÚ™ÍÜÒ÷Ê%·Í*¯¹G­ŸÂHÖL «RïëÊlÌ—s{(œ|ÄrÇ;æ•ÉÃÈŽ±H°·-Ú±eYq'ãµk$²+š|ºÈ×fSçå&+PÀÞTB†|i”W ›×2žÐá(Ó5~ìꆧ²â+Ípð±FZ ‘¤/„"Jc8‚hWç¡ôÌSÕÀ@&À6u€;š‚Çãg†)”Uo ¼lloÞ6¿L”TÔ&ˆ¹i á«xL…TVË@7<“Ø0ß²õ¶0–_|´|Á¨íé0ܬ*c¢£áO~èó]¢ §&:mïre×èºÀŠ¤]ñÉ÷•‹wÛU×.:Tn/ãÓÝ8ˆ0•{B¡´ï\UíO¥÷Únú[§*1Ã,æ´5ì7ÒÅýÛ_ÿöóÕåýů׿]]þ©ƒÄCzPqÝÕäzrÃGãÜë×Ü滸Ü5®«ƒŠ êá+Ïè•™6 â†_öí»™õ"­À[“^ÂdŸcœÚÕ¨—¾IçRÝÊl¹l¾ÞãQ÷‹es,\WM¥ˆ‚°Ý±·²ØZªà®XpwQw â8òË<¢”~âf|ð øjÒIäi?ЉܿûõÃïo>¼ÐÇåÇÙûë7³KzŽ£B·o>OÆwÏëÉØO#Ùï¬L·ä¯Á`œêּɿ Ü;£Ëë––}§±ù©ˆC9Òí‘‘À;ŧIrÜÚ“Å#óY›×Øy«NãPg…Ë—R$qw·›ÒzŬɬ£ª $€<Á•–uÞ€IñÔÆã%k½¢¯E,ÕÀó—8?§»Êuc¥Eº3>n*½ _&NÛ·‚^î ¡z¯öå¦xĸä{›²É覎#PrÝsèÓ0QÞÇ|YU¤ì%š’Þ‘Óm°“øÞº¬ìЄÕ*ŸóGzŸƒX¶¥àa´t›´Ó ~‰®É²”¼íkÞ¬:·Âº[gk3v¨žêÙSÌßåìòMcLžöýÍìƒ ž:סӸ%q7€ò&°§nþN&iµ¿Áx­ýØûÌVjЊ:U^— ¤¥ ,ž·ÊšZÇ#‹ü“/S™ îjxÒ:.Ú`x¨0Ùü‘-Š`„‰9o6Õ2›Ûõ(‘Ú˜1îÔQ*àDÆ 4aOÂœÀ ŒŽôÑh»Ú#à@¤ãcHxÍܺ¸r"¤õ±¥ØðÛ‹Ó1wqÒšQ}‡zœ¤Þ6{°3IÓ‘uÛíêio¹ƒÁ1Ñ¹Ø }]ÅíW|«l×”kp7sH(- ¦8Ò\Zó(¿7tÊŠ±`linï³I‚cð•ûìlƒœEå*ê0½@ÌxÅ›`ÂÖ¸^˜}žTâ'\ãc„.€öN5ßÌ|טö`</2yZ­{QñèMv“µU }Ž¯µ­ÓiEUolóÍΤT¾Ð©3(`Ã>*PÀøÂÔó*ÿ¼/6aŽ­kÒ¾•¥ta×s¸šïÖ6'l=§¦ü’‹TÐçÂzçþr(C*Ìd΢ ‚˜¦ºâÎï ’R{#{Îù"M´ËùÆé8I­µâµV´õ‹˜dÀIó $k«+È>ÉÀÖ]¬„ªÜØ*Ä ÷J9—¡S6ee<ú¹—@EXe ¨²¶F”ÀiÈo ßT÷¶0ôºÚ! ™ F^^om>Y-­{Â~ì‰l`†XÒ›­ì¦V.°Ú]Ü ¢HD{€ò„§.ö5üɦ­¾=IÒòDÝÛ·¬eè´Ð± EÝ:¡dÀá q X âlä+@¾ˆƒS„:iB­O"?j‰«³n:aØ`á êm"á¡­"@ŸÊŽÎcɱ€+÷õÚÓøo%•Dw™5°˜7¶­¹ÍæÍÎV;à @ä‚ê¨ AN–ëô‚t¦—Õûm ÁÎíÉìuë*é‡IbãÂ9zž˜à(~Ö¶åÆVºâ…ÝLïi±ž=™€¡i7 K ªAͯå„ea bëÃ2:´ä<å®j/_€«jÑ&„ÎÈ×f]VXI‰ÚÒNè­³o9…5´¿t`PÊèÇ`\ 2¨_hЄÔA½¼zÀÛ§½ý®âÁðaí/‹ï~û¦÷º\Š¬d8¦8Jl÷d‘ox¬ëî&¶J–Ú ö‘ — !˜&j,©"¡A+e Ï„GÊ íêio¹ƒÿ1Ñ®¼ðg¼FÔÖ‡ls0ÎgÐòv‚ì Z“Õ@Ø…•Yb¹&®°õ$eEEXzœç•uq¸¥E±z‰ E»Õ³åû (íÿäs=\…ØéQ¬¼-ñ‡j Žqõë ü=—/AÊNâ &P—'ç-Ç,ÓA³Ï1Ê&grß»<Ü›ÉïÏúž›ëÁ×=õâþãû\ÞÓ¯3‡$Ô!íùýòÞJÏÁ›ö­ú~pL(î çÊûdá¶ËP:‚h—ÿ_ÈîÒ endstream endobj 1274 0 obj << /Length 3470 /Filter /FlateDecode >> stream xÚ­ZmsÛ6þž_áo•;!J¼dgîCë:¹´iâqÔ¹7ãc$ÊæTU’ªãûõ·‹I­Äsã  `±/Ï.ÈObøã'Y|b¤d™JOVw/bÛ[ßœPãòõ îè" Œ<Ê—/¾{¥õ Ygüd¹ñY-×'W‹4>ý¸üùDÁ’8.¶÷Õ«å?N#)åB2øC’çË~-Ä.)¯(ñWÄeÌ„V'Iª—Š°ßæ»S­£ŸÍ¦½¿ÆÞëõ¦½®ÅêZ®uüîå¯/ß¾$šr÷²:´î·p}þaùæ×–ç@Éí>A>\2®iÆب—S÷*ßn½Y‹ÏÅêÐÝÄ8+.d˜”0õ£ixxžà4ë¢iëêÁn²›cŽádè¦. ²ÎÛ¼òÝ+¯˜T‰@1GŠÄ¸ã$#>Î6Ò‰èÜ€ÜN3¹¨ê¶ÆÀŽöTŠÅÃÄ¢ ®ºØu±[4ÑXÃp©ãÂ)Øò©X,6Õv[Štq_în°‹/òÚ½k;¢»ªié]yS5,VÖ‘bûÞDºªêºhùn kˆø‘Ù1÷ø¯(v£ù⌞Kš£¨7ùÊ­%ß­C²ušvúžöŸ(ÿ)èº>‰DÌ´14ä÷˜›¨8ô8£»€oÈêð0Ý€‘­H0-º½M„xkÍxÊ;RÜ‹åü×©Ö ÷ÒÉ´±êbßàa¦«ÌLǦ}Ø[E;»¾X^š6ž9IX"û]!yÉ 2Uja•JªEµÏÿ<4±;¸þ/ I6Z2I8„¥‰úáZ5Le†çÊGé*oA£ñüD5h™Y€2äçÀ ïù¯«Ã§mÐàÌd";º—A^°-öǺ………8‰› œ¢,Ñ–NY¢µ§ª£ÝVhr(ìùuð Ã"&6EBk«ÐTÀ7c Wc3©‹|KªòÓûß~|{>§-iìëiX<`J:ëHƯ޾ÿa9Ç×(¦´'ö(Éâ9aÁ¥©~<ÁÛ÷ï^Y>S–r3™§]±jË¿¬ãÙ> ß3€x[uDßÞÆgz,;ÐQmœÁÝ~[|Æ9ÖHà²jeMÎG¢ã²Ô”až”}Æ2‘øü6O1|,t)f….™Q#ÖÛ§Xƒž$ƒ#šª'ÉR§ 8ú1Œ:JÓÄwÚÔVåÖdø0‰]¯Å×gï½x{þÏYuP0wú´6+&ÂSX}>:ƒ4óX̺×écÛÍê)ùѹ e÷=°~ZÙ9Ï?0¨ƒ‡[ä ýâ±àïýàBš¢˜ÂL¥1–Š•h–¥™…šLÃ;ƒ7ˆa1ç¸ÇŸÝøÈc€ÄÓiìÞ>àf*·LˆºL8S€™§/Ï«‹ºX• ÐD0ñcŸß ŠÈÒ…àÓm 0ù8–0¯`àU èŽ:òÈ›˜2íb/w "¼ë!¤‡­æð©>åˆ[ËLÐgq6F0«{Y;ûPWŸ G|_¶·ÔúOH …ɘù¤ãÑí› í.ƒh*³‘“Wú’Ð5k)?ácwâ©8@‚U5A"I¾ éu ¢‘œÀ·ûN?o–|[îþp*ú]Ùôò>¤þTx#Dr‚Öî>+ãv·£zYD[Ü¡ nŒé –0a¦œ6AVÜ…ç #‘™9ù‹„e™špݹ‚¹&ý  Š4]ü¶+?;}ÇY>ZìχS • É‚èÚÛý¤ã8ŽºÒÇ—&¿ ‚$aØ°Èr·ÚÖŽ/Ä6ˆ 5¸$õ5K²M”½ a©žÀ2ÊTέbŽHô±¥Û¼O g×t‰UCÜQql^‰Õ†~ïoËÝÒ£K¨nÐÛ×dùÐ Z>‡è#³Ì/+|‰íS¥‰Rúü‹Ï«bO.Ë0( 3‚½ä¶Sˆš@3’ë˜=cš÷*Ö[/úÒ8!£WÐH8Æ ÇðŶ«½kn+T½í:¨m³ì#ß6UP,`³ÒLÕ-ä¼2§cÛÎéX"˜~vŒ¼ b‰1¤Ë}}ä¥ñ‘Ç yÇÓŒ#/JŽ“ñ.ò¦Òˆ‘ë¿:ÿÜñ»ä¶Ïrÿ<äëú‡²Íþémé~=PˆqÛæ¢6ncˤS¡H“0)ÒÔ¥8ÕGâvGyäL™öq{&6?ÒçgÀ?©â¡úŠÇMQSÇ$äÍ‹û®·Ü…ÓTˆƒ2ι{à”öAä°kÊ›]±2@7Ĩœph,ð€%“%B« »n#˜Ûç7®´Îkì-©<Õ33+Ç]b!mAkä=Þ¼›Í=…ÀfÇK€À¬ü%7(üÉ;Æ•­ñ'¦ž­×åfsÝΔâ˜ûR+Ø[‰Õœðö‘á#¼5`‰D›¯¶ûnx4Œéüx’±ÕãÂÓ,¡æU¢¢Oe‹Éâõ¡>P÷¿\÷ñ%“ ãr/¥Ìø9AØËrxˆyú\/·€¬ÁŒÄS‹úúrôy›Äàg'%o©²¹¸-   À?d“Å_·»ñ‘Ç ”lN¦™X0¬À´µ`h^]ù6j«©k‰¢nþûŽº~I46 ë> C·ŒÝ=q0/@SþŸ¸rÔG64e:جu,*Ç¢'þÿ˜¢*¬%‰ %[|).¯yóî§/,ô€ïÖ*uQ$IûB~uh¨ÃGô=IØòP8 ÄôøÛv<ÎèSL³8Üuäòʔۘ;ÎBZu•|t£ÄlÒ aH .ÉíûïgÁˆ@ȨÇNÁJ;Œ àv My·ß>P{ d cšÛ(—ΨîÚä«]~W4=§>¯6}¶ðèÚgÆix†Ïx‡¢‡L"‚×p¤uŽÇñ@ÓË‘ŽèÆô;Ìx<„ø÷‘0_*&wæ~Ì°¥_¦cÀià(ð zÀQëVˆ Wö-¥æ„¦°Ô¢&U ´^ ÆÞ=j ^ WZVP ˆXS„£Å#ÖÝcƒŽX+Ž kúÚ9&q™‚£ g¦Õ“{p)D·Ür7_ æ~¡¤Ö÷@­ËoÝ=¨¿‘è29{Ü(Ïl¡Ç=xÿK­§´'p¼<Uú]QÂÔ ,M Ã\ŽÎe¢SÂÒ Nœ÷ úÙ÷Ôèów·‹{;†pFG9.ÜÛmuv!%‚!a' )SºŠ& ” v5«|›»>›ƒ·ûCÛг„“ئAsÑàºÇ`f³­öMP*Ùϧ­Š¨Øe™©_AÁ¬ò¦ ¤Ø7sõ‰L²XéçXŠa‰êV'%’©ëPÂf"ö¸8Õ_’à ”( ÄeDŸ/1yRhìŠûÈsZØ徘 ‡Ía·¢s·‡íVö¨d@3ð|ÆÈ3 ôëÑL7>ò„ûdš1šI„4“`C.®Þ7‹}çÝf‘ÒŠ¨ÛðG{˜88ƒ¯å£¢‚Šäÿz´Çy8ÓQÙÑ”©¯š¸ *V ¬ì±NŠMëŒÜÆÚYÞhC@˜èáºXRáÛgÅÔ5|‘}B™ñª]+å%6*ãvöK 9KÂÖʪ®îC=ÜUöÉ–æ•E1š²`7¸Û¸Vµq8 KXf¦©F´ßÚoDpa¢¥÷Í„=?X^¾ÿäp¿-ÃÉøÿᦵíÐq36ŒdRóç ‹Ó1 ”'eÆÔÌÕ§ÂT{¨w3w,føÐÀ‡œJôþvšï= “à >4 ibô+¿_ñ1·‘­Þlh:§j¢» ÁÈ„ï醌'ÕTé¼XSàöfgBR¢Y:HÙF$›]_ãÇ ÁërÀX±õ‚ÌìÑèaEßâÄ°ÿ/\~Q·Ô û¼#ºO¹ _¸þŠz{8|ÊÝG=|Ø=zrAM³·Y§î»†FŒ=¹H¥ëÔÊQŠ ÄYçõº/Ý"³º²¢®.¾„ï…>:¯®7×û ½N=SÏ†Ë 6èÇ´ JeLò.(=:Þ²·\%g±ÈfkŠ€‡ÒN„ÐqCe3¥¦_¹!ÕLìÅêÀd˜XC€UϨÒøÈc,Ž§™V”­ ZêÕë~#ô™ÁªE4"»ûT|agÔÕ¹½S]SÅk^TqÝ}=vKõø¾^Ù¯Î2ªZ½®'âh ÞÖXú%BXÂÑ!g\fSŸ6hÍlé ${µEšÍKl­0R„îøb à›ä¼v6½"~Œ„]¸„¶ûªI@‘2™ýpË<ø1`¾­nìg€ðd«ð{öMøîÐt’€o\M•ü$„JZ endstream endobj 1284 0 obj << /Length 3402 /Filter /FlateDecode >> stream xÚ¥ZëÛÆÿî¿Bí—REs|¥Ÿœ‹/q‘ºFª (œàÀ£Vw„)R]R‘¯}gvfùÐQuœÀ€¹»íÎÌÎã7ëþ‰U­R¥Â\g«òð"r«öaEƒ¿{!˜n„› å7Û¯nãx%¢0r±Úî§[mw«ÁÍcqì]o”RAúõz£uÜu]5´x{»ýö¶=àH_‹ ]odìŒmhívé µ½-`A¤2IƒL¬ÙþíÅ›íÀa,åo)ŸË’LeB„É*Ét(”&qÚO¼è4¬DDó„”:8Wý#Œ¢4è ’ +<:˜¢ t4+:"¬šß„ôÜÂé¸9_yFRùóöûþ|Wµ»êè%îRŠ8óÔý´úttÊefü±Ìg$zs_õ´øp²'O¶ÁíWðË(LT²ÚÜ™·5v_”¯* ~Žâ¨3æòrD’‡‘¦À\¦Ý…1¼ £þÒÝÔ|ezeþ÷›É &xyŒã쫶!þt‡ ?xYqò“GoT õËK\Ì‚cñ`h¤ÒKÙX‰Ì²™l×ÏSF’ËM$ \ALV=ݽáí‹y…|*ͱ÷4ª¾ãkLe˜ îí^¶²¶1¤ƒŽœ³°†ü°ÝÓÂÜ–ÌTÄ¡Ôƒ™VMoŒEÛ¸¹{ûnû~ûãÝÖ ³äQQf"õ¿ ‰&ÑSšTI˜‚‚a€ ±WHwaGý©sñGÇ‚­çªÛaÒqT^ ¶e> c"ÜCt:5Ξ:^±ôCQJ¹(4¨p­bª‚äZpHo«r-Óàã†6 Y?Ú͵«:ò»hæpå©ëŒ#ÕA‡5ϽM„QÆ#Ó0ϳßámôûÍdƒEo›3÷6CÔG_C6?¼.KÓyÍ+¯y=Ѽ ^¿K¯ÓYäÿÅiI³+"Y¦.×à-I’Ͽú3b^nÊf¹É¢ƒ…cåÞVfÿ’œKÅ„pÛ©sõËñ\Âd¹7ÍÛ·?¼!ûjÉŽUJái ûp:Œ~ŠLŠ#dзKgå:L¢á÷ýÓÑC‚7^óÅD™maÇ,”ù@Q‚KôÄ}ùXXŠ‹r@Ö™YdV³jðYÜ°({Bn¾íXö ÒP‰¹¾!%š¦CÄ_|EÄŽ7ŒEM«Us5©<Ì¢!o"±Sô©®7ÀÄ¡jŠÞìþÄ»j™°ëFÙ˜HLcÐÉݪ5ÅnSâ È«³­z3LIäÔÏ÷c¤AÎÝõ¢R—5–áõ&ϯ÷ö§wÿç†!Ÿ‹LÌ‚-ð%¡ÊR ³a+z÷oÞ0IFxq)€&÷Qàä¼vÀÁ‘qåp¬Í§%žµ Ç…§,©C„B$Hä³›ÍN‡˜\ ‡ÉEÓÙDÜÜ¡1¨iÿŠƒlbé°ZG]à¢y<óñ®¯ê¶W€+:Cƒ‰qÁŒàæùG½Ëøh7JNØd³BVS@z¯;"@3k=µ«5”ƒ=þ¨¹œ´v(zrúbVnèQâI "8"â±°}‡¡0ŽÞU7SضÒ=¶§MAg&À^ìGž-·§ž&>òà±b6PûYlñÔǵS¾§@c]Î)£\ø=êÛ  þdÉ´DHŒ&|ÀÝòˆìVs¼Æߣœšºúè.¶~b¬ $ì $+-9 ìX]ó©ë V³q ádÍ7h#Y)Tý8&B9:­3Ã8ÜHXß,âÃTKip`*¿8ZÌÒj7‘|>ˆ€{?1ÄúqUQUE ¹žkîì|ABØ|¨Hé4Ü '=ž™þæPuÓEcmkñÂaiëìCB«ÐR!Jº¼ôlû]Ûü… Cº³–Á›ŠòPŒº«ºö®¼»‡Ëø¾T§áX†à%wóóI1{SlPL‚„Ï• çJ©%×w8ægíÔPt<%ÅÓ µ¦¯Ê¢¦®iŽßWuÕ™œÃ‰}¢iA ¦¥œcäN½d Þ‡ÎVÕ”°”B„ØœfRÒ 8®È¿ÊÔ;ÆwöDh¾:0i€ `@ €‰•N$šœ.ó0V|zêõ&|ï°EáB3$ì8€Ä +¦@ t舆ºé¸NáŸöDvJîÐZ{„ÐYý×]Rs?,jPíbÐ6c7yw»:É=4qx/ÿžÀ†ãtÝ#®ö‘fæѸRXK5rççLÚ´Lî‚óŽ)˜Ò|:ÖU yùyçïYÓ'¿Ò H rçâ‹Âf2ýˆ‘{Ùà„Î ÎL%êêUxÎKÜøEÓÚ ×¬1„¦ºn1Æi:1Íüùv…JÃ\¤ó;¥œäÂÛ‘gV:\?½$º­Häîß¹»À•ØìèÍ=¡¢ÅæCª(þ}ºE0 ¦ùºcÞøtDFÿ9™¦4¾¥¥ò0ºìi•Æö¡*1Ú.MMÃݔӑÈë ùt]†rÂÃBé#.-šT1¤^ù»mHE]ÖpôŠ\Úøþ¸…¥±¸ø Q1HÚµC}ŠS¸´GZš0‡RþCˆšø]èÈE4†1[×ÏçÕì‘0øu eka«öÔÑcÁk([ ç©ã SWãRPÈj¦š]±7/R_+ìbt-µ~`+.ìÝ®-½¥ƒS·Xw\D±ÏÝL Ùdl¥:ûÉ’©ë‚Q&`o¨[‹^]ž3ˆËbÖ,übÚŽ„éÎt¥­¸÷·#ÚêÙ§_„^Á„iòåþüï7“ –zôÇ\à `W»Ï~Àã‡wæ¼q•Ò$0°J>lªÞ7¬Yp¦˜´æõ€(Ef)£ ¦²g(¶”Qº’Q [«Ï¡¦ÞLÈ—PÊŦÞå$Èy~`.rÔ>Ë”ïC˜P¤LãYÊrs~º„´æ¾OÕÓƳTäjìxȸ½Y–©2ͧýmLäÔÞÄ`ÀHVpK ž{ט…AA¿ ‚éÔƒ‘»Îk:8²$øövË?sq#4Æ]h<=‚ÙH¦ŠXÂ\)ÆñûEµX=A­6 ž×{“˜ám,픆b<󃆺(Ÿ#ı稷רÛæ"èDw»}ïäKz]ùðýE ’CðìËrÁbÎ>œÓŸHLCüø™?1ÁO÷§š&íþ%‘>Nð*cÕ¯—¾ƒòt<çê·}<>‚h•TòÉÅt•úö­G50Zøúò/uñéc·£óQQX¢Çf,%Œqû”î؇Ñþ„ÜGÞê—À/ù6Œ!Sßp—,K’÷h)ÜF }óÇWò’„$ž_.ígI»¹˜~æC™µ ~4E½qŠŽCó²dÉüqaž„:W3?`ž¢ çHs>ƒÅ+ÌçyIðïwpþòŸÿ€ë©Jð‡|< Û©›¡Î‡¥”–›Ë°ðí-„­-ŸsNBÂB_Ò«WB­yU 8úºzRá¹JÅy¬"+-½ô&’NL$½j""‹ÂL-ösîì2sÎD’ÙJL°ÿ)XÆ[ endstream endobj 1293 0 obj << /Length 3598 /Filter /FlateDecode >> stream xÚ¥ZësÛÆÿî¿‚ßBuÄ3îÜé'–2îØJ+“¶®G…HÄ„Œ¬üõݽ½ (ÙéhF¸ÇÞko¿Ý#ŸEðÇgi4K¤d©2³åþEd[ëÍŒ ?}ÿ‚;º.Êoo_¼¼ÖzÆ#–F)ŸÝ®Ã©nW³s#.>Ýþ}&Áâ(†Ylëõõí/ )å\2øC’W·Ý:Zˆ/ÜR>³#Î9‹g±QŒKå–¿0r^Õí±ÌÚ|÷xaÔü¶ÃÓy±Æo<¼f^©rlr*´[Wx¨«rCŪ¼à󼕸QzþPìvÔ¼É[*dôYVûC±Ëm±÷ãr‡‹Œ‡&>_ì‹fŸµKlßM^×Um™ —À%ãšóŸHGÇrWüŠ´9qu—o²%ÎóHõk<"¸ÎJ ç §yy«O W¬[pÍdšøÉyBÄ#¦­»ÎY~:n5Ûê¸[Q¹¬ê}¶Û=Rí5Í€‘ɼÉ< Šòpl_VÇ>Ž[uY¾tƒ²–F=Ø£Öyw=n¹‡m^Ų㶀; ùÅi·´8på°ËJv±PJÏo·ECþ›íð6û@[“­ó?Å;sžwoדüU,J:š^ãâÄ´J3Í•²*`y™×¤³l§ÖPŠÅi·ÆˆÿfÄCb›Õ–×Pk+÷ÝVV=Lw¦»ä/ÂçöFÜÎVË)ˤ&zÚohÚu¯‰n}{³R%p³jx³8³åÀ2³"Ï…7)5Ö¸cqâ| ?§×B‚ÚÆñLÇÌUüÕšíÇ/‚ &a¼ÌH»a÷ Ӵѯƒ£‘]…n+cäù|oľª©^¸9ê)@`µÞ:åãAµ7”ˆdpü'°˜£~æ°ãI­·ª"¸Hœ&¼Ð×»6¯­¤lˆÙ¤Ål: ¼£VÀ•’ $—Zyå $¹‰éÝÏ7¯ß½ýþæêÍ”P€Œ”ôôà]u¶¡œ·ÃÕ{o§½zCÁy6ÜlQú1…wdŸ$X÷‘?\fx¥Rç°”±QÓªrÚOÕ²j‰b•“êçh °§*©ÃÍ‘íúÛ$™ß[¿RáfÜw^*M¶¬P¨¥ÛHs¼oZ´ ˆØ­ ˜åöQ¯–8P ü•ø‰­—8x[´ æÇÏݨ‡=”Å–Ï&&“…•‡¢ÝÒøoß¿¡¶¢?ô¦ÁO2¡ÌJj–.83 |­.»á‹`ü„tj2n8D…ödÄbàÿF&lè6$x´à+*#º³QÃ?æo½ÕŠ“PudXÊ­÷ýø)š­ n‰©dö`)÷3 vEz7ûðâ“–)ŠMIÛ÷ÆÕÚÒdC Äõ‰ 5’Å°ˆŸãiâ©ùÔ¶F“mÈËk€Î`e¤Š¡óÔÍ*à4Ž¿Ö¢êXÌ_£µ´GˆØUÀC½Q…Ê>wÊV‡á{}‘ú¨jÊ` >¤ EؼEEÔ¬–[˜‹>Ô6Ý·Yáº÷Ùç‚:=‚ËïT ;ü4]—TìP¬ø@Ì–ÛªÎغ¦ŽUÖfT¢éö/¦Ô¹G>J ‡’[ÛþûE®ØâpI^RµqX âðŽµHP8B‹ò›±~F% ƒ4Ò¾¤,M¿^ýøE0Á$8.3ÔdÜŸÕd< i²mi­=+Xb²ñ±[x¼ŸWcŽê«b0) ƒéÿ£ÆbÎÔ˜*fF©IEV÷ŠlÒ Eæ)gb[?ÇÓŠì©ùÔÆF“†!@¥ó+ú<ÈÒ…&àpBp›šŠJ.ƒR¡–>%•Ö‘aoa±1'ãÁ¸E….\u§ 5¢ Ñšèô ÄÓ˜I ÷ À ¥òë#E7~L0ÉÕá2§^Bç…F¡"¶]…; Eê ¾ugæÔͦ”ôàŒçåÄS?s¢ñ¤œÐ.ÚŠ¾Ùá°s‡Èè³qˆ±ôÑ‚[n€TUqE™kϹkÙú¾ªÌ©óµkZÓ·(1qB½}ð†"l’ùÏ(¿}ªÑÂV@ sÄO« EV42HÒ§¢eVÕTÞg"œðE"b’w¹žC]Y£m3)2FÕe Ä ò€¦ADˆÓ2²áùTz œ­ˆºµªr÷8¹#ˆú„°ÍèQD†+rRU»™a$‡ÁN< vbìrwá±µaBÄÃ;_Y¶Ë¤ ‹ˆíØ`#\øÚ-$Òo!™ÛŒ*ø=Ê÷BOÑ%ÁR,ÑlƒWvÔú}¼@Ô6HÙˆ3a BC¥ÕiÜ|¡Æn2N•xzç|Ƴ¢81%FQs09æúvùçÉÔ$@ÂT¥ÝŽ¼m@©ˆâNŠ‡¡Ï@¾1Q$]€„Qie“) “Èm5ûš·â¢&'&$«š¿YÓT o«iƒPª éA$tì’êZÅkEsuPOk³H醣֒ìéPv´UYTi/Èx4ÊÄN{j¢sY&‚l>zÕ}ôÚºR[ÿ4ǺÌ:l€uÒ:ãø öØ3ÛÔL¯­M“ÚÌë®wo#DìÍN2 ÚP>{ƒ¼ 4²Ä8:̉iô§nŸ‚¶u…K Êé懰Ï^Ã$Ö,‹ßŽù+ÛÑ×ÓYlGó7ùr—Ù3)¼ž 1!è"üÜò*ŽÞ¶#¼œœ ¹N©êÉY`܈~Ø—ƒ*DÝv·Anª6ƒ©ü³4€”i‡Jk¥©{BBÊó!a÷Ø8õS&ê’<¶7ÿœí½VL„%Üóɳ‚³Ð=;ÐÎ œùîîÍ?ûî ½®ûKmDúê•{8AâW—¯¬wžb“pÒ'RÖv³˜ØÇŸ‹U­y³-=³œ-ÞªžÌé€êó¨›Ú±pów"9ÂsßýðþÇwWÿüÒó=qÂX‚ÍŸ L$WTõ6–úÂCºLO |´yàUI8é§äxÍK,ÂyÕÔa¥ ¾ù®Kq7ÔBž-:q®ÐÔ?¶`í¾:¤.'…EaŒ+ûg¤6ßä5ñüÃÛ_ÝÝžcSš0ðÔ~ º0|ýS"¬Hæ0NéM´@Ï»% †+”¥LVA°‰mø``jCÚ"»ß¹ËdƦ£l¶ `—Ô–³ ›”²NÌùs'¶×JBNìéWÔ„i¥B“!Ò.³îböAn݆='—M§"°П^"{= €qnÓèÚ€àòiBÆh‚GGO«šóiU¸UÕ£`2†1æTê9¡à‘+Œ·Û® 0rKZØØ®¿%ˆH°xœFkòvr›Rw2 —c÷ö7º«wTû }ÞS£«ÝLí9á.þ&Àë´Ý9îS@Ç#þ„K*Z²ÛÀÍ%µ>Ë9PˆXêS«ãÔ6–ƒmpˆ·'~7õ¨EÚç»#15=Õû©yjò‘~ÙD7g¬Ià…O,&D,IÔÐbÊ1^iz/H&û˜¤}<ä¤Õ?ÞþtªÒ‡I›±„w§°a),¶­0‡%Rä(÷Öë’ÉF펮k]W{êÀÌÞ7ÛpˆÒ*7²$V”¹÷èÈá‡3âoÀHt·ˆÿ(Ä°™Œ?Î gÐ5:ÁÉúvN\L8q¡Î8q`¡’ɳ{ œûÛSúI1±= KFcÂ-# ËãÙ¬YlF¿õPNÆ^7»%Š8É臵“x`êzdÊ’T=ƒ7XÊþ)RÁ¬Œ%ô›ˆApƒm.¸™|K™î5ð0}Ý´w–ÎAýÁš]|4uJ¡*IGü¾[ß=Íã„݉Cs¼w2ÿjRQý:^RƒP«_æàL*0Ö¹¥à¡,Íاs"%” úlÑûFXÓgN]ê}ùjŒgx䈒¸M3Øo'ž¶ŠÐ¿ö6°Ð/2I Â@D»”{QŽ&96G÷·pÛÜß(NéÃÿŒÔüŽJÁÁÄòÏ\Ö»Ë÷—7Ÿž 0â Aa•>ŽÎ9Ba!O9B(FêYÿÅ!ôàÏ9BiÒ¯›èœ#4‘|Òpe£ø™yî=æ¦|:æøðy¬Úc– ›H â‘TaB®Þ/üûO"!þš¿ÁÌ‚Þƒ·Ùlš}T¦„š}Ú¬Ól3‡È'~ý’>ìŠeÑÚ'%ÌÁÙŒ –( p?þ;±r¹£ž½_Ê3rþ¶Ô…'¸åJ›ºz Ê܇±iò\ A¢Þ54cF6†ôжP”¶k‹î*¼†eBnþTB0aÿ?Íä‹ endstream endobj 1302 0 obj << /Length 2598 /Filter /FlateDecode >> stream xÚÅYmsÛ¸þž_¡éÓ7&L|õL?äT»õí¤‰nÜ6ɨŒÅl$R!¥sÜ_ß]ì’"e(vnrÓÉL.€Åîb_ž…å(€r”£Dk‘…éh¶zXjýiDƒ7}!y ýÞÊŸ'/N/¢h$‘™M}V“ùè7¾Ë×SûZk/9;öÃ0òÆùrY”Ÿˆxq1¹¥Ñ¢®V8 ½ëcéUǾJ½¹©K¢]§¡WÕ›:‚LTœx©>þ0ùåÅù¤“0RꙪàʧt‰D&£Qœ†Bêô‰Ä±Çw{gP.™zÇ fµÅÄËkCÔyUòh º4 ±W”0úDÄÍÏæu#ƒÔîhJ‚Î`R–oÌ`Gâ­Ìªªˆø‘¶‚@ö¸/»–‘Hµ]­Ø3àÈF^lژ÷A¬á?ÉÛ‡ª§"ÑI»»*]GøV¥i4ò¥`'»tíâ‰DuÌ®ù*ˆ¾‚e™¦)¸b WL‚š¯ùj½4'ôug¬eÁŠ†)*OÄ{š¸§ YÎiŠm™´¶Dëë¾”JÅ"è,vÅKÂYDfíŠ÷A \|hØ®ºvDf"£v‰š“à5Ü; ØzÅ™ÓðJE"J20|*4üZ.¸ït<ýË«_¾:Ç«ãE*òÖUQÚ Ä³3ú…3pùÙÉYë Oÿ*7kC _OÞØ5ƒýëÁÙ^¿û3ýX_CûϦ­t þ\Ñ‚ŸèçšåOß^þë|:Áà %ûžŒ«gÓÅ”5³¾|ÒéÅ£wW'×v«¾‡º !¶éð œLUœügg†¼¡ßm³Í—°ÝuQE^úa T¯Ù 厷['òI—NŸãÒrÏ£‡l$äÈ,þ.>7ÎÈH„Úå ͺÎ*L_Ÿ™†Ä—°Ñ¸ ñþ~xLǯ®__ÿãùaâùDJOVw° ¾nþ¸Ð9¹ùÁÁsòùÿ>oÙÇ 2H ‰Q†"‹3‹Ds³¸Êˆ!¥$Úý~MìÃâÌ6”W_éÀקÁ»7æ7[ÐêÆÂ$¤‘×'Pííç¼X™²­Í¸2Þ:ÿd˜Oº¯¢ÖPmå@ÃÃȈ?¡ŽÖ9d6TgÅÛJË*qgbÕìô¢T´[×a"B !Ôw9Ê‚1$ƒ¦áên!0e¡%L~´§U›;úäúëÝö]b‰{‰…>×–/D Ò R"ß-¿bƒà$‘¤¶¥”þz™Ï ‚çù›Ê·¢áA šÁñWg`Û§UÃÁD˜Uu¡GÁI†1­ND ±+ï%2œÍLƒ0“@u1Z„Á}ÑÌ-ª†ñË×—4`  £‹ãÌ".iwEJ™Ës&}‰VHÀ5¿‚¢EiÐŽI†Øv?ªbØ%Áõ%t~Tñ~¿ÇÀU{Ç £ %ÕB“Œïn©ƒ˜7ô]-hÁ­5‘ÿÖÞýolÂÌ{½Ì)¸tšqp!Y> ®P¥Pû†šŽ®võzí3µz¡0YÐvL úº®Hè¹á{É)f´æSúöºò‡?¡ƒ*vˆ¸s`å5}#ÀÄ@S›ŠHó¢ùÌ‹+ž¹Ë74n÷@´f$Q ¢ÖfA±±^s_¶Å Ïû¼|@ä/0»gÞäŽWiJùRª O¤0¯9© ëBãÈô ttàÉSàsúû}’÷û=.ŸÜ;f/Óƒ|¡€PF©!×ßmƒ8Ç®wEäó¯œ~êMðÜLSŽ‡ù0x”ãc=[4PîIžW?¡Ê>SFwXÅQ›GQÔa…:tRïÝÎ=º0–"Kõï¾vÿJìã¾Å‘“0Ï~ä¼»íÒ¨ö.Wƒ«Ñà¹mzÐŽ{Q2±Î `*‘…ê‰{iWû½å•ö™öî¥Øe}EÝКjeî)RákSsè=ðwE¿ÈᶶFè¶+ÇaÕU]ë Fh Ïp´ûdK©X™!Œ£3ÖÛ ˆ~^¼¤†~¦œîeŠ’~a"\rŸsyúªmt¸/h@ zG ½-Ý” Z.,ΆŅ¡áפñ§ž’½ò«Á_Àñ¸ü‚±ìÊ/0à[ <°üv…ö¢Xrùµa¬r>µ~£%Že¸{Û:PƒáDÍ®kdºÊ›Â4À-Œ2ºŽ¿ ¯Ô®Ú%Hš°JãL±Ú›iýÆN6ôkÙÀÜœÙ5ü½µhÃõô„†J»‡ ÖA8dJü§i¼nˆ‚ªé¬Wæ+ã²Dœˆ êX£"”±ÊçóÁhñoò—!Ê´{ªò£¨-HU¹| Ä×…O?Æd[ ¬‰†"IBу:W3‹öµ­–"x·g°¯‚L ÐG"#aL_äó'ø’ ~GrÜÑœ½3XŠÅ×þî-Nc¼!y*N ªH¨4lí0þÛË7/Ç“ó7Ôäâ硦K>£®Ç'ƒà™+@zU”„~ø ²²¾/Ž§7¿^]Yœþé0Ëz÷í‰÷OÛ[~JÈ9ìžDá®V4ZŸmRkéEsèN$a4ì,±›þDýôxzy3yüPC"r$?z L/*vW0=Z=ð1Ï7G´ûô´}è ÕÞV¥/»§ò(IT‰ôâºÝ\Š=ñ¨6£Å‹¿sñ¨/55Ö2 Á}› :åë4æ‹ Q°3G³©øñãÈÔuÅ擄ÍæøÙf"cDXÀ ŽPÐi@Èm=ÿÇ™°t§ø2¡²ßcÁ8Q¬ŒIc§©Ò lø¬×’› ŸÌU¬9Ô»um8s”sy@ü·+Ldˆ [›EÈä®ÃCÄw]Ò8¢#1c ïÒ˜9f +FE´†E˜‹‡}meÉ[~»ƒ¡ÂŽ3áœû\($!ÀôÝŸÆÓŽ†>=îÒå ewlá²Äñleà¾M30¶ï ÛåÒïR#C²¶ìt('Ñmº´ô|=¼¤9/BJŒvz c~{"—&©ØýãHÚ™´u ⶴšñ%¸ì¶.¹´›(Û+í¹ó¯H2éDs&Q‡l) ³¤÷l.Þ!DgÇš³¯ãÁY¦²Ç [›8îWVGè2®‘†Ý½NI‹té)ÄÜ>WŽ^$ñ+–}+Â?Ö5ù|ìÍ·­çF},w:@o8IÝì(«ò¿¦®„C7ªÓpx8W ˆ.ŸöžË/ÄHþÂÄ?¡ÙV_vob¶]mŠü¢…3^Aà w§swÏzÿÔ ï endstream endobj 1210 0 obj << /Type /ObjStm /N 100 /First 993 /Length 2783 /Filter /FlateDecode >> stream xÚÍ[mo¹þî_ÁíS|™á à‹Ï‡ —ÂH\ô劼N…ÊR*­.Éýú>Ã9–\G+[ ‰Zr—³Ã™gž2Ö椌²6ge£UÖ™¨¼õÒ0Š˜¤‘gt娢‹Š#©ßI%Snq*‘CòÊÉžà .Ûr³MÊz[:qB–>Æ¢…Ž0¿søȦÜá•3¥î/÷’rå6çœrÞdÌápŒŒs-’{=z¹{+ré Ê%–Ù ¼7"^Ö[ÊŠñpoS@#(¶·hÂÄÊûäƤÐ(À©¨Çå3K¯'EÖˆ˜ž¹2ä'êƨ1ŠH”E[ÄDoâ sdEK€kв%ÇF±Kå^ˆHNæ`«˜£Ì!­¼áZòòȽX‘ÀP™¬-ÞOÊÒHXÔŒ.Y] —V‘8+ô‘XES‹çDʤAEo&`1ÈRº`UÄ?—TLåµ#zsºL¶´"Z¢#\ *y–§D´Ø‰e…¨RH sàQ)uÀìr(½YekÄ* dve1`”™pML0‹)@1hK€þrðòfÐnŽEÙ°ùœ2Ë x1EQ¥)Fì>Œw"~Â솂¨Ê4³„®Eå¦ÒL2Q–fŒå*>L&RºÐö˜+AÛY o±îòÞ)‹×°L›Ñe£ ³¸g‚ñ8ÛeL+†XÒÝ9-º|Øvt†éN^¼8\}ùبÁÙt:kOo—ïÛòû—ñô?'ƒgóëf®~5@óîdð¦µêW½m >5¸<¿Pƒ«æs«ÞÝ]…Ëá‡ædðb5Óv!î_fÅ/fËù¨Ytѳ\{Ý\‡?Î>w˃88vXËáw ¶øn`YÛ&.‘\ä)|ÕX_ɵáV gê•N—ïö7OQð€¬–(/¶²ÓÞn·™³Éd6¶Õ•&ãÓ‹n›ÛÙüËA§ ›Z"X…–% À¾á|0Z8@p¬…HÕyÖy]\\ýucriÊ$ò=onØƨé¯.k†A¸ÐQ¸ ÐÑK AF0ÅXÄB "ÆrÞ.Ñ夣ùÜŒ ÞžÀ @® Ñ ÂEœþÙe-¤Ö¼£¨ˆ`Y´=þÿ¥†±z‹Œ‹{p Öÿ oØ;@zaaG¶^:/ý6¼Ü|AçöxÁ/u÷Í-žóýx²÷«û-¯îíþ¡Âq ¡6bß0€H¢ Tà"€µBƒˆïá¬~«%¾N&+[¬Hr3ŸÝÊ÷/͇áèËáü$xP;$‡ÄÚ#Mö7ƒÅó4²×­Ò77ãiÓM!og×ËIO`³€Y€,О‘šxÐöˆäy·&k¶Åʽ{¡Úg«-Ðïà¬ko/¡-„Šz*Wy”«ôÉWÖä©6ª'Q¬WªKQ½Ý×.Z?§vñúµ‹+g£ÚEõ WòFµÁ¾^© ¦¾~ÜÙe%s>9-™÷C„îõrÒŽOשÐp"yVS$Xëí°íç+àÊÚå¼frDÑðƒLnå%§cØÇüf8êüd† \T„¼+{÷”+#EáT“]­@ÚÊvƒA JöIí/™‘@bæ€<µ”Ä1+{> •Ì¨¬¾lFá;+p$ú§¤XÄÈþDêÿÿ8 GŽp«GÔ86+à;+:ë¬nï|-˜ûù§Çåkà_çk\s±°Îj.Æ5 ½ËÝÂhËFr&Ù¹ê¢+Ð!ÙgÏCdõ‘§E–=5§“‹¥ê?{Fä”>ÁP1”ø.{ÔäWÓC{~×CÎ/®ò=»‘Ï7Íêê°öñìÞ¦¨w”Èæ¤yA{ö4ÄFÒldÃ;”Î&—® ç}*ŸwPaGås!ö†…H÷a!ºýa!TXˆk|¨°k#T|ˆk ¨%šP+3±dbê!œtFŽ@ÁkA«BAa6Ú'ŽãÄ’T‚ÖðLã!“xÊIå@‚lz$ÖbáÝN;­ûX§ '¤p ß1€ —LáûSv[£ðq$»ž: èääí«èôý¸”v~^ΗòýªÂ}!Ï´r‚Fd6xv §öyåšér^ÌB¸¯d½¥ÒïBÏ®ÃÃÉi;;­ñå ˲¸'“O¯}wå‚r0ÜÕÂæÉ8ÖlvØý×¼ÔIä„ D‘Ó^%"u;TTWé7Ü=›oXy»©«Åø÷˜ÃwHÂ9ûYþžb•°s¶ä«AIˆœÊøÆÁ-Cïr¤-á1õ±n/ÄèR t©v% œ(¬+øµr¶î¡ þ“A†• ·”cBR£gÎrô²=;AÓµ|ßþ»xæ§ñâºÛ¬<ë6ÒëÞåÁâ‘œÀ’ã„ÀÕ,e…õÖõ¦þ½Ýü}ê!‰ƒ¼6hã;¬p ¦™Â‘Š«¨•JùÅi9÷,E˜¹¿4ÅçÓc€á«AäÍcÓé]{žý3è¼å\@¶ûcAª|8Uœ* ÎuÇ3U˜È•Ngs˜=GFBf¬íöÂåx4XbvÇ°8‡Åed žÎ’.®ÒYÙŠ¡ï"e$9%&ut!«"„rŽ·×Wêüåðš‚·è€PÊ@´£ñ,2yŽ:V$[¶¡S”d_6WQR±VÙY-GÞ‰yØe3r†rœÑ¦O^6ç­NñPH›sÐä«A„,Œ€Lßxúl³\±“›õ>•æ»ÁÞ˜ý!8W΂ó‚WÈëMÿC€Vq"ÄS@­÷Q—ÿ< [b¶×†ÿ͉¨ endstream endobj 1308 0 obj << /Length 2182 /Filter /FlateDecode >> stream xÚµ]sã¶ñý~…ßL56ŽHt'­{wqç榓¨“iÓŒJQÍ9‰T@òÿûîb~™L.=‹Åîb¿Å¯Bøñ«,¼J¤dY”^ç7¡]5W4øþÃîànðvù×í›·ïãøŠ‡, 3~µ=ŽQmW?i´ùyû÷+™¦BXìêû÷Û7·RÊ@2ø!È›wÛþžXˆ¯$!Ÿ"•FŒËÈݽIeP›ÖäÕæV(ä ÞÇÔøXÌY,{‚Û—‹þO‡÷»l¿‡wG&7)Á²,óGØæ6"ø®Þˆ4xÞˆ$Ð_p¨Í Ü›dÁ ®Õ#Âh´OÚÑetsQÔUSîOnÿXFçmY=â,cè%FD¤€ù¤ýÃÇwZ¢^чzAJëÎ4útdöoe&˜à n. ì_V–ÀB”¦AÒ,8ÔômŸÊ†¶öÈê »‰^ ”KÁÀìH(›zWìöeuà''KÒ¼u—àxTiºlòظ]VÈY«y¡Z²E`=­ÜÓç"-÷&7/$àžÅŽN·»ªhKx %†¢„…ÒÓw¬/ð°Kr±&=X^‘[Ù{\Å©nô"²ˆIÕ#»!{~* |€'šâ£à7ÇOo\¶4nPPÚÁUôÈö]ë[ 6Ÿ71(0h(#’ I²hB’@^K“ä1•dDZ›ÛX‰<G¢8(Î'qðCk¬fãøݯ`R oÎ8æ8¦Í£ñGÞo²ÈS¹ p\îôNào:?‘íÄ"x.Áz`Ä­_ºnhõž>#·õ¯Î ”òÜi³tçrúìK·~®­iÃÒ%/«cwBÇÀ“à¡ò«¦-‹î”[¿ H —´)O©z =Ûçé¥ß[& ”ížË„ºkëãcA4\*&Ó#v* bÍ$8¡_ºÒ ©à̹š Åá÷²¤ G¹Ûz©òsYä§Ó‹[?j+¶"oõÖî‰ OÆľˆj˜Še°µŠ¸A~ƒGÝÒ lœtõØ>Ý × ߊ;g¤ºkÜ‘ƒ€™õ ¡ìý‚AÇ@K­»ŠÜBè?‹ä ÝƒÉÁ§;OY¢æBvä¢Õû;ÉÙÆ’)!§Â@AÄç¬ÔÙbÏ1NŽÑª64€‘!ã0®4­{{žñŽ[ömå ËBÄ*^Œ^ôEÂa˜z¨»%L·"L†”Œ¥YB€]£A*±ˆ.|ýœ—Þݹåy`°Æ=¨>s¬àç,ŠgZ…ÚP@ø´µÝâ”gô˜Ö x±=é¨ ×ß^èµO èÚY°—Þâ5ů² ‡úK~êô”ùÆËzJ´ÆÄäVo\ÝÞÔäî|H$ãò{Ì"à ««œò?Ìêå+X]batlæý©‰†ÌÕì {r¤”JÊÝ*gOÞ¤g†Î«”6¸éMÒyK8Rxk£G£pââ©õÖS ÌW‚pÈŸr“À‹³gcr\}¹[Ž¯Ù*<z’ûïþ2~“Km5S@½ë?ü Ëúááßïv#ãêQÌõq-Ç^P‡Ù}ÚûÖ½>Æĕȇ˜W©>Òy„aUÝ2š»¼iê¢ÄPÕÛÅ@XÓÖîúkmLíäD×Û”§DÆõ»V cêosY##©§í¸sïcO¸'kÜ÷'Äþ ÿy cl¤ýÔ|­•|ªmT¨äy»¨„³'}‰Ô¬„J¡X$£¡ð8rŠô8†Œï8Û¼§=:"z1›°NQ ñV`Åë˜zs¢i¹˜ã‹”3žð1Gß, L&,M{Î÷Öðu‘wöåUl뀩bb ymY§î€Yμ÷±cmÎeå AX°±¼ÃwÄÙÄK`â)¹Õò¾p³@¶ÆôùL:!àҙΠÌBû2ÃÎ;˜kÇÈ‘¾½kÄ æ'uuíx}Æ- ÒGÌØ÷‘¬Âr)1‹,è0Ó$ø\¹ŠÛÎr”4&F-Íéua0È gS9‘«0¶®@Ÿ GËÍrqI°…¾:»Åºò¢H-¤Û=Ú~>7´öü¤j³˜?IÅ}–N¬ø²tiÆTO,JE6¶¨±Ê@í¸˜Ô'…hC‰pEdj‹H8S›9,F·©l6¯°œ´&³TߧŠÅ©š0¤LQ t}žŠ„סd4n«Ñ¼¡I¯**dŠóYú€îzp€#Ö—PpN*<\ðáÇTxF q¡w8Qœ]£—%Àk‰‰(Uò@´¿%ƒD8™Å»»¦ËOƦ…*ôÁŽuÉ1Ëœ¬cçø²Df»³¦ç¦t…b80´gã‚* ÈÂ<²µ‚!Mzõ[É:9K‡&Ê@½'Ìû?} ù ~ 25Ž•âšCq–öí~÷éŸ?ZÂû|Š©¡58q7”‰ÙÎÑç ‰Å%}KצÙØG6ˆ;*úÇ! Uˆõw_ìm+˜4b"¯;0;02øqÚ}ù +m°êú¿µ_¬[@õ¹F-@ú£¿—½¥áÆzÓE$£þ›=gëcrþÂv}‡ƒIM0Cò*š6lôÃÃf_¶¶Ûg—1Úç䉗R¨Á$lÓ±« œa~xûlÊVÓz]iÚ˜å븇ºŒ[¹CQžléæ•{l\¤¸8Zb¨r f‚omçYçtÒyá»»t´g4™1ú*@ñM?qžú4'@ pÝjg›‡Üò†ÓÓ­ 9 ˆŒ'wbÄø]›­ Û ÆÄŒ`V,G@¦ÎÓéëÔ5G!:V}¯‚F æÿÝø*oˆT endstream endobj 1311 0 obj << /Length 1985 /Filter /FlateDecode >> stream xÚ½XKo举ϯèÛª7#’’( °‡]'³lÁ¤ƒ H ZbÛBÔbGRãüúT±¨W›¶gö°ÐA|ë]É71||SÄ%%+’|SžÞÅn´»ßPãÓï¸_·ƒ…»ÅÊï÷ïþð!M7ló$²ÝÐiàJd*ÊÓí¯ûŸßýi?q˜ ñ…¢àÊ·eÉò„q™,?9Nò謻¡./î€Í+¢ÁÒ\ ¢Qkx0ØPѽiMW—4úX÷I™Gæ¿g'£E:/m9Ô¶EÁ€ŸlÉP’ɤþ;Çãðx “ƒ'K›VBðX²xÚã¹}Úâ±Ïuš ·ovRe,‘ÅfÇ%ã)m{ìêÁ‘4Yå¼Ó%ný7Mô—»Î^†ºõ í‘Vz=€ymwz’+‰™H80g9/èÄ%µT¤Ñééà˜8”ºûWœÆå Mœ¡Ã©yW·Nݺ!†s–8ʳ(—Þøu;tuÛ£e°ûþ½îí¡< -tQ¤{œ( º§vgnüîè©~ÖÍŬ)–!*ÃÓÙ¿ìßØ&¥eb©´x%cÌÛÈÝ¢w+}xFL[Ñ®WUr­ÝÊp>à¨?£SˆQ ê|äzzœOPº±˜pú±üzç;Ø(m×™ÃEE–,ãÆ멃-ŽºôK+P+OÚ£ÈëUÆñìòDÝ}<¢ ÙgÎÜwŽ\ N¼íË^%”P¥ÏCÖ¹ò¢¾±%¶¯ýœ¯ÜüM#øèÙ»˜K (ÝVº«B¬f9ã¹3ÃÈD茤`°Õ/ô*CO)­ëÈtCO£š~˼à²+(}bãY ?ãE²\Êqù*éÀaž9»âà–~ËÔºÛ§L µŠ"oñÉ©Øv—ʼn×f®0ñë“Ár„LNZ:‡x/ Ø¤“{"ªƒ©ž3!&)ŸåŠñL1ÁÕ´ÅzêN–'ìd‘îîj°C÷Ds•4M z ±¹2@ÇU¢à§‰lšè{¯G®˜©ÖÉ͆M͸˜,ýfzQŠåjª\(¿ +^1YLj½=|üÛ/¿ ¢‚JŠYVLJª±Dñ$jm‹úå`ï "õѺ‚Ç“Q7Кuñ¶Ô¸7~ZÓï–~>W¡/©ˆCýqr ½¬€2-×jô%Ì=1ðhÁP{‹A§LÀ+Š9È_ˆp‘³lvÞ¢®±2àé.¾UD‰ÝåC¬ýß¡åP€ð8þžï¥œnà:Ä<­0Þ泂±ÃìdªXšdWõÞ—ñgÔs·vN‡3°5÷Úo+ŠIl»‚•t}ÿZŸj‡Í ³Ö ÀlŒðFFx ˜Œ ™®_}úÀ—W¶œ}W,ŸãÀWLϘ·ÜÀƒ‘)ºeï(þ%¤=±ÆX¿ ò¡òïGTõÓÇý ìæ“=.錮&$øû#À¯Ån/"ÈN|ç¡ì´^ûmkj/¿+–Z¤ßzé¯öl ÊT¯aÈ—-5êAdLÆ|’†­ÆÅ[ž0ªªƒ¤»T,íÃÑ¥(Ï&·–¾ô m]k +.À½ÓiÌü‡Q+ž9í{¦Ö7¦ë¬×11ãr9v‰©o¾\ÿÃ¥Ó K8eC%õ®&aX»ta¸zñ”övA(ÁSVÈ|Y%‘¨‹jl8kcÃ^ÏÔíùBwY×ëGn¨D<˜ •RÑw÷ºno‚Y9†:ÈÅ«0 š‚ó%ŽÂSFÙgºÉŒJS¯Äó °¡àSöˆî¶ÃÁá©Æû. Òûß±ô÷é·£øh ²¨ªÆ§Ž6±!jª¾6ýgLÉIä }ˆ¼à'ãÂ8 Xå,ó#bI`]¢ò«D@®îîÿ®Ò—ånn•;h½ŽpýÏtÖ/÷é@ ã• €…„Éò ¥-U=÷!‚ÙIªÑ ´Å›žº­¨ a ‡¾Ô5ô¦9Ò„ö+5M-h£px-èê“îê¼÷‰–Þ¹%¦Ô¨‹P˜‚Óˆ9N_7sÊòÅ•Þ„?¤ðqÇ€× ¹‡òGAü Û(ÓÒLUƒ¢%˜˜î44¸z–ÅÒžÎucºžºuão+ møILºµßçp°n¨‚:Bº…ã .ï?Z<“øñ÷‡l9ÞCB…º§äïs­›e6–KÀ ½un¶Nîöúmú„Éš¼ÁeJÿÇÿN„ŒzgwÌ'ÑÏ—Þ?"¹œßÑJÕêQUNAèXéMe€Bù5oM'Kì!º@îåsÌ+ Ƈöÿù:’ endstream endobj 1315 0 obj << /Length 564 /Filter /FlateDecode >> stream xÚTMÓ0½÷W䶎´1þLŽ ŠÄŠ„Ë!4NH”¢ý÷Œ=NÛ-9¬*µöóxæÍ›çòˆÁ‡G9‹2)i®Lt8ï˜G‡c„‹/w<Ä%˜ÜD¾+voöZGœÑœå<*êÛTE}'&Ÿ"™ š²²xt¿/¾Æ‰”’H ²ûP\êh!^IÈEþÏ(½eħ3QjåR!¦;´secH ÉC]O‹¤5“žlÂ9UZÁ¯¤\ãÛU|µ.î¾õD1*üµœrúߟÿ4Í*2šÑ­4yb\­hÓÅ"#S»ΰÔ×ò@9âù<Îe‹P݈=ÇÂ~>`ÉpöÄ4³ôHãD)C–f:î/D’"§¸‡ù!õ4”Î(9lµ›*ªµYo È`iÚ•”¹¢Š‹ %Ùíq&ó½štcƒ–J©ÎÅZÇØ¢£9ÕJ¯¡Ð5‡–S¦ÈçÞQ[AÆ@ìq²e…õûz«h.iz-zqŠSâÎ)÷8Ki&..%—“íì_GÂÏê^Å‚Íâ6 5|Y¼@—y´¨¢È h­_ªx}CC?OMgÇfÀ£Ã:Ã_ó8áyÕ¿Ý”züŽ€7½«%_íy´ª¼ZU­V•dIœ¹bN›B¯ð:`ÎޮhÞoþFègœœµü²³¶ ϪG¤mºß¸Z‘éÞØU¬¶ù9”óKúˆ¸µ‚4Zꬳþ!ý˜=® endstream endobj 1324 0 obj << /Length 3646 /Filter /FlateDecode >> stream xÚÉŽÜÆõ®¯è[ØÆ4Ã*7û8ƒ(HN,ÀYpjšÕ3´Ød‹‹Fó÷y[qéáXR`X]|µ¼zûR£vü§vE´Ëâ8,L¾;ž_Eíîw<øÏß_)Yw€…‡ÅÊ¿¾yõç×I²SQXD…Ú½9-zSîÞ·ö2¸nˆã8È¿ÝŒI‚[[×UsÏÀׯßü£Sמqd‚¹{{|’ù}m7t¶ÙT–ª4ȳý»7ÿ|õ·7ÓÍ­¿\ùŒ† hÈB­#4¤¹ Ul˜†|HR½¸u–ø[ÃHn £éÖ8¿/¦[Ã]¯Ù|Ôy'‚áÍCÕþ4 Ž{{x¶¥*(]쪻ý¦ð µÃƒãAÕìuÀú“=:Þ3´<%÷Ä“áúö®–-w¸ã‰×‡ÍÄaZ ›aºœ§um]ò•pÅùRÕ®ë‘8¤FÅ¡JSÓ´¬RQЗ ÝgÎ)k¬Ñ@GQÌ£[ø‰ ¢+CºZ>Æuö®ª«ÁßVžœÆÎõüõk”D½s×Ú Ü<ߥF‡©.HCÂ樀”&UYC–:ã÷lèþ5š+€ûeL×ÛY•”ˆÙs‚‡¸ÉûE½Ò,øÁK¥k°`à»T»4¸Ø{Ç“Ù3ѨS鮚¬øŒøÕ‡Åò š¯¥;ƒTF¦î5;›…é•ôÜÊœhrœ-lK‚ßÛŽ¿Ëª·åÇ=€,ëøDßð`9ØÿÊa¨~ÄN• Kådå¹+Ýáˆ_Ž¬î=Ž] œ49ØpÛ$*¨Nüû„³íˆ:8Wý`y=Ï8c»ûñìø®2…g?‰÷¼ Éù¶i’"È! Ï 2YuéÚûΞù㱪ë—.5ÀTÜüQ.‡f{ Z/ø˜š¼IÇ®ëÚ®'=ïÓ”¼—ÐѨ† Õ{$ÆÕþ¬Îö‚C°w#Ó^j@œ?·ræ}¶ÊÄ€7êäv,¶8hœ+]nÑö}ÝÓ!ºýa‘5ú3¼|-ý›úñV0³ö4wº¾j›ž!ä/a×­œ(à¦möäoA:‡¥íJ¹ÊÝ8Lèi拾ǡ"åè#K`ISs% BH;¿a(jM#™X=äp‘|]‘cïdá§tm1¼ãÀsœ{òËp"ËTÎz¬†Y< Þ7-î{,>ÃN n×⚃‡u2fˆêçYL:îl×Ï"Ï!>":´˜ t”ƒ*ŽW©®‚Êê×HÅ„?Ù}âhð'Þ¾¸·mê§ir½éÒ98×|â/Œ3ÿ•¸ž®B@’…Iž5DEy: ¿m%± Aéýº?y¬ýàl)Á P”£,U§Egd¢`·JÃT䋘!ïHãtDzM\e;RŠ€° J˜<ıBKA¯Î«È;àËß'÷Èß-ð«cعjÈË´tƒ­jó!;爀ð•YÊ-–"Ác*Î0Àõ…:Y«YÕë±$Ņȇþ¬á![ æ$­®î:ÛUd;01e?0.ÝÉŽèÜ´ë< Ʀv}Ïû¬´T<„{Ü}ó§A²A1e€¶²ö‰, œù†©ôO qôìQ‚Ƹ!Uˆ¤Q1‰ëpÀˆ":œ^N.U¤Â,›”‹Ø™¿I È¬ ´yjÈ”#oÑÍ©º‡O$Ö^)O=#øƒì"÷¹5 zBSvš¢d ¤ùóâ,äÝsšÜ6%gU¿åu‚ såŠÉÝ'ψ…åéŒx3ZÁ¾Ôä«ÃG€ÙÚžæmä¡îRƒ‚ ³S1úû Ž7oS˜0Qú«Ã¶(³ÙÖfqŽ+Ò0J¡â—(*§6]4Äö+Ëy©^½f×2æH ·®Ûæþ0Eg˜r9,óç:£Eä‡<Ø$X›,ÌÀ×%ýSl¾€ÿtUÌb¶ñÅPO.§þRß æÁ¤ŠÌI+í—yŸ8qwúƒ{(­C•«¥Ú¾ˆn:ò34aš/O®¿„ÄïàDHç?Œ¶ìÆ‹Te’rR±Vù „„­!³Àôs“¸¨£Ù±C½U£Ò­;`Õ^¨9f'Ê Ô"Y'èH4Ïý=‚éRðÛ·äÁ€Njü.ùÝ–F¢ÃL/Yöá …'¾Ôì‰ „’|§:ÌUþõÍÙX°Õì¹BC7ûÙQ˜àû™P…š‡o"kõÒl¹_ÅÒ¯A¦iuMˆIŠ0×銗û5~õg®}}è²_c °"Sl‘­OήåHÛÒ âx/ŽjC’)p-š”ŸRd“Ÿò”“¸’ŠSîiìÀäØRóÃE.ø=·½ŒòèpWÉØ}\3eŠ%Ÿ3¹Ì›­¬HE˜è/ÈŠÔl¼b´L*p%=-̳¸ÄXÃÉÌ|ª—¯â8&3¹/^ OŒÐ)ØŽÊ“"’ºÀ³á+q¿iÙ~êeº•_^“/“Þâ®0c¢ È$šÓvæ&ŽOm]sOó‚ÜÖ}:º áø–™Æ»,. °MÖé.„îlÛ_‡i2ÕÓÓ7+ðZ*Êü*öQ]…>Š2ö’í$ÜšiǺäOIãùƒ3¡dÑ'Ì·]nÇa–f«æ¥ûA‰”™¹ kÅ-*Ä ÏT ôÃ|£ÌߨtÇ„$—µB‘½¾¦o6´ÁFÀuÅ´kMpèAµ³ŒXÉé:¸«î×Ë?ÒΟŠ†"wmYB”î¹I€@²í\-z™=[îMAòˆýÁ¼À1fR±ÝpÄv ÎŽªh½èØ5ÿ£ënÀàî]÷MþÿAßI²ôDÕÚ —+€fª£¸£þA´ØšÔ†Rÿ'™Ÿ5?ˆg°Ò5íxÿniþ!ÍÂEOëCÉf¢,Œ×†0•ÕÌ)˜Ôg£›ªfoÜÞWÞ™”¹q@µþ–e…±õfp‰á 3c±Péúí‡&Ÿòr¬ùÏnz@%盥†g˜+µˆŠ‹òªF$ÖèM­4VYT*§à§¾B+¹j¹~¦EÀT•Ó—û4%ð 8#Çz¨eÖÞ3ëhÊvÝdeF: ¨ž¼"~TÍŸcO·;’ÒÔã¹9œíïSƒm­Ã*LfzèEæƇsš«çñ¡b†¯D¬*øËï?Ô(©i»'œ0ßëÎh 4â ëÞg~E4̽”Ýf1Tªä!šºn_ÝÊþÃ‭ìö Í:»ÅûƘÛâàíϤÈa(…1-¬ï{O$JrÒ`Î;*äŒÏ‡aN%ÏÞ/µ‚”|¤žQú¹çK^|X¬Þz½\¹J†µAå–›–-ÿ²áéL(¹ÌÓ±7à‡í+ÄœG?`¹–úê|ñ㻉üÐGzé|œƒ;à µºz¶äWólÙU\ïâŒ,Á'EÏÛó[‘·¥rêÞ0ˆß±èò*Ê?Ì"¶F΄Ûs/¨¬ÝÄà#{÷at<7AyûñæŽÕíµÅ£?ÃJõ€=õÚ}’Œì…Ýgz‘Ä’¢êcßOùü}%ÝŽ·â­8Þš¯ñl¿T}ÙRk^ õôw ྠد-f¸Û®©æŒÓDÏ´ÜïF½aòþ‰¿…ˆ£å~M¤˜ç8u´ ½êÛWõ/@yn1£ g¼±Ô˜× øÀ¼ÎW(iä÷I'!ðÖG³òYÇö;‰4ÎÿhR`õyšþW!û‹¶œÑšµ3¢?, &ã­—Ú-W´ü娓ü¡ŒŠg…w…‰1ÅŠ¬?øƒ"Yý"®óDýa:¯€´ “"ÁƒÞ¾‹v%ÀAÖPYæ»GZuÞ)ƒYW½ûùÕ¿Ù2ìç€UÇ;¥Ò0^õ 5V^iÊäFY† È<ò È}žWÏë¹™ëÒ9`id{üŸ†U‰^jD|{5ôöŠ`+ˆ8fEÖT¾¶¢~= "<¬ÂÆz15ÖqF ×<0æ(Hyñ×>X|twô‡ xM~tÄ>Ðð^'#Éé[ÍO×Dy1¡±ðß)¾`™$£´» Ç$/ŸýŒöEÀ]®êH¦GpîXÄA]«ÁŠžÂ … [Þ âz×Mïò¸¹õ[›÷kºì$ï ’H²U’÷qÓûBš óí©’HSö×”@#þM}‘‘N¾4Y+. ç΢ Ö?™à‡ÏœtN˵¬ g'ýmüàÎn¸eÀ¢ä¡KsúŽcHß3Jßû›-™²O~ä¿Í–ún¹ ðÏ]7 endstream endobj 1331 0 obj << /Length 3465 /Filter /FlateDecode >> stream xÚZÝsܶ÷_q3}ðiªƒ Éæ)uåÔ™¦“ZJóàdêÈÓ1æ‘’gY™þñÝüOr2žñ X,öã· ÉUÿä* V±R"ÕÉj{xµ½[qãÝ7/¤·‰›ÑÌ¿ß¼xõ&ŠV2iÊÕÍn¼ÔM¾z¿N’‹Ÿo¾]©8&0° Qß¼¹ùñb£”Z+ÿpÊ‹«¿O†_ÈÎ|Ä‘ÑŽŒV&ÑB*ÍûÿÈ7s/K X.ÿUÜeÛ‡‹M¨ãõ›‹D­›¶o³š Û¬®›žÛ§®àF¿/ì²f²¬Ž„”~ÝÝ®¿ÿpȪªÙ.1a"aR?9¨³C¹ÝàtØ$û²±L´Í©/ëB\l´ŒÖowL}¸ãusâÎý|LÃÿ–Û¾¡+Ù„°g2”Jo´YñùX5%ÎKC: 4äúúíwÿ`ßÒ~ ¢ +ŠùÍ…‘)ÃÄpžDÓí‰ÆDÂ2¤kœRÆ÷é¾ßŒXвù6Äüu±µ’NáLÌèûû¬*ïêCáe£Yóh„êqàÞé7Ï•VZ„*†Ý¤eŠ<½ÿ9Xå0ø-AÇ«{šyX)Æ ZÕêúÅ–×*¡)ÆÃÝþ| fëcvg…/£GÖ!µˆÃx"ã',ÄÎ~F¢óE­™Dò’•E Q\†´Eòø ¬9½¬*ä5Z×E‘s«oøĪïNrÝL…åF× ­å&++þ÷0]§¨»“[¢ßg´†vŠÞr/kÛ ëÛ2»eUdmJ³¹¥mzRaÕä`¤)E.Ø<§þdcÄFDñq(×e½žt¡?ÑÞŸ !Žù7gM(:ìê5ù¤ïI,Ÿð?»Džõ·º¾=m{Pê£urÎ…Q"â8œø¡²ÉËâ’ÐôÛoÁ;îÚæÀ Ú­VÂÄÉTw§üŠƒuÉfW´» e€¤s>$ÐB+ðØ2Ûú>Ä~¿-°äCfÛL}ò§E$Bn¾ÿÆ„¯`Û7-÷É…`ï²Û5í»]ù{Ñ‘=Ç©µg$+=?¯Ž#$“ãž7g;ù™³Í–ôÆÌ,ð'4ìݲ¾ÙK4JèdºcE⟽Æ{]qÌÚ¬/¸—µw'ç[; HÐx[óÌëËí©ÊÚK4?eãÑâŽ*)d¡ŽBE©›Ë€}Çœ#úåÐ×6¦RDÊ,én=Ò]$Y©mÙ‹¡ß“—+†/ Ò r}GN§ƒr]ËbÌ, bÙvC¡o»õ’8"rávÆåÒ*‘HUâf”ݢ˒BGzXeš/L®óE¾Œýìæ Vd=´B²¢+øìS#t¬˜x3i4½¡É½ ¥)ðúŒY6;þ½ß7Ïp¬r÷§¹ë÷“»}sªrží¬:Åo§Ì®†Js‰W$l›ü푦7u^ÖwKÒQ! ¿áÖà`—$BŒt4èñ ¸Ë!HMÊð'bÐ @Î5CëXÑõÖ¨¤ØŸ»j&Ñï䆔JGáŠwN¦.6–^g!–-Üxì .ˆ;9ÁSß÷n¹Û·Ý:¼J¼S‹ˆqÈÞŸ…r|•péµµó0 >C1Û¦:êÍ‚idÖ¿¢_סb¿Ž®mÐ>ïyB„¾‚“8›¦sh„éñ…š£uWÞVvdÇKƒÂ} ã(Sû¶Ì]Ü/úm7Ù9ÍJ™© "ǘ5ˆ6k¹7½7 x_I¾Z™õ?›AdŸð`:k%6wöþéí-„VA²¾ExwÇm†f8ËÉž÷²#°[´ŸèèØGŒÏÅǺͶøñGƒmÞñÍ‚«=ÂEß–ŽH5àHÈ¢–Ö*aÒ/Ëé䇑wŒ¨îPû?´a»¤±û«—,À·•RHÔH™#9ÇTí5À%o–mÁ¸¢íìIü‘š6w†É.-]—}·tNi(Â4jÉDzÎÏøë¥Ø™"^i³* å#Ð:ç%ç\È¢!ô<¹¥:³%Ž¶`¼žºì–;ÎIcÛ;iìP_,0¥[ÙÍ 7  œ}"''g:Fξd/‡(„úƒõpŸ,9ûAÆ,6fÇ°'\â- ýŸÄvoJZ#N'œUˆâ7arçרÙG—ÀÈ–]šAÃãT@Â4Ë‚Íp ‚!ì¥Èé¡èqt‹‡iZ›¸ÁîÎåÓðà¥ÀDl úc{f)"`ÌÔÒ„™–‘Ø,6uÈL2£™lêGŸá/çÑ[#€Šµ Ád'’Žmpž¡Aj`íÒ®0óœ¥eøeé·€È_=,y]pÓ™ó€àÀœ—ÂL<ÈcÉH …Žoñe݈ ìDƒkð–9ãµã8"Ì·@Ø?Ä[ØaÄIä“01I® @Ð’5ÇB›?Í¥øU“-ÂDeF8T«Ä‡Å‘5Az% LM/‰QŽL×5„ —4+™Ì¡Dƒñ bô«âRUl HmKÒÊ; £OU|æÎD)y]!åÈ=ဠ[•e辄LaA(X`QíSÞ3”^pdlSŽ%åᬘç½ZNt4)Fî><±Þ4 ¾ØëÅè=Ãô®Ä¯±-GŠ;Óêx„ë10=ë:À·9S§I4ŠÏ®f?ºeU1³vaŽÒ&ä‚ŒLEŠaŽž*æÂ3¦î‘q5KGë×€Æz¸w%)ÛBEõË1é,ªOd€.c˜òÚ?î ü.µã`(‡=D7K;w¿äfÆ?„kO”ƒ[Eiy€¢®iÙ´y+}J ÖõA‹püZ›Ý=ØuŽ‡;ʇànüMÑB‹Àaꯑåß]}}ýû«E…"6>Ö7í¢¯ ÇshÅ«ë›·ß}}sõe:¶á0NmÜÀÖL4HŠسê¶Èè •X5CR›^þ8pccñA”ˆ4™•²ŠtT¿ˆóRE$ah‘ c'†‰Û¯yl_d„,„õPÖmèÜï 2‰$åêAáÖš2I¬éÀX^à 5+Á·€ú¥ådÀDÐ)k0C댱D“/²¼O×WâøÚ"Ÿ p@¶!ÉæÚh)‘Ñ¿t=œ4$ ûƒ¤¦eK¤€‘lI‰!"O½1 ãw YVV • ãHc‹ðS6khyyÅ”’ÅJ©$;`Ùãø{¾‰ç­ý£»g=í%;’û}éíÍÐ+7¼bbgל(]3–/ãø‚†IÐÊK‚_ÅzÂb<°lÄaˆåŸdìו؟å[Ž¬-Òš”™ä&žéEY1lÁ„#ÂUgéttªçý—õmqlíKGÑuTAUw|ü$@N >µþ˃`šá+5ÁH¢ÐÉ8¾²_\N •Œ°×¶:åÅ¢p”0‰~Æ»A@ö5¿<µX*´ô’¶ø3L­ò}e…˜*1ó5ô$r_v\ÔRC¥*Ol“XVíz[Ù -’ž²éÑÍRýŠSšs£fýú’ý-`t€¢çf@ÈK|>gæŸÖ2°(»ë\' ªÒf-KMøx0¾wu}õî¿WÞþûûn– ‹R¤FÌêå e Lª'K?ÝL2.¢Úb%œàÔQ˜Çæðä‘H„À“«[¬@(1œíç*·É¨â€Û469N%'ÈîmÛLíòÑù\Ú¿rlL"ûJ ¤_–¥ˆáiè¯Ï–—_R”æzõ rãsRR± ?`‘ÔþÍòœùqg±ñÒ¸8ïu‡§›UBƉ0ˆfÉ%¥¶f”èi pþ»=0hL²þº‚„aH*W7Òá8ö+lpúœF³Ž¢¡ŒÞ,&q˜å+5x2¶¬;¬bâê»Sm›°‡!,‡<2ãœ)¥+Zà’©œªÙ¬«Öi$HYr€kˆXI‡ŒÕ_}¦âÅöä¿ô«/GÆom•ìõå´jæ‹+¸º"w‡*4}ÂUÚ!,K7íRq.BóLôKD’NŸŠÏx¢ây+·UÄ{1 á1…pù­ÒÍÊrG˜WeöxþO*ÒX˜pšª9i_‡Xo&/†&æ\sŒÊúxê_5§þˆx“B?•dü37’ pR úØWÞÅ-w¬\碅]€r^!>÷&›Â*!dWô96%<å¿]þ›!7aöȵ¦'Xp}ïèláX®¿ÇD‹ÿV"qo«0% ÿ= ò#'ÌŸ\u³Ÿáw¾¨TJ׶øÈ¥7"â³;Ýâ5ºò&’¶ünpBŒ”ÊþõÒfå{˜Ôým±œDpzËyéÿb`lxÓ¨ƒ×˜žrI™¡·’ÑZãWT1[iÅ›€ÔÝmøÇð' êþÕRáðá…ÍÄéO§èé€Zh‚–¼ÍÀOØ6`G€y‡Ž—ãœé7 ÀOám(>”¿³àµ‡³›¥tÈ•„´Ç¸ü4’¸b’v p§¤ [úÌ‹8å¡Þ}03`©§•ciÿH'i°gòƹagâßf•Û²w*þæìý endstream endobj 1339 0 obj << /Length 2612 /Filter /FlateDecode >> stream xÚ¥koÛÈñ{~…>R…Åã.—¯+ú!uœƒ‹;MT¹À É•DœD*|Äv}çEŠ”©:¹Â€9;;»;ï™]©…j‘x‹È÷ÝÄÄ‹ìðÆ#l½]0ðá§7JèV@¸Qþ}ýæ‡÷A°Pž›x‰Z¬7ã­Öùâ“s½K­­—+ß÷øÇåʘÀ¹N÷û¢Ü2òýûõ¯ mêê€q~¶Û4{–ùeì;UÝÖi¹\©(T¡'ËÏë¼¹YœZ£Hùº al\å–!­·ÝÁ–K;m³\éD9m…ÀÂp¼0 Ü@Ep-Ë7›öñÞ>Ù¬k­OΉŒkÓ“_áΉÓî,žÓØH¬#§-²†QÕ†¿• âw/ðºr_ü¥èÿè[*qŠ ~…-Ž{¬% 7˜­Hù“aqž½Eè  9iKd¡óîýZ–aà!@±œÊ®YmÓÖægGáXsi/ Ü8Ö£@š‰5eÜ8h‹¹R¢\O ­1·SäªÓiRŒŽ¡(NuûLÙªãÜ÷ˆ>Puû¼ÏŒ?Î1)9TÑ43rÖÀœ“„rŸoZ,ðh¢+ž/z8§2?ÃýjîI¾È—bÈæ©t_ö‘JdúŠq’dxÁÒM·çAµ¹bÒ´H÷•þ×"yhÆ|úL \%ÌÕf  Ý0b€ßÈÇAËÆœj¹£lîGœ*K @*O·äRÇ^œÇV ¿}Ò&º>¿‚D&4Î!m¹;áRqÚ>b#Kþâ#úƒ9êý¾¯§Ä&A±„±s]a`>!â9|" H*L8jvâQ_¨IÚóÜ(f›t¡9'‘ë‡þ¨1ðãÀù`ÓýŠÒžg¨¹¬uÆídÏ#Å‚NbW™x È8¦¯@ \#L&Òv:–J9³RG:¼,Ä=02ï)ž«#5nq@³”‚ððEêûRD L™®¥Eé—쳨 ?œŠzÁèÚ&¹,ü 2P¿j0b9¾¾r/Ä=F^­kž–7ø À—¸ ÁèÞ|óx_ϳ6õw.¬¡O]h°ACõ’&²± "\Võ’â3ŽÐ²2$õ"‚Nï •ãìî1­%²ˆºVI’ªe%èIŽ{ÜÙ’)(B1ìÇÁ+VçÃAÁ¨w—o´ë]!è¿é~”#zë@_:¯<ÑÇ%ÝÅ—uw»¹P½h 9Õ§X´8³­ÁËëPóŽ÷ájq`ËÙäa ¾$Á©OôŸé?ækTŽœGTY𻫨ÏCă!pQo!¶ÒâxÆÕç)âdØøkEÜ× í¤En|¸r¨ö"R2¹†¤Ñ´uÑ7ÜHÃEcDRNYB Š7ì8‡Y§±ö¼SUJ¹áÂx`{ÿûûs^½-Ÿé[§GL{sã)îͧÏ{sšúp$xчh'¹GÚ—}¸jWAÿ1–ïrÞS¿"Ïù¦0‚x MÂYù9Tdw€È£Ã¬Û·,Ó£×ÆŠØk—’ðйš^HF76¬0QBÑõJÚû¯ö_º¢¶}ÞϦ-O ]€žÃ˜Iëã¾öW‡ñNv´h_lËñ!_ÆœTu±-Êt?—ÂN‡R¥ô¨¿‹›¥\¦ ©¥\öжâïÐâ $7<â8ˆ ƒ÷~ÎCxúv5lé9Õƒ¼°nÛ¥8ÒY›¯H\›ÿoIð='¦KAêjËu·˜}BQÐ.…C±¥"ˆ¥¦šm/ÂÐ ¢pr‡¯Åáþ ÂR, 2ò_'äd„©ºš¼ ‘|±×qŸÎ^Vhßõü¡Üâéý¿ïÞþ|ûÓÝÍ»ÙVß÷¡u€nzì¬àíAŠySÅCa‹ÇùTÅ}ûÔék6¯)¸û•…icg14~þ!?î!~‰mÝÉ¿øJ™¹<ÛjâÐWs†¦W?LèåûPúsñ54iVc„›Ì2]Ó=4­8Q‘î™JîõÃÞ¢$j˜ Ž`UZó¾h!=v5Iht«Óó*.xŠCéˆaú±hwŒþxûË»9ö‹ŠI7¼Þ@3p©$`|,|Á7ùþr ëW£ æ ÂÙ1gÏ5À0Ïíå' ÔõNn‹ÓÔC­™ÕÇ›‚ò[¾ŠT_Ne®£ é@ Ý°õé³·Èa”åšhñH”‡…]0:ÿ~ññÍ¿f“»\Ì'¼Sð’u³ÏWÒŸQUBH/^Å‚Àƒ©š/W¥žú¥žo:ªJµÐ‘Acæ›PsÔ†²+±„mì¼b¡MN‰¯]7ü‚ÓÌeô¶\Aμ¥ å9×Wü=5€08Ûý®ÀÃiÁALS¬¼âpä÷ðÝj'𦂦  w{詳Ã/pÞl Êx˜ò§*í*/Àð4 KÍ?DàøåUyÈﯿ¢$£TÞo(¯%Ÿî>ŸL`ð×™“h§4$Dh Ó!\™ô©Ëd’¿ñgXH—;•c`ß]åœuEÉûý??üúöÃ;Ü|\ßþòv}ƒ~q‘é"ûgŸ—´jDœc£Z=ß åtÅ7= Ý–/~k“=mõ¯ò“ ¦¼É‚Ÿ†ø òFF6)šÝdß_á|y“ÿf;çU÷€‘h‚3sŸl=£ßÁùzj²êkò¦Æ¨ùTÞÚ­­ÿŸeâ'jþ5ïÌC¨â~£›Ìqñÿ=b*ü/Žy¨Ð endstream endobj 1343 0 obj << /Length 1464 /Filter /FlateDecode >> stream xÚ½XÛnÛF}÷W}(¨Vd¸Üå-AÒÔ.RÄ.à Çhq%-B“/±¯ïÌÎ’¢(ÊV‚¢0lîevv.gÎf~Ø$v'!çN,¢ÉòþÌÕ«åzBƒë?Ϙ‘³AÐîIþ>?{uáûæ:±³É|ÕW5O'7VìNoçMxè9€½zq1ÿ4µ9çwàEÎÎçÝ=¾çhJZô-bpN‚H8Œ 2`™dÙNûVºZÕ‹TVuY<-¶Y’v}×<Z6tÑI¡Í"'à¤ïª¨ÕRNm„Ö¦˜z‘õ@}«,šZå²2Ë¥¤ååÔfZˆY‡)-'Fîbq«(a·.“œÖªæ®U6\×Jò”¶jÜ3Щʘ ë¦Ì¥‘ýªH)ãóÉ‘s?´>»L”U0§\7÷2G¿jÚ® šýX»Â!kÓL±ECéª^°t,œqèp£9â€GLtñÄ«’¬*Ì¥›Äô€¦˜ý%Ž7I¾–é¨M"t"ßk/Ð&ÉG¹lj9f’k¤ÇÝ´…ç9Üå&r¬ï­Q©g.ý¢ÕX«¤€Ç«.`–ÇŽ÷`8>ì9n ˜§«a¥_íy»§`¤X‡×h“>Êe­ €]+‚â´y[7¦daí\{H"°£Œ¨*âôp)\nm“µ4z¢¡‹"p!¼=×y+ý‚CC¥mŒÀ*`Üškɤ´ ûñX6Íý¤VÉ=¸à r*_ƒc^À¬»¦¦Å ®ÂºU0È×´µʶª“”äpã…Ïm‚ÛPú2£…#Xàà âIàŽ 8þ^,´çíž‚‘Ð ¯ÑÞ¿Û$ÛZ–dŸnÝ\6Y­lðN&)’ .":0ëäØZÒªˆ˜záž3Ï°»‘~Áô¡Ò.ë3J¬ðŒšÏ=!i 5!`“&YJÓJÝo³'oK‰Løˆ3aøU'\Ò>6’ÊhQõæõ]ìõ ƒ+•×r‘Ť€™G¸xؤT®êŽB°à±¶IuGÙøYMÂhá"ïiˆ¾§ÇátAج{µÞ˜– xLœ =Ç i%BöÔ;ÄÐR?¾Ð’î38ø:Åþ’5fZ¬F9<öÃã–5Ûxð™‡ÃVê5Ô¼Á]‘feŒÒÝæßdY O¾ÙB*K±Ž©'׺Wsí’ÞmòL}Ñ}Á+²,¡7kŒy¨0ˆ¡)’¢ƒ¢NcL¤J2õ-Auör¸‹wGVxXµZå=MGÐ)íTAk®@Q’™Ý²LÐΧö6²¯»)œA Ðú»=Ÿý4éå§iñ2¢{}uY@åÈG˜qóáöæòöæêöÍF;`’<Ž:9(eî¹ÈoôØ­à)"óÃìrv5ënžá¯VÆxèЫ‘¯_z{ý‡9¤—Î?Îß_¾ŸcJÈ{ÿüáKDÿuñ¸›Ç^0ßœZ}ïsÊH¯Íîò×R™ "ÜQYSÉ~ª" •W5 ñÔ¼¦Es—I Ñ0½cÞvh줺ìã*Ã…¿D;ŒÁ`”ÒvÀ©IåQ,Ø~¯µ?Cµ{n蟄 ­Å?è&CÓz/Ó´)½ÈœDÞÿÁ×IýÝŬn“ªjEŽrH—ÕD­>#ÿ>­ò¾/’ì')Ê´}ý)["¢ç¼H;27æ)„ ƒpç‡h±ÈGHØwŒw®ô ìžÑæÞkÂ7Š“¸‘¼Q,Õ®@T>ÒùE¤tÖÖÏÑ…¯ÄJ¤´úWÍ¥;¸î´jõ^¨ÖÒ[—`WW3•ÏÀ„Ùx­¹§ÖêÔšjäáÿRj'4múïâê8JE¥P ˜O-oIT¸Ø×O€j¯ ˜ÁªÈ²ÿƒB¸¯¤> stream xÚ•XmÛ¸þž_ár»–E½+÷¡¸.’"E“.òáî°àJ”W,º¢gûë;Ãj%¯6Éa_†Ã!ç™g†›þĦ6YùEœoÊÓ«ÀŽöÇ 5~ýÇ+Ár;ÜÍ$ÿ~xµ›$øEPˆÍ¡ž«:T›ß¼ÛyT¿ÝEQä寷»8N¼[Ù¶Mw¤Á·oŸ¨U÷ú„­Øû—:Êò‘ç·yäé~èe·Ý‰,©Wˆí‡¾zs˜,KÂð€’ÏÏÎÏ ÒØOE¶IóØQL©ôxߪ-¬O¼RŸÎ­úJ=h h߉È ‹7'Õ™Fw“ÐïA¼ß‡Ôÿ+}Ä }?À¤Xh¤¦éuTý_r’;·p ,¦~°X JU]—;”»«êá®Ë»°Â­qèæý͇›¦»knðÚïÞ|<¼{ÿóáÍKÛ_ëU_U9Ê©žô’LãtÚÒX)3ôúÑZìÔ¹µà—lî—]F~Ù‰Â!ìÝé¼Ý…9"îÃk^»Àe’ûE‚5vÑ=4%83@þAÔjÜ׸Ūò<‚ÌH®ß÷fXÛ4ŒÁE‘œAµëúiÛ˜PegÇá<²=²ï%žìñ™ÁlhÉ܉<ñ !®1R‘ðd‹[FÁ42 -ð\œ$^/Áœž¤`–#¡ÑJ3P«Ú ÎéuÐ4Øh ¡y‰4ôu[Ýú¤ôðиþJú”º3ê¿£êJÞK×Î µ_ ùZõµÄöpgM§ ÷Ô—m˜yª7Ì4¤É6t_%v—ÅÜì²#<§s«=KcTÅk4}ƒžÑNº(²W—¢Ö"aÁkHtç^{y¢NéFG£ˆ›ÁÐXßt²gz,u;žºÝ ®1I½ÿè~R½O Ì”€Ä8 ”H™Ÿ|‘¹ŸàÊÐûd]hÀÜÊÒp–ÐåÀ÷í¶p¦þm ç‡dÁ! ¦‡IgGc˨r€;…Yì]ð~ÍT)GÃbg.ÔÑ‹Ø«ÀFcC=CKöêëö4INÀÖ¥1öØFJ1ÖŸ_ì¦ #ËK3d¨ÞÛÜóf7ÌÑWñä+`_ý½"}:Ì>Æ· a–¼p¨WÒbííà’G©SËSÓ6’E. ÈŽÓ©ðóÂFN{–}sooζš@®S%ÒWf›8‹ü(Llºô˜óƒþ‰ÐæÍåÈ<ºõ»™‚•BàzkÍGÂAÆ Ù¿}¢ë­ õ)&#ï“=òî#ùo*~AçýqC²gÀ« ÀÎ à[…Ï^áO®Ÿ¿Qyš!är‚\šä¤Ë.8mT ©ò„,öBù‰²“ÐÆZrïÛ:eßûÛ²=͘4©mã ¾ª5ê‚ûЛ½kÜP!¾0¸tª!¦Ö~n5?dÂ8á GÕ©ªµx‚°ÂUZü…®â‰w^"§ª¢üÏÒßBó”²°†÷Š¥!<ÐåN~y ®»#B'¸=»Bç FcÙ)¬žÏí#µ¥ºç·$õ‰ÙR[»ï/}3ð¸&¨oìö}%ÿÔª†…ÊaÂB€'!ׄ0å^¿%šöÙrÁ‘§á]¼[y˜ ²éì«)õèy”9:¦AÆ׊+ `üð)Oèr_g{Á¯×üPÄP‡LåE¥Nà€ò`Ó=n†i=£wÝšÓ 1®X3D@$Óï(  MiX@z*–þÃðz±4KÐ)¾‡Fá'ù$´¬¿Qö‰0öáÕåÖ=ÁÄP6ô€÷Â?'@™ ®‡¢Â2¸ßFLKE\/S.Æ*0¸º.×±ØÊìïF`1¦ ‘ÏÓ„p?.¡ ‡Úæžh‡›Á¨¶Æ<’±©Ñ-Û† »éÝkhvf¾l¡HïÉ ËÏAÛ6Ýg—ùñ]‹Ì=@`¸añÌÆwû“¼3Šh¸\ŒO½ÿ†´õ endstream endobj 1352 0 obj << /Length 49 /Filter /FlateDecode >> stream xÚ3T0BCKscc=K …ä\.°hQº„äÎeUgbl¤±*OS endstream endobj 1357 0 obj << /Length 2996 /Filter /FlateDecode >> stream xÚ•YYÛÈ~÷¯l– F4»Ù¼6A€a'’EÕblÃáP­×©ð˜ñäק®æ¡¡l »«¯êê:¾*©›þÔMÜ$aèg&½)N/¢¶÷7Üø×_^(™·…‰ÛÙÌ?ï^¼|E7*ð³ S7»Ã|«Ýþæ÷ꘟ{Ûn¶azÙ›­1‘÷Ëù¾Í÷e}ÏäCÛœ¸õæÍîWn=ltâÙ¶+›š z³Ui¦/ 7v{ñz7²iýüãÌgHà‰¯Udäqj|¾@¶ÙF±^°œDŽeh ËÐzØ„jb½H0Ô~„¼ý[œE^,;lÅ^±Ñ©ÇR»J¬¼G¤Xž× }UÖ–§öG¡žÛf³…I…í:¦š–ç gI¼}Þ3óxBÃ$+'B£¼¯í~±2r»Ç^Sí­Ð䶑\o£B_E|™²FNñC^àÚ$óú†¿øžMûIzeä–_æ…>H:UÞî(Kq·d¹ó¶\&“;þÖMÏ»œù‰„—·ûn[4§3H᮲ Öó.,NªÐ¸¼˜h ‚y½ç†ÍÛª´2힎t¢ûTÄÒOEú0å±-ûÞÖrvÃßa£<Wôì¾EÓÙqKg>–Uµv¥C^V0#MyÛ4ñ@u>1E® -¹ Œâ ÄAêýD—„!*~{žp&Λ®#!Îv†ýZà¸ì…úÌlj»¬`9,{ró3ïwønÉÞ][îïíŠbÑ“˜@y÷ù÷>ˆ‚¼gbeóNš5pÑHÛ¾T\”–_±z‚% ,ʨԻcŠ|è,ïÇRÀe!ŽyÇý} û„¶•}˜ØÙSÎý²yìpé¹µe3tko2W··ükƒ/ÊF.ÃÀ¦Î4Øîèttª½îØ ÕžÛwd–;¹Œ÷m^Þ‰-PT§ù<¸t³-DH!1c¼…ì ûœò~õ*ykÙ5—{˦Š*P Öæ°Q@ƒm tW¹e^É´Ž„t*4^s`ò¹ÊënÿÒ~¶ÅС€Ô¾ê:y˜¶{º8­ËO4ÉùÏ|ŠRýRWå'q­:ßì“`GÓܪ¥qÃ÷űºaBgÏy›÷2|´9¹Iœc*Ûñù l´å‚ã$$¡‘Tö3wàùêÞ¹]¢À­ͨWŒƒ}’žk*Ïh˜$„¢$­C3@OD–¦ˆ€Vˆ>JÃUée€PnЦ&7ÜñJ~z<ÌpìÀé5ãpì‹ç±¶ö³ØÿÄø‡þ1ôZ‹“Qäaè¦:áâ!,\Ÿað6~hb‹ÐóûY–Ü€Ì|ˆR¼ø§ád[TL Ê;ŒÒOgäӳݕ j¢tÆ«)ˆ('Òl–ÈÖ¤sy2¥t#d?Ha͇×ñ(ê/Áƒ­ˆI+íë,uwG)}tŠ²Âop&JÜlbaŒxœf»aaön8æÈù;Àzìk8s®tàk†¢_c'L|LŽŽ3: ½S^´M·v@¬à©#·¢øØÚµ}uì«` ZКè?If[•§«[cûó#»[6}¦£Ý¾,Où}YçíÓÀò{rà Fh§;Âo8Â_î¼!.Ø’ÍnU¢ü(M–>t†I¶a6†å,qF ´ äý Ëqç‰;'F9G†„@°Ÿ {îy… ³&;=`xl¹Ý5'Yãð,yÕæhíOwK9…²› Ö±_ñÙ¯cÁ#·ÿʇ ȼš|ÐEü7íûº0Öþ°úÆ©Ÿ.=ùïöö@ˆtDúƒ/R0CHÂîHz| ‘çŽôb›òôÅmԸͥzmŸsÈêaÀ‰fò*`¼Št2OÈØ+ü}"êÀòº¸h`%E##h¼Ê2nŒ!…†G÷“p@ºipàsàÈPW„ž^Öm…ÙU?q¡ûfµþýš€Lè'z´¿…‹,ÊQ”ç¸Ó\„ôœ±ñÿ࣠ÂGꬽÌà ñÓ$½ÈéGð2˜+úŒùAÿ”¦¤qI™gnýv¶ÁJ |y ]òg[LÒK_1Ûï^¹çÂ2’á„ÿïFþp+h(¿—«jõ,†#C¬ó›}! –Ù_¹Çå¦tÔyFS ²Í\Ѭº¡Sâ`…¦\‘>ÇQOdÐ…­è "ÈËõ"ìdZCóµYÕ 0b>ŽÂ+7†>$Ãáü­JGB´ÆP$£ +ô•§ŒróPûJ9DÖ£«œÉ0 µo@Íænèm}Å!ŽæHÐV’ÿmpFƒP~í&€¦Â M}á&éüÎ5?‚$­ä¢Œî.Ï83S^`A³ÉNŸÀ]—~†ÙTs ÌöOÀþߌÎBpÎ'\ ðß Gà )bQýäù(ëÖcUCs2¯Ÿ#žFBþPîßš'ôîüÕd…øx`A=7˜ç˜òWçÓZ „[Ô -LGLªè”ŽI| Ð;4òÌÍ”bÿ¾t¹uÇ,ƒ»]ù_+ÓÝ sºd8/’É:š q‚ædûUie¼9?iº¨¦~œLÀîi²¡Œ&–·ˆSRÂ)tH-L–œ^@dõÿ@>x¨ú¸/¤€ñ@/äº|¦~' i¢À8ŽùÑ‚d¬$$^xËF®UìDZ^Êk2p•‹~åá‰)¬WÐ@ YC•·Üç«Ds¢[®w9Þdð&ƒ¶/(ú)ç)GR~hàc¥6¸84MV Ö^­ÉD<“[eçtÄåœ2À6AZ½+¤ Ø¿™`•¼Ö R¼4^U…ˆ C#†‰ÔÄæýÐÂæ²®7çüç^€lÌ’Œt©)ðTøUFQ üÄĈy|œµ½#%jŒM]­ê*xÀ$Ã* Jèæ.kPäɲ€c.ö±$7©fã²íwÜkð]µ{¾Ž7áH:M–-ºgiÚLUIï“Ž/‹Fò\$ÏDÉê{Ë ÕÝÐóðˆnEÇ€äü ­ÁÂ'Q'—£ø™Ì®PaMðh[9r®0ˆek™Ì,t’ÈÄ ø;º7Ž•ÝíŠ °“à MEsé*ÑÎŒ`aÍ4ùQ@;YBkD%aÂiâƒm«¢&‘‰ ~çg#éÈÛfopÿòäŽîø[K}‘÷›å¼Ÿ{ëiÛµ4O¢›€q8cµ¤ö÷© –¹ð¥A3ô÷“ÜñT%ÄRŠ(¢ž‘W°ÚŽÊp™ËÅç{†+«Y(óä4¨8M*­³`Ú”ý<ô/á@ø¬:'A€ƒÅyñ0Ø]À²k,0~dF§…·øøúçÝÛü¸{½ZÓ1¾‰ÂÉÀ¥‚úY–‰ÈÍ´+²æä,žÌc,O"YJRJ]–¤Ö¸a™lÄÉ׸‚‹ùvm;p‚ T_ï¯î‡ß¼Sûe¶_Os%·½V4ošÄ‹}Eèiì«T?·†ÕȇF‹²OÆ„œ«7{np1˜ ;v{18¢íL­FüȳÓY‘˜~Ë0Ó"›žçÍ_™I ð÷œN 1¼¸èç¨Äû÷ª¸"ßÄcž³o›Õ(e0Õ÷¼åa¨)¦ÚYÄÉΈšBFMÛý4Õ i¡;ÌùÇ®D² h«EL¹¸è•rj”}£6?Íô¼Èȇ2/ñt:_›e.Ÿ ýàB¢wïVÓ2˜µLxEœÄ tù®Ö™T¨À¢K/ab?M’¥—ø•Æ£dæ®"W<Þ’~A:…Ð5ŽáñU¸d™¤½*Ò,ò£P­n«|³tÁ,V• ŒUM™×b|€-~3h Ù ôƒ$Ü'gÚ¬Ž‰ÎßO/žJ‘d“‘‘E3™eRH—ä;Ë(¼¦^¡sÍýÇ1¸óL?îýGµ¿"¶,M÷„ýåž.".xEh”÷öÀôÏî=±hô?MÚ­„ endstream endobj 1360 0 obj << /Length 3363 /Filter /FlateDecode >> stream xÚZY“Û6~÷¯Ð#§Ê¢‰ƒ±yò:ãd¶ìØk+åMeS*†¤$n(R!)Ë“_¿ÝhðÒ€žqjª†¸èãëFClÀ[é`¥„ðµŒWéñY`Z›ýŠ ~xÆì¸5 \OFþsóìÅë0\±À×f«ÍnºÔ&[ýêiyóÛæ_+¡¸¬bZ_¿Þ|ºY !<áÃyv»è„œ?qC8òñE±ô™D»;ä@ZjoŸWyS¤HæDÓ9aã‡ÝîvÝÅŽš­Ì#_ñaйJ»¢®hí®¦oþˆ1/=w7ÌCºa0n k’ªÝÕÍ‘ª—¢;Ј㠽sÙ§²H‹k÷4æ¿A¸6ÌCîÇá°ãC}9&Õ½kÓ2ð±dÂzŒVNªŒˆ·]SÀf³Ün-TèÇX­™ðYH³NI“ó.oÚç7ë0°;®¼ú|³–*ð.¸ßú\fT­j¬_°Â¼3¬Ü"Grê4¼ÀB’}¾ #ØEšg4´¨pÙ%iî<³Š|é©”¶§2©¶xôm¶ë\ÇgAìóAh(&¤ÞžnÖ@,O‹Ý½=6¨Lk{lÖ«NÝ¢ìD4aK„ÂÛúØBKŤ±m°:ú€ÇÄš‘ºS¢ÌžÎND@"FvÎ# q?ó9’QÞåš{8Õ.ùky¿Ñ²´Ç¥ªù©ë™jÑ™ÑrrvêÀ UZž³¢ÚSé8NªNçîE}îàC-IÓ$¨.÷­U>hÛ5õq6Q6úfS€9,ö#Íi;wÕú@%Ö\‡£ µTN›oYWû¼¡2j’BM:5yg§òÂöö{ÃrÒìÏÇœ϶ØuÛ´Iº×9؆ì™p jÖŠ¼¢–oòaó£á$•KÔ,ôåh’ˆÛ»Ÿ¶ïß¼|uë’¯Š}ÉõhÁ,Lö „¡ˆ½_nbiìPD°Ãâx*7"YÛR‰ŒÝ CÛi”¿ ŽúwuCÝ¿S·§é0žy$V¨~D²ç§Q' ÀÌçÚ4e1ÚŽb¡Q ÷Mþ¹¨Ï-ì8¨µRF¸˜,ú€yJùZ‡=;j„¯,w"ŸôØŒûu™÷Ùdµ¬h0‡p;5É%! j€Ù{±ý0¨]†«¦µZ8È“]ÛœÕp!åDÿ#¤¶`‚º'•W uJc8÷4P #ª’H8Ðú¹H†vboØc/¶Ì°×Šœàhfí4±6óPåóφ TãÂð­`¡P ÖŒµ4i‘œT Ðö¼?Ð<"ŽCÚƒu-P¾ºlOþç9)ûáän±œÖM“[VÔUf´vßï¦×ÔàŠý#º÷({ Cˆ÷.ýÔbfË·7wo_nܶ¬ªõc:€ËšãÃÛÛ—þà^’ùb K–$÷xI ÔbäÀA °éh"–Ùdà{»AAÚˆb⛈©_Î(5ÓÏxø@©«Š\Dx\ü A• p?¦-CmyƒpƦp£ŽIA%)Ûš¦¤uÕ‚Á7Ô~nq}#ðçQ8câû—›»ÛŸ6.&F1„vjêCÙ¦>(#Ø#Ͳ _ê"i_Â0T‰Aœ4¬93•¯UŠé~ò/i~êz‡™tä1‹¾eêÛ¥w¬ë‹-ÇW¦¾º!ÎagMß ?a@B—bGPÑù–Bc"$å¾n  9¶¾=&í{f9¯àÇz u×ìH×1v¥5ø¯ü U¾½Á`“‡m—'µe'Šk?”á4zª²mÚäI—›Ñ7…°I5 Ø¥ËåD#©éâ|X¼£˜3ýPý×°Õ $¹áß„¤xÂy Œõ4ÔGÆè„ÊÒF¤¨(ÈJ…ÆmêÒâÈ»âf]9½l¨|>j¼%xOW¢3¹`õÜ:-®"÷ q=©dñã²[Ã(¸Î0Çê[ž¹hÀÆ•’ ¤sÏ[á^•ù‘§"òû8a‘²"_‘‚û_èº.<o–äÝŽˆL„%P:NŸAœ4?jå<àZhGT3 ë±†¾6PS¦èyðò¹»·&˜³×Œ¼KÑ%'Ï©oئ¨Q§ Vý=ÓÞ»ÁQe4rvÕ믭ÅEŒ‡°wÆßzÓœ^Ãz<ìˆáZ÷ ”X‡Ü:ݤ¬÷g<3¶‡ûA­‘ƒ=6f?Oùhô‹a­œ/ü;çÃÐÅ[,WÏÜ߇ۗo¶›wÛWïÞ¾sûç^MÀó"‚Êr@š+K6$ìÒH©9I ÿG ÏŠ&7ÙT5¥M„ëRú®â¡4¼Ù2·}‡¾S6Ãþ(Ü@=5gÿñ•Sà­Tü ?Ž>,ï~|Å?8ýÇâѵvá“ŸÔ>ú›ð)šâ“ÈL¸f¨ÜƒÕf—Kjr, ý`~ÿî­k{1XLͳD¥°;­›¬¿ƒaÎ'Þе˜ÂÅœþ;T)5Ì3'¤ 9ø1)¯ä,ljæR*Á<¿À7) :P°Ò ™½è"ˆþb’Pæ’˼tÌF´…ÍH%hÈØB)R(X–C‰Ó4"½.ó¶í3(Æ(†ðçÂä¸ÆÄŠ)§IY.¤(çys£u»ºÙçÝö²¨,ˆ¦ˆ•ìº>O‡ywÌ{És|°‰1±½îyx)°„}«‡’ OÏq§*Žtu¯›îEþÅ&Ù›Žö¶M5Â,”Ǻíh>^ωPÜ?Há;e’n¦”ܘqˆ ü¶«›do ˆã›ËÑ$I™ ¤-´ÙKÜûÌôô¹gÛÓçb„BVþÓñRl ó'åuÔ<.Fï' ß á2‚CÈs‘ßk ‰Ø˜5Δþê6%Vã[6eÕ¬” ñçÖ>6<”—ù¥€u}rAL·6Ej?†ÀEøº÷zΟ/¬ûkˆa¢~GŸzÉl >ÄÞâã"¶çÕ¡3â!WjÊI–aZ±³¯ fÃòcQŽ¿¯‰íÃ|­V@IüôE ï‚ @„Çnã °a´·‡zÞäkãéð—ƒÜ*&úìq‹d§µàÓoŸÈÜ!|4vj[üelò}Òdä‚°n^bëµÌP›,Á2ħ¾ùÉ‚ Õ‚Þ©a Ñ=`³ÚÙ_fý9üäJÿ"˜ÈüÞê×ß‚Uípf_@äs1£Ž l ŸÊÕÇgÿ¦ßdM5!as±Ò‡h¦ |C¼¡¦ÎünskáÇA$ÝÄ‘y¸ÚË0†ÈsûЪ½KE•ßÏ¿ô€ ÃPS4„« ÅÐòÃO?S¡,¨3¥êÝ‹w´Æ¨ØL©h>ÔM3˜g¢)hhK"Mµ=l7Úô`˜4Ij<³¥ðœ'§Sbü Øuy?¨NàËhÈH4ýc=¢ÉºMvøð¦é÷`!¦¥¬¥äÆ;‚šZƒ•&“¯éªo¸ôjÏíédU;½'˜@ÎÏÊþÿ!Ã]~ endstream endobj 1367 0 obj << /Length 3100 /Filter /FlateDecode >> stream xÚYÝ“Û¶÷_qÓë:C‚ HÖ4ɵî$íLrž<Ø™+„NŒ)R!HŸ/}÷ ©£l·£ `±À~ýˆ®BøEWyx•Æqëìª<¼‰Ú?\q㧿¿ˆdÜng#ÿvûâë›$¹ŠÂ óèêv7gu[]½Ý|»/Žƒí¯·qoò¿\oµN6oŽ}QÕí“w}wàÖÍÍí/Üúp­Òí]ݵLP×Û(Ë£t“'׿ÞþóÅ÷·“X‰R_(?ŽüüL¦ƒ(Ö¼®mžX‚ÑÙ [zóX{¦ {Ë 7ôue÷Û¡ p§z³ëdëûºÄ-ɼzòµÊ6…L+{[ ¶’©oZ˜<Œ-@‚,Þ|å—¬·¶he* lsÅA”°ð1ôæÑfO+@cèð?”•˜ÖÚGnfMÑŒt%œû¦¿Ž6Å0m臮—ùÝn°­°ßÒEÂ@Å„waÖgÑFÁžt˜ÐÒ-¬&1ßEÄÛ°ÁƒJ×ÒyÃR}J·ƒF׸£Ûñ¿«ÿ¡Q¨4·ŠÆuÜ’£æÇÚUL…áØF i…z¼Þ‚Â:Ô©,t¹¬úŠÿïÇAuÁAwk{kºöÁûGGÜËrìœWLã³@±G9<•r•ö›«Hl•@aÝ¥ïÆ¡†³†`ðfnð‘I‚D£‹ÿÝnx¼«G°½;÷ä{¸“óá¹ g‰´ R5ÍD˜;IA)¦QĶ€ZÐþ¶q¤eÔÂpÝP´UÑWâ\$Çö<ŒÍ½éˆ?Ùf1Ô¨ÿ6`SP±† žJ+fû£»EwNT0Š ¶ ʬl>"%±ì5دNÅÙ¨ƒ¶“([@Z²émÙ¶­x/ƒj…e0â!ñ ÇwcÏ3Š¾/ð丗ÂËŠ¦TY/4…ÆÙ•k›,ˆRí^žÊžKÄJsËÓti›`G *eEJtdKÛÙõÖ®É $:ñC__ã½Wðÿm7`#‡sü}¬{[aˆˆsv(@â£ÿ9ׇN;Ù?mn|(ÚÒŠô{ŒPØúùõßñ†£Ð€fÍÂꊲ´àêl ©0öC2yF‘ìvß9á=L± cÂnlKd 3‹rAä'þ²kGqGC&«e±'h)Ù,k+òŒû÷d_BCï^ $®8Ø)©¤¶¨YÞˆûÀ%®DçäJŒbÛ؉0 ­!òåó5¶HrmP8ä¬ÁѧuÖ‚Ž‚vªVLùnßuï×ì'7Jâ¹=¯¹ˆ ¢\?³Ç‹L!Ó§y>Ÿ©Na¦-n±Í¥²¥J\Fé@EÙ3—ÁàfÐbznù¨Ø[Hs¥tãɽtÜ>PTÚ.ü’`ôqà†g1Ö9æ„&ès¢þ£nc½ò˜TÕïÂ(̉¶ï»~»‚ {W D jJAeÊôæ7ÿÄíÊøC ƒ2ö ) w"ÀZG¬sŃõýSKغ[Sg¬U`Â)¾€÷ kZTêòxž…ˆý='ë~àU¼ß@ôAëz©¸cß(ÅÌ”gî.»ìd³—#oœAPÌü¸Þcß®o/ T6…Ľùá‡ÕÝ)hNìlMh6·{‹QÞ³åtïäÞ™„S¤¸ñ8%e>ÐI€ úS@+Ƹ!‚è<ë`Ž˜°¸864s‡|! ’ÐÈá§á‡w°¢©ŽO4¡DKÁÆo£¸UƒrhdÀ·WÕ;!›QºÅÂÒÉH>nÅÜëvYthò+à,xŸÏ#£/Ò­4R·«M§^EïmßÚæk²`ÕB²0ˆNisÂzçèe«¼àÀˆ¹¹Î5j±'° øB  YÃË—á˜86ÓÆA­©/g XË=b~ADj!¿8nÎ3ˆÙt÷CQËé ¤WlsÄ“:/ÙügÕB(¶–‰à‚‹›Øøa/E©Ã.±p2 î” Òt»N‚,šô÷òŸ+ú Œ&UKÃ÷‡—ŒuèA{²1.Â!™bW¶.kæÍÿsùtÙÅs0f:âê"o›øŒ7ú×D¸ƒiꉹû4s}ÆüÓŒñÔ'¥4_*¶h£²‚[Ñ(NÊ·–W?%e‚ÜD‹èp„øSãõÃÅJŒÜ Ï h˜>Øò©"ˆ5Öe˜?Ì(œ™ìÃ(,¦šEÕ‰`ì;™•Æ¢eîÓ*[6« JÙò¥O.0°³cŠº~LÊ]€JB¼XQŽ =‡>”ªØ?gŸþc_BÑ×¥%qñZ€Ùià—(°XĪ®Ú—·!±Õ¥Ð©’ÂFݺ¡K ÿ@ˆ1äp‘mÏt*BW£Q6?Û×Ú®ðþ"È£3ô×­úv@Þø²³J ÞŸ¢¡¿=2z{ÏWFΈ7‚YíзPî%`»Ïaºöù]ëŒÌS§æT° ~} ê‡ë›ÐÎ:v«ÁTÏ¡ý;˜¸·E…hWÃTÚøàø'^›FÅ„¦‡»(>¥$ a±ÿ ßÃlãh£Ø%ƒPHápé:Ú|×G£¾¶ÏfÄFaðBÌá9GÝ8¾4š_ëGQõVõ@qDWûA}AA#Þñ/)óË~?;c°rá¾ ‰ô³•$‚ž‘•}ãí/ÓMq,7¾ñ|oýUîZŽTãÒ“Gt¾-PEؘoëò‹…ý™Mœ3ýl]á¹ÎêŠÛ}qÚ}AñÃÉÒÌó E–fRÿfb¤NÆýhËQŽ4ÉOeu2z¡á(ÂÖ;à%D²ù†®á¡ëˆI£h[ºU1§ÂN–´C‰~X|µz5¶ïÆ33>2ðG)á^Aë ¤óZ0Á0kd )ÝÓdØPS˜ÈÒK6 €,š_E&FÔü¿š´Ÿ¾=Í_3†å"KƒF1“@³”oo§½ÀáÔ¹*G3V¹3¦ÏÌ8{JU:ßÉ'ÞÝdð§å>cynÃI¶À_Zyž§HBAÐ^(Òx%þEÞkè®ëY Ñ{¢÷úw¢*¬#£4ÈNyßÝþã§ï¿ùîçon¾_-ÿÌot¢¤xàµ8Ä*¼6a˜€mz³¾C¤š®Áå“ÿj.3¤rU ò,yŽíå)±ã¸v?U ð1ºâ¾‘èu*æcÌh)e´£ï$`ø[šŽõ.dEð«æSN–þ¿•ËÔåjY8ëå9-+׫Ø-M åeÃ'Ä-b+l‘tÌ}×:Ž@0…Â7ü/_*³Ó»ª¯œ« $ ç¸5x ¨n)ëK„é—ëœ >l!o©+ÁŠcHlæk›L¨ôHXÍèÔác†K¨køD É“=—fã§c{z…fcÏÅáŸúšf‰ ÙÉšOæfv+AðßáMÉ];\ÎtQAT1³jÕD™+‰éO ¥c(¸õ—Yi³Tfý‡LIý#,|äRŒ#™ª L^ˆ’™Vßkã É—on"ÖJ´º«Paœ.ùÝ «F>=Æö¶¹¼øn,K~…p¯|6ø/p£µj endstream endobj 1371 0 obj << /Length 627 /Filter /FlateDecode >> stream xÚTmoÓ0þÞ_‘ŽDŒÛI,@Œ XHÛ4…Õm"²¤r]Fùõœ}N—V•†"%çóã»{î‡' žh–”BP-«äî~Á‚×®4¾¾[ðˆË˜Í¯ëÅó ¥Ψfš'õjª^&WDéMý1eN V@”ླྀ¨¿§™‚ ‡,ÞÖ‡<*Ïÿ³ |º¢¢’” ‰¹]k u•“Á< aÇë†èí¶ø½kúÞ,}e°˜ä9”¬äÄeµr·Ýй[×ZÓ,·ñÌQœqªÄt¤–1³q;;ÄŒÑü®š®ßYCÓLiF.9m”,UJúfQ{”ixñ›8ÓÂÓ4Ä›¶Ù8c±R¦ fŒÇå qWŸv½ë2lŽYâؽnž¡½iÖ¾÷B©O æ²ôU$E ÕTò 5Lèl?Cç4h(“"®b®¡YQÄuk,”(eä!¥&Èë­Ù4¶q½“ÂVëùZ´¯—}uúðxK¼€ ³‚€ä¦8)'¾Ú w®ã¡ÝèÒ¼"ûMšÁÇÄl} F´,΄#£—^µ‚¶¯Î‰UV¥šÄ êÓ¹$ïÇ4/ɃÏl~‡·…ù)˜ßÞoŒ;Ȯٺ.³m< ±Þ#‰q§ï†_h¡G’&ž~ìf€ý´Ýãâš)vŽX^)Ê«ÃÕÊúÀìpƒÂ²¨NQÐfzÔ ÄG\¶­x5e—À߆îÅ£Š›-j;ä{0÷zž~?Ÿ`øúºaûH†¾Çµgkèšú‘Jòåóå‡ós1êi)tº>ÿçNh¿ endstream endobj 1378 0 obj << /Length 2829 /Filter /FlateDecode >> stream xÚYmÛ¸þž_±È—Ê@̈%J—¶ÀÕí¦[\¯E³APÜÝ­,ÛBdɧ—l¶¿¾óFYòÊ—m±ÀŠ’Ãápæ™Zßøð§oRÿƆ¡JMr“_ùDm÷7Üø×ûWZæ­aâz2óO÷¯ÞÞFÑöUê§úæ~7eu¿½ùÉÛ²S_´«u†žö¿Ã†ñîê®Ïª*ë˦桬Þrc3t}s,ÿ#c:üÐKíê—û¿½úËý(J/”g>Ú‚ÐV:2"tœ¥CÃBkµŽâàRL‰˜Ð˜‹ Â]*4…{jË,ïe+ÓÐËWõ@++í±bRãm‹.oˇÕ:H¼Bæõ‡‚å\ŠÔ) ÖåseáX³ãïííý§73NÆE×3ñË ÷j»éB8Ê­C¥#ûñPæ8ñÀ÷vÌ°óÄ7%ârgÛàÈc]5Ù¶Øòì]Û/ï-0‘ÒIp‡Vù±¥»SŒ)?…: KœS¦·éÖ¯' Ìðr: ë!dÝ°ˆ‡æ(´S¶/.… #­“΄½nhnö7D»dJ¢)Òý:ˆ…E§·!ÝÁz§L¦¬óòTp³ÆørìÍPm± ·€ÖàìÏ<؇¬–ëÃYÝS×Gx,ûÃ8…F¿ÿñÃS6Lț㩬Ð`‘ø³ùbõñìpÜ5B’ïó|É5­,ø¢LB¿`–ÚÔ0Ö¢Œ0T%bšgÿ•ílÅ–§¿A[ô½S•Õ5ÊÄÖëK¼[láøݶY×—98ÓuDl¶ÐriòšYÅØÊQ7Ø(¾žª¦ì¹“ñçµ[’*keÓü)¯ Þ&oÒ<øû»™“É¡d¸5Ð2úÄr'DDNë &wÃéD.Ø´=/Ù52'äƱ¡[wN[Ï=µE^vg íÉ$ø vqéŸÌÛ¹‰§¬…X7T™‹ -ßYÙƒ®ÀÄ»%±Ñ3YÊÈXïÃÝßÿ,nBDÛκ\üM2 ZqäÔ2ÚçE‡ñ'´1ÇsƒëÀå b Fwi¢@XAè¸LÚœ‰ôè‘ »ÉE!5ÕK¾õ7øo[TEß-!µñU0æ—¬1 åÞB‡ö@¹ÔEmVw0L† 8éE™„$I„Á¼–_¤ƃ|UÇy)ž‹Ö€Tw8Ú„ŠäÐüA–Èd)A€¾-»¾-×àYò©IƒAÒ1±l…jâ@4¤ ¸¹ªUéku«û’‹.W«:…&™&¤ad\B‚rBŠÄl)”jß(ß7/ ¥Q¤ ¦L<œußfGæî‰_ÖlK)€õ¨Öª@tòíå)a*™&N=—¦(m'ë)>²žu”*Š›»\÷ÝbŽ¨D¼I²3õvvºK=fŸ¯Ó×Qð¿”eL¶ ço© ¶†²ÚrSÊ{ÒºœB(Ù=úiƒçþÊy:H½¶È„ÕèܭʇVÀŠT FJä/pÞ˜o<Ça¹QÀ¢µ……U>!öŒžښ㱨 {ð¬?ÏX)$c`éjÂXY=ÚœSü³|Ôª$ Æ‚sÇ<Ë^¾Ýqc‡¸ïರ©%6Ð[µ$¼IÍn oçW¼Ã„s3übÍ ˜±pH@c=ʶ¿z„DÅᘧr¾¸†…`oñ³ÍÒ&P& g*âdõº Æñt³×r¸2¶ždÌ_ätý(ä1=¾>Kÿ‰ÞPîtŒûí„wÆŸg™'Š½p‹-ˆ“÷M‹L¬qþdÃkÕIªU˜Œ¶±^Cn±[Æo“ª$ÜL*Êö¼ äžKï4Fÿ6Tk°î gÊ£8öþ½J kvaä…§Ht&Ôà“¤Ð¿LZ ¶ÍŸ¾m¢ÎÁý5ïðÀv°ãäehŸäü€·û¬¬ßqáÚ©”ºcêL9m­< îZåPáî6“Rü¬¨µ·Mý;¡_Ü Ž¢”Cƒç o¤*¤J¯Š0hd=Fònzæ朼-¾”ÍÐIõ?+—ð7Ž¢Æ©_ÎÐï¯%¿/¡óÐ ‡‹X‚–·· …{¼‡OgãÖø³›Ä.Ê"ÿO×#“¹â~–v4¯Ç¶ìÅYb°F`w4‚ð©ÿòu¿8Mç^‡oçÿ‚,û¬ endstream endobj 1392 0 obj << /Length 3696 /Filter /FlateDecode >> stream xÚµZmoã6þ¾¿"ßêkU|•ÔÃ}è¶Ýb»½ÅmÅa[ô[N„Ú’+ÙÍå~ýÍp†¥Ðq’^ ¦F9CÎ<óBŠ‹þÄE‘^dJ%…Î/V»W©£v7ÔøÇ÷¯÷[BÇeÐóÍÕ«/ßs!Ò¤H qqµ ‡ºZ_|^ùåÏW»P™LljaG}ûöêÇË¥Rj¡øÃ.¯¾»æ1R>‘!ìyž#›ëD(Ms—ͦ–Åâp[5ÈDºXµ»}½­è¡ìéõ±?–Ûär©µZ¼ÛлûK™/Ú#uX·Dü)ÚYòˆíþPïÊ-ûêÀÔ J ÚA ÜåRûµùæíû¯¿ÿÄ'’¨,)LîûmÚŽ†¦2`ŠŸúûþPí^ã„b±ßVe_¹€—2ÉrX.¡ahŒ­ã,Õ ,þšÅ¯M‹"Þõî¶<ýp{)°@fs¯Ê®¢ÆO©IËmÛÜð'5tåO¸G{<ì<ωÐlà «¶ÙÔ7ÉͱêûØ:Ø<ÉäÐ4xßú™Kžî…©ˆºr{ĺYmk¿*"OlæWEÐÀúŽÔ³vû©›ãá+à„†Ç®r Û'1iÀFdš(©CiŽ]ŘD…EéûýåU¬;à†hP¨í–´œ&]ôÔ`;°&èŸ)o¸ÿºrêX­éñàžúÑ@üþ‡Rã›ÖͶ®ÝîᎠ÷ÁO2ÐÜjü6"ê’ùuêÝŸ®¾~ÿ>&­ÖIªƒÚåu¶ÒP@»g¨ …qúo¥†I ‡ ‰wIZÀ?!DL)!Vøï—Á¼˜Oã5‚Ø¢µNIµ®‰º/oª9»*Í™ª »§Ì÷>ÃÜ|PÇ€”QrñC{`&«žÔ¨ZÕ ?÷±Ì) ¿ËåmµÝG!'OÀB|¿CKãoëþ@«@ê ¤Q ‘ Ó¥Eb„˜ÛrY5åõ¶Zö·*ëy‘X­&\h¹XuUé$€¦oõ´Yth°c½8hè 0C½â^Ûúº+»ú„K)•ª)$6b9¬Q¹í™¿r…x¹ªödË@`ȱ6‘€_“UØTw„8hËqÿÐHV:¸Ä(®²X¼¦îû²‘ŽÛ²ÛÞEÌY=Y;@V3E$(‹É,ÀG`4ìËf۞ƶL,¾ÿWà*¡ƒÊqU¡Ü8¦Çv 470꾫Vž@oÚ†úýî,«ëB»¡_h¡Ó‰ïZ‘H‘ .òç@_Í=Œ[7à<ËõtZÆ>Ø8°†ˆíÖÕ¦z uùyPÞ›ŠZRÌÑ`U& 9 Ÿ¾÷¶çƒ2|F̼žM2V ¿*{‘å`X´<½¯BZ”I=7 )4 ü ‡‚^Š†ï'†ƒÔ`ü%ÃÁÖIÃø¼Ðƒá¸xŽt>Ï]Ì ûŒªª™aå¡aålX¹`§ шðá¡åX|*mŠ'‘š‘ÉóÅU<´Q6±#ªŸÃÿ,•C·êê=ËrWƒdЗàúÌtoKZ«‚Ãf­S—A ¥ê: ð‘´ƒ8ØY Òëh-4@ÓtûDŠÑ¦ëž¦òq>¶ûrç[õyö²Í^XPù„=(@öaM[–tÌ_ðiW.yÅ )ȳi¾²B ¿­›êKNÙ:‡„&Š„i0W\™Tñ$äï—Á1$œMs «§C  B؉DÜû ¿óA@i³DdâAàoÇrØr •Ó9@AS¤Œ‚Ù€‚©u(?8bwÜO!_üÎÉžpàShHþBFBhûÚcœç Ô4‡MZèEÓ6aÆ ¥Ubíýò‹‹:„Ì£V3åâ\vƱM·ï\þ{½ÆN …)\¶ˆÇðäÃ`7«U4Ô £:éÄÒ<[`ô-„Г"ý'·¯Ñ5¶–Vóõ  Âñ ”½ƒ5ª5K‰Å6ÔØ€fCHmG/ %9'ììVƒp°Þ¡#ð¡"2‰ïOúƒR ýBw ÓÓîü£‘Or:äInfX?¸•“[p-„Ûr;]íZTFʽP*C'€kÐ3^dè\ò¢ÐðHO¸dØ­iñ¢À€ÌÎ)Ošävè4 Ìa§MØðL²61‰`<çPLƒæ>,=Ã8ʇ=ø¢’v ÇC¸ØA6µF¦”Þsá§'ÄnPçÔFõ…„·ZŸ¿ÑiH¬Á“%)èásŽÿ~ ñÙ4ÿG§c`¿$d%¡Ô_¹÷~çƒ>âtt ¨f½ÓÉ_ät·ÀL´vðl9æ¬i±øÎ};.­b3*i)‘à ÆÒjÊ] o5Á˜U cÖÛD"6܃×.gÌEhhýC…JÁÊ™ `3Ö>[¡ü÷Ë`€Èͧ¡Jôm¹'{\4ôóù@tÍ+쪌@D‘nA›u VU­”2Ï'Â<Ðpï3¬Ïwvw[“7"®öXÆmѧ¢í‰é’~úz·wIh× ~âoÊ thÙÊ"±ÈXˆïû²ClØ¢§§Ö•M¿i»]OÏ®XO>Rƒêô=B‰0‹7÷DeˆþÒJÖ)nÐ$xèÚã¡n|FÍDWÇâ½k°;XÇÍŒ-ç„ví¾jvѪan ™Ff­]¼¯¥-æ]±‚ö|;¾À%‹k×rÀúï—Á±ÊÓlšÀ³f,d8^à€®„”׃.¶ t1Ä~º£Y\(ð½¹<º¾÷2èËzgƒN¢hÀ³¡bâA)rçNÞ°qì™Ëk—ÛµCT6­&L&ΣMž˜1§:qö‘¬Uz•×<°?¢lÚ$z 2kÐäuá‚€{j1ú~É1#‡#z䢴ÖXçWUW‘Ó!bSîðpbY@ºó†‡ýƒQÊWkœ ÌÓD„ öa¸G®Ãy=x*ýqfÓrûÏó(놿†é×ûÃr@¦ ’qo”RÈxù¸Ïò%¡(1)…嬬Â%¢n:ÎTµïa,®¬} ?ªºÚÝ—âÛZÐ- è@: „¦¢Ãm1:a‡áb®ÏÅZtPélÆ{R«Âc@°)W=bê¬4$.êEúl›¡váéô„Ž.³tqË×/ðÕ´žÅøÒŠÅáÅ`Ž³ £ I†ÆŒ«‹oêž~ýYê ï2œøL¿÷‘¦1þ÷A]¡¾üšÔIœAkPEÙßLãX—ñ…¿8ÈöžyÖ¢ 0xÜÒëcC4„Ћ¯w?4 æ¨èeÓÒï BOAè0XçïªßŽuçNŒ%°ôA¾£w[ÌŸy}4ߊe¼Bû ‘g˜çç»îëeðy,ÛL1 Å Ô¶ôóùÛº?t5d`â»j׺Í}[\²ƒ¸X—ÃgÁÐ2n·© VL$|$ ¦Îgš 9I‹b’#sc¹÷ìròJˆÕü–F˜ÐªÌòí3h¬ckt05´ÖƒÏSý‡¼b®Ø»„Z¨ oø»Ñë)âï—Á±˜y6ÍâM—ìÆ ’$84ÞF7/ä鈚°‰„¸$”ê‘tî}F†ù  CDªðzæK y Ut÷â'мhÔ–…|0!R±ø–>ì±æ ÈEQ½ÚZŠ:¼­nÊÕýò­KRhG¿ëJÛ ð‹1`Æ÷#®yj_ŒXã秋ûL +9ý|þs|wņþäÜf‡ul½wÒó{Œäh P§LêsI7höº2ðÅDÖÇ°‹zŸ‘m>hpu¹ Ð•¯/"¡çKŠ„ [x¤1KÂð¯']ÝV½+ç»-¨Æ)Âô(MQy:w}Ö=ñuV1‰Çñ™DZËpÙ2Ås*¶µ¤Ÿ¦ºÙÖ7õõ–;—»öHqÃëhÈ ‹Ñ9g”o‚"º!iŒ»ðÉOŠíá ð3~G—FJ~ê2§t¶mžvå!KÌXñg[88݆æ™Jb!Ð2(I^1=ÚO¯,Ebtή™Já› Ô3Æ+gÁÔuŒ¨7÷Ô½ô—3ý9"‡Ü8Â05 šOÏýòDwV7í¥ÍúEAERvîË>zLª.ƃµ·Yö×ØäRbó sƒ©YHÿç>»§2 8Œ§í²ï'‘eDøß`ë endstream endobj 1398 0 obj << /Length 3329 /Filter /FlateDecode >> stream xÚµÙrÜ6òÝ_¡ÊK8µ&C\<’ʃ­ØŽ·6‰7RŽª$É™a™CÎò¢|ýv£^‚bÅ›-WyÀFhôÝ ±‹þ±‹4¼ˆ…R™\ä§g¡†v‡ |ÿæ3x> ú Ì—×Ï>{­Ô ƒ4LÙÅõ~¹Õuqñ‹wyÌÎCÙí|!„ÇÂÏw¾”Ê{ÛôCV×ÙPµ MeMAƒË±ÚSõ‡™c©H¹—¦»ß®ÿùìÕõDŠâü‰4#æ¢#¹":Iu%2`Bῆ,ÆCá‚Ñ—ñ ‚+‡„åûwÕpôqìßuÙù\v½Y´â “ a±]…\`Ò{y¿óy̽¢Ügc=<‡¯DzwDz!øë]"¼¶ºÌðˆóΈèáQšu% ª&¯Ç¢,ÌFñœ—Ík€¼ ßä30ET•§sÝî È’*¡õRÅ^]5ï«æ@Ф8·øÊcÕf»ß,]]–ŸÎU ÓE9”ùP4w³9ÜÁ}¡€î™ý@̾:Œ]ébºo‘é’Œ–ôyW‡`çG‚yo^‘*8­êq-èC¸…¾ùö§.¨(åD hSøqÀn‘ž£ì™wÔœ¾Ó\¼E.–™Pþ´M²Ü&N‚8šTh¸k]gqHÅ, ©ëI@x‘vºpI«1é‚2bçEÖºæŸU2Ðá˜)¯mPÒ\ìÀÅ#°­ˆÿ5}ÚÓ–UQµý}“wà r¥Ní‹@OÓZÿ`å¤ïMAŠ °±5Ã[ŒMwÎÛ®ìiõÐã`*HµqTE©O*p Ö´”ˆÉoIä¬fr#ÃdÚÎŒ@½À§ !”¿6wJñ 0ÂÔøÒbN_D;vø=f]Aþ·33ù}^ì¼)±¡Ï’ÜÆ@ÝCÝêM¦ýØ{³{㺛o(]º‚C9àÅÚ}‘Ý».ó@Å“Õ¶N¹Áv*žt=¯Ûü½Ó…C& ¢¥O "Ô"%¼CWfF4ü0¤/3kUtí”5y‰ŽÒDP+­ÓˆÒÛ±6ãCÙ”pÇl‡¾Á„vÄÉ&HPtB $¤+‘TD…c2ìéÛºUV´Í§ÁÙÒe#$£#ID'I‚'%Àˆ8šÕý £'€D±°1™¤o6m«=ƒÀª?¬5#è\SÚ†ý®<€zÙÊþ„© ´mƒèáš,Àh*½«²ÜÖY\BŠ ù› ã ¥ÁZ+P0„Æuѵ†,«/»Þ_là(·Çhò¯ 6˜Ü ¢þ—KËW¨' éúßžìœJÂda¸½2ˆV·y¼r$ÜP¾ÞШ¶Ó ¥]ËG¹¡²Aoâ÷½ÓFPðE°qE±@HáÚ;­^B!?bÃìö÷¿•À¬ªÛ2wº]îD}pOù`Ϧl›_úsr+ÖÙÓ+½ hXb»Hl]k 9xü½zûÍWEû™Â,¥¨ ð««WA‡ónYŒ'šx«ÚÿX›’ˆ8#æRhôMÀ_] }Å$`q5¤¥âG§¦† ‡ˆýÕ£ s«<^;­;Uð?ïÐþ4AWð˜î äeWôÿRMÖúëìÐööV0ñçGL>Ê|ÞèUSáݽ»¤É7ÒÉU~³†=©pü·¯¾ûv>·ou¾éÅ÷ßøܵ¾­²“ Œ7LÈÁgCø3¾9JbÔ’çÔ62W“p5„e˜2ÇÂȇ0ZÔdøõH™Îh?h€ûºÍ×Íp{i²¸z˜™*oŽ†kò¹©ÓD Bg+“²4bœâ˜ÁÓï\6êOSßê1õ aP´ãž;Hä{2AøÔDY¼ª·›œÎ:Ș,oqàò„i³)@2ÈVJwœíG æä²2tEå±£¡“‚vE1€”*+4KX…"sU–©N`lBÖБeÖÕ•Î-àÀËw?ô®MVhH³}˜±n –þ>@–eZD€¢-nš­lÃPr¸ïFK³qhOØœ±y‹€ÂQç½ÅsŠÀY½Ã‚°ÇMêA½M½ü¨LM1¥E0Ÿu‡²#)8«ÒwU)ð?¨¢ '“Aƒ01 WÌ©ˆðÑ•ÿ«ÎÌdô³h ÂW?š*b- Ï`4s¹ msÛµ¿R'¢kÚ¯Ž@Ö_/ú4F(7Õ@ßN¨ìý=ö—Uèõ¥É> Ó£‰éؘE/û7û7Í™*nÚ´®z³+¥«Þ(@7ÍE} ¹Š­yldíý@INý²&$FÍp§PF~!럨Ç›ÀS5åê[ÂÒÉòüI­]ÅÄ\'àîºÝ—ÙÆ6ÖÌq¢ lœ¡Ú0Žk0×çNÄ)»©¾ü“)N!0OŽâlšŠ°ÿé0ƒ/†šLJ±‰=³8dÌL¢©—® ç´Äõ¾ôØåøœ@UC ¶+tëÂÂÇ °¸Ö–w:há¡uu0k¨vX?oö…Ãht#ð®êua°Ñô[jTBª“wY$(½² PÛ@†®ÂrAºå -èÚ +åÁ‰=Ô¬£îªÚXo Z@®Ìn¶øÜXESbŸÕ9Òû&Ë¿»"ÐÏ:½ÿ‹%H0ú1ï3™5àK`!ª$yŠ)°…)¶,™Á.tóˆÝzœH Æ£ ż„˜ÎLÅ óx¨ÆáÔy­5zÔšž¦aÉŸjÛžt ›—¡}sÃÛãÄÀ¥mç'àkÀDŸm€ÓS‚‘/Ó —fùôÚçj…CBœÎ­”GXͱʞ¼Re©,è(¨C¦®8—P—§+[Y{nÎæZsƒÁ‚E¥'pS–ôÄ(™éb töy€j1½5&xЇyHmçÙì`ÉÆjüïúGÞ½RÉ8;ZÕð÷‘í£Š¾tö©õíË®+ ïÿÞ¿iǦ@#u–»N5ׂQf ÄŒ§©ŽÝ—¡@ª Œ6Éè•i}È¿;¾,óL›—R›`ž‹¥¤|‹†·;…”W¥~<»_.TËà,Q[„)ZHÊû¾xùó q#vœ-ZzsDk3 ¨+‡¡<‡“×VóaÔ--n p®«kZ>8C¹‚Pþa›ËJò¨CO;ëÄÏ¢¦"e|8AÕ_–ö¸.Ø™×b©‰,HãÙPpCz8C¤UV0=B*¶Õ > stream xÚÅZmo¹þî_Áסø2’EpE.9\ #ö!m¯ÁA±ÖŽPY2¤ÕÙî¯ï3\­k[kKÖ*18\-ÅÎ<3ó )ërTFY—“²ÁÊ +—0ð†™¨þÞ)¶FÎøŸÊŒâPfF™13¨ÄtÀ6©l¬bçT&'Xåä1ï6˜ÊÊÚ0G9Ȥ éžeá‹2Â#9ykñubw€—ä3Lá,dJ¢ò]¼È¹¼ÅF¬E!Ã9–·xtdE>6ò=B rÐX8ŒXÔ•—1eù¢U.;nô†ËgIA™€Å0ð²/l†±kOe/.ce/ò2Ä °¦eWd”O,Úzk³B²A–ÍŠ<ÖflŒ<É|1J².$¥òÍ ˆÅP)Âx‹5²‘ÕÈbäE>EŒ‘ma©`e—âR=܈AqŽ‡¬à‹Wñ'PW¯Bbg™’ì!`‹S<¦„èe5|þÇjøD€w”½h€S¼Æ@†-¦“Ùbh8HY ¤ù,n`ÑË‹N03—lËYv¾˜d$³ë`'ÑD§¢-Vb£+€ÂctIV‰¤"Ù1¨Ênáш2*&#»HÀ|ÌòIÕ#«˜­ø>[•Œ Ö#V’kFN%_æë) ‰NÄc)’¨ÿ¥TÂ!“Š¨Žd[T‡Os‰FàeòEšW9” ÉFeöâ X!Gç^¿>œÜ\Vjðf:Õƒãå—º<ÿ:žþç`ðÓl>ªæêwƒ07Ÿ«ÓZýŽ¸Ñƒ}43P咶ؔOF‡È˜÷F½~­ÇjðËìd¦ïÔŸ°Òâ߆ÌìLþ~/F³‹WÇÃ?ÇÓsùàh2œ.þ¢~üñÿvT+%®Èkq6à #i©[­¢‡(ðóõål^÷Ó¡¤%WµÄÉvÒß_ô—÷"éhL8Ë‘E¬¶îùÂßÁšAhÔG5øÇ?ÿ¤ÿFM—“Éç§&!ékdÚÍ‘›t¢ïM<œMë¢ç!b?5ó‘,"’Só qmÛ1"ƒÚq©~õ$‘Y†Xwp4ŸW°½;Tƒ“êºVŸï›ùhx^ ÞB|5­²T‘.–]Ì–óÓjÑ¢òÙ‡j4þ4»nìÏóÙÁÄGÃ9¾:ÄE«û»ÛØòþ$¦ éfãDKìo%k;¬”·³’Ç'­dÍVz0Éf’6ö Ýš)¬ïa zÚ›v¸óF­oĹçm¤I[ .²ï™µ}Gsy÷dçl›Èn³^ê›ÑÐj"¥•HÚôañŠõ9¢ê?’=Žßx'¾NÆçSÀ³.ÓQI#gõÕH/³Ó~àDs¯ ü¦ÑÒœø·tœÎkÝzý²œ/E‡?±Àl~G)Aèâ #‹ñ«Eÿ@vQi…‘ý­“6 ±ÃÝ\ø·ÅcmHÁ·ß†—<˜„þ i•7O´)¡m [DË]ð?.ý¹ŒÏëQÑDÚnQáÝ*ȬÞ·êd½&äk6(øèŽKÁw(ø9tºúïÕÕ«a›£¾®N—uUÒärzZ#‹÷^Ȭ“í8§å¬ˆÑEzmÐ¥kFS浚ICl‘J ÄÚ—Q({mvÐLíAÏ(² G`h\í¾’Ygi›¨½? ðEòz~[z72÷Œä׃‘l`liøJù>•Ip$÷ŒÎÀ ,…)jŸºñÕ’ÙZ”<Œ÷Èf-²½œ4£ IËq[-=içé WWÃQ5j)÷žÈ½PKF‡´åa„§¼ }=o+V%4 Az‡n¤R²:˜¸ŸÀÆèÛœ<˜¼ÓááÁÉæ FØÇUKº÷(nÝ…Á<Ï…ww6œÍ=Pr£z±C½°;ÂBï²îQ4åÜyD'°J JËi½Gq¶aog—“êZÀ4]^|©æëp ¼œ6MÚµ•Ù=l:Ü“z '÷Ëd÷v{·=ÛCKÆPäPä–|²ïõÀ¦RÌ v|ʃ,€Ì°Uݹµnçn|‡è T\¹[DÅåìo«œà{¦[““RßJ5<«ê›µØ`Þ#*öûØÑsîZA!ö†ÉÑ•\K&§õMב-RÓŽ c·óÄÕ‘€áÊ&?y$°%ɉæ…HNì 9q[’“Ìš³ã­³[Âó^„áRÐ6¹Â3ÉU"2ÅDšIT¸ç*õ»ÿöñ}ó燯u}ù×ÁàêêJK·¢góó¾·ŒÎ(‰‘tBÛ!Ø (“Ñùî²xs:)=ÜÛÙæžï¡Ãe¤D“³ÝR†&B¸vÞ@nŠ·Ë$2¡êe‹wî ~ÙììÔž1¦¶²¥ö4%µ˜O-æS{³’o_µ'“©½†I;ÄŸË/p¬ÎžÊm®TDöåù¹qq>]JX .çÕå`Q£·¢W|­/&Lg£j0žŽªk-Ï=,sÖÆ'xZŽ¶ €#‡oyÙáLFk„X) ”Ãi`”nFì|SѨ…èÒäA¬å®-.çA„²áü7m½=„"‹üˆ‰l¹ž# :Oñ»KB®,¿¾  nÔ\5y€À†ô= ¶÷QôËÎçò뮤ù~fhÎ-Š…C^ôˆ†(W\@bìVâÝxQÏÇ_–u5zuQ]Ìæ7wo$¯Æõ×r#yô¾gp8¯Ë/˜BÒ™™ÔÜ{qã~ú$ ÂËX¦H“É°îÙÆH.—{@¹PŽ®üJ甞{Yûku><½yäx«­lûÒîGNw¯XT,÷ bùÎyÓÿ“|¹ endstream endobj 1405 0 obj << /Length 3586 /Filter /FlateDecode >> stream xÚkoÜÆñ»…€|Uè(.÷AýRGµSMmXr6 R긧#Ì#¯$Ïò¥èïÌÎðuZ)v @Ü].ggfç='Î"øgYt–Hf*=[ï^Dnµ½?£Á»ï_Þ·‚«ÙÎïn_\¾ÖúLDaeâìv3u[œýˆ(:_ÉØh¼~}ûÆR2„¿óŸoÿúâÕí\Çñb;¡‘*”ÊÄŒ†IU(¤Ðãó•6qp]w}^Uy_65`“è`xÖM½z_—ŸiÖ»Þî:Äò”F`‚ÈB3‰×ýù*΢ Û6‡ª ñ<’ÀÒ¤µxܧó8 luäÍ}›—÷[\ë7Mû€ƒ¼å¯û†žëf·/+†BìÑ%P5Íý8Ífè;øŒ¾›Tù¿øXÖ÷H#@†BþýOHÓ .׶/mG³fCÏœY`æ,H’P$rÖM½)ï­õqK«Ð¤é°µ[·å¾ÏW*΂ïò®\ÃeÏS\ ®IpDL›Ót.‰ð¯ dgài‹²ç•­õ!(Ò(Œ¢d‰a¸õ!¨Ò03bعµya[‚ýS¤£u³GŒŽ´§"WÒDa*5óQЧ›¶Ùùp‰u˜Åú•°¬}Ø9l„ÓÈ¢2Hµ¬U:RøMa7eí½%ÃÔ¨a£»n„úé”2oËæÐÑB³G¥àI^4`!ly†ÌØæm¾îm[v}¹î.ðM6}ÔÄ%w'fÉ¥Q¨ÑÀåã@1V°òo¯¸™0) ×>"pCD~KÀ~Š„ªP ñˆ²¦Åñ(ÐKP%­u§‡´Z”­]÷M ŠiªX ôÒз~±‹ã0–_$tY?:)ß„M^º«€ZTŠžvM…Ó… '$n±³=ވȋjŸ[Þ\[ÚXÖe_:­#èµ—, “Ih¿9Ô _>B$ÝIfç'ƒœÿT¤B Öbnt §ÈÁú°³D.j¸ý({D8 ¤·' .%b–4xµkºžFh¬ð¸Û­lò²u$âÛ;<âJ:X²ð-NÓ>†ˆÃ ³É‚Ð|Â]Ý%ù`V*ÂLœ¨?~®e°ÏÛ¾bÖhÝ0¨ò–ÞYª–¦hƒmxz“8LÄx`Õ€ vfÆ<®d‡Z™¥‹v.hO¢ß9K“ÀÉö±·¯;©†õ_mÛÐÈÝ5>QÊщu4¯›žÝaO°›ôýªId‚¨&–6=”ÎVè*?Ž± ÎùÛn€Öì,Ôå ^¶ÈÔ„bº‡ç,åjØ»£Ñ²Ç‰6H3. æY ÷´Ý€.LIäÅ ¯oͧé˜b=¯”0Œ²qËAÊ/Pþw‰°§ê< óòþ ÉÈSpš)>™­Ç›z=Ä[ÛÙyÑqØÐqÅ–ã”®|Aì8Æ¡È ™EúÄ©»Ô­]lp¡Cñ[›‰$JÞNðٹSÄÒÒlß6÷m¾{64Q‰™DT% E 8 ØßCßÓCü4¦¼EQb$Fi;¾ŸA2¤5¥ÕOQ\‡÷¾@$Q £´[îl½ÞÆ—~àÝ‹4žE`ªàqå(ÆXÕ]^Ì™‡{BÐL©|C[rZŸ]ˆ+ ¦X_©éÝi¥ÉP«dBG¨\Oï}„¢Ì‹Ñc"—;/‘24Zœš÷#_FöH%¡ŽNrG­æXQ? AÊ®Í#Þ?QB•S¼Hhè1¤…Q…ŒE’#Rq„R§F`báÓظÝ)nÛlú‡ÕCÙcEsi¹Ð”Ï •Vi0™ÖX ðG€cª˜[R‡¸t’É騳ö´ D 8C*×FŸ¡é‚Pcg`¹2o _¯ÆÏ=½ŠåTB€û¦„GëÄ=T4üññ&*x߃ÇBùÏh= äqù½¥WJœÒ!#™O76†½Ïb½8÷¥ÒL-_B…#Þ¾iªîrvÏO¥!ñ¼|O¥7­à $%ñT„ZŠeÝ¢™J®ÞDîM|™ hH+Ä <ô””3!çXņ°ì­Ò$”¨©˜|Erl Ij ]1—_»ü6ÑÁUs¨Ïe –¢}¶”DãÛŽšû*¯k,æâ$_÷¶Ù0³ŸíúÐ[ÞèˆÀA_¢ÿ(mk‡ì×AäúJWÞ9ƒKp ©îÁ÷ÛpžM‹eO³o$˜%µ„hÆõÎ,Øä¹ ·(áe*•fT“Í‚ûrhA9Þ@„½°$2L¦Ùß嵚`•®ù0Bðbl‹fÀÓÕ®2¢Ó!ÄónëBãŽÈÕåž/†%ˆ@tØl¹¤‚$~Bwe[,®)f”ŠƒÞÒp‡ûÎoÇÆ[Ò€¾¢•œıöÈÛ°ùçF­íšê€V‡Ï`¹ Š¾°F­®ÊŽ¿ .àèavŒ…- µª<Á 7=ìÝÜmt2[Ü \ÍT&^I%$zÜWêý*‘Dó.×PÚI8ÆÃ1û-e•ß pµ•cFWoß;/¦Mp…e¶¶åú?6 • Ùd:ªþ²|Ùñµ¢#?5Fîî@И(7œåV|~N Þ‡I‚·.Ò›$¢}{uùö ?¼zwA»_VûmÎûßW}›ß¼Å`æå;ÜvE{ÐÑy–?el¾ñÛë—Fy› ôÝêÝõÍÕ·ÎøÃõÛ›1 ¥®Šíº¦í¼ÑèË5¾%gáÔ Ÿ[ê³:Â7î|ÉÄç^Lp¨Áô‡:ïíصÅõ’!ºê~6ïââP(„qC5O*ÈÞ¯¨K¯ ËÅÚzå1\Åd#ÐüûxA‘b³é1TTQÊÝW^ÅMeÑó®³NQïœ1†0n÷pÜ wî+>b—sß'csÀîdòUQz5î@u¾«¤²;÷²ìŽ¡¨€ Í¬"êJÚƒ®¦ KòQY°¯d…±ýþòPw}ñ KpWùõÐW ÕÂnòCÕÓ‹ %k°|—ND³"ÃjÙw´Ž¤Üå?³NbÀ’ÉÇÑŸu>o=ÿ‰zôCWŽ-’e/ÎHâþË«›Ûë^Þ¾ò¦4œêsäÉægç帖ï"WpßW@:Ÿ¤ÉRþé²9¬Çq°FÇó3,+5¸ç/òHcüÙUÿcn&ဠëÞ_1ÄalNŠ¸—NEý5ðf³0éX Ž@.ÚýÊÂÏXiv!x8—»aDù­»ƒ¦ºD\¾ï§¢•‰C¥ôB9©·¬¸Oþ‘PüÙõ€:šåUkóâH“=x¾Q©qÁùt€âÆ<=LʽB½Æ›cݾëŽÜDÿLsºvp¡ŒgW>ýtøÔβj@TØÔ$£rìwÒNz€ÙÜcC“WK°áèvË¥1‰Eo·ò²&Cë(dþÔ3ŒO„ ßNâL;jûà#i4³‹ûûÇY¨…^¸ý'ªxÜÅ£ò•LÄíÀT°¦Å²acwàëixàŸaäý£ ,Tf0 ÉB•f_ì ߯f(q:É6–ÇŒ‰îŸð_Ø´÷Ãnq,ˆ=“½ñîÇh,ó· S fe4çÀ3à 9:h¡æšÜuÀ“\ŒÞ˜ Þ;S-ä\`u”Oõe±·t… fÊF. ÿF« Ë dHMEÊŠ4êÔ¢©¿í©€ábÛ|ʃ"Ž‡\`c ZGKîºàhi:Ú’Ó«*ï†ÛmO«cÕ'Ζã/†°È—¸r9”µ`µÂ¢ýª«š‡%,Hÿm¨|c endstream endobj 1411 0 obj << /Length 2450 /Filter /FlateDecode >> stream xÚ•koä¶ñûýŠEú¡ÚÂK‹z«Ÿz½WÜ%‡‹£è-Q»l$qCiãs‚ü÷ÎpH­´ÖõX’Ã!g8ï‘ù&„?¾)ÃMǬLŠMÕ= -Ôì74ùðæwx;@ÜÍ0ÿyûìúušnxÈÊ°ä›Ûf~Õm½ùwðâ Ž£4Û]Çÿ¾Ý%IÜôÃ(ÚVŒJ÷´%úš&/Nè;õ«ÛãEQFpoÿsûͳW·/iýA¦óË\gEÂxœ×úHÔ“< F¤áD6?‘†,Î `ñ+Ý7j2Ò¡..O3Ô=êÇ0 ýÕ?ôê¬9®Ò@›5BYÁâ(ñ§ÿRËFõ’Duwsûõß¿ýîîÇÛ›w¯>¬Ñ†±â|\õk$x¢‹§°ÃêK8 ÃlþÙòá Äg0<³ÙÅyÊÒ$ÝìxÌxJ¸·5lwQžªmqV§Ah<¸I·òàTm£"8¤Õ¸x@¸4;#Ýž¬jÖžQä,)rÏÞ^Ž£ê¤njñ¸ö˜œ³<†<©lêrXêq„Wv¨”( nFYûïÈŽÒ!+7~wîü^¦(-¸wVtíª;D™ÆeÔ 4%Z œ––ZCÁñÝÛ+šÌÔ‚Kf`\¸ã Ý*Æˇ+²d“¦‹Âܦa– Køá‘ÍÇKÈ<1ûó»Ù”œþ~Iƾú0ŽG(&À ¯á7Ä߸ îZ»dv©z£„E¢ucf¸Ú!s·V}-?Í£2ìIë¹w¼Îíæ0 BáXó$s[á¢|R”Nlj½’@î2üÙJ@T0cÚ×SG°å6¡¸kÇ3vÿ‚ùÄ埜lLõî'Î Ÿ.Õ³ŒŽ¿ÕÍxeàç÷ëß*í²1Äêú÷ë¿]¿khT+¤—FÐ3äqrÖ OZ›Zó( ^o‹˜D¥è7<ø„á²Ãf­ÖÂ00'È1hõŒÈ6,£®q!j|Y”PAó¬Rދȵ 5}xõÒ¦+h‹.ć [Àè œbA@S/>œwî°j±Ž[‘d„d“É+PŠ×^†×&2_’‡J-,JÓT‚ÏÛ" Þü]õ[²(¾ðÙ“tZôô„µ–¬d Ÿ"ÈxËr–‡é…jQ['¯D‚Ö+“8 ag•kôË„eçîh8¸â[ëq€²íȆ՞ª(xê¬A³TÉת>îZ×…´´·(ŽAnp  «ot;uNp°¦Žàþó¢ËY”OL|.HF™íë–…ø@#wçJÉXÞqgrË'x ±‹‚ÿÈ«5¶>Å Ôå6a©ÚatÚx\ …|;ÐBP£Oã‚~rvÏ ûÚ@WãV%‡áŠ@ƒê+{.uBû Wóªùi1F¡7.н_~ÁÊ•Óå|_À^ZÍlÄPJÓ™,qã;6’`´5$Íö´°„ŽI‚{¥àS_S?gŸà²°«ø)áÆH¿&ˆm µÕÄ0!<­úÊ2gi™üé‚?ÍήÔz³ËÉ›ñáøôÙ?)µ²²X²`çÿ”Zû)A«@Þ\`׎=kvÆ™¹¢ â0zWSî³Öä Ãé\k9‡È-¡EÝö/›T@’Ú¬Rh‰Þd;}h6ÕèÆijÆ“Œt c+…éiJÞ5Í5f-„º“4óaVú®'îÜ¥)Рª¤W?Ÿ ›¬4Q®5Ëýçšé#OM3ÛÔäåùÓÇÔ(kKðÁöT+[ù´Ó½¨m'f\¹r**Ã@ÚR!ÇRrÞ#£-ÅYe<€D7Ÿi^Í©³Ò‰}…³ofM‘Ó}ïô¬Q«©Õê:G—½½,¥x¶™YŒj±…utDM4 uS5}¶¦Ù)[µ5[‰@¶öŠÓül=öôѾ‹õ‘a £?6¯)2Ó“½w ±–6ÀUFÝ“øë‰îüFêá…'±(‹í’ç„g )1Œ³•>(Íš~nqÛt˜åÁ Ý!Ï›«¯°yãç1+‚wŒpì­FíõáÌš_ܶ5gVO¾°ÑvctG._Ä¡$Ä}A -bü§£¢;¾›_‰KDüw‹;bñ`]Y<ŠýSËÒ~Džsøù@鱿ÀPB<]J=Vï:Œ„Óî²kŒMaöà <¢&µÏe}¹È % £©øx?%p áb?ÐQkÌî ;áeYÒì9ªöÅ;Z|óæýÛçßÒâ´™@)ûã%Ž÷3çµOrÌí­è÷'¸ñ/å öý"BÄÁMwl'wžýÃÃÝû·/oì÷ÿµb òöõá¬ÙGþûMƒ/ endstream endobj 1415 0 obj << /Length 49 /Filter /FlateDecode >> stream xÚ3T0BCKscc=K …ä\.°hQº„äÎeUgbl¤±*OS endstream endobj 1420 0 obj << /Length 2791 /Filter /FlateDecode >> stream xÚ¥ÙrÛFòÝ_¡òË‚µ"„ûÈ-YŠbÉV‰Ê:Uë}C+Ì°¤|ýö13)8IÕ–ª„žž«¯é‹þ‰þIî¤aèæQvRÔï<ÂªÍ ÷Wï|½n 磕Þ]Æñ‰ï¹¹—û'ëñQ«“;ç[±ë¥šÍÃ0t|ÿ§Ù<ŠbgQÌ‚ÌyjZüÏ‚Ôºa·n3§U½\1®lø»ªG(tQ,¯Œí·’‘Kܘ:²@b˙'îe'…"éns§ÚµÞzùA“æ!ˈúBXà–_çÈò㇮3;÷8óå~ñtèçN»æ/Ñž³º^•¢aô²(eSH÷‘34+=úzÉ;„ÁÞ¾½~`à¦lŠ¶ÒÇ݈G-"Ñ·êÏsÑ„<–l«xÙNÉïe;täûLÿw¼Dª®l|˜zÌ|//¾ž2HÌÀ×*#{‰Â^ñê ¿¬'„,/÷æ¸ ¹–M§‹Õ÷Yœ÷…¤Sü#máИaìüWý$7‹ ˆ÷©òo^ì]àæÅýJcÿç<”:¾:\v¥ÄþmÁø³çy~4Ï£¹?÷ò,f,)w^†^âó<…iäØ RÍGkQL¯ ÞQnÊÌð­}ü}(wæ}³#8oÕÎêÃXR©¬*ö ݶܹæ¸Às}?äó–½d¶ Îœ+—¿¿´Û¦k5Ò*Z‘©yiˆ+õBÖ B–)€.ä®×§£Ñ0J«Ÿ/–¯$['©*°–#Pviè}Úù„^? O8L<ç¥ Èš*²ïÄyÕ»žÄ”xbB#NRý®`¹µÅÄØ" Á-Ix§UÇxk{)Ù`öO7ñÍYžó±Ù”„ÍfŠîs­Wvñ¼Ä½»Á1½5Ð-“ÀgR8ú+œÞßiôß‚åŽlBÛ1‚$Ö«Z1êâöìØóà¨CS 4u-YÙ /ˆrgMòŒ2M `ÎAŽ }P­X=2Û0D¶F¯ÛûؽQ¢@oS½òä ø!»ÂAßê/_‘‘—aÔŽwB›1VQ´î“S2¾þpË좫%Ï’;ò0K¾§˜ À€'­†•´#þ~óü°'óœ‡n@î4"_ò°-;^RŒäÃ=÷0P²&ËûŽÈœÑÙ¡‹iê ôåë%ʵи½¾[òÜN”J®æØ_¥'—×· ª Œ£(ïxÆû)ÈÂôXW8¶ºÂ•¨«ƒkIWØíÝ2¾píŬöÈ£iŽ–í  ‰¡*tNyîº)&­ó+=dÜ—ƒI@ÆÕËõP1ºoù»nËBµÝk×˺c N‰²×s+ôD~Nˆ /†×ñÌ£ë¥:œËù“͉kPr!»Ùæã!M¨z%¦ÄôóÝ9FøÀ‰!Ä0´¼½ëB "„Cç ì$nwz†õ'Åš‹ï[Iqׂ­ÜÞÐIÅ`Åc2ÍÀ8•`¸j‰†²æPE>]íï8ÒÁ[ÛDwþXW é6²§ˆD¼²#ÐLc<+WÖé"§n{mߢ`HbkßYwl€Ï\sHcØ&÷¶‹2ù GñÀîéØ‚Ðõ ÕÚ†b` š0Š0s»¬Jâ$×*?Œð0„ƒ$•Xm ¥*¡¢mx0t,%¿´Õý´ÓZV#x BP5wü™ KöO&ÈR.5©  ›r³%;<åkM,KtÄ µ†ÄÐoái¢}¹³ÐǪ`¸Ó6K˜2–GwØô/ALÅÍÎÿË8y>ë+FUÁ’=%PtÊ«Ó ‘ÐÚ΄}È #Ȥɲ¸${ÛF´ž»ßš]O¥©yL&v»ª,ˆ÷IÿÊA/H"tϨµ$ŸÑEÀˆS§­ô,½CžáR`H}ø˜quwÊQŸë/]9¿y]©‘ÝJtÜÖâINÉ=ŠÝÌÏÌÊIãM\/IÌŠ^¾”ͺ< ‚¿?ŠèD9EaâÆ%°*û¶­þúPíΣ c3ªx ß\»ØWˆ°1¬*z›0øHH(a)î–ßwÜíÐú„%ÜÔræpMQ5ÁëPMMâÆð,\¼¶7­>[ô…˜«{К§²¶Ô/o¸&küñÁ\* endstream endobj 1424 0 obj << /Length 880 /Filter /FlateDecode >> stream xÚUMoÜ8 ½çWÌQÖ^Ëò×›nf±ÅvÑ-È¡íAcË3BlÉ•å$í¯/)jÒI‘¢…›|¢(ò‘¢ù&‡‡o¶ù¦"Û–í¦›®ò€ºã†„÷_ñh—‚azay½¿úsWUžgÛ|Ë7ûáÒÕ¾ß|`Hí$H¤ÒM¹ÈxEItvšGåÕøh¬ëõÇœ å”Až@ OT%“ãÑ:p}"ä-‹›wÂjzIê;®†¬üIF_=æ×2µjlôìÔçU;ð¬âI´¾¨Y:é#y?ËS@ÚÑɉÁ[ü–!3‰"(R›ÅgO»·/b;âì8¥úˆœ³÷JOXY”;k0Cïôa yéÄ@A~‡ËŠŒô:Å]_º$ålünÖÙ•lÎŽ¥é£@ƒz  P>~r~&ÖOjœ ›ƒÉòò Çô%N´§Â@n/fýV›cúúD9sÁþÕ+ Ô¯ Ýsòr!eX—´ú€KjÓ“î{e FÚÐ746 Ô^uÒ@Ì 23€`Ý„âBNì@VñJƒTdy<¨z!ÅeçQ‡eŸÁ±ô†yºlŒÅ9Õy2Ó/÷Â/ê&F¤„sÎ^!ÿò€2^wR ,…öù(ê–á1èFH}9BxUdmS©á´Ã`ÎfÏ&MQe-Ì£hõ1¯òkýU:wŽ-4¬›°3P¿ÕÐAúÏƯœl㤅NCa›B£º.Ð…”7mVoëçüR#ƒØqãBz¯î© ,:*ÁÅ,›¤!ðì¾[¬¡ˆ"º%ÿ%ô^¸ë#á—Œg‘Ÿö’Ÿ/0Î7óÍ%U¸o¨¨e& ^–c’‚€SáV™»a£fãþh„ZúQ»:`Žº…Ä9R:Šžd×fEZ†`çÿ~XƒzŠª?à[vwa©úY—kƒ#7Ñnü ¢j(ºª{¬Ç¤L´ ÷ ¾; “ío³ó¿ñµÛï. endstream endobj 1430 0 obj << /Length 1932 /Filter /FlateDecode >> stream xÚÅ]oÛ8ò½¿ÂØ—•Z)êkûri‘¤^¤i8(‚ë=(’ls+‹^IÞ Ø?¿3œ¡-¹ZìÝÓÁ€53ç›C1 à'fY0KÂÐÏT:+voKm73î¯ßæ[ãbÀùaõæÝUÍDàgA&f«õPÔªœýÛû¸Í÷}ÕÎazBþ2_(y7º¨š®"jÞ”|4û¹L¼×Vo¶s™zý|!e¥ž¢ùV¿¾¹\U‰¤ü/uFΔN@éÄ—"R¬tœ*_„Š”r¾ˆbéÝ̳ªÎ…‡Ú&k j Ö€U±\2· ˆ& !H«ÏyßW†à«yz –aF«iz®© „Ÿˆèÿ i·° @æÄ :š-Dè‹Hж./PG«è âÓw”f˦ëuè9ýÌš¾+´·¢m©Íæ ¾=Â(2_ÈtF%C #|×mUÔ™u2^0¬y[½r nCš9O‘7´U©»¾ÕÏV+°§/¤Ù;Ó¼ƒˆÀîR¯_ÇL‡¦¬˜§ß²(¶+…–)©`ͦ¯3dà¦ëÛG4>ò®«¦jóš»Ãs­ ‚• ¹o={ûÃ3H¬u·­J¢<ÛÔ .« ”Ac'Ǥè˜Cò^›æ=I«4Èh‰ád¯ÚVGpz c¨ÜÑ¡Sû-/·´ü-ˆ‚¼'„;“Ùãñ°.ˆä ½:?v°?0ª¤Í838©W[̈PDÞ¾™M›ïwtpXím^¢…òtCë, o¡ä0ò±ô[«< ë‰ò¢ëš(ÏCFÈøõ¡FãEæa¨¼ïërõéË㊋Û'bÿ:Bx÷‹ ß_Ü®ž0,*Óú­±¢Ĉƒ4D!çŸÞíkm3¨\-y´wn64àø|yÔŸà(̆‹Ë›åꉙZbºZ®n/ˆx…ü_î§2ú‚âtgÝãÿjùññ@¢?Þß}y¸ô { ÜT'§Ì5kBAM‡+ 5¾ÀÖ†³c‡Þ0íIVYõ¹®»É&òD™wœò:pl]¼ÍÑ<ç^)#¨¡¢ÒD`žœ îÉDD‡JÉU€À5êƒæXÄ™ƒˆ+q„¶XÁµi6$Ãíbö"´omJ¿Ÿr½Æx™×˜þ-§g«m—×Z¤$özQÜ›hÀƒ¶Z²òæ{MUò°¼¬tø2÷•^£AVpM-Õ´¼öÁt=ž0aÖgN©@ ÀJÀ-âñŽDĶv;ÕÔq]G(7â ùg“ŒÌ )"9‹‚Ð…°ÓŒÁšdð'¤kÆ”á|ãö/&³óc¨ÿßÝpÿkÎè± ^ŽeþL„czfƒ’±‡ÉȆ¿ŸÈ÷?h|.ÔjìŸ.}+bXQK¼Ucè6e©9a¢@-À¥ïº)ëWf뾟ßÍ@ÅDµ«6®ï±iYWå†Eðå,U4Ž()?ØNbZƵÓÈ]'ˆp M4[{áCzîmuÚ¤'‚æïËV[µ¶„žÄŽ9 ‚Ðà”^xOXŽÉ–®@‘ò𕎲$Ž}™dnújçö2øûý ‘„”Ss›’~˜ç6ëF8°äogpR)Í.hRÇjïéë6¼¢}»mË¡R?NÆ5Y˜Ý}#³Ôë¸yˆ2ð[V….™ö²­h˜@Ä §Ûk÷Òìñ{ÌËE]W© ÔV¿ªŽ7– õi€PP%Ôºxiíw¢» BÝA¾ª(8å+èÚ6{"mµ®Úª)t³wT o7¯ãêq^G4æäúCÏWuCøµOß_Ͷél‘à¡ßàñ³Úò©%Ý`j}ÓLi}|²áU_í*r"',ö7Tc.†ošJ5K¦S—0w6× þ¨axËËËKÞ™ wÆÂYâ6fáä+Bù‘ŠÖÄàð@(Eü§ Åi*„·E„ëÜ^d˜ú"ŽÇíåÖ4 šóÃTçÒŽ(ÖføRƒ@ˆ.2èF ,ºÎŸk&óüŽ í–nŒ#Í©‰Äukv̹5]5>‘º6׉Æîu3½ÊÜ…^ΨGTþÆbE½V.~±S@Y¯­EjÐzé™5/°rN³3’†ˆ+$Š¼ßª¢'BÁ/zè Ñ oC˜õãDòQEÚ^m6ŽANÛV ˦)5Ž0ÈÐÁ€_yø,®ÒTE48Ø®KÒ:îñy ñ@U ¥iä'îq,£KcGïÌž<jŸ‘ùQÞ”õq†L_€>/W?wN> stream xÚ3T0BCKscc=K …ä\.°hQº„äÎeUgbl¤±*OS endstream endobj 1639 0 obj << /Length 6834 /Filter /FlateDecode >> stream xÚí][—·q~ׯØÇÙãL»nô%o1)Zt,G‰'çX~Ω‰æBÏΊҿи4Pƒ*`AY"—stŽvI6ê«þPU(.Ínjù»뛾iª±nÖû/êéoOooô/ÿõÇ/˜yn)\zOþáÕ¿!Ä ««±ÙÍ«7¾¨Ww7[<û~õî¼9Ý.›¦Y°æ_Õ/íâÙñ°Þ¼;ë¿}y¸Ûüt»ä¯‡«ûÛ¿¿úÓ_¾r¨‚óLõÔ“úõR¿¾âL´F¿nh+Ö´Z?ÖÜ.EÇ'n—l¡µê[£•TEKh«¦í¸’ )h몯{- ÓÏ𛡻®5OÈçyÏÍíòõVIm›Åê–-Në[>,¾ßž7ëóÃi£Û^û^T¢¤˜©¹¤ŒõLÿàÁŸXúßþY @q1TŒ7+& Bu“|‡›ºªGù?Ƨþ ÿÆï8Û~é ˜:ϧ‚Lì4滾jè‚›Œ}šF"'ä¹î¹jQN‚iO«@&U„¸Ðeàx. æiˆDHúj`M9 ¦}B•dRe¨/tûªæc. æé äÐýeãP5Öÿÿ-éÿ«»oE¿XÉw§¢_l*È°øfµŽù¿T¤é>Üÿ!u/c#dü«›‚®1í—ž€Ë® :È^¨Âx%d@öU!zÆ¡J2*œWý˜Íyš"øPµ\”s`Ú'T A´\’ÐôUײ\ÌÓ42‰Ð6“‡“`ÚÓª=b]zƒàU; ¹$˜§id !ÁŒ¨Å$Øa›V%Ñ#:lg’`‡mˆ¤‡íbì°Mª@tH±a;“;l“È@$=l“`‡mZ•dR¥gØ°I‚¶)d5h÷MÕŒ£’wo§íùûýåà+sûºeuòý5¸è&b8sÄ>ÚdLû¥'àÒdÈÔ)µO•‹!€•Åtr4ík1Û·‡ýF&nãâ±V‚,Vs‘ô4•è÷BOztû¥' ’ô„ º§8šõxºäd=2™ÈzJIpY©J¢IÀ³ž<\ÖC!‘‰¬§”—õPªõŒhÖ“G‚Ëz(d 2‘õ”’à²R•Døšõä‘ಠˆLd=¥$¸¬‡R€è@³ž<\ÖC!‘‰¬§”—õª„ ºXay$¸1Œ@Vc˜ôÆ;SªØ·½ºÅf}9‚5¬â׬'{Hµjø-·œYn9~'2ZŽ§KŽåÈÊrdJÓŽ¶ÆÅX»øëÿ^Z —yÏpµ›Ü˜i®ËȵÇmÈëN„fEr,(„½¬“2^ufåÉ2éëÕýv­WI¶:ÍÆê£b¬x7>£C‚VÕ˜»áñ6aSOš"Øg"m/»å*„U˜§i` ËMz9•hË90íª„ :KÅR“LljB‘tjRÌ ´*!ˆž®´XtÈ$Á† YF‡V‚÷ 3ο;ÞÊ–‹µrüt¸ÛÞŸOÛ×·lqÞ‘u’F>í§¸Njß½xêá @§Á¢C#§KÆÔƒD"é©G1 vêAª@°ª³uÆ<œSÈ@d¿KIpþMª‚èIh‡úw οòÅøߊ®3ãy–Lži·ß«ð §*Ë»í~s¸—®¿ÚéuÓÕé´R+§?ß_ÆäÐR÷Y4Àz½•™û^Ÿà½nžÁ–Þl¯{ºäô:¬ú\Î#Ç^}.md~¦«:áÖ¶÷»HGÖU7>Í)ÅE°©ê¦qœ=ÚLû¥' R At Ä*’¾*I8IW$‹9°IR‚Í‚Lù0“[‘$‘Hº"YL‚K HUBMº›I‚K (d 2‘”’àÒJ¢3^†¦y$¸´€B"iA) n€ U A°¥;@ä‘àY œIc0•ˆ5?E™Ö7×X²HÀ«ž÷ŽËÇ tû¥' âµ!ˆ.Þ£áÂS%'\À¡ÈD¸(åÀ… J‚­çYßÎ#Á… ˆL„‹R\¸ U A4 .òHpá‚@Vá‚©½Öv1Ž‘pÑË÷k?óp%†½ò„b€'~ …§KN„ ÈD„(%ÁEJ’J(òHp‚B"¢”!HUBTB‘G‚‹²ŠµÔ¼3»±Ÿ­ö»È*US±_bCõ“ ]Sõ‚9úŸCèöKO@ÄQC"j†Vü=e¨¥[ߧ HÌ=äÌ°å,˜ö´*İРþ‘Ç‚óZúG3ŒUk‹k]f¿Ûè‚ÚÛÍasZ‘,¼««‘OËgÂ2T<@xÐ"èŽ9€¯J†ÐÀ¡HÚŠ9°@ª@Ð0`L:“ë´r€~¨úQ@`‹Ý&²“³©¥,ñD—›¡R¯iy|ÁQ·_z"u¿1xS’óU!ºÚVùH` +8Êõp`Ú'T A&UÆKÓ3I°I9‰ D"$˜q¹˜;ø“ªÔàŸÉ‚‹}4™ˆ}¥,¸ØG©@Rƒ& .öÐ*öuÒ„­¶w»Íú¼ýQ­°môjÛ›‡ÃY]Såú·„XÜ0/V¼^à À#‰Ïž@× |]2Ö Hd ’^/(&ÁER•D/ lzŸI‚Þ“È@$=½/&Áeo”*ÛÆmæâ™$Øé=‰ DÒÓûbìôžV%Ñ•`ô¸u& .‚È*‚ Êyã"ØÃþ°Ü¯n—¢_ü_lÞ"­vÝ'µ)ÈFó¦å!m€‡4ŸÎȦ<]rB… D&BZ) .¤‘ª„ ØyòHp!B"!­”Ò(Uˆ>€îŒÊ$Á…4 ˆL„´R\H#U At‚ŽOHóHp!@V!­Uû¢:ÒöﶻM$’5¢êFöDg¢6*ÊcÞ,y>ß‘™¨4…!ЄšˆN'`X¼k*ÁX9¦=­ Á¦¡f¥7“»xL"‘Ø4t¨1”“`Ú'T A4 #: Í#ÁÍB)d 21 -%ÁÍB)Uˆ™…â9\ .àÐ*à5]ňwSú»š‰ÕÛÈvNÑÉ_›Ïb 5Iåc¾kù^?Œøˆ?+’3à°¡ÀÄp_øún´§ô °`M?ëõåã°Êð¹ÃYk-ÿ ͽ}ûpZM—ÉNLJóö°‰xÀ í?ÝáëfÃHy?ÏðŽöioñ¬ÎÓ%§«r°­Yu6ke ®³ßlßF¯÷kEUÝÓ^Y³-CIùln€Ïæ|ÞÛÍyºäÌæ(d 21›+%ÁEvJ¢K•è9¶L\p§ÈDx/%Áù=©J‚ pÖïóHp~O +·¯Õ¡s†eýózg*ëëãƒ=ÆIoDÕÖÍÓŽ6ï6ü”'÷³<¹÷;!Òó6÷tÉIî)d 2‘Ü—’à’{J’LîóXpö /Îyñ¾©ÚÑ ·Ï“ç¼îVçÕ|ÂS]ˆ»}ý€¬AIjø?k rKWÇË9/¿%É€Þ’dŸÁµ¹ÒÈ×%ã–$ˆ¤oI*&ÁÞ’D«‚`ËkæQ& öL‰ DÒg’ŠI°Å)R‚Ò5ÕóLlAžD"é‚|1 6…£U AôòZ¥p™$ØŽD"鮘›Â‘ªMšÂe’`S8ˆ¤S¸bl G«‚`dšA)“;„QÈjk‡JتØóg¯¢÷Üôãg{& †¶òÐ0 ÀCƒß7Œ¡¡ÁÓ%'4PÈ@d"4”’àB¥ Ñ—”ÔhhÈ#Á… ˆL„†R\h U Aô<Ý@•I‚ ² 2c°»¶î6«»éž’a1}Âá‡Èš]ÿÑï>ø­Ï.ÚÞÔÄ–[ŒkŒ×wz畧HŽ¹à°ÊZ¸ ¶ðø|óã6vEºt+®&òaÅs/°%¼øòaOzùpЫxÜñTÉ1$XYëªÎ–²ž¿x…\½×\íˆá—¨Y‹/Qóài‚ßSè%j¾*—¨‘À@$}‰Z1¶F«‚è¡MOò8pé  D&Ò“RÜ`Cª‚hÐÙ$¸0A «0QwÓuõ:L|õ*~öq¯a‚ÈHl’m¸,Ïägx&ïwÃ'ùž.9®B!‘ W)%Á¹ ©J¢]ÏÌòHp®B KWa£Zs±©üö~}ÚœÍ1Èõñ~{0Ë6çÓêpÿæxÚÇn¦âUßüªWS!3oû.Å3oO:óÃgÞ¾.3oˆ¤gÞÅ$X%U ‰™w& Ö_Id ’ö×b¬¿Òª„ ‰™w& Ö_)då¯C[õÝó×·ƒXN[»ŒtZ9€6ë—9Lþdßµ8Ió IZ@(zC…¯JŽ%SÀ@dÂ’K9p–Lª‚$’´Lœ%ÈÊ’{õÁÑ&fÉ_­NçÝæ猡Gêÿ+KÆÆóåcÎ,s|ª"é‘ <]rÆ ˆLŒ9¥$¸1‡R€¤Æœ<œ§RÈ@dÂSKIpžJª‚$rÄLœ§ÈÊS;µ“¡yjV†8¶ÿ-o¤Æ\×¼V¹ëÎp×õ¹#ÒEO—ץȄ떒à\—R€`_z´~–G‚s] ˆL¸n) ÎuIUBíºèuò™$8×%•ë ^ s®¹ÚƒM[µ¯•GL°Ì–;Î,w¿ûŽ:Ž§KŽãPÈ@dÂqJIpŽCª‚hð1/ç8²rœ–WÜ‹xþm|¥¡æíuóC8¦ÚÊÇÔY>¦ú}CŒ©ž.9c*… D&ÆÔR\h T ©15(d 2JIp¡T%I©y$¸Ð/¶öª¯*4½Ê¿Lníýr½Þmî·«ûsìD“:+®7µFHöû2ÞÜ5!óˆŽ%èõ½ž"DÀJ뺪µÅ–ÍO›õÃyÿð`}]ˆn˜+^¸žÛ£ÞïtÙÚÓ#cÕšB ÒkÖ¥oo«¡¤½`÷övD¡PCôxRúö6zzDECAÞëÛP`áDŸ¡b7]+'qÜŒ'/âãI#ÿdt{ñâÕÿD¯hb¸Näùñ!Á¤ãü¦ãcÕŠÇ헞€ ƒ ñð"x¨n^êIÐYSç¥5Þw5kÎj‡æû[¶X¾ßÞß÷ºv÷pÞî¶gu,ïçÈi£¾ê>Ń¹€ÜVzäØ –‘Çv³m¾œÛ_ t'_lÖµÌDã+Bt´y˜‚ãaT}u´DéëÛæ¤!rÌ|2ò¬WŸl‡s&^U;pÜÆ—çãR<Ï°u&eõÃGD±5ï_Þ¥³¼O}’‘NõôHõ*h»U(g²Ç'Ñ QØ8üõû F—º‘SÆÖiýèðbÚ/=—ñ€ÄÓ´vZå*xاÀ¡ÈxˆieŽÒvM9¦=­ ‰§•a£öŒ(Áy:ŠD8àÍôU†bL{Z¢oðº(Œ¶ ¯F6ä’`ž¦‘H„„f©h9ºyB‘ùøNÛöU߉\ÌÓ40‰0 ™ïõå˜ö´*Y—sªcÙñÀIÊ#ÁM’(d 21I*%ÁM’HUB=?¨ÑÄ!‹„9s ¡ÈDîPHÂÎ)UUóÉ#  â D;0Ùò«äÆ•ïW»7êK»ÍOæ;{ÇÓ~9“Еù§4mj圲‡)HŽ]‰ÏëöKO@ÄçCdñËø§¯J†Ë“À@$íòÅX—§U AMàÖ?óHp.O!C‘´Ë—’à\žT€3†L”ËSˆÎáe2ÖöÆã¿_óS|M¾»^Á”¿%ÉrZî/³Ü_üŽ#üÅÓ%Ç_d(2á/…$ÌþB©@ôzqõ—<&!¿´]5fùWÏ_ˆÈA¢nÚÑpõ—ÇnÜ4}j.7œYn8~7Š¸áxz¤ ‡@t†£J3vÝW›Ó~{Þ®"—Á·¬êëîZ© *•–¼âR¥'­U=Ô¡ÕS%' ’À¡ÈD@-åÀù¥ ¡<&¿ _ÈpY×vBqT…£÷{é|Xü¬ï$˜.£Çw‘JÒXõ-ÿ|œ3 Ãa¹iÌpÓð;êrGÎdž)Ó i0Q NXÆ»ÕiµßD?=3Èiê§E1Ößš˜òîvíñÞö¸oâq`V"Õ×!\¼òÐ ¢j[c^/“¥‡—ñ ì%Pýù¥^H½²éYÕååʹ=Z­4èIÇ*ž"… 6H—-J_ßÎÂH=ä̶ß³^ße ,Hç …¯ï‚¥GœÖV¾Ÿ÷ê*Hp.Dtu5Ž&(mËw»èòRÓÈwúöh€×i†a:$nµx,±¶ýÒpùª$¾pÞH7Pó _œ^û4 DÆm«‡ªc]9¦}B•©ŠšÃy¸ó0‰„—z˜¾~PÊmŸP%ÁŽ˜C ™$Øs$2IŸ‹(&Áž‹ U Ø©sˆ!“{.‚D"ÉsŘs´"„f`ÄÎEd2`ÏEÀ@$}.¢˜{.‚T€$ÎEd’`æId ’>QL‚ÝHL«‚`GÌ‚v& vœD"é5òb\ÒI©@t¥]#Ï$Áå2™ÈÌ Yå60 ÀmÀï‘&>IõôHÙèl€•àfZ¼½?ZÈÌÓ(µD¨¾é.Ÿìù´@ÿ·¿×7wòÿ$_·º›÷Ó£û>eaõÍîæÛ/þ3‚)% | 1×Ħ­DÿKª/ƒõðõöp·=¼Þz%>ýϲ`©Ž¡ <ՙ੎Ï3~fÊ×%'Õ¡ÈDªSJ‚KuHUBÄ™©<æT‡@†"©N! s £T Ø9XÈò˜@Œåù¨®$2Áóßñ¢¼Ñêôèö]-ê?É,zgl‡Ï¾JoŒ›«…á<È€z}FêÑ µ¾.D!C‘´•’à<ˆT€ å™(‚ˆˆ‰vª$M˜N.kí¶‡ÔH:9Ññ —9ÿû°ý)rƺ­ú橬ua¾1…Šñ|c€û†yFïÂ}ÃÓ%Ç7d(2á…$̾A©@4 Ñ49“€É7DçM]µv¯ëŸ_~ûMô#²Œ±ëNºÇn¨‡Šõ½cøñ»¦tû¥' bÃ!ˆYš¨åxŠ¤,‡€t–ÃFùÓìúÜW&çx½Ú­ë[¶ˆNE:ùÂO{Zí♡§Nß ºØä­Œ¬Òî?Àž<¨=Ùgt,n±óõHØ…謩S»®Ì>èýñ°ºÛ®µ¹¼;e(}{Zí÷Ñx:ò©öé˜2HZŠIO:H4³:zߦ¯Hªg H׳jâ!ÌNÞ¯¿y=×3ô×ì*`µ”¬žt` úM4ØÀêë’1°RÈP$=°–’0BJBÜ#“IÀä.¢s—¶•yç0»Ë×Çý^-Ú<¶ëÕùÛg,_´û /–Aö"Ø×)Þ‹à @÷"œ]šªÝ8àë’±D"é½Å$Ø=© Ñ9à€è3I°zˆ¤ ôÅ$¸ Eª‚èÚ4ZBÉ#aZ2™Z…$ÌA‹R€ K5SÐÊ#` Z¢ ZÌôaÌzùûÿˆ$ú’­v¼ž* ]CÓWî®=î^ ôL™§HŽ[à°@`Â)Ê^ö BÙ§5yDÖ«OÂ×Z9‡àêë™fr¼·]¯#EŶêºë2süðYî ³ÜCü.#ò]O—!¡È„—’0» ¥ !ªï™Lžã…¤±‘SSù—dé°y¿\N+ý¡i Ñß:‹î|ÔÆmþ|ÆÁdªýXñº)φçöh2lA>6eWO‘ŒT˜‚ Ò‰péëÛ<˜Ò#„HdÁy¯o“` 6H§À¥¯oéG¡S?4ˆe½¾‹a,HG°Â×wŒÒ#„Я­[R¯þÿÒ(Àô endstream endobj 1400 0 obj << /Type /ObjStm /N 100 /First 995 /Length 2501 /Filter /FlateDecode >> stream xÚÅZ]o¹|ׯàãå!\’ÍæÇÁ8äΆƒ `Ø>äÃñƒ"¯}‹È»†´‚}ÿ>Õ£©•,¯Æ’v„Ò‚3;SÕdW5É™ÒÕ¥wS·Fq±à?‡ì’VkD'M­‘œÆæRWÿ¸¬ôàJ ®¦áJqµ´£˜cp­‰R0Ü“‹±;‡KE†VÇbvQ{%ÙéÐ-ŽáNœ«:´€Ö´À[Ïv_ôhwÄêR|DŵÀŒ.[Ø)ÛõI\’­…ˆÛsJh©u¡&C*.•b}@×RúSÝ:Þp}ïÖ„*aè»$'8¶Vv’²}+¹Tñm\'ÁXE†(`‚áÆA+°Vs*e¸®;ÍÕb”Úà Ö¡? Ò–,º,N» ºW‚  J¬)Jm@©®XÑjHß0b¹#‰b£¢6Zj1+²ÐhBz›Ý>×Їde¤Ú?ƒ¨J²dcઅVuU‡L!©€Öê®Ö!ú™ô!zt°…!ú’\‹bÈê–Ôb)Ù5©KQ×r+à@@­ Ñ—êZJ3‘ (ÝA Æ‹Ûz(†R£ë±ç’ëiÐJ×s´X ³®ƒÎd/ÙòÁîµ/RÖÛ=Ôû½A…8(ÖTR²ï¾ ’í©ǘôƒV ­7”fq˜‡B &.ËHèƒ nˆ&[4£·á†˜Š5aˆì¬‚Oëf§b†È_EäšÙL–-Hø:FôìèÉ“£Åëß?-Ýâçõz³=Z¼ºøÏv8þëjýߣÅ/›³wË3÷& „·G‹—Ë“­{“rñ!™VÕGU{Ó+ÎöŠ«~vOž¸Å+·øóæõÆ-ž¹žþ~rºüwÈáéæb½]žÿÁýôÓþ,€gî JKÅK·øÇ?ÿ…,AHÁ­/NOßò¢ç›õv€}qµËkŸÉÛ¦4¸f<€ŒGDÜ¿xq¶9yµDänñâÙs·x½ü²u;èËxqüay´x šåz{nEÆŠßKëóùæâìdy~YĆs[¾[ÿ²ùr92õ«ö„ž¿8>ÃÝ(a-\^8 ë9ˆ‡Riñ ¥Ò7øß#ïN²ø)I¨ÞÊ©höµ¥½)øûêüÝæ£åà×íêtµ]-¿JÂWAìg•|·ª ¶ûazðV2!0;~{#xwį/ÿrùñÃÇãÕévóãû÷Ûϲ¿9ûpS 6»LjáÁIo¦ú¾²`ðºOX¨D%ò‹n5pFÅŸGqýÅ•Qq1Qz•6¡Á[ÒšopœªšÈ0+á¿è(¸ï§þ·íöÓ‹ÅÉñÇS¿ZŸ­Žýû³ÅÆ«õ»å¿\û߶Oï©J¤U³ú`S4lÑ1`ªÑ×$ÿ稰lò6•ìmjD&q8‹IîÀ{ÌÑXaøŠõæuÖ(Òï32Ÿ?ö·:5†;˜ðÆE˜P}Õ»˜ð+ç]wä•ñîµ=Õ=Þ³ºÕÉA~/¼ö„×î^Nñ›Rë½ýž0»w[Þv¯¶V->åam‹Ã‡[ë[ õ»H¨?`u°¿†?X1)›’”î—’ë}Jò€>;ʑ=BK-qÎIPœULT- â &C€ùbûŸ±Î¬÷­ZÖV´§«“åú|y¾øðéôɇ‡Ôsì#¼ícEQ@mK˜,$Ø¢4/ç[û¤»Ø!h‡¯ÖÍW"Â6iôÚ,¦ùV^nIs–Ù"ý›ðLpV¿è¨~¡ „+.ilô±‘Üd®Ó²°‘Ù r&r&r&r&²Y‰¬DV"+‘•ÈJd%²Y‰\ˆ\ˆ\ˆ\ˆ\ˆ\ˆ\ˆ\ˆ\ˆ\ˆ\‰\‰\‰\‰\‰\‰\‰\‰\‰\‰ÜˆÜˆÜˆÜˆÜˆÜˆÜˆÜˆÜˆÜˆÜ‰Ü‰Ü‰Ü‰Ü‰Ü‰Ü‰Ü‰Ü‰ÜGd ÈFbCØÈl(…ÊFcƒÈ‘È‘ÈÜh$r$r$r$r$r$r$r"r"r"r"r"r"2' MDNDND" ‘…ÈBd!2=¨ô ÒƒJ*=¨ô ÒƒJ*=¨ô ÒƒJ*=¨ô ÒƒJ*=¨ô ÒƒJ*=¨ô ÒƒJ*=¨ô ÒƒJ*=¨ô ÒƒJ*=¨ô ÒƒJ*=¨ô ÒƒJ*=¨ô ÒƒJ*=¨ô ÒƒJ*=¨ô ÒƒJ*=¨ô ÒƒJ*=¨ô ÒƒJ*=¨ô ÒƒJ*=XèÁBz°Ðƒ…,ô`¡ =XèÁBz°Ðƒ%üL(ió^,ØòÙ“„T*V&ꊨ—[ž I½ut7Ί ·=é'gKX¥INÕ9[õÍÞ2³Woo¦8[83TßáuˆÖ'è,ÅäͼŸO%îåŒRF¤úßQ¦äkŸ¦,R¦æíÅÇŽRª/öZaª—‡rfñPѧ&oí§:´£Svœ£S¦8u&§ì8G§L¦³Ïã”çè”)ÎzèØvä/Ù«¨è fµgí¾$ÝË™uŠx›q0ú€´îœ‚­Yo}?ešÉ*ä¤U¦8ç² 9i• Né3Y…œ´Êgi3Y…œ´Êgm3Y…œ”ígK3q"ÑÞ(“c.*ûËíL“'–»>_ëf†|c}ÊQ2Øö d—Pº½¢Ùcaö(sg­ÞÖ©;J*h‚RæØ%Gv‚2åỷ=+DÚÜÕH‹ïÚöÏcesî8Ç¡ä¬3-9DZâ,e&Î|+×8sóMÚ㬅D|À~6 æglp9©dAžÃ-«éyæ”å8§LQ\([rR¶Sœ3-¿vœ”ígŠ3Iˆœ”Ðç¡sJIÃcê»·=înl±YŠ·ø¹†–”Ú Ê4—;ÉÉ¡âl3É6`I WK¡ *Â㪖”ÚÛ)g-)9²”kÖÞG'ëûgïP°m¹edÃÁýDá zEÊ~N’Ê<ù riu×Sû!MÑú(›#ö“”c?§(\T¸RÐ/ô/ ³œexHñQú‰y¹‡k”£[§({I¸äd:'8ç.I™ÐIR™g™ ¥xûå— R³ñã<çcÉ9ÖÛ)΢3”œcÁäœië¹ãdB'8k˜)ŸöS®zõ(ALѹ?Îê–ù$'ó9ÁYgz”°ãd>'8ÛL»•'ó9ÁÙãLùÌêcÆl†ý¼¥SÅ~&ú(õÐöqÇ8.<§(¯·ö`¤]q²ÞNqö™Ê-9™Í ÎÂL²•ìíUîN¶6­õq—²%';:Å9ÓjûYn«×æ2‰¾åýsYžéá¿ØoEÛÕÉ~Ü\oYå]?ÿSß endstream endobj 1640 0 obj << /Type /ObjStm /N 100 /First 1001 /Length 1873 /Filter /FlateDecode >> stream xÚµšM‹$¹†ïý+t´/j)¾‚aaÁŸ¼{[|±=‡Å0c–ñÁÿÞo¨RYsè’Á©„¦;»J©'¾Rf•n©¤*½¥Z-.<S\ôÄâ©j)ILã¢&õ”,nÐÂÉ+Ç…¤N=.0°ˆ¾áÊbÆ1C©ñ˜û¸¡Ï_x#‰+ nÈ \c\ÌÞm|&‰ŠqšˆJ‡ØZ[" ¹µB‚מÈ,î ’ @Ш&òÚ%.CâÄ•‚1˜8$%MÌÊ`Á6ƵÄêã^OÜJè¥ÙkÜÁ%Aµ˜…k’"q’:l‡/„¼Ä•$yØ€5 ÌLLãÛ–¤qÌ CˆË˜¹'éóII°f|+ðE²%H1¾å„ÏC_‘¤!$®4©Iè!–´éø¬Á-$€™´ûø¬'+c-ɨÆÌZ“ñEáp²À)¦C•d6dQMÖ<ôPKæ=æÓ–` ÐÈV‡q[#Ž{­¤Æ H“1‹Q‚x!•qj6fóš`ÄÖ)fÉ‹ÄÌÖ}:®<9YxŽr±ÖJrñ az·á­FÉÛX¨ï>"¶Iò.!KÓÔ‹60š¥^[̇©:ù˜ÏS—2îí©GpU…C»qȇ¼èmÈ┺[Ä$”éÝc–[ÄZ¨>+T#zƒ ÓÐ"m3aTÑÇT`suü_Z‹:4)9.q/& Y­c²ø ñ<€VYÚÛ§Ooï?ÿç_ŸÓû_¾|ýööþÓ¿ÿömüÿç_¿üóíý_ûÇçßÒ/Õ üõíý/Ÿÿþ-ýBré‘¡R‡r…FL%7m÷cúô)½ÿ”Þÿôõç¯éýéwÊ¿O?üð†Ÿÿ“é-GQ9™½åi¶`öv‘iž‘Ú(”Y€b‘#0¸–Œº‡éœ åòdN=Løú"´rŽH&mYâ/Qnˆ2Í"ý‡’gA==™Ü²¡~-™z‘)0.ød*eAf-™v‘©žk¬9“i-c Y2íª?=™G‚.™›ôd»döM¶Ë.ô´­JFAÿ8Yê&ÛNæ´í‚Ée“m'sÚvÁ”]þD^b½8‘"¹6ûØWÓ³#£#¸1»Éd”'ùزW+ü0UrÇÒH­æFô¢ÐÕøy,"'rÆÏ)›\9‘Ó®+$mJ“"™úw%ëw5ÿز»Bv2§ž ¦èžT;g³ïJP©Ù˜n-A'sÚvÁÜU‚Næ´í‚y¹©RÑzö™µ—ܻݒS͉<Ô\!/§Ê !¸•ûAÕKf»'QfÄi××Ämás§Y_·%¦UläiV빑ܛ˜“9 »`ò®ˆÌiÚSd“žZaãgqÇÎ2ÿ¸Ð2mÒs2§ž+æ®’‚}³¸W‰áÍÅ}2§mÌm14™Ó¶ ¦ljJ€ªåÙ_ÖÂ0qýØ´¾§Ûstͱó<š=t$íÎ^oW.€›Œ:‡Q@¾hS­°i§dÔ³¨§8@5Ö¬~K÷,Ш³'Û„8÷Ì5NIsí/œØ¯ªH¹¹>™¡â‚'u¯XnX,­{+ýž åˆ$i·Œ ˜ÔÈ袩Þ¦R9‹ñI¬žãüzE¼x¬&Ä°ß“È(²XDVDº¸x{V=‰Òr3]/æ¢(jgo'Ò([];òj_'æ(-ÏØAÁíkäU»FkÞë‰DqãbkdÛSuäYtVñê{ŠÀDÒÚî›üèŠ0¥Ó‘Ší^{µµMŽœÌÓ+f“Mž<˜Ó•KfÛäÊÉ„/W<ßÅkš ÝÇàÅcùx7Я†j<ªx‰³,»ò"gþ/¼+ÿæ« ¦—MÁ3™ÿÓ—k96ñ \ KG}ÚUš`ûá·´W‡Y'ršu…¬›Úä“I´ä]­§ŽŠ2ŽMÕ©£¡”ß«á$††/id›hb¹û£åˆŽ²VºåÁè\6„QÃ#§=µïÅÁÕ.'2T\àv-BèÜì0)¶ÄzKk|ò°¿jí±L ÕÜ^mþiOç8kŒw”ŽJÃX<¬¼HŠMi8™3lVLÞ”ˆ'3Þ[𮞱{Ž×•X;º¶ž¸ÓHIŽ6 ö;:*ÃNä<X!¯.ÅÑr|‡<ŽVÈË›òãàdK¦o98‘ÇIÀÙ÷œœÌ£Ÿ\1wœÌ£ƒ\2Û¦4™³-˜¶©…œÌ³­2eÓäɤµ/wµ¬¬%Ç[¼ƒ§qYoyëíäIÉ¥>ösŒ>„„ïÙ̸!ä¢Ë3nXã!Ú­G“yÆÍŠ¹Ë®“v]ñ|Žã-FyÄ›=ô¢¬›bõd-y—ëÜÔ‘‘‹å;Q×_½¥»tœÌÐqÅ»z¾‚¥¹¢!.Y¬>xlY_íå o2jì‹Ÿ{"Ô€5<#õ¿/ËI endstream endobj 1738 0 obj << /Length 4426 /Filter /FlateDecode >> stream xÚí]]o·}÷¯ÐãK/Éå.·E⢂±Ð<}¸±®cµú0®”é¯/¹üXrvgHQA€&®É9Gg‡3ÃY.ÅÏ:ó?›º³QJ6õúìÝÍ«nþO?Ÿ¹ÿøá/¯¸·7÷ÉÈ?_¼zýF©3Þ±©›øÙÅûÔÔÅåÙ?v¼Óç{)5ìÞ¼¹øÑü·”;ÉÌ?çÿ¼øÛ«o.¢q%D% ;rECœi6 CïI ºg\öŽÄíÝéæp}õŸÃÃÕÝ­Å}ýF'ÃÕÀø íyðŸÎ÷|äî_Âü ÐäCÇt?DK•)ów¬›Ìÿp1sÎÿŸ”|˜¿O ¬~2ë×Të•Ì¨àª…ÑàÜä üÕ xä¬Óª]?Ÿ¦@f*ÓšŠbÃT­]ÎM"hΤèÛ5ðói*Ä-)¾â2 ¾V7˜ÆÍ " LfÝŒ¼]?¿À$q ÈEtŠI=VJF“ÈÐ䶂/åºY„0Ÿ¦@œ«*DÇ„¬õƒ0šF&„bzha~J2SjÅE*6x­~4 L""ôÍÖ,‚ŸOS N½æ2±IèZüèrnA)¦ÔÔ.‚ŸOS 3Ù­¸ =GU+‚M#“ˆcÇú®½Dói*Ä•ëÀ8Nl¢V?º€œ›DDЊ Pœ,¸éûeþFTÈ f-_…‘ubJ‰P1Á †°&ðžÉÞ<4ï Òdû' ú{Â)7³6:í=ô»ÆG¦uˆÛo¿_;„Ô¬ëÇ/Ë!ÈíµPšqnžŠ)<‡Q7nþ>1°Q@ä .XLëbdRªŒ U@¸Ñ420‰ÂÎhÁϧ©·LÖU”ŒkQ+‚M#“XáWq«1RTrR¬¨! ÙÆ Þ-µÃÝ/û»÷ûׇwÇuèµÙûô_r@²EpR/UûJX à+!}b@WBÂ¥f%PÈÀda%´ŠWI%q"LèJ¨!®€¼Ê›Êl*FSOÏ°ßóæÇÃååÕíÏ ³gƒRŸÃBYõP§þL ¦Ôª¡‡âæï=”diíèb4[KžQ¡úHnt87‰8º”¬“²]?Ÿ¦@f*늴[›j üèpnk#ÙŒh×Àϧ©¬¤ÌþGëZüh˜ÄÚH®@k!T4•‹x¾d«!T$20IWÍ"„ÜGR ." U`¥!÷‘ÈÀ$ûšE¹¦’ƒÌTF4÷UŠr…l3ŸRŒ›°àóÚép}}¼¶­±{8nïßßn6:’õ“úl BÌÕ½í®¾À]=|c×ü2áRãê20YpõV¢«“Tr·á‘¨«×‰]@¶®Þ÷LMctõ‡«ÃµoŽóÝÉ8ÇîÒüÇñ´ñŽ^˜ŸY=ƒ Ðc6G“É¥‚GYí0~þ>1°ÑèÎA’1øKB¥Æ_`ë/¦šï”ïà™ñÆÙ ³ÃîžÅÆø‰° ç~|¬ ao1€‡½ô1®{+z`£ž2*T¼u£ À¹I$ÞÏnÕ .Š Ù>É—OqùÀvùiØOËò¹ÝŠ¬’›%¨^VÐcWPx°^ävïY àÞ“>I´;™R©ñØz·å’/†ÝkŠ»ÓÃÁä矮®¯ì ®ß6•&  ^ð&øRNõÌîÃ<à£ßºéûdþúõ`át\¿äüN=eB¨s£i``yE:L¬“c»~~JâDØ8Q§çwê•"øÑ420‰¨“Lë¾]?Ÿ¦@ܦãDd¼V‚y,š™ÃŽÓi6JÞþãûù"9¶êF6õc¥a4‰ MbÇéLLu³a>M€¸íþˆvdëDˆýW ˜ÄŽÓiõU»~~JâÚ íƒÖ‰û 20Y胶Šû âDÑ>h±J!“…>h«±JRÉAœíƒÖ‰û 20Y胶Š›CâN›stßT'BÜ%QÈÀd¡9Ô*Bl‘Tr'ˆ֛u"Äz“@¶õfçj]WožŽï®î7¿Ö0ÑÃ,ž'ìW°¢Ïã?¾ö eàb/Ór£þÔ¬³ú'T¨úÓ.ç&±:¤ ɪ]?Ÿ¦@\ÚEËÏ: bùIç& åg«±ü¤¨l#âêÅ: |ùI¡fæ ågëËO’H‚Tß±V¬`)? dh²P~6Š°”ŸâDhùY'B,?)d`²P~¶ŠËO’Jâb–Ÿu"Äò“B& åg«±ü¤¨×[ÇËÏ:bùI!“…ò³U„X~’Tr—ðò³N„X~RÈÀd¡ül!–Ÿâjpü5|±ü¤ÉBùÙ*B,?I*9ˆ+?5Z~Ö‰ËO€¼:‚Ö›gЩÁÁþP<‚vïÖÅ© í½ñågt^ÛŸ¦ê•É¯&”5ÙJ  G¶Â˜íÊ1ö”JE® s“t®hÖ ä ’ A>½R„+Hd`’ÎÍ"„\ASÉA\é4b¹¢R„+Hd`’ÎÍ"„\AR ®‡‹W®!ä ˜¤sE³!WÐTrÐWc•"„\A!ÛLÑ&•è˜^¿ö#/Ùïn毿®®ö—W7Ç[ÛÁ˜¸˜¿:œNû·¿¹?Úc]‡‡ ¢&&§õm+ï˜C¤òh‘ýü}b`Ãésä¨ìгÉFç„ ýí4|87I|g«Ÿ ŸOS ØQY{~½ãµ"øÑäÜ$ým³á;[’ AHºb+EßÙ’ÈÀ$ým/3slÿÎ61€~gÆ -âðQlÊ¥â;[ Ù†1˜Ÿ_†óa«öÔóæþå Çïñyjл=y-ðä•>TŽ'¯„KMò"­'qÅô8OÚ8dûôüËüèiuMš[]‰´Õ• Ï;tû’p©Ù¾PÈÀdaûÒ*BܾPTö±EØkÔ‰·/20Yؾ´Š·/$•»³#ì5êDˆÛ ˜,l_ZEˆÛŠ qMOn_êDˆÛ ˜,l_ZEˆ€¤’ƒ¸‚ =†_)B̲Íöf+é_ïž·ÿÞH“)u^j‰ß'ÉøCœAöæ“¢‰ô¤hölêO •"€?IS£“ïמŽ‡ëýñW»Ó=Þº;¾~s±ö0³µÔ~Õ‰tAŸæpš@Ãiö8zp%åRNId`’§Í"„pJSÉAÐ{ œCWŠÜŸB¶î¯{&ú1qÿ»ùâŸËKÂýÕÈø$ŸEÔËÔ¾ ø*HŸ± .5«€B& « U„¸ H*9rX\u"ÄU@ KÞ³ ÞÒùÃ7›ž/í ˜á½"ëG³ÿÓí7á…ùä}tY½q‘aª½2Œ¦‘ÉíÐOÒ”í¸é4‘Éö5°êko… £I`hr[Õ™Íß86Kæ¨ä ¾×À·.Û®T td\ûöxŸ oÏMù÷ÝöòWlìÅsÝY¬VÅÈ&©ƒv_nú~™¿±*2$.¨ÎÔ#J¦D¨EáS°À ²$¸`£½â»íÇÓI9’çåPõ£Ï«‡‹‹Á^ðNwv9|ÚßÎ÷Jïþu·ÑsíIù?ÍŠÐ…dc'#óGû¢Ÿ¿O ¬€ Ç_{i;Ð:ãBd)?šF&‘,%5Sª]7½@$ƒ@n±¶¯»ÆAÕ*àGÓÀÀ$¢€2ItÛ%ðói*¹À¹ìe×u"øÑ420‰ˆ0hSêöí"øù*9ˆaí ãÈz5ÕŠàGÓÈÀ$"‚–ÌþRfü|š qm«q'“ö²ë:üh˜ÄÊVŸ=[Eˆš¤’ƒ¸Ž8ž£«DX’4 MÒt£Kž¦¨äðœ©ë± ?ô³µÙ“rß5z»}ðSš?ùSF÷ó1ž_ím5W·ç|÷³kóØ îýÏwîþ¦Ë«ûÞºîÌÓþ?¾^Å‚Ÿïõ‚Ýbv~ŒKü v —š`G!“…`×*B vâŠÑ vu"Ä`G!“…`×*B v$•Ä» vU",ÁŽ@†& Á®Q„%ØQTˆ;X·ù+±*˜ƒC]§™ Š>’}8œŽ—û›ãÍÝiã.ÁÞ"Ÿß¥Ãñéx­Ú]`1€»@ú@Ö71Ì.ð(¹\@M¶Gäîí_¿ýzãV-õ¥÷hÓ¤ñ ΠÛãsƒï¢&Ð.jöp$–R*©&éÔЬAH 4•é„8^'BL 24I§†Vb\ ©¢®ÀÆ 1Æ…ÙCjøhïxš«Y줺êf“/A"ó /b»o,pßHŸ”Üìf¦sX¨@1…Ž³QŽé’¿ß¼ë[ê‹>t /H{5´À«¡Tõuó6–. —šjˆB& ÕP«1ìQT²âcŒª!†= ˜,„½VbØ#©ä nÅhØ«a {24Y{",a¢@‹›æ°W'ÀöÄöú‰Ï·º°g]’¿©þ~cg('Ƨ?ZoaûZø¹ZO %ó±h™tJc'ÐR&'ÐH``’>Ö,A8FR È‘›p\¬R„pD&éhÍ"„,FSÉA/qCÊ©!d1˜¤³X³!‹‘TÀCÊ©!d1˜¤³X³!‹ÑTr¬nõ)§N„˜Å(dh’Îb­"Ä,FR È7Ù67U 6bÁöž¾øœ›fÌ¿÷ô¿Úß¾r|wÎw[‡Éí ãË»˜ÇæÞè&O 9<Á˸—ù1Ä;¼”GÉË"âeö€P¸màÇ¢—}ºº¿¼»ÙÜEÝô,v‘öËwÁŸ°‹L àù×)ŠL¹Ôä_ ˜,äßVbþ¥¨Â¡ÈJbþ¥ÉBþm!æ_’JR8Y'Â’ dh²EX"#E€‡"+˜##ââü¿Ì¢ÞWþP÷éî'“e¯7þh÷§«‡ëˆ8)6ôŸÓIIäù!šŸob}¾™Ú|³Kò(<_ 1>ßôƒøìùÞÿvÿp¼Ù›Güióe²X÷YRHÔj½o F-ê«‚üB¹bêîˆQ‹B†&é¨Õ*ÂâÕâò×æ›@R€ÿƒµÃ{ endstream endobj 1641 0 obj << /Type /ObjStm /N 100 /First 1007 /Length 2336 /Filter /FlateDecode >> stream xÚµšA‹¹Çïó)tL.zR©J%YØq. ˜µÙï–,3a<†Í·O•^ÿûyaž Q?clMO÷ÿ×%U•JRgé5¤¥kÈ©{£¢q¥‡RJÈ5¥Àܽ‘ƒÔì UÕ%h÷pèiÜ#¦CùÎZ5ä\Æ]öZÈ¥ŽìGVoe‡µê-×î~_¦›ýCÙí”CõºØãÕ%{—»¬Ö¢ÔÇ“%ªþÛÌöÖmèJ(ÜǵŠßb- EiZ(Å[fgf™Ùœ†]FäÜ‹YA¸ ˨X7?AX8yKW×j`ÕqM·aµ ÉúÊZ=HÎãíS*ì-ëÍ"É…‚˜Œ_+A¤¹ŠÝ"uØV$ˆ÷…µjN®Â6‰]…K0ûü ûÑØþ„™UKw‹¸†*‰Áª›o­fƒW\φ¤6ök’‚Á¼‡$=÷PÐ<Wl”Ï} ôÜf´ ;CjÐ*j Ñ :ÆÔJ[OôÐRòkö’-7°ÎnTü+…VÆ(¼ñ°£rh¢ãš„Vû¸VCkÉƼV ­Ó¸ÖÌåʸ¯‡ž‡ÙtªÞ/šC/:®QèÜǵúp¼ªæ®Jãš„ÞÆXZ×™[Uc¸ù)©w‘Zo§Üü­],•aŠP2ñ¦uA’ñÍHÞÖ´»’V¬%5ÿ]m}¸¤û‡ýñ‘qñ|îŽÖ&= ëM"¥—“xiÇãÜBq<*Ï@Œà¨ à´œÁ~7I½l¶i&÷c8ýí§¿[ÔJš‡¯¿þúñºéÍãÃó€½±rl»÷U4º5­‚+h7«ÃÊÖ¶Ò³YÅî?˜ÒéíÓãçw÷fT8½}ý&œÞßÿö>þ¾oÞ~úçýÝéO¼xþbSäùݼ;¾<~}ú|?®Õóµ¿Þÿü˧;wšWóÚ½gÞ~z²§=uåó£Ç¿x”ðþ>£‚ß‚FECÑhhô­A Œ¡e‚2A™ LP&(” ” ” ” ” ” ” ”73nfÜ̸™÷›ñŒ×`¼†à5¯!P( ”Êe²@Y \¡\¡\¡\¡\¡\¡\¡\¡\¡\¡¬PV(+”Ê e…²BY¡¬PV(7(7(7(7(7(7(7(7(7(7(w(w(w(w(w(w(w(w(w(÷MYSB#£Ah4 A£¢¡h44 œ¡œ¡œ¡Œ°R„•"¬a¥+EX)ÂJVŠ°R„•"¬a¥+EX)ÂJVŠ°R„•"¬a¥+EXéVÿÿ|mkäØ|9lK‰óÚXl´ÅsO±Ô—§ÝÅY×—É|oGÚŠÆ}h†ì‹È–c±ñØ‘†²¤0#æÅõLîÕPyGR’X,¾¦LZcR¶ž5gÝ™”¢oªL™‹>‘ÄV/£IEltó”I‹‹Sòš×Bygr¾4e.VR$Eú…Y9ªåæ³,®iHSät NÒ}þžçªy½æ ÌÎVÉõ9sчrj‘-ßÙ|ǦƱEgæ²¼\:òj"[òkÛ‘…"{š!W½¶il6ïXõí?²Ò  R,!Xh29hÆÔÕŒ°e”âùC–Ó8‹ußzyþZÝ *Ñ÷{w"rÐu¢”@LÕòÍe§”mýWošÛôÅÙº²ÔKþ±`½òÉÛALd  syû)L¤  sùD9Lä  ³ÒA.&\hÂÔƒ¶bØOú¾Ù„f«à¥Þ¤|F×+b_ŠNã ð6¡²uí΄™æêÑn§Ù'©ìv–nYXËmís³sÆ\=•b[þp»¾±½ƒ1á‡ot£ƒj¶q,¤$3—eŠ\N|\¼3·3Í«&õqñ8íK—áü¸OUoÒ¯øngº‘Þ²ï©íÇ‹ùœ†l¶ôßý|1W¾ÉÚ–K‹"$kTÿ`‚\=ûç±¹®fµ¢À¿˜0WÏáñèÎܾ2w;ÿ Î?m8 endstream endobj 1970 0 obj << /Length 6974 /Filter /FlateDecode >> stream xÚí]ko7’ýž_¡6°—Û|ôk¿9~L¼qâÀ‘wfw±¸ÐHr"@/H 2™_¿}»ù(V³Š3‰u%c€LìÜ®S<çT‘ì§üêßß´ílÄØŒòàð uxrð¿Ï^þ|t}wzó|£µ~&ÍìþÅ<{wö÷›çòÙÑÍoËŸß^žœþãùFumÛ<“Íøüÿÿó«×‡¸Uª0ÃÝ/W)öSŠ½P²56Ån0Bj³¤(ÍóMÛ©gïžoä³)±£)³9±¾µ‰MÙ,AŒÐ¦S» ¦}ÓÛa.¿Ñò`c×û“é€Fêå'ÇÛ£ÛÛ«ã³£»Ó“å×øq;ˆ¡ÑÓ1óo'šd/—ÿSÑŸdþ¿íÅH`5ôb˜˜êÔ(Œ4³È¢þ›hÆéRÍjÇewÇo@€Yz8Æ 3ÕƒYå2ö¢Qc” m8÷ërä…çäôÇÆ¢¾Ü¾zÿñëw¯×FУeÿtŒ@ø¡W;FÝ%Ê\Š¶ß§«EÄ)87îêSp y ."w©Sp0—‚Sp,2 ÉŸ‚«&ÁûšM%ɵ·2¼¯2öµ;16”¯·7êø1îŽÜ¨«HA#jiAa.‚bä• C'ôˆ½¾º¹Ûþzv{ru±–s”¢ëäþߧëh·ÔkÐÚB–GòêÌ¥D[„¼Ò¶oÅ0DÚ~ººùéôéhk ¨×6 µ…,3Ú‚\J´EÈ+m»v¾«h{v1×íío·w§¤ÄRñ9¯PRÙñÔKÐRAÒ©@.%R!ä•T­C—’ê©”¡% ^Û€Ö²Ìh r)Ñ!¯´5Fè6Ööòìn{÷óÍéÑÉíÙÏ;êh!ÕÌúäR"0B^ ¬µL$ðîDͼڕ‰g=‡fJV>’==%·å¤^ÏÈ r)‘!¯äVZh–[?U¹-'õr‡´ÜxFnK‰Üy%·TbPi¹§½l²ÂÁvÖŽº^ЀRË r)!¯m”Ð’T=NAí¨ë hA!µÌ äR"(BÆ‚ªi-|ŽOìÌ"b•b7Q0‘‚}‹‹Bòû²jܾŒO%Y:µ¤V…$¸5"‹ŒBòû²jÜtŦ‚@2ïŸ($ÁMW,2 ÉOWÕ$¸ÆƧƒdžï*$Á56Œ¼jlƒºQQc»M<Ìa´h[ùåa@æʧ£²Þ%!í¨Wbeç\r)q B^¹¤×¢ؽ¿zýãá‡÷ÿ½}ûýÏŽÝCzÇ Õàí¨ê|@7xHêÈr)ið2 ™iðµ$øÏ¥‚@–¥xG6ø2|ƒçQÈLƒ¯%Á—.›J ’kðe$øÒEÈ«ÒÝM`ƒÏ<œ+§…ðçϱ(µƒ©^”†ãÉE)ä‹^”‚DJ¥. ™Y”Ö2à{›J BM2®Á”‘à{‡ŒBfzV- ¾gq© \Ï*#Á÷,…Ìô¬Z|ÏbS‰A2/¨($Á÷,„¼êYSIô}´Üøæð˧gø‹Ž´ú¢袀ÊHòÄ"Ì¥¤(8d2Sµ$ø¢`S‰A–ûÝ艼Œ_yUF ÕÁ¢˜–ào¿{q˜~——y Oäw¸ÑUß wDj2ads+ 6 HMìÓÎzÚ W`ç3A Kù®O‰w“U_Jý5ŒB$tÃüÄR5 öøL*1uQÌ^?,$Á]’d‘QHª‡í®{·õ$ØãùTÈ℆nbe,ø&† WML/÷›…Ýù&™?ãÒ t³{´fôYÞ[{üX+€@ˆfÔôb4}” ÿýu8I¸ÐÞ?WÍ»IMP÷§Ù;ê Ip7é±È($“^5 î&=>•dYN ÔMz…$¸›ôXd’¿I¯š?5s© e·¥É}w ~ßÍ!£™}w- ~ßͦƒP[N·I.#Áï»9d2³ï®%Áo1¸TȲåÈ-F ~‹Á!£™-F- ~‹Á¦ƒd^Ý_H‚ŸòjvVR(“˜Ó/øê»ñªmÙÁÔ·­€n[1=’m äRÒ¶8d2Ó¶jIðm‹M%Y*¶%ÛV ¾mqÈ(d¦mÕ’àÛ— É]Ã.#Á·-…Ì´­Z|ÛbS‰Arm«Œ߶òªmÉFôZ§ÛÖöXÝ$o¹ÚÝÄõ°¿_êœe‡Wo߀¶/äP“lÁ\JìË!£ûÖ’àí˦ƒäì[F‚·/B^Ù·i„R”}“¯ÛÏ{75ÛØ‘ÔÏ6!=Û@ºNuSÈ¥d¶áQÈÌlSK‚/W.’y¾²_®2 ™)×Z|¹²©Ä ¹r-#Á—+BÆå:Œ¢—©j½IM4ì ðvðÕÞÇ“Ö†üÒH¤ÀØl·uíð«Ù<"ˆŒ§Ë†ï,`±£ûQ¨&åèÛëó³»ôÞO6»—È=ì·¾"ÚáV‹Ž'E„Œ&÷ Pµ`骯…€®¨‰”d-€\JjAÆ!3µPIB¨.B¼P`®…2æZ@ˆéZhZÑD÷Ì/ßo}¾ÑF?³'Œ?Ý\=—Ï.¸ 3í4r©þàóÆž'›t½!-d†8qòȉ“b˜qê8C¿#ûÝc9íAõŸÚ©ˆ^áFPÝ+@²WD4É‘ê0—‚^Á!ã|¯¨%ÁÛ“M0—n ØÙ#¦í9hѹ¶çc=Åç\ïæ€v3d•¾>s)q3ƒŒCfÜ\IBp3— YšmrÏ_HÀìf„˜vs¿»jÒRnÞMzÔ-#ZO«Lýû­çÍ¢^œ€2E´GN„˜§SÂhC­Fä´É¬zÚ‚êK>õÀO÷0dúL?H¤¤ а(`¦Ô ?¸ŒÉ#†à燰G»A(cü0îíQ{üX{_y0}?úåÂœ'´¿æ‘QHâ<¡½å¢šw_› YH O–’àN–²È($²´šß­ØTbå¾òÒX ¡[1È8d¦[U’º— !>e3w«2æn…ÓÝJOÉKø.˜³«“³‹§y—#£¾ÜCºÜ!ãš.wKI¹sÈ(d¦ÜkIðåΦƒPw0¹Ú,"!”;ƒŒCfʽ’„Pî\*d!!]îeÌåŽÓå.G1t=.÷Î<åo¶ySZrêз hz#r)q>ƒŒCfœ_IBp>— an`,$`v>BL;¿…Š>Øv10;a|=íÏybQ B‡TïíÃåð 8~mÃbqá*­Ä(‡(f²¿æqQHbÒƒhÛz–Ã3‰DÄÅyczÑwm)ö×<0 I00¹O}=öx>²Ð¯·‚JtSüBì¯yd’ ÁmkIð{P6•„¸CÁoËHð{P…ÌìAkIð‹R.²ìA{rQZF‚_”rÈ(dfQZK‚ŸšÙTbjûåæÑ"ÂÔÌ ã™©¹’„05s© nj.#@%zArjÖãÔ’ûÜ°óÝë?~üðšøêæ>íBWÝDÍßOwC¾·Ÿíñ Ñ_bâ&<Û…`*ŽCò­š×ØØTHæIÝB\cc‘QH¾±U“àŸJ B½qÂv¡2|cãqH¾±Õ’à› Yº{òR@!»Æ†Óm˜øWð½×gÛ¿ß\ÝÞíÁ ?nõö h{Bš:òFY˜K‰=d2cÏJ‚=¹TÈòhHÚžeÌöDˆi{NíZ ÙÓ~ÇüѾãлΎ¾ÞÚ!mmHñúÚŽ÷!È¥ÄÚ 2™±v% ÁÚ\*„¹¬QHÀlm„˜¶v׉NÃ{À¿ûáíöÕë7/>¾;Ü~ýîýËoÓ/;Ô­ùœm×­ZlúõK£€^AŽè=Ì¥diÄ!£™¥Q- ¾@ÙTbâƒ$¾šŠHÊ ã™­$!(— Yæžä¸…ÌŠÓÚ¶bš{ØïPìÃËܽ­ìðê½ÐÞ…Òç+`.%ÞeqÈŒw+IÞåRA ÄW)fï–0{!¦½kZ¡MG{—xùªlEÿp>dJ‰fW/Z@‹$V œh1-š6¢o$#ZúM¬û!š\½h!-d ä‘ !¦ESF4mKˆv³¯Ÿüð ܯ~–èYrÈÌ —’Y‚AÆ!3³D% Á»\*d™%’O{0{!¦½+õüØ»?Ýý|zCŸùiÚiídØǶ¼—ì˜ê І…Ä1§ƒ@.%†eqÈŒa+I†åRA Üé 2fÃ"Ä´a› ºÃ§ƒv7/'¾<½ûš‚|p6µ×µÝHª/žƒäÅóˆ®Äm»öJ7Ì¥àâ9‹ŒBòÏ«IpÏùTb…„–¼ÆTF‚¿ÆÄ!£™kLµ$ø)\*„hXþ¬G þD ‡ŒBfN¤Ô’à»6›J ’¹»Œ„еd2Óµ+I]›KpKä2T¢$»¶š¬ 债öùÕñÑùöö쟉ÇçFYþ’å|ßqî¨ …Xî’¥€yd„ňia)L¯Ha·òdïÖnLÕ ;QD\K¾?æRЉ8d’ïDµ$Ãr© eý˜¼§€Ù°1mؾƒêh꓇|ÍÅ¥_½TÈ¥BÄQK.`.K…ä— Õ$øeS‰Aˆ O¾šŠHÊ ã™­$!(— áf”2æEˆéíšy×Æèöîæèòöúêö4U«»Osš¾îó N{›C½ÁBÚ`p ôE=˜K‰Ád2c°J‚Á¸Tg°2fƒ!Ä´ÁÌ(ZÝÒÓ'⃞b;ÞzCZGH*¡#È#§#BLë¨1 #§ãß(lõ" (³T¹”4 ‡Ì4ŠJ‚Á¸Tg°2fƒ!Ä´ÁÔîô‘¦ vqtù[êbœFŽ‹q”Ií8êMÐ&…d1Ë%K‰Id2cÒJ‚I¹TgÒ2f“"Ä´Ie/ÚqàMšÞ…OáG¥>¿O={v(õ…´D/bË òÈI„Ó5[•‘ˆªº]J)ªì-l}Ù‡tÙñ1‹XKIÙ3È8d¦ì+IžâRA \Ù—0{ !&=%ÇnÞ0“žbídÔ4Þ?x1K⯠‰ØI óÈ‚Ó‚ ­è°»W÷§o¦°”óE%^……’z¥ýñ´Ð€õt3IädŽáÒ*÷­hdKÈœœf'ó4Þ߇=Õvèõr†´ž_BPGNQ„˜–´3Bw#%©:Ùï»·ÜðªW ¹ ˆ8lÉ›a`.«‡äWµ$ïr© λeÌÞEˆiï¶ZôJSÞÕ¹Ù¡×KÐ’B~ IA9IbZR£EÓ÷”¤é{¶›F˜®Ý×\{¢í¨ëÕ h5!µéH`95bZM­„ÑŠQ3}½q/îçvƒ«- Eƒ ¢> stream xÚµ›A·…ïû+xL.²Š,V‚;‚rIÁò!Ž ƒ"/#Æn ­çߧŠ;¯wìòàé ‰êi¾¯YUMv·«ØH%U1MµJ4,Sª£”Ä­E£¦&%”úhp›ç´¤ežÓ“ÕÙðß Ù·†+¶Ç\›úˆ–¥Ê­ê‚mX´jªÝj´¦äü•SUâh¹¨ñ<¯'*}ž'‰ª gÔ‘ˆtž§~Ù´êè!:¨$ =ª‰F‹óˆió\žlĸ¨%.£ž˜J\Ib®ê àÆ16ÒĽÍó,±tÖ`Òñ+×Äj¡çák%1˜S«už×R#ž¿öäróWK­õ`°x+®ÞÄõ9¢V“´WÚ(IŸÊ>‰X{«%QšÇº'ƒCÏEýǸú6’G)¸M“_wsF³4ü²üX/iôH¶ÿ™†Ìl{·1æUuNCekiؘ‹QvIJuªŒ¤L³‡&mœÑ-yp£‡ŒÊŒ‹Ô¤c^½‹z€¢‡p²BÁæ…ó‡—õÙC’ñÌŒ—Šµ™ŒI †GÈÆŒ©Ÿb:33j2ëÑc¸¨Ç>º bñ2Šf›U9;yôJ›µµ\úLÊð*)2Ç.~­ŽŠú,:# ~V1‹Ä¨×V­QÊ#Ä+Q$$rV¹EÝi›•…§þSícv“p˜Îna yÔðQuZüD…£[„‹jná "‰na9âPq!Ôlõ“Ô‹ ¢(Do:Íó>Åü2™GÆe´›W¯nN?ýï¿·éôýÝÝýÃÍéÝ×=Ìÿí×»ÿÜœ~¸ÿüËíçô¾øäP>Üœ~¼ýôÞ7å\ü‚Djnî©f”Ù*b¹Jóó¾O¯^¥Ó»túëýO÷éô:ý©?§ï¾»ñÿþ ÓF6/0{9"³bJ½ŽÙ+åHÝÆôR]ñ†^9FÑLáè:ró²hÃÿv/ õÌ£?ˬ×2‘K0‘˳•ryfn¹\0¥í”K0#— žòN¼"Ù ;qµeµç#j×fqä¸ww“Üø"‹¥ùßÏ÷I"[ÈÚ÷ êÆŒ¨.x×rãiÏöˆëÆžÏçºS‰v·=ù}eâÜžÌrl4Er,~&oPÖ^áazë½åXå`zëB¹É #ûLoólŒ“i'gœ™pÆ’¹W퀹\ðFÝ)®µäX¿oq­š} ûü`§Û˜[\WÌ–3âºàÉØÇ#¾¦ËæKºÍ#>V{aÙqýÍã1—¹\0¯+r æ–ËSdŸ\nÌÈå‚g×.­X³ï˜šZî%öx#Çî&†^øù)¶_‹ì~»ð…þÆô)@|?½b^»JÞJÌsÉ.™´SÉ‚y.Ù%Sv*Ù3%»dŽJLZçrì´jZ%–Gž×T©íÛÖV;…}zå§Ú©¾ãj|èŽgc¢v̽v<`nµ³`îµãÙ˜‘Ëo¯{­<7†FGCÐh(vnô‚”;”;”;”;”;”;”;”;”Êe²@Y ,P( ”Êååååååååååe…²BY¡¬PV(+”Ê e…²BÙ lP6(” Êeƒ²AÙ lge-Š¡Áh44:‚Æ@CÑ€r…r…r…r…r…r…r…r…r…r…2A™ LP&(” Êe‚2A™ ÌPf(3”Ê e†2C™¡ÌPf(7(à *<¨ð Âƒ *<¨ð Âƒ *<¨ð Âƒ *<¨ð Âƒ *<¨ð Âƒ *<¨ð Âƒ *<¨ð Âƒ *<¨ð Âƒ *<¨ð Âƒ *<¨ð Âƒ *<¨ð Âƒ *<¨ð Âƒ *<¨ð Âƒ *<¨ð Âƒ *­3·LõùW„¶Ó}¥óÓ½3^xa¯ºÏ<ÔJÏñÙæ¡Vý^f/0m§q‚‰q.˜{Å–Mrá‹Øº_Efj›_ЀÏ{ëxþ«;1çØðnL÷«ÈÁLáÌzÁ~ÿì3}%#6¦ø½¬ÌôÕ|Ó§ºå^|=$ÇÖ׊6&›¯‡fRÍñÁíÆ$óõÐÁÌês<_0«úÚdË,%ǹ³ø½ÌÆ¡5DjY/vöñw׃Æé1¬ê¬Ö3‹&"ÊñdšºÏ·lG|ë†Mö†›ìyÞd¯˜×~Ø‹[öÆ<ß²W̱Ó-{c¢lÌ«&ør ¨žkOeÛü–MÕêLÔí‚yu ¡nÁDÝ.˜×~‚²Õ-˜¨Ûs§‡Cu»`ߘ¨ÛóÚq"–Ôr7Ï£/ƒ¬E~)bþükïkCëwXI‰Ð.;mÊ6$"»@Rß©‚|fˆ·xÛ0}¡"å&ï4N01ÎSt'fñ5]0«/¨ùR”š5ÞødÿØy ò•Q?¦bÏÐF„i^&Ò6çýäG" endstream endobj 1972 0 obj << /Type /ObjStm /N 100 /First 1001 /Length 1828 /Filter /FlateDecode >> stream xÚµš=$·†óýZ —d«Hà @†'Š$e‚Ûîáèßû-nWÏ; Ül`1ÛÓÝäÓo}ENƒ2•žñQšª<ÏTb™g˜Dç¡fóL#óŒ’_ÆÑ(Ýð‘™_p4Уø¹‚[«úm¥Pa›ç*ÅpAÑŽ„ŠÕyÎ)l~„ £U?2ªYgÛNµt£ X’š©ŠßÜÑAmì-j¥ªó *S5?ª½Ïê³…—2[q­³E'fqFn‡á”‰›y/\ˆµÏs°Ò› ˜‰Gñgf!ÉÓ†ÜHŠxϬ$x¬7YÂ:Ûvé „·•Lbnº.…¤Ok ÓÂÔòìEà‹¢n&ø¨Õ©H”O›ŠQky¶èÔ´º?dÀ{ì-Z¦æ8*Ô†z ˜X³¹5н–ÙKRžÚ.Ó)MIÛÔÖŒTe¶è¤Ö:0“öi!ŒŽîWÑ=ù³h%¸×ûC0WïN1§AŒµ©M•L§652ëój'ëÃuÀtˆV·eBùU+;Ã*ÁêÞ‹1Á"ó>!<㼯ÚÏû¹3ôºõ1Ÿ9²¸?l œg/0ì¨ê-ÐÁàî¾ÄÇáQÒ™`>×Û…†M»ôF£O»@þmÞ‡r¶Ù¡'Hé ĉ£=›€vŽ›>ö{hYg|‡\Ä/÷ÙÙÀ¥ÿëýZ­¦¡ž‚5Ídë5!#ˆ9µÛ~ OŸèõgzýû×_¾Òëßè/fßÑ÷ß¿àïÿDvKÓ1–2rpìy²s$~ ËÈ)÷‘•/"[Oê0Þ¥ŒEˆ%öä´‘š}ŒÔ‹Èðe _.ª› È0ìsäe_ z-©"/j«Iæ÷‘²~H”Mž bxò9ñrð„'ž|ŽÜæÉ@†'Ÿ#/¡²•· TÂà­}ˆ,¼Ie Cås$_ŽWNPŠbÛwGüJGÆܯ Pyµ&Cݳ –~©=±Ïƒ4K³ŽzŽ¬íª'aVŸ ž\¨ÌW]™™cÚÈ>cY¨ÄPñ¦Æåg8Óëé@V8Ó냲֋ÌÚ1K¶,•²drß”%ÁŒa}ÁÔ]St0c`_1wÍÑÁŒ1oÁ´¾iÐ f¤Ê‚ÙË&–‘d¼óguö“›üÌðç‚im“?ƒþ\1wÐÁ .˜—ý:sOã}ERÔ'›{ü:ƒ:ÌË:í]eEêÜ$ó0m ò ¤m²l ð ä®yèCåhá{Ã'ˆ‡Èñ²ÆƒˆêÇ÷¨`Dï=[A5ÐçÖMöm¢Šãá…ѽ<Ì$†ŒœÇI¹ªðˆiÉ£qX¶1æ‘[6zZ®©ùCò4êi›ŒH7êwu—çä±ø¨öÆœ“'»ö—óß’Äi•ä?(^e”«¹ÞáÅ35äéŲì2k0ݬ ÞÕ¥Scaüouë“ß JÙ¤1˜®qÁã]3UÆrŒ¨%c=Pn‹g„Žw³FΩè“ÝóMåÿÉt žlš5¤dücÖ>|Ï÷Ö‚ãdBãŠ×7¥¢XAúñÏP'›ÜkSÍž ›*jÈÚoŠO¦k\ð¶iÄRØßH956‘½Þ7Át ÞåJÕí'¨Pk¸âñcÉ_bAÝÁ¯ÇÛ¦uÕÉy®mÁz¼lñ?íÙÈÆ endstream endobj 2071 0 obj << /Length 4417 /Filter /FlateDecode >> stream xÚí][o·}÷¯ÐcTÌò²·¾9±œª°$CVEñÁˆíD¨¬²‚¢ÿ¾{á’Ã!gHÑnm}äâ|;gxΙ!¹WyÐLɃ±9èµ£~yÿ¤Yþôæ׃õ_Î|"í蓼_~ñä»çm{ 16£<¸xC]¼9øÇ7R6ßjÕµÝ7ÏŸ_ü<ý»Ößh1ýõí?/þúäèÂo•*Ìbþe”†–ƒ»ÎØ,ºÁ©ÍšÅ»w·ÿÞ½ÿp¹ûpõúz÷þõõvoÞÝî~Q7sß=À±RѪ i9òÏßÊ^®ÿPÁÉüÿ«:Ñ¢†^ R»ñÌÔˆvú¢§¿Iµpþ $k;þXS`Ìdz¯¢\Æ^4j r¡eÚ~!GRõZ´zd¥ºQ¿Ü©ìxê¥òh© iŒT —©r$U§¦ª£«êöæõõÇ¿|›K17‰Ï¬w¯fD—è5°Ç‚±dP§c?(!äÂh`Í#£ òŸH#Ö’àŒÈ¦‚¬FÔ¤ËHpFDÈ‘[%´Iö ΃ÓLÕèöë雑ìhêÝêÐn…”1n¹”¸•CF!3n­%Á¹•M%ɹµŒçV„¹ÕHÑ·¶FæuÐîäåñîÕçOO¾qôlw|{uÅøE&7LI;)G7†;Ëc?byÈêÑ&ʥ녞êæÂÈcÍ#£„G·ª%Á*— Y=Ú…ZF‚+T…Ìj- ®PÙTB•Ij ®PrT¨Z ÙŽd¡žýt‘˜U#ôôÿXªTÚôëkÔ krÄÔ(È¥¤F9d2S£µ$¸åRA ¹-#ÁÕ(‡ŒBfj´–W£l*!H®FËHp5Š£USþžL/Οž¾zyöŠ˜M¿¢µé×ר@×(ä¨È¹”Ô(‡ŒBfj´–W£\*$·à-#ÁÕ(‡ŒBfj´–W£l*!H®FËHp5Š£m¦Õk7Ò5šžH§íÙÐèÿm‘Rei3®/K€.KH S– —’²äQÈLYÖ’àÊ’KäʲŒW–2 ™)ËZ\Y²©„ ¹²,#Á•%BÆeÙN›JÝkP–§g»‹ã“£Ç'lj‚¦z”òË.ù¬roœTË rÄ«Ž’æR 7FŽäÑ÷ûåÓ‹ã£Ó„ÒSƒú¢'ÆŠZµÖS’Ú ëΊÙãA€X1²¶ê8•^L{É F0ûë p’hRF‹©‘Ös`çSA 땼xºj•0ÃPJ‚ý5ŒB$Øé´š„mÎæS A¨Òµl! ÛœÍ"£üœ]MÂ6g³© µ$5g’°ÍÙ,2 ÉÏÙÕ$¸&Φ‚¬ C6ñ2\GÈQï'\阯rÄ|®Ù±Ý“¹ºhï”èUﺳööøC Q‹!ÈB®!›H¥¤ °ÀaÈL¨åÀ5.²vBE62\àQÈL¨%Á56•dIehÈ&PF‚k9j“¦FšÀ|gDܺ™3¹— rÿÄJo;wwÿzü!pB,7§‚TJÌÏ£ó×ràÌϦ‚¬€žËHpæGÈ‘ùÛNŒ£N˜'ßÄþŸô2Zß[ÿS·$Ô{Ü =™6¤ÇA*%ç€QÈŒÇk9pgS Ar/#Áy!G7ÐcªÁïTÊãí½ð8µh³ã­_´ùô¢ ’Ú’‹6JÉ¢Cfmµ¸E— ¡ì¼` ®¦9d2SÓµ$¸šfS AÖ³­=YÓe$¸šFÈQMëv¹œ‘¨i½—5mÇ[_Ó>]ÓTº¦A*%5͇!35]Ë«i.’«é2\MsÈ(d¦¦kIp5ͦ‚¬5=’5]F‚«i„Õ´j…j’ótòi‚qÅrR/·@Ë ‰W)7È¥Dn„É-覴Ór'·c+=~ÕŠS[ ;Øú-†@o1 £=Ùº@*%­‹F!3­«–çe6•$çå2œ—räåƈFö”—SÛŒ±MwŸ¶Ò”ªvèõªú´ª_FUK‰ª«jF-Œ";”ÞcU·¡W« ªüÒªÂ\ TÅÈ‘ªSÓT²V“O R ÚÜ¿e1m@r& 8&g"˜JÁLÄ£üLTÍs7›J²º›Ü’àÜ#wOû¥áî=[Umƒ­÷²@{2J{¤Râe…Ìx¹–çe6•$çå2œ—räåNŠNS:½ªº¿^¶ƒ­÷²@{2J{¤Râe…Ìx¹–çe6•$çå2œ—räåVŠÆ}Y§½<4ý÷2qFrPõI€<#°ÖSg$a*g$yà0$F²šƒíŒ$› ¡üºUX ®h9d2S´µ$¸¢eS A2g$ IpE‹£¢50&š€~ýãæôý!S¶ªÛ›Ëã÷–…z}Z`HµnIA.%#äH`5Š¡•”ÀÄ[l#†®»W‹Œe;Þz)}ZJHª&oȇ¹”H‰#)å(TÛ‘R¦ßrs¥´ã­—Ò ¥„¤2R‚\J¤DÈ‘”Í º.]•7É+A{Ùv- õû´ÀjM/†A.%#d,°ÑtéZýøáêò6=»JÙŠ^ö_‹Ð„nÛàªuHÝééæR FŽt›ÐM/y݈ISk¡•þ’‰oÔÙAÔëãÐú@¦èÆ s)Ñ!GúôúL]3á×£D½>>­dŠÑäR¢BŽôé:¡†dýt&Ýóöå‚Ò6ôzU}ZUȯ&/(Á\JTEÈ‘ªm+º!ªºíýŠd¹²±P/°@ ©Öä£Ý0—r$°içÔ }÷n“h‡[/¤;žÖ0ªÉ7ÅDJT a#µflI÷m¸·^F€Ö’Ê r)Q!GR*-Òåøpv‡+õân‡ÓÒz’a]%²˜‘¨rZ€6Q}&í¦@½ÚÓ;?íiÿêk ym! ]’¯Zƒ¹\[`‘QHþÚB5 Îçl*!Èz…¼¶PH‚s;BŽüÞ(o¥oˆ;_û^Ⱦ߃›¬ïÖÁ×{ÛO[ðË8Û'Rbl6 ˜±uåð«¹<ˆœ§‹†ï,ÂbG«q¢@¦:xú9»ý²ô6újOƒ¤©ŠiWÃ\ lÍ"£¼±«I؜ͧ‚d¼]HÂfnŒ¹{óæÖÀÜëàë½íŽ§­ øeœí)16ÌغrøÎÕ\DÎÓEÃw–a#G÷UÔ¯ÿ}yûÛîúö·›·¯ß|¼?ïùßFSïQ€6)¤¬¥] r)±)‡ŒBfŒZK‚s*›J²¾«‰¼fXH‚3+Bܪ¥™v|2üˆÎúÖ½ó£WGç;ÚŸ¾L½÷tlD§‡{øäo;-¶ZÓVœh;žý8Ißg×6ýòøtÙ·‰¶_g€Ãi[·rZpŸÀ=žO/£›¿ŒUHÀü -öÃLÎν]œÊº¹¼¾%ÞB6õ ¦¡œ$–ŸzÝ}Zw(BüÄ÷¢;È#§;BLëÞiÿ¢«¥«¿üxïœZŽûüÞ9ÖfšÆ©yotÝÕÛñ‡ @d b?­Gö@KIdqÈL¬$Á×— YOû6ÉZ(#`©„˜®…v‚nà=¶óžê_—×o’¯_ëF}ïwUfÐËçÚ¶¡ßÙ×öøC ö5!6VfÚöê¦ r¡eÝ~Í#£i_»º«%Á7›J².YYÜE$øâfqÈLqW’à‹›K¯W\Š»Œ¥bÄtq›ù / 'º£gÏ/š&ù–l5Ê}¹LçªÏP_â>]âe©È¹””8‡ŒBfJ¼–Wâl*!qÅÕc ¾Äd2Sâ•$øçRA Üü]FÀRâ1]âzþìW•¸L”x/Z¥ÂƙҒSï|€v>T@Ò“È¥Äù 2™q~% Þù\*„¸3lq~Jňiç+)T?bçËæÑùƒ#§Þù>í|¨ã|K‰ód2ãüJ¼ó¹Tçü2ç#Ä´óíǤ°ó{þlJKN½ó}ÚùPÆù —ç3È8dÆù•$xçs© Îùe,ÎGˆiç7Ð ~öçüŒÚÐ<<ç[rêïз 0ι”8ŸAÆ!3ί$Á;ŸKpÎ/#`q>BL:¿æçoUäüÇž¯GNµóAÒù´óa.ÎçqHÞùµ$8糩 Æù…(#¦ßÏŸ‚ë°ó×ù‹)-9õÎ÷hçCçƒ\JœÏ ãçW’àÏ¥‚@8ç—°8!¦ß bŒ|ÿØñgK.ÔÔ»~;œö¼çžq¼Ë¢Äï$f.ãõª{§ÓIœËK½x<ÀJ;¼„Q\}üøövw{ùþíÕåûË}vÓQnY¨WÕ u…T§ï¯yä´EˆiuÍTKîÖ2Ÿ²ßOu- õêú´ºê.]· œº1­®žß† Òê&?‚>JÑuêÞß?à8·Ô ëÐÂB– aA9abZX5ЩÂþtúôÅñ§GÏ÷„ôbøê4Å‹†i"RƸÝyÙ`?â…!ú¬™¿áÖŽA.ÌÕ`ûk…$®Û«ÕÕ$l—ÄÙTȺŒ©Kâ…$l—ÄYd’¿$^M‚[@²©„ ë ô"²ˆ¿ŒdqÈÌR²’ß¿¸TÈJ˜ì_e,ý !¦û—ì¦í¼ëåçãWÏÎNvg§/þžü8•iöì.OJqËL½â>­8¤ŸXh‚ÜÞ¼¹|÷n—0Óü:°¡ßëeõ¶ÛßÛ:~Ê]æ ¹¥Þ~C¼\qÛÿÂ\ ¶Ô,2 Éo©«IØ&>•„ª ÛíËHp‡ŒCòH- ®1°© æA’BæÆ€ÓAO‹Ÿnk çùÆ~ RIÑý8ÃÔ½šÌé?íMúO2™@›ÌþfQʨ¤É@9“!Ää*EãßvO>¨¤§HfxôPñIŸÖz»ø´] vF'íòÈÙ!¦í¢õ Ý"ÝòénYY­7‹;žö Î$Oú€$rN áÒFQó« pŠ|ì+ŸÁ)–Öz«ø´W vD_yäÜ‚Óv™÷@ 6ùØX>ƒ],­õvñh»@íˆÞòÈÙ!¦íÒL»¤íŽXòiÝŠVvÕ.Ô–ÑRW¿eôè-#Ô'a‰mr)Ù22È8dfËXI‚¯ .²’<—TH€R1b².Ô(ý'>ÈgÛèÛèFkµ]@Ò.vi»À<2vÁˆi» R`–Ç%Úç0ËBj½U¶Ãi£xÕ›¸ r& °Ò/>$ï…4É]MRùÒJo楕„³ ÷ÂÊÿ§ÍöÌ endstream endobj 1973 0 obj << /Type /ObjStm /N 100 /First 1007 /Length 2213 /Filter /FlateDecode >> stream xÚµšO·Åïó)xL.²ŠU$Á€A¹$€`éDÐA‘c7V€óíSÅî×3vI†m+î4ûýXÅ¿³¹õRÈ=ÙL^Ȩx—ñˆC‘ñ¨Ñæ ÚÆ# µ«jèËëöJV¹¦ñTBn£Ù_ãñn ”F;r ”µ#wklõ§”ñP! @’¿KHÉÛI¨òxZµÑ*’@]º—4pª£^ œG«¨¦î êîÃhF ,äO9VvåZÆSÜÄ[o¸«+³„’Z^B(öÓK5íãJÉåD9ÌA{²‚zKÔ?ê²5¬hº=H®X¿hSrÐ<¼-”ª·®˜7ÃUó¢(Öe‹ÖX­ÃÑR­‡£¥{ôKéÁºÀ’B%O‹.9Ta•kaG8TqK U5C$TûÏ?³¼h#Z©¡%O…nÖ´Lã³±Ó,ÔÆâ4Í¡•ÑJÁlteåÐjòöi ­åj •ÐÜj+i°ŒteõäSW©%XϹJ•`~yK-=,ÚfçÖzýW{èMü {­÷‘UÞa)U—ö\HæŠM"qUí÷TFO4ËÔ$<êZ›“š V4Ráx?¤ÖƧ>"RòFun9‹Ñ¼‘æ‹×õüÍ\¼®›—‹ŽºÆÉRǧö#{–XÑh¹¥ñ©ÑFÞ[Ñh”FïÚÀΔ­Ç¬}’ÖÓ«W§óûÿë!œ¿||z>ß}ûûóøýO??þótþáéËO_‡dsBúx:ÿøðù9|L±Z\8zoù`ÉѵÕù>¼zÎïÂùOïŸÂùuø¶ß‡ï¾;ÙÿwòØþµù`ðJŠ¥öCx¥×胑‰b²ô“Tc¶· 6æccÓcÜáUžcαØTºÅ˜{Ì­½£LŠLq‡WË1Šå§ N)ÖšmÚ¤¨¶ÐpjQ ½È”{™Ú"Ù4»1k6qï3ïôµ4ŽÉ¦Ù)rÒ]æ½}¹åÏÊÜògY'嘴ߗ³Æ$õäsÛà‘Åݺ¾Ìës<¥Ú£­À°”Z7fVé“"\‰àmZMsò”lŒ4[m‘§¶c‹šä˜±OWæfêsRžnL·u‡7-o¤ù\z‰Q%Úæêe&OŠ̱¹ÞáÍÊÔR}¼/<‘È™õ”ÕÆûe}´¹=ªÖcG#˜ãoZŒ¤6ä/s8q‰‰èر¦Ç¸Ã›–7Y|ü-<²}AÕcæ8xšŠ…‹§Þ>v(é!îà&9j§ }ëì–l×Úʱ<[=T—™ÆNú±p?¤7^eÏÊ…×lÇÑùЌɶ;µ¬Ü2&WÛu”zȪ¸Å¦Ç¸Ã›æ©+/£>K’ò±¼’-/—QŸ‹­ÆrcÔçIîäÂã·_~ùø?Uzóôø<`oÄâ^꾩֨µhc´¥²þÒ¬Ìk™B奒)ß~yúüîÁ‚ 緯߄óû‡_ŸÃÇßzóöÓ?Nç?ðáñù«_`fßíøúôíËçÿliïéü營~þôÃÓ¯‹i~X¯Ý7o?}±·ý~n}y8þÕÀãÔÛ3nA×BFP` ‚‚¢PQh(@Y ,P( ”Êe²@Y ,PV(+”Ê e…²BY¡¬P®¨SQ§¢NÝê€^Ao 7Ðè ôååååååååååååååå¾(Ë˵Q …‚‚  (T PÎPÎPÎPÎPÎPÎPÎPÎPÎPÎP&(” Êe‚2A™ LP&(”Ê e†2C™¡ÌPf(3”Ê ååååååååååe²@Y ,P( ”Êe²@Y¡¬PV(+”Ê e…²BYåÿÿ,ï×(ͲFÙ’¹Üy¯ië¶lÞXÇ辕ecÖlgäzaÖn;õv ³RôéH5EŸÉãµIÄvê±Ò1ž-N0çóÞûò-NI‘-·8ýþãÆNaZœ`"ÎæÝqº¶Í[ZRôy´F¶yCýÞ…oœÑÓ$oÁ„·;Ì{O›·`ÂÛ=fžä-õX²\¼eÛJs=Ö[0áísš·`ÂÛ=æ,o³òèb-iä[_»´IÖ®H8»ƒœ6#¬H»ƒœfl²ã,]%m–Xo]OÌrLX»Ãœf-˜ðv‡y··+Sü {!&‰Y^îMÒ;‰ÌÑ7†bdßר¶Á“^¢ÈÞ¼Y8úß¡lH¡Xüƒdns’vc®I»Ç¼ÛÚ5i7æš´»ÖÎJ 0×Úcêßp“’í²jª1÷ËA[ü/§m™d-°vI“6ÓÖî0ÛÝ‹µm>ªÿUD±³ðgåȪ‡ŒN„ $ÂÜAR™¦\ϪšÉ„(WdG³H)±¤«ÉÇ&¾\éE  áê’fÙ &¬ÝaÖ:ÉZ[ÆÚÕÔSìÀ{ãu–³+Æ"áë²öIH²µújÇ.l'ìv#Ú¤Y SÔËöYÈ»ýF˜“¢\‰r‡xw@ú¤W»JIþÍÊ,ãKÒ Yz”ò¡}YZŠíjïcû¼›õ2©37$ÂÜARfM‘¸_¬-¦CSv#"ÊÛÄiAJ·£A¹i“­æHD¹ƒ¤IKI)=úÕ4Ö’"jGëC¢\—’coi–±@ÂØ '-%…[,rÕ—Eý\ý"“'mb ù7¸Wa²D}ù>˜u2×Hz&‰ŸÄŽe&ªWafÛ½üÍ?OJ îS½ 3ÙÞäsëÍÿeo‰ endstream endobj 2079 0 obj << /Length 183 /Filter /FlateDecode >> stream xÚ31Õ3Q0P0SÐ52V0¶T0³TH1ä*ä26Š(XA¥’s¹œ<¹ôÃŒ¸ô=€â\úž¾ %E¥©\úNÎ †\ú. ц ±\ž. ö Ìÿø™ u@¡? Œ?€Â ˜„ЃœÂ‚åÀ.9‹ñÄ)ÿ@0ˆ ;¤¤¦ÿ}ÝÌþ^ °c€;¬âP3Øs¹zrrâÐb¿ endstream endobj 2080 0 obj << /Length 179 /Filter /FlateDecode >> stream xÚ31Õ3Q0P0SÐ52V0¶T0³TH1ä*ä26Š(XA¥’s¹œ<¹ôÃŒ¸ô=€â\úž¾ %E¥©\úNÎ †\ú. ц ±\ž. þ```ÿÀÀÀâÔ±=Ë700ð€â@ôÄ@Ç0C1ã càk€8†ù?ƒØ ±:ˆZ˜>ôt3È)°ð…=ÄÉ 09Æ _p¹zrrÿÙd9 endstream endobj 2101 0 obj << /Length1 1738 /Length2 9937 /Length3 0 /Length 11040 /Filter /FlateDecode >> stream xÚ¶Tê6 Ò­Ò 0HwIw‡ÒÝ ÃÃÐ ’Ò%Ý-! HKw‡tJ·tˇžsîñÞÿ_ëûÖ¬5ó>{?»Þ½÷»††REEÜÌÖ,c sbá`eH*Jp°ØÙ¹XÙÙ9Ñih4 NPðßbt-°ƒ#Ä&ðAÒ tz”IyŠ¶0€‚3ÀÁàààà`gp²³óÿC´uH] fEV€‚- ìˆN#ikçî±°tz óÏ@bpðóó2ÿ6ˆÛ€  t²ÛÎ2@ÿïè°s³ƒ¿8þŸà·ÉÿßÜÿòòýÿMHÆ ý­¦ÿ­ÿÿ¨6¨ûß„ÇQvvz\ EÛÇå€ý/Uü×*KØBÍþW'ï|\q˜ô?—q”¸ÍT N Ë¿&èŸ<:‡B``[Gȯ§ÀÂÁÎþ?ºÇuY?>'Žú­?nÓ‡”†lÍ~­'7èàtGlü#âxr<î§Øí÷`ØXa¶N&€Çâ¼æ¶è¿úÉÃ`“þ%úx¹lòÿAü¼6à¿è‘ úâæ{D¶ÐÇ´ÿ‘üÊžÍìÈ`ÿËôlqÿ¡`³ürØ ÀGÿÐ?àcp›áã̳ý™û1OØãmý¡Œmûä°Ùý#;ü#;þë컀ÿ(ŒƒÀæô/áÑ·“¥øÏp 8ÿ 9³sû>¸ÿ†ÿÕ9³ƒÃãƒö{³Ûúþýz‚Án`úì”-H0Ъ*°éªBœÔ•ecXx‚fC;‰ÅsÖ¡Ùù%áSšÿ²Ã…xB_ûÓ…5iús±9Š{Ͻúj”à†8ÕÆ[¯;ãµ±Fô™Q‚=ñÏ]dh/X4Ä6½îí½´ü¬êá[h²ìù°Urp¯\;eÝ>w}œ šÚPÝüÄóãîã8K„f¸_á$M¶iú7¢—ÈN,d¨ŒÏ¸áLž_L<Ïy PˆaB÷ÞàÊ÷Ô[ጼþæ±XªÁéØJLM¬GD†pþ|pŒÖSbûƒá´gQAø²ðtohTï\`gì·×ZÄ/ߪ9³§ú¾iÁ¥nÍt¥:@µ@#JZ:ŠÏ¨žïhŠdÍãÊ©úŠŒÓ_@|®ï©ÈÔcuà͘šÛÈ[÷5[˜Øp®eÄGÕO1¿Ð>M˜ÔÓðm“Ê_ÊÈ0¿Y$¹å»Ÿ…§DsÈÄU–²2¡è®åD])®x¨7áw+˜¼¢¬¼ H{ã|¸÷´”Ó+( h ¿m0“¨ç¢#CÜ@X}gBØ~ø^9ŸüØð-¯Ðô,¨îÿõ¢Q±VÝáòŠ¶|èó;½ï6š½Ágý_ë\ˆ°Ï #%Öõøýnêj¢hË©OX^aaÞù…+ÏŒ½XwÖ¶uYãÂa¯Gç™°^^y@a„Éê­{„M¤wŽš¯øè‡/1G1E+xàybÜvô"ûm#ÆEµº˜®}Wxûëz·0ºœ®wƨÈ$‚C$Ó TÈŠäó‘™{©(±HÙg{Êõµèg9¹h‰G“Oµxß\iz5 Xâ§F­%|É ê˜ò··c„ÙH9UGÇK)äû\0YYêVljOVÈËfŠs¾ Ý¡UØò!ÁDÙ‹šŠ0¾"¾äÄ92U§©˜’xu¬’}]Öû, ºˆ SÕ¢éìñ1dIͶ §ïÆù»!òì¹°iå O•»y}˜œyGÅm^áÄ»m`+Ï9—töÎ{[Ù¨‹ À‹ø }¡ MVxˆ–>cË£ÉtúbÜ2q[tª`P…œÑÁ•œÓL΋Àaèôƒ-1oêJêE/Uø­Jy»rΫŸù\ÉiÎ|Oq7‹}vž±?ñHè\ûë~'¼Å Ý ñm°>”nÎëqÈ51¸œQsq¤Ø–Ýû&B‘Ä>Am.½qâÃr™à5¤Ñ߀ÿ¾-Žc—>vF,ðÕ–êêfjÜ… 1;d‚üB…,‚ ÿ˜"»`7E iièÀ²£^g”èœ óÒ˜¥qª×Å)-ö}ð’÷m¢tgÚ柵ÊÇñœC£bz/›…› ®‘F2æ{̘£3-ío±Õ*y¼ Bö Íµ%’OÒÍA/ÖL’“ö‘ - 8zQÛ¬U¥:’¶¡©b—°t$Œ@wÂA[®~ÜF€$™w™BcüÒÖsœHQâÛ.ï 燣ȱ¤D”eí÷?¬¿{Êò—#ÖÆŽ2ĆK,&‰´ËpxU¡bÕº·v†ªØÓI¡±9V`Õèäø+<µí7iÝ]ýø¦è¡†!€ÁóVúc£Ï^èÌø,ý©‘,_Àš[y¹ö3-’ë9å¥8Ó91…Ã7˜ª)ñÇOy¤< ÑdÖ·+*ûVi®íç@ÇæàúH¢$aU ÷pÕöó±¸Äù¡5š¼<­Ô_º¼€cÇVñ :bò «ÛñcÞ$¤vsO`’0Íñ¦ÁÔüŠ—¦zÓÏ ¤P^žƒˆLÆDñ–M¯üõO±å'‹L¨Õ¶òÝÙß©¾°)ÚQ+¯¥ÚðLð%b<2¯: µ-4Wz›¾.vtU@¨KCæGb™ž¯Í³¹ü@_ZÃfÅ…qy.v'þráÁq +øƒ¥ÍýnÝ•Ý+“Hî×H}/´Óª—Ã&Ôµ…ØAú€ÂX¡ÏìÜ)qÝ9Ä/Â/¡u\íDj×™£,mÏ[ÅftðâÞá½)„^²¢"Œ3fɉÅT'˜R?X“VyûžO¢>ävqãä9Û² 3óiJêŒÖ¸ ’_HÈÇðx£ðl8çIÈ(€Ñ" ì9,?‹¬Ëz.ª%7;¼.ê¼+‹z8H:O›8c{ÐÐŒŒÔ™¤<å£ÜË7xD`kªDàÈ)Ñ!QkSë#ÔKX¸&볯üäZCc³½‘ójä­+q[¢<)æÎŒÊF¤½ZåƒI<Ê=‰A¼ùˆ ý†žÖ/˜ØŠËˆ‰ ,aCMÇ…Rˆ*P#-ŸöÆíº§$¨1õú®xv› ÒZuPNÑݸÙàƒŒ(%êÏ*RVÄÖÞœS,ízg!š“û Ú-jíêÚâwt# L ¹¹}9e_–Ëex¹ùîA„±ÞñLY!ö±¼euc‰åÌ6ëù’ø3ºéÅ¢œÀýŸ."à¤+f×¥ž¤t+i§V*W¢ºšŽËá`—am8‹Tù€:õ]+Ýðºz˜‹²roÙDßß‹éÆÿ­Ô ͦ{ÃÐs­ìË‹÷)‘dª)ý$¢¥RÇŒ¤–{ïÙ࣯ªIÈ–XßÛ¡Yà@·Š¡Ô[t jñU¿¥9¾VdUÁó.‹ý3cÏ2ÇŸ¹&T.§Ê P¢–§–Ëœc_óiÛw“ÍYJKqœYa}ÈlŽÊ .9Öe3Ükô£îüÞçU8øõK?Ø‚hOª‹óSPXÔ3”íÝä¾o¿¼¶˜D<"ËIn™SdçÓœ¡Ðöj*Ë/¬a‰»Ó‘â‰Ð" ½'â-Ï•õ³|í¹ìîç ÉS A2åtë° ˜~¢±¯ô4AwqX*5à×ÿÁ»ÑéY=O’…6#ò}$Ù¨l¡GÍ—8 àœíýŒjÊDÖžÛNòxv}WÙc”Õ´’l#’ÈVJ&¿-×Q.‚ø—Y¦hzì¶oW·¾È•ÇÚfÌFO×Cå[bd3æ¯+ˆ©¬è ò4U:ðó.{ÜŒ.9$-a¬T—<š®v{»Ò&Ò/ŒÝ¿êjÃ@—VåÝ…ò˜ù}Lxx§WFqÉs ÑQáü ¦ø-!a¾^ßÏîçû¬Nü­ÜüGÑ÷Mgj±là~PÊ–\aî-ÊŠ¸&ú§’ÌÉƺç*ÆäW»-ó IeAh9`#G ë¢×BM¹O¾t_ö~§§§JæŠøú P,›‡¿ÿ¡ý]+±°ü3Ñ-Én°åT艬 #¹ŸK£µi,"#Vùv7Ú¨ˆŽ+_OC!66ØŸâ‰!¥’yÄO¦’< ¨ …KR:eÉrîžWÖ½´ÚNBw$êlD.ÂYR<»()Wº=Y'Ñõ´RÛ×û‘‡†\±4UTïøŸA Õ>š‘Ui‹Ì£iré½ý©¢³“"*¸’M³Ï¥Ký3¯…Â'³LÔú¢‰”¦=Ô/bcérw¶ò¦Iv5/ BGˆÊ´¨zZð ƒ#€_~¥«„‹¥Èðd™M!™£) úÍߤFÐÄhŽ¤¨¡^|X­óÛ”Õø´Œi°Ù,?iw¤n–^ñàÞ ½ÂBª°Ã÷Ý2 –Æïì-ôÚ¡ÞÙ*;î³{W0L}(¹õ!þÐPÄå<>¶¸Xæn1àú6¯nhKèy9fCûD}.R:¡œE–Êe;’Ú\ìƒóÚ~ÌøìK÷•#BÉÆjõÉâØzM‹AÐwÎj*Ä@ôÔÏJŽ I¹DVðµ,I}N"¬‰"{M{Ôz~*ˆ·_ÇÏ&^&|ñvÀ¢a܆f©Æô¸k™Í÷á°Ðïi¹H÷ £QšArÌÁ¸ëÞá1\d¨9Å€ÒruZ,5Î"‚j‚níÓV™e¶)¬Ÿb ÷ HAŽ”³ïP~¨‰Ä9¦øoQ³O`!½Kôž ~É"©Ä2çÓ2KB¾z©¯Ž Í¿(ÉmL…^¿aðÚò,@ßM;¾¼àÝïlÃWTéKŠ Ý¢hæ ',àÏRóE­)¬sEOÂÀû4ÁÓÖ¿Úÿópðã‹8 ™¦ÐÇB†‚z“k—ÇÕ5tôÖKü­Ê³^Û)ëÂdÑuƒ°QäúÌ»iÊ"ËìÎEöCüC5²oÕŒÎiê  ƒi>黬0ÊÜ#_!¬É nx‡”ÕfÔ7¹úóµø/ù°ãé湃À+J‘Ú…^%Hê컃Hƒf‡´'Å43µÑ ófiÀ'R­‚1ý’Ô”ìƒñÁo‰-D%€drÛãÕ—\s?¸,£UqCÊîHBàÇu¥vôët\HêwRÔu륗꾆A'ÙøgµÂ‡3CÝð$íBt´BÞ€žßNl¿‘¸~ÉrHÉ^r5….øRáü¢ ùxI7D²Þ¶µ>WòN‹/ÿJ¾÷5u¸àdp¨±ÉKˉÍóÃ÷÷p_?}&ÕÆv–Ñ úÈíôäì}1–›âá` ™lÓ§)åY–éÂU û^øà ÁƒÔÔÊܾklÚ¢^æ­ÍÊ`Œª‚ÑÔSžAÆÎφNˆ˜IÊÏ*&wÞ‡%,\[y¶zQ\ªbÔ¢åÜ\<ºÔ·o[·ß?ÃÇÊHÎÞ UÞï+é”#º„£¥yµÝP@DC”ÕX Ó2@®vRĹ»„Ez3[Öç¢Á’ت/D›÷‰ÑÔbXïDÌŒ-qdU\˜ÚÓ–Ö'4×Næ Q²Š:Fb–ÃÅíê¡[¶’¡ÁŸCåн>§5›¥0j‰òÅçJkDѦÓéOŠ:»>ß’ÙÚM|ªyG‡ŸD4Ï*#C¿fz˜C’¦¦R–¤3wLÒƯÈw>FMˆGv‰gÿÕ2T%"ú¾Ü\È„·m.2fµÞ虘 }ã!’‡»)i„QjLpP ¬×©/(Ò0ì'±Z#¦'œ»=Á»ÝÝèÔz[Ž$xÛk!#Я%„a÷“ës¸±–Ýëè™,¿óò¡Ÿ˜¾˜#ߣƒJIîE2x»iµˆØ¢Š˜Rmü¢šâ\é^Ôœ]`3fK0)¨åj¼ú¦-ƇCÿÄæŽ3³îd =xgªý¼V+½wã=h0íyXÎü!)+§Qt!§ƒ¯rò^¶Âvý=©™ð@ùÏ-§B­7Ã}­gÄÖ­+DH»t=Q&;c[´Xüø2…ŸO1 rZÂú²RâföbßV1s‰g•iÚ››Žiìö]ìM™‡Us«¢ ¬ŽPu00{jK}®s¿{‡åÖT»JІ¯ëwUYûÒ¸:a%ÖSb‰îتpY—×Ô¢ Œƒ7Èb >©â5"bʶ· txq°¯ûý¼±P‚¨Écú}pi_¿5·§é±ðÙ çdÈ°P-Ç?°:hÅƉˆ<èîáÎð%LhÖòeߘ^àð²Ö§ðwëÇú‡f£dŠE§Àç“ÀßÐØ0k_æ„=øO®­¿Ä¡Câ)²ÌÉ7Ðò¸Tõ4nË7 ÉNù¼ÊGå/D~LGmö£«_¶œ^‘Ê5ï7EýPF|˜`å+Ÿ‹Éwó4žº rÕáMÓ·iëN˜'Xçα¶vJtû£]3Yõºò@1Õ|ô"æëËIuà´%T1¦–±#Ž>'ЯŽ×û=àûBcðÏnÀ1Æ*¶Õ%ÓA°{„žÎ§Æ% kíç™ÄidƒúÔÔÐÛª¹Óâ…ù¼±ÿ;ÿµs5yÙšgþ’¤MOhå¦Çkí¶«N x“…Ötúdµï:œrèL•û¤#}³8`7íú^õÎÄF½UrÒ,éhíŸoÙ Å­¿ðy1aW˜R•¡¦Ö·P~AI¥x¹|ú›:,‘k$qœ/Þ¢ƒäËOðN ™Åç»N÷ZYºð@H¨J<ÎÛ˜óšq„š½Q®^)á,b¼좓Íþü‹U,-éyÖ#oSÏš¹wsdãpi-×ã‹ø™×”¸&5„G9 ‘Ñqžp\>2Á–`¯âˆîÒþ§‚héí§7µmY/v9Oì­GÆ1#Ç<„]ÃxÚNÂCjñ xgJ³‡¢å¿üZ k"GŸYÂ@Ùn5-^¾ò!òTfUhí  VI%¯Ì¡¢-»]”Ø+ÎØÙpnHÊÜ%z?^ôeä#Š f½i5x3GÍ!õ§&<ǵ‡n9º¯¶G]'¨®…¹ƒCç9–Õ¨_é@¯ Üè+ô󆬧Պôš ò£hš÷Ý.¶l,éç:b …;2Ú¡»†-Êd­‘R?œÈZäºãIpbVÁ˜ÆÅëq‘ááÂÄ‚2j­ZjÌ5J¿zªHbóÁÅ<™=§1B¾z[¾S.iá-˜×Ó*±M5~Z«6í|ÂFeùÅ^n8÷¬±-ÕI«HõäÇ«ín(ïl0ÚÄŒ´>eÀ!uçv+q.²½Žûd#±Ò€Í ³È–PÉ$%šl}¶WÈ¥3T«F¹…‘Ó®4ÝQ´ÈcnŨÜ5aX’•km6~ö4Y¢›HJê±q‘hÁÉúE-¬ i øÔw:~KR… Y%#…ŸêÓ†å}´·™½·ªC| ?Bè”~¢éoR(ES4ê4ÓûHH¸4ÐTªæ£á9?‹*K%2©îBÛÎ+áA¬DÑØw³åÉ¿¡‹§Ìà`{ZÓé°VYKH•ô_ÖƒÝØ”(®6y= [ÐT£µV3§‰0ÌhÂ<µwL%¡Ïøp´µ…ËDŸiÅWnà’1½düͻ#zý†ÌOÇwh¤Ti¿Ø’¯%4“¡ „áx×ó~ŒevÍåõQuÓuòs-à´OÀ§¬®“Ïß^µ˜äø²éC¿®wbƒæëJQ¥ˆdÀÚ³B…Yû»Y¿H™ª×wz¯«^£H‡!‘r+ Æ“¥cO¸)[<¥^\`*nè-WoÓª~$BuÉ,&Ñi cêÚ»‘‡0N© Ñ…›Î4eò6´ë+«w@8êI«-l]ÿªhþ¢OÃŽÆ\ÊÌÜX†9$žôˆö´&m/„x7µ×²N}&‰XÊéfíÛØ(¤=¿UÓâƒYe|z¸HbxŒO[B÷ÒÛføpãÇéõTL¡3òFô’7mƦđNÎŽ§–Ät—˜³äQ9aÕk• «ŠÙõz™ -5*Ä0/¼c#çäPžrGM0S®#´”ñ´<•¢Šú8+¨T]pEZﾂåÄѨ† îÎ’§1sß›ÚpÓR"/Ï&"+\ê"ðf,ãV”«}oX ‹Î“šÃ}ý¥xwt|_¦3äRÔA”ˆ†¶™û~ ÑÚÁÆÅ>HL¨s Ëêú¸wš9!õzGì“?’~Þ;!a:—œ+6K…øs†[€ºl“óí0æò³MÆôH †yû‚³ë³û½ ³†Wú C-&çáY÷VáìÍNbL^åI0á •Ádü·÷þø?ÅUí,›ð=l4K-…'™GŸ¬Xq®¼d|Ž1&²³„; ~^|ÞÅe0饾Y{NúEÛUÒSÍ”¸ÑçªD{fVœ³­åÛÇ9(;“tZTÄÄôim‘––™œ k«bHsÄ%vZbcƒ‘MWam=¸X} »p62;G„ÌW¯Øø‡â®XÛ3Òéë—f|° v/È«œ‹´„õAZHÂÂÙÍ z¬-—³·1hûå2çñÑRqtìg¥¥ü¯’8ÍÉ ÚæNd¥ÕX%ûć+ȾûÝfˆÁº*½Õ’3ôó9Lk;:˜ÏèÜÊÈ@¯Rÿ15»„kÊŒnÖ?v„ôý™Éùªø'¤3 7rüÐô1‹#ûß}¾Ån›–”žÃtì+ÈióœmáüVòdrd[Ÿ~mËãIV)ÙÜ1>wr‘2_;Z몦´Cî´ëPã‹0f’úµHÞ5FÊWø·y@ýI”AÅL¯‚Êþ¶xñ†'b5ñ}«–¬[_>i¿wÝ>©ðÇÄÍeaYy¯òÆŽ˜Ø‰û‰ã!«"7+mÝ531gnÖ5ÚöÄ,tnwWýí+QX†&nPéÖpaÛT㮿ŠeÑ¢K‹,ØŽ²v¿±jw6˜ØkKìs¬‰v“¢CœULV¨î^ä) ¥T/7×É+' 厥ù´’Ï1÷³æóJýíx9Î ºìˆÜE}0a Y»ì¬’lŸ(¯ôQIuœq¦r*†SC«.>gBMÍŒ”­…™±&É$SÃÐÛ+Êö§dâûœ-e¾j&Hô/Ò&ëß&Ë0²‘-©ÊõmŸmº©Â‰&r [Xµ¯Ð/3Ðq"–€?£‘ÈÌSàßƉK0,í~Ç(wþ†ØoŸÃiÅç…¤ðµOˆ½u27 /fÛÞ×ä$æ”Ðùs‹ÖNøs¦NçU©ãEâŒFÖ%F†è´3fÊ’àÔié»´§ï4§š¥ë‰þ±w%3úƒW‘3£+@‘—ä¡®ËH ‰ôÉ7ÃÒó¾amzI·ïÒðÔ3y‹]ŽÆÌÙ;,ö“ºxƒò_Ö≜Oy»cÖŠá^±K›T¹b×F6õšŒ%ÆúÖ¾]Ž^Fý¬ËšØÓ œ3¤¼Š°0ÅrE?áÕ¦/G±¼ Êá{«ÖÎÖíbß=ØÈì&¾Š@¬©žžÿzSGI‘EÒÀx}ûÛôOûÄë©L©A%ã=toàíºÖèb/æfá¾%ü¼åJÀT#i­„Hºß%Ü ë·¸´dnÅøx5vÂgNÁ¸ZÁÁû1bÏSÁ5áfJ#ßtÔ)ƒ~hI)nÆ5¼"=D˜1º´ éÂxI+ˆ¼tq‚ïè¥Ý¾>“K„OûþwKºÀ#‡ÙÜéM•'–‘ìµÈ‰ÕzBŠ%Ò%eFëâ5ðÒósãš~‘8ÜÆÊÌz#"?@ñ ´õú¸`¹QåÓ9ǃI²hÐP5‘/!ÔžÇK0LÃrr±+_ûÅI[ΗØø$¥0¦„—ècJäåvÃ(¤£4òµœûÕMvÂÑ] ÎÌ“˜îooaÉ3Ñùªï)äÞò§dY‘CƒrãWjB-¡ì‰¢Mþ™UŸõÿ<(HX`4jOvrÃö'?ô’ì×XšSWóТQ”ôß…ì—xñ2ÏÐÉA.Ì4Û o¸eÂö¹;Ÿé—ÉÒn&P†È(SlGUÑŸR½ð©hq3WÎXä#D¦Ðô¢SaõîëŠVî±8š[é4Ô@â < ѹ«ËHЃ¥ò¡DàNùã•|j}#ó*#¡å@ë£så͈?VÎ vÍÅkÙ×>ïIJf¼xùWžF¯¸HËw'P“ÖtÝì8Ù0ÎT—&gŸ=Ùß@•QñúÜtO¢ f.ˆ ¸M£] JSÎâ[LuJYÉ®wù3Ä×Êé›*6¯mÂkõ¡·‡mFõo8[ÈUôm¶»N½8qØ—R=´d|‰ŸQ(\Õ_¤s“û“ŸqoéÉõ%Ôòuh=CËFttµ’Ò’§"¬r‘-ÙTÆÿ˼²Ô“„·­½ÛàyòYÿŽzË¥†£ˆÜãË^*ÁÜîÚøM@;oHÏ ªV.Ó‹q°þ’õùgoΚò§Æ騣Ý^¡}p[H›úÉrêSQȺÁP`WâÀÒâØ"“g‰´~Sq»ÈÓ=CÈ›h’ö|ì¡ãŒ²„¸ù|QåãGh0[½bŸƒ…7¹}bQÑ•É"øW4òt ."6¦V}l¥Àå¦ÿ£äjF!P{ €} Ã=î9N‰G¡Û®gªrX<ÒûɈ€[I‰œLmáÓïáb˜i»é?YöèÁ¤É ‘›Õ£ˆâ¥{}‰nQHp_èŒ&£ühµ?b–§êïرœà¬éd‚8ÚñÖQ2ú ‰„7¥Ï  ¶w@bÜ;ÁÙüÏ/½ÈgŸÁY\ç8´ëR'i4ŠÒQ°ÌÜe¶ºù«¤Òž²v)¯1ø ø.ŠÉåÊé89 <¥ëËÈ>ý•ÍA'øÈ‘D 0²n×öùè£õ¶yY5…ÆØÖÀÝmÐÌTiÊN»ÞzNZ{§òøáQø“ôúT éê aëc‘F]”"ƒçˆÛ+ó¹.a1¶_š¡.AOÅbÕ†ZÓD¾Á~Š}EÌ+)}FoXÄqÑš®‰œ„m/ØG2¦ÂqÖÓï#$õn饫aÝ-þdoÓI£QYÛ :&ºjþJ.l—v}Av‡²7k-ÐH•Í‚®»êk„eó;Íá¶&â›ï­™#ø_íï9ƒº•X—ßàò w2@´Ä¼údP²Ã_ËÛ#Q'†®ÁÎÚÍ÷Í·¨pD\¸ÇD5ô5¶¨e‘a puo› `ŠЇCNÑéo±¯òñ&<€V® íÚS†Tå+JPŽÐ´œÑìZ>y³Ý6:¬O`ITË°èÆ*-8y/ 8x2<±Þ+'¦vnO‚3Œ:‚.…&!l ;Ñ^—çŒPp?ìè:ÀËÀ»?ùlšÓÈËë³¼ÄËOƒi÷ìÇç4’Ø·´aeÔw![èNZ+L^—ãÌìö1±p5ù±úîG;UÆ6¢|uèØ”f»Õ´$gêÓ§)urÚæ–û8`ê­Ê>u&¸K€ÅW0cÁªq/mãÝiŽKŠ«YWDÒíËÓ(ÝôÚдÈP2Eï 1îx-Ò@u9y$MÜ®FFãUº¶Íâ’8Â/Ôq…> stream xÚí[ÛŽ·}߯à£@ݼo€`@²¬D€–Øô°Y¤E¤]awlØŸ:UìŽ4Ó–½=È‹€]‡]ÍúÙ=ÞZg¬ñÖzã\B'<:dt¢¡Dè$‹g“ªSœWS=ˆ5ÎqÏaF3…óô.*npd\´rG4.yÌÏø.“Ü‘+Qî(ÆÕ,wTnJæÆ3‰«•1¼ãÅ2÷˜„¼\eb1&M1¢ÇMÎBÇ“–ÐcâZ…®žDèª ^f Ö„@1‚c9È,,–e–LH2K`)å >B4¡ZÌ’!´!rAæ+†|¼4 ÷2«D|3+f&o([ÜAÁP¦‰ Õ¹P4$Üc]¸©Q6Ñg¹ZL R£jb´ ‹ÖÄä èLÌ`ßò2b Xs &VO•lš™Õä2VÀ H,,ôXád‹I °±š”ÄRXÉ)‹¬’3©ˆ¬’‡‰Ye†0 Qv÷2PVè’É$†—²ÉQl(““¬Ÿs”;*÷J>sžEœ ¯Ê¹ÊóU 3cËÌÕa}c¢ÁMñ›^‰³0[%Š­ñ‚ ³i*k§¤°âL©61Ås¶eYÇ¥Š ™*eKäž0β®A”ÀvY“˜)×JX@©ì )€KfÚ9'¾S±Þ"ÃTø…“©ìž€.Á[‚tÙ5B(v7öœ¬ÊÎAQ¸eÃ㮨ƒ-Ù±Eð¢e´˜!|·fÍå³û÷ÏÆg¿½ß˜ñÁÕÕõöl|úó¶òýÛË«ÿž¯o^nnÌsËÑÀ¾8Ø\lÍs_òPØiC‰A‡5â$5 ,[¦{`îß7ãS3þõúÙµ™/BúÒ|õÕÿÝ3Ó ž:a?¤H§ÅL4ˆÅO˜Ùlè‡1ËJ˜1 ©v|&7p¼=ŒéV¤0ÀxfÌh‡’ŽðV ~ˆ¶ã“ìà3Ó»¡ÚŽO_‡”OÄ'Û "§’!0¶g{uI‚«Ç‘ƒ˜Î®Äç„9ñ¹€éëJ|Z;dŽ 3Ÿ¶ TÊiùœ0'>0×âÓW;8”O`—zZ}ΘÏ%ÌÕøÌuˆ~§O¬Å×ÓêsÆœø\À\ÏT†:}æ4d{X¶q->'̉ÏÌ”ï†Éuß¹PLÌ_àj#Ú<8.ôR©Cu‡1ï éüKɦ»çÓJxÙ ‰U*xlJòa<º±Ú*¤ ©™ë÷Y¤Pkr‡]Ä­Äã„ ðÂM• J®ß¹¨\³Se]rAš"BlÖŸ´TÄ!47¼†ì—ªi-ö´ªec¤0øèN‚G\’zB€‰ŒÁe#›*EÆ ¼ŽO"Ó©¤š1[Iµ„êJ®Ñ0g×XÀ,k¹Æ„ ].á­TÂEÇQ”š+zÏ;òzRß Þ™Ç¶#ÞÆywØõ­„—<Û±ùf:ìú)®c2Ä;,OnÔdˆw>ÎÃ1¬Äâ„ —ðâJxä4 ‡º#g¬w…›ÿÿÉ–`ÈÈìj&W®ÍL"ι¦^Íà"CFÉ$iµŸ#ï= Þ”ÉBÛè¤u–/ú½¡$“{Þ×¢ô C=ÐïG01ñ,¥õò€pGÝU¥–1^µïe^ÖŽR¯“EzÜ Ì@î®,„T8îiù[ J™ÿ Ó &™â9‹gã¿._nßÜê[7ì®gã7W×//¯^yýF†@ „[}cð|‚¢¬‘žùiQ—¯^mn6WXæsϹa\Ž’/7¯° Æ9` ǵªÐ¸T;£Ši(+”Þ£5`¨?ö}Yä9Jd´âY…÷Q ž!’pFaw*\=²¯QåqÞJÂEøÞ),UÞ&TöPÇ3d<öîuð\ÌŸäŒg´}nôù^ž‡cÞ¨«ˆI‚ Æ‰¾uV&hä ¼¨+㉈¯ê±H”C< *ÒZÁ‚»‡èx&ò2"·¼¹á‘j!«ä-bO%„⟌ˆU!$ Íg¬x ÈE^Ô´—© z~‰|Kà4üÏ‹ŽC˜¿&¼%ƒð ragêØ•~?¢±JÛ~Diúñþ^÷ÉK<«ClKSʽ™ÔG®â^l4ñr#Fp•ERD½¨èñ¶ú8$=Ôÿ”–Ãd™oû£7ÒÔíãø•{ÿÊÞ­ûc™ËáÐÚ…›°¡ùž™j묋 Ö¾DH²Ú*MG@•ÚÚŽ Å®î”c€û±)²›hX¸ä} s‚ŒìZ9àVpK¦ÌN* ´.pîÅ š‘§ƒ b„•c¤šÊ‰©VÍNÛª džኅ÷bnßm'ÁèƒØ÷Gm§ÛÕV­¯¿Ú摾^å=Òõs¼n ‡gÞì]¤,6žJŠ8È ;pq©ðB…©x¡Wî¥P¥Èp;uüPì×w—¶Ÿó˜Ç%‡0V"BiIA­Ô€VtŸ=$샚¢£•ÐXͱJ` \•Ä"†–$jåÀ@ˆ¦$‰ãÖÁÌðÚ(Þ†eà9ßœIæO­ÖŠr™öô§í¾äH'QX/É É V‰Päå!‚“V_)rJ£±5f1Ž÷9bd4‹¼“ä¸"¾î¥b ˆÙP42O«T³ku«¶]e«yòH«™³UÉ:¢¥x—ckKX¨äeøãR»•Ì]¿§ìÓ_ŸÂÛULû…ûÚ~m}é@ê‰êqºûТ™›™.±¨±F„¡û¼»iýFÛYŒ–2ŸKÞÿS™-Gú#ÿ^ Hmƒ­ˆ-{•˜Žä(Wª¦ìe<%ÙyƹèÍ™‘E¥Ô`4:’YG¿á×iîå(¥sU’²S}$Jà J(KP"Yˆç}jcÒÞ ‘,G'%¡j}Y¸²+‚Ñ^Z¬êˆn‘•fG[u.÷È‚Cµ;:‡×‚5Oå/+ƒ¸HA 6&Dç»ðò …¤® óµ»´mήe$˜ƒ®PÚþ`ÿ¦R¸ ™åhídr¾W9ŽôTÛp¯ ™BE9÷zª IL[™¾¿Ö-pÉš—¬[ÛãÖm§ª­Ùv*Jk¢œ³äjÐ!»9=âîľVÕÒã´r½·Ív†Øæ›·Ó.oúÚ¬zfôÃ^'‚­ZúkXu²¹Yµ*äw¬:9™-É•"óÌQtÙ{âg*¼Ã-¶(¡¬¼vÀâtt]) ”Q9køñ²2EˆN t­ß“”‹­¯þŽ–¬dÒêŒýÕUZ:ÛÒJ?¥msv­bõ4:’IÜÁŠI¹å¸È(r:!æ—QÝ)~D•$Àg©`|ñC³n& ÍrCÝšj¨!(¥´µJ%ܨF{¹öRéyÒèLô6kòžÇú{0ªšS 5Ëéä"óy7k¬YO˜îN²#qSyâ[¢T' ’†Hbsꬒ¶œl+åEÎyb•m¨TY¸?$È=E”áÎâ§pûÍGù'Ó.‚îz¨âk‹ Øàf‡àS´ ¥Ð8I¸~òšP%®“ž2A‡‚pî„ç%ó(OR¨KxŠµé~K²¡ö²9« jÁËqznR¼ÕÚî¤BQ³F|m‡.ˆV$-š*X)eèëzºÅzW:üT¦H¾ÂÛœØø“ßhƒœÍIeýµ·.¦€œE—¦åè ç„{¡ü;2ÏÝ•åVTùé·}¤Òð>°ÂÉ4ƒh{ÔçØ(5FH[D‹ÚâÀ¦¯”¢Ëø]Þo©N#Þ•óÐ"GØéCH=O¦°áeÑ·ži&'éïÖÇ~|öŸžüåëï:ËãoÏ_ßêŸßc¡Üóâ£ÎÁ;ñÉíÅæj‹C³ñëó÷Û\¾~³ÅùöÙ\»çpñÉöüíåŃ«×o7†§ºÝ¼û'Î{ÏÆÛMì4ú,Oš¿¿ŸŒçãÅxqýöúj|9nÆW—¿lÆ7ãåøv|7^W—W›ñz|?ÞŒ·ãíæ—ÍÕ¸·on6›ñçñ×ñ·/uí/Ñ;ë¦G¼ÿP2 endstream endobj 2103 0 obj << /Length1 2429 /Length2 17938 /Length3 0 /Length 19337 /Filter /FlateDecode >> stream xÚŒ÷PÚÒ€‹âîî înàîÜa€!¸»»;àîîîîÜÁ=Øeï#Ùç¯êÞ¢ ækï^½ÖTdʪL"¦vÆ I;[g&6fV>€˜‚¨;€••ƒ™••ŠJÍÂÙô9•ÈÑÉÂΖïbŽ  ó»Lèün¨`g u±°qظùØxøXY쬬¼ÿ1´säˆ]-L ÌY;[•˜½‡£Ågsç÷<ÿù 5¡°ñòò0þí±9Z˜m @gsÍ{F 5@ÕÎÄäìñ?!hùÍíùXXÜÜܘ6NÌvŽŸénÎæÈÑd ø«e€"ÐôïÖ˜¨jæNÿR¨Ú™9»A€wµ… ÈÖéÝÅÅÖäxÏP•‘(Ùƒlÿe,ÿ/FÀ¿‡`cfûo¸{ÿÈÂöog ‰‰=ÐÖÃÂö3ÀÌÂP’”gvvwfmMÿ2Z;Ù½û]Ö@ãwƒ¿K$E>€ïþ»?'G {g'f' ë¿zdù+Ìû˜%lMÅìll@¶ÎNÕ'ná2yŸ»Ë¿×ÊÖÎÍÖë?dfakjöW¦.ö,ê¶. ñÛ¼‹þÈ>ƒœ\¬¬¬<ܼänbÎòW5{ÐßJ¶¿Äï=øxÙÛÙÌÞÛùX˜Þÿ x9]AgG×?ÿKllS g€1è³…-ŸèïbÙ¿øýü-ܺ¬ïëÇ`ýëç¿Ÿôß7ÌÔÎÖÚãùßGÌ"/¥,!£Ãðï–ÿ«µsx1qq˜Ø¹Øll¼.V€ÏÿÆùïþÓýßRe Å¿«cýQÆÖÌÀû¯&Þ§÷ŸF\ÿ½´ÿ¾6t€ÿÍ h÷¾Ï íŸõ×cåb5yÿÅöÿùüíòÿo÷ÿŠòÿºþÿ·"Ikë¿õ´ÿ2øÿÑm,¬=þmñ¾Ï.ÎïwCÁîý†Øþ_SMп.´¨µéÿÕÉ8ßoˆˆígëÿŽÑÂIÒÂdªlálbþ¯%úÏ)¼·¶°)Û9Yüõà˜ØXYÿîýΙX½?*Nïgõ· ô~¥þ7¥„­‰é_w‹ttz °¾/;À‹íý’š‚ÜÿÞm ³­ó» à½9€™#Â_'ÊÍ`ùKô/â°ˆþ!‹Øú`ÿC¼‰ÿ+€Eò±X¤þ;€EúqXdþ'€Eö½×"÷‡Þk‘ÿCïµ(ü¡÷ZÿÐ{-Jÿ¥ïµ(ÿ¡÷ì*è=»êzÏ®ö‡Þ³«ÿ¡÷ìè=»æ‰÷€è½ã?ô^‹Éé¯e1ý¾Oô_|?0–-Ó ØßKÙ˜ÌÿáôÞ‚Ù|aöüKiñ'$Ç_èú'Û_ë?ú¿Ìí\ÿýÝàó?ð½€?¹9ßgbîaoþþÝðÇâ]fñ|oÂòø>«àûp¬ÿïÓ±ùG+ïÓù™ëÝÕö}éÿ¡ïÕîO1ïÎvÿ£~oÆþú=˜ýû—œ­5ÈìÏD9Ùþ-uüŸAs¾WmÿþúØýãpþ•ÃŸÿ‹\@N_Í?6ïãùÇðØÞgáô§¿äúaq½›;½?ïÿMûÞ¢“õÿœ.Û{%êz YœÍAÿX’÷¶œÝìþáðÃåø>a×àûÜþ±ïÞîÿÀ÷ðÿÀ÷zþ)î=’'Èñ_©þç‘1qq|Ÿ óß_ï/Ðøïo{Èd‚°²hgò1ز>¸ó¡V„ÐiR`Žj_3•ŽÉkűËå7 l2]Mfà¦ãHòh?ÚÚ®í­ð*é‹×I[#lX{Ò§Ž'ïgÕ™ý„åiœá©Â‘†!bx"&5áïo+È6ðYª\—(Êù˜nƒRî Cå?ÆC÷?ÔpË!>—Ï2ŨGë”ÌSåg-à‘Ã83ÃÑc\¸£ÎßÞÍaäL½‘Ê&0 øüŠá(òÒÙb}\ð\¯TcwêŧÄ×Á#†¼ÅŸ¡ö=J‘Å]ò*-Ž“ *6#6XlìB:pÖ‘õ¬ð—Mmáý4 +GÝ 3ù´C%:úÓZ5§#·§dö…Í°ÁéS÷á ŽgTÕ|¦cа•ó©2{XÚ@¢5Ið-­L÷QØQÊ÷2~þð¯ï,Mc⛟pÂ2»qž³nŸÐù´‚çáÒgjJÊ.[÷FMûTPmŒh”øÍœ,ç#ŒÑi¶æÈp5Æ›• 8aYÌ „œåxݦl'ÊÛ€bö‰IЪÿÁ£0ÄÖ’ƒÚ¾Kð²i¹.h&6Ý’O¶ÏjLî)’‚]5¿V÷ËÉP–¨1†Ðx¥x ]p=OM)X˜‘OCÐöù¦ä^™˜o²q¦l6ñ§i¨i¢÷ŒLùôYˆ´[¸juňB\üœ-«‚C3© 7‚y|úºG–‡{©zž˜?U _aôy“`°#-~¥bà7ìì¤ Š Á˾;óüÑ&†TõW·š·Ë]„çFÊø½é½mMEf±"ZY½sŃ¢Ü®þVtèy¡+kC4§¾2…På–ûrÄîX3UÒíÅã-ô‰#^bûØá;å‘øZEÇE:yó²Ø…ÀÙñÉ•¼ö>åù(ƒ¼–VRèÞÞy„=¦ÒIOuú‰$•‡¾w»ÒEn¸AžûŸø Fa;HQ·'˜¿Ì`›¸·—ã¸â}-' MÄà×áä^Nç’û1ŸÞ>ÇàùŒ.Õ•Ç N×ëç¨]™É·Tƒ5j’Ž#Š\–Ç#£±ð½¸Ÿ«ùÖ–¡¢Ü”É;­£“Alãm‚kV0(£gªêk›öåxÿé.ŸUÿÅL)Ãjy*¹3@iðÍnu}â46£ªò#x0íwšÕÒ€¯}fuAÈÏ,E&š_ƒÄS¿½à—~úþ[e˜Vµ4¸}3jpe¶'ý™Òþ3‰lšœxíš·!Z À£såS–¹¯Ïpâ ÁÓ€ÕMûOcÓÄÔM¿¹Û»ÅYŒE¹òEC\‡©&¥¯:m^sU.ž©é,¿µ9ÏWÍå¾èbQé¤ ;¡’ì1€È˜AÍÃYtýsm¶g–Q a¬\C×ä€cˆ£v ÿÛ«!Dó_†Ð§0Ö؇"¶pñ›Àôý¦À–;ÅÞoÈó=|y%ºi”9¦|Ó&Ø‹i–/bO ì{Fã\ŽÕòq×|x{ …#¿·ªº'|·^Mr1–ŒàK·"0ÉÌx •k|à$ Ë—R1“›Sa÷á8“:‰?x‚ ÔádñN¡ñ­Ðñ‘ø¥}« *FÀÕk|Ñ16|é xqS¤r˜° ƒ¦æÃ`¾²?tÆÛ¤[–€ß€ÏdŒXߺÖ|bòP>M2\Àòrhxº“?ÅX/žHU75ï ó€‘}$…÷p`Ø>¨¸Dž•mLö9†?p½·dqœ^XÍ Ò!VF£ã¬5[dP²×2 wöp¥èÆ©èPò›çŠm¸¥v‘^>q*‹`~q37îñ“~ä°ð lÒE-²ùåyŸ¬‚ý@I<÷ ©+‘Ëé…2hï©Â¢ùø… ¾# h(}ДdcžT2»¢élõ?ÔuÈh)cÎZcu!RÏ¿'þÕE]¾ÿƒQdÞ‰œ³r£Hp¶­0U€¢$,G¯ Ñ)Qã e8Ê$Ûõå¢1ñdAXÁµŽiÜE™yXJ|}TÜ  »âiYþ†ýnœdqV›”¼ÿŠR‘¹àuÎœ]ÌÍÜIniA/تìÇ Y=„IR"Yü'n¼ïýVø˜±‹6¼ä0ŠH#ñÙ>Ý® À‘9·›—ª ¿ÀFµ«OÆ8ª2˜XÑ&Ü,}¤·“ÈÒ×_frà 7UldU1%Iª¶…C»¾1~†èô¡úÄÙMzdز™ßÆRmdr ³®ÜpzÕGpŸ» ) n7` 8î6¥M}¨+?š§Î'Bù…CVÅc‡ô;²ô<£.òJ“Æ~¾gφë²ð/b2Bü\—EЈjHžN%§>.òF) ËþåÝä<ÙrÝsˆg÷‡W{¡ë<”Q^WWñ…Þ¹®Ft$ýÌâ¡DDy+Ér”ÉøXGðiF¿úô”Ùu§,qÜŽ­9•·ˆ¤¢~5äfÿq“:Ø ö_ ATÏâz 誼ÒËŸb:²Ín¤J žõ¸¸Â’)vW\±uC´§é°qäöT‹ ˆ}mÛÒª´ž?Pð‘‡M)_¿®'#8³hb,{Lõøõ â•À]Wr¶~’ó¢[˜ Ñï4ì°#¯êdد$X3P…•åߌã¼+ï(ì¡É¹{äbPäI•µÆ6=QŠstBMfï}ÞNé¨på—åÌ”ØQ׫v¶´Šóêñ5L ·.G¿œè+$•Š6h5+kÁvš­zéÓ9¤ Þ™@¦/t=O.u½. XÖ¦·{œM?[=ËLäÂ’pÔdatù +]' žC¢…4#¢°éÛ‘k¡-ÎÏäÍ®Í$çYJ„ ÀÆ¿Ø¡MŽ'Èr×wDõo=õºp3 +7¾¨°¤9ržà(ü.MÁYËX’tó9qÑÄê_ù5êô*OØ€Ç.-5;Ö(ǃGQ'å“h˜ ‡¦L°ÑË£Öî9¤|ŠÔÀñiÒWóâl{)F !âfsir™2Z¡ü7ü×$ºÝ”ÔâkÄdSÇ>¶7þî ò‚… èRÛèn¹a¯ÖÀO¨VtÚœ}ºBõë¦ Z5¾úAI´Ä¸}-˜sFpm÷Sîë*¹tÖT ÊøíðfT€E玭ókßëÚŽ|m®ºù¶{‡5·Õ½½ôá˜òw9nZcÁŠ‘øÒlÃ,8/%.:g ±£,°IÙP¢‰·ävŠR±#x•?ÿ¥ô-œÜR ºCË òEÁ‘`ùX¢¦ñWÖ‚ü3ùÙ/z3<† 2pu‘ˆ„©ïHŒB½»î<)—Ü„=o@²ëHð]Ş̌ÄÑäQ—"Ü(ãüV–¤M©ž†5 €´ò½^™ž‘ñöîå£@7Ý„P¦ß{áF–î|—w~’Õ@rv¼äÌÄßvD›)صœú÷(Ì1Ð0¦¢œq²üÐô2_ž=e‰˜I?·D’®)¢ºªe§³5#‘RŸÑCƒ  ¨K\d¼oƒÞYËšÈùÁãMqUa/ª*ú¹«ÕÝn•È%–žüˆ§/®N$ãø‡IsQgtÉCÜY±ÀÐw_Ónu¼¼cn6H™ƒñN?Œ}£iÖoì›0R}6Eöíð‹7Þ=’Óœ¯8QÏõ ½¿Ëú®˜šÇ(˜YÀBX 6?Vm›ì„öB3°X‹æÎøbAz†Ü3ä"v+Bgjbcjšt8ÆMcþ ‰Àòë×`4õ×£üX\éL¤¦ÒK…¥ð¡Oü—_ë´|Ž“¶W„x#'âC¸°LÏòc ¥mšé÷š‰ 9 ¤û¤}$àÉÝš¡:.8Û¢) Êó÷g „Ê'yO1¿îñh¥4?Õr{Ú¡¿ »L³UMU‰¤1ÿ¸h2Æ÷·Ú; ˆÕ=Ö½’^ËåŒØw™E¼˜4Œ‚AÖ'D:µF8xXbQµgŒ(Üø«ÞWêøGõºìßʼZÇ÷¼n ˆ"ß—Û:«S·—ö6uM¢r7$>“ºQŽŒ™Õù']eQ<šÙÅíÂü5O2L¬õŒmº‚Û¾&¢÷(rÏC^ëņ;EÇw>ï±éNî¼%NéÇÚ¢y+ÇK_™`0‰Ø"››ùËtתˆcÆ~Ýt÷òÀ€í4íá–Ô-G"yÜ–^šß¤c|FE"16º#å–:Dx9M&y1Õ:‹oºmÓ’ð«Ë< 9ibtåX[*Ãä~±°+ H”äoˆ'à˜O{Ö|]ßQ¤ÂÙøcà6‰·5à•A-QÍ—o%@øIx™N$ZZ×"ã]”Œã»°šQc&Wfaå^ˆ;óì³ij Îqü^iðµ9µBP)`*òkÙuì®;qJYX}Ö)Ù¯,g†ÒíÓÇÍç'kÆ6§ xg˜éJc—[¯…rüø®»æ9Œ2Ä€)9Ò~)éHa<2 Ó»îᶋI=­™_î=j#èÿ&µ ‹6‹‰bÖXâ!ìÏ£¥›‹„ûÓC“†»CV¤ôIêV–å‘ÒWŒ8–¬qÇ'y×JA®Q¤E‡ï.¿»9!ÁÎk¼ZòSÞ_”j4NR‹²¹*,ÎäÁªkÛKºzEø‡ ¼ÕóFn}ƒû/2ã•IA.¨!ôé*ü}›W;Íäç|Ì'¾“ipÔUO¡+Mjð_Àä y¤|€“]À<–ßK¡´o=ßçÞyËÍ[§Fˆw l»f©å|ù×ÆÜSw\<€Š—ŸLp?LÛ>R¦èöWµ¿º…F¨ëö)˜žž|êN{Ý0kŒ€nrã[_§C¡Ö?Rhÿ˜AßÂpû¼c Ð?ëñ$eÚv+OTÂMVOE"·a¿•À˜¾kk߇C"ï«)ÎàíôÝTSŠ»ºÞƒ Ö¸ V‹ZNª¨žbðŒgæcÉ -“ÜyR$Eözƒ&y›û#³/xœ¸ÜøØš] NñôÇ7c÷ëÎœ ²çà~M›äþï=Ä_%¶:ÙœÏq°Äp]õ8¢. ¸ï—§îͳ‡ºÉUýd¶äÕZ‹q³{2™Ãa–Ï×07@…ëW»:~ìÚ抸©Ûi÷Œ‹'lÿ©´—ÙÔfƒÖ9}‘ã@Ë DÌŽXzœ¶tCw’¡õᄜŕ ŽJ'ol®` kÝê?‘2ÜÃé†"…)ôa9ø¡… ;êöh½ËB.QíHû¥YÕÕù ùjÈ@CĦü¸OU}®_ÓÆĤ¯2‚µxÞ¬Ékÿy@ßF>—ú¥²ñ¬hÆÿبÀ ‡G&K,Â>L*À÷þÑ܃íxl{î‚ZÅ…òcßÛ÷ ée‰ é V0¬N¥ˆb@ɬ'¾£8vü ÕÍÊ›Ž„þZTQ`)¥V.UŠB”Y¨K:^b}ƒJ¬_‘‹Y{Ž žy¤_°öéyœZ6Y![3Ü’23ò¨ –ÍÉLª£;T'ÞåNÝýc_°ÕÅ/CüP4ûú(]b4„¦Nl®_#Ž6AûÍSàI•å“©Úä·ã4êeý­8YÝ‘×ÚuÝ#áo=Ô’,vá x3šù“¢ÁƒÝf`A yL%Ö-]¢y¦™®˜ç‚{¢e°ñìöÛ$ºä¸8×x–óÛ?%wT'ìJÃ&€Y³/‘¼y¿ÉÙZÖr2Œd ÄBÉyù¿ˆ(Ýɺ;+ÏeSÜ£%¢ZyøŽê3g)ôp»,çœszÉVCv‡Ìì<両“ÂJ ¶@ö×»K 36‡:{à"lÝtí¦o Æ=¸\Û=¼à^²I7§D Qj+üúXo·mqò”Ìø‘J¦Ê*™z,ÔeéÊ8!‰;v©ûv,Æã¸V¯! ÂÔ½…k+/óßom¥b ZBcÞ>'ظ£´ÕÅÌÿÖ´™y%Ëûd˜):Ÿêüe8,P¸ '„ßç™åd7—^G;¬I7Uwìþ¡²^kâÛiÅ);¡ÝàÅ“v’W”„Бø&êñ’Kù%Jt”$®ÚòôbU¯ÇÖ¹ý0ÝQêëªL톮ªS ñËDàÇ„jǹCnÍÁg—thåÀ8šï‘v§%¤gXýýk{¨â€ØzcŠêQÛ¼Lý£[™«±ñ¸ºnúy¹‹T8ØÛÎZ€"”«»ªn¦´Öñƒ#’±Üsá-80Z¼îõÇä~,»§,amœ!>k3`ï›7›p›‚Æyw.œˆ$˜áP{‹¥—ZðWÇmLg–}±ßEñ{×úôcϦ‰qü@° º“çB²ƒƒ|]·‚”ŒMùçIš:as>åçy#¿~¤qƳÛþúý¤i’-{zâ&wŠ•&­Ijž¿ŸîØ#jS,Táo¿=Q<™òþ\ø3‰”6•4™3–8CÅ¥šgÃ~ãÎÌ?~DÖŸ"†ïî4ýmhÁ¦ò„ë”FE¯–»ˆÊÒ>ï¦ è¾Èº}¥-щIlmU»³òhÖÅ^SÐðó÷´b̤™Ãne¬Ïˆ>åÛÖSNª I©%/óOÆâ(¦^ƒ‘®ÀÑ]ê¢CU'»ƒóZçâ|@¾Õ;^ 5²ÅF{(¦\Ê°×—ƒÇßqý'ÃX .ò_¹&¹4ÍûéËëUËèú ½Ã~%cs§`0Öó ØRQ¸4À,é'[2uôÎ!É6ÃtWÖ†A++8͇û'„~Wöîgä“õúWëÚ"%&Å”çâ«hˆ`ë‡\Ügú £ÍSF¼–Ù$˜½âhÔ½â÷×G†±à»L:’ì`$TßE2ådxe»žÄhþeq¢àÉ{è·Ê7SU³âE8Q(ßáŸX¼çÏ^ØNªnEA‚×Æ›Ž®ÜS% ßžkƆ†Ø$>õ ™µœ^ºp÷oP½Hœ“¼²¥G¬·Fw«¹%Â!–(õ.cÔ¤ƒÚ'PÓ¾‚‡‡ BÜ‘„§bàá[ÿ%[ªÉá%jfîA–áËì%”«'ƒKÒXµ»?»1¾ÐÈ…cŽ¼DæožÄçKÎe&ky}^ã¢!$ªœ¹5Ðámd»{'Z¢õ•Rgª“^ë5äðùÁ^þr¼ë M¬<.7Q^„lV‚Y]Ãâ•a|†}]ÝŠÄ>Z$Ÿì(Ȳ`STiL±˜ëý]ÜÝÿã ¬Ž{2L&%›éÂnmì‡åA÷½)R•¨&­Qúq“›ð›nC”)Ë"ãÝàcK]ž[šda9MBé—´XÀˆPÂÙî*„¾Vž1#ŸÃÔÇ´o¹§ ¢“ú´³Iߌ5©”MÀoL—¦lø“Œù§¿b@0ƒÓ4œ£sGäô},øÏ:Ž­ØÝÆvå³;¢¿xÐ{ÂÅZ.Þw=Vˆê tû ŸÚ† £Ž¶Ôò* –-ýü!žÀÆ•.ik>‰*:ÄMb=|©–¨hf‘³1Äã([#ÏérT°VÆÅ%ï'Z…çA‡é¸ž™pšCêvÜÍüãÞ§Éãk,øŒ/Üm¾yþ*Çxä–NlÜŠýâkÖd¥g„i¶ìä:?"s:rg.ôg>CŒ<Ó‘ØOWSº+ûÀ¶ó86Ò†DM›Œï•F.Çw‰¡ a½–Hdk™gÍYgzœÏ]q®CR¿ÂW3¾&8eÊë ÝO9£t•s`ßJ/<‹ŠšmßúŽ,y•#¡TM—ŽÌ¯2l¤£fC{ò„}ªÚ}–oºB§m± Lïo{Ë“5S'в¯ìÖÒfs*T*éºâoÑñŸï…º{¡ÇìµÈ<5·Ž•ß–@o}finréˆÓ+‰—ö]iGb°ô»9ÞÔJ xÃ]´ Q]ÈÖB©ôµB”1 ÉÍœ-Êׄ*æQkZ+úþ>Eéî\àû<ÿÛ*Þìl–Êãî7ÁÀÏD(9ÿ­¥<®èŒµ™ÞZ*á3µ–¨²\è!ª§Mê+ŸÐÑQ&Á’V“;Ÿ ¬çõèÀÆx6;w©ˆˆl6 êƒõ‰äH¾ó-Tàïs1Öÿþ‚–¨B'Ï®©·ZwœÌ´{´~±Y¢ýÅð…ÌàuAªþ¨?–Þ`š êô8ƒ°ÕÕìÓL‚±Í%Yù9í{žZà‰x»Üü¨Ÿˆ²RáBþƒ¬ ºû:䤋þ¼}ð‡ŸÚCÞ]í/”ùîü=E›rý²-jjÝS%Éœ^Ì‘U.À¦ÈO, g⊺¥P¤|î÷Qõ–JÜÒÙj~Nù Õb×þø}9€÷M¦HfÏÞ7Hƒ8O MÎö¬±yÑÖžáï5¯Î“!Mú‚Y×—>Œ ÂTÃÊÀÌñÝt¦Û1)ñÁ¡É'D’…ÌZ†AåùW º®íuÖÉúÒÖdzŸ˜óo·l !D\õËó nB—P"Z·ã½=rá¼Èe+ ,>\"4F…HŽ½YEŒÂp cˆgýañv•}‘¦gˆÙ1•FtµÏÑLVØ™ÓæR£ÁÎ-&v}:M——Ïæ#tÝ>Üê ,SüwÈÞñdŒ‘ñXtLȘ[hÛŽnŠûøÍøA5¦Üš J1ÍÃ!»ã”Tùûú› ÖDMQ’u‰Ü†ÎE©ØÅb‰D÷¸¼D+ý«ªc |mÁ­¼ òØúXè›'Ùà†jªvÆéØX_é•Œ?Þ œHSâ/‰…¶¨Í»;ñZêE|±|ÔØ‹F¶Ô, p©qÊá0‡N™êa¸)ži`ýÙ˜™ÍÅ€‚p? 5’d¢¤ht¸ÑÏÃÔ)è…¨ ÛHô˜;KÄ [QCZܪ}Òó%MЗeä ½,»þ¡š<«œšá™øëÑîø×Ð^mN4¸S†d[óò œÏø…rƒa7Üsi>9„ ô”F—üi Iò¢òE¹½L·q‰&h+X´MYõ„fIq1i]UŽàæ0pŸ™´×ó´@ú™SN èamq"¢)þ=ö!ø@_úØ`ª Š€)~9²CQXA’~ÂŽ°äÈ×›z{×±}¾ÖP-ø±ÏüDiò¦^÷9¨s r½j0û è ·üzœ#±¯Ã#Á"à°kÈ}šs¤é~øŒŸ‹n±„¸Þlï¸ýé~ö6fŒ"î ¼t…?wVO¦.n’iœ€¦Ü\H+G¤&dšÚHŽ›xìY…—C P1¡³NÕý«Á[Ú_o«C/ 7Ùgag/5îÔ&QHYEêøÐô× ~šqky¯ OÕñ‹c‹-%UÇË7Ì0ܘîBÐémZçéŒÅégXYgpÒvÚÉv´Êzô©Å}öëê.‡D|öcç#C-h€mÀE.®L 0;î{’YÔ=Û·úU}]+ó%™¦ÙôÆðá¸82wªø¸oòá·7×>4•ÀRì U*;Ú“û€è²Ê&pKÞýåZ¸èGÒKŽÐg0¯sÌî~8{¢¢ó¸-‚´@»bA|WqÝO2ÆCY·Û$¤7µÍpúòJ›ÏíÓ:KôBežÚAs'È®\Rª›ôà¿‚÷ÓâÅ´‚‘?+n*f~¡Ã+éKßw:$¬/ÏƯ­hL`Ȥ©Ñæ Ú:W{±>šaƒ°(ž ”¤£|Ýô¦Cº~Ö³5æ(ÈÅûpîײñÌküu¸Ã…#+3¬#‰dl.bßK§¸¾x)xr|¥ì{xeûÏÕj÷@ã=sšÎ:òÓ:„1凷Oo„„eQ„ttdÙÏjâÀîµ ;°d²¸TaÐ¥[&©~±r2À^iÁ­üïhÍP2ÎoóëóqÝ-ñÊÕ~…§¡…‰5ÑN’’Ù p#uì"õògXOr¨OzPÌ+V;ñŽÌwƒœoÙ2¤6`zù.Þ%žïÿuèKƒ©´ãÞÁò³o, !Ñ‹ðb°ÍvÍ<³ŸKœÍð»W.àð6ÒŽ‡ëNCÛ¬we­I—U_áGL7}(¬ì»®"]8¶1£ƒC/ƳåäFè·¤mˆL†ÔDžÀÛ3W-tŸ;Œ®$7<0‚«†¤¡ujh+ ¸ ÚÛ€Ö=µ`F€-¼¼=ªd/Òà ¹Ú>p­ÃøÛò9 ߪÉ5WIZêA lŠ¡WLs×9 ÔTuÓz2D#†‡Tk«¤üB*)ûs„å§\/jŠ®™ÃjE&܉âܺCYòr ŒtYHª=8…Ži ûbèˆ}h#­ÂóC˜ ÅëàTvmöv㫃óú´=ü¡mAåvI§õ†}뜧Œ†‘ Ü°J o,"ŸÂM¬ ¿e6qwŒRHäâžtÿ†áÃyåN#wdŒõ9ÅÑŽ&G¤t>Æã(ŸchoD¬PCxy‹9ÒApó»Å±a^É©Ã^-Å­kÓ±s¨\›Ec ¸@~Àé¶é¸oq9I¸ €à©ÚU‡îÔÏÉ<äa8^Õ«?ݲŠÓ( | JÃ¥B{©Ux#ôåc¸§ ß~‹¹ªT•øÜhkÞ„U'%O=‰RRÅÜ°¬sEOLñËå6;tB›‹ð>\ÔŠîú,¸nI—;v‹l ð¤áÍBäøÍd1,—ù¬×òàg8ý÷ãe5HºéýÉ0NúÐ~ùô®ÅnsÄ&IY×L¨ˆhWÞ‚MôÍ úà-=„àèuÏ j&qõ ½ØÌÜIZix­(û8NkŸ°ªÔpHFßnxÊFðVè¡Pe“ãcÜOdWT1шûóÇ×w¨6×0–6?mMn;û5*=’h¨«X¼ë«"“YÑڧŠ„d†gšŽ}8„g¢9¹’ †ð¿ŸÄÉŠv6ö²÷«2´é¸ÆÁ,Ì]à‰‹ dçr‰Þ JkîÝÈ’~™w:ò[h²—ô{Æô÷ú ‹O‘`åf Æ;ôF}žs+:Ù³EíŸ Aã7¢îü «@Aè$µ¼kñR­è\ÜEpëOC«¼$H‚BÕRP½ žÓìÃ0—2íà§ÌvN,g7 M¥ªÒ÷q׊À‘À> ˜*e´' Gy«ðtÈ’¡:Âà?ÒhtÇ„ó1ó4T?w“çÈWzx!Þ©²$)Ä €²´¿dœdtr†ª'bwG’”¶äŒAØ/â8ü°f},†, —hÆ\§IyŠÂ#P‹!#/^ümê)=œ¶ é½¸ :× Eï¾èøMµ£‚7>Å2>paä›nWÂy¬ë«;zRÝ]ôkŽi'èJÌ„t¨–Õ|Þ++Sd3;/èãe?ŽRƒªq,êØIªn2x†'w‡ô;ioÄ–W“})Aq‹iJy/L%´Ø§Ûkqª³´væD”áFÝG{e‹&¶-o&«q¶ø£ã›ØòÚx³å}§rKù(×½ òr[¬¬\Üšó}¨h§é»s×%v=¢­Ð\ÓÉ.Â4àô†#šõþ\zbõ1„!Š1V¼\¢R¬Gè«»xE…ßÕí_°¥Ø^¡»\f]ü¥2_ l¡³µÆÜÉSMÌ<Ó†UGôÙ挭BŽmc.<¨&2G !@¦FîQ?¼¶õ"ÓªÖ³¤Y«qܦ}c¢ ,>.÷^È•Ÿ/Ž¨®ö3¢uyØ屈~bSE|[qåõþ8 e þ©¹C²¢@-?46oá°Û³°á|™%KJ{Y_ ©7çK5ãª{ÞxùZ©kµ¯™ޓʺ…t‚¿Ø‡ËbÉp¯š¼¢…¡Â#g4NÔѦ’6ÿŒ[’mòøe+ž;Ôy±íÍ"£êü² l¬p„ÕÿKOjÁà¢OQŽÑÓ¹‘©øÕ¦Ñö²°⤂䤈Zï1›$u連¶ˆ¤ZKµB&:§cµNN¿ÃåÔ@~ê¦b™Š^‰â*0¼ùÄ » óc»±FÀoP+„p5x}ðTy\…Ëò ¶ëVØÁÍ,†.Ø׎›¾Š ªËo„??žéÀ<Ûoeaä1ˆÃÿÎŒ‡B¼54Ÿ(lâãßé O-Jx2è§Ý…-NdØŠ0S¼ôɳm¡¥è‡{¢pÄ­^ž­óžt†R»ÙÈ´õ›á "ÞûÊß~0#:XÅ5ùÎ%Þè¸Ae|à<‰”ÚÎÂä¨B]’ 5Ìè+%$äz¦ê:?¡#6é#JpÃÓó|ÓNàÒV #6‚ e?-¿F¦`\ÿD¾ë£M½­Ã%4/y¦å@âiãý-tÕãÙOÌïÌT%qñêíKÀ¶°ñ£&°Éð?®š< ñi E¤b×2„/¶rzHÙûí"Òè• …_`ZhñVªC5÷þ]ˆ¹Ë׿ܱjeqÇ|†KŠ¤ïÝAkº§ŸåªÐŽvÖÿìs_4q\ЈËíñ•úeÎ9ÕMG~ÓÜ{G]ç2q)w©{¯˜Ý^:cÑÎ'ТÄ΂Èx+Ò”÷ꮃ©óm» &Ø0~„Õͯ7 wj<2¶kVü$ÔGjE.³©æ:62–² 3Ò$#ê"MÇ1A18T®ì3Þܯ²FN–ׯŠÈÖ¡µ(:‰–ŽLÖ!¢ï7›Z×ëŸZ6´bé¤nms·Z0’&™ý>œÝ;ÝïÞ~¾öý@H¢­4 á>È„äS/Ögì:÷í{6mcÂï䵿ô²^Næ¥ùzêNñ=9ê2»>šjÉ-m¢ÿ:.©Ôâ–uC,éŸC\“©ä"|u8 ÕÌÔ¶9ÃUÎg6•¾³VmG|aX¦­ ?³e)5ÉÜyŠÿbªÍ&©z×òIzõyCß ê÷4 T©K@á“k ‹¡ÝD÷f]X~nD˜‚¿ä/Dàg‘8ªØB¬œ×‘±Ãž}<„ÉiÅ/^q\ clÜ¡CÒš€ «("‡ƒÜˆ/ÛÉË+G¬#“ª¨0Ñ÷µ¥û™ƒ!4ÄO¯ÜÉý­>ÞÉ9{2¸Þ)ŒO[\ȬYªÐÄ!åýr9˜'ÌÂÛÛ¯0T‘¬ð«pjІYÃ.¿†ç"vª‹nè0‰1§Pð«:L[åhûò6ØÄ© 5¢C½È«%ž…Ùè?‚~´Æ5²Ë `¸!;Z¬&Šî~ž”¤±&¤ k&³oÌ“¤üêO6ë²ddQçIÜ唬M&Ûrðw6Eéä8W5¾wËM©ÎƲuYBêi\=qÃcÚYb8pz¥ ä‘=.£Üæ}¡gúÃÈîÕœåT<Á`ÝülÙQÀͯ,»]ˆçsAù2µàNøU©0¡[°·öÓm”°õÒ¬.±£²E8„ «66ÊýÙØN†<è¶LÝÆ8›+~>ª2Ž&|EúÎ,Q½–²ý ÿ¾o1M’Þzs&h¼̲y ÊïÞ¦=_Ù]XÊQÕYßíe¥#SÆŠŒDj‘¿s5¦Ô×ç§é)[_ù=ç3dªlå^W&|‰žÜ¶õ“À¬ç2ñ%çOŒGD÷/’/?_é˜ÂZ崶߅ª à„CX„¦œØÿ„Š‘8ïKmôTø‘çºV¹?ÌAÏ/ÊE#/µ½›'ïv·AO$ò‹4E¯”à›á<&UÈñ=„º°Ö"yïhþ ¹ýiFdCЪâ1j”è{óaß΃à#Í)¿Ú½/¿3c»Xð •c¬ØÍî½ì¾Eâu ¢%#`¬ÿ¹“‹íç3kHðö»´ˆYë…äV\°ÊO-ø0Ó 4Ì Zþ Wx ?¤¶ùrß¾qÍØq’žn u°79{ýG ܧ€ÄPäTK±Ï÷BPÞA»ŽÔ3z褩Èä¥þ?ꤕüÃ\+‚–Z+j2jךoêÞÐ^§ØýJ9G[jˆ)‹€ Ã[ý69Ö o0?,Âjâ5„¯Ž‘Þèlý9ÆH,Ô†O?v«ih—Ò½()8i;·B:Ñèå éÖÛ¨jï¾ã£p¸³¼F¤J_Ù=Ï—ÑÖ†÷cbÉe}>!\òpð…´Ñj .ÿþƒ?œ#ek‚¥¬õ:?»ö³A–=â ¯JË67n;Øôš£Ùh /ËŒši,ÎL}†|ŒWH·Œj Xõú0ŽB•9m,ýKG•xMR­Ú8À-Åé·ôEÑIëïø·ä6dùHA¬®Ï/‘yžf3 Ÿèë†|~ŸK­çíí7i~füÊ7÷r~ )ëA¯«•+¤Ú‹‡H%{­ñTêÀÜ[½”"vãÀÛfiÚl… ÊW%áµ@¢wUrñ&&„:n1×#åxA`çºÙU˜rjeŠ™1žRŸ)¬>"„ÿ>ùbûÒÐAñ-ß—à 5¹ö´ô°]‹Ï–n'O]iƒK|~K'^}ÏWtÎ<²p‘h½WžÌbÌ\ÄÈh¸Ïs½ðHêå…>ðMA‰£oýÍÝÛjq ùOÏAºz“˜&›¦I0?•Ò6<‚o§º¹9á4‰mi—:Òmû¼³ƒ‰WG&ô1t'³.Søl&·®oŠÍòÃ>Y<õÀ œ• Á´¢¦&¤þ…¸_]ú-¼UMuVÆb,%ºòs †H û,¿:ŒÃÚ²=]Ö, ¥·£ù¨–vSL]?"¡Í¥E×_pÁS+Ð8Ugó&”ì¹¾Ìv¹¶ÄóžSµ7â^÷— Ü4_%€ðÃòŒkà¨U© šÀQßÇʃªu#÷S)PñÙWÞGûSé´M›¿4çC¸[ Ãý˜‰_훃4¨%ÚvÎ1<A¿ÃÃÙ~ùB}ô¾3X8dÒÛ~!äݵXöÙª¯ÛF25«Ü†Gëö7\y6•BÑᥠ50±ÒosèQ¾©ºDºÒtD,nž®7‘ÄY¶‘Š(¬ ݉ùº a¢î~‡cØ%ßÕsº½êz£@yþ½6°“ƒe†,9‰Ù;´·¨–y@‹ÇGÄÕ´ Ý=§£î”JQ '¦ô;æž±ÈaÄÄâ  ¶ºñ ~°pMñVFòÕ«ô¡¦`1¨L"ww5VŽ7-–FFãÅ5²É„z2il¯K˜+jxVv’cº¯›±ðñF3Á¢Áõ©­Ïò ͬ5ZQª”c$qB¦Éãž<ÕAžÖ_¿Lï íjÿU+-7\4©ëdœÛáȨ}è£Iå¥äKnæî¦6èÖ^Àšw+÷%‘j1Oˆ) y!á~šCmÛí ½ž©ÇO³Ê  Cº@o®´<T°ö.‚Å«ô,Å>‰,ob ›\ó¸a†Ð‹DWvZbs­Ä§—­%âÇE—Ý@‡ÝZÍPJ2U9ßKößØMüö]0²ÜÝj±úJlŒü‹y fÚþÁ„¿ôýÊûLœr#ïCm;o˜¯{9WQÞ,t'lòÓ-XÝDƒ/NÎö:;iŽú}¨"ƒYž¿)ž€ÉÖ”PÉùêçrdúÁcòeˆ¬y¦n  ÌRqªqh]Ä3Áb‰ ÿÒ×¥›vƒ†žÒ…ùÁº¶túµ9Kß(ó 5\CÉ?ˆ…%Åð¥;HA•FžM¦ µŠØ´»öíËœ&;.LEä^½ªåÜšeERRŒÓeÞUúp)èbŒha-âË©Û!Z¦X›rÃóTÌƤºÎÕRçwÒ*lG1%¼¾ky;bH¯Î»§°ŸöÙÀómEáð ‚Üø9]‡AÏfцr­ó@¿ì[Ek[°0Š’´´·7î4­keŒã°e"i]$_(zQF(3yƒ;&¥þ®ËÈ–y÷ä/¯©ÀT¹B8˜*/Šé>1Á™‡ˆägZì]/^шQ:a}‘PÞµþÞ‘äÛ¸òͼ 1ümµ8&-¹ÚE¤ÂqX@E5(@,ÍÆ@2@®ŽSü¸Ñõ³–k›üí¤o¼óWw“%÷#…6 î¯mÂWÉú’×·)1òÆ%J£ˆz•/覑Zµ®Ô+‘—=ºíÕo¦Áyùù¿V‘ƒ?}Üt^â2 ‡*e'¨‰‡#Ší¯V§;‚•|À¸V#|è3“Ê]@â>P“ ßÞѯ(aØÁ*¿Ê—˜w5$z“}açºX(³¥÷èÒUà8»vNL47Ò¸]6žRÒÁiI­o‘[ɽ ìsÆ9„K^_ ç¤=8ù6çf6¨áÊ 2|Îg ÐöFi€So='C=’ii–Žy2ì¿&§=ØÝ•½Ç¼9åM8c®›Aš|ë³i3ûXXÛ¸AYX*àˆ&G‘l¬^ÜqÌ$Jañ)1 s$ŠîÇÛ¯·e`¶nð,¼I›ÊqŠN¤hÈDpÚ ;Þò„X ì°8µ Ø`R¤É«/` â¹a#€Ú- Å…ûx ÍW:*6c hŸ’‰C£Èö³¯½£W¢ymëè4Æûa”0K‡ÂƒÑ…™µÑŠ£àSÌ,y-yå'ù£NÒ(ÙRaF‡ÒX”#,?V|í_׳íÛa ÍuEóèW:>P;ãü]SÜþ‡²ýNáSµ¤÷Þt¨t\¹ìñªs öæ,lŸG‚S~Á¹šÎº >tÃ䡿Пˆ-¥¼^¾À-©Jœà˜Â¨Ž‘ë!ür=zþY~4ˆr6ý"¥¬Z„oSæ® ¨Â‰sÓñ·©…WOg8Û1ÊV–ËX¢è»]’[Í—ßôSä#‰·ëKȳÖõŬC€ë¶•‘ô†7­„.L%ož6ž6f æç“ÈŠÑä?½ âÔn€S|h›¶ðÏ:$dÀÃáß}ÀO$öŒª(žYNE+g>W!ëµo–bþÆ]Ù…—µÎšrο»ò óg]\Hen8'œ_O;t ùŒŒ7††d>ø–%Ù阃I§G[‡ÁVØõ "MFOç}ôüð[±G!Ûa:­’dôI‘,íöŽ¿2ðX5!´C¨‡c*ÿ¹ø‰æ›õ4Xe=«Éèf¨zÈéÅ‹|‘5J÷Z»€ÉËålÖhaã½j“(Ü×/o¥¬e8ãÁ 'AÝ>,䙂ÊêGÒæç|“wߺ”Õƒñêò (màò’ô$«Ùœ ‹üÈOPä½jÁé‚¡N³Cx/¾?ªH{n¦¯Þ0uÆ] ð:1ÖÀ63j O×$笴ƒ;Û†f~Y ‘ºœ­¶–§±ç¢y­EójQüé”÷P*ÙD*9òÒ”E¼ú›»B‚4ejîº×šu²ŠëºlägsŠF5{ÞzÐ9œÃ¾r‡y\}XÒ‚vÞIÓWÚ]„…üzÜÈA ÐÇâMj‘ÖŒ Á>Uþ 8C‰We»ømèñµ}ŽO„ï)çUÞ%åÞ¥PFŽ[B†?,\±ìaúŒ³Föu>õÔûX¬6쫉5¼œõ+OŠ¢K´©¡1sEˆ×Ed)ºå7 Ñ6ÑÚB™#m%þ>{ž’'lš«[`?%=yzÊÒBégÂbÝÒ‰yNñ쪳²b›éÔ kƒ­y+Í…yicCZ”ôõá™Ï®y›ŠXiË›M¯¿T8õš¿Ô@yã¤7gàhàÂLD.§Ù"åüŸ vümìjËTs À‡úmïÇ-á<ëÁÉ ƒ¿¡Uå«ÆÉ +:(+ñ±jœ˜Ñ7š^ï£Îó"}hÞìxÊgÂþ1Ýø¡êžØeÙÕá9Š–9å “§D-ù¾©¤J™ÆÿÛ $öÉÐg×àíaX}|Þ¶ÓŠUi6k¬ë €)¶>v=]rãƒØ;ÞëQ„ÜÈ{UE†F©z“4ü"ÔãÔú¸ÏËRIücûD$æž'ýåhæ:­6³ƒJQÞèrM½ª69âSú‰ãüUÅ)Uˆ"-‘mÕC]<âÓñPÞŒXV+ ¬*¸KjîÈÖ÷«SŠ]辜Ë'ÚάÝô¸€üõ‘A ìj£VÄ°p ]´-qYé.Lóº`#´Œ&"r„Õ ÐVËöàN&è Š±MeõB!¾b•Wë¬Z ïeuÛkE„„†" ”ÖÍmŠw5 õ¡}Òt^æ°¬ Ç›4´‰09ë§#Y ܳVž¿=–¬ ¯¼hUAœg*×1›»*¢`Þx=ð_šÑ*ãã•'ýó»ià8¤§½ÒÄ%5ºrjgÕå¯GìbI ‹”µ;Ûåäjn8Á²š¢½4ž—¢å ½Íï8Q‡âš)7°ê Aê§öÁ©¶'á'ðöò YÜ]^±„Lp&¢µâ%fPäµ™&ú}‰ä¯Ž<³·µ4z½æ|¥Á_aÔ†äYÂÇM» ýÒ Th\‡{D7C5åÊ- ?Ûèõ3&&¼.¿›º¦0 ú±Šny¦’qUï 7†÷8WþT)| zØáå·@ŸØ7ßá‹°Øö5èà6WP¯0õ´€pŸÈKqÁ× ·ÏÞD(›üg(ªø ^ï2AǤq#uF˜¹˜çXi³ó/OÂP™'~)v>s©^U¥\IZ ß²Zð„žO?Ú A1Ǽœ[×.¿übß-H#¬„t€Óœòà—qr ¥Ú?O˜ÒKHë j…™¦ò¥£ óÓ£ˆÀ¯6 aZÍfPá˜ãÈAfZ 7‰|+¯íCÍw_±H9ñ“8¸PÙ×  ¹Êƒ¸ÍÉ1_Ègõ¿"i¤ ‘õ<¦ô×w&´(·+Ør5ÉGÀÀyQ6ayB–Cwο¹ø%dîücÇÈ2С–Íç\RÀ"WpI„ÐèVÎ%Жd>Ÿ¨‘%äBX„kb˜¸›OŒ,íüãmÅ»z̃1ÃænßÒáé뼶o®M¢0J°3Ï2ÄÄ_`ñ–˜L‰¥UPcð­°'¢ãåüë™ÃžpˆÝÏA¶Ò€úý·M°÷[Ký¿+5hÃΪ¢8†«§³º`á3H'´ÿéνrmòS¹µ;oH–߀Tñßýü‰:nÎUBÓ¹ žÝ¾1EYÃÈ^OØ|£°pBéå\nx­Ëô›€,àÄ‘¦­èÉÑc’J’b»”W´t %fð-D­²"ÑEU¶u¡VSCòœg“i¸W€ `—?a+qx\õ°Œí};²û¡]Ýð•fZ7ÃÛ/ ŠûÏæFýN“íÔ†âbo€áébòÏÍ·.€ÑÐÿÃp6Ö(ÇG cžÊ[ÔÀÇ ^ê“© qúø.3¦ÄÍrÿ r‘åê3™÷*TI ¦zâU*ž¬ï›Ì).DÜGÑÖ´;l Ýb 4çú:"ééÈ™Ydrz{iµ',Íû8÷z”h-mÑ :$¢¼T»còCÀ‚Ž _Õ1Ú‘ 1½iwaWˆÈÚÓh[iáW%å7&mùâr7[W¦Ú@ê—Ô£Bm÷M(E˜·¼u$L†¶VÖo+‰’–u»£G2­Ü+Æ>åS‘Xíå•óá?­dp‰røcÚÌÔ7›à¯â–ho2ð^þäÛ’™ CÝ*ÏD•–^w±=öb(©cœ³HNˆëÕ’@5=¬ÌXV‰áÍÎ :™_ºJB2ë,rÿ´ÓVÕÝÁó1{œ ¶g momÌÂx¨IÅG¸¥ÏRÙš®/Í’3q"JBàÚ[pë3T}Áñ«v âÓút¢¹à\æ¹ÚýlŽp;tåASܱdŒ­íÛ@E „A$…Ž§5s{L¡K¯°³ÿ!+£írÝ& g PBqRÒTÕ¡) “ÍýƳbÅÓoH’= (Sœ ïNúè!Vpì½ ÚÊ­²Á‘7¬¯`×–n=¶5ýlÔz¥‘wLõ ÷Ðe}X40dΰé‡n¿ÉîcC÷Çö•âñA'`¨z“ÏŠ Ø8ûP¥»èŒPø ˜ ß Cî;¬ÆO¢VÔÓ3/t qV Âá¡€ð½ŽÀ`4E uóÖZ°id$:™ãR?àTb°iäŠ[´aQ‘ ~ÃÓ…òhàdžÖ$­ßIX½«(o« 7ª4D]öØ2+„}¥ÜðÚç¡ûŽ®¡ŠàêeÙô„"‡#"'Ä$.\?Zí÷º°è4®‘wÑS]•\ÁªXÖ*NÃS„üª”¤C¬ìgäî‡v$aÕuQxôtZƾ0WdÆÚÚôÁ±3aíŸ"μ0¾ ¾Õ½4¾JjhÉì9ï‹~Ùž‚Ôo2â®ZaIZHgšÃ×áP‘XeÐÖšpàÐPSb""AJl›ÙòŒ´ðÌ;g’¼ïäÉc˜N![Å•N½f.“ªš"õ¡•5c²õÛìë ãÉÕÈçîLLšåȵjÇc™C`lQOš'\>ºÇ"?Wù„ùÁïpÙŠä›&λ5%Ä> ÑiäŠÝTŸÿ ¬i¡Ú…–v²¸@fsß~° Q#k¸é!IÅÛ)Chg¾¡ &K¦@> > stream xÚuT”íÖ6]Ò%H Ò1twˆÒÝ0  1Ã#Ý)"¥t R‚„RR’Ò„´ !͇¾¾çœ÷üÿZß·žµæyöÞ×Þû¾vÜÃƬ£Ï«`·…¨ÂaH^ Ÿ€$@ISÅ(â$`c3€"] ë ØŒ (&ù%„¼Õ)ƒ·@M8 ðÈÓE%b’A‰¿p„$@äµhòÁa6%¸›/êàˆ¼Íó÷'€Ì JHˆñüv(¸BP0Ð!!®·Á €> … }ÿ‚CÚ‰t“äç÷ööæ¹zðÁ²œ<o(Ò ñ€ ¼ v€_”Z WÈj|lG¨Ç_}¸=Ò„€n.P0æqëâ ³ƒ ·Ùúêm7ì/°Æ_ÀŸâ€|À…ûãý+öÛÃ]Ý@0_(Ì`u´U5ø>Hf÷ rñ€ßúƒ¼@Pí-à÷ÑAU]è–á~`Ô éÁçuùÅ‘ÿW˜Û2«Àì”à®®Òƒà×ù”¡ø¶î¾üšë ƒ{ÃPKöP˜ý/vžnü†0¨»'D]ùæVEðo @Ü°#ÿ¯¾nßFà/õ-”Ü `KⵇܾP /‰ð„ø£þÓðO‰ØAÁH€-Ä #øwô[5Äþ/ù¶ÿ¨À\àvü€_Ï¿¾,o'Ìsñý7üw‹ùµÌ µ ¹ÿPþ—QQî@ñ x%D@aa€˜˜ ÀÿŸþU‚¿éÿÖꀠŽ÷!Õaöp€Ä_,nË÷7¯?£Áñgo8ÿÌ ¿h€ãßóo! "¾ýþŸ·à·ËÿoøEù_çÿ¿O¤êéâòÛÎñàÿ±ƒ\¡.¾·퉼]MøíŠÀþj ùk£5!vPO×ÿ¶ª#A·K¢spùW!¡ªPˆ vükŽþîÃmx( ¢÷€þºs¼@ÿ²Ý®Øùö^ñ¸íÖoäv«þ™R†ÛýZ?AQùÜ6ÿV €·{jñù=Þ~>y븥簇#~õTàwCÀí<ÁH$ÄùËø—^Àïáéêúû:´ƒz¸¹€|ÿ‘ì‰@Ü.çï ¹=Úßòï›ñ€ ¦'à`©0§·aͧU tÞ¼kƒ2_ØÖŒÓ9yQÓˆÏsbÜÎ7™!‹ˆ…”ÞÒ¹Žcù¦+ÔvC-ndãsݦ‹'—ÖÏôFÖš¦†©»‡ ·jºðéy äן\¹?1 vÆl@o}Ä–ëî)N¬“OqêýQͧ¦«tösÄÄšîúÑÇw.KGyã ã,‚‹ÇØòl³ÆiXp¼ x\äû>$cÇ'_Ès†n˜=ã&ð߉*B™- &œûÍ—z´ÑÞ§5£aÀ<&ÿ<ò¥¸‘úèî$ª¤h¡`Œ9ÁÞíFMÎüÄ éì™É.B­M¨þ0ç«TìULÙQ„ $}6‚ãí«IüðÎÙ¦;.WÒ´{ŒKñÀÂÔÅêUNäÚ hÆQkT XÌ#ä2ï* )'Ô•E-ùÀS+ÏRqÿ„ìNáàµ×Eù×( ±GhêĪÚüØÎfœJ¯œƒŸû•½à+2ñôò~ù‘(®x:âÔáÁDl¹áå%·r…±BQ×ÒâX ã.ÃjϸN–T›%"‹L¾p ÷&!‡,O 7¥r)hw-Àþü‘pضøÔÊ+YUó(ÿ4ðôë;–Œ€òoØI[Þ/½Ê¿öŒE^S붯?n«£¡º>2q¾‡ïŒ•@9¾Ù%¤RÎŽZ é6ï7†»¹¬Xqì~—¢õY–6¾Ò6™#Å/¡v 2} Ù¥öÎë™\–Ž N{Í®N5²›|ÞJžïû.¸º‹sÁÆS½€L!e׾Æ{ K]{ÊÊvIé¡ó¦R:ê½m±õëgìü'}h‹“ùÕ÷c5FášIN¶âEÚSÈc3ŵ—§Ðå"ûLÞæà:s|µBÎö S,k‚ïñK~z#ÖÙû˜œd‰-Tq&ŽúÕý~> ]å{?-ÿ>›°–nš¿ÚÇi@\(ƒ÷xvoY6ÔOöÞz ‰ÄO­‘L/Þ™^Kœ—Ê\ Mêþ9õîôU_Kfm}&ež0xû·$V>ؔܖq±gÜy1—|÷®6r«Påylt¶x×ü©‚UêÐÕ QTy?ÉšŠÞ€Ô§½Íx©db–nQŽ®×5é䔘d¥9?Œª¾ãKÀt_3Í·I‡Èß]ßÈâ `’žAŠÏ {ÂÁé?®ÓÝI)øàuµcª»+ør‘uEØñmýFçšê Dý‚7J¿‹!{š0Œ ¶úíóT˜)ÊÞ×+^X ò¢cÛÒK;Rî`>ÖŒ|(ºO‹U}¯’ÝBbλ÷¢3\§Óíéá5v£ÍV ´b‹cU°_ ›ÝϾ(ÿÇ›8…æ¾gærÏÏúnüÓ¿ˆiÍrvñ"9 ­ÊTÔ9¬æ’4ÍI£-ï· ìKÌ?l·ÒõûÍo¢rçî H•&+[MÁ_•º {°‹ºÊ+0Îì×®îžyõÃ* ö ¦­ìô) Ëï„“~z<ïð­깿1õ"Kwr¦Õž8™rJ^þŠé§{ pJŸ¯"Ÿûi³À§è@éåŸÅ–õŸ½²+'\qÍ~îåöf¼aºi16r4%y wª` œkÞ`ßÖBˆ2ôV?µÏu.´¥™y=²‚·YuMs·íõ½åmö6ù(Z`7Õ²PŠx_óñ|bõÉ…õôG5мA«¦-z_Ì€€Y‰"jµ0Ènb­GðÑIjåªì@Lï ^ôàãr±¸e¬ôɧ‰x#SwÛ H¥>UmìTp› '*r¡Ñ§´â¹ H‰åÖÜþac†Ëaèÿlz0)”R2²p`®±¶{Vv²oÅŽŠ»Ôá4ú\§È“0³d0¸ˆW¤„HÉÈ[€¿%xB›s<Î8Ç£Bœ[wÂÝ×ÄöæКX‘h°QŒƒóNÄE¸ÕcÑÉ—îºWû.ó’w+ñ_õÚ¸ïŽù~‡æZk|ÀÛ¤üªù˜¹U¤ŠÃc‹Æ^ºOE¯é¡÷‚Ù¶¬CmñLJ8EÔº@&­X æ5)ñ;ò3‹/$ƒ†Êlñ>åPx·_²oï¬Ä€KiœS™:çC7Ï.ãÑnøðÇë;Ðâârª«rïȳøø‚í– £­Ê¨phÆØ£%š/çlI)´¨qßùt U4ÚKI‚Ê) ùÜuç÷,s"Ø„G8${Ù’Wy/õGz&ùÀ:U©ë~öJ–º¥òTzwY){þÓ°gƒßO”Õ"õq>êÞ rC§+kÅ•¾ùvZ€»Ï¢¼Ùê0ž ¤\¢†añs‹V/ËΙfµ„`Ü3Éä®Þªª8 ÿý¡]Òþn™ ™fûÜq^yq'À'îãƸ„»Ôæ!øÙª%K&÷0&óºÃJ&ƒ )›óÆ`kHÜK¯ü÷l³š¯;°’D'"´PáŸØS,ãi”/ÞÚH­Ëªo7׬t¯9Ô3gæ)éC7¦:ÓáöÃ=wc\h<%áB¹¾© †ûóù“û lêv)ìöçKÑöeòË/¼¦¾–Ç\›]½qº£¥øÖ2´ˆrŸÏ÷=ôØX¨ó_÷é> …Ýæ¤é¥eY‚M.3x€º|[mV#yrˆVf¤ÏãõFt®ë+_Èõþ“6Êiّѯ»œJ­ßRÇœY ž 1Q®ÀßÊséNò$LˆF­JÓ± wãÙÄù(wïD‡KªL gQša¢¬’‘i•/-½¿-dg()”f§°;YæÉ\\<ŽÂ>bØSQè”)¬”ì×Ôur™ôš¯¼²å™gpMs5ÄÐ$ ˆJ¤™ÅTYö65G…>ycÃØ7$GI‹ô:PÝÈ÷žDrv30E3Nð¯ò®`j{U‹`v‘µËû¾Ž ‘X6²ß OŽ|:;D÷ ˆ@¡}gÆï/ðë•6C;†'ð!m{íïíëÖ/d÷uäW±–ëE=± T„>æÒbÛØf .CY>°Î¶GHÓ¯"0BîÓìõŸ÷|<;Cƒ“—Ý* ¸ÃÊ”ð¾á1ìb¼DÝÜh,@‹}%Tu‘]#÷0^ëî;ÔRÏî:¦#]N·zÚ« ´öû†oÍó»Óx÷o‹ÒµHÚŒ¨ :¨3wwˆG0Gúª&5 Ò3û2$h ú©újøÑë¥Iª¸¸5K†šªvLÕß’Uó¦’&\L‰TSêîQAé?=»ÓF+™¢FiºrtÖøC­Ýo-el†¦4_ýXoÂRBf0ÓGUw.i×ú¾šÒ ™2¹û¦ï—–ÁW´ä)­d«A‰´¾üh›s‡[Fîpc?3ð×æÔ 5Zuøx*I¥ ãwŒÜ^$nºñ®â/÷ç¦DÆÓ­ÞU‰¬‹«H§®ÖXI`$÷çáÏgI®6žv±m&,Ì!·×£Ïê§5Š“£2д2´n_ÓH ìÚ¹ãd Ÿw`ÈÿˆÀÙå*|rÚÚì„ÿÒ!™0X*ÊÍák¿D±çh¦ŒFˆ¶oÜÚ(°Ï²»ÞDèYeÇóÌCg¦ÄžÓaO>2CMû±ßq AÐêz[¾~¾Õ~òñh“Þ¸}âNýµJ‡EFåM½ŠÅëµaZ¿”äÚwsݽ/kSôÎhv"μÛRú,¿A½)&h¿ŠÑuZ®ªYíÒµìfêúÛäÊÅB ªŸ¤]n¥~ë ñnVðôü$‹2èR¶k¼ÙîÝ·[ÞA=?«À¥]1FRýÕt£ˆÈ…ÎsPäýä¶à¹Käff)þ®µ3¿¡Âìb­rogê²íý`+ÝÙäØ—êöŽzò­Š—2–¦bä\™ ˆPÝk³±îÜ—0D×qé‰ú’¦cÛØ=®&·€Úô ™K1¶‹Bae›Š>X$RÂy ž£¶öi85^ø™ÅÏôE_°7qºuqwŸ€RW÷åËÃI¹Ä®MòZüüJ·ÍÏ*!ô$«¯¨øÊ{ÔŸW.…¤'6›î®Î*n5ž™m[´b(R1–…‹8†d!ê#HhU rèì¬(º¼K;b0¦¡uï\ùR¸Ç‚„4,r][–6ÖÕÔ¯Í4¸â:Œè›ƒGÞ陜Áô¡l¹dw3ìÂôEÇhMâåü#â¸W²5zMK YùÊÏú6ˆ™ÛŽ(÷÷.‡!oi]ËvyTùíVÓbÆXŠÝ“ëp"Ì4ò9I$Xw!Юu±åP5P#;_¬Þ®âë"‚è㡽µ,‘®~›Di+úÜ[¨À„ΧEÔk/o;Ñ(èÎMãŸÿÖšÜÙE¡ý°¢°\‰Õs4ÈJ*é´·FP+v§¨M´Ž„ÞtµmØJÏ‹Q»)‡]{WëtÂhT}dÚX÷ÊÇ*KGaspüf1 |ý4ûq¶'2Iÿæ™Yè’šž £AŠƒÐÎÃváå¬IöÈc="›Ò¹`ç,îwe ´Ë ¤dNO–ÄæàKMµtpÜÉ£É9 ãîn. ÎXâ¾aháàÎ?ú>3Å·ypˆy´4̤7?ñ¥ðª~Ƨž¥ßÌØ#AJå4hX«aÇU÷ÜÞ3¹¹¿èc€«n^µ=c§‡Úæ í0ç–ž©àòGÙJ_A»h¬÷¡ƒ´M†$)XyÝí©Ë hâò\Ò\ru«ëÌÇOùv|s,uîjM —)>»u~_G˜R8ì½­ÒL¨ÕëùÊKL¾ÕNpdmÒ;5¸ÚôaáÉ®°èæ ¡àî¿gcŠ+ë0þa¢Ö,j÷=êç3Ò$¢Öê2bQæEä ^HòñÁy¦AÃi~ïîwëÈð'«XÞõvsX!Ì—JÛ¬@R{õzLÎ0!¹4°8ñƒŸ”*´\JŠUìŸÈoM”Á»4¬œÃËL¯½ªËÛW­1x"ŸNøzŠÙÐgøG DÏIÆPf°HÙ}½úñ«Å¤Ùž›XîƒÊÒ’‹€º7±d˶-|BÍ7aàÕkl¶˲Éh)¬ÊŠö˜„ûw[Ö@»õéz¾9D³f3’ãñ.Ï>œXŸÁ&ýH‰™ûù—†q¯0Æ_í sv9j¿KÒÄÕ´û8@Ú°žx=A¸ >®å7¹ˆ~窢`ÙÃR(Û.¯ë9† ^,×'¹$`*j‘Ää"rijFæ«L¿ |%bc»ä¶ÀŽÓÙ©ÌçaL`JS¬oqáÚq¤-³çàBòã9w’"Ø9i©õ穾ãÌÔO¯^Y}ÞÆäÐ(®¿švòŠô;rŒ¥ß”Ø|—\³žF'6Aav¬dz‚ƒ¾_ÔðY»}©T’‡ù]Ò•!e¸Ò}®Gsµà6táø 툣¬Ü…¾ÏSEY†eõôxÉ$Y{M4ØZM³©ò{•£íÊ*xÌp.Ë++#Uó ¶r ƒ£úMÀ!ÃÕ­-’…´p‰žÈ Ô^XÅþM³#›—Ä1;’wÞQ‘²¨¶H·+åAè¹ê[×XÑì¸qìÚvZýɶjÃjNpúó6^ÝBãæae•·~#"Ù¬|8Ê@R=°ù˜òjÉ‹ûóþÜ¡®Í ãäÀìŒì …JCÛÖL K•/Ü'rÎð·… Ú•…c×ÐÕø蟔ñh?àò•]h繸7.>_Ö½}Û –Ì*¶ÇxVP‘ðN‰v¥'?-}[Ÿ+œç†têŠé|¦”œþ)298v€àcŽ15&ÈFô? €+˜õ%þ¢<3ˆJ“)Âö‡JµÀAE{Æ(®¡Ü À²­iá—–íÈŸê¡…®}iÇ|ä=6>´?¸§(Bž º›Ib9­T¿ÚDkÒ´T=.•êÓýJ!“Ðcçû¥ܽ¤1ÁÈ [À­rNB»;><¦a§ì”‰Saåê$062"lÉU_ÒÑߦ :Z¹Â”[ „Ž ÓغâwºB/×^ö}áÓ±yk,00Â.x!°>1‡6èuáíÂg ‰uy1¸Ëp6ÏÐÅ\iš–æ—1ÁÒªºè¬7ð¢~|p§â¥ó2[lºº§-èfq­7éIcG»€M4 ‰œ“{WµcúJ?áü"ßØÄ…Z{ó)q¥e;ªÄ)‘ ­Ú>-%Ê:ÚíIòÚ–Œ™vFÇwîÒ+›í‡ø"D/3G#Åìñ>a´N”YÍkÇÇÂv2óɆ•9Í;ù)‘jÆ€dŽ:¢¿ÜœÇ§Xo&â7ÚÉJ…̘cd4Šc'AĪÅR•!Þù «àÖš …Ôªä¹9 2pp“ Hj¥ŠP1¥o…Eüë5Á܈Ǧ&#ú#Yz„D¡ÅN}G)Åzƽ1ŒÜ†X¯ÀAÞì›Êª¤¶2.´Ÿ.´é×ñ½ûü¤žJ?Þ}ebE0@üõÁSÓ‡gü=ôŒeÌ×1×ÝžªC;¥+è‡$oÍ™º8ï¨Ç)>åÅPx`›€RôhãÛÄž[0Žìì@=ÌZð mß+®jŸã›8sKüPƒ~2ÄÿücGº Ôš¨t Ô¾Ó©ªëÕã?X× |·÷N´×ˆ|pëŸz£VÌßY¶v]/æzt2ýÄP9fáEð«Ñh€GÙÑ^]F´8ÈrÅ&à¤óýý¬È&ßÄfnd¤E5H c×ÊVz"1¾þÐBCÉë›úF‚sÄ^(ŸÏ‡¸%Í}b-XŽÒ„ëv öoœî±ìhk˜-ߎÀdÙs÷Dt›¦±ÑîÒÔ9¡J}Uø·YåGæsÈàä©1 tÀ¥ÜCš†‘µ±9QHMÐO—þ;h&àÞ=¹v0­39ñ²Žè‹æ@q¦›!aˆ.[ö§÷gy«6SÃM=]Fr|úº;ûõL‘h%ëÉUªà—Ò[Æ¿VÑ NŸŽŒZ:þœY^D.zèý„Û_‘ˆ,=Eç]+m!:­ah™ê‹\nÖ¸Öüž¶D ä’å VV« zHj2ú Ûãý3ïO\§iyIý«Rt[«ñ¬:ŠÓŽpm$[Ì£§Mø4ž#h'Ä |’—G©Q³¡¡D7 +kŒoT »×„@ŸS¸^\ëlex%s9[≜ )_ˆ®ÇË•û¬ -¡fÙÞ ¿Q9,©ƒÝ´åÞ8u)yÓPœ¬SʨåcþòäÎ÷á×°£@2É «1î°WQ†‚øB«hžßó÷Y»Hµ=[Ë,o¾¬–§¬M8eáq×ê’Óm{˜ûÂW’žŠF?åAôÚk)‚E~/aÌíwlÜš ¼Aajsb×Éó2D¶›~[cÃ…-…©¦òeñõþ˜1(‰KÅ=Ñû}icÅÒÌ1 »õ+K=2ÉRK¡4úà¬Am…Õy“ÅŠ 5í¶ƒt#¬âF24g/ÙáXÙ²>j˜° 5ë“Ç«€³…áz8Ö2oF×>ãÓÄ3c/cçTYèÓÔÚ:Onnâ³­1z€›ç‡K‡l xfU{Ñ=Y]Tö~ÊŠSD.‰H™?Çé´××½^þˆôÑ"¾©A?§8Œå5•ï¬õ”`ðh¾ùfÚô¶TQ@\Ï€çÕ•h_¡ôØEfQÆ”õ‡Õ>%§CóŽ7sÓŠcoûm} øô75„*ÚåH.‘|8I¦Ó{1ì|8…7I ž:ª=àˆƒ6ÉúG/1£!v8Ÿ{6#ǽ¾Ø×NÜÛÑ:Ùùhijh-¿]¤l"Û‚<ÕÁl(­=Ü°ññŸ!N3gúö4[*Š1}[D®û)Qõ® àÄJïsuOžÝY8¶*0VïV‹n?hðÐòäÑYGqÝAúž]€õ6¹ýÍ*~D PÎ2CÞy`¯Ë»3ëjÉí™ú\Ñ,8¯Žõ -0c5ŸEI~]½z?%*SRš\i}Ô¹üYßEÄ\†&P«o´ASK¼Å'„ç”2HN¨ Q‘"qÚÝíò™›O<¯Û óŒµ-Z&¿9õŸÓÕ¯³(©ðoÓûGõ§ endstream endobj 2108 0 obj << /Length1 1840 /Length2 11511 /Length3 0 /Length 12669 /Filter /FlateDecode >> stream xÚ·Pœé-Šk4$x îîîî—kh‡àîîA‚; 4xpw n.™™33ç¼WuouU÷¿¶ïo¯ýÕß”ï”ÕDÌÀ&@I°”…‘™ ¦  à`ffcdffE£¤TAm€ÿ‘£Qj!Ž °ï¿,Ä @cè³LÜúl¨¶È:ÙXØ,œ¼,\¼ÌÌVffžÿ‚!¼qcg@ ¶:¢QŠíÝ Kèsžÿ<¨Mi,<<\ô¸Dl©±@Áj ´}ÎhjlP›‚€P·ÿ AÍo …Úó21¹¸¸0Û:2‚!‚4ôÔ  tBœf€ß-mµÆˆF P·9þ©P›C]Œ!@À³Àd ´s|vq²3BÏÙj2ò%{ ÝŸÆòÐþ: #Ëßáþòþd÷‡³±©)ØÖÞØÎ dg0ÙJ’òŒPW(=ÀØÎì·¡±#øÙßØÙdclòlðGéÆI€ñs‡õçh ÙCA6¿{dúæù˜%ìÌÄÀ¶¶@;¨#ÚïúÄA éó¹»1ý5\k;°‹Ç9ÈÎÌüwfNöLv ' Œø_6Ï"´d@(€ƒ™›‹›t]M-™~'Pw³þ¡dù-~îÁËÃl0nè2>ÿ y8;PˆÐËãߊÿFh,,3)`´Ù¡ýýY 4ÿ?Ïrè2?ÓÀüûó÷“þ3ÃÌÀv6nÿ˜ÿ1b&eqquuº¿Zþ[)* vx0°±X9˜,ÌìÜ®ç¯ÿŽó÷ ü§û?¤ÊÆ ¿ªûWD;s0àwŠß]<ß:qþ‹Ôí à¿S(‚Ÿ PÿÃ=ffÓç/–ÿç-øÃåÿü¿£ü_ùÿ¿I:ÙØü¡§þÓàÿ£7¶Ù¸ýeñLh'èór(€ŸWÄîMµ€n´Ð ädû¿Z¨ñó’ˆØY<…‘™ýO9ÈQä 4SAM-ÿ$Ó¦ñœÃdT;‚~_<Ï^ÌÌÿ£{Þ=SëçËÅñyfªŒŸúÇxcàóªýwv¦`³ß;ÉÊÁ 0†@ŒÝО)ñŒ8,ÏËktýƒó&F;0ôÙðܳÀ Aû=h.N“üoÑŸˆ À¤ðâ0)þƒxLJ#nægzþƒØLêÿ g?íгŸÎ߈ç9ƒñ?èÙÒäôliú7âøžo„¬Ÿ“Ù¿ € ø/È `2ÿä0þŸµú|®Ãú_𹛿!çs›ßSø[ÿ¼Lvÿ‚Ï™ÁC6¶gd ´ø§RŽçpöÏœÿS-ës9öÿª‡åùÈ ÿ‚Ïêòq>Çw´1v´üÇ€õ¹×?àÍßÔ y&È‹ûLŽÿà?.h ÐhŠ676å °ª h½©!taØåGTø[¿Bê`1ioÎ-5ö4j·Œ,2ñ¦érÑñ?1Õ¥b¶JøY£ Ãåvc†“b‰¯Ö˜G…èí…mÿÈ0±o>c0¾Ÿ1Dë%´aw×Ü;ßµS=_í 6Œ WÖÒ¹{&Å&jy†i¢úSÓEƒÔ¢VÙÆ¢£­sè "LâÒ~º5££²Š"mìã b/Þ÷Þ±œÐÏcMÂbjÕ×°Úƒ3Äï3>‡Â·» šÏJ×Þé’ê«K\uY¥©„}d¢»GßÜÄÛ¤¥`.Ùpoºk/ÆÙ×\d§gJ¿Ó¼^ô5kx×àÃgÆž—¯^ˆ1ÇÁK¿qçŒKn*M»8ÃGç>öY;¢Ûý¥5àMÊäŠK¯ ±CŸ çK¤Ì]ó¨&U ðûôu¬@ȾysÿO+¹_Q“ÖM6WË”†¶ú˜©< =þ9[o¬å‡¼ oSÔÇ”Y<¾ûâàÚå„FMgIÎ{˜¨· ·š#­å”{„ÊOP…tv¢ jgòl ~Z½»>®_òř߬SÏù1>VuI›\ãE &I6Ÿ 1¡†¶ìu½Ås»?jzå Ýž×Îóo!ãò9s€¬šßyubo'ÐTq#¨©Nj'º½ÁæÏp¬cµJT2h]óoÁ¤ì|­‘Ù¸ê!8É”SVY"f ^Ùb¥,óʵ_xc¶ãa‰Ç.††là"þé§0PåÂìUü¾1ÈLgÜ`U¸£^Åþˆ¨ÆW´·Y„xOÒqù~œ¬\ÎtoÛ íBð×·_êàqZ"œ¦oCwn>säö8H©£DÆŸ|j|29TX?°ñ¾Íl$àב÷`¹y2Õ”Á¯‚ç¤6„(Ý„BÍ||ü—J‚XõMAÆkWìçO'Ä#Ô/În •–´]ÒrsÕžz?tB]pê òÙèië{Âœ9¼K y>½©Ê&ã4n­ÃzÔÙKóŠç,aõ’óg>6Ô¬Âý¼0ÌTÄÙùq(”ùÌZ4.¤ïÑæXÐ%nY{Ÿ]Z'²ªÁÕ» ÇËgS’",Ý÷u½»µ¸Šª ”¯¶ÿÕžUASO¦5ú¦"þ{ù$ûmêÅJíâ“•mº‘ÅD°Jiu *õ§¾!ß*•—¤bö&»#Â|7Ýb0<À=*I]êp´¼_·1·Ÿ„FßFB`ÂŽRZ[ /*‘±&Ì•P´–¿Ãmg;Ý2/]ÇFZ3ÅUPUr\¾ø6Q|[?ýÕ·šaE^GóC©É!»v–î0³Ó°ž Ä”uOÍ;íÞNù˜ö†›ÜþX䪫ô €B°Íº4Ñ€ÝÄ£[zP£¯ïé—c ÂbF8tÔSŸP©Ÿº¹MèÞ커XA•H 5­Á"\ÄÈÄÇ×?.»?&éž@ñ©ÝG㓼ͬ9§«Ãj1ªZ‹WµŸ´4S--åÝ‚I¢u|å/j7°ë÷«Šãd×üµ¶x=·z»7B-Yš»~² ø$·Ht¤„ÖÆ9ºxî‹Ãæ)Õ>Øɳæ2´Kºæøp Øxl“4”­§%˜ñ¥ JÔˆz ÞEà“dÈv£IÓAT®bê‰U8®ò½ÕßšÈ{1 ŽrXc¾½PqÉSzäÂãøBNÈ›8KŸ§aˆ'¯,[ß?øFQ¬b[Májt2ýr^0ñVßØuÈÖ‹­S6ÚY+r­îLñ}îÝpŠ³Iϳ¬w’a+ƒÃaM{óòñ|†Þhf*I²2 -IÝ:Ô?ÞâÙª:ìâ\çËê¾ûh>ƒ ÏCŸô™³vý!Ë]þäGú—ƒ"5zD!–h!ôó‘©Þ7j¿ yØæ1è—RøìãÃ]ÔNü϶´x÷1ê¸<1 Íp…ZJ óI3]UÒ{öÛo_K#‘¡è¹<|Ådš˜$ °ÒDôÈd =göW¯NÔ”˜àpÓ’ót÷™8ßHþé9ßE.°5Û Ô¨pŸc›ïTŸ‘ýغ­[Ó^ÿ8Ífôàœóhoñ×x¡ ©Uš~Ÿƒ£ÃÝMMÙ«*¯auÏœ> („ FãWㄲÔÛô,}1IÛƒÉn‘9À'6¤Éüz>eÄ”x¥}Y¨aÊq8"¶&îîåÀìw•éI kð~èUüDü] ~”aÞçºÍ*Nä7l—S3?“±ZŒ,/‰p|Ù=×é¿~ÇYF©‚'9—Ð-a€/{h<¼|µë_ìëÞ–´ ©@ÄØ_ (-E—‡c ‹ÙámÅ»Ì,ÏRùHfMŒ²8q2ë;ù¢Å]$z¦µÄSWã±g¯Qïý­y`À²nýҜ՗d]¸vqÿEŽ¤Â6|IxÙ°@äýöl燖ÉNlz U ªí”¸.|ydg¤ýÇ¡Ÿ?CÑ‘´ãE¸mEÜç.w2ÌÚùÈ6S?¾E+n·_i½J~Ý„/&õé•O¼â÷쪨žO>€ò1'f§ÈÊõT¿‘Ñf»/:ܘµÑ>OŸ(iï8 õ~œ7ˆn¯ÅÏÇòîo)GPØÃ[•ûtöí¹tCé@ÅË ŠÙÒ©¬Æ_S€6ÂÞ¡ŸŽ»:-™V^yhz_¡ÒbûÈ'ƒ·$©Lhâk[€/ƒ¨ `ÏÆ‹Tø2þÜJÂöý±ϯŒɦUÁ67ÌQù!ÔExóÞb()tÇ#DMDsší)T§˜ÆEu¥q’Ò‘¶04+o– ØË +ª[ t fJWëwb©¼><{;†£ _¼¬?Ê ï•}}9$%ÐcðÙòÇ>uez3•Ûd2öFØÑàÍ«Ôcl ÆÁ[¥ë ´áÇV•’¸ê‹­ß_ÒSŸø¨.TÝ9š'Ôåüгõ1µfWÓm{£o¯ øtI웜„Ó‹U8œwS=hãÈä§÷M¥ @Úô¾ú}{Z¿W‚œãÖZ foì…Ǧ**3̧ÅF_#U—Ì4k!^ÙM[#åùˆäbRÏé+lâÏba bâ^-Æø–O+%VÛ,}w º(6pçˆðÅw§xô2=Zf²V ³nù gYqDl 1B”‰´om–’°œ¢ÚzíÈÇY­ 4JÌÜæ —Ö–7VOÖW…¢b¸wT¾' íºÖà b'Ø ŒQ7Y_êàجlìøÕÆšÐ.øùPÎÕ­»põ›!´ˆ¸<}/æ»ç¸Ö,esþ°À¶D¬peKã «Á*¤¶7›šXZ° ¸;¯<¼T;ÅWyý£67ÀËV+cá´^aMdÔÝÑÈÑÄ+&+bî¡Û±I×0Þ±• .9kC2¤#C)Nâëçd›ãðeK=)#Ø+cî‚´F êå³ 6œJ‡´º©Ë“s1D²ÏOÎáM—VX0‘B)Ùw‹C"ēیŎ¼íúý„|Ò¬m5B6u5¦¸Ø›0Ê…×ï€(ù û P–¨LLý±&üA‰ñ‹Ö÷L[+H{ã àôÑœ€è‘>øÃBn)J¬PûÑÖo} §¼“‰wv_)LÕ|9Þ•Æeµ0'†­Ï ¾Kjø¸)Ž:IÎÍ÷Â;÷¬îéÜ8Û`¿JÀBìg^lï’ÐíîÁ¬;2_¤¡~|+äãóréøÊP&è/ôîó]N±.)‹¨æ”Ûo„¼·¢<—ȧýo)Òi9c/ï,!—ïüQIe|»C¯|&´z‹Í>^K=2K¤¬¢Â, ³l6Fº‚ªÂ4K^Ä7U~N(Çœ±ß±(P„uqÓ_™ÑËXtdu°=~ÊbüQ(×pwÅçqy©ÞÌî‡Ô™ý5ºH1ïd s©Ýú`µÇlÕµKÆRB ß¿(¤P?NPëôzÙr,ãw”κ„_Aþ²0!õq¿Wµ*m:ÎÔéÁ­YCÕ×z>æù¤~0”ðóˆoGP€g«¶òTŽf[/õžûòø¾­‡s[T·"”6{ø4ÍF<µz˜ Ýj=µÌž¹ÏNLYTÔ€(Y|{’Šÿâf騆^D)›º"Èñ›y5%õâ½ý¨ø‚4+b ƒ}ºÜÉǪP“ ¶wNç*{ É(Ôâü|µVU=ƒw+s2[·¬Ž¸Äy +÷"uARñ#Šjýa¦}¬S¿Zf!ØŒAK\ÓbQb¢9ká»U¸K˜q£¤3Ü ‡rjVý¥ŽTE a¦Üns-ñ¡PÑ„=KÒõ)ºAdÊÙ+Ü×XFÊÕžHä–ž„$n·åÓÓ©jéÎCó!z¬¨º³L®3ölí0ë:¹eþˆõ´pF9@ɵzy¼w_/8†ÖÞñÑíK‰Â+ë"a;ºóa>ÕØ»¬¸PV·± J¼€¹/’/»r{?ä4ÕýÉ?s¬Ed%ÚK𶻠ffãI˜b£Çgw—u È\î…Éöýü– M£êz²·ßãŠÊ1Î8oYLŽÒ".¿Ÿ=Ç<gZÿ±«O¶ðÅFPÓ‰o!¯^߀¥›ÜRÝ£$¿ÎSö÷À_»=Ï>7® Naá\¨3A:>‘Vžˆô˜ì·öç×ÝoÚ”QÚ¸_0§t܉WSd—ÓkX:Û’ ?Uµ›¸N$Û_³&_ʾˆV˹ó”E§yºÖC DÂõ¢O˜pÖç§ñJ:Ñ/H(8Nüò3®l}yOê{?VW<"Lcš:Å®ŒÂÕš[8L7;“±gUcy¶€'A"b,ô%Î(·õ (o‰ßä61#»m#ûm ŸÀËt©oownL.5ˆ=UP E¨õåæˆðJý§”pcŸ¯IopÉk€ž÷)&•ë$©è/ òÁÜãÕÌɃ!žc“ûíJ$4+ß·0%¡¨IJÙpù®5…-½ˆáó/|2ÜîxŠÇ| K®œPâÙ ñ~›«‡¾æf€A/àjÑõYñKtçJ.§J°•èü·P PÀΖ8â—H˜X"ݨ 7ñ„¹d[­O†äq&áÆOIž.4X2F«åÝ ccOã>Ç6FÙLu{2¶aªm7á¶#¼õI?&$¤@¸Î$0ÓU4tœÝ=ñü’n–«áDzÌðF…ÿâ-‹³£’O³¿:lí‚õ4G²™¶ã]’tTµ¿™µ–=×–ƒ›ê?Éš+ºÀC†sÂ/©1\¯*æcwÓÔ©‹Ê1Ü©­æÌ u™|ZÒ†–Ex ì#dT€©£ƒŠERÝKBÿ´•-w¥H’ˆJ¡É-2_G¬ Õäv£éS$ªˆËk¹|´ë6Qylj¾ç&±$˜‰ýñFaÐ@_=$õãûçJâ{!74«ÕÀïYk Þ]è)ž U¦ò˜uêXÙÇî°°à2žå¦ Ë¥è>te‹FT¯„uì 4¡ÐR ûªs}æþèCž2±Ú×a“ø•›ôuiMo[“¥ì}㧮¢)»‚aœ/7zú?Ò¤ *Å¡}.é×váa"Zó´N-¦âS0½]–ýÜ.…Þ´OÒÙ•EĬp &·´Wߣì9Û‚ePÍ¡,¤½C(>`å±lÝ=iñ”aäpê2RQοžñ£1:-IVQéoØ£þ…|ÍÏñ6±Ö È9ÀWa¥BŒÂ!Ek$4¢*©PürµÛçqzØb#jß¾Ó|~Îþ¦£쬠9öH®ç[WͭɵN´³¢[ø=~È çÞ;­ƒU´Ñ˜;‰× }2Í'œ7ô÷2ê]Á‘£nésÁÇñ°ÒIö æí'MSË,ý·6T +¹’ÝK¼Í_1• ö(ÈÂPÞ‰óÄcÁRÒÀ!ã¿ÝHWpè'3”O¾"új†M{Ç<Ã|ªÏåJ}ØÂn^¡ÌÝJKĈàÌAœå8dÂU«/¯Ä|B{Ó w”8K×(,dd˜—NFï>ŠáWàc $åöùØ bžôë}ÄW¥{MáÜ N/L·ZAytó½ÝÏÙ5ÆYð´ËŽ§dIÊ­")ÉG6¼õ´tç¨!{­ %´ÕÞÊ+•'qŒ¢&;¦5”8\J øRáÃú8$ðv}ãÝ7ðQÞaY¶¿öƒäD7ÃÍM…ΫµjOP4@µ¤OãIKÉbDÌΘ۶Œu øÈY°ˆ¬?åG’anJ•+#k'ÐÆ”'çþzüiÔm¸Žm潟ð/ÍAU* ð«FãÒ¾…‡ÍñîßvÖídÚ ×ÓR»Ð_Œx¾ p5÷Í#ý*ÜV?Ñž[y7gF|!u0sÉåŠn¥qÅ]ùBUl²C¨@$ÑU›çk䯥vr”¶/þ*Ï,ëžœõl蜩ҋÐB‹ž}¡×¨ß]€}ïÃ>Å„§š*Mb}ÅqäÖï$-½P¡gÙh•Ž,{òç1ÄÅ ”m•)¥ãÀ ¤Ré'oͺG))-Ssx(ug½‹ÛA–|?áêgä`Ó{s¿[çϽjw©:ëà‰‚åáM™\jæT“8ñ~¯G_¬»Á-’-c2‘ËÍ-j ž_Ó„N\©o¼æ/!H .Ä—ʯˆ [Ÿ}6ï0"xê øI=Ù/J an ŒäÐÑ·Æ› œ­/ìPµÒvaë(лªz¡ÿdŠeÁÏ«/ûáÚ¡†”w8›£Ä ¹ršg¬z_y×9’-Y.Snóµ”êû ’õþïðø׆ȎŸŒ©§?yãLER~ÇsÉN°ù%-h)O®9Ë; y4rèjÊš«dH}ß׳ŒU˜¸/áìJ½'p»NWÆ¢¸Â_åm5úh‰. PÙyxÿ*fU\Úﵞälöåž¡c€/æ2×`b¥zª±øå)Áþe›WÆ´Ç"˜aïí %EòO+‡ÓÒMOrÜH<·ç»…'î¤FÉ*º’Ä™)òAñá@|QnӦݢ{×¹ðc*•q¦4j Å…êŸä˜tòSνÙ*“£…¡„Òâ6}£þ¥Œ{ˆT€¡’.»hŸ\8†A±X·:ÇÙ”¡L\[…Çåkl®”–ï)è¼È9²ü£.˾=rÁêpÇjø/É#NÔŒ¤wlÞˆ+¶ÝaTY‘‚™{îPaø¦."Ñ;ÆØ0à«KȪ¬­±ÚÒÓ/xEÊfÙYwùLë¢JuTõ:¾Y®î¸ÎòðÀj¶1Èfv!Ž,¶Ü"v9xjL.9^î,f;ÁuÆ«ÄàȦÔ1ãOë‹Ù½>½ahutÉ7·“Ó“‰¥QFÒÂÕÛ4ÉfÏÃÍW7¼}$•Õ»ºú|õȯ$á¦Pé¡‘Ô6(ÊÔ®îNY1ÇßÛ•óm¨ß‚ò Íàæ8*TtlcŠð ^i¿^©®/½°¡ŠKŠŽ¾Q¤½lßï¤[B-•%Í ¹*u¸§aó³çòµ»°} ÜŠ,7Öë÷#cßÔ ¹C°f^v6Ҧ+˜W(uÍjB2\{V¢í˜§}n{ãÒÛˆï$Ô¾˜åÔs« mh'4ä]Mg².ªø(°5¤îÞŽSq~$=x3x(?zyñÒmn\^hï°Ó™÷Ò>€knkùÝ63ö‡UÿÀ$Å,Ï»~]—dã–R0Ô<•ô¬ô ojÆcIÍ*!âÝ¶É üLÉÔ,ä‘sÿ“ô‘ }Ä:h_@ U0SòM —¾`õÅ*HŒcßV‹*Èu>Ââ[ÈÐp“…ªDb¿¶ •ææy®ù]Ûp‹øÓ»¦ñ± üû´‡ÓxZ,Ö) {RÇèd"jÊÜÍRZ˜úü?¹@«–C³1&´y™¬íÐ4Îö|Ô–ÚÒï'þgŸGxCó0‡Üyq†£M {ô^•Žåa¼m•Dÿ4ÓÈÍXáÒô%e¡¤}¿FLéÜrmò¹Ø´¼ØÑOúfw£)è·î³"|IÿM¨””p5xÅ áKae†QŒ6nY$JËáíÝ~³ÝSN~1/õ6lÿ2×C­³}Ösýõ$™ÙÊ®ŽÉrµÂL]¦`Ÿ$ ²å_rÞŠ|³!UÕD>kÁF¸‡S -NÓPáî,w: Ïodù9êmHzÃÆ]ÂÎ"!JÞC¾«‚6¥Á¼þ§xd5ÒÊÝlÊêµ|¾±Ü×(»Œ.ëõ){2:ƒ‚t-Ñr[o&¢o‹|­d‘Ę™Éa‹äì+•Œðì(üwõ^ËG>¿Á5‹r½ýY§•é"ZÊŸ>–uáQB<âÑÈþì¼Zâdé¬h¡þ3ö†¿7¦«ç×ån!OÏ‘žK}Dyáž•êvå^Žó>þrŸµŠvYÝ„ì#2Ê+À.:YFSšÎLÌâqãõ)ÉÁÕÜ°¾üæ,4Ú*'faøÕÙ@ø÷b œÓ ?µ†d>ïÎY9®_S„d[¢7Í*ö´³¼3òËß…m“ù4í½Fuî´i>÷)¿Â&žŽß\¦AËíüº×Ðé[/Qtú§SÊ…ƒªT8’hIWEqWw—ãÅÌ©§KÕ1^ûË&·‚É.Àô¦bñ6^M“Ã%ÏZ²~×¾x]Žø¸ÅÍÚ>ÒÁ×1²"¯ÊÅBá±l5ðõ½ª9GHüê€]º^bÐB_TC*7…øµÚÍÖtpúŒ¤º©iñ+sk ‘iCúmFPLO¾ïõ6;Ö=X?å ¬›§ æ 1¿¼ôû÷f’HåB F©¨Î†ïf_œžI#1åÚ¸¹aµ=Ñá.U…r]šÒ¸ÝUl¡rÃðƒ­”§Ý7æ‘dCIç§U¾.âñ«&YZçOÓ•`yªô9(nPë|®ê'!Á_-hä/E—ô%X‚B—.x4Ž<3 ò£®«qBK]êochÍø*¬´D¾ªo‘ éï/Þ¶t(·£{¸DáM¬LË|ôh@¾Hþ¹ãº*-¡ *ïµÈl3Ï´F+‚û½Rûà:úWa<ß‘]t}Þýêib¿…^c™[:¯=þŠŒ‘ßP×2‰aòtÙQ™‚{•dÖI*× ÄÆs® ]Üáý˜+£ ›†ƒJ8t²iÑ°Íå|1š‡"â'ö ›Š+atKs´H‘nn†íð³ •z9šhˆKi<ßô#Ùá ž)…ðÇMç¬(O±B{ÛC3Ò¥l|\õÛFô?ò1.ÞêƒX0®í9’ , ÞWwéÃP¢e7½$AÐ::á*ÕãBy!¼ Y0x€ ÀϱÔÿz¯në˜;HgÖTˆ!Þ±·ÑéYû¦˜uWSjFé;Øcᨱí9“_9XiYûùžÏG<ÉÜj>ÇH;TÞ¥«QØÆoÿ*бyI2Ò -éÖ †¿jZpkiÐõV,‡„µñÕûMEO|Í$†JÖ“!¯`*«ùß·Ñt+KZW¹D°P—6p5bÖICXúß¼Îtñ±·ZÍbêFa2i¥È)”‘݉üÀšMjEÓ ‚,”‡õv4UÖdȤ›ï¤?5Ä)!ÕžÞËÂ,::‰õè9_wÊ÷ O~+DÇu4²€Åщ=î—¯H0È%WVã¦rZŸ~ÀˆLÔ÷¹±…ÕFî{€=Ù‚H…$Ö>ŽÎìh'µÂQÎñÿÄŸ úÓþ–ŸæÈñS+ÆˬAd¸"¹%úùžÐ\Å„C„ÌuHå²Ð\»;&0VúdÂTu«(ÝX"ßË´ã´‘ë$Ú §GQ_ë÷c;ì”Å_†OW Ú¸Ðêxª"«˜%î„qV½SÔë çÊcñÈ“ÒÈ*àËÖ¥_Ežj£×ëE’ž}Ž§10éì »“üŠ3'TÓ:¦ ¶5‰Òðà$P¿~ÿ†ö§~«©=[ö»ÏŽ:ÑÞ¨â[@¹ê Dª‚H×klÞöÁÅéZ"Â7Í’‰Sú~Ù3ÍkÂeš¡fB˜ç庼ã2º2õ¾Ëõt:_¬tò¨Á 3wÄu*Qï\% o`„o'Üìfú÷°c' ]¶±¿1tðXÊ‘ÐÁ—5c'äl™j‹9¸2<"vQ»¦ÍAÀg’mö>ÅL#vy_c&Ó=’êž6`ß“èàh#n6 Ç=Ìdø*ïm©g(SL°+ñfx÷c 0¼T)ßåÝ}¯:xÑÿäp’â˪\¦øCöÀƤjðñ˜®DúÑù2œ£™{î$uH`õe4IY‹î8CÇl …uÓYñ⬂°óh‡#9bB«†ß²í "3rÍökE³ý¾ƒº÷¾.=‚4EŸ¥y’Ç|vï¥ûu… •‡ Ÿ¾›õ0»Ã Ù¸D;Ç€“lœ%?P3áA…U̶…ðøÕoô Ñéæ+É·ýàý›ÚX×mݼóôØŽS&§×Ü=K³de¨ÅÓç­° MNúƒö¾þŸóõIÈìœw·ÚQï«3”öƒõ .<…Öå”M”eõ°„y •ær–Ö=TBªõôa(<â.$}¤Xµ²¾7ƒn0™bÇ`ÙDùzscæ½Gªb¶Kúæä^:ªµ¾1bt,ñûªî´¾U…³¾—ž2»Ø¤E ±¸ÂáŽ:gˆǤÿdË¡{ÃR¹êŽ‘4H¡ï²ƒ0?Kyáqô  ¢KºzSà:Ë"F¯yžWÿ³5Æ’@ë;U<ÛÆl‘c¿6² ëG5³$ÏkR.±¦(R[´LÉ"UGÈ¿ØløÝF²è.¤Wý‡Ýd'š(=EJRÝwÍFŸ¹w:§Â,h‘E)©ÐBéRm˲8“cR±Nõb䚣щ»þ·TýKŒq¡n–ßp†ÝôÞóö¡ûKWûœ1ææc]ªêX;PS—ÉDòº=€„ S—‡Ö !5å¨!)ùзB"›à-¸´Â)Õ0¹|VJjHƒ#ï‰Ä À¦Ÿ¼¾³¶p rÃù)n€ó}4ëkj3CÃŒ~ûI}Žôç«Û“µæx3°8„°|^œ¡Møµ”žRØ‹†à ¦ÑšGâ´ƒŠ’‘.2ê_} â‡Wj6Ꭽ›J94Ôyˆ¦¬€¯Q)Fâ®ö0¾£û~oÍK[¼°X´9,*›WÑÙsçHÀ·ÃÇ rêË~Ã= £¡ /Ù¤CíÌ5>•=nø ""À7ö²ã@ÈkIöa%3ÏŒñåO½öUÇøì¥fª¥*|Ì&ž-Þ,z=öïfÇ áT5²ÎŸ¥ Ï£QEžR jF´¯Ät9H Éí²ßÚ9a?ZV»°:×]ÕÒѵÝØëþ¤Lr¹o7 ÈH¸NòvKpÖÑÖ¦a¿+”ï4Ú¯çT£ö$°æ—GRö©N‘”‡YalRS…½RƒP”ˆ¡Yø ¶W™¬£Y±f?>cȹ÷[Þ­ Ÿÿ·ßÙ,HÞpªþJ¤|ÿªÝ›'Öjœ›(ûƒÏE;ØÒFŸLÐÇÀ<žãS¬ ùöCv2ìôøÖÍ\z¿š(¨Ô纓PÛ¯ s×B»bnt Xµ–x²#ÚV$ÓB»À2,mÊo¾pùèœ6*[°Òá”Ê¡öL∰–¸çŒ\Òô—ƒa#¢}O©*k¥£+/dk×^!æà÷°5Ó³? Irš!XJ©qV ç·êGЇÊKF¸ÝwÇØ_x³Õ`¨Í#-ïÐøÚÔ~½1 þZÒ‡þ8àU~ë»ð=f঵ÉúÓClxÁó;BOɃ\5‰Ç—D ‘峡s®Ê¾äÕxu²ï”ƒuR‘_À&K¨…»}·×ieÕ莴PükÆø X&ü![ö”Mnƒ÷x¡Ç¯:¦ÿ±J±ªÕãúSëI!FêÆùRHÑÁk¦.–%Q%Ë ØòãPíeÿPšå'ÊöáãŠWi—»æ1ý™ñ˜•è#Y[©†ª¨´Õéß©ôW5”O¶?‹ý´°ÄËÚ¼žËòÄÂØž,ðXÀ¥*>njˆ R9ÑŽ?Q¸Û¶þì&ÇEɬªEª\Ü{Ÿoî…®Šýim¼UDÅvÕžó}G‘¢ö§´é| @ö#šì"m»E˜7cæÇ_&äNUص¹!¦#¦O/Ãø&N‘\_2Jÿ0¬@ñ´}íni·ÑãëVˆ$Îè”9]sev¿rÌ ˜Þˆüv »mõ݇~a¤wÛõ¥È—< FËr]nºEy:ÙNÁ+ˆËºžÑJPRŸèOÖ=ZuãÍî0Q¡sÍÉ6Ÿ¯ ŒnØ—iõ—>øÚÝbÍüºÛ(S5þÆ–Ù!h¶$ ôXSš^p5‡A¡[-È€jÎncãÍÐìžÉ´N,åŠ*½ö¨"èëI&—tEO³ôÑö…¸Åuw´öÅèù7îZÉ“x.9>e 8¾§7軘|øLº1¯¸ öºûm\\tC”÷ø%¿u‘XÆJX±Ê]ú3K—¨,hбx"AÒÊê\W¨©»­»qÕåŽWÝDºÆòÓYׂÎ!±i…Ô?'ü"à™|Ž0âîöÖ{ÈÖåʚךB×øø­qê²J^ÈâÍllj«ëËŸ¼I•5Ôr±õûöY¬µrøÀ6¾ˆ¹0mÿʽ­K/E+ìÇŽDÇL[µ$¬:|xÝ3Bˆû1/ÃdM”Dù!N¸Ë®Á¾žØ×q›ÝÁMmËéé}Ïöÿg5µÉ endstream endobj 2110 0 obj << /Length1 1420 /Length2 5888 /Length3 0 /Length 6852 /Filter /FlateDecode >> stream xÚuT“kÓ-("DAŠJ‡Ð”¡7iR¤wAš„$$AH ¤7Þ{•"M8€4AŠ(H‘^Dzïˆáå|ç?ß½kÝ»²VòÎÌž™g?³ç '«ž!¿"mSE£pü !ià}mmu0PHHD@HHÀÉi„ÄÙÃþö8Â0X$%ý¿÷100ïSãð@m4 ¨ál‰AâÒ i!! °Ôß@4F¨ vABÚ@ 4 †pÞG;ºcpßçïG „’’’àû•Tt€a0 ¨ Æ!`øŽ°=Ð AÂpîÿ*Á%‹Àá¥]]]ÀX4.ÇÍtEâ@†qA?)uÀ°?Ôœ@#û;`ˆ¶Å¹‚10 Þa„ÀPX|Š3 Ãñ݆êZ@]Gê7Xë7€øçr€ ÐÊýÉþY‰ú• †@ÐŽ`”;Ú"ía@]U-œŽFAÁöX4>ìFÚƒmð€_GUõ`<Ã?ü° Ò‡À"írüYÍ*(è}´ƒ …Ã~žO‰Að÷î.øg¸OPhW”çß–-µýIêì(hŒB:9ÃÔ•ÿ`ð.À?>8 ’”‘Âœ€07Bðg#wGد è§ÏÁÛÓí´ÅÓ€y#maø€'ìâ0Î0oÏÿø·€P$´Á‘(À?Õñn˜ío? Ò h.„—(ôóóŸ'K¼Â h”½û?ð_#TÓSÔ12åýCù?A%%´Г_ä‚„„Åøï×ùÏ üÍþ—WŒüs:¡*ª£lÑø:¿Yà¯ïo&.¤Áõgo¸ÿn¡ƒÆ äúGÿBbBüèÿ{ ~¥üßÄÿ³ÊÿSÿÿ}"Ug{û_q®ß€ÿ#v@Ú»ÿAàíŒÃ/‡6¿"¨ÿ†šÀ~o´6 Štvøï¨:Œ_E/t~¨€èo?«ŠtƒAõ8â·˜þž¾‡=ÓCc‘?_<ø,!¡ÿŠáwòÿrÁâgö;Æâ÷k¼?m~Õþ} ý¹“Âbâ@0và%·Ä€ž üòBan¿4@¡qø ž³7Ðü´˜$PÐ?4ô§ð¯Úg ßü—(ðÿ¶-? æƒÆGЙ@»ªÀÆïŠ ®ü‹}²W¶“¾› ó÷åY‘à:T1LO›Ð,VRµ²kÑQrúž13ºï¹PÍòÚ]ô?‹ê œÅæÅÈÅ7ÂϱžGŒl#7ê²M”˜¥ 1zLá­¤òPx‡!ç ¯»Uíc±®wn©Jh5ê4·V¥iÝf”z¸1g€«·é6Z; ½›9ç0ÿÜ",Q'…MS¶ùãZN1ÛÕÓ=EùŽém¸Ÿ†‹ÿâG0ŽRE¦›QYôjùéò³ÌB—õ*O5êû· µ³Y hÝ÷sâê·i-ÎدšÚ¥…!bë´7šrR–óhßÄc±¶ÉÄÌÏl!ðgžPÙiUßæúÑžAš£ñ+³= ö×Zûßp¥±{!J‹ö|ž~P4 ·Kz1ã0É‚Pg¿£âq…];ªW%ʺetäí%5I#Çce8ZÿªhgXY46Uoë·<>[TW_€©ìŸt%àÆ‹ NϤûJ$Þ`¥â®ÆÅÇÑÌmEy ZÕÊÇF!H3Ò®‘ `éB¿/­íI8¹à(®+¹/YX6©Yì'³]¸u-¨Ê£—ø´¹¿½7Ò|ý{²d•Ñ¡b¨¨mù’u}F^ï ]mæÞ¤f¸ù=ŽQR­†‚;k§ÃðäáÅÀÀ;â·îÈíל(s¥zÇH2=‚®tóe›œÅ+asþb›hGKU¨>üXÏz|«ß´Ös‰Õ îUær0òôot‰¡,ûXÙ‘¥HºT3™îJp)¥+3YÐæu„-Ì–Ñü!^båJO"¥Rƒ©Ü0^ï'÷‰(¿Ôd¸‘X ©"Øúåæê6•z¢LÔhf,A6×òüÐKüÕ¢Ê1¨s¾Ë’iœåîÉ7}gä+<†™ 8(ÛdMïǬMìõò0:&Íjà„92º%õbŽþ<®ð¶.¶ô`ª¢ îsúó¢‰Ðä ¢â«Ff/÷cæWæ2O›ÏâïL ¡µr.Ã% %÷èÆt¤Û7Ùwd¾œÝ ½:6JÖú>×–I.E¬éà×Kêª6O·aô™¤¡(…iGeK˜¡‡¾ŒwJ-cThvk0íz 3a‘(.÷zY™ÔÛbyßIuE´ôþ#ÊK—-ìšÃ§‡+=©UÊ×'Øeœô¬’0“íæ¦LI5ŽµøMä?Ù}óh‡ªŒ÷2 qµñŽËvzjªÊKŠ X‰€òpŠ+NÖ?kEsr—™© Ò¤Î>«¸ûnÝîZ=©÷I –>ÕamÒ÷t˜kÃ&*™ `ªcDü+Yâ׆ÐÔr$qãðO´ð'‚OHtšÅè™/õš8&áK¯ú. ¸^ðŠÛÔm·1vTìúô¾Ïè5/åÏYkyÑ_ØôI“µ¢ ZÞ–døŤ¯úÌéB1¸ªO$$–šFy“ä²pq¦ü|GÏoWÌ‘ XZfæ}Xs¿¹öÝ`g\Æ©ö¹~„㙄ËÉðÚ󦦚s=O÷*µºšÛjin½øXl²ïæÆîå;DÕa×<&srëR×£D!uÆy©ûíNt×ÁdºAZÛ†Ž¹‚t¶ð² -½Ï ›—ÄJà²æ-ËÄž)FDÿ…–œÒE§Ò©²8<}ÊôÉ›”UD>OƒR7Ônlx*:ë]ºL¦j®˜Š&ɇàq_Upqì ÷‹Ô­›KGO-©kä®èî,Œ©Š®•pÖïÿBt¥ _PrdQ[dØŽèj†€'”¬½—€ûÿØí|¼±TëBz°Pm´‹H…£².j6rú˜8§Ÿè ùƒžP€Ä÷>[«¬ìOÊ¥•·N¸¿‘¸n„¹|CäÆ-ë ‹Úü¨‘žÎʃ ò ïK=ðNsW¾³IŸ{¦ŽææûZ¡>ŸáyBúHAXŽ¥ÔOÙ5$õšåCåiô!ä‘äã{Ñêwy¦gà1ߦÍê-ê?âVß3ªv²vôÌ'ÑÐMŽ‹kHù TйÇ;бûMjàä=†Ý{'ås`™{kç#µRls߯‡°W™s7ð`1m“øBËpXäÁ'z{±4 ÓDt»BÚî¥^Œ›Yz”㧘c¥¹¿*ZZHïc :½¿˜­†Ah ³Ã(fŒ“vOGnP%¿$Vµÿ*xàpý¦™Ø$IM g{_Aì-ÙƒÙu;µ„ -M­ªc—ƒÉGÞ«}ËTOUèzG=&Ý÷Ï Ùw¾0äße4Êò¥C\J9.t ØvZŠØ´f©Ph7N:] ¢h æŠ1 07sÊ÷š×’.芩/K¯»ëYh§®ù놮[ Y(\ü^ª3)ñrîɸ¾B眂‘&C…}rÃRЇ‘—ñg î‡èj8¡ôN™Ê6Bh¤bV}ãÙ&ߺÄ{ÖÜÇ Ü¯„Æí'\ØŒ…#¯­IÌžÛPJ?Ýçæ׆¶aN–_Þ ÷.XÝÕ.æ)S»ÓhñÞ~†ÚËç!Ë´øBŠËÞbÕŒ=¤ L_ÎËä‹æl.ÞA¾Ö`äŽúqΗ­ kRP6t–è6ûTü| “õ–%.Œy»¡}D_¡²1”¿ÒÂÚž0Wz)hžÃ¦àÛkÀ³¡žù[}é ú&ÁyA2Eº®J±¨¢ÀöŒriQU«v•GcÔŸ±ƒH :«…Èaã×Ô:uFñ­°—;ë Ïgxqeùöx ÄD´†P¾NŸnZeÇÇË„œ ‚ÞÇ‘¨æ<9²‹^´Ú¡Ð¡Ù[–0Ûž¸Î#Fâ„Ó¯pˆêA¶*NÕÝæÒˆo#Î"4xP€+kŠ†–ñˆÀ¦Á‘Ç%³i…k±‘b®ÇTñ÷=Þ”ú,I³8«N •ìƒ·0ýÝf录:-ÈBQ½Ÿ‰ãïb «ņÑ!¿“Œ”ê:¹$ÒM,n¹ Ò:DÃSYgëþù¥R~^`‰âê¢ä Es¶¸]\ —nþn"1o¾ÿÐ-"‰"÷¡ÜÈŸ—(±÷ x³s•ø˜‹)øõ1%úÒÕ=Ï¿(¾á  bþÅË/:ö=µa [¢FìÑCo#I#͹hÚ ìW½¨êk·n!ÜXÏÊž¹=•ÿÁ¦§vS7@ÅËÝã…f¿RÆi‘þ&Gb¤áجgÁÆ^‡TðÊâÙ°B9qõº¬#î‹U=»“ÉŒõmbƒ0¥zn%ö1Ɖþ·äè‘Ð{V?¾0¥ï‚AdÛ7 Ô–üÓ’J)ôïn‘O?Ó DB±êµIäp›Wô/{Ül)åvp‚ôßÄeÂÞ„e;ó÷r]°¿Pø|›>8#4(ã­võ#†ÎÜÁ®x@ ÓåáNõ7K‡{†åÌëqæ>\’óOî$5(ó#mÒüÎH9ëLt’ÓêÇË(Î’–ŠˆÐsŠh2Føš$d®”VÙ•øxãÊð}I‹„ c ZabyM›W…pº¬ªÓ³!˜ˆô[d`‘¸ÞIJøºUÙΪºoá2aÑÆÇÊ©Ûû«^$Ç `ìÍ¢kƒO±meâ›LºѶ¶›öû- 1ÍêÚ™n‘í¹¦b²ÖþM@nõK¹6Ù+!nëã&ü£!ë윥š«etå%¤…W‘w]¶ º•<=ßê9~%ÕVzÒNvbܼ+»ã’ómû²ûÉîydƆÇÈË«wø™y“&9o|VGrgmN²å°Ñ&ƒ¥!³ìZ±&Á$Ìþû‘ê{‚凮Û)oîXMˆ§žtÒx‡†KécdÇ-ÔTž‡˜®w<º´×øtM%ývš©W§²,¸l½bô AïªÁÌRC8³(µN›WÃæ CÓºDÿ,|XÎa¸`k|¥¿øõŽ@2ùp3Aʘoè›go›®GÒû——Z«ÔöJó9‡›ø™eDõ׌û¯K6ÎSñëYiï¯Â.ÞV–ù:´—‚ ®t.^Æ üà?bdgµmÚ'¦ˆ˜%ø6Z´Ùxyr»X`þéIˆõ ¿=ås:.K;¢i§|Êö ú ”Ÿ ­Âû Kò¶­1CËQ»ñ”œ¼TéYå³7SbBÆUBaz’;ES˜M ¬‰÷W®IftUó& OÌÖv¯Mò™Öh/\/«ofñ¥/ðæáµ)Où } «bßtù\zA/¶ëù†JÖ`ÜĪa¡‰8öè À•™Ðûö˜1ðy[-,UL%:Tñƒµ^Ê¿é“ Â“à„JÇQî¤%®¹§LîlC‹ùcôi;^Øæ¶Üüb>“×Í &qWÌ"þRC‰°ç@SzÖ¶Gæ.èYU„O˜Zð*ÝÍB¹v²y+ÈòûŽÓ"}øÖŠ_¹¿VLØé‚r+N©å¶fC6†š¼·–LZ´ù'[4VS<Ë-?æ»fæ Èú–$Ω@.SŽWœ (-oˆ‰Ä$X],½ePÍ^˜lÄ›¯¨ú±‡*æ¯*]Ás7&žæaÚm°¿Qä³ò"üL|Ãþ ~‘ª ~i²s¢Ö^×­¬ðœÎ䂤PÞœiï>Ú‘¿7cÑNY’þÒÅûè~á×F†ÒIâZÐYÕ˜ó¨' Ÿf²5Øi·v,ž/zÍ÷ñÖ¼/˾W›>+€¶…$¬¨hgð¥äÑ0ÓrûmJA=\çD'.“À³gÇ€ R¶<Ý?.ÞRW¸èÔ|oë–°5 „Kž—J8ç$8\>ë )]•”©ù’ïd6' ½¢|} FâU³»"ÓA¿µš®¦MÐÖ·Iªýê¡T(&*KÝöÂÊ%fôÉ\ËO¦«ç›d´‘m y ¿@ô’ ÿGðX«¾ô›Dð! û®½ƒàÁµ·Q­÷cÉ>IEËW‹PóÁ¢“LâsŸ®•k®cì-Ò¾äêˆ7×Y[+n>îºuyí™6n '”ö S‘ùuƒÄ,gþ©I§].#u/}eí¾ý•Òí©_¨¡ÿ.èÒd­@$šÑÑäô·Áô¡Z¯?÷ü¤•l8ù’Ãïý¨ÕÅ”È<,%¿Tw[+¼ª²ùÛÌÿÝq endstream endobj 2112 0 obj << /Length1 1422 /Length2 6382 /Length3 0 /Length 7344 /Filter /FlateDecode >> stream xÚuTj×.Ý) H ]Cww7HÊŒÀ 1tJwJHHƒ’¢ÒHJ‡¤t‡ƒÔ?ê9ßùÏwïZ÷®YkæݽŸw?ûf]n9;˜ X…sóñÅ ZZjB P€äÇcf6„ÀÁ©ñ˜ÁîTü9(¸ƒAp¤NGúiÁ uOgŸ€OXœODðb;ÂÜÅŠ /ˆ@‹ ƒ‚=ð˜`®¾îG8²ÌßG›-;€OLL„ëw8@Îì±AZ ¸#ØYÑä 0€ÙBÀpߥ`“t„Ã]Åyy½½½y@.<0wiv.€7îÐ{€Ý½Àv€_€Ú ðdÿ9Y"éeƒ:ûþãþ{¾¼††fŠª*œÿÇ&/óø ˆ¸ù…€Hº ŠD‡Àgùþ¿±ÿÖê‚ õü'¡ÔàþÁ€¼¼¿qxýE ¶¿V†ðïÚ0$—Á¶¨oÚ"¿øþ¿àwÈÿ÷¿²ü¿¨ÿß ){:;ÿ6³ý¶ÿf ÄÙ÷/$•=áȵЂ!—úß®&à?«¬¶ƒxºü·U B®‡ÔIqn>A à=ÄCâ¶Ó…ÀmÿéïY k8C `]˜ä׋ƒŒÿˆÜ:['ä«âœØȹ‚ðßÃý%ƒ‘Köï>” ¶0»_ÛÈ/$ ¹»ƒ|ñ„@JB>äÚÚ}~óÀË…Á‘!$æ@€=ÌïטõÀk÷K÷GäðzüÿUÆÖÓÝÙÇov {ø[þý€Á>`[¼¹i˜­DøÓºð÷ˆZ9joîIÌ£,Äc~î‘+xÒø“ÕTƒ¼Üy×ÊsÝ|ÊVO?jË»!^.Í|÷_¯§oð<ç¦WÞv ·Iž¾;GLó¿ aœ&nF)4ɧ/wïAÑ¥#ù„+cçÐcÀLÀZ×5›æÍrÿ•²ˆzÄ{íŸj_åj’ÓˆïÓ‡·Øôný dÍÿæ’¶o“©Í¨!ùá/Õ»†d¥o ´›¶#WèÆñ£H8©’D?¢ vñBÑu~¹×^ö=2Ù=»žÊeö Dö©ãúéž%‡[÷Ë&ÊnœýRM•Ô ìÂÄTÙá”Úï)ÓUQÐ¥û>|#e C cãW‚¤›ZhAz¾ø_Çã7.:&= Ï+gNÅ_/N.aÅ„‡'6…ê¨Ü²¹ ’&a¿¨%PRS|D›é‰;ÞßÙzƒã>_9¹æ|l"µfò´G*µXðÄr~o7Á&˜¿üyi/âKleW×;¹twjöÏ+"WWã®& :+o€~ß ÞuzØu¸‹ÓúÏÙÂLŽKdøj×⢣!ÕÚ'$±.g±ã„›ºy¦r¯ÅÔ,у3Ro’)Ÿ`¢3´]áŸ&ð°àWvÕ‚÷)³zD»î?Ä.þÑâU R»1ˆ©¤wôAGkY~ËKEÔúôýøÇõÌò¡¸Ss¼UçzOÒ œ²ÔO)Nù[—æ,$h‡#n,À=ÙO¦?”FjUÐNûTât¦ORÞ ÖÉ D¬9H…›Q(¿ ÜÍ{–7šPŸ4 Å:ÂÄÇz)íÕq§œ`ƒÛÐ6л–@r„œ©Úäšü¾Ö•íz–÷ˆWËŠ¥¶}zò*iøÜpú½|Nbü±8ù¥gªÑ™šþú¤IMA<Ò‚žxsöæ÷䵚 ÎØtµó?r–©åˆndk#4'§ç¾XIô·_ øbÈm)EZ‡ªx±båÔ ßQúœµ™9°ìpT¨3§O°mðÆÑ ?ÆõòOÜbßÝõ‰úaÄSo»– }blRÚ¢ü¶}͆kd³d£ô‚±¶1jG©3úžŸš<šüiêãpeæ÷<šT¶;~&¥÷À$ì7̸¾qXVÓYpS|©q"c?RBåäាN2E]DW¸Rµ¦h ]»æ 9›’=àuB·Hêý[KSEw•kJM9Ù-LtÛîm BÜÇ_@¸b:Ý]Úš.¥BLœ'¦ì9ÒÞÎWΑû?ÂK¿4/³;$>{‡;Ž)hÇæx«ÞÎàvº~GZ¬[Rhýîbk'Í*œ@A õŠðÛ|çðhà%Hè5=÷IÓR·Dë õ£Òd…ùž®Ÿ‡Å³ž&–n<ê!¨eÌÆæ«( Ѩýo™óÈ -íQ@xìå¡$k·¥ãø|Pœò÷3ÊßAŠÖµ’Âïâ¥Fcë?®šSFPkxa§ Œ)àGΆÝÚÆF4‘¿ ¬þy Î…jÿög¥àMhãýÈW¢5³èVì÷Q:°g׋L<Ò_>ÓIÞ×{Ù]Ö°ÛD9k2ïeV0“ˆ9Þ-;iŽòÁŠà*tÁ,zÃî‰íà*oÐG˜>›â–‰!¾OqPýB§êÇ´žu~ óê  KlË‹U¥©/V^ªmè":6f¦NK”uüƒRz¦C‰Ì³'®\ÚÄ¥=+¹Þ¾Çb$O¸ò#§ ™Î_3õ›ˆ Îo7ë½N®GXWZ=¦ß¹w†µUÌÜqßž\Z§¼§ `~K@ýà¦+¯8†C„9aÔUÌ)ÚÛƒÚ\ÎðñOmÊR‰E–~§Rà(;íFí«9´ÅR!\µÊÙNºÝÞ„íµ(«–"ŠÜî°¾Þ•>aj6YA5#x‘fb=±q,†£)‹A6ìsŒ¹Jd*Ž)ÊÙP‰ùL½è<¸ Ì÷€1)gÓN4¸r¡ÔüF¤; וî­ÂháYˆÉ…6»çîùË ²ÌèÇ6DÚ­1VˆŠ÷!¹!I©õmõUÓ®‹ 6L< ßFôŸ¨æ°M¡(Œž áòBq°7­ŸÙ¾/ÝžKê=”qã³hâ¿Ò¹ÓøáßE¡Û j-å÷5{‡·‚j¯mŸpØ>‹b4ÒÀ|€MjþhØ .'/G~·7¥á,a¿Ìjdoˆ ,º«e©ä÷ኅ—5«ꈺ.êù* 'oªv´Å:‚6d¯/g̳4tqß y|`w•ï”]´‘#uÀwˆ†m^Åô6ô'é-O‘"è¸RqíÜ]ÔˆÆüÔØí6`ôÜS‘´ñZ€îÌ%`z\åç:O8¬3UìÂF0Ýú­w–†Sz‡Tø'ªqÊÜ}…¶·’ÜÖ™E™ÃuáòˆÜˆòrø¼‰¶îCƒóô ri(Õ“”Ê”·Y-u4 ß~ò6Ôl]¶q‹Þ³ž(Âü¨'^¸M4_äßs`Æ h~Œßy™Yo<§~""Ü;Q ’nü1œQqÛènO;at `'cyíñVŠüHà Ån_U}>/i~?ºŠng"¯B®Š‘dé'ªʘç\³f\І8BåƒÅô·üzI[SÞhø¡‰0HZr%˜~ÆNgBW¼–Ô-âœÔ0ªÃPIÕÍ×=ë¤7˜»Ü9¶}óa¢òÅ7“€Îö²kpzpž¯ûÁWzU©å.׃"oSöšcä]/(8%þ°ÉFs–ÎÞœÇPÕnŒXûØÍ(Ù3mh ðU ݇)5ŠîîU†-U¡÷sN[¼¢äÈwÒ•š\¬µR½„(Ç°¼¡4l‹ˆæÛû©Ó¬Ð’$h4hjrºRpf#¬ôõ]±fpi ”uÓ‚XúÕ¦¿úzêV÷^˜Ý­Á™\ùHR˜&—?n½Aé¦D«ÿ÷¦ªˆ]Wüñ¨Îv™~6þÚ6åôë´à³#¬ëÙÀ—­´Î·‡$›JèúÞ,ž0áS˜®šÕuLe$ê÷㌇ìäI…tÅ}œ³F-b¹=ð =jh –×0ò‹òð-40ô!&*À@i©ÏË¢q2â¸aÆ«¦ü™½9{_ræalxè œóˆMKª`°@ ®™¬›D{ʆùTâ±]û0<·SK}"?~fºŸ,ÓØO½4N>ÒûxÕºZÕ¶8-6†ú]î´6%jê’ºÆÍØFÚ½ˆK+Éäœ8ûo€–“ÐrÖ|’ \É<ƒ)ùc·!<3„à-¥G’å4‡‘ÇvøêÝ8ç,Ih,]Ip(e‡Oæ`/nj`ñgOIÁÌž‰M™•–)æ¥gæ,eQÏŒ?vÜlŒ¾”˜NgS$ ŸF˜éS¦^¹geWî­ppÝF—ûuæ[&,¼þîË’t Í¿(SP´ùÐTøA£T5YF`ôîÓë Ö%QÚÛ§š…>t»ý®š Î I,^´±Áê€e?"&͸í 6?˜ÔºèÏ)þŠhh¢s³åDš]%ú¾&á¸e‡ïÐÃêoükQÛ&}aü³‚)é«ÄtþQM.ÅuLIc«2ß’Œž¦–ëö1ä¤|ËÌ°¡ Jˆðߟ™Ve–å""Êq粊`?VÝð²›ë®Ú¡S/%az.®fWil2M"nˆy49:¶WþHú®d†ƒ¯ƒE£µá!Z7ým·EGëöãZB~™AZVÅcÐOòUPV”\ñ÷ÆÚ­ ¸:ùhÅ,ñ•ëºcöå¶glj†Æê4‘M`Åž³ªå•™Ç_#Íh®/oã§.»ê^Ä+'Mq#­{Ó̤éªûÜ´%€dn}‹µâÈ™’8UZ'ŠzÐh¢]Îh f—ý~‘ªõ·øBV††ïr˜ f ›'Œ;uìæÞY‘w ’?Ó¡‡xJ|ŠŒŠž:ÃO3øÂè?²úAøE ã'Ïœi§uEyÝÇù¹[ö³-²ÞÍéè;Ìõ(ƒûlã êì8LW~"Ù­ÕCX¾ji¬lÛ9‡´k59Ç®-Ÿ€‡?ãñŒúEã_ób>H±ÈÚ .HoËʼ_»PSaãÍCô±Ïq3¢j™m¶´³ñwëõ9ïUò¡Ž‹þÂå¸cÊ]À¾[x'Iš³¥!83“ÃmÔB-9~vþ¡ÆUÌý™µ™ÉŽÄÓw³@é¼ ?·fœæöÓjeßGK?uEÄ=`Û·ã"†oq¦ˆÛWô±(fæsùj£ RC#œ¸Öçü¬4n&íÇæðü)›k6GÐSý±F/ÃeŸ¦pñ—O¿í‘'E©at¦˜ghpkåÃ+LĽióýû¶/žÛPŸ¿Q¡ â>;œu⮉k~¤{e¶¥˜Zî!¿(pzò]VdÚº‰ì3ŽK2#&u¨>‹ácÖbŽ±ZK°Vãæ8¨a»Wà =ñh€Ž*€ZîʼnuW¤CeÑâÉH£Qý>?ËÊ[ÆLèî‹$?}¹c› õ±åé•ÏÙ /ñv `¦øF,Ê_è}¿O†ûƒ¥¬¯3\`ý\íýÔ|™œ÷c"ÌîôCÿÁ¡'I´j„ :ÝÈ"Áæ,—j³Q‡õK¦úrÆÄS©êJE…]*Òεo&x¿q×ûäÔ–Í€Žñ2šuàøÔéHàܸ,™Éò°—iH/ûgèÇ*¬%߇xf–°?ÇZýHøû ¥¼ØÖ&¨Î6ž¨÷ÎHÙjw²éôme¼y2cÇOô5™$ÙœG)TÍG¡?^­?`‘ôÉÁ’vû”B؉®¢œÒtM·592ó£´úLu]Yl~éÓr¥oÒEnr¨ª«)d±v+iÄï•ûÚ¢ci•Ûs0qJìrlxÏÕa)’2ß´°…xþýò!H^w6úðæG¡°{I¯Þœ*ÍÁ×x f¯Òì»'A_ä««yƒžUÈ&šJut ì²Îñ·$a.Œ¥O?0Æ^ˆf5…2Thê·ëUO®I6æ+$)X`Bä ˜|c¼ êê.zµ‘–íˆ×,ÊÍÜèl~H‘ÜîÑÅÁO9_í"‹­Ìv]6e-®Isƒ0 õ£¸ã–­Þý¥ªý蠨俴sÇ \«'…¤Oïq™Ã¿1:ƒµø¢zGgC|)B1˜æ¥;–^£ž²(êãòÊtx;zãw'"ܶ>Fç³)=Ϥ ³¶Â—ƒLIo%ëé3-G=ÂÔžŸÃ<ó¥¼ÆïÕÉÐÓиXÈ)é%ÉøÙû ¤ü–嶰µ¼õ”ìÞE9syâ#n¸ Û×Ç ²±M‘ëå»yBkD’ ¯˜K“ÇW\b<ž¹ìyCkö1Ž.+Ý k,jjº­Õmpzv]ߺÊp¹~$:^ûÀ[[žúèN{dØ©^4Þ+ÒÈJªE’šWY“i]‹k*aúvLFmOÅ}i9qAÇšŸ‚I¦éC‡$›]bv)­Á }sFX~fõb´àûRÉZ4ªúÙ+DÁs\ñ3ãÛ|ÝXÙòµŽ wy«åJ·Ù=ÃV‰$)÷å¯ xR}üÖK1‘T¸ ÀyûC}²`K“¶o8Íò ¬,ý¬85ÒÏä,Nw=¾p;ØcIWzÊç…úyyç`lWÙ†ä®óÂ|*.ëõîãå±á‹ù¡Mþùùû¾r)þQ)b+䵋{áÊ­¨Áý‡fBJÓµaO,Ñ;§B:œBê¿X|æ¬x»çã=¨›¿ŽY˜P«GÐb’÷±ìY¬›ÅFÌ ]eסÞļÙx“jF!™\vœ= –34_žªn>4 8ýIȨöivS­kîÍ™OÑúçaáYEéþcÖFš£ÔÞ²&îɪÖ!òuöŠâ³Ó\r>a"~DÇgñAFI³ ;¤\•XfƒP;ìÓä¢:‘háìñÙ~({ߘ8<Ó<@u°~Rþt¦i;ÞoË1Ñ4)¥Û­\*¿ãb÷v÷$¿á„I§;ž»ÊFZùUÓ{ŠjNyåò%øÌxSÁôl%×:ä& %Ô‹GuX¿ˆ|¥‰N‹[›Wpiù½þ+7§ù“×öÝ6,S+}®ö‚™®zÚÆžFÄUiϯX‡¹x3uè§â1ŠéÎÄcëë\4Ge6l«ªxظ Ίí» ÑÕÊ©–mï8ŒŠë%Ù;0iY¶Ôíaô[W2ѵ™Iœ¹ Mmñ®7.²og@(Ë—äo°ªLdŒqšsD·h ö,æež? Ýõõ°'uEÐö'.š§òË(#–ËMH‚^žs¨ éþ,lLœêZYÅâ/°šëŸÁ—Ȳ ‰mfz¾sS"eØûÔáI|¥´Ù;P;'X8‚Ó@ï»e\÷âu±ÅiîÔÚi,œ¾Jë@8<˜‡ËùÇ> Õû¹Ì74=¾u"*)›¡þ9¯8_”‘Îó=Ý`Zåy-±±ÞW¹ûJ!âyŠ˜šÙÐȉAù–°È‹÷÷5ЖšŸÊê±x äwßë¦ $céТùÅUu-)=€«?禎:é.w"ÔrD­öåXCÉ95p¬§~:¥k—¸Ë4=R–>ÛJ$iúh‡bÎñ(š,§0$"ª"…áË¢'ÕÔƒmÖ$*lÓmËXÅä¶ZÀ.üµ°h ™_–.v´J{5άô£´¸GÓ®ÔÐýDI´Î%:–ÏZf²ûa8ûµ¼‹û$˜E ðɨ<éïjÛlçù16¼ð©ùúçÏ—®8U#ðO‘¥èÏŽ¨j_Çí»JNXO²JŸöy½mÇQè•9Ge] w ÕPpÁî†^ØiÙb ÓoZK÷Ãˬ)Û¶X7&½c¤ÂOL°5uH'¨>ë[Ž] Û+˜ÍT:JkA{„ŒQÚ9mñ*‹GÙ•knð9ݵOÃÒÎÍww±/_užÄ-¹­1ħ/ ŸjD@(6í¡®@œÛ—ŽÓ(m‰ç(C bÉŒÙæHEjêTmsk†ûGʾE¼X«¥bƒå2U=_:™=nyùÁ¼ù'6y"ÔsFæë‰ØXŒ˜6»¹›¿ÇU±1ça'£ü„5¯f"ê%õ*¡w³|Ñ+.Kçðöq1Ô`—”á]äúPñveƒ~êÃ;T£ÔS²))&³´&‚ªA¢+,K·âõÜÜÂáê!â+. žc1çP%Ê{‘…'W‘Ù(£"æFÍÉÎÙx½ßö'™]} §Ùo«$^R5rHÁŽQÔ³Sû6†c ©¦øÅg«–1þ䪷0&(æ JéF?â¼G›r)`O\`ù££:–ÛK$“t¨¼‘mtÞYŸ©×ó3M¨mÒiîtý}½déD)t^¾WÙˬK••¬ôçá÷Ñ…܈PÛ‰d2Óòjç‚6Š?”hdD°ŸÔýÑêÁ#ÙwöÚ t¶ÜaSœ)™§mËVÔ~8RxR´ÞNMF»q8oFÿÚJ@ endstream endobj 2114 0 obj << /Length1 1614 /Length2 8884 /Length3 0 /Length 9942 /Filter /FlateDecode >> stream xÚµPœÙ-Œ[pw:¸CãîÜ´iÜÝ5@‚ îî—`‚kB ¸[pydfîÌÜûÿUïUWukëYû¬ý5ÝKu-6I 3°œÔ• ÈÎ)VQQäprr³srr¡ÑÑiC\íÀ™ÑètÁÎ.¨Ð¿¤Á ×g› Èõ9NÅ Pr³¹@>! ¿''€‹“Sð?ÎB;Ä ÂPr€‚]Ð褽œ!VÖ®Ïmþó`4gùYÿHHÚƒ!æ (@äj ¶îh²h9˜CÀ®^ÿU‚QÄÚÕÕQˆƒÃÃÃdïÂîàl%ÆÄ ð€¸Z4Á.`gw°à7a€*Èü'3v4:€¶5ÄåO»–ƒ¥«È x6ØAÌÁP—ç 7¨ØðÜ ¥¨ PsCÿ Vþ3€ð×l@vàßåþÊþ]ý#dnî`ï‚zA VKˆ &§ÌîêéÊ A-~‚ì\žóAî ˆÈì9à“ƒr’Ð3Á¿è¹˜;C]]Ø] v¿)rü.óÿvü7BsW€Ø Eû§ú³lù'~¾|gˆ'À€óY{@çïÏßOFÏò²p€ÚyýþÇýrHª+½ÑÕ`ù“ñß>))O€ÀÆÅË ùþç¿ÿ®ò7ÿÿpÿê‚üu6Î *B-@Î?9<ï?<Üÿ’ã_+ÃøïªÏZÿ‘¾!'/§ùóðÿyþHùÿÓýï*ÿ7éÿïäÜììþp3þáÿÿ¸Aö;¯¿ž¥ìæú¼*ÏËýßP=🫬¶€¸Ùÿ¯WÑô¼’P«g‰³yØ9yþ´C\ä ž` uˆ«¹õŸBúÏ]<÷°ƒ@Áê.ßoœç,NÎÿñ=o¹íó[ÅåùÆþt\žWÐõËýÁÏKößç…š;XüÞF.^>ÈÙä…ö,ˆgÄ ð>¯­Øó½8Ø¡®Ï)€gÎ~Kg´ß×ÌÏàPþmúñ8TþAÕ¿‘à³ôzö™ýƒæ£ß¼8,þð¿ €Ãò_ðù6ÿ‚ÏmlÿÏ’ä€þ ¹xŽyŸ±Ëß﹋ÈÅúËÉÜÍÙùyŽ¨ûy†ÿÁ¼ÁÀ`O°9Ú✃¹p¨M]hûu$™Ûæ¸âqÚµ>Ûx1Šë€ì”ÉÚ{­¬Ì¥Wer‹ý@9c›.U)§ëìïóç>õT ^<—lTr;VTf sO—°3‰>WäÔsX-0¹z)R”BÅÎ0ê1ØݨâVZtX¾ u} ‰ôø%rüJaíªÝ5%™Ê„䂺ëš®­fCÚ›¼Ä 9ëöñ!?c £RU?P¿é¼A{ïQýcpdh§Ÿ¢'&HÉ=xó„6ÜGVxˆ\†¹Zw%ó>§Ø}ÿ®™<ž”“ÁSbIÂfÆ€Þ>B@DõˆýÖÛ-·|áÞ@›ý²ûLÉo¡ |XŽ™¦_>oÕמ­(]P=Ã!µÇWqŸHdSœ)´s’WÛõ—ctÖÉEµn¡@¡‹¨óTÞ:\q/+æGªÑ6¿´R Âàð­Óê¿´H2(óé³JüV3uSjµÿæQÛ‚H¼Á™8JêÞvºdˆN,pRzÔ<ì [¤džŠÊB[’É[üVAAuÄ,Ífªv§–ùµlFÁ˜¨«•²:dý&“Dþ–ߨ²ö™om®ð+oÞa|Œò7mc´}ç"±Ãˆò!áµBÓNicd§Ïn©X·ËÌBIAŽ`ý÷«‰T †$ºÞãØ¥Q •¸›ðNÙÚ¨3Ê¿¤ÞÑ…Õ GBNõìŽ;„Y?(‹ô«›gD±¢)«ÃÇn²9!¾¥;=úüTêšB Óýx"«½QïESº0åÌN1×*xrP2T}ï`̺„#Gš†è“AâY×½«éó ]vÊRÏq1oîÝÝÍŠ(Ρ“-)·"Wó=òærþF¡1·1-Õg~k•6%šýþ‰Dâ§!nœ•1àiœ\–çF^x‹ëÃìêKÉé¥Ë°b ¿¬¯5ù(‹Ž©ê*é]éi‚¥“ÏÂÆç›À³â¸ CŽ”Hr\„„PüäWwûý^#{ ÔKŸurQY2“—„*;Bò ðÅaîq2}‡í…EVU8–æ¬@ïŽ1¬3ÑÒ…ôÜ¿4 ¯P?UÔa1Ù¤åNi´ï¾½/ñhüûÜ#\¼ §ŽÏ§P§‚üÖ^þþ“ †;œ…Óã6¿´A)ÏPf¶žužY"0:Ó\õ>ZñƒŽí(›|æN ³¾5ƒ~ô²©½-Ó`,ñ^¹ø¾áõØp÷Ë.,eöNLÙ5°Ižtö³ù¹"õ ¯£]8-fûFOÞ}aÄþÃÅü@Už.3i¦1Ûk£„›‡H•>‹ úK‡Æ7õ&žéMjyÓ×òº‰ktyoÓl|‘Áðöp<­Dƒ'>Œ6ì¯4_Çþêº' q ?Jníý™T” ÕŸKŠìW7yîwÍnç<¢„æbëQ‘up©ö¾Xi±'|ø¬‡¼ærÞûâ(§Õ]mYŒw•7ÈY\]ÂkÆ&ÌÕæG{JNDÜ‹ƒ/MàDH`v›«0BÙÃI®˜Õýã¢GMÑqdØŒo8¯´S^ljpaÐíYäØ<_º»ß~—T¤&Ü@ÉÔ—s§ ºSÿŸ"••žc<¦#’}ã¹vÕ]Éš ªHóYD$;} ,„=tlE‰á©Ÿ.u³¦œ1cJsõ˜- ¥®™³•k[¤Ä26aÀðûÎÄË%¢PŒÚÊtèWvœR•¶vX”äjÚ8%­ôÒUë¼µûjŒÏ‹ƒàR/‡ô¥Y£8hñ⩨Nl‰;Ð'åxÇûéÈ$7»è à/nþ1Ó\눹)‘»Váúô‚e&®8.ªÜ¿ëó§1Æ'V¢´qêPš4èËÆŽ_…£ÚNŠÒ´Ìµ˜üÎ×T£­¤Í„|Äaº-~*÷öœ"û{j€WàM†bH¥_Ì1©·ºZŠZ¶·AÖç==3§'^LÏop^:„‹.b’ø«jìpfM-QR/-Tú:^½š‡Ï}Ù 9wc˜‡Ív)¸O˜!LV ;¨†åÑ©LgS-²f ?eïJÊš6/L~"¬.:R$ÿe†G’ó;N·â›t<øÙkñù "wø%Ænr³çˆÀºÛã·xÑ‚×-à`•¦HšÇñ•¸½3g{,åö.[Be[¡pœttˆà;ÃÈÙ[ßGû.K·A Ã`y¡õïFû]B@«¯?' yrîŠ L¼*˜ 8 Å5iy¨)]ŠÛh' ¦èm=ì·§R]ÔªDjÑC雊Ær¸}DˆžSIûxaúÅÝÁçú)7fÚôGó½-³¸±–™h Ü{¼õ²ÂO¶¬Ly¸TÓƒ#SèÛ &Ò͈$ªñBS[¯ý£è,0 R—X—ØáK›™xÀú¾ãóªkÁ/š5Ó♳îNn/ð}%›ç[½ÛU·djðy ÈŒñKî.Ç[ƒáä?~Ø^£ô5ü1M©žO› Îú yŠ*¤¸èXöæƒæíJÚŠ-E+ûÚ,rg‹‘3¯‰5t"tø)q¤d#Ú6žµ c?If‘ñEAe¦ÉÞ<¤H÷êAãçi°0¶ bÍÚïŒíšÙ¾¼*¢-À´;–R[}é"öq<±ç²ÄßœNMKcË?Ô¸·íÎjÆMyIÇ·Wé¹,þÄè«Y§ª RÔ&Ñ T:ìžcí¼¥g«Óê~œ°¥#ªCí_!ËÞôœû¡g÷l†qkX>·Låß»ë=£i#éqíV qU©qÊ=FœÄ,. Å†ýž‚ÿNK-ó«AޞĻʮ˜ÆOd^8>ÀC¹ò/È)}!LÑÆr¯‚ý”د¿JPŒðxc­.;“~ÊpÓÂÙ’°qžôd4*‚ærËI¢oªÉÉß…;ªÓpn”î…T¤-Ø9IùSцàGV=d'6«këú1´ñb~Þà .IçxgŃEŽ[¿t#’»Øæì8œ"7¾ ½^Ç÷ L]—mLjÄ®4è‚í[Ÿ}Ù¢$Fÿƒ:J*ø•XËüÃ|ô åˆ_¯LЦ½ŒNJ»õÌî«  “íç“&õK:6;_×}Ä–´GóŽ;¢€Ä5mtëîUëZØÓ·¶g©]í+ï(š"†¥kí?@z[ÎŽã´ÆÔôøã÷ב¶‚ô# Ü—õøU¼\ úƒÎ÷c»F|Û@fY"¸øü}ÛPôw˃H¦l.z?Þ™î"NÅ0|~Ü=ìx¹.×0;™nY+ Õ2g¹Ö¿RÌÈS_N­ñöÙ¤ñ€Ë{cõf>ɯnTeý~?BÕ…ämB–˜•ü@Ô¦šK3&9®˜Ãmò0ä™E(û>*«T2Ìuš¬×tlåU8šsßçL&òÔ½4¾Øn®"8!ͨ0éº :· ÷®0ê í^Åç¼…« vå§1ÚØÏ+ÃõTM–Ô•Q·qÐh”¬‰ÇÑ^˜D~GÁ(“"–k Û41¯BÔŒÃÎàrWÅýl'÷=C‡Š¡eÃËwkD&ÿo‘›ìž¢öµœÃGÓ^#G´hŒ¨dºö‹áMïeŒN>؈‚éŠá´ƒIoຠ}êÚöeØÕ†Eâ&Û¾ù¥}Eþ4¤A¯ƒ^°14DˆÄÞ:+ìUœÿÐ? &õ–SNó¢KºØŽU>¾±ªR]ÊúTHtE$&@(êxE|I³×Y~s1gÒ%„‰)o\Ë„¸®…ƒ+0Ô>ò†¦´AÒš©vO„´ÊpŽ(†dË0ŸþýRŠ–{Ûå#Ù«T½èuÀV~@-ò)à[ÚŽÙ· (2Ãœ¯½¬™G¶…O „SJÒZk&ïgp¶½XeœÞü˜Õ½ø:¤…9v5pŽ|»¡Æ¶Â¡·Cr¥„¸ÆÓ.Ô?†¡’çΰFúmÖ‰ü¸5˜Ø õ¢ÌY*e‰ab|èéÄùê×FÐnÿÃÊä|‡keÔG׌Ą¢vjl×P• ÄÊ9ÜQOK¿nÐÀÃü,äzTþy¿Ìù&#ÅÍMÆmònÑŒ–;tg­EÙò‚ïMËb©„L}ÆÄ‹ž¤ºtîç‘ésÛ`‘6W..gQút¸ÎnØva€©æEí'2M/ H¤7»5¯yÝ_mÉÞBt¿¿¼mÙ¾im¨vJÆzO‰…²åšp¯Fǯçxžzhã_Û{¬cõÚ©=tóp;>N«‡<¥aõDo _Ärv¢ÖþåP¬PÿUœò1õ€˜`5mT…з“JáŸHéÓ[4Y^ ¹šXrb)» _™ ù[‘2·†ÎWKdÇ$Ô¢÷°,ìÜDŸÚŒpØ÷y¾^™f’l%I{¿÷ŸŒZG"ž¹®Xæî…÷¡ÆÓió¿+I‡lβóÐŒ…Ž7Ñ‹’Ú' Nh×]-³"nŸÄNŠŽõ–”À•Î"4Î[·Á¸Ûú!·nUwÚâózäÆ®<çÎá±ê8îóëd|ßtâ“ߧŒ4îF*¹žº>nÆVóóD2Eêñrñœ|QЉÄJxƒÅçYr¿Œ·êQ̴̚ó3æ÷=K‹8½Š èl ý‡çê“_¨]©{Dæs’•Z½’·?³«¤¡Ý˜.W_ÌHÄg˜Ô“:}Š±´õrU®:0sѯÅÊå 3)P 8b‰£L FO/iƤ#PŽùî{KédçQ˜þ$LxæÆÆJ7ö5‘ÐŽŒZÃ)ˆµá²ûeEEáQ>Læ2*úƒÂ‡¸ÑÖs ¡üál |ôι£f™ïÀõõWBÈ,¢÷Œ ‘{¡úWØå µÜ?Ø !¼ÔvE5ý ùÂXƒ·Jo†ø!Êí€!§‹ ®/Ô#fqtFV\üºQÀ'>ƒù[X_´>F]V~jÒ)0˜uU©÷¬4RíV³¡âümsa{gÎW¡3rLâ3!ÌÎÓ‚]$X]M¯íOiÜhZ£•MWšÕ`‹%ËÄ.ÀBhc’¶Oª#»°•EmÆǨ÷äg¡K¹$[#¥c¦ã-ÇÏgeß܈)Õ}ï`yfÎ.\Ðóõ¨%ú9è†#¦Ú‘nh;À®äãÓ2*³5±Êë¾4Öû"²(|¾ï…b6–š8Å¿Hµ{Ž”·Š—0ðf|P\0©àŠmâa{¶Jˉš¸èv9‚ÙIÞ™:Ôû¨#0]’€Oܬðšvã¥Ö‰nÍ^†2‰¤„F¼”bÜ,²¦¢a±ñÞŠ Wû„z;w'o›‡2ÒYÃzŒEÖ¬$WŽ+‡—>™èò'VŽ€™÷h.N¯ÔO+ÍØMbƒ<ÏF:ñ‚#OXi2 Ï䫾,yÑ"6WDiØ0·«‚û˜¸åûw[0 ¡@™ü«1“eoÒHóf—õøˆäLÔY“EµÂ›—x‰¾Üeê¡ì6 [W@Ÿ£{ÅlÎeœ# y"–¹D}Y áµr Ž±c¸ª i[ÊsªA‹•g††™¹]ý5jÙzVzÕPàø¥×ºÕ‚¶XùÅ5íµAe ®~¢ pוàDüU•û vÔç1o¾mÊæ{fúã.Aõ0ô¾ܹÕ@™:q+dNW;W…ÚDŽ]AUG.m!1ât"²¿?2y…S5]™…î÷§ÆèA¥S§>þ8÷ʼntÍ}§÷cGŽy˜34!¦ÜOO]B¡ûbCÀ„ò嵈ÏXF7`g è‰Øå¦ø%(¥Äcâ‚ö&9´«ÎµŠËãÛ YYÕ) ç •ô¡ };–Jþ跊xýô^–¢ÀF ¡É+3ï/[6ö JWèo+|øÞ¦~”‡Æï]Ä^#Á6¼Xgù’Cê””˜ä©Õ3±Þ¥ð®0' ð¤›)å–vJðHZe¯ó.9Ò’Ž96Yñœ€©¬´çÔš2þÁ]%¾–€È6j´¦ &– ‰üSä>ðZ×·#èR¾˜2b„#Ž²LFËljEôK*œoÌ€7ði¹³ÂüD-µ S™i¡/ƒN«1Œî ŸÞ‰PàN,7:'çüRIcÒ±âme§‹‰Bp¿IÝé‚5.NBû9ÏÌhZìJLEâû‡e±ûDÉžW®`+nìé2Ü[ÝDV${ö˜w²L@#·IûÎP”¢7mTK`»ô¶1¥¸^p´Í_©ÑOtPÞ¿Ä„"û»€Ÿ†üÍÖÅI‘‚ç7ÈZ]ßH¾»SË´¡wž¤¡[­J8¦,uÊ}ö¦ÌÃè8fhÙ…Š‹»ð‚Î( Õù­ ¼È¿Px¥ýä4j¦7äCÈÕ(íå-„ª™XDe¬hÍÊ{p`FsK°©ó~{zéPUÚ`%3r +Ým)²Ïæ`öªq çàðnsþ§HG¸n'Ø-FôõT|† a˜¥QÃÐâò×!r$Ù‘…Âñ֚ÎÙð½$"«¶©‡#È$É•O•sWX¼JËT=Üöà¸Îƒ¤¹Hö›lÛ"Ý雦¨ö'´½ènÄ1½· £ðâæ~–&@¸¸µ±*ì=­'“„Y30D#ÿÓ!ÑÈ|óõd††°ßQÇÆ«°â#z€èm‹‚>‘ÙýóIÄ/s¹ÚΛ_½]?ê‘â6ñ“;¸š«Š*ç9HúÃQ sñ"XGºàܘ™çÄiÅI»A¤øÍýbºZøµ£~6©![íš6.ÓˆŸ"yl÷, xº£H¤l~¶L¾™5²i"SÍ”WLšv¾›–â:ç-\âä¼#:Èê ê1g#ãýâ…M‚øæ~?|u4‰—-ì¥o¿P¼Ñ7Ù}ÂÌÇëµÂ„“ñ =“¹ô |l½þÝf«öV‹^YÛ郄ç¢M.¤²Š¶§Â5»’%Úá>¯ßÓZŒ- v®`J8(ˆu1Ëò ¼.$ó¸LÖ‰›ݯ,Ó•6 P|Ë£=<Šg$»&°Pœ SÆrúN¶nO™Â|íò¶R˜ £Ö¤ /aûñzk: ýÅO¼9búJ`Û&Ç[8‡¢îÐ(Fü×…>ÊÝ“^Å¡4MêH€Åp¥4¨þåÁ·B|K½ݳ›ZU‚UÙ»úËü]jPFàpÝHƒß Õvd˜xŒÛŽ°` lóCdO} ¸E«%}²˜wut¸ƒö Ÿô¯KëˆtgJà}»H⌉®QUÐÝQ°²F oŠ’ü}r;l0y=Üšaê‚/Õc*äú9<³Î†òkÝ­øgFÒ’zŒ¿\ªŸ@¹LsØ Q¤x%ä¢òßIMúúVм-³I±m[Ò.¬=ÝÆH·–Ò÷ËOkOïÊ0®Mz7« „âÏ/tN{„ÇÄfB%Š»ŠC|Žûܹà^ ¢ãh‹[’ÞáÛµ6¸Þ¾IÀ³>{4Ï‚†C“s—V§Zk¬•Ô‚°§štwZë‹ÃšT$è¯æù6æ&>³Þ4ÌF~Kå›m©¾ì3N‘zSxŠn•>I9 š£4 /.èX©u!]i!Ç|¶ÓÇñ©MÙUv°þ¤>CÚù‰-Êk~Ÿ8]£ö BŠËà‘\ùÜ b»jO‡5¹m||ˆÅó—- ¹ó©IÂPé!W–µù%yíÕ1åÐ!K“!¶šZ@±s–­ùžŽç?†>!͈Ôt±¶/SÓ d½–å„€¯°v[\¾dí^ï!Ÿ9ÛE׉•«NíwCZ"oO*¢íOXò¥SŒC&6¯•i­‡¶Gp@ù]Tv¨àzë!Ãå€Õ#ÖvóÁòæc~àAE/ÙÛlü‹¼\5íOžGV?gJ¤tZaUÚý¾rŸÈ i¼ÏÐlø…ð}g_÷Œ„3Žœà¢`Ú.JÀ〆&ôlÌ&#,£äÉ2r³»¼sj£ >ù²ø‘tu{߀!/h©0¥Ç_óȯ±ˆÕL© ÔO~Ñ-N14³2Ý.²6Ìž¶‚¾¦/öÓÑq¦îÌxLQŒÙT(:vÓzõ4HŸ@tÒîÝ\7[}Ù¾iÕ;a:)¯8¦ˆ –!j² §xƒóØñVD{Dl9ý…'YÉ~Ù¶¯›$;Üš½êë$-=bïCÒÚYž„ô-Ñ üFƒ´éòž½)³G¦\Áàòݾ6.Â/”±,¦oÓ¬²Ó4Þ(mpþÌ÷cä<ã‹‹Î㧾çÚmÂé\V Jö<í(3ÿÈÕŠ­7ÒDB#ä^HLõ]¿š•–O))©xÝÚ“ê0;blŸRÃS&§û¢"ªNÞ§)ÈÈ(Nž¹ï[B³•+têÄaQÑ €º¬Û{çO“ü¥È¯zÖ9uñ#ü4DíýOÙ´ŽkR/õYgÄm±qî@A¯ÄÜ~mÚ»Ö&™k{íVÃ`²°;Ö €%®ÿüË”M2CþU+¾˜°Ÿëå›]W-ØMW–ê3X¨¢d&è’¸¸¥ß|½§¥8uÓÞ¨•®ç i7*ÍÏ,©)VEÎ ^*.^qMx|sÆC Õ ²žÒî@_Ùu^6(6à»õ[wägý*©hU¯Aà¡70b/Яa#2ÑÄz¸÷Ò.éÛ¨ÿ Þ™·zÅA¶Ï3ÚŸæŽEàÉkÏ8“ø¥èWvIÊ›ða%‰«¡elîi¥[¬œu^¢Øþð¤ÄSÒÓ{"DY™úÖexuýY5ð¡ø•VŒMõü(Ýé'¼£ª—bŠËƦW »œ …(Å;˜¶g>—ê¿Àej¢¶À÷F?:ëfLK(V±Ì®²Ÿšÿ_h J¹§&DÿÞ\Rß…ÅÓ?;}Ül«'ð>«ˆ&<Z6 à´Œï×­?²’/(žïdN# á³ëž#W(€©ÚÂÞ˜@e,,q]$_GR0ÜdåšÄñ!éˆì =¶}Gn $–*éï^îìßóˆ7ÞðP…ž‹é;<ö³ U>„ÊSqÛÖº4.äù!Šø?ÿ1ß ±"m=‚Bá³ßÝZ+_Þõ¼1¥AÓ£  +•³Ap€—° d¬TÓ`r')Fõ±õwìŽÏÕÞÄ%R¥e“â¶Öïn,%p5üà¶a¶™í¯&¾ëEuTùÃ뛋fLÎmt‰š/ßõD0yÿN¶yô§*#~oÔÉlO륧;1ÍVðW£Iv¬àÙ5â!&•³f¥ûÏ5Œa5¬' ±>ñcvç» H® ø¤žÃ(EGžANäèÖ†-˜ÔÙµºßR)̉?8ϽF.©Þÿ•/àNõñõlõ K¾~Óã—1…¹«/å&’Ûv3c[PO9ô÷ØT4Ù,iw É?…o·à-Ñâ‘«3‡U"ÆbráОװC>ˆÈd¹Ç“Äâ@ÙFræ6”|ÖÓÊŠ¦ážlºÉà] ±©mEE&' bøMŇf\U¿HDf·”Êòmµ×–Ö79²q`P‰¹âNæÜ8äÍi¹ Y0)ÙA€õ:ë—A<œùŒ»+[3霭õ²¢g/꺌ûBk}“Ùä[Omò=F„ 1\@\²Z'„4ëÖOœ%ìëyL.“[•äZ‡Ã•G'_í†p¼T°_%ÔÁÞ™¿~¨Ÿ a/–‚ùÁ~M0.…!]¥Hp»|¢PX ŒEv|3õ%Øê¡a+ûÿÝiJT endstream endobj 2116 0 obj << /Length1 1441 /Length2 6359 /Length3 0 /Length 7336 /Filter /FlateDecode >> stream xÚtTÔkׯ¤„€ 1¤  %!Ý‚4Â03À3À Ò‚4è¡SA)éRº éîA@éøÆ8çýÎ{ïZ÷®Yë?Ïîý{öo?œ¬ºú@9(Ò¦ŒD | €‚––š8âI89 àhGØ5 §!ÌG"$þ—ƒ‚+ ŒÆêÁh¬ŸPwsHˆJ€@AHüoG¤«@쇴øêH E©€tör…ÛÚ¡±eþ>¸!<qqÑû¿ÂrN0W8ŒhÑv0'lEØ „Àah¯¥à–´C£%øù=<<øÀN(>¤«­4Ï}€mxCÁ\ÝaPÀOÀm°ì72>N€õ[¯´A{€]a¬Â!PØ7æ Àè«itœaˆßΚ¿îþÜ @€OàŸt¢&‚#~ƒ!¤“3áGØlàŽ0€Ž²&Ú}F@:‚QHl<Ø w[c~u(ËéÀX€à¡ ®pg4Šwü ‘ÿgì-+!  H''"ùÙŸ"ÜÁ^»ÿïÉ: Ÿ?‚ µù êæÌÿwqƒ©)þqÁªHþ£³…¡" 1Q!1aÌó„ØñÿLoàå ûeø©Æ"ðóqF:l° `~pöÄv‡Юn0?Ÿÿmø·D" €Â!h€5ÌŽ ùOv¬fó[Æßî 0a¹'ýüýs²ÀÒ ŠD8zýÇý×|ùtT”ÕMy#þÇ&/ôøÅ@A@$*ÅüþæŸ øü/­.þ§9Ð2ª!lØ<¿A`oïo îxÁýggxÿ.¡Ä’àþ÷ÍA" ö#ðÿ½¿BþoÄÿ™åÿÅýÿnHÙÍÑñ—™û—ýÿ0ƒàŽ^°\vCc÷B ‰ÝÄ»Á~ï² wsúo«ŒÝ9„-–ã@a>Пۅ£”áž0¨. ±ûͤ¿g­áGÀt‘(øÏ'ý— »v쳂ÂNì· ŒÂî ú×pÊ0ì–ý»% ý¹Ž‚"`WW° –XIà#€Ý[(Ìóáü|$ÀböØ ]I~ŽYT À¯ýSõKÀ2€ñ(‚5:c‡„þRý«0ÄÍÕÛÙ/¾`»ú[þõ(À`ž0ÉÔ8ò0ľ"¤á¸LîŽpu@’`'ùØX8û„Ý©4l¹” Ÿ‘>­Q¨<Õ! üľY[Þå8s~⻦’¥ÊKøÈ¢¼nËb7~uˆ3úÂ爑mœ¢îZŽQ¢<³Ä[×ÎkºL‘7[ˆe ¶úœ¾w+Ú'_xpQç+‹ª‡6h7µ”å§kÒ2Šn-?B¿·î1X9 »›µì¼eþ^†˜§öh–qÈ ð,¥ï3ˆˆ9y–×*/¤!Ѐ•Ñrl.M•l£DÒÞâ;r• î{vv7Û>˜·KÞë²Õ7çQžôåê~r‰ûD_‹‡ç~vdа±þQf°®È\ÉØü†&qæ5BÆÈ8]¹¯$þÖÅ6„£·•®éìïc’î,ÖLóBQáß>2iÑ"\g¢ö}>}YÚøLðí¬¸*-#¬ùøjÈË´™tµÓdüµ[(Üfq™…[!ä¨J£vBÃi†¸' \©“~Cª½UxRM0+ò¼N²ÎÈM©SkåщÓO›x8­>œ+Œ/äðý4 '(œ‘ˆ˜QeQ~%ÜwíÈ&Ô«uvàÍßËVê K¾YK·ûxÝëíéÓ…žÚ}æe[Æ ¿Åªý|&üb®®¢æ›ìx”~_ù~Z¬P’”åØýêòw–Æ+Jni€†9Í}‰,¦ÜÕ¢NHôâ—í\}E³Œú*S¡Ê!^N™ý&E³'ëóN ·)zùV·FΔÑ Mþ<„ïsX§»ç÷騙$Jؾ¥Ýi¿®7•VûÁk¤õâ&¢>BE¼÷yñÚcö܃0óX™D–î’w©ê×Ûˆöž>¾g¶TŸcäm>°T¬Äˆ¸µ9¼D¿÷!»dÛIé¾RršQ÷#o2cJG¨eP·åìôœvÎ% e‹å(åq4Ô>Õ=H!ìêüŠOè3+ÈÊ6 ½IQà”’Ç„ÕÉ2¼¿;ØiµµÆw Žm@Fò¨ÕáÍs¼íKç•c§8å¹õF ŽVühØ|èkI®¨#ÖZÔÞhHé¼8^B•ªºÌÆ·œÓ^;;’x@ŽDvjü‚» -+ªG¹ž‘^ø)‰6í Öpâqf“ƒ™%ô[{Ýl×n&¾ØBO ô[©³èjÞéåãÓ&ÞõÐô¸æVN¯PÇK"öê.3Öé´:•~¨îˆAÃ×îÅŠÝLÖ\~ÙU­OíÚ]@>W<8TšüÙœm%¬€‚ÊH7Ÿr@Ú8$s=CAZ^0 1]NaÅ5¼­3ø±[  bóÈõSQ,F ÿó|< ¥·ý¢~†[φºMêDßVÌ­ËžÙ~ÛvvN©Ïa^:3ÇMÖM&&Ç<²Ø¹‚ lÑ躬Žù+å/TîPÓO0¸¿=óÎx [´Ø/Är- Hq¹ï¶ÈÙ!—\áoxÒÓyŸëu|Ëkùˆ’fÏD–[XêË‘Å i/¾xI•÷Mĺpå•}÷*Õ‘ØŸ”%Ù©ß´_®†ÚåÛ,ç,VžV•Àë=PÓVf¹ãylßÁnþ šÊuö&¦ Èt½0‘ªYRCV;ÆO.£)t“ qŸfX¼&Ü (]­X³_ƒ Ì–»ŸX¦«-ýP-ýö¥HrcÍX·ÏÃñ]üRz4wžnÕè¤öV ;ï 0N›\™•KDgÑür¢j„OVÿ†Öžx Bx™YÇuÝ6x4Ÿò‰OŸB?ð‡f¬p¶ƒ¿Jo%æã°îjRïx[Úze~óË[ç[Îzv€§ÇJÉwgKœ‹dÏ%‹¸V†ŠxêZpІm»Ù2ÕªDî ç"˜rü"8>aBaØ œ¢Þ]ø6=oð ¾ _@ò¼ˆb>fZ>TÃö3°·h,1ÓÚníè ̹cÅ6ã:B¡ÁˆðÄÈ‘ÌïÂC|ûç£8¯˜(¾ÞñšS2Ó>:Ÿ¯Ý:¦g¦,­sÿž«-Æd•}zxùlü©·õw7d èDè¨D‚¯=æa„ÚP ÝW_Ååik •¯ß»Ž¸y3y®CïÕ%ÖŠT©0-E+ÛU9%̉»ê ¯Mš³B'çmW**)F _ëE\ó†]¯2˹_7oìYkëÑ\ÆøôÒX°jLp—Ÿ#…VHAú¨gK{´=Óî¾Mæ~‡ ß¹ è(p¥=&²|þ¼POUb¿rp|'s¢±ôÇõ|’Ð#Ù·Å=qƒƒ]ÁÝ•Éi,”´A„›Zgßžøs¥7úê&×ôãK3âÓŸ{Æ ‰Q»Ž_-zœoÈ¢‡×ÄIðZò¶3Äe]—XòAkH^Ž®D–â²Òµ˜–T UÜø*XÑhiƒÃV½»t(Ù³õFJ/<ŸYÆ_!fNn‰ÅÒ-m¥³uGL÷’·JíÙ{ï;)c<‚Ä =bîúþ’õäÌ’óƒ×· bÞÐ7â‚9Hª¤†fÓy^  ÂÖ]¥œŽ6^Q߸Šç]j½„ÕÆÔ*¸Å9SH™>ÿÆÅÓþ~ŠæïÄ·ï~Xidm‘ØHf;äŠ,¤*ž~”ê7ã% ~­‹ÿˆ¤+,M¶„)Û’7*K·B ü÷ÙG,i€4¨CJÞ<~Ânõ¼è³r>Úµ>ø¹ég af&o ‰ëktEw³íq »a6Ž˜ˆQ÷ûX[cÈ»)-£êÑü ‘z±}ª+C^ÓÙÓ7‘wÙfæªæF’°p ø›µµÇFdÞð¤%HÈöÑÓ”ÎSž¸hDŸRCèwlÒÈ«öSé!/¸î¿¯¦ú&çP|'L°£½œµö0ÓÒ’núúäÈîé¨ìŠÄ©ÛÞA´z¸Ë@¬?¯ê‹”ß”û8Ãíòê¢,4¨švwÒݳ îw>Onâ°Q9tÕพÞð®…:¦„=˜=(TRb2ïT³OîÃùí—pÆ2ÉÄÚC•QL«Þ¦v‹îÉ5„Ø üd|}aä¹ÿcÓ ‹ÎŸS¾÷æjãÁ¯*ùî¢bñ[ûs¨û‹®ñ…°J’n…ƺ+Ñ̾}æŽÓ¸jqÃa¶¯eÜ5HÖ8&„Àþñ`˜ÚpNb¸5÷~mmzœþ~Ca‡ÒšrC“æ¬]0H“¬mÍ>ÉQf4ytójíga®/§ø|Úe^y%q9þ Á›hû¤ò˜6VÆÄ¥g©“†–QGÄþ0¤ó×8è‚R|Ú”ˆÿƒb¡µ¯O ýüÌœ‘mµ›rêÁl™Ð‘&Ék6‡ÉÏ.ièNC}]xºÝÄDžç8ñVÔiÀ’E`¹Õ*yøUÒJ è/ã!h ";ç<™|‹«nT±9þV¢)¤x¶bi`pÏr‘CŸDZœíý¦Ú„/f÷š^äX¡ÓQ0BÉÊÉtlS}ß3­Ç7G´`-K‚sÖ~5”ü”fÞGO¡Ö>8£2 T¯ûp1çÔBeJ¥ÌLä|Ü/!™4š¾Sˆ·Ùté{E÷ÀmÄ2?-¼X1O²TÖ9ÛxÉá—®×ÑO+÷KÝ j'Þx«‡¥Þé´Ž°±Ãœ á-–S+È Ú¤;î©j‘&°=<·(yÝH£PÈ"Ä›éh;nïNl“ J„ÌsÄMªY²Öñ— á2vgçË'å`Tžài»IUèµ?L­Ôw0:O”­§ÔTrú€&¤–8@íaüµ½‰Ã UÁY Ô¹HTElšÆ¾ùhºH:B-=é¯1š~¿†)”ì†Ì¹ÚËÞ5ZPÖ¿'ˉt5RÃ`#—/AØ•ZK‘çâØ_s²k`æìóž.& tËøæ¨ ¡EGŠ©ÕöÔ@ )õ™þž}AwØÐWÿûƒqÜ%—•¹ê¥mæï ]ô>ÌHjÏc%UÖ©U›#—þzãö°¦ó³å Ä›Uàc£]0RHøŒ¤õ\D|;Ì®í„lyE„ …3ZúdÀsX]0Ä@³zn¿´ :;±-ñ÷}Áë^é#γ֬.³§¦/Þ©Ó+O ÜRýl¦,ï2FŠ4%1:?ÁŽÇ”µŽám‹ ø51gÜ(Y¹’¬)¼z‡§ÖV¦ŠqAGXD½|£DófŽŽÓ<9šÉñ©9)“3È­@~“¾ˆåŽ¾Nߌ‰$D77î’b)wრ‰?rÕ÷“'É\¿]ˆ¶IÈåVÞ˜0sN‰làLvüÉÁT“¡n &Šiàã½Î,N[§YÈ™Š­Ön©³™›œµ¦îFç•CýË_ Ò˜4|ña ÷7ÏÎçpàèkÆ—LñÒ¨@Ø£<$Å ƒ´¾ép×sɸövØk:Ê`d·Ñc#¡ðÏ[©xç3Œ$H_MÒ­ò¿žë,毗?JÐ ¸yEb£^MX¾¬¾z¹ÙjÐ÷/•¢™tÇývXø !Õ|”Öªîô9"~ñ3«OÈÎóµ UŒ§¼Ha{/’õ=f¯{MßÓPª”26Ú‚D%¥‚#‹º7U¥êkRC.(ÂTi71­Á7WÛµ–œ_y5ÅË.ñ(®SN®Lêt·<ÜËæ*TQ¬÷ªâÍë™=gOòºÁèñ ¡K/.¼È•¶y›9²à=÷PïpŽ šÆ]ñ´ÛAÅ"N¼‹MÁn=6Ö>N¬Ê[σÜs]aaißäÒkÞÎ¥šŽºEgÞ´Q ì1ÖUÄÙ ûþ~vº…ë‡l ¢yš#x¨Ðp¼ªä’„ûÆðwž¥jÏÓ»ŽëAs÷¤?Ž%VÈhm¿zù¾~.H²vRÜîèQä…øžæ$¦¬€t1÷ ³ê‘B8Ñ=GB)׋˜<.vJf°I2 èñÀÕØr¼·ƒ£µ(Ú¾‘ZíJ’ÞtFv¤Yª _OÔ™ø”Ðéê2ïÒóÚ§dFã~ é²&eׄ»8 ƒz ¼™ùôa¤³š áW í3I, -ìwW‚N éÓ˜ø:òù»S ùšÁµTµ¾ç nÁÜŒ©ºtÞhLv+:=ðöe9CšXÓ(()(å.HÞ+Ž‰x ›m‘¨¨ÙupwpÖéŽY¢ì–ÎÙÇÌ,µsDÉBsãëbÆ^ IBUuË­eÞë—î˜v¸»˜€¨;Iú ­‰÷ó ½ù·½”Ÿ_ÅÅŠïö™›B?¥¥^Þª,Jâm ‡¸Ë!meÝ/‘ÿkhºÀ½D`rÓÁ¬eèó^g 7X¥Ò»µè8J}6¡]9Àò¦êµ2cM܇י,ý!éËNÂeŸ·yßèƒ"¦Ò]¾ê„jø¬:È|TŽÒ'QmíÝšÈi³¾>Joß6`Ÿf¥²ì¹G(×@ÖiXò(˜Z“bïܤưÂÙ Wž–Ž]þ¦=ÅÀæG±½:¥Kç—|ÿ£²£§ã±ŽG\á³›çw{R3í¯?÷nï´-¤NQnEý°œZX 8ü ¹­ÿ(™elôcÎKÉ š¬çdÝçül€—}Üy’Iª”%2WÅ\,n¸Ejú¹¯]MËÞ=l³zèy0æ,O$Æþ"›ÌQ¶Š-~ ï/~ÌÆ&Ù’{¡‹’ukÁ±pŘO(-ÆÚŒ„#Ÿ,ÕŸWõkÖÈú>ù¡aå jÑä@6¤m´…­Ý­å Š–µL”“>mÐ3w᪫&´UéÞ²xá½¢ß}¯Å/QsÏöŒƒº¨–E3½Ù½hú\*ýcsÍŽµa†è‡»7ÖoU·K[ÑþÐUoÄ€ˆÝøl˜ò',ƒÑ:Ìb—ÊyL¼Õ¥Cû:Û¦Iþ@œÕÙWì_v>Z¥Øo^™®7“…yÊ.'^e̦ ÕÈãeÛû¥ö{ç~¿b±g@2~óoÜV¤kÚ}>‘R­~eÎx:ãWÓAp+L5>iç®Ö_ôû7…oуŠ¼cÍ_­G8ÛÓY³ d wbb¹oöy0ãǨÀç_{*Ô%Ï—èŽ2¼Û>¨ÃùjÂFj-¤.-Ĉô`ø]Üq^oÁËK}àÙ'Íw2ÎãFú?\J2<øÔàĸ…>yBSÉ™{â“GžWñuSq[ݧ–ÍžÖd|9_Ò•Ô~}gsœM–ç{8‹iªÒªãƒ› Qq0[Ö6qš””ð½ìŽÏ*‚ߌL¢çiÏô#˜Tà=šÅ¹ I—[|Uì`ߺ[;…ßÏFcærØí ½ŸÔIå_ÖD>ä4Æ%S%½ßT‘þÝ/g¤M©YJUS ÉrµðU´P+63º§“ÄÕV½Èº>=ÆJžéží˜8ëï Çy8“ø‹¬€ ¸#j¢ša™=pýžyñ·;µ±â`:dzØWnûÿÈA» endstream endobj 2118 0 obj << /Length1 2779 /Length2 23861 /Length3 0 /Length 25404 /Filter /FlateDecode >> stream xÚŒ÷T” Û ÒÝ CƒtwKwKçÐÍТ¤tw‡„tHJ—tIww‡”Ä÷ÞïÖýýÿZç,Ö‚¹î¼îzžš\EIÌÜÑ(åèbbcfåH(ª±±XY9˜YYÙ©©5¬AvÀĈÔZ@WkGþ? $\€& °ì l§èès³°qظùÙxøYY쬬|ÿ3ttá¼1q·6(2䀮ˆÔŽN^.Ö–V pšÿ}ЙÑØøøxÿrˆÙ]¬ÍLŠ& + =8£™‰@ÝÑÌòúO:A+ȉŸ…ÅÃÃÙÄÞ•ÙÑÅR˜žàa ²¨].î@sÀ¯‚J&öÀ¿+cF¤hXY»þ-Ww´y˜¸`µÐÁìáæ`t€“ÔeÊN@‡¿þ6`üÓ3Û¿áþñþÈÚá/g33G{'/kK€…µ ,¥À ò1L̚ع:‚ýMÜM¬íLLÁ17H‰©LÀþSž«™‹µÈ•ÙÕÚîW‰,¿Â€»,é`.áhot¹"þâ÷ÆÚhn»Ëß“µupôpðùXX;˜[ü*ÂÜ͉EÓÁÚÙ (ûæ°ñ·Ìp±²²òðq€Î §™˯ð^NÀ¿”l¿Äà |}œà"€¾Ö@ðDWw äâôõùSñ_„ÈÆ0·6L–Öˆ¿£ƒÅ@‹¿1xø.Öž=Vðî±Xýüûɼ^æŽv^¿Íÿš/‹†Ž´‚®Ãßÿ«wôø0q²˜Ø¹Xl¿–ŒüÁ÷¿aþmÀÿŠÿKªbbý¹?"Ê:X8øþ®ܼÿÕáþÏZÐýs2ô€ÿfPrï2@÷{õõY¹XÍÀ¿Øþ?À_.ÿÿöþW”ÿ·Õÿ¿„¤ÜììþRÓý¥ÿÿQ›Ø[Ûyýc^e7ø,ÁÇáðMµŸ²"ÐÜÚÍþÿjeA&àós°´û·Ö®RÖž@sk™Õß;ô¿)€ÃÛY;U]­=lLàýøàÌlÁWð¬þRÁ÷ôß”’fŽæ¿‹`ââbâ…=q|ØÀjôükµ,ÌŽ ° \ž/ÀÂÑñ×D¹¹,b¿D#n‹øoÄ`‘øx,o~#>‹ä¿ˆ‡À"õ±X¤#v‹ÌoÄ`‘ý8,r¿˜‹üoæ¢ð¹(þF`.J¿˜‹ò¿ˆÌEå7sQýÀ\Ô~#0õßÌEã7sÑüÀ\´~#0íßÌåíoæ¢ó/â[šü‹8ÁÙM\Á»eíjûÛìnú隺˜˜ÙÁ¯( Ðo9Ç¿ò¿ê_8ŸÙ¿ˆ ÌÌѼLÿæäü%±·ÿÍâ×–±˜ÿÁ)¿#€+þ';8ÐÞÜÄÕê˜xÿ”qÿŠãì¾àÿI8~yšÙ™Øÿ‘ Ü‹ßìcñäü­ÿHÆ¿!Ç/èþ›,Û/Áï|\¿ÌÝ\þÈ6°ü‚ãÿ&Ì ±•—“Ðá °ìü¬ànØüÁã´ý‚»m÷âJÁOf–ß‘¹À®à‹þC®Ýñ7°³ãÔàbœ~«ÁÁœÀ¯o‡ÿl'Û?Òÿî'˜µxËO™Ü';7×?âƒ%ο‡÷ ¹]ÿzý»[ì¿„Ž  ¹©ÝòrpþVüŸuäþGó_{¾¤ÿ5fû5ï?¦Ånþo†\`'W ½õ÷™ë— Ðý™qƒ¸‚ß ÿ– ݶ• ÌêwZðKˆdåücéÁÝy8þáŽáöÚýfæñÇ’‚½=ÿ€àð^@p½“Gòºüê?r37ð A½jÁ—õ?ü××) Ðh†¸0ëh&dSÔzW-FäÁ´3&4E½£JÏä³àÒæö€ —D_•°ær#–4ؾ¼%I÷Ct‘ìÉ稹.äk‚jËÏwFqj“;-ˆó¸ßÆ ŽÄ¾ô‘ 3iˆî¾{r~§åo Ý Ù!GëìÆ‹ª’uçÑ+íù¥¯tiäãìŽên·<Òcéw¦(ÍH}ÿâiê<Ó¬| X ükÌsO´é7S˜9ã/drq ˆ¾ÇQ…>ºëìÑ÷3Þ+åì®Tºø$Ð?0G&i|Ä÷“åðæ|>.›÷l,$ËAaLYfBgÞgϨ²V‹phì©qYhgÛÎMÑ}ÃÚN¨)iÄ6r¡PÆ©j1ŠÀÕpØIöÛ-ÄWZo´º,lRˆù¾yL£èXù5uùü¬s^úÆtû1åëÝ`ãÐV—¯v±HŸ±¥ÛkI+½ðå%þ ’˜RÐ(T§6PqÑ=‰÷Æ$XÐ'àBüm³€•U—¦r„ø”+÷ÝE´oEý¹Rûú4Æœ¦AàK ·Ò‹iGÊùy‡¤qæGº¼ É®°¼ÕZUZ`ýt`)Ëùz.ó›/2…øM®ù‹~ªBNYéž¹ÊÁpý/›{¾O̧Zš-±\ŸTÉ·#˜F÷ëýă{ùnÌ÷7oÊÊÄ qµîÏ+ݧ=hûò—¯Þã‘R`0íJïË(ºå„+¯æµÄc´Å‰}÷´èc_-«,Y0æù¯Ÿ©aâ6ᶵHEõC—¶©d1ï·)¼šôó¼{U„o¤7öV+¯ÿêZ5· ECåå‡{`î="pˆ1Þéy‘0ÐåƒUÂCS‚îElP¤1ƒ!ü-Põv6ÖHÔ›ƒPfMÅÄ}Ò›@sǨ:gû›néëÒíþ(̪E†~S'¿>+¥#þª¸ûZZóy­£¡S¨(˜•ÊÌÊÙëÖŒ­¼¼e³¤Ç=ƒ-s„ô…ÖE)¤Öä¥î4òL¦ÆµkùGÛÑÙYVve-š~¯ˆZ!oÚ/IžÖßxº7áTeëy(ãÛtÖ*l*)Ä»ڞ仚Gg8 zÇuÁÃU|Ú+àŸ%VÈö&0ˆÆ8/uA /Ì/ tWzÓ¡(ÍÃG%«`ýL“ 6oI´á`Çí"‡#¥\)(°s¤2Z -&†?äþ1áÃêw»|š½ú;¹:Ú…1íʼn@šÎ|VyÀ3N:‹5‚õÂw7ÓlÂn DÞþ^o1™Â™HD/æK½\ÏæÍ›˜[Ei²“Ø8³ç'õyvSD3¨¤àT HˆoEüf¤v©"œôýv™“~‡'¦Î¦4¢á —dåèò~&qŽÑrs²týíÆù©²]#Öä)›ñ2¯ó•ÃéÖÅÕ°À9ãïÁºIjDÇ–ªµ\OèÖT׌ÙLm6)ÏñÓéý”5€å³}\T½§–×úEŒÉ, Jï÷ø)ÅNî—](Y$ãÒnæû>3»ˆó…Qá84éHTH°=ô‚JÍE)o?óBLØ*…LGÈWCæN |ê<œN°ó—z ƒEšTº /‚Xß(4.ïþy_þ¸öº5\ð[Q^ÒVVV)´kgŸ`½t Š„¤q°ÖA¾ÕzÛ`/5Žj˜T¡£J[ïæ$FàªÒØ6ŠIŽÖ[¾9¾cÝ(W¾HZã­ R”c—Eí{~¥ÄÂãשý‰d¢°È™†$Ú+Á4윘%vÞH»TÓìÚgXµ+Áߎ•ùL_>íP(å` j÷¼–Â4Xa)2ö*T­ALÔWMê&’z]ºl­{µo¨¸17ú¤hXðQZã‹ÒWyž˜*xµ´Ùc=—ÂŒÇ}X':Î@ßÖšybäˆÛy•‡kÄ1)HMÉüã«çÈó)ùÀÎ4™HÚÚzaH]UÑ14Ù…uõ§í—äÌN½Ïô<½4í¹VcYÈ5:´ÓK&Ý =´o]>L¯¥±89$AË)äá"ÁÓ%‹"ÖîåO¤#VÔ+‘øÈKÕ¯-Úf˜®v;¾Ç=1V¨úd¿dR ™ÙÜçóŠé¾y¦°¶üu7Ÿ3áš-•ê–‹½ØË©kv÷·«KÝ5è×i-B%·C®?Î3°ME}â¼f‘àÉ%$È"|ðÌà™µ?è6Ÿ‘î²â9 ­›:[ºª¤šS1MêßTMgÐ(:u¼ÂlB³ÃB~×ü>£Žwÿy°7¹Áb¬ßÛë³{ Ù†Q­¸wåPHÎÇkv÷éVxX¨Ã«Ös!=o!M>‘óÌØ#&/…à[z,Ã!ÂßñìH!èˆz©¼¿ÎJ«¸·sŸ=(O@ KÎr'g½ßo —J @^Y{8-‘\3Ñ °¸Sñäô“ ÎÅyw§ýX'°$M:‹®ƒEÁ!~͵põ¸{·!£O1U¦{Ç–<–vbÆ¥tÇWê$ÌÎ#×"þtUïÊ}Þ`Iw]≣ʈ Ö«ÜäMvíkòç'S£â3àƦ }°÷ª¤·®'w<°é¾Ñþ¡Ð«&nH¢&nb×qäùÅ(wN'‹¿–K±¸êÕÚîéöY8VtKZ=чÏ?Yl ‚×ûòÈ}WOÞÐhû ”nϪ¾m¦‚•Y+ÞCÉT´"|˜a^Ð!¾wÀ)sIF7º>vXöï†GiW~܉)Á0) 8$G¤ˆÀ­â/YbÇ׿ÏÆý#eLZ=qòY^„¸'Ÿ†”…ŸÖÙù<•C O½oV(YÕtLú3CH|aDÖ€ÈO–í@y5õà uDOõ£‚U©ÓNÀ{u—g/.€¢ŠŽtÉÌ„)ß«(bƒW%“x:¸Á¢\~ñ+k“t„ˆw™qÂHY=ƒodlBòÑ·\q\p<—“w™»…'f«ªòÆmS*ízh’«(÷Ó«/±ô“ÛRÞÆý"ù[²f¢ê24É‚/0­ñ§¯€î¼XþYRÄú[Z5¥}W&bä^^­[üS ØÙfäŸ<Ïa’LÇŸgÃ6X,|9Ö’3Ž†‹ú·(¤K’1*öh-]'á¦ÉÞjk>ù>ÞÔ76éôÖG"ùMÏQÛŒö“Ë´ŸE(˼ÀŽF“Ý7Dì‡ÔŽ]~ˆÛ”krÏC­õ7X¦D¤Z€+E[¢•°ñçJ RÇ"i9®îºM¨ÀÏBH7)¯ócÂNû`þñÀÎQ²N‘î2Ö¢±˜rÉxnÂAÇÇfɘã”;I¥ñWZY»Ü§ë5ä;uy?¦Tº‡Ž}óa%«RÏ„ÆœLö ‘üç­Ž”m¡³O3¶œî+xݼ–©\.“W®’Ü-|Íœ{¸º¼Þ“?­Ìd=Ýu£j2V@¶ƒréw­ùe.sÛ•ƒ{—è*€Êþ$RZíÏΰ¸Âånûë½LÒü¡½ÞÙÁ2ºúÔ1š'îó¤(Þ[†øëÃâQ^Ô~J›M Bø•oÖEZí±åâïp™¦8‚nI¬·ædD—cý-ÆŸl›Ô˜2~<¹ÓúõÛ3cÒŽªou¿iY°óz$ÃçIÀÍGÖù1+ñ¥¢Ïþ Þ°¡JÚËH:î®~­è'ˆ%réy=jdóq3¬†­Â!Ôèça_'X†]‹RK>+*’‡Ï}¢gŠ÷Æó†ß›¥åo¸Ê` <úZ^m˺8Ÿ‰}æ° XÓÁ°v+÷ë^Ùî¹l›<;þŽ©ã‚~@òæÇ LÔ¢9)œZ!ô A „|ÄTˈM¿ùÕ¤´`ð«Õæ8ŸÞt±Û ¼ú>Ÿ|©m|V2µfXjm_ Êv¤@ˆZ(ÖDhõž¼²<´5ÙŽŽu«“ÊLìÁzûàµÒTZ>õðºjÀòæ…Sä-»Aw,çg™ôÑüŒ¹Ëtî g cmfªæÉÕ{äÛÂøø ácúoWi2±ÙP_?WˆñPø¼ù¨&ÈÂzRì뿾wÝ)òûiË'ûö:ÞâSIæ$ôæËøÓO¬…¢·’ÇÅÉè̇ä=¯å–ÔyP;uH«à‰÷aÎl–PHh¾/ˆíË!3”í†v4·Àc.b ýHVÊÛ¡âv¯U'ÔÐÁ£\¯THyá+>ÝÉ€§½9¾í¹ÿ …g®%­0ËÈ°‡ëŠS.ý\°nÈÒTâ\×ójðÒ*²­üþ^—´Ü;èM H™ãˆ™ž‘³¯8/*lC/¨me¼­Ø*ü2—?øò~¸~¢õFnæ]ÊׯoÑ|'‹%\Âð¾i ˆ ì#Yà?A¥~ŠOTñ€ §˜j•¿™¢‹ÓÈmD•º}Ðß#|º5œN€&!„FHCè|w¨x£ß€"ã6Íý¢Ú÷4‘J›¥Ê1`*àIÙMÆqE¦>Â_jcŠq²DÝ’‹ƒcÊS~9Í$qÑZõï&ð4¨n†m¦ß<Ëéª7ð³Ëº‡ærlG×ÛË›@gT¢ùL‹e]®_£¤Ú%{)3=z=AÖ)¢Á":/u†·§ô‚§þ¶k…µF°œ¾Åj]{Ö¿Ó¢YêM£UÌóü (zCu˜I&Ò`ß'F¤P]¥&Iœ Èëª[$¤(5`½w÷ôÍ9Gëw}ĆyÐÉó³Î—q žÔ{£ Ï®¨@hÍìw´oãëûC‚¤EÑ4gѹe‘ôÉÒ‚ö–o„ùµŒ~B Ïñ½–eõŪקð™Ç?ãÐ\ôþv1è¶-•i=²PüCeíó½dl(/ZÀ£faæƒ}å¹%Vø0Žó4ÿº¹ easÕ2N¨œs?Raf««# S³h<¢<´4OÌEÅ;Õ‰¨ÝÒ‡€î¥·ë»àßO½ 3uêXá…ü~¥3ö­ÅÝbg1³çêHì½Bq³z_Ìçå2õ—¿à’V_âš/âû=Þ£½ƒµÛ°M~ËÛsyǩ윬|&‘sïf”'{‰pY<ž¤µ›‰˜ú˜¡T§·¼*¨Ùõ–ŸøUÀâ’ˆà õz ‰R˲¸:ßÇ»À®/Q¥Ù=³½™_,÷5ZW¶ MK´ûàa—èj_S"=`ËïöéqÂ#¶‚ ±•È]*zt¼  ?Ò๤ˆ™ËJ³„úû"2Ù ¹½õ‡í.ÑFžÂ3å9 _úË3Ì­hªï¨DŒVŠ•JÓöâö.?à}™ø>"yÆ>?!£ÂÓ…WEz¢³­=ÈT+5ˆ)©7g> ¶™C… apÁN-¨Ú¼»ù~ÚÊpƒtë`ä¸úç~ÊÅ~ó"'É ëëC2Gx>—„QMïa_nuƒ+!£­\ÿ°z @«ª†“j¡’š^qø­²cÁ“¶2w)³ù· BQò½¹”}1CÂJ¼=y¤Êe‰FjÐòÈlp Ì®:C“vŠ^VåÙvLFœÒtŠ‘ì½ô=’¤@t)é@Ú+¥”ïR­8×ÛæŽw÷V}oæ½ %e8T6wGü¶GP­ôp%hPúr ¬sÞ=?_ê@ka”œL”•¡¤U¡›¥³âQF§™r- þû$É‘TÄ26gùç§õ3#“7hS‹=°ÉD[¢6MLÆvª‘*¼Ú$°býî‘©Pº_^´µÕ±§TŸbU~Û“Ðò½š<$ÊÒmy¾¦ZЮ§ñÀ  hôŒk]‚@•küáÕ.*ŸïÀÅØ‘•K³ì= ùšƒr{sU2º½".îtî%|ÔlÈG‰†’¤1žÕ1’\£ï–:;‡=i²ÁÞK¤®)ÕéÛϵŸào¶g{^ñèkÈu¾ٽÀ¶È¶ÞŠ=šÙe‡ó‡z`G;™†pú„j¿í‰ j䦛R,øæäj˜€Ç|8£Yj²ÈñöÊ‚‰N¯ñùd«à>Á)Å€¶J{#ÑR6Lˆ4C£äø’± UfY¼H8ƒÇ ¤ç¡’VSI8ü>†>ÐbE–ȱÆóÛlÓÝçÊñëؤƒ¶K·˜c å›Ñ›{շͪH/ùÆý£´Ô š%º¸ 1Gx‘>1Ókdâ¬ià§ÅÓ™ƒœLÝ–q†Yþë5mÅQÐùI­k™°¬ËyîZC±ÜSMf´dÍŸ¬:Y*zWroJܯåËâ-”Ã*î …Æ=Ì·‰XŒAfªjÊl‰Móí*¹j{•S‚Æw_ã3uÞÙ&xÓ24˜ûéÚ´LÄ~k£Ú+B¼i ÎDyå7©„+öF0"‘aÝ.ËK„z+^õ$äç¦Êä›õ5ý7¼pÏ­|› m5Ï@–|2•ô«^ìT¥ž^ ô,oÁ#_$“š…íoÛ4:/ôœœ`7#CÙáViÌ/VÛ»G§ÜÀ ¿ïÿ’ß³¤¥ÁãÅ(Û£¹tØEÆŽ•¤œçMOŠã§Û`“;Û¹Cû–¿UïCÕǢȩG;?„"rx ÜÒ}m/¯ùTF áQh¢Ý».¤0á›×H•±¤\„4s¡Y˜·¨ßT#Ý5ú)ø¸'<;… ±“Æ® ÅØi_’0aœ'hÎà¾,7*·H­g›¿J?×[ [ÂÛ®øsáÇ'EÏí÷Ã=Ü÷„å9Z¨N͉<$âXbžµ~äggX¾,U_o`Ãâg›Z½-ö4Þn"¼{¢”Vô’¹)v²Ò„®ÔѹЭ?¹yÎE¹¡e{ Éò='Óëý£ù•å 9³]d^‡é+o~¶EJG½Ó0ŽÂ7 éV›É+ˆé4åþØ·Û»'IË÷à‡¿ÑZŸ0Öcà'G¦CãH˜Ëñ†ó›ùÇâ '1+†‹£^"¾«Ùª®_- ÚöàŸi&^Kì­"uD_¿¿ßîXq-Žâͬþ¼¼GÕðÒðƒlN³®ú áùæíûøù¸Ž]7!%ͬLó°5|Ùã¾Ý·¡ Æå)”®˜¹Ü΂ƒNV€óNêÊþ,m? Q[[q"ÝG|’°»CŸ<Ó­aÏp½ %N¼Øþ9õ=.ç«ŸÄ-W±ö¨Nû(L9±©a-Fì°r»¾V5t%͘ëBmUìæœnæí^y¹oü9Ëxë¿´ Ö0)¨ýðÏ=ß –§Kˆa>8ÿ>>z½£†f‹(ÇOÚyxÔôAF²pﳆ2@óæâ¨ÌÝ›­xãxŠï—\YO¯Ôýr®¤µ¼«‰êêâ Þ}\† ªMu&nÄóøøw'ÆôTÛ¼Í8*ªç_>0gxyYŽ[ñš:]u,gÃ÷'*a¸n¼{ä¦6¦U@Z÷F¦ ßç#}½Og}Ÿ:Ò)–ïÁ‚ÜÆ5¨KŒ¨”·C™µ þ‡aàRµíM×Ù8ôý~eï„*õ|^ØÏ G œv“lúã ¤W*Ò Œ;K!q1Ì“ì~¶ÎŒNxfs3Gëܱ{ýHçµCg­¤oØ~Š³”ôצ’f“¨³W½uâëØX-/6BžS]÷ËÃ2vîG†w6ðß×G×MÝçX¯©;J]CÇÛTömšL² |‹5Q‡t3žÉ Öª×ÃïzÇï]Ò&åBë9Ö—X¨¼c”ôt“>vŠ Ú’DDöª¿ÅIiÈv£ñí¶d+›Bäfz¨r[\9VŸÛ)OÆô“e ¶3Eº÷J8º†D£ª«^l[~~KùàÎÏNLCÌÄC»%*ѾÚÞÆ•Ï9?Ó›/ÌÓÉ w1wnöY,•5À—ÜS,玑5;çGÒ…Ñܤ ê¼»xŠ wc©SÈK¡·h~ÚHùö©àÀÞÀœ"Ï ÛùÞ)”_±ÝÒ75='i?¹¡ðá!5´¥x}YaKoÇdÕºlÒ½+‰Å³ñP2³¤@…˜'¶µêº,ª=L¢ [µºÙº?Nç¾— ͆•Wä$"&|ÿ¢Wû@瓈®›ºSÑ1X3æ™®ÈcF/>g*CÒ¸ Ú©9F!‘!CÜþ¹ýVÌÝåm«·û6¡ýÛÞiswÿi~¹ÂÈÛ2zýµ†Íny9MŠã …Y¯ÎR¯èd„ëCRXâkQMø’Ö»²S†;ï¬ûî뙜 EÍ.t>ôp|Óó¯ÎÊØMIjÁkËÅc®É¾Š.]Æ¡-fè[¡2zμŽYŸVwµ/—e4âî!C³íÞƒ,óìÈ…ô#ÂÎo7pl¨Ö6^¡ÐmBy§vÊ5<"ñ…SÝ™6)Ÿ±Æì¢Ë„G±ð·ÓVÉ3jâfs–B‰·ÛÓ¡ß™››Íßâ ³âë¶@ÚñïÀ¦~æù\Üd„NMg•£a=³KK„o€7fÙdÖ³–/ÞžúöÓ‰`)‰Ç9A+.&ñõùaÑ’)æ‡( òÛ²¡$<æÜm9‡‹+\ÁîÀËywØ,ä(¡”‘^œ=Ó˜ ¡2+;ã»ñÅ“ç}š„Ÿ…Q³*jçqoòð˜Ö*M¥4d)Æ“¥JRÃ,2üñÃm+›Ø™G™*²!³)¼7$Æ=Òï,ØÛŸ…;ä/¼Çó–Å3”ºD?i͑ןü،ԬC¯æ(žF›ÉãÐ`"ánÒ8”/ÔÇñeÎtÒt¿¦‡›W\¢«îMFPk<†^¸cžø˜AH êXòN Ð÷m}ûh ›DUíR*Œâì½Böó•ç¤+!9+?ÅÑÿ²ÝG’ëÑmyú«/î¡û‰«À”ôÛ9ý½/ŒÊ\áNîX=Búýé=µ·›BWfÔÇÜN¨– uëcÙzÑDVÄœ¯Ñ}ãÕa“ë Fú²ß÷÷õñéxŠÞ!׉ó¾)çÓH2Ó[»Mmh"ËÑö”1íÙ⬠„²DÔ óäš{¯ÍûJK›Å²\ °“Ax®Ùø–fÓ}CqBjƒŽ¹G´Õ?½™^úlXGImt¸º#öÎXZˆIø©ÈâÝ“,Ee §íB¨¿áBr€fÝ¢w(±iP–2m›eðýé2EËxš²„æí(½÷BCý Ÿƒ8Mÿ׆À‹ï¶½»ê]H\?œ&üwE½¶ 9Rl¦œôô)«¼éÃÂf_çLC„ò€ š8v]6þt|dNÔw=•4t»óæ‚~#»5 šg' wº ´Ãhæe ;]É šIã\ÎÁïq»¶¸ï¾øX–¸Ü"vZi©Ú‘ˆ²¼h“Üs…y'ËÇÚÝ¥îˆ{Iˆ¤yWCõä0zY¿½^±S4Ë{þ¾)îÙ>@W§l»æî¤Íú:éfûŽb¢.WqsX6[‚rv©Y7ø•~úûDx°³ém¼£Ôá¢(eã0©q~:<–# ¾Ô(2И(6H*Â*@z­$†µ!)ðNßßwq%›.dcÌê»XÕqÄ …3ŸËW0”#m­ìÑeáaÁCÛÛ ¤Tp×Gϱ7]+½½²|¹ÈÝ:"µ6–¦³šñš÷¹Ä‰T«31ˆ,™( :‘ÒìÞ²¥M¶“†?Ãäiz²de d,‹øû«Z”ä~ù>kúUŠÊjx¦C$ÁèfÉ&:úðÆ–™wÔkG«UÖ0'„ ÛЃ:Uð©W<ÒÂzGo^¬£ñ"¨gk|>Ü<#£y‰àÊÑñG/Œw1¦m¼ß8¯¾Ì-¼UÆÆÎzwP:õEQÇƾM}Öס¸K“¡e3Û™9yñjÞÒÌ›6F™XáÚìò8L–€ìàPaOmù½øTnX£¹o¹k±ðŸèt97},> C ì;™*nElÀæ wi&M‘9¢|[“U4RL\C|O pêr>ëNI-Qú ìÝÿ^uc»ï –2£`ýrÂ2×üÑÌí~ï•ùº“Úí™Ñ,´¡“Ef ¼'D0oïÖ9½OÌÎPú+á.äªÆHÌÊ d¢âÍÜ!Á`[õ½)#®&ʇّ¦‚"Äsþj¿›Ö ú¸)`¾ó„µåÐH¼ªïúdÃçÎÁÅ´Z\¾…‰.kžÂæ0·w¹i wÄÎÌzû  òõÐá¼šÓ ²Þ·Þ¾7«4ËbáÐ}ÙÐI#PÍ|§pø%7Ñ®oiiÔ6,Ná ÝJ}¯w–ˆ%t b<7æhSU3voªõ'»yÈRÒïiäZÇ}ÄŸÊ; (:ŠÆ–#çmܽ-t'…>ì-ËñµTŠY¼¶@½RÀ¶ÅÐÍ!´š _<¶²í¤x¥‘ø`ušÅù¥ä1Û+œC%ÛÂW¬§ÙvÕ22Šg‘åJÑTúÈñK¾ J®[i<ñó¶Té«Œc©©amÅƾP…ÔÃ9ÛEmú»Õ3{´©<9J$³ njspÏ€z÷Ÿ .æ8QO,¶ø_öÞjf…Ù°¦iQ‡”§pÈ: &Ù,³'G9š p¿èÇ€ÉL 9:¨µÅÁ"WPæÓgšP g ´J•NL²‡ÖÎgÊt•rWSÛœrij[{ÍÖK0—º™>¿~Û\Ë*7³×\¯víÒÔ0õ< AZz}@cÄ…Qç×zçïw& Æ>£É¾œ9;Ã~D£I BÛÞЬZFŒí ä‘ðïýLkJt× 4ã›ì9’'–7ÂZi|ÃïbK[ïWô›AQÖš6T-¢!åp¤Õ¸MÍ‹`NÅîq²£û“n{®%z¼tí9¦–Ê€˜|ôÞtXt†'gxè7Ù¹àrÊÀæú>ââóIÕ3T ¨”â+Äõ­é£ ƒ‚×L›^çö–_ZØTVðå–”]_± „§<Ï™éõ›ÀH”âì7o××,Ö0zY+™Ä:º¬"†bMŽõå3ãÄØ[“|˜”kLu7)€Ï–Œ{ó†_@G‰ù7i¶†"_ˆÈ÷ÛGm$:}Ïd –Q Á€óàÜ'öð¨G»ØÏþÔ“™Ü{[šµzä‡8´ê8~üÎèýÍx DR³µåÒ[Øžœ#`‘;*Ê™qÀÝH¡¸RÊ´{¼·Ù # Ë΋ÿ±ŒÝO¯:¬èÊü¡4þøXÆîì2}Õãó@êÌgvÙñ¶t ’§µz·.eª£Hþ( TLy€W›Ä’äÝunòçf¶žâ× Ÿaçí ï¼Ì3\“Öd¬ˆì²¯yÞY·èµdõU¦ÈSŠÃÙlšG(­R+>¯JhC²Êö_Ÿa @È>ŒçÚo[§¤z ÓÏ8¢ËWI(©)cSUx—ܸ™Ò2Û}{å÷daXmÚ˜>ä—ôR™fã2C©›"z­0«±,@˜“åÍÄkûõàS¹ÿÀ°’|×\7U8Åà–`?\VAþw¸5«sœôj¸æÌœ·Õ”Þ¢‡÷-Õfð#‰–j§­7ˆ´pŠ3¦<ÏKä?a<Šh…›2€:/(DÛëÃØN¶’jÔ (mÅÕ]ïiçÔ—I£ öXfSý ¤öÃ9uTÐü.¡2Iç«\ƒ$?ý¸?µŽíú|ux@C]ìÔ¯5vÓ‹'3lŸ Ëpedå+ÙÍsòÎå¨ëéM_= ÙÞø˸a)5hé 0MÑË B[m‡4aCÔÞ’ZCuøÞüŽøÛIø4ÿÒòæ ¨û\1õËãÖzGÜ•h);䟹ÝÍ|¥G:¥.È;'aFíT¦b÷!5ɘbI!a¨b>šè iˆ~>òOè’A²ƒU€Êa4ù¸Õ()©ºÜP8=ò@I›‡=ês¸¥Nz(5}æ†Ll¹kvEýí@{÷ÞÈDoX8 åINžJïº`Á€”Áí3R|% ]´Z Í–•HÇùAŽš´wíw¬'ÎUÖÐÃNÈýRÃh‚?l¯nÚëÙ'AîÍs¬í y¥Ê9k&â‘1Š·ËOE…/è¸XYY–EÖxä¯i¨‘Ûí¬­·=e¿e+c¿:¯#Ï7õ•Â~å¯Ý>ŸòºW'$n(—}¥ ¬Y1;-QvMûmß$Ü<ÄSΉý¶O]Æ©ED_óº±.›Ú‹Ô7á*›6YYV7*ö±ð­/v¹›Ñ5$·¨ jt˜a+ËzÙKaÀA$ת@d)|Y($:B1>ÔiTs^ƒ$öÍ‘$òPL¨Eù¬¥´Yµž€ðìu´ ¼ Gî×]!œvö¢j]'ïî$AŠ¼zÁ×OèSº™Õ9DDãw‰©ÚZ ¶Ž6HÓ0ß9!^Ÿøeø*ôðS:óäÎ5vÜ?† ó"$¤·—ºéép)騒:‹íÿÔwÞfvÃS\h/8³‡ªî{ƒ¤.ߘÍlo½Òh×}ÔÈ­ÿŠùîàÖ9„ÿçÖ =6ŸŒæ‹úzk€¢l}~\l÷·älu¨Èe'[~ò†ÞÅdÿJ®+ IØ´KZóôú>óU@÷‡hohR6"BÜR¡¼Þ¸-1IDÇšnØ—I¥ÅEh¶¢ÁƒL}âñO?Å °™ÌöÆ/ùó ìS¥Õ!DOryðc¾)êßÓ¼Á¿%xt=°‡É“¡ìµ_’ÉñâÿÆ’(‡ÑD<é–û õ¬h'z‘^ÍÔ¶½a’Œ“æüÔÝ'¡»¹I•„+X{!j23R”Š‹`ÿFQˆÞñÀqFEÂwØöúKÁÒ;OšŸ­¥!>b‚¾Ç2wÉ;« ƒVÍ(YÜ×5¯18xb5ŸÈ°ÝVí4­¾ª`MnAÄØqÓõ¬fª8ešïàK5lW&‘|'dŠëYŽO'…S¿?P7ˆ 5Y ~÷Toa[;‡?WÜi桱Pm›”Ì-ÂFæ šØ‘OyÿK0’êD…íG°[Ø ¹éON=Lìr†*ɺçÝBpPÏBß[P’¤ì7òmIȵz£&‚u-ŒI'nøD_ƒÍÄÀú6=ßW>ÔÇŒé^Ë%üPÍ­â>•ðIÀ?!£¶‡T—‘»þàü=úÒiušÍôÆš‰ßBQ¼g~äÓí<^w³ñ–Ϊ잣”Á¬ìÔsqÒ} 'ô­çÈ‹†ôSûAG߇EtœîóÕ =š x…÷hå7ŒÌ|ߣCN^¨èH’vVÄ&9Š5—«èp_Ìq£Ã5;ªçëþX;êh¸!ÇUß ²d´_ …ÆNÏ_ïË”‡=1¼˜Q»mŠO@N=µxÌ=¿&⸣a{ɬGP±Hj.ô¯Š{§YyØOØ'æZ#µ„ÎÑõ:Â&FUÕàHò(”+sûs›G&ähx8Ä×ë°›ìEЪ–99;i4ÄC˽Ì’GxÓËòù0å ^ƒØïšSŠE%h+ÑÔ¦â ÔÖo‰ÕUy.¡N†=S\Ê&„gŠ:gÊÙu:‘¢‡”g›Åå™vŽ^ø“©Wþï?–uzíéµ×¶°Lcc<æy·‰9‹zBNpx^¯í…¯5sÔì,¬‚Ê÷1Š¾u©¦“”á?áßÃy¯ukHE4ÈóÙôS©@Åè'vYïKÙ‹ª bÞ\Ø9O ˆ‹›Œm~ÎY•T>¢.[–M­߈¦Í(³X’*üÚ‚ØÝ›èZí“!‡0+JYáÏÃqцMéH|2ŒÐÎÉ¥³Âï NœÇü8–±b_¿.ÂŽ°4™Õ?$‘P䢘v³ËÞ "Þ€S‘¬} É8R±ó`T5[9ÉCLl膩¦ ú:d§>‡ð‡SØD]ä¼v‡¡dò<Ÿ{:­µÉ:ÔNM¬Z›w˜€†!¬ë è¤Ò5‘MÊ®ü¸~ù6&M «pVhGS„¡eöX–Yw¤ý@ö 0 × o9Œý_[‰1og›œ¦©O‰šÎ½nnrCÕyŒ°y9j|6Jûl¡îOÏEO"·y*¾ÈIòS«`ÅJõrŠLᤊĉÇ=žHŒk ½J‚J@h°j|¨ß“d"ú33K´Œ-;'“›¥p]‹B…¼SyÛ2P5U¦z­6M­¨åp‚Rð£v+ «™&Pü«Ýâ¥H¹ö «O]C‹{5!ÌW^ÞÕÎø’)íl_Ö¤¢ôiÊëÂÃu?Ùný(õy&Fã½›÷ÊQs×C\L]g›°[¬ŸÍß—qcàbeÊÙÆ:ìjPóÒ|‘‚P?OÒ…‰¹Ýˆ*(-xí¨™j#έ&Fcd4îIÁ„}ózËÐÛ°†MÔ$q±«1ÆÂå64 ô¤¡ÝH-äÓ\‰ÜZ4kƒžª”pƒ±‘®Ö~Éá·â*àM²üe½Ìmú©âŠ¸º³i¹õpeëMs)Ìí¬“Ù—ÁMf'·Ûg‹Ÿ÷°?)`öó²Ùãb¡2ù>Sj×ë:q&™^_ù!ê)%–BªM—„µ}@ál6º)²|ÙZì->~'dé2ê&šLèåäÇPX™Ì™ )úcÇ'`ÝhŨ—‹ï^"Úö-æ{nÔå ÖF_ 9„êæjYï. 5Ðçîá·ö’j µn—³„ïIõˆBNÉë׶fYæ^1oèõB{û?÷¹ÚN'U­¬, {Tº'!wºˆÁxØS‰ù”9ýAbNÛՠ˾ZÝ‹vâ]üzÂæ'zM1—îí.’·öûóS²|°8øôbÅþ''S©Ð=l'BÚEË·_tŽqµ| “3pÅE=ˆBXµuürºË5‡Çà€<ûÖ=£@§ª¦«•‹& GA:[I/?Ѷ̶ªÔwP{D6 Ü^1U3-¶’É–Ç›£Þr3:è„î1)ê…—SÚ4q#VԹчŸ°>ŒôwR¸Ù«¨åÁ½ff‰JÅjp;ψ±ú@‡^ :trCŒ‰ê,Ès¯N“OÓ­k×®L϶d.¯c½ù”{FÜ’Z³Î#¹¾#Tø×cU×ò­àkÇ–ï´BV}¶Î‚¯i,Ë>—×» ÂÞ3<¾ïÅŠEI³Ó„özùV£¡Mæô.êX„y†š©ÐaÆ_å[å¡/¢ÉVyÀöô”MÜqtvÙ1ìë>fχÆ3·¥u­‚èJ‹=œŽj¿Ú Ö·êÏá·æ86~í©ÝÔë6–ª.Œûγ«£6Ù[ ý„r /_/KÛômö8¤ZFð¢IzÝ0ú•×LŠ‚BŒ¡sGÑfâ9AkFJ°ÈTpó¯Nr¡3Å,@ íÂ9"E³þ-ÔøýeƒèéQqÊz^¥R³Ì–Õ¾´JCÿàeáÕ…ÆÚÇçK¾ŸQË؉Á&R Ál§”ð)ã8z-§d ž B_{§²½ÉáŸç ÑèˆÖ²æ¹IÚ58}OvÉ• &jù‚÷Õ¹?Éx•c/ê|ëB˜_MIÄjz4}Pu<ÐßzÕ7ÐMJ¯#Ž’;Øõ Ø!ÙÄí¿lóÙá ´AúB ÓOé]nÒ="å¹ÏÅî Æt^¥™? Ä ã£È“Ú µà®J‰5àæjlê¨ÎèH ì.2tÊþxt½‚\^s6ó§h[^`ë+ uñïTô6#D}Бº´.SðØõdʈ龓!vÛ,˜~·é ”KÁ;Ø‹ D‘ãb&j¿P9ŸVþF÷XÄîþæc‚ð‚‹k”Ôžì+[g¿wñíÓ4q,ÉDW$ eŒz½–¶Iwë[¯M3äÁ¸NóGFº†Mwç´ð~g|{']8`ëÞ=žLÀOš÷Ì¡]qѶ|óÊz“LoÍ‚·uh&\wOÕlæ™ 15)éÉ1`ÜàoZJõ½ð1úþFefD.Gû5cr–ètV|2Ûk¼0ycœ×ÚÎöJìÇ’­ÂoEº e·“¿¶)ÎaǸ¢#‡;¥â¾s¤©õa+Ø€•µ­Iàøhò”:»|ä–ê% ·-æyÛ09ÀúÊÛ<æ̾i¯¡O…m11J›¬!ú&ˆ ÁHÐ$&"º4a®ïBÎÒBUSeÈèÛR@¶ÌÔ©¸³77×-Ø%Qz™ù.ÚÙf¼ÏÔþåµÊqN¾Î©IËóÇPºÙöïø¦¼!ÜkqwGCÛ$HÈsä¢ ¯û²Ã/œºS€n}Rû{#º¨æbê >Ê,EÏâ¼|\ ø fÕqHë½Â®ëÏ×;M®Ù¬ aDMÀe$|T¦t6Ó¥¯ð"–õÄü0}ùäÏqô œ±J»ÙðCï/,Pfó( X«Ëgg蹌‘;…èR_èÍ®‰¾Ns[G«D<žûÑ0ïçXF öéWL.Ô¼†&ç)3¡‹ ƒ»t½ºz¿*%-"òT8Ì–W qZaíwÌ'ŽZ¾à0°|^ÌìUÒ»AýR/Iô厥k|‰Šqéë^\ÚRo3A(ÇzÞ:‰@€kkm>Ô“|—Sê†lü­¨„ÂÐLÚÑxuš=yí$ðµŠT‚˜|Ÿ#ük®ÏÄØ’»2岴בe« "ó|åð<È'yædƒK‘>ÍV\q‰2Ë5Æ›Œ ‡úOhò˜i&Y|™RÎ1‘O#:š4Ùéؤ²ÿ³„‡S»Vû†„ð—àòfa‰°ºb¼"gëu"cÚþ*þ8ÿr’Z;£&È×[ñ• ñ½gõ1:qrºKÂ,öB«¦X1ác%Lë6ÿ7$b¦Ñ¦¨åŒäRxŒä÷eã¦+^mÛ>”Z§µIÐÂ¥«¡¢xß4Œó мfõß×£8Iôž×”²ÐE ± eiVé§Ò©|e/ÑÔ,DÖÍ’„ñ”b帹U˜ûÐuLn²Lš*§¹CH'üZoâ†J7° ÍxçU† K<b> .DüUæbaÉÈxÃY‰/ÄÙÜWÜ/uêe×þ³ÝÙ?Ö9ÊÇ'–lüEðXÊEnƒ Ÿ‚ÎØ\nÖ63&NZf›S/uðgÜ::}4² Teöµíjp+gÙþpQÓ+Íxf¥R£‰ÇÅê­÷N1Š/Wì?„);3{ÂwΦ÷¯¬éPm_ @j¬Ær)p¦­lAؤÞWÆ•s-j¶uëû6ß)ÖßÍ#ÔìD1\í_Ž'hù°ŸÏˆ`ÖÛhXŸÏÒƒcp$q°v"ðò¶¢^v-»¢@fÅÌšñj?µàwÒp›zû3Þã:m3~n ˆ9¥ l Fº¥U–U‘ ”Íãap !ÍÁ†s0Ä\Ï„Á¼_âØméÑ©ùªn0#zçC+x@´{2EZÏ) Tû®+2÷ôñ%œ¼÷•cþ09ƒÂé‘p«_}‘™À~;A<¾­Gr:AE‡|lõòÅRú~ •êÇÒ„hŸI~„•¢;gr=LÞ¢ …‹\#¢pÅ~ÕuT"ºÎ4œ¼Žk_‹4¶@a( O…„_{ù)Gª.“qé¥ÀF&ÿH<Ÿ÷Œ‚~_A™ð›_?æMkžÙæ=Ý2¡yD÷ RßD<õÂé§tí-+5+‘{sYúUQw;ÄÝãÓ^¿û²òT«€eV[l«Ï%+“¨"Ùpê™~„ ­®­ÿOR!­ÞÚÍŒŸøf| ðÊäg‹ß]•Å«ck]~5B3è¨,2:¥L‚?Ùßw¥ ¨¦c³iC\ˆDÝÇêŠåŒˆØ’[K_¹f×Àˆ5Ûþøò‘LÄ7o³˜Ù4¯èðN#«†l¥š.¢ív†˜ÿ4Uýì6->¼8·…8³¥c"vü”£6œœŽŒ3$’¨É=ä@ Þ7E$l“ßó8aÂ_Õ¬ñÉ™áñ銻‡ÔÎÎ4Uùš¥>¦Ã¨P- °7‘(ç¹¥îÙ,håKÓSKÀ·{–íƒýø16Z¸x‹L5e°Õì°ñ•uZ#;¨Ü”’™/=v=ûͨ2ÌdaàNk)9t¦aZn† ø;ˆ0ÛÔV ‚Ü?î­!Ð}â~m¨•ƒkó©ß®=™ Æ()ù‘º6yÿš“ôa‚m&èD}4RáEËÔŒƒ˜"zìþðñõïÅþ¸lòÆãî LAßßwžò•aCÊ&Å‚p,!©Ä§>Ahj7<…´O¢°Ã<ñ°û4¢8“Ê篹æ«ûYêÆEåå• ó_60+gà“f£wÁ}’×Üq•}¶2Ä­jÏ°7sdjÁ/U_)æFvQ‰—{Iiµ¹æaS£vŽ¶¦ªÿôU¾&*Ñ÷Æë¬ù_Úœ!Hï'ÂB&™…— ï3º‹€$IÎ…À= éÕÇqrâÃý?“Ê?,=%4clæÀ©•“Š&wÊ1R¸ogQ´Ý;qì-hìÇf©ð#ÏzŸ¬¢Q‰+d°LÙYw6*M۾ƽ•v;üüT•a;ÜíC‡³Èp!Ç£¨Ðv®½z°*`ãÂûweKžÿ‰õÇÄ¥äpŽ„KéRHSHÿ^¿iÈÐ_ÅilîAšnƒ°: íR¢, Ã1Ûk3D‡ `Ÿºërñ…!¼)¢‰vjš[÷ þæ3dÁЬºç#£5ÚÙ¢GcœÝÖdyÈCF¹çT:`<úùÑ(9r:Îkj/óËù Á×Oò¶u=·¾?2<–ç³Õã`ˆ"Rèuû ð@´ìö>¢½5z·y‡õ<½{RF¤†G®„sïW~yÀ3°SzÒþˆYÚb´#ÀÓšÌ póéZ²lÞaX›–ú[ÕªøÃ@ t®F~üœ|gJAÔºGã-'*©Tß² ‰"ªYzÞ÷ã»5¾Ÿß=…_'ínÉìeuË{ƒÜZhþ4‚ê!j2÷H€’†š³óôa4XZP¤zŠêIe|ùÃ<½ò#_×~Œÿ¿˜û¶·Bx® § ƒ &{Ï RS—án¬&ÃðžšÊ½T¬áèñ˦¶–3£T›oø½à»©üLέɤ²I§~­3„ÚŸÆ®ÕSCf2^Ì×"»‚ B¨†ú]_Â7"õŸPb_®U9(¼Ϻ=óðrãžI(¯pÏ-Cmß ï7ÊP”©3®K%Jžo`ÅIô2±·Ý'|yíD+=Æι§$ôÊ¿;<žw¹´Z˜­u Ø“ãÎóg“åˆß·r븹5š^61¾¬=øB%z?~j5­ì*ñ##;_5tä°¾ó¾{¿ŒðY0 } 1„©ž•¡Âs ±Ýê¿–k Fœˆ+b4fÐ"†é=WvQbjý%ø‚'Ô»¼ †‹ÿÄ›zšÕÖ,ǯ̌rÈ?Æì½+°ÖF¸ÇäƒÄW1 qù2æ!†|•¾S>[mÅ~FÒ†Îï»qå·¯†#Lo+ÝýpüΗ‘3¢eÛóFÇ&¼h®ýHŽfÛ¡ý> $¥’§Åš!Lø›ìÇÚŸðΦ¤á¿³êKÎ>zðKÌ ´›Æ;Y%y!pèY:£_Ï×™9Õ&—%˜Áa”¹œûç•[4&©µá´‚`÷iŽ7‡‰úMK¾ê¡‘±Ý?y7;ijGWKöÓ#®¹=ûÜl4Ó^÷ØÚ7ÐÊr‚TÇãG×¥…ÏM ¿#Vàî·eð™îe·‘c~$ßàLãÞÿ|®EPþŒ(«¶œ¨Uw¥³_ˆÜ ñR˜ǼñAË:'û¡U¹‹Ú`6ÕdgŽ“ kІE_*±ô´úNúÐañåd%#/ž‚±Q±Ï÷É ,âñÉâAfN/>£§blI)¿Ó*è(-¬þ·_þÆ({OŠîâÁʵ¡+éȳß Eòe\S #ó$ô¸ÉÏGePD  ™R$E»qaž‘±×zíì—ýÈfPÛñ&Ý¢vÓ2ßüñé$õU‚œ2_`K4Ç!»AMrh>¼Yâ¥r›‡ß¡ý¶í¢.Àü›±¯z-À¾W-)ÏSÈTI¦´´®…†r´!>5*wþsUI2ÖòÑw œ¨@´š5×ìÄòÚ‚/ñÄÜÁ¯Mˆ–èÆhÿúM°Ÿd䎄ãPH±Á|‡¨j3©®Xcí¯Ñç¾: wqSë™Z‰^Rt1£ÔAÀÚÒžHòOœT3ÔSUÓ×½Íy:?UµLÛ^9„ ¯<§š€âÖ«cq.1C„n°b=÷´Ïò aõoEægI¹°!•ƒm7J >Œß¾1îȶ/6 Ãa¹Mƒõ@Êxá›6'·2Aí\¢rUÝ‹¢Žc´­} ~ÂdeS! ÜþùÊ~òk§“%ù†Ûž¦4bò £$»1’;}4C†‹„YouwæÒªƒw°€Õ‹ù½*Â{n`hFâ ›„žFÙ^3Y}ã«öSÐLÓìÏ®7G w¦?´¡Æã7rUÁçGéËþÉdlž¡Àíd¤7÷C¡½-ž•~õ?.N$0DiÀÏèúNû L©–ßZ… ýú(ANl`D!hö»WÐf TÄO}›…¬‚²Ï@Å% ×IRñM(\ŽDwž_üÝ^Tï82KP J‚"JXÚ~}1ÒãÐQàÀžË20\[h!¨úÇp£d>óf6%)Á9¹v¬+ëgëµ³BB{k ¨¹Ç‚XßF4†3Qñn:Ø‘S,¾VXÍ=ª;Ј\ÇŽ]©WRÈEÞÜ%ÉŒîЉ«|rÓóúŸÙ]býXïŒEmÿq'z–f¸¨N³ß ™>¦RÑñ=ypÊ•žÐöí0ÝÍ_Ù¦¯éå˜)ûTši¾Ztaþuã]Ÿùc7¿Q Ü5FòÜàõRn·šhPÕ/uEK‹;ñíé÷øxÊB¢ü“‚HFúú ØaîÃ(är9x¥N ’É€Ò2ãðܽ•ì;%N!TÐ):¤t?nÄU±0²óÚ4âV{CRÅFà¸ø#˜>¹@ÖÑ"7F•:óhò©zÓÈ[jqX’5p=§YàWâ“g¾m;ìTQÏáÿ1Öl‘.~›d?R¡Î/=m6}¿NdNdñïӱeƒ°ðO;„S|.¬ÌžWOK×ÏHn3n¶Pš‡ŽÎ`LØUH4Ü*t’x ý༻U ÑHÿ«Q‰Ò°ã!µ$Àá%²pËS<%hp>/CŽÖhÁn°]î$gOUpö‹o¼äT||]Û¦®KДÓXÕ÷dÕxË:ž»C³ÌòQ­µ©xjq.É~í˜Þã/ Î¸ÜBkœKÇ'ãZ_¸ =O_‘+¤‹•;iF9}Û%rW”äu²2Ùô‡ 8µI3ý ƒp£'@5ú.ÒÁ€ó’£&Yx7'Z±ˆùÖIßÍÓ¹èÊÒùà¸ø¤¡¸£?ÌM©62R35Ò_̱ÕéJÚTY¨î (”ÜR8H;q®[³W§¹Là&ó¹hÄã/ýûK¾©£¶EÄ|ü¬˜d*Áܯ²æ" ¼È‘¥òøÜ„Q*Y/:M_œÞÔµ€Á,«…37¯¼!ô uP“[§Ðún¥+xž6ýþ™2‰æA·„Œaú¯}Pü×XZçqÁrö@ÝLGžŒå<99|R½Ôu“§Ÿ˜OŒ°v~§‘¥ÍâfA”_1Vv’ëÙD®• _p·$©†ù5ÁP± t¹ì‘†1ÁêØ™ š‡Á©°äëÉášQ“Qé¹ï…í‡TÞ“­“Ëùc}u[]FA|\4û)VAˆô› [(™O‰0Bèg8y&Há™f¼$ÜHˆ…˳"Öy>k]J³LîjÈ3çÂɼ&!8oyMôq;i,B½дÉ¡gºÀˆþOŸµ6’ ¹âv7ò—`x­'±Tçë¯Ê]K]ƒb£ìË%¿­Ú¨ã¨mkJBO0g-#ˆ¯ïjSm®¯vB…üâfŽ};¬œË:ƒ`¶]¹…e1)=D+䲊ŸëÐ餃/T„.u ïCdCܹ?DÊòeãþ`sǨK — ©éþ©•†"G ×Jb‹A &þÂ*µÉ.XÒåTÅåt&ë‡%ÒÞPe”ž§nH†AàÐup‡xR­S»OéUšxʧãvÙ‡ìÌf02qSxÑd>¢PÛ¦T1À¶N£QxoᆂJ½Jwì?­>Ï•b€Ÿ‰¯¦€ÉPâ{‡ß” þ-ôh·Jyv´ˆM¥Î¼F¸ÕZÇš!VIþý[Ü&FðE› (qcõ§$?[4Ûã{ο‚, EogÜZOŸ.Y€Ø?q<ßè+·®‘,!=8X"nñËW¿G`ÖšX}ÛÒhܦŽµºo!îª_Ž ^’ßâj`©Xë&úýÅ ¼7~'3¢]Š4} -ä_„‹áà¹l*D¡¥IÃ8«D˜å»‘ß–©æúñV‹²(tåC&Zû›YFcöíÛ]êT©€täºF˜\òvVf êãMbOýfÐÆ8ÚÉlL=¤k ×Ô‡ªÉÑZ‹f¶ð Dê±—ÅtG³ P½?rKã÷ÂÜ„cÊÂl@ۊ䎕ñŸ %¯õôª ýÐ3·»­Zì´£pf^ÇûžCŽ@œ’'È©‹yúq®)éüʤuó.WBdøA8~:5œE©¦G1æ)Ò ‚ÂYÍû–/þ^A©9±³®Â2°5ìoUÆt=¿ATj ’e›SïbKÜ:l“Ϛɩ£)fyö5g/RPR©ã“ññ„ÏÞÏÕœaœ#I svTl8Æ`C½&ÀÜrj¾Êеvv»psÃÈš¹Sm’¦ÿæÄ"Þ„Šã냓P¬—¢|—ÖyàŒö÷+ðÈÞXÓäcþØ=}d˜²Ýߧ6è Æíq®¹ŽŒ0­'< ¢…IÀ-_ ùí9Ž©½ãf ,=@‡Ë•÷”þecŠBbÓÐ}¥ÉÎäï±ëG)Â;X&“˜Šœ.ø'Y…Cz®±'ÓG¯á±M *L+㯌Øü€¼ÝÂ9rí·½¬f]cÇÎ:$§x´²äëÀ¡³g" ëßåˆÜz1¬»`o}Eh=ÐxŽŒCËKO»¡7N¿žrc³ `÷{žFG|sxÍpãƒã¶ÀÊkœë0ªÍ Y­þÜ €K¿ó-ø¶Ö Oêø—ã–^òþUî¥dÊ̺°{šFk^سŒ—ÉÔŽõSY´©O×÷âÑvÅÔ¬T=­ÓŽ&1Ÿ3²ÏÒ¬æŠ\‰Ç‘t2Vâ¯Kÿdš©e ^ÅõßÊ"%Bn¢Œ]Å,Öóóà˜]_¤,‘zu²sÍ™òxãzgàœ›ÍêùXf—‘pÝ!-“×¹ à×ÑŠ#1´Sw±?ÐgjlCk7Ź¤®Ù5óÁtžn.4®PÚ§ AÝg%§sd²n[êµË…ïN­D-¨‹HÖ6"£Ã•"U˵•@'9õ%íÿʹñ<ƈ ãŒ;PŒ¶*ö2KàÞY(‚â—¬ ž1`)wÁÜ´½:î˜uý+ÝÙ=ØŒÙB³Q™&‡ Þèámf÷2)£¡¡‘¢x4?~r]úg&´’_/Q´½âïÚŽÀ~ïBeˆ_ qѱ ïsÕ »ñöšØ3ã‘Án`®SÉ}d†U)½ŒIÃS…“)ˆÓièrMÎéâDgÉSpL«•Å#*†Ÿmœµ»¶Ô]{£ÔŽvû¢ä¨óE'Qõ˜N§ï1Iˆå_¼RžáU!€GÐ fÀ9)Å:e<àPÐnæÇ ƒ³„‚PŒÔÂ*Y§›5ú²oX¦·@0˜.V9×éi´Aùz%‹Åd€‰rÎd7ÇŒÔ)Í'Þ°tÕÊaa.,‘0.¶ÿÄ.œ›è«Zú'F‰š¬ÛF£ÌxLx'" #¸>G &ßUÝÔ®G×ô&«ø Þ=Îu)¶]¬ŒQòG2eüºÆ$cD{Î9­ç«wgŽ‡ìõÞOJ¨1ËÏʤ® Âê¤TÉŽŠ.”½÷¢ýmFl'ÁÙÆyUú"7I ð„:GÈî’x®*¯x£™±I¿Ö•©ÞÚf>wÜyøTÁ²,öø˜e9½5UQ¾-iÑk#Ïàwìó:‚—TÎc‡>¿žÄ«Ç¼ÜoÁ ¡ì*ÙÍS²mÿ«öÐ_ôø´ÌKñUDž ©¾ìçáœíQY*Û¿mÑ(I;Ç ©šXó-Q7ÂÝÎIz¾7-ÚD2?XzÔ ÔWŽmÔafS-…Åí]Tx•¤›<âÝÉ£jvQ½ºu}#oñ;Kè³@‚%âkm*Á¯%ho&v°ý±1ikI(H|‘èo¡AàÿàK¾šºW`þHTËÔô'­ª0aã„rÊ"/±¯m³½WÙ e«kub·²,…𹵫Oâ#†ŒìŒPЇoÍ á±á ®H³©n=ÞõÌŸôÑí‡*›i,ÏïhRtGåæêÈ$ÚÀÄÅÁËÁJž¥wþ÷)µÝ TQŽB%VJp8E>½I›(ÇbJ³K[Ê+¨­A~ÒxÝa½ ê½ÎÇC%ÓW>tEŽ9Ý÷¶ByŸ$áÃä[ÖºCäs?$›pfŠ×ÐT®®í¿«‹< — à9@@F¨yëÆhÄ9ËŸ <ÓnxϦí0Â*T0Lt|’S榭ª|¡¯e4·.çv7×¹¥\yظtw‡KJAµjQkÚ…GZ§äVav …}¨°GÎ0ô»y5Wçô+¤Ö>õÄî†bŽgxA"\­(¡ÔäàRVþO.çà@LÅ}¯HÛ—£Û˸Øƾ ~“ÃeȆ¼×†8pùYÀ‰”év“ð’wåøÙˆ«ì2øXø*ó¯Ùu¬bBó°qO®2M^MvµÉ¨dÎIfÔ#÷±Å÷mð¢{´c-TÀ¼hê-k¾ð7 )YdĬXÔb÷;ݳ¼sSå_>0ù0ùhY¡OÝN- Ž±ru¯«Äį„-±J:"±³¨‹wü›'H”^…7ù®ó’mÚ§œ÷[‡BrœëwRCÅtW¦[–ÏxkmÎSM¹u½\wAF#þuCùÉäÖx—šäþ rêe‡Ê*ZßjîO¬ÊÏ'ªz®É¬ TJfö¸ÇŒX¨ú{ð„ƒT¦?s¶ÚI0ÇòÞb¿‰³I?»¼Õ¾lªñ¸´Ó\JÅñ¸°OFX·ØwáËû™kÑ*Œ„=¦Ï²Þ0²÷q(ŽÛ¤(,U*•ã‚q¶¢8€¬^ƒbaÞÉyÎ#Ä)‘„ÔÙËþAørp>ÜJ çC­oWŠSï–KÂåÊx¤Ê›øqKφˈ³h~yì‰Î,s5¥ÛîìÔ8 ‘–a²j8ºÎsüöõÙ5~œÉ¸ÔgD•nkÉaÓ%3ûtì•aéÏKµSEʲ”¿‰b4¯Wó3®um¹$¦X¾EáÇíaÎ{—ŠE™ àºëͺHÕ"“ªa wàç>LÛ®-¶‘& õGÒ-SâG1ö]Õ>3sôècÒ.mL¦óÀo]f0UI$CŽ)é©Æ®eÙUMç¥:ñHé¨Yz×ÿåVk1¾0•}örN} $Øï7‰DC;Üíø¶¬ŠRÊh*fâmž˜Å³¼¸u(r`àúà´ŽfûnØ‚ÁàE)ÔCP¶[msú@)k;ì62-5(«ÆgÌeÙ¹¦ì­ævä6³`8Õf¿!ï4"”Ê|̦ Ï„tô¡¶@n2fµ Q9)tyÓƒ?Á¾†Ó?²GÂðÔ>úèïHi†®KÄðÉhš³ß¡,.ú)ö•&ï&· ªNï³N­Ô†½›œxóà T¾Ë 5:«8-Ã#Ó5?¿Á-ëZ:ú8hAÕ|?ex²Œé|œ¯¬E|K´Û¬ÂÚ#ÿïU¾:ÇÿŒ,#y§"qý‡ºØ«ËdÞÀܳÑqâg%¤·Ÿ*Õƒ³B5‚|»=p¹¸qùªAý¯*jPXל—zÄþDbrG[®Ã¤ø9Ä5‡Š²–W—`“jY—sƒùz˜n¸wœHÌl‰2žçu¿,¢P6·F[­;€!D¡*¤-vÈ&³ÄB"üU’ð½eÎL[£ªÝ•Ž|vÙW 8H@lÆ»ñ© à%Ùõ´¤GßèþBO±ÆWk–—E~olB<“ïdLX\Ä\6­%’¨«ŠÝy¸ÿh6ð1Ç_š¸Z<.p;5÷'ƒ¤TJ V.]^zîg ¥<ÏOû.Š•FNLþ;ÄuˆþWʽ¾ ª¯R"z„+žÕPS‡ÎŠ=}¦œ!!²ª»Xã^ Mím.5±Â½W¢ë7¯R‚5k‹&ÄOMùvÊzíø¹Gë ð§½þñÕ½Lg¡Ú¶*+ÅY(ã <Ç-÷&šLÏ…Ã1R¦­gO¥ex«öè+¹{ ª}D÷vP·ÞÁe»’?‡[˜½ sEÑ%W!0m#°½B´}ø_¦×\¹YÅæH‚Z翦Œe2–ÿIÍŒ28XÞ]H@¯L¬ Ï_}Lƒ ÀãôÊ2æ%gÇ_Ó"š6M.«À­AIl©GÐ ='¦1pöFÙµQWš¾Hí©"Wn" ªó ÙäÍ´^Úƒ~èGû¡nAaÍ·&,ÉG¹£ÓãÍ“š­_@»69CÒ!:ýpêä–žic”½¤wÕÛí×v+<“2X»ý>WÑÀ}½Yrª/ò@ ÃHJP ›o)^ò®vpL{åè‘|mð:mbQ‰159v>°BAõÁW˜Å¼¦YO3#µË¨÷…–†dÒ 4Þ"?lÞùþm$áÑßÖeK5¯e~äJðMÛ³P€ÿI@.Uï8«8˜9¹‡Èm)ç&ÅٜȻ•0%Î'Eú8PV¡¼|Ý{r“¼†ý~Ôç©„%cJÙ’ƒïs ëé'³{C¯fCÇÑõ±'öO3Ü4AÉv]íd¯¢c³ÿb͹”ýrÈRnß¾îq$矯n ú/ÏiüÕ|0ú–Íbçœ9[2>¼Mi üBDúsªo#ÂRÓ…¿¨ó^²®½ÅœèˆútÙ) QUôaqQšx9gúUO±Z,< J;wä Vû(£«Û{§µÕ‡NbŸpª ±âŠ:¥ýþdÅ^8øÌάR%Êê·b³ÔŠ“q°!X‚‘énèß‹û^¿3T¹VôáÕ6œÀ¼Ôö2‘oÅ JÆYXLqeK²~°¸vw®Aß±ÅôöÏ?^‰ñöÆhijº Š¡mh†"­/÷à`ìŒ7ã•;Ú®²Z–Öož¾ Ý·¤6ÛAÈ"[ÛŽ›ÆkŠ‰ÈdŠ9š³M—Nµ ßôLäýPÐŽ½ù2+%Ø.¡áš7YDíãA`]PËuâòkY 0U&ÇQ »¨‹hzô3ô*ݳuº4"G ‰—«®é¾³¥$d Ït‘ZÒv~߇ÀáyU‹MêwÉ(q*«e¦íûTúk^¾ž¤T´€¸ƒ äÔb8¶GÛ˜ ׇ^IµÉBÉ”ZgóŽºK­Ü ÝYþEÓß—Ù‹ˆ+9™ŠF3è9>kbõdsseaËÑÄ×7ß´‡î¿ïVŸŒÔGê_iÜ0YY«![Ó£+Kù@¹rm·ÆÓˆJ0‘[àç*ÞJ•ú2_OEœÍ#gâî–Í°Ç~H»-Õu‘"lHÀªŠ-¦)n£ÕÄ[w´;«º´Óyins¸KzÉ\š“»I5f•k2ÝO!ÀŠçö ãÙ ñBRZç‹Z€â:RÈgƒI¬~OœŽq¥$œÓ|B’!r À$*‘WÀÛ (PÅ6TH\¼ZA• 1ì+"¼JH*ƒ›]vÈÖè´dà–,›¢$3 jd˜)úX5*œm Àh¾ü::OqëGc$Uú\E™¯Ä1âÖ_Á»“ xÂÙ endstream endobj 2120 0 obj << /Length1 1395 /Length2 6101 /Length3 0 /Length 7050 /Filter /FlateDecode >> stream xÚVTì»'¤CJRp4H  éF¤ 6`°Ñ ©„tƒˆHƒ4Òt Ò!%J+]wúÕýþ÷žsïÙ9Ûûôó{Ÿßóž±3ëèóÉÛ"¬ÁÊ8ŠO( x¤¥'…ù@!Bvv(Ê ü‡–Ý슄"à’ÿÍþÈ B¡uŠ ÚM ¨»9…‚¢’‚b’@ @”øËá* P¹CmZüuŒ$d„pör…ÚCPè*\6ÜA 1ÞßáyØj‚´@(†®hrè#l `”׿RpICP(gI~ ÉpµÈÍ ð€¢ =0ìê¶ü‚ x ‚ã'd@ È?Ôú;”È @+œ 6`8à·»еújšmg0ügÍ?x^ @_ðïtFÿJ…ÿÙØ `Î ¸n°ƒ:ÚÊšü(O/·ýårB"Ðñ wÔ dvøÝ8 ,¯ ¡ñý‰iã uF!ù‘P§_~¥A_²ÜöÃQHÂ_ý)B]Á6è[÷ø=VG8ÂîóÇÙ ·µûÁÖÍYÀuq«)þéVþ£³£"@ PL\v€=m ¿’x9ƒ©Ñýûù8#œvh`?¨ýC胹ƒ(W7°ŸÏ7ü["ØBmPk°=NøOv´l÷‡Œž¼+Ô` DOüõùûdŽæ–-îäõûïá 詪j«ñüü·IAá ðáø„D€h®Äпgùÿ_Øku@Ð?{þ“Q n‡Hü}wÁpÿ“\® 7àß#Ð<¸þ¡½Phƒþü“ÿwÈÿÆù_YþÚÿg?ÊnNN¿­\¿Ìÿà ‚A¼þ´£Yì†Bo„½ðÿt}þc‰µÀ¶P7ØZÕP ôfÈÃíþ¾D(Rê ¶Õ¢l 诠Ó;Aá`úë™ Çüz×lÑO =©ß&0z•þ]R nƒ°ýµsB"¢«+È‹=x´$ðD/§-Øó7¯üp @ÃóØ!\ Íó@ýXýRþ–ÑuPˆßò¿ Ù¸¹º¢—ï7Ð]ü%ÿÞt0ØlC8;…°‘ q¨ i>«gðà[‘ùľþ$•›ÏgÖµÅí‚/‰»<#hÉõD>©¿“laM‰ëXnîÞ•ÏvÃ{¼ÐÆݦKߟ–qzcëM„3©{Fs·å«» îòÈ}õ½rñ5 tÄnÀlSgÏrq'Õɦ<óèRñ¬î.žz1µ®ûµ\Tƒègñ8_”a¤Y`Áû[ëד´,¸(>Füû?¶ÓãSrÈm¸•òvpöª¨/öbÜ4£¨¨0}*¬ªØù%Gt…ùá´QÏ7z1qÒV¢Ä¶§Ãö:Û¶i’~bÞüÌy% «OâR¹ENŸ«1¢êw7x31Ò‡ó »?ónònç¼’§êKHú;.%vü'bß)ž Ú›xz¤å¿½ˆ÷“—l^NC:bÄsTÜXö >Xû^plm.Ÿ- ×msp 4’æ›…ÐOáÖ?dgßq[(ÿ¤Ç°)rLoHöÊ”¼Hëêê}AS|°Æ)þ›'tâ1®8±Ï¬‹òôøùŠ “¦û¥°*QˆÏPª(ù¾•gy¶=åë?Ž÷Ýs±ª¢nwš“y× ¯Krô4vPÑ8êþòÙØEÕÌZ>¹l@90׳"f1®6Ý#ÃÏ,iŒvªHZ.ඣg‡-×J´o{» †7Øçë§ÁlÞ7;‘×?Ô¦ØöÍkïi¤ÆŽFc™2ŠÛ×æ’ƒ“·¥„â-&ñæ•,u2®«dŽ”¾ýȱv+™W;ªyäθx›ÆØUS#sŠ±IÔWLð@ï`¿¥xR:æU„ÂÃq´à«²pY‹{°)ºÊÂJ˜½‡:(ÑÚê1Á2AH[»˜¥JàÙ¤®–íe­È0äêøߧB¾~iƘcõ £¸éêø¾C#cDáþ‚Åî„}ñm¯7öåØ)x£bÌV«ƒ…Ç"nkó,ãÍðpb¶¨<È¿÷Š/ÿ:¦¸­¹S<ŠÆ3«»Þ~Òó¹….§ðfï«ŸTçôÒgŠÏjË—Ä6–Œ;^“ÌXø±Õ™K|3ãÛƒt ÕºÍÛØ¥œŠÜtkÛįwãËÀYO/4õØh4±ÔA”:ÞfM?%/x¦HèVu’àôñ?¡9¾nÃ"\þ–ïÈ|ë&î@h °ú>ç‹(½`áÚåaÑWŠŽÎèá–½ÇøêVŽùƒ]»¿¢ÄüãzF3Ëlç#äyd…)Leua0…|vŽ7êÁó¯É ½]zÀ’%ÚO¸Qµ¹x}‘kÁÉÅ3ìßÎéí‡I•å[(sÓI¶6â5ýšÂ‚aÄ\æjÛL£w´Èlsß?¾p,à~Ù*·ÖVµo§_%òê&Iø‘ØÞºÔæ<7Îéd”àpI‹©Zkr•ç3Í+3ê;sŸq X?M_ÐÐݽ&åµu%scÖfxÃÙ–;T™®ŠÎ‹l9ðE^VØÓ@r’ÖUÚN“ɘûé8o»œDÈï=ÆrÁk“P$ƒ+;ëoB‰õ{f»†gwŒnÅ,Kl¢ÿ %óŸÛçvX©¦â¿Ž­jÁÛ$¼½fZƒ(uºÝ߈ð¼9s'^Œé ½ò4Ð7Ïdß¡eÕ|mÆúÂ2ÊñÑ„©gˆñmQ‡¤ 2w ¦A>HÔèÙðæ|Ùœ†#ñËü¤ÏÅ.î€ËÍt/LnjNƒäæd:gyû„ëËÕ 3óFÊDhgC\…4ðÛrWóëN,£Ž› µøádúXÁ>vÌÌøIwÞ–µ;ŠGõ‚cÙÌaoMhÀìÖâ¶sC]h‘Ú? çÜ Áw y¶¥vOöÉ^0ƒgãQÆq‰Æ#×WO²…üq^ت4¸®ƵÆ-‰Ôœ¿ÜsXx•9®¨^¥Ýë x?X.•«®ÖÍ!RñPAèM§ÖbˆT7꜓r™ÄŸã¦I_ȾAO$Âú¼¶ãº¨°ÐÖV\DöÑźÒU­2¥"£«ë½µ^Ü©~~L#šÇqT”é"y(–Ä[ïNQà ñˆ IÆþâM°pçÎiº“¤ ݽí’jóZ‡ò5e*ƒÅm’8}CV ­ ñ럷U…>]<öˆ;/yFŽ -ý‘9Qý¤A$’¬Ðß”¹É;'Vg®Žœ;Žœ{­Ã²™Ú1î:òÈ7½´”Ì&Ãâéã9<þ§göÚsÅ1 ä0&UómÉ£U±\%w»V5›«òétbZûfæËêÓš̱…šǧŠ×V4`OL‰F¬AX|ÄãÒq;·ìg9:Ód!çú\§ÐЭz¥‚8øi'-rÆ4*±bÔnü„Ê^Óîà¨({ñõùµóšÉ§ÐÉ;µ/HMÃ-0Í `û–ç˜`†DZ”uJ |ÄÛ”•Eæ•dH`h\vó%CñÙ’ð+^R)¹ÝÏ·GY=æs¶IΩ̊I·–zô$Pqêî Âz¿“8{…dp'¶3–[[侫Ûì-'¤ýYÒ Nºœ#§ Ð¶ÉùáX¦Ï*ƒ–%ÐïðÐÛ”Ù#ךÝÒ=1¢³+¤„cáÂ<ˇ@¡šk¨\û§4Nð‡‹ƒë0Á õû;]Üš‡3)LeùõÉRVäœÔaÂ]w&'o©Ý• yà;؆ϕEŸ9{²9ª7hÀû‚¸ÝYL?0‘‘‰1ª3CN€‰{Sß’¦7{»wWyÜ n°6&ÛLû‚²²¤ïªúß?lG~c!o+÷¹!ðN/§É¨×á)»õ !¥4Zq‡.Ù,¼mN….¡·[¬quãs©üƒC.Õ).Æõº*ÇŸ‘IDöóÎw®QÂt"Äw/g!ͪÝ=&oÉë¬ÓJ˜f‰/δ™03ÝíÝgܕþ_[Ö‹§ˆd{¥ˆ ‹)?tÞÙ»¡uÞÓŠÏz}d˜i#0¾+ÚÿÃiOã:ÆeççÌGµžc†ŠE:ØK¾TéÓí c¶ùKÜð UëÉ ÔwDÁT`Ú|VÖ³*—ÈpÖC¼š³•,Øü!\²œáçc6N½îßojæÇiŠ•r暦²Q¼úÝù X;&Se¤êW\ØÒ½Û;êÕøáOÆ·ÂsSKÔžPîíz~Q±ÄɼFPØ=·]2° 9´Ÿ¼ƒgGÿm3Ÿ(YvÀ/NîcxÌåÉàÙ’(GFºgÖ1éR/$’Pý¸Y=“XA;Ì‹a[[Ë_è:Áž©›¿…0Šèê|Y¹KU¿QÝ©æ ’—eUŽ”zúÄdõõ²,߀]ºnÜ`_„þ,¶ñí¶Ç4!Ž|Áb³¦Å6Ýõ\þôø|A°«N"ϲ%`oΉËÔcP™Hw‘.ƒóöÖ1>ÂéE}ý¦Ùžª~ ×°u7ªŠV’8tþV÷sO‚9ßðpBJŒÙÀ¢8ͤic©JFìva7g‰iÖ‹_ß‹»¤ôäL>¥v9—Õh…ËÒ™cðº¯äJug9nû›˜¿ˆÆåå;já–…gë×å5œŠñµ;6ÁªÉpˆ5nèy…ýŸµÛèÈž:Å,(¾%î4æ¤ Ò/=Û z“í³T1ì°&¦›Ø0é¯!³nÿjY©B†j_wÂì±uÅý@[æÝHÎÎTã£ì½K¥z‡^5£yv ñ¢Ñ8µNÙÈŠÙ)Söø ¾4ÞhÏ2†ÕÆnÍm¿óŠºýÌ¢)è°H–»m§¿Y„Ò)ðÒZXè¯{ž×8óÚa/(¿ê ÅÚ]er7ÓBÆù3)½ÎÚâ2­B­ˆ¯¢YŠøäÔú`‰Ã?TAý¤í>»xThå`Äììš~KdßTçG¤~Ä´îÝ£FˆÈɉ8¡<235#ß©pq9kɽ7§"¯)Œ½Ö©9¡?PÈ÷Œ¾Òtëâ9p±3“® >zAëH''ÎòÌe&ÂÁñÖèÇK1Gp¾ÂAn»ü9]©÷UgTÏú£|cepJ ƒ«ªý¸ýÌeqÞû+¶1–ÕÀòüúa…J6÷- ¯j{\– ãùÀS¼À€©èú¨!wC­‘ulb|»{Ûª‘ØòoNjR:ë?rHïÆ„Ô¶¿ Ácø@·‚3=¸?(ýªáKÁAÛ…õž.aÜý×}Ÿ[ÈI¶â,—¾äá©I–Yñá¦ÌR5‡ ±^ §ëä\ø5þpG¥GæZ£sðC’t™ d‘È—cñ,ÄdÏ…KøÜ9³R‚›“žH,8 fñ(\ÀqƒŒ?àÌŽ|)ŒOY¾VûQœžg¦û;ROÍ}Ô‰çîõÇ¢.=DÆj+˜Wí*jje­ô0»@ç¼Kmy–ýÛ¢iÌß^³´>h*û°-iÎon°L'|ôÎ;Ä~áúJ,_¾ÊèÃاSbg¦C ÷û\†.IèPÍÈÀQ«‹Ìý*…cd!ÇýÙ ³¤•ðÒݤªo>½Où/oá¦ÉX¾ØÀ›¨®ª\{Ó„á1pm0Þ¿Š((¬,f¦9-:£m 'W{TçÛm{NåZé| ŠonKfWçÿ0ÍWƒó–¨‡CË©)»ò'­†%[Ü(“Ó.7Y=MlqÈì½{¾6å*³…l$zæ4œö*™×-D~I"Mß²‡Šænw¼kù²Øß´ÉC§D½¸“*²Ïž_?+ 3„©Ô^Ž¥Æ?Jþ\X/ÚS[Çؤ@4eù°±©¤û3ð*­7ït)3ßôag/ꛊéã£æÃUçÁu!vœ—%/v.RæÙÆW1úbm3’= ”¾~8¥Rß–ñOè}áþŠ4Ô^þZzº-Õý›µ{@½õS0<È»¾\ûWjÒ~zÕ>ZøR2qÞ4€ç·Õ\æ_­4÷†Ê™Œê¿F…ƒñ=—„Îú[·GZò§oçW×ÈÍqãæßlá^:eY´s"®sfž,¥JêIGôBéËʼ ÄŽµ¹˜c“¸á…ð’gWQ¾¾§Ùy£.9SµœˆòVå =ƒ­Ì‘. À~NµG¤gZ{ekªA«nõyëX CAD@ÙPËŠ3oéqàldöÕÉŸµñëTÝ*ý·`/ƒA1«à´QözŠÞž„?&Eˆ´nº,6^ÎëZ±Ž4;-–éBO5£\ˆ\~7êyÙ»ð ØÈ0?îhɬ¥í…4¦Ç¶'÷þáÞ}ð-Ç•t¾¦“,ªâ0j.Q›ê¢#I.oÇ4ö÷»Íq8^nˆ0¾-ÿ$nœ:©÷¤‚íÍÚ¤Ã=¸œ³âƒ­¯Ú`òÄíªÔÖ…ötŸ°«O¸ñDAµ çv›&G]—S?< i¬%b,iDW?¶‰q4ˆ‰>eò™©ýŒ¼ÁR÷Áiê]Ÿû} 6IØ|Ào²aë;v{²XÓ¾CoršãžncU?úñhsËÍ 7ΟL?£\-®oò±SI/yÖÔ% ¢ XÕ}êa~T.‹_ñ1ô°ñÞAî²Ý;ónË$§!“Bîsæ…ñz%Ò@‰”úö¯:ŸUì÷sC[p­ð²>êú| ¤¤§‚g…GOf*ÏùX´ 5Ûá¾XÝÍú* âîØgœn0”ûêxgšgõî¨R@ßÍ®åéž/%ײcþ¡ÜÕšAÕj’[»òXÑ]~jzê˜t‚Ë#Áàø1”åQô"U,,§éV‡ÛDü©:kЮÂÉò6Ž˜$õåG7¥%Ý*¥‘À‚ ³;M¼âóiRÊ‚‚»bëÑ:ƒ˜kéK ðر¡Ó®“ÈÞ"|úJµ‰Íç㟕¾ up‡-a˜Ìf2 EB¦ßáfÓ´:Ý`N±³ »•Màï³PÛrµ²ñ\æD–âš »CÌm#Gsà!ä7'|m{ªY>i¦Ü걉Q¯Šg¹Ì”‘èÙ{…;Á7šê‹Ô÷íyŒ©£VNSöŠâ(ñó `qAG¤Õf5âû°Îî?>\2‰þJŽ+Ê:„ÄJš?`hfjÄÖÍnuÜðŒ5EA1þ-÷&„yó¡£(Í£ hM%PŠ Œè›&Vºò)èÑúŒœ/R÷Š;ÕUF,:*|Àë·jJ)O~î{X^¾ëZ…yºÇ=²0µM"eð‘¶c–asb¤tÓñÛL5hQI§ñ-ë—ðï×»7ÊësSÊÂÌêÎê%ø5Éw3È^wãT*½?º±Vú ûËn¨8+ëmßl%ʧò˜1pg˜§Zªk"%“ˆÂyåªÉ:±‰“sût[Y¯ô ¤„ϸ”ÙÃðêÌòŠ?"åŒnÞ Ïø/&ZEZ+Zà‡Q ¾qã7˜Žþ/-óÅ— endstream endobj 2122 0 obj << /Length1 1544 /Length2 7225 /Length3 0 /Length 8249 /Filter /FlateDecode >> stream xÚvTìû6%1ZVFIÃFƒtw·ÔŒØ`Œ•‰$¥‘îFJB é–iþÓ·þ¿ß÷ó}gçlÏsÝõ\wq±š*9¡aê($F,’ªèKA Q!HÀÅeŠÀ¸ÃþD\æ0´7…”ù_r4 ‚Ábª VM…jû¸Á¢@°„ XRŠ€@Ò+¢Ð2@Uˆ/ ¨'ÔF!aÞ.”gáì‚ÁFùûäòÁÒÒ’¿ÍJ04 Aõ ˜6"â4AA0LÀ¸à‘uÁ`Äö›˜€ hê‚ðþ6AÁ1~4 ˆÜPÒkàƒt‚¡ØØ@-] ' ù§²îŸ À¿R ÿq÷—õ/GäocŠòð„ Hg á¨ë aü1@Òé—"ÄÝ…µ‡øBîG¬Âï‡C€êJF@–ß_켡h„'Æ[Èáþ‹¡ð/7Ø$«!TP0$Æðë}ª4 ŠÍz€ðﲺ!Q~È ?ÏpÒ þ‹‚“§°áåÓRýK þÅœa 8’”a^@˜?ÔEø—sÓOØo!øŒ}H'ÊÇR€… à0ì Èâ bÐ>° ÿ-øÏ :!  #Ìüë ÃàÞ±•G#ü÷AØÆA¿>ÿœl±½å„Bºü«þ»¸Â†V–*zü¿ ÿ#RVFùƒE$‚"â ,"”ÄBþÓË?üÿæþ5„ þzè_ZH8 (ý'lîþ¦áûWOðü5.¼ÀÿŒ Âö1 ÈóoÛÛ€ÄAPìøÿ»ù›üßzþ——ÿGÛÿ÷{Ô}ÜÝKy~‰ÿ)Äáð—ÛÅ>ìDè¡°süoU ØŸC¬sBøxü·T ÁN†ÒÙýŸ$"¼Õþ0'Cêògý]¬{wfˆòFüZ3@A0ô_2ì¬AÝ°«Ä[©ß"v”þ3¤Šrú5s"â@ ` ½‰ƒÀØát‚ùÿîk °…Áš±ôB€pð«ž` 0ÌË›u,ü ÃØÖþÁ(ô?€˜4P»àþ½ƒ€ÂžØµ‚t‡Á1ÿ¢à¿Ð?ëÿŒuïéîãý¯{¬Æ û_±þ0~¨M¤€Â0ôŸÀ°‡ú ±0¿»›š¿ï¿× 惦&PÐ{]+7”+1û ®Ér­Z¼à šB7ûœQ%ó¾ËŸG+%÷uP}]Vã9Rœf½ Úª¯"ŠlH2j<¾°O4þ´Ú˜¡û0üzK©²û6 ‹ ©âZð¥W°ùC7üzÜVm®l/) Ã\š¿. ÿÊî’™«Fkï$tH/J> >3‹µyX8Æ•ãøjœ#x›˜ïæ¾?åØÑñèͬákVíD~@Èö3Ñü 둸ÓñÀÙ7¦"ÞmŒœŒÖ ·ñn~üt7Hy#E›þKPQþ\Þ˜OA´è×6[Љ%IÙD¢Ô{…^|¶5çÅŸÛ­¤uH¨²é”×k—¥Ê<ñÄ×Ö/@GY½¨vƒ&n—ܬ©`o½x-©rR•vkI":L›ƒINîˆû9ë>úé´/«ü4«Æ!cJ[„09«¹:…^Uh7\Øã3_â)O&;¢¸=  &N#OÖwAâ”$ mWZý„™‘@`Ø6Ü.¸NOÏô™™ä š?ÊÍ¥xÏÕÀo£ù«×6ɾÛ]¤5{…°´W}QÐ;ìIH…ã‚¡q‹¨$Þ̤”ñ§¬4mLC·ñ Hž¹u¦¬Å$;ì<³~[û¼Sø0Þä&9i ß Br5áh0ÑÊóU‹ngTKânâå8SY…c;BÊæ}ÒRnD‰8{Ü× Æ«ž“xJã[×ü2ÌÌ|wµˆn2¡Íá–Xçàº7¾¿âЖ^Ý®VjÆ‹"ÞýVm±OꪬÚ>ÓV÷È°Ï4v’°¡à3Äžso“ê-yŸªmĤhœñqª2£µ <ÀGŒl#³È¿c¡ aqþ“þ©ìUV,x’ùóRŽå¢[²$ý¨Vz½<óÅFÈèÀ$÷̪iÚ5oqè-N-®¡k¯…? µ8*õn& o˜û» Qtøwo+Î5ÍdÐXÏkK¸„í½]š–sQš.ŸúH(CÕÊ‘€rœðm˜£ÁZl.Y)svp;rºoƒ’ú6nA ß)u3åÂqäæOjœÀ1ßf5^ru‚¿´): ©†X¢[’UË ¬ŠÔú„†¾Êm%ñpÙ5 ‹˜·ÕÑóÖµ0½úö¹»g…g~Çc¯ÞÀHtã&‹qœˆÁa.ºÖT×ãP÷xň&N÷~–pò@I½ÛïYóßçgì_L’ðÛmÖMø*U6PjWþ¬%¾,Û;–µZþJx}?,ª²^õžËûO«t”´n‰"xôüpÍ/gÔn3‰ì”v²+Ò÷9þá'1Á¼pXÇ÷¦ÑÆ¿‡ W¸‡t¯7v9Rª²#¤K˜îª”nð¾4   •Œ ùjWíáì¾›1Ù#r“óüM¹¨‡D¡¬¹Ž;>ëP y|Àϲ×ìjùR”T'9kÝ+]µ]-ÊQ®ÍëM1&/lì(™®_žóçÛfrBº-¹ U\ŸOk)ÁøÒVœŒŽç%5_†F’_ìQPf}ÙbÑ\sþ¿ ìä—+¡Ø»~ˆYø˜a›$ºyEšf2‚:ÏQŒ%ùqŸBcÖÖ?¯W\ºâeM³rHtž}ÊÞß œ¦qê&¡¶·ÍüKÊ5M¶š±0Ùí'HÜm³dÚC/G¿J>‘_)¥d»K¬Ñ­€Î· –H/¢HëÕå…ùu0$~í|¡ÌSN€®Ïdf‹>i÷¤§.°U½‘ù0êîtæŽ(Ý!šq òµ°SÆ)æ~5ËsIÃ`•Óë‚MXtoÄ›,*àøÜ<Å™î"»ÄÛNµîf-¯IŽŠG$"«–Çû妚릳{^ðn%U{¿“Ûð° žÜêCín§á¢ßË:I¨x¾µÀÃ’šàMîOÐx&q+iŠå=ã™^äCî›|߈k¸ñK º%VVV·­JÙõA™åi¬iਠtµã¨m¡ c}1Æ·ù5Z?šôzwQÞ)U ~kÖÌÚ¢N c%p½W5Oœväy¹3ˆÙ ˜%ašGíPÊŸpƒ ¯ ö©ïjMŽÛÚK÷UÓJ…Rv1Ÿ´¨_0¨6ïâFÛ>|ï>[7ÿV6žT:ç*i±‡ôµý8b âØÝ¡_+d|xv p,³Îâh:sÿ;/õ®æÛ&¢1%K¼iµ­Æç]×…?Íœ³dݪ½ qt;+•K×&úù\ߌ6P»PLhÅr“qÞhQœ–ñŠöJå*ñt*‡ 3“ÛÀ7W}´¡wR›)t²V÷_º<æëoë÷\œlg½{’RÛ{ÏFjOÜ'twõ¾ +¿†J˜Ìa…:?»>`d”éÁV¤õ ÆT%Œ¢Épðöñp–Ç ¿\y‰y&¡E¶áÈ‚èAð9^¶§Gˆ½1¿ÅÌIr½æK cï}sË̽d¡Ló;õþ–ôxM@½Uhv^|ººydœëV@Ö£'z F¡crq†{ªÃî8+‡ºÔ¦ž=V<µ—É£ì¡sF¯dvçq\“6—‹æeÆÏ€C¥™ÙÑT6â2Ò–fŸ€®1xÆéóζ{x^Õ{ÖuE‰Ëa/Û££Ù*ø¯6öØ3£‰·*\9ø{–zzmÒ&‚’Ÿ?H^/I¶i¦48)`šñSƒýáIl~þã<ø.øBËÖÚéÍz¨ºÕ²æd®Í©X¼÷ôNs},)ß·"¦–ô¿}ºK*j2,Êa¶ƒÜq PÙÆ–è ¾—/Aªz™ýmK¥”õã§âÊvxT!±òc£"É^k……¼ÊåÅ2³gŸ\¡µúBû¾²rÊ{¯ÍíåÞ•[)6ô‰ŸoPÕi5ÇMâÉÓ| ¸±^Q©šÝ~Ī$"צ|ª°›×ÔÎnE]hw¶ï­9ÿé%¤ D¢>§šJÜC(XÆ8~{4{‚ù Ël»ÂgÀ›¾®Õ‡nÝMÒu² 9‰°ø‘N-´%V÷Hh·ÑŽk/äv%Ý€êTðžµBåÎcÉÇqà}Ü·d|䪲†ËàWÛÓèÒ@íš“äìYÂoÈßþN«Ê”¥`h½‡Aå þЉINÐᆖÄÕÞÆ9š#ùÊ2™ýÔƒÖ„Pž¶]ðUì$GpbÙ½CK÷Áø«Ó¹O“ÕŠŠ¨š!ä‰ Ë~“ÐX·ùĬ:ÿ€Ÿ(ÍlVê ¯±%%Çíº‚Ó\é|ï@±ï–ÑìŽK”~øºA™Pœõ0ý·88‹®ÄÜéb³bÆ1õÇ,§"Ì_L§‘¡»B;¥€[˜“ƒKì»ïÐì3Ì'Ô]¹ ó¼U¿ö —¥ø2/X¬pDOÙ…œqB7EM”'ß° ¦ zÏI5‘ð-ÏF§$#^ÔMmªJÖ÷Ñ©^4ž–ýàX}w4Ó†R;×3‚<‹‘%ŸªÙ<ìš*;矼-™ÒoOÚhô“8ð‘qW4ù0B¥üG†\pv6’_Äñ–ygz‹ëá^5­ñ¾ÏþK»ízÛ*IŽˆ·­‘ûoþ’´«µ&jtâåyÉ™“#{fý2 ¨·^d•T6”Wý·¾¯²fyÅàÐÓÎãôcdn¹à.Wwô . ø¾“|J©æˆÞ½Ö`-N+Çt$ß™~šJ”¶ß=‘A{’Çq«ÑÅ䦱R嶼H×á#bƒqwJv¯9ÊOªß ""|ÜŽðÅìCXgz29ŠÉpòô[£Î@ú¥DáTö*«o¦”e—äJ·t (ÀÂgf¶¶È±$í·_¥…"{ŒóŠÀD®DßÝ‘©º•?Îà§Ãta%á o'BK|c´ß»ú©3€¥÷{uì,$u¯qyª?·ºinH¡»Ðo«ÇÒõÚ?)Ýü´ÌFCÖü8ìÜéì¬ýìøýFO IÍù’¹Ê´Þ¾T3¢0ø馮µÊÖCÇo TCñ ¼…>W°¼Ù¢ >Ù\*ÂîfÂë;!ÑOeXWÕ‚)Þž$ÛJÖ¤.#7ž\&¯µRÈQIDoACïp8¼%@é9Ôi-¬ü¡¡Ãí^ÐzfÜÁ§Æ­"S*,%oÊóYé1«XÓ>á*yîMËëùyøçœÿþñµ¦½ªñ—Ǻçxѱ‘3ÔKÅÅ uoÇŠ3µ /{:Ý—¸ü)hª-|Ö?èâµ™¦f§DIB²Å«Ÿ{¯¢å–P'ŸâèꊠÏ2¢!W€5´¡•WMímš8ó­¬IQ3XA¦‘„ëD>T»ÑlÕ䉘«Vºjß®| L\ÖI)„Q*ŽÈnR_{(ïu>~¹ÔQw9Újµ›Jß&9ngŠìù®çÐ2e)0ÕwŒÂ ,xIB6£oQtõõœBÅd?×Óqb`´µ÷Ú¦w8(ê´›AD<ŸgzWÛxOÅç.É¥°ê3ŽrXËPRðÀyÈíóÇ›K£‚¨W éCV¦)*¶kˆ±á³ßÎ:ŒvÕ‘F„^•;ˆœ XÂ{¸‹¯‹o=v(ŒÜ¹P¬´P™øé-b¥¹@õJ,å‡Jé®çÜp‡ª;¨ÆÙF…dÞZ²þ™XÁû&^Ê3¼à~$…ßfÝ O){ãý»¥>¹2}|ø¯Ð/}vIbä^U³£U”‘ÞÎãÌÇ÷3GO†‘Å%jÓ²ŽC¥ÒS‘KÌü7+°ÿq®‡_Í«‹%%lìFÈÕK¼A½/zCpúzœ¢ì6|lÐ>)Áj«d©YÌ%Vί7£‰ñÛaŽÄ¶?'æ 9uûÑ5Ý Uλ´ó‡&Š"ƒø‹‚^8ÿHu½,ƒDÄkÑ€W\ –mè¨nâ»Ô“˜}r%É×Äè*Ž®¥n4qñ,Ç…wãþÐPÑ Ýr‰Ž<)¨S`¿LÎç`PH,#ÞD­·ùr9$1ødÕZ¿Ñ¬çgÜÇ<¹ÑçfJ•ò”L­V@}èŠù')08“ f‰SLƒì ö—ÎhˆÚ—~½Ht1L¨v¼A¯C/ §öøDÈ ŠW¸€K{yH%JR™ õ28ã}£z!´®á”»Añ4Œ•II{ý—‰ìÌHh‘ûè½ÈÀÎs‹j×’F'=ƒÎVÖœ˜U“X1s{gó<é°\®n‰tR ô¼Ï1øI¤Â<Ùjr À궡bÛ#<Ûô£*FC—|á/[á¼–ß}gZ“¬}øSümI÷:óÒuºÛibËøhKŸÉÂøßã·¤¦ï1”¹v|œÑ<øa±‰ý¹ûaä@£MûÜq{µeK)í~—‡ˆ«ªPqKÔ/¹;'´\&×ð‰í çðѼ`YæÔ[`.-¸òXrâ$ªL*â¼ Š%€­M¸KB _/Ê¢Èy˜wýD£l6Ól,3—•z[Óˆ¥;|´}·“œä±ühZÜÊOšyù$Ë©Y‚@HR½ÁMeMx<©-ÆËÐè|}ÆôÈ+¼,¥>ŒuÇ7Ü”y¾m>S¿gx#XѦ™ ÿQLÅ~[¾“†ÓòH ª|PÇ "ËÏBfÝ©°Öÿ€&1®ŒGf„7'€çˆ”É×6ÓèíG2b¿¼\ïX¤%VË;rs]Þúêtcÿƒù‹x¡ë±°Hlè›eǧñGjeO¤Dq´«§xXͯ‡ë«z¾M5e1•°æú~>¢I¥ºèƸ–ŠfeLÛÜò=§ª5°äKß¾Ÿþ!W´s\xº)65Q²“ˆ£3wñήò風ã—2“"g‰žSÍ6§Ï–wÒSÃ~ hÕõ¾ŒHò'ý»^Ôƒ_Å^%pæŠÌ­|+;>hÖÖÓ›Ã5Фü ñ¥Å}+òKÞAûwkËŠÉ-©ÉVÎ+8Vüojjò«rÅúZ‡ZbmXÊÙךy‡S.U÷%+œZÀãšÎÌ7x8ªljþ½fñŒ=›§vˆ `¾¯ÂóE›Q4¢¨ÏêÉ%$×›l6«ƒ šÏÄÈJ·AÙÐþ~¸¦òÍ.#gøÐa¿…%”™º(nû¾ÿÇüxa¼Ïe@¢o6‚´N>=Ñ¿ÿædϪÏÈ€§]W‹b‡ßˆ¬Ø²#•À ¾Ï·Oôƒá‡ÙjaQŠpýìgŠ¡ÛNìÝÕn2½A“.­RO2y£ÄÄcM;Íhû¨èíý_â-\;|Tç ˜_mM‰—5›+2Ë÷†=àcA½×éü4SÀãæPvö¾Ö#¿\ædù¨û†û|PÜóÅAQ ÁH÷Ò1ú‡ûSlŸá‘ãvWý¸þQ?¨%d¢—fc{¿GïAŠz^É´²³ÁqðyÆ8ûùƒ³bO¡û+»fï9\C9(„­xdðm'ÆçÅ9ϾFs ì¶C8%N  myÞ‚/>m_ 0ZÓ²>nÉShmyÍ2;zqã]†>“‚x÷cœÊ¬úR«ýtÙ«Šñº}K7 NûŽÃf§†‡S<ÇÜð𸽹ˈûãŸ+U¬›¿Ë×EØàj*P-s¶)ì¿gž†y‚¡Ã­v*‡n Ú(5“~¸@öê’GŒø´™ªÄ<‹×m¯.-0­"æÁ6Bzjl…×k“¨™k^™0ò§&Ÿ9uEWM¦ß¬?4ë Á= {ñ¾·HÆàRl/½íÖõC:Æ/ÙûEÊ[š¼ŠkbŒ©ülQGçœ2ßቺ¢,Q¹'îjiZ Š©r…†3Í_œïJÌ%´7·ÄÇ=ok›¬^3©bú×hÚžp?ƒ˜Ë™›MÔl à ù\3”¹ñiš FÂ×ê¢ ™±FùO¾·\æŸIæS¥íÚŸ3·qì9?}Oô KöÃB³—6SóëúFL¡¯ãÄ9fSàÐðHÔSAŽ‹ÿ%ld›…“¸X-+Såyêú!<²°ýÃëå‘u[ø¢²ÈÙϘƒÕÈ7zJ‚Œú×….­E]DII|Ëêß È!’Wû[zF¿näÜ÷q®¯ºz-ð¾dÐŽŸB5°zÜÛ¯š“hT†¦ø9„+$ůu·lÆ4…a²(05Àï÷††Û ÇÉc µ šž{¹4“ìGWôÙ¢é»yÓX®„’xËwÈ{ Rk|Ýœ€%¿ù0ŽU,:EóÇ{³u™"ö§GmÓ/8K­úJyo|¨‚t%3>ÕŽãÈå+oÙ 9\—£¹Y ÎʦèZó¥ûô-©]Ù5ÿ3›Ž áE±eÇ—N*¶V€É¹pa†E0Ér,Ë÷³`4²-/"rÛhl]ì>Ê™ìŽ+ꙣX‰¦i$Óõ„v=ùÕy(…÷y0[Kpù”x!õÄ+”Í¢ù>» ™5:Žçx¸àÁc|Ã9§ý/Ä* xåßGíQ“=Ûd9øîˆÇâGjt°îÂí›ë¹ÃÔ7;wŠxêÝyÍ–ÚñeÖ_¶VŒ¿î,¨ØíúBù³_ÛC7 EZÿ£aáÍòOÂLF³*ªT —7Øzßì;„´5Ú™…\Ÿô§©¾‰â»µ¼#÷fBt+éΤ٘ë&O?Œ:Å[¾H2RÙKwÈß²æ®è\Ñ'öˆ{pÅ6n´NùO)‚©¢4—£ˆ1€Î ¾"ëãŠá‹Ë›s1´ðψ¦¹ÂSÑ;RÓMó富w¦¸Ò½ä= ¾­;x—ûÐêìÔܬ­¸ÊF:Ñ Ähe; Ñÿ:q/ñìaêc£•yŒhµAˉÔr@]Œ"õÀÁ~Óןu*·ú_7D6·¥¢Z«j¹q.¦öôÉÞr}È;zO³2Ä÷ýp*Ú¿ÓmPtF•º¿490,¡F¾gºë_Õ?…ô Œ »¡ÜdÐàþDÿ®ŽCÄ*^EÙ¬¹šÎ½É½E‚7Ìl$”xQ…Z–¢i®ƒ½3BKáO/Œ8ì9öv¯ékÁS“iÂYX9„Povâ¾s]·‹8ˆqåFJhðÅÒZjí!4o§Š.@¨6w(Ÿ‚¼•/ çîå91K˜ŸïÙ`r óËA8Œ¡¾cŸh[éhÂs?6ƒÓ¬ïvÉ^ÝzŒGMëkß¹ˆòå•ü;˜=g«"P5ùèÐxyn®¹Ø[¤qL§Ží”]”^ÝÙª˜oå›»ˆŽû;*ù•F/‹¼î»Ê, 3Š*; ÁùÐ*êtæ¡Ò\¤zããAF¥áì¸À“úçÏ,6Gó§õê<§q ØþZ³þå endstream endobj 2124 0 obj << /Length1 2555 /Length2 21012 /Length3 0 /Length 22472 /Filter /FlateDecode >> stream xÚŒ÷P\ÛÚŠÜ!¸Ó@pww÷àî4и» î$¸…àNp'hp·àîìvö>ggŸÿ½ª{««è5>ŸÎ¹J²÷jŒ¢æ¦@){WFV&>€¸¢*/€……‰…… ‘’Räj ü[ŠH© tv9ØóýK/î 4qË$L\ÁfŠö97[+;€•‹•›…ÀÆÂÂû_Cg>€„‰;È Ès°º RŠ;8z9ƒ,­\ÁQþû 1£°òòr3üåµ:ƒÌLìŠ&®V@;pD3[€šƒèêõ?4V®®Ž|ÌÌL&v.LΖB´ «@ètvš~— P2±þU"%@Ý äò·XÍÁÂÕÃÄ lAf@{°ƒ›½9ÐŽ P“U(;íÿ6VøÛ€ðŸÖX™Xÿ¡û÷o"ý_Î&ffvŽ&ö^ {K€ÈP–R`rõte˜Ø›ÿ64±uqû›¸›€lMLÁ%nU˜€ëûOu.fÎ GW&íï ™Ó€›,io.î`g´wuAüŸÈhîºó_cµ±wð°÷ùûÙdonñ»s7Gf {“PVâ?`â™%ÐÀÉÂÂÂÍË:€žfVÌ¿ÉÕ½)Y‹Áùûù8:8,À%ý@@𢋉;àêìôóù·â"++Àdæ 0Z‚ìÿ°ƒÅ@‹¿1xòÎ O€ xñX,¿?ÿ<€wËÜÁÞÖëù_Ãe~/¡®"£FÿWÁÿ¨ÄÄ<>Œì¼F6N+ ;€üà÷¿,ÿÔÿßÚÿ’¾7ý'7–?Œ²öÞ¿K÷î¿e¸ÿg'hþs\hÿAɼÇ@ÍŸµ×gád1ÿaýÿ¼ü¹üÿÛùß,ÿ/kÿó‘r³µýKKó[ýÿ£5±ÙzýGÞb7Wð‰PtŸ ûÿkªüû+ÍAnvÿW+ëj>¢ö–¶ÿ4ä"òš¿¹šYý½@ÿ˜Þd|ïàú}ÍYYXþ|ÖÌlÀW‰ xR©€à£ô¿!%íÍÌŸ96N.€‰³³‰"xð`Ä ðaNs ç_{ `f²wp»Àåù,œÏ“‹À,ú[ô7â0‹ÿA<f‰?ˆÀ,ùâf0KýA¬fé?ˆ À,ó±˜eÿ ³ÜŽ.ÿq˜þ p.Š8¥?œ‹ò?ˆœËû?]õGWûƒÀÑÕÿ pt?]óG×úƒÀѵÿ pt/ØÒäÄŽnâÞ‹Í°»év7ûq‚uf¶àùÿCÁñ[bg÷‡ô÷b0›ÿ ‚Û üÃNüï­üÇ€ L´37q±ú—83‹?Lañ/Èñ‚þÅÆ 'ûoèþ'&ëoíýos7çEXþ ‚ùÿäÂn¼•—£Ðþ_`Ù¿â³€‹²þ7Ùæ_Ü4ÛApGíþU¸[˜9Á®öà³ô/=¸v‡?É€þG .ÆñLæ~gÚÛ-þt˜ƒõ?Rçÿi<8kGðìþ5¬ß­rú3|0¹“›ƒ+ÐÜÔöÙ9þ(þ—|›ÿ­ù_1ëï™ý«ã¬àºü©ú7ºÿ«Ãœ`sðkæŸ\Á}q±ýŸa§ÿ'øªfvµrþkÓÀ½põpø—˜Ãí_<÷ApâÿZ)°·ç¿ ˜Þë_Üï?É™¼Î‡úŸ ÏÌÍÜv׿^HàÛð¿ø¯_@ 'Ð qqÎÁŒ?ĺ6¤í¾Z”Ѓqg\pšrGë-£Ï¢s»Û/T¸Úª¬ uç[Ñ”ï=è+[’47"K¤Ï>G-õpaß’TZ}ŸŒT'wZ~à L‰Öõ#1ª‹ìú>;ùjÚ@µ¼é”£ÌurãA}ŸyïÑ'íY×_¶<úqnGe·ŠKé©lŠ1F#Z?°d†2Ïôó,9¬+#1<ƹ'ÚÌÍí4FÎÄ+©\=¢ßq {‘î[ìì÷j¹:›Kþ;|]­@‰óé^ (¯»KÊHA Ùå>x÷R¿sñIYDñ±¬@X?s–†P+úç—Ø xoÊèYy%¹§$㾯…S~sZ\Íù}¹x¢ð1ï ѹ”¢3É#¹[ÛV’Üf• ÔNµE.é@Ó} /ž²ËYtÊÌä©]HMˆ®¬üïlƒ†F_)Qú±ysŽybôù~óKZﱩjj‹UÕ s‹p¦ß4¤›„IÝTÃÑtø®Å«¿E¥@méú²ŠU¾ÜR¶ F3)KD'Òó6gwkl›Ô/‚qóŸÏŽ—(Ä»f;¾{è êa¿âéLkCýÃ×¾oB[.Ï{ÀÔ!ËþMo2†•†Uæd·ó&"Æ7¼n~‰EÀ„Y/˜ ÇéF—wıæ{ÄCóJ*Ì?¦yˈrP§ŒÏºQŒŒCÒíø—8ÜâMÌð =+È:WáÄÅ»ÚâÝÈQÃÞâÉKƾGwHl3{ð°Q·íŒÆÏÚ4 {\MŸ_jjYÛz$‰Yh ± I.˜µ"b­ H–µãýEfü¨K•§‘÷[M'KeÄøvg¨a¾s™EºXð\“)ÌwÌt±[8P2È3HÅî¼×Hä¬úZ&xZÈ7=Ëø…puÆ›„” -á3´å •ÊVµ®¬^ƒ€VŠ±pã»Þ|L;²í÷ýoÆN$ƒ¯ÆcPÓÜ_VÉw·o~öu/á´ÔŒ–r`+í»éê6 FSPuT9¹l+°º¯ $¶æß92»6Ò¬4Œ…âïT"/òS-Ù»Yjn”®ÜC ºäÐUré×ä¾ ®Ê&‚½²x\OOmËí¶û¾«ªï!çŽ0¡êx[}±Cãÿñ `Ç”£vh hª®ýÉ·6¶™˜—fê¸!¹H‚Åò¶â¨™¹öË›Ï_30-Œä¯åONžcå˜ÞØÆÄøöx‘/ÙUv‘Q,•xšNÝ9r*é½Ð©z¹¤%¿;â3e>©褋Ëÿ‚Æô±(]\÷ÿÄ]9Mz‡ð¿)”EÛý…fáÍ(2–J?BD—!q‹´ØD#­8žEÖ€mD—e³à‚ˆè纾€ò¢oˆƒŽ˜’¹¡G±ßâR(A —”vu"øwJ*®rYò÷nB¶ÈSÅö™~gÙ9?·Ê áæÝá÷­aòV¼²Ñd ƒ²3ã…?ÇðˆÖô~|ݱ&ÿ‚aeT_­5³â´;Ö³B£õå”ñ¦™€#Æ%Îh­ãóWòræÃï¦ [“+8e°[ûÞÙG;»šA=–¶ ÷øËÔ~÷¨ìUŠþÊÎ92cÚ-SÑOÉ ´k„zÊê‹ÑêR[´Ì¤þ}mp“„ý¾ýú2.<ÕfÏù¥ 4²›9ö‚;ÏõuÑXÜc 4d~cÓ…ÂȨìmÈž…Ÿ£dè ÙÔöøEÝpÉ“û†=¨]Ø€-DôG[&œép¡S%ÿbö³íÇËÞN÷ARÍð°Èí¸çúÒM줽±3N+¬éièÜw:èKG'jÒóÁµG Ì¢Jæ` ¹¦•hÅ~Ï,ú~ªM¢2sõLy VŸbm ½ W=úZèsÇSQyá<8Ó[ ù)©µPi~y}h!sÌT\šO’qXÞ–pܨ÷¸Z\iç1U"Ì‚ÝÍ…£ª“…| CÅk—/ô°È: "ù®U•Ö++<Ò_7TÌFø±'å\‡ßk:ÈÒça±ä[Íc+¤uh¯Ù ¨BëQ¥²Ç˨“ _>—uÀH[ÆÇØ3p=ÎóªÑȲÖR¿_¶É5â”4³&6*œÓ®2HtçËrÒEÍÀ®òö8†Þ>qòN-RºÙ}(yÙ¨{y“M$+ ”¾C¤ŽÕ‡„™Dç)>rˆ6ž} (‹îx¤¥ÓØÑ éõÝKóî×WŸv›1™rKÂ,†BúUtT`©]ÒHÎŽÉùÿ@U5 õD06µh ;¯N©Ch’~{Dœæ F[›ƒŽÊM–¨|8S<þ²nì€, ÇûPQbå³'cÃ+öQµòî¨ÂËFO:áKú|µ{Å!$ƒ²\1¯ÿæ)yôÙÄiËì`¯r­ƒ%M_«çþOñuË“äVX¿÷$@$4*Û6<–Äá;ê.”&Ê¡_ãš­Jǃ Q!׶«¨ðÖÙ‚üžgÑ/}×™Ã=ìh;´°‹dÒW®µ­ rø¡4ð7Ãl>+´ gÖæTó,þ!VQK‡ùäQ«¿ë-çæF(CFEúÈñUÓ­:GÛåN&ÔÎ^tõvø8‘9Ú¯½$¹dèFæÔ¼³µÜx&(úܯ‚|’g…ªI‰e<ÔÓ; ÎñwVpÁüs¤v4<¿Ðú¦÷ï¨;Z ÐéH¤È`bv¹ 0#’Òíuô 2$¿ã€ñ>ÚAµ 1µ9'r—S—ôfY×6²T#_S_í”ì˜I³X%yªÿ¤“†D¤†sjÌ›…ž¶TWè …âøzo&Ð?\º ,18smPõp›±ƒ°Ó‰“ñ*6(-ŒSËC^OPê:Å+y¯†G;„1Õ/¶íÉç`²yj±_%A•fD(ðBƒÂ}óš©”ª@ýrÚˆõ±Sj³’‹ÊËQ*²ù²Ohý\ZÜÊ_A©•vªÿLs&n-ù¦îŒD™e•ü±:/ã:b_ ŸÒÿå-`¨Ï#ãeìÀëÆ[‰œ=)8!ÞDçL–œA/â‹F‡‡tÏ|X*ÍøæfvÝBÞ ¤¹€V;4¨ÞÑæ#çÐÊéÕ—±Å¹|´jÅäúüZÖyéã+fÙR)©Iæ%]8*–*½ù~ƒGJîàOÍ»Üú8ÿ¶+ÑÉ—Ú“{_Ùm¯¬¢€¦ B\ÄüøWQʘ {k |X£@,»¡Z*a” ±üùL™®­h‡ÂëÎ ZŽêÈfÓmîÐ%àÅrA f•@RT•× ƒ†¶÷—º”Îzàxp7‹®B8œzÚÚHìqìq/™-—#ÔÍ}¼ªGÄTÉÎx˜·ÍPë&þ E²Y {ÒBc*£Ä{rXˆ! 1~Ãæô]1.›<\ÇñŽp½£ 4×£0ïÉ{ÄJƒdâãiø^þfþlfJ{ÕRŽèDF;õδHùå-tÑ••®‘réN¡¦ Eö¡_9~Ù÷7ìèN\ òÚ*_†GžFÑKj~}z:@{ßlþ£Ïj]î#Q{P´ƒ†+7´$¢ã GVŠÀßY„Í™´è] õšpÓQL\ñ¼rPUIð5™+S½é’Pa5r ³é¯  Ú_JP4Æ|ùÊ̲³S‡oÉo|c×€÷Ý£g‰SD%Š¼Ð‰†á4•°Õ[ÓjFoͲ>’Ûq» ‰S])Ýø‘ÖÖ«øxšÅ¬!Uýãd½áËB+—™kÍU×IW¬ŠÝ­B+fýæ·vP#«óÆêûDHRÚ,•ì³€ˆ· õ]uãGÌÉ>‡­W… O5Øá²ìü±–›×. 5?vw"ð>w§w/Ÿ˜«¸ ï¶Ù„‡©Ñ{E½I¬.½òY <'•(Â’g‡êg$òƤ? gH0%#ú‚*c¡£ :éÿì¼7‡Ër²Ìd‘ËfcæÀUi¹ÊYåüâcó­ïAˆ] çÜÈΔƒÜµcŸLô…x}“ƒ›Ò 1FáË{V?œ9•:ŠMw>Ù¹.°mtØÉòðáŠ'ÔÄÎ)uÁ6£…, m£þü%¨¿D'#3¨ÕÎ/­Z2GA¢JzBŽVD‰'‚“V|`™!Ïãæ¾a*ƒC°}†‹\ÑÛ56Û`YTŽÖ@!) Ö9 šNÖ>µ*GÜpO…L¶£GE31«ú<=r'Z·ö Ã,’mi6^ýcá,ÿ8¼XtZ'ª”é…ÔlUJÏu.´½vNCš,q"´¨À0ú&®'(3Äé Uÿ—Úéæ-ÁÎ"tœŽ¬Û±ž¶K…0&É(k6qWçÕ8¾Bi˜Jóª=§7¥×Jïä,Ë`~Ù+¸lûˆ9é Xã§4X¼èˆ-9ÅFž„Ü\‚:ðF®8kÕ¢a¿v‹×Õæ:ñ!AñòzÍ` T§}¼Åk Â¾ {)S›cÚ¥ßP§0~’£«¢+­Tv›çCbe¬m“6A'—Ò6mƒcûÔ¯,E/Ø"~çˆ5¨N«Ac¤ù*HQTÓª›8™7ó(¾Ú¡Ì÷¹vÛ¹Æ0v—O±˜×¥ ù%4gÒI$„N"ñ—Î'#ô€iYS!ñ  Þil‹ÈÊ ââ¡„òÜQ¤Ÿgí©Z'dÙ°‰f íPÖ®îã«zO Z¼8H¥½·´]M¢ö 3{o.PµUd8Ë0M†Ö›ç>»c·€$G–PC'PYìV 9Ô)¨8Êë§M¬·¬Ê¾ùK”^)Øüˆü§;„‚äM¢ÞZÞø´®’R!ñVÏ©vt/)ƹüžçT'üæ/ìi!z¸Áô­^³"¥Øk0ŸcˆÍžêé6·%ß ”3?½N‰HN¸K +*aôýr£ª©èÌ…`þî­uÌÝWJ AηŠ˜J&c¦ÛK×®$Dc¼‹¡€ðìýÛú8{Ö+š½÷ä2PõóÆ û»m¬5åjÈ—Åx=ÊntJƒ÷¨6žf‘ŸÓ_5 fºyËezœxªÙ¾N¼e¿È¿¼œ˜Í¼Îf TË€¥Â¡BkLÐï›4èúlªµžG–¯´V丈mÒ{Ê’h×f·jß:()ƒ\ï·/îÌwQ}EüvgæXZ°`"aëçèg­B(’;q…&þ&×­Ïz•‘ºëeo­2ê}u T8H‡vÅ°u §¸Û;¿$yîH9e¨…Z¶Ç+Ó {ž!¯nÓÐÌ`¹û–vòiœ)²m¥+™Ñ—&Ã¥"’O@CPÃXœ$¹wq¤%ÙßGc·Ç!Ÿk#9àî¶MÄMŒ¿ "Š±Ÿ&ÓÕÇŽh™—<`ÙûÈð•Œa~!Qÿ2œÇŒ»Ÿ´ GopTK5µ4}$ýÈÏm¦ì‹ÅkäÈHï¯\œöA€T"ÙA|Ÿ‡«èÀî)ëñÄåhe“Cü~LsZE¦í;?`ž¥ž+aSV$ }›Nš7ÑõqèM7‰WôÙ˜¶iRöšQ~d4 §Ô| ü²‡õ¹Iq[%Ãë6Ñ'Ižó}V½c/Oï·v0¸ôâá žö^8 Ï?Ôž§lƒŸÞ›P–¶+h©Ì,BÄRíºö0ÏõۼɉOÂxݱÖØŒEŠP%<Ñ¡š~Y¢ÊWèåd~anÛ·ªÞAIë›ÈćpùÑòÃ6¥K2Ñ7‹o.·`4Gt$ªûIÿƒ#°X‚«ÓQFÉð¶eågÚÕ²]Õ¦YLê C€Rêñ™Y«½‰ßñ&†¢ûZ–ðœäSSç¼;~ÿM½ûå Q;‹Ôò’9tößÐÑ´7&GPôq¼œ”0D†3\Z}é­«k^‘:oçÓÑ–Å’¨—A¶¨¤¤Lf¬û¨»$ùµ`“â¶reòZ5Óè„úÄ¢]ÀVIo'þÙIÉ3juÞ;°!ãy86ðæž|túæ,‹Ð]èWûC-€Ê.¥‚´K²œZÍHorºÈˉŠîA„6Å̆4r…%¶ÿu®¦™0[ƒ¿ÛÏ\Û™OÓé–¨Äyç('4}`â#MH@ú†Óâ݃—á.” ¡8>1otœÛÍ#·¡þZ6g/iÁ†³ûVF ùû;Õðžð²æ]Ý<Çï_ex{­×·{æ^4ƒ{)í·EíboÂò¶¶îÆ®¿ŽHœ^*G‹z@³ÍñÅ i!ýd…§Pî×Y%õ Õ:<„a¤pÍc_?wv&}_F½TÝh¯Ü.¼Ò6äËコ²p>-ðâÁ¡Ë6º?/H´ß{+Éíç´© ³c¸âÈ™¸~EÝö£Á¨±th%Èe*íñÔ§¼SjaËžó®·.í8Zr€Ô–ÿÖ¼ý àCa;ÞZ†æ'f¦~)Ï »Ucá<ÝOÿ}Z²0º“ÑMÿÏŸ4wZ:ÅH‚º¹7/£f´çOûÎòÙ•Û$fçICÅ?asȲj²Ä6Qß‘¯ø:( %–Ñr›Õ÷tÈŽÍÚ½í©ÚZ]‡.œ¯/ ÖÑ6(U ¥ã ‚'I®ìt`Ah0Rk-e¯Ì™]—vÎ0ý‚/WW¹¤]—AžÈ bfïÎVŸáž¹w9§j“Iä.¦ æ÷|XÍÃ[(%Þšº¥{!Ü°TT;œ bg §Ë±ÉûDO9–¯„XzÜûöAåkôÛù²TR3ÌiÙ-½±úGtŸ ˜·T¬AþèW¨/Q¿N…ôÑ_aMŠÞ¾E®SÇ C¬÷ÑqÓïÜŸ©ËHÞ‹ã@32pî,ï†E:î*¶ðï–-ÎGï\ó÷Ô¤“úÚØ>Œqô:Ó—üµ´ŠÆ­þçÏlŽ'FöOS6îQÔŠÊÂS'ơη‚yÎ|¾'R²+óWð1…±Ñäš^°/aw´úE€[ŠI]xŠcŒy,È~ü±79TF¬ÎìßSॄ¶9½Å°ÝÈï]ñs§äÄƯàF.’õÈë£.H£7{BZn7†²>­íÛP¥Æ9¼a6ÌZßñ²Öfwããø¼Ýþ4ÇAM¦î]¢ŒSy4Í-Ï 1v¸çiDöx°J˜¦ùÚ·Gcð¥£Õű~ rIZ챓?¶Þá›ÆYTÌ)v¬:Õ7B“ZÎtDt5ú¨žýµ?¯`†Ëê_'±Cl¢çV£»Q1XZô‚é8¨4 *D Ãç‰G‡Økü$6ï9Cýó˺-Wq¸[aŽj¤ðo ÉãìœGŠúŠ•ÚŒsxõnKGç£KYöÚݘƒ ŽÝoW+¸ë®Æ©žò9{ÚPÞvœ›äWÂQÓˆVz-¤„Ú¼9–ù©Ü#_Ù-rgrÇ¢4.ƒ¨5c¾óIR"Cº{ ¾·çùò5š«!ñ„º*ULB”ñ‚4. O%Itp !v¸1¸°»\,fÕ;•K•'â±_—F©uà—"+h³P’›ærfif¸Z=Vô`Oˆˆª±$G˜Â­aZ›ù Y³LÛÔ†Z6ji¶ —„ØdÙ’¹&ãVN“VÇnÚÏð礗\>²&©áò²ÐñØ‘Cmºa äk–ÊW»pB”1;dI¨ˆà‡¨—…òd{нþˆúîhW~Þ~YË%|Ü%ãÆ<^[~–5þ”ëC›ÊY)íRz\éŒ(Þ}ŠÍRcàÉðF½@¨ƒ³ˆyL#`·KóÍ{-@ù 0‹µ8Nuôr!Ïv'µ—+˵6í¶S£†½Ã˜ÏR 1º¿.‚ÛÍ!D)²q†¬x¨Ü¿áîJµpãtFã$‚ð-7vÍÈ7!“/öŽj3؆™B‹ÙÏr²N¿„z^b=ëëC=^tóÇ¥!8ÏxXN¥^™à?GŠ-á'n„4“$2ßÑrG(ØÛ54 Õ!¹`…cºjÍœgÓ±äõ-RêC¿·lsC«ëȈRd² ˜ájËŠ–ã]M¸Ì˜ û¹gÖ¼œ«ûn ·¥ Õß`å0N¯2áÆP½ÑGjàØñx‰ÞÙI§o” Àÿìþnì@ê4šs¾Çï[)*Mù7?nõȽ;Eè6ibËÒ…E›M¥ËŒz»Ýª»ðI«h ´zÝú’ÞÛ0€š£×wºôÜH4^…!²4ª‹ý]±º\5C^,ÃMWó‡(©ëÁû¯h*³bk[]…ÐÎg±6üL |“Û¡ óŒúToÐœër©@¤‹En2CA9P¸Ã{ÐYS²~¡Jm7 ÔÕëàw³Ïö¢åÚzj¯3¥8¡ˆíÝù>Ÿ)ßs´ ëû"B€‘¦-õÈ$¡%Ãc8hUÑ\4µ¼²È5†·ÈŠRŽ¼½ÚÜÓ?i¡[î· cìw’¼t5±ð_ øÏZ]ÌJMú*¶|ÝÔ#æ`µ°^Â9ì¹Cë¾Sba®‘Fº!pEÖ'Nî±à,Ëó¦æ I[þxж ‘±i!ÞZeðe ªåŠ¶‹û™— äÏæç3?Y±Î+A5ìmv@ñÉ;¬RixZ)"¯-,¹‰€Î}d4hñ’Üå°ˆ„SUŒh#üÔà8 l ÕÞ¼§NÁL…!ÏPlÆ5ž”ˆHJ^ZkSÖһءGç‘«´1æo{eë’ &½w@Gƒ?.pÉP±é_M»¢*Mð“Œéd˜¶Õ<=Ìac ÷.>y0S'^ºi;Fmÿ•¼áÏ{ˆ]j²Ö)>0ºÆãቸþv-†pÆÁáÙ?)ùk£/°ëQIê¸!„ç=$…‡'=d|Nˆ ‘ñcÏÎ(€0Nä@þ5ÐÉCtçLÕDÐLS°@EpÎOFùC¶Æ玚(“„ÓØ-g—%U8T½˜h×{nÜkÆÉ™=OÙ»óÇì¨m¥âPüyjŠªAŽÍ,·y¯‹B»1û6tÃzã.~sÖdOXl€ödJ´oÚjSi ä!œŸ0 üÊm?…½sÁ‡,(J¼'ÃPß:Ì8ü°KSøšKû6.ìËÞakØBõ9BÀd)+aqªaÃâöÇÚ¹öÅÀPs×(_¾\qd´„ Æ“EÄO„¤ž ¥Ã¹-¬jX…·w*;¯Î\n«E+Szêßi—_ÙêsÕüÐü„2jSÄ” †º»´GÞGý’N,Bö]·Û*Ý©þQ¾¹¥h»Š®H×x£—@¸^=Úw\ŠœØô;äíxVzï6€ se¢'3õ­‰"…Æ7c–Ûöº„ÎÖýQ]›Ñ\`Ýw”ì1ð»\ Iaì×f3éÛn§%9ÞNX¯$Ñ1ø^n<ÄNsÊgã P«ªÖ¨W€¸óÔþ²—É |çò Å„KF¿^ïX÷ihçÙYdëÞÙVÛÓyós¥ l píY½uo×nö ÑŒÎÞÀXÒ}äU®™÷ü]ÍŠû[ì´_ (´ïÚP´`YÏÛ%MIãgì˜nÀÿ½Ëvdi¿™³ã J Æò`deM×ö¸Ù/ö›W¼Ub3ôüÞËE°‡™˜GûÅ:­q†&t‘€è-R¼ f¾¡­G;¤‡vTX¬ŠÚmUÛ ”¯:¤.¡„óê „A­®Ñ\8‡€ƒöV¢‘õ‡nþÁÊbÛ®(lÎ#®.$ª !ƒk±mhy“1ÖµhýfÔ¾ûˆ(† æω? ÒÇ·Ò„¾¼â¦_§Áî{èµ1U?ÓIÜÙÑmZºUÿbïŸÇ§µEÜ«§ó}Æ'IÁ¬bÿö3Þl:fÖü„)š.¿+!÷ aü—¯ ¾Ã¥¯»úHÙ"lM–|ƒdƒ,×”þ7E7}íë ÆôAÒî3:êÚR+,lžîM'éÞÊä üzÿ,RZ™Ãn¯‘H˘æXØi͹ãþsü+jàðO£áì—‹Jÿ…àÕÛ ¤1šÃ.~´¨SrŒŠÖ ûÒüb³…ÌUî¶TOÐ#ª‡%U”j» ©aPNÔÃ[õ+eÔŸÌ#eÞ×Çô¶ñ!Ü8Ií­#çsB‘¨¶¶§_Ý7íP2¡ëS_ V9ïŸJð¬ýW>9z$boÊ‹©¾ùr%BgMáèÊ—ÊÚ˜FH<&1쬟â¯9‡\Ì™.sOÓm3íÈFq6ĺ­~ÿãt÷1„™_ŒM±†d©ay½€€ëuúS?ñ§o¡!lÏ[˜øèFÏR0/ŸéÑFˆïÈ2æAÂZò™†$ØDº•ýuˆ‚ù$Ã_z/‹w¨iùUÙ>7y=i&Ž§NœàÚ¦RD Á÷Ð;Hà/0DåÛ]vœQÿÓk6ÂŽù´Õƒ/fbyªS ©'.ñÔtµ.¼gS?n‡…\®vA‘†ÉðÅ#~y¢òGj꽂 çï&™ƒí)?Ù¾ZR›RBæä`G¿åYì'ôo÷¹+! y㾟Gy¶ÇÂ#$NÕ;”qåú-O’ õë"ÒáM99F6^öLx_È\ö\THEF¥îa‡š>m|‘¾kT+˜­¶þõ(‡‚æÎOÆù*º ­é¸¼½z˜È6bÛ ÕdlRé~횶‰è >¥ž3Tgùjöö‡$Üù} õÁÔ…ñÏV9'´Ûþîy"·27+Ÿ¿g•ø´úW•WhïR¢¯žy28âéÍnªQøª¤ê*&¿.lø™h6c~fž§E9/ H–Ë—e›OPÞOZÀ'õt'¼¢uËùâܱÖ;€ ñΡòn¦†OÌ0Y ¿ øÆyfÐ níl-Ƭ¦õÜ~RT]4˜”q½3|ê¯+Š¨¾ DžéY¨ef  ÐÕ1à­·»ÛØ#oúÀýÅ’).†eiÔ6”ÀQþ¶£Ñó’·ÊaœîU¥¼Š¹~zá[ƒ™kcGˆhK ® äùôp5”F¶V¤}ªÎ¹c <ݸ‡àcÌÙ£K¼‘Ìí]L|þ÷½±C”TÚÑWŠåœÃÈÖÑÁñ—C(vî·Ð°°E©‡nÅf_!\ûF¹3g~#ÅZ®=ï ÌÖŽob+_C¦$S1Éb6}òÿ½Vÿ9’«ýŽ"­˜Þk+Na•÷—%É~³ÁúŒh0Ù\'åãà;“ócR{) ?‡•@6ók콄0'ÈûÉ žÑ¯Áª~/ßmÏãË9âÞ&‡’ÆIÊ7Œ.M­´pCê÷…8¤]•Ruì«W‡Rka-ÌûÐU\øa"ÃbôÚÑ=u\N[äYÕתÌ.â,a}å)Œ;ñî`[Gâl’ )BÌ*òO×5ô9aáÉ„eØòìXÎPo¬» x¨jÜ%´ã«8ZgY¥@Ü…Lë X)裉¿L>r9ÒÏß[£f’®L.d ÕËp”_>2™­Ÿšät姭(›…@Áo¦EzãpÅkND}I…³ç›¼å*¢oþAãº{qÚ¤Ý":brÂÖ ;ßKèªØÉýHÐnŽá+ËëÞ¥Š\¶yvïÝ«ª#LÖ}Ö²8RÏÇ< mUâÕ2e¢î"áé²âäIO‘tzÏ•í,5L¦#ä–™1'U{I’AkæûÕÍñv{?‡ÊÕ͆µüêl6¯‹€ß†92^» úI-vìmÑ I=³„þØDI°¤›Af¹³Á¦¤¶2]Ÿ‰ïOÓ(—o<¯PgÞrÜÅ[jË;Ò‡ñ âꬃe EŽÞæÍ]ömª§|mRŸ—¼1Ù‰æ"±®-n?€*ï]LB0]Øð‰4Y˜“U„2¼û”OŽš©Nü„W6BùZ1S¸rÌ%YÕ©eæQøaû}:lÚ‹KQ@WÎy®Än4W}=™xXGLª–rèÀ·ëÂð/z"ëŠW}”,•Ú9¥5‚n#&Á³/ŸãL xmièEñ·ßO<ÚŒOÖ;LÌd†ùÙÄ Y¨Íª*VxcZ3nÞ-X¬¤'°ä>D ÷}#д iti»‰±ÁAoC—?^ïÜÿtza{¨' $ yS|%ã–.b][wDá߬gŒšg€,ÁÂmX ô^¶?ÃèצSæ •ž›ö\ËBlFþA[$œUkû%Š.(kñn‹¯èK±­Â£Å†i:‡÷¦Ÿê1âQ3óvVÓ ‚‰¸¼ úf¹…':ï4í¼ª¦ÒçãÿL X·ÂprÁBl³â-Ä®½ã°nQóMy¡ò€‹?}„CÀ“Æ'±ÐÆf}]*sÈ;_“¦yJ§_"ñɘô-6º8Ðúª÷¼v@ž¾ãjÛæ¶sæãè Göö‚©y*Ÿ.d‘ѵ´Æ«ö梳dé ]÷îÏŽÖØè¿é}ñàÍéßyŸÊäÂþö.ÙJi"wýÄ:)".)´›†©÷Úäk>Ù—wC¯¤8$Ý^…’ß!%¼jô [vÏ‹vĜֺL«bÚ6¾Ò²,M\0|ÇeãC’EòöðÃͱ9»÷é}PBèêÆÚ õ7¥þÆ¥˜E ûùWÜ«È#?âwE“¸èÊ•5M˯ãm²©}Òå)Þ¶øùÕ>vþ¨Á ClUÎäRŒ¢¦YRÕOP3®ŠƒÜ <ƒ¬Ò±Ýùâ†JBëÁ¶9~PØ>wbNÙ­g<²0.£nê5~ÅiŸ°¾£Å®Až™Ìr8³úÃÛHyÉ&tï2æf„g^È?VÁ j».’IX£›óMÞÎAÑù½ @Æ]Ó®5ì|mWM²ô·|ETÎÑ’²+l+™æ h÷tC{ì4†þÕ•Þ†ezI˜føð©Zf›W~”U§IŒaÍËEy¦9ĽÀRˆ€´Çx%„ëUvÝ9ô-¸Ž%¹›AÇŒ&©uKY­]s.ÒÖÖ#ê¥ÿÙ·¤Š±)Ü\Óԟ˲&L^_µ…LhOâG¹Ñ-wèpY —ØB« ƒ/ÈS5¤™“ ÂRhK¿­Zfƒ­"½¦µr<ÿ½¾–ŽÌýJ:±‡Z¶+)6^IØk£mœ/ÑßøgØ ǹU ¾ÞP*ˆÉ&¿Óˆpå0^ï{~Ž*ºÃE*Êm–þMµÓ}û{ÎSqdŽr{*Å6¾ˆ dO Ìò½Ÿw#)ÃÏë*ñ}<¨¾1 ƒ¡aøÚg䧱®Ã¥2¾‘L®'úo¾.¹&ßkÈ!‹ãÇ(<ðmÌ7Om¿uÂF“þ‘“•ï,é–D–ž'Œ¼‘žû‰7oâ¡´fÛú9‡n«Ž…e&Õü † „`éè<Ò¥ E®BŒ„ÀÍ~g€Ï¢™º‘·Èíƒq[SAÿP´¾(•²„î* oî ¥u§2\†6ÎB´qŒ½êúéK<Ì’&Å:fêZu‚‚Ú˜·¢F¼5•Üûþ’'P‘•Ï«×ð÷¸Ä¢:Ä͆²¶öoÓ¸¢§}űͷÇéúÆc|~èLz(Q#“ž?¡»2“_Rgë *åä¯ý4BíhDç„mcŒ+Ìk øH'*t½¶u™7¥Ñ5è™Aû”\¼Ÿ×ÄAA¥{\ZþÌÇGOJ±Ô.˜¢U(Ö¨`[ˤÂýý™)Õ¾¾á”÷Z~ÌŒóbG&3å˜ß±}*9æ(1A«LÅíÏljÈ.B´T¶èF¼ì׳UyfTŸ­Ö›§ >'ì„c´3±î½ì1±zØÕj|̳ìùö&KÔ2üaþq×÷5Xö²z "¼Úsèú¢R¶9‡&Ì>i×CØ=ãlð?õ‘P‹o® È~°^E°äùU.$3%EàöÜýÈXí/]ñ‰*~U©ø¶úË2Š¡‘´–ü$–TQÿ}ÒwYE²@_Úgº–Š ªB|¾M&‰i‚äØ@¤áÏ;¨íO³ÇÅkA ^ Û÷Á°Šü¦Îð{ýÎFA·Ýãel±T0n{ö_zO QÒ“àv ’âËíó?!"Ê|')„›F4ÔÓõGzs‰Ý¦å«à“PRìoà 8Q‘µL÷–hõ<ÕªÃÇØÎÑ(vr•u#,7!”„ÑC±¢°+Eaçy‚1Šñ›B<×Õ7ú5Qó¥é3ÛÜñŸÕŒêåÛžê‡N,‚¿m4' czóp³>0Ü„Ð ù$¤÷ù'i~öèñÞL1ù®zI'¤¡™ס„”dw,øµîO?ãjaH*J¯º‹#¤ ´×3œ·•:°òÑ4󻳑æ’d’êsà³SÔJîž‹L«q~1k®N‹‹©Xésá2m/‰˜äø~X€±´ú:³à>4³€¡Üã—hI†ÀR#Nô Ó\d´L£föƒ½S-× éˆÃþžÔ ¥Nƒ ¹ùÄiùÖt@Às#q¤K•{[ãq‘C¯©)šÑ-Çݸ‹3Ã'¯RÄD¹Ñ…h̪2­w™›¢‚b€„•ýs¡Qzß^~T;^Ÿy ô‰H~$‰:eSðÕò%ï"âTY-»FMUÔ¹+jókØt±Y2ãspDt¿@ %îRÑI‹5s`RoX³Sƒò¨•ÂE!Dk]Gͧ£ÚË ³»ÛX¼»$lÊðÎYù¨ùËÕÒ´Ž[Ø3Có^c”¦¦¾‡‘?Ù ®ìàË…F,c»ï¬nWõãkÊ-~CMÏ\#Í><Þ@ì·Ú·je_({/x}Ý„è·-èÍ\O²èÙôQ)mž-ù"¯ŸjЦCå2Ú«H>Qd^é{~Íœ;ªRn‘Ž¨%: £nö^Õ¹I­_|÷£Ô.tåá§Û¦J!en;^˜ª¦!dDÊIO•…ïL™˜ýq=¥[+5Ð{\f.Ö{’=!i‰‹W÷YŒó¡œyô¾Eó3ü›K¤a1©PÅ!ëQ±:cÚäÀ#nœBbý_¿Æû² L ©tˆ\;rÑ#Þ H7ßV¥å>ýôç_;iÈ|“ÏÒŒ-ìþ^§`•(èsþÈx‡ðîeE G‚î½JjŠ<¿kÈ"^Õ‡úœZ¾É’½’Vë×ð8îzo$Siì÷¬ 5”6 µ%ä:YO½Ñ’“žÎ¦{9ƒ ‡fó ¼mI©e<s{w î(â7¢cÓk•³©U4–]1ó+ô|õ§Qõ¼f±ã×ñQðTÒU°ÎÍ=}õ9Ñ?`ÐÛ÷g£Ûo¼ÎÆCL˜Qä­ïø)ñˆø±‚˜c:$Îâq¶©³e¥ÓŽ>V+xè„vÚ"2<Ü:œêê¨X¾éÇVž.×Uhx•*3Øý8ný#Éfí™ÔR0\Ûf Ãÿ z÷a•ÞèiÿºaSˆ}Yûb¹nh/œ¹uÁEËÆÖb9]¥ðšR¤š…¼¢’8™‰äÞëÔÔΆ²WWålcg6ÓðÉǾI3ŽO_÷ú¿Ïß-g¨°„k¨ÿBÆ9 4¼¥Ìüî»Ê¶´¾÷ÕsâÒdp†½ö+#ü¯´‘`ñ~ºÅ˜ « Æ+lŽâM/t êFÑ#Œrùg´$èµ2nî*ø=(~d~7ÜÂÇ×ïÌ$®j.òÚxŠ„èçH Þÿ£àÛr~Þü;æjäR~r:Äý@›Ò+.'¼o ¿=¢lAùb ŒcŒK7ñ¡‡ÃêYË€¿ÎÎÄUæѵ“Ãðîmô5Ûiç¨né€ÃèzJ¼í–Wü ]ï~Bæ©uqµˆ³ªaØjµ€aw—?GëZ#¸¼WèVHúÔ‰ Ð ¤µ(Ô]~:ypˆð· Ÿ™DÜOWA˜Ê’uÛƒ˜¥Þ-(:2Ð:Ëû eÑÔ?fÀgîdÏ0Ôó)èH®ŽE8?“©ÄHô(_"(4·|Ø,‰ J}n‰§¿(­ßnEDV)Þi`³°Ó‘–1zxƒ¸Å?¾<®ˆ¤"[ÒdÖ˜ÍÙ–/©óÈÄû/9žg Xõ[‰TԣʖPÚä{¿‚†¨DãŪöìDܾ¯TÔµ»Á¶µ£žÍÃb(!¦™w†OZzÖœ7ó)øÌL¹ÔϪ_¨c`:U%ør¸|&lß^Ö™ÃÈÆâ’Æ*†AB׌վ弓Íâ €9²¹iS¼“ÿàì½ð¹Y\MõW=Š€¿¤òmR¶t]@îËT3§?â;¨À‡.ÍCV©ÏöE´™™k²n~{oPWÆÉÒ›°Qö­¢‰@½ômô|P½ï' t/BÀ*‘ “Ø+—$RÒ,Ï1{ÏlHÏ|Ê‘ÝacW"‹éÁüA¨ÝR®™p1„.ae64Íe4ÿðêÜj–«;ðöx¡¨2C9a>nèøä0öœqߥiw~Y‡“ˆ¨i‚æeB¿˜H462_ÖIš™†e×V鬳ÿz>3GƒMìV7¿ÛHä6êìÍf>!žÖ”Ç\û*ªtžÂ“\ê‰XÝM]äÝú±_Ü•ìRì©ãun÷—1–³µF"%’@]Ù‰&ó'ŸüÂÖˆÔŸdû¾×÷ùæWo·|yºu#gß_ßñ_•ÚAZºËÈ·ÌVÀä6ÍÁÞòÉz3ªq+VŽ^à{î¦f‹¿°æqq=DÛ¨&Ñð)Ñx³\àßd ³»MrOûe¹÷<ÀÈ_KF ÷` |¯òaZ_í¼0.ì|É<ÿÂ/Rfëz·t62±zbkï·Tð€¤ðÁµûœzN«Q&5¢Ö¬\ìùÚ t.ˆšP.6żN\ú“ìd4ÿш‡W¸sõHfã¹SSsðCWè›—Ãw)‹TÏ8 a8 TS‘æ¡ø‹él_P3]¼dÌi÷£ñ[ö”«i”•­ðR[%Å>x|>'fðõo„‡7ÈÜ¥žÔ°“ŠÜ:Íl÷¢%ö¦Ôßh¿t˜È0ùÃ’×Ãå;{ ¡zõeþÝE¸ÙqÏ7…³2'iH‡ ·9Qg*nÊ uD³eVR·©Œ0%шÐ{¶©˜´ØíZºÞšö¢Í\ÇåzÁR¹Cð)M%·|ËTÂß|–©ÚX¾êª† ãÈDŒ¼öçÜ–¾^Ñeåðëxy³d•»Š6‡8X lÊë· b}¦õQ`7ùeÄðÛ䤷{öpÙèoÁÞuÁ½Ñº1q\¨XÞFâèòŽb?·ÆÄÍÝ܇9$ÜÒ¾Mì|YvóU†ÒŸ^ùÁ©t¡–%Ôn]cy¸ó*1°DD="¹k× ßšBúæЕô^:3$à ï‘<¥Ž§žçVI9Í`½¥‰ÏsæN—ì¬/[ƒ¨FZ_p߉•"ù éEN.<8èÁ‹V¶o®V“húé- „èºÌ;‘yP»¼ý€³÷1V%,ã– ÔJß«›Ü€ü.<Š²¦ï.MÊÅWt¾8À1©N  ó’$»jöÁ‘ÖMx b+Øo‡BGWçÖÖf"ÄžFÃNù|g(]ßW…ÉᜦˆÂIë¨Tô[ §»›yoÞÃÿúé(‘à¥;¦…‹¾ú´ð')i>žu?+Í›ß+•[ž÷†DömΛ˜š¶—ìz˜BŒšjì*ow·ãfE»œÅÓa §òÒ‚“Ä×ߺH ¬IzÀT´fö¶ÎS#hUÚeÁkn§ƒ DÃÊ=5*8_Ki™°fT [BOºWR¹k>ÉDö7ëÇë–Pê ¬6_F©úB¼,9Ö gN›Å¾UîbUV?‡\H\±S-¤§ŽwdÃéò{€p| %? ²!4ÿaü*^Uä8´N™AÛÐÿ'aŸ3rËïtú¹.«t³}B£¡Žk¼ïËÏKgº±hhºx»#" °ÇŽéP8aYÔßs£÷Ð|·lˆ´ä¦B{ÏѦžI7 Uö!ËÑÌô–B~e+ý3ƒ˜EÁÛ´™ «5%_Yò)á<8q}˜C¶Ç]Ñ´ çЇÑÊyl‚œÇ–¬ÇêßuC}‰˜/ùôbs‰Fs¬ ÞU5þ°R5ý€‹vŒíÆâ † ½Ý/,3ol©C®¼GS±Ê$Á¢9mœ¡+óÊò^ñËWÊãò/µ3{p=*y©_›¶½6‹üp_tþÆzä¿+Gám10Yȯjòe½È½ÚONµß3­V ­¹ Ýñ²R©„ Jz??ª¤–·c‡³+±8Œ§}Bô{}ñøòZaJêæ}ʾƒÐlW®ïqÞ—€„@On2æùÕÅ@t|§2¸ZÑ9Ø}ifºñ]å¦>N‹mƒ8}±óíå.„ö-ž©3‚/ï~òz˜¹šºã##ûµ1Žh¸MGÓ÷Ï;Çà?Q¢Vr!3úkS©' ñ{¦HLçôáYxŽ˜rÔÊæˆ:ÉÌ9 ŽÆÄ&(<ß‹w¼‡k fuÆÇZ!ÉÝÓxXEüMâSpe-må­Øáq±6KvÏcS¯“tx[!#…ŠÌì`Ô&:ÃwÄJ~»èÃŒefêO¡Á»“Ï”q;¥3ÀØ‚?ˆÕ*ï6oN?ºòt'ÄTnB2 3Ý ·yù&Ìr¼ÂÌHHtœs[ö o?§XYÏ+œÖ硱™=²{Ó®)ZºPÿ«® ›P¯–PŸå•Ïç=‹¹éû–žu9HcCÿá˜Ë<@iKê&øí¢½¡ ü"ðXŒ¾í±K£Î¦%"G¬JµòƲUnH…N ›£íƒMæÑ¿ãÛ®Šð4ïµà G¸]ýúÒ´ùƒR%SÊüT¡錃¼gÞëeÒ‹-& lÈh9oÀS(£­ø’¶Øî&Eî•­ ¾ò_i¸âÀxŠÝär~Ò ž0¸/¬³S¸fÞ¨D;ù —¶Í¢ã†8û8†ldÓ‚|`›=#´Rè#½%vÊÖc@ÞJ“F*U¸…ŸJe†ðWñ…øÀP°s% ú(§ä†jˆxciw[¢1z>gÙ9-eéþl;Ú‡µ~=ñn¢H¨Cå{yŠNsF&€Rü}å –Àv×yý"‹¹!¾äº «JáÎÛ?y­pÀ$YÖ&W‹Òü³ÔV¯És+¯È©l›Ðyˆåºnè9« ŽZb öw®‡§ÅՃ˭”Þì¯Xžh†ËŸÔ ç¢0f’Ê/ßÎÂ.ñòû¾|Oõ@´NNÏ©ZÁég¢©1Œ Èj<¶¬½y+"û5B:©'ÚšjYsëšÉ•8ݺÃéTF83»׊¥}FR‘ 3½tºF1³ÜÌûlÐm»[áRRâùoñ¤ô†Œ KƒÒJÁ¥BûàäžÑoîŸm‚ýª©¢h…ÂRž ð“#€ù0¥|Šãe3À©Ôk¯&ðùräç1Àuª#8Ä+Ä¥WGem’œ”šä=ó‹U1!C™3h£Ý·_\®í2 %ÌƦËÊ"06΂·â§]ãá슾noɃ#$7ènÖx›²Ugw XÓ›daÝ?xç|ÏòvTÖÑf;V« gz§†ç»Éf®û¨™¸M“¥¸¯,°šGqó’àï ±‡;{âÏ#¬~ %:«ªPL¢d…moêA`à[s·­¢ôÒiG…M¦å„™%ß½í&Ê Ä¸O¾åöàpŒâêm‰?¹wŒf³ÿTÁ\‹‘5e³6¿h1ÊyÖÏVõ£¦¸E@.?EsèvÀ*, ¬à‡7¯*‚¿%Éæ´¦`IÅýVÞ9H)qš^Å rgAÍ(=×æs¾+uܨ.3ŒÈ“‹~㵈gJGf—Q|>+R>ô$aúÎŒ¬Í—‚ÖqØbÕ¯É"–€H¥ÌÏÌŸÓq?ǾYÒûhÏï;©¨Éc¥‰ kÒXû$ãè ’ðDÂ<_®rÊ©htÚY°$ÀçÀˆºG³µïTä?tîWþaËëöQºIÇà;s( íµ83UÛÓÿ\aoM‘É7<‰Ü O¯‰œ¿Þá;,d:…”Øæô/Ÿâ…aó'aLb|þq½Å´0ÇwÃ{wFÁÞ& —$)ýœ4™ÒS¹EÙÒŠWqë’/J¥¶³flN2¦ :GÂM%ÒðºÚ56‰9ýJö»F}7(þiÚk´çÈjÞªû/Ÿs€ÌGÂx+ÊUÛ£Ï>ð”Ð! ý ö‰þ•·ù`¸Ùî7ì ¥œæTG®Xbâ鿯„Ý*öH4)H$@ASBsÃK\JXôQÒ=2€¦“Ž’Ú}];mר2.©~ª²²÷èáÅ÷ú|:Ó©øDùÙ÷·áÖbLÔ=»NÈÕlõÊ@a©DOG ²H µPš¾•lœl•9ÿRÁ˜è-^EìX“~4Éài…õŠäf‹>õÁÞ–³p‰¡¸F"ßšÅ)|m8Ø—¤°!_³úЫJ€me™ƒÉâ'È÷èÒ&™°gÒ[‰úz1¨ÍÇb6‚Úë†r9õ AKÚÏJ;=û&­1}÷¼C^䜋Ì&âúKc<¶Y#é ËCy–T*j¾Qª¤’ô[ÜB3³-RÈÎÙü"íW?'•÷Ä&J·¤o¢µ!lô~xLÊÀ£^ùcAÓĆ,’gm¤ÿ—%hjUƒX¦gº .ŒAîùŽ¡®¶š9øêê€oKIäIÙc­€ºÀ‘õHnLoQÝ3­ ‹…µ ÔCI½lþvÝTG@GçxænÝF›È¤§(,ýx÷v=ó©u¦;x(Ÿt ½›tñ¬U óøÍÔ•½»Š ‚´ËµÔ4mlL÷ísZNW;þl0ß›Lz—¬Ù¬¿›¨®#þ«Ú!¡àÖ~bðˆè®·¥‰õ›=Y]ÓG£ãÝWLê*wñùu{ó.¥Æ* Ò䯰kK”©Üô ·~?o°žqÈ8@O²RêßêNí-µ¤¸úf\´è”f)Ñ<—ŽHtù+ˆëÖVðuéúƒað*³gD{™cgÄôÆ9í´Å\2àÌ=E0¶êšbÊ cD!Hš$vsk`[8« ÎroôuùzÞYÌwÐäÞÚ¡^•‘h O€OÍ•\Åè*¼/?)'ž N-LšÄ ‡}€ #bÜËqø±–ÐÖ| B™9 û—ò*”¨(Ê#3Ũñžq¯ª_;å6²þ¼Î»à(™M¥„»êïEcº¸'Õ‹‰¾GÑ~„AYò³‘4êþì]qÅ}°‚Å?Ö‹×ÏaPñabÿÝ!=7#•Ÿ—|2œÐl´•ÑË?}Zq Pà–{€´·¿hM…°ê@.:*âK6b'‰¶4±´ Û?n Ñ÷+C¥i˜Loµ>ã4k¨¡lWÉh¤H±¯“ê$9õÃAÑýœ#©xXC‡°dP …£GÌ­$G–º92øÙFm® pÛÆF„ìf<΃Ï<°…jÌlcC JJ+S²›#K'‘1 =­ŸWÈîhGaQ »@]|ܤե :äEÏ´•èÊ–õ¬R‡x\ñÎnäj )¸÷”º·‘š™ÐKRŦq¦#™ÛÁh%˜âáЬ‡ b)m бû$oî§J÷¶"Ñ26~ÔÕrÎbSÊ©wqFzWéQuÛOø?ñ•:?YâàÉñç9¿d€¼ëÎë®­$g=Jƒp^‘¢ûG]Œ™¡å14H³"¼ÏÛ?¨K>ªRú ÜËFÞ ß‰+‹(Ìù)ã>Fçô‰èÂzÔŠ‡Ù£‚»‹e½ýGiRpjw-uæŽG[­zéÔ%õøرS;3Ö˜+qÑÚjd…õçB,vÀÀ0þÑô=M/IFšÃ‘’÷Ê“·(‹$‘š6ìíÍó—(‡Ç…kzP8÷QÑ>ý@«âµž$¶æ‡ázO~ ÌñÕz¦•« Í›-˜d´ nZ‘é–êáx •|ŽCoì×>6îØpÛ˜E˜M+Ô7nCDnUÅý²2ÇrU'Ôn(I Ù8„ÄK‘QsS49ºá÷ENÛµ>©¢2 þepþë÷׈ÏMvÍÛ½65Žr.@ÿ‘*7UÍ~Žü¸´íWqf‡gÕŽÛ¹½iÚÖC ¨níâ¾ï¬Hé'únëµ1¡ºœû›fõ»I2'¨kAU‹T¯šçç“´g%ÀÉ¥®^·4CEv!í™­½ýÓbe§ÿ)ÓÊ8DRgˆûhKeúîð&D¯ˆ¥*°ÂÕŽ6{”ÏÚ'ø§ Äâ?U„€Cu}UHq‰´Ftîùag IÞýv°ÆÜߘ‚ö“V ØuS$Ã(ÃÍòÇ“-[7ü¿t÷÷4À•ÑÁ¨¡±Lß0*òÐ*·<®úpüZëÃKU”Ö§ï«Ù’¦ƒDÇq8>I@š–¦ÃžI6§ƒ*•Í×:Õþ„VŸZ®=8ub¾AÚvv˜£ š6[{àïPTŸù•Èƒ=` j5Î].ª! ®Ôj¶ë–ô”`ÓB~sÿÄsZˆ– Ž"(¼±ty£<›5Ÿ?ÕâˆKpôº™Íó¢ƒð+i1Sçø$ºJ‘økΗb8ê Q$©Ìÿ|ë…äîûpJí¨šÓ-qŒíƒ*õXëŽ1Z÷ø‘A7†ŒVHx³^õØÞ+Ü/ªXIñ³ÅÌoAµ¿úk¶ºc”%Í¥07­àtÒî(|Sf•\Ó|ÜÁáp9êºo§.¼Ä¹-%X[m¸(qE\4aÏq.)¬àÊÜ×:qªýñžh]‡·ì¯Ï³½OÿÊȯw¹¿6×0“·þ¢ÂËVˆŒnaé©Áõt(°ólaAAMçÎ'ä¼²0Ï6”Þÿò#„ÈCÐX”S-Ôe$Þ[  FÃfÕÕ¯ãhç³|ÏFÏÎ å”Íýdüq¹§kI†míåa¶¤Höw (~˜eG/£•¦¤Cd:½m˜%´ÛâæдÕT*Jå„M+>V¬}t~Ç}öQÇÂ3k_MY.ªÓù6,M²Ò?²O ì±5Ñ·ª;VÄ‹uVù7IÍçÐoz¶me ó8ˆû“ŠÞß’öƒ³Á¡ºŠû.žÏ¤k—ˆë6¼.išxÊ~¥.„•Ím+·Ž7¼£×–uzò”ö®Í°( “¾\ÓGóg£×[vˆ&ƒ¡44;Ë'o渴£Wî`‚pX,0 OÝb;À‘:i¬)Óßœ† dvýÛ' Ê­>”%4½¸¼éù8’²™€NZc ‰8—ì2†v—øöë°ï˜n Å8–ŠC|Óõ×è<ó·ºYŒ½uæ´=5RA¿i€ÆcLæpó§îW±Ö¤Ÿj{ cMÌâ5ch»¬ÿ *ÀC¼eÁÂ: (t%â°¼X5âál§¤¥Šó½©JÁyVqµ‹ÓŽŽ°%’ŒXæº2p&ZØo‰ÊyÜ2ÕG,¾H‡Léödþaiäßæè`üÎ@’þÍoj™I>qX•¦2ST¦nÚ¶Y{‚·×© ?凘j•&EäcÔ.Jz‚$-ŽŸ×Ü{€–ØCÎy‹—~§z6Ðø¤IåÌjõ†ÅÖP“ Õb'Nqö@ â½ ^†„ã$4fš>ÿ~e‰úŠ¼íîéî®1c®S…Gu”´H¿Ò+£5­ó„¬LZ÷ß?x™±ò ð§ú#pü¯$!Úk$èŠý; ‡ ùC4šY˜B0"’× Q Ñ…Ê endstream endobj 2126 0 obj << /Length1 2138 /Length2 15750 /Length3 0 /Length 17025 /Filter /FlateDecode >> stream xÚŒöP\ÛÖ ãNp¤ Ü݃÷à 4Ò¸[ðàî<¸»;w÷àî.s®äÜïÿ«Þ«®êÞcú˜kε›‚TA™^ÈØÚ(n v gf`âˆÈ*Ë03˜˜X˜˜X((T@–ÀÿÈ‘(Tvö k0Ï?,Dì€o2Q‡7CYk0@ÊÑÀÌ `æàaæäab°01qÿÇÐÚŽ jà2È2¤¬Á@{$ kW;©™Ã[žÿ<¨Œ¨ÌÜÜœt»„¬€v #0@ÖÀÁ hõ–ÑÈÀ lm:¸þO*>3FFggg+{k;Sj:€3ÈÁ  ´Ú9QÈXÿM‰ b²ÿ—BÙÚÄÁÙÀxX‚Œ€`û7G°1Ðð– ,)·‚ÿe,ó/:À¿›`f`þo¸{ÿþÛÙÀÈÈÚÊÆì ›L@–@€¼¸ ƒƒ‹Àlü—¡¥½õ›¿“ÈÒÀðÍàïÒ âBŠƒ7†ÿægod²q°g°YþÅ‘ñ¯0om‹X[YÁöHÕ' ²½õÝ•ñ߇k¶v»ÿ™€ÀÆ&Ñ0v´aü Ù:%Eÿmó&Bú#3:Ø™˜˜8¹¹@[ÐÅÈŒñ¯*®6À¿•Ì‰ß8xºÛXÛLÞh=A&À·$w{' ÀÁÎèéþOÅÿ"$ff€1ÈÈ`4‘þDMþ…ßÎßäÐbz?fÓ_Ÿÿ>é¼M˜±5ØÒõùßGÌ(ª,"­¢FûoÊÿU [»Üé9Xô,ìLffVçÛƒçÿÆùoþÃþo©‚èßÕý#¢$ØÄÀý/oÝû§Oտ׆ð¿ä¬ßæ ú3þÚLìLFo_ÌÿŸ—ào—ÿ³ÿW”ÿ×ñÿ¿‰;ZZþ­§ú—ÁÿÞÀ déúo‹·yvtxÛ Yë· ÿ_S5à¿Zh r´ú¿ZIƒ·›¾Í9=7Ç¿Ä {q ÐXä`dö¯QúÏY¼¥° Öö ¿®=3Óÿѽmž‘ÅÛÕbÿvb«€o‹õ¿iÅÀFÖÆm ;ÀÀÎÎÀémÞ;ÀùmU.O8€‘líðæx£è 0±¶Cúë\9ØŒB‰þ…8Œ"×ÛÀýAÜF±ÿ"N&£øÄ `üò±%ÿ £Ìô–AözË ÷½eÿ/âzË ð±•þ · Ê€Q忈û-ƒÁô–ÁðzË`ô_Äöæ÷v…Yý±þëÿßxÿߊ0ùÿB hߊ2ý|‹nö'×[ŸÍ\mÌÞîÚ?o²ú¿5Êüð‡Å?àËÀ7&VàÛeÀø'2û›+øm|þ¡#bý§˜7gëÿQ¿Q±ù£~ fóöÒ[MþH™ÿ-ý×^ÿWüVµÍÛ6[ÿië[x[Gë·EúKfæ·ŽØý¾Ñ·ÿ| ôÇømÞœ­ÿ¡~ãäøøÖ§À·”Îÿ8™7o—À·Ò]ÿ†ÿ³;FŽvotþ¾ãÞë?øïWè4BZ˜µ6âõ7¯òo¹«"r¦ßeaCí»¹BPÛ#pÚUôỿdÍ ê˜Þ ½_Ç+ ”«PZk™p¥ÐÔt›$¹ýÅçÑãXbñýûäÞ!ö;Å":HÐF2ÑÁªôUÚ@Ò^¡´þ8`Šb«K ÿ1ŠÊ÷ªÅ߇ÊÐ~XYŒ7]rküBââ$]½ýu"2è^†¹†›R<Û¬./ÝÁº%Û-kzz‹Ó/ÁÖh¥¤?ƒ óU£)ìó ùþ÷KzhÖ_ȧM=‚xš_ýÊ™)Õ|ë&—JF'G…dK @‰!Œ³Œ„Ш?™çÓ4øJj6‚û­£Ên\­•ä\gÄz:į–ÁòÞ¨ :Â\bÔäŸruñ¨¬PäFaâ}W&Ô¤Cv*ZÀ™¹~׺jvúµ¦èÀ 횦p¿» böôy5ûêÕOØÍÆ9ZÛZ2ö õ€ª +š ³×—Ì•^»Ð ~÷Õhµyjœ5½ºn”êM]"ïN–©‡~þºT6= e™lû¡ w¥eÉßµqÉîÊp†¤()àû4Y²†*z³}žl²Ae÷á´É3öáaS¶°¬µŸÌðôæá{¶š<%¤–$º‹,‚þ\jQ•’‚î‡ÔŸ½³HZŸÑú$µ|M÷ðÅ8 ÏAÌ䋧ʵÀvõ)üë÷™Áå¥U”]© f§T÷ÒÝ^øKAâpž)·•ãÒW’gãj÷r×^Ü)ñ°'m;ÁÔ´k ŸÞ‚Ëe4Úßýˆ&ŽÖ u]9Ej$r*ãTâ^ij’ÙÁn[¡@“*Ø0Àå$.Â9,(êä¨BdOÝP¿»pÒ"!rq£&‚öìu9P i…ˆû1²À™Š $§ê…$j øÚ‚ 'ÕF¸å“±­ôjå…Ÿî&ŽíŒü¹¶ÿUw!E/Vhq§¥PŠ04H_ˆ~#˜Ÿå‰M¿UÕÃ0CNªï3’Êw]ŽgåL®LŸpÜH•Ðâq¤{›0Dµ"úØÃñ(®Ë ?ôS{Ü›9Á~¼°1Hg¹gó>ûzF-‘F˜™×ä{¤­nªó- §{í‚ÒM¬ðõ8Œ¯”/È”·‘äkÊgÌV׎{À_9„ÁÀN“0̽B\a¸¿pÕšÍÉÄ­ñJ2 ¿_>Žú/r·µN2“aÁQÔ£•gJMÛLÝÄË|1± h=¡>$Îi~?UunT†ÕùTðã$¤âðóUú^€%]‰ÚQKS2?ôÅÛœ zS|Ö©¿IÅCJ§¦¥ZmŠ3¶Öe˜³Vó·K5l>7Q#…§b/†=‡´sut⣽A˜(Á#â#ò™ì—0!*k¸Sx˦$•êÙV¨=¥ÒºoXªót=¯±|í©0öàuÔEC†óóg+æ}Í—Ô¸] `½ûÉ1Õ¶gzÊ ïªX7îq ¶”Ç íÇ VìãbdÒqT)h~µwÏö»Öˆv!®«K4Q-ºïÍðÎj?Ak¯Mè+MòÄoŸ°9/¤™—Ñë§ÍÑOœ¸©ë•ié`+X “Ô/ó½B{UâB‡0¨‰3Í ÷ q.!üø¹‰Þú”ùÝb‰a%¥Ñí}Ü܉³ˆK”úÙ ÝÔy¯;Ü\ÑĤ PôC‡Íû´v…!ßÚ¥BŒâC9è8hšuoh¡Ý°ÐëÌ'H]͌Иî‹Q^š‘xkëÆÎa¢ ýòÇ]Sépô–†~a¾Dø²2ÉûU[¾`®œþ"¥ó‹?;ýcÄoœtÐÌ=*6R#š¹°&¾´&Š½æ€tíµ¥-9T$6Ñ<_“|ÍXÔ×pÔ™”`Ãß?ï÷íç¨_–R˜áy=$Ï,UúÞ‰FYuÅ".ëĸ÷ŠhÓ†gµULËí; KµDhÞ ŽH"ÃþÙãÑÅärøa~Ùz¥Œf¤qª—ˆ–C#ýCW¦Š+Mzj®!}›¦j€I¡+wa&à ›ñBY•ðsóg•÷á.0è›>›³ 6[£à›]…y€‹-§ÐÇÎ5ï…zåØO"îM¥˜‡¡ w~Ô×ÕŸÃØarÝܤ‰ô¿V¯„ñŸÖC+ˆ.trp«ïíâÕrÁÄá$ EÈŽ®~½Ï9ÛÜ=rÚþj8v X6„bB«z³âû]xfêñ^QSÉwÈJ:q9{€êQª¨ºÝJ®”<Ëî¹æÌx­ê7E¢u]÷%ÍÊÊ?2×R;FgÏSŒêëN»ë¶zá°ãèsª©û÷|TMþŸ=¸ãr&Ö’¾S¬"Ùu ¥öÕÛd IDÁýÜF“³ØeÛ-ˆ=4úG«cí ‰äwãÜ(å fS3ïÉœ*Û9ûDqn2Žäý9hÍ[ªæIçD ŸFîνª„&6ÞGhôyoûÁ*dÀTÃù>U¶´d\¨t,Ÿ L¥°Ô¬{(tƒ/ûá;ýqZËi!,6Ç´lÜ,Y[|mYƨŸEª7»7°}ŠuÍ’â»p¹9(—ܲ “î/„6KxO£º&?çüxÊ'ËöS:l× žü ê~.ûã±å´»éí¨>U+,x:]æ®{´sõ(ð»"Öø/÷cŠ‡\B[v D1\¶‚XcŠæ8vÙ/ iÙ‚ô~Þ¤~8l+NæÓzþîœÅ‡†,„žãjxH9¢Hc£Øºâ‡n9º·hÀ™. ê\Äeú Ϩ-LíM÷+‘ÒQ÷F(ˆMY^cHÃÏ/Ÿé8éd$s¶-õ£äÞŸÀekn Ǹ„†Íéhç]ieÎÚùÞÞ<É„0ø#3ó?ÆÃ$åœ Ø!ÙÖ}ïê8ASh0ï1f2¬d|¿Ê‹z½z«ÒfeÈ{r²+¥ª¾T0܈EÆ]ÏË ~âÞAø¹Ñ—§ÍÓü«‹ÏÕšÕäh´W^e” Ö¿µ v¿-a™pH¹Aî©ýÙ‡_º¾‹T»`ѼÉWqßd3Ý/bYìyô¢þ2TÖ“Ö>Uuø–¥~Fy×0¶D8°$$éãœ}ö*Q ìb§"¦L‡0äñ— 5ªVƃ åf›ê¶èLb5ÖøB7/ß| 2Ï4ûöNҊ׷øéIv`D„3”v¯NKïPdfÒ†uÙBû>æ›ú4#b!YÖºŸ™ÓóŽÇÏŸ‘I±„ U™Ë=ûÑ{î–æåŽ@Ä_UŒ"Ü>Ö¼ÅÀýQÇ+\5i10ÕW¾!ãÊÒu ×JOFI–¼«½^Ra›Ó¡R‰€ )·Ã¸ è$‚Kؼ»¬$öØB ŽIzÆ1ìŠñå ÓøÑŽu»¾,1-')3…Þê‹>ptßN:/þ^Þ]éÈøJü{¨ÐÑLÁnˆÄüK}œ³…³¼ªû-Í6Z™o¬÷y=[¤"(¬¿¸5b“]bìUŸÎ@ª¤·…"—Û¼׺°´>×nQV,vCT‚Dÿ”á.hògê ÑÞ<&™ûŒõywØ´/“å}zjáa†™ÒÊ_¦§Pè½B{™åÜÊêMè= †Õ_…{à F„³eÇß%áQíô·„ºåà‡­¸½ß]†òió©€neGïúÑ(Ô/b ùuÞWÿ1:¡.6ƒU‚êÕmØÄ 9Ûñ²{:}gÄ媰¨õÙ—åƒíµa`˦˜LOÐnðÄUæÿßèÊ—MìÝ"vMÆ*šâzi_=Ó9† —佑?÷ç[¦%’EWá$ß¾:ã> CѺ»1ÈXa”·»œŽ7ÁÌuÇç± ¡ÐæÕé¿Ëßr<Íâ¿&íRn}ú,³æw6XEÛÉ-ïè’—hΰ‹ëyYænm±˜±qyÄëÔîÔ‰ÏG\ ö-Å‹zNLs¨ÉcÙ›ÌÿV±Š`S•3:’7 qs!!Æ~³µ†EFªqIÑǪe/¿à‹MO>*áÆÈKÅó[Ý%‰Q,‘¯è·W-# ˜eâû}f· ZYdÞB¤¹a•í á ç9¤F‡:ÿ§(ÿl”˜ç3ß¡8¼ƒ‘!^~o‰S¢îe«Ï’O#iÌJ™¤‡µPBþÆþé2'Ïôi:³09245Q¿%Éðöáø'¯Â–ù/ÂË€ÍÒûÆÊBšðò4"‹àü˜*B ·nÄW»Ã9æñÁ¤Ÿlö8çѵ’ðc‰é– Xï!bZÍ ÞlØ"Kú`ëÏ«×Ä«8V_ëÊoøD»b‹Uÿ×ò¥ÖuW7CÙ³9Wœß(cCÝiW`öw钺ߟÑýU p]°µãË…©'cÌhLé–¬:m°?¤ãlêк€tšéî©â>îÀåÕéü ±j± *cÉBñ’$-[ ·ïê£Æ­ÁOénÙ‘ŽðÈÓa–„Q†ë40£Ée¤òäprX BýÒo}~“w¦·/²±™5^±ß²Xqž0È¢#MÚEÊdã£ù[å˜rì(×UΟ_¯Ð Ø(ƒ˜E‡ à5§Î‘l€VEãYÔߢ)Î$Ôå·ÎZ]Àu²þj+ [K•Ã̃J¯xâp=ϳq©^•â¶˜íc1V¤ðb%¼{xyJIñ3ìÇ£=&ºúZ“›ø²òFê«‹—¸³Í«â$9ÇîP1¶ÝçyÛd-°†_ƒx¶éÊ6e‹óÜMÆñ1>ÖSS\‚OÝú5ãÁê™ÓÝNº«#°nä<ߟ?íR±¶wÓ+`!©/â6Ùb Ì À­¶Ðâ!žP 醒*NY3äD9Õ5ÇþÂÆšðì‹>^.$Ä*—”~´}›Ú¹B6å&¹òáÀ³²¯öÁ w¤»r¾nÿl™ð(êUR÷ž³ºdÂP%Ïy·ÒdA³eúÓÇ{º}"#ýuXHL_i¸e'ù<‹šÿˆA׉¹ çøë¨Þ‘ÝnuÔ¬ÿçb!~b ÏžX U>EYœ©êŒ»ôcurÞeû‚„^à‰äøÀ¤tÿÇgTÛV0ñ4FLs”øÁf¸èòR]9Éò"ÚôÒMÔÕ¼_þn™¾ÑžFq_ ¹ŒÖ¼Q«óŸŽc˜~e³Á6MºÆØÇßô¦?n/–Å&Èþ˜zÚk Ù?WvùŒX°Ó‹¸¬òQz²£ê:¨1_–÷ÔᔪÚù½!”Ýüî€ß·i~“}e<…þc] sVÂV7oÿ" ±ISš·Õr¨ËíU< ºø<`UŠ£‹ÇØtÄN/Ð)Í!€c&A‡{;ÏðX‡Eò*Ú®+¶ÆnC‡êãáôÒ+½¤îO¸¬6<~TgË›ù°UÄן²YÀýZçX™€Ç QV›žÜ|dR;P|Ÿ‡ÛÊ׉K. -µ3þßcŽ²æ I¼x¨ ¸ycàž¦j™™üê÷‚ÅÇ{¿fV¾gÅ¥v¨­Z‰y_K®O=ß;på*D$µÉqêØ?!MŸÖh¹×¢|:RJÜWˆø˜­ß¬Ê³î;ÜŸÐhƒ Ÿy›Š#X5ÞDóRŠN+~—ÍŠs“ªëçŒEI!Éž-ªhŸM—o”¯˜Äù:°(Qn·’î l¥ÁÐÙúpo$…=BåÃÀÙ«Ä[¸Iù纰ãMÙnÓê3·¹NÖ¡ ýb“(I:¨’¥Î³Ê»?ç!ÏôÏ6Ø(D*}ýIPÒÁÒ-Þƒãî~›cbÐæWD_·À{W{/t$°8zc5vùÍj„Hd3˜¤Ó’ÛvÓ²æ^!*>…ÿÍD§µÕœåÂX8Ùô°¥ïíWj"‹R`M'ŽúTµï(ô(ê¿xu€©Sœì«:f/è¬1RW}&RNê “~+ÑI¢`J1ó û ƒYµ=–펿ý¥Þ­BÇ9n4¼“hÙKÐ&`]»;u2»›W`,ïZ`£ˆ >™êðVÍÍåTO€óÅe픨ÍT"0¯p걉,¦^û™`Ì…×ñi€Ö¯0Ð8ýl ˜Öá‘:·HIÁÏ ™eå]"PcZ¤Òð›RV µqUF|¬ <€I^;5‘Õ›¥—µ|>},Ú(‹$…¦ÞŽšt ð}§ Ã(vóŒÙ.ŽÜS·AŽÉÅ_J³·é,{èúµšè‰¶oøÃKÖ½½  ²K»Ç‹´Ãp‘jÞÆ\ïüÀ}J¶<ËQ\Ì‚näÀ¬^וiDŽ.i>… œz×LÒ%ìWWd¿EÉØ5æ8¼yÄ„ìv®²úN¡^ŸQ¦íàèôj™*Ý­9ìÃáÛ?m.ì©ÖH8“;ì»v?Õèpä ñÅ©£^µ9uÞ'N¦‰ÍK pÛè/ɵó(d²óµ¦µVŒšˆR_˜?Ì-LÍÔUÇ@mÔõbrí…”léjá ¡ÙôœXën6+B"­è.¯ÓˆUGí†L·3”Œ‰1ƒîqlÊø×l¤?!XîH/ú"qOZÑ©Zð­X©Ü-a¤hÊ5´b&%óïfé]øOˆ›"l.”è9£·3uG[5ÉIßÇÈÜMÝ>ÇúþÈ¿ ÉÀu]2ûòj`LÂ%ž°Õ·dC?¦luåbÜ´¼œR®·`¤Å¤œvCó~¯æAøôµ…µ"왹%X9É’´>¾Tº¹KTý¤š]Á7—±ËpžNñ«€ý`7ðÆÞ’±“4@"# £çY#{]çùþÚ¯A‡iŠÌøév7¤Êîda%êÑ#Btü»:cþ ¨`îIœì·¹³à…2EA¿¿;æ磢€Oaïµ¼°è¯¯Lx)\+“LKÈA]]L‡À§¤Ö1A–žÒ¼««\zMŒ; %©¸”Ÿ?iª8ã+ÎÊÓÇ»¨K>W!»3á\)ñ@gt ´ˆÙ„ãxˆü(/:U3 ztgd- öq«Ü†GF·EyÅ(†”óiìÑß±7Ö-:µlÝo¹Ç1bO5Ç¿ÀaTv£ñË# q~ÓO¯Å"D;Yjîx¦ ¨V8Õº5¦jfBìîç1˜·Íu¤ãò›UÝ´ñý5ÕCRxäaªaEzW#›¸Ô–$q>RÛŒi¡£+Ž ø RŸ·!jïË^ù²ÆRíªÌ'²4|)”ûÕçÔÆPF¯RážöÊd“ 9V§@'š8x/“ñÍ„[½À­––õýzôÊ…Å0Ô!ÅH«ôŽ_Ë9Ô첈K‹´ëöŒÞKýqapW„aK-¡á,N,IÓfÙ™&ýŠax¥Ùé2RFÀî´“¤‹þFÀ¶Å#ø3¯\þ³ÂÈ€‹·A{ÌÓ}#1Ë4º7íðÞö±"6ñ¶¹2]aÛ*"ÑájôÒUÏ3µðÔõ‹Á>\üŠóù6+eQç·¡€gzb…îr| ‚^ü`$­ºÕÊ‹j3Qïd¶h|bþ_kªF´-…Ú[7J „ÒÍx©§¼*¤;Š£½°ÉƒAŸÀs™ëýóý¾ñ=cÌËlJl6à mï-?2.¹‘Ij5Ïd”+ªà‡Ø#tœâš U¿í †|–|ÝUâÊh“áâH\åçY¬ýË®*ƒƒOz#"\¶)"S`©†W—ëºp…×qŸ÷ÔS9¡:ö_Ì5k0‰¬Ê>vËô—{ú|H牽&—Ëe«l:îG*¡\Þq¨Ùh µK¦'jŠ ´L9:Ì6 -Ô*çÀŒ»þ ‡l‹¤ÑhõÊN™ŒbípT·²7%-D ½?”ôÒÒáç·ÔB<y™ê$}¼^!_xKûn‡0Ñ'^›Û9tü$Åd´âke¦|mýòof;öÕ0UqÉ)ü€Ýû†ìù ­¸ÑúÕlåì43Ý=èN>¹ì¬!p*µàu‡ðŒ% ©ù~M—ådÒ£GËßÇë°åÆ“wSFª a¢>ÝѬ鄭]_Í]FVâR€0÷5#ý›P¤mè ²ÌæÙ&G‡‘|›®%é©ŠÄ`£Ž?#²z­ê•ˆÆ&Ïl+¼Ç §4¢í¨~Îz­OyÁƒxí¸e‘HF’²å•R?hÀ(R~]y"¿ì~ÚÓÆJ›lhª]èW XÌs¾ü¶+º‚­YP÷ýÙJ>³·Pñ0½ó 9xňwÊKh¡‰Yþ,·Lú¾j(–qþÖã㦠ÊXØ8Š,Œƒ’¡¾üF¤Ð ZÄCáÂe¿¼­±rdÆÙ™²#â¹ÝÈÁ`ÈLhzüé|9ž:qnΗªØó¶X <ì[èIaä÷ñ5ÎNqV4jÎäDr6u¨ûŸ²ýQa{ÍL(¬?BCA¹ÏAérDæ¬,"jÅF~÷kÑ~¥7)Ÿü$ž#š¥/ß©e!ò>5l ‹‰ßõ·CtjÃ;Äâ›FžC„ô.âDQˆ‚®n4e¢;êàá<”ÖE¾15âF¾!.¸~}Ô”5bfƒµ+’ô;nh Ë„ËrI˜§Ý¢ä™²Á¡uƒŠ“²`^ǘÎæ;±€ûÃ–È ?bUa–Ôša>Is· ߊI=6ÁHž&?òÅ ;ζºùm×oS‡+ WƒŒ 6ûL>‡Iæ:ÍÃmÐUwÁú »8Ž’ÏÈÍ®wN>É8ÈXÚ™¯%ƃ‹ü¾Ø[ï¬üzø.Á‘»jûÓ1‰§ôÍõ¸é çÆOÓãlOw¥1ïüï’5|EŠ§5EøTÊøX#ü4õÖõØü1FÚ™38Ì®6£åì?ÏÞW¿N~P VÉCgž:\(éJ+ô}6?vÒŽ}•^© €jòvAO¬×B`qˆ—PßçTì(úq¹,ìð*>Àšo?ÄE$ãHu|V·Ól¡å¤Ë¹¼ª4T¯Nußáà¡á9»`CÉÔ|=¼ÝWUŠÞ;2hzÒ¯`ÄŠò®ûŽF>\f÷»a.vC5‹ÿŸÎëz¹­ß·ûí©—¡Nà9+,«zAtX_ǽºù1°(2º‡ëöÞ³YK;þÐ~}—°QLvÚ M-úéAôZ±ä

ëE¿Œ’·ÃcÓk[ÛõFm¶%N½‚‚µFwµâ$ðµýèð‡×0¸¶%¾ãÂ5‘ìÁ˜š¿ŸÍ‡2&œ¯µû&#*1@BÌê0¦>·çþí=Š²%z{s)žB3PÔ¢ÎÍV5»tQh›Gÿ.÷‰øÇ­×Å âsÁõä®2"¬ÖsÁ<ŒôÇ’=åínÄÁ_ç<ÙåŠÇ®š)4¬¾s±a¸½ÄÒ(…µCLVrž?Nž)¬‡ùj\Xb=ÖoÉ÷˜c’MÏe‘}úÄ%ïÔ«Ìi/êP¨‹BÆ;/>+h‰ŠYïÃ%0¥é÷ˆŠ¸[ ©¯w‹Ú¥Å¦RH*¿ïÃ,VL€Ñђϸ-¸ÐåÀ¡žìÐbU©‡Œˆb"‚´ÀŠL7§sß+*ÐhûâgÄMaK²uò业ÕØî!»å }‹>^+[vã¿æÉÍ” ŒIÐdàÀÏ÷då:îcgêµ]Eî­@qï4gÿ&W® UCl‘ÝÑ#$‹žË-l?ð}e;>˜·Çf¤ðN ŽM½RŽe8´rž™IºxYxäåFõˆ¨°ºlôš½T@Æ}äœ/˜ÍÌ›ž•Í0ÃVb¶U/z]ŠGé¼&þ„­N9d°fHÎê2»ƒËª_EÞñ4øE´IŒî@ëÇ}‰”½I©e¡Qu?ÏJšÖYnZajCæª$¸–H§¸5õ'½bs&ËÛ¶V¯™5g —ÿ´›ÎðË€%ˆõ‚ânª÷Ù7ƒ¹$D½ˆ™&ˆI!7QøÁ†nì!+Õñ*¥÷÷ð®ç5}:pºFt º†Rz®k9Å=0W€iÏ¡SñœÏ†ZÞî(SŠ‹çž ÒeÆÜø×ìÁá׊¯¥“õòkäªó9Ñ6釙0Å$¿ßs3:I*K¶û˜›³»÷4˜sÕ.,¬×k:úúÏïW/Ap`|/n º',înN3y:L°4ie¤OI%i‹V?õðVô]ÒçÈrM;Ùr,¬ ÏÞ{Œô²ƒ³ž•Ä=§¶U>;ίØeåhÀÜÔ‘cþ¯t7çFZ¢’"aZ}ÐmM¶°KXå½1›º„³ŒO¸ݬq–Yи_â¸/‡*aÄ…Uø„ÈÂ\Q¡†Û8Xr £/ºF¤9m’a =kµâ¨U6Ún*Uê‚š ós1 M _W³¯òR;ºVsE×?„z³a/5Zt±³Û{ ›ø"ª€Ö‰Nk¢)\’νŽ·\fñ¼Œß)ML}4Gr2å¾!â-©æô\ç”RÄuž•ãx§D"y}aú[ Öù–¸«Òèÿ>21#±n 9¼‰oÀ¨Eò69=;¹¢îÊцÊæù}3j &ê’µìjc*!KiÝ]­ãôÔo]¯|äŽî8¢"+òT*–ç©‚J}+G&1*ØL|¸ˆOø0§Vò‚@i ž™ï°Ë€E…ߺ‰x­½…þÑñýõT¿ƒÀ­4ˆ’TêCn„bQTØ_uü©Ûé§&FÌÆÍÒÔêåêêÑx¦¼‡M.Yã— r'¹(C>@CÜð©‹Eú »õYõ?õ ÿfõÞý|'$½"!Xö"b­W-o+™×NTR½ÀrG‘^(¯|Ä%Áeº¡×™­¨ôtX…Âéf™WtÑàcz"‡ugpª?ÍRÙuÀÍG´¹Àp‘]\o‚µ"‰O¡ÜP]o;è¼ZCûr–Rîò‹Ú8§dLG@Ò/{&m› 9Šˆôèûá*ì6º8èwèÓÖ ºš¾ŒVhyœmm–Xjµ`=çz<¡q[üVZº`k" ¥Ù¯;Wk7ÎèÁTE·Lsw·(jk=¼±­–­©¼|»ðäSŒ ¹ËÒ™s•Êµ&½ºÕvJçL&{ÃûèÝ…œž¦á&FGE qâU¶Ýkðò,ê&ÚT çÐ5ï¥s¼>GÐÐFÌîÉOn:)µ:Pè]—ü<&ï×fß vÕzs‡ãvï§ É8©ß»Vðùwkb1éB\î?}dN™Ø¤†°©ÑZöÝ6¡÷ÑÍ:4ã‘KñM­°øÌáLÒŸˆ”¢æ`ê,s€¿qö/)^3Oï%¢.è:Õ ÖÕ&UÆSPÂíô»Ê4HÌoÏh™™1'Uê÷óµ$Æzú~·*K$nO#¬»'¶ÔæöŒ¢ßú{áÊ=wiºâ«Ë†›.*[KÈc’/«¸˜<+8!ö‡›[]#Œõ=ÖË“`-¥º¬lÙ-+N"ƒ‘ S¡v¤*´hù[Î3“lvÓãHº%•½¯5óy~EêcT¬ŽI@ÐÍZÀQ¿ÂøDIæ·\ì’"þ4^²Qb&äÞ%ȨëPº…Î!f°üÕVÌQîÖ.ß·,»3š ÃP}‚,Ýú ³ú4KÜ^h^V’õufÚ·³jú—Xê4ž–í-¿°¨O÷ ³É)_ƒzë Anâ(ÏSwƒm¡ŒC²ü¹÷×·ð)–:Ïó´?)VôMâãÕ€•áÚFK`ýpCú²sX¼d…uz™GãJ;)öózP¡WÈÏ^g÷§Ü^ýñXwÚ„d †ÖÜA°%â)Š-“lûVI5² |B»Pýñ©ß‚+)G½]À®ˆ'=cÇêrÕ‹´ÒX?*ô„ #íÔj(í7IJ÷Ÿ§ƒB éãdâÎU«…ší·Ï–Ú»^cügã‘evt¼EìXdc ~ç†vPù¼B@FQó1˜Ÿ>95`_wòšë?t$y—/îfPò=å&-h3IXE9í/xY!ðv­ÚØfË®+c>ÓؘA—¶µÂëã&)ù§ô­®|&xiÆkgý÷³Iªß³|<†z}˃V|ȹ¾ä›=wBâÖwg»OåØ^t hMÖÍÑP i<(Ý~þͲßGêMÍ¢<$È]©¦ÊNà |½|LÜþ½/3hÄtXÚ~º½ cÔíäó5¸ ±–¤;4Ö¤Ñ\Ð ­DšÀuÞ  8¼¡¡÷ÏÌU—q¼z‘é=³ÁúÜÈWâ„Q»³ù9¤¨ZHÊLçBÿ»…rSÛ´ŠQaÎ{©6è-‚¥nñ¥G¿¯jfZñ(A9§Ž$†»q$"šóžï|¿‘ Ù}îòú‰0¶Ä.úœ²@æG´µ¬ZñIŸ »‡á )ôEœ¶×=æ,»&wI&å¡¿ó™ì9Êè2’P“rDìtϤ…”Ooì0õJ ‚º [Ýœ‡ÑidøáB(Ù/ë •\—²aå~×Âñð~öüä|^&j¸¶|üÒ`DïâÏ‘u;†®ThÌHÎ):Ñ£³mbà¤ýÆm./Ù­ånWPO^ÑDd<é%è‘ì‹ÓÑûÂj§@\^Úø¨óFKæùæy+[§ñ#ĘØseI;GH6œR†Oí>w¿ ŸÉ3»á+ˆ2§Òeçâ;ÓÐ_n`v¾1”ˆg²‡o}V=$[°-ñÐ `fZ#¦²!–¦“ˆ³öŸ«À9vߎùØέ))ª”\/£%$súôˬªM°Râ]e¿DH»°š·{ ÀçU¼œ­€›§¬&ªaÊ×l8_ZƘ–ï9(ÐQ)¾Á¥«ê\xÑfãâƒä=Íþ­vyÅel¨(~p5…{-’½-ÛŠ†‹z±drMIAXG¯~ÙnïÛåNÒw>©¤Ðx´tAäßÂ]?yí¸oôÔöáœ-Òuµ §”ÔNÄ‚âcÌ旆ĩx/L#àip<Â{pâ9ìh5˜­¦kucW«.ä"yØsÁQz“e¦Mþ7²xªw6SvüaÊ Chµø<ŽóÇΣk*9ÌÇÁ4Åtk#l³qú«×.iù8V¦’Ä.c«gk ="ŸhÚù_ÌÍÛ3˜s 6+«i¤d¤@ëaÅ;ÙŸZÔÜèëÕ:Wãrf¦W6XÏÜËîAÎå ]%iÁ9­Øƒ?oD-Ÿ~â6é.^pñþö^˜&ý‘ª V|Îi‚–†ªuC§8Lv“'ô—ÿóº)Š–ñ—5N i:—s0òâê«Ižð†¥Ú£Ð&U¾˜2ßuÔN×?I510³Ž!ñ–ru‹²/m,ŒêÖÕO&•æ̪ˈǤƒ,¾Yèá1ÐÖ„|§;Y UIÝË šÂ'áË;ZhóKi¤'Øã­/Ÿ‰¥G'²{XÆ4 pOzP‘£ 4pn!ç=rս듖è”ãj§àT)ƒq̄׎©YëÜéB6È”T]_„ÒRÓQ6@왓<¹¦õ@Žt¦ ­L¿µädq;“ö1C•‚+0 ´ÝWFB8?ØL$‹"øNöknúF„'ÝYèä=Õ^ømåÏAd–åy…z:_OWà¯_¾V²/—Ï Çå²T–è£'%2Ì6 à9žDžIJÔ¥¥_X~ÐÈF¼2eê´ïE_ó°ß¶ü*ñ«·A~ï\s¼’gÿQØ]wèç:f8?×~õ’}üؾGÝléð­“RŒ%ºì’ۊŧ øæMRÝ=Õ¨è|9´nj×1ºÛV1Ãǯ!§†—–fŒÞâ‡ÈÔòœ?c€7)¾Lµr#x¾XJ'ýHB0œÑé²Øø>ˆÏ)Âkæa„y÷Æìª!÷{ ü9'‡dÍÝß3ÿ˜÷V3ìbÝÌxâ‘dªZ[’³ô~r°9 牘5¦,žrn7g³PiȼêQø Óöˆ–Üñ*îp¤$%¶%éǨMô5eò²Hå›Áƒ/ªðèá0Ÿ¡Î\Îö.ï—§#wç ‰.«Ä9îÔ{Ïl@hæúý»&õ$ÆMæ6wBUùå @5ó™Ž1PØÒ$¯C‹öu(‘þ4c‰ ‘Ù—IO_!n¥»­o¤¦ñÜs!E ¥îàN ³sϘ”ê+ç;¨Ñ[=."5-ËdLȨÑÑo\ð: _aM΂5/®00;ùñ™ƒ‡þ½^ ¬>eÙk­9nƒë¦‹V^Ç¥ a'…¶h.ǡæ^ ŽÚJ’}¨JUr´NQ”Kþrx<4"DG´‚ÇFŸB3Ž×Ã;Eßs+s›°‰’&hØ‚„œD¸·¿¤QÚemDÂ\;ËÝÿó*±–ȱ 3‚m°´¾åBÕ$˜õô€Ü*ý$:rþžU3@ª \ <«G¿ÇèTë”a›¾$ 6áÇzf^6±«âBmÌ*±zÀ ¿Òý0Ô O˜V¦´³ºi/¡Am_ßy‘—´*;ž›òb¾¸íÑGø)j¸¿Äö’ÇI•”BLl…õÐíw¢„û7ôãy{ Þ-úÁ¶–xã<¸Èls±•`úfÇLo§ã¤‹¦ ¬‰™’ñº¢ý(¿Ï—-ý®„·w ÌÇyMøŒ?vÓñè·òûz<ú^%P84D]ÈÈoF‚F(Ãü˜èGÃϺӱˆûË[çßÚ.Ú7ÍùWß,aP;ðÕ ·{$Ÿ¨v¹hw˜ qJ¾Û©MyÑ¿«“×s ¶¡>éCH) h©Paù1È·À†VÕg]2ÊÎ{‰[V‘[~>+‚5o¶]§o] ^º,kíÂçÇz|–771:ëŽmþIB¯‰™y„s§§{ÏÝJ2à&1ÕÁ6Ùn¾¾˜±7, x—ï!绕*‰±Pv·T̯̆ËV½%ãK}fÆÒ«ó2[j%&)¤-Þ›UâÚ°” ˜ѱΰ¯Pºç×÷µƒ¤âXøa‚[•¦íŽƒ7ƒ„‹ñp”ÛTü“ð¦¢Ã Ž#JiP„ˆµñg/ÙX[šÁNÇžÛ”W»Aoñ@#Ê)±ŠR:ٶƊÃÅ(ÆïkäÂd,3#!³­ã{Áúy?ž/¸£ÝÈ:dŠ‹5Z»åY¢´­‘7ê ‰ïÙa¾š"%jÆE#€˜-^ö.H1>«ìtc"Ý~«³‹Žô Ìd=¯F ãÅ‘¥Ñ˜|¦Aw`¨fŽJìI^êíÜ„0[ Ï—?ÿA»¸6wQ–ã±JŠ¸cÍáþÁ¹cV×uãØ9‹à_ò!}RÉÖ1l ±¡Öb‹’´Ðª èo.âÌ‘a2ºŒ«DÜÌí ÛoË×)Úþ’™Ãv÷ñ¶W¾Á“Æ I'•äì]|Tw}ž!ž9ñ-+¢JQJ£**!ÃÒG˜ÌÂJ™dA‘kÃ>Œ2´½{:¹ÕÐK}ø°qŽ¼‹ÕXlá7éü’ ¿°Öá}1:ý}¡¾~QûTvæG°b$÷âŸüzž²pßD§cœ˜×¬ô ãþ¼2õ‘I?–¶à†Ü*ëÇXG*+–ôDz.o6g¹-æMâG˜,r£Qs„Š•` ø¡Ïg!NñNn¬ñSðgV˜ì»R³ÖÑzv&]hËŠß+AÁlg"X}öt—5“˜i¿øŽºm©V˜…öoÈû-±êƒFϲß;r¾‡»–É/°±P3|õš.~%Ëxª°†nÌ SJIK ¼­Îë9¤ÞHÉ‚T™©Mô0)ÞdÞÓ̳ÄLgBf–°øÊm)¢ß Å¡Ôê¾ý%5Æÿ²{„©íàgÙ缉CØ ¿“½Ÿ(8p‚fM§Ø)Ö—?!maà‰ìÓ8\2þñcàÂ7’cûI+)Û‹ð‘òÆ\ò9ŵº€D:ò‚ÎÄ5†­Z‰¹ ǹGòŸ}õ‹ÕžßÃ$¦¬Ucó™F šL“Å!C9BÖp¿4yEQ-‚U¡»Âß#˜S«ãüŠˆV.â¯]Pq§%±+1´ËÂ|B+drHÐUGÔ­ã€æ–3%èMTaÄY¼íD«Fïô´ ˜½‡uÓ¤NÒåÂ.eªàvBÙw ò¿BåÉHîºÝ~+âžœW@ÅÅO„›m4IϽª.j½žU©³!çëD«X1¯åV-zìbîãHû0inÆhŸJ`§T«5H‰­á¾­} PzðŽ®Ë92ãò‚F±ÿþÓdz«KÁH:²íÁmüÄ®DíŽ-^ÖZ¢WÝ„AÂT¾¶CÁMf0{ÉÒ†uœ€!q!¥ñÈ€ßéjœ¶¬‰†Ê×Üb´ ÅBW‘ %ü«(ìïú¹øÅ€-4„Í r”)¹~ˆkûñE@S¦ý¶íÑø—?#ýË2÷2¡™ºáOTX¯HÂÂËРÀ'çUO«¢PºôñAP*ÑÑ£Éàxbîÿúþê‡ÒxaÒ¬[ŽÊN“Æ}FŽeʽ¥s¸"£žï¶SK‰´®ïLzTœ[¿­üÃ"¹7ŽdUW¾±—Z&ÓJ Ëÿ8î?I¤sT|€P K§z 5ñÓGÑðÒÎp!Û,h§ø…R;ˆš¯ :Ë6 :³Iˆ‚àhóÐnUfR3;Ì4¢h1Ü50ð&®Zƒ+¤NŽ3pDŸ‚?…ÆB¨Ó¡–gù520ø-ÙU\=BSι¬´]Ѷº¬Çvåé.&,!´9½Õî¼õñ+üKô@tÕ… Ybx²Îv€²“¹hÒçïŠF¥;ß\ܽ *U?[b‚µátícU3}«vAï•‘ä endstream endobj 2128 0 obj << /Length1 1596 /Length2 9062 /Length3 0 /Length 10105 /Filter /FlateDecode >> stream xÚµT”í6L·tJt3tww H0À3ÄPÒÝÒÒHH H7Ò‚‚„t§´”À‡¾ï9ç=çÿ×ú¾5kžy®÷ÞûÚ÷0>ÓÖã”±YaP8'P §¡§.ù¸€@^,FF}Ü ü·‹ñ9Ø̓ŠþÃ@Î ‚?ÊäAðG;  êáàáðŠò‰^ Pä_†07Q€<ÈbÐà¨Â `w,F9˜‹ÄÎþ˜æ_¯kVˆˆÇw€Œ3Ø b ‚4@p{°ócFk@f Ã}þ+‹¸=î"ÊÍíååÅrv点ÙI²r¼ p{€.Øìæ ¶ü.  rÿU#@ßâþ—\f ÷¹'ˆ5êþèáµ»“ôTÔZ.`è_Æêpþî €‡‹çßáþöþýã ²¶†9»€ >¨Àâh)ªsÁ½áÔæ·!ÈÉöèòAœ@VN(Êè@þ]ž»µÄîÎåqú]"÷ï0]V€ÚÈÁœÁP¸;ÖïóÉCÜÀÖm÷áþk²ŽP˜Ô÷o` ÚØþ.ÂÆÃ…Û qõ«Èÿmò(ÂúÌ @!!Øö¶¶çþ^ßÇüGÉó[üX¿¯ Ì`ûXØb ~üÁòuy‚p7°¿ï?ÿ°xx6k8À lbý'ú£lû~¾Ä`|äøûóï7³GzÙÀ N>ÿ1ÿ3_ny]#Yö¿*þ·NVæ ðåäpò <<¡Ç¯ÿ‡ùwþUü©6ò÷á€ÿ‰¨µ…Dþªá±yÿªÃóoZ°ü½2¬€ÿΠ {ä2Àòꛀ֞ÿçøãòÿÇûßQþoÔÿß)z89ýQ³üÑÿÔ gˆ“ÏßTö€?®…ìq9 ÿkjþk•5À6çÿÕªÀAë!µ{¤8§¿à_bˆ»"Äl£ [ÛÿÅ£Mâ1… Ö†¹C~_8N ðtKgíøx©¸?Îë ü¸SÿVj ³ù½|¼‚›ÈëqüHàËó¸¥6`ï?ôpsAaðGÀc‰þ[˜Öï© ñ¸U~‹þBBn#a €[ûßè‘,Ü #žGšrƒÿãØýò¸íÿÜ@a·Ó? €Ûù?ç1-ôð1ìð1²Û?  €þøÊëü¯.Y{¸¹=^!¸üØÂá?÷ì ¶Æšû³ s¨ k»ª‘¡ôâÜçå2psŽa¸¡@á¹­"'~}ÆW;+Ó5µ{½B¸Ï[öÐähD t4_öËŒ%ûІjkµÎá7A(q³ï®îÁ»¥ÖÙùÂ’jcGzôQwÐë\ú++‡ëÉætö®’8ÓǪË÷öGèÌ\8Hãñ±èäÂÂ|8Haí2ÙíU²µ2_§:TÜO¥ÇS¨w¯3ûGxÖæË9!«™”{{O\Ôγ÷$3vzŸªYN¾2nô(ß&²„œ·…³X¹êIÑ‹å¨lLœ*ŸþÈ1ú€ÿ0™y­ÎýA„Ùú­ÕHK¸‹oC£?IÃ1öè’ä“2³³®å4 †ÖOã–8éEQ‡Ý¨3¸Y²ßöQKŸ92S~•væ([r1ÝK1jò<„©Êºxõ×Kkþq'ª(Zè ™™]79‡ÐÕ¨M2êÈÁƒëPsÖ´êmq×µg‡ù2î+¥Á OÏX5 J ~>“÷¥´®çÆ÷ÈΙݧè}Ë·ïMmA¹p’ ë—•eÀó!y‰ûIS‰®Ì“hšUú2”=ù»~(•Rv¦†=³ò˜Þ×cÕºÉ?}dÎrtÿw­¢ŽÝý%óÞtÄN‹Èá÷I†Ásý”Xý_+üÍJ/™9^"Qß·ºóc„YÖI/×ÌG~âŠb.©ÈªªKk®Ž'xRW°ÇAp:v—;±ÃÞ§»O+E 3Zp•Ñ>B,@4žú†íSLœÀ‚“¡“ Þ,ó þ•Æù£8e“¬-(JÍ).wÕœ¾I˜z0“PhքĈa®ˆu0VÂM̋ŲO"RNâ uÁê¤Z½LÆܯ•ŒW¦÷º%µªJûÊ⨼6LQ»&ÔÔi›­Ç|ðb“§6'Y˼ʞeª”ÐâŠ*N4üðÞsó ¬‰_泚÷¡Tõr¯âL›ÌaÆ2+¤êXšD—Y—êi'÷€ùqìWÈ>m8cæ=»ØóØb™I’1ôÞy27Ç]È$çjêœ!MàZ,Äp1@f¡uXF\[JqM¼K'†)­ÀKÂf¿q¦¡ö#@Z­>äÓ‚õ‚b?ÿ[:%]Áà^4խʦLèÌTô• •„ï|(†ì!a?B¿¹M½ô²6ÕÍaÒ0ñ|ƒ¨l“¯ÂÊ.j2˜%÷Ä 'p°ô¹ócŠ‚Ìü§ð ¥ž-Fëþi\®ÐÐ¥o«_ÛÒ*}ŹS7ÕBCy9ý•W°`º|Ê8k€î‚N1¶tF· aAÔãÕ¥~™J$½ÂtÔÜ®¥§ñ!; Cªfþ”ñ+"*tDAèR ?,Ú-­^Lu ã©œ¯ÝâëR.7Í ù¹s]ÏË…×üþ|•Æ«˜Wb9lx_º€í^,ˆ,¤¢ŒØöïKV$ªêùöO`,2S‚¯Ä害wª:’X Ø¢ÂõŽ£Q•ÈZÎ/½7úókYÑOtJ]O^•½ŠW^OG³¹ÌÌ1Ø ¼744D¼-éu—M ~¬êDM>{9€äJo][ÊNckÔJ²Dqñbú¨Ý¶Ÿ‹ŽS×ͧÞEm=õ«[Ñì>É‘XÌ"U7ýYGëxmÕÐú "8u]Ýò+ìµüuþŽ©ýzBôtû­7|x’;çzœ<ôäãÀ}Ôiz‚Úáz'¾EI—%+ö òÝÚ…kÕB×h)­»â¡—uë÷^?<¹o™=w,³5c;AڟǶ`D…²­Ê>Æ7˜è":3þKP$Æyàò)°1àÓîa\2=òB¢Û™oZ“CVQÏ Ƶ ó©}*Ũà.“5TåX¨º3«L,÷Ú)„ÄåeJP½Ð˜ú õŒXu[ŠCŸQz%#ÜçF2Âó>Çx@CsŽŸ˜„5¡ÆçÇ–Þ•%¤CORìÀ=ÿ®ÁÀÔ16½¯pZS~LÞ¦M÷½9€9‡LPÙ‘eݹ|0!«ÃóÆïjd¨û›šîÄ]“êóž„ ÚØ^Ê:ðz*N²ÅÉ”øX’U*CŠšW•Å†˜x¬È1>lÇâh^m¥W߇J6h%]Ižw }´d‹º7Œt¶”V Ž+åÞ+!syˆëš9º %à~ðÓõôÛ¹Åôº¬)&gg§‘õ¶‰Ëð3I*cJ*ÂRÓÏ D÷–JÒæë>f¸GVôc8Ñ©ÿ…¿1·ïr¶ ”!¡”¬'÷¦M¢gpÌM ½pÆ­ÎÍ.Q}Ì¢§RôòÒÛmbz>;¦™ñ×ÝžK¾ÿ]ÑÞ;ýVZüÌvÊˇûq}¦CDtùØŸÕg³Ä<äÇæüß”ÇQê¸âF5«.¯6Ê•swßñ´”R>,½‹¢ÕüµD¡žÂÏ’¼7³N,É•%iÞïëv‹úÚ20jc8ýÐiœ^g¡ ˆø‹óžÄ,Én#ÄU3ƒ–ˆ•pJ¼òå’ñžêæ“vnÂP[;ʹ»á‘ÍÛŒNàpŠ—•x"°’|SÚׇ*\¦úïz%Œ‘¼ë„‡ÙºWîû¦íÜj/[¼[IðÛöðkjtS¾œÁñ#ž;QW« Û>DF$+QÞ/äaŠ0ÆnšDQdŒ;SÏÄÝÓù”tûÛÄ} Æl·÷B $çÚ ’Cå¦ë¶ ž×•ù$dµ8ÏVª_zñÿ<Áñ;„­çT an‰b)óÄ¢Œ[šÄïÐ(“TþD+×C}sÄ— ¥‘¤é!Òð®V¾S»&6œýºîÏÊ{Ã5Ýk…ïᢀEÜÙ8ŠÊ<ü"¶Ï1†“7 ŸàyKþHŸµ5nÕ"5Kž¨Fͺw!‰™»ôcFçáx‡QšðUQH§P{œ¢ß6©ÐcwôSU°¡ˆÏìøÂ2ÏZ4VØk¼éÝl؆2(—íѯpð>ƒ¢Ä¾å…¹“ÏgÉÇ–Ï„ì6Ûð3Ó¯"å9–¦ŒðæéPe™ŸºñÔ¤½1¡šSb°Üo/X´\Òs¶Š´e“ÖÌá&D*ÌÛcd^¢DõkÂ|Ʋs_Ò3ÞS{ïª;² Ö+àÔ¶·±Z"æöäH!¨A$_•¥7§O9È-©—R«sS1Øs@Ë;ÏÏ| Z¢È–ÒL]ÂÔqë‘ÕÛöx—³kœ3aë÷þåÁæ>ÈCqŒTTœÊm¯>nV6öRˆ&AŸíÙ%ç_7­âaÄt÷€~g´q²þQB‡äYÏ‚SÒS:-Ö¡÷‚KD‰+2’¯îç…ÒÔa~ÃVíõéX#ãû6)³‰(Õ~fíÞCAî_D¼Óù†YØP×qŒºØ ê &C9ÚV—^{ŽCgžW(îã⹺ÓÞÈòô¯X™å¹ýWE€‹IÇ?ZLÛyht2’³Â¦=Y}t¦µ…Èo/0c(fÞH~K|Ë ýÞÈõæ£Tô ¼DAî¢[«õ›ùÀXÎíÞ]ò™~Fçw±_1ã£ï…ëyžì4Â=ت¾²“ZsÝ&ŸÇøŽd}¦QÚدìÍßUïÝ‹ûy¸Ú¨„“*uqŠst±8=Dd'dßWǨR‡8yxŠE˜ªŸa톨ؤ6Sÿ4q†¨’¥ÙZzÌ› ‹“³RªÆœM+®"© ¡[Áþî?_È»èVœå2~èsî(LÝ—nÉ}·E)HÒ>2ù3·ØðXÚXu†¤ÈY€áá)ïE:6~ЛMœóa¤s#Îáê kK×Цñ«Ë¿._£Þ%ë7ð•ÎoZTfBp¢te:º úæÊÈ?bZ‹V×s”Òݯ)n(CÎ1#݈øš‹ø¢ïÜy3o~M†‡}kV^ºŸýΚ,ðëve8¥8@®°±ßŽÈm£ø\/ûyS¶rŠ‚Ô{xñyüyò¯ê,â“ÙÝsk¦gØ Eé:(vy(-£OZ+ÚŠŸþJµéå[D«[ôkKûÖ¹—Ìä»â¢:­%ËÈÅ /Sƒû¤—_õ  ¥7V…½ }ÛKPõæ@*© ÔÄ?ô^šR<5eៅ¦–OÁ±̆(P\õ.Ãzq Š€/l{¥l«òÕ~ëþ, ¶ú~y(vRÎ’=aŒjwÔ{ø¶›cäsBv¦¾Lí]¢ˆo+Ñq|Ðd¶X}Þ÷[Ý¢Ê ­ñB%&oûÔh_£æ°·t6Óa8…Or2¢0ÏÞ½f9ˆ&»UàPÅܲôãzŽjBMK¿U@¶gðʾ*ähØ(©“tUã’cuU*?çäjÒ_hz]$Çò؆°e ³˜Â"úµM%è´ áñ†uøO3Aý WŠo C`»À.ÇamuCw„‰s¥Í/Bp²ª<á°ò•3 纽°Ô¹˜"¾…*žŠžl^pq–ˆ¶‰:fæO¾+rðŽTk 4ðE²LÍe…ry'³ðÕ1߉…Bu¼qÕ×ïkÁ˜Èz€ÜÖìCÑÉêظ72‰Ò'Õ=.¨mà´ŽÐl [:v€‰ò ÒðÑtA×ñrÙIjÅ ’0ç ’ƒ—ì¹Ájá]•Ÿh¼@f üyò÷FpäœU5½“GE+«Iíº´nìn1ü¼{©êy2Î{öê7ÚO«ß §togJ?Y­«!¤Æ?U2-:\&n÷Ía xV"æaX“„À~ÒaÍYÚ/_îàûÆ!º65cÔmUÍ<äýÐRùšá%wgû´|½ú±ùÒ^?î…û栗G˜ØÍéø©“Èð„ô;#|ßòIî.m_á6k¯§ã0;mÒ9Û—UXNÌ'v…xs?(‘m)ôÖŸÄöûVl'äo8øǼ½CÿÞdvt ¹yNcâ¥%u&>>#.—v™5•{õ‚#‡µcð:ä!(°}ÚŽæÔÚñ£2¢RRj—ˆLW|•„ུíìlàà—H°zîq^/Á02wÃwû¯ì‹g!öw@¹IÂÞý ÇUQ³ÐÇ¿¯¬<J¥âÂä䣌I:Íè5öhØýÇÁš?ŽpVtSåûth:i)Fz4)ì6bÁÉ.¹ËÊf—’=o tÝ¢ÁC¹Æ==q•Hè÷¥ÊwdŠ’øº·WG }æ×?G) ¦Õÿˆ¶ßC–ÝùÙ¶ cdõvÅk|Ò'k*6»ùä$ò½Ó×~ÿ]ƒÌ”|_sb5®$O¬oËïé1¶Ÿ"!#6Êô^ô4„RmVãf/W¢ !Ë7Ü2L¡ï¸dROo›HÁ´‚GiÍ©©Yã]'õ\c­e2ìd.ñ䨥ÎÌî…?)QMÀv-(Ø€@k¼#GÏ·#P‘00«©B]æ—BœÏ…Í51鬀|yñ&Ê>Õê¥9çOO>òÏ1Ë¿TÒýŠzYpSÇ+*ó²¶ß¢&ÂäMá© TX«…®¸IúSµû¥ßóD êA™æÅt„ñ¨N@dÂ÷o9¾Âx„>³þ3Ãd˜ˆí£žZLÈèy:÷ÄûÏÊȼçÐG3^q”¿X¥=ƒ–;÷8×l›°ÜVß×j6Öç9qoç‹Ì+˜JÌmfj•Œ¨>\Èm_Yå9 ëê;yeŽfHLáÀH<Ä¢x‹Nk0Jà°øò¦m¡m(¬AÚ88p¹ÝÅûùŠÓ®=—>Òþye¿¬U ó¬`Ž)KNÈW;0×ù,/~ £ñ®`Ë‘Ñ·{åR=Íib™@6tSþÙ#Jˆ´ ƒäjVvWˆs³ÚK(GT1'«U$<ÐĬ ’þ±ÓÂÑ!ù©Û8zc/òÇçeÁ|WÃ3¡¾±Ð»]ã“R¼®Q=äë¾zˆwNQžž{'Y°©åÅç?´ÂR¥¡~YÁ$ÔQ·Ò+N8hlJ³=&#d™ »67w!Œ7ŒåOÅ}à¾Ñ/Ò˜ƒ§3mdé£rä½Ññ§>Ÿ(Ëù¿è%Ѫ0;c/†VdhfŠl,ëVÙ @ç¡ŸšFÊ[:6¼ÎzóÎ>+s»a_šË;~xi”Ù¼¥Vi|õ Ä÷.6^œü~)ˆóê4šâVÿjÏ:Ûœ«U%3žp¸1/9õ=ëä ¸9ƒßÈJÝ4DòÓuÜopVa$MŠ%‡]… T¹M÷€ƒ°Õê¹F¬$ÖÖ<_[2½`E«vº›xÎ×;Ø‘‚ßЊ_Ò/•ÑÛ5ipDï—Âfeæ92µ Ã?3S+|ø—d=à#aoõœŠ¼'Ux"|î5wn=¢l\ó^KOº{oÊð¶%’)d©àrÊ2 Ÿb_©ñÅnCL3þÚ’'X,¼?‚…e/\i< ù-Œ¤Ý=eÆKíB¥?·w”Oç„Uà×<¾÷gª/Rºy·RLËžX!‚FI‰¸±™™µ8\q¶å®ÐÄ&ÈûØ¥œ—j°höÕ»ŸF9UX¼7\%’«©åJßxï糺äZ°&«ëL·>ùî~ûb¥dWBâV®¬$iñIöæÚ“' %b‰™g)›¢KHû_Æ2&íÂ̶@s¿LÌõâ±UÂPЇ;½STñêÅFر®¢»zî´@^;h8á0Ù©©â6´ “Hv›8ŠÉ“)rYÄÄËÈxFO|†½Šõ+cïÒ(<îj×¾Ò3؃J5x£™"Ò;.ìÏœH¾RÒ{ªæÒ‰kkM[ÛQØhاY4Øq]νeü1þK½MZB1òÕHC¬\*€áÂÝöNë*_ím˜Œë„I=Ñaš´õ]êPCYx¿÷…äø£Ëò*NÇ‹æ“‹ë¡#¾lË[ò–yòw¶?׳\Öƒ>²é‡ÜÙZ¯^ç|˜‰­X†xJ¥m;µ”lœÅ?s|'F£R-ŸIû•?Œ@ú(_cV™à­Ã]ØÔÙ“ÖŸiˆç¨ç3d6§ô€×ÌOÞ\bçŸUzÈÃ;¼ýR¥á»fom;“㿪gÄ(n)ûlg÷n÷¸@T»/~ ŒøKÚÄ:7Ý}YZ4úØ•ÅÃÓÚðqö.´…Å;³&ƒ¯¦¦ïÛX ܽ¤ùcÙιa±úØÅ&ö³vÙœ¶ ~‰þ>4¤£ïÂ먆N‰Áç\’_Ù>Q¬ ¦úðíR,¢9/«Ü‰bÉ‚Çñ¼òÌŽ± -V^°f™ô5^:ÿä©^ào½jµ°óùðºõkÑsEå%ÀK bµZÙD#Ƀ«³þÕjÙÔby¶ÏÊN·Ž0_Þì`­O¾ŽÚ[L±»gi©ÈJvk†vƒÁ‡œNÙ“ ™`|Ÿ©¥„ŽMtƒÈg&°}Û×øËZQ©QdËø]:Oo)•»'{æY°lIâ¨À]ÚŒên°Ç§]+œÌçcôÊLZº^m7Ç;:9òËCƒùK& –×Lu_¨÷ÄpÖð:± €‹$·UÞƒR&¢ø=¤œtW•d#' [¾7¾Ë™£ïPcâ|™¾\=ƒ»—Àf¸Øû°·ó Oòvîµç<¯ñc ”wÏb %Q’²e¼ Mà0jÉ¥÷­ß™Óñ· 7?“¸žr˜ù>?Ò#ýÁ¡#g´Û£É¸,÷%¡Gpú»èNÆ”Çù…ËO´Ìa'Z_¤.ÌÁþË·D©¤Ü4ÜöT9Á5³eÍ2fú3ÑÌÈ9FÌî©ûÄ+A÷Ü×·”Wá)áö‡ü\Ažæ”r*A¹Â>Tœ4Ã5‰Újš*ÇFÕÖ»Óg¸ˆoŸOí„ï!fx±Ï¾h+³OC0|MÒŽ±o¿--DhU°jt·eiÏk¿kVú¦UŸŸ?Ïá{ÊRX5‰ÿù-s¯PÞR}¥Þhð˜Š!|±Ý>jP ÀUiåÄ)ì*pÖùcÖMWÞ0áÉñ'~òÌŠÒÈÂ%ÅŠå®ãØÏÍÌ¿Mã7ÄPc©nÃ&SHü½ÍŠE„«š>æ=£ý˜Zo9+XËÁ÷FòkûðÏ}k^{îz]ù‹ÊFçz]+×Ixtø¶IðÞøÝÕéø·JÔk¶Êpu[J¢xÉ—ŽEÍ~[÷ƒýnl»‡¡O[uN? [°èÔˆ§ÊÀÆæ8ãZïg„ðµV³&çØ`’e¯%pNKÞ2Êf˜Rߢ¸íà¨ëyGÊeçÇK0²$Ú¨™ Æ ,¡ž·öþ˜húÒÀÐÓ*þETË$¹%Ë9f"áä K_Èvû†_]KË'ˆËÞZØQPj{FjºÄÙB·û¶™>ÛBb/P‰Ë°Ol’1Åî)±çtªîÍÞõ)6:›õëgó—`¹Ì¿¾ñcN9ØùÙ-ã¡öpÙâõ­ÅpÞÌKüã˜E7‡/±)”v^]v.AoIg*7d°ä}=ž<Ò5òã.,£/iÉ8œE} ù„²Wwº5¢:gÖrl¢‰$G.ç ,S½IòåRÄúQô‹¿pvvÌRèa±ïêM¹jò»µPVÏd4¤C©ÙYþšùé³UqçLȨtd)Ä©Æ$0Çô|újb×Er¡Ú°?4zyŠˆQ¯f–ÌkQ/ßHÑñ»’8)/0%z²ùs§ž‡âÁÁMßÞÔ*ý~2žªN=v%ŠÄ”T‘˜öä=Ž ¬Q‚ëákñù*©_”ß»I/síŠÅ%æG›Ä ÓR+ (´A=d»°D¹Ï'5k%5^E¯žÈÄy ¤Þ,ípË1ÿ¤ß!öŠ-Eî$t~ËêÈSä ò6 ™ 0ÜóVò´—ðnf?°#G¥ oKšŠÊX˜øAÀ…HY]uç1¹«O8äwÆïsŠ¡B ñ•€Ñ´³ò™ït ž^§R¿ÿŽJLhôãĤâ¶(ZÓ‘¥ßâ"¼ßFN5;ªïçDî4ß<4T¹ïÙËŽ\;&¢$¹ˆ7Ôb•þM›£UØ.î‰=#],ò'3 Í,é…uaþÓÕôâÉîµBC ®½ìñèÓWÊFU‚ì›=W-!–†d»eý–ÚbD·~×±óÞs@2ïu¬éFTûxÄ<ÜM´ƒÊ†„"(å0§tÑdníyv¥ëªæLû„‘qrNe= “¬ƒUt{^áíÍ›ŸŽ–ªé¨o™ôÕ½Nä\䔢°y9vXÃéJJèŽ&½(^VlíÆN»qyfÂGo>ëg›ÐÒÌÉãÓJïÔåc f6§ VáßàT¥®á‘ÝmˆÑÙ˜ —…å™÷nïϵ|yo;Ýbit[¢6À¼<|kÛµÿ=jÚn"jL½ö„&ÿN¦»¿Ö ‹§‡ÏGçÞ¯Y!R$%‰Ñ(!{ÔŒ »ä`q @›sÀb^®J/bÇö’²%¯y´¬b™Nšr/£o8-³ô˜wc`ë>˜.k‰6Zƒ‚ôÙ­„+•&¿ú:ÌýØÐe¯sÖßíQ±íüF•.Pwµ¹›‘ÆLIVtéLšÃ|ÿªb±I endstream endobj 2130 0 obj << /Length1 1702 /Length2 8772 /Length3 0 /Length 9851 /Filter /FlateDecode >> stream xÚvuTTo´6- !Ý1”4 Ý ) ÒÂÀ 0Ä 1tHKH‡€RÒÝJˆ€Ò ‚tƒH£wô×÷ûþ¸kÖ:sžÝûÝûy×agÖ5àS€ l ª8’O$TÒ2Ð44A a~HÀÎnC:CÿÑØAÝ=`¸Ô¿l”Ü¡`$J¦ F¢Lµp †§3PP((&%(.…@ É¿ îR@e° Ôâj àP»ÂÕ×fï€DeúëÈiË””çýíTpºÃlÁp éuAe´; ¶0(Ò÷?!8eHW)ooo~°‹?ÂÝ^Ž‹è C:õ¡Pw/(ø«m 6Øúwsüv ¡Ìã•Âé v‡Qg˜-îrò„C î@T~ º&PÇ ÿÃXó^àŸÇäü;ÜŸÞ¿Áà¿Á¶¶W0Ü·ÚÁœ¡@UM~¤’†C~‚=(°æ ¶Aü. TUЂQ=þÙ¡‡­;ÌéÁïsþÕ¥À¯0¨ƒVC”..P8Òð«>e˜;Ôuò¾Ø Žð†ûÿƒí`pˆÝ¯V ž®Fp˜›'T]ùO+”ðÌŠŠ‚@ Ô`¡n@¨­ƒÀ¯$†¾®ÐßJÁ_bTþ®W ªh ÌŠúø{€½ @¤»'4Ðÿߊÿ"€  ³Em ö08àŸè(1ÔîŒÚw˜Ð„ZBA è×ïï7KÔžApgßÌZ@]UßPÁŒçï¦ÿV+*"|€þ|B Ÿ°0PLP(&) üo ¿à¯öKuÁ°?ËûW@u¸(ùG¨ãû«¯?׃óOöpÿ›AZk(óX€DA¶¨‡àÿ™ ¿]þøåÿÀ‚ÿ]“ª§³óo οLþ ° ÌÙ÷OÔb{"Q$ÑB ¨ÿߦÆÐ?¸­…À<]þ·V F‘EnZx>I~±?Ä0U˜¢ CÚ:ü^ž¿ÆÊà ƒCu°_Oð¯¡ü­C1ÐÖ uÉx †ö[Eì¿YUà¶È/& ‰ŠÁîî`_µjB¢¢@Ae!PŸß[à‡#( ªÃ@ Âðk´’â@ð/ÑH(`û7A!ÿÑÿªPò/(€þ ¢þE°A1 €ã¿ *³Ó¿ PÀùˆÚoø?• +ê:ƒ;CíÿHÿ”þ±l‹Qy\Q+†øW¡‚¨ÊÜÿQ•yü ¢<ÿ‚¨Ê<ÿQ‡âýþçäm=ÝQ©‘¿I‚Ë_ø÷……ú@ms3[é0Ǻ°¶Ë:o¾án1£ó[O&@œ3ú?Æ{mj§R:‘¼3‰ÔeÝA¹*sL-Lôˆçt*qâÞN’ŠíüwHÉÂI1Ô¢i÷ùëe´¥¾árŽºic©Ó_î(ù~P;d.?7e—S&5ˆ@P*?Ï+(zÑ»ŒØÖ,ï敳FܨBzLÈqü®ª¡‚Ÿ_c+)v„EMÄÖë}ìHá‡%ãg„n69—Þžh··¶ÑgË7Ý”u‰â¾ŽŠõ`Íp—ÈC·°kwrxFän>< Yü¿º;EvRWß`Š “äå>g¢Óл˦ÏçãRcx°›ðâØí$ïë…Ÿ£¿¦E4‡vòm9­èÓæÞñKèÊÜ :P6=]¯‹iÐÖ»›"ÀLRq² Q¼¿kïm†i5§« ×j‰¯)÷HØ®}'³òc2륪x\ñXMÞHÉW6ònWÖ±Y ‘¤íMAãS–ÇùÅ48wÎYw-îÊSÅB9;\¿tìÏ›ôÒHÞ,]~hXÉÐYx›YìC“™ÍÆW0/Â…;8nz±Çï§ÀM;MŽ‘‡ñäù‡³^¶&Û ßuIÅò3zš¿Û<,‹[V×Í?Zâ,ˆÓµg’Ÿ¿.Y]tÐPÙ¯kx„åH°—®wF’£­Ø,õâÓOÆqÆ›Ÿ~"8×’ZŠùnIådúq–‡ý”Ò®Wx,ÏEDs|$Ì9ªß—9ò«äApòþk€³k}Tµ¯vèç´ ®,?’Ôv/e¢›A[GU&nÁºï ™¸ªWææ¡!Ûn›žs ’²Û¼¾¶RN«ÝUbñº¯Gz–åÛgõú|!ƒc@„°–Ú¼st¯ñŽ>‰ž¯R…Ú†wê‹U^™¡ùŽ-Z1$ÑÐ6Ö.4>8’ñ¥yn¾ú}u¬ÇãuVoK´"µÚ…<˜Ú®VŸjÞ(Ý­LÞÝá?(®üºm[¨ÐÅ&9þR?Ñ+&qQ¥½ ‚Û§Kò´òeAs‘ä˜ Ã@xä6ìhNÓ(÷ý £7€ÒUA(†Öž©¡Ê±e,ò\ô­®-õ)ÚKÆ´ÙšÑÛ;üŸ«ƒü¥Ð*çL‹£(ûÂGŸ{èÄC?†²foBª®´•¢9'] '¿¡¡½Ïªù‡Ž»¤ÅŒw™+Âj¯†…üu-o¥(¬Šä·ã¼_|Wë`ÚQÑÊm9Éeo8œþ è¼ÉŸÞ5ÍÓ,Ò5|ìÛm»×¡wÓ2€FúϬ6ƒWÆào3N:Q-£;Í„šýÚT–*é)»…CY*W­†ºƒ…f¸3Çö­6'‰ó•Ú\Ù)jl GÁ7ÔØ8ê½|®õ¿ Vó’cKX7¯ÃïׂžDM_yÂU˜#Žë|Ç]—}ºÔEYø^ÛXÎûvÉîpºZ.ØÄÏ>Äœ'ñÓÒ±±p!!²Žª}º )B‚•Z ¦xÙ&ª_8DfUñ¥b~u{oh*7ú¤÷bi!nŽæŸECvx³BûêÇ äŒó ¦'÷š£*Œúa”Qôq—]QˆœánKøInz)ÁÌ£"6ÄB´Îñk¡F\{tæïM$÷xžù|Q9íªŒ¼£¨×(Î?“ ©ˆNÏ7ö¸éYnš4úûeç¡ÏêŒQg0õrçwdçğȱ¤,©)ËV unÛHñΪ„0¯¼Ó)ÇÑ?zL/¹°DÖÜ6ÍÕ÷úÉ1³Ç›¯ÌAKW.‰Ë;Ðêò>-Å‚„ÇQñZR°Ðl'•wâŒÊñ 7p`œÐ}ÁY¹{x? õÄzi@ºÉš¶YQn2r,`~:zhù?ðÚ"aÿºõM£úfåÊßZ“Eï¶BÃË8½9ÚwöÃrÔ¸|}fŪôô‚²µ>ˆ°fqÁ÷q›æ󺥒Ò1û|ÁkMéù –>/ {ûÓ%‡åÀ\ã5܆BiƒQáïÝMœwšƒ!I‘“ŠëaS߶ü‰­áë£ìŸã¡^LÆû]\«o(Õæå|J·Œð߉ð7˜‰²™[Ã$¯™äù)ð—°'ü©ÈrÔ¦;Sõ‰¾ãrÅíñåœ\ž¸<}³¾KÄ“¤f%2\ÞŽù±Ž×a–šnÐGLyûÌÜ᱕…‡MyagZ«jÊé"ÝóõþQ‘ç$w0λ—®—ëfÚ”ê …Jïå”wµäðà‰0åÎåÅL/š“FP;^”ëÂïò¬þ˜˜Õ3±ÕB⧿”O´§ 'bd€ì7¶Š÷ /£!“²*V噊"Ñ)“#b!f½@4w:S´Ô9ž­±ÒW½ð•›hŒ/=l>ñ‹~yÀúÖtrGŒËˆo ób¬C ­Ù¬“D›»@g_ÃÙìLûÝ Y”å=þÛþ›ób$Ck >AOÝSpøÃ/<Ï¥Ê(Uw÷è“Ð×%CîëÆÊX<³ÉâÂk\z1ž’ú¨‹Ö*×Ýñ¡pªñ{èÙ­§ºNËH,K‹ äK½37®­õDå@üåF-þŽ,óBufz~à”ÚGå·åjŒ‘É{´ÏÈS×xäK‘–üH¾Æy¨y±ài¸æqºšz‹>µóáÞ„´‰- ×óà ñŒo_¿'^WŽ BgõÙ"/X à µ)Áí[Tɼpƒ /ãhz“åÇ“óS:—]p¼Ø§ÞÏ¡µÖ}qçÐùzC©ÇÔf¸ÑÍþ¹³í¿<ÙùæŠ.ãÎ=iž ¯:„p€k6¯Èš¼aňúÁ~‘½[ºþØYLÿE#†³*÷‡ïF}y©A>×ÉÃ$‡rÉæDŸh}b§¸êg†d¿c÷¥ õŽoÔ}FÚšI—E£ùÆo¢`€@ê™øye•ÙåÓ®¾Ÿ"‘Üü:8߀êä©w&–x›çzí®ü@xÈ«{êk ëêK¯ü÷Ù'y>Û~ìä`J$ݨª"E㢯û=0bðÙyØŠ°á2ANÕ’žKÁUQW¾ZòÛ°Ðo„»Ö $Tê«;J3kÒáO/¥ö- …³7];“5.MLš±sýü‡ ë?÷ Š+ °^ Ä>²¿=’Ã{SMh%kö¡ ß“ÿL¸¯(Ò8×hzßìì'Çp5W;ˆ›$Ðü5}Ë‘ë—±žK*SèlsòkhŽ¼fÂÖ•*6ÇZt9Ó–R’jJ‡Ñ#ž;²è? oh±åcžˆÕÏÏ&«~êÙZ 1èg›LeaßdJÓ=J© Ë¾6;F:{߯À™ÒÑj7˜Ÿ~T*Ô÷¨}9ð«ŽïlœÔ÷ ÛÒåbyµÛòH>²),º¨xt៤RŽkéàk,SÇă‚B ¢5ã«Gž!ª³v3UiÑÄá'µ+¼ç]öBÂJ!”zßµQgPÌ2=ú¬6]~Ñ´¸ŒÚ£ƒá½u?"^-cúµ }#IXJÝÓÁXžD¿ì`^4L×-$¶Øò%eDsæÃÔö{,ÿi9Ÿ…dåhrµØ;c™K€ã™8ÿ”ïŠÆ\ßê¦IœÂg&÷¤nËüMâÉÏÜU(ŽàôÌB?µúéI¾¤%íx~ã=«à‚¶o)ÐWlû7Q r¶„ÌD^2†7¯¨¦œoF”J·•Â߸;€¬^•ƒ@¸ú»G=E²³œ Âã‹õúÆÖpã šyÏë»J1žgh,{kLósÎå2•;‚OKÒÕK©¦Ð¢ø~Ïå ýø´–ûè­zûGñO.'mŸ1âf<TœIK˜! ÷ü >Ñ¡m¯½ò·à69’t;‰Ôåé¥ðÞeuSøÄ`‘=ë;LáR}Ë=Ãb™·×õH@›úFŽNìÇýP¹^‚M—Fœ­Æý¶¾>ÏÎmZg‹”B并¯˜¹GGÒþòÜ€¿…Û·©‘$w%>³Z¯\‰m¢£d:­s ¯‡q—.ò' Ξ™eKI匵qînœZzh};ÀuÙ\)¬UäÄ'¼Hn>õxmÓ@R¦VE°5Öë¢ÕñÇîÖÛjnjúéOO3ø”gÇ—&u Éf±bÍŽ`EB$[êdº­ï#ÍVr cš;Mcwhê6oÄÓ¾ç£õ½hn?`RŒÉƪ§"Õ]ïBJ‡uéºzˆæŽ13t#mÙŸ}ëmÔ/® ÉŸÛßØÔX¤ÕûÎ{U-®pÝ1Œ\¢€Íï±úó^F`u2KÎÍâbãTäìÈyû+Q4ÏÑfa—l[¸ C<±A¾~1“ûp™ªé$#õ9-a‘âáeLûÙþ’•¢_P~²â^÷ˆÅ)]&•à’ÉA(@ÝUN°Ç÷ÛÚ·)mùë>Á«,#n…n 6míâàöSü€kÚ-ý×Oó 8à²úi—3q­6 `µ¨ ¬¤òÐ%ëAVt‡jpØ'â'UØŒ—Äžt­q¦ÜNÒ²ÃWmáE ¾i5î2³†×[bÍÙ£ÏFð%øKVE®ƒn$2¦'ºzÞó|ÑhnDD.„KÝZæ8ÍñA´)_òµø;(¨ËVæKk©}c7—ä“s>Ò`ðšTk|YS$¯°$@ƒ¶‰¥ûÑxÚt' LzÝTC$!ä‰ÉsÔ\$™“h‘4gªÑ–íÌú"yk?爅 Ja1Ì|yòÒãõð½[8•ðÕ*oUk‡hÓvkžÖÄ,sKÙ^òÒ.-¼f|â›ÊÆÌ%:ÅNBûê=¨rÂQKZò—$ÜÜj5pû±¶édØ 4Ì…ècB8O5<â ËDyr[K·ræ9*]yK÷S„åað’d9WEVÅûùm¿(%`Íšš r,î/?[ZMó 9m`†7f!…%ÿÀ”Á§’Ìã¨È¬~N+Z6tÿÇ‘BènƒäîŠâ/¯ô+o)¨8ÌÑྞsËI î.SƘ¯>pÅåòó¡q?‰š…½·xyfßÆè„zˆjíÈcîÑ$~|‘9ÀLߣ/›Ñwœ S•˜½§”Ôju5;]~_z%rÓ¢Ó0Ò#5+oº±,˜ÄÝ=Á%Ö&×çÈøq „ÚÁj\ÈT:hxW$oäÛ¦o’fpÿ ‚5^Æ‚økb–ÀÁ‘åÊIµÇNJh }Bïú×;þÏͳð— ô㧗çùÈð×2_AÚ¢N‘Dá_ ÷)sš/–µíÇ»˜æûkð¾ÝYÞ‘¯®EgÒ{0Ü>d¬«> ²*ðö둹¹Èž%¥2Áð|nPˆ_ùšØ ¨ÀµI^løa”A:ó‘7~y|5³°I·E’‰@HGWtêæ|e¤¤¿„ê«À½ ‹âì‚¡¢û£½>„¬jZ» 墾.ËŸêmËæÐ{ÓC^ªâÛê$£Ž¾Ó´o ò‘຾·Ï**FI¡d”oe±*g`ñ¦XVàÅl·®V¯=´NbtÇd¸£jŸ~ù \OKÑqm·!ïuêåwUiMËO*u-™~Ï :º°;ᵓ8æx!`Mzòjz­”À7s•µ*ÙÅ„ØÓL¨ÌÞ ‹¦ ŽOψ<íð­1wUDnë´Ä{l/DµœÖ:¶f\£”ÐÃÛÛÏfn†©§õ+ÜnÍmA¥p=ëÙ>3‹ißavÜ•UÎXÙ 1ö½:Éö··EAðëç$BXÀ]¹ú·¡”{âQ•…ÏÜ eŸŽ„uíaæŽ;8=˜XŽQ¦õ‹êPï`zE ªPY‰“?èyŠqQļ1/ ÷^6]/z§ªjbpòAôÎ…xs¦Öò! )j'åûXeÎ4»†I§OårŒéšÒIüìüF¯¯’Wo~ðQÚ±Æ{ðÁ¾ËÇ>Ñ}bùÜ‘ߢ€Ð K>vë(2K•[v£íã$ç¾×ˇš’IÕ¸{ÐÏšjôÄÏk²×ükÜÐ,Û„†ZóÄó¯H —3_îÛá’‹Šâ>*øõ½Ã»º+7½Hxb°HŠÆ8ä®×K¦±·©õÇH(7˜9âUx|ûñàg©ÒEÅüñ&¦ö÷ÏÊ›R}`£ª ôÚóUW®».bU)/†~Z~îk©n‹}ñ\dÏ7e^ÊeLš->Ƚmý)A!b¬nd¯CRºDÿ>¬í€‘g­Ø¾ýû°1cu½5Ø­uÙ=?ÏÀÊ·7OÕ{­|¦Úð~Î ©XÄs€™0¢qÇöIͼPµÆ}ù« úâ~§ì\%ß‹/:÷­W­8\Xø<­^,½½³õ›ŠƱô/¹ëÞÀà’É´K9ÆÆî)áñï Ž'DéP6šÙ~“ž…—¸p9~Ëùe)-?VŽž$¾{œpëfÛÈ诳ßôEŸpHC™! kîåõ®JrjU-:e'Õ”ÿ: £Ü Ú5ê@ŽÍÊåµÆl¨³iÐM6ɛݖ "¢7‘©²±•ûFÆãöËåG­Á_­mCæ?ȋɈ}Œ§>-æD6YR»ÐÎmžH´ÅÐù!ú°S9MQo';.ƒvÌ›ê/÷x—Ï…í +{ÿÉÌF‚r,^Rªxgi3LsyçÒéSÿÔj5'æÝ™GŽÃÊU¾Tb@SOrÃS ^Ýiµ<Ýeðþ÷³ùÑrWükn—ÆãuÌÃ=~éO3<—ëVdl ²È8¨åâ~"uò›Qü;tÆWk\§xá)nËÒ¼à‡ŠJ’‘okèOŸ«=YàiR=€ †¸¡&jʸ¬ä‹&dK¼e‡;=æª/†Î9ãÊå+›è=–B®Ã•¢óÄÛš½ŠpUÇ£ÍÉTS#Ó{Œ86[°¨©®õœ2ƒ¨#w&ë&I•xì«+á$ÜßX”â `¸ûMuE>:xQD0¸=6 $¶P—"«^;LjѧVÅ$4k"Èã( ‡ñåôWT¨C岧ǹ<‡t÷p$Êa’D¤ÒgtoÛ⫦>m¨…yñ (îí¼ûj#Õð ž­Ê#ýüY›uãçš´ú%ë*PZt×"vVJÛA2î§J@©Sß;ìÁ nR— wýVg-ψϹX,ÖõzG¨U@¿Z?…“6.Æ`>Ó(íAM³°‘sžF­º§´×R»Èå~9Z‹yë«)ùYo'É4>܆|u<ÉÕðþè1ª }ÀTõ©&b'è¨[˜êù8Ò÷¦*·pÆó„c¸Œ€¥ÇHmoÂo¯)Q.Ž“æ0€³Ëÿ±»…×YE'ª‹qßæMœ®f£02•íp­¢ˆóCïÚâüD×db «º×#/÷ê·¤¨²wªb|ï†RL7±>`ÐÉÕCÜ3Ò•|¼d7>1æ{®E‡û°¤4o[Û†¼ècØ#›´O¾_ï% .­[ Á¦XÞö¾}:¿8e;a®¹Šq©yàôžÉËÑÆ„€)¢ “Á¶^K“if¡M¼«ðù±H̺ƒö$›°<àÏ<}®c‹ˆ´¸öÃÏð(k/%]šOw¿Ðc ƒŒ?ÍXd½(ý3§g¼B¾½XÆë?÷ñƒsôþêboŽ›Pë[ﻎÐÈX ôOµØîófÎ]Mè v5$×ä›…©­N‰xŒÒâzÎÎ;‚ [­#ŽHâ5¶4´¢Ñ4cÇêÂwÆ|ئò¸rS?ÅÆôo‹+Œ'«³FÔaF&oC¿m$¤°2tŒ¨ï†õím&¿´Œ}WÆŠÖœà$’¸UÜÔ|{Ç/ÆvöT÷e¥Ñ1} ¯tõãÚäÀ/€©1Ž ÷¤Ð7¶n¸^¸`ÚÄ›íßS¾©)Îõé?ŠðÇ®‡ {I#6ßü¬ñž†*èõ^C&–ïµÇ”x\|ÇБtújÓ =H¾­7;ÈêD¿;è±iÊ_¸I²ŒY¶Œl ŸTæáí±‘±d(^M¼#ö·.,á…[ÃæØLCŠÆKacÂ[ M38lݾŒÃT†‰¯ï±]ˆ½[ef6·¯÷Ì´Âò¶+ëŽË*‡kÞú?Sís˜™ø!1¤4«‹—”[?X/²1 ›L?jëÏ<ö¼ÇÒxq¼“$!ãè_¦e}|bpͽ¥nÊE'–Gk3*ÚÓOsA¥ü\4ËÑÁ5ÙÎ陘=¿Ïn„q%‡/ÏÙó$»‡DÆ/dŠÄ?óÅu~ejù×mœÏ}HJõ®%9ëdÞØ´C²Ë>øàùµÿÌûRg¥YlüƺE§b‡;ýÞoú^Mð=zÆ©ÛÉý¼›ÆQÈ6àE< ‰6ôŠÐÑzG…°}“RV©¶0ØH3Ȥ”B_¥ViçÙÛ¨G"hˤ-»LÒ) åxÍDfd¿¶¿ÉE~ШÀý,oþè}K‡¹0UˆU†Uo›Ú-WZôzÁV]Êþ­zc#,Œnn=Tõ ®ûzIpVêÚ¤Ëv}í»ÆpʶCs±£âÚËð†1EÀ”n<Ã:O$‘/¦„E ;uú›ÆÞÂÜ—?øHdH}礿UÞ«ãîZ¬«vR: L¾Xs•Ø‹sÆÈþôÅų­÷AÉSƒfÈv¹[“â+]Mب [SBXFh&sj®w­ÕÆ:¶K2èÄܱŸN¸w{Ã!¦«X°M%¾'abÊú¨ï‚5ü“Õ±ž: oŠ®·Ê¹qø@øcùåS:%çŇ»ÄßM…|·{`sÛ*†°½Kˆ…{lÁ3–y™—?ñSUÝÝJCcÝ<5_—”lØ‘xÍ«p“¼`p‡x[zO¸ª‰ßÝü4>[‰+%±Ë´l',5)Ðÿ–_‘F0H:^ÍB’wØä˜;8Jò©Ì}>fû‹ÑxKo¹Z¬n«nu†Ú”ûÀ<§iÒsŒDw{rÍ.|T—.cqCŸÃÁÅj_šôpÆ´r,l’Iøãî¨ùî}$w©®ÛÃ}r4KÎõ¥ºeó÷"°ÀzªRvK|Š‰$§c£ÒÈË—¹òìçòð™±Àì‡wBÂÔôç틦õ•Ž§Ý·¹Š]èægI}†ÃÅ©vj—8{ÎŒ¢§Ÿh=ç w UÿtîÔ-¯Éé¨E &ùˆfx,ryª‰´œ‘)ïyʵ”.“¬}Q ø±y²Q¢g”‘zJÐþ?cDÑ endstream endobj 2132 0 obj << /Length1 1682 /Length2 7792 /Length3 0 /Length 8899 /Filter /FlateDecode >> stream xÚ¸TÔ[6Ž€t§ô€t]ÒÝ]"9 0Ä 2Cƒ¤”4Ò‚tH7‚€tIJH—4"à7z½÷}ïûÿ¯õ}kÖš™óìgïsöÞÏ>¿YèkÀ#g³+à~^ @AËÀŒy@CÂü7ŽÃb v‡C`P‰ÿb(¸ƒmHLÑ$jÁ u¿ €_D‚_TâaîEOˆ@‹ ƒ‚á8, 07wˆƒ#¹Ïß_ì ¿¸¸(÷ow€œ+ز´lŽ`WäŽ € #|þ‚ý‘#á&ÁÇçååÅkã ç…¹;Hsp¼ G€>v÷Û~¥ жqÿI‡`èÿe0€Ù#¼lÜÁ$à¡p¤‹Ôì@î0PÓ踡‘5ÿ"pþÀÏËÿO¸?Þ¿A ¿m@ ˜«› ÔuØC\ÀeM^„7‚`µûE´qÃþ6ž6[$á÷ÑmÊrzd†òƒƒÜ!n8/âò+G¾_aeV‚Ú)À\]ÁPç×ù!î`²î>|šë …yAýþ^ÙC vö¿Ò°ópã3‚Bžz€ÕÿpÎ00 怟ÀÞ G¾_ú¸ÃÈüÜ`n{dàˆ=ùã·ñîà¿ÿ6ü{…ÃÏ°ƒ€[°ŠóŸèHlÿ×Ùwˆ7Àˆ”?øëõÏ7 ¤Âì`PŸÿз˜ÏXEÎÌÔëOÊÿååaÞ?q¸ÀÏÏ/ü;Î?ø;ûߨ® äÏéþ+¢Ôÿ+ dõþNÄó2ØÿŒ àß;hÃzØÿ#ÿ'@a ùÆÿÿ<¿]þÿ´ÿ+ÊÿUþÿ{"e—ßvö¿ÿ»+ÄÅç©gr6´`È þ/Õü×@kí ®ÿkUCØ gDê€Ô9¿/Pè/W†xƒít!ã_Zú»È=\ P°. ùuï ½€Àÿ±!G䌼[àÈ–ý6‘“õï}•  ˜Ý¯ظ»Ûøà €\ üø‘³jöþ-q/†@º9ìaî8¿Ë<'Ÿ ÔÁÖÝä FÞ‘öˆ_„?6áÿ¶ýÕп‚HY,ÜÙ@ÇÿrðýòÿO,Á?ð¿Ãð#·°Eöü_’ ‚¸ƒ\À _2úY>dÔ_WÂ?˜sAüÔ)¡¿Q$è zÀÿCC.ˆ²lÿìàsCžf‡¼Ù‰ Õô…Éw·±û-Ë_à¿JòpwGzüž d?þ^ÿ¾Á`o0g~’ szÖvU-Gãų9‚±²Ú•`Ö)Œ`}í稉™©2ùTÞÚ®òÁPšî\iøÄŒ? çáÕ¨÷“÷ÙÁË©ª¹¥Þuùcž¼ïgïÀôv0j‡1Ñ´¨uâ£D=:Bksts¶d“™›ç¢Ý‚Ûo· n;àÛÅ÷®€¢ "ÆTD¬\NÇ>ê 3FæUÄ Ù¢8wótC5£±[Ô¸Kù¶ªï>ëæqƒKK³ûzÓOó5zªD*ÃI©È|Õs†’ú©È³Õá(¬ò ¥) LT…®á¬ËhÖ¢Ë?¦òûZû=ê´‚quž 7xÔ„g¤rº—Oi­6² ¿óá?hJŠrä®Nc–g©²C¹L?ªtÉíA§ `Ôv|þ©5+.¹4ÓD]hˆÁ|*h¹5þàÊ:63‘_À"EÊE¬ÿËy]z·!{“…›G)RjR]54mF§U¡‡Ü¼ßÞôBÊM’¡ç;MÛ»#Žȧtä“Á ¦í™ûöÕ„/>s*‰ j4faVñ9ñSú¹9躭*’eÅçÆ™â ó ‰g 5·QÒµ±Í$žËM:–Ò½]¨~-Ò$•"|öyüøž¢Þ×x¢šu\½îï÷-lÝù°,5r;;Zk³©#„ û 7;òf“ÁÇ¢3˹òwFV¤%ÏžäÚÇþˆû¶}^”k*WùØ%êeçdàºÕ›aºöÇ}q´é]ã_{ÚÙR¶½Uë¯/ÐImU$ú™Â ]Êâ…[–úÈ£ŒÌûl4ÛΛ°Ü»‡íοãñ›_J m»YîÔk¹ý({*¯ !d›{..¸^©Ä*i¨ˆÿMŒßf¼>D`Ê‹âh߀³á"£±@‘¿oÎü¢)š¸öZÅíø¢‡v‰´¥†–O½ßt‰V¢,öÃSWpñ Ÿ©ŽÈG9XgÍá³ým “§4Š!ZÍ‘fB+ÅgáMÄùfñÓÃù¦ºJÇš·ëV͸•Í¥0Õ?~JïI*²=¶Î6*ÿVþÕŵ‰6~¦ÒËÓ¢ƒŒæø)Ôó¬/×Ó]”R< Ô$FÕwäK¯œc{RT9ÀႯpe(*üÜÍŽFëi? å²{­µÀPpé"C â\Žd:¬´ ԯز(s ß ú)Ó(Ó˜ŒcTÃy7õîÇöƒoúÄwVˆ)ÉŒÎá“ô~Yö‡eÎábÒ)\£×Ù~þëß¾éØn*—¸ ¥tœ‰y£WÐ`L"’TMäJº+×$kŒ—ÇïðZÉBaµiŸ=[r¾&Ét×¾d|™]#5´(ÍÖ·^\Þñ„©ü{CJ&å„þÁ}O}…”¹Bôl=EG*‚)Y+îœ~Š¡lÂvgŠX[{—m9µ+]KÔ²Ú ¯ùLÒ";µU_gvÐ¥`¨÷úè௃Fºjµö~£¥#pêçx‡#fUmy|_=ßÇ™ç5ÿ˜£oÒdÑÉ“ÝÏP O&y´¬Ë§þfès gâ4ws“£Q±–Ìyí¾Ìj²çL˜¨'Á,¥Ôg]~xÀº a²lv’°å±`J”áf­˜}K8€ÚG¯("´Y‡#¬¿(:µ5ÉBð;L¯à:Iß÷ éÙ™ý®_ĸ%±¢ãâù£|¿ŒŒ5ù”¡½™€ë®¾ù¸€=›Â™ÏòZµUÿÛ©·² z±ä(µèS¶Pª­{‘7BÎ:R”wå²3F€¤’Eî%]˜Ò‹eÊ-gÔOœ¢XOÑÌ1¾^^7ŽÎíL=µ†XÎ27²ÆlžZ‹¹¤Äo4›YÃÂÞó%¬­Yrd2n~²Á?C]­8œ¯]¥n7`Ê“ŒßI¹ñ°D1ˆÂ‘>Í‘—"þœï ýœa¸ŽU™_3ÖÆô|â²”Ô‚c„'0À/r¢o@6w*•ëm&„àÞm)ÈfÔ :ðy½nêË´_¾viÊ‘# ü{Ùîq¥yvU»€"dc³Âxac´?´Ó¦4Óâ|GfšõàÔ‘ç’œîH$’øS'Ò³ÛªCw¸V<Á`‰¦{À¨çv‰ý>æ²Ù]ä²í½daíãÔξ Ï9zqg<â8ŠzÕ$7IjÅÙu@²Õö¸Ä±~˜EÅR–­¯Š‚ Kˆ?x©Ð‹q§&uD$v¨P_w‡y«±Yâgû'Žö8¥‡s;…vl=›4<Í xPøhz,'qó41s‹”x®CÏ«ßPÈ×$ë:Ÿ ÕÁÈæQ+¢³4U ?µh,žŸ5äp¤Ã|šÃ5ûdæ“Ò;ÒZ˜›u‡ÔJk8æÖšG¶HbŒqKÃvž=V3~QŠ0(( ´“Yª1%G'e5 }þáÒYÛ¤Þ¬ë@yîÑŽ|@mê®síçÌýö…é1K9" ªÜÍíW×°V”˜QÂ]×PØÔǯ¥BTz›-ä•ay>¯Hý“ã<HŽ¸ž%äÒx©TŠûôn-I«o–¯ò‚Žb1÷íFœ>¯«P·ë©}¸Í¤"€Wôÿ37Ç¢ž?OŽ2¡+æ%—ÿv5ûF³ñôÉJ´JI…‹hˆ†= ™ï²{®ÕÔ&?¶2¡¡pS-@{烞”šð#*_¾ÅÚ†yßIÍžÇaþÍQŠ÷èû„x¿¯•Ü¢©\7Éõ¯FT`EÖ~\¥šk’ÛxÙü¾áÒQ3ªz›gŽbì Õʧ{®ÅY÷ ./²/ê@ñr^ˆx@¥ÊÏ!ŠUaŠï9 aÒÍÝו÷Ë@:IH¢|϶Åúzý; ¥í~:ùfÆ÷Ô»öo¥­R=-vP¤J èKY¦ò–n•áiTcü¾M©‰Þ.‹oØÔõwÍs¦ÔùÝ‚5à‡Ž7¼èS£zÍ“Ð…*>Tú’äæV¥ÌY^ò­f9ÝóÈ]gÊn* Å- žD³嬢ùj*ñ­c[º¿-S‚sdY“Îu$–¼G<‰Õ&ÉI’‹Î¼Í1¾ —äÐS­Á¹lªÑúÙ¸ÍwþñÙ°Ý6ôðb ·vÚ°í%qOFÂgÍÑõý£Ó]KEHÉ­yѱüÑ«Å&nƲe£ˆkÿ,”ԇ߅Ãöí'Œ¢÷žê\ÞLúKýŒÉDþžéÖñ×íÁØgŠŸÓÇll&sâÐq9¸ÉqWs¿—òw*&VÍú6—t”ZUŸ6GKȻ˧\BC3Ð<ÌOEèg¯¶D£ò2Ñzñ©©: tƒ^ƒ7f\ÝNÕ#`u²Õ5k©Á?­Ž÷ä&‹×ôÈw;åéÏŸ1‹+å·P$8Iá¼°”ï¼òóº_eE(,xY¨|è0ÕY+ôJ>cg#ë«\I9ÉjûÓwcÇC„ +]ËBõ˜Ñô ã‡üouí|®8ãž1èÍ9_o»OûÚ½ Ø9SN'Þäs•­íoós¯°¤½—·¨ï¡]Wø!ÒEàQ“³œšE¥Öa`ƒAþÖO³Ôyh=uH,FFü4k §e]"Û H…píÁ kã€ÿ£•ÑÞ‚´Ëú êƒj’Ÿ{óe\/A•A òžù°¹~…?Ê0Ú wc´ê@-Ök[ø¶ÁV¤¥Rw´z?]šµ·d¤ËÉ¢Žz&h*­žç+3¯x$¥}õü^»-Ÿ.Öï)¥­»4Ó¼ç΋ŽçÁ+ØZôÊ" è,Ju½?^Ñk­ùÊ¢ ¯&|ßRù`aÏ ¯àyªÕ b^frãÕmGjDö…¯3³³3³…Ð~9ú1Ⱦ)ﺊ¨UÝ$Uötèì‡à«‹47½ª×Éz[Q趫%ä|ng÷êÏ^Ó&?íPžC·º›–#Nê„î“ê}Âü™ïªçÇê+ùhº3#ØõtÝ": +|Å»ùÜ~[¾œ«bÚýÅ‘ÒÊ‘9bbÁAGƒ6Vï ó5ó;)2ê/!³Ù2{v²‚ " Á2êT‚½iÚiN­UÚêŒÝÂI»7«é÷ÜÝÚ©ªùÉœns•‰©‰:#AÏátˆ”¶Ë5…'©-nè•wÙV ™Äù5ÓìÍ×Y…õ¸ZZ-8õµaøOûiÍdºË,ÊR­!Ýý}òóy*sºŽµÏ´ëÌ›ÐÕ­ö"ƒ°Ž )|—'í±·±M¥¾Œ¹ ¾g-„|*+ñϨ³µÝ®.1[ècc†`.>) "^Óp¶Ï·9ƒy¨‹ ¾ $Ú«;õ\Ýë<Ôølâ—­5µ,(3[ªujñ]„J’ŒØ“!á9Qú}Ó5Éûaò IØQŦ¤ñ¼º«j’2¹Œ†‘^ÜF8ÅW´ #‹·¡IR¦Ñw¸·IÂu]ÑAîAZi©ªpÎõ`[´Žg–øíˆ^9÷š<¶†«Þ¢”›³{OÎ/çrh5ƒQ÷Cót“YEBÔ _I ¡…³‰XzÉF¦ÝZÉëEƒ”ÔnÏö¡ã›o Ïje:ÕC[ù’”™DŒx«ÒÛ¨ç\‹zSa@*¦Á©ò ©ýàøîòA¨7LÝbu¶ŒqØëFDg‰7àENb(P˜ø8ßxm¨´–«‚¸°±¢²ë†N4ºÖ’ùv¦´Ïg{sðC2¦anŒ‚ðŠ»ÃšÆ"áaæøˆ´VÿêøúÎh¬g)K*esb9i{øyUZž›+ìN”ˆâÐõþÉX>ôä ËbÔ¢WuØäÁs¹DêHWÞ˜7q•ÇØúæ”ãé0íGN·°îØB¶?´ñW’ÎtI0?¢ø 4Ã.1Ž¯‘¹x¾Nú8Cˆ^#Ò!½´w§ »t›T6&Û°â%`šÍ悉¥OÃßZO#àKþÇõSÅðìä(‚€3Ö Ûbâ¼·3Å­Û¶Ôžò ®@¥N.¼á¢HVñsÓf]éq¨Ã9¶šò¶„L7âå\†‘Î(¨ÀI@ÁÕ¼ðK¹Ž¨t°³Éµsa~i;êi[è»Cº"ŠlBòîGFÅïfµ?‹iÂìœ>Ãg ‰"¤ñ2ÒPŸ¤N ØLSìÈŹ”DkŽµ70°2ìbÄüz-Nc\›_ñEPáöå´ÎÆ‘ÚQ{R½0C Ç0K½¸cð‚¹‘žïf0f(}Ð}%ª,{0öÁI—P|í´’)ë½C²|ô«˜ÑÞyÝøsIyK½xÒdÁïœò¸‚µ¡ÐV4æƵ>‚½Ÿ¹V:à;ëO8 éÌ]+Ö!<âäz| 7µjÁA92sïûö&«ëïÑ}uÜ@³ˆ.²=îľÝmÝ+·Úx4q5’ÚÇ…NôAÛ]Vm>–]I¥žwWª#[À¢•¨²¼‹(ªŸóÄ Ÿ~BÝÌg»>Ñsªe¬L>þ9Ê嬶)`".mŸPÚ ¿úÐp.ˆˆh(ÿÙ’çT^ÔãÏT\·u~˜0ƒ“Û?`N>¥혋gFB^Ë1u&_¹K~Ȉ©šô…»Â{Ôÿ^÷>(TÍy¤¬A¼¾1wäÀ±·ù±Èøƒç(®ÜËŽ<&Iâ ÐFÕ°5V_J•l,zio¾ªµq)“ˆb¾Ò^ÑÝóê±RF3j, F1YnN|wxz'¬±Wêc—tc’º´Íòf9«À¦†öƒ×£±u¾Ÿì[“ b»%¥û5­”¨‰P*¤jÜ* ñyU¿r%Ã,_4øÄüVïý„‚æyþYb…wºö—Ž"ºçÊMËY÷YñŸ1…m(%K„û7HZ£öâyþ`¹bX¨24?Í`X…€Q£™TÒŒ¨bWÊ1Â|¢žÞÓvz‰}¬ÈÞÚ1%—½±åÚ~¯n1çRåý…MÄët]ÄôQ§¢½&Õûû,vY/ û"㌡nó½Àgû1üÊßwé- å=+°û×t²<'«hîyâáø^o6-_ç—Èl^dÝ3ÏòõKšfFóW|vžN³Ó—ûjÍÞû\Ö/A꾉tçqä†ôubÕ¶ìVïÏf£Õk¬ß)ô¨JöƒüRùÈÛŠ–º0²zÇ·ß—-øt”‚Åc—‡þ¹Ýccï J÷êSzà%›Çá ?9£ôgÁÀºXÊž¶7–Óo_>à˜ÎàñkÔ‰o®Ö™¹ Ò/é®#ôþ¸³Å©×‹uî%ï0ž¸ÖvkpÛš“Å4[æ’e{|á+Ù\“®æ3ýDiöþû¡½(7÷g-¼FY Ùn0k×kÁ”t—žÁ(„ÁåÖºàŽ>7ýþ«åù®2¦õ8éJà$á+®óÕðâǧlÔXd5ÍÅßÒ¿lÛFê½£>?®6p!Éuê½$PQûbzR¡-­=cá%Õð}¢M–¶}³~I¡´‡É½6˜$–Ò¢„‘rñ«ƒú¿HLRÍ L4IðGÎ~Lf"ìÔ3iyªsûm"QˆSÇ÷G /¨éÌ ô²ÜŒõfeqnnŒ4þëžÔ¾(ÉK³uW÷¸å„¡ƒj+ÊÑ)|Ù27ÃÑ«ø6pÞ:m¾;…éVTåâ­êsa w˜ÎV+Í€Šõ{vÖn0pÓ½ãK¹z1½jPâýàsTíÊ'2–Bl‡Ù;¶¥­c ]l“Ö^n 1CÒÔxÆÅÏV?ñ.TÓiª¯¢›c—¢ÃÏîèüR{)îT;]#`/¿$,w¬$æêåô7vÌÛB¤³ëÓ¿‹N¶k±Sžîp˜™ÊÓqýIéÕõ<±äÙ¨Ÿ¼B«!céÞx'¸S¥Î)â@ùŽË°û4ª¬ë°ø ¥-|0Öº‚Åj9ӽ؆àÖ€s´soµf-ÅSŠ!$”(”UQÊ#Ås>Fèèç¨_–ÒåTrÖêÇ>§]ÛÆbœ$Çã=àñøâþD\DVp݉õ` Í`¼Dgr™]¢–bB^cV} Ü¯³žQèÚWÊ@„Ä_>zÖOhàVÓš¾¿T¡m¯fOLž©!óf§Åþ Ÿw³@ò3e,Ã(ßÖ–ôc`ÑÇé½ù\ AÔ“³ÊõÇo¸ýsZ¬#;š¿þ¨Ã–\+êgI Wê°ÊÇeöø(¡¤·ïvÀï­†¦Ûªt_»±­f_†¾ÖJ°¶¼``*¸«’$™:P cTPÁÁ1VÕ[ÿ¸›w€‘k¹ù€F7Ò´ËÈ¥„Ð8çiæú\ém¼Î˜ž\<ËR•s BÈ•nÕP¸êjvöæÓˆXÌþA¨ä¥föDV‹ÄćÏTu<¬–Q¦2wÿNJþ’„ÝD¾÷úÁO*¼"´"Eçþv÷9Û@áÐx©»ŠÅê{œÐ‰}èñ:aãÛ–‰…n±°¢ÀAÓúƆ.üZœb/¥—~ÚÝ{›zØÎY^j˜Ö»ë2…ùÞTê͇‰þ†€+”!йkåÕâùÅÈ€wo÷Ï×°˜'£°[ÃúS–ðß'mÒûo‹TU±íZÜ}áÞíý€a ¿éÑRεqî&þ¬YÕ½ÌJ÷ÒPƒŠ®o¦BµE¯nny6àìY;ßOsœq¾­<«4°6f`cTÀá6ö©»#c®k/Þ:Ô"šÞ=W5#x¤aŠm_ú‹dÞƒCKübOV+r¥çñÊMÅzCÅ6÷ßr£Ãý{¶‡5/ìäÜB‹†KqLê4]]Í1ó~p`Ëbpc~ùéYšÒ}ûu•´öuq’Ç×¾ÑëZoE²ò$¼Yoëù­XÌš'ÓŒ)FcÍ"ý{Þëa¥"9?¦Ê­Q‹Ø¡¨ôÍdÑ–Ñ€FÙÚr1*½‚ÀFA²7)ëø»2aÁ®@ÿ€u•*׶#3¢þÔýã3Çf~g'!ÓñrÜUó©¦_5g)F¼±üÖ^Pp¶1od)Ûö”9i(YëÚÎvLéÆ©Å‚´UÅét~ÉM¦mmBŒâø·hxrQvP1÷PÙ𯟔ȼ†"RæôƬqnKÚBnR™ ŒI¢‡¬DM$¢“¹”œ¶OVëƒm&‡›L«ª†šB"T Bx¢ÅîRây/ ‰ý„Ÿã½ «ÿÒuåÐóvlÓ^ýM@cäî!Á)ºýÒ;ÝÙéÀž ocôå¶÷qûo=g“Âê†â±­ž×>âìðd¼õiã+ÃÅë©Z”ì9f®’•G®ˆ%F0ôp9IO5 S¦­ÝŒ0¾ ~xʳä/º®}þm¸rdýÕIX]/ ¦—´Ðjòu«ÁjÅn6îñø0EÙGT1t•xôĺ i“F±¿5¶×˜–ò‘8ÈT!QφÊLè8ô£#ø9뭗ɹp!‚¢À7Í­@©Æq¡y—ÒŸW]"s1F\á·´ÂÓn ­»ãÌ÷YÙ/ugŒK±·¼_ÒõgÜ4ª°eâñšóšRæyÔ;ºO9¿ýjYcùõÉp.Žà]è|+XrlU„|šCH&®Kɨ_Eg!ngLJdI'ùÐ'ôÝe~,È4¡’É2‰{<@‹À•ýŽ¶ÙKíöÙžöäPkïâ1Ž¤ƒÓe g$;©—ÔŽ¶la¿MÉ>}•¬,ô¶òeá–[´À®ÜÕaôUém&`šúÀ®WØúíÌ—ÂÞ=–oÂ$èâô^)}EX€öìø}‡AWµZ)?­É= ÝSÄ$ìMNÇX1´È{sRìrj†ØŸ¾ô‡¢ØÓª±e+²OÜ«à%v`Ï&ÌôÜ!‚uéÞ`ÒÌjæ^å×Í«Mx›Û˸™xöêú+°4Ú¾!•éŸÏHŽV_Ñ-` ZåŒä©ôé†æiñ•Ý}‚ó³’;’&Å~¤nÞê^C|>á~¦?_;PÚ׳œ‹ö:kþZÁRM+.]cX5Ë»"ìÕæN7FÓõ‡æÈÁ7ÆËÜN[¨´]~ï>º§çŒF²µÕY4}Q¥¡+öó^ YÒ‡¶KÔ½¤Š ‘·RïLÃ9ú>:ʢɒÓ Ÿ´Ò÷_RóûüÀÀÑÊrJR5¿w:ø ‰¶k]Uo·xÞ¤hy}¸7Æ—ÄÞF#JÕ¨*¡×SIÊË.ÑMB.ÃûÐ \˜T•ÛêøUèÛ2ŒY÷2Jƒ³¸Üú¸™fÜs’‚æ±CÏOÞ>cÌ× ¼*^ U/¢¸‰Vý¢ûF!)ÏYªRivh*µV±1óâ]‡Ñûž½ùìM"KEÎœåÒ \“õ¼¡«Ðý”€]Ù!•-¢óõO$š:Ä3=Â1Bóþ·:OÙ#ñŸßaJ^i¾E\Ú&†^–gåÆá°Ü›êf$®6·m{DžcÏu¸!àkpˆ9N& äâx=§MÜvæK64!CÖ,ÃIÏÀTï«cÄÓE:~ûc)úÿL´[ÿ endstream endobj 2134 0 obj << /Length1 1438 /Length2 6260 /Length3 0 /Length 7242 /Filter /FlateDecode >> stream xÚtT”kÛ.Ý(ô 4Â0´HH‡4H ÃCÌÀ0tw#!%%R"!%!)ÝHÒÝ!ˆ Â?êÞûû÷wÎZç¬Yë÷¹î¼žûº_vVm=>Yk„T Gñø$òzÆb!~A"vv}ÊúLÄnEºÂp‰ÿå „‚QhLŒBûi à57GH•‰Iüíˆ@JÀî0k€?@ ‡º±Ë#œ½0[;ºÌ߯.7ôàØýßáY'(Ã`”Ô ]vè! 0(Êë_)¸$íP(g ÐÃÃìäÊ@ÚJsßxÀPv]¨+éµü" Ð;Aÿ0ã'bèÛÁ\ÿàz”  G wEG¸Á­¡Hº8@OU å …ÿqVÿãpð×Ý@ü Òýý+ þ; œœÁp/Ü`s„´”ÔùQž¨û0Üú—#ØÑŽ»ƒaŽ`+´ÃïÎÁ%YMð/z®$ÌåÊï süEø+ ú–áÖò''(åJô«? A_»ðÏdà¸Ï_ÜÚæ k7gà8ÌÅ ªªð— "úf EDˆŠ ‹ .¨'Äø+½¾—3ô·ô F3ðóqF8lÐ$ ~0(úÈÇì nP?Ÿÿmø÷‰Xà (€Ô'úOv4 µùsF ó˜  µüúýóö-/kÜÑë?î¿ç TÖ•ÕVTàýÃø›œÂàÃð=@‚臘¸ Àïßiþ¹€¿ÉÿFµÁ°¿šøOFU¸ ðàôåýÍÃý/Ypýµ2Ü€WÐD µ pýGúf"ôôÿ½¿Cþoºÿ•åÿ%ýÿnHÉÍÑñ·™ë·ýÿ0ƒ`Ž^9 ¥ì†B¯…½ðÿv5„þYe ¨5ÌÍé¿­ª(0z=dá¶h‰ó„ù„ÿà0W%˜'ÔZ†‚ØýÒߣ@×p„Á¡ÚWد/:J@à¿lè­ƒ8 ¿*®èý6AÑKõﺊpÂú×ö ŠˆÀH$Ø‹H-2A€½¦ÖPÏßúùá:€æè°A ‰~U»¢9Ã\ÐS°ûeü €N0¸›ë?$"ÁÖ¿õ üW7$½¤¿Õ‚îñïóï/ê …ÍL! Cí«C›¿WÊ2xð­ã-.·F%wGŠ 8¦_úØ©ã¿Pw‘³´.¿3˜¦ý¹$|lÊ—çàûˆ§Ù‡¬ …T•\†b÷ªÜ_^ÿÅ׶Pfk½í¨XZÔ*ùa¢ÓmKSSÎdé«0±¡ÍŠ‚Ÿ­®›&vN€¢od¬© ¨8ÙœÖ]¬cV¶eÔ€±Y<Ò4ÝM_õÉèO¬Xü¹\s,݅תi|ÆÀü|_VOwšÐÉ«ÇoEËÃ)騼ÕXs“z騳Ôê]18äD¾É—¤$Ü¥+t çXÀvªž12²ÅwÑ 8Šíea‘˜IH­Ûj£µMÍcC‚}’¬WÞ^x/Ù¥†o5#õS´ÀXà>1X9dÛ«x ±u£õ8J¹v}q:Å}$plÌ–­×{tgp¿ƒK>wÁ(ÆåHå–¦ÖeôÒ懌ç&)ØÈS …1“Y½9õj_ÑyhlÁ'»¯fòtÏ((îÈßu<6Ú˜~¡|¼»‰4%ç ŽÏ¸k˜‰û–CqÓtYÃâREã$r°S´„B¹l5뜳hbU§ø‘¼- ÷pÐñEx­v{dà%I¬r¥Þ’À³ksà—¯O”è_©gù#€øá ï¥#Åø7öå¬Zuc±ª|Öß*bVëQuñ×> Å  ˜Æá{4œ0D*IÒß«ÛΚNdn*¶÷nñvYc¾O£Š—RX7<›Oã½óÖsùšÓ*Ê3 ÚdšC p± ¹ŽÔÍž² «4k»çð˜rÑmÅzDl ©ê“qsUM‚ž¶›pó×êŸ|,Åä {-A ¼Å.N§‹ø;“ÁÝŸ+˜ ‰òj€@«ÇÚm%s6̲Z_-e=p½'I>|}ù%ÔÓOóKUoÎîìP*ÄÝtÓ<ü8Œÿä.ÖÅ5ÇÒÀ¶p/ã&sêðá 09›ÐeËY±o…pšçÒÙb ÉÍ Æ¢âPò®¢¸ôçàG}Ý¡9¸¨³Ýç&t6ÞÞ'2ш©fè.Ù"/êÔÀh—“!²¶ç‡á8±Ã'ü_A6ÇmçïCI³“Ô¨äÈ'<8…ÞÉü$óS8fp„ö™H¤¥¦­ ¿Y«W „úøš¾A;/}+Zc:+ EïÌÞD†ã%ç§×¨ˆwÍ oñRâ|*§fÒœaê€B~¦ŒÓ¼ïŽe/t‰CŠ*2´„¹rI{·Ÿ¼wÂMhüÄsKŒÀå1°ëMÜCâ#œr®D>ªpq! LQII`hWF‡1KÎÅωjZœš]ø™Jâî@x÷ƒÆlŽö\z¹1‰~"ýüÆhÃø}úÜ’ |E1õ%û•Vã/Í}Wa26cÛìÏYËAkß9 hXéd&WnÕ‰¼S@ô| ÇÿqãÊå(¤°†ïYod#Ÿ]Rev¦F†‚@sdÕâ3ÑNÙ‹íü ¦~’ç«b?Ig›âðxÔÄü \ss›¤U/+ûãbßú. dÑ~û¬p;ª?÷Û&.êœáf-¨÷ÍîAØvWÛÐë”ÑQENÉmæ”'îƒxµV²a=aV¡‘Ú½ý²åá÷Œ¥K95#8C£5aái;j©ë.÷C⟔›t†Î›fª}jÑyLG›Ã£ž&þv%¬f"Ÿî‘@\ejõëÙ9[|ß¾J¥*Kýåu‡ÃÊg©‚éÙã™MÓq=Gå÷¢Y†˜³û·=­º-CŒƒ–öçÀ>9“4=kàs÷F°„ª0¶ì>ŠJ>®>§üóå8ÚŠû®·³9¤ïm¼Ð2-Wb%OT9!Øæƒð~çgtr½O¢¯ŠÔ}@¢mòZoQ5vL6’§I…ú;ÂÇëþØZõ‹;_nm×=H¯hJbÌœx]2¾õ28}ëc">Pž†²Ü5çÂ…1;Ù|oå˜ F6ÝË–pB‹ÍNui¿ÇÃ4Å*x=/\Š¡—±~·ŒÒSOã¤HyК!ÿ4øoätó­ç ÖôüqÈL ƒ;r³v:Þ[±ŠmìcÐía¿³RŽê „ #.=ÏŽÜØž.àÄr–C§ž„Š¸—æÂHžÀÃ_DšuRW¶I‰… Ñ.'(í³ÌZ³leÖ”)±Ž>Z<¤Ç÷ó{í¢Hs{˜!¥Ì˪ l¸Ä »ôvm}ü­Gø–[›öº‰±/&‹ ÚûøÕèÛ¡ŸéÃL©F(>?¶-¡zu¯‚YÛ¢OFùØ-ÔÂóS£Hå Ó˜ÙÃÕ¯þ†Ç“$ô˜BÖ’d ù0Z”™ExeáH°$7:q¨(þàVnÑ‹A0W.þ¡àl³&ˆ´Â©úwkÚA·šba8þì½Ï|…¯¥tŸ¸Ù0·T?¥­ð#hM’ü³ÙI%~ Ë’Q<Â*ZegåØË{›–"d#ž;åx¹mr¥¡›É·}_ DéŽÅD…ÁiüMÙcl×í Ï‚'îÆj¹Æ·Â£\JÖØEú§ÙC¹ü™î!²ÚRžQm×M‰ó©õ­¥^ ›/,‹qzÒ¢˜š« ¸}Ò\¬Ü¬/™7ø¶¢Ôã4ËÝaÜÆ#Ò¢YwHמú 8}Ãǵ*ó™ JÝNØ*+ípIüH_úètêΆ"\J±™ÅÙ€Eä »|•á‹¶Î·ŽÉôKà‘µ_y\H…›Ô9’ü 2D †?Ù°@=™ëʸ˜ÌÔ§æ»&“©÷×j‹½ÜΛ“vž|+œ”WR(\¸ÕqÄôÔô&ÌQŸÚcŽ±gæ«éœòh“ÞÔ²¼ËÙy×ùB*—8Žru­ÑË›õåD&‰t'ãÂᛃ>TvØŒïàð E<-^@?õ7ryÑt«”êÆe–CÇ¡ö;àsLÜç÷toIÛ–Òˆ™X«ŸFŸ¥W¯`;.OÒn× GóöÚº\ú("ßÆ(¿ÎPn!Õ”m<ÛV-ÕtZð7¬^ ~er©{—I6¯âzã))M™Ù|¾IJÎÊqÒ·E¹[ÅþÑ ¦Ðô3ðkÌñà·LŸ $ÛFÜÉíò¹˜,‘ H™Á©-z5ã4¸l¶ Á3^/üÚØeQÐ{6[ruÀ+¢ÆôU‹LEÎ&‘X$:côÔæzŸ6ÐÚí3¾îÒ$)ÏI¦\#ít˜”OŸg.S»a¬Ú&&ÙR·e\ÒÖú©±c&XÓœMG¸bË<È+™ä×%³jï[Dðp%KË›…±½Öü¹]wÏJ{LFaâ°Ðç&6Ñtñh%[ jño—>§gnvŸ=ÙÇ…W|Ûð™åæ7"Nñ–eaì&ØuÄM»/£oê†ÅHê¢mwõ}ÆEbýŸ9 ”©æ<&#)OµJ…Oè ·lî$ѽ{§q¶¸ø/QQ«³,<þu qÄB¬önripŒ ¦ÕCN¦OGgoßOPîO’ˆtÈj\ßë\üˆõö£¦ž]Çrz~uRÀ=¦É…O>Êh’ÿp½f¶.põûèO8ÌÍ/,~’U«ŽUj9¿(iÏ¿& —Ûeïùüàõü§;&¬ ÔŸgܼÌ!Ç© z#L§í„o5Žõ‘†}®¨Û›‰Ðµ6mÅ¥QåyùvLÞÃûçîcwú}Î÷}æ¶ø"xU´Èq+ÝDŸh I( ÒzÅJW¸”qáÎ/= ÆÃgârhÅ•ûzm^Ðÿ¶¿0d ô¯b”œ'©>ÙÂÞgw Ô¨ Yò‰¯sùŒq‡1´ìSÀ¦È#¨G},£ÌSóy挖oCõõ ÖÔöÞð%/¥¹½ÐÑÒ#Mà4É]‘‘}ð®£G2½¾FÝJ}‡95SvÆ€yï×'(–ÊËöt¬-LpØçGŽOš|îsè •æ£‘bÀ©ÚO©¶Á§+Ë£ý¨žÕ­-ê¨nŸ{`®sÊç)49ÿÝÏŸlÞ?,ÂÒÛ*^"˲®#j$¿S*’vK¨K¹Ñ(n¨.ž ´¦T]m˳ôd­yÏö¶m•N‰žÄÎv ¸¯êÒ›Z6Ö2Ç^ÏvsÇñ6Iø^>+3¥¢ª\ã½F±AÅþnD¾z9yQÆ}0Ëío¤R§½²¢¹¡J̱ ®[¿ñ™I¾ê·<ß/¬]¨FªLw¢;ÍGI#§#ÉÖš1€3^´´]‰¤sœúIU°JIã×p ù,ƒåp£BhÉ.o_Ç¡ªgÄåçÕ䤨ÖX.­ÃE6ìxÙp#5aýMãÌ7w9øC˜Œ}¾¹ Ãí[ÞŽÉÏÞCr´Ñ¯Ü*{¯ŽØU-chVÐÈ:Òix˜ñ5a.œÁ”¼X2ƒ!§ßÁ¦O]PêZ!±q©@Œ;RüdLç‡p,Ôd[hÔ4<äµXt\kÌÿùí"fn o‹Ì°;?>H¿n©‘¸òJR'‡í@UßÑÏ+Ýx:뜥ܯƒih¾>Þ¼†¼|Au2È`kX³Jù8I’ùšì¸åÕÍΓ®r-$ÓÖ:_4ß÷ÃW(¢Hû³ Cl]1mŠæuáŸïtqc5.Dƒ Zf˜'½1Âî„ÔLšãpÉ|Xy§I§£ô?¼z?|§·q¶¸fƒ›Mp2Möº‚hÜ#ñŒ3—Êw©>ESä.å¡?I@Ž:é݉JP85ÏÌEÞjü:ÄxcÕ#Õ,¢·IûSá)ØŽq FÝ2/´„¹»’0Ó°¡Ž¥dz_ÍŸ?„†ÎMò÷˜]ñ `T)“8´³ò`‹¸“m}Í! z?ÓF©YP ÝmRýØáÊÜû8¿,wPÏzy1C½ÉBdM¿p ªÑQþ±6ìÃ(°Â›Ž¿£ã2ÆòLþ2÷ºCÃÁFçY£•Ÿ:‡tA0®_ìagˆ9fsMU%8¾ÿ3öX†Àž–ŠDL½y¬½ZÞ‡h’61‡1ÄÔpƒà]*©©‹bRÌóÕÁ7¾Ý#¡]dõýÓ¡r]¾Çïtc>'K&í<-z›ÝQû²ÙÈK›½K'ì˜`’§1þµp¶FLVøƒ~²v–seVŽÆHÉÝöÊÔ.ü@›½Hü9Þ¦”ÄÖXÐzJ[/O“ÇHið ×­ai›{8ÅÕ#áÜË$~úðôŠ6u£Õ^íÌVãdìq–§§ù¢aØÜ´Ð=ÁXDqf£±ã«L7Sß[8öf­_$Ë“~dGJšVãbdì="õ8áîï :Bdµ}!$œmãâÞWᶕÚúà³ø}ì]nãSEý€º2®³ëã£H ÎÚhKý˶•å´×ùw; ûdÑÅõØô6—ùVŸ4w²íFÁX×Ú/óµC ¼ÒL¨¨È·q]§(“ÜÈu—Û£ý~^ÌpQ2¡ÞÓíɤFÂ,¶¾€éjÞúF‡–²«£ ›ó™m¬8S»õµšïsÍže²g•ÂUV†¥ó!1#ïWÀ×[]U+ôöZ|š"e¸˜UJÊ­«Ù-)m{×Å)Ñ<ç~tåG˜õ:¬ñW.39ÕèAV}‘×ÏLŸ’ §üÐáˆ(‹a¿ê3ª@•gµ›"_ýu‘<ļ`ë~ën¶Œvñ¦4òchÎ?ßàhÙúSÇꞶþ“rñ¼áÔT×’#jX»|ÊÅc]¼ ÚŽ›GËŽÎÌœG+Áˆ`r““ÚÚÚ/ð»^Èí[¶#ЙåüŽ2« ¶~îb Áœ²Bñ¹lf]ú˜õï} ÖòFã™DµJÝ;*ÞÃ%TKÃù«øg6‘UìÚj:äes!ŒÉÏX{Mæ0£.Hé軽¿ÈZПR =¾bÜî§'2$—½—g=¹Ye¶f.]‘èlNÔ;$(ÓÅâÚ¾>ÆÙߎ= `‹GùK2©à^"‹Fð¼ê¥møSzÉ’ŽKt{WÂʨ޳ؗú½¡æ%KЮ8¶WäÏ,ª)žünÔ@÷²lÐVü±2•Å@ kÐâÔ»È0äŠGø©gÓ^×›MÌ5„»3BNÇ-¢¯ÎÈ-û¼ÆáôM7÷7ҎꆅŒÚš.ÙëÊ qn‚§„×ÏnœÎ'xw‘¨ÖØHí©¶«Õ3ÍÕ½ÊjÏÚ0Â¥å­c ÁÈ®`#WêšÒz—áŒ+ôI`Ñ®ØVÇœøTŸø ßPH_ˆÎo’÷4TZ=3‚ž¥ŽãÄúðœìÂ{<©Ç½^»·§ÒVl¼îóš+Eé—Nn<-Ú„ŸmbÍäô»»<¶XÉn¿ä[*Ÿshð ÞÌÈ·Û…]Tb3›{o˜KëÔuÖ÷z¨¶–Nø#\‹‡Cœ™œ 1ч[êKÅšiZËs6Ây!ÉFc _¿‹Å6XwS`hfØÕƒ;Øòï4éûÝÄø毟ñ¡ˆ"1dpõ÷òO_²ßßàâ…ÅÙÇboµÍ÷ Ôݦ“ó´º/öÄ£N¥«Ù$—"ý›5·Êu*v©¤õ{Ùá¡{7“4ÂÌÛ) ~œOrÜë±…‹Ëpæ4 µîsk2¯£žÎ¿Ö Z}"—†¡Ýz[-Š“JrÔ¼’ºWõî¥ñ1`±ˆg˜r°!e[oã¤Vú©øöÁ Ì]~´“>ºÔ¿°‡Šz©Å.QY¸–Z ,cóó[ )qÏ1!ÀÏŸV>âÛýÅÑ›A±z'^«–ª “såç1iÄô,ÒGëE­ÛêÔËO|c’#˳Båð5ÒtrýkÁ xQª³\Buêc%It?bœŠ5W߈pR]Ñqš¶à~? ×£™ÖÀ«9}µ3Ó8x“‚yYñPŽ7{ uˆ“ÍRnå çÉöc½/|ÍœD›Ò½ìÕþæà~sZÖó¾…žŽi69³¦)UWøþ(æÆñ…ö¤Ój Ê÷Œ%µÚÂy¨†€ =¢sèü›R©ƒü±ž¢m>jÕüþGì$“§PÇͽxuj“ªÀÅóұ€ž¡ËÄUò·²¢Û‡n†“gR|‚%.LºTÎŽ¤º¬¬Ãœ6Éÿ8Kb endstream endobj 2136 0 obj << /Length1 1396 /Length2 5968 /Length3 0 /Length 6914 /Filter /FlateDecode >> stream xÚtX“íû¶”â)銾ÄèninÆ6`0¶±%!H§´ÒÝ *%!Ò**JwHH úŸñþÞÿûû¾ãø¾cÇñ칯¼Ïë:χ—ËÀH@Šr€©£8aA PU×ÈB‰ ‚@"^^c8ûcðšÂ0X8 )ó¿T100Ž`Sãqº($PË KÈKÊ€@@Húï@F¨ö„Cº‚@-†ðª¢Ðx ÜÉGhó÷+r (,--Éÿ+¨ìÃÀ!`$PŒs†¹:BÀ  ‡áðÿ*Á'çŒÃ¡e„„¼¼¼ÁnXAÆIá?Ð ŽsÞƒaaOø0PìûLÀ 4v†cÛPŽ8/0$p ‰%dx ¡0 Ðh¤©ÔGÿƒu~ðÿÌ(,(üŸr²‚#%ƒ!”ŒÄÑN@G8ÔW×Äyãø`$ôg EòÁž`8ì@øus0P]Ù&ü ÁÀÑ8¬ Žø QègÂ”ï ¡ª(77‡ü¼ŸƒÆŽú½YW$Ê éûçàGB‚€z …Lpw˜¦ÚŸ‚ ðÍ †Šƒ¤%$ÄA@˜;æ qúYÞ†ýr ÿ4øû¢Qh #Ìî#ü|±`O‡ñ€ùûþoÇ¿Oaa Á`Np$àŸê3Ìñ÷™°| Üh"pOúùûÏ› ^Pÿ'ü×~…TŒ´ÍTÿúø?>”7ÐW@D( -.“JJJýÿ]æ?øü/«þçr *j"Q@éßÃû‡çZðý‘Ì-à¿;è¡\†ùþ¡¾5H!<„ÿ¿ð+åÿÆûŸUþ_Ôÿï ©{ ¿Ü|¿üÿ‡ìGàÿ¨ì#ÈBEò¿CÍ`¿¥¬ ƒÂ=ÜþÛ«‰ä¡Œt"P\@XL$öÛǪýaP8âü›H¯‚ÐG PXøÏ/! ú/AuWÂW ü³+ATÿî{ AAªOD\Æ`Àxˆ@2qq ¯0A¦P˜÷/~…‘(!HÀètDa?×*" 3c~:ÿ*ñÀ`ÒûÅBç¿Ï¿tƒyà €ïPÙ.uZk”Y½–†.Î̵GÄ[ô†‹ãnLæø:ë\ÊÐwW±‡V2 ¦¼/ {çÇzûËñ°·õ‹ÇAÓ)w³qîô.¨ìäöŸ|}á〢XœF$S#h¶ Ù©ì­H­n&™½; •ì]©Z.8oÇ®XJoÝ€E‡Ô\)¬¸å¬ö â( .ž9Ü€…5u,-Æ*ÍÃXÓdäœ8 {¤ÒÍ|‚_°ŠMøüùÕã—½©¢{yÚÝÕ•atÌ×|´¸²û˜ék5c/ÜP?T-MŽ¿Î\èvcšÄ^rúÛDÞËç }¢NzªAú¯›<ê‘™‚þ(¥&>Îf’®‰ú›ÊRýû ½mŽ3ûO“_ç† k¢S(tÐu¡Þ³-]¥Žêß$KŸ³ ù7Je^>÷²tbÞ}Ëø0G¥vÛã¤>T…:ÅDœ,ià{;öØ.0t´-æä7Âõ²Ñv>Gד¹³ë¸2GÓ¶KìK.ÜHPdñ wðÞ›Ñfûêíç4°Ë–•Ì´1ºÊ5Y’2³Ë}-dL›)Šï¬8Pz^ÙFûÜþª"‡š+oãúêx—èO×bû\÷Û ü¶j&ÇZ½o<Úêa¡õ@ÈíÖèPÄ]+ûTYeFj\‹†6Ì HB8$µDê/_°â¸kío«1¾¡äïË+dON{±"Áf³˜ÉÆ% ÷± ·d‚öctô“®µæÇÂJ­4óU/)-9PCŽÜ–’y¥ýßä¸Õ ½4 Ë}-×¢ºê`ºÐcy&4¥ül<éŽÆÃJ;üZ †dLáÚN4äjÒM?µûŽ‡ä”݉„‡:)qlõ‚ÁØ¥½nÛá"›¹“àXÁ¥\ÍÌ ý8Õ§àLbZ…Iµyœ¶϶ž#oÊ›µB”‹9¿*•F5sê Ïf–B”— IdžٹP` ¯VôN*-xRÝ„$&t…{8ggÚyB¦…³…©D26WcZ|•ÒªŠf’ç¥ÍÆñä§Ëþ·DÐxÝÌ^EQ<çË×1]l©»$ß1)cΊMÒìŽfjÿHW1|~!ã\£t|ÍÒð¼;Øv{ÊÐøTLÅ‚9¿:(ê‰I3¼Äž´Y¼Ž‹–¸‰?±S×ç¥×X àC…ãå’Àî÷÷ÛžéúUà"ãÞ/ÍçkÊ]ä÷ã°à $¹—¶ÊÔbÍÈÂg¯ÒÂ;ëÜlÀ¯³ÉMRL!Â=9NWÅ%Ïê{"ãøcÄÏO—Ö-nÓ²KR¯EYR»Wô }ÂK¤«½è3€W‘,ß h-ã/¨ôá‹ä¿WM\Ô‹i›uðéÒ†êq}ÏÔ5‘}hüÙ†íd,;Ù•™ßB„·‘ |yb‚|ø ų½>Äy1/|éÃ’ûrmzNTáÁ#룧\-@–@y_Ãæ6†&U%Dd‡Y”Tx¡iüíhý-¬[ÖÜìb•cö}üÑ#ÿÂqrY¼¬SÊU=@Á¹“ûÊó³DÚ]¶ÑÒû°‡ÇŽ=¥6Y²›ãÜÉNjƒéëÍŒ¿üRÆΛ{kÏ^²X¬a¨c4)çY÷ =-¦Ï‚ÄwñcÖ%{3³~›]¨TQk—ÎEcñ÷…Ôòî•Œ`æ6ñím¯b‰gf®¿š>4šèŒÖX°ô¶èªEpåZ媸y]µó“˜– ©RØ|½¬µ²ú* bËÜ<Ä„BE…óÁÓ¢v³ËÇÚMZ%fé¡âw¹÷5¥jû®g¢¥JÞGûø,ÒiŽ”ìUoq0ŒzãüÁD§ÿÑtÈYÚq©º4Ìo…ù£Áv9$^|§¿7<'|å]jj°Òúî£J6@EéE;×òÍ+OžÝ¡ý|¥ºZrÿYzþ+ ¼$«³8ü Yp¾ÖÆ-y„QZ¹ÃI‘²Qâ }Ú=©¼Åù]¸i¥ãqÇÝìš|° zWÚ¼ëQòøínE0±#Ì’Ö†xR´ïfŒ´R#µÁ1다{eZN¶®DÚؽ:Àõ ãć½ŒŒý‡ŸNXnÖËyeÒðèå…t4k}\É’º½;É.*”IQ&®%›ÒE‡š{›þ™®Öœjè©òŠ–€­8ƒ€? ¢Ül&Dg/öÖ•¶í倚;UM§ªRŸÄ„äÌÄ÷¾ 'Þ¿¢Ê}mñz\¤oÿkÜΦ¦“Ã%Ldò՜ґ ·o °¾f"—ð›}ÞÔYQô'Ä %ÃÖ·*yfœå ß&$¨a~^ƒ©J?L¤t×zç»­¦'0xk{æ*³Yl4U©^®ö‚ƒ°ùý“:õnbT]š‹e20«ÎÖtnuQ£- Ò/#bͧ¶˜ÉØÖ@w†¹ëïxÅïÃϤˑ n€¸T]ëäìÞÐ[þ£Wµ÷èLLpã'‘Ný2ÏŒ Ùn‡QÞÕs•’¤à£Îxå"³¨*0¦:I=ÞÛ£öÏxùµ€ò8° φ£(´k ”‘8{E’{Œy=žâöyËŽ.¶V8â=(é¤3ÍI5™v]/QÏD¼ëÚâ+²kq—ëº'êï£Ö¨$cçñ¨f%a!’ý„ÇãÙþDmÀ.5g@¦ÀŽ[u¯îo† ÙzK0¹öî,ÑÕUò€ä´/£^b]x3ÖÄÓ…Æ@ÊÅEòyVzÜà{m®Â8 -( ‚9=ŽgÕY„1G:{ža}L|r’ 6zƒ]bå]p%¦ 9éÓ bZä2Þ⪶4èpØ*ƒ¦ @µº¦¹â-´cËø ë©·Ðç@*Ý;ýj{‘hA±ãž×ÄÙ%WFÍMÜ^hu8f˜.|mºTí¼²^z™Œ± -qþÖXu£,ôÊž¶?¹¢Î.ÕÊõG¢\=Ë‚ûâòCí°…UNGjzVÉ_NÇ'EлקÍ]üÅ3c'k¾RßT†Ô§lŸF$Me öÍÜ÷ƒP~у5Qã`Cû†ùç[s1åÓǦFǵ.e+ÇQò¢·¸©È/>œØ™gßaåöbfÚ“Ų3O¨‘Äi„÷G,à_ <`´A‰’Ÿàˆ>5¥…®ÞŸg{R«v½/îvÅû3n³6ù$ÍA–Wm‰ŒèÙôè„sQVÂ/ÅÕËtò+ŠØ‹ÏîÄ¡HÖ”ÞHNûí«ÙEÙ^ù™åž¡[L’,Ý;ÌÌüŸ>–ZeÜË=HŸ& ₺w=6þ6¿ØbàæºäyëìAn@ix­ä•æQkĪîPÐ'õ˜p²ipõ+xRÈ­Ñø:¼l¸n߯³rÕ/ï„Ç…ê°_GÈ{ï^Íý{½QÜÜ¢ž†GŽÁÎõG«&nœUŒ¤ ã»ÈMž+à8øEŸ¡¦õ㢱°Â¢®W ‹uó2=îEé§üº ©Õ ˆ8ÆšþK»ŠZ/³·Ÿ‡°‡¤^;ënq¥™íc½Æf½úEdÛ퓾KU½óò ·J(ö¥°ÜD=ª®¼4M‚§Ì)bþƃŒG<"ÏîhûÁôüjÁgò{ÛãFµ=NLS{ñØD”IšÜ°Æ»ŽœÃlû´Éf Ç×(› ÃÁ\UP—lœd5²NpÇ|þ|}RéÕ{—DUjF×µ‰Z‹`®Ü¹òÑÑQÄjýð·¦9&q‰[ÚH Ý@ìÅä“7§Cšo[>Ѳ»š]³QûþfôîÒéAÜw!-ò‹Ë©ù¨p¬,êJ·@¥œïRµçqðFMÒW0Ùê!!YÓè,d˜<°:£HÒ­„¨«JÿDʨ[2¾Ùá¯^H{jodA5]f®.w®í#X ‚ŽÎäf£-ÌSŽXy¹ENfتI~¸ª®]˜#Ä(Ãy÷dÎ;ÛûD–œfpªZúãÙV?äGl!¶ƒAs§—ÜßL‰é»Ô1ŒÌ¢¸ïû‚›,L ŸÒâ¼V‹®#í/~ÔFå‘ôisyôìUÉAYnù79ª ¯àRŠöTPǽ¸»ÇƒŒe©ÒÖØ ó}6¯2+Ÿˆü‹ßZfRbDÀšÃÓòïèy¢ä3¿Zîª z›ÇŠé™Í£Òk¾Ðo¬“«"s¯ >¡s¿½xÍ'¡Øð}“TÚ¼:l„‰ssöJ6r¡@ÊlZKn Wt°¼¼'‘e7 ‹Ë(“ÝöˆwÇ$Úå ¾w&6v*x¤ÜD$J§t<³•2<³ËtYPi¤Œ˜L‘(±9“„Éþ¥Àñê&õ¥ª—C ‹º÷'}ÙÀ“—IÙú/zÐ ŸÝ(?Ø´õ•q‘‰7ËÏnpî’‘ÝX8[^xSŸ>ãP|‰E[{ËReÙx„–ü.º¿­¸àŒ¡#ù~rsçÃ:ý¶oýqäÃÃ@ˆÖ¤©°ºn-+ÙÒÝ”³Ú¿öiýIC[‰%ÓsäXÐóïYIr?xÐLÿ‡ž;¼YY¿(1[:Ù™3¤Ì@O66Jhr~}LѧK?¡ÝWÜÆGñ›I^¼ƒýóËwùÞÝ=Ž¸î±îš—ý¨T¿æKT<Í  ÑÅ¿îMaTÌŒ‰l<\ÏÆ"E¶éœxŒ(}­œ?Ò¼}êêÓ),ý-ý± üTáP9i¤µƒù$‘¼õõz®õBåuLB[×®†Y3Õ³™²e˵›û™îa/r0WÏŸÒš°Oð|ÆæjH(I«-Ä_l”p§îIKæýÚëÍý±'dÓí]I8Ÿò—KkÇ' šÔM–OóoKž$=X%]:ý~ÖÒfcäíµ= L§gœ6y«<&àºÖÍ©°p¸WD©i(P¼úxü‚ÝQM’^Fä Ç7Ìó&ñ%·éµÒêÙšN¶ª>5Þ{y«yßÙr§ûBÞ§€”Äk&l¨àÒIK˜š©œmz˜‡³4yìà—BÄÄD”M):Œ¥h.^–l‘Ó~—›%ò9s(ê‚ƪõ\˜Mì JÎØ{mÙÆ=T%ñí¨1ÊÜ«‹Ú-rà§6Õý#>FË "‘/ ºQ“å)óµ§;c wʺ¦)·”…Œ«W£½ õÄît˜„¬ôÿ ± 䨈1V*Gô\ã¾ãyLÙYt\93oõÂs«Ë”lMp¾7ó¯«ê¡OCxX›_6¸^Z©‰¾|p¿(ú ÙîO‡•2Eöîǃ2ÏÕÓúL3TáÓ Œ¤Î—9"ÀL:$æè«©·'”û†GutéÄ™9÷üÊëò¾q#T: +pFz©™­ôoouÔfZžÁ„Çm«f·ùìQ¡y{A÷˜Jû›Î#B8õ&²…ƒ/·Ø‰µ~cfqWN¯½t#Šâ‰Á »N›bÓî1»öuøÍVöo„C7ÃÓ"Cl­wš¿|Ę0+ç„4¬Ùó›[a(Ë·n~-Ç÷c}F’2ÚëÔéVP—1Ì‘øéO0¥;þlZCç…ï¦âŸW;\K'Ž´B«o XË–ÏVC7 ·Íe#þR¶( Ÿl'÷²À m#V-áRDÖb …Šå%:]?Þ(>üHÇ\¿†ËŒ™?ùB1‘%Ù’Ç]I<7Þ¨d ÝG•3ê èïµvG£"=#£+­» Û]Íb̵o¢7L*s&J)‡vÃ"7¶diüµ#“×ØA+v.X:D“DÜ:ââÇúéL¶ BËzm™šwä™û¿ÆË'ÞÎmpÇ—Ž¸=+©hÕ€çU¬÷+yúûý GOLœ ü¿ÝdºìJ˜eß/[ÈõØÚÙ÷í•!‡s‘ólùíöŸQ_]O¾Kˆ¯q?Újå,äaŒú¢ëÜœÿ‘ÑäærnIþí3n’Å ,Ú<%úO8¹è¥èÃGÊh·—_]ë°Ôw¯Nõ¿œ³\?/ÛÇð-Af˜«÷1Ÿë$‰º µ¥½ë©Ù_3Sbn¤bGS^ð—ÃbÈ­<11«Šµ÷νM6=ô£rûÛ¤ÞuJÄÓõ ÈõÀ´n‚lƒ2%’tÂú‘¬•¶,Ã#…¤{¥¤~O€_Žù[OÚâ)ZÓ¤—‰á(šÙõ³™„× ¬ˆ[Iõ¸|qÏV>¬Z0 O‘Ë+¢ŒXf)Ø v>ÑÆÅ#¹é³¯ùÈ—æ)W;«ÒýS[;Çú\b“ÉÒµ*ÈûŠ¾¦)JÖL¶²„d¹d‚uüV’5^ͼ@¦åÔ¼~!²Uï茭Ž9;|Æk.m²M¥MÓ‹´è·Ò³¨+Ágºš|(žÎ™¼÷Û…ºÆsW‡¡j>ZÁ¡²OÖ×È߶µ4pA"Õm:+µ3Ë0§ºj —/Êd â=þu¯n÷ endstream endobj 2138 0 obj << /Length1 2132 /Length2 17124 /Length3 0 /Length 18399 /Filter /FlateDecode >> stream xÚŒ÷p¤k× ÇžØtlÛ6&¶“Žm'cbMlÛ¶&˜Ø˜$ÛÖÉÞ/ö¼ßÿWS]Õý\Ë׺×zîj2"e:!;# ¸­3=#7@DNEŠ‰ÀÈÈBÏÈÈ GF¦bál üŽL èèdagËý‡…ˆ#ÐÐùC&jèüa(gg v±0±˜Ø¹™8¸ÌŒŒ\ÿ1´s䈺Z˜äèÒv¶@'82;{G 3sç<ÿyPS˜¸¸8hÿvÙ-Œ mr†Îæ@›ŒÆ†Öe;c  ³Çÿ„ ä5wv¶çf`pss£7´q¢·s4㧢¸Y8›”€N@GW  à/ÊyCà¿©ÑÑTÌ-œþ¥P¶3uv3t>ÖÆ@[§[ #à#;@YJðÅhû/cÙÐþÝ=ÓÃýÛû¯@¶;ÛÙØÚzXØšL-¬€/â²ôÎîδC[“¿ ­ì>ü ] -¬ > þ.Ý .¤0ü`øo~NÆŽöÎNôNÖqdø+ÌG›ÅlMDìll€¶ÎNpÕ'já4þè»ÿ×ÊÖÎÍÖë?ÈÔÂÖÄô/&.ö ª¶.@)ÑÛ|ˆàþ‘™lŒŒŒœÌœ ènlÎðW{àßJ¦¿Ä|¼ìíì¦4€>¦À8/'CW ÀÙÑèãõ§âÀÄÂØ`4³°…û'ú‡hú/üqþŽîmÆñc0þõùï“îÇ„™ØÙZ{ücþ÷3Hˆiˆˆ‰Ñü›ò•ÂÂvî/:63#€‰‰™Àññàó¿qþÛÿ°ÿ[ª`hñïêþˆ(ekjàú‰îý‡ˆë¿'ƒòßkCøß òvó Pþ3þ:ŒlŒÆ_LÿŸ—ào—ÿ³ÿW”ÿ×ñÿ¿‰»X[ÿ­§ü—ÁÿÞÐÆÂÚãßóìâü±rvbûMÕÿZh9 ‰…‹ÍÿÕJ9~숭ÙÇœÓ1±Ò3²þKná$ná4Q°p66ÿ×,ýç0>rX[Øìœ,þzï|x12þÝÇê[}¼[œ>Žìoðc³þ7¯˜­±É_+ÈÌÆ0tt4ô€û˜€ÄðbúØU ûß#` ·µsþp|pô˜Ú9Âýu°ìl¡¿DÿB‘€A쿈ƒÀ þbú˜Æ €AêÄ`ýqäÿ‹8?¢(üƒ>ü”ÿA¬•ÐGeªÿ ÊÔÿA•iþq}è ÿAùŒþA–ÆÿEl:c;ëfÿGÂÊú—ÄÆæÿ¿NÁäøÁød0˜þ?”¦À¿”ð3û~$3ÿ'õAs{s í²?ý?:iùü jõü`cýü jóGe4þˆüW¥vÿäþ°ý¸´þPTnÿúÃ×þãÞ°µš:ÿ#eú·ô_«ý_ñG‘ö m÷GÓ>®M‡?àqÇ?àK§?àGç?àK—?àK×?àGån4ü£R÷?àG‰ÀVžÃÿYcÇοÝ>6ê?øïK tÃ-/ØóYÖu<Ô}v£Û`fý4tq£¾#†ãº¯ ÂûxÍ’¹$Ô3·ù¸‰zÌ\òÞl¥Éh¥·áɹÜÐij´Më=…Ê÷tì$ë¼'ÓÝ­eÀbkb†yÞ¢ôÃíFp– ‹Öárw>ãP‚—¼¥ê¾Òüš¢é„l"*›““,¨S(£³J¸Vhv®KŠÔéJpâ4_üð1upŒ™í÷ÐJ-¨ÅVê磣Oö27GüßúØqe ¦³d;}bØϱ”7A_)œÆ•Hx2¥v¦®$¯v¢¤”ÌZ$‡7ÁSÀE6c‰/6a…²Ý(~ëñPw‹–´E,'©z²°Q”õ•—K ñ‰ÃÂ,ÁŠOŸÕÑg£ÈxËÖŽLÇë©8%eΦÒŽ¥£5)(_{-2MÃŽE:F[Ð wâoû)‚dEA[.Zë¯L:תtô#I“!E\>“¨ày³OžKpYäb¤Œ?3tP0´þ@g8©=Ãî7yʱ[`ºP‹Ô—X€ë Ϫ5Dp•%¶7W~J¦+)ÆÎ3Ûo õq¾jßôÂ!O*Òœ½Í4—‚ ô^ÍsB™¶0}mÝÆ Õ}ªç Ãö|©=±xOO¶Î¼POyÆW­7‹€/÷™ÌuˆÒ3¦ ¬Hò «cïó=߈Vî®+¨w½%M¬¿±¨'&½¬ûò ./0[mY1Nàdý†䪼‚…½Ì¥´dlÂÀµÎÅ7¿®ä7Aåv~2×m÷n-ÉO9¾hSz#Ô«ÙåÛ¸~íŸ$fšKOv¾æÎËæF]yÊ.KJ¥êÈ×Sw\(欞Rƒ£ó H°âB!££â0jQ’NÎk¸âR«sŠ:V€x[¶­‚5×xSåÁ-¤4$äОaóò…–KõÜ—‚ïVÙBJºµ,JðÇxÆßJƒÒƒ›—5¯¬*z–rÆôAÆNCo£“wíÓ )4= )Ý^œÈ{ðü¹ä¶¶§öDL‘#݉y¿%¬løÊ›WF1ž¬„ŠVînÒ„F;œ]®Ø"ðÄ1\Þi4æ”dÈ´ ¸i™ÔX°œ÷©E¦îÓR²Í#zÓ'*ÝŸE,Ih—c¸O.áÉq9‘7 s`;ÁPó-6Ý%¼f#ÁxÆçK³ˆ ‘ ÆZåDÏn^Ϙ0øƒ•ëf™]| `».âÕy'ÛH°T§ Dgs‘ÂÜ Ÿùe’/[`ª^Óá\®›ò¨ ÙuÄÂLx ftuXš©†Ö8­ìëÿ嶴ªñÓy¢Áë¡ù¦†“„qý6Ù>RÜÐÏâió³vÞ÷Æh5Ô*s’œ$’Ï©Ð*““¹Î²" f™smJJ¯,y«>rÆ>}WÉq¯_S%J«u̧muÜë¹H@3-°,}â¦I³wà? wß™”N©méV5•æk–ð_P‹gH37à6û¬h–V°oí,¿™üƒ‹¸•¤T”ú’t¸¬#õ·‡‡N. „ë H²—ql¦¢Û Ãåt‘E.íBJ[Z>H;;w´€²?²# só-h0,J4͵îi×à¨+fEHÝ·ã¥Þ75>]UD,Nßñ‰­(tÀêÕk(1¤?Û>þ5fw+åÒ¶tàÍÃR«]&Qä~Y1•G<½/ÌôQöûœž™¼µ÷åQ$üÚeúá´%)}ìcšÈ*Öÿ̃‰ãJf®gKâOQ\=bBÊçõ2øR²êÕ'§r'Ñ lç*V6¿†VǤ(Þ”ÄÌ×Ú†ÁŸ ðÚýAm‰ìÏò},Ú»¿Y;HRÄóG€1?Ý"&wÃÔ­}™´Nb.T$4Ü‘Á÷ŸðKáöYFîäb[Ö5=m‡ÒX¶6½Ê§{Ÿð^ôkà…˜®°C‡’j˜ß8ʇi»í¥G z)³&jÂ0Ú†#(Òžõ-’‰Œ1.™;þµSÊÍ«vß4  )̨ØÐG˹ujÌ‘²5%:‰’Z=\µóŽýæ Æ™ †fžeUv8“ßIÃt9áó*²•ö›Änn.ÕÍ@–~vzêŠ#cH-åv{®ÖìÉNx‹ˆÕ¸ùïe vV1V±Û_»³*r» Ë|w뼡±™óuí¾[p¶ɪàN1¬§ýä@¨£‰ÍyÖ) pڭ뽘œ¬ÆôÅ2ëA ìˆë³/â³€¥¼DˆÛáãïñ;œ¶ ÷]öÌ•g{ ƒEÅ-@@œ¹¦#ef >AB¿6õ+ÎfÏ%$£R"5Zç\¶,Í d‡Ž'´9ÇHot€S¤!üôÓ)Y¦yî]ôkåU£NEŸ ÖÆ“ ?‘æêDô1W»\˜bÔDƒÚ¼IýV(ÌzÍ›ðþFõ£:y ÔåyÀžÔ­}¦Ðm1š9uk) ŸV²¦ÿåóÕ>RA ŠªU‰:ó‚ùn”³U¹/i©¾ŠY±ìñR“Ë+ı'—À¹^ìÇK-%OE.síÚéÑÍ8ÝÒƒ/G%Ö .òíןGŽ•ÅäŠI‹ŒÚ]éÓBrºY¦ó{M­©{¶ªüOç_óÖÈF[^)§{TâpSÒ£2d’… opk·0 þ%YYþvxO"Eg`Ƈ‹]eºN§’P'&â÷ŠL—rC´ÿl?Ë–ÚëÄUÓ€Hä3%¡éØ@º»8Ͼ–Ýé|@ïäL0ç>¶Ä·€ÍÝ5Æ å´Â£8³>­îCò¶öFA‰C(­g(XKã¡H¥dVÐå…ݲ½Ÿ­{€³ÖyYòßäÎOÚú–AáM7ëàŒ±À Ü-N$¢<‹…Oþ¬cYIA,,ÊÃKz`ÌcK²%l+õw­ä¡&2|‚}±o¢a÷Ý<Ÿìv¡â‘æiM›Û”fn<šñÐ¥gz„Ù¡æ}£`ÜßÑ)9B•ëtæÆ‹ùjÖ³14t Ncpä˜È²€YWJîì¬!óÙE \–0rù8`¬¤pPxŽt"z\Á9sÀ<h½óÖ­]iÄ5ÚòëQ^©Ô1¨†ŒQwó|˜~»#úĈs@%/UÄ+‚VÃå” ‚Œ®k( õçÀRÊ^€Œ*ŠÔªgkIöòŽéãVÓ>&H7Eiª¸…›9ó¨Q!Ž(¯F Òí”äÄ}7¸ëÜlÀ<¿F kÈöçÁ9ò-bg]×.U–«‡ÜÖ¾X Õ¦V¼ó£ßE«¤e8“Ãk{¹¾3¦÷º!çIáˆoØظUÂÖ«Ýc‘þñ~(ò Þo>¾]Œ(ØëÔB×ô¹¸Ðë(äëò‘ÕŸœªlÚ­ú³ˆGð½ÓT‡7@(>›Ñ­§Ñî÷צ¹(d£f°O¢E\¯X>câ¡à Ý•Œ÷ÆûÍ;¶yJk=Rwaö ÊãžlÀ„ze‡«wàEÄû'ÅLi5öè­‚T~;ç6ræËËã<ÌŽf-õIÒ·n9En…ÉŽñ³ž ·¢»v†kWA˾‰÷M}Ù$ÁÂû2·~ù™¨ ñ5 ˜Ö¸Ñˆ3·Èá½ó ²6C¹ãþ¿c†¼3¼K##?Œôtf‡_妦(ŠÑAȽÄp¶è²H‹v !nQûð0zÔêhy‰†u%…aávM—Ï[øˆI‰HÖgÌ8T ä@´'Œ@š¥†àû.³·üô0ø×÷ £.¢:³†Oþ¦ýõníá΋2yZóÏÎ÷·äÓ`9´¼$ ¯T¿5¿Óz~ôŒœêŽ‚„Þ+ª&c"J*Îiª©Q'¶w™|$ÚÜ*92•ll;¾¼Ô  Š„RÂëe6rnÚzWÒc, ‰ü)ãŒzî5ßd&\ÿL<¿Q|.jú@´¹Áç_ª¸¾Íýr™@Ù «Ñï["åäJ`é—:š`U®o†·Íp]ϵ€t}«¯« BÓÔ›äyD˜€ rNKªV:Í…bùólUk£ñ&æÿÝíw×Fs±sÐü˜¢ÃxÕérꌄ•ñÄÝÖ™WA&¬w±RÛ#Ø«,(2üÖ.º«ˆùé%[®?ç±'§ÝÓØCbMÙðð\åÑ/?„\"Ss6ÚçÁ¶[ãÔ¬œø¬]~M惂qŽx×ýè×éÔ'óº˜)©ç''÷Ä“‚Æȸ­ATRëT~²PŠ:W˧Y™'ý!êÐq+qûr ä3•õFÐœRmÚ!Gfê‰Ü5—5®I¹Æ£î|>ã8÷7E‘騧qjÏõÌavº„,yk:b¥é=ôëê ßåæ²õ³RaXXn!×xž)\Ï]U‰¦T æq+‡»/ÃÉ>«)óüÑ1[¼‰9yä œ¿‡_·Š7æó¾¸5Üy”4˜7 ÅJ¾c@yB‚›ŒÜD•\–ÐË` p†QW}ÙŸpÁc£6‡Ææz‘ÿ=Õ@>Wí³ÒL’I ³·Àñ·ÊàýÔ(R¨Áœá?7¦ŽÙCä&”ròÙêÇ…Š¯ òt@PTU§•§)Ny+¿@ùÁueð8`W„Ù_†¡qk¿[¦[çÿ~iÊÜòHÄâÒi­¬²ü'¶>q¾_TáCD$‹œ @ï‘¥Ÿ sfµIPúòU Š`-h·kܸ¹%™Å «:ªB/ÝÑ&¼‚úX£ ëUp[jA|ºš1{ä å½ܬg–eÒëïš/¶Ó‹âÜDg î¹(ÿþû®ˆ×£ÙýØd÷E¥94—ÒýÕÉ oÙrh²ý@±n µè:¤‘µØ‘úóø¥w1 ¹‹O@|{“(}‰Né­DÏwóé ú¬ã¨¨(Ëq|Ž¶Ü~% °_ÉÕ*DŠ‹_èzì?yö";ÁËýºM㩾_ŽðA4zßfùÔ•ÇÀ¼æDE¥‡V²qgqWšü+2ƒ Cipé@9‰=˜ – Ë¨mK›îÍ©KP-’C‹Õ!<ª6BL-P¨5×<<’b–Xý‹Ú Aàe5:³P!$B^$¨=ëêšJv^Lt¹omæWYyÄ×ù‹Q8l™N´ Ãòóͳ¿ÓéÁÛÁÂÈ»&çr‘¢zSq¬â»´1Ø5zÃÓdJ¢¥ PQ£Þ[¼ñt(±RwÜeK²Ó–iÍÓÁ¾sõ×x¦«ô¨_Éû¹ šV΂]9ô\~lWn¼ýŠx:cep?ï·ÔfjUý˜qã’þJ2*hÉÞkž î£sÉTœ¦FÑöŠ¼Ï}6 ”Ïs¢^+µ8_±õ¦Av²üPø%gÄ»ŒVäE1D.~-€Y:%L_l5%a×…MVk¡r ±'&ÊhÉ™N¶»—°×Ț˹ÛÎÑçiÇÝqÁÇbpÖ'ï†ï¨«ÒÙHÉÉäÏxïÂNOx3-sûl¤Ëïmðþ‚—6>"î´/qß$º2—éµã9•ƒÙÁk3ÿ™û kÚÎX×þ['áZäÝïF Ky„[£ßöøïC–—,ê{=JXÏü7ob9ˆFÁ¦ˆ#:T^í*¦ ƒë~w_)­!² ƒ³&Ã…1Îꆅ»gxs ¡_¬ÚŸžbIÐåR‹ VyN©¿™¿ñè¶$I1éNëÿðÝÂDÊÇ¡k¨ú) 3Kº›Ë5ÜRá:²­öyã!Žs)0õ(Eç".(ëñ¹º#uâ;Ñj_æs "gJHÎyã¡Ô×ãxÝ7¦Tðï(G=öê­(Q/|çívž6³ª$÷^Ÿ–BOq·ûŒò äöügxRéü\D=~²§àP¹b ¥æˆQâ7¤)n¢Ì‹Þ ‚e[x!]¬§- L(sCRö åwG(*û>ÑEÃ_;KÐèx5ƒ‚:¿INVìf­$擦cQnÑkkуys»ñTêbØâzk´&ùÄ“h¤b ¸ÈŒÓÏe pÉ1ÜÆß¾H=Þ† Ì¼b’P9Iò3Ñ®ôZ瘖K<ðkžxâÞÐßvIüÒ|Ý/féžÑ®O ÿ‘g/WïG¼iéŠkxI ‰ƒ¯>°bO+±I9é?á߃¦ %#š V‹U“þKZÞ[ɨI²ñ²|ý¢¤Á¡zaŠ«eÆ&—keÆ”͉ûˆcÆØqÄÝ’°¬\®DÂc¦»”(óÎBÀ9aMž´h:>>¼ÄÔ‹ú‡ÉNo«ƒ±u´Ëù0^/!Ô(IÍýêÑ\ãGÛóÆÍÄdç~ ó Ð?ÔË6·ªàœJôyÚˆõ«€Œáœ,SíµË½#Z´Bëð8v¦Xm‡×܃›þP ó6´œ-ÝNº¹çY­¬Ù8犣ަR´H³r5ÑAô¸·°5Á¥ *É•›k¢ ù>¶„> ú–R‰“ç»ìRÆC‰…^¨Õ¼gb)é9›ñÆ OH.„iÖ¶-êñƒñ|X²¹øôسg‚ë²Q‰-ow6’Œ%â¯G†ÚwÄc/txçÄá1˜\–!\FÝçÞ?}­ßòiÉ£†ÞÜz0·í¯µ Ð%4E‰Ïœu/UÀfKâ²¹42L"š¡£ R6tΆÀ¥HSÚÆä1dp^ˆ#smø9ÇȦÚ[QRŸg8ØbÃÝH\9<ÙlN,/‘á(“X‰ w»òj6›­í(|ÈÍ+˜¼{²æ5öÊBö…X‚ƒoÝŒh4'¹é#a'õúoVÊáñç#ˆÁˆ;o{ýì3ÊsÁø+*Ì5|a;³Â¿¦jÞã]óðš/øcÌî·uÜ¢]ØC7<üJÜ%LÍhù…ͱík2Üçgþ ß⯭…€Ý~T[Ä=ÒÎ|{<žIP朗ñþžbÖ×ðyÜÊrXì—Pcv ”Ÿ³­6¾7åZÑœ<®aÛÂÜÛí_§†Œihö`lA¹bt šj­±â ñE-¼N"#}9+’âÝ)³+T¶(Xr}Œö¯c‡Cy2ëÎb˜Äô°Æ¡½ÜØ ¯ˆü…+ãPÅa¨»ÙBø;"Ž=7›ô;Q9ÈzIjIŒº-lcÜ:Lä]ÃפkZtaéÕ'™e¤lÍ-ÿ¥ozoså“ÖÜXÈÝ(Ùa¥Ô^úŽƒƒÉ!îcœ÷r¿“V­ëc:ÝÔp¶ã³Ö d< *cý| øf]„¶EÇdÚ®¸!.“bU u}§#ÖMrŠ«eB\ÊïínHi@g<Á ¯^üí]XØÔ[EÓ7·ê¬o|ïü-.ÞX…g/8Ž£ÎÌŠ»ø·Mº¸• ªçšÛvqIº¸z]½‡©ùKi¯òSMT<®ùvÕÊ“f·v»Ì,6|¢}ÅŸ{ VK•A*^ßîY(Óð´&Î~Šðøb†BÛÅl—;dç–¹ÌÊ\èYÌά¯ #Ã"ç+iªë–""IBçÙ€³…ðƒsGDêÑ«âI ìiå­)KPÕb±S!¤@/Ú!¿šìbûÌ+‰cùPÅÅìq¶S%X&Žg¬©<¿ ‘†ô7J³`&gM¾»#V¾—s„IßÓì<‰WŠÅ.¤#k] ÑÞ +2N}’5º ÓÊl´({IJKÃõì¬nS‰jŠ—Ö$4(Gnz|àÕt5Ék£„ò‰Ö}kÛ¾ >*«ë®sì,¹y¹ç*EÂT1^ÈIÃh±Å³.wÊ»ª8U¢è“Ÿ Ð¨˜ÆY·"RW’ãÎ'L9–ªærûšÂå©+ò4’_écu¢Eã“ÀZÅÉárÐÊÒtÈXU•5‹æÑÑo;Ðx(Fô‹–†n æ`tž¢äìvf±ð G½Ž}‹\Þ½¿t*(éUéá³A?ÇŠYí7°§sr‚G'vQPyõôŠ×‰òÊÙHÐ[Ýž;ÒFsÑP±w½ò3™¦0Uëqæí*9&Ò²nœu‹H˜®l† 9ýìØšÚ<Ö¨âZw„Ì÷ë8ã %ü±ªˆðÉéÍu·äÉ û½k&ð<öÒU ½Gã졳<§•@2t4ÚË%‹ÊlŸ-³â}¶ê†ßî`5§­îÇ›ÐôÀ^·öú)È$‚×—>ñ1§X'WÈ)Ř·Iq¢+@—ÝRŸËò _Òob˜`×s,ÕuÙ8Ýk€è1•å9rS¿×%U¾›íM×j㬷C„Þz¦¦•)PCÉ¿E„ÑÅBÜqÂ{H¹©`¾Iʼn&†=F7§.ùäç΋Ð ïÔi‚êºXj…ÄzkÚLŽõk´øzWÆ·ÇЉËtñW—G+ÃxfÒ“OÇÚû¦‘QK£õ1D“„ )|ƒ˜â3j×Ñ·ì㎛/à¥ñ ú,’ZEžû‹·^]rwE¹„SÖ`š†m¡Åq·}ügY¯›$»Ëö[EÍ ïò+ËixÊ~;,Žä2H†ôXô¼UƒI’ÔP¾ø·ò{àÝQ¥ø.úUL.ÛnýdÚ~Ðb$«†5T>ØïÓ}‰<ã"5”fÙjþiÑ £]½3ŠK’B3ñ#œ\ì7f4wé¸þŠÆ¥vKºp{%»Õr‡ „]qWaÿ­ »¼»çI ¿ú#H,¹"gD|¾xbã««,§|.h¶ y½]º âJV&iFý¦Øç»i{2›užÂ[1§—uÜ`ùæd˜]¨˜ðÜ—k¿Ï±ø$=ÍN•n–wM˜¬¯l5¯Â±CÞ)?Ë7³êü‰xCn/*ß/Á}ÒTØaèÈgÕÓ?wÑÐÇu­ñ€¥IŠfÉÍÛâ@C8=féÉ×q¨%^xm›^q”±;6¤ÔĦK´C´5Hµ‚–0Uv™òìK8<ÅæHœiçcDõp¿v|aªNgp~ß@Ç£ýþÚ7W ~îZø(£ å]¡–×&nš¼Ét8œ9%ò­/f#q*:Fõêi¾•Ý?# õþ:'"z‘¿4±zŸ\…6 ÖÚ##£SdéE2-O5S'íZ.Yx=G‰$ʆ7VûãMºÖ,±Ù?JDó2—¥~ÀJ»G$‹álïÇX{‹ ,Êa[)Ú§ÊJ0ÅÇúi£va9„ſtz*ùןrš6#¡Q°ÖáíB¦Â3KhË0ú ÐV+ÄË•[œ“'`áíû– {_¢ÃÏT“{ô`ùsvåðFáú¯ ˆ{†x‘4–¾#~mNº×š1Ó/q ˜Œ'¤áqɺëÝC¿d[ä”3¦ÄÕ« §ªïó¿T.£š¬`8‘ÓÔQB–0¬\–°¨Ñú䎈zÜ«Íø;³iÀ±5ÔÊt„“Bè^NP姣›ï§3"dÚˆ2zï°3PlXpÎŒl0Y'!VÐÜ´±Ù‚µwù5|‘ö8ý…Á3ÔÏo79 ¥¹¿Ðº¨»ZÕù§eõwH%0AýŒC{2âr€ƒuÎß> >Ÿ?;Œ~VÇð7Go$tõ$€b¬£¶—¬_¹>Û‘¦X«¦ä2éµy9¡‚]ïCͳã…>SãH¡a“jœðßU¾‰"§j®ÚŒJ°·ÀYÀŽJ¤Ì‘à оŽ«Ú j9Q-¨|,¤G/ö(Ÿ í¯ Ш{ªMÔ÷ðD9”Ä< ë&L$•žÀ§H sy/ \Ç´D)u'5—™Zèšô›ü/´*fª½‚Îœ[3¤Ûå£ ¡';³Õ!‰Á¸eP%€0Zþb5Wµ9’9@±¹+Ó†ï‹ÒãæÝ<ƒFrzæÇ"ü08SÍa*ª,)©äJ#eت˜ƒ ÇG%£Ž£ð §§ÇòS4¨{h—OúÝ+Ö#;æ±pYéä~^:~;5f£ßÈ©z¤ ÚWÛ›S… ¶¨"úJŒú™eLL²jŠ-£§ó[óbÜfX;¿ñ–jlÌØ°‹/¨—‘%(Xôã»p]ª¹NM¢ö„Fl˜ßC²eÜgê³ SÙV:ÏsêA™öö )s>5jÍWp£dFóå£#bÚuf}ŠuŠé-3wªÀ’n‚ác¾²L ãònÕa·‡øIbÔVX3÷eåØZ☬,Ð ÝØš¨rËMM^W ôh>Ùãqw1r3õW£ë˜)iY¾·àŠ-²¶¥ÔFØÛIg„[o°FÈûº¬û ‚ ÅA.”hΗŸ^LØú€u×ï3aê-óûz¹ bFÃKÃ)µL°µ¡ÏÊÜ䚟ÉÓÉâô$°·Lõ¬{GRðR@ÇÄ¿hÏã‚èúZƒÚn\ Ò´«ÄUÖöÖJÓQOo„ÚcAµOŽEË™¹òŸGï:TÀßÞ°ÿtº&SWŠî·¡Eö³ŒªÂfCq·vbvý¬Ã…X¼¹ƒ>»B³§ãÇ  òUVÓ1]©ø5}Á<žõîíßzúI[Î¥xüBø•wSWa‘ÀqG[+’”«þØ¥Íɵ4]O!‰£M7rÀϪàÇIlY<\Ø–<´Én÷u…ÈÉä@ë XzlGŽc¡“‹»TW.Ó ªÍ”ú3Ñ<ª't‡‘÷í{æ¼*z-è`1(ŒÁÌ_öv*í›Wp–Æœ>©®úõpah}|ÕYžy)l(^kÜ‹gkiÄÄ[Gka{T£Z…ôE°Îa'f⦂‡V¯=œFë¸Â.1S5õ6HñŽn9L¶Í B[$–钃܉ÞtÎ:†XéU™o™®¿-G`—Ëy65\è6L®0jVŸ˜U¤A8(×®_R|ïýŠˆ»ÆÙ_™ ;Ó®pH—Y}^dÿ¦.ðí8ãó8¬’õÁS¥b€›:tD.¿zIgÔü•J—;Änœ£“=C[w–ÊüéØh¤4ívÒ}ÞS‡o³¥©0…¸¸òžêF—¨¾= Vx"ÌÕÌM¸UÅÉûû&æŠë/=2„e`¢[$2³w¾ªð(¦“Ýæ~'Oöµ1Y,<‘Ÿ=Órùï[sèëØwK715Ôxy׆ ùÚ/üõ:%¢ÎX‡(ŒµXöuä·ea¢x­VðhÃòacáÁ£Ùc~§†»'¼'¹‡™L¹sí:G/»üø@Ååuá¼/‡^V# ñD“\¤¡ÝDPht¹N\,Ù¿‰ JÆ"ëßè#ÍVv†ãP³f÷™†¹É­˜€Ðá¼sõ,\•%åÍè‘„ví9mH&¨ŒRûs }ÝœÎ'™(y8¶KåÖ;`‚g¿0ðmlöƳU’†Cœ Ãcä,u7>jsH¶·Á*=ô\# 6kõQ7æ‡]aôt¹9c)¾SUŸHÑÉä‚œÇ Â Ñs6 •ºÆÓ_Š»dÃø„µ8ïäÜÑE —mf¢æ)PkuUEÅb:ÐW,s£ËV¡tîÿd9ˆspùùgÌš% Wç…Ãü¯´ŸÊäô§ûñŒ'_bÊ®³ªYq*‚oÉàŠHšU/±_ÅuŬ‰Åêï¹÷͈¨³ž6w‹®¿W·…öì7šØÿF?"• ¦~ ´jøõj¿déjÑ'pÔÛ’b}Ó«ýc4¥5ÍËãXYæV¢ÔëñwÛEV«pÓi¯1$$bzÒ|©Pỉf[¸º kÔàT á·¡$æ&vÈÌF›*V»cù|ϨÈüó¦i‡Rv¸f3qÝ»J½ÖyÙlõë?îü¶dÅ||ŠhÚI3'êï¿!|„¿÷2Už¸MŽŽ ­gØ¢†™×±»Ð/Šµ*¡ÎÜo æý®Ó`M'bâÔ 3ìÇÑ@qăAîè¼é Võ;L^cÛ%—bÞ/‰I„\±ˆšQJýn›Mfcö’6ØA¶–Ÿ˜|3oIæxuêöãÆÈí>cøë”d¬PE‘Çd鲦J.Ä8´½Âà‚‰®Xw©1\$v–šXÁ;T‚0ò«Öë …q û¢êÄ3k'}2ç)Ýï°Â/ë ß"9­¤ú¿±Ì»0>§SŒê º-V»bôXÙì±·¬FàGX¹`—š»n¨öû{Ê,·@|â»LCš ½ W+z „àGpk­c[jKwÇwíÑ|~›°ðMëîž”ÿì¶Þ"äHk¯˜mêçÕ/Sq*@›0²Y@„4½'$ç}R0qS¸ý¶o§nñ{ÂqsSï8à›A¼q,plÅÖÇw®ã;¦ +Î6ÏsW•i†Þ¢û¯!EÉ–† )•*r!nªSýbÕs2t¿ o5¯åY¿ï`c®‰s;ˆ£ðl2Š>5m¥|2ú`TBŠèq=’RœuC+o4g&kûÞ5ÊÄÚ.ˆ©ðË.zá‹uJTá_U4ZeñHMd#3&¯:g„,¦.wû©Ó˜®yÆ”yÑ0ÂÉå$§¾ ×\Ê¥&H­I ×F}7p’óSH'U$Áˆ~¨E¹µXæbq¹ NŽç-±+/$`åeS'LVtY'¿MÄO£4¥‰¥«‘äø¢µÁmìü` 6´¬Û«¿„8»;»ør²gÂ<ê1u)®ÎH!Ž­õ\Je_®·ã )OTÈŸ]¯ ²ò©ÕiäÎY×q&m‹Ùé—µ|3ÌS[Ë—¡ø‡'âßP÷¸?`ÄN~'³'ÐvQ¥CÇxG4ŽÎ>ËIú”´aX˜æ2¼V“uŸ¦é%»|{FŒaDwèÂ6°ÖÇ™“ÎOÆ~º¶º°[XT²ýÑ Àgò½ùF1”¯³äÈ©rêšàÒôùéwû¨6b“ ¢kå<€? Ή$³ Þ÷ð¨JŸr;ddAnØfæc†œá[ú,Û+ŠKO'AKŠ¬UƒàºÖ…ï›ÅÄ,¦àÊÆ\å«bˆt»Ë¼® ×V«ÂZ‹µyé¾øjMÚñ~~m—ã°0غfï{¸5yµxÖ÷!¬¹â3⨬{5*÷ž’»»E°NÁŸßê| ýŸ_;óJdkQm&×ý=Ò~ÅÏâÙrF=ÏÅ/?3çÆ»pgŽ°â@öž½¡É’”&ñ4rŽ‡;Eà­2ÎÈٱܡ¶Âö³Ù”¨×|£V…ÒÊ“qTbO%ÆÚcÀjcôý:…ó‚š;ýPêŒF„0‚\ŸúÑœùäÂâÝ‘wºãÒÜ‹»ët²»žÏ?Êå“I¼ òwÈTspm£ø—¢ÙÓœB´4"ò"pµš0)ø|‰œã;f¨j/6óÖµu2ôŸäŠÈÆŽJM…Þþ•Y;à‡ýÖ'Dv‹+'Ý#ðGäŸsdì¸R‹ð.s ÷gKæXàoÍŸÂ¥¼âqÓRC°ˆ£6êã•¥yˆMç€_7aó–¹qSN^ ó^Ç r‰M0ׂ©ˆ} eæ°zâñvŽ÷U_š"œ¶¥Í«šŸ¶2´C‡ˆÑ\·A;…Q¬¡.1šEµ¸’šP«…ã·¡‘ùyê©AD_Ÿ¯”] ƒ;nnåw`æ©õË•}adªò¶½J`±¨VîPk6?£6¤Ó<äÝ#¹x«‘Ž]Ù:V™š;A ï”Ô )Ž mhP4N+TF¹x‡UüºÛ!„.Ôä2$¯a˜DŸÎ˜Ò¯ !õÁ'¨…%çÝX­à°ž?CbW3ÐiÌa—MP÷%wq4(} P4\æ·îT•E*Ê4€Ñ§Y¥"HVšD¨7º#Ÿ 7à¡3œúM¦ó&…; ™å@ïÝ\rJÂÝÍ£ŠՇߊބ÷Εäd(qE1@c¿€4‘‚ŠÍ[ EFñåíuÆgnü¡C çdFÎá!âÃÓŒ×Ü«sôÙ?±#‘ʸ݆$Z|wä ÛÚû¹Gê0¨z\to[Uo3mçlG‰?¬ðíù2r‡Ã‰O^RW…Ù Añ7€ºëõgº»…k?š›CV¾Ó*ì<’iNÖÖ_Éœ{ 5C‘WRÊr!?†§{²ß^^•9õ³´œÝ’o½²ë¼iìí”/¶_~UrU»aBÙ‰T©ï“Ïñê©à÷7ÇO)uÖŒÊïÔÒc¶ü©ÚW5(zš ΚHfP~q·£OK‹ÌZAÑÖ¯çèß¾ËËÍQÙ_ìrC—Ä_²Ú~ˆFœè¬ÄìWߟÔÍBïyËPr¨&ÔÈ|ûœ8ôZªð#€±WÝ\Ñ÷<É>ùòB¯­ 5 4ªyRžN¨¸ór1þ;¹Ç0ôž…ç·!;ShuÝTýuxòF²Ëà€jõßS濵e~¼ÆeÿŽôNäK!g‰ú¤8¾T@.…$Ä(/Ÿ4CSÅö<”PTO n’¼ˆaµXÎä[‘RÔø›{×h7ƒôÙ¨˜%‡D(ì~ºvÉꇓçp™+ÒµÎí¾¤"ºJÈ»Q4ìAC²6T“½°¨ Zc©â\Kâ-Þˆ¡ÃÁ½:èÝæÃ¥W"Á  P+òdù:ëžÔ†ye‘ŒÚÖMƒÖ$!«'1T:዆o)ÅoÈó®åF ÞŒ¦1ª’*ªà=8‚äàý™" U)y~ä[ Úk²_f”œIÒÁ×{>C·DÝR™ Ü4vÖ;²yxý2»íœ9bê¡þò+IÖ%7Üg4äªÕ&‡ßâ p—«:’JÁ9A½´*Nn)ƒW-{Ô3Têvć9˱PYœØêëщ%Õ•â;rSV_›Cç¡îTYK‡ZovÅ#¿PL±­d|¥tT]”µé§ü¡…_Õ´ÙÖTøﲦ‹:¦jøg„›™0?©W8ã‰'Z§ãW±£–Œ­ րŊ(ûsòû-¶°Ÿru!váw ¶¨)rÛí$ÃÔ£'3ºRÝÎä"Ùã¤5Ú›|² ³Ñ.T$jE.4Õ«Ðñ‘…´ÜD) (²VK7?[¦/®ñY"u€èR1‰ÒkÀ¬>+o_¬÷Ûáa‚ëZ´¾æÌ= ;¾“¬–fl†!”kþêåŒ/ò¤A¯ ɪøì\ nìœ:ÅYç|RεÈ'ÙØRSÃÔóJ>oü‡¾v†lc埧 è‘íl­ë3òœ*´„â‰>áî¢n:/Ï=û²K¹Ó´~35ŸG%O’4ñD{ùü`ý³pWÙÂeÀœ&G-%¸U#¤¨0æ6|òm“âVÖÌÓ@škÓSÊ#}­?£T¼‘©ÑMýÍ,·zLrâ³ä)&ÏÉʘ*\Àc„t`‚óž)ÙûSÈÃBå^v‚IÇûˆ>2©ejÑÏܦ¶Ž´˜Oݯ“»Ì²?‹Ð)·iÕCràxž&ø£p‡˜Ìl1'%¦©ùáU&LB$êz²Z#äh;¯õ'ob—èö­¾üð÷º‘°úÎ5Š9»Ø¼",ƒCÓzûKùs¡|:<ãP¸—Sô vVK¦ ~M†]<’Ý_LFmÁ7öø>ëÉ6¬Ê³E!3ÎtÆÔ­Ô¥†íì›ÿ¢Ï^/XõfqÌ(sR§ÂàqýïPù‹Òn+™¨¸°ÆB7ÉŸÄj®9©±-»xk6úklëeˆÆGí1×¾öÂ0j”sBxèëP%.ýlº°ð%[‚ =JzUT{Ä«ó"ý +i©Úäj’RqõP6¶}dFƒ|Äía:ϯÌî¤ÊÜZ"?¯ ¿a˜ ñ»™d’B5é§â¿½¸jï&|ZÖUåÞM©‹r,„–î÷ek¤ŒÇÞƒ"¨üY·ÄFÜ¥¼6ê=¤~_îjUÞˆw®°_ÊBEñ¨NZ¦€õÜËà©RfÖòÛ!–ÿ§A“œÀeÙz°NFù|:·=²GL¤rºÃwUÔ–œ)=eûH4*¶‡§LO¶Û$¤êóY4H#¤{;ØÒ×Ï¥*×êÈ9MdúÈ„ÇZþ ÃO1Qú=ø¹øR=ù©+Ìé|jûk±±°ü6’[s‚:yözú/œ0ìDÛ§1“†t¦ _ºë±ó)o!Ñ¡ýaÁ¨®Üã©ú”ýF ΰ‡Äbfã$}-GP0 ƒ̤ìÓ<9­ÈÐg/ÝÛ À¥jÔŽé Z—…n®8Vˆ™h¨DuÁ„4{4*ĹÖÛvÂÏÞØvYüYýµrMIƒ!ª’<§6ËH5SktÔ9“:g­:ušº™ÓWå5kÁñdÿ«ƒ·>™þî0·BË”6n¹WðÃZ­ÇŸR8éÓtÖÛ½¾9ðl*-0ê7Ë.8eé7>Ãj{Wá|ZÐBUW=Ji—1ïùò-Øv0©ˆ( ^Vž8_ý&kÊž„‹¥Î󬩺9ðŽV´°«ž_OM;C£Å Ô{›.³Ü¿½ió4¤H@‹Æá°–~Ñ)v•]ÞÛÌ»š7JÈ.»f¨RŠeEuùôÑaôñÀP ÙȆ‰HEË÷Ù-”¢9Ò:fGß›VÀóõ‘Ô-ÆÚo_Aî;Æ•Ázú!ØrV9S;¬Îlºô[’Fìr"&:Ààvlô·F»£ëY¿àÚǵ˜ëwó¿<¾º”’ÊÕ&ZOë¸|­¦ñc|E…„Z/Iˆ‚ _*¶àÕç$)Qµ;Ó­6F°rEΔÕÇwû-Ìea´ý’íªƒÂ––õ~S½4²ÈKº•¿†)žSùët°_ƒ§³×ÆoQM¹Ü©'é‹a'K‹[»w†¾Æ±%S™åÞŠ¤J—_ó7÷ñ›¶ê ·ŽjN0‰„°D lXEDÅn³Ì%F"”xÅ“BYžÄß‘Øóî`~€Ø”\áž‚):4íì'ë´±1ã³$ycHæ‚M˜Nàô¼€D²£Z™B†—Á±M¹Ì€‚Щ{NÚÀ®çéΧ…Ìcóž‚àwJÈÎMV}>ØôuÜ–án¤Iì_ì›ô‘2ÜHsUôÂC(wá~ê¶^‡“€nÛS˜1ëjwsôjDæš—(aLÈŸ2ɧ%‰`Á¥§|â‹àlDJjœj0[ë2‡×*MàØÁqÁ¦³NÝÌ&‰‚@ZŠÅ‹`ñNL©©Ï·Ÿu" =x…Ò|`N;…¥7BŽ«°Q½Â¿î\´B†Ã9mÜÆ1h+ÝÀþŒPÌîAš.¶U…œ?³Ð릵î€3´HH .ØÈž$å “"1–¡ÊÚ!´ð›-êfåÖ'ÃÀGWg¾aÎÞgÐþVät(7~ÓžF?IÒ\À‡£eìKþ=Öò`cþ¹· 3ó<ØÀ@ôÊ Av’cê6¨a¦3+É_Óp²¾ã½}iÏírìØ)Ø.Öø’•Ï¶0™n%ºdÓ„^ÞÚµ'³*%Qb¡æ®z ç—¹_š¨^g•ñäXü|ä„/jY Ð{ž@3ÀgPÍä6giv~0"9÷À·_R v„–QéŽyØÖ[gëiÔËV¾ªâ<ŸOziFÜîÿ:[X#ì$¹B‚—z§Ûœ,p Þ»'rNát¾.OEUâ8·¤÷}Utø.¼5°¬ô*:¿˜?N&B×Û>h(šœôn€0|·öž~@Y\k@1À9,æ–­§<Àµ„Ó ¯ÔøÍn^)JíU—È)ŒT«~¯X‘!x=^d³­¿=éÊ|1¶e:…ãGàÛXÞöõ+8áRti´QÚzõœQC6§©5]ÇáK€ãÊDro™ŒfǸ' kP´ÏéeKÂŒSSìVÒ×AbŠÌ‘š¾ëVâa×ǸàZ«MØîB»cÂC;×áÛb¼»¨Jíϼ'ÉF±âÑ4QùjW„•(Nè‹aKïØ'^? ²§ë 4qþŸbù*¶¤MÜŒ;ˆ«bhü˜¿ã-|mË™Uc.ým@4#¢º,ÏÔ,éÿ™4λ³Q9¤5¶ (æTäëáúȃ¦^Û¦— ¨ßv’o&1°–‡ôÌ%+øÝ ›ÃÌÖ²J,²mïøŸ&° 'AšÐq4‹ãQ_ú…úÀ µ=À@êRbÔŸ‰8÷…f‰Þ'ªd% AÖÿeNÎ(i_aî‡xqdé©=aU n~4m6‡l«ð²)ÖÉ+û3Z HV¹"9Y‰•#™ÒìÌ·rVœë„²î.™ƒõT)‘-ØV·œä0§ÃB/˜%*nvêæ®øy#öS)hÑ’ V.¯›yÅ#¨%ÅÊÅÐÙáa[4±cœÆÎʱnI>ûþ2D¾Ì $öÈÍ‚Š‹wñœâXW¾lX¯Ç'j´´K?ŠI0C-ò‚¿ÍÚó üù’½³6ÚJˆU!…FæMç—{áÁæì~'ý"çuë™ãÈ£QǤxÎ$†Â‡7FÜ %~Soï2Òc³B0ÈÝÓX5ɦ]/Üé¸úŠÚ0’¡ä±Yö‹UªÄ´ÖÐÆàvJN´½@†V¬Þ7вÆr^äoÖ’ç¡O3˜ÌKo«ïMà• ¿© [V R·¤üè¤ÃÃ2l)“ò8WfÙ}ÏÁ‡3”à$ªzþªfQg¦‡¤`º‚éiI,Á³¬L‘­ÆF…¦FT~òaê¼µ4þõËUí8“JÇâ˜QdIÞˆ©ö¢\HzoYnëgÁSò•?9§ÎYš·ÏCó®îà,:)éªÑ»IP²“9ÊBfÑÙæ 0ìi6kâÁ7c¬ú7ñÈy€ÌRp»ðSU½¤dp£@F£ä™`4ýÌ*iBsü´¥5Ýóx`¨É°ps–Äì€ÿD÷Yäü•Â›¢öW6kYÃb# qÔ==,š–ëãÛi!…Ú|~ƒt¤¸hxÝIøwÁwþÆEºé…^pNºò8´çˆÁÆ%æ6~·­¹Ÿáœk„dØ‹óȃ¿p7 ²L-úªÂâ±uu•ô[•¼J¨.•Vr¦d™v; "³¹¢húH® Ä+¯½Ž¥1f+þ¹/Ä¡4«/$‹5·`J ·ñ‹‹„š;d¢´ ]¿Ðî™ È;’/nÁ0zÄÒ˜¾+‡à* }NpSóMàÖb¤µTÛí˜.iÖø¢ºt˜, &kcÎÒ=$? ºî(Ž¾GvXÛèæ#Ç& ÙkÒF,œ)?={2¥'–ýqÍñ¿6ñ‰tˆ‹^>Aº“ºsk©z::œ7J…J}6X eë«MhºÔêî ,Ôk·¤6u§ùk,…\fl;¢¯e;`LòMÏŠØ9µµTmë*ýe>Vçß!‹a=31æHÒ3×’¦M¹YÈkÓ!¥JöÏêÞrçy\„¬½Ö†%¨#yívØ›<=÷ý.úå;l·ªivˆ_þeËÆõÍæX} Ä°¡à%R î]œñÉøç7å*ÍJú~ü9‘ Þ#l¼ÖÍðùRLÉT»h&º–»Äô`7nþ6pà|­ùh·Xš@¬gü&²æ›GBCáÑÀ¥Å^¨aÿrƒoG¯gÚpR±F%S>¸qåxeÎ`Ñ>‚_341¤LPNm˜ŽØèçÜ’“zjÒéÎ÷£|Cªq-tÈl¡´zRX~“Kólm¬Gý¦ Ña6´®¶·/0F€§ãÍîm\ö1´ ¦*?õ<ÞIü9Ç’AÉBkÜå!Ô'R[ÿÉ21Z̹¢¿+¡w’€\þªÞšl#‰>(¢Qh6¦ÓKô©Ï–”èÓúe±P;¢w7¹ªÙq証µt(5gà*ýôW[œÁéç²VdM¤ÆOù[â² ?L­]}ÛÂú.u·ò;ÈÝphc»ƒÒ‹TL¦çÔ—ð Ü·^Úú¾ #yföY²¦ñ¬ðÛÇ{©¤˜©Èï*·]‚Øh ›÷y8Åcù0 c¤î¹Rn}‘ùã •ìòN¨Þ™ämŒÁ`ÐÜ›óãcŽŠ©½Îx~"Q?#ÈÞÆàd¦šWêI/Š3 úê<`äø*ˆuU®uk†9‰%?Ά C‰ŠTVEï˜q R:(¤ endstream endobj 2140 0 obj << /Length1 1584 /Length2 8649 /Length3 0 /Length 9684 /Filter /FlateDecode >> stream xÚ´T”í6,Ò¥H7C‡Ô 1RÒÝ€à 1C ÝÝ!JJ HJ·tw Ò Ý ”ȇ¾ï9ç=çÿ×ú¾5k=s_;ï½÷µofz m.I „TGr¹y…Òª:ŠB^^>n^^3³ iý[ŒÃ¬ur†!àÂÿ0v‚B÷2òÞN(¹Ø€|  0,ÌË ñò ýËá$ ¸Â,ªÜ%êŒÃ,pðp‚YY#ïÓüë`3g…„ÀœÜ’öP'˜9P… ­¡ö÷Í!vm„9 Šôø¯l¢ÖH¤ƒ0››7ÄÞ™ádõœàCZ´ ÎP'W¨àwÁ5ˆ=ô¯Ê¸q˜:Ö0ç¿äÚK¤Ä ¸ØÁÌ¡pç{¸Ô pŸ ­¨Pw€Âÿ2VùË€ðwo@nà¿Ãýíý; þÇbnŽ°w€À=`p+€%Ì P—SáFº#9¸ÅoCˆ3âÞâ ÙAÌî þÜ“Ô@î ü»^€å}P˜%ôþÇËâ  \ >^ÿTü7Â0s$À jƒãü'ú½jù¾¾Ì`Ä{Ï= €÷÷ïߧ—÷ô²@Àí<þcþg¾<ò²ÊúŠOÿªøß:))„;À‹‹OÀàüÏàûƒÏ‡ùwþUü©ö÷åxÿQn‰ýUÃ}óþU‡ëß´`û{eØÿA qÏe(€í?Ô7æà5¿ÿÿŸàËÿïGù¿Qÿ/$çbg÷GÍöGÿÿQCìavÜSÙy¿ªˆûå€ÿ¯©>ô¯UV…ZÀ\ìÿW«ˆ„ܯ‡$Üêžâ\@~n^þ¿ä0g9˜;ÔB†4·þ‹HÿÅ};ªp†ý~qî½xyÿGw¿uæ¶÷¯ŠóýÀþ¨ ÷Kõßyeáæ‹ßÛ@œœ 8÷ó¿G/àýšZ@ÝÿðÀà G ï]÷5ú,N8¿Ç ÞSê·èx ÿF¿ïÃcñxo ýðÀþŸxìþ…<öÿÀûPðÀûPˆ@€ÇáÀãôxŸÈùPÀƒü¼ÏëúþWcÌ]œœîŸ?ü½ïÚ¿ðŸ7 u‡šãÌÍ ÌE‚mª‚›/+$©Ü¸¾‚øñ{O#ϱô7d)\·4¥E¯Îør£f%Û?oF]­î ïêm HymM–½S£„IçjšÑ-ÍÖ9½Ç Å®÷œU\@›Êmm†¯øàV¤G ZýnçSìdœŽ'ߦÓwäEYÊ.J­0Yëö¹ñŽÆDa’?{Ƈ÷0¸E2½¥LªRrês«"“ó©ÄèÁ¹«Ôž!ÀZï|1' l5•jwßAù<}÷yÊv§ ¥ò«qÀóF§,ùM<[àysp›™ó°¶8£H†âÆø©ÂéFŒ¢–^p ëi£ð“`µùS6d‹ àóö(†Ž¸™–Ìæ8µ8öGï"~œXŽŽ.þ,‘wèÙJ ÞÇ©|úM3bêCR“½åß$i% q‘]Ú彜}Ÿ?¦»ó5ó߬†Ë€–š˜ V ­æœŠÄÒ¼ò‡‰hSŽ¿Ø¹Ð:¬£j”ëBô´7mŠ9êÄùw4dæ•~z+÷ÔÖ”¾«Ÿ%³ªðíÁi¢¥¬˜ó,Ô7–šz3)n?àJMG ‘wR»bx­NVìT õÌ“×c¢î…1óÏS+¦(ÞEPq%—´4MRy‘ ©va}ç¨ä_¸6-Ï{¬®•¡«„W1r«þô’áv%4eòõûWõ] 2¾]šÃn ¥²²¯î?íP€EŸÓèLú|Ê®v¿ Hö+?ž•Ÿ{i<>Y6¸èÆýkS‚ó—âülƳáéT﹂YœprôX´£IçIµ1Ë(Þt~Éç ý䩲“™Ø‡Õ5’Ø|ßùÆF;놔/lE]0m6÷Jhž}æÛHPØ.9 µ%¹‹yi¥º®ïÓg‰F ì&²pvR<¥îñtÿÁ8°fú «Ý|ižÂüÝh‚Oþ:P7ô®çx¯ègôGÔ'¸}¾õs«¤åãâ’Çâü»ÌE‘­ÙDÖ„2£yK‚èJK¿† ÛËm x/1kdª$a1©D ”_Òpû°TGSÑcÏW©Yh$Çíã¯öíc§óÔN³9|B¿¥ðzDo½k´x58q(~Þ ¼îêšW*îBCx-]@âáÒªæoÄGÓç*s\*ŸÊHñÄÒ•·j©¦)=¡ÝX›ã DHV'ïÜÙc—,ž3Iì^Ï·|Å{(ÍPhZÃô²Á@6¸2Èm‘˦všzRõÞfYÝV0ÝZ*àclOøçŸgxýe¢ÎR à˜aºòTãŸQ[rë±Æã ´¡Ÿï\úuÑö½øo‡L”Œê’ü`âWõ‰õ„A[Ñ«]Gάlyèw”ZòæÍk$ëZOØNß뎋w&³²,øUï'?éz©Mú}U•÷ö-(8†ÙûG‡ôΖáÃÑ0q"VAfv³k0ûÄ;ðÓyH²#ت4׃ò±¤6ûú-8€ªUc©ˆ¹¦²õ¶§s5L£Ør„Uó%¸¸Øk Å£²jS®Ô?›¬E8ë¯åÌ:ƒºß&6ÞZ‰F©d©?EC}Y1+’ÜøIy-lp _àN™\™«•„T’ÆêWµwvö©ˆyK“0ÎϺ®ë0Rðj%VH˲í]¢Aó§dë7õŠQF‘^©«ëgK“·‰]¬ÊåLK ¬ïwGCçżåDñãRŽ0Œ˜3I‚ÑðJêÊé‡Þ[ÚmGÖMP÷¾5x禜1ùVœà³ÕNrïÈâ‹Äüj¿· á ±9e#ü©*¶9£ÅÀl¬ê®àØŸÇ©…³eº›ý|%Ñb20v{ŸmÖäú¨2鎭}õg Öp(𠦬4í× d܆€ÆÇE¥²\ºT@ xêSÐg•%ŸÛò‚-˜DdÜŒmŽ’Vìª6ž¥Ô×®I<â|¨ºV\Ž²~_fºh;ºÎ8„׆Ùý]ÝGT'CÐnw繬Mgùx|ßÒºü³›ѺÆxPÕrnÉJ5‰‹Ï/%{e aóéÐsx·/#þÎpwäŠrï¯Ú˜/¬¯ä¨ÞEqjÔÊõ[ðwpØó“á˺=/õ½îWd“ßìfmt{žûîÄšN±dgÜÝ'+žòDÝ“EÃÒ{°xº^õêU¹Í¯aý‹ýö°- Њ_/?çëwÓzbGxjW.¹ LšI@*’Œº¬AA4ý“¶J5..i–ç…ëª$¸[¹Ãt 5c< òù¬O…_äMµì¥û[¾¡ ò½óo2óbÖ\?Ç›_X¼l<Ä||ÚÇX¢cö±ÔÈSMvƒ^£kœ6Ö€ªi€¤8(߃8‡{Çä)½žö°±³$õ9\Ö›r,xíȉÕßÝðú uËÔ £—`ãŽÅ?“:‚Ó*îCïÛ¥~ÆÜ­Êy–ak±F=ÞŽM¡³7§©í²™ïnEwÀ+Â;¡Ù‘¥d’¨Wæmœ‡ÚÓá:‘siº7§‹Š‡ZØLúÆ+¨£iê—ùfE¦ãíÈçÔv8ú‘Þí#®g=„"NÎmŠ½Ýž’ªéoÁ7?({ZTŠkYbbctmeÀW™|2+'Ayñ «îzïÚ/|pI=qûj¿~‰+…uVc^ƒ3ÕP±»+üÇ`û‰Jc…9yq×’þ¸Xá*¿€¶LEåÍw¢’$ \ž¬Ã[¬vî™ùì¹± (c¯G’=Zý Ž¹À$hçµy3¿žæ›äÌ7>¬è-üríµ¥Á$fŸÖ©#prŠôñ«>Ž‹êÿzèî>Hf_Sô "q¬è02d Õ”»2x¦Ü(éMcóšÉò ¢uùÚŽßjôVØùÂ¥ÂV¦Cƒ `Š›k?ùTà9Ž­hSMXÊÇü3-Æ`lä¯}?àǷϻۜšâÊ]9ÅnÞñ[ÚelyVölpc᩺¿Œ ] â%ëýÞÀë£]D1½¶JéÉ|[!iT~HÎp08|ê× bˆNŒB4¾ ÅÐY÷ŠÍnƒm`'œîJ¶šqè1 [¤{Æ¿kÌÅÆÉeÒÌ8JXe(ÔmFáû£áå-¦ábÝGžÍåßôúê2Ý3õG©Sû ½ ~…{Y WZú½Æ $t±/Ò2Zªba“DP"Ç®$@Eu¦„M5s¬u³·vZô ¯§çµñ”Sçâ”m“r#ˆ"-ÏZ¦öNkBÑEÔ¹QÜXùÜÕÏO¨Æ†ºZ\Ÿr¦¾ $¸¦}/óËt ¤/sèb³¥WüJ¾Šò1h¨ø–a$ÁÕx…vUÒ–‡b$ª^èfS¨õ™Ù£IrC"ýxl²´v©$›HÖbâA»ëâ>óòSÕÇR§Ù¦E6&¯ ¶ v{a¤ãÙ7®`¾šP}ð41SR˜[O; Šö"ØKÜ‚Ÿ—Ïgk¿pxdó6”~÷õðf˜qjñIðáàæI÷ÒcVë@‡KuÛù×z·íÅ9RÓ@òuSÜDDÀçɯŽÅx¹®)úÁè?}çÒL D¸×"OƒÔÆcÉÞ¼t¨¦¿+Xº0.BÅêrC!r[’[ä@ùñþ¼§CTí~ÈM ‡6‡Lë kæ½n™õ°—d*¬òU¨nŠË?24D9sö®TÐõiK+Œ³ðU %ϵ2ç[åMæ¸É·úÝÔÎ]bÐxºª6U‚s–×G4zŒˆç$ã{`GZó%ªW‰>õ‹– ë²ë¹¢0…{=-ôò_±Ôšëv\ršd ÉdëGS”É ¹CÞ%á0 ÎÊ+I4™u‘5%‚{ŠÄ>æçªî)ü è×Vø敼G#âë‚Ýy”¬Y`x>’ø°ìŽ/[6ØŽmà9Õ4ré¨Ûº!xÂ톰[)VrÇ[ßãx)Éc?þrèbåNTç a¸NˆÀ„Fœu.Ç‘%=æÎ"´¡B¥îY.üFÝèÄ÷£ ½Ï…”4Y­~ —æ{Ðpu1KRB]ÝøÞÙídêTïY§ –ÄmÅgž)“V ½K …ç9ŒÉá)OØ€¥ïw(;›èÛé"C¾é]ŠV¯›J'.‘üý¶•­…n:v )’Ú_êc6iEIù33¬ê)m(=³l1À>Àî©j OæcoÓF/àS ·1½ ‰ç›@ðãzéžðÄ4±+O=M}—ßõœÒ’8p´ðkúÒi²Çœ¯Iª+F&¶í«Ïø¿¼[èBʤ””y©1žÙL¾ =‹ßãz½‚³Ç?Äü€Ûñ£æãgeî•Ç{r”šPÀ|Í©~šºÝ¦¼Ð\B(™fÿ4t7¦êÿuÚ‹›ã’g­ é¨ö²D““ýU9?K‘$”Ë7Õ[B2bWVEÌàâf2ÏΦ@¯òWŠÀ× ~æÔ+4Eö’ Õ@D× {Ù –ÎêÚòŸLNj‘X'"ã3çœúàMª.S’Nê4Ç’lÙkÖèPb˜UŠ »ºïm u•å²%ÆŒù”ì¯U¥æGÞèþÒ´ˆþ¶;[NYvDW(ŠÂð,Ð}ÛVy\BòåíÑý.ÕÏ™þˆœ@ŒD‚´ùðdk+Xá>4¸ÆaåíY~˜ø<ù oªf±"+½6ûò¬ã°ÿãDªtÚkþµNTï‹×†RÃZ…v*1Í]Ò òÕþ¸Êù€z 7“4)}Í[ 9¥5æô(#½™Ó·V:«ÛŽ™»)'z‡L+“•“ý0ÌÕú43r1¹e‘Zãܱ)®¯>Ç"¾¨{&w!^ý9‚âÿë7[?Žµ¯ô UƒÐ`-"e.”|XÚ‹Z“ªr( HKÇùgõã"û(BoW®Ú«#â³ëøC ¥ÝÒ€Ç×Z—G ¥°(½5˜Ã6 èpÅ¡Ì;H(B÷¨‰·äê桨'>ö^É°[ê¾"z˜¢´µŸ2ÖˆIj·å‰V€Nør7ƒ»7£‚ö…GŒZ.Äl ꀄxk=lðÙyÒ{n³×uÆñuMÜô]v2¨',MYÒ“>¿Ö븘G:(©–ËF=§U¬3(ñA+M~¯*\¿Û;¹9»ñ=§tý¤Cß~6™~8Ol\¿ønÒ½óÜèŒmƒ ªVÞ d³'Þh¯JÄ F›/ÖÇÛ§Û±îEöº–<•ñS¬6 Iµ?` Ýs‘нòt³…Ô‡$O©¼FÙO6þ)~øH‘Œóá|xÌŠ ©#d#HÂmóE]|A°¶©–GÓ¡Áñ†ùÓYÓ™»$4Àzíåû@¸ïÁ²Zï4;[†/`ßqÒ×ß~½_³¨íÓú.Rdƒ8×ÈknƒÜÑVÅ.Dâ`FRY‚ŸóÔù'ú“7‚Kï.Ÿ0‚OÄbTZ1š ¾hŸUvú±È­xR3Ü¥ÂèÖK4J~ëî˜F®Ì ýZƒg0AÄ¥:øRðý¨+Vݨ˜üñÒ.Ç ?»©¤ÍW"›Ä•B²”ãœöÔpî·Ûž·º¡ÙYÝÖ®yòë)Øgèf§\LíSD\–zASBEQô;ëßì1¾± ]º¦ÀNt‰·Úe–çÏ·3g¢ò\ìšæýÄÜf^1ú•4nêÔ­V<ùM¼b`Ó\0Pé˜íÇù´½¾Å}­¸jï¸Ót:ÀL.m˜#tþKEaק¹—®Ö ïð¡ô¸xnÜZÄÖÇñ&y£.1rV²É']aM$jø_‡›ýém¹¦2YWü÷«ã棑Îåöž-é4´Á•™¼Kž;A±.ÚXë–Ë“ÝsɸdZC5UüÄ¥ë¨Þ¹ù²Œô¤èôT`p/EîÑŠý Í~£Õmººã'w_–•|b{ÔÈ%£¼nšnžPØÓ´yMK|² -nêä…Ef“4wh§áÅ… nI®, ê=¯¨VÒmÀº{^†:>f­®ï➆àiôâÔ1x̵edOÖÄÎ$4䔚Hòn¢2uôB—O]’áê1»RÕ:A«è£Š:Â],¯°¸í¯ug…hl£}~†½fÏ áEŽÄOU%Lp>4ý¤é½–©6ÇGÞЪIf l7”[Pÿ4µ¥=œV–—¹=vùAÐß&/V^ñ+0ÇÀK ñxu¢l³ºº_ÝiÎ ˆ áeq8~”±Ùè Õ/à‘-lè鳉qúf¡¸Ÿ·ÜC šLQ‰FdÛ Sólèv ×ÅŒ`©´›s[©QX·‘®Îÿ³$)cÀ”b+éú×ô­á¼ƒ;GûE}‘ßØîÖó°‚Døvá{¯òäV\n|)‰x,¯ ^Z|ûA´|ª¤nà±™)5²N9ÑæTT$Z yŽÛ›ÂòãP’†¨ãpì“’uHТ(»˜ûäM)mñ<F¾=SnW7ÝÕG†|)ÊêGÀNz!âÊœ91µ‰}˜ÝÁàvlñó]ÕÞÚ/Œ¼Éý`h€c iîÇÓ­ Hûj¾h»xëÒŵSæ@¯¹¿´"ø¢’×mß‚ÞMjÍâ‘·ŠuÈÙã_Gߊ+2¿½±œ(µøî\~)M ³í’'5 [Ñ|¿\ŒéWÆ[cHH`ÑÓO¬¾ïÅœæщ÷ظP¾|=âUÇD8!¨ã<õÎPÐýí§ãÊ4‰lgPÔºS¬¡`Σ ”䟑¾Ë€ÅCå4±1k]¤ìÝ8Ñ‹á±±…Úº»ÆaÞßÎÆ^㶖Žù{[ôÓè{¯d P]?™uÒeðËS×Z>tO\˜ž‘|#X¿c˼÷èbXW0'Ö„M(ͯ‘ â„WcjÍÝeäs…¥¡§Ø>€“,£‰³Á€³ñ½´ÅÊ¥p-³(ăš(¥Yü)•sÅa:Â~gd=¡p¦Œ‰%(öídæKq;pšvÞ~fGÁo²4Ô•X %J|ß·ó·-¼å{´™ XO—ü< »ø8€}L+Z5îøÜå5}&ß]iˆáÅÖÑ·šržkÌ€ðî3¢ï¤u3¾"YCÅy,`Ъ±9‹J‰«|Ìx'Bl ·VŠñP2?ÊäU@îj…CÂ~Z7NNùgªbï#Å\P}Riú’°bI+Òƒ|é…ÆL% Œ3qÎh»€M ”³jÇšÁôns¡Èk¾85%=n"Ð*\ãÕBG7?1ˆÜPBÜúìSoZŽïÔÛ¶ˆ±úË•¾¤”ð>kØX» ùÀ«z§<÷¬Ýϼ]­Ú:@¨í•+mܤZRY^H¨({0wè8ZÖeQ(¢ ÌÂ^O¦p/RdlñúÁ“Æ<ŒòÛšÃþM¶äb«[yÊh®¯+ÓhVm!pk«%âV×Nœüíf5Ý°ª"¤|mø ‘Cjç‘r)Žm¶n}˜RM;G§Õ}ÜDÀS¨îMsîI@»Î•ãvß(ko ¶Ts€wÌc¹ä(Yr†€áŠ‚E̼¡È6õ‰$;1³Ê;¿ºÍô‘4ãÔø€Drac‚äÌBöW щݚ]±óv±ìŒ-Ì|¢lé/ ÎKGbœ˜#lhäï–›¤…S.ÔwÌZ1)k£Ç—æ|tâ­ðìOŸ®µ&˜Åhé`ZçØ’Ž^ɉ¾ÿYHðFÿQnC軣y!›ÙI oÿ”îaùH¸òlçYÇ*:¢ÀLÉÔ¥j^“à4S;wQY »“Üa™ arÁéÆã›6"¢¿fžÇ·/÷Þx-àV®\ïż µnKɨìY£qÙñ¬•s>¢¾¢fw,y#©«lõôå#B$8L5c1è°–0eßgX~ŽÉ¸Ö"\&¹„^ÐeW÷ã>X«ÈobZ¼IB,å1»ñ2³ÓýhÞ¡OÞÜж»³Š·V,'T’UY?˜å )ŒA}Ï0fô„‚šŸoÓ©’kW“¤dwÚßã9û¥5¦Å,ê.™È†ÃÖ±g&•ùw¢ñÁO0¼ª¢lb®†ë.÷,øʃ»¯‚Q¸v_žë,ÂcƒmïJT3ÆÔåq¸kG±ûƒCdÞ½­ÛýÐm•%^ VxDTd¸± ‡KØź“çcÜ#« `êùuEX[ø²+'ª’þN8ÈDÞ‰4Lv§µñ§‘–ýU'¿É­˜w3+à—_k`¸7³"}ú“ÚÕô ®Rà;A—#2†ËŸÝ«Æ$l­„‡˜â‰ ±^Œ˜vSÃXö ç–œU/§f~ùªÜºë?4"u>Ànbƒ÷Kl‰ËÍäN|$vUH=ÎÇ’™nÌij– Y² y_«Kt ØgÁ·ºê Oz=ŸˆìpTxÑ7ŒVb,q“¦½%àú°%ùaƒ …iõ5¾Û áV½ž»cëjwBÖÍð’?ʱþ¹¹¬½²#³ÌÓ‘òØí®ÿgPåÓ endstream endobj 2142 0 obj << /Length1 2785 /Length2 19656 /Length3 0 /Length 21232 /Filter /FlateDecode >> stream xÚŒ÷PÚÒ c‚»Ëww îÜ}pw·àîNpwî.ÁÝ .A‚xsÎ=÷$÷ûÿª÷Š*˜Õ¶º{wï=PT¦2±3ŠÛÙ:Ó330ñDäTT˜™LL¬ LL,pää*ÎÖÀÿÊáÈÕ€ŽNv¶<Xˆ8 A2QCg¡œ-@ÚÅÀÌ `æàaæäab°01qÿ×ÐΑ jèjacHÛÙàÈEìì=-ÌÌA<ÿý 2¦0sssÒýí²:ZÚä Í6 FcCk€²±ÐÙãBPñ™;;Ûó02º¹¹1Ú81Ø9š}¤¦¸Y8›”€N@GW  ௒ò†6ÀJc€#¨˜[8ýG¡lgêìfè€ÖÆ@['‹‹­ Ðb(KÉì¶ÿ1–ýàŸæ˜˜ÿ ÷÷_,lÿv646¶³±7´õ°°5˜ZX â² ÎîÎtC[“¿ ­ì@þ†®†Ö†F ƒ¿S7ˆ }‚*ü§>'cG {g'' ë¿jdü+ ¨Íb¶&"v66@[g'¸¿òµpƒúîÁøÏáZÙÚ¹Ùzý™ZØš˜þU†‰‹=£ª­…ƒ PJôî·Ì è `gbbâbe@wcsÆ¿T<ì+™ÿƒjðñ²·³˜‚ÊúX˜A༜ ]gG ×ŸŠÿEpÌÌ cg€ÐÌÂîwthú :G w€6hü˜LýüûI4a&v¶Ö¿Íÿ>bFIUyE1ÚJþW),lçð¢gг°²ØY9ÜŸÿòoýÿ­ýo©¢¡Å?¹ýOÊÖÔÀýŸ@½ûo®ÿÌÕ?KC ø_y;Ð4T¿‡_‡‰Éô‹ùÿó üíòÿoòÿŠòÿ:üÿ7#qkë¿õTÿ1øÿÑÚXX{ücšfgÐfÈÙöÃöÿšªÿ³Îr@ ›ÿ«•r6mˆ­™õ¿´p·pš(Z8›ÿ=1ÿ=Ptk [ ¢“Å_÷ €ž™‰éÿè@+glºSœ@‡õ· Ú¨ÿe³5¶3ùkõXØ9†ŽŽ†pL ùbagx1ƒvÔèþ÷hlíœA.Pu>S;G¸¿Ž”ƒÀ(ô—è?ˆÀ(üqE~#.£èoÄ `ûq2Å#f£ÄoÄš×߈À(õØe~#»ìob—û@ìò¿ˆ]á_ÄbWü@|J¿ˆOù7b0ªüF vÕßÄ®öØÕ#»Æob×üqƒØµ~#Ÿá¿ˆäghcÚ—¿®ÂÿJÙ@:FÌÂÉêwCA©:ÿŽò4ú@F†ÆVNÖ†NæÿJ™YØþ;þ!Ukähh ´š:ÿ!fÿGüŸmü7*óÄV@çÿ±çfýWþ@…ÿ‹ØA)ÛYƒ&÷ßÂØþ’ØØünÀ_#ÍhòQGµø? é\@÷À¿-%š_kC›?¢€Úaú; ÈÂÔÂõ°©í\þè ÈÄì7 Hoö×ã üÓ”ûïÖ²:fîao´ýÃ$³ø‚’·ü‚ÞêjÎï"8@]°þkëA­ü£"Ð-Îø›ŠÄdëbcô×=köG ç„ÑîwŽ  /¨AuÙÿVƒBÚƒÞ|Ûÿ9[6æ¤ÿ{²l b@cja÷Ç”‚zfoíòGΠï@Œ¿“5ÌÁÅÎhbôÇiqÿ#üßðÌÌ ã?šÍ *ðw`v“ÐÆ⇉ý/ ë'À âz+ÿÍÔòÿ³Ì :~Ó‚ÞFgsGà£jŠ³›Ý .¿Äù÷×"'c;Ç?; :M×? (a·?Ö Ôýbõø‚NÅówΠHž@Çÿdð?7¹±‹#èXœÿ~lA×üñßߨ€@w 1ÜÊ¢1oe}PÇC­¾ýþÿù¾z5½×Šc§ËL2uMVà–ãPòhÊúžÕ­à*ñ/¯ÓÖF˜Ð¶ÄOíÏÞ/úñJ3ûípËÓXCß O… ßЫxÿrðV °‚lï–&ÏupáBRÌGppo,_›YÜÿtPÃ!ÿR>K­¥P2Ožg”½€CíLOKƒvéŽ<{7‡–óíX:žÎç,šµÈKk›%æqÁs£R…Å©— W ‡òmb†ÂKø(E{É«´8j‹é 5ØÖÄ ±÷x÷‡smY)4l&kÇn‰ 2ŸöÆ~„÷YûÊö4õö§HK8´¦nnâù _‹ôFBß߇9ҚȹÆÂÈ{xì¬@ÝjžÞÊœž×ÓöÓ6Ý}·‹¹Þ"¼!ú+ÔÁCšàçWw¡efYW¿å·3û9÷wåΡ ± $8˜ÂätdôÂŒfNòÑøÛ:t6÷5ÜíF`ïuz©µGqWÅU×®^ÙÎó&´Ò¸dž=ÝÜš<¬“¾K±¡µ[ÇŸkaVB ¥GÐ%ræ.Ž–ËÔñ< J¯™T δ‹ŸOÊ…Þ÷ê—ç=› ’šÑÖ?ò·}<€š×Ê5ôûÚ’ªN!‹!ɮɸºÇ§U?TxT˜:˜Ž…/6Y¤=>E‘-ßÇv´p‡õ:ÇÜÞÊ´%L^»*‹KøetìºF¼OÀL)]¦¬A7|2ŽOyN1ðkÏrleÿmù®öAâŠn-©oÚ„Ö\¯Ö²Ç˜ûHMT}hÝp€:…¨vüÜ1X&ZÓ~Û/ 7A{°¢Ï£OWüVwÿŒáí¢%ƒþ¯©èÜÕqI⦠ۃh½“‡!·ØÑ9¹–9w›¯ XšÕ©ÏÖ÷Ñ+ãìM½˜D{¼9+sŽ³yuOkºÌ½¿M3<”( AmŽÖÊÌCÚdYßé­S¨¯ûcGn {µä_þ²ÀÛøQA˜Ý˜Ú¿º>Ê'.ã+…RÄ)ˆ¥Â)´.÷¬ ˆÁÀ³%9Fk¢Q÷…ק>áá Æ堦ƭؒKqGûp»ýZœ°TGñN^„2ë¥,—ÓþSg4“1¸dúüÛ¥‹ö¢›fgJÆ—ýl–mçéÙѱ”Õ‚¨Æ,óŸé_²v^iV!ÈüáL†Š "ÈŠÁÁ™á<($¹WeÔ1eÞª3?†ÊN£L±¡+ÐE› PæMRûJrù+3³Ø'o:ühñ› /,9{ÉO-ßR±vÑGP©qÁgÂÆUÁ¦ÂðTקÅëîãZ™“œÅ¾ζlä "Îäûc3ãöÔ‹ë) ¤Ðg'g_÷;`Áèe}xŽ‚ a6s»?déÉcc,ÎÍùÊÚ-X±«SËé¹fèy2ÃLpLÞ°~y)æ¨IÙ6R~Â&nìÇokÙŲ۹½hŒqçÆŒýò0qÄêo”#¥šÑëñµž'fâ;¢éb^ÙúIdâþÉUñs©©Ùy·}Þ3ŸÅ±¥WYìBý8’[Å«ò20“’Kôíj|·åÛTÏѽx•¿ðB©Ý\t[ü9ºÔçÞÔ ­âhøÓ•7Ê个$p¯ñ´ð~ÃLõ‘€ç‡o¼ìY…Š6Bó#´¹jKcžc ™ÆXÄ¡8¬z*Hp9Š®pxµµ{½'•ñº÷JJ™å"E"Pz5쿸I ïÐ Ÿ’®SŸ&Ôó”P©SÐA†‚½ J”#àÒ;&L¨ä…$/`iZøC! ÓYeC–R øMUˆ‘dº\µ¯hn?•üøˆ:£g@ÖH•\5G m“þD/C‰,³;‡Hù15Kl[QžŽGîÈ+B™U‚E IC'g[€þrAK#šÒíèÖO‹ÁÁ<­ÿݽ&~Îõ³u ay©ÛÄŠ|µªýd‰AUÕÏÝ 9î:èïúÒf³FnvS&:œÆTf“ÀªD{ GyÔ>èúY å¹µ'¼–AÁDQî÷l›ëA÷!¨¨ÄJUØJ†­¿7ç£vN&\ÇÜ÷ç mÜ-î¤ùß-m(K‰ìö¨ø¡*4™[î*ÂsU»g*KÏžÞ©À‘PœxQáÈ ò0ˆœYLÊpçw»KÒ£»îŠ½Ì³¹ë³…íÇæ®´½Dd ÐýüRb²kÍNi4n´á@—"yÇPH¬=âÜy//o{¯5 qýÆÁe’¨e¸nQ. ¢ëg¯Ùóñ¼ #$ŒÝ¹/Í@g ÚN[òX7Ãþa¼ Žù¢ÞÇÙ_ìŒ*凗Þ´“Ü¡áKÓ¸.Š7ÚuZ©Þìj Õó*s·kŽ¸Æ˲eÜ5¨›ÏVìü'C™šygD+n7!ãÞØ¥8Ù!®"5‘Š@Ï=q¨ºÍècñÏ»Ò0 z*^=‡ò†û Ĩ¹ „²œ hS§ìNVN’`««ý?{b|ð~õ«2¤©ž49T…ƒ‚¡fp–"0wˆ_z #¢a‚Ìu¸ðo«ó¯ùx$9¬ú~uˆx$v›yÀŸ[Á¹Y›Õ;ú’Ý‘ ò¦Õ~hùõy™ f+Y=ÀATv°³#ÆkýÜ #qÄÄ- ß{)QBúÍ“±%ëH’²K‹,(|,}Ynª+ºp;&vH e",¼–sË«FÖ~µÿ^šmçÈjlõsÐYºŠ]ወ­ÛÇ­7Ç‘Z¯vGì^U­7n±òœÅy‰ê—x6îÕÛv¸]“Œ…;¾ÇHî»)±ÊÁÞY¦ÜÄK•=W .#Âî¦Fb/ nê¼ÏÁévt=ﱋ„e­¬dš¼`¸MaÍÌ^èÀ­½+(:ðýgqäF i æÉî »]I×›?žqªžûMö7ÃT€Q¶@`„Ú~•GžNŸbŽÊ8,Vߪx…)]yS[G(K“w¦å~[EfãDô@¦šé{ÝÚB, @ìë‰ ©òÛE²SèÓOò®4Öz%ð.<ø,Ø¢å/D¼Ñõ¸-hXtOÙÇ7ˆ;üH2…™ þ6 £Ùüã"k;üX¥ØL¸Ñ }%š[õ—ø‰t š¶½nüê»1û)ó¨åÞÜ©ž„Ò«U.ÿo ‰žy‡›]²! þO‰Ø/ß®c0^ï¶ENX©l¸˜}÷Kú•¹xâCÉf<´/øXÓ­¯ µÕʉ›ùÑ+NÊÓƒÞÄqÇÑ?¯s‚ø Szkfñê!hi®ç¾Q9»l2ßÓ ·õÔ¨©hÎØJÑÖw(lŽ)¯®¯ÿ¡xÒOj®êC¢Ê篑JådstED¥gŒ¹GÂ`…g»biÐ\½ éÃ/rïÝTï³ÎØR—W貶-ŠC5 ¶Zi€Ëºš´Ñ?åâ iZ²}Ûð{<×Ûø6=TD¦CR5 BðsDuÛÆ·l\²C¦8ž}ŽqÅTŒš´PZ}3PõY#Ôò/ps¡ñÙ”ŽIb ± 맛N.wËìB¤ScÔ{)º1[½Õµ¨ƒN¦%eß „Nqš(ôŒ9¨G†Æ_ãÖ¬WaáqÂ&‡q%¢!‘㛀&Jó8^ڵ䖗Òy—6Äm¾°L¯“ñø]–Èh³—®ÆLÁ–’ sôe¤¼ªú6žØ|%/ák03 t/$kÌ÷ÆPk`}HP:WyÇùH’7eä¯2+˜Bò ?ÉÍ~Ê23qÁ\ƒù¾—ò¶tú]£^±2ß7ª7~O½ì=ðA@.ý8Ã)=x|bý÷&©´Ö*$^Dµú®Žj kXù!°àq¦Ð)‡9å柷ÌÓÝÛójåg·æTY®ÍD.0>T`ÚÃùhJÁÈh>†ÑuÍÝ ¥ýPRœ †Q´²à»WWßO,ŒCÜ>5sžRKPÊ°žˆ½Sñ¢ÑaRŸ$"^ÝÏ•ÏE¨xÇ””jûvþ*{Ù˜¤ÑX‡¸ºù(šd(½ÿÑ0ͱƒm)qA÷Ѥ«<µ7éе†£â¤-‹ˆ9 jªDÛb̯hñæT±¦ÏEøš1?mc韱Æ+VU#{2 ›%é]¥œØW[Ú ½é†é.¹|'l,Ù?¬íiB.·>Dº­ÇLߌܗ^)s a=F\ /z»?°‚ûà kª˜¡¤AA¯B×ù`©ä©¦M…†åT|ƒ0ÅÌœ7Ë)nŸ½Ô(Ö§uNbÎc¡ZDòÚ*}ø±güy©67¾¢ÔS‘æÓxYªò’Ò¾ ñ¡µü9ör¬ºåhh„[§‹þ` $Ók‘~·M??$ãіߧ«Ÿ .¦?wóá®(¥ÞËCå·ï Œ¢;X–_¥SÃÅœfÌÕÓ~iÎBž!œvÌßm/ã œNá^´q|çÝÏéXv#ñHÎz-íøÚÌ·IUa&¡×7ÐØà[ã®x ­ç/³»9‹Ý»­ù¤,:žHòð5‰ CÀ´…“(u5c˵5Acü»ÝQØ!ëKå.­J0ЛóÖ7Çý[¼+az,Œá7ñ#-ÉbÖ Ž[Þ—0!˜;°ÝÅÂøÀiò»´¿œ±–ws/Æ;2‘í"‹ˆ%àågøY’{°çúí®!^q†ÑØŽÁ´`u‰RBc…PÉt`7ò&ò=}Jà;fyîáÄ5ò ÈIÇ+’Õ˜Wà8|fD¡„0à!þ6Õ ÑØÝ_JÔŽyÅe%Nv+A—ü6!gàá#qRœ…=&@mœ°þnó&%ãsGþ—ðXgŒ5ø(Ã:„Ý-Ñ7¬¦—0W8™– eI&-¾rÂŽk9½/vb;‹?’ù4ÂL 2ˆz¾ÃÛ-þ´ï? umóÙH´ºF{˜1q,i$!b>.>שNUJ5#(;¡Ž K©´D-Ì»^£­šòˆß+µ[¹Z“yWYÜŸ19Ø_vÂ}r›q´Û“|úÖîêyu˜9øù¶rÅkP:<,3ºOk¨P£¬1t¿Ä$Bè£Ù™Äz?œÞî)·/"ò3?¼:é–i–Ä1ß)sJY¤ú‚¿oÖc áY{þy^ï}13Y-»lÒ¦*ªýÇW¬…$© 0æPÒ,4iî Št©XÈ`Ò:À·A³˜­fcšvaTpÚÒ*Ìuqu[“9ŸoJš.éFŠ-U®‡ÆúfrVìZvvß Js>Ïšæ†@úboKô}œYn: £ËAÉ­LÛ3³Õ…ãBo€LËY60 “g4°‡.A>œÚ=fa±€,}rOŨ= ¸çÍ+R¼¦kèȱ· q‹¤þ•¢C$2 Lë.üÓY¸¦Aé¡´¾pÍ6‘yɼÃÕ1:ÂÞQ†ëoÃøÝ4ݲr÷•Ã?3º“ŠådpS ½Œìd2¥lgTƒì¾â»-˜^³&ô.D{Ç+u9  ^VP>•!€+?-ÚI-ªé À¬éZ,ª˜DFkëoi5éŸY†T¼&±Ö”RÊs¹!Ë…‘ÃЊñ¸FÁEŒ‘A~+ÿQæ·“hXìŽÅûM!Æ Áá¾nj}ýãG‰‘Ó¶…ÃGº<-Û1ÂþÒLÖƒ¾(àôUQÙвÝs1f€ŸìLé±å-§éu½òé]M%ä®4±e9kq ) œ|B«ˆH¯ʘ+ ŸÖHƒ7÷|¨aÂΈÆÓ“Üx{Dó…ì­Kv±NÛZŽšöÄ6-C¾~°ÑI)™ó]Eó C³Ïèâð\¯ñÌK×÷ >¾L׳ûõ¯[Xž‹?Cùžà'Á®RF¦ 6À¸'¡7qšH²Œ›ÓÇÒÈ<Ž§ËóÔκê)8}‹³ó 6ö˜ ¿ Þú+ '´¨Ój‰ŒeF(|9QðÊÔpì¿`‹Ý8µØÓ½¯TÇ>•|OïÃ;?„µBß›IƒýUJ¤:Ö‹y¤5D("mKk°á;L5®£B<•,<žmÇ1FGl>·Åû–ªè/ùûzb"J7Ù>DDð}:´zK’äc0u×¹)#S×ÌÖÂG º…I˜êÏ(3ÍÚ¦Á瘉eŠ=î_)¤Ðp2ÒßN c-ø©ÇSsñiqO–3‘ºªÁìο% dé&~2¾™„!¬§ã͆"–ù‹f¹[nì.‡“ æÖÏ&3 ¡üCe1T7k«;Æ>Ÿ¶#àâ{ÔEx.î©m܆^ºSÛ(OŽ^ªœW?·Xâ` Æ Ä×̆ðž3ߪ:éiƒºÓÚ{ÖŸ>”€_ÝoòB¿fÜ?#œ›&]’¿Î¦ ?Ôºõ­ úŸ…ö7Ð-wZÜ‹jT﯌,Ÿh/ª¨cmÜÕjajœ³ø0¹CîúŽÀœ#Àâ qè°úýašƒ¨! €jºRcQèsÃãîÀN‚‹Þu4§³ŒÅ«ö°3¼¹¬•l6ë)FQÙ?–îVÚH^*Vš¾¬ _”( *T0ˆñ,çŠ\Š]sV[Sî®Ð „^ª{]ÒÇ‘ôHA!èx×´ÇéHU:jª“Ž¾vxO}éñþ¢´ o—¦~ì¾´½8s‘µ‹wW{´ZɵF¨vC+õÃù&*ÊuV‘vͱC¥ÎÿWó³þGaÓ Þëåµ(¤J¸£Ÿ9WvgÓøº” _›ÍIFŽ¥™¬Ò#äÞI€þ{Òo-¦jÂ;¼Þ›´ÃîM~Y¬íq«xû”j/ 2A—±Þ΄ >/ÙŸcª7N…ZÓß’s‹(@þiþ^àæ¼u Ôà·½c§£ ¾ìnýBkèý*·O·©ë×£ˆþÅk“G[%œý¾=q7:ÜAbl‡·pÉH9aùòý  SoXùcP£‡±¢Š×íàçR÷êu&oEd¹áOÌÂHd7?sœŸÌ¶?ÉP‡bÔŠˆÅ7ë¢CŽZ æ|q§[z ¼·7ƒï³5ÃÓ^ª¡ò  \D3ÓÐLo´ÝtK;-œç8Vc‘è 1ß2â6î««!;ÐÜŠ•W–±oá iÎÒD6œ›­%"1äüµe1‚el{TQ”~ÌÜrÉißÕB`€.6bc‘ H0\fZƒ}n³$ØÌ´°NÚ®f•Òüõ{8½Ù",­ †i,2›ôÞ8Ã{D¾8*²EÜŽ`Ìä\¾A©±Ýú¡º;9ôP6§ïãöco†äN2CŒ_:Ec·Š÷\[}¿©,.$ xTæâ HŠ„ʲ¾-!±&²<À2ÿÁ)31 íŠ-1´šÝºï!ÏïxúyoSájî«;EÕ`ȵJÓÅÂÔàþ!8³.~¡_‚™°äÕmÙ–·së|ûÎÊ4£8ÜY;ÔÇsbémœŽuÏp·ÓÈd|ƒHJû,ÛÉ^~ecyžêëÎ:ƒu㯤Ÿ±¼Åv•³ÊËäèΤ7|9ðÞGQ2Õê/‘[„RO 5—òþ-='˜¾,³ 1~d¯°®|»Û+Žp¶7Ù¥¸âîN¦¬Žý©¡ŽÈìAe Ð-p¿Žù>hÆжS;@êxÝ'©›¦Ë.¡@÷tù|s›¦ðú±³ÅAâdÜÇQu›‚aÎ/D>ñõr¥µD0¶’ïâNô|˜Ô_‰ƒ!-÷g#ªP®þ½æaüþH‘!ï´Èîš_ÐÖ²'¶íÑ蜄ÙÔÌQû´l iaZ®¸¾­N´Ø|ýE…Õαï)ÑTÑ*2—C„PË”n5Ô„ŽpŒe‹Bž#: 8—ÄéøC†ÊÜ•ÄÅ·Hž™¿}Üœ@{ øüŒ,)|œŸ£·Ii~ÝsÏ{¡Æ-å}>¾OŠ]÷‚1qHŽÍYBöŠìaÊ`Æh“/®zÅš“G÷VuçÈ7þB©ç2H×2AÊÞL—·œÞÔ‹Æœ2‹O/~rô ¶sÒÎÆÂóî%QÓI`:ªD€½çÂ9è–Öõ*}«cƻۻ`'Öõ±®Œ´°’' ú_ô9%d_]Ñ“IñÝÂ+ÏB^pµy¹ÄBæfÚÖ}뻞'Z“ðªa²YZ¸‚%x\‘hÍÄ·ùàë‚'Úëä)bôýkÚ¹ÙƨY· œÅ…iÊöAoä$ãÝ ¤,ƒ'¤ÐvÉ ‡Ô‚õ4ÄûÍøŠÓwÃÄ ÎˆyÕJ–øóèbkïZùoDZµÉRnºýŽh±ƒ¡ÝÏ´ù#é Õæã-p0‰'h #%!ö6êÆ÷Ñ›ì8.E“©Ÿë‹•nÍq†NÈ—XgRbç‰%}¾ø›gÓR>¾¦!µ5ƒšfÜäå¼ëÒg7$.ǸàxÿådÙ-£}#Òó¿€47h²zÓ7tñM3™B~¼éa´åÏ\g èÑ–8Û· è@;yqŸÔÊòñ/óZRÚÀ¥CB¥r¼]¼§ùH87Mà„ª xMóŠ—Ò0“,]KMnbwøŒèjÑ ”µƒµ“&+‡ŠsÉ;™#…XÞ'V{ °VËu^Lxað>Œö­,u2 w|‹WÙ›81ÏmâKu‚™º)=uúÒ«JËyù&í-³÷´ÊÊýÂÃö3$×#êýÅp®ðˆ¥{14äS‹ÜƒØi^U^íå‚ÑImD¹$FdA×]J‹C©%ªõ¤^7ìx]‰»q>Í‘<ý7L¸¨´W¸q@›!ÌOCÍ5®Ñ`‰lWl)ïàË…FÉ xÊúuŽÔD%ÌÙEÎO¦f²^yec1ë*d·(EÖã „–®DzXX–Q–*³âmE!Ö±ôß‘\‰Â0Ú11Z$ajM³g‚ÝÐ!œ1ÂM”Ûã+m"‹3ŒóSU²Á ÐÃÚáå&yJ½ú]˜¾D)xhbbM z™eKü’3ì.¨ãÞ£Îëd*·Çh>¶ÉÖ„¬b~¼Š ÃR»èŒÃîÁØš0šý‚è¹3?.É·œb³8£ªšýù° ²®SsX>¡. ÉP:-…} mûE¥V†ÜÚÍKÜp|:Ld3¿W_æD]XUãü .źåázÃ1·åy¾ ‰ëz7U5@õØg¼02£!]¦oVw_^ÖÉsü_E…“üV¼x_áùáÀM¤*až–ʪpÞÁ?Œ U'ìˆá6(¯™¡Ji¹ÇÿJÛù<þ2T6¡ ¥—–Se=ÚÄ´`‚ÄNÐ ìlݹÃkz××úLx[L«ÃkºU#ܨ]u ‡&™ÍÃÏ}kâ”5åД{2&ªkY¶æªÂ˜á´î¨ÝÑÀ»­V„Lr“WN «@"“dëd ½ÏÀROvÁv¡€6.ã½ÆÙEçï Ѷ~A`Ä3³Ñl©Ý¦`ÐþõèݲÍmlHJä5£æ逭7»À5Õ¯‡áȶÑÙ+>Ô¦3$°‘”É=cs¨ÓÐ%M±IÿZ££ŸÁlãØ”UŒÒ ¬fFˆóÄöÕ% „)͉jä65è¥û”)ž|ÿø‚ a|Äá MÕÍ#„ˆœu9 “·¸p¯$•)§~Mò ¯‚}kÙH[z0%CÕô‡¢[ÚaïN“]øk™‡ ¶ÛR¾nAÜuˆÿt„2Á5†?ÿ²h”žÖÆ2šàÞÚ¶T”÷qÔÖÆët\U]WQ áYO™…#pý&\ M´…ž÷…ŽOSZJ1»ï Ðå±–Þ§ûÇ͉R DÄKºÇÎÄÐÇðn0– ¬k{°,0ÆÓOô¨+7x›çH=öë"ç°wåÅ=)öÙµŽˆÐp} ï.§ Ä”ÈRw¢ïoÉÈ÷6¹DÀ¢n¾J¼‹­Xº cíúB‚w¯~-1F:¬Y^A1mö§§hCeµÙ&4“—OwSâÒÑŒý@¦dÍðÝ Ì©ï×~L¶ìWiSñm J2H%‡”,(&ŠXA;^2µO]Ö¾-l÷ ±C®îrÞ`åjGPQf#šäëˆ÷Ìæ¸Ð`“¿òðvq"–äê òJ‹‡Ë$/ʯW¦ž¡Öh“(òRMìV’KÖû'¿‹ð{¹Ë¯«Ç 5¿(ìÈ  _ŠÝö&ÌíÃZy•ŠiHu¾Ã¥ÆAaÀ‰¦¸Ój|,É÷}D0#(R[8ùœ‚'Óü)6šÔÔÙ?§cÓHÍ2IÏÛ†Ñ4öMï{rºú„›s³ø“kç{èÀ‰ù=|Ü6¸7]Ic~󋶰ŸS³×÷Û±–ÎY<Õ0à-lMe;rï6œ¦i ý¿êÙmð*3ç÷Htèñl‡G ¹&{ ÐÚ= ’F¿˜Ï›Ä«³¸nO2 ¯¬sSè%:ÛÁ€/Þ.A”d%ž`D§ ’êohVì=ŒIÀ/iû}dÍŽ_9턵î;îJÉ…~qöÅÞ!îdvTr%~…cŒ„5æLZÉÙ6¡+F“:á•IÍŸ.j*1ëZþ¤Ý¤m°dði«•°§ñgRϧ{/XBB²—ÕJQ–…Dp:÷å ù-ŒÙ,Y-¸·<Ã}uëŽùrÏ`R“%üY|8£að`.½bXøJqéôårëÔµŽôŠLŸâðÅ`YcÇ6áx¬"Šu%pš6œ)ÂÆéåênúå›qäŠnßCì¦Mg¶3>’gÀøý(Q¨Žr¤¡X Ì—{Ý”¦¥rÝ”1I(-$ùYy^Ãb7îrŸ1VÁ):¯Háå—Ï[Â0¬´›”Òg€hÌ®Zß³DM‚¶Äi˜€é§:Ý…`Ìž… ëÙ*×’p‘ȵ·gs…1ÜȤúqN‡¡¡áÝ Qr‘äî$-«ÓòY~ÄûÄ’y¯åÙ/¾MÃ~Òo±\òÑôfD%ºtå åwêœíöúJ~ãy ¾_Ú]Ô—6ä/µÝâ ­/Ž…•q 8î–“„¾ hg¾?¢Ê¶¤ÈÝÄ´&Tk¾*t½ ‰ <¡RÈ›!Ù}è+a¹óïB ½ûLq0˜ñ¥·Ãõ™ñ¤Nƒ‹yçEâز¢ ŽjÛÆÛÃ’%¯q§x¢o›iÊÆ'ŽƒMX«ßŽô7T»bX¸\s£Kst€~dlíêäLá%m×gvì›…N³Î$Gƹ·ùàߜ߇G-Œ¹!Sñùâ…fhRüü6Ó·ª¸ šRÔu(.{”÷“(ÝÍàÑZUu0ýÚ†·ƒž„'ìýWVÎHáɲ•ÔLòèØHžshBï幂ôz£ƒª7ÏUäÜ^߃­žÇM·ë=a½€HŸXíKo‘Ù‹«T\åMX•ÇÏ ubÑÊ.:òš6îSç¾X² õÚ‹'ë¼±Æy]qc?‘(ð›%þ€ï®”ðÁ£"Y©£Û|¾· Ð!þK_°'˜§þZeeH1 Êým²6oþ—»ü¢+c’ö³c¯5âï‚È®¸ü*¥ pÙ;ùõ¢9ùÁÛdéD¬~Ñ®cÎ_8BSÀ¥B7«K ü5ˆñÂ`4ßSçí†Æ}ä‰Øuß“¦JrézIÙA@t•lT<Õ´,ìí‰X™à=xèH§¨¢‹[¬`[.W†æŒœóˆ>bnÔ³Å7^É KÁã —Rƒ¶çÔî±\åšEJ|8kŽ2?ó'æÊã¡»V[¤Pu¤{Ë2WΔ.v"ÊhVxÂmzçÚÍt§1ãá+›ågÃ¥ÈýÔŸãC÷-ÍSŽSÍ'…¾DTé§ÔíȶWúXîN«|ŽwãÎÛÙÞv‘çóÉf‘/6ã õAHæ\~tÔ˵aLGé~;ìuæd4¿ÅHÒuúv¿î®ß`ãx P¸ªŠ¸€^?k«c>5Ë÷‰## Ü£ÖˆSÌÐ0/-ásÄtŠ¡é¯x²ÝÛvëEéO‚Aœ.šeÑeñÚ´P[±‹`¶þ„·?ÒŽÔJ2¸Œ8Å^z¡ôÞ_‰*9M&GË;nI2­½QZe”5ƒÉíú~åèŒK¡#è¦Î|ZY)`mrZ¼fTUéýBp‘õ­B.²”‡Ÿl•TÚdhü@¾Ü†Ûhé-!àÅ—òä+¼V û¯jd’Ÿ·:¸Äa³.-YG˜½´¶{ cÅÝÂd<¢s2ÐÚ³& KÙ“õ$8f¶éÃDíÔ焵ú)Âw_M( j7 ¸Ýmr=u±?~QCT¸p„éNÜ]7`†ñ7ÅÑóLºýåYkЉ=cŒ¨¥0ûÉß>i¯ B™;òæò¢’%‰I‹+j™3‰A0é\$ákÌÛxgɸÐö žLö‹Åë( íd`æ¤ûes9ÿ–w-h[ÈÏ\}\10 »P„qè4ÏøqÔètRIL ÊPôf÷’/àMê¬Ù³ˆ[“ŒŸºBzXUÑ°šÂ÷Ž†Ìç sèõ…¿„Ww¡œ¤M4î“ Ûw¥ f„>tFýyƒçäO$ WV¡Zx½D¤œ ¢LD¬ŽZŒ|hRãõ!&œÈ8ÞÚÌ3Þðj›ëχ<7dÙÇБ&'ÔEÃÝ9‚ ˜ìÛo?É_™î1²… =ªZè3Þ/ø2ÚBH¥Ì¶bË°Nt3 Ê¿„´L y…æè‚ʾ …DYÿÄeW§4j íkpÌŒØfúõ¤T˜»QçKá ÁÂØËøGŠÓ5`qHëD>¿z4¬2]‹nu?»Y¦Fg=îqˆ†„þÍKº—›¦‰8Î$9çx5­µ?4\¼¹¡‡¡ "¥õ«‹Ä±6½ÙôPûrÔBÇ$57}ôøÇÁ%ƒ½/ïSàÜ®~Ê´É×ÍÊSŸ5!î“ÉØFg¿È¶Ñ¤TÊmÄ¥ü¸¬sA(.Ñ(–ûZdÀÉZs¾ÏW4ËеóMŠWAzù«þ×Lb¦Jx'gƒVܹ}(ÜI DNݯ|þþ¤FšÄ—äwÐ@Õ¤ï½,IÝÂ.Ô44Õß]=,ý=ß6óå{¦ú¾¸£ |Z¡MN'ÛÏhìTñ©,Ä”]X)V#Âq8±wŠÓÕ—¼»g]sIbŠ4ØxÛ%L!åoÁsN*i|¥ÏR‚yfÁ"Ù@i¯”A…9L ‰ýá]9H›X€PÇH¡#ý±àîíHÞX}ë[‰m£ÑQ ¢ø såü[R@:ÚDýY ÔÍã´ÑOˆ¦F¸¯OîõM»¨œ">xj¤þ0†…f£óó‰p!ák ÉRAÞ¶LÓMw²[)òA™„.«6ð‹ñcœ$³{É0Tr—”W˜'£i†õ©¸§+äFSžíg…“ ÆÉk9 þ¿HfuŸø¨qn¤ôa¯3+òÆÐ\BÍŸezòl8)©ƒä·Óè6JþÜwN=ïSQ×f!{leX7ôfÂ0×ÕÎ $Ì£‰Q¢‰Ñ›Êz‡¤ÏKõ4Å3{8ÕÂUÑaÚõÿL“UK¸•ãˆ }B4eÖïÿx§q,kföÍ!lAH±–âàgN)8¡U^ÎHOHè[ýÕ2b~,bùžš¤Ö L×0c¾äûÅy.¥ÚÑ0ÝZÑHÿn6{z¹xH:#½^¡©h¨ |gõáÕÆlŒs²ùš/(sqxn…ß)Ujl}ˆ(ÇÑ—·ª"g)ó l/ªæ¢Ö›‚™p «£m<5Ñ£FÉGò¹Ë?Ô‚íJܱDŽ. ::'^.¤Fø7í3dYk¤íÈÞqféµÐ;ï=$iC¸*‰ª}E{iÕ¡Äð‹, Ðô5²¿™½"YÈ­p5K5ôˆø‚˜)Ìιµ»í&‘ª»ú¦ç^-³ŠvWØÀgDÅ1ð2Þ#PÿÉ^6zm«Í ¬•þ.r¬¾ýÎ&øµWz¹=‹‘¾‘£‹Ö•µbt'ð>kÏ;˜‹ÿ²&zL¿Z£®³¾+5´.8c§Z¹û¼÷;}zJf¦/5Ãn$¿ü¶ ô6§î·kô& E˜-%9TÚ-(Ç"²9gNqaÚ°†n ]qäé9:hò¢¤ˆ¸¬Zñœàr?V¬lÇ>^ÀÆH‚ÿÂoR÷yo­dóÉb÷)å“‘â!­ž ëòU(ÕH]H†JYŠ4?©®§£øL ¦n-m‡QHýâ{þ—"Ûý£ÈK ø<šóD†Ô{|lTéôUé…•DX¿F„w7bîi«éšD󘎫õñºŸ g7&´$5„.°|¾çV N4Éö#]Vº)¼ˆÅ;8e= 9—äpõ*TánÕl=¥È‡šoÍdÛ^ ³µöEqyÜÄ+êØ>†:Lµƒƒ:9%BdªñpèjÇ|f{$ÿúöQ¸G𣲑"ãñ±é»«4 X±Ñ ¾×ÂLFJgPøGáåž TXÄÆÎÜ…¼ M§¬ÆªñÀx¼{Q VúS¹Ö,ÍÉx2Ú8i>7¼)Gþè^̤>X=‘Æ«I,F¸$5S.-õÒ¾tu£ c}¿>Õ"%ÒWsl·FêR-ÂX¡$îK%xέ_Ÿª Ïú´%k62ÿ[Áèw¯V§Q´.C(Æw‡.‚'í|¢kȃÁ±Ñ³*jÎa«½èõí\@ˆô{T¬Yîúu^e¹òÜäók&ÅCP/xµåŒêc0NŸ’›¯¢_9ž¡YM>Û˜ëAe«Rgѱ° !n“ŸÃé`§‹–ijÙÁºKhÖ·u¢›ãI+É«ÇY9™ý)ºƒ˜°õ;W«Ú!FÃ.¾·?Ûå›áG.Óöxƒ],ñÛ!¾ Ž©ÌWq‡Œ¢ˆ#ÜãxÛ)DÒuLêÈ„z„ŸSÁÏùæ$C+éŽìí§ð®tÉlf„¼´h›ó÷¨¬ý¼Ä¸ˆ˜;‹"¸Y¶ôyhÌÉê¬ïÛä…ÊZ£àÃOãï­>§¯?z”–Â[¦MÜ$œ²ÛýD­¼ñ"–?I?•-ÑÏÛcµgÏrJJ4RÙL­P“ ó¾ŒÌÿ´ ]ÂßÂj󅀸$ßäž5Ðu9¿ãWR¢Øh+›³)Æô)/&ᯆ¾LzÍZ”{ꄤÔŠt>k…0;ÊK0üç1òõTm¾™0ÕSœVÍlS}u(ˆœ°`xë­IÆ|ỽ–ûƾeÃ*[µ”0D ñì[4€‰uø¼<úÅy;Û·¥Ù$áYêú(QÄ~„‰B+¿ 41W…”mÕU¶ø†!>ëÏœ€_¡uͪٔù#Xª§ßW¨2·6}õÙvDå V¹ßÇí¸ëÒÜÇ·Ìב(ILz,ÑA‹k|e$áF+Pµâsz$ãhÕ÷h§iœÅV%•i ÉèzEAu†› ¥èŽö¦¸Üè_,m/媞Ñ+ Þ/”ÆÁ1°`ä&«5©A……ÏH§¦l¼à;ö`"Å¿;|S™q—nÂðu@gÔ~?'…äì ¹«±W35K]O¾ÆÞ¿ju ñßåàFbÑ‚…Dó{A¹ }OrʉCõ–r$8™b@\œÈbt©¬éÁaTy}ˆ8¤‡Ì7‡›„­…¶$CÒ±öK%Ù³_Ú„UX„:od'‚XFäéBMª ‡|†ùëÄÁ¾¡ÆØ`Ϲqréíì_?Uï^¾’ºŒøSE©›(:ÅK]o¦èBt*˜@W› ¢_Ô{©.S½ðëѶš©né¡MšF1jÒTM^ÿƒA“eßÇßÉÄ—Ox;T冃Áô³†* ÛuÊut&}­^ía'y,åE2Ã~<™’YÓÇôUL}}óQḠ,cú^‘—›†h&¼ÉûTì¾E|Ò—ƒñÇï»4óçy'ä bß La¼çƒ?Oª¸‘šLÀëI&iÓw¡dìé-_džÏÖ,s°}êü !02Ьje3ܱ0&3ê"DñV& Ž–ƒî1"/õ¡$œ=ÅÅ&s‚„[«tpÿã"6ClEe)‚iÙN¥Š…‚3ŠWÚ#D}’éè΢*¯ì§(¿”ñ ˜ü0xÎO7V¾®ŸÛiGö¾Ïrñuáéªjm‘§žF:ùõÙ†7s|Šhy+»ô!h£óœî®&FØ cq¢Q-:ëN$”¸ÅkG  ŠZšï§Æ‘ WéÿÖ˜©T.S•¬×ûXn7uhÿ˜z{0¬½á•Œ k|ÂSuHÿ°'éwêxÀš}u‚ú¡±à^m0‘ŸJP»5'KÇçÅì‡ô pÚR®)ÿ‘Ѐ§tg€`_ÈbÛ‡ •œRë'¤¯¢ÃCï½Á ,e¹~覞ÅEº‰“91>Aö » £>ACR$Ü"?Õpyˆ#¬j3úªêñò׶Ó‘÷e¾Ñá8 Ì™óx@Eô"áñ~˺>]ë½À^Çf´r¸†$í‡s†LØf~GÑÉn×F‡ø…Þe¯”}³IÃ5YÏg’ÃîóØs‹œôLŠú(Nó+ºFfðªqAÒ>>¦‚ê„¥¼®Û$¸XÍ0„ùÕwð­ÖÌQ–…µ}$Ç~²T™î4Í'‘’¶C‚ý[ªó>‘~Gk`I³EòÙ!ð_¼z“hssÅ9…Þ!Þ;ÅúóæšS\/ÕúÉßéLXÿŸûï%ý[šàÖºÎô=ä{Z Dô77Ìž˜và™8òã;H,3p¤• o?½ßJ “z†€P*€Cî…àŸ‚XŒû— (¡%î+ z9àø~zÕ2îØ‚Ü„£F¯ß,jH7'žç–³z)YhUêT•:GK ML™¢SçÚ¤n¢Ä_`HÄÒÜ3¹ÜzÇœ3Ê)@œÑi~rÝÆ–R´-xq ?¿îêhÞ­¦“ö$,¡¿?,¹z¢S’—º† œ?Èøk»]¼ê©½¡Ì!aÚsÜàRAgX”w Å-«Z~!ƒíDn>.tÀñ“Vx†M ¿ã|•ñz€ªxx0Ø f!è9dÇÃ;ªG¥ß7NÉ$û5O·òNK1j}(4Ý#{>Åië_–f#Ò^*µÄ¼#ƒÖÌÔ ÷~€¿ˆ3I?¢w9V…¬$ËûbBOË›D”ä ®AJ¨þOø„LWÑ« ò´¦áE„2 t“Hý~­_¦~hË*ámÎrÒ<‹‘cNëaÓï_Šá¶ö\p*ÅËeOpøX¥M{2ùAC¸Û3 $Ú‚‹uåÍÏ(Ú8ù©:Þqál•~F%%ðYË7:Àø¸ì½°?ÄÉš8VÙs;Õ›¥óÄHk°¿ÓP;.NÛ“»¾âÑ >epû®C* ΂Øhõž:°“¨-vn@hFeÖbDÐÓqOñʵ#ÅÝ|` ]£±;2åcá1Î[VÇ kó5ZŠ e„¾.júõnä²®6Z%íö¥ð”ê½|53Cµ2H{ø­ZT-ûù {ÿœµ•8ʉe•Õ2Ì ÖkäP0Hï8ý+ ,;µÓÝRB¹cŸ‡ëEì©—‘LÚ.É–—ûyæ¡¥¶¯¦Õ­èxõÁa‰¼7ælÏ”?ÓÃT›õš©T98QÝÅ›03`Ñéò($äט•K‰8{¦±Ta½žÓ©Ü6)çd# Œ¦„->Ó_¹§ƒ:jjBXÝ¡a”t”ãÇýÆ„H–Ø•ÃA'gLÓ©;Qý%gWkqÊu[ž+ìÜœ”¶Îz¡«_;´¥?{·GBKèŽg7¦ŽÏÇHâ”Ú ky;ÒÙX3ø¬òUqô5vÕ„QTÕ"Þ`)¦×.s5Mqë§N­¬ùR¯Ê-ÇÂO¦=«Í¿æ¢§PCÀ/îµ TÂ`ˆ¦?ÿï‹´Ôýª‹§#k½{ÐHÊ‹Í<µ‰"åÓÐhoÀå2ܽŠÜ/s,vduuPwÙG@aÞgPã‹ÑxÓÓÑÆT7BÒÇŒayIÈYi!¦“ɧÔ| ^§¨üqâÀþ”+7êÎ:y]"¾+Ô4ö"[N˜=û‚ýcQQ˜åå Ö7’¾üËÑŸ^ýPBåI’„,53ÒdétÖ´¸EíŸðjøAÐŒÁ~'eј;×|Ä@‘¢—aA²øyæÈ…*Æ´Ï oG•@ |fì¬"ÏEÚvàðnuÙ8ÌÖ¯Š« ÷+VrJG¹l_µ\ºÕ¾Jo ‰®ÿ]DP:Þòò¼¨=ä‚J×[ì‘’ç|À`äkiØ¥\@­‚>Gb|Ș®MŒÓ]mK$)ÿ;2K¹—íbömëjÿËäЖ+ˆ¨þW†„²¨[Œ»þ&ùUÿ ÂîÒ*¯¸ÍÆ,µײNe»†eRùy(úr¤”$~ûå,Ü:ÚØ-*eOY6>(Öâ£Ú徶,s.Ý$ž2?çúÓofœ“[ïõ€ Ò¡sŸA»?’ßhKŸ¬ù…-€eÞ1’ûðgëŠÑÝöœÖö\"óÙamŽÚ‚Þk§ˆ ¹xün¥Ç éÌ«]'«EŸ–²mÒÃ^x–Q<ᣠ±$åñS¨½Þ"æïsh«VWŒƒëÝe’¨43-D32O‚ÿ0‰PaYÞ„R5–Ë 5œ ¢Z½×\ù-°^ë_·nÒŠ½ {šp‡n¬|öÓˆo–ÆžêöÊDC )w90¦îbÿè…È9ý}Yg˜ÆÉõC (ÆÒ¨ðwź´˜š3¸ŠÑÇX¾Sg…t%5Éä˜ëë’+†ù-ƒá£­úŸò ½P##'×Ø¿¨m÷²†¿œ]Œ›uÅOÇé¹=2WævnÂ;Ähö\Aþ:4©Û";5VûîJÚÍ)g)sã jjÓcGLÉÓNL÷?;T|¯S)¼­ñ÷·+n¸÷znƪ›]< ³<ð¦xÜØŸB¶Ò.LEÜËžŽ–]Ÿ]nÝÏ7Ì`È­¬áFîü°EÖ½!ê ³¸ü%=~™ÕŽÚ,Èv€ò\lÏšr¤PÜÓ;OÑ+ï÷®ö*÷v÷>í&œz°ü‹_oè¦×¥vÓë®ìRÅ#ƒñ«h@!QÇ`à÷é—ª«ÜLDü‘ãÂÿÁgµØW1hØ–šyá%­yæçvM iÏQå¹hW¨V ß.Ö Û¦]áDTºo£`£¬ 9ž¬¤%ŽfËk®/eÉÑÿVðÐTÝgJ:Í>POz¿ 䧫†÷vühÒsà×æ½I‹í3„?¨4'sܪS^¨` x-îÖÒëö >M5t‡!¥i2-ßgU4Ã-µ^R¼Û6ÙT³mþéÀÒ#ðÅo«}…ÉB„Ý>Æ';WÐ öú ºèó&¡Î•·%/¤„Ðmšu?ô#^BÀïÊUBæ2Þî'œO*gÆQÇwùɶã{·ÆQTïÊ/äZ7!¦Õ šÏ"©úSØÐ1 ·H%½cä0•'X«^}6ýq_²¾«ïÃW%åÞ¸k"þbq‡ÐEïVB?û "@$5…†‚×­–IüVkŸú‹ßzs’QéÿÞHj§`ÎÕEN®ö s—$\ »\.…²±Á 3Gof÷Itnƒ5¦©víÉ+‚Â9±ŽFB'š—º&.LóÚâIâ"íÔLkb+ÊÿZ/鉎Ÿ€nÿ"&`¶3`2wÇí¼ñZc‘© N—¶´äH8)xs(­.ÅàdîâqY¢çr4w[&ž­J-–¯MÐE§Ò?"Rx”»¾Ñ—-o÷Ëãb¡ØN¥x”Å `ÙÑŠý'p‹'’óÒH0•Ì&$-‘* ¥6iêjU~?òæ:¶£B๗‰±‹Çê#(ªLP@ð·j5pØ.|gOÉtŒp/Å«]~³|³ÌÚÌ)eð÷fn¬A!µ•…2Ò @©’—”›jÀ¶ÊÊßJ«¸5³˜­«ÔMKÏpðH=Jb)‚ÍÑ”FH:4¾fsíÊ,j¾´/MqášR sý+¹vÂÒÊÈòßm¥!(ï; »W dÌÏèP£a3ÿUÛ©A%šñZ˜t.dÊý7yoç8!G%JŽ'Ofú‹{{ ÕBú ß,” 0• ¢ÞõI/g° ôiÖS¹ÿ¹RŠ»\ùÐ1•+ˆ(Ô­¨‡¾‚öÈ´×ñGF~NÛQiÂΓ^A6< N4FÞ¤Ì ‚_‹…«Š?a Ä÷88éŒ-³æ¾á#‘³ó o'9œö £YItH0D"®PñQC*q-e!~ÛÒå®*t¤&{+Næ±¥ò>¯2|‚ËI/á=ù¼5Åz],šÕFtéMc¡šÙ—ÁÒøŸf)µs´Ÿù`µSk{é¨oŽ6 ÔH®R"Ÿ¹|uù÷Ìœ >1@Ç{ßß}@™0 ªë[)è$UâÓŬ4f¨ý&{Xâÿß½#|¸Ïü¼cáC‘üfk­P+q#S+aÛì @ߎ¯âuã$£¤ûÀǼ þ-½Ú2[ ç:ÿÖŸÈ¡B~«ÛâtˆCM7º¼Ž(«´ŸpBŠÀ®®µƒ³…„¯ð»gÍ;Ö"ÀÎmAÏàgM§:§"kš¬·]¡Õ:ÏùÁš“˜q>B„˜ 4»”nؒ݉QUr“\”ÅÓ¬i˜jd=É‘‘RØP¿o]‹ÀŽ4Rf !¿ÞèXºž@ bLâˆ,rÅ<ê-nm³ ÷5׊«LÒÒö8(~°^lœ¾ê¦Ÿ Uxà›âÿeíw¢o|[ûn[–.ÙƒUiäd\S¾ZvNm S¸æ»ö©ãªÉÔßyX¸øîë¾–Âñd8ŸP$~7±,bg–B ˜f9Zèªs|®",åÂP¤?75À¨mÖtè^¹Ô4÷1¨®p’çtNïÓŸ.­ÀžèaŸÁÆ Ä‚þ¡e%Æídö€k[`lx߬ÐNÓ%º÷Ï<ŠIÎsSÂ*í­4AžöA|¶naA£Sdås25ÝE;B.}!iZGW;SÔºŸãDøFÁصZö…Lï1F-ô·.íMfÕ€†©Œ¡»á<È‹É;¸–ôS•´$Ø3,Q˜þd6ìôЊŒ)ßžð™Fåef3?x"Ž»ùÝ¢;ç|Å–HEÜ«vÏp°L®8}¢³…H2Cî`|ÅèÇ×4ï=q€kh`<=Ò„¥Ýt.‰Pl»,6œ¸Š¾1Ûâí°6Ä¡˜¿ 0¶$Ô*½óqx³ú|PòNcÈN¬vŽ¾Ý Â1£W°§íL¢„Äᧃ¾päù7%‡qçý_ y |¬pV¿(Ò¢6¬¸½}2ØRÍcÒJz¤÷LT¦hµV%îfëŽR®(97ó¼7X©¾ÐXNÈ„<+Ò¹);ÓgÝÆ{6è»ñDVðæQm¬™D 95¢ÆpO•n$_?èsшã^©ù &…>Y; /6UL,§;ãpjíÖÄTw”=”ƒì‰ JÉ#æ7âœ?¶´ËXbºÉ0«µ‚ «–Áƒò©4>BÖë$è,#q,Xá³M#Á íåŒ!(×Ýn¿iâædn:#êÔLTwU[\è‰Ý3„µEGüu¶òé¡<ÑOå]½ˆzÉSœi¬r`‚jËuÍ™ÊûB›‘‹UÛ’„оV™&¼wå•Ï]à_êTPýúñüƒ‡ô»1,bq!!ÌùŠ q ¯–™D9ÀZæt¢µû/H4×ã)Ÿ†Ái¯­u[þ•õFÚÌhÊ>e-º,0¢a=Š›´“¯ endstream endobj 2144 0 obj << /Length1 2242 /Length2 14314 /Length3 0 /Length 15632 /Filter /FlateDecode >> stream xÚöP\ÛÖ ãîî4ÜÝ‚»»{ãNcÁ]‚»' î.ÁÝà XÐðúœ+9÷ûÿª÷ª«º÷XSÆk͹vÓP¨j0‹Y:›¥@Ìì,l %MM~'  ¦-Èøïe$m ›»­³“À?$Ü€f ðš¤ì§äì÷p°sØyØyØØllüÿqtvHšyÚZ”XòÎN@w$ g7[k˜æ?: z;??/Óßá1G ›­…™@É dt3Z˜94œ-l ŸÿIA'd¹°²zyy±˜9º³8»Y‹Ð3¼lA6u ;ÐÍh øK0@ÙÌø/e,H4M[÷­k8[¼ÌÜ€ð‚ƒ­ÐÉáád t€ÉrŠ Ó¿œÿåÀø÷ÞØYØÿ›îßÑ%²uú;ØÌÂÂÙÑÅÌÉÇÖÉ`eë¨H+²€¼AL3'Ë¿ÍÜÁñfžf¶fæ`‡¿+7H‹©ÌÀÿ-ÏÝÂÍÖäÎânëð—DÖ¿Ò€wYÊÉRÂÙÑèrGú«>I[7 xÛ}Xÿu²öNÎ^N¾ÿV¶N–V‰°ôpaÕr²uõÊIþÛ¼„ôgÍp³±±ñq²€® ·… ë_é5}\€ÿ^+ð÷uqvXEým­€à$_w3O äæô÷ý§á;;ÀÒÖ0ZÛ:!ýÉ^Zý ƒßÍÖ`Àî=vÛ_Ÿÿ>ÛËÒÙÉÁçûßç˪£¢,%¥Éø/Åÿµ‰‹;{|™yÌœœn.?Àÿ“üWþ¤ÿ½ªjfûïÒØþä“s²rðÿKxëþ£ÂóßMA÷ï¡ü/ƒ²3¸“º?oÈÆÍfþbÿÿÜþ‡üÿëú¿²ü¿5þÿ-HÚÃÁáo3Ýßöÿ³™£­ƒÏ¿Àì…’3x4œþ¯«ð_ƒ¬´´õpü¿V9x8Äœ¬þ»¶îÒ¶Þ@KU[…ÍßíòŸCgw°uª:»ÛþuÓ˜ÙÙØþ ¬ /ç€å{ü90ÃßoBw g·*Kóú9ÀYþAʦñùËþðG#xC?ÝþEù?#kááæ~‘ý}§‚çù?øï·&è ´@Z]r¶ ³«ë|ø*FìÅ|0%{д2ƒ7<]r&Ö0DŠH¬)zè÷êê§lÝÙ#OSàêÁ‡¦Z„ýà5(ãÝ0T±>±t vXË£€üR1ǧküy¦ÐäÛüÉg‹ŽŸ“æ|qnÔ¦Ò½û[ŸYëòžÜîot†o>ô¥ˆßÝ@ÈRd/$5žMG…¶Ågâ­Žã<ŒäŽôP쟦×÷Õ©õ6j0À˜(NùìëŽY3tQP‰ J[âø<ñ£zJ³´{÷ÓêláOÍYD+ÌtW?ÙÝV#…nüD­ßÖÌe¶±kÍ\€éMFXª?0·š/ò&~ ˜±&³'€a Ì6.ÖO7š*o¤åä´ŽýÉyxêåÈñ ¦7ÞE<6Wn"-Z ÁµJmç$$ÚChÓ˜âJNJxšaV~†¥Zß+bÎâ†3s‰ø\¯n‰…ÃNAÌ=N°­ôj°#þ—锺ñ‚¹„‚OGTüÖÚJó7Z_B„+5ki¤ÂRO‡zÕÂËÖ/_‹&úz‘Æ€GƒÂ2>^HÞFž%í˜1pÅ´O=˜ÎüV-k¶j.ýC[³>ðŽj n¶xv‰ãÒÐ놓½ZðXâˆî'G¤‘ÕYaß|xÏö‹à;òÔßhvÁGw„Ï@ÍøZ»å²^™ŸÇ|°ò˜ð¹k݆'`6QxµY øpÇ»O*®j¯1£Ëø–þÑêÕkY¸ç¦¾Gç6„þ!¥ö€–Ôhþ“WõÃ+ó-$ „hG Üà ¯yÕ°ÿl&ǘ‰¬~Š1ÑS§šþVÖ-¨Û)àv‘æô2^ØE—Ç šzyâß3Û%C}7¸ÄY#D"²ŽŽÿÊÿ#š³ee |_òKQòcT4¢ J]ëJÖºO­¼„ð—°§©?ð1ç @Ϭ¬Ô±/"!Ú‡CœåŽYo5üø}!WyÂ:.eQŒ{K }NË‘äœÖåT¼ÃcjÁ}¢MªsÌç¢óu’—Ú†Äî_’{ "…Ê×TM£Ö¼sIÈ€¼Fj÷y?ñ;ÁXbôëúQµ™:QVF¯4oJ„†€~b¸Ž¯˜kGójU•;íýíä\.~-Öí|s£»‡í· ð'Á¯ûAJ³r"Ή“&J$DÇá­†„ŽÛ ¹4c1\¿ãï³Ê˜G7D¹^æ¦ç³üíÜ"õ%[–òòá(_¯Ø~ÜÞè&–]’ºÜçÃLn6w°É`G=ŽÉiVÊ®MIßÝ!±i¢ÀÈù"’‡óØ >AMxŬ Žv•ö¬E#tŒñF>Œd¾vÃM¯Ÿc=® ø>•á|A¿Ëo|ï ÛØ™EqÄ— RcURä¨ëÄÓ@–9¾ªš Õekû{eÔH—-í‡ø³·/Q&7é·*+©î¨­šnIbÐÎч%~§qŸBñp²¡A@ΈqŸ+6 æ÷.:SË2u‘¨7í«Ùå`9—lÖE¬{Ô?ŽõÝ–—7Ü]ZqR¤nÊÀ˜Ò ªEy‹SDé¼{2½OÎqþ=b™eSØÏó»ŸáÇéÏ^­¥F°,Ô-3¬!Únó™L¤1rŠý°=ö ª¾ãÜ/¨É1[àR»×¦ïÀ¨_Æ=ÁHŸd9ñ<»n3Ê˾—›Ôø‚®ÍœÉ,c£ùÍ[ )D}RR‚=}|‘Ƨ~¡ô‰hžƒó0'xË!î«q±}â“@e5:²"9ÝÚ¥âÀÊYˆp̯OnvÏi?‘_}¡ß|ó!W·ñUõvírÑC¢¶ ¸¸ÆÜvA_ªf:’³¨^ešT[6×C¹ ”Ò ´Oa~†¼«ŽJ:ò1ç )éOü¶7Sê k³˜‰<°ãoþ`=Â’5ÒÍÚþKŸ|?»$¦¨¤t}-UŒ÷FÃ?‘Ž`ùȯ¿Õ« “Ö[‹ ÚàéØÔlß–Ü’ I¯Ä«ˆ Š¬}ÝúÅN:/6Ý0ï—êÚº@ “Kð*ÀMþs¨K²Éý:î[Áøê8päv+k‚X&œb\<+ÍÅ8[è%sX]²À¯ˆ„Ö¥þj3l4CØ…àÊý¸Kx3†ó›äÁ›çÉHzõÍ;Òð™Ùœú{§ÉÏ×RÄA|Dˆ!ìÓ}o½KH•1òGâä £€6{ìqé †.ÈÁÂlyYÈ€¢Á˜òö)e©qûsEjŸ~ìÂ}‡8¸–;#SÀ[I ÏFÑÅ|t?÷eÖ :b¯»kÁnÉ´´ï~„^€­Ð_õ뤗âç§æ)ÿQúùIù= L¿WAÿ´¸Äó¸°±~3þ4rì¶ÞXÀt¯Ø‚³#¹r™mê ÙâU¡3D±­ñÉòW5\~+@õ6hû”er¼/ª µá}q•³¢ºïCÊK}¹mÖîk’Õ÷ÝtP«TϹ~=Ãõa°n25µ”dÃ&±7ß¾­êg®Ý'Ä=dV›Z#t–ô6몶|&·/ãzÉ›71ŽÝOØŸ¨7®ÓmƒöLjõI6ýÄdp7ÜaIŸó(ÄŒëRò¾Ùߟ>Ê H £Žöv%û†)¼®'˜Qט¥£4o¯Tðw#øa ã¶8ëígŽø »ªÊªð"Á#cj7…Œ@tévÉRk”¾§:²}ÝÔ¸µ©nG®rƒZiTÏÂúŽ­#ñî"×M\µóë80Ê×Nj<æhÆÃÖdzKÔÛ¥Ï;c**jÞ¸.ÉpŒímëqŸæJÙîýÏíe1TõÐØÑ@¿r…^ÕG .1Ö(Œë :fÝ©¼´¼ÙsXÆ›8úûrÌüÅKCqªµØ„©¶a)ÉrWz{ßÝÈ+ÑT”ê¸'1ÜÞ*)"-¡ó­'wg‚¼™üØ~­ãý(Ê––u’ê"kÉÖ%“¤|œG3*‡|u ¦EÅÚYöìbÑñÆ*¨3×z–#ÐðÍåà]þ ýeîEZ<Ç>lëRÙQ š%€rp$ðv”ÿC…à3§ú3zÌáj‹þ¤AFFR íÜÞ¨äï_~_ÅÔ¶e$d†ïbZMÍâfC÷8Úî}… D`=`fhãÙ^Û¶;R5lìFÚSCå%¹ñ ¦Ù)ƒòìÁX¬ù³Šµ:¿Q3õ,ìGëKJ¢g`Iïü->ó­A*Jø•× Y‡ÒtêU’ó³Ä&]7½Ï¹!ȆìF×Ï0…,¢ ¨Îß_Êq1ùSø@H­DfÝÑhNˆL¼†^­aù3õi –P2ö±Ü”  ¦dzf‚fx#~Ý⪴÷“שÃïöõÃoÛG퓆“¸±Ð7§94Þ¦x›Ô㓽R«Ö™åä‡Ï¦s(ß;C%9œõ VY6g–N?š5bI'©QÏS•ÄMZ¨Û“|VÙ‘„mÍ1’ì¯éW„%†a=È)êE»ëS¸ÀøvWýµÄtigT:ÔZÍdLé;.¬Ó³«jmG{K¥ ]¦õŠ ±­ã`;¨õCÿÉKµ d…3ò²)×ç\%Ìq¼ìö—–èîi‰^å†Þ¨úùÉ|Ш'>†YÙêÍPÄEÛsçWÓ¢ dAë9á7‡ýMÔ0åµW“TG‰ûÍÞµø.„$Ú6`(8ØÍh‹_ü¼Hù ÕȪˎ¢Q¸ï*ÜúÛ†–×ã¦FæV»üAÌd¡sPéåñûn@¿qxèŠ-² J†VèÛÈóÂršÿÁÕƒÍVÓÀÇý'¶°Ñ‹Õ¹¯g i"èaÍ·IÎýу&ìK\ —¡Ÿ? ~-8d°y¥¿!q*¸"i~ܘŽÃªž0¬‚œïüpŸì@RiDc^ ãòj·AÎ+ƒ@LßúþÕ`G;zV¹°{åwÀ7ã²ÈÌ£ØTR(OÓ˜ÊOÓÙHGt¤¡(Ý›Öò× 7¼ºP~ 1»Æ­²ÄùµÅáÊ J4úÎ5y“f¡GÌ´„Êæ‹v!0_8íá›ï®GqpšIBš¯w.jÚõ¹•þµm7ûXd…C.²ê@³±ŸÚó§Â¯ŠÓQ=´„3²NÉôˆÓåc#~6@‚‚MÔ^ÿŸh9ríJÓ·EßÒt¯Ð™õ«Î¶™,·*|¬ Ë=rŒ@ˆ¿aH‰M×0ƒWRctd«°¾þØуÞôǧªlÚ®á!n…GÍ”w¹˜6¯x¿°Ó"üä…îØÀæH¬€±p€˜îœ…¥©â",æYV¢ÔÝ#¹ví<Œ·bstçO:ø)Û$;áÐÑIÖtO,\lÓq.»ii£:Å,ù²(â8öÂÎÊiBke=+v@âSN›Y1¨…5vãhð° âÚs}cy{MÕ•DŠZYÕØd5<Äòñس®õþ3%ºf>„„àw_yZ™×\0»/Öm†8Íü[ó±Uy1¾Nu¥eï.‚¼œYׯ¹¼´!)–]¡n<ÛÜǾ ìÅ,zìEƦOÈ43C0z¦‚ˆ 15Ü‘ ΊÊ‹dÉ÷6ª6û’WkÍêIYÌÝa5mD×'‡ÃŽ:Þ#_—,–hð=ÃÇPTd0Mã!‡UzN+ã“Ô@wLG)ÅáÍ“û¹—Èû©â -½ÚáOthô¢»ððXCDjôË Ò}¾ÊT¾s?üÇÞóúîMå Ù>•{±RN ºZé\SÆg#Ôì±QcZ;‡]ÖšéÜzž:²¼L‰¯÷²Zˆ~ò豤 {j!ðœ~0Rõ—€¿b …·÷¼-´âÙˆ˜Có …¢i_D5Öå˜UŒµ6*L~ ¾ ŸÐv¬m©†,S²ñþµŸ¯J{I7ã›O|xbý\4c¼TŒ?òŠ8X²<î÷£®îÜÒ~䀨眵pj£=N6h!ÚuZþW•øE1¨¦‹øsPFpà'k•MS•º‚Òm‹…Û¶”Ã\†ÖÌIÌçFŒšV„¬NëÄz~ý{!6Ï}ñ¤–[2áŒÀ}µ$#3~ ¡é^m!æj% ¹±ßVP±+qq/å ñôÄ2­vßI‚û©µíÛ¬šp3y/æÉÛ<ÌÀÑ‚”pNÒM‘n¾õ&<ÌL·?7{@Ρ©ó0ÿä5Ü…7 ý^¿FÙ1DB+0¢ñiõ%i)!MÐ!¨×FÐXøsůNVˆ1©`ª³pWj —EbDbÍ^ø+| +‹«‚˜Ô$2Ù!OzúÏê²ñÓÓI5I|±éŠÒwKNe<0ð*å¯2s{Ô»’“É·äjG|îBÊ’<™ÖµÊiåþ | “])þÅfûz­ºg¿) ÓQȯAü=˜k]4”ªíHQ.°|âûò_±cµîâ*³K8s¯Ä²iO˜œuW¸·_—ÅŠCÖ'O¦ç:¹¶wÖRœrš*¾ì·¼ªwntÁéýþÐÑï-[ǯí;£NŸN9+P¿NŽf“aÆbl¼C0=­jvE÷ŽÖzûS§W¢·“â°ÔÛ¢©#SvîPÔ÷œS¼†Wʳç<ì£å3½eÍ*"Š:bˆÚêè¹é¾é°,:^ÇÃS°Ü… Nâ›n̽’Eýê¯@¬¯Iáüt•?K—ž¸4!4x‚™/e zHcimˆ¡¥ë?O­á­b\ÂÚYø:½Éj-c¸:ñïÎ7äèc0”Ãy© *T–tU<ÍÊͤéÜ2†€öÔúò̼bÐëXM2ˆÌSëô‡ÓPM¨ ™úbk ;.ËíªUI,µ£ê; ÊAv?æËi"¾ÂadÇR»ZS1ŠØ÷ùv‹öôêLÑ‚pëb-tH›ÙÙÞ»&½oV¼iÏtk, E)OýÅÛØ–‚ wœ«u‡¦ßË#0=ò×Ôü ØëäV{ß~µ}“zúÒÛi‘3–°ìØ:[êm>Ðóù!Ì mtÚ/`±D’æâ®c4í—E|]ù¹ßà£fW¡IÔñ |PSæ\óiO}.çÇŠFnÖ m…û(tF1§þŠH^“>„ïw`U©ñ¦~C!N1œ-Z“9oÈÁšÛYîU3ùH.T"»eQ>Fš‚ˆcË|â)"šqotr6ˆPÍjðè&K\¨e“ùÌÃYÌ\´°Í" ôúà®/¤ø_÷K?Ħ˜:Z ìÍŒK¦ŸFîX&Œã_7]G›ä‰LYl E^—lºè–Š~3ˆ)Zy,ϧ` Úþ^qB¤£,m“yñZ¥¦i•gÂÅ«)œ«À¯{MÀ}„ëõØ®@=H—åŠRÇ`iXT&Ñì ?* ºQjá»GÑîOYŽ+s­ÏK›fxæ‰Úµpìé>¥jvoKŒvìèÚLlÍÿ>»Œ<Þýº4ÌÄŒG kI;¿v1ñ îEµ¾ü‹fÜJåJ–‚?Ã/·WȸCŸ@'ñÚ›ãàS²©”œ¬ÌÓ1‘€’»Éd7$†¢g‚îu¯P¸‹µ½KܦÈÜ1,)2<0]°´Q¿‹9‹)†Ktíë™>C¸æ Ê‹òíURóvÚ>YU0EÝNÄwå½ŸØ ˜ìŒ¦06#¥lÌ8G°,èˆúQsýnä YTkÓ Ê9„ZÉÅ=öÓ qAßË>Ö#‡Û\[º…Á’ÍY2Î÷Tž#—ÎxC0Bé¡@ôD‚ÆÑ‹q.œzÅlr®Zl4->„ ƒyñäȼghcɉØH¹6?ÍÚ´1ßû[çU¿ñ‘¦UK”k=#´®á~~U…¼ê‹© †]!“â4[œS=,«W– ÌÛüIWºŽÖ¸)éÍѸ»þ™:Ÿñä“=âüòzj©o´§ê‡SY£l#~øB´F`)Oš– ëo‡n˜fÌï[í_­Ÿ.·Š‰zf†‚ñ#ŽC¤'‹½bT>¨ˆe|Q»…T^‹ß\·¹ÊpwùD>5Þºèóë|á°Û*s834X.äR_˜¸éG£N¶/>‘{nÚåëØ-jüïútø,^~l×öœH48rŽš.§uèïµÀìZô™I…‚¼Žƒ5g—ˆÊÛûärj¼Ü> |skÛhN ÉÍØê«~‰_ë¾Â«Àõ[]> ‘ñæ“{ˆ#Eë  "ÙÍ>÷br UŠžù–É|Œõ%ëå™îY•?`cÇ3“ŽrYþ% âð§ïOÆitÙísŸ¾ïcѺäwþÖçŠo¡BÎéÖ÷I¤UIÙ +zXöè=*K¥q~¡˜"íRGw ¬zœõö¯ÄãËÝZüê–aEÜOú¹ nÛ0$W¾óédâ\§‘ô*º4­å^Ó{g™A•×á.w:ò†ó)ãZÂK‘wç$¯—'dxémÛ®ñ˜öU³ã£#¢z¼Ò'+Ÿ»… qägÙ"y³(±™‡ß‹¿Ít±ed°èÁTI޻ʷF4E–áv‘ |y ñ³Æ«~‡:$H—nÜ®ù)æÁË—™™Û× v v‰øâ¾æwžt@†æS »*.÷á𢳩lj¡þ ?Ó/Âð1$©ôݬÄß3³|°„Mfö.kŸz€7–úY‹¯…ßÚ>j’3–ð8~Ömê#³Èd¢PªÕO!`”ãŽiM-y$O–õ5‡$÷ög…xWWÄÜ*£ê,ŠòGÓù©fchÂ$IJò‰’™y9$¼Œj# –Ù«ïÔÃM>Pq?Ø_`M¦î“è’¬ßTÎߡ˽kZÀ.Âr’\ã3V$À jkN\ϾÎ|ÏBœõûóø– ø悆qRq5|ó9jÍ>?˜ÚÆødÄÀÓÂKuRŬlhˆ"=Ù¥-ñÕÉù÷.Ó‡P¾Š¾ödÜæÑ5š ¡Þ±¬<>òRóCívÇÅXB_7µŽ$&?{îDEÇA)óhQÒÉ»ÔËMÅl“Ó5Rzx„rÊ_nÐJO"(Ü…›)K aR“ÂŽ¸}W?hmçÎß²‚—÷ͯ̚ö¨ê5"I‰]ñ±XmŒ8{*S‡ûJ°nxÜõâzÓ‡Ùww¥‘oxŠ$ã,:LÞ#XsUæÁú^*ÜtüPÉ™šÈm‘Ñ#Ò'–ž YBgIŽõ7‰h„£JÚ½~íŽÇbU&‰×KûŒÒý¥í«è¯k×S x­áw¥ 8&Íx\]äþ¶K£ëtâ”Sêa\éP8(/%è)J¾Ô›yIÏP‘.Ö©¡k9(åprTyý;shœ—™@þϵœjj÷³fã®+P*dÒfÒÏ]²BM¿óWT˨Y3{üö©tºªEÞQ¾Ê[Éøü"¹u)]]ˆÛËSøÑx}ŽNU¡'g“kŠ†±g’ý ¹fÓÁJ+~åü\F.ãÌDÚ/¥è¥pŸ0jIÊ£ï›CüUÃÉ”S-̬¾@^{Çç·ãç®ok,¡ç?‰¦ÓmÌã5†QF"¯‚ð]´×PÞb84¤ [Æ„ë¨ãóPk5‹MŒLßùpŸ5J<[O¼Ù¹`l$ÊA?ÛD±SÆ‹/žÓv°ëôÄY ÔÏÙOw7¶K+y/™ã¤Š×oY‹Á ŒÛI "ß,(A,„ìËøNoV“p{ÐÏЋX F¨9Yûu¸=>pÛ¸9ÛJî¤ÅÃÆOu_7 ‡é0¾§DqârùïU©øe‚`&Ç\æûÇˤz[Œn±oiÉf·ªï]«Z¢«oa¡»E¾C³~[è³'n‚šÜIÔëEî™O\ ½IîQJ/À»kwe`µvæLëm]^¹°¸+9p¢Nÿ`aÀgi9j{ïd¶ž$d˜-‡e~„°òéÀч“#ͧ¢G~ìÚäàE´(×[ôYù†ßZ2s–ËLãEØ€dÙ€GŒäX/¢ÖAîí«Æ9x Ï€+ýD÷vôÿìRT¹¦ ‡~m‚=¥Ž¦•1´ðÓ¹A ÕÈ•]ƽˆçû Î&éoªÛMMßa®Äò"‘¥Æ¤¢à„ÛV‰¥qò`zû餻zÈ4¥õðëj¼ F_ŽÁ[kíˆø_æ–Ñ»¨ýÂ"_-¼OXnµ®CW™®Ñ&Ù[’ö~– n í¨õQèÄ;ÙÞ´³Å¹Å uEÔPÉ$× 'à§ÝÐ÷!ˈ¾í÷`ò'¥ø}áE—#^n i˜€xµ´ÂOÇÝY'•'ùtüɾ½¤¾´·.Z{?Ævæ™ë…(õ6Å4,–[eºÜGI8ó„Ódº²eš gíso·Ñ/@ŸÅ@ÄCé!ÚÌ.³Ý¬Øf_VÃB3¥ô*|§0q\׫2.?½f¼ ÊÍÂ…ª-ÌSg`É…ÛDyPbKñw˜µ}¼($߆õL€2Çøb|áxª‘—ˆZNž+-/L¥ŸÎÞø¿ŸÅ1­ «8å<:®t0Ò½ð𛊘âƒÔŽl!¯\®v½çd"þÇ°*—AOjºœ¯Áuñ…Æ‚ÊUÑŠ_ð¸9$kMqô §MæK˘ìÔ?ïesÓ4)cß|Ì„„fRÏï:bú@þ-B_X¨’ÿG}+aU%Õñ8ô)¾#âΗìó¥™ —mRÄçRUTRI²*Õ¯^âéù‡ÙðÜñ-Š¦–Ú'nŸ¦”ñ;mÓ2W|Åï5G"g=…;vå{ù,wE®plȾ¥pHï.´1bl”neà¯}T)y^Ûv3ךœ“±¼ï”aTA;ËZ_º÷ó>g¶³–E|/CR‹<©$¸ï÷(.…HtóÁ…Ä÷­ãÜ¡©Ðg”½©u B˜°EŒ“¾Q‘%Q_yõ2A^nÄáîñyŽJ¬,­óRQüÂ!JÇNÌ‘…`‘TeV¢jÉdÁ#ÞÊgŸïò$•ÔP"2m ÛrÆYlõ@Ê)œ/eº†y8Kzn4Y§FM•†žÃ˜»§c*]‚¹Xf6l„aÆ~l,Ü "äé(k Qõ–pþïc¸%òwŠm Ê~oŠ‘4ÚæˆVßbÒE—­ö³ ³¥0r<ÊVQi+±HÓóuæ^¹l#qº\!‚*µ-òhXZ”Q·eÖéó^~íjä8â¡ÀÄÅ¢2H@xΞQÍŽ) aÊë½TÂÍÌ^å餢 Sê]–…t~âÇOoÉ5Ì*Î "E—û£?›ål†*Ÿî|ÏȨ6f) Ž"Öáˆì7"”ë*p’Èä}Š»dá×»ZaÞÀ„˜b®ÛE$Ä—Õ”9‡ [ Å€½’”ySæB ç!ˆqš$§õ?ï¶ÔιJà–Ü âBûŠzYjÂùµ?¦p€ì¨^E öa܀߂Ÿàžê¶øfÜ)¸M•I˜{WªµLË¥ݺÐyêÓc™§·èðv!V.ò^™,š»ï„ÚŤG£ø“Ð*«î¾³ñí³ }¸=nìþz#4] âèÂxU™´I¾VȃÓdÌÏääMZ­Ö¸ÊFn…LÂÂXnNï•Tf{ø˜e]:ƒPŸÜìE©¾ëû2„.*¶žCÝ>ÂÏß‹B'_×f ê|7"¢¾ÄM¨üuO7¼{SŠZÙÇTäîʲ.у¿ŒŸq,Iò‘Ü3;Ö‡~=ó'# |ö{7üd5]ÅIâý(qbÀcm¿(#†%ÒçúB·€»Œ‡w4ôÂ+ñ‹ÈVP÷Kê éÂîÞž«@7!1í„¥¶ó{˜­Ÿ¶â|Å ¥)¬ë¥óQy4S‰•.ÝOL*½?ßž©è(f3ÓyŽép©7OçÀI{º.ÍÁ¶Ëó[4~º¸ðm¯õSÈqQiC%‘Úi­R:v^=`vQøi‹xU‡¹¦k¡\Ö$ðëî5â¶Ò…+~dÖßâ·Î4çï–šh[Þ‡ôE8Š#¹sŒÅÌË ö™éRÁÓù©Ô"£?¢ƒ½|ÃJúŸia ŸÇ)‚õ´iµûãÕpçï_Eô“¦wh‰]ƒ§ W(놜¿ë&B´—É%Öåµ WÖ˜Uí:ËÖ1ž=¨Õ¹O`ª÷‰"žãɪñƒˆ˜\Õr~‡íª÷R„ Ä4ö@à«ïøÜÁD)DÆ(L8ÉòR¬wÉÿ€ q;õ¹ÿ‚”Ì,¢ µæž+köZ€¥ÇU™rýxb,{ ']—Aü¡÷T})h²1q¼ }{Îw#º{zW àŠa³ ×I""ñW¨ø‘aî`Ós² ®óGËgç€ù²¦G³ú`Ø°Ïš^Ö±~ \ÍL¼yxÖÚ^On»c÷[4S£X BBîOJ9H‡»ÇáQçŽÍOŽýjDQ’HqA ¦3¹æמ.ãs-Vrª#w  ËÁúE6FXB¾—&9ç§MOý7³3õ³v±~)ê;{ïç‚^>Ç!¥:ë€ùÊÄ®ÙÄÌ<*š%QMÜËæÅñò ’TEîâ²=ÒÇŸ. ‰évÝöÖt¾ÉFþeÝCÙ¾(&y‘ 0&É)BnoÓòïvð}_c± ¾]ÇE¸vöÉ¢Ø õÞVcT“?/°t€¹K‡Å=—@ |G:XÜÏAã ºå¶: ”–Œccìß*në'pŦ¢;Rä‡t8†á”ço§Jnuܬ…È鹂°ªN€Ýî¾&wdÈ0e?]]|à§5âÒP‡…Š+ƒÕŸ¡c%UK=Dj®ó”ø–Ï[’Íè>¢uÅí{C²•Û…½˜Ïõè„(ó壗ùŽÎö´ú!\(¨^¹<}g#\â4³²¤V˜e@á>ŸE È~ûä@†vþAŒÏtõ!D×;®îøð&=¦…sCNREKÇ÷tñ«j=UÁnª´ñdüš,ˆ·=çhCi–yczæƒÓ–X®ú§Ùé<þz/$TSYEW¨ñîù2v÷cÎÇY „ã¥z†™'KE ŸZä÷lZÉfr‹mÕýêèOƒ]nó37¹ØºIÕk<ýÑ.hï0¾E©YB¼ý°r\—¸DFáD¡Î ó( ¢°?ªÿ9n’E a6£’V,ŒÍÍˆë ¢‘âȸð™+Í…UJƒãš=x‹+«t £îТ0(d–ÝdtUغ¡0YKæ#‰òáÞy`)ì r ù‘øÅ •ÀðâœþÍÐ@v* zÍz«*—YÁªÃgŽrFñõ X(,æ”Bjwz‚Ð ^ö›'GãóTòµ ž‹D”$¤dÙÇζÓUD¢JÕÊPR=ŒÏÓ©!eÍ5šQ4,±%\¹ÎùÂO»€† ?úJ^gß|û侃é,Dª³âξ}«y&f”°Ñ‰ä’#ƒ4\% šóðÝÇ·‰©@;Uf®FÜ&–Gm“!¥2âÓõ™ @ëj³®’wú!Ï –¥ÿ&Í*ñ6‡¯zÏ~™,ÚÅ„ÈãVw–ý5{ÅSè‹ZÊÇñ ê&Ê’r,ë#ì³â¯cj“nÏ*g#¾Ÿyvé—×yÐx‰ëhÐœ„w‚·9+nÞ%ûAÚÝgÏ#bá Àcœ³wW„ÿ ðSì ÄäôN—¡»#ó,Ö’ÅW[Ž±Ê¼£6]ãPv±â †òÃËw(ôšZ“Ž耊žßê©J’]{CÈGÿéŒ#œŸèšS)=5“³ãîºÆÙ¾=›L£gN³ømoe˜$o™€ è_3îhÍì¼,—ã8)ÚjÀmÓ aÈ©BÖ«’[bgfÁÕÇä|à:Hr^Ùñ0±™çãT8åMºVjâl=ÉÚ¤ZÄíˆKùÕT 0·{ í(ñwÑÚþ±ÒQñO ¬Q_áDõÕ=dk^ÏoiC™6·/Ú(¡Yâ­3wÕžâPªËÓ}“/ }ïîÛ,ïåï//ªz:0±­J^”.yUë&Ú5'WöÈÓû…c–6ÀÔ=‘¥ÜݤßûÆèÄËO’õõRCØl3ÇðU÷«óapdi«‘)ðà:ìVZ·™U–—‘ß­Íô ªmnP‹G¾~¾[I7Ù¡ÑÐfªŸ ë5ˆæ>8ÿm:Åžáï_¬VKÆJã»<.œá}›•x‘DŠ?šé?~‚ï¬ú!nÅç‚XÊÅŠo\“ïÖþÂÈ!ôg:ŠjÜF]@;SôGì)ÉÌÏ ±õ{ÒBôÅ¢5ZÌ8~•Ëf4·âè¯5"ézq§ï+EožwÆ<>©MQ´êI|ÐK6Í祕X…äAjV«oPˆdvºù6ÿ¶éÛ§¤ÿ!r&W¶t¤µ"õrÓ€A›‡Ê¼Í 0ç’…µw2ÿE}ñ¦2ßH5ÖcÏ4åxŽªÄ%¸‹ŒÔ~wë÷ŒzÑÅÂO’ztˆÈ³lr~èb¿Î™‚– ¯ìPàÕù¯Òüò8[J3q•Ý{¿Ó:1§Å§ñ§Œt°ZÛÑÀ{ ÐÒÊúYi+äsö:ÃÄk¼ný“T¹ê¬GV†KþØáˆÇÉQo-¿ø+ñôcºØí‚ÿƒµ:ÍÆO±0Æ* Òe=ù4¥ýiÍl#RÆÆ–UÜÕRÕzîíÓË›ñ K’—÷@‰q(AÉ93Xú—O9Iƒˆâ*ÌH)s‡êÞÚù‚Bd „ß±ƒ=m|Ô¹O5ÓáGæí="°¾Ó ÉV.·_­ÞCCÏ9¾ÃúøÙäáêÞw™BQ~¨u†¸å“hxÖYZ…9îâÑîDiUe‡ËZ¿°|öJ$3-ètôfÃóØ.=%ËÜúfŒOµ}h̘—¦Ë©ÝTòU|z«¤âœ­vpV‰E)¹Íá‰gÕaíqZ1{5u*R‹®Ss!ÆbõÏ& Û®?øè\:Ý;‘¤®5$ÃRñ}òþ¤Šx99ý®c© $RL×b&îww E úà õYTÕeùH$’ü×È›ž·*=ïŠàj m²‚e²¬ÀvɆuúK²JE î×Ç"v87ÆÌÉ‚qæ:¼ táˆÏf==¤™*nl•c{ ÃH¢S’&ÕkÚ¤áwÕ~þ9¼Ë0eøãwi^ej#Uãåì©­u“¡Ü¾h;û¼G.«+æ†y­ ÷y.?>VuE‡Ç£L”AúbK?Q2»_PÆQ$»AÐ4kZáÏ\RÅßïÞ96‘ $—)©Þ"c²†7zÔÉø¦ùó´êÞ/H‚øR SxµC˜©‡pŠao–d‡í¿GFü~eÌæCܽåéæûÑêó¯o{îpg@ÍÅœ¹Ò?›e4=Ó–üŠ4Æ]|É°u¯.=XÇÁ½åãÑéEønž43Vÿ1¨×£s_dôaìƒ,}6‰Šè’¥”q»ëNy‡v‰ÛzASp~à/äž ¢Î'Èhþ£ë*Ücƒð³joÆk›¹&Õ©~›áÀ|Q”„³YÙC„‹‹!iU‹®TúY#Æöpy‡:rS¢‰OZvΦäPb©‚uò'Ö\šá—tÈD“~È}wf“Ø3€G¸ágpz×Î{’²’ÂEß~EhÊ^>é«Ö!:NÑûŠëªš—ói¸y1RzÜÕOo’ô˜œØBgE ï¨ ¿c:¯Ô<©8Í4"V †~ «¤. E¾QÉsn¿,ÕwœÔRb?Éÿßʉä} EnW6üÊšÒP©êžÈŒ¶Á£L¢º›‹ÛªgnÁ +£ýõ-ˆ\'ß9\ƒô~´‚äc(ë˜Á2 õ«Ò<µÖ4ëɽN7léã4thb¥SÏáœ0Q&d¾å´ýµkuMX.o§¼ÛbÅ…Õ‘nPÖ´ vŽ·sgj[D`7ë“+3‰oØÀW— U9¦AÎ|§VK#«íX±e¶ ¾dñ•ƒöÅf9…£m¦D®è¾j×U/ÂXYвŸúJüøyæWáma´)=¬&‰÷›¼ØnãÝÕ LÍÏô|k5‹é’ÔúÞ¤‚û+ßÉ^õµâ 0ÝøV1 ]ŒøÅNÎæ=ä|M˜~9õÑèèQnÅ¡ˆ"q> stream xÚµZYo9~ׯàã.ï  àØŽãIâ8±’8̃bwlíèptä˜_¿UMR}©ÛnY ØjöÁ:¾úXE²[p¦#‚3K×ØðÄ( Ή’À¿Ã–&\†›–pí8¶<áÖcO§Ra!‰ÐZ 77xÍ)öžH#á.9‘Î`_8U<\ÓD)‹Ú$˜dYÞòDyP$¸âD ƒ+I4\Jô"–£é`¶%FkC¼wÄX&ˆáÄ8Éà)b¼{™`$[b­'V€}ŽIbÑ5G¬²b€nZ—¬ÂÛpâ 蹎å¨ÀCŽ£D ýA…"D(-àIFœÆÿà5Þç¤80Éy¸e%‰9èXs°Œxæï‰× ÂI¼A¡ p€°GH‡‡=Xè!6œ1îÞ‚, C[˜G95ð€(ãU%á—3&ÀÆ1¸xË(ýħ4ÜŸuF€w~˜7è¨òØךP„sî ‡(F€EÜÛ\¢G(P ?B°ècò¢c1ì¹0ë0î…9Ñ(Á ˆžÏ[Î ‹ ÀÔ—“È@ð8xpY´WÀƒÚ…@ ƒ=Ð0c°‡òH ì¡áסŸAmÂj10h#°U‹°l$öðtH$?S`Ð € b*d”‡–C²!JH%† ÞåØC9ÐÁ°‡F[ö0h챇,´Ç¡ÑséÐ=—ïæH3J£çÀ-èV=W‚Áo¿ èð×}Fè‹ùlu”-¯ãûÕ|1ÈÏÏFS¸óúäüøôê?‡ož_r7&£Û%Qá‰çÏç?ÉŸÏ´$Ï„Á$V³¿ô`yÍVxòptÿ2ßÞÁ©3ŠjðÞ3Ž7OW£Éøú`v;ÉЋU6ýˆ ÐËØI)”q7Z\d+ò/z@ŸÓCzDé zB_ÒSú}E_Ó7ôŒ¾¥çô=½ Cú~¤Ÿèˆ~¡×ô†f4E³éÍhyG¿Ò¯ð7†¿ïý:¡_çë½¥wôî×ý]6£cú_ú7Ð)ÑÙx–Ñ9Ãï=½-²Ù$ûº ­E.ô>[Œç7ôý¶Î–«ñ|FtI—Ùw´ÿ¤Ë ê\ÑÕÝ"ËèêÇœ®éwúƒþ¤¿è?ôŸl1ÿw@óÅ0€´*!;üþû£Âsvuðáì„çø’³íá Âã!¥AþƒÜgEÅÊáa¥àÆZ‚£l)6’—cs¿˜ß¬¯W«ìçŠ.×Óéá¸/ï'£_5'M'ÏŽ†o†àä›Ó6'¥@bƒ¤ÑÍAÙÅÁg\%G!¹·9ZÐmH/éçH´ë9¸œÓíë†C“l¹ƒ¦Ùí(‘å~H’Sãg ך“óƒ³áe€¦exBŠÐ³'hŒnç@îaÕ%(V=\¯Ž^žä.é­A a‰ÙC Ûº¡Ëš7¢7ç\}|—{c·Ç'…‡C^Þ7Ž?ÀÜ"1~Íy:CFF>ÖœU}œýôöäÅW¹³¾%ùDF¨G{§ªÕÛ3ôl!{¥Ÿáç“×W¯Á«÷mÙG±ä{#7IÖøVŸ*€ï*%0d§Ñr(,ÿ†°YŒ®ÿÎVy‹íPÇ0M zY¬R4³Y8|[&4ûy=Mc!Í+é~‹éýd½,WÔoëù*»ù2ÉO'¡G~V\CžŽƒG»×bÞ+¿ùáåÛSä‰né‰&8ÿï¦ {M\;Mdp°ê蕈Ï?_ž¼AZ2—H‰¦Ô{¡½m÷'/° )–“h+…;E€›A¢ßÜcøîå‚Ð’Ñä&£I³TÇØìÔ·1îÞÿ·¡üíÁÁ[²»TÑ«L]¾~‚˜^¼nËèFl¨-ŸT§ü#¸]ëi# ÍÉ%ÄèÉa©"¿ªb\·WµiÍnËâ4øžò"¬PïòYP¶™ÅAñe ù &EãÅõ$»žßÿZ@ªœdS0*_Ó…r4Ï ØLדÕø~ò+BŠ†g‹  ÝŒ¯G“°½ªÏÉûƒóã£ضՃ\ó)ø±Nì׎• Þo÷²W>>¼¸xõé0÷²cÙàµË7 =v^ÊV/—ãéx2ZÔ\ê•OŽ/Á¥aû–…Þ”µ¯u¹ë¬1E…9ƒœ& ŸJÛÕùAÚ&k)5+þ)½lŠw Ô~›Çç¯>æ ú‡0Ýß6P{–9¡©¢ªÜG‡¿WÝT½²éËᇳóœ;­uE…²¢å“ÊÊnËÍ®)çN:§ýåhvSL?G+\yBóº/£EKÖí½8­-G7˃[z»ÈF ¾IÞ|K.Ÿ,­§_ÀÒñí¬ß¢4Ì`w]o®g7 ôz¾È:g´JôÛx9;>æ”i›tEÆàk ïžÌ˜ÎEJ_Æ›QÍ%G#Ý”‚Q]3´ÂÜrK™/×_Vù)^÷ž–Þ!ôàrøö3&¡çù଄·oCÜè‹ñb¹BPˆ†î¯Gñ„ 8û4¾YÝ-‰`>><œ˜yn2ÂûTCÓ°HÕ-âU‹TaglÕ_J4,²u‹\Ù"çJY×_ý}AC¿¯ëº‚+ÇÈô7 ¾+_7€óI\É]ÀÛþúk[è õ²Á|ÝSÀu9;PÛõn »PÑï|ýµè†~Û‰?çe¸PÛ2nè÷’Ìû= ÉêÞdÝÑà#® >–á°;ÀQÝIlhoÐQ•Ùh*Sî’Ê[x íºO,ö‘ëÛO ‹ìÄ$ÒVBÔ>,ªîÙ4 jÐÕÊÊpñe~è=ÔÜ\©Û$y'ij65@ÚFõMŽ†E W å¨íPÕj{ õúñêÝêk‹ô†úiqàlôWF‘ÛaÚU_P7ô?”R+ð78*v±¨²­¤µƒx¹È;»ƒêëƆE ‚ÊÊ0åªëù†LmYÒ0¨AY%[cæô# :ÝfKXŒÌ׸´Àù2ùfyïWã›%ù3E‹|¢m8 …ð0)'"¼,ûk'&0ŒXaÛØÜšˆG»»k‚Å6”d‚ŸÇ…£GÏà b7%.LÖ‰ _À1w¡n=sò ž8…«(\Gá& 7ñzðtG%.*‰öa­GŽá«âÃˤG+±e%>¤aâU®ÒyðÀë`n…ï®ÄD‹m<ºèQ ¼÷Þì®? ‚E ࣩáb#ÌÝ°aŸ¢J&1¶ü3®ÐÐ"5Ò£ž¢Ê& qÔpæ’W.]ñé™0ŒwTÓ¸œð±+€ðz16ú RWUiÇcNÅ%Oˆ~6»ªr‘ ‚EÓOùTÄX ™J?A•Ð%a“*—qs™X*™}‚*ÉYª1D2EFÊtKF$eØœÛU•J'…Hj“Ña™b%-{Š*›Tù¤ÊÇ+ŠE?Uò\ñ~ãÊWT©Äh•pS"]Il0 @ûp±íª|¬‚= Ž,›¬Y5äRŒÕÒÈXàx¬E±ðù§(ñ2¥j•’ijÈ”mÒ-üÐþ îÄ)xø4? “¸ÌÓhQ‰`îA¯\»*%Ò\Éò‡Ä@ÞIýREÞ¤Ê c7A®Iz»^MÆ3–ÏaRg8IÌ¿úÏÏ¢&µé<^M2¸·Wˆ‰›H%b×óEöWÝã Ýã*)ï®»»Ç™õYö³°ª&LÂx§0ë˶[Æ6Ââî\›0]&·‹:¹0Y¶é Ûæ—-06¾«·±eKâ.AU–)7ºSVï­v™oÃ·Ë ½Í6€M°~`]1¥¼ -ì lM;#­ÏêlÕE´ìÔ™ö>‚Ní·8 ‹°¨ZXdE–Úæ¾*¡tgïJ ô¶@¨"ŠwÉ’•°l·«‹´²*QQÛh/ ¸¥Ü.+ô–Û𕾢3µèjnкBYIgº’ΞéšNQDET£²Q¼çÛâ Š8ˆZx;¾Í_Q Ïmgï òbò¼@žËí²Bï­–ðyæ»P`•„SÙT-Ì À§è¼¬©dð¬xV‰õÖ0°" ŒwÈÒ¾2ضáÀ6AÑÞvʪ…m Šö¦%»d¹2¬ÉÊš¬Mˆ´«‡h‹º(´ÚµàµUFNå?¢®NžñºU(é,À*íÇáÏ+:“û‘8¶¢ÓÖuªÕé’ªf[`) ³®æô|ì½­ê¢0ëZa®õ®f]æÿ.{:p endstream endobj 2152 0 obj << /Type /ObjStm /N 100 /First 898 /Length 2278 /Filter /FlateDecode >> stream xÚ}XËŽ$·¼÷WÔQsXˆÌL¾A€!A–[ä½yuk[Àû0ffùïYE;Yì:HMÉÈŒŒ$£{CL‹[Bä%Å%„¼ø@ø”…(ãÓ-T°,qðI‹d¿Öµ„OY"3>Ý’\¹ŠKJX'Z²`Ý祸‚OY¼#ÁÀ-Þ; ºˆ†t´xžÄfl”‚ÍŒ¨‚ÿ¼ºHÆæ€D%cs^$asòlNHI6g`HÄæq<ÒBŽ‘9X‘÷H]iyœåb1b00#eA*ö€… £€v”@I@²FÇ„²FGÞTPAvÌAv% n´0韑¨02 f È‚²°þ5 , ‚PAŽHŒ‘.'¯ÿ[—Ë…AsñXÇrAYX;*°¦‚Í‚äHSd@Š*È’Q flVr áDaœ”è„Í1'ÍEO°_‹0`ÕÃ+' tIûÁEàà¯Z;Š“²hK¬µ)Ú+îBY›5$ížÀØ“°9:ìIØ“Ö›6¢ØKÐö¡ˆÍ‹@0:¤A =FùB  iFà€`¯Âb³R©—¸õ46¯±¼%lÀ=œ5k°Z#ô±Ëé‚QZÐgQÿ–—„šê¨,)“<Ú,ŵ«jÒŠaDKÊIQôR•¬Í‡ü2.€ŽÂ’ɯ‹Kfv—o¾¹|ýöéõÃeȸ‹¿^¾þ†aþòø|ýôºÞÄmþ|ýs½®ëìçë_XÃÓÙ·ß,éXþ +ä[¬õ9±@¸a¡2gXb°x†:ͱ¶Ó5¯á´ßOK6§õ%¸ÉD¢©P%ðÃÓóËkOó:K5£ï>Λ8„”ÔCÊiH2é»IúÂËa±¥’±XÜEáxŠeD©YX]¦3,oòâ™DÜ%"+Ñ~zâ+êÕ&9=mªÍ³jS¯6¹9Övzš‰ïõõ§õõ¦¾­ÚµÕZš[«Õ4[«ñ²ËàÏe07ÅÏdð]wvSÔeo±d‚åº(n¸ÅÜ7k.×epîä´~¸ÉÄMD‘²‹:Å2¢Ìò’:ͱÖÓ-¯áô^_ɧõÍæ%jÊmí±§¹µG¥ÜÚC†9õr’Lú“»!™;ÖÙK¤ß´n°òL”ÔEIwnJ==“!u’•NëoS±îÒÔJ¶<ë›^L%=A»zqP/úqF8v!¢œž6B¤™± Ý–ñéy^ݧeði[»–×pº N…Я~·´Ì‹·ç¹ Ñ8ßkéîç2ú9Å[ö3 ”nÍ2ZópÚè0³féÖ,£5[,cÍó¼º5ËhÍ–QefÍÒ­YFk°Ì›5³féÖ,t§ÚõôĤ[³ Ö¼7Ä–I{ÆjC´Ä¶†îìÛ–tÏ–Á³wØ-ÈÌô¤{¶Œž=œ6…§Yá»ËhÆ–)ü<¯^øÑŒ)üÌ~¥Û¯¸ó›6'+ƒ72´ï%wîe÷l<Û†dûÀÏ<›»góàÙl¬†‹ÌN‡~šNOßÊÀ3ÏæîÙû&MÑ|“ö-htªcÛ‡yŒ¹ DÅ ´ÙB–IY¨ä~:œNF çgXÒ±¬@ûé +—Ùé] ÊéôôméôŸœX9v,>ź}•æyå^ßTΰR2X“W‰R¯v s¬zzVßÔë›ü™V±˜ ™—dOsm®V¾Ö\a ÙE‰ir ’f2Ä.C<¿öuæ ît¶+èüÙXŠ]®0È%·/…Ùuè^Nƒ—§\q&W7iLzÇÚNÏ3é•LšŒ‘Øb…Û÷ª%vç½¢îÞ4¸÷ÄÜ“™{SwoÜ{À2îM3÷¦îÞ4¸÷ˆed˜¹7u÷¦Á½m×Ñ̽©»7 î}has¢UÅØ÷.ÿK×í›èÞµÙ¢XÇþùñãõeù÷WoüòõþòÕzXüÆd{¥1‘Ð6Á‰Ø†ã•êoÈ÷éãÓkÃÔ¥ßÆ`Þa?yn‡½Î¤Í4zýE®SÖÅÐf¢‹´Oƒ.Æ1ðŠ¿®NBGEà=ZR„Ôf‹‰K›]Ìu¦|÷±pÛ¯9>lËÇÈà•·‡;À*¶0à”w¾`´= +µÈsŽÖ¥c0I­¾`‘¥á¿çÄ}ÿôòûóõõªãŸÑAÿÓáÛçÇO/|~þø¢³w_}ÿãÛ—w‹H¸=ýëõñÛëŸ×O+ÔoÛîïênN‡ÝŸß¿6ÿ«AW‚¤\-¤<°tàG¨í(”IUPRœš-¡È¹PåT‹@ZäC,ÅÔ¥c0èÛ ¢¶i•«Š¤U®ÂÑJöŸ¿ü¤׿?þ÷ÃTsÍKµçCtm>geZ90ˆæzåD‹´‰2­r³¶SM–ƒÞdc)¦®ƒéåà–fÒI-ëÍhÂ2¸–CëÚúCœîʇxÊM&wB¼î¯”õâ× ¯¯@l·¾NÂzëd½Í‚Š«kÇ€Êɵ÷Cò:kñ‹bÕ%‰•`PR®…š²žÖµCÄ ¬\hX+­ZÆ°Òª½”V}Q0[yÕÄšt>´ŒBëâ1¦ò µ’Aiµ«ØXF¥j¼¨¬¼k3Í9ãéeŒ“Û•U£•UhAäÍž^õüýËóýüéÓëõùÇßq!¸ÔBDåj§E¥ÚŠR‡†Ò€º¶fò!o4S endstream endobj 2206 0 obj << /Producer (pdfTeX-1.40.13) /Creator (TeX) /CreationDate (D:20121125074251-05'00') /ModDate (D:20121125074251-05'00') /Trapped /False /PTEX.Fullbanner (This is pdfTeX, Version 3.1415926-2.4-1.40.13 (TeX Live 2012/Debian) kpathsea version 6.1.0) >> endobj 2168 0 obj << /Type /ObjStm /N 38 /First 371 /Length 2481 /Filter /FlateDecode >> stream xÚ¥ËrÛ8òî¯à19¨,‚[S3åµãk’I*±+‡ÍÔ"B27©%)ÛÊ×O7€¦H‚’åÌÅ„~¿‰…)æ S„a ;&qQÌpÁ‚8Iq ‡<¸à¸H‹…¸Jƒ0I 8B!å¬ Í)LÃVðeŠDÐgQŒ+DIbN£ ’1DÄŒ#W‘q:Gn" ’yhpyD B€L 2HCs*çAs”E†AÊì’|Α›Œ§æÔH"¤,“€§ %iÀ¹1„ä ä<R\é¥ Ä»Ê!ˆ›ÓV¥žˆJ}˜yÀåvýMרRØBÈnZã™~¿ b²[Ú–³-…©vΠW®,Ýî6ºéYó:/0Ä#Ûª;IÃg{v»ÒËίªì»c]eÛò&eN«^Äï„ÀóöÞd(wzµ+ÕÚ&ÌE]«]//1¼Äó`³[Ûk¾Ô•‹^µTéìÿæ©Õ`÷¬§êÿ·*«·kûM­yÓU‘aªHç RœÙ´ªì'M/=—!œ7OzÑ+€Cib6Ý˜æ ®·}©k]šâ'ÅÜåp«W¡> }ƒ:…B™ÐÉ|ê#ÁúŒM æSÂIàO=co";Ň)4¬ƒ¼¥²ò…ÓÊP©L¢hxÞ<š»ið:‚Ühòú`ÀdU¯tÛŽêqÜ ±wË›'…yµ#NÙgB¡I¾E,ЬkZ&n*°uiXFNÞÿèR×]§ÜU[Ó&ªGC~MKÁnëíôG34Aô‡ªt65¼ÑÅfshp³CÂÀZÇ•=…¬g@ÂyЋÖöNW„ZB †„QÀ%ÁÛêѸM•»^ 5h•íÔú7Vûy×cq¹mÚjÿpù˜&â(øçíf°¥ò­ª³GUëóÏղŰKø4¾ý{WæØÜRv ¨¬Ê¥wM«Ñiêî‹Ìö^žLˆZ= :D³)rÓ²•ÉQ ¿N÷¶†Ú½°&ínÈïòoµ² º<õŽwùÚ°îû¥Úìê|ußâ,ÀªüÞu w{ÉÛfÎé S|A‘ò®Rú›*  (Ž¿|lεœ æŒ†yʼÞŽb¹ <e”Ø–Æ‚XXéƒôÃ9¤'ä`^pº>w-ÇYÉyBAJͦ°%î2ÈAЈ5¨à® ­·e¾è½7‡£Ê?GÐ7e7$ÑEÉDŒä«jMÅÍDõ’ŠÉ»œ>8òÆŒû˜ÙÓô¦Á¤O¯ §€ã šè.9‰C¶<"câ¿ÁLÛìg9ø.™¢Óoq,û:µ¦2Û¬r™Iú‡~œ)²‹Pô(öczü:pÇ8ì»ùo$lÿbŒùBaÚ{m éÍ«¼²—øÛj“/ð.“ú±þœ)Ž“ôLÛAO´óþ¸â˜%¡3ë‡]?äúqo›Þ Ó¿ÎÓ-ë4²îÕ/xEuC›±é¹K«Üu¡VæF븤ë^ ½6÷ÏlÒ#îY›F?èò¼Ê²~H}}µ¨àª¯ÏñÏ ©6__C,Ð[ÙxÄš, ÝSçÔ@FqJÓ§ñÄ„›ß󟢄=l( ·Ò{E—f1b¤ÒÿÀ(Ï3ô]P=N¼˜pgÜÏ7ﯺW%zSs=o #×_k…SÓ©ìî¹aæZÏ¢Ö®¢À¸Û8öë+tÓ«Ü€ÄØaG—×ga[èÝ€ 7`¼UÅraçúÙ²ë†8K¼GžI =Oʳ´ÍH£–ºÝáhLÓ•'›,Ž®þÞnaÊ„1':gí»ÍªVô‚Ù=[»û” [Ö.îµÒû:s°EÐcu÷àÖVÌHäYÊ×ð²g›I0jÐ2 â$ˆh®ýro;¿ÕBTîèö »s-¦»¦ÎØþ‡äø¾=ȹí%¾yð2úˆv?óޞ†¹áÚ'¨$el¬¦…Æ;|4>ú 1iµ8ÅbŒ~×æ0ÇçÖáøtÐò~Ã`uû¥ª³Þ„a1fŸÕÃ0˜ôºÖOåDŽž–ùê_5–Ä2ÃïÙ <1AC2D4BAA678A3496AE0E10EA834C735>] /Length 5323 /Filter /FlateDecode >> stream xÚ%šYlÉyǧæ(%R<$QJÔA‰Eê u·DQ”DQQ÷AQÔEI­B ÛNÐ ç°ã ŒN`g×/±áÙ|!ÙØJà†a;Ñq‚Zl`?$›l¼È$ëÄ'ÞmŸ“ùý÷凮¯›Ã™þ}_O}US*•JÿgJ¥ž’)•¿ÐFír©\*uÐv|±=tÉ2¥ñ75çl”M©s@±í kÀ‚à@'XV‚.Ð VÐ ú@?X Ö€µ`¬ëA ‚ `#›Àf°lÃ`Ðû;ÀN0 v1 µL€= jJ]_Ö-Ùͧ¬ƒöÛ(Oœm£Uk@lûÁ1 K.ƒ»`¹\ª4,[Á0΃`±ft‚~ O> öƒà 8ƒ#à(HÀ1pL`œ3à8 ΀YpÌsàpœüYÁ}.Úo£ºX ¬ƒ`+P <Á“6rÞZN}äe@äU@òç$Nòç$Nòç$Nòç$Nòç$Nòç$Nòç$Nòç$Nòç$Nòç¤@Nòç$NòçÊû‡¦4üS½«¼çC -¾šÕLéÐï•8âeÁ<¸ –Kµzt‚U`ØvƒÃ`´ïs-µ ¬í*³¾ýÆ;u”Stù¦ö¿¬è½PˆùV*¥Mî)ySù6j¡JŒÉ©Ë|„X™“S¡ùNbk‰‘R¹þÇ.bƒÄȵœªÍljm&Fæ»Ábˆ‘ù^°Øqb¤mN‰ç“Ħˆ‘Ï9µ• ¦÷L¢çTY~ˆØbT@N½åGˆ]%FiäT^ž»FŒšÉ©Áü8±ëÄ(¦œjÌO»GŒ*Ë©Ëü$±ûÄ(¿œ ÍO{DŒºÌ©Õü ±ÇÄ(Ø\  Ĩä\/pŽ˜#¦?Ó?o[GŒÚÏõÆ牭'¦©M‘4xÖ=åi‘ë†]!¦{¯·«›}Ø01ž/¹D鶘>ª$ß"¶—O¤œçA~›Ø>bºM<ò;ÄNã–óŒÈïkÓ-æi‘ódhÜ$ÆS/繑?$v‡˜ôð@Éñd¨ã9™SÄùb+ˆµÕšw·¿ä¦^*qd8±’–A…ØjbûVÏ—æb“ -è ¦6ÁÐ=½uÃö1\VÛ@l/Ã.ÐML7lÃU ‡Ø(±Ý {A±CÄÆöƒÕÄc¸¬%¦DßÅp¬#¦¤e¸Ô‰#¶“á Ø@ì<± 7‚!b—‰0Ü6SAlg¸ð…Ö¼AlÃa°˜Di¨‹GˆÝ"¶•¡þ‘ðw‰é¥ô&wSòof¨8΃ÛÓÛÐÍá;´°Ä†êÆꡯ‡‡>‚¤Lc&ótC =@L¡¯Œ8Dl„Xáa ïÄt뎂„رu ãÄ&‰é¶OÄŽ[Ëpœ$–“²pŠØb«žgˆÍ“îY@±òÑÇpPì…|(UΊ½Ð½ïaxPìÅ1¥ÙEp‰/ I7ÃËà 1=¤•¢WÁ5bz ¬dxÜ ÖILé}Ü"¶ŠX'Ãp›X1•Æ"Ð÷¯ê¨ƒá¸KLõ¡²ºîÛH¬ÆðxHL•ä2xDLuTaøøœ.¡fZIk5 ½[À:Súè?ë’!@ž¶6šÒ³ÿ¢;¹Ç”žÔYf¥-2¶¥¹(yÚÒ ”ìliÞ9jJ1Ü~Z|ý5ý…昻Lé›ÒpŒ×SaëÏ4Û$ëZ»Mé•›º„ÄliÊIJµÈµÖ¤)}ç7t–”jixÈ”¾»A1Í'5ÔLPS¿iSz³_ggøoLõÑÒL|i6¥Ÿ¿K×ñ]‘ðE•è/-²¤uÞ˜¿¬KÈœ9Ô"KZ—Œéy]'È—º[¤J‹,iÝ4fè[:K"µH¤ýVƒÇfK)ðИ±Ðy5hZUÀܧQ2fæt‚iT£fÌ߯ÓYPƒiOc…1+ŠuñžùÚnô‚UÆÜø˜N0jÐ7èéÌP}ÆÜìÑY&@ f< ¦) æ% Ú¾ÆFc–?¢K6-``ªÑ`nÑ1æ=Cºd0·hìûÁ3&ýu]B÷Ù`jÐ`.ÐàË¿A3Ú˜“ÆüîWu3ŠÆap0KôR|½7NóGu]7øzoÌÍà5e¿Nó'ÿ£ëæÀEp\t½ Í©çùÂu¦ìš£kR®Ùõ= [ Æ„wÞß ¹2“ã†fÚùÊÛ]cþj^×iŠ­%‹Çœ-™rúN4YÁhj¾ë@h«ý§‹TYyIÂL«©)l·1o êÏVMa5?í5æ¿ætBSÓ`À˜/)†²&Êš¸l™rùºNhÞ©‰&.›š"Ž˜r÷¸Îb°Ét°¹pÛ›¸l¢¬9fÊ«ÿS×!¹I®5õNi]‡Õ&B›¬É4šò–mú l5Y?hž0åñ·Ch“‰\ƒM¦VMæ\Í3¦|àóºoM”5/˜òô;7b\3åùšKúøZòÁVó¶)ßÞ¨ëhe²©‰²¦,Ü7å'»uƒ…‰ž˜òû^/™Ê¾÷iõåW¬ˆ**¦ü;oë¿ÑŽ$ÜÝo…3å¿¡?£$ ”ݦü‰QÅPV ¬ ÊŠõ ·}öK:«/4ª¢,ÐSÔMù3úø…þ[PœÞ :‚‚,¶˜ò—Þ¥ëXã)ÆÀ¨)ÿå”b¨-Uh~žb¼}öë:‹ßBÿ\³u ±8`ÊßTJ”_ÁBsjM¶›òËï¼2~ šÌD«éôŒ)¿º¬³šIkꬹòœ)ï+:A]¬Ò-.šr1¬TcA5×L¥²^1ʯÀjAûZ,˜JO®”U!y-î™ÊÐS  Lš¹1õ«ŸgòÉ΢ִž˜ÊŽGSùÚ't´XS9ú²†] ô‚•¦ræE‰fÈdq±ÏT®ÖÅÌ׃:¦²ô¦În[À°l2•_ú†^ïÃQ° Œq°ì`ÄTÞÿm½ÞNp°µ¨˜4•ßü”Î&à8 fÀ pÔTžù‰.9 fÁ8t‡XÓZœ6•?ìÑuÀ)pœ  Fç2P{Ã*ØâupÑT>ýŠþV=ÎU°Ô¶¨OYêDn˜Êç&tñ= ngо&ôyu½5µ(Ëà1óq­ÃÑud´Þ2ÇÙ¦òâ³z=50t,«’íW}ÎTì×ߢ6[P›!9ë4•¿ÓYõ j8Ôa¨¥Po°ÚT¾= KÖõ8Ïpž!9CrÆzgF díxµWAÒdÛÀÀ`†î ÓÙ°`:Ãy†Õ¬%ÿ^Ó 4é“‘*Y’‘ Ù>0 Ȉ û™Ö,§)µÅ¿õ¼^e™ªVã7Ão†ßŒ;ž‘ ¦3üfèÎfLåßÕ«>é“‘>Y’a:ÃyFd8Ï.›j××ô4·j3R #2œg¶²Û¦:¸¤‹é\3${B›Q”ZÆǦ:š©zÔ)ULõ§t]p¦:ý! {A·©ž»«‹1\ úLõÊ{uI?X XŠ­¯ƒ` Ø֘굷uñP¨®mì›MuùúdŒ5]@-Ô^°ÓTßûPשMSµÏTÓ7Ûè£êGÁASýàôÆéêBbªÏ<§‹õah’ê3`ÊTP'΀S¦úÉ-žgÁ¸ÎƒYSýìë’sàh˜êŸU»æÁpÜm«þ-]Âòl2ÅÑm°î€%p܀ɗÁc°`ª_Ý©WyîÓì1L%ù‰©þÝ÷54¦¶§[G,°§5À*{Úvþ¯?Ô‰• °òžö¦úßs:KF¤k@¿©þlL1R ÅeŠËtmûÄŸê>«çÜM­ö=]LoŸ’)Ý{ŠótÈÔ:?¯³tåé6Sëûˆ†Û]yJ¤HNI†”dHÙ Hé»S’!¥ÛNI”I±Ÿ’éA ~š}ƒT]4‘ªwF|ªŽ™{Ÿê=³µž$Cz >%ÒY@ ¤¤@Šî”H€ŒH1â<%ÒËñ)âÓk@;7ÀM@2¤ û)öSì§ØO±Ÿb?•ZR %RR ÅyJ¤í·|m[V,kK–E'Ëj”e™Ê²~eYز¬xY–Â,kd–Å3˪še¹Í²gY ³¬ÜY–ô,k}–E@Ëê eÙвžhYh´¬@Z–&-k––ÅLË*§eùÓ².jY0µ¬¤Z–X-k¯–EYËj­eײ¾kYøµ¬[–Š-kÈ–Åe˪³e9Ú²NmYÀ¶¬l[–¼-ká–Er˺emݲènY·,Ó[Öï- û–ËV€eÀ²y`ÙU°l7Xö!,– Ë–†e¯Ã² bÙ±l›XöS,-–ËÖŒeÏƲ™cÙå±lÿXö…,F–$Ë“eïÉ>]·Á"¸–À]pÜÀC° Ççççççççççççççççççç¿¿¿¿¿¿¿¿¿¿¿^ûoøõøõøõøõøõøõøõøõøõøõøõøõz ì0µþÎ’©™mãê ú€ˆ÷ˆ÷ˆ÷ˆ÷ˆ÷ˆ÷ˆ÷ˆ÷ˆ÷ˆ÷ˆ÷ˆ÷ˆ÷ˆ÷ˆ÷ˆ÷ˆ÷ˆ÷Z C¼G¼G¼G¼G¼G¼G¼G¼G¼G¼G¼G¼G¼G¼G¼G¼G¼G¼G¼G¼G¼G¼G¼G¼G¼G¼G¼G¼G¼G¼G¼G¼G¼G¼G¼G|ä¾DÄGÄGÄGÄGÄGÄGÄGÄGÄGÄGÄGÄGÄGÄGÄGÄGŠ=Rì‘b{$"ÉI†H2D’!’ ‘dˆ$C$"ÉI†H2D’!’ ‘dˆ$C$"É•kLí¤Ó‡Ùij¿ò7:"-"iI‹ˆýˆýˆýˆýˆýˆýˆýˆýˆýˆýˆýˆýˆýˆýˆýˆýˆýˆýˆýˆýˆýˆýˆýˆýˆýˆýˆýˆýˆýˆýˆýˆýˆýˆýˆýˆýˆýˆýˆýˆýˆýˆýˆýˆýˆýˆýˆýˆýˆýˆýˆýˆýˆýˆýضßÁþ`ÇK”ATA XÐè+ÀJкÁ*Ðz[ç/õƒÕ` X À:°ÔÁ Ø6‚!° l[ÀV0 ¶í`ì;Á(ØÆÀ8Ø ö€½`˜“`?8‚Cà08Ž‚ÇÁ8¦ÁI0Xò~é48fÁY0Îó .€yp\—Áp\×Á pÜ à6XwÀ¸ îûàx–Á#ðà<à<à<à<à<à<à<à<à<à<à<à<à<à<à<à<à<à<èç8888888888888888888888ø ø ø ø ø ø ø ø ø ø ø ø ø ø ø ø ø ø ø ø ø ÚÒÀoÀoÀoÀoÀoÀoÀoÐãf—©-ÑìuÚÔ~ë+íïžOý›î8ööööööööööööööööööööööööööÃún-f²}AÿæêlÖÕ³‡¦M3z:Wg›ˆÐ8mši—ŒvÎÕ»» Ú%£§sõ>Àº휫³÷IçèÐœ£“sôjŽvÎñƒ§-5í¡Ñع:;5ôyŽvÎÑÎ9~³ã´¥¦=4º;WgëˆÆÎѵ9º;GwçømӾݫ³ JÇçèø휣Åsõƒ€}ɺV£Øu«k}ƒk:9GwçêlÒÓ9~+ä´Gwçê§ÀiÀ¶v]É0njÏ\ÕÝ5µWš:â×´xŽÖÍÑĹú%SûìÏÛiñ|U—°·M;çÔ¿ñ{$§í=íçÑâ¹ú¸ ë§ôtŽžÎÑÓ96Gcçh죓sõǦöB§þÑS{ý£JÌ’é|mUÉØñihŒ=÷•ýµguT1ö“ßÑQÕØ/ÿ¯Žjƾ:¯#kìÞ££Ó±4¬#g:~uDG¦ãÙÐÑ Óñ%\²Ò¸ýktÔeÜ’ÕQ·q¿ý3­2î…Ã:ê1î?ÕQ¯q¯}@G}¦sûË:ê7óƒ:Zm:?ýa‘k‰öHµ“„™3 f’YÓùýsÓdÇ4Ù1MvOõ{ý^@?ÎÐO–ô%n"ÓdÇ4Ù=ÕÏ4ô£$ª'mߺ·ø ŸK)!ºc—RB´ÃŽvØñc2§Wí´Ò»”¢1vô¿ŽîØÑ;~tæ´ûJŸìRJˆvØÑ"»”¢1vü.ÍiG–Ù¥”°£;v)%DOì艿_sÚª¥;v)%D'ìÒÓùÃÿh„_¬-ý?v-œ endstream endobj startxref 654869 %%EOF fftw-3.3.3/doc/html/0000755000175000017500000000000012054411111011166 500000000000000fftw-3.3.3/doc/html/index.html0000644000175000017500000004207512054411111013113 00000000000000 FFTW 3.3.3

FFTW 3.3.3

Table of Contents

Next: , Previous: (dir), Up: (dir)

FFTW User Manual

Welcome to FFTW, the Fastest Fourier Transform in the West. FFTW is a collection of fast C routines to compute the discrete Fourier transform. This manual documents FFTW version 3.3.3.

fftw-3.3.3/doc/html/Introduction.html0000644000175000017500000002553612054411111014470 00000000000000 Introduction - FFTW 3.3.3

Next: , Previous: Top, Up: Top

1 Introduction

This manual documents version 3.3.3 of FFTW, the Fastest Fourier Transform in the West. FFTW is a comprehensive collection of fast C routines for computing the discrete Fourier transform (DFT) and various special cases thereof.

  • FFTW computes the DFT of complex data, real data, even- or odd-symmetric real data (these symmetric transforms are usually known as the discrete cosine or sine transform, respectively), and the discrete Hartley transform (DHT) of real data.
  • The input data can have arbitrary length. FFTW employs O(n log n) algorithms for all lengths, including prime numbers.
  • FFTW supports arbitrary multi-dimensional data.
  • FFTW supports the SSE, SSE2, AVX, Altivec, and MIPS PS instruction sets.
  • FFTW includes parallel (multi-threaded) transforms for shared-memory systems.
  • Starting with version 3.3, FFTW includes distributed-memory parallel transforms using MPI.

We assume herein that you are familiar with the properties and uses of the DFT that are relevant to your application. Otherwise, see e.g. The Fast Fourier Transform and Its Applications by E. O. Brigham (Prentice-Hall, Englewood Cliffs, NJ, 1988). Our web page also has links to FFT-related information online.

In order to use FFTW effectively, you need to learn one basic concept of FFTW's internal structure: FFTW does not use a fixed algorithm for computing the transform, but instead it adapts the DFT algorithm to details of the underlying hardware in order to maximize performance. Hence, the computation of the transform is split into two phases. First, FFTW's planner “learns” the fastest way to compute the transform on your machine. The planner produces a data structure called a plan that contains this information. Subsequently, the plan is executed to transform the array of input data as dictated by the plan. The plan can be reused as many times as needed. In typical high-performance applications, many transforms of the same size are computed and, consequently, a relatively expensive initialization of this sort is acceptable. On the other hand, if you need a single transform of a given size, the one-time cost of the planner becomes significant. For this case, FFTW provides fast planners based on heuristics or on previously computed plans.

FFTW supports transforms of data with arbitrary length, rank, multiplicity, and a general memory layout. In simple cases, however, this generality may be unnecessary and confusing. Consequently, we organized the interface to FFTW into three levels of increasing generality.

  • The basic interface computes a single transform of contiguous data.
  • The advanced interface computes transforms of multiple or strided arrays.
  • The guru interface supports the most general data layouts, multiplicities, and strides.
We expect that most users will be best served by the basic interface, whereas the guru interface requires careful attention to the documentation to avoid problems.

Besides the automatic performance adaptation performed by the planner, it is also possible for advanced users to customize FFTW manually. For example, if code space is a concern, we provide a tool that links only the subset of FFTW needed by your application. Conversely, you may need to extend FFTW because the standard distribution is not sufficient for your needs. For example, the standard FFTW distribution works most efficiently for arrays whose size can be factored into small primes (2, 3, 5, and 7), and otherwise it uses a slower general-purpose routine. If you need efficient transforms of other sizes, you can use FFTW's code generator, which produces fast C programs (“codelets”) for any particular array size you may care about. For example, if you need transforms of size 513 = 19*33,you can customize FFTW to support the factor 19 efficiently.

For more information regarding FFTW, see the paper, “The Design and Implementation of FFTW3,” by M. Frigo and S. G. Johnson, which was an invited paper in Proc. IEEE 93 (2), p. 216 (2005). The code generator is described in the paper “A fast Fourier transform compiler”, by M. Frigo, in the Proceedings of the 1999 ACM SIGPLAN Conference on Programming Language Design and Implementation (PLDI), Atlanta, Georgia, May 1999. These papers, along with the latest version of FFTW, the FAQ, benchmarks, and other links, are available at the FFTW home page.

The current version of FFTW incorporates many good ideas from the past thirty years of FFT literature. In one way or another, FFTW uses the Cooley-Tukey algorithm, the prime factor algorithm, Rader's algorithm for prime sizes, and a split-radix algorithm (with a “conjugate-pair” variation pointed out to us by Dan Bernstein). FFTW's code generator also produces new algorithms that we do not completely understand. The reader is referred to the cited papers for the appropriate references.

The rest of this manual is organized as follows. We first discuss the sequential (single-processor) implementation. We start by describing the basic interface/features of FFTW in Tutorial. Next, Other Important Topics discusses data alignment (see SIMD alignment and fftw_malloc), the storage scheme of multi-dimensional arrays (see Multi-dimensional Array Format), and FFTW's mechanism for storing plans on disk (see Words of Wisdom-Saving Plans). Next, FFTW Reference provides comprehensive documentation of all FFTW's features. Parallel transforms are discussed in their own chapters: Multi-threaded FFTW and Distributed-memory FFTW with MPI. Fortran programmers can also use FFTW, as described in Calling FFTW from Legacy Fortran and Calling FFTW from Modern Fortran. Installation and Customization explains how to install FFTW in your computer system and how to adapt FFTW to your needs. License and copyright information is given in License and Copyright. Finally, we thank all the people who helped us in Acknowledgments. fftw-3.3.3/doc/html/Tutorial.html0000644000175000017500000001100312054411111013572 00000000000000 Tutorial - FFTW 3.3.3

Next: , Previous: Introduction, Up: Top

2 Tutorial

This chapter describes the basic usage of FFTW, i.e., how to compute the Fourier transform of a single array. This chapter tells the truth, but not the whole truth. Specifically, FFTW implements additional routines and flags that are not documented here, although in many cases we try to indicate where added capabilities exist. For more complete information, see FFTW Reference. (Note that you need to compile and install FFTW before you can use it in a program. For the details of the installation, see Installation and Customization.)

We recommend that you read this tutorial in order.1 At the least, read the first section (see Complex One-Dimensional DFTs) before reading any of the others, even if your main interest lies in one of the other transform types.

Users of FFTW version 2 and earlier may also want to read Upgrading from FFTW version 2.

Footnotes

[1] You can read the tutorial in bit-reversed order after computing your first transform.

fftw-3.3.3/doc/html/Complex-One_002dDimensional-DFTs.html0000644000175000017500000002422712054411111017637 00000000000000 Complex One-Dimensional DFTs - FFTW 3.3.3

Next: , Previous: Tutorial, Up: Tutorial

2.1 Complex One-Dimensional DFTs

Plan: To bother about the best method of accomplishing an accidental result. [Ambrose Bierce, The Enlarged Devil's Dictionary.]

The basic usage of FFTW to compute a one-dimensional DFT of size N is simple, and it typically looks something like this code: 

     #include <fftw3.h>
     ...
     {
         fftw_complex *in, *out;
         fftw_plan p;
         ...
         in = (fftw_complex*) fftw_malloc(sizeof(fftw_complex) * N);
         out = (fftw_complex*) fftw_malloc(sizeof(fftw_complex) * N);
         p = fftw_plan_dft_1d(N, in, out, FFTW_FORWARD, FFTW_ESTIMATE);
         ...
         fftw_execute(p); /* repeat as needed */
         ...
         fftw_destroy_plan(p);
         fftw_free(in); fftw_free(out);
     }

You must link this code with the fftw3 library. On Unix systems, link with -lfftw3 -lm.

The example code first allocates the input and output arrays. You can allocate them in any way that you like, but we recommend using fftw_malloc, which behaves like malloc except that it properly aligns the array when SIMD instructions (such as SSE and Altivec) are available (see SIMD alignment and fftw_malloc). [Alternatively, we provide a convenient wrapper function fftw_alloc_complex(N) which has the same effect.]

The data is an array of type fftw_complex, which is by default a double[2] composed of the real (in[i][0]) and imaginary (in[i][1]) parts of a complex number. The next step is to create a plan, which is an object that contains all the data that FFTW needs to compute the FFT. This function creates the plan: 

     fftw_plan fftw_plan_dft_1d(int n, fftw_complex *in, fftw_complex *out,
                                int sign, unsigned flags);

The first argument, n, is the size of the transform you are trying to compute. The size n can be any positive integer, but sizes that are products of small factors are transformed most efficiently (although prime sizes still use an O(n log n) algorithm).

The next two arguments are pointers to the input and output arrays of the transform. These pointers can be equal, indicating an in-place transform.

The fourth argument, sign, can be either FFTW_FORWARD (-1) or FFTW_BACKWARD (+1), and indicates the direction of the transform you are interested in; technically, it is the sign of the exponent in the transform.

The flags argument is usually either FFTW_MEASURE or FFTW_ESTIMATE. FFTW_MEASURE instructs FFTW to run and measure the execution time of several FFTs in order to find the best way to compute the transform of size n. This process takes some time (usually a few seconds), depending on your machine and on the size of the transform. FFTW_ESTIMATE, on the contrary, does not run any computation and just builds a reasonable plan that is probably sub-optimal. In short, if your program performs many transforms of the same size and initialization time is not important, use FFTW_MEASURE; otherwise use the estimate.

You must create the plan before initializing the input, because FFTW_MEASURE overwrites the in/out arrays. (Technically, FFTW_ESTIMATE does not touch your arrays, but you should always create plans first just to be sure.)

Once the plan has been created, you can use it as many times as you like for transforms on the specified in/out arrays, computing the actual transforms via fftw_execute(plan): 

     void fftw_execute(const fftw_plan plan);

The DFT results are stored in-order in the array out, with the zero-frequency (DC) component in out[0]. If in != out, the transform is out-of-place and the input array in is not modified. Otherwise, the input array is overwritten with the transform.

If you want to transform a different array of the same size, you can create a new plan with fftw_plan_dft_1d and FFTW automatically reuses the information from the previous plan, if possible. Alternatively, with the “guru” interface you can apply a given plan to a different array, if you are careful. See FFTW Reference.

When you are done with the plan, you deallocate it by calling fftw_destroy_plan(plan): 

     void fftw_destroy_plan(fftw_plan plan);

If you allocate an array with fftw_malloc() you must deallocate it with fftw_free(). Do not use free() or, heaven forbid, delete. FFTW computes an unnormalized DFT. Thus, computing a forward followed by a backward transform (or vice versa) results in the original array scaled by n. For the definition of the DFT, see What FFTW Really Computes.

If you have a C compiler, such as gcc, that supports the C99 standard, and you #include <complex.h> before <fftw3.h>, then fftw_complex is the native double-precision complex type and you can manipulate it with ordinary arithmetic. Otherwise, FFTW defines its own complex type, which is bit-compatible with the C99 complex type. See Complex numbers. (The C++ <complex> template class may also be usable via a typecast.) To use single or long-double precision versions of FFTW, replace the fftw_ prefix by fftwf_ or fftwl_ and link with -lfftw3f or -lfftw3l, but use the same <fftw3.h> header file.

Many more flags exist besides FFTW_MEASURE and FFTW_ESTIMATE. For example, use FFTW_PATIENT if you're willing to wait even longer for a possibly even faster plan (see FFTW Reference). You can also save plans for future use, as described by Words of Wisdom-Saving Plans. fftw-3.3.3/doc/html/Complex-Multi_002dDimensional-DFTs.html0000644000175000017500000001606412054411111020210 00000000000000 Complex Multi-Dimensional DFTs - FFTW 3.3.3

Next: , Previous: Complex One-Dimensional DFTs, Up: Tutorial

2.2 Complex Multi-Dimensional DFTs

Multi-dimensional transforms work much the same way as one-dimensional transforms: you allocate arrays of fftw_complex (preferably using fftw_malloc), create an fftw_plan, execute it as many times as you want with fftw_execute(plan), and clean up with fftw_destroy_plan(plan) (and fftw_free).

FFTW provides two routines for creating plans for 2d and 3d transforms, and one routine for creating plans of arbitrary dimensionality. The 2d and 3d routines have the following signature: 

     fftw_plan fftw_plan_dft_2d(int n0, int n1,
                                fftw_complex *in, fftw_complex *out,
                                int sign, unsigned flags);
     fftw_plan fftw_plan_dft_3d(int n0, int n1, int n2,
                                fftw_complex *in, fftw_complex *out,
                                int sign, unsigned flags);

These routines create plans for n0 by n1 two-dimensional (2d) transforms and n0 by n1 by n2 3d transforms, respectively. All of these transforms operate on contiguous arrays in the C-standard row-major order, so that the last dimension has the fastest-varying index in the array. This layout is described further in Multi-dimensional Array Format.

FFTW can also compute transforms of higher dimensionality. In order to avoid confusion between the various meanings of the the word “dimension”, we use the term rank to denote the number of independent indices in an array.1 For example, we say that a 2d transform has rank 2, a 3d transform has rank 3, and so on. You can plan transforms of arbitrary rank by means of the following function: 

     fftw_plan fftw_plan_dft(int rank, const int *n,
                             fftw_complex *in, fftw_complex *out,
                             int sign, unsigned flags);

Here, n is a pointer to an array n[rank] denoting an n[0] by n[1] by ... by n[rank-1] transform. Thus, for example, the call 

     fftw_plan_dft_2d(n0, n1, in, out, sign, flags);

is equivalent to the following code fragment: 

     int n[2];
     n[0] = n0;
     n[1] = n1;
     fftw_plan_dft(2, n, in, out, sign, flags);

fftw_plan_dft is not restricted to 2d and 3d transforms, however, but it can plan transforms of arbitrary rank.

You may have noticed that all the planner routines described so far have overlapping functionality. For example, you can plan a 1d or 2d transform by using fftw_plan_dft with a rank of 1 or 2, or even by calling fftw_plan_dft_3d with n0 and/or n1 equal to 1 (with no loss in efficiency). This pattern continues, and FFTW's planning routines in general form a “partial order,” sequences of interfaces with strictly increasing generality but correspondingly greater complexity.

fftw_plan_dft is the most general complex-DFT routine that we describe in this tutorial, but there are also the advanced and guru interfaces, which allow one to efficiently combine multiple/strided transforms into a single FFTW plan, transform a subset of a larger multi-dimensional array, and/or to handle more general complex-number formats. For more information, see FFTW Reference.

Footnotes

[1] The term “rank” is commonly used in the APL, FORTRAN, and Common Lisp traditions, although it is not so common in the C world.

fftw-3.3.3/doc/html/One_002dDimensional-DFTs-of-Real-Data.html0000644000175000017500000002012212054411111020352 00000000000000 One-Dimensional DFTs of Real Data - FFTW 3.3.3

Next: , Previous: Complex Multi-Dimensional DFTs, Up: Tutorial

2.3 One-Dimensional DFTs of Real Data

In many practical applications, the input data in[i] are purely real numbers, in which case the DFT output satisfies the “Hermitian” redundancy: out[i] is the conjugate of out[n-i]. It is possible to take advantage of these circumstances in order to achieve roughly a factor of two improvement in both speed and memory usage.

In exchange for these speed and space advantages, the user sacrifices some of the simplicity of FFTW's complex transforms. First of all, the input and output arrays are of different sizes and types: the input is n real numbers, while the output is n/2+1 complex numbers (the non-redundant outputs); this also requires slight “padding” of the input array for in-place transforms. Second, the inverse transform (complex to real) has the side-effect of overwriting its input array, by default. Neither of these inconveniences should pose a serious problem for users, but it is important to be aware of them.

The routines to perform real-data transforms are almost the same as those for complex transforms: you allocate arrays of double and/or fftw_complex (preferably using fftw_malloc or fftw_alloc_complex), create an fftw_plan, execute it as many times as you want with fftw_execute(plan), and clean up with fftw_destroy_plan(plan) (and fftw_free). The only differences are that the input (or output) is of type double and there are new routines to create the plan. In one dimension: 

     fftw_plan fftw_plan_dft_r2c_1d(int n, double *in, fftw_complex *out,
                                    unsigned flags);
     fftw_plan fftw_plan_dft_c2r_1d(int n, fftw_complex *in, double *out,
                                    unsigned flags);

for the real input to complex-Hermitian output (r2c) and complex-Hermitian input to real output (c2r) transforms. Unlike the complex DFT planner, there is no sign argument. Instead, r2c DFTs are always FFTW_FORWARD and c2r DFTs are always FFTW_BACKWARD. (For single/long-double precision fftwf and fftwl, double should be replaced by float and long double, respectively.)

Here, n is the “logical” size of the DFT, not necessarily the physical size of the array. In particular, the real (double) array has n elements, while the complex (fftw_complex) array has n/2+1 elements (where the division is rounded down). For an in-place transform, in and out are aliased to the same array, which must be big enough to hold both; so, the real array would actually have 2*(n/2+1) elements, where the elements beyond the first n are unused padding. (Note that this is very different from the concept of “zero-padding” a transform to a larger length, which changes the logical size of the DFT by actually adding new input data.) The kth element of the complex array is exactly the same as the kth element of the corresponding complex DFT. All positive n are supported; products of small factors are most efficient, but an O(n log n) algorithm is used even for prime sizes.

As noted above, the c2r transform destroys its input array even for out-of-place transforms. This can be prevented, if necessary, by including FFTW_PRESERVE_INPUT in the flags, with unfortunately some sacrifice in performance. This flag is also not currently supported for multi-dimensional real DFTs (next section).

Readers familiar with DFTs of real data will recall that the 0th (the “DC”) and n/2-th (the “Nyquist” frequency, when n is even) elements of the complex output are purely real. Some implementations therefore store the Nyquist element where the DC imaginary part would go, in order to make the input and output arrays the same size. Such packing, however, does not generalize well to multi-dimensional transforms, and the space savings are miniscule in any case; FFTW does not support it.

An alternative interface for one-dimensional r2c and c2r DFTs can be found in the ‘r2r’ interface (see The Halfcomplex-format DFT), with “halfcomplex”-format output that is the same size (and type) as the input array. That interface, although it is not very useful for multi-dimensional transforms, may sometimes yield better performance. fftw-3.3.3/doc/html/Multi_002dDimensional-DFTs-of-Real-Data.html0000644000175000017500000001707312054411111020736 00000000000000 Multi-Dimensional DFTs of Real Data - FFTW 3.3.3

Next: , Previous: One-Dimensional DFTs of Real Data, Up: Tutorial

2.4 Multi-Dimensional DFTs of Real Data

Multi-dimensional DFTs of real data use the following planner routines: 

     fftw_plan fftw_plan_dft_r2c_2d(int n0, int n1,
                                    double *in, fftw_complex *out,
                                    unsigned flags);
     fftw_plan fftw_plan_dft_r2c_3d(int n0, int n1, int n2,
                                    double *in, fftw_complex *out,
                                    unsigned flags);
     fftw_plan fftw_plan_dft_r2c(int rank, const int *n,
                                 double *in, fftw_complex *out,
                                 unsigned flags);

as well as the corresponding c2r routines with the input/output types swapped. These routines work similarly to their complex analogues, except for the fact that here the complex output array is cut roughly in half and the real array requires padding for in-place transforms (as in 1d, above).

As before, n is the logical size of the array, and the consequences of this on the the format of the complex arrays deserve careful attention. Suppose that the real data has dimensions n0 × n1 × n2 × … × nd-1 (in row-major order). Then, after an r2c transform, the output is an n0 × n1 × n2 × … × (nd-1/2 + 1) array of fftw_complex values in row-major order, corresponding to slightly over half of the output of the corresponding complex DFT. (The division is rounded down.) The ordering of the data is otherwise exactly the same as in the complex-DFT case.

For out-of-place transforms, this is the end of the story: the real data is stored as a row-major array of size n0 × n1 × n2 × … × nd-1 and the complex data is stored as a row-major array of size n0 × n1 × n2 × … × (nd-1/2 + 1).

For in-place transforms, however, extra padding of the real-data array is necessary because the complex array is larger than the real array, and the two arrays share the same memory locations. Thus, for in-place transforms, the final dimension of the real-data array must be padded with extra values to accommodate the size of the complex data—two values if the last dimension is even and one if it is odd. That is, the last dimension of the real data must physically contain 2 * (nd-1/2+1)double values (exactly enough to hold the complex data). This physical array size does not, however, change the logical array size—only nd-1values are actually stored in the last dimension, and nd-1is the last dimension passed to the plan-creation routine.

For example, consider the transform of a two-dimensional real array of size n0 by n1. The output of the r2c transform is a two-dimensional complex array of size n0 by n1/2+1, where the y dimension has been cut nearly in half because of redundancies in the output. Because fftw_complex is twice the size of double, the output array is slightly bigger than the input array. Thus, if we want to compute the transform in place, we must pad the input array so that it is of size n0 by 2*(n1/2+1). If n1 is even, then there are two padding elements at the end of each row (which need not be initialized, as they are only used for output).

The following illustration depicts the input and output arrays just described, for both the out-of-place and in-place transforms (with the arrows indicating consecutive memory locations): rfftwnd-for-html.png

These transforms are unnormalized, so an r2c followed by a c2r transform (or vice versa) will result in the original data scaled by the number of real data elements—that is, the product of the (logical) dimensions of the real data.

(Because the last dimension is treated specially, if it is equal to 1 the transform is not equivalent to a lower-dimensional r2c/c2r transform. In that case, the last complex dimension also has size 1 (=1/2+1), and no advantage is gained over the complex transforms.) fftw-3.3.3/doc/html/More-DFTs-of-Real-Data.html0000644000175000017500000001513512054411111015633 00000000000000 More DFTs of Real Data - FFTW 3.3.3

Previous: Multi-Dimensional DFTs of Real Data, Up: Tutorial

2.5 More DFTs of Real Data

FFTW supports several other transform types via a unified r2r (real-to-real) interface, so called because it takes a real (double) array and outputs a real array of the same size. These r2r transforms currently fall into three categories: DFTs of real input and complex-Hermitian output in halfcomplex format, DFTs of real input with even/odd symmetry (a.k.a. discrete cosine/sine transforms, DCTs/DSTs), and discrete Hartley transforms (DHTs), all described in more detail by the following sections.

The r2r transforms follow the by now familiar interface of creating an fftw_plan, executing it with fftw_execute(plan), and destroying it with fftw_destroy_plan(plan). Furthermore, all r2r transforms share the same planner interface: 

     fftw_plan fftw_plan_r2r_1d(int n, double *in, double *out,
                                fftw_r2r_kind kind, unsigned flags);
     fftw_plan fftw_plan_r2r_2d(int n0, int n1, double *in, double *out,
                                fftw_r2r_kind kind0, fftw_r2r_kind kind1,
                                unsigned flags);
     fftw_plan fftw_plan_r2r_3d(int n0, int n1, int n2,
                                double *in, double *out,
                                fftw_r2r_kind kind0,
                                fftw_r2r_kind kind1,
                                fftw_r2r_kind kind2,
                                unsigned flags);
     fftw_plan fftw_plan_r2r(int rank, const int *n, double *in, double *out,
                             const fftw_r2r_kind *kind, unsigned flags);

Just as for the complex DFT, these plan 1d/2d/3d/multi-dimensional transforms for contiguous arrays in row-major order, transforming (real) input to output of the same size, where n specifies the physical dimensions of the arrays. All positive n are supported (with the exception of n=1 for the FFTW_REDFT00 kind, noted in the real-even subsection below); products of small factors are most efficient (factorizing n-1 and n+1 for FFTW_REDFT00 and FFTW_RODFT00 kinds, described below), but an O(n log n) algorithm is used even for prime sizes.

Each dimension has a kind parameter, of type fftw_r2r_kind, specifying the kind of r2r transform to be used for that dimension. (In the case of fftw_plan_r2r, this is an array kind[rank] where kind[i] is the transform kind for the dimension n[i].) The kind can be one of a set of predefined constants, defined in the following subsections.

In other words, FFTW computes the separable product of the specified r2r transforms over each dimension, which can be used e.g. for partial differential equations with mixed boundary conditions. (For some r2r kinds, notably the halfcomplex DFT and the DHT, such a separable product is somewhat problematic in more than one dimension, however, as is described below.)

In the current version of FFTW, all r2r transforms except for the halfcomplex type are computed via pre- or post-processing of halfcomplex transforms, and they are therefore not as fast as they could be. Since most other general DCT/DST codes employ a similar algorithm, however, FFTW's implementation should provide at least competitive performance. fftw-3.3.3/doc/html/The-Halfcomplex_002dformat-DFT.html0000644000175000017500000001405512054411111017332 00000000000000 The Halfcomplex-format DFT - FFTW 3.3.3

Next: , Previous: More DFTs of Real Data, Up: More DFTs of Real Data

2.5.1 The Halfcomplex-format DFT

An r2r kind of FFTW_R2HC (r2hc) corresponds to an r2c DFT (see One-Dimensional DFTs of Real Data) but with “halfcomplex” format output, and may sometimes be faster and/or more convenient than the latter. The inverse hc2r transform is of kind FFTW_HC2R. This consists of the non-redundant half of the complex output for a 1d real-input DFT of size n, stored as a sequence of n real numbers (double) in the format:

r0, r1, r2, ..., rn/2, i(n+1)/2-1, ..., i2, i1

Here, rkis the real part of the kth output, and ikis the imaginary part. (Division by 2 is rounded down.) For a halfcomplex array hc[n], the kth component thus has its real part in hc[k] and its imaginary part in hc[n-k], with the exception of k == 0 or n/2 (the latter only if n is even)—in these two cases, the imaginary part is zero due to symmetries of the real-input DFT, and is not stored. Thus, the r2hc transform of n real values is a halfcomplex array of length n, and vice versa for hc2r.

Aside from the differing format, the output of FFTW_R2HC/FFTW_HC2R is otherwise exactly the same as for the corresponding 1d r2c/c2r transform (i.e. FFTW_FORWARD/FFTW_BACKWARD transforms, respectively). Recall that these transforms are unnormalized, so r2hc followed by hc2r will result in the original data multiplied by n. Furthermore, like the c2r transform, an out-of-place hc2r transform will destroy its input array.

Although these halfcomplex transforms can be used with the multi-dimensional r2r interface, the interpretation of such a separable product of transforms along each dimension is problematic. For example, consider a two-dimensional n0 by n1, r2hc by r2hc transform planned by fftw_plan_r2r_2d(n0, n1, in, out, FFTW_R2HC, FFTW_R2HC, FFTW_MEASURE). Conceptually, FFTW first transforms the rows (of size n1) to produce halfcomplex rows, and then transforms the columns (of size n0). Half of these column transforms, however, are of imaginary parts, and should therefore be multiplied by i and combined with the r2hc transforms of the real columns to produce the 2d DFT amplitudes; FFTW's r2r transform does not perform this combination for you. Thus, if a multi-dimensional real-input/output DFT is required, we recommend using the ordinary r2c/c2r interface (see Multi-Dimensional DFTs of Real Data). fftw-3.3.3/doc/html/Real-even_002fodd-DFTs-_0028cosine_002fsine-transforms_0029.html0000644000175000017500000002514012054411111024033 00000000000000 Real even/odd DFTs (cosine/sine transforms) - FFTW 3.3.3

Next: , Previous: The Halfcomplex-format DFT, Up: More DFTs of Real Data

2.5.2 Real even/odd DFTs (cosine/sine transforms)

The Fourier transform of a real-even function f(-x) = f(x) is real-even, and i times the Fourier transform of a real-odd function f(-x) = -f(x) is real-odd. Similar results hold for a discrete Fourier transform, and thus for these symmetries the need for complex inputs/outputs is entirely eliminated. Moreover, one gains a factor of two in speed/space from the fact that the data are real, and an additional factor of two from the even/odd symmetry: only the non-redundant (first) half of the array need be stored. The result is the real-even DFT (REDFT) and the real-odd DFT (RODFT), also known as the discrete cosine and sine transforms (DCT and DST), respectively.

(In this section, we describe the 1d transforms; multi-dimensional transforms are just a separable product of these transforms operating along each dimension.)

Because of the discrete sampling, one has an additional choice: is the data even/odd around a sampling point, or around the point halfway between two samples? The latter corresponds to shifting the samples by half an interval, and gives rise to several transform variants denoted by REDFTab and RODFTab: a and b are 0 or 1, and indicate whether the input (a) and/or output (b) are shifted by half a sample (1 means it is shifted). These are also known as types I-IV of the DCT and DST, and all four types are supported by FFTW's r2r interface.1

The r2r kinds for the various REDFT and RODFT types supported by FFTW, along with the boundary conditions at both ends of the input array (n real numbers in[j=0..n-1]), are:

  • FFTW_REDFT00 (DCT-I): even around j=0 and even around j=n-1.
  • FFTW_REDFT10 (DCT-II, “the” DCT): even around j=-0.5 and even around j=n-0.5.
  • FFTW_REDFT01 (DCT-III, “the” IDCT): even around j=0 and odd around j=n.
  • FFTW_REDFT11 (DCT-IV): even around j=-0.5 and odd around j=n-0.5.
  • FFTW_RODFT00 (DST-I): odd around j=-1 and odd around j=n.
  • FFTW_RODFT10 (DST-II): odd around j=-0.5 and odd around j=n-0.5.
  • FFTW_RODFT01 (DST-III): odd around j=-1 and even around j=n-1.
  • FFTW_RODFT11 (DST-IV): odd around j=-0.5 and even around j=n-0.5.

Note that these symmetries apply to the “logical” array being transformed; there are no constraints on your physical input data. So, for example, if you specify a size-5 REDFT00 (DCT-I) of the data abcde, it corresponds to the DFT of the logical even array abcdedcb of size 8. A size-4 REDFT10 (DCT-II) of the data abcd corresponds to the size-8 logical DFT of the even array abcddcba, shifted by half a sample.

All of these transforms are invertible. The inverse of R*DFT00 is R*DFT00; of R*DFT10 is R*DFT01 and vice versa (these are often called simply “the” DCT and IDCT, respectively); and of R*DFT11 is R*DFT11. However, the transforms computed by FFTW are unnormalized, exactly like the corresponding real and complex DFTs, so computing a transform followed by its inverse yields the original array scaled by N, where N is the logical DFT size. For REDFT00, N=2(n-1); for RODFT00, N=2(n+1); otherwise, N=2n.

Note that the boundary conditions of the transform output array are given by the input boundary conditions of the inverse transform. Thus, the above transforms are all inequivalent in terms of input/output boundary conditions, even neglecting the 0.5 shift difference.

FFTW is most efficient when N is a product of small factors; note that this differs from the factorization of the physical size n for REDFT00 and RODFT00! There is another oddity: n=1 REDFT00 transforms correspond to N=0, and so are not defined (the planner will return NULL). Otherwise, any positive n is supported.

For the precise mathematical definitions of these transforms as used by FFTW, see What FFTW Really Computes. (For people accustomed to the DCT/DST, FFTW's definitions have a coefficient of 2 in front of the cos/sin functions so that they correspond precisely to an even/odd DFT of size N. Some authors also include additional multiplicative factors of √2for selected inputs and outputs; this makes the transform orthogonal, but sacrifices the direct equivalence to a symmetric DFT.)

Which type do you need?

Since the required flavor of even/odd DFT depends upon your problem, you are the best judge of this choice, but we can make a few comments on relative efficiency to help you in your selection. In particular, R*DFT01 and R*DFT10 tend to be slightly faster than R*DFT11 (especially for odd sizes), while the R*DFT00 transforms are sometimes significantly slower (especially for even sizes).2

Thus, if only the boundary conditions on the transform inputs are specified, we generally recommend R*DFT10 over R*DFT00 and R*DFT01 over R*DFT11 (unless the half-sample shift or the self-inverse property is significant for your problem).

If performance is important to you and you are using only small sizes (say n<200), e.g. for multi-dimensional transforms, then you might consider generating hard-coded transforms of those sizes and types that you are interested in (see Generating your own code).

We are interested in hearing what types of symmetric transforms you find most useful.

Footnotes

[1] There are also type V-VIII transforms, which correspond to a logical DFT of odd size N, independent of whether the physical size n is odd, but we do not support these variants.

[2] R*DFT00 is sometimes slower in FFTW because we discovered that the standard algorithm for computing this by a pre/post-processed real DFT—the algorithm used in FFTPACK, Numerical Recipes, and other sources for decades now—has serious numerical problems: it already loses several decimal places of accuracy for 16k sizes. There seem to be only two alternatives in the literature that do not suffer similarly: a recursive decomposition into smaller DCTs, which would require a large set of codelets for efficiency and generality, or sacrificing a factor of 2 in speed to use a real DFT of twice the size. We currently employ the latter technique for general n, as well as a limited form of the former method: a split-radix decomposition when n is odd (N a multiple of 4). For N containing many factors of 2, the split-radix method seems to recover most of the speed of the standard algorithm without the accuracy tradeoff.

fftw-3.3.3/doc/html/The-Discrete-Hartley-Transform.html0000644000175000017500000001240412054411111017634 00000000000000 The Discrete Hartley Transform - FFTW 3.3.3

Previous: Real even/odd DFTs (cosine/sine transforms), Up: More DFTs of Real Data

2.5.3 The Discrete Hartley Transform

If you are planning to use the DHT because you've heard that it is “faster” than the DFT (FFT), stop here. The DHT is not faster than the DFT. That story is an old but enduring misconception that was debunked in 1987.

The discrete Hartley transform (DHT) is an invertible linear transform closely related to the DFT. In the DFT, one multiplies each input by cos - i * sin (a complex exponential), whereas in the DHT each input is multiplied by simply cos + sin. Thus, the DHT transforms n real numbers to n real numbers, and has the convenient property of being its own inverse. In FFTW, a DHT (of any positive n) can be specified by an r2r kind of FFTW_DHT. Like the DFT, in FFTW the DHT is unnormalized, so computing a DHT of size n followed by another DHT of the same size will result in the original array multiplied by n. The DHT was originally proposed as a more efficient alternative to the DFT for real data, but it was subsequently shown that a specialized DFT (such as FFTW's r2hc or r2c transforms) could be just as fast. In FFTW, the DHT is actually computed by post-processing an r2hc transform, so there is ordinarily no reason to prefer it from a performance perspective.1 However, we have heard rumors that the DHT might be the most appropriate transform in its own right for certain applications, and we would be very interested to hear from anyone who finds it useful.

If FFTW_DHT is specified for multiple dimensions of a multi-dimensional transform, FFTW computes the separable product of 1d DHTs along each dimension. Unfortunately, this is not quite the same thing as a true multi-dimensional DHT; you can compute the latter, if necessary, with at most rank-1 post-processing passes [see e.g. H. Hao and R. N. Bracewell, Proc. IEEE 75, 264–266 (1987)].

For the precise mathematical definition of the DHT as used by FFTW, see What FFTW Really Computes.

Footnotes

[1] We provide the DHT mainly as a byproduct of some internal algorithms. FFTW computes a real input/output DFT of prime size by re-expressing it as a DHT plus post/pre-processing and then using Rader's prime-DFT algorithm adapted to the DHT.

fftw-3.3.3/doc/html/Other-Important-Topics.html0000644000175000017500000000554212054411111016275 00000000000000 Other Important Topics - FFTW 3.3.3

Next: , Previous: Tutorial, Up: Top

3 Other Important Topics

fftw-3.3.3/doc/html/SIMD-alignment-and-fftw_005fmalloc.html0000644000175000017500000001255112054411111020136 00000000000000 SIMD alignment and fftw_malloc - FFTW 3.3.3

Next: , Previous: Other Important Topics, Up: Other Important Topics

3.1 SIMD alignment and fftw_malloc

SIMD, which stands for “Single Instruction Multiple Data,” is a set of special operations supported by some processors to perform a single operation on several numbers (usually 2 or 4) simultaneously. SIMD floating-point instructions are available on several popular CPUs: SSE/SSE2/AVX on recent x86/x86-64 processors, AltiVec (single precision) on some PowerPCs (Apple G4 and higher), NEON on some ARM models, and MIPS Paired Single (currently only in FFTW 3.2.x). FFTW can be compiled to support the SIMD instructions on any of these systems.

A program linking to an FFTW library compiled with SIMD support can obtain a nonnegligible speedup for most complex and r2c/c2r transforms. In order to obtain this speedup, however, the arrays of complex (or real) data passed to FFTW must be specially aligned in memory (typically 16-byte aligned), and often this alignment is more stringent than that provided by the usual malloc (etc.) allocation routines.

In order to guarantee proper alignment for SIMD, therefore, in case your program is ever linked against a SIMD-using FFTW, we recommend allocating your transform data with fftw_malloc and de-allocating it with fftw_free. These have exactly the same interface and behavior as malloc/free, except that for a SIMD FFTW they ensure that the returned pointer has the necessary alignment (by calling memalign or its equivalent on your OS).

You are not required to use fftw_malloc. You can allocate your data in any way that you like, from malloc to new (in C++) to a fixed-size array declaration. If the array happens not to be properly aligned, FFTW will not use the SIMD extensions. Since fftw_malloc only ever needs to be used for real and complex arrays, we provide two convenient wrapper routines fftw_alloc_real(N) and fftw_alloc_complex(N) that are equivalent to (double*)fftw_malloc(sizeof(double) * N) and (fftw_complex*)fftw_malloc(sizeof(fftw_complex) * N), respectively (or their equivalents in other precisions). fftw-3.3.3/doc/html/Multi_002ddimensional-Array-Format.html0000644000175000017500000000716212054411111020346 00000000000000 Multi-dimensional Array Format - FFTW 3.3.3

Next: , Previous: SIMD alignment and fftw_malloc, Up: Other Important Topics

3.2 Multi-dimensional Array Format

This section describes the format in which multi-dimensional arrays are stored in FFTW. We felt that a detailed discussion of this topic was necessary. Since several different formats are common, this topic is often a source of confusion.

fftw-3.3.3/doc/html/Row_002dmajor-Format.html0000644000175000017500000001114112054411111015545 00000000000000 Row-major Format - FFTW 3.3.3

Next: , Previous: Multi-dimensional Array Format, Up: Multi-dimensional Array Format

3.2.1 Row-major Format

The multi-dimensional arrays passed to fftw_plan_dft etcetera are expected to be stored as a single contiguous block in row-major order (sometimes called “C order”). Basically, this means that as you step through adjacent memory locations, the first dimension's index varies most slowly and the last dimension's index varies most quickly.

To be more explicit, let us consider an array of rank d whose dimensions are n0 × n1 × n2 × … × nd-1. Now, we specify a location in the array by a sequence of d (zero-based) indices, one for each dimension: (i0, i1, i2,..., id-1). If the array is stored in row-major order, then this element is located at the position id-1 + nd-1 * (id-2 + nd-2 * (... + n1 * i0)).

Note that, for the ordinary complex DFT, each element of the array must be of type fftw_complex; i.e. a (real, imaginary) pair of (double-precision) numbers.

In the advanced FFTW interface, the physical dimensions n from which the indices are computed can be different from (larger than) the logical dimensions of the transform to be computed, in order to transform a subset of a larger array. Note also that, in the advanced interface, the expression above is multiplied by a stride to get the actual array index—this is useful in situations where each element of the multi-dimensional array is actually a data structure (or another array), and you just want to transform a single field. In the basic interface, however, the stride is 1. fftw-3.3.3/doc/html/Column_002dmajor-Format.html0000644000175000017500000000760012054411111016240 00000000000000 Column-major Format - FFTW 3.3.3

Next: , Previous: Row-major Format, Up: Multi-dimensional Array Format

3.2.2 Column-major Format

Readers from the Fortran world are used to arrays stored in column-major order (sometimes called “Fortran order”). This is essentially the exact opposite of row-major order in that, here, the first dimension's index varies most quickly.

If you have an array stored in column-major order and wish to transform it using FFTW, it is quite easy to do. When creating the plan, simply pass the dimensions of the array to the planner in reverse order. For example, if your array is a rank three N x M x L matrix in column-major order, you should pass the dimensions of the array as if it were an L x M x N matrix (which it is, from the perspective of FFTW). This is done for you automatically by the FFTW legacy-Fortran interface (see Calling FFTW from Legacy Fortran), but you must do it manually with the modern Fortran interface (see Reversing array dimensions). fftw-3.3.3/doc/html/Fixed_002dsize-Arrays-in-C.html0000644000175000017500000001023012054411111016472 00000000000000 Fixed-size Arrays in C - FFTW 3.3.3

Next: , Previous: Column-major Format, Up: Multi-dimensional Array Format

3.2.3 Fixed-size Arrays in C

A multi-dimensional array whose size is declared at compile time in C is already in row-major order. You don't have to do anything special to transform it. For example: 

     {
          fftw_complex data[N0][N1][N2];
          fftw_plan plan;
          ...
          plan = fftw_plan_dft_3d(N0, N1, N2, &data[0][0][0], &data[0][0][0],
                                  FFTW_FORWARD, FFTW_ESTIMATE);
          ...
     }

This will plan a 3d in-place transform of size N0 x N1 x N2. Notice how we took the address of the zero-th element to pass to the planner (we could also have used a typecast).

However, we tend to discourage users from declaring their arrays in this way, for two reasons. First, this allocates the array on the stack (“automatic” storage), which has a very limited size on most operating systems (declaring an array with more than a few thousand elements will often cause a crash). (You can get around this limitation on many systems by declaring the array as static and/or global, but that has its own drawbacks.) Second, it may not optimally align the array for use with a SIMD FFTW (see SIMD alignment and fftw_malloc). Instead, we recommend using fftw_malloc, as described below. fftw-3.3.3/doc/html/Dynamic-Arrays-in-C.html0000644000175000017500000001105412054411111015404 00000000000000 Dynamic Arrays in C - FFTW 3.3.3

Next: , Previous: Fixed-size Arrays in C, Up: Multi-dimensional Array Format

3.2.4 Dynamic Arrays in C

We recommend allocating most arrays dynamically, with fftw_malloc. This isn't too hard to do, although it is not as straightforward for multi-dimensional arrays as it is for one-dimensional arrays.

Creating the array is simple: using a dynamic-allocation routine like fftw_malloc, allocate an array big enough to store N fftw_complex values (for a complex DFT), where N is the product of the sizes of the array dimensions (i.e. the total number of complex values in the array). For example, here is code to allocate a 5 × 12 × 27 rank-3 array: 

     fftw_complex *an_array;
     an_array = (fftw_complex*) fftw_malloc(5*12*27 * sizeof(fftw_complex));

Accessing the array elements, however, is more tricky—you can't simply use multiple applications of the ‘[]’ operator like you could for fixed-size arrays. Instead, you have to explicitly compute the offset into the array using the formula given earlier for row-major arrays. For example, to reference the (i,j,k)-th element of the array allocated above, you would use the expression an_array[k + 27 * (j + 12 * i)].

This pain can be alleviated somewhat by defining appropriate macros, or, in C++, creating a class and overloading the ‘()’ operator. The recent C99 standard provides a way to reinterpret the dynamic array as a “variable-length” multi-dimensional array amenable to ‘[]’, but this feature is not yet widely supported by compilers. fftw-3.3.3/doc/html/Dynamic-Arrays-in-C_002dThe-Wrong-Way.html0000644000175000017500000001024512054411111020423 00000000000000 Dynamic Arrays in C-The Wrong Way - FFTW 3.3.3

Previous: Dynamic Arrays in C, Up: Multi-dimensional Array Format

3.2.5 Dynamic Arrays in C—The Wrong Way

A different method for allocating multi-dimensional arrays in C is often suggested that is incompatible with FFTW: using it will cause FFTW to die a painful death. We discuss the technique here, however, because it is so commonly known and used. This method is to create arrays of pointers of arrays of pointers of ...etcetera. For example, the analogue in this method to the example above is: 

     int i,j;
     fftw_complex ***a_bad_array;  /* another way to make a 5x12x27 array */
     
     a_bad_array = (fftw_complex ***) malloc(5 * sizeof(fftw_complex **));
     for (i = 0; i < 5; ++i) {
          a_bad_array[i] =
             (fftw_complex **) malloc(12 * sizeof(fftw_complex *));
          for (j = 0; j < 12; ++j)
               a_bad_array[i][j] =
                     (fftw_complex *) malloc(27 * sizeof(fftw_complex));
     }

As you can see, this sort of array is inconvenient to allocate (and deallocate). On the other hand, it has the advantage that the (i,j,k)-th element can be referenced simply by a_bad_array[i][j][k].

If you like this technique and want to maximize convenience in accessing the array, but still want to pass the array to FFTW, you can use a hybrid method. Allocate the array as one contiguous block, but also declare an array of arrays of pointers that point to appropriate places in the block. That sort of trick is beyond the scope of this documentation; for more information on multi-dimensional arrays in C, see the comp.lang.c FAQ. fftw-3.3.3/doc/html/Words-of-Wisdom_002dSaving-Plans.html0000644000175000017500000001356612054411111017717 00000000000000 Words of Wisdom-Saving Plans - FFTW 3.3.3

Next: , Previous: Multi-dimensional Array Format, Up: Other Important Topics

3.3 Words of Wisdom—Saving Plans

FFTW implements a method for saving plans to disk and restoring them. In fact, what FFTW does is more general than just saving and loading plans. The mechanism is called wisdom. Here, we describe this feature at a high level. See FFTW Reference, for a less casual but more complete discussion of how to use wisdom in FFTW.

Plans created with the FFTW_MEASURE, FFTW_PATIENT, or FFTW_EXHAUSTIVE options produce near-optimal FFT performance, but may require a long time to compute because FFTW must measure the runtime of many possible plans and select the best one. This setup is designed for the situations where so many transforms of the same size must be computed that the start-up time is irrelevant. For short initialization times, but slower transforms, we have provided FFTW_ESTIMATE. The wisdom mechanism is a way to get the best of both worlds: you compute a good plan once, save it to disk, and later reload it as many times as necessary. The wisdom mechanism can actually save and reload many plans at once, not just one.

Whenever you create a plan, the FFTW planner accumulates wisdom, which is information sufficient to reconstruct the plan. After planning, you can save this information to disk by means of the function: 

     int fftw_export_wisdom_to_filename(const char *filename);

(This function returns non-zero on success.)

The next time you run the program, you can restore the wisdom with fftw_import_wisdom_from_filename (which also returns non-zero on success), and then recreate the plan using the same flags as before. 

     int fftw_import_wisdom_from_filename(const char *filename);

Wisdom is automatically used for any size to which it is applicable, as long as the planner flags are not more “patient” than those with which the wisdom was created. For example, wisdom created with FFTW_MEASURE can be used if you later plan with FFTW_ESTIMATE or FFTW_MEASURE, but not with FFTW_PATIENT.

The wisdom is cumulative, and is stored in a global, private data structure managed internally by FFTW. The storage space required is minimal, proportional to the logarithm of the sizes the wisdom was generated from. If memory usage is a concern, however, the wisdom can be forgotten and its associated memory freed by calling: 

     void fftw_forget_wisdom(void);

Wisdom can be exported to a file, a string, or any other medium. For details, see Wisdom. fftw-3.3.3/doc/html/Caveats-in-Using-Wisdom.html0000644000175000017500000001134412054411111016314 00000000000000 Caveats in Using Wisdom - FFTW 3.3.3

Previous: Words of Wisdom-Saving Plans, Up: Other Important Topics

3.4 Caveats in Using Wisdom

For in much wisdom is much grief, and he that increaseth knowledge increaseth sorrow. [Ecclesiastes 1:18]

There are pitfalls to using wisdom, in that it can negate FFTW's ability to adapt to changing hardware and other conditions. For example, it would be perfectly possible to export wisdom from a program running on one processor and import it into a program running on another processor. Doing so, however, would mean that the second program would use plans optimized for the first processor, instead of the one it is running on.

It should be safe to reuse wisdom as long as the hardware and program binaries remain unchanged. (Actually, the optimal plan may change even between runs of the same binary on identical hardware, due to differences in the virtual memory environment, etcetera. Users seriously interested in performance should worry about this problem, too.) It is likely that, if the same wisdom is used for two different program binaries, even running on the same machine, the plans may be sub-optimal because of differing code alignments. It is therefore wise to recreate wisdom every time an application is recompiled. The more the underlying hardware and software changes between the creation of wisdom and its use, the greater grows the risk of sub-optimal plans.

Nevertheless, if the choice is between using FFTW_ESTIMATE or using possibly-suboptimal wisdom (created on the same machine, but for a different binary), the wisdom is likely to be better. For this reason, we provide a function to import wisdom from a standard system-wide location (/etc/fftw/wisdom on Unix): 

     int fftw_import_system_wisdom(void);

FFTW also provides a standalone program, fftw-wisdom (described by its own man page on Unix) with which users can create wisdom, e.g. for a canonical set of sizes to store in the system wisdom file. See Wisdom Utilities. fftw-3.3.3/doc/html/FFTW-Reference.html0000644000175000017500000000651012054411111014440 00000000000000 FFTW Reference - FFTW 3.3.3

Next: , Previous: Other Important Topics, Up: Top

4 FFTW Reference

This chapter provides a complete reference for all sequential (i.e., one-processor) FFTW functions. Parallel transforms are described in later chapters.

fftw-3.3.3/doc/html/Data-Types-and-Files.html0000644000175000017500000000573612054411111015562 00000000000000 Data Types and Files - FFTW 3.3.3

Next: , Previous: FFTW Reference, Up: FFTW Reference

4.1 Data Types and Files

All programs using FFTW should include its header file: 

     #include <fftw3.h>

You must also link to the FFTW library. On Unix, this means adding -lfftw3 -lm at the end of the link command.

fftw-3.3.3/doc/html/Complex-numbers.html0000644000175000017500000001111212054411111015050 00000000000000 Complex numbers - FFTW 3.3.3

Next: , Previous: Data Types and Files, Up: Data Types and Files

4.1.1 Complex numbers

The default FFTW interface uses double precision for all floating-point numbers, and defines a fftw_complex type to hold complex numbers as: 

     typedef double fftw_complex[2];

Here, the [0] element holds the real part and the [1] element holds the imaginary part.

Alternatively, if you have a C compiler (such as gcc) that supports the C99 revision of the ANSI C standard, you can use C's new native complex type (which is binary-compatible with the typedef above). In particular, if you #include <complex.h> before <fftw3.h>, then fftw_complex is defined to be the native complex type and you can manipulate it with ordinary arithmetic (e.g. x = y * (3+4*I), where x and y are fftw_complex and I is the standard symbol for the imaginary unit);

C++ has its own complex<T> template class, defined in the standard <complex> header file. Reportedly, the C++ standards committee has recently agreed to mandate that the storage format used for this type be binary-compatible with the C99 type, i.e. an array T[2] with consecutive real [0] and imaginary [1] parts. (See report http://www.open-std.org/jtc1/sc22/WG21/docs/papers/2002/n1388.pdf WG21/N1388.) Although not part of the official standard as of this writing, the proposal stated that: “This solution has been tested with all current major implementations of the standard library and shown to be working.” To the extent that this is true, if you have a variable complex<double> *x, you can pass it directly to FFTW via reinterpret_cast<fftw_complex*>(x). fftw-3.3.3/doc/html/Precision.html0000644000175000017500000001113412054411111013727 00000000000000 Precision - FFTW 3.3.3

Next: , Previous: Complex numbers, Up: Data Types and Files

4.1.2 Precision

You can install single and long-double precision versions of FFTW, which replace double with float and long double, respectively (see Installation and Customization). To use these interfaces, you:

  • Link to the single/long-double libraries; on Unix, -lfftw3f or -lfftw3l instead of (or in addition to) -lfftw3. (You can link to the different-precision libraries simultaneously.)
  • Include the same <fftw3.h> header file.
  • Replace all lowercase instances of ‘fftw_’ with ‘fftwf_’ or ‘fftwl_’ for single or long-double precision, respectively. (fftw_complex becomes fftwf_complex, fftw_execute becomes fftwf_execute, etcetera.)
  • Uppercase names, i.e. names beginning with ‘FFTW_’, remain the same.
  • Replace double with float or long double for subroutine parameters.

Depending upon your compiler and/or hardware, long double may not be any more precise than double (or may not be supported at all, although it is standard in C99).

We also support using the nonstandard __float128 quadruple-precision type provided by recent versions of gcc on 32- and 64-bit x86 hardware (see Installation and Customization). To use this type, link with -lfftw3q -lquadmath -lm (the libquadmath library provided by gcc is needed for quadruple-precision trigonometric functions) and use ‘fftwq_’ identifiers. fftw-3.3.3/doc/html/Memory-Allocation.html0000644000175000017500000001044012054411111015326 00000000000000 Memory Allocation - FFTW 3.3.3

Previous: Precision, Up: Data Types and Files

4.1.3 Memory Allocation

     void *fftw_malloc(size_t n);
     void fftw_free(void *p);

These are functions that behave identically to malloc and free, except that they guarantee that the returned pointer obeys any special alignment restrictions imposed by any algorithm in FFTW (e.g. for SIMD acceleration). See SIMD alignment and fftw_malloc.

Data allocated by fftw_malloc must be deallocated by fftw_free and not by the ordinary free.

These routines simply call through to your operating system's malloc or, if necessary, its aligned equivalent (e.g. memalign), so you normally need not worry about any significant time or space overhead. You are not required to use them to allocate your data, but we strongly recommend it.

Note: in C++, just as with ordinary malloc, you must typecast the output of fftw_malloc to whatever pointer type you are allocating.

We also provide the following two convenience functions to allocate real and complex arrays with n elements, which are equivalent to (double *) fftw_malloc(sizeof(double) * n) and (fftw_complex *) fftw_malloc(sizeof(fftw_complex) * n), respectively: 

     double *fftw_alloc_real(size_t n);
     fftw_complex *fftw_alloc_complex(size_t n);

The equivalent functions in other precisions allocate arrays of n elements in that precision. e.g. fftwf_alloc_real(n) is equivalent to (float *) fftwf_malloc(sizeof(float) * n). fftw-3.3.3/doc/html/Using-Plans.html0000644000175000017500000001605312054411111014141 00000000000000 Using Plans - FFTW 3.3.3

Next: , Previous: Data Types and Files, Up: FFTW Reference

4.2 Using Plans

Plans for all transform types in FFTW are stored as type fftw_plan (an opaque pointer type), and are created by one of the various planning routines described in the following sections. An fftw_plan contains all information necessary to compute the transform, including the pointers to the input and output arrays. 

     void fftw_execute(const fftw_plan plan);

This executes the plan, to compute the corresponding transform on the arrays for which it was planned (which must still exist). The plan is not modified, and fftw_execute can be called as many times as desired.

To apply a given plan to a different array, you can use the new-array execute interface. See New-array Execute Functions.

fftw_execute (and equivalents) is the only function in FFTW guaranteed to be thread-safe; see Thread safety.

This function: 

     void fftw_destroy_plan(fftw_plan plan);

deallocates the plan and all its associated data.

FFTW's planner saves some other persistent data, such as the accumulated wisdom and a list of algorithms available in the current configuration. If you want to deallocate all of that and reset FFTW to the pristine state it was in when you started your program, you can call: 

     void fftw_cleanup(void);

After calling fftw_cleanup, all existing plans become undefined, and you should not attempt to execute them nor to destroy them. You can however create and execute/destroy new plans, in which case FFTW starts accumulating wisdom information again.

fftw_cleanup does not deallocate your plans, however. To prevent memory leaks, you must still call fftw_destroy_plan before executing fftw_cleanup.

Occasionally, it may useful to know FFTW's internal “cost” metric that it uses to compare plans to one another; this cost is proportional to an execution time of the plan, in undocumented units, if the plan was created with the FFTW_MEASURE or other timing-based options, or alternatively is a heuristic cost function for FFTW_ESTIMATE plans. (The cost values of measured and estimated plans are not comparable, being in different units. Also, costs from different FFTW versions or the same version compiled differently may not be in the same units. Plans created from wisdom have a cost of 0 since no timing measurement is performed for them. Finally, certain problems for which only one top-level algorithm was possible may have required no measurements of the cost of the whole plan, in which case fftw_cost will also return 0.) The cost metric for a given plan is returned by: 

     double fftw_cost(const fftw_plan plan);

The following two routines are provided purely for academic purposes (that is, for entertainment). 

     void fftw_flops(const fftw_plan plan,
                     double *add, double *mul, double *fma);

Given a plan, set add, mul, and fma to an exact count of the number of floating-point additions, multiplications, and fused multiply-add operations involved in the plan's execution. The total number of floating-point operations (flops) is add + mul + 2*fma, or add + mul + fma if the hardware supports fused multiply-add instructions (although the number of FMA operations is only approximate because of compiler voodoo). (The number of operations should be an integer, but we use double to avoid overflowing int for large transforms; the arguments are of type double even for single and long-double precision versions of FFTW.) 

     void fftw_fprint_plan(const fftw_plan plan, FILE *output_file);
     void fftw_print_plan(const fftw_plan plan);

This outputs a “nerd-readable” representation of the plan to the given file or to stdout, respectively. fftw-3.3.3/doc/html/Basic-Interface.html0000644000175000017500000000751012054411111014716 00000000000000 Basic Interface - FFTW 3.3.3

Next: , Previous: Using Plans, Up: FFTW Reference

4.3 Basic Interface

Recall that the FFTW API is divided into three parts1: the basic interface computes a single transform of contiguous data, the advanced interface computes transforms of multiple or strided arrays, and the guru interface supports the most general data layouts, multiplicities, and strides. This section describes the the basic interface, which we expect to satisfy the needs of most users.

Footnotes

[1] Gallia est omnis divisa in partes tres (Julius Caesar).

fftw-3.3.3/doc/html/Complex-DFTs.html0000644000175000017500000001706512054411111014212 00000000000000 Complex DFTs - FFTW 3.3.3

Next: , Previous: Basic Interface, Up: Basic Interface

4.3.1 Complex DFTs

     fftw_plan fftw_plan_dft_1d(int n0,
                                fftw_complex *in, fftw_complex *out,
                                int sign, unsigned flags);
     fftw_plan fftw_plan_dft_2d(int n0, int n1,
                                fftw_complex *in, fftw_complex *out,
                                int sign, unsigned flags);
     fftw_plan fftw_plan_dft_3d(int n0, int n1, int n2,
                                fftw_complex *in, fftw_complex *out,
                                int sign, unsigned flags);
     fftw_plan fftw_plan_dft(int rank, const int *n,
                             fftw_complex *in, fftw_complex *out,
                             int sign, unsigned flags);

Plan a complex input/output discrete Fourier transform (DFT) in zero or more dimensions, returning an fftw_plan (see Using Plans).

Once you have created a plan for a certain transform type and parameters, then creating another plan of the same type and parameters, but for different arrays, is fast and shares constant data with the first plan (if it still exists).

The planner returns NULL if the plan cannot be created. In the standard FFTW distribution, the basic interface is guaranteed to return a non-NULL plan. A plan may be NULL, however, if you are using a customized FFTW configuration supporting a restricted set of transforms.

Arguments
  • rank is the rank of the transform (it should be the size of the array *n), and can be any non-negative integer. (See Complex Multi-Dimensional DFTs, for the definition of “rank”.) The ‘_1d’, ‘_2d’, and ‘_3d’ planners correspond to a rank of 1, 2, and 3, respectively. The rank may be zero, which is equivalent to a rank-1 transform of size 1, i.e. a copy of one number from input to output.
  • n0, n1, n2, or n[0..rank-1] (as appropriate for each routine) specify the size of the transform dimensions. They can be any positive integer.
    • Multi-dimensional arrays are stored in row-major order with dimensions: n0 x n1; or n0 x n1 x n2; or n[0] x n[1] x ... x n[rank-1]. See Multi-dimensional Array Format.
    • FFTW is best at handling sizes of the form 2a 3b 5c 7d 11e 13f,where e+f is either 0 or 1, and the other exponents are arbitrary. Other sizes are computed by means of a slow, general-purpose algorithm (which nevertheless retains O(n log n) performance even for prime sizes). It is possible to customize FFTW for different array sizes; see Installation and Customization. Transforms whose sizes are powers of 2 are especially fast.
  • in and out point to the input and output arrays of the transform, which may be the same (yielding an in-place transform). These arrays are overwritten during planning, unless FFTW_ESTIMATE is used in the flags. (The arrays need not be initialized, but they must be allocated.)

    If in == out, the transform is in-place and the input array is overwritten. If in != out, the two arrays must not overlap (but FFTW does not check for this condition).

  • sign is the sign of the exponent in the formula that defines the Fourier transform. It can be -1 (= FFTW_FORWARD) or +1 (= FFTW_BACKWARD).
  • flags is a bitwise OR (‘|’) of zero or more planner flags, as defined in Planner Flags.

FFTW computes an unnormalized transform: computing a forward followed by a backward transform (or vice versa) will result in the original data multiplied by the size of the transform (the product of the dimensions). For more information, see What FFTW Really Computes. fftw-3.3.3/doc/html/Planner-Flags.html0000644000175000017500000002130112054411111014422 00000000000000 Planner Flags - FFTW 3.3.3

Next: , Previous: Complex DFTs, Up: Basic Interface

4.3.2 Planner Flags

All of the planner routines in FFTW accept an integer flags argument, which is a bitwise OR (‘|’) of zero or more of the flag constants defined below. These flags control the rigor (and time) of the planning process, and can also impose (or lift) restrictions on the type of transform algorithm that is employed.

Important: the planner overwrites the input array during planning unless a saved plan (see Wisdom) is available for that problem, so you should initialize your input data after creating the plan. The only exceptions to this are the FFTW_ESTIMATE and FFTW_WISDOM_ONLY flags, as mentioned below.

In all cases, if wisdom is available for the given problem that was created with equal-or-greater planning rigor, then the more rigorous wisdom is used. For example, in FFTW_ESTIMATE mode any available wisdom is used, whereas in FFTW_PATIENT mode only wisdom created in patient or exhaustive mode can be used. See Words of Wisdom-Saving Plans.

Planning-rigor flags
  • FFTW_ESTIMATE specifies that, instead of actual measurements of different algorithms, a simple heuristic is used to pick a (probably sub-optimal) plan quickly. With this flag, the input/output arrays are not overwritten during planning.
  • FFTW_MEASURE tells FFTW to find an optimized plan by actually computing several FFTs and measuring their execution time. Depending on your machine, this can take some time (often a few seconds). FFTW_MEASURE is the default planning option.
  • FFTW_PATIENT is like FFTW_MEASURE, but considers a wider range of algorithms and often produces a “more optimal” plan (especially for large transforms), but at the expense of several times longer planning time (especially for large transforms).
  • FFTW_EXHAUSTIVE is like FFTW_PATIENT, but considers an even wider range of algorithms, including many that we think are unlikely to be fast, to produce the most optimal plan but with a substantially increased planning time.
  • FFTW_WISDOM_ONLY is a special planning mode in which the plan is only created if wisdom is available for the given problem, and otherwise a NULL plan is returned. This can be combined with other flags, e.g. ‘FFTW_WISDOM_ONLY | FFTW_PATIENT’ creates a plan only if wisdom is available that was created in FFTW_PATIENT or FFTW_EXHAUSTIVE mode. The FFTW_WISDOM_ONLY flag is intended for users who need to detect whether wisdom is available; for example, if wisdom is not available one may wish to allocate new arrays for planning so that user data is not overwritten.
Algorithm-restriction flags
  • FFTW_DESTROY_INPUT specifies that an out-of-place transform is allowed to overwrite its input array with arbitrary data; this can sometimes allow more efficient algorithms to be employed.
  • FFTW_PRESERVE_INPUT specifies that an out-of-place transform must not change its input array. This is ordinarily the default, except for c2r and hc2r (i.e. complex-to-real) transforms for which FFTW_DESTROY_INPUT is the default. In the latter cases, passing FFTW_PRESERVE_INPUT will attempt to use algorithms that do not destroy the input, at the expense of worse performance; for multi-dimensional c2r transforms, however, no input-preserving algorithms are implemented and the planner will return NULL if one is requested.
  • FFTW_UNALIGNED specifies that the algorithm may not impose any unusual alignment requirements on the input/output arrays (i.e. no SIMD may be used). This flag is normally not necessary, since the planner automatically detects misaligned arrays. The only use for this flag is if you want to use the new-array execute interface to execute a given plan on a different array that may not be aligned like the original. (Using fftw_malloc makes this flag unnecessary even then.)
Limiting planning time
     extern void fftw_set_timelimit(double seconds);

This function instructs FFTW to spend at most seconds seconds (approximately) in the planner. If seconds == FFTW_NO_TIMELIMIT (the default value, which is negative), then planning time is unbounded. Otherwise, FFTW plans with a progressively wider range of algorithms until the the given time limit is reached or the given range of algorithms is explored, returning the best available plan.

For example, specifying FFTW_PATIENT first plans in FFTW_ESTIMATE mode, then in FFTW_MEASURE mode, then finally (time permitting) in FFTW_PATIENT. If FFTW_EXHAUSTIVE is specified instead, the planner will further progress to FFTW_EXHAUSTIVE mode.

Note that the seconds argument specifies only a rough limit; in practice, the planner may use somewhat more time if the time limit is reached when the planner is in the middle of an operation that cannot be interrupted. At the very least, the planner will complete planning in FFTW_ESTIMATE mode (which is thus equivalent to a time limit of 0). fftw-3.3.3/doc/html/Real_002ddata-DFTs.html0000644000175000017500000002160112054411111015034 00000000000000 Real-data DFTs - FFTW 3.3.3

Next: , Previous: Planner Flags, Up: Basic Interface

4.3.3 Real-data DFTs

     fftw_plan fftw_plan_dft_r2c_1d(int n0,
                                    double *in, fftw_complex *out,
                                    unsigned flags);
     fftw_plan fftw_plan_dft_r2c_2d(int n0, int n1,
                                    double *in, fftw_complex *out,
                                    unsigned flags);
     fftw_plan fftw_plan_dft_r2c_3d(int n0, int n1, int n2,
                                    double *in, fftw_complex *out,
                                    unsigned flags);
     fftw_plan fftw_plan_dft_r2c(int rank, const int *n,
                                 double *in, fftw_complex *out,
                                 unsigned flags);

Plan a real-input/complex-output discrete Fourier transform (DFT) in zero or more dimensions, returning an fftw_plan (see Using Plans).

Once you have created a plan for a certain transform type and parameters, then creating another plan of the same type and parameters, but for different arrays, is fast and shares constant data with the first plan (if it still exists).

The planner returns NULL if the plan cannot be created. A non-NULL plan is always returned by the basic interface unless you are using a customized FFTW configuration supporting a restricted set of transforms, or if you use the FFTW_PRESERVE_INPUT flag with a multi-dimensional out-of-place c2r transform (see below).

Arguments
  • rank is the rank of the transform (it should be the size of the array *n), and can be any non-negative integer. (See Complex Multi-Dimensional DFTs, for the definition of “rank”.) The ‘_1d’, ‘_2d’, and ‘_3d’ planners correspond to a rank of 1, 2, and 3, respectively. The rank may be zero, which is equivalent to a rank-1 transform of size 1, i.e. a copy of one real number (with zero imaginary part) from input to output.
  • n0, n1, n2, or n[0..rank-1], (as appropriate for each routine) specify the size of the transform dimensions. They can be any positive integer. This is different in general from the physical array dimensions, which are described in Real-data DFT Array Format.
    • FFTW is best at handling sizes of the form 2a 3b 5c 7d 11e 13f,where e+f is either 0 or 1, and the other exponents are arbitrary. Other sizes are computed by means of a slow, general-purpose algorithm (which nevertheless retains O(n log n) performance even for prime sizes). (It is possible to customize FFTW for different array sizes; see Installation and Customization.) Transforms whose sizes are powers of 2 are especially fast, and it is generally beneficial for the last dimension of an r2c/c2r transform to be even.
  • in and out point to the input and output arrays of the transform, which may be the same (yielding an in-place transform). These arrays are overwritten during planning, unless FFTW_ESTIMATE is used in the flags. (The arrays need not be initialized, but they must be allocated.) For an in-place transform, it is important to remember that the real array will require padding, described in Real-data DFT Array Format.
  • flags is a bitwise OR (‘|’) of zero or more planner flags, as defined in Planner Flags.

The inverse transforms, taking complex input (storing the non-redundant half of a logically Hermitian array) to real output, are given by: 

     fftw_plan fftw_plan_dft_c2r_1d(int n0,
                                    fftw_complex *in, double *out,
                                    unsigned flags);
     fftw_plan fftw_plan_dft_c2r_2d(int n0, int n1,
                                    fftw_complex *in, double *out,
                                    unsigned flags);
     fftw_plan fftw_plan_dft_c2r_3d(int n0, int n1, int n2,
                                    fftw_complex *in, double *out,
                                    unsigned flags);
     fftw_plan fftw_plan_dft_c2r(int rank, const int *n,
                                 fftw_complex *in, double *out,
                                 unsigned flags);

The arguments are the same as for the r2c transforms, except that the input and output data formats are reversed.

FFTW computes an unnormalized transform: computing an r2c followed by a c2r transform (or vice versa) will result in the original data multiplied by the size of the transform (the product of the logical dimensions). An r2c transform produces the same output as a FFTW_FORWARD complex DFT of the same input, and a c2r transform is correspondingly equivalent to FFTW_BACKWARD. For more information, see What FFTW Really Computes. fftw-3.3.3/doc/html/Real_002ddata-DFT-Array-Format.html0000644000175000017500000001252112054411111017154 00000000000000 Real-data DFT Array Format - FFTW 3.3.3

Next: , Previous: Real-data DFTs, Up: Basic Interface

4.3.4 Real-data DFT Array Format

The output of a DFT of real data (r2c) contains symmetries that, in principle, make half of the outputs redundant (see What FFTW Really Computes). (Similarly for the input of an inverse c2r transform.) In practice, it is not possible to entirely realize these savings in an efficient and understandable format that generalizes to multi-dimensional transforms. Instead, the output of the r2c transforms is slightly over half of the output of the corresponding complex transform. We do not “pack” the data in any way, but store it as an ordinary array of fftw_complex values. In fact, this data is simply a subsection of what would be the array in the corresponding complex transform.

Specifically, for a real transform of d (= rank) dimensions n0 × n1 × n2 × … × nd-1, the complex data is an n0 × n1 × n2 × … × (nd-1/2 + 1) array of fftw_complex values in row-major order (with the division rounded down). That is, we only store the lower half (non-negative frequencies), plus one element, of the last dimension of the data from the ordinary complex transform. (We could have instead taken half of any other dimension, but implementation turns out to be simpler if the last, contiguous, dimension is used.)

For an out-of-place transform, the real data is simply an array with physical dimensions n0 × n1 × n2 × … × nd-1 in row-major order.

For an in-place transform, some complications arise since the complex data is slightly larger than the real data. In this case, the final dimension of the real data must be padded with extra values to accommodate the size of the complex data—two extra if the last dimension is even and one if it is odd. That is, the last dimension of the real data must physically contain 2 * (nd-1/2+1)double values (exactly enough to hold the complex data). This physical array size does not, however, change the logical array size—only nd-1values are actually stored in the last dimension, and nd-1is the last dimension passed to the planner. fftw-3.3.3/doc/html/Real_002dto_002dReal-Transforms.html0000644000175000017500000001726112054411111017443 00000000000000 Real-to-Real Transforms - FFTW 3.3.3

Next: , Previous: Real-data DFT Array Format, Up: Basic Interface

4.3.5 Real-to-Real Transforms

     fftw_plan fftw_plan_r2r_1d(int n, double *in, double *out,
                                fftw_r2r_kind kind, unsigned flags);
     fftw_plan fftw_plan_r2r_2d(int n0, int n1, double *in, double *out,
                                fftw_r2r_kind kind0, fftw_r2r_kind kind1,
                                unsigned flags);
     fftw_plan fftw_plan_r2r_3d(int n0, int n1, int n2,
                                double *in, double *out,
                                fftw_r2r_kind kind0,
                                fftw_r2r_kind kind1,
                                fftw_r2r_kind kind2,
                                unsigned flags);
     fftw_plan fftw_plan_r2r(int rank, const int *n, double *in, double *out,
                             const fftw_r2r_kind *kind, unsigned flags);

Plan a real input/output (r2r) transform of various kinds in zero or more dimensions, returning an fftw_plan (see Using Plans).

Once you have created a plan for a certain transform type and parameters, then creating another plan of the same type and parameters, but for different arrays, is fast and shares constant data with the first plan (if it still exists).

The planner returns NULL if the plan cannot be created. A non-NULL plan is always returned by the basic interface unless you are using a customized FFTW configuration supporting a restricted set of transforms, or for size-1 FFTW_REDFT00 kinds (which are not defined).

Arguments
  • rank is the dimensionality of the transform (it should be the size of the arrays *n and *kind), and can be any non-negative integer. The ‘_1d’, ‘_2d’, and ‘_3d’ planners correspond to a rank of 1, 2, and 3, respectively. A rank of zero is equivalent to a copy of one number from input to output.
  • n, or n0/n1/n2, or n[rank], respectively, gives the (physical) size of the transform dimensions. They can be any positive integer.
    • Multi-dimensional arrays are stored in row-major order with dimensions: n0 x n1; or n0 x n1 x n2; or n[0] x n[1] x ... x n[rank-1]. See Multi-dimensional Array Format.
    • FFTW is generally best at handling sizes of the form 2a 3b 5c 7d 11e 13f,where e+f is either 0 or 1, and the other exponents are arbitrary. Other sizes are computed by means of a slow, general-purpose algorithm (which nevertheless retains O(n log n) performance even for prime sizes). (It is possible to customize FFTW for different array sizes; see Installation and Customization.) Transforms whose sizes are powers of 2 are especially fast.
    • For a REDFT00 or RODFT00 transform kind in a dimension of size n, it is n-1 or n+1, respectively, that should be factorizable in the above form.
  • in and out point to the input and output arrays of the transform, which may be the same (yielding an in-place transform). These arrays are overwritten during planning, unless FFTW_ESTIMATE is used in the flags. (The arrays need not be initialized, but they must be allocated.)
  • kind, or kind0/kind1/kind2, or kind[rank], is the kind of r2r transform used for the corresponding dimension. The valid kind constants are described in Real-to-Real Transform Kinds. In a multi-dimensional transform, what is computed is the separable product formed by taking each transform kind along the corresponding dimension, one dimension after another.
  • flags is a bitwise OR (‘|’) of zero or more planner flags, as defined in Planner Flags.
fftw-3.3.3/doc/html/Real_002dto_002dReal-Transform-Kinds.html0000644000175000017500000001412612054411111020323 00000000000000 Real-to-Real Transform Kinds - FFTW 3.3.3

Previous: Real-to-Real Transforms, Up: Basic Interface

4.3.6 Real-to-Real Transform Kinds

FFTW currently supports 11 different r2r transform kinds, specified by one of the constants below. For the precise definitions of these transforms, see What FFTW Really Computes. For a more colloquial introduction to these transform kinds, see More DFTs of Real Data.

For dimension of size n, there is a corresponding “logical” dimension N that determines the normalization (and the optimal factorization); the formula for N is given for each kind below. Also, with each transform kind is listed its corrsponding inverse transform. FFTW computes unnormalized transforms: a transform followed by its inverse will result in the original data multiplied by N (or the product of the N's for each dimension, in multi-dimensions).

  • FFTW_R2HC computes a real-input DFT with output in “halfcomplex” format, i.e. real and imaginary parts for a transform of size n stored as:

    r0, r1, r2, ..., rn/2, i(n+1)/2-1, ..., i2, i1

    (Logical N=n, inverse is FFTW_HC2R.)
  • FFTW_HC2R computes the reverse of FFTW_R2HC, above. (Logical N=n, inverse is FFTW_R2HC.)
  • FFTW_DHT computes a discrete Hartley transform. (Logical N=n, inverse is FFTW_DHT.)
  • FFTW_REDFT00 computes an REDFT00 transform, i.e. a DCT-I. (Logical N=2*(n-1), inverse is FFTW_REDFT00.)
  • FFTW_REDFT10 computes an REDFT10 transform, i.e. a DCT-II (sometimes called “the” DCT). (Logical N=2*n, inverse is FFTW_REDFT01.)
  • FFTW_REDFT01 computes an REDFT01 transform, i.e. a DCT-III (sometimes called “the” IDCT, being the inverse of DCT-II). (Logical N=2*n, inverse is FFTW_REDFT=10.)
  • FFTW_REDFT11 computes an REDFT11 transform, i.e. a DCT-IV. (Logical N=2*n, inverse is FFTW_REDFT11.)
  • FFTW_RODFT00 computes an RODFT00 transform, i.e. a DST-I. (Logical N=2*(n+1), inverse is FFTW_RODFT00.)
  • FFTW_RODFT10 computes an RODFT10 transform, i.e. a DST-II. (Logical N=2*n, inverse is FFTW_RODFT01.)
  • FFTW_RODFT01 computes an RODFT01 transform, i.e. a DST-III. (Logical N=2*n, inverse is FFTW_RODFT=10.)
  • FFTW_RODFT11 computes an RODFT11 transform, i.e. a DST-IV. (Logical N=2*n, inverse is FFTW_RODFT11.)
fftw-3.3.3/doc/html/Advanced-Interface.html0000644000175000017500000000667012054411111015410 00000000000000 Advanced Interface - FFTW 3.3.3

Next: , Previous: Basic Interface, Up: FFTW Reference

4.4 Advanced Interface

FFTW's “advanced” interface supplements the basic interface with four new planner routines, providing a new level of flexibility: you can plan a transform of multiple arrays simultaneously, operate on non-contiguous (strided) data, and transform a subset of a larger multi-dimensional array. Other than these additional features, the planner operates in the same fashion as in the basic interface, and the resulting fftw_plan is used in the same way (see Using Plans).

fftw-3.3.3/doc/html/Advanced-Complex-DFTs.html0000644000175000017500000001710412054411111015707 00000000000000 Advanced Complex DFTs - FFTW 3.3.3

Next: , Previous: Advanced Interface, Up: Advanced Interface

4.4.1 Advanced Complex DFTs

     fftw_plan fftw_plan_many_dft(int rank, const int *n, int howmany,
                                  fftw_complex *in, const int *inembed,
                                  int istride, int idist,
                                  fftw_complex *out, const int *onembed,
                                  int ostride, int odist,
                                  int sign, unsigned flags);

This routine plans multiple multidimensional complex DFTs, and it extends the fftw_plan_dft routine (see Complex DFTs) to compute howmany transforms, each having rank rank and size n. In addition, the transform data need not be contiguous, but it may be laid out in memory with an arbitrary stride. To account for these possibilities, fftw_plan_many_dft adds the new parameters howmany, {i,o}nembed, {i,o}stride, and {i,o}dist. The FFTW basic interface (see Complex DFTs) provides routines specialized for ranks 1, 2, and 3, but the advanced interface handles only the general-rank case.

howmany is the number of transforms to compute. The resulting plan computes howmany transforms, where the input of the k-th transform is at location in+k*idist (in C pointer arithmetic), and its output is at location out+k*odist. Plans obtained in this way can often be faster than calling FFTW multiple times for the individual transforms. The basic fftw_plan_dft interface corresponds to howmany=1 (in which case the dist parameters are ignored).

Each of the howmany transforms has rank rank and size n, as in the basic interface. In addition, the advanced interface allows the input and output arrays of each transform to be row-major subarrays of larger rank-rank arrays, described by inembed and onembed parameters, respectively. {i,o}nembed must be arrays of length rank, and n should be elementwise less than or equal to {i,o}nembed. Passing NULL for an nembed parameter is equivalent to passing n (i.e. same physical and logical dimensions, as in the basic interface.)

The stride parameters indicate that the j-th element of the input or output arrays is located at j*istride or j*ostride, respectively. (For a multi-dimensional array, j is the ordinary row-major index.) When combined with the k-th transform in a howmany loop, from above, this means that the (j,k)-th element is at j*stride+k*dist. (The basic fftw_plan_dft interface corresponds to a stride of 1.)

For in-place transforms, the input and output stride and dist parameters should be the same; otherwise, the planner may return NULL.

Arrays n, inembed, and onembed are not used after this function returns. You can safely free or reuse them.

Examples: One transform of one 5 by 6 array contiguous in memory: 

        int rank = 2;
        int n[] = {5, 6};
        int howmany = 1;
        int idist = odist = 0; /* unused because howmany = 1 */
        int istride = ostride = 1; /* array is contiguous in memory */
        int *inembed = n, *onembed = n;

Transform of three 5 by 6 arrays, each contiguous in memory, stored in memory one after another: 

        int rank = 2;
        int n[] = {5, 6};
        int howmany = 3;
        int idist = odist = n[0]*n[1]; /* = 30, the distance in memory
                                          between the first element
                                          of the first array and the
                                          first element of the second array */
        int istride = ostride = 1; /* array is contiguous in memory */
        int *inembed = n, *onembed = n;

Transform each column of a 2d array with 10 rows and 3 columns: 

        int rank = 1; /* not 2: we are computing 1d transforms */
        int n[] = {10}; /* 1d transforms of length 10 */
        int howmany = 3;
        int idist = odist = 1;
        int istride = ostride = 3; /* distance between two elements in
                                      the same column */
        int *inembed = n, *onembed = n;
fftw-3.3.3/doc/html/Advanced-Real_002ddata-DFTs.html0000644000175000017500000001255712054411111016551 00000000000000 Advanced Real-data DFTs - FFTW 3.3.3

Next: , Previous: Advanced Complex DFTs, Up: Advanced Interface

4.4.2 Advanced Real-data DFTs

     fftw_plan fftw_plan_many_dft_r2c(int rank, const int *n, int howmany,
                                      double *in, const int *inembed,
                                      int istride, int idist,
                                      fftw_complex *out, const int *onembed,
                                      int ostride, int odist,
                                      unsigned flags);
     fftw_plan fftw_plan_many_dft_c2r(int rank, const int *n, int howmany,
                                      fftw_complex *in, const int *inembed,
                                      int istride, int idist,
                                      double *out, const int *onembed,
                                      int ostride, int odist,
                                      unsigned flags);

Like fftw_plan_many_dft, these two functions add howmany, nembed, stride, and dist parameters to the fftw_plan_dft_r2c and fftw_plan_dft_c2r functions, but otherwise behave the same as the basic interface.

The interpretation of howmany, stride, and dist are the same as for fftw_plan_many_dft, above. Note that the stride and dist for the real array are in units of double, and for the complex array are in units of fftw_complex.

If an nembed parameter is NULL, it is interpreted as what it would be in the basic interface, as described in Real-data DFT Array Format. That is, for the complex array the size is assumed to be the same as n, but with the last dimension cut roughly in half. For the real array, the size is assumed to be n if the transform is out-of-place, or n with the last dimension “padded” if the transform is in-place.

If an nembed parameter is non-NULL, it is interpreted as the physical size of the corresponding array, in row-major order, just as for fftw_plan_many_dft. In this case, each dimension of nembed should be >= what it would be in the basic interface (e.g. the halved or padded n).

Arrays n, inembed, and onembed are not used after this function returns. You can safely free or reuse them. fftw-3.3.3/doc/html/Advanced-Real_002dto_002dreal-Transforms.html0000644000175000017500000000715112054411111021203 00000000000000 Advanced Real-to-real Transforms - FFTW 3.3.3

Previous: Advanced Real-data DFTs, Up: Advanced Interface

4.4.3 Advanced Real-to-real Transforms

     fftw_plan fftw_plan_many_r2r(int rank, const int *n, int howmany,
                                  double *in, const int *inembed,
                                  int istride, int idist,
                                  double *out, const int *onembed,
                                  int ostride, int odist,
                                  const fftw_r2r_kind *kind, unsigned flags);

Like fftw_plan_many_dft, this functions adds howmany, nembed, stride, and dist parameters to the fftw_plan_r2r function, but otherwise behave the same as the basic interface. The interpretation of those additional parameters are the same as for fftw_plan_many_dft. (Of course, the stride and dist parameters are now in units of double, not fftw_complex.)

Arrays n, inembed, onembed, and kind are not used after this function returns. You can safely free or reuse them. fftw-3.3.3/doc/html/Guru-Interface.html0000644000175000017500000000770312054411111014623 00000000000000 Guru Interface - FFTW 3.3.3

Next: , Previous: Advanced Interface, Up: FFTW Reference

4.5 Guru Interface

The “guru” interface to FFTW is intended to expose as much as possible of the flexibility in the underlying FFTW architecture. It allows one to compute multi-dimensional “vectors” (loops) of multi-dimensional transforms, where each vector/transform dimension has an independent size and stride. One can also use more general complex-number formats, e.g. separate real and imaginary arrays.

For those users who require the flexibility of the guru interface, it is important that they pay special attention to the documentation lest they shoot themselves in the foot.

fftw-3.3.3/doc/html/Interleaved-and-split-arrays.html0000644000175000017500000000707512054411111017437 00000000000000 Interleaved and split arrays - FFTW 3.3.3

Next: , Previous: Guru Interface, Up: Guru Interface

4.5.1 Interleaved and split arrays

The guru interface supports two representations of complex numbers, which we call the interleaved and the split format.

The interleaved format is the same one used by the basic and advanced interfaces, and it is documented in Complex numbers. In the interleaved format, you provide pointers to the real part of a complex number, and the imaginary part understood to be stored in the next memory location.

The split format allows separate pointers to the real and imaginary parts of a complex array.

Technically, the interleaved format is redundant, because you can always express an interleaved array in terms of a split array with appropriate pointers and strides. On the other hand, the interleaved format is simpler to use, and it is common in practice. Hence, FFTW supports it as a special case. fftw-3.3.3/doc/html/Guru-vector-and-transform-sizes.html0000644000175000017500000001317512054411111020131 00000000000000 Guru vector and transform sizes - FFTW 3.3.3

Next: , Previous: Interleaved and split arrays, Up: Guru Interface

4.5.2 Guru vector and transform sizes

The guru interface introduces one basic new data structure, fftw_iodim, that is used to specify sizes and strides for multi-dimensional transforms and vectors: 

     typedef struct {
          int n;
          int is;
          int os;
     } fftw_iodim;

Here, n is the size of the dimension, and is and os are the strides of that dimension for the input and output arrays. (The stride is the separation of consecutive elements along this dimension.)

The meaning of the stride parameter depends on the type of the array that the stride refers to. If the array is interleaved complex, strides are expressed in units of complex numbers (fftw_complex). If the array is split complex or real, strides are expressed in units of real numbers (double). This convention is consistent with the usual pointer arithmetic in the C language. An interleaved array is denoted by a pointer p to fftw_complex, so that p+1 points to the next complex number. Split arrays are denoted by pointers to double, in which case pointer arithmetic operates in units of sizeof(double).

The guru planner interfaces all take a (rank, dims[rank]) pair describing the transform size, and a (howmany_rank, howmany_dims[howmany_rank]) pair describing the “vector” size (a multi-dimensional loop of transforms to perform), where dims and howmany_dims are arrays of fftw_iodim.

For example, the howmany parameter in the advanced complex-DFT interface corresponds to howmany_rank = 1, howmany_dims[0].n = howmany, howmany_dims[0].is = idist, and howmany_dims[0].os = odist. (To compute a single transform, you can just use howmany_rank = 0.)

A row-major multidimensional array with dimensions n[rank] (see Row-major Format) corresponds to dims[i].n = n[i] and the recurrence dims[i].is = n[i+1] * dims[i+1].is (similarly for os). The stride of the last (i=rank-1) dimension is the overall stride of the array. e.g. to be equivalent to the advanced complex-DFT interface, you would have dims[rank-1].is = istride and dims[rank-1].os = ostride.

In general, we only guarantee FFTW to return a non-NULL plan if the vector and transform dimensions correspond to a set of distinct indices, and for in-place transforms the input/output strides should be the same. fftw-3.3.3/doc/html/Guru-Complex-DFTs.html0000644000175000017500000001372412054411111015130 00000000000000 Guru Complex DFTs - FFTW 3.3.3

Next: , Previous: Guru vector and transform sizes, Up: Guru Interface

4.5.3 Guru Complex DFTs

     fftw_plan fftw_plan_guru_dft(
          int rank, const fftw_iodim *dims,
          int howmany_rank, const fftw_iodim *howmany_dims,
          fftw_complex *in, fftw_complex *out,
          int sign, unsigned flags);
     
     fftw_plan fftw_plan_guru_split_dft(
          int rank, const fftw_iodim *dims,
          int howmany_rank, const fftw_iodim *howmany_dims,
          double *ri, double *ii, double *ro, double *io,
          unsigned flags);

These two functions plan a complex-data, multi-dimensional DFT for the interleaved and split format, respectively. Transform dimensions are given by (rank, dims) over a multi-dimensional vector (loop) of dimensions (howmany_rank, howmany_dims). dims and howmany_dims should point to fftw_iodim arrays of length rank and howmany_rank, respectively.

flags is a bitwise OR (‘|’) of zero or more planner flags, as defined in Planner Flags.

In the fftw_plan_guru_dft function, the pointers in and out point to the interleaved input and output arrays, respectively. The sign can be either -1 (= FFTW_FORWARD) or +1 (= FFTW_BACKWARD). If the pointers are equal, the transform is in-place.

In the fftw_plan_guru_split_dft function, ri and ii point to the real and imaginary input arrays, and ro and io point to the real and imaginary output arrays. The input and output pointers may be the same, indicating an in-place transform. For example, for fftw_complex pointers in and out, the corresponding parameters are: 

     ri = (double *) in;
     ii = (double *) in + 1;
     ro = (double *) out;
     io = (double *) out + 1;

Because fftw_plan_guru_split_dft accepts split arrays, strides are expressed in units of double. For a contiguous fftw_complex array, the overall stride of the transform should be 2, the distance between consecutive real parts or between consecutive imaginary parts; see Guru vector and transform sizes. Note that the dimension strides are applied equally to the real and imaginary parts; real and imaginary arrays with different strides are not supported.

There is no sign parameter in fftw_plan_guru_split_dft. This function always plans for an FFTW_FORWARD transform. To plan for an FFTW_BACKWARD transform, you can exploit the identity that the backwards DFT is equal to the forwards DFT with the real and imaginary parts swapped. For example, in the case of the fftw_complex arrays above, the FFTW_BACKWARD transform is computed by the parameters: 

     ri = (double *) in + 1;
     ii = (double *) in;
     ro = (double *) out + 1;
     io = (double *) out;
fftw-3.3.3/doc/html/Guru-Real_002ddata-DFTs.html0000644000175000017500000001447012054411111015762 00000000000000 Guru Real-data DFTs - FFTW 3.3.3

Next: , Previous: Guru Complex DFTs, Up: Guru Interface

4.5.4 Guru Real-data DFTs

     fftw_plan fftw_plan_guru_dft_r2c(
          int rank, const fftw_iodim *dims,
          int howmany_rank, const fftw_iodim *howmany_dims,
          double *in, fftw_complex *out,
          unsigned flags);
     
     fftw_plan fftw_plan_guru_split_dft_r2c(
          int rank, const fftw_iodim *dims,
          int howmany_rank, const fftw_iodim *howmany_dims,
          double *in, double *ro, double *io,
          unsigned flags);
     
     fftw_plan fftw_plan_guru_dft_c2r(
          int rank, const fftw_iodim *dims,
          int howmany_rank, const fftw_iodim *howmany_dims,
          fftw_complex *in, double *out,
          unsigned flags);
     
     fftw_plan fftw_plan_guru_split_dft_c2r(
          int rank, const fftw_iodim *dims,
          int howmany_rank, const fftw_iodim *howmany_dims,
          double *ri, double *ii, double *out,
          unsigned flags);

Plan a real-input (r2c) or real-output (c2r), multi-dimensional DFT with transform dimensions given by (rank, dims) over a multi-dimensional vector (loop) of dimensions (howmany_rank, howmany_dims). dims and howmany_dims should point to fftw_iodim arrays of length rank and howmany_rank, respectively. As for the basic and advanced interfaces, an r2c transform is FFTW_FORWARD and a c2r transform is FFTW_BACKWARD.

The last dimension of dims is interpreted specially: that dimension of the real array has size dims[rank-1].n, but that dimension of the complex array has size dims[rank-1].n/2+1 (division rounded down). The strides, on the other hand, are taken to be exactly as specified. It is up to the user to specify the strides appropriately for the peculiar dimensions of the data, and we do not guarantee that the planner will succeed (return non-NULL) for any dimensions other than those described in Real-data DFT Array Format and generalized in Advanced Real-data DFTs. (That is, for an in-place transform, each individual dimension should be able to operate in place.)

in and out point to the input and output arrays for r2c and c2r transforms, respectively. For split arrays, ri and ii point to the real and imaginary input arrays for a c2r transform, and ro and io point to the real and imaginary output arrays for an r2c transform. in and ro or ri and out may be the same, indicating an in-place transform. (In-place transforms where in and io or ii and out are the same are not currently supported.)

flags is a bitwise OR (‘|’) of zero or more planner flags, as defined in Planner Flags.

In-place transforms of rank greater than 1 are currently only supported for interleaved arrays. For split arrays, the planner will return NULL. fftw-3.3.3/doc/html/Guru-Real_002dto_002dreal-Transforms.html0000644000175000017500000001032412054411111020414 00000000000000 Guru Real-to-real Transforms - FFTW 3.3.3

Next: , Previous: Guru Real-data DFTs, Up: Guru Interface

4.5.5 Guru Real-to-real Transforms

     fftw_plan fftw_plan_guru_r2r(int rank, const fftw_iodim *dims,
                                  int howmany_rank,
                                  const fftw_iodim *howmany_dims,
                                  double *in, double *out,
                                  const fftw_r2r_kind *kind,
                                  unsigned flags);

Plan a real-to-real (r2r) multi-dimensional FFTW_FORWARD transform with transform dimensions given by (rank, dims) over a multi-dimensional vector (loop) of dimensions (howmany_rank, howmany_dims). dims and howmany_dims should point to fftw_iodim arrays of length rank and howmany_rank, respectively.

The transform kind of each dimension is given by the kind parameter, which should point to an array of length rank. Valid fftw_r2r_kind constants are given in Real-to-Real Transform Kinds.

in and out point to the real input and output arrays; they may be the same, indicating an in-place transform.

flags is a bitwise OR (‘|’) of zero or more planner flags, as defined in Planner Flags. fftw-3.3.3/doc/html/64_002dbit-Guru-Interface.html0000644000175000017500000001226612054411111016276 00000000000000 64-bit Guru Interface - FFTW 3.3.3

Previous: Guru Real-to-real Transforms, Up: Guru Interface

4.5.6 64-bit Guru Interface

When compiled in 64-bit mode on a 64-bit architecture (where addresses are 64 bits wide), FFTW uses 64-bit quantities internally for all transform sizes, strides, and so on—you don't have to do anything special to exploit this. However, in the ordinary FFTW interfaces, you specify the transform size by an int quantity, which is normally only 32 bits wide. This means that, even though FFTW is using 64-bit sizes internally, you cannot specify a single transform dimension larger than 231−1numbers.

We expect that few users will require transforms larger than this, but, for those who do, we provide a 64-bit version of the guru interface in which all sizes are specified as integers of type ptrdiff_t instead of int. (ptrdiff_t is a signed integer type defined by the C standard to be wide enough to represent address differences, and thus must be at least 64 bits wide on a 64-bit machine.) We stress that there is no performance advantage to using this interface—the same internal FFTW code is employed regardless—and it is only necessary if you want to specify very large transform sizes.

In particular, the 64-bit guru interface is a set of planner routines that are exactly the same as the guru planner routines, except that they are named with ‘guru64’ instead of ‘guru’ and they take arguments of type fftw_iodim64 instead of fftw_iodim. For example, instead of fftw_plan_guru_dft, we have fftw_plan_guru64_dft. 

     fftw_plan fftw_plan_guru64_dft(
          int rank, const fftw_iodim64 *dims,
          int howmany_rank, const fftw_iodim64 *howmany_dims,
          fftw_complex *in, fftw_complex *out,
          int sign, unsigned flags);

The fftw_iodim64 type is similar to fftw_iodim, with the same interpretation, except that it uses type ptrdiff_t instead of type int. 

     typedef struct {
          ptrdiff_t n;
          ptrdiff_t is;
          ptrdiff_t os;
     } fftw_iodim64;

Every other ‘fftw_plan_guru’ function also has a ‘fftw_plan_guru64’ equivalent, but we do not repeat their documentation here since they are identical to the 32-bit versions except as noted above. fftw-3.3.3/doc/html/New_002darray-Execute-Functions.html0000644000175000017500000001630712054411111017666 00000000000000 New-array Execute Functions - FFTW 3.3.3

Next: , Previous: Guru Interface, Up: FFTW Reference

4.6 New-array Execute Functions

Normally, one executes a plan for the arrays with which the plan was created, by calling fftw_execute(plan) as described in Using Plans. However, it is possible for sophisticated users to apply a given plan to a different array using the “new-array execute” functions detailed below, provided that the following conditions are met:

  • The array size, strides, etcetera are the same (since those are set by the plan).
  • The input and output arrays are the same (in-place) or different (out-of-place) if the plan was originally created to be in-place or out-of-place, respectively.
  • For split arrays, the separations between the real and imaginary parts, ii-ri and io-ro, are the same as they were for the input and output arrays when the plan was created. (This condition is automatically satisfied for interleaved arrays.)
  • The alignment of the new input/output arrays is the same as that of the input/output arrays when the plan was created, unless the plan was created with the FFTW_UNALIGNED flag. Here, the alignment is a platform-dependent quantity (for example, it is the address modulo 16 if SSE SIMD instructions are used, but the address modulo 4 for non-SIMD single-precision FFTW on the same machine). In general, only arrays allocated with fftw_malloc are guaranteed to be equally aligned (see SIMD alignment and fftw_malloc).

The alignment issue is especially critical, because if you don't use fftw_malloc then you may have little control over the alignment of arrays in memory. For example, neither the C++ new function nor the Fortran allocate statement provide strong enough guarantees about data alignment. If you don't use fftw_malloc, therefore, you probably have to use FFTW_UNALIGNED (which disables most SIMD support). If possible, it is probably better for you to simply create multiple plans (creating a new plan is quick once one exists for a given size), or better yet re-use the same array for your transforms.

If you are tempted to use the new-array execute interface because you want to transform a known bunch of arrays of the same size, you should probably go use the advanced interface instead (see Advanced Interface)).

The new-array execute functions are: 

     void fftw_execute_dft(
          const fftw_plan p,
          fftw_complex *in, fftw_complex *out);
     
     void fftw_execute_split_dft(
          const fftw_plan p,
          double *ri, double *ii, double *ro, double *io);
     
     void fftw_execute_dft_r2c(
          const fftw_plan p,
          double *in, fftw_complex *out);
     
     void fftw_execute_split_dft_r2c(
          const fftw_plan p,
          double *in, double *ro, double *io);
     
     void fftw_execute_dft_c2r(
          const fftw_plan p,
          fftw_complex *in, double *out);
     
     void fftw_execute_split_dft_c2r(
          const fftw_plan p,
          double *ri, double *ii, double *out);
     
     void fftw_execute_r2r(
          const fftw_plan p,
          double *in, double *out);

These execute the plan to compute the corresponding transform on the input/output arrays specified by the subsequent arguments. The input/output array arguments have the same meanings as the ones passed to the guru planner routines in the preceding sections. The plan is not modified, and these routines can be called as many times as desired, or intermixed with calls to the ordinary fftw_execute.

The plan must have been created for the transform type corresponding to the execute function, e.g. it must be a complex-DFT plan for fftw_execute_dft. Any of the planner routines for that transform type, from the basic to the guru interface, could have been used to create the plan, however. fftw-3.3.3/doc/html/Wisdom.html0000644000175000017500000000622312054411111013241 00000000000000 Wisdom - FFTW 3.3.3

Next: , Previous: New-array Execute Functions, Up: FFTW Reference

4.7 Wisdom

This section documents the FFTW mechanism for saving and restoring plans from disk. This mechanism is called wisdom.

fftw-3.3.3/doc/html/Wisdom-Export.html0000644000175000017500000001133012054411111014513 00000000000000 Wisdom Export - FFTW 3.3.3

Next: , Previous: Wisdom, Up: Wisdom

4.7.1 Wisdom Export

     int fftw_export_wisdom_to_filename(const char *filename);
     void fftw_export_wisdom_to_file(FILE *output_file);
     char *fftw_export_wisdom_to_string(void);
     void fftw_export_wisdom(void (*write_char)(char c, void *), void *data);

These functions allow you to export all currently accumulated wisdom in a form from which it can be later imported and restored, even during a separate run of the program. (See Words of Wisdom-Saving Plans.) The current store of wisdom is not affected by calling any of these routines.

fftw_export_wisdom exports the wisdom to any output medium, as specified by the callback function write_char. write_char is a putc-like function that writes the character c to some output; its second parameter is the data pointer passed to fftw_export_wisdom. For convenience, the following three “wrapper” routines are provided:

fftw_export_wisdom_to_filename writes wisdom to a file named filename (which is created or overwritten), returning 1 on success and 0 on failure. A lower-level function, which requires you to open and close the file yourself (e.g. if you want to write wisdom to a portion of a larger file) is fftw_export_wisdom_to_file. This writes the wisdom to the current position in output_file, which should be open with write permission; upon exit, the file remains open and is positioned at the end of the wisdom data.

fftw_export_wisdom_to_string returns a pointer to a NULL-terminated string holding the wisdom data. This string is dynamically allocated, and it is the responsibility of the caller to deallocate it with free when it is no longer needed.

All of these routines export the wisdom in the same format, which we will not document here except to say that it is LISP-like ASCII text that is insensitive to white space. fftw-3.3.3/doc/html/Wisdom-Import.html0000644000175000017500000001176612054411111014521 00000000000000 Wisdom Import - FFTW 3.3.3

Next: , Previous: Wisdom Export, Up: Wisdom

4.7.2 Wisdom Import

     int fftw_import_system_wisdom(void);
     int fftw_import_wisdom_from_filename(const char *filename);
     int fftw_import_wisdom_from_string(const char *input_string);
     int fftw_import_wisdom(int (*read_char)(void *), void *data);

These functions import wisdom into a program from data stored by the fftw_export_wisdom functions above. (See Words of Wisdom-Saving Plans.) The imported wisdom replaces any wisdom already accumulated by the running program.

fftw_import_wisdom imports wisdom from any input medium, as specified by the callback function read_char. read_char is a getc-like function that returns the next character in the input; its parameter is the data pointer passed to fftw_import_wisdom. If the end of the input data is reached (which should never happen for valid data), read_char should return EOF (as defined in <stdio.h>). For convenience, the following three “wrapper” routines are provided:

fftw_import_wisdom_from_filename reads wisdom from a file named filename. A lower-level function, which requires you to open and close the file yourself (e.g. if you want to read wisdom from a portion of a larger file) is fftw_import_wisdom_from_file. This reads wisdom from the current position in input_file (which should be open with read permission); upon exit, the file remains open, but the position of the read pointer is unspecified.

fftw_import_wisdom_from_string reads wisdom from the NULL-terminated string input_string.

fftw_import_system_wisdom reads wisdom from an implementation-defined standard file (/etc/fftw/wisdom on Unix and GNU systems).

The return value of these import routines is 1 if the wisdom was read successfully and 0 otherwise. Note that, in all of these functions, any data in the input stream past the end of the wisdom data is simply ignored. fftw-3.3.3/doc/html/Forgetting-Wisdom.html0000644000175000017500000000537712054411111015360 00000000000000 Forgetting Wisdom - FFTW 3.3.3

Next: , Previous: Wisdom Import, Up: Wisdom

4.7.3 Forgetting Wisdom

     void fftw_forget_wisdom(void);

Calling fftw_forget_wisdom causes all accumulated wisdom to be discarded and its associated memory to be freed. (New wisdom can still be gathered subsequently, however.) fftw-3.3.3/doc/html/Wisdom-Utilities.html0000644000175000017500000000740112054411111015211 00000000000000 Wisdom Utilities - FFTW 3.3.3

Previous: Forgetting Wisdom, Up: Wisdom

4.7.4 Wisdom Utilities

FFTW includes two standalone utility programs that deal with wisdom. We merely summarize them here, since they come with their own man pages for Unix and GNU systems (with HTML versions on our web site).

The first program is fftw-wisdom (or fftwf-wisdom in single precision, etcetera), which can be used to create a wisdom file containing plans for any of the transform sizes and types supported by FFTW. It is preferable to create wisdom directly from your executable (see Caveats in Using Wisdom), but this program is useful for creating global wisdom files for fftw_import_system_wisdom.

The second program is fftw-wisdom-to-conf, which takes a wisdom file as input and produces a configuration routine as output. The latter is a C subroutine that you can compile and link into your program, replacing a routine of the same name in the FFTW library, that determines which parts of FFTW are callable by your program. fftw-wisdom-to-conf produces a configuration routine that links to only those parts of FFTW needed by the saved plans in the wisdom, greatly reducing the size of statically linked executables (which should only attempt to create plans corresponding to those in the wisdom, however). fftw-3.3.3/doc/html/What-FFTW-Really-Computes.html0000644000175000017500000000711512054411111016472 00000000000000 What FFTW Really Computes - FFTW 3.3.3

Previous: Wisdom, Up: FFTW Reference

4.8 What FFTW Really Computes

In this section, we provide precise mathematical definitions for the transforms that FFTW computes. These transform definitions are fairly standard, but some authors follow slightly different conventions for the normalization of the transform (the constant factor in front) and the sign of the complex exponent. We begin by presenting the one-dimensional (1d) transform definitions, and then give the straightforward extension to multi-dimensional transforms.

fftw-3.3.3/doc/html/The-1d-Discrete-Fourier-Transform-_0028DFT_0029.html0000644000175000017500000001003612054411111021676 00000000000000 The 1d Discrete Fourier Transform (DFT) - FFTW 3.3.3

Next: , Previous: What FFTW Really Computes, Up: What FFTW Really Computes

4.8.1 The 1d Discrete Fourier Transform (DFT)

The forward (FFTW_FORWARD) discrete Fourier transform (DFT) of a 1d complex array X of size n computes an array Y, where:

.
The backward (FFTW_BACKWARD) DFT computes:
.

FFTW computes an unnormalized transform, in that there is no coefficient in front of the summation in the DFT. In other words, applying the forward and then the backward transform will multiply the input by n.

From above, an FFTW_FORWARD transform corresponds to a sign of -1 in the exponent of the DFT. Note also that we use the standard “in-order” output ordering—the k-th output corresponds to the frequency k/n (or k/T, where T is your total sampling period). For those who like to think in terms of positive and negative frequencies, this means that the positive frequencies are stored in the first half of the output and the negative frequencies are stored in backwards order in the second half of the output. (The frequency -k/n is the same as the frequency (n-k)/n.) fftw-3.3.3/doc/html/The-1d-Real_002ddata-DFT.html0000644000175000017500000001120012054411111015663 00000000000000 The 1d Real-data DFT - FFTW 3.3.3

Next: , Previous: The 1d Discrete Fourier Transform (DFT), Up: What FFTW Really Computes

4.8.2 The 1d Real-data DFT

The real-input (r2c) DFT in FFTW computes the forward transform Y of the size n real array X, exactly as defined above, i.e.

.
This output array Y can easily be shown to possess the “Hermitian” symmetry Yk = Yn-k*,where we take Y to be periodic so that Yn = Y0.

As a result of this symmetry, half of the output Y is redundant (being the complex conjugate of the other half), and so the 1d r2c transforms only output elements 0...n/2 of Y (n/2+1 complex numbers), where the division by 2 is rounded down.

Moreover, the Hermitian symmetry implies that Y0and, if n is even, the Yn/2element, are purely real. So, for the R2HC r2r transform, these elements are not stored in the halfcomplex output format.

The c2r and H2RC r2r transforms compute the backward DFT of the complex array X with Hermitian symmetry, stored in the r2c/R2HC output formats, respectively, where the backward transform is defined exactly as for the complex case:

.
The outputs Y of this transform can easily be seen to be purely real, and are stored as an array of real numbers.

Like FFTW's complex DFT, these transforms are unnormalized. In other words, applying the real-to-complex (forward) and then the complex-to-real (backward) transform will multiply the input by n. fftw-3.3.3/doc/html/1d-Real_002deven-DFTs-_0028DCTs_0029.html0000644000175000017500000001507712054411111017312 00000000000000 1d Real-even DFTs (DCTs) - FFTW 3.3.3

Next: , Previous: The 1d Real-data DFT, Up: What FFTW Really Computes

4.8.3 1d Real-even DFTs (DCTs)

The Real-even symmetry DFTs in FFTW are exactly equivalent to the unnormalized forward (and backward) DFTs as defined above, where the input array X of length N is purely real and is also even symmetry. In this case, the output array is likewise real and even symmetry.

For the case of REDFT00, this even symmetry means that Xj = XN-j,where we take X to be periodic so that XN = X0. Because of this redundancy, only the first n real numbers are actually stored, where N = 2(n-1).

The proper definition of even symmetry for REDFT10, REDFT01, and REDFT11 transforms is somewhat more intricate because of the shifts by 1/2 of the input and/or output, although the corresponding boundary conditions are given in Real even/odd DFTs (cosine/sine transforms). Because of the even symmetry, however, the sine terms in the DFT all cancel and the remaining cosine terms are written explicitly below. This formulation often leads people to call such a transform a discrete cosine transform (DCT), although it is really just a special case of the DFT.

In each of the definitions below, we transform a real array X of length n to a real array Y of length n:

REDFT00 (DCT-I)

An REDFT00 transform (type-I DCT) in FFTW is defined by:

.
Note that this transform is not defined for n=1. For n=2, the summation term above is dropped as you might expect.
REDFT10 (DCT-II)

An REDFT10 transform (type-II DCT, sometimes called “the” DCT) in FFTW is defined by:

.
REDFT01 (DCT-III)

An REDFT01 transform (type-III DCT) in FFTW is defined by:

.
In the case of n=1, this reduces to Y0 = X0. Up to a scale factor (see below), this is the inverse of REDFT10 (“the” DCT), and so the REDFT01 (DCT-III) is sometimes called the “IDCT”.
REDFT11 (DCT-IV)

An REDFT11 transform (type-IV DCT) in FFTW is defined by:

.
Inverses and Normalization

These definitions correspond directly to the unnormalized DFTs used elsewhere in FFTW (hence the factors of 2 in front of the summations). The unnormalized inverse of REDFT00 is REDFT00, of REDFT10 is REDFT01 and vice versa, and of REDFT11 is REDFT11. Each unnormalized inverse results in the original array multiplied by N, where N is the logical DFT size. For REDFT00, N=2(n-1) (note that n=1 is not defined); otherwise, N=2n.

In defining the discrete cosine transform, some authors also include additional factors of √2(or its inverse) multiplying selected inputs and/or outputs. This is a mostly cosmetic change that makes the transform orthogonal, but sacrifices the direct equivalence to a symmetric DFT. fftw-3.3.3/doc/html/1d-Real_002dodd-DFTs-_0028DSTs_0029.html0000644000175000017500000001450712054411111017140 00000000000000 1d Real-odd DFTs (DSTs) - FFTW 3.3.3

Next: , Previous: 1d Real-even DFTs (DCTs), Up: What FFTW Really Computes

4.8.4 1d Real-odd DFTs (DSTs)

The Real-odd symmetry DFTs in FFTW are exactly equivalent to the unnormalized forward (and backward) DFTs as defined above, where the input array X of length N is purely real and is also odd symmetry. In this case, the output is odd symmetry and purely imaginary.

For the case of RODFT00, this odd symmetry means that Xj = -XN-j,where we take X to be periodic so that XN = X0. Because of this redundancy, only the first n real numbers starting at j=1 are actually stored (the j=0 element is zero), where N = 2(n+1).

The proper definition of odd symmetry for RODFT10, RODFT01, and RODFT11 transforms is somewhat more intricate because of the shifts by 1/2 of the input and/or output, although the corresponding boundary conditions are given in Real even/odd DFTs (cosine/sine transforms). Because of the odd symmetry, however, the cosine terms in the DFT all cancel and the remaining sine terms are written explicitly below. This formulation often leads people to call such a transform a discrete sine transform (DST), although it is really just a special case of the DFT.

In each of the definitions below, we transform a real array X of length n to a real array Y of length n:

RODFT00 (DST-I)

An RODFT00 transform (type-I DST) in FFTW is defined by:

.
RODFT10 (DST-II)

An RODFT10 transform (type-II DST) in FFTW is defined by:

.
RODFT01 (DST-III)

An RODFT01 transform (type-III DST) in FFTW is defined by:

.
In the case of n=1, this reduces to Y0 = X0.
RODFT11 (DST-IV)

An RODFT11 transform (type-IV DST) in FFTW is defined by:

.
Inverses and Normalization

These definitions correspond directly to the unnormalized DFTs used elsewhere in FFTW (hence the factors of 2 in front of the summations). The unnormalized inverse of RODFT00 is RODFT00, of RODFT10 is RODFT01 and vice versa, and of RODFT11 is RODFT11. Each unnormalized inverse results in the original array multiplied by N, where N is the logical DFT size. For RODFT00, N=2(n+1); otherwise, N=2n.

In defining the discrete sine transform, some authors also include additional factors of √2(or its inverse) multiplying selected inputs and/or outputs. This is a mostly cosmetic change that makes the transform orthogonal, but sacrifices the direct equivalence to an antisymmetric DFT. fftw-3.3.3/doc/html/1d-Discrete-Hartley-Transforms-_0028DHTs_0029.html0000644000175000017500000000656312054411111021537 00000000000000 1d Discrete Hartley Transforms (DHTs) - FFTW 3.3.3

Next: , Previous: 1d Real-odd DFTs (DSTs), Up: What FFTW Really Computes

4.8.5 1d Discrete Hartley Transforms (DHTs)

The discrete Hartley transform (DHT) of a 1d real array X of size n computes a real array Y of the same size, where:

.

FFTW computes an unnormalized transform, in that there is no coefficient in front of the summation in the DHT. In other words, applying the transform twice (the DHT is its own inverse) will multiply the input by n. fftw-3.3.3/doc/html/Multi_002ddimensional-Transforms.html0000644000175000017500000001315512054411111020177 00000000000000 Multi-dimensional Transforms - FFTW 3.3.3

Previous: 1d Discrete Hartley Transforms (DHTs), Up: What FFTW Really Computes

4.8.6 Multi-dimensional Transforms

The multi-dimensional transforms of FFTW, in general, compute simply the separable product of the given 1d transform along each dimension of the array. Since each of these transforms is unnormalized, computing the forward followed by the backward/inverse multi-dimensional transform will result in the original array scaled by the product of the normalization factors for each dimension (e.g. the product of the dimension sizes, for a multi-dimensional DFT).

The definition of FFTW's multi-dimensional DFT of real data (r2c) deserves special attention. In this case, we logically compute the full multi-dimensional DFT of the input data; since the input data are purely real, the output data have the Hermitian symmetry and therefore only one non-redundant half need be stored. More specifically, for an n0 × n1 × n2 × … × nd-1 multi-dimensional real-input DFT, the full (logical) complex output array Y[k0, k1, ..., kd-1]has the symmetry: Y[k0, k1, ..., kd-1] = Y[n0 - k0, n1 - k1, ..., nd-1 - kd-1]*(where each dimension is periodic). Because of this symmetry, we only store the kd-1 = 0...nd-1/2+1elements of the last dimension (division by 2 is rounded down). (We could instead have cut any other dimension in half, but the last dimension proved computationally convenient.) This results in the peculiar array format described in more detail by Real-data DFT Array Format.

The multi-dimensional c2r transform is simply the unnormalized inverse of the r2c transform. i.e. it is the same as FFTW's complex backward multi-dimensional DFT, operating on a Hermitian input array in the peculiar format mentioned above and outputting a real array (since the DFT output is purely real).

We should remind the user that the separable product of 1d transforms along each dimension, as computed by FFTW, is not always the same thing as the usual multi-dimensional transform. A multi-dimensional R2HC (or HC2R) transform is not identical to the multi-dimensional DFT, requiring some post-processing to combine the requisite real and imaginary parts, as was described in The Halfcomplex-format DFT. Likewise, FFTW's multidimensional FFTW_DHT r2r transform is not the same thing as the logical multi-dimensional discrete Hartley transform defined in the literature, as discussed in The Discrete Hartley Transform. fftw-3.3.3/doc/html/Multi_002dthreaded-FFTW.html0000644000175000017500000001045112054411111016061 00000000000000 Multi-threaded FFTW - FFTW 3.3.3

Next: , Previous: FFTW Reference, Up: Top

5 Multi-threaded FFTW

In this chapter we document the parallel FFTW routines for shared-memory parallel hardware. These routines, which support parallel one- and multi-dimensional transforms of both real and complex data, are the easiest way to take advantage of multiple processors with FFTW. They work just like the corresponding uniprocessor transform routines, except that you have an extra initialization routine to call, and there is a routine to set the number of threads to employ. Any program that uses the uniprocessor FFTW can therefore be trivially modified to use the multi-threaded FFTW.

A shared-memory machine is one in which all CPUs can directly access the same main memory, and such machines are now common due to the ubiquity of multi-core CPUs. FFTW's multi-threading support allows you to utilize these additional CPUs transparently from a single program. However, this does not necessarily translate into performance gains—when multiple threads/CPUs are employed, there is an overhead required for synchronization that may outweigh the computatational parallelism. Therefore, you can only benefit from threads if your problem is sufficiently large.

fftw-3.3.3/doc/html/Installation-and-Supported-Hardware_002fSoftware.html0000644000175000017500000001126312054411111023160 00000000000000 Installation and Supported Hardware/Software - FFTW 3.3.3

Next: , Previous: Multi-threaded FFTW, Up: Multi-threaded FFTW

5.1 Installation and Supported Hardware/Software

All of the FFTW threads code is located in the threads subdirectory of the FFTW package. On Unix systems, the FFTW threads libraries and header files can be automatically configured, compiled, and installed along with the uniprocessor FFTW libraries simply by including --enable-threads in the flags to the configure script (see Installation on Unix), or --enable-openmp to use OpenMP threads.

The threads routines require your operating system to have some sort of shared-memory threads support. Specifically, the FFTW threads package works with POSIX threads (available on most Unix variants, from GNU/Linux to MacOS X) and Win32 threads. OpenMP threads, which are supported in many common compilers (e.g. gcc) are also supported, and may give better performance on some systems. (OpenMP threads are also useful if you are employing OpenMP in your own code, in order to minimize conflicts between threading models.) If you have a shared-memory machine that uses a different threads API, it should be a simple matter of programming to include support for it; see the file threads/threads.c for more detail.

You can compile FFTW with both --enable-threads and --enable-openmp at the same time, since they install libraries with different names (‘fftw3_threads’ and ‘fftw3_omp’, as described below). However, your programs may only link to one of these two libraries at a time.

Ideally, of course, you should also have multiple processors in order to get any benefit from the threaded transforms. fftw-3.3.3/doc/html/Usage-of-Multi_002dthreaded-FFTW.html0000644000175000017500000001522312054411111017527 00000000000000 Usage of Multi-threaded FFTW - FFTW 3.3.3

Next: , Previous: Installation and Supported Hardware/Software, Up: Multi-threaded FFTW

5.2 Usage of Multi-threaded FFTW

Here, it is assumed that the reader is already familiar with the usage of the uniprocessor FFTW routines, described elsewhere in this manual. We only describe what one has to change in order to use the multi-threaded routines.

First, programs using the parallel complex transforms should be linked with -lfftw3_threads -lfftw3 -lm on Unix, or -lfftw3_omp -lfftw3 -lm if you compiled with OpenMP. You will also need to link with whatever library is responsible for threads on your system (e.g. -lpthread on GNU/Linux) or include whatever compiler flag enables OpenMP (e.g. -fopenmp with gcc).

Second, before calling any FFTW routines, you should call the function: 

     int fftw_init_threads(void);

This function, which need only be called once, performs any one-time initialization required to use threads on your system. It returns zero if there was some error (which should not happen under normal circumstances) and a non-zero value otherwise.

Third, before creating a plan that you want to parallelize, you should call: 

     void fftw_plan_with_nthreads(int nthreads);

The nthreads argument indicates the number of threads you want FFTW to use (or actually, the maximum number). All plans subsequently created with any planner routine will use that many threads. You can call fftw_plan_with_nthreads, create some plans, call fftw_plan_with_nthreads again with a different argument, and create some more plans for a new number of threads. Plans already created before a call to fftw_plan_with_nthreads are unaffected. If you pass an nthreads argument of 1 (the default), threads are disabled for subsequent plans.

With OpenMP, to configure FFTW to use all of the currently running OpenMP threads (set by omp_set_num_threads(nthreads) or by the OMP_NUM_THREADS environment variable), you can do: fftw_plan_with_nthreads(omp_get_max_threads()). (The ‘omp_’ OpenMP functions are declared via #include <omp.h>.)

Given a plan, you then execute it as usual with fftw_execute(plan), and the execution will use the number of threads specified when the plan was created. When done, you destroy it as usual with fftw_destroy_plan. As described in Thread safety, plan execution is thread-safe, but plan creation and destruction are not: you should create/destroy plans only from a single thread, but can safely execute multiple plans in parallel.

There is one additional routine: if you want to get rid of all memory and other resources allocated internally by FFTW, you can call: 

     void fftw_cleanup_threads(void);

which is much like the fftw_cleanup() function except that it also gets rid of threads-related data. You must not execute any previously created plans after calling this function.

We should also mention one other restriction: if you save wisdom from a program using the multi-threaded FFTW, that wisdom cannot be used by a program using only the single-threaded FFTW (i.e. not calling fftw_init_threads). See Words of Wisdom-Saving Plans. fftw-3.3.3/doc/html/How-Many-Threads-to-Use_003f.html0000644000175000017500000000707712054411111016740 00000000000000 How Many Threads to Use? - FFTW 3.3.3

Next: , Previous: Usage of Multi-threaded FFTW, Up: Multi-threaded FFTW

5.3 How Many Threads to Use?

There is a fair amount of overhead involved in synchronizing threads, so the optimal number of threads to use depends upon the size of the transform as well as on the number of processors you have.

As a general rule, you don't want to use more threads than you have processors. (Using more threads will work, but there will be extra overhead with no benefit.) In fact, if the problem size is too small, you may want to use fewer threads than you have processors.

You will have to experiment with your system to see what level of parallelization is best for your problem size. Typically, the problem will have to involve at least a few thousand data points before threads become beneficial. If you plan with FFTW_PATIENT, it will automatically disable threads for sizes that don't benefit from parallelization. fftw-3.3.3/doc/html/Thread-safety.html0000644000175000017500000001132312054411111014474 00000000000000 Thread safety - FFTW 3.3.3

Previous: How Many Threads to Use?, Up: Multi-threaded FFTW

5.4 Thread safety

Users writing multi-threaded programs (including OpenMP) must concern themselves with the thread safety of the libraries they use—that is, whether it is safe to call routines in parallel from multiple threads. FFTW can be used in such an environment, but some care must be taken because the planner routines share data (e.g. wisdom and trigonometric tables) between calls and plans.

The upshot is that the only thread-safe (re-entrant) routine in FFTW is fftw_execute (and the new-array variants thereof). All other routines (e.g. the planner) should only be called from one thread at a time. So, for example, you can wrap a semaphore lock around any calls to the planner; even more simply, you can just create all of your plans from one thread. We do not think this should be an important restriction (FFTW is designed for the situation where the only performance-sensitive code is the actual execution of the transform), and the benefits of shared data between plans are great.

Note also that, since the plan is not modified by fftw_execute, it is safe to execute the same plan in parallel by multiple threads. However, since a given plan operates by default on a fixed array, you need to use one of the new-array execute functions (see New-array Execute Functions) so that different threads compute the transform of different data.

(Users should note that these comments only apply to programs using shared-memory threads or OpenMP. Parallelism using MPI or forked processes involves a separate address-space and global variables for each process, and is not susceptible to problems of this sort.)

If you are configured FFTW with the --enable-debug or --enable-debug-malloc flags (see Installation on Unix), then fftw_execute is not thread-safe. These flags are not documented because they are intended only for developing and debugging FFTW, but if you must use --enable-debug then you should also specifically pass --disable-debug-malloc for fftw_execute to be thread-safe. fftw-3.3.3/doc/html/Distributed_002dmemory-FFTW-with-MPI.html0000644000175000017500000001436712054411111020407 00000000000000 Distributed-memory FFTW with MPI - FFTW 3.3.3

Next: , Previous: Multi-threaded FFTW, Up: Top

6 Distributed-memory FFTW with MPI

In this chapter we document the parallel FFTW routines for parallel systems supporting the MPI message-passing interface. Unlike the shared-memory threads described in the previous chapter, MPI allows you to use distributed-memory parallelism, where each CPU has its own separate memory, and which can scale up to clusters of many thousands of processors. This capability comes at a price, however: each process only stores a portion of the data to be transformed, which means that the data structures and programming-interface are quite different from the serial or threads versions of FFTW.

Distributed-memory parallelism is especially useful when you are transforming arrays so large that they do not fit into the memory of a single processor. The storage per-process required by FFTW's MPI routines is proportional to the total array size divided by the number of processes. Conversely, distributed-memory parallelism can easily pose an unacceptably high communications overhead for small problems; the threshold problem size for which parallelism becomes advantageous will depend on the precise problem you are interested in, your hardware, and your MPI implementation.

A note on terminology: in MPI, you divide the data among a set of “processes” which each run in their own memory address space. Generally, each process runs on a different physical processor, but this is not required. A set of processes in MPI is described by an opaque data structure called a “communicator,” the most common of which is the predefined communicator MPI_COMM_WORLD which refers to all processes. For more information on these and other concepts common to all MPI programs, we refer the reader to the documentation at the MPI home page.

We assume in this chapter that the reader is familiar with the usage of the serial (uniprocessor) FFTW, and focus only on the concepts new to the MPI interface.

fftw-3.3.3/doc/html/FFTW-MPI-Installation.html0000644000175000017500000001135712054411111015633 00000000000000 FFTW MPI Installation - FFTW 3.3.3

Next: , Previous: Distributed-memory FFTW with MPI, Up: Distributed-memory FFTW with MPI

6.1 FFTW MPI Installation

All of the FFTW MPI code is located in the mpi subdirectory of the FFTW package. On Unix systems, the FFTW MPI libraries and header files are automatically configured, compiled, and installed along with the uniprocessor FFTW libraries simply by including --enable-mpi in the flags to the configure script (see Installation on Unix).

Any implementation of the MPI standard, version 1 or later, should work with FFTW. The configure script will attempt to automatically detect how to compile and link code using your MPI implementation. In some cases, especially if you have multiple different MPI implementations installed or have an unusual MPI software package, you may need to provide this information explicitly.

Most commonly, one compiles MPI code by invoking a special compiler command, typically mpicc for C code. The configure script knows the most common names for this command, but you can specify the MPI compilation command explicitly by setting the MPICC variable, as in ‘./configure MPICC=mpicc ...’.

If, instead of a special compiler command, you need to link a certain library, you can specify the link command via the MPILIBS variable, as in ‘./configure MPILIBS=-lmpi ...’. Note that if your MPI library is installed in a non-standard location (one the compiler does not know about by default), you may also have to specify the location of the library and header files via LDFLAGS and CPPFLAGS variables, respectively, as in ‘./configure LDFLAGS=-L/path/to/mpi/libs CPPFLAGS=-I/path/to/mpi/include ...’. fftw-3.3.3/doc/html/Linking-and-Initializing-MPI-FFTW.html0000644000175000017500000001016712054411111017753 00000000000000 Linking and Initializing MPI FFTW - FFTW 3.3.3

Next: , Previous: FFTW MPI Installation, Up: Distributed-memory FFTW with MPI

6.2 Linking and Initializing MPI FFTW

Programs using the MPI FFTW routines should be linked with -lfftw3_mpi -lfftw3 -lm on Unix in double precision, -lfftw3f_mpi -lfftw3f -lm in single precision, and so on (see Precision). You will also need to link with whatever library is responsible for MPI on your system; in most MPI implementations, there is a special compiler alias named mpicc to compile and link MPI code.

Before calling any FFTW routines except possibly fftw_init_threads (see Combining MPI and Threads), but after calling MPI_Init, you should call the function: 

     void fftw_mpi_init(void);

If, at the end of your program, you want to get rid of all memory and other resources allocated internally by FFTW, for both the serial and MPI routines, you can call: 

     void fftw_mpi_cleanup(void);

which is much like the fftw_cleanup() function except that it also gets rid of FFTW's MPI-related data. You must not execute any previously created plans after calling this function. fftw-3.3.3/doc/html/2d-MPI-example.html0000644000175000017500000001761712054411111014371 00000000000000 2d MPI example - FFTW 3.3.3

Next: , Previous: Linking and Initializing MPI FFTW, Up: Distributed-memory FFTW with MPI

6.3 2d MPI example

Before we document the FFTW MPI interface in detail, we begin with a simple example outlining how one would perform a two-dimensional N0 by N1 complex DFT. 

     #include <fftw3-mpi.h>
     
     int main(int argc, char **argv)
     {
         const ptrdiff_t N0 = ..., N1 = ...;
         fftw_plan plan;
         fftw_complex *data;
         ptrdiff_t alloc_local, local_n0, local_0_start, i, j;
     
         MPI_Init(&argc, &argv);
         fftw_mpi_init();
     
         /* get local data size and allocate */
         alloc_local = fftw_mpi_local_size_2d(N0, N1, MPI_COMM_WORLD,
                                              &local_n0, &local_0_start);
         data = fftw_alloc_complex(alloc_local);
     
         /* create plan for in-place forward DFT */
         plan = fftw_mpi_plan_dft_2d(N0, N1, data, data, MPI_COMM_WORLD,
                                     FFTW_FORWARD, FFTW_ESTIMATE);
     
         /* initialize data to some function my_function(x,y) */
         for (i = 0; i < local_n0; ++i) for (j = 0; j < N1; ++j)
            data[i*N1 + j] = my_function(local_0_start + i, j);
     
         /* compute transforms, in-place, as many times as desired */
         fftw_execute(plan);
     
         fftw_destroy_plan(plan);
     
         MPI_Finalize();
     }

As can be seen above, the MPI interface follows the same basic style of allocate/plan/execute/destroy as the serial FFTW routines. All of the MPI-specific routines are prefixed with ‘fftw_mpi_’ instead of ‘fftw_’. There are a few important differences, however:

First, we must call fftw_mpi_init() after calling MPI_Init (required in all MPI programs) and before calling any other ‘fftw_mpi_’ routine.

Second, when we create the plan with fftw_mpi_plan_dft_2d, analogous to fftw_plan_dft_2d, we pass an additional argument: the communicator, indicating which processes will participate in the transform (here MPI_COMM_WORLD, indicating all processes). Whenever you create, execute, or destroy a plan for an MPI transform, you must call the corresponding FFTW routine on all processes in the communicator for that transform. (That is, these are collective calls.) Note that the plan for the MPI transform uses the standard fftw_execute and fftw_destroy routines (on the other hand, there are MPI-specific new-array execute functions documented below).

Third, all of the FFTW MPI routines take ptrdiff_t arguments instead of int as for the serial FFTW. ptrdiff_t is a standard C integer type which is (at least) 32 bits wide on a 32-bit machine and 64 bits wide on a 64-bit machine. This is to make it easy to specify very large parallel transforms on a 64-bit machine. (You can specify 64-bit transform sizes in the serial FFTW, too, but only by using the ‘guru64’ planner interface. See 64-bit Guru Interface.)

Fourth, and most importantly, you don't allocate the entire two-dimensional array on each process. Instead, you call fftw_mpi_local_size_2d to find out what portion of the array resides on each processor, and how much space to allocate. Here, the portion of the array on each process is a local_n0 by N1 slice of the total array, starting at index local_0_start. The total number of fftw_complex numbers to allocate is given by the alloc_local return value, which may be greater than local_n0 * N1 (in case some intermediate calculations require additional storage). The data distribution in FFTW's MPI interface is described in more detail by the next section.

Given the portion of the array that resides on the local process, it is straightforward to initialize the data (here to a function myfunction) and otherwise manipulate it. Of course, at the end of the program you may want to output the data somehow, but synchronizing this output is up to you and is beyond the scope of this manual. (One good way to output a large multi-dimensional distributed array in MPI to a portable binary file is to use the free HDF5 library; see the HDF home page.) fftw-3.3.3/doc/html/MPI-Data-Distribution.html0000644000175000017500000001414612054411111015753 00000000000000 MPI Data Distribution - FFTW 3.3.3

Next: , Previous: 2d MPI example, Up: Distributed-memory FFTW with MPI

6.4 MPI Data Distribution

The most important concept to understand in using FFTW's MPI interface is the data distribution. With a serial or multithreaded FFT, all of the inputs and outputs are stored as a single contiguous chunk of memory. With a distributed-memory FFT, the inputs and outputs are broken into disjoint blocks, one per process.

In particular, FFTW uses a 1d block distribution of the data, distributed along the first dimension. For example, if you want to perform a 100 × 200 complex DFT, distributed over 4 processes, each process will get a 25 × 200 slice of the data. That is, process 0 will get rows 0 through 24, process 1 will get rows 25 through 49, process 2 will get rows 50 through 74, and process 3 will get rows 75 through 99. If you take the same array but distribute it over 3 processes, then it is not evenly divisible so the different processes will have unequal chunks. FFTW's default choice in this case is to assign 34 rows to processes 0 and 1, and 32 rows to process 2.

FFTW provides several ‘fftw_mpi_local_size’ routines that you can call to find out what portion of an array is stored on the current process. In most cases, you should use the default block sizes picked by FFTW, but it is also possible to specify your own block size. For example, with a 100 × 200 array on three processes, you can tell FFTW to use a block size of 40, which would assign 40 rows to processes 0 and 1, and 20 rows to process 2. FFTW's default is to divide the data equally among the processes if possible, and as best it can otherwise. The rows are always assigned in “rank order,” i.e. process 0 gets the first block of rows, then process 1, and so on. (You can change this by using MPI_Comm_split to create a new communicator with re-ordered processes.) However, you should always call the ‘fftw_mpi_local_size’ routines, if possible, rather than trying to predict FFTW's distribution choices.

In particular, it is critical that you allocate the storage size that is returned by ‘fftw_mpi_local_size’, which is not necessarily the size of the local slice of the array. The reason is that intermediate steps of FFTW's algorithms involve transposing the array and redistributing the data, so at these intermediate steps FFTW may require more local storage space (albeit always proportional to the total size divided by the number of processes). The ‘fftw_mpi_local_size’ functions know how much storage is required for these intermediate steps and tell you the correct amount to allocate.

fftw-3.3.3/doc/html/Basic-and-advanced-distribution-interfaces.html0000644000175000017500000001774612054411111022175 00000000000000 Basic and advanced distribution interfaces - FFTW 3.3.3

Next: , Previous: MPI Data Distribution, Up: MPI Data Distribution

6.4.1 Basic and advanced distribution interfaces

As with the planner interface, the ‘fftw_mpi_local_size’ distribution interface is broken into basic and advanced (‘_many’) interfaces, where the latter allows you to specify the block size manually and also to request block sizes when computing multiple transforms simultaneously. These functions are documented more exhaustively by the FFTW MPI Reference, but we summarize the basic ideas here using a couple of two-dimensional examples.

For the 100 × 200 complex-DFT example, above, we would find the distribution by calling the following function in the basic interface: 

     ptrdiff_t fftw_mpi_local_size_2d(ptrdiff_t n0, ptrdiff_t n1, MPI_Comm comm,
                                      ptrdiff_t *local_n0, ptrdiff_t *local_0_start);

Given the total size of the data to be transformed (here, n0 = 100 and n1 = 200) and an MPI communicator (comm), this function provides three numbers.

First, it describes the shape of the local data: the current process should store a local_n0 by n1 slice of the overall dataset, in row-major order (n1 dimension contiguous), starting at index local_0_start. That is, if the total dataset is viewed as a n0 by n1 matrix, the current process should store the rows local_0_start to local_0_start+local_n0-1. Obviously, if you are running with only a single MPI process, that process will store the entire array: local_0_start will be zero and local_n0 will be n0. See Row-major Format.

Second, the return value is the total number of data elements (e.g., complex numbers for a complex DFT) that should be allocated for the input and output arrays on the current process (ideally with fftw_malloc or an ‘fftw_alloc’ function, to ensure optimal alignment). It might seem that this should always be equal to local_n0 * n1, but this is not the case. FFTW's distributed FFT algorithms require data redistributions at intermediate stages of the transform, and in some circumstances this may require slightly larger local storage. This is discussed in more detail below, under Load balancing.

The advanced-interface ‘local_size’ function for multidimensional transforms returns the same three things (local_n0, local_0_start, and the total number of elements to allocate), but takes more inputs: 

     ptrdiff_t fftw_mpi_local_size_many(int rnk, const ptrdiff_t *n,
                                        ptrdiff_t howmany,
                                        ptrdiff_t block0,
                                        MPI_Comm comm,
                                        ptrdiff_t *local_n0,
                                        ptrdiff_t *local_0_start);

The two-dimensional case above corresponds to rnk = 2 and an array n of length 2 with n[0] = n0 and n[1] = n1. This routine is for any rnk > 1; one-dimensional transforms have their own interface because they work slightly differently, as discussed below.

First, the advanced interface allows you to perform multiple transforms at once, of interleaved data, as specified by the howmany parameter. (hoamany is 1 for a single transform.)

Second, here you can specify your desired block size in the n0 dimension, block0. To use FFTW's default block size, pass FFTW_MPI_DEFAULT_BLOCK (0) for block0. Otherwise, on P processes, FFTW will return local_n0 equal to block0 on the first P / block0 processes (rounded down), return local_n0 equal to n0 - block0 * (P / block0) on the next process, and local_n0 equal to zero on any remaining processes. In general, we recommend using the default block size (which corresponds to n0 / P, rounded up).

For example, suppose you have P = 4 processes and n0 = 21. The default will be a block size of 6, which will give local_n0 = 6 on the first three processes and local_n0 = 3 on the last process. Instead, however, you could specify block0 = 5 if you wanted, which would give local_n0 = 5 on processes 0 to 2, local_n0 = 6 on process 3. (This choice, while it may look superficially more “balanced,” has the same critical path as FFTW's default but requires more communications.) fftw-3.3.3/doc/html/Load-balancing.html0000644000175000017500000001046112054411111014571 00000000000000 Load balancing - FFTW 3.3.3

Next: , Previous: Basic and advanced distribution interfaces, Up: MPI Data Distribution

6.4.2 Load balancing

Ideally, when you parallelize a transform over some P processes, each process should end up with work that takes equal time. Otherwise, all of the processes end up waiting on whichever process is slowest. This goal is known as “load balancing.” In this section, we describe the circumstances under which FFTW is able to load-balance well, and in particular how you should choose your transform size in order to load balance.

Load balancing is especially difficult when you are parallelizing over heterogeneous machines; for example, if one of your processors is a old 486 and another is a Pentium IV, obviously you should give the Pentium more work to do than the 486 since the latter is much slower. FFTW does not deal with this problem, however—it assumes that your processes run on hardware of comparable speed, and that the goal is therefore to divide the problem as equally as possible.

For a multi-dimensional complex DFT, FFTW can divide the problem equally among the processes if: (i) the first dimension n0 is divisible by P; and (ii), the product of the subsequent dimensions is divisible by P. (For the advanced interface, where you can specify multiple simultaneous transforms via some “vector” length howmany, a factor of howmany is included in the product of the subsequent dimensions.)

For a one-dimensional complex DFT, the length N of the data should be divisible by P squared to be able to divide the problem equally among the processes. fftw-3.3.3/doc/html/Transposed-distributions.html0000644000175000017500000001654012054411111017024 00000000000000 Transposed distributions - FFTW 3.3.3

Next: , Previous: Load balancing, Up: MPI Data Distribution

6.4.3 Transposed distributions

Internally, FFTW's MPI transform algorithms work by first computing transforms of the data local to each process, then by globally transposing the data in some fashion to redistribute the data among the processes, transforming the new data local to each process, and transposing back. For example, a two-dimensional n0 by n1 array, distributed across the n0 dimension, is transformd by: (i) transforming the n1 dimension, which are local to each process; (ii) transposing to an n1 by n0 array, distributed across the n1 dimension; (iii) transforming the n0 dimension, which is now local to each process; (iv) transposing back.

However, in many applications it is acceptable to compute a multidimensional DFT whose results are produced in transposed order (e.g., n1 by n0 in two dimensions). This provides a significant performance advantage, because it means that the final transposition step can be omitted. FFTW supports this optimization, which you specify by passing the flag FFTW_MPI_TRANSPOSED_OUT to the planner routines. To compute the inverse transform of transposed output, you specify FFTW_MPI_TRANSPOSED_IN to tell it that the input is transposed. In this section, we explain how to interpret the output format of such a transform.

Suppose you have are transforming multi-dimensional data with (at least two) dimensions n0 × n1 × n2 × … × nd-1. As always, it is distributed along the first dimension n0. Now, if we compute its DFT with the FFTW_MPI_TRANSPOSED_OUT flag, the resulting output data are stored with the first two dimensions transposed: n1 × n0 × n2 ×…× nd-1, distributed along the n1 dimension. Conversely, if we take the n1 × n0 × n2 ×…× nd-1 data and transform it with the FFTW_MPI_TRANSPOSED_IN flag, then the format goes back to the original n0 × n1 × n2 × … × nd-1 array.

There are two ways to find the portion of the transposed array that resides on the current process. First, you can simply call the appropriate ‘local_size’ function, passing n1 × n0 × n2 ×…× nd-1 (the transposed dimensions). This would mean calling the ‘local_size’ function twice, once for the transposed and once for the non-transposed dimensions. Alternatively, you can call one of the ‘local_size_transposed’ functions, which returns both the non-transposed and transposed data distribution from a single call. For example, for a 3d transform with transposed output (or input), you might call: 

     ptrdiff_t fftw_mpi_local_size_3d_transposed(
                     ptrdiff_t n0, ptrdiff_t n1, ptrdiff_t n2, MPI_Comm comm,
                     ptrdiff_t *local_n0, ptrdiff_t *local_0_start,
                     ptrdiff_t *local_n1, ptrdiff_t *local_1_start);

Here, local_n0 and local_0_start give the size and starting index of the n0 dimension for the non-transposed data, as in the previous sections. For transposed data (e.g. the output for FFTW_MPI_TRANSPOSED_OUT), local_n1 and local_1_start give the size and starting index of the n1 dimension, which is the first dimension of the transposed data (n1 by n0 by n2).

(Note that FFTW_MPI_TRANSPOSED_IN is completely equivalent to performing FFTW_MPI_TRANSPOSED_OUT and passing the first two dimensions to the planner in reverse order, or vice versa. If you pass both the FFTW_MPI_TRANSPOSED_IN and FFTW_MPI_TRANSPOSED_OUT flags, it is equivalent to swapping the first two dimensions passed to the planner and passing neither flag.) fftw-3.3.3/doc/html/One_002ddimensional-distributions.html0000644000175000017500000001240212054411111020364 00000000000000 One-dimensional distributions - FFTW 3.3.3

Previous: Transposed distributions, Up: MPI Data Distribution

6.4.4 One-dimensional distributions

For one-dimensional distributed DFTs using FFTW, matters are slightly more complicated because the data distribution is more closely tied to how the algorithm works. In particular, you can no longer pass an arbitrary block size and must accept FFTW's default; also, the block sizes may be different for input and output. Also, the data distribution depends on the flags and transform direction, in order for forward and backward transforms to work correctly. 

     ptrdiff_t fftw_mpi_local_size_1d(ptrdiff_t n0, MPI_Comm comm,
                     int sign, unsigned flags,
                     ptrdiff_t *local_ni, ptrdiff_t *local_i_start,
                     ptrdiff_t *local_no, ptrdiff_t *local_o_start);

This function computes the data distribution for a 1d transform of size n0 with the given transform sign and flags. Both input and output data use block distributions. The input on the current process will consist of local_ni numbers starting at index local_i_start; e.g. if only a single process is used, then local_ni will be n0 and local_i_start will be 0. Similarly for the output, with local_no numbers starting at index local_o_start. The return value of fftw_mpi_local_size_1d will be the total number of elements to allocate on the current process (which might be slightly larger than the local size due to intermediate steps in the algorithm).

As mentioned above (see Load balancing), the data will be divided equally among the processes if n0 is divisible by the square of the number of processes. In this case, local_ni will equal local_no. Otherwise, they may be different.

For some applications, such as convolutions, the order of the output data is irrelevant. In this case, performance can be improved by specifying that the output data be stored in an FFTW-defined “scrambled” format. (In particular, this is the analogue of transposed output in the multidimensional case: scrambled output saves a communications step.) If you pass FFTW_MPI_SCRAMBLED_OUT in the flags, then the output is stored in this (undocumented) scrambled order. Conversely, to perform the inverse transform of data in scrambled order, pass the FFTW_MPI_SCRAMBLED_IN flag.

In MPI FFTW, only composite sizes n0 can be parallelized; we have not yet implemented a parallel algorithm for large prime sizes. fftw-3.3.3/doc/html/Multi_002ddimensional-MPI-DFTs-of-Real-Data.html0000644000175000017500000002215312054411111021414 00000000000000 Multi-dimensional MPI DFTs of Real Data - FFTW 3.3.3

Next: , Previous: MPI Data Distribution, Up: Distributed-memory FFTW with MPI

6.5 Multi-dimensional MPI DFTs of Real Data

FFTW's MPI interface also supports multi-dimensional DFTs of real data, similar to the serial r2c and c2r interfaces. (Parallel one-dimensional real-data DFTs are not currently supported; you must use a complex transform and set the imaginary parts of the inputs to zero.)

The key points to understand for r2c and c2r MPI transforms (compared to the MPI complex DFTs or the serial r2c/c2r transforms), are:

  • Just as for serial transforms, r2c/c2r DFTs transform n0 × n1 × n2 × … × nd-1 real data to/from n0 × n1 × n2 × … × (nd-1/2 + 1) complex data: the last dimension of the complex data is cut in half (rounded down), plus one. As for the serial transforms, the sizes you pass to the ‘plan_dft_r2c’ and ‘plan_dft_c2r’ are the n0 × n1 × n2 × … × nd-1 dimensions of the real data.
  • Although the real data is conceptually n0 × n1 × n2 × … × nd-1, it is physically stored as an n0 × n1 × n2 × … × [2 (nd-1/2 + 1)] array, where the last dimension has been padded to make it the same size as the complex output. This is much like the in-place serial r2c/c2r interface (see Multi-Dimensional DFTs of Real Data), except that in MPI the padding is required even for out-of-place data. The extra padding numbers are ignored by FFTW (they are not like zero-padding the transform to a larger size); they are only used to determine the data layout.
  • The data distribution in MPI for both the real and complex data is determined by the shape of the complex data. That is, you call the appropriate ‘local size’ function for the n0 × n1 × n2 × … × (nd-1/2 + 1)

    complex data, and then use the same distribution for the real data except that the last complex dimension is replaced by a (padded) real dimension of twice the length.

For example suppose we are performing an out-of-place r2c transform of L × M × N real data [padded to L × M × 2(N/2+1)], resulting in L × M × N/2+1 complex data. Similar to the example in 2d MPI example, we might do something like: 

     #include <fftw3-mpi.h>
     
     int main(int argc, char **argv)
     {
         const ptrdiff_t L = ..., M = ..., N = ...;
         fftw_plan plan;
         double *rin;
         fftw_complex *cout;
         ptrdiff_t alloc_local, local_n0, local_0_start, i, j, k;
     
         MPI_Init(&argc, &argv);
         fftw_mpi_init();
     
         /* get local data size and allocate */
         alloc_local = fftw_mpi_local_size_3d(L, M, N/2+1, MPI_COMM_WORLD,
                                              &local_n0, &local_0_start);
         rin = fftw_alloc_real(2 * alloc_local);
         cout = fftw_alloc_complex(alloc_local);
     
         /* create plan for out-of-place r2c DFT */
         plan = fftw_mpi_plan_dft_r2c_3d(L, M, N, rin, cout, MPI_COMM_WORLD,
                                         FFTW_MEASURE);
     
         /* initialize rin to some function my_func(x,y,z) */
         for (i = 0; i < local_n0; ++i)
            for (j = 0; j < M; ++j)
              for (k = 0; k < N; ++k)
            rin[(i*M + j) * (2*(N/2+1)) + k] = my_func(local_0_start+i, j, k);
     
         /* compute transforms as many times as desired */
         fftw_execute(plan);
     
         fftw_destroy_plan(plan);
     
         MPI_Finalize();
     }

Note that we allocated rin using fftw_alloc_real with an argument of 2 * alloc_local: since alloc_local is the number of complex values to allocate, the number of real values is twice as many. The rin array is then local_n0 × M × 2(N/2+1) in row-major order, so its (i,j,k) element is at the index (i*M + j) * (2*(N/2+1)) + k (see Multi-dimensional Array Format).

As for the complex transforms, improved performance can be obtained by specifying that the output is the transpose of the input or vice versa (see Transposed distributions). In our L × M × N r2c example, including FFTW_TRANSPOSED_OUT in the flags means that the input would be a padded L × M × 2(N/2+1) real array distributed over the L dimension, while the output would be a M × L × N/2+1 complex array distributed over the M dimension. To perform the inverse c2r transform with the same data distributions, you would use the FFTW_TRANSPOSED_IN flag. fftw-3.3.3/doc/html/Other-Multi_002ddimensional-Real_002ddata-MPI-Transforms.html0000644000175000017500000001246412054411111024143 00000000000000 Other Multi-dimensional Real-data MPI Transforms - FFTW 3.3.3

Next: , Previous: Multi-dimensional MPI DFTs of Real Data, Up: Distributed-memory FFTW with MPI

6.6 Other multi-dimensional Real-Data MPI Transforms

FFTW's MPI interface also supports multi-dimensional ‘r2r’ transforms of all kinds supported by the serial interface (e.g. discrete cosine and sine transforms, discrete Hartley transforms, etc.). Only multi-dimensional ‘r2r’ transforms, not one-dimensional transforms, are currently parallelized.

These are used much like the multidimensional complex DFTs discussed above, except that the data is real rather than complex, and one needs to pass an r2r transform kind (fftw_r2r_kind) for each dimension as in the serial FFTW (see More DFTs of Real Data).

For example, one might perform a two-dimensional L × M that is an REDFT10 (DCT-II) in the first dimension and an RODFT10 (DST-II) in the second dimension with code like: 

         const ptrdiff_t L = ..., M = ...;
         fftw_plan plan;
         double *data;
         ptrdiff_t alloc_local, local_n0, local_0_start, i, j;
     
         /* get local data size and allocate */
         alloc_local = fftw_mpi_local_size_2d(L, M, MPI_COMM_WORLD,
                                              &local_n0, &local_0_start);
         data = fftw_alloc_real(alloc_local);
     
         /* create plan for in-place REDFT10 x RODFT10 */
         plan = fftw_mpi_plan_r2r_2d(L, M, data, data, MPI_COMM_WORLD,
                                     FFTW_REDFT10, FFTW_RODFT10, FFTW_MEASURE);
     
         /* initialize data to some function my_function(x,y) */
         for (i = 0; i < local_n0; ++i) for (j = 0; j < M; ++j)
            data[i*M + j] = my_function(local_0_start + i, j);
     
         /* compute transforms, in-place, as many times as desired */
         fftw_execute(plan);
     
         fftw_destroy_plan(plan);

Notice that we use the same ‘local_size’ functions as we did for complex data, only now we interpret the sizes in terms of real rather than complex values, and correspondingly use fftw_alloc_real. fftw-3.3.3/doc/html/FFTW-MPI-Transposes.html0000644000175000017500000000751612054411111015335 00000000000000 FFTW MPI Transposes - FFTW 3.3.3

Next: , Previous: Other Multi-dimensional Real-data MPI Transforms, Up: Distributed-memory FFTW with MPI

6.7 FFTW MPI Transposes

The FFTW's MPI Fourier transforms rely on one or more global transposition step for their communications. For example, the multidimensional transforms work by transforming along some dimensions, then transposing to make the first dimension local and transforming that, then transposing back. Because global transposition of a block-distributed matrix has many other potential uses besides FFTs, FFTW's transpose routines can be called directly, as documented in this section.

fftw-3.3.3/doc/html/Basic-distributed_002dtranspose-interface.html0000644000175000017500000001252612054411111021765 00000000000000 Basic distributed-transpose interface - FFTW 3.3.3

Next: , Previous: FFTW MPI Transposes, Up: FFTW MPI Transposes

6.7.1 Basic distributed-transpose interface

In particular, suppose that we have an n0 by n1 array in row-major order, block-distributed across the n0 dimension. To transpose this into an n1 by n0 array block-distributed across the n1 dimension, we would create a plan by calling the following function: 

     fftw_plan fftw_mpi_plan_transpose(ptrdiff_t n0, ptrdiff_t n1,
                                       double *in, double *out,
                                       MPI_Comm comm, unsigned flags);

The input and output arrays (in and out) can be the same. The transpose is actually executed by calling fftw_execute on the plan, as usual.

The flags are the usual FFTW planner flags, but support two additional flags: FFTW_MPI_TRANSPOSED_OUT and/or FFTW_MPI_TRANSPOSED_IN. What these flags indicate, for transpose plans, is that the output and/or input, respectively, are locally transposed. That is, on each process input data is normally stored as a local_n0 by n1 array in row-major order, but for an FFTW_MPI_TRANSPOSED_IN plan the input data is stored as n1 by local_n0 in row-major order. Similarly, FFTW_MPI_TRANSPOSED_OUT means that the output is n0 by local_n1 instead of local_n1 by n0.

To determine the local size of the array on each process before and after the transpose, as well as the amount of storage that must be allocated, one should call fftw_mpi_local_size_2d_transposed, just as for a 2d DFT as described in the previous section: 

     ptrdiff_t fftw_mpi_local_size_2d_transposed
                     (ptrdiff_t n0, ptrdiff_t n1, MPI_Comm comm,
                      ptrdiff_t *local_n0, ptrdiff_t *local_0_start,
                      ptrdiff_t *local_n1, ptrdiff_t *local_1_start);

Again, the return value is the local storage to allocate, which in this case is the number of real (double) values rather than complex numbers as in the previous examples. fftw-3.3.3/doc/html/Advanced-distributed_002dtranspose-interface.html0000644000175000017500000001043612054411111022447 00000000000000 Advanced distributed-transpose interface - FFTW 3.3.3

Next: , Previous: Basic distributed-transpose interface, Up: FFTW MPI Transposes

6.7.2 Advanced distributed-transpose interface

The above routines are for a transpose of a matrix of numbers (of type double), using FFTW's default block sizes. More generally, one can perform transposes of tuples of numbers, with user-specified block sizes for the input and output: 

     fftw_plan fftw_mpi_plan_many_transpose
                     (ptrdiff_t n0, ptrdiff_t n1, ptrdiff_t howmany,
                      ptrdiff_t block0, ptrdiff_t block1,
                      double *in, double *out, MPI_Comm comm, unsigned flags);

In this case, one is transposing an n0 by n1 matrix of howmany-tuples (e.g. howmany = 2 for complex numbers). The input is distributed along the n0 dimension with block size block0, and the n1 by n0 output is distributed along the n1 dimension with block size block1. If FFTW_MPI_DEFAULT_BLOCK (0) is passed for a block size then FFTW uses its default block size. To get the local size of the data on each process, you should then call fftw_mpi_local_size_many_transposed. fftw-3.3.3/doc/html/An-improved-replacement-for-MPI_005fAlltoall.html0000644000175000017500000001132312054411111022100 00000000000000 An improved replacement for MPI_Alltoall - FFTW 3.3.3

Previous: Advanced distributed-transpose interface, Up: FFTW MPI Transposes

6.7.3 An improved replacement for MPI_Alltoall

We close this section by noting that FFTW's MPI transpose routines can be thought of as a generalization for the MPI_Alltoall function (albeit only for floating-point types), and in some circumstances can function as an improved replacement.

MPI_Alltoall is defined by the MPI standard as: 

     int MPI_Alltoall(void *sendbuf, int sendcount, MPI_Datatype sendtype,
                      void *recvbuf, int recvcnt, MPI_Datatype recvtype,
                      MPI_Comm comm);

In particular, for double* arrays in and out, consider the call: 

     MPI_Alltoall(in, howmany, MPI_DOUBLE, out, howmany MPI_DOUBLE, comm);

This is completely equivalent to: 

     MPI_Comm_size(comm, &P);
     plan = fftw_mpi_plan_many_transpose(P, P, howmany, 1, 1, in, out, comm, FFTW_ESTIMATE);
     fftw_execute(plan);
     fftw_destroy_plan(plan);

That is, computing a P × P transpose on P processes, with a block size of 1, is just a standard all-to-all communication.

However, using the FFTW routine instead of MPI_Alltoall may have certain advantages. First of all, FFTW's routine can operate in-place (in == out) whereas MPI_Alltoall can only operate out-of-place.

Second, even for out-of-place plans, FFTW's routine may be faster, especially if you need to perform the all-to-all communication many times and can afford to use FFTW_MEASURE or FFTW_PATIENT. It should certainly be no slower, not including the time to create the plan, since one of the possible algorithms that FFTW uses for an out-of-place transpose is simply to call MPI_Alltoall. However, FFTW also considers several other possible algorithms that, depending on your MPI implementation and your hardware, may be faster. fftw-3.3.3/doc/html/FFTW-MPI-Wisdom.html0000644000175000017500000001563412054411111014436 00000000000000 FFTW MPI Wisdom - FFTW 3.3.3

Next: , Previous: FFTW MPI Transposes, Up: Distributed-memory FFTW with MPI

6.8 FFTW MPI Wisdom

FFTW's “wisdom” facility (see Words of Wisdom-Saving Plans) can be used to save MPI plans as well as to save uniprocessor plans. However, for MPI there are several unavoidable complications.

First, the MPI standard does not guarantee that every process can perform file I/O (at least, not using C stdio routines)—in general, we may only assume that process 0 is capable of I/O.1 So, if we want to export the wisdom from a single process to a file, we must first export the wisdom to a string, then send it to process 0, then write it to a file.

Second, in principle we may want to have separate wisdom for every process, since in general the processes may run on different hardware even for a single MPI program. However, in practice FFTW's MPI code is designed for the case of homogeneous hardware (see Load balancing), and in this case it is convenient to use the same wisdom for every process. Thus, we need a mechanism to synchronize the wisdom.

To address both of these problems, FFTW provides the following two functions: 

     void fftw_mpi_broadcast_wisdom(MPI_Comm comm);
     void fftw_mpi_gather_wisdom(MPI_Comm comm);

Given a communicator comm, fftw_mpi_broadcast_wisdom will broadcast the wisdom from process 0 to all other processes. Conversely, fftw_mpi_gather_wisdom will collect wisdom from all processes onto process 0. (If the plans created for the same problem by different processes are not the same, fftw_mpi_gather_wisdom will arbitrarily choose one of the plans.) Both of these functions may result in suboptimal plans for different processes if the processes are running on non-identical hardware. Both of these functions are collective calls, which means that they must be executed by all processes in the communicator.

So, for example, a typical code snippet to import wisdom from a file and use it on all processes would be: 

     {
         int rank;
     
         fftw_mpi_init();
         MPI_Comm_rank(MPI_COMM_WORLD, &rank);
         if (rank == 0) fftw_import_wisdom_from_filename("mywisdom");
         fftw_mpi_broadcast_wisdom(MPI_COMM_WORLD);
     }

(Note that we must call fftw_mpi_init before importing any wisdom that might contain MPI plans.) Similarly, a typical code snippet to export wisdom from all processes to a file is: 

     {
         int rank;
     
         fftw_mpi_gather_wisdom(MPI_COMM_WORLD);
         MPI_Comm_rank(MPI_COMM_WORLD, &rank);
         if (rank == 0) fftw_export_wisdom_to_filename("mywisdom");
     }

Footnotes

[1] In fact, even this assumption is not technically guaranteed by the standard, although it seems to be universal in actual MPI implementations and is widely assumed by MPI-using software. Technically, you need to query the MPI_IO attribute of MPI_COMM_WORLD with MPI_Attr_get. If this attribute is MPI_PROC_NULL, no I/O is possible. If it is MPI_ANY_SOURCE, any process can perform I/O. Otherwise, it is the rank of a process that can perform I/O ... but since it is not guaranteed to yield the same rank on all processes, you have to do an MPI_Allreduce of some kind if you want all processes to agree about which is going to do I/O. And even then, the standard only guarantees that this process can perform output, but not input. See e.g. Parallel Programming with MPI by P. S. Pacheco, section 8.1.3. Needless to say, in our experience virtually no MPI programmers worry about this.

fftw-3.3.3/doc/html/Avoiding-MPI-Deadlocks.html0000644000175000017500000000765312054411111016061 00000000000000 Avoiding MPI Deadlocks - FFTW 3.3.3

Next: , Previous: FFTW MPI Wisdom, Up: Distributed-memory FFTW with MPI

6.9 Avoiding MPI Deadlocks

An MPI program can deadlock if one process is waiting for a message from another process that never gets sent. To avoid deadlocks when using FFTW's MPI routines, it is important to know which functions are collective: that is, which functions must always be called in the same order from every process in a given communicator. (For example, MPI_Barrier is the canonical example of a collective function in the MPI standard.)

The functions in FFTW that are always collective are: every function beginning with ‘fftw_mpi_plan’, as well as fftw_mpi_broadcast_wisdom and fftw_mpi_gather_wisdom. Also, the following functions from the ordinary FFTW interface are collective when they are applied to a plan created by an ‘fftw_mpi_plan’ function: fftw_execute, fftw_destroy_plan, and fftw_flops. fftw-3.3.3/doc/html/FFTW-MPI-Performance-Tips.html0000644000175000017500000001034512054411111016344 00000000000000 FFTW MPI Performance Tips - FFTW 3.3.3

Next: , Previous: Avoiding MPI Deadlocks, Up: Distributed-memory FFTW with MPI

6.10 FFTW MPI Performance Tips

In this section, we collect a few tips on getting the best performance out of FFTW's MPI transforms.

First, because of the 1d block distribution, FFTW's parallelization is currently limited by the size of the first dimension. (Multidimensional block distributions may be supported by a future version.) More generally, you should ideally arrange the dimensions so that FFTW can divide them equally among the processes. See Load balancing.

Second, if it is not too inconvenient, you should consider working with transposed output for multidimensional plans, as this saves a considerable amount of communications. See Transposed distributions.

Third, the fastest choices are generally either an in-place transform or an out-of-place transform with the FFTW_DESTROY_INPUT flag (which allows the input array to be used as scratch space). In-place is especially beneficial if the amount of data per process is large.

Fourth, if you have multiple arrays to transform at once, rather than calling FFTW's MPI transforms several times it usually seems to be faster to interleave the data and use the advanced interface. (This groups the communications together instead of requiring separate messages for each transform.) fftw-3.3.3/doc/html/Combining-MPI-and-Threads.html0000644000175000017500000001473212054411111016463 00000000000000 Combining MPI and Threads - FFTW 3.3.3

Next: , Previous: FFTW MPI Performance Tips, Up: Distributed-memory FFTW with MPI

6.11 Combining MPI and Threads

In certain cases, it may be advantageous to combine MPI (distributed-memory) and threads (shared-memory) parallelization. FFTW supports this, with certain caveats. For example, if you have a cluster of 4-processor shared-memory nodes, you may want to use threads within the nodes and MPI between the nodes, instead of MPI for all parallelization.

In particular, it is possible to seamlessly combine the MPI FFTW routines with the multi-threaded FFTW routines (see Multi-threaded FFTW). However, some care must be taken in the initialization code, which should look something like this: 

     int threads_ok;
     
     int main(int argc, char **argv)
     {
         int provided;
         MPI_Init_thread(&argc, &argv, MPI_THREAD_FUNNELED, &provided);
         threads_ok = provided >= MPI_THREAD_FUNNELED;
     
         if (threads_ok) threads_ok = fftw_init_threads();
         fftw_mpi_init();
     
         ...
         if (threads_ok) fftw_plan_with_nthreads(...);
         ...
     
         MPI_Finalize();
     }

First, note that instead of calling MPI_Init, you should call MPI_Init_threads, which is the initialization routine defined by the MPI-2 standard to indicate to MPI that your program will be multithreaded. We pass MPI_THREAD_FUNNELED, which indicates that we will only call MPI routines from the main thread. (FFTW will launch additional threads internally, but the extra threads will not call MPI code.) (You may also pass MPI_THREAD_SERIALIZED or MPI_THREAD_MULTIPLE, which requests additional multithreading support from the MPI implementation, but this is not required by FFTW.) The provided parameter returns what level of threads support is actually supported by your MPI implementation; this must be at least MPI_THREAD_FUNNELED if you want to call the FFTW threads routines, so we define a global variable threads_ok to record this. You should only call fftw_init_threads or fftw_plan_with_nthreads if threads_ok is true. For more information on thread safety in MPI, see the MPI and Threads section of the MPI-2 standard.

Second, we must call fftw_init_threads before fftw_mpi_init. This is critical for technical reasons having to do with how FFTW initializes its list of algorithms.

Then, if you call fftw_plan_with_nthreads(N), every MPI process will launch (up to) N threads to parallelize its transforms.

For example, in the hypothetical cluster of 4-processor nodes, you might wish to launch only a single MPI process per node, and then call fftw_plan_with_nthreads(4) on each process to use all processors in the nodes.

This may or may not be faster than simply using as many MPI processes as you have processors, however. On the one hand, using threads within a node eliminates the need for explicit message passing within the node. On the other hand, FFTW's transpose routines are not multi-threaded, and this means that the communications that do take place will not benefit from parallelization within the node. Moreover, many MPI implementations already have optimizations to exploit shared memory when it is available, so adding the multithreaded FFTW on top of this may be superfluous. fftw-3.3.3/doc/html/FFTW-MPI-Reference.html0000644000175000017500000000712412054411111015065 00000000000000 FFTW MPI Reference - FFTW 3.3.3

Next: , Previous: Combining MPI and Threads, Up: Distributed-memory FFTW with MPI

6.12 FFTW MPI Reference

This chapter provides a complete reference to all FFTW MPI functions, datatypes, and constants. See also FFTW Reference for information on functions and types in common with the serial interface.

fftw-3.3.3/doc/html/MPI-Files-and-Data-Types.html0000644000175000017500000000713412054411111016177 00000000000000 MPI Files and Data Types - FFTW 3.3.3

Next: , Previous: FFTW MPI Reference, Up: FFTW MPI Reference

6.12.1 MPI Files and Data Types

All programs using FFTW's MPI support should include its header file: 

     #include <fftw3-mpi.h>

Note that this header file includes the serial-FFTW fftw3.h header file, and also the mpi.h header file for MPI, so you need not include those files separately.

You must also link to both the FFTW MPI library and to the serial FFTW library. On Unix, this means adding -lfftw3_mpi -lfftw3 -lm at the end of the link command.

Different precisions are handled as in the serial interface: See Precision. That is, ‘fftw_’ functions become fftwf_ (in single precision) etcetera, and the libraries become -lfftw3f_mpi -lfftw3f -lm etcetera on Unix. Long-double precision is supported in MPI, but quad precision (‘fftwq_’) is not due to the lack of MPI support for this type. fftw-3.3.3/doc/html/MPI-Initialization.html0000644000175000017500000000763412054411111015420 00000000000000 MPI Initialization - FFTW 3.3.3

Next: , Previous: MPI Files and Data Types, Up: FFTW MPI Reference

6.12.2 MPI Initialization

Before calling any other FFTW MPI (‘fftw_mpi_’) function, and before importing any wisdom for MPI problems, you must call: 

     void fftw_mpi_init(void);

If FFTW threads support is used, however, fftw_mpi_init should be called after fftw_init_threads (see Combining MPI and Threads). Calling fftw_mpi_init additional times (before fftw_mpi_cleanup) has no effect.

If you want to deallocate all persistent data and reset FFTW to the pristine state it was in when you started your program, you can call: 

     void fftw_mpi_cleanup(void);

(This calls fftw_cleanup, so you need not call the serial cleanup routine too, although it is safe to do so.) After calling fftw_mpi_cleanup, all existing plans become undefined, and you should not attempt to execute or destroy them. You must call fftw_mpi_init again after fftw_mpi_cleanup if you want to resume using the MPI FFTW routines. fftw-3.3.3/doc/html/Using-MPI-Plans.html0000644000175000017500000001202512054411111014557 00000000000000 Using MPI Plans - FFTW 3.3.3

Next: , Previous: MPI Initialization, Up: FFTW MPI Reference

6.12.3 Using MPI Plans

Once an MPI plan is created, you can execute and destroy it using fftw_execute, fftw_destroy_plan, and the other functions in the serial interface that operate on generic plans (see Using Plans).

The fftw_execute and fftw_destroy_plan functions, applied to MPI plans, are collective calls: they must be called for all processes in the communicator that was used to create the plan.

You must not use the serial new-array plan-execution functions fftw_execute_dft and so on (see New-array Execute Functions) with MPI plans. Such functions are specialized to the problem type, and there are specific new-array execute functions for MPI plans: 

     void fftw_mpi_execute_dft(fftw_plan p, fftw_complex *in, fftw_complex *out);
     void fftw_mpi_execute_dft_r2c(fftw_plan p, double *in, fftw_complex *out);
     void fftw_mpi_execute_dft_c2r(fftw_plan p, fftw_complex *in, double *out);
     void fftw_mpi_execute_r2r(fftw_plan p, double *in, double *out);

These functions have the same restrictions as those of the serial new-array execute functions. They are always safe to apply to the same in and out arrays that were used to create the plan. They can only be applied to new arrarys if those arrays have the same types, dimensions, in-placeness, and alignment as the original arrays, where the best way to ensure the same alignment is to use FFTW's fftw_malloc and related allocation functions for all arrays (see Memory Allocation). Note that distributed transposes (see FFTW MPI Transposes) use fftw_mpi_execute_r2r, since they count as rank-zero r2r plans from FFTW's perspective. fftw-3.3.3/doc/html/MPI-Data-Distribution-Functions.html0000644000175000017500000002277412054411111017727 00000000000000 MPI Data Distribution Functions - FFTW 3.3.3

Next: , Previous: Using MPI Plans, Up: FFTW MPI Reference

6.12.4 MPI Data Distribution Functions

As described above (see MPI Data Distribution), in order to allocate your arrays, before creating a plan, you must first call one of the following routines to determine the required allocation size and the portion of the array locally stored on a given process. The MPI_Comm communicator passed here must be equivalent to the communicator used below for plan creation.

The basic interface for multidimensional transforms consists of the functions: 

     ptrdiff_t fftw_mpi_local_size_2d(ptrdiff_t n0, ptrdiff_t n1, MPI_Comm comm,
                                      ptrdiff_t *local_n0, ptrdiff_t *local_0_start);
     ptrdiff_t fftw_mpi_local_size_3d(ptrdiff_t n0, ptrdiff_t n1, ptrdiff_t n2,
                                      MPI_Comm comm,
                                      ptrdiff_t *local_n0, ptrdiff_t *local_0_start);
     ptrdiff_t fftw_mpi_local_size(int rnk, const ptrdiff_t *n, MPI_Comm comm,
                                   ptrdiff_t *local_n0, ptrdiff_t *local_0_start);
     
     ptrdiff_t fftw_mpi_local_size_2d_transposed(ptrdiff_t n0, ptrdiff_t n1, MPI_Comm comm,
                                                 ptrdiff_t *local_n0, ptrdiff_t *local_0_start,
                                                 ptrdiff_t *local_n1, ptrdiff_t *local_1_start);
     ptrdiff_t fftw_mpi_local_size_3d_transposed(ptrdiff_t n0, ptrdiff_t n1, ptrdiff_t n2,
                                                 MPI_Comm comm,
                                                 ptrdiff_t *local_n0, ptrdiff_t *local_0_start,
                                                 ptrdiff_t *local_n1, ptrdiff_t *local_1_start);
     ptrdiff_t fftw_mpi_local_size_transposed(int rnk, const ptrdiff_t *n, MPI_Comm comm,
                                              ptrdiff_t *local_n0, ptrdiff_t *local_0_start,
                                              ptrdiff_t *local_n1, ptrdiff_t *local_1_start);

These functions return the number of elements to allocate (complex numbers for DFT/r2c/c2r plans, real numbers for r2r plans), whereas the local_n0 and local_0_start return the portion (local_0_start to local_0_start + local_n0 - 1) of the first dimension of an n0 × n1 × n2 × … × nd-1 array that is stored on the local process. See Basic and advanced distribution interfaces. For FFTW_MPI_TRANSPOSED_OUT plans, the ‘_transposed’ variants are useful in order to also return the local portion of the first dimension in the n1 × n0 × n2 ×…× nd-1 transposed output. See Transposed distributions. The advanced interface for multidimensional transforms is: 

     ptrdiff_t fftw_mpi_local_size_many(int rnk, const ptrdiff_t *n, ptrdiff_t howmany,
                                        ptrdiff_t block0, MPI_Comm comm,
                                        ptrdiff_t *local_n0, ptrdiff_t *local_0_start);
     ptrdiff_t fftw_mpi_local_size_many_transposed(int rnk, const ptrdiff_t *n, ptrdiff_t howmany,
                                                   ptrdiff_t block0, ptrdiff_t block1, MPI_Comm comm,
                                                   ptrdiff_t *local_n0, ptrdiff_t *local_0_start,
                                                   ptrdiff_t *local_n1, ptrdiff_t *local_1_start);

These differ from the basic interface in only two ways. First, they allow you to specify block sizes block0 and block1 (the latter for the transposed output); you can pass FFTW_MPI_DEFAULT_BLOCK to use FFTW's default block size as in the basic interface. Second, you can pass a howmany parameter, corresponding to the advanced planning interface below: this is for transforms of contiguous howmany-tuples of numbers (howmany = 1 in the basic interface).

The corresponding basic and advanced routines for one-dimensional transforms (currently only complex DFTs) are: 

     ptrdiff_t fftw_mpi_local_size_1d(
                  ptrdiff_t n0, MPI_Comm comm, int sign, unsigned flags,
                  ptrdiff_t *local_ni, ptrdiff_t *local_i_start,
                  ptrdiff_t *local_no, ptrdiff_t *local_o_start);
     ptrdiff_t fftw_mpi_local_size_many_1d(
                  ptrdiff_t n0, ptrdiff_t howmany,
                  MPI_Comm comm, int sign, unsigned flags,
                  ptrdiff_t *local_ni, ptrdiff_t *local_i_start,
                  ptrdiff_t *local_no, ptrdiff_t *local_o_start);

As above, the return value is the number of elements to allocate (complex numbers, for complex DFTs). The local_ni and local_i_start arguments return the portion (local_i_start to local_i_start + local_ni - 1) of the 1d array that is stored on this process for the transform input, and local_no and local_o_start are the corresponding quantities for the input. The sign (FFTW_FORWARD or FFTW_BACKWARD) and flags must match the arguments passed when creating a plan. Although the inputs and outputs have different data distributions in general, it is guaranteed that the output data distribution of an FFTW_FORWARD plan will match the input data distribution of an FFTW_BACKWARD plan and vice versa; similarly for the FFTW_MPI_SCRAMBLED_OUT and FFTW_MPI_SCRAMBLED_IN flags. See One-dimensional distributions. fftw-3.3.3/doc/html/MPI-Plan-Creation.html0000644000175000017500000004242212054411111015057 00000000000000 MPI Plan Creation - FFTW 3.3.3

Next: , Previous: MPI Data Distribution Functions, Up: FFTW MPI Reference

6.12.5 MPI Plan Creation

Complex-data MPI DFTs

Plans for complex-data DFTs (see 2d MPI example) are created by: 

     fftw_plan fftw_mpi_plan_dft_1d(ptrdiff_t n0, fftw_complex *in, fftw_complex *out,
                                    MPI_Comm comm, int sign, unsigned flags);
     fftw_plan fftw_mpi_plan_dft_2d(ptrdiff_t n0, ptrdiff_t n1,
                                    fftw_complex *in, fftw_complex *out,
                                    MPI_Comm comm, int sign, unsigned flags);
     fftw_plan fftw_mpi_plan_dft_3d(ptrdiff_t n0, ptrdiff_t n1, ptrdiff_t n2,
                                    fftw_complex *in, fftw_complex *out,
                                    MPI_Comm comm, int sign, unsigned flags);
     fftw_plan fftw_mpi_plan_dft(int rnk, const ptrdiff_t *n,
                                 fftw_complex *in, fftw_complex *out,
                                 MPI_Comm comm, int sign, unsigned flags);
     fftw_plan fftw_mpi_plan_many_dft(int rnk, const ptrdiff_t *n,
                                      ptrdiff_t howmany, ptrdiff_t block, ptrdiff_t tblock,
                                      fftw_complex *in, fftw_complex *out,
                                      MPI_Comm comm, int sign, unsigned flags);

These are similar to their serial counterparts (see Complex DFTs) in specifying the dimensions, sign, and flags of the transform. The comm argument gives an MPI communicator that specifies the set of processes to participate in the transform; plan creation is a collective function that must be called for all processes in the communicator. The in and out pointers refer only to a portion of the overall transform data (see MPI Data Distribution) as specified by the ‘local_size’ functions in the previous section. Unless flags contains FFTW_ESTIMATE, these arrays are overwritten during plan creation as for the serial interface. For multi-dimensional transforms, any dimensions > 1 are supported; for one-dimensional transforms, only composite (non-prime) n0 are currently supported (unlike the serial FFTW). Requesting an unsupported transform size will yield a NULL plan. (As in the serial interface, highly composite sizes generally yield the best performance.)

The advanced-interface fftw_mpi_plan_many_dft additionally allows you to specify the block sizes for the first dimension (block) of the n0 × n1 × n2 × … × nd-1 input data and the first dimension (tblock) of the n1 × n0 × n2 ×…× nd-1 transposed data (at intermediate steps of the transform, and for the output if FFTW_TRANSPOSED_OUT is specified in flags). These must be the same block sizes as were passed to the corresponding ‘local_size’ function; you can pass FFTW_MPI_DEFAULT_BLOCK to use FFTW's default block size as in the basic interface. Also, the howmany parameter specifies that the transform is of contiguous howmany-tuples rather than individual complex numbers; this corresponds to the same parameter in the serial advanced interface (see Advanced Complex DFTs) with stride = howmany and dist = 1.

MPI flags

The flags can be any of those for the serial FFTW (see Planner Flags), and in addition may include one or more of the following MPI-specific flags, which improve performance at the cost of changing the output or input data formats.

  • FFTW_MPI_SCRAMBLED_OUT, FFTW_MPI_SCRAMBLED_IN: valid for 1d transforms only, these flags indicate that the output/input of the transform are in an undocumented “scrambled” order. A forward FFTW_MPI_SCRAMBLED_OUT transform can be inverted by a backward FFTW_MPI_SCRAMBLED_IN (times the usual 1/N normalization). See One-dimensional distributions.
  • FFTW_MPI_TRANSPOSED_OUT, FFTW_MPI_TRANSPOSED_IN: valid for multidimensional (rnk > 1) transforms only, these flags specify that the output or input of an n0 × n1 × n2 × … × nd-1 transform is transposed to n1 × n0 × n2 ×…× nd-1. See Transposed distributions.
Real-data MPI DFTs

Plans for real-input/output (r2c/c2r) DFTs (see Multi-dimensional MPI DFTs of Real Data) are created by: 

     fftw_plan fftw_mpi_plan_dft_r2c_2d(ptrdiff_t n0, ptrdiff_t n1,
                                        double *in, fftw_complex *out,
                                        MPI_Comm comm, unsigned flags);
     fftw_plan fftw_mpi_plan_dft_r2c_2d(ptrdiff_t n0, ptrdiff_t n1,
                                        double *in, fftw_complex *out,
                                        MPI_Comm comm, unsigned flags);
     fftw_plan fftw_mpi_plan_dft_r2c_3d(ptrdiff_t n0, ptrdiff_t n1, ptrdiff_t n2,
                                        double *in, fftw_complex *out,
                                        MPI_Comm comm, unsigned flags);
     fftw_plan fftw_mpi_plan_dft_r2c(int rnk, const ptrdiff_t *n,
                                     double *in, fftw_complex *out,
                                     MPI_Comm comm, unsigned flags);
     fftw_plan fftw_mpi_plan_dft_c2r_2d(ptrdiff_t n0, ptrdiff_t n1,
                                        fftw_complex *in, double *out,
                                        MPI_Comm comm, unsigned flags);
     fftw_plan fftw_mpi_plan_dft_c2r_2d(ptrdiff_t n0, ptrdiff_t n1,
                                        fftw_complex *in, double *out,
                                        MPI_Comm comm, unsigned flags);
     fftw_plan fftw_mpi_plan_dft_c2r_3d(ptrdiff_t n0, ptrdiff_t n1, ptrdiff_t n2,
                                        fftw_complex *in, double *out,
                                        MPI_Comm comm, unsigned flags);
     fftw_plan fftw_mpi_plan_dft_c2r(int rnk, const ptrdiff_t *n,
                                     fftw_complex *in, double *out,
                                     MPI_Comm comm, unsigned flags);

Similar to the serial interface (see Real-data DFTs), these transform logically n0 × n1 × n2 × … × nd-1 real data to/from n0 × n1 × n2 × … × (nd-1/2 + 1) complex data, representing the non-redundant half of the conjugate-symmetry output of a real-input DFT (see Multi-dimensional Transforms). However, the real array must be stored within a padded n0 × n1 × n2 × … × [2 (nd-1/2 + 1)]

array (much like the in-place serial r2c transforms, but here for out-of-place transforms as well). Currently, only multi-dimensional (rnk > 1) r2c/c2r transforms are supported (requesting a plan for rnk = 1 will yield NULL). As explained above (see Multi-dimensional MPI DFTs of Real Data), the data distribution of both the real and complex arrays is given by the ‘local_size’ function called for the dimensions of the complex array. Similar to the other planning functions, the input and output arrays are overwritten when the plan is created except in FFTW_ESTIMATE mode.

As for the complex DFTs above, there is an advance interface that allows you to manually specify block sizes and to transform contiguous howmany-tuples of real/complex numbers: 

     fftw_plan fftw_mpi_plan_many_dft_r2c
                   (int rnk, const ptrdiff_t *n, ptrdiff_t howmany,
                    ptrdiff_t iblock, ptrdiff_t oblock,
                    double *in, fftw_complex *out,
                    MPI_Comm comm, unsigned flags);
     fftw_plan fftw_mpi_plan_many_dft_c2r
                   (int rnk, const ptrdiff_t *n, ptrdiff_t howmany,
                    ptrdiff_t iblock, ptrdiff_t oblock,
                    fftw_complex *in, double *out,
                    MPI_Comm comm, unsigned flags);
MPI r2r transforms

There are corresponding plan-creation routines for r2r transforms (see More DFTs of Real Data), currently supporting multidimensional (rnk > 1) transforms only (rnk = 1 will yield a NULL plan): 

     fftw_plan fftw_mpi_plan_r2r_2d(ptrdiff_t n0, ptrdiff_t n1,
                                    double *in, double *out,
                                    MPI_Comm comm,
                                    fftw_r2r_kind kind0, fftw_r2r_kind kind1,
                                    unsigned flags);
     fftw_plan fftw_mpi_plan_r2r_3d(ptrdiff_t n0, ptrdiff_t n1, ptrdiff_t n2,
                                    double *in, double *out,
                                    MPI_Comm comm,
                                    fftw_r2r_kind kind0, fftw_r2r_kind kind1, fftw_r2r_kind kind2,
                                    unsigned flags);
     fftw_plan fftw_mpi_plan_r2r(int rnk, const ptrdiff_t *n,
                                 double *in, double *out,
                                 MPI_Comm comm, const fftw_r2r_kind *kind,
                                 unsigned flags);
     fftw_plan fftw_mpi_plan_many_r2r(int rnk, const ptrdiff_t *n,
                                      ptrdiff_t iblock, ptrdiff_t oblock,
                                      double *in, double *out,
                                      MPI_Comm comm, const fftw_r2r_kind *kind,
                                      unsigned flags);

The parameters are much the same as for the complex DFTs above, except that the arrays are of real numbers (and hence the outputs of the ‘local_size’ data-distribution functions should be interpreted as counts of real rather than complex numbers). Also, the kind parameters specify the r2r kinds along each dimension as for the serial interface (see Real-to-Real Transform Kinds). See Other Multi-dimensional Real-data MPI Transforms.

MPI transposition

FFTW also provides routines to plan a transpose of a distributed n0 by n1 array of real numbers, or an array of howmany-tuples of real numbers with specified block sizes (see FFTW MPI Transposes): 

     fftw_plan fftw_mpi_plan_transpose(ptrdiff_t n0, ptrdiff_t n1,
                                       double *in, double *out,
                                       MPI_Comm comm, unsigned flags);
     fftw_plan fftw_mpi_plan_many_transpose
                     (ptrdiff_t n0, ptrdiff_t n1, ptrdiff_t howmany,
                      ptrdiff_t block0, ptrdiff_t block1,
                      double *in, double *out, MPI_Comm comm, unsigned flags);

These plans are used with the fftw_mpi_execute_r2r new-array execute function (see Using MPI Plans), since they count as (rank zero) r2r plans from FFTW's perspective. fftw-3.3.3/doc/html/MPI-Wisdom-Communication.html0000644000175000017500000000652212054411111016471 00000000000000 MPI Wisdom Communication - FFTW 3.3.3

Previous: MPI Plan Creation, Up: FFTW MPI Reference

6.12.6 MPI Wisdom Communication

To facilitate synchronizing wisdom among the different MPI processes, we provide two functions: 

     void fftw_mpi_gather_wisdom(MPI_Comm comm);
     void fftw_mpi_broadcast_wisdom(MPI_Comm comm);

The fftw_mpi_gather_wisdom function gathers all wisdom in the given communicator comm to the process of rank 0 in the communicator: that process obtains the union of all wisdom on all the processes. As a side effect, some other processes will gain additional wisdom from other processes, but only process 0 will gain the complete union.

The fftw_mpi_broadcast_wisdom does the reverse: it exports wisdom from process 0 in comm to all other processes in the communicator, replacing any wisdom they currently have.

See FFTW MPI Wisdom. fftw-3.3.3/doc/html/FFTW-MPI-Fortran-Interface.html0000644000175000017500000002455712054411111016511 00000000000000 FFTW MPI Fortran Interface - FFTW 3.3.3

Previous: FFTW MPI Reference, Up: Distributed-memory FFTW with MPI

6.13 FFTW MPI Fortran Interface

The FFTW MPI interface is callable from modern Fortran compilers supporting the Fortran 2003 iso_c_binding standard for calling C functions. As described in Calling FFTW from Modern Fortran, this means that you can directly call FFTW's C interface from Fortran with only minor changes in syntax. There are, however, a few things specific to the MPI interface to keep in mind:

  • Instead of including fftw3.f03 as in Overview of Fortran interface, you should include 'fftw3-mpi.f03' (after use, intrinsic :: iso_c_binding as before). The fftw3-mpi.f03 file includes fftw3.f03, so you should not include them both yourself. (You will also want to include the MPI header file, usually via include 'mpif.h' or similar, although though this is not needed by fftw3-mpi.f03 per se.) (To use the ‘fftwl_’ long double extended-precision routines in supporting compilers, you should include fftw3f-mpi.f03 in addition to fftw3-mpi.f03. See Extended and quadruple precision in Fortran.)
  • Because of the different storage conventions between C and Fortran, you reverse the order of your array dimensions when passing them to FFTW (see Reversing array dimensions). This is merely a difference in notation and incurs no performance overhead. However, it means that, whereas in C the first dimension is distributed, in Fortran the last dimension of your array is distributed.
  • In Fortran, communicators are stored as integer types; there is no MPI_Comm type, nor is there any way to access a C MPI_Comm. Fortunately, this is taken care of for you by the FFTW Fortran interface: whenever the C interface expects an MPI_Comm type, you should pass the Fortran communicator as an integer.1
  • Because you need to call the ‘local_size’ function to find out how much space to allocate, and this may be larger than the local portion of the array (see MPI Data Distribution), you should always allocate your arrays dynamically using FFTW's allocation routines as described in Allocating aligned memory in Fortran. (Coincidentally, this also provides the best performance by guaranteeding proper data alignment.)
  • Because all sizes in the MPI FFTW interface are declared as ptrdiff_t in C, you should use integer(C_INTPTR_T) in Fortran (see FFTW Fortran type reference).
  • In Fortran, because of the language semantics, we generally recommend using the new-array execute functions for all plans, even in the common case where you are executing the plan on the same arrays for which the plan was created (see Plan execution in Fortran). However, note that in the MPI interface these functions are changed: fftw_execute_dft becomes fftw_mpi_execute_dft, etcetera. See Using MPI Plans.

For example, here is a Fortran code snippet to perform a distributed L × M complex DFT in-place. (This assumes you have already initialized MPI with MPI_init and have also performed call fftw_mpi_init.) 

       use, intrinsic :: iso_c_binding
       include 'fftw3-mpi.f03'
       integer(C_INTPTR_T), parameter :: L = ...
       integer(C_INTPTR_T), parameter :: M = ...
       type(C_PTR) :: plan, cdata
       complex(C_DOUBLE_COMPLEX), pointer :: data(:,:)
       integer(C_INTPTR_T) :: i, j, alloc_local, local_M, local_j_offset
     
     !   get local data size and allocate (note dimension reversal)
       alloc_local = fftw_mpi_local_size_2d(M, L, MPI_COMM_WORLD, &
                                            local_M, local_j_offset)
       cdata = fftw_alloc_complex(alloc_local)
       call c_f_pointer(cdata, data, [L,local_M])
     
     !   create MPI plan for in-place forward DFT (note dimension reversal)
       plan = fftw_mpi_plan_dft_2d(M, L, data, data, MPI_COMM_WORLD, &
                                   FFTW_FORWARD, FFTW_MEASURE)
     
     ! initialize data to some function my_function(i,j)
       do j = 1, local_M
         do i = 1, L
           data(i, j) = my_function(i, j + local_j_offset)
         end do
       end do
     
     ! compute transform (as many times as desired)
       call fftw_mpi_execute_dft(plan, data, data)
     
       call fftw_destroy_plan(plan)
       call fftw_free(cdata)

Note that when we called fftw_mpi_local_size_2d and fftw_mpi_plan_dft_2d with the dimensions in reversed order, since a L × M Fortran array is viewed by FFTW in C as a M × L array. This means that the array was distributed over the M dimension, the local portion of which is a L × local_M array in Fortran. (You must not use an allocate statement to allocate an L × local_M array, however; you must allocate alloc_local complex numbers, which may be greater than L * local_M, in order to reserve space for intermediate steps of the transform.) Finally, we mention that because C's array indices are zero-based, the local_j_offset argument can conveniently be interpreted as an offset in the 1-based j index (rather than as a starting index as in C).

If instead you had used the ior(FFTW_MEASURE, FFTW_MPI_TRANSPOSED_OUT) flag, the output of the transform would be a transposed M × local_L array, associated with the same cdata allocation (since the transform is in-place), and which you could declare with: 

       complex(C_DOUBLE_COMPLEX), pointer :: tdata(:,:)
       ...
       call c_f_pointer(cdata, tdata, [M,local_L])

where local_L would have been obtained by changing the fftw_mpi_local_size_2d call to: 

       alloc_local = fftw_mpi_local_size_2d_transposed(M, L, MPI_COMM_WORLD, &
                                local_M, local_j_offset, local_L, local_i_offset)

Footnotes

[1] Technically, this is because you aren't actually calling the C functions directly. You are calling wrapper functions that translate the communicator with MPI_Comm_f2c before calling the ordinary C interface. This is all done transparently, however, since the fftw3-mpi.f03 interface file renames the wrappers so that they are called in Fortran with the same names as the C interface functions.

fftw-3.3.3/doc/html/Calling-FFTW-from-Modern-Fortran.html0000644000175000017500000001102112054411111017700 00000000000000 Calling FFTW from Modern Fortran - FFTW 3.3.3

Next: , Previous: Distributed-memory FFTW with MPI, Up: Top

7 Calling FFTW from Modern Fortran

Fortran 2003 standardized ways for Fortran code to call C libraries, and this allows us to support a direct translation of the FFTW C API into Fortran. Compared to the legacy Fortran 77 interface (see Calling FFTW from Legacy Fortran), this direct interface offers many advantages, especially compile-time type-checking and aligned memory allocation. As of this writing, support for these C interoperability features seems widespread, having been implemented in nearly all major Fortran compilers (e.g. GNU, Intel, IBM, Oracle/Solaris, Portland Group, NAG). This chapter documents that interface. For the most part, since this interface allows Fortran to call the C interface directly, the usage is identical to C translated to Fortran syntax. However, there are a few subtle points such as memory allocation, wisdom, and data types that deserve closer attention.

fftw-3.3.3/doc/html/Overview-of-Fortran-interface.html0000644000175000017500000001723112054411111017557 00000000000000 Overview of Fortran interface - FFTW 3.3.3

Next: , Previous: Calling FFTW from Modern Fortran, Up: Calling FFTW from Modern Fortran

7.1 Overview of Fortran interface

FFTW provides a file fftw3.f03 that defines Fortran 2003 interfaces for all of its C routines, except for the MPI routines described elsewhere, which can be found in the same directory as fftw3.h (the C header file). In any Fortran subroutine where you want to use FFTW functions, you should begin with: 

       use, intrinsic :: iso_c_binding
       include 'fftw3.f03'

This includes the interface definitions and the standard iso_c_binding module (which defines the equivalents of C types). You can also put the FFTW functions into a module if you prefer (see Defining an FFTW module).

At this point, you can now call anything in the FFTW C interface directly, almost exactly as in C other than minor changes in syntax. For example: 

       type(C_PTR) :: plan
       complex(C_DOUBLE_COMPLEX), dimension(1024,1000) :: in, out
       plan = fftw_plan_dft_2d(1000,1024, in,out, FFTW_FORWARD,FFTW_ESTIMATE)
       ...
       call fftw_execute_dft(plan, in, out)
       ...
       call fftw_destroy_plan(plan)

A few important things to keep in mind are:

  • FFTW plans are type(C_PTR). Other C types are mapped in the obvious way via the iso_c_binding standard: int turns into integer(C_INT), fftw_complex turns into complex(C_DOUBLE_COMPLEX), double turns into real(C_DOUBLE), and so on. See FFTW Fortran type reference.
  • Functions in C become functions in Fortran if they have a return value, and subroutines in Fortran otherwise.
  • The ordering of the Fortran array dimensions must be reversed when they are passed to the FFTW plan creation, thanks to differences in array indexing conventions (see Multi-dimensional Array Format). This is unlike the legacy Fortran interface (see Fortran-interface routines), which reversed the dimensions for you. See Reversing array dimensions.
  • Using ordinary Fortran array declarations like this works, but may yield suboptimal performance because the data may not be not aligned to exploit SIMD instructions on modern proessors (see SIMD alignment and fftw_malloc). Better performance will often be obtained by allocating with ‘fftw_alloc’. See Allocating aligned memory in Fortran.
  • Similar to the legacy Fortran interface (see FFTW Execution in Fortran), we currently recommend not using fftw_execute but rather using the more specialized functions like fftw_execute_dft (see New-array Execute Functions). However, you should execute the plan on the same arrays as the ones for which you created the plan, unless you are especially careful. See Plan execution in Fortran. To prevent you from using fftw_execute by mistake, the fftw3.f03 file does not provide an fftw_execute interface declaration.
  • Multiple planner flags are combined with ior (equivalent to ‘|’ in C). e.g. FFTW_MEASURE | FFTW_DESTROY_INPUT becomes ior(FFTW_MEASURE, FFTW_DESTROY_INPUT). (You can also use ‘+’ as long as you don't try to include a given flag more than once.)
fftw-3.3.3/doc/html/Extended-and-quadruple-precision-in-Fortran.html0000644000175000017500000001037112054411111022304 00000000000000 Extended and quadruple precision in Fortran - FFTW 3.3.3

Previous: Overview of Fortran interface, Up: Overview of Fortran interface

7.1.1 Extended and quadruple precision in Fortran

If FFTW is compiled in long double (extended) precision (see Installation and Customization), you may be able to call the resulting fftwl_ routines (see Precision) from Fortran if your compiler supports the C_LONG_DOUBLE_COMPLEX type code.

Because some Fortran compilers do not support C_LONG_DOUBLE_COMPLEX, the fftwl_ declarations are segregated into a separate interface file fftw3l.f03, which you should include in addition to fftw3.f03 (which declares precision-independent ‘FFTW_’ constants): 

       use, intrinsic :: iso_c_binding
       include 'fftw3.f03'
       include 'fftw3l.f03'

We also support using the nonstandard __float128 quadruple-precision type provided by recent versions of gcc on 32- and 64-bit x86 hardware (see Installation and Customization), using the corresponding real(16) and complex(16) types supported by gfortran. The quadruple-precision ‘fftwq_’ functions (see Precision) are declared in a fftw3q.f03 interface file, which should be included in addition to fftw3l.f03, as above. You should also link with -lfftw3q -lquadmath -lm as in C. fftw-3.3.3/doc/html/Reversing-array-dimensions.html0000644000175000017500000001655212054411111017233 00000000000000 Reversing array dimensions - FFTW 3.3.3

Next: , Previous: Overview of Fortran interface, Up: Calling FFTW from Modern Fortran

7.2 Reversing array dimensions

A minor annoyance in calling FFTW from Fortran is that FFTW's array dimensions are defined in the C convention (row-major order), while Fortran's array dimensions are the opposite convention (column-major order). See Multi-dimensional Array Format. This is just a bookkeeping difference, with no effect on performance. The only consequence of this is that, whenever you create an FFTW plan for a multi-dimensional transform, you must always reverse the ordering of the dimensions.

For example, consider the three-dimensional (L × M × N) arrays: 

       complex(C_DOUBLE_COMPLEX), dimension(L,M,N) :: in, out

To plan a DFT for these arrays using fftw_plan_dft_3d, you could do: 

       plan = fftw_plan_dft_3d(N,M,L, in,out, FFTW_FORWARD,FFTW_ESTIMATE)

That is, from FFTW's perspective this is a N × M × L array. No data transposition need occur, as this is only notation. Similarly, to use the more generic routine fftw_plan_dft with the same arrays, you could do: 

       integer(C_INT), dimension(3) :: n = [N,M,L]
       plan = fftw_plan_dft_3d(3, n, in,out, FFTW_FORWARD,FFTW_ESTIMATE)

Note, by the way, that this is different from the legacy Fortran interface (see Fortran-interface routines), which automatically reverses the order of the array dimension for you. Here, you are calling the C interface directly, so there is no “translation” layer.

An important thing to keep in mind is the implication of this for multidimensional real-to-complex transforms (see Multi-Dimensional DFTs of Real Data). In C, a multidimensional real-to-complex DFT chops the last dimension roughly in half (N × M × L real input goes to N × M × L/2+1 complex output). In Fortran, because the array dimension notation is reversed, the first dimension of the complex data is chopped roughly in half. For example consider the ‘r2c’ transform of L × M × N real input in Fortran: 

       type(C_PTR) :: plan
       real(C_DOUBLE), dimension(L,M,N) :: in
       complex(C_DOUBLE_COMPLEX), dimension(L/2+1,M,N) :: out
       plan = fftw_plan_dft_r2c_3d(N,M,L, in,out, FFTW_ESTIMATE)
       ...
       call fftw_execute_dft_r2c(plan, in, out)

Alternatively, for an in-place r2c transform, as described in the C documentation we must pad the first dimension of the real input with an extra two entries (which are ignored by FFTW) so as to leave enough space for the complex output. The input is allocated as a 2[L/2+1] × M × N array, even though only L × M × N of it is actually used. In this example, we will allocate the array as a pointer type, using ‘fftw_alloc’ to ensure aligned memory for maximum performance (see Allocating aligned memory in Fortran); this also makes it easy to reference the same memory as both a real array and a complex array. 

       real(C_DOUBLE), pointer :: in(:,:,:)
       complex(C_DOUBLE_COMPLEX), pointer :: out(:,:,:)
       type(C_PTR) :: plan, data
       data = fftw_alloc_complex(int((L/2+1) * M * N, C_SIZE_T))
       call c_f_pointer(data, in, [2*(L/2+1),M,N])
       call c_f_pointer(data, out, [L/2+1,M,N])
       plan = fftw_plan_dft_r2c_3d(N,M,L, in,out, FFTW_ESTIMATE)
       ...
       call fftw_execute_dft_r2c(plan, in, out)
       ...
       call fftw_destroy_plan(plan)
       call fftw_free(data)
fftw-3.3.3/doc/html/FFTW-Fortran-type-reference.html0000644000175000017500000002243512054411111017074 00000000000000 FFTW Fortran type reference - FFTW 3.3.3

Next: , Previous: Reversing array dimensions, Up: Calling FFTW from Modern Fortran

7.3 FFTW Fortran type reference

The following are the most important type correspondences between the C interface and Fortran:

  • Plans (fftw_plan and variants) are type(C_PTR) (i.e. an opaque pointer).
  • The C floating-point types double, float, and long double correspond to real(C_DOUBLE), real(C_FLOAT), and real(C_LONG_DOUBLE), respectively. The C complex types fftw_complex, fftwf_complex, and fftwl_complex correspond in Fortran to complex(C_DOUBLE_COMPLEX), complex(C_FLOAT_COMPLEX), and complex(C_LONG_DOUBLE_COMPLEX), respectively. Just as in C (see Precision), the FFTW subroutines and types are prefixed with ‘fftw_’, fftwf_, and fftwl_ for the different precisions, and link to different libraries (-lfftw3, -lfftw3f, and -lfftw3l on Unix), but use the same include file fftw3.f03 and the same constants (all of which begin with ‘FFTW_’). The exception is long double precision, for which you should also include fftw3l.f03 (see Extended and quadruple precision in Fortran).
  • The C integer types int and unsigned (used for planner flags) become integer(C_INT). The C integer type ptrdiff_t (e.g. in the 64-bit Guru Interface) becomes integer(C_INTPTR_T), and size_t (in fftw_malloc etc.) becomes integer(C_SIZE_T).
  • The fftw_r2r_kind type (see Real-to-Real Transform Kinds) becomes integer(C_FFTW_R2R_KIND). The various constant values of the C enumerated type (FFTW_R2HC etc.) become simply integer constants of the same names in Fortran.
  • Numeric array pointer arguments (e.g. double *) become dimension(*), intent(out) arrays of the same type, or dimension(*), intent(in) if they are pointers to constant data (e.g. const int *). There are a few exceptions where numeric pointers refer to scalar outputs (e.g. for fftw_flops), in which case they are intent(out) scalar arguments in Fortran too. For the new-array execute functions (see New-array Execute Functions), the input arrays are declared dimension(*), intent(inout), since they can be modified in the case of in-place or FFTW_DESTROY_INPUT transforms.
  • Pointer return values (e.g double *) become type(C_PTR). (If they are pointers to arrays, as for fftw_alloc_real, you can convert them back to Fortran array pointers with the standard intrinsic function c_f_pointer.)
  • The fftw_iodim type in the guru interface (see Guru vector and transform sizes) becomes type(fftw_iodim) in Fortran, a derived data type (the Fortran analogue of C's struct) with three integer(C_INT) components: n, is, and os, with the same meanings as in C. The fftw_iodim64 type in the 64-bit guru interface (see 64-bit Guru Interface) is the same, except that its components are of type integer(C_INTPTR_T).
  • Using the wisdom import/export functions from Fortran is a bit tricky, and is discussed in Accessing the wisdom API from Fortran. In brief, the FILE * arguments map to type(C_PTR), const char * to character(C_CHAR), dimension(*), intent(in) (null-terminated!), and the generic read-char/write-char functions map to type(C_FUNPTR).

You may be wondering if you need to search-and-replace real(kind(0.0d0)) (or whatever your favorite Fortran spelling of “double precision” is) with real(C_DOUBLE) everywhere in your program, and similarly for complex and integer types. The answer is no; you can still use your existing types. As long as these types match their C counterparts, things should work without a hitch. The worst that can happen, e.g. in the (unlikely) event of a system where real(kind(0.0d0)) is different from real(C_DOUBLE), is that the compiler will give you a type-mismatch error. That is, if you don't use the iso_c_binding kinds you need to accept at least the theoretical possibility of having to change your code in response to compiler errors on some future machine, but you don't need to worry about silently compiling incorrect code that yields runtime errors. fftw-3.3.3/doc/html/Plan-execution-in-Fortran.html0000644000175000017500000001433212054411111016707 00000000000000 Plan execution in Fortran - FFTW 3.3.3

Next: , Previous: FFTW Fortran type reference, Up: Calling FFTW from Modern Fortran

7.4 Plan execution in Fortran

In C, in order to use a plan, one normally calls fftw_execute, which executes the plan to perform the transform on the input/output arrays passed when the plan was created (see Using Plans). The corresponding subroutine call in modern Fortran is: 

      call fftw_execute(plan)

However, we have had reports that this causes problems with some recent optimizing Fortran compilers. The problem is, because the input/output arrays are not passed as explicit arguments to fftw_execute, the semantics of Fortran (unlike C) allow the compiler to assume that the input/output arrays are not changed by fftw_execute. As a consequence, certain compilers end up repositioning the call to fftw_execute, assuming incorrectly that it does nothing to the arrays.

There are various workarounds to this, but the safest and simplest thing is to not use fftw_execute in Fortran. Instead, use the functions described in New-array Execute Functions, which take the input/output arrays as explicit arguments. For example, if the plan is for a complex-data DFT and was created for the arrays in and out, you would do: 

      call fftw_execute_dft(plan, in, out)

There are a few things to be careful of, however:

  • You must use the correct type of execute function, matching the way the plan was created. Complex DFT plans should use fftw_execute_dft, Real-input (r2c) DFT plans should use use fftw_execute_dft_r2c, and real-output (c2r) DFT plans should use fftw_execute_dft_c2r. The various r2r plans should use fftw_execute_r2r. Fortunately, if you use the wrong one you will get a compile-time type-mismatch error (unlike legacy Fortran).
  • You should normally pass the same input/output arrays that were used when creating the plan. This is always safe.
  • If you pass different input/output arrays compared to those used when creating the plan, you must abide by all the restrictions of the new-array execute functions (see New-array Execute Functions). The most tricky of these is the requirement that the new arrays have the same alignment as the original arrays; the best (and possibly only) way to guarantee this is to use the ‘fftw_alloc’ functions to allocate your arrays (see Allocating aligned memory in Fortran). Alternatively, you can use the FFTW_UNALIGNED flag when creating the plan, in which case the plan does not depend on the alignment, but this may sacrifice substantial performance on architectures (like x86) with SIMD instructions (see SIMD alignment and fftw_malloc).
fftw-3.3.3/doc/html/Allocating-aligned-memory-in-Fortran.html0000644000175000017500000001524712054411111021006 00000000000000 Allocating aligned memory in Fortran - FFTW 3.3.3

Next: , Previous: Plan execution in Fortran, Up: Calling FFTW from Modern Fortran

7.5 Allocating aligned memory in Fortran

In order to obtain maximum performance in FFTW, you should store your data in arrays that have been specially aligned in memory (see SIMD alignment and fftw_malloc). Enforcing alignment also permits you to safely use the new-array execute functions (see New-array Execute Functions) to apply a given plan to more than one pair of in/out arrays. Unfortunately, standard Fortran arrays do not provide any alignment guarantees. The only way to allocate aligned memory in standard Fortran is to allocate it with an external C function, like the fftw_alloc_real and fftw_alloc_complex functions. Fortunately, Fortran 2003 provides a simple way to associate such allocated memory with a standard Fortran array pointer that you can then use normally.

We therefore recommend allocating all your input/output arrays using the following technique:

  1. Declare a pointer, arr, to your array of the desired type and dimensions. For example, real(C_DOUBLE), pointer :: a(:,:) for a 2d real array, or complex(C_DOUBLE_COMPLEX), pointer :: a(:,:,:) for a 3d complex array.
  2. The number of elements to allocate must be an integer(C_SIZE_T). You can either declare a variable of this type, e.g. integer(C_SIZE_T) :: sz, to store the number of elements to allocate, or you can use the int(..., C_SIZE_T) intrinsic function. e.g. set sz = L * M * N or use int(L * M * N, C_SIZE_T) for an L × M × N array.
  3. Declare a type(C_PTR) :: p to hold the return value from FFTW's allocation routine. Set p = fftw_alloc_real(sz) for a real array, or p = fftw_alloc_complex(sz) for a complex array.
  4. Associate your pointer arr with the allocated memory p using the standard c_f_pointer subroutine: call c_f_pointer(p, arr, [...dimensions...]), where [...dimensions...]) are an array of the dimensions of the array (in the usual Fortran order). e.g. call c_f_pointer(p, arr, [L,M,N]) for an L × M × N array. (Alternatively, you can omit the dimensions argument if you specified the shape explicitly when declaring arr.) You can now use arr as a usual multidimensional array.
  5. When you are done using the array, deallocate the memory by call fftw_free(p) on p.

For example, here is how we would allocate an L × M 2d real array: 

       real(C_DOUBLE), pointer :: arr(:,:)
       type(C_PTR) :: p
       p = fftw_alloc_real(int(L * M, C_SIZE_T))
       call c_f_pointer(p, arr, [L,M])
       ...use arr and arr(i,j) as usual...
       call fftw_free(p)

and here is an L × M × N 3d complex array: 

       complex(C_DOUBLE_COMPLEX), pointer :: arr(:,:,:)
       type(C_PTR) :: p
       p = fftw_alloc_complex(int(L * M * N, C_SIZE_T))
       call c_f_pointer(p, arr, [L,M,N])
       ...use arr and arr(i,j,k) as usual...
       call fftw_free(p)

See Reversing array dimensions for an example allocating a single array and associating both real and complex array pointers with it, for in-place real-to-complex transforms. fftw-3.3.3/doc/html/Accessing-the-wisdom-API-from-Fortran.html0000644000175000017500000000766512054411111020750 00000000000000 Accessing the wisdom API from Fortran - FFTW 3.3.3

Next: , Previous: Allocating aligned memory in Fortran, Up: Calling FFTW from Modern Fortran

7.6 Accessing the wisdom API from Fortran

As explained in Words of Wisdom-Saving Plans, FFTW provides a “wisdom” API for saving plans to disk so that they can be recreated quickly. The C API for exporting (see Wisdom Export) and importing (see Wisdom Import) wisdom is somewhat tricky to use from Fortran, however, because of differences in file I/O and string types between C and Fortran.

fftw-3.3.3/doc/html/Wisdom-File-Export_002fImport-from-Fortran.html0000644000175000017500000001146412054411111021674 00000000000000 Wisdom File Export/Import from Fortran - FFTW 3.3.3

Next: , Previous: Accessing the wisdom API from Fortran, Up: Accessing the wisdom API from Fortran

7.6.1 Wisdom File Export/Import from Fortran

The easiest way to export and import wisdom is to do so using fftw_export_wisdom_to_filename and fftw_wisdom_from_filename. The only trick is that these require you to pass a C string, which is an array of type CHARACTER(C_CHAR) that is terminated by C_NULL_CHAR. You can call them like this: 

       integer(C_INT) :: ret
       ret = fftw_export_wisdom_to_filename(C_CHAR_'my_wisdom.dat' // C_NULL_CHAR)
       if (ret .eq. 0) stop 'error exporting wisdom to file'
       ret = fftw_import_wisdom_from_filename(C_CHAR_'my_wisdom.dat' // C_NULL_CHAR)
       if (ret .eq. 0) stop 'error importing wisdom from file'

Note that prepending ‘C_CHAR_’ is needed to specify that the literal string is of kind C_CHAR, and we null-terminate the string by appending ‘// C_NULL_CHAR’. These functions return an integer(C_INT) (ret) which is 0 if an error occurred during export/import and nonzero otherwise.

It is also possible to use the lower-level routines fftw_export_wisdom_to_file and fftw_import_wisdom_from_file, which accept parameters of the C type FILE*, expressed in Fortran as type(C_PTR). However, you are then responsible for creating the FILE* yourself. You can do this by using iso_c_binding to define Fortran intefaces for the C library functions fopen and fclose, which is a bit strange in Fortran but workable. fftw-3.3.3/doc/html/Wisdom-String-Export_002fImport-from-Fortran.html0000644000175000017500000001224712054411111022263 00000000000000 Wisdom String Export/Import from Fortran - FFTW 3.3.3

Next: , Previous: Wisdom File Export/Import from Fortran, Up: Accessing the wisdom API from Fortran

7.6.2 Wisdom String Export/Import from Fortran

Dealing with FFTW's C string export/import is a bit more painful. In particular, the fftw_export_wisdom_to_string function requires you to deal with a dynamically allocated C string. To get its length, you must define an interface to the C strlen function, and to deallocate it you must define an interface to C free: 

       use, intrinsic :: iso_c_binding
       interface
         integer(C_INT) function strlen(s) bind(C, name='strlen')
           import
           type(C_PTR), value :: s
         end function strlen
         subroutine free(p) bind(C, name='free')
           import
           type(C_PTR), value :: p
         end subroutine free
       end interface

Given these definitions, you can then export wisdom to a Fortran character array: 

       character(C_CHAR), pointer :: s(:)
       integer(C_SIZE_T) :: slen
       type(C_PTR) :: p
       p = fftw_export_wisdom_to_string()
       if (.not. c_associated(p)) stop 'error exporting wisdom'
       slen = strlen(p)
       call c_f_pointer(p, s, [slen+1])
       ...
       call free(p)

Note that slen is the length of the C string, but the length of the array is slen+1 because it includes the terminating null character. (You can omit the ‘+1’ if you don't want Fortran to know about the null character.) The standard c_associated function checks whether p is a null pointer, which is returned by fftw_export_wisdom_to_string if there was an error.

To import wisdom from a string, use fftw_import_wisdom_from_string as usual; note that the argument of this function must be a character(C_CHAR) that is terminated by the C_NULL_CHAR character, like the s array above. fftw-3.3.3/doc/html/Wisdom-Generic-Export_002fImport-from-Fortran.html0000644000175000017500000001225012054411111022363 00000000000000 Wisdom Generic Export/Import from Fortran - FFTW 3.3.3

Previous: Wisdom String Export/Import from Fortran, Up: Accessing the wisdom API from Fortran

7.6.3 Wisdom Generic Export/Import from Fortran

The most generic wisdom export/import functions allow you to provide an arbitrary callback function to read/write one character at a time in any way you want. However, your callback function must be written in a special way, using the bind(C) attribute to be passed to a C interface.

In particular, to call the generic wisdom export function fftw_export_wisdom, you would write a callback subroutine of the form: 

       subroutine my_write_char(c, p) bind(C)
         use, intrinsic :: iso_c_binding
         character(C_CHAR), value :: c
         type(C_PTR), value :: p
         ...write c...
       end subroutine my_write_char

Given such a subroutine (along with the corresponding interface definition), you could then export wisdom using: 

       call fftw_export_wisdom(c_funloc(my_write_char), p)

The standard c_funloc intrinsic converts a Fortran bind(C) subroutine into a C function pointer. The parameter p is a type(C_PTR) to any arbitrary data that you want to pass to my_write_char (or C_NULL_PTR if none). (Note that you can get a C pointer to Fortran data using the intrinsic c_loc, and convert it back to a Fortran pointer in my_write_char using c_f_pointer.)

Similarly, to use the generic fftw_import_wisdom, you would define a callback function of the form: 

       integer(C_INT) function my_read_char(p) bind(C)
         use, intrinsic :: iso_c_binding
         type(C_PTR), value :: p
         character :: c
         ...read a character c...
         my_read_char = ichar(c, C_INT)
       end function my_read_char
     
       ....
     
       integer(C_INT) :: ret
       ret = fftw_import_wisdom(c_funloc(my_read_char), p)
       if (ret .eq. 0) stop 'error importing wisdom'

Your function can return -1 if the end of the input is reached. Again, p is an arbitrary type(C_PTR that is passed through to your function. fftw_import_wisdom returns 0 if an error occurred and nonzero otherwise. fftw-3.3.3/doc/html/Defining-an-FFTW-module.html0000644000175000017500000000700212054411111016141 00000000000000 Defining an FFTW module - FFTW 3.3.3

Previous: Accessing the wisdom API from Fortran, Up: Calling FFTW from Modern Fortran

7.7 Defining an FFTW module

Rather than using the include statement to include the fftw3.f03 interface file in any subroutine where you want to use FFTW, you might prefer to define an FFTW Fortran module. FFTW does not install itself as a module, primarily because fftw3.f03 can be shared between different Fortran compilers while modules (in general) cannot. However, it is trivial to define your own FFTW module if you want. Just create a file containing: 

       module FFTW3
         use, intrinsic :: iso_c_binding
         include 'fftw3.f03'
       end module

Compile this file into a module as usual for your compiler (e.g. with gfortran -c you will get a file fftw3.mod). Now, instead of include 'fftw3.f03', whenever you want to use FFTW routines you can just do: 

       use FFTW3

as usual for Fortran modules. (You still need to link to the FFTW library, of course.) fftw-3.3.3/doc/html/Calling-FFTW-from-Legacy-Fortran.html0000644000175000017500000001107112054411111017665 00000000000000 Calling FFTW from Legacy Fortran - FFTW 3.3.3

Next: , Previous: Calling FFTW from Modern Fortran, Up: Top

8 Calling FFTW from Legacy Fortran

This chapter describes the interface to FFTW callable by Fortran code in older compilers not supporting the Fortran 2003 C interoperability features (see Calling FFTW from Modern Fortran). This interface has the major disadvantage that it is not type-checked, so if you mistake the argument types or ordering then your program will not have any compiler errors, and will likely crash at runtime. So, greater care is needed. Also, technically interfacing older Fortran versions to C is nonstandard, but in practice we have found that the techniques used in this chapter have worked with all known Fortran compilers for many years.

The legacy Fortran interface differs from the C interface only in the prefix (‘dfftw_’ instead of ‘fftw_’ in double precision) and a few other minor details. This Fortran interface is included in the FFTW libraries by default, unless a Fortran compiler isn't found on your system or --disable-fortran is included in the configure flags. We assume here that the reader is already familiar with the usage of FFTW in C, as described elsewhere in this manual.

The MPI parallel interface to FFTW is not currently available to legacy Fortran.

fftw-3.3.3/doc/html/Fortran_002dinterface-routines.html0000644000175000017500000001753512054411111017676 00000000000000 Fortran-interface routines - FFTW 3.3.3

Next: , Previous: Calling FFTW from Legacy Fortran, Up: Calling FFTW from Legacy Fortran

8.1 Fortran-interface routines

Nearly all of the FFTW functions have Fortran-callable equivalents. The name of the legacy Fortran routine is the same as that of the corresponding C routine, but with the ‘fftw_’ prefix replaced by ‘dfftw_’.1 The single and long-double precision versions use ‘sfftw_’ and ‘lfftw_’, respectively, instead of ‘fftwf_’ and ‘fftwl_’; quadruple precision (real*16) is available on some systems as ‘fftwq_’ (see Precision). (Note that long double on x86 hardware is usually at most 80-bit extended precision, not quadruple precision.)

For the most part, all of the arguments to the functions are the same, with the following exceptions:

  • plan variables (what would be of type fftw_plan in C), must be declared as a type that is at least as big as a pointer (address) on your machine. We recommend using integer*8 everywhere, since this should always be big enough.
  • Any function that returns a value (e.g. fftw_plan_dft) is converted into a subroutine. The return value is converted into an additional first parameter of this subroutine.2
  • The Fortran routines expect multi-dimensional arrays to be in column-major order, which is the ordinary format of Fortran arrays (see Multi-dimensional Array Format). They do this transparently and costlessly simply by reversing the order of the dimensions passed to FFTW, but this has one important consequence for multi-dimensional real-complex transforms, discussed below.
  • Wisdom import and export is somewhat more tricky because one cannot easily pass files or strings between C and Fortran; see Wisdom of Fortran?.
  • Legacy Fortran cannot use the fftw_malloc dynamic-allocation routine. If you want to exploit the SIMD FFTW (see SIMD alignment and fftw_malloc), you'll need to figure out some other way to ensure that your arrays are at least 16-byte aligned.
  • Since Fortran 77 does not have data structures, the fftw_iodim structure from the guru interface (see Guru vector and transform sizes) must be split into separate arguments. In particular, any fftw_iodim array arguments in the C guru interface become three integer array arguments (n, is, and os) in the Fortran guru interface, all of whose lengths should be equal to the corresponding rank argument.
  • The guru planner interface in Fortran does not do any automatic translation between column-major and row-major; you are responsible for setting the strides etcetera to correspond to your Fortran arrays. However, as a slight bug that we are preserving for backwards compatibility, the ‘plan_guru_r2r’ in Fortran does reverse the order of its kind array parameter, so the kind array of that routine should be in the reverse of the order of the iodim arrays (see above).

In general, you should take care to use Fortran data types that correspond to (i.e. are the same size as) the C types used by FFTW. In practice, this correspondence is usually straightforward (i.e. integer corresponds to int, real corresponds to float, etcetera). The native Fortran double/single-precision complex type should be compatible with fftw_complex/fftwf_complex. Such simple correspondences are assumed in the examples below.

Footnotes

[1] Technically, Fortran 77 identifiers are not allowed to have more than 6 characters, nor may they contain underscores. Any compiler that enforces this limitation doesn't deserve to link to FFTW.

[2] The reason for this is that some Fortran implementations seem to have trouble with C function return values, and vice versa.

fftw-3.3.3/doc/html/FFTW-Constants-in-Fortran.html0000644000175000017500000000754212054411111016541 00000000000000 FFTW Constants in Fortran - FFTW 3.3.3

Next: , Previous: Fortran-interface routines, Up: Calling FFTW from Legacy Fortran

8.2 FFTW Constants in Fortran

When creating plans in FFTW, a number of constants are used to specify options, such as FFTW_MEASURE or FFTW_ESTIMATE. The same constants must be used with the wrapper routines, but of course the C header files where the constants are defined can't be incorporated directly into Fortran code.

Instead, we have placed Fortran equivalents of the FFTW constant definitions in the file fftw3.f, which can be found in the same directory as fftw3.h. If your Fortran compiler supports a preprocessor of some sort, you should be able to include or #include this file; otherwise, you can paste it directly into your code.

In C, you combine different flags (like FFTW_PRESERVE_INPUT and FFTW_MEASURE) using the ‘|’ operator; in Fortran you should just use ‘+’. (Take care not to add in the same flag more than once, though. Alternatively, you can use the ior intrinsic function standardized in Fortran 95.) fftw-3.3.3/doc/html/FFTW-Execution-in-Fortran.html0000644000175000017500000001331012054411111016516 00000000000000 FFTW Execution in Fortran - FFTW 3.3.3

Next: , Previous: FFTW Constants in Fortran, Up: Calling FFTW from Legacy Fortran

8.3 FFTW Execution in Fortran

In C, in order to use a plan, one normally calls fftw_execute, which executes the plan to perform the transform on the input/output arrays passed when the plan was created (see Using Plans). The corresponding subroutine call in legacy Fortran is: 

             call dfftw_execute(plan)

However, we have had reports that this causes problems with some recent optimizing Fortran compilers. The problem is, because the input/output arrays are not passed as explicit arguments to dfftw_execute, the semantics of Fortran (unlike C) allow the compiler to assume that the input/output arrays are not changed by dfftw_execute. As a consequence, certain compilers end up optimizing out or repositioning the call to dfftw_execute, assuming incorrectly that it does nothing.

There are various workarounds to this, but the safest and simplest thing is to not use dfftw_execute in Fortran. Instead, use the functions described in New-array Execute Functions, which take the input/output arrays as explicit arguments. For example, if the plan is for a complex-data DFT and was created for the arrays in and out, you would do: 

             call dfftw_execute_dft(plan, in, out)

There are a few things to be careful of, however:

  • You must use the correct type of execute function, matching the way the plan was created. Complex DFT plans should use dfftw_execute_dft, Real-input (r2c) DFT plans should use use dfftw_execute_dft_r2c, and real-output (c2r) DFT plans should use dfftw_execute_dft_c2r. The various r2r plans should use dfftw_execute_r2r.
  • You should normally pass the same input/output arrays that were used when creating the plan. This is always safe.
  • If you pass different input/output arrays compared to those used when creating the plan, you must abide by all the restrictions of the new-array execute functions (see New-array Execute Functions). The most difficult of these, in Fortran, is the requirement that the new arrays have the same alignment as the original arrays, because there seems to be no way in legacy Fortran to obtain guaranteed-aligned arrays (analogous to fftw_malloc in C). You can, of course, use the FFTW_UNALIGNED flag when creating the plan, in which case the plan does not depend on the alignment, but this may sacrifice substantial performance on architectures (like x86) with SIMD instructions (see SIMD alignment and fftw_malloc).
fftw-3.3.3/doc/html/Fortran-Examples.html0000644000175000017500000001555212054411111015173 00000000000000 Fortran Examples - FFTW 3.3.3

Next: , Previous: FFTW Execution in Fortran, Up: Calling FFTW from Legacy Fortran

8.4 Fortran Examples

In C, you might have something like the following to transform a one-dimensional complex array: 

             fftw_complex in[N], out[N];
             fftw_plan plan;
     
             plan = fftw_plan_dft_1d(N,in,out,FFTW_FORWARD,FFTW_ESTIMATE);
             fftw_execute(plan);
             fftw_destroy_plan(plan);

In Fortran, you would use the following to accomplish the same thing: 

             double complex in, out
             dimension in(N), out(N)
             integer*8 plan
     
             call dfftw_plan_dft_1d(plan,N,in,out,FFTW_FORWARD,FFTW_ESTIMATE)
             call dfftw_execute_dft(plan, in, out)
             call dfftw_destroy_plan(plan)

Notice how all routines are called as Fortran subroutines, and the plan is returned via the first argument to dfftw_plan_dft_1d. Notice also that we changed fftw_execute to dfftw_execute_dft (see FFTW Execution in Fortran). To do the same thing, but using 8 threads in parallel (see Multi-threaded FFTW), you would simply prefix these calls with: 

             integer iret
             call dfftw_init_threads(iret)
             call dfftw_plan_with_nthreads(8)

(You might want to check the value of iret: if it is zero, it indicates an unlikely error during thread initialization.)

To transform a three-dimensional array in-place with C, you might do: 

             fftw_complex arr[L][M][N];
             fftw_plan plan;
     
             plan = fftw_plan_dft_3d(L,M,N, arr,arr,
                                     FFTW_FORWARD, FFTW_ESTIMATE);
             fftw_execute(plan);
             fftw_destroy_plan(plan);

In Fortran, you would use this instead: 

             double complex arr
             dimension arr(L,M,N)
             integer*8 plan
     
             call dfftw_plan_dft_3d(plan, L,M,N, arr,arr,
            &                       FFTW_FORWARD, FFTW_ESTIMATE)
             call dfftw_execute_dft(plan, arr, arr)
             call dfftw_destroy_plan(plan)

Note that we pass the array dimensions in the “natural” order in both C and Fortran.

To transform a one-dimensional real array in Fortran, you might do: 

             double precision in
             dimension in(N)
             double complex out
             dimension out(N/2 + 1)
             integer*8 plan
     
             call dfftw_plan_dft_r2c_1d(plan,N,in,out,FFTW_ESTIMATE)
             call dfftw_execute_dft_r2c(plan, in, out)
             call dfftw_destroy_plan(plan)

To transform a two-dimensional real array, out of place, you might use the following: 

             double precision in
             dimension in(M,N)
             double complex out
             dimension out(M/2 + 1, N)
             integer*8 plan
     
             call dfftw_plan_dft_r2c_2d(plan,M,N,in,out,FFTW_ESTIMATE)
             call dfftw_execute_dft_r2c(plan, in, out)
             call dfftw_destroy_plan(plan)

Important: Notice that it is the first dimension of the complex output array that is cut in half in Fortran, rather than the last dimension as in C. This is a consequence of the interface routines reversing the order of the array dimensions passed to FFTW so that the Fortran program can use its ordinary column-major order. fftw-3.3.3/doc/html/Wisdom-of-Fortran_003f.html0000644000175000017500000001063012054411111016001 00000000000000 Wisdom of Fortran? - FFTW 3.3.3

Previous: Fortran Examples, Up: Calling FFTW from Legacy Fortran

8.5 Wisdom of Fortran?

In this section, we discuss how one can import/export FFTW wisdom (saved plans) to/from a Fortran program; we assume that the reader is already familiar with wisdom, as described in Words of Wisdom-Saving Plans.

The basic problem is that is difficult to (portably) pass files and strings between Fortran and C, so we cannot provide a direct Fortran equivalent to the fftw_export_wisdom_to_file, etcetera, functions. Fortran interfaces are provided for the functions that do not take file/string arguments, however: dfftw_import_system_wisdom, dfftw_import_wisdom, dfftw_export_wisdom, and dfftw_forget_wisdom.

So, for example, to import the system-wide wisdom, you would do: 

             integer isuccess
             call dfftw_import_system_wisdom(isuccess)

As usual, the C return value is turned into a first parameter; isuccess is non-zero on success and zero on failure (e.g. if there is no system wisdom installed).

If you want to import/export wisdom from/to an arbitrary file or elsewhere, you can employ the generic dfftw_import_wisdom and dfftw_export_wisdom functions, for which you must supply a subroutine to read/write one character at a time. The FFTW package contains an example file doc/f77_wisdom.f demonstrating how to implement import_wisdom_from_file and export_wisdom_to_file subroutines in this way. (These routines cannot be compiled into the FFTW library itself, lest all FFTW-using programs be required to link with the Fortran I/O library.) fftw-3.3.3/doc/html/Upgrading-from-FFTW-version-2.html0000644000175000017500000003355612054411111017257 00000000000000 Upgrading from FFTW version 2 - FFTW 3.3.3

Next: , Previous: Calling FFTW from Legacy Fortran, Up: Top

9 Upgrading from FFTW version 2

In this chapter, we outline the process for updating codes designed for the older FFTW 2 interface to work with FFTW 3. The interface for FFTW 3 is not backwards-compatible with the interface for FFTW 2 and earlier versions; codes written to use those versions will fail to link with FFTW 3. Nor is it possible to write “compatibility wrappers” to bridge the gap (at least not efficiently), because FFTW 3 has different semantics from previous versions. However, upgrading should be a straightforward process because the data formats are identical and the overall style of planning/execution is essentially the same.

Unlike FFTW 2, there are no separate header files for real and complex transforms (or even for different precisions) in FFTW 3; all interfaces are defined in the <fftw3.h> header file.

Numeric Types

The main difference in data types is that fftw_complex in FFTW 2 was defined as a struct with macros c_re and c_im for accessing the real/imaginary parts. (This is binary-compatible with FFTW 3 on any machine except perhaps for some older Crays in single precision.) The equivalent macros for FFTW 3 are: 

     #define c_re(c) ((c)[0])
     #define c_im(c) ((c)[1])

This does not work if you are using the C99 complex type, however, unless you insert a double* typecast into the above macros (see Complex numbers).

Also, FFTW 2 had an fftw_real typedef that was an alias for double (in double precision). In FFTW 3 you should just use double (or whatever precision you are employing).

Plans

The major difference between FFTW 2 and FFTW 3 is in the planning/execution division of labor. In FFTW 2, plans were found for a given transform size and type, and then could be applied to any arrays and for any multiplicity/stride parameters. In FFTW 3, you specify the particular arrays, stride parameters, etcetera when creating the plan, and the plan is then executed for those arrays (unless the guru interface is used) and those parameters only. (FFTW 2 had “specific planner” routines that planned for a particular array and stride, but the plan could still be used for other arrays and strides.) That is, much of the information that was formerly specified at execution time is now specified at planning time.

Like FFTW 2's specific planner routines, the FFTW 3 planner overwrites the input/output arrays unless you use FFTW_ESTIMATE.

FFTW 2 had separate data types fftw_plan, fftwnd_plan, rfftw_plan, and rfftwnd_plan for complex and real one- and multi-dimensional transforms, and each type had its own ‘destroy’ function. In FFTW 3, all plans are of type fftw_plan and all are destroyed by fftw_destroy_plan(plan).

Where you formerly used fftw_create_plan and fftw_one to plan and compute a single 1d transform, you would now use fftw_plan_dft_1d to plan the transform. If you used the generic fftw function to execute the transform with multiplicity (howmany) and stride parameters, you would now use the advanced interface fftw_plan_many_dft to specify those parameters. The plans are now executed with fftw_execute(plan), which takes all of its parameters (including the input/output arrays) from the plan.

In-place transforms no longer interpret their output argument as scratch space, nor is there an FFTW_IN_PLACE flag. You simply pass the same pointer for both the input and output arguments. (Previously, the output ostride and odist parameters were ignored for in-place transforms; now, if they are specified via the advanced interface, they are significant even in the in-place case, although they should normally equal the corresponding input parameters.)

The FFTW_ESTIMATE and FFTW_MEASURE flags have the same meaning as before, although the planning time will differ. You may also consider using FFTW_PATIENT, which is like FFTW_MEASURE except that it takes more time in order to consider a wider variety of algorithms.

For multi-dimensional complex DFTs, instead of fftwnd_create_plan (or fftw2d_create_plan or fftw3d_create_plan), followed by fftwnd_one, you would use fftw_plan_dft (or fftw_plan_dft_2d or fftw_plan_dft_3d). followed by fftw_execute. If you used fftwnd to to specify strides etcetera, you would instead specify these via fftw_plan_many_dft.

The analogues to rfftw_create_plan and rfftw_one with FFTW_REAL_TO_COMPLEX or FFTW_COMPLEX_TO_REAL directions are fftw_plan_r2r_1d with kind FFTW_R2HC or FFTW_HC2R, followed by fftw_execute. The stride etcetera arguments of rfftw are now in fftw_plan_many_r2r.

Instead of rfftwnd_create_plan (or rfftw2d_create_plan or rfftw3d_create_plan) followed by rfftwnd_one_real_to_complex or rfftwnd_one_complex_to_real, you now use fftw_plan_dft_r2c (or fftw_plan_dft_r2c_2d or fftw_plan_dft_r2c_3d) or fftw_plan_dft_c2r (or fftw_plan_dft_c2r_2d or fftw_plan_dft_c2r_3d), respectively, followed by fftw_execute. As usual, the strides etcetera of rfftwnd_real_to_complex or rfftwnd_complex_to_real are no specified in the advanced planner routines, fftw_plan_many_dft_r2c or fftw_plan_many_dft_c2r.

Wisdom

In FFTW 2, you had to supply the FFTW_USE_WISDOM flag in order to use wisdom; in FFTW 3, wisdom is always used. (You could simulate the FFTW 2 wisdom-less behavior by calling fftw_forget_wisdom after every planner call.)

The FFTW 3 wisdom import/export routines are almost the same as before (although the storage format is entirely different). There is one significant difference, however. In FFTW 2, the import routines would never read past the end of the wisdom, so you could store extra data beyond the wisdom in the same file, for example. In FFTW 3, the file-import routine may read up to a few hundred bytes past the end of the wisdom, so you cannot store other data just beyond it.1

Wisdom has been enhanced by additional humility in FFTW 3: whereas FFTW 2 would re-use wisdom for a given transform size regardless of the stride etc., in FFTW 3 wisdom is only used with the strides etc. for which it was created. Unfortunately, this means FFTW 3 has to create new plans from scratch more often than FFTW 2 (in FFTW 2, planning e.g. one transform of size 1024 also created wisdom for all smaller powers of 2, but this no longer occurs).

FFTW 3 also has the new routine fftw_import_system_wisdom to import wisdom from a standard system-wide location.

Memory allocation

In FFTW 3, we recommend allocating your arrays with fftw_malloc and deallocating them with fftw_free; this is not required, but allows optimal performance when SIMD acceleration is used. (Those two functions actually existed in FFTW 2, and worked the same way, but were not documented.)

In FFTW 2, there were fftw_malloc_hook and fftw_free_hook functions that allowed the user to replace FFTW's memory-allocation routines (e.g. to implement different error-handling, since by default FFTW prints an error message and calls exit to abort the program if malloc returns NULL). These hooks are not supported in FFTW 3; those few users who require this functionality can just directly modify the memory-allocation routines in FFTW (they are defined in kernel/alloc.c).

Fortran interface

In FFTW 2, the subroutine names were obtained by replacing ‘fftw_’ with ‘fftw_f77’; in FFTW 3, you replace ‘fftw_’ with ‘dfftw_’ (or ‘sfftw_’ or ‘lfftw_’, depending upon the precision).

In FFTW 3, we have begun recommending that you always declare the type used to store plans as integer*8. (Too many people didn't notice our instruction to switch from integer to integer*8 for 64-bit machines.)

In FFTW 3, we provide a fftw3.f “header file” to include in your code (and which is officially installed on Unix systems). (In FFTW 2, we supplied a fftw_f77.i file, but it was not installed.)

Otherwise, the C-Fortran interface relationship is much the same as it was before (e.g. return values become initial parameters, and multi-dimensional arrays are in column-major order). Unlike FFTW 2, we do provide some support for wisdom import/export in Fortran (see Wisdom of Fortran?).

Threads

Like FFTW 2, only the execution routines are thread-safe. All planner routines, etcetera, should be called by only a single thread at a time (see Thread safety). Unlike FFTW 2, there is no special FFTW_THREADSAFE flag for the planner to allow a given plan to be usable by multiple threads in parallel; this is now the case by default.

The multi-threaded version of FFTW 2 required you to pass the number of threads each time you execute the transform. The number of threads is now stored in the plan, and is specified before the planner is called by fftw_plan_with_nthreads. The threads initialization routine used to be called fftw_threads_init and would return zero on success; the new routine is called fftw_init_threads and returns zero on failure. See Multi-threaded FFTW.

There is no separate threads header file in FFTW 3; all the function prototypes are in <fftw3.h>. However, you still have to link to a separate library (-lfftw3_threads -lfftw3 -lm on Unix), as well as to the threading library (e.g. POSIX threads on Unix).

Footnotes

[1] We do our own buffering because GNU libc I/O routines are horribly slow for single-character I/O, apparently for thread-safety reasons (whether you are using threads or not).

fftw-3.3.3/doc/html/Installation-and-Customization.html0000644000175000017500000001076212054411111020051 00000000000000 Installation and Customization - FFTW 3.3.3

Next: , Previous: Upgrading from FFTW version 2, Up: Top

10 Installation and Customization

This chapter describes the installation and customization of FFTW, the latest version of which may be downloaded from the FFTW home page.

In principle, FFTW should work on any system with an ANSI C compiler (gcc is fine). However, planner time is drastically reduced if FFTW can exploit a hardware cycle counter; FFTW comes with cycle-counter support for all modern general-purpose CPUs, but you may need to add a couple of lines of code if your compiler is not yet supported (see Cycle Counters). (On Unix, there will be a warning at the end of the configure output if no cycle counter is found.)

Installation of FFTW is simplest if you have a Unix or a GNU system, such as GNU/Linux, and we describe this case in the first section below, including the use of special configuration options to e.g. install different precisions or exploit optimizations for particular architectures (e.g. SIMD). Compilation on non-Unix systems is a more manual process, but we outline the procedure in the second section. It is also likely that pre-compiled binaries will be available for popular systems.

Finally, we describe how you can customize FFTW for particular needs by generating codelets for fast transforms of sizes not supported efficiently by the standard FFTW distribution.

fftw-3.3.3/doc/html/Installation-on-Unix.html0000644000175000017500000003130712054411111015774 00000000000000 Installation on Unix - FFTW 3.3.3

Next: , Previous: Installation and Customization, Up: Installation and Customization

10.1 Installation on Unix

FFTW comes with a configure program in the GNU style. Installation can be as simple as: 

     ./configure
     make
     make install

This will build the uniprocessor complex and real transform libraries along with the test programs. (We recommend that you use GNU make if it is available; on some systems it is called gmake.) The “make install” command installs the fftw and rfftw libraries in standard places, and typically requires root privileges (unless you specify a different install directory with the --prefix flag to configure). You can also type “make check” to put the FFTW test programs through their paces. If you have problems during configuration or compilation, you may want to run “make distclean” before trying again; this ensures that you don't have any stale files left over from previous compilation attempts.

The configure script chooses the gcc compiler by default, if it is available; you can select some other compiler with: 

     ./configure CC="<the name of your C compiler>"

The configure script knows good CFLAGS (C compiler flags) for a few systems. If your system is not known, the configure script will print out a warning. In this case, you should re-configure FFTW with the command 

     ./configure CFLAGS="<write your CFLAGS here>"

and then compile as usual. If you do find an optimal set of CFLAGS for your system, please let us know what they are (along with the output of config.guess) so that we can include them in future releases.

configure supports all the standard flags defined by the GNU Coding Standards; see the INSTALL file in FFTW or the GNU web page. Note especially --help to list all flags and --enable-shared to create shared, rather than static, libraries. configure also accepts a few FFTW-specific flags, particularly:

  • --enable-float: Produces a single-precision version of FFTW (float) instead of the default double-precision (double). See Precision.
  • --enable-long-double: Produces a long-double precision version of FFTW (long double) instead of the default double-precision (double). The configure script will halt with an error message if long double is the same size as double on your machine/compiler. See Precision.
  • --enable-quad-precision: Produces a quadruple-precision version of FFTW using the nonstandard __float128 type provided by gcc 4.6 or later on x86, x86-64, and Itanium architectures, instead of the default double-precision (double). The configure script will halt with an error message if the compiler is not gcc version 4.6 or later or if gcc's libquadmath library is not installed. See Precision.
  • --enable-threads: Enables compilation and installation of the FFTW threads library (see Multi-threaded FFTW), which provides a simple interface to parallel transforms for SMP systems. By default, the threads routines are not compiled.
  • --enable-openmp: Like --enable-threads, but using OpenMP compiler directives in order to induce parallelism rather than spawning its own threads directly, and installing an ‘fftw3_omp’ library rather than an ‘fftw3_threads’ library (see Multi-threaded FFTW). You can use both --enable-openmp and --enable-threads since they compile/install libraries with different names. By default, the OpenMP routines are not compiled.
  • --with-combined-threads: By default, if --enable-threads is used, the threads support is compiled into a separate library that must be linked in addition to the main FFTW library. This is so that users of the serial library do not need to link the system threads libraries. If --with-combined-threads is specified, however, then no separate threads library is created, and threads are included in the main FFTW library. This is mainly useful under Windows, where no system threads library is required and inter-library dependencies are problematic.
  • --enable-mpi: Enables compilation and installation of the FFTW MPI library (see Distributed-memory FFTW with MPI), which provides parallel transforms for distributed-memory systems with MPI. (By default, the MPI routines are not compiled.) See FFTW MPI Installation.
  • --disable-fortran: Disables inclusion of legacy-Fortran wrapper routines (see Calling FFTW from Legacy Fortran) in the standard FFTW libraries. These wrapper routines increase the library size by only a negligible amount, so they are included by default as long as the configure script finds a Fortran compiler on your system. (To specify a particular Fortran compiler foo, pass F77=foo to configure.)
  • --with-g77-wrappers: By default, when Fortran wrappers are included, the wrappers employ the linking conventions of the Fortran compiler detected by the configure script. If this compiler is GNU g77, however, then two versions of the wrappers are included: one with g77's idiosyncratic convention of appending two underscores to identifiers, and one with the more common convention of appending only a single underscore. This way, the same FFTW library will work with both g77 and other Fortran compilers, such as GNU gfortran. However, the converse is not true: if you configure with a different compiler, then the g77-compatible wrappers are not included. By specifying --with-g77-wrappers, the g77-compatible wrappers are included in addition to wrappers for whatever Fortran compiler configure finds.
  • --with-slow-timer: Disables the use of hardware cycle counters, and falls back on gettimeofday or clock. This greatly worsens performance, and should generally not be used (unless you don't have a cycle counter but still really want an optimized plan regardless of the time). See Cycle Counters.
  • --enable-sse, --enable-sse2, --enable-avx, --enable-altivec, --enable-neon: Enable the compilation of SIMD code for SSE (Pentium III+), SSE2 (Pentium IV+), AVX (Sandy Bridge, Interlagos), AltiVec (PowerPC G4+), NEON (some ARM processors). SSE, AltiVec, and NEON only work with --enable-float (above). SSE2 works in both single and double precision (and is simply SSE in single precision). The resulting code will still work on earlier CPUs lacking the SIMD extensions (SIMD is automatically disabled, although the FFTW library is still larger).
    • These options require a compiler supporting SIMD extensions, and compiler support is always a bit flaky: see the FFTW FAQ for a list of compiler versions that have problems compiling FFTW.
    • With AltiVec and gcc, you may have to use the -mabi=altivec option when compiling any code that links to FFTW, in order to properly align the stack; otherwise, FFTW could crash when it tries to use an AltiVec feature. (This is not necessary on MacOS X.)
    • With SSE/SSE2 and gcc, you should use a version of gcc that properly aligns the stack when compiling any code that links to FFTW. By default, gcc 2.95 and later versions align the stack as needed, but you should not compile FFTW with the -Os option or the -mpreferred-stack-boundary option with an argument less than 4.
    • Because of the large variety of ARM processors and ABIs, FFTW does not attempt to guess the correct gcc flags for generating NEON code. In general, you will have to provide them on the command line. This command line is known to have worked at least once: 
                     ./configure --with-slow-timer --host=arm-linux-gnueabi \
                 --enable-single --enable-neon \
                 "CC=arm-linux-gnueabi-gcc -march=armv7-a -mfloat-abi=softfp"

To force configure to use a particular C compiler foo (instead of the default, usually gcc), pass CC=foo to the configure script; you may also need to set the flags via the variable CFLAGS as described above. fftw-3.3.3/doc/html/Installation-on-non_002dUnix-systems.html0000644000175000017500000001222512054411111020737 00000000000000 Installation on non-Unix systems - FFTW 3.3.3

Next: , Previous: Installation on Unix, Up: Installation and Customization

10.2 Installation on non-Unix systems

It should be relatively straightforward to compile FFTW even on non-Unix systems lacking the niceties of a configure script. Basically, you need to edit the config.h header (copy it from config.h.in) to #define the various options and compiler characteristics, and then compile all the ‘.c’ files in the relevant directories.

The config.h header contains about 100 options to set, each one initially an #undef, each documented with a comment, and most of them fairly obvious. For most of the options, you should simply #define them to 1 if they are applicable, although a few options require a particular value (e.g. SIZEOF_LONG_LONG should be defined to the size of the long long type, in bytes, or zero if it is not supported). We will likely post some sample config.h files for various operating systems and compilers for you to use (at least as a starting point). Please let us know if you have to hand-create a configuration file (and/or a pre-compiled binary) that you want to share.

To create the FFTW library, you will then need to compile all of the ‘.c’ files in the kernel, dft, dft/scalar, dft/scalar/codelets, rdft, rdft/scalar, rdft/scalar/r2cf, rdft/scalar/r2cb, rdft/scalar/r2r, reodft, and api directories. If you are compiling with SIMD support (e.g. you defined HAVE_SSE2 in config.h), then you also need to compile the .c files in the simd-support, {dft,rdft}/simd, {dft,rdft}/simd/* directories.

Once these files are all compiled, link them into a library, or a shared library, or directly into your program.

To compile the FFTW test program, additionally compile the code in the libbench2/ directory, and link it into a library. Then compile the code in the tests/ directory and link it to the libbench2 and FFTW libraries. To compile the fftw-wisdom (command-line) tool (see Wisdom Utilities), compile tools/fftw-wisdom.c and link it to the libbench2 and FFTW libraries fftw-3.3.3/doc/html/Cycle-Counters.html0000644000175000017500000001077612054411111014646 00000000000000 Cycle Counters - FFTW 3.3.3

Next: , Previous: Installation on non-Unix systems, Up: Installation and Customization

10.3 Cycle Counters

FFTW's planner actually executes and times different possible FFT algorithms in order to pick the fastest plan for a given n. In order to do this in as short a time as possible, however, the timer must have a very high resolution, and to accomplish this we employ the hardware cycle counters that are available on most CPUs. Currently, FFTW supports the cycle counters on x86, PowerPC/POWER, Alpha, UltraSPARC (SPARC v9), IA64, PA-RISC, and MIPS processors.

Access to the cycle counters, unfortunately, is a compiler and/or operating-system dependent task, often requiring inline assembly language, and it may be that your compiler is not supported. If you are not supported, FFTW will by default fall back on its estimator (effectively using FFTW_ESTIMATE for all plans). You can add support by editing the file kernel/cycle.h; normally, this will involve adapting one of the examples already present in order to use the inline-assembler syntax for your C compiler, and will only require a couple of lines of code. Anyone adding support for a new system to cycle.h is encouraged to email us at fftw@fftw.org.

If a cycle counter is not available on your system (e.g. some embedded processor), and you don't want to use estimated plans, as a last resort you can use the --with-slow-timer option to configure (on Unix) or #define WITH_SLOW_TIMER in config.h (elsewhere). This will use the much lower-resolution gettimeofday function, or even clock if the former is unavailable, and planning will be extremely slow. fftw-3.3.3/doc/html/Generating-your-own-code.html0000644000175000017500000001115412054411111016566 00000000000000 Generating your own code - FFTW 3.3.3

Previous: Cycle Counters, Up: Installation and Customization

10.4 Generating your own code

The directory genfft contains the programs that were used to generate FFTW's “codelets,” which are hard-coded transforms of small sizes. We do not expect casual users to employ the generator, which is a rather sophisticated program that generates directed acyclic graphs of FFT algorithms and performs algebraic simplifications on them. It was written in Objective Caml, a dialect of ML, which is available at http://caml.inria.fr/ocaml/index.en.html.

If you have Objective Caml installed (along with recent versions of GNU autoconf, automake, and libtool), then you can change the set of codelets that are generated or play with the generation options. The set of generated codelets is specified by the {dft,rdft}/{codelets,simd}/*/Makefile.am files. For example, you can add efficient REDFT codelets of small sizes by modifying rdft/codelets/r2r/Makefile.am. After you modify any Makefile.am files, you can type sh bootstrap.sh in the top-level directory followed by make to re-generate the files.

We do not provide more details about the code-generation process, since we do not expect that most users will need to generate their own code. However, feel free to contact us at fftw@fftw.org if you are interested in the subject.

You might find it interesting to learn Caml and/or some modern programming techniques that we used in the generator (including monadic programming), especially if you heard the rumor that Java and object-oriented programming are the latest advancement in the field. The internal operation of the codelet generator is described in the paper, “A Fast Fourier Transform Compiler,” by M. Frigo, which is available from the FFTW home page and also appeared in the Proceedings of the 1999 ACM SIGPLAN Conference on Programming Language Design and Implementation (PLDI). fftw-3.3.3/doc/html/Acknowledgments.html0000644000175000017500000001434412054411111015127 00000000000000 Acknowledgments - FFTW 3.3.3

Next: , Previous: Installation and Customization, Up: Top

11 Acknowledgments

Matteo Frigo was supported in part by the Special Research Program SFB F011 “AURORA” of the Austrian Science Fund FWF and by MIT Lincoln Laboratory. For previous versions of FFTW, he was supported in part by the Defense Advanced Research Projects Agency (DARPA), under Grants N00014-94-1-0985 and F30602-97-1-0270, and by a Digital Equipment Corporation Fellowship.

Steven G. Johnson was supported in part by a Dept. of Defense NDSEG Fellowship, an MIT Karl Taylor Compton Fellowship, and by the Materials Research Science and Engineering Center program of the National Science Foundation under award DMR-9400334.

Code for the Cell Broadband Engine was graciously donated to the FFTW project by the IBM Austin Research Lab and included in fftw-3.2. (This code was removed in fftw-3.3.)

Code for the MIPS paired-single SIMD support was graciously donated to the FFTW project by CodeSourcery, Inc.

We are grateful to Sun Microsystems Inc. for its donation of a cluster of 9 8-processor Ultra HPC 5000 SMPs (24 Gflops peak). These machines served as the primary platform for the development of early versions of FFTW.

We thank Intel Corporation for donating a four-processor Pentium Pro machine. We thank the GNU/Linux community for giving us a decent OS to run on that machine.

We are thankful to the AMD corporation for donating an AMD Athlon XP 1700+ computer to the FFTW project.

We thank the Compaq/HP testdrive program and VA Software Corporation (SourceForge.net) for providing remote access to machines that were used to test FFTW.

The genfft suite of code generators was written using Objective Caml, a dialect of ML. Objective Caml is a small and elegant language developed by Xavier Leroy. The implementation is available from http://caml.inria.fr/. In previous releases of FFTW, genfft was written in Caml Light, by the same authors. An even earlier implementation of genfft was written in Scheme, but Caml is definitely better for this kind of application.

FFTW uses many tools from the GNU project, including automake, texinfo, and libtool.

Prof. Charles E. Leiserson of MIT provided continuous support and encouragement. This program would not exist without him. Charles also proposed the name “codelets” for the basic FFT blocks.

Prof. John D. Joannopoulos of MIT demonstrated continuing tolerance of Steven's “extra-curricular” computer-science activities, as well as remarkable creativity in working them into his grant proposals. Steven's physics degree would not exist without him.

Franz Franchetti wrote SIMD extensions to FFTW 2, which eventually led to the SIMD support in FFTW 3.

Stefan Kral wrote most of the K7 code generator distributed with FFTW 3.0.x and 3.1.x.

Andrew Sterian contributed the Windows timing code in FFTW 2.

Didier Miras reported a bug in the test procedure used in FFTW 1.2. We now use a completely different test algorithm by Funda Ergun that does not require a separate FFT program to compare against.

Wolfgang Reimer contributed the Pentium cycle counter and a few fixes that help portability.

Ming-Chang Liu uncovered a well-hidden bug in the complex transforms of FFTW 2.0 and supplied a patch to correct it.

The FFTW FAQ was written in bfnn (Bizarre Format With No Name) and formatted using the tools developed by Ian Jackson for the Linux FAQ.

We are especially thankful to all of our users for their continuing support, feedback, and interest during our development of FFTW. fftw-3.3.3/doc/html/License-and-Copyright.html0000644000175000017500000001017312054411111016066 00000000000000 License and Copyright - FFTW 3.3.3

Next: , Previous: Acknowledgments, Up: Top

12 License and Copyright

FFTW is Copyright © 2003, 2007-11 Matteo Frigo, Copyright © 2003, 2007-11 Massachusetts Institute of Technology.

FFTW is free software; you can redistribute it and/or modify it under the terms of the GNU General Public License as published by the Free Software Foundation; either version 2 of the License, or (at your option) any later version.

This program is distributed in the hope that it will be useful, but WITHOUT ANY WARRANTY; without even the implied warranty of MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the GNU General Public License for more details.

You should have received a copy of the GNU General Public License along with this program; if not, write to the Free Software Foundation, Inc., 51 Franklin Street, Fifth Floor, Boston, MA 02110-1301 USA You can also find the GPL on the GNU web site.

In addition, we kindly ask you to acknowledge FFTW and its authors in any program or publication in which you use FFTW. (You are not required to do so; it is up to your common sense to decide whether you want to comply with this request or not.) For general publications, we suggest referencing: Matteo Frigo and Steven G. Johnson, “The design and implementation of FFTW3,” Proc. IEEE 93 (2), 216–231 (2005).

Non-free versions of FFTW are available under terms different from those of the General Public License. (e.g. they do not require you to accompany any object code using FFTW with the corresponding source code.) For these alternative terms you must purchase a license from MIT's Technology Licensing Office. Users interested in such a license should contact us (fftw@fftw.org) for more information. fftw-3.3.3/doc/html/Concept-Index.html0000644000175000017500000016531412054411111014446 00000000000000 Concept Index - FFTW 3.3.3

Next: , Previous: License and Copyright, Up: Top

13 Concept Index

fftw-3.3.3/doc/html/Library-Index.html0000644000175000017500000022254612054411111014460 00000000000000 Library Index - FFTW 3.3.3

Previous: Concept Index, Up: Top

14 Library Index

fftw-3.3.3/doc/html/equation-dft.png0000644000175000017500000000210212054411111014207 00000000000000‰PNG  IHDR«:Jû„Z0PLTE³³³¨¨¨œœœ„„„xxxlll```TTTHHH<<<000$$$ êYËžtRNS@æØfÀIDATxœíW=hA~{û“¹ìýD°Ø4!E¢k¥Ø¸ÂŠ…-B@ÝΠÁDçüÁ5¨I;OEˆB xÍi%hq  )D,”Q²¾ÙÙ¹ÛÜíy{fƒ ~Çμyóí73oçïb@òSº]w2ŠE¡OÔè^ˆåÄÑŠFd¸”1l æl©áwC¤énU/G»KAEôZÆfTÊ€m¹þÐ %ÄÎAš½Õn{å4£Ýmà*n]ªì[Kll 6: ©YBUôhT@g„q$üq\€~|ù¢Ö¹µIÏs[m1O#Ý%°»ù-HñhšK¡ŽèëLéÏk a¾†ÉT£lŽVúK+ü3P°æª SCTîõSõ0ðÏã öÀ=_˜s1 µ§8+$K¿²„”° Ú¦;ùk‘ ¬hj9FgR“Hl¦.Ä”d} åxÜK¿Û¡?ˆTܘóW=ŸK²fl®±Ml(4¶†`do`˧FQ:ˆ ò—,¡M ÷µ7«?!…a¾­SoG bjé|­ú›'qì‘t†Ê9hÔlv°üðS­Y¸ {¤¬åy+{³lÂ9 d Éz¥‚}ÏûýO›`äÏø,ùVÕ¥‘ºªß Óñµ£È€n‚R(àÀ²¬,CÌZí¾ŸJã®ð¸í§›Å3Í»-U`ŸÌèéHõ·Ãåδ ÆI‹É½²ß™Öcr1¨z\®Ï©xÔ\5¦æ$òÙ•b4ŽÚЋ9´›?Rmk¯‰ŽvÏ9gRlmR[ýØó6škãí^Á+ˆiÒÙ‹ôW9ÃCâÅ`ب"Úâbˆ"O¸~ó‚ˆ@µ:!ç)c?Ÿ\¦Hm7"¤ÉE&VáÜZ'Ìi°1m0ú€p^³0J¬à¡Œ¬¨ §Å‘×>/òﱧ…Ô,´H-s æˆõyjJ–PQ1;í¤:¢¿b#J ¡X# ™ ýEýDD§¸W·™öº-ü¿!O™ú×›ØãÓb†¥íé0£(u›h@\îøµ~3ËhvkAÑcVܲQ|ÙA¥„4¢I­ëx“^Ó·•…t׬~©†@ˆy½Ã‹ôZ丞.Ń.Ì,–Õn%Z¸ýüBj±[6+¢™˜xSË‘½T²ÂL†Å2ÆÜ w9š,µ÷â¥þ4ûÜe¾¦r༺ÿÀ†„—é²n“û¼¸N”ˆMÌï•MwfÀ3ôþ`ý7¢õl{Þf\§$‘ðPÉ”QTíåè)è6ÔÖV2†ß…>V`†4?óåôjX7tî‚Kùl´$™Q±‘Ïõà±C(¤˜©ŠŒ- Á‡­ íD°ï·‘ùÑ6ä^Ž î›ä#ÈŠI®Éú‰Ÿ(kvdÆ Ýî¿!ÆÄFåÛ~ËC÷6®øÔDW³(ÄQª„+uYh@_¥åõ›éXšïÒ'› 4kõi0ãÒõjÝVh~ÜÂ-Ô vj¦ÐöòCw6(ý·BS‚Ú—rY>ŸwL:׶Ù/°ŸY&·0ª.Ï4¼¡ 3‘z¬›ÛpÛǺkc^®¿D>´sÈ]9(æúè¹µw~¥KÊG8î'Ãküjœ6—å˜+íØ7¬¢ÿ& ZÁ¯ožzLlËP:?0¢öYx8;ŠL2«œÁA«V_ä¿*A2¢Žá„Oð˜“ü³VUüb:øÝ™TyCø±U øĪ# e~FH&vöy{Eë7Ò‘§œ'Å?ÂAseµ:µÛŶӄîJï¯^žCÊ&sOë™]ìÜ µLròRìjµ_¨$Šy"wì$dT’ä¸ãàyCÃ^-òå<Ö[`Œg/¾n 7é…Šäµi®^a¶ Úìl¢F4X~{ož>®Ü7‡tºa8â-Ž!›d¶PÁ²ÃT±ê/ì]%2¬LÿNMâ «$nê/5.`½ ©ËdL$Ê0^1ï.Ò3sXö‰ìD²°7™þ/Þ¢‡éfØ3«¤O‰Ì"#B)gЩ;B^V¤Þñ>e ‹mDFëé¾=4Æ1Ò騤}PÄ®«2RÏ®5Ûž;ÁN®°›M¹Z¤¶IçÐìtohä*Þ|³1L¦ûåÃm´ïÚ{Þ–/¡¹Z-Çût.TäËì”#â-¾Û–*Â3¹’磸`׸C¤µcX>ºï¾ °&êÊáàia4¯Î% -;$3/¸Á3áFo.|Mžu>ÅjM)‹ÃÂOá³éúVGOÕT2a¯evú:À½gü‰µýüßÿfõ&¢5‚3ÓLY-3Kc$4®»3ÍØüž,"Fçzýç»uC­hb1Ö à%rPì ån b¯Í4ƒFÀ?õoŽË·Z…MIEND®B`‚fftw-3.3.3/doc/html/equation-idft.png0000644000175000017500000000206712054411111014372 00000000000000‰PNG  IHDR¡:]Ù“0PLTE³³³¨¨¨œœœ„„„xxxlll```TTTHHH<<<000$$$ êYËžtRNS@æØfµIDATxœíWMhÔ@~ÙMÒÙÍþ!A£‡RÐbzÒ›öPAhAO‚š[»‚QÁ©m1¶Úî­‚WD¨Y a/‹'AíADoÁ?*J×7™Ìv›MÛ©ÝŠ‚dæÍË7_^æçe Å/K Â*Õ¦wŒzË•ÑŠ@â:3*mx½Æø¦ôF"½e.M¡*üªÀ§xø>H7 ÖAµME× Úê²¢œ‘ðT6N$§a™2ÙܨlPöejûIŽƒnB F/ä 6ïzòlä¡rÃ\Ñf<ü zÝ »ÖÀLTìep$»ÿ’îQX5Ëë ·1e=d눰o‹bЈk³»Ôœ ’š¸Ž.j ê¿ÔN0bö-&³nÖ†AKé†æ)¶ ~a)£<ŽäV`¤„ÇÓ¿2ŒÌBѶ8M`‹sNJŽ=Vd˜7·ПDL–(³6µc’D’¶äˆñ–¾MÂ=ái«\ßv~Nœv’ä›Ù mé~\<ºhäæ«åQ\ïlI é1HƒX2ORÝ~²Óà>ÄRæ¢E\J ˜ÖYÎøé—þ¶±û½ pLÚ[ÂÆD: @ŠÈ4ªÕ¤ 0bÂ3Ãzgj|•[J7³BQóŸmaYh}ý¨ù<¾Išµã0OJvä i40 ·Æ"²y¥[ÂáDóºñ:ÈŒŽõyÀØ· 8Ûå” q´Ï7dˆ8x†ÑÁë¼­ÿØmJ¼½ìÂß=—†¡Î§⺮Ó6Åtñ°o*m“ä¿€ºÀ—–[GyÝÝÔ7y{^WÖVìYæP™2CwÈbJºöؾ#hâÆÊܼ°LáâPÇ¡¥À`ü»ËΙŸ¤>®ÄÏðð4\)cnƒ†bô³ö=´§r0RYøÜÔzÚôŸ2s&àQþ«[LÒ¤Hç¨8‹eü5:RµË?"ù<'R§+Š™w¼ÖæüˆöÆrñ¤&þ Ì*§‰óî¹!îdËáôœ¬X4N;¼N8Âóáù–w¿a³ž¿½ð}K2MÈ ØíRÜtñJ¤…¶A§†Ó&©áéEƒRÿ-P=m½Ô®Yà†µ.O"ü-ø+½ÒjË|fˆIEND®B`‚fftw-3.3.3/doc/html/equation-redft00.png0000644000175000017500000000341012054411111014701 00000000000000‰PNG  IHDR\:yåÌz0PLTE³³³¨¨¨œœœ„„„xxxlll```TTTHHH<<<000$$$ êYËžtRNS@æØf†IDATxœíYmˆTU~æÞ;wÎÎÙ{!!?²nСÞ(hA›î…Å\HK0jú#e¢CIý9¶~\Ûµ]úi’Jû“CYX5”Š`ņ˜A?œ>)ÉÑT¤Âé=ç~ìÙ5gu†qcŸ…¹÷žóžç}Î{ÏyÏ9wV á ,î‹ôB[jˆÂ"£{ê>¼¡e-7)a_Kd5ÇLºq DVD öj(r ƒ´ñ1lº˜2b}òQ›œŽ·Sk66ë\jtb¡ÉP*vÁb$™1Yd (]H,îqûîÔ²øø*ÌÑtqMMEصŠèúÐoÞÖÑs-õ¹6që$Õz_4x%¸œ~âýõ ”r2I‰Â¬êB·Q‚ª3B„˜!IóRIåíÁ¥XsçЋl䈛\ÿ î³])nsŒ²\¹‹ 3nàL… B²å·žÅàŠŽ÷J=5kW˜N× Öç5o¬Ãh“ŒN ᶞs…È$¼õ¼ŽÓzÎë ~—L̶ÚYijYUůB$nBåvøtÉ/ªþµz‰ceÔorÖC³Ò,G¶ W³fÍ–¬f9¶Xš4«áëUMäŠðppæm#^áÿÅ@Øç¿°Ç[™+ØHMFÕYëÀe¨V°Êã"ˆü`n‹ô¯h[ðŸü »ÛïDŸÓÁlÅÑ Ñrj™ô:»‘öS€4ŠòN³Pȃ"ã4…üÛIp` ‰¦yØíY O·sõЬ«Û\ x¹å”z•#šç­Ä ¥€”Ýz^´)¶74 ‹öÑ0¹V>EþÚÁÂTS,yñÔ+7Œr—n9Eû-wÅÈ"L÷¢$S4ð¥ñø±©XÎ7ØŠä¦Ûtj(D[¯£ ¶f”õ’Š_ZAù΋ …!&Ù¥¥p3°xÜV³ga©àpP¼]Q}K––§£TËð,É› D‹î¾Œ &r¯Jz ¹‘­ŽÊ9ç”ÎÒ¯º;qjwÂÊ+cÅ€—Ž‡ºm†=*€çØvÏarU-Š¥A *9¼×ÅÊ‚l‰Û@õ3†•ÓTÇ5'ÏË/3á@ëíµÀ P45)Yê%…N6Ú‘ÑÓFá‰K?nFõ&çT’)‹ˆhÈî‚›aV W±Nò(†£9I:àÇ0z,âÈŒÇ[£ƒÍ h'`ß…¤MÆ#x#԰ײ¯þCÃN¯"Ü ÓB+S”•ãËŸò¶¼[H¶EoZ›(>¯ÕÚvT™Nƒ¤ÜÞ¹Ìã¼ ¼b<´ƒWLåmà}¡ œ3˜ÁŒÄæ¡é°«Ûµñ¼&œì@‡¥4ƒe8L» o+’£v§µ4ú9´ZÍî°”´ x3Vf-„Û^9WÃmMùWNþìßÌ>×·ë.s¬ÿ»¾Œü5yæPCY.¼ë¡Sðü`Øq9÷è?“g¨ÄÐqÑkÜN]7ú¤¥î†ñ»l2sVï´±°©±2]aÑŽ_óý½UÑN²€ ž8z½Á Bý÷¹iÐùµØÙ!ÎÛ(W(–-²VÌY½S’›)?úé%Ü/oÿp¼kåxK„-`Ý¿L ¨HŸ•#ñƒuÈ›<þˆ³º÷ËÄþe®‡ä™Z­¢ çÐUð¾>ÎæÞ²§A.9%¹Ñ¥>‡ê#ärÞÊ·èýyÛxAÆ‘¤|/B: Éq-ìÆŸø¥ÊÉŸêy± Ø`ƒa®üíiËÃ@šb~O–cù‡T›*g_À”\n åzB.£¸-_ ¾#Ðq»TúHD&¯_öâÓñ`¥x]GƒoÏ÷bÅäÄp0|½Xtéa…_ú@.léóÎ~›äZè*Ã\cÍ-P£-}ð^Wq ¥³Ô˜úg|RøNû©Ã´ÿêƒxlŸü ìÌfÚޗƽ+gÃÚ¾o¶"!?èoðÒ%WËäE!W'wa­V–·D_B^P´µãk­ud‡W;I3¨É÷Pγ#öKg&O½æÓ~ÿ}U½ñàEȬñÙòwQ¼j^|Âdóëü¹ÊJ<Ö û±‘W/Ô=ú}œÿlð¯}´ÑYÌiÿdå“#vîQxÓ­’ß7¯„%ÝÃvÅuIXm58‹9eG›ö{mH|-²Moß®Ú$ÿ–¿ñ s£D§•ü/¡ää%awPÃÔA‰:€ÍÉßA×&r±OÑRYœ&rw;­`Jxx:|]hþëî—&+ #IEND®B`‚fftw-3.3.3/doc/html/equation-redft01.png0000644000175000017500000000277612054411111014720 00000000000000‰PNG  IHDR :Íï:u0PLTE³³³¨¨¨œœœ„„„xxxlll```TTTHHH<<<000$$$ êYËžtRNS@æØf|IDATxœíY]ŒUþæ§ím§Û•Q'â/„1ë!1¤˜•P£¨cLHD"á˜h®²àà‚»oþ$&Õ·‰ ®öFeC‚5‘@„¨‘‚AÔsïtú³Ý¥À,‘o·3gÎœûÝïž{çÎÙ.pMP€Ž«’iˆk~W¨:À[â’£ÍÈFOÛU×Öe0Ø\6 -ìÑðCï& æa˜Ù£‡ùó)ÃéabÍì~†r†Åh($VSSE[[9—R<¤¦pÀíi’”Ço^zè ¹<’áì]°>÷wõèølîš±(ŽI‡1aqx9¨¨¤–‰e‘´Q†¦Ë°:0;¦GGù—É“ÊÁkÈc(lÉ¡[ÓÆ‹ûÁ)Â;¤›=hâ—2Ö ÏÞLaõØ£Õ¯`ÙÆÏéõ€¯yâÄ™¦Åk?oø›>©NG¸dÔvŽ†äÒª‚V¦ÆíÈ{Qi)rr‡‹ÉätX€Ê ‘MµL%ÌSoÙIضK4ŠRh&¼þ$—A¦ô:uÑÄ€?Ít¹#mpÀÆ|Áእ¯£^æÛ-}Ä-ïJä#Mï2UÁ0 ŽÓœ¯}¨Q7î\ù ›ø3“òïCè­ÄäsÈãQäâj¸aÍ×|‹ÏÁ†ãsˆ:ìféAyc"‡~âuÛqyä]û=ƒjCQßQÝfÐu04ëZg|N‹¶¤]Ti5 ˆmÑl½‰KbÛtaV¤o­ }XquGpØ&m†E5 ê£T—ªq°"]årAìœ $‡Ü+/4µQ±b.Êqïݬ §?ir(¨D¹M­Õ0ÍbŒê íð*ˆ€ç¥ïɘF§ 3ü øïo%Ž ý0”t§ ÞÏ[à¨v&)QeT%-¥h*Ã9^à‹Ld1f7u¤±…gÉÃ,‘Ïo%z“c!d¢‡Vmœû¨‚¢Z+#ßm®ð? ‚4Kƒäè(dÛ3÷bd,oÏËb§cºE΋YÈy™*5Qåëò±9"úÃl¼¾aÿ Ð’c<…RòýØûX ñ÷Âe®‰¿B0;¦†[øŸb'`çm9qŸåXX)ø¾¿}°*Ž%‡ïš±Éª¸…6ÎRÅ’Ka‰§ŒÛóK¼ñÇ+uûØÉS‰3‡º\ÊØnqÚt±y]œ=¬ã^nc;Œ_ºnjâfrB¥BSºG¨ÐµðæLi™k}©KÚ‡ÜÌ9!æäôa=*èvÚ¹N’ð5;dI²g¤©ÈV6Gýå +¤ù›Ûâ ¨x>+yí^Úa‰£Ï:ëòß*ûW»Œ'Î45¬šÂ£@ºäÿ0SÅ,ë sáÔ¶ÝB…NI딧Vw¯ýX|á»cÒß*|_…·–\²»Ãð4M”-:0×þüªÅx!CÒÜã rxÍ>ÙyD"T¤J!%Ì.àC¡‚U(x±(ž©¶/—¿/&öàëŽaøx"LÍ©–ˆ0ìÎOI……tæz뮑o3h˜¬Š-ÀÞtP˜‘‹¥M2VfªªMFh<´ ~ÿyg&^úìÌ(&Qn,m4*⬙†Ÿ(æèkã“ß±©u㩪AM†8–ZêyvØi“è;O{Øß$»Â4yêgDò©u¿s¶›È®K6»ÓùnÏŸøÉR×e¸:•c¯S®:J¾Ió±€²¯eb¤·óËâ9/<§½È“ÛÙ¥ì„<=(¼³UŽ¹æ0·Mv¤7ìr¸çGñåwœüãêÚÍ/òÜAª¸A záùöªˆ0¯ÿ¼ºZ°…Åû媨 ¬ÁË´¿ƒUá|6ÈÞ#<5_ŽÞ°øç`òøb>ÇIEND®B`‚fftw-3.3.3/doc/html/equation-redft10.png0000644000175000017500000000265712054411111014716 00000000000000‰PNG  IHDRã:Ç–Ã0PLTE³³³¨¨¨œœœ„„„xxxlll```TTTHHH<<<000$$$ êYËžtRNS@æØf-IDATxœíXMˆEþúgzj¦w§[ôøñà%¤e=ˆ"ttñv£b;’Á ’C(e²IÚltöè€?1¦‰!DçA“%Ñ $dÅÃŽrPp;n"þf|Õ?³3“dp;Ù A¿eºª_õ{ß«Wo«^7°@HPTX€â2½ ê À¬!Y¬1‚n•/”äÊ`Iç”ñ‘ªß–‰ŒéBÚê¦ö®’Íî7h`ôòr†Z‰Á§ÉÙ²¥<Ì}•âVÀ‹8Ôõ ›!Ôi6î¿tJ‹¬ÏØë}Gû£ ~YyÓ¤‹-B0!né! hD÷¡Yz´¦“¸yn|œçÄ3&¸3mRx·ªEÒàPH6·¡R-Lòo|äüÐP@."âý6ü¸=Ò“. ê}Üw\Omùš@N£”yýSàŠP”4ZW‰¿é—íˆ<êvŸÇùTƒÂ­H¤ðJ´î§Ë-X)­ZÈ:'´àãSƒeÑ:’©Š§Œ°oâN‡7ÇTÁÀ{s;IòôÇ´÷hC š"Œ]ó¾Ô“ñÃd U%âî¡_žÂØ$ã=¹C"|ÞÉ..hªÒS©4)óôTÉ_¨§â»1a¦Öäi”6±´šl$­&• 7 Ò+“i4Õstq;%tmÀ q•†“ÓGTOG¢nÐÞ=9ÄéÔ O¯@y„ûTä!6 #ÙÚ F7õ°»F˜}HrT›4eSŽLèÐh—33 tYl5¬Lªz­&ÎCàBd„|.Ë Üåû6gaù±æ0dž-ì¶n’¾C'°“Vé÷lØÛjû†Þä9Ÿv0ï0'M*ÂÀ蟸ªÚeCÐm¨Å"²^ˆ=ŗʆ4Rà%¾Œj¢\â­‹i_e09ÄL‡„ÞŽ]>ùGQEì;騷P2âP…[c ²b*5£½4ÄsIçþŽØnÛžHÝH2Ô­$ ‹P†Ä£XŒÁ] ÒõC‰b¿Ð)†ªžÇò©8›”éJ£*-–›F‘Y©øþǃ=|PLïn9uØÀ(Wc žçí åíx´ïNå­çF¹übTzHë[Þx‘ÞªÿìI{E³é·ø^¼“³™³™Ùc=ºI.3—èNÇî—1,ù¼Œ-t>ý$nµ¨.S©F©˜xµÛÒƒNþÁ<3×!Ì7Ù%ßnÊÚ’UQ~z8©œÂ~+S…"¬aØlSÕ7ÇÏh@Aþ\h±Å;ìÿìDmæôÓöÚâ—ÒÁÕã™ÙV«©Ö‰kª—Rw©(Üoâ(rUï$pK§vì )UŠ_¾ÜíþgQ³üâ|´´Šà’âmù0Öüø’Éx)Os§É*Î4zÞ;žgÇ”[í¤ø.6ú)u³U¬»€·CJF>ŽÞ-ê´š@I¨â±(gÛŒJYûXò>QšÈÕaÏŽÛ8(^ô³oÙjTxË×¹j Þ¸äL«vÝ&éó€tæå¨7_âÅímQÒiÉ{„êãÊxÆÚœ›H {¬nÊïâζ{»ßÇ>Ä^÷÷ô¡¼ýT˜Á¯ÍüÞÇ­k‹bg`”ײ6R1ý{}J¬ïù¨¼ˆ`KËw‰•<ŒÚ (1Šh7òÇ7Kûø32(F<1ð¬Y<ü×DÏš fIEND®B`‚fftw-3.3.3/doc/html/equation-redft11.png0000644000175000017500000000305312054411111014706 00000000000000‰PNG  IHDR:êA»0PLTE³³³¨¨¨œœœ„„„xxxlll```TTTHHH<<<000$$$ êYËžtRNS@æØf©IDATxœíY[hUþæ²³';›Ì}h©—EòÖ‘ø ú2¥ FÚšõ¡*‚²}Œ–º`jA<¶i;ØJò$Þ [Qkì`k)4h UÙB±º*ˆ Ø±±E¼­ÿ93³™Mºiì®ÙPú…™sùÏùþÿüç2ÿžW ¦ÐJÌ4ÁLè€~Ôò² Ç ­pV__±Úy /ɼ/ž,ðŽÎ€¯§H/˜Ei}¶]~g qçÑÆÂsŠƃ>dÈÏjßTM[Åò³BCôdšÅa/­ÎˆI (ó(E ÞLâ¹`Ä4 vÙ/ΣÍ|`à™9¤5[¼Mzv‰ ‡XÞUò«XP”ŠUc8§„–ΉwŒðÑÒÆN£&Sœ¡g Ô» ÞîöѲdãG%¥@¤Æ´›966ÛÂ9«~±!7 ûïÍ£•Ù]ÿÔö)2÷b>‹í²Ð´ö9:Š!Í;Vìÿær¨Mt䊡q'® é°þ+êÍé¥;K°Jp„â\ñ$_ŽXªm)©²BŠ‡Š¡âE<9z2òhuÅ;4~ÌtaAÿåmÔ/GŒá 7U×Ö…µsÚʽI&¤£[S¦*¨h‘=ÀíR[ÑM—3¨‘b¯Ù?ýô¬™ìå<¹0 ×)w€‡fÔìÀ†í›hŸãª‡Ý!Þ>ÅS¬3[}[Aj&gj3KÏQ¬¡µâŸl!Ïïa,›Ü¾OX£“Ê\¹yÐfä9õìw*cT8â!Êœ—lÝw3ãÅyŒ\¤ygHêû¯ù¾kÛóøâ~è ZÉ©w¶‚õ𪰆ÑȆú«¿GÆÇ?”3‰q_)1F+iA½IÖü ¼\\¯¼KDÛMON=«ÎðÆ<;ÅÒÙß[ÈYY\*e«ÚTq5™ŽÚÏ?H{F­×£û´ëáüð`çF᫵ì5Žmðv?vñlÕäPɾ^ž²fðKZ·^Ä£„Xa«¿±ãî¥å¸ 8$®î²¯¸«#õ¡Ï1ìÛ™5F¥¹AÜënåô³ñ@²á8]îB›þ2[®Lv”\Å-‘ù*μ”Ôôívš­ù&Îl½£Y@ˆG"~ˆ^2ÆMFÚO´ñõ ;ÑÚå ±Å ƒ{êc­-ä~—è¶%‹jI&J¡S·rÀãtRvlYùVZ4Þùª®cOÒÙ / ߸oÇ™®šãÁÅ´x»‚a}ÙÆQ~IEND®B`‚fftw-3.3.3/doc/html/equation-rodft00.png0000644000175000017500000000312012054411111014711 00000000000000‰PNG  IHDR :•©l0PLTE³³³¨¨¨œœœ„„„xxxlll```TTTHHH<<<000$$$ êYËžtRNS@æØfÎIDATxœíY]ˆUþæ'“»™ìÎ<¶ÔŸ,ø°oÙ}(öeJƒ®¬ºñA$},J èC„ÛmwwÚ]ÙÅk-…ˆ?õç¡¡UY\Ð m)TK„b+>tU¡àF«El5ž{g&™$uÛqt?ÈÜ;÷žóÝsîï™àF¡À´C¦ßAºp@¿GÔ°X™ÞL_Ъ¶éêü†›íõŒê§oë ø ã”u)Ç]YZk·Ë‹×§ ·­XÍ°X@ÐÈ™­{xòic‰ü@ÅÌë-Í¢;…y‘þ‰Ý†¤YE°ë‹¼Ø‰PG0ðÜJÕK¶xŠ¹;ÐÑ aŸœË%<#’ÊŒ !«ç1åÙ#›L¸ô.F£T`»(á{€ç‘¥þ¯@Ÿ¤©•ö…®¡»Ó¤ ~´„€h‹Y™dˆ”†clÕ.ß´ó£÷^'bfLÍýPcàHÜ‹ŽæÞu¡i1Ä‹qÍ«XÌ+¾Ó™L§žìoÔ€[ÐxXW¥Y|Uæ,!±UU5!J"þ´ä¨¢jKf6ò–äuE  T×'JÑ/!7\GñvÉ4ÉÚƒ$8 6ã "ÎiÇ O”¸îÂ3ÅD¦=›µqžè7ùŒ%¥`¤‘º™N‰ùÐ"·ÌŽ†®•ûà¦YxÊ颿¼1Lå Õýc†)*ÅëÕ`(Š^Ž¦;Af‡‡“ìx·Î#¾[Dx»Þ²¡Æ7Žó$µ‰[xËN¿D¿OEæAÜÌ!–™Ý{È‘{7Ù=JÑóX*"1j,–ˆ»H‹<ËÄ@²†5´c–wÛ‡vžð3ŒwÕŽ:îÇIxž7Õ33Nù¡íÌB—-YC‡°j!Ün›âC½ðŸY÷óª¶ûÔŸ;®4—­ûH„¾Fû¯"Üf¾O,ŸöËg¸üXžÂ±LDŸ¢Æ”LÁ¼ØÄ«q$…Ž1/ušá+s6v5WigqDðßZ‹Ü§–XëÝqhЛõ’Q?™¡³ˆTZйº/É(ÄW ývD§Y‰ J—Ÿ j(.Îf·@Ü–‘+~î·Hó?º~º!{–]dW'A‰åZmI/éÉnÇ}˽/Z J͇ õ''·+‘A:µšj¹­6çe¢^ø¶^dÌ ÒG€Ç’ò,}0žKQ·QGiîyªô¼ƒžçÜž¸Û?&íðö.x»ƒH:YÏm¿âÕv%2ˆ‘ƒãCÈo™ÅÅäX$üîx Ú£ŒË4¹Þ¶)ï’òŒ4“¦+§«´:;-fÄ;}¥\K±‚¸’"dE»”kS#Ïi#þìh´ÔZÍ¿¤»%üÄößÁ–÷Ôc³3ÎL™˜{‰ñdÅäP¢YÁ#ž™‚RÅF[ý…ré³?äq;ð>„ŽÅ“œ6¥1ß‚ªÇÑöMk‰Qlª”6+ç^\©×mp •¢™0£kŒt"þG•Æ®aGˆÄ—Aæå°d`6`_™‰‘¦rш°õZ!ñ±È3Ñ^GJìôßÛ£|.]vôPm~«Wrè¶%= 5/%Û_ó³³wÿ:c¢Qeüyà1˜ØHŒéCT6#ñA‡P‚ûÀј\hi[]¢‰Š†81*âƒNDIÍ>¸1@HTR£XbH02(¢„õÜ{gfg·­íB‡Rí×î̹wÎýî7gîÜsﶳ (2£€? 8¤^P—°:Š–•¡E5Új¨ŸŸœP™-i3EÉ P9ÕE¹ d>PÔ5Râ¢*С@l¿ÕJ&g#^ê‚E?ãÝù;ÙS^Ó ÔÊá%@ÆÛ…õ[•UÔWY­)ÉYØ?9{‹‹*tBý•ÁŸÞ媡§Ò·¥9³®  Ã­S{5 ~¢<‹[½€cÖuì<¢¼BÞ¦Ôöƒ€îݜӹ³N¾rKà²ý:yÜ·q¯G ׺lc@Ã|ÎE±ÜÍîO…od«–ï4çÀ\ìŽQäÆ<3¼Èùyü!à©Vņ",PûúÂŒÎÖ`•â êÞÂzã‡àFp<Æ}n/sQLÔ"Œ˜¡îúag‘6‹¢x¬P1 %s3£øÛB5L(ëÑ(çyüÜ…Ÿ¼b(À¹E]4rB²ŽÕ%†Ø(^ËwÛ@t•&ÖsuüTŠ ÚéÀ4E“ÝÕìO,_4›©î›æ#Øàù,¥9üfâRê¥ÃÍÆ MiÍÆ^ûCéPö”x°€ùipd>­wo9ÈÏ’U*ϱ”nðÎd®ëAf§9×ZºÀQí0j³iα”\},£3ñÒ›ìtåÌKí™xÉ'~ÆM¤¨%ÄÆK6ýÝ^GœÈêÅ}Àóîé.9O\œ|f“Fö°“Ö¤¢Ìöbäį™Ó#×/mñßk&D70…SmÄ|ÿ£ídüYaòj×uU¶vÈx/¶FpÝêÁ9—m?‚˜€~v¸ÏÎóßÒLì^ò ÒùMy$»Åߟì&_%|Pªð$lÝY28ÿãÜŒeë›CŽ>¼ Ö“²á˜|;ÖóËDÕ¢_ÜVŸá¥jrûÿ»-΋û¾'§ÈÅí»Pvæt³Ù`üý‡:dc7ä<÷ÛNÙûÑfæ¢CÛ÷pÙ*v•oûBƒ’Á¼ÁBÌøÜ°z}R9ÌŽ®û–벘~.jå?ÅZ•u½à‰=·!Ï(è6¡å<>äWìcÉæ¢ìhõ>NîîsBÙYOðc+4׃7¹l‚÷¾zûF¤Æ€£7KÛ¢~×~<ë@îèh8¼ÝD´±°Fv¢–JE;oç»ë¤‘g„ˆí›ó‡lND{˜ ï½9¿Üíå!?©03;®üYFÙZu àëO’Ñ–ÏDVÿ‘!ø×ÚÛ¼|ÁVÇ‹Áá´É^ÞlÖyiI¼ÚðÓƒýѧöiG†Žl«¾Bhv¼@9‰&d‡Ý,7ä³ä°Õ’­ŽžtàÓ)0S§Ù]–*º‘·`ÄÊdaqòÅ*U6Š÷—ÔD 9éJ~„h^Ûõ°—•ÒÑ £9v£$6~fÌþI_Ú› N±CLìÂd:¹Ë$Èü¯E5=£fÒøïn«º!µŒ"\/k™0:›[é>õïnNý?× Ô¾s­äÙá'©o>íÉb¬Ÿâ?q®QE•;q`ÛcP›k)]a5<ƒiÖß0¿¢m½+æTF·xl^¾‰WÿWÀ¿C?¿ý*IEND®B`‚fftw-3.3.3/doc/html/equation-rodft10.png0000644000175000017500000000306312054411111014720 00000000000000‰PNG  IHDR:ÉêH0PLTE³³³¨¨¨œœœ„„„xxxlll```TTTHHH<<<000$$$ êYËžtRNS@æØf±IDATxœíXMˆUþº{¦çezgº ñ§ö"iÙƒÐ!ƒ¬Ä¸#(Š™€‚‹ûà!„—˜ÝídVvñ`Œ!0âOÔ2$QVtjØ@0¼Ìh4ˆYë½îž¿¸ŽÙÞuBL-Û]]¯^Õ÷ªëU×<`Q¤@3aƒÁÎÀA¦L¢Äf1À<°W«œjf½{b‚/ÎaObM.«Šë; ø£‚wipF0«ðøKÀé¥`ⶅÇ*e†$ ú»ø r “zãœÕîEcØÚ>Ÿ_P]X~b=Æ_ê©Ñ›t<¿àXMÍ¡,@ ¯éÓ¡š`óÐmT(‹ë";&|kÝ°°jHˆùžñ°9 î4ðqÊ•©‘â¸"4îeÇ@óöˆŠ@Š…ø¡Øl\Ž½Ú„¨§ŽÛËÒ’kvò>ôN¤&M‹5}©hÔBç¾yÂÛ´LÛyŠZFT ‡“þ¼dM1k£ªjb:©@æ‡a2·¾uKÚå°PPUª+nÓ’²:âú1cp¹Kbl·ïДýq®¿Auq–Ó3)˜YaÝœlá|½¹U‚㶕ÆP€šØt¥¤ º\]>KÂ:hþ× &õo£·DáP‹g‚^“oÃe«ñÜpdõVéE<ÖìgYll(¶ ê°n JS©¹ªíº6ú¥‚ {vó1ÙŠF X𹨸 *A¦+Ùºü¤%Y¹…ÊHiôíwàI%zv¨Ò¸}z¨´B Vñ0ùQ6hš#ÆÖ‚)£0l*¨'Ãr\¦ú^¶©û` •J^8Ã'5 À†i²ô^×SÈÄUFg÷Ø¡ ®XwR}Ñm3„QB´’=b¡— ªùØO©6VôÅ—8§ª"+:™HccÊ †ÄpÈ嬔hx‡ƒ0•OO êzš7¢Ð‘)­²#ÑBª+PÝ!ª\’·^A5¼“ŠÀÕš×z|±Ò$UÃJwÁQd¡t›„’èrqwëµ±Áè%êRÒú`¦¤Ç®Z[£ÿOó IFƒ(fv^”§Å,: ’”’ï³t<5zk1;ÐmÀB, ÌŽ‡à&ýohŠ÷Ï÷mdžõÄfœ€ïû}qV?´}ÅÙþ¸I™ˆÜþPÏ0+^v_ÏüɱõJ‡H)7½ÁƒgÑ~²ó?$/4öÊk´mX»Í€ì±Æ°ò#Ûˆ›€q¡ÃÁ°‘¼ØhÔUò~¢„Q –ú°…Ïp?u}%ÿtˆ2÷V?(@$ÈSÚë\ã‡ÁM=ÿ]S¤O \CV¦¨5f<Ÿ¦ðP@4÷ úþ~ßwB'ÛqæKMðèQ9÷¬ÿb8N³R%lù¯ Œ2:8Ù½W*ï kx(ðü@! yúeÊÑ7-Ê{d¥(A`Üpåëes]‘PEÌ“.&EZ¼»¢œïŠóÄÑHjN»”ßD+¤‚øÅ‘ö8¬m4ª’¹¥‰Æ~»¸  ÈL=yT?5qjçôËŒ§æ •Þ]†·9E©èBm4æ¡Ô±ÖRa'1ŽÛcâ )µÏÙ„½îs\MìÛn‰^ê'­Wξp­ãÀ±ð¾ÚVnM±#F9í ™˜ Rò«y%’d§ìvöMÈì\×!âWοkÉZx"Á>_ƒò¥Ø,¹‘™…•–›ä†•Ô? × ©ySrKpš—ž¢ÂÖ7b«¼;(!ÜYTú£xš*hy¬Ÿ‘pÞ™¡þaÀ#×A>þ—ô^4O+Þýä<IEND®B`‚fftw-3.3.3/doc/html/equation-rodft11.png0000644000175000017500000000315412054411111014722 00000000000000‰PNG  IHDR:vž0PLTE³³³¨¨¨œœœ„„„xxxlll```TTTHHH<<<000$$$ êYËžtRNS@æØfêIDATxœíY]ˆUþæ'“Û™lf|k©?Sða_dG¶Å"Li¬AQú éƒà¢Ö<øP Âmín§Íʬµ"þTíCC«²Ð ¶ôAK ÅV|hT¡`£Õ"¶Ͻ3ù]ݦMÚ,¥ß’Ì™ûsÎwÏ=sæä.p­P ÙpÁàŽÀCB+Z€¾Aô°,ØëeŽç`‰[»¶`®Î¯Ùl`M)©b ]ÞӉȷ˜M>#Vy!Õò KÄëaŒ;íf(b€¦ëwðeÔÈì4ªb™ gœb»9Ctcs»Š 5 ˆ«€!ÕÜ0°+Žx¥‡1½ÀÀ‹‹ôVúòDÌæ ÍÜVeɽYÒà¢DQO½yÌÎê5|À©Bc*nÖÊ[‘"ŸW OSH%ä˜8Ç%©‡?Àh7ÜO"ˆ7ÖÙê±ë²v=8ØÃ(k ¶®ÔþUÄD/ÁvEhZÿ:ŠIO¬&ÞÞ@«vÝæ]Exs„6»GͧI—¥h‹©ëUU:lŠbÜ–ÍüÚ*Ô©Ü‚AsdTUd—šâ7M!&S«'¾?c >÷©ÙÛx4µ@œoQ¾'­¤ÑŸo'e–²w¶h¿Ù\*ir×11ò&×BLî†\l•>é$utg»QúŒ !Þ‹7¯?ÈEÊÆè¡]´ðÀ&Ëý븹á F ï[Ãól jØø@ÔP xóÁ¤¬5±°4¼jüÎ|ÄN·y;Y<·—ˆá«É¦®(Y”Ìl_Š5ùÎ1ÃA’òáOר‘•‰PPIs<ʯ”"íÆ«ŽÒ6´pE‘=e¦y¢o• #„"2klµXgªGÑõ(7giˆU*¥…Q|qD†é–m3sŸU­†š ¥Ú,Oîv#5·Ñ®8÷X:NvD§€Æªv Î,2†}4z,5>ÝpNi©ý…ab}\§Éò §RN\ÔêkB·•¬a 5êZš›ØÉÁ/³ÃEåiNyÊÛÄmŒ·_Ž®Ô.x1E¥‚»mˆXuŒ¨¯{á)òÖojEuŽûZÛÉFlÈ–Ö[:eDZ[¨Òç !<‚>`Õ ˆâ¥…4-¬/·H(… `fÿlª´£xý(2}kanß*ná"Ìòa3 ìßr4*p AÌ, 6§ãaQš›2“[Xv½ØTêÙCaùo7Îè³ÿpl¾ÔѤ䈋QçὨ”ÙÙŸbç6Fì‘߇e¨®rÛÈÒo Ë?ÍÊ#ÞXç: hb¢‘vâ¡(›u*­çlë$x¿oJwÜQo«ÇÍ*ë>Àm°y§Ù2^râë¤v ‡ÜXfZä^=œ!Nü0âH;¥Ød“(¿颲6•Z'Î/Ì_eÃÜŸm¶ñÃëÊÔ)vŽ]žÞKlbçëõª^&ǺÙXúpÈÁ—xˆÊÒBp²›ÍÇ¡„/ÄǦH6:™4³«þ$¼¨gh6sBÓÀ“V‘¹Y*èO›ä0r‘柡ΠØ^Äf«`”ìT¶O‘$‚]óÁË›x!´C³IÜôÞl-mr´ù¤TúH¨Å£!…‡3 2ZÖ¸Hý6°Iù€Ìå$L[¾ÜzVéò*J{ó±S„ÎûËŠé.ߌEvXVˆñŠv!MlŒ9Jµ_n÷ÌX½^–Âíͧ©¿ÀÎïÈW³O1NÌœØ>÷*ãñŠÅ¡Ò¶Žð66'(n}¨õúe(5Œ9êïì¸×b£ÏþœÁ‡‘» D›Ç÷zzhN­ Á¾ïn1 ýѤµÊé—B©u¸:]WzÐÚNÑܦ¦ð1‚r¤)B_ìŸ ±o"áµFKrÖmÀ¾‹„í«;Ú¢…¨‚ð¸Éè`f´%bv6Ê×âKMì¯çÿÐ ƒ|Þ%†Íd BÍÈ‹’ÐÙÛ@ð4¥ÊჭÈÞMAãÏ£4l*“x†rsqjIøÆ{7ƇJ#ÂãK)x‡‚¥ýký˜žÂUIEND®B`‚fftw-3.3.3/doc/html/rfftwnd-for-html.png0000644000175000017500000003751312054411111015025 00000000000000‰PNG  IHDR«‚à8aY 6iCCPdefault_rgb.iccxœ•‘gP”‡†Ï÷}Û ì²tXz“*eé½I¯¢Kï,K±!b"Šˆ4E €£R$VD±± Y$(1ETPòÃ;çÞñÇ}~=óÎ;眙@@ERR|?{NHh¾!’—™nçãã ßåý( ÷V}¿ó](Ñ1™<X€|^:_€ä€fŽ ]€VTRº9 ,~HhrXq_}XQ_}Xü?@¢Å}ãQßøö(Ûñ ±1¹ÿ´XAN$?†“éçbÏqspàøðÓb’c¾9øÿ*AL®À!-}?!.^ÀùŸ¡F††ðï/Þú{ð¿ÿßôÒ¸ ؾ³¨j€î]RÿÍT0 ºîð²øÙ_3(ÀHƒ¨€&肘%Ø‚¸ƒ7@(lÄC ð!òaA 샃PõÐ-Ч¡Îø·á.ŒÂ¼‚yxK‚:ÂD¤ED ÑAŒ.b8!žˆŠD qH*’…ä#;‘¤©Aäär¹‰ # dùù„b( e¡ò¨:ªrQ;Ô @×£qhš‡¢{Ñ*´=‰v¡WÐÛè(*D_¡ `TŒ)aºsÀ¼±0,ãc[±b¬kÄÚ±^l»‡ ±9ì#Ž€câ88]œ%Έãá2p[q¥¸Ü \®w7›Ç}ÁÓñrx¼Þ ‚Ãçà‹ð•øf|'þ~?…O Ø ‚Á•JH$l&”:— ÄI‘H”&ê­ˆÞÄH¢€XD¬&ž$^"Ž§ˆHT’"ɈäL #¥’ H•¤VÒEÒiš´D%«‘-ÈÞähò&r¹‰ÜK¾Cž"/QÄ(+J%‘²ƒREi§\£ŒSÞR©Teª9Õ—š@ÝN­¢ž¢Þ NP?ÒÄiÚ4Z8-‹¶—vœv™öˆö–N§«Óméat}/½…~•þŒþA„)¢'â&-²M¤V¤KdDä5ƒÌPcØ160ò•Œ3Œ;Œ9Q²¨º¨ƒh¤èVÑZÑs¢c¢ bL1C1o±±R±V±›b3âDquq'ñhñBñcâWÅ'™S…éÀä1w2›˜×˜S,KƒåÆJd•°~f ±æ%Ä%Œ%‚$r%j%.HÙ[íÆNf—±O³°?IÊKÚIÆHî‘l—‘\”’•²•Š‘*–ê•ú$Í‘v’N’Þ/Ý-ýT'£-ã+“#sDæšÌœ,KÖR–'[,{Zö±*§-ç'·Yî˜Ü Ü‚¼‚¼‹|º|µüUù9¶‚­B¢B…ÂE…YE¦¢µb‚b…â%Å— Ž'™SÅéçÌ+É)¹*e)5( )-)k(*(w(?U¡¨pUbU*TúTæUU½TóUÛT«‘Õ¸jñj‡ÔÔÕ5ÔƒÕw«w«ÏhHi¸iäi´iŒkÒ5m4345ïk´¸ZIZ‡µîj£Ú&ÚñÚµÚwtPSÃ:ëð«ÌW¥®j\5¦KÓµÓÍÖmÓÐcëyêèuë½ÖWÕÓ߯? ÿÅÀÄ Ù Éà‰¡¸¡»aa¯áßFÚF<£Z£û«é«Wo[ݳú±ŽqŒñã‡&L/“Ý&}&ŸMÍLù¦í¦³fªffufc\ׇ[ʽaŽ7·7ßf~Þü£…©…Àâ´Å_–º–I–­–3k4ÖĬiZ3i¥liÕ`%´æXGXµÚ(ÙDÚ4Ú<·U±¶m¶¶Ó²K´;i÷ÚÞÀžoßi¿è`á°Åá²#æèâXì8ä$îèTãôÌYÙ9ιÍyÞÅÄe³ËeW¼«‡ë~×17y7ž[‹Û¼»™û÷~š‡¿GÇsOmO¾g¯êåîuÀk|­ÚÚÔµÝÞàíæ}Àû©†O†Ï¯¾_ßZß~†~ù~þLÿþ­þïìÊžjfö1‚ƒZ‚ƒƒËƒ…!ú![Bn‡Ê„&„ö„šÃÖ9­;¸n*Ü$¼(üÁzõ¹ëonÙ¼áÂFÆÆÈg"ðÁ­Ë‘Þ‘‘ QnQuQó<Þ!Þ«hÛèŠèÙ«˜ò˜éX«ØòØ™8«¸q³ñ6ñ•ñs  5 o]듼“Ž'­$'w¤R"RÎ¥Š§&¥ö§)¤å¦ §ë¤¥ 3,2fÌó=øÍ™HæúÌK.ÌÒÌÚ•5‘m]›ý!'(çL®Xnjîà&íM{6Mç9çý´·™·¹/_)GþÄ»- [‘­Q[û¶©l+Ü6µÝeû‰”I;~+0((/x·3xgo¡|áöÂÉ].»ÚŠDŠøEc»-w×ÿ€û!ᇡ=«÷TïùR]|«Ä ¤²d¹”WzëGë~\Ù»w¨Ì´ìÈ>¾Ô}öÛì?Q.VžW>yÀë@W§¢¸âÝÁoVWÖ¢Ê:$¬ò¬ê©V­ÞW½\_3Zk_ÛQ'W·§nñpôá‘#¶GÚëåëKê?M8ú°Á¥¡«Q½±òáXö±MAM?qji–i.iþ|<õ¸ð„߉þ³––V¹Ö²6´-«mödøÉ»?;þÜÓ®ÛÞÐÁî(9§²N½ü%â—§=N÷áži?«v¶®“ÙYÜ…tmêšïŽïö„ö Ÿs?××kÙÛù«Þ¯ÇÏ+¯½ q¡ì"åbáÅ•Ky—.§_ž»we²ocß“«!Wï÷ûö]ó¸vãºóõ«v—nXÝ8Óâæ¹[Ü[Ý·Mow š vþfò[çéP׳;=wÍïö¯¾8b3råžã½ë÷Ýîß];:ü ðÁñð1áÃè‡3’½yœýxéÉöqüxñSѧ•Ïäž5þ®õ{‡ÐTxaÂqbð¹ÿó'“¼ÉWdþ±ÕšK>‚£n9ǦjXÕ Ø.¥×eM!Ù×¹:23hÏ—,o믊·\—î·:Ói]õ›Ÿ)0íÎ8¨g®«“m>àÝ›]«†böhaeÔoA49êYãšÂíqåw¦À„—[rrÕ9vȳ(%à wϾµv ¾ŒÅî4V“½Ì8ýЄÜ^æÕ§­*0oiË‹-Mz°ºàƒŒ²_XX` RÆ‚›‹²Ñ¸­~¹¥ìYPT£½ÝãÅ_“ñ:`!3¢ûÓdAn"Ú4#ºUNãKðAv{ÜŸÏ瘷ˆ5 4ðÞU8ŽÁ<¯þï¶ÿ­©ÇûßÊ2elª´%˜ÿ[T10%O§‘Ÿ}°¯^~ò2cOïž2˜·´ì>½{ªâlÎ+ï9Sø!Î^`-çÌŠë€ýî­®.pÆ‚¸ì ĵ¢Ü_PÞm}@€ê¬ž=3"º âÊsWp$/>xÑ~ñüõó}k²·Õ 8ÚÕ5=éðQ¨ \×Eas¬4Ì3¯¥m–<²ü§þ‘fÌDáïŠk¾Ã»ðï›û‚/ò}ÁKÚÇh fë—ܲ8Fêõå ìxÁ×1Œ8pêõ%&àåÕ@mC8Œ ñפÝ2:røàxÒÇ‚û>ñÔ(ñ¾`}^ଌĵ:ÛÖßåž!Z…@uÒ¯®ýÙ4¨Ž^0W”?lƒ‘ÓY¢Š{ÉptúKÉajÇô¥Uq6çye„,¥UqÎd› PÕ (þ€ÃH¼/¸ä/ÀB‡ ÖV®Ê0#º@…W(D¶ÑÍž]ãÁtëÓéÔýôùíwwª pÕô‚Ïçó`5×"ª\%¿ßÄ´õJ¨ÔîðÛwª\%¿ß y°ŸÃGa} ¨ÿ ·wÔ(\š&ÝX3# %“Ó!»ˆ;¦žÓGK«âlÎ+ò}ÁvLêÚw%7ô€]#õú´+Ž| Œ8pêõU3#¸™ ñ×H@ 2 Ä%¸$ —â’€@\ˆ+Ê:+¶Á†«â^ò( pI/ˆKqI@ . Ä%¸$ —â’€@\+pfŸÌägÜ‘6 —â’€@\ˆKqI@ŠÖŸWpêYõƒ0å¿ï]Xätú~-Ë«é&þXHã©ßšÖö&žuYHrmƒ®Ë©óù<šhývß ÔúÿÔâ#™äfÚhýÔ›zÂ5"(שF›}z¸d$)ÂeGxпîG„#i$ e™I±Ëoµ\í8à HF‡Œ¥Û­øã¹ü/­¿É$pÚp{Æ‚¸$ EÐä.$ EÜ…â2¼N7©Í7Ó^$éOÈõÔ\­½¡Õ½YpƒÛÃó6>´Wpµö05Ósð‰ÛãÓ§ Ü_׸ë:XS Ž¼ m@ \—+ãæm j®àòÜÆTC/{GXp5ËÀ®F›ƒ{Ãs}%à:‚²›Z |þ[Yèe¹å¤ ”å–=- Ä%¸$ W ±à¿}e*_ºÿ4â xáçÌïÿðêÞU˜ó«ßÞ» Wüø§ç(«ä?¼~¸w€²<¾y Ôüáÿ~œ±´¿ÿëßK{ùóþøå_3˜½zÛ€ûêTøAQ½-~ô?7¡zÁ‡_ü¬ýâñO9Pv¿øYá5d =dÝ ÖLœc_üæ£îë_ÿî/·¨ÙwÖ°iš/~óQ±uË(èHÈãŸþ:ø]—¦äº1ªÿ«=ÁfžüëßýåÆùÒ¬©a?£ocÕo/£pmÀZV…t@ƒ¶IÿDœTWO°ý²opÎl©aöÌU·¼‚¶!A¿mÒý‰šú›zõoíNÝÌ®’Ûk¡n‡jÚă{4 ›ïÚ2ƒ†Oymóê.WÙ.k8UÉB궷‘̲ÞT!{gS[þÌtœAÿ¶˜:—úMÅîÁþhæÍRæj »¯³¶^µ¤n{IÀvq®-Quƒµ­7–ã¿3ÒèµÚ®s×=gð„¶Ø]eÛ5 Ójx…Ôm¼¼=Šš#]KðÕRÏ F›u ¿Û·wäMUcy [Ùë™±n Då_N)¹†…Ôm23î‰q› M<¸±>ÀóJ®a!u› ÎÒ¾qÿ×&A®ûj®Ì†Ù‚7Ž?ÁmIÃëó“C°¨Ñ ‚µi¸hFtBÖw¹틦išrïû¢iš&æ5ÖŠêÍÖáýûFqé=!«B°Þøk_  µ”Õ±²ˆvÎìZ½å·-Ï·WÜ·0«Ž?à–_ <Ô}Á@Xi#Âëðj3P¸vý®í7®¬¾'dfâ±øn&Ë}{)wņ øª“x_ð ÅP£ô•º@¥6g_DkW¿ü°Üé£MñÕëo¸üøæqÉ·–üÈàñu[X±Åßzõ«‹>(…Ÿ3…W¯Ýo¾šüÚèáõƒý‚“™E´s¦äê5Mó£üùñÍ£õ¸$à¸nµîî¿©gÞ~gU  8âîK÷3wêk`»ˆ 8hÙõ3ejSç)·ßWÈ(è}Áý½\»ý3—lÞ|›Í .w2¼üØ.hθº±tû­?8€ˆ½àQË7–ž¿2TD6ÍDïr0|óJ»‹Ø Þ²õõ僺ÃP/mÀ¦)fófàÆ$`Óc…¨$ WÄë€@í–섹„ª·v£ôŽeUJÀ÷ÜkÞßMÓ4åN9ü¢išf|{¬£+ü ¨ÞWÓP¾gù p¨å$­šE´sf×ê-_–i¾(£Y~)PG6,úäºÕŒh . Ä%¸],yÿæßÿ¡ô͹í˜^šÂϙ«×|·czsŽáÕ§¯î]…9…Wï\C ³+ü-«ÞFm õ‚¸$ ×é|q·ÓÃë‡,—œ€ch3AˆKq¥$àé4²PÒèƒ%[1°Ÿqò8½à"ô'è–©üfWø[V½,²%`1eàHVô‚»Œ;¢Ì¡Ž-¥ (þ 8›2à[hõÊ3c ’1™k~Î^½Ž‘ . ĵº<ÚÕ5=¨‘6 T©– w…“€E(ÿl.¿†Ù|ËÕòÛË9 P÷„qå a£ò'O•_Ãì¾åŒjùíÝÿ:`Û­>}§ÿàå×@«–ˆ)\ân™—©´¥mØÝh|Y¬{ý±BêeÆÏç¶I(þ€ý¤÷³iŽj‚Ù‹è$^¼AÓLü{»ÿHȨ¶ÿ«ìju^Y´_ïÔ*‚À~Ò¯‚©ÿÏUiØr÷õèƒyÚ ¸÷„q1à.ì˜Äå: —Õ±€¸ì˜ÄeÇt .;¦q â’€E(c­òk˜]áoYõ²°c:—6 —âZ=³¯Ò±à‡×öÖ:m&¸'ˆËêX@\®qåÙ1]èQžRÛÝÍsÔ'¨ò§–_Ãì ˪—Eb°y©˜L•¸©ÕmÀ˽1uJå ‚‡p6eÀ·PbÎlL6Ö*üœ)¼z”ù€mÞI= vÙÖˆÖª“8,ï€pOT©– w…“€E(ÿl.¿†Ù|ËÕòÛ“€@\é÷„Tço_•˜èÐÿkßÿÖÔãýoe™3±¶bË¿UöA)üœ)¼zMÓ¼jVH­ÔÃë‡þïÇ üû¿þ±´—?ÿè_þ5cÙ«÷㟆8OúþöÕ©äƒò£üÙ)½Åþñçœ+¤·ô«ßÞ» G ¸$ —â’€@\ˆKqI@ ®@÷„‡ñÅo>j¿øõïþ²¥ T¬‹Â&) % p i(¤á¿ùh4%àÐÃ/~vûý¢išæ¯»ÐMÓ4…¯ó±“Šê-ÒOà 8ôø§E Z„ZHÃÚ0YD;gv­ÞL¨µô‚XŒ„±˜ „³1ø:î â’€@\ˆKqI@ . Äh6Ì/?,åQ…W/a÷ñúwLõ«‹>(…Ÿ3…Wï÷°cúuÝB´‘»â²ˆvΔ\½ÆŽéˆKqI@ ®Ã&àésEO`…c& ø–¨#O=Íw7sA¦øÕ‘€MÓœÏç~®NQf2û©òžóùÜ5-A±„Þ1½m¶ÿyY¸ÓõQ‡Ý1½‹¼~öÍç ”„Ȧa xKvL¿dÇôÕpΔR½6 GsPÙ1ý’µa²ˆvÎÜwÇôfYGXDZv0D‹8ÎŽé[I@ ®­ h ^›°»;-Wmn)=»»Ó„ P©ÄÜœ+¥¬20µ6AÉkôwÈhšæñÍãêÌš¹’C`Àl ®u xµ‰ç‚ P‘ ¸°‡+Z,MÀUø„ P…E ˜0¾!ò]OÀäá]!îJnœÝ"’Í%`–É}B(ÖdšÛ Þ-fDkeOÀì @!h$éf¢ja,n/àFvJš ©U8I”ÏÊ@\ˆKqI@ ®‘‘ZÙÖ|ó˜·@úþöUÓ­~üS#„e‘€s^~ò2WQOïžrÅ”?~ù׌¥ýý_ÿÎXZÓ4/þQÞÙN/ˆKqI@ . Ä%¸$ —¬LÚ"c£?e½²ZœN§î` ¾¾_¥®(¹n}*Ð_lÉÂK™]‡©¿¨Ýãݧ¢¿ºíàëÁ3çË¡í¡iåè×Í‚£?SòàÇ ^®¿Ôqûœþ™yua½´zîJV`4È.¿zu§ìà™åœŽ\jC§=(£_·ýâµGðòGf éþ9õœ©SkæGVÕv?ðþ.Ûwƒ“/Ë«ŒžÐSõ‚·4\5Õ6Ìxç‹Z~juOè7*sU2¼¿…m·ËÍ“~tÏ–Z®[ÒíS`p¸§N‰µ›ÚhôÁBö’€=óž@wÿSÌ}íšG ÿ„ß…¼¿ùNÐòSsù>VºÀEÙØ n.(®ž0ƒçLý‰]U‡©rFŸ0zañ.$àý-¼n½üñ„Ù…üAŽiK\G,/sf,xm’Ÿ0¹= X©ócôñ„5ú 7mK¾^RæüO]}•„³«œéfDqI@ . ¤«ýŠâ’€@\ˆKq™8Ç&¿uùå‡?»wæØ.½@+fc/Ÿº}÷;]–Ð â’€@\ˆKqI@ . Ä%¸$ —âró—6 —â’€@\ˆKqI@ .kDsÝ??üI÷õÿ|ùyËì üç‡?žåuO§S÷u7÷ërß´e}»Ÿ}•%?¸åÕÙNrE?˜ú‘´][ìeðe|Ý~²t!Õ~‘1tF_eþùY^—íô‚£û`·š‹OZ–Þ|µßýç‡?iÿ»|~÷àeso¾•—%|»ßÀLpû¯ñê.“÷|>‹Å»€‡Ò~’»ØòÏÕL¾ ‚éjo´Í²îiÿóåÿ]Þ=8|k_wÔ kÚöÿ?jᯱËÍÑWéž“PmnF/øP®~ª·¾°6Uý¼kCp¾#|õu/óeê=®J¢ä_£¼«Žäº6ƒrt¦Æ=—ÿf^wá˜ÃÂgnq›W!; xpW;Â]îLÅÍöøí÷Ëï7 »G²Än®`ºü5^^:¼¼¶˜·ìAF1õ \ÒÍÕú[[ÎÕç/éçž©Ò¦^×4—ÂI@挎çfì’wÐ ½úºóá2:’›=nó*ìĨã+¡ru²ËÝ]ý-íúk,áÅd6 —¿<WBã¢ü`kæwU¯‘=8®@\zÁ@\ð?ú÷‡tù—€ßko‚pH£mó›Æun¨ÐÔÌ©óhûF¨U·„O—nk×j”€Àõ×yœyšlëSÂ!ô€]Ûpþ—ü^ûçÂA¨QÚ'×8ˆ„Œ6 p‰»ýiaiqI@ . Ä%¸¢ÜüðúáÞU€@Þ~ööÕ§¯î]‹+ß½{z|󘫴Z<¼~ˆsˆ³øö³·¯_ü0WiMÓ¼yþ&{^0YÄ|ñÁ‹ö¿ìÅæ--o=wz×ÌËø Ï~³ŸcMîOA^?øíÈ¥°@½à¾ç¯Ÿ›¦yñÁ‹ö‹íö8ð]%·Õ§ß5óò&ËG0ã9–±œ=ŒÆ_0wúü?ýœ÷ðç­¤ÈÛ¨pÛc}5ÚÇgΊÁ·æ \rWxµÌ„·<ó~ôíçÝãÝ׫¬-ðÛÏGCðP½àãíõ‘7¯ÛþŽ4Lóüõs÷«ë>êŸ]™ ¼z×8ÿ„¼oùÛÏÛlêÂ(9þ–8Õè8Z°½Íù‹\µgÛ-A!˜Ýeã« ‹&é>ÕšK>‚£n9ǦjXˆ.›^ópÔqpaê Úó%ËÛú«â-×¥û­ÎtZWýæg L;‚3ꙫÀå66ç l[…KÊ?T/x¹Âûƒ]Û!×8]›ûÆ‚3Ê~ò̘p—\æ[UfÆ·œ7û¶xœ6àr;_Æbw«É^fƒ~hÂn/óêÓV˜·´«Åö[gËK›qµÀËÇGŸyœ6 ½>¸¥ì}ˆPîÚ^åhmÀ6k¡|7î¬Àíñ×êí…*457p‰C% À y8ÿäêçÍ-´pC¥pTƒ4|óüM¬Õ±–X~¥ÃêX‡gu¬-^|ð"ïXÇ`u¬…·|4qfD´–‡¯ŽÀl Šå÷½Í;ÎŒh ”,×% WbVâs;P£”Í;!Tguv7Þ™J Øê{BúKô%-|Yf;¦g4˜Ýýn/gM_ýV–‰ÖS¯Ò?è uf‹ZvLÏ“€å¯È’w~ãžÜ²ÅÞ÷„d)0宸v¾hc ƒ}³r•fÕR¸¯l3¢Knf‘kgÕ…-^qûÖ6À¥”ì÷|› ³/m#Ô\Ê/(Gz°ºàë4©ºÍ­ç£p~gÕUηæúO} l—’€ƒÖ_ùà kõ7{ÍÒý¼,ÐÖ½P‚lwÅÕ>$Ò£.¤¶´¿¦ ´+”cuð²Åw€MÚ惩߂koE8U „rä ®+—l„ºªõ—½@à6òô‚+Š¿Q%ï¬ ìgëŒèc(ygU`?‰ó/ðHcÁ@FÝ®Fyï“Ë"ñ: ¼ÖZµ“ïmØ'¸ƒBÒ0qI«Ñë€%7 —¯óaª ÜÑT–²6LSÿÈï¼…ƒVÇ:<«cmÑ_kÉîæ÷j¦ÏˆžZ*`¡„໚§«ÊÜ´2B¿1(%Òš{o?{ûvÙ3W =G¼'¸±ÜµÃ&íóÛÿÏ??}Fô`•@€Q[â¯Ë¾WŸ¾Z>Ò¾b—ƒ3ØÔÔôöÖæ×›çoÒ~pþºaúÑÝ×®;Ù>F<_B†•ÄP©¥mÀ%½]QÔei¶ö´ vFvLçªvLûÙÂ)1ÿ1_‡Ç7Í9†WŸ¾*¹À«·Gåsˆ·hš]²%!4§Šj߯•€jœß¿m`ð­öŸóCµÙVFH+çf^?¼³îe§)"Ww¦ ÁÑç·™2‡¸XS¹6x|Ik1ý:àÌ‹EKCà6ºÙ03é6h¥ÍÇ‘ù€@M.»¡ƒoÍ?2²>àÚÈëj-LªM³+ºhF4ÜRù‹æ¥Ïˆ8ךä]§J¨FöÛÏ´Êd¼Â¨ ˜ÈtÙÃsˆ#€@Mò1K@ ®M ˜¼n5@š¼ƒ!÷ŸµØÎœÜÄןN™ej¥ÑpKû͈Îu#F› ‰cÁ oÊ[^ÚÔ½~åÏ,nfje–å¡‘’€Ù§äŒÞÏ\Ñ-w@¥ÒWÇÚ[¿k ÐìÐ$Jìß i&þ€½:¦½ü§ì*e}ÀÓét³Šöå\ „˜ÖîûÑÚq$df,&-§ú?Õ}=ú @^VFʵvôµ¬’ÄUèHÀ H@ ®¥½ànDæê6¹¤Í‡³J>ÀuKÛ€£óTvu83+$,&  Tc>㦖˜š!ãXÛI•€ÀÑ,oJÀD6;<‡¸FkÇŽ% p4û®}©Ÿ»‹]M­•0y0çÊ—›X´}—¢½_ÊÔÝ”‘ëÖŒiÕØ+nk×–P–øÛ´WÀ½dÌÖ”‘Ѷœ=PmÀ~ÆÉ;àô‚rU³6L•Žé²‡çG°¢ Øß®)ZÞwQâÚ0⸋ìkäï–9õòAKÈçÚŠIDATD›]íý6ñÞò1Þïü:ö—Ü\„ì'ß1Îæ#qˆoÉÚ0@DkÇŽóìlz P;%‡5ÕäJhŠIÀcŠ9·+”˜‡¸?'ïÒà9K¬NÀ™¥ðé²Ù•_Ãìâ2ÍÜÚ)ÙîŠ 5¸£ŒiåîŽZþ¾Ua0£ÿ»íkêñþ·²Lé˜*mIæ¿E²·Ÿ½}õé«{×âŠÇ7‰ xÙó-`á!C›qY¤ß×Ï»˜ƒ!À-̾åÍÏÄ‘Í=àŽÚŒtxöûÌ*3•qýX\hu^®?“°&@ RfDw_ô§CŸ ÜF¿xÙ\;­ ÔdI'wyG8¥|ÕÚ2ûÁo2Î_-?ùgµ[hs0K^–Óýsíppâl˜Ñ‹}ZÀUýðZ{ÙîÛÏÙ×=’–­Ùî X+! GG?’Suõ­l•î”´üϼػ2À¨™ðZ~_ðBé÷7e‡ÝFÏ_?/yš•ÏÊ[ôWF˜ïŸæ]@a­lë,´=õFS5¡Ø”’ÚiÏn{ÈÒÜ›iQ&Üœ¾6LuKw‘±Ÿ;3âÑŸ³ûl€{ ¸„eš\ûÔÈ=!Àq¬m J@ ®¥½ànÀJgD¬Öª1ßÉM¸5xi°kßièwÔÞÜ…Ýe&î{_p¥lƒ‘Ó¹j§Óß~ö6᧦j’¾czu옾±4÷oTø!Î^àN;¦§-5Z“M+#Ü^öeŒ„qI@ \ÛWÕŸ/A/(Za ã*KÒó° hý.8€n²ËÛÏÞ¾í=2cÕ̘c& LàH¾ýüñÍó7íT˜…ýâ…c&u$`?ÑÚ;óºÿ>þî= FK–Ý_;X\G6ßÝ‹Ò_ŸU'ª°jÉÒ%2–VMöõ›xƒ¶žX„mÙ#xWU&`¿,ò .E¥a} ¨ÿ ‡q÷4Œ’&vL‡*Lå`Y;¦˜ÓGK³2ÂF…âìÚ1`Òݯ¶$ p#…¤^ŸöU`ðu¬ Ä%¸$ —â’€@\ˆKqº/øÞU€@vÚ1=¯@;¦\ú苨“ûh?IEND®B`‚‰PNG  IHDR«‚à8aY 6iCCPdefault_rgb.iccxœ•‘gP”‡†Ï÷}Û ì²tXz“*eé½I¯¢Kï,K±!b"Šˆ4E €£R$VD±± Y$(1ETPòÃ;çÞñÇ}~=óÎ;眙@@ERR|?{NHh¾!’—™nçãã ßåý( ÷V}¿ó](Ñ1™<X€|^:_€ä€fŽ ]€VTRº9 ,~HhrXq_}XQ_}Xü?@¢Å}ãQßøö(Ûñ ±1¹ÿ´XAN$?†“éçbÏqspàøðÓb’c¾9øÿ*AL®À!-}?!.^ÀùŸ¡F††ðï/Þú{ð¿ÿßôÒ¸ ؾ³¨j€î]RÿÍT0 ºîð²øÙ_3(ÀHƒ¨€&肘%Ø‚¸ƒ7@(lÄC ð!òaA 샃PõÐ-Ч¡Îø·á.ŒÂ¼‚yxK‚:ÂD¤ED ÑAŒ.b8!žˆŠD qH*’…ä#;‘¤©Aäär¹‰ # dùù„b( e¡ò¨:ªrQ;Ô @×£qhš‡¢{Ñ*´=‰v¡WÐÛè(*D_¡ `TŒ)aºsÀ¼±0,ãc[±b¬kÄÚ±^l»‡ ±9ì#Ž€câ88]œ%Έãá2p[q¥¸Ü \®w7›Ç}ÁÓñrx¼Þ ‚Ãçà‹ð•øf|'þ~?…O Ø ‚Á•JH$l&”:— ÄI‘H”&ê­ˆÞÄH¢€XD¬&ž$^"Ž§ˆHT’"ɈäL #¥’ H•¤VÒEÒiš´D%«‘-ÈÞähò&r¹‰ÜK¾Cž"/QÄ(+J%‘²ƒREi§\£ŒSÞR©Teª9Õ—š@ÝN­¢ž¢Þ NP?ÒÄiÚ4Z8-‹¶—vœv™öˆö–N§«Óméat}/½…~•þŒþA„)¢'â&-²M¤V¤KdDä5ƒÌPcØ160ò•Œ3Œ;Œ9Q²¨º¨ƒh¤èVÑZÑs¢c¢ bL1C1o±±R±V±›b3âDquq'ñhñBñcâWÅ'™S…éÀä1w2›˜×˜S,KƒåÆJd•°~f ±æ%Ä%Œ%‚$r%j%.HÙ[íÆNf—±O³°?IÊKÚIÆHî‘l—‘\”’•²•Š‘*–ê•ú$Í‘v’N’Þ/Ý-ýT'£-ã+“#sDæšÌœ,KÖR–'[,{Zö±*§-ç'·Yî˜Ü Ü‚¼‚¼‹|º|µüUù9¶‚­B¢B…ÂE…YE¦¢µb‚b…â%Å— Ž'™SÅéçÌ+É)¹*e)5( )-)k(*(w(?U¡¨pUbU*TúTæUU½TóUÛT«‘Õ¸jñj‡ÔÔÕ5ÔƒÕw«w«ÏhHi¸iäi´iŒkÒ5m4345ïk´¸ZIZ‡µîj£Ú&ÚñÚµÚwtPSÃ:ëð«ÌW¥®j\5¦KÓµÓÍÖmÓÐcëyêèuë½ÖWÕÓ߯? ÿÅÀÄ Ù Éà‰¡¸¡»aa¯áßFÚF<£Z£û«é«Wo[ݳú±ŽqŒñã‡&L/“Ý&}&ŸMÍLù¦í¦³fªffufc\ׇ[ʽaŽ7·7ßf~Þü£…©…Àâ´Å_–º–I–­–3k4ÖĬiZ3i¥liÕ`%´æXGXµÚ(ÙDÚ4Ú<·U±¶m¶¶Ó²K´;i÷ÚÞÀžoßi¿è`á°Åá²#æèâXì8ä$îèTãôÌYÙ9ιÍyÞÅÄe³ËeW¼«‡ë~×17y7ž[‹Û¼»™û÷~š‡¿GÇsOmO¾g¯êåîuÀk|­ÚÚÔµÝÞàíæ}Àû©†O†Ï¯¾_ßZß~†~ù~þLÿþ­þïìÊžjfö1‚ƒZ‚ƒƒËƒ…!ú![Bn‡Ê„&„ö„šÃÖ9­;¸n*Ü$¼(üÁzõ¹ëonÙ¼áÂFÆÆÈg"ðÁ­Ë‘Þ‘‘ QnQuQó<Þ!Þ«hÛèŠèÙ«˜ò˜éX«ØòØ™8«¸q³ñ6ñ•ñs  5 o]듼“Ž'­$'w¤R"RÎ¥Š§&¥ö§)¤å¦ §ë¤¥ 3,2fÌó=øÍ™HæúÌK.ÌÒÌÚ•5‘m]›ý!'(çL®Xnjîà&íM{6Mç9çý´·™·¹/_)GþÄ»- [‘­Q[û¶©l+Ü6µÝeû‰”I;~+0((/x·3xgo¡|áöÂÉ].»ÚŠDŠøEc»-w×ÿ€û!ᇡ=«÷TïùR]|«Ä ¤²d¹”WzëGë~\Ù»w¨Ì´ìÈ>¾Ô}öÛì?Q.VžW>yÀë@W§¢¸âÝÁoVWÖ¢Ê:$¬ò¬ê©V­ÞW½\_3Zk_ÛQ'W·§nñpôá‘#¶GÚëåëKê?M8ú°Á¥¡«Q½±òáXö±MAM?qji–i.iþ|<õ¸ð„߉þ³––V¹Ö²6´-«mödøÉ»?;þÜÓ®ÛÞÐÁî(9§²N½ü%â—§=N÷áži?«v¶®“ÙYÜ…tmêšïŽïö„ö Ÿs?××kÙÛù«Þ¯ÇÏ+¯½ q¡ì"åbáÅ•Ky—.§_ž»we²ocß“«!Wï÷ûö]ó¸vãºóõ«v—nXÝ8Óâæ¹[Ü[Ý·Mow š vþfò[çéP׳;=wÍïö¯¾8b3råžã½ë÷Ýîß];:ü ðÁñð1áÃè‡3’½yœýxéÉöqüxñSѧ•Ïäž5þ®õ{‡ÐTxaÂqbð¹ÿó'“¼ÉWdþ±Â¾Ô}öÛì?Q.VžW>yÀë@W§¢¸âÝÁoVWÖ¢Ê:$¬ò¬ê©V­ÞW½\_3Zk_ÛQ'W·§nñpôá‘#¶GÚëåëKê?M8ú°Á¥¡«Q½±òáXö±MAM?qji–i.iþ|<õ¸ð„߉þ³––V¹Ö²6´-«mödøÉ»?;þÜÓ®ÛÞÐÁî(9§²N½ü%â—§=N÷áži?«v¶®“ÙYÜ…tmêšïŽïö„ö Ÿs?××kÙÛù«Þ¯ÇÏ+¯½ q¡ì"åbáÅ•Ky—.§_ž»we²ocß“«!Wï÷ûö]ó¸vãºóõ«v—nXÝ8Óâæ¹[Ü[Ý·Mow š vþfò[çéP׳;=wÍïö¯¾8b3råžã½ë÷Ýîß];:ü ðÁñð1áÃè‡3’½yœýxéÉöqüxñSѧ•Ïäž5þ®õ{‡ÐTxaÂqbð¹ÿó'“¼ÉWdþ±ÕšK>‚£n9ǦjXÕ Ø.¥×eM!Ù×¹:23hÏ—,o믊·\—î·:Ói]õ›Ÿ)0íÎ8¨g®«“m>àÝ›]«†böhaeÔoA49êYãšÂíqåw¦À„—[rrÕ9vȳ(%à wϾµv ¾ŒÅî4V“½Ì8ýЄÜ^æÕ§­*0oiË‹-Mz°ºàƒŒ²_XX` RÆ‚›‹²Ñ¸­~¹¥ìYPT£½ÝãÅ_“ñ:`!3¢ûÓdAn"Ú4#ºUNãKðAv{ÜŸÏ瘷ˆ5 4ðÞU8ŽÁ<¯þï¶ÿ­©ÇûßÊ2elª´%˜ÿ[T10%O§‘Ÿ}°¯^~ò2cOïž2˜·´ì>½{ªâlÎ+ï9Sø!Î^`-çÌŠë€ýî­®.pÆ‚¸ì ĵ¢Ü_PÞm}@€ê¬ž=3"º âÊsWp$/>xÑ~ñüõó}k²·Õ 8ÚÕ5=éðQ¨ \×Eas¬4Ì3¯¥m–<²ü§þ‘fÌDáïŠk¾Ã»ðï›û‚/ò}ÁKÚÇh fë—ܲ8Fêõå ìxÁ×1Œ8pêõ%&àåÕ@mC8Œ ñפÝ2:røàxÒÇ‚û>ñÔ(ñ¾`}^ଌĵ:ÛÖßåž!Z…@uÒ¯®ýÙ4¨Ž^0W”?lƒ‘ÓY¢Š{ÉptúKÉajÇô¥Uq6çye„,¥UqÎd› PÕ (þ€ÃH¼/¸ä/ÀB‡ ÖV®Ê0#º@…W(D¶ÑÍž]ãÁtëÓéÔýôùíwwª pÕô‚Ïçó`5×"ª\%¿ßÄ´õJ¨ÔîðÛwª\%¿ß y°ŸÃGa} ¨ÿ ·wÔ(\š&ÝX3# %“Ó!»ˆ;¦žÓGK«âlÎ+ò}ÁvLêÚw%7ô€]#õú´+Ž| Œ8pêõU3#¸™ ñ×H@ 2 Ä%¸$ —â’€@\ˆ+Ê:+¶Á†«â^ò( pI/ˆKqI@ . Ä%¸$ —â’€@\+pfŸÌägÜ‘6 —â’€@\ˆKqI@ŠÖŸWpêYõƒ0å¿ï]Xätú~-Ë«é&þXHã©ßšÖö&žuYHrmƒ®Ë©óù<šhývß ÔúÿÔâ#™äfÚhýÔ›zÂ5"(שF›}z¸d$)ÂeGxпîG„#i$ e™I±Ëoµ\í8à HF‡Œ¥Û­øã¹ü/­¿É$pÚp{Æ‚¸$ EÐä.$ EÜ…â2¼N7©Í7Ó^$éOÈõÔ\­½¡Õ½YpƒÛÃó6>´Wpµö05Ósð‰ÛãÓ§ Ü_׸ë:XS Ž¼ m@ \—+ãæm j®àòÜÆTC/{GXp5ËÀ®F›ƒ{Ãs}%à:‚²›Z |þ[Yèe¹å¤ ”å–=- Ä%¸$ W ±à¿}e*_ºÿ4â xáçÌïÿðêÞU˜ó«ßÞ» Wüø§ç(«ä?¼~¸w€²<¾y Ôüáÿ~œ±´¿ÿëßK{ùóþøå_3˜½zÛ€ûêTøAQ½-~ô?7¡zÁ‡_ü¬ýâñO9Pv¿øYá5d =dÝ ÖLœc_üæ£îë_ÿî/·¨ÙwÖ°iš/~óQ±uË(èHÈãŸþ:ø]—¦äº1ªÿ«=ÁfžüëßýåÆùÒ¬©a?£ocÕo/£pmÀZV…t@ƒ¶IÿDœTWO°ý²opÎl©aöÌU·¼‚¶!A¿mÒý‰šú›zõoíNÝÌ®’Ûk¡n‡jÚă{4 ›ïÚ2ƒ†Oymóê.WÙ.k8UÉB궷‘̲ÞT!{gS[þÌtœAÿ¶˜:—úMÅîÁþhæÍRæj »¯³¶^µ¤n{IÀvq®-Quƒµ­7–ã¿3ÒèµÚ®s×=gð„¶Ø]eÛ5 Ójx…Ôm¼¼=Šš#]KðÕRÏ F›u ¿Û·wäMUcy [Ùë™±n Då_N)¹†…Ôm23î‰q› M<¸±>ÀóJ®a!u› ÎÒ¾qÿ×&A®ûj®Ì†Ù‚7Ž?ÁmIÃëó“C°¨Ñ ‚µi¸hFtBÖw¹틦išrïû¢iš&æ5ÖŠêÍÖáýûFqé=!«B°Þøk_  µ”Õ±²ˆvÎìZ½å·-Ï·WÜ·0«Ž?à–_ <Ô}Á@Xi#Âëðj3P¸vý®í7®¬¾'dfâ±øn&Ë}{)wņ øª“x_ð ÅP£ô•º@¥6g_DkW¿ü°Üé£MñÕëo¸üøæqÉ·–üÈàñu[X±Åßzõ«‹>(…Ÿ3…W¯Ýo¾šüÚèáõƒý‚“™E´s¦äê5Mó£üùñÍ£õ¸$à¸nµîî¿©gÞ~gU  8âîK÷3wêk`»ˆ 8hÙõ3ejSç)·ßWÈ(è}Áý½\»ý3—lÞ|›Í .w2¼üØ.hθº±tû­?8€ˆ½àQË7–ž¿2TD6ÍDïr0|óJ»‹Ø Þ²õõ僺ÃP/mÀ¦)fófàÆ$`Óc…¨$ WÄë€@í–섹„ª·v£ôŽeUJÀ÷ÜkÞßMÓ4åN9ü¢išf|{¬£+ü ¨ÞWÓP¾gù p¨å$­šE´sf×ê-_–i¾(£Y~)PG6,úäºÕŒh . Ä%¸],yÿæßÿ¡ô͹í˜^šÂϙ«×|·czsŽáÕ§¯î]…9…Wï\C ³+ü-«ÞFm õ‚¸$ ×é|q·ÓÃë‡,—œ€ch3AˆKq¥$àé4²PÒèƒ%[1°Ÿqò8½à"ô'è–©üfWø[V½,²%`1eàHVô‚»Œ;¢Ì¡Ž-¥ (þ 8›2à[hõÊ3c ’1™k~Î^½Ž‘ . ĵº<ÚÕ5=¨‘6 T©– w…“€E(ÿl.¿†Ù|ËÕòÛË9 P÷„qå a£ò'O•_Ãì¾åŒjùíÝÿ:`Û­>}§ÿàå×@«–ˆ)\ân™—©´¥mØÝh|Y¬{ý±BêeÆÏç¶I(þ€ý¤÷³iŽj‚Ù‹è$^¼AÓLü{»ÿHȨ¶ÿ«ìju^Y´_ïÔ*‚À~Ò¯‚©ÿÏUiØr÷õèƒyÚ ¸÷„q1à.ì˜Äå: —Õ±€¸ì˜ÄeÇt .;¦q â’€E(c­òk˜]áoYõ²°c:—6 —âZ=³¯Ò±à‡×öÖ:m&¸'ˆËêX@\®qåÙ1]èQžRÛÝÍsÔ'¨ò§–_Ãì ˪—Eb°y©˜L•¸©ÕmÀ˽1uJå ‚‡p6eÀ·PbÎlL6Ö*üœ)¼z”ù€mÞI= vÙÖˆÖª“8,ï€pOT©– w…“€E(ÿl.¿†Ù|ËÕòÛ“€@\é÷„Tço_•˜èÐÿkßÿÖÔãýoe™3±¶bË¿UöA)üœ)¼zMÓ¼jVH­ÔÃë‡þïÇ üû¿þ±´—?ÿè_þ5cÙ«÷㟆8OúþöÕ©äƒò£üÙ)½Åþñçœ+¤·ô«ßÞ» G ¸$ —â’€@\ˆKqI@ ®@÷„‡ñÅo>j¿øõïþ²¥ T¬‹Â&) % p i(¤á¿ùh4%àÐÃ/~vûý¢išæ¯»ÐMÓ4…¯ó±“Šê-ÒOà 8ôø§E Z„ZHÃÚ0YD;gv­ÞL¨µô‚XŒ„±˜ „³1ø:î â’€@\ˆKqI@ . Äh6Ì/?,åQ…W/a÷ñúwLõ«‹>(…Ÿ3…Wï÷°cúuÝB´‘»â²ˆvΔ\½ÆŽéˆKqI@ ®Ã&àésEO`…c& ø–¨#O=Íw7sA¦øÕ‘€MÓœÏç~®NQf2û©òžóùÜ5-A±„Þ1½m¶ÿyY¸ÓõQ‡Ý1½‹¼~öÍç ”„Ȧa xKvL¿dÇôÕpΔR½6 GsPÙ1ý’µa²ˆvÎÜwÇôfYGXDZv0D‹8ÎŽé[I@ ®­ h ^›°»;-Wmn)=»»Ó„ P©ÄÜœ+¥¬20µ6AÉkôwÈhšæñÍãêÌš¹’C`Àl ®u xµ‰ç‚ P‘ ¸°‡+Z,MÀUø„ P…E ˜0¾!ò]OÀäá]!îJnœÝ"’Í%`–É}B(ÖdšÛ Þ-fDkeOÀì @!h$éf¢ja,n/àFvJš ©U8I”ÏÊ@\ˆKqI@ ®‘‘ZÙÖ|ó˜·@úþöUÓ­~üS#„e‘€s^~ò2WQOïžrÅ”?~ù׌¥ýý_ÿÎXZÓ4/þQÞÙN/ˆKqI@ . Ä%¸$ —¬LÚ"c£?e½²ZœN§î` ¾¾_¥®(¹n}*Ð_lÉÂK™]‡©¿¨Ýãݧ¢¿ºíàëÁ3çË¡í¡iåè×Í‚£?SòàÇ ^®¿Ôqûœþ™yua½´zîJV`4È.¿zu§ìà™åœŽ\jC§=(£_·ýâµGðòGf éþ9õœ©SkæGVÕv?ðþ.Ûwƒ“/Ë«ŒžÐSõ‚·4\5Õ6Ìxç‹Z~juOè7*sU2¼¿…m·ËÍ“~tÏ–Z®[ÒíS`p¸§N‰µ›ÚhôÁBö’€=óž@wÿSÌ}íšG ÿ„ß…¼¿ùNÐòSsù>VºÀEÙØ n.(®ž0ƒçLý‰]U‡©rFŸ0zañ.$àý-¼n½üñ„Ù…üAŽiK\G,/sf,xm’Ÿ0¹= X©ócôñ„5ú 7mK¾^RæüO]}•„³«œéfDqI@ . ¤«ýŠâ’€@\ˆKq™8Ç&¿uùå‡?»wæØ.½@+fc/Ÿº}÷;]–Ð â’€@\ˆKqI@ . Ä%¸$ —âró—6 —â’€@\ˆKqI@ .kDsÝ??üI÷õÿ|ùyËì üç‡?žåuO§S÷u7÷ërß´e}»Ÿ}•%?¸åÕÙNrE?˜ú‘´][ìeðe|Ý~²t!Õ~‘1tF_eþùY^—íô‚£û`·š‹OZ–Þ|µßýç‡?iÿ»|~÷àeso¾•—%|»ßÀLpû¯ñê.“÷|>‹Å»€‡Ò~’»ØòÏÕL¾ ‚éjo´Í²îiÿóåÿ]Þ=8|k_wÔ kÚöÿ?jᯱËÍÑWéž“PmnF/øP®~ª·¾°6Uý¼kCp¾#|õu/óeê=®J¢ä_£¼«Žäº6ƒrt¦Æ=—ÿf^wá˜ÃÂgnq›W!; xpW;Â]îLÅÍöøí÷Ëï7 »G²Än®`ºü5^^:¼¼¶˜·ìAF1õ \ÒÍÕú[[ÎÕç/éçž©Ò¦^×4—ÂI@挎çfì’wÐ ½úºóá2:’›=nó*ìĨã+¡ru²ËÝ]ý-íúk,áÅd6 —¿<WBã¢ü`kæwU¯‘=8®@\zÁ@\ð?ú÷‡tù—€ßko‚pH£mó›Æun¨ÐÔÌ©óhûF¨U·„O—nk×j”€Àõ×yœyšlëSÂ!ô€]Ûpþ—ü^ûçÂA¨QÚ'×8ˆ„Œ6 p‰»ýiaiqI@ . Ä%¸¢ÜüðúáÞU€@Þ~ööÕ§¯î]‹+ß½{z|󘫴Z<¼~ˆsˆ³øö³·¯_ü0WiMÓ¼yþ&{^0YÄ|ñÁ‹ö¿ìÅæ--o=wz×ÌËø Ï~³ŸcMîOA^?øíÈ¥°@½à¾ç¯Ÿ›¦yñÁ‹ö‹íö8ð]%·Õ§ß5óò&ËG0ã9–±œ=ŒÆ_0wúü?ýœ÷ðç­¤ÈÛ¨pÛc}5ÚÇgΊÁ·æ \rWxµÌ„·<ó~ôíçÝãÝ׫¬-ðÛÏGCðP½àãíõ‘7¯ÛþŽ4Lóüõs÷«ë>êŸ]™ ¼z×8ÿ„¼oùÛÏÛlêÂ(9þ–8Õè8Z°½Íù‹\µgÛ-A!˜Ýeã« ‹&é>ÕšK>‚£n9ǦjXˆ.›^ópÔqpaê Úó%ËÛú«â-×¥û­ÎtZWýæg L;‚3ꙫÀå66ç l[…KÊ?T/x¹Âûƒ]Û!×8]›ûÆ‚3Ê~ò̘p—\æ[UfÆ·œ7û¶xœ6àr;_Æbw«É^fƒ~hÂn/óêÓV˜·´«Åö[gËK›qµÀËÇGŸyœ6 ½>¸¥ì}ˆPîÚ^åhmÀ6k¡|7î¬Àíñ×êí…*457p‰C% À y8ÿäêçÍ-´pC¥pTƒ4|óüM¬Õ±–X~¥ÃêX‡gu¬-^|ð"ïXÇ`u¬…·|4qfD´–‡¯ŽÀl Šå÷½Í;ÎŒh ”,×% WbVâs;P£”Í;!Tguv7Þ™J Øê{BúKô%-|Yf;¦g4˜Ýýn/gM_ýV–‰ÖS¯Ò?è uf‹ZvLÏ“€å¯È’w~ãžÜ²ÅÞ÷„d)0宸v¾hc ƒ}³r•fÕR¸¯l3¢Knf‘kgÕ…-^qûÖ6À¥”ì÷|› ³/m#Ô\Ê/(Gz°ºàë4©ºÍ­ç£p~gÕUηæúO} l—’€ƒÖ_ùà kõ7{ÍÒý¼,ÐÖ½P‚lwÅÕ>$Ò£.¤¶´¿¦ ´+”cuð²Åw€MÚ惩߂koE8U „rä ®+—l„ºªõ—½@à6òô‚+Š¿Q%ï¬ ìgëŒèc(ygU`?‰ó/ðHcÁ@FÝ®Fyï“Ë"ñ: ¼ÖZµ“ïmØ'¸ƒBÒ0qI«Ñë€%7 —¯óaª ÜÑT–²6LSÿÈï¼…ƒVÇ:<«cmÑ_kÉîæ÷j¦ÏˆžZ*`¡„໚§«ÊÜ´2B¿1(%Òš{o?{ûvÙ3W =G¼'¸±ÜµÃ&íóÛÿÏ??}Fô`•@€Q[â¯Ë¾WŸ¾Z>Ò¾b—ƒ3ØÔÔôöÖæ×›çoÒ~pþºaúÑÝ×®;Ù>F<_B†•ÄP©¥mÀ%½]QÔei¶ö´ vFvLçªvLûÙÂ)1ÿ1_‡Ç7Í9†WŸ¾*¹À«·Gåsˆ·hš]²%!4§Šj߯•€jœß¿m`ð­öŸóCµÙVFH+çf^?¼³îe§)"Ww¦ ÁÑç·™2‡¸XS¹6x|Ik1ý:àÌ‹EKCà6ºÙ03é6h¥ÍÇ‘ù€@M.»¡ƒoÍ?2²>àÚÈëj-LªM³+ºhF4ÜRù‹æ¥Ïˆ8ךä]§J¨FöÛÏ´Êd¼Â¨ ˜ÈtÙÃsˆ#€@Mò1K@ ®M ˜¼n5@š¼ƒ!÷ŸµØÎœÜÄןN™ej¥ÑpKû͈Îu#F› ‰cÁ oÊ[^ÚÔ½~åÏ,nfje–å¡‘’€Ù§äŒÞÏ\Ñ-w@¥ÒWÇÚ[¿k ÐìÐ$Jìß i&þ€½:¦½ü§ì*e}ÀÓét³Šöå\ „˜ÖîûÑÚq$df,&-§ú?Õ}=ú @^VFʵvôµ¬’ÄUèHÀ H@ ®¥½ànDæê6¹¤Í‡³J>ÀuKÛ€£óTvu83+$,&  Tc>㦖˜š!ãXÛI•€ÀÑ,oJÀD6;<‡¸FkÇŽ% p4û®}©Ÿ»‹]M­•0y0çÊ—›X´}—¢½_ÊÔÝ”‘ëÖŒiÕØ+nk×–P–øÛ´WÀ½dÌÖ”‘Ѷœ=PmÀ~ÆÉ;àô‚rU³6L•Žé²‡çG°¢ Øß®)ZÞwQâÚ0⸋ìkäï–9õòAKÈçÚŠIDATD›]íý6ñÞò1Þïü:ö—Ü\„ì'ß1Îæ#qˆoÉÚ0@DkÇŽóìlz P;%‡5ÕäJhŠIÀcŠ9·+”˜‡¸?'ïÒà9K¬NÀ™¥ðé²Ù•_Ãìâ2ÍÜÚ)ÙîŠ 5¸£ŒiåîŽZþ¾Ua0£ÿ»íkêñþ·²Lé˜*mIæ¿E²·Ÿ½}õé«{×âŠÇ7‰ xÙó-`á!C›qY¤ß×Ï»˜ƒ!À-̾åÍÏÄ‘Í=àŽÚŒtxöûÌ*3•qýX\hu^®?“°&@ RfDw_ô§CŸ ÜF¿xÙ\;­ ÔdI'wyG8¥|ÕÚ2ûÁo2Î_-?ùgµ[hs0K^–Óýsíppâl˜Ñ‹}ZÀUýðZ{ÙîÛÏÙ×=’–­Ùî X+! GG?’Suõ­l•î”´üϼػ2À¨™ðZ~_ðBé÷7e‡ÝFÏ_?/yš•ÏÊ[ôWF˜ïŸæ]@a­lë,´=õFS5¡Ø”’ÚiÏn{ÈÒÜ›iQ&Üœ¾6LuKw‘±Ÿ;3âÑŸ³ûl€{ ¸„eš\ûÔÈ=!Àq¬m J@ ®¥½ànÀJgD¬Öª1ßÉM¸5xi°kßièwÔÞÜ…Ýe&î{_p¥lƒ‘Ó¹j§Óß~ö6᧦j’¾czu옾±4÷oTø!Î^àN;¦§-5Z“M+#Ü^öeŒ„qI@ \ÛWÕŸ/A/(Za ã*KÒó° hý.8€n²ËÛÏÞ¾í=2cÕ̘c& LàH¾ýüñÍó7íT˜…ýâ…c&u$`?ÑÚ;óºÿ>þî= FK–Ý_;X\G6ßÝ‹Ò_ŸU'ª°jÉÒ%2–VMöõ›xƒ¶žX„mÙ#xWU&`¿,ò .E¥a} ¨ÿ ‡q÷4Œ’&vL‡*Lå`Y;¦˜ÓGK³2ÂF…âìÚ1`Òݯ¶$ p#…¤^ŸöU`ðu¬ Ä%¸$ —â’€@\ˆKqº/øÞU€@vÚ1=¯@;¦\ú苨“ûh?IEND®B`‚‰PNG  IHDR«‚à8aY 6iCCPdefault_rgb.iccxœ•‘gP”‡†Ï÷}Û ì²tXz“*eé½I¯¢Kï,K±!b"Šˆ4E €£R$VD±± Y$(1ETPòÃ;çÞñÇ}~=óÎ;眙@@ERR|?{NHh¾!’—™nçãã ßåý( ÷V}¿ó](Ñ1™<X€|^:_€ä€fŽ ]€VTRº9 ,~HhrXq_}XQ_}Xü?@¢Å}ãQßøö(Ûñ ±1¹ÿ´XAN$?†“éçbÏqspàøðÓb’c¾9øÿ*AL®À!-}?!.^ÀùŸ¡F††ðï/Þú{ð¿ÿßôÒ¸ ؾ³¨j€î]RÿÍT0 ºîð²øÙ_3(ÀHƒ¨€&肘%Ø‚¸ƒ7@(lÄC ð!òaA 샃PõÐ-Ч¡Îø·á.ŒÂ¼‚yxK‚:ÂD¤ED ÑAŒ.b8!žˆŠD qH*’…ä#;‘¤©Aäär¹‰ # dùù„b( e¡ò¨:ªrQ;Ô @×£qhš‡¢{Ñ*´=‰v¡WÐÛè(*D_¡ `TŒ)aºsÀ¼±0,ãc[±b¬kÄÚ±^l»‡ ±9ì#Ž€câ88]œ%Έãá2p[q¥¸Ü \®w7›Ç}ÁÓñrx¼Þ ‚Ãçà‹ð•øf|'þ~?…O Ø ‚Á•JH$l&”:— ÄI‘H”&ê­ˆÞÄH¢€XD¬&ž$^"Ž§ˆHT’"ɈäL #¥’ H•¤VÒEÒiš´D%«‘-ÈÞähò&r¹‰ÜK¾Cž"/QÄ(+J%‘²ƒREi§\£ŒSÞR©Teª9Õ—š@ÝN­¢ž¢Þ NP?ÒÄiÚ4Z8-‹¶—vœv™öˆö–N§«Óméat}/½…~•þŒþA„)¢'â&-²M¤V¤KdDä5ƒÌPcØ160ò•Œ3Œ;Œ9Q²¨º¨ƒh¤èVÑZÑs¢c¢ bL1C1o±±R±V±›b3âDquq'ñhñBñcâWÅ'™S…éÀä1w2›˜×˜S,KƒåÆJd•°~f ±æ%Ä%Œ%‚$r%j%.HÙ[íÆNf—±O³°?IÊKÚIÆHî‘l—‘\”’•²•Š‘*–ê•ú$Í‘v’N’Þ/Ý-ýT'£-ã+“#sDæšÌœ,KÖR–'[,{Zö±*§-ç'·Yî˜Ü Ü‚¼‚¼‹|º|µüUù9¶‚­B¢B…ÂE…YE¦¢µb‚b…â%Å— Ž'™SÅéçÌ+É)¹*e)5( )-)k(*(w(?U¡¨pUbU*TúTæUU½TóUÛT«‘Õ¸jñj‡ÔÔÕ5ÔƒÕw«w«ÏhHi¸iäi´iŒkÒ5m4345ïk´¸ZIZ‡µîj£Ú&ÚñÚµÚwtPSÃ:ëð«ÌW¥®j\5¦KÓµÓÍÖmÓÐcëyêèuë½ÖWÕÓ߯? ÿÅÀÄ Ù Éà‰¡¸¡»aa¯áßFÚF<£Z£û«é«Wo[ݳú±ŽqŒñã‡&L/“Ý&}&ŸMÍLù¦í¦³fªffufc\ׇ[ʽaŽ7·7ßf~Þü£…©…Àâ´Å_–º–I–­–3k4ÖĬiZ3i¥liÕ`%´æXGXµÚ(ÙDÚ4Ú<·U±¶m¶¶Ó²K´;i÷ÚÞÀžoßi¿è`á°Åá²#æèâXì8ä$îèTãôÌYÙ9ιÍyÞÅÄe³ËeW¼«‡ë~×17y7ž[‹Û¼»™û÷~š‡¿GÇsOmO¾g¯êåîuÀk|­ÚÚÔµÝÞàíæ}Àû©†O†Ï¯¾_ßZß~†~ù~þLÿþ­þïìÊžjfö1‚ƒZ‚ƒƒËƒ…!ú![Bn‡Ê„&„ö„šÃÖ9­;¸n*Ü$¼(üÁzõ¹ëonÙ¼áÂFÆÆÈg"ðÁ­Ë‘Þ‘‘ QnQuQó<Þ!Þ«hÛèŠèÙ«˜ò˜éX«ØòØ™8«¸q³ñ6ñ•ñs  5 o]듼“Ž'­$'w¤R"RÎ¥Š§&¥ö§)¤å¦ §ë¤¥ 3,2fÌó=øÍ™HæúÌK.ÌÒÌÚ•5‘m]›ý!'(çL®Xnjîà&íM{6Mç9çý´·™·¹/_)GþÄ»- [‘­Q[û¶©l+Ü6µÝeû‰”I;~+0((/x·3xgo¡|áöÂÉ].»ÚŠDŠøEc»-w×ÿ€û!ᇡ=«÷TïùR]|«Ä ¤²d¹”WzëGë~\Ù»w¨Ì´ìÈ>¾Ô}öÛì?Q.VžW>yÀë@W§¢¸âÝÁoVWÖ¢Ê:$¬ò¬ê©V­ÞW½\_3Zk_ÛQ'W·§nñpôá‘#¶GÚëåëKê?M8ú°Á¥¡«Q½±òáXö±MAM?qji–i.iþ|<õ¸ð„߉þ³––V¹Ö²6´-«mödøÉ»?;þÜÓ®ÛÞÐÁî(9§²N½ü%â—§=N÷áži?«v¶®“ÙYÜ…tmêšïŽïö„ö Ÿs?××kÙÛù«Þ¯ÇÏ+¯½ q¡ì"åbáÅ•Ky—.§_ž»we²ocß“«!Wï÷ûö]ó¸vãºóõ«v—nXÝ8Óâæ¹[Ü[Ý·Mow š vþfò[çéP׳;=wÍïö¯¾8b3råžã½ë÷Ýîß];:ü ðÁñð1áÃè‡3’½yœýxéÉöqüxñSѧ•Ïäž5þ®õ{‡ÐTxaÂqbð¹ÿó'“¼ÉWdþ±Ù•b4ŽÚЋ9´›?Rmk¯‰ŽvÏ9gRlmR[ýØó6škãí^Á+ˆiÒÙ‹ôW9ÃCâÅ`ب"Úâbˆ"O¸~ó‚ˆ@µ:!ç)c?Ÿ\¦Hm7"¤ÉE&VáÜZ'Ìi°1m0ú€p^³0J¬à¡Œ¬¨ §Å‘×>/òﱧ…Ô,´H-s æˆõyjJ–PQ1;í¤:¢¿b#J ¡X# ™ ýEýDD§¸W·™öº-ü¿!O™ú×›ØãÓb†¥íé0£(u›h@\îøµ~3ËhvkAÑcVܲQ|ÙA¥„4¢I­ëx“^Ó·•…t׬~©†@ˆy½Ã‹ôZ丞.Ń.Ì,–Õn%Z¸ýüBj±[6+¢™˜xSË‘½T²ÂL†Å2ÆÜ w9š,µ÷â¥þ4ûÜe¾¦r༺ÿÀ†„—é²n“û¼¸N”ˆMÌï•MwfÀ3ôþ`ý7¢õl{Þf\§$‘ðPÉ”QTíåè)è6ÔÖV2†ß…>V`†4?óåôjX7tî‚Kùl´$™Q±‘Ïõà±C(¤˜©ŠŒ- Á‡­ íD°ï·‘ùÑ6ä^Ž î›ä#ÈŠI®Éú‰Ÿ(kvdÆ Ýî¿!ÆÄFåÛ~ËC÷6®øÔDW³(ÄQª„+uYh@_¥åõ›éXšïÒ'› 4kõi0ãÒõjÝVh~ÜÂ-Ô vj¦ÐöòCw6(ý·BS‚Ú—rY>ŸwL:׶Ù/°ŸY&·0ª.Ï4¼¡ 3‘z¬›ÛpÛǺkc^®¿D>´sÈ]9(æúè¹µw~¥KÊG8î'Ãküjœ6—å˜+íØ7¬¢ÿ& ZÁ¯ožzLlËP:?0¢öYx8;ŠL2«œÁA«V_ä¿*A2¢Žá„Oð˜“ü³VUüb:øÝ™TyCø±U øĪ# e~FH&vöy{Eë7Ò‘§œ'Å?ÂAseµ:µÛŶӄîJï¯^žCÊ&sOë™]ìÜ µLròRìjµ_¨$Šy"wì$dT’ä¸ãàyCÃ^-òå<Ö[`Œg/¾n 7é…Šäµi®^a¶ Úìl¢F4X~{ož>®Ü7‡tºa8â-Ž!›d¶PÁ²ÃT±ê/ì]%2¬LÿNMâ «$nê/5.`½ ©ËdL$Ê0^1ï.Ò3sXö‰ìD²°7™þ/Þ¢‡éfØ3«¤O‰Ì"#B)gЩ;B^V¤Þñ>e ‹mDFëé¾=4Æ1Ò騤}PÄ®«2RÏ®5Ûž;ÁN®°›M¹Z¤¶IçÐìtohä*Þ|³1L¦ûåÃm´ïÚ{Þ–/¡¹Z-Çût.TäËì”#â-¾Û–*Â3¹’磸`׸C¤µcX>ºï¾ °&êÊáàia4¯Î% -;$3/¸Á3áFo.|Mžu>ÅjM)‹ÃÂOá³éúVGOÕT2a¯evú:À½gü‰µýüßÿfõ&¢5‚3ÓLY-3Kc$4®»3ÍØüž,"Fçzýç»uC­hb1Ö à%rPì ån b¯Í4ƒFÀ?õoŽË·Z…MIEND®B`‚fftw-3.3.3/doc/equation-idft.png0000644000175000017500000000206712054410050013427 00000000000000‰PNG  IHDR¡:]Ù“0PLTE³³³¨¨¨œœœ„„„xxxlll```TTTHHH<<<000$$$ êYËžtRNS@æØfµIDATxœíWMhÔ@~ÙMÒÙÍþ!A£‡RÐbzÒ›öPAhAO‚š[»‚QÁ©m1¶Úî­‚WD¨Y a/‹'AíADoÁ?*J×7™Ìv›MÛ©ÝŠ‚dæÍË7_^æçe Å/K Â*Õ¦wŒzË•ÑŠ@â:3*mx½Æø¦ôF"½e.M¡*üªÀ§xø>H7 ÖAµME× Úê²¢œ‘ðT6N$§a™2ÙܨlPöejûIŽƒnB F/ä 6ïzòlä¡rÃ\Ñf<ü zÝ »ÖÀLTìep$»ÿ’îQX5Ëë ·1e=d눰o‹bЈk³»Ôœ ’š¸Ž.j ê¿ÔN0bö-&³nÖ†AKé†æ)¶ ~a)£<ŽäV`¤„ÇÓ¿2ŒÌBѶ8M`‹sNJŽ=Vd˜7·ПDL–(³6µc’D’¶äˆñ–¾MÂ=ái«\ßv~Nœv’ä›Ù mé~\<ºhäæ«åQ\ïlI é1HƒX2ORÝ~²Óà>ÄRæ¢E\J ˜ÖYÎøé—þ¶±û½ pLÚ[ÂÆD: @ŠÈ4ªÕ¤ 0bÂ3Ãzgj|•[J7³BQóŸmaYh}ý¨ù<¾Išµã0OJvä i40 ·Æ"²y¥[ÂáDóºñ:ÈŒŽõyÀØ· 8Ûå” q´Ï7dˆ8x†ÑÁë¼­ÿØmJ¼½ìÂß=—†¡Î§⺮Ó6Åtñ°o*m“ä¿€ºÀ—–[GyÝÝÔ7y{^WÖVìYæP™2CwÈbJºöؾ#hâÆÊܼ°LáâPÇ¡¥À`ü»ËΙŸ¤>®ÄÏðð4\)cnƒ†bô³ö=´§r0RYøÜÔzÚôŸ2s&àQþ«[LÒ¤Hç¨8‹eü5:RµË?"ù<'R§+Š™w¼ÖæüˆöÆrñ¤&þ Ì*§‰óî¹!îdËáôœ¬X4N;¼N8Âóáù–w¿a³ž¿½ð}K2MÈ ØíRÜtñJ¤…¶A§†Ó&©áéEƒRÿ-P=m½Ô®Yà†µ.O"ü-ø+½ÒjË|fˆIEND®B`‚fftw-3.3.3/doc/equation-redft00.png0000644000175000017500000000341012054410050013736 00000000000000‰PNG  IHDR\:yåÌz0PLTE³³³¨¨¨œœœ„„„xxxlll```TTTHHH<<<000$$$ êYËžtRNS@æØf†IDATxœíYmˆTU~æÞ;wÎÎÙ{!!?²nСÞ(hA›î…Å\HK0jú#e¢CIý9¶~\Ûµ]úi’Jû“CYX5”Š`ņ˜A?œ>)ÉÑT¤Âé=ç~ìÙ5gu†qcŸ…¹÷žóžç}Î{ÏyÏ9wV á ,î‹ôB[jˆÂ"£{ê>¼¡e-7)a_Kd5ÇLºq DVD öj(r ƒ´ñ1lº˜2b}òQ›œŽ·Sk66ë\jtb¡ÉP*vÁb$™1Yd (]H,îqûîÔ²øø*ÌÑtqMMEصŠèúÐoÞÖÑs-õ¹6që$Õz_4x%¸œ~âýõ ”r2I‰Â¬êB·Q‚ª3B„˜!IóRIåíÁ¥XsçЋl䈛\ÿ î³])nsŒ²\¹‹ 3nàL… B²å·žÅàŠŽ÷J=5kW˜N× Öç5o¬Ãh“ŒN ᶞs…È$¼õ¼ŽÓzÎë ~—L̶ÚYijYUůB$nBåvøtÉ/ªþµz‰ceÔorÖC³Ò,G¶ W³fÍ–¬f9¶Xš4«áëUMäŠðppæm#^áÿÅ@Øç¿°Ç[™+ØHMFÕYëÀe¨V°Êã"ˆü`n‹ô¯h[ðŸü »ÛïDŸÓÁlÅÑ Ñrj™ô:»‘öS€4ŠòN³Pȃ"ã4…üÛIp` ‰¦yØíY O·sõЬ«Û\ x¹å”z•#šç­Ä ¥€”Ýz^´)¶74 ‹öÑ0¹V>EþÚÁÂTS,yñÔ+7Œr—n9Eû-wÅÈ"L÷¢$S4ð¥ñø±©XÎ7ØŠä¦Ûtj(D[¯£ ¶f”õ’Š_ZAù΋ …!&Ù¥¥p3°xÜV³ga©àpP¼]Q}K––§£TËð,É› D‹î¾Œ &r¯Jz ¹‘­ŽÊ9ç”ÎÒ¯º;qjwÂÊ+cÅ€—Ž‡ºm†=*€çØvÏarU-Š¥A *9¼×ÅÊ‚l‰Û@õ3†•ÓTÇ5'ÏË/3á@ëíµÀ P45)Yê%…N6Ú‘ÑÓFá‰K?nFõ&çT’)‹ˆhÈî‚›aV W±Nò(†£9I:àÇ0z,âÈŒÇ[£ƒÍ h'`ß…¤MÆ#x#԰ײ¯þCÃN¯"Ü ÓB+S”•ãËŸò¶¼[H¶EoZ›(>¯ÕÚvT™Nƒ¤ÜÞ¹Ìã¼ ¼b<´ƒWLåmà}¡ œ3˜ÁŒÄæ¡é°«Ûµñ¼&œì@‡¥4ƒe8L» o+’£v§µ4ú9´ZÍî°”´ x3Vf-„Û^9WÃmMùWNþìßÌ>×·ë.s¬ÿ»¾Œü5yæPCY.¼ë¡Sðü`Øq9÷è?“g¨ÄÐqÑkÜN]7ú¤¥î†ñ»l2sVï´±°©±2]aÑŽ_óý½UÑN²€ ž8z½Á Bý÷¹iÐùµØÙ!ÎÛ(W(–-²VÌY½S’›)?úé%Ü/oÿp¼kåxK„-`Ý¿L ¨HŸ•#ñƒuÈ›<þˆ³º÷ËÄþe®‡ä™Z­¢ çÐUð¾>ÎæÞ²§A.9%¹Ñ¥>‡ê#ärÞÊ·èýyÛxAÆ‘¤|/B: Éq-ìÆŸø¥ÊÉŸêy± Ø`ƒa®üíiËÃ@šb~O–cù‡T›*g_À”\n åzB.£¸-_ ¾#Ðq»TúHD&¯_öâÓñ`¥x]GƒoÏ÷bÅäÄp0|½Xtéa…_ú@.léóÎ~›äZè*Ã\cÍ-P£-}ð^Wq ¥³Ô˜úg|RøNû©Ã´ÿêƒxlŸü ìÌfÚޗƽ+gÃÚ¾o¶"!?èoðÒ%WËäE!W'wa­V–·D_B^P´µãk­ud‡W;I3¨É÷Pγ#öKg&O½æÓ~ÿ}U½ñàEȬñÙòwQ¼j^|Âdóëü¹ÊJ<Ö û±‘W/Ô=ú}œÿlð¯}´ÑYÌiÿdå“#vîQxÓ­’ß7¯„%ÝÃvÅuIXm58‹9eG›ö{mH|-²Moß®Ú$ÿ–¿ñ s£D§•ü/¡ää%awPÃÔA‰:€ÍÉßA×&r±OÑRYœ&rw;­`Jxx:|]hþëî—&+ #IEND®B`‚fftw-3.3.3/doc/equation-redft01.png0000644000175000017500000000277612054410050013755 00000000000000‰PNG  IHDR :Íï:u0PLTE³³³¨¨¨œœœ„„„xxxlll```TTTHHH<<<000$$$ êYËžtRNS@æØf|IDATxœíY]ŒUþæ§ím§Û•Q'â/„1ë!1¤˜•P£¨cLHD"á˜h®²àà‚»oþ$&Õ·‰ ®öFeC‚5‘@„¨‘‚AÔsïtú³Ý¥À,‘o·3gÎœûÝïž{çÎÙ.pMP€Ž«’iˆk~W¨:À[â’£ÍÈFOÛU×Öe0Ø\6 -ìÑðCï& æa˜Ù£‡ùó)ÃéabÍì~†r†Åh($VSSE[[9—R<¤¦pÀíi’”Ço^zè ¹<’áì]°>÷wõèølîš±(ŽI‡1aqx9¨¨¤–‰e‘´Q†¦Ë°:0;¦GGù—É“ÊÁkÈc(lÉ¡[ÓÆ‹ûÁ)Â;¤›=hâ—2Ö ÏÞLaõØ£Õ¯`ÙÆÏéõ€¯yâÄ™¦Åk?oø›>©NG¸dÔvŽ†äÒª‚V¦ÆíÈ{Qi)rr‡‹ÉätX€Ê ‘MµL%ÌSoÙIضK4ŠRh&¼þ$—A¦ô:uÑÄ€?Ít¹#mpÀÆ|Áእ¯£^æÛ-}Ä-ïJä#Mï2UÁ0 ŽÓœ¯}¨Q7î\ù ›ø3“òïCè­ÄäsÈãQäâj¸aÍ×|‹ÏÁ†ãsˆ:ìféAyc"‡~âuÛqyä]û=ƒjCQßQÝfÐu04ëZg|N‹¶¤]Ti5 ˆmÑl½‰KbÛtaV¤o­ }XquGpØ&m†E5 ê£T—ªq°"]årAìœ $‡Ü+/4µQ±b.Êqïݬ §?ir(¨D¹M­Õ0ÍbŒê íð*ˆ€ç¥ïɘF§ 3ü øïo%Ž ý0”t§ ÞÏ[à¨v&)QeT%-¥h*Ã9^à‹Ld1f7u¤±…gÉÃ,‘Ïo%z“c!d¢‡Vmœû¨‚¢Z+#ßm®ð? ‚4Kƒäè(dÛ3÷bd,oÏËb§cºE΋YÈy™*5Qåëò±9"úÃl¼¾aÿ Ð’c<…RòýØûX ñ÷Âe®‰¿B0;¦†[øŸb'`çm9qŸåXX)ø¾¿}°*Ž%‡ïš±Éª¸…6ÎRÅ’Ka‰§ŒÛóK¼ñÇ+uûØÉS‰3‡º\ÊØnqÚt±y]œ=¬ã^nc;Œ_ºnjâfrB¥BSºG¨ÐµðæLi™k}©KÚ‡ÜÌ9!æäôa=*èvÚ¹N’ð5;dI²g¤©ÈV6Gýå +¤ù›Ûâ ¨x>+yí^Úa‰£Ï:ëòß*ûW»Œ'Î45¬šÂ£@ºäÿ0SÅ,ë sáÔ¶ÝB…NI딧Vw¯ýX|á»cÒß*|_…·–\²»Ãð4M”-:0×þüªÅx!CÒÜã rxÍ>ÙyD"T¤J!%Ì.àC¡‚U(x±(ž©¶/—¿/&öàëŽaøx"LÍ©–ˆ0ìÎOI……tæz뮑o3h˜¬Š-ÀÞtP˜‘‹¥M2VfªªMFh<´ ~ÿyg&^úìÌ(&Qn,m4*⬙†Ÿ(æèkã“ß±©u㩪AM†8–ZêyvØi“è;O{Øß$»Â4yêgDò©u¿s¶›È®K6»ÓùnÏŸøÉR×e¸:•c¯S®:J¾Ió±€²¯eb¤·óËâ9/<§½È“ÛÙ¥ì„<=(¼³UŽ¹æ0·Mv¤7ìr¸çGñåwœüãêÚÍ/òÜAª¸A záùöªˆ0¯ÿ¼ºZ°…Åû媨 ¬ÁË´¿ƒUá|6ÈÞ#<5_ŽÞ°øç`òøb>ÇIEND®B`‚fftw-3.3.3/doc/equation-redft10.png0000644000175000017500000000265712054410050013753 00000000000000‰PNG  IHDRã:Ç–Ã0PLTE³³³¨¨¨œœœ„„„xxxlll```TTTHHH<<<000$$$ êYËžtRNS@æØf-IDATxœíXMˆEþúgzj¦w§[ôøñà%¤e=ˆ"ttñv£b;’Á ’C(e²IÚltöè€?1¦‰!DçA“%Ñ $dÅÃŽrPp;n"þf|Õ?³3“dp;Ù A¿eºª_õ{ß«Wo«^7°@HPTX€â2½ ê À¬!Y¬1‚n•/”äÊ`Iç”ñ‘ªß–‰ŒéBÚê¦ö®’Íî7h`ôòr†Z‰Á§ÉÙ²¥<Ì}•âVÀ‹8Ôõ ›!Ôi6î¿tJ‹¬ÏØë}Gû£ ~YyÓ¤‹-B0!né! hD÷¡Yz´¦“¸yn|œçÄ3&¸3mRx·ªEÒàPH6·¡R-Lòo|äüÐP@."âý6ü¸=Ò“. ê}Üw\Omùš@N£”yýSàŠP”4ZW‰¿é—íˆ<êvŸÇùTƒÂ­H¤ðJ´î§Ë-X)­ZÈ:'´àãSƒeÑ:’©Š§Œ°oâN‡7ÇTÁÀ{s;IòôÇ´÷hC š"Œ]ó¾Ô“ñÃd U%âî¡_žÂØ$ã=¹C"|ÞÉ..hªÒS©4)óôTÉ_¨§â»1a¦Öäi”6±´šl$­&• 7 Ò+“i4Õstq;%tmÀ q•†“ÓGTOG¢nÐÞ=9ÄéÔ O¯@y„ûTä!6 #ÙÚ F7õ°»F˜}HrT›4eSŽLèÐh—33 tYl5¬Lªz­&ÎCàBd„|.Ë Üåû6gaù±æ0dž-ì¶n’¾C'°“Vé÷lØÛjû†Þä9Ÿv0ï0'M*ÂÀ蟸ªÚeCÐm¨Å"²^ˆ=ŗʆ4Rà%¾Œj¢\â­‹i_e09ÄL‡„ÞŽ]>ùGQEì;騷P2âP…[c ²b*5£½4ÄsIçþŽØnÛžHÝH2Ô­$ ‹P†Ä£XŒÁ] ÒõC‰b¿Ð)†ªžÇò©8›”éJ£*-–›F‘Y©øþǃ=|PLïn9uØÀ(Wc žçí åíx´ïNå­çF¹übTzHë[Þx‘ÞªÿìI{E³é·ø^¼“³™³™Ùc=ºI.3—èNÇî—1,ù¼Œ-t>ý$nµ¨.S©F©˜xµÛÒƒNþÁ<3×!Ì7Ù%ßnÊÚ’UQ~z8©œÂ~+S…"¬aØlSÕ7ÇÏh@Aþ\h±Å;ìÿìDmæôÓöÚâ—ÒÁÕã™ÙV«©Ö‰kª—Rw©(Üoâ(rUï$pK§vì )UŠ_¾ÜíþgQ³üâ|´´Šà’âmù0Öüø’Éx)Os§É*Î4zÞ;žgÇ”[í¤ø.6ú)u³U¬»€·CJF>ŽÞ-ê´š@I¨â±(gÛŒJYûXò>QšÈÕaÏŽÛ8(^ô³oÙjTxË×¹j Þ¸äL«vÝ&éó€tæå¨7_âÅímQÒiÉ{„êãÊxÆÚœ›H {¬nÊïâζ{»ßÇ>Ä^÷÷ô¡¼ýT˜Á¯ÍüÞÇ­k‹bg`”ײ6R1ý{}J¬ïù¨¼ˆ`KËw‰•<ŒÚ (1Šh7òÇ7Kûø32(F<1ð¬Y<ü×DÏš fIEND®B`‚fftw-3.3.3/doc/equation-redft11.png0000644000175000017500000000305312054410050013743 00000000000000‰PNG  IHDR:êA»0PLTE³³³¨¨¨œœœ„„„xxxlll```TTTHHH<<<000$$$ êYËžtRNS@æØf©IDATxœíY[hUþæ²³';›Ì}h©—EòÖ‘ø ú2¥ FÚšõ¡*‚²}Œ–º`jA<¶i;ØJò$Þ [Qkì`k)4h UÙB±º*ˆ Ø±±E¼­ÿ93³™Mºiì®ÙPú…™sùÏùþÿüç2ÿžW ¦ÐJÌ4ÁLè€~Ôò² Ç ­pV__±Úy /ɼ/ž,ðŽÎ€¯§H/˜Ei}¶]~g qçÑÆÂsŠƃ>dÈÏjßTM[Åò³BCôdšÅa/­ÎˆI (ó(E ÞLâ¹`Ä4 vÙ/ΣÍ|`à™9¤5[¼Mzv‰ ‡XÞUò«XP”ŠUc8§„–ΉwŒðÑÒÆN£&Sœ¡g Ô» ÞîöѲdãG%¥@¤Æ´›966ÛÂ9«~±!7 ûïÍ£•Ù]ÿÔö)2÷b>‹í²Ð´ö9:Š!Í;Vìÿær¨Mt䊡q'® é°þ+êÍé¥;K°Jp„â\ñ$_ŽXªm)©²BŠ‡Š¡âE<9z2òhuÅ;4~ÌtaAÿåmÔ/GŒá 7U×Ö…µsÚʽI&¤£[S¦*¨h‘=ÀíR[ÑM—3¨‘b¯Ù?ýô¬™ìå<¹0 ×)w€‡fÔìÀ†í›hŸãª‡Ý!Þ>ÅS¬3[}[Aj&gj3KÏQ¬¡µâŸl!Ïïa,›Ü¾OX£“Ê\¹yÐfä9õìw*cT8â!Êœ—lÝw3ãÅyŒ\¤ygHêû¯ù¾kÛóøâ~è ZÉ©w¶‚õ𪰆ÑȆú«¿GÆÇ?”3‰q_)1F+iA½IÖü ¼\\¯¼KDÛMON=«ÎðÆ<;ÅÒÙß[ÈYY\*e«ÚTq5™ŽÚÏ?H{F­×£û´ëáüð`çF᫵ì5Žmðv?vñlÕäPɾ^ž²fðKZ·^Ä£„Xa«¿±ãî¥å¸ 8$®î²¯¸«#õ¡Ï1ìÛ™5F¥¹AÜënåô³ñ@²á8]îB›þ2[®Lv”\Å-‘ù*μ”Ôôívš­ù&Îl½£Y@ˆG"~ˆ^2ÆMFÚO´ñõ ;ÑÚå ±Å ƒ{êc­-ä~—è¶%‹jI&J¡S·rÀãtRvlYùVZ4Þùª®cOÒÙ / ߸oÇ™®šãÁÅ´x»‚a}ÙÆQ~IEND®B`‚fftw-3.3.3/doc/equation-rodft00.png0000644000175000017500000000312012054410050013746 00000000000000‰PNG  IHDR :•©l0PLTE³³³¨¨¨œœœ„„„xxxlll```TTTHHH<<<000$$$ êYËžtRNS@æØfÎIDATxœíY]ˆUþæ'“»™ìÎ<¶ÔŸ,ø°oÙ}(öeJƒ®¬ºñA$},J èC„ÛmwwÚ]ÙÅk-…ˆ?õç¡¡UY\Ð m)TK„b+>tU¡àF«El5ž{g&™$uÛqt?ÈÜ;÷žóÝsîï™àF¡À´C¦ßAºp@¿GÔ°X™ÞL_Ъ¶éêü†›íõŒê§oë ø ã”u)Ç]YZk·Ë‹×§ ·­XÍ°X@ÐÈ™­{xòic‰ü@ÅÌë-Í¢;…y‘þ‰Ý†¤YE°ë‹¼Ø‰PG0ðÜJÕK¶xŠ¹;ÐÑ aŸœË%<#’ÊŒ !«ç1åÙ#›L¸ô.F£T`»(á{€ç‘¥þ¯@Ÿ¤©•ö…®¡»Ó¤ ~´„€h‹Y™dˆ”†clÕ.ß´ó£÷^'bfLÍýPcàHÜ‹ŽæÞu¡i1Ä‹qÍ«XÌ+¾Ó™L§žìoÔ€[ÐxXW¥Y|Uæ,!±UU5!J"þ´ä¨¢jKf6ò–äuE  T×'JÑ/!7\GñvÉ4ÉÚƒ$8 6ã "ÎiÇ O”¸îÂ3ÅD¦=›µqžè7ùŒ%¥`¤‘º™N‰ùÐ"·ÌŽ†®•ûà¦YxÊ颿¼1Lå Õýc†)*ÅëÕ`(Š^Ž¦;Af‡‡“ìx·Î#¾[Dx»Þ²¡Æ7Žó$µ‰[xËN¿D¿OEæAÜÌ!–™Ý{È‘{7Ù=JÑóX*"1j,–ˆ»H‹<ËÄ@²†5´c–wÛ‡vžð3ŒwÕŽ:îÇIxž7Õ33Nù¡íÌB—-YC‡°j!Ün›âC½ðŸY÷óª¶ûÔŸ;®4—­ûH„¾Fû¯"Üf¾O,ŸöËg¸üXžÂ±LDŸ¢Æ”LÁ¼ØÄ«q$…Ž1/ušá+s6v5WigqDðßZ‹Ü§–XëÝqhЛõ’Q?™¡³ˆTZйº/É(ÄW ývD§Y‰ J—Ÿ j(.Îf·@Ü–‘+~î·Hó?º~º!{–]dW'A‰åZmI/éÉnÇ}˽/Z J͇ õ''·+‘A:µšj¹­6çe¢^ø¶^dÌ ÒG€Ç’ò,}0žKQ·QGiîyªô¼ƒžçÜž¸Û?&íðö.x»ƒH:YÏm¿âÕv%2ˆ‘ƒãCÈo™ÅÅäX$üîx Ú£ŒË4¹Þ¶)ï’òŒ4“¦+§«´:;-fÄ;}¥\K±‚¸’"dE»”kS#Ïi#þìh´ÔZÍ¿¤»%üÄößÁ–÷Ôc³3ÎL™˜{‰ñdÅäP¢YÁ#ž™‚RÅF[ý…ré³?äq;ð>„ŽÅ“œ6¥1ß‚ªÇÑöMk‰Qlª”6+ç^\©×mp •¢™0£kŒt"þG•Æ®aGˆÄ—Aæå°d`6`_™‰‘¦rш°õZ!ñ±È3Ñ^GJìôßÛ£|.]vôPm~«Wrè¶%= 5/%Û_ó³³wÿ:c¢Qeüyà1˜ØHŒéCT6#ñA‡P‚ûÀј\hi[]¢‰Š†81*âƒNDIÍ>¸1@HTR£XbH02(¢„õÜ{gfg·­íB‡Rí×î̹wÎýî7gîÜsﶳ (2£€? 8¤^P—°:Š–•¡E5Új¨ŸŸœP™-i3EÉ P9ÕE¹ d>PÔ5Râ¢*С@l¿ÕJ&g#^ê‚E?ãÝù;ÙS^Ó ÔÊá%@ÆÛ…õ[•UÔWY­)ÉYØ?9{‹‹*tBý•ÁŸÞ媡§Ò·¥9³®  Ã­S{5 ~¢<‹[½€cÖuì<¢¼BÞ¦Ôöƒ€îݜӹ³N¾rKà²ý:yÜ·q¯G ׺lc@Ã|ÎE±ÜÍîO…od«–ï4çÀ\ìŽQäÆ<3¼Èùyü!à©Vņ",PûúÂŒÎÖ`•â êÞÂzã‡àFp<Æ}n/sQLÔ"Œ˜¡îúag‘6‹¢x¬P1 %s3£øÛB5L(ëÑ(çyüÜ…Ÿ¼b(À¹E]4rB²ŽÕ%†Ø(^ËwÛ@t•&ÖsuüTŠ ÚéÀ4E“ÝÕìO,_4›©î›æ#Øàù,¥9üfâRê¥ÃÍÆ MiÍÆ^ûCéPö”x°€ùipd>­wo9ÈÏ’U*ϱ”nðÎd®ëAf§9×ZºÀQí0j³iα”\},£3ñÒ›ìtåÌKí™xÉ'~ÆM¤¨%ÄÆK6ýÝ^GœÈêÅ}Àóîé.9O\œ|f“Fö°“Ö¤¢Ìöbäį™Ó#×/mñßk&D70…SmÄ|ÿ£ídüYaòj×uU¶vÈx/¶FpÝêÁ9—m?‚˜€~v¸ÏÎóßÒLì^ò ÒùMy$»Åߟì&_%|Pªð$lÝY28ÿãÜŒeë›CŽ>¼ Ö“²á˜|;ÖóËDÕ¢_ÜVŸá¥jrûÿ»-΋û¾'§ÈÅí»Pvæt³Ù`üý‡:dc7ä<÷ÛNÙûÑfæ¢CÛ÷pÙ*v•oûBƒ’Á¼ÁBÌøÜ°z}R9ÌŽ®û–벘~.jå?ÅZ•u½à‰=·!Ï(è6¡å<>äWìcÉæ¢ìhõ>NîîsBÙYOðc+4׃7¹l‚÷¾zûF¤Æ€£7KÛ¢~×~<ë@îèh8¼ÝD´±°Fv¢–JE;oç»ë¤‘g„ˆí›ó‡lND{˜ ï½9¿Üíå!?©03;®üYFÙZu àëO’Ñ–ÏDVÿ‘!ø×ÚÛ¼|ÁVÇ‹Áá´É^ÞlÖyiI¼ÚðÓƒýѧöiG†Žl«¾Bhv¼@9‰&d‡Ý,7ä³ä°Õ’­ŽžtàÓ)0S§Ù]–*º‘·`ÄÊdaqòÅ*U6Š÷—ÔD 9éJ~„h^Ûõ°—•ÒÑ £9v£$6~fÌþI_Ú› N±CLìÂd:¹Ë$Èü¯E5=£fÒøïn«º!µŒ"\/k™0:›[é>õïnNý?× Ô¾s­äÙá'©o>íÉb¬Ÿâ?q®QE•;q`ÛcP›k)]a5<ƒiÖß0¿¢m½+æTF·xl^¾‰WÿWÀ¿C?¿ý*IEND®B`‚fftw-3.3.3/doc/equation-rodft10.png0000644000175000017500000000306312054410050013755 00000000000000‰PNG  IHDR:ÉêH0PLTE³³³¨¨¨œœœ„„„xxxlll```TTTHHH<<<000$$$ êYËžtRNS@æØf±IDATxœíXMˆUþº{¦çezgº ñ§ö"iÙƒÐ!ƒ¬Ä¸#(Š™€‚‹ûà!„—˜ÝídVvñ`Œ!0âOÔ2$QVtjØ@0¼Ìh4ˆYë½îž¿¸ŽÙÞuBL-Û]]¯^Õ÷ªëU×<`Q¤@3aƒÁÎÀA¦L¢Äf1À<°W«œjf½{b‚/ÎaObM.«Šë; ø£‚wipF0«ðøKÀé¥`ⶅÇ*e†$ ú»ø r “zãœÕîEcØÚ>Ÿ_P]X~b=Æ_ê©Ñ›t<¿àXMÍ¡,@ ¯éÓ¡š`óÐmT(‹ë";&|kÝ°°jHˆùžñ°9 î4ðqÊ•©‘â¸"4îeÇ@óöˆŠ@Š…ø¡Øl\Ž½Ú„¨§ŽÛËÒ’kvò>ôN¤&M‹5}©hÔBç¾yÂÛ´LÛyŠZFT ‡“þ¼dM1k£ªjb:©@æ‡a2·¾uKÚå°PPUª+nÓ’²:âú1cp¹Kbl·ïДýq®¿Auq–Ó3)˜YaÝœlá|½¹U‚㶕ÆP€šØt¥¤ º\]>KÂ:hþ× &õo£·DáP‹g‚^“oÃe«ñÜpdõVéE<ÖìgYll(¶ ê°n JS©¹ªíº6ú¥‚ {vó1ÙŠF X𹨸 *A¦+Ùºü¤%Y¹…ÊHiôíwàI%zv¨Ò¸}z¨´B Vñ0ùQ6hš#ÆÖ‚)£0l*¨'Ãr\¦ú^¶©û` •J^8Ã'5 À†i²ô^×SÈÄUFg÷Ø¡ ®XwR}Ñm3„QB´’=b¡— ªùØO©6VôÅ—8§ª"+:™HccÊ †ÄpÈ嬔hx‡ƒ0•OO êzš7¢Ð‘)­²#ÑBª+PÝ!ª\’·^A5¼“ŠÀÕš×z|±Ò$UÃJwÁQd¡t›„’èrqwëµ±Áè%êRÒú`¦¤Ç®Z[£ÿOó IFƒ(fv^”§Å,: ’”’ï³t<5zk1;ÐmÀB, ÌŽ‡à&ýohŠ÷Ï÷mdžõÄfœ€ïû}qV?´}ÅÙþ¸I™ˆÜþPÏ0+^v_ÏüɱõJ‡H)7½ÁƒgÑ~²ó?$/4öÊk´mX»Í€ì±Æ°ò#Ûˆ›€q¡ÃÁ°‘¼ØhÔUò~¢„Q –ú°…Ïp?u}%ÿtˆ2÷V?(@$ÈSÚë\ã‡ÁM=ÿ]S¤O \CV¦¨5f<Ÿ¦ðP@4÷ úþ~ßwB'ÛqæKMðèQ9÷¬ÿb8N³R%lù¯ Œ2:8Ù½W*ï kx(ðü@! yúeÊÑ7-Ê{d¥(A`Üpåëes]‘PEÌ“.&EZ¼»¢œïŠóÄÑHjN»”ßD+¤‚øÅ‘ö8¬m4ª’¹¥‰Æ~»¸  ÈL=yT?5qjçôËŒ§æ •Þ]†·9E©èBm4æ¡Ô±ÖRa'1ŽÛcâ )µÏÙ„½îs\MìÛn‰^ê'­Wξp­ãÀ±ð¾ÚVnM±#F9í ™˜ Rò«y%’d§ìvöMÈì\×!âWοkÉZx"Á>_ƒò¥Ø,¹‘™…•–›ä†•Ô? × ©ySrKpš—ž¢ÂÖ7b«¼;(!ÜYTú£xš*hy¬Ÿ‘pÞ™¡þaÀ#×A>þ—ô^4O+Þýä<IEND®B`‚fftw-3.3.3/doc/equation-rodft11.png0000644000175000017500000000315412054410050013757 00000000000000‰PNG  IHDR:vž0PLTE³³³¨¨¨œœœ„„„xxxlll```TTTHHH<<<000$$$ êYËžtRNS@æØfêIDATxœíY]ˆUþæ'“Û™lf|k©?Sða_dG¶Å"Li¬AQú éƒà¢Ö<øP Âmín§Íʬµ"þTíCC«²Ð ¶ôAK ÅV|hT¡`£Õ"¶Ͻ3ù]ݦMÚ,¥ß’Ì™ûsÎwÏ=sæä.p­P ÙpÁàŽÀCB+Z€¾Aô°,ØëeŽç`‰[»¶`®Î¯Ùl`M)©b ]ÞӉȷ˜M>#Vy!Õò KÄëaŒ;íf(b€¦ëwðeÔÈì4ªb™ gœb»9Ctcs»Š 5 ˆ«€!ÕÜ0°+Žx¥‡1½ÀÀ‹‹ôVúòDÌæ ÍÜVeɽYÒà¢DQO½yÌÎê5|À©Bc*nÖÊ[‘"ŸW OSH%ä˜8Ç%©‡?Àh7ÜO"ˆ7ÖÙê±ë²v=8ØÃ(k ¶®ÔþUÄD/ÁvEhZÿ:ŠIO¬&ÞÞ@«vÝæ]Exs„6»GͧI—¥h‹©ëUU:lŠbÜ–ÍüÚ*Ô©Ü‚AsdTUd—šâ7M!&S«'¾?c >÷©ÙÛx4µ@œoQ¾'­¤ÑŸo'e–²w¶h¿Ù\*ir×11ò&×BLî†\l•>é$utg»QúŒ !Þ‹7¯?ÈEÊÆè¡]´ðÀ&Ëý븹á F ï[Ãól jØø@ÔP xóÁ¤¬5±°4¼jüÎ|ÄN·y;Y<·—ˆá«É¦®(Y”Ìl_Š5ùÎ1ÃA’òáOר‘•‰PPIs<ʯ”"íÆ«ŽÒ6´pE‘=e¦y¢o• #„"2klµXgªGÑõ(7giˆU*¥…Q|qD†é–m3sŸU­†š ¥Ú,Oîv#5·Ñ®8÷X:NvD§€Æªv Î,2†}4z,5>ÝpNi©ý…ab}\§Éò §RN\ÔêkB·•¬a 5êZš›ØÉÁ/³ÃEåiNyÊÛÄmŒ·_Ž®Ô.x1E¥‚»mˆXuŒ¨¯{á)òÖojEuŽûZÛÉFlÈ–Ö[:eDZ[¨Òç !<‚>`Õ ˆâ¥…4-¬/·H(… `fÿlª´£xý(2}kanß*ná"Ìòa3 ìßr4*p AÌ, 6§ãaQš›2“[Xv½ØTêÙCaùo7Îè³ÿpl¾ÔѤ䈋QçὨ”ÙÙŸbç6Fì‘߇e¨®rÛÈÒo Ë?ÍÊ#ÞXç: hb¢‘vâ¡(›u*­çlë$x¿oJwÜQo«ÇÍ*ë>Àm°y§Ù2^râë¤v ‡ÜXfZä^=œ!Nü0âH;¥Ød“(¿颲6•Z'Î/Ì_eÃÜŸm¶ñÃëÊÔ)vŽ]žÞKlbçëõª^&ǺÙXúpÈÁ—xˆÊÒBp²›ÍÇ¡„/ÄǦH6:™4³«þ$¼¨gh6sBÓÀ“V‘¹Y*èO›ä0r‘柡ΠØ^Äf«`”ìT¶O‘$‚]óÁË›x!´C³IÜôÞl-mr´ù¤TúH¨Å£!…‡3 2ZÖ¸Hý6°Iù€Ìå$L[¾ÜzVéò*J{ó±S„ÎûËŠé.ߌEvXVˆñŠv!MlŒ9Jµ_n÷ÌX½^–Âíͧ©¿ÀÎïÈW³O1NÌœØ>÷*ãñŠÅ¡Ò¶Žð66'(n}¨õúe(5Œ9êïì¸×b£ÏþœÁ‡‘» D›Ç÷zzhN­ Á¾ïn1 ýѤµÊé—B©u¸:]WzÐÚNÑܦ¦ð1‚r¤)B_ìŸ ±o"áµFKrÖmÀ¾‹„í«;Ú¢…¨‚ð¸Éè`f´%bv6Ê×âKMì¯çÿÐ ƒ|Þ%†Íd BÍÈ‹’ÐÙÛ@ð4¥ÊჭÈÞMAãÏ£4l*“x†rsqjIøÆ{7ƇJ#ÂãK)x‡‚¥ýký˜žÂUIEND®B`‚fftw-3.3.3/doc/FAQ/0000755000175000017500000000000012054411114010634 500000000000000fftw-3.3.3/doc/FAQ/Makefile.am0000644000175000017500000000110212054410050012600 00000000000000BFNNCONV_SRC = bfnnconv.pl m-ascii.pl m-html.pl m-info.pl m-lout.pl m-post.pl FAQ = fftw-faq.ascii fftw-faq.html EXTRA_DIST = fftw-faq.bfnn $(FAQ) $(BFNNCONV_SRC) html.refs html.refs2: html.refs cp -f ${srcdir}/html.refs html.refs2 $(FAQ): $(BFNNCONV_SRC) fftw-faq.bfnn html.refs2 @echo converting... perl -I${srcdir} ${srcdir}/bfnnconv.pl ${srcdir}/fftw-faq.bfnn @echo converting again... perl -I${srcdir} ${srcdir}/bfnnconv.pl ${srcdir}/fftw-faq.bfnn faq: $(FAQ) clean-local: rm -f *~ core a.out *.lout *.ps *.info *.ascii *.xrefdb *.post rm -rf *.html html.refs2 fftw-3.3.3/doc/FAQ/Makefile.in0000644000175000017500000002735112054410200012624 00000000000000# Makefile.in generated by automake 1.11.6 from Makefile.am. # @configure_input@ # Copyright (C) 1994, 1995, 1996, 1997, 1998, 1999, 2000, 2001, 2002, # 2003, 2004, 2005, 2006, 2007, 2008, 2009, 2010, 2011 Free Software # Foundation, Inc. # This Makefile.in is free software; the Free Software Foundation # gives unlimited permission to copy and/or distribute it, # with or without modifications, as long as this notice is preserved. # This program is distributed in the hope that it will be useful, # but WITHOUT ANY WARRANTY, to the extent permitted by law; without # even the implied warranty of MERCHANTABILITY or FITNESS FOR A # PARTICULAR PURPOSE. @SET_MAKE@ VPATH = @srcdir@ am__make_dryrun = \ { \ am__dry=no; \ case $$MAKEFLAGS in \ *\\[\ \ ]*) \ echo 'am--echo: ; @echo "AM" OK' | $(MAKE) -f - 2>/dev/null \ | grep '^AM OK$$' >/dev/null || am__dry=yes;; \ *) \ for am__flg in $$MAKEFLAGS; do \ case $$am__flg in \ *=*|--*) ;; \ *n*) am__dry=yes; break;; \ esac; \ done;; \ esac; \ test $$am__dry = yes; \ } pkgdatadir = $(datadir)/@PACKAGE@ pkgincludedir = $(includedir)/@PACKAGE@ pkglibdir = $(libdir)/@PACKAGE@ pkglibexecdir = $(libexecdir)/@PACKAGE@ am__cd = CDPATH="$${ZSH_VERSION+.}$(PATH_SEPARATOR)" && cd install_sh_DATA = $(install_sh) -c -m 644 install_sh_PROGRAM = $(install_sh) -c install_sh_SCRIPT = $(install_sh) -c INSTALL_HEADER = $(INSTALL_DATA) transform = $(program_transform_name) NORMAL_INSTALL = : PRE_INSTALL = : POST_INSTALL = : NORMAL_UNINSTALL = : PRE_UNINSTALL = : POST_UNINSTALL = : build_triplet = @build@ host_triplet = @host@ subdir = doc/FAQ DIST_COMMON = $(srcdir)/Makefile.am $(srcdir)/Makefile.in ACLOCAL_M4 = $(top_srcdir)/aclocal.m4 am__aclocal_m4_deps = $(top_srcdir)/m4/acx_mpi.m4 \ $(top_srcdir)/m4/acx_pthread.m4 \ $(top_srcdir)/m4/ax_cc_maxopt.m4 \ $(top_srcdir)/m4/ax_check_compiler_flags.m4 \ $(top_srcdir)/m4/ax_compiler_vendor.m4 \ $(top_srcdir)/m4/ax_gcc_aligns_stack.m4 \ $(top_srcdir)/m4/ax_gcc_version.m4 \ $(top_srcdir)/m4/ax_openmp.m4 $(top_srcdir)/m4/libtool.m4 \ $(top_srcdir)/m4/ltoptions.m4 $(top_srcdir)/m4/ltsugar.m4 \ $(top_srcdir)/m4/ltversion.m4 $(top_srcdir)/m4/lt~obsolete.m4 \ $(top_srcdir)/configure.ac am__configure_deps = $(am__aclocal_m4_deps) $(CONFIGURE_DEPENDENCIES) \ $(ACLOCAL_M4) mkinstalldirs = $(install_sh) -d CONFIG_HEADER = $(top_builddir)/config.h CONFIG_CLEAN_FILES = CONFIG_CLEAN_VPATH_FILES = SOURCES = DIST_SOURCES = am__can_run_installinfo = \ case $$AM_UPDATE_INFO_DIR in \ n|no|NO) false;; \ *) (install-info --version) >/dev/null 2>&1;; \ esac DISTFILES = $(DIST_COMMON) $(DIST_SOURCES) $(TEXINFOS) $(EXTRA_DIST) ACLOCAL = @ACLOCAL@ ALLOCA = @ALLOCA@ ALTIVEC_CFLAGS = @ALTIVEC_CFLAGS@ AMTAR = @AMTAR@ AR = @AR@ AS = @AS@ AUTOCONF = @AUTOCONF@ AUTOHEADER = @AUTOHEADER@ AUTOMAKE = @AUTOMAKE@ AVX_CFLAGS = @AVX_CFLAGS@ AWK = @AWK@ CC = @CC@ CCDEPMODE = @CCDEPMODE@ CFLAGS = @CFLAGS@ CHECK_PL_OPTS = @CHECK_PL_OPTS@ CPP = @CPP@ CPPFLAGS = @CPPFLAGS@ CYGPATH_W = @CYGPATH_W@ C_FFTW_R2R_KIND = @C_FFTW_R2R_KIND@ C_MPI_FINT = @C_MPI_FINT@ DEFS = @DEFS@ DEPDIR = @DEPDIR@ DLLTOOL = @DLLTOOL@ DSYMUTIL = @DSYMUTIL@ DUMPBIN = @DUMPBIN@ ECHO_C = @ECHO_C@ ECHO_N = @ECHO_N@ ECHO_T = @ECHO_T@ EGREP = @EGREP@ EXEEXT = @EXEEXT@ F77 = @F77@ FFLAGS = @FFLAGS@ FGREP = @FGREP@ FLIBS = @FLIBS@ GREP = @GREP@ INSTALL = @INSTALL@ INSTALL_DATA = @INSTALL_DATA@ INSTALL_PROGRAM = @INSTALL_PROGRAM@ INSTALL_SCRIPT = @INSTALL_SCRIPT@ INSTALL_STRIP_PROGRAM = @INSTALL_STRIP_PROGRAM@ LD = @LD@ LDFLAGS = @LDFLAGS@ LIBOBJS = @LIBOBJS@ LIBQUADMATH = @LIBQUADMATH@ LIBS = @LIBS@ LIBTOOL = @LIBTOOL@ LIPO = @LIPO@ LN_S = @LN_S@ LTLIBOBJS = @LTLIBOBJS@ MAINT = @MAINT@ MAKEINFO = @MAKEINFO@ MANIFEST_TOOL = @MANIFEST_TOOL@ MKDIR_P = @MKDIR_P@ MPICC = @MPICC@ MPILIBS = @MPILIBS@ MPIRUN = @MPIRUN@ NEON_CFLAGS = @NEON_CFLAGS@ NM = @NM@ NMEDIT = @NMEDIT@ OBJDUMP = @OBJDUMP@ OBJEXT = @OBJEXT@ OCAMLBUILD = @OCAMLBUILD@ OPENMP_CFLAGS = @OPENMP_CFLAGS@ OTOOL = @OTOOL@ OTOOL64 = @OTOOL64@ PACKAGE = @PACKAGE@ PACKAGE_BUGREPORT = @PACKAGE_BUGREPORT@ PACKAGE_NAME = @PACKAGE_NAME@ PACKAGE_STRING = @PACKAGE_STRING@ PACKAGE_TARNAME = @PACKAGE_TARNAME@ PACKAGE_URL = @PACKAGE_URL@ PACKAGE_VERSION = @PACKAGE_VERSION@ PATH_SEPARATOR = @PATH_SEPARATOR@ POW_LIB = @POW_LIB@ PRECISION = @PRECISION@ PREC_SUFFIX = @PREC_SUFFIX@ PTHREAD_CC = @PTHREAD_CC@ PTHREAD_CFLAGS = @PTHREAD_CFLAGS@ PTHREAD_LIBS = @PTHREAD_LIBS@ RANLIB = @RANLIB@ SED = @SED@ SET_MAKE = @SET_MAKE@ SHARED_VERSION_INFO = @SHARED_VERSION_INFO@ SHELL = @SHELL@ SSE2_CFLAGS = @SSE2_CFLAGS@ STACK_ALIGN_CFLAGS = @STACK_ALIGN_CFLAGS@ STRIP = @STRIP@ THREADLIBS = @THREADLIBS@ VERSION = @VERSION@ abs_builddir = @abs_builddir@ abs_srcdir = @abs_srcdir@ abs_top_builddir = @abs_top_builddir@ abs_top_srcdir = @abs_top_srcdir@ ac_ct_AR = @ac_ct_AR@ ac_ct_CC = @ac_ct_CC@ ac_ct_DUMPBIN = @ac_ct_DUMPBIN@ ac_ct_F77 = @ac_ct_F77@ acx_pthread_config = @acx_pthread_config@ am__include = @am__include@ am__leading_dot = @am__leading_dot@ am__quote = @am__quote@ am__tar = @am__tar@ am__untar = @am__untar@ bindir = @bindir@ build = @build@ build_alias = @build_alias@ build_cpu = @build_cpu@ build_os = @build_os@ build_vendor = @build_vendor@ builddir = @builddir@ datadir = @datadir@ datarootdir = @datarootdir@ docdir = @docdir@ dvidir = @dvidir@ exec_prefix = @exec_prefix@ host = @host@ host_alias = @host_alias@ host_cpu = @host_cpu@ host_os = @host_os@ host_vendor = @host_vendor@ htmldir = @htmldir@ includedir = @includedir@ infodir = @infodir@ install_sh = @install_sh@ libdir = @libdir@ libexecdir = @libexecdir@ localedir = @localedir@ localstatedir = @localstatedir@ mandir = @mandir@ mkdir_p = @mkdir_p@ oldincludedir = @oldincludedir@ pdfdir = @pdfdir@ prefix = @prefix@ program_transform_name = @program_transform_name@ psdir = @psdir@ sbindir = @sbindir@ sharedstatedir = @sharedstatedir@ srcdir = @srcdir@ sysconfdir = @sysconfdir@ target_alias = @target_alias@ top_build_prefix = @top_build_prefix@ top_builddir = @top_builddir@ top_srcdir = @top_srcdir@ BFNNCONV_SRC = bfnnconv.pl m-ascii.pl m-html.pl m-info.pl m-lout.pl m-post.pl FAQ = fftw-faq.ascii fftw-faq.html EXTRA_DIST = fftw-faq.bfnn $(FAQ) $(BFNNCONV_SRC) html.refs all: all-am .SUFFIXES: $(srcdir)/Makefile.in: @MAINTAINER_MODE_TRUE@ $(srcdir)/Makefile.am $(am__configure_deps) @for dep in $?; do \ case '$(am__configure_deps)' in \ *$$dep*) \ ( cd $(top_builddir) && $(MAKE) $(AM_MAKEFLAGS) am--refresh ) \ && { if test -f $@; then exit 0; else break; fi; }; \ exit 1;; \ esac; \ done; \ echo ' cd $(top_srcdir) && $(AUTOMAKE) --gnu doc/FAQ/Makefile'; \ $(am__cd) $(top_srcdir) && \ $(AUTOMAKE) --gnu doc/FAQ/Makefile .PRECIOUS: Makefile Makefile: $(srcdir)/Makefile.in $(top_builddir)/config.status @case '$?' in \ *config.status*) \ cd $(top_builddir) && $(MAKE) $(AM_MAKEFLAGS) am--refresh;; \ *) \ echo ' cd $(top_builddir) && $(SHELL) ./config.status $(subdir)/$@ $(am__depfiles_maybe)'; \ cd $(top_builddir) && $(SHELL) ./config.status $(subdir)/$@ $(am__depfiles_maybe);; \ esac; $(top_builddir)/config.status: $(top_srcdir)/configure $(CONFIG_STATUS_DEPENDENCIES) cd $(top_builddir) && $(MAKE) $(AM_MAKEFLAGS) am--refresh $(top_srcdir)/configure: @MAINTAINER_MODE_TRUE@ $(am__configure_deps) cd $(top_builddir) && $(MAKE) $(AM_MAKEFLAGS) am--refresh $(ACLOCAL_M4): @MAINTAINER_MODE_TRUE@ $(am__aclocal_m4_deps) cd $(top_builddir) && $(MAKE) $(AM_MAKEFLAGS) am--refresh $(am__aclocal_m4_deps): mostlyclean-libtool: -rm -f *.lo clean-libtool: -rm -rf .libs _libs tags: TAGS TAGS: ctags: CTAGS CTAGS: distdir: $(DISTFILES) @srcdirstrip=`echo "$(srcdir)" | sed 's/[].[^$$\\*]/\\\\&/g'`; \ topsrcdirstrip=`echo "$(top_srcdir)" | sed 's/[].[^$$\\*]/\\\\&/g'`; \ list='$(DISTFILES)'; \ dist_files=`for file in $$list; do echo $$file; done | \ sed -e "s|^$$srcdirstrip/||;t" \ -e "s|^$$topsrcdirstrip/|$(top_builddir)/|;t"`; \ case $$dist_files in \ */*) $(MKDIR_P) `echo "$$dist_files" | \ sed '/\//!d;s|^|$(distdir)/|;s,/[^/]*$$,,' | \ sort -u` ;; \ esac; \ for file in $$dist_files; do \ if test -f $$file || test -d $$file; then d=.; else d=$(srcdir); fi; \ if test -d $$d/$$file; then \ dir=`echo "/$$file" | sed -e 's,/[^/]*$$,,'`; \ if test -d "$(distdir)/$$file"; then \ find "$(distdir)/$$file" -type d ! -perm -700 -exec chmod u+rwx {} \;; \ fi; \ if test -d $(srcdir)/$$file && test $$d != $(srcdir); then \ cp -fpR $(srcdir)/$$file "$(distdir)$$dir" || exit 1; \ find "$(distdir)/$$file" -type d ! -perm -700 -exec chmod u+rwx {} \;; \ fi; \ cp -fpR $$d/$$file "$(distdir)$$dir" || exit 1; \ else \ test -f "$(distdir)/$$file" \ || cp -p $$d/$$file "$(distdir)/$$file" \ || exit 1; \ fi; \ done check-am: all-am check: check-am all-am: Makefile installdirs: install: install-am install-exec: install-exec-am install-data: install-data-am uninstall: uninstall-am install-am: all-am @$(MAKE) $(AM_MAKEFLAGS) install-exec-am install-data-am installcheck: installcheck-am install-strip: if test -z '$(STRIP)'; then \ $(MAKE) $(AM_MAKEFLAGS) INSTALL_PROGRAM="$(INSTALL_STRIP_PROGRAM)" \ install_sh_PROGRAM="$(INSTALL_STRIP_PROGRAM)" INSTALL_STRIP_FLAG=-s \ install; \ else \ $(MAKE) $(AM_MAKEFLAGS) INSTALL_PROGRAM="$(INSTALL_STRIP_PROGRAM)" \ install_sh_PROGRAM="$(INSTALL_STRIP_PROGRAM)" INSTALL_STRIP_FLAG=-s \ "INSTALL_PROGRAM_ENV=STRIPPROG='$(STRIP)'" install; \ fi mostlyclean-generic: clean-generic: distclean-generic: -test -z "$(CONFIG_CLEAN_FILES)" || rm -f $(CONFIG_CLEAN_FILES) -test . = "$(srcdir)" || test -z "$(CONFIG_CLEAN_VPATH_FILES)" || rm -f $(CONFIG_CLEAN_VPATH_FILES) maintainer-clean-generic: @echo "This command is intended for maintainers to use" @echo "it deletes files that may require special tools to rebuild." clean: clean-am clean-am: clean-generic clean-libtool clean-local mostlyclean-am distclean: distclean-am -rm -f Makefile distclean-am: clean-am distclean-generic dvi: dvi-am dvi-am: html: html-am html-am: info: info-am info-am: install-data-am: install-dvi: install-dvi-am install-dvi-am: install-exec-am: install-html: install-html-am install-html-am: install-info: install-info-am install-info-am: install-man: install-pdf: install-pdf-am install-pdf-am: install-ps: install-ps-am install-ps-am: installcheck-am: maintainer-clean: maintainer-clean-am -rm -f Makefile maintainer-clean-am: distclean-am maintainer-clean-generic mostlyclean: mostlyclean-am mostlyclean-am: mostlyclean-generic mostlyclean-libtool pdf: pdf-am pdf-am: ps: ps-am ps-am: uninstall-am: .MAKE: install-am install-strip .PHONY: all all-am check check-am clean clean-generic clean-libtool \ clean-local distclean distclean-generic distclean-libtool \ distdir dvi dvi-am html html-am info info-am install \ install-am install-data install-data-am install-dvi \ install-dvi-am install-exec install-exec-am install-html \ install-html-am install-info install-info-am install-man \ install-pdf install-pdf-am install-ps install-ps-am \ install-strip installcheck installcheck-am installdirs \ maintainer-clean maintainer-clean-generic mostlyclean \ mostlyclean-generic mostlyclean-libtool pdf pdf-am ps ps-am \ uninstall uninstall-am html.refs2: html.refs cp -f ${srcdir}/html.refs html.refs2 $(FAQ): $(BFNNCONV_SRC) fftw-faq.bfnn html.refs2 @echo converting... perl -I${srcdir} ${srcdir}/bfnnconv.pl ${srcdir}/fftw-faq.bfnn @echo converting again... perl -I${srcdir} ${srcdir}/bfnnconv.pl ${srcdir}/fftw-faq.bfnn faq: $(FAQ) clean-local: rm -f *~ core a.out *.lout *.ps *.info *.ascii *.xrefdb *.post rm -rf *.html html.refs2 # Tell versions [3.59,3.63) of GNU make to not export all variables. # Otherwise a system limit (for SysV at least) may be exceeded. .NOEXPORT: fftw-3.3.3/doc/FAQ/fftw-faq.bfnn0000644000175000017500000007411012054410050013135 00000000000000\comment This is the source for the FFTW FAQ list, in \comment the Bizarre Format With No Name. It is turned into Lout \comment input, HTML, plain ASCII and an Info document by a Perl script. \comment \comment The format and scripts come from the Linux FAQ, by \comment Ian Jackson. \set brieftitle FFTW FAQ \set author Matteo Frigo and Steven G. Johnson / fftw@fftw.org \set authormail fftw@fftw.org \set title FFTW Frequently Asked Questions with Answers \set copyholder Matteo Frigo and Massachusetts Institute of Technology \call-html startup html.refs2 \copyto ASCII FFTW FREQUENTLY ASKED QUESTIONS WITH ANSWERS `date '+%d %h %Y'` Matteo Frigo Steven G. Johnson \endcopy \copyto INFO START-INFO-DIR-ENTRY * FFTW FAQ: (fftw-faq). FFTW Frequently Asked Questions with Answers. END-INFO-DIR-ENTRY  File: $prefix.info, Node: Top, Next: Question 1.1, Up: (dir) FFTW FREQUENTLY ASKED QUESTIONS WITH ANSWERS `date '+%d %h %Y'` Matteo Frigo Steven G. Johnson \endcopy This is the list of Frequently Asked Questions about FFTW, a collection of fast C routines for computing the Discrete Fourier Transform in one or more dimensions. \section Index \index \comment ###################################################################### \section Introduction and General Information \question 26aug:whatisfftw What is FFTW? FFTW is a free collection of fast C routines for computing the Discrete Fourier Transform in one or more dimensions. It includes complex, real, symmetric, and parallel transforms, and can handle arbitrary array sizes efficiently. FFTW is typically faster than other publically-available FFT implementations, and is even competitive with vendor-tuned libraries. (See our web page for extensive benchmarks.) To achieve this performance, FFTW uses novel code-generation and runtime self-optimization techniques (along with many other tricks). \question 26aug:whereisfftw How do I obtain FFTW? FFTW can be found at \docref{the FFTW web page\}. You can also retrieve it from \ftpon ftp.fftw.org in \ftpin /pub/fftw. \question 26aug:isfftwfree Is FFTW free software? Starting with version 1.3, FFTW is Free Software in the technical sense defined by the Free Software Foundation (see \docref{Categories of Free and Non-Free Software\}), and is distributed under the terms of the GNU General Public License. Previous versions of FFTW were distributed without fee for noncommercial use, but were not technically ``free.'' Non-free licenses for FFTW are also available that permit different terms of use than the GPL. \question 10apr:nonfree What is this about non-free licenses? The non-free licenses are for companies that wish to use FFTW in their products but are unwilling to release their software under the GPL (which would require them to release source code and allow free redistribution). Such users can purchase an unlimited-use license from MIT. Contact us for more details. We could instead have released FFTW under the LGPL, or even disallowed non-Free usage. Suffice it to say, however, that MIT owns the copyright to FFTW and they only let us GPL it because we convinced them that it would neither affect their licensing revenue nor irritate existing licensees. \question 24oct:west In the West? I thought MIT was in the East? Not to an Italian. You could say that we're a Spaghetti Western (with apologies to Sergio Leone). \comment ###################################################################### \section Installing FFTW \question 26aug:systems Which systems does FFTW run on? FFTW is written in ANSI C, and should work on any system with a decent C compiler. (See also \qref runOnWindows, \qref compilerCrashes.) FFTW can also take advantage of certain hardware-specific features, such as cycle counters and SIMD instructions, but this is optional. \question 26aug:runOnWindows Does FFTW run on Windows? Yes, many people have reported successfully using FFTW on Windows with various compilers. FFTW was not developed on Windows, but the source code is essentially straight ANSI C. See also the \docref{FFTW Windows installation notes\}, \qref compilerCrashes, and \qref vbetalia. \question 26aug:compilerCrashes My compiler has trouble with FFTW. Complain fiercely to the vendor of the compiler. We have successfully used \courier{gcc\} 3.2.x on x86 and PPC, a recent Compaq C compiler for Alpha, version 6 of IBM's \courier{xlc\} compiler for AIX, Intel's \courier{icc\} versions 5-7, and Sun WorkShop \courier{cc\} version 6. FFTW is likely to push compilers to their limits, however, and several compiler bugs have been exposed by FFTW. A partial list follows. \courier{gcc\} 2.95.x for Solaris/SPARC produces incorrect code for the test program (workaround: recompile the \courier{libbench2\} directory with \courier{-O2\}). NetBSD/macppc 1.6 comes with a \courier{gcc\} version that also miscompiles the test program. (Please report a workaround if you know one.) \courier{gcc\} 3.2.3 for ARM reportedly crashes during compilation. This bug is reportedly fixed in later versions of \courier{gcc\}. Versions 8.0 and 8.1 of Intel's \courier{icc\} falsely claim to be \courier{gcc\}, so you should specify \courier{CC="icc -no-gcc"\}; this is automatic in FFTW 3.1. \courier{icc-8.0.066\} reportely produces incorrect code for FFTW 2.1.5, but is fixed in version 8.1. \courier{icc-7.1\} compiler build 20030402Z appears to produce incorrect dependencies, causing the compilation to fail. \courier{icc-7.1\} build 20030307Z appears to work fine. (Use \courier{icc -V\} to check which build you have.) As of 2003/04/18, build 20030402Z appears not to be available any longer on Intel's website, whereas the older build 20030307Z is available. \courier{ranlib\} of GNU \courier{binutils\} 2.9.1 on Irix has been observed to corrupt the FFTW libraries, causing a link failure when FFTW is compiled. Since \courier{ranlib\} is completely superfluous on Irix, we suggest deleting it from your system and replacing it with a symbolic link to \courier{/bin/echo\}. If support for SIMD instructions is enabled in FFTW, further compiler problems may appear: \courier{gcc\} 3.4.[0123] for x86 produces incorrect SSE2 code for FFTW when \courier{-O2\} (the best choice for FFTW) is used, causing FFTW to crash (\courier{make check\} crashes). This bug is fixed in \courier{gcc\} 3.4.4. On x86_64 (amd64/em64t), \courier{gcc\} 3.4.4 reportedly still has a similar problem, but this is fixed as of \courier{gcc\} 3.4.6. \courier{gcc-3.2\} for x86 produces incorrect SIMD code if \courier{-O3\} is used. The same compiler produces incorrect SIMD code if no optimization is used, too. When using \courier{gcc-3.2\}, it is a good idea not to change the default \courier{CFLAGS\} selected by the \courier{configure\} script. Some 3.0.x and 3.1.x versions of \courier{gcc\} on \courier{x86\} may crash. \courier{gcc\} so-called 2.96 shipping with RedHat 7.3 crashes when compiling SIMD code. In both cases, please upgrade to \courier{gcc-3.2\} or later. Intel's \courier{icc\} 6.0 misaligns SSE constants, but FFTW has a workaround. \courier{icc\} 8.x fails to compile FFTW 3.0.x because it falsely claims to be \courier{gcc\}; we believe this to be a bug in \courier{icc\}, but FFTW 3.1 has a workaround. Visual C++ 2003 reportedly produces incorrect code for SSE/SSE2 when compiling FFTW. This bug was reportedly fixed in VC++ 2005; alternatively, you could switch to the Intel compiler. VC++ 6.0 also reportedly produces incorrect code for the file \courier{reodft11e-r2hc-odd.c\} unless optimizations are disabled for that file. \courier{gcc\} 2.95 on MacOS X miscompiles AltiVec code (fixed in later versions). \courier{gcc\} 3.2.x miscompiles AltiVec permutations, but FFTW has a workaround. \courier{gcc\} 4.0.1 on MacOS for Intel crashes when compiling FFTW; a workaround is to compile one file without optimization: \courier{cd kernel; make CFLAGS=" " trig.lo\}. \courier{gcc\} 4.1.1 reportedly crashes when compiling FFTW for MIPS; the workaround is to compile the file it crashes on (\courier{t2_64.c\}) with a lower optimization level. \courier{gcc\} versions 4.1.2 to 4.2.0 for x86 reportedly miscompile FFTW 3.1's test program, causing \courier{make check\} to crash (\courier{gcc\} bug #26528). The bug was reportedly fixed in \courier{gcc\} version 4.2.1 and later. A workaround is to compile \courier{libbench2/verify-lib.c\} without optimization. \question 26aug:solarisSucks FFTW does not compile on Solaris, complaining about \courier{const\}. We know that at least on Solaris 2.5.x with Sun's compilers 4.2 you might get error messages from \courier{make\} such as \courier{"./fftw.h", line 88: warning: const is a keyword in ANSI C\} This is the case when the \courier{configure\} script reports that \courier{const\} does not work: \courier{checking for working const... (cached) no\} You should be aware that Solaris comes with two compilers, namely, \courier{/opt/SUNWspro/SC4.2/bin/cc\} and \courier{/usr/ucb/cc\}. The latter compiler is non-ANSI. Indeed, it is a perverse shell script that calls the real compiler in non-ANSI mode. In order to compile FFTW, change your path so that the right \courier{cc\} is used. To know whether your compiler is the right one, type \courier{cc -V\}. If the compiler prints ``\courier{ucbcc\}'', as in \courier{ucbcc: WorkShop Compilers 4.2 30 Oct 1996 C 4.2\} then the compiler is wrong. The right message is something like \courier{cc: WorkShop Compilers 4.2 30 Oct 1996 C 4.2\} \question 19mar:3dnow What's the difference between \courier{--enable-3dnow\} and \courier{--enable-k7\}? \courier{--enable-k7\} enables 3DNow! instructions on K7 processors (AMD Athlon and its variants). K7 support is provided by assembly routines generated by a special purpose compiler. As of fftw-3.2, --enable-k7 is no longer supported. \courier{--enable-3dnow\} enables generic 3DNow! support using \courier{gcc\} builtin functions. This works on earlier AMD processors, but it is not as fast as our special assembly routines. As of fftw-3.1, --enable-3dnow is no longer supported. \question 18apr:fma What's the difference between the fma and the non-fma versions? The fma version tries to exploit the fused multiply-add instructions implemented in many processors such as PowerPC, ia-64, and MIPS. The two FFTW packages are otherwise identical. In FFTW 3.1, the fma and non-fma versions were merged together into a single package, and the \courier{configure\} script attempts to automatically guess which version to use. The FFTW 3.1 \courier{configure\} script enables fma by default on PowerPC, Itanium, and PA-RISC, and disables it otherwise. You can force one or the other by using the \courier{--enable-fma\} or \courier{--disable-fma\} flag for \courier{configure\}. Definitely use fma if you have a PowerPC-based system with \courier{gcc\} (or IBM \courier{xlc\}). This includes all GNU/Linux systems for PowerPC and the older PowerPC-based MacOS systems. Also use it on PA-RISC and Itanium with the HP/UX compiler. Definitely do not use the fma version if you have an ia-32 processor (Intel, AMD, MacOS on Intel, etcetera). For other architectures/compilers, the situation is not so clear. For example, ia-64 has the fma instruction, but \courier{gcc-3.2\} appears not to exploit it correctly. Other compilers may do the right thing, but we have not tried them. Please send us your feedback so that we can update this FAQ entry. \question 26aug:languages Which language is FFTW written in? FFTW is written in ANSI C. Most of the code, however, was automatically generated by a program called \courier{genfft\}, written in the Objective Caml dialect of ML. You do not need to know ML or to have an Objective Caml compiler in order to use FFTW. \courier{genfft\} is provided with the FFTW sources, which means that you can play with the code generator if you want. In this case, you need a working Objective Caml system. Objective Caml is available from \docref{the Caml web page\}. \question 26aug:fortran Can I call FFTW from Fortran? Yes, FFTW (versions 1.3 and higher) contains a Fortran-callable interface, documented in the FFTW manual. By default, FFTW configures its Fortran interface to work with the first compiler it finds, e.g. \courier{g77\}. To configure for a different, incompatible Fortran compiler \courier{foobar\}, use \courier{./configure F77=foobar\} when installing FFTW. (In the case of \courier{g77\}, however, FFTW 3.x also includes an extra set of Fortran-callable routines with one less underscore at the end of identifiers, which should cover most other Fortran compilers on Linux at least.) \question 26aug:cplusplus Can I call FFTW from C++? Most definitely. FFTW should compile and/or link under any C++ compiler. Moreover, it is likely that the C++ \courier{\} template class is bit-compatible with FFTW's complex-number format (see the FFTW manual for more details). \question 26aug:whynotfortran Why isn't FFTW written in Fortran/C++? Because we don't like those languages, and neither approaches the portability of C. \question 29mar:singleprec How do I compile FFTW to run in single precision? On a Unix system: \courier{configure --enable-float\}. On a non-Unix system: edit \courier{config.h\} to \courier{#define\} the symbol \courier{FFTW_SINGLE\} (for FFTW 3.x). In both cases, you must then recompile FFTW. In FFTW 3, all FFTW identifiers will then begin with \courier{fftwf_\} instead of \courier{fftw_\}. \question 28mar:64bitk7 --enable-k7 does not work on x86-64 Support for --enable-k7 was discontinued in fftw-3.2. The fftw-3.1 release supports --enable-k7. This option only works on 32-bit x86 machines that implement 3DNow!, including the AMD Athlon and the AMD Opteron in 32-bit mode. --enable-k7 does not work on AMD Opteron in 64-bit mode. Use --enable-sse for x86-64 machines. FFTW supports 3DNow! by means of assembly code generated by a special-purpose compiler. It is hard to produce assembly code that works in both 32-bit and 64-bit mode. \comment ###################################################################### \section Using FFTW \question 15mar:fftw2to3 Why not support the FFTW 2 interface in FFTW 3? FFTW 3 has semantics incompatible with earlier versions: its plans can only be used for a given stride, multiplicity, and other characteristics of the input and output arrays; these stronger semantics are necessary for performance reasons. Thus, it is impossible to efficiently emulate the older interface (whose plans can be used for any transform of the same size). We believe that it should be possible to upgrade most programs without any difficulty, however. \question 20mar:planperarray Why do FFTW 3 plans encapsulate the input/output arrays and not just the algorithm? There are several reasons: \call startlist \call item It was important for performance reasons that the plan be specific to array characteristics like the stride (and alignment, for SIMD), and requiring that the user maintain these invariants is error prone. \call item In most high-performance applications, as far as we can tell, you are usually transforming the same array over and over, so FFTW's semantics should not be a burden. \call item If you need to transform another array of the same size, creating a new plan once the first exists is a cheap operation. \call item If you need to transform many arrays of the same size at once, you should really use the \courier{plan_many\} routines in FFTW's "advanced" interface. \call item If the abovementioned array characteristics are the same, you are willing to pay close attention to the documentation, and you really need to, we provide a "new-array execution" interface to apply a plan to a new array. \call endlist \question 25may:slow FFTW seems really slow. You are probably recreating the plan before every transform, rather than creating it once and reusing it for all transforms of the same size. FFTW is designed to be used in the following way: \call startlist \call item First, you create a plan. This will take several seconds. \call item Then, you reuse the plan many times to perform FFTs. These are fast. \call endlist If you don't need to compute many transforms and the time for the planner is significant, you have two options. First, you can use the \courier{FFTW_ESTIMATE\} option in the planner, which uses heuristics instead of runtime measurements and produces a good plan in a short time. Second, you can use the wisdom feature to precompute the plan; see \qref savePlans \question 22oct:slows FFTW slows down after repeated calls. Probably, NaNs or similar are creeping into your data, and the slowdown is due to the resulting floating-point exceptions. For example, be aware that repeatedly FFTing the same array is a diverging process (because FFTW computes the unnormalized transform). \question 22oct:segfault An FFTW routine is crashing when I call it. Did the FFTW test programs pass (\courier{make check\}, or \courier{cd tests; make bigcheck\} if you want to be paranoid)? If so, you almost certainly have a bug in your own code. For example, you could be passing invalid arguments (such as wrongly-sized arrays) to FFTW, or you could simply have memory corruption elsewhere in your program that causes random crashes later on. Please don't complain to us unless you can come up with a minimal self-contained program (preferably under 30 lines) that illustrates the problem. \question 22oct:fortran64 My Fortran program crashes when calling FFTW. As described in the manual, on 64-bit machines you must store the plans in variables large enough to hold a pointer, for example \courier{integer*8\}. We recommend using \courier{integer*8\} on 32-bit machines as well, to simplify porting. \question 24mar:conventions FFTW gives results different from my old FFT. People follow many different conventions for the DFT, and you should be sure to know the ones that we use (described in the FFTW manual). In particular, you should be aware that the \courier{FFTW_FORWARD\}/\courier{FFTW_BACKWARD\} directions correspond to signs of -1/+1 in the exponent of the DFT definition. (\italic{Numerical Recipes\} uses the opposite convention.) You should also know that we compute an unnormalized transform. In contrast, Matlab is an example of program that computes a normalized transform. See \qref whyscaled. Finally, note that floating-point arithmetic is not exact, so different FFT algorithms will give slightly different results (on the order of the numerical accuracy; typically a fractional difference of 1e-15 or so in double precision). \question 31aug:nondeterministic FFTW gives different results between runs If you use \courier{FFTW_MEASURE\} or \courier{FFTW_PATIENT\} mode, then the algorithm FFTW employs is not deterministic: it depends on runtime performance measurements. This will cause the results to vary slightly from run to run. However, the differences should be slight, on the order of the floating-point precision, and therefore should have no practical impact on most applications. If you use saved plans (wisdom) or \courier{FFTW_ESTIMATE\} mode, however, then the algorithm is deterministic and the results should be identical between runs. \question 26aug:savePlans Can I save FFTW's plans? Yes. Starting with version 1.2, FFTW provides the \courier{wisdom\} mechanism for saving plans; see the FFTW manual. \question 14sep:whyscaled Why does your inverse transform return a scaled result? Computing the forward transform followed by the backward transform (or vice versa) yields the original array scaled by the size of the array. (For multi-dimensional transforms, the size of the array is the product of the dimensions.) We could, instead, have chosen a normalization that would have returned the unscaled array. Or, to accomodate the many conventions in this matter, the transform routines could have accepted a "scale factor" parameter. We did not do this, however, for two reasons. First, we didn't want to sacrifice performance in the common case where the scale factor is 1. Second, in real applications the FFT is followed or preceded by some computation on the data, into which the scale factor can typically be absorbed at little or no cost. \question 02dec:centerorigin How can I make FFTW put the origin (zero frequency) at the center of its output? For human viewing of a spectrum, it is often convenient to put the origin in frequency space at the center of the output array, rather than in the zero-th element (the default in FFTW). If all of the dimensions of your array are even, you can accomplish this by simply multiplying each element of the input array by (-1)^(i + j + ...), where i, j, etcetera are the indices of the element. (This trick is a general property of the DFT, and is not specific to FFTW.) \question 08may:imageaudio How do I FFT an image/audio file in \italic{foobar\} format? FFTW performs an FFT on an array of floating-point values. You can certainly use it to compute the transform of an image or audio stream, but you are responsible for figuring out your data format and converting it to the form FFTW requires. \question 09apr:linkfails My program does not link (on Unix). The libraries must be listed in the correct order (\courier{-lfftw3 -lm\} for FFTW 3.x) and \italic{after\} your program sources/objects. (The general rule is that if \italic{A\} uses \italic{B\}, then \italic{A\} must be listed before \italic{B\} in the link command.). \question 15mar:linkheader I included your header, but linking still fails. You're a C++ programmer, aren't you? You have to compile the FFTW library and link it into your program, not just \courier{#include \}. (Yes, this is really a FAQ.) \question 22oct:nostack My program crashes, complaining about stack space. You cannot declare large arrays with automatic storage (e.g. via \courier{fftw_complex array[N]\}); you should use \courier{fftw_malloc\} (or equivalent) to allocate the arrays you want to transform if they are larger than a few hundred elements. \question 13may:leaks FFTW seems to have a memory leak. After you create a plan, FFTW caches the information required to quickly recreate the plan. (See \qref savePlans) It also maintains a small amount of other persistent memory. You can deallocate all of FFTW's internally allocated memory, if you wish, by calling \courier{fftw_cleanup()\}, as documented in the manual. \question 16may:allzero The output of FFTW's transform is all zeros. You should initialize your input array \italic{after\} creating the plan, unless you use \courier{FFTW_ESTIMATE\}: planning with \courier{FFTW_MEASURE\} or \courier{FFTW_PATIENT\} overwrites the input/output arrays, as described in the manual. \question 05sep:vbetalia How do I call FFTW from the Microsoft language du jour? Please \italic{do not\} ask us Windows-specific questions. We do not use Windows. We know nothing about Visual Basic, Visual C++, or .NET. Please find the appropriate Usenet discussion group and ask your question there. See also \qref runOnWindows. \question 15oct:pruned Can I compute only a subset of the DFT outputs? In general, no, an FFT intrinsically computes all outputs from all inputs. In principle, there is something called a \italic{pruned FFT\} that can do what you want, but to compute K outputs out of N the complexity is in general O(N log K) instead of O(N log N), thus saving only a small additive factor in the log. (The same argument holds if you instead have only K nonzero inputs.) There are some specific cases in which you can get the O(N log K) performance benefits easily, however, by combining a few ordinary FFTs. In particular, the case where you want the first K outputs, where K divides N, can be handled by performing N/K transforms of size K and then summing the outputs multiplied by appropriate phase factors. For more details, see \docref{pruned FFTs with FFTW\}. There are also some algorithms that compute pruned transforms \italic{approximately\}, but they are beyond the scope of this FAQ. \question 21jan:transpose Can I use FFTW's routines for in-place and out-of-place matrix transposition? You can use the FFTW guru interface to create a rank-0 transform of vector rank 2 where the vector strides are transposed. (A rank-0 transform is equivalent to a 1D transform of size 1, which. just copies the input into the output.) Specifying the same location for the input and output makes the transpose in-place. For double-valued data stored in row-major format, plan creation looks like this: \verbatim fftw_plan plan_transpose(int rows, int cols, double *in, double *out) { const unsigned flags = FFTW_ESTIMATE; /* other flags are possible */ fftw_iodim howmany_dims[2]; howmany_dims[0].n = rows; howmany_dims[0].is = cols; howmany_dims[0].os = 1; howmany_dims[1].n = cols; howmany_dims[1].is = 1; howmany_dims[1].os = rows; return fftw_plan_guru_r2r(/*rank=*/ 0, /*dims=*/ NULL, /*howmany_rank=*/ 2, howmany_dims, in, out, /*kind=*/ NULL, flags); } \endverbatim (This entry was written by Rhys Ulerich.) \comment ###################################################################### \section Internals of FFTW \question 26aug:howworks How does FFTW work? The innovation (if it can be so called) in FFTW consists in having a variety of composable \italic{solvers\}, representing different FFT algorithms and implementation strategies, whose combination into a particular \italic{plan\} for a given size can be determined at runtime according to the characteristics of your machine/compiler. This peculiar software architecture allows FFTW to adapt itself to almost any machine. For more details (albeit somewhat outdated), see the paper "FFTW: An Adaptive Software Architecture for the FFT", by M. Frigo and S. G. Johnson, \italic{Proc. ICASSP\} 3, 1381 (1998), also available at \docref{the FFTW web page\}. \question 26aug:whyfast Why is FFTW so fast? This is a complex question, and there is no simple answer. In fact, the authors do not fully know the answer, either. In addition to many small performance hacks throughout FFTW, there are three general reasons for FFTW's speed. \call startlist \call item FFTW uses a variety of FFT algorithms and implementation styles that can be arbitrarily composed to adapt itself to a machine. See \qref howworks. \call item FFTW uses a code generator to produce highly-optimized routines for computing small transforms. \call item FFTW uses explicit divide-and-conquer to take advantage of the memory hierarchy. \call endlist For more details (albeit somewhat outdated), see the paper "FFTW: An Adaptive Software Architecture for the FFT", by M. Frigo and S. G. Johnson, \italic{Proc. ICASSP\} 3, 1381 (1998), available along with other references at \docref{the FFTW web page\}. \comment ###################################################################### \section Known bugs \question 27aug:rfftwndbug FFTW 1.1 crashes in rfftwnd on Linux. This bug was fixed in FFTW 1.2. There was a bug in \courier{rfftwnd\} causing an incorrect amount of memory to be allocated. The bug showed up in Linux with libc-5.3.12 (and nowhere else that we know of). \question 15oct:fftwmpibug The MPI transforms in FFTW 1.2 give incorrect results/leak memory. These bugs were corrected in FFTW 1.2.1. The MPI transforms (really, just the transpose routines) in FFTW 1.2 had bugs that could cause errors in some situations. \question 05nov:testsingbug The test programs in FFTW 1.2.1 fail when I change FFTW to use single precision. This bug was fixed in FFTW 1.3. (Older versions of FFTW did work in single precision, but the test programs didn't--the error tolerances in the tests were set for double precision.) \question 24mar:teststoobig The test program in FFTW 1.2.1 fails for n > 46340. This bug was fixed in FFTW 1.3. FFTW 1.2.1 produced the right answer, but the test program was wrong. For large n, n*n in the naive transform that we used for comparison overflows 32 bit integer precision, breaking the test. \question 24aug:linuxthreads The threaded code fails on Linux Redhat 5.0 We had problems with glibc-2.0.5. The code should work with glibc-2.0.7. \question 26sep:bigrfftwnd FFTW 2.0's rfftwnd fails for rank > 1 transforms with a final dimension >= 65536. This bug was fixed in FFTW 2.0.1. (There was a 32-bit integer overflow due to a poorly-parenthesized expression.) \question 26mar:primebug FFTW 2.0's complex transforms give the wrong results with prime factors 17 to 97. There was a bug in the complex transforms that could cause incorrect results under (hopefully rare) circumstances for lengths with intermediate-size prime factors (17-97). This bug was fixed in FFTW 2.1.1. \question 05apr:mpichbug FFTW 2.1.1's MPI test programs crash with MPICH. This bug was fixed in FFTW 2.1.2. The 2.1/2.1.1 MPI test programs crashed when using the MPICH implementation of MPI with the \courier{ch_p4\} device (TCP/IP); the transforms themselves worked fine. \question 25may:aixthreadbug FFTW 2.1.2's multi-threaded transforms don't work on AIX. This bug was fixed in FFTW 2.1.3. The multi-threaded transforms in previous versions didn't work with AIX's \courier{pthreads\} implementation, which idiosyncratically creates threads in detached (non-joinable) mode by default. \question 27sep:bigprimebug FFTW 2.1.2's complex transforms give incorrect results for large prime sizes. This bug was fixed in FFTW 2.1.3. FFTW's complex-transform algorithm for prime sizes (in versions 2.0 to 2.1.2) had an integer overflow problem that caused incorrect results for many primes greater than 32768 (on 32-bit machines). (Sizes without large prime factors are not affected.) \question 25may:solaristhreadbug FFTW 2.1.3's multi-threaded transforms don't give any speedup on Solaris. This bug was fixed in FFTW 2.1.4. (By default, Solaris creates threads that do not parallelize over multiple processors, so one has to request the proper behavior specifically.) \question 03may:aixflags FFTW 2.1.3 crashes on AIX. The FFTW 2.1.3 \courier{configure\} script picked incorrect compiler flags for the \courier{xlc\} compiler on newer IBM processors. This is fixed in FFTW 2.1.4. \comment Here it ends! fftw-3.3.3/doc/FAQ/fftw-faq.ascii0000644000175000017500000011213212054411114013301 00000000000000 FFTW FREQUENTLY ASKED QUESTIONS WITH ANSWERS 25 Nov 2012 Matteo Frigo Steven G. Johnson This is the list of Frequently Asked Questions about FFTW, a collection of fast C routines for computing the Discrete Fourier Transform in one or more dimensions. =============================================================================== Index Section 1. Introduction and General Information Q1.1 What is FFTW? Q1.2 How do I obtain FFTW? Q1.3 Is FFTW free software? Q1.4 What is this about non-free licenses? Q1.5 In the West? I thought MIT was in the East? Section 2. Installing FFTW Q2.1 Which systems does FFTW run on? Q2.2 Does FFTW run on Windows? Q2.3 My compiler has trouble with FFTW. Q2.4 FFTW does not compile on Solaris, complaining about const. Q2.5 What's the difference between --enable-3dnow and --enable-k7? Q2.6 What's the difference between the fma and the non-fma versions? Q2.7 Which language is FFTW written in? Q2.8 Can I call FFTW from Fortran? Q2.9 Can I call FFTW from C++? Q2.10 Why isn't FFTW written in Fortran/C++? Q2.11 How do I compile FFTW to run in single precision? Q2.12 --enable-k7 does not work on x86-64 Section 3. Using FFTW Q3.1 Why not support the FFTW 2 interface in FFTW 3? Q3.2 Why do FFTW 3 plans encapsulate the input/output arrays and not ju Q3.3 FFTW seems really slow. Q3.4 FFTW slows down after repeated calls. Q3.5 An FFTW routine is crashing when I call it. Q3.6 My Fortran program crashes when calling FFTW. Q3.7 FFTW gives results different from my old FFT. Q3.8 FFTW gives different results between runs Q3.9 Can I save FFTW's plans? Q3.10 Why does your inverse transform return a scaled result? Q3.11 How can I make FFTW put the origin (zero frequency) at the center Q3.12 How do I FFT an image/audio file in *foobar* format? Q3.13 My program does not link (on Unix). Q3.14 I included your header, but linking still fails. Q3.15 My program crashes, complaining about stack space. Q3.16 FFTW seems to have a memory leak. Q3.17 The output of FFTW's transform is all zeros. Q3.18 How do I call FFTW from the Microsoft language du jour? Q3.19 Can I compute only a subset of the DFT outputs? Q3.20 Can I use FFTW's routines for in-place and out-of-place matrix tra Section 4. Internals of FFTW Q4.1 How does FFTW work? Q4.2 Why is FFTW so fast? Section 5. Known bugs Q5.1 FFTW 1.1 crashes in rfftwnd on Linux. Q5.2 The MPI transforms in FFTW 1.2 give incorrect results/leak memory. Q5.3 The test programs in FFTW 1.2.1 fail when I change FFTW to use sin Q5.4 The test program in FFTW 1.2.1 fails for n > 46340. Q5.5 The threaded code fails on Linux Redhat 5.0 Q5.6 FFTW 2.0's rfftwnd fails for rank > 1 transforms with a final dime Q5.7 FFTW 2.0's complex transforms give the wrong results with prime fa Q5.8 FFTW 2.1.1's MPI test programs crash with MPICH. Q5.9 FFTW 2.1.2's multi-threaded transforms don't work on AIX. Q5.10 FFTW 2.1.2's complex transforms give incorrect results for large p Q5.11 FFTW 2.1.3's multi-threaded transforms don't give any speedup on S Q5.12 FFTW 2.1.3 crashes on AIX. =============================================================================== Section 1. Introduction and General Information Q1.1 What is FFTW? Q1.2 How do I obtain FFTW? Q1.3 Is FFTW free software? Q1.4 What is this about non-free licenses? Q1.5 In the West? I thought MIT was in the East? ------------------------------------------------------------------------------- Question 1.1. What is FFTW? FFTW is a free collection of fast C routines for computing the Discrete Fourier Transform in one or more dimensions. It includes complex, real, symmetric, and parallel transforms, and can handle arbitrary array sizes efficiently. FFTW is typically faster than other publically-available FFT implementations, and is even competitive with vendor-tuned libraries. (See our web page for extensive benchmarks.) To achieve this performance, FFTW uses novel code-generation and runtime self-optimization techniques (along with many other tricks). ------------------------------------------------------------------------------- Question 1.2. How do I obtain FFTW? FFTW can be found at the FFTW web page. You can also retrieve it from ftp.fftw.org in /pub/fftw. ------------------------------------------------------------------------------- Question 1.3. Is FFTW free software? Starting with version 1.3, FFTW is Free Software in the technical sense defined by the Free Software Foundation (see Categories of Free and Non-Free Software), and is distributed under the terms of the GNU General Public License. Previous versions of FFTW were distributed without fee for noncommercial use, but were not technically ``free.'' Non-free licenses for FFTW are also available that permit different terms of use than the GPL. ------------------------------------------------------------------------------- Question 1.4. What is this about non-free licenses? The non-free licenses are for companies that wish to use FFTW in their products but are unwilling to release their software under the GPL (which would require them to release source code and allow free redistribution). Such users can purchase an unlimited-use license from MIT. Contact us for more details. We could instead have released FFTW under the LGPL, or even disallowed non-Free usage. Suffice it to say, however, that MIT owns the copyright to FFTW and they only let us GPL it because we convinced them that it would neither affect their licensing revenue nor irritate existing licensees. ------------------------------------------------------------------------------- Question 1.5. In the West? I thought MIT was in the East? Not to an Italian. You could say that we're a Spaghetti Western (with apologies to Sergio Leone). =============================================================================== Section 2. Installing FFTW Q2.1 Which systems does FFTW run on? Q2.2 Does FFTW run on Windows? Q2.3 My compiler has trouble with FFTW. Q2.4 FFTW does not compile on Solaris, complaining about const. Q2.5 What's the difference between --enable-3dnow and --enable-k7? Q2.6 What's the difference between the fma and the non-fma versions? Q2.7 Which language is FFTW written in? Q2.8 Can I call FFTW from Fortran? Q2.9 Can I call FFTW from C++? Q2.10 Why isn't FFTW written in Fortran/C++? Q2.11 How do I compile FFTW to run in single precision? Q2.12 --enable-k7 does not work on x86-64 ------------------------------------------------------------------------------- Question 2.1. Which systems does FFTW run on? FFTW is written in ANSI C, and should work on any system with a decent C compiler. (See also Q2.2 `Does FFTW run on Windows?', Q2.3 `My compiler has trouble with FFTW.'.) FFTW can also take advantage of certain hardware-specific features, such as cycle counters and SIMD instructions, but this is optional. ------------------------------------------------------------------------------- Question 2.2. Does FFTW run on Windows? Yes, many people have reported successfully using FFTW on Windows with various compilers. FFTW was not developed on Windows, but the source code is essentially straight ANSI C. See also the FFTW Windows installation notes, Q2.3 `My compiler has trouble with FFTW.', and Q3.18 `How do I call FFTW from the Microsoft language du jour?'. ------------------------------------------------------------------------------- Question 2.3. My compiler has trouble with FFTW. Complain fiercely to the vendor of the compiler. We have successfully used gcc 3.2.x on x86 and PPC, a recent Compaq C compiler for Alpha, version 6 of IBM's xlc compiler for AIX, Intel's icc versions 5-7, and Sun WorkShop cc version 6. FFTW is likely to push compilers to their limits, however, and several compiler bugs have been exposed by FFTW. A partial list follows. gcc 2.95.x for Solaris/SPARC produces incorrect code for the test program (workaround: recompile the libbench2 directory with -O2). NetBSD/macppc 1.6 comes with a gcc version that also miscompiles the test program. (Please report a workaround if you know one.) gcc 3.2.3 for ARM reportedly crashes during compilation. This bug is reportedly fixed in later versions of gcc. Versions 8.0 and 8.1 of Intel's icc falsely claim to be gcc, so you should specify CC="icc -no-gcc"; this is automatic in FFTW 3.1. icc-8.0.066 reportely produces incorrect code for FFTW 2.1.5, but is fixed in version 8.1. icc-7.1 compiler build 20030402Z appears to produce incorrect dependencies, causing the compilation to fail. icc-7.1 build 20030307Z appears to work fine. (Use icc -V to check which build you have.) As of 2003/04/18, build 20030402Z appears not to be available any longer on Intel's website, whereas the older build 20030307Z is available. ranlib of GNU binutils 2.9.1 on Irix has been observed to corrupt the FFTW libraries, causing a link failure when FFTW is compiled. Since ranlib is completely superfluous on Irix, we suggest deleting it from your system and replacing it with a symbolic link to /bin/echo. If support for SIMD instructions is enabled in FFTW, further compiler problems may appear: gcc 3.4.[0123] for x86 produces incorrect SSE2 code for FFTW when -O2 (the best choice for FFTW) is used, causing FFTW to crash (make check crashes). This bug is fixed in gcc 3.4.4. On x86_64 (amd64/em64t), gcc 3.4.4 reportedly still has a similar problem, but this is fixed as of gcc 3.4.6. gcc-3.2 for x86 produces incorrect SIMD code if -O3 is used. The same compiler produces incorrect SIMD code if no optimization is used, too. When using gcc-3.2, it is a good idea not to change the default CFLAGS selected by the configure script. Some 3.0.x and 3.1.x versions of gcc on x86 may crash. gcc so-called 2.96 shipping with RedHat 7.3 crashes when compiling SIMD code. In both cases, please upgrade to gcc-3.2 or later. Intel's icc 6.0 misaligns SSE constants, but FFTW has a workaround. icc 8.x fails to compile FFTW 3.0.x because it falsely claims to be gcc; we believe this to be a bug in icc, but FFTW 3.1 has a workaround. Visual C++ 2003 reportedly produces incorrect code for SSE/SSE2 when compiling FFTW. This bug was reportedly fixed in VC++ 2005; alternatively, you could switch to the Intel compiler. VC++ 6.0 also reportedly produces incorrect code for the file reodft11e-r2hc-odd.c unless optimizations are disabled for that file. gcc 2.95 on MacOS X miscompiles AltiVec code (fixed in later versions). gcc 3.2.x miscompiles AltiVec permutations, but FFTW has a workaround. gcc 4.0.1 on MacOS for Intel crashes when compiling FFTW; a workaround is to compile one file without optimization: cd kernel; make CFLAGS=" " trig.lo. gcc 4.1.1 reportedly crashes when compiling FFTW for MIPS; the workaround is to compile the file it crashes on (t2_64.c) with a lower optimization level. gcc versions 4.1.2 to 4.2.0 for x86 reportedly miscompile FFTW 3.1's test program, causing make check to crash (gcc bug #26528). The bug was reportedly fixed in gcc version 4.2.1 and later. A workaround is to compile libbench2/verify-lib.c without optimization. ------------------------------------------------------------------------------- Question 2.4. FFTW does not compile on Solaris, complaining about const. We know that at least on Solaris 2.5.x with Sun's compilers 4.2 you might get error messages from make such as "./fftw.h", line 88: warning: const is a keyword in ANSI C This is the case when the configure script reports that const does not work: checking for working const... (cached) no You should be aware that Solaris comes with two compilers, namely, /opt/SUNWspro/SC4.2/bin/cc and /usr/ucb/cc. The latter compiler is non-ANSI. Indeed, it is a perverse shell script that calls the real compiler in non-ANSI mode. In order to compile FFTW, change your path so that the right cc is used. To know whether your compiler is the right one, type cc -V. If the compiler prints ``ucbcc'', as in ucbcc: WorkShop Compilers 4.2 30 Oct 1996 C 4.2 then the compiler is wrong. The right message is something like cc: WorkShop Compilers 4.2 30 Oct 1996 C 4.2 ------------------------------------------------------------------------------- Question 2.5. What's the difference between --enable-3dnow and --enable-k7? --enable-k7 enables 3DNow! instructions on K7 processors (AMD Athlon and its variants). K7 support is provided by assembly routines generated by a special purpose compiler. As of fftw-3.2, --enable-k7 is no longer supported. --enable-3dnow enables generic 3DNow! support using gcc builtin functions. This works on earlier AMD processors, but it is not as fast as our special assembly routines. As of fftw-3.1, --enable-3dnow is no longer supported. ------------------------------------------------------------------------------- Question 2.6. What's the difference between the fma and the non-fma versions? The fma version tries to exploit the fused multiply-add instructions implemented in many processors such as PowerPC, ia-64, and MIPS. The two FFTW packages are otherwise identical. In FFTW 3.1, the fma and non-fma versions were merged together into a single package, and the configure script attempts to automatically guess which version to use. The FFTW 3.1 configure script enables fma by default on PowerPC, Itanium, and PA-RISC, and disables it otherwise. You can force one or the other by using the --enable-fma or --disable-fma flag for configure. Definitely use fma if you have a PowerPC-based system with gcc (or IBM xlc). This includes all GNU/Linux systems for PowerPC and the older PowerPC-based MacOS systems. Also use it on PA-RISC and Itanium with the HP/UX compiler. Definitely do not use the fma version if you have an ia-32 processor (Intel, AMD, MacOS on Intel, etcetera). For other architectures/compilers, the situation is not so clear. For example, ia-64 has the fma instruction, but gcc-3.2 appears not to exploit it correctly. Other compilers may do the right thing, but we have not tried them. Please send us your feedback so that we can update this FAQ entry. ------------------------------------------------------------------------------- Question 2.7. Which language is FFTW written in? FFTW is written in ANSI C. Most of the code, however, was automatically generated by a program called genfft, written in the Objective Caml dialect of ML. You do not need to know ML or to have an Objective Caml compiler in order to use FFTW. genfft is provided with the FFTW sources, which means that you can play with the code generator if you want. In this case, you need a working Objective Caml system. Objective Caml is available from the Caml web page. ------------------------------------------------------------------------------- Question 2.8. Can I call FFTW from Fortran? Yes, FFTW (versions 1.3 and higher) contains a Fortran-callable interface, documented in the FFTW manual. By default, FFTW configures its Fortran interface to work with the first compiler it finds, e.g. g77. To configure for a different, incompatible Fortran compiler foobar, use ./configure F77=foobar when installing FFTW. (In the case of g77, however, FFTW 3.x also includes an extra set of Fortran-callable routines with one less underscore at the end of identifiers, which should cover most other Fortran compilers on Linux at least.) ------------------------------------------------------------------------------- Question 2.9. Can I call FFTW from C++? Most definitely. FFTW should compile and/or link under any C++ compiler. Moreover, it is likely that the C++ template class is bit-compatible with FFTW's complex-number format (see the FFTW manual for more details). ------------------------------------------------------------------------------- Question 2.10. Why isn't FFTW written in Fortran/C++? Because we don't like those languages, and neither approaches the portability of C. ------------------------------------------------------------------------------- Question 2.11. How do I compile FFTW to run in single precision? On a Unix system: configure --enable-float. On a non-Unix system: edit config.h to #define the symbol FFTW_SINGLE (for FFTW 3.x). In both cases, you must then recompile FFTW. In FFTW 3, all FFTW identifiers will then begin with fftwf_ instead of fftw_. ------------------------------------------------------------------------------- Question 2.12. --enable-k7 does not work on x86-64 Support for --enable-k7 was discontinued in fftw-3.2. The fftw-3.1 release supports --enable-k7. This option only works on 32-bit x86 machines that implement 3DNow!, including the AMD Athlon and the AMD Opteron in 32-bit mode. --enable-k7 does not work on AMD Opteron in 64-bit mode. Use --enable-sse for x86-64 machines. FFTW supports 3DNow! by means of assembly code generated by a special-purpose compiler. It is hard to produce assembly code that works in both 32-bit and 64-bit mode. =============================================================================== Section 3. Using FFTW Q3.1 Why not support the FFTW 2 interface in FFTW 3? Q3.2 Why do FFTW 3 plans encapsulate the input/output arrays and not ju Q3.3 FFTW seems really slow. Q3.4 FFTW slows down after repeated calls. Q3.5 An FFTW routine is crashing when I call it. Q3.6 My Fortran program crashes when calling FFTW. Q3.7 FFTW gives results different from my old FFT. Q3.8 FFTW gives different results between runs Q3.9 Can I save FFTW's plans? Q3.10 Why does your inverse transform return a scaled result? Q3.11 How can I make FFTW put the origin (zero frequency) at the center Q3.12 How do I FFT an image/audio file in *foobar* format? Q3.13 My program does not link (on Unix). Q3.14 I included your header, but linking still fails. Q3.15 My program crashes, complaining about stack space. Q3.16 FFTW seems to have a memory leak. Q3.17 The output of FFTW's transform is all zeros. Q3.18 How do I call FFTW from the Microsoft language du jour? Q3.19 Can I compute only a subset of the DFT outputs? Q3.20 Can I use FFTW's routines for in-place and out-of-place matrix tra ------------------------------------------------------------------------------- Question 3.1. Why not support the FFTW 2 interface in FFTW 3? FFTW 3 has semantics incompatible with earlier versions: its plans can only be used for a given stride, multiplicity, and other characteristics of the input and output arrays; these stronger semantics are necessary for performance reasons. Thus, it is impossible to efficiently emulate the older interface (whose plans can be used for any transform of the same size). We believe that it should be possible to upgrade most programs without any difficulty, however. ------------------------------------------------------------------------------- Question 3.2. Why do FFTW 3 plans encapsulate the input/output arrays and not just the algorithm? There are several reasons: * It was important for performance reasons that the plan be specific to array characteristics like the stride (and alignment, for SIMD), and requiring that the user maintain these invariants is error prone. * In most high-performance applications, as far as we can tell, you are usually transforming the same array over and over, so FFTW's semantics should not be a burden. * If you need to transform another array of the same size, creating a new plan once the first exists is a cheap operation. * If you need to transform many arrays of the same size at once, you should really use the plan_many routines in FFTW's "advanced" interface. * If the abovementioned array characteristics are the same, you are willing to pay close attention to the documentation, and you really need to, we provide a "new-array execution" interface to apply a plan to a new array. ------------------------------------------------------------------------------- Question 3.3. FFTW seems really slow. You are probably recreating the plan before every transform, rather than creating it once and reusing it for all transforms of the same size. FFTW is designed to be used in the following way: * First, you create a plan. This will take several seconds. * Then, you reuse the plan many times to perform FFTs. These are fast. If you don't need to compute many transforms and the time for the planner is significant, you have two options. First, you can use the FFTW_ESTIMATE option in the planner, which uses heuristics instead of runtime measurements and produces a good plan in a short time. Second, you can use the wisdom feature to precompute the plan; see Q3.9 `Can I save FFTW's plans?' ------------------------------------------------------------------------------- Question 3.4. FFTW slows down after repeated calls. Probably, NaNs or similar are creeping into your data, and the slowdown is due to the resulting floating-point exceptions. For example, be aware that repeatedly FFTing the same array is a diverging process (because FFTW computes the unnormalized transform). ------------------------------------------------------------------------------- Question 3.5. An FFTW routine is crashing when I call it. Did the FFTW test programs pass (make check, or cd tests; make bigcheck if you want to be paranoid)? If so, you almost certainly have a bug in your own code. For example, you could be passing invalid arguments (such as wrongly-sized arrays) to FFTW, or you could simply have memory corruption elsewhere in your program that causes random crashes later on. Please don't complain to us unless you can come up with a minimal self-contained program (preferably under 30 lines) that illustrates the problem. ------------------------------------------------------------------------------- Question 3.6. My Fortran program crashes when calling FFTW. As described in the manual, on 64-bit machines you must store the plans in variables large enough to hold a pointer, for example integer*8. We recommend using integer*8 on 32-bit machines as well, to simplify porting. ------------------------------------------------------------------------------- Question 3.7. FFTW gives results different from my old FFT. People follow many different conventions for the DFT, and you should be sure to know the ones that we use (described in the FFTW manual). In particular, you should be aware that the FFTW_FORWARD/FFTW_BACKWARD directions correspond to signs of -1/+1 in the exponent of the DFT definition. (*Numerical Recipes* uses the opposite convention.) You should also know that we compute an unnormalized transform. In contrast, Matlab is an example of program that computes a normalized transform. See Q3.10 `Why does your inverse transform return a scaled result?'. Finally, note that floating-point arithmetic is not exact, so different FFT algorithms will give slightly different results (on the order of the numerical accuracy; typically a fractional difference of 1e-15 or so in double precision). ------------------------------------------------------------------------------- Question 3.8. FFTW gives different results between runs If you use FFTW_MEASURE or FFTW_PATIENT mode, then the algorithm FFTW employs is not deterministic: it depends on runtime performance measurements. This will cause the results to vary slightly from run to run. However, the differences should be slight, on the order of the floating-point precision, and therefore should have no practical impact on most applications. If you use saved plans (wisdom) or FFTW_ESTIMATE mode, however, then the algorithm is deterministic and the results should be identical between runs. ------------------------------------------------------------------------------- Question 3.9. Can I save FFTW's plans? Yes. Starting with version 1.2, FFTW provides the wisdom mechanism for saving plans; see the FFTW manual. ------------------------------------------------------------------------------- Question 3.10. Why does your inverse transform return a scaled result? Computing the forward transform followed by the backward transform (or vice versa) yields the original array scaled by the size of the array. (For multi-dimensional transforms, the size of the array is the product of the dimensions.) We could, instead, have chosen a normalization that would have returned the unscaled array. Or, to accomodate the many conventions in this matter, the transform routines could have accepted a "scale factor" parameter. We did not do this, however, for two reasons. First, we didn't want to sacrifice performance in the common case where the scale factor is 1. Second, in real applications the FFT is followed or preceded by some computation on the data, into which the scale factor can typically be absorbed at little or no cost. ------------------------------------------------------------------------------- Question 3.11. How can I make FFTW put the origin (zero frequency) at the center of its output? For human viewing of a spectrum, it is often convenient to put the origin in frequency space at the center of the output array, rather than in the zero-th element (the default in FFTW). If all of the dimensions of your array are even, you can accomplish this by simply multiplying each element of the input array by (-1)^(i + j + ...), where i, j, etcetera are the indices of the element. (This trick is a general property of the DFT, and is not specific to FFTW.) ------------------------------------------------------------------------------- Question 3.12. How do I FFT an image/audio file in *foobar* format? FFTW performs an FFT on an array of floating-point values. You can certainly use it to compute the transform of an image or audio stream, but you are responsible for figuring out your data format and converting it to the form FFTW requires. ------------------------------------------------------------------------------- Question 3.13. My program does not link (on Unix). The libraries must be listed in the correct order (-lfftw3 -lm for FFTW 3.x) and *after* your program sources/objects. (The general rule is that if *A* uses *B*, then *A* must be listed before *B* in the link command.). ------------------------------------------------------------------------------- Question 3.14. I included your header, but linking still fails. You're a C++ programmer, aren't you? You have to compile the FFTW library and link it into your program, not just #include . (Yes, this is really a FAQ.) ------------------------------------------------------------------------------- Question 3.15. My program crashes, complaining about stack space. You cannot declare large arrays with automatic storage (e.g. via fftw_complex array[N]); you should use fftw_malloc (or equivalent) to allocate the arrays you want to transform if they are larger than a few hundred elements. ------------------------------------------------------------------------------- Question 3.16. FFTW seems to have a memory leak. After you create a plan, FFTW caches the information required to quickly recreate the plan. (See Q3.9 `Can I save FFTW's plans?') It also maintains a small amount of other persistent memory. You can deallocate all of FFTW's internally allocated memory, if you wish, by calling fftw_cleanup(), as documented in the manual. ------------------------------------------------------------------------------- Question 3.17. The output of FFTW's transform is all zeros. You should initialize your input array *after* creating the plan, unless you use FFTW_ESTIMATE: planning with FFTW_MEASURE or FFTW_PATIENT overwrites the input/output arrays, as described in the manual. ------------------------------------------------------------------------------- Question 3.18. How do I call FFTW from the Microsoft language du jour? Please *do not* ask us Windows-specific questions. We do not use Windows. We know nothing about Visual Basic, Visual C++, or .NET. Please find the appropriate Usenet discussion group and ask your question there. See also Q2.2 `Does FFTW run on Windows?'. ------------------------------------------------------------------------------- Question 3.19. Can I compute only a subset of the DFT outputs? In general, no, an FFT intrinsically computes all outputs from all inputs. In principle, there is something called a *pruned FFT* that can do what you want, but to compute K outputs out of N the complexity is in general O(N log K) instead of O(N log N), thus saving only a small additive factor in the log. (The same argument holds if you instead have only K nonzero inputs.) There are some specific cases in which you can get the O(N log K) performance benefits easily, however, by combining a few ordinary FFTs. In particular, the case where you want the first K outputs, where K divides N, can be handled by performing N/K transforms of size K and then summing the outputs multiplied by appropriate phase factors. For more details, see pruned FFTs with FFTW. There are also some algorithms that compute pruned transforms *approximately*, but they are beyond the scope of this FAQ. ------------------------------------------------------------------------------- Question 3.20. Can I use FFTW's routines for in-place and out-of-place matrix transposition? You can use the FFTW guru interface to create a rank-0 transform of vector rank 2 where the vector strides are transposed. (A rank-0 transform is equivalent to a 1D transform of size 1, which. just copies the input into the output.) Specifying the same location for the input and output makes the transpose in-place. For double-valued data stored in row-major format, plan creation looks like this: fftw_plan plan_transpose(int rows, int cols, double *in, double *out) { const unsigned flags = FFTW_ESTIMATE; /* other flags are possible */ fftw_iodim howmany_dims[2]; howmany_dims[0].n = rows; howmany_dims[0].is = cols; howmany_dims[0].os = 1; howmany_dims[1].n = cols; howmany_dims[1].is = 1; howmany_dims[1].os = rows; return fftw_plan_guru_r2r(/*rank=*/ 0, /*dims=*/ NULL, /*howmany_rank=*/ 2, howmany_dims, in, out, /*kind=*/ NULL, flags); } (This entry was written by Rhys Ulerich.) =============================================================================== Section 4. Internals of FFTW Q4.1 How does FFTW work? Q4.2 Why is FFTW so fast? ------------------------------------------------------------------------------- Question 4.1. How does FFTW work? The innovation (if it can be so called) in FFTW consists in having a variety of composable *solvers*, representing different FFT algorithms and implementation strategies, whose combination into a particular *plan* for a given size can be determined at runtime according to the characteristics of your machine/compiler. This peculiar software architecture allows FFTW to adapt itself to almost any machine. For more details (albeit somewhat outdated), see the paper "FFTW: An Adaptive Software Architecture for the FFT", by M. Frigo and S. G. Johnson, *Proc. ICASSP* 3, 1381 (1998), also available at the FFTW web page. ------------------------------------------------------------------------------- Question 4.2. Why is FFTW so fast? This is a complex question, and there is no simple answer. In fact, the authors do not fully know the answer, either. In addition to many small performance hacks throughout FFTW, there are three general reasons for FFTW's speed. * FFTW uses a variety of FFT algorithms and implementation styles that can be arbitrarily composed to adapt itself to a machine. See Q4.1 `How does FFTW work?'. * FFTW uses a code generator to produce highly-optimized routines for computing small transforms. * FFTW uses explicit divide-and-conquer to take advantage of the memory hierarchy. For more details (albeit somewhat outdated), see the paper "FFTW: An Adaptive Software Architecture for the FFT", by M. Frigo and S. G. Johnson, *Proc. ICASSP* 3, 1381 (1998), available along with other references at the FFTW web page. =============================================================================== Section 5. Known bugs Q5.1 FFTW 1.1 crashes in rfftwnd on Linux. Q5.2 The MPI transforms in FFTW 1.2 give incorrect results/leak memory. Q5.3 The test programs in FFTW 1.2.1 fail when I change FFTW to use sin Q5.4 The test program in FFTW 1.2.1 fails for n > 46340. Q5.5 The threaded code fails on Linux Redhat 5.0 Q5.6 FFTW 2.0's rfftwnd fails for rank > 1 transforms with a final dime Q5.7 FFTW 2.0's complex transforms give the wrong results with prime fa Q5.8 FFTW 2.1.1's MPI test programs crash with MPICH. Q5.9 FFTW 2.1.2's multi-threaded transforms don't work on AIX. Q5.10 FFTW 2.1.2's complex transforms give incorrect results for large p Q5.11 FFTW 2.1.3's multi-threaded transforms don't give any speedup on S Q5.12 FFTW 2.1.3 crashes on AIX. ------------------------------------------------------------------------------- Question 5.1. FFTW 1.1 crashes in rfftwnd on Linux. This bug was fixed in FFTW 1.2. There was a bug in rfftwnd causing an incorrect amount of memory to be allocated. The bug showed up in Linux with libc-5.3.12 (and nowhere else that we know of). ------------------------------------------------------------------------------- Question 5.2. The MPI transforms in FFTW 1.2 give incorrect results/leak memory. These bugs were corrected in FFTW 1.2.1. The MPI transforms (really, just the transpose routines) in FFTW 1.2 had bugs that could cause errors in some situations. ------------------------------------------------------------------------------- Question 5.3. The test programs in FFTW 1.2.1 fail when I change FFTW to use single precision. This bug was fixed in FFTW 1.3. (Older versions of FFTW did work in single precision, but the test programs didn't--the error tolerances in the tests were set for double precision.) ------------------------------------------------------------------------------- Question 5.4. The test program in FFTW 1.2.1 fails for n > 46340. This bug was fixed in FFTW 1.3. FFTW 1.2.1 produced the right answer, but the test program was wrong. For large n, n*n in the naive transform that we used for comparison overflows 32 bit integer precision, breaking the test. ------------------------------------------------------------------------------- Question 5.5. The threaded code fails on Linux Redhat 5.0 We had problems with glibc-2.0.5. The code should work with glibc-2.0.7. ------------------------------------------------------------------------------- Question 5.6. FFTW 2.0's rfftwnd fails for rank > 1 transforms with a final dimension >= 65536. This bug was fixed in FFTW 2.0.1. (There was a 32-bit integer overflow due to a poorly-parenthesized expression.) ------------------------------------------------------------------------------- Question 5.7. FFTW 2.0's complex transforms give the wrong results with prime factors 17 to 97. There was a bug in the complex transforms that could cause incorrect results under (hopefully rare) circumstances for lengths with intermediate-size prime factors (17-97). This bug was fixed in FFTW 2.1.1. ------------------------------------------------------------------------------- Question 5.8. FFTW 2.1.1's MPI test programs crash with MPICH. This bug was fixed in FFTW 2.1.2. The 2.1/2.1.1 MPI test programs crashed when using the MPICH implementation of MPI with the ch_p4 device (TCP/IP); the transforms themselves worked fine. ------------------------------------------------------------------------------- Question 5.9. FFTW 2.1.2's multi-threaded transforms don't work on AIX. This bug was fixed in FFTW 2.1.3. The multi-threaded transforms in previous versions didn't work with AIX's pthreads implementation, which idiosyncratically creates threads in detached (non-joinable) mode by default. ------------------------------------------------------------------------------- Question 5.10. FFTW 2.1.2's complex transforms give incorrect results for large prime sizes. This bug was fixed in FFTW 2.1.3. FFTW's complex-transform algorithm for prime sizes (in versions 2.0 to 2.1.2) had an integer overflow problem that caused incorrect results for many primes greater than 32768 (on 32-bit machines). (Sizes without large prime factors are not affected.) ------------------------------------------------------------------------------- Question 5.11. FFTW 2.1.3's multi-threaded transforms don't give any speedup on Solaris. This bug was fixed in FFTW 2.1.4. (By default, Solaris creates threads that do not parallelize over multiple processors, so one has to request the proper behavior specifically.) ------------------------------------------------------------------------------- Question 5.12. FFTW 2.1.3 crashes on AIX. The FFTW 2.1.3 configure script picked incorrect compiler flags for the xlc compiler on newer IBM processors. This is fixed in FFTW 2.1.4. fftw-3.3.3/doc/FAQ/fftw-faq.html/0000755000175000017500000000000012054411114013312 500000000000000fftw-3.3.3/doc/FAQ/fftw-faq.html/index.html0000644000175000017500000001647712054411114015246 00000000000000 FFTW Frequently Asked Questions with Answers

FFTW Frequently Asked Questions with Answers

This is the list of Frequently Asked Questions about FFTW, a collection of fast C routines for computing the Discrete Fourier Transform in one or more dimensions.

Index

Matteo Frigo and Steven G. Johnson / fftw@fftw.org - 25 November 2012

Extracted from FFTW Frequently Asked Questions with Answers, Copyright © 2012 Matteo Frigo and Massachusetts Institute of Technology. fftw-3.3.3/doc/FAQ/fftw-faq.html/section1.html0000644000175000017500000000711412054411114015650 00000000000000 FFTW FAQ - Section 1

FFTW FAQ - Section 1
Introduction and General Information

Question 1.1. What is FFTW?

FFTW is a free collection of fast C routines for computing the Discrete Fourier Transform in one or more dimensions. It includes complex, real, symmetric, and parallel transforms, and can handle arbitrary array sizes efficiently. FFTW is typically faster than other publically-available FFT implementations, and is even competitive with vendor-tuned libraries. (See our web page for extensive benchmarks.) To achieve this performance, FFTW uses novel code-generation and runtime self-optimization techniques (along with many other tricks).

Question 1.2. How do I obtain FFTW?

FFTW can be found at the FFTW web page. You can also retrieve it from ftp.fftw.org in /pub/fftw.

Question 1.3. Is FFTW free software?

Starting with version 1.3, FFTW is Free Software in the technical sense defined by the Free Software Foundation (see Categories of Free and Non-Free Software), and is distributed under the terms of the GNU General Public License. Previous versions of FFTW were distributed without fee for noncommercial use, but were not technically ``free.''

Non-free licenses for FFTW are also available that permit different terms of use than the GPL.

Question 1.4. What is this about non-free licenses?

The non-free licenses are for companies that wish to use FFTW in their products but are unwilling to release their software under the GPL (which would require them to release source code and allow free redistribution). Such users can purchase an unlimited-use license from MIT. Contact us for more details.

We could instead have released FFTW under the LGPL, or even disallowed non-Free usage. Suffice it to say, however, that MIT owns the copyright to FFTW and they only let us GPL it because we convinced them that it would neither affect their licensing revenue nor irritate existing licensees.

Question 1.5. In the West? I thought MIT was in the East?

Not to an Italian. You could say that we're a Spaghetti Western (with apologies to Sergio Leone).
Next: Installing FFTW.
Return to contents.

Matteo Frigo and Steven G. Johnson / fftw@fftw.org - 25 November 2012

Extracted from FFTW Frequently Asked Questions with Answers, Copyright © 2012 Matteo Frigo and Massachusetts Institute of Technology. fftw-3.3.3/doc/FAQ/fftw-faq.html/section2.html0000644000175000017500000003272712054411114015661 00000000000000 FFTW FAQ - Section 2

FFTW FAQ - Section 2
Installing FFTW

Question 2.1. Which systems does FFTW run on?

FFTW is written in ANSI C, and should work on any system with a decent C compiler. (See also Q2.2 `Does FFTW run on Windows?', Q2.3 `My compiler has trouble with FFTW.'.) FFTW can also take advantage of certain hardware-specific features, such as cycle counters and SIMD instructions, but this is optional.

Question 2.2. Does FFTW run on Windows?

Yes, many people have reported successfully using FFTW on Windows with various compilers. FFTW was not developed on Windows, but the source code is essentially straight ANSI C. See also the FFTW Windows installation notes, Q2.3 `My compiler has trouble with FFTW.', and Q3.18 `How do I call FFTW from the Microsoft language du jour?'.

Question 2.3. My compiler has trouble with FFTW.

Complain fiercely to the vendor of the compiler.

We have successfully used gcc 3.2.x on x86 and PPC, a recent Compaq C compiler for Alpha, version 6 of IBM's xlc compiler for AIX, Intel's icc versions 5-7, and Sun WorkShop cc version 6.

FFTW is likely to push compilers to their limits, however, and several compiler bugs have been exposed by FFTW. A partial list follows.

gcc 2.95.x for Solaris/SPARC produces incorrect code for the test program (workaround: recompile the libbench2 directory with -O2).

NetBSD/macppc 1.6 comes with a gcc version that also miscompiles the test program. (Please report a workaround if you know one.)

gcc 3.2.3 for ARM reportedly crashes during compilation. This bug is reportedly fixed in later versions of gcc.

Versions 8.0 and 8.1 of Intel's icc falsely claim to be gcc, so you should specify CC="icc -no-gcc"; this is automatic in FFTW 3.1. icc-8.0.066 reportely produces incorrect code for FFTW 2.1.5, but is fixed in version 8.1. icc-7.1 compiler build 20030402Z appears to produce incorrect dependencies, causing the compilation to fail. icc-7.1 build 20030307Z appears to work fine. (Use icc -V to check which build you have.) As of 2003/04/18, build 20030402Z appears not to be available any longer on Intel's website, whereas the older build 20030307Z is available.

ranlib of GNU binutils 2.9.1 on Irix has been observed to corrupt the FFTW libraries, causing a link failure when FFTW is compiled. Since ranlib is completely superfluous on Irix, we suggest deleting it from your system and replacing it with a symbolic link to /bin/echo.

If support for SIMD instructions is enabled in FFTW, further compiler problems may appear:

gcc 3.4.[0123] for x86 produces incorrect SSE2 code for FFTW when -O2 (the best choice for FFTW) is used, causing FFTW to crash (make check crashes). This bug is fixed in gcc 3.4.4. On x86_64 (amd64/em64t), gcc 3.4.4 reportedly still has a similar problem, but this is fixed as of gcc 3.4.6.

gcc-3.2 for x86 produces incorrect SIMD code if -O3 is used. The same compiler produces incorrect SIMD code if no optimization is used, too. When using gcc-3.2, it is a good idea not to change the default CFLAGS selected by the configure script.

Some 3.0.x and 3.1.x versions of gcc on x86 may crash. gcc so-called 2.96 shipping with RedHat 7.3 crashes when compiling SIMD code. In both cases, please upgrade to gcc-3.2 or later.

Intel's icc 6.0 misaligns SSE constants, but FFTW has a workaround. icc 8.x fails to compile FFTW 3.0.x because it falsely claims to be gcc; we believe this to be a bug in icc, but FFTW 3.1 has a workaround.

Visual C++ 2003 reportedly produces incorrect code for SSE/SSE2 when compiling FFTW. This bug was reportedly fixed in VC++ 2005; alternatively, you could switch to the Intel compiler. VC++ 6.0 also reportedly produces incorrect code for the file reodft11e-r2hc-odd.c unless optimizations are disabled for that file.

gcc 2.95 on MacOS X miscompiles AltiVec code (fixed in later versions). gcc 3.2.x miscompiles AltiVec permutations, but FFTW has a workaround. gcc 4.0.1 on MacOS for Intel crashes when compiling FFTW; a workaround is to compile one file without optimization: cd kernel; make CFLAGS=" " trig.lo.

gcc 4.1.1 reportedly crashes when compiling FFTW for MIPS; the workaround is to compile the file it crashes on (t2_64.c) with a lower optimization level.

gcc versions 4.1.2 to 4.2.0 for x86 reportedly miscompile FFTW 3.1's test program, causing make check to crash (gcc bug #26528). The bug was reportedly fixed in gcc version 4.2.1 and later. A workaround is to compile libbench2/verify-lib.c without optimization.

Question 2.4. FFTW does not compile on Solaris, complaining about const.

We know that at least on Solaris 2.5.x with Sun's compilers 4.2 you might get error messages from make such as

"./fftw.h", line 88: warning: const is a keyword in ANSI C

This is the case when the configure script reports that const does not work:

checking for working const... (cached) no

You should be aware that Solaris comes with two compilers, namely, /opt/SUNWspro/SC4.2/bin/cc and /usr/ucb/cc. The latter compiler is non-ANSI. Indeed, it is a perverse shell script that calls the real compiler in non-ANSI mode. In order to compile FFTW, change your path so that the right cc is used.

To know whether your compiler is the right one, type cc -V. If the compiler prints ``ucbcc'', as in

ucbcc: WorkShop Compilers 4.2 30 Oct 1996 C 4.2

then the compiler is wrong. The right message is something like

cc: WorkShop Compilers 4.2 30 Oct 1996 C 4.2

Question 2.5. What's the difference between --enable-3dnow and --enable-k7?

--enable-k7 enables 3DNow! instructions on K7 processors (AMD Athlon and its variants). K7 support is provided by assembly routines generated by a special purpose compiler. As of fftw-3.2, --enable-k7 is no longer supported.

--enable-3dnow enables generic 3DNow! support using gcc builtin functions. This works on earlier AMD processors, but it is not as fast as our special assembly routines. As of fftw-3.1, --enable-3dnow is no longer supported.

Question 2.6. What's the difference between the fma and the non-fma versions?

The fma version tries to exploit the fused multiply-add instructions implemented in many processors such as PowerPC, ia-64, and MIPS. The two FFTW packages are otherwise identical. In FFTW 3.1, the fma and non-fma versions were merged together into a single package, and the configure script attempts to automatically guess which version to use.

The FFTW 3.1 configure script enables fma by default on PowerPC, Itanium, and PA-RISC, and disables it otherwise. You can force one or the other by using the --enable-fma or --disable-fma flag for configure.

Definitely use fma if you have a PowerPC-based system with gcc (or IBM xlc). This includes all GNU/Linux systems for PowerPC and the older PowerPC-based MacOS systems. Also use it on PA-RISC and Itanium with the HP/UX compiler.

Definitely do not use the fma version if you have an ia-32 processor (Intel, AMD, MacOS on Intel, etcetera).

For other architectures/compilers, the situation is not so clear. For example, ia-64 has the fma instruction, but gcc-3.2 appears not to exploit it correctly. Other compilers may do the right thing, but we have not tried them. Please send us your feedback so that we can update this FAQ entry.

Question 2.7. Which language is FFTW written in?

FFTW is written in ANSI C. Most of the code, however, was automatically generated by a program called genfft, written in the Objective Caml dialect of ML. You do not need to know ML or to have an Objective Caml compiler in order to use FFTW.

genfft is provided with the FFTW sources, which means that you can play with the code generator if you want. In this case, you need a working Objective Caml system. Objective Caml is available from the Caml web page.

Question 2.8. Can I call FFTW from Fortran?

Yes, FFTW (versions 1.3 and higher) contains a Fortran-callable interface, documented in the FFTW manual.

By default, FFTW configures its Fortran interface to work with the first compiler it finds, e.g. g77. To configure for a different, incompatible Fortran compiler foobar, use ./configure F77=foobar when installing FFTW. (In the case of g77, however, FFTW 3.x also includes an extra set of Fortran-callable routines with one less underscore at the end of identifiers, which should cover most other Fortran compilers on Linux at least.)

Question 2.9. Can I call FFTW from C++?

Most definitely. FFTW should compile and/or link under any C++ compiler. Moreover, it is likely that the C++ <complex> template class is bit-compatible with FFTW's complex-number format (see the FFTW manual for more details).

Question 2.10. Why isn't FFTW written in Fortran/C++?

Because we don't like those languages, and neither approaches the portability of C.

Question 2.11. How do I compile FFTW to run in single precision?

On a Unix system: configure --enable-float. On a non-Unix system: edit config.h to #define the symbol FFTW_SINGLE (for FFTW 3.x). In both cases, you must then recompile FFTW. In FFTW 3, all FFTW identifiers will then begin with fftwf_ instead of fftw_.

Question 2.12. --enable-k7 does not work on x86-64

Support for --enable-k7 was discontinued in fftw-3.2.

The fftw-3.1 release supports --enable-k7. This option only works on 32-bit x86 machines that implement 3DNow!, including the AMD Athlon and the AMD Opteron in 32-bit mode. --enable-k7 does not work on AMD Opteron in 64-bit mode. Use --enable-sse for x86-64 machines.

FFTW supports 3DNow! by means of assembly code generated by a special-purpose compiler. It is hard to produce assembly code that works in both 32-bit and 64-bit mode.

Next: Using FFTW.
Back: Introduction and General Information.
Return to contents.

Matteo Frigo and Steven G. Johnson / fftw@fftw.org - 25 November 2012

Extracted from FFTW Frequently Asked Questions with Answers, Copyright © 2012 Matteo Frigo and Massachusetts Institute of Technology. fftw-3.3.3/doc/FAQ/fftw-faq.html/section3.html0000644000175000017500000003503212054411114015652 00000000000000 FFTW FAQ - Section 3

FFTW FAQ - Section 3
Using FFTW

Question 3.1. Why not support the FFTW 2 interface in FFTW 3?

FFTW 3 has semantics incompatible with earlier versions: its plans can only be used for a given stride, multiplicity, and other characteristics of the input and output arrays; these stronger semantics are necessary for performance reasons. Thus, it is impossible to efficiently emulate the older interface (whose plans can be used for any transform of the same size). We believe that it should be possible to upgrade most programs without any difficulty, however.

Question 3.2. Why do FFTW 3 plans encapsulate the input/output arrays and not just the algorithm?

There are several reasons:
  • It was important for performance reasons that the plan be specific to array characteristics like the stride (and alignment, for SIMD), and requiring that the user maintain these invariants is error prone.
  • In most high-performance applications, as far as we can tell, you are usually transforming the same array over and over, so FFTW's semantics should not be a burden.
  • If you need to transform another array of the same size, creating a new plan once the first exists is a cheap operation.
  • If you need to transform many arrays of the same size at once, you should really use the plan_many routines in FFTW's "advanced" interface.
  • If the abovementioned array characteristics are the same, you are willing to pay close attention to the documentation, and you really need to, we provide a "new-array execution" interface to apply a plan to a new array.

Question 3.3. FFTW seems really slow.

You are probably recreating the plan before every transform, rather than creating it once and reusing it for all transforms of the same size. FFTW is designed to be used in the following way:
  • First, you create a plan. This will take several seconds.
  • Then, you reuse the plan many times to perform FFTs. These are fast.
If you don't need to compute many transforms and the time for the planner is significant, you have two options. First, you can use the FFTW_ESTIMATE option in the planner, which uses heuristics instead of runtime measurements and produces a good plan in a short time. Second, you can use the wisdom feature to precompute the plan; see Q3.9 `Can I save FFTW's plans?'

Question 3.4. FFTW slows down after repeated calls.

Probably, NaNs or similar are creeping into your data, and the slowdown is due to the resulting floating-point exceptions. For example, be aware that repeatedly FFTing the same array is a diverging process (because FFTW computes the unnormalized transform).

Question 3.5. An FFTW routine is crashing when I call it.

Did the FFTW test programs pass (make check, or cd tests; make bigcheck if you want to be paranoid)? If so, you almost certainly have a bug in your own code. For example, you could be passing invalid arguments (such as wrongly-sized arrays) to FFTW, or you could simply have memory corruption elsewhere in your program that causes random crashes later on. Please don't complain to us unless you can come up with a minimal self-contained program (preferably under 30 lines) that illustrates the problem.

Question 3.6. My Fortran program crashes when calling FFTW.

As described in the manual, on 64-bit machines you must store the plans in variables large enough to hold a pointer, for example integer*8. We recommend using integer*8 on 32-bit machines as well, to simplify porting.

Question 3.7. FFTW gives results different from my old FFT.

People follow many different conventions for the DFT, and you should be sure to know the ones that we use (described in the FFTW manual). In particular, you should be aware that the FFTW_FORWARD/FFTW_BACKWARD directions correspond to signs of -1/+1 in the exponent of the DFT definition. (Numerical Recipes uses the opposite convention.)

You should also know that we compute an unnormalized transform. In contrast, Matlab is an example of program that computes a normalized transform. See Q3.10 `Why does your inverse transform return a scaled result?'.

Finally, note that floating-point arithmetic is not exact, so different FFT algorithms will give slightly different results (on the order of the numerical accuracy; typically a fractional difference of 1e-15 or so in double precision).

Question 3.8. FFTW gives different results between runs

If you use FFTW_MEASURE or FFTW_PATIENT mode, then the algorithm FFTW employs is not deterministic: it depends on runtime performance measurements. This will cause the results to vary slightly from run to run. However, the differences should be slight, on the order of the floating-point precision, and therefore should have no practical impact on most applications.

If you use saved plans (wisdom) or FFTW_ESTIMATE mode, however, then the algorithm is deterministic and the results should be identical between runs.

Question 3.9. Can I save FFTW's plans?

Yes. Starting with version 1.2, FFTW provides the wisdom mechanism for saving plans; see the FFTW manual.

Question 3.10. Why does your inverse transform return a scaled result?

Computing the forward transform followed by the backward transform (or vice versa) yields the original array scaled by the size of the array. (For multi-dimensional transforms, the size of the array is the product of the dimensions.) We could, instead, have chosen a normalization that would have returned the unscaled array. Or, to accomodate the many conventions in this matter, the transform routines could have accepted a "scale factor" parameter. We did not do this, however, for two reasons. First, we didn't want to sacrifice performance in the common case where the scale factor is 1. Second, in real applications the FFT is followed or preceded by some computation on the data, into which the scale factor can typically be absorbed at little or no cost.

Question 3.11. How can I make FFTW put the origin (zero frequency) at the center of its output?

For human viewing of a spectrum, it is often convenient to put the origin in frequency space at the center of the output array, rather than in the zero-th element (the default in FFTW). If all of the dimensions of your array are even, you can accomplish this by simply multiplying each element of the input array by (-1)^(i + j + ...), where i, j, etcetera are the indices of the element. (This trick is a general property of the DFT, and is not specific to FFTW.)

Question 3.12. How do I FFT an image/audio file in foobar format?

FFTW performs an FFT on an array of floating-point values. You can certainly use it to compute the transform of an image or audio stream, but you are responsible for figuring out your data format and converting it to the form FFTW requires.

Question 3.13. My program does not link (on Unix).

The libraries must be listed in the correct order (-lfftw3 -lm for FFTW 3.x) and after your program sources/objects. (The general rule is that if A uses B, then A must be listed before B in the link command.).

Question 3.14. I included your header, but linking still fails.

You're a C++ programmer, aren't you? You have to compile the FFTW library and link it into your program, not just #include <fftw3.h>. (Yes, this is really a FAQ.)

Question 3.15. My program crashes, complaining about stack space.

You cannot declare large arrays with automatic storage (e.g. via fftw_complex array[N]); you should use fftw_malloc (or equivalent) to allocate the arrays you want to transform if they are larger than a few hundred elements.

Question 3.16. FFTW seems to have a memory leak.

After you create a plan, FFTW caches the information required to quickly recreate the plan. (See Q3.9 `Can I save FFTW's plans?') It also maintains a small amount of other persistent memory. You can deallocate all of FFTW's internally allocated memory, if you wish, by calling fftw_cleanup(), as documented in the manual.

Question 3.17. The output of FFTW's transform is all zeros.

You should initialize your input array after creating the plan, unless you use FFTW_ESTIMATE: planning with FFTW_MEASURE or FFTW_PATIENT overwrites the input/output arrays, as described in the manual.

Question 3.18. How do I call FFTW from the Microsoft language du jour?

Please do not ask us Windows-specific questions. We do not use Windows. We know nothing about Visual Basic, Visual C++, or .NET. Please find the appropriate Usenet discussion group and ask your question there. See also Q2.2 `Does FFTW run on Windows?'.

Question 3.19. Can I compute only a subset of the DFT outputs?

In general, no, an FFT intrinsically computes all outputs from all inputs. In principle, there is something called a pruned FFT that can do what you want, but to compute K outputs out of N the complexity is in general O(N log K) instead of O(N log N), thus saving only a small additive factor in the log. (The same argument holds if you instead have only K nonzero inputs.)

There are some specific cases in which you can get the O(N log K) performance benefits easily, however, by combining a few ordinary FFTs. In particular, the case where you want the first K outputs, where K divides N, can be handled by performing N/K transforms of size K and then summing the outputs multiplied by appropriate phase factors. For more details, see pruned FFTs with FFTW.

There are also some algorithms that compute pruned transforms approximately, but they are beyond the scope of this FAQ.

Question 3.20. Can I use FFTW's routines for in-place and out-of-place matrix transposition?

You can use the FFTW guru interface to create a rank-0 transform of vector rank 2 where the vector strides are transposed. (A rank-0 transform is equivalent to a 1D transform of size 1, which. just copies the input into the output.) Specifying the same location for the input and output makes the transpose in-place.

For double-valued data stored in row-major format, plan creation looks like this: 

fftw_plan plan_transpose(int rows, int cols, double *in, double *out)
{
    const unsigned flags = FFTW_ESTIMATE; /* other flags are possible */
    fftw_iodim howmany_dims[2];

    howmany_dims[0].n  = rows;
    howmany_dims[0].is = cols;
    howmany_dims[0].os = 1;

    howmany_dims[1].n  = cols;
    howmany_dims[1].is = 1;
    howmany_dims[1].os = rows;

    return fftw_plan_guru_r2r(/*rank=*/ 0, /*dims=*/ NULL,
                              /*howmany_rank=*/ 2, howmany_dims,
                              in, out, /*kind=*/ NULL, flags);
}
(This entry was written by Rhys Ulerich.)
Next: Internals of FFTW.
Back: Installing FFTW.
Return to contents.

Matteo Frigo and Steven G. Johnson / fftw@fftw.org - 25 November 2012

Extracted from FFTW Frequently Asked Questions with Answers, Copyright © 2012 Matteo Frigo and Massachusetts Institute of Technology. fftw-3.3.3/doc/FAQ/fftw-faq.html/section4.html0000644000175000017500000000546712054411114015664 00000000000000 FFTW FAQ - Section 4

FFTW FAQ - Section 4
Internals of FFTW

Question 4.1. How does FFTW work?

The innovation (if it can be so called) in FFTW consists in having a variety of composable solvers, representing different FFT algorithms and implementation strategies, whose combination into a particular plan for a given size can be determined at runtime according to the characteristics of your machine/compiler. This peculiar software architecture allows FFTW to adapt itself to almost any machine.

For more details (albeit somewhat outdated), see the paper "FFTW: An Adaptive Software Architecture for the FFT", by M. Frigo and S. G. Johnson, Proc. ICASSP 3, 1381 (1998), also available at the FFTW web page.

Question 4.2. Why is FFTW so fast?

This is a complex question, and there is no simple answer. In fact, the authors do not fully know the answer, either. In addition to many small performance hacks throughout FFTW, there are three general reasons for FFTW's speed.
  • FFTW uses a variety of FFT algorithms and implementation styles that can be arbitrarily composed to adapt itself to a machine. See Q4.1 `How does FFTW work?'.
  • FFTW uses a code generator to produce highly-optimized routines for computing small transforms.
  • FFTW uses explicit divide-and-conquer to take advantage of the memory hierarchy.
For more details (albeit somewhat outdated), see the paper "FFTW: An Adaptive Software Architecture for the FFT", by M. Frigo and S. G. Johnson, Proc. ICASSP 3, 1381 (1998), available along with other references at the FFTW web page.
Next: Known bugs.
Back: Using FFTW.
Return to contents.

Matteo Frigo and Steven G. Johnson / fftw@fftw.org - 25 November 2012

Extracted from FFTW Frequently Asked Questions with Answers, Copyright © 2012 Matteo Frigo and Massachusetts Institute of Technology. fftw-3.3.3/doc/FAQ/fftw-faq.html/section5.html0000644000175000017500000001342012054411114015651 00000000000000 FFTW FAQ - Section 5

FFTW FAQ - Section 5
Known bugs

Question 5.1. FFTW 1.1 crashes in rfftwnd on Linux.

This bug was fixed in FFTW 1.2. There was a bug in rfftwnd causing an incorrect amount of memory to be allocated. The bug showed up in Linux with libc-5.3.12 (and nowhere else that we know of).

Question 5.2. The MPI transforms in FFTW 1.2 give incorrect results/leak memory.

These bugs were corrected in FFTW 1.2.1. The MPI transforms (really, just the transpose routines) in FFTW 1.2 had bugs that could cause errors in some situations.

Question 5.3. The test programs in FFTW 1.2.1 fail when I change FFTW to use single precision.

This bug was fixed in FFTW 1.3. (Older versions of FFTW did work in single precision, but the test programs didn't--the error tolerances in the tests were set for double precision.)

Question 5.4. The test program in FFTW 1.2.1 fails for n > 46340.

This bug was fixed in FFTW 1.3. FFTW 1.2.1 produced the right answer, but the test program was wrong. For large n, n*n in the naive transform that we used for comparison overflows 32 bit integer precision, breaking the test.

Question 5.5. The threaded code fails on Linux Redhat 5.0

We had problems with glibc-2.0.5. The code should work with glibc-2.0.7.

Question 5.6. FFTW 2.0's rfftwnd fails for rank > 1 transforms with a final dimension >= 65536.

This bug was fixed in FFTW 2.0.1. (There was a 32-bit integer overflow due to a poorly-parenthesized expression.)

Question 5.7. FFTW 2.0's complex transforms give the wrong results with prime factors 17 to 97.

There was a bug in the complex transforms that could cause incorrect results under (hopefully rare) circumstances for lengths with intermediate-size prime factors (17-97). This bug was fixed in FFTW 2.1.1.

Question 5.8. FFTW 2.1.1's MPI test programs crash with MPICH.

This bug was fixed in FFTW 2.1.2. The 2.1/2.1.1 MPI test programs crashed when using the MPICH implementation of MPI with the ch_p4 device (TCP/IP); the transforms themselves worked fine.

Question 5.9. FFTW 2.1.2's multi-threaded transforms don't work on AIX.

This bug was fixed in FFTW 2.1.3. The multi-threaded transforms in previous versions didn't work with AIX's pthreads implementation, which idiosyncratically creates threads in detached (non-joinable) mode by default.

Question 5.10. FFTW 2.1.2's complex transforms give incorrect results for large prime sizes.

This bug was fixed in FFTW 2.1.3. FFTW's complex-transform algorithm for prime sizes (in versions 2.0 to 2.1.2) had an integer overflow problem that caused incorrect results for many primes greater than 32768 (on 32-bit machines). (Sizes without large prime factors are not affected.)

Question 5.11. FFTW 2.1.3's multi-threaded transforms don't give any speedup on Solaris.

This bug was fixed in FFTW 2.1.4. (By default, Solaris creates threads that do not parallelize over multiple processors, so one has to request the proper behavior specifically.)

Question 5.12. FFTW 2.1.3 crashes on AIX.

The FFTW 2.1.3 configure script picked incorrect compiler flags for the xlc compiler on newer IBM processors. This is fixed in FFTW 2.1.4.
Back: Internals of FFTW.
Return to contents.

Matteo Frigo and Steven G. Johnson / fftw@fftw.org - 25 November 2012

Extracted from FFTW Frequently Asked Questions with Answers, Copyright © 2012 Matteo Frigo and Massachusetts Institute of Technology. fftw-3.3.3/doc/FAQ/bfnnconv.pl0000644000175000017500000002176012054410050012726 00000000000000#!/usr/bin/perl -- # Copyright (C) 1993-1995 Ian Jackson. # This file is free software; you can redistribute it and/or modify # it under the terms of the GNU General Public License as published by # the Free Software Foundation; either version 2, or (at your option) # any later version. # It is distributed in the hope that it will be useful, # but WITHOUT ANY WARRANTY; without even the implied warranty of # MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the # GNU General Public License for more details. # You should have received a copy of the GNU General Public License # along with GNU Emacs; see the file COPYING. If not, write to # the Free Software Foundation, Inc., 59 Temple Place - Suite 330, # Boston, MA 02111-1307, USA. # (Note: I do not consider works produced using these BFNN processing # tools to be derivative works of the tools, so they are NOT covered # by the GPL. However, I would appreciate it if you credited me if # appropriate in any documents you format using BFNN.) @outputs=('ascii','info','html'); while ($ARGV[0] =~ m/^\-/) { $_= shift(@ARGV); if (m/^-only/) { @outputs= (shift(@ARGV)); } else { warn "unknown option `$_' ignored"; } } $prefix= $ARGV[0]; $prefix= 'stdin' unless length($prefix); $prefix =~ s/\.bfnn$//; if (open(O,"$prefix.xrefdb")) { @xrefdb= ; close(O); } else { warn "no $prefix.xrefdb ($!)"; } $section= -1; for $thisxr (@xrefdb) { $_= $thisxr; chop; if (m/^Q (\w+) ((\d+)\.(\d+)) (.*)$/) { $qrefn{$1}= $2; $qreft{$1}= $5; $qn2ref{$3,$4}= $1; $maxsection= $3; $maxquestion[$3]= $4; } elsif (m/^S (\d+) /) { $maxsection= $1; $sn2title{$1}=$'; } } open(U,">$prefix.xrefdb-new"); for $x (@outputs) { require("m-$x.pl"); } &call('init'); while (<>) { chop; next if m/^\\comment\b/; if (!m/\S/) { &call('endpara'); next; } if (s/^\\section +//) { $line= $_; $section++; $question=0; print U "S $section $line\n"; $|=1; print "S$section",' 'x10,"\r"; $|=0; &call('endpara'); &call('startmajorheading',"$section", "Section $section", $section<$maxsection ? "Section ".($section+1) : '', $section>1 ? 'Section '.($section-1) : 'Top'); &text($line); &call('endmajorheading'); if ($section) { &call('endpara'); &call('startindex'); for $thisxr (@xrefdb) { $_= $thisxr; chop; if (m/^Q (\w+) (\d+)\.(\d+) (.*)$/) { $ref= $1; $num1= $2; $num2= $3; $text= $4; next unless $num1 == $section; &call('startindexitem',$ref,"Q$num1.$num2","Question $num1.$num2"); &text($text); &call('endindexitem'); } } &call('endindex'); } } elsif (s/^\\question \d{2}[a-z]{3}((:\w+)?) +//) { $line= $_; $question++; $qrefstring= $1; $qrefstring= "q_${section}_$question" unless $qrefstring =~ s/^://; print U "Q $qrefstring $section.$question $line\n"; $|=1; print "Q$section.$question",' 'x10,"\r"; $|=0; &call('endpara'); &call('startminorheading',$qrefstring, "Question $section.$question", $question < $maxquestion[$section] ? "Question $section.".($question+1) : $section < $maxsection ? "Question ".($section+1).".1" : '', $question > 1 ? "Question $section.".($question-1) : $section > 1 ? "Question ".($section-1).'.'.($maxquestion[$section-1]) : 'Top', "Section $section"); &text("Question $section.$question. $line"); &call('endminorheading'); } elsif (s/^\\only +//) { @saveoutputs= @outputs; @outputs=(); for $x (split(/\s+/,$_)) { push(@outputs,$x) if grep($x eq $_, @saveoutputs); } } elsif (s/^\\endonly$//) { @outputs= @saveoutputs; } elsif (s/^\\copyto +//) { $fh= $'; while(<>) { last if m/^\\endcopy$/; while (s/^([^\`]*)\`//) { print $fh $1; m/([^\\])\`/ || warn "`$_'"; $_= $'; $cmd= $`.$1; $it= `$cmd`; chop $it; print $fh $it; } print $fh $_; } } elsif (m/\\index$/) { &call('startindex'); for $thisxr (@xrefdb) { $_= $thisxr; chop; if (m/^Q (\w+) (\d+\.\d+) (.*)$/) { $ref= $1; $num= $2; $text= $3; &call('startindexitem',$ref,"Q$num","Question $num"); &text($text); &call('endindexitem'); } elsif (m/^S (\d+) (.*)$/) { $num= $1; $text= $2; next unless $num; &call('startindexmainitem',"s_$num", "Section $num.","Section $num"); &text($text); &call('endindexitem'); } else { warn $_; } } &call('endindex'); } elsif (m/^\\call-(\w+) +(\w+)\s*(.*)$/) { $fn= $1.'_'.$2; eval { &$fn($3); }; warn $@ if length($@); } elsif (m/^\\call +(\w+)\s*(.*)$/) { eval { &call($1,$2); }; warn $@ if length($@); } elsif (s/^\\set +(\w+)\s*//) { $svalue= $'; $svari= $1; eval("\$user_$svari=\$svalue"); $@ && warn "setting $svalue failed: $@\n"; } elsif (m/^\\verbatim$/) { &call('startverbatim'); while (<>) { chop; last if m/^\\endverbatim$/; &call('verbatim',$_); } &call('endverbatim'); } else { s/\.$/\. /; &text($_." "); } } print ' 'x25,"\r"; &call('finish'); rename("$prefix.xrefdb-new","$prefix.xrefdb") || warn "rename xrefdb: $!"; exit 0; sub text { local($in,$rhs,$word,$refn,$reft,$fn,$style); $in= "$holdover$_[0]"; $holdover= ''; while ($in =~ m/\\/) { #print STDERR ">$`##$'\n"; $rhs=$'; &call('text',$`); $_= $rhs; if (m/^\w+ $/) { $holdover= "\\$&"; $in= ''; } elsif (s/^fn\s+([^\s\\]*\w)//) { $in= $_; $word= $1; &call('courier'); &call('text',$word); &call('endcourier'); } elsif (s/^tab\s+(\d+)\s+//) { $in= $_; &call('tab',$1); } elsif (s/^nl\s+//) { $in= $_; &call('newline'); } elsif (s/^qref\s+(\w+)//) { $refn= $qrefn{$1}; $reft= $qreft{$1}; if (!length($refn)) { warn "unknown question `$1'"; } $in= "$`\\pageref:$1:$refn:$reft\\endpageref.$_"; } elsif (s/^pageref:(\w+):([^:\n]+)://) { $in= $_; &call('pageref',$1,$2); } elsif (s/^endpageref\.//) { $in= $_; &call('endpageref'); } elsif (s/^(\w+)\{//) { $in= $_; $fn= $1; eval { &call("$fn"); }; if (length($@)) { warn $@; $fn= 'x'; } push(@styles,$fn); } elsif (s/^\}//) { $in= $_; $fn= pop(@styles); if ($fn ne 'x') { &call("end$fn"); } } elsif (s/^\\//) { $in= $_; &call('text',"\\"); } elsif (s,^(\w+)\s+([-A-Za-z0-9.\@:/]*\w),,) { #print STDERR "**$&**$_\n"; $in= $_; $style=$1; $word= $2; &call($style); &call('text',$word); &call("end$style"); } else { warn "unknown control `\\$_'"; $in= $_; } } &call('text',$in); } sub call { local ($fnbase, @callargs) = @_; local ($coutput); for $coutput (@outputs) { if ($fnbase eq 'text' && eval("\@${coutput}_cmds")) { #print STDERR "special handling text (@callargs) for $coutput\n"; $evstrg= "\$${coutput}_args[\$#${coutput}_args].=\"\@callargs\""; eval($evstrg); length($@) && warn "call adding for $coutput (($evstrg)): $@"; } else { $fntc= $coutput.'_'.$fnbase; &$fntc(@callargs); } } } sub recurse { local (@outputs) = $coutput; local ($holdover); &text($_[0]); } sub arg { #print STDERR "arg($_[0]) from $coutput\n"; $cmd= $_[0]; eval("push(\@${coutput}_cmds,\$cmd); push(\@${coutput}_args,'')"); length($@) && warn "arg setting up for $coutput: $@"; } sub endarg { #print STDERR "endarg($_[0]) from $coutput\n"; $evstrg= "\$${coutput}_cmd= \$cmd= pop(\@${coutput}_cmds); ". "\$${coutput}_arg= \$arg= pop(\@${coutput}_args); "; eval($evstrg); length($@) && warn "endarg extracting for $coutput (($evstrg)): $@"; #print STDERR ">call $coutput $cmd $arg< (($evstrg))\n"; $evstrg= "&${coutput}_do_${cmd}(\$arg)"; eval($evstrg); length($@) && warn "endarg running ${coutput}_do_${cmd} (($evstrg)): $@"; } fftw-3.3.3/doc/FAQ/m-ascii.pl0000644000175000017500000001115612054410050012435 00000000000000## ASCII output # Copyright (C) 1993-1995 Ian Jackson. # This file is free software; you can redistribute it and/or modify # it under the terms of the GNU General Public License as published by # the Free Software Foundation; either version 2, or (at your option) # any later version. # It is distributed in the hope that it will be useful, # but WITHOUT ANY WARRANTY; without even the implied warranty of # MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the # GNU General Public License for more details. # You should have received a copy of the GNU General Public License # along with GNU Emacs; see the file COPYING. If not, write to # the Free Software Foundation, Inc., 59 Temple Place - Suite 330, # Boston, MA 02111-1307, USA. # (Note: I do not consider works produced using these BFNN processing # tools to be derivative works of the tools, so they are NOT covered # by the GPL. However, I would appreciate it if you credited me if # appropriate in any documents you format using BFNN.) sub ascii_init { open(ASCII,">$prefix.ascii"); } sub ascii_startmajorheading { print ASCII '='x79,"\n\n"; $ascii_status= 'h'; &ascii_text($_[0] ? "Section $_[0]. " : ''); } sub ascii_startminorheading { print ASCII '-'x79,"\n\n"; $ascii_status= 'h'; } sub ascii_italic { &ascii_text('*'); } sub ascii_enditalic { $ascii_para .= '*'; } sub ascii_email { &ascii_text('<'); } sub ascii_endemail { &ascii_text('>'); } sub ascii_ftpon { } sub ascii_endftpon { } sub ascii_ftpin { } sub ascii_endftpin { } sub ascii_docref { } sub ascii_enddocref { } sub ascii_courier { } sub ascii_endcourier { } sub ascii_newsgroup { } sub ascii_endnewsgroup { } sub ascii_ftpsilent { $ascii_ignore++; } sub ascii_endftpsilent { $ascii_ignore--; } sub ascii_text { return if $ascii_ignore; if ($ascii_status eq '') { $ascii_status= 'p'; } $ascii_para .= $_[0]; } sub ascii_tab { local ($n) = $_[0]-length($ascii_para); $ascii_para .= ' 'x$n if $n>0; } sub ascii_newline { return unless $ascii_status eq 'p'; &ascii_writepara; } sub ascii_writepara { local ($thisline, $thisword, $rest); for (;;) { last unless $ascii_para =~ m/\S/; $thisline= $ascii_indentstring; for (;;) { last unless $ascii_para =~ m/^(\s*\S+)/; unless (length($1) + length($thisline) < 75 || length($thisline) == length($ascii_indentstring)) { last; } $thisline .= $1; $ascii_para= $'; } $ascii_para =~ s/^\s*//; print ASCII $thisline,"\n"; $ascii_indentstring= $ascii_nextindent; last unless length($ascii_para); } $ascii_status= ''; $ascii_para= ''; } sub ascii_endpara { return unless $ascii_status eq 'p'; &ascii_writepara; print ASCII "\n"; } sub ascii_endheading { $ascii_para =~ s/\s*$//; print ASCII "$ascii_para\n\n"; $ascii_status= ''; $ascii_para= ''; } sub ascii_endmajorheading { &ascii_endheading(@_); } sub ascii_endminorheading { &ascii_endheading(@_); } sub ascii_startverbatim { $ascii_vstatus= $ascii_status; &ascii_writepara; } sub ascii_verbatim { print ASCII $_[0],"\n"; } sub ascii_endverbatim { $ascii_status= $ascii_vstatus; } sub ascii_finish { close(ASCII); } sub ascii_startindex { $ascii_status= ''; } sub ascii_endindex { $ascii_status= 'p'; } sub ascii_endindexitem { printf ASCII " %-11s %-.66s\n",$ascii_left,$ascii_para; $ascii_status= 'p'; $ascii_para= ''; } sub ascii_startindexitem { $ascii_left= $_[1]; } sub ascii_startindexmainitem { $ascii_left= $_[1]; print ASCII "\n" if $ascii_status eq 'p'; } sub ascii_startindent { $ascii_istatus= $ascii_status; &ascii_writepara; $ascii_indentstring= " $ascii_indentstring"; $ascii_nextindent= " $ascii_nextindent"; } sub ascii_endindent { $ascii_indentstring =~ s/^ //; $ascii_nextindent =~ s/^ //; $ascii_status= $ascii_istatus; } sub ascii_startpackedlist { $ascii_plc=0; } sub ascii_endpackedlist { &ascii_newline if !$ascii_plc; } sub ascii_packeditem { &ascii_newline if !$ascii_plc; &ascii_tab($ascii_plc*40+5); $ascii_plc= !$ascii_plc; } sub ascii_startlist { &ascii_endpara; $ascii_indentstring= " $ascii_indentstring"; $ascii_nextindent= " $ascii_nextindent"; } sub ascii_endlist { &ascii_endpara; $ascii_indentstring =~ s/^ //; $ascii_nextindent =~ s/^ //; } sub ascii_item { &ascii_newline; $ascii_indentstring =~ s/ $/* /; } sub ascii_pageref { &ascii_text("Q$_[1] \`"); } sub ascii_endpageref { &ascii_text("'"); } 1; fftw-3.3.3/doc/FAQ/m-html.pl0000644000175000017500000002242112054410050012306 00000000000000## HTML output # Copyright (C) 1993-1995 Ian Jackson. # This file is free software; you can redistribute it and/or modify # it under the terms of the GNU General Public License as published by # the Free Software Foundation; either version 2, or (at your option) # any later version. # It is distributed in the hope that it will be useful, # but WITHOUT ANY WARRANTY; without even the implied warranty of # MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the # GNU General Public License for more details. # You should have received a copy of the GNU General Public License # along with GNU Emacs; see the file COPYING. If not, write to # the Free Software Foundation, Inc., 59 Temple Place - Suite 330, # Boston, MA 02111-1307, USA. # (Note: I do not consider works produced using these BFNN processing # tools to be derivative works of the tools, so they are NOT covered # by the GPL. However, I would appreciate it if you credited me if # appropriate in any documents you format using BFNN.) %saniarray= ('<','lt', '>','gt', '&','amp', '"','quot'); sub html_init { $html_prefix = './'.$prefix; $html_prefix =~ s:^\.//:/:; system('rm','-r',"$html_prefix.html"); system('mkdir',"$html_prefix.html"); open(HTML,">$html_prefix.html/index.html"); print HTML "\n"; print HTML "\n"; $html_needpara= -1; $html_end=''; chop($html_date=`date '+%d %B %Y'`); chop($html_year=`date '+%Y'`); } sub html_startup { print HTML < $user_title

$user_title

END &html_readrefs($_[0]); if (length($user_copyrightref)) { local ($refn) = $qrefn{$user_copyrightref}; if (!length($refn)) { warn "unknown question (copyright) `$user_copyrightref'"; } $refn =~ m/(\d+)\.(\d+)/; local ($s,$n) = ($1,$2); $html_copyrighthref= ($s == $html_sectionn)?'':"section$s.html"; $html_copyrighthref.= "#$qn2ref{$s,$n}"; } } sub html_close { print HTML $html_end,"
\n$user_author\n"; print HTML "- $html_date\n

\n"; print HTML "Extracted from $user_title,\n"; print HTML "" if length($html_copyrighthref); print HTML "Copyright © $html_year $user_copyholder."; print HTML "" if length($html_copyrighthref); print HTML "\n\n"; close(HTML); } sub html_startmajorheading { local ($ref, $this,$next,$back) = @_; local ($nextt,$backt); $this =~ s/^Section /section/; $html_sectionn= $ref; $next =~ s/^Section /section/ && ($nextt= $sn2title{$'}); $back =~ s/^Section /section/ ? ($backt= $sn2title{$'}) : ($back=''); if ($html_sectionn) { &html_close; open(HTML,">$html_prefix.html/$this.html"); print HTML "\n"; print HTML "\n"; $html_end= "
\n"; $html_end.= "Next: $nextt.
\n" if $next; $html_end.= "Back: $backt.
\n" if $back; $html_end.= ""; $html_end.= "Return to contents.

\n"; print HTML < $user_brieftitle - Section $html_sectionn END print HTML "" if $next; print HTML "" if $back; print HTML <

$user_brieftitle - Section $html_sectionn
END $html_needpara= -1; } else { print HTML "\n

\n"; $html_needpara=-1; } } sub html_endmajorheading { print HTML "\n

\n\n"; $html_needpara=-1; } sub html_startminorheading { local ($ref, $this) = @_; $html_needpara=0; $this =~ m/^Question (\d+)\.(\d+)/; local ($s,$n) = ($1,$2); print HTML "\n

\n"; } sub html_endminorheading { print HTML "\n

\n\n"; $html_needpara=-1; } sub html_newsgroup { &arg('newsgroup'); } sub html_endnewsgroup { &endarg('newsgroup'); } sub html_do_newsgroup { print HTML "$_[0]"; } sub html_email { &arg('email'); } sub html_endemail { &endarg('email'); } sub html_do_email { print HTML "$_[0]"; } sub html_courier { print HTML "" ; } sub html_endcourier { print HTML ""; } sub html_italic { print HTML "" ; } sub html_enditalic { print HTML "" ; } sub html_docref { &arg('docref'); } sub html_enddocref { &endarg('docref'); } sub html_do_docref { if (!defined($html_refval{$_[0]})) { warn "undefined HTML reference $_[0]"; $html_refval{$n}='UNDEFINED'; } print HTML ""; &recurse($_[0]); print HTML ""; } sub html_readrefs { local ($p); open(HTMLREFS,"<$_[0]") || (warn("failed to open HTML refs $_[0]: $!"),return); while() { next if m/^\\\s/; s/\s*\n$//; if (s/^\\prefix\s*//) { $p= $'; next; } elsif (s/^\s*(\S.*\S)\s*\\\s*//) { $_=$1; $v=$'; s/\\\\/\\/g; $html_refval{$_}= $p.$v; } else { warn("ununderstood line in HTML refs >$_<"); } } close(HTMLREFS); } sub html_ftpsilent { &arg('ftpsilent'); } sub html_endftpsilent { &endarg('ftpsilent'); } sub html_do_ftpsilent { if ($_[0] =~ m/:/) { $html_ftpsite= $`; $html_ftpdir= $'.'/'; } else { $html_ftpsite= $_[0]; $html_ftpdir= ''; } } sub html_ftpon { &arg('ftpon'); } sub html_endftpon { &endarg('ftpon'); } sub html_do_ftpon { #print STDERR "ftpon($_[0])\n"; $html_ftpsite= $_[0]; $html_ftpdir= ''; print HTML ""; &recurse($_[0]); print HTML ""; } sub html_ftpin { &arg('ftpin'); } sub html_endftpin { &endarg('ftpin'); } sub html_do_ftpin { #print STDERR "ftpin($_[0])\n"; print HTML ""; &recurse($_[0]); print HTML ""; } sub html_text { print HTML "\n

\n" if $html_needpara > 0; $html_needpara=0; $html_stuff= &html_sanitise($_[0]); while ($html_stuff =~ s/^(.{40,70}) //) { print HTML "$1\n"; } print HTML $html_stuff; } sub html_tab { $htmltabignore++ || warn "html tab ignored"; } sub html_newline { print HTML "
\n" ; } sub html_startverbatim { print HTML "

\n"   ;                       }
sub html_verbatim      { print HTML &html_sanitise($_[0]),"\n";         }
sub html_endverbatim   { print HTML "
\n" ; $html_needpara= -1; } sub html_endpara { $html_needpara || $html_needpara++; } sub html_finish { &html_close; } sub html_startindex { print HTML "
    \n"; } sub html_endindex { print HTML "
\n"; } sub html_startindexitem { local ($ref,$qval) = @_; $qval =~ m/Q(\d+)\.(\d+)/; local ($s,$n) = ($1,$2); print HTML "
  • Q$s.$n. "; $html_indexunhead=''; } sub html_startindexmainitem { local ($ref,$s) = @_; $s =~ m/\d+/; $s= $&; print HTML "

    " if ($s > 1); print HTML "
  • Section $s. "; $html_indexunhead=''; } sub html_endindexitem { print HTML "$html_indexunhead\n"; } sub html_startlist { print HTML "\n"; $html_itemend="
      "; } sub html_endlist { print HTML "$html_itemend\n
    \n"; $html_needpara=-1 } sub html_item { print HTML "$html_itemend\n
  • "; $html_itemend=""; $html_needpara=-1; } sub html_startpackedlist { print HTML "\n"; $html_itemend=""; } sub html_endpackedlist { print HTML "$html_itemend\n\n"; $html_needpara=-1; } sub html_packeditem { print HTML "$html_itemend\n
  • "; $html_itemend=""; $html_needpara=-1; } sub html_startindent { print HTML "
    \n"; } sub html_endindent { print HTML "
    \n"; } sub html_pageref { local ($ref,$sq) = @_; $sq =~ m/(\d+)\.(\d+)/; local ($s,$n) = ($1,$2); print HTML "Q$sq \`"; } sub html_endpageref { print HTML "'"; } sub html_sanitise { local ($in) = @_; local ($out); while ($in =~ m/[<>&"]/) { $out.= $`. '&'. $saniarray{$&}. ';'; $in=$'; } $out.= $in; $out; } 1; fftw-3.3.3/doc/FAQ/m-info.pl0000644000175000017500000001277312054410050012306 00000000000000## Info output # Copyright (C) 1993-1995 Ian Jackson. # This file is free software; you can redistribute it and/or modify # it under the terms of the GNU General Public License as published by # the Free Software Foundation; either version 2, or (at your option) # any later version. # It is distributed in the hope that it will be useful, # but WITHOUT ANY WARRANTY; without even the implied warranty of # MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the # GNU General Public License for more details. # You should have received a copy of the GNU General Public License # along with GNU Emacs; see the file COPYING. If not, write to # the Free Software Foundation, Inc., 59 Temple Place - Suite 330, # Boston, MA 02111-1307, USA. # (Note: I do not consider works produced using these BFNN processing # tools to be derivative works of the tools, so they are NOT covered # by the GPL. However, I would appreciate it if you credited me if # appropriate in any documents you format using BFNN.) sub info_init { open(INFO,">$prefix.info"); print INFO <'); } sub info_ftpon { } sub info_endftpon { } sub info_ftpin { } sub info_endftpin { } sub info_docref { } sub info_enddocref { } sub info_courier { } sub info_endcourier { } sub info_newsgroup { } sub info_endnewsgroup { } sub info_ftpsilent { $info_ignore++; } sub info_endftpsilent { $info_ignore--; } sub info_text { return if $info_ignore; if ($info_status eq '') { $info_status= 'p'; } $info_para .= $_[0]; } sub info_tab { local ($n) = $_[0]-length($info_para); $info_para .= ' 'x$n if $n>0; } sub info_newline { return unless $info_status eq 'p'; print INFO &info_writepara; } sub info_writepara { local ($thisline, $thisword, $rest, $output); for (;;) { last unless $info_para =~ m/\S/; $thisline= $info_indentstring; for (;;) { last unless $info_para =~ m/^(\s*\S+)/; unless (length($1) + length($thisline) < 75 || length($thisline) == length($info_indentstring)) { last; } $thisline .= $1; $info_para= $'; } $info_para =~ s/^\s*//; $output.= $thisline."\n"; $info_indentstring= $info_nextindent; last unless length($info_para); } $info_status= ''; $info_para= ''; return $output; } sub info_endpara { return unless $info_status eq 'p'; print INFO &info_writepara; print INFO "\n"; } sub info_endheading { $info_para =~ s/\s*$//; print INFO "$info_para\n\n"; $info_status= ''; $info_para= ''; } sub info_endmajorheading { &info_endheading(@_); } sub info_endminorheading { &info_endheading(@_); } sub info_startverbatim { print INFO &info_writepara; } sub info_verbatim { print INFO $_[0],"\n"; } sub info_endverbatim { $info_status= $info_vstatus; } sub info_finish { close(INFO); } sub info_startindex { &info_endpara; $info_moredetail= ''; $info_status= ''; } sub info_endindex { print INFO "$info_moredetail\n" if length($info_moredetail); } sub info_endindexitem { $info_indentstring= sprintf("* %-17s ",$info_label.'::'); $info_nextindent= ' 'x20; local ($txt); $txt= &info_writepara; if ($info_main) { print INFO $label.$txt; $txt =~ s/^.{20}//; $info_moredetail.= $txt; } else { $info_moredetail.= $label.$txt; } $info_indentstring= $info_nextindent= ''; $info_status='p'; } sub info_startindexitem { print INFO "* Menu:\n" if $info_status eq ''; $info_status= ''; $info_label= $_[2]; $info_main= 0; } sub info_startindexmainitem { print INFO "* Menu:\n" if $info_status eq ''; $info_label= $_[2]; $info_main= 1; $info_moredetail .= "\n$_[2], "; $info_status= ''; } sub info_startindent { $info_istatus= $info_status; print INFO &info_writepara; $info_indentstring= " $info_indentstring"; $info_nextindent= " $info_nextindent"; } sub info_endindent { $info_indentstring =~ s/^ //; $info_nextindent =~ s/^ //; $info_status= $info_istatus; } sub info_startpackedlist { $info_plc=0; } sub info_endpackedlist { &info_newline if !$info_plc; } sub info_packeditem { &info_newline if !$info_plc; &info_tab($info_plc*40+5); $info_plc= !$info_plc; } sub info_startlist { $info_istatus= $info_status; print INFO &info_writepara; $info_indentstring= " $info_indentstring"; $info_nextindent= " $info_nextindent"; } sub info_endlist { $info_indentstring =~ s/^ //; $info_nextindent =~ s/^ //; $info_status= $info_lstatus; } sub info_item { &info_newline; $info_indentstring =~ s/ $/* /; } sub info_pageref { &info_text("*Note Question $_[1]:: \`"); } sub info_endpageref { &info_text("'"); } 1; fftw-3.3.3/doc/FAQ/m-lout.pl0000644000175000017500000001373012054410050012330 00000000000000## Lout output # Copyright (C) 1993-1995 Ian Jackson. # This file is free software; you can redistribute it and/or modify # it under the terms of the GNU General Public License as published by # the Free Software Foundation; either version 2, or (at your option) # any later version. # It is distributed in the hope that it will be useful, # but WITHOUT ANY WARRANTY; without even the implied warranty of # MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the # GNU General Public License for more details. # You should have received a copy of the GNU General Public License # along with GNU Emacs; see the file COPYING. If not, write to # the Free Software Foundation, Inc., 59 Temple Place - Suite 330, # Boston, MA 02111-1307, USA. # (Note: I do not consider works produced using these BFNN processing # tools to be derivative works of the tools, so they are NOT covered # by the GPL. However, I would appreciate it if you credited me if # appropriate in any documents you format using BFNN.) sub lout_init { open(LOUT,">$prefix.lout"); chop($dprint= `date '+%d %B %Y'`); $dprint =~ s/^0//; } sub lout_startup { local ($lbs) = &lout_sanitise($user_brieftitle); print LOUT <0)*40+5); $lout_plc= !$lout_plc; } sub lout_startlist { &lout_endpara; print LOUT "\@RawIndentedList style {\@Bullet} indent {0.5i} gap {1.1vx}\n"; $lout_styles .= 'l'; $lout_status= ''; } sub lout_endlist { &lout_endpara; print LOUT "\@EndList\n\n"; $lout_styles =~ s/.$//; } sub lout_item { &lout_endpara; print LOUT "\@ListItem{"; $lout_styles.= 'I'; } sub lout_startindex { print LOUT "//0.0fe\n"; } sub lout_endindex { $lout_status='p'; } sub lout_startindexmainitem { $lout_marker= $_[0]; $lout_status= ''; print LOUT "//0.3vx Bold \@Font \@HAdjust { \@HContract { { $_[1] } |3cx {"; $lout_iiendheight= '1.00'; $lout_styles .= 'X'; } sub lout_startindexitem { $lout_marker= $_[0]; print LOUT "\@HAdjust { \@HContract { { $_[1] } |3cx {"; $lout_iiendheight= '0.95'; $lout_styles .= 'X'; } sub lout_endindexitem { print LOUT "} } |0c \@PageOf { $lout_marker } } //${lout_iiendheight}vx\n"; $lout_styles =~ s/.$//; } sub lout_email { &lout_courier; &lout_text('<'); } sub lout_endemail { &lout_text('>'); &lout_endcourier; } sub lout_ftpon { &lout_courier; } sub lout_endftpon { &lout_endcourier; } sub lout_ftpin { &lout_courier; } sub lout_endftpin { &lout_endcourier; } sub lout_docref { } sub lout_enddocref { } sub lout_ftpsilent { $lout_ignore++; } sub lout_endftpsilent { $lout_ignore--; } sub lout_newsgroup { &lout_courier; } sub lout_endnewsgroup { &lout_endcourier; } sub lout_text { return if $lout_ignore; $lout_status= 'p'; $_= &lout_sanitise($_[0]); s/ $/\n/ unless $lout_styles =~ m/[fhX]/; print LOUT $_; } sub lout_tab { local ($size) = $_[0]*0.5; print LOUT " |${size}ft "; } sub lout_newline { print LOUT " //1.0vx\n"; } sub lout_sanitise { local ($in) = @_; local ($out); $in= ' '.$in.' '; $out=''; while ($in =~ m/(\s)(\S*[\@\/|\\\"\^\&\{\}\#]\S*)(\s)/) { $out .= $`.$1; $in = $3.$'; $_= $2; s/[\\\"]/\\$&/g; $out .= '"'.$_.'"'; } $out .= $in; $out =~ s/^ //; $out =~ s/ $//; $out; } sub lout_endpara { return if $lout_status eq ''; if ($lout_styles eq '') { print LOUT "\@LP\n\n"; } elsif ($lout_styles =~ s/I$//) { print LOUT "}\n"; } $lout_status= ''; } sub lout_startverbatim { print LOUT "//0.4f\n\@RawIndentedDisplay lines \@Break". " { {0.7 1.0} \@Scale {Courier Bold} \@Font {\n"; } sub lout_verbatim { $_= $_[0]; s/^\s*//; print LOUT &lout_sanitise($_),"\n"; } sub lout_endverbatim { print LOUT "}\n}\n//0.4f\n"; } 1; fftw-3.3.3/doc/FAQ/m-post.pl0000644000175000017500000001074212054410050012332 00000000000000## POST output # Copyright (C) 1993-1995 Ian Jackson. # This file is free software; you can redistribute it and/or modify # it under the terms of the GNU General Public License as published by # the Free Software Foundation; either version 2, or (at your option) # any later version. # It is distributed in the hope that it will be useful, # but WITHOUT ANY WARRANTY; without even the implied warranty of # MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the # GNU General Public License for more details. # You should have received a copy of the GNU General Public License # along with GNU Emacs; see the file COPYING. If not, write to # the Free Software Foundation, Inc., 59 Temple Place - Suite 330, # Boston, MA 02111-1307, USA. # (Note: I do not consider works produced using these BFNN processing # tools to be derivative works of the tools, so they are NOT covered # by the GPL. However, I would appreciate it if you credited me if # appropriate in any documents you format using BFNN.) sub post_init { open(POST,">$prefix.post"); } sub post_startmajorheading { print POST '='x79,"\n\n"; $post_status= 'h'; &post_text($_[0] ? "Section $_[0]. " : ''); } sub post_startminorheading { print POST '-'x77,"\n\n"; $post_status= 'h'; } sub post_italic { &post_text('*'); } sub post_enditalic { $post_para .= '*'; } sub post_email { &post_text('<'); } sub post_endemail { &post_text('>'); } sub post_ftpon { } sub post_endftpon { } sub post_ftpin { } sub post_endftpin { } sub post_docref { } sub post_enddocref { } sub post_courier { } sub post_endcourier { } sub post_newsgroup { } sub post_endnewsgroup { } sub post_ftpsilent { $post_ignore++; } sub post_endftpsilent { $post_ignore--; } sub post_text { return if $post_ignore; if ($post_status eq '') { $post_status= 'p'; } $post_para .= $_[0]; } sub post_tab { local ($n) = $_[0]-length($post_para); $post_para .= ' 'x$n if $n>0; } sub post_newline { return unless $post_status eq 'p'; &post_writepara; } sub post_writepara { local ($thisline, $thisword, $rest); for (;;) { last unless $post_para =~ m/\S/; $thisline= $post_indentstring; for (;;) { last unless $post_para =~ m/^(\s*\S+)/; unless (length($1) + length($thisline) < 75 || length($thisline) == length($post_indentstring)) { last; } $thisline .= $1; $post_para= $'; } $post_para =~ s/^\s*//; print POST $thisline,"\n"; $post_indentstring= $post_nextindent; last unless length($post_para); } $post_status= ''; $post_para= ''; } sub post_endpara { return unless $post_status eq 'p'; &post_writepara; print POST "\n"; } sub post_endheading { $post_para =~ s/\s*$//; print POST "$post_para\n\n"; $post_status= ''; $post_para= ''; } sub post_endmajorheading { &post_endheading(@_); } sub post_endminorheading { &post_endheading(@_); } sub post_startverbatim { $post_vstatus= $post_status; &post_writepara; } sub post_verbatim { print POST $_[0],"\n"; } sub post_endverbatim { $post_status= $post_vstatus; } sub post_finish { close(POST); } sub post_startindex { $post_status= ''; } sub post_endindex { $post_status= 'p'; } sub post_endindexitem { printf POST " %-11s %-.66s\n",$post_left,$post_para; $post_status= 'p'; $post_para= ''; } sub post_startindexitem { $post_left= $_[1]; } sub post_startindexmainitem { $post_left= $_[1]; print POST "\n" if $post_status eq 'p'; } sub post_startindent { $post_istatus= $post_status; &post_writepara; $post_indentstring= " $post_indentstring"; $post_nextindent= " $post_nextindent"; } sub post_endindent { $post_indentstring =~ s/^ //; $post_nextindent =~ s/^ //; $post_status= $post_istatus; } sub post_startpackedlist { $post_plc=0; } sub post_endpackedlist { &post_newline if !$post_plc; } sub post_packeditem { &post_newline if !$post_plc; &post_tab($post_plc*40+5); $post_plc= !$post_plc; } sub post_startlist { &post_endpara; $post_indentstring= " $post_indentstring"; $post_nextindent= " $post_nextindent"; } sub post_endlist { &post_endpara; $post_indentstring =~ s/^ //; $post_nextindent =~ s/^ //; } sub post_item { &post_newline; $post_indentstring =~ s/ $/* /; } sub post_pageref { &post_text("Q$_[1] \`"); } sub post_endpageref { &post_text("'"); } 1; fftw-3.3.3/doc/FAQ/html.refs0000644000175000017500000000053112054410050012376 00000000000000\ References for the FFTW FAQ \ the FFTW web page \ http://www.fftw.org FFTW Windows installation notes \ http://www.fftw.org/install/windows.html Categories of Free and Non-Free Software \ http://www.gnu.org/philosophy/categories.html the Caml web page \ http://caml.inria.fr pruned FFTs with FFTW \ http://www.fftw.org/pruned.html fftw-3.3.3/doc/fftw3.info0000644000175000017500000001405112054411105012054 00000000000000This is fftw3.info, produced by makeinfo version 4.13 from fftw3.texi. This manual is for FFTW (version 3.3.3, 25 November 2012). Copyright (C) 2003 Matteo Frigo. Copyright (C) 2003 Massachusetts Institute of Technology. Permission is granted to make and distribute verbatim copies of this manual provided the copyright notice and this permission notice are preserved on all copies. Permission is granted to copy and distribute modified versions of this manual under the conditions for verbatim copying, provided that the entire resulting derived work is distributed under the terms of a permission notice identical to this one. Permission is granted to copy and distribute translations of this manual into another language, under the above conditions for modified versions, except that this permission notice may be stated in a translation approved by the Free Software Foundation. INFO-DIR-SECTION Texinfo documentation system START-INFO-DIR-ENTRY * fftw3: (fftw3). FFTW User's Manual. END-INFO-DIR-ENTRY  Indirect: fftw3.info-1: 1076 fftw3.info-2: 297310  Tag Table: (Indirect) Node: Top1076 Node: Introduction1749 Node: Tutorial8085 Ref: Tutorial-Footnote-19329 Node: Complex One-Dimensional DFTs9423 Node: Complex Multi-Dimensional DFTs15179 Ref: Complex Multi-Dimensional DFTs-Footnote-118611 Node: One-Dimensional DFTs of Real Data18746 Node: Multi-Dimensional DFTs of Real Data23191 Node: More DFTs of Real Data27121 Node: The Halfcomplex-format DFT30623 Node: Real even/odd DFTs (cosine/sine transforms)33232 Ref: Real even/odd DFTs (cosine/sine transforms)-Footnote-138842 Ref: Real even/odd DFTs (cosine/sine transforms)-Footnote-239031 Node: The Discrete Hartley Transform39964 Ref: The Discrete Hartley Transform-Footnote-142149 Node: Other Important Topics42398 Node: SIMD alignment and fftw_malloc42691 Node: Multi-dimensional Array Format44951 Node: Row-major Format45572 Node: Column-major Format47265 Node: Fixed-size Arrays in C48349 Node: Dynamic Arrays in C49785 Node: Dynamic Arrays in C-The Wrong Way51423 Node: Words of Wisdom-Saving Plans53171 Node: Caveats in Using Wisdom55846 Node: FFTW Reference57934 Node: Data Types and Files58422 Node: Complex numbers58854 Node: Precision60595 Node: Memory Allocation62157 Node: Using Plans63728 Node: Basic Interface67608 Ref: Basic Interface-Footnote-168352 Node: Complex DFTs68416 Node: Planner Flags72383 Node: Real-data DFTs77740 Node: Real-data DFT Array Format82736 Node: Real-to-Real Transforms84991 Node: Real-to-Real Transform Kinds88961 Node: Advanced Interface91429 Node: Advanced Complex DFTs92169 Node: Advanced Real-data DFTs96428 Node: Advanced Real-to-real Transforms98755 Node: Guru Interface99861 Node: Interleaved and split arrays100784 Node: Guru vector and transform sizes101827 Node: Guru Complex DFTs104392 Node: Guru Real-data DFTs107228 Node: Guru Real-to-real Transforms110151 Node: 64-bit Guru Interface111470 Node: New-array Execute Functions113793 Node: Wisdom117791 Node: Wisdom Export118150 Node: Wisdom Import120124 Node: Forgetting Wisdom122146 Node: Wisdom Utilities122518 Node: What FFTW Really Computes123885 Node: The 1d Discrete Fourier Transform (DFT)124710 Node: The 1d Real-data DFT126069 Node: 1d Real-even DFTs (DCTs)127723 Node: 1d Real-odd DFTs (DSTs)130932 Node: 1d Discrete Hartley Transforms (DHTs)133874 Node: Multi-dimensional Transforms134550 Node: Multi-threaded FFTW137153 Node: Installation and Supported Hardware/Software138622 Node: Usage of Multi-threaded FFTW140447 Node: How Many Threads to Use?143755 Node: Thread safety144779 Node: Distributed-memory FFTW with MPI146947 Node: FFTW MPI Installation149526 Node: Linking and Initializing MPI FFTW151318 Node: 2d MPI example152548 Node: MPI Data Distribution156784 Node: Basic and advanced distribution interfaces159662 Node: Load balancing164097 Node: Transposed distributions165783 Node: One-dimensional distributions169555 Node: Multi-dimensional MPI DFTs of Real Data172124 Node: Other Multi-dimensional Real-data MPI Transforms176772 Node: FFTW MPI Transposes178945 Node: Basic distributed-transpose interface179785 Node: Advanced distributed-transpose interface181969 Node: An improved replacement for MPI_Alltoall183257 Node: FFTW MPI Wisdom185233 Ref: FFTW MPI Wisdom-Footnote-1187976 Node: Avoiding MPI Deadlocks188889 Node: FFTW MPI Performance Tips189918 Node: Combining MPI and Threads191387 Node: FFTW MPI Reference194858 Node: MPI Files and Data Types195437 Node: MPI Initialization196433 Node: Using MPI Plans197532 Node: MPI Data Distribution Functions199358 Node: MPI Plan Creation204814 Node: MPI Wisdom Communication215491 Node: FFTW MPI Fortran Interface216417 Ref: FFTW MPI Fortran Interface-Footnote-1222446 Node: Calling FFTW from Modern Fortran222853 Node: Overview of Fortran interface224204 Node: Extended and quadruple precision in Fortran227656 Node: Reversing array dimensions229037 Node: FFTW Fortran type reference232572 Node: Plan execution in Fortran237059 Node: Allocating aligned memory in Fortran239955 Node: Accessing the wisdom API from Fortran243319 Node: Wisdom File Export/Import from Fortran244096 Node: Wisdom String Export/Import from Fortran245758 Node: Wisdom Generic Export/Import from Fortran247746 Node: Defining an FFTW module249976 Node: Calling FFTW from Legacy Fortran251045 Node: Fortran-interface routines252602 Ref: Fortran-interface routines-Footnote-1256260 Ref: Fortran-interface routines-Footnote-2256463 Node: FFTW Constants in Fortran256596 Node: FFTW Execution in Fortran257751 Node: Fortran Examples260507 Node: Wisdom of Fortran?263926 Node: Upgrading from FFTW version 2265606 Ref: Upgrading from FFTW version 2-Footnote-1275229 Node: Installation and Customization275412 Node: Installation on Unix277056 Node: Installation on non-Unix systems285719 Node: Cycle Counters287934 Node: Generating your own code289686 Node: Acknowledgments291721 Node: License and Copyright295441 Node: Concept Index297310 Node: Library Index333952  End Tag Table fftw-3.3.3/doc/fftw3.info-10000644000175000017500000110453612054411105012223 00000000000000This is fftw3.info, produced by makeinfo version 4.13 from fftw3.texi. This manual is for FFTW (version 3.3.3, 25 November 2012). Copyright (C) 2003 Matteo Frigo. Copyright (C) 2003 Massachusetts Institute of Technology. Permission is granted to make and distribute verbatim copies of this manual provided the copyright notice and this permission notice are preserved on all copies. Permission is granted to copy and distribute modified versions of this manual under the conditions for verbatim copying, provided that the entire resulting derived work is distributed under the terms of a permission notice identical to this one. Permission is granted to copy and distribute translations of this manual into another language, under the above conditions for modified versions, except that this permission notice may be stated in a translation approved by the Free Software Foundation. INFO-DIR-SECTION Texinfo documentation system START-INFO-DIR-ENTRY * fftw3: (fftw3). FFTW User's Manual. END-INFO-DIR-ENTRY  File: fftw3.info, Node: Top, Next: Introduction, Prev: (dir), Up: (dir) FFTW User Manual **************** Welcome to FFTW, the Fastest Fourier Transform in the West. FFTW is a collection of fast C routines to compute the discrete Fourier transform. This manual documents FFTW version 3.3.3. * Menu: * Introduction:: * Tutorial:: * Other Important Topics:: * FFTW Reference:: * Multi-threaded FFTW:: * Distributed-memory FFTW with MPI:: * Calling FFTW from Modern Fortran:: * Calling FFTW from Legacy Fortran:: * Upgrading from FFTW version 2:: * Installation and Customization:: * Acknowledgments:: * License and Copyright:: * Concept Index:: * Library Index::  File: fftw3.info, Node: Introduction, Next: Tutorial, Prev: Top, Up: Top 1 Introduction ************** This manual documents version 3.3.3 of FFTW, the _Fastest Fourier Transform in the West_. FFTW is a comprehensive collection of fast C routines for computing the discrete Fourier transform (DFT) and various special cases thereof. * FFTW computes the DFT of complex data, real data, even- or odd-symmetric real data (these symmetric transforms are usually known as the discrete cosine or sine transform, respectively), and the discrete Hartley transform (DHT) of real data. * The input data can have arbitrary length. FFTW employs O(n log n) algorithms for all lengths, including prime numbers. * FFTW supports arbitrary multi-dimensional data. * FFTW supports the SSE, SSE2, AVX, Altivec, and MIPS PS instruction sets. * FFTW includes parallel (multi-threaded) transforms for shared-memory systems. * Starting with version 3.3, FFTW includes distributed-memory parallel transforms using MPI. We assume herein that you are familiar with the properties and uses of the DFT that are relevant to your application. Otherwise, see e.g. `The Fast Fourier Transform and Its Applications' by E. O. Brigham (Prentice-Hall, Englewood Cliffs, NJ, 1988). Our web page (http://www.fftw.org) also has links to FFT-related information online. In order to use FFTW effectively, you need to learn one basic concept of FFTW's internal structure: FFTW does not use a fixed algorithm for computing the transform, but instead it adapts the DFT algorithm to details of the underlying hardware in order to maximize performance. Hence, the computation of the transform is split into two phases. First, FFTW's "planner" "learns" the fastest way to compute the transform on your machine. The planner produces a data structure called a "plan" that contains this information. Subsequently, the plan is "executed" to transform the array of input data as dictated by the plan. The plan can be reused as many times as needed. In typical high-performance applications, many transforms of the same size are computed and, consequently, a relatively expensive initialization of this sort is acceptable. On the other hand, if you need a single transform of a given size, the one-time cost of the planner becomes significant. For this case, FFTW provides fast planners based on heuristics or on previously computed plans. FFTW supports transforms of data with arbitrary length, rank, multiplicity, and a general memory layout. In simple cases, however, this generality may be unnecessary and confusing. Consequently, we organized the interface to FFTW into three levels of increasing generality. * The "basic interface" computes a single transform of contiguous data. * The "advanced interface" computes transforms of multiple or strided arrays. * The "guru interface" supports the most general data layouts, multiplicities, and strides. We expect that most users will be best served by the basic interface, whereas the guru interface requires careful attention to the documentation to avoid problems. Besides the automatic performance adaptation performed by the planner, it is also possible for advanced users to customize FFTW manually. For example, if code space is a concern, we provide a tool that links only the subset of FFTW needed by your application. Conversely, you may need to extend FFTW because the standard distribution is not sufficient for your needs. For example, the standard FFTW distribution works most efficiently for arrays whose size can be factored into small primes (2, 3, 5, and 7), and otherwise it uses a slower general-purpose routine. If you need efficient transforms of other sizes, you can use FFTW's code generator, which produces fast C programs ("codelets") for any particular array size you may care about. For example, if you need transforms of size 513 = 19 x 3^3, you can customize FFTW to support the factor 19 efficiently. For more information regarding FFTW, see the paper, "The Design and Implementation of FFTW3," by M. Frigo and S. G. Johnson, which was an invited paper in `Proc. IEEE' 93 (2), p. 216 (2005). The code generator is described in the paper "A fast Fourier transform compiler", by M. Frigo, in the `Proceedings of the 1999 ACM SIGPLAN Conference on Programming Language Design and Implementation (PLDI), Atlanta, Georgia, May 1999'. These papers, along with the latest version of FFTW, the FAQ, benchmarks, and other links, are available at the FFTW home page (http://www.fftw.org). The current version of FFTW incorporates many good ideas from the past thirty years of FFT literature. In one way or another, FFTW uses the Cooley-Tukey algorithm, the prime factor algorithm, Rader's algorithm for prime sizes, and a split-radix algorithm (with a "conjugate-pair" variation pointed out to us by Dan Bernstein). FFTW's code generator also produces new algorithms that we do not completely understand. The reader is referred to the cited papers for the appropriate references. The rest of this manual is organized as follows. We first discuss the sequential (single-processor) implementation. We start by describing the basic interface/features of FFTW in *note Tutorial::. Next, *note Other Important Topics:: discusses data alignment (*note SIMD alignment and fftw_malloc::), the storage scheme of multi-dimensional arrays (*note Multi-dimensional Array Format::), and FFTW's mechanism for storing plans on disk (*note Words of Wisdom-Saving Plans::). Next, *note FFTW Reference:: provides comprehensive documentation of all FFTW's features. Parallel transforms are discussed in their own chapters: *note Multi-threaded FFTW:: and *note Distributed-memory FFTW with MPI::. Fortran programmers can also use FFTW, as described in *note Calling FFTW from Legacy Fortran:: and *note Calling FFTW from Modern Fortran::. *note Installation and Customization:: explains how to install FFTW in your computer system and how to adapt FFTW to your needs. License and copyright information is given in *note License and Copyright::. Finally, we thank all the people who helped us in *note Acknowledgments::.  File: fftw3.info, Node: Tutorial, Next: Other Important Topics, Prev: Introduction, Up: Top 2 Tutorial ********** * Menu: * Complex One-Dimensional DFTs:: * Complex Multi-Dimensional DFTs:: * One-Dimensional DFTs of Real Data:: * Multi-Dimensional DFTs of Real Data:: * More DFTs of Real Data:: This chapter describes the basic usage of FFTW, i.e., how to compute the Fourier transform of a single array. This chapter tells the truth, but not the _whole_ truth. Specifically, FFTW implements additional routines and flags that are not documented here, although in many cases we try to indicate where added capabilities exist. For more complete information, see *note FFTW Reference::. (Note that you need to compile and install FFTW before you can use it in a program. For the details of the installation, see *note Installation and Customization::.) We recommend that you read this tutorial in order.(1) At the least, read the first section (*note Complex One-Dimensional DFTs::) before reading any of the others, even if your main interest lies in one of the other transform types. Users of FFTW version 2 and earlier may also want to read *note Upgrading from FFTW version 2::. ---------- Footnotes ---------- (1) You can read the tutorial in bit-reversed order after computing your first transform.  File: fftw3.info, Node: Complex One-Dimensional DFTs, Next: Complex Multi-Dimensional DFTs, Prev: Tutorial, Up: Tutorial 2.1 Complex One-Dimensional DFTs ================================ Plan: To bother about the best method of accomplishing an accidental result. [Ambrose Bierce, `The Enlarged Devil's Dictionary'.] The basic usage of FFTW to compute a one-dimensional DFT of size `N' is simple, and it typically looks something like this code: #include ... { fftw_complex *in, *out; fftw_plan p; ... in = (fftw_complex*) fftw_malloc(sizeof(fftw_complex) * N); out = (fftw_complex*) fftw_malloc(sizeof(fftw_complex) * N); p = fftw_plan_dft_1d(N, in, out, FFTW_FORWARD, FFTW_ESTIMATE); ... fftw_execute(p); /* repeat as needed */ ... fftw_destroy_plan(p); fftw_free(in); fftw_free(out); } You must link this code with the `fftw3' library. On Unix systems, link with `-lfftw3 -lm'. The example code first allocates the input and output arrays. You can allocate them in any way that you like, but we recommend using `fftw_malloc', which behaves like `malloc' except that it properly aligns the array when SIMD instructions (such as SSE and Altivec) are available (*note SIMD alignment and fftw_malloc::). [Alternatively, we provide a convenient wrapper function `fftw_alloc_complex(N)' which has the same effect.] The data is an array of type `fftw_complex', which is by default a `double[2]' composed of the real (`in[i][0]') and imaginary (`in[i][1]') parts of a complex number. The next step is to create a "plan", which is an object that contains all the data that FFTW needs to compute the FFT. This function creates the plan: fftw_plan fftw_plan_dft_1d(int n, fftw_complex *in, fftw_complex *out, int sign, unsigned flags); The first argument, `n', is the size of the transform you are trying to compute. The size `n' can be any positive integer, but sizes that are products of small factors are transformed most efficiently (although prime sizes still use an O(n log n) algorithm). The next two arguments are pointers to the input and output arrays of the transform. These pointers can be equal, indicating an "in-place" transform. The fourth argument, `sign', can be either `FFTW_FORWARD' (`-1') or `FFTW_BACKWARD' (`+1'), and indicates the direction of the transform you are interested in; technically, it is the sign of the exponent in the transform. The `flags' argument is usually either `FFTW_MEASURE' or `FFTW_ESTIMATE'. `FFTW_MEASURE' instructs FFTW to run and measure the execution time of several FFTs in order to find the best way to compute the transform of size `n'. This process takes some time (usually a few seconds), depending on your machine and on the size of the transform. `FFTW_ESTIMATE', on the contrary, does not run any computation and just builds a reasonable plan that is probably sub-optimal. In short, if your program performs many transforms of the same size and initialization time is not important, use `FFTW_MEASURE'; otherwise use the estimate. _You must create the plan before initializing the input_, because `FFTW_MEASURE' overwrites the `in'/`out' arrays. (Technically, `FFTW_ESTIMATE' does not touch your arrays, but you should always create plans first just to be sure.) Once the plan has been created, you can use it as many times as you like for transforms on the specified `in'/`out' arrays, computing the actual transforms via `fftw_execute(plan)': void fftw_execute(const fftw_plan plan); The DFT results are stored in-order in the array `out', with the zero-frequency (DC) component in `out[0]'. If `in != out', the transform is "out-of-place" and the input array `in' is not modified. Otherwise, the input array is overwritten with the transform. If you want to transform a _different_ array of the same size, you can create a new plan with `fftw_plan_dft_1d' and FFTW automatically reuses the information from the previous plan, if possible. Alternatively, with the "guru" interface you can apply a given plan to a different array, if you are careful. *Note FFTW Reference::. When you are done with the plan, you deallocate it by calling `fftw_destroy_plan(plan)': void fftw_destroy_plan(fftw_plan plan); If you allocate an array with `fftw_malloc()' you must deallocate it with `fftw_free()'. Do not use `free()' or, heaven forbid, `delete'. FFTW computes an _unnormalized_ DFT. Thus, computing a forward followed by a backward transform (or vice versa) results in the original array scaled by `n'. For the definition of the DFT, see *note What FFTW Really Computes::. If you have a C compiler, such as `gcc', that supports the C99 standard, and you `#include ' _before_ `', then `fftw_complex' is the native double-precision complex type and you can manipulate it with ordinary arithmetic. Otherwise, FFTW defines its own complex type, which is bit-compatible with the C99 complex type. *Note Complex numbers::. (The C++ `' template class may also be usable via a typecast.) To use single or long-double precision versions of FFTW, replace the `fftw_' prefix by `fftwf_' or `fftwl_' and link with `-lfftw3f' or `-lfftw3l', but use the _same_ `' header file. Many more flags exist besides `FFTW_MEASURE' and `FFTW_ESTIMATE'. For example, use `FFTW_PATIENT' if you're willing to wait even longer for a possibly even faster plan (*note FFTW Reference::). You can also save plans for future use, as described by *note Words of Wisdom-Saving Plans::.  File: fftw3.info, Node: Complex Multi-Dimensional DFTs, Next: One-Dimensional DFTs of Real Data, Prev: Complex One-Dimensional DFTs, Up: Tutorial 2.2 Complex Multi-Dimensional DFTs ================================== Multi-dimensional transforms work much the same way as one-dimensional transforms: you allocate arrays of `fftw_complex' (preferably using `fftw_malloc'), create an `fftw_plan', execute it as many times as you want with `fftw_execute(plan)', and clean up with `fftw_destroy_plan(plan)' (and `fftw_free'). FFTW provides two routines for creating plans for 2d and 3d transforms, and one routine for creating plans of arbitrary dimensionality. The 2d and 3d routines have the following signature: fftw_plan fftw_plan_dft_2d(int n0, int n1, fftw_complex *in, fftw_complex *out, int sign, unsigned flags); fftw_plan fftw_plan_dft_3d(int n0, int n1, int n2, fftw_complex *in, fftw_complex *out, int sign, unsigned flags); These routines create plans for `n0' by `n1' two-dimensional (2d) transforms and `n0' by `n1' by `n2' 3d transforms, respectively. All of these transforms operate on contiguous arrays in the C-standard "row-major" order, so that the last dimension has the fastest-varying index in the array. This layout is described further in *note Multi-dimensional Array Format::. FFTW can also compute transforms of higher dimensionality. In order to avoid confusion between the various meanings of the the word "dimension", we use the term _rank_ to denote the number of independent indices in an array.(1) For example, we say that a 2d transform has rank 2, a 3d transform has rank 3, and so on. You can plan transforms of arbitrary rank by means of the following function: fftw_plan fftw_plan_dft(int rank, const int *n, fftw_complex *in, fftw_complex *out, int sign, unsigned flags); Here, `n' is a pointer to an array `n[rank]' denoting an `n[0]' by `n[1]' by ... by `n[rank-1]' transform. Thus, for example, the call fftw_plan_dft_2d(n0, n1, in, out, sign, flags); is equivalent to the following code fragment: int n[2]; n[0] = n0; n[1] = n1; fftw_plan_dft(2, n, in, out, sign, flags); `fftw_plan_dft' is not restricted to 2d and 3d transforms, however, but it can plan transforms of arbitrary rank. You may have noticed that all the planner routines described so far have overlapping functionality. For example, you can plan a 1d or 2d transform by using `fftw_plan_dft' with a `rank' of `1' or `2', or even by calling `fftw_plan_dft_3d' with `n0' and/or `n1' equal to `1' (with no loss in efficiency). This pattern continues, and FFTW's planning routines in general form a "partial order," sequences of interfaces with strictly increasing generality but correspondingly greater complexity. `fftw_plan_dft' is the most general complex-DFT routine that we describe in this tutorial, but there are also the advanced and guru interfaces, which allow one to efficiently combine multiple/strided transforms into a single FFTW plan, transform a subset of a larger multi-dimensional array, and/or to handle more general complex-number formats. For more information, see *note FFTW Reference::. ---------- Footnotes ---------- (1) The term "rank" is commonly used in the APL, FORTRAN, and Common Lisp traditions, although it is not so common in the C world.  File: fftw3.info, Node: One-Dimensional DFTs of Real Data, Next: Multi-Dimensional DFTs of Real Data, Prev: Complex Multi-Dimensional DFTs, Up: Tutorial 2.3 One-Dimensional DFTs of Real Data ===================================== In many practical applications, the input data `in[i]' are purely real numbers, in which case the DFT output satisfies the "Hermitian" redundancy: `out[i]' is the conjugate of `out[n-i]'. It is possible to take advantage of these circumstances in order to achieve roughly a factor of two improvement in both speed and memory usage. In exchange for these speed and space advantages, the user sacrifices some of the simplicity of FFTW's complex transforms. First of all, the input and output arrays are of _different sizes and types_: the input is `n' real numbers, while the output is `n/2+1' complex numbers (the non-redundant outputs); this also requires slight "padding" of the input array for in-place transforms. Second, the inverse transform (complex to real) has the side-effect of _overwriting its input array_, by default. Neither of these inconveniences should pose a serious problem for users, but it is important to be aware of them. The routines to perform real-data transforms are almost the same as those for complex transforms: you allocate arrays of `double' and/or `fftw_complex' (preferably using `fftw_malloc' or `fftw_alloc_complex'), create an `fftw_plan', execute it as many times as you want with `fftw_execute(plan)', and clean up with `fftw_destroy_plan(plan)' (and `fftw_free'). The only differences are that the input (or output) is of type `double' and there are new routines to create the plan. In one dimension: fftw_plan fftw_plan_dft_r2c_1d(int n, double *in, fftw_complex *out, unsigned flags); fftw_plan fftw_plan_dft_c2r_1d(int n, fftw_complex *in, double *out, unsigned flags); for the real input to complex-Hermitian output ("r2c") and complex-Hermitian input to real output ("c2r") transforms. Unlike the complex DFT planner, there is no `sign' argument. Instead, r2c DFTs are always `FFTW_FORWARD' and c2r DFTs are always `FFTW_BACKWARD'. (For single/long-double precision `fftwf' and `fftwl', `double' should be replaced by `float' and `long double', respectively.) Here, `n' is the "logical" size of the DFT, not necessarily the physical size of the array. In particular, the real (`double') array has `n' elements, while the complex (`fftw_complex') array has `n/2+1' elements (where the division is rounded down). For an in-place transform, `in' and `out' are aliased to the same array, which must be big enough to hold both; so, the real array would actually have `2*(n/2+1)' elements, where the elements beyond the first `n' are unused padding. (Note that this is very different from the concept of "zero-padding" a transform to a larger length, which changes the logical size of the DFT by actually adding new input data.) The kth element of the complex array is exactly the same as the kth element of the corresponding complex DFT. All positive `n' are supported; products of small factors are most efficient, but an O(n log n) algorithm is used even for prime sizes. As noted above, the c2r transform destroys its input array even for out-of-place transforms. This can be prevented, if necessary, by including `FFTW_PRESERVE_INPUT' in the `flags', with unfortunately some sacrifice in performance. This flag is also not currently supported for multi-dimensional real DFTs (next section). Readers familiar with DFTs of real data will recall that the 0th (the "DC") and `n/2'-th (the "Nyquist" frequency, when `n' is even) elements of the complex output are purely real. Some implementations therefore store the Nyquist element where the DC imaginary part would go, in order to make the input and output arrays the same size. Such packing, however, does not generalize well to multi-dimensional transforms, and the space savings are miniscule in any case; FFTW does not support it. An alternative interface for one-dimensional r2c and c2r DFTs can be found in the `r2r' interface (*note The Halfcomplex-format DFT::), with "halfcomplex"-format output that _is_ the same size (and type) as the input array. That interface, although it is not very useful for multi-dimensional transforms, may sometimes yield better performance.  File: fftw3.info, Node: Multi-Dimensional DFTs of Real Data, Next: More DFTs of Real Data, Prev: One-Dimensional DFTs of Real Data, Up: Tutorial 2.4 Multi-Dimensional DFTs of Real Data ======================================= Multi-dimensional DFTs of real data use the following planner routines: fftw_plan fftw_plan_dft_r2c_2d(int n0, int n1, double *in, fftw_complex *out, unsigned flags); fftw_plan fftw_plan_dft_r2c_3d(int n0, int n1, int n2, double *in, fftw_complex *out, unsigned flags); fftw_plan fftw_plan_dft_r2c(int rank, const int *n, double *in, fftw_complex *out, unsigned flags); as well as the corresponding `c2r' routines with the input/output types swapped. These routines work similarly to their complex analogues, except for the fact that here the complex output array is cut roughly in half and the real array requires padding for in-place transforms (as in 1d, above). As before, `n' is the logical size of the array, and the consequences of this on the the format of the complex arrays deserve careful attention. Suppose that the real data has dimensions n[0] x n[1] x n[2] x ... x n[d-1] (in row-major order). Then, after an r2c transform, the output is an n[0] x n[1] x n[2] x ... x (n[d-1]/2 + 1) array of `fftw_complex' values in row-major order, corresponding to slightly over half of the output of the corresponding complex DFT. (The division is rounded down.) The ordering of the data is otherwise exactly the same as in the complex-DFT case. For out-of-place transforms, this is the end of the story: the real data is stored as a row-major array of size n[0] x n[1] x n[2] x ... x n[d-1] and the complex data is stored as a row-major array of size n[0] x n[1] x n[2] x ... x (n[d-1]/2 + 1) . For in-place transforms, however, extra padding of the real-data array is necessary because the complex array is larger than the real array, and the two arrays share the same memory locations. Thus, for in-place transforms, the final dimension of the real-data array must be padded with extra values to accommodate the size of the complex data--two values if the last dimension is even and one if it is odd. That is, the last dimension of the real data must physically contain 2 * (n[d-1]/2+1) `double' values (exactly enough to hold the complex data). This physical array size does not, however, change the _logical_ array size--only n[d-1] values are actually stored in the last dimension, and n[d-1] is the last dimension passed to the plan-creation routine. For example, consider the transform of a two-dimensional real array of size `n0' by `n1'. The output of the r2c transform is a two-dimensional complex array of size `n0' by `n1/2+1', where the `y' dimension has been cut nearly in half because of redundancies in the output. Because `fftw_complex' is twice the size of `double', the output array is slightly bigger than the input array. Thus, if we want to compute the transform in place, we must _pad_ the input array so that it is of size `n0' by `2*(n1/2+1)'. If `n1' is even, then there are two padding elements at the end of each row (which need not be initialized, as they are only used for output). These transforms are unnormalized, so an r2c followed by a c2r transform (or vice versa) will result in the original data scaled by the number of real data elements--that is, the product of the (logical) dimensions of the real data. (Because the last dimension is treated specially, if it is equal to `1' the transform is _not_ equivalent to a lower-dimensional r2c/c2r transform. In that case, the last complex dimension also has size `1' (`=1/2+1'), and no advantage is gained over the complex transforms.)  File: fftw3.info, Node: More DFTs of Real Data, Prev: Multi-Dimensional DFTs of Real Data, Up: Tutorial 2.5 More DFTs of Real Data ========================== * Menu: * The Halfcomplex-format DFT:: * Real even/odd DFTs (cosine/sine transforms):: * The Discrete Hartley Transform:: FFTW supports several other transform types via a unified "r2r" (real-to-real) interface, so called because it takes a real (`double') array and outputs a real array of the same size. These r2r transforms currently fall into three categories: DFTs of real input and complex-Hermitian output in halfcomplex format, DFTs of real input with even/odd symmetry (a.k.a. discrete cosine/sine transforms, DCTs/DSTs), and discrete Hartley transforms (DHTs), all described in more detail by the following sections. The r2r transforms follow the by now familiar interface of creating an `fftw_plan', executing it with `fftw_execute(plan)', and destroying it with `fftw_destroy_plan(plan)'. Furthermore, all r2r transforms share the same planner interface: fftw_plan fftw_plan_r2r_1d(int n, double *in, double *out, fftw_r2r_kind kind, unsigned flags); fftw_plan fftw_plan_r2r_2d(int n0, int n1, double *in, double *out, fftw_r2r_kind kind0, fftw_r2r_kind kind1, unsigned flags); fftw_plan fftw_plan_r2r_3d(int n0, int n1, int n2, double *in, double *out, fftw_r2r_kind kind0, fftw_r2r_kind kind1, fftw_r2r_kind kind2, unsigned flags); fftw_plan fftw_plan_r2r(int rank, const int *n, double *in, double *out, const fftw_r2r_kind *kind, unsigned flags); Just as for the complex DFT, these plan 1d/2d/3d/multi-dimensional transforms for contiguous arrays in row-major order, transforming (real) input to output of the same size, where `n' specifies the _physical_ dimensions of the arrays. All positive `n' are supported (with the exception of `n=1' for the `FFTW_REDFT00' kind, noted in the real-even subsection below); products of small factors are most efficient (factorizing `n-1' and `n+1' for `FFTW_REDFT00' and `FFTW_RODFT00' kinds, described below), but an O(n log n) algorithm is used even for prime sizes. Each dimension has a "kind" parameter, of type `fftw_r2r_kind', specifying the kind of r2r transform to be used for that dimension. (In the case of `fftw_plan_r2r', this is an array `kind[rank]' where `kind[i]' is the transform kind for the dimension `n[i]'.) The kind can be one of a set of predefined constants, defined in the following subsections. In other words, FFTW computes the separable product of the specified r2r transforms over each dimension, which can be used e.g. for partial differential equations with mixed boundary conditions. (For some r2r kinds, notably the halfcomplex DFT and the DHT, such a separable product is somewhat problematic in more than one dimension, however, as is described below.) In the current version of FFTW, all r2r transforms except for the halfcomplex type are computed via pre- or post-processing of halfcomplex transforms, and they are therefore not as fast as they could be. Since most other general DCT/DST codes employ a similar algorithm, however, FFTW's implementation should provide at least competitive performance.  File: fftw3.info, Node: The Halfcomplex-format DFT, Next: Real even/odd DFTs (cosine/sine transforms), Prev: More DFTs of Real Data, Up: More DFTs of Real Data 2.5.1 The Halfcomplex-format DFT -------------------------------- An r2r kind of `FFTW_R2HC' ("r2hc") corresponds to an r2c DFT (*note One-Dimensional DFTs of Real Data::) but with "halfcomplex" format output, and may sometimes be faster and/or more convenient than the latter. The inverse "hc2r" transform is of kind `FFTW_HC2R'. This consists of the non-redundant half of the complex output for a 1d real-input DFT of size `n', stored as a sequence of `n' real numbers (`double') in the format: r0, r1, r2, r(n/2), i((n+1)/2-1), ..., i2, i1 Here, rk is the real part of the kth output, and ik is the imaginary part. (Division by 2 is rounded down.) For a halfcomplex array `hc[n]', the kth component thus has its real part in `hc[k]' and its imaginary part in `hc[n-k]', with the exception of `k' `==' `0' or `n/2' (the latter only if `n' is even)--in these two cases, the imaginary part is zero due to symmetries of the real-input DFT, and is not stored. Thus, the r2hc transform of `n' real values is a halfcomplex array of length `n', and vice versa for hc2r. Aside from the differing format, the output of `FFTW_R2HC'/`FFTW_HC2R' is otherwise exactly the same as for the corresponding 1d r2c/c2r transform (i.e. `FFTW_FORWARD'/`FFTW_BACKWARD' transforms, respectively). Recall that these transforms are unnormalized, so r2hc followed by hc2r will result in the original data multiplied by `n'. Furthermore, like the c2r transform, an out-of-place hc2r transform will _destroy its input_ array. Although these halfcomplex transforms can be used with the multi-dimensional r2r interface, the interpretation of such a separable product of transforms along each dimension is problematic. For example, consider a two-dimensional `n0' by `n1', r2hc by r2hc transform planned by `fftw_plan_r2r_2d(n0, n1, in, out, FFTW_R2HC, FFTW_R2HC, FFTW_MEASURE)'. Conceptually, FFTW first transforms the rows (of size `n1') to produce halfcomplex rows, and then transforms the columns (of size `n0'). Half of these column transforms, however, are of imaginary parts, and should therefore be multiplied by i and combined with the r2hc transforms of the real columns to produce the 2d DFT amplitudes; FFTW's r2r transform does _not_ perform this combination for you. Thus, if a multi-dimensional real-input/output DFT is required, we recommend using the ordinary r2c/c2r interface (*note Multi-Dimensional DFTs of Real Data::).  File: fftw3.info, Node: Real even/odd DFTs (cosine/sine transforms), Next: The Discrete Hartley Transform, Prev: The Halfcomplex-format DFT, Up: More DFTs of Real Data 2.5.2 Real even/odd DFTs (cosine/sine transforms) ------------------------------------------------- The Fourier transform of a real-even function f(-x) = f(x) is real-even, and i times the Fourier transform of a real-odd function f(-x) = -f(x) is real-odd. Similar results hold for a discrete Fourier transform, and thus for these symmetries the need for complex inputs/outputs is entirely eliminated. Moreover, one gains a factor of two in speed/space from the fact that the data are real, and an additional factor of two from the even/odd symmetry: only the non-redundant (first) half of the array need be stored. The result is the real-even DFT ("REDFT") and the real-odd DFT ("RODFT"), also known as the discrete cosine and sine transforms ("DCT" and "DST"), respectively. (In this section, we describe the 1d transforms; multi-dimensional transforms are just a separable product of these transforms operating along each dimension.) Because of the discrete sampling, one has an additional choice: is the data even/odd around a sampling point, or around the point halfway between two samples? The latter corresponds to _shifting_ the samples by _half_ an interval, and gives rise to several transform variants denoted by REDFTab and RODFTab: a and b are 0 or 1, and indicate whether the input (a) and/or output (b) are shifted by half a sample (1 means it is shifted). These are also known as types I-IV of the DCT and DST, and all four types are supported by FFTW's r2r interface.(1) The r2r kinds for the various REDFT and RODFT types supported by FFTW, along with the boundary conditions at both ends of the _input_ array (`n' real numbers `in[j=0..n-1]'), are: * `FFTW_REDFT00' (DCT-I): even around j=0 and even around j=n-1. * `FFTW_REDFT10' (DCT-II, "the" DCT): even around j=-0.5 and even around j=n-0.5. * `FFTW_REDFT01' (DCT-III, "the" IDCT): even around j=0 and odd around j=n. * `FFTW_REDFT11' (DCT-IV): even around j=-0.5 and odd around j=n-0.5. * `FFTW_RODFT00' (DST-I): odd around j=-1 and odd around j=n. * `FFTW_RODFT10' (DST-II): odd around j=-0.5 and odd around j=n-0.5. * `FFTW_RODFT01' (DST-III): odd around j=-1 and even around j=n-1. * `FFTW_RODFT11' (DST-IV): odd around j=-0.5 and even around j=n-0.5. Note that these symmetries apply to the "logical" array being transformed; *there are no constraints on your physical input data*. So, for example, if you specify a size-5 REDFT00 (DCT-I) of the data abcde, it corresponds to the DFT of the logical even array abcdedcb of size 8. A size-4 REDFT10 (DCT-II) of the data abcd corresponds to the size-8 logical DFT of the even array abcddcba, shifted by half a sample. All of these transforms are invertible. The inverse of R*DFT00 is R*DFT00; of R*DFT10 is R*DFT01 and vice versa (these are often called simply "the" DCT and IDCT, respectively); and of R*DFT11 is R*DFT11. However, the transforms computed by FFTW are unnormalized, exactly like the corresponding real and complex DFTs, so computing a transform followed by its inverse yields the original array scaled by N, where N is the _logical_ DFT size. For REDFT00, N=2(n-1); for RODFT00, N=2(n+1); otherwise, N=2n. Note that the boundary conditions of the transform output array are given by the input boundary conditions of the inverse transform. Thus, the above transforms are all inequivalent in terms of input/output boundary conditions, even neglecting the 0.5 shift difference. FFTW is most efficient when N is a product of small factors; note that this _differs_ from the factorization of the physical size `n' for REDFT00 and RODFT00! There is another oddity: `n=1' REDFT00 transforms correspond to N=0, and so are _not defined_ (the planner will return `NULL'). Otherwise, any positive `n' is supported. For the precise mathematical definitions of these transforms as used by FFTW, see *note What FFTW Really Computes::. (For people accustomed to the DCT/DST, FFTW's definitions have a coefficient of 2 in front of the cos/sin functions so that they correspond precisely to an even/odd DFT of size N. Some authors also include additional multiplicative factors of sqrt(2) for selected inputs and outputs; this makes the transform orthogonal, but sacrifices the direct equivalence to a symmetric DFT.) Which type do you need? ....................... Since the required flavor of even/odd DFT depends upon your problem, you are the best judge of this choice, but we can make a few comments on relative efficiency to help you in your selection. In particular, R*DFT01 and R*DFT10 tend to be slightly faster than R*DFT11 (especially for odd sizes), while the R*DFT00 transforms are sometimes significantly slower (especially for even sizes).(2) Thus, if only the boundary conditions on the transform inputs are specified, we generally recommend R*DFT10 over R*DFT00 and R*DFT01 over R*DFT11 (unless the half-sample shift or the self-inverse property is significant for your problem). If performance is important to you and you are using only small sizes (say n<200), e.g. for multi-dimensional transforms, then you might consider generating hard-coded transforms of those sizes and types that you are interested in (*note Generating your own code::). We are interested in hearing what types of symmetric transforms you find most useful. ---------- Footnotes ---------- (1) There are also type V-VIII transforms, which correspond to a logical DFT of _odd_ size N, independent of whether the physical size `n' is odd, but we do not support these variants. (2) R*DFT00 is sometimes slower in FFTW because we discovered that the standard algorithm for computing this by a pre/post-processed real DFT--the algorithm used in FFTPACK, Numerical Recipes, and other sources for decades now--has serious numerical problems: it already loses several decimal places of accuracy for 16k sizes. There seem to be only two alternatives in the literature that do not suffer similarly: a recursive decomposition into smaller DCTs, which would require a large set of codelets for efficiency and generality, or sacrificing a factor of 2 in speed to use a real DFT of twice the size. We currently employ the latter technique for general n, as well as a limited form of the former method: a split-radix decomposition when n is odd (N a multiple of 4). For N containing many factors of 2, the split-radix method seems to recover most of the speed of the standard algorithm without the accuracy tradeoff.  File: fftw3.info, Node: The Discrete Hartley Transform, Prev: Real even/odd DFTs (cosine/sine transforms), Up: More DFTs of Real Data 2.5.3 The Discrete Hartley Transform ------------------------------------ If you are planning to use the DHT because you've heard that it is "faster" than the DFT (FFT), *stop here*. The DHT is not faster than the DFT. That story is an old but enduring misconception that was debunked in 1987. The discrete Hartley transform (DHT) is an invertible linear transform closely related to the DFT. In the DFT, one multiplies each input by cos - i * sin (a complex exponential), whereas in the DHT each input is multiplied by simply cos + sin. Thus, the DHT transforms `n' real numbers to `n' real numbers, and has the convenient property of being its own inverse. In FFTW, a DHT (of any positive `n') can be specified by an r2r kind of `FFTW_DHT'. Like the DFT, in FFTW the DHT is unnormalized, so computing a DHT of size `n' followed by another DHT of the same size will result in the original array multiplied by `n'. The DHT was originally proposed as a more efficient alternative to the DFT for real data, but it was subsequently shown that a specialized DFT (such as FFTW's r2hc or r2c transforms) could be just as fast. In FFTW, the DHT is actually computed by post-processing an r2hc transform, so there is ordinarily no reason to prefer it from a performance perspective.(1) However, we have heard rumors that the DHT might be the most appropriate transform in its own right for certain applications, and we would be very interested to hear from anyone who finds it useful. If `FFTW_DHT' is specified for multiple dimensions of a multi-dimensional transform, FFTW computes the separable product of 1d DHTs along each dimension. Unfortunately, this is not quite the same thing as a true multi-dimensional DHT; you can compute the latter, if necessary, with at most `rank-1' post-processing passes [see e.g. H. Hao and R. N. Bracewell, Proc. IEEE 75, 264-266 (1987)]. For the precise mathematical definition of the DHT as used by FFTW, see *note What FFTW Really Computes::. ---------- Footnotes ---------- (1) We provide the DHT mainly as a byproduct of some internal algorithms. FFTW computes a real input/output DFT of _prime_ size by re-expressing it as a DHT plus post/pre-processing and then using Rader's prime-DFT algorithm adapted to the DHT.  File: fftw3.info, Node: Other Important Topics, Next: FFTW Reference, Prev: Tutorial, Up: Top 3 Other Important Topics ************************ * Menu: * SIMD alignment and fftw_malloc:: * Multi-dimensional Array Format:: * Words of Wisdom-Saving Plans:: * Caveats in Using Wisdom::  File: fftw3.info, Node: SIMD alignment and fftw_malloc, Next: Multi-dimensional Array Format, Prev: Other Important Topics, Up: Other Important Topics 3.1 SIMD alignment and fftw_malloc ================================== SIMD, which stands for "Single Instruction Multiple Data," is a set of special operations supported by some processors to perform a single operation on several numbers (usually 2 or 4) simultaneously. SIMD floating-point instructions are available on several popular CPUs: SSE/SSE2/AVX on recent x86/x86-64 processors, AltiVec (single precision) on some PowerPCs (Apple G4 and higher), NEON on some ARM models, and MIPS Paired Single (currently only in FFTW 3.2.x). FFTW can be compiled to support the SIMD instructions on any of these systems. A program linking to an FFTW library compiled with SIMD support can obtain a nonnegligible speedup for most complex and r2c/c2r transforms. In order to obtain this speedup, however, the arrays of complex (or real) data passed to FFTW must be specially aligned in memory (typically 16-byte aligned), and often this alignment is more stringent than that provided by the usual `malloc' (etc.) allocation routines. In order to guarantee proper alignment for SIMD, therefore, in case your program is ever linked against a SIMD-using FFTW, we recommend allocating your transform data with `fftw_malloc' and de-allocating it with `fftw_free'. These have exactly the same interface and behavior as `malloc'/`free', except that for a SIMD FFTW they ensure that the returned pointer has the necessary alignment (by calling `memalign' or its equivalent on your OS). You are not _required_ to use `fftw_malloc'. You can allocate your data in any way that you like, from `malloc' to `new' (in C++) to a fixed-size array declaration. If the array happens not to be properly aligned, FFTW will not use the SIMD extensions. Since `fftw_malloc' only ever needs to be used for real and complex arrays, we provide two convenient wrapper routines `fftw_alloc_real(N)' and `fftw_alloc_complex(N)' that are equivalent to `(double*)fftw_malloc(sizeof(double) * N)' and `(fftw_complex*)fftw_malloc(sizeof(fftw_complex) * N)', respectively (or their equivalents in other precisions).  File: fftw3.info, Node: Multi-dimensional Array Format, Next: Words of Wisdom-Saving Plans, Prev: SIMD alignment and fftw_malloc, Up: Other Important Topics 3.2 Multi-dimensional Array Format ================================== This section describes the format in which multi-dimensional arrays are stored in FFTW. We felt that a detailed discussion of this topic was necessary. Since several different formats are common, this topic is often a source of confusion. * Menu: * Row-major Format:: * Column-major Format:: * Fixed-size Arrays in C:: * Dynamic Arrays in C:: * Dynamic Arrays in C-The Wrong Way::  File: fftw3.info, Node: Row-major Format, Next: Column-major Format, Prev: Multi-dimensional Array Format, Up: Multi-dimensional Array Format 3.2.1 Row-major Format ---------------------- The multi-dimensional arrays passed to `fftw_plan_dft' etcetera are expected to be stored as a single contiguous block in "row-major" order (sometimes called "C order"). Basically, this means that as you step through adjacent memory locations, the first dimension's index varies most slowly and the last dimension's index varies most quickly. To be more explicit, let us consider an array of rank d whose dimensions are n[0] x n[1] x n[2] x ... x n[d-1] . Now, we specify a location in the array by a sequence of d (zero-based) indices, one for each dimension: (i[0], i[1], ..., i[d-1]). If the array is stored in row-major order, then this element is located at the position i[d-1] + n[d-1] * (i[d-2] + n[d-2] * (... + n[1] * i[0])). Note that, for the ordinary complex DFT, each element of the array must be of type `fftw_complex'; i.e. a (real, imaginary) pair of (double-precision) numbers. In the advanced FFTW interface, the physical dimensions n from which the indices are computed can be different from (larger than) the logical dimensions of the transform to be computed, in order to transform a subset of a larger array. Note also that, in the advanced interface, the expression above is multiplied by a "stride" to get the actual array index--this is useful in situations where each element of the multi-dimensional array is actually a data structure (or another array), and you just want to transform a single field. In the basic interface, however, the stride is 1.  File: fftw3.info, Node: Column-major Format, Next: Fixed-size Arrays in C, Prev: Row-major Format, Up: Multi-dimensional Array Format 3.2.2 Column-major Format ------------------------- Readers from the Fortran world are used to arrays stored in "column-major" order (sometimes called "Fortran order"). This is essentially the exact opposite of row-major order in that, here, the _first_ dimension's index varies most quickly. If you have an array stored in column-major order and wish to transform it using FFTW, it is quite easy to do. When creating the plan, simply pass the dimensions of the array to the planner in _reverse order_. For example, if your array is a rank three `N x M x L' matrix in column-major order, you should pass the dimensions of the array as if it were an `L x M x N' matrix (which it is, from the perspective of FFTW). This is done for you _automatically_ by the FFTW legacy-Fortran interface (*note Calling FFTW from Legacy Fortran::), but you must do it manually with the modern Fortran interface (*note Reversing array dimensions::).  File: fftw3.info, Node: Fixed-size Arrays in C, Next: Dynamic Arrays in C, Prev: Column-major Format, Up: Multi-dimensional Array Format 3.2.3 Fixed-size Arrays in C ---------------------------- A multi-dimensional array whose size is declared at compile time in C is _already_ in row-major order. You don't have to do anything special to transform it. For example: { fftw_complex data[N0][N1][N2]; fftw_plan plan; ... plan = fftw_plan_dft_3d(N0, N1, N2, &data[0][0][0], &data[0][0][0], FFTW_FORWARD, FFTW_ESTIMATE); ... } This will plan a 3d in-place transform of size `N0 x N1 x N2'. Notice how we took the address of the zero-th element to pass to the planner (we could also have used a typecast). However, we tend to _discourage_ users from declaring their arrays in this way, for two reasons. First, this allocates the array on the stack ("automatic" storage), which has a very limited size on most operating systems (declaring an array with more than a few thousand elements will often cause a crash). (You can get around this limitation on many systems by declaring the array as `static' and/or global, but that has its own drawbacks.) Second, it may not optimally align the array for use with a SIMD FFTW (*note SIMD alignment and fftw_malloc::). Instead, we recommend using `fftw_malloc', as described below.  File: fftw3.info, Node: Dynamic Arrays in C, Next: Dynamic Arrays in C-The Wrong Way, Prev: Fixed-size Arrays in C, Up: Multi-dimensional Array Format 3.2.4 Dynamic Arrays in C ------------------------- We recommend allocating most arrays dynamically, with `fftw_malloc'. This isn't too hard to do, although it is not as straightforward for multi-dimensional arrays as it is for one-dimensional arrays. Creating the array is simple: using a dynamic-allocation routine like `fftw_malloc', allocate an array big enough to store N `fftw_complex' values (for a complex DFT), where N is the product of the sizes of the array dimensions (i.e. the total number of complex values in the array). For example, here is code to allocate a 5 x 12 x 27 rank-3 array: fftw_complex *an_array; an_array = (fftw_complex*) fftw_malloc(5*12*27 * sizeof(fftw_complex)); Accessing the array elements, however, is more tricky--you can't simply use multiple applications of the `[]' operator like you could for fixed-size arrays. Instead, you have to explicitly compute the offset into the array using the formula given earlier for row-major arrays. For example, to reference the (i,j,k)-th element of the array allocated above, you would use the expression `an_array[k + 27 * (j + 12 * i)]'. This pain can be alleviated somewhat by defining appropriate macros, or, in C++, creating a class and overloading the `()' operator. The recent C99 standard provides a way to reinterpret the dynamic array as a "variable-length" multi-dimensional array amenable to `[]', but this feature is not yet widely supported by compilers.  File: fftw3.info, Node: Dynamic Arrays in C-The Wrong Way, Prev: Dynamic Arrays in C, Up: Multi-dimensional Array Format 3.2.5 Dynamic Arrays in C--The Wrong Way ---------------------------------------- A different method for allocating multi-dimensional arrays in C is often suggested that is incompatible with FFTW: _using it will cause FFTW to die a painful death_. We discuss the technique here, however, because it is so commonly known and used. This method is to create arrays of pointers of arrays of pointers of ...etcetera. For example, the analogue in this method to the example above is: int i,j; fftw_complex ***a_bad_array; /* another way to make a 5x12x27 array */ a_bad_array = (fftw_complex ***) malloc(5 * sizeof(fftw_complex **)); for (i = 0; i < 5; ++i) { a_bad_array[i] = (fftw_complex **) malloc(12 * sizeof(fftw_complex *)); for (j = 0; j < 12; ++j) a_bad_array[i][j] = (fftw_complex *) malloc(27 * sizeof(fftw_complex)); } As you can see, this sort of array is inconvenient to allocate (and deallocate). On the other hand, it has the advantage that the (i,j,k)-th element can be referenced simply by `a_bad_array[i][j][k]'. If you like this technique and want to maximize convenience in accessing the array, but still want to pass the array to FFTW, you can use a hybrid method. Allocate the array as one contiguous block, but also declare an array of arrays of pointers that point to appropriate places in the block. That sort of trick is beyond the scope of this documentation; for more information on multi-dimensional arrays in C, see the `comp.lang.c' FAQ (http://c-faq.com/aryptr/dynmuldimary.html).  File: fftw3.info, Node: Words of Wisdom-Saving Plans, Next: Caveats in Using Wisdom, Prev: Multi-dimensional Array Format, Up: Other Important Topics 3.3 Words of Wisdom--Saving Plans ================================= FFTW implements a method for saving plans to disk and restoring them. In fact, what FFTW does is more general than just saving and loading plans. The mechanism is called "wisdom". Here, we describe this feature at a high level. *Note FFTW Reference::, for a less casual but more complete discussion of how to use wisdom in FFTW. Plans created with the `FFTW_MEASURE', `FFTW_PATIENT', or `FFTW_EXHAUSTIVE' options produce near-optimal FFT performance, but may require a long time to compute because FFTW must measure the runtime of many possible plans and select the best one. This setup is designed for the situations where so many transforms of the same size must be computed that the start-up time is irrelevant. For short initialization times, but slower transforms, we have provided `FFTW_ESTIMATE'. The `wisdom' mechanism is a way to get the best of both worlds: you compute a good plan once, save it to disk, and later reload it as many times as necessary. The wisdom mechanism can actually save and reload many plans at once, not just one. Whenever you create a plan, the FFTW planner accumulates wisdom, which is information sufficient to reconstruct the plan. After planning, you can save this information to disk by means of the function: int fftw_export_wisdom_to_filename(const char *filename); (This function returns non-zero on success.) The next time you run the program, you can restore the wisdom with `fftw_import_wisdom_from_filename' (which also returns non-zero on success), and then recreate the plan using the same flags as before. int fftw_import_wisdom_from_filename(const char *filename); Wisdom is automatically used for any size to which it is applicable, as long as the planner flags are not more "patient" than those with which the wisdom was created. For example, wisdom created with `FFTW_MEASURE' can be used if you later plan with `FFTW_ESTIMATE' or `FFTW_MEASURE', but not with `FFTW_PATIENT'. The `wisdom' is cumulative, and is stored in a global, private data structure managed internally by FFTW. The storage space required is minimal, proportional to the logarithm of the sizes the wisdom was generated from. If memory usage is a concern, however, the wisdom can be forgotten and its associated memory freed by calling: void fftw_forget_wisdom(void); Wisdom can be exported to a file, a string, or any other medium. For details, see *note Wisdom::.  File: fftw3.info, Node: Caveats in Using Wisdom, Prev: Words of Wisdom-Saving Plans, Up: Other Important Topics 3.4 Caveats in Using Wisdom =========================== For in much wisdom is much grief, and he that increaseth knowledge increaseth sorrow. [Ecclesiastes 1:18] There are pitfalls to using wisdom, in that it can negate FFTW's ability to adapt to changing hardware and other conditions. For example, it would be perfectly possible to export wisdom from a program running on one processor and import it into a program running on another processor. Doing so, however, would mean that the second program would use plans optimized for the first processor, instead of the one it is running on. It should be safe to reuse wisdom as long as the hardware and program binaries remain unchanged. (Actually, the optimal plan may change even between runs of the same binary on identical hardware, due to differences in the virtual memory environment, etcetera. Users seriously interested in performance should worry about this problem, too.) It is likely that, if the same wisdom is used for two different program binaries, even running on the same machine, the plans may be sub-optimal because of differing code alignments. It is therefore wise to recreate wisdom every time an application is recompiled. The more the underlying hardware and software changes between the creation of wisdom and its use, the greater grows the risk of sub-optimal plans. Nevertheless, if the choice is between using `FFTW_ESTIMATE' or using possibly-suboptimal wisdom (created on the same machine, but for a different binary), the wisdom is likely to be better. For this reason, we provide a function to import wisdom from a standard system-wide location (`/etc/fftw/wisdom' on Unix): int fftw_import_system_wisdom(void); FFTW also provides a standalone program, `fftw-wisdom' (described by its own `man' page on Unix) with which users can create wisdom, e.g. for a canonical set of sizes to store in the system wisdom file. *Note Wisdom Utilities::.  File: fftw3.info, Node: FFTW Reference, Next: Multi-threaded FFTW, Prev: Other Important Topics, Up: Top 4 FFTW Reference **************** This chapter provides a complete reference for all sequential (i.e., one-processor) FFTW functions. Parallel transforms are described in later chapters. * Menu: * Data Types and Files:: * Using Plans:: * Basic Interface:: * Advanced Interface:: * Guru Interface:: * New-array Execute Functions:: * Wisdom:: * What FFTW Really Computes::  File: fftw3.info, Node: Data Types and Files, Next: Using Plans, Prev: FFTW Reference, Up: FFTW Reference 4.1 Data Types and Files ======================== All programs using FFTW should include its header file: #include You must also link to the FFTW library. On Unix, this means adding `-lfftw3 -lm' at the _end_ of the link command. * Menu: * Complex numbers:: * Precision:: * Memory Allocation::  File: fftw3.info, Node: Complex numbers, Next: Precision, Prev: Data Types and Files, Up: Data Types and Files 4.1.1 Complex numbers --------------------- The default FFTW interface uses `double' precision for all floating-point numbers, and defines a `fftw_complex' type to hold complex numbers as: typedef double fftw_complex[2]; Here, the `[0]' element holds the real part and the `[1]' element holds the imaginary part. Alternatively, if you have a C compiler (such as `gcc') that supports the C99 revision of the ANSI C standard, you can use C's new native complex type (which is binary-compatible with the typedef above). In particular, if you `#include ' _before_ `', then `fftw_complex' is defined to be the native complex type and you can manipulate it with ordinary arithmetic (e.g. `x = y * (3+4*I)', where `x' and `y' are `fftw_complex' and `I' is the standard symbol for the imaginary unit); C++ has its own `complex' template class, defined in the standard `' header file. Reportedly, the C++ standards committee has recently agreed to mandate that the storage format used for this type be binary-compatible with the C99 type, i.e. an array `T[2]' with consecutive real `[0]' and imaginary `[1]' parts. (See report `http://www.open-std.org/jtc1/sc22/WG21/docs/papers/2002/n1388.pdf WG21/N1388'.) Although not part of the official standard as of this writing, the proposal stated that: "This solution has been tested with all current major implementations of the standard library and shown to be working." To the extent that this is true, if you have a variable `complex *x', you can pass it directly to FFTW via `reinterpret_cast(x)'.  File: fftw3.info, Node: Precision, Next: Memory Allocation, Prev: Complex numbers, Up: Data Types and Files 4.1.2 Precision --------------- You can install single and long-double precision versions of FFTW, which replace `double' with `float' and `long double', respectively (*note Installation and Customization::). To use these interfaces, you: * Link to the single/long-double libraries; on Unix, `-lfftw3f' or `-lfftw3l' instead of (or in addition to) `-lfftw3'. (You can link to the different-precision libraries simultaneously.) * Include the _same_ `' header file. * Replace all lowercase instances of `fftw_' with `fftwf_' or `fftwl_' for single or long-double precision, respectively. (`fftw_complex' becomes `fftwf_complex', `fftw_execute' becomes `fftwf_execute', etcetera.) * Uppercase names, i.e. names beginning with `FFTW_', remain the same. * Replace `double' with `float' or `long double' for subroutine parameters. Depending upon your compiler and/or hardware, `long double' may not be any more precise than `double' (or may not be supported at all, although it is standard in C99). We also support using the nonstandard `__float128' quadruple-precision type provided by recent versions of `gcc' on 32- and 64-bit x86 hardware (*note Installation and Customization::). To use this type, link with `-lfftw3q -lquadmath -lm' (the `libquadmath' library provided by `gcc' is needed for quadruple-precision trigonometric functions) and use `fftwq_' identifiers.  File: fftw3.info, Node: Memory Allocation, Prev: Precision, Up: Data Types and Files 4.1.3 Memory Allocation ----------------------- void *fftw_malloc(size_t n); void fftw_free(void *p); These are functions that behave identically to `malloc' and `free', except that they guarantee that the returned pointer obeys any special alignment restrictions imposed by any algorithm in FFTW (e.g. for SIMD acceleration). *Note SIMD alignment and fftw_malloc::. Data allocated by `fftw_malloc' _must_ be deallocated by `fftw_free' and not by the ordinary `free'. These routines simply call through to your operating system's `malloc' or, if necessary, its aligned equivalent (e.g. `memalign'), so you normally need not worry about any significant time or space overhead. You are _not required_ to use them to allocate your data, but we strongly recommend it. Note: in C++, just as with ordinary `malloc', you must typecast the output of `fftw_malloc' to whatever pointer type you are allocating. We also provide the following two convenience functions to allocate real and complex arrays with `n' elements, which are equivalent to `(double *) fftw_malloc(sizeof(double) * n)' and `(fftw_complex *) fftw_malloc(sizeof(fftw_complex) * n)', respectively: double *fftw_alloc_real(size_t n); fftw_complex *fftw_alloc_complex(size_t n); The equivalent functions in other precisions allocate arrays of `n' elements in that precision. e.g. `fftwf_alloc_real(n)' is equivalent to `(float *) fftwf_malloc(sizeof(float) * n)'.  File: fftw3.info, Node: Using Plans, Next: Basic Interface, Prev: Data Types and Files, Up: FFTW Reference 4.2 Using Plans =============== Plans for all transform types in FFTW are stored as type `fftw_plan' (an opaque pointer type), and are created by one of the various planning routines described in the following sections. An `fftw_plan' contains all information necessary to compute the transform, including the pointers to the input and output arrays. void fftw_execute(const fftw_plan plan); This executes the `plan', to compute the corresponding transform on the arrays for which it was planned (which must still exist). The plan is not modified, and `fftw_execute' can be called as many times as desired. To apply a given plan to a different array, you can use the new-array execute interface. *Note New-array Execute Functions::. `fftw_execute' (and equivalents) is the only function in FFTW guaranteed to be thread-safe; see *note Thread safety::. This function: void fftw_destroy_plan(fftw_plan plan); deallocates the `plan' and all its associated data. FFTW's planner saves some other persistent data, such as the accumulated wisdom and a list of algorithms available in the current configuration. If you want to deallocate all of that and reset FFTW to the pristine state it was in when you started your program, you can call: void fftw_cleanup(void); After calling `fftw_cleanup', all existing plans become undefined, and you should not attempt to execute them nor to destroy them. You can however create and execute/destroy new plans, in which case FFTW starts accumulating wisdom information again. `fftw_cleanup' does not deallocate your plans, however. To prevent memory leaks, you must still call `fftw_destroy_plan' before executing `fftw_cleanup'. Occasionally, it may useful to know FFTW's internal "cost" metric that it uses to compare plans to one another; this cost is proportional to an execution time of the plan, in undocumented units, if the plan was created with the `FFTW_MEASURE' or other timing-based options, or alternatively is a heuristic cost function for `FFTW_ESTIMATE' plans. (The cost values of measured and estimated plans are not comparable, being in different units. Also, costs from different FFTW versions or the same version compiled differently may not be in the same units. Plans created from wisdom have a cost of 0 since no timing measurement is performed for them. Finally, certain problems for which only one top-level algorithm was possible may have required no measurements of the cost of the whole plan, in which case `fftw_cost' will also return 0.) The cost metric for a given plan is returned by: double fftw_cost(const fftw_plan plan); The following two routines are provided purely for academic purposes (that is, for entertainment). void fftw_flops(const fftw_plan plan, double *add, double *mul, double *fma); Given a `plan', set `add', `mul', and `fma' to an exact count of the number of floating-point additions, multiplications, and fused multiply-add operations involved in the plan's execution. The total number of floating-point operations (flops) is `add + mul + 2*fma', or `add + mul + fma' if the hardware supports fused multiply-add instructions (although the number of FMA operations is only approximate because of compiler voodoo). (The number of operations should be an integer, but we use `double' to avoid overflowing `int' for large transforms; the arguments are of type `double' even for single and long-double precision versions of FFTW.) void fftw_fprint_plan(const fftw_plan plan, FILE *output_file); void fftw_print_plan(const fftw_plan plan); This outputs a "nerd-readable" representation of the `plan' to the given file or to `stdout', respectively.  File: fftw3.info, Node: Basic Interface, Next: Advanced Interface, Prev: Using Plans, Up: FFTW Reference 4.3 Basic Interface =================== Recall that the FFTW API is divided into three parts(1): the "basic interface" computes a single transform of contiguous data, the "advanced interface" computes transforms of multiple or strided arrays, and the "guru interface" supports the most general data layouts, multiplicities, and strides. This section describes the the basic interface, which we expect to satisfy the needs of most users. * Menu: * Complex DFTs:: * Planner Flags:: * Real-data DFTs:: * Real-data DFT Array Format:: * Real-to-Real Transforms:: * Real-to-Real Transform Kinds:: ---------- Footnotes ---------- (1) Gallia est omnis divisa in partes tres (Julius Caesar).  File: fftw3.info, Node: Complex DFTs, Next: Planner Flags, Prev: Basic Interface, Up: Basic Interface 4.3.1 Complex DFTs ------------------ fftw_plan fftw_plan_dft_1d(int n0, fftw_complex *in, fftw_complex *out, int sign, unsigned flags); fftw_plan fftw_plan_dft_2d(int n0, int n1, fftw_complex *in, fftw_complex *out, int sign, unsigned flags); fftw_plan fftw_plan_dft_3d(int n0, int n1, int n2, fftw_complex *in, fftw_complex *out, int sign, unsigned flags); fftw_plan fftw_plan_dft(int rank, const int *n, fftw_complex *in, fftw_complex *out, int sign, unsigned flags); Plan a complex input/output discrete Fourier transform (DFT) in zero or more dimensions, returning an `fftw_plan' (*note Using Plans::). Once you have created a plan for a certain transform type and parameters, then creating another plan of the same type and parameters, but for different arrays, is fast and shares constant data with the first plan (if it still exists). The planner returns `NULL' if the plan cannot be created. In the standard FFTW distribution, the basic interface is guaranteed to return a non-`NULL' plan. A plan may be `NULL', however, if you are using a customized FFTW configuration supporting a restricted set of transforms. Arguments ......... * `rank' is the rank of the transform (it should be the size of the array `*n'), and can be any non-negative integer. (*Note Complex Multi-Dimensional DFTs::, for the definition of "rank".) The `_1d', `_2d', and `_3d' planners correspond to a `rank' of `1', `2', and `3', respectively. The rank may be zero, which is equivalent to a rank-1 transform of size 1, i.e. a copy of one number from input to output. * `n0', `n1', `n2', or `n[0..rank-1]' (as appropriate for each routine) specify the size of the transform dimensions. They can be any positive integer. - Multi-dimensional arrays are stored in row-major order with dimensions: `n0' x `n1'; or `n0' x `n1' x `n2'; or `n[0]' x `n[1]' x ... x `n[rank-1]'. *Note Multi-dimensional Array Format::. - FFTW is best at handling sizes of the form 2^a 3^b 5^c 7^d 11^e 13^f, where e+f is either 0 or 1, and the other exponents are arbitrary. Other sizes are computed by means of a slow, general-purpose algorithm (which nevertheless retains O(n log n) performance even for prime sizes). It is possible to customize FFTW for different array sizes; see *note Installation and Customization::. Transforms whose sizes are powers of 2 are especially fast. * `in' and `out' point to the input and output arrays of the transform, which may be the same (yielding an in-place transform). These arrays are overwritten during planning, unless `FFTW_ESTIMATE' is used in the flags. (The arrays need not be initialized, but they must be allocated.) If `in == out', the transform is "in-place" and the input array is overwritten. If `in != out', the two arrays must not overlap (but FFTW does not check for this condition). * `sign' is the sign of the exponent in the formula that defines the Fourier transform. It can be -1 (= `FFTW_FORWARD') or +1 (= `FFTW_BACKWARD'). * `flags' is a bitwise OR (`|') of zero or more planner flags, as defined in *note Planner Flags::. FFTW computes an unnormalized transform: computing a forward followed by a backward transform (or vice versa) will result in the original data multiplied by the size of the transform (the product of the dimensions). For more information, see *note What FFTW Really Computes::.  File: fftw3.info, Node: Planner Flags, Next: Real-data DFTs, Prev: Complex DFTs, Up: Basic Interface 4.3.2 Planner Flags ------------------- All of the planner routines in FFTW accept an integer `flags' argument, which is a bitwise OR (`|') of zero or more of the flag constants defined below. These flags control the rigor (and time) of the planning process, and can also impose (or lift) restrictions on the type of transform algorithm that is employed. _Important:_ the planner overwrites the input array during planning unless a saved plan (*note Wisdom::) is available for that problem, so you should initialize your input data after creating the plan. The only exceptions to this are the `FFTW_ESTIMATE' and `FFTW_WISDOM_ONLY' flags, as mentioned below. In all cases, if wisdom is available for the given problem that was created with equal-or-greater planning rigor, then the more rigorous wisdom is used. For example, in `FFTW_ESTIMATE' mode any available wisdom is used, whereas in `FFTW_PATIENT' mode only wisdom created in patient or exhaustive mode can be used. *Note Words of Wisdom-Saving Plans::. Planning-rigor flags .................... * `FFTW_ESTIMATE' specifies that, instead of actual measurements of different algorithms, a simple heuristic is used to pick a (probably sub-optimal) plan quickly. With this flag, the input/output arrays are not overwritten during planning. * `FFTW_MEASURE' tells FFTW to find an optimized plan by actually _computing_ several FFTs and measuring their execution time. Depending on your machine, this can take some time (often a few seconds). `FFTW_MEASURE' is the default planning option. * `FFTW_PATIENT' is like `FFTW_MEASURE', but considers a wider range of algorithms and often produces a "more optimal" plan (especially for large transforms), but at the expense of several times longer planning time (especially for large transforms). * `FFTW_EXHAUSTIVE' is like `FFTW_PATIENT', but considers an even wider range of algorithms, including many that we think are unlikely to be fast, to produce the most optimal plan but with a substantially increased planning time. * `FFTW_WISDOM_ONLY' is a special planning mode in which the plan is only created if wisdom is available for the given problem, and otherwise a `NULL' plan is returned. This can be combined with other flags, e.g. `FFTW_WISDOM_ONLY | FFTW_PATIENT' creates a plan only if wisdom is available that was created in `FFTW_PATIENT' or `FFTW_EXHAUSTIVE' mode. The `FFTW_WISDOM_ONLY' flag is intended for users who need to detect whether wisdom is available; for example, if wisdom is not available one may wish to allocate new arrays for planning so that user data is not overwritten. Algorithm-restriction flags ........................... * `FFTW_DESTROY_INPUT' specifies that an out-of-place transform is allowed to _overwrite its input_ array with arbitrary data; this can sometimes allow more efficient algorithms to be employed. * `FFTW_PRESERVE_INPUT' specifies that an out-of-place transform must _not change its input_ array. This is ordinarily the _default_, except for c2r and hc2r (i.e. complex-to-real) transforms for which `FFTW_DESTROY_INPUT' is the default. In the latter cases, passing `FFTW_PRESERVE_INPUT' will attempt to use algorithms that do not destroy the input, at the expense of worse performance; for multi-dimensional c2r transforms, however, no input-preserving algorithms are implemented and the planner will return `NULL' if one is requested. * `FFTW_UNALIGNED' specifies that the algorithm may not impose any unusual alignment requirements on the input/output arrays (i.e. no SIMD may be used). This flag is normally _not necessary_, since the planner automatically detects misaligned arrays. The only use for this flag is if you want to use the new-array execute interface to execute a given plan on a different array that may not be aligned like the original. (Using `fftw_malloc' makes this flag unnecessary even then.) Limiting planning time ...................... extern void fftw_set_timelimit(double seconds); This function instructs FFTW to spend at most `seconds' seconds (approximately) in the planner. If `seconds == FFTW_NO_TIMELIMIT' (the default value, which is negative), then planning time is unbounded. Otherwise, FFTW plans with a progressively wider range of algorithms until the the given time limit is reached or the given range of algorithms is explored, returning the best available plan. For example, specifying `FFTW_PATIENT' first plans in `FFTW_ESTIMATE' mode, then in `FFTW_MEASURE' mode, then finally (time permitting) in `FFTW_PATIENT'. If `FFTW_EXHAUSTIVE' is specified instead, the planner will further progress to `FFTW_EXHAUSTIVE' mode. Note that the `seconds' argument specifies only a rough limit; in practice, the planner may use somewhat more time if the time limit is reached when the planner is in the middle of an operation that cannot be interrupted. At the very least, the planner will complete planning in `FFTW_ESTIMATE' mode (which is thus equivalent to a time limit of 0).  File: fftw3.info, Node: Real-data DFTs, Next: Real-data DFT Array Format, Prev: Planner Flags, Up: Basic Interface 4.3.3 Real-data DFTs -------------------- fftw_plan fftw_plan_dft_r2c_1d(int n0, double *in, fftw_complex *out, unsigned flags); fftw_plan fftw_plan_dft_r2c_2d(int n0, int n1, double *in, fftw_complex *out, unsigned flags); fftw_plan fftw_plan_dft_r2c_3d(int n0, int n1, int n2, double *in, fftw_complex *out, unsigned flags); fftw_plan fftw_plan_dft_r2c(int rank, const int *n, double *in, fftw_complex *out, unsigned flags); Plan a real-input/complex-output discrete Fourier transform (DFT) in zero or more dimensions, returning an `fftw_plan' (*note Using Plans::). Once you have created a plan for a certain transform type and parameters, then creating another plan of the same type and parameters, but for different arrays, is fast and shares constant data with the first plan (if it still exists). The planner returns `NULL' if the plan cannot be created. A non-`NULL' plan is always returned by the basic interface unless you are using a customized FFTW configuration supporting a restricted set of transforms, or if you use the `FFTW_PRESERVE_INPUT' flag with a multi-dimensional out-of-place c2r transform (see below). Arguments ......... * `rank' is the rank of the transform (it should be the size of the array `*n'), and can be any non-negative integer. (*Note Complex Multi-Dimensional DFTs::, for the definition of "rank".) The `_1d', `_2d', and `_3d' planners correspond to a `rank' of `1', `2', and `3', respectively. The rank may be zero, which is equivalent to a rank-1 transform of size 1, i.e. a copy of one real number (with zero imaginary part) from input to output. * `n0', `n1', `n2', or `n[0..rank-1]', (as appropriate for each routine) specify the size of the transform dimensions. They can be any positive integer. This is different in general from the _physical_ array dimensions, which are described in *note Real-data DFT Array Format::. - FFTW is best at handling sizes of the form 2^a 3^b 5^c 7^d 11^e 13^f, where e+f is either 0 or 1, and the other exponents are arbitrary. Other sizes are computed by means of a slow, general-purpose algorithm (which nevertheless retains O(n log n) performance even for prime sizes). (It is possible to customize FFTW for different array sizes; see *note Installation and Customization::.) Transforms whose sizes are powers of 2 are especially fast, and it is generally beneficial for the _last_ dimension of an r2c/c2r transform to be _even_. * `in' and `out' point to the input and output arrays of the transform, which may be the same (yielding an in-place transform). These arrays are overwritten during planning, unless `FFTW_ESTIMATE' is used in the flags. (The arrays need not be initialized, but they must be allocated.) For an in-place transform, it is important to remember that the real array will require padding, described in *note Real-data DFT Array Format::. * `flags' is a bitwise OR (`|') of zero or more planner flags, as defined in *note Planner Flags::. The inverse transforms, taking complex input (storing the non-redundant half of a logically Hermitian array) to real output, are given by: fftw_plan fftw_plan_dft_c2r_1d(int n0, fftw_complex *in, double *out, unsigned flags); fftw_plan fftw_plan_dft_c2r_2d(int n0, int n1, fftw_complex *in, double *out, unsigned flags); fftw_plan fftw_plan_dft_c2r_3d(int n0, int n1, int n2, fftw_complex *in, double *out, unsigned flags); fftw_plan fftw_plan_dft_c2r(int rank, const int *n, fftw_complex *in, double *out, unsigned flags); The arguments are the same as for the r2c transforms, except that the input and output data formats are reversed. FFTW computes an unnormalized transform: computing an r2c followed by a c2r transform (or vice versa) will result in the original data multiplied by the size of the transform (the product of the logical dimensions). An r2c transform produces the same output as a `FFTW_FORWARD' complex DFT of the same input, and a c2r transform is correspondingly equivalent to `FFTW_BACKWARD'. For more information, see *note What FFTW Really Computes::.  File: fftw3.info, Node: Real-data DFT Array Format, Next: Real-to-Real Transforms, Prev: Real-data DFTs, Up: Basic Interface 4.3.4 Real-data DFT Array Format -------------------------------- The output of a DFT of real data (r2c) contains symmetries that, in principle, make half of the outputs redundant (*note What FFTW Really Computes::). (Similarly for the input of an inverse c2r transform.) In practice, it is not possible to entirely realize these savings in an efficient and understandable format that generalizes to multi-dimensional transforms. Instead, the output of the r2c transforms is _slightly_ over half of the output of the corresponding complex transform. We do not "pack" the data in any way, but store it as an ordinary array of `fftw_complex' values. In fact, this data is simply a subsection of what would be the array in the corresponding complex transform. Specifically, for a real transform of d (= `rank') dimensions n[0] x n[1] x n[2] x ... x n[d-1] , the complex data is an n[0] x n[1] x n[2] x ... x (n[d-1]/2 + 1) array of `fftw_complex' values in row-major order (with the division rounded down). That is, we only store the _lower_ half (non-negative frequencies), plus one element, of the last dimension of the data from the ordinary complex transform. (We could have instead taken half of any other dimension, but implementation turns out to be simpler if the last, contiguous, dimension is used.) For an out-of-place transform, the real data is simply an array with physical dimensions n[0] x n[1] x n[2] x ... x n[d-1] in row-major order. For an in-place transform, some complications arise since the complex data is slightly larger than the real data. In this case, the final dimension of the real data must be _padded_ with extra values to accommodate the size of the complex data--two extra if the last dimension is even and one if it is odd. That is, the last dimension of the real data must physically contain 2 * (n[d-1]/2+1) `double' values (exactly enough to hold the complex data). This physical array size does not, however, change the _logical_ array size--only n[d-1] values are actually stored in the last dimension, and n[d-1] is the last dimension passed to the planner.  File: fftw3.info, Node: Real-to-Real Transforms, Next: Real-to-Real Transform Kinds, Prev: Real-data DFT Array Format, Up: Basic Interface 4.3.5 Real-to-Real Transforms ----------------------------- fftw_plan fftw_plan_r2r_1d(int n, double *in, double *out, fftw_r2r_kind kind, unsigned flags); fftw_plan fftw_plan_r2r_2d(int n0, int n1, double *in, double *out, fftw_r2r_kind kind0, fftw_r2r_kind kind1, unsigned flags); fftw_plan fftw_plan_r2r_3d(int n0, int n1, int n2, double *in, double *out, fftw_r2r_kind kind0, fftw_r2r_kind kind1, fftw_r2r_kind kind2, unsigned flags); fftw_plan fftw_plan_r2r(int rank, const int *n, double *in, double *out, const fftw_r2r_kind *kind, unsigned flags); Plan a real input/output (r2r) transform of various kinds in zero or more dimensions, returning an `fftw_plan' (*note Using Plans::). Once you have created a plan for a certain transform type and parameters, then creating another plan of the same type and parameters, but for different arrays, is fast and shares constant data with the first plan (if it still exists). The planner returns `NULL' if the plan cannot be created. A non-`NULL' plan is always returned by the basic interface unless you are using a customized FFTW configuration supporting a restricted set of transforms, or for size-1 `FFTW_REDFT00' kinds (which are not defined). Arguments ......... * `rank' is the dimensionality of the transform (it should be the size of the arrays `*n' and `*kind'), and can be any non-negative integer. The `_1d', `_2d', and `_3d' planners correspond to a `rank' of `1', `2', and `3', respectively. A `rank' of zero is equivalent to a copy of one number from input to output. * `n', or `n0'/`n1'/`n2', or `n[rank]', respectively, gives the (physical) size of the transform dimensions. They can be any positive integer. - Multi-dimensional arrays are stored in row-major order with dimensions: `n0' x `n1'; or `n0' x `n1' x `n2'; or `n[0]' x `n[1]' x ... x `n[rank-1]'. *Note Multi-dimensional Array Format::. - FFTW is generally best at handling sizes of the form 2^a 3^b 5^c 7^d 11^e 13^f, where e+f is either 0 or 1, and the other exponents are arbitrary. Other sizes are computed by means of a slow, general-purpose algorithm (which nevertheless retains O(n log n) performance even for prime sizes). (It is possible to customize FFTW for different array sizes; see *note Installation and Customization::.) Transforms whose sizes are powers of 2 are especially fast. - For a `REDFT00' or `RODFT00' transform kind in a dimension of size n, it is n-1 or n+1, respectively, that should be factorizable in the above form. * `in' and `out' point to the input and output arrays of the transform, which may be the same (yielding an in-place transform). These arrays are overwritten during planning, unless `FFTW_ESTIMATE' is used in the flags. (The arrays need not be initialized, but they must be allocated.) * `kind', or `kind0'/`kind1'/`kind2', or `kind[rank]', is the kind of r2r transform used for the corresponding dimension. The valid kind constants are described in *note Real-to-Real Transform Kinds::. In a multi-dimensional transform, what is computed is the separable product formed by taking each transform kind along the corresponding dimension, one dimension after another. * `flags' is a bitwise OR (`|') of zero or more planner flags, as defined in *note Planner Flags::.  File: fftw3.info, Node: Real-to-Real Transform Kinds, Prev: Real-to-Real Transforms, Up: Basic Interface 4.3.6 Real-to-Real Transform Kinds ---------------------------------- FFTW currently supports 11 different r2r transform kinds, specified by one of the constants below. For the precise definitions of these transforms, see *note What FFTW Really Computes::. For a more colloquial introduction to these transform kinds, see *note More DFTs of Real Data::. For dimension of size `n', there is a corresponding "logical" dimension `N' that determines the normalization (and the optimal factorization); the formula for `N' is given for each kind below. Also, with each transform kind is listed its corrsponding inverse transform. FFTW computes unnormalized transforms: a transform followed by its inverse will result in the original data multiplied by `N' (or the product of the `N''s for each dimension, in multi-dimensions). * `FFTW_R2HC' computes a real-input DFT with output in "halfcomplex" format, i.e. real and imaginary parts for a transform of size `n' stored as: r0, r1, r2, r(n/2), i((n+1)/2-1), ..., i2, i1 (Logical `N=n', inverse is `FFTW_HC2R'.) * `FFTW_HC2R' computes the reverse of `FFTW_R2HC', above. (Logical `N=n', inverse is `FFTW_R2HC'.) * `FFTW_DHT' computes a discrete Hartley transform. (Logical `N=n', inverse is `FFTW_DHT'.) * `FFTW_REDFT00' computes an REDFT00 transform, i.e. a DCT-I. (Logical `N=2*(n-1)', inverse is `FFTW_REDFT00'.) * `FFTW_REDFT10' computes an REDFT10 transform, i.e. a DCT-II (sometimes called "the" DCT). (Logical `N=2*n', inverse is `FFTW_REDFT01'.) * `FFTW_REDFT01' computes an REDFT01 transform, i.e. a DCT-III (sometimes called "the" IDCT, being the inverse of DCT-II). (Logical `N=2*n', inverse is `FFTW_REDFT=10'.) * `FFTW_REDFT11' computes an REDFT11 transform, i.e. a DCT-IV. (Logical `N=2*n', inverse is `FFTW_REDFT11'.) * `FFTW_RODFT00' computes an RODFT00 transform, i.e. a DST-I. (Logical `N=2*(n+1)', inverse is `FFTW_RODFT00'.) * `FFTW_RODFT10' computes an RODFT10 transform, i.e. a DST-II. (Logical `N=2*n', inverse is `FFTW_RODFT01'.) * `FFTW_RODFT01' computes an RODFT01 transform, i.e. a DST-III. (Logical `N=2*n', inverse is `FFTW_RODFT=10'.) * `FFTW_RODFT11' computes an RODFT11 transform, i.e. a DST-IV. (Logical `N=2*n', inverse is `FFTW_RODFT11'.)  File: fftw3.info, Node: Advanced Interface, Next: Guru Interface, Prev: Basic Interface, Up: FFTW Reference 4.4 Advanced Interface ====================== FFTW's "advanced" interface supplements the basic interface with four new planner routines, providing a new level of flexibility: you can plan a transform of multiple arrays simultaneously, operate on non-contiguous (strided) data, and transform a subset of a larger multi-dimensional array. Other than these additional features, the planner operates in the same fashion as in the basic interface, and the resulting `fftw_plan' is used in the same way (*note Using Plans::). * Menu: * Advanced Complex DFTs:: * Advanced Real-data DFTs:: * Advanced Real-to-real Transforms::  File: fftw3.info, Node: Advanced Complex DFTs, Next: Advanced Real-data DFTs, Prev: Advanced Interface, Up: Advanced Interface 4.4.1 Advanced Complex DFTs --------------------------- fftw_plan fftw_plan_many_dft(int rank, const int *n, int howmany, fftw_complex *in, const int *inembed, int istride, int idist, fftw_complex *out, const int *onembed, int ostride, int odist, int sign, unsigned flags); This routine plans multiple multidimensional complex DFTs, and it extends the `fftw_plan_dft' routine (*note Complex DFTs::) to compute `howmany' transforms, each having rank `rank' and size `n'. In addition, the transform data need not be contiguous, but it may be laid out in memory with an arbitrary stride. To account for these possibilities, `fftw_plan_many_dft' adds the new parameters `howmany', {`i',`o'}`nembed', {`i',`o'}`stride', and {`i',`o'}`dist'. The FFTW basic interface (*note Complex DFTs::) provides routines specialized for ranks 1, 2, and 3, but the advanced interface handles only the general-rank case. `howmany' is the number of transforms to compute. The resulting plan computes `howmany' transforms, where the input of the `k'-th transform is at location `in+k*idist' (in C pointer arithmetic), and its output is at location `out+k*odist'. Plans obtained in this way can often be faster than calling FFTW multiple times for the individual transforms. The basic `fftw_plan_dft' interface corresponds to `howmany=1' (in which case the `dist' parameters are ignored). Each of the `howmany' transforms has rank `rank' and size `n', as in the basic interface. In addition, the advanced interface allows the input and output arrays of each transform to be row-major subarrays of larger rank-`rank' arrays, described by `inembed' and `onembed' parameters, respectively. {`i',`o'}`nembed' must be arrays of length `rank', and `n' should be elementwise less than or equal to {`i',`o'}`nembed'. Passing `NULL' for an `nembed' parameter is equivalent to passing `n' (i.e. same physical and logical dimensions, as in the basic interface.) The `stride' parameters indicate that the `j'-th element of the input or output arrays is located at `j*istride' or `j*ostride', respectively. (For a multi-dimensional array, `j' is the ordinary row-major index.) When combined with the `k'-th transform in a `howmany' loop, from above, this means that the (`j',`k')-th element is at `j*stride+k*dist'. (The basic `fftw_plan_dft' interface corresponds to a stride of 1.) For in-place transforms, the input and output `stride' and `dist' parameters should be the same; otherwise, the planner may return `NULL'. Arrays `n', `inembed', and `onembed' are not used after this function returns. You can safely free or reuse them. *Examples*: One transform of one 5 by 6 array contiguous in memory: int rank = 2; int n[] = {5, 6}; int howmany = 1; int idist = odist = 0; /* unused because howmany = 1 */ int istride = ostride = 1; /* array is contiguous in memory */ int *inembed = n, *onembed = n; Transform of three 5 by 6 arrays, each contiguous in memory, stored in memory one after another: int rank = 2; int n[] = {5, 6}; int howmany = 3; int idist = odist = n[0]*n[1]; /* = 30, the distance in memory between the first element of the first array and the first element of the second array */ int istride = ostride = 1; /* array is contiguous in memory */ int *inembed = n, *onembed = n; Transform each column of a 2d array with 10 rows and 3 columns: int rank = 1; /* not 2: we are computing 1d transforms */ int n[] = {10}; /* 1d transforms of length 10 */ int howmany = 3; int idist = odist = 1; int istride = ostride = 3; /* distance between two elements in the same column */ int *inembed = n, *onembed = n;  File: fftw3.info, Node: Advanced Real-data DFTs, Next: Advanced Real-to-real Transforms, Prev: Advanced Complex DFTs, Up: Advanced Interface 4.4.2 Advanced Real-data DFTs ----------------------------- fftw_plan fftw_plan_many_dft_r2c(int rank, const int *n, int howmany, double *in, const int *inembed, int istride, int idist, fftw_complex *out, const int *onembed, int ostride, int odist, unsigned flags); fftw_plan fftw_plan_many_dft_c2r(int rank, const int *n, int howmany, fftw_complex *in, const int *inembed, int istride, int idist, double *out, const int *onembed, int ostride, int odist, unsigned flags); Like `fftw_plan_many_dft', these two functions add `howmany', `nembed', `stride', and `dist' parameters to the `fftw_plan_dft_r2c' and `fftw_plan_dft_c2r' functions, but otherwise behave the same as the basic interface. The interpretation of `howmany', `stride', and `dist' are the same as for `fftw_plan_many_dft', above. Note that the `stride' and `dist' for the real array are in units of `double', and for the complex array are in units of `fftw_complex'. If an `nembed' parameter is `NULL', it is interpreted as what it would be in the basic interface, as described in *note Real-data DFT Array Format::. That is, for the complex array the size is assumed to be the same as `n', but with the last dimension cut roughly in half. For the real array, the size is assumed to be `n' if the transform is out-of-place, or `n' with the last dimension "padded" if the transform is in-place. If an `nembed' parameter is non-`NULL', it is interpreted as the physical size of the corresponding array, in row-major order, just as for `fftw_plan_many_dft'. In this case, each dimension of `nembed' should be `>=' what it would be in the basic interface (e.g. the halved or padded `n'). Arrays `n', `inembed', and `onembed' are not used after this function returns. You can safely free or reuse them.  File: fftw3.info, Node: Advanced Real-to-real Transforms, Prev: Advanced Real-data DFTs, Up: Advanced Interface 4.4.3 Advanced Real-to-real Transforms -------------------------------------- fftw_plan fftw_plan_many_r2r(int rank, const int *n, int howmany, double *in, const int *inembed, int istride, int idist, double *out, const int *onembed, int ostride, int odist, const fftw_r2r_kind *kind, unsigned flags); Like `fftw_plan_many_dft', this functions adds `howmany', `nembed', `stride', and `dist' parameters to the `fftw_plan_r2r' function, but otherwise behave the same as the basic interface. The interpretation of those additional parameters are the same as for `fftw_plan_many_dft'. (Of course, the `stride' and `dist' parameters are now in units of `double', not `fftw_complex'.) Arrays `n', `inembed', `onembed', and `kind' are not used after this function returns. You can safely free or reuse them.  File: fftw3.info, Node: Guru Interface, Next: New-array Execute Functions, Prev: Advanced Interface, Up: FFTW Reference 4.5 Guru Interface ================== The "guru" interface to FFTW is intended to expose as much as possible of the flexibility in the underlying FFTW architecture. It allows one to compute multi-dimensional "vectors" (loops) of multi-dimensional transforms, where each vector/transform dimension has an independent size and stride. One can also use more general complex-number formats, e.g. separate real and imaginary arrays. For those users who require the flexibility of the guru interface, it is important that they pay special attention to the documentation lest they shoot themselves in the foot. * Menu: * Interleaved and split arrays:: * Guru vector and transform sizes:: * Guru Complex DFTs:: * Guru Real-data DFTs:: * Guru Real-to-real Transforms:: * 64-bit Guru Interface::  File: fftw3.info, Node: Interleaved and split arrays, Next: Guru vector and transform sizes, Prev: Guru Interface, Up: Guru Interface 4.5.1 Interleaved and split arrays ---------------------------------- The guru interface supports two representations of complex numbers, which we call the interleaved and the split format. The "interleaved" format is the same one used by the basic and advanced interfaces, and it is documented in *note Complex numbers::. In the interleaved format, you provide pointers to the real part of a complex number, and the imaginary part understood to be stored in the next memory location. The "split" format allows separate pointers to the real and imaginary parts of a complex array. Technically, the interleaved format is redundant, because you can always express an interleaved array in terms of a split array with appropriate pointers and strides. On the other hand, the interleaved format is simpler to use, and it is common in practice. Hence, FFTW supports it as a special case.  File: fftw3.info, Node: Guru vector and transform sizes, Next: Guru Complex DFTs, Prev: Interleaved and split arrays, Up: Guru Interface 4.5.2 Guru vector and transform sizes ------------------------------------- The guru interface introduces one basic new data structure, `fftw_iodim', that is used to specify sizes and strides for multi-dimensional transforms and vectors: typedef struct { int n; int is; int os; } fftw_iodim; Here, `n' is the size of the dimension, and `is' and `os' are the strides of that dimension for the input and output arrays. (The stride is the separation of consecutive elements along this dimension.) The meaning of the stride parameter depends on the type of the array that the stride refers to. _If the array is interleaved complex, strides are expressed in units of complex numbers (`fftw_complex'). If the array is split complex or real, strides are expressed in units of real numbers (`double')._ This convention is consistent with the usual pointer arithmetic in the C language. An interleaved array is denoted by a pointer `p' to `fftw_complex', so that `p+1' points to the next complex number. Split arrays are denoted by pointers to `double', in which case pointer arithmetic operates in units of `sizeof(double)'. The guru planner interfaces all take a (`rank', `dims[rank]') pair describing the transform size, and a (`howmany_rank', `howmany_dims[howmany_rank]') pair describing the "vector" size (a multi-dimensional loop of transforms to perform), where `dims' and `howmany_dims' are arrays of `fftw_iodim'. For example, the `howmany' parameter in the advanced complex-DFT interface corresponds to `howmany_rank' = 1, `howmany_dims[0].n' = `howmany', `howmany_dims[0].is' = `idist', and `howmany_dims[0].os' = `odist'. (To compute a single transform, you can just use `howmany_rank' = 0.) A row-major multidimensional array with dimensions `n[rank]' (*note Row-major Format::) corresponds to `dims[i].n' = `n[i]' and the recurrence `dims[i].is' = `n[i+1] * dims[i+1].is' (similarly for `os'). The stride of the last (`i=rank-1') dimension is the overall stride of the array. e.g. to be equivalent to the advanced complex-DFT interface, you would have `dims[rank-1].is' = `istride' and `dims[rank-1].os' = `ostride'. In general, we only guarantee FFTW to return a non-`NULL' plan if the vector and transform dimensions correspond to a set of distinct indices, and for in-place transforms the input/output strides should be the same.  File: fftw3.info, Node: Guru Complex DFTs, Next: Guru Real-data DFTs, Prev: Guru vector and transform sizes, Up: Guru Interface 4.5.3 Guru Complex DFTs ----------------------- fftw_plan fftw_plan_guru_dft( int rank, const fftw_iodim *dims, int howmany_rank, const fftw_iodim *howmany_dims, fftw_complex *in, fftw_complex *out, int sign, unsigned flags); fftw_plan fftw_plan_guru_split_dft( int rank, const fftw_iodim *dims, int howmany_rank, const fftw_iodim *howmany_dims, double *ri, double *ii, double *ro, double *io, unsigned flags); These two functions plan a complex-data, multi-dimensional DFT for the interleaved and split format, respectively. Transform dimensions are given by (`rank', `dims') over a multi-dimensional vector (loop) of dimensions (`howmany_rank', `howmany_dims'). `dims' and `howmany_dims' should point to `fftw_iodim' arrays of length `rank' and `howmany_rank', respectively. `flags' is a bitwise OR (`|') of zero or more planner flags, as defined in *note Planner Flags::. In the `fftw_plan_guru_dft' function, the pointers `in' and `out' point to the interleaved input and output arrays, respectively. The sign can be either -1 (= `FFTW_FORWARD') or +1 (= `FFTW_BACKWARD'). If the pointers are equal, the transform is in-place. In the `fftw_plan_guru_split_dft' function, `ri' and `ii' point to the real and imaginary input arrays, and `ro' and `io' point to the real and imaginary output arrays. The input and output pointers may be the same, indicating an in-place transform. For example, for `fftw_complex' pointers `in' and `out', the corresponding parameters are: ri = (double *) in; ii = (double *) in + 1; ro = (double *) out; io = (double *) out + 1; Because `fftw_plan_guru_split_dft' accepts split arrays, strides are expressed in units of `double'. For a contiguous `fftw_complex' array, the overall stride of the transform should be 2, the distance between consecutive real parts or between consecutive imaginary parts; see *note Guru vector and transform sizes::. Note that the dimension strides are applied equally to the real and imaginary parts; real and imaginary arrays with different strides are not supported. There is no `sign' parameter in `fftw_plan_guru_split_dft'. This function always plans for an `FFTW_FORWARD' transform. To plan for an `FFTW_BACKWARD' transform, you can exploit the identity that the backwards DFT is equal to the forwards DFT with the real and imaginary parts swapped. For example, in the case of the `fftw_complex' arrays above, the `FFTW_BACKWARD' transform is computed by the parameters: ri = (double *) in + 1; ii = (double *) in; ro = (double *) out + 1; io = (double *) out;  File: fftw3.info, Node: Guru Real-data DFTs, Next: Guru Real-to-real Transforms, Prev: Guru Complex DFTs, Up: Guru Interface 4.5.4 Guru Real-data DFTs ------------------------- fftw_plan fftw_plan_guru_dft_r2c( int rank, const fftw_iodim *dims, int howmany_rank, const fftw_iodim *howmany_dims, double *in, fftw_complex *out, unsigned flags); fftw_plan fftw_plan_guru_split_dft_r2c( int rank, const fftw_iodim *dims, int howmany_rank, const fftw_iodim *howmany_dims, double *in, double *ro, double *io, unsigned flags); fftw_plan fftw_plan_guru_dft_c2r( int rank, const fftw_iodim *dims, int howmany_rank, const fftw_iodim *howmany_dims, fftw_complex *in, double *out, unsigned flags); fftw_plan fftw_plan_guru_split_dft_c2r( int rank, const fftw_iodim *dims, int howmany_rank, const fftw_iodim *howmany_dims, double *ri, double *ii, double *out, unsigned flags); Plan a real-input (r2c) or real-output (c2r), multi-dimensional DFT with transform dimensions given by (`rank', `dims') over a multi-dimensional vector (loop) of dimensions (`howmany_rank', `howmany_dims'). `dims' and `howmany_dims' should point to `fftw_iodim' arrays of length `rank' and `howmany_rank', respectively. As for the basic and advanced interfaces, an r2c transform is `FFTW_FORWARD' and a c2r transform is `FFTW_BACKWARD'. The _last_ dimension of `dims' is interpreted specially: that dimension of the real array has size `dims[rank-1].n', but that dimension of the complex array has size `dims[rank-1].n/2+1' (division rounded down). The strides, on the other hand, are taken to be exactly as specified. It is up to the user to specify the strides appropriately for the peculiar dimensions of the data, and we do not guarantee that the planner will succeed (return non-`NULL') for any dimensions other than those described in *note Real-data DFT Array Format:: and generalized in *note Advanced Real-data DFTs::. (That is, for an in-place transform, each individual dimension should be able to operate in place.) `in' and `out' point to the input and output arrays for r2c and c2r transforms, respectively. For split arrays, `ri' and `ii' point to the real and imaginary input arrays for a c2r transform, and `ro' and `io' point to the real and imaginary output arrays for an r2c transform. `in' and `ro' or `ri' and `out' may be the same, indicating an in-place transform. (In-place transforms where `in' and `io' or `ii' and `out' are the same are not currently supported.) `flags' is a bitwise OR (`|') of zero or more planner flags, as defined in *note Planner Flags::. In-place transforms of rank greater than 1 are currently only supported for interleaved arrays. For split arrays, the planner will return `NULL'.  File: fftw3.info, Node: Guru Real-to-real Transforms, Next: 64-bit Guru Interface, Prev: Guru Real-data DFTs, Up: Guru Interface 4.5.5 Guru Real-to-real Transforms ---------------------------------- fftw_plan fftw_plan_guru_r2r(int rank, const fftw_iodim *dims, int howmany_rank, const fftw_iodim *howmany_dims, double *in, double *out, const fftw_r2r_kind *kind, unsigned flags); Plan a real-to-real (r2r) multi-dimensional `FFTW_FORWARD' transform with transform dimensions given by (`rank', `dims') over a multi-dimensional vector (loop) of dimensions (`howmany_rank', `howmany_dims'). `dims' and `howmany_dims' should point to `fftw_iodim' arrays of length `rank' and `howmany_rank', respectively. The transform kind of each dimension is given by the `kind' parameter, which should point to an array of length `rank'. Valid `fftw_r2r_kind' constants are given in *note Real-to-Real Transform Kinds::. `in' and `out' point to the real input and output arrays; they may be the same, indicating an in-place transform. `flags' is a bitwise OR (`|') of zero or more planner flags, as defined in *note Planner Flags::.  File: fftw3.info, Node: 64-bit Guru Interface, Prev: Guru Real-to-real Transforms, Up: Guru Interface 4.5.6 64-bit Guru Interface --------------------------- When compiled in 64-bit mode on a 64-bit architecture (where addresses are 64 bits wide), FFTW uses 64-bit quantities internally for all transform sizes, strides, and so on--you don't have to do anything special to exploit this. However, in the ordinary FFTW interfaces, you specify the transform size by an `int' quantity, which is normally only 32 bits wide. This means that, even though FFTW is using 64-bit sizes internally, you cannot specify a single transform dimension larger than 2^31-1 numbers. We expect that few users will require transforms larger than this, but, for those who do, we provide a 64-bit version of the guru interface in which all sizes are specified as integers of type `ptrdiff_t' instead of `int'. (`ptrdiff_t' is a signed integer type defined by the C standard to be wide enough to represent address differences, and thus must be at least 64 bits wide on a 64-bit machine.) We stress that there is _no performance advantage_ to using this interface--the same internal FFTW code is employed regardless--and it is only necessary if you want to specify very large transform sizes. In particular, the 64-bit guru interface is a set of planner routines that are exactly the same as the guru planner routines, except that they are named with `guru64' instead of `guru' and they take arguments of type `fftw_iodim64' instead of `fftw_iodim'. For example, instead of `fftw_plan_guru_dft', we have `fftw_plan_guru64_dft'. fftw_plan fftw_plan_guru64_dft( int rank, const fftw_iodim64 *dims, int howmany_rank, const fftw_iodim64 *howmany_dims, fftw_complex *in, fftw_complex *out, int sign, unsigned flags); The `fftw_iodim64' type is similar to `fftw_iodim', with the same interpretation, except that it uses type `ptrdiff_t' instead of type `int'. typedef struct { ptrdiff_t n; ptrdiff_t is; ptrdiff_t os; } fftw_iodim64; Every other `fftw_plan_guru' function also has a `fftw_plan_guru64' equivalent, but we do not repeat their documentation here since they are identical to the 32-bit versions except as noted above.  File: fftw3.info, Node: New-array Execute Functions, Next: Wisdom, Prev: Guru Interface, Up: FFTW Reference 4.6 New-array Execute Functions =============================== Normally, one executes a plan for the arrays with which the plan was created, by calling `fftw_execute(plan)' as described in *note Using Plans::. However, it is possible for sophisticated users to apply a given plan to a _different_ array using the "new-array execute" functions detailed below, provided that the following conditions are met: * The array size, strides, etcetera are the same (since those are set by the plan). * The input and output arrays are the same (in-place) or different (out-of-place) if the plan was originally created to be in-place or out-of-place, respectively. * For split arrays, the separations between the real and imaginary parts, `ii-ri' and `io-ro', are the same as they were for the input and output arrays when the plan was created. (This condition is automatically satisfied for interleaved arrays.) * The "alignment" of the new input/output arrays is the same as that of the input/output arrays when the plan was created, unless the plan was created with the `FFTW_UNALIGNED' flag. Here, the alignment is a platform-dependent quantity (for example, it is the address modulo 16 if SSE SIMD instructions are used, but the address modulo 4 for non-SIMD single-precision FFTW on the same machine). In general, only arrays allocated with `fftw_malloc' are guaranteed to be equally aligned (*note SIMD alignment and fftw_malloc::). The alignment issue is especially critical, because if you don't use `fftw_malloc' then you may have little control over the alignment of arrays in memory. For example, neither the C++ `new' function nor the Fortran `allocate' statement provide strong enough guarantees about data alignment. If you don't use `fftw_malloc', therefore, you probably have to use `FFTW_UNALIGNED' (which disables most SIMD support). If possible, it is probably better for you to simply create multiple plans (creating a new plan is quick once one exists for a given size), or better yet re-use the same array for your transforms. If you are tempted to use the new-array execute interface because you want to transform a known bunch of arrays of the same size, you should probably go use the advanced interface instead (*note Advanced Interface::)). The new-array execute functions are: void fftw_execute_dft( const fftw_plan p, fftw_complex *in, fftw_complex *out); void fftw_execute_split_dft( const fftw_plan p, double *ri, double *ii, double *ro, double *io); void fftw_execute_dft_r2c( const fftw_plan p, double *in, fftw_complex *out); void fftw_execute_split_dft_r2c( const fftw_plan p, double *in, double *ro, double *io); void fftw_execute_dft_c2r( const fftw_plan p, fftw_complex *in, double *out); void fftw_execute_split_dft_c2r( const fftw_plan p, double *ri, double *ii, double *out); void fftw_execute_r2r( const fftw_plan p, double *in, double *out); These execute the `plan' to compute the corresponding transform on the input/output arrays specified by the subsequent arguments. The input/output array arguments have the same meanings as the ones passed to the guru planner routines in the preceding sections. The `plan' is not modified, and these routines can be called as many times as desired, or intermixed with calls to the ordinary `fftw_execute'. The `plan' _must_ have been created for the transform type corresponding to the execute function, e.g. it must be a complex-DFT plan for `fftw_execute_dft'. Any of the planner routines for that transform type, from the basic to the guru interface, could have been used to create the plan, however.  File: fftw3.info, Node: Wisdom, Next: What FFTW Really Computes, Prev: New-array Execute Functions, Up: FFTW Reference 4.7 Wisdom ========== This section documents the FFTW mechanism for saving and restoring plans from disk. This mechanism is called "wisdom". * Menu: * Wisdom Export:: * Wisdom Import:: * Forgetting Wisdom:: * Wisdom Utilities::  File: fftw3.info, Node: Wisdom Export, Next: Wisdom Import, Prev: Wisdom, Up: Wisdom 4.7.1 Wisdom Export ------------------- int fftw_export_wisdom_to_filename(const char *filename); void fftw_export_wisdom_to_file(FILE *output_file); char *fftw_export_wisdom_to_string(void); void fftw_export_wisdom(void (*write_char)(char c, void *), void *data); These functions allow you to export all currently accumulated wisdom in a form from which it can be later imported and restored, even during a separate run of the program. (*Note Words of Wisdom-Saving Plans::.) The current store of wisdom is not affected by calling any of these routines. `fftw_export_wisdom' exports the wisdom to any output medium, as specified by the callback function `write_char'. `write_char' is a `putc'-like function that writes the character `c' to some output; its second parameter is the `data' pointer passed to `fftw_export_wisdom'. For convenience, the following three "wrapper" routines are provided: `fftw_export_wisdom_to_filename' writes wisdom to a file named `filename' (which is created or overwritten), returning `1' on success and `0' on failure. A lower-level function, which requires you to open and close the file yourself (e.g. if you want to write wisdom to a portion of a larger file) is `fftw_export_wisdom_to_file'. This writes the wisdom to the current position in `output_file', which should be open with write permission; upon exit, the file remains open and is positioned at the end of the wisdom data. `fftw_export_wisdom_to_string' returns a pointer to a `NULL'-terminated string holding the wisdom data. This string is dynamically allocated, and it is the responsibility of the caller to deallocate it with `free' when it is no longer needed. All of these routines export the wisdom in the same format, which we will not document here except to say that it is LISP-like ASCII text that is insensitive to white space.  File: fftw3.info, Node: Wisdom Import, Next: Forgetting Wisdom, Prev: Wisdom Export, Up: Wisdom 4.7.2 Wisdom Import ------------------- int fftw_import_system_wisdom(void); int fftw_import_wisdom_from_filename(const char *filename); int fftw_import_wisdom_from_string(const char *input_string); int fftw_import_wisdom(int (*read_char)(void *), void *data); These functions import wisdom into a program from data stored by the `fftw_export_wisdom' functions above. (*Note Words of Wisdom-Saving Plans::.) The imported wisdom replaces any wisdom already accumulated by the running program. `fftw_import_wisdom' imports wisdom from any input medium, as specified by the callback function `read_char'. `read_char' is a `getc'-like function that returns the next character in the input; its parameter is the `data' pointer passed to `fftw_import_wisdom'. If the end of the input data is reached (which should never happen for valid data), `read_char' should return `EOF' (as defined in `'). For convenience, the following three "wrapper" routines are provided: `fftw_import_wisdom_from_filename' reads wisdom from a file named `filename'. A lower-level function, which requires you to open and close the file yourself (e.g. if you want to read wisdom from a portion of a larger file) is `fftw_import_wisdom_from_file'. This reads wisdom from the current position in `input_file' (which should be open with read permission); upon exit, the file remains open, but the position of the read pointer is unspecified. `fftw_import_wisdom_from_string' reads wisdom from the `NULL'-terminated string `input_string'. `fftw_import_system_wisdom' reads wisdom from an implementation-defined standard file (`/etc/fftw/wisdom' on Unix and GNU systems). The return value of these import routines is `1' if the wisdom was read successfully and `0' otherwise. Note that, in all of these functions, any data in the input stream past the end of the wisdom data is simply ignored.  File: fftw3.info, Node: Forgetting Wisdom, Next: Wisdom Utilities, Prev: Wisdom Import, Up: Wisdom 4.7.3 Forgetting Wisdom ----------------------- void fftw_forget_wisdom(void); Calling `fftw_forget_wisdom' causes all accumulated `wisdom' to be discarded and its associated memory to be freed. (New `wisdom' can still be gathered subsequently, however.)  File: fftw3.info, Node: Wisdom Utilities, Prev: Forgetting Wisdom, Up: Wisdom 4.7.4 Wisdom Utilities ---------------------- FFTW includes two standalone utility programs that deal with wisdom. We merely summarize them here, since they come with their own `man' pages for Unix and GNU systems (with HTML versions on our web site). The first program is `fftw-wisdom' (or `fftwf-wisdom' in single precision, etcetera), which can be used to create a wisdom file containing plans for any of the transform sizes and types supported by FFTW. It is preferable to create wisdom directly from your executable (*note Caveats in Using Wisdom::), but this program is useful for creating global wisdom files for `fftw_import_system_wisdom'. The second program is `fftw-wisdom-to-conf', which takes a wisdom file as input and produces a "configuration routine" as output. The latter is a C subroutine that you can compile and link into your program, replacing a routine of the same name in the FFTW library, that determines which parts of FFTW are callable by your program. `fftw-wisdom-to-conf' produces a configuration routine that links to only those parts of FFTW needed by the saved plans in the wisdom, greatly reducing the size of statically linked executables (which should only attempt to create plans corresponding to those in the wisdom, however).  File: fftw3.info, Node: What FFTW Really Computes, Prev: Wisdom, Up: FFTW Reference 4.8 What FFTW Really Computes ============================= In this section, we provide precise mathematical definitions for the transforms that FFTW computes. These transform definitions are fairly standard, but some authors follow slightly different conventions for the normalization of the transform (the constant factor in front) and the sign of the complex exponent. We begin by presenting the one-dimensional (1d) transform definitions, and then give the straightforward extension to multi-dimensional transforms. * Menu: * The 1d Discrete Fourier Transform (DFT):: * The 1d Real-data DFT:: * 1d Real-even DFTs (DCTs):: * 1d Real-odd DFTs (DSTs):: * 1d Discrete Hartley Transforms (DHTs):: * Multi-dimensional Transforms::  File: fftw3.info, Node: The 1d Discrete Fourier Transform (DFT), Next: The 1d Real-data DFT, Prev: What FFTW Really Computes, Up: What FFTW Really Computes 4.8.1 The 1d Discrete Fourier Transform (DFT) --------------------------------------------- The forward (`FFTW_FORWARD') discrete Fourier transform (DFT) of a 1d complex array X of size n computes an array Y, where: Y[k] = sum for j = 0 to (n - 1) of X[j] * exp(-2 pi j k sqrt(-1)/n) . The backward (`FFTW_BACKWARD') DFT computes: Y[k] = sum for j = 0 to (n - 1) of X[j] * exp(2 pi j k sqrt(-1)/n) . FFTW computes an unnormalized transform, in that there is no coefficient in front of the summation in the DFT. In other words, applying the forward and then the backward transform will multiply the input by n. From above, an `FFTW_FORWARD' transform corresponds to a sign of -1 in the exponent of the DFT. Note also that we use the standard "in-order" output ordering--the k-th output corresponds to the frequency k/n (or k/T, where T is your total sampling period). For those who like to think in terms of positive and negative frequencies, this means that the positive frequencies are stored in the first half of the output and the negative frequencies are stored in backwards order in the second half of the output. (The frequency -k/n is the same as the frequency (n-k)/n.)  File: fftw3.info, Node: The 1d Real-data DFT, Next: 1d Real-even DFTs (DCTs), Prev: The 1d Discrete Fourier Transform (DFT), Up: What FFTW Really Computes 4.8.2 The 1d Real-data DFT -------------------------- The real-input (r2c) DFT in FFTW computes the _forward_ transform Y of the size `n' real array X, exactly as defined above, i.e. Y[k] = sum for j = 0 to (n - 1) of X[j] * exp(-2 pi j k sqrt(-1)/n) . This output array Y can easily be shown to possess the "Hermitian" symmetry Y[k] = Y[n-k]*, where we take Y to be periodic so that Y[n] = Y[0]. As a result of this symmetry, half of the output Y is redundant (being the complex conjugate of the other half), and so the 1d r2c transforms only output elements 0...n/2 of Y (n/2+1 complex numbers), where the division by 2 is rounded down. Moreover, the Hermitian symmetry implies that Y[0] and, if n is even, the Y[n/2] element, are purely real. So, for the `R2HC' r2r transform, these elements are not stored in the halfcomplex output format. The c2r and `H2RC' r2r transforms compute the backward DFT of the _complex_ array X with Hermitian symmetry, stored in the r2c/`R2HC' output formats, respectively, where the backward transform is defined exactly as for the complex case: Y[k] = sum for j = 0 to (n - 1) of X[j] * exp(2 pi j k sqrt(-1)/n) . The outputs `Y' of this transform can easily be seen to be purely real, and are stored as an array of real numbers. Like FFTW's complex DFT, these transforms are unnormalized. In other words, applying the real-to-complex (forward) and then the complex-to-real (backward) transform will multiply the input by n.  File: fftw3.info, Node: 1d Real-even DFTs (DCTs), Next: 1d Real-odd DFTs (DSTs), Prev: The 1d Real-data DFT, Up: What FFTW Really Computes 4.8.3 1d Real-even DFTs (DCTs) ------------------------------ The Real-even symmetry DFTs in FFTW are exactly equivalent to the unnormalized forward (and backward) DFTs as defined above, where the input array X of length N is purely real and is also "even" symmetry. In this case, the output array is likewise real and even symmetry. For the case of `REDFT00', this even symmetry means that X[j] = X[N-j], where we take X to be periodic so that X[N] = X[0]. Because of this redundancy, only the first n real numbers are actually stored, where N = 2(n-1). The proper definition of even symmetry for `REDFT10', `REDFT01', and `REDFT11' transforms is somewhat more intricate because of the shifts by 1/2 of the input and/or output, although the corresponding boundary conditions are given in *note Real even/odd DFTs (cosine/sine transforms)::. Because of the even symmetry, however, the sine terms in the DFT all cancel and the remaining cosine terms are written explicitly below. This formulation often leads people to call such a transform a "discrete cosine transform" (DCT), although it is really just a special case of the DFT. In each of the definitions below, we transform a real array X of length n to a real array Y of length n: REDFT00 (DCT-I) ............... An `REDFT00' transform (type-I DCT) in FFTW is defined by: Y[k] = X[0] + (-1)^k X[n-1] + 2 (sum for j = 1 to n-2 of X[j] cos(pi jk /(n-1))). Note that this transform is not defined for n=1. For n=2, the summation term above is dropped as you might expect. REDFT10 (DCT-II) ................ An `REDFT10' transform (type-II DCT, sometimes called "the" DCT) in FFTW is defined by: Y[k] = 2 (sum for j = 0 to n-1 of X[j] cos(pi (j+1/2) k / n)). REDFT01 (DCT-III) ................. An `REDFT01' transform (type-III DCT) in FFTW is defined by: Y[k] = X[0] + 2 (sum for j = 1 to n-1 of X[j] cos(pi j (k+1/2) / n)). In the case of n=1, this reduces to Y[0] = X[0]. Up to a scale factor (see below), this is the inverse of `REDFT10' ("the" DCT), and so the `REDFT01' (DCT-III) is sometimes called the "IDCT". REDFT11 (DCT-IV) ................ An `REDFT11' transform (type-IV DCT) in FFTW is defined by: Y[k] = 2 (sum for j = 0 to n-1 of X[j] cos(pi (j+1/2) (k+1/2) / n)). Inverses and Normalization .......................... These definitions correspond directly to the unnormalized DFTs used elsewhere in FFTW (hence the factors of 2 in front of the summations). The unnormalized inverse of `REDFT00' is `REDFT00', of `REDFT10' is `REDFT01' and vice versa, and of `REDFT11' is `REDFT11'. Each unnormalized inverse results in the original array multiplied by N, where N is the _logical_ DFT size. For `REDFT00', N=2(n-1) (note that n=1 is not defined); otherwise, N=2n. In defining the discrete cosine transform, some authors also include additional factors of sqrt(2) (or its inverse) multiplying selected inputs and/or outputs. This is a mostly cosmetic change that makes the transform orthogonal, but sacrifices the direct equivalence to a symmetric DFT.  File: fftw3.info, Node: 1d Real-odd DFTs (DSTs), Next: 1d Discrete Hartley Transforms (DHTs), Prev: 1d Real-even DFTs (DCTs), Up: What FFTW Really Computes 4.8.4 1d Real-odd DFTs (DSTs) ----------------------------- The Real-odd symmetry DFTs in FFTW are exactly equivalent to the unnormalized forward (and backward) DFTs as defined above, where the input array X of length N is purely real and is also "odd" symmetry. In this case, the output is odd symmetry and purely imaginary. For the case of `RODFT00', this odd symmetry means that X[j] = -X[N-j], where we take X to be periodic so that X[N] = X[0]. Because of this redundancy, only the first n real numbers starting at j=1 are actually stored (the j=0 element is zero), where N = 2(n+1). The proper definition of odd symmetry for `RODFT10', `RODFT01', and `RODFT11' transforms is somewhat more intricate because of the shifts by 1/2 of the input and/or output, although the corresponding boundary conditions are given in *note Real even/odd DFTs (cosine/sine transforms)::. Because of the odd symmetry, however, the cosine terms in the DFT all cancel and the remaining sine terms are written explicitly below. This formulation often leads people to call such a transform a "discrete sine transform" (DST), although it is really just a special case of the DFT. In each of the definitions below, we transform a real array X of length n to a real array Y of length n: RODFT00 (DST-I) ............... An `RODFT00' transform (type-I DST) in FFTW is defined by: Y[k] = 2 (sum for j = 0 to n-1 of X[j] sin(pi (j+1)(k+1) / (n+1))). RODFT10 (DST-II) ................ An `RODFT10' transform (type-II DST) in FFTW is defined by: Y[k] = 2 (sum for j = 0 to n-1 of X[j] sin(pi (j+1/2) (k+1) / n)). RODFT01 (DST-III) ................. An `RODFT01' transform (type-III DST) in FFTW is defined by: Y[k] = (-1)^k X[n-1] + 2 (sum for j = 0 to n-2 of X[j] sin(pi (j+1) (k+1/2) / n)). In the case of n=1, this reduces to Y[0] = X[0]. RODFT11 (DST-IV) ................ An `RODFT11' transform (type-IV DST) in FFTW is defined by: Y[k] = 2 (sum for j = 0 to n-1 of X[j] sin(pi (j+1/2) (k+1/2) / n)). Inverses and Normalization .......................... These definitions correspond directly to the unnormalized DFTs used elsewhere in FFTW (hence the factors of 2 in front of the summations). The unnormalized inverse of `RODFT00' is `RODFT00', of `RODFT10' is `RODFT01' and vice versa, and of `RODFT11' is `RODFT11'. Each unnormalized inverse results in the original array multiplied by N, where N is the _logical_ DFT size. For `RODFT00', N=2(n+1); otherwise, N=2n. In defining the discrete sine transform, some authors also include additional factors of sqrt(2) (or its inverse) multiplying selected inputs and/or outputs. This is a mostly cosmetic change that makes the transform orthogonal, but sacrifices the direct equivalence to an antisymmetric DFT.  File: fftw3.info, Node: 1d Discrete Hartley Transforms (DHTs), Next: Multi-dimensional Transforms, Prev: 1d Real-odd DFTs (DSTs), Up: What FFTW Really Computes 4.8.5 1d Discrete Hartley Transforms (DHTs) ------------------------------------------- The discrete Hartley transform (DHT) of a 1d real array X of size n computes a real array Y of the same size, where: Y[k] = sum for j = 0 to (n - 1) of X[j] * [cos(2 pi j k / n) + sin(2 pi j k / n)]. FFTW computes an unnormalized transform, in that there is no coefficient in front of the summation in the DHT. In other words, applying the transform twice (the DHT is its own inverse) will multiply the input by n.  File: fftw3.info, Node: Multi-dimensional Transforms, Prev: 1d Discrete Hartley Transforms (DHTs), Up: What FFTW Really Computes 4.8.6 Multi-dimensional Transforms ---------------------------------- The multi-dimensional transforms of FFTW, in general, compute simply the separable product of the given 1d transform along each dimension of the array. Since each of these transforms is unnormalized, computing the forward followed by the backward/inverse multi-dimensional transform will result in the original array scaled by the product of the normalization factors for each dimension (e.g. the product of the dimension sizes, for a multi-dimensional DFT). The definition of FFTW's multi-dimensional DFT of real data (r2c) deserves special attention. In this case, we logically compute the full multi-dimensional DFT of the input data; since the input data are purely real, the output data have the Hermitian symmetry and therefore only one non-redundant half need be stored. More specifically, for an n[0] x n[1] x n[2] x ... x n[d-1] multi-dimensional real-input DFT, the full (logical) complex output array Y[k[0], k[1], ..., k[d-1]] has the symmetry: Y[k[0], k[1], ..., k[d-1]] = Y[n[0] - k[0], n[1] - k[1], ..., n[d-1] - k[d-1]]* (where each dimension is periodic). Because of this symmetry, we only store the k[d-1] = 0...n[d-1]/2 elements of the _last_ dimension (division by 2 is rounded down). (We could instead have cut any other dimension in half, but the last dimension proved computationally convenient.) This results in the peculiar array format described in more detail by *note Real-data DFT Array Format::. The multi-dimensional c2r transform is simply the unnormalized inverse of the r2c transform. i.e. it is the same as FFTW's complex backward multi-dimensional DFT, operating on a Hermitian input array in the peculiar format mentioned above and outputting a real array (since the DFT output is purely real). We should remind the user that the separable product of 1d transforms along each dimension, as computed by FFTW, is not always the same thing as the usual multi-dimensional transform. A multi-dimensional `R2HC' (or `HC2R') transform is not identical to the multi-dimensional DFT, requiring some post-processing to combine the requisite real and imaginary parts, as was described in *note The Halfcomplex-format DFT::. Likewise, FFTW's multidimensional `FFTW_DHT' r2r transform is not the same thing as the logical multi-dimensional discrete Hartley transform defined in the literature, as discussed in *note The Discrete Hartley Transform::.  File: fftw3.info, Node: Multi-threaded FFTW, Next: Distributed-memory FFTW with MPI, Prev: FFTW Reference, Up: Top 5 Multi-threaded FFTW ********************* In this chapter we document the parallel FFTW routines for shared-memory parallel hardware. These routines, which support parallel one- and multi-dimensional transforms of both real and complex data, are the easiest way to take advantage of multiple processors with FFTW. They work just like the corresponding uniprocessor transform routines, except that you have an extra initialization routine to call, and there is a routine to set the number of threads to employ. Any program that uses the uniprocessor FFTW can therefore be trivially modified to use the multi-threaded FFTW. A shared-memory machine is one in which all CPUs can directly access the same main memory, and such machines are now common due to the ubiquity of multi-core CPUs. FFTW's multi-threading support allows you to utilize these additional CPUs transparently from a single program. However, this does not necessarily translate into performance gains--when multiple threads/CPUs are employed, there is an overhead required for synchronization that may outweigh the computatational parallelism. Therefore, you can only benefit from threads if your problem is sufficiently large. * Menu: * Installation and Supported Hardware/Software:: * Usage of Multi-threaded FFTW:: * How Many Threads to Use?:: * Thread safety::  File: fftw3.info, Node: Installation and Supported Hardware/Software, Next: Usage of Multi-threaded FFTW, Prev: Multi-threaded FFTW, Up: Multi-threaded FFTW 5.1 Installation and Supported Hardware/Software ================================================ All of the FFTW threads code is located in the `threads' subdirectory of the FFTW package. On Unix systems, the FFTW threads libraries and header files can be automatically configured, compiled, and installed along with the uniprocessor FFTW libraries simply by including `--enable-threads' in the flags to the `configure' script (*note Installation on Unix::), or `--enable-openmp' to use OpenMP (http://www.openmp.org) threads. The threads routines require your operating system to have some sort of shared-memory threads support. Specifically, the FFTW threads package works with POSIX threads (available on most Unix variants, from GNU/Linux to MacOS X) and Win32 threads. OpenMP threads, which are supported in many common compilers (e.g. gcc) are also supported, and may give better performance on some systems. (OpenMP threads are also useful if you are employing OpenMP in your own code, in order to minimize conflicts between threading models.) If you have a shared-memory machine that uses a different threads API, it should be a simple matter of programming to include support for it; see the file `threads/threads.c' for more detail. You can compile FFTW with _both_ `--enable-threads' and `--enable-openmp' at the same time, since they install libraries with different names (`fftw3_threads' and `fftw3_omp', as described below). However, your programs may only link to _one_ of these two libraries at a time. Ideally, of course, you should also have multiple processors in order to get any benefit from the threaded transforms.  File: fftw3.info, Node: Usage of Multi-threaded FFTW, Next: How Many Threads to Use?, Prev: Installation and Supported Hardware/Software, Up: Multi-threaded FFTW 5.2 Usage of Multi-threaded FFTW ================================ Here, it is assumed that the reader is already familiar with the usage of the uniprocessor FFTW routines, described elsewhere in this manual. We only describe what one has to change in order to use the multi-threaded routines. First, programs using the parallel complex transforms should be linked with `-lfftw3_threads -lfftw3 -lm' on Unix, or `-lfftw3_omp -lfftw3 -lm' if you compiled with OpenMP. You will also need to link with whatever library is responsible for threads on your system (e.g. `-lpthread' on GNU/Linux) or include whatever compiler flag enables OpenMP (e.g. `-fopenmp' with gcc). Second, before calling _any_ FFTW routines, you should call the function: int fftw_init_threads(void); This function, which need only be called once, performs any one-time initialization required to use threads on your system. It returns zero if there was some error (which should not happen under normal circumstances) and a non-zero value otherwise. Third, before creating a plan that you want to parallelize, you should call: void fftw_plan_with_nthreads(int nthreads); The `nthreads' argument indicates the number of threads you want FFTW to use (or actually, the maximum number). All plans subsequently created with any planner routine will use that many threads. You can call `fftw_plan_with_nthreads', create some plans, call `fftw_plan_with_nthreads' again with a different argument, and create some more plans for a new number of threads. Plans already created before a call to `fftw_plan_with_nthreads' are unaffected. If you pass an `nthreads' argument of `1' (the default), threads are disabled for subsequent plans. With OpenMP, to configure FFTW to use all of the currently running OpenMP threads (set by `omp_set_num_threads(nthreads)' or by the `OMP_NUM_THREADS' environment variable), you can do: `fftw_plan_with_nthreads(omp_get_max_threads())'. (The `omp_' OpenMP functions are declared via `#include '.) Given a plan, you then execute it as usual with `fftw_execute(plan)', and the execution will use the number of threads specified when the plan was created. When done, you destroy it as usual with `fftw_destroy_plan'. As described in *note Thread safety::, plan _execution_ is thread-safe, but plan creation and destruction are _not_: you should create/destroy plans only from a single thread, but can safely execute multiple plans in parallel. There is one additional routine: if you want to get rid of all memory and other resources allocated internally by FFTW, you can call: void fftw_cleanup_threads(void); which is much like the `fftw_cleanup()' function except that it also gets rid of threads-related data. You must _not_ execute any previously created plans after calling this function. We should also mention one other restriction: if you save wisdom from a program using the multi-threaded FFTW, that wisdom _cannot be used_ by a program using only the single-threaded FFTW (i.e. not calling `fftw_init_threads'). *Note Words of Wisdom-Saving Plans::.  File: fftw3.info, Node: How Many Threads to Use?, Next: Thread safety, Prev: Usage of Multi-threaded FFTW, Up: Multi-threaded FFTW 5.3 How Many Threads to Use? ============================ There is a fair amount of overhead involved in synchronizing threads, so the optimal number of threads to use depends upon the size of the transform as well as on the number of processors you have. As a general rule, you don't want to use more threads than you have processors. (Using more threads will work, but there will be extra overhead with no benefit.) In fact, if the problem size is too small, you may want to use fewer threads than you have processors. You will have to experiment with your system to see what level of parallelization is best for your problem size. Typically, the problem will have to involve at least a few thousand data points before threads become beneficial. If you plan with `FFTW_PATIENT', it will automatically disable threads for sizes that don't benefit from parallelization.  File: fftw3.info, Node: Thread safety, Prev: How Many Threads to Use?, Up: Multi-threaded FFTW 5.4 Thread safety ================= Users writing multi-threaded programs (including OpenMP) must concern themselves with the "thread safety" of the libraries they use--that is, whether it is safe to call routines in parallel from multiple threads. FFTW can be used in such an environment, but some care must be taken because the planner routines share data (e.g. wisdom and trigonometric tables) between calls and plans. The upshot is that the only thread-safe (re-entrant) routine in FFTW is `fftw_execute' (and the new-array variants thereof). All other routines (e.g. the planner) should only be called from one thread at a time. So, for example, you can wrap a semaphore lock around any calls to the planner; even more simply, you can just create all of your plans from one thread. We do not think this should be an important restriction (FFTW is designed for the situation where the only performance-sensitive code is the actual execution of the transform), and the benefits of shared data between plans are great. Note also that, since the plan is not modified by `fftw_execute', it is safe to execute the _same plan_ in parallel by multiple threads. However, since a given plan operates by default on a fixed array, you need to use one of the new-array execute functions (*note New-array Execute Functions::) so that different threads compute the transform of different data. (Users should note that these comments only apply to programs using shared-memory threads or OpenMP. Parallelism using MPI or forked processes involves a separate address-space and global variables for each process, and is not susceptible to problems of this sort.) If you are configured FFTW with the `--enable-debug' or `--enable-debug-malloc' flags (*note Installation on Unix::), then `fftw_execute' is not thread-safe. These flags are not documented because they are intended only for developing and debugging FFTW, but if you must use `--enable-debug' then you should also specifically pass `--disable-debug-malloc' for `fftw_execute' to be thread-safe.  File: fftw3.info, Node: Distributed-memory FFTW with MPI, Next: Calling FFTW from Modern Fortran, Prev: Multi-threaded FFTW, Up: Top 6 Distributed-memory FFTW with MPI ********************************** In this chapter we document the parallel FFTW routines for parallel systems supporting the MPI message-passing interface. Unlike the shared-memory threads described in the previous chapter, MPI allows you to use _distributed-memory_ parallelism, where each CPU has its own separate memory, and which can scale up to clusters of many thousands of processors. This capability comes at a price, however: each process only stores a _portion_ of the data to be transformed, which means that the data structures and programming-interface are quite different from the serial or threads versions of FFTW. Distributed-memory parallelism is especially useful when you are transforming arrays so large that they do not fit into the memory of a single processor. The storage per-process required by FFTW's MPI routines is proportional to the total array size divided by the number of processes. Conversely, distributed-memory parallelism can easily pose an unacceptably high communications overhead for small problems; the threshold problem size for which parallelism becomes advantageous will depend on the precise problem you are interested in, your hardware, and your MPI implementation. A note on terminology: in MPI, you divide the data among a set of "processes" which each run in their own memory address space. Generally, each process runs on a different physical processor, but this is not required. A set of processes in MPI is described by an opaque data structure called a "communicator," the most common of which is the predefined communicator `MPI_COMM_WORLD' which refers to _all_ processes. For more information on these and other concepts common to all MPI programs, we refer the reader to the documentation at the MPI home page (http://www.mcs.anl.gov/research/projects/mpi/). We assume in this chapter that the reader is familiar with the usage of the serial (uniprocessor) FFTW, and focus only on the concepts new to the MPI interface. * Menu: * FFTW MPI Installation:: * Linking and Initializing MPI FFTW:: * 2d MPI example:: * MPI Data Distribution:: * Multi-dimensional MPI DFTs of Real Data:: * Other Multi-dimensional Real-data MPI Transforms:: * FFTW MPI Transposes:: * FFTW MPI Wisdom:: * Avoiding MPI Deadlocks:: * FFTW MPI Performance Tips:: * Combining MPI and Threads:: * FFTW MPI Reference:: * FFTW MPI Fortran Interface::  File: fftw3.info, Node: FFTW MPI Installation, Next: Linking and Initializing MPI FFTW, Prev: Distributed-memory FFTW with MPI, Up: Distributed-memory FFTW with MPI 6.1 FFTW MPI Installation ========================= All of the FFTW MPI code is located in the `mpi' subdirectory of the FFTW package. On Unix systems, the FFTW MPI libraries and header files are automatically configured, compiled, and installed along with the uniprocessor FFTW libraries simply by including `--enable-mpi' in the flags to the `configure' script (*note Installation on Unix::). Any implementation of the MPI standard, version 1 or later, should work with FFTW. The `configure' script will attempt to automatically detect how to compile and link code using your MPI implementation. In some cases, especially if you have multiple different MPI implementations installed or have an unusual MPI software package, you may need to provide this information explicitly. Most commonly, one compiles MPI code by invoking a special compiler command, typically `mpicc' for C code. The `configure' script knows the most common names for this command, but you can specify the MPI compilation command explicitly by setting the `MPICC' variable, as in `./configure MPICC=mpicc ...'. If, instead of a special compiler command, you need to link a certain library, you can specify the link command via the `MPILIBS' variable, as in `./configure MPILIBS=-lmpi ...'. Note that if your MPI library is installed in a non-standard location (one the compiler does not know about by default), you may also have to specify the location of the library and header files via `LDFLAGS' and `CPPFLAGS' variables, respectively, as in `./configure LDFLAGS=-L/path/to/mpi/libs CPPFLAGS=-I/path/to/mpi/include ...'.  File: fftw3.info, Node: Linking and Initializing MPI FFTW, Next: 2d MPI example, Prev: FFTW MPI Installation, Up: Distributed-memory FFTW with MPI 6.2 Linking and Initializing MPI FFTW ===================================== Programs using the MPI FFTW routines should be linked with `-lfftw3_mpi -lfftw3 -lm' on Unix in double precision, `-lfftw3f_mpi -lfftw3f -lm' in single precision, and so on (*note Precision::). You will also need to link with whatever library is responsible for MPI on your system; in most MPI implementations, there is a special compiler alias named `mpicc' to compile and link MPI code. Before calling any FFTW routines except possibly `fftw_init_threads' (*note Combining MPI and Threads::), but after calling `MPI_Init', you should call the function: void fftw_mpi_init(void); If, at the end of your program, you want to get rid of all memory and other resources allocated internally by FFTW, for both the serial and MPI routines, you can call: void fftw_mpi_cleanup(void); which is much like the `fftw_cleanup()' function except that it also gets rid of FFTW's MPI-related data. You must _not_ execute any previously created plans after calling this function.  File: fftw3.info, Node: 2d MPI example, Next: MPI Data Distribution, Prev: Linking and Initializing MPI FFTW, Up: Distributed-memory FFTW with MPI 6.3 2d MPI example ================== Before we document the FFTW MPI interface in detail, we begin with a simple example outlining how one would perform a two-dimensional `N0' by `N1' complex DFT. #include int main(int argc, char **argv) { const ptrdiff_t N0 = ..., N1 = ...; fftw_plan plan; fftw_complex *data; ptrdiff_t alloc_local, local_n0, local_0_start, i, j; MPI_Init(&argc, &argv); fftw_mpi_init(); /* get local data size and allocate */ alloc_local = fftw_mpi_local_size_2d(N0, N1, MPI_COMM_WORLD, &local_n0, &local_0_start); data = fftw_alloc_complex(alloc_local); /* create plan for in-place forward DFT */ plan = fftw_mpi_plan_dft_2d(N0, N1, data, data, MPI_COMM_WORLD, FFTW_FORWARD, FFTW_ESTIMATE); /* initialize data to some function my_function(x,y) */ for (i = 0; i < local_n0; ++i) for (j = 0; j < N1; ++j) data[i*N1 + j] = my_function(local_0_start + i, j); /* compute transforms, in-place, as many times as desired */ fftw_execute(plan); fftw_destroy_plan(plan); MPI_Finalize(); } As can be seen above, the MPI interface follows the same basic style of allocate/plan/execute/destroy as the serial FFTW routines. All of the MPI-specific routines are prefixed with `fftw_mpi_' instead of `fftw_'. There are a few important differences, however: First, we must call `fftw_mpi_init()' after calling `MPI_Init' (required in all MPI programs) and before calling any other `fftw_mpi_' routine. Second, when we create the plan with `fftw_mpi_plan_dft_2d', analogous to `fftw_plan_dft_2d', we pass an additional argument: the communicator, indicating which processes will participate in the transform (here `MPI_COMM_WORLD', indicating all processes). Whenever you create, execute, or destroy a plan for an MPI transform, you must call the corresponding FFTW routine on _all_ processes in the communicator for that transform. (That is, these are _collective_ calls.) Note that the plan for the MPI transform uses the standard `fftw_execute' and `fftw_destroy' routines (on the other hand, there are MPI-specific new-array execute functions documented below). Third, all of the FFTW MPI routines take `ptrdiff_t' arguments instead of `int' as for the serial FFTW. `ptrdiff_t' is a standard C integer type which is (at least) 32 bits wide on a 32-bit machine and 64 bits wide on a 64-bit machine. This is to make it easy to specify very large parallel transforms on a 64-bit machine. (You can specify 64-bit transform sizes in the serial FFTW, too, but only by using the `guru64' planner interface. *Note 64-bit Guru Interface::.) Fourth, and most importantly, you don't allocate the entire two-dimensional array on each process. Instead, you call `fftw_mpi_local_size_2d' to find out what _portion_ of the array resides on each processor, and how much space to allocate. Here, the portion of the array on each process is a `local_n0' by `N1' slice of the total array, starting at index `local_0_start'. The total number of `fftw_complex' numbers to allocate is given by the `alloc_local' return value, which _may_ be greater than `local_n0 * N1' (in case some intermediate calculations require additional storage). The data distribution in FFTW's MPI interface is described in more detail by the next section. Given the portion of the array that resides on the local process, it is straightforward to initialize the data (here to a function `myfunction') and otherwise manipulate it. Of course, at the end of the program you may want to output the data somehow, but synchronizing this output is up to you and is beyond the scope of this manual. (One good way to output a large multi-dimensional distributed array in MPI to a portable binary file is to use the free HDF5 library; see the HDF home page (http://www.hdfgroup.org/).)  File: fftw3.info, Node: MPI Data Distribution, Next: Multi-dimensional MPI DFTs of Real Data, Prev: 2d MPI example, Up: Distributed-memory FFTW with MPI 6.4 MPI Data Distribution ========================= The most important concept to understand in using FFTW's MPI interface is the data distribution. With a serial or multithreaded FFT, all of the inputs and outputs are stored as a single contiguous chunk of memory. With a distributed-memory FFT, the inputs and outputs are broken into disjoint blocks, one per process. In particular, FFTW uses a _1d block distribution_ of the data, distributed along the _first dimension_. For example, if you want to perform a 100 x 200 complex DFT, distributed over 4 processes, each process will get a 25 x 200 slice of the data. That is, process 0 will get rows 0 through 24, process 1 will get rows 25 through 49, process 2 will get rows 50 through 74, and process 3 will get rows 75 through 99. If you take the same array but distribute it over 3 processes, then it is not evenly divisible so the different processes will have unequal chunks. FFTW's default choice in this case is to assign 34 rows to processes 0 and 1, and 32 rows to process 2. FFTW provides several `fftw_mpi_local_size' routines that you can call to find out what portion of an array is stored on the current process. In most cases, you should use the default block sizes picked by FFTW, but it is also possible to specify your own block size. For example, with a 100 x 200 array on three processes, you can tell FFTW to use a block size of 40, which would assign 40 rows to processes 0 and 1, and 20 rows to process 2. FFTW's default is to divide the data equally among the processes if possible, and as best it can otherwise. The rows are always assigned in "rank order," i.e. process 0 gets the first block of rows, then process 1, and so on. (You can change this by using `MPI_Comm_split' to create a new communicator with re-ordered processes.) However, you should always call the `fftw_mpi_local_size' routines, if possible, rather than trying to predict FFTW's distribution choices. In particular, it is critical that you allocate the storage size that is returned by `fftw_mpi_local_size', which is _not_ necessarily the size of the local slice of the array. The reason is that intermediate steps of FFTW's algorithms involve transposing the array and redistributing the data, so at these intermediate steps FFTW may require more local storage space (albeit always proportional to the total size divided by the number of processes). The `fftw_mpi_local_size' functions know how much storage is required for these intermediate steps and tell you the correct amount to allocate. * Menu: * Basic and advanced distribution interfaces:: * Load balancing:: * Transposed distributions:: * One-dimensional distributions::  File: fftw3.info, Node: Basic and advanced distribution interfaces, Next: Load balancing, Prev: MPI Data Distribution, Up: MPI Data Distribution 6.4.1 Basic and advanced distribution interfaces ------------------------------------------------ As with the planner interface, the `fftw_mpi_local_size' distribution interface is broken into basic and advanced (`_many') interfaces, where the latter allows you to specify the block size manually and also to request block sizes when computing multiple transforms simultaneously. These functions are documented more exhaustively by the FFTW MPI Reference, but we summarize the basic ideas here using a couple of two-dimensional examples. For the 100 x 200 complex-DFT example, above, we would find the distribution by calling the following function in the basic interface: ptrdiff_t fftw_mpi_local_size_2d(ptrdiff_t n0, ptrdiff_t n1, MPI_Comm comm, ptrdiff_t *local_n0, ptrdiff_t *local_0_start); Given the total size of the data to be transformed (here, `n0 = 100' and `n1 = 200') and an MPI communicator (`comm'), this function provides three numbers. First, it describes the shape of the local data: the current process should store a `local_n0' by `n1' slice of the overall dataset, in row-major order (`n1' dimension contiguous), starting at index `local_0_start'. That is, if the total dataset is viewed as a `n0' by `n1' matrix, the current process should store the rows `local_0_start' to `local_0_start+local_n0-1'. Obviously, if you are running with only a single MPI process, that process will store the entire array: `local_0_start' will be zero and `local_n0' will be `n0'. *Note Row-major Format::. Second, the return value is the total number of data elements (e.g., complex numbers for a complex DFT) that should be allocated for the input and output arrays on the current process (ideally with `fftw_malloc' or an `fftw_alloc' function, to ensure optimal alignment). It might seem that this should always be equal to `local_n0 * n1', but this is _not_ the case. FFTW's distributed FFT algorithms require data redistributions at intermediate stages of the transform, and in some circumstances this may require slightly larger local storage. This is discussed in more detail below, under *note Load balancing::. The advanced-interface `local_size' function for multidimensional transforms returns the same three things (`local_n0', `local_0_start', and the total number of elements to allocate), but takes more inputs: ptrdiff_t fftw_mpi_local_size_many(int rnk, const ptrdiff_t *n, ptrdiff_t howmany, ptrdiff_t block0, MPI_Comm comm, ptrdiff_t *local_n0, ptrdiff_t *local_0_start); The two-dimensional case above corresponds to `rnk = 2' and an array `n' of length 2 with `n[0] = n0' and `n[1] = n1'. This routine is for any `rnk > 1'; one-dimensional transforms have their own interface because they work slightly differently, as discussed below. First, the advanced interface allows you to perform multiple transforms at once, of interleaved data, as specified by the `howmany' parameter. (`hoamany' is 1 for a single transform.) Second, here you can specify your desired block size in the `n0' dimension, `block0'. To use FFTW's default block size, pass `FFTW_MPI_DEFAULT_BLOCK' (0) for `block0'. Otherwise, on `P' processes, FFTW will return `local_n0' equal to `block0' on the first `P / block0' processes (rounded down), return `local_n0' equal to `n0 - block0 * (P / block0)' on the next process, and `local_n0' equal to zero on any remaining processes. In general, we recommend using the default block size (which corresponds to `n0 / P', rounded up). For example, suppose you have `P = 4' processes and `n0 = 21'. The default will be a block size of `6', which will give `local_n0 = 6' on the first three processes and `local_n0 = 3' on the last process. Instead, however, you could specify `block0 = 5' if you wanted, which would give `local_n0 = 5' on processes 0 to 2, `local_n0 = 6' on process 3. (This choice, while it may look superficially more "balanced," has the same critical path as FFTW's default but requires more communications.)  File: fftw3.info, Node: Load balancing, Next: Transposed distributions, Prev: Basic and advanced distribution interfaces, Up: MPI Data Distribution 6.4.2 Load balancing -------------------- Ideally, when you parallelize a transform over some P processes, each process should end up with work that takes equal time. Otherwise, all of the processes end up waiting on whichever process is slowest. This goal is known as "load balancing." In this section, we describe the circumstances under which FFTW is able to load-balance well, and in particular how you should choose your transform size in order to load balance. Load balancing is especially difficult when you are parallelizing over heterogeneous machines; for example, if one of your processors is a old 486 and another is a Pentium IV, obviously you should give the Pentium more work to do than the 486 since the latter is much slower. FFTW does not deal with this problem, however--it assumes that your processes run on hardware of comparable speed, and that the goal is therefore to divide the problem as equally as possible. For a multi-dimensional complex DFT, FFTW can divide the problem equally among the processes if: (i) the _first_ dimension `n0' is divisible by P; and (ii), the _product_ of the subsequent dimensions is divisible by P. (For the advanced interface, where you can specify multiple simultaneous transforms via some "vector" length `howmany', a factor of `howmany' is included in the product of the subsequent dimensions.) For a one-dimensional complex DFT, the length `N' of the data should be divisible by P _squared_ to be able to divide the problem equally among the processes.  File: fftw3.info, Node: Transposed distributions, Next: One-dimensional distributions, Prev: Load balancing, Up: MPI Data Distribution 6.4.3 Transposed distributions ------------------------------ Internally, FFTW's MPI transform algorithms work by first computing transforms of the data local to each process, then by globally _transposing_ the data in some fashion to redistribute the data among the processes, transforming the new data local to each process, and transposing back. For example, a two-dimensional `n0' by `n1' array, distributed across the `n0' dimension, is transformd by: (i) transforming the `n1' dimension, which are local to each process; (ii) transposing to an `n1' by `n0' array, distributed across the `n1' dimension; (iii) transforming the `n0' dimension, which is now local to each process; (iv) transposing back. However, in many applications it is acceptable to compute a multidimensional DFT whose results are produced in transposed order (e.g., `n1' by `n0' in two dimensions). This provides a significant performance advantage, because it means that the final transposition step can be omitted. FFTW supports this optimization, which you specify by passing the flag `FFTW_MPI_TRANSPOSED_OUT' to the planner routines. To compute the inverse transform of transposed output, you specify `FFTW_MPI_TRANSPOSED_IN' to tell it that the input is transposed. In this section, we explain how to interpret the output format of such a transform. Suppose you have are transforming multi-dimensional data with (at least two) dimensions n[0] x n[1] x n[2] x ... x n[d-1] . As always, it is distributed along the first dimension n[0] . Now, if we compute its DFT with the `FFTW_MPI_TRANSPOSED_OUT' flag, the resulting output data are stored with the first _two_ dimensions transposed: n[1] x n[0] x n[2] x ... x n[d-1] , distributed along the n[1] dimension. Conversely, if we take the n[1] x n[0] x n[2] x ... x n[d-1] data and transform it with the `FFTW_MPI_TRANSPOSED_IN' flag, then the format goes back to the original n[0] x n[1] x n[2] x ... x n[d-1] array. There are two ways to find the portion of the transposed array that resides on the current process. First, you can simply call the appropriate `local_size' function, passing n[1] x n[0] x n[2] x ... x n[d-1] (the transposed dimensions). This would mean calling the `local_size' function twice, once for the transposed and once for the non-transposed dimensions. Alternatively, you can call one of the `local_size_transposed' functions, which returns both the non-transposed and transposed data distribution from a single call. For example, for a 3d transform with transposed output (or input), you might call: ptrdiff_t fftw_mpi_local_size_3d_transposed( ptrdiff_t n0, ptrdiff_t n1, ptrdiff_t n2, MPI_Comm comm, ptrdiff_t *local_n0, ptrdiff_t *local_0_start, ptrdiff_t *local_n1, ptrdiff_t *local_1_start); Here, `local_n0' and `local_0_start' give the size and starting index of the `n0' dimension for the _non_-transposed data, as in the previous sections. For _transposed_ data (e.g. the output for `FFTW_MPI_TRANSPOSED_OUT'), `local_n1' and `local_1_start' give the size and starting index of the `n1' dimension, which is the first dimension of the transposed data (`n1' by `n0' by `n2'). (Note that `FFTW_MPI_TRANSPOSED_IN' is completely equivalent to performing `FFTW_MPI_TRANSPOSED_OUT' and passing the first two dimensions to the planner in reverse order, or vice versa. If you pass _both_ the `FFTW_MPI_TRANSPOSED_IN' and `FFTW_MPI_TRANSPOSED_OUT' flags, it is equivalent to swapping the first two dimensions passed to the planner and passing _neither_ flag.)  File: fftw3.info, Node: One-dimensional distributions, Prev: Transposed distributions, Up: MPI Data Distribution 6.4.4 One-dimensional distributions ----------------------------------- For one-dimensional distributed DFTs using FFTW, matters are slightly more complicated because the data distribution is more closely tied to how the algorithm works. In particular, you can no longer pass an arbitrary block size and must accept FFTW's default; also, the block sizes may be different for input and output. Also, the data distribution depends on the flags and transform direction, in order for forward and backward transforms to work correctly. ptrdiff_t fftw_mpi_local_size_1d(ptrdiff_t n0, MPI_Comm comm, int sign, unsigned flags, ptrdiff_t *local_ni, ptrdiff_t *local_i_start, ptrdiff_t *local_no, ptrdiff_t *local_o_start); This function computes the data distribution for a 1d transform of size `n0' with the given transform `sign' and `flags'. Both input and output data use block distributions. The input on the current process will consist of `local_ni' numbers starting at index `local_i_start'; e.g. if only a single process is used, then `local_ni' will be `n0' and `local_i_start' will be `0'. Similarly for the output, with `local_no' numbers starting at index `local_o_start'. The return value of `fftw_mpi_local_size_1d' will be the total number of elements to allocate on the current process (which might be slightly larger than the local size due to intermediate steps in the algorithm). As mentioned above (*note Load balancing::), the data will be divided equally among the processes if `n0' is divisible by the _square_ of the number of processes. In this case, `local_ni' will equal `local_no'. Otherwise, they may be different. For some applications, such as convolutions, the order of the output data is irrelevant. In this case, performance can be improved by specifying that the output data be stored in an FFTW-defined "scrambled" format. (In particular, this is the analogue of transposed output in the multidimensional case: scrambled output saves a communications step.) If you pass `FFTW_MPI_SCRAMBLED_OUT' in the flags, then the output is stored in this (undocumented) scrambled order. Conversely, to perform the inverse transform of data in scrambled order, pass the `FFTW_MPI_SCRAMBLED_IN' flag. In MPI FFTW, only composite sizes `n0' can be parallelized; we have not yet implemented a parallel algorithm for large prime sizes.  File: fftw3.info, Node: Multi-dimensional MPI DFTs of Real Data, Next: Other Multi-dimensional Real-data MPI Transforms, Prev: MPI Data Distribution, Up: Distributed-memory FFTW with MPI 6.5 Multi-dimensional MPI DFTs of Real Data =========================================== FFTW's MPI interface also supports multi-dimensional DFTs of real data, similar to the serial r2c and c2r interfaces. (Parallel one-dimensional real-data DFTs are not currently supported; you must use a complex transform and set the imaginary parts of the inputs to zero.) The key points to understand for r2c and c2r MPI transforms (compared to the MPI complex DFTs or the serial r2c/c2r transforms), are: * Just as for serial transforms, r2c/c2r DFTs transform n[0] x n[1] x n[2] x ... x n[d-1] real data to/from n[0] x n[1] x n[2] x ... x (n[d-1]/2 + 1) complex data: the last dimension of the complex data is cut in half (rounded down), plus one. As for the serial transforms, the sizes you pass to the `plan_dft_r2c' and `plan_dft_c2r' are the n[0] x n[1] x n[2] x ... x n[d-1] dimensions of the real data. * Although the real data is _conceptually_ n[0] x n[1] x n[2] x ... x n[d-1] , it is _physically_ stored as an n[0] x n[1] x n[2] x ... x [2 (n[d-1]/2 + 1)] array, where the last dimension has been _padded_ to make it the same size as the complex output. This is much like the in-place serial r2c/c2r interface (*note Multi-Dimensional DFTs of Real Data::), except that in MPI the padding is required even for out-of-place data. The extra padding numbers are ignored by FFTW (they are _not_ like zero-padding the transform to a larger size); they are only used to determine the data layout. * The data distribution in MPI for _both_ the real and complex data is determined by the shape of the _complex_ data. That is, you call the appropriate `local size' function for the n[0] x n[1] x n[2] x ... x (n[d-1]/2 + 1) complex data, and then use the _same_ distribution for the real data except that the last complex dimension is replaced by a (padded) real dimension of twice the length. For example suppose we are performing an out-of-place r2c transform of L x M x N real data [padded to L x M x 2(N/2+1) ], resulting in L x M x N/2+1 complex data. Similar to the example in *note 2d MPI example::, we might do something like: #include int main(int argc, char **argv) { const ptrdiff_t L = ..., M = ..., N = ...; fftw_plan plan; double *rin; fftw_complex *cout; ptrdiff_t alloc_local, local_n0, local_0_start, i, j, k; MPI_Init(&argc, &argv); fftw_mpi_init(); /* get local data size and allocate */ alloc_local = fftw_mpi_local_size_3d(L, M, N/2+1, MPI_COMM_WORLD, &local_n0, &local_0_start); rin = fftw_alloc_real(2 * alloc_local); cout = fftw_alloc_complex(alloc_local); /* create plan for out-of-place r2c DFT */ plan = fftw_mpi_plan_dft_r2c_3d(L, M, N, rin, cout, MPI_COMM_WORLD, FFTW_MEASURE); /* initialize rin to some function my_func(x,y,z) */ for (i = 0; i < local_n0; ++i) for (j = 0; j < M; ++j) for (k = 0; k < N; ++k) rin[(i*M + j) * (2*(N/2+1)) + k] = my_func(local_0_start+i, j, k); /* compute transforms as many times as desired */ fftw_execute(plan); fftw_destroy_plan(plan); MPI_Finalize(); } Note that we allocated `rin' using `fftw_alloc_real' with an argument of `2 * alloc_local': since `alloc_local' is the number of _complex_ values to allocate, the number of _real_ values is twice as many. The `rin' array is then local_n0 x M x 2(N/2+1) in row-major order, so its `(i,j,k)' element is at the index `(i*M + j) * (2*(N/2+1)) + k' (*note Multi-dimensional Array Format::). As for the complex transforms, improved performance can be obtained by specifying that the output is the transpose of the input or vice versa (*note Transposed distributions::). In our L x M x N r2c example, including `FFTW_TRANSPOSED_OUT' in the flags means that the input would be a padded L x M x 2(N/2+1) real array distributed over the `L' dimension, while the output would be a M x L x N/2+1 complex array distributed over the `M' dimension. To perform the inverse c2r transform with the same data distributions, you would use the `FFTW_TRANSPOSED_IN' flag.  File: fftw3.info, Node: Other Multi-dimensional Real-data MPI Transforms, Next: FFTW MPI Transposes, Prev: Multi-dimensional MPI DFTs of Real Data, Up: Distributed-memory FFTW with MPI 6.6 Other multi-dimensional Real-Data MPI Transforms ==================================================== FFTW's MPI interface also supports multi-dimensional `r2r' transforms of all kinds supported by the serial interface (e.g. discrete cosine and sine transforms, discrete Hartley transforms, etc.). Only multi-dimensional `r2r' transforms, not one-dimensional transforms, are currently parallelized. These are used much like the multidimensional complex DFTs discussed above, except that the data is real rather than complex, and one needs to pass an r2r transform kind (`fftw_r2r_kind') for each dimension as in the serial FFTW (*note More DFTs of Real Data::). For example, one might perform a two-dimensional L x M that is an REDFT10 (DCT-II) in the first dimension and an RODFT10 (DST-II) in the second dimension with code like: const ptrdiff_t L = ..., M = ...; fftw_plan plan; double *data; ptrdiff_t alloc_local, local_n0, local_0_start, i, j; /* get local data size and allocate */ alloc_local = fftw_mpi_local_size_2d(L, M, MPI_COMM_WORLD, &local_n0, &local_0_start); data = fftw_alloc_real(alloc_local); /* create plan for in-place REDFT10 x RODFT10 */ plan = fftw_mpi_plan_r2r_2d(L, M, data, data, MPI_COMM_WORLD, FFTW_REDFT10, FFTW_RODFT10, FFTW_MEASURE); /* initialize data to some function my_function(x,y) */ for (i = 0; i < local_n0; ++i) for (j = 0; j < M; ++j) data[i*M + j] = my_function(local_0_start + i, j); /* compute transforms, in-place, as many times as desired */ fftw_execute(plan); fftw_destroy_plan(plan); Notice that we use the same `local_size' functions as we did for complex data, only now we interpret the sizes in terms of real rather than complex values, and correspondingly use `fftw_alloc_real'.  File: fftw3.info, Node: FFTW MPI Transposes, Next: FFTW MPI Wisdom, Prev: Other Multi-dimensional Real-data MPI Transforms, Up: Distributed-memory FFTW with MPI 6.7 FFTW MPI Transposes ======================= The FFTW's MPI Fourier transforms rely on one or more _global transposition_ step for their communications. For example, the multidimensional transforms work by transforming along some dimensions, then transposing to make the first dimension local and transforming that, then transposing back. Because global transposition of a block-distributed matrix has many other potential uses besides FFTs, FFTW's transpose routines can be called directly, as documented in this section. * Menu: * Basic distributed-transpose interface:: * Advanced distributed-transpose interface:: * An improved replacement for MPI_Alltoall::  File: fftw3.info, Node: Basic distributed-transpose interface, Next: Advanced distributed-transpose interface, Prev: FFTW MPI Transposes, Up: FFTW MPI Transposes 6.7.1 Basic distributed-transpose interface ------------------------------------------- In particular, suppose that we have an `n0' by `n1' array in row-major order, block-distributed across the `n0' dimension. To transpose this into an `n1' by `n0' array block-distributed across the `n1' dimension, we would create a plan by calling the following function: fftw_plan fftw_mpi_plan_transpose(ptrdiff_t n0, ptrdiff_t n1, double *in, double *out, MPI_Comm comm, unsigned flags); The input and output arrays (`in' and `out') can be the same. The transpose is actually executed by calling `fftw_execute' on the plan, as usual. The `flags' are the usual FFTW planner flags, but support two additional flags: `FFTW_MPI_TRANSPOSED_OUT' and/or `FFTW_MPI_TRANSPOSED_IN'. What these flags indicate, for transpose plans, is that the output and/or input, respectively, are _locally_ transposed. That is, on each process input data is normally stored as a `local_n0' by `n1' array in row-major order, but for an `FFTW_MPI_TRANSPOSED_IN' plan the input data is stored as `n1' by `local_n0' in row-major order. Similarly, `FFTW_MPI_TRANSPOSED_OUT' means that the output is `n0' by `local_n1' instead of `local_n1' by `n0'. To determine the local size of the array on each process before and after the transpose, as well as the amount of storage that must be allocated, one should call `fftw_mpi_local_size_2d_transposed', just as for a 2d DFT as described in the previous section: ptrdiff_t fftw_mpi_local_size_2d_transposed (ptrdiff_t n0, ptrdiff_t n1, MPI_Comm comm, ptrdiff_t *local_n0, ptrdiff_t *local_0_start, ptrdiff_t *local_n1, ptrdiff_t *local_1_start); Again, the return value is the local storage to allocate, which in this case is the number of _real_ (`double') values rather than complex numbers as in the previous examples.  File: fftw3.info, Node: Advanced distributed-transpose interface, Next: An improved replacement for MPI_Alltoall, Prev: Basic distributed-transpose interface, Up: FFTW MPI Transposes 6.7.2 Advanced distributed-transpose interface ---------------------------------------------- The above routines are for a transpose of a matrix of numbers (of type `double'), using FFTW's default block sizes. More generally, one can perform transposes of _tuples_ of numbers, with user-specified block sizes for the input and output: fftw_plan fftw_mpi_plan_many_transpose (ptrdiff_t n0, ptrdiff_t n1, ptrdiff_t howmany, ptrdiff_t block0, ptrdiff_t block1, double *in, double *out, MPI_Comm comm, unsigned flags); In this case, one is transposing an `n0' by `n1' matrix of `howmany'-tuples (e.g. `howmany = 2' for complex numbers). The input is distributed along the `n0' dimension with block size `block0', and the `n1' by `n0' output is distributed along the `n1' dimension with block size `block1'. If `FFTW_MPI_DEFAULT_BLOCK' (0) is passed for a block size then FFTW uses its default block size. To get the local size of the data on each process, you should then call `fftw_mpi_local_size_many_transposed'.  File: fftw3.info, Node: An improved replacement for MPI_Alltoall, Prev: Advanced distributed-transpose interface, Up: FFTW MPI Transposes 6.7.3 An improved replacement for MPI_Alltoall ---------------------------------------------- We close this section by noting that FFTW's MPI transpose routines can be thought of as a generalization for the `MPI_Alltoall' function (albeit only for floating-point types), and in some circumstances can function as an improved replacement. `MPI_Alltoall' is defined by the MPI standard as: int MPI_Alltoall(void *sendbuf, int sendcount, MPI_Datatype sendtype, void *recvbuf, int recvcnt, MPI_Datatype recvtype, MPI_Comm comm); In particular, for `double*' arrays `in' and `out', consider the call: MPI_Alltoall(in, howmany, MPI_DOUBLE, out, howmany MPI_DOUBLE, comm); This is completely equivalent to: MPI_Comm_size(comm, &P); plan = fftw_mpi_plan_many_transpose(P, P, howmany, 1, 1, in, out, comm, FFTW_ESTIMATE); fftw_execute(plan); fftw_destroy_plan(plan); That is, computing a P x P transpose on `P' processes, with a block size of 1, is just a standard all-to-all communication. However, using the FFTW routine instead of `MPI_Alltoall' may have certain advantages. First of all, FFTW's routine can operate in-place (`in == out') whereas `MPI_Alltoall' can only operate out-of-place. Second, even for out-of-place plans, FFTW's routine may be faster, especially if you need to perform the all-to-all communication many times and can afford to use `FFTW_MEASURE' or `FFTW_PATIENT'. It should certainly be no slower, not including the time to create the plan, since one of the possible algorithms that FFTW uses for an out-of-place transpose _is_ simply to call `MPI_Alltoall'. However, FFTW also considers several other possible algorithms that, depending on your MPI implementation and your hardware, may be faster.  File: fftw3.info, Node: FFTW MPI Wisdom, Next: Avoiding MPI Deadlocks, Prev: FFTW MPI Transposes, Up: Distributed-memory FFTW with MPI 6.8 FFTW MPI Wisdom =================== FFTW's "wisdom" facility (*note Words of Wisdom-Saving Plans::) can be used to save MPI plans as well as to save uniprocessor plans. However, for MPI there are several unavoidable complications. First, the MPI standard does not guarantee that every process can perform file I/O (at least, not using C stdio routines)--in general, we may only assume that process 0 is capable of I/O.(1) So, if we want to export the wisdom from a single process to a file, we must first export the wisdom to a string, then send it to process 0, then write it to a file. Second, in principle we may want to have separate wisdom for every process, since in general the processes may run on different hardware even for a single MPI program. However, in practice FFTW's MPI code is designed for the case of homogeneous hardware (*note Load balancing::), and in this case it is convenient to use the same wisdom for every process. Thus, we need a mechanism to synchronize the wisdom. To address both of these problems, FFTW provides the following two functions: void fftw_mpi_broadcast_wisdom(MPI_Comm comm); void fftw_mpi_gather_wisdom(MPI_Comm comm); Given a communicator `comm', `fftw_mpi_broadcast_wisdom' will broadcast the wisdom from process 0 to all other processes. Conversely, `fftw_mpi_gather_wisdom' will collect wisdom from all processes onto process 0. (If the plans created for the same problem by different processes are not the same, `fftw_mpi_gather_wisdom' will arbitrarily choose one of the plans.) Both of these functions may result in suboptimal plans for different processes if the processes are running on non-identical hardware. Both of these functions are _collective_ calls, which means that they must be executed by all processes in the communicator. So, for example, a typical code snippet to import wisdom from a file and use it on all processes would be: { int rank; fftw_mpi_init(); MPI_Comm_rank(MPI_COMM_WORLD, &rank); if (rank == 0) fftw_import_wisdom_from_filename("mywisdom"); fftw_mpi_broadcast_wisdom(MPI_COMM_WORLD); } (Note that we must call `fftw_mpi_init' before importing any wisdom that might contain MPI plans.) Similarly, a typical code snippet to export wisdom from all processes to a file is: { int rank; fftw_mpi_gather_wisdom(MPI_COMM_WORLD); MPI_Comm_rank(MPI_COMM_WORLD, &rank); if (rank == 0) fftw_export_wisdom_to_filename("mywisdom"); } ---------- Footnotes ---------- (1) In fact, even this assumption is not technically guaranteed by the standard, although it seems to be universal in actual MPI implementations and is widely assumed by MPI-using software. Technically, you need to query the `MPI_IO' attribute of `MPI_COMM_WORLD' with `MPI_Attr_get'. If this attribute is `MPI_PROC_NULL', no I/O is possible. If it is `MPI_ANY_SOURCE', any process can perform I/O. Otherwise, it is the rank of a process that can perform I/O ... but since it is not guaranteed to yield the _same_ rank on all processes, you have to do an `MPI_Allreduce' of some kind if you want all processes to agree about which is going to do I/O. And even then, the standard only guarantees that this process can perform output, but not input. See e.g. `Parallel Programming with MPI' by P. S. Pacheco, section 8.1.3. Needless to say, in our experience virtually no MPI programmers worry about this.  File: fftw3.info, Node: Avoiding MPI Deadlocks, Next: FFTW MPI Performance Tips, Prev: FFTW MPI Wisdom, Up: Distributed-memory FFTW with MPI 6.9 Avoiding MPI Deadlocks ========================== An MPI program can _deadlock_ if one process is waiting for a message from another process that never gets sent. To avoid deadlocks when using FFTW's MPI routines, it is important to know which functions are _collective_: that is, which functions must _always_ be called in the _same order_ from _every_ process in a given communicator. (For example, `MPI_Barrier' is the canonical example of a collective function in the MPI standard.) The functions in FFTW that are _always_ collective are: every function beginning with `fftw_mpi_plan', as well as `fftw_mpi_broadcast_wisdom' and `fftw_mpi_gather_wisdom'. Also, the following functions from the ordinary FFTW interface are collective when they are applied to a plan created by an `fftw_mpi_plan' function: `fftw_execute', `fftw_destroy_plan', and `fftw_flops'.  File: fftw3.info, Node: FFTW MPI Performance Tips, Next: Combining MPI and Threads, Prev: Avoiding MPI Deadlocks, Up: Distributed-memory FFTW with MPI 6.10 FFTW MPI Performance Tips ============================== In this section, we collect a few tips on getting the best performance out of FFTW's MPI transforms. First, because of the 1d block distribution, FFTW's parallelization is currently limited by the size of the first dimension. (Multidimensional block distributions may be supported by a future version.) More generally, you should ideally arrange the dimensions so that FFTW can divide them equally among the processes. *Note Load balancing::. Second, if it is not too inconvenient, you should consider working with transposed output for multidimensional plans, as this saves a considerable amount of communications. *Note Transposed distributions::. Third, the fastest choices are generally either an in-place transform or an out-of-place transform with the `FFTW_DESTROY_INPUT' flag (which allows the input array to be used as scratch space). In-place is especially beneficial if the amount of data per process is large. Fourth, if you have multiple arrays to transform at once, rather than calling FFTW's MPI transforms several times it usually seems to be faster to interleave the data and use the advanced interface. (This groups the communications together instead of requiring separate messages for each transform.)  File: fftw3.info, Node: Combining MPI and Threads, Next: FFTW MPI Reference, Prev: FFTW MPI Performance Tips, Up: Distributed-memory FFTW with MPI 6.11 Combining MPI and Threads ============================== In certain cases, it may be advantageous to combine MPI (distributed-memory) and threads (shared-memory) parallelization. FFTW supports this, with certain caveats. For example, if you have a cluster of 4-processor shared-memory nodes, you may want to use threads within the nodes and MPI between the nodes, instead of MPI for all parallelization. In particular, it is possible to seamlessly combine the MPI FFTW routines with the multi-threaded FFTW routines (*note Multi-threaded FFTW::). However, some care must be taken in the initialization code, which should look something like this: int threads_ok; int main(int argc, char **argv) { int provided; MPI_Init_thread(&argc, &argv, MPI_THREAD_FUNNELED, &provided); threads_ok = provided >= MPI_THREAD_FUNNELED; if (threads_ok) threads_ok = fftw_init_threads(); fftw_mpi_init(); ... if (threads_ok) fftw_plan_with_nthreads(...); ... MPI_Finalize(); } First, note that instead of calling `MPI_Init', you should call `MPI_Init_threads', which is the initialization routine defined by the MPI-2 standard to indicate to MPI that your program will be multithreaded. We pass `MPI_THREAD_FUNNELED', which indicates that we will only call MPI routines from the main thread. (FFTW will launch additional threads internally, but the extra threads will not call MPI code.) (You may also pass `MPI_THREAD_SERIALIZED' or `MPI_THREAD_MULTIPLE', which requests additional multithreading support from the MPI implementation, but this is not required by FFTW.) The `provided' parameter returns what level of threads support is actually supported by your MPI implementation; this _must_ be at least `MPI_THREAD_FUNNELED' if you want to call the FFTW threads routines, so we define a global variable `threads_ok' to record this. You should only call `fftw_init_threads' or `fftw_plan_with_nthreads' if `threads_ok' is true. For more information on thread safety in MPI, see the MPI and Threads (http://www.mpi-forum.org/docs/mpi-20-html/node162.htm) section of the MPI-2 standard. Second, we must call `fftw_init_threads' _before_ `fftw_mpi_init'. This is critical for technical reasons having to do with how FFTW initializes its list of algorithms. Then, if you call `fftw_plan_with_nthreads(N)', _every_ MPI process will launch (up to) `N' threads to parallelize its transforms. For example, in the hypothetical cluster of 4-processor nodes, you might wish to launch only a single MPI process per node, and then call `fftw_plan_with_nthreads(4)' on each process to use all processors in the nodes. This may or may not be faster than simply using as many MPI processes as you have processors, however. On the one hand, using threads within a node eliminates the need for explicit message passing within the node. On the other hand, FFTW's transpose routines are not multi-threaded, and this means that the communications that do take place will not benefit from parallelization within the node. Moreover, many MPI implementations already have optimizations to exploit shared memory when it is available, so adding the multithreaded FFTW on top of this may be superfluous.  File: fftw3.info, Node: FFTW MPI Reference, Next: FFTW MPI Fortran Interface, Prev: Combining MPI and Threads, Up: Distributed-memory FFTW with MPI 6.12 FFTW MPI Reference ======================= This chapter provides a complete reference to all FFTW MPI functions, datatypes, and constants. See also *note FFTW Reference:: for information on functions and types in common with the serial interface. * Menu: * MPI Files and Data Types:: * MPI Initialization:: * Using MPI Plans:: * MPI Data Distribution Functions:: * MPI Plan Creation:: * MPI Wisdom Communication::  File: fftw3.info, Node: MPI Files and Data Types, Next: MPI Initialization, Prev: FFTW MPI Reference, Up: FFTW MPI Reference 6.12.1 MPI Files and Data Types ------------------------------- All programs using FFTW's MPI support should include its header file: #include Note that this header file includes the serial-FFTW `fftw3.h' header file, and also the `mpi.h' header file for MPI, so you need not include those files separately. You must also link to _both_ the FFTW MPI library and to the serial FFTW library. On Unix, this means adding `-lfftw3_mpi -lfftw3 -lm' at the end of the link command. Different precisions are handled as in the serial interface: *Note Precision::. That is, `fftw_' functions become `fftwf_' (in single precision) etcetera, and the libraries become `-lfftw3f_mpi -lfftw3f -lm' etcetera on Unix. Long-double precision is supported in MPI, but quad precision (`fftwq_') is not due to the lack of MPI support for this type.  File: fftw3.info, Node: MPI Initialization, Next: Using MPI Plans, Prev: MPI Files and Data Types, Up: FFTW MPI Reference 6.12.2 MPI Initialization ------------------------- Before calling any other FFTW MPI (`fftw_mpi_') function, and before importing any wisdom for MPI problems, you must call: void fftw_mpi_init(void); If FFTW threads support is used, however, `fftw_mpi_init' should be called _after_ `fftw_init_threads' (*note Combining MPI and Threads::). Calling `fftw_mpi_init' additional times (before `fftw_mpi_cleanup') has no effect. If you want to deallocate all persistent data and reset FFTW to the pristine state it was in when you started your program, you can call: void fftw_mpi_cleanup(void); (This calls `fftw_cleanup', so you need not call the serial cleanup routine too, although it is safe to do so.) After calling `fftw_mpi_cleanup', all existing plans become undefined, and you should not attempt to execute or destroy them. You must call `fftw_mpi_init' again after `fftw_mpi_cleanup' if you want to resume using the MPI FFTW routines.  File: fftw3.info, Node: Using MPI Plans, Next: MPI Data Distribution Functions, Prev: MPI Initialization, Up: FFTW MPI Reference 6.12.3 Using MPI Plans ---------------------- Once an MPI plan is created, you can execute and destroy it using `fftw_execute', `fftw_destroy_plan', and the other functions in the serial interface that operate on generic plans (*note Using Plans::). The `fftw_execute' and `fftw_destroy_plan' functions, applied to MPI plans, are _collective_ calls: they must be called for all processes in the communicator that was used to create the plan. You must _not_ use the serial new-array plan-execution functions `fftw_execute_dft' and so on (*note New-array Execute Functions::) with MPI plans. Such functions are specialized to the problem type, and there are specific new-array execute functions for MPI plans: void fftw_mpi_execute_dft(fftw_plan p, fftw_complex *in, fftw_complex *out); void fftw_mpi_execute_dft_r2c(fftw_plan p, double *in, fftw_complex *out); void fftw_mpi_execute_dft_c2r(fftw_plan p, fftw_complex *in, double *out); void fftw_mpi_execute_r2r(fftw_plan p, double *in, double *out); These functions have the same restrictions as those of the serial new-array execute functions. They are _always_ safe to apply to the _same_ `in' and `out' arrays that were used to create the plan. They can only be applied to new arrarys if those arrays have the same types, dimensions, in-placeness, and alignment as the original arrays, where the best way to ensure the same alignment is to use FFTW's `fftw_malloc' and related allocation functions for all arrays (*note Memory Allocation::). Note that distributed transposes (*note FFTW MPI Transposes::) use `fftw_mpi_execute_r2r', since they count as rank-zero r2r plans from FFTW's perspective.  File: fftw3.info, Node: MPI Data Distribution Functions, Next: MPI Plan Creation, Prev: Using MPI Plans, Up: FFTW MPI Reference 6.12.4 MPI Data Distribution Functions -------------------------------------- As described above (*note MPI Data Distribution::), in order to allocate your arrays, _before_ creating a plan, you must first call one of the following routines to determine the required allocation size and the portion of the array locally stored on a given process. The `MPI_Comm' communicator passed here must be equivalent to the communicator used below for plan creation. The basic interface for multidimensional transforms consists of the functions: ptrdiff_t fftw_mpi_local_size_2d(ptrdiff_t n0, ptrdiff_t n1, MPI_Comm comm, ptrdiff_t *local_n0, ptrdiff_t *local_0_start); ptrdiff_t fftw_mpi_local_size_3d(ptrdiff_t n0, ptrdiff_t n1, ptrdiff_t n2, MPI_Comm comm, ptrdiff_t *local_n0, ptrdiff_t *local_0_start); ptrdiff_t fftw_mpi_local_size(int rnk, const ptrdiff_t *n, MPI_Comm comm, ptrdiff_t *local_n0, ptrdiff_t *local_0_start); ptrdiff_t fftw_mpi_local_size_2d_transposed(ptrdiff_t n0, ptrdiff_t n1, MPI_Comm comm, ptrdiff_t *local_n0, ptrdiff_t *local_0_start, ptrdiff_t *local_n1, ptrdiff_t *local_1_start); ptrdiff_t fftw_mpi_local_size_3d_transposed(ptrdiff_t n0, ptrdiff_t n1, ptrdiff_t n2, MPI_Comm comm, ptrdiff_t *local_n0, ptrdiff_t *local_0_start, ptrdiff_t *local_n1, ptrdiff_t *local_1_start); ptrdiff_t fftw_mpi_local_size_transposed(int rnk, const ptrdiff_t *n, MPI_Comm comm, ptrdiff_t *local_n0, ptrdiff_t *local_0_start, ptrdiff_t *local_n1, ptrdiff_t *local_1_start); These functions return the number of elements to allocate (complex numbers for DFT/r2c/c2r plans, real numbers for r2r plans), whereas the `local_n0' and `local_0_start' return the portion (`local_0_start' to `local_0_start + local_n0 - 1') of the first dimension of an n[0] x n[1] x n[2] x ... x n[d-1] array that is stored on the local process. *Note Basic and advanced distribution interfaces::. For `FFTW_MPI_TRANSPOSED_OUT' plans, the `_transposed' variants are useful in order to also return the local portion of the first dimension in the n[1] x n[0] x n[2] x ... x n[d-1] transposed output. *Note Transposed distributions::. The advanced interface for multidimensional transforms is: ptrdiff_t fftw_mpi_local_size_many(int rnk, const ptrdiff_t *n, ptrdiff_t howmany, ptrdiff_t block0, MPI_Comm comm, ptrdiff_t *local_n0, ptrdiff_t *local_0_start); ptrdiff_t fftw_mpi_local_size_many_transposed(int rnk, const ptrdiff_t *n, ptrdiff_t howmany, ptrdiff_t block0, ptrdiff_t block1, MPI_Comm comm, ptrdiff_t *local_n0, ptrdiff_t *local_0_start, ptrdiff_t *local_n1, ptrdiff_t *local_1_start); These differ from the basic interface in only two ways. First, they allow you to specify block sizes `block0' and `block1' (the latter for the transposed output); you can pass `FFTW_MPI_DEFAULT_BLOCK' to use FFTW's default block size as in the basic interface. Second, you can pass a `howmany' parameter, corresponding to the advanced planning interface below: this is for transforms of contiguous `howmany'-tuples of numbers (`howmany = 1' in the basic interface). The corresponding basic and advanced routines for one-dimensional transforms (currently only complex DFTs) are: ptrdiff_t fftw_mpi_local_size_1d( ptrdiff_t n0, MPI_Comm comm, int sign, unsigned flags, ptrdiff_t *local_ni, ptrdiff_t *local_i_start, ptrdiff_t *local_no, ptrdiff_t *local_o_start); ptrdiff_t fftw_mpi_local_size_many_1d( ptrdiff_t n0, ptrdiff_t howmany, MPI_Comm comm, int sign, unsigned flags, ptrdiff_t *local_ni, ptrdiff_t *local_i_start, ptrdiff_t *local_no, ptrdiff_t *local_o_start); As above, the return value is the number of elements to allocate (complex numbers, for complex DFTs). The `local_ni' and `local_i_start' arguments return the portion (`local_i_start' to `local_i_start + local_ni - 1') of the 1d array that is stored on this process for the transform _input_, and `local_no' and `local_o_start' are the corresponding quantities for the input. The `sign' (`FFTW_FORWARD' or `FFTW_BACKWARD') and `flags' must match the arguments passed when creating a plan. Although the inputs and outputs have different data distributions in general, it is guaranteed that the _output_ data distribution of an `FFTW_FORWARD' plan will match the _input_ data distribution of an `FFTW_BACKWARD' plan and vice versa; similarly for the `FFTW_MPI_SCRAMBLED_OUT' and `FFTW_MPI_SCRAMBLED_IN' flags. *Note One-dimensional distributions::.  File: fftw3.info, Node: MPI Plan Creation, Next: MPI Wisdom Communication, Prev: MPI Data Distribution Functions, Up: FFTW MPI Reference 6.12.5 MPI Plan Creation ------------------------ Complex-data MPI DFTs ..................... Plans for complex-data DFTs (*note 2d MPI example::) are created by: fftw_plan fftw_mpi_plan_dft_1d(ptrdiff_t n0, fftw_complex *in, fftw_complex *out, MPI_Comm comm, int sign, unsigned flags); fftw_plan fftw_mpi_plan_dft_2d(ptrdiff_t n0, ptrdiff_t n1, fftw_complex *in, fftw_complex *out, MPI_Comm comm, int sign, unsigned flags); fftw_plan fftw_mpi_plan_dft_3d(ptrdiff_t n0, ptrdiff_t n1, ptrdiff_t n2, fftw_complex *in, fftw_complex *out, MPI_Comm comm, int sign, unsigned flags); fftw_plan fftw_mpi_plan_dft(int rnk, const ptrdiff_t *n, fftw_complex *in, fftw_complex *out, MPI_Comm comm, int sign, unsigned flags); fftw_plan fftw_mpi_plan_many_dft(int rnk, const ptrdiff_t *n, ptrdiff_t howmany, ptrdiff_t block, ptrdiff_t tblock, fftw_complex *in, fftw_complex *out, MPI_Comm comm, int sign, unsigned flags); These are similar to their serial counterparts (*note Complex DFTs::) in specifying the dimensions, sign, and flags of the transform. The `comm' argument gives an MPI communicator that specifies the set of processes to participate in the transform; plan creation is a collective function that must be called for all processes in the communicator. The `in' and `out' pointers refer only to a portion of the overall transform data (*note MPI Data Distribution::) as specified by the `local_size' functions in the previous section. Unless `flags' contains `FFTW_ESTIMATE', these arrays are overwritten during plan creation as for the serial interface. For multi-dimensional transforms, any dimensions `> 1' are supported; for one-dimensional transforms, only composite (non-prime) `n0' are currently supported (unlike the serial FFTW). Requesting an unsupported transform size will yield a `NULL' plan. (As in the serial interface, highly composite sizes generally yield the best performance.) The advanced-interface `fftw_mpi_plan_many_dft' additionally allows you to specify the block sizes for the first dimension (`block') of the n[0] x n[1] x n[2] x ... x n[d-1] input data and the first dimension (`tblock') of the n[1] x n[0] x n[2] x ... x n[d-1] transposed data (at intermediate steps of the transform, and for the output if `FFTW_TRANSPOSED_OUT' is specified in `flags'). These must be the same block sizes as were passed to the corresponding `local_size' function; you can pass `FFTW_MPI_DEFAULT_BLOCK' to use FFTW's default block size as in the basic interface. Also, the `howmany' parameter specifies that the transform is of contiguous `howmany'-tuples rather than individual complex numbers; this corresponds to the same parameter in the serial advanced interface (*note Advanced Complex DFTs::) with `stride = howmany' and `dist = 1'. MPI flags ......... The `flags' can be any of those for the serial FFTW (*note Planner Flags::), and in addition may include one or more of the following MPI-specific flags, which improve performance at the cost of changing the output or input data formats. * `FFTW_MPI_SCRAMBLED_OUT', `FFTW_MPI_SCRAMBLED_IN': valid for 1d transforms only, these flags indicate that the output/input of the transform are in an undocumented "scrambled" order. A forward `FFTW_MPI_SCRAMBLED_OUT' transform can be inverted by a backward `FFTW_MPI_SCRAMBLED_IN' (times the usual 1/N normalization). *Note One-dimensional distributions::. * `FFTW_MPI_TRANSPOSED_OUT', `FFTW_MPI_TRANSPOSED_IN': valid for multidimensional (`rnk > 1') transforms only, these flags specify that the output or input of an n[0] x n[1] x n[2] x ... x n[d-1] transform is transposed to n[1] x n[0] x n[2] x ... x n[d-1] . *Note Transposed distributions::. Real-data MPI DFTs .................. Plans for real-input/output (r2c/c2r) DFTs (*note Multi-dimensional MPI DFTs of Real Data::) are created by: fftw_plan fftw_mpi_plan_dft_r2c_2d(ptrdiff_t n0, ptrdiff_t n1, double *in, fftw_complex *out, MPI_Comm comm, unsigned flags); fftw_plan fftw_mpi_plan_dft_r2c_2d(ptrdiff_t n0, ptrdiff_t n1, double *in, fftw_complex *out, MPI_Comm comm, unsigned flags); fftw_plan fftw_mpi_plan_dft_r2c_3d(ptrdiff_t n0, ptrdiff_t n1, ptrdiff_t n2, double *in, fftw_complex *out, MPI_Comm comm, unsigned flags); fftw_plan fftw_mpi_plan_dft_r2c(int rnk, const ptrdiff_t *n, double *in, fftw_complex *out, MPI_Comm comm, unsigned flags); fftw_plan fftw_mpi_plan_dft_c2r_2d(ptrdiff_t n0, ptrdiff_t n1, fftw_complex *in, double *out, MPI_Comm comm, unsigned flags); fftw_plan fftw_mpi_plan_dft_c2r_2d(ptrdiff_t n0, ptrdiff_t n1, fftw_complex *in, double *out, MPI_Comm comm, unsigned flags); fftw_plan fftw_mpi_plan_dft_c2r_3d(ptrdiff_t n0, ptrdiff_t n1, ptrdiff_t n2, fftw_complex *in, double *out, MPI_Comm comm, unsigned flags); fftw_plan fftw_mpi_plan_dft_c2r(int rnk, const ptrdiff_t *n, fftw_complex *in, double *out, MPI_Comm comm, unsigned flags); Similar to the serial interface (*note Real-data DFTs::), these transform logically n[0] x n[1] x n[2] x ... x n[d-1] real data to/from n[0] x n[1] x n[2] x ... x (n[d-1]/2 + 1) complex data, representing the non-redundant half of the conjugate-symmetry output of a real-input DFT (*note Multi-dimensional Transforms::). However, the real array must be stored within a padded n[0] x n[1] x n[2] x ... x [2 (n[d-1]/2 + 1)] array (much like the in-place serial r2c transforms, but here for out-of-place transforms as well). Currently, only multi-dimensional (`rnk > 1') r2c/c2r transforms are supported (requesting a plan for `rnk = 1' will yield `NULL'). As explained above (*note Multi-dimensional MPI DFTs of Real Data::), the data distribution of both the real and complex arrays is given by the `local_size' function called for the dimensions of the _complex_ array. Similar to the other planning functions, the input and output arrays are overwritten when the plan is created except in `FFTW_ESTIMATE' mode. As for the complex DFTs above, there is an advance interface that allows you to manually specify block sizes and to transform contiguous `howmany'-tuples of real/complex numbers: fftw_plan fftw_mpi_plan_many_dft_r2c (int rnk, const ptrdiff_t *n, ptrdiff_t howmany, ptrdiff_t iblock, ptrdiff_t oblock, double *in, fftw_complex *out, MPI_Comm comm, unsigned flags); fftw_plan fftw_mpi_plan_many_dft_c2r (int rnk, const ptrdiff_t *n, ptrdiff_t howmany, ptrdiff_t iblock, ptrdiff_t oblock, fftw_complex *in, double *out, MPI_Comm comm, unsigned flags); MPI r2r transforms .................. There are corresponding plan-creation routines for r2r transforms (*note More DFTs of Real Data::), currently supporting multidimensional (`rnk > 1') transforms only (`rnk = 1' will yield a `NULL' plan): fftw_plan fftw_mpi_plan_r2r_2d(ptrdiff_t n0, ptrdiff_t n1, double *in, double *out, MPI_Comm comm, fftw_r2r_kind kind0, fftw_r2r_kind kind1, unsigned flags); fftw_plan fftw_mpi_plan_r2r_3d(ptrdiff_t n0, ptrdiff_t n1, ptrdiff_t n2, double *in, double *out, MPI_Comm comm, fftw_r2r_kind kind0, fftw_r2r_kind kind1, fftw_r2r_kind kind2, unsigned flags); fftw_plan fftw_mpi_plan_r2r(int rnk, const ptrdiff_t *n, double *in, double *out, MPI_Comm comm, const fftw_r2r_kind *kind, unsigned flags); fftw_plan fftw_mpi_plan_many_r2r(int rnk, const ptrdiff_t *n, ptrdiff_t iblock, ptrdiff_t oblock, double *in, double *out, MPI_Comm comm, const fftw_r2r_kind *kind, unsigned flags); The parameters are much the same as for the complex DFTs above, except that the arrays are of real numbers (and hence the outputs of the `local_size' data-distribution functions should be interpreted as counts of real rather than complex numbers). Also, the `kind' parameters specify the r2r kinds along each dimension as for the serial interface (*note Real-to-Real Transform Kinds::). *Note Other Multi-dimensional Real-data MPI Transforms::. MPI transposition ................. FFTW also provides routines to plan a transpose of a distributed `n0' by `n1' array of real numbers, or an array of `howmany'-tuples of real numbers with specified block sizes (*note FFTW MPI Transposes::): fftw_plan fftw_mpi_plan_transpose(ptrdiff_t n0, ptrdiff_t n1, double *in, double *out, MPI_Comm comm, unsigned flags); fftw_plan fftw_mpi_plan_many_transpose (ptrdiff_t n0, ptrdiff_t n1, ptrdiff_t howmany, ptrdiff_t block0, ptrdiff_t block1, double *in, double *out, MPI_Comm comm, unsigned flags); These plans are used with the `fftw_mpi_execute_r2r' new-array execute function (*note Using MPI Plans::), since they count as (rank zero) r2r plans from FFTW's perspective.  File: fftw3.info, Node: MPI Wisdom Communication, Prev: MPI Plan Creation, Up: FFTW MPI Reference 6.12.6 MPI Wisdom Communication ------------------------------- To facilitate synchronizing wisdom among the different MPI processes, we provide two functions: void fftw_mpi_gather_wisdom(MPI_Comm comm); void fftw_mpi_broadcast_wisdom(MPI_Comm comm); The `fftw_mpi_gather_wisdom' function gathers all wisdom in the given communicator `comm' to the process of rank 0 in the communicator: that process obtains the union of all wisdom on all the processes. As a side effect, some other processes will gain additional wisdom from other processes, but only process 0 will gain the complete union. The `fftw_mpi_broadcast_wisdom' does the reverse: it exports wisdom from process 0 in `comm' to all other processes in the communicator, replacing any wisdom they currently have. *Note FFTW MPI Wisdom::.  File: fftw3.info, Node: FFTW MPI Fortran Interface, Prev: FFTW MPI Reference, Up: Distributed-memory FFTW with MPI 6.13 FFTW MPI Fortran Interface =============================== The FFTW MPI interface is callable from modern Fortran compilers supporting the Fortran 2003 `iso_c_binding' standard for calling C functions. As described in *note Calling FFTW from Modern Fortran::, this means that you can directly call FFTW's C interface from Fortran with only minor changes in syntax. There are, however, a few things specific to the MPI interface to keep in mind: * Instead of including `fftw3.f03' as in *note Overview of Fortran interface::, you should `include 'fftw3-mpi.f03'' (after `use, intrinsic :: iso_c_binding' as before). The `fftw3-mpi.f03' file includes `fftw3.f03', so you should _not_ `include' them both yourself. (You will also want to include the MPI header file, usually via `include 'mpif.h'' or similar, although though this is not needed by `fftw3-mpi.f03' per se.) (To use the `fftwl_' `long double' extended-precision routines in supporting compilers, you should include `fftw3f-mpi.f03' in _addition_ to `fftw3-mpi.f03'. *Note Extended and quadruple precision in Fortran::.) * Because of the different storage conventions between C and Fortran, you reverse the order of your array dimensions when passing them to FFTW (*note Reversing array dimensions::). This is merely a difference in notation and incurs no performance overhead. However, it means that, whereas in C the _first_ dimension is distributed, in Fortran the _last_ dimension of your array is distributed. * In Fortran, communicators are stored as `integer' types; there is no `MPI_Comm' type, nor is there any way to access a C `MPI_Comm'. Fortunately, this is taken care of for you by the FFTW Fortran interface: whenever the C interface expects an `MPI_Comm' type, you should pass the Fortran communicator as an `integer'.(1) * Because you need to call the `local_size' function to find out how much space to allocate, and this may be _larger_ than the local portion of the array (*note MPI Data Distribution::), you should _always_ allocate your arrays dynamically using FFTW's allocation routines as described in *note Allocating aligned memory in Fortran::. (Coincidentally, this also provides the best performance by guaranteeding proper data alignment.) * Because all sizes in the MPI FFTW interface are declared as `ptrdiff_t' in C, you should use `integer(C_INTPTR_T)' in Fortran (*note FFTW Fortran type reference::). * In Fortran, because of the language semantics, we generally recommend using the new-array execute functions for all plans, even in the common case where you are executing the plan on the same arrays for which the plan was created (*note Plan execution in Fortran::). However, note that in the MPI interface these functions are changed: `fftw_execute_dft' becomes `fftw_mpi_execute_dft', etcetera. *Note Using MPI Plans::. For example, here is a Fortran code snippet to perform a distributed L x M complex DFT in-place. (This assumes you have already initialized MPI with `MPI_init' and have also performed `call fftw_mpi_init'.) use, intrinsic :: iso_c_binding include 'fftw3-mpi.f03' integer(C_INTPTR_T), parameter :: L = ... integer(C_INTPTR_T), parameter :: M = ... type(C_PTR) :: plan, cdata complex(C_DOUBLE_COMPLEX), pointer :: data(:,:) integer(C_INTPTR_T) :: i, j, alloc_local, local_M, local_j_offset ! get local data size and allocate (note dimension reversal) alloc_local = fftw_mpi_local_size_2d(M, L, MPI_COMM_WORLD, & local_M, local_j_offset) cdata = fftw_alloc_complex(alloc_local) call c_f_pointer(cdata, data, [L,local_M]) ! create MPI plan for in-place forward DFT (note dimension reversal) plan = fftw_mpi_plan_dft_2d(M, L, data, data, MPI_COMM_WORLD, & FFTW_FORWARD, FFTW_MEASURE) ! initialize data to some function my_function(i,j) do j = 1, local_M do i = 1, L data(i, j) = my_function(i, j + local_j_offset) end do end do ! compute transform (as many times as desired) call fftw_mpi_execute_dft(plan, data, data) call fftw_destroy_plan(plan) call fftw_free(cdata) Note that when we called `fftw_mpi_local_size_2d' and `fftw_mpi_plan_dft_2d' with the dimensions in reversed order, since a L x M Fortran array is viewed by FFTW in C as a M x L array. This means that the array was distributed over the `M' dimension, the local portion of which is a L x local_M array in Fortran. (You must _not_ use an `allocate' statement to allocate an L x local_M array, however; you must allocate `alloc_local' complex numbers, which may be greater than `L * local_M', in order to reserve space for intermediate steps of the transform.) Finally, we mention that because C's array indices are zero-based, the `local_j_offset' argument can conveniently be interpreted as an offset in the 1-based `j' index (rather than as a starting index as in C). If instead you had used the `ior(FFTW_MEASURE, FFTW_MPI_TRANSPOSED_OUT)' flag, the output of the transform would be a transposed M x local_L array, associated with the _same_ `cdata' allocation (since the transform is in-place), and which you could declare with: complex(C_DOUBLE_COMPLEX), pointer :: tdata(:,:) ... call c_f_pointer(cdata, tdata, [M,local_L]) where `local_L' would have been obtained by changing the `fftw_mpi_local_size_2d' call to: alloc_local = fftw_mpi_local_size_2d_transposed(M, L, MPI_COMM_WORLD, & local_M, local_j_offset, local_L, local_i_offset) ---------- Footnotes ---------- (1) Technically, this is because you aren't actually calling the C functions directly. You are calling wrapper functions that translate the communicator with `MPI_Comm_f2c' before calling the ordinary C interface. This is all done transparently, however, since the `fftw3-mpi.f03' interface file renames the wrappers so that they are called in Fortran with the same names as the C interface functions.  File: fftw3.info, Node: Calling FFTW from Modern Fortran, Next: Calling FFTW from Legacy Fortran, Prev: Distributed-memory FFTW with MPI, Up: Top 7 Calling FFTW from Modern Fortran ********************************** Fortran 2003 standardized ways for Fortran code to call C libraries, and this allows us to support a direct translation of the FFTW C API into Fortran. Compared to the legacy Fortran 77 interface (*note Calling FFTW from Legacy Fortran::), this direct interface offers many advantages, especially compile-time type-checking and aligned memory allocation. As of this writing, support for these C interoperability features seems widespread, having been implemented in nearly all major Fortran compilers (e.g. GNU, Intel, IBM, Oracle/Solaris, Portland Group, NAG). This chapter documents that interface. For the most part, since this interface allows Fortran to call the C interface directly, the usage is identical to C translated to Fortran syntax. However, there are a few subtle points such as memory allocation, wisdom, and data types that deserve closer attention. * Menu: * Overview of Fortran interface:: * Reversing array dimensions:: * FFTW Fortran type reference:: * Plan execution in Fortran:: * Allocating aligned memory in Fortran:: * Accessing the wisdom API from Fortran:: * Defining an FFTW module::  File: fftw3.info, Node: Overview of Fortran interface, Next: Reversing array dimensions, Prev: Calling FFTW from Modern Fortran, Up: Calling FFTW from Modern Fortran 7.1 Overview of Fortran interface ================================= FFTW provides a file `fftw3.f03' that defines Fortran 2003 interfaces for all of its C routines, except for the MPI routines described elsewhere, which can be found in the same directory as `fftw3.h' (the C header file). In any Fortran subroutine where you want to use FFTW functions, you should begin with: use, intrinsic :: iso_c_binding include 'fftw3.f03' This includes the interface definitions and the standard `iso_c_binding' module (which defines the equivalents of C types). You can also put the FFTW functions into a module if you prefer (*note Defining an FFTW module::). At this point, you can now call anything in the FFTW C interface directly, almost exactly as in C other than minor changes in syntax. For example: type(C_PTR) :: plan complex(C_DOUBLE_COMPLEX), dimension(1024,1000) :: in, out plan = fftw_plan_dft_2d(1000,1024, in,out, FFTW_FORWARD,FFTW_ESTIMATE) ... call fftw_execute_dft(plan, in, out) ... call fftw_destroy_plan(plan) A few important things to keep in mind are: * FFTW plans are `type(C_PTR)'. Other C types are mapped in the obvious way via the `iso_c_binding' standard: `int' turns into `integer(C_INT)', `fftw_complex' turns into `complex(C_DOUBLE_COMPLEX)', `double' turns into `real(C_DOUBLE)', and so on. *Note FFTW Fortran type reference::. * Functions in C become functions in Fortran if they have a return value, and subroutines in Fortran otherwise. * The ordering of the Fortran array dimensions must be _reversed_ when they are passed to the FFTW plan creation, thanks to differences in array indexing conventions (*note Multi-dimensional Array Format::). This is _unlike_ the legacy Fortran interface (*note Fortran-interface routines::), which reversed the dimensions for you. *Note Reversing array dimensions::. * Using ordinary Fortran array declarations like this works, but may yield suboptimal performance because the data may not be not aligned to exploit SIMD instructions on modern proessors (*note SIMD alignment and fftw_malloc::). Better performance will often be obtained by allocating with `fftw_alloc'. *Note Allocating aligned memory in Fortran::. * Similar to the legacy Fortran interface (*note FFTW Execution in Fortran::), we currently recommend _not_ using `fftw_execute' but rather using the more specialized functions like `fftw_execute_dft' (*note New-array Execute Functions::). However, you should execute the plan on the `same arrays' as the ones for which you created the plan, unless you are especially careful. *Note Plan execution in Fortran::. To prevent you from using `fftw_execute' by mistake, the `fftw3.f03' file does not provide an `fftw_execute' interface declaration. * Multiple planner flags are combined with `ior' (equivalent to `|' in C). e.g. `FFTW_MEASURE | FFTW_DESTROY_INPUT' becomes `ior(FFTW_MEASURE, FFTW_DESTROY_INPUT)'. (You can also use `+' as long as you don't try to include a given flag more than once.) * Menu: * Extended and quadruple precision in Fortran::  File: fftw3.info, Node: Extended and quadruple precision in Fortran, Prev: Overview of Fortran interface, Up: Overview of Fortran interface 7.1.1 Extended and quadruple precision in Fortran ------------------------------------------------- If FFTW is compiled in `long double' (extended) precision (*note Installation and Customization::), you may be able to call the resulting `fftwl_' routines (*note Precision::) from Fortran if your compiler supports the `C_LONG_DOUBLE_COMPLEX' type code. Because some Fortran compilers do not support `C_LONG_DOUBLE_COMPLEX', the `fftwl_' declarations are segregated into a separate interface file `fftw3l.f03', which you should include _in addition_ to `fftw3.f03' (which declares precision-independent `FFTW_' constants): use, intrinsic :: iso_c_binding include 'fftw3.f03' include 'fftw3l.f03' We also support using the nonstandard `__float128' quadruple-precision type provided by recent versions of `gcc' on 32- and 64-bit x86 hardware (*note Installation and Customization::), using the corresponding `real(16)' and `complex(16)' types supported by `gfortran'. The quadruple-precision `fftwq_' functions (*note Precision::) are declared in a `fftw3q.f03' interface file, which should be included in addition to `fftw3l.f03', as above. You should also link with `-lfftw3q -lquadmath -lm' as in C.  File: fftw3.info, Node: Reversing array dimensions, Next: FFTW Fortran type reference, Prev: Overview of Fortran interface, Up: Calling FFTW from Modern Fortran 7.2 Reversing array dimensions ============================== A minor annoyance in calling FFTW from Fortran is that FFTW's array dimensions are defined in the C convention (row-major order), while Fortran's array dimensions are the opposite convention (column-major order). *Note Multi-dimensional Array Format::. This is just a bookkeeping difference, with no effect on performance. The only consequence of this is that, whenever you create an FFTW plan for a multi-dimensional transform, you must always _reverse the ordering of the dimensions_. For example, consider the three-dimensional (L x M x N ) arrays: complex(C_DOUBLE_COMPLEX), dimension(L,M,N) :: in, out To plan a DFT for these arrays using `fftw_plan_dft_3d', you could do: plan = fftw_plan_dft_3d(N,M,L, in,out, FFTW_FORWARD,FFTW_ESTIMATE) That is, from FFTW's perspective this is a N x M x L array. _No data transposition need occur_, as this is _only notation_. Similarly, to use the more generic routine `fftw_plan_dft' with the same arrays, you could do: integer(C_INT), dimension(3) :: n = [N,M,L] plan = fftw_plan_dft_3d(3, n, in,out, FFTW_FORWARD,FFTW_ESTIMATE) Note, by the way, that this is different from the legacy Fortran interface (*note Fortran-interface routines::), which automatically reverses the order of the array dimension for you. Here, you are calling the C interface directly, so there is no "translation" layer. An important thing to keep in mind is the implication of this for multidimensional real-to-complex transforms (*note Multi-Dimensional DFTs of Real Data::). In C, a multidimensional real-to-complex DFT chops the last dimension roughly in half (N x M x L real input goes to N x M x L/2+1 complex output). In Fortran, because the array dimension notation is reversed, the _first_ dimension of the complex data is chopped roughly in half. For example consider the `r2c' transform of L x M x N real input in Fortran: type(C_PTR) :: plan real(C_DOUBLE), dimension(L,M,N) :: in complex(C_DOUBLE_COMPLEX), dimension(L/2+1,M,N) :: out plan = fftw_plan_dft_r2c_3d(N,M,L, in,out, FFTW_ESTIMATE) ... call fftw_execute_dft_r2c(plan, in, out) Alternatively, for an in-place r2c transform, as described in the C documentation we must _pad_ the _first_ dimension of the real input with an extra two entries (which are ignored by FFTW) so as to leave enough space for the complex output. The input is _allocated_ as a 2[L/2+1] x M x N array, even though only L x M x N of it is actually used. In this example, we will allocate the array as a pointer type, using `fftw_alloc' to ensure aligned memory for maximum performance (*note Allocating aligned memory in Fortran::); this also makes it easy to reference the same memory as both a real array and a complex array. real(C_DOUBLE), pointer :: in(:,:,:) complex(C_DOUBLE_COMPLEX), pointer :: out(:,:,:) type(C_PTR) :: plan, data data = fftw_alloc_complex(int((L/2+1) * M * N, C_SIZE_T)) call c_f_pointer(data, in, [2*(L/2+1),M,N]) call c_f_pointer(data, out, [L/2+1,M,N]) plan = fftw_plan_dft_r2c_3d(N,M,L, in,out, FFTW_ESTIMATE) ... call fftw_execute_dft_r2c(plan, in, out) ... call fftw_destroy_plan(plan) call fftw_free(data)  File: fftw3.info, Node: FFTW Fortran type reference, Next: Plan execution in Fortran, Prev: Reversing array dimensions, Up: Calling FFTW from Modern Fortran 7.3 FFTW Fortran type reference =============================== The following are the most important type correspondences between the C interface and Fortran: * Plans (`fftw_plan' and variants) are `type(C_PTR)' (i.e. an opaque pointer). * The C floating-point types `double', `float', and `long double' correspond to `real(C_DOUBLE)', `real(C_FLOAT)', and `real(C_LONG_DOUBLE)', respectively. The C complex types `fftw_complex', `fftwf_complex', and `fftwl_complex' correspond in Fortran to `complex(C_DOUBLE_COMPLEX)', `complex(C_FLOAT_COMPLEX)', and `complex(C_LONG_DOUBLE_COMPLEX)', respectively. Just as in C (*note Precision::), the FFTW subroutines and types are prefixed with `fftw_', `fftwf_', and `fftwl_' for the different precisions, and link to different libraries (`-lfftw3', `-lfftw3f', and `-lfftw3l' on Unix), but use the _same_ include file `fftw3.f03' and the _same_ constants (all of which begin with `FFTW_'). The exception is `long double' precision, for which you should _also_ include `fftw3l.f03' (*note Extended and quadruple precision in Fortran::). * The C integer types `int' and `unsigned' (used for planner flags) become `integer(C_INT)'. The C integer type `ptrdiff_t' (e.g. in the *note 64-bit Guru Interface::) becomes `integer(C_INTPTR_T)', and `size_t' (in `fftw_malloc' etc.) becomes `integer(C_SIZE_T)'. * The `fftw_r2r_kind' type (*note Real-to-Real Transform Kinds::) becomes `integer(C_FFTW_R2R_KIND)'. The various constant values of the C enumerated type (`FFTW_R2HC' etc.) become simply integer constants of the same names in Fortran. * Numeric array pointer arguments (e.g. `double *') become `dimension(*), intent(out)' arrays of the same type, or `dimension(*), intent(in)' if they are pointers to constant data (e.g. `const int *'). There are a few exceptions where numeric pointers refer to scalar outputs (e.g. for `fftw_flops'), in which case they are `intent(out)' scalar arguments in Fortran too. For the new-array execute functions (*note New-array Execute Functions::), the input arrays are declared `dimension(*), intent(inout)', since they can be modified in the case of in-place or `FFTW_DESTROY_INPUT' transforms. * Pointer _return_ values (e.g `double *') become `type(C_PTR)'. (If they are pointers to arrays, as for `fftw_alloc_real', you can convert them back to Fortran array pointers with the standard intrinsic function `c_f_pointer'.) * The `fftw_iodim' type in the guru interface (*note Guru vector and transform sizes::) becomes `type(fftw_iodim)' in Fortran, a derived data type (the Fortran analogue of C's `struct') with three `integer(C_INT)' components: `n', `is', and `os', with the same meanings as in C. The `fftw_iodim64' type in the 64-bit guru interface (*note 64-bit Guru Interface::) is the same, except that its components are of type `integer(C_INTPTR_T)'. * Using the wisdom import/export functions from Fortran is a bit tricky, and is discussed in *note Accessing the wisdom API from Fortran::. In brief, the `FILE *' arguments map to `type(C_PTR)', `const char *' to `character(C_CHAR), dimension(*), intent(in)' (null-terminated!), and the generic read-char/write-char functions map to `type(C_FUNPTR)'. You may be wondering if you need to search-and-replace `real(kind(0.0d0))' (or whatever your favorite Fortran spelling of "double precision" is) with `real(C_DOUBLE)' everywhere in your program, and similarly for `complex' and `integer' types. The answer is no; you can still use your existing types. As long as these types match their C counterparts, things should work without a hitch. The worst that can happen, e.g. in the (unlikely) event of a system where `real(kind(0.0d0))' is different from `real(C_DOUBLE)', is that the compiler will give you a type-mismatch error. That is, if you don't use the `iso_c_binding' kinds you need to accept at least the theoretical possibility of having to change your code in response to compiler errors on some future machine, but you don't need to worry about silently compiling incorrect code that yields runtime errors.  File: fftw3.info, Node: Plan execution in Fortran, Next: Allocating aligned memory in Fortran, Prev: FFTW Fortran type reference, Up: Calling FFTW from Modern Fortran 7.4 Plan execution in Fortran ============================= In C, in order to use a plan, one normally calls `fftw_execute', which executes the plan to perform the transform on the input/output arrays passed when the plan was created (*note Using Plans::). The corresponding subroutine call in modern Fortran is: call fftw_execute(plan) However, we have had reports that this causes problems with some recent optimizing Fortran compilers. The problem is, because the input/output arrays are not passed as explicit arguments to `fftw_execute', the semantics of Fortran (unlike C) allow the compiler to assume that the input/output arrays are not changed by `fftw_execute'. As a consequence, certain compilers end up repositioning the call to `fftw_execute', assuming incorrectly that it does nothing to the arrays. There are various workarounds to this, but the safest and simplest thing is to not use `fftw_execute' in Fortran. Instead, use the functions described in *note New-array Execute Functions::, which take the input/output arrays as explicit arguments. For example, if the plan is for a complex-data DFT and was created for the arrays `in' and `out', you would do: call fftw_execute_dft(plan, in, out) There are a few things to be careful of, however: * You must use the correct type of execute function, matching the way the plan was created. Complex DFT plans should use `fftw_execute_dft', Real-input (r2c) DFT plans should use use `fftw_execute_dft_r2c', and real-output (c2r) DFT plans should use `fftw_execute_dft_c2r'. The various r2r plans should use `fftw_execute_r2r'. Fortunately, if you use the wrong one you will get a compile-time type-mismatch error (unlike legacy Fortran). * You should normally pass the same input/output arrays that were used when creating the plan. This is always safe. * _If_ you pass _different_ input/output arrays compared to those used when creating the plan, you must abide by all the restrictions of the new-array execute functions (*note New-array Execute Functions::). The most tricky of these is the requirement that the new arrays have the same alignment as the original arrays; the best (and possibly only) way to guarantee this is to use the `fftw_alloc' functions to allocate your arrays (*note Allocating aligned memory in Fortran::). Alternatively, you can use the `FFTW_UNALIGNED' flag when creating the plan, in which case the plan does not depend on the alignment, but this may sacrifice substantial performance on architectures (like x86) with SIMD instructions (*note SIMD alignment and fftw_malloc::).  File: fftw3.info, Node: Allocating aligned memory in Fortran, Next: Accessing the wisdom API from Fortran, Prev: Plan execution in Fortran, Up: Calling FFTW from Modern Fortran 7.5 Allocating aligned memory in Fortran ======================================== In order to obtain maximum performance in FFTW, you should store your data in arrays that have been specially aligned in memory (*note SIMD alignment and fftw_malloc::). Enforcing alignment also permits you to safely use the new-array execute functions (*note New-array Execute Functions::) to apply a given plan to more than one pair of in/out arrays. Unfortunately, standard Fortran arrays do _not_ provide any alignment guarantees. The _only_ way to allocate aligned memory in standard Fortran is to allocate it with an external C function, like the `fftw_alloc_real' and `fftw_alloc_complex' functions. Fortunately, Fortran 2003 provides a simple way to associate such allocated memory with a standard Fortran array pointer that you can then use normally. We therefore recommend allocating all your input/output arrays using the following technique: 1. Declare a `pointer', `arr', to your array of the desired type and dimensions. For example, `real(C_DOUBLE), pointer :: a(:,:)' for a 2d real array, or `complex(C_DOUBLE_COMPLEX), pointer :: a(:,:,:)' for a 3d complex array. 2. The number of elements to allocate must be an `integer(C_SIZE_T)'. You can either declare a variable of this type, e.g. `integer(C_SIZE_T) :: sz', to store the number of elements to allocate, or you can use the `int(..., C_SIZE_T)' intrinsic function. e.g. set `sz = L * M * N' or use `int(L * M * N, C_SIZE_T)' for an L x M x N array. 3. Declare a `type(C_PTR) :: p' to hold the return value from FFTW's allocation routine. Set `p = fftw_alloc_real(sz)' for a real array, or `p = fftw_alloc_complex(sz)' for a complex array. 4. Associate your pointer `arr' with the allocated memory `p' using the standard `c_f_pointer' subroutine: `call c_f_pointer(p, arr, [...dimensions...])', where `[...dimensions...])' are an array of the dimensions of the array (in the usual Fortran order). e.g. `call c_f_pointer(p, arr, [L,M,N])' for an L x M x N array. (Alternatively, you can omit the dimensions argument if you specified the shape explicitly when declaring `arr'.) You can now use `arr' as a usual multidimensional array. 5. When you are done using the array, deallocate the memory by `call fftw_free(p)' on `p'. For example, here is how we would allocate an L x M 2d real array: real(C_DOUBLE), pointer :: arr(:,:) type(C_PTR) :: p p = fftw_alloc_real(int(L * M, C_SIZE_T)) call c_f_pointer(p, arr, [L,M]) _...use arr and arr(i,j) as usual..._ call fftw_free(p) and here is an L x M x N 3d complex array: complex(C_DOUBLE_COMPLEX), pointer :: arr(:,:,:) type(C_PTR) :: p p = fftw_alloc_complex(int(L * M * N, C_SIZE_T)) call c_f_pointer(p, arr, [L,M,N]) _...use arr and arr(i,j,k) as usual..._ call fftw_free(p) See *note Reversing array dimensions:: for an example allocating a single array and associating both real and complex array pointers with it, for in-place real-to-complex transforms.  File: fftw3.info, Node: Accessing the wisdom API from Fortran, Next: Defining an FFTW module, Prev: Allocating aligned memory in Fortran, Up: Calling FFTW from Modern Fortran 7.6 Accessing the wisdom API from Fortran ========================================= As explained in *note Words of Wisdom-Saving Plans::, FFTW provides a "wisdom" API for saving plans to disk so that they can be recreated quickly. The C API for exporting (*note Wisdom Export::) and importing (*note Wisdom Import::) wisdom is somewhat tricky to use from Fortran, however, because of differences in file I/O and string types between C and Fortran. * Menu: * Wisdom File Export/Import from Fortran:: * Wisdom String Export/Import from Fortran:: * Wisdom Generic Export/Import from Fortran::  File: fftw3.info, Node: Wisdom File Export/Import from Fortran, Next: Wisdom String Export/Import from Fortran, Prev: Accessing the wisdom API from Fortran, Up: Accessing the wisdom API from Fortran 7.6.1 Wisdom File Export/Import from Fortran -------------------------------------------- The easiest way to export and import wisdom is to do so using `fftw_export_wisdom_to_filename' and `fftw_wisdom_from_filename'. The only trick is that these require you to pass a C string, which is an array of type `CHARACTER(C_CHAR)' that is terminated by `C_NULL_CHAR'. You can call them like this: integer(C_INT) :: ret ret = fftw_export_wisdom_to_filename(C_CHAR_'my_wisdom.dat' // C_NULL_CHAR) if (ret .eq. 0) stop 'error exporting wisdom to file' ret = fftw_import_wisdom_from_filename(C_CHAR_'my_wisdom.dat' // C_NULL_CHAR) if (ret .eq. 0) stop 'error importing wisdom from file' Note that prepending `C_CHAR_' is needed to specify that the literal string is of kind `C_CHAR', and we null-terminate the string by appending `// C_NULL_CHAR'. These functions return an `integer(C_INT)' (`ret') which is `0' if an error occurred during export/import and nonzero otherwise. It is also possible to use the lower-level routines `fftw_export_wisdom_to_file' and `fftw_import_wisdom_from_file', which accept parameters of the C type `FILE*', expressed in Fortran as `type(C_PTR)'. However, you are then responsible for creating the `FILE*' yourself. You can do this by using `iso_c_binding' to define Fortran intefaces for the C library functions `fopen' and `fclose', which is a bit strange in Fortran but workable.  File: fftw3.info, Node: Wisdom String Export/Import from Fortran, Next: Wisdom Generic Export/Import from Fortran, Prev: Wisdom File Export/Import from Fortran, Up: Accessing the wisdom API from Fortran 7.6.2 Wisdom String Export/Import from Fortran ---------------------------------------------- Dealing with FFTW's C string export/import is a bit more painful. In particular, the `fftw_export_wisdom_to_string' function requires you to deal with a dynamically allocated C string. To get its length, you must define an interface to the C `strlen' function, and to deallocate it you must define an interface to C `free': use, intrinsic :: iso_c_binding interface integer(C_INT) function strlen(s) bind(C, name='strlen') import type(C_PTR), value :: s end function strlen subroutine free(p) bind(C, name='free') import type(C_PTR), value :: p end subroutine free end interface Given these definitions, you can then export wisdom to a Fortran character array: character(C_CHAR), pointer :: s(:) integer(C_SIZE_T) :: slen type(C_PTR) :: p p = fftw_export_wisdom_to_string() if (.not. c_associated(p)) stop 'error exporting wisdom' slen = strlen(p) call c_f_pointer(p, s, [slen+1]) ... call free(p) Note that `slen' is the length of the C string, but the length of the array is `slen+1' because it includes the terminating null character. (You can omit the `+1' if you don't want Fortran to know about the null character.) The standard `c_associated' function checks whether `p' is a null pointer, which is returned by `fftw_export_wisdom_to_string' if there was an error. To import wisdom from a string, use `fftw_import_wisdom_from_string' as usual; note that the argument of this function must be a `character(C_CHAR)' that is terminated by the `C_NULL_CHAR' character, like the `s' array above.  File: fftw3.info, Node: Wisdom Generic Export/Import from Fortran, Prev: Wisdom String Export/Import from Fortran, Up: Accessing the wisdom API from Fortran 7.6.3 Wisdom Generic Export/Import from Fortran ----------------------------------------------- The most generic wisdom export/import functions allow you to provide an arbitrary callback function to read/write one character at a time in any way you want. However, your callback function must be written in a special way, using the `bind(C)' attribute to be passed to a C interface. In particular, to call the generic wisdom export function `fftw_export_wisdom', you would write a callback subroutine of the form: subroutine my_write_char(c, p) bind(C) use, intrinsic :: iso_c_binding character(C_CHAR), value :: c type(C_PTR), value :: p _...write c..._ end subroutine my_write_char Given such a subroutine (along with the corresponding interface definition), you could then export wisdom using: call fftw_export_wisdom(c_funloc(my_write_char), p) The standard `c_funloc' intrinsic converts a Fortran `bind(C)' subroutine into a C function pointer. The parameter `p' is a `type(C_PTR)' to any arbitrary data that you want to pass to `my_write_char' (or `C_NULL_PTR' if none). (Note that you can get a C pointer to Fortran data using the intrinsic `c_loc', and convert it back to a Fortran pointer in `my_write_char' using `c_f_pointer'.) Similarly, to use the generic `fftw_import_wisdom', you would define a callback function of the form: integer(C_INT) function my_read_char(p) bind(C) use, intrinsic :: iso_c_binding type(C_PTR), value :: p character :: c _...read a character c..._ my_read_char = ichar(c, C_INT) end function my_read_char .... integer(C_INT) :: ret ret = fftw_import_wisdom(c_funloc(my_read_char), p) if (ret .eq. 0) stop 'error importing wisdom' Your function can return `-1' if the end of the input is reached. Again, `p' is an arbitrary `type(C_PTR' that is passed through to your function. `fftw_import_wisdom' returns `0' if an error occurred and nonzero otherwise.  File: fftw3.info, Node: Defining an FFTW module, Prev: Accessing the wisdom API from Fortran, Up: Calling FFTW from Modern Fortran 7.7 Defining an FFTW module =========================== Rather than using the `include' statement to include the `fftw3.f03' interface file in any subroutine where you want to use FFTW, you might prefer to define an FFTW Fortran module. FFTW does not install itself as a module, primarily because `fftw3.f03' can be shared between different Fortran compilers while modules (in general) cannot. However, it is trivial to define your own FFTW module if you want. Just create a file containing: module FFTW3 use, intrinsic :: iso_c_binding include 'fftw3.f03' end module Compile this file into a module as usual for your compiler (e.g. with `gfortran -c' you will get a file `fftw3.mod'). Now, instead of `include 'fftw3.f03'', whenever you want to use FFTW routines you can just do: use FFTW3 as usual for Fortran modules. (You still need to link to the FFTW library, of course.)  File: fftw3.info, Node: Calling FFTW from Legacy Fortran, Next: Upgrading from FFTW version 2, Prev: Calling FFTW from Modern Fortran, Up: Top 8 Calling FFTW from Legacy Fortran ********************************** This chapter describes the interface to FFTW callable by Fortran code in older compilers not supporting the Fortran 2003 C interoperability features (*note Calling FFTW from Modern Fortran::). This interface has the major disadvantage that it is not type-checked, so if you mistake the argument types or ordering then your program will not have any compiler errors, and will likely crash at runtime. So, greater care is needed. Also, technically interfacing older Fortran versions to C is nonstandard, but in practice we have found that the techniques used in this chapter have worked with all known Fortran compilers for many years. The legacy Fortran interface differs from the C interface only in the prefix (`dfftw_' instead of `fftw_' in double precision) and a few other minor details. This Fortran interface is included in the FFTW libraries by default, unless a Fortran compiler isn't found on your system or `--disable-fortran' is included in the `configure' flags. We assume here that the reader is already familiar with the usage of FFTW in C, as described elsewhere in this manual. The MPI parallel interface to FFTW is _not_ currently available to legacy Fortran. * Menu: * Fortran-interface routines:: * FFTW Constants in Fortran:: * FFTW Execution in Fortran:: * Fortran Examples:: * Wisdom of Fortran?::  File: fftw3.info, Node: Fortran-interface routines, Next: FFTW Constants in Fortran, Prev: Calling FFTW from Legacy Fortran, Up: Calling FFTW from Legacy Fortran 8.1 Fortran-interface routines ============================== Nearly all of the FFTW functions have Fortran-callable equivalents. The name of the legacy Fortran routine is the same as that of the corresponding C routine, but with the `fftw_' prefix replaced by `dfftw_'.(1) The single and long-double precision versions use `sfftw_' and `lfftw_', respectively, instead of `fftwf_' and `fftwl_'; quadruple precision (`real*16') is available on some systems as `fftwq_' (*note Precision::). (Note that `long double' on x86 hardware is usually at most 80-bit extended precision, _not_ quadruple precision.) For the most part, all of the arguments to the functions are the same, with the following exceptions: * `plan' variables (what would be of type `fftw_plan' in C), must be declared as a type that is at least as big as a pointer (address) on your machine. We recommend using `integer*8' everywhere, since this should always be big enough. * Any function that returns a value (e.g. `fftw_plan_dft') is converted into a _subroutine_. The return value is converted into an additional _first_ parameter of this subroutine.(2) * The Fortran routines expect multi-dimensional arrays to be in _column-major_ order, which is the ordinary format of Fortran arrays (*note Multi-dimensional Array Format::). They do this transparently and costlessly simply by reversing the order of the dimensions passed to FFTW, but this has one important consequence for multi-dimensional real-complex transforms, discussed below. * Wisdom import and export is somewhat more tricky because one cannot easily pass files or strings between C and Fortran; see *note Wisdom of Fortran?::. * Legacy Fortran cannot use the `fftw_malloc' dynamic-allocation routine. If you want to exploit the SIMD FFTW (*note SIMD alignment and fftw_malloc::), you'll need to figure out some other way to ensure that your arrays are at least 16-byte aligned. * Since Fortran 77 does not have data structures, the `fftw_iodim' structure from the guru interface (*note Guru vector and transform sizes::) must be split into separate arguments. In particular, any `fftw_iodim' array arguments in the C guru interface become three integer array arguments (`n', `is', and `os') in the Fortran guru interface, all of whose lengths should be equal to the corresponding `rank' argument. * The guru planner interface in Fortran does _not_ do any automatic translation between column-major and row-major; you are responsible for setting the strides etcetera to correspond to your Fortran arrays. However, as a slight bug that we are preserving for backwards compatibility, the `plan_guru_r2r' in Fortran _does_ reverse the order of its `kind' array parameter, so the `kind' array of that routine should be in the reverse of the order of the iodim arrays (see above). In general, you should take care to use Fortran data types that correspond to (i.e. are the same size as) the C types used by FFTW. In practice, this correspondence is usually straightforward (i.e. `integer' corresponds to `int', `real' corresponds to `float', etcetera). The native Fortran double/single-precision complex type should be compatible with `fftw_complex'/`fftwf_complex'. Such simple correspondences are assumed in the examples below. ---------- Footnotes ---------- (1) Technically, Fortran 77 identifiers are not allowed to have more than 6 characters, nor may they contain underscores. Any compiler that enforces this limitation doesn't deserve to link to FFTW. (2) The reason for this is that some Fortran implementations seem to have trouble with C function return values, and vice versa.  File: fftw3.info, Node: FFTW Constants in Fortran, Next: FFTW Execution in Fortran, Prev: Fortran-interface routines, Up: Calling FFTW from Legacy Fortran 8.2 FFTW Constants in Fortran ============================= When creating plans in FFTW, a number of constants are used to specify options, such as `FFTW_MEASURE' or `FFTW_ESTIMATE'. The same constants must be used with the wrapper routines, but of course the C header files where the constants are defined can't be incorporated directly into Fortran code. Instead, we have placed Fortran equivalents of the FFTW constant definitions in the file `fftw3.f', which can be found in the same directory as `fftw3.h'. If your Fortran compiler supports a preprocessor of some sort, you should be able to `include' or `#include' this file; otherwise, you can paste it directly into your code. In C, you combine different flags (like `FFTW_PRESERVE_INPUT' and `FFTW_MEASURE') using the ``|'' operator; in Fortran you should just use ``+''. (Take care not to add in the same flag more than once, though. Alternatively, you can use the `ior' intrinsic function standardized in Fortran 95.)  File: fftw3.info, Node: FFTW Execution in Fortran, Next: Fortran Examples, Prev: FFTW Constants in Fortran, Up: Calling FFTW from Legacy Fortran 8.3 FFTW Execution in Fortran ============================= In C, in order to use a plan, one normally calls `fftw_execute', which executes the plan to perform the transform on the input/output arrays passed when the plan was created (*note Using Plans::). The corresponding subroutine call in legacy Fortran is: call dfftw_execute(plan) However, we have had reports that this causes problems with some recent optimizing Fortran compilers. The problem is, because the input/output arrays are not passed as explicit arguments to `dfftw_execute', the semantics of Fortran (unlike C) allow the compiler to assume that the input/output arrays are not changed by `dfftw_execute'. As a consequence, certain compilers end up optimizing out or repositioning the call to `dfftw_execute', assuming incorrectly that it does nothing. There are various workarounds to this, but the safest and simplest thing is to not use `dfftw_execute' in Fortran. Instead, use the functions described in *note New-array Execute Functions::, which take the input/output arrays as explicit arguments. For example, if the plan is for a complex-data DFT and was created for the arrays `in' and `out', you would do: call dfftw_execute_dft(plan, in, out) There are a few things to be careful of, however: * You must use the correct type of execute function, matching the way the plan was created. Complex DFT plans should use `dfftw_execute_dft', Real-input (r2c) DFT plans should use use `dfftw_execute_dft_r2c', and real-output (c2r) DFT plans should use `dfftw_execute_dft_c2r'. The various r2r plans should use `dfftw_execute_r2r'. * You should normally pass the same input/output arrays that were used when creating the plan. This is always safe. * _If_ you pass _different_ input/output arrays compared to those used when creating the plan, you must abide by all the restrictions of the new-array execute functions (*note New-array Execute Functions::). The most difficult of these, in Fortran, is the requirement that the new arrays have the same alignment as the original arrays, because there seems to be no way in legacy Fortran to obtain guaranteed-aligned arrays (analogous to `fftw_malloc' in C). You can, of course, use the `FFTW_UNALIGNED' flag when creating the plan, in which case the plan does not depend on the alignment, but this may sacrifice substantial performance on architectures (like x86) with SIMD instructions (*note SIMD alignment and fftw_malloc::).  File: fftw3.info, Node: Fortran Examples, Next: Wisdom of Fortran?, Prev: FFTW Execution in Fortran, Up: Calling FFTW from Legacy Fortran 8.4 Fortran Examples ==================== In C, you might have something like the following to transform a one-dimensional complex array: fftw_complex in[N], out[N]; fftw_plan plan; plan = fftw_plan_dft_1d(N,in,out,FFTW_FORWARD,FFTW_ESTIMATE); fftw_execute(plan); fftw_destroy_plan(plan); In Fortran, you would use the following to accomplish the same thing: double complex in, out dimension in(N), out(N) integer*8 plan call dfftw_plan_dft_1d(plan,N,in,out,FFTW_FORWARD,FFTW_ESTIMATE) call dfftw_execute_dft(plan, in, out) call dfftw_destroy_plan(plan) Notice how all routines are called as Fortran subroutines, and the plan is returned via the first argument to `dfftw_plan_dft_1d'. Notice also that we changed `fftw_execute' to `dfftw_execute_dft' (*note FFTW Execution in Fortran::). To do the same thing, but using 8 threads in parallel (*note Multi-threaded FFTW::), you would simply prefix these calls with: integer iret call dfftw_init_threads(iret) call dfftw_plan_with_nthreads(8) (You might want to check the value of `iret': if it is zero, it indicates an unlikely error during thread initialization.) To transform a three-dimensional array in-place with C, you might do: fftw_complex arr[L][M][N]; fftw_plan plan; plan = fftw_plan_dft_3d(L,M,N, arr,arr, FFTW_FORWARD, FFTW_ESTIMATE); fftw_execute(plan); fftw_destroy_plan(plan); In Fortran, you would use this instead: double complex arr dimension arr(L,M,N) integer*8 plan call dfftw_plan_dft_3d(plan, L,M,N, arr,arr, & FFTW_FORWARD, FFTW_ESTIMATE) call dfftw_execute_dft(plan, arr, arr) call dfftw_destroy_plan(plan) Note that we pass the array dimensions in the "natural" order in both C and Fortran. To transform a one-dimensional real array in Fortran, you might do: double precision in dimension in(N) double complex out dimension out(N/2 + 1) integer*8 plan call dfftw_plan_dft_r2c_1d(plan,N,in,out,FFTW_ESTIMATE) call dfftw_execute_dft_r2c(plan, in, out) call dfftw_destroy_plan(plan) To transform a two-dimensional real array, out of place, you might use the following: double precision in dimension in(M,N) double complex out dimension out(M/2 + 1, N) integer*8 plan call dfftw_plan_dft_r2c_2d(plan,M,N,in,out,FFTW_ESTIMATE) call dfftw_execute_dft_r2c(plan, in, out) call dfftw_destroy_plan(plan) *Important:* Notice that it is the _first_ dimension of the complex output array that is cut in half in Fortran, rather than the last dimension as in C. This is a consequence of the interface routines reversing the order of the array dimensions passed to FFTW so that the Fortran program can use its ordinary column-major order.  File: fftw3.info, Node: Wisdom of Fortran?, Prev: Fortran Examples, Up: Calling FFTW from Legacy Fortran 8.5 Wisdom of Fortran? ====================== In this section, we discuss how one can import/export FFTW wisdom (saved plans) to/from a Fortran program; we assume that the reader is already familiar with wisdom, as described in *note Words of Wisdom-Saving Plans::. The basic problem is that is difficult to (portably) pass files and strings between Fortran and C, so we cannot provide a direct Fortran equivalent to the `fftw_export_wisdom_to_file', etcetera, functions. Fortran interfaces _are_ provided for the functions that do not take file/string arguments, however: `dfftw_import_system_wisdom', `dfftw_import_wisdom', `dfftw_export_wisdom', and `dfftw_forget_wisdom'. So, for example, to import the system-wide wisdom, you would do: integer isuccess call dfftw_import_system_wisdom(isuccess) As usual, the C return value is turned into a first parameter; `isuccess' is non-zero on success and zero on failure (e.g. if there is no system wisdom installed). If you want to import/export wisdom from/to an arbitrary file or elsewhere, you can employ the generic `dfftw_import_wisdom' and `dfftw_export_wisdom' functions, for which you must supply a subroutine to read/write one character at a time. The FFTW package contains an example file `doc/f77_wisdom.f' demonstrating how to implement `import_wisdom_from_file' and `export_wisdom_to_file' subroutines in this way. (These routines cannot be compiled into the FFTW library itself, lest all FFTW-using programs be required to link with the Fortran I/O library.)  File: fftw3.info, Node: Upgrading from FFTW version 2, Next: Installation and Customization, Prev: Calling FFTW from Legacy Fortran, Up: Top 9 Upgrading from FFTW version 2 ******************************* In this chapter, we outline the process for updating codes designed for the older FFTW 2 interface to work with FFTW 3. The interface for FFTW 3 is not backwards-compatible with the interface for FFTW 2 and earlier versions; codes written to use those versions will fail to link with FFTW 3. Nor is it possible to write "compatibility wrappers" to bridge the gap (at least not efficiently), because FFTW 3 has different semantics from previous versions. However, upgrading should be a straightforward process because the data formats are identical and the overall style of planning/execution is essentially the same. Unlike FFTW 2, there are no separate header files for real and complex transforms (or even for different precisions) in FFTW 3; all interfaces are defined in the `' header file. Numeric Types ============= The main difference in data types is that `fftw_complex' in FFTW 2 was defined as a `struct' with macros `c_re' and `c_im' for accessing the real/imaginary parts. (This is binary-compatible with FFTW 3 on any machine except perhaps for some older Crays in single precision.) The equivalent macros for FFTW 3 are: #define c_re(c) ((c)[0]) #define c_im(c) ((c)[1]) This does not work if you are using the C99 complex type, however, unless you insert a `double*' typecast into the above macros (*note Complex numbers::). Also, FFTW 2 had an `fftw_real' typedef that was an alias for `double' (in double precision). In FFTW 3 you should just use `double' (or whatever precision you are employing). Plans ===== The major difference between FFTW 2 and FFTW 3 is in the planning/execution division of labor. In FFTW 2, plans were found for a given transform size and type, and then could be applied to _any_ arrays and for _any_ multiplicity/stride parameters. In FFTW 3, you specify the particular arrays, stride parameters, etcetera when creating the plan, and the plan is then executed for _those_ arrays (unless the guru interface is used) and _those_ parameters _only_. (FFTW 2 had "specific planner" routines that planned for a particular array and stride, but the plan could still be used for other arrays and strides.) That is, much of the information that was formerly specified at execution time is now specified at planning time. Like FFTW 2's specific planner routines, the FFTW 3 planner overwrites the input/output arrays unless you use `FFTW_ESTIMATE'. FFTW 2 had separate data types `fftw_plan', `fftwnd_plan', `rfftw_plan', and `rfftwnd_plan' for complex and real one- and multi-dimensional transforms, and each type had its own `destroy' function. In FFTW 3, all plans are of type `fftw_plan' and all are destroyed by `fftw_destroy_plan(plan)'. Where you formerly used `fftw_create_plan' and `fftw_one' to plan and compute a single 1d transform, you would now use `fftw_plan_dft_1d' to plan the transform. If you used the generic `fftw' function to execute the transform with multiplicity (`howmany') and stride parameters, you would now use the advanced interface `fftw_plan_many_dft' to specify those parameters. The plans are now executed with `fftw_execute(plan)', which takes all of its parameters (including the input/output arrays) from the plan. In-place transforms no longer interpret their output argument as scratch space, nor is there an `FFTW_IN_PLACE' flag. You simply pass the same pointer for both the input and output arguments. (Previously, the output `ostride' and `odist' parameters were ignored for in-place transforms; now, if they are specified via the advanced interface, they are significant even in the in-place case, although they should normally equal the corresponding input parameters.) The `FFTW_ESTIMATE' and `FFTW_MEASURE' flags have the same meaning as before, although the planning time will differ. You may also consider using `FFTW_PATIENT', which is like `FFTW_MEASURE' except that it takes more time in order to consider a wider variety of algorithms. For multi-dimensional complex DFTs, instead of `fftwnd_create_plan' (or `fftw2d_create_plan' or `fftw3d_create_plan'), followed by `fftwnd_one', you would use `fftw_plan_dft' (or `fftw_plan_dft_2d' or `fftw_plan_dft_3d'). followed by `fftw_execute'. If you used `fftwnd' to to specify strides etcetera, you would instead specify these via `fftw_plan_many_dft'. The analogues to `rfftw_create_plan' and `rfftw_one' with `FFTW_REAL_TO_COMPLEX' or `FFTW_COMPLEX_TO_REAL' directions are `fftw_plan_r2r_1d' with kind `FFTW_R2HC' or `FFTW_HC2R', followed by `fftw_execute'. The stride etcetera arguments of `rfftw' are now in `fftw_plan_many_r2r'. Instead of `rfftwnd_create_plan' (or `rfftw2d_create_plan' or `rfftw3d_create_plan') followed by `rfftwnd_one_real_to_complex' or `rfftwnd_one_complex_to_real', you now use `fftw_plan_dft_r2c' (or `fftw_plan_dft_r2c_2d' or `fftw_plan_dft_r2c_3d') or `fftw_plan_dft_c2r' (or `fftw_plan_dft_c2r_2d' or `fftw_plan_dft_c2r_3d'), respectively, followed by `fftw_execute'. As usual, the strides etcetera of `rfftwnd_real_to_complex' or `rfftwnd_complex_to_real' are no specified in the advanced planner routines, `fftw_plan_many_dft_r2c' or `fftw_plan_many_dft_c2r'. Wisdom ====== In FFTW 2, you had to supply the `FFTW_USE_WISDOM' flag in order to use wisdom; in FFTW 3, wisdom is always used. (You could simulate the FFTW 2 wisdom-less behavior by calling `fftw_forget_wisdom' after every planner call.) The FFTW 3 wisdom import/export routines are almost the same as before (although the storage format is entirely different). There is one significant difference, however. In FFTW 2, the import routines would never read past the end of the wisdom, so you could store extra data beyond the wisdom in the same file, for example. In FFTW 3, the file-import routine may read up to a few hundred bytes past the end of the wisdom, so you cannot store other data just beyond it.(1) Wisdom has been enhanced by additional humility in FFTW 3: whereas FFTW 2 would re-use wisdom for a given transform size regardless of the stride etc., in FFTW 3 wisdom is only used with the strides etc. for which it was created. Unfortunately, this means FFTW 3 has to create new plans from scratch more often than FFTW 2 (in FFTW 2, planning e.g. one transform of size 1024 also created wisdom for all smaller powers of 2, but this no longer occurs). FFTW 3 also has the new routine `fftw_import_system_wisdom' to import wisdom from a standard system-wide location. Memory allocation ================= In FFTW 3, we recommend allocating your arrays with `fftw_malloc' and deallocating them with `fftw_free'; this is not required, but allows optimal performance when SIMD acceleration is used. (Those two functions actually existed in FFTW 2, and worked the same way, but were not documented.) In FFTW 2, there were `fftw_malloc_hook' and `fftw_free_hook' functions that allowed the user to replace FFTW's memory-allocation routines (e.g. to implement different error-handling, since by default FFTW prints an error message and calls `exit' to abort the program if `malloc' returns `NULL'). These hooks are not supported in FFTW 3; those few users who require this functionality can just directly modify the memory-allocation routines in FFTW (they are defined in `kernel/alloc.c'). Fortran interface ================= In FFTW 2, the subroutine names were obtained by replacing `fftw_' with `fftw_f77'; in FFTW 3, you replace `fftw_' with `dfftw_' (or `sfftw_' or `lfftw_', depending upon the precision). In FFTW 3, we have begun recommending that you always declare the type used to store plans as `integer*8'. (Too many people didn't notice our instruction to switch from `integer' to `integer*8' for 64-bit machines.) In FFTW 3, we provide a `fftw3.f' "header file" to include in your code (and which is officially installed on Unix systems). (In FFTW 2, we supplied a `fftw_f77.i' file, but it was not installed.) Otherwise, the C-Fortran interface relationship is much the same as it was before (e.g. return values become initial parameters, and multi-dimensional arrays are in column-major order). Unlike FFTW 2, we do provide some support for wisdom import/export in Fortran (*note Wisdom of Fortran?::). Threads ======= Like FFTW 2, only the execution routines are thread-safe. All planner routines, etcetera, should be called by only a single thread at a time (*note Thread safety::). _Unlike_ FFTW 2, there is no special `FFTW_THREADSAFE' flag for the planner to allow a given plan to be usable by multiple threads in parallel; this is now the case by default. The multi-threaded version of FFTW 2 required you to pass the number of threads each time you execute the transform. The number of threads is now stored in the plan, and is specified before the planner is called by `fftw_plan_with_nthreads'. The threads initialization routine used to be called `fftw_threads_init' and would return zero on success; the new routine is called `fftw_init_threads' and returns zero on failure. *Note Multi-threaded FFTW::. There is no separate threads header file in FFTW 3; all the function prototypes are in `'. However, you still have to link to a separate library (`-lfftw3_threads -lfftw3 -lm' on Unix), as well as to the threading library (e.g. POSIX threads on Unix). ---------- Footnotes ---------- (1) We do our own buffering because GNU libc I/O routines are horribly slow for single-character I/O, apparently for thread-safety reasons (whether you are using threads or not).  File: fftw3.info, Node: Installation and Customization, Next: Acknowledgments, Prev: Upgrading from FFTW version 2, Up: Top 10 Installation and Customization ********************************* This chapter describes the installation and customization of FFTW, the latest version of which may be downloaded from the FFTW home page (http://www.fftw.org). In principle, FFTW should work on any system with an ANSI C compiler (`gcc' is fine). However, planner time is drastically reduced if FFTW can exploit a hardware cycle counter; FFTW comes with cycle-counter support for all modern general-purpose CPUs, but you may need to add a couple of lines of code if your compiler is not yet supported (*note Cycle Counters::). (On Unix, there will be a warning at the end of the `configure' output if no cycle counter is found.) Installation of FFTW is simplest if you have a Unix or a GNU system, such as GNU/Linux, and we describe this case in the first section below, including the use of special configuration options to e.g. install different precisions or exploit optimizations for particular architectures (e.g. SIMD). Compilation on non-Unix systems is a more manual process, but we outline the procedure in the second section. It is also likely that pre-compiled binaries will be available for popular systems. Finally, we describe how you can customize FFTW for particular needs by generating _codelets_ for fast transforms of sizes not supported efficiently by the standard FFTW distribution. * Menu: * Installation on Unix:: * Installation on non-Unix systems:: * Cycle Counters:: * Generating your own code::  File: fftw3.info, Node: Installation on Unix, Next: Installation on non-Unix systems, Prev: Installation and Customization, Up: Installation and Customization 10.1 Installation on Unix ========================= FFTW comes with a `configure' program in the GNU style. Installation can be as simple as: ./configure make make install This will build the uniprocessor complex and real transform libraries along with the test programs. (We recommend that you use GNU `make' if it is available; on some systems it is called `gmake'.) The "`make install'" command installs the fftw and rfftw libraries in standard places, and typically requires root privileges (unless you specify a different install directory with the `--prefix' flag to `configure'). You can also type "`make check'" to put the FFTW test programs through their paces. If you have problems during configuration or compilation, you may want to run "`make distclean'" before trying again; this ensures that you don't have any stale files left over from previous compilation attempts. The `configure' script chooses the `gcc' compiler by default, if it is available; you can select some other compiler with: ./configure CC="" The `configure' script knows good `CFLAGS' (C compiler flags) for a few systems. If your system is not known, the `configure' script will print out a warning. In this case, you should re-configure FFTW with the command ./configure CFLAGS="" and then compile as usual. If you do find an optimal set of `CFLAGS' for your system, please let us know what they are (along with the output of `config.guess') so that we can include them in future releases. `configure' supports all the standard flags defined by the GNU Coding Standards; see the `INSTALL' file in FFTW or the GNU web page (http://www.gnu.org/prep/standards/html_node/index.html). Note especially `--help' to list all flags and `--enable-shared' to create shared, rather than static, libraries. `configure' also accepts a few FFTW-specific flags, particularly: * `--enable-float': Produces a single-precision version of FFTW (`float') instead of the default double-precision (`double'). *Note Precision::. * `--enable-long-double': Produces a long-double precision version of FFTW (`long double') instead of the default double-precision (`double'). The `configure' script will halt with an error message if `long double' is the same size as `double' on your machine/compiler. *Note Precision::. * `--enable-quad-precision': Produces a quadruple-precision version of FFTW using the nonstandard `__float128' type provided by `gcc' 4.6 or later on x86, x86-64, and Itanium architectures, instead of the default double-precision (`double'). The `configure' script will halt with an error message if the compiler is not `gcc' version 4.6 or later or if `gcc''s `libquadmath' library is not installed. *Note Precision::. * `--enable-threads': Enables compilation and installation of the FFTW threads library (*note Multi-threaded FFTW::), which provides a simple interface to parallel transforms for SMP systems. By default, the threads routines are not compiled. * `--enable-openmp': Like `--enable-threads', but using OpenMP compiler directives in order to induce parallelism rather than spawning its own threads directly, and installing an `fftw3_omp' library rather than an `fftw3_threads' library (*note Multi-threaded FFTW::). You can use both `--enable-openmp' and `--enable-threads' since they compile/install libraries with different names. By default, the OpenMP routines are not compiled. * `--with-combined-threads': By default, if `--enable-threads' is used, the threads support is compiled into a separate library that must be linked in addition to the main FFTW library. This is so that users of the serial library do not need to link the system threads libraries. If `--with-combined-threads' is specified, however, then no separate threads library is created, and threads are included in the main FFTW library. This is mainly useful under Windows, where no system threads library is required and inter-library dependencies are problematic. * `--enable-mpi': Enables compilation and installation of the FFTW MPI library (*note Distributed-memory FFTW with MPI::), which provides parallel transforms for distributed-memory systems with MPI. (By default, the MPI routines are not compiled.) *Note FFTW MPI Installation::. * `--disable-fortran': Disables inclusion of legacy-Fortran wrapper routines (*note Calling FFTW from Legacy Fortran::) in the standard FFTW libraries. These wrapper routines increase the library size by only a negligible amount, so they are included by default as long as the `configure' script finds a Fortran compiler on your system. (To specify a particular Fortran compiler foo, pass `F77='foo to `configure'.) * `--with-g77-wrappers': By default, when Fortran wrappers are included, the wrappers employ the linking conventions of the Fortran compiler detected by the `configure' script. If this compiler is GNU `g77', however, then _two_ versions of the wrappers are included: one with `g77''s idiosyncratic convention of appending two underscores to identifiers, and one with the more common convention of appending only a single underscore. This way, the same FFTW library will work with both `g77' and other Fortran compilers, such as GNU `gfortran'. However, the converse is not true: if you configure with a different compiler, then the `g77'-compatible wrappers are not included. By specifying `--with-g77-wrappers', the `g77'-compatible wrappers are included in addition to wrappers for whatever Fortran compiler `configure' finds. * `--with-slow-timer': Disables the use of hardware cycle counters, and falls back on `gettimeofday' or `clock'. This greatly worsens performance, and should generally not be used (unless you don't have a cycle counter but still really want an optimized plan regardless of the time). *Note Cycle Counters::. * `--enable-sse', `--enable-sse2', `--enable-avx', `--enable-altivec', `--enable-neon': Enable the compilation of SIMD code for SSE (Pentium III+), SSE2 (Pentium IV+), AVX (Sandy Bridge, Interlagos), AltiVec (PowerPC G4+), NEON (some ARM processors). SSE, AltiVec, and NEON only work with `--enable-float' (above). SSE2 works in both single and double precision (and is simply SSE in single precision). The resulting code will _still work_ on earlier CPUs lacking the SIMD extensions (SIMD is automatically disabled, although the FFTW library is still larger). - These options require a compiler supporting SIMD extensions, and compiler support is always a bit flaky: see the FFTW FAQ for a list of compiler versions that have problems compiling FFTW. - With AltiVec and `gcc', you may have to use the `-mabi=altivec' option when compiling any code that links to FFTW, in order to properly align the stack; otherwise, FFTW could crash when it tries to use an AltiVec feature. (This is not necessary on MacOS X.) - With SSE/SSE2 and `gcc', you should use a version of gcc that properly aligns the stack when compiling any code that links to FFTW. By default, `gcc' 2.95 and later versions align the stack as needed, but you should not compile FFTW with the `-Os' option or the `-mpreferred-stack-boundary' option with an argument less than 4. - Because of the large variety of ARM processors and ABIs, FFTW does not attempt to guess the correct `gcc' flags for generating NEON code. In general, you will have to provide them on the command line. This command line is known to have worked at least once: ./configure --with-slow-timer --host=arm-linux-gnueabi \ --enable-single --enable-neon \ "CC=arm-linux-gnueabi-gcc -march=armv7-a -mfloat-abi=softfp" To force `configure' to use a particular C compiler foo (instead of the default, usually `gcc'), pass `CC='foo to the `configure' script; you may also need to set the flags via the variable `CFLAGS' as described above.  File: fftw3.info, Node: Installation on non-Unix systems, Next: Cycle Counters, Prev: Installation on Unix, Up: Installation and Customization 10.2 Installation on non-Unix systems ===================================== It should be relatively straightforward to compile FFTW even on non-Unix systems lacking the niceties of a `configure' script. Basically, you need to edit the `config.h' header (copy it from `config.h.in') to `#define' the various options and compiler characteristics, and then compile all the `.c' files in the relevant directories. The `config.h' header contains about 100 options to set, each one initially an `#undef', each documented with a comment, and most of them fairly obvious. For most of the options, you should simply `#define' them to `1' if they are applicable, although a few options require a particular value (e.g. `SIZEOF_LONG_LONG' should be defined to the size of the `long long' type, in bytes, or zero if it is not supported). We will likely post some sample `config.h' files for various operating systems and compilers for you to use (at least as a starting point). Please let us know if you have to hand-create a configuration file (and/or a pre-compiled binary) that you want to share. To create the FFTW library, you will then need to compile all of the `.c' files in the `kernel', `dft', `dft/scalar', `dft/scalar/codelets', `rdft', `rdft/scalar', `rdft/scalar/r2cf', `rdft/scalar/r2cb', `rdft/scalar/r2r', `reodft', and `api' directories. If you are compiling with SIMD support (e.g. you defined `HAVE_SSE2' in `config.h'), then you also need to compile the `.c' files in the `simd-support', `{dft,rdft}/simd', `{dft,rdft}/simd/*' directories. Once these files are all compiled, link them into a library, or a shared library, or directly into your program. To compile the FFTW test program, additionally compile the code in the `libbench2/' directory, and link it into a library. Then compile the code in the `tests/' directory and link it to the `libbench2' and FFTW libraries. To compile the `fftw-wisdom' (command-line) tool (*note Wisdom Utilities::), compile `tools/fftw-wisdom.c' and link it to the `libbench2' and FFTW libraries  File: fftw3.info, Node: Cycle Counters, Next: Generating your own code, Prev: Installation on non-Unix systems, Up: Installation and Customization 10.3 Cycle Counters =================== FFTW's planner actually executes and times different possible FFT algorithms in order to pick the fastest plan for a given n. In order to do this in as short a time as possible, however, the timer must have a very high resolution, and to accomplish this we employ the hardware "cycle counters" that are available on most CPUs. Currently, FFTW supports the cycle counters on x86, PowerPC/POWER, Alpha, UltraSPARC (SPARC v9), IA64, PA-RISC, and MIPS processors. Access to the cycle counters, unfortunately, is a compiler and/or operating-system dependent task, often requiring inline assembly language, and it may be that your compiler is not supported. If you are _not_ supported, FFTW will by default fall back on its estimator (effectively using `FFTW_ESTIMATE' for all plans). You can add support by editing the file `kernel/cycle.h'; normally, this will involve adapting one of the examples already present in order to use the inline-assembler syntax for your C compiler, and will only require a couple of lines of code. Anyone adding support for a new system to `cycle.h' is encouraged to email us at . If a cycle counter is not available on your system (e.g. some embedded processor), and you don't want to use estimated plans, as a last resort you can use the `--with-slow-timer' option to `configure' (on Unix) or `#define WITH_SLOW_TIMER' in `config.h' (elsewhere). This will use the much lower-resolution `gettimeofday' function, or even `clock' if the former is unavailable, and planning will be extremely slow.  File: fftw3.info, Node: Generating your own code, Prev: Cycle Counters, Up: Installation and Customization 10.4 Generating your own code ============================= The directory `genfft' contains the programs that were used to generate FFTW's "codelets," which are hard-coded transforms of small sizes. We do not expect casual users to employ the generator, which is a rather sophisticated program that generates directed acyclic graphs of FFT algorithms and performs algebraic simplifications on them. It was written in Objective Caml, a dialect of ML, which is available at `http://caml.inria.fr/ocaml/index.en.html'. If you have Objective Caml installed (along with recent versions of GNU `autoconf', `automake', and `libtool'), then you can change the set of codelets that are generated or play with the generation options. The set of generated codelets is specified by the `{dft,rdft}/{codelets,simd}/*/Makefile.am' files. For example, you can add efficient REDFT codelets of small sizes by modifying `rdft/codelets/r2r/Makefile.am'. After you modify any `Makefile.am' files, you can type `sh bootstrap.sh' in the top-level directory followed by `make' to re-generate the files. We do not provide more details about the code-generation process, since we do not expect that most users will need to generate their own code. However, feel free to contact us at if you are interested in the subject. You might find it interesting to learn Caml and/or some modern programming techniques that we used in the generator (including monadic programming), especially if you heard the rumor that Java and object-oriented programming are the latest advancement in the field. The internal operation of the codelet generator is described in the paper, "A Fast Fourier Transform Compiler," by M. Frigo, which is available from the FFTW home page (http://www.fftw.org) and also appeared in the `Proceedings of the 1999 ACM SIGPLAN Conference on Programming Language Design and Implementation (PLDI)'.  File: fftw3.info, Node: Acknowledgments, Next: License and Copyright, Prev: Installation and Customization, Up: Top 11 Acknowledgments ****************** Matteo Frigo was supported in part by the Special Research Program SFB F011 "AURORA" of the Austrian Science Fund FWF and by MIT Lincoln Laboratory. For previous versions of FFTW, he was supported in part by the Defense Advanced Research Projects Agency (DARPA), under Grants N00014-94-1-0985 and F30602-97-1-0270, and by a Digital Equipment Corporation Fellowship. Steven G. Johnson was supported in part by a Dept. of Defense NDSEG Fellowship, an MIT Karl Taylor Compton Fellowship, and by the Materials Research Science and Engineering Center program of the National Science Foundation under award DMR-9400334. Code for the Cell Broadband Engine was graciously donated to the FFTW project by the IBM Austin Research Lab and included in fftw-3.2. (This code was removed in fftw-3.3.) Code for the MIPS paired-single SIMD support was graciously donated to the FFTW project by CodeSourcery, Inc. We are grateful to Sun Microsystems Inc. for its donation of a cluster of 9 8-processor Ultra HPC 5000 SMPs (24 Gflops peak). These machines served as the primary platform for the development of early versions of FFTW. We thank Intel Corporation for donating a four-processor Pentium Pro machine. We thank the GNU/Linux community for giving us a decent OS to run on that machine. We are thankful to the AMD corporation for donating an AMD Athlon XP 1700+ computer to the FFTW project. We thank the Compaq/HP testdrive program and VA Software Corporation (SourceForge.net) for providing remote access to machines that were used to test FFTW. The `genfft' suite of code generators was written using Objective Caml, a dialect of ML. Objective Caml is a small and elegant language developed by Xavier Leroy. The implementation is available from `http://caml.inria.fr/' (http://caml.inria.fr/). In previous releases of FFTW, `genfft' was written in Caml Light, by the same authors. An even earlier implementation of `genfft' was written in Scheme, but Caml is definitely better for this kind of application. FFTW uses many tools from the GNU project, including `automake', `texinfo', and `libtool'. Prof. Charles E. Leiserson of MIT provided continuous support and encouragement. This program would not exist without him. Charles also proposed the name "codelets" for the basic FFT blocks. Prof. John D. Joannopoulos of MIT demonstrated continuing tolerance of Steven's "extra-curricular" computer-science activities, as well as remarkable creativity in working them into his grant proposals. Steven's physics degree would not exist without him. Franz Franchetti wrote SIMD extensions to FFTW 2, which eventually led to the SIMD support in FFTW 3. Stefan Kral wrote most of the K7 code generator distributed with FFTW 3.0.x and 3.1.x. Andrew Sterian contributed the Windows timing code in FFTW 2. Didier Miras reported a bug in the test procedure used in FFTW 1.2. We now use a completely different test algorithm by Funda Ergun that does not require a separate FFT program to compare against. Wolfgang Reimer contributed the Pentium cycle counter and a few fixes that help portability. Ming-Chang Liu uncovered a well-hidden bug in the complex transforms of FFTW 2.0 and supplied a patch to correct it. The FFTW FAQ was written in `bfnn' (Bizarre Format With No Name) and formatted using the tools developed by Ian Jackson for the Linux FAQ. _We are especially thankful to all of our users for their continuing support, feedback, and interest during our development of FFTW._  File: fftw3.info, Node: License and Copyright, Next: Concept Index, Prev: Acknowledgments, Up: Top 12 License and Copyright ************************ FFTW is Copyright (C) 2003, 2007-11 Matteo Frigo, Copyright (C) 2003, 2007-11 Massachusetts Institute of Technology. FFTW is free software; you can redistribute it and/or modify it under the terms of the GNU General Public License as published by the Free Software Foundation; either version 2 of the License, or (at your option) any later version. This program is distributed in the hope that it will be useful, but WITHOUT ANY WARRANTY; without even the implied warranty of MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the GNU General Public License for more details. You should have received a copy of the GNU General Public License along with this program; if not, write to the Free Software Foundation, Inc., 51 Franklin Street, Fifth Floor, Boston, MA 02110-1301 USA You can also find the GPL on the GNU web site (http://www.gnu.org/licenses/gpl-2.0.html). In addition, we kindly ask you to acknowledge FFTW and its authors in any program or publication in which you use FFTW. (You are not _required_ to do so; it is up to your common sense to decide whether you want to comply with this request or not.) For general publications, we suggest referencing: Matteo Frigo and Steven G. Johnson, "The design and implementation of FFTW3," Proc. IEEE 93 (2), 216-231 (2005). Non-free versions of FFTW are available under terms different from those of the General Public License. (e.g. they do not require you to accompany any object code using FFTW with the corresponding source code.) For these alternative terms you must purchase a license from MIT's Technology Licensing Office. Users interested in such a license should contact us () for more information. fftw-3.3.3/doc/fftw3.info-20000644000175000017500000022611012054411105012214 00000000000000This is fftw3.info, produced by makeinfo version 4.13 from fftw3.texi. This manual is for FFTW (version 3.3.3, 25 November 2012). Copyright (C) 2003 Matteo Frigo. Copyright (C) 2003 Massachusetts Institute of Technology. Permission is granted to make and distribute verbatim copies of this manual provided the copyright notice and this permission notice are preserved on all copies. Permission is granted to copy and distribute modified versions of this manual under the conditions for verbatim copying, provided that the entire resulting derived work is distributed under the terms of a permission notice identical to this one. Permission is granted to copy and distribute translations of this manual into another language, under the above conditions for modified versions, except that this permission notice may be stated in a translation approved by the Free Software Foundation. INFO-DIR-SECTION Texinfo documentation system START-INFO-DIR-ENTRY * fftw3: (fftw3). FFTW User's Manual. END-INFO-DIR-ENTRY  File: fftw3.info, Node: Concept Index, Next: Library Index, Prev: License and Copyright, Up: Top 13 Concept Index **************** [index] * Menu: * 64-bit architecture <1>: FFTW Fortran type reference. (line 53) * 64-bit architecture <2>: 2d MPI example. (line 69) * 64-bit architecture: 64-bit Guru Interface. (line 6) * advanced interface <1>: MPI Plan Creation. (line 43) * advanced interface <2>: MPI Data Distribution Functions. (line 47) * advanced interface <3>: Basic and advanced distribution interfaces. (line 45) * advanced interface <4>: Advanced Interface. (line 6) * advanced interface <5>: Row-major Format. (line 26) * advanced interface <6>: Complex Multi-Dimensional DFTs. (line 62) * advanced interface: Introduction. (line 67) * algorithm: Introduction. (line 99) * alignment <1>: Allocating aligned memory in Fortran. (line 6) * alignment <2>: Overview of Fortran interface. (line 50) * alignment <3>: Using MPI Plans. (line 24) * alignment <4>: New-array Execute Functions. (line 36) * alignment <5>: Planner Flags. (line 76) * alignment: Memory Allocation. (line 12) * AltiVec: SIMD alignment and fftw_malloc. (line 13) * AVX: SIMD alignment and fftw_malloc. (line 13) * basic interface <1>: Basic Interface. (line 6) * basic interface <2>: Tutorial. (line 14) * basic interface: Introduction. (line 67) * block distribution <1>: FFTW MPI Performance Tips. (line 14) * block distribution <2>: Basic and advanced distribution interfaces. (line 72) * block distribution: MPI Data Distribution. (line 21) * C multi-dimensional arrays: Fixed-size Arrays in C. (line 6) * C++ <1>: Memory Allocation. (line 24) * C++ <2>: Complex numbers. (line 35) * C++ <3>: Dynamic Arrays in C. (line 32) * C++ <4>: SIMD alignment and fftw_malloc. (line 33) * C++: Complex One-Dimensional DFTs. (line 114) * c2r <1>: Real-data DFTs. (line 89) * c2r <2>: Planner Flags. (line 74) * c2r: One-Dimensional DFTs of Real Data. (line 37) * C99 <1>: Precision. (line 30) * C99 <2>: Complex numbers. (line 22) * C99: Dynamic Arrays in C. (line 32) * Caml <1>: Acknowledgments. (line 47) * Caml: Generating your own code. (line 12) * code generator <1>: Generating your own code. (line 6) * code generator: Introduction. (line 80) * codelet <1>: Acknowledgments. (line 54) * codelet <2>: Generating your own code. (line 7) * codelet <3>: Installation and Customization. (line 29) * codelet: Introduction. (line 80) * collective function <1>: MPI Plan Creation. (line 27) * collective function <2>: Using MPI Plans. (line 10) * collective function <3>: Avoiding MPI Deadlocks. (line 12) * collective function <4>: FFTW MPI Wisdom. (line 39) * collective function: 2d MPI example. (line 61) * column-major <1>: Fortran Examples. (line 93) * column-major <2>: Fortran-interface routines. (line 28) * column-major <3>: Reversing array dimensions. (line 6) * column-major: Column-major Format. (line 6) * compiler <1>: Cycle Counters. (line 14) * compiler <2>: Installation on Unix. (line 160) * compiler <3>: Installation and Customization. (line 16) * compiler: Introduction. (line 86) * compiler flags: Installation on Unix. (line 29) * configuration routines: Wisdom Utilities. (line 26) * configure <1>: Installation on Unix. (line 7) * configure <2>: FFTW MPI Installation. (line 10) * configure: Installation and Supported Hardware/Software. (line 12) * cycle counter <1>: Cycle Counters. (line 6) * cycle counter: Installation and Customization. (line 16) * data distribution <1>: MPI Data Distribution Functions. (line 6) * data distribution <2>: Basic distributed-transpose interface. (line 33) * data distribution <3>: Multi-dimensional MPI DFTs of Real Data. (line 34) * data distribution <4>: MPI Data Distribution. (line 6) * data distribution <5>: 2d MPI example. (line 81) * data distribution: Distributed-memory FFTW with MPI. (line 14) * DCT <1>: 1d Real-even DFTs (DCTs). (line 24) * DCT <2>: Real-to-Real Transform Kinds. (line 32) * DCT: Real even/odd DFTs (cosine/sine transforms). (line 16) * deadlock: Avoiding MPI Deadlocks. (line 6) * Devil: Complex One-Dimensional DFTs. (line 8) * DFT <1>: The 1d Discrete Fourier Transform (DFT). (line 6) * DFT <2>: Complex One-Dimensional DFTs. (line 106) * DFT: Introduction. (line 9) * DHT <1>: 1d Discrete Hartley Transforms (DHTs). (line 6) * DHT: The Discrete Hartley Transform. (line 17) * discrete cosine transform <1>: 1d Real-even DFTs (DCTs). (line 24) * discrete cosine transform <2>: Real-to-Real Transform Kinds. (line 32) * discrete cosine transform: Real even/odd DFTs (cosine/sine transforms). (line 16) * discrete Fourier transform <1>: The 1d Discrete Fourier Transform (DFT). (line 6) * discrete Fourier transform: Introduction. (line 9) * discrete Hartley transform <1>: 1d Discrete Hartley Transforms (DHTs). (line 6) * discrete Hartley transform <2>: Real-to-Real Transform Kinds. (line 29) * discrete Hartley transform: The Discrete Hartley Transform. (line 17) * discrete sine transform <1>: 1d Real-odd DFTs (DSTs). (line 24) * discrete sine transform <2>: Real-to-Real Transform Kinds. (line 46) * discrete sine transform: Real even/odd DFTs (cosine/sine transforms). (line 16) * dist <1>: Guru vector and transform sizes. (line 39) * dist: Advanced Complex DFTs. (line 30) * DST <1>: 1d Real-odd DFTs (DSTs). (line 24) * DST <2>: Real-to-Real Transform Kinds. (line 46) * DST: Real even/odd DFTs (cosine/sine transforms). (line 16) * Ecclesiastes: Caveats in Using Wisdom. (line 7) * execute <1>: New-array Execute Functions. (line 6) * execute <2>: Complex One-Dimensional DFTs. (line 91) * execute: Introduction. (line 43) * FFTW: Introduction. (line 33) * fftw-wisdom utility <1>: Wisdom Utilities. (line 15) * fftw-wisdom utility: Caveats in Using Wisdom. (line 40) * fftw-wisdom-to-conf utility: Wisdom Utilities. (line 26) * flags <1>: FFTW Constants in Fortran. (line 19) * flags <2>: Overview of Fortran interface. (line 67) * flags <3>: Guru Real-to-real Transforms. (line 27) * flags <4>: Guru Real-data DFTs. (line 58) * flags <5>: Guru Complex DFTs. (line 25) * flags <6>: Real-to-Real Transforms. (line 78) * flags <7>: Real-data DFTs. (line 69) * flags <8>: Complex DFTs. (line 75) * flags <9>: One-Dimensional DFTs of Real Data. (line 61) * flags: Complex One-Dimensional DFTs. (line 65) * Fortran interface <1>: Calling FFTW from Legacy Fortran. (line 6) * Fortran interface <2>: Calling FFTW from Modern Fortran. (line 6) * Fortran interface <3>: FFTW MPI Fortran Interface. (line 6) * Fortran interface: Column-major Format. (line 20) * Fortran-callable wrappers: Installation on Unix. (line 93) * frequency <1>: The 1d Discrete Fourier Transform (DFT). (line 14) * frequency: Complex One-Dimensional DFTs. (line 87) * g77: Installation on Unix. (line 114) * guru interface <1>: Fortran-interface routines. (line 44) * guru interface <2>: FFTW Fortran type reference. (line 53) * guru interface <3>: Guru Interface. (line 6) * guru interface <4>: Complex Multi-Dimensional DFTs. (line 62) * guru interface: Introduction. (line 67) * halfcomplex format <1>: The 1d Real-data DFT. (line 20) * halfcomplex format <2>: The Halfcomplex-format DFT. (line 9) * halfcomplex format: One-Dimensional DFTs of Real Data. (line 75) * hc2r <1>: Planner Flags. (line 74) * hc2r: The Halfcomplex-format DFT. (line 9) * HDF5: 2d MPI example. (line 90) * Hermitian <1>: The 1d Real-data DFT. (line 9) * Hermitian: One-Dimensional DFTs of Real Data. (line 7) * howmany loop: Guru vector and transform sizes. (line 39) * howmany parameter: Advanced Complex DFTs. (line 30) * IDCT <1>: 1d Real-even DFTs (DCTs). (line 51) * IDCT <2>: Real-to-Real Transform Kinds. (line 40) * IDCT: Real even/odd DFTs (cosine/sine transforms). (line 41) * in-place <1>: FFTW Fortran type reference. (line 37) * in-place <2>: Reversing array dimensions. (line 53) * in-place <3>: An improved replacement for MPI_Alltoall. (line 34) * in-place <4>: Guru Real-data DFTs. (line 48) * in-place <5>: Real-to-Real Transforms. (line 66) * in-place <6>: Real-data DFT Array Format. (line 31) * in-place <7>: Real-data DFTs. (line 62) * in-place <8>: Complex DFTs. (line 62) * in-place <9>: One-Dimensional DFTs of Real Data. (line 47) * in-place: Complex One-Dimensional DFTs. (line 58) * installation: Installation and Customization. (line 6) * interleaved format: Interleaved and split arrays. (line 13) * iso_c_binding <1>: Extended and quadruple precision in Fortran. (line 17) * iso_c_binding <2>: Overview of Fortran interface. (line 12) * iso_c_binding: FFTW MPI Fortran Interface. (line 6) * kind (r2r) <1>: Real-to-Real Transform Kinds. (line 6) * kind (r2r): More DFTs of Real Data. (line 51) * linking on Unix <1>: Linking and Initializing MPI FFTW. (line 11) * linking on Unix: Usage of Multi-threaded FFTW. (line 16) * LISP: Acknowledgments. (line 47) * load balancing <1>: FFTW MPI Performance Tips. (line 14) * load balancing: Load balancing. (line 6) * MIPS PS: SIMD alignment and fftw_malloc. (line 13) * monadic programming: Generating your own code. (line 29) * MPI <1>: Installation on Unix. (line 87) * MPI: Distributed-memory FFTW with MPI. (line 6) * MPI communicator <1>: FFTW MPI Fortran Interface. (line 32) * MPI communicator <2>: MPI Plan Creation. (line 27) * MPI communicator <3>: Using MPI Plans. (line 10) * MPI communicator: Distributed-memory FFTW with MPI. (line 34) * MPI I/O <1>: FFTW MPI Wisdom. (line 10) * MPI I/O: 2d MPI example. (line 90) * mpicc <1>: Linking and Initializing MPI FFTW. (line 11) * mpicc: FFTW MPI Installation. (line 23) * new-array execution <1>: FFTW MPI Fortran Interface. (line 50) * new-array execution <2>: MPI Plan Creation. (line 193) * new-array execution <3>: Using MPI Plans. (line 14) * new-array execution: New-array Execute Functions. (line 6) * normalization <1>: 1d Discrete Hartley Transforms (DHTs). (line 8) * normalization <2>: 1d Real-odd DFTs (DSTs). (line 63) * normalization <3>: 1d Real-even DFTs (DCTs). (line 68) * normalization <4>: The 1d Real-data DFT. (line 29) * normalization <5>: The 1d Discrete Fourier Transform (DFT). (line 9) * normalization <6>: Real-to-Real Transform Kinds. (line 18) * normalization <7>: Real-data DFTs. (line 96) * normalization <8>: Complex DFTs. (line 82) * normalization <9>: The Discrete Hartley Transform. (line 21) * normalization <10>: Real even/odd DFTs (cosine/sine transforms). (line 68) * normalization <11>: The Halfcomplex-format DFT. (line 23) * normalization <12>: Multi-Dimensional DFTs of Real Data. (line 65) * normalization: Complex One-Dimensional DFTs. (line 106) * number of threads: How Many Threads to Use?. (line 6) * OpenMP <1>: Thread safety. (line 6) * OpenMP <2>: Usage of Multi-threaded FFTW. (line 11) * OpenMP: Installation and Supported Hardware/Software. (line 14) * out-of-place <1>: Real-data DFT Array Format. (line 27) * out-of-place: Planner Flags. (line 64) * padding <1>: Reversing array dimensions. (line 53) * padding <2>: Multi-dimensional MPI DFTs of Real Data. (line 23) * padding <3>: Real-data DFT Array Format. (line 31) * padding <4>: Real-data DFTs. (line 67) * padding <5>: Multi-Dimensional DFTs of Real Data. (line 44) * padding: One-Dimensional DFTs of Real Data. (line 17) * parallel transform <1>: Distributed-memory FFTW with MPI. (line 6) * parallel transform: Multi-threaded FFTW. (line 6) * partial order: Complex Multi-Dimensional DFTs. (line 56) * plan <1>: Complex One-Dimensional DFTs. (line 44) * plan: Introduction. (line 42) * planner: Introduction. (line 41) * portability <1>: Installation and Customization. (line 16) * portability <2>: Wisdom of Fortran?. (line 11) * portability <3>: Fortran-interface routines. (line 22) * portability <4>: FFTW Fortran type reference. (line 69) * portability <5>: Calling FFTW from Modern Fortran. (line 14) * portability <6>: Installation and Supported Hardware/Software. (line 14) * portability <7>: Complex numbers. (line 35) * portability <8>: Caveats in Using Wisdom. (line 9) * portability: SIMD alignment and fftw_malloc. (line 22) * precision <1>: Installation on Unix. (line 46) * precision <2>: FFTW Fortran type reference. (line 12) * precision <3>: Extended and quadruple precision in Fortran. (line 6) * precision <4>: MPI Files and Data Types. (line 18) * precision <5>: Linking and Initializing MPI FFTW. (line 11) * precision <6>: Memory Allocation. (line 36) * precision <7>: Precision. (line 6) * precision <8>: SIMD alignment and fftw_malloc. (line 13) * precision <9>: One-Dimensional DFTs of Real Data. (line 41) * precision: Complex One-Dimensional DFTs. (line 118) * r2c <1>: MPI Plan Creation. (line 82) * r2c <2>: Multi-dimensional Transforms. (line 14) * r2c <3>: Real-data DFTs. (line 18) * r2c <4>: The Halfcomplex-format DFT. (line 6) * r2c: One-Dimensional DFTs of Real Data. (line 37) * r2c/c2r multi-dimensional array format <1>: Fortran Examples. (line 93) * r2c/c2r multi-dimensional array format <2>: Reversing array dimensions. (line 37) * r2c/c2r multi-dimensional array format <3>: Real-data DFT Array Format. (line 6) * r2c/c2r multi-dimensional array format: Multi-Dimensional DFTs of Real Data. (line 26) * r2hc: The Halfcomplex-format DFT. (line 6) * r2r <1>: MPI Plan Creation. (line 146) * r2r <2>: Other Multi-dimensional Real-data MPI Transforms. (line 6) * r2r <3>: The 1d Real-data DFT. (line 20) * r2r <4>: Real-to-Real Transforms. (line 6) * r2r: More DFTs of Real Data. (line 13) * rank: Complex Multi-Dimensional DFTs. (line 31) * real-even DFT <1>: 1d Real-even DFTs (DCTs). (line 9) * real-even DFT: Real even/odd DFTs (cosine/sine transforms). (line 16) * real-odd DFT <1>: 1d Real-odd DFTs (DSTs). (line 9) * real-odd DFT: Real even/odd DFTs (cosine/sine transforms). (line 16) * REDFT <1>: Generating your own code. (line 20) * REDFT <2>: 1d Real-even DFTs (DCTs). (line 9) * REDFT: Real even/odd DFTs (cosine/sine transforms). (line 16) * RODFT <1>: 1d Real-odd DFTs (DSTs). (line 9) * RODFT: Real even/odd DFTs (cosine/sine transforms). (line 16) * row-major <1>: Reversing array dimensions. (line 6) * row-major <2>: Multi-dimensional MPI DFTs of Real Data. (line 86) * row-major <3>: Basic and advanced distribution interfaces. (line 32) * row-major <4>: Guru vector and transform sizes. (line 48) * row-major <5>: Real-to-Real Transforms. (line 47) * row-major <6>: Complex DFTs. (line 47) * row-major: Row-major Format. (line 6) * saving plans to disk <1>: Accessing the wisdom API from Fortran. (line 6) * saving plans to disk <2>: FFTW MPI Wisdom. (line 6) * saving plans to disk <3>: Wisdom. (line 6) * saving plans to disk: Words of Wisdom-Saving Plans. (line 6) * shared-memory: Multi-threaded FFTW. (line 24) * SIMD <1>: Overview of Fortran interface. (line 50) * SIMD <2>: SIMD alignment and fftw_malloc. (line 13) * SIMD: Complex One-Dimensional DFTs. (line 38) * split format: Interleaved and split arrays. (line 16) * SSE: SIMD alignment and fftw_malloc. (line 13) * SSE2: SIMD alignment and fftw_malloc. (line 13) * stride <1>: MPI Plan Creation. (line 43) * stride <2>: Guru vector and transform sizes. (line 28) * stride <3>: Advanced Complex DFTs. (line 48) * stride: Row-major Format. (line 31) * thread safety <1>: Combining MPI and Threads. (line 53) * thread safety <2>: Thread safety. (line 6) * thread safety: Usage of Multi-threaded FFTW. (line 49) * threads <1>: Installation on Unix. (line 64) * threads <2>: Combining MPI and Threads. (line 6) * threads <3>: Thread safety. (line 6) * threads: Multi-threaded FFTW. (line 24) * transpose <1>: MPI Plan Creation. (line 181) * transpose <2>: Combining MPI and Threads. (line 75) * transpose <3>: FFTW MPI Performance Tips. (line 19) * transpose <4>: FFTW MPI Transposes. (line 6) * transpose <5>: Multi-dimensional MPI DFTs of Real Data. (line 93) * transpose: Transposed distributions. (line 15) * vector: Guru Interface. (line 10) * wisdom <1>: Accessing the wisdom API from Fortran. (line 6) * wisdom <2>: FFTW MPI Wisdom. (line 6) * wisdom <3>: Wisdom. (line 6) * wisdom: Words of Wisdom-Saving Plans. (line 6) * wisdom, problems with: Caveats in Using Wisdom. (line 6) * wisdom, system-wide <1>: Wisdom Import. (line 37) * wisdom, system-wide: Caveats in Using Wisdom. (line 33)  File: fftw3.info, Node: Library Index, Prev: Concept Index, Up: Top 14 Library Index **************** [index] * Menu: * c_associated: Wisdom String Export/Import from Fortran. (line 36) * C_DOUBLE <1>: FFTW Fortran type reference. (line 12) * C_DOUBLE: Overview of Fortran interface. (line 34) * C_DOUBLE_COMPLEX <1>: FFTW Fortran type reference. (line 12) * C_DOUBLE_COMPLEX: Overview of Fortran interface. (line 34) * c_f_pointer <1>: Wisdom Generic Export/Import from Fortran. (line 27) * c_f_pointer <2>: Wisdom String Export/Import from Fortran. (line 36) * c_f_pointer <3>: Allocating aligned memory in Fortran. (line 37) * c_f_pointer <4>: FFTW Fortran type reference. (line 48) * c_f_pointer: Reversing array dimensions. (line 63) * C_FFTW_R2R_KIND: FFTW Fortran type reference. (line 32) * C_FLOAT: FFTW Fortran type reference. (line 12) * C_FLOAT_COMPLEX: FFTW Fortran type reference. (line 12) * c_funloc: Wisdom Generic Export/Import from Fortran. (line 25) * C_FUNPTR: FFTW Fortran type reference. (line 61) * C_INT <1>: FFTW Fortran type reference. (line 27) * C_INT: Overview of Fortran interface. (line 34) * C_INTPTR_T: FFTW Fortran type reference. (line 27) * c_loc: Wisdom Generic Export/Import from Fortran. (line 27) * C_LONG_DOUBLE: FFTW Fortran type reference. (line 12) * C_LONG_DOUBLE_COMPLEX: FFTW Fortran type reference. (line 12) * C_PTR: Overview of Fortran interface. (line 34) * C_SIZE_T: FFTW Fortran type reference. (line 27) * dfftw_destroy_plan: Fortran Examples. (line 25) * dfftw_execute: FFTW Execution in Fortran. (line 11) * dfftw_execute_dft <1>: Fortran Examples. (line 25) * dfftw_execute_dft: FFTW Execution in Fortran. (line 28) * dfftw_execute_dft_r2c: Fortran Examples. (line 75) * dfftw_export_wisdom: Wisdom of Fortran?. (line 16) * dfftw_forget_wisdom: Wisdom of Fortran?. (line 16) * dfftw_import_system_wisdom: Wisdom of Fortran?. (line 16) * dfftw_import_wisdom: Wisdom of Fortran?. (line 16) * dfftw_init_threads: Fortran Examples. (line 36) * dfftw_plan_dft_1d: Fortran Examples. (line 25) * dfftw_plan_dft_3d: Fortran Examples. (line 60) * dfftw_plan_dft_r2c_1d: Fortran Examples. (line 75) * dfftw_plan_dft_r2c_2d: Fortran Examples. (line 88) * dfftw_plan_with_nthreads: Fortran Examples. (line 36) * fftw_alloc_complex <1>: Allocating aligned memory in Fortran. (line 6) * fftw_alloc_complex <2>: Reversing array dimensions. (line 63) * fftw_alloc_complex <3>: Basic and advanced distribution interfaces. (line 43) * fftw_alloc_complex <4>: Memory Allocation. (line 33) * fftw_alloc_complex <5>: SIMD alignment and fftw_malloc. (line 35) * fftw_alloc_complex: Complex One-Dimensional DFTs. (line 38) * fftw_alloc_real <1>: Allocating aligned memory in Fortran. (line 6) * fftw_alloc_real <2>: FFTW Fortran type reference. (line 48) * fftw_alloc_real <3>: Other Multi-dimensional Real-data MPI Transforms. (line 44) * fftw_alloc_real <4>: Multi-dimensional MPI DFTs of Real Data. (line 86) * fftw_alloc_real <5>: Memory Allocation. (line 33) * fftw_alloc_real: SIMD alignment and fftw_malloc. (line 35) * FFTW_BACKWARD <1>: Complex DFTs. (line 71) * FFTW_BACKWARD <2>: One-Dimensional DFTs of Real Data. (line 39) * FFTW_BACKWARD: Complex One-Dimensional DFTs. (line 61) * fftw_cleanup <1>: MPI Initialization. (line 21) * fftw_cleanup: Using Plans. (line 36) * fftw_cleanup_threads: Usage of Multi-threaded FFTW. (line 61) * fftw_complex <1>: FFTW Fortran type reference. (line 12) * fftw_complex <2>: Overview of Fortran interface. (line 34) * fftw_complex <3>: Complex numbers. (line 11) * fftw_complex: Complex One-Dimensional DFTs. (line 42) * fftw_cost: Using Plans. (line 61) * FFTW_DESTROY_INPUT <1>: FFTW Fortran type reference. (line 37) * FFTW_DESTROY_INPUT <2>: FFTW MPI Performance Tips. (line 24) * FFTW_DESTROY_INPUT: Planner Flags. (line 62) * fftw_destroy_plan <1>: Overview of Fortran interface. (line 24) * fftw_destroy_plan <2>: Avoiding MPI Deadlocks. (line 19) * fftw_destroy_plan <3>: Using Plans. (line 27) * fftw_destroy_plan: Complex One-Dimensional DFTs. (line 100) * FFTW_DHT <1>: Real-to-Real Transform Kinds. (line 28) * FFTW_DHT: The Discrete Hartley Transform. (line 17) * FFTW_ESTIMATE <1>: Cycle Counters. (line 18) * FFTW_ESTIMATE <2>: Planner Flags. (line 28) * FFTW_ESTIMATE <3>: Words of Wisdom-Saving Plans. (line 22) * FFTW_ESTIMATE: Complex One-Dimensional DFTs. (line 71) * fftw_execute <1>: Plan execution in Fortran. (line 11) * fftw_execute <2>: Overview of Fortran interface. (line 57) * fftw_execute <3>: Avoiding MPI Deadlocks. (line 19) * fftw_execute <4>: Basic distributed-transpose interface. (line 17) * fftw_execute <5>: New-array Execute Functions. (line 8) * fftw_execute <6>: Using Plans. (line 13) * fftw_execute: Complex One-Dimensional DFTs. (line 85) * fftw_execute_dft <1>: Plan execution in Fortran. (line 28) * fftw_execute_dft <2>: Overview of Fortran interface. (line 24) * fftw_execute_dft <3>: FFTW MPI Fortran Interface. (line 50) * fftw_execute_dft: New-array Execute Functions. (line 80) * fftw_execute_dft_c2r <1>: Plan execution in Fortran. (line 31) * fftw_execute_dft_c2r: New-array Execute Functions. (line 80) * fftw_execute_dft_r2c <1>: Plan execution in Fortran. (line 31) * fftw_execute_dft_r2c <2>: Reversing array dimensions. (line 46) * fftw_execute_dft_r2c: New-array Execute Functions. (line 80) * fftw_execute_r2r <1>: Plan execution in Fortran. (line 31) * fftw_execute_r2r: New-array Execute Functions. (line 80) * fftw_execute_split_dft: New-array Execute Functions. (line 80) * fftw_execute_split_dft_c2r: New-array Execute Functions. (line 80) * fftw_execute_split_dft_r2c: New-array Execute Functions. (line 80) * FFTW_EXHAUSTIVE <1>: Planner Flags. (line 43) * FFTW_EXHAUSTIVE: Words of Wisdom-Saving Plans. (line 22) * fftw_export_wisdom <1>: Wisdom Generic Export/Import from Fortran. (line 12) * fftw_export_wisdom: Wisdom Export. (line 10) * fftw_export_wisdom_to_file: Wisdom Export. (line 10) * fftw_export_wisdom_to_filename <1>: Wisdom File Export/Import from Fortran. (line 6) * fftw_export_wisdom_to_filename <2>: Wisdom Export. (line 10) * fftw_export_wisdom_to_filename: Words of Wisdom-Saving Plans. (line 29) * fftw_export_wisdom_to_string <1>: Wisdom String Export/Import from Fortran. (line 6) * fftw_export_wisdom_to_string: Wisdom Export. (line 10) * fftw_flops <1>: FFTW Fortran type reference. (line 37) * fftw_flops <2>: Avoiding MPI Deadlocks. (line 19) * fftw_flops: Using Plans. (line 67) * fftw_forget_wisdom <1>: Forgetting Wisdom. (line 7) * fftw_forget_wisdom: Words of Wisdom-Saving Plans. (line 48) * FFTW_FORWARD <1>: Complex DFTs. (line 71) * FFTW_FORWARD <2>: One-Dimensional DFTs of Real Data. (line 39) * FFTW_FORWARD: Complex One-Dimensional DFTs. (line 61) * fftw_fprint_plan: Using Plans. (line 81) * fftw_free <1>: Memory Allocation. (line 8) * fftw_free <2>: SIMD alignment and fftw_malloc. (line 25) * fftw_free: Complex One-Dimensional DFTs. (line 101) * FFTW_HC2R <1>: Real-to-Real Transform Kinds. (line 25) * FFTW_HC2R: The Halfcomplex-format DFT. (line 9) * fftw_import wisdom_from_filename: Wisdom File Export/Import from Fortran. (line 6) * fftw_import_system_wisdom <1>: Wisdom Import. (line 10) * fftw_import_system_wisdom: Caveats in Using Wisdom. (line 36) * fftw_import_wisdom <1>: Wisdom Generic Export/Import from Fortran. (line 37) * fftw_import_wisdom: Wisdom Import. (line 10) * fftw_import_wisdom_from_file: Wisdom Import. (line 10) * fftw_import_wisdom_from_filename <1>: Wisdom Import. (line 10) * fftw_import_wisdom_from_filename: Words of Wisdom-Saving Plans. (line 35) * fftw_import_wisdom_from_string <1>: Wisdom String Export/Import from Fortran. (line 44) * fftw_import_wisdom_from_string: Wisdom Import. (line 10) * fftw_init_threads <1>: MPI Initialization. (line 11) * fftw_init_threads <2>: Combining MPI and Threads. (line 35) * fftw_init_threads <3>: Linking and Initializing MPI FFTW. (line 13) * fftw_init_threads: Usage of Multi-threaded FFTW. (line 22) * fftw_iodim <1>: Fortran-interface routines. (line 44) * fftw_iodim <2>: FFTW Fortran type reference. (line 53) * fftw_iodim: Guru vector and transform sizes. (line 15) * fftw_iodim64 <1>: FFTW Fortran type reference. (line 53) * fftw_iodim64: 64-bit Guru Interface. (line 46) * fftw_malloc <1>: FFTW Fortran type reference. (line 27) * fftw_malloc <2>: Using MPI Plans. (line 24) * fftw_malloc <3>: Basic and advanced distribution interfaces. (line 43) * fftw_malloc <4>: Memory Allocation. (line 8) * fftw_malloc <5>: Dynamic Arrays in C. (line 15) * fftw_malloc <6>: SIMD alignment and fftw_malloc. (line 25) * fftw_malloc: Complex One-Dimensional DFTs. (line 34) * FFTW_MEASURE <1>: An improved replacement for MPI_Alltoall. (line 43) * FFTW_MEASURE <2>: Planner Flags. (line 33) * FFTW_MEASURE <3>: Words of Wisdom-Saving Plans. (line 22) * FFTW_MEASURE: Complex One-Dimensional DFTs. (line 66) * fftw_mpi_broadcast_wisdom <1>: MPI Wisdom Communication. (line 9) * fftw_mpi_broadcast_wisdom: FFTW MPI Wisdom. (line 29) * fftw_mpi_cleanup <1>: MPI Initialization. (line 19) * fftw_mpi_cleanup: Linking and Initializing MPI FFTW. (line 24) * FFTW_MPI_DEFAULT_BLOCK <1>: MPI Plan Creation. (line 43) * FFTW_MPI_DEFAULT_BLOCK <2>: Advanced distributed-transpose interface. (line 23) * FFTW_MPI_DEFAULT_BLOCK: Basic and advanced distribution interfaces. (line 72) * fftw_mpi_execute_dft <1>: FFTW MPI Fortran Interface. (line 50) * fftw_mpi_execute_dft: Using MPI Plans. (line 19) * fftw_mpi_execute_dft_c2r: Using MPI Plans. (line 19) * fftw_mpi_execute_dft_r2c: Using MPI Plans. (line 19) * fftw_mpi_execute_r2r <1>: MPI Plan Creation. (line 193) * fftw_mpi_execute_r2r: Using MPI Plans. (line 19) * fftw_mpi_gather_wisdom <1>: MPI Wisdom Communication. (line 9) * fftw_mpi_gather_wisdom: FFTW MPI Wisdom. (line 29) * fftw_mpi_init <1>: MPI Initialization. (line 9) * fftw_mpi_init <2>: Combining MPI and Threads. (line 35) * fftw_mpi_init <3>: FFTW MPI Wisdom. (line 55) * fftw_mpi_init <4>: 2d MPI example. (line 50) * fftw_mpi_init: Linking and Initializing MPI FFTW. (line 18) * fftw_mpi_local_size: MPI Data Distribution Functions. (line 16) * fftw_mpi_local_size_1d <1>: MPI Data Distribution Functions. (line 66) * fftw_mpi_local_size_1d: One-dimensional distributions. (line 18) * fftw_mpi_local_size_2d <1>: MPI Data Distribution Functions. (line 16) * fftw_mpi_local_size_2d <2>: Basic and advanced distribution interfaces. (line 19) * fftw_mpi_local_size_2d: 2d MPI example. (line 81) * fftw_mpi_local_size_2d_transposed <1>: MPI Data Distribution Functions. (line 16) * fftw_mpi_local_size_2d_transposed: Basic distributed-transpose interface. (line 39) * fftw_mpi_local_size_3d: MPI Data Distribution Functions. (line 16) * fftw_mpi_local_size_3d_transposed <1>: MPI Data Distribution Functions. (line 16) * fftw_mpi_local_size_3d_transposed: Transposed distributions. (line 53) * fftw_mpi_local_size_many <1>: MPI Data Distribution Functions. (line 47) * fftw_mpi_local_size_many: Basic and advanced distribution interfaces. (line 55) * fftw_mpi_local_size_many_1d: MPI Data Distribution Functions. (line 66) * fftw_mpi_local_size_many_transposed <1>: MPI Data Distribution Functions. (line 47) * fftw_mpi_local_size_many_transposed: Advanced distributed-transpose interface. (line 23) * fftw_mpi_local_size_transposed: MPI Data Distribution Functions. (line 16) * fftw_mpi_plan_dft: MPI Plan Creation. (line 11) * fftw_mpi_plan_dft_1d: MPI Plan Creation. (line 11) * fftw_mpi_plan_dft_2d <1>: MPI Plan Creation. (line 11) * fftw_mpi_plan_dft_2d: 2d MPI example. (line 61) * fftw_mpi_plan_dft_3d: MPI Plan Creation. (line 11) * fftw_mpi_plan_dft_c2r: MPI Plan Creation. (line 85) * fftw_mpi_plan_dft_c2r_2d: MPI Plan Creation. (line 85) * fftw_mpi_plan_dft_c2r_3d: MPI Plan Creation. (line 85) * fftw_mpi_plan_dft_r2c: MPI Plan Creation. (line 85) * fftw_mpi_plan_dft_r2c_2d: MPI Plan Creation. (line 85) * fftw_mpi_plan_dft_r2c_3d: MPI Plan Creation. (line 85) * fftw_mpi_plan_many_dft: MPI Plan Creation. (line 11) * fftw_mpi_plan_many_dft_c2r: MPI Plan Creation. (line 132) * fftw_mpi_plan_many_dft_r2c: MPI Plan Creation. (line 132) * fftw_mpi_plan_many_transpose <1>: MPI Plan Creation. (line 185) * fftw_mpi_plan_many_transpose: Advanced distributed-transpose interface. (line 15) * fftw_mpi_plan_transpose <1>: MPI Plan Creation. (line 185) * fftw_mpi_plan_transpose: Basic distributed-transpose interface. (line 14) * FFTW_MPI_SCRAMBLED_IN <1>: MPI Plan Creation. (line 65) * FFTW_MPI_SCRAMBLED_IN <2>: MPI Data Distribution Functions. (line 76) * FFTW_MPI_SCRAMBLED_IN: One-dimensional distributions. (line 43) * FFTW_MPI_SCRAMBLED_OUT <1>: MPI Plan Creation. (line 65) * FFTW_MPI_SCRAMBLED_OUT <2>: MPI Data Distribution Functions. (line 76) * FFTW_MPI_SCRAMBLED_OUT: One-dimensional distributions. (line 43) * FFTW_MPI_TRANSPOSED_IN <1>: MPI Plan Creation. (line 72) * FFTW_MPI_TRANSPOSED_IN <2>: Basic distributed-transpose interface. (line 28) * FFTW_MPI_TRANSPOSED_IN: Transposed distributions. (line 26) * FFTW_MPI_TRANSPOSED_OUT <1>: MPI Plan Creation. (line 72) * FFTW_MPI_TRANSPOSED_OUT <2>: Basic distributed-transpose interface. (line 28) * FFTW_MPI_TRANSPOSED_OUT: Transposed distributions. (line 26) * FFTW_NO_TIMELIMIT: Planner Flags. (line 96) * FFTW_PATIENT <1>: An improved replacement for MPI_Alltoall. (line 43) * FFTW_PATIENT <2>: How Many Threads to Use?. (line 20) * FFTW_PATIENT <3>: Planner Flags. (line 38) * FFTW_PATIENT <4>: Words of Wisdom-Saving Plans. (line 22) * FFTW_PATIENT: Complex One-Dimensional DFTs. (line 122) * fftw_plan <1>: FFTW Fortran type reference. (line 9) * fftw_plan <2>: Using Plans. (line 8) * fftw_plan: Complex One-Dimensional DFTs. (line 50) * fftw_plan_dft <1>: Complex DFTs. (line 18) * fftw_plan_dft: Complex Multi-Dimensional DFTs. (line 39) * fftw_plan_dft_1d <1>: Complex DFTs. (line 18) * fftw_plan_dft_1d: Complex One-Dimensional DFTs. (line 50) * fftw_plan_dft_2d <1>: Overview of Fortran interface. (line 24) * fftw_plan_dft_2d <2>: Complex DFTs. (line 18) * fftw_plan_dft_2d: Complex Multi-Dimensional DFTs. (line 21) * fftw_plan_dft_3d <1>: Reversing array dimensions. (line 22) * fftw_plan_dft_3d <2>: Complex DFTs. (line 18) * fftw_plan_dft_3d: Complex Multi-Dimensional DFTs. (line 21) * fftw_plan_dft_c2r: Real-data DFTs. (line 89) * fftw_plan_dft_c2r_1d <1>: Real-data DFTs. (line 89) * fftw_plan_dft_c2r_1d: One-Dimensional DFTs of Real Data. (line 35) * fftw_plan_dft_c2r_2d: Real-data DFTs. (line 89) * fftw_plan_dft_c2r_3d: Real-data DFTs. (line 89) * fftw_plan_dft_r2c <1>: Real-data DFTs. (line 18) * fftw_plan_dft_r2c: Multi-Dimensional DFTs of Real Data. (line 17) * fftw_plan_dft_r2c_1d <1>: Real-data DFTs. (line 18) * fftw_plan_dft_r2c_1d: One-Dimensional DFTs of Real Data. (line 35) * fftw_plan_dft_r2c_2d <1>: Real-data DFTs. (line 18) * fftw_plan_dft_r2c_2d: Multi-Dimensional DFTs of Real Data. (line 17) * fftw_plan_dft_r2c_3d <1>: Reversing array dimensions. (line 46) * fftw_plan_dft_r2c_3d <2>: Real-data DFTs. (line 18) * fftw_plan_dft_r2c_3d: Multi-Dimensional DFTs of Real Data. (line 17) * fftw_plan_guru64_dft: 64-bit Guru Interface. (line 36) * fftw_plan_guru_dft: Guru Complex DFTs. (line 17) * fftw_plan_guru_dft_c2r: Guru Real-data DFTs. (line 29) * fftw_plan_guru_dft_r2c: Guru Real-data DFTs. (line 29) * fftw_plan_guru_r2r: Guru Real-to-real Transforms. (line 12) * fftw_plan_guru_split_dft: Guru Complex DFTs. (line 17) * fftw_plan_guru_split_dft_c2r: Guru Real-data DFTs. (line 29) * fftw_plan_guru_split_dft_r2c: Guru Real-data DFTs. (line 29) * fftw_plan_many_dft: Advanced Complex DFTs. (line 12) * fftw_plan_many_dft_c2r: Advanced Real-data DFTs. (line 18) * fftw_plan_many_dft_r2c: Advanced Real-data DFTs. (line 18) * fftw_plan_many_r2r: Advanced Real-to-real Transforms. (line 12) * fftw_plan_r2r <1>: Real-to-Real Transforms. (line 19) * fftw_plan_r2r: More DFTs of Real Data. (line 39) * fftw_plan_r2r_1d <1>: Real-to-Real Transforms. (line 19) * fftw_plan_r2r_1d: More DFTs of Real Data. (line 39) * fftw_plan_r2r_2d <1>: Real-to-Real Transforms. (line 19) * fftw_plan_r2r_2d: More DFTs of Real Data. (line 39) * fftw_plan_r2r_3d <1>: Real-to-Real Transforms. (line 19) * fftw_plan_r2r_3d: More DFTs of Real Data. (line 39) * fftw_plan_with_nthreads <1>: Combining MPI and Threads. (line 35) * fftw_plan_with_nthreads: Usage of Multi-threaded FFTW. (line 32) * FFTW_PRESERVE_INPUT <1>: Planner Flags. (line 66) * FFTW_PRESERVE_INPUT: One-Dimensional DFTs of Real Data. (line 61) * fftw_print_plan: Using Plans. (line 81) * FFTW_R2HC <1>: Real-to-Real Transform Kinds. (line 20) * FFTW_R2HC: The Halfcomplex-format DFT. (line 6) * fftw_r2r_kind <1>: FFTW Fortran type reference. (line 32) * fftw_r2r_kind <2>: Other Multi-dimensional Real-data MPI Transforms. (line 12) * fftw_r2r_kind: More DFTs of Real Data. (line 51) * FFTW_REDFT00 <1>: Real-to-Real Transform Kinds. (line 31) * FFTW_REDFT00 <2>: Real-to-Real Transforms. (line 32) * FFTW_REDFT00: Real even/odd DFTs (cosine/sine transforms). (line 35) * FFTW_REDFT01 <1>: Real-to-Real Transform Kinds. (line 38) * FFTW_REDFT01: Real even/odd DFTs (cosine/sine transforms). (line 41) * FFTW_REDFT10 <1>: Real-to-Real Transform Kinds. (line 34) * FFTW_REDFT10: Real even/odd DFTs (cosine/sine transforms). (line 38) * FFTW_REDFT11 <1>: Real-to-Real Transform Kinds. (line 42) * FFTW_REDFT11: Real even/odd DFTs (cosine/sine transforms). (line 43) * FFTW_RODFT00 <1>: Real-to-Real Transform Kinds. (line 45) * FFTW_RODFT00: Real even/odd DFTs (cosine/sine transforms). (line 45) * FFTW_RODFT01 <1>: Real-to-Real Transform Kinds. (line 51) * FFTW_RODFT01: Real even/odd DFTs (cosine/sine transforms). (line 49) * FFTW_RODFT10 <1>: Real-to-Real Transform Kinds. (line 48) * FFTW_RODFT10: Real even/odd DFTs (cosine/sine transforms). (line 47) * FFTW_RODFT11 <1>: Real-to-Real Transform Kinds. (line 54) * FFTW_RODFT11: Real even/odd DFTs (cosine/sine transforms). (line 51) * fftw_set_timelimit: Planner Flags. (line 90) * FFTW_TRANSPOSED_IN: Multi-dimensional MPI DFTs of Real Data. (line 93) * FFTW_TRANSPOSED_OUT: Multi-dimensional MPI DFTs of Real Data. (line 93) * FFTW_UNALIGNED <1>: FFTW Execution in Fortran. (line 52) * FFTW_UNALIGNED <2>: Plan execution in Fortran. (line 54) * FFTW_UNALIGNED <3>: New-array Execute Functions. (line 27) * FFTW_UNALIGNED: Planner Flags. (line 76) * FFTW_WISDOM_ONLY: Planner Flags. (line 48) * MPI_Alltoall: An improved replacement for MPI_Alltoall. (line 9) * MPI_Barrier: Avoiding MPI Deadlocks. (line 12) * MPI_COMM_WORLD <1>: 2d MPI example. (line 61) * MPI_COMM_WORLD: Distributed-memory FFTW with MPI. (line 34) * MPI_Init: 2d MPI example. (line 50) * ptrdiff_t <1>: FFTW Fortran type reference. (line 27) * ptrdiff_t <2>: 2d MPI example. (line 69) * ptrdiff_t: 64-bit Guru Interface. (line 23) * R2HC: The 1d Real-data DFT. (line 20) * REDFT00: 1d Real-even DFTs (DCTs). (line 11) * REDFT01: 1d Real-even DFTs (DCTs). (line 47) * REDFT10: 1d Real-even DFTs (DCTs). (line 40) * REDFT11: 1d Real-even DFTs (DCTs). (line 56) * RODFT00: 1d Real-odd DFTs (DSTs). (line 11) * RODFT01: 1d Real-odd DFTs (DSTs). (line 44) * RODFT10: 1d Real-odd DFTs (DSTs). (line 38) * RODFT11: 1d Real-odd DFTs (DSTs). (line 51) fftw-3.3.3/tools/0000755000175000017500000000000012054411114010620 500000000000000fftw-3.3.3/tools/fftw-wisdom-to-conf.10000644000175000017500000000676412054410050014446 00000000000000.\" .\" Copyright (c) 2003, 2007-11 Matteo Frigo .\" Copyright (c) 2003, 2007-11 Massachusetts Institute of Technology .\" .\" This program is free software; you can redistribute it and/or modify .\" it under the terms of the GNU General Public License as published by .\" the Free Software Foundation; either version 2 of the License, or .\" (at your option) any later version. .\" .\" This program is distributed in the hope that it will be useful, .\" but WITHOUT ANY WARRANTY; without even the implied warranty of .\" MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the .\" GNU General Public License for more details. .\" .\" You should have received a copy of the GNU General Public License .\" along with this program; if not, write to the Free Software .\" Foundation, Inc., 51 Franklin Street, Fifth Floor, Boston, MA 02110-1301 USA .\" .TH FFTW-WISDOM-TO-CONF 1 "February, 2003" "fftw" "fftw" .SH NAME fftw-wisdom-to-conf \- generate FFTW wisdom (pre-planned transforms) .SH SYNOPSIS \fBfftw-wisdom-to-conf\fR [< \fIINPUT\fR] [> \fIOUTPUT\fR] .SH DESCRIPTION .PP ." Add any additional description here .I fftw-wisdom-to-conf is a utility to generate C .B configuration routines from FFTW .B wisdom files, where the latter contain saved information about how to optimally compute (Fourier) transforms of various sizes. A configuration routine is a C subroutine that you link into your program, replacing a routine of the same name in the FFTW library, that determines which parts of FFTW are callable by your program. The reason to do this is that, if you only need transforms of a limited set of sizes and types, and if you are statically linking your program, then using a configuration file generated from wisdom for those types can substantially reduce the size of your executable. (Otherwise, because of FFTW's dynamic nature, all of FFTW's transform code must be linked into any program using FFTW.) FFTW is a free library to compute discrete Fourier transforms in one or more dimensions, for arbitrary sizes, and of both real and complex data, among other related operations. More information on FFTW can be found at the FFTW home page: .I http://www.fftw.org .I fftw-wisdom-to-conf reads wisdom from standard input and writes the configuration to standard output. It can easily be combined with the .I fftw-wisdom tool, for example: fftw-wisdom -n cof1024 cob1024 -o wisdom .br fftw-wisdom-to-conf < wisdom > conf.c will create a configuration "conf.c" containing only those parts of FFTW needed for the optimized complex forwards and backwards out-of-place transforms of size 1024 (also saving the wisdom itself in "wisdom"). Alternatively, you can run your actual program, export wisdom for all plans that were created (ideally in FFTW_PATIENT or FFTW_EXHAUSTIVE mode), use this as input for \fIfftw-wisdom-to-conf\fR, and then re-link your program with the resulting configuration routine. Note that the configuration routine does not contain the wisdom, only the routines necessary to implement the wisdom, so your program should also import the wisdom in order to benefit from the pre-optimized plans. .SH OPTIONS .TP \fB\-h\fR, \fB\--help\fR Display help on the command-line options and usage. .TP \fB\-V\fR, \fB\--version\fR Print the version number and copyright information. .SH BUGS Send bug reports to fftw@fftw.org. .SH AUTHORS Written by Steven G. Johnson and Matteo Frigo. Copyright (c) 2003, 2007-11 Matteo Frigo .br Copyright (c) 2003, 2007-11 Massachusetts Institute of Technology .SH "SEE ALSO" fftw-wisdom(1) fftw-3.3.3/tools/fftwf-wisdom.10000644000175000017500000001507512054411105013246 00000000000000.\" .\" Copyright (c) 2003, 2007-11 Matteo Frigo .\" Copyright (c) 2003, 2007-11 Massachusetts Institute of Technology .\" .\" This program is free software; you can redistribute it and/or modify .\" it under the terms of the GNU General Public License as published by .\" the Free Software Foundation; either version 2 of the License, or .\" (at your option) any later version. .\" .\" This program is distributed in the hope that it will be useful, .\" but WITHOUT ANY WARRANTY; without even the implied warranty of .\" MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the .\" GNU General Public License for more details. .\" .\" You should have received a copy of the GNU General Public License .\" along with this program; if not, write to the Free Software .\" Foundation, Inc., 51 Franklin Street, Fifth Floor, Boston, MA 02110-1301 USA .\" .TH FFTW-WISDOM 1 "February, 2003" "fftw" "fftw" .SH NAME fftwf-wisdom \- create wisdom (pre-optimized FFTs) .SH SYNOPSIS .B fftwf-wisdom [\fIOPTION\fR]... [\fISIZE\fR]... .SH DESCRIPTION .PP ." Add any additional description here .I fftwf-wisdom is a utility to generate FFTW .B wisdom files, which contain saved information about how to optimally compute (Fourier) transforms of various sizes. FFTW is a free library to compute discrete Fourier transforms in one or more dimensions, for arbitrary sizes, and of both real and complex data, among other related operations. More information on FFTW can be found at the FFTW home page: .I http://www.fftw.org Programs using FFTW can be written to load wisdom from an arbitrary file, string, or other source. Moreover, it is likely that many FFTW-using programs will load the \fBsystem wisdom\fR file, which is stored in .I /etc/fftw/wisdomf by default. .I fftwf-wisdom can be used to create or add to such wisdom files. In its most typical usage, the wisdom file can be created to pre-plan a canonical set of sizes (see below) via: .ce fftwf-wisdom -v -c -o wisdomf (this will take many hours, which can be limited by the .B -t option) and the output .I wisdomf file can then be copied (as root) to .I /etc/fftw/ or whatever. The .I fftwf-wisdom program normally writes the wisdom directly to standard output, but this can be changed via the .B -o option, as in the example above. If the system wisdom file .I /etc/fftw/wisdomf already exists, then .I fftwf-wisdom reads this existing wisdom (unless the .B -n option is specified) and outputs both the old wisdom and any newly created wisdom. In this way, it can be used to add new transform sizes to the existing system wisdom (or other wisdom file, with the .B -w option). .SH SPECIFYING SIZES Although a canonical set of sizes to optimize is specified by the .B -c option, the user can also specify zero or more non-canonical transform sizes and types to optimize, via the .I SIZE arguments following the option flags. Alternatively, the sizes to optimize can be read from standard input (whitespace-separated), if a .I SIZE argument of "-" is supplied. Sizes are specified by the syntax: .ce <\fItype\fR><\fIinplace\fR><\fIdirection\fR><\fIgeometry\fR> <\fItype\fR> is either \'c\' (complex), \'r\' (real, r2c/c2r), or \'k\' (r2r, per-dimension kinds, specified in the geometry, below). <\fIinplace\fR> is either \'i\' (in place) or \'o\' (out of place). <\fIdirection\fR> is either \'f\' (forward) or \'b\' (backward). The <\fIdirection\fR> should be omitted for \'k\' transforms, where it is specified via the geometry instead. <\fIgeometry\fR> is the size and dimensionality of the transform, where different dimensions are separated by \'x\' (e.g. \'16x32\' for a two-dimensional 16 by 32 transform). In the case of \'k\' transforms, the size of each dimension is followed by a "type" string, which can be one of f/b/h/e00/e01/e10/e11/o00/o01/o10/o11 for R2HC/HC2R/DHT/REDFT00/.../RODFT11, respectively, as defined in the FFTW manual. For example, \'cif12x13x14\' is a three-dimensional 12 by 13 x 14 complex DFT operating in-place. \'rob65536\' is a one-dimensional size-65536 out-of-place complex-to-real (backwards) transform operating on Hermitian-symmetry input. \'ki10hx20e01\' is a two-dimensional 10 by 20 r2r transform where the first dimension is a DHT and the second dimension is an REDFT01 (DCT-III). .SH OPTIONS .TP \fB\-h\fR, \fB\--help\fR Display help on the command-line options and usage. .TP \fB\-V\fR, \fB\--version\fR Print the version number and copyright information. .TP \fB\-v\fR, \fB\--verbose\fR Verbose output. (You can specify this multiple times, or supply a numeric argument greater than 1, to increase the verbosity level.) Note that the verbose output will be mixed with the wisdom output (making it impossible to import), unless you write the wisdom to a file via the .B -o option. .TP \fB\-c\fR, \fB\--canonical\fR Optimize/pre-plan a canonical set of sizes: all powers of two and ten up to 2^20 (1048576), including both real and complex, forward and backwards, in-place and out-of-place transforms. Also includes two- and three-dimensional transforms of equal-size dimensions (e.g. 16x16x16). .TP \fB\-t\fR \fIhours\fR, \fB\--time-limit\fR=\fIhours\fR Stop after a time of .I hours (hours) has elapsed, outputting accumulated wisdom. (The problems are planned in increasing order of size.) Defaults to 0, indicating no time limit. .TP \fB\-o\fR \fIfile\fR, \fB\--output-file\fR=\fIfile\fR Send wisdom output to .I file rather than to standard output (the default). .TP \fB\-m\fR, \fB\--measure\fR; \fB\-e\fR, \fB\--estimate\fR; \fB\-x\fR, \fB\--exhaustive\fR Normally, .I fftwf-wisdom creates plans in FFTW_PATIENT mode, but with these options you can instead use FFTW_MEASURE, FFTW_ESTIMATE, or FFTW_EXHAUSTIVE modes, respectively, as described in more detail by the FFTW manual. Note that wisdom is tagged with the planning patience level, and a single file can mix different levels of wisdom (e.g. you can mostly use the patient default, but plan a few sizes that you especially care about in .B --exhaustive mode). .TP \fB\-n\fR, \fB\--no-system-wisdom\fR Do not import the system wisdom from .I /etc/fftw/wisdomf (which is normally read by default). .TP \fB\-w\fR \fIfile\fR, \fB\--wisdom-file\fR=\fIfile\fR Import wisdom from .I file (in addition to the system wisdom, unless .B -n is specified). Multiple wisdom files can be read via multiple .B -w options. If .I file is "-", then read wisdom from standard input. .SH BUGS Send bug reports to fftw@fftw.org. .SH AUTHORS Written by Steven G. Johnson and Matteo Frigo. Copyright (c) 2003, 2007-11 Matteo Frigo .br Copyright (c) 2003, 2007-11 Massachusetts Institute of Technology .SH "SEE ALSO" fftw-wisdom-to-conf(1) fftw-3.3.3/tools/Makefile.am0000644000175000017500000000174512054410050012601 00000000000000AM_CPPFLAGS = -I$(top_srcdir)/libbench2 -I$(top_srcdir)/api bin_SCRIPTS = fftw-wisdom-to-conf bin_PROGRAMS = fftw@PREC_SUFFIX@-wisdom BUILT_SOURCES = fftw-wisdom-to-conf fftw@PREC_SUFFIX@-wisdom.1 EXTRA_DIST = fftw-wisdom-to-conf.in dist_man_MANS = fftw-wisdom-to-conf.1 fftw@PREC_SUFFIX@-wisdom.1 EXTRA_MANS = fftw_wisdom.1.in fftw@PREC_SUFFIX@-wisdom.1: fftw_wisdom.1 rm -f $@ cp fftw_wisdom.1 $@ if THREADS fftw@PREC_SUFFIX@_wisdom_CFLAGS = $(PTHREAD_CFLAGS) if !COMBINED_THREADS LIBFFTWTHREADS = $(top_builddir)/threads/libfftw3@PREC_SUFFIX@_threads.la endif else if OPENMP fftw@PREC_SUFFIX@_wisdom_CFLAGS = $(OPENMP_CFLAGS) LIBFFTWTHREADS = $(top_builddir)/threads/libfftw3@PREC_SUFFIX@_omp.la endif endif fftw@PREC_SUFFIX@_wisdom_SOURCES = fftw-wisdom.c fftw@PREC_SUFFIX@_wisdom_LDADD = $(top_builddir)/tests/bench-bench.o \ $(top_builddir)/tests/bench-fftw-bench.o $(LIBFFTWTHREADS) \ $(top_builddir)/libfftw3@PREC_SUFFIX@.la \ $(top_builddir)/libbench2/libbench2.a $(THREADLIBS) fftw-3.3.3/tools/Makefile.in0000644000175000017500000006575312054410201012621 00000000000000# Makefile.in generated by automake 1.11.6 from Makefile.am. # @configure_input@ # Copyright (C) 1994, 1995, 1996, 1997, 1998, 1999, 2000, 2001, 2002, # 2003, 2004, 2005, 2006, 2007, 2008, 2009, 2010, 2011 Free Software # Foundation, Inc. # This Makefile.in is free software; the Free Software Foundation # gives unlimited permission to copy and/or distribute it, # with or without modifications, as long as this notice is preserved. # This program is distributed in the hope that it will be useful, # but WITHOUT ANY WARRANTY, to the extent permitted by law; without # even the implied warranty of MERCHANTABILITY or FITNESS FOR A # PARTICULAR PURPOSE. @SET_MAKE@ VPATH = @srcdir@ am__make_dryrun = \ { \ am__dry=no; \ case $$MAKEFLAGS in \ *\\[\ \ ]*) \ echo 'am--echo: ; @echo "AM" OK' | $(MAKE) -f - 2>/dev/null \ | grep '^AM OK$$' >/dev/null || am__dry=yes;; \ *) \ for am__flg in $$MAKEFLAGS; do \ case $$am__flg in \ *=*|--*) ;; \ *n*) am__dry=yes; break;; \ esac; \ done;; \ esac; \ test $$am__dry = yes; \ } pkgdatadir = $(datadir)/@PACKAGE@ pkgincludedir = $(includedir)/@PACKAGE@ pkglibdir = $(libdir)/@PACKAGE@ pkglibexecdir = $(libexecdir)/@PACKAGE@ am__cd = CDPATH="$${ZSH_VERSION+.}$(PATH_SEPARATOR)" && cd install_sh_DATA = $(install_sh) -c -m 644 install_sh_PROGRAM = $(install_sh) -c install_sh_SCRIPT = $(install_sh) -c INSTALL_HEADER = $(INSTALL_DATA) transform = $(program_transform_name) NORMAL_INSTALL = : PRE_INSTALL = : POST_INSTALL = : NORMAL_UNINSTALL = : PRE_UNINSTALL = : POST_UNINSTALL = : build_triplet = @build@ host_triplet = @host@ bin_PROGRAMS = fftw@PREC_SUFFIX@-wisdom$(EXEEXT) subdir = tools DIST_COMMON = $(dist_man_MANS) $(srcdir)/Makefile.am \ $(srcdir)/Makefile.in $(srcdir)/fftw-wisdom-to-conf.in \ $(srcdir)/fftw_wisdom.1.in ACLOCAL_M4 = $(top_srcdir)/aclocal.m4 am__aclocal_m4_deps = $(top_srcdir)/m4/acx_mpi.m4 \ $(top_srcdir)/m4/acx_pthread.m4 \ $(top_srcdir)/m4/ax_cc_maxopt.m4 \ $(top_srcdir)/m4/ax_check_compiler_flags.m4 \ $(top_srcdir)/m4/ax_compiler_vendor.m4 \ $(top_srcdir)/m4/ax_gcc_aligns_stack.m4 \ $(top_srcdir)/m4/ax_gcc_version.m4 \ $(top_srcdir)/m4/ax_openmp.m4 $(top_srcdir)/m4/libtool.m4 \ $(top_srcdir)/m4/ltoptions.m4 $(top_srcdir)/m4/ltsugar.m4 \ $(top_srcdir)/m4/ltversion.m4 $(top_srcdir)/m4/lt~obsolete.m4 \ $(top_srcdir)/configure.ac am__configure_deps = $(am__aclocal_m4_deps) $(CONFIGURE_DEPENDENCIES) \ $(ACLOCAL_M4) mkinstalldirs = $(install_sh) -d CONFIG_HEADER = $(top_builddir)/config.h CONFIG_CLEAN_FILES = fftw_wisdom.1 fftw-wisdom-to-conf CONFIG_CLEAN_VPATH_FILES = am__installdirs = "$(DESTDIR)$(bindir)" "$(DESTDIR)$(bindir)" \ "$(DESTDIR)$(man1dir)" PROGRAMS = $(bin_PROGRAMS) am_fftw@PREC_SUFFIX@_wisdom_OBJECTS = \ fftw@PREC_SUFFIX@_wisdom-fftw-wisdom.$(OBJEXT) fftw@PREC_SUFFIX@_wisdom_OBJECTS = \ $(am_fftw@PREC_SUFFIX@_wisdom_OBJECTS) am__DEPENDENCIES_1 = fftw@PREC_SUFFIX@_wisdom_DEPENDENCIES = \ $(top_builddir)/tests/bench-bench.o \ $(top_builddir)/tests/bench-fftw-bench.o $(LIBFFTWTHREADS) \ $(top_builddir)/libfftw3@PREC_SUFFIX@.la \ $(top_builddir)/libbench2/libbench2.a $(am__DEPENDENCIES_1) fftw@PREC_SUFFIX@_wisdom_LINK = $(LIBTOOL) --tag=CC $(AM_LIBTOOLFLAGS) \ $(LIBTOOLFLAGS) --mode=link $(CCLD) \ $(fftw@PREC_SUFFIX@_wisdom_CFLAGS) $(CFLAGS) $(AM_LDFLAGS) \ $(LDFLAGS) -o $@ am__vpath_adj_setup = srcdirstrip=`echo "$(srcdir)" | sed 's|.|.|g'`; am__vpath_adj = case $$p in \ $(srcdir)/*) f=`echo "$$p" | sed "s|^$$srcdirstrip/||"`;; \ *) f=$$p;; \ esac; am__strip_dir = f=`echo $$p | sed -e 's|^.*/||'`; am__install_max = 40 am__nobase_strip_setup = \ srcdirstrip=`echo "$(srcdir)" | sed 's/[].[^$$\\*|]/\\\\&/g'` am__nobase_strip = \ for p in $$list; do echo "$$p"; done | sed -e "s|$$srcdirstrip/||" am__nobase_list = $(am__nobase_strip_setup); \ for p in $$list; do echo "$$p $$p"; done | \ sed "s| $$srcdirstrip/| |;"' / .*\//!s/ .*/ ./; s,\( .*\)/[^/]*$$,\1,' | \ $(AWK) 'BEGIN { files["."] = "" } { files[$$2] = files[$$2] " " $$1; \ if (++n[$$2] == $(am__install_max)) \ { print $$2, files[$$2]; n[$$2] = 0; files[$$2] = "" } } \ END { for (dir in files) print dir, files[dir] }' am__base_list = \ sed '$$!N;$$!N;$$!N;$$!N;$$!N;$$!N;$$!N;s/\n/ /g' | \ sed '$$!N;$$!N;$$!N;$$!N;s/\n/ /g' am__uninstall_files_from_dir = { \ test -z "$$files" \ || { test ! -d "$$dir" && test ! -f "$$dir" && test ! -r "$$dir"; } \ || { echo " ( cd '$$dir' && rm -f" $$files ")"; \ $(am__cd) "$$dir" && rm -f $$files; }; \ } SCRIPTS = $(bin_SCRIPTS) DEFAULT_INCLUDES = -I.@am__isrc@ -I$(top_builddir) depcomp = $(SHELL) $(top_srcdir)/depcomp am__depfiles_maybe = depfiles am__mv = mv -f COMPILE = $(CC) $(DEFS) $(DEFAULT_INCLUDES) $(INCLUDES) $(AM_CPPFLAGS) \ $(CPPFLAGS) $(AM_CFLAGS) $(CFLAGS) LTCOMPILE = $(LIBTOOL) --tag=CC $(AM_LIBTOOLFLAGS) $(LIBTOOLFLAGS) \ --mode=compile $(CC) $(DEFS) $(DEFAULT_INCLUDES) $(INCLUDES) \ $(AM_CPPFLAGS) $(CPPFLAGS) $(AM_CFLAGS) $(CFLAGS) CCLD = $(CC) LINK = $(LIBTOOL) --tag=CC $(AM_LIBTOOLFLAGS) $(LIBTOOLFLAGS) \ --mode=link $(CCLD) $(AM_CFLAGS) $(CFLAGS) $(AM_LDFLAGS) \ $(LDFLAGS) -o $@ SOURCES = $(fftw@PREC_SUFFIX@_wisdom_SOURCES) DIST_SOURCES = $(fftw@PREC_SUFFIX@_wisdom_SOURCES) am__can_run_installinfo = \ case $$AM_UPDATE_INFO_DIR in \ n|no|NO) false;; \ *) (install-info --version) >/dev/null 2>&1;; \ esac man1dir = $(mandir)/man1 NROFF = nroff MANS = $(dist_man_MANS) ETAGS = etags CTAGS = ctags DISTFILES = $(DIST_COMMON) $(DIST_SOURCES) $(TEXINFOS) $(EXTRA_DIST) ACLOCAL = @ACLOCAL@ ALLOCA = @ALLOCA@ ALTIVEC_CFLAGS = @ALTIVEC_CFLAGS@ AMTAR = @AMTAR@ AR = @AR@ AS = @AS@ AUTOCONF = @AUTOCONF@ AUTOHEADER = @AUTOHEADER@ AUTOMAKE = @AUTOMAKE@ AVX_CFLAGS = @AVX_CFLAGS@ AWK = @AWK@ CC = @CC@ CCDEPMODE = @CCDEPMODE@ CFLAGS = @CFLAGS@ CHECK_PL_OPTS = @CHECK_PL_OPTS@ CPP = @CPP@ CPPFLAGS = @CPPFLAGS@ CYGPATH_W = @CYGPATH_W@ C_FFTW_R2R_KIND = @C_FFTW_R2R_KIND@ C_MPI_FINT = @C_MPI_FINT@ DEFS = @DEFS@ DEPDIR = @DEPDIR@ DLLTOOL = @DLLTOOL@ DSYMUTIL = @DSYMUTIL@ DUMPBIN = @DUMPBIN@ ECHO_C = @ECHO_C@ ECHO_N = @ECHO_N@ ECHO_T = @ECHO_T@ EGREP = @EGREP@ EXEEXT = @EXEEXT@ F77 = @F77@ FFLAGS = @FFLAGS@ FGREP = @FGREP@ FLIBS = @FLIBS@ GREP = @GREP@ INSTALL = @INSTALL@ INSTALL_DATA = @INSTALL_DATA@ INSTALL_PROGRAM = @INSTALL_PROGRAM@ INSTALL_SCRIPT = @INSTALL_SCRIPT@ INSTALL_STRIP_PROGRAM = @INSTALL_STRIP_PROGRAM@ LD = @LD@ LDFLAGS = @LDFLAGS@ LIBOBJS = @LIBOBJS@ LIBQUADMATH = @LIBQUADMATH@ LIBS = @LIBS@ LIBTOOL = @LIBTOOL@ LIPO = @LIPO@ LN_S = @LN_S@ LTLIBOBJS = @LTLIBOBJS@ MAINT = @MAINT@ MAKEINFO = @MAKEINFO@ MANIFEST_TOOL = @MANIFEST_TOOL@ MKDIR_P = @MKDIR_P@ MPICC = @MPICC@ MPILIBS = @MPILIBS@ MPIRUN = @MPIRUN@ NEON_CFLAGS = @NEON_CFLAGS@ NM = @NM@ NMEDIT = @NMEDIT@ OBJDUMP = @OBJDUMP@ OBJEXT = @OBJEXT@ OCAMLBUILD = @OCAMLBUILD@ OPENMP_CFLAGS = @OPENMP_CFLAGS@ OTOOL = @OTOOL@ OTOOL64 = @OTOOL64@ PACKAGE = @PACKAGE@ PACKAGE_BUGREPORT = @PACKAGE_BUGREPORT@ PACKAGE_NAME = @PACKAGE_NAME@ PACKAGE_STRING = @PACKAGE_STRING@ PACKAGE_TARNAME = @PACKAGE_TARNAME@ PACKAGE_URL = @PACKAGE_URL@ PACKAGE_VERSION = @PACKAGE_VERSION@ PATH_SEPARATOR = @PATH_SEPARATOR@ POW_LIB = @POW_LIB@ PRECISION = @PRECISION@ PREC_SUFFIX = @PREC_SUFFIX@ PTHREAD_CC = @PTHREAD_CC@ PTHREAD_CFLAGS = @PTHREAD_CFLAGS@ PTHREAD_LIBS = @PTHREAD_LIBS@ RANLIB = @RANLIB@ SED = @SED@ SET_MAKE = @SET_MAKE@ SHARED_VERSION_INFO = @SHARED_VERSION_INFO@ SHELL = @SHELL@ SSE2_CFLAGS = @SSE2_CFLAGS@ STACK_ALIGN_CFLAGS = @STACK_ALIGN_CFLAGS@ STRIP = @STRIP@ THREADLIBS = @THREADLIBS@ VERSION = @VERSION@ abs_builddir = @abs_builddir@ abs_srcdir = @abs_srcdir@ abs_top_builddir = @abs_top_builddir@ abs_top_srcdir = @abs_top_srcdir@ ac_ct_AR = @ac_ct_AR@ ac_ct_CC = @ac_ct_CC@ ac_ct_DUMPBIN = @ac_ct_DUMPBIN@ ac_ct_F77 = @ac_ct_F77@ acx_pthread_config = @acx_pthread_config@ am__include = @am__include@ am__leading_dot = @am__leading_dot@ am__quote = @am__quote@ am__tar = @am__tar@ am__untar = @am__untar@ bindir = @bindir@ build = @build@ build_alias = @build_alias@ build_cpu = @build_cpu@ build_os = @build_os@ build_vendor = @build_vendor@ builddir = @builddir@ datadir = @datadir@ datarootdir = @datarootdir@ docdir = @docdir@ dvidir = @dvidir@ exec_prefix = @exec_prefix@ host = @host@ host_alias = @host_alias@ host_cpu = @host_cpu@ host_os = @host_os@ host_vendor = @host_vendor@ htmldir = @htmldir@ includedir = @includedir@ infodir = @infodir@ install_sh = @install_sh@ libdir = @libdir@ libexecdir = @libexecdir@ localedir = @localedir@ localstatedir = @localstatedir@ mandir = @mandir@ mkdir_p = @mkdir_p@ oldincludedir = @oldincludedir@ pdfdir = @pdfdir@ prefix = @prefix@ program_transform_name = @program_transform_name@ psdir = @psdir@ sbindir = @sbindir@ sharedstatedir = @sharedstatedir@ srcdir = @srcdir@ sysconfdir = @sysconfdir@ target_alias = @target_alias@ top_build_prefix = @top_build_prefix@ top_builddir = @top_builddir@ top_srcdir = @top_srcdir@ AM_CPPFLAGS = -I$(top_srcdir)/libbench2 -I$(top_srcdir)/api bin_SCRIPTS = fftw-wisdom-to-conf BUILT_SOURCES = fftw-wisdom-to-conf fftw@PREC_SUFFIX@-wisdom.1 EXTRA_DIST = fftw-wisdom-to-conf.in dist_man_MANS = fftw-wisdom-to-conf.1 fftw@PREC_SUFFIX@-wisdom.1 EXTRA_MANS = fftw_wisdom.1.in @OPENMP_TRUE@@THREADS_FALSE@fftw@PREC_SUFFIX@_wisdom_CFLAGS = $(OPENMP_CFLAGS) @THREADS_TRUE@fftw@PREC_SUFFIX@_wisdom_CFLAGS = $(PTHREAD_CFLAGS) @COMBINED_THREADS_FALSE@@THREADS_TRUE@LIBFFTWTHREADS = $(top_builddir)/threads/libfftw3@PREC_SUFFIX@_threads.la @OPENMP_TRUE@@THREADS_FALSE@LIBFFTWTHREADS = $(top_builddir)/threads/libfftw3@PREC_SUFFIX@_omp.la fftw@PREC_SUFFIX@_wisdom_SOURCES = fftw-wisdom.c fftw@PREC_SUFFIX@_wisdom_LDADD = $(top_builddir)/tests/bench-bench.o \ $(top_builddir)/tests/bench-fftw-bench.o $(LIBFFTWTHREADS) \ $(top_builddir)/libfftw3@PREC_SUFFIX@.la \ $(top_builddir)/libbench2/libbench2.a $(THREADLIBS) all: $(BUILT_SOURCES) $(MAKE) $(AM_MAKEFLAGS) all-am .SUFFIXES: .SUFFIXES: .c .lo .o .obj $(srcdir)/Makefile.in: @MAINTAINER_MODE_TRUE@ $(srcdir)/Makefile.am $(am__configure_deps) @for dep in $?; do \ case '$(am__configure_deps)' in \ *$$dep*) \ ( cd $(top_builddir) && $(MAKE) $(AM_MAKEFLAGS) am--refresh ) \ && { if test -f $@; then exit 0; else break; fi; }; \ exit 1;; \ esac; \ done; \ echo ' cd $(top_srcdir) && $(AUTOMAKE) --gnu tools/Makefile'; \ $(am__cd) $(top_srcdir) && \ $(AUTOMAKE) --gnu tools/Makefile .PRECIOUS: Makefile Makefile: $(srcdir)/Makefile.in $(top_builddir)/config.status @case '$?' in \ *config.status*) \ cd $(top_builddir) && $(MAKE) $(AM_MAKEFLAGS) am--refresh;; \ *) \ echo ' cd $(top_builddir) && $(SHELL) ./config.status $(subdir)/$@ $(am__depfiles_maybe)'; \ cd $(top_builddir) && $(SHELL) ./config.status $(subdir)/$@ $(am__depfiles_maybe);; \ esac; $(top_builddir)/config.status: $(top_srcdir)/configure $(CONFIG_STATUS_DEPENDENCIES) cd $(top_builddir) && $(MAKE) $(AM_MAKEFLAGS) am--refresh $(top_srcdir)/configure: @MAINTAINER_MODE_TRUE@ $(am__configure_deps) cd $(top_builddir) && $(MAKE) $(AM_MAKEFLAGS) am--refresh $(ACLOCAL_M4): @MAINTAINER_MODE_TRUE@ $(am__aclocal_m4_deps) cd $(top_builddir) && $(MAKE) $(AM_MAKEFLAGS) am--refresh $(am__aclocal_m4_deps): fftw_wisdom.1: $(top_builddir)/config.status $(srcdir)/fftw_wisdom.1.in cd $(top_builddir) && $(SHELL) ./config.status $(subdir)/$@ fftw-wisdom-to-conf: $(top_builddir)/config.status $(srcdir)/fftw-wisdom-to-conf.in cd $(top_builddir) && $(SHELL) ./config.status $(subdir)/$@ install-binPROGRAMS: $(bin_PROGRAMS) @$(NORMAL_INSTALL) @list='$(bin_PROGRAMS)'; test -n "$(bindir)" || list=; \ if test -n "$$list"; then \ echo " $(MKDIR_P) '$(DESTDIR)$(bindir)'"; \ $(MKDIR_P) "$(DESTDIR)$(bindir)" || exit 1; \ fi; \ for p in $$list; do echo "$$p $$p"; done | \ sed 's/$(EXEEXT)$$//' | \ while read p p1; do if test -f $$p || test -f $$p1; \ then echo "$$p"; echo "$$p"; else :; fi; \ done | \ sed -e 'p;s,.*/,,;n;h' -e 's|.*|.|' \ -e 'p;x;s,.*/,,;s/$(EXEEXT)$$//;$(transform);s/$$/$(EXEEXT)/' | \ sed 'N;N;N;s,\n, ,g' | \ $(AWK) 'BEGIN { files["."] = ""; dirs["."] = 1 } \ { d=$$3; if (dirs[d] != 1) { print "d", d; dirs[d] = 1 } \ if ($$2 == $$4) files[d] = files[d] " " $$1; \ else { print "f", $$3 "/" $$4, $$1; } } \ END { for (d in files) print "f", d, files[d] }' | \ while read type dir files; do \ if test "$$dir" = .; then dir=; else dir=/$$dir; fi; \ test -z "$$files" || { \ echo " $(INSTALL_PROGRAM_ENV) $(LIBTOOL) $(AM_LIBTOOLFLAGS) $(LIBTOOLFLAGS) --mode=install $(INSTALL_PROGRAM) $$files '$(DESTDIR)$(bindir)$$dir'"; \ $(INSTALL_PROGRAM_ENV) $(LIBTOOL) $(AM_LIBTOOLFLAGS) $(LIBTOOLFLAGS) --mode=install $(INSTALL_PROGRAM) $$files "$(DESTDIR)$(bindir)$$dir" || exit $$?; \ } \ ; done uninstall-binPROGRAMS: @$(NORMAL_UNINSTALL) @list='$(bin_PROGRAMS)'; test -n "$(bindir)" || list=; \ files=`for p in $$list; do echo "$$p"; done | \ sed -e 'h;s,^.*/,,;s/$(EXEEXT)$$//;$(transform)' \ -e 's/$$/$(EXEEXT)/' `; \ test -n "$$list" || exit 0; \ echo " ( cd '$(DESTDIR)$(bindir)' && rm -f" $$files ")"; \ cd "$(DESTDIR)$(bindir)" && rm -f $$files clean-binPROGRAMS: @list='$(bin_PROGRAMS)'; test -n "$$list" || exit 0; \ echo " rm -f" $$list; \ rm -f $$list || exit $$?; \ test -n "$(EXEEXT)" || exit 0; \ list=`for p in $$list; do echo "$$p"; done | sed 's/$(EXEEXT)$$//'`; \ echo " rm -f" $$list; \ rm -f $$list fftw@PREC_SUFFIX@-wisdom$(EXEEXT): $(fftw@PREC_SUFFIX@_wisdom_OBJECTS) $(fftw@PREC_SUFFIX@_wisdom_DEPENDENCIES) $(EXTRA_fftw@PREC_SUFFIX@_wisdom_DEPENDENCIES) @rm -f fftw@PREC_SUFFIX@-wisdom$(EXEEXT) $(fftw@PREC_SUFFIX@_wisdom_LINK) $(fftw@PREC_SUFFIX@_wisdom_OBJECTS) $(fftw@PREC_SUFFIX@_wisdom_LDADD) $(LIBS) install-binSCRIPTS: $(bin_SCRIPTS) @$(NORMAL_INSTALL) @list='$(bin_SCRIPTS)'; test -n "$(bindir)" || list=; \ if test -n "$$list"; then \ echo " $(MKDIR_P) '$(DESTDIR)$(bindir)'"; \ $(MKDIR_P) "$(DESTDIR)$(bindir)" || exit 1; \ fi; \ for p in $$list; do \ if test -f "$$p"; then d=; else d="$(srcdir)/"; fi; \ if test -f "$$d$$p"; then echo "$$d$$p"; echo "$$p"; else :; fi; \ done | \ sed -e 'p;s,.*/,,;n' \ -e 'h;s|.*|.|' \ -e 'p;x;s,.*/,,;$(transform)' | sed 'N;N;N;s,\n, ,g' | \ $(AWK) 'BEGIN { files["."] = ""; dirs["."] = 1; } \ { d=$$3; if (dirs[d] != 1) { print "d", d; dirs[d] = 1 } \ if ($$2 == $$4) { files[d] = files[d] " " $$1; \ if (++n[d] == $(am__install_max)) { \ print "f", d, files[d]; n[d] = 0; files[d] = "" } } \ else { print "f", d "/" $$4, $$1 } } \ END { for (d in files) print "f", d, files[d] }' | \ while read type dir files; do \ if test "$$dir" = .; then dir=; else dir=/$$dir; fi; \ test -z "$$files" || { \ echo " $(INSTALL_SCRIPT) $$files '$(DESTDIR)$(bindir)$$dir'"; \ $(INSTALL_SCRIPT) $$files "$(DESTDIR)$(bindir)$$dir" || exit $$?; \ } \ ; done uninstall-binSCRIPTS: @$(NORMAL_UNINSTALL) @list='$(bin_SCRIPTS)'; test -n "$(bindir)" || exit 0; \ files=`for p in $$list; do echo "$$p"; done | \ sed -e 's,.*/,,;$(transform)'`; \ dir='$(DESTDIR)$(bindir)'; $(am__uninstall_files_from_dir) mostlyclean-compile: -rm -f *.$(OBJEXT) distclean-compile: -rm -f *.tab.c @AMDEP_TRUE@@am__include@ @am__quote@./$(DEPDIR)/fftw@PREC_SUFFIX@_wisdom-fftw-wisdom.Po@am__quote@ .c.o: @am__fastdepCC_TRUE@ $(COMPILE) -MT $@ -MD -MP -MF $(DEPDIR)/$*.Tpo -c -o $@ $< @am__fastdepCC_TRUE@ $(am__mv) $(DEPDIR)/$*.Tpo $(DEPDIR)/$*.Po @AMDEP_TRUE@@am__fastdepCC_FALSE@ source='$<' object='$@' libtool=no @AMDEPBACKSLASH@ @AMDEP_TRUE@@am__fastdepCC_FALSE@ DEPDIR=$(DEPDIR) $(CCDEPMODE) $(depcomp) @AMDEPBACKSLASH@ @am__fastdepCC_FALSE@ $(COMPILE) -c $< .c.obj: @am__fastdepCC_TRUE@ $(COMPILE) -MT $@ -MD -MP -MF $(DEPDIR)/$*.Tpo -c -o $@ `$(CYGPATH_W) '$<'` @am__fastdepCC_TRUE@ $(am__mv) $(DEPDIR)/$*.Tpo $(DEPDIR)/$*.Po @AMDEP_TRUE@@am__fastdepCC_FALSE@ source='$<' object='$@' libtool=no @AMDEPBACKSLASH@ @AMDEP_TRUE@@am__fastdepCC_FALSE@ DEPDIR=$(DEPDIR) $(CCDEPMODE) $(depcomp) @AMDEPBACKSLASH@ @am__fastdepCC_FALSE@ $(COMPILE) -c `$(CYGPATH_W) '$<'` .c.lo: @am__fastdepCC_TRUE@ $(LTCOMPILE) -MT $@ -MD -MP -MF $(DEPDIR)/$*.Tpo -c -o $@ $< @am__fastdepCC_TRUE@ $(am__mv) $(DEPDIR)/$*.Tpo $(DEPDIR)/$*.Plo @AMDEP_TRUE@@am__fastdepCC_FALSE@ source='$<' object='$@' libtool=yes @AMDEPBACKSLASH@ @AMDEP_TRUE@@am__fastdepCC_FALSE@ DEPDIR=$(DEPDIR) $(CCDEPMODE) $(depcomp) @AMDEPBACKSLASH@ @am__fastdepCC_FALSE@ $(LTCOMPILE) -c -o $@ $< fftw@PREC_SUFFIX@_wisdom-fftw-wisdom.o: fftw-wisdom.c @am__fastdepCC_TRUE@ $(CC) $(DEFS) $(DEFAULT_INCLUDES) $(INCLUDES) $(AM_CPPFLAGS) $(CPPFLAGS) $(fftw@PREC_SUFFIX@_wisdom_CFLAGS) $(CFLAGS) -MT fftw@PREC_SUFFIX@_wisdom-fftw-wisdom.o -MD -MP -MF $(DEPDIR)/fftw@PREC_SUFFIX@_wisdom-fftw-wisdom.Tpo -c -o fftw@PREC_SUFFIX@_wisdom-fftw-wisdom.o `test -f 'fftw-wisdom.c' || echo '$(srcdir)/'`fftw-wisdom.c @am__fastdepCC_TRUE@ $(am__mv) $(DEPDIR)/fftw@PREC_SUFFIX@_wisdom-fftw-wisdom.Tpo $(DEPDIR)/fftw@PREC_SUFFIX@_wisdom-fftw-wisdom.Po @AMDEP_TRUE@@am__fastdepCC_FALSE@ source='fftw-wisdom.c' object='fftw@PREC_SUFFIX@_wisdom-fftw-wisdom.o' libtool=no @AMDEPBACKSLASH@ @AMDEP_TRUE@@am__fastdepCC_FALSE@ DEPDIR=$(DEPDIR) $(CCDEPMODE) $(depcomp) @AMDEPBACKSLASH@ @am__fastdepCC_FALSE@ $(CC) $(DEFS) $(DEFAULT_INCLUDES) $(INCLUDES) $(AM_CPPFLAGS) $(CPPFLAGS) $(fftw@PREC_SUFFIX@_wisdom_CFLAGS) $(CFLAGS) -c -o fftw@PREC_SUFFIX@_wisdom-fftw-wisdom.o `test -f 'fftw-wisdom.c' || echo '$(srcdir)/'`fftw-wisdom.c fftw@PREC_SUFFIX@_wisdom-fftw-wisdom.obj: fftw-wisdom.c @am__fastdepCC_TRUE@ $(CC) $(DEFS) $(DEFAULT_INCLUDES) $(INCLUDES) $(AM_CPPFLAGS) $(CPPFLAGS) $(fftw@PREC_SUFFIX@_wisdom_CFLAGS) $(CFLAGS) -MT fftw@PREC_SUFFIX@_wisdom-fftw-wisdom.obj -MD -MP -MF $(DEPDIR)/fftw@PREC_SUFFIX@_wisdom-fftw-wisdom.Tpo -c -o fftw@PREC_SUFFIX@_wisdom-fftw-wisdom.obj `if test -f 'fftw-wisdom.c'; then $(CYGPATH_W) 'fftw-wisdom.c'; else $(CYGPATH_W) '$(srcdir)/fftw-wisdom.c'; fi` @am__fastdepCC_TRUE@ $(am__mv) $(DEPDIR)/fftw@PREC_SUFFIX@_wisdom-fftw-wisdom.Tpo $(DEPDIR)/fftw@PREC_SUFFIX@_wisdom-fftw-wisdom.Po @AMDEP_TRUE@@am__fastdepCC_FALSE@ source='fftw-wisdom.c' object='fftw@PREC_SUFFIX@_wisdom-fftw-wisdom.obj' libtool=no @AMDEPBACKSLASH@ @AMDEP_TRUE@@am__fastdepCC_FALSE@ DEPDIR=$(DEPDIR) $(CCDEPMODE) $(depcomp) @AMDEPBACKSLASH@ @am__fastdepCC_FALSE@ $(CC) $(DEFS) $(DEFAULT_INCLUDES) $(INCLUDES) $(AM_CPPFLAGS) $(CPPFLAGS) $(fftw@PREC_SUFFIX@_wisdom_CFLAGS) $(CFLAGS) -c -o fftw@PREC_SUFFIX@_wisdom-fftw-wisdom.obj `if test -f 'fftw-wisdom.c'; then $(CYGPATH_W) 'fftw-wisdom.c'; else $(CYGPATH_W) '$(srcdir)/fftw-wisdom.c'; fi` mostlyclean-libtool: -rm -f *.lo clean-libtool: -rm -rf .libs _libs install-man1: $(dist_man_MANS) @$(NORMAL_INSTALL) @list1=''; \ list2='$(dist_man_MANS)'; \ test -n "$(man1dir)" \ && test -n "`echo $$list1$$list2`" \ || exit 0; \ echo " $(MKDIR_P) '$(DESTDIR)$(man1dir)'"; \ $(MKDIR_P) "$(DESTDIR)$(man1dir)" || exit 1; \ { for i in $$list1; do echo "$$i"; done; \ if test -n "$$list2"; then \ for i in $$list2; do echo "$$i"; done \ | sed -n '/\.1[a-z]*$$/p'; \ fi; \ } | while read p; do \ if test -f $$p; then d=; else d="$(srcdir)/"; fi; \ echo "$$d$$p"; echo "$$p"; \ done | \ sed -e 'n;s,.*/,,;p;h;s,.*\.,,;s,^[^1][0-9a-z]*$$,1,;x' \ -e 's,\.[0-9a-z]*$$,,;$(transform);G;s,\n,.,' | \ sed 'N;N;s,\n, ,g' | { \ list=; while read file base inst; do \ if test "$$base" = "$$inst"; then list="$$list $$file"; else \ echo " $(INSTALL_DATA) '$$file' '$(DESTDIR)$(man1dir)/$$inst'"; \ $(INSTALL_DATA) "$$file" "$(DESTDIR)$(man1dir)/$$inst" || exit $$?; \ fi; \ done; \ for i in $$list; do echo "$$i"; done | $(am__base_list) | \ while read files; do \ test -z "$$files" || { \ echo " $(INSTALL_DATA) $$files '$(DESTDIR)$(man1dir)'"; \ $(INSTALL_DATA) $$files "$(DESTDIR)$(man1dir)" || exit $$?; }; \ done; } uninstall-man1: @$(NORMAL_UNINSTALL) @list=''; test -n "$(man1dir)" || exit 0; \ files=`{ for i in $$list; do echo "$$i"; done; \ l2='$(dist_man_MANS)'; for i in $$l2; do echo "$$i"; done | \ sed -n '/\.1[a-z]*$$/p'; \ } | sed -e 's,.*/,,;h;s,.*\.,,;s,^[^1][0-9a-z]*$$,1,;x' \ -e 's,\.[0-9a-z]*$$,,;$(transform);G;s,\n,.,'`; \ dir='$(DESTDIR)$(man1dir)'; $(am__uninstall_files_from_dir) ID: $(HEADERS) $(SOURCES) $(LISP) $(TAGS_FILES) list='$(SOURCES) $(HEADERS) $(LISP) $(TAGS_FILES)'; \ unique=`for i in $$list; do \ if test -f "$$i"; then echo $$i; else echo $(srcdir)/$$i; fi; \ done | \ $(AWK) '{ files[$$0] = 1; nonempty = 1; } \ END { if (nonempty) { for (i in files) print i; }; }'`; \ mkid -fID $$unique tags: TAGS TAGS: $(HEADERS) $(SOURCES) $(TAGS_DEPENDENCIES) \ $(TAGS_FILES) $(LISP) set x; \ here=`pwd`; \ list='$(SOURCES) $(HEADERS) $(LISP) $(TAGS_FILES)'; \ unique=`for i in $$list; do \ if test -f "$$i"; then echo $$i; else echo $(srcdir)/$$i; fi; \ done | \ $(AWK) '{ files[$$0] = 1; nonempty = 1; } \ END { if (nonempty) { for (i in files) print i; }; }'`; \ shift; \ if test -z "$(ETAGS_ARGS)$$*$$unique"; then :; else \ test -n "$$unique" || unique=$$empty_fix; \ if test $$# -gt 0; then \ $(ETAGS) $(ETAGSFLAGS) $(AM_ETAGSFLAGS) $(ETAGS_ARGS) \ "$$@" $$unique; \ else \ $(ETAGS) $(ETAGSFLAGS) $(AM_ETAGSFLAGS) $(ETAGS_ARGS) \ $$unique; \ fi; \ fi ctags: CTAGS CTAGS: $(HEADERS) $(SOURCES) $(TAGS_DEPENDENCIES) \ $(TAGS_FILES) $(LISP) list='$(SOURCES) $(HEADERS) $(LISP) $(TAGS_FILES)'; \ unique=`for i in $$list; do \ if test -f "$$i"; then echo $$i; else echo $(srcdir)/$$i; fi; \ done | \ $(AWK) '{ files[$$0] = 1; nonempty = 1; } \ END { if (nonempty) { for (i in files) print i; }; }'`; \ test -z "$(CTAGS_ARGS)$$unique" \ || $(CTAGS) $(CTAGSFLAGS) $(AM_CTAGSFLAGS) $(CTAGS_ARGS) \ $$unique GTAGS: here=`$(am__cd) $(top_builddir) && pwd` \ && $(am__cd) $(top_srcdir) \ && gtags -i $(GTAGS_ARGS) "$$here" distclean-tags: -rm -f TAGS ID GTAGS GRTAGS GSYMS GPATH tags distdir: $(DISTFILES) @list='$(MANS)'; if test -n "$$list"; then \ list=`for p in $$list; do \ if test -f $$p; then d=; else d="$(srcdir)/"; fi; \ if test -f "$$d$$p"; then echo "$$d$$p"; else :; fi; done`; \ if test -n "$$list" && \ grep 'ab help2man is required to generate this page' $$list >/dev/null; then \ echo "error: found man pages containing the \`missing help2man' replacement text:" >&2; \ grep -l 'ab help2man is required to generate this page' $$list | sed 's/^/ /' >&2; \ echo " to fix them, install help2man, remove and regenerate the man pages;" >&2; \ echo " typically \`make maintainer-clean' will remove them" >&2; \ exit 1; \ else :; fi; \ else :; fi @srcdirstrip=`echo "$(srcdir)" | sed 's/[].[^$$\\*]/\\\\&/g'`; \ topsrcdirstrip=`echo "$(top_srcdir)" | sed 's/[].[^$$\\*]/\\\\&/g'`; \ list='$(DISTFILES)'; \ dist_files=`for file in $$list; do echo $$file; done | \ sed -e "s|^$$srcdirstrip/||;t" \ -e "s|^$$topsrcdirstrip/|$(top_builddir)/|;t"`; \ case $$dist_files in \ */*) $(MKDIR_P) `echo "$$dist_files" | \ sed '/\//!d;s|^|$(distdir)/|;s,/[^/]*$$,,' | \ sort -u` ;; \ esac; \ for file in $$dist_files; do \ if test -f $$file || test -d $$file; then d=.; else d=$(srcdir); fi; \ if test -d $$d/$$file; then \ dir=`echo "/$$file" | sed -e 's,/[^/]*$$,,'`; \ if test -d "$(distdir)/$$file"; then \ find "$(distdir)/$$file" -type d ! -perm -700 -exec chmod u+rwx {} \;; \ fi; \ if test -d $(srcdir)/$$file && test $$d != $(srcdir); then \ cp -fpR $(srcdir)/$$file "$(distdir)$$dir" || exit 1; \ find "$(distdir)/$$file" -type d ! -perm -700 -exec chmod u+rwx {} \;; \ fi; \ cp -fpR $$d/$$file "$(distdir)$$dir" || exit 1; \ else \ test -f "$(distdir)/$$file" \ || cp -p $$d/$$file "$(distdir)/$$file" \ || exit 1; \ fi; \ done check-am: all-am check: $(BUILT_SOURCES) $(MAKE) $(AM_MAKEFLAGS) check-am all-am: Makefile $(PROGRAMS) $(SCRIPTS) $(MANS) installdirs: for dir in "$(DESTDIR)$(bindir)" "$(DESTDIR)$(bindir)" "$(DESTDIR)$(man1dir)"; do \ test -z "$$dir" || $(MKDIR_P) "$$dir"; \ done install: $(BUILT_SOURCES) $(MAKE) $(AM_MAKEFLAGS) install-am install-exec: install-exec-am install-data: install-data-am uninstall: uninstall-am install-am: all-am @$(MAKE) $(AM_MAKEFLAGS) install-exec-am install-data-am installcheck: installcheck-am install-strip: if test -z '$(STRIP)'; then \ $(MAKE) $(AM_MAKEFLAGS) INSTALL_PROGRAM="$(INSTALL_STRIP_PROGRAM)" \ install_sh_PROGRAM="$(INSTALL_STRIP_PROGRAM)" INSTALL_STRIP_FLAG=-s \ install; \ else \ $(MAKE) $(AM_MAKEFLAGS) INSTALL_PROGRAM="$(INSTALL_STRIP_PROGRAM)" \ install_sh_PROGRAM="$(INSTALL_STRIP_PROGRAM)" INSTALL_STRIP_FLAG=-s \ "INSTALL_PROGRAM_ENV=STRIPPROG='$(STRIP)'" install; \ fi mostlyclean-generic: clean-generic: distclean-generic: -test -z "$(CONFIG_CLEAN_FILES)" || rm -f $(CONFIG_CLEAN_FILES) -test . = "$(srcdir)" || test -z "$(CONFIG_CLEAN_VPATH_FILES)" || rm -f $(CONFIG_CLEAN_VPATH_FILES) maintainer-clean-generic: @echo "This command is intended for maintainers to use" @echo "it deletes files that may require special tools to rebuild." -test -z "$(BUILT_SOURCES)" || rm -f $(BUILT_SOURCES) clean: clean-am clean-am: clean-binPROGRAMS clean-generic clean-libtool mostlyclean-am distclean: distclean-am -rm -rf ./$(DEPDIR) -rm -f Makefile distclean-am: clean-am distclean-compile distclean-generic \ distclean-tags dvi: dvi-am dvi-am: html: html-am html-am: info: info-am info-am: install-data-am: install-man install-dvi: install-dvi-am install-dvi-am: install-exec-am: install-binPROGRAMS install-binSCRIPTS install-html: install-html-am install-html-am: install-info: install-info-am install-info-am: install-man: install-man1 install-pdf: install-pdf-am install-pdf-am: install-ps: install-ps-am install-ps-am: installcheck-am: maintainer-clean: maintainer-clean-am -rm -rf ./$(DEPDIR) -rm -f Makefile maintainer-clean-am: distclean-am maintainer-clean-generic mostlyclean: mostlyclean-am mostlyclean-am: mostlyclean-compile mostlyclean-generic \ mostlyclean-libtool pdf: pdf-am pdf-am: ps: ps-am ps-am: uninstall-am: uninstall-binPROGRAMS uninstall-binSCRIPTS uninstall-man uninstall-man: uninstall-man1 .MAKE: all check install install-am install-strip .PHONY: CTAGS GTAGS all all-am check check-am clean clean-binPROGRAMS \ clean-generic clean-libtool ctags distclean distclean-compile \ distclean-generic distclean-libtool distclean-tags distdir dvi \ dvi-am html html-am info info-am install install-am \ install-binPROGRAMS install-binSCRIPTS install-data \ install-data-am install-dvi install-dvi-am install-exec \ install-exec-am install-html install-html-am install-info \ install-info-am install-man install-man1 install-pdf \ install-pdf-am install-ps install-ps-am install-strip \ installcheck installcheck-am installdirs maintainer-clean \ maintainer-clean-generic mostlyclean mostlyclean-compile \ mostlyclean-generic mostlyclean-libtool pdf pdf-am ps ps-am \ tags uninstall uninstall-am uninstall-binPROGRAMS \ uninstall-binSCRIPTS uninstall-man uninstall-man1 fftw@PREC_SUFFIX@-wisdom.1: fftw_wisdom.1 rm -f $@ cp fftw_wisdom.1 $@ # Tell versions [3.59,3.63) of GNU make to not export all variables. # Otherwise a system limit (for SysV at least) may be exceeded. .NOEXPORT: fftw-3.3.3/tools/fftw-wisdom-to-conf.in0000644000175000017500000000437412054410050014707 00000000000000#! /bin/sh if test "x$1" = "x--help" || test "x$1" = "x-h"; then cat < OUTPUT] Convert wisdom (stdin) to C configuration routine (stdout). Options: -h, --help: print this help -V, --version: print version/copyright info EOF exit 0 fi if test "x$1" = "x--version" || test "x$1" = "x-V"; then cat <&2 exit 1 ;; esac cat <<\fIinplace\fR><\fIdirection\fR><\fIgeometry\fR> <\fItype\fR> is either \'c\' (complex), \'r\' (real, r2c/c2r), or \'k\' (r2r, per-dimension kinds, specified in the geometry, below). <\fIinplace\fR> is either \'i\' (in place) or \'o\' (out of place). <\fIdirection\fR> is either \'f\' (forward) or \'b\' (backward). The <\fIdirection\fR> should be omitted for \'k\' transforms, where it is specified via the geometry instead. <\fIgeometry\fR> is the size and dimensionality of the transform, where different dimensions are separated by \'x\' (e.g. \'16x32\' for a two-dimensional 16 by 32 transform). In the case of \'k\' transforms, the size of each dimension is followed by a "type" string, which can be one of f/b/h/e00/e01/e10/e11/o00/o01/o10/o11 for R2HC/HC2R/DHT/REDFT00/.../RODFT11, respectively, as defined in the FFTW manual. For example, \'cif12x13x14\' is a three-dimensional 12 by 13 x 14 complex DFT operating in-place. \'rob65536\' is a one-dimensional size-65536 out-of-place complex-to-real (backwards) transform operating on Hermitian-symmetry input. \'ki10hx20e01\' is a two-dimensional 10 by 20 r2r transform where the first dimension is a DHT and the second dimension is an REDFT01 (DCT-III). .SH OPTIONS .TP \fB\-h\fR, \fB\--help\fR Display help on the command-line options and usage. .TP \fB\-V\fR, \fB\--version\fR Print the version number and copyright information. .TP \fB\-v\fR, \fB\--verbose\fR Verbose output. (You can specify this multiple times, or supply a numeric argument greater than 1, to increase the verbosity level.) Note that the verbose output will be mixed with the wisdom output (making it impossible to import), unless you write the wisdom to a file via the .B -o option. .TP \fB\-c\fR, \fB\--canonical\fR Optimize/pre-plan a canonical set of sizes: all powers of two and ten up to 2^20 (1048576), including both real and complex, forward and backwards, in-place and out-of-place transforms. Also includes two- and three-dimensional transforms of equal-size dimensions (e.g. 16x16x16). .TP \fB\-t\fR \fIhours\fR, \fB\--time-limit\fR=\fIhours\fR Stop after a time of .I hours (hours) has elapsed, outputting accumulated wisdom. (The problems are planned in increasing order of size.) Defaults to 0, indicating no time limit. .TP \fB\-o\fR \fIfile\fR, \fB\--output-file\fR=\fIfile\fR Send wisdom output to .I file rather than to standard output (the default). .TP \fB\-m\fR, \fB\--measure\fR; \fB\-e\fR, \fB\--estimate\fR; \fB\-x\fR, \fB\--exhaustive\fR Normally, .I fftw@PREC_SUFFIX@-wisdom creates plans in FFTW_PATIENT mode, but with these options you can instead use FFTW_MEASURE, FFTW_ESTIMATE, or FFTW_EXHAUSTIVE modes, respectively, as described in more detail by the FFTW manual. Note that wisdom is tagged with the planning patience level, and a single file can mix different levels of wisdom (e.g. you can mostly use the patient default, but plan a few sizes that you especially care about in .B --exhaustive mode). .TP \fB\-n\fR, \fB\--no-system-wisdom\fR Do not import the system wisdom from .I /etc/fftw/wisdom@PREC_SUFFIX@ (which is normally read by default). .TP \fB\-w\fR \fIfile\fR, \fB\--wisdom-file\fR=\fIfile\fR Import wisdom from .I file (in addition to the system wisdom, unless .B -n is specified). Multiple wisdom files can be read via multiple .B -w options. If .I file is "-", then read wisdom from standard input. .SH BUGS Send bug reports to fftw@fftw.org. .SH AUTHORS Written by Steven G. Johnson and Matteo Frigo. Copyright (c) 2003, 2007-11 Matteo Frigo .br Copyright (c) 2003, 2007-11 Massachusetts Institute of Technology .SH "SEE ALSO" fftw-wisdom-to-conf(1) fftw-3.3.3/tools/fftw-wisdom.c0000644000175000017500000002273212054410050013156 00000000000000/* Re-use libbench2 and the test program, but override bench_main so that we can have different command-line syntax. */ #include "my-getopt.h" #include "bench.h" #include #include #include #include #include #include #if defined(HAVE_THREADS) || defined(HAVE_OPENMP) # define HAVE_SMP extern int threads_ok; #endif #define CONCAT(prefix, name) prefix ## name #if defined(BENCHFFT_SINGLE) #define FFTW(x) CONCAT(fftwf_, x) #elif defined(BENCHFFT_LDOUBLE) #define FFTW(x) CONCAT(fftwl_, x) #elif defined(BENCHFFT_QUAD) #define FFTW(x) CONCAT(fftwq_, x) #else #define FFTW(x) CONCAT(fftw_, x) #endif /* from bench.c: */ extern unsigned the_flags; extern int usewisdom; extern int nthreads; /* dummy routines to replace those in hook.c */ void install_hook(void) {} void uninstall_hook(void) {} int verbose; static void do_problem(bench_problem *p) { if (verbose) printf("PLANNING PROBLEM: %s\n", p->pstring); /* BENCH_ASSERT(can_do(p)); */ problem_alloc(p); setup(p); done(p); } static void add_problem(const char *pstring, bench_problem ***p, int *ip, int *np) { if (*ip >= *np) { *np = *np * 2 + 1; *p = (bench_problem **) realloc(*p, sizeof(bench_problem *) * *np); } (*p)[(*ip)++] = problem_parse(pstring); } static int sz(const bench_problem *p) { return tensor_sz(p->sz) * tensor_sz(p->vecsz); } static int prob_size_cmp(const void *p1_, const void *p2_) { const bench_problem * const *p1 = (const bench_problem * const *) p1_; const bench_problem * const *p2 = (const bench_problem * const *) p2_; return (sz(*p1) - sz(*p2)); } static struct my_option options[] = { {"help", NOARG, 'h'}, {"version", NOARG, 'V'}, {"verbose", NOARG, 'v'}, {"canonical", NOARG, 'c'}, {"time-limit", REQARG, 't'}, {"output-file", REQARG, 'o'}, {"impatient", NOARG, 'i'}, {"measure", NOARG, 'm'}, {"estimate", NOARG, 'e'}, {"exhaustive", NOARG, 'x'}, {"no-system-wisdom", NOARG, 'n'}, {"wisdom-file", REQARG, 'w'}, #ifdef HAVE_SMP {"threads", REQARG, 'T'}, #endif /* options to restrict configuration to rdft-only, etcetera? */ {0, NOARG, 0} }; static void help(FILE *f, const char *program_name) { fprintf( f, "Usage: %s [options] [sizes]\n" " Create wisdom (pre-planned/optimized transforms) for specified sizes,\n" " writing wisdom to stdout (or to a file, using -o).\n" "\nOptions:\n" " -h, --help: print this help\n" " -V, --version: print version/copyright info\n" " -v, --verbose: verbose output\n" " -c, --canonical: plan/optimize canonical set of sizes\n" " -t , --time-limit=: time limit in hours (default: 0, no limit)\n" " -o FILE, --output-file=FILE: output to FILE instead of stdout\n" " -m, --measure: plan in MEASURE mode (PATIENT is default)\n" " -e, --estimate: plan in ESTIMATE mode (not recommended)\n" " -x, --exhaustive: plan in EXHAUSTIVE mode (may be slow)\n" " -n, --no-system-wisdom: don't read /etc/fftw/ system wisdom file\n" " -w FILE, --wisdom-file=FILE: read wisdom from FILE (stdin if -)\n" #ifdef HAVE_SMP " -T N, --threads=N: plan with N threads\n" #endif "\nSize syntax: \n" " = c/r/k for complex/real(r2c,c2r)/r2r\n" " = i/o for in/out-of place\n" " = f/b for forward/backward, omitted for k transforms\n" " = [x[x...]], e.g. 10x12x14\n" " -- for k transforms, after each dimension is a :\n" " = f/b/h/e00/e01/e10/e11/o00/o01/o10/o11\n" " for R2HC/HC2R/DHT/REDFT00/.../RODFT11\n" , program_name); } /* powers of two and ten up to 2^20, for now */ static char canonical_sizes[][32] = { "1", "2", "4", "8", "16", "32", "64", "128", "256", "512", "1024", "2048", "4096", "8192", "16384", "32768", "65536", "131072", "262144", "524288", "1048576", "10", "100", "1000", "10000", "100000", "1000000", "2x2", "4x4", "8x8", "10x10", "16x16", "32x32", "64x64", "100x100", "128x128", "256x256", "512x512", "1000x1000", "1024x1024", "2x2x2", "4x4x4", "8x8x8", "10x10x10", "16x16x16", "32x32x32", "64x64x64", "100x100x100" }; #define NELEM(array)(sizeof(array) / sizeof((array)[0])) int bench_main(int argc, char *argv[]) { int c; unsigned i; int impatient = 0; int system_wisdom = 1; int canonical = 0; double hours = 0; FILE *output_file; char *output_fname = 0; bench_problem **problems = 0; int nproblems = 0, iproblem = 0; time_t begin; verbose = 0; usewisdom = 0; bench_srand(1); #ifdef HAVE_SMP /* do not configure FFTW with threads, unless the user requests -T */ threads_ok = 0; #endif while ((c = my_getopt(argc, argv, options)) != -1) { switch (c) { case 'h': help(stdout, argv[0]); exit(EXIT_SUCCESS); break; case 'V': printf("fftw-wisdom tool for FFTW version " VERSION ".\n"); printf( "\n" "Copyright (c) 2003, 2007-11 Matteo Frigo\n" "Copyright (c) 2003, 2007-11 Massachusetts Institute of Technology\n" "\n" "This program is free software; you can redistribute it and/or modify\n" "it under the terms of the GNU General Public License as published by\n" "the Free Software Foundation; either version 2 of the License, or\n" "(at your option) any later version.\n" "\n" "This program is distributed in the hope that it will be useful,\n" "but WITHOUT ANY WARRANTY; without even the implied warranty of\n" "MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the\n" "GNU General Public License for more details.\n" "\n" "You should have received a copy of the GNU General Public License\n" "along with this program; if not, write to the Free Software\n" "Foundation, Inc., 51 Franklin Street, Fifth Floor, Boston, MA 02110-1301 USA\n" ); exit(EXIT_SUCCESS); break; case 'v': verbose = 1; break; case 'c': canonical = 1; break; case 't': hours = atof(my_optarg); break; case 'o': if (output_fname) bench_free(output_fname); if (!strcmp(my_optarg, "-")) output_fname = 0; else { output_fname = (char *) bench_malloc(sizeof(char) * (strlen(my_optarg) + 1)); strcpy(output_fname, my_optarg); } break; case 'm': case 'i': impatient = 1; break; case 'e': the_flags |= FFTW_ESTIMATE; break; case 'x': the_flags |= FFTW_EXHAUSTIVE; break; case 'n': system_wisdom = 0; break; case 'w': { FILE *w = stdin; if (strcmp(my_optarg, "-") && !(w = fopen(my_optarg, "r"))) { fprintf(stderr, "fftw-wisdom: error opening \"%s\": ", my_optarg); perror(""); exit(EXIT_FAILURE); } if (!FFTW(import_wisdom_from_file)(w)) { fprintf(stderr, "fftw_wisdom: error reading wisdom " "from \"%s\"\n", my_optarg); exit(EXIT_FAILURE); } if (w != stdin) fclose(w); break; } #ifdef HAVE_SMP case 'T': nthreads = atoi(my_optarg); if (nthreads < 1) nthreads = 1; threads_ok = 1; BENCH_ASSERT(FFTW(init_threads)()); break; #endif case '?': /* `my_getopt' already printed an error message. */ cleanup(); return EXIT_FAILURE; default: abort (); } } if (!impatient) the_flags |= FFTW_PATIENT; if (system_wisdom) if (!FFTW(import_system_wisdom)() && verbose) fprintf(stderr, "fftw-wisdom: system-wisdom import failed\n"); if (canonical) { for (i = 0; i < NELEM(canonical_sizes); ++i) { unsigned j; char types[][8] = { "cof", "cob", "cif", "cib", "rof", "rob", "rif", "rib" }; for (j = 0; j < NELEM(types); ++j) { char ps[64]; if (!strchr(canonical_sizes[i],'x') || !strchr(types[j],'o')) { #ifdef HAVE_SNPRINTF snprintf(ps, sizeof(ps), "%s%s", types[j], canonical_sizes[i]); #else sprintf(ps, "%s%s", types[j], canonical_sizes[i]); #endif add_problem(ps, &problems, &iproblem, &nproblems); } } } } while (my_optind < argc) { if (!strcmp(argv[my_optind], "-")) { char s[1025]; while (1 == fscanf(stdin, "%1024s", s)) add_problem(s, &problems, &iproblem, &nproblems); } else add_problem(argv[my_optind], &problems, &iproblem, &nproblems); ++my_optind; } nproblems = iproblem; qsort(problems, nproblems, sizeof(bench_problem *), prob_size_cmp); if (!output_fname) output_file = stdout; else if (!(output_file = fopen(output_fname, "w"))) { fprintf(stderr, "fftw-wisdom: error creating \"%s\"", output_fname); perror(""); exit(EXIT_FAILURE); } begin = time((time_t*)0); for (iproblem = 0; iproblem < nproblems; ++iproblem) { if (hours <= 0 || hours > (time((time_t*)0) - begin) / 3600.0) do_problem(problems[iproblem]); problem_destroy(problems[iproblem]); } free(problems); if (verbose && hours > 0 && hours < (time((time_t*)0) - begin) / 3600.0) fprintf(stderr, "EXCEEDED TIME LIMIT OF %g HOURS.\n", hours); FFTW(export_wisdom_to_file)(output_file); if (output_file != stdout) fclose(output_file); if (output_fname) bench_free(output_fname); cleanup(); return EXIT_SUCCESS; }